Solar Light Harvesting with Nanocrystalline Semiconductors Lecture Notes in Chemistry

Lecture Notes in Chemistry 99

Oleksandr Stroyuk Solar Light Harvesting with Nanocrystalline Semiconductors Lecture Notes in Chemistry

Volume 99

Series editors Barry Carpenter, Cardiff, UK Paola Ceroni, Bologna, Italy Barbara Kirchner, Bonn, Germany Katharina Landfester, Mainz, Germany Jerzy Leszczynski, Jackson, USA Tien-Yau Luh, Taipei, Taiwan Eva Perlt, Bonn, Germany Nicolas C. Polfer, Gainesville, USA Reiner Salzer, Dresden, Germany The Lecture Notes in Chemistry

The series Lecture Notes in Chemistry (LNC) reports new developments in chemistry and molecular science-quickly and informally, but with a high quality and the explicit aim to summarize and communicate current knowledge for teaching and training purposes. Books published in this series are conceived as bridging material between advanced graduate textbooks and the forefront of research. They will serve the following purposes: • provide an accessible introduction to the ﬁeld to postgraduate students and nonspecialist researchers from related areas, • provide a source of advanced teaching material for specialized seminars, courses and schools, and • be readily accessible in print and online. The series covers all established ﬁelds of chemistry such as analytical chemistry, organic chemistry, inorganic chemistry, physical chemistry including electrochemistry, theoretical and computational chemistry, industrial chemistry, and catalysis. It is also a particularly suitable forum for volumes addressing the interfaces of chemistry with other disciplines, such as biology, medicine, physics, engineering, materials science including polymer and nanoscience, or earth and environmental science. Both authored and edited volumes will be considered for publication. Edited volumes should however consist of a very limited number of contributions only. Proceedings will not be considered for LNC. The year 2010 marks the relaunch of LNC.

More information about this series at http://www.springer.com/series/632 Oleksandr Stroyuk

Solar Light Harvesting with Nanocrystalline Semiconductors

123 Oleksandr Stroyuk Laboratory of Organic Photovoltaics and Electrochemistry L.V. Pysarzhevsky Institute of Physical Chemistry Kiev Ukraine

ISSN 0342-4901 ISSN 2192-6603 (electronic) Lecture Notes in Chemistry ISBN 978-3-319-68878-7 ISBN 978-3-319-68879-4 (eBook) https://doi.org/10.1007/978-3-319-68879-4

Library of Congress Control Number: 2017955261

© Springer International Publishing AG 2018 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Printed on acid-free paper

This Springer imprint is published by Springer Nature The registered company is Springer International Publishing AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland Preface

The solar light conversion and storage is currently one of the most blossoming interdisciplinary ﬁelds of science converging the physical chemistry, physics of solid state, optics, photochemistry, electrochemistry, catalysis, and many other research directions. The present textbook is intended to give a perspective on the current state of photochemical systems for the solar light harvesting based on nanocrystalline semiconductor materials and assemblies. The book chapters provide an account on various aspects of such systems, including the solar water splitting and evolution of molecular hydrogen, the photosynthetic processes of CO2 and N2 reduction, and the photoelectrochemical solar cells based on nanoparticulate semiconductor materials. A special focus is made on a “nano” aspect of semiconductor photocatalysis—the role of nanocrystals and size effects in the solar energy conversion, the design of semiconductor nanostructures with tailored photochemical properties, and the perspectives of nanophotocatalysis and photovoltaic systems based on semiconductor quantum dots. The introduction provides a brief account on various concepts of the solar light harvesting using the bulk and nanocrystalline semiconductors as well as a short historical account on the development of various photochemical and photovoltaic light conversion technologies. The ﬁrst chapter is an introduction to the photochemistry of semiconductor nanoparticles (NPs). It highlights basic principles of the selection of semiconductor materials for the applications in the solar light harvesting and requirements to the optical and electrophysical properties of photoactive semiconductor NPs. The chapter is focused on special features of the nanocrystalline semiconductors, in particular, on the quantum size effects and a unique capability of semiconductor NPs for the photoinduced charging. We discuss the most prominent size effects in the photochemistry of semiconductor NPs such as a dramatic enhancement of the photocatalytic/photoelectrochemical activity of nanocrystalline semiconductors as compared to their bulk counterparts, a crucial role of the surface charge traps in the photochemical processes, the effects of NP shape and porosity, the charging- induced changes in the NP photoreactivity, etc.

v vi Preface

The second and third chapters provide a review of the current state of the art in the semiconductor-based light-harvesting systems for the water splitting and the reduction of carbon dioxide and dinitrogen. The semiconductor-catalyzed photochemical water splitting for the hydrogen production as a green and sustainable fuel is discussed in detail. A review of the photocatalytic systems for the photosynthetic reduction of water, CO2, and N2 encompasses the systems based on the dye-sensitized oxide nanocrystalline semiconductors, binary semiconductor heterostructures, a survey of the visible-light-sensitive metal-chalcogenide nanophotocatalysts, and new and emerging nanostructured photocatalysts and cocatalysts of these photosynthetic processes. The fourth chapter introduces the reader to the semiconductor-based photoelectrochemical solar cells designed for the conversion of solar light into electric current. As the topic of dye-sensitized liquid-junction semiconductor solar cells has recently been broadly covered elsewhere, the discussion is limited mostly to the semiconductor nanoparticle-sensitized solar cells with liquid electrolytes, where the light conversion occurs as a result of a cyclic series of photochemical/photocatalytic processes and secondary “dark” redox reactions. The fifth chapter provides a concise account on typical synthetic approaches used for the preparation of various semiconductor nanomaterials—the colloidal NPs, nanocrystalline powders, thin films, binary and more complex nanoheterostructures, and nanocomposites of semiconductors with other functional components, such as metal NPs, carbonaceous compounds, etc. The final sixth chapter has a methodological character and acquaints the readers with the experimental methods using light as a probe of the structure, electrophysical, photophysical, and photochemical properties of nanocrystalline semiconductors and related heterostructures. The chapter discusses the methods of absorption and photoluminescence spectroscopy, flash photolysis, and other spectroscopic techniques that can be used to gain insights into the photochemical behavior of semiconductor NPs. I would like to thank all persons who helped me in writing this book, particularly, to my wife, Dr. Alexandra Raevskaya, a trusted friend and colleague, for her steady support and discussions on the book subject. Also, I appreciate deeply the experience and skills acquired by my coauthoring with senior peers from L.V. Pysarzhevsky Institute of Physical Chemistry, National Academy of Sciences of Ukraine—Prof. Anatoliy Kryukov, Prof. Stepan Kuchmiy, and Prof. Vitaliy Pokhodenko. Recently, we have published a comprehensive book on semiconductor nanophotocatalysis (Publishing House “Akademperiodika”, Kyiv, Ukraine, 2014) and the coauthoring of this book has been a source of invaluable experience and constant inspiration for me. As this book was written during my stay in Technical University of Dresden as a Marie-Skłodowska Curie Fellow, the support of European Union’s Horizon 2020 Research and Innovation Program (Grant No. 701254) is deeply appreciated. Preface vii

I hope that the present book will be useful both for a novice reader who starts a journey into the exciting world of the solar light-harvesting science and for an advanced reader who is already familiar with the ﬁeld and seeks an informative review on principal topics of the solar light conversion, such as the solar cells and semiconductor-based artiﬁcial photosynthesis.

Kiev, Dresden Dr. Oleksandr Stroyuk 2016–2017 Contents

1 Basic Concepts of the Photochemistry of Semiconductor Nanoparticles ...... 1 1.1 Light Absorption by Bulk and Nanocrystalline Semiconductors ...... 6 1.2 Influence of Surface States on the Photochemical Properties of Semiconductor NPs ...... 14 1.3 Influence of Size Dependences of CB and VB Levels ...... 19 1.4 Photoinduced Charging of Semiconductor NPs ...... 23 References ...... 29 2 Semiconductor-Based Photocatalytic Systems for the Solar-Light-Driven Water Splitting and Hydrogen Evolution ...... 39 2.1 Photocatalytic Systems Based on the Wide-Band-Gap Semiconductors and Sensitizers ...... 41 2.2 Photocatalytic Systems Based on the Binary and More Complex Semiconductor Heterostructures ...... 48 2.3 Photocatalytic Systems Based on the Metal-Doped Wide-Band-Gap Semiconductors ...... 55 2.4 Photocatalytic Systems Based on the Nonmetal-Doped Wide-Band-Gap Semiconductors ...... 58 2.5 Photocatalytic Systems Based on the Metal-Sulfide Semiconductors ...... 61 2.6 Emerging Semiconductor Photocatalysts for the Solar Hydrogen Production ...... 71 2.7 New-Generation Co-Catalysts for the Photocatalytic Hydrogen Production ...... 85 2.8 Stoichiometric Water Splitting Under the Illumination with the Visible Light ...... 88 References ...... 93

ix x Contents

3 Semiconductor-Based Photocatalytic Systems for the Reductive Conversion of CO2 and N2 ...... 127 3.1 Photocatalytic Reduction of Carbon Dioxide ...... 128 3.2 Photocatalytic Fixation of Dinitrogen ...... 146 References ...... 152 4 Semiconductor-Based Liquid-Junction Photoelectrochemical Solar Cells ...... 161 4.1 Principles and Designs of Semiconductor NP-Sensitized Solar Cells ...... 161 4.2 Basic Photoelectrochemical Characteristics of SSSCs ...... 166 4.3 Nanocrystalline Photoanodes Produced by the Ex Situ Deposition of Sensitizer NPs ...... 170 4.4 Nanocrystalline Photoanodes Produced by the In Situ Deposition of Sensitizer NPs ...... 182 4.5 Making Progress in SSSCs—Toward More Efﬁcient and Less Toxic Photoelectrodes ...... 203 4.6 Nanocrystalline Semiconductor Counter-Electrodes for SSSCs ...... 217 References ...... 227 5 Synthesis of Nanocrystalline Photo-Active Semiconductors ...... 241 5.1 Colloidal Semiconductors ...... 242 5.2 Nanocrystalline Powdered Semiconductors ...... 251 5.3 Nanocrystalline Films of Photo-Active Semiconductors ...... 260 5.4 Mesoporous Photo-Active Semiconductor Nanomaterials ...... 264 5.5 Spatially Organized Nanocrystalline Photo-Active Semiconductors ...... 268 5.6 Nanocrystalline Photo-Active Semiconductors on Carriers ...... 271 5.7 Doped Semiconductor Nano-Photocatalysts ...... 273 5.8 Bi- (Multi-) Component Photo-Active Semiconductor Nanostructures ...... 275 5.9 Photo-Active Semiconductor/Metal Nanostructures ...... 281 References ...... 284 6 Probing with Light—Optical Methods in Studies of Nanocrystalline Semiconductors ...... 319 6.1 A Brief Characterization of the Spectral Studies of Nano-Semiconductors ...... 319 6.2 Studies of Nano-Photocatalysts by the Electron Absorption Spectroscopy ...... 321 6.3 Luminescence Spectroscopy as a Tool for the Studies of Nanocrystalline Semiconductors ...... 329 6.4 Studies of Nanocrystalline Semiconductors by the Time-Resolved Photolysis Techniques ...... 339 Contents xi

6.5 Studies of Nanocrystalline Semiconductors Using Raman Scattering Spectroscopy ...... 350 6.6 Studies of Colloidal Semiconductor-Based Systems Using Dynamic Light Scattering ...... 356 References ...... 364 Index ...... 373 Abbreviations and Symbols

AIS Silver indium sulfide, Ag-In-S CB Conduction band CBD Chemical bath deposition CE Counter electrode CIS Copper indium sulfide, Cu-In-S CNP Carbon nanoparticle CVD Chemical vapor deposition Cys Cystein CZTS Copper zinc tin sulfide, Cu2ZnSnS4 DDT Dodecanethiol DEL Double electric layer DLS Dynamic light scattering DMSO Dimethyl sulfoxide DSSC Dye-sensitized solar cell EDTA Ethylenediaminetetraacetic acid EDX Energy-dispersive X-ray spectroscopy EIS Electrochemical impedance spectroscopy EMA Effective mass approximation EPR Electron paramagnetic resonance FTO Fluorine-doped tin oxide (transparent conductive glass) FWHM Full width on half-maximum GCN Graphitic carbon nitride (g-C3N4) GO Graphene oxide GSH Glutathione HDA Hexadecyl amine HOMO Highest occupied molecular orbital HS Hollow sphere HTT Hydrothermal treatment ICL Inorganic complex ligands IE Ion exchange

xiii xiv Abbreviations and Symbols

IPCE Incident photon to current efﬁciency IR Infrared ITO Indium tin oxide (transparent conductive glass) LO Longitudinal optical phonon LUMO Lowest unoccupied molecular orbital MAA Mercaptoacetic acid MOF Metal–organic framework MPA Mercaptopropionic acid NB Non-stationary bleaching NHE Normal hydrogen electrode NP Nanoparticle NR Nanorod NT Nanotube NS Nanosheet NW Nanowire OA Oleic acid ODE Octadecene OLA Oleylamine OTE Optically transparent electrode PCE Power conversion efﬁciency PEC Photoelectrochemical PEG Polyethylene glycol PEI Polyethyleneimine PL Photoluminescence PVA Polyvinyl alcohol PVP Polyvinylpyrrolidone QSE Quantum size effect QY Quantum yield RGO Reduced graphene oxide SEM Scanning electron microscopy SILAR Successive ionic layer adsorption and reaction SPP Sodium polyphosphate SPR Surface plasmon resonance SO Surface optical phonon SSSC Semiconductor-sensitized solar cell STEM Scanning transmission electron microscopy TBT Titanium tetrabutoxide TEA Triethanolamine TEM Transmission electron microscopy TGA Thioglycolic acid TMD Transition metal dichalcogenide TOP Trioctylphosphine TOPO Trioctylphosphine oxide TTIP Titanium tetraisopropoxide UV Ultraviolet Abbreviations and Symbols xv

VB Valence band Vis Visible light WZ Wurtzite XRD X-ray diffraction ZAIS Zinc-doped silver indium sulﬁde, Zn-Ag-In-S ZB Zinc blende aB Bohr radius of exciton Eg Bandgap EF Fermi energy ECB Conduction band potential EVB Valence band potential e-CB Conduction band electron + h VB Valence band hole – + e tr/h tr Trapped electron/trapped hole m*e Effective mass of CB electron m*h Effective mass of VB hole hv Light quantum energy kbe Wavelength of fundamental absorption band edge of a semiconductor Jsc Short-circuit current density Voc Open-circuit voltage RCT Charge transfer resistance η Total light conversion efﬁciency D Donor of electron A Acceptor of electron FF Fill factor of a voltage–current characteristic Introduction

…nature is not in a hurry and mankind is.

Giacomo Ciamician 1912. A solar light-harvesting system can be defined in the broadest terms possible as a combination of light-sensitive moieties (molecules, metal complexes, supramolecular complexes, nanodimensional inorganic or organic particles, biomolecules, and their assemblies) with various electron mediators, cocatalysis, etc. that serves to absorb the incoming solar light and converts the luminous energy into the form of electrical or chemical energy available for the future utilization or for the immediate chemical/electrochemical reactions. According to this definition, the artificial light-harvesting systems (that is, the systems devised by humans inspired by the natural light-harvesting systems) can be tentatively divided into three major classes: (i) the systems for the light-to-current conversion or the so-called solar cells, (ii) the systems for the artificial photosynthesis and accumulation/storage of the luminous energy in the form of stable chemical products, and (iii) the systems for immediate utilization of the solar energy as a driving force for various, mostly destructive, photochemical/photocatalytic tranformations of chemical species. The latter systems are broadly used in the environmental photocatalysis, where the solar light energy is applied to decompose various persistent organic and inorganic contaminants and the harmful biota both in the gas phase and waters. In the present book, we will focus on the former two types of systems that produce photocurrent or/and stable energy-rich products and, therefore, can be used for the conversion of the light energy in the form suitable for a long-term storage and a broad distribution to potential consumers. The efficient harvesting of solar light was an ever-inspiring dream of many great scientists, from philosophers to middle-age alchemists to academic chemists and physicist. The idea of mimicking the natural photosynthesis to accumulate the enormous energy flux supplied annually by Sun was discussed already by the ancient philosophers, but, probably, the first to put it into correct words and to

xvii xviii Introduction present it to a broad audience was Italian chemist Giacomo Ciamician, often honored as the Father of Photochemistry [1−4]. In his famous speech published in Science (1912) [5], he formulated the idea of photocatalysis to be applied for the solar light harvesting: “… the solar energy that reaches a small tropical country … is equal annualy to the energy produced by the entire amount of coal mined in the world!”, “… By using suitable catalyzers, it should be possible to transform the mixture of water and carbon dioxide into oxygen and methane, or to cause other endo-energetic processes” [1, 2]. Ciamician recognized the main problem of the solar light harvesting that still restricts a broad implementation of the photocatalytic/photoelectrochemical technologies, that is, a relatively low intensity of the solar light flux reaching the Earth surface as compared to the “concentrated” energy that can be produced from traditional fossil fuels. The Ciamician’s speech on the future of solar light conversion technologies is so vivid and precise that a large portion of its deserved to be cited unchanged and uninterrupted: “Where vegetation is rich, photochemistry may be left to the plants and by rational cultivation, as I have already explained, solar radiation may be used for industrial purposes. In the desert regions, unadapted to any kind of cultivation, photochemistry will artificially put their solar energy to practical uses. On the arid lands there will spring up industrial colonies without smoke and without smokestacks; forests of glass tubes will extend over the plants and glass buildings will rise everywhere; inside of these will take place the photochemical processes that hitherto have been the guarded secret of the plants, but that will have been mastered by human industry which will know how to make them bear even more abundant fruit than nature, for nature is not in a hurry and mankind is. And if in a distant future the supply of coal becomes completely exhausted, civilization will not be checked by that, for life and civilization will continue as long as the sun shines! If our black and nervous civilization, based on coal, shall be followed by a quieter civilization based on the utilization of solar energy, that will not be harmful to progress and to human happiness.” The progress of the mankind confirmed decisively the correctness of Ciamician’s prophesies. Indeed, the steady development by using sustainable raw sources and regenerative energy sources, like the solar energy, wind energy, geothermal energy, etc. nowadays not only determines the competitiveness of a country’s economy but can be the sole way to deal with a grave challenge of the global climatic changes caused by CO2 and heat emissions. It is expected that the energy demands of the humankind will increase to 50 TW in 2050 requiring the alternative non-carbon energy sources because the energy production in such a scale will invariably cause dramatic disruptions in the global climate balance [6]. At the same time, more energy is sent by the Sun to the Earth’s surface in one hour that the humankind consumes in a year [3] and this energy is potentially available to every country and every person and waits to be harvested and used. The solar light harvesting now comes by two parallel and sometimes intertwined ways of progress (Fig. 1). One is the realization of the Ciamician’s dream of artificial photosynthesis, that is, the endothermic conversion of extensively Introduction xix

Fig. 1 A scheme illustrating the most important semiconductor-based systems for the solar light harvesting

abundant compounds—water, N2, and CO2 into the chemical products that can be processed later releasing the accumulated energy or, alternatively, used as valuable chemical raw materials for more complex syntheses. Such photosynthetic processes require a series of concerted reactions involving many electrons, protons, and other components and, therefore, they can be realized with acceptable efficiency only in the presence of catalysts and photocatalysts. The second way is the direct solar light energy conversion into the electric power that occurs in the so-called solar cells. The main component of the solar cell is a light absorber which is light-excited and supplies nonequilibrium electrons and electron vacancies—holes into an electric circuit, thus resulting in the photocurrent generation. Numerous studies carried out around the globe in the last three–four decades showed that both types of the light-harvesting systems can be successfully realized by using photosensitive semiconductor materials. Indeed, some semiconductors show characteristics ideal for the solar light harvesting. The semiconductors have intense and continuous absorption bands that cover entire UV, visible, and, sometimes, near IR (NIR) ranges of the solar spectrum. This feature arises from the possibility of the electron transition from a continuous filled valence band (VB) into a continuous vacant conduction band (CB) under the excitation with the light of any energy higher than the distance between CB and VB, that is, the bandgap Eg. Typically, the semiconductors are crystalline and robust and reveal uncomparably higher photochemical stability than the light-harvesting molecular species and xx Introduction metal coordination compounds. Finally, some of the semiconductor materials used to convert/store the solar light are extremely abundant in the nature, such as silicon or iron oxide. In comparison to the semiconductors, the molecular and metal-complex light harvesters typically have much narrower absorption bands and are prone to various photochemical transformations resulting in the deterioration of their light-absorption capability. Also, typically quite a complex chemistry is needed for the preparation of efficient light-harvesting molecules, complexes, and assemblies that can compete with the natural photosynthesizing species and semiconductor materials. From the historical perspective, the semiconductor-based photocatalytic, photovoltaic, and photoelectrochemical (PEC) systems stem from the same roots. For example, the photocatalytic water splitting can be realized in a combined photochemical/photoelectrochemical regime in a water-splitting solar cell [7]. Today, probably nobody knows for sure the exact priority of the first photocatalytic experiments using semiconductors, but it is generally agreed upon that a real breakthrough in the area was inspired by the works of A. Fujishima and K. Honda on the PEC splitting of water on a biased titanium dioxide electrode. Indeed, TiO2 was an ideal material for the semiconductor-based light-harvesting systems due to availability, low cost, stability, and the millennia-long story of utilization as a pigment [8]. However, titania is only sensitive to the UV and a tiny (around 5%) portion of the visible light. This utter limitation on the background of so many unique positive properties inspired and continues to inspire numerous concepts and methods for the sensitization of TiO2 to the visible light by coupling it with the highly absorbing species or by altering its band structure imparting the TiO2 crystals with the capability to absorb the visible light. In recent two–three decades, the photochemistry of semiconductor materials experiences a real explosive growth associated with the miniatuarization of the semiconductor crystals to the nanometer dimensions. A transition from the micro- to nanoscale opens huge perspectives and potential of tailoring/designing the properties of semiconductor light harvesters via variations in the crystal size, shape, nanoparticle (NP) association mode, NP surface chemistry, doping, etc. Evolution of photocatalytic synthetic processes with the participation of semiconductors. The present book focuses on the photocatalytic processes resulting in the accumulation of the light energy in the form of endothermic chemical substances such as the molecular hydrogen as a main product of the water reduction and various products of multi-electron reduction of atmosphere-abundant CO2 and N2. The details of working principle and examples of corresponding photocatalytic systems are discussed in Chap. 2 (the water reduction) and Chap. 3 (the reduction of CO2 and N2). Here, we provide only a very general description of basic principles of the semiconductor-based photocatalytic systems introducing the reader into this field of photochemistry. The photocatalytic process starts when a semiconductor crystal absorbs a light quantum with the energy equal to or (typically) higher than the width of the forbidden band (or the bandgap Eg, process 1 in Fig. 2). The photogenerated charge Introduction xxi

Fig. 2 A general layout of a semiconductor-based photocatalytic system carriers can then recombine either via a radiative pathway 2 emitting a photoluminescence quantum or via a non-radiative pathway 3 providing a vibrational energy to the crystal lattice and adsorbed species. Some portion of the charge carriers migrates in the crystal and can reach the surface (processes 4) where the carriers get “trapped” by various lattice defects (vacancies, adatoms, undercoordinated atoms, etc.) as well as by the adsorbed species. The photogenerated VB hole typically has a high oxidation potential and gets filled with an electron from various donor species present in the system, for example, from water molecules or OH– ions, the process resulting in the evolution of molecular oxygen as a final product. The photogenerated CB electrons “seek” for + accepting species present in the system, such as protons or H3O ions reducing them to the molecular hydrogen (or CO2– to CO, CH2O, formate, CH4, and other products). After the withdrawal of both CB electron and VB hole to the accepting and donating species, respectively, the semiconductor crystal regains its original state thus finishing the photocatalytic cycle. Provided that fresh donors and acceptors are constantly supplied to the surface of semiconductor crystal, it can function as a “pump” transferring electrons from the donors to the acceptors at the expense of the solar light energy. Pioneer reports on the semiconductor-mediated photocatalytic processes appeared as early as in 1920–1930s dealing mostly with the photobleaching of dyes in the presence of titania crystals [7−9]. In the late 1960s, the studies of the water photoelectrolysts on titania electrodes were started by Fujishima and Honda, who reported in 1969–1972 on the photoelectrochemical water splitting on a single-crystal rutile electrode illuminated with the UV light [8, 10, 11]. In the proposed “artificial photosynthesis” scheme, the processes of water reduction and oxidation were separated in space—the oxidation of H2OtoO2 took place on the illuminated and externally biased titania electrode with the participation of VB holes, while the reduction of H2OtoH2 occurred in a separate vessel on the platinum foil that accepted electrons from the titania CB (Fig. 3a, b). xxii Introduction

Fig. 3 A layout (a) and a schematic energy diagram (b) of the pioneer PEC cell for the water splitting proposed by Fujishima and Honda; (c) illustration of a photocatalytic water-splitting system based on a suspended semiconductor photocatalyst. Reprinted with permissions from [10]. Copyright (2012) American Chemical Society

The miniatuarization of semiconductor crystals to several microns and the direct deposition of Pt NPs on the titania surface allowed to compose the suspension-based photocatalytic systems where both water reduction and oxidation took place on the suspended microparticle aggregates. Such systems can function without an external bias as the photogenerated electrons can migrate through the net of contacting crystals, thus avoiding the electron–hole recombination, contrary to the single-crystal titania electrodes (Fig. 3c). The next step in the development of the water-splitting systems was the introduction of “fuels”—the so-called “sacrificial” electron donors that provided electrons for the water reduction, being oxidized to CO2 and other products. When the process is performed in the presence of air, a water-splitting system converts into a system for the oxidation of sacrificial donors, as the electrons are transferred to O2 and the net result of the photocatalytic process is the total oxidation of the introduced organic species to carbon dioxide and other inorganic compounds (nitrates, sulfates, phosphates, etc.)—the so-called photocatalytic “mineralization” of organic compounds. Such systems started to live an independent life and constituted a special direction of the semiconductor photocatalysis dealing with the photocatalytic destruction of harmful organic compounds, the water decontamination, the air purification, the mitigation of harmful microorganisms, etc. [4, 8−10]. The development of such environment-oriented photocatalytic systems took place in 1980–1990s simultaneously with the advancement in the semiconductor-based light-energy-accumulating systems. A future progress of the environmental semiconductor photocatalysis can be readily grasped by “reviewing” the current reviews on this subject [4, 8−10, 12, 13]. This progress encompasses the systems for drinking water and air purification, the development of self-cleaning and photoactive building materials, various anti-bacterial coatings, etc. The environmental photocatalysis blossomed in 1990– 2000s resulting in the first real commercial implementations of the photocatalytic technologies. As noted in [4], starting from 2000, a steady flux of more than 1300 Introduction xxiii international patents on the photocatalytic technologies was observed, contrary to a few tens per year before 1990. An ever higher attention is currently paid to the semiconductor-mediated photocatalytic “synthetic” reactions allowing to form C–C, C–N, C–S, and other bonds and to realize the photo-driven syntheses of valuable industrial chemical products [3, 4]. The implementation of semiconductor photocatalysis in the organic synthesis is typically impeded by two obstacles. The first one is a low selectivity of the photocatalytic reactions on the semiconductor materials. Indeed, the VB holes in the most extensively studied semiconductor photocatalyst—TiO2—exist in the form of highly reactive HO• radicals that can oxidize indiscriminately almost every organic compound introduced into the system. Recently, various approaches were developed to selective oxidative and reductive processes on the semiconductor photocatalysts, including the doping, surface state engineering, introduction of metal-complex cocatalysts and enzymes, etc. [4]. The second problem consists in a typically low light intensity of the conventional light sources. To enhance the photochemical transformations using such low-intensity light fluxes, a special attention is paid to the design of new reactor types including the fluidized bed reactors and the continuous-flow systems [4]. Simultaneously, the technologies of solar light concentration are under the development resulting in the first promising results [4]. The issues of the photochemical production of non-carbonaceous fuels and the utilization of renewable energy sources advanced greatly starting from 2000s, when the perils of the over-abundant CO2 emission as a result of enormous consumption of the fossil fuels, were truly realized. The realization and apprehension of clear evidence of the hazardous global climatic changes resulted in a shift of the energy policy of most developed economies and the eve of 2010s was marked by a strongly renewed interest in all the spectrum of light-harvesting technologies, including the semiconductor-based solar cells and the photocatalytic systems producing hydrogen from water or mimicking the natural photosynthesis by converting CO2 and N2 by using the solar light energy. The need in efficient photochemical light conversion systems was strongly supported by the development of chemistry and photochemistry of nanocrystalline semiconductors [11]. The studies of nanoparticulate semiconductors and various related composites showed an unprecedented and virtually unlimited variability of the properties and functional characteristics that can be achieved by varying the size, shape, and composition of semiconductor nanocrystals, as well as a way of the spatial organization of NP-based systems on the nano-, micro-, and macroscale levels [11]. The so-called “Holy Grail” of the photocatalytic technologies of water splitting is an apparent quantum efficiency of 30% at 600 nm [11]. Currently, the several-percent efficiency at wavelengths as long as 500 nm was achieved with the total light conversion efficiency below 0.1%. Therefore, new-generation semiconductor photocatalysts with a bandgap of around 2 eV and lower are needed to create photocatalytic water-splitting systems feasible for the practical implementation. xxiv Introduction

Evolution of semiconductor-based solar cells. The attempts of the light energy conversion into the electric power stem from the photoelectric effect discovered by Henri Bequerel in 1839 [14, 15]. It should be noted that Bequerel observed the illumination-induced electric current between two electrodes immersed in a liquid electrolyte. Later, the photoeffect was also observed for solid electrolytes and a focus of the solar cell studies was turned to solid photovoltaic materials for many decades to come, till the dye-sensitized liquid-junction solar cells appeared on the research scene and started their fast progress. The first to appear was the silicon-based solar cells. Silicon has a bandgap of 1.1 eV which is close to the optimal value of *1.3 eV (peak of the solar irradiation spectrum) necessary for the achievement of the highest photoconversion efficiency. It is a photochemically stable (at least in the crystalline state), low-toxic, and earth-abundant material [6, 16]. The first efficient solar cell on the crystalline silicon was produced by Bell laboratory in 1954 and showed an efficiency of 6% [14, 16]. Today, such cells dominate the solar light converter market along with the amorphous Si-based sells [14]. Together, such cells occupy around 80% of the global solar cell market [16, 17]. The Si-based solar cells, typically named the first-generation semiconductor solar cells (Fig. 4), though being unrivalled in terms of the ratio of conversion efficiency versus production cost, have a number of shortcomings. First, the silicon solar cell technology requires a huge amount of very pure silicon. Typically, the photovoltaic technologies used rejected materials from the semiconductor industry and the necessary amount of raw materials can be maintained only if both industries are developed with the same rate, which is doubtful in view of a recent drastic growth of interest to the solar energy harvesting. The solar cells based on amorphous silicon, which is much less expensive, emerged in 1960–1970s with a first commercial cell available in 1981 [14]. Another fundamental shortcoming of the silicon is the indirect character of the interband electron transitions (see Chap. 1) resulting in a comparatively low linear absorption coefficient. At least a 100-lm-thick silicon layer is required for the complete solar light absorption on the Earth surface, thus putting limitations on the minimal thickness (and, therefore, the cost) of the solar cells. The above shortcomings stimulated a search for new conceptions/materials, for example, the application of new crystalline forms of Si with a reduced thickness, like Si nanoribbons, amorphous silicon, cadmium chalcogenides, ternary copper– indium–chalcogenides, and more complex quaternary compositions [14]. It was found already in mid-1980s that some of the materials, in particular, Cu2S and

Fig. 4 Evolution of semiconductor-based solar cells Introduction xxv

CdTe (studied since 1960s), CuInSe2, CuGaSe2, CuInS2 and their alloys Cu(In, Ga) (S, Se)2 (CIGS/CIGSe, studied from 1070s), and kesterite Cu2ZnSnS4 (studied from 1990s), sputtered as thin films can be used as the solar cell light absorbers [14, 17]. Most of such compounds are the direct-bandgap semiconductors with high absorption coefficients and, therefore, much thinner (1–2 lm) absorber layers are required for the efficient solar light harvesting [6]. Additionally, the alloying of several components can be used for a precise variation of the absorber bandgap. The thin-film cells can be formed on a variety of substrates (flexible or rigid, conductive, or insulating) with a broad spectrum of techniques, including plasma vacuum deposition, chemical vapor deposition, electrodeposition, sputtering methods, etc. [18]. A typical thin-film cell is designed as a p–n junction and includes a 1-lm-thick sputtered molybdenum contact on soda-lime glass, 1–2-lm p-type-conducting metal-chalcogenide absorber layer and a thin layer (*50 nm) n-type-conducting semiconductor material (for example, CdS), a transparent layer of undoped ZnO that prevents the shunting, and, finally, a conductive layer of Al-doped zinc oxide with a thickness of around 120 nm (Fig. 5a). The thin-film cells were christened as the second-generation semiconductor cells (Fig. 4) and showed efficiencies of 20% and higher [19]. However, the cell production puts very rigorous requirements to the purity of sources used for the thin-film sputtering, and therefore, the second-generation cells are very expensive and found applications mostly in the aerospace industry [14, 18], where the cell cost is not so critical. Recently, the thin-film solar cell technology gained a new impetus by discov- ering the possibility of using NP or molecular precursor “inks” for the preparation of metal-chalcogenide thin-film absorbers. The ternary and quaternary compounds can be prepared by using the well-established methods of the colloidal chemistry and concentrated to the form of inks. The inks can be deposited on any desirable substrate by the conventional inkjet printing and annealed in a non-oxidative atmosphere resulting in the film solidification and formation of good-quality absorber layers [6]. In a similar way, other components of the solar cell (metal contact, n-type component, barrier layers, etc.) can be prepared as the NP inks so

Fig. 5 A typical layout of (a) thin-film CIGS-based solar cell; (b) organic/inorganic perovskite-based solar cell. (c) an energy level diagram for the CH3NH3PbI3-based cell presented in (b). Reprinted with permissions from [6] (a) and [20]. Copyright (2010, 2015) American Chemical Society xxvi Introduction that the entire solar cell can be produced by using the inkjet printing technologies making such solar cells competitive to the conventional silicon-based devices [6]. A breakthrough in the efficiency of thin-film bulk-heterojunction solar cells is expected by using special light harvesters—semiconductor NPs capable of the multi-exciton generation. When some of the narrow-bandgap semiconductor NPs, such as PbX (X = S, Se, Te), CdTe, InP, Si, etc. [21], are illuminated with the light energy exceeding considerably the NP bandgap, one absorbed light quantum can be accommodated by generating several electron–hole pairs. This effect is possible due to a special feature of semiconductor NPs—the so-called “phonon bottleneck”. The term is used to refer to the slow thermalization of the primary electron–hole pair favoring to the channelling of lattice energy to other routes, in particular, into the generation of additional electron–hole couples till the energy excess (hv − Eg)is completely accommodated in the excited NP [21]. The phenomenon of multi-phonon generation gives a hope to surpass the fundamental Shockley– Queisser limit of 31−33% light conversion efficiency achievable for a single-junction solar cell [21]. In recent years, the thin-film solar cells based on the organo-inorganic perovskites have emerged coming the way to around 20% efficiency in mere 5 years from the publication of the first reports [20, 22, 23]. Such perovskites combine high absorption coefficients, a variable bandgap that can be tuned across the entire visible and NIR ranges, an unprecedently large electron mobility and charge diffusion coefficient, the tolerance to point defects and grain boundaries, and relative simplicity of the cell formation [22]. The most popular materials are CH3NH3PbX3 (X = Cl, Br, I). Typically, a perovskite layer is sandwiched between an electron transport layer (an inorganic wide-bandgap semiconductor like TiO2, SnO2, and ZnO) and a transparent hole transporting layer, for example, a derivative of spir- obifluorene (spiro-OMeTAD, Fig. 5b, c) [20]. However, for a successful implementation of such solar cells, a number of quite critical issues should be addressed, including the efficient recycling of Pb-containing cells after their utilization, as well as a low chemical and photochemical stability of the lead-based organo-inorganic perovskites. The idea of devising solar cell with two electrodes—a light-sensitive photoanode/photocathode and a catalytically active counter electrode connected by a liquid electrolyte—stems directly from the Bequerels experiments on the photoelectric effect [15, 24]. As discussed above, this idea was realized by A. Fujishima and K. Honda in their PEC cell for the water splitting on a rutile single crystal. However, TiO2 can absorb only a fraction of the solar irradiation and, therefore, the spectral sensitivity range of TiO2 crystals should be extended to longer wavelengths either by modifying the band structure of the crystal or by introducing visible-light-absorbing species on the crystal surface. The idea of the sensitization of wide-bandgap semiconductors, like TiO2, with external molecular absorbers is also one of the oldest conceptions of the solid-state photochemistry, introduced by Vogel in 1883 for silver halide emulsions used in the photographic process [15, 24]. A combination of a liquid electrolyte (“liquid-junction”)-based PEC cell with the dye sensitization approach gave rise to the dye-sensitized solar Introduction xxvii cells (DSSCs) belonging to the third-generation solar cells (Fig. 4), according to the generally accepted classification. The sensitization effect was also observed long ago for the narrow-bandgap semiconductors deposited to/formed on the surface of the wide-bandgap semiconductors, for example, for Ag2S produced by an ion exchange on the surface of AgBr (AgI) [25]. It is, therefore, logically inevitable that the conception of DSSCs was later extended to the semiconductor-sensitized solar cells (SSSCs) with the same liquid-junction architecture as in the DSSCs [26−29]. The DSSC research was strongly stimulated by the fuel crisis of 1973 and over a thousand papers on DSSCs emerged in a few years from the start of the studies attesting to an explosive growth of the DSSC field [15, 24, 27, 30]. A regenerative DSSC is designed to return to its original state after a cycle of the photoinduced and secondary (“dark”) chemical reactions on the electrodes with a net result of generating the electric power from the light energy (Fig. 6). The solar light is absorbed in the DSSCs by a molecular sensitizer—a dye or a metal complex. Ref. 30 provides an extensive review of the basic classes of organic dyes and metal coordination compounds that were tested as sensitizers of the liquid-junction DSSCs. It was found that the highest efficiency is observed for the DSSCs with monolayer-adsorbed sensitizer molecules because competitive processes of the intermolecular interactions (charge transfer, formation of excimers, etc.) result in a loss of the light-harvesting efficiency at higher coverages [30]. To achieve a high-light absorption with a single dye molecule layer on the wide-bandgap semiconductor surface, it was suggested to use mesoporous electrodes with a high specific surface area. As a result, the typical DSSCs comprise a mesoporous dye-sensitized TiO2 (or ZnO or SnO2) photoanode that supplies the photogenerated electrons into the electric circuit and regenerates the original state via the oxidation of a redox shuttle in the electrolyte [15, 24, 27, 30]. Typically, the mesoporous layer has a thickness of *10 lm and consists of loosely aggregated 10−30-nm particles resulting in 50−60% porosity and complete permeability with the liquid electrolyte. The mesoporous layer is formed on the optically transparent electrodes (OTEs), such as indium tin oxide (ITO) or

Fig. 6 (a) Scheme of a dye-sensitized liquid-junction solar cell. (b) A 900-cm2 glass-based sandwich DSSC module. The device consists of six serial-connected so-called meander-type current-collecting parts; (c) A building-integrated DSSC demonstrator from Dyesol. Reprinted with permissions from [30]. Copyright (2010) American Chemical Society xxviii Introduction

fluorine-doped tin oxide (FTO). The enormous experimental work on the selection and modification of the sensitizers to achieve the highest light conversion efficiency pushed to the summit a class of Ru–bipyridyl complexes as the most efficient and stable light harvesters [15, 24, 27, 30]. Also, a large variety of counter electrodes, electrolyte compositions, and redox shuttles were probed, and the highest light conversion efficiencies were achieved for the Ru–bpy complex sensitizers, Pt counter electrodes, and iodide/iodate redox couples in mixed water/polar organic solvents [30]. A dye absorbs the solar light and injects an electron into CB of the wide-bandgap support or, alternatively, directly into the subbandgap states of the metal oxide NP originating from the surface defects. The one-electron-oxidized dye regenerates its original state by accepting an electron from a reducing component of the redox shuttle, for example, iodide ions (Fig. 6a) [30]. The oxidized form of À * l redox shuttle, I3 ions, diffuses through the electrolyte layer ( 50 m) to the Pt counter electrode, where it is reduced to I− by the electrons coming from the photoanode through the electric circuit. If no appropriate redox couple is present in a DSSC, the photogenerated holes can oxidize water to O2, while the electrons transferred to the counter electrode can reduce water to hydrogen and the DSSCs perform as a photochemical solar cell for the water splitting as discussed above. The light conversion efficiency of such cells can be boosted by introducing a hole scavenger that acts as a consumable fuel enhancing the H2 generation and suppressing the oxidation of water [15, 24]. As discussed in Ref. [32], the DSSC concept is a good example of a system, where the performance of the overall device is better than that of the separate components. Indeed, the mesoporous titania cannot absorb efficiently the solar light and also does not conduct electric current. The conventional Ru–bpy sensitizers degrade very quickly when illuminated in solutions without any oxide support and redox shuttles. However, a combination of all the components into a united system results in a solar cell that can generate the electric current densities of up to 20 mA/cm2 and exhibits stable performance for more than 15 years in the outdoor solar illumination [30]. The DSSC technology is the one that has come the way from the early laboratory concepts to the small pilot cells and, finally, to the large-scale commercial realization (see some examples in Fig. 6b, c). The toughest challenge for the DSSCs still to be met is to surpass a threshold of 15% efficiency [30]. The thermodynamics of the DSSC design allows to achieve this value; however, quite spectacular efforts applied in the field of DSSC in two recent decades resulted in only *11% efficiency for the best-performing cells. A small ratio of applied efforts to the achieved efficiency increments observed in the recent years stimulated the studies of other liquid-junction cell designs, in particular, the above-mentioned SSSCs with the semiconductor sensitizer introduced in the form of NPs. The SSSCs with liquid electrolyte are the main subject of Chap. 4. The working principle of the SSSCs is essentially based on a combination of the light-driven photocatalytic processes on a photoanode (photocathode) and the electrocatalytic processes on a counter electrode (Fig. 7), very similar to the Introduction xxix above-discussed DSSCs. Here also, the light is absorbed by the NPs of a narrow-bandgap semiconductor, for example, CuInS2 (the photosensitizer NPs), resulting in the electron transfer to the wide-bandgap porous semiconductor metal oxide layer (TiO2 or ZnO). The VB hole of the sensitizer NPs is then filled at the expense of the oxidation of sulfide ions—one of the components of the redox couple present in the liquid electrolyte and having a very high adsorption affinity to the surface of sensitizer NPs. The elemental sulfur produced as a result of the S2– photooxidation gets bound by the polysulfide species and diffuses to the CuxS counter electrode where it is reduced by the electrons arriving from the photoanode and with this, the PEC cycle is finished leaving the cell in exactly the same state as before the light adsorption. The SSSCs started with a modest few-percents efficiency of the light harvesting but showed an accelerated growth and achieved in 2015−2016, a promising efficiency higher than 11%. The potential of such cells is still to be realized to a full extent. There exists a general optimism toward such semiconductor NP-sensitized liquid-junction solar cells in the research community, which was vividly expressed by P. Kamat in his essay “Quantum Dot Solar Cells. The Next Big Thing in Photovoltaics” [31]. The NP-sensitized SSSC field is one of the principal focuses of the present book and comes under a detailed discussion in Chap. 4. It is instructive to conclude the introduction to the semiconductor-based solar cells with some numerical data on the current top efficiencies. As reported in 2015 in a regularly updated solar cell efficiency table [19], the highest light conversions achieved are 25.6% and 11.4% for polycrystalline and amorphous silicon, respectively, 28.8%—for the gallium arsenide thin-film cells, 21%—for the thin-film CdTe-based cells, and 11.9%—for the DSCCs [19]. The highest reported efficiency for the NP-sensitized solar cells is currently around 12% [32]. For comparison, the top efficiency of a solar cell based exclusively on the organic semiconductors and the charge transport layers is around 11%, while the highest

Fig. 7 A layout of the photoelectrochemical cycle in the liquid-junction SSSC based on light-harvesting CuInS2 NPs and polysulfide redox couple xxx Introduction light conversion efficiency of 37.9% was reported for the multi-junction thin-film semiconductor cells [19].

References

1. Venturi M, Balzani V, Gandolfi MT (2005) Fuels from solar energy. A dream of Giacomo Ciamician, the father of photochemistry. Proceedings of the Solar World Congress, Orlando, Florida, 2005. 2. Albini A, Dichiarante V (2008) The ‘belle époque’ of photochemistry. Photochem Photobiol Sci 8:248–254. doi:10.1039/b806756b 3. Protti S, Fagnoni M (2009) The sunny side of chemistry: green synthesis by solar light. Photochem Photobiol Sci 8:1499–1516. doi:10.1039/b909128a 4. Ravelli D, Dondi D, Fagnonia M, Albini A (2008) Photocatalysis. A multi-faceted concept for green chemistry. Chem Soc Rev 38:1999–2011. doi:10.1039/b714786b 5. Ciamician, G., The Photochemistry of the Future, Science, 1912, 36, 385. 6. Habas SE, Platt HAS, van Hest MF, Ginley DS (2010) Low-Cost inorganic solar cells: From ink to printed device. Chem Rev 110:6571–6594. doi:10.1021/cr100191d 7. Walter MG, Warren EL, McKone JR, Boettcher SW, Mi Q, Santori EA, Lewis NS (2010) Solar water splitting cells. Chem Rev 110:6446–6473. doi:10.1021/cr1002326 8. Hashimoto K, Irie H, Fujishima A (2007) TiO2 Photocatalysis: A Historical Overview and Future Prospects. AAPPS Bulletin 17:12–28. 9. Serpone N, Emeline AV, Horikoshi S, Kuznetsov VN, Ryabchuk VK (2012) On the genesis of heterogeneous photocatalysis: a brief historical perspective in the period 1910 to the mid-1980s. Photochem Photobiol Sci 11:1121–1150. doi:10.1039/c2pp25026h 10. Teoh WY, Scott JA, Amal R (2012) Progress in heterogeneous photocatalysis: From classical radical chemistry to engineering nanomaterials and solar reactors. J Phys Chem Lett 3:629 −639. doi:10.1021/jz3000646 11. Maeda K (2011) Photocatalytic water splitting using semiconductor particles: History and recent developments. J Photochem Photobiol C 12:237–268. doi:10.1016/j.jphotochemrev. 2011.07.001 12. Chen H, Nanayakkara CE, Grassian VH (2012) Titanium dioxide photocatalysis in atmospheric chemistry. Chem Rev 112:5919−5948. doi:10.1021/cr3002092 13. Ljubas D (2005) Solar photocatalysis—a possible step in drinking water treatment. Energy 30:1699–1710. doi:10.1016/j.energy.2004.11.010 14. Goetzberger A, Luther J, Willeke G (2002) Solar cells: past, present, future. Sol Energy Mater Sol Cells 74:1–11. 15. M Grätzel (2001) Photoelectrochemical cells. Nature 414:338–344. doi:10.1038/35104607 16. Green MA (2009) The path to 25% silicon solar cell efficiency: History of silicon cell evolution. Prog Photovolt: Res Appl 17:183–189. doi:10.1002/pip.892 17. Unold T, Schock HW (2011) Nonconventional (Non-Silicon-Based) photovoltaic materials. Annu Rev Mater Res 41:297–321. doi:10.1146/annurev-matsci-062910-100437 18. Chopra KL, Paulson PD, Dutta V (2004) Thin-Film solar cells: An overview. Prog Photovolt: Res Appl 12:69–92. doi:10.1002/pip.541 19. Green MA, Emery K, Hishikawa Y, Warta W, Dunlop ED (2015) Solar cell efficiency tables (Version 45). Prog Photovolt: Res Appl 23:1–9. doi:10.1002/pip.2573 20. Ke W, Fang G, Liu Q, Xiong L, Qin P, Tao H, Wang J, Lei H, Li B, Wan J, Yang G, Yan Y (2015) Low-Temperature Solution-Processed Tin Oxide as an Alternative Electron Transporting Layer for Efficient Perovskite Solar Cells. J Am Chem Soc 137:6730−6733. doi:10.1021/jacs.5b01994 Introduction xxxi

21. Nozik AJ, Beard MC, Luther JM, Law M, Ellingson RJ, Johnson JC (2010) Semiconductor Quantum Dots and Quantum Dot Arrays and Applications of Multiple Exciton Generation to Third-Generation Photovoltaic Solar Cells. Chem Rev 110:6873–6890. doi:10.1021/cr900289f 22. Song TB, Chen Q, Zhou H, Jiang C, Wang HH, Yang Y, Liu Y, Youab J, Yang Y (2015) Perovskite solar cells: ﬁlm formation and properties. J Mater Chem A 3:9032–9050. doi:10. 1039/c4ta05246c 23 Grätzel M (2014) The light and shade of perovskite solar cells. Nature Mater 13:838–842. doi:10.1038/nmat4065 24. M Grätzel (2003) Dye-sensitized solar cells. J Photochem Photobiol C 4:145–153. doi:10. 1016/S1389-5567(03)00026-1 25. James TH (1977) The Theory of The Photographic Process. Macmillan, New York. 26. Rühle S, Shalom M, Zaban A (2010) Quantum dot sensitized solar cells. ChemPhysChem 11:2290–2304. doi:10.1002/cphc.201000069 27. Hodes G (2008) Comparison of dye- and semiconductor-sensitized porous nanocrystalline liquid junction solar cells. J Phys Chem C 112:17778–17787. doi:10.1021/jp803310s 28. Kamat PV (2008) Quantum dot solar cells. Semiconductor nanocrystals as light harvesters. J Phys Chem C 112:18737–18753. doi:10.1021/jp806791s 29. Kamat PV, Tvrdy K, Baker DR, Radich JG (2010) Beyond photovoltaics: semiconductor nanoarchitectures for liquid-junction solar cells. Chem Rev 110:6664–6688. doi:10.1021/ cr100243p 30. Hagfeldt A, Boschloo G, Sun L, Kloo L, Pettersson H (2010) Dye-Sensitized Solar Cells. Chem Rev 110:6595–6663. doi:10.1021/cr900356p 31. Kamat PV (2013) Quantum dot solar cells. The next big thing in photovoltaics. J Phys Chem Lett 4:908–918. doi:10.1021/jz400052e 32. Sharma D, Jha R, Kumar S (2016) Quantum dot sensitized solar cells: recent advances and future perspectives in photoanode. Sol Energy Mater Sol Cells 155:294–322. doi:10.1016/j. solmat.2016.05.062 Chapter 1 Basic Concepts of the Photochemistry of Semiconductor Nanoparticles

A strong interest to photochemical (photocatalytic, photoelectrochemical, and photoelectro-catalytic) processes with the participation of semiconductors was observed starting from 1970ths and resulted in the rise of semiconductor photochemistry as an independent discipline with exciting perspectives of applications in the chemical industrial synthesis, solar energy conversion, environmental protection, etc. [1–6]. The semiconductor photochemistry combined conceptions and knowledge of “classic” molecular photochemistry, catalysis, solid state physics, spectroscopy and other disciplines. In recent years, a renaissance of the semiconductor photochemistry is observed associated with successful developments in the physics and chemistry of nanocrystalline semiconductors, in particular the semiconductor nanoparticles (NPs) displaying size dependences of optical and electrophysical characteristics deemed before to be fundamental and invariable for the corresponding “bulk” semiconductor materials [7–12]. The so-called “quantum size effects” (QSEs) are of a special importance for the semiconductor photochemistry, the term QSEs referring to all possible size dependences of fundamental electro-physical properties of semiconductors with a crystal size smaller than a certain “critical” value. The QSEs originate from the spatial conﬁnement of the photogenerated electrons and holes (or a bound electron-hole pair—exciton) in the volume of NPs typically smaller than the exciton diameter (or doubled exciton Bohr radius aB) for a given semiconductor material [8, 10, 13–15]. Among the typical QSEs are size dependences of the bandgap Eg, spectral parameters and intensity of absorption and photoluminescence (PL) bands, oscillator strengths of optical excitonic transitions, exciton binding energy, as well as a gradual transformation of continuous energy bands [conduction band (CB) and valence band (VB)] into a spectrum of discrete electron states as the NP size is decreased. The latter effect makes possible another important size effect—the photoinduced charging of semiconductor NPs, that can affect strongly their photochemical behavior.

© Springer International Publishing AG 2018 1 O. Stroyuk, Solar Light Harvesting with Nanocrystalline Semiconductors, Lecture Notes in Chemistry 99, https://doi.org/10.1007/978-3-319-68879-4_1 2 1 Basic Concepts of the Photochemistry of Semiconductor …

The critical size for the QSEs to be observable can be associated with de Broglie wavelength of electron or exciton diameter 2aB [7, 8, 10, 16]. Two cases of spatial exciton confinement can be distinguished—a weak confinement, when the NP size d is close to 2aB, and a strong confinement at d <2aB. The aB depends on the chemical nature of a semiconductor and can vary in a relatively broad range (Table 1.1). As a result, the range of NP sizes, where QSEs can be observed, is broad. The dynamics of photochemical and photo-electrochemical (PEC) processes with the participation of semiconductors depends largely on their electro-physical parameters, in particular, the bandgap and energies (electrochemical potentials) of conduction and valence bands—ECB and EVB, respectively [1, 9]. A photochemical process starts after the absorption of a light quantum hv, which is possible under a condition hv Eg. The absorption results in an electron transition from VB to CB, leaving a hole in VB (Fig. 1.1). Typically, the electron-hole pair has an excess of vibrational energy (hv–Eg) which is accommodated by the crystal lattice till the electron and hole reach the CB bottom and VB top corresponding to the ECB and EVB levels (“thermalization” process).

Table 1.1 Bohr exciton Semiconductor aB,nm radius aB for some semiconductors [7, 10] CdS 2.4 ZnS 1.5 CdSe 3.9

TiO2 0.8–1.9 ZnO 4.8 CuCl 0.7

PbI2 1.9 PbS 18 PbSe 46

Fig. 1.1 Photoinduced charge transfers between a semiconductor, acceptor A, and donor D 1 Basic Concepts of the Photochemistry of Semiconductor … 3

Afterwards two key processes occur—the migration of charge carriers to the crystal surface and their interfacial transfer to other components of the photochemical system. The dynamics of interfacial charge transfer, in turn, depends on relative positions of the electron and hole potentials and the redox potentials of an electron acceptor A and an electron donor (a hole acceptor) D. The charge transfer is possible when ECB is more negative than the redox-potential of the acceptor 0 •− 0 •+ E A/A ) and EVB is more positive than the redox potential of the donor E (D/D ) [9] (Fig. 1.1), that is,

0 À ECB\E ðÞA=A ð1:1Þ

and

0 þ EVB [ E ðÞD=D ð1:2Þ

In “classic” photochemical systems based on bulk semiconductors, where ECB and EVB are the characteristic constants for a given semiconductor (when measured in the bulk but varying near the crystal surface as a result of double electric layer variations), only those acceptors and donors, which comply with the conditions (1.1) and (1.2), respectively, can participate in the photochemical transformations on the semiconductor crystal surface. A free energy of the interfacial electron transfer can be expressed as [14, 17]. ÂÃ 0 0 À DG ¼ eECBÀE ðÞA=A ð1:3Þ

A similar expression can be written also for the hole transfer. Thus, the design of photochemical systems based on bulk semiconductors with size-invariable Eg, ECB, and EVB is limited to the selection of appropriate electron acceptors and donors. Some influence can have also the adsorption of potential-determining ions. For example, the CB potential of metal oxide semiconductors, such as TiO2 and ZnO (pH 0) can be tuned by pH variations as ECB(pH) = ECB − 0.059pH. In the case when both conditions (1.1) and (1.2) are satisfied, a photoexcited semiconductor donates an electron to A and accepts an electron from D and thus regenerates its original (prior to the light absorption) state. Such events can, therefore, happen many times in a cyclic manner and the photochemical process occurs in a photocatalytic regime. In the case of neutral A and D species, the electron transfers generate an anion-radical A•− and a cation-radical D•+. If A and D are ionic species they decrease and increase the oxidation state, respectively. The electron transfers to A and from D should occur at a comparable rate, otherwise, the semiconductor typically undergoes reductive or oxidation photocorrosion, depending on the carrier type that gets accumulated in the crystal. The character of secondary (“dark”) processes depends on the nature and reactivity of the intermediary A•− and D•+ species generated in the primary charge transfer events. The dark stages include the formation of neutral free radicals or stable final products, their interaction and reactions with original A and D species, etc. 4 1 Basic Concepts of the Photochemistry of Semiconductor …

Apart from the selection of A and D species, the design of a photochemical system based on bulk semiconductors includes also the selection of a semiconductor that absorbs in an appropriate spectral “window”. For example, the solar light harvesting applications require the absorption spectrum of the semiconductor to be maximally overlapping with the solar irradiation spectrum near the Earth surface, that is, the materials with Eg range of 1.5–2.5 eV (see Table 1.2)[1]. In other applications, such as photopolymerization and photolithography, the light sensitivity range of semiconductors may be limited by the UV spectral domain (Eg =3–4 eV, Table 1.2). Another criterion for the selection of photo-active semiconductors for photocatalytic applications, is the chemical/photochemical stability of the semiconductor. For example, ZnO dissolves both in acidic and alkaline media and can corrode interacting with the products of photo-catalytic reactions. Also, it easily transforms into ZnS upon a contact with sulfide anions and thus additional precautions should be taken when using ZnO-based materials in liquid-junction solar cells with sulfide/polysulfide electrolytes. Cadmium sulfide is one of the “universal” photocatalysts and sensitizer materials for solar cells. However, it is prone to oxidative photocorrosion resulting in dissolution (CdS + 2O2 = CdSO4) when exposed to illumination in tha absence of strong electron donors, such as Na2SorNa2SO3. Finally, when large-scale applications are anticipated for the semiconductor-based photo-chemical systems, the factors of a low cost and availability as well as a low toxicity can be weighted and considered as well. The latter

Table 1.2 Band gap Eg and Semiconductor Eg, eV (approx. kbe, nm) the absorption band ed CdS (cubic) 2.4 (520) position (kbe) of some bulk semiconductors [7, 10, 26] ZnS (cubic) 3.6 (350) ZnS (hexagonal) 3.8 (330) PbS 0.41 (3030)

In2S3 2.0 (620)

Bi2S3 1.3 (1000)

MoS2 1.23 (1010) MnS 3.0 (415)

CuInS2 1.55 (800) CdSe 1.74 (715) AgCl 3.3–3.5 (355–380)

TiO2 (anatase) 3.2 (390)

TiO2 (rutile) 3.0 (415) ZnO 3.2 (390)

Fe2O3 2.0–2.2 (565–620)

SnO2 3.5 (355)

CeO2 3.4 (365)

BiVO4 2.4 (520)

Bi2MoO6 2.6 (480) 1 Basic Concepts of the Photochemistry of Semiconductor … 5 factor largely undermines possible applications of CdS- and PbS-based systems, as well as recently emerging Pb-based organo-inorganic perovskite materials for solar cells [18, 19]. The CuInS2-based materials have almost perfect bandgap for the solar light harvesting, but they contain relatively rare indium and therefore, alternatives are constantly probed using Earth crust abundant materials for solar cell absorbers, such as kesterite Cu2ZnSnS4 [20–22], that combines availability, low toxicity and a bandgap matching to the solar spectrum. As a result of the above-discussed criteria, that sometimes become contradictory, the selection of appropriate semiconductor photocatalysts or solar cell absorber materials is often a complex task that is still to be fulfilled in a satisfactory manner. A typical example is the photo-catalytic systems for the stoichiometric water splitting, where there is no “universal” semiconductor photocatalyst that combines simultaneously a high photoactivity, appropriate band gap and CB/VB positions and a high chemical stability (see a detailed discussion of such systems in Chap. 2). Going down to nanometer dimensions of semiconductor crystals opened new possibilities for the design of semiconductor-based light conversion systems associated both with the size/volume ratio effects and with the changes of fundamental photophysical/electrophysical properties of semiconductors. The transition from microcrystalline to nanocrystalline state of semiconductors results in a number of quantitative changes, including an increase in the total surface area and the surface-to-volume ratio and an increase in the population of various defects—dangling bonds, dislocations, undercoordinated atoms, vacancies, adventitious doping, etc. These changes, that can be designated as geometrical/ morphological ones, affect invariably the physical and chemical properties of semiconductor particles. In particular, an increase of surface atoms from around 1% for a 10-nm particle to *50% for 1–2-nm NP, can influence strongly the thermodynamic properties of the NP as a whole, such as melting and phase transition temperatures, heat capacitance, solubility, etc. Formation of various surface defects impacts the adsorption capacity and, therefore, the catalytic and photocatalytic processes with the participation of semiconductor NPs that obligatorily include intermediary steps of the reactant adsorption and product desorption. Besides, the NP surface defects can participate directly in the interfacial charge transfers and act as recombination sites thus influencing the photochemical and PL properties of semiconductor NPs. The nano-dispersed semiconductors can sometimes crystallize in the phases unstable for the bulk counterparts (phase size effect) and the relative stability of various lattice polymorphs and phase transition temperatures can differ drastically for bulk and nanocrystalline semiconductors. The size effects can also affect magnetic properties of semiconductor NPs, electric conductivity, diffraction of X-rays, Raman scattering and other properties. As the crystal size comes into the range of *10 nm and smaller, the above-mentioned morphological effects are joined by the quantum size effects—[7, 14, 23–25] altering the fundamental electron structure of the semiconductor. First, an expansion of the bandgap is observed, resulting in a corresponding increase of the photogenerated charge carrier potentials. As the NP size decreases, the 6 1 Basic Concepts of the Photochemistry of Semiconductor …

Fig. 1.2 Size-dependent variation of the electron excitation energy for CdS-based molecular and nanocrystalline species. The energies of electron transitions for molecular ions and clusters are provided in [156], the bandgap of nanocrystalline CdS taken from [64] continuous energy bands transform into sets of quantized discrete occupied and vacant levels. This phenomenon can be readily visualized by imagining the reverse process of formation of semiconductor NPs from molecular species, for which the electron transition energy should be much higher than in the bulk solids. Figure 1.2 illustrates schematically a change of the excitation energy for a typical semiconductor 2− photocatalyst—cadmium sulfide during the transition from molecular [Cd(SR)4] 4− ion (SR—aliphatic thiol) to larger polynuclear [Cd10S4(SR)16] and [Cd32S14(SR)36] clusters, then to *5-nm CdS NPs and, finally, to the bulk CdS. As the number of structural CdS units increases, the number of binding and non-binding orbitals increases as well. As a result, the CdS NPs with discrete (quantized) energy levels appear in an intermediary size range, then the distance between the quantized levels decreases gradually and, finally the adjacent levels melt into continuous energy bands, typical for larger CdS NPs and bulk cadmium sulfide crystals. The photochemical reactions with the participation of both bulk and nanocrystalline semiconductors start from the light absorption and formation of the elec- tronically excited state. The above discussion shows that the character of light absorption can be strongly affected by size effects. The following section introduces the reader to the basics of light absorption phenomena in semiconductors.

1.1 Light Absorption by Bulk and Nanocrystalline Semiconductors

The electromagnetic irradiation excites both atomic and electronic subsystems of a semiconductor crystal. The extinction of a light flux as a result of absorption in the −al semiconductor crystal can be described by the Lambert-Beer equation I = I0e , 1.1 Light Absorption by Bulk and Nanocrystalline Semiconductors 7 where I0 and I is an incident light intensity and the light intensity at a distance l from the crystal surface, respectively, a is a linear absorption coefficient showing the probability of light absorption per a length unit (cm−1). In the case of optically transparent colloidal semiconductor NPs the absorption coefficient can be expressed as a = ecl, where e is the molar absorption coefficient, c is the molar semiconductor concentration (in moles per L), l is the optical pass in the cuvette. Principal types of electron transitions in semiconductor crystals are summarized in Fig. 1.3a. Fundamental absorption. Transition type 1 corresponds to the fundamental absorption in the semiconductor crystal and results in the generation of free charge — À þ carriers a conduction band electron (eCB) and a valence band hole (hVB). This transition can occur if the light quantum energy is larger than or equal to the bandgap of the semiconductor, hv Eg. The fundamental absorption can originate from the electron transition of two types—direct transitions and indirect transitions. The absorption intensity is determined by the transition probability which can be assessed by the selection rules [7]. The direct (or vertical) interband transitions occur in the semiconductors having the lowest points of the potential curves E(k) on VB and CB, where E and k are the energy and the quasi-impetus of the electron, one above the other (Fig. 1.3b, transition 1). For the realization of the direct electron transitions, the light quantum energy should be equal to or higher than the direct bandgap. The indirect electron transitions can occur in semiconductors having displaced (in the k space) minimums of the potential curves of the ground and excited states. The electron comes from the VB maximum to the CB minimum (Fig. 1.3b, curve 1//). The indirect electron transitions require an additional energy supplied from the

Fig. 1.3 Photoinduced electron transitions in a semiconductor. In a 1—fundamental absorption; 2, 3—absorption on free charge carriers; 4–7—absorption on defects/impurities; 8—intra-bandgap absorption; 9—exciton absorption; 10—absorption as a result of exciton dissociation. In b direct (1) and indirect (1/,1//) electron transitions induced by the fundamental light absorption 8 1 Basic Concepts of the Photochemistry of Semiconductor … vibrational energy pool of the lattice, that is by the vibrational quanta of lattice— phonons with typical energies of a single phonon varying from 20 to 70 meV [7, 8, 26]. Thus, the direct electron transitions involve two particles (electron and hole) and at hv = Eg result only in the electron excitation of the semiconductor. The indirect transitions involve three particles (electron, hole, and phonon) and require both light and vibrational energy. As a process involving three particles is much less probable than a process with the participation of only two particles, the probability of indirect electron transitions and the intensity of corresponding absorption bands is much lower than the corresponding parameters of a direct transition. Indirect electron transitions can be realized also in the direct-bandgap semiconductors (Fig. 1.3b, transition 1/) as a result of absorption of light quanta with the energy much higher than Eg. Absorption on free charge carriers. The photoexcitation of electrons and holes delocalized in the corresponding energy bands results in the absorption of free carriers and transitions within the range of electron states available in the corresponding bands (Fig. 1.3a, transitions 2 and 3). The free-carrier absorption bands are continuous and reside in the IR range of the spectrum. Absorption on defects/impurities. The light absorption can occur as a result of localized electron transitions from defect- or impurity atom-related levels into the conduction band (Fig. 1.3a, transitions 4 and 5) or from the valence band—on the localized intra-bandgap levels (transitions 6 and 7). The defect/impurity absorption can be observed as a “tail” below the absorption band edge (at hv = Eg). As the density of impurity/defect-related states is much lower than the density of states in the allowed bands, the absorption coefficients of the impurity/defect-related bands are typically by several orders of magnitude lower than corresponding coefficients of the fundamental absorption. If the concentration of donor and acceptor defects in the semiconductor is relatively high, the so-called donor-acceptor couples can form that can absorb light resulting in the donor-acceptor electron transitions (Fig. 1.3a, transition 8). Exciton absorption. The photoexcited electron can come free into the CB, or, alternatively, remain bound with the hole by the Coulomb interactions forming a hydrogen-like e−…h+ quasi-particle or exciton (Fig. 1.3a, transition 9) that can migrate through the crystal. The exciton has an own discrete set of levels situated below the CB bottom. The exciton can dissociate either as a result of light absorption or under the influence of the thermal lattice energy. Upon the exciton dissociation the electron becomes free and delocalized in CB (Fig. 1.3a, transition 10). The exciton radius can be estimated using Eq. (1.4) which is similar to the Bohr equation for the hydrogen atom: 2e ¼ h 1 þ 1 ; ð : Þ aB 2 Ã Ã 1 4 e me mh 1.1 Light Absorption by Bulk and Nanocrystalline Semiconductors 9 where ћ is the reduced Planck constant (h/2p); e is the dielectric constant of the Ã Ã semiconductor; e is the electron charge; me , mh are the effective masses of CB electron and VB hole. The effective masses do not correspond to any measure of the inertia of the charge carriers and reflect the influence of the periodic potential of the semiconductor lattice on the movement of charge carriers. The effective masses of electron and hole depend on the semiconductor composition (Table 1.3) and typically are presented as a portion of the electron rest mass me. The doubled aB can be regarded as a “borderline” NP diameter for the QSEs to be observable. As mentioned earlier, two regimes of the spatial exciton confinement can be distinguished—weak confinement at 2aB < d k, where k is the average excitation wavelength) and strong confinement at d <2aB. In the former regime, the exciton experiences spatial confinement but the movement of charge carriers still results in a displacement of the mass center of the exciton (i.e. the NP size is larger than the exciton diameter). In the strong confinement regime, when the NP size becomes smaller than the exciton diameter, the charge carriers move at a steady exciton mass center resulting in a strong dependence of the electron properties of NPs on their size. The spatial exciton confinement in semiconductor NPs results in an increase of nano bulk the exciton energy and bandgap (Eg ) as compared to the bulk material (Eg ). nano bulk The size-dependent energy gain (DE = Eg − Eg ) can be estimated by using an effective mass approximation (EMA) based on the assumptions of parabolic band edges and size-independ effective masses [7, 8, 16, 25, 27, 28]: p2h2 1 1 1:786e2 DE ¼ þ À À 0:248RÃ ð1:5Þ 2 Ã Ã e y 2R me mh R

The ﬁrst term in Eq. (1.5) depends on R2 (R is the NP radius) and corresponds to the exciton energy increase due to the spatial conﬁnement in a “potential box”, that is in the NP volume. The second term describes the energy of Coulombic interactions between electrons and holes and increases in a reverse proportion to the NP Ã size. Ry is a Rydberg energy accounting to the correlation of the electron and hole

Table 1.3 Effective electron Ã Ã Ã Ã Semiconductor me mh (me /me) and hole (mh/me) masses for some CdS 0.2 0.8 semiconductors [7, 26] ZnS 0.27 0.58 PbS 0.1 0.1

Ag2S 4.55 7.8 CdSe 0.13 0.44 ZnSe 0.17 0.06 PbSe 0.05 0.05 CdTe 0.11–0.14 0.35–0.8

TiO2 *30 *3 ZnO 0.27 0.50 10 1 Basic Concepts of the Photochemistry of Semiconductor … movement. The two latter terms in Eq. (1.5) counter-weight the size-dependent Eg increment, however, in the case of medium and strong confinement the first term dominates and the latter two terms are typically neglected. The results of experimental and theoretical studies of size dependences of DE for various semiconductors, in particular, for CdS [16, 29–34], CdSe [33, 35–37], CdTe [33, 38–40], ZnO [41–44], PbS [45, 46], PbSe [46–48], etc. showed that EMA and Eq. (1.5) describe adequately only the case of weak exciton confinement in semiconductor NPs. The semiconductor NPs in the strong confinement regime experience a gradual transformation of bulk-like continuous band structure into sets of molecular-like discrete energy levels. Such NPs are similar to large molecular clusters and therefore are often referred to as semiconductor nanoclusters [13, 15]. For these NPs the basic EMA assumption of the parabolic bands is not valid anymore and EMA fails to predict adequately the size-dependence of Eg. For a correct description of the size dependences of electronic parameters of semiconductor nanoclusters other models are applied instead of the infinitely deep potential well model, in particular, the model of a finite-depth potential well or various semi-empirical quantum chemical calculations [32, 44, 48–53]. Therefore, the EMA model should be applied with a caution and when the discrepancy between predicted and experimental values becomes too high it is better to use empirical calibration curves plotted on the basis of numerous measurements by electron microscopy, X-ray diffraction, and other techniques. Examples of such calibration curves are discussed in Chap. 6. An increase of the bandgap due to the QSEs can be observed in the absorption spectra of semiconductor NPs as a “blue” (hypsochromic) shift of the absorption band edge (kbe), which is the larger the smaller NP size is. Figure 1.4 exemplifies

Fig. 1.4 Normalized absorption spectra of colloidal CdS NPs (a) and CdSe NPs (b). The NP size is a 8 nm (curve 1) and 1.8 nm (curve 2); b 2.5 nm (curve 1) and 1.8 nm (curve 2) 1.1 Light Absorption by Bulk and Nanocrystalline Semiconductors 11 this phenomenon for two photo-active semiconductors—CdS [54–56] and CdSe [57], showing strong size-dependent blue kbe shifts, especially for ultra-small NPs (d < 2 nm). The blue shift of the absorption band edge is observed for all semiconductor NPs experiencing the spatial exciton conﬁnement, its magnitude depending on the NP size d and the electron and hole effective masses. The highest bandgap increments D Ã Ã E are typical for semiconductor NPs with a small me (me 1) and therefore, – Ã relatively large aB (aB 2 5 nm). For example, in the case of PbSe (me = 0.05me [58], aB =4–6nm[59]) DE reaches a record value of 2.8 eV (relative to the bulk material) as the NP size in decreased to 2–3nm[58]. A considerable, up to 2 eV, increment of the bandgap can be achieved by reducing the size of PbS NPs from *20 to 2 nm [60]. A moderate Eg increment (DE = 0.1–0.2 eV) is typical for the semiconductors with small exciton radii (see Table 1.1)—CuCl (aB =0.7nm [7]), PbCl2 (aB = 1.9 nm [7]), TiO2 (aB = 0.8–1.9 nm [61]), ZnS (aB = 1.5 nm [62]), etc. The spatial conﬁnement of electron/hole movement in semiconductor NPs results also in an increase of the energy of Coulomb interaction between electron and hole in the exciton, that is, to an increase of the exciton binding energy Eex in inverse proportion to the NP radius R [7, 8, 63]:

2 À1 Eex ¼ e Â ðÞ3eR ð1:6Þ

The exciton binding energy in bulk semiconductors does not exceed several meV and the exciton can easily dissociate under the influence of lattice vibrations: ð À... þ Þ! À þ þ e h eCB hVB. By this reason, the absorption bands corresponding to the energy states of the exciton can be observed in bulk crystals only at very low temperatures. According to Eq. (1.6), a decrease of the NP size is accompanied by an increase of the exciton binding energy and Eex can become larger than the vibrational lattice energy (kT 25 meV at 300 К) already in the weak exciton confinement regime, the NPs revealing excitonic absorption peaks even at room temperature (Fig. 1.4). The intensity of light absorption by semiconductor NPs is also affected by the spatial exciton confinement. As Fig. 1.4 shows for CdS and CdSe NPs, a decrease of the NP size is accompanied by the “concentration” of absorbance within the excitonic peak which becomes more and more narrow as the NP size is reduced. This effect originates from increased overlapping of the wave functions of electron and hole and a corresponding increase in the exciton generation probability. In spectroscopic terms, this effect results in an increase of the oscillator strength of the first excitonic transition that can be calculated using Eq. (1.7)[7, 8]:

2mÃ f ¼ e DEjjl 2jjUðRÞ 2; ð1:7Þ h2 where DE is the transition energy, |l|2 is the dipole transition moment, |U(R)|2 is the 3 electron and hole wave function overlapping factor proportional to (aB/R) . 12 1 Basic Concepts of the Photochemistry of Semiconductor …

Equation (1.7) anticipates an increase of the oscillator strength of the excitonic transition proportionally to a decrease of the NP volume (R3). This model found experimental evidence, in particular, in the optical properties of size-selected CdS NPs [30, 31, 33, 36, 56, 64]. Figure 1.5 shows that a decrease of the CdS NP size from around 5 to 2 nm results in an increase of the oscillator strength of the excitonic transition by more than an order of magnitude. Similar results were reported for CdSe NPs [33, 63, 65–67] and CdTe NPs [33, 39]. As the bandgap of semiconductor NPs increases with a size decrease the positions of CB and VB levels change as well shifting to more negative and to more positive values, respectively. The shifts—DECB and DEVB can be estimated for a given NP diameter d by using Eqs. (1.8) and (1.9)[9, 16].

2 D ¼ h ð : Þ ECB Ã 2 1 8 2me d

2 D ¼ h ð : Þ EVB Ã 2 1 9 2mhd

It is obvious that for the quantum-sized semiconductor NPs the free energy of electron/hole transfer (Eq. 1.3) depends on the NP size. Therefore, the NP size becomes an additional “ﬁtting” parameter allowing for varying the energy level alignment in the photochemical system without actual changes in its chemical composition. Up to date, a considerable massive of experimental evidence was accumulated on the size dependences of the photochemical activity of many semiconductor NPs [9–12, 14]. The reported results showed that the QSEs can not only accelerate the photochemical/photocatalytic processes with the participation of semiconductor NPs but can even result in an expansion of a range of semiconductor materials that can be used as photocatalysts as well as of substrates that can be involved in photochemical transformations for a given semiconductor. Despite the fact that an increase in the photochemical activity is observed almost routinely when nanocrystalline semiconductors are used instead of bulk

Fig. 1.5 Relative oscillator strength of the first excitonic transition (f/fex) of CdS NPs on the NP radius R and R−3 (insert); fex is the oscillator strength of the first excitonic transition in bulk CdS crystals, fex = 0.0256 [157] 1.1 Light Absorption by Bulk and Nanocrystalline Semiconductors 13 counterparts, the interpretation of this phenomenon is often fragmentary or biased. In particular, the acceleration of photochemical reactions is typically associated only with the size-dependence of CB/VB energies. At that, other important factors that can affect both thermodynamics and kinetics of the interfacial charge carrier transfers are discarded or not taken into account. A large volume of data on the size dependences of photophysical and primary photochemical processes that is accumulated to date using spectral and kinetic methods, in particular, by the PL spectroscopy and the flash photolysis, is used quite rarely in the interpretations of the size dependences of photocatalytic and/or photo-electrochemical activity of nanocrystalline semiconductor materials. In this view, the present book is intentionally focused on the potential of spectroscopic methods for the studies of nanocrystalline semiconductors in an attempt to fill the gap between photophysics and photochemistry of semiconductor NPs (see Chap. 6 for details). The studies of semiconductor NPs by means of the stationary and time-resolved PL measurements showed that the photogenerated free charge carriers get “trapped” quite rapidly by various defects of the NP lattice. As the NPs are characterized by a high surface-to-volume ratio and largely disordered surface with a lot of undercoordinated surface atoms, the most part of the lattice defects belong to the NP surface. The defect-related electron states are located in the bandgap and have, therefore, a localized character. Depending on the distance (in terms of energy) between the defect-related states and CB/VB edges one distinguishes between “shallow” and “deep” charge traps. The trapping of charge carriers results in a decrease of their energy depending on the trap “depth” and the carriers do not move freely anymore and become localized. Further electron-hole recombination processes can occur either at the encounter between freely moving electron/hole and a trapped hole/electron or via the electron tunneling between the trapped electrons and holes. As the charge carriers get trapped their chemical potential decreases as compared to original ECB/EVB level and, therefore, the trapping can affect strongly further chemical reactions with the participation of the photogenerated electrons and holes. Typically, the trapping of carriers in semiconductor NPs occurs very fast, in a femtosecond/picosecond time range and, as a result, almost exclusively trapped carriers participate in the chemical reactions while having electrochemical potentials different from those that one can expect from Eqs. (1.8) and (1.9). However, despite the obvious influence of the charge trapping on the photochemical activity of semiconductor NPs only scattered reports on the analysis of possible consequences of this phenomenon on the photochemical properties of NPs can be found. The phenomenon of photoinduced charging of semiconductor NPs is another example of a QSE that draws undeservedly low attention with respect to the photochemical NP behavior, despite quite elaborate studies of this effect by spectral and kinetic methods. As the NP size decreases the continuous energy bands transform into sets of discrete levels. If the rate of interfacial transfer of electrons is much lower than the rate of hole transfer the discrete levels in CB get filled with excessive electrons and thus the photogeneration of new electron-hole pairs requires 14 1 Basic Concepts of the Photochemistry of Semiconductor … a higher energy, as compared to Eg. Simultaneously with the charging-induced increase in the optical bandgap, the surface traps become filled as well thus resulting in inhibition of the electron-hole recombination and in a considerable increase of the potential of the surface double charged layer. All these factors are favorable for the interfacial electron transfer and produce a unique photo-chemical behavior of charged NPs differing strongly from that of “regular” non-charged NPs. A vivid example of such behavior is the participation of the photocharged semiconductor NPs in the processes prohibited thermodynamically for regular NPs (discussed below).

1.2 Inﬂuence of Surface States on the Photochemical Properties of Semiconductor NPs

The surface states of semiconductor NPs inﬂuence both radiative electron-hole recombination processes and the interfacial electron transfers. Both types of processes are competing and, therefore, by observing the evolution of PL properties of semiconductor NPs we can make some conclusions on the photoinduced charge transfer dynamics and the photochemical reactions in general. The discussion of these interrelations between PL and charge transfer requires a concise characterization of the PL phenomena prior the examination of the role of surface states in the photochemistry of semiconductor NPs. The photoexcited semiconductor NPs relax to the ground state by several competing routes, in particular by non-radiative electron-hole recombination (1.10), charge carrier trapping by NP defects (1.11, 1.12), direct radiative recombination between free charge carriers (1.13), and interfacial transfer of free carriers to acceptor and donor substrates.

À þ þ ! ðÞðÞÀ ð: Þ eCB hVB nph ph phonon 1 10 ÀÁ À ! À ÀÀ ð : Þ eCB etr etr trapped electron 1 11 ÀÁ þ ! þ þ À ð : Þ hVB htr htr trapped hole 1 12 À þ þ ! ðÞðÀ : Þ eCB hVB hvlum hvlum PL quantum 1 13

The photoexcitation of semiconductor NPs as small as *2aB results in the generation of charge carriers both in the NP volume and on the NP surface [68, 69]. The charge carriers produced in the NP volume can then migrate to the defect-rich surface and get captured by the surface traps. The migration process is typically very fast, for example around 10 fs for 3–5-nm CdS NPs [68]. The charge carriers can be trapped (localized) by the lattice defects, undercoordinated surface atoms, admixtures and adsorbed substrates. The charge trapping results in the generation of active species, typically of a radical (ion-radical) nature that can participate in the 1.2 Influence of Surface States on the Photochemical … 15 following redox reactions on the NP surface. For example, the electron trapping on 3+ the TiO2 NP surface produces Ti ions that can be detected by the electron paramagnetic resonance (EPR) and optically, by a characteristic absorption band with a peak at 600–900 nm (see Chap. 6)[70–80]. In oxygen-containing systems, •− the electron can be trapped on the NP surface as a O2 anion-radical [72, 81] which is one of the main actors in numerous photocatalytic processes of the oxidation of organic compounds. The trapped hole can exist on the TiO2 NP surface in the form of a OH• radical [81]oraO•− radical [70–72] that can also be detected by EPR [70, 72, 75–77] and by characteristic absorption bands [71, 79, 80]. The hole trapping in metal-sulfide NPs produces S•− anion-radicals with characteristic absorption features [82, 83]. A special feature of semiconductor NPs differing them from bulk counterparts is a very short time of the charge carrier migration to the NP surface (sm), as compared to the characteristic electron-hole recombination time (sr). For example, sm can be as small as 10 ps for a 10-nm TiO2 NPs, while sr is by several orders of magnitude larger—around 100 ns [14, 84]. Owing to such drastic difference in the characteristic times, the primary separation and trapping of the photogenerated electrons and holes in NPs occur very efficiently. As the NP size is increased from *10 nm to *1 lm, the migration time (which depends on R2) grows to around 100 ns [14, 84] and the micro-crystal thus loses the favorable conditions for the charge separation existing in the case of a nanocrystal. By this reason, an external electric field should be applied to the micro-crystalline semiconductors for the charge migration to be competitive to the recombination. The semiconductor NPs also reveal a different structure of the semiconductor/electrolyte interface as compared to the corresponding bulk materials. A typical length of a depleted layer in the semiconductor microcrystals is by an order of magnitude larger than the linear size of quantum-sized semiconductor NPs [14] (Fig. 1.6). A distribution of the potential from the center of a semiconductor crystal to the distance r from the center, Du(r), can be described by Eq. (1.14)[14] kT r Àðr À WÞ 2 2ðr À WÞ DuðrÞ¼ 0 1 þ 0 ; ð1:14Þ 6e LD r where r0 is the crystal radius, W is a band bending area length (Fig. 1.6); k is Boltzmann constant, T is temperature, LD is a Debye length depending on the 2 1/2 charge density ND, LD =(e0ekT/2e ND) (e, e0 is the dielectric constant of semiconductor and electrolyte, respectively). 2 For low-doped semiconductors with e * 10, LD is on the order of 10 nm [85]. Therefore, for a 1–10-nm particle r0 LD and Eq. (1.14) can be simplified to Eq. (1.15), indicating that Du is the same both on the surface and in the volume of semiconductor NPs, and no appreciable band bending takes place, that can impede the interfacial charge transfer (Fig. 1.6). 16 1 Basic Concepts of the Photochemistry of Semiconductor …

Fig. 1.6 Potential variation on the semiconductor/electrolyte interface for semiconductor microcrystals (a) and nanocrystals (b)

kT r 2 Du ¼ 0 ð1:15Þ 6e LD

The lack of the band bending on the semiconductor/electrolyte interface results in a dramatic acceleration of the charge transfer from semiconductor NPs, as compared to the bulk crystals of the same composition. For example, hole migration − to the surface of colloidal TiO2 NPs and the interfacial transfer to SCN ions occur within 50 fs after the photoexcitation [86]. The photoinduced electron transfer from CdSe and CdS NPs to adsorbed methylviologen (4,4/-dimethylbipyridyl cation, MV2+) requires 70 fs and 200–300 fs, respectively [87, 88]. The charge transfer competes with the radiative electron-hole recombination that can occur via two different mechanisms [69, 89]. The ﬁrst route of PL generation is through direct interband recombination of the photogenerated carriers. As this recombination occurs very often between the exciton-bound charge carriers, this PL type is typically referred to as the excitonic PL [process (1.13)]. The excitonic PL band maximum position is close to the absorption band edge and approximately corresponds to Eg (Fig. 1.7a). For the direct-bandgap semiconductor NPs, the excitonic PL decays in a nanosecond time range and reveals a quantum yield of 10−3–100 depending on the NP composition and synthesis mode. The second-type PL originates from the recombination of free charge carriers with counterparts trapped by the NP lattice defects—either between a free hole and a trapped electron (1.16) or between a free electron and a trapped hole (1.17) [69, 90]. 1.2 Inﬂuence of Surface States on the Photochemical … 17

Fig. 1.7 a Normalized absorption (curve 1) and PL (curve 2) spectra of colloidal 2.5-nm CdSe NPs [158–160]. b Absorption (curve 1) and PL (curves 2–4) spectra of colloidal 1.8-nm CdS NPs [54]. The PL spectra were registered in 5 ns (2), 20 ns (3), and 50 ns (4) after the laser pulse

À þ þ ! = ð : Þ etr hVB hvlum 1 16

À þ þ ! == ð : Þ eCB htr hvlum 1 17

The energy states corresponding to the traps reside in the bandgap—the electron traps De are located lower than the CB bottom, while the hole traps Dh typically are above the VB top. As a result, the bands of defect-related (or donor-acceptor, DA) PL are typically shifted to lower wavelengths as compared to kbe (Fig. 1.7a). As the trap states can differ by “depth” and the local surrounding, the trap state energy spectrum is typically quite broad and mirrored by a large spectral width of the defect-related PL. The DA luminescence emission at room temperature is a phenomenon typical for semiconductor NPs with a high and disordered surface area, while for the bulk semiconductors the DA PL can typically be observed only in cryogenic conditions [7]. Mechanisms and dynamics of the defect-related PL depend strongly on the synthesis conditions, size and surface chemistry of semiconductor NPs. The recombination is supposed [90–93] to occur at an encounter of two opposite charge carriers, of which one is localized in a deep trap (De/h > kT), while other migrating in CB/VB or trapped by “shallow” traps (De/h kT) that can be ionized under the inﬂuence of lattice vibrations. A shift between the maximum of excitonic PL band (or bandgap) and defect-related PL band characterizes a depth of the traps relative to the corresponding band edges. If both charge carriers are trapped the PL originates, most probably, from the electron tunneling between the electron and hole trap states. Some examples of analytical extraction of trap energies from PL spectra of semiconductor NPs are discussed in Chap. 6. 18 1 Basic Concepts of the Photochemistry of Semiconductor …

In real semiconductor NPs, both electron and hole have quite a broad spectrum of traps differing both by the depth (energy) and the distance r between electron and hole traps. The pairs of opposite charges that are closer to each other experience a stronger Coulomb interaction than the pairs separated by a larger distance. The distance distribution affects the emitted PL spectrum according to Eq. 1.18 [8, 90]:

2 Elum ¼ EgÀðÞþDhÀDe e =er ð1:18Þ

The probability of the radiative recombination is inversely proportional to r and, therefore, the average distance between trapped electrons and holes gradually increases in the course of radiative recombination. As a result, the third member of Eq. (1.15) and the energy of emitted PL quanta gradually decrease. This effect can be observed as a red shift of the PL band maximum in the course of PL decay (Fig. 1.7b). The surface traps can sometimes affect the dynamics and even the mechanism of photochemical and photocatalytic reactions in a rather decisive manner. For example, CdSe NPs can act as a photocatalyst of the one-electron reduction of MV2+ to cation-radical as well as of the reduction of MV+• to a neutral form MV0 [94]. At the same time, CdxZn1−xS NPs with roughly the same CB potential (at x = 0.25) can photocatalyze only the first of these processes and reveal no activity in the photochemical generation of MV0. A detailed analysis of PL properties of both CdSe and CdxZn1−xS NP presented in Chap. 6 allowed to conclude that the difference in the photochemical behavior originates from a different depth of the electron traps in both semiconductors. Thus, the electron trapping in CdSe NPs decreases the electron energy only slightly and does not impede it from the inter- +• facial transfer to MV , while in the case of CdxZn1−xS NPs the electrons “fall” too deeply into the traps losing considerably in the chemical potential and the capability of MV+• reduction [94]. −• Colloidal ZnS NPs cannot reduce CO2 to CO2 despite the fact the CB potential is negative enough for this process to occur [95]. This fact is interpreted as a result of a deep trapping of the photogenerated electron resulting in a loss of energy of around 1 eV. As additional HS– ions are introduced into the system, they fill the surface vacancies and block the trapping, and the ZnS NPs gain the ability to reduce À fi CO2 at the expense of the oxidation of H2PO2 as a sacri cial donor [95]. This example demonstrates vividly the possibility of influencing the dynamics of photochemical/photocatalytic processes by a proper modification of the semiconductor NP surface. During the photocatalytic reduction of CdII and ZnII on the surface of CdS and ZnS NPs, respectively, the ions adsorb selectively on the NP surface and create additional electron traps capable of participation in the radiative recombination [96, 97]. As a result, an increase in the CdII/ZnII concentration results in the deterioration of photocatalytic activity of CdS (ZnS) NPs [96, 97]. Doping of CdS and CdxZn1 III II −xS NPs by small amounts of Bi or Cu introduces additional deep electron traps mediating and increasing the electron transfer from NPs to molecular oxygen [98]. 1.2 Influence of Surface States on the Photochemical … 19

These examples show that an intentional modiﬁcation of the surface of semiconductor NPs via adsorption or implantation of ionic species can be used to inﬂuence the dynamics of the photochemical processes on the NP surface.

1.3 Inﬂuence of Size Dependences of CB and VB Levels

A basic condition of photochemical/photocatalytic processes with the participation of semiconductor NPs is a correspondence between the CB and VB levels and the redox potentials of the electron acceptors and donors adsorbed on the NP surface. The VB potentials of typical photochemically active bulk semiconductors, such as CdS (1.6 V versus normal hydrogen electrode (NHE) [99]), ZnS (1.8 V vs. NHE [95]), TiO2, ZnO, SnO2,WO3 (EVB > 2.5 V vs. NHE [1, 14]), and Fe2O3 (1.6 V vs. NHE [100]) are relatively high. As a result, an increase of the EVB level induced by QSEs, as a rule, does not affect strongly the photochemical activity of these semiconductors in oxidative reactions with the participation of VB holes. At the same time, the CB potentials of many photoactive semiconductors are located only slightly above the NHE [1, 14] and even small variations in ECB can affect quite spectacularly their photochemical activity. Also, for most photo-active semicon- Ã Ã ductors me < mh, and, according to Eqs. (1.8) and (1.9), we can expect that a size-dependent variation of the CB level will be much larger than the corresponding change in the EVB potential. Two very important consequences of the QSEs can be envisaged for semiconductor NPs as a result of the size-dependence of ECB and EVB levels, in particular, (i) demonstration of photochemical properties by NPs of a semiconductor that is absolutely passive in the bulk form and (ii) enhancement of the photochemical/photocatalytic activity of semiconductors with a decrease of the crystal size. It should be noted, however, that the increase of CB and VB potential per se does not guarantee realization of these two effects, because the photochemical activity of semiconductor NPs depends not only on the charge carrier energy but also on the dynamics of primary photophysical/photochemical processes, recombination rate, and many other factors. Photochemical activity of nanocrystalline semiconductors passive in the form of bulk materials. The photochemical activity and photocatalytic properties are revealed by a comparatively narrow number of inorganic semiconductors [1] and most reported semiconductor photocatalysts have a relatively wide bandgap (Eg > 2.5 eV). At the same time, the range of reactions potentially possible for the narrow-bandgap semiconductors (Eg < 2.5 eV) is limited by substrates with the redox-potentials intermediary between ECB and EVB levels of semiconductor photocatalysts. Therefore, at a low Eg the primary photoinduced charge transfers are limited and invariably characterized by a low free Gibbs energy (Eq. 1.3). The bandgap expansion with a decrease of the size of semiconductor NPs can overcome these limitations. In particular, the QSEs can result in an expansion of the 20 1 Basic Concepts of the Photochemistry of Semiconductor … range of semiconductor photocatalysts due to the introduction of new materials, that have no photochemical activity in the bulk form. For example, as the size of PbSe NPs is reduced from d > 100 nm to 5–10 nm, this semiconductor becomes sus- ceptible to the reductive photocorrosion with the formation of Pb0 and can also participate in the photocatalytic redox-processes thermodynamically forbidden for bulk PbSe, such as the MV2+ and water reduction [58]. Also, as opposite to the bulk materials, CdSe NPs smaller than 5 nm can act as a photocatalyst of the water and CO2 reduction [58]. Molybdenum disulfide is photocatalytically passive in the oxidation of phenol and its derivatives when introduced in the form of either bulk crystals and 8–10-nm particles [101, 102]. At the same time, the formation of the oxidation products was detected chromatographically in the presence of 4–5-nm MoS2 NPs (Fig. 1.8a). The rise of photocatalytic activity of small MoS2 NPs was assigned [102]toa size-dependent increase of the VB potential because the generation of very active • OH radicals is only possible for 4–5-nm MoS2 NPs (Fig. 1.8b). This system can also be used for the illustration of another special feature of the photochemistry of semiconductor quantum-sized NPs. A size-dependent increase of the photogenerated charge carrier energies, though being extremely positive for the photoinduced charge transfers, is invariably accompanied by a blue shift of the absorption NP threshold. Figure 1.8c shows that molybdenum disulfide looses strongly the ability for the visible light absorption as the NP size is reduced from 8– 10 nm to 4–5 nm. This limitation has a general character—one should weight gains in charge energies and losses in the visible light harvesting when designing a photo-catalytic/photoelectrochemical system based on the quantum-sized semiconductor NPs. The photocatalytic reduction of benzophenone in acetonitrile can occur only in the presence of CdS NPs smaller than 4 nm [103]. Cadmium sulfide reveals also a

Fig. 1.8 a Kinetic curves of the photocatalytic oxidation of phenol in the presence of MoS2 NPs and nanocrystalline TiO2 Evonik P25; b band positions for various semiconductors with respect to the water oxidation redox-potential (at pH 7); c absorption spectra of colloidal MoS2 NPs of different sizes. Reprinted with permissions from [102]. Copyright (1999) American Chemical Society 1.3 Influence of Size Dependences of CB and VB Levels 21 photocatalytic activity in the reductive decomposition of sodium selenosulfate [104] and reduction of Ni(II) with sodium sulfide [105] when present in the form of 6–8-nm particles, while bulk CdS is inert in these processes. PbS NPs smaller than 3 nm reveal photocatalytic properties in the MV2+ reduction [106]. A similar effect is observed for CdSe and CdTe as the NP size is reduced to 3–5nm[94, 107]. ZnS nanocrystals can be used as a photocatalyst of the CO2 reduction while bulk zinc sulfide reveals no activity in this reaction [95, 108]. Analysis of the band edge positions of differently sized ZnS crystals [95] showed that the photoactivity of ZnS NPs originates from a size-dependent increase of the CB potential that becomes •− more negative than the redox potential of CO2/CO2 reduction (−1.9 V vs. NHE). A size decrease of Si NPs from 3–4to1–2 nm renders them active in the photocatalytic reduction of some organic dyes and CO2 [109]. As opposite to the microcrystalline zinc oxide, ZnO NPs can initiate the methylmethacrylate photopolymerization [110]. Similarly, MnO2 NPs revealed photocatalytic properties in the oxidative coupling of b-naphtol, untypical for the bulk material [111]. Acceleration of photocatalytic processes as a result of size-dependent increase of the CB and VB energies. A phenomenon of the acceleration of photocatalytic processes with a decrease of the semiconductor crystal size is broadly observed. In many cases, this effect stems not only from an increase of the specific surface area and the generation of surface defects participating in the interfacial charge transfers but also from changes in the band edge energies due to the QSEs. As discussed earlier, an increase in absolute ECB and EVB values results in a corresponding increase in the free energy of electron transfers, which affects the rate of photocatalytic processes. For example, CdS NPs [112] and ZnS NPs [113] were found to be much more efficient photocatalysts of the Rhodamine B degradation as compared to the corresponding bulk materials. A reduction of CdS and ZnS NP size from 5 to 2 nm and from *3 to 1.6 nm, respectively, results in a *5-fold acceleration of the photocatalytic dehalogenation of polyhalogenated aromatics [114]. Zeolite-hosted In2S3 NPs revealed a much higher photocatalytic activity in the hydrogen evolution from aqueous solutions, as compared to the bulk indium sulfide (Fig. 1.9a) as well as a high stability and reusability due to NP-host interactions [115]. Similarly, CdS nanocrystals formed in or anchored to the zeolites and mesoporous hosts revealed an enhanced photocatalytic activity in the hydrogen evolution from aqueous sulfide/sulfite solutions as compared to bulk CdS [116– 119]. A decrease of the CdS NP size from 5 to 3.8 nm, though being comparatively small, results in almost 40-fold acceleration of the photocatalytic reduction of some nitro-aromatic compounds (Fig. 1.9b) [120]. The specific rate of photoinduced buthylmethacrylate polymerization in the presence of CdS NPs was found to depend on NP size increasing by 60–70% as the NP size is reduced from *8 to 3.8 nm [121]. This effect is a result of the direct monomer photoreduction by CB electrons which is possible only in the case of the smallest 3.8-nm CdS NPs. For larger NPs, the photopolymerization can be initiated only by radicals generated via the oxidation of 2-propanol (solvent) with VB holes. 22 1 Basic Concepts of the Photochemistry of Semiconductor …

Fig. 1.9 a Kinetic curves of the hydrogen evolution over platinized zeolite/In2S3 NPs (■), non-platinized zeolite/In2S3 NPs (▲), and platinized (♦) and non-platinized (▼) bulk In2S3 under visible light illumination (k > 430 nm); b a ratio of rate constants of the photocatalytic and non-catalytic reduction of nitrotoluene as a function of CdS NP size. Reprinted with permissions from [115](a) and [120](b). Copyright (2006, 2008) Elsevier

The photocatalytic methylviologen reduction rate was also found to depend considerably on the size of colloidal CdS NPs [103, 122, 123] and In2S3 NPs [123]. A reduction of CdS NP size from 5 to 3 nm results in a *0.2 eV increment of the CB potential providing 4–5-fold acceleration of the photoinduced electron transfer from CdS NPs to MV2+. The dependence of the photoinduced electron transfer rate constant on ECB is linear when presented in the coordinates of Tafel equation, which is a typical relation between the rate of an electrochemical reaction and an over-voltage of the electrode charge transfer. In the current case, the over-voltage DE is provided by a difference between the size-dependent CB potential of CdS NPs and the MV2+/MV•+ pair redox-potential. Therefore, the Tafel equation can be bulk 0 2+ •+ bulk written as lg(k/k )=−aDE = −a(ECB(R) − E (MV /MV )), where k is the rate constant for bulk CdS, a is a constant. A study of the photocatalytic MV2+ reduction on the surface of a broad series of semiconductor NPs differing both by the composition and the size [94] showed that similar Tafel-like dependences between the methylviologen reduction rate (or +• quantum yield of MV radical) and ECB of colloidal semiconductor NPs are typical allowing to predict the efficiency of this process for any given NPs with the known ECB (Fig. 1.10a). A Tafel-like dependence was also observed between the rate of the photocatalytic nitrate reduction to NH3 and ECB of size-selected CdS NPs (Fig. 1.10b) [124]. The NPs studied in [124] belong to a very narrow size range of 2.0–2.2 nm, corresponding to the regime of strong spatial confinement. As a result of strong QSEs, a NP size reduction by mere 0.2 nm supplies an appreciable 0.25 V increment of the CB potential and a rate increase by a factor of 5–6[124]. A direct relationship between the electrochemical characteristics of the photogenerated charge carriers and NP size was experimentally proven for CdS NPs [125, 126] and Bi2S3 NPs [127]. In particular, as the CdS NP size is reduced from 1.3 Influence of Size Dependences of CB and VB Levels 23

Fig. 1.10 a A relationship between the quantum efﬁciency of the photocatalytic methylviologen •+ reduction Ф(MV ) and ECB of 5.0-nm ZnS NPs (point No. 1), 3.0-nm CdTe NPs (2), 5.5–5.6-nm CdSe NPs (3), Cd0.25Zn0.75S NPs (4), Cd0.50Zn0.50S NPs (5), Cd0.63Zn0.33S NPs (6), and CdS NPs with the size d = 6.5–6.6 nm (7), Cd0.75Zn0.25S (8), CdS NPs with d =10–11 nm (9), and 4.8-nm ZnO NPs (10). The solid line represents a linear ﬁt of the presented data; b Tafel plot of the natural log of the measured current density versus the reaction over-voltage for amine-capped CdS NPs. Reprinted with permissions from [94](a) and [124](b). Copyright (2010 a and 1997 b) Elsevier (a) and American Chemical Society (b)

4.5 to 3.9 nm the gap between anodic (oxidative) and cathodic (reductive) current peaks increases from 2.63 to 3.39 eV in line with a corresponding broadening of the optical bandgap from 3.06 to 3.23 eV [125]. A similar correlation between a size-dependent increment of the bandgap and a distance between the anodic and cathodic current peaks was established also for colloidal CdTe NPs [128].

1.4 Photoinduced Charging of Semiconductor NPs

Under very intense illumination several electron-hole couples (excitons) can be generated simultaneously in each semiconductor NPs. Interactions of two excitons give rise to various non-linear optical phenomena, their amplitude depending non-linearly on the light flux intensity. Typically, the non-linear effects depend quite strongly on the NP size [7, 8, 38, 63, 129–137]. Some of the non-linear effects can also be observed in the case of relatively low-intensity excitation, for example, for the AM1.5 light flux. In particular, PbX NPs (X = S, Se) display a pronounced tendency to multi-exciton generation [138– 141]. This phenomenon is observed when PbX NPs are excited at hv * nEg (n =3–10). As the thermalization of hot carriers is slow for small PbX NPs, the excitation energy can be accommodated by the generation of several (at least two) electron-hole pairs and, thus the light harvesting efficiency of PbX NP-based can theoretically be higher than 100–200% [140]. Another non-linear optical phenomenon can be observed when the rates of interfacial transfers of electrons and holes are strongly different resulting in the 24 1 Basic Concepts of the Photochemistry of Semiconductor … population of semiconductor NPs with an excessive charge. Typically for the metal chalcogenide NPs, the VB hole trapping and subsequent reactions both with chalcogenide (oxide) lattice and adsorbed substrates are quite fast, while electron transfer can be obstructed by many factors, such as a relatively low CB potential, resistance of NPs to the reductive corrosion, non-availability of suitable adsorbed acceptors, etc. As a result, the semiconductor NPs are typically populated with excessive electrons that fill the available states near the CB bottom. The excessive electrons are relatively long-lived and the NPs can be excited many times while being in the charged state. At that, the transition of each new electron from VB to CB requires a higher energy, because a portion of the lowest states near the CB edge gets occupied by the excessive electrons (the so-called “electrons-spectators”) and this portion increases with an increase of the excessive charge density. Therefore, the excitation of a charged NP requires a higher energy, than for the “normal” NP, Eg + DEB, where DEB—is an excess necessary to push an electron to the nearest available free electron state in CB. Obviously, DEB depends on the excessive charge density and, thus, on the light intensity and NP size (volume). This phenomenon is often referred to as Burstein-Moss effect [142–144], while DEB is called a Burstein shift [143, 144]. The Burstein-Moss effect was observed for the first time in strongly doped InSb [7]. Typically, bulk semiconductor crystals are moderately doped and have quite a high density of states near the CB edge. Therefore, an excessive charge density high enough to induce an appreciable optical shift cannot be achieved for bulk semiconductors even at intense photoexcitation. The situation changes dramatically for the semiconductor NPs, which have a tiny volume and partially quantized CB as a result of the QSEs. When a strong light pumping is applied to excite the quantum-sized NPs, an excessive charge with a density of the order of ne * 1026 m−3 can be created comparable to the free electron gas density of typical metals, *1028 m−3 [69, 143, 144]. A partial filling of CB states with excessive electrons can be observed as a blue shift of the fundamental absorption band edge (Fig. 1.11a) or as a negative “bleaching” band in differential absorption spectra of charged semiconductor NPs. The photoinduced blue shift of kbe as well as the intensity and spectral width of the non-stationary bleaching (NB) bands of semiconductor NPs depend on the excessive charge density ne. When ne becomes comparable with the free charge density in metals the Burstein-Moss effect can be described by Eq. (1.19) that relates the optical shift DEB = EF − ECB with ne [142, 144]. "# 3 Ã 2 2 D ¼ð þ me ÞÁ h 3ne À ð : Þ EB 1 Ã Ã p 4kT 1 19 mh 2me 8

In view of the obvious dependence of ne on the NP volume, the Burstein-Moss effect in semiconductor NPs has a pronounced size-dependent character. For the semiconductor NPs residing in the strong conﬁnement regime, that is, at R < aB,a high DEB can be reached already at a comparatively moderate photoexcitation 1.4 Photoinduced Charging of Semiconductor NPs 25

Fig. 1.11 a Absorption spectra of colloidal ZnO NPs in ethanol prior to (curve 1) and after the illumination (curve 2), kexc = 310–370 nm, curve 3 is a difference between curves 1 and 2 [148, 151, 161]; b NB bands of colloidal CdSe NPs with an average size of 2.7 nm (curve 1), 3.0 nm (2), and 3.2 nm (3) [104, 162]; c normalized kinetic curves of the NB decay in the NP band maximum (k = 345 nm) for colloidal ZnO NPs with an average size of 3.7 nm (curve 1) and 4.4 nm (2) [161, 163]

power, even under the stationary illumination [56, 144]. Apparently, at an equal ne the shift DEB will increase with a NP size decrease. As kbe of quantum-sized NPs shifts to lower values with a size decrease, the NP band maximum reveals a size-dependent blue shift as well (Fig. 1.11b). Finally, as the rate of interfacial transfer of excessive electrons depends also on the NP size, the NB relaxation rate also reveals a size-dependence, the smaller NPs discharging faster (Fig. 1.11c) [56, 64, 130, 143, 145, 146]. The relaxation of NB bands in the differential spectra (and the return of kbe to the original position in conventional absorption spectra) corresponds to the interfacial transfer of excessive charge to other components of the system—the solvent, dissolved oxygen or other electron acceptors. For example, kinetic curves of the NP decay presented in Fig. 1.11creflect gradual consumption of the excessive elec- − •− trons in the charge transfer to oxygen molecules: e +O2 ! O2 . Oppositely, the NB amplitude increases, when additional electron donors are introduced into the system. For example, the introduction of Na2SO3,Na2S, N2H4 or (CH3)2CHOH, a into aqueous CdS and CdxZn1−xS colloids results in 3–4-fold increase of the NB band intensity as a result of the efficient VB hole capture [56]. The excessive charge density depends on the light intensity and the NP size and composition. In the case of deaerated aqueous CdS colloids the charge accumulation results in the cathodic (reductive) photocorrosion and a partial transformation of CdS into metallic Cd [97, 99, 125] and, therefore, the Burstein-Moss effect for CdS NPs can be observed only under pulse photoexcitation in air- (oxygen-) saturated colloidal solutions, where the excessive charge is withdrawn by O2 after each light pulse, preserving the NP stability [56, 147]. In the case of colloidal ZnO NPs larger than 5 nm resistant to the reductive photocorrosion, the charge accumulation 26 1 Basic Concepts of the Photochemistry of Semiconductor … results in a blue kbe shift even under the stationary photoexcitation (Fig. 1.11a) and persists for many hours [148–150]. Using Eq. (1.19) one can show [150–152] that the case of DEB = 0.2 eV presented in Fig. 1.11a corresponds to the accumulation of 3–4 excessive electrons by each ZnO NP. The air admission into the illuminated solution results in the instant backward shift of kbe to the original position due to the transfer of excessive electrons to oxygen. The same mechanism of NB relaxation is valid for the case of pulse photoexcitation of air-saturated ZnO colloids (Fig. 1.11b). The charge transfer to oxygen occurs much faster for 3.7-nm ZnO NPs as compared with larger 4.4-nm NPs indicating a higher photochemical activity of smaller ZnO NPs. The Burstein-Moss effect can result in a considerable enhancement of the photocatalytic activity of semiconductor NPs [11, 12, 143–145]. In this view, studies of the characteristics and decay dynamics of NB band can provide unique information on the photoinduced charge transfer kinetics on the NP/electrolyte interface. Some examples of the application of pulse photolysis and NB phenomenon for probing of the photochemical behavior of semiconductor NPs are discussed in details in Chap. 6. The accumulation of an excessive charge alters significantly the photophysical and electrophysical properties of nanocrystalline semiconductors as well as the dynamics of interfacial charge transfer. In particular, it results in a cathodic polarization of NPs, that is, in a change of the potential of the double electric layer (DEL) on the NP surface, especially of its dense Helmholzian component [143]. A charging-induced increase of the optical bandgap can be used to calculate an excessive charge density ne and an average number of excessive electrons per NP, Ne, from Eq. (1.19). Then, a charging-altered DEL potential E* can be estimated using the reported typical values of the DEL capacity C of colloidal semiconductor NPs (around 0.06–0.10 F/m2 [153]) as [143]

Ã E ¼ ECB þ Ne=C ð1:20Þ

The inﬂuence of the Burstein-Moss effect on the interfacial charge transfer can be illustrated by photochemical processes occurring in mixed aqueous colloidal solutions containing *10-nm CdS NPs and size-selected 3.6–6.6-nm CdTe NPs [154]. The pulse photoexcitation of such colloids results in the electron transfer from CdS NPs to CdTe NPs evidenced by quenching of the NB band of CdS NPs. By using the above-discussed methodology, an average number of transferred electrons per CdTe NP DNe can be estimated from the reduction in the NB band intensity. As shown in Table 1.4, DNe increases with an increase of the CdTe NP size. An energy level scheme for the CdS–CdTe system (Fig. 1.12a) shows that the photoinduced electron transfer from the stationary CB level of CdS NPs to the CB level of CdTe NPs of any size meets a thermodynamic barrier. Therefore, it was supposed [154] that electrons come to CdTe NPs not from the stationary CdS CB level, but from a non-equilibrium higher-energy state E* generated as a result of the photoinduced charging of CdS NPs (Fig. 1.12a). Indeed, estimations performed using Eq. (1.20) showed that the accumulation of an excessive charge on the CdS 1.4 Photoinduced Charging of Semiconductor NPs 27

Table 1.4 Bandgap Eg and CB potential ECB of size-selected CdTe NPs, the electron transfer * over-voltage E − ECB and the average number of transferred electrons per CdTe NP DNe [154] * dCdTe, нм Eg,eV ECB, V (NHE) E − ECB,V DNe 3.0 2.24 −1.3 0.3 0.5 3.2 2.12 −1.2 0.4 0.6 5.0 1.82 −1.0 0.6 1.6 6.6 1.74 −0.9 0.7 4.5

NP/electrolyte interface under pulsed photoexcitation can result in an increment of the DEL potential as large as *0.7 eV, making possible the electron transfer from charged CdS NPs to CdTe NPs of any size studied (Fig. 1.12a). The dynamics of photoinduced electron transfer from the charged CdS NPs to * CdTe NPs is governed by the over-voltage, E − ECB(CdTe) depending on the size of CdTe NPs. Table 1.4 and Fig. 1.12 show that a decrease of the CdTe NP size from 6.6 to 3.0 nm results in a shift of the CB potential from around −0.9 V (versus NHE) to around −1.3 V (versus NHE), thus decreasing the electron transfer over-voltage and reducing the efﬁciency of this process.

Fig. 1.12 a Energy level scheme for a colloidal system containing CdS NPs and size-selected CdTe NPs; b Rate of the photoinduced reductive corrosion of ZnO NPs as a function of the NP Eg and size [148, 151] 28 1 Basic Concepts of the Photochemistry of Semiconductor …

In some cases, the above-discussed transition of a semiconductor from a photochemically-passive to a photochemically-active state can originate from a simultaneous contribution of the size-dependence of ECB energy and the photoinduced NP charging. For example, colloidal ZnO NPs in ethanol are resistant to the reductive photocorrosion if the NP size is larger than *5 nm. However, the stationary illumination of ZnO colloids with the smaller NPs results in the formation À ! 0 − of metallic Zn [148, 151]: ZnO + 2etr +H2O Zn + 2OH . The photoreduction starts only after the development of a photoinduced Burstein shift of 0.15–0.18 eV. In the size range of 3.7–4.4 nm the photocorrosion rate is directly proportional to the of ZnO NP bandgap (Fig. 1.12b). As the photocorrosion does not demand any reactants to diffuse to the ZnO NP surface, the sole reason for the increased photoactivity of ZnO NPs smaller than 4.8 nm can be a size-dependent increase of the energy of photogenerated charge carriers [148, 151]. Indeed, the CB potential of ZnO NPs shifts from −0.60 V (versus NHE) to −0.74 V (versus NHE) as the NP size is reduced from 4.8 to 3.7 nm. At the same time, the photocorrosion occurs only after the photoinduced charging of ZnO NPs, that is, under a cathodic polarization. According to Eq. (1.19), a Burstein shift of DEB = 0.15 eV corresponds to an additional shift of the CB potential of around 0.20 B [148]. Therefore, the total size- and polarization shifts can increase the CB level in ZnO NPs to E* = −0.80 V for 4.8-nm particles and to −0.94 V for the smallest 3.7-nm particles (Table 1.5). It was reported that the reductive dissolution of 5–6-nm ZnO NPs starts at the potentials more negative than Ecorr = −0.8 V (versus NHE) [155]. As shown in Table 1.5, the reductive photocorrosion is impossible for 4.8-nm ZnO NPs, even in * the state of the photoinduced polarization (|E |<|Ecorr|). For smaller ZnO NPs, * however, the condition |E |>|Ecorr| is valid and the NPs become unstable when illuminated by UV light in deaerated solutions. The above-discussed examples show that the capability of semiconductor NPs for the photoinduced charging can have a number of far-reaching consequences for their photochemical behavior. First, an increase of the over-voltage of the interfacial charge transfer results in an acceleration of the photochemical reactions. Additionally, the ﬁlling of the surface traps with excessive electrons blocks the radiative and non-radiative recombination channels and adds to an increased photoactivity of the charged NPs. Second, the accumulation of an excessive negative charge by semiconductor NPs should favor to multi-electron processes allowing to avoid one-electron reduction steps that sometimes can require very high redox

Table 1.5 Some * * Eg, ECB,V d, E ,V E − Ecorr, characteristics of size-selected eV (NHE) nm (NHE) V ZnO NPs 3.43 −0.58 4.8 −0.78 0.02 3.48 −0.61 4.4 −0.81 −0.01 3.50 −0.63 4.1 −0.83 −0.03 3.57 −0.68 3.9 −0.88 −0.08 3.63 −0.72 3.7 −0.92 −0.12 1.4 Photoinduced Charging of Semiconductor NPs 29 potentials. This factor is of a special importance for the photocatalytic multi-electron CO2 and N2 reduction discussed in Chap. 3. Finally, the photoinduced charging of semiconductor NPs creates a new high-energy excited state E* that can participate in processes impossible for “conventional” uncharged NPs,in particular, with the participation of substrates with the redox potential more negative than the CB level of uncharged semiconductor NPs. Concluding the brief and basic discussion of the photochemical behavior of semiconductor NPs, we can outline a number of special features differing nanocrystalline semiconductors from their bulk counterparts. These special features arise from the phenomena of the spatial exciton conﬁnement, the participation of surface states in photochemical reactions and the photoinduced charging of semiconductor NPs. An extremely high surface-to-volume ratio of semiconductor NPs with a size of a few nanometers creates numerous surface states (corresponding to surface defects, vacancies, dangling bonds, etc.) that can actively participate in the photophysical processes, in particular, by trapping the photogenerated charge carriers. At that, the nature, energy, and density of the surface states can sometimes dictate the possibility and rate of photochemical/photocatalytic processes with the participation of semiconductor NPs. The quantum size effects in semiconductor NPs, that is, size-dependent variation of basic electrophysical parameters, such as the bandgap, CB and VB levels, etc., can result in a dramatic enhancement of the photochemical processes as the semiconductor crystal size is reduced to a few nanometers. Also, semiconductors passive in the form of microcrystals, can become photochemically active when introduced as nanocrystals as a result of altered energies of the photogenerated charge carriers. A fundamental difference between the photochemical properties of semiconductor NPs and bulk counterparts is, therefore, in the occurrence of photochemical processes, when there is no mutual correspondence between the redox potentials of reactants and CB/VB levels of the semiconductor crystal. This feasibility originates not only from the size-dependent increase of the energy of charge carriers but also from the phenomenon of photoinduced charging of semiconductor NPs resulting in a radical change of the thermodynamics and kinetics of the interfacial charge transfers.

References

1. Grätzel M (ed) (1983) Energy resources through photochemistry and catalysis. Academic Press, Inc., New York 2. Fox MA, Dulay MT (1993) Heterogeneous photocatalysis. Chem Rev 93:341–357. doi:10. 1021/cr00017a016 3. Linsebigler AL, Lu G, Yates JT Jr (1995) Photocatalysis on TiO2 surfaces: principles, mechanisms, and selected results. Chem Rev 95:735–758. doi:10.1021/cr00035a013 30 1 Basic Concepts of the Photochemistry of Semiconductor …

4. Shiraishi Y, Hirai T (2008) Selective organic transformations on titanium dioxide-based photocatalysts. J Photochem Photobiol C 9:157–170. doi:10.1016/j.jphotochemrev.2008.05. 001 5. Imahori H, Umeyama T (2009) Donor-acceptor architecture on semiconducting electrodes for solar energy conversion. J Phys Chem C 113:9029–9039. doi:10.1021/jp9007448 6. Liu G, Wang L, Yang HG et al (2010) Titania-based photocatalysts—crystal growth, doping and hetero-structuring. J Mater Chem 20:831–843. doi:10.1039/B909930A 7. Gaponenko SV (1996) Optical properties of semiconductor nanocrystals. University Press, Cambridge 8. Wang Y, Herron N (1991) Nanometer-Sized semiconductor clusters: materials synthesis, quantum size effects, and photophysical properties. J Phys Chem 95:525–532. doi:10.1021/ j100155a009 9. Kryukov AI, Kuchmiy SY, Pokhodenko VD (2000) Energetics of electron processes in semiconductor photocatalytic systems. Theor Experim Chem 36:63–81. doi:10.1007/ BF02529022 10. Rogach A (ed) (2008) Semiconductor nanocrystal quantum dots: synthesis, assembly, spectroscopy and applications. Springer-Verlag GmbH, Vienna 11. Stroyuk OL, Kryukov AI, Kuchmiy SY, Pokhodenko VD (2005) Quantum size effects in photonics of semiconductor nanoparticles. Theor Experim Chem 41:67–91. doi:10.1007/ s11237-005-0025-9 12. Stroyuk OL, Kryukov AI, Kuchmiy SY, Pokhodenko VD (2005) Quantum size effects in semiconductor photocatalysis. Theor Experim Chem 41:207–228. doi:10.1007/s11237-005- 0042-8 13. Henglein A (1989) Small-particle research: physicochemical properties of extremely small colloidal metal and semiconductor particles. Chem Rev 89:1861–1873. doi:10.1021/ cr00098a010 14. Hagfeldt A, Grätzel M (1995) Light-induced redox reactions in nanocrystalline systems. Chem Rev 95:49–68. doi:10.1021/cr00033a003 15. Eychmüller A (2000) Structure and photophysics of semiconductor nanocrystals. J Phys Chem B 104:6514–6528. doi:10.1021/jp9943676 16. Brus L (1986) Electronic wave functions in semiconductor clusters: experiment and theory. J Phys Chem 90:2555–2560. doi:10.1021/j100403a003 17. Adams DM, Kamat PV, Marcus RA et al (2003) Charge transfer on the nanoscale: current status. J Phys Chem B 107:6668–6697. doi:10.1021/jp0268462 18. Zhao Y, Zhu K (2016) Organic–inorganic hybrid lead halide perovskites for optoelectronic and electronic applications. Chem Soc Rev 45:655–689. doi:10.1039/c4cs00458b 19. Grätzel M (2014) The light and shade of perovskite solar cells. Nat Mater 13:838–842. doi:10.1038/nmat4065 20. Sharma D, Jha R, Kumar S (2016) Quantum dot sensitized solar cells: recent advances and future perspectives in photoanode. Sol Energy Mater Sol Cells 155:294–322. doi:10.1016/j. solmat.2016.05.062 21. Aldakov D, Lefrançois A, Reiss P (2013) Ternary and quaternary metal chalcogenide nanocrystals: synthesis, properties and applications. J Mater Chem C 1:3756–3776. doi:10. 1039/C3TC30273C 22. Thomas SR, Chen CW, Date M et al (2016) Recent developments in the synthesis of nanostructured chalcopyrite materials and their applications: a review. RSC Adv 6:60643– 60656. doi:10.1039/C6RA05502H 23. Beydoun D, Amal R, Low G, McEvoy S (1999) Role of nanoparticles in photocatalysis. J Nanopart Res 1:439–458. doi:10.1023/A:1010044830871 24. El-Sayed MA (2004) Small is different: shape-, size-, and composition-dependent properties of some colloidal semiconductor nanocrystals. Acc Chem Res 37:326–333. doi:10.1021/ ar020204f 25. Alivisatos AP (1996) Perspectives on the physical chemistry of semiconductor nanocrystals. J Phys Chem 100:13226–13239. doi:10.1021/jp9535506 References 31

26. Landolt-Börnstein (1982) Numerical data and functional relationships in science and technology, Group III, vol 17b—Semiconductors, Sect. 3.10.1. Springer, Berlin 27. Khairutdinov RF (1998) Chemistry of semiconductor nanoparticles. Russ Chem Rev 67:109–122 28. Alivisatos AP (1996) Semiconductor clusters, nanocrystals, and quantum dots. Science 271:933–937 29. Shankar SS, Chatterjee S, Sastry M (2003) Synthesis of CdS nanoparticles within thermally evaporated aerosol OT thin films. Phys Chem Commun 6:36–39. doi:10.1039/B303919F 30. Murakoshi K, Hosokawa H, Saitoh M et al (1998) Preparation of size-controlled hexagonal CdS nanocrystallites and the characteristics of their surface structures. J Chem Soc Faraday Trans 94:579–586. doi:10.1039/A707192B 31. Vossmeyer T, Katsikas L, Giersig M et al (1994) CdS nanoclusters: synthesis, characterization, size dependent oscillator strength, temperature shift of the excitonic transition energy, and reversible absorbance shift. J Phys Chem 98:7665–7673. doi:10.1021/ j100082a044 32. Nosaka Y, Shigeno H, Ikeuchi T (1995) Formation of polynuclear cadmium-thiolate complexes and CdS clusters in aqueous solution studied by means of stopped-flow and NMR spectroscopies. J Phys Chem 99:8317–8322. doi:10.1021/j100020a066 33. Yu WW, Qu L, Guo W, Peng X (2003) Experimental determination of the extinction coefficient of CdTe, CdSe, and CdS nanocrystals. Chem Mater 15:2854–2860. doi:10.1021/ cm034081k 34. Choi KM, Shea KJ (1994) Amorphous polysilsesquioxanes as a confinement matrix for quantum-sized particle growth. Size analysis and quantum size effect of CdS particles grown in porous polysilsesquioxanes. J Phys Chem 98:3207–3214. doi:10.1021/j100063a026 35. Murray CB, Norris DJ, Bawendi MG (1993) Synthesis and characterization of nearly monodisperse CdE (E = sulfur, selenium, tellurium) semiconductor nanocrystallites. J Am Chem Soc 115:8706–8715. doi:10.1021/ja00072a025 36. Soloviev VN, Eichhöfer A, Fenske D, Banin U (2000) Molecular limit of a bulk semiconductor: size dependence of the “band gap” in CdSe cluster molecules. J Am Chem Soc 122:2673–2674. doi:10.1021/ja9940367 37. Rivera-Marquez A, Rubin-Falfan M, Lozada-Moralez R et al (2001) Quantum confinement and crystalline structure of CdSe nanocrystalline films. Phys Stat Sol (a) 188:1059–1064 38. Brysch A, Bour G, Neuendorf R, Kreibig U (1999) Nonlinear optical spectroscopy of embedded semiconductor clusters. Appl Phys B 68:447–451. doi:10.1007/s003400050647 39. Rajh T, Micic OI, Nozik AJ (1993) Synthesis and characterization of surface-modified colloidal cadmium telluride quantum dots. J Phys Chem 97:11999–12003. doi:10.1021/ j100148a026 40. Sharma SN, Kohli S, Rastogi AC (2005) Quantum confinement effects of CdTe nanocrystals sequestered in TiO2 matrix: effect of oxygen incorporation. Physica E 25:554–561. doi:10. 1016/j.physe.2004.08.110 41. Wong EM, Hörtz PG, Liang CJ et al (2001) Influence of organic capping ligands on the growth kinetics of ZnO nanoparticles. Langmuir 17:8362–8367. doi:10.1021/la010944h 42. Viswanatha R, Sapra S, Satpati B et al (2004) Understanding the quantum size effects in ZnO nanocrystals. J Mater Chem 14:661–668. doi:10.1039/B310404D 43. Pesika NS, Stebe KJ, Searson PC (2003) Relationship between absorbance spectra and particle size distributions for quantum-sized nanocrystals. J Phys Chem B 107:10412–10415. doi:10.1021/jp0303218 44. Green M, Wakefield G, Dobson PJ (2003) A simple metalorganic route to organically passivated mercury telluride nanocrystals. J Mater Chem 13:1076–1078. doi:10.1039/ B210563B 45. Miyoshi H, Yamachika M, Yoneyama H, Mori H (1990) Photochemical properties of PbS microcrystallites prepared in nafion. J Chem Soc Faraday Trans 86:815–818. doi:10.1039/ FT9908600815 32 1 Basic Concepts of the Photochemistry of Semiconductor …

46. Kang I, Wise W (1997) Electronic structure and optical properties of PbS and PbSe quantum dots. J Opt Soc Am B 14:1632–1646. doi:10.1364/JOSAB.14.001632 47. Sashchiuk A, Langof L, Chaim R, Lifshitz E (2002) Synthesis and characterization of PbSe and PbSe/PbS core–shell colloidal nanocrystals. J Cryst Growth 240:431–438. doi:10.1016/ S0022-0248(02)01156-9 48. Allan G, Delerue C (2004) Confinement effects in PbSe quantum wells and nanocrystals. Phys Rev B 70:245321 49. Joswig JO, Springborg M, Seifert G (2000) Structural and electronic properties of cadmium sulfide clusters. J Phys Chem B 104:2617–2622. doi:10.1021/jp993512c 50. Gurin VS (1996) Electronic structure of CdS cores in Cd thiolate complexes and clusters. J Phys Chem 100:869–872. doi:10.1021/jp9527620 51. Li J, Wang LW (2003) High energy excitations in CdSe quantum rods. Nano Lett 3:101– 105. doi:10.1021/nl020231z 52. Hu J, Wang L, Li L et al (2002) Semiempirical pseudopotential calculation of electronic states of CdSe quantum rods. J Phys Chem B 106:2447–2452. doi:10.1021/jp013204q 53. Woggon U, Wind O, Gindele F et al (1996) Optical transitions in CdSe quantum dots: from discrete levels to broad gain spectra. J Lumin 70:269–280. doi:10.1016/0022-2313(96) 00060-9 54. Rayevska OE, Grodzyuk GY, Dzhagan VM et al (2010) Synthesis and characterization of white-emitting CdS quantum dots stabilized with polyethyleneimine. J Phys Chem C 114:22478–22486. doi:10.1021/jp108561u 55. Raevskaya AE, Grodzyuk GY, Stroyul OL et al (2010) Preparation and spectral properties of high-efficiency luminescent polyethyleneimine-stabilized CdS quantum dots. Theor Exp Chem 46:233–238. doi:10.1007/s11237-010-9145-y 56. Raevskaya AE, Stroyuk OL, Kuchmiy SY (2003) Optical characteristics of colloidal nanoparticles of CdS stabilized with sodium polyphosphate and their behavior during pulse photoexcitation. Theor Exp Chem 39:153–160. doi:10.1023/A:1024933023783 57. Raevskaya AE, Grodzyuk GY, Korzhak AV et al (2010) Preparation and optical properties of polyethyleneimine-stabilized colloidal CdSe and CdSxSe1−x quantum dots. Theor Experim Chem 46:397–402. doi:10.1007/s11237-011-9173-2 58. Nedeljkovic JM, Nenadovic MT, Micic OI, Nozik AJ (1986) Enhanced photoredox chemistry in quantized semiconductor colloids. J Phys Chem 90:12–13. doi:10.1021/ j100273a005 59. Steckel JS, Coe-Sullivan S, Bulovic V, Bawendi MG (2003) 1.3 lm to 1.55 lm tunable electroluminescence from PbSe quantum dots embedded within an organic device. Adv Mater 15:1862–1866. doi:10.1002/adma.200305449 60. Nenadovic MT, Rajh T, Mi OL (1985) Size quantization in small semiconductor particles. J Phys Chem 89:397–399. doi:10.1021/j100249a004 61. Kavan L, Stoto T, Grätzel M et al (1993) Quantum size effects in nanocrystalline semiconducting titania layers prepared by anodic oxidative hydrolysis of titanium trichloride. J Phys Chem 97:9493–9498. doi:10.1021/j100139a038 62. van Dijken A, Janssen AH, Smitsmans MH et al (1998) Size-selective photoetching of nanocrystalline semiconductor particles. Chem Mater 10:3513–3522. doi:10.1021/ cm980715p 63. Klimov VI (2000) Optical nonlinearities and ultrafast carrier dynamics in semiconductor nanocrystals. J Phys Chem B 104:6112–6123. doi:10.1021/jp9944132 64. Raevskaya AE, Stroyuk OL, Kuchmii SY (2004) Photocatalytic oxidation of hydrosulfide ions by molecular oxygen over cadmium sulfide nanoparticles. J Nanopart Res 6:149–158. doi:10.1023/B:NANO.0000034719.30620.d3 65. Schmelz O, Mews A, Basché T et al (2001) Supramolecular complexes from CdSe nanocrystals and organic fluorophors. Langmuir 17:2861–2865. doi:10.1021/la0016367 66. Leatherdale CA, Woo WK, Mikulec FV, Bawendi MG (2002) On the absorption cross section of CdSe nanocrystal quantum dots. J Phys Chem B 106:7619–7622. doi:10.1021/ jp025698c References 33

67. Striolo A, Ward J, Prausnitz JM et al (2002) Molecular weight, osmotic second virial coefficient, and extinction coefficient of colloidal CdSe nanocrystals. J Phys Chem B 106:5500–5505. doi:10.1021/jp020170t 68. Zhang JZ, O’Neil RH, Roberti TW (1994) Femtosecond studies of photoinduced electron dynamics at the liquid-solid interface of aqueous CdS colloids. J Phys Chem 98:3859–3864. doi:10.1021/j100065a049 69. Zhang JZ (2000) Interfacial charge carriers dynamics of colloidal semiconductor nanoparticles. J Phys Chem B 104:7239–7253. doi:10.1021/jp000594s 70. Cropek D, Kemme PA, Makarova OV et al (2008) Selective photocatalytic decomposition of nitrobenzene using surface modified TiO2 nanoparticles. J Phys Chem C 112:8311–8318. doi:10.1021/jp712137x 71. Yoon M, Chang JA, Kim Y et al (2001) Heteropoly acid-incorporated TiO2 colloids as novel photocatalytic systems resembling the photosynthetic reaction center. J Phys Chem B 105:2539–2545. doi:10.1021/jp003736r 72. Scotti R, Bellobono IR, Canevali C et al (2008) Sol–gel pure and mixed-phase titanium dioxide for photocatalytic purposes: relations between phase composition, catalytic activity, and charge-trapped sites. Chem Mater 20:4051–4061. doi:10.1021/cm800465n 73. Fitzmaurice DJ, Frei H (1991) Transient near-infrared spectroscopy of visible light − sensitized oxidation of I at colloidal TiO2. Langmuir 7:1129–1137. doi:10.1021/ la00054a019 74. Boschloo G, Fitzmaurice D (1999) Spectroelectrochemical investigation of surface states in nanostructured TiO2 electrodes. J Phys Chem B 103:2228–2231. doi:10.1021/jp984414e 75. Dimitrijevic NM, Saponjic ZV, Rabatic BM et al (2007) Effect of size and shape of nanocrystalline TiO2 on photogenerated charges. An EPR study. J Phys Chem C 111:14597– 14601. doi:10.1021/jp0756395 76. Ke SC, Wang TC, Wong MS, Gopal NO (2006) Low temperature kinetics and energetics of the electron and hole traps in irradiated TiO2 nanoparticles as revealed by EPR spectroscopy. J Phys Chem B 110:11628–11634. doi:10.1021/jp0612578 77. Berger T, Sterrer M, Diwald O et al (2005) Light-induced charge separation in anatase TiO2 particles. J Phys Chem B 109:6061–6068. doi:10.1021/jp0404293 78. Rabani J, Yamashita K, Ushida K et al (1998) Fundamental reactions in illuminated titanium dioxide nanocrystallite levels studied by pulsed laser. J Phys Chem B 102:1689–1695. doi:10.1021/jp973411j 79. Yoshihara T, Katoh R, Furube A et al (2004) Identification of reactive species in photoexcited nanocrystalline TiO2 films by wide-wavelength-range (400–2500 nm) transient absorption spectroscopy. J Phys Chem B 108:3817–3823. doi:10.1021/jp031305d 80. Tang J, Durrant JR, Klug DR (2008) Mechanism of photocatalytic water splitting in TiO2. Reaction of water with photoholes, importance of charge carrier dynamics, and evidence for four-hole chemistry. J Am Chem Soc 130:13885–13891. doi:10.1021/ja8034637 81. Fu H, Zhang L, Zhang S et al (2006) Electron spin resonance spin-trapping detection of radical intermediates in N-doped TiO2-assisted photodegradation of 4-chlorophenol. J Phys Chem B 110:3061–2065. doi:10.1021/jp055279q 82. Kamat PV, Dimitrijević NM, Fessenden RW (1988) Photoelectrochemistry in particulate systems. 7. Electron-transfer reactions of indium sulfide semiconductor colloids. J Phys Chem 92:2324–2329. doi:10.1021/j100286a029 83. Kamat PV, Ebbesen TW, Dimitrijević NM, Nozik AJ (1989) Primary photo-chemical events in CdS semiconductor colloids as probed by picosecond laser flash photolysis. Chem Phys Lett 157:384–389. doi:10.1016/0009-2614(89)87267-7 84. Jortner K, Rao CNR (2002) Nanostructured advanced materials. Perspectives and directions. Pure Appl Chem 74:1491–1506 85. Nozik AJ, Memming R (1996) Physical chemistry of semiconductor–liquid interfaces. J Phys Chem 100:13061–13078. doi:10.1021/jp953720e 34 1 Basic Concepts of the Photochemistry of Semiconductor …

86. Yang X, Tamai N (2001) How fast is interfacial hole transfer? In situ monitoring of carrier dynamics in anatase TiO2 nanoparticles by femtosecond laser spectroscopy. Phys Chem Chem Phys 3:3393–3398. doi:10.1039/B101721G 87. Matylitsky VV, Dworak L, Breus VV et al (2009) Separation in multiexcited CdSe quantum dots mediated by adsorbed electron acceptors. J Am Chem Soc 131:2424–2425. doi:10. 1021/ja808084y 88. Logunov S, Green T, Marguet S, El-Sayed MA (1998) Interfacial carriers dynamics of CdS nanoparticles. J Phys Chem A 102:5652–5658. doi:10.1021/jp980387g 89. Rogach A (ed) (2008) Semiconductor nanocrystal quantum dots: synthesis, assembly, spectroscopy and applications. Springer, Vienna 90. Chestnoy N, Harris TD, Hull R, Brus LE (1986) Luminescence and photophysics of cadmium sulfide semiconductor clusters: the nature of the emitting electronic state. J Phys Chem 90:3393–3399. doi:10.1021/j100406a018 91. Ramsden JJ, Grätzel M (1984) Photoluminescence of small cadmium sulphide particles. J Chem Soc Faraday Trans 1(80):919–933. doi:10.1039/F19848000919 92. Kamat PV, Patrick B (1992) Photophysics and photochemistry of quantized ZnO colloids. J Phys Chem 96:6829–6834. doi:10.1021/j100195a055 93. Ramsden JJ, Webber SE, Grätzel M (1985) Luminescence of colloidal cadmium sulfide particles in acetonitrile and acetonitrile/water mixtures. J Phys Chem 89:2740–2743. doi:10. 1021/j100259a007 94. Stroyuk O, Raevskaya AE, Kozytskiy AV, Kychmiy SY (2010) Electron energy factors in photocatalytic methylviologen reduction in the presence of semiconductor nanocrystals. J Photochem Photobiol A 210:209–214. doi:10.1016/j.jphotochem.2009.11.019 95. Kanemoto M, Shiragami T, Pac C, Yanagida S (1992) Semiconductor photocatalysis. Effective photoreduction of carbon dioxide catalyzed by ZnS quantum crystallites with low density of surface defects. J Phys Chem 96:3521–3526. doi:10.1021/j100187a062 96. Stroyuk OL, Raevskaya AE, Korzhak AV, Kuchmiy SY (2007) ZnS nanoparticles: spectral properties and photocatalytic activity in metals reduction reactions. J Nanopart Res 9:1027– 1039. doi:10.1007/s11051-006-9183-5 97. Stroyuk OL, Shvalagin VV, Raevskaya AE et al (2003) Photocatalysis of the reduction of Cd2+ ions by CdS nanoparticles in isopropyl alcohol. Theor Experim Chem 39:341–346. doi:10.1023/B:THEC.0000013985.94005.c3 98. Dzhagan VM, Stroyuk OL, Raevskaya AE et al (2010) Spectroscopic and photochemical + 2+ 2+ 3+ study of Ag -, Cu -, Hg -, and Bi -doped CdxZn1−xS nanoparticles. J Colloid Interface Sci 345:515–523. doi:10.1016/j.jcis.2010.02.001 99. Henglein A (1997) Nanoclusters of semiconductors and metals: colloidal nanoparticles of semiconductors and metals: electronic structure and processes. Ber Bunsenges Phys Chem 101:1562–1572. doi:10.1002/bbpc.19971011103 100. Faust BC, Hoffmann MR, Bahnemann DW (1989) Photocatalytic oxidation of sulfur dioxide in aqueous suspensions of a-Fe2O3. J Phys Chem 93:6371–6381. doi:10.1021/j100354a021 101. Wilcoxon JP (2000) Catalytic photooxidation of pentachlorophenol using semiconductor nanoclusters. J Phys Chem B 104:7334–7343. doi:10.1021/jp0012653 102. Thurston TR, Wilcoxon JP (1999) Photooxidation of organic chemicals catalyzed by nanoscale MoS2. J Phys Chem B 103:11–17. doi:10.1021/jp982337h 103. Yanagida S, Ogata T, Shindo A et al (1995) Semiconductor photocatalysis: size control of surface-capped CdS nanocrystallites and the quantum size effect in their photocatalysis. Bull Chem Soc Jpn 68:752–758. doi:10.1246/bcsj.68.752 104. Raevskaya AE, Stroyuk OL, Kuchmiy SY (2006) Preparation of colloidal CdSe and CdS/ CdSe nanoparticles from sodium selenosulfate in aqueous polymers solutions. J Colloid Interface Sci 302:133–141. doi:10.1016/j.jcis.2006.06.018 105. Stroyuk OL, Korzhak AV, Raevskaya AE, Kuchmiy SY (2004) Photocatalysis of the release of molecular hydrogen from aqueous solutions of sodium sulfite by composite CdS/Ni nanoparticles. Theor Experim Chem 40:1–6. doi:10.1023/B:THEC.0000020758.50103.8f References 35

106. Torimoto T, Sakata T, Mori H, Yoneyama H (1994) Effect of surface charge of 4-aminothiophenol-modified PbS microcrystal photocatalysts on photoinduced charge transfer. J Phys Chem 98:3036–3043. doi:10.1021/j100062a048 107. Kovalenko MV, Bodnarchuk MV, Stroyuk OL, Kuchmiy SY (2004) Synthesis, optical, and photocatalytic properties of quantum-sized particles of CdTe. Theor Experim Chem 40:220– 225. doi:10.1023/B:THEC.0000041806.60632.04 108. Kanemoto M, Hosokawa H, Wada Y et al (1996) Semiconductor photocatalysis. Part 20. Role of surface in the photoreduction of carbon dioxide catalyzed by colloidal ZnS nanocrystallites in organic solvent. J Chem Soc Faraday Trans 92:2401–2411. doi:10.1039/ FT9969202401 109. Kang Z, Tsang CHA, Wong NB et al (2007) Silicon quantum dots: a general photocatalyst for reduction, decomposition, and selective oxidation reactions. J Am Chem Soc 129:12090– 12091. doi:10.1021/ja075184x 110. Hoffman AJ, Yee H, Mills G, Hofmann MR (1992) Photoinitiated polymerization of methyl methacrylate using Q-sized ZnO colloids. J Phys Chem 96:5540–5546. doi:10.1021/ j100192a066 111. Jana S, Pande S, Sinha AK, Pal T (2008) Synthesis of superparamagnetic b-MnO2 organosol: a photocatalyst for the oxidative phenol coupling reaction. Inorg Chem 47:5558– 5560. doi:10.1021/ic800499q 112. Guo Y, Zhang H, Wang Y et al (2005) Controlled growth and photocatalytic properties of CdS nanocrystals implanted in layered metal hydroxide matrixes. J Phys Chem B 105:21602–21607. doi:10.1021/jp054400q 113. Zhao Q, Xie Y, Zhang Z, Bai X (2007) Size-selective synthesis of zinc sulfide hierarchical structures and their photocatalytic activity. Cryst Growth Des 7:153–158. doi:10.1021/ cg060521j 114. Yin H, Wada Y, Kitamura T, Yanagida S (2001) Size-selective synthesis of zinc sulfide hierarchical structures and their photocatalytic activity. Environ Sci Technol 35:227–231. doi:10.1021/cg060521j 115. Shen S, Guo L (2006) Structural, textural and photocatalytic properties of quantum-sized In2S3-sensitized Ti-MCM-41 prepared by ion-exchange and sulfidation methods. J Sol State Chem 179:2629–2635. doi:10.1016/j.jssc.2006.05.010 116. Sathish M, Viswanathan B, Viswanath RP (2006) Alternate synthetic strategy for the preparation of CdS nanoparticles and its exploitation for water splitting. Inter J Hydrogen En 31:891–898. doi:10.1016/j.ijhydene.2005.08.002 117. Sathish M, Viswanath RP (2007) Photocatalytic generation of hydrogen over mesoporous CdS nanoparticle: effect of particle size, noble metal and support. Catal Today 129:421–427. doi:10.1016/j.cattod.2006.12.008 118. Shen S, Guo L (2008) Growth of quantum-confined CdS nanoparticles inside Ti-MCM-41 as a visible light photocatalyst. Mater Res Bull 43:437–446. doi:10.1016/j.materresbull.2007. 02.034 119. Bao N, Shen L, Takata T et al (2007) Facile Cd-thiourea complex thermolysis synthesis of phase-controlled CdS nanocrystals for photocatalytic hydrogen production under visible light. J Phys Chem C 111:17527–17534. doi:10.1021/jp076566s 120. Datta A, Priyam A, Bhattacharyya SN et al (2008) Temperature tunability of size in CdS nanoparticles and size dependent photocatalytic degradation of nitroaromatics. J Colloid Interface Sci 322:128–135. doi:10.1016/j.jcis.2008.02.052 121. Stroyuk OL, Granchak VM, Korzhak AV, Kuchmii SY (2004) Photoinitiation of buthylmethacrylate polymerization by colloidal semiconductor nanoparticles. J Photochem Photobiol A 162:339–351. doi:10.1016/S1010-6030(03)00371-X 122. Matsumoto H, Uchida H, Matsunaga T et al (1994) Photoinduced reduction of viologens on size-separated CdS nanocrystals. J Phys Chem 98:11549–11556. doi:10.1021/j100095a041 123. Nosaka Y, Ohta N, Miyama H (1990) Photochemical kinetics of ultrasmall semiconductor particles in solution: effect of size on the quantum yield of electron transfer. J Phys Chem 94:3752–3755. doi:10.1021/j100372a073 36 1 Basic Concepts of the Photochemistry of Semiconductor …

124. Korgel BA, Monbouquette HG (1997) Quantum confinement effects enable photocatalyzed nitrate reduction at neutral pH using CdS nanocrystals. J Phys Chem B 101:5010–5017. doi:10.1021/jp971016n 125. Haram SK, Quinn BM, Bard AJ (2001) Electrochemistry of CdS nanoparticles: a correlation between optical and electrochemical band gaps. J Am Chem Soc 123:8860–8861. doi:10. 1021/ja0158206 126. Weib H, Fernandez A, Kisch H (2001) Elektronische Halbleiter-träger-wechselwirkung – ein neuer Effekt in der Halbleiterphotokatalyse. Angew Chem 113:3942–3945 127. Riley DJ, Waggett JP, Wijayantha KGU (2004) Colloidal bismuth sulfide nanoparticles: a photoelectrochemical study of the relationship between bandgap and particle size. J Mater Chem 14:704–708. doi:10.1039/B311517H 128. Haram SK, Kshirsagar A, Gujarathi YD et al (2011) Quantum confinement in CdTe quantum dots: investigation through cyclic voltammetry supported by density functional theory (DFT). J Phys Chem C 115:6243–6249. doi:10.1021/jp111463f 129. Burda C, El-Sayed MA (2000) High-density femtosecond transient absorption spectroscopy of semiconductor nanoparticles. A tool to investigate surface quality. Pure Appl Chem 72:165 130. Haase M, Weller H, Henglein A (1988) Photochemistry of colloidal semiconductors. 26. Photoelectron emission from cadmium sulfide particles and related chemical effects. J Phys Chem 92:4706–4712. doi:10.1021/j100327a030 131. Jacobsohn M, Banin U (2000) Dependence of second harmonic generation in CdSe nanocrystal quantum dots. J Phys Chem B 104:1–5. doi:10.1021/jp9925076 132. Wuister SF, de Mello Donega C, Meijerink A (2004) Influence of thiol capping on the exciton luminescence and decay kinetics of CdTe and CdSe quantum dots. J Phys Chem B 108:17393–17397. doi:10.1021/jp047078c 133. Tan GL, Yang Q, Hömmerlich U et al (2004) Linear and non-linear optical properties of capped CdTe nanocrystals prepared by mechanical alloying. Opt Mater 27:579–584. doi:10. 1016/j.optmat.2004.06.005 134. Wu F, Lewis JW, Kliger DS, Zhang JZ (2003) Unusual excitation intensity dependence of fluorescence of CdTe nanoparticles. J Chem Phys 118:12–16. doi:10.1063/1.1533733 135. Zheng MJ, Zhang LD, Zhang JG (2001) Size dependence of non-linear optical properties of SiO2 thin films containing InP nanocrystals. Appl Phys A 73:183. doi:10.1007/ s003390000615 136. Yang C, Yu P, Wang J et al (2000) Synthesis and characterization of poly(vinylpyrrolidone)- modified zinc oxide nanoparticles. Chem Mater 12:2268–2274. doi:10.1021/cm9907817 137. Ganeev RA, Baba M, Morita M et al (2004) Thermally induced and Kerr-induced optical nonlinearities of a pseudoisocyanine solution at 532 nm. J Opt A 6:1076 138. Shabaev A, Efros AL, Nozik AJ (2006) Multiexciton generation by a single photon in nanocrystals. Nano Lett 6:2856–2863. doi:10.1021/nl062059v 139. Ellingson RJ, Beard MC, Johnson JC et al (2006) Highly efficient multiple exciton generation in colloidal PbSe and PbS quantum dots. Nano Lett 5:865–871. doi:10.1021/ nl0502672 140. Nozik AJ (2008) Multiple exciton generation in semiconductor quantum dots. Chem Phys Lett 457:3–11. doi:10.1016/j.cplett.2008.03.094 141. Midgett AG, Luther JM, Stewart JT et al (2013) Size and composition dependent multiple exciton generation efficiency in PbS, PbSe, and PbSxSe1−x alloyed quantum dots. Nano Lett 13:3078–3085. doi:10.1021/nl4009748 142. Kamat PV, Dimitrijević NM, Nozik AJ (1989) Dynamic Burstein-Moss shift in semiconductor colloids. J Phys Chem 93:2873–2875. doi:10.1021/j100345a003 143. Savinov EN, Nagorky VE, Parmon VN (1994) Influence of excessive charge of colloidal cadmium sulfide particles on the rate of interfacial charge transfer. Khim Fiz 13:56–65 144. Liu C, Bard AJ (1989) Effect of excess charge on band energetics (optical absorption edge and carrier redox potentials) in small semiconductor particles. J Phys Chem 93:3232–3237. doi:10.1021/j100345a069 References 37

145. Albery WJ, Brown GT, Darwent JR, Saeivar-Iranizad E (1985) Time-resolved photoredox reactions of colloidal CdS. J Chem Soc Faraday Trans 1(81):1999–2007. doi:10.1039/ F19858101999 146. Bavykin DV, Savinov EN, Parmon VN (2000) Studies on the kinetics of interfacial electron transfer sensitized by colloidal CdS. J Photochem Photobiol A 130:57–61. doi:10.1016/ S1010-6030(99)00196-3 147. Kamat PV, Shanghavi B (1997) Interparticle electron transfer in metal/semiconductor composites. Picosecond dynamics of CdS-capped gold nanoclusters. J Phys Chem B 101:7675–7679. doi:10.1021/jp9709464 148. Shvalagin VV, Stroyuk OL, Kuchmiy SY (2004) Role of quantum size effects in the cathodic photocorrosion of ZnO nanoparticles in ethanol. Theor Experim Chem 40:378–382. doi:10.1007/s11237-005-0003-2 149. Cao G (2004) Nanostructures & nanomaterials: synthesis, properties & applications. Imperial college press, London 150. Stroyuk OL, Shvalagin VV, Kuchmiy SY (2004) Photochemical synthesis, spectral-optical and electrophysical properties of composite nanoparticles of ZnO/Ag. Theor Exp Chem 40:98–104. doi:10.1023/B:THEC.0000028904.52818.e7 151. Stroyuk OL, Shvalagin VV, Kuchmii SY (2005) Photochemical synthesis and optical properties of binary and ternary metal-semiconductor composites based on zinc oxide. J Photochem Photobiol A 173:185–194. doi:10.1016/j.jphotochem.2005.02.002 152. Stroyuk OL, Shvalagin VV, Kuchmiy SY (2007) Photochemical synthesis of ZnO/Ag nanocomposites. J Nanopart Res 9:427–440. doi:10.1007/s11051-006-9086-5 153. Liver N, Nitzan A (1992) Redox properties of small semiconductor particles. J Phys Chem 96:3366–3373. doi:10.1021/j100187a035 154. Bodnarchuk MI, Kovalenko MV, Stroyuk OL, Kuchmiy SY (2004) Photoinduced electron transfer between CdS and CdTe in colloidal solutions. Theor Exp Chem 40:287–292. doi:10. 1023/B:THEC.0000049074.31907.81 155. Hoyer P, Weller H (1995) Potential-dependent electron injection in nanoporous colloidal ZnO films. J Phys Chem 99:14096–14100. doi:10.1021/j100038a048 156. Frenzel J, Joswig JO, Seifert G (2007) Optical excitations in cadmium sulfide nanoparticles. J Phys Chem C 111:10761–10770. doi:10.1021/jp071125u 157. Leistikow MD, Johansen J, Kettelarij AJ et al (2009) Size-dependent oscillator strength and quantum efficiency of CdSe quantum dots controlled via the local density of states. Phys Rev B 79:045301 158. Dzhagan VM, Valakh MY, Raevskaya AE et al (2007) Resonant Raman scattering study of CdSe nanocrystals passivated with CdS and ZnS. Nanotechnology 18:285701 159. Raevskaya AE, Stroyuk AL, Kuchmiy SY et al (2007) Optical study of CdS- and ZnS-passivated CdSe nanocrystals in gelatin films. J Phys Condens Matter 19:386237 160. Dzhagan VM, Valakh MY, Raevskaya AE et al (2008) Size effects on Raman spectra of small CdSe nanoparticles in polymer films. Nanotechnology 19:305707 161. Stroyuk OL, Dzhagan VM, Shvalagin VV et al (2010) Size-dependent optical properties of colloidal ZnO nanoparticles charged by photoexcitation. J Phys Chem C 114:220–225. doi:10.1021/jp908879h 162. Raevskaya AE, Stroyuk OL, Kuchmiy SY et al (2006) Growth and spectroscopic characterization of CdSe nanoparticles synthesized from CdCl2 and Na2SeSO3 in aqueous gelatine solutions. Colloids Surf A 290:304–309. doi:10.1016/j.colsurfa.2006.05.038 163. Stroyuk OL, Shvalagin VV, Kotenko IE et al (2010) Photochemical reduction of sulfur in ethanol in the presence of ZnO nanoparticles. Theor Exp Chem 46:218–224. doi:10.1007/ s11237-010-9143-0 Chapter 2 Semiconductor-Based Photocatalytic Systems for the Solar-Light-Driven Water Splitting and Hydrogen Evolution

The research and development of new technologies for the conversion and storage of inexhaustible solar light energy were boosted several decades ago by the 1970th fuel crisis and a strategic need for sustainable power sources that can serve as alternatives to the fossil fuels. The basic idea was to accumulate the solar light energy as the electricity as well as to store it in the form of highly endothermic and eco-friendly fuels, in particular, molecular hydrogen produced by the photochemical splitting of water. Direct photochemical water splitting to gaseous hydrogen and oxygen can occur only under the illumination with highly energetic quanta at the wavelength k shorter than 240 nm [1]. However, such irradiation is completely absorbed by the atmosphere and does not reach the Earth surface. To overcome this obstacle, the water splitting is realized in the presence of photocatalysts—the substances capable of absorbing longer-wavelength light quanta (k > 300 nm) and inducing chemical transformations of water molecules. Inorganic semiconductors are probably the most broadly studied photocatalysts of water splitting. The semiconductor photocatalysts combine a high photosensi- tivity with a photochemical activity, stability, availability and relative simplicity of practical implementation. It should be noted that the photocatalytic and electro-photocatalytic (photoelectrochemical) processes with the participation of semiconductor nanomaterials are very similar by the nature and start with the same primary act of light quantum absorption resulting in the generation of an electron-hole couple. Differences between photocatalytic and photoelectrochemical/ photoelectrocatalytic processes arise mainly on the secondary steps of the charge carrier migration to the reaction participants. By this reason, both types of processes can be regarded as photocatalytic ones occuring in “usual” and electrochemical regimes and discussed together. Molecular hydrogen can be produced in photocatalytic systems of two types: (a) water splitting systems where stoichiometric amounts of H2 and O2 are produced simultaneously, and (b) systems with a so-called “sacriﬁcial” donor which is consumed irreversibly supplying electrons for the water reduction.

© Springer International Publishing AG 2018 39 O. Stroyuk, Solar Light Harvesting with Nanocrystalline Semiconductors, Lecture Notes in Chemistry 99, https://doi.org/10.1007/978-3-319-68879-4_2 40 2 Semiconductor-Based Photocatalytic Systems …

Stoichiometric (total) water splitting is accompanied by the energy accumulation and a free energy increment DG = 238 kJ/mole [2, 3]. Such process requires a semiconductor photocatalyst with a valence band (VB) potential more positive than the water oxidation potential (1.23 V vs. normal hydrogen electrode (NHE) at pH 0) and a conduction band (CB) potential more negative than the NHE potential (E = 0.0 V at pH 0). Therefore, a minimal light quantum energy required for the semiconductor-driven water splitting is 1.23 eV. Invariable losses accompanying interfacial charge transfers as well as over-voltages of the H2 and O2 formation increase this minimal energy to 1.7–1.9 eV [2, 3]. Therefore, the photocatalytic conversion of solar light energy should be the most favorable from the energetic viewpoints for semiconductors with a band gap (Eg) around 1.7–1.9 eV and a corresponding fundamental absorption band edge at kbe = 650–730 nm. The wider-band-gap semiconductors with kbe < 400 nm can also be used for the water splitting. However, due to a relatively small fraction of the UV light in the solar flux at the Earth surface, the conversion efficiency in such systems is typically not higher than 1–2%. Therefore, successful application of wide-band-gap semiconductors for the water splitting can be achieved only by expansion of their light sensitivity range to the visible domain of the spectrum. This effect can be achieved either by doping with metal/non-metal additives during the semiconductor synthesis or by various post-synthesis modifications. It should be noted that the semiconductors-based systems for the total water splitting have not yet showed reasonably high conversion efficiency as a result of a fast recombination of the oppositely charge photogenerated charge carriers as well as of primary intermediates—hydrogen atoms and hydroxyl radicals. A much higher conversion efficiency was achieved in the photo-catalytic systems with sacrificial donors. The range of sacrificial donors is very broad including inorganic sulfur compounds (H2S and alkali metal sulfides, sulfites, thiosulfates, thionates, etc.), hydrazine and aliphatic amines (triethylamine, triethanolamine (TEA), etc.), aliphatic alcohols (methanol, ethanol, 2-propanol), carboxylic acids (formic acid, ethylenediaminetetraacetic (EDTA) acid, etc.), carbohydrates and other organic substances, in particular those abundant in the broadly available and sustainable source—the fermented bio-mass. In the donor-based systems the photocatalytic process includes following typical stages: (i) excitation of a semiconductor photocatalyst by a light quantum with a proper (typically above-band-gap) energy, (ii) the interfacial transfer of a CB electron to an adsorbed water molecule followed by its reduction − • • (e +H2O ! H +OH), (iii) filling of a VB hole with an electron from a sacrificial donor (h+ + D ! D+•). This cycle requires the CB potential of a semiconductor photocatalyst to be more negative than the water reduction potential in given conditions and the VB potential—to be more positive than the oxidation potential of a sacrificial donor (or water molecules). Figure 2.1 provides a graphic review of band edge positions for a series of semiconductor materials relative to the standard potentials of water reduction and oxidation. The figure shows separately the semiconductors suitable (a) and unsuitable (b) for the evolution of the solar hydrogen from water. 2 Semiconductor-Based Photocatalytic Systems … 41

Fig. 2.1 CB and VB energy levels for some semiconducting photocatalysts with respect to NHE (ENHE) and vacuum (Evac). Reprinted with permission from Ref. [4]. Copyright (2015) The Royal Society of Chemistry

Typically, the semiconductor-based photocatalytic systems for the hydrogen production include aco-catalyst, that has no inherent photochemical activity but is capable of increasing dramatically the efﬁciency of semiconductor photocatalysts. Metal particles (Pt, Pd, Rh) deposited either on the semiconductor surface or on the surface of an inert carrier are typical co-catalysts for the semiconductor-based photocatalytic systems. The co-catalyst accepts and accumulates the charge carriers photogenerated in the semiconductor crystals inhibiting their recombination as well as contributes to a lowering of the water reduction overvoltage. In recent years the studies of new light energy conversion systems based on semiconductor photocatalysts and photoelectrodes have bloomed in leading research centers [5–34]. The research focused also on the photosynthetic microorganisms and other photoactive bio-systems capable of the molecular hydrogen evolution [35–37]. The present chapter obviously cannot encompass the whole variety of papers reporting on the photochemical water splitting. It aims mainly to highlight typical and most important directions of the recent research as well as to give the reader a notion of the current state of the area and its future development.

2.1 Photocatalytic Systems Based on the Wide-Band-Gap Semiconductors and Sensitizers

The wide-band-gap semiconductors, mostly metal oxides, belong to a large group of light-sensitive materials broadly studied as photocatalysts of the water reduction. The spectral sensitivity range of such materials can be expanded to longer wavelengths by combining them with dyes-sensitizers that absorb strongly UV and near IR light. 42 2 Semiconductor-Based Photocatalytic Systems …

Upon absorption of the visible and near IR light a sensitizer gets excited from the ground singlet state S0 into the first (or a higher) singlet excited state S1 (Sn). The S1 state can either return to S0 via emitting fluorescence or via the radiationless internal conversion. It can also convert into the first triplet excited state T1 or inject an electron into the conduction band (CB) of a semiconductor. After that, the water reduction occurs either on the semiconductor surface or (most often) on the surface of a metal co-catalysts (Fig. 2.2). The role of spectral sensitizers is typically played by organic dyes or metal complexes (Fig. 2.3). The basic operation principles and the state-of-the-art of the photocatalytic H2 evolution with the dye-sensitized semiconductors are comprehensively outlined in a recent review [38]. The most studied sensitized systems are based on titanium(IV) dioxide. For example, the hydrogen evolution under the illumination with the visible light (Vis-illumination) was observed in the presence of TiO2/Pt heterostructures modified by eosin [39, 40], derivatives of phenothiazine [41, 42], triphenylamine [43] and perylene [44], by various complexes of PtIV [45], ZnII [46] and NiII [47], copper phthalocyanine and ruthenium bipyridyl complexes [39]. Eosins adsorbed on the surface of Na2Ti2O4(OH)2 nanotubes (NTs) or MCM-41 zeolite modified by TiO2 nanoparticles (NPs) in the presence of the photodeposited Pt NPs act as spectral sensitizers of the hydrogen evolution from aqueous TEA solutions [48, 49]. A sensitization effect was also observed in a similar system based on eosin Y and N-doped TiO2 NPs [50]. Hydrogen generation from water/acetonitrile/КI occurs at the expense of I− oxidation under the Vis-illumination of the platinized titania and layered K4Nb6O17 sensitized by adsorbed coumarin and merocyanine dyes [51]. In the latter case, an effect of Pt NP localization on the photocatalyst activity was observed. The hydrogen formation rate over the K4Nb6O17/Pt composites with Pt NPs formed inside the interlayer space was found to be much higher than in similar systems where the metal NPs were distributed evenly between the inner and outer surface of the semiconductor or deposited only onto the outer semiconductor surface. The À effect is caused by a side reaction of I3 complex with the CB electrons. The eosin Y acts as a “universal” sensitizer for a series of layered wide-band-gap magnesium, calcium and strontium titanates [52]. The highest photocatalytic activity in the hydrogen evolution from aqueous diethanolamine solutions was

Fig. 2.2 Scheme of a photocatalytic system for the hydrogen evolution based on a TiO2/Pt heterostructure and a sensitizer (S). S0,S1,S+•—sensitizer in the ground state, excited state and oxidized state, respectively, D—sacriﬁcial donor 2.1 Photocatalytic Systems Based on the Wide-Band-Gap … 43

Fig. 2.3 Structure of some molecular sensitizers used in the semiconductor-based photocatalytic systems for hydrogen evolution

observed for SrTiO3 modiﬁed by 0.5 wt.% Pt. Co3O4 NPs sensitized by eosin Y showed a high activity in the water reduction under the Vis-illumination in the absence of any additional co-catalysts [53]. / / Adsorption of 1,1 -dinaphtyl-2,2 -diol on the surface of TiO2 NPs results in the formation of a charge-transfer complex with an intense absorption band centered at 550–600 nm. The photoexcitation of the complex into a charge-transfer absorption band leads to the hydrogen evolution from aqueous TEA solutions with a quantum yield (QY) of 0.02% [54]. The photocatalytic hydrogen evolution from aqueous glycerol solutions was observed for TiO2/Pt nanoheterostructures sensitized by inorganic tungsten-containing heteropolyacids [55, 56]. Molecular and metal complex dyes were successfully used to sensitize not only metal oxide photocatalysts but also semiconductors of other types, such as cadmium sulﬁde [57] and graphitic carbon nitride (g-C3N4, GCN) [58]. The Vis-illumination of aqueous GCN suspensions in the presence of eosin Y, TEA, and Pt NPs resulted in the hydrogen evolution with a QY of around 19% [58]. In similar photocatalytic systems, g-C3N4 was sensitized by erythrosin [59, 60] and copper phtalocyanine [61]. GCN sensitized by ZnII phthalocyanines revealed a comparatively high quantum yield of H2 evolution reaching 3.05% and a spectral sensitivity of up to 750 nm [62]. Starting from 1980th, various RuIII/II complexes with bipyridyl ligands were broadly studied as sensitizers of the hydrogen production and the studies in this direction are still advancing. For example, a photocatalytic system for the hydrogen 44 2 Semiconductor-Based Photocatalytic Systems …

2+ production comprising Ru tris-bipyridyl complexes, TiO2 NPs and hydrogenase as a co-catalyst was reported [63]. The hydrogen evolution under the Vis-illumination of aqueous solutions of sacrificial donors (methanol [64, 65]or TEA [66]) was observed in the presence of mesoporous TiO2 modified by Pt NPs and mono- and bidentate Ru2+ bipyridyl complexes. 2+ A strong electrostatic interaction between Ru(bpy)3 cation and the negatively charged surface of K4Nb6O17 nanoscrolls produced by the exfoliation of the bulk potassium niobate results in efficient electron phototransfer from the excited sensitizer to the semiconductor CB. The rate of photocatalytic hydrogen evolution from aqueous EDTA solutions is by an order of magnitude higher in the case of K4Nb6O17 nanoscrolls than for the bulk semiconductor [67]. The H2 evolution QY from EDTA solutions in the presence of H4Nb6O17 and HCa2Nb3O10 nanoscrolls 2+ / 2+ modified by platinum NPs and Ru(bpy)3 and Ru(bpy)2(4,4 -(PO3H2)2bpy) complexes reached 20–25% [68]. New sensitizers of titanium dioxide—binuclear RuIII complexes with separate fragments connected by an azobenzene “bridge” were reported in [69]. As opposite to “classical” sensitizers of such type that typically adsorb strongly on the semiconductor surface, the bonding between the sensitizer and the photocatalyst is weak in this case. The weak coupling allows for the photooxidized sensitizer to desorb from the semiconductor surface inhibiting a reverse electron transfer and accelerating the photocatalytic hydrogen evolution from aqueous solutions of methanol or TEA. A recent extensive review of the sensitized H2 evolution in the semiconductor- based systems [38] outlined principal challenges that still need to be met in this area. Most dyes have relatively narrow absorption bands, typically in the Vis range and an expansion of the light-harvesting range into the near IR is a vital challenge to be addressed. Some strategies aimed at resolving this problem include co-sensitization of semiconductor nanomaterials with combinations of dyes having complementary absorption spectra; fabrication of heterostructures with dyes, narrow-band-gap semiconductors, and conductive polymers; search for ligands capable of bonding to the semiconductor surface and forming intense ligand-to- metal charge transfer absorption bands, etc. The second challenge lies in a typically low stability of the molecular sensitizers. The organic dyes suffer from the photodegradation as a result of alternative reactions involving the singlet and triplet excited dyes, while the metal complexes are prone to photoinduced ligand exchange and photosolvation reactions resulting in the deterioration of their light-harvesting ability. Attempts of abating this problem include a proper modification of the semiconductor surface to mitigate secondary reactions as well as a rational design of the dye structure to reduce the possibility of the excited state relaxation pathways competing with the charge injection. In recent years, a new research direction formed focusing on the visible-light-induced photocatalytic activity of heterostructures of wide-bandgap semiconductors with noble metal NPs, the latter exhibiting a surface plasmon resonance in the visible spectral range. This effect was christened as “plasmonic photocatalysis” [16, 70, 71] and was first accepted sceptically, but a number of reports on various photocatalytic transformations and photoelectrochemical 2.1 Photocatalytic Systems Based on the Wide-Band-Gap … 45 processes that can be performed by illuminating the semiconductor/metal NPs with the visible light was growing steadily, showing good perspectives of this phenomenon for the solar light harvesting [16, 34, 70–73]. The NPs of noble metals—gold, and silver reveal intense absorption bands in the visible spectral range as a result of electron gas oscillations in a surface layer of the metal NPs that is referred to as surface plasmon resonance (SPR). The SPR effect can be observed only for NPs (roughly smaller than 100 nm) and not for the corresponding bulk metals. The spectral parameters of SPR absorption band depend on the metal type, NP size and shape, dielectric parameters of the dispersive medium (solvent), nature of species adsorbed on the NP surface, on the proximity of neighboring metal NPs and many other factors [70–72]. For spherical non-aggregated silver and gold NPs the SPR maxima can be found around 390–400 and 530–550 nm, respectively. The SPR absorption of gold NPs, though being quite intense and fitting to the solar spectrum, does not result in an interband electron transition and generation of additional free charge carriers, as it happens at the above-bandgap photoexcitation of semiconductors. Therefore, the Au NPs cannot act similarly to conventional molecular spectral sensitizers that inject an electron into the wide-bandgap semiconductor after the photoexcitation. The fact fed the skepticism concerning the reality of the “plasmon photocatalysis” phenomenon when it was only emerging in the field of solar light harvesting. Meanwhile, more and more reports on the photocatalytic transformations occurring under excitation into the SPR band of various gold/semiconductor heterostructures were steadily accumulated, some reports providing photoaction spectra (dependences of the QY of a photoreaction on the excitation wavelength) coinciding with the absorption spectra of Au NPs [74–79]. In attempts to interpret these processes, several alternative mechanisms were proposed including the heat transfer from Au NPs to the semiconductor resulting in the interband electron transition, ionization of the surface states of semiconductor NPs under the influence of the electromagnetic field of SPR-excited Au NPs, and others. However, a number of recently reported scrupulous and sophisticated studies showed that Au NPs excited into the SPR band can indeed inject “hot” electrons into the CB of wide-bandgap semiconductors, such as titania, in the cases when the Fermi level of photoexcited metal NPs shifts higher than the Schottky barrier on the semiconductor-metal interface (Fig. 2.4)[13, 34, 70–73]. For the plasmonic NPs smaller than 20 nm the hot electrons exhibit a broad spectrum of energies falling within the range from EF,M to EF,M + hv, while larger particles exhibit much smaller hot electron energies close to EF,M and therefore for the larger metal NPs the probability of the hot electron injection is much lower. The electrons with an energy lower than the Schottky barrier relax through the electron-electron and electron-phonon interactions. After the hot electron injection, a metal NP recompenses via a hole transfer to a water molecule (resulting in the O2 evolution) or to another sacrificial donor, similarly as it happens with the photoexcited molecules of dye sensitizers or the photoexcited semiconductor NPs. The hot electron injection probability depends also on the distance to the semiconductor surface that should be covered by a hot electron before the internal relaxation 46 2 Semiconductor-Based Photocatalytic Systems …

Fig. 2.4 Plasmonic energy conversion: electrons from occupied energy levels are excited above the Fermi energy. Hot electrons with energies high enough to overcome the Schottky barrier uSB = uM − vS are injected into the conduction band Ec of the neighboring semiconductor, where uM is the work function of the metal and vS is the electron afﬁnity of the semiconductor. DOS is the density of states, EF,M and EF,S–Fermi level of the metal and metal/semiconductor heterojunction, Ev—valence band of semiconductor. Reprinted with permissions from Ref. [72]. Copyright (2014) Macmillan Publishers Limited occurs, as well as on the density of states on the semiconductor surface that can accommodate a hot electron [71]. In this chapter, the effect of plasmonic light absorption in the semiconductor-based photocatalytic for the hydrogen evolution will be discussed only concisely. A series of recent reviews covers the issue of plasmonic photocatalysis much more extensively and can serve as a perfect guide for further development of this area [13, 16, 34, 70–73, 80]. The most popular plasmonic photocatalyst for hydrogen production is probably a TiO2/Au combination. The SPR-enhanced H2 evolution under illumination with the visible light (typically with k > 420–450 nm) was observed in the presence of nanocrystalline TiO2/Au heterostructures [74, 81–84], N-doped TiO2 decorated with Au NPs [78], mesoporous TiO2/Au composites [75] and aerogels [77], porous ﬂat TiO2/Au electrodes [85], TiO2/Au photonic crystals [86]. Mixed Au/Pt NPs deposited onto the surface of TiO2 nanosheets can play a double role, the gold providing SPR for the visible light harvesting, while Pt acting as a co-catalyst of hydrogen evolution [87]. The photoaction spectrum of TiO2/Au composite as a photocatalyst of H2 evolution was found to be very similar to the absorption spectrum (Fig. 2.5) indicating unambiguously on the participation of SPR-excited gold NPs in the photochemical transformations. Direct participation of Au NPs in the photocatalytic reaction was clearly demonstrated for a mesoporous TiO2/Au heterostructure evolving hydrogen from aqueous solutions of ascorbic acid when excited into narrow spectral windows of 500 ± 20 and 550 ± 20 nm [75]. No H2 was detected in such conditions for the pure titania. It is notable that the excitation into the 500 ± 20 nm window results in a higher rate of hydrogen evolution because the energy of hot electrons depends on 2.1 Photocatalytic Systems Based on the Wide-Band-Gap … 47

Fig. 2.5 Absorption and photoaction spectra of TiO2 Evonik P25 and a P25/Au heterostructure. Reprinted with permissions from Ref. [74]. Copyright (2016) American Chemical Society

the excitation energy and the probability of injection is higher for the shorter-wavelength light. The effect of SPR-induced enhancement of the photocatalytic/ photoelectrochemical H2 evolution is of general nature and can be observed for other photoactive semiconductors, such as nanocrystalline CdS [88] and Ta2O5/ Ta3N5 [89], ZnO nanorods (NRs) [76, 90], La2Ti2O7 nanosheets [77]. The CdS/Au heterostructures exhibited not only an enhanced activity in the photocatalytic water reduction but also a much higher photostability in aqueous Na2S/Na2SO3 solutions as compared to the individual CdS [88]. A spectacular plasmon enhancement of the photocatalytic/photoelectrochemical H2 evolution was also observed for branched ZnO nanowires (NWs) decorated with gold NPs [76]. The deposition of Au NPs onto a highly developed surface of branched ZnO NWs resulted in a much broader spectral response extending to 700–750 nm. The incident-photon-to-current-efﬁciency (IPCE) spectra (analogs of photoaction spectra) of ZnO and ZnO/Au NWs excited by UV light (Fig. 2.6, panel 1) are roughly the same revealing no appreciable spectral differences and corresponding to the direct interband electron excitation of the semiconductor photocatalyst. However, the ZnO/Au heterostructures, as opposite to bare ZnO NWs, revealed a spectral response in the visible range with the band shape mimicking closely the SPR band shape of gold NPs (Fig. 2.6, panel 2). Recently, the family of “plasmonic” photocatalysts was joined by GCN/Au nanoheterostructures. Graphitic carbon nitride absorbs only a limited portion of the visible light up to 460–470 nm and can be sensitized to longer-wavelength irradiation by the deposition of Au NPs [91, 92]. Similarly to gold, Ag NPs exhibit an intense SPR band in the visible spectral range and can induce the effect of spectral sensitization when excited into the SPR band, however, in this case the sensitization effect is not so obvious, as for gold, because the SPR band maximum of Ag NPs is closer or even overlapped with the absorption spectra of the most photoactive semiconductors. The effect of 48 2 Semiconductor-Based Photocatalytic Systems …

Fig. 2.6 IPCE spectra of P-ZnO, B-ZnO, Au/P-ZnO, and Au/B-ZnO NW photoanodes collected at 1.23 V versus NHE in a wavelength window of 300–420 nm (panel 1) and 420–850 nm (panel 2). Reprinted and adapted with permissions from Ref. [76]. Copyright (2014) American Chemical Society

plasmon-enhanced H2 evolution was reported for N-doped TiO2/Ag heterostructures [78], ZnO/Ag [93], GCN/Ag [94]. The ZnO NRs decorated with triangular Ag nanoprisms revealed a higher plasmon-activated photoactivity as compared with similar heterostructures based on regular spherical Ag NPs as a result of a strong electromagnetic ﬁeld generated on the prism edges [94].

2.2 Photocatalytic Systems Based on the Binary and More Complex Semiconductor Heterostructures

Absorption of the visible light by a narrow-band-gap component of binary semiconductor composites also results in the electron injection to the CB of a wide-band-gap component, where, with the participation of a co-catalyst, hydrogen formation occurs. The photogenerated hole remains separated from the electron and reacts with a donor. Such spatial separation of the charge carriers is a reason for typically high rates of the photocatalytic hydrogen evolution over binary hetero-structures composed of narrow-band-gap metal sulfides and wide-band-gap metal oxides [95–100]. Figure 2.7 shows a scheme of charge transfers in a 2.2 Photocatalytic Systems Based on the Binary … 49 photocatalytic system based on a very popular TiO2/CdS composite. In the further discussion we will define binary and more complex heterostructures by listing their components one after other separated by a “/” symbol. Typically we will put to the left of the slash a “basic” component of the heterostructure, for example, a wide-bandgap semiconductor (TiO2) onto which another component, such as a narrow-bandgap sensitizer (CdS) is deposited or attached. One of the most broadly studied semiconductor sensitizers for the hydrogen evolution is cadmium sulfide as well as related solid solutions, such as cadmium zinc sulfide. For example, Vis-sensitive photocatalysts of the hydrogen evolution from aqueous solutions of 2-propanol or Na2S–Na2SO3 were formed by the deposition of CdS NPs on the surface of nanocrystalline titania [101, 102]. The photoactivity of the heterostructures increases remarkably with a decrease of the CdS NP size as a result of a size-dependent increase of the CB energy of CdS NPs [103, 104]. The photocatalytic activity of such systems can be further boosted by modification with fullerenes acting as photoelectron acceptors [105]. Ternary TiO2/CdS/Pt heterostructures can be used for the photocatalytic H2 evolution directly from the sea water after addition of sacrificial donors (Na2S and Na2SO3)[106]. An important factor governing the photocatalytic properties of ternary TiO2/CdS/Pt composites in the water reduction is a “correct” spatial organization of components [107–109]. A photocatalyst produced by the Pt NP photodeposition on the surface of preliminarily formed binary TiO2/CdS heterostructure showed by an order of magnitude lower photoactivity than similar composites prepared by the CdS NP deposition onto pre-formed TiO2/Pt heterostructure [107]. The same photocatalytic behavior is typical for a broad range of ternary TiO2/CdS/M composites, where M = Au, Ag, Pd, Pt [110]. TiO2/CdS/Pt heterostructures produced by the impregnation of TiO2/CdS composites with chloroplatinic acid followed by the thermal Pt(IV) reduction exhibited a higher photoactivity in the H2 evolution than similar composites produced via the photocatalytic Pt(IV) reduction [108]. In this case, the difference in photoactivity also owes to the fact that the thermally deposited Pt NPs are attached mostly to the

Fig. 2.7 Scheme of spatial separation of the photogenerated charge carriers in a CdS/TiO2 heterostructure and the H2 formation under the Vis-illumination 50 2 Semiconductor-Based Photocatalytic Systems …

TiO2 surface, where the water reduction takes place, while the photo-deposited metal NPs are distributed randomly between the CdS and TiO2 NPs. Ternary WO3/CdS/Au heterostructures built on the basis of inverted WO3 opals are more active photocatalysts of the water splitting than their analogs produced from randomly structured tungsten oxide. An advanced photoactivity of the opal-based photocatalysts stems from a more efficient light absorption due to the multiple scattering and refraction of light in the regular pores of the opals [111]. A shape anisotropy of zinc oxide NRs [112] and nanobelts [113] favors to the spatial charge carrier separation in ZnO/CdS heterostructures reflecting in a high photocatalytic activity in the H2 evolution from water/methanol mixtures. The ion exchange capability of a Ti(IV)-modified MCM-41 zeolite was used to form 2.5-nm CdS NPs in the zeolite pores [114]. After the Pt NP photodeposition such heterostructure exhibits a high photocatalytic activity in the hydrogen evolution from aqueous sodium sulfite solutions exceeding strongly that of bulk cadmium sulfide. The photocatalytic H2 evolution from aqueous TEA solutions was also observed in the presence of CdS NPs immobilized on MCM-41 with a fraction of Si atoms replaced with Zr and Ti [115]. The heterostructures of CdS NPs [116–118] and Cd0.5Zn0.5S NPs [119] with TiO2 NTs are efficient Vis-sensitive photocatalysts of the hydrogen evolution from aqueous Na2S/Na2SO3 solutions. The CdS NP deposition the surface of TiO2 nanoplates [120] and meso-porous microspheres [121] with prevailingly exposed {001} facets yields efficient photocatalysts of the water reduction by lactic acid [120]. The photoactivity of such heterostructures exceeds that of similar composites produced from conventional titania crystals because the {001} lattice face of titania exhibits a relatively higher efficiency of the interfacial electron transfer [120, 121]. Spatial separation of the photogenerated charge carriers between the host titanosilicate matrices ETS-4 and ETS-10 comprising ultra-thin (–O–Ti–O–Ti–O–)x “quantum wires” and CdS NPs deposited into the host pores results in a high photoactivity of such heterostructures in the H2 evolution from aqueous Na2S/Na2SO3 solutions [122]. Similar approaches were used to introduce CdS NPs into the interlayer galleries of layered titanates [123–127], niobates [128–130] and tantalates [129, 131, 132], as well as layered mixed ZnII and CrIII hydroxides [133]. In such composites, the water reduction to H2 occurs on co-catalyst NPs (Pt, Ni or RuO2) deposited on the outer photocatalyst surface, while the oxidation of a sacrificial donor (Na2Sor Na2SO3) involves CdS NPs attached to the inner surface of the layered host material. Due to the spatial separation of the charge carriers the photocatalytic activity of the composites exceeds strongly that of individual cadmium sulfide or a mechanical mixture of CdS and a layered metallate [123, 124, 128]. To achieve favorable conditions for the formation of CdS NP-based heterostructures and to promote photocatalytic processes with their participation, a preliminary treatment of layered host materials is often performed aimed at an expansion of the interlayer galleries. For example, the intercalation of propylamine and [Pt(NH3)4]Cl2 complex into the interlayer space of HNbWO6 expands considerably the inner voids between the layers favoring to the secondary intercalation with CdII and ZnII [134]. The annealing and sulfurization of such material resulted 2.2 Photocatalytic Systems Based on the Binary … 51 in a HNbWO6/Cd0.8Zn0.2S/Pt heterostructure exhibiting a Vis-light-driven photocatalytic activity in the hydrogen evolution from aqueous solutions of sodium sulfite. A treatment of co-deposited CdS and TiO2 NPs with titanium(IV) chloride followed by he annealing [135] assures the formation of TiO2/CdS heterostructures with a good mechanical and electronic contact between the CdS and TiO2 NPs favoring to the charge transfers between the components. The highest photocurrent and photocatalytic activity in the hydrogen generation were observed at 80 wt.% titania content [135]. Directed migration of the photogenerated charge carriers—from a layer of cadmium selenide to TiO2 NTs through an intermediary CdS layer in ternary TiO2/ CdS/CdSe heterostructures contributes to their high photoelectrochemical activity in the hydrogen evolution from aqueous solutions of Na2S/Na2SO3 or ethylene glycol with QY reaching *9.5% [136]. A similar effect was observed for nanoheterostructures formed by CdS “nanoflowers” grown on the surface of TiO2 NT arrays (Fig. 2.8a) [137]. A very efficient charge transport from the visible-light-sensitive CdSe NPs to the thin (*5 nm thick) titania NSs results in a strong non-additive enhancement of the photocatalytic hydrogen evolution from aqueous Na2S/Na2SO3 solutions [138]. Coupling of the TiO2 NSs to CdS NPs via a molecular bridge—bifunctional mercaptopropionic acid (MPA) anion allows to double the H2 evolution efficiency as compared to the bare CdSe NPs, while direct (without linkers) deposition of the sensitizer NPs onto the TiO2 NSs increases the efficiency by another *100% (Fig. 2.8b). An electron paramagnetic resonance (EPR) study showed that Ti4+ ions can be converted into Ti3+ by the photogenerated CB electrons and act as charge transfer mediators to the CdSe NPs. Due to the fact, the annealing of TiO2 NSs that

Fig. 2.8 a A scheme of the photoelectrochemical H2 evolution with “TiO2 nanotube/CdS nanoflower” heterostructures; b The rate of photocatalytic hydrogen evolution in the presence of TiO2 nanosheets and NS/CdSe heterostructures. Reprinted with permissions from Ref. [137] (a) and [138](b). Copyright (2015, 2016) Elsevier (a) and American Chemical Society (b) 52 2 Semiconductor-Based Photocatalytic Systems … caused their aggregation and a loss of the surface area had a detrimental effect on the photocatalytic activity of both bare TiO2 NSs and TiO2 NS/CdSe nanocomposites [138]. The sol-gel deposition of 10–20-nm titania NPs on the surface of microcrystalline cadmium sulfide followed by the photodeposition of Pt NPs results in a ternary CdS/TiO2/Pt composite that reveals photocatalytic properties in the hydrogen evolution from aqueous Na2S/Na2SO3 solutions [139]. Other platinum group metals can also act as co-catalysts of this process forming the following activity sequence: Pt > Rh > Pd > Ru. Isotopic studies in a similar system, where H2S was used as a sacrificial electron donor, showed that H2 is evolved at the expense of the decomposition of both H2O and H2S[140]. CdS/TiO2 heterostructures based on cadmium sulfide NWs [141] exhibited a much higher photocatalytic activity in the H2 evolution from aqueous Na2S/Na2SO3 solutions than non-modified CdS NWs. Spatial separation of the photogenerated charge carriers between the heterostructure components results in the separation of oxidative and reductive steps of the process—the water reduction to H2 occurs on the TiO2 NPs, while the sacrificial donors are oxidized on the surface of CdS NWs. Despite the fact that sacrificial donors, especially sodium sulfide and sulfite can efficiently quench the oxidative photocorrosion of cadmium sulfide, some inevitable release of inherently toxic CdII ions can be expected for the CdS-based photocatalysts. This hazard stimulates a constant search for other less toxic narrow-bandgap sensitizers capable of competing with cadmium sulfide in the hydrogen evolution efficiency. A particular attention in this search is paid to ternary and quaternary metal-chalcogenide NPs, such as indium-based chalcopyrite CuInS2 and AgInS2 (AgIn5S8) NPs and quaternary kesterite Cu2ZnSnS4 NPs. These compounds have relatively narrow bandgaps of around 1.4–1.8 eV and reveal strong absorption bands covering the entire visible spectral range thus making such NPs ideal light harvesters for the photocatalytic hydrogen evolution systems. The CuInS2/TiO2 [142, 143] and TiO2/AgIn5S8 [144] heterostructures revealed a photocatalytic activity under the photoexcitation over almost the whole visible spectral range. The sensitization of Ag NP-decorated ZnO NW arrays with CuInS2 NPs results in *100-fold enhancement of the photoelectrochemical hydrogen production efficiency under the Vis-illumination as compared to the original NWs [145]. The quaternary kesterite NPs were successfully employed as a light harvester for the photoelectrochemical hydrogen production over a ZnO/CdS/Cu2ZnSnS4 heterostructure based on ZnO NWs [146]. The mutual positions of the CB and VB levels of the components are ideally suitable for a cascade transfer of the photogenerated electrons from the outer kesterite layer to the CdS buffer layer to the ZnO NW layer (Fig. 2.9a). After the cascade the electrons are collected into the electric circuit and transferred to a Pt counter electrode, where the H2 evolution occurs, while the CdS/Cu2ZnSnS4 (CZTS) light-harvesting layer is regenerated via the oxidation of a sacrificial donor (Na2S/Na2SO3)[146]. The photocurrent (and correspondingly, H2 on the counter electrode) is generated under the illumination in the 2.2 Photocatalytic Systems Based on the Binary … 53

Fig. 2.9 a A scheme of charge transfers in ZnO NW/CdS/Cu2ZnSnS4 (CZTS) system; b IPCE spectra of ZnO NW-based heterostructures with CdS and CZTS NPs. Reprinted with permissions from Ref. [146]. Copyright (2015) The Royal Society of Chemistry entire visible range (400–700 nm) with the light-to-current conversion efficiency reaching *45% (Fig. 2.9b). Quaternary NPs of other types, such as Cu–Ga–In–S NPs [147], are also currently probed as spectral sensitizers with the aim of combining a high absorptivity in the visible spectral range and a “suitable” band positions for the efficient charge transfer to TiO2. Among the binary non-toxic semiconductor sensitizers, a special attention is focused on bismuth and antimony chalcogenides that combine a high sensitivity to the visible light, a relative stability and band positions favorable for the charge injection into TiO2, ZnO, and other wide-bandgap semiconductor materials. Thermal hydrolysis of thiourea in the presence of Bi(NO3)3 and nanocrystalline TiO2 yields TiO2/Bi2S3 heterostructures manifesting a photocatalytic activity in the Vis-light-driven H2 production from aqueous Na2S/Na2SO3 solutions [148]. The photoactivity of the composite was found to be much higher than that of bismuth sulfide alone and maximal—at the equimolar content of the components [148]. The TiO2/Bi2S3 composites produced by a solvothermal method from 10 to 15-nm titania NPs exhibited photocatalytic properties in the hydrogen evolution from water/methanol mixtures [149]. Spatial separation of negative and positive charge carriers in the nanoheterostructures of titania and copper(I,II) oxides as well as the capability of copper oxides of accumulating electrons and decreasing the water reduction over- voltage allowed to carry out the photocatalytic H2 evolution under the illumination with the visible light [138, 150–160]. A p/n heterojunction also forms on the interface between TiO2 and copper phosphide Cu3P NPs enabling efficient separation of the photogenerated charge carriers and the water reduction with an apparent QY (measured at a certain wavelength) of 4.6%, which is by an order of magnitude higher than for sole titania NPs [161]. 54 2 Semiconductor-Based Photocatalytic Systems …

The photocatalytic Vis-light-driven formation of hydrogen was observed also in the presence of In2O3/In2S3 [162], CuO/ZnO [163], In1−xGaxN/ZnO [164], CuFeO2/ SnO2 [165], RuO2/TiO2 [166], and CuAlO2/TiO2 [167] nanoheterostructures. Along with the development of photocatalytic systems based on traditional semiconductors, a search is also performed for new photosensitive semiconducting materials combining the visible light sensitivity with a capacity to act as spectral sensitizers for wide-band-gap semiconductors. At that, a special attention is paid to carbon materials—fullerenes, carbon NTs, etc. For example, a composite of multiwall carbon NTs with titania modified by Ni NPs exhibited a photocatalytic activity in the water reduction when excited by the visible light [113, 168]. It was assumed that the photoexcitation of carbon NTs results in the electron injection into the TiO2 CB followed by the electron transfer to the Ni NPs where the final act of the water reduction occurs. The oxidized NTs are then regenerated at the expense of methanol oxidation. Recently, good perspectives were shown for the sensitization of wide-bandgap semiconductor materials with carbonaceous nanostructured species, such as carbon NPs and nanodispersed carbon nitride. The carbon NPs can be produced by thermal/electrochemical decomposition of a variety of organic precursors and contain a partially aromatic carbon core and an outer shell abundant with various functional groups [169, 170]. They absorb light in broad and intense bands extending throughout the visible range and can strongly bind to the most of the photoactive wide-bandgap semiconductors typically used for the photocatalytic processes. A comprehensive account of recent successes and challenges associated with the utilization of carbon NPs in the photocatalysis can be found in [169]. For example, the nanocrystalline titania can be sensitized to the visible light by the carbon NPs [171, 172] produced via hydrothermal treatment of vitamin C [171]or by the electrochemical destruction of graphite [172, 173]. Such heterostructures exhibited almost by an order of magnitude higher photocatalytic activity in the H2 evolution from water/methanol mixtures under the illumination with the “white” light (k > 400 nm) as compared to the bare titania. Graphitic carbon nitride is often called a “rising star” of the semiconductor photocatalysis as it combines a unique set of properties including chemical stability, sensitivity to the visible light, “appropriate” positions of CB and VB energies allowing both for the water reduction and oxidation to occur simultaneously. This material will be discussed in details later in the section devoted to new photoactive materials. Here, we only mention the role of GCN as a component of composite H2 evolution photocatalysts. It was found that spatial separation of the photogenerated charg carriers imparts TiO2/GCN heterostructures with a photocatalytic activity in the Vis-light-driven hydrogen evolution from water with no sacrificial donors [174– 176] as well as from aqueous solutions of methanol [177] or TEA [178]. The exfoliation of GCN into a-few-layer or even single-layer CN sheets increases strongly its activity as a hydrogen evolution photocatalyst, both in the individual state and when incorporated into complex nanoheterostructures. It is reported that the photocatalytic activity of composites of titania NRs with the GCN 2.2 Photocatalytic Systems Based on the Binary … 55 nanosheets produced by an ultrasound treatment of the bulk GCN is by far higher than the photoactivity of a mixture of TiO2 NRs and the unexfoliated GCN [179]. As GCN has a bulk bandgap of 2.7 eV it can also be a subject to the spectral sensitization. Such effect was achieved for CdS/GCN [180] and ZnIn2S4/GCN [181] composites, as well as for CdSe NP-decorated hollow GCN spheres [182]. The GCN NSs can be used as a “mat” to accommodate wide-bandgap semiconductor NPs. For example, GCN/TiO2 heterostructures produced by the solvothermal deposition of titania NPs onto GCN NSs demonstrated the rates of photocatalytic hydrogen evolution by *10 and *20 times higher than those observed in the presence of sole TiO2 and the bulk GCN [183]. A similar effect was also achieved for *20-nm InVO4 nanocrystals grown on the GCN sheets [184].

2.3 Photocatalytic Systems Based on the Metal-Doped Wide-Band-Gap Semiconductors

Doping of the wide-band-gap semiconductors with metal ions introduces new occupied local states in the band gap that can be excited by the visible light and supply electrons to CB (Fig. 2.10). The CB electrons participate then in the water reduction while the holes localized on the dopant states get filled by electrons from a sacrificial donor or water [137, 185–189]. Visible-light-sensitive photocatalysts of the hydrogen evolution from aqueous solutions of sodium sulfite were prepared by doping ZnS with PbII [190], NiII [191] and CuII [192]. Such photocatalysts can function without additional co-catalysts. The visible-light absorption by these compounds originates from the photoinduced electron transition from the local dopant states in the CB of zinc sulfide. The photocatalytic activity of ZnS:PbII is maximal at 1.4 wt.% lead content and decreases considerably at a higher dopant concentration (more than 2%) as a result of the formation of a separate PbS phase. Additional co-doping of the photocatalyst with halogen anions results in a 3-fold increase of the photoactivity, the effect originating from a relaxation of the lattice strain and a decrease of the number of non-radiative recombination sites [192]. At an optimal dopant concentration of 4.3 wt.% the

Fig. 2.10 A photocatalytic system for the hydrogen production based on NiII- doped titania 56 2 Semiconductor-Based Photocatalytic Systems …

ZnS:CuII-based photocatalytic system exhibits an apparent hydrogen production QY of 3.7% [192]. A maximal increment of the photocatalytic H2 evolution rate after doping of zinc sulﬁde with Cu(II)—by a factor of 11 is observed after the introduction of *5 mol% copper [193]. At the same time, for ZnS:NiII the peak photocatalytic activity in the water reduction was achieved already at 0.1 mol% Ni content [191]. Doping of titania with BiIII imparts this semiconductor with a photocatalytic activity in the H2 evolution from water/ethanol mixture under the Vis-illumination [194]. In a similar way, doping of SrTiO3 with Cr [195, 196] and Rh ions [197, 198] yields Vis-light-sensitive photocatalysts of the hydrogen evolution from aqueous methanol [195, 196] and pure water splitting [197, 198]. The substitution of Ti4+ with Cr3+ or Fe3+ in titania crystals requires a compensation of the excessive negative lattice charge and induces self-oxidation of CrIII to CrVI and the release of molecular oxygen [199–202]. The recombination of charge carriers at anion vacancies forming after the O2 subtraction decreases the photocatalytic activity. To balance the charge and to increase the stability and activity of Cr-doped titania an equimolar amount of Ta5+ or Nb5+ should additionally be introduced into the lattice. An increase of the photocatalytic activity of a doped semiconductor as a result of the charge compensation was also observed 2+ 5+ for the co-doping of TiO2 and SrTiO3 with combinations of Ni /Ta [199] and Cr3+/Sb5+ [200]. A strong doping-induced photoactivity enhancement of a semiconductor host in 3+ the water reduction was observed after the introduction of Bi into NaTaO3 [203], 2+ + Zn into SrTiO3 and BaTiO3 [204], Ag into BiVO4 [205], cations of Y, La, Ce or 2+ Yb into NaTaO3 [206], and Zn into Ga2O3 [207]. The photocatalytic water reduction to H2 under Vis illumination was reported for 3+ 2+ ZnS and SrTiO3 doped with La [208], Ni -doped InTaO4 and InNbO4 [209]. After the deposition of a co-catalyst (Pt, RuO2, NiOx) the latter two systems demonstrated an apparent QY of up to 0.66% (at k = 400 nm). Besides doping with 2+ Ni , InTaO4 can be turned into a Vis-sensitive photocatalyst of the H2 evolution by introducing Mn, Fe, Co, and Cu cations [210]. Doping with chromium turns a Ba2In2O5/In2O3 heterostructure into a “universal” photocatalyst capable of the hydrogen evolution from water and water/methanol mixtures in the presence of Pt or Ni as well as of the O2 evolution from aqueous AgNO3 (electron acceptor) solutions [211]. A broad range of dopants—CrVI,FeIII,CoII,NiII,RuII, and PdII was used to convert nanocrystalline Bi2O3 (Eg = 2.8 eV) into a Vis-sensitive photocatalyst of the water reduction [212]. Doping with palladium(II) resulted in the best characteristics, the fact apparently originating from in situ Pd(II) photoreduction Pd0 which can act as a co-catalyst. Almost in each system based on doped semiconductors there exists an optimal dopant concentration range where the maximal photoactivity is observed, while a higher dopant amount deteriorates the semiconductor activity in the water reduc- 2+ 3+ tion. For TiO2 doped with Ni [213, 214]orBi [194], the maximal rate of the photocatalytic hydrogen evolution from water/alcohol mixtures was observed at a 2.3 Photocatalytic Systems Based on the Metal-Doped … 57

1% mol. dopant content [213]. This effect is typically associated with a hindrance of the free migration of photogenerated charge carriers in the bulk of highly-doped semiconductor crystals because of an abundance of the local dopant states acting as charge traps. In some cases doping results in a fusion of the dopant states and a “top” of the host VB. The effect narrows the band gap (and increases the Vis-light sensitivity of a semiconductor) without the emergence of additional local states in the forbidden band. For example, doping of indium titanate with a mixture of nickel and chromium cations results in a fusion of Ni3d, Cr3d, Ti3d/In5sp, and O2p orbitals yielding a Vis-light-sensitive In12NiCr2Ti10O42 photocatalyst (Eg = 2.14 eV) of the H2 evolution from water/methanol mixtures [215, 216], more efficient than “mono-doped” In6NiTi6O22 (Eg = 2.48 eV) and In3CrTi2O10 (Eg = 2.00 eV) [215]. Sometimes a variation in the metal dopant nature allows switching the semiconductor activity between the water reduction and the water oxidation. For 2+ 3+ example, doping of SrTiO3 with Mn or Ru impart this semiconductor with a photocatalytic activity for the Vis-light-driven oxygen evolution from aqueous AgNO3 solutions [217]. At the same time, doping of strontium titanate with Ru, Rh, or Ir (1 wt.%) cations and the deposition of 0.1 wt.% Pt converts this wide-band-gap semiconductor into a Vis-light-sensitive photocatalyst of the water reduction by methanol demonstrating an apparent H2 evolution QY of 5.2% at 420 nm [217]. The radio-frequency magnetron sputtering technique is typically used [218–224] to produce titania films that exhibit a visible light sensitivity originating from a stoichiometry deviation, that is a gradient of the O/Ti atomic ratio from the surface to the bulk of the films. A post-synthesis hydrothermal treatment of the films enhances considerably their photocatalytic activity in the water reduction as a result of an increase of the film crystallinity and the specific surface area [222]. This method was also applied to produce Ti foils decorated with titania nano-columns oriented normally to the film surface [225] with an O/Ti ratio changing from 2.00 on the column top to 1.93 at the site of the column contact with the substrate. After the modification of an opposite side of a Ti/TiO2 foil with Pt NPs, it was used as a Vis-light-sensitive photocatalyst of the total water splitting in a combined reactor with membrane-separated compartments for the water reduction to H2 and the water oxidation to O2 (Fig. 2.11a) [218, 225]. Such design allows avoiding the recombination between primary products of the reduction (H atoms) and the oxidation (OH radicals), which is one of the main factors limiting the H2 evolution efficiency. Recently, a so-called “black” titania emerged as a new visible-light-sensitive photocatalyst of the water reduction [29]. The “black” TiO2 is typically produced by treating titania with hydrogen or aluminium resulting in a massive reduction of Ti4+ to Ti3+, the latter imparting the material with a characteristic blackish-gray to black color (Fig. 2.11b, insert). According to [226], reduction with Al yields much deeper reduced TiO2−x samples with the absorbance extending to longer wavelengths as compared to the hydrogen-processed titania (HP-TiO2), the absorption band encompassing the entire visible and near IR ranges (Fig. 2.11b). A strong light-harvesting capability of the “black” TiO2 results in much higher photocurrents/H2 evolution rates in the 58 2 Semiconductor-Based Photocatalytic Systems …

Fig. 2.11 a A combined photocatalytic reactor for simultaneous evolution of H2 and O2 from water in the presence of a “Pt/Ti foil/Ti NTs” heterostructure; b absorption spectra of original TiO2 and products of titania reduction with hydrogen (HP-TiO2) and aluminium (TiO2−x). Yellow background—solar AM1.5 irradiation spectrum. Insert: photographs of conventional nanocrystalline titania (Evonik P25) and black TiO2−x produced by the reduction with Al. Reprinted with permissions from Refs. [222](a) and [226](b). Copyright (2008, 2015) Elsevier (a) and American Chemical Society (b) photoelectrochemical/photocatalytic systems as compared to those with conventional nanocrystalline TiO2 powders [226], NTs [227] or mesoporous TiO2 [228].

2.4 Photocatalytic Systems Based on the Nonmetal-Doped Wide-Band-Gap Semiconductors

A partial oxygen substitution in a metal oxide semiconductor lattice by other non-metals—nitrogen, carbon, sulfur, etc., was found to be one of the most versatile methods of tailoring the band gap of semiconductor photocatalysts. The p-orbitals of a dopant typically have a higher energy than the p-orbitals of oxygen, so the dopant introduction results in a narrowing of the band gap without appreciable shifts of the CB edge (Fig. 2.12). The effect is explored by a so-called “band design” concept, that is, tailoring of the band gap and the VB position of semiconductor photocatalysts by non-metal dopings [2, 137, 185, 187–189]. The introduction of nitrogen into the lattice of titanium dioxide NPs achieved by TiO2 synthesis in the presence of ammonia [50], results in a retardation of the NPs growth during the calcination, a decrease of the average NP size from 20 to 14 nm, and a shift of the light sensitivity threshold of titania to longer wavelengths. Also, the doping generates oxygen vacancies on the TiO2 NP surface that promote adsorption of a sensitizer—eosin [50]. Such sensitized TiO2:N NPs showed a 3-times higher photocatalytic activity in the hydrogen evolution from aqueous TEA solutions as compared to undoped TiO2 NPs. 2.4 Photocatalytic Systems Based on the Nonmetal-Doped … 59

Fig. 2.12 A scheme of a photocatalytic system for the hydrogen production based on N-doped titania

N-doped TiO2 produced by the titania calcination with urea [229, 230] absorbs the visible light with k < 600 nm and exhibits a photocatalytic activity in the Vis-light-driven H2 production from aqueous solutions of Na2SO3 [229] and methanol [230]. Of two forms of surface nitrogen—the chemisorbed N and the nitrogen substituting O atoms in the oxide lattice, it is the latter that imparts titania with a Vis-sensitivity and the enhanced photocatalytic activity. Among N-doped titania materials a higher photocatalytic activity in the Vis-light-induced water splitting is typically observed for the mesoporous TiO2 [231–233]. The photocatalytic activity of TiO2:N in the water reduction can be further enhanced by combining it with Pt NPs [234] or other electron acceptors such as graphene derivatives [235, 236]. Annealing of tantalum oxide in a stream of ammonia and water vapors (to prevent the formation of tantalum nitride) yields tantalum oxynitride TaON absorbing visible light in a range of k < 500 nm (Fig. 2.13a) [236–239]. The material demonstrates a high photocatalytic activity in the water oxidation to O2 (with an apparent QY of 30% at 420–500 nm excitation), but possesses a negligible photoactivity in the water reduction, even in the presence of Pt (QY of 0.2% at 420–500 nm). Oppositely to pure oxide semiconductors, for which platinum is typically the best co-catalyst of hydrogen evolution, for the N-doped semiconductors a much higher activity is observed for a ruthenium co-catalyst (QY of 0.8% and 2.1% in the presence of methanol and ethanol, respectively). The photocatalytic deposition of Ru NPs produces 2–4-nm particles exhibiting a higher catalytic activity than 20–50-nm Ru NPs formed by the conventional impregnation/annealing [237]. Complete substitution of O with nitrogen yields tantalum nitride Ta3N5 with Eg = 2.1 eV (Fig. 2.13a) that is also an active photocatalyst of the water splitting [238]. The annealing in ammonia stream was used to produce zirconium oxynitride Zr2ON2 from ZrO2 [240]. The fusion of N2p-orbitals and O2p-orbitals in the VB of zirconium oxynitride results in the bandgap shrinking to 2.6 eV. After the photodeposition of 5 wt.% Pt, Zr2ON2 crystals exhibited a photocatalytic activity in the H2 evolution from aqueous solutions of methanol as well as the O2 evolution from silver nitrate solutions under the Vis-illumination. Solar-light-sensitive photocatalysts of the water reduction/oxidation were prepared by a partial nitridation of ZrO2/ Ta2O5 composite [241]. In a similar way, layered LaTaON2 and Y2Ta2O5N2 60 2 Semiconductor-Based Photocatalytic Systems …

Fig. 2.13 Diffuse reflectance spectra of Ta2O5, TaON, and Ta3N5; b a layout of the photoelectrocatalytic system for water splitting based on C-doped and Pt-decorated TiO2 NTs. Reprinted with permissions from Refs. [238](a) and [251](b). Copyright (2003, 2007) Elsevier (a) and American Chemical Society (b) perovskites were produced [242] exhibiting a photocatalytic activity in the hydrogen evolution from water/ethanol mixtures in the presence of Pt and Ru NPs. The introduction of a nitrogen dopant into the Sr2Nb2O7 perovskite yields a series of Sr2Nb2O7−xNx compounds that preserve the layered structure and, due to a contribution of the N2p-orbitals into the VB, exhibit a photocatalytic activity in the hydrogen evolution from water/methanol mixtures under the Vis-illumination [243]. A nitrogen-doped solid solution of gallium and zinc oxides (Ga1−xZnx)(N1−xOx) with x = 0.18 was used as a photocatalyst of the water reduction that, in a combination with a mixed co-catalyst Rh2–yCryO3, exhibited a QY of 6% at 420– 440 nm [244]. The co-catalyst was produced in situ via the photocatalytic reduction of KCrO4 over a (Ga1−xZnx)(N1−xOx)/Rh composite that, in turn, was synthesized by the photocatalytic deposition of Rh NPs [245]. A layer of chromium oxide on the metal surface prevents reverse reactions between H2 and O2 allowing the (Ga1–xZnx)(N1−xOx)/Rh2−yCryO3 heterostructure to function as a photocatalyst of the total water splitting. Active co-catalysts for this system were also produced by the semiconductor impregnation with a mixture of rhodium salts and ruthenium carbonyl Ru3(CO)12 followed by annealing [246–248]. The nitridation of a mixture of germanium and zinc oxides yields a compound (Zn1+xGe)(N2Ox) that exhibits a photocatalytic activity in the water reduction under illumination with the visible light [249]. N-doping of indium oxide narrows its band gap from 3.5 to 2.0 eV and imparts the semiconductor with a sensitivity to the visible light [250]. The photocurrent (proportional to the water reduction rate) generated under the Vis-illumination by the In2O3:N electrode is by a factor of 2 higher than that for the undoped indium oxide, and by a factor of almost 50—than the photocurrent generated by a TiO2:N reference photoelectrode [250]. The photoelectrochemical hydrogen generation was realized in a system, where a TiO2 NT array incorporated with Pt NPs acted as a cathode, while a photoanode 2.4 Photocatalytic Systems Based on the Nonmetal-Doped … 61 was formed by the carbon-doped TiO2 NTs (Fig. 2.13b) [251]. The photoanode was produced by the sonoelectrochemical anodization of titanium foil in a mixture of NH4F with ethylene glycol followed by the annealing in the H2 atmosphere. The cell demonstrated a photocurrent QY of 8.5% [251]. By the calcination of titanate NTs at 600 °C in a CO stream, 8–42 mol% carbon can be introduced without the formation of a separate titanium carbide phase [252]. The fusion of O2p- and C2p-orbitals in the VB results in the bandgap shrinking to 2.2 eV and a corresponding expansion of the spectral sensitivity range. By combining the carbon-doped TiO2 NTs (a photoanode) with Pt (a cathode) the Vis-light-driven water splitting to H2 and O2 was achieved [252, 253]. An alternative approach to the C-doped TiO2 consists in the burning of Ti foils in the carbon-enriched flame [254, 255]. The carbon doping results in a bandgap reduction from 3.20 to 2.65 eV as well as in the formation of a filled sub-band 1.6 eV above the VB top. This material was tested as a photoanode for the photoelectrochemical water splitting and showed a QY of 13% under the illumination with “white” light in aqueous 5.0 M NaOH solution. The C-doping increases electrode surface porosity favoring additionally to the photoelectrochemical reaction [254, 255]. Sulfur-doped TiO2 nanocrystals produced by a mechanochemical treatment of a mixture of titania with S8 were used for the photoelectrochemical water splitting under the Vis-illumination [256]. The photocatalytic water reduction or oxidation (depending on the type of co-catalyst—Pt or IrO2) under the Vis-illumination (420–480 nm) with the participation of indium-lanthanum oxysulfides was observed in [257].

2.5 Photocatalytic Systems Based on the Metal-Sulﬁde Semiconductors

Among the “solar” hydrogen production systems based on narrow-band-gap semiconductors a leading role is evidently played by metal-sulfide photocatalysts, mainly CdS, that is, however, photochemically unstable and liable to the photocorrosion. By this reason, a further development of the photocatalytic systems for the hydrogen production based on metal-sulfide semiconductors requires new methods of the photocorrosion mitigation. Also, a search is performed for new metal-sulfide materials that do not contain cadmium, lead, and other acutely toxic metals. Among the challenges in this field is also a search for new co-catalysts that do not contain precious platinum group/noble metals. New ways of photoactivity enhancement of the metal-sulfide semiconductors are constantly probed via a careful design of composite materials and harnessing of the quantum size effects inherent in the nanocrystalline metal-chalcogenide semiconductors. The anti-corrosion stability can be achieved via a combination of metal-sulfide NPs with various photochemically passive carriers. For example, CdS NPs formed in the pores of zeolites were found to be photochemically stable when used as a 62 2 Semiconductor-Based Photocatalytic Systems … photocatalyst of the hydrogen evolution from water/alcohol mixtures [128, 258– 263]. The photoactivity of such heterostructures depends considerably on the carrier structure and increases from zeolite L to SBA-15 to zeolite Y [258]. A combination of a high photocatalytic activity and photostability was observed for CdS NPs stabilized by the colloidal silica [264] and organic polymers [265, 266], as well as for CdS NPs deposited on the surface of carbon nanofibers [267], aluminium oxide [268], silica gels [269–271], and glasses [272, 273]. The glass-incorporated CdS NPs revealed a high photostability and can be used as visible-light-sensitive photocatalysts for the hydrogen evolution from aqueous H2S solutions with a QY of 17–18% [272]. Interaction between cadmium(II) salts and polyvinylidene sulfide yields 5–30-nm CdS NPs regularly dispersed over the polymer surface [274]. Such composite exhibits photochemical stability and a high (up to 20%) QY of the hydrogen evolution from aqueous H2S solutions. The interest to nanocrystalline cadmium sulfide is greatly stimulated by a strong dependence of the electron properties of CdS NPs on the crystal size (d)atd <5– 6 nm. A variation of the CdS NP size in this range is accompanied by pronounced changes of the optical and photochemical NP properties allowing for a tuning of the spectral sensitivity range and efficiency of the NP-based photocatalytic systems. This feature is excellently exemplified by the photocatalytic systems for the hydrogen production based on size-selected CdS NPs decorated with Pt NP co-catalyst [275]. The H2 evolution QY decreases from around 17 to *11%, as the CdS NP size increases from 2.8 to 4.6 nm. The dependence was interpreted in terms of a size-dependent driving force of photoinduced charge transfer from CdS to vacant states of Pt NPs (Fig. 2.14a). However, the photoactivity increase is counter-weighted by a “blue” shift of the absorption band edge of CdS NPs as their size is reduced (Fig. 2.14b), resulting in a partial loss of the solar light harvesting ability. To compensate for this detrimental effect, a double-chamber photocatalytic reactor was proposed [275], where shorter-wavelength light is absorbed selectively by smaller and more active 2.8-nm CdS NPs, while a longer-wavelength portion of the light passes through the first chamber and is absorbed in the second chamber by larger CdS NPs. In this way, a 50% increment of the H2 evolution efficiency can be achieved for the double-chamber systems as compared with a single photocatalytic reactor with a mixture of smaller and larger CdS NPs [275]. Similarly to cadmium sulfide, a reduction in the size of MoS2 NPs from 15–25 to *5 nm was found to result in almost doubled photocatalytic activity in the H2 evolution, both types of NPs being much active than bulk molybdenum sulfide [276]. Alternatively to CdS, bismuth, indium, iron, ruthenium, zirconium and silver sulfides, as well as related solid solutions can be used as water reduction photocatalysts. For example, RuS2 NPs immobilized on thiol-modified polystyrene beads exhibited a photocatalytic activity in the hydrogen evolution from water/2-propanol [277]. The earth abundant pyrite FeS2 was shown to be a suitable candidate for the Vis-light photo-electrochemical H2 generation [278]. 2.5 Photocatalytic Systems Based on the Metal-Sulfide … 63

Fig. 2.14 Band diagram (a) and absorption spectra (b) of 2.8–4.6-nm CdS NPs; c a scheme of a dual photocatalytic reactor for the H2 evolution. Reprinted with permissions from Ref. [275]. Copyright (2015) The Royal Society of Chemistry

A solvothermal synthesis of tin sulfide starting from Sn(II) salts yields a mixed SnS2/SnS nanostructure with the intertwined tin(IV) and tin(II) sulfide layers. A lattice mismatch between the layers generates inherent sulfur vacancies resulting in a Eg narrowing by around 0.3 eV (to *2 eV) as compared with pure SnS2 [279]. Such composite showed an excellent activity in the H2 evolution under illumination with a “blue” LED source (400–500 nm). Layered zinc indium sulfide and its composites with metal NPs exhibited a photocatalytic activity in the Vis-light-driven H2 evolution from aqueous Na2S/Na2SO3 solutions [280–292]. The photoactivity of ZnIn2S4 was found to increase proportionally to the post-synthesis hydrothermal treatment duration as well as to the concentration of cetyltrimethyl ammonium bromide acting as a template. The dependence was assumed to originate from a deformation of the ZnIn2S4 crystal lattice resulting in a dipole moment in the semiconductor interlayer space that favors to the photogenerated charge carriers separation. The copper(II) doping of ZnIn2S4 expands its spectral sensitivity to around 800 nm, the maximal rate of photocatalytic hydrogen production registered at 0.5 wt.% copper content [293]. Good perspectives as Vis-sensitive photocatalysts of the water reduction could be envisaged for a number of ternary/multinary metal sulfides with a narrow band gap that suit perfectly as visible light harvesters and can potentially induce the water photodecomposition—CaIn2S4 with a band gap of 1.76 eV [294, 295], AgGaS2 (Eg = 2.48 eV) [296], CuGaS2 [297], CuIn1−xGaxS[298], (CuGa)1−xZn2xS2 [299, 300], Zn1−2x(CuGa)xGa2S4 [301], Cu3SnS4 (1.38 eV) [302], and Cu2ZnSnS4 (Eg = 1.75 eV) [303–307]. Some of these materials were studied as nanocrystalline materials, while for others the effects of nano-scaling are still to be explored. A photocatalytic activity in the hydrogen evolution from aqueous Na2S/Na2SO3 solutions under the Vis-illumination was observed for mesoporous agglomerates of 64 2 Semiconductor-Based Photocatalytic Systems …

CuInS2 NPs (Eg = 1.53 eV) modified by Pt NPs [308], for CuIn5S8 [309] and CuIn0.7Ga0.3S2 films [310], CuInS2/NaInS2 nanoheterostructures [311], (CuIn)xZn2 (1−x)S2 (x = 0.01–0.50) microspheres [312–317], non-stoichiometric Cu–In–Zn–S NPs attached to reduced graphene oxide (RGO) sheets [318], as well as for the nanocrystalline (CuAg)xIn2xZn2(1−2x)S2 solid solutions [319–322]. The Vis-light sensitivity of these materials originates from a contribution of Cu3d- and S3p-orbitals into the VB and In5s5p- and Zn4s4p-orbitals—into the CB of mixed sulfide semiconductor. Nanoporous ZnS–In2S3–Ag2S solid solutions demonstrated a photocatalytic activity in the water reduction without additional co-catalysts [323]. Similarly, ternary sulfide AgInZn7S9 (Eg = 2.3 eV) is capable of the photocatalytic hydrogen evolution from water with no co-catalysts and electron donors present in the system [324]. When used together with Pt NPs and a sacrificial donor (Na2S/Na2SO3), this quaternary photocatalyst showed an H2 production QY of *15%. CdIn2S4 NTs revealed the photocatalytic properties in the water reduction with no additional sacrificial donors with a QY of up to 17% [325]. A high photocatalytic activity in this process was also observed for ZnIn2S4/CdIn2S4 [326, 327], In2S3/ZnIn2S4 [328], and CdS/ZnIn2S4/RGO heterostructures [329]. A mixed sulfide AgIn5S8 (Eg = 1.77 eV) modified by Pt NPs was used as a Vis-sensitive photocatalyst of the hydrogen evolution from aqueous sulfide/sulfite solutions with a QY of 5.3% [330]. Various layered metal-sulfide nanomaterials, both in the form of nanometer grains and especially as a few-layer (single-layer) NSs are of a great potential for the solar H2 production. For example, layered NaInS2 (Eg = 2.3 eV) was found to be an efficient photocatalyst of the water reduction with a QY of 6% [331], while bulk indium sulfide remains inactive in this reaction. The 2.5-nm In2S3 NPs produced by ion exchange/sulfidation in the pores of titania-containing Ti-MCM-41 zeolite exhibited pronounced photocatalytic properties in the water reduction [332]. The activity stems from an efficient photoinduced electron transfer from In2S3 NPs to the zeolite host and inhibition of the subsequent electron-hole recombination. First-principles calculations showed perspectives of single-layer and a-few-layer zirconium sulfide as a visible-light-sensitive photocatalyst (Eg = 1.9–2.0 eV) of the water splitting [333], though these predictions still require an experimental verification. A comprehensive review of the photochemical water splitting on layered transition metal dichalcogenides (TMDs) can be found in [334]. The exfoliation of some of TMDs into single-layer sheets can result in a dramatic increase of the optical bandgap thus providing an additional driving force for the water splitting processes as illustrated in Fig. 2.15a for MoS2. A band diagram, shown in Fig. 2.15b for a series of reported single-layer TMDs, provides a notion of possible candidates for the utilization in the photocatalytic/photoelectrochemical systems for the water splitting. Ultra-thin sheets of MnSb2S4 with Eg = 1.9 eV and a thickness of 0.76 nm produced by the spontaneous thermal exfoliation of hydrazine-intercalated bulk materials revealed promising properties as a photocatalyst of the water reduction with a peak QY of 0.14% [335]. 2.5 Photocatalytic Systems Based on the Metal-Sulfide … 65

Fig. 2.15 a Energy diagram for bulk and single-layer MoS2; b band edge positions of single-layer + sulﬁde TMDs relative to the vacuum level, including the redox potentials for the H /H2 and O2/ H2O couples at pH = 0. Reprinted with permissions from Ref. [334]. Copyright (2015) The Royal Society of Chemistry

The films of layered Bi2S3 with Eg = 1.28 eV produced by the electrodeposition can be used as an efficient and stable photocatalyst of the Vis-light-driven hydrogen evolution from aqueous Na2S solutions [336]. In a similar photocatalytic system based on a Bi2S3/zeolite Y composite the QY of H2 evolution reached 0.12% [337]. Mixed sulfides CdxZn1−xS are typically capable of the photocatalytic water reduction without additional co-catalysts [338–341]. For these semiconductors, as a rule, a dome-shaped relationship between the photocatalyst composition (the parameter x) and the rate of hydrogen evolution is observed. Such dependence is quite non-trivial because both CB and VB potentials of CdxZn1−xS solid solutions increase as the Cd is gradually substituted with Zn (Fig. 2.16) and, therefore, a monotonous dependence of the photocatalytic activity of cadmium-zinc sulfide crystals on their composition should be expected. The exact position of the maximum on this relationship still remains contro- versial, most probably, due to differences in synthesis methods of CdxZn1−xS solid solutions that can affect considerably their photochemical behavior. For example, it was found [338] that a maximal apparent QY of hydrogen evolution from aqueous sulfide/sulfite solutions, 10.2% at 420 nm, corresponds to x = 0.8. According to [339], the highest H2 production QY with the participation of CdxZn1–xS crystals can be observed in a range of x =0.25–0.30. Studies of the photocatalytic activity of cadmium-zinc sulfides precipitated on paper revealed two distinct photoactivity maxima corresponding to x = 0.5 and 0.2 [342]. The peak activity of CdxZn1−xS microspheres produced by hydrothermal synthesis [343] was found at x = 0.1. A detailed transient flash photolysis study [344] showed that a dependence between the capability of CdxZn1−xS NPs to accumulate an excessive negative charge (photoinduced polarization of NPs) and the NP composition also has a dome-shaped character. The maximum position on this dependence corresponds to a maximum position on the dependence between the composition of CdxZn1−xS NPs and the QY of hydrogen evolution (Fig. 2.17). Therefore, a direct relationship 66 2 Semiconductor-Based Photocatalytic Systems …

Fig. 2.16 CB and VB potentials of CdxZn1−xS solid solutions with various Cd/Zn ratios. Reprinted with permissions from Ref. [8]. Copyright (2010) Elsevier

Fig. 2.17 a Dependence between the quantum yield Ф(H2) of the photocatalytic H2 evolution from aqueous Na2SO3 solution and the Cd molar fraction x in CdxZn1−xS NPs. b Effect of the photoetching time of nanocrystalline Pt/CdS on the photocatalytic H2 evolution rate. Reprinted and adapted with permission from Ref. [348]. Copyright (2008) Elsevier

between the photocatalytic activity of CdxZn1−xS NPs and their electric capacitance can be concluded from these dependences. The photocatalytic water reduction by electron donors was reported for more complex semiconductors based on cadmium-zinc sulfide, such as Cd0.1SnxZn0.9 −2xS solid solutions [345]. The compound with x = 0.01 exhibited a 1.5-fold higher III photocatalytic activity than undoped Cd0.1Zn0.9S. Doping of Cd0.5Zn0.5S with Bi increased considerably the QY of photocatalytic H2 evolution from aqueous Na2S/Na2SO3, solutions that reached *10% at 0.1 mol% dopant [346]. Along with the stability issue, various strategies are probed to enhance the photoactivity of metal-sulfide narrow-bandgap semiconductors suitable for the solar water splitting. At that, the most fruitful approaches include using (i) loosely 2.5 Photocatalytic Systems Based on the Metal-Sulfide … 67 aggregated nanocrystalline and mesoporous metal sulfides; (ii) biphase nanocrystalline metal-sulfides and materials with a graded (gradient) composition; (iii) chemical/photochemical treatment of a nanocrystalline metal-sulfide aimed at the elimination of the surface defects/ligands; (iv) metal-sulfide NPs with an anisotropic shape; (v) composites with the water oxidation and reduction processes separated in space. This list and selected examples given below provide a mere illustration of a variety of the possible ways of influencing/enhancing the photoactivity of metal-sulfide semiconductors. The attractiveness of the metal-sulfide photocatalytic systems for the solar H2 production can be also enhanced by using broadly available raw materials and contaminants as sacrificial donors. An ultrasound treatment of reaction mixtures during deposition of cadmium sulfide on the surface of aluminum and magnesium oxides favors to the formation of mesoporous CdS with an average pore diameter of 5.5 nm and a particle size of 4–6nm[347]. Such materials exhibited a high photocatalytic activity in the hydrogen evolution from Na2S/Na2SO3 solutions in the presence of Pt group metals with the catalytic activity of metals increasing from Rh to Pd to Pt. The photocatalytic activity of nanocrystalline CdS can be boosted by a photochemical treatment in aqueous air-saturated solutions of formic acid [348]. The treatment decreases the NP size as a result of the oxidative photocorrosion. Simultaneously, the cleavage of a surface NP layer eliminates the surface defects participating in the electron-hole recombination and thus the overall photocatalytic activity of CdS NPs is increased by more than 2 times (Fig. 2.17b) [348]. A ligand shell on the surface of colloidal cadmium sulfide NPs is necessary to ensure the individual character of each NP and prevent their aggregation. However, the ligands even as small as MPA can present an obstacle for the photoinduced electron transfer to water molecules and inhibit the photocatalytic hydrogen evolution. As shown in [349], the CdS NPs “stripped” from the surface ligands and stabilized only by a surface charge reveal a two orders of magnitude higher photoactivity in the H2 evolution from aqueous Na2SO3 solutions as compared to the MPA-capped NPs with the same NP size. The colloidal CdS NPs stabilized elec- trostatically by an outer shell of sulfide ions revealed a 8–9 times higher efficiency of the photocatalytic H2 evolution from water/hydrazine mixtures than similar NPs capped with MPA [350]. Typically, different crystalline modifications of a semiconductor differ in the band energies and can be combined to produce heterostructures, where efficient separation of the photogenerated charge carriers becomes possible. The most well-known example of such a heterostructure is the commercially available nanocrystalline titania Evonik P25, consisting of 70–80% anatase and 20–30% rutile, that is extensively used as a benchmark photocatalyst for comparing the photoactivity of different (and not only oxide) semiconductor materials. The anatase and rutile have slightly offset CB potentials enabling a one-way migration of the photogenerated electrons from anatase to rutile. Similar heterostructures can be composed of different-phase metal sulfide semiconductors. For example, cubic CdS NPs (Eg = 2.6 eV) can be deposited on 68 2 Semiconductor-Based Photocatalytic Systems … the surface of hexagonal microcrystalline cadmium sulfide with Eg = 2.3 eV [351]. The charge separation occurring in this system due to a difference in the CB positions imparts the composite with a high photocatalytic activity in the hydrogen evolution from aqueous sodium sulfite solutions. Similarly to the above-discussed CdS/TiO2/Pt system [107], here also the co-catalyst localization plays an important role. The highest activity was observed for a composite produced by the photocatalytic deposition of Pt NPs on the microcrystalline CdS prior to the formation of a layer of cubic CdS NPs. A dramatic acceleration (by around 500 times) of the photocatalytic H2 production was observed after the deposition of a thin (*2.5 nm) shell of cubic CdS on the surface of hexagonal CdS NRs [352]. The concentric core/shell CdS NRs also exhibited unrivaled photostability and even the possibility of hydrogen evolution in aerobic conditions, indicating a very efficient spatial separation of H2 and O2 generation events on different locations of the photocatalyst surface. Mixed-phase elongated CdxZn1−xS NRs with thin hexagonal wurtzite layers “sandwiched” between thicker cubic zinc blend domains were reported to be an efficient hydrogen evolution photocatalyst [341]. The CB and VB edge offsets and an internal electric field existing in such heterostructures result in a directed flow of the photogenerated CB electrons from wurtzite (WZ) to zinc blend (ZB) and the VB holes—in the opposite direction (Fig. 2.18a). In this way, the water reduction and sulfite oxidation occur at different sites and the electron-hole recombination is efficiently suppressed without additional metal co-catalysts [341]. In the case of ZnIn2S4, the CB potential of a cubic modification (−1.5 V vs. NHE) is more negative than that of a hexagonal phase (–1.1 V vs. NHE) and thus, a combination of the ZB and WZ NPs results in a flow of the photogenerated electrons from the cubic to the hexagonal zinc indium sulfide phase [291]. As a result, the ZB/WZ ZnIn2S4 composite is a highly superior photocatalyst of the water reduction as compared to individual ZB and WZ phases (Fig. 2.18b). A number of photochemically active metal chalcogenides can form solid solutions with almost ideally mixed components. As discussed above, cadmium and zinc sulfides form solid-solution compounds with the composition varying from pure CdS to pure ZnS via intermediate CdxZn1−xS phases with intermixed metal cations and a joint S2− sublattice. As CdII and ZnII have a different reactivity to sulfide-ions and the sulfides of cadmium and zinc have a different solubility, it is possible, by properly adjusting the synthesis conditions, to form CdxZn1−xS crystals with a gradient of cadmium or zinc concentration, for example, the crystals enriched with CdII on the surface. The cadmium and zinc sulfide possess quite different bandgaps (2.4 and 3.8 eV for bulk cubic CdS and ZnS, respectively) and differ also in the CB and VB energies, the CB level changing from −0.8 V versus NHE for CdS to −1.8 V versus II NHE for ZnS [8]. In a graded CdxZn1−xS crystal with a Cd -enriched outer layer the CB potential on the crystal surface is, therefore, lower (closer to that of pure CdS) than in the bulk of the crystal, where it grows and shifts closer to the CB potential of pure ZnS. The CB level gradient directs the photogenerated electrons to the surface and prevents their recombination with the photogenerated VB holes. 2.5 Photocatalytic Systems Based on the Metal-Sulfide … 69

Fig. 2.18 a Upper panel: HRTEM image of a typical CdxZn1−xS NR. The blue squares and green arrows index the segments of WZ and ZB structures. Lower panel: migration of charge carriers in CdxZn1−xS ZB/WZ nanojunctions; b kinetic curves of the hydrogen evolution in the presence of sole ZB and WZ ZnIn2S4 as well as a ZB/WZ nanojunction. Insert: an HRTEM image of the ZB/WZ nanojunction. Reprinted and adapted with permissions from [341](a) and [291](b). Copyright (2016) Americal Chemical Society (a) and Elsevier (b)

This charge separation principle was realized for a CdxZn1−xS/SiO2 heterostructure produced by the sulfurization of a graded CdxZn1−xO/SiO2 composite [271]. The surface area of CdxZn1−xS crystals can be cleaved layer by layer via the bombardment with heavy Ar+ ions thus allowing to reveal with the XPS a graded structure of such crystals, the Cd to Zn molar ratio decreasing from *1to2 on the surface to *1–3 in the bulk of the crystals (Fig. 2.19a). The graded crystal structure favors to the directed ﬂow of photogenerated CB electrons along a ECB downfall, that is, toward the crystal surface, where they can react with water evolving gaseous hydrogen. Due to this effect, the graded CdxZn1−xS/SiO2 heterostructure showed a high photocatalytic activity in the water

Fig. 2.19 a Illustration of a graded structure of CdxZn1−xS crystals; b photocatalytic H2 evolution activity of ZnS/SiO2, CdS/SiO2,CdxZn1−xS/SiO2 and CdxZn1−xS with different structures under the visible light irradiation. Reprinted and adapted with permissions from Ref. [271]. Copyright (2016) The Royal Society of Chemistry 70 2 Semiconductor-Based Photocatalytic Systems … reduction leaving behind both the individual sulfides and cubic/hexagonal CdxZn1 −xS NPs with a regular non-graded structure (Fig. 2.19b). The electron flow arriving on the crystal surface reduces undercoordinated surface cadmium species to Cd0 that acts as a metal co-catalyst of the photoprocess and therefore, no additional co-catalyst is necessary [271]. The liability of metal-sulfide semiconductors to the photocorrosion can be used to transform the surface layer of the sulfide crystals into a metal oxide or a mixed oxide/sulfide composition with a tunable band structure. For example, the photolysis of ZnS microcrystals under 254-nm UV light at the ambient air pressure and moisture converts the surface layer of the microcrystals into a graded ZnSxO1−x interface [353]. A combination of the surface etching with Auger spectroscopic probing revealed that the surface layer of such photochemically treated ZnS microcrystals is composed of a zinc oxysulfide solid solution NPs with the oxygen content decreasing from the surface to the crystal bulk (Fig. 2.20a). The graded region extends to 80–100 nm matching the depth of light penetration into the zinc sulfide microcrystals. As the CB level of pure ZnO (*−0.5 V vs. NHE at pH 7) is lower than the CB level of ZnS, the CB electrons photogenerated within the graded ZnSxO1−x layer are directed toward the surface, where they can participate in the water reduction. Similarly to the case of CdxZn1−xS, this effect results in a drastic enhancement of the hydrogen evolution, that proceeds also without additional co-catalysts due to a partial reduction of surface Zn species to Zn0 [353]. The photocatalytic activity of the photoproduced ZnSxO1−x layer depends on the duration of photolysis (Fig. 2.20b), most probably, due to a balance between the thickness of oxidized

Fig. 2.20 a Distribution of Zn, S, C, and O atoms in the photolyzed single ZnS crystal derived from the Auger spectroscopic data. The etching time of 10 min roughly corresponds to a 100-nm etching depth. The signal of adventitious carbon stems from surface contaminations. b The rate of the photocatalytic H2 evolution (RH) from water/ethanol mixture under UV illumination (k > 320 nm) in the presence of microcrystalline ZnS, ZnS/ZnSxO1−x heterostructures produced by the photolysis and a mechanical mixture of ZnS and ZnO. Reprinted with permissions from Ref. [353]. Copyright (2016) Elsevier 2.5 Photocatalytic Systems Based on the Metal-Sulfide … 71 layer and the light penetration depth, and supersedes by an order of magnitude the activity of a mechanical mixture of ZnS and ZnO of the same composition. A photocatalytic activity increase can be achieved by using semiconductor crystals with an anisotropic (non-spheric) shape, where the charge migration rate is different depending on the crystal axis. For example, CdS NWs with a length of 3–4 lm and a diameter of 50 nm act as a more efficient photocatalyst of H2 evolution from aqueous Na2S/Na2SO3 solution than isotropic nanocrystalline CdS [354]. The commercial attractiveness of the photocatalytic systems for the hydrogen production based on the narrow-band-gap semiconductors can be enhanced by a cut in the costs of sacrificial donors. For example, a CdS/LaMnO3 heterostructure can be used as a Vis-sensitive photocatalyst for the hydrogen production either from conventional sulfide/sulfide donors [355] or from the broadly available biomass [356]. The rate of hydrogen evolution from the biomass under the poly–chromatic illumination is comparable to that observed in the sulfide/sulfite/containing systems due to the presence of perfect donors—methanol and formic acid in the partially fermented biomass. A special interest is evoked by the issue of the utilization of various industrial wastes as sacrificial donors for the water reduction. This approach can be illustrated by the photocatalytic hydrogen evolution from hydrogen sulfide solutions in ethanola- mine and other aliphatic amines in the presence of CdS/Pt nanoheterostructures [357]. Such solutions are abundantly produced as wastes of the carbon and natural gas industries as well as in the technologies of crude oil desulfurization. The amines readily dissolve hydrogen sulfide and favor its ionization and proton release that can be reduced to the molecular hydrogen. At the same time, a high solubility of poly- fi 2− 2À sul de anions as products of the S oxidation promotes desorption of Sx from the photocatalyst surface and prohibits reverse reactions between polysulfide and the photogenerated CB electrons. The sulfide-capped colloidal CdS NPs were found to be an excellent photocatalyst of H2 evolution by using aqueous hydrazine as a sacrificial electron donor [350]. The hydrazine contains 12 wt.% of hydrogen and is considered as a promising liquid hydrogen carrier with N2 as a sole product of the oxidative decomposition.

2.6 Emerging Semiconductor Photocatalysts for the Solar Hydrogen Production

A constant search for new photocatalysts of the water splitting is currently under way. Each new compound emerging within the focus of attention of possible photocatalytic applications is typically tested as a hydrogen evolution photocatalyst. The emerging photocatalysts are of exceedingly broad scope making almost impossible their rigorous classification, the most trend-making of them being new narrow-bandgap semiconductors with a lattice formed by metal ions and oxygen (oxides, metallates, 72 2 Semiconductor-Based Photocatalytic Systems … etc.), metal-chalcogenide semiconductors, carbonaceous species like graphitic carbon nitride and carbon nanoparticles, and metal-organic frameworks. Some of the new semiconductor photocatalysts can be referred to as “exotic” because they have only recently entered the attention spotlight of the photocatalysis community and only a little is known about the potential and perspectives of such materials. The following subsection provides an account of the photocatalytic hydrogen evolution systems based on the semiconductor materials that can be currently characterized as “emerging” photocatalysts. Photocatalysts with a chalcogenide lattice. The examples of hydrogen-evolving photocatalysts with a chalcogenide-based lattice are confined mostly to the above-discussed sulfide semiconductors because metal selenide and telluride semiconductors have typically too low bandgaps to induce the water reduction/oxidation reactions and reveal an unacceptably low photostability even in the presence of sacrificial electron donors. The photocatalytic hydrogen evolution from sulfide/sulfite solutions under the Vis-illumination of CdSe nanobelts [358] was a rare example of the photochemical activity of cadmium selenide and, probably, the first evidence of its capability to induce the water reduction. Later, size dependences of the photocatalytic activity of CdSe NPs in the water reduction were reported [359, 360]. The bandgap of CdSe NPs is broadened considerably as the NP size decreases from 3.1 to *1.8 nm due to the QSEs resulting in a steep growth of the photocatalytic activity with a NP size reduction (Fig. 2.21a). No activity is typically observed for the CdSe NPs larger than 3–4 nm. As the size of CdSe NPs is reduced, a strong “blue” shift of the NP absorption band edge is observed (Fig. 2.21a, insert) resulting in a drastic loss of visible light harvesting capability. Nevertheless, smaller CdSe NPs still reveal a high photocatalytic activity (when the H2 evolution rate is normalized to the light absorbance) as a result of a larger driving force of the CB electron transfer to water molecules [361]. This example again illustrates that a reasonable trade-off between the light harvesting capacity and the CB energy should be maintained for the quantum-sized semiconductor NPs to attain the maximal QY of the solar hydrogen production [359, 360]. Also, the control of surface chemistry of the nanocrystalline CdSe plays a crucial role in the photocatalytic water reduction. For example, passivation of the surface defects of CdSe NRs tipped with Pt NPs with an atomically-thick CdS layer increases the H2 production rate by a factor of 6–7[362]. Nanocrystalline CdSe was used as a light-harvesting material in a photoanode designed for the water oxidation to O2 with the hydrogen evolution occuring on a counter electrode [363]. The cadmium selenide was protected against the corrosion/ photocorrosion by a sputtered layer of metallic cobalt that was partially converted into a water oxidation co-catalyst—CoPO4 by the oxidative etching in a phosphate buffer [363]. The reported assortment of other (Cd-free) metal selenide photocatalysts of water reduction is apparently limited to the nanocrystalline c-In2Se3 (a Pt NP co-catalyst, TEA as a sacrificial donor, Eg * 1.6 eV) [364]. 2.6 Emerging Semiconductor Photocatalysts … 73

Fig. 2.21 a The absorption-corrected rate of Xe lamp spectrum 2.0 nm the photocatalytic H2 2.8 nm 4.0 nm evolution with the participation of size-selected CdSe NPs. Insert: absorption spectra of 2.0–4.0 nm CdSe NPs overlapped with a typical spectrum of a Xe lamp (used for the simulation of the solar spectrum). Reprinted and adapted with permissions from Ref. [359]. Copyright (2012) The Royal Society of Chemistry

Photocatalysts with oxygen-containing lattice. A broad range of various visible-light-sensitive oxides, both stoichiometric and mixed-valence, as well as numerous metallates (ferrites, aluminates, cuprates, stannates, borates, etc.) can be used as photocatalysts of the solar H2 production in the bulk and nanocrystalline forms. Among these compounds a special attention was paid to layered perovskites (tantalates, niobates, titanates, ferrites, etc.) used as bulk materials, NPs, NRs, nanofibers, hollow spheres, etc. and exhibiting a lucrative combination of a high stability and activity in the hydrogen evolution. The CuGa2O4 and CuGa2−xFexO4 spinels were recently introduced as Vis-sensitive photocatalysts of the H2 evolution from aqueous H2S solutions in the presence of a NiO/RuO2 co-catalyst [365]. A photocatalytic activity in the hydrogen evolution from aqueous sulfide/sulfite solutions was observed for CuLaO2 (Eg = 2.33 eV) modified with the photodeposited Pt NPs [366], as well as for CuLaO2.62 [367] and CuAlO2 [368]. Ultra-thin (*3 nm) tin niobate NSs produced by the hydrothermal exfoliation of thicker 2D SnNb2O6 particles showed a remarkable photocatalytic activity in the visible-light-driven water reduction, which is 4 and 14 times higher than the activity of the 50-nm thick particles and bulk niobate, respectively [369]. Nanocrystals of CuFe2O4 and ZnFe2O4 [370, 371] were successfully tested as photocatalysts of the water reduction under the illumination with the visible light. The photocatalytic properties in similar systems were reported for MnO2 [372], Ga2O3 [373], NiO [374], Sn3O4 [375], nanocrystalline LaFeO3 [376, 377] and 1 LaMgxTa1−xO1+3xN2−3x (x /3) perovskites [378], nanoparticles of SrSnO3 [379] and MnCo2O4 [380], copper borate [381], a family of M2BiSbO7 (M = Ga, Fe, Gd) [382], as well as for ZnAg3SbO4 [383]. Also, a number of new UV-light-sensitive semiconductor photocatalysts for the water reduction was reported in recent years, such as titanium phosphate [384], gallium borate [385], Zn2GeO4 [386, 387], LaCO3OH [388], Sm2GaTaO7 [389], that can potentially be sensitized by molecular dyes and narrow-band-gap semiconductor NPs. 74 2 Semiconductor-Based Photocatalytic Systems …

A new family of In-based photocatalysts of the water reduction under the Vis-illumination [370] includes InVO4 (Eg = 1.9 eV), InNbO4 (Eg = 2.5 eV), and InTaO4 (Eg = 2.6 eV). Differences in the band gap among these compounds originate from a contribution of the V3d-, Nb4d- and Ta5d-orbitals in the respective CB positions. Mesoporous InVO4 produced by a template synthesis exhibited a higher photocatalytic activity in the hydrogen evolution than the non-porous nanocrystalline indium vanadate [390]. A recently reported series of homologous compounds ZnxIn2O3+x (x =4,5,or7) showed a unique combination of a strong light harvesting capability with a high mobility of the photogenerated charge carriers [391]. A layered structure of such compounds comprised of light-absorbing Zn(In)O4(5) layers alternated with InO6 layers provides abundant charge collection sites and transport channels (Fig. 2.22a). An enhanced sensitivity of these compounds to the visible light as compared to individual ZnO and In2O3 (Fig. 2.22b) originates from a hybridization between the O2p and In4d orbitals in the Zn(In)O4(5) layers resulting in an upward shift of the VB edge and a reduced bandgap (2.57–2.67 eV depending on x)[391]. Under the full-spectrum illumination the mixed compounds revealed an almost 5-fold higher photocatalytic activity in the water reduction than the parent zinc and indium(III) oxides (Fig. 2.22c). Photocathodic production of solar hydrogen. In conventional photoelectrochemical systems for the solar hydrogen production a visible-light-sensitive semiconductor/heterostructure attached to a conductive substrate acts as a photoanode. The light-excited photoanode oxidizes water or a sacrificial donor present in the electrolyte and injects a photogenerated electron into the conductive substrate-collector, that transfers it to an outer electrical circuit. Afterward, the electrons come to a cathode, where the water reduction to molecular hydrogen takes place. In such a system, the reduction and oxidation half-reactions are separated in space and the H2 evolution efficiency is determined by the catalytic properties of the cathode and the photoactivity of the anode responsible for the donor oxidation. In recent years, alternative photoelectrochemical systems emerged where the target reaction—the water reduction to H2 takes place as a direct consequence of the light absorption by a photocathode, while the photogenerated VB holes are transferred into the electric circuit to a counter anode, where the water oxidation and O2 evolution occur. An account of the current progress in this field can be found in [392]. The assortment of semiconductor materials suitable for the photocathode applications is rather limited, because they should comply with a set of rigorous requirements, in particular, the sensitivity to the visible light (that is, have a relatively narrow enough bandgap) and a high conduction band potential, that is, ECB negative enough to induce the water reduction (Fig. 2.23)[392]. Also, a photocathode should be coupled with an appropriate oxygen-evolving electrocatalyst to ensure the cyclic performance of the photoelectrochemical system and the efficient regeneration of the photocathode. As compared to the photoanodes comprised typically of n-type semiconductors and prone to the photocorrosion, the photocathodes usually use p-type conducting 2.6 Emerging Semiconductor Photocatalysts … 75

Fig. 2.22 a Schematic representation of a Zn4In2O7 photocatalyst with different functional parts; b photographs of mixed zinc indates and pure zinc and indium oxides; c the average photocatalytic hydrogen production rate under the full range irradiation (k 250 nm) and the visible light irradiation (k 420 nm). Reprinted with permissions from Ref. [391]. Copyright (2016) Americal Chemical Society

Fig. 2.23 Band edge positions of several typical photocathodic semiconductor materials. Reprinted with permissions from Ref. [392]. Copyright (2015) The Royal Society of Chemistry

semiconductor materials that appear much more stable toward the photocorrosion provided that the photogenerated electrons are efficiently transferred to water molecules. Copper oxides are very promising visible-light-sensitive materials for the photocathodes of the hydrogen-evolving photoelectrochemical systems [393]. To achieve efficient light harvesting with copper oxides it is suggested to use Cu2O (CuO) NW arrays rather than conventional planar semiconductor electrodes [393, 394]. A mixed Cu2O/CuO heterostructure can be easily sensitized with a layer of copper sulfide via the ion exchange reaction and additionally decorated with a Pt NP co-catalyst exhibiting quite spectacular 3.6% efficiency of the solar light conversion [395]. The reduced graphene oxide (RGO) was found to be an efficient 76 2 Semiconductor-Based Photocatalytic Systems … co-catalyst for 1D Cu2O photocathodes and a promising candidate to replace the conventional noble metal co-catalysts [394]. Another promising and stable photocathode material is nickel oxide. As NiO has a weak absorbance in the visible spectral range it can be coupled with other robust and strongly-absorbing semiconductors, for example, graphitic carbon nitride. Due to a band edge offset in the NiO/GCN heterostructure, the nickel oxide layer accepts the photogenerated VB holes from GCN while the CB electrons of GCN reduce water to H2 [396]. A similar cascade hole transfer from GCN to a p-type semiconductor is observed in a GCN/CoSe2 composite attached to the top of a Si microwire array photocathode [397]. As Fig. 2.23 shows, silicon can also be utilized as a photocathode material for the solar hydrogen production. The Si photocathodes are typically designed as NW arrays [397–400] that can be additionally decorated with other semiconductors [397, 399] and metal co-catalysts (Pt) [399], and covered with a titania layer to protect the photocathode from the photocorrosion [399]. The activity of silicon can be boosted by making porous photocathodes with a highly developed surface area from Si NPs [401]. Amorphous silicon coupled to a triple Ni–Mo–Zn alloy as a hydrogen evolution catalyst and a Co-containing water oxidation catalyst forms a photoelectrochemical cell for the hydrogen evolution with a QY of 4.7% under the 1 Sun illumination [402]. The visible-light-induced H2 evolution was observed on a photocathode formed by the CuInS2 nanodisks grown epitaxially on cubic Au NPs (Fig. 2.24a) [403] coupled to a Pt counter-electrode. The water reduction process was assumed to involve both the photogenerated CB electrons of CuInS2 and the “hot” electrons injected into the semiconductor from the plasmon-excited Au NP seeds. A search into other possible ternary/quaternary narrow-bandgap metal-chalcogenide materials for the H2-evolving photocathodes is still performed resulting in ever more complex and highly tunable compositions, like Cu–In–Ga–Se–S[404] and Zn–Cu– In–Ga–Se [405]. A combination of a photocathode and a photoactive anode allows constructing a water splitting system functioning without any externally applied bias. For example, a titania oxygen-evolving photoanode can be combined with a hydrogen- evolving photocathode comprised of the visible-light-sensitive composite of Zn phthalocyanine and fullerene C60 decorated with Pt NPs (Fig. 2.24b) [406]. A photoelectrochemical system for the H2 evolution with no external bias was assembled from a CuGaS2/RGO photocathode and a BiVO4/CoOx photoanode [297]. The RGO layer was deposited by the direct photocatalytic reduction of graphene oxide on the surface of CuGaS2 photocathode enhancing dramatically its photoresponse to the visible light. New “exotic” inorganic photocatalysts. Some new photocatalysts for water reduction were recently reported that can be referred to as “exotic” because of their quite rare (or otherwise unreported) photoactivity in the redox-processes. Many of such new compounds were only tested as bulk (microcrystalline) powders to date, but are nevertheless discussed here because a strong enhancement of the 2.6 Emerging Semiconductor Photocatalysts … 77

II Fig. 2.24 Schemes of the photoelectrochemical water splitting using Au/CuInS2 (a) and Zn phthalocyanine (ZnPc)/fullerene C60/Pt (b) heterostructured photocathodes. Reprinted with permissions from Refs. [403](a) and [406]. Copyright (2016) American Chemical Society (a) and The Royal Society of Chemistry (b) photocatalytic activity can be anticipated for such compounds in the nanocrystalline state thus posing a challenge for further studies. For example, a recently discovered fibrous modification of the red phosphorus was reported to have photocatalytic properties in the water reduction [407]. The photoactivity of such materials can be enhanced considerably by decreasing the crystal dimensions via the growth restriction on the silica fibers or by an ultrasound treatment of the bulk material [407]. The red P was reported to have the CB potential around −0.25 V versus NHE at pH 0 which is sufficient for the water reduction, while the VB holes (EVB > 1.5 eV) can oxidize either water or a broad range of sacrificial electron donors [408]. A community of rare photocatalysts of hydrogen evolution was recently joined by silicon carbide [409–414] capable of the water reduction even in the absence of sacrificial donors, as well as by gallium nitride with a CB potential by 0.5 V more negative than the water reduction potential [415]. Such difference appeared to be sufficient to overcome the hydrogen evolution overvoltage and produce H2 from aqueous solutions of methanol and Na2S/Na2SO3 under the Vis-illumination without additional co-catalysts. Iron silicide b-FeSi2 revealed a photocatalytic activity in the hydrogen evolution from aqueous dithionite solutions even under the illumination with the near-IR light [416]. A Vis-sensitive photocatalyst of the H2 evolution from aqueous solutions of formic acid was prepared via the deposition of Pt NPs on the surface of polycrystalline Si [417]. The photocatalytic water reduction was also observed for macroporous silicon [418], mesoporous Si coupled to a noble-metal-free non-stoichiometric cobalt phosphate co-catalyst [419], and Si nanowires loaded with an iron phosphite co-catalyst [420]. Layered siloxene NSs (Fig. 2.25a) produced by a topotactical transformation of calcium silicide in water revealed photocatalytic properties in the water reduction without additional co-catalysts and sacrificial electron donors [421]. The siloxene is 78 2 Semiconductor-Based Photocatalytic Systems …

Fig. 2.25 A schematic molecular structure (a) and band structure (b) of siloxene. Reprinted and adapted with permissions from Ref. [421]. Copyright (2016) The Royal Society of Chemistry a direct bandgap semiconductor with a strong visible light harvesting capability and a suitable bandgap of 2.5 eV. The position of both CB and VB bands of the siloxene are respectively negative and positive enough for both water reduction and oxidation to occur on the photocatalyst surface under the photoexcitation with the visible light (Fig. 2.25b) [421]. Iron silicide was reported to be able to evolve hydrogen from aqueous solutions of sodium dithionite even when illuminated with the NIR light [416]. It is a unique semiconductor material because it combines a narrow band gap of 0.8 eV with a CB position of around −0.65 V (vs. NHE) which is unprecedently high for such a narrow-bandgap material. Titanium disilicide is a narrow-bandgap semiconductor that can harvest light over the entire visible spectral range and serve as a stable photocatalyst for the hydrogen production without additional sacriﬁcial donors [422]. Carbon NPs. Carbon NPs (CNPs) have recently shown a high promise for applications concerning with the light emission and absorption, including the photocatalytic solar light harvesting [169, 170]. The CNPs can be synthesized quite easily by the thermal decomposition of a single or mixed organic precursor at 200– 300 °C. The structure of CNPs is still a subject of discussions, as numerous XPS studies showed them to contain simultaneously aliphatic sp3-hybridized and aromatic sp2-hybridized carbon, sometimes also the amine- and pyridine-like nitrogen. The surface of CNPs is typically decorated with hydroxyl and carboxyl groups. The CNPs are often called carbon quantum dots (QDs), however, the usage of this term seems not to be justiﬁed as no information on the possibility and character of QSEs is available for CNPs. The CNPs can emit bright and broadband PL in the visible spectral range with a broad gamut—from blue to red, the fact determining the perspectives of the CNPs for luminescent bio-labeling. Similarly to the exact internal structure of the CNPs, the PL origins and mechanisms are also a subject of vivid discussions. A special class of CNPs is constituted by graphene quantum dots, that is, small pieces of graphene functionalized with oxygen-containing groups and N dopants that reveal a strong dependence of the spectral properties on their lateral size [169, 170]. The synthesis of CNPs typically requires very simple and available precursors making them an excellent candidate for the mass-scale photocatalytic applications. 2.6 Emerging Semiconductor Photocatalysts … 79

For example, the carboxylate-terminated 2–3-nm CNPs can be produced via the pyrolysis of sodium salt of ethylenediaminetetraacetic acid (EDTA) at 350 °C. The CNPs can be then coupled to Pt NPs and the CNP/Pt heterostructure and used as an efficient visible-light-sensitive photocatalyst of the H2 evolution from aqueous solutions of dihydronicotinamide adenine dinucleotide (NaDH) which is a convenient electron/proton sources frequently used in a biochemical research practice [423]. The CNPs produced by a hydrothermal treatment of flower pollens was successfully used to sensitize GCN NSs, the GCN/CNPs assembly evolving hydrogen under the illumination with the visible light from methanol/water mixtures with a Pt NP co-catalyst [424]. Pure CNPs produced from multi-wall carbon NTs were found to possess the capability of reducing water to hydrogen in water/methanol mixtures without additional co-catalysts [425]. In the presence of Pt NPs, the CNPs revealed a superior photoactivity as compared with the nanocrystalline TiO2 Evonik P25. The electrochemical etching of graphite electrodes is another convenient method of the CNPs production yielding stable suspensions of 4–5-nm CNPs. Such NPs are crystalline with a lattice regularity of 0.321 nm typical for the parental graphite indicating the CNPs to be small fragments of graphite stabilized by an outer shell of functional groups (mostly COOH as revealed by the XPS) [426]. The CNPs can be coupled with layered MoS2 into a material with a pronounced photoelectrochemical activity in the water reduction process [427]. The scalable and benign carbonization of vegetable raw materials (such as spinach, peas, and others) was reported to produce CNPs capable of strong adsorption on the surface of nanocrystalline titania [428]. The CNPs/TiO2 composites were tested as a photocatalyst of the hydrogen evolution from methanol/ water mixtures. Graphitic carbon nitride. The graphitic carbon nitride is probably one of the first artificially synthesized organic polymers but it emerged only relatively recently as a universal and visible-light sensitive photocatalyst with brilliant perspectives in the domains of the solar energy harvesting and the environmental photocatalysis [429– 433]. Similarly to graphite, GCN is formed by planar infinite single layers that have an aromatic character and are stacked by the van-der-Waals forces (pp-interactions) with an interplanar distance of around 0.34 nm. The term “carbon nitride” does not reflect exactly the structure of single layers as they are composed of heptazine (tris-s-triazine) heterocycles bounded through tertial amine N atoms into an infinite network with an intra-network periodicity of around 0.68 nm. Two alternative structures of the single layer carbon nitride (SLCN) are proposed, one describing its as a regular net-like polyheptazine (Fig. 2.26a), the other postulating that a SLCN is composed on infinite 2D polyheptazine ribbons bound together by numerous hydrogen bonds (Fig. 2.26b) [429–435]. The exact structure of SLCN still remains a subject of discussion, however, the second structure seems to be more realistic. It can be seen that the term “carbon nitride” is only a convenient approximation to describe the stacked multilayer polyheptazine structure 80 2 Semiconductor-Based Photocatalytic Systems …

Fig. 2.26 Alternative models of the single layer carbon nitride structure. Reprinted with permissions from Ref. [435]. Copyright (2014) American Chemical Society

because the composition of the bulk undoped material is close C3N4, however, this term does not reflect the exact chemical nature of this material, though used broadly by the historical reasons and convenience. GCN is a direct bandgap semiconductor with the VB and CB formed respectively by filled and vacant C2p and N2p orbitals. The bandgap of bulk GCN is around 2.7 eV and can vary by 0.1–0.2 eV as a result of differences in the synthesis, possible adventitious doping, and structural defects and, therefore, GCN absorbs the UV and a portion of the visible light in a range of k < 460 nm. Also, the GCN has a very “suitable” position of both CB and VB levels that are respectively negative and positive enough to allow the photogenerated charge carriers to participate in a variety of redox-processes, including the reduction and oxidation of water [429, 431–434]. Reported data on the exact position of CB and VB levels of GCN reveal some scatter, most probably due to the above-mentioned differences in the synthesis nuances and structural imperfection and, most often, can be found at around −1.0 and +1.7 V versus NHE [431–433]. The GCN has comparatively high thermal, chemical and photochemical stability as well as a low toxicity that distinguish this material from other inorganic semiconductors with similar Eg, ECB and EVB parameters. The bulk GCN can be produced in copious amounts from a variety of affordable precursors, such as melamine (1, 3, 5-triaminotriazine), dicyandiamide, and urea by the thermal treatment at 400– 600 °C. By introducing various heteroatomic additives GCN can be doped with P, S,B etc. on the stage of the material formation [436]. Alternatively, the GCN can be annealed or treated with aggressive oxidizing/reducing agents after the synthesis to vary the C/N ratio and thus to modify its spectral and photochemical properties [429, 431–434, 436]. Various supramolecular assemblies of triazine derivatives with other carbocyclic and heterocycles compounds can be used for the synthesis of GCN with tailored electron properties, for example with an increased hydropho- bicity or a more extended aromatic system [436–439]. By performing the thermal condensation of precursors in the presence of templates the GCN can be produced as porous solids. Finally, similarly to graphite, GCN can be exfoliated to form 2.6 Emerging Semiconductor Photocatalysts … 81 a-few-layer and single-layer moieties or, alternatively, subjected to a partial cleavage to produce GCN NPs. A comprehensive account on the photocatalytic systems for hydrogen evolution and other processes based on GCN can be found in reviews [430–432, 436, 440]. The compact GCN has a relatively low specific surface area of 5–15 m2/g thus limiting the efficiency of the charge carrier transfer from the photoexcited GCN to other components of the photocatalytic system. To increase the GCN surface area various “hard” (for example silica NPs [182, 441, 442]) and “soft” (organic polymers [443], sucrose [444]) templates are used resulting in the mesoporous GCN samples active in the photocatalytic H2 evolution. Using of uniform SiO2 micro-beads as a hard template allows producing hollow GCN spheres with a wall thickness of around 50 nm and an inner void of several hundred nm in diameter [182, 442] (Fig. 2.27a). The wall can then be modified with a co-catalyst, like layered MoS2 [442] or Pt NPs [182] as well as with a sensitizer, for example, CdS NPs [182], forming a complex photocatalyst for the solar hydrogen evolution. The “hollow sphere” architecture is favorable for the photocatalytic process as the co-catalyst is typically localized on the outer walls and thus the water reduction/water oxidation half-reactions can be spatially separated. Also, the hollow spheres (HSs) are well known for the ability of a more efficient light harvesting as a result of multiple light scattering in the inner voids of the spheres. Alternatively, the GCN can be grafted to the developed surface of a photochemically inert carrier, such as mesoporous [445] and macroporous silica [446]or a zeolite [447, 448]. As in the case of inorganic semiconductor photocatalysts of the water reduction, GCN typically requires a co-catalyst to efficiently evolve hydrogen. For example, metal NP (Pt [444, 449–451], Au [91, 92], Cu [452], Ni [453]) and alloys (Au–Pt [448], Ni-Pt [454]) co-catalysts can be attached to the GCN surface or formed in situ by the chemical/photochemical deposition. Similarly to titania, GCN can be coupled with Au NPs to form plasmonic photocatalysts for the H2 production [91, 92]. GCN can interact with MoS2 NSs [442, 455, 456], RGO [457], and multiwall

Fig. 2.27 TEM image (a) and EDX element mapping (b) of a GCN hollow sphere decorated with MoS2 nanosheets. c TEM image of a CdS-decorated GCN hollow sphere. Reprinted with permissions from Refs. [442](a) and [182](b). Copyright (2015–2016) Elsevier 82 2 Semiconductor-Based Photocatalytic Systems … carbon NTs [458–460] via the pp interactions forming visible-light-sensitive layered “dyads” for the hydrogen production. The assortment of sacrificial donors that can be applied in the GCN-based systems for H2 evolution is also quite broad [436] including traditional TEA [449–451, 453], carbon acids [448], hydrazine [454], and Na2S/Na2SO3 [180]. Similarly to other wide-bandgap semiconductors that can harvest only a part of the visible spectrum, GCN can be sensitized to longer wavelengths by organic dyes and narrower-bandgap semiconductor NPs. The interactions between organic dyes and GCN are especially strong due to the possibility of a multiple bonding between a dye and GCN via functional (carboxyl) groups present of the edges and outer planes of the GCN particles and via the pp-stacking between the extensive aromatic systems of the dye and outer sheets of the multilayer GCN particles. In this way, visible/ NIR-light-sensitive photocatalysts for the H2 evolution were produced by coupling GCN with eosin Y [461, 462], erythrosin B [463], thiazole orange [464], and Zn phthalocyanines (an apparent QY of around 3% was observed at 730 nm) [62]. As GCN has a band structure similar to that of cadmium sulfide it can be introduced into various binary heterostructures with inorganic semiconductors, where offsets between the CB and VB edges of the components favor to an efficient spatial separation of the photogenerated charge carriers [436]. For example, the photocatalytic activity of TiO2 NRs increases by an order of magnitude after the decoration with GCN NSs [179]. A large variety of oxide/metallate semiconductor photocatalysts can be produced by a solvothermal treatment in supercritical conditions. Introduction of GCN into the reaction mixtures offers a surface for the nucleation/deposition of oxide semiconductors and typically results in a decreased crystal size as compared to the conventional solvothermal synthesis. In this way, the heterostructures of GCN with C,-N-doped TiO2 [183], CuFe2O4 [449], CdS [180, 465], ZnIn2S4 [181] were produced with an enhanced photocatalytic activity in the H2 evolution as compared to both individual GCN and the inorganic component. A prolonged thermal treatment of GCN at 550–650 °C can also generate numerous lattice defects (“pinholes”) as a result of the residual ammonia elimination and splitting of the intra-layer bonds. The distortion of single layers results also in the expansion of the GCN and a partial exfoliation to NSs [466]. The formation of point defects, new surface edges and reduction in the GCN particles thickness typically result in an increase of the photocatalytic activity of this material in the water reduction [466]. The ultrasound-assisted exfoliation of bulk GCN in 2-propanol produces 2-nm thick NSs revealing an enhanced photocatalytic activity in the solar hydrogen production as compared to the compact g-C3N4 [467]. The exfoliation of GCN can be facilitated by the preliminary intercalation with sulfuric acid [468] in this case leading to the predominant (around 60 mass%) formation of SLCN. The SLCN showed a 3-times higher photoactivity than the non-treated GCN. The GCN nanoribbons around 2 lm in length, *200 nm in width and 3 nm thick were produced by “chemical scissors”, that is by the treatment with a mixture of concentrated HNO3/H2SO4 [469]. The molar C/N ratio decreases to *0.63 2.6 Emerging Semiconductor Photocatalysts … 83 versus 0.76 for the stoichiometric C3N4 indicating the formation of carbon vacancies that can act as charge traps retarding the electron-hole recombination. As a result, the nanoribboned GCN showed a 20 times higher photocatalytic activity in the H2 evolution as compared to the starting bulk material [469]. GCN can be effectively disintegrated by adding water to the bulk GCN mixed with the concentrated H2SO4 due to a strong exothermic effect (Fig. 2.28a) [450]. Following ultrasound-assisted exfoliation produces 2–3-nm thick NSs exhibiting a much higher photocatalytic activity in the H2 evolution from aqueous TEA solutions (Fig. 2.28b) [450]. Atomically thin carbon nitride nanomeshes can be produced by the solvothermal exfoliation of mesoporous GCN intercalated with 2-propanol [451]. Along with a highly developed surface area, the holey SLCN revealed a higher bandgap of 2.75 eV (as compared to 2.59 eV for the starting material) and a CB level by *0.5 eV more negative than that of mesoporous GCN. Due to these favorable factors, the SLCN nanomeshes showed an apparent QY of the H2 production of 5.1% at k > 420 nm, which is the highest reported for the exfoliated carbon nitrides [451]. Alternatively to the conventional polyheptazine-based GCN, formed by polyheptazine networks, a special attention is currently brought to polytriazine networks that can be produced by versatile synthetic approaches and additionally doped to obtain visible-light-harvesting photocatalysts of the hydrogen production [470–474]. Metal-organic frameworks (MOFs). The metal-organic frameworks can also be rated as an “emerging star” of the semiconductor photocatalysis with a number of reports on various redox-processes catalyzed by the photoexcited MOFs increasing drastically in recent years [475, 476]. MOFs are formed by metal-organic complex units linked by “bridge” bi- or tri-functional ligands into the 2D/3D continuous networks. Typical bridge ligands for the construction of the 2D and 3D frameworks are aromatic bi- and tri-carboxylic acids. MOFs combine a high light sensitivity with an almost unlimited versatility of possible building blocks and substituents that can alter/modify the MOF structure in a desirable manner. High-intensity ligand-to-metal (metal-to-ligand) electron transitions impart them with a strong

Fig. 2.28 a Schematic illustration of the GCN disintegration and exfoliation with H2O/H2SO4; b kinetic curves of the H2 evolution over the bulk and nanosheet GCN. Reprinted and adapted with permissions from Ref. [450]. Copyright (2015) The Royal Society of Chemistry 84 2 Semiconductor-Based Photocatalytic Systems … light harvesting capability, while the 2D/3D networks are favorable for the directed charge transport and form a system of regular pores that can be accessed by the water and sacrificial donors. The current state-of-the-art in the photocatalytic applications of MOFs is highlighted by comprehensive reviews [475, 476]. Similarly to the photocatalysis with inorganic semiconductors, no definite models were proposed allowing to predict a photocatalytic activity for a given MOF or a class of MOFs and therefore the quest for new photoactive MOFs is performed mostly empirically. As with photoactive inorganic semiconductors, there is a “club” of selected MOFs exhibiting a high photocatalytic activity and by this reason appearing most frequently in the researchers’ spotlight. One of such MOFs is UiO-66 formed by ZrO6 octahedra linked with p-dibenzoic acid and its derivatives (Fig. 2.29a) [477–480]. The UiO-66-type MOFs combine regular 3D porous structure, a high stability and sensitivity to the visible light. The introduction of functionalities into the bridge molecules opens a way of affecting the pore characteristics. For example, 3 nm Pt NPs can be introduced into the inner voids of UiO-66-NH2 MOF acting as a co-catalyst of the photocatalytic water reduction [480]. Such MOF/Pt heterostructure reveals a much higher photoactivity as compared to an analog with Pt NPs attached to the outer surface of the MOF grains. A spectral response of the UiO-66 MOF can be extended by the sensitization with dyes [483], CdS NPs [481, 482] and CdxZn1−xS NPs [484]. After modification with RGO, a CdS/UiO-66 assembly becomes a much more active photocatalyst of the H2 production from aqueous Na2S/Na2SO3 solutions as compared with the “classical” TiO2/CdS heterostructure (Fig. 2.29a). Similar 3D structures can be assembled using oxo-zirconium units with a tetracarboxylate-derived zinc porphyrine ZnTCPP (Fig. 2.30)[482]. The MOF contains a relatively large inner channel where a hydrogenase-biomimetic metal-organic co-catalyst can be placed. The assembly shows a photocatalytic activity in the solar H2 production from the aqueous ascorbic acid solutions [482].

Fig. 2.29 a Schematic structure of UiO-66 MOF; b rate of the H2 evolution using various photocatalysts. Reprinted with permissions from Refs. [483](a) and [479](b). Copyright (2014– 2015) The Royal Society of Chemistry 2.6 Emerging Semiconductor Photocatalysts … 85

Fig. 2.30 a Structural building units of Zr6O8(CO2)8(H2O)8, and b structural building unit of ZnTCPP. c Model MOF structure. d Structure of the co-catalyst. e Model structure of the MOF/ co-catalyst assembly. Color scheme: Zr, green; Zn, dark gray; C, light gray; O, red; N, blue; Fe, light green; S, yellow. Reprinted with permissions from Ref. [484]. Copyright (2014) The Royal Society of Chemistry

A key step in the photoprocess is a charge transfer from the singlet excited state of Zn porphyrinate bridges to the incorporated co-catalyst, where the subsequent proton reduction occurs. A popular family of photoactive MOFs includes a number of MIL-100 complexes formed by various central ions and tricarboxylic acids (like 1,3,5- benzenetricarbioxylic acid) [484, 485]. AFe3+-based MIL-100 has a broad spectral response in the visible range and can be used as a photocatalyst of the H2 evolution from water/CH3OH mixtures enhanced by the additional deposition of Pt NPs [484]. A La3+-based 3D MOF is a wide-bandgap compound with Eg = 3.7 eV that can be sensitized to the visible light by CdSe NPs [485]. The MOFs with a high sensitivity to the visible light and photocatalytic activity in the H2 evolution can be produced by using derivatives of azo-dyes [486] and rhodamines [487] as bridge ligands. Another face of the application of MOFs in the photocatalysis is in using them as precursors for the preparation of highly dispersed visible-light-sensitive H2-evolving materials, for example, C,H-doped iron oxides [488].

2.7 New-Generation Co-Catalysts for the Photocatalytic Hydrogen Production

The most reported semiconductor-based photocatalytic systems for the hydrogen production contain obligatorily a co-catalyst, typically, a noble metal (Pt, Pd, Rh) that is characterized by a much lower water reduction over-voltage as compared with the semiconductor photocatalysts. The co-catalyst accepts the photogenerated CB electrons from the photocatalyst and then participates in the water/protons reduction, atomic hydrogen accumulation and recombination to the molecular hydrogen. In the photoelectrocatalytic systems where the photoexcitation of a 86 2 Semiconductor-Based Photocatalytic Systems … semiconductor anode results in the oxidation of water with simultaneous hydrogen evolution on a counter electrode the photoanode is typically coupled with appropriate oxygen evolution catalysts [489–491]. Recent studies showed that the “conventional” noble metal co-catalysts can be successfully substituted with much less expensive and more available compounds, in particular, iron group metal NPs [492], molybdenum sulfide [426, 455, 456, 493– 504], oxide [505], and nitride [506], nickel oxide [507, 508] and hydroxide [408, 509], nickel sulfide [510, 511], nitride [512] and phosphide [513], tungsten carbide [514], as well as with cobalt oxide [515, 516], phosphate [176], and phosphide [517]. For example, in the CdS-based systems, the introduction of mere 0.2 wt.% MoS2 results in a rather drastic 36-fold increase of the rate of photocatalytic hydrogen evolution from aqueous lactic acid solutions [493]. The catalytic activity of MoS2 in the water reduction is typically associated with an efficient charge carriers separation between CdS and MoS2, as well as with a well-known capability of molybdenum disulfide to the hydrogen activation. The catalytic activity of MoS2 can be boosted by the exfoliation of bulk layered material into single or a-few-layer NSs [502, 518]. As the layers of bulk MoS2 are kept together by relatively weak van-der-Waals forces the exfoliation can be achieved by a conventional ultrasound treatment. The sonication of bulk MoS2 in the presence of CdS NRs results in a composite photocatalyst of the H2 evolution from aqueous lactic acid solutions that is far more active than bare CdS NRs or a mechanical NR mixture with unexfoliated molybdenum disulfide (Fig. 2.31a) [502]. Apart from the most stable semiconducting 2H-phase of MoS2, it can form an allotropic 1T modification that is characterized by the metallic conductance. The exfoliation of 1T-MoS2 was found to produce an even more efficient co-catalyst of the hydrogen evolution for the CdS NR photocatalyst, than conventional exfoliated 2H-MoS2 [503] (Fig. 2.31b). A similar catalytic activity in the H2 evolution in the presence of nanoparticulate CdS was observed for ultrathin composite cobalt selenide/reduced graphene oxide NSs that revealed a half-metallic character [519]. The in situ photodeposition of Ni, Co, or Cu NPs on the surface of Cd0.4Zn0.6S resulted in a 5-fold acceleration of photocatalytic hydrogen evolution from aqueous Na2S/Na2SO3 solutions [492]. The modification of nanocrystalline Cd0.2Zn0.8S with 3 wt.% CuS increases strongly the photocatalytic water reduction QY up to *37% at k = 420 nm [520]. The catalytic properties of copper sulfide were also observed in a photocatalytic hydrogen production system based on a visible-light- sensitive CuO/Al2O3 composite [521]. Nickel sulfide can act as a co-catalyst of the photocatalytic water reduction on the nanocrystalline CdS [522, 523], Cd0.5Zn0.5S [524], and a ZnS1−x−0.5yOx(OH)y/ZnO heterostructure [525]. Cobalt sulfide revealed a catalytic activity in the photocatalytic water reduction on GCN [526]. Nickel phosphide Ni2P NPs were proven to act as a “universal” co-catalyst of the hydrogen evolution for a broad range of semiconductor photocatalysts including TiO2, CdS and GCN (Fig. 2.31c) and different sacrificial agents, such as lactic acid, TEA, and methanol [527]. 2.7 New-Generation Co-Catalysts for the Photocatalytic … 87

Fig. 2.31 a, b Rate of the photocatalytic hydrogen production in the presence of CdS NRs, a mechanical mixture of bulk 2H-MoS2 (BM) and CdS NRs and a composite of CdS NRs with ultrathin 2H-MoS2 NSs (UM) and 1T-MoS2 NSs produced by the sonication. c Rate of the photocatalytic H2 production from aqueous solutions of sacriﬁcial donors with the participation of nanocrystalline TiO2, CdS, GCN and their composites with Ni2P NPs. Reprinted with permission from Refs. [502](a), [503](b), and [527](c). Copyright (2016) The Royal Society of Chemistry

Various biological molecules, for example, hydrogenase [63, 528] as well as iron complexes mimicking an active center of the hydrogenase [482, 529], have also good perspectives as co-catalysts in the photocatalytic systems for the H2 generation. A recent account on the hybrid artiﬁcial photosynthesis systems based on the semiconductor light harvesters and biomimetic metal-complex co-catalysts can be found in [530]. The reduced graphene oxide is often used as a hydrogen evolution co-catalyst, however, it can play other unique roles, such, for example, as a conductive 2D “mat” for assembling of various components of a photocatalytic system [531]. GO is typically produced by the ultrasound-assisted exfoliation of layered graphite oxide which, in turn, can be obtained by treating graphite with strong oxidizing agents such as KMnO4/H2SO4 or KClO3/HNO3. The single (or a-few-layer) GO derived by the exfoliation can then be reduced by a variety of agents, like NaBH4 and hydrazine, or via a photochemical/electrochemical/microwave treatment [531]. 88 2 Semiconductor-Based Photocatalytic Systems …

RGO bears residual functional groups such as –COH and –COOH that can interact with semiconductor NPs and metals bringing then together on the surface of RGO NSs. At the same time, the RGO sheets typically possess a good conductivity close to that of pristine graphite. Thus RGO enables a good electric contact among the components of a photocatalytic system that are assembled but spatially separated on the 2D RGO mat. The intermediatory role of RGO NSs ensuring the electric contact between spatially separated semiconductor and metal NPs was observed for TiO2/RGO/Ag [532] and TiSi2/RGO/RuO2 [422]. Thiolated RGO NSs can strongly interact both with CdS NPs and dendritic Pt nanocrystals assembling them into a photocatalytic system for the hydrogen production from aqueous lactic acid solutions [533]. RGO can also substitute noble metal co-catalysts revealing in some cases an appealing catalytic activity for the hydrogen generation. For example, the rate of hydrogen evolution in binary TiO2 NS/RGO NS heterostructure is more than 40 times higher than for the individual titania [534]. A catalytic effect of RGO was also observed in the photocatalytic H2 evolution systems based on nanocrystalline CdS [500, 535, 536], Zn-doped CIS NPs [500], ZnIn2S4 [292], Cu2O[537], CaIn2O4 [538], BiPO4 [539], Bi2WO6 [540], K4Nb6O17 [541], and GCN [457].

2.8 Stoichiometric Water Splitting Under the Illumination with the Visible Light

One of the conditions for the successful total water splitting consists in the spatial separation of the water reduction and water oxidation processes necessary to prevent reverse reactions between hydrogen and oxygen [2]. This challenge is typically addressed by using two separate electrode cells connected with a membrane. In the ﬁrst cell the water reduction proceeds at the expense of a mediator oxidation, then the oxidized mediator diffuses through the membrane into the second cell where it participates in the water oxidation to O2 (Fig. 2.32).

Fig. 2.32 Scheme of a photocatalytic system for total water splitting based on tantalum oxynitride, tungsten À À oxide and I IO3 mediator couple 2.8 Stoichiometric Water Splitting Under the Illumination … 89

Such coupled systems are often compared to the Z-scheme of the natural photosynthesis [2, 137, 187–189, 542, 543]. The scheme is very convenient as it allows to design separately the cathode and anode cells and then combine them by an 3+ 2+ À À appropriate redox-pair, for example, Fe /Fe or I IO3 . The photocatalytic 3+ 5+ hydrogen evolution can occur on the surface of SrTiO3/Pt doped with Cr /Ta at À the expense of the iodide ions oxidation to IO3 . The oxygen is evolved in a − complementary cell with the regeneration of I on the surface of a WO3/Pt heterostructure. Such system showed an apparent QY of around 0.1% at 420 nm [544, 545]. In the same manner, a hydrogen evolution cell based on a visible-light-sensitive GaInP2/Pt composite was coupled with an oxygen evolution cell based on AgCl sensitized with silver bromide [546]. A number of cathode/anode cells connected by the iodide/iodate mediator was proposed, for example TiO2(anatase)/Pt (H2 evolution)—TiO2(rutile)/Pt (O2 evolution) [547], TaON/Pt (H2 evolution)—WO3/Pt (O2 evolution) [548], ATaO2N/Pt, A = Ca, Sr, Ba (H2)—WO3/Pt (O2)[549], TaON/Pt (H2)—TaON-RuO2 (O2)[550], TaON/ZrO2/Pt,Ru (H2)—WO3/Pt (O2)[241], sensitized layered H4Nb6O17/Pt (H2) —WO3/Pt (O2)[551], BiVO4 (H2)—Rh-doped SrTiO3/Ru (O2)[198, 552], BaZrO3/BaTaO2N(H2)—WO3/TiOx (O2)[553]. In the most part of such systems, ﬁ À the selective water oxidation takes place as a result of ef cient IO3 adsorption on the semiconductor surface despite the presence of a large excess of I− [547]. In a cell, where the photocatalytic water reduction occurs on the surface of rhodium-doped SrTiO3/Pt and the water oxidation proceeds with the participation 3+ 2+ of BiVO4,Fe /Fe redox couple is used as a mediator [554, 555]. The system performs under the Vis-illumination (k < 500 nm) with an apparent QY of 0.3% at 440 nm. The Fe3+/Fe2+ couple was also employed as an electron carrier in a total water splitting system based on Rh-doped SrTiO3 and WO3 [556]. Alternatively, the Z-scheme can be realized by combining an H2-evolving and an O2-evolving cell by a conducting bridge. For example, the RhCrOx-loaded LaMg1/ 3Ta2/3O2N crystals, acting as an H2 evolution photocatalyst can be combined with Mo-doped BiVO4 crystals as an oxygen-evolving photocatalyst on a shared gold substrate (Fig. 2.33a) [557]. Both components are capable of absorbing the visible light. Such composite exhibits a 5-times higher photocatalytic activity in the water splitting than a combination of corresponding suspensions. The Au substrate acts as a transport layer enabling ﬁlling of the tantalate holes with the electrons photogenerated in bismuth vanadate [557]. A similar role of an electron mediator in a Z-scheme photocatalyst can be played by the photoreduced GO (Fig. 2.33b) [558]. As opposite to the RGO, produced by a conventional reduction with hydrazine, the photoreduced RGO showed a much more expressed hydrophilic character binding strongly both to an H2-evolving photocatalyst (Rh-doped SrTiO3 decorated with Ru NPs) and to an O2-evolving photocatalyst (BiVO4). By using spatially organized semiconductor materials—NRs, NTs, layered substances, etc. a spatial separation of the water reduction and water oxidation sites can be achieved within a single photocatalyst, and in such a way a short-circuited 90 2 Semiconductor-Based Photocatalytic Systems …

Fig. 2.33 a Schematic band diagram of the (RhCrOx/LaMg1/3Ta2/3O2N)/Au/BiVO4:Mo photocatalyst; b scheme of the water splitting in a photocatalytic system based on Rh:SrTiO3/Ru and BiVO4 coupled via RGO sheets. Reprinted with permissions from Refs. [502](a) and [558](b). Copyright (2011, 2016) American Chemical Society photo-electrochemical cell can be designed. For example, in a system based on a composite of titania NTs with Pt NPs [559, 560] the water reduction and oxidation half-reactions occur on different components—on Pt NPs and TiO2 NTs, respectively, allowing to reach a water splitting QY of 0.6% [559]. In the case of a Ni/NiOx/In0.9Ni0.1TaO4 heterostructure, the co-catalyst (Ni/NiOx) particles act as a hydrogen evolution cathode, while the surface of the Ni-doped indium tantalate—as an oxygen evolution anode [561]. A QY of the total water splitting in such system reached 0.66%. A separation of the cathodic and anodic water cleavage processes was also realized in the case of a composite photo-catalyst produced by the intercalation of Fe2O3 NPs into the interlayer voids of HTiNb(Ta)O5 [562]. The hydrogen evolution and oxygen evolution processes become naturally separated in the case of two-sided nanocrystalline titania films produced by the magnetron sputtering with one side of the film decorated by Pt NPs. The photocatalytic hydrogen evolution from aqueous H2SO4 solution occurs on the TiO2/Pt side of the film under the Vis-illumination, while on another side the water oxidation in the presence of NaOH takes place [221]. By oxidizing titanium foil in the presence of water vapors and NaF thin films of F-doped titania were prepared, exhibiting a photoelectrochemical activity in the total Vis-light-driven water splitting [563]. The photoelectrochemical water splitting under the Vis-illumination was also observed in the case of nanocrystalline TiO2 films etched in HF solution [564] and was attributed to the formation of Vis-light-absorbing titanium oxyfluoride species on the film surface. Along with the above-described complex systems where the spatial separation of cathodic and anodic processes is organized intentionally, alternative semiconductor compounds capable of simultaneous water oxidation and reduction under the Vis-excitation are explored continuously. One of the first “universal” photocatalysts of the kind was a solid solutions of gallium nitride and zinc oxide (Ga1−xZnx)(N1−xOx) with Eg varying in a range of 2.58–2.76 eV. Gallium-zinc oxy-nitride exhibited a high 2.8 Stoichiometric Water Splitting Under the Illumination … 91 photochemical stability and, in the presence of RuO2, acted as a photocatalyst of the water splitting to H2 and O2 under the Vis-illumination [565]. A solid solution of bismuth and yttrium tungstates, BiYWO6 (Eg = 2.71 eV) was also used as a total water splitting photocatalyst that showed an apparent QY of 0.17% at 420 nm in the presence of RuO2 and Pt/Cr2O3 co-catalysts [566]. Porous films of BiVO4 produced by the thermal decomposition of vanadium oxyacetyl acetonate [567] and BiVO4/ Cu2O heterostructures [568] showed a photocatalytic activity in the water splitting under external bias [567]. The photoelectrochemical water splitting was realized on the surface of nanocrystalline hematite a-Fe2O3 films synthesized by the thermal decomposition of ferrocene or iron pentacarbonyl [569, 570]. A visible-light-sensitive photoelectrochemical cell for the total water splitting was tested [571], where a RuII bipyridyl complex served simultaneously as an “antennae” and as a “bridge” connecting to the titania NPs via phosphate groups and simultaneously—to the IrO2 NPs via COOH groups. The sensitizer photoexcitation results in the electron transfer through the following chain (Fig. 2.34): water molecules (water oxidation to O2) ! IrO2 NPs ! sensitizer ! TiO2 NPs ! Pt cathode ! water molecules (water reduction to H2). Three new types of total water-splitting photocatalysts functioning under the UV III and a portion of visible light were proposed in [572–574]: BiMNbO7 (M = Al , III III V V V V Ga ,In ), InMO4 (M = Nb ,Ta) and BiMO4 (M = Nb ,Ta)withEg in a range of 2.4–2.7 eV. The photoactivity of the compounds increases considerably in the presence of NiO or Pt co-catalysts. The stoichiometric water splitting under the Vis-illumination was also observed in the presence of CaTaO2N perovskite [575], complex solid solutions In–Ni–Ta–O–N[576] and Bi–Y–V–O[577], as well as gallium borate Ga4B2O9 [385]. A feasibility of the photocatalytic decomposition of water confined in an inner volume of the single-wall carbon NTs was shown in [578]. Illumination of the NTs results in the evolution of a gaseous mixture with 80% hydrogen fraction. Currently, studies are underway aimed at the search and development of new semiconductor materials capable of acting as photocatalysts of the stoichiometric water splitting, for example, GCN [579, 580], GCN/TiO2 nanoheterostructures [176], composites of the CNPs with BiVO4 [581], etc. Concluding the discussion of the large massif of experimental data on the photocatalytic systems for hydrogen production based on nanocrystalline semiconductor materials, we outline very generally the principal directions, where the highest efforts are currently applied and where a future progress could be expected. New principles of functioning of the photocatalytic systems can open rich and unexpected directions of progress. For example, utilization of the quantum size effects allowed to create much more efficient systems for the solar water splitting based on conventional semiconductor materials, to design highly efficient nanoheterostructures based on the same materials but in different phase compositions and grain size, and to “invoke life” into some semiconductor materials that are passive in the bulk form but reveal pronounced photocatalytic properties in the 92 2 Semiconductor-Based Photocatalytic Systems …

Fig. 2.34 Scheme of a photoelectrochemical system for the stoichiometric water splitting based on a heterostructure of TiIV and IrIV oxides and a visible-light-sensitive RuII complex (S). Details of the system—in [571]

water reduction/splitting when introduced in a nanocrystalline form. Also, the phenomenon of plasmonic photocatalysis is a vivid example of the new principle of light energy harvesting making a great impact on the development of semiconductor-based photocatalytic light-harvesting systems. New photocatalysts should be continuously searched for, in particular, among the available and Earth-abundant materials. The example of graphitic carbon nitride, that was known since the middle of 19th century but discovered as an excellent photocatalyst only at the end of 20th century, shows that new solutions for the challenges of the solar light harvesting can be just before our eyes and wait to be discovered and realized. Very high expectations are currently associated with inexpensive and abundant materials based on carbon NPs, that can be produced from a variety of available natural sources, as well as with ternary and more complex metal chalcogenides based on broadly available copper, tin, zinc and other elements, that can act as excellent harvesters of the visible and near IR solar irradiation. New co-catalysts are continuously discovered and such materials can enhance dramatically the performance of conventional semiconductor-based photocatalytic systems as well as to reduce a need for expensive noble and platinum-group metals. Finally, new sacriﬁcial donors derived from sustainable sources, such as biomass, when coupled to the above-discussed beneﬁts of new photocatalysts and co-catalyst can make the photocatalytic water splitting a really competitive and lucrative process and assist to its broad implementation in our everyday life. This road should be paved by simultaneous development of the theoretical backgrounds of the solar-light-induced water splitting and predictive modelling of the most optimized designs and constructions of the photochemical reactors and water splitting solar cells as well as their operational regimes [582]. References 93

References

1. Calvert JG, Pitts, JN Jr (1966) Photochemistry. Wiley, New-York, London, Sydney 2. Maeda K, Domen K (2007) New non-oxide photocatalysts designed for overall water splitting under visible light. J Phys Chem C 111:7851–7861. doi:10.1021/jp070911w 3. van de Krol R, Liang Y, Schoonman J (2008) Solar hydrogen production with nanostructured metal oxides. J Mater Chem 18:2311–2320. doi:10.1039/B718969A 4. Babu VJ, Vempati S, Uyar T, Ramakrishna S (2015) Review of one-dimensional and two-dimensional nanostructured materials for hydrogen generation. Phys Chem Chem Phys 17:2960–2986. doi:10.1039/c4cp04245j 5. Walter MG, Warren EL, McKone JR et al (2010) Solar water splitting cells. Chem Rev 110:6446–6473. doi:10.1021/cr1002326 6. Kitano M, Hara M (2010) Heterogeneous photocatalytic cleavage of water. J Mater Chem 20:627–641. doi:10.1039/B910180B 7. Maeda K, Domen K (2010) Photocatalytic water splitting: recent progress and future challenges. J Phys Chem Lett 1:2655–2661. doi:10.1021/jz1007966 8. Jing D, Guo L, Zhao L et al (2010) Efﬁcient solar hydrogen production by photocatalytic water splitting: from fundamental study to pilot demonstration. Inter J Hydrogen En 35:7087–7097. doi:10.1016/j.ijhydene.2010.01.030 9. Maeda K (2011) Photocatalytic water splitting using semiconductor particles: history and recent developments. J Photochem Photobiol, C 12:237–268. doi:10.1016/j.jphotochemrev. 2011.07.001 10. Abe R (2010) Recent progress on photocatalytic and photoelectrochemical water splitting under visible light irradiation. J Photochem Photobiol C 11:179–209. doi:10.1016/j. jphotochemrev.2011.02.003 11. Moriya Y, Takata T, Domen K (2013) Recent progress in the development of (oxy)nitride photocatalysts for water splitting under visible-light irradiation. Coord Chem Rev 257:1957– 1969. doi:10.1016/j.ccr.2013.01.021 12. Du P, Eisenberg R (2012) Catalysts made of earth-abundant elements (Co, Ni, Fe) for water splitting: recent progress and future challenges. En Environ Sci 5:6012–6021. doi:10.1039/ C2EE03250C 13. Warren SC, Thimsen E (2012) Plasmonic solar water splitting. En Environ Sci 5:5133–5146. doi:10.1039/C1EE02875H 14. Valdes A, Brillet J, Grätzel M et al (2012) Solar hydrogen production with semiconductor metal oxides: new directions in experiment and theory. Phys Chem Chem Phys 14:49–70. doi:10.1039/C1CP23212F 15. Zhou H, Qu Y, Zeid T, Duan X (2012) Towards highly efﬁcient photocatalysts using semiconductor nanoarchitectures. En Environ Sci 5:6732–6743. doi:10.1039/C2EE03447F 16. Wang P, Huang B, Dai Y, Whangbo MH (2012) Plasmonic photocatalysts: harvesting visible light with noble metal nanoparticles. Phys Chem Chem Phys 14:9813–9825. doi:10. 1039/C2CP40823F 17. Horiuchi Y, Toyao T, Takeuchi M et al (2013) Recent advances in visible-light-responsive photocatalysts for hydrogen production and solar energy conversion—from semiconducting TiO2 to MOF/PCP photocatalysts. Phys Chem Chem Phys 15:13243–13253. doi:10.1039/ C3CP51427G 18. Prevot MS, Sivula K (2013) Photoelectrochemical tandem cells for solar water splitting. J Phys Chem C 117:17879–17893. doi:10.1021/jp405291g 19. Zhang K, Guo L (2013) Metal sulphide semiconductors for photocatalytic hydrogen production. Catal Sci Technol 3:1672–1690. doi:10.1039/C3CY00018D 20. Osterloh FE (2013) Inorganic nanostructures for photoelectrochemical and photocatalytic water splitting. Chem Soc Rev 42:2294–2320. doi:10.1039/C2CS35266D 94 2 Semiconductor-Based Photocatalytic Systems …

21. Fresno F, Portela R, Suarez S, Coronado JM (2014) Photocatalytic materials: recent achievements and near future trends. J Mater Chem A 2:2863–2884. doi:10.1039/ C3TA13793G 22. Zhang P, Zhang J, Gong J (2014) Tantalum-based semiconductors for solar water splitting. Chem Soc Rev 43:4395–4422. doi:10.1039/C3CS60438A 23. Ran J, Zhang J, Yu J et al (2014) Earth-abundant cocatalysts for semiconductor-based photocatalytic water splitting. Chem Soc Rev 43:7787–7812. doi:10.1039/C3CS60425J 24. Huang ZF, Pan L, Zou JJ et al (2014) Nanostructured bismuth vanadate-based materials for solar-energy-driven water oxidation: a review on recent progress. Nanoscale 6:14044– 14063. doi:10.1039/C4NR05245E 25. Ida S, Ishihara T (2014) Recent progress in two-dimensional oxide photocatalysts for water splitting. J Phys Chem Lett 5:2533–2542. doi:10.1021/jz5010957 26. Hisatomi T, Kubota J, Domen K (2014) Recent advances in semiconductors for photocatalytic and photoelectrochemical water splitting. Chem Soc Rev 43:7520–7535. doi:10.1039/C3CS60378D 27. Vaneski A, Schneider J, Susha AS, Rogach AL (2014) Colloidal hybrid heterostructures based on II–VI semiconductor nanocrystals for photocatalytic hydrogen generation. J Photochem Photobiol C 19:52–61. doi:10.1016/j.jphotochemrev.2013.12.001 28. Zong X, Wang L (2014) Ion-exchangeable semiconductor materials for visible light-induced photocatalysis. J Photochem Photobiol C 18:32–49. doi:10.1016/j.jphotochemrev.2013.10. 001 29. Chen X, Liu L, Huang F (2015) Black titanium dioxide (TiO2) nanomaterials. Chem Soc Rev 44:1861–1885. doi:10.1039/C4CS00330F 30. Moniz SJA, Shevlin SA, Martin DJ et al (2015) Visible-light driven heterojunction photocatalysts for water splitting—a critical review. En Environ Sci 8:731–759. doi:10. 1039/C4EE03271C 31. Wang M, Han K, Zhang S, Sun L (2015) Integration of organometallic complexes with semiconductors and other nanomaterials for photocatalytic H2 production. Coord Chem Rev 287:1–14. doi:10.1016/j.ccr.2014.12.005 32. Gholipour MR, Dinh CT, Beland F, Do TO (2015) Nanocomposite heterojunctions as sunlight-driven photocatalysts for hydrogen production from water splitting. Nanoscale 7:8187–8208. doi:10.1039/C4NR07224C 33. Tachibana Y, Vayssieres L, Durrant JR (2012) Artiﬁcial photosynthesis for solar water-splitting. Nat Photonics 6:511–518. doi:10.1038/nphoton.2012.175 34. Zhou N, Lopez-Puente V, Wang Q et al (2015) Plasmon-enhanced light harvesting: applications in enhanced photocatalysis, photodynamic therapy and photovoltaics. RSC Adv 5:29076–29097. doi:10.1039/C5RA01819F 35. Ghirardi ML, Dubini A, Yu J, Maness PC (2009) Photobiological hydrogen-producing systems. Chem Soc Rev 38:52–61. doi:10.1039/B718939G 36. Szacilowski K, Macyk W, Drzewecka-Matuszek A et al (2005) Bioinorganic photochemistry: frontiers and mechanisms. Chem Rev 105:2647–2694. doi:10.1021/cr030707e 37. Allakhverdiev SI, Kreslavski VD, Thavasi V et al (2009) Hydrogen photoproduction by use of photosynthetic organisms and biomimetic systems. Photochem Photobiol Sci 8:148–156. doi:10.1039/B814932A 38. Zhang X, Peng T, Song S (2016) Recent advances in dye-sensitized semiconductor systems for photocatalytic hydrogen production. J Mater Chem A 4:2365–2402. doi:10.1039/ c5ta08939e 39. Nada AA, Hamed HA, Barakat MH et al (2008) Enhancement of photocatalytic hydrogen 2+ production rate using photosensitized TiO2/RuO2-MV . Inter J Hydrogen En 33:3264– 3269. doi:10.1016/j.ijhydene.2008.04.027 40. Jin Z, Zhang X, Lu G, Li S (2006) Improved quantum yield for photocatalytic hydrogen generation under visible light irradiation over eosin sensitized TiO2–Investigation of different noble metal loading. J Mol Catal A 259:275–280. doi:10.1016/j.molcata.2006.06.035 References 95

41. Tiwari A, Mondal I, Pal U (2015) Visible light induced hydrogen production over thiophenothiazine-based dye sensitized TiO2 photocatalyst in neutral water. RSC Adv 5:31415–31421. doi:10.1039/C5RA03039K 42. Lee J, Kwak J, Ko KC et al (2012) Phenothiazine-based organic dyes with two anchoring groups on TiO2 for highly efficient visible light-induced water splitting. Chem Commun 48:11431–11433. doi:10.1039/C2CC36501D 43. Tiwari A, Pal U (2015) Effect of donor-donor-p-acceptor architecture of triphenylamine-based organic sensitizers over TiO2 photocatalysts for visible-light-driven hydrogen production. Inter J Hydrogen En 40:9069–9079. doi:10.1016/j.ijhydene.2015.05.101 44. Chen S, Li Y, Wang C (2015) Enhancement of visible-light-driven photocatalytic H2 evolution from water over g-C3N4 through combination with perylene diimide aggregates. RSC Adv 5:15880–15885. doi:10.1016/j.apcata.2015.03.026 45. Zhang J, Du P, Schneider J et al (2007) Photogeneration of hydrogen from water using an integrated system based on TiO2 and platinum(II) diimine dithiolate sensitizers. J Am Chem Soc 129:7726–7727. doi:10.1021/ja071789h 46. Astuti Y, Palomares E, Haque SA, Durrant JR (2005) Triplet state photosensitization of nanocrystalline metal oxide electrodes by zinc-substituted cytochrome c: application to hydrogen evolution. J Am Chem Soc 127:15120–15126. doi:10.1021/ja0533444 47. Bala S, Mondal I, Goswami A et al (2014) Synthesis, crystal structure and optical properties of a naphthylbisimide-Ni complex: a framework on TiO2 for visible light H2 production. Dalton Trans 43:15704–15707. doi:10.1039/C4DT02006E 48. Li Q, Jin Z, Peng Z et al (2007) High-efficient photocatalytic hydrogen evolution on Eosin Y-sensitized Ti − MCM41 zeolite under visible-light irradiation. J Phys Chem C 111:8237– 8241. doi:10.1021/jp068703b 49. Li Q, Lu G (2007) Visible-light driven photocatalytic hydrogen generation on Eosin Y-sensitized Pt-loaded nanotube Na2Ti2O4(OH)2. J Mol Catal A 266:75–79. doi:10.1016/j. molcata.2006.10.047 50. Li Y, Xie C, Peng S et al (2008) Eosin Y-sensitized nitrogen-doped TiO2 for efficient visible light photocatalytic hydrogen evolution. J Mol Catal A 282:117–123. doi:10.1016/j.molcata. 2007.12.005 51. Abe R, Sayama K, Arakawa H (2004) Dye-sensitized photocatalysts for efficient hydrogen production from aqueous I− solution under visible light irradiation. J Photochem Photobiol A 166:115–122. doi:10.1016/j.jphotochem.2004.04.031 52. Puangpetch T, Sommakettarin P, Chavadej S, Sreethawong T (2010) Hydrogen production from water splitting over Eosin Y-sensitized mesoporous-assembled perovskite titanate nanocrystal photocatalysts under visible light irradiation. Inter J Hydrogen En 35:12428– 12442. doi:10.1016/j.ijhydene.2010.08.138 53. Zhang N, Shi J, Niu F et al (2015) A cocatalyst-free Eosin Y-sensitized p-type of Co3O4 quantum dot for highly efficient and stable visible-light-driven water reduction and hydrogen production. Phys Chem Chem Phys 17:21397–21400. doi:10.1039/c5cp02983j 54. Ikeda S, Abe C, Torimoto T, Ohtani B (2003) Photochemical hydrogen evolution from aqueous triethanolamine solutions sensitized by binaphthol-modified titanium(IV) oxide under visible-light irradiation. J Photochem Photobiol A 160:61–67. doi:10.1016/S1010- 6030(03)00222-3 55. Fu N, Lu G (2009) Hydrogen evolution over heteropoly blue-sensitized Pt/TiO2 under visible light irradiation. Catal Lett 127:319–322. doi:10.1007/s10562-008-9681-4 56. Fu N, Lu G (2009) Photo-catalytic H2 evolution over a series of Keggin-structure heteropoly blue sensitized Pt/TiO2 under visible light irradiation. Appl Surf Sci 255:4378–4383. doi:10. 1016/j.apsusc.2008.11.056 57. Kuchmiy SY, Korzhak AV, Guba NF et al (1995) Sensitization of cadmium sulfide by cyanine dyes in the photocatalytic production of hydrogen. Theoret Experim Chem 31:370– 374. doi:10.1007/BF00531245 96 2 Semiconductor-Based Photocatalytic Systems …

58. Xu J, Li Y, Peng S et al (2013) Eosin Y-sensitized graphitic carbon nitride fabricated by heating urea for visible light photocatalytic hydrogen evolution: the effect of the pyrolysis temperature of urea. Phys Chem Chem Phys 15:7657–7665. doi:10.1039/C3CP44687E 59. Wang Y, Hong J, Zhang W, Xu R (2013) Carbon nitride nanosheets for photocatalytic hydrogen evolution: remarkably enhanced activity by dye sensitization. Catal Sci Technol 3:1703–1711. doi:10.1039/C3CY20836B 60. Xu J, Li Y, Peng S (2015) Photocatalytic hydrogen evolution over erythrosin B-sensitized graphitic carbon nitride with in situ grown molybdenum sulfide cocatalyst. Inter J Hydrogen En 40:353–362. doi:10.1016/j.ijhydene.2014.10.150 61. Takanabe K, Kamata K, Wang X et al (2010) Photocatalytic hydrogen evolution on dye-sensitized mesoporous carbon nitride photocatalyst with magnesium phthalocyanine. Phys Chem Chem Phys 12:13020–13025. doi:10.1039/C0CP00611D 62. Song S, Guo Y, Peng T et al (2016) Effects of the symmetry and carboxyl anchoring group of zinc phthalocyanine derivatives on g-C3N4 for photosensitized H2 production. RSC Adv 6:77366–77374. doi:10.1039/c6ra15890k 63. Reisner E, Fontecilla-Camps JC, Armstrong FA (2009) Catalytic electrochemistry of a [NiFeSe]-hydrogenase on TiO2 and demonstration of its suitability for visible-light driven H2 production. Chem Commun 550–552. doi:10.1039/B817371K 64. Peng T, Ke D, Cai P et al (2008) Influence of different ruthenium(II) bipyridyl complex on the photocatalytic H2 evolution over TiO2 nanoparticles with mesostructures. J Power Sources 180:498–505. doi:10.1016/j.jpowsour.2008.02.002 65. Lakshminarasimhan N, Bae E, Choi W (2007) Enhanced photocatalytic production of H2 on mesoporous TiO2 prepared by template-free method: role of interparticle charge transfer. J Phys Chem C 111:15244–15250. doi:10.1021/jp0752724 66. Li J, E Y, Lian L, Ma W (2013) Visible light induced dye-sensitized photocatalytic hydrogen production over platinized TiO2 derived from decomposition of platinum complex precursor. Inter J Hydrogen En 38:10746–10753. doi:10.1016/j.ijhydene.2013.02.121 67. Maeda K, Eguchi M, Youngblood WJ, Mallouk TE (2008) Niobium oxide nanoscrolls as building blocks for dye-sensitized hydrogen production from water under visible light irradiation. Chem Mater 20:6770–6778. doi:10.1021/cm801807b 68. Maeda K, Eguchi M, Lee SH et al (2009) Photocatalytic hydrogen evolution from hexaniobate nanoscrolls and calcium niobate nanosheets sensitized by ruthenium(II) bipyridyl complexes. J Phys Chem C 113:7962–7969. doi:10.1021/jp900842e 69. Peng T, Dai K, Yi H et al (2008) Photosensitization of different ruthenium(II) complex dyes on TiO2 for photocatalytic H2 evolution under visible-light. Chem Phys Lett 460:216–219. doi:10.1016/j.cplett.2008.06.001 70. Ueno K, Misawa H (2013) Surface plasmon-enhanced photochemical reactions. J Photochem Photobiol C 15:31–52. doi:10.1016/j.jphotochemrev.2013.04.001 71. Valenti M, Jonsson MP, Biskos G et al (2016) Plasmonic nanoparticle-semiconductor composites for efficient solar water splitting. J Mater Chem A 4:17891–17912. doi:10.1039/ c6ta06405a 72. Clavero C (2014) Plasmon-induced hot-electron generation at nanoparticle/metal-oxide interfaces for photovoltaic and photocatalytic devices. Nat Photonics 8:95–103. doi:10.1038/ NPHOTON.2013.238 73. Nguyen BH, Nguyen VH (2015) Recent advances in research on plasmonic enhancement of photocatalysis. Adv Nat Sci Nanosci Nanotechnol 6:043001. doi:10.1088/2043-6262/6/4/ 043001 74. Mendez-Medrano MG, Kowalska E, Lehoux A et al (2016) Surface modification of TiO2 with Au nanoclusters for efficient water treatment and hydrogen generation under visible light. J Phys Chem C 120:25010–25022. doi:10.1021/acs.jpcc.6b06854 75. Fang J, Cao SW, Wang Z et al (2012) Mesoporous plasmonic Au–TiO2 nanocomposites for efficient visible-light-driven photocatalytic water reduction. Inter J Hydrogen En 37:17853– 17861. doi:10.1016/j.ijhydene.2012.09.023 References 97

76. Zhang X, Liu Y, Kang Z (2014) 3D branched ZnO nanowire arrays decorated with plasmonic Au nanoparticles for high-performance photoelectrochemical water splitting. ACS Appl Mater Interfaces 6:4480–4489. doi:10.1021/am500234v 77. DeSario PA, Pietron JJ, DeVantier DE et al (2013) Plasmonic enhancement of visible-light water splitting with Au–TiO2 composite aerogels. Nanoscale 5:8073–8083. doi:10.1039/ c3nr01429k 78. Ingram DB, Linic S (2011) Water splitting on composite plasmonic-metal/semiconductor photoelectrodes: evidence for selective plasmon-induced formation of charge carriers near the semiconductor surface. J Am Chem Soc 133:5202–5205. doi:10.1021/ja200086g 79. Meng F, Cushing SK, Li J et al (2015) Enhancement of solar hydrogen generation by synergistic interaction of La2Ti2O7 photocatalyst with plasmonic gold nanoparticles and reduced graphene oxide nanosheets. ACS Catal 5:1949–1955. doi:10.1021/cs5016194 80. Hou W, Cronin SB (2012) A review of surface plasmon resonance-enhanced photocatalysis. Adv Funct Mater 23:1612–1619. doi:10.1002/adfm.201202148 81. Wu B, Liu D, Mubeen S et al (2016) Anisotropic growth of TiO2 onto gold nanorods for plasmon-enhanced hydrogen production from water reduction. J Am Chem Soc 138:1114– 1117. doi:10.1021/jacs.5b11341 82. Chen JJ, Wu JCS, Wu PC, Tsai DP (2011) Plasmonic photocatalyst for H2 evolution in photocatalytic water splitting. J Phys Chem C 115:210–216. doi:10.1021/jp1074048 83. Mubeen S, Hernandez-Sosa G, Moses D et al (2011) Plasmonic photosensitization of a wide band gap semiconductor: converting plasmons to charge carriers. Nano Lett 11:5548–5552. doi:10.1021/nl203457v 84. Priebe JB, Radnik J, Lennox AJJ et al (2015) Solar hydrogen production by plasmonic Au − TiO2 catalysts: impact of synthesis protocol and TiO2 phase on charge transfer efficiency and H2 evolution rates. ACS Catal 5:2137–2148. doi:10.1021/cs5018375 85. Zhan Z, An J, Zhang H et al (2014) Three-dimensional plasmonic photoanodes based on Au-Embedded TiO2 structures for enhanced visible-light water splitting. ACS App Mater Interfaces 6:1139–1144. doi:10.1021/am404738a 86. Zhang Z, Zhang L, Hedhili MN et al (2013) Plasmonic gold nanocrystals coupled with photonic crystal seamlessly on TiO2 nanotube photoelectrodes for efficient visible light photoelectrochemical water splitting. Nano Lett 13:14–20. doi:10.1021/nl3029202 87. Chen C, Kuai L, Chen Y et al (2015) Au/Pt co-loaded ultrathin TiO2 nanosheets for photocatalyzed H2 evolution by the synergistic effect of plasmonic enhancement and co-catalysis. RSC Adv 5:98254–98259. doi:10.1039/c5ra17732d 88. Yu G, Wang X, Cao J et al (2016) Plasmonic Au nanoparticles embedding enhances the activity and stability of CdS for photocatalytic hydrogen evolution. Chem Commun 52:2394–2397. doi:10.1039/c5cc10066f 89. Luo Y, Liu X, Tang X et al (2014) Gold nanoparticles embedded in Ta2O5/Ta3N5 as active visible-light plasmonic photocatalysts for solar hydrogen evolution. J Mater Chem A 2:14927–14939. doi:10.1039/c4ta02991g 90. Wu M, Chen WJ, Shen YH et al (2014) In situ growth of matchlike ZnO/Au plasmonic heterostructure for enhanced photoelectrochemical water splitting. ACS Appl Nater Interfaces 6:15052–15060. doi:10.1021/am503044f 91. Zeng J, Song T, Lv M et al (2016) Plasmonic photocatalyst Au/g-C3N4/NiFe2O4 nanocomposites for enhanced visible-light-driven photocatalytic hydrogen evolution. RSC Adv 6:54964–54975. doi:10.1039/c6ra08356k 92. Patnaik S, Martha S, Madras G, Parida K (2016) The effect of sulfate pre-treatment to improve the deposition of Au-nanoparticles in a gold-modified sulfated g-C3N4 plasmonic photocatalyst towards visible light induced water reduction reaction. Phys Chem Chem Phys 18:28502–28514. doi:10.1039/c6cp04262g 93. Zhang X, Zhao J, Wang S et al (2014) Shape-dependent localized surface plasmon enhanced photocatalytic effect of ZnO nanorods decorated with Ag. Inter J Hydrogen En 39:8238– 8245. doi:10.1016/j.ijhydene.2014.03.153 98 2 Semiconductor-Based Photocatalytic Systems …

94. Yi J, She X, Song Y et al (2016) A silver on 2D white-C3N4 support photocatalyst for mechanistic insights: synergetic utilization of plasmonic effect for solar hydrogen evolution. RSC Adv 6:112420–112428. doi:10.1039/c6ra23964a 95. Kryukov AI, Kuchmiy SY, Korzhak AV et al (1993) Synergism in an organized photocatalytic system based on cadmium sulfide and titanium complexes. Theoret Exp Chem 29:312–316. doi:10.1007/BF00532100 96. Kryukov AI, Kuchmiy SY, Kulik SV et al (1994) New photocatalysts: dispersed semiconductor materials with microheterojunctions. Theoret Exp Chem 30:182–186. doi:10.1007/BF00531180 97. Kryukov AI, Kuchmiy SY, Pokhodenko VD (1994) Molecular design in photo-catalysis: physicochemical principles for designing high-efficiency photocatalytic oxidation-reduction systems. Theoret Exp Chem 30:141–157. doi:10.1007/BF00534653 98. Kryukov A, Kuchmiy SY, Pokhodenko VD (1997) Nanostructural composite photocatalysts based on polycrystalline cadmium sulfide. Theoret Exp Chem 33:306–321. doi:10.1007/ BF02522707 99. Kryukov AI, Kuchmiy SY, Pokhodenko VD (2000) Energetics of electron processes in semiconductor photocatalytic systems. Theoret Exp Chem 36:69–89. doi:10.1007/BF02529022 100. Fedoseev BI, Savinov EI, Parmon VN (1987) Photocatalytic hydrogen evolution from Na2S solution in the presence of highly disperse cadmium and copper sulfides. Kinetics Catal 28:1111–1115 101. Ogisu K, Takanabe K, Lu D et al (2009) CdS nanoparticles exhibiting quantum size effect by dispersion on TiO2: photocatalytic H2 evolution and photoelectrochemical measurements. Bull Chem Soc Jpn 82:528–535. doi:10.1246/bcsj.82.528 102. Hirai T, Suzuki K, Komasawa I (2001) Preparation and photocatalytic properties of composite cds nanoparticles-titanium dioxide particles. J Colloid Interface Sci 244:262–265. doi:10.1006/jcis.2001.7982 103. Stroyuk OL, Kryukov AI, Kuchmiy SY, Pokhodenko VD (2005) Quantum size effects in the photonics of semiconductor nanoparticles. Theoret Exp Chem 41:67–91. doi:10.1007/ s11237-005-0025-9 104. Stroyuk OL, Kryukov AI, Kuchmiy SY, Pokhodenko VD (2005) Quantum size effects in semiconductor photocatalysis. Theoret Exp Chem 41:207–228. doi:10.1007/s11237-005-0042-8 105. Lian Z, Xu P, Wang W et al (2015) C60-decorated CdS/TiO2 mesoporous architectures with enhanced photostability and photocatalytic activity for H2 evolution. ACS Appl Mater Interfaces 7:4533–4540. doi:10.1021/am5088665 106. Ji SM, Jun H, Jang JS et al (2007) Photocatalytic hydrogen production from natural seawater. J Photochem Photobiol A 189:141–144. doi:10.1016/j.jphotochem.2007.01.011 107. Park H, Choi W, Hoffmann MR (2008) Effects of the preparation method of the ternary CdS/ TiO2/Pt hybrid photocatalysts on visible light-induced hydrogen production. J Mater Chem 18:2379–2385. doi:10.1039/B718759A 108. Jang JS, Choi SH, Kim HG, Lee JS (2008) Location and state of Pt in platinized CdS/TiO2 photocatalysts for hydrogen production from water under visible light. J Phys Chem C 112:17200–17205. doi:10.1021/jp804699c 109. Cui X, Jiang G, Zhu M et al (2013) TiO2/CdS composite hollow spheres with controlled synthesis of platinum on the internal wall for the efficient hydrogen evolution. Inter J Hydrogen En 38:9065–9073. doi:10.1016/j.ijhydene.2013.05.062 110. Zhou H, Pan J, Ding L et al (2014) Biomass-derived hierarchical porous CdS/M/TiO2 (M = Au, Ag, pt, pd) ternary heterojunctions for photocatalytic hydrogen evolution Inter J Hydrogen En 39:16293–16301. doi:10.1016/j.ijhydene. 2014.08.032 111. Cui X, Wang Y, Jiang G et al (2014) A photonic crystal-based CdS–Au–WO3 heterostructure for efficient visible-light photocatalytic hydrogen and oxygen evolution. RSC Adv 4:15689–15694. doi:10.1039/C4RA01415D 112. Vaishnav JK, Arbuj SS, Rane SB, Amalnerkar DP (2014) One dimensional CdS/ZnO nanocomposites: an efficient photocatalyst for hydrogen generation. RSC Adv 4:47637– 47642. doi:10.1039/C4RA08561B References 99

113. Yang G, Yan W, Zhang Q et al (2013) One-dimensional CdS/ZnO core/shell nanofibers via single-spinneret electrospinning: tunable morphology and efficient photocatalytic hydrogen production. Nanoscale 5:12432–12439. doi:10.1039/C3NR03462C 114. Shen S, Guo L (2008) Growth of quantum-confined CdS nanoparticles inside Ti-MCM-41 as a visible light photocatalyst. Mater Res Bull 43:437–446. doi:10.1016/j.materresbull.2007. 02.034 115. Liu Z, Shen S, Guo L (2012) Study on photocatalytic performance for hydrogen evolution over CdS/M-MCM-41 (M = Zr, Ti) composite photocatalysts under visible light illumination. Inter J Hydrogen En 37:816–821. doi:10.1016/j.ijhydene.2011.04.052 116. Zhang YJ, Yan W, Wu YP, Wang ZH (2008) Synthesis of TiO2 nanotubes coupled with CdS nanoparticles and production of hydrogen by photocatalytic water decomposition. Mater Lett 62(2008):3846–3848. doi:10.1016/j.matlet.2008.04.084 117. Xing C, Jing D, Liu M, Guo L (2009) Photocatalytic hydrogen production over Na2Ti2O4(OH)2 nanotube sensitized by CdS nanoparticles. Mater Res Bull 44:442–445. doi:10.1016/j.materresbull.2008.04.016 118. Kim JC, Choi J, Lee YB et al (2006) Enhanced photocatalytic activity in composites of TiO2 nanotubes and CdS nanoparticles. Chem Commun 48(2006):5024–5026. doi:10.1039/ B612572G 119. Chen Y, Guo L (2012) Highly efficient visible-light-driven photocatalytic hydrogen production from water using Cd0.5Zn0.5S/TNTs (titanate nanotubes) nanocomposites without noble metals. J Mater Chem 22:7507–7514. doi:10.1039/C2JM16797B 120. Qi L, Yu J, Jaroniec M (2011) Preparation and enhanced visible-light photocatalytic H2- production activity of CdS-sensitized Pt/TiO2 nanosheets with exposed (001) facets. Phys Chem Chem Phys 13:8915–8923. doi:10.1039/C1CP20079H 121. Gao B, Yuan X, Lu P et al (2015) Enhanced visible-light-driven photocatalytic H2- production activity of CdS-loaded TiO2 microspheres with exposed (001) facets. J Phys Chem Sol 87:171–176. doi:10.1016/j.jpcs.2015.08.018 122. Guan G, Kida T, Kusakabe K et al (2005) Photocatalytic activity of CdS nanoparticles incorporated in titanium silicate molecular sieves of ETS-4 and ETS-10. Appl Catal A 295:71–78. doi:10.1016/j.apcata.2005.08.010 123. Shangguan W, Yoshida A (2001) Synthesis and photocatalytic properties of CdS-intercalated metal oxides. Sol En Mater Sol Cells 69:189–194. doi:10.1016/S0927- 0248(01)00020-4 124. Shangguan W, Yoshida A (2002) Photocatalytic hydrogen evolution from water on nanocomposites incorporating cadmium sulfide into the interlayer. J Phys Chem B 106:12227–12230. doi:10.1021/jp0212500 125. Dinh CT, Pham NH, Kleitz F, Do TO (2013) Design of water-soluble CdS–titanate–nickel nanocomposites for photocatalytic hydrogen production under sunlight. J Mater Chem A 1:13308–13313. doi:10.1039/C3TA12914D 126. Xiao J, Peng T, Ke D et al (2007) Synthesis, characterization of CdS/rectorite nanocomposites and its photocatalytic activity. Phys Chem Miner 34:275–285. doi:10. 1007/s00269-007-0146-x 127. Parayil SK, Baltrusaitis J, Wu CM, Koodali RT (2013) Synthesis and characterization of ligand stabilized CdS-Trititanate composite materials for visible light-induced photocatalytic water splitting. Inter J Hydrogen En 38:2656–2669. doi:10.1016/j.ijhydene.2012.12.042 128. Ryu SY, Choi J, Balcerski W et al (2007) Photocatalytic production of H2 on nanocomposite catalysts. Ind Eng Chem Res 46:7476–7488. doi:10.1021/ie0703033 129. Shangguan W (2007) Hydrogen evolution from water splitting on nanocomposite photocatalysts. Sci Technol Adv Mater 8:76–81. doi:10.1016/j.stam.2006.09.007 130. Choi J, Ryu Y, Balcerski W et al (2008) Photocatalytic production of hydrogen on Ni/NiO/ KNbO3/CdS nanocomposites using visible light. J Mater Chem 18:2371–2378. doi:10.1039/ B718535A 100 2 Semiconductor-Based Photocatalytic Systems …

131. Chen W, Gao H, Yuan J et al (2013) Structure characteristics of CdS/ H1.9K0.3La0.5Bi0.1Ta2O7 and photocatalytic activity for hydrogen evolution under visible light. Inter J Hydrogen En 38:10754–10760. doi:10.1016/j.ijhydene.2013.02.067 132. Yu J, Lei SL, Chen TC et al (2014) A new CdS/Bi1−xInxTaO4 heterostructured photocatalyst containing solid solutions for H2 evolution from water splitting. Inter J Hydrogen En 39:13105–13113. doi:10.1016/j.ijhydene.2014.06.148 133. Zhang G, Lin B, Yang W et al (2015) Highly efficient photocatalytic hydrogen generation by incorporating CdS into ZnCr-layered double hydroxide interlayer. RSC Adv 5:5823–5829. doi:10.1039/C4RA11757C 134. Wu J, Lin J, Yin S, Sato T (2001) Synthesis and photocatalytic properties of layered HNbWO6/(Pt, Cd0.8Zn0.2S) nanocomposites. J Mater Chem 11:3343–3347. doi:10.1039/ B103838A 135. So WW, Kim KJ, Moon SJ (2004) Photo-production of hydrogen over the CdS–TiO2 nanocomposite particulate films treated with TiCl4. Inter J Hydrogen En 29:229–234. doi:10. 1016/S0360-3199(03)00211-8 136. Wang H, Zhu W, Chong B, Qin K (2014) Improvement of photocatalytic hydrogen generation from CdSe/CdS/TiO2 nanotube-array coaxial heterogeneous structure. Inter J Hydrogen En 39:90–99. doi:10.1016/j.ijhydene.2013.10.048 137. Kudo A, Miseki Y (2009) Heterogeneous photocatalyst materials for water splitting. Chem Soc Rev 38:253–278. doi:10.1039/B800489G 138. Moon GD, Joo JB, Lee I, Yin Y (2014) Decoration of size-tunable CuO nanodots on TiO2 nanocrystals for noble metal-free photocatalytic H2 production. Nanoscale 6:12002–12008. doi:10.1039/C4NR03521F 139. Jang JS, Ji SM, Bae SW et al (2007) Optimization of CdS/TiO2 nano-bulk composite photocatalysts for hydrogen production from Na2S/Na2SO3 aqueous electrolyte solution under visible light (k 420 nm). J Photochem Photobiol, A 188:112–119. doi:10.1016/j. jphotochem.2006.11.027 140. Jang JS, Kim HG, Borse PH, Lee JS (2007) Simultaneous hydrogen production and decomposition of H2S dissolved in alkaline water over CdS–TiO2 composite photocatalysts under visible light irradiation. Inter J Hydrogen En 32:4786–4791. doi:10.1016/j.ijhydene. 2007.06.026 141. Jang JS, Kim HG, Joshi UA et al (2008) Fabrication of CdS nanowires decorated with TiO2 nanoparticles for photocatalytic hydrogen production under visible light irradiation. Inter J Hydrogen En 33:5975–5980. doi:10.1016/j.ijhydene.2008.07.105 142. Li C, Xi Z, Fang W et al (2015) Enhanced photocatalytic hydrogen evolution activity of CuInS2 loaded TiO2 under solar light irradiation. J Sol State Chem 226:94–100. doi:10. 1016/j.jssc.2015.02.011 143. Guo K, Liu Z, Han J et al (2014) Hierarchical TiO2–CuInS2 core–shell nanoarrays for photoelectrochemical water splitting. Phys Chem Chem Phys 16:16204–16213. doi:10.1039/ C4CP01971G 144. Li K, Xu J, Zhang X et al (2013) Low-temperature preparation of AgIn5S8/TiO2 heterojunction nanocomposite with efficient visible-light-driven hydrogen production. Inter J Hydrogen En 38:15965–15975. doi:10.1016/j.ijhydene.2013.09.147 145. Cheng Z, Zhan X, Wang F et al (2015) Construction of CuInS2/Ag sensitized ZnO nanowire arrays for efficient hydrogen generation. RSC Adv 5:81723–81727. doi:10.1039/c5ra14188e 146. Choi Y, Baek M, Zhang Z et al (2015) A two-storey structured photoanode of a 3D Cu2ZnSnS4/CdS/ZnO@steel composite nanostructure for efficient photoelectrochemical hydrogen generation. Nanoscale 7:15291–15299. doi:10.1039/c5nr04107d 147. Kandiel TA, Takanabe K (2016) Solvent-induced deposition of Cu–Ga–In–S nanocrystals onto atitanium dioxide surface for visible-light-driven photocatalytic hydrogen production. Appl Catal B 184:264–269. doi:10.1016/j.apcatb.2015.11.036 148. Brahimi R, Bessekhouad Y, Bouguelia A, Trari M (2007) Visible light induced hydrogen evolution over the heterosystem Bi2S3/TiO2. Catal Today 122:62–65. doi:10.1016/j.cattod. 2007.01.030 References 101

149. Kim J, Kang M (2012) High photocatalytic hydrogen production over the band gap-tuned urchin-like Bi2S3-loaded TiO2 composites system. Inter J Hydrogen En 37:8249–8256. doi:10.1016/j.ijhydene.2012.02.057 150. Senevirathna N, Pitigaka P, Tennakone K (2005) Water photoreduction with Cu2O quantum dots on TiO2 nanoparticles. J Photochem Photobiol A 171:257–259. doi:10.1016/j. jphotochem.2004.10.018 151. Zhang S, Zhang S, Peng F et al (2011) Electrodeposition of polyhedral Cu2O on TiO2 nanotube arrays for enhancing visible light photocatalytic performance. Electrochem Commun 13:861–864. doi:10.1016/j.elecom.2011.05.022 152. Kumar DP, Reddy NL, Kumari MM et al (2015) Cu2O-sensitized TiO2 nanorods with nanocavities for highly efficient photocatalytic hydrogen production under solar irradiation. Sol En Mater Sol Cells 136:157–166. doi:10.1016/j.solmat.2015.01.009 153. Zhang S, Peng B, Yang S et al (2013) The influence of the electrodeposition potential on the morphology of Cu2O/TiO2 nanotube arrays and their visible-light-driven photocatalytic activity for hydrogen evolution. Inter J Hydrogen En 38:13866–13871. doi:10.1016/j. ijhydene.2013.08.081 154. Li L, Xu L, Shi W, Guan J (2013) Facile preparation and size-dependent photocatalytic activity of Cu2O nanocrystals modified titania for hydrogen evolution. Inter J Hydrogen En 38:816–822. doi:10.1016/j.ijhydene.2012.10.064 155. Sharma D, Upadhyay S, Satsangi VR et al (2014) Improved photoelectrochemical water splitting performance of Cu2O/SrTiO3 heterojunction photoelectrode. J Phys Chem C 118:25320–25329. doi:10.1021/jp507039n 156. Jung M, Scott J, Ng YH et al (2014) CuOx dispersion and reducibility on TiO2 and its impact on photocatalytic hydrogen evolution. Inter J Hydrogen En 39:12499–12506. doi:10.1016/j. ijhydene.2014.06.020 157. Bandara J, Udawatta C, Rajapakse C (2005) Highly stable CuO incorporated TiO2 catalyst for photocatalytic hydrogen production from H2O. Photochem Photobiol Sci 4:857–861. doi:10.1039/B507816D 158. Yu Z, Meng J, Li Y et al (2013) Efficient photocatalytic hydrogen production from water over a CuO and carbon fiber comodified TiO2 nanocomposite photocatalyst. Inter J Hydrogen En 38:6649–16655. doi:10.1016/j.ijhydene.2013.07.056 159. Xu S, Du AJ, Liu J et al (2011) Highly efficient CuO incorporated TiO2 nanotube photocatalyst for hydrogen production from water. Inter J Hydrogen En 36:6560–6568. doi:10.1016/j.ijhydene.2011.02.103 160. Kumar SP, Shankar MV, Kumari MM et al (2013) Nano-size effects on CuO/TiO2 catalysts for highly efficient H2 production under solar light irradiation. Chem Commun 49:9443– 9445. doi:10.1039/C3CC44742A 161. Yue X, Yi S, Wang R et al (2016) A novel and highly efficient earth-abundant Cu3P with TiO2 “P–N” heterojunction nanophotocatalyst for hydrogen evolution from water. Nanoscale 8:17516–17523. doi:10.1039/c6nr06620h 162. Yang X, Xu J, Wong T et al (2013) Synthesis of In2O3–In2S3 core–shell nanorods with inverted type-I structure for photocatalytic H2 generation. Phys Chem Chem Phys 15:12688– 12693. doi:10.1039/C3CP51722E 163. Liu Z, Bai H, Xu S, Sun DD (2011) Hierarchical CuO/ZnO “corn-like” architecture for photocatalytic hydrogen generation. Inter J Hydrogen En 36:13473–13480. doi:10.1016/j. ijhydene.2011.07.137 164. Rajaambal S, Mapa M, Gopinath CS (2014) In1−xGaxN@ZnO: a rationally designed and quantum dot integrated material for water splitting and solar harvesting applications. Dalton Trans 43:12546–12554. doi:10.1039/C4DT01268B 165. Derbal A, Omeiri S, Bouguelia A, Trari M (2008) Characterization of new heterosystem CuFeO2/SnO2 application to visible-light induced hydrogen evolution. Inter J Hydrogen En 33:4274–4282. doi:10.1016/j.ijhydene.2008.05.067 102 2 Semiconductor-Based Photocatalytic Systems …

166. Nguyen-Phan TD, Luo S, Vovchok D et al (2016) Visible light-driven H2 production over highly dispersed ruthenia on rutile TiO2 nanorods. ACS Catal 6:407–417. doi:10.1021/ acscatal.5b02318 167. Brahimi R, Bessekhouad Y, Bouguelia A, Trari M (2007) CuAlO2/TiO2 heterojunction applied to visible light H2 production. J Photochem Photobiol A 186:242–247. doi:10.1016/ j.jphotochem.2006.08.013 168. Ou Y, Lin J, Fang S, Liao D (2006) MWNT–TiO2: Ni composite catalyst: a new class of catalyst for photocatalytic H2 evolution from water under visible light illumination. Chem Phys Lett 429:199–203. doi:10.1016/j.cplett.2006.08.024 169. Yu H, Shi R, Zhao Y et al (2016) Smart utilization of carbon dots in semiconductor photocatalysis. Adv Mater 28:9454–9477. doi:10.1002/adma.201602581 170. Cao S, Yu J (2016) Carbon-based H2-production photocatalytic materials. J Photochem Photobiol C 27:72–99. doi:10.1016/j.jphotochemrev.2016.04.002 171. Wang J, Gao M, Ho GW (2014) Bidentate-complex-derived TiO2/carbon dot photocatalysts: in situ synthesis, versatile heterostructures, and enhanced H2 evolution. J Mater Chem A 2 (2014):5703–5709. doi:10.1039/C3TA15114J 172. Yu H, Zhao Y, Zhou C et al (2014) Carbon quantum dots/TiO2 composites for efficient photocatalytic hydrogen evolution. J Mater Chem A 2:3344–3351. doi:10.1039/ C3TA14108J 173. Tang Y, Hao R, Fu Y et al (2016) Carbon quantum dot/mixed crystal TiO2 composites via a hydrogenation process: an efficient photocatalyst for the hydrogen evolution reaction. RSC Adv 6:96803–96808. doi:10.1039/c6ra17597j 174. Wang J, Huang J, Xie H, Qu A (2014) Synthesis of g-C3N4/TiO2 with enhanced photocatalytic activity for H2 evolution by a simple method. Inter J Hydrogen En 39:6354– 6363. doi:10.1016/j.ijhydene.2014.02.020 175. Zang Y, Li L, Xu Y et al (2014) Hybridization of brookite TiO2 with g-C3N4:a visible-light-driven photocatalyst for As3+ oxidation, MO degradation and water splitting for hydrogen evolution. J Mater Chem A 2:15774–15780. doi:10.1039/C4TA02082K 176. Li Y, Wang R, Li H et al (2015) Efficient and stable photoelectrochemical seawater splitting with TiO2@g-C3N4 nanorod arrays decorated by Co-Pi. J Phys Chem C 119:20283–20292. doi:10.1021/acs.jpcc.5b05427 177. Pany S, Parida KM (2015) A facile in situ approach to fabricate N, S-TiO2/g-C3N4 nanocomposite with excellent activity for visible light induced water splitting for hydrogen evolution. Phys Chem Chem Phys 17:8070–8077. doi:10.1039/C4CP05582A 178. Zhong X, Jin M, Dong H et al (2014) TiO2 nanobelts with a uniform coating of g-C3N4 as a highly effective heterostructure for enhanced photocatalytic activities. J Sol State Chem 220:54–59. doi:10.1016/j.jssc.2014.08.016 179. Jiang Y, Guo S, Hao R et al (2016) A hybridized heterojunction structure between TiO2 nanorods and exfoliated graphitic carbon nitride sheets for hydrogen evolution under visible light. CrystEngComm 18:6875–6880. doi:10.1039/c6ce01442a 180. Cheng F, Yin H, Xiang Q (2017) Low-temperature solid-state preparation of ternary CdS/ g-C3N4/CuS nanocomposites for enhanced visible-light photocatalytic H2-production activity. Appl Surf Sci 391:432–439. doi:10.1016/j.apsusc.2016.06.169 181. Liu H, Jin Z, Xu Z et al (2015) Fabrication of ZnIn2S4–g-C3N4 sheet-on-sheet nanocomposites for efficient visible-light photocatalytic H2-evolution and degradation of organic pollutants. RSC Adv 5:97951–97961. doi:10.1039/c5ra17028a 182. Zheng D, Zhang G, Wang X (2015) Integrating CdS quantum dots on hollow graphitic carbon nitride nanospheres for hydrogen evolution photocatalysis. Appl Catal B 179:479– 488. doi:10.1016/j.apcatb.2015.05.060 183. Chen W, Liu TY, Huang T et al (2015) A novel yet simple strategy to fabricate visible light responsive C, N-TiO2/g-C3N4 heterostructures with significantly enhanced photocatalytic hydrogen generation. RSC Adv 5:101214–101220. doi:10.1039/c5ra18302b References 103

184. Hu B, Cai F, Chen T et al (2015) Hydrothermal synthesis g-C3N4/Nano-InVO4 nanocomposites and enhanced photocatalytic activity for hydrogen production under visible light irradiation. ACS Appl Mater Interfaces 7:18247–18256. doi:10.1021/acsami.5b05715 185. Kudo A (2007) Recent progress in the development of visible light-driven powdered photocatalysts for water splitting. Inter J Hydrogen En 32:2673–2678. doi:10.1016/j. ijhydene.2006.09.010 186. Ni M, Leung M, Leung D, Sumathy K (2007) A review and recent developments in photocatalytic water-splitting using TiO2 for hydrogen production. Renew Sustain En Rev 11:401–425. doi:10.1016/j.rser.2005.01.009 187. Matsuoka M, Kitano M, Takeuchi M et al (2007) Photocatalysis for new energy production: recent advances in photocatalytic water splitting reactions for hydrogen production. Catal Today 122:51–61. doi:10.1016/j.cattod.2007.01.042 188. Lee JS (2005) Photocatalytic water splitting under visible light with particulate semiconductor catalysts. Catal Surv Asia 9:217–227. doi:10.1007/s10563-005-9157-0 189. Kudo A (2003) Photocatalyst materials for water splitting. Catal Surv Asia 7:31–38. doi:10. 1023/A:1023480507710 190. Tsuji I, Kudo A (2003) H2 evolution from aqueous sulﬁte solutions under visible-light irradiation over Pb and halogen-codoped ZnS photocatalysts. J Photochem Photobiol A 156:249–252. doi:10.1016/S1010-6030(02)00433-1 191. Kudo A, Sekizawa M (2000) Photocatalytic H2 evolution under visible light irradiation on Ni-doped ZnS photocatalyst. Chem Commun 1371–1372. doi:10.1039/B003297M 192. Kudo A, Sekizawa M (1999) Photocatalytic H2 evolution under visible light irradiation on Zn1−xCuxS solid solution. Catal Lett 58:241–243. doi:10.1023/A:1019067025917 193. Mei Z, Ouyang S, Zhang Y, Kako T (2013) Ultraﬁne Zn1−xCuxS(0 x 0.066) nanocrystallites for photocatalytic H2 evolution under visible light irradiation. RSC Adv 3:10654–10657. doi:10.1039/C3RA41076E 194. Wu Y, Lu G, Li S (2009) The doping effect of Bi on TiO2 for photocatalytic hydrogen generation and photodecolorization of rhodamine B. J Phys Chem C 113:9950–9955. doi:10. 1021/jp9009433 195. Ishii T, Kato H, Kudo A (2004) H2 evolution from an aqueous methanol solution on SrTiO3 photocatalysts codoped with chromium and tantalum ions under visible light irradiation. J Photochem Photobiol A 163:181–186. doi:10.1016/S1010-6030(03)00442-8 196. Wang D, Ye J, Kako T, Kimura T (2006) Photophysical and photocatalytic properties of SrTiO3 doped with Cr cations on different sites. J Phys Chem B 110:15824–15830. doi:10. 1021/jp062487p 197. Iwashina K, Kudo A (2011) Rh-doped SrTiO3 photocatalyst electrode showing cathodic photocurrent for water splitting under visible-light irradiation. J Am Chem Soc 133:13272– 13275 198. Kato H, Sasaki Y, Shirakura N, Kudo A (2013) Synthesis of highly active rhodium-doped SrTiO3 powders in Z-scheme systems for visible-light-driven photocatalytic overall water splitting. J Mater Chem A 1:12327–12333. doi:10.1039/C3TA12803B 199. Niishiro R, Kato H, Kudo A (2005) Nickel and either tantalum or niobium-codoped TiO2 and SrTiO3 photocatalysts with visible-light response for H2 or O2 evolution from aqueous solutions. Phys Chem Chem Phys 7:2241–2245. doi:10.1039/B502147B 200. Kato H, Kudo A (2002) Visible-light-response and photocatalytic activities of TiO2 and SrTiO3 photocatalysts codoped with antimony and chromium. J Phys Chem B 106:5029– 5034. doi:10.1021/jp0255482 201. Hwang DW, Kim HG, Lee JS et al (2005) Photocatalytic hydrogen production from water over M-doped La2Ti2O7 (M = Cr, Fe) under visible light irradiation (k > 420 nm). J Phys Chem B 109:2093–2102. doi:10.1021/jp0493226 202. Hwang DW, Kim HG, Jang JS et al (2004) Photocatalytic decomposition of water–methanol solution over metal-doped layered perovskites under visible light irradiation. Catal Today 93–95:845–850. doi:10.1016/j.cattod.2004.06.084 104 2 Semiconductor-Based Photocatalytic Systems …

203. Kanhere P, Zheng J, Chen Z (2012) Visible light driven photocatalytic hydrogen evolution 3+ and photophysical properties of Bi doped NaTaO3. Inter J Hydrogen En 37:4889–4896. doi:10.1016/j.ijhydene.2011.12.056 204. Zou JP, Zhang LZ, Luo SL et al (2012) Preparation and photocatalytic activities of two new Zn-doped SrTiO3 and BaTiO3 photocatalysts for hydrogen production from water without cocatalysts loading. Inter J Hydrogen En 37:17068–17077. doi:10.1016/j.ijhydene.2012.08. 133 205. Xue Y, Wang X (2015) The effects of Ag doping on crystalline structure and photocatalytic properties of BiVO4. Inter J Hydrogen En 40:5878–5888. doi:10.1016/j.ijhydene.2015.03. 028 206. Jana P, Montero CM, Pizzarro P et al (2014) Photocatalytic hydrogen production in the water/methanol system using Pt/RE:NaTaO3 (RE = Y, La, Ce, Yb) catalysts. Inter J Hydrogen En 39:5283–5290. doi:10.1016/j.ijhydene.2013.12.182 207. Shimura K, Yoshida H (2012) Effect of doped zinc species on the photocatalytic activity of gallium oxide for hydrogen production. Phys Chem Chem Phys 14:2678–2684. doi:10.1039/ C2CP23220K 208. Kudo A, Niishiro R, Iwase A, Kato H (2007) Effects of doping of metal cations on morphology, activity, and visible light response of photocatalysts. Chem Phys 339:104–110. doi:10.1016/j.chemphys.2007.07.024 209. Zou Z, Arakawa H (2003) Direct water splitting into H2 and O2 under visible light irradiation with a new series of mixed oxide semiconductor photocatalysts. J Photochem Photobiol A 158:145–162. doi:10.1016/S1010-6030(03)00029-7 210. Zou Z, Ye J, Sayama K, Arakawa H (2002) Photocatalytic hydrogen and oxygen formation under visible light irradiation with M-doped InTaO4 (M = M n, Fe Co, Ni and Cu) photocatalysts. J Photochem Photobiol A 148:65–69. doi:10.1016/S1010-6030(02)00068-0 211. Wang D, Zou Z, Ye J (2005) Photocatalytic water splitting with the Cr-doped Ba2In2O5/ In2O3 composite oxide semiconductors. Chem Mater 17:3255–3261. doi:10.1021/ cm0477117 212. Gurunathan K (2004) Photocatalytic hydrogen production using transition metal ions-doped c-Bi2O3 semiconductor particles. Inter J Hydrogen En 29:933–940. doi:10.1016/j.ijhydene. 2003.04.001 213. Jing D, Zhang Y, Guo L (2005) Study on the synthesis of Ni doped mesoporous TiO2 and its photocatalytic activity for hydrogen evolution in aqueous methanol solution. Chem Phys Lett 415:74–78. doi:10.1016/j.cplett.2005.08.080 214. Liu Q, Ding D, Ning C, Wang X (2015) Black Ni-doped TiO2 photoanodes for high-efficiency photoelectrochemical water-splitting. Inter J Hydrogen En 40:2107–2114. doi:10.1016/j.ijhydene.2014.12.064 215. Wang D, Ye J, Kitazawa H, Kimura T (2007) Photophysical and photocatalytic properties of three isostructural oxide semiconductors In6NiTi6O22,In3CrTi2O10, and In12NiCr2Ti10O42 with different 3d transition metals. J Phys Chem C 111:12848–12854. doi:10.1021/ jp0678599 216. Wang D, Zou Z, Ye J (2005) Photocatalytic H2 evolution over a new visible-light-driven photocatalyst In12NiCr2Ti10O42. Chem Phys Lett 411:285–290. doi:10.1016/j.cplett.2005. 05.124 217. Konta R, Ishii T, Kato H, Kudo A (2004) Photocatalytic activities of noble metal ion doped SrTiO3 under visible light irradiation. J Phys Chem B 108:8992–8995. doi:10.1021/ jp049556p 218. Kitano M, Takeuchi M, Matsuoka M et al (2007) Photocatalytic water splitting using Pt-loaded visible light-responsive TiO2 thin film photocatalysts. Catal Today 120:133–138. doi:10.1016/j.cattod.2006.07.043 219. Fukumoto S, Kitano M, Takeuchi M et al (2009) Photocatalytic hydrogen production from aqueous solutions of alcohol as model compounds of biomass using visible light-responsive TiO2 thin films. Catal Lett 127:39–43. doi:10.1007/s10562-008-9769-x References 105

220. Kitano M, Tsujimaru K, Anpo M (2008) Hydrogen production using highly active titanium oxide-based photocatalysts. Topics Catal 49:4–17. doi:10.1007/s11244-008-9059-2 221. Selli E, Chiarello GL, Quartarone E et al (2007) A photocatalytic water splitting device for separate hydrogen and oxygen evolution. Chem Commun 5022–5024. doi:10.1039/ B711747G 222. Matsuoka M, Kitano M, Fukumoto S et al (2008) The effect of the hydrothermal treatment with aqueous NaOH solution on the photocatalytic and photoelectrochemical properties of visible light-responsive TiO2 thin films. Catal Today 132:159–164. doi:10.1016/j.cattod. 2007.12.032 223. Kitano M, Takeuchi M, Matsuoka M et al (2005) Preparation of visible light-responsive TiO2 thin film photocatalysts by an RF magnetron sputtering deposition method and their photocatalytic reactivity. Chem Lett 34:616–617. doi:10.1246/cl.2005.616 224. Dholam R, Patel N, Adami M, Miotello A (2008) Physically and chemically synthesized TiO2 composite thin films for hydrogen production by photocatalytic water splitting. Inter J Hydrogen En 33:6896–6903. doi:10.1016/j.ijhydene.2008.08.061 225. Matsuoka M, Kitano M, Takeuchi M et al (2005) Photocatalytic water splitting on visible light-responsive TiO2 thin films prepared by a RF magnetron sputtering deposition method. Topics Catal 35:305–310. doi:10.1007/s11244-005-3838-9 226. Wang Z, Yang C, Lin T et al (2013) Visible-light photocatalytic, solar thermal and photoelectrochemical properties of aluminium-reduced black titania. En Environ Sci 6:3007–3014. doi:10.1039/c3ee41817k 227. Cui H, Zhao W, Yang C et al (2014) Black TiO2 nanotube arrays for high-efficiency photoelectrochemical water-splitting. J Mater Chem A 2:8612–8616. doi:10.1039/ c4ta00176a 228. Zhang K, Zhou W, Zhang X et al (2016) Large-scale synthesis of stable mesoporous black TiO2 nanosheets for efficient solar-driven photocatalytic hydrogen evolution via an earthabundant low-cost biotemplate. RSC Adv 6:50506–50512. doi:10.1039/c6ra06751d 229. Yuan J, Chen M, Shi J, Shanguang W (2006) Preparations and photocatalytic hydrogen evolution of N-doped TiO2 from urea and titanium tetrachloride. Inter J Hydrogen En 31:1326–1331. doi:10.1016/j.ijhydene.2005.11.016 230. Sreethawong T, Laehsalee S, Chavadej S (2009) Use of Pt/N-doped mesoporous-assembled nanocrystalline TiO2 for photocatalytic H2 production under visible light irradiation. Catal Commun 10:538–543. doi:10.1016/j.catcom.2008.10.029 231. Sreethawong T, Laehsalee S, Chavadej S (2008) Comparative investigation of mesoporous- and non-mesoporous-assembled TiO2 nanocrystals for photocatalytic H2 production over N-doped TiO2 under visible light irradiation. Inter J Hydrogen En 33:5947–5957. doi:10. 1016/j.ijhydene.2008.08.007 232. Liu SH, Syu HR (2013) High visible-light photocatalytic hydrogen evolution of C, N-codoped mesoporous TiO2 nanoparticles prepared via an ionic-liquid-template approach. Inter J Hydrogen En 38:13856–13865. doi:10.1016/j.ijhydene.2013.08.094 233. Kim H, Monllor-Satoca D, Kim W, Choi W (2015) N-doped TiO2 nanotubes coated with a thin TaOxNy layer for photoelectrochemical water splitting: dual bulk and surface modification of photoanodes. En Environ Sci 8:247–257. doi:10.1039/C4EE02169J 234. Lin WC, Yang WD, Huang IL et al (2009) Hydrogen production from methanol/water photocatalytic decomposition using Pt/TiO2−xNx catalyst. En Fuels 23:2192–2196 235. Pei F, Liu Y, Xu S et al (2013) Nanocomposite of graphene oxide with nitrogen-doped TiO2 exhibiting enhanced photocatalytic efficiency for hydrogen evolution. Inter J Hydrogen En 38:2670–2677. doi:10.1016/j.ijhydene.2012.12.045 236. Pei F, Xu S, Zuo W et al (2014) Effective improvement of photocatalytic hydrogen evolution via a facile in-situ solvothermal N-doping strategy in N-TiO2/N-graphene nanocomposite. Inter J Hydrogen En 39:6845–6852. doi:10.1016/j.ijhydene.2014.02.173 237. Hara M, Nunoshige J, Takata T et al (2003) Unusual enhancement of H2 evolution by Ru on TaON photocatalyst under visible light irradiation. Chem Commun 3000–3001. doi:10. 1039/B309935K 106 2 Semiconductor-Based Photocatalytic Systems …

238. Hara M, Hitoki G, Takata T et al (2003) TaON and Ta3N5 as new visible light driven photocatalysts. Catal Today 78:555–560. doi:10.1016/S0920-5861(02)00354-1 239. Suzuki TM, Nakamura T, Saeki S et al (2012) Visible light-sensitive mesoporous N-doped Ta2O5 spheres: synthesis and photocatalytic activity for hydrogen evolution and CO2 reduction. J Mater Chem 22:24584–24590. doi:10.1039/C2JM33980C 240. Mishima T, Matsuda M, Miyake M (2007) Visible-light photocatalytic properties and electronic structure of Zr-based oxynitride, Zr2ON2, derived from nitridation of ZrO2. Appl Catal A 324:77–82. doi:10.1016/j.apcata.2007.03.017 241. Maeda K, Terashima H, Kase K et al (2008) Surface modification of TaON with monoclinic ZrO2 to produce a composite photocatalyst with enhanced hydrogen evolution activity under visible light. Bull Chem Soc Jpn 81:927–937. doi:10.1246/bcsj.81.927 242. Liu M, You W, Lei Z et al (2004) Water reduction and oxidation on Pt–Ru/Y2Ta2O5N2 catalyst under visible light irradiation. Chem Commun 2192–2193. doi:10.1039/B407892F 243. Ji SM, Borse PH, Kim HG et al (2005) Photocatalytic hydrogen production from water– methanol mixtures using N-doped Sr2Nb2O7 under visible light irradiation: effects of catalyst structure. Phys Chem Chem Phys 7:1315–1321. doi:10.1039/B417052K 244. Maeda K, Teramura K, Domen K (2008) Effect of post-calcination on photocatalytic activity of (Ga1−xZnx)(N1−xOx) solid solution for overall water splitting under visible light. J Catal 254:198–204. doi:10.1016/j.jcat.2007.12.009 245. Maeda K, Teramura K, Lu D et al (2007) Roles of Rh/Cr2O3 (core/shell) nanoparticles photodeposited on visible-light-responsive (Ga1−xZnx)(N1−xOx) solid solutions in photocatalytic overall water splitting. J Phys Chem C 111:7554–7560. doi:10.1021/jp071056j 246. Maeda K, Teramura K, Lu D et al (2006) Characterization of Rh–Cr mixed-oxide nanoparticles dispersed on (Ga1−xZnx)(N1−xOx) as a cocatalyst for visible-light-driven overall water splitting. J Phys Chem B 110:13753–13758. doi:10.1021/jp061829o 247. Teramura K, Maeda K, Saito T et al (2005) Characterization of ruthenium oxide nanocluster as a cocatalyst with (Ga1−xZnx)(N1−xOx) for photocatalytic overall water splitting. J Phys Chem B 109:21915–21921. doi:10.1021/jp054313y 248. Maeda K, Teramura K, Takata T et al (2005) Overall water splitting on (Ga1−xZnx)(N1−xOx) solid solution photocatalyst: relationship between physical properties and photocatalytic activity. J Phys Chem B 109:20504–20510. doi:10.1021/jp053499y 249. Lee Y, Terashima H, Shimodaira Y et al (2007) Zinc germanium oxynitride as a photocatalyst for overall water splitting under visible light. J Phys Chem C 111:1042–1048. doi:10.1021/jp0656532 250. Reyes-Gil KR, Reyes-Garcia EA, Raftery D (2007) Nitrogen-doped In2O3 thin film electrodes for photocatalytic water splitting. J Phys Chem C 111:14579–14588. doi:10.1021/ jp072831y 251. Mohapatra SK, Misra M, Mahajan VK, Raja KS (2007) Design of a highly efficient photoelectrolytic cell for hydrogen generation by water splitting: application of TiO2−xCx nanotubes as a photoanode and Pt/TiO2 nanotubes as a cathode. J Phys Chem C 111:8677– 8685. doi:10.1021/jp071906v 252. Park JH, Kim S, Bard AJ (2006) Novel carbon-doped TiO2 nanotube arrays with high aspect ratios for efficient solar water splitting. Nano Lett 6:24–28. doi:10.1021/nl051807y 253. Liu Z, Pesic B, Raja KS et al (2009) Hydrogen generation under sunlight by self ordered TiO2 nanotube arrays. Inter J Hydrogen En 34:3250–3257. doi:10.1016/j.ijhydene.2009.02. 044 254. Shaban YA, Khan SU (2008) Visible light active carbon modified n-TiO2 for efficient hydrogen production by photoelectrochemical splitting of water. Inter J Hydrogen En 33:1118–1126. doi:10.1016/j.ijhydene.2007.11.026 255. Shaban YA, Khan SU (2009) Carbon modified (CM)-n-TiO2 thin films for efficient water splitting to H2 and O2 under xenon lamp light and natural sunlight illuminations. J Sol State Electrochem 13:1025–1036. doi:10.1007/s10008-009-0823-4 References 107

256. Randeniya LK, Murphy AB, Plumb IC (2008) A study of S-doped TiO2 for photoelectrochemical hydrogen generation from water. J Mater Sci 43:1389–1399. doi:10.1007/ s10853-007-2309-z 257. Ogisu K, Ishikawa A, Teramura K et al (2007) Lanthanum-indium oxysulfide as a visible light driven photocatalyst for water splitting. Chem Lett 36:854–855. doi:10.1246/cl.2007. 854 258. Ryu SY, Balcerski W, Lee TK, Hoffmann MR (2007) Photocatalytic production of hydrogen from water with visible light using hybrid catalysts of CdS attached to microporous and mesoporous silicas. J Phys Chem C 111:18195–18203. doi:10.1021/jp074860e 259. Hirai T, Nanba M, Komasawa I (2003) Dithiol-mediated incorporation of CdS nanoparticles from reverse micellar system into Zn-doped SBA-15 mesoporous silica and their photocatalytic properties. J Colloid Interface Sci 268:394–399. doi:10.1016/j.jcis.2003.09.011 260. Khatamian M, Oskoui MS, Haghighi M (2014) Photocatalytic hydrogen generation over CdS–metalosilicate composites under visible light irradiation. New J Chem 38:1684–1693. doi:10.1039/C3NJ01348K 261. Peng R, Zhao D, Baltrusaitis J et al (2012) Visible light driven photocatalytic evolution of hydrogen from water over CdS encapsulated MCM-48 materials. RSC Adv 2:5754–5767. doi:10.1039/C2RA20714A 262. Peng R, Wu CM, Baltrusaitis J et al (2013) Ultra-stable CdS incorporated Ti-MCM-48 mesoporous materials for efficient photocatalytic decomposition of water under visible light illumination. Chem Commun 49:3221–3223. doi:10.1039/C3CC41362D 263. Peng R, Lin C, Baltrusaitis J et al (2014) Insight into band positions and inter-particle electron transfer dynamics between CdS nanoclusters and spatially isolated TiO2 dispersed in cubic MCM-48 mesoporous materials: a highly efficient system for photocatalytic hydrogen evolution under visible light illumination. Phys Chem Chem Phys 16:2048–2061. doi:10.1039/C3CP52801D 264. Henglein A, Gutierrez M (1983) Photochemistry of colloidal metal sulfides. 5. Fluorescence and chemical reactions of ZnS and ZnS/CdS co-colloids. Ber Bunsenges Phys Chem 87:852–858. doi:10.1002/bbpc.19830871005 265. Deshpande A, Shah P, Gholap RS, Gupta NM (2009) Interfacial and physico-chemical properties of polymer-supported CdS-ZnS nanocomposites and their role in the visible-light mediated photocatalytic splitting of water. J Colloid Interface Sci 333:263–268. doi:10.1016/ j.jcis.2009.01.037 266. Lunawat PS, Senapati S, Kumar R, Gupta NM (2007) Visible light-induced splitting of water using CdS nanocrystallites immobilized over water-repellant polymeric surface. Inter J Hydrogen En 32:2784–2790. doi:10.1016/j.ijhydene.2007.04.001 267. Kim YK, Kim M, Hwang SH et al (2015) CdS-loaded flexible carbon nanofiber mats as a platform for solar hydrogen production. Inter J Hydrogen En 40:136–145. doi:10.1016/j. ijhydene.2014.11.011 268. Hirai T, Bando Y, Komasawa I (2002) Immobilization of CdS nanoparticles formed in reverse micelles onto alumina particles and their photocatalytic properties. J Phys Chem B 106:8967–8970. doi:10.1021/jp020386v 269. Hirai T, Nanba M, Komasawa I (2002) Dithiol-mediated immobilization of CdS nanoparticles from reverse micellar system onto Zn-doped silica particles and their high photocatalytic activity. J Colloid Interface Sci 252:89–92. doi:10.1006/jcis.2002.8430 270. Yu G, Geng L, Wu S et al (2015) Highly-efficient cocatalyst-free H2-evolution over silica-supported CdS nanoparticle photocatalysts under visible light. Chem Commun 51:10676–10679. doi:10.1039/C5CC02249E 271. Yu G, Zhang W, Sun Y et al (2016) A highly active cocatalyst-free semiconductor photocatalyst for visible-light-driven hydrogen evolution: synergistic effect of surface defects and spatial bandgap engineering. J Mater Chem A 4:13803–13808. doi:10.1039/ c6ta03803d 108 2 Semiconductor-Based Photocatalytic Systems …

272. Bb Kale, Baeg JO, Apte SK et al (2007) Confinement of nano CdS in designated glass: a novel functionality of quantum dot–glass nanosystems in solar hydrogen production. J Mater Chem 17:4297–4303. doi:10.1039/B708269J 273. Apte SK, Garaje SN, Valant M, Kale BB (2012) Eco-friendly solar light driven hydrogen production from copious waste H2S and organic dye degradation by stable and efficient orthorhombic CdS quantum dots–GeO2 glass photocatalyst. Green Chem 14:1455–1462. doi:10.1039/C2GC16416G 274. Kanade KG, Baeg JO, Mulik UP et al (2006) Nano-CdS by polymer-inorganic solid-state reaction: visible light pristine photocatalyst for hydrogen generation. Mater Res Bull 41:2219–2225. doi:10.1016/j.materresbull.2006.04.031 275. Li W, O’Dowd G, Whittles TJ et al (2015) Colloidal dual-band gap cell for photocatalytic hydrogen generation. Nanoscale 7:16606–16610. doi:10.1039/c5nr04950d 276. Baldovi HG, Latorre-Sanchez M, Esteve-Adell I et al (2016) Generation of MoS2 quantum dots by laser ablation of MoS2 particles in suspension and their photocatalytic activity for H2 generation. J Nanopart Res 18:240–248. doi:10.1007/s11051-016-3540-9 277. Hirai T, Nomura Y, Komasawa I (2003) Immobilization of RuS2 nanoparticles prepared in reverse micellar system onto thiol-modified polystyrene particles and their photocatalytic properties. J Nanoparticle Res 5:61–67. doi:10.1023/A:1024422226598 278. Barawi M, Ferrer IJ, Flores E et al (2016) Hydrogen photoassisted generation by visible light and an earth abundant photocatalyst: pyrite (FeS2). J Phys Chem C 120:9547–9552. doi:10. 1021/acs.jpcc.5b11482 279. Li G, Su R, Rao J et al (2016) Band gap narrowing of SnS2 superstructures with improved hydrogen production. J Mater Chem A 4:209–216. doi:10.1039/c5ta07283b 280. Shen S, Zhao L, Guo L (2008) Cetyltrimethylammoniumbromide (CTAB)-assisted hydrothermal synthesis of ZnIn2S4 as an efficient visible-light-driven photocatalyst for hydrogen production. Inter J Hydrogen En 33:4501–4510. doi:10.1016/j.ijhydene.2008.05. 043 281. Lei Z, You W, Liu M et al (2003) Photocatalytic water reduction under visible light on a novel ZnIn2S4 catalyst synthesized by hydrothermal method. Chem Commun 2142–2143. doi:10.1039/B306813G 282. Guijun M, Hongjian Y, Xu Z et al (2008) Photocatalytic splitting of H2S to produce hydrogen by gas-solid phase reaction. Chin J Catal 29:313–315. doi:10.1016/S1872-2067 (08)60029-7 283. Shen S, Zhao L, Guo L (2009) Crystallite, optical and photocatalytic properties of visible-light-driven ZnIn2S4 photocatalysts synthesized via a surfactant-assisted hydrothermal method. Mater Res Bull 44:100–105. doi:10.1016/j.materresbull.2008.03.027 284. Shen J, Zai J, Yuan Y, Qian X (2012) 3D hierarchical ZnIn2S4: the preparation and photocatalytic properties on water splitting. Inter J Hydrogen En 37:16986–16993. doi:10. 1016/j.ijhydene.2012.08.038 285. Fan WJ, Zhou ZF, Xu WB et al (2010) Preparation of ZnIn2S4/fluoropolymer fiber composites and its photocatalytic H2 evolution from splitting of water using Xe lamp irradiation. Inter J Hydrogen En 35:6525–6530. doi:10.1016/j.ijhydene.2010.04.036 286. Li Y, Zhang K, Peng S et al (2012) Photocatalytic hydrogen generation in the presence of ethanolamines over Pt/ZnIn2S4 under visible light irradiation. J Mol Catal A 363–364:354– 361. doi:10.1016/j.molcata.2012.07.011 287. Chaudhari NS, Bhirud AP, Sonawane RS et al (2011) Ecofriendly hydrogen production from abundant hydrogen sulfide using solar light-driven hierarchical nanostructured ZnIn2S4 photocatalyst. Green Chem 13:2500–2506. doi:10.1039/C1GC15515F 288. Li F, Luo J, Chen G et al (2014) Hydrothermal synthesis of zinc indium sulfide microspheres + with Ag doping for enhanced H2 production by photocatalytic water splitting under visible light. Catal Sci Technol 4:1144–1150. doi:10.1039/C3CY00952A References 109

289. Shang L, Zhou C, Bian T et al (2013) Facile synthesis of hierarchical ZnIn2S4 submicrospheres composed of ultrathin mesoporous nanosheets as a highly efficient visible-light-driven photocatalyst for H2 production. J Mater Chem A 1:4552–4558. doi:10. 1039/C3TA01685D 290. Xu Z, Li Y, Peng S et al (2012) NaCl-assisted low temperature synthesis of layered Zn-In-S photocatalyst with high visible-light activity for hydrogen evolution. RSC Adv 2:3458– 3466. doi:10.1039/C2RA01159J 291. Chen Y, He J, Li J et al (2016) Hydrilla derived ZnIn2S4 photocatalyst with hexagonal-cubic phase junctions: A bio-inspired approach for H2 evolution. Catal Commun 87:1–5. doi:10. 1016/j.catcom.2016.08.031 292. Tian F, Zhu R, Zhong J et al (2016) An efficient preparation method of RGO/ZnIn2S4 for photocatalytic hydrogen generation under visible light. Inter J Hydrogen En 41:20156– 20171. doi:10.1016/j.ijhydene.2016.08.063 293. Shen S, Zhao L, Zhou Z, Guo L (2008) Enhanced photocatalytic hydrogen evolution over Cu-doped ZnIn2S4 under visible light irradiation. J Phys Chem C 112:16148–16155. doi:10. 1021/jp804525q 294. Ding J, Sun S, Yan W et al (2013) Photocatalytic H2 evolution on a novel CaIn2S4 photocatalyst under visible light irradiation. Inter J Hydrogen En 38:13153–13158. doi:10. 1016/j.ijhydene.2013.07.109 295. Ding J, Li X, Chen L et al (2016) Au–Pt alloy nanoparticles site-selectively deposited on CaIn2S4 nanosteps as efficient photocatalysts for hydrogen production. J Mater Chem A 4:12630–12637. doi:10.1039/c6ta04468a 296. Zhou Q, Kang SZ, Li X et al (2015) AgGaS2 nanoplates loaded with CuS: an efficient visible photocatalyst for rapid H2 evolution. Inter J Hydrogen En 40:4119–4128. doi:10.1016/j. ijhydene.2015.01.143 297. Iwase A, Ng YH, Amal R, Kudo A (2015) Solar hydrogen evolution using a CuGaS2 photocathode improved by incorporating reduced graphene oxide. J Mater Chem A 3:8566– 8570. doi:10.1039/C5TA01237F 298. Yu X, An X, Shavel A et al (2014) The effect of the Ga content on the photocatalytic hydrogen evolution of CuIn1−xGaxS2 nanocrystals. J Mater Chem A 2:12317–12322. doi:10. 1039/C4TA01315H 299. Kato T, Hakari Y, Ikeda S et al (2015) Utilization of metal sulfide material of (CuGa)1 −xZn2xS2 solid solution with visible light response in photocatalytic and photoelectrochemical solar water splitting systems. J Phys Chem Lett 6:1042–1047. doi:10.1021/acs.jpclett. 5b00137 300. Kandiel TA, Huttona GA, Reisner E (2016) Visible light driven hydrogen evolution with a noble metal free CuGa2In3S8 nanoparticle system in water. Catal Sci Technol 6:6536–6541. doi:10.1039/c6cy01103a 301. Quintans CS, Kato H, Kobayashi M et al (2015) Improvement of hydrogen evolution under visible light over Zn1−2x(CuGa)xGa2S4 photocatalysts by synthesis utilizing a polymerizable complex method. J Mater Chem A 3:14239–14244. doi:10.1039/C5TA02114F 302. Chen F, Zai J, Xu M, Qian X (2013) 3D-hierarchical Cu3SnS4 flowerlike micro-spheres: controlled synthesis, formation mechanism and photocatalytic activity for H2 evolution from water. J Mater Chem A 1:4316–4323. doi:10.1039/C3TA01491F 303. Kush P, Deori K, Kumar A, Deka S (2015) Efficient hydrogen/oxygen evolution and photocatalytic dye degradation and reduction of aqueous Cr(VI) by surfactant free hydrophilic Cu2ZnSnS4 nanoparticles. J Mater Chem A 3:8098–8106. doi:10.1039/ C4TA06551D 304. Yu X, Shavel A, An X et al (2014) Cu2ZnSnS4-Pt and Cu2ZnSnS4-Au heterostructured nanoparticles for photocatalytic water splitting and pollutant degradation. J Am Chem Soc 136:9236–9239. doi:10.1021/ja502076b 305. Gonce MK, Dogru M, Aslan E et al (2015) Photocatalytic hydrogen evolution based on Cu2ZnSnS4,Cu2ZnSnSe4 and Cu2ZnSnSe4−xSx nanofibers. RSC Adv 5:94025–94028. doi:10.1039/c5ra18877f 110 2 Semiconductor-Based Photocatalytic Systems …

306. Yu X, An X, Genç A et al (2015) Cu2ZnSnS4–PtM (M = Co, Ni) nanoheterostructures for photocatalytic hydrogen evolution. J Phys Chem C 119:21882–21888. doi:10.1021/acs.jpcc. 5b06199 307. Zhang ZX, Chong RF, Meng YN et al (2015) High temperature recrystallization of kersterite Cu2ZnSnS4 towards enhanced photocatalytic H2 evolution. Inter J Hydrogen En 40:13456– 13462. doi:10.1016/j.ijhydene.2015.08.032 308. Zheng L, Xu Y, Song Y et al (2009) Nearly monodisperse CuInS2 hierarchical micro-architectures for photocatalytic H2 evolution under visible light. Inorg Chem 48:4003–4009. doi:10.1021/ic802399f 309. Gannouni M, Assaker IB, Chtourou R (2015) Photoelectrochemical cell based on n-CuIn5S8 film as photoanodes for photocatalytic water splitting. Inter J Hydrogen En 40:7252–7259. doi:10.1016/j.ijhydene.2015.04.057 310. Guan Z, Luo W, Feng J et al (2015) Selective etching of metastable phase induced an efficient CuIn0.7Ga0.3S2 nano-photocathode for solar water splitting. J Mater Chem A 3:7840–7848. doi:10.1039/C5TA01259G 311. Hu P, Ngaw CK, Tay YY et al (2015) A “uniform” heterogeneous photocatalyst: integrated p–n type CuInS2/NaInS2 nanosheets by partial ion exchange reaction for efficient H2 evolution. Chem Commun 51:9381–9384. doi:10.1039/C5CC02237A 312. Zhang X, Du Y, Zhou Z, Guo L (2010) A simplified method for synthesis of band-structure-controlled (CuIn)xZn2(1−x)S2 solid solution photocatalysts with high activity of photocatalytic H2 evolution under visible-light irradiation. Inter J Hydrogen En 35:3313– 3321. doi:10.1016/j.ijhydene.2010.01.111 313. Zhang X, Yang M, Zhao J, Guo L (2013) Photocatalytic hydrogen evolution with simultaneous degradation of organics over (CuIn)0.2Zn1.6S2 solid solution. Inter J Hydrogen En 38:15985–15991. doi:10.1016/j.ijhydene.2013.10.014 314. Huang Y, Chen J, Zou W et al (2015) Enhanced photocatalytic hydrogen evolution efficiency using hollow microspheres of (CuIn)xZn2(1−x)S2 solid solutions. Dalton Trans 44:10991–10996. doi:10.1039/C5DT01269D 315. Lin y, Zhang F, Pan D (2012) A facile route to (ZnS)x(CuInS2)1−x hierarchical microspheres with excellent water-splitting ability. J Mater Chem 22:22619–22623. doi:10.1039/ C2JM35166H 316. Tang X, Tay Q, Chen Z et al (2013) Cu–In–Zn–S nanoporous spheres for highly efficient visible-light-driven photocatalytic hydrogen evolution. New J Chem 37:1878–1882. doi:10. 1039/C3NJ00266G 317. Xu M, Zai J, Yuan Y, Qian X (2012) Band gap-tunable (CuIn)xZn2(1−x)S2 solid solutions: preparation and efficient photocatalytic hydrogen production from water under visible light without noble metals. J Mater Chem 22:23929–23934. doi:10.1039/C2JM35375J 318. Tang X, Tay Q, Chen Z et al (2013) CuInZnS-decorated graphene nanosheets for highly efficient visible-light-driven photocatalytic hydrogen production. J Mater Chem A 1:6359– 6365. doi:10.1039/C3TA01602A 319. Zhang G, Zhang W, Wang P et al (2013) Stability of an H2-producing photo-catalyst (Ru/ (CuAg)0.15In0.3Zn1.4S2) in aqueous solution under visible light irradiation. Inter J Hydrogen En 38:1286–1296. doi:10.1016/j.ijhydene.2012.11.033 320. Zhang G, Zhang W, Minakata D et al (2013) The pH effects on H2 evolution kinetics for visible light water splitting over the Ru/(CuAg)0.15In0.3Zn1.4S2 photocatalyst. Inter J Hydrogen En 38:11727–11736. doi:10.1016/j.ijhydene.2013.06.140 321. Tsuji I, Kato H, Kudo A (2006) Photocatalytic hydrogen evolution on ZnS − CuInS2 − AgInS2 solid solution photocatalysts with wide visible light absorption bands. Chem Mater 18:1969–1975. doi:10.1021/cm0527017 322. Tsuji I, Kato H, Kobayashi H, Kudo A (2005) Photocatalytic H2 evolution under visible-light irradiation over band-structure-controlled (CuIn)xZn2(1−x)S2 solid solutions. J Phys Chem B 109:7323–7329. doi:10.1021/jp044722e References 111

323. Li Y, Chen G, Zhou C, Sun J (2009) A simple template-free synthesis of nanoporous ZnS– In2S3–Ag2S solid solutions for highly efficient photocatalytic H2 evolution under visible light. Chem Commun 2020–2022. doi:10.1039/B819300B 324. Kudo A, Tsuji I, Kato H (2002) AgInZn7S9 solid solution photocatalyst for H2 evolution from aqueous solutions under visible light irradiation. Chem Commun 1958–1959. doi:10. 1039/B204259B 325. Kale BB, Baeg JO, Lee SM et al (2006) CdIn2S4 nanotubes and “marigold” nanostructures: a visible-light photocatalyst. Adv Func Mater 16:1349–1354. doi:10.1002/adfm.200500525 326. Yu Y, Chen G, Wang G, Lv Z (2013) Visible-light-driven ZnIn2S4/CdIn2S4 composite photocatalyst with enhanced performance for photocatalytic H2 evolution. Inter J Hydrogen En 38:1278–1285. doi:10.1016/j.ijhydene.2012.11.020 327. Chen X, Li L, Zhang W et al (2016) Fabricate globular flower-like CuS/CdIn2S4/ZnIn2S4 with high visible light response via microwave-assisted one − step method and its multipathway photoelectron migration properties for hydrogen evolution and pollutant degradation. ACS Sustainable Chem Eng 4:6680–6688. doi:10.1021/acssuschemeng. 6b01543 328. Mei Z, Ouyang S, Tang DM et al (2013) An ion-exchange route for the synthesis of hierarchical In2S3/ZnIn2S4 bulk composite and its photocatalytic activity under visible-light irradiation. Dalton Trans 42:2687–2690. doi:10.1039/C2DT32271D 329. Hou J, Yang C, Cheng H et al (2013) Ternary 3D architectures of CdS QDs/graphene/ ZnIn2S4 heterostructures for efficient photocatalytic H2 production. Phys Chem Chem Phys 15:15660–15668. doi:10.1039/C3CP51857D 330. Chen D, Ye J (2007) Photocatalytic H2 evolution under visible light irradiation on AgIn5S8 photocatalyst. J Phys Chem Sol 68:2317–2320. doi:10.1016/j.jpcs.2007.07.059 331. Kudo A, Nagane A, Tsuji I, Kato H (2002) H2 evolution from aqueous potassium sulfite solutions under visible light irradiation over a novel sulfide photocatalyst NaInS2 with a layered structure. Chem Lett 31:882–883. doi:10.1246/cl.2002.882 332. Shen S, Guo L (2006) Structural, textural and photocatalytic properties of quantum-sized In2S3-sensitized Ti-MCM-41 prepared by ion-exchange and sulfidation methods. J Sol State Chem 179:2629–2635. doi:10.1016/j.jssc.2006.05.010 333. Li S, Wang C, Qiu H (2015) Single- and few-layer ZrS2 as efficient photocatalysts for hydrogen production under visible light. Inter J Hydrogen En 40:15503–15509. doi:10.1016/ j.ijhydene.2015.08.110 334. Wang F, Shifa TA, Zhan X et al (2015) Recent advances in transition-metal dichalcogenide based nanomaterials for water splitting. Nanoscale 7:19764–19788. doi:10.1039/c5nr06718a 335. Lin Z, Ning S, Yang Z et al (2016) Large-scale preparation of heterometallic chalcogenide MnSb2S4 monolayer nanosheets with a high visible-light photocatalytic activity for H2 evolution. Chem Commun 52:13381–13384. doi:10.1039/c6cc07127a 336. Bessekhouad Y, Mohammedi M, Trari M (2002) Hydrogen photoproduction from hydrogen sulfide on Bi2S3 catalyst. Sol En Mater Sol Cells 73:339–350. doi:10.1016/S0927-0248(01) 00218-5 337. Abdi A, Denoyelle A, Commenges-Bernole M, Trari M (2013) Photocatalytic hydrogen evolution on new mesoporous material Bi2S3/Y-zeolite. Inter J Hydrogen En 38:2070–2078. doi:10.1016/j.ijhydene.2012.11.085 338. Zhang K, Jing D, Xing C, Guo L (2007) Significantly improved photocatalytic hydrogen production activity over Cd1−xZnxS photocatalysts prepared by a novel thermal sulfuration method. Inter J Hydrogen En 32:4685–4691. doi:10.1016/j.ijhydene.2007.08.022 339. del Valle F, Ishikawa A, Domen K et al (2009) Influence of Zn concentration in the activity of Cd1−xZnxS solid solutions for water splitting under visible light. Catal Today 143:51–56. doi:10.1016/j.cattod.2008.09.024 340. Chan CC, Chang CC, Hsu CH et al (2014) Efficient and stable photocatalytic hydrogen production from water splitting over ZnxCd1−xS solid solutions under visible light irradiation. Inter J Hydrogen En 39:1630–1639. doi:10.1016/j.ijhydene.2013.11.059 112 2 Semiconductor-Based Photocatalytic Systems …

341. Du H, Liang K, Yuan CZ et al (2016) Bare Cd1−xZnxS ZB/WZ heterophase nano-junctions for visible light photocatalytic hydrogen production with high efficiency. ACS Appl Mater Interfaces 8:24550–24558. doi:10.1021/acsami.6b06182 342. Roy AM, De GC (2003) Immobilisation of CdS, ZnS and mixed ZnS–CdS on filter paper: effect of hydrogen production from alkaline Na2S/Na2S2O3 solution. J Photochem Photobiol A 157:87–92. doi:10.1016/S1010-6030(02)00430-6 343. Zhang X, Jing D, Liu M, Guo L (2008) Efficient photocatalytic H2 production under visible light irradiation over Ni doped Cd1−xZnxS microsphere photocatalysts. Catal Commun 9:1720–1724. doi:10.1016/j.catcom.2008.01.032 344. Stroyuk AL, Raevskaya AE, Korzhak AV et al (2009) Photocatalytic production of 0 hydrogen in systems based on CdxZn1−xS/Ni nanostructures. Theoret Exp Chem 45:12–22 345. Kimi M, Yuliati L, Shamsuddin M (2011) Photocatalytic hydrogen production under visible light over Cd0.1SnxZn0.9−2xS solid solution photocatalysts. Inter J Hydrogen En 36:9453– 9461. doi:10.1016/j.ijhydene.2011.05.044 346. Peng S, An R, Li Y et al (2012) Remarkable enhancement of photocatalytic hydrogen evolution over Cd0.5Zn0.5S by bismuth-doping. Inter J Hydrogen En 37:1366–1374. doi:10. 1016/j.ijhydene.2011.09.140 347. Sathish M, Viswanath RP (2007) Photocatalytic generation of hydrogen over mesoporous CdS nanoparticle: effect of particle size, noble metal and support. Catal Today 129:421–427. doi:10.1016/j.cattod.2006.12.008 348. Li Y, Du Y, Peng S et al (2008) Enhancement of photocatalytic activity of cadmium sulfide for hydrogen evolution by photoetching. Inter J Hydrogen En 33:2007–2013. doi:10.1016/j. ijhydene.2008.02.023 349. Chang CM, Orchard KL, Martindale CM, Reisner E (2016) Ligand removal from CdS quantum dots for enhanced photocatalytic H2 generation in pH neutral water. J Mater Chem A 4:2856–2862. doi:10.1039/c5ta04136h 350. Jana MK, Gupta U, Rao CNR (2016) Hydrazine as a hydrogen carrier in the photocatalytic generation of H2 using CdS quantum dots. Dalton Trans 45:15137–15141. doi:10.1039/ c6dt02505f 351. Silva LA, Ryu SY, Choi J et al (2008) Photocatalytic hydrogen production with visible light over Pt-interlinked hybrid composites of cubic-phase and hexagonal-phase CdS. J Phys Chem C 112:12069–12073. doi:10.1021/jp8037279 352. Li K, Han M, Chen R et al (2016) Hexagonal@Cubic CdS Core@Shell nanorod photocatalyst for highly active production of H2 with unprecedented stability. Adv Mater 28:8906–8911. doi:10.1002/adma.201601047 353. Baran MP, Korsunskaya NE, Stara TR et al (2016) Graded ZnS/ZnSxO1−x heterostructures produced by oxidative photolysis of zinc sulfide: structure, optical properties and photocatalytic evolution of molecular hydrogen. J Photochem Photobiol A 329:213–220. doi:10.1016/j.jphotochem.2016.07.003 354. Jang JS, Joshi UA, Lee JS (2007) Solvothermal synthesis of CdS nanowires for photocatalytic hydrogen and electricity production. J Phys Chem C 111:13280–13287. doi:10.1021/jp072683b 355. Kida T, Guan G, Minami Y et al (2003) Photocatalytic hydrogen production from water over a LaMnO3/CdS nanocomposite prepared by the reverse micelle method. J Mater Chem 13:1186–1191. doi:10.1039/B211812B 356. Kida T, Guan G, Yamada N et al (2004) Hydrogen production from sewage sludge solubilized in hot-compressed water using photocatalyst under light irradiation. Inter J Hydrogen En 29:269–274. doi:10.1016/j.ijhydene.2003.08.007 357. Ma G, Yan H, Shi J et al (2008) Direct splitting of H2S into H2 and S on CdS-based photocatalyst under visible light irradiation. J Catal 260:134–140. doi:10.1016/j.jcat.2008. 09.017 358. Frame FA, Carroll EC, Larsen DS et al (2008) First demonstration of CdSe as a photocatalyst for hydrogen evolution from water under UV and visible light. Chem Commun 2206–2208. doi:10.1039/B718796C References 113

359. Holmes MA, Townsend TK, Osterloh FE (2012) Quantum confinement controlled photocatalytic water splitting by suspended CdSe nanocrystals. Chem Commun 48:371– 373. doi:10.1039/C1CC16082F 360. Grigioni I, Bernareggi M, Sinibaldi G et al (2016) Size-dependent performance of CdSe quantum dots in the photocatalytic evolution of hydrogen under visible light irradiation. Appl Catal A 518:176–180. doi:10.1016/j.apcata.2015.09.021 361. Zhao J, Holmes MA, Osterloh FE (2013) Quantum confinement controls photocatalysis: a free energy analysis for photocatalytic proton reduction at CdSe nanocrystals. ACS Nano 7:4316–4325. doi:10.1021/nn400826h 362. Rasamani KD, Li Z, Sun Y (2016) Significant enhancement of photocatalytic water splitting enabled by elimination of surface traps in Pt-tipped CdSe nanorods Nanoscale 8:18621– 18625. doi:10.1039/c6nr06902a 363. Costi R, Young ER, Bulović V, Nocera DG (2013) Stabilized CdSe-CoPi composite photoanode for light-assisted water oxidation by transformation of a CdSe/Cobalt metal thin film. ACS Appl Mater Interfaces 5:2364–2367. doi:10.1021/am400364u 364. Yang S, Xu CY, Yang L et al (2016) Solution-phase synthesis of g-In2Se3 nanoparticles for highly efficient photocatalytic hydrogen generation under simulated sunlight irradiation. RSC Adv 6:106671–106675. doi:10.1039/c6ra21784b 365. Gurunathan K, Baeg JO, Lee SM et al (2008) Visible light active pristine and Fe3+ doped CuGa2O4 spinel photocatalysts for solar hydrogen production. Inter J Hydrogen En 33:2646–2652. doi:10.1016/j.ijhydene.2008.03.018 366. Saadi S, Bouguelia A, Derbal A, Trari M (2007) Hydrogen photoproduction over new catalyst CuLaO2. J Photochem Photobiol A 187:97–104. doi:10.1016/j.jphotochem.2006.09. 017 367. Koriche N, Bouguelia A, Trari M (2006) Photocatalytic hydrogen production over new oxide CuLaO2.62. Inter J Hydrogen En 31:1196–1203. doi:10.1016/j.ijhydene.2005.08.015 368. Koriche N, Bouguelia A, Aider A, Trari M (2005) Photocatalytic hydrogen evolution over delafossite CuAlO2. Inter J Hydrogen En 30:693–699. doi:10.1016/j.ijhydene.2004.06.011 369. Zhou C, Zhao Y, Shang L et al (2016) Facile synthesis of ultrathin SnNb2O6 nanosheets towards improved visible-light photocatalytic H2-production activity. Chem Commun 52:8239–8242. doi:10.1039/c6cc03739a 370. Ye J, Zou Z, Arakawa H et al (2002) Correlation of crystal and electronic structures with photophysical properties of water splitting photocatalysts InMO4 (M = V5+,Nb5+,Ta5+). J Photochem Photobiol A 148:79–83. doi:10.1016/S1010-6030(02)00074-6 371. Zhu H, Fang M, Hyang Z et al (2016) Novel carbon-incorporated porous ZnFe2O4 nanospheres for enhanced photocatalytic hydrogen generation under visible light irradiation. RSC Adv 6:56069–56076. doi:10.1039/c6ra05098k 372. Pai YH, Tsai CT, Fang SY (2013) Enhanced photocatalytic hydrogen generation with Pt Nanoparticles on multi-phase polycrystalline microporous MnO2 photocatalyst. J Power Sources 223:107–113. doi:10.1016/j.jpowsour.2012.09.024 373. Jin S, Wang X, Ju M et al (2015) Effect of phase junction structure on the photocatalytic performance in overall water splitting: Ga2O3 photocatalyst as an example. J Phys Chem C 119:18221–18228. doi:10.1021/acs.jpcc.5b04092 374. Hu C, Chu K, Zhao Y, Teoh WY (2014) Efficient photoelectrochemical water splitting over anodized p-Type NiO porous films. ACS Appl Mater Interfaces 6:18558–18568. doi:10. 1021/am507138b 375. Manikandan M, Tanabe T, Li P et al (2014) Photocatalytic water splitting under visible light by mixed-valence Sn3O4. ACS Appl Mater Interfaces 6:3790–3793. doi:10.1021/ am500157u 376. Tijare SN, Joshi MV, Padole PS et al (2012) Photocatalytic hydrogen generation through water splitting on nano-crystalline LaFeO3 perovskite. Inter J Hydrogen En 37:10451– 10456. doi:10.1016/j.ijhydene.2012.01.120 114 2 Semiconductor-Based Photocatalytic Systems …

377. May KJ, Fenning DP, Ming T et al (2015) Thickness-dependent photoelectrochemical water splitting on ultrathin LaFeO3 films grown on Nb:SrTiO3. J Phys Chem Lett 6:977–985. doi:10.1021/acs.jpclett.5b00169 378. Pan C, Takata T, Nakabayashi M et al (2015) A complex perovskite-type oxynitride: the first photocatalyst for water splitting operable at up to 600 nm. Angew Chem Int Ed 54:1–6. doi:10.1002/anie.201410961 379. Lee CW, Kim DW, Cho IS et al (2012) Simple synthesis and characterization of SrSnO3 nanoparticles with enhanced photocatalytic activity. Inter J Hydrogen En 37:10557–10563. doi:10.1016/j.ijhydene.2012.04.063 380. Shibli SMA, Arun PS, Raj AV (2015) Exploration of octahedrally shaped MnCo2O4 catalyst particles for visible light driven photocatalytic water splitting reaction. RSC Adv 5:19393– 19399. doi:10.1039/C4RA12646G 381. Liu J, Wen S, Zou X et al (2013) Visible-light-responsive copper(II) borate photocatalysts with intrinsic midgap states for water splitting. J Mater Chem A 1:1553–1556. doi:10.1039/ C2TA00522K 382. Luan J, Guo N, Chen B (2014) Hydrogen production with Ga2BiSbO7,Fe2BiSbO7 and Gd2BiSbO7 as photocatalysts under visible light irradiation. Inter J Hydrogen En 39:1228– 1236. doi:10.1016/j.ijhydene.2013.11.020 383. Mahapure SA, Palei PK, Nikam LK et al (2013) Novel nanocrystalline zinc silver antimonate (ZnAg3SbO4): an efficient & ecofriendly visible light photocatalyst with enhanced hydrogen generation. J Mater Chem A 1:12835–12840. doi:10.1039/ C3TA12883K 384. Guo S, Han S (2014) Constructing a novel hierarchical 3D flower-like nano/micro titanium phosphate with efficient hydrogen evolution from water splitting. J Power Sources 267:9–13. doi:10.1016/j.jpowsour.2014.05.011 385. Wang G, Jing Y, Ju J et al (2015) Ga4B2O9:anefficient borate photocatalyst for overall water splitting without cocatalyst. Inorg Chem 54:2945–2949. doi:10.1021/ic5031087 386. Liang J, Xu J, Gu Q et al (2013) A novel Zn2GeO4 superstructure for effective photocatalytic hydrogen generation. J Mater Chem A 1:7798–7805. doi:10.1039/C3TA11374D 387. Yan S, Wan L, Li Z, Zou Z (2011) Facile temperature-controlled synthesis of hexagonal Zn2GeO4 nanorods with different aspect ratios toward improved photocatalytic activity for overall water splitting and photoreduction of CO2. Chem Commun 47:5632–5634. doi:10. 1039/C1CC10513B 388. Pan B, Xie Q, Wang H et al (2013) Synthesis and photocatalytic hydrogen production of a novel photocatalyst LaCO3OH. J Mater Chem A 1:6629–6634. doi:10.1039/C3TA01553J 389. Ruiz-Gomez MA, Torres-Martinez LM, Figueroa-Torres MZ et al (2013) Hydrogen evolution from pure water over a new advanced photocatalyst Sm2GaTaO7. Inter J Hydrogen En 38:12554–12561. doi:10.1016/j.ijhydene.2012.11.131 390. Ye J, Zou Z, Oshikiri M et al (2002) A novel hydrogen-evolving photocatalyst InVO4 active under visible light irradiation. Chem Phys Lett 356:221–226. doi:10.1016/S0009-2614(02) 00254-3 391. Lv M, Liu G, Xu X (2016) Homologous compounds ZnnIn2O3+n (n = 4, 5, and 7) containing laminated functional groups as efficient photocatalysts for hydrogen production. ACS Appl Mater Interfaces 8:28700–28708. doi:10.1021/acsami.6b10951 392. Huang Q, Ye Z, Xiao X (2015) Recent progress in photocathodes for hydrogen evolution. J Mater Chem A 3:15824–15837. doi:10.1039/c5ta03594e 393. Luo J, Steier L, Son MK et al (2016) Cu2O nanowire photocathodes for efficient and durable solar water splitting. Nano Lett 16:1848–1857. doi:10.1021/acs.nanolett.5b04929 394. Dubale AA, Su WN, Tamirat AG et al (2014) The synergetic effect of graphene on Cu2O nanowire arrays as a highly efficient hydrogen evolution photocathode in water splitting. J Mater Chem A 2:18383–18397. doi:10.1039/c4ta03464c 395. Dubale AA, Tamirat AG, Chen H-M (2016) A highly stable CuS and CuS–Pt modified Cu2O/CuO heterostructure as an efficient photocathode for the hydrogen evolution reaction. J Mater Chem A 4:2205–2216. doi:10.1039/c5ta09464j References 115

396. Dong Y, Chen Y, Jiang P et al (2016) A novel g-C3N4 based photocathode for photoelectrochemical hydrogen evolution. RSC Adv 6:7465–7473. doi:10.1039/c5ra23265a 397. Basu M, Zhang ZW, Chen CJ et al (2016) CoSe2 embedded in C3N4:anefficient photocathode for photoelectrochemical water splitting. ACS Appl Mater Interfaces 8:26690– 26696. doi:10.1021/acsami.6b06520 398. Li S, Zhang P, Song X, Gao L (2015) Photoelectrochemical hydrogen production of TiO2 passivated Pt/Si-nanowire composite photocathode. ACS Appl Mater Interfaces 7:18560– 18565. doi:10.1021/acsami.5b04936 399. Bao XQ, Cerqueira MF, Alpuimab P, Liu L (2015) Silicon nanowire arrays coupled with cobalt phosphide spheres as low-cost photocathodes for efficient solar hydrogen evolution. Chem Commun 51:10742–10745. doi:10.1039/c5cc02331a 400. Lewis NS (2016) Developing a scalable artificial photosynthesis technology through nanomaterials by design. Nat Nanotechnol 11:1010–1019. doi:10.1038/nnano.2016.194 401. Chandrasekaran S, McInnes SJP, Macdonald TJ (2015) Porous silicon nanoparticles as a nanophotocathode for photoelectrochemical water splitting. RSC Adv 5:85978–85982. doi:10.1039/c5ra12559f 402. Reece SY, Hamel JA, Sung K, Jarvi TD, Esswein AJ, Pijpers JJH, Nocera DG (2011) Wireless solar water splitting using silicon-based semiconductors and earth-abundant catalysts. Science 34:645–648. doi:10.1126/science.1209816 403. Patra BK, Khilari S, Pradhan D, Pradhan N (2016) Hybrid dot-disk Au-CuInS2 nanostructures as active photocathode for efficient evolution of hydrogen from water. Chem Mater 28:4358–4366. doi:10.1021/acs.chemmater.6b01357 404. Chae SY, Park SJ, Han SG et al (2016) Enhanced photocurrents with ZnS passivated Cu(In, Ga)(Se, S)2 photocathodes synthesized using a nonvacuum process for solar water splitting. J Am Chem Soc 138:15673–15681. doi:10.1021/jacs.6b09595 405. Kaneko H, Minegishi T, Nakabayashi M et al (2016) Enhanced hydrogen evolution under simulated sunlight from neutral electrolytes on (ZnSe)0.85(CuIn0.7Ga0.3Se2)0.15 photocathodes prepared by a bilayer method. Angew Chem Int Ed 55:15329–15333. doi:10.1002/anie. 201609202 406. Abe T, Fukui K, Kawai Y et al (2016) A water splitting system using an organophotocathode and titanium dioxide photoanode capable of bias-free H2 and O2 evolution. Chem Commun 52:7735–7737. doi:10.1039/c6cc01225f 407. Hu Z, Yuan L, Liu Z et al (2016) An elemental phosphorus photocatalyst with a record high hydrogen evolution efficiency. Angew Chem Int Ed 55:9580–9585. doi:10.1002/anie. 201603331 408. Dang H, Dong X, Dong Y et al (2014) Enhancing the photocatalytic H2 evolution activity of red phosphorous by using noble-metal-free Ni(OH)2 under photoexcitation up to 700 nm. RSC Adv 4:44823–44826. doi:10.1039/c4ra06867j 409. Gao Y, Wang Y, Wang Y (2007) Photocatalytic hydrogen evolution from water on SiC under visible light irradiation. React Kin Catal Lett 91:13–19. doi:10.1007/s11144-007- 5064-x 410. Wang M, Chen J, Liao X et al (2014) Highly efficient photocatalytic hydrogen production of platinum nanoparticle-decorated SiC nanowires under simulated sunlight irradiation. Inter J Hydrogen En 39:14581–14587. doi:10.1016/j.ijhydene.2014.07.068 411. Wang D, Guo Z, Peng Y, Yuan W (2015) Visible light induced photocatalytic overall water splitting over micro-SiC driven by the Z-scheme system. Catal Commun 61:53–56. doi:10. 1016/j.catcom.2014.12.008 412. Wang Y, Guo X, Dong L et al (2013) Enhanced photocatalytic performance of chemically bonded SiC-graphene composites for visible-light-driven overall water splitting. Inter J Hydrogen En 38:12733–12738. doi:10.1016/j.ijhydene.2013.07.062 413. Liu H, She G, Mu L, Shi W (2012) Porous SiC nanowire arrays as stable photocatalyst for water splitting under UV irradiation. Mater Res Bull 47:917–920. doi:10.1016/j. materresbull.2011.12.046 116 2 Semiconductor-Based Photocatalytic Systems …

414. Wang B, Wang Y, Lei Y et al (2016) Mesoporous silicon carbide nanofibers with in situ embedded carbon for co-catalyst free photocatalytic hydrogen production. Nano Res 9:886– 898. doi:10.1007/s12274-015-0971-z 415. Kida T, Minami Y, Guan G et al (2006) Photocatalytic activity of gallium nitride for producing hydrogen from water under light irradiation. J Mater Sci 41:3527–3534. doi:10. 1007/s10853-005-5655-8 416. Yoshimizu M, Kobayashi R, Saegusa M et al (2015) Photocatalytic hydrogen evolution over b-iron silicide under infrared-light irradiation. Chem Commun 51:2818–2820. doi:10.1039/ C4CC08093A 417. Yoneyama H, Matsumoto N, Tamura H (1986) Photocatalytic decomposition of formic acid on platinized n-type silicon powder in aqueous solution. Bull Chem Soc Jpn 59:3302–3304. doi:10.1246/bcsj.59.3302 418. Jang YJ, Ryu J, Hong D et al (2016) A multi-stacked hyperporous silicon flake for highly active solar hydrogen production. Chem Commun 52:10221–10224. doi:10.1039/ c6cc04775k 419. Zhang H, Li A, Wang Z et al (2016) Decorating mesoporous silicon with amorphous metal– phosphorous-derived nanocatalysts towards enhanced photoelectrochemical water reduction. J Mater Chem A 4:14960–14967. doi:10.1039/c6ta05725j 420. Lv C, Chen Z, Chen Z et al (2015) Silicon nanowires loaded with iron phosphide for effective solar-driven hydrogen production. J Mater Chem A 3:17669–17675. doi:10.1039/ c5ta03438h 421. Li S, Wang H, Li D et al (2016) Siloxene nanosheets: a metal-free semiconductor for water splitting. J Mater Chem A 4:15841–15844. doi:10.1039/c6ta07545b 422. Mou Z, Yin S, Zhu M et al (2013) RuO2/TiSi2/graphene composite for enhanced photocatalytic hydrogen generation under visible light irradiation. Phys Chem Chem Phys 15:2793–2799. doi:10.1039/c2cp44270a 423. Wu W, Zhan L, Ohkubo K et al (2015) Photocatalytic H2 evolution from NADH with carbon quantum dots/Pt and 2-phenyl-4-(1-naphthyl)quinolinium ion. J Photochem Photobiol, B 152:63–70. doi:10.1016/j.jphotobiol.2014.10.018 424. Liu Q, Chen T, Guo Y et al (2016) Ultrathin g-C3N4 nanosheets coupled with carbon nanodots as 2D/0D composites for efficient photocatalytic H2 evolution. Appl Catal B 193:248–258. doi:10.1016/j.apcatb.2016.04.034 425. Yang P, Zhao J, Wang J et al (2015) Pure carbon nanodots for excellent photocatalytic hydrogen generation. RSC Adv 5:21332–21335. doi:10.1039/c5ra01924a 426. Ming H, Ma Z, Liu Y et al (2012) Large scale electrochemical synthesis of high quality carbon nanodots and their photocatalytic property. Dalton Trans 41:9526–9531. doi:10. 1039/C2DT30985H 427. Zhao S, Li C, Wang L et al (2016) Carbon quantum dots modified MoS2 with visible-light-induced high hydrogen evolution catalytic ability. Carbon 99:599–606. doi:10.1016/j.carbon.2015.12.088 428. Wang J, Ng YH, Lim YF, Ho GW (2014) Vegetable-extracted carbon dots and their nanocomposites for enhanced photocatalytic H2 production. RSC Adv 4:44117–44123. doi:10.1039/c4ra07290a 429. Thomas A, Fischer A, Goettmann F et al (2008) Graphitic carbon nitride materials: variation of structure and morphology and their use as metal-free catalysts. J Mater Chem 18:4893– 4908. doi:10.1039/b800274f 430. Zhou L, Zhang H, Sun H et al (2016) Recent advances in non-metal modification of graphitic carbon nitride for photocatalysis: a historic review. Catal Sci Technol 6:002–7023. doi:10.1039/c6cy01195k 431. Wen J, Xie J, Chen X, Li X (2017) A review on g-C3N4-based photocatalysts. Appl Surf Sci 391:72–123. doi:10.1016/j.apsusc.2016.07.030 432. Zheng Y, Lin L, Wang B, Wang X (2015) Graphitic carbon nitride polymers toward sustainable photoredox catalysis. Angew Chem Int Ed 54:12868–12884. doi:10.1002/anie. 201501788 References 117

433. Wang X, Blechert S, Antonietti M (2012) Polymeric graphitic carbon nitride for heterogeneous photocatalysis. ACS Catal 2:1596–1606. doi:10.1021/cs300240x 434. Zhu J, Xiao P, Li H, Carabineiro SA (2014) Graphitic carbon nitride: synthesis, properties, and applications in catalysis. ACS Appl Mater Interfaces 6:16449–16465. doi:10.1021/ am502925j 435. Liu J, Wang H, Antonietti M (2016) Graphitic carbon nitride “reloaded”: emerging applications beyond (photo)catalysis. Chem Soc Rev 45:2308–2326. doi:10.1039/ C5CS00767D 436. Patnaik S, Martha S, Parida KM (2016) An overview of the structural, textural and morphological modulations of g-C3N4 towards photocatalytic hydrogen production. RSC Adv 6:46929–46951. doi:10.1039/c5ra26702a 437. Zhang G, Wang X (2013) A facile synthesis of covalent carbon nitride photocatalysts by Co-polymerization of urea and phenylurea for hydrogen evolution. J Catal 307:246–253. doi:10.1016/j.jcat.2013.07.026 438. Chuang PK, Wu KH, Yeh TF, Teng H (2016) Extending the p-Conjugation of g-C3N4 by Incorporating Aromatic Carbon for Photocatalytic H2 Evolution from Aqueous Solution. ACS Sustainable Chem Eng 4:5989–5997. doi:10.1021/acssuschemeng.6b01266 439. Fan X, Zhang L, Cheng R et al (2015) Construction of graphitic C3N4-based intramolecular donor − acceptor conjugated copolymers for photocatalytic hydrogen evolution. ACS Catal 5:5008–5015. doi:10.1021/acscatal.5b01155 440. Cao S, Yu J (2014) g-C3N4-based photocatalysts for hydrogen generation. J Phys Chem Lett 5:2101–2107. doi:10.1021/jz500546b 441. Wang X, Maeda K, Chen X et al (2009) Polymer semiconductors for Artificial photosynthesis: hydrogen evolution by mesoporous graphitic carbon nitride with visible light. J Am Chem Soc 131:1680–1681. doi:10.1021/ja809307s 442. Zheng D, Zhang G, Hou Y, Wang X (2016) Layering MoS2 on soft hollow g-C3N4 nanostructures for photocatalytic hydrogen evolution. Appl Catal A 521:2–8. doi:10.1016/j. apcata.2015.10.037 443. Liang Q, Li Z, Yu X et al (2015) Macroscopic 3D porous graphitic carbon nitride monolith for enhanced photocatalytic hydrogen evolution. Adv Mater 27:4634–4639. doi:10.1002/ adma.201502057 444. He F, Chen G, Zhou Y et al (2015) The facile synthesis of mesoporous g-C3N4 with highly enhanced photocatalytic H2 evolution performance. Chem Commun 51:16244–16246. doi:10.1039/c5cc06713h 445. Shen S, Zhao D, Chen J et al (2016) Enhanced photocatalytic hydrogen evolution over graphitic carbon nitride modified with Ti-activated mesoporous silica. Appl Catal A 521:111–117. doi:10.1016/j.apcata.2015.11.004 446. Qiao S, Mitchell RW, Coulson B et al (2016) Pore confinement effects and stabilization of carbon nitride oligomers in macroporous silica for photocatalytic hydrogen production. Carbon 106:320–329. doi:10.1016/j.carbon.2016.05.039 447. Chen X, Jun YS, Takanabe K et al (2009) Ordered mesoporous SBA-15 type graphitic carbon nitride: a semiconductor host structure for photocatalytic hydrogen evolution with visible light. Chem Mater 21:4093–4095. doi:10.1021/cm902130z 448. Xu L, Jin B, Zhang J et al (2016) Efficient hydrogen generation from formic acid using AgPd nanoparticles immobilized on carbon nitride-functionalized SBA-15. RSC Adv 6:46908– 46914. doi:10.1039/c6ra06071d 449. Cheng R, Fan X, Wang M et al (2016) Facile construction of CuFe2O4/g-C3N4 photocatalyst for enhanced visible-light hydrogen evolution. RSC Adv 6:18990–18995. doi:10.1039/ c5ra27221a 450. Tong J, Zhang L, Li F et al (2015) Rapid and high-yield production of g-C3N4 nanosheets via chemical exfoliation for photocatalytic H2 evolution. RSC Adv 5:88149–88153. doi:10. 1039/c5ra16988g 118 2 Semiconductor-Based Photocatalytic Systems …

451. Han Q, Wang B, Gao J et al (2016) Atomically thin mesoporous nanomesh of graphitic C3N4 for high-effciency photocatalytic hydrogen evolution. ACS Nano 10:2745–2751. doi:10.1021/acsnano.5b07831 452. Fan M, Song C, Chen T et al (2016) Visible-light-drived high photocatalytic activities of Cu/ g-C3N4 photocatalysts for hydrogen production. RSC Adv 6:34633–34640. doi:10.1039/ c5ra27755h 453. Bi L, Meng D, Bu Q et al (2016) Electron acceptor of Ni decorated porous carbon nitride applied in photocatalytic hydrogen production. Phys Chem Chem Phys 18:31534–31541. doi:10.1039/c6cp05618k 454. Xu L, Liu N, Hong B et al (2016) Nickel–platinum nanoparticles immobilized on graphitic carbon nitride as highly efficient catalyst for hydrogen release from hydrous hydrazine. RSC Adv 6:31687–31691. doi:10.1039/c6ra01335j 455. Ge L, Han C, Xiao X, Guo L (2013) Synthesis and characterization of composite visible light active photocatalysts MoS2–g-C3N4 with enhanced hydrogen evolution activity. Inter J Hydrogen En 38:6960–6969. doi:10.1016/j.ijhydene.2013.04.006 456. Jin X, Fan X, Tian J et al (2016) MoS2 quantum dot decorated g-C3N4 composite photocatalyst with enhanced hydrogen evolution performance. RSC Adv 6:52611–52619. doi:10.1039/c6ra07060d 457. Sun Q, Wang P, Yu H, Wang X (2016) In situ hydrothermal synthesis and enhanced photocatalytic H2-evolution performance of suspended rGO/g-C3N4 photocatalysts. J Mol Catal A 424:369–376. doi:10.1016/j.molcata.2016.09.015 458. Suryawanshi A, Dhanasekaran P, Mhamane D et al (2012) Ioubling of photocatalytic H2 evolution from g-C3N4 via its nanocomposite formation with multiwall carbon nanotubes: electronic and morphological effects. Inter J Hydrogen En 37:9584–9589. doi:10.1016/j. ijhydene.2012.03.123 459. Ge L, Han C (2012) Synthesis of MWNTs/g-C3N4 composite photocatalysts with efficient visible light photocatalytic hydrogen evolution activity. Appl Catal B 117–118:268–274. doi:10.1016/j.apcatb.2012.01.021 460. Wang N, Li J, Wu L et al (2016) MnO2 and carbon nanotube co-modified C3N4 composite catalyst for enhanced water splitting activity under visible light irradiation. Inter J Hydrogen En 41:22743–22750. doi:10.1016/j.ijhydene.2016.10.068 461. Gao LF, Wen T, Xu JY et al (2016) Iron-doped carbon nitride-type polymers as homogeneous organocatalysts for visible light-driven hydrogen evolution. ACS Appl Mater Interfaces 8:617–624. doi:10.1021/acsami.5b09684 462. Min S, Lu G (2012) Enhanced electron transfer from the excited Eosin Y to mpg-C3N4 for highly efficient hydrogen evolution under 550 nm irradiation. J Phys Chem C 116:19644– 19652. doi:10.1021/jp304022f 463. Zhang H, Li S, Lu R, Yu A (2015) Time-resolved study on xanthene dye-sensitized carbon nitride photocatalytic systems. ACS Appl Mater Interfaces 7:21868–21874. doi:10.1021/ acsami.5b06309 464. Mori K, Itoh T, Kakudo H et al (2015) Nickel-supported carbon nitride photo-catalyst combined with organic dye for visible-light-driven hydrogen evolution from water. Phys Chem Chem Phys 17:24086–24091. doi:10.1039/c5cp04493f 465. Cao SW, Yuan YP, Fang J et al (2013) In-situ growth of CdS quantum dots on g-C3N4 nanosheets for highly efficient photocatalytic hydrogen generation under visible light irradiation. Inter J Hydrogen En 38:1258–1266. doi:10.1016/j.ijhydene.2012.10.116 466. Zhang Z, Zhang Y, Lu L et al (2017) Graphitic carbon nitride nanosheet for photo-catalytic hydrogen oduction: the impact of morphology and element composition. Appl Surf Sci 391:369–375. doi:10.1016/j.apsusc.2016.05.174 467. Yang S, Gong Y, Zhang J et al (2013) Exfoliated graphitic carbon nitride nanosheets as efficient catalysts for hydrogen evolution under visible light. Adv Mater 25:2452–2456. doi:10.1002/adma.201204453 References 119

468. Xu J, Zhang L, Shi R, Zhu Y (2013) Chemical exfoliation of graphitic carbon nitride for efficient heterogeneous photocatalysis. J Mater Chem A 1:14766–14772. doi:10.1039/ C3TA13188B 469. Bu X, Bu Y, Yang S et al (2016) Graphitic carbon nitride nanoribbon for enhanced visible-light photocatalytic H2 production. RSC Adv 6:112210–112214. doi:10.1039/ c6ra23218c 470. Schwinghammer K, Mesch MB, Duppel V (2014) Crystalline carbon nitride nanosheets for improved visible-light hydrogen evolution. J Am Chem Soc 136:1730–1733. doi:10.1021/ ja411321s 471. Li L, Fang W, Zhang P et al (2016) Sulfur-doped covalent triazine-based frameworks for enhanced photocatalytic hydrogen evolution from water under visible light. J Mater Chem A 4:12402–12406. doi:10.1039/c6ta04711d 472. Lin L, Ou H, Zhang Y, Wang X (2016) Tri-s-triazine-based crystalline graphitic carbon nitrides for highly efficient hydrogen evolution photocatalysis. ACS Catal 6:3921–3931. doi:10.1021/acscatal.6b00922 473. Bhunia MK, Melissen S, Parida MR (2015) Dendritic tip-on polytriazine-based carbon nitride photocatalyst with high hydrogen evolution activity. Chem Mater 27:8237–8247. doi:10.1021/acs.chemmater.5b02974 474. Liu H, Chen D, Wang Z et al (2017) Microwave-assisted molten-salt rapid synthesis of isotype triazine-/heptazine based g-C3N4 heterojunctions with highly enhanced photocatalytic hydrogen evolution performance. Appl Catal B 203:300–313. doi:10.1016/j.apcatb. 2016.10.014 475. Li Y, Xu H, Ouyang S, Ye J (2016) Metal–organic frameworks for photocatalysis. Phys Chem Chem Phys 18:7563–7572. doi:10.1039/c5cp05885f 476. So MC, Wiederrecht GP, Mondloch JE et al (2015) Metal–organic framework materials for light-harvesting and energy transfer. Chem Commun 51:3501–3510. doi:10.1039/ c4cc09596k 477. Xiao JD, Shang Q, Xiong Y et al (2016) Boosting photocatalytic hydrogen production of a metal-organic framework decorated with platinum nanoparticles: the platinum location matters. Angew Chem Int Ed 55:9389–9393. doi:10.1002/anie.201603990 478. Zhou JJ, Wang R, Liu XL et al (2015) In situ growth of CdS nanoparticles on UiO-66 metal-organic framework octahedrons for enhanced photocatalytic hydrogen production under visible light irradiation. Appl Surf Sci 346:278–283. doi:10.1016/j.apsusc.2015.03. 210 479. Lin R, Shen L, Ren Z et al (2014) Enhanced photocatalytic hydrogen production activity via dual modification of MOF and reduced graphene oxide on CdS. Chem Commun 50:8533– 8535. doi:10.1039/c4cc01776e 480. He J, Wang J, Chen Y et al (2014) A dye-sensitized Pt@UiO-66(Zr) metal–organic framework for visible-light photocatalytic hydrogen production. Chem Commun 50:7063– 7066. doi:10.1039/c4cc01086h 481. Su Y, Zhang Z, Liu H, Wang Y (2017) Cd0.2Zn0.8S@UiO-66-NH2 nanocomposites as efficient and stable visible-light-driven photocatalyst for H2 evolution and CO2 reduction. Appl Catal B 200:448–457. doi:10.1016/j.apcatb.2016.07.032 482. Sasan K, Lin Q, Mao CY, Feng P (2014) Incorporation of iron hydrogenase active sites into a highly stable metal–organic framework for photocatalytic hydrogen generation. Chem Commun 50:10390–10393. doi:10.1039/c4cc03946g 483. Ragon F, Campo B, Yang Q et al (2015) Acid-functionalized UiO-66(Zr) MOFs and their evolution after intra-framework cross-linking: structural features and sorption properties. J Mater Chem A 3:3294–3309. doi:10.1039/C4TA03992K 484. Wang D, Song Y, Cai J et al (2016) Effective photo-reduction to deposit Pt nanoparticles on MIL-100(Fe) for visible-light-induced hydrogen evolution. New J Chem 40:9170–9175. doi:10.1039/c6nj01989g 120 2 Semiconductor-Based Photocatalytic Systems …

485. Vaddipalli SR, Sanivarapu SR, Vengatesan S (2016) Heterostructured Au NPs/CdS/LaBTC MOFs photoanode for efficient photoelectrochemical water split-ting: stability enhancement via CdSe QDs to 2D-CdS nanosheets transformation. ACS Appl Mater Interfaces 8:23049– 23059. doi:10.1021/acsami.6b06851 486. Sun X, Yu Q, Zhang F et al (2016) A dye-like ligand-based metal—organic framework for efficient photocatalytic hydrogen production from aqueous solution. Catal Sci Technol 6:3840–3844. doi:10.1039/c5cy01716e 487. Dong XY, Zhang M, Pei RB et al (2016) A crystalline copper(II) coordination polymer for the efficient visible-light-driven generation of hydrogen. Angew Chem Int Ed 55:2073– 2077. doi:10.1002/anie.201509744 488. Xu JY, Zhai XP, Gao LF et al (2016) In situ preparation of a MOF-derived magnetic carbonaceous catalyst for visible-light-driven hydrogen evolution. RSC Adv 6:2011–2018. doi:10.1039/c5ra23838b 489. Nellist MR, Laskowski FAL, Lin F, Mills TJ, Boettcher SW (2016) Semiconductor- electrocatalyst interfaces: theory, experiment, and applications in photoelectrochemical water splitting. Acc Chem Res 49:733–740. doi:10.1021/acs.accounts.6b00001 490. Orlandi M, Dalle Carbonare N, Caramori S, Bignozzi CA, Berardi S, Mazzi A, El Koura Z, Bazzanella N, Patel N, Miotello A (2016) Porous versus compact nanosized Fe(III)-based water oxidation catalyst for photoanodes functionalization. ACS Appl Mater Interfaces 8:20003–20011. doi:10.1021/acsami.6b05135 491. Ronconi F, Syrgiannis Z, Bonasera A, Prato M, Argazzi R, Caramori S, Cristino V, Bignozzi CA (2015) Modification of nanocrystalline WO3 with a dicationic perylene bisimide: applications to molecular level solar water splitting. J Am Chem Soc 137:4630– 4633. doi:10.1021/jacs.5b01519 492. Wang J, Li B, Chen J et al (2012) Enhanced photocatalytic H2-production activity of CdxZn1−xS nanocrystals by surface loading MS (M = Ni Co, Cu) species. Appl Surf Sci 259:118–123. doi:10.1016/j.apsusc.2012.07.003 493. Zong X, Yan H, Wu G et al (2008) Enhancement of photocatalytic H2 evolution on CdS by loading MoS2 as cocatalyst under visible light irradiation. J Am Chem Soc 130:7176–7177. doi:10.1021/ja8007825 494. Ji J, Guo L, Li Q et al (2015) A bifunctional catalyst for hydrogen evolution reaction: the interactive influences between CdS and MoS2 on photoelectro-chemical activity. Inter J Hydrogen En 40:3813–3821. doi:10.1016/j.ijhydene.2015.01.075 495. Jia T, Kolpin A, Ma C et al (2014) A graphene dispersed CdS–MoS2 nanocrystal ensemble for cooperative photocatalytic hydrogen production from water. Chem Commun 50:1185– 1188. doi:10.1039/C3CC47301E 496. Lu Y, Wang D, Yang P et al (2014) Coupling ZnxCd1−xS nanoparticles with graphene-like MoS2: superior interfacial contact, low overpotential and enhanced photocatalytic activity under visible-light irradiation. Catal Sci Technol 4:2650–2657. doi:10.1039/C4CY00331D 497. Zhu B, Lin B, Zhou Y et al (2014) Enhanced photocatalytic H2 evolution on ZnS loaded with graphene and MoS2 nanosheets as cocatalysts. J Mater Chem A 2:3819–3827. doi:10. 1039/C3TA14819J 498. Zhao H, Dong Y, Jiang P et al (2015) In situ light-assisted preparation of MoS2 on graphitic C3N4 nanosheets for enhanced photocatalytic H2 production from water. J Mater Chem A 3:7375–7381. doi:10.1039/C5TA00402K 499. Laursen AB, Kegnaes S, Dahl S, Chorkendorff I (2012) Molybdenum sulfides—efficient and viable materials for electro- and photoelectrocatalytic hydrogen evolution. En Environ Sci 5:5577–5591. doi:10.1039/C2EE02618J 500. Liu M, Li F, Sun Z et al (2014) Noble-metal-free photocatalysts MoS2–graphene/CdS mixed nanoparticles/nanorods morphology with high visible light efficiency for H2 evolution. Chem Commun 50:11004–11007. doi:10.1039/C4CC04653F 501. Redman DW, Kim HJ, Stevenson KJ, Rose MJ (2016) Photo-assisted electrodeposition of MoSx from ionic liquids on organic-functionalized silicon photoelectrodes for H2 generation. J Mater Chem A 4:7027–7035. doi:10.1039/C5TA09684G References 121

502. Kumar DP, Hong S, Reddy DA, Kim TK (2016) Noble metal-free ultrathin MoS2 nanosheet-decorated CdS nanorods as an efficient photocatalyst for spectacular hydrogen evolution under solar light irradiation. J Mater Chem A 4:18551–18558. doi:10.1039/ c6ta08628d 503. Du P, Zhu Y, Zhng J et al (2016) Metallic 1T phase MoS2 nanosheets as a highly efficient co-catalyst for the photocatalytic hydrogen evolution of CdS nanorods. RSC Adv 6:74394– 74399. doi:10.1039/c6ra10170d 504. Sasikala R, Gaikwad AP, Jayakumar OD et al (2015) Nanohybrid MoS2-PANI-CdS photocatalyst for hydrogen evolution from water. Colloids Surf A 481:485–492. doi:10. 1016/j.colsurfa.2015.06.027 505. Du H, Xie X, Zhu Q et al (2015) Metallic MoO2 cocatalyst significantly enhances visible-light photocatalytic hydrogen production over MoO2/Zn0.5Cd0.5S hetero-junction. Nanoscale 7:5752–5759. doi:10.1039/C4NR06949H 506. Ma B, Liu Y, Li J et al (2016) Mo2N: An efficient non-noble metal cocatalyst on CdS for enhanced photocatalytic H2 evolution under visible light irradiation. Inter J Hydrogen En 41:22009–22016. doi:10.1016/j.ijhydene.2016.08.133 507. Khan Z, Khannam M, Vinothkumar N et al (2012) Hierarchical 3D NiO–CdS heteroar- chitecture for efficient visible light photocatalytic hydrogen generation. J Mater Chem 22:12090–12095. doi:10.1039/C2JM31148H 508. Lin CY, Lai YH, Mersch D, Reisner E (2012) Cu2O|NiOx nanocomposite as an inexpensive photocathode in photoelectrochemical water splitting. Chem Sci 3:3482–3487. doi:10.1039/ C2SC20874A 509. Bi G, Wen J, Li X et al (2016) Efficient visible-light photocatalytic H2 evolution over metal-free g-C3N4 co-modified with robust acetylene black and Ni(OH)2 as dual co-catalysts. RSC Adv 6:31497–31506. doi:10.1039/c6ra03118h 510. Yuan YP, Cao SW, Yin LS et al (2013) NiS2 Co-catalyst decoration on CdLa2S4 nanocrystals for efficient photocatalytic hydrogen generation under visible light irradiation. Inter J Hydrogen En 38:7218–7223. doi:10.1016/j.ijhydene.2013.03.169 511. Parida KM, Biswal N, Das DP, Martha S (2010) Visible light response photocatalytic water splitting over CdS-pillared zirconium–titanium phosphate (ZTP). Inter J Hydrogen En 35:5262–5269. doi:10.1016/j.ijhydene.2010.03.017 512. Sun Z, Chen H, Zhang L et al (2016) Enhanced photocatalytic H2 production on cadmium sulfide photocatalysts using nickel nitride as a novel cocatalyst. J Mater Chem A 4:13289– 13295. doi:10.1039/c6ta04696g 513. Cao S, Chen Y, Wang CJ et al (2014) Highly efficient photocatalytic hydrogen evolution by nickel phosphide nanoparticles from aqueous solution. Chem Commun 50:10427–10429. doi:10.1039/C4CC05026F 514. Yang JS, Ham DJ, Lakshminarasimhan N et al (2008) Role of platinum-like tungsten carbide as cocatalyst of CdS photocatalyst for hydrogen production under visible light irradiation. Appl Catal A 346:149–154. doi:10.1016/j.apcata.2008.05.020 515. Yan Z, Wu H, Han A et al (2014) Noble metal-free cobalt oxide (CoOx) nano-particles loaded on titanium dioxide/cadmium sulfide composite for enhanced photocatalytic hydrogen production from water. Inter J Hydrogen En 39:13353–13360. doi:10.1016/j. ijhydene.2014.04.121 516. Yuan J, Wen J, Gao Q et al (2015) Amorphous Co3O4 modified CdS nanorods with enhanced visible-light photocatalytic H2-production activity. Dalton Trans 44:1680–1689. doi:10.1039/C4DT03197K 517. Cao S, Chen Y, Hou CC et al (2015) Cobalt phosphide as a highly active non-precious metal cocatalyst for photocatalytic hydrogen production under visible light irradiation. J Mater Chem A 3:6096–6101. doi:10.1039/C4TA07149B 518. Yuan YJ, Ye ZJ, Lu HW et al (2016) Constructing anatase TiO2 nanosheets with exposed (001) facets/layered MoS2 two-dimensional nanojunctions for enhanced solar hydrogen generation. ACS Catal 6:532–541. doi:10.1021/acscatal.5b02036 122 2 Semiconductor-Based Photocatalytic Systems …

519. Cao S, Chen Y, Kang L et al (2015) Enhanced photocatalytic H2 evolution by immobilizing CdS nanocrystals on ultrathin Co0.85Se/RGO—PEI nanosheets. J Mater Chem A 3:18711– 18717. doi:10.1039/c5ta04910e 520. Zhang L, Jiang T, Li S et al (2013) Enhancement of photocatalytic H2 evolution on Zn0.8Cd0.2S loaded with CuS as cocatalyst and its photogenerated charge transfer properties. Dalton Trans 42:12998–13003. doi:10.1039/C3DT51256H 521. Zhang L, Liu YN, Zhou M, Yan J (2013) Improving photocatalytic hydrogen evolution over CuO/Al2O3 by platinum-depositing and CuS-loading. Appl Surf Sci 282:531–537. doi:10. 1016/j.apsusc.2013.06.006 522. Wang X, Liu M, Chen Q et al (2013) Synthesis of CdS/CNTs photocatalysts and study of hydrogen production by photocatalytic water splitting. Inter J Hydrogen En 38:13091– 13096. doi:10.1016/j.ijhydene.2013.03.016 523. Zhang J, Qiao SZ, Qi L, Yu J (2013) Fabrication of NiS modified CdS nanorod p–n junction photocatalysts with enhanced visible-light photocatalytic H2-production activity. Phys Chem Chem Phys 15:12088–12094. doi:10.1039/C3CP50734C 524. Li N, Zhou B, Guo P et al (2013) Fabrication of noble-metal-free Cd0.5Zn0.5S/NiS hybrid photocatalyst for efficient solar hydrogen evolution. Inter J Hydrogen En 38:11268–11277. doi:10.1016/j.ijhydene.2013.06.067 525. Li Y, Lin S, Peng S et al (2013) Modification of ZnS1−x−0.5yOx(OH)y–ZnO photocatalyst with NiS for enhanced visible-light-driven hydrogen generation from seawater. Inter J. Hydrogen En 38:15976–15984. doi:10.1016/j.ijhydene.2013.09.149 526. Zhu Y, Xu Y, Hou Y et al (2014) Cobalt sulfide modified graphitic carbon nitride semiconductor for solar hydrogen production. Inter J Hydrogen En 39:11873–11879. doi:10. 1016/j.ijhydene.2014.06.025 527. Chen Y, Qin Z (2016) General applicability of nanocrystalline Ni2P as a noble-metal-free cocatalyst to boost photocatalytic hydrogen generation. Catal Sci Technol 6:8212–8221. doi:10.1039/c6cy01653g 528. Yasomanee JP, Bandara J (2008) Multi-electron storage of photoenergy using Cu2O–TiO2 thin film photocatalyst. Solar En Mater Solar Cells 92:348–352. doi:10.1016/j.solmat.2007. 09.016 529. Jian JX, Ye C, Wang XZ et al (2016) Comparison of H2 photogeneration by [FeFe]- hydrogenase mimics with CdSe QDs and Ru(bpy)3Cl2 in aqueous solution. Energy Environ Sci 9:2083–2089. doi:10.1039/c6ee00629a 530. Wen F, Li C (2013) Hybrid artificial photosynthetic systems comprising semicon-ductors as light harvesters and biomimetic complexes as molecular cocatalysts. Acc Chem Res 4:2355– 2364. doi:10.1021/ar300224u 531. Yeh TF, Cihlar J, Chang CY et al (2013) Roles of graphene oxide in photocatalytic water splitting. Mater Today 16:78–84. doi:10.1016/j.mattod.2013.03.006 532. Yang Y, Liu E, Dai H et al (2014) Photocatalytic activity of Ag-TiO2-graphene ternary nanocomposites and application in. hydrogen evolution by water splitting. Inter J Hydrogen En 39:7664–7671. doi:10.1016/j.ijhydene.2013.09.109 533. Fang Z, Wang Y, Song J et al (2013) Immobilizing CdS quantum dots and dendritic Pt nanocrystals on thiolated graphene nanosheets toward highly efficient photocatalytic H2 evolution. Nanoscale 5:9830–9838. doi:10.1039/c3nr03043a 534. Xiang Q, Yu J, Jaroniec M (2011) Enhanced photocatalytic H2-production activity of graphene-modified titania nanosheets. Nanoscale 3:3670–3678. doi:10.1039/C1NR10610D 535. Hong Y, Shi P, Wang P, Yao W (2015) Improved photocatalytic activity of CdS/reduced graphene oxide (RGO) for H2 evolution by strengthening the connection between CdS and RGO sheets. Inter J Hydrogen En 40:7045–7051. doi:10.1016/j.ijhydene.2015.04.005 536. Yu X, Du R, Li B et al (2016) Biomolecule-assisted self-assembly of CdS/MoS2/graphene hollow spheres as high-efficiency photocatalysts for hydrogen evolution without noble metals. Appl Catal B 182:504–512. doi:10.1016/j.apcatb.2015.09.003 References 123

537. Xu H, Li X, Kang S et al (2014) Noble metal-free cuprous oxide/reduced graphene oxide for enhanced photocatalytic hydrogen evolution from water reduction. Inter J Hydrogen En 39:11578–11582. doi:10.1016/j.ijhydene.2014.05.156 538. Ding J, Yan W, Sun S et al (2014) Fabrication of graphene/CaIn2O4 composites with enhanced photocatalytic activity from water under visible light irradiation. Inter J Hydrogen En 39:119–126. doi:10.1016/j.ijhydene.2013.10.077 539. Pan B, Wang Y, Liang Y et al (2014) Nanocomposite of BiPO4 and reduced graphene oxide as an efficient photocatalyst for hydrogen evolution. Inter J Hydrogen En 39:13527–13533. doi:10.1016/j.ijhydene.2014.02.031 540. Sun Z, Guo J, Zhu S et al (2014) A high-performance Bi2WO6—graphene photocatalyst for visible light-induced H2 and O2 generation. Nanoscale 6:2186–2193. doi:10.1039/ c3nr05249d 541. Hong Z, Li X, Kang S et al (2014) Enhanced photocatalytic activity and stability of the reduced graphene oxide loaded potassium niobate microspheres for hydrogen production from water reduction. Inter J Hydrogen En 39:12515–12523. doi:10.1016/j.ijhydene.2014. 06.075 542. Alexander BD, Kulesza PJ, Rutkowska I et al (2008) Metal oxide photoanodes for solar hydrogen production. J Mater Chem 18:2298–2303. doi:10.1039/B718644D 543. Kudo A, Kato H, Tsuji I (2004) Strategies for the development of visible-light-driven photocatalysts for water splitting. Chem Lett 33:1534–1539. doi:10.1246/cl.2004.1534 544. Sayama K, Mukasa K, Abe R et al (2002) A new photocatalytic water splitting system under visible light irradiation mimicking a Z-scheme mechanism in photosynthesis. J Photochem Photobiol A 148:71–77. doi:10.1016/S1010-6030(02)00070-9 545. Sayama K, Mukasa K, Abe R et al (2001) Stoichiometric water splitting into H2 and O2 À À using a mixture of two different photocatalysts and an IO3 I shuttle redox mediator under visible light irradiation. Chem. Commun 2416–2417. doi:10.1039/B107673F 546. Schürch D, Currao A, Sarkar S et al (2002) The silver chloride photoanode in photoelectro– chemical water splitting. J Phys Chem B 106:12764–12775. doi:10.1021/jp0265081 547. Abe R, Sayama K, Sugihara H (2005) Development of new photocatalytic water splitting into H2 and O2 using two different semiconductor photocatalysts and a shuttle redox À À – mediator IO3 I . J Phys Chem B 109:16052 16061. doi:10.1021/jp052848l 548. Abe R, Takata T, Sugihara H, Domen K (2005) Photocatalytic overall water splitting under À À visible light by TaON and WO3 with an IO3 I shuttle redox mediator. Chem Commun 3829–3831. doi:10.1039/B505646B 549. Higashi M, Abe R, Takata T, Domen K (2009) Photocatalytic Overall Water Splitting under À À Visible Light Using ATaO2N (A = Ca, Sr, Ba) and WO3 in a IO3 I Shuttle Redox Mediated System. Chem Mater 21:1543–1549. doi:10.1021/cm803145n 550. Higashi M, Abe R, Ishikawa A et al (2008) Z-scheme overall water splitting on modified-TaON photocatalysts under visible light (k < 500 nm). Chem Lett 37:138–139 551. Abe R, Shinmei K, Koumura N et al (2013) Visible-light-induced water splitting based on two-step photoexcitation between dye-sensitized layered niobate and tungsten oxide photocatalysts in the presence of a triiodide/iodide shuttle redox mediator. J Am Chem Soc 135:16872–16884. doi:10.1021/ja4048637 552. Jia Q, Iwase A, Kudo A (2014) BiVO4–Ru/SrTiO3: Rh composite Z-scheme photo-catalyst for solar water splitting. Chem Sci 5:1513–1519. doi:10.1039/C3SC52810C 553. Maeda K, Lu D, Domen K (2013) Solar-driven Z-scheme water splitting using modified BaZrO3–BaTaO2N solid solutions as photocatalysts. ACS Catal 3:1026–1033. doi:10.1021/ cs400156m 554. Kato H, Hori M, Konta R et al (2004) Construction of Z-scheme type heterogeneous photocatalysis systems for water splitting into H2 and O2 under visible light irradiation. Chem Lett 33:1348–1349. doi:10.1246/cl.2004.1348 124 2 Semiconductor-Based Photocatalytic Systems …

555. Kato H, Sasaki Y, Iwase A, Kudo A (2007) Role of iron ion electron mediator on photocatalytic overall water splitting under visible light irradiation using Z-scheme systems. Bull Chem Soc Jpn 80:2457–2464. doi:10.1246/bcsj.80.2457 556. Bae SW, Ji SM, Hong SJ et al (2009) Photocatalytic overall water splitting with dual-bed system under visible light irradiation. Inter J Hydrogen En 34:3243–3249. doi:10.1016/j. ijhydene.2009.02.022 557. Pan Z, Hisatomi T, Wang Q et al (2016) Photocatalyst sheets composed of particulate LaMg1/3Ta2/3O2N and Mo-Doped BiVO4 for Z-scheme water splitting under visible light. ACS Catal 6:7188–7196. doi:10.1021/acscatal.6b01561 558. Iwase A, Ng YH, Ishiguro Y et al (2011) Reduced graphene oxide as a solid-state electron mediator in Z-scheme photocatalytic water splitting under visible light. J Am Chem Soc 133:11054–11057. doi:10.1021/ja203296z 559. Paulose M, Mor GK, Varghese OK et al (2006) Visible light photoelectrochemical and water-photoelectrolysis properties of titania nanotube arrays. J Photochem Photobiol A 178:8–15. doi:10.1016/j.jphotochem.2005.06.013 560. Khan MA, Akhtar MS, Woo SI, Yang OB (2008) Enhanced photoresponse under visible light in Pt ionized TiO2 nanotube for the photocatalytic splitting of water. Catal Commun 10:1–5. doi:10.1016/j.catcom.2008.01.018 561. Zou Z, Ye J, Arakawa H (2002) Surface characterization of nanoparticles of NiOx/ In0.9Ni0.1TaO4: effects on photocatalytic activity. J Phys Chem B 106:13098–13101. doi:10. 1021/jp0216225 562. Jang JS, Kim HG, Reddy VR et al (2005) Photocatalytic water splitting over iron oxide nanoparticles intercalated in HTiNb(Ta)O5 layered compounds. J Catal 231:213–222. doi:10.1016/j.jcat.2005.01.026 563. Wu G, Chen A (2008) Direct growth of F-doped TiO2 particulate thin films with high photocatalytic activity for environmental applications. J Photochem Photobiol A 195:47–53. doi:10.1016/j.jphotochem.2007.09.005 564. Kitano M, Iyatani K, Tsujimaru K et al (2008) The effect of chemical etching by HF solution on the photocatalytic activity of visible light-responsive TiO2 thin films for solar water splitting. Topics Catal 49:24–31. doi:10.1007/s11244-008-9064-5 565. Maeda K, Takata T, Hara M et al (2005) GaN:ZnO solid solution as a photocatalyst for visible-light-driven overall water splitting. J Am Chem Soc 127:8286–8287. doi:10.1021/ ja0518777 566. Liu H, Yuan J, Shangguang W, Teraoka Y (2008) Visible-light-responding BiYWO6 solid solution for stoichiometric photocatalytic water splitting. J Phys Chem C 112:8521–8523. doi:10.1021/jp802537u 567. Sayama K, Nomura A, Arai T, Sugita T et al (2006) Photoelectrochemical decomposition of water into H2 and O2 on porous BiVO4 thin-film electrodes under visible light and significant effect of Ag ion treatment. J Phys Chem B 110:11352–11360. doi:10.1021/ jp057539+ 568. Bornoz P, Abdi FF, Tilley SD et al (2014) A bismuth vanadate-cuprous oxide tandem cell for overall solar water splitting. J Phys Chem C 118:16959–16966. doi:10.1021/jp500441h 569. Saremi-Yarahmadi S, Upul Wijayantha KG, Tahir AA, Vaidhyanathan B (2009) Nanostructured a-Fe2O3 electrodes for solar driven water splitting: effect of doping agents on preparation and performance. J Phys Chem C 113:4768–4778. doi:10.1021/jp808453z 570. Saremi-Yarahmadi S, Tahir AA, Vaidhyanathan B, Upul Wijayantha KG (2009) Fabrication of nanostructured a-Fe2O3 electrodes using ferrocene for solar hydrogen generation. Mater Lett 63:523–526. doi:10.1016/j.matlet.2008.11.011 571. Youngblood WJ, Lee SHA, Kobayashi Y et al (2009) Photoassisted overall water splitting in a visible light-absorbing dye-sensitized photoelectrochemical cell. J Am Chem Soc 131:926–927. doi:10.1021/ja809108y 572. Zou Z, Ye J, Arakawa H (2003) Photocatalytic water splitting into H2 and/or O2 under UV and visible light irradiation with a semiconductor photocatalyst. Inter J Hydrogen En 28:663–669. doi:10.1016/S0360-3199(02)00159-3 References 125

573. Oshikiri M, Boero M, Ye J et al (2002) Electronic structures of promising photocatalysts InMO4 (M = V, Nb, Ta) and BiVO4 for water decomposition in the visible wavelength region. J Chem Phys 117:7313–7318 574. Zou Z, Ye J, Arakawa H (2001) Photophysical and photocatalytic properties of InMO4 (M = Nb5+,Ta5+) under visible light irradiation. Mater Res Bull 36:1185–1193. doi:10. 1016/S0025-5408(01)00607-9 575. Xu J, Pan C, Takata T, Domen K (2015) Photocatalytic overall water splitting on the perovskite-type transition metal oxynitride CaTaO2N under visible light irradiation. Chem Commun 51:7191–7194. doi:10.1039/C5CC01728A 576. Li Y, Jiang S, Xiao J, Li Y (2014) Photocatalytic overall water splitting under visible light over an In–Ni–Ta–O–N solid solution without an additional cocatalyst. Inter J Hydrogen En 39:731–735. doi:10.1016/j.ijhydene.2013.10.092 577. Liu H, Yuan J, Jiang Z et al (2011) Novel photocatalyst of V-based solid solutions for overall water splitting. J Mater Chem 21:16535–16543. doi:10.1039/C1JM11809A 578. Guo DZ, Zhang GM, Zhang ZX et al (2006) Visible-light-induced water-splitting in channels of carbon nanotubes. J Phys Chem B 110:1571–1575. doi:10.1021/jp055929q 579. Maeda K, Wang X, Nishihara Y et al (2009) photocatalytic activities of graphitic carbon nitride powder for water reduction and oxidation under visible light. J Phys Chem C 113:4940–4947. doi:10.1021/jp809119m 580. Zhao G, Huang X, Fina F et al (2015) Facile structure design based on C3N4 for mediator-free Z-scheme water splitting under visible light. Catal Sci Technol 5:3416–3422. doi:10.1039/C5CY00379B 581. Wu X, Zhao J, Guo S et al (2016) Carbon dot and BiVO4 quantum dot composites for overall water splitting via a two-electron pathway. Nanoscale 8:17314–17321. doi:10.1039/ c6nr05864g 582. Haussener S, Hu S, Xiang C, Weber AZ, Lewis NS (2013) Simulations of the irradiation and temperature dependence of the efﬁciency of tandem photoelectrochemical water-splitting systems. Energy Environ Sci 6:3605–3618. doi:10.1039/C3EE41302K Chapter 3 Semiconductor-Based Photocatalytic Systems for the Reductive Conversion of CO2 and N2

Semiconductor-based photocatalytic systems aimed at the reduction of carbon dioxide and dinitrogen are continuously studied for more than 30 years [1]. A gradual shift from micro- to nanocrystalline semiconductor photocatalysts, which is, probably, the main trend in modern semiconductor photocatalysis/ photoelectrochemistry, allowed to achieve attractively high quantum efficiencies of the CO2 and N2 conversion as well as to apply a potent array of spectral methods for the elucidation of mechanistic aspects of these important photoreactions. A decrease of the photocatalyst crystal size to a few nanometers allows not only to intensify the photocatalytic synthetic reactions but also to engineer the surface and band structure of the nano-photocatalysts to direct the reactions toward desirable products. In recent years, the photochemical conversion of CO2 got under a renewed spotlight focus because of the global climatic changes induced by the over-abundant anthropogenic CO2 emissions. Also, understanding of the photocatalytic CO2 transformations on the surface of semiconductor NPs can shed light on the pre-biotic photosynthesis of simplest organic molecules that could happen billions of years ago [2, 3]. In those ages, the oceans were saturated with H2S, the Earth atmosphere was of reductive character and the solar irradiation was a lot stronger. Such conditions favored to the formation of colloidal metal sulfide NPs and their participation in the CO2 and N2 reduction to compounds with C–C and C–N bonds—acetates, propionates, ethane, ethanol, urea, amines, etc. [4, 5] that could be used as a feedstock for the synthesis of more complex organic compounds. Molecular nitrogen is very inert in ambient conditions and can be fixed typically in biochemical processes occurring in the roots of legumes, some bacteria, as well as during electric discharges in the atmosphere, thus enriching the soils with water-soluble nitrogen compounds. Alternatively, N2 can be reduced to ammonia at high pressures and temperatures over a catalyst in the chemical industry. A possible substitution of this process with new energy-saving ways of N2 fixation in mild (ambient) conditions can have enormous economic and social effect and stimulates

© Springer International Publishing AG 2018 127 O. Stroyuk, Solar Light Harvesting with Nanocrystalline Semiconductors, Lecture Notes in Chemistry 99, https://doi.org/10.1007/978-3-319-68879-4_3 128 3 Semiconductor-Based Photocatalytic Systems … the incessant search for new photochemical processes and photocatalysts that could make such technology truly sustainable and competitive one [6–8].

3.1 Photocatalytic Reduction of Carbon Dioxide

The photocatalytic reduction of CO2 into hydrocarbons or oxygenates is often called the artiﬁcial photosynthesis, because it combines the processes of solar light accumulation in the form of the products of photo-transformations—CO, CH4, CnH2n+2,CH3OH, HCHO, HCOOH, etc., with the oxidation of water to oxygen, similarly as it happens in green plants [9–13]. The idea of mimicking the natural photosynthesis aimed at the conversion of CO2 and at the release of oxygen was constantly in the focus of attention for more than three decades. Various approaches were probed including biological conversion, thermal hydrogenation processes, electrochemical reduction and photocatalytic reduction of CO2 [9–17]. The latter approach is very attractive because the breaking of C=O bonds in the very stable CO2 molecule needs a high supply of energy that can be easily obtained with the light quanta of UV and visible spectral range. The reduction of CO2—one the most stable forms of C(IV) requires also the presence of electron-supplying agents, typically water or other sacriﬁcial and abundant electron donors—hydrogen, H2S, SO2, amines, etc. (Fig. 3.1a). The photocatalytic reduction of CO2 into CH4 and CH3OH is a highly endothermic reaction with a free Gibbs energies equal to 702.2 kJ/mol 3 (CO2 +2H2O=CH3OH + /2O2) and 818.3 kJ/mol (CO2 +2H2O=CH4 +2O2), respectively [9, 11, 13, 16]. Among the advantages of the photocatalytic CO2 conversion are relatively mild conditions, simultaneous mitigation of climatic changes caused by ever-increasing anthropogenic CO2 emission, the possibility of formation of C–C bonds in the form

Fig. 3.1 Schematic illustration of different steps in the photocatalytic CO2 reduction with H2O over a heterogeneous photocatalyst. The dotted lines indicate the thermodynamic potentials for water oxidation and CO2 reduction into CO, CH4, and CH3OH. Reprinted with permissions from Ref. [16]. Copyright (2016) The Royal Society of Chemistry 3.1 Photocatalytic Reduction of Carbon Dioxide 129 of hydrocarbons, ethers and carbonic acids and others. From the other side, the efficiency of the photocatalytic CO2 reduction is restricted by a relatively low light flux intensity near the Earth surface as well as a low selectivity of the photoprocesses toward the most desirable products [9–11, 13, 14, 16]. In the light of recent global awareness of perils associated with climate changes, the problems of CO2 conversion and utilization find a new appreciation. According to the decisions made on the Paris Global Climat Conference in 2012 [18], each of 192 participating countries will work up its own strategy of diminishing of CO2 production and conversion of already emitted carbon dioxide and the efforts aimed at the photocatalytic conversion will undoubtedly have a new impetus [10, 11, 13, 14, 16]. Historically, the first studies of the photocatalytic reduction of CO2 were performed on ZnO, GaP, ZnS, then the circle of semiconductor photocatalysts was extended to TiO2, CdS, SiC, as well as various niobates, tungstates and germanates (Fig. 3.1b) [16, 19]. The efficiency and route of the photocatalytic CO2 reduction depend on a variety of factors, including the composition and band structure of photocatalyst, surface chemistry, composition of the reactant mixture, the spectral composition of the exciting light, etc. In this view, multiple approaches are developed simultaneously to the design of the photocatalytic CO2-converting systems. In general, the photocatalyst design ideology is similar to that used for the water-splitting system with account to the specifics of the chemistry and photochemistry of carbon dioxide. The design includes: (1) band structure engineering, such as doping, using quantum size effects and solid state alloying for the band edge manipulating; (2) combination of various semiconductors with matching CB and VB levels; (3) introduction of surface vacancies/defects enabling adsorption and conversion of CO2 molecules; (4) expansion of the spectral sensitivity range of the photocatalysts by sensitization with organic dyes, metal complexes and inorganic narrow bandgap NPs; (5) introduction of additional co-catalysts/co-adsorbents; (6) morphological design of semiconductor photocatalysts on the nano-level (shape/lattice face/phase engineering) creating favorable conditions for the separation and directed transport of the photogenerated charge carriers; (7) combination of various subsystems into the Z-schemes where the CO2 reduction and the water/hydrogen oxidation are separated in space [9, 15, 16, 20, 21]. Also, an important role is attributed to the constant search of new photocatalysts, in particular, among layered materials such as inorganic perovskites and carbonaceous compounds, metal-organic frameworks (MOFs) and other classes of compounds [9–12, 20–22]. The photocatalytic conversion of carbon dioxide with the participation of nanocrystalline semiconductors is a multi-faced phenomenon that can be analyzed and systematized from various aspects, for example, from the viewpoint of material science or from the viewpoint of catalysis. In the former case, a classification of reported of semiconductor-based systems for CO2 reduction can be based on composition and structure of the photoactive semiconductor phase. Such approach has been used in Chap. 2 for the classification of the photocatalytic water-splitting systems. It allows grasping the versatility of photocatalytic systems and a scope of 130 3 Semiconductor-Based Photocatalytic Systems … materials that can be used in the design of the solar-light-driven systems for CO2 fixation. We continue with such kind of classification in the present chapter. Nanocrystalline semiconductors as photocatalysts of CO2 conversion. Zinc sulfide has relatively high CB potential (around −1.8 V vs. NHE [1]) and therefore this semiconductor is one of the most viable candidates for the role of a photocatalyst of the CO2 reduction. The nanocrystalline ZnS-based photocatalysts can be produced by a variety of methods, including the controlled precipitation [23, 24], ion-exchange and hydrothermal synthesis, the latter producing the most photo-active materials [25]. Zinc sulfide NPs attached to the surface of montmorillonite were used as a working photosensitive body in reactors of various geometry for CO2 conversion to methanol, methane, and CO [26]. This study reported a distinct dependence of the yield of various reduction products on the reactor geometry. Cadmium sulfide NPs with a much lower CB potential (ECB = −0.8 V vs. NHE) can also be used for the CO2 reduction provided a suitable cocatalyst is introduced into the photocatalytic system. In particular, the efficient Vis-light-induced conversion of CO2 into CO was observed for CdS NP assemblies with carbon monoxide dehy- drogenase [27]. Hexagonal colloidal CdS NPs in N,N′-dimethylformamide (DMF) were found to be a visible-light-sensitive photocatalyst of the CO2 reduction with carbon monoxide as a main product [28]. The addition of excessive Cd2+ ions was found to affect positively the efficiency of CO2-to-CO conversion and the maximal CO yield was observed at a molar ratio of excessive Cd2+ to CdS close to 0.2 (Fig. 3.2a). A combined PL and EXAFS (extended X-ray absorption fine structure spectroscopy) showed that Cd2+ adsorbs on CdS NPs surface building up the metal sublattice and creating in this way sulfur vacancies [28]. The sulfur vacancies (□) serve as selective sites for the CO2 reduction facilitating the formation of surface 2+ complexes between Cd ions, sulfur vacancies and two CO2 molecules and facilitating the two-electron reduction of carbon dioxide to CO (Fig. 3.2b).

Fig. 3.2 a Effect of Cd2+ excess on the CO formation in the presence of CdS NPs in DMF; b Scheme of CdS NP-photocatalyzed formation of CO with the participation of sulfur vacancies introduced by excessive Cd2+. Reprinted with permissions from Ref. [28]. Copyright (1997) American Chemical Society 3.1 Photocatalytic Reduction of Carbon Dioxide 131

The visible light energy can be harvested and used for the reduction of CO2 by ternary CdIn2S4 microspheres produced by a hydrothermal synthesis [29]. The microspheres are characterized by a comparatively narrow band gap (1.68 eV) which is not favorable for a deep reduction of CO2 but allows to produce dime- thoxymethane and methyl formate when the reaction is performed in methanol [29]. The productivity of metal sulfide microspheres can be boosted by the modification with metal NPs or conjugated polymers enhancing the electron-hole separation. In particular, the decoration of Bi2WO6 microspheres with polyaniline, polypyrrole, or polythiophene allows to increase the rate of the photocatalytic CO2 reduction to methanol and ethanol [30]. The highest activity increment, by a factor of 2.8 as compared to bismuth tungstate, was observed for polythiophene. The molar ratio of CH3OH–C2H5OH produced from carbon dioxide was around 3:1 [30]. Recently, open-framework zeolite-like structures consisting of nanosized metal-chalcogenide nanoclusters were tested as visible-light-sensitive photocatalysts of the CO2 reduction. The mixed zinc-germanium-sulfide-based frameworks were found to photocatalyze the carbon dioxide reduction with water to methane and the conversion efficiency is affected by the incorporation of third metal cations (Au3+,Pd2+) into the framework [31]. The photocatalytic CO2 reduction was quite broadly studied in titania-based systems. In particular, three nanocrystalline TiO2 polymorphs—anatase, rutile, and brookite were subjected to comparative studies both in the pristine form and after a treatment with ionized helium flow creating oxygen vacancies on the NP surface [32]. Such treatment resulted in a remarkable (up to 10-times) increase of the rate of photocatalytic CO2 reduction to CO and CH4 on anatase and brookite, while rutile retained a low activity in these processes even after the treatment. In situ diffuse reflectance infrared Fourier transform spectroscopy revealed that the enhancement fi À effect originates from a higher ef ciency of CO2 anion radical formation on oxygen vacancies and Ti3+ sites of the plasma-treated titania NPs [32]. A detailed mechanistic study of photoinduced events on the surface of titania NPs in contact with CO2-saturated aqueous solutions [33] revealed multiple in- À termediates produced with CB electrons (H atoms and CO2 radicals) and with VB À holes (OH radicals and CO3 radicals). Among the secondary intermediates, CH3O and CH3 radicals were detected by the electron paramagnetic resonance. In view of the versatility of intermediary species, the principal pathway of the CO2 reduction can be affected by a broad variety of factors. In particular, the TiO2 NP size can affect the selectivity of the CO2 reduction. As the titania NP size is decreased the number of undercoordinated Ti(IV) ions on the NP surface increases making the deoxygenation of CO2 to CO more and more favorable [34]. Also, the nature and formation rate of the reduction products can be strongly affected by preferably exposed crystal faces because different TiO2 facets were found to have different band structure and band edge positions [35]. Thus, local heterojunctions between different facets can form in the nanocrystalline TiO2-based systems favoring to various reduction pathways. 132 3 Semiconductor-Based Photocatalytic Systems …

The photoactivity of titania in the carbon dioxide conversion can be increased by introducing co-catalysts, such as metal NPs, that can collect the photogenerated electrons and favor to multi-electron processes. As shown in [36], the rate of photocatalytic CO2 reduction with water vapor to methane increases from Ag to Rh to Au to Pd to Pt in line with an increase of the electron-accepting capability of the metal NPs. For a given metal, the CO2 conversion increases with a decrease of the metal NP size and typically shows a dome-shaped dependence on the metal content [36]. The CO2 reduction efficiency is limited by a competing reaction of water reduction to H2 that can be suppressed by the deposition of an additional MgO layer on the photocatalyst surface. The attachment of titania to the pores of zeolites and mesoporous silica is a traditional method of increasing the TiO2 NP stability and the photocatalytic CO2 reduction efficiency [19]. A sol-gel transformation of Ti(IV) precursors in the pores of HZSM-5 zeolite [37] or TUD-1 mesoporous silica [38] resulted in highly dispersed TiO2 NPs, the titania-based composites revealing remarkable photocatalytic activity in the CO2 reduction. The photocatalytic deposition of Ag NPs onto the titania surface results in an almost 10-fold increase of the efficiency of CO2 conversion into methanol due to the photoinduced electron accumulation of the metal NPs [39]. A sol-gel transformation of Ti(IV) precursor around the Au NPs was used to produce the so-called “yolk-shell” nanostructures (Fig. 3.3a, b) with a sole Au NP encapsulated into a mesoporous hollow titania sphere [40]. Excitation of the surface plasmon resonance (SPR) in the Au NPs creates a local electromagnetic field (Fig. 3.3c) affecting the photophysical and charge transfer processes in the TiO2 shell. As a result, the yolk-shell structure exhibited an increased rate of the photocatalytic CO2 conversion to methane as well as in the generation of C2H6, contrary to the bare TiO2 shells and TiO2 P25 [40]. Titania 12-nm NPs anchored to the reduced graphene oxide (RGO) sheets revealed a superior photocatalytic activity in the CO2 reduction to methane over the bare TiO2 NPs and a similar heterostructure based on the non-exfoliated graphite oxide [41].

Fig. 3.3 a, b TEM images of Au/TiO2 “yolk-shell” nanostructures; c simulated spatial distribution of the local electromagnetic ﬁeld enhancement on the x–y plane for the yolk-shell Au/TiO2 heterostructure. Reprinted with permissions from Ref. [40]. Copyright (2015) The Royal Society of Chemistry 3.1 Photocatalytic Reduction of Carbon Dioxide 133

A solvothermal treatment of titania nanofibers imparts them with the meso- porosity (Fig. 3.4a) resulting in a spectacular 6- and 25-fold enhancement of the photocatalytic activity in the CO2 reduction to CH4 as compared with the non-treated titania nanofibers and commercial nanocrystalline TiO2 Evonik P25, respectively [42]. The single-crystalline anatase nanocubes with exposed {100} and {001} facets (Fig. 3.4c, d) were produced by a combined hydrothermal/calcination synthesis and applied as a UV-sensitive photocatalyst of the reductive CO2 conversion to methane and methanol [43]. According to the Mott-Schottky measurements, the TiO2 nanocubes have a more negative CB potential as compared to titania nanowires and TiO2 P25 (Fig. 3.4e) favoring to the CO2 reduction not only to methane but also to CH3OH (Fig. 3.4f). The sodium niobate and tantalate perovskites were used as UV-sensitive photocatalysts of the CO2 reduction to CO, CH4, and CH3OH [44]. The lattice type influences quite strongly the photoactivity of nanocrystalline sodium niobate, the rate of photocatalytic CO2 reduction being 2-fold higher for the cubic NaNbO3 as compared to the orthorhombic polymorph [45]. Ultra-thin WO3 nanosheets (NSs) produced by the oriented attachment of tungsten oxide NPs (Fig. 3.5a–c) revealed enhanced photocatalytic performance in the CO2 reduction to methane under the illumination with simulated solar light [46]. Since the bulk WO3 is passive in this process (Fig. 3.5e), the photoactivity of WO3 NSs consisting of only six repeating unit cells of monoclinic WO3 was ascribed to the quantum size effects resulting in an increase of the CB potential above the

Fig. 3.4 a, b SEM images of mesoporous TiO2 nanoﬁbers; c, d TEM (c) and SEM (d) images of TiO2 nanocubes; e energy diagram for titania nanocubes (TC) and titania nanowires (TW); f rates of CH4 and CH3OH production with the participation of TC, TW and P25. Reprinted with permissions from Refs. [42](a, b) and [43](c–f). Copyright (2014, 2015) The Royal Society of Chemistry 134 3 Semiconductor-Based Photocatalytic Systems …

Fig. 3.5 a–c TEM images of ultra-thin WO3 NSs; d energy diagrams for NS and bulk WO3; e kinetic curves of the CH4 accumulation for WO3 NSs and commercial bulk tungsten oxide. Reprinted with permissions from Ref. [46]. Copyright (2012) American Chemical Society

standard potential of CO2 reduction to CH4 (Fig. 3.5d) [46]. Similar effects were observed for thin (*10 nm) Bi2WO6 NSs [47]. Gallium oxide NPs decorated with ultra-small (*1 nm) Ag NPs were found to be an efficient photocatalyst for the CO2 reduction to carbon monoxide [48]. The layered double zinc-gallium hydroxy-carbonates [Zn3Ga(OH)8]2CO3 Â mH2O revealed a photocatalytic activity in the CO2 conversion to CO and CH3OH under the UV illumination [49]. The photocatalyst modification with Ag or Au NPs increased the photoconversion efficiency by around 70–80%. Ultra-thin ZnGa2O4 NSs assembled into mesoporous microspheres [50] as well as Zn2GeO4 nanobelts [51] revealed a high photocatalytic activity in the CO2 reduction to CH4 with the simultaneous water oxidation to O2. Mixed CuFeO2/CuO films can be relatively easily prepared by the electrodeposition and applied as a photocatalyst of the CO2 reduction to formate at the expense of water oxidation to O2 [52, 53]. The participation of carbon dioxide in the formate generation was unambiguously confirmed by 13C isotopic studies [52]. The photocatalysts gradually degrade due to partial copper reduction, but can be easily recov- ered via the oxidative annealing and used continuously for more than a month [52]. The BiOCl nanoplates can be rendered photocatalytically active in the CO2 conversion by generating oxygen vacancies under the UV illumination [54]. The O 3.1 Photocatalytic Reduction of Carbon Dioxide 135 vacancies favored to enhanced CO2 adsorption and capture of the photogenerated charge carriers suppressing their recombination. Silicon NPs (1–4 nm) produced by the ball milling can be used as a visible-light-sensitive photocatalyst of the CO2 reduction with water exhibiting a high selectivity towards the formation of HCHO [55]. The photoreduction is observed only on Si NPs with an open surface, while the surface passivation with alkyl derivatives deteriorates the photocatalytic properties of Si NPs completely, indicating on a crucial role of CO2 adsorption on the active surface sites [55]. Similarly to the water-splitting systems discussed in Chap. 2, the photocatalytic conversion of CO2 is constantly probed on new semiconductor compounds, in particular on graphitic carbon nitride (GCN) or metal organic frameworks (MOFs). In particular, GCN modified with a Co-bipyridylate and a CoOx co-catalysts was used as a photocatalyst of the CO2 deoxygenation into CO (Fig. 3.6a) [56]. GCN is abundant with nitrogen vacancies that act as strong binding sites for CO2 molecules, while the Co-based species accelerate electron transfers between GCN, CO2, and triethanolamine [56]. The photoaction spectra of CO and H2 (a by-product) generation follow the absorption spectrum of GCN attesting to the photocatalytic character of the photoinduced transformation of CO2 in this system (Fig. 3.6b). Similarly to the H2-evolving systems, the exfoliation or nano-structuring of GCN typically results in an enhancement of the photocatalytic CO2 conversion efficiency. For example, the ammonia-assisted thermal exfoliation of GCN yields a nanoplate-like material with a plate thickness of around 3 nm [57]. A decrease in the GCN particle thickness is accompanied by a blue shift of the absorption band and an increase in the CB potential (Fig. 3.6b) as revealed by the Mott-Schottky measurements. The higher reducing potential of the photogenerated CB electrons of nanostructured GCN is reflected in an increased rate of the photocatalytic methanol and methane generation from carbon dioxide (Fig. 3.6c) [57]. Porphyrin-incorporated Zr-based MOFs were successfully tested as a solar-light-driven photocatalyst of the CO2 reduction with water to HCOOH [58]. A hybrid of MOF-253 with Ru(CO)2Cl2 complex displayed a visible-light-driven photocatalytic activity in the CO2 reduction to formate anions [59]. Cobalt imidazole zeolitic MOF (Co-ZIF-9) was shown to act as a co-catalyst for the CO2 adsorption and activation in a combination with a TiO2 photocatalyst of the CO2 reduction to CO and methane, increasing the photoconversion efficiency by a factor of 2 as compared to the bare titania [60]. MOFs can also be used as a “host” for various molecular species—electrocatalysts, photosensitizers, etc., allowing to transfer the CO2 reduction into the heterogeneous regime and increase the stability and turnover numbers of the catalysts. For example, a manganese bipyridyl complex, Mn(bpydc)(CO)3Br (bpydc = 5,5′-dicarboxylate-2,2′-bipyridine) was incorporated into a highly robust Zr (IV)-based MOF UiO-67 (Fig. 3.7a). 2+ The assembly was then sensitized with a [Ru(dmb)3] (dmb = 4,4′- dimethyl-2,2′-bipyridine) complex and used as a photocatalyst of the CO2 reduction to formate with 1-benzyl-1,4-dihydronicotinamide as a sacrificial electron donor [61]. The cyclic sequence of photoinduced electron transfers in this system can be 136 3 Semiconductor-Based Photocatalytic Systems …

Fig. 3.6 a Scheme of GCN/Co–bpy–CoOx photocatalyst; b diffuse reflectance spectrum of GCN (dashed line) and the rates of CO and H2 generation (bars). Insert in (b): kinetic curves of CO and H2 accumulation; c, d Energy diagram (c) and the rates of photocatalytic CH4 and CH3OH generation (d) for bulk and thermally-exfoliated nanostructured (NS) GCN. Reprinted with permissions from Refs. [56](a, b) and [57](c, d). Copyright (2014) American Chemical Society (a, b) and (2017) The Royal Society of Chemistry (c, d) presented by Fig. 3.7b. As compared to homogeneously soluted Mn complex, the MOF-incorporated assembly retained prolonged stability, partially because the rigid framework prohibited dimerization of a one-electron-reduced Mn complex. Similar strategy of incorporating a molecular metal-complex photocatalyst into a robust MOF structure to prevent undesirable side reactions and to enhance the I photocatalyst stability was realized for Re (CO)3(bpydc)Cl, bpydc = 2,2′- bipyridine-5,5′-dicarboxylate incorporated into the Zr-based MOF UiO-67 [61]. The combination of a molecular photocatalyst with a MOF host affords an unprecedented flexibility in structure variation. In particular, the number of Re complexes per unit cell of the MOF can be varied as n = 0, 1, 2, 3, 5, 11, 16, and 24. The highest photocatalytic activity in the CO2 reduction to CO was observed for n =3 [62] (Fig. 3.8a), as a result of a fine balance of the proximity between photoactive centers needed for the cooperatively enhanced photocatalytic activity. The most active Re complex/MOF composite photocatalyst with n = 3 was deposited onto the surface of Ag nanocubes (Fig. 3.8b) with a strong SPR on the cube edges. The SPR-induced local electromagnetic field promoted charge 3.1 Photocatalytic Reduction of Carbon Dioxide 137

Fig. 3.7 a Scheme of the synthesis of UiO-67-Mn(bpy)(CO)3Br; b Mechanism of the photocatalytic reduction of CO2 with sensitized UiO-67-Mn(bpy)(CO)3Br. Reprinted with permissions from Ref. [61]. Copyright (2015) American Chemical Society separation processes in the Re complex/MOF layer resulting in a 7-fold enhancement of the photocatalytic CO2 deoxygenation [62]. Sensitizer-based systems for CO2 photoreduction. Nanocrystalline titania sensitized with a well-known Ru(II)-bipyridyl complex (N719) coupled with a Pt counter electrode was used for the visible-light-driven photoreduction of CO2 ro HCOOH, HCHO, and CH3OH in a two-vessel geometry (Fig. 3.9a) [63]. A portion fi “ ” À À of the TiO2 lm was stained with the dye and connected by an I I3 -containing electrolyte to a Pt counter electrode. This part of the film acted as a DSSC providing electrons for the rest of the TiO2 film where the CO2 reduction to oxygenates took place (Fig. 3.9b). The photogenerated holes were transferred by iodine/iodide redox-shuttle (see Chap. 4 for details on the DSSC design) from TiO2/dye surface to the Pt counter electrode and then went via the electric circuit to another TiO2 film placed into a second vessel and separated from the first one by a Nafion membrane. Water was oxidized on this second TiO2 film to O2 providing electrons for the CO2 reduction in the first vessel thus completing the photoelectrocatalytic cycle [63]. In such way, the products of CO2 reduction were shielded from the re-oxidation on the 138 3 Semiconductor-Based Photocatalytic Systems …

Fig. 3.8 a Scheme of “Re complex/MOF/Ag nanocube” composite formation; b TEM of the composite particles. Reprinted with permissions from Ref. [62]. Copyright (2017) American Chemical Society

anode surface. The rate of CO2 conversion in such system can be increased considerably by an external bias. Nanocrystalline titania decorated with a series of Nile-Red-type dyes and Pd NPs was used as a photocathode for the conversion of CO2 into methanol with the simultaneous water oxidation on a W-doped BiVO4-based photoanode modified with a cobalt phosphate co-catalyst (Fig. 3.9c) [64]. Both electrodes were immersed into aqueous KHCO3 solution and the formation of CH3OH from carbonate ions and O2—from water molecules was confirmed by isotopic studies. The photosynthetic system was additionally biased with a voltage of *0.6 V from an independent silicon solar cell [64]. A tandem principle was realized for a nickel(II) oxide photocathode sensitized with a supramolecular Ru(II)-Re(I) complex assembly and coupled to a tantalum oxynitride photoanode modified with a cobalt oxide co-catalyst (Fig. 3.9d) [65]. The nickel oxide-based photocathode showed a selectivity toward the formation of CO. The system also used water as an electron donor and a low external bias of *0.3 eV. The p-type CuGaO2 photocathode senstized by a similar supramolecular assembly of Ru(II)-Re(I) complexes revealed a photoelectrochemical activity in the CO2 reduction by water without an additional external bias [66]. 3.1 Photocatalytic Reduction of Carbon Dioxide 139

Fig. 3.9 a Photocatalytic reduction of CO2 in a two-vessel reactor; b working principle of the two-vessel reactor; c, d schematic design of photoelectrochemical systems for the CO2 conversion with TiO2-based (c) and NiO-based (d) photo-electrodes. Reprinted with permissions from Refs. [63](a, b), [64](c), and [65](d). Copyright (2013) Elsevier (a, b), (2017) The Royal Society of Chemistry (c), and (2016) American Chemical Society (d)

A Ru-bipyridyl sensitizer was also applied to extend the spectral sensitivity range of nanocrystalline CuCo2O4 [67]. The sensitized spinel acts as a photocatalyst of the CO2 deoxygenation to CO. Mesoporous N-doped Ta2O5 microspheres sensitized with a Ru-bipyridyl complex showed a photocatalytic activity in the visible-light-driven CO2 conversion into HCOOH [68]. Composite TiO2/Zn phthalocyanine NPs produced by a combined microwave/ hydrothermal process can be applied as a selective photocatalyst of the CO2 reduction to methanol under simulated solar light illumination [69]. The functionalization of TiO2 NPs with various aminosalicylic acids (ASA) results in the formation of surface charge transfer complexes that extend the absorption range of titania far into the visible range (Fig. 3.10a, blue curve) [70]. The photoinduced charge transfer in such complexes occurs directly from the 140 3 Semiconductor-Based Photocatalytic Systems … highest occupied molecular orbital of ASA molecules into the CB of TiO2 NPs and then—to adsorbed CO2 molecules. As a result, the absorption band edge of a ternary TiO2-ASA-CO2 systems shifts even further into the visible range as compared to the binary TiO2-ASA charge transfer complex (Fig. 3.10a, red curve). The light sensitivity range of GCN can be greatly extended by coupling it with Co porphyrin [71] or Co bipyridyl complexes [72]. The sensitized GCN showed a 5-fold (Co bipyridylate) and a 13-fold (Co porphyrinate) increments of the rate of photocatalytic CO2 reduction to CO as compared to the individual GCN. Also, by decreasing the lateral size and thickness of GCN particles the photoconversion efﬁciency can be additionally enhanced to a CO yield of 17 lmol/g/h [71]. A dependence of the reaction rate on the photoexcitation wavelength (the photoaction spectrum) mimics the GCN/Co-porphyrin absorption spectrum (Fig. 3.10b) showing two distinct sensitivity ranges of GCN (k < 470 nm) and Co-porphyrin (k > 500 nm). In tandem systems comprising GCN and Ru(II)-bipyridyl complexes, the rate and pathway of the CO2 reduction depend on the bipyridyl substituent X in the 4-position [73]. In particular, the tandems based on X = COOH and X = PO3H2 the main product of the CO2 reduction was HCOOH, while for X = CH2PO3H2 the photoprocess yielded CO and HCOOH with a relatively high selectivity toward carbon monoxide (40–70%). The difference arises from a photoinduced transformation of the CH2PO3H2-substituted Ru bipyridyl complex into a polymeric species active speciﬁcally toward the CO formation [73]. Polymeric Ru-bipyridyl complexes were also used as sensitizers/electrocatalysts to increase the rate of photocatalytic CO2 reduction to formate over indium phosphide [74]. Doped semiconductor nano-photocatalysts of CO2 conversion. Doping of the wide-bandgap semiconductors, such as titanium dioxide, with metal ions/non-metal atoms is one of the most frequent and productive strategies both of increasing the

Fig. 3.10 a Absorption spectra of TiO2 NPs, binary TiO2-3ASA charge transfer complex, and ternary TiO2-2ASA-CO2 system; b Rate of CO production over GCN/Co-porphyrin hybrid (bars) compared with the hybrid absorption spectrum (red line). Reprinted with permissions from Refs. [70](a) and [71](b). Copyright (2013) American Chemical Society (a) and (2017) Elsevier (b) 3.1 Photocatalytic Reduction of Carbon Dioxide 141 photoreaction efficiency due to the recombination inhibition and of extending the spectral sensitivity range of the photocatalysts as a result of the participation of localized/delocalized dopant levels in the light absorption. In Chap. 2 we have discussed doping methods for the photocatalytic H2-evolving systems. This approach is also broadly used in the photosynthetic CO2-conversion systems [19]. For example, the incorporation of Pd, Cu, and Mn ions into the TiO2 lattice imparts the semiconductor with the sensitivity in the visible range (Fig. 3.11a) [34, 35]. The dopants are present in the titania lattice as –O–M–O– fragments and can actively participate in the trapping of the phogenerated charge carriers—the VB holes in the case of Pd and Cu and the CB electrons in the case of Mn (Fig. 3.11b). As a result, doping increases the efficiency of the photocatalytic CO2 reduction to methane [75]. In-doped nanocrystalline TiO2 was applied as a photo-active phase for the photocatalytic CO2 conversion in microchannel monolith photoreactors [76]. These systems produce a broad range of reduction products with a product population decreasing in the following sequence: CO > CH4 >C2H6 >C2H4 >C3H6. After a multi-parameter optimization of the photoreactor performance, the quantum yields of CO and methane reached 0.1 and 0.022%, respectively. Doping of the mesoporous titania with In was also reported to change the basic CO2 reduction product from CO to CH4 increasing the light harvesting efficiency by a factor of around 8 [77]. 2+ Doping with Ni prohibits the growth of TiO2 nanocrystals and the anatase- to-rutile conversion during the thermal synthesis and results in a partial substitution of Ti4+ with nickel ions [78]. The dopant extends the absorption edge of titania to longer wavelength and provides traps for the photogenerated charge carriers thus decreasing the recombination efficiency. The Ni-doped nanocrystalline titania loaded onto the quartz optical fibers can be used in a monolith reactor for the CO2 reduction with water vapors under the UV/Vis illumination [78]. A similar effect on the growth of titania NPs was observed for Ce doping [79]. After the deposition on mesoporous SBA-15 silica TiO2-Ce was applied as a visible-light-sensitive photocatalyst of the CO2 conversion into CO and CH4. The photocatalytic conversion of CO2 in the presence of TiO2-Cu/SBA-15 composite with a 45 wt.% loading of

Fig. 3.11 a Diffuse reflectance spectra of TiO2 and Cu, Pd, and Mn-doped (1 mol%) titania; b, c mechanistic diagrams of the CO2 photoreduction on TiO2–Cu (b) and TiO2–Mn (c). Reprinted with permissions from Ref. [75]. Copyright (2017) American Chemical Society 142 3 Semiconductor-Based Photocatalytic Systems … the photoactive component yields methanol as a principal product [80]. The nanocrystalline titania doped with Ce(IV) revealed an extended spectral sensitivity range and the highest photocatalytic activity in the CO2 reduction at 0.28 mol% dopant content [81]. Nanocrystalline ZnS doped with Ni(II) showed a high selectivity toward the formation of methyl formate as a result of the photocatalytic CO2 reduction in methanol [25]. The highest yields were observed at a *0.3 mol% dopant content. The melamine pyrolysis in the presence of ammonium molybdate results in the formation of mesoporous Mo-doped GCN [82]. The Mo doping enhanced the GCN absorptivity in the visible range proportionally to the dopant content (Fig. 3.12a). The Mo-doped GCN revealed an enhanced photocatalytic activity in the CO2 reduction to CO and CH4 compared with the pure g-C3N4 [82]. Nitrogen-doped anatase NPs (10–20 nm) with predominantly exposed (001) faces covered with RGO sheets were tested as a photocatalyst of the CO2 reduction with water to methane [83]. Coupling with RGO resulted in an 11-fold increase of the CO2 conversion efficiency. Oxygen-enriched titania NPs were produced by the thermal decomposition of a peroxy-titanium complex [84]. The oxygen doping shifted the absorption threshold of TiO2 NPs from 390 nm to around 420 nm favoring to the visible-light-driven photocatalytic conversion of CO2 into methane. Titania NT arrays produced by the anodization of a Ti foil followed by the hydrothermally-assised doping with vanadium and nitrogen (Fig. 3.13a) were reported to be an efficient photocatalyst of the CO2 reduction to CH4 [85]. The calcination of a zeolite-like zinc-imidazolate framework on air results in the formation of mesoporous ZnO that retains the zeolitic structure up to 300 °C, while converting to wurtzite zinc oxide at higher temperatures [86]. Along with this conversion, carbon doping of ZnO takes place as well as the deposition of a

Fig. 3.12 Diffuse reﬂection spectra of a Mo-doped GCN with a different dopant content (values on ﬁgure correspond to the molar% of Mo with respect to melamine) and b carbon-doped ZnO nanostructures (values correspond to the calcination temperature). Reprinted with permissions from Refs. [82](a) and [86](b). Copyright (2016) Elsevier (a) and The Royal Society of Chemistry (b) 3.1 Photocatalytic Reduction of Carbon Dioxide 143

Fig. 3.13 a SEM image of V,N-doped titania NT arrays [85]; b, c SEM and TEM images of carbon nanosphere template (b) and C-doped titania hollow spheres (c); d STEM image and element distribution profiles for a TiO2 hollow sphere (b–d). Reprinted with permissions from Ref. [134]. Copyright (2017) The Royal Society of Chemistry carbonaceous layer on the ZnO surface imparting the material with the visible light sensitivity (Fig. 3.13b). As a result, the C-doped mesoporous ZnO can be used as a solar-light-driven photocatalyst of the CO2 reduction to methanol, the conversion efficiency being around 6-times higher than for the undoped reference ZnO NRs [86]. Similarly, the annealing of titania layers deposited onto a carbon nanosphere template yielded hollow TiO2 spheres with a surface carbon layer and carbon-doped inorganic matrix [84]. The doped titania hollow spheres showed twice as high photocatalytic activity in the CO2 reduction to CH4 as the reference TiO2 P25. Binary semiconductor nanoheterostructures for CO2 photoreduction. The photocatalytic conversion of CO2 in the presence of methanol over a CuO/TiO2 heterostructure results in the preferential formation of methyl formate [22]. At that, methanol serves as a reactant and as a VB hole scavenger, while CO2 is reduced by the photogenerated CB electrons [22]. Deposition of Cu2O NPs onto titania was reported to result in enhanced adsorption of CO2 and simultaneous inhibition of water adsorption [87]. Simultaneously, Pt NPs deposited onto TiO2 acts as electron “pools” promoting multi-electron photoinduced reactions. The summary effect of Cu2O and Pt NPs results in complete suppression of the water reduction on the titania surface and the selective reduction of CO2 with CH4 as a sole product of the photoreaction [87]. Titania and Cu2O can be separated in space and used as a photoanode and a cathode, respectively, in a photoelectrochemical system for the CO2 reduction. In this way, the Cu2O can be protected against the oxidative photocorrosion with the photogenerated VB holes [88]. Ordered and hierarchically porous CeO2/TiO2 heterostructures were produced by using SBA-15 zeolite as a sacrificial template and used as a visible-light-sensitive photocatalyst of the CO2 reduction with water [89]. The enhancement factor of the photocatalytic reduction of CO2 to CO over the TiO2/CeO2 nanocomposites depends on the crystalline structure of TiO2, being the highest for the rutile/CeO2 composites of all the titania polymorphs [90]. 144 3 Semiconductor-Based Photocatalytic Systems …

By using self-ordered polystyrene microspheres as a sacrificial template ordered macroporous titania can be produced. The decoration of such materials with a nanolayer of ceria extends its spectral sensitivity to the visible range [91]. The macroporous TiO2/CeO2 retains an ordered character of the original polystyrene photonic crystal favoring to a higher light absorptivity and can be used as a solar-light-driven photocatalyst of the CO2 reduction to CO with water [91]. Titania coupling with GCN allows for the spatial separation of charge carriers since the photogenerated electrons are collected in the lower-positioned titania CB and the VB holes (photogenerated in both semiconductors) are supplied to the reactants through the VB of carbon nitride. As a result of the charge separation, the titania/GCN composites displayed an ehnanced photocatalytic activity in the CO2 reduction [92]. A similar enhancement effect in the CO2 reduction to methane was observed also for GCN/KNbO3 [93] and GCN/NaNbO3 [94]. Efficient separation of the photogenerated charge carriers can be realized in a binary heterostructure of boron carbide B4C with GCN [95]. The p-type conducting semiconductor B4C forms with the n-type GCN a heterojunction with a favorable band edge offsets, allowing for the interfacial electron transfer from the photoexcited B4C to GCN. The VB hole in B4Cisfilled by oxidizing H2 to atomic hydrogen and by transferring the photogenerated CB electrons to GCN (Fig. 3.14a), while CO2 is reduced primarily on the surface of platinum particles deposited onto the graphitic carbon nitride. The chain electron transfers result in the photocatalytic reduction of CO2 to CO and further hydrogenation of carbon monoxide to methane and ethane [95]. The sheet-like GCN can be used as a “mat” for the growth of other nanocrystalline semiconductors, such as indium oxide. The GCN/In2O3 heterostructures revealed an enhanced photocatalytic activity in the reduction of water and CO2 due to spatial separation of the CB electrons and VB holes between the heterostructure components [96]. On the other hand, the GCN sheets can be deposited as a shell around inorganic core crystals, such as NRs and NWs. The core/shell LaPO4/GCN NWs (Fig. 3.14b, c) displayed an enhanced photocatalytic activity in the CO2 conversion into CO as compared with individual components [97]. The GCN NPs with a size of around 3 nm formed via the urea condensation on the surface of TiO2 brookite nanocubes (Fig. 3.14d, e) acted as a spectral sensitizer allowing to reduce CO2 selectively to CH4 under Vis-light illumination [98]. Vanadium-doped TiO2 sensitized by graphene NSs was used as a photocatalyst for the model endothermic conversion of potential environmental pollutants into photosynthetic products [99]. On the first step, methylene blue dye as a model persistent pollutant was mineralized to CO2 and then carbon dioxide reduced photocatalytically into CH4,CH3OH, and CH3CH2OH with an apparent quantum efficiency of *5% at 420 nm [99]. By combining Fe2O3 NTs with branched SnO2 NRs with predominantly exposed {110} and {101} faces the photoelectrocatalytic reduction of CO2 to methanol can be accelerated by more than 7 times [100]. 3.1 Photocatalytic Reduction of Carbon Dioxide 145

Fig. 3.14 a Scheme of photoinduced charge separation in p-B4C/n-GCN heterojunction (potentials are given versus NHE). b, c TEM images of LaPO4/GCN core/shell nanowires. d, e TEM images of TiO2/GCN nanocomposite. Reprinted with permissions from Refs. [95](a), [97] (b, c), and [98](d, e). Copyright (2016, 2017) Elsevier

In metal chalcogenide/titania composites with a proper CB and VB level positions, the CO2 reduction and the donor (for example, water) oxidation occur on the surface of titania and metal chalcogenide NPs, respectively. Both branches of the photoprocess can be additionally separated in space to avoid the re-oxidation of CO2 reduction products and to promote the formation of C–C bonds. For example, by combining two subsystems—the Moorella thermoacetica bacteria decorated with CdS NPs and TiO2 NPs loaded with Mn(III) phthalocyanine into a Z-system, the CO2 reduction to acetic acid can be achieved [101]. Both subsystems are connected by a donor/acceptor cysteine/cystine couple. The cysteine (Cys) gets oxidized on the surface of CdS NPs to cystine (CySS) and then CySS is regenerated to Cys on the surface of the phthalocyanine-functionalized titania (Fig. 3.15a). Titania NTs can be sensitized to the visible light by CdS and Bi2S3 NPs [102]. The decoration of TiO2 NTs with bismuth sulﬁde NPs favors to the formation of methanol and increases the total CO2 reduction rate by a factor of 2.2 as compared to the bare titania NTs [102]. Titania NTs sensitized by mixed CdSeTe NPs were used as a photoelectrocatalyst of the CO2 reduction to methanol in a two-cell reactor [103]. The bandgap of mixed cadmium selenide-telluride NPs was adjusted to 1.24 eV (corresponding to kbe * 1000 nm), thus allowing for harvesting the entire visible and near-IR irradiation. The efﬁciency of photocatalytic reduction of CO2 with water vapor over TiO2/ CdSe nanoheterostructures was found to depend on the CdSe NP size as a result of a size-dependence of the cadmium selenide CB position (Fig. 3.15b) [104]. The main reduction product was methane with CH3OH, CO, and H2 present as secondary admixtures. 146 3 Semiconductor-Based Photocatalytic Systems …

Fig. 3.15 a Scheme of a “Moorella thermoacetica—CdS /TiO2–Mn phthalocyanine” tandem system for the CO2 conversion into acetic acid; b Energy diagram of TiO2/CdSe heterostructures with bulk and nanocrystalline cadmium selenide. Reprinted with permissions from Refs. [101] (a) and [104](b). Copyright (2016, 2010) American Chemical Society

Mixed cadmium zinc sulﬁde Cd0.2Zn0.8S NPs were deposited onto a UiO-66 MOF resulting in a nanocomposite with enhanced spatial separation of the photogenerated charge carriers [105]. The heterostructure can be used as a photocatalyst of the CO2 reduction to methanol at the expense of water oxidation, showing remarkable chemical/photochemical stability.

3.2 Photocatalytic Fixation of Dinitrogen

The photochemical reduction of dinitrogen (N2) is probably second by importance, after the CO2 fixation, photosynthetic process that can be used directly for the accumulation and storage of the solar energy in a chemical form. Also, the N2 reduction to ammonia can provide the ways to valuable raw materials of chemical industry and fertilizers. Nowadays, the most important process of the N2 fixation is a conventional catalytic Haber-Bosch reaction between N2 and H2. This process is though energy-demanding, requires non-renewable feedstocks for the hydrogen generation and suffers from the catalyst poisoning. Therefore, alternative ways of the N2 fixation are constantly developed including thermal and non-thermal plasma-based processes, biological dinitrogen fixation, metal-complex catalysis and photocatalytic N2 transformations [106]. The dinitrogen reduction results in energy accumulation of 678 kJ/mol and the realization of this process using the solar light can potentially allow substituting modern energy-demanding catalytic technologies with mild photosynthesis-like processes thus contributing to the alleviation of global climate and energy diver- sification problems. The feasibility of photochemical reduction of N2 to ammonia and traces of N2H4 over desert sands of various origins was first shown in 1983 [107]. The N2 fixation efficiency was found to depend on the content of TiO2 in the sand samples. This work indicated that close to *107 tons of dinitrogen per year can be converted into 3.2 Photocatalytic Fixation of Dinitrogen 147 ammonia in the semiarid desert conditions under the solar light illumination of the desert sands [107]. A special feature of the photocatalytic nitrogen fixation is a strong dependence of the conversion efficiency on the nature of lattice defects of semiconductor photocatalysts. In some cases, the presence of defects—anionic and cationic vacancies, is an obligatory condition for a semiconductor to reveal photocatalytic properties in the N2 reduction [108–110]. For example, the microwave treatment of GCN results in the formation of numerous pores as well as in the generation of nitrogen vacancies via the NH3 elimination [108]. The newly generated vacancies act as perfect N2 adsorption sites because of matching physical sizes of a dinitrogen molecule and a nitrogen vacancy [110]. Also, the vacancies can trap the photogenerated charge carriers thus allowing to avoid the recombination processes. Finally, the generation of vacancies with corresponding mid-bandgap electronic states results in an increase of the light absorbance of GCN in the visible spectral range (Fig. 3.16a) [108, 110]. Typically, there exists an optimal vacancy density producing the highest photocatalytic activity in the N2 reduction. For example, for the microwave-treated GCN such density can be created at a 25-min treatment providing the best performance for the photoinduced ammonia generation (Fig. 3.16b) [108]. Figure 3.16b shows that the dinitrogen reduction kinetics is very similar to the kinetics of the water reduction to H2, the process rate remaining roughly constant for many hours of illumination. Also, the microwave-treated GCN retains a steady photocatalytic activity for a long illumination period 20 h and more (Fig. 3.16b, insert). Similarly to the heat treatment, the porosity and nitrogen vacancies in GCN can be introduced by a treatment with concentrated acids, such as HCl [109] and HNO3 [111]. The HCl treatment, in particular, results in a spectacular *13-fold increment

Fig. 3.16 a Diffuse reflectance spectra of GCN produced at 550 °C (CN-550) and the products of microwave treatment with a varied duration (denoted as MCN-x, x—duration); b kinetic curves of the ammonia accumulation over different GCN samples. Reprinted with permissions from Ref. [108]. Copyright (2016) Elsevier 148 3 Semiconductor-Based Photocatalytic Systems … of photocatalytic GCN activity in the N2 reduction to ammonia as compared with the bulk material [109]. The thermal condensation of an adduct produced in methanol at the interaction between melamine and concentrated nitric acid was found to yield a sponge-like porous GCN material abundant with nitrogen vacancies. The vacancies impart such materials with advanced photocatalytic properties in the dinitrogen conversion into ammonia [112]. As compared to GCN produced by the conventional melamine condensation, the porous sponge-like GCN exhibited a *27-times higher photocatalytic activity. In a similar way, a series of metal-sulfide NPs and metal-sulfide/GCN nanoheterostructures were prepared, in particular, Cd–Zn–Sn–S[113] and Cd–Zn–Sn–S/GCN [114], Cd–Zn–Mo–S/GCN [115], Cd–Ni–Mo–S[116]. Such composites are abundant with the sulfur vacancies in metal chalcogenide NPs that favor to the adsorption and activation of N2 molecules. 3+ Aqueous chalcogels comprising ultra-small nanosized [Mo2Fe6S8(SPh)3] and 4− [Sn2S6] sub-units (SPh is a thiophenolate anion) were found to photocatalyze the reduction of N2 to ammonia under the solar light illumination thus mimicking Fe–Mo–S active centers of the N2-fixating micro-organisms [117]. Oxygen vacancy-rich BiOCl NPs revealed photocatalytic properties in the N2 reduction [118]. As the lattice of bismuth oxychloride is strongly anisotropic, the oxygen vacancies on {001} and {010} lattice facets have non-equal energies and can coordinate N2 molecules in a different way (Fig. 3.17a), thus favoring to different pathways of the N2 reduction. The dinitrogen photoreduction on {001} faces, where N2 molecules are coordinated by a sole nitrogen atom, results in the preferential formation of ammonia (Fig. 3.17b), while the photoreduction of N2 coordinated to the {010} face by both N atoms yields hydrazine as a principal intermediate, followed by the N2H4 conversion into NH3 (Fig. 3.17c) [118]. Similarly, BiOBr NSs with preferentially exposed {001} facets revealed a high photocatalytic activity in the N2 reduction to ammonia due to abundant presence of oxygen vacancies [119]. The process requires no additional hole scavengers or co-catalysts and proceeds at the ambient humidity, pressure, and temperature. The 2–5-nm bismuth monoxide NPs were reported to be an efficient photocatalyst of N2 reduction to ammonia capable of producing up to *1230 mmol NH3 per (g Â h) which is by three orders of magnitude faster than for conventional Fe-doped titania photocatalysts [120]. The most probable reason for the high photoactivity of BiO NPs is the strong adsorption and activation of N2 molecules on Bi-rich (and, therefore, rich with oxygen vacancies) NP surface. The introduction of co-catalysts capable of coordinating N2 molecules and weakening of the triple N–N bond results in a pronounced enhancement of the photocatalytic dinitrogen reduction. For example, mixed Ru(II) complexes with EDTA and chloride anions can coordinate molecular nitrogen and facilitate N2 reduction by the photoelectrons generated in silver-doped CdS/RuO2/Pt 15 nanoheterostructure [121]. The isotopic N studies confirmed that N2 was the sole nitrogen source in the photoproduced ammonia. 3.2 Photocatalytic Fixation of Dinitrogen 149

Fig. 3.17 a Crystal structure of BiOCl and the corresponding cleaved {001} and {010} facets (left part); the terminal end-on adsorption structure of N2 on {001} surface and side-on bridging adsorption structure of N2 on {010} surface of BiOCl (right part); b, c kinetic curves of NH3 (b) and N2H4 (c) formation of {001} and {010} facet exposed BiOCl. Reprinted with permissions from Ref. [118]. Copyright (2016) The Royal Society of Chemistry

A modification of partially exfoliated and protonated GCN with RGO NSs increases the rate of photocatalytic N2 reduction to ammonia by factors of around 42, 8, and 4 as compared to the bulk GCN, exfoliated GCN and a composite of bulk GCN with RGO, respectively [122]. The strong PL quenching in the composites of exfoliated GCN and RGO attests to an efficient charge transfer from the photocatalyst (GCN) to the co-catalyst (RGO). A composite of titania with poly(3-methylthiophene) revealed a photocatalytic activity in the visible-light-driven conversion of N2 into ammonia and ammonium salts when exposed to solar-like “white” light at the ambient humidity and temperature [123]. Similarly to the CO2 photoconversion, doping is also a potent tool for influencing the photocatalytic properties of semiconductor nanomaterials in the nitrogen fixation. At that, the most spectacular efficiency increments were observed for doping with iron in its various forms. For example, FeIII-doping of the 150 3 Semiconductor-Based Photocatalytic Systems … nanocrystalline titania with preferentially exposed {101} facets results in an almost 4-fold increase in the photocatalytic N2 reduction rate [124]. Though typically the role of iron dopants is explained in a conventional way by trapping of the photogenerated charge carriers, a more probable reason for the selective activation with Fe dopants can be in the formation of donor-acceptor bonds between N2 and Fe centers thus resulting in an activation of otherwise inert dinitrogen molecule. In particular, by doping of GCN with iron the photocatalytic ammonia generation rate can be increased by a factor of 13 and higher [125]. Iron ions were found to be incorporated into the interstitial positions of GCN and stabilized by electron-rich heptazine fragments via donor-acceptor interactions. The Fe centers participate in the N2 chemisorption and the photoinduced electron transfers from GCN to adsorbed dinitrogen as confirmed by the density functional theory calculations [125]. In particular, HOMO of N2 become delocalized while LUMO is hybridized with the iron-related orbitals as a result of the Fe–N2 interactions, thus facilitating the electron transfers from GCN to N2. + Copper(I)—nitrogen vacancy couples in Cu -doped GCN serve as the N2 adsorption and activation sites promoting photoinduced electron transfers from the semiconductor to adsorbed dinitrogen molecules converting them to ammonia ions [126]. Doping with iron enhances the photocatalytic N2 reduction over mesoporous Ta2O5 by a factor of two with an optimal Fe loading of around 1 wt.% [127]. The iron can be introduced as Fe2O3 deposited onto other wider-bandgap semiconductors to facilitate the photoinduced charge transfers. In particular, TiO2/ Fe2O3 systems displayed a much higher efficiency of the photocatalytic N2 reduction to ammonia and hydrazine as compared to sole titania (Fig. 3.18a) [128]. The formation of both products can be imagined as a step-wise dinitrogen reduction and the addition of H atoms produced from the water reduction with CB electrons to a N2 molecule (Fig. 3.18b). The photocatalytic fixation of dinitrogen was also observed in the presence of nanocrystalline iron titanate and confirmed unambiguously by the isotopic studies [129]. Iron titanate also exhibited a suppressed activity in the undesirable process of the photocatalytic NH3 oxidation as compared with pure titania. Hydrated iron oxide NPs stabilized in the cavities of Nafion membranes were found to photocatalyze a nitrogen fixation cycle involving both the N2 reduction and oxidation in aerated aqueous solutions [130]. The photogenerated CB electrons reduce dinitrogen to ammonia, while the VB holes oxidize water to O2 and N2—to nitrite, as shown by the following brutto-equations [130]:

À þ ! ; þ ! þ ; 6eCB +N2 +6H 2NH3 2H2O+4hVB O2 +4H þ ! À þ ; þ ! þ À: 6hVB +N2 +4H2O 2NO2 +8H 2hVB +N2 +O2 +2H2O 4H + 2NO2

The photocatalytic N2 reduction is promoted in conditions allowing for multi-electron processes to occur, as the total conversion of a dinitrogen molecule to two ammonia molecules requires six electrons. To create such conditions various 3.2 Photocatalytic Fixation of Dinitrogen 151

Fig. 3.18 a Rate of the photocatalytic generation of NH3 and N2H4 from N2 over different photocatalysts. Reprinted with permissions from Ref. [128]. Copyright (2017) Elsevier approaches are probed including the introduction of electron-collecting co-catalysts, for example, noble metal NPs. The introduction of a co-catalyst is probably the most straightforward way of facilitating the multi-electron N2 reduction. In particular, the decoration of TiO2/ Fe2O3 composites with Pd NPs that can collect the photogenerated electrons results in an appreciable increase of the photocatalytic NH3 and N2H4 generation rate (Fig. 3.18a) [128]. Nanostructures of titania with ruthenium, rhodium, palladium, and platinum NPs were reported to have photocatalytic properties in the N2 reduction to ammonia, the catalytic activity of metal NPs decreasing as Ru > Rh > Pd > Pt [131]. The activity sequence mirrors the efficiency of primary separation of the photogenerated electrons and holes between metal NPs and semiconductor crystals, respectively. The silicon NR arrays decorated with Au NPs were shown to reduce atmospheric N2 to ammonia in a photoelectrochemical regime, the ammonia yield increasing under an elevated N2 pressure [132]. In the presence of sulfite ions as a sacrificial electron donor, the photoprocess directly yielded ammonia sulfate which is a fer- tilizer of industrial importance. Exfoliated layered semiconductor materials with a high electron density can also favor to multi-electron transfers as exemplified for the ultra-thin MoS2, while bulk molybdenum disulfide is passive in the nitrogen reduction [133]. Partially exfoliated GCN displayed around 5-times higher photocatalytic activity in the N2 reduction to NH3 as compared to the pristine bulk graphitic carbon nitride [122]. Concluding the discussion of the photosynthesis-like systems for the photocatalytic CO2 and N2 conversion, we define the “hottest” pathways of further progress of this very promising field. Similarly to the hydrogen photoproduction systems, here a search for new semiconductor nanomaterials with unexpected properties is of the paramount importance. For example, a high potential can be noted for the metalorganic frameworks allowing for a precise molecular design and control of the geometry 152 3 Semiconductor-Based Photocatalytic Systems … and composition of the active sites, which is even more important for multi-electron CO2 and N2 reduction to desired products, than for the reduction of water to H2. A spatial design of semiconductor-based nanoheterostructures and nanoassem- blies aimed at the separation of the reduction and oxidation sites and the inhibition of the re-oxidation of CO2 and N2 reduction products is, therefore, a next important thing in the progress of such photocatalytic synthetic systems. Such design can include studying of various hierarchical structures, like hollow spheres with incorporated co-catalysts and sensitizers, multi-faceted materials with a different reactivity of faces toward the CO2 and N2 reduction, spatial separation of the reduction and oxidation semi-reactions by transforming a photocatalytic process into a photoelectrochemical one, etc. Similarly to the water splitting systems and even to a much higher extent, the efficiency of the photocatalytic reduction of CO2 and N2, as well as the selectivity of these processes, can be influenced by the co-catalysts of multi-electron reactions. The challenge of creating efficient catalysts of concerted 4–8-electron processes is of the multi-disciplinary nature, requiring a convergence of efforts in the photochemistry, electrochemistry, and catalysis and promising in the future to make the CO2 and N2 conversion technologies competitive to the presently used catalytic processes. The photocatalytic systems for CO2 and N2 conversion can be enhanced and modified by doping and/or creating of additional lattice defects—vacancies. The vacancies can favor to the CO2 and N2 adsorption and, at the same time, vary in the adsorption geometry, thus providing possibilities for the formation of different products and determining not only the efficiency but also the selectivity of the photocatalytic transformations. Also, as the natural processes of the CO2 and N2 photofixation occur mostly in the living microorganisms, the photocatalytic systems utilizing both the potential of semiconductor nanomaterials and that of bio-mimicking approaches (ferments and the analogs of the active ferment centers, bacteria, etc.) can pave the way to very efficient and selective phototransformations of dinitrogen and carbon dioxide.

References

1. Grätzel M (ed) (1983) Energy resources through photochemistry and catalysis. Academic Press, Inc, New York 2. Zhang XV, Martin ST, Friend CM et al (2004) Mineral-assisted pathways in prebiotic synthesis: photoelectrochemical reduction of carbon (IV) by manganese sulﬁde. J Am Chem Soc 126:11247–11253. doi:10.1021/ja0476415 3. Zhang XV, Ellery SP, Friend CM et al (2007) Photodriven reduction and oxidation reactions on colloidal semiconductor particles: implications for prebiotic synthesis. J Photochem Photobiol A 185:301–311. doi:10.1016/j.jphotochem.2006.06.025 4. Chittenden GJ, Schwartz AW (1981) Prebiotic photosynthetic reactions. Biosystems 14:15–32 5. Olson JM, Blankenship RE (2004) Thinking about the evolution of photosynthesis. Photosynth Res 80:373–386 References 153

6. Shiavello M, Sclafani A (ed) (1985) Photoelectrochemistry, photocatalysis and photoreaction. NATO-ASI Series, Reidel, Dordrecht 7. Shiavello M (ed) (1988) Photocatalysis and environment trends and application. Kluwer 8. Davies JA, Boucher OL, Edwards JG (1995) The question of artificial photosynthesis of ammonia on heterogeneous catalysts. In: Neckers OC, Volman DH, von Bünou G (eds) Advances in photochemistry. New York, Wiley 9. Tu W, Zhou Y, Zou Z (2014) Photocatalytic conversion of CO2 into renewable hydrocarbon fuels: state-of-the-art accomplishment, challenges, and prospects. Adv Mater 26:4607–4626. doi:10.1002/adma.201400087 10. Nikokavoura A, Trapalis C (2017) Alternative photocatalysts to TiO2 for the photocatalytic reduction of CO2. Appl Surf Sci 391:149–174. doi:10.1016/j.apsusc.2016.06.172 11. Wen F, Li C (2012) Hybrid artificial photosynthetic systems comprising semiconductors as light harvesters and biomimetic complexes as molecular cocatalysts. Acc Chem Res 46:2355–2364. doi:10.1021/ar300224u 12. Kim W, McClure BA, Edri E, Frei H (2016) Coupling carbon dioxide reduction with water oxidation in nanoscale photocatalytic assemblies. Chem Soc Rev 45:3221–3243. doi:10. 1039/c6cs00062b 13. Roy SC, Varghese OK, Paulose M, Grimes CA (2010) Toward solar fuels: photocatalytic conversion of carbon dioxide to hydrocarbons. ACS Nano 4:1259–1278. doi:10.1021/ nn9015423 14. Ozin GA (2015) Throwing new light on the reduction of CO2. Adv Mater 27:1957–1963. doi:10.1002/adma.201500116 15. Qu Y, Duan X (2013) Progress, challenge and perspective of heterogeneous photocatalysts. Chem Soc Rev 42:2568–2580. doi:10.1039/c2cs35355e 16. Chang X, Wang T, Gong J (2016) CO2 photo-reduction: insights into CO2 activation and reaction on surfaces of photocatalysts energy. Environ Sci 9:2177–2196. doi:10.1039/ c6ee00383d 17. Chen Y, Lewis NS, Xiang C (2015) Operational constraints and strategies for systems to effect the sustainable, solar-driven reduction of atmospheric CO2. Energy Environ Sci 8:3663–3674. doi:10.1039/c5ee02908b 18. Atwater HA (2016) Photonics and light science wish list from COP21 in Paris. ACS Photonics 3:155–157. doi:10.1021/acsphotonics.5b00749 19. Izumi Y (2013) Recent advances in the photocatalytic conversion of carbon dioxide to fuels with water and/or hydrogen using solar energy and beyond. Coord Chem Rev. doi:10.1016/j. ccr.2012.04.018 20. Dhakshinamoorthy A, Navalon S, Corma A, Garcia H (2012) Photocatalytic CO2 reduction by TiO2 and related titanium containing solids. Energy Environ Sci 5:9217–9233. doi:10. 1039/c2ee21948d 21. Sun H, Wang S (2014) Research advances in the synthesis of nanocarbon-based photocatalysts and their applications for photocatalytic conversion of carbon dioxide to hydrocarbon fuels. Energy Fuels 28:22–36. doi:10.1021/ef401426x 22. Qin S, Xin F, Liu Y et al (2011) Photocatalytic reduction of CO2 in methanol to methyl formate over CuO–TiO2 composite catalysts. J Colloid Interface Sci 356:257–261. doi:10. 1016/j.jcis.2010.12.034 23. Kanemoto M, Hosokawa H, Wada Y et al (1996) Role of surface in the photoreduction of carbon dioxide catalysed by colloidal ZnS nanocrystallites in organic solvent. J Chem Soc Faraday Trans 92:2401–2411. doi:10.1039/FT9969202401 24. Zhou R, Guzman MI (2014) CO2 reduction under periodic illumination of ZnS. J Phys Chem C 118:11649–11656. doi:10.1021/jp4126039 25. Chen J, Xin F, Qin S, Yin X (2013) Photocatalytically reducing CO2 to methyl formate in methanol over ZnS and Ni-doped ZnS photocatalysts. Chem Eng J 230:506–512. doi:10. 1016/j.cej.2013.06.119 26. Kočí K, Reli M, Kozák O et al (2011) Influence of reactor geometry on the yield of CO2 photocatalytic reduction. Catal Today 176:212–214. doi:10.1016/j.cattod.2010.12.054 154 3 Semiconductor-Based Photocatalytic Systems …

27. Chaudhary YS, Woolerton TW, Allen CS et al (2012) Visible light-driven CO2 reduction by enzyme coupled CdS nanocrystals. Chem Commun 48:58–60. doi:10.1039/c1cc16107e 28. Fujiwara H, Hosokawa H, Murakoshi K et al (1997) Effect of surface structures on photocatalytic CO2 reduction using quantized CdS nanocrystallites. J Phys Chem B 101:8270–8278 29. Jiang W, Yin X, Xin F et al (2014) Preparation of CdIn2S4 microspheres and application for photocatalytic reduction of carbon dioxide. Appl Surf Sci 288:138–142. doi:10.1016/j. apsusc.2013.09.165 30. Dai W, Xu H, Yu J et al (2015) Photocatalytic reduction of CO2 into methanol and ethanol over conducting polymers modified Bi2WO6 microspheres under visible light. Appl Surf Sci 356:173–180. doi:10.1016/j.apsusc.2015.08.059 31. Sasan K, Lin Q, Mao C, Feng P (2016) Open framework metal chalcogenides as efficient photocatalysts for reduction of CO2 into renewable hydrocarbon fuel. Nanoscale 8:10913– 10916. doi:10.1039/c6nr02525k 32. Liu L, Zhao H, Andino JM, Li Y (2012) Photocatalytic CO2 reduction with H2O on TiO2 nanocrystals: comparison of anatase, rutile, and brookite polymorphs and exploration of surface chemistry. ACS Catal 2:1817–1828. doi:10.1021/cs300273q 33. Dimitrijevic NM, Vijayan BK, Poluektov OG et al (2011) Role of water and carbonates in photocatalytic transformation of CO2 to CH4 on titania. J Am Chem Soc 133:3964–3971. doi:10.1021/ja108791u 34. Lee D, Kanai Y (2012) Role of four-fold coordinated titanium and quantum confinement in CO2 reduction at titania surface. J Am Chem Soc 134:20266–20269. doi:10.1021/ ja309871m 35. Yu J, Low J, Xiao W et al (2014) Enhanced photocatalytic CO2-reduction activity of anatase TiO2 by coexposed 001 and 101 facets. J Am Chem Soc 136:8839–8842. doi:10.1021/ ja5044787 36. Xie S, Wang Y, Zhang Q et al (2014) MgO- and Pt-promoted TiO2 as an efficient photocatalyst for the preferential reduction of carbon dioxide in the presence of water. ACS Catal 4:3644–3653. doi:10.1021/cs500648p 37. Tong Y, Chen L, Ning S et al (2017) Photocatalytic reduction of CO2 to CO over the Ti– Highly dispersed HZSM-5 zeolite containing Fe. Appl Catal B 203:725–730. doi:10.1016/j. apcatb.2016.10.065 38. Hamdy MS, Amrollahi R, Sinev I et al (2014) Strategies to design efficient silica-supported photocatalysts for reduction of CO2. J Am Chem Soc 136:594–597. doi:10.1021/ja410363v 39. Liu E, Kang L, Wu F et al (2014) Photocatalytic reduction of CO2 into methanol over Ag/ TiO2 nanocomposites enhanced by surface plasmon resonance. Plasmonics 9:61–70. doi:10. 1007/s11468-013-9598-7 40. Tu W, Zhou Y, Li H et al (2015) Au@TiO2 yolk–shell hollow spheres for plasmon-induced photocatalytic reduction of CO2 to solar fuel via a local electromagnetic field. Nanoscale 7:14232–14236. doi:10.1039/c5nr02943k 41. Tan LL, Ong WJ, Chai SP et al (2013) Reduced graphene oxide-TiO2 nanocomposite as a promising visible-light-active photocatalyst for the conversion of carbon dioxide. Nanoscale Res Lett 8:465. doi:10.1186/1556-276X-8-465 42. Fu J, Cao S, Yu J et al (2014) Enhanced photocatalytic CO2-reduction activity of electrospun mesoporous TiO2 nanofibers by solvothermal treatment. Dalton Trans 43:9158–9165. doi:10.1039/c4dt00181h 43. Xu Q, Yu J, Zhang J et al (2015) Cubic anatase TiO2 nanocrystals with enhanced photocatalytic CO2 reduction activity. Chem Commun 51:7950–7953. doi:10.1039/ c5cc01087j 44. Fresno F, Jana P, Reñones P et al (2017) CO2 reduction over NaNbO3 and NaTaO3 perovskite photocatalysts. Photochem Photobiol Sci 16:17–23. doi:10.1039/c6pp00235h 45. Li P, Ouyang S, Xi G et al (2012) The effects of crystal structure and electronic structure on photocatalytic H2 evolution and CO2 reduction over two phases of perovskite-structured NaNbO3. J Phys Chem C 116:7621–7628. doi:10.1021/jp210106b References 155

46. Chen X, Zhou Y, Liu Q et al (2012) Ultrathin, single-crystal WO3 nanosheets by two-dimensional oriented attachment toward enhanced photocatalystic reduction of CO2 into hydrocarbon fuels under visible light. ACS Appl Mater Interfaces 4:3372–3377. doi:10. 1021/am300661s 47. Zhou Y, Tian Z, Zhao Z et al (2011) High-yield synthesis of ultrathin and uniform Bi2WO6 square nanoplates benefitting from photocatalytic reduction of CO2 into renewable hydrocarbon fuel under visible light. ACS Appl Mater Interfaces 3:3594–3601. doi:10. 1021/am2008147 48. Yamamoto M, Yoshida T, Yamamoto N et al (2015) Photocatalytic reduction of CO2 with water promoted by Ag clusters in Ag/Ga2O3 photocatalysts. J Mater Chem A 3:16810– 16816. doi:10.1039/C5TA04815J 49. Kawamura S, Puscasu MC, Yoshida Y et al (2015) Tailoring assemblies of plasmonic silver/gold and zinc–gallium layered double hydroxides for photocatalytic conversion of carbon dioxide using UV-visible light. Appl Catal A 504:238–247. doi:10.1016/j.apcata. 2014.12.042 50. Liu Q, Wu D, Zhou Y et al (2014) Single-crystalline, ultrathin ZnGa2O4 nanosheet scaffolds to promote photocatalytic activity in CO2 reduction into methane. ACS Appl Mater Interfaces 6:2356–2361. doi:10.1021/am404572g 51. Liu Q, Zhou Y, Kou J et al (2010) High-yield synthesis of ultralong and ultrathin Zn2GeO4 nanoribbons toward improved photocatalytic reduction of CO2 into renewable hydrocarbon fuel. J Am Chem Soc 132:14385–14387. doi:10.1021/ja1068596 52. Kang U, Park H (2017) A facile synthesis of CuFeO2 and CuO composite photocatalyst films for the production of liquid formate from CO2 and water over a month. J Mater Chem A 5:2123–2131. doi:10.1039/c6ta09378g 53. Zhang T, Wang X, Huang X et al (2016) Bifunctional catalyst of a metallophthalocyanine-carbon nitride hybrid for chemical fixation of CO2 to cyclic carbonate. RSC Adv 6:2810–2818. doi:10.1039/c5ra21058e 54. Zhang L, Wang W, Jiang D et al (2015) Photoreduction of CO2 on BiOCl nanoplates with the assistance of photoinduced oxygen vacancies. Nano Res 8:821–831. doi:10.1007/ s12274-014-0564-2 55. Peng F, Wang J, Ge G et al (2013) Photochemical reduction of CO2 catalyzed by silicon nanocrystals produced by high energy ball milling. Mater Lett 92:65–67. doi:10.1016/j. matlet.2012.10.059 56. Lin J, Pan Z, Wang X (2014) Photochemical reduction of CO2 by graphitic carbon nitride polymers. ACS Sustainable Chem Eng 2:353–358. doi:10.1021/sc4004295 57. Xia P, Zhu B, Yu J et al (2017) Ultra-thin nanosheet assemblies of graphitic carbon nitride for enhanced photocatalytic CO2 reduction. J Mater Chem A 5:3230–3238. doi:10.1039/ c6ta08310b 58. Xu HQ, Hu J, Wang D et al (2015) Visible-light photoreduction of CO2 in a metal–organic framework: boosting electron–hole separation via electron trap states. J Am Chem Soc 137:13440–13443. doi:10.1021/jacs.5b08773 59. Sun D, Gao Y, Fu J et al (2015) Construction of a supported Ru complex on bifunctional MOF-253 for photocatalytic CO2 reduction under visible light. Chem Commun 51:2645– 2648. doi:10.1039/C4CC09797A 60. Yan S, Ouyang S, Xu H et al (2016) Co-ZIF-9/TiO2 nanostructure for superior CO2 photoreduction activity. J Mater Chem A 4:15126–15133. doi:10.1039/C6TA04620G 61. Fei H, Sampson MD, Lee Y et al (2015) Photocatalytic CO2 reduction to formate using a Mn (I) molecular catalyst in a robust metal-organic framework. Inorg Chem 54:6821–6828. doi:10.1021/acs.inorgchem.5b00752 62. Choi KM, Kim D, Rungtaweevoranit B et al (2017) Plasmon-enhanced photocatalytic CO2 conversion within metal-organic frameworks under visible light. J Am Chem Soc 139:356– 362. doi:10.1021/jacs.6b11027 156 3 Semiconductor-Based Photocatalytic Systems …

63. Qin G, Zhang Y, Ke X et al (2013) Photocatalytic reduction of carbon dioxide to formic acid, formaldehyde, and methanol using dye-sensitized TiO2 film. Appl Catal B 129:599– 605. doi:10.1016/j.apcatb.2012.10.012 64. Jia Y, Xu Y, Nie R et al (2017) Artificial photosynthesis of methanol from carbon dioxide and water via a Nile red-embedded TiO2 photocathode. J Mater Chem A 5:5495–5501. doi:10.1039/c6ta10231j 65. Sahara G, Kumagai H, Maeda K et al (2016) Photoelectrochemical reduction of CO2 coupled to water oxidation using a photocathode with a Ru(II)–Re(I) complex photocatalyst and a CoOx/TaON photoanode. J Am Chem Soc 138:14152–14158. doi:10.1021/jacs. 6b09212 66. Kumagai H, Sahara G, Maeda K, Higaghi M, Abe R, Ishitani O (2017) Hybrid photocathode consisting of a CuGaO2 p-type semiconductor and a Ru(II)–Re(I) supramolecular photocatalyst: non-biased visible-light-driven CO2 reduction with water oxidation. Chem Sci 8:4242–4249. doi:10.1039/C7SC00940B 67. Jiang M, Gao Y, Wang Z, Ding Z (2016) Photocatalytic CO2 reduction promoted by a CuCo2O4 cocatalyst with homogeneous and heterogeneous light harvesters. Appl Catal B 198:180–188. doi:10.1016/j.apcatb.2016.05.055 68. Suzuki TM, Nakamura T, Saeki S et al (2012) Visible light-sensitive mesoporous N-doped Ta2O5 spheres: synthesis and photocatalytic activity for hydrogen evolution and CO2 reduction. J Mater Chem 22:24584–24590. doi:10.1039/c2jm33980c 69. Qian W, Wei W, Jianfeng C et al (2012) Novel synthesis of ZnPc/TiO2 composite particles and carbon dioxide photo-catalytic reduction efficiency study under simulated solar radiation conditions. Colloid Surfaces A 409:118–125. doi:10.1016/j.colsurfa.2012.06.010 70. Finkelstein-Shapiro D, Petrosko SH, Dimitrijevic NM et al (2013) CO2 preactivation in photoinduced reduction via surface functionalization of TiO2 nanoparticles. J Phys Chem Lett 4:475–479. doi:10.1021/jz3020327 71. Zhao G, Pang H, Liu G et al (2017) Co-porphyrin/carbon nitride hybrids for improved photocatalytic CO2 reduction under visible light. Appl Catal B 200:141–149. doi:10.1016/j. apcatb.2016.06.074 72. Walsh JJ, Jiang C, Tang J, Cowan AJ (2016) Photochemical CO2 reduction using structurally controlled g-C3N4. Phys Chem Chem Phys 18:24825–24829. doi:10.1039/ c6cp04525a 73. Kuriki R, Ishitani O, Maeda K (2016) Unique solvent effects on visible-light CO2 reduction over ruthenium(II)-complex/carbon nitride hybrid photocatalysts. ACS Appl Mater Interfaces 8:6011–6018. doi:10.1021/acsami.5b11836 74. Sato S, Arai T, Morikawa T et al (2011) Selective CO2 conversion to formate conjugated with H2O oxidation utilizing semiconductor/complex hybrid photocatalysts. J Am Chem Soc 133:15240–15243. doi:10.1021/ja204881d 75. Yan Y, Yu Y, Huang S et al (2017) Adjustment and matching of energy band of TiO2-based photocatalysts by metal ions (Pd, Cu, Mn) for photoreduction of CO2 into CH4. J Phys Chem C 121:1089–1098. doi:10.1021/acs.jpcc.6b07180 76. Tahir M, Amin NS (2013) Photocatalytic CO2 reduction and kinetic study over In/TiO2 nanoparticles supported microchannel monolith photoreactor. Appl Catal A 467:483–496. doi:10.1016/j.apcata.2013.07.056 77. Tahir M, Amin NS (2015) Indium-doped TiO2 nanoparticles for photocatalytic CO2 reduction with H2O vapors to CH4. Appl Catal B 162:98–109. doi:10.1016/j.apcatb.2014.06. 037 78. Ola O, Maroto-Valer MM (2014) Role of catalyst carriers in CO2 photoreduction over nanocrystalline nickel loaded TiO2-based photocatalysts. J Catal 309:300–308. doi:10.1016/ j.jcat.2013.10.016 79. Zhao C, Liu L, Zhang Q et al (2012) Photocatalytic conversion of CO2 and H2O to fuels by nanostructured Ce–TiO2/SBA-15 composites. Catal Sci Technol 2:2558–2568. doi:10.1039/ c2cy20346d References 157

80. Yang HC, Lin HY, Chien YS et al (2009) Mesoporous TiO2/SBA-15, and Cu/TiO2/SBA-15 composite photocatalysts for photoreduction of CO2 to methanol. Catal Lett 131:381–387. doi:10.1007/s10562-009-0076-y 81. Matějová L, Kočíb K, Reli, et al (2014) Preparation, characterization and photocatalytic properties of cerium doped TiO2: on the effect of Ce loading on the photocatalytic reduction of carbon dioxide. Appl Catal B 152–153:172–183. doi:10.1016/j.apcatb.2014.01.015 82. Wang Y, Xu Y, Wang Y et al (2016) Synthesis of Mo-doped graphitic carbon nitride catalysts and their photocatalytic activity in the reduction of CO2 with H2O. Catal Commun 74:75–79. doi:10.1016/j.catcom.2015.10.029 83. Ong WJ, Tan LL, Chai SP et al (2014) Self-assembly of nitrogen-doped TiO2 with exposed 001 facets on a graphene scaffold as photo-active hybrid nanostructures for reduction of carbon dioxide to methane. Nano Res 7:1528–1547. doi:10.1007/s12274-014-0514-z 84. Tan LL, Ong WJ, Chai SP, Mohamed AR (2014) Band gap engineered, oxygen-rich TiO2 for visible light induced photocatalytic reduction of CO2. Chem Commun 50:6923–6926. doi:10.1039/C4CC01304B 85. Lu D, Zhang M, Zhang Z et al (2014) Self-organized vanadium and nitrogen co-doped titania nanotube arrays with enhanced photocatalytic reduction of CO2 into CH4. Nanoscale Res Lett 9:272. doi:10.1186/1556-276X-9-272 86. Liu S, Wang J, Yu J (2016) ZIF-8 derived bimodal carbon modified ZnO photocatalysts with enhanced photocatalytic CO2 reduction performance. RSC Adv 6:59998–60006. doi:10. 1039/C6RA11264A 87. Xiong Z, Lei Z, Kuang CC et al (2017) Selective photocatalytic reduction of CO2 into CH4 over Pt-Cu2O TiO2 nanocrystals: the interaction between Pt and Cu2O cocatalysts. Appl Catal B 202:695–703. doi:10.1016/j.apcatb.2016.10.001 88. Chang X, Wang T, Zhang P et al (2016) Stable aqueous photoelectrochemical CO2 reduction by a Cu2O dark cathode with improved selectivity for carbonaceous products. Angew Chem Int Ed 55:8840–8845. doi:10.1002/anie.201602973 89. Wang Y, Li B, Zhang C et al (2013) Ordered mesoporous CeO2-TiO2 composites: highly efficient photocatalysts for the reduction of CO2 with H2O under simulated solar irradiation. Appl Catal B 130–131:277–284. doi:10.1016/j.apcatb.2012.11.019 90. Zhao J, Wang Y, Li Y et al (2016) Phase-dependent enhancement for CO2 photocatalytic reduction over CeO2/TiO2 catalysts. Catal Sci Technol 6:7967–7975. doi:10.1039/ C6CY01365A 91. Jiao J, Wei Y, Zhao Z et al (2014) Photocatalysts of 3D ordered macroporous TiO2- supported CeO2 nanolayers: design, preparation, and their catalytic performances for the reduction of CO2 with H2O under simulated solar irradiation. Ind Eng Chem Res 53:17345– 17354. doi:10.1021/ie503333b 92. Reli M, Huo P, Šihor M et al (2016) Novel TiO2/C3N4 photocatalysts for photocatalytic reduction of CO2 and for photocatalytic decomposition of N2O. J Phys Chem A 120:8564– 8573. doi:10.1021/acs.jpca.6b07236 93. Shi H, Zhang C, Zhou C, Chen G (2015) Conversion of CO2 into renewable fuel over Pt– g-C3N4/KNbO3 composite photocatalyst. RSC Adv 5:93615–93622. doi:10.1039/ c5ra16870h 94. Shi H, Chen G, Zhang C, Zou Z (2014) Polymeric g-C3N4 coupled with NaNbO3 nanowires toward enhanced photocatalytic reduction of CO2 into renewable fuel. ACS Catal 4:3637– 3643. doi:10.1021/cs500848f 95. Zhang X, Wang L, Du Q et al (2016) Photocatalytic CO2 reduction over B4C/C3N4 with internal electric field under visible light irradiation. J Colloid Interface Sci 464:89–95. doi:10.1016/j.jcis.2015.11.022 96. Cao SW, Liu XF, Yuan YP et al (2014) Solar-to-fuels conversion over In2O3/g-C3N4 hybrid photocatalysts. Appl Catal B 147:940–946. doi:10.1016/j.apcatb.2013.10.029 97. Li M, Zhang L, Fan X et al (2017) Core-shell LaPO4/g-C3N4 nanowires for highly active and selective CO2 reduction. Appl Catal B 201:629–635. doi:10.1016/j.apcatb.2016.09.004 158 3 Semiconductor-Based Photocatalytic Systems …

98. Li K, Peng B, Jin J et al (2017) Carbon nitride nanodots decorated brookite TiO2 quasi nanocubes for enhanced activity and selectivity of visible-light-driven CO2 reduction. Appl Catal B 203:910–916. doi:10.1016/j.apcatb.2016.11.001 99. Zou JP, Wu DD, Luo J et al (2016) A strategy for one-pot conversion of organic pollutants into useful hydrocarbons through coupling photodegradation of MB with photoreduction of CO2. ACS Catal 6:6861–6867. doi:10.1021/acscatal.6b01729 100. Yang Z, Wang H, Song W et al (2017) One dimensional SnO2 NRs/Fe2O3 NTs with dual synergistic effects for photoelectrocatalytic reduction CO2 into methanol. J Colloid Interface Sci 486:232–240. doi:10.1016/j.jcis.2016.09.055 101. Sakimoto KK, Zhang SJ, Yang P (2016) Cysteine–cystine photoregeneration for oxygenic photosynthesis of acetic acid from CO2 by a tandem inorganic–biological hybrid system. Nano Lett 16:5883–5887. doi:10.1021/acs.nanolett.6b02740 102. Li X, Liu H, Luo D et al (2012) Adsorption of CO2 on heterostructure CdS(Bi2S3)/TiO2 nanotube photocatalysts and their photocatalytic activities in the reduction of CO2 to methanol under visible light irradiation. Chem Eng J 180:151–158. doi:10.1016/j.cej.2011. 11.029 103. Wei W, Yang Z, Song W et al (2017) Different CdSeTe structure determined photoelectrocatalytic reduction performance for carbon dioxide. J Colloid Interface Sci 496:327–333. doi:10.1016/j.jcis.2016.11.054 104. Wang C, Thompson RL, Baltrus J, Matranga C (2010) Visible light photoreduction of CO2 using CdSe/Pt/TiO2 heterostructured catalysts. J Phys Chem Lett 1:48–53. doi:10.1021/ jz9000032 105. Su Y, Zhang Z, Liu H, Wang Y (2017) Cd0.2Zn0.8S@UiO-66-NH2 nanocomposites as efficient and stable visible-light-driven photocatalyst for H2 evolution and CO2 reduction. Appl Catal B 200:448–457. doi:10.1016/j.apcatb.2016.07.032 106. Cherkasov N, Ibhadon AO, Fitzpatrick P (2015) A review of the existing and alternative methods for greener nitrogen fixation. Chem Eng Proc 90:24–33. doi:10.1016/j.cep.2015.02. 004 107. Schrautzer GN, Strampach N, Hui LN et al (1983) Nitrogen photoreduction on desert sands under sterile conditions. Proc Natl Acad Sci USA 80:3873–3876 108. Ma H, Shi Z, Li Q, Liu N (2016) Large-scale production of graphitic carbon nitride with outstanding nitrogen photofixation ability via a convenient microwave treatment. Appl Surf Sci 379:309–315. doi:10.1016/j.apsusc.2016.04.085 109. Ma H, Shi Z, Li Q, Li S (2016) Preparation of graphitic carbon nitride with large specific surface area and outstanding N2 photofixation ability via a dissolve-regrowth process. J Phys Chem Sol 99:51–58. doi:10.1016/j.jpcs.2016.08.008 110. Dong G, Ho W, Wang C (2015) Selective photocatalytic N2 fixation dependent on g-C3N4 induced by nitrogen vacancies. J Mater Chem A 3:23435–23441. doi:10.1039/c5ta06540b 111. Raevskaya AE, Panasiuk YV, Korzhak GV et al (2017) Photocatalytic H2 production from aqueous solutions of hydrazine and its derivatives in the presence of nitric-acid-activated graphitic carbon nitride. Catal Today 284:229–235. doi:10.1016/j.cattod.2016.12.024 112. Wu G, Gao Y, Zheng B (2016) Template-free method for synthesizing sponge-like graphitic carbon nitride with a large surface area and outstanding nitrogen photofixation ability induced by nitrogen vacancies. Ceram Internat 42:6985–6992. doi:10.1016/j.ceramint.2016. 01.086 113. Hu S, Chen X, Li Q et al (2016) Effect of sulfur vacancies on the nitrogen photofixation performance of ternary metal sulfide photocatalysts. Catal Sci Technol 6:5884–5890. doi:10. 1039/c6cy00622a 114. Hu S, Li Y, Li F et al (2016) Construction of g-C3N4/Zn0.11Sn0.12Cd0.88S1.12 hybrid heterojunction catalyst with outstanding nitrogen photofixation performance induced by sulfur vacancies. ACS Sustainable Chem Eng 4:2269–2278. doi:10.1021/acssuschemeng. 5b01742 References 159

115. Zhang Q, Hu S, Fan Z et al (2016) Preparation of g-C3N4/ZnMoCdS hybrid heterojunction catalyst with outstanding nitrogen photofixation performance under visible light via hydrothermal post-treatment. Dalton Trans 45:3497–3505. doi:10.1039/c5dt04901f 116. Cao Y, Hu S, Li F et al (2016) Photofixation of atmospheric nitrogen to ammonia with a novel ternary metal sulfide catalyst under visible light. RSC Adv 6:49862–49867. doi:10. 1039/c6ra08247e 117. Banerjee A, Yuhas BD, Margulies EA et al (2015) Photochemical nitrogen conversion to ammonia in ambient conditions with FeMoS-chalcogels. J Am Chem Soc 137:2030–2034. doi:10.1021/ja512491v 118. Li H, Shang J, Shi J et al (2016) Facet-dependent solar ammonia synthesis of BiOCl nanosheets via a proton-assisted electron transfer pathway. Nanoscale 8:1986–1993. doi:10. 1039/c5nr07380d 119. Li H, Shang J, Ai Z, Zhang L (2015) Efficient visible light nitrogen fixation with BiOBr nanosheets of oxygen vacancies on the exposed 001 facets. J Am Chem Soc 137:6393–6399. doi:10.1021/jacs.5b03105 120. Sun S, An Q, Wang W et al (2017) Efficient photocatalytic reduction of dinitrogen to ammonia on bismuth monoxide quantum dots. J Mater Chem A 5:201–209. doi:10.1039/ c6ta09275f 121. Taqui Khan MM, Rao NN (1991) Stepwise reduction of coordinated dinitrogen to ammonia via diazinido and hydrazido intermediates on a visible light irradiated Pt /CdS Â Ag2S/RuO2 particulate system suspended in an aqueous solution of K[Ru(EDTA-H)Cl]2H2O. J Photochem Photobiol A 56:101–111 122. Hu S, Zhang W, Bai J et al (2016) Construction of a 2D/2D g-C3N4/rGO hybrid heterojunc– tion catalyst with outstanding charge separation ability and nitrogen photofixation performance via a surface protonation process. RSC Adv 6:25695–25702. doi:10.1039/ c5ra28123g 123. Ogawa T, Kitamura T, Shibuya T, Hoshino K (2004) Characterization and material conditions of conducting polymer/titanium oxide hybrid systems used for dinitrogen fixation under ordinary pressure and temperature. Electrochem Commun 6:55–60. doi:10.1016/j. elecom.2003.10.015 124. Zhao W, Zhang J, Zhu X et al (2014) Enhanced nitrogen photofixation on Fe-doped TiO2 with highly exposed (1 0 1) facets in the presence of ethanol as scavenger. Appl Catal B 144:468–477. doi:10.1016/j.apcatb.2013.07.047 3+ 125. Hu S, Chen X, Li Q et al (2017) Fe doping promoted N2 photofixation ability of honeycombed graphitic carbon nitride: the experimental and density functional theory simulation analysis. Appl Catal B 201:58–69. doi:10.1016/j.apcatb.2016.08.002 126. Hu S, Qu X, Bai J, Li P, Li Q, Wang F, ong L (2017) Effect of Cu(I)–N active sites on the N2 photofixation ability over flowerlike copper-doped g-C3N4 prepared via a novel molten salt-assisted microwave process: the experimental and density functional theory simulation analysis. ACS Sustainable Chem Eng 5:6863–6872. doi:10.1021/acssuschemeng.7b01089 127. Yue C, Trudeau ML, Antonelli D (2005) Mesoporous tantalum oxide photocatalysts for Schrauzer-type conversion of dinitrogen to ammonia. Canadian J Chem 83:308–314. doi:10. 1139/v05-018 128. Lashgaria M, Zeinalkhania P (2017) Photocatalytic N2 conversion to ammonia using efficient nanostructured solar-energy-materials in aqueous media: a novel hydrogenation strategy and basic understanding of the phenomenon. Appl Catal A 529:91–97. doi:10.1016/ j.apcata.2016.10.017 129. Linnik O, Kisch H (2006) On the mechanism of nitrogen photofixation at nanostructured iron titanate films. Photochem Photobiol Sci 5:938–942. doi:10.1039/b608396j 130. Tennakone K, Ileperuma OA, Bandara JMS et al (1991) Simultaneous reductive and oxidative photocatalytic nitrogen fixation in hydrous iron(iii) oxide loaded nafion films in aerated water. J Chem Soc Chem Commun 579–580 131. Ranjit KT, Varadarajan TK, Viswanathan B (1996) Photocatalytic reduction of dinitrogen to ammonia over noble-metal-loaded TiO2. J Photochem Photobiol, A 96:181–185 160 3 Semiconductor-Based Photocatalytic Systems …

132. Ali M, Zhou F, Chen K et al (2016) Nanostructured photoelectrochemical solar cell for nitrogen reduction using plasmon-enhanced black silicon. Nat Commun 7:11335. doi:10. 1038/ncomms11335 133. Sun S, Li X, Wang W et al (2017) Photocatalytic robust solar energy reduction of dinitrogen to ammonia on ultrathin MoS2. Appl Catal B 200:323–329. doi:10.1016/j.apcatb.2016.07.025 134. Wang W, Xu D, Cheng B et al (2017) Hybrid carbon@TiO2 hollow spheres with enhanced photocatalytic CO2 reduction activity. J Mater Chem A 5:5020–5029. doi:10.1039/ c6ta11121a Chapter 4 Semiconductor-Based Liquid-Junction Photoelectrochemical Solar Cells

4.1 Principles and Designs of Semiconductor NP-Sensitized Solar Cells

The photoelectrochemical light-harvesting systems constitute an important part of the assay of available solar light conversion approaches, along with the photovoltaic light conversion and endothermal photochemical reactions such as the hydrogen production, CO2 reduction, etc. [1–12]. Today, the realm of semiconductor-based solar cells is dominated (up to 85%) by “classic” photovoltaic systems based on single-crystal and polycrystalline silicon with a light conversion efficiency reaching 14–19 and 8–10%, respectively [2, 4, 12]. At the same time, a high price of the single-crystalline Si stimulates a search for alternative technologies based on more available materials, such as amorphous silicon, thin-film CdTe-based heterostructures [2, 12], organic conjugated polymers [2, 13–15], liquid-junction solar cells [2, 4–11], etc. The photovoltaic light-harvesting systems are contingently categorized into the first-generation, second-generation, and third-generation solar cells depending on the operating principles [4, 8]. The first-generation group encompasses “classic” Si-based cells [4, 12] with a p/n junction responsible for the separation of the photogenerated charge carriers. Due to fundamental reasons, such as an indirect nature of electron transitions, the fabrication of inexpensive thin-film Si-based solar cells that can be implemented on the broadest scale, is impossible. This problem is solved partially in the second-generation solar cells based on semiconductor thin films coupled to an optically transparent electrode (OTE) and a counter electrode. The most vivid example of the second-generation solar cell is an “n-CdS/p-CdTe” system [2, 12]. Today, the second-generation cells occupy around 15% of the solar cell market. The high light conversion efficiencies can be achieved in such systems by combining several p/n junctions, however, the multi-layer cells are expensive and can be rationally used in specific applications, for example, in the aerospace field [2]. Also, the production of thin-film solar cells requires

© Springer International Publishing AG 2018 161 O. Stroyuk, Solar Light Harvesting with Nanocrystalline Semiconductors, Lecture Notes in Chemistry 99, https://doi.org/10.1007/978-3-319-68879-4_4 162 4 Semiconductor-Based Liquid-Junction Photoelectrochemical … high-precision technologies and corresponding equipment, for example, the gas-phase or molecular-beam epitaxy, the magnetron sputtering, etc. [2, 7, 12]. The third-generation solar light harvesters include solar cells with nanocrystalline semiconductor electrodes, in particular, the nanostructured metal oxides, such as TiO2, ZnO, and SnO2 attached to the surface of various OTEs [4, 6]. The metal oxide semiconductors have relatively large band gaps (Eg > 3 eV) and thus they are capable of harvesting only a small portion of the solar light. To cover the visible spectral range the metal oxide electrodes should be sensitized by various compounds that can efficiently absorb visible (and near-IR) light and transfer the photogenerated charge carriers to a wide-bandgap component [7, 8]. Typically the metal oxides can be sensitized by organic dyes and strongly-absorbing metal complexes in the so-called dye-sensitized solar cells (DSSCs) or, alternatively, by NPs of a narrow-bandgap semiconductor capable of the visible light absorption—in the semiconductor-sensitized solar cells (SSSCs). Both types of cells are composed of a light-harvesting photoanode, a counter electrode and liquid electrolyte containing a redox couple capable of the electron donation to the photoanode and recovering its original state on the counter electrode by accepting an electron. Alternatively, an acceptor can be reduced on a photoexcited photocathode and then oxidized on a counter electrode (CE), thus completing the light-harvesting cycle. The cells of both types are typically referred to as “the liquid-junction solar cells”. Primary photoinduced processes in such solar cells include the photoinduced charge separation and the oxidation/reduction of a dissolved substrate, resembling the photocatalytic semiconductor-driven processes. As a result, the range of substances used both as the wide-bandgap matrices (scaffolds), the dyes and complexes used as sensitizers and the narrow-bandgap NPs are typical and similar both for the photocatalytic reactions and for the photoelectrochemical (PEC) third-generation liquid-junction solar cells. The energy criteria used for the selection of appropriate components of a PEC system are also very similar to those applied in the semiconductor-based photocatalytic systems and require an energy correspondence between the CB level of a narrow-bandgap sensitizer (or a LUMO level of the photoexcited dye-sensitizer), the CB level of a wide-bandgap oxide scaffold, and the redox level of the electron “shuttling” couple present in the electrolyte (Fig. 4.1a). Alternatively, the nanocrystalline semiconductors can be coupled with other materials, such as the conjugated polymers, fullerene derivatives, organic oligo-dyes, organic-inorganic perovskites, etc., forming solid p/n-junctions where the photoinduced separation of electron and hole can occur, similarly to the second-generation thin-film solar cells. Such cells are referred to as “the bulk heterojunction solar cells” [8]. The highest reported light-conversion efficiency for the bulk-heterojunction solar cells is around 11% [13]. As we focus in this book on the photochemical light-harvesting, that is, the processes involving chemical transformations of the participants, we will focus predominantly on the liquid-junction SSSCs, where the solar light energy conversion occurs as a result of concerted photochemical/chemical transformations of the cell components. 4.1 Principles and Designs of Semiconductor … 163

Fig. 4.1 a Energy diagram illustrating the band positions of some semiconductor materials typically used in the third-generation solar cells relative to the redox levels of some popular electron-shuttling redox-couples; b A working principle of the dye-sensitized liquid-junction solar cell with a titania-based photoanode. Reprinted with permissions from Ref. [11]. Copyright (2001) Nature Publishing Group 164 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

The DSSC action is based on the photoinduced electron transfer from a photoexcited dye-sensitizer into the conduction band (or some mid-bandgap surface-related states) of the wide-band-gap metal oxide (TiO2, ZnO, SnO2) and then—to the electric circuit (Fig. 4.1a) [11, 16]. The one-electron oxidized dye-sensitizer is reduced by a component of the redox couple present in solution. À À Typically, the iodide/iodate I IO3 couple is used in the DSSCs. In this case, the À dye recovers its original state by oxidizing iodide to iodate and then IO3 ions are reduced to I− again on a counter electrode (typically Pt, Au, Ag, etc.), thus finishing the PEC cycle. The highest light conversion efficiency, 12%, was achieved for the mesoporous titania scaffolds sensitized by RuL(NCS)3 complexes, where L is a bipyridyl-based ligand [16]. In SSSCs the solar light is harvested by NPs of a narrow-bandgap semiconductor (the term “narrow” here is relative, it is used rather to distinguish such visible-light-sensitive NPs from the wide-bandgap metal oxide materials), for example, CdS, CdSe, CuInS2, or InP [4–9, 17, 18]. The sensitizer NPs absorb visible light quanta with the energy higher than the bandgap resulting in an electron coming from VB to CB (Fig. 4.2a). Then, the CB electron is transferred from the sensitizer to the metal oxide scaffold and then—to the electric circuit and, finally, to the counter electrode. The transfer is only possible if the CB level of the sensitizer NPs is higher than the CB of the wide-bandgap component (Fig. 4.2b). A photogenerated VB hole of the sensitizer NP is filled with an electron from a redox couple component (in this case from sulfide ions) producing an oxidized form of the shuttle (elemental sulfur). The shuttle is then regenerated on the counter-electrode accepting an electron and finishing the PEC cycle. The hole transfer from the sensitizer to the scaffold VB is impossible, as the VB level of the metal oxide resides deeper than the sensitizer VB level (Fig. 4.2b). In this way, the photogenerated electron and hole are reliably separated between the photoanode components [8, 18].

Fig. 4.2 Schemes illustrating a action principle of a liquid-junction solar cell sensitized by visible-light-sensitive semiconductor NPs and b energy diagram of a cell comprising TiO2, CdS, S2−/S0 electron-shuttling couple and a counter electrode 4.1 Principles and Designs of Semiconductor … 165

The SSSCs have a number of advantages over DSSCs, in particular, (i) a broad variability of the electron parameters (Eg, CB and VB level positions) as a result of size and composition variations of the sensitizer NPs [4–9, 17, 18]; (ii) the possibility of multi-exciton generation in some narrow-bandgap NP materials, in particular PbS, PbSe, PbTe, CdSe, InAs, InP, CdTe, and Si [10]; (iii) a more robust electron contact between the scaffold and the sensitizer NPs as compared to molecular sensitizers, that are typically adsorbed via the surface bridge OH groups [5]. Efficient spatial separation of the photogenerated electrons and holes is the most important condition to be met by a binary semiconductor heterostructure to be an efficient SSSC photoanode. For this, the heterostructure should have the mutual CB and VB positions similar to those depicted in Fig. 4.2b[4–9]. In real systems, even if the basic energy condition is met, concurrent recombination processes always took place, resulting in a loss of the light conversion efficiency. The losses are accounted for by the electron-hole recombination in the sensitizer NPs preceding the electron transfer, by the charge capturing in the surface traps of sensitizer NPs, by the recombination of the injected electron with components of the electrolyte, by side photocatalytic reactions, etc. The liquid-junction SSSCs provide also a number of advantages over the bulk heterojunction solar cells, where a donor and an acceptor contact directly. In particular, the liquid-junction SSSCs do not suffer from the “non-ideality” of the heterojunction due to the presence of a liquid electrolyte that envelops the entire surface of the electrodes and ensures a good electric contact between the photoanode and the counter-electrode. Also, the liquid-junction SSSCs can be produced in a relatively simple way without any high-precision equipment or elaborate/unique laboratory techniques, such as the high vacuum, ultra-clean environment, ultra-high-pure semiconductors, etc. At the same time, the liquid-junction SSSCs suffer from a relatively low chemical stability [4]. The most widely used wide-bandgap scaffold is mesoporous/nanocrystalline titania that is characterized by a chemical stability, a high electron mobility and low recombinational losses [3–9, 18, 19]. Another popular metal oxide scaffold is zinc oxide, that exhibits a number of quite unique properties including spectacular quantum size effects, the capability of photoinduced charge accumulation, a high photoactivity and relative simplicity of preparation. However, ZnO is chemically unstable as compared to TiO2 and suffers from degradation in acidic/basic solutions as well as in the presence of sulfide/polysulfide electrolyte. At the same time, this instability can be exploited to modify ZnO scaffolds or to convert them in binary heterostructures, as will be shown in this chapter later. The available syntheses of nanocrystalline ZnO also provide a virtually unlimited variety of morphologies and geometries—from single-crystalline nanorods (NRs) to intricate ordered 3D structures [19, 20]. The most popular narrow-bandgap sensitizers for the liquid-junction SSSCs are metal sulfide NPs (CdS, PbS, CuInS2, AgInS2), metal selenide NPs (CdSe, CdSxSe1−x, CdSexTe1−x, PbSe), and binary metal chalcogenide nanocomposites (CdS/PbS, CdS/CdSe, CdS/CuInS2, etc.). Starting from 2009 the feasibility of using 166 4 Semiconductor-Based Liquid-Junction Photoelectrochemical … organic-inorganic Pb-based perovskites CH3NH3PbHal3 (Hal—Br, I) as spectral sensitizers for the liquid-junction solar cells was shown [21]. Today, the perovskite solar cells form an independent and rapidly developing branch of the photovoltaics reaching the light conversion efficiencies of around 18% [22–26]. However, the perovskite-based materials also suffer from chemical/photochemical instability and contain toxic lead making the solar cells recycling a challenge still to be properly met. In the SSSCs based on the narrow-bandgap metal chalcogenide NPs the highest fi fi fi ef ciencies of light conversions were achieved with aqueous sul de/polysul de – 2À 2À electrolytes [4 9, 18]. The S Sx electron-shuttling couple ensures high photocurrents and photovoltages in such SSSCs and simultaneously inhibits the 2À 2À undesirable photocorrosion of the sensitizer NPs. Along with S Sx ., other redox À À couples are constantly probed, including IÀ IÀ,FeCNðÞ2 FeðÞ CN 3 , . 3 6 6 ð ðÞ2 þ ð ðÞ3 þ Co o - phen 3 Co o - phen 3 , etc. [4]. The alternative redox couples are typically introduced when the sulfide/polysulfide shuttle is impossible to use, for example, in the case of CH3NH3PbHal3 perovskite or Sb2S3, which dissolves in the presence of S2− ions [21, 27]. The counter electrodes for the liquid-junction SSSCs are typically selected for a particular redox couple, because a CE should be catalytically active with respect to the reduction/oxidation of the electron-shuttling species [28]. The Pt-based counter À À electrodes are used for the electrolytes with I I3 redox-couple, however, in the polysulfide media platinum is rapidly deactivated as a result of poisoning [4, 18, 28]. The highest electrocatalytic activities with respect to the polysulfide electrolytes of SSSCs were found for a number of transition metal sulfides, in particular, CoS, CuxS, PbS, and NiS, attached to conductive substrates [28, 29]. Typically, 2− such materials are stable in the presence of S and reveal a high electrocatalytic 2À 2À activity toward oxidation/reduction of the S Sx shuttle. A search for new and more efficient CE materials is constantly performed [30, 31] biasing to more complex structures [32, 33], for example, the composites of metal sulfides with graphene derivatives [34, 35].

4.2 Basic Photoelectrochemical Characteristics of SSSCs

The illumination of a SSSC with the light corresponding to the absorption band of sensitizer NPs results in the photocurrent generation. The photocurrent density (the current per a surface area) is limited by the radiative recombination in the sensitizer NPs and several types of non-radiative recombination processes involving the sensitizer and metal oxide NPs and the electrolyte components [6, 9]. The photogenerated valence band holes in the sensitizer NPs are ﬁlled with electrons from a donating component of the shuttling couple—sulﬁde anions: 4.2 Basic Photoelectrochemical Characteristics of SSSCs 167

S2À +2hþ ! S ð4:1Þ

2À ! 2ÀðÞ... ð : Þ S+SxÀ1 Sx x=1 7 4 2

The shuttling couple is then regenerated on the surface of a counter electrode:

2À À ! 2À 2À ð : Þ Sx +2e SxÀ1 +S 4 3

The electron migration through the electric circuit that connects a photoanode and a counter electrode results in the current characterized typically by the short-circuit photocurrent density, Jsc, measured at a zero voltage. The open-circuit photovoltage, Voc, is the second important characteristic of SSSCs measured at J = 0 and corresponding to a difference between the Fermi level of the photoanode, EF, and the redox-potential of the shuttling couple in the electrolyte E(Red/Ox) [7, 18]:

Voc ¼ EFÀEðÞRed/Ox : ð4:4Þ

The Fermi level of the photoanode resides between the work function of the conductive OTE and the CB potential of the wide-bandgap oxide scaffold [4, 7, 18]. Typically, the SSSCs are illuminated with a solar light simulator emitting the so-called AM1.5 light flux. The AM1.5 flux is characterized by an intensity of 100 mW/cm2 and has a spectral distribution very similar to that of the solar irradiation near the Earth surface [4–11], but xenon and mercury high-pressure lamps are also used for the SSSC characterizations similarly to the photocatalytic light-harvesting systems. Figure 4.3a illustrates the solar irradiation spectra just outside the Earth atmosphere (AM0) and near the Earth surface (AM1.5). The “wells” in the AM1.5 spectrum are associated with the selective absorption of some wavelengths by the atmosphere gases (oxygen, water, CO2)[36]. The figure shows also a spectrum of a black body heated to 5800 K which is an ideal irradiation spectrum for a solar simulator. The basic parameter of a SSSC is a total light power conversion efficiency η [7, 8]:

g ¼ðJsc Â Voc Â FFÞ=Pin; ð4:5Þ where FF is the fill factor of the voltage-current characteristics, Pin is the incoming light flux intensity, mW/cm2. All parameters necessary for the calculation of η can be determined from the voltage-current curve for a given SSSC (Fig. 4.3b) [7, 18]. The fill factor can be calculated as

FF = Pactual=Ptheoretical; ð4:6Þ where Ptheoretical =Jsc Â Voc is a theoretically highest possible power for the given SSSC, while Pactual is the experimentally measured cell power that can be determined as a maximum on the dependence of the cell power on the applied voltage (Fig. 4.3c). 168 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

Fig. 4.3 a Solar spectrum outside the Earth atmosphere (AM0) and near the Earth surface (AM1.5) and a black body irradiation at 5800 K (the ideal spectrum for a solar simulator), duE/dk is a normalized light ﬂux [36]. b, c Characteristics of a SSSC that can be derived from the voltage– photocurrent (b) and power—voltage (c) dependences. (d) IPCE spectra of a TiO2/CdSe photoanode and a ternary TiO2/CdSe/ZnS heterostructure. Reproduced with permissions from Refs. [36](a) and [18](b–d). Copyright (2011, 2012) The Royal Society of Chemistry

A ratio of the number of photogenerated electrons to the number of absorbed photons is referred to as a photocurrent quantum yield c or incident-photon-to-current-efﬁciency, IPCE [7, 18]:

c ¼ 1241 Â Jsc=ðÞÂk Â Pactual 100%; ð4:7Þ

−2 −2 where Jsc is presented in A Â cm ,Pactual—in W Â cm , k—in nanometers. The value of 1241 is a combination of fundamental constants (h Â c Â 109/e), where h is the Planck constant (6.62 Â 10−34 J Â s−1), c is the light velocity in a vacuum (3 Â 108 m Â s−1), e is the electron charge (1.602 Â 10−19 C). A dependence of IPCE on the excitation wavelength is referred to as the photocurrent quantum yield spectrum (IPCE spectrum or photo-action spectrum). The IPCE spectrum typically coincides with the absorption spectrum of the solar cell photoanode (or the photocathode) [6, 7, 18] and supplies information on the spectral sensitivity range and the efficiency of a photo-electrode. As an example, IPCE spectra of a TiO2/CdSe photoanode are presented in Fig. 4.3d showing that 4.2 Basic Photoelectrochemical Characteristics of SSSCs 169 such heterostructure can harvest the solar light up to *700 nm with the efficiency that increases substantially, by around 50%, when a protective ZnS shell is deposited onto the surface of CdSe NPs [18]. The ways of the modification of wide-bandgap metal oxide scaffolds with narrow-bandgap sensitizer NPs are quite versatile and affect considerably the light conversion efficiency of the final SSSCs. The sensitizer NPs can be produced separately (ex situ) and adsorbed/deposited from a colloidal solution onto the scaffold surface. Typically the oxide surface is preliminary functionalized by a “bridge” bifunctional molecule that can interact simultaneously with the hydroxy- lated oxide surface and with sensitizer NPs (Fig. 4.4)[7, 18]. In particular, metal oxide surfaces can be modified with mercapto-carboxylic acids HOOC–R–SH (mercaptoacetic, mercaptopropionic, etc.) simply by immers- ing the photoanode into an aqueous acid solution. The carboxylic group is attached to the oxide surface via hydrogen bonding between –COOH and surface –COH groups while –SH groups can efficiency interact with the undercoordinated metal ions on the surface of metal chalcogenide sensitizer NPs, such as CdTe, CdSe, CdS, PbS, PbSe, etc. Alternatively, the ex situ synthesized sensitizer NPs can be deposited onto oxide surfaces by the electrophoretic deposition [7, 18, 37]. Chemical bath deposition (CBD) is performed by the immersion of a metal oxide film into a hot solution containing metal and chalcogenide precursors that form sensitizer NPs during the slow decomposition. Hydrolytically unstable compounds are typically used as a chalcogenide source, such as thiourea or thioacetamide (release of S2−) or sodium selenosulfate (release of Se2−). Slow chalcogenide release allows for the uniform nucleation and controlled growth of the sensitizer NPs. Successive ionic layer adsorption and reaction (SILAR) is also broadly used for the preparation of chalcogenide/oxide heterostructures. In this method, a metal oxide film is immersed consecutively into a solution containing metal ions and into

Fig. 4.4 Scheme illustrating the deposition of ex-situ synthesized sensitizer NPs onto the surface of oxide scaffold pre-modified with mercaptopropionic acid. Yellow and red circles correspond to the metal oxide and metal chalcogenide sensitizer NPs, respectively. Reprinted with permissions from Ref. [7]. Copyright (2010) American Chemical Society 170 4 Semiconductor-Based Liquid-Junction Photoelectrochemical … a solution containing chalcogenide ions. As a result of multiple repetitions of such procedure, the content and size of semiconductor NPs can be increased in a controlled manner. Despite the seemingly trivial character of the method, the photoanodes produced by SILAR show quite high light conversion efficiencies thus favoring to the actual domination of this method in the preparation of various SSSC components. Electrochemical deposition is a convenient and potent method for the formation of chalcogenide/oxide heterostructures based on the electrochemical decomposition of chalcogenide precursors with the release of X2− ions, similarly to the CBD method. Metal chalcogenide films produced by the electrodeposition typically show a high adhesion to the oxide surface, while the sensitizer NP size can be tailored by varying the current density, temperature, and the electrolyte composition. Chemical vapor deposition (CVD) is based on the gas-phase interaction of precursors and nucleation of the sensitizer NPs on a substrate [38]. The method can be used both for the deposition of metal oxide scaffolds with a precisely controlled morphology and for the formation of narrow-bandgap sensitizer NPs. The SSSC photoanodes can also be prepared by spray pyrolysis [39], molecular beam epitaxy [40], and ultrasound-assisted deposition [41]. Recently, photocatalytic deposition was introduced as an emerging method for the formation of metal chalcogenide (sulfide, selenide) NPs using inherent photochemical activity of the most popular TiO2 and ZnO scaffolds [42]. The photodeposition showed a broad variability of the sensitizer NP parameters and good perspectives for the SSSC-related applications. Finally, good perspectives can be envisaged for various chemical transformations of unstable ZnO scaffolds, for example, ion exchange reactions that can produce a variety of binary and more complex chalcogenide/oxide heterostructures such, for example, as reported in [43] for the preparation of ZnO/ZnxCd1−xSe composites. Below in this chapter we will discuss the most popular methods used for the formation of SSSC components—the photoanodes, photocathodes, and counter electrodes. We should note that the literature on the synthesis of metal-chalcogenide NPs is of enormous volume and the reports discussed below are only a small fraction of it confined to the examples of using the ex situ produced NPs as spectral sensitizers of SSSCs.

4.3 Nanocrystalline Photoanodes Produced by the Ex Situ Deposition of Sensitizer NPs

Deposition of the ex situ synthesized sensitizer NPs. The ex situ deposition is broadly used for the preparation of photoanodes for the liquid-junction SSSCs [44]. Table 4.1 summarizes the PEC parameters of some of the reported SSSCs produced by using the ex situ deposition of sensitizer metal-chalcogenide NPs. In this approach, the sensitizer NPs are synthesized separately by using well-known 4.3 Nanocrystalline Photoanodes Produced by the Ex Situ … 171 synthetic protocols and then deposited onto the surface of oxide scaffold most often by using bifunctional molecules-linkers. A typical and very popular linker is mercaptopropionic acid (MPA), HS–CH2CH2–COOH, that can bind strongly to the surface of titanium (zinc) oxide via the carboxyl group and simultaneously to form a coordination bond with the NP surface cations via the mercapto-group [44]. The MPA molecule is short enough to allow electron transfer from the photoexcited NPs to the wide-bandgap scaffold. Recently, linear aminoalkanoic acids [45] and phosphonoalkanoic acids [46] were introduced as alternatives to the SH-based molecular linkers for the attachment of ex situ produced sensitizer NPs. The attractiveness of the ex situ deposition appears, in the first place, in broad possibilities of the variation of composition and size of metal-chalcogenide NPs as well as in the selection of an appropriate molecule-linker. The synthetic approaches typically used to produce NPs are “heating up” and “hot injection” methods [44]. The syntheses occur is organic solvents with high boiling temperatures capable of the coordination to the surface of growing NPs passivating them against the growth and aggregation. Both methods allow for a precise control over the size and size distribution of NPs and, in the case of a shell formation, also over the thickness of the shell. The heating-up method consists in the thermal decomposition of metal and chalcogen precursors (or a single precursor) in high-boiling-point solvents at 180– 280 °C [44]. Oleylamine (OLA) is very often used as a reaction medium as it can serve both as a high-boiling-T solvent and as a passivating ligand capable of coordination to the NP surface in the form of a monolayer. At the same time, it can dissolve sulfur and selenium or other chalcogen precursors, thus acting as a universal reaction medium. A similar role can be played by combinations of oleic acid (OA) and trioctylphosphine (TOP). Another popular composition for the heating up procedure combines the paraffin as an inert medium, Cd oleate and solutions of elemental chalcogens (S, Se, Te) in TOP as the NP precursors. The size of growing NPs is determined by the pyrolysis duration and the NP growth can be quenched at any desirable moment by a sharp temperature reduction. In such a way, the size-selected CdSe [47, 48], CdSexTe1−x [49], CuInS2 [50–53] and alloyed ZnSe-AgInSe2 NPs [54] can be prepared (Fig. 4.5a, b). In the case of metal sulfide NPs, dodecanthiol (DDT) is often used in various roles—as a solvent, coordinating ligand and sulfur source [50–53]. For example, by varying the duration of heat treatment of CuInS2 (CIS) NPs in DDT from 10 to 90 min the NP size can be smoothly increased from 2.9 to 5.3 nm (Fig. 4.5c, d) [51]. In the hot-injection approach, the metal precursors are dissolved in OLA (or mixtures of OLA with octadecene (ODE) or TOP) and kept at an elevated temperature [44]. Additional ligands can also be added to the reaction mixture to allow a more precise control of the NP characteristics, such as trioctylphosphine oxide (TOPO) or hexadecylamine (HDA). Then the temperature of the solution is increased (up to 320 °C) to promote the decomposition of a chalcogen precursor dissolved in TOP, which is then rapidly injected thus creating favorable conditions for the homogeneous NP nucleation. The mixture is then cooled down to 250–270 °C and Table 4.1 Examples of SSSCs produced by the ex-situ sensitizer NP deposition Photoelectrochemical Liquid-Junction Semiconductor-Based 4 172

2 Photoanode material Molecular linker Eg(NPs), eV Counter electrode Jsc, mA/cm Voc, V FF, % η, % Reference

TiO2/CdTe/CdS MAA 1.56 Au 13.60 0.682 41 3.80 [85]

TiO2/CdTe/CdSe MPA *1.4 PbS 7.77 0.602 52 2.42 [66] ZnO NW/CdSe MPA *2.0 Pt 2.10 0.5–0.6 *30 0.4a [55]

TiO2/Au/CdSe/P3OT MAA *2.0 Pt 0.91 0.610 38 0.66 [56]

TiO2/CdSe MAA *2.0 CuxS 8.53 0.56 46 2.21 [73]

TiO2/CdSe Cys *2.3 PbS 2.3 0.48 46 0.83 [77] a TiO2/CdTe MPA 2.48 Pt 3.61 850 66 2.02 [80]

TiO2/CdS0.17 Se0.87 MAA *1.9 Pt 8.72 650 39.3 2.23 [86]

TiO2/CdSe0.45 Te0.55 MPA 1.55 CuxS 19.35 571 57.5 6.36 [49] 3− TiO2/CdSe SbS4 *2.0 CuxS 6.17 510 53 1.67 [57]

TiO2/CdSe MPA 2.0 CuxS 15.93 619 65.8 6.49 [58] ZnO NR/CdS Cys *2.6 Pt 2.42 550 50 0.67 [78] 2− TiO2/CuInS2 S *1.8 Pt 4.14 543 49.4 1.11 [50]

TiO2/CdS/CuInS2 MAA *1.6 Carbon 8.21 489 37 1.47 [82]

TiO2/CuInS2:Zn MPA *1.4 CuxS 19.73 580 58 6.66 [71]

TiO2/CIS DDT *1.6 CuxS/RGO 10.10 501 47 2.38 [51]

TiO2/CIS (ex situ/in situ) MPA *1.8 CuxS 7.72 570 42 1.84 [53]

TiO2/CIS/CdS MPA *1.55 CuxS 16.9 560 45 4.2 [69]

TiO2/CIS Cys + MAA *1.8 CuxS 12.82 640 54 4.44 [83]

TiO2/ZnSe-AgInSe2/CdS MPA *1.6 CuxS 8.8 *500 43 1.9 [54]

TiO2/CdSxSe1−x no – CuxS/RGO 11.2 557 51 3.20 [65]

TiO2/CdSe no *1.83 Pt 3.0 524 27 0.4 [47]

TiO2/CdSe NRs no *1.83 PbS 9.7 564 49 2.7 [37]

TiO2/CuInS2/CdS no *1.83 CuxS/RGO 15.65 529 47 3.91 [52] 2 2− 2− Note the table reports the highest η values achieved in the corresponding papers; the cells were illuminated with AM1.5 light (100 mW/cm ) if not stated otherwise; redox couple is S /Sx if not stated otherwise; in some cases a scattering layer was applied on top of the photoanodes to increase efficiency and a ZnS layer was deposited onto the photoanode by SILAR to increase PEC efficiency (see original refs.) − … I2/I redox couple was used; P3OT is poly(3-octylthiophene) The values of total light conversion efficiency η are intentionally highlighted in bold characters 4.3 Nanocrystalline Photoanodes Produced by the Ex Situ … 173

Fig. 4.5 a, b Absorption spectra of colloidal CdSeTe, CdTe, and CdSe NP (a) and sensitized TiO2 ﬁlms (b). Insets: Photographs of colloidal NPs (a) and TiO2-based sensitized photoanodes. c Absorption spectra of the size-selected CuInS2 NP in toluene. d Photographs of TiO2/CuInS2 photoanodes with the size-selected sensitizer NPs. Reprinted with permissions from Refs. [49](a, b) and [51](c, d). Copyright (2013, 2014) American Chemical Society kept at this T for a certain time to let the NPs grow and rapidly cooled to room T to quench the NP growth. The hot-injection approach allows for a better size and size distribution control, as compared to the heating-up method, as the steps of nuclei formation and subsequent growth are separated in time. However, for the synthesis of the NPs of a desirable size a precise control over the reaction duration and temperature is required. The method is successfully applied to produce the size-selected NPs of CdSe [55–64], CdSxSe1−x [65], CdTe [66], PbS [67, 68], CuInS2 [69], AgInS2 [70], Zn-doped CuInS2 [71], and Cu2ZnSnS4 [72]. For example, in this way, by con- trolling the duration of post-injection aging the size-resolved series of 2.3, 2.6, 3.0, and 3.7-nm CdSe NPs can be produced (Fig. 4.6)[60] as well as the size-selected fractions of 2.9–6.6 nm PbS NPs [67]. After the NP growth quenching a shell of another semiconductor with a wider bandgap can be grown on the NP surface by the second round of injections of the shell material precursors [44]. The repetition of such injections allows to precisely 174 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

Fig. 4.6 Absorption spectra of toluene NP colloids (a) and IPCE spectra of TiO2/CdSe heterostructures (c) both containing 3.7, 3.0, 2.6, and 2.3 nm CdSe NPs. b, d Photographs of colloidal solutions (b) and TiO2/CdSe films (d) containing the size-selected CdSe NPs. Reprinted with permissions from Ref. [60]. Copyright (2008) American Chemical Society tune the thickness and composition of the protective shell. In this way, a ZnS shell is typically grown on the CdSe NPs [59], and CdSe shell—on CdTe NPs [66]. The metal-chalcogenide NPs produced by both methods are tightly covered with hydrophobic organic ligands (OLA or TOP) and interact weakly with the polar surface of oxide wide-bandgap materials. To achieve efficient adsorption of the NPs on the oxide surface typically a bifunctional bridge-ligand is used as the above-discussed MPA [44]. The linker is introduced in two ways. The first and a more straightforward way is soaking of the TiO2 scaffold with MPA (or another linker, like MAA [56]) solution followed by a prolonged (60–70 h) incubation of the MPA-modified TiO2 electrode in a colloidal NP solution in non-polar solvents (toluene, CHCl3, etc.). In this way, the titania surface can be decorated by CdSe [56, 60, 61, 64], CdTe [66], and AgInS2 NPs [70] grown by the hot-injection synthesis. The amount of adsorbed CdSe NPs and, therefore, the light harvesting capability of the TiO2/CdSe photoanode, can be increased considerably (in 5–6 times) by a multiple precipitation/redispersion of HDA/TOPO-stabilized CdSe NPs in toluene. This procedure results in the elimination of residual unbound ligands and partial desorption of the ligands from the NP surface thus enabling the subsequent interaction with the MPA-treated TiO2 scaffold [64]. However, removal of the ligand by the washing procedure also results in the NP aggregation on the titania surface and therefore some of the adsorbed NPs are not really attached to TiO2 and cannot participate efficiently in the charge transfer. Thus, an optimal precipitation/ redispersion cycle number exists (2 as reported in [64]) providing a balance between the amount of the adsorbed NPs and their aggregation state. The second way consists in the ex situ attachment of MPA to the NP surface via a ligand exchange [44]. In this approach, the organic phase containing NPs is brought into the contact with a polar solution (methanol, dimethylformamide (DMF), water) containing a molecule-linker and then the bi-phase mixture is subjected to vigorous mechanical or ultrasonic shaking. The MPA gradually substitutes the OLA rendering the NP surface polar and transferring the MPA-capped 4.3 Nanocrystalline Photoanodes Produced by the Ex Situ … 175

NPs into the polar solvent. Then oxide (TiO2, ZnO) film is immersed into the polar NP-MPA solution and the NPs are efficiently adsorbed. This approach was used to decorate the nanocrystalline TiO2 films with CdSe [58, 63], CdSe/ZnS [59], CdSexTe1−x [49], PbS [67], CIS [69], Zn-doped CIS [71], and alloyed ZnSe– AgInSe2 NPs [54]. The adsorption of mercaptocarboxylate-capped NPs on the surface of TiO2 depends strongly on pH of the NP solution. At neutral pH (around 8) the carboxyl groups of the stabilizers are mostly protonated and the NPs can bind tightly to the mesoporous oxide scaffold, thus blocking the surface layer and hindering further portions of the NPs from penetration into the bulk of the oxide film [73]. Also, as pH becomes lower the hydrodynamic size of MPA-capped CdSe NPs is reported to increase considerably indicating that agglomeration of the NPs takes place, further lowering the NP absorption efficiency [58]. At an elevated pH (higher than 10) the mercaptocarboxylate ligands are mostly ionized and charged negatively thus experiencing electrostatic repulsion from the surface of TiO2 that is also negatively charged. This repulsion, however, favors to the NP diffusion deeper into the mesoporous oxide scaffold and results in better adsorption and higher loadings of the NPs, which can be further increased by elevating the temperature of NP solution [73]. Also, it is reported that the deprotonated thiolate group can form much stronger (by around 40 times) coordination bonds with Cd(II) ions on the surface of CdSe NPs [58] and, therefore, the NPs appear to be much more resistant to the agglomeration in such conditions. Due to these factors the light conversion efficiency of the SSSCs based on the MPA-terminated CdSe NPs generally increases considerably with an increase of pH of the solution used to deposit NPs onto the titania surface (Fig. 4.7a). The same ligand exchange methodology can be applied to the ternary cadmium-free NPs, such as CIS and AgInS2 (AIS) chalcopyrites. The OLA-capped CIS NPs can be rendered water-soluble by the ligand exchange with MPA or sulfide ions [50]. The light conversion efficiency on TiO2/CIS heterostructures depends strongly on the size of a capping ligand and expectedly increases when bulky OLA or DDT is substituted with smaller MPA and S2− (Fig. 4.7b, gray bars) [50]. Even more dramatic changes can be observed in the rate constant of the electron transfer from the capped-CIS NPs to TiO2—kET increases by around an order of magnitude after OLA (or DDT) is exchanged to smaller MPA and S2− species (Fig. 4.7b, red bars). The changes in the electron transfer dynamics are also illustrated by a drastic decrease of the charge transfer resistance between CdSe NPs and TiO2 as OLA or DDT are substituted with MPA and sulfide ions (Fig. 4.7b, blue circles). In a similar way, the nanocrystalline ZnO-based photoanodes can be produced. For example, TOPO/HDA ligands on the surface of CdSe NPs can be substituted by MPA in methanol solutions resulting in good adsorption of the ligand-exchanged CdSe NPs on the surface of ZnO nanowires (NWs) [55]. The coverage of ZnO NWs with NPs can be increased substantially via a preliminary treatment of the NWs with oxygen plasma [55]. The treatment can influence the ZnO NWs in multiple ways—it can charge the surface attracting the negatively charged CdSe– 176 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

Fig. 4.7 a Dependence of the light conversion efﬁciency for SSSCs based on the MPA-capped CdSe NPs on pH of colloidal solution used for the photoanode preparation. Plotted using data presented in [58]. b Light conversion efﬁciency η, electron transfer rate constant kET and charge transfer resistance RCT calculated from the electrochemical impedance spectra of TiO2/CIS photoanodes based on differently capped CIS NPs. Plotted using data reported in [50]

MPA NPs, eliminate surface impurities that prevent efficient NP adsorption, and produce dangling (non-compensated) bonds tending to interact with the NPs [55]. Similar approaches were applied to anchor PbS NPs onto the mesoporous ZnO [68] and CdSe NPs—on ZnO nanotubes (NTs) [62]. A series of different molecules were tested as linkers for the attachment of PbS NPs to ZnO films [68], in particular oxalic, malonic and thioacetic acids, MAA (which is often referred to as thioglycolic acid, TGA) and MPA as well as hexanedithiol. The highest photoresponses were obtained for the ZnO/PbS systems with the ligands having a free –SH group available for binding to the undercoordinated Pb atoms on the PbS NP surface— TGA, MPA, and hexanedithiol (Fig. 4.8a) [68]. Iodide-capped 6–7 nm PbSe NPs (Fig. 4.8b) were attached to the ZnO surface by cysteine (HS–CH2–CH(NH2)–COOH, Cys) [74]. Such NPs impart the zinc oxide films with the spectral sensitivity to 1800–1900 nm as confirmed by both absorption and IPCE spectral measurements (Fig. 4.8c, d). The 6–7 nm PbSe NPs reside in the strong quantum confinement regime resulting in a considerable increase of the CB potential and making possible the photoinduced electron transfer into CB of the ZnO scaffold, contrary to the PbSe bulk materials (Fig. 4.8e) [74]. The attempts to apply the ligand exchange with MPA to the ternary CIS NPs resulted in strong NP aggregation during the phase transfer into water. To cir- cumvent the aggregation effect a two-step ligand exchange was proposed [53]. On the first stage, the original DDT ligands were partially replaced with OA and the resulting NPs dispersed in water with the help of an ultrasound treatment. Oleic acid replaces partially DDT and enters the ligand shell of CIS NPs interacting with the alkyl chains of neighboring thiol molecules, while –COOH group remains in the outer part of the shell. In alkaline solutions, the carboxyl group of OA is deprotonated and protects the CIS NPs against aggregation via the electrostatic repulsion 4.3 Nanocrystalline Photoanodes Produced by the Ex Situ … 177

Fig. 4.8 a Scheme of a supposed binding of different ligands to the ZnO surface; b TEM of as-synthesized OA-capped PbSe NPs; c, d absorption (b) and IPCE (c) spectra of colloidal PbSe NPs (c) and ZnO/PbSe heterostructures (d); e Band energy level diagram depicting the relevant energy levels of bulk PbSe and 6–7 nm PbSe NPs and ZnO crystal. Reprinted with permissions from Refs. [68](a) and [74](b–e). Copyright (2011, 2016) American Chemical Society between the NPs [53]. On the second step, the CIS NPs are brought into the contact with a large excess of MPA that substitutes both residual DDT and OA, producing the non-aggregated water-soluble NPs that can easily be attached to TiO2 [53]. Kesterite Cu2ZnSnS4 NPs prepared by the hot-injection method can be rendered water-soluble by a ligand exchange with graphene oxide reduced by aromatic thiols [72]. The sheets of reduced graphene oxide (RGO) produced by this method are decorated with C–SH and C=S groups that can coordinate to the NP surface, similarly to MPA and MAA. Along with the organic bifunctional linkers, other types of small molecules and metal complexes are probed as potential linkers for the attachment of metal-chalcogenide NPs to oxide surfaces. Of particular interest are 3− 3− metal-chalcogenide inorganic complex ligands (ICL), such as SnS4 ,SbS4 and 3− AsS3 [57]. Such ICL can be relatively easily produced by the dissolution of corresponding metal sulfides in an excess of Na2S, they have a high affinity both to the NPs and the oxide surface and can efficiency stabilize the NPs because of a relatively high negative charge. Also, the ICL-capped NPs can readily self-assemble in tightly packed single layers which is very favorable for the formation of uniform SSSC photoanodes. Similarly to MPA (MAA), ICLs can be introduced by a simple ligand exchange and promote the phase transfer of CdSe NPs into stable aqueous solutions [57]. A detailed study of the CdSe-ICL systems with a combination of the UV photoelectron spectroscopy and time-resolved photoluminescence (PL) spectroscopy revealed a correlation between the LUMO position of the complexes relative to ECB 178 4 Semiconductor-Based Liquid-Junction Photoelectrochemical … of CdSe NPs, the rate of photoinduced charge transfer from CdSe NPs to the titania scaffold, and the PEC efficiency of such TiO2/NPs assemblies [57]. In particular, the charge transfer rate constant was found to be the highest (1.4 Â 1011 s−1) for the Sn-ICL, that has the lowest LUMO level relative to the ECB of CdSe NPs, and decreasing for Sb-ICL (5.1 Â 1010 s−1) and As-ICL (3.4 Â 1010 s−1) because the barrier for the electron transfer (a delta between LUMO and ECB) increases from 0.21 eV for Sn-ICL to 0.77 eV for Sb-ICL to 1.26 eV for As-ICL (Fig. 4.9a). However, the PEC light conversion efficiency does not follow this trend. It is maximal for the Sb-ICL (1.67%), lower—for As-ICL (1.24%) and much lower (0.61%)—for the Sn-ICL. The authors of Ref. [57] hypothesized that the PEC performance of CdSe-ICL-TiO2 assemblies is determined not only by the efficiency of electron transfer from CdSe NPs but also by the rate of valence band hole transfer to the electrolyte that meets the highest barrier for the Sn-ICL (Fig. 4.9a). Similarly to the above-discussed polyelectrolyte-assisted multi-layer NP adsorption, the ICL-terminated CdSe NPs can be deposited as multilayers by alternating the deposition of negatively charged NPs with the adsorption of Cd or 2+ Zn cations [57]. In this way, by using Cd the PEC efficiency of a TiO2/CdSe heterostructure with the Sb-ICL linker can be increased to 1.84% for a four-layer deposition of the sensitizer NPs (Fig. 4.9b). Alternatively, the NPs can be attached to the titania surface without linkers. For this, the native ligands (OLA, TOPO, etc.) can be partially or even completely eliminated by multiple washing of the NP precipitate with methanol [61]orCH2Cl2 [48]. The direct adsorption can result in even closer interaction between the NPs and titania scaffold. For example, the charge transfer rate constant measured by the time-resolved PL for CdSe NPs treated with methanol was found to be more than 3 9 −1 9 −1 times higher (7.2 Â 10 s vs. 2.3 Â 10 s ), than for similar TiO2/CdSe composites produced using the MPA linker [61]. However, this method also suffers from the aggregation of CdSe NPs devoid of their ligand shell and, therefore, the washing conditions should be chosen very carefully to achieve high PEC characteristics.

Fig. 4.9 a Energy diagram for CdSe NPs, TiO2 scaffold, and three ILCs. b Jsc and light power conversion efﬁciency (PCE) for SSSCs based on 1–4 layer CdSe/Sb-ICL NPs. Reprinted with permissions from Ref. [57]. Copyright (2015) American Chemical Society 4.3 Nanocrystalline Photoanodes Produced by the Ex Situ … 179

The preparation of photoanodes is typically finalized by the deposition of a protective layer of ZnS [48, 49, 58, 66, 69, 71, 73] or a CdS layer [52, 54, 69], most frequently, using the SILAR technique. A ZnS (CdS) layer protects the NPs from the photochemical and “dark” corrosion and prevents the “leakage” of photogenerated charge carriers from NPs into the electrolyte. Direct aqueous synthesis. As shown by the above discussion, the mercaptocarboxylate-capped CdX (X = S, Se, Te) NPs can easily adsorb on the oxide surface, being in this way very similar to the conventional Ru-complex sensitizers of the dye-sensitized solar cells, where the dyes bind to the TiO2 surface via the –COOH groups [73]. In view of the complexity of the ex situ synthesis and the post-synthesis ligand exchange, a direct synthesis of the mercaptocarboxylate-terminated NPs in water and other polar solvents is greatly preferable. As the synthesis temperature is restricted by the solvent boiling point (100 °∁ for water) the direct synthesis does not allow such precise size variations and structural perfection of the NPs as the above-discussed heating-up and hot-injection approaches. However, the direct syntheses provide another, quite powerful methods of size variation and thus can be strong competitors to other ex situ synthetic protocols in view of their simplicity and a “green” nature. Typically, CdSe NPs can be synthesized directly in water via the interaction between a chalcogen precursor and Cd(II) complexes with ligands-stabilizers. In this way, 2.3-nm CdSe NPs stabilized by MAA were produced in aqueous solutions [73] that can be directly adsorbed on the surface of mesoporous TiO2. Ultra-small colloidal core/shell CdSe/CdS NPs can be produced by a direct aqueous synthesis [75] and used as a sensitizer of mesoporous TiO2 (Fig. 4.10)ina SSSC with polysulfide electrolyte and a copper sulfide-based counter electrode [76]. The photoanodes were prepared by simply soaking the titania film with colloidal solutions of 1.8–2.0-nm CdSe/CdS NPs (Fig. 4.10a, b). The sensitizer NPs penetrate the bulk of mesoporous titania film very uniformly showing identical atomic cadmium, selenium, and sulfur contents both near the FTO transparent electrode, in the bulk of TiO2 film and on the film surface (Fig. 4.10c–e). The CdSe/CdS NPs absorb light in a spectral range of k < 450−460 nm and reveal a high chemical and photochemical stability. The total light conversion efficiency in a SSSC with the ultra-small NP-based FTO/TiO2/CdSe/CdS photoanode and an FTO/ TiO2/Cu2S CE formed by the sulfidation of photocatalytically deposited Cu NPs is as high as 6.3% [76]. An aqueous synthesis of cadmium selenide NPs using cysteine anions as a capping agent results in the formation of ultra-small CdSe NPs with a well-resolved absorption maximum at 422 nm [77] that allows to identify these NPs ad the so-called “magic-size clusters”—stable ultra-small CdSe NPs with a well-defined number of monomeric CdSe units in each NPs. At 80 °C the synthesis yields “regular” CdSe NPs with a bandgap around 2.32–2.34 eV, corresponding to the average size of around 2.5 nm. Both NP types can easily be attached to the titania surface upon the TiO2 film immersion into the colloidal CdSe solution [77]. Cysteine can be applied also for the aqueous synthesis of CdS NPs with the NP size 180 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

Fig. 4.10 a Cross-sectional SEM image of mesoporous TiO2 ﬁlm soaked with colloidal CdSe/CdS NP solution (a), the elemental analysis was performed in the points numbered 1–5 and the results presented for Cd, Se, and S in (e); b TEM/HRTEM images of colloidal CdSe/CdS NPs; c, d Cd (c) and Se (d) atom distribution in the cross section of the TiO2/CdSe/CdS ﬁlm varying from 2.3 to 2.8 nm [78]. The cysteine-capped CdS NPs readily adsorb on the surface of ZnO NRs producing uniform ZnO/CdSe nanoheterostructures. The MPA-capped CdTe NPs can be produced directly in aqueous solutions via the sodium tellurite reduction with NaBH4 in the presence of Cd(II) salts under the microwave heating and then adsorbed onto the surface of ZnO NRs [79]. A variation of the heating duration (7–30 min) allows tunig of the average size of CdTe NPs from 4 to 9 nm. Alternatively, CdTe NPs can be formed at the expense of Te reduction by NaBH4 in boiling aqueous solutions containing cadmium perchlorate and MPA [80, 81]. Such NPs can be deposited onto the TiO2 surface either by spontaneous adsorption from the solution [80] or by the drop-casting and evaporation of CdTe colloid on the TiO2 scaffold surface [81]. As the MPA-capped CdTe NPs are charged negatively, the multi-layer deposition of the NPs onto titania is possible together with a positively charged polyelectrolyte— poly(dimethyl diallyl ammonium chloride) [80]. The procedure includes a cyclic adsorption of the polyelectrolyte and MPA-capped NPs attracted to each other by the electrostatic forces. The ternary CIS NPs formed directly in aqueous solutions in the presence of TGA can be used as “inks” to sensitize porous TiO2 substrates via a simple immersion technique [82]. 4.3 Nanocrystalline Photoanodes Produced by the Ex Situ … 181

Similarly to CdTe, CIS NPs can be produced directly in aqueous solutions of various sulfur-containing ligands (TGA, MPA, glutathione (GSH), Cys) under the microwave heating [83]. By applying a co-linker (for example, TGA to the Cys-capped CIS NPs) the amount of NPs adsorbed on titania can be increased considerably resulting in a drastic (by more than 20 times) increase in the light conversion efficiency [83]. The additional ligand is supposed to participate in the chemical reduction of S–S fragments that form on the NP surface as a result of partial oxidation of the primary ligand and hinder the NP adsorption on the TiO2 surface. A PL quenching study showed that the rate constant of electron transfer 10 −1 from CIS NPs to TiO2 is the highest for Cys linker, 9.5 Â 10 s , decreasing to 7.1 Â 1010 s−1 for a more bulky GSH. The same tendencies were found also for AIS and CdSxSe1−xS NPs [83]. The MAA-stabilized CIS/ZnS NPs penetrate uniformly the volume of mesoporous TiO2 films revealing a homogeneous composition of the resulting TiO2/CIS both along the cross section of the films (Fig. 4.11a, b) and across the outer film surface (Fig. 4.11c). The stability and PEC activity of CIS NPs both increase upon the deposition of a thin ZnS shell on the CIS NP surface. The TiO2/CIS/ZnS composites act as visible-light-sensitive photoanodes in the SSSCs with polysulfide electrolyte and copper sulfide-based counter electrodes with a total conversion efficiency of around 8% [84]. Mixed ex situ/in situ approach. In this approach, the TiO2 films are immersed into polar solutions where the primary nuclei of sensitizer NPs form. The growth and attachment of the metal-chalcogenide to the oxide surface occur simultaneously during the following heat treatment in the solvothermal conditions resulting in a uniform NP distribution over the film volume. The NP size can be tailored by varying the duration or/and temperature of the heat treatment. For example,

Fig. 4.11 Cross-sectional SEM images of FTO/TiO2/CIS/ZnS photoanode (a, b); results of the energy-dispersive X-ray spectroscopic determination of elemental composition of the photoanode cross section (b) and the outer surface (c)[84] 182 4 Semiconductor-Based Liquid-Junction Photoelectrochemical … a hydrothermal treatment (HTT) of the nanocrystalline TiO2 films in a solution of primary CdTe nuclei, produced by the injection of NaHTe into aqueous solution of cadmium(II) mercaptoacetate, yields CdTe NPs anchored to the titania surface via the MAA bridge [85]. The average size of CdTe NPs can be varied from around 3 to 6 nm by increasing the HTT temperature from 80 to 160 °C. Additionally, a CdS shell was spontaneously deposited onto the CdTe NPs as a result of MAA hydrolysis rendering the NPs stable toward the air oxidation [85]. The nuclei for the preparation of CdSe NPs can be produced by the injection of NaHSe into alkaline aqueous Cd(II)-MAA solution under an inert atmosphere [73]. In a similar way, a mixture of Na2S and NaHSe, prepared via the reduction of selenium with NaBH4, was used to form the nuclei of MAA-stabilized CdSxSe1−x NPs [86]. A photoanode is then produced by a HTT of a TiO2 film immersed into the nuclei solution. Electrophoretic deposition. The metal-chalcogenide NPs can be deposited onto the surfaces of oxide substrates indiscriminately of the NP surface chemistry by the electrophoretic deposition technique. In this method, two electrodes—FTO/TiO2 and bare FTO are immersed into the NP solution and a voltage of 60– −1 200 V Â cm is applied between the two electrodes, with the FTO/TiO2 film typically connected to the positive terminal of the power supply unit [51, 65]. The magnitude of applied voltage depends on the solvent polarity and increases from 60 V/cm for aqueous solutions to 200 V/cm for toluene [37, 47]. The NPs move in the electrostatic field and deposit as a uniform layer on the polarized titania surface. Most probably, the NPs are stripped from a portion of their protecting ligand layer in the process of the field-stimulated adsorption, coming, therefore in a close contact with the oxide scaffold. The electrophoretic deposition was successfully applied to decorate porous titania scaffolds with CdSe [47], CdSxSe1−x [65], and CIS NPs [51, 52]. The method can be also used to form multi-layer structures as shown by the layer-by-layer deposition of composition-selected CdSxSe1−x NPs with an increasing bandgap simply by switching between the NP solutions [65]. The electrophoretic deposition was found to be especially attractive to produce uniform sensitizer layers when conventional adsorption of MPA-capped NPs is inefficient for some reasons, for example, in the case of elongated CdSe NRs that meet difficulties in penetrating the mesopores of titania scaffolds [87].

4.4 Nanocrystalline Photoanodes Produced by the In Situ Deposition of Sensitizer NPs

The in situ formation of sensitizer NPs occurs directly on the surface of wide-bandgap oxide as a result of chemical reactions taking place in the oxide surface layer with the participation of surface functional groups or the charge carriers generated in/injected into the wide-bandgap material. In the former case, 4.4 Nanocrystalline Photoanodes Produced by the In Situ … 183 metal ions are first adsorbed on the oxide surface, then the interaction between the adsorbed metal ions and chalcogen anions takes place resulting in a layer of metal chalcogenide NPs. The two ideologies of NP deposition—in situ and ex situ are in constant com- petition, as each of them provides unique possibilities of the control of morphology and properties of the NPs and/or NP-oxide interface, but, at the same time, each has inherent drawbacks and limitations. In particular, the in situ deposition by SILAR or CBD ensures a good contact between the metal oxide scaffold and the NP sensitizer, because the NP formation starts immediately on the surface oxide layer and typically an intermediary thin oxide/chalcogenide layer forms between the oxide and NPs ensuring perfect electron transport from the photoexcited NPs to the oxide scaffold. The NPs deposited by ex situ methods are invariably separated from the oxide surface by a ligand shell, which can be thin, for example, in the case of using MAA or MPA ligands, but nevertheless affecting negatively the efficiency of electron transfer from NPs to the porous oxide layer. However, the ex situ formation allows tuning the size, shape, and composition of the NPs by well-established synthetic protocols with a precision typically unachievable for the in situ methods. As a result, the properties of NP/oxide photoanodes depend strongly on the method of preparation, in particular, the way of NP deposition and both in situ and ex situ methods are constantly developing and brought to the comparison. Deposition of sensitizer NPs by SILAR. The method is very simple from the experimental viewpoint but, at the same time, it allows to produce a variety of metal-chalcogenide NP-based heterostructures exhibiting quite high efficiencies of the light conversion. Typically, the SILAR procedure consists in the immersion of a wide-bandgap porous oxide film (TiO2, ZnO) into a solution of a metal precursor (soluble salts) for some time necessary for the adsorption/desorption equilibrium to settle, then the film is extracted, washed with pure water and immersed into another solution containing chalcogenide X2−/HX− ions (X = S, Se, Te) or a chalcogenide precursor that can readily decompose producing the chalcogenide ions. At that, a thin (ideally—a monomolecular) layer of metal-chalcogenide forms on the surface of the oxide film. The procedure is them repeated many times each cycle producing an additional layer of the metal chalcogenide. In this way, the metal chalcogenide layer thickness (and in some cases—the NP size) is determined by a number N of the SILAR cycle repetitions that can vary from 2–3 to tens for manual preparation and even to hundreds if a mechanically-controlled setup is used for the film preparation. Typically, the SILAR produces smooth layers of sensitizer NPs covering the entire surface of the wide-bandgap material and does not block the pores in the oxide scaffold that can still be freely penetrated by the electrolyte after the NP deposition. The SILAR can be applied to different wide-bandgap scaffold materials (TiO2, ZnO, In2O3) and morphologies (mesoporous films, NRs, NWs, nanosheets, and nanoplates, etc.), NP sensitizers (CdX, PbX, CuX, ternary/multinary NPs) and performed from different solvents (water, methanol, ethanol, acetone). Table 4.2 demonstrates some examples of the SSSCs where the visible-light-sensitive component was produced by the SILAR technique. 184 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

Table 4.2 Some examples of SSSCs produced by the SILAR deposition of sensitizer NPs

Photoanode material Eg(NPs), Counter Jsc, Voc, FF, η, Reference eV electrode mA/cm2 V % % ZnO/CdS (ZnO 2.4 Pt 7.2 n/r n/r 3.53 [94] NWs)

TiO2/CdS/CdSe *1.8 Au 11.9 0.51 53 3.20 [116] (meso-TiO2)

TiO2/CdS (TiO2 2.4 Pt 6.4 0.46 41 1.19 [92] NTs)

TiO2/CdS (TiO2 *2.1 CuxS 9.3 0.59 42 2.29 [93] nanosheets)

TiO2/ZnS/CdS/ZnS *2.4 CuxS 10.3 0.64 57 3.69 [88]

TiO2/CdS/Bi2S3/ZnS *1.5 CuxS 9.3 0.50 54 2.52 [89]

TiO2/CdSe *1.8 CuxS 15.9 0.58 57 5.21 [90]

ZnO/Ag2S (ZnO 1.0 Au 11.4 0.26 38 1.10 [95] NWs)

TiO2/Ag2S n/r Pt 7.3 0.33 41 0.98 [105]

TiO2/AgSbS2 1.7 Au 2.4 0.32 n/r 0.34 [101]

TiO2/AgBiS2 1.32 Pt 7.6 0.18 39 0.53 [99]

TiO2/PbS/CdS n/r CuxS 10.9 0.44 46 2.21 [103]

TiO2/PbS/ *2.2 CuxS 10.3 0.42 32 1.37 [104] Pb0.2Cd0.8S/CdS 2+ TiO2/PbS:Hg *1.0 CuxS 30.0 0.40 47 5.58 [108]

TiO2/CuxTe *1.0 Carbon 8.6 0.50 16 0.69 [118] a TiO2/CuInS2/Bi2S3 *1.5 CuxS n/r n/r n/r 0.44 [100]

TiO2/Cu2Se/CuInS2 1.5 Pt 6.5 0.59 32 1.22 [107] Note The table reports the highest η values achieved in corresponding papers; the cells were 2 2− 2− illuminated with AM1.5 light (100 mW/cm ) if not stated otherwise; redox couple is S /Sx if not stated otherwise; in some cases a scattering layer was applied on top of the photoanodes tpo increase efﬁciency (see original refs.) n/r—not reported a electrolyte contained Na2S and Na2SO3

As the most frequent case, CdS NPs are deposited by SILAR using aqueous or alcohol solutions of Cd(II) nitrate and Na2S on the mesoporous titania [88–91], TiO2 NTs [92] and nanosheets [93], ZnO NWs [94, 95], nanoplates [91, 96] and NRs [97], producing visible-light-sensitive TiO2/CdS and ZnO/CdS heterostructures. Typically, cadmium sulﬁde forms a dense layer on the TiO2 (ZnO) surface shielding it from the electrolyte. According to the optical data, the CdS layer consists of separate NPs with the NP size and the layer thickness in general depending on the number of SILAR cycles. At primary steps of the SILAR procedure the thickness and absorbance of the CdS layer depend almost linearly on the SILAR cycle number. Figure 4.12 shows that the thickness of CdS layer deposited by the SILAR onto the ZnO NWs increases continuously from 3 to around 12 nm as N grows from 10 to 120 [94]. A comparison of TEM and optical absorption data indicates that for a given 4.4 Nanocrystalline Photoanodes Produced by the In Situ … 185

Fig. 4.12 TEM images of ZnO NWs decorated with a CdS layer by the SILAR with a different cycle number N. Reproduced with permissions from Ref. [94]. Copyright (2009) The Royal Society of Chemistry

ZnO/CdS nanoheterostructure the CdS layer thickness is comparable to the size of CdS NPs [94] indicating a monolayer coverage of the ZnO NWs with CdS NPs of a different size. The PEC activity of the TiO2/CdS and ZnO/CdS heterostructures produced by the SILAR also depends on the cycle number N and increases at first, then decreases revealing a distinct maximum. For example, Jsc produced by the NW-based ZnO/CdS heterostructures increases till N = 30 and then comes to a saturation value for a much higher N (up to 120) [94]. At the same time, for TiO2/CdS composites produced from the anodized titania NTs the light conversion efficiency grows up to N = 5 and falls considerably at a higher SILAR cycle number (Fig. 4.13a, curve 1) [92]. The efficiency decrease is associated with the blockage of NT openings that prohibits the electrolyte penetration and the regeneration of sensitizer NPs.

Fig. 4.13 a Efﬁciency of SSSCs based on TiO2 NT/CdS (curve 1) and meso-TiO2/Ag2S (curve 2) heterostructures as a function of the SILAR cycle number N, plotted using data reported in [92] (curve 1) and [105] (curve 2); b Photocurrent generation spectra of ZnO nanoplates (curve 1) and ZnO/CdS heterostructures produced at N = 5 (curve 2), 10 (3), 20 (4), 60 (5), and 200 (6). Y is the photocurrent quantum yield [98] 186 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

On the contrary, for the mesoporous TiO2 (ZnO) with no particular spatial arrangement the light conversion efficiency typically saturates at a certain N (varying for the scaffold nature) and almost does not change at a higher number of the SILAR cycle repetitions [91, 92, 94]. These examples show that the photoanode efficiency dependence on the amount of SILAR-deposited NPs can vary quite strongly reflecting differences in the shape and surface chemistry of the wide-bandgap scaffolds. The CdS NP-based heterostructures produced by the SILAR reveal another intriguing property that is often overseen by the researchers. As discussed above, the thickness of cadmium sulfide layer and the size of CdS NPs both increase as the cycle number N is elevated. The size increase is accompanied by a lowering of the bandgap of CdS NPs and a corresponding “red” shift of the band edges both in optical absorption and photocurrent spectra of TiO2/CdS (ZnO/CdS) heterostructures. The bandgap shrinking indicates a continuous weakening of the quantum size effects in the growing CdS NPs and one can expect that for NPs larger than the doubled Bohr exciton radius in cadmium sulfide (typically for d > 10 nm) no quantum size effects will be observed and the bandgap of CdS NPs will reach the value of Eg = 2.4 eV typical for the bulk cadmium sulfide. However, the TiO2/CdS (ZnO/CdS) heterostructures produced at a relatively high SILAR cycle number, N >20–30, reveal distinctly lower Eg values, that can be as small as 2.0–2.2 eV corresponding to the band edge of around 600 nm (Fig. 4.13b) [98]. The phenomenon of the bandgap of SILAR-produced CdS NPs being lower than the bulk value appears to be registered quite frequently but not paid due attention [92, 93, 95]. Gary Hodes et al. were the first group to assess this phenomenon systematically [97], probing several alternative explanations, including formation of a type II heterojunction between the wide-bandgap scaffold (ZnO in this case) and CdS NPs, possible effects of adsorbates and surface states on the band structure, contribution of Cd-enriched faces of CdS NPs into the band edge positions, and, finally, the participation of sub-bandgap states in the light absorption and the photocurrent generation. The analysis of Hodes et al. allowed to conclude definitely [97] that the extension of the edges of absorption and IPCE bands in the spectra of ZnO/CdS (TiO2/CdS) heterostructures originates from a large contribution of sub-bandgap states introduced by a high structural disorder of the lattice of CdS NPs deposited by the SILAR, unlikely all other deposition methods. This conclusion found support in the studies of CdS-based photoanodes on various scaffolds (ZnO, TiO2, In2O3) by the resonant Raman spectroscopy [91] confirming a high structural disorder of the SILAR-produced CdS NPs as indicated by the appearance of characteristic disorder-activated vibrational modes. Apart from cadmium sulfide, the SILAR technique was successfully applied for the preparation of photoanodes comprising NPs of ZnS [88, 89], Bi2S3 [89, 99, 100], Sb2S3 [101, 102], PbS [103, 104], Ag2S[95, 99, 101, 105, 106] and others. Recently, this method was extended for the synthesis of ternary metal-chalcogenide NPs, such as CIS [100, 107]. In this case, each SILAR cycle includes the successive adsorption of copper(II) and indium(III) followed by their interaction with sulfide ions. 4.4 Nanocrystalline Photoanodes Produced by the In Situ … 187

Figure 4.14 shows how the absorbance of some toxic-metal-free sensitizer NPs grows as the SILAR cycle number is continuously increased. The morphology of oxide/chalcogenide heterostructures produced by the SILAR depends strongly on the interaction between the oxide surface and metal ions in the forming NPs as well as on the metal chalcogenide layer thickness. In particular, the SILAR-deposited CdS NPs typically tend to form a uniform and dense layer on the surface of both titanium and zinc oxides. On the contrary, Ag2S[95, 105, 106] and PbS [103] NPs formed by the SILAR are typically larger and distributed randomly over the oxide surface (Fig. 4.14a, insert) thus leaving a portion of the surface accessible for the electrolyte and for the potential recombination between the reduced/oxidized electrolyte species and the photogenerated holes/electrons, respectively. This recombination is one of possible reasons for a generally lower light conversion efficiency of Ag2S-based (Fig. 4.13a, curve 2) and PbS-based (Table 4.2) heterostructures as compared to their CdS-based counterparts. A larger size of the SILAR-deposited Ag2S and PbS NPs with respect to CdS NPs reflects faster aggregation of the less-soluble silver sulfide particles already on the stage of the primary nuclei formation. At the same time, the Ag2S-based photoanodes also reveal a typical volcano-shaped dependence of the light conversion efficiency on N (Fig. 4.13a, curve 2) indicating on a general character and reasons for such dependence for various sensitizer NPs. Apart from individual metal-chalcogenide NPs, the SILAR allows to deposit mixed metal solid-solution chalcogenide NPs and to dope NPs with another metal [108]. For example, by using Pb(II) and Cd(II) precursors individually and as a mixture, a graded multi-layer TiO2/PbS/PbxCd1−xS/CdS photoanodes were formed revealing a light conversion efficiency of around 1.4% [104] (Table 4.2). By varying the composition of a mixture of Cd(II) and Zn(II) salts, a series of ZnO/ CdxZn1−xS photoanodes can be produced with a tunable spectral response and positions of the CB and VB levels [109]. As the reactivity of both metals towards

Fig. 4.14 Absorption (a, c) and transmission (b) spectra of TiO2-based photoanodes produced at a different SILAR cycle number (given on figures) with Ag2S NPs (a), AgBiS2 NPs (b), and CuInS2 NPs (c). Insert in (a) TEM of Ag2S NPs on the TiO2 surface. Reprinted with permissions from Refs. [105](a), [99](b), and [100](c). Copyright (2010 (a), 2013 (b), 2015 (c)) Elsevier (a), American Chemical Society (b), The Royal Socienty of Chemistry (c) 188 4 Semiconductor-Based Liquid-Junction Photoelectrochemical … sulfide anions is typically different, the composition of mixed NPs can differ quite notably from the solution composition set during the SILAR procedure. In particular, the alloyed layer in the above-mentioned TiO2/PbS/PbxCd1−xS/CdS composite contains 20 mol% lead while only 5 mol% Pb(II) was present in the precursor solution [104]. Therefore, the composition of a mixed photoanode should be ver- ified in each specific case as, for example, in the SILAR-produced ZnO/CdxZn1−xS system, where a correlation between the real and nominal Zn content was determined independently by optical, Raman and energy-dispersive X-Ray spectroscopies [109]. The CdxZn1−xS solid solution is a perfect “polygon” for probing dependences of the photochemical/PEC activity on the CB and VB energies because both values can be easily varied by changing the NP composition at more or less constant size and lattice parameters. As mentioned above, the photovoltage in a liquid-junction SSSC depends on the difference between the redox-potential of the electron shut- ting couple in the electrolyte and the Fermi energy of the photoanode. In the cadmium-zinc-sulfide-based SSSCs the latter can be approximately assumed to be linearly dependent on ECB of CdxZn1−xS NPs. The conduction band potential of mixed CdxZn1−xS NPs varies from ECB(CdS) = −0.8 V (versus normal hydrogen electrode, NHE) to ECB(ZnS) = −1.8 V (versus NHE). The details of calculations of composition-dependent band potentials of CdxZn1−xS NPs can be found in Chap. 6. The open-circuit photovoltage in the SSSCs based on the ITO/ZnO/CdxZn1−xS photoanodes, Voc(x), grows with an increase in the Zn content (Table 4.3). The photovoltage increment with respect to Voc of a CdS-based photoanode, DVoc(x)=Voc(CdS) − Voc(x) generally follows the corresponding increment of the CB potential, DECB(x), showing a considerable deviation between the two parameters only for the smallest studied Cd content at x = 0.62. A decrease in the Cd content is also accompanied by an increase in the potential corresponding to the maximal power of the CdxZn1−xS-based SSSCs. As the sensitizer is changed from CdS to mixed cadmium-zinc-sulfide NPs with x = 0.62 this potential grows by around 250 mV [109], similarly to the corresponding increase in the CB potential of the sensitizer NPs (220 mV, Table 4.3), as well as to the open-circuit photovoltage of the corresponding cells (280 mV, Table 4.3). A close

Table 4.3 Molar Cd fraction x in CdxZn1−xS NPs (determined by the energy-dispersive X-ray spectroscopy), bandgap Eg of CdxZn1−xS determined by the optical absorption spectroscopy), the photocurrent density at a dark immersion potential, Jph, the open-circuit photovoltage Voc, DECB(x) and DVoc(x) parameters for the solar cells based on CdxZn1−xS NPs [109] −2 xEg,eV Jph,mAÂ cm Voc,mV DECB(x), mV DVoc(x), mV 0.62 2.69 0.43 1.46 220 280 0.74 2.62 0.47 1.37 170 190 0.83 2.53 0.40 1.30 105 120 0.95 2.47 0.29 1.24 56 60 1.09 2.40 0.23 1.18 0 0 4.4 Nanocrystalline Photoanodes Produced by the In Situ … 189 increment of the photovoltage, around 310 mV, was also observed for the second-generation thin-film solar cells based on CdxZn1−xS/CdTe heterostructures, where x was varied from 1.00 to 0.65 [110]. The photocurrent density Jph of the CdxZn1−xS-based photoanodes grows as x is decreased from 1.00 to 0.75–0.80, but then falls at a higher Zn(II) content (Table 4.3, Fig. 4.15a, curve 1) [109]. Such behavior was also observed for the SILAR-produced TiO2/CdxZn1−xS heterostructures at x < 0.75 and explained by a strong blue shift of the absorption band edge, kbe, of the sensitizer NPs, resulting in a partial loss of the solar light harvesting capability [111]. A similar reason is responsible for a Jph decrease in the case of ZnO/CdxZn1−xS system because the dependence between the absorbance-normalized Jph and the CB potential shows the expected monotonous ascending behavior (Fig. 4.15a, curve 2). The figure shows that the absorbance-normalized photocurrent density increases by a factor of almost 4asx is lowered from 1.0 to 0.62 indicating a crucial role of the energy band positions of sensitizer NPs for the SSSC performance. The relationship between the composition-variable CB potential increment of CdxZn1−xS NPs and the normalized photocurrent density can be described by Tafel equation DE = a + blogi, where a and b are coefficients and DE is an over-voltage of the interfacial electron transfer from cadmium-zinc-sulfide NPs to the ZnO scaffold. Accordingly, the dependence of log(i/Aint) on the energy gap between a donor level (ECB of sensitizer NPs or the related ECB(x) − ECB(CdS) difference) and an acceptor level (ECB of ZnO scaffold) is linear (Fig. 4.15b). The linear Tafel dependences are typical for systems where no barrier exists for the interfacial charge transfer and the transfer rate is determined predominantly by the difference between the donor and acceptor level energies. In the cadmium-zinc-sulfide-based photoanodes a photogenerated electron migrates from

Fig. 4.15 a Photocurrent density Jph (curve 1) and Jph normalized to the total absorbance of ITO/ ZnO/CdxZn1−xS ﬁlms at k > 400 nm (curve 2) as a function of the composition-dependent conduction band potential ECB(x)ofCdxZn1−xS NPs; b ECB(x) − ECB(CdS) versus logarithm of the normalized photocurrent density 190 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

CdxZn1−xS NPs to ZnO NPs to ITO, thus decreasing its energy. By using reported data about the free electron energy versus the vacuum level for CdS, ZnO, and ITO [16, 112–115], the conduction band potentials of CdS and ZnO in 0.01 M Na2S aqueous electrolyte can be estimated as −0.9 V (versus NHE) and −0.6 to −0.5 V (NHE), respectively, while the accepting level of ITO is estimated to be at around −0.1 to 0 V (NHE). Therefore, in the TIO/ZnO/CdS system there exists a favorable thermodynamic band alignment for the cascade electron transfer from CdS to ZnO to ITO and the efﬁciency of this process is expected to increase as ECB becomes more negative with an increase in the Zn(II) content, in accordance with the above-discussed experimental results. The CdSe NPs can also be deposited by SILAR, typically under an inert atmosphere to avoid the photoanode contamination with elemental selenium. The 2− Se ions come from SeO2 reduced in situ by NaBH4 [90, 116] or directly from Na2Se [96]. Similarly to CdS, an increase in the SILAR cycle number results both in the growth of CdSe absorbance and in a decrease of the average Eg of the deposited CdSe NPs. For example, the bandgap of CdSe NPs deposited on the mesoporous TiO2 decreases from *2.5 eV to around 1.8 eV as the SILAR cycle number is elevated from 3 to 8–10 (Fig. 4.16a). Using a well-established correlation curve between Eg and average size of CdSe NPs the size of CdSe NPs can be estimated to be around 2.6 nm at N = 3, increasing to *8 nm for N = 7, and to higher size values at the further repetition of the SILAR deposition cycles. To protect CdSe NPs during the PEC experiments, the photoanode was covered by a ZnS shell, also using simple SILAR technique with a small (1–2) number of cycles [116]. Such a thin ZnS layer prohibits the photocorrosion of CdX NPs as

Fig. 4.16 a Bandgap and average size of cadmium selenide NPs in TiO2/CdSe heterostructure as a function of the SILAR cycle number N (plotted basing on data reported in [90]); b–e TEM images of TiO2 crystals before PbSe deposition (b) and after (c), 2 (d), and 3 (e) SILAR cycles of PbSe deposition. The scale bar is 5 nm. The red figures is a Pb-to-Se ratio as determined by the energy-dispersive X-ray spectroscopy. Reprinted with permissions from Ref. [117]. Copyright (2012) American Chemical Society 4.4 Nanocrystalline Photoanodes Produced by the In Situ … 191 well as the recombination of the photogenerated charge carriers with the electrolyte species as discussed in details below in this chapter. The less-soluble PbSe NPs can be deposited using sodium selenosulfate as a Se2 − source that is quite stable on air in the absence of metal ions allowing for a simple and reliable SILAR procedure to be performed in the ambient conditions [117]. Surprisingly, even a single SILAR cycle produced well defined 2–3 nm PbSe NPs randomly distributed over the surface of mesoporous TiO2. As the cycle number is elevated to 3 the NP size increases to 4–5 nm and the PbSe NPs become enriched with Pb (Fig. 4.16b). Frequently, the SILAR deposition is followed by a heat treatment, either to promote the crystallization of the NP deposit or to induce a chemical transformation of the deposited precursors. For example, CuxTe NPs can be produced on the mesoporous TiO2 by a thermal treatment of a copper(II) tellurite layer deposited by the SILAR [118]. Annealing of a layered ZnO/CuS/Sb2S3 heterostructure results in the copper(II) reduction with sulfide ions and simultaneous formation of ternary CuSbS2 NPs [102]. Similarly, the thermal treatment of two separate metal sulfide layers pre-deposited by the SILAR on the mesoporous titania yields AgSbS2 [101] and AgBiS2 NPs [99]. Electrodeposition of sensitizer NCs. The electrodeposition methods are based on the electrochemical reduction of chalcogenide precursors resulting in the release of X2− anions (S2−,Se2−,Te2−). The chalcogenide then interacts with metal ions adsorbed on the surface of a wide-bandgap scaffold which serves as a working electrode. The method is typically fast and can potentially be applied to prepare a broad variety of metal sulfide, selenide, and telluride NPs. For example, CdS NPs can be electrochemically deposited from hot aqueous electrolytes or water/DMSO mixtures containing Cd(II) nitrate, and thiourea or elemental sulfur. The electrodeposition is typically performed under the galvanostatic control and elevated temperature (around 90 °C). In this way, CdS NPs were successfully electrodeposited onto the surface of ZnO NRs (Fig. 4.17a–c) [119, 120], ZnO NTs [121], and hierarchical TiO2 microspheres [122]. Cadmium selenide and telluride NPs were electrodeposited in a similar way by using Na2SeSO3 [122, 123] and K2TeO3 [124] as selenium and tellurium sources, respectively. As a rule, the electrodeposition yields relatively large NPs in the form of dense NP layers with a thickness of several tens of nanometers showing bulk-like bandgaps [119–121, 125]. In cases of oxide scaffolds with a largely anisotropic morphology, like NW or NT arrays, a homogeneous distribution of the electrodeposited metal-chalcogenide NPs can be achieved by applying frequency-controlled electrodeposition regimes. For example, the electrodeposition of CuInS2 NPs onto ZnO NR arrays (Fig. 4.17d–f) by using square current pulses with a frequency of 1 kHz produces a relatively smooth CIS layer distributed evenly from the top of ZnO NR down to the NR contact site with the FTO surface (Fig. 4.17f) [125]. At the same time, with a 1 Hz pulse or under the continuous electrodeposition the CIS layer deposits predominantly on the ZnO NR tops forming a dense layer that hinders the electrolyte penetration to the inter-NR space (Fig. 4.17e). The 192 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

Fig. 4.17 SEM (a, b) and TEM (c) images of a ZnO NR tip (a), ZnO NR tip with electrodeposited CdS (b) and a fragment of the ZnO/CdS NR heterostructure (c). d–f SEM images of ZnO NR array (d) and ZnO NR/CuInS2 heterostructures produced by the electrodeposition with a pulse frequency of 1 Hz (e) and 1 kHz (f). Reprinted with permissions from Refs. [120](a–c) and [125](d–f). Copyright (2012) Elsevier (a–c) and (2015) The Royal Society of Chemistry (d–f) reason for such morphology of the CIS layer is the depletion of the inter-NR space with the reactants at the continuous/quasi-continuous deposition resulting in the preferred CIS NP deposition on the outer border of the NR array. By applying a 1 kHz pulses the rates of the CIS NP growth and the diffusion of fresh portions of reactants to the ZnO NR surface can be equilibrated favoring to the formation of a smooth sensitizer NP layer. It should be noted, however, that the optimal frequency for the electrodeposition is unique for a given scaffold morphology and electrolyte composition [125]. The electrodeposition is one of the most frequently used methods for the formation of Cu2O NPs both on bare conducting substrates and on the surface of oxide scaffolds. Typically, the electrodeposition is performed from alkaline solutions containing Cu(II) complexes with lactic acid anions [126, 127]. Chemical Bath Deposition (CBD) of sensitizer NPs. The CBD is also a relatively simple deposition method requiring the immersion of a wide-bandgap substrate into a hot bath containing metal and chalcogen precursors. The X2− anions are 4.4 Nanocrystalline Photoanodes Produced by the In Situ … 193 slowly released as a result of the hydrolytic (solvolytic) decomposition of the chalcogen precursor, while the metal ions are bound by a complexing agent to prevent rapid formation of a deposit and to level off the rates of X2− release and precipitate formation. The metal sulfide NP deposits are typically produced using Na2S2O3, thiourea or thioacetamide as sulfur sources that can be slowly decomposed in alkaline media in the ambient conditions. Similarly to the SILAR, the CBD deposition can be used for the preparation of complex ternary metal sulfide-based heterostructures. In particular, the CBD from aqueous acidic solutions of copper(II) and bismuth(III) nitrates in the presence of Na2S2O3 can be used to produce meso-TiO2/CuBiS2 heterostructures [128]. A solvothermal treatment of a trilayer meso-TiO2/Ag2S/ In2S3 heterostructure, where the indium sulfide layer was deposited by the CBD using thiourea as a sulfur source results in TiO2/AgInS2 composites. Metal selenides can be deposited by the CBD using sodium selenosulfate as a Se2− source that can be easily prepared by dissolving elemental Se in hot aqueous Na2SO3 solutions. Typically, to deposit uniform layers of CdSe NPs onto the wide-bandgap scaffold a thin layer of “seed” CdS NPs is preliminarily formed by the SILAR [129–132]. Then the TiO2/CdS (ZnO/CdS) heterostructure is immersed into an alkaline (pH 11–12) aqueous solution containing Cd(II), nitrilotriacetic acid as a Cd(II) complexing agent and Na2SeSO3 at room or lowered temperature resulting in the growth of a uniform CdSe NP layer. The deposited CdSe NP layer thickness can be controlled by varying the CBD duration. As an example, a variation of the CDB time from 5 to 50 h can be used to tailor the thickness of a CdSe layer deposited on TiO2/CdS heterostructure from 20 to 180 nm [132]. This effect was used [132] to probe the light-to-current conversion efficiency of multi-layer CdSe NP coverings on the surface of compact titania. The compact scaffolds were chosen to minimize possible distortions of the multi-layer uniformity caused by the curvature of TiO2 mesopores. It was found that Jsc generated by such multilayer TiO2/CdS/CdSe heterostructures increases as the CdSe layer thickness is increased to around 100 nm as a result of an enhancement of the light absorbance by the photoanode (Fig. 4.18a, red bars). At higher thicknesses, however, the photocurrent generation efficiency drops most probably due to an increase in the distance the photogenerated carriers need to pass before they can be collected by the TiO2 layer. The normalization of Jsc to the CdSe layer absorbance showed that, indeed, for the film thickness of 20–100 nm, the photocurrent generation efficiency is almost the same, while lowering noticeably for higher thicknesses of the CdSe layer (Fig. 4.18a, blue bars). This result illustrates a fundamental difference between the dye-sensitized and NP-sensitized PEC solar cells—the maximal efficiency of the DSSCs is typically observed at a monolayer coverage of the TiO2 surface with a dye-sensitizer, while in the SSSCs a multilayer absorber can be used and thus the light harvesting efficiency can be increased dramatically. Another important conclusion drawn from the results of [132] was that the SSSCs do not necessarily need a mesoporous wide-bandgap scaffold with a highly developed surface area, as is the case for DSSCs. Using relatively compact TiO2 194 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

Fig. 4.18 a Photocurrent density JSC and normalized JSC as functions of the CdSe NP layer thickness on the TiO2 ﬁlm surface; b Illustration of an optimal TiO2 scaffold structure for SSSCs. Adapted (a) and reprinted (b) with permissions from Ref. [132] Copyright (2012) American Chemical Society

films composed of around 120 nm titania crystals the authors of [132] obtained a much higher light conversion efficiency than in the case of the mesoporous titania. The calculations performed in [132] basing on an optimized CdSe NP absorber thickness of 100 nm and the maximal TiO2 layer thickness providing 100% collection of all injected charge carriers, 4 lm, showed that the best hypothetical configuration of the titania scaffold for the SSSCs is a periodic array of vertically aligned hexagonally packed TiO2 NRs with a diameter of *80 nm and a distance between the neighbouring NRs of *250 nm (Fig. 4.18b). Such a layer can simultaneously serve as a light scattering layer further increasing the light harvesting efficiency. Photochemical deposition of sensitizer NPs. The most papers on the photocatalytic deposition of narrow-bandgap NPs onto wider-bandgap photoactive scaffolds are focused on metal-sulfide NPs. The metal sulfide photodeposition is typically achieved via the photocatalytic decomposition of sulfur-containing metal complexes or via the photocatalytic reduction of elemental sulfur [42]. The former approach can be exemplified by the formation of TiO2/MoS2 and TiO2/WS2 heterostructures via the photocatalytic decomposition of (NH4)2MoS4 and (NH4)2WS4 complexes on the surface of nanocrystalline titania (Fig. 4.19a) [133, 134]. The photoprocess involves the central ion reduction with the photogenerated titania CB electrons followed by the deposition of nanocrystalline MoS2 or WS2. The photocatalytic reduction of elemental sulfur in ethanol solutions in the presence of the nanocrystalline ZnO or TiO2 and a corresponding metal salt was used for the preparation of colloidal ZnO/CdS [135] and ZnO/ZnS NPs [136], as well as composite films of TiO2/PbS [137–139] (Fig. 4.19b), TiO2/CdS [138, 140–142] (Fig. 4.19c), TiO2/Ag2S[143, 144] (Fig. 4.19d), and TiO2/CuxS[138]. When a 0 TiO2/Au nanostructure is used as a photocatalyst, a ternary TiO2/Au/CdS 4.4 Nanocrystalline Photoanodes Produced by the In Situ … 195

Fig. 4.19 TEM/HRTEM images of the photocatalytically deposited MoS2 NPs (a), PbS NPs (b), CdS NPs (c), and Ag2S NPs (d) on the nanocrystalline titania surface. Reprinted with permissions from Refs. [134](a), [137](b), [140](c), and [143](d). Copyright (2011 (a, b, d) and 2009 (c)) Elsevier (a), The Royal Society of Chemistry (b), and American Chemical Society (c, d) heterostructure can be easily produced where cadmium sulfide is deposited as a thin, 1–2 nm, layer on the surface of gold nanocrystals [145] (Fig. 4.20a, b). The size of photodeposited metal-sulfide NPs can be controlled by varying the photocatalytic reaction conditions, in particular, the illumination intensity and duration, the reactant concentrations, the composition and morphology of the photocatalyst, etc. [42, 139, 144–146]. For example, by changing the duration of the photocatalytic CdS deposition on TiO2/Au heterostructure the size of CdS shells grown on the Au cores can be varied in a broad range (Fig. 4.20a) [145]. The size of photodeposited NPs can be also tuned by introducing a stabilizer that restricts the growth of the photoreaction product [42, 139]. In particular, by decreasing the MAA concentration from 0.04 M to zero the average size of photocatalytically deposited PbS NPs can be increased by more than an order of magnitude—from around 5 to 70 nm [139]. The introduction of MAA during the photodeposition of cadmium sulfide on the CdS nuclei (pre-deposited by the SILAR) reduced the average size of final CdS NPs from 6 to 4 nm [142]. A detailed study of a TiO2/Au/CdS heterostructure produced by the photocatalytic deposition [145] showed that cadmium sulfide is deposited as a thin layer predominantly on the surface of Au NPs resulting in Au/CdS core/shell composites (Fig. 4.20b). This phenomenon was explained by efficient separation of the photogenerated electrons and holes between the Au NPs and the nanocrystalline titania support, respectively [145]. At that, the reduction of sulfur (or cadmium) with the 196 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

Fig. 4.20 a TEM images of TiO2/Au/CdS heterostructure produced at a different duration of the photocatalytic CdS deposition tp; b HRTEM image of a Au/CdS core/shell NP on titania surface; c–e Atomic force microscopic images of TiO2/CdS (c, d) and TiO2/CuxS(e) composite films produced by the photocatalytic deposition (c, e) and CBD (d). Reprinted with permissions from Refs. [145](a, b) and [138](c–e). Copyright (2006 (a, b), 2009 (c–e)) Nature Publishing Group (a, b) and Elsevier (c–e) photogenerated electrons takes place mostly on the surface of Au NPs resulting in the formation of a CdS shell. Similar effects of the charge carrier separation were supposed to account for the formation of spatially-organized ZnO/CdS [135, 146] and TiO2/CdS [138] heterostructures, where CdS is present as CdS NTs [135, 146] or NRs [138] (Fig. 4.20c). When cadmium sulfide starts to deposit on the oxide surface, a ZnO– CdS (TiO2–CdS) heterojunction forms where the photogenerated electrons and holes can be spatially separated. As the oppositely charges carriers are attracted to each other, the photoinduced redox reactions occur predominantly at the oxide-CdS interface resulting in the growth of new CdS NPs at the same place and geometrical environment yielding ordered NTs and NRs. As a result, the morphology of photocatalytically produced TiO2/CdS composites differs drastically from the morphology of similar heterostructures synthesized by the conventional CBD procedure (Fig. 4.20c, d). Also, no effects of spatial organization of the photodeposited NPs were observed when the deposited metal-sulfide is photochemically-passive and cannot supply the photogenerated charge carriers to the oxide-sulfide heterojunction as, for example, in the case of the photodeposited TiO2/CuxS heterostructures (Fig. 4.20e). The mechanism of metal-sulfide (MS) NPs deposition on the surface of metal-oxides (M/Os) via the photocatalytic sulfur reduction in ethanol can be presented as follows [42]: 4.4 Nanocrystalline Photoanodes Produced by the In Situ … 197 ÀÁ = ! = À þ þ ; M O+hm M O eCB hVB þ ! þ 2hVB +CH3CH2OH 2H +CH3CHO,

À þ ðÞ!¼ ... 2À; ð : Þ 2eCB Sx x 1 8 SxÀ1 +S 4 8

2 þ þ À ! 0; ð : Þ M 2eCB M 4 9

M2 þ +S2À ! MS, M0 +S0 ! MS:

The photoprocess can proceed by two routes—via the direct reduction of sulfur to S2− with the photogenerated CB electrons [an “ionic” route (8)] or, alternatively, via the reduction of metal cations to M0 [“atomic” route, (9)], or via both routes simultaneously [42]. The feasibility of the photocatalytic deposition for the formation of SSSC photoanodes was first shown for the TiO2-based systems [137, 141, 143, 144]. Recently, nanocrystalline ZnO/CdS heterostructures were also shown to be quite efficient photoanodes of the liquid-junction SSSCs [147, 148]. In both cases, the photoanodes produced by the photocatalytic deposition revealed an increased PEC activity as compared to similar heterostructures formed using the ex situ synthesized CdS NPs or by the SILAR, as shown in Table 4.4 on the example of titania-based SSSCs [141]. Figure 4.21 shows some time-resolved PEC responses from the TiO2/CdS and ZnO/CdS heterostructures produced by the SILAR and photocatalytic deposition and having similar composition and optical properties [147, 148]. The illumination of ITO/TiO2/CdS or ITO/ZnO/CdS photoanodes immersed into aqueous 0.01 M Na2S electrolyte by the “white” light with k > 400 nm results in a rise of photovoltage and photoinduced current between the photoanode and a Pt counter electrode. The photovoltage is roughly the same for the ITO/TiO2/CdS films produced by both methods (Fig. 4.21a) which is expected for the systems with the similar chemical composition. At the same time, the sensitization of both TiO2 and ZnO via the photocatalytic deposition of CdS NPs results in much higher photocurrent densities as compared to the SILAR-produced analogs (Fig. 4.21b, c). In the case of ITO/TiO2/CdS, the absorption-normalized photocurrent density generated by the photochemically produced anode is 5 times higher than for the SILAR-produced heterostructure (Fig. 4.21b), while in the case of ITO/ZnO/CdS the photodeposited cadmium

Table 4.4 Performance of SSSCs based on TiO2/CdS photoanodes produced by different methods [141] −2 CdS NP deposition method Jsc,mAÂ cm Voc,V FF η (%) Photocatalytic deposition 6.5 0.7 0.7 2.5 SILAR 2.7 0.7 0.7 1.2 Adsorption of ex situ synthesized CdS NPs 0.5 0.6 0.6 0.15 198 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

Fig. 4.21 Temporal changes of voltage (a) and current (b, c) induced by the illumination (“hv on” moments)/extinction (“hv off” moments) registered for ITO/TiO2/CdS (a, b) and ITO/ZnO/CdS (c) photoanodes produced by the SILAR (curves 1) and photocatalytic deposition (curves 2). Curve 3 (c) corresponds to the bare ITO/ZnO film sulfide NPs reveal twice as high efficiency as CdS NPs formed by SILAR (Fig. 4.21c). The photoexcitation of non-sensitized ZnO films does not produce any appreciable photo-response (Fig. 4.21c, curve 3). The observations show that the metal oxide—cadmium sulfide heterostructures produced by the photocatalytic deposition of CdS NPs are capable of more efficient spatial separation of the photogenerated electrons and holes between the components as compared to the products of SILAR procedure. The conclusion is strongly supported by the results of a comparative time-resolved laser photolysis study of TiO2/CdS composites produced by the photodeposition and CBD [149] and discussed in details in Chap. 6. The report showed that primary separation of the photogenerated charge carriers and formation of intermediates—Ti3+ in the nanocrystalline titania (a trapped electron) and S•− in the CdS NPs (a trapped hole) occurs by an order of magnitude more efficient for the photochemically-formed TiO2/CdS as compared to the analog produced by CBD (the atomic force microphotographs of both heterostructures are depicted in Fig. 4.20c and d, respectively). The photocurrent density generated by the illuminated ZnO/CdS heterostructures increases in a direct proportion to the amount of sensitizer NPs, which, in turn, depends on the SILAR cycle number N and the photodepositon duration (Fig. 4.22a). However, after the normalization to the light absorbance, the photoanodes differing in the CdS NP content show more or less the same efficiency of light conversion (Fig. 22b). As shown earlier in many examples, as well as in [147, 149] for ITO/ZnO/CdS heterostructures, an increase in the SILAR cycle number results in a considerable growth of the CdS NP size. The results presented in Fig. 4.22b show, therefore, that the size factor is of low importance for the ZnO/CdS heterostructures and the overall light conversion efficiency is affected rather by the sensitizer content than by the NP dimensions. The reason for the lack of size 4.4 Nanocrystalline Photoanodes Produced by the In Situ … 199

Fig. 4.22 Dependence of the photocurrent density (a) and absorbance-normalized photocurrent density (b) generated by the ITO/ZnO/CdS photoanodes on the SILAR cycle number (curves 1) and the photodeposition duration (curves 2) dependence can be in very favorable conditions for the spatial charge separation that exists even in the bulk ZnO–CdS and TiO2–CdS heterojunctions and should be even more advantageous in the case of nanocrystalline semiconductors. The TiO2/Ag2S nanoheterostructures produced by the photocatalytic silver sulfide deposition were successfully tested as photoanodes for the hydrogen-evolving PEC solar cells [143]. The bandgap of Ag2S NPs was found to become narrower with an increase in the photodeposition duration indicating an increase in the silver sulfide NP size. The highest efficiency of the solar light harvesting was observed for Ag2S NPs with Eg = 1.75 eV [143]. The silver sulfide NPs can also be produced by the sulfidation of Ag NPs deposited + via the photocatalytic Ag reduction. The TiO2/Ag2S films produced by this method from titania NTs showed a light conversion efficiency of 1.23% when applied as a photoanode in a liquid-junction SSSC with the polysulfide electrolyte [144]. The TiO2/PbS heterostructures can be prepared by the photocatalytic lead sulfide deposition from ethanol solutions of lead perchlorate and S8 [137]. Such films showed a photocurrent density of 1.71 mA/cm2 (at 0 V vs. Ag/AgCl) when illuminated by the AM1.5 light in aqueous solutions of Na2S and Na2SO3. Additional examples of SSSCs based on the photocatalytically-produced photoanodes are presented in Table 4.5. The UV-illumination of titania films immersed into degassed ethanol solutions of S8 and SbCl3 results in the deposition of amorphous antimony sulfide, the amount of Sb2S3 growing with an increase in the photodeposition duration [150]. The annealing in N2 atmosphere yields crystalline TiO2/Sb2S3 heterostructures that can be used as visible-light sensitive SSSC components. 200 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

Table 4.5 Some examples of SSSCs produced by the photodeposition of sensitizer NPs Photoanode CE I,mWÂ cm−2 Redox-couple species η (%) Reference

TiO2/PbS Pt 100 Na2S/Na2SO3 0.16 [137] − TiO2/CdS Pt 100 I /I2 2.51 [141]

TiO2/Ag2S Pt 100 Na2S/Na2SO3 0.29 [143] 2− 2− TiO2 NR/Ag2S Pt 100 S /Sx 0.19 [144] 47 1.27 2− 2− ZnO NR/CdS/CdSe CuxS 100 S /Sx 2.03 [241]

The amorphous antimony sulfide is deposited in the form of spheroidal particles with a size of 150–300 nm making Sb2S3 NPs clearly visible in the background of much smaller titania nanocrystals (Fig. 4.23a, b). According to the energy-dispersive X-ray spectroscopy (EDX) analysis, the atomic Sb-to-S ratio, 1:1.3, is close to the expected stoichiometric value. The TiO2/Sb2S3 films retain their morphology after the annealing (Fig. 4.23c). Along with the nanometer-sized deposits, some much larger spherical particles can be observed on the surface of TiO2/Sb2S3 films at a lower SEM magnification

Fig. 4.23 SEM images of starting nanocrystalline TiO2 film (a) and TiO2/Sb2S3 (b–d) films with amorphous (b) and crystalline Sb2S3 (b). Image (d) was taken at a lower magnification. e SEM image of photodeposited metallic Sb; f Photocurrent density Jphoto at 0.1 V versus Ag/AgCl for FTO/TiO2 and FTO/Sb2S3 electrodes. Photodeposition time is indicated on the bars. Reprinted from Ref. [150] with permissions. Copyright (2015) Elsevier 4.4 Nanocrystalline Photoanodes Produced by the In Situ … 201

(Fig. 4.23d). The EDX analysis of such spheres showed them to be strongly enriched with antimony (Sb:S is 13:1 in the bulk of the spheres and 5:1—on their surface). Also, the photodeposition in the absence of sulfur produced similar but strongly aggregated microparticles of metallic antimony (Fig. 4.23e). The presence of such particles in the photodeposited TiO2/Sb2S3 composites can, therefore, be taken as an indication that the photocatalytic deposition of antimony sulfide proceeds, most probably, via the atomic route of Sb0 formation followed by the antimony sulfidation with S8. Because of the ready dissolution of antimony sulfide in the polysulfide electrolytes, the TiO2/Sb2S3 composites require alternative electron-shuttling couples to be used or, alternatively, a sacrificial electron donor that is consumed irreversibly supplying electrons to the photoanode, such as ascorbic acid [150]. The light conversion efficiency of the composites with crystalline antimony sulfide under the illumination with the “white” light in aqueous solutions of ascorbic acid grows with an increase in the sensitizer content, that is with an increase in the photodeposition duration (Fig. 4.23f). At the same time, the amorphous antimony sulfide deposits revealed no photoactivity even showing an adverse light-shielding effect on the titania. Preparation of SSSC photoanodes by ions exchange. The ion exchange (IE) is a quite straightforward method for the formation of various metal oxide/metal chalcogenide nanoheterostructrures, that is typically applied to chemically unstable zinc oxide. Both Zn2+ ions and O2− anions in the ZnO lattice can be substituted by other metal cations and chalcogenide anions, respectively, producing less soluble metal chalcogenides. For example, the nanocrystalline ZnO films immersed into 2− aqueous solutions of Se -generated species (Na2SeO3 + NaBH4) gain yellow color indicating the formation of zinc selenide [151]. The extinction of ZnO/ZnSe films in the visible spectral range increases with an increase of the ZnO layer thickness (proportional to the electrodeposition duration of original zinc oxide films [151]) as more and more ZnO is converted into zinc selenide (Fig. 4.24a). The presence of a ZnO-related shoulder in the extinction spectra of ZnO/ZnSe heterostructures indicates a partial character of the ion exchange. An analysis of the spectral curves showed that the bandgap corresponding to the new spectral feature in the visible range is 2.74–2.75 eV, which is typical for bulk hexagonal zinc selenide [152, 153]. The partial transformation of ZnO into ZnSe does not induce appreciable change in the morphology of zinc oxide films (Fig. 4.24b, c). An EDX analysis of the film cross-section (Fig. 4.24d) showed that selenium is distributed evenly in the volume of the films (Fig. 4.24e) mimicking the distribution of zinc and oxygen (Fig. 4.24f, g) and indicating that the entire ZnO film is accessible for the IE reaction. The IE procedure can be repeated again producing ternary and more complex oxide/chalcogenide nanocomposites. For example, ZnO nanowires (NWs) were converted into ZnO/ZnSe heterostructures by a partial IE of oxygen (Fig. 4.25(I)a, b), then Zn2+ ions in the ZnSe layer were partially substituted with Cd2+ thus producing ternary ZnO/ZnSe/CdSe nanocomposites retaining the wire-like shape of the original zinc oxide NWs (Fig. 4.25(I)c–e) [43, 154]. 202 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

Fig. 4.24 a Extinction spectra of ZnO (curve 1) and ZnO/ZnSe (curves 2–5) films produced by the IE from zinc oxide films electrodeposited during 2 min (curve 2), 4 min (3), 6 min (4), and 10 min (5). See deposition conditions in [151]; b–d SEM images of ZnO/ZnSe film, e–g atom distribution maps for Se (e), Zn (f), and O (g) produced by the EDX analysis [151]. The size of images in (d–g) is around 10 Â 10 lm

Fig. 4.25 I SEM images of (a) a ZnO NW array, b ZnO/ZnSe NW, c, d Zn0.7Cd0.3Se NWs prepared by reacting a ZnO/ZnSe nanocable with Cd2+ at 50 °C (c), 90 °C (d), and 140 °C (e). II Diffused reﬂectance spectra of (a) as-prepared ZnO NR; b CdS/ZnO nanocable, c ZnO/CdS0.61/ CdSe0.39 (50 °C) nanocable, and d ZnO/CdS0.24/CdSe0.76 (90 °C) nanocable arrays; III Scheme of the fabrication process of ZnO/AgInS2 NR arrays [158]. Reprinted with permissions from Refs. [43](I), [155](II), and [158](III). Copyright (2011–2014) American Chemical Society

The morphology of heterostructures and amount of the incorporated cadmium selenide depend on the temperature of the ZnSe conversion into CdSe [43]. By substituting O2− in ZnO nanocables with sulﬁde ions ZnO/ZnS composites were produced, which then were transformed into ZnO/CdS by exchanging Zn2+ with 4.4 Nanocrystalline Photoanodes Produced by the In Situ … 203

Cd2+ (Fig. 4.25(II), curves a, b). Finally, sulfur in CdS was partially substituted with selenium resulting in a considerable red shift of the absorption edge of the prospective ZnO/CdSxSe1−x photoanode (Fig. 4.25(II), curves c, d) due to the formation of an alloyed cadmium sulfoselenide layer [155]. The IE transformation of ZnO is often used to form a thin blocking zinc chalcogenide layer preventing the charge leakage and the recombination in the photoanodes prepared by the following deposition of ex situ synthesized sensitizer NPs. In this way, ZnS and ZnSe blocking layers were formed on the ZnO surface prior to the deposition of the visible-light-harvesting Cu2ZnSnS4 NPs [156]. The ion exchange methods were probed for the preparation of Cd- and Pb-free photoanodes of the liquid-junction SSSCs. For example, by substituting Zn2+ in + ZnO/ZnS heterostructures by Ag the ZnO/Ag2S composites can be produced, 3+ which can then be subjected to a partial IE with Sb to produce ZnO/AgSbS2 3+ heterostructures [157]. In a similar way, by introducing In into ZnO/Ag2S heterostructure produced by the IE, visible-light-sensitive ZnO/AgInS2 photoanodes were formed (Fig. 4.25(III)) [158].

4.5 Making Progress in SSSCs—Toward More Efﬁcient and Less Toxic Photoelectrodes

The photoelectrochemical SSSCs are nowadays in a constant progress and steady efforts are applied to increase their efficiency in the solar light harvesting and their attractiveness as compared with competing photovoltaic technologies. These efforts can be categorized into several main trends that will be discussed in this subsection. The photocurrent generation efficiency is limited to a far extent by the electron-hole recombination in the sensitizer NPs as well as on the interfaces between the wide-bandgap scaffold and the electrolyte and between the metal oxide and the sensitizer NPs. To suppress the recombination and minimize losses of the photogenerated charges various approaches are developed, one of the most simple and, at the same time, efficient being the formation of additional “protective” or “buffer” semiconductor layers either on the sensitizer NP surface, or between the sensitizer and the oxide scaffold, or between the metal oxide and OTE [s066, s067, s068, s070]. As discussed in details in Chap. 1, the properties of sensitizer NPs can change dramatically in a critical size range of around 1–10 nm as a result of the quantum size effects. This feature opens possibilities of enhancing the photoresponse of sensitizer NPs, that is increasing the energy and transfer rate of CB electrons and VB holes through variations of the sensitizer NP size. By combining several size-selected sensitizer NPs in a “cascade” structure with an outer NP layer having the highest CB potential and an inner, closest to the scaffold, layer having the lowest CB potential, the directed transport of the photogenerated CB electrons can be organized, from the smaller NPs in the outer NP layer to the larger NPs in the inner layer to the metal oxide scaffold [159–161]. Such a cascade structure allows 204 4 Semiconductor-Based Liquid-Junction Photoelectrochemical … suppressing the recombinative processes as the CB electron gets physically separated from the VB hole and cannot come in the reverse direction through the potential barrier. Similar cascade structures can also be organized by using different sensitizer materials with correspondingly matched CB levels [159–161]. Finally, the attractiveness of SSSCs can be enhanced by utilizing new nontoxic and Earth-abundant sensitizer materials instead of Cd- and Pb-based NPs that dominate today in the photoelectrochemical SSSCs [159, 160, 162]. The low toxicity of cell components is an important issue that can even counterweight the high efficiency of such heavy-metal containing materials as CdSe or PbS. As a vivid example, the organo-inorganic Pb-containing perovskites can be mentioned, such as CH3NH3PbHal3 (Hal = Cl, Br, I). The perovskites became a rapidly rising star of the photovoltaics making progress in the light harvesting efficiency from several percents in 2009 to more than 21% in the recent years [22, 26, 163–171]. However, the organo-inorganic perovskites suffer from chemical and photochemical instability and the problem of possible lead leakage is a grave concern that can impede a broad implementation of the photovoltaic technologies based on such materials despite their high efficiency. In this view, constant efforts are applied in the screening and testing of new narrow-band-gap semiconductor materials, in particular among the more complex ternary and quaternary metal chalcogenides in the hope of finding reasonably efficient, abundant and low-toxic light-harvesting materials [159–162, 172, 173]. Finally, the metal oxide scaffold, though not participating directly in the light harvesting, can strongly influence the rate of secondary charge transfer processes and limit the total light conversion efficiency of a SSSC. Also, the enhanced light scattering from specially designed metal oxide nano-architectures can influence in a positive way the light harvesting efficiency of the cell as a whole. In this view, constant efforts are applied for the design of new morphologies of metal oxide scaffolds favoring to the accommodation of sensitizer NPs and affecting their light absorption. Suppression of the recombination by barrier layers. Charge losses in the photoanodes consisting of the mesoporous wide-bandgap scaffold and sensitizer NPs can originate from several recombination processes, in particular, (i) the electron-hole recombination in the volume of NPs, (ii) the recombination of an electron injected into the metal oxide scaffold with a hole left in a sensitizer NP, and (iii) “leakage” of electrons migrating along the mesoporous network of TiO2 (ZnO) toward OTE as a result of the interactions with the electrolyte that permeates the whole photoanode volume. Each of the above-discussed recombination pathways can be addressed separately by introducing a special barrier layer that hinders the recombinative charge losses but at the same time does not impede the direc- tional electron transfer from the sensitizer NPs to the metal oxide to the OTE. Suppression of the recombination in the sensitizer NPs. The recombination in the sensitizer NPs occurs predominantly via structural defects introducing additional states in the NP bandgap and allowing for the thermal or radiative dissipation of the light excitation energy. As opposite to bulk counterparts, NPs have a much larger surface-to-volume ratio and a lot of structural defects reside on the NP surface. 4.5 Making Progress in SSSCs—Toward More Efficient … 205

These defects (unsaturated bonds, cation/anion vacancies, etc.) can be passivated by appropriate ligands or a shell of other metal chalcogenide semiconductor, typically with a larger bandgap, such as ZnS. The passivating layers can be very conveniently deposited by the SILAR procedure allowing for a quite precise control over the thickness of the barrier layer. For example, the light conversion efficiency of TiO2/PbS heterostructures can be enhanced by the deposition of a CdS layer, both lead sulfide and cadmium sulfide layers deposited by the SILAR [103, 174]. As the SILAR cycle number is increased to 4–5 the photocurrent increases, reaches a saturation value and then decreases because a too thick protective CdS shell exerts a light-shielding effect on the PbS NP sensitizer (Fig. 4.26a) [103]. An additional *60% enhancement of the light conversion efficiency on TiO2/ PbS photoanodes can be achieved by the SILAR deposition of a mixed PbxCd1−xS layer prior to the deposition of the passivating CdS shell [104]. A passivation effect can be achieved by the deposition of a ZnS layer onto CuInSe2 sensitizer NPs anchored to the titania surface [175]. The zinc sulfide layer increases the photostability of sensitizer NPs and shifts the NP absorption edge to longer wavelength allowing for a more efficient harvesting of the solar light. The effect is caused by a partial penetration of the wavefunctions of photogenerated charge carriers from the 4-nm CuInSe2 NPs into the ZnS layer resulting in a weakening of the exciton confinement effect in the sensitizer NPs. In this case, there exists an optimal number of the SILAR deposition cycles (Fig. 4.26b), because a thicker ZnS shell impedes the electron transfer from the electrolyte species to CuInSe2 NPs. A similar passivation effect of a ZnS shell was observed also for CdS/CdSe NPs [176]. CdTe NPs can be passivated with a CdSe shell that contributes to the light absorbance and serves as an electron acceptor resulting in photogenerated electron and hole separation in the CdSe shell and CdTe core, respectively [66].

Fig. 4.26 a Total light conversion efﬁciency of TiO2/PbS/CdS-based (a), TiO2/CuInSe2-based (b) and TiO2/CuInS2-based (c) SSSCs as a function of the SILAR number of PbS and CdS deposition (a) and ZnS deposition (b) as well as the nature of protective shell (c). Reprinted/ adapted with permissions from Refs. [103](a), [175](b), and [107](c). Copyright (2011–2015) American Chemical Society 206 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

By choosing appropriate shell materials and deposition sequence the light conversion efficiency of TiO2/CuInS2-based SSSCs was increased from *1% for the bare sensitizer NPs to *4.5% for a layered CuInS2/CdSe/ZnSe heterostructure (Fig. 4.26c) [107]. A higher efficiency of the ZnSe-based photoanode as compared to the ZnS-based counterpart originates from the enhanced visible light absorbance of zinc selenide contributing to the overall solar light harvesting. Similar effects of increased light absorbance and recombination suppression account for a higher light conversion efficiency of TiO2/CdS/CdSe photoanodes covered with a ZnSe protective layer (η = 6.4%) as compared to a similar heterostructure passivated with a ZnS layer (η = 4.9%) [177]. Alternatively, the surface states of sensitizer NPs responsible for the recombination losses can be passivated by molecular ot ionic species, thus leaving the surface of sensitizer NPs fully open for interactions with redox species in the electrolyte. For example, a passivation effect was observed upon the adsorption of aliphatic amines on CdS NPs, which increases the light conversion efficiency from 1.45 to 2.35% [178]. Adsorption of a layer of 4-tert-butylpyridine on a TiO2/CdS/CdSe photoanode passivated with a ZnS shell adds around 1% to the cell efficiency [176]. However, the amine adsorption prior to the sensitizer NP deposition results in the deterioration of the cell performance due to intervention of the amine layer into the charge transfer between TiO2 and CdS/CdSe NPs. A similar adverse effect on the cell performance has also the adsorption of a layer of electron-withdrawing molecules, such as 4-cyanopyridine [176]. A layer of mercaptophenol deposited onto TiO2/PbSe heterostructures was reported [117] to improve the hole tunneling to the electrolyte from lead selenide NPs thus contributing to the light harvesting efficiency. The recombination of photogenerated charge carriers in TiO2/Sb2S3 heterostructures can be suppressed by the deposition of an outer shell of a conjugated polymer—poly-3-hexylthiophene resulting in an enhancement of the light conversion efficiency from 3.2 to 4.2% [27]. Lead sulfide NPs can be passivated by adsorption of halogenide ions [179, 180], most probably due to the formation of a thin surface layer of lead halogenides. Suppression of the recombination on the interface between the sensitizer NPs and the metal oxide scaffold. Provided the CB levels of the sensitizer NPs and TiO2 (ZnO) are favorable for the electron transfer from NPs to oxide, the sensitizer NP photoexcitation results in the extremely fast electron transfer from NPs to the neighboring TiO2 (ZnO) layer, leaving a hole in the NP valence band. Potentially, the electron in the titania CB can recombine with the hole in the sensitizer VB, similarly that it occurs in the DSSCs after the electron injection from the photoexcited dye-sensitizer to the titania scaffold. To prevent such charge losses a blocking layer is often introduced between the sensitizer NPs and TiO2 (ZnO) constituting a potential barrier for the injected electron on its way back to the parental metal chalcogenide NPs. Similarly to the above discussed passivation of TiO2/CdS photoanodes with an outer ZnS shell, the deposition of an intermediary zinc sulfide layer between titania and CdS or CdxZn1−xS NPs results in an increment of the light conversion 4.5 Making Progress in SSSCs—Toward More Efficient … 207 efficiency from 3.06 to 3.69% as a result of shielding of the titania layer both from the charge leakage to the electrolyte and against the reverse electron transfer to the sensitizer NPs [88]. Improved charge collection in the TiO2/ZnS/CdS/ZnS heterostructure with the inner and outer ZnS protective shells can be clearly exemplified by the electrical impedance spectra (Fig. 4.27a). The larger is the radius of the Nyquist plots for a given photoanode the larger is the electric capacitance Cl on the photoanode/electrolyte interface, which is a quantitative measure of the charge collection efficiency. The Cl values obtained for TiO2/CdS, TiO2/CdS/ZnS, and TiO2/ZnS/CdS/ZnS by a simulation of the impedance spectra with an equivalent circuit were 1516, 2217, and 2586 lF, respectively, indicating on the better charge collection for the double passivated photoanode due to lower electron-hole recombination [88]. A screening search for potential materials for a barrier layer between TiO2 and CuInS2 (CIS) NPs showed that a number of metal chalcogenides can suppress the reverse electron transfer from TiO2 to CIS. These include cadmium, copper and indium chalcogenides, the most efficient being CuxSe and In2Se3 allowing to increase the light conversion efficiency by a factor of 3 and higher (Fig. 4.27b) [107]. The electron-hole recombination suppression by buffer layers was proved unambiguously by an increase of the open-circuit voltage. The highest efficiency of indium selenide barrier layer was explained by a combination of favorable factors, including a proper band alignment (the CB of In2Se3 stays between the CB levels of TiO2 and CIS NPs), formation of an intermediate CuInSxSe1−x layer allowing for a better orientation of crystalline planes of CIS NPs with respect to those of titania NPs, and, finally, by a layered character of indium selenide that “smears” uniformly on the titania surface. Similarly to zinc sulfide, CdS layers can be placed between the titania scaffold and CIS NPs [82] and on top of binary TiO2/CIS heterostructure [52] resulting in both cases in an improvement of the cell performance.

Fig. 4.27 a Electrical impedance spectra of SSSCs with different photoelectrodes in the form of Nyquist-plots. b Light conversion efﬁciency (η) and open-circuit voltage (Voc) for TiO2/CuInS2 photoanodes with different buffer layers placed between TiO2 and CuInS2 NPs. Reprinted/adapted with permissions from Refs. [88](a) and [107](b). Copyright (2016) The Royal Society of Chemistry (a) and (2013) American Chemical Society (b) 208 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

The light conversion efficiency on Ag2S NP-decorated TiO2 NT arrays can be increased from 0.22 to 0.28% by placing a barrier recombination-blocking ZnO layer between titania and sensitizer NPs [106]. Alternatively, a TiO2 NP layer can be inserted between ZnO NR scaffold and CdS/CdSe NPs to suppress the electron leakage from zinc oxide to the electrolyte. This approach results in a spectacular increase of the light conversion efficiency from 1.54 to 3.14% [130]. A Mg-doped ZnO layer placed between the ZnO scaffold and PbS NPs can efficiently suppress the backward electron transfer from zinc oxide to the metal chalcogenide NPs due to a favorable cascade CB level positions of all three components [181]. Suppression of the charge leakage from the sensitizer NPs and the metal oxide scaffold to the electrolyte. After the injection from photoexcited sensitizer NPs to adjacent mesoporous TiO2 (ZnO) layer, the electron migrates through the network of contacting metal oxide NPs till it reaches OTE and comes into the electric circuit. At that, a possibility exists for the electron to be captured by the components of the + electrolyte, for example, by water or H3O ions that can reduce the photoanode performance considerably. To avoid such losses the TiO2/NPs heterostructures are typically covered with an additional protective layer of wide-bandgap materials, most often, zinc sulfide that creates a barrier for the electron to reach the electrolyte. In the case of ZnO-based photoanodes such passivation can occur directly in the polysulfide electrolyte as a result of a partial anion exchange and in situ ZnO transformation into ZnS. A more general approach consists in the formation of a thin ZnS layer by several SILAR deposition cycles. The protective ZnS layer also provides a stronger contact between the sensitizer NPs and TiO2 (ZnO) as well as protects the light-harvesting metal chalcogenide NPs from corrosive processes that can occur during the photoelectrochemical events in the SSSC. In the case of titania scaffolds, the photoanode can be relatively easily “insu- lated” from the charge leakage by soaking with TiCl4, followed by the hydrolysis and annealing. As a result, a thin and evenly distributed layer of TiO2 NPs is formed on the photoanode surface. The titania layer prevents charge transfer to the electrolyte from both the sensitizer NPs and the metal oxide scaffold as well as provides a better contact between the sensitizer NPs and the metal oxide transport layer. This procedure resulted in more than 150% increment of the light conversion efficiency when applied to TiO2/CdS photoanodes [182]. Cascade designs of SSSCs. The cascade design of photoanodes/photocathodes of SSSCs can be realized in several ways, in particular, (i) by using several different semiconductor photo–electrode materials with favorable CB and VB level offsets; (ii) by using alloyed solid solution compounds with a varied or spatially gradient structure to create a CB (VB) offset from the outer surface of the photoelectrode toward OTE, and (iii) by combining NPs of the same semiconductor but of different size that reveal a strong size-dependence of the CB and VB levels and placing them in the order of decreasing CB (VB) potential from the outer photoelectrode surface toward OTE. Cascades of different semiconductor NPs. The formation of a cascade of two and more different semiconductor NPs aimed to produce a descending gradient of the 4.5 Making Progress in SSSCs—Toward More Efficient … 209

CB level (or an ascending gradient of the VB level) is, probably, the most straightforward way of achieving enhanced spatial separation of the photogenerated charge carriers and suppressing their recombination in the sensitizer NPs. Such cascade effects were reported for TiO2/ZnO/CdS [183], TiO2/CdS/CdSe [116, 122], SnO2/TiO2/CdS/CdSe [184], ZnO/CdS/CdSe [185], TiO2/PbS/CdS [103, 104], TiO2/CuInS2/CdS [69, 186], and TiO2/ZnIn2S4/CdS [187]. Typical CB/VB level alignments in some successful cascade systems are presented in Fig. 4.28. A ternary cascade system of PbSe, CdS, and carbon NPs (the so-called carbon dots) was applied as a sensitizer for the titania scaffolds (Fig. 4.28b) [188] resulting in almost 5% efﬁciency of the solar light conversion and the spectral sensitivity extending over 1000 nm.

Fig. 4.28 Energy schemes of some successfully realized cascade photoelectrode designs—SnO2/ TiO2/CdS/CdSe (a), TiO2/ZnIn2S4/CdS (b), TiO2/PbSe/CdS/C (c), and TiO2/CuInS2/CdS (d). Reprinted with permissions from Refs. [184](a), [178], (c)[188](d)[186]. Copyright (2011, 2015). The Royal Society of Chemistry (b–d) 210 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

As a rule, the cascade effect results in a non-additive enhancement of the light conversion efficiency as compared to the sum of efficiencies of the SSSCs based on separate components. For example, in a TiO2/CdS/CdSe cascade the reported light conversion efficiency is much higher (3.2% [116], 4.81% [122]) than the sum of efficiencies in TiO2/CdS-based (0.39% [116], 1% [122]) and TiO2/CdSe-based SSSCs (2.29% [116], 2.69% [122]). Cascades of alloyed metal chalcogenide NPs. Some metal chalcogenide semiconductors, for example, CdS and ZnS, CdS and CdSe, can form solid solutions of any varied composition and, as a consequence, with varied CB and VB positions. By combining several composition-selected sensitizer NPs a cascade structure can be arranged for the directed electron migration from the metal-chalcogenide layer to the metal oxide scaffold and then—to the electric circuit. By changing the Se-to-S ratio in the mixed CdSxSe1−x NPs one can produce multi-colored, green to orange-yellow to red, light absorbers with a varied bandgap (Fig. 4.29a) [65]. Arranging of the CdSxSe1−x NPs by increasing Eg from the TiO2 surface to the outer photoanode zone results in a cascade structure reaching the light conversion efficiency of 3%. The quantitative data on relative band edge positions of titania and composition-selected CdSxSe1−x determined by the UV photoelectron spectroscopy were reported (Fig. 4.29b) [189], allowing for a precise design of such photoanodes. The SSSCs with the composition-selected ZnO/CdxZn1−xSe nanocable array photoanodes revealed the light conversion efficiencies of up to 4.74% [43]. Some synthetic approaches, for example, the temperature-gradient chemical vapor deposition [190], or the photochemical transformation [191], were applied to prepare heterostructures with a continuous gradient structure, where the directed charge flow can be realized as well, for example, ZnO/CdxZn1−xSe [190], ZnO/

Fig. 4.29 a Absorption spectra and photographs (taken under the UV illumination) of colloidal composition-selected CdSxSe1−x NPs; b energy diagram of the TiO2/CdSxSe1−x heterojunctions with different x. Reprinted with permissions from Refs. [65](a) and [189](b). Copyright (2012, 2013) American Chemical Society (a) and The Royal Society of Chemistry (b) 4.5 Making Progress in SSSCs—Toward More Efﬁcient … 211

ZnOxS1−x [191]. Such heterostructures were [190] or potentially can be [191] applied as the photoanode materials in the SSSCs. Cascades of the size-selected metal-chalcogenide NPs. In a similar way, differently-sized NPs of the same semiconductor can be arranged in layers on the oxide scaffold surface in order of increasing Eg enabling the photogenerated electrons to migrate from the outer sensitizer NP layer to the metal oxide transport layer and into the circuit. Such a “rainbow-cell” design was successfully realized for TiO2/CdSe photoanodes with the ex situ produced 2.3–3.7 nm NPs [60, 63] as well as with the in situ deposited CdSe NPs [73]. Differently sized (2.0–4.5 nm) CdTe NPs were also tested in a cascade structure with ZnO NR array scaffolds both as direct sensitizers and as energy relays for the indirect excitation of an additional dye sensitizer [79]. Lox-toxic alternative sensitizer NPs for SSSCs. A search of new narrow-bandgap semiconductor materials capable of efficient light harvesting and their testing as the light-sensitive components of the photoanodes and photocathodes is, probably, the current hot spot in the SSSC field. Each and every new compound that potentially can be used as a light absorber is probed, including new (and sometimes quite exotic) Cd-free and Pb-free metal chalcogenides, ternary and quaternary chalcopyrites and kesterites, emerging materials such as carbon dots and many others. Some of new absorber materials were discussed in the sections devoted to the formation of photoanodes using the ex situ and in situ synthesized sensitizer NPs. This subsection collects recent reports on new (promising and potentially promising) sensitizer materials for the SSSCs that were not scrutinized before in this chapter. Some of the SSSCs examples are summarized in Table 4.6. The most straightforward way to the low-toxic SSSCs is to probe the sensitizers which are similar to the dominating cadmium and lead selenides. In this way, alternative binary chalcogenide sensitizers were studied, including Sb2S3 [27, 150], Bi2S3 [89, 100], FeS2 [192], ZnSe [151, 154, 193] and Sb2Se3 [194]. In the SSSC design with new sensitizer materials, not only a correspondence between the absorption spectrum of the sensitizer NPs and the solar irradiation spectrum should be taken into account, but also a proper alignment of the CB and/or VB levels of the sensitizer NPs and other photoelectrode components, including the metal oxide scaffold and the passivating layers. For example, in the bismuth sulfide-based systems, the deposition of an intermediary CdS layer between Bi2S3 NPs and mesoporous TiO2 film creates a cascade structure (Fig. 4.30a), favoring to the photoinduced electron transfer from the sensitizer NPs into the circuit, while in the TiO2/Bi2S3/CdS structure the sensitizer blocks the electron transfer from cadmium sulfide and thus the efficiency of the light conversion with such photoanode is only 0.56% as compared to 2.52% for the TiO2/CdS/Bi2S3 photoanode [89]. As discussed in the previous subsections, great expectations in the SSSC development are associated with the ternary and more complex metal chalcogenide NPs that combine a high absorptivity in the visible and near-IR ranges with an unprecedented flexibility of properties via the variations in the NP composition, size, shape, doping, etc. [173, 195]. Table 4.6 Some examples of SSSCs based on various Cd- and Pb-free NP sensitizer NPs Photoelectrochemical Liquid-Junction Semiconductor-Based 4 212 −2 Photoanode material Eg, eV CE/redox-couple Jsc,mAÂ cm Voc, V FF, % η, % Reference 2−/0 ZnO/ZnSe/CdS *2.3 CuxS/S 2.29 440 27 0.27 [193] 2+/3+ TiO2/Sb2S3 1.65 Pt/Co 12.0 530 50 3.20 [27] TiO2/Sb2S3/P3HT 12.2 667 51 4.20 2−/0 TiO2/Ag2S 1.0 Pt/S 10.3 290 33 0.98 [144] −/0 ZnO/ZnS/FeS2 n/r Pt/I 0.87 390 36 0.12 [192] −/0 TiO2/BiOI 1.85 Pt/I 1.52 490 51 0.38 [242] 2−/0 TiO2/CuInS2/CdS *1.6 CuxS/S 16.9 560 45 4.20 [243] −/0 ZnO/AgInS2 1.8 Pt/I 3.8 540 35 0.72 [196] 2−/0 TiO2/AgInS2 1.8 Au/S 4.62 450 39 0.80 [197] 2−/0 TiO2/AgInS2/In2S3 *1.7 Pt/S 7.87 320 28 0.70 [198] 2−/0 TiO2/CuInSe2/ZnS 1.22 CuxS/S 26.93 528 57 8.10 [175] 2−/0 TiO2/CdS/CuInS2 *1.5-1.6 C/S 8.12 489 37 1.47 [82] 2−/0 TiO2/AgSbS2 1.7 Au/S 2.42 320 n/r 0.34 [101] 2−/0 TiO2/CuBiS2 2.1 CuxS/S 6.87 250 36 0.62 [128] 2−/0 TiO2/AgBiS2 1.32 Pt/S 7.61 180 39 0.53 [99] −/0 TiO2/Cu2ZnSnS4 n/r Pt/I 0.41 560 58 0.13 [72] 2−/0 ZnO/ZnSe/Cu2ZnSnS4 1.5 CuxS/S 10.46 490 43 2.20 [156] 2−/0 TiO2/CuInS2:Zn n/r CuxS/S 20.65 586 58 7.04 [71] 2−/0 TiO2/CuInSe1.4 S0.6 n/r CuxS/S 10.5 550 60 3.45 [199] 2−/0 TiO2/CuInS2/CdS *1.8 CuxS/S 15.65 529 47 3.91 [52] 2−/0 TiO2/CuInS2 CuxS/S 12.82 640 54 4.20 [83] TiO2/AgInS2 9.75 432 65 2.62 Note 1 sun (AM1.5) if not stated otherwise; with outer ZnS layer typically … 4.5 Making Progress in SSSCs—Toward More Efficient … 213

Fig. 4.30 Schematic energy diagrams of TiO2/CdS/Bi2S3 and TiO2/Bi2S3/CdS heterostructures (a), and TiO2/In2S3/Ag-In-S/In2S3/ZnS photoanode (b). Reprinted with permissions from Refs. [89](a) and [198](b). Copyright (2015) Elsevier (a) and The Royal Society of Chemistry (b)

For example, the ternary CuInS2 and AgInS2, and corresponding non-stoichiometric CIS and AIS compounds are used in a constantly broader manner [52, 53, 69–71, 82, 83, 100, 125, 196, 197], tending to gradually substitute CdS and CdSe in the SSSCs. In the case of CIS(AIS)-based photoanodes formed by the attachment of ex situ synthesized sensitizer NPs on the surface of metal oxide scaffolds the size of NPs can be finely tuned by adjusting the synthesis conditions and so the size dependences of the light conversion efficiency can be conveniently probed. In the case of CIS/AIS NPs, similarly to other earlier discussed examples, such dependences are typically volcano-shaped (Fig. 4.31a, blue bars) as a result of a counter-balance of the CB level increase and a blue absorption band edge shift both observed with a NP size decrease (Fig. 4.31b), the latter effect resulting in a partial loss of the solar light absorption [51]. Typically the size-dependence of η mimics the size variation of PL emission intensity (Fig. 4.31a, red bars) because both the light emission and the photocurrent generation compete with the non-radiative electron-hole recombination. Therefore, the PL spectroscopy can be used as a diagnostic tool for assessing/predicting the PEC activity of CIS/AIS-based photoanodes as discussed later in Chap. 6. The efficiency of AIS-based SSSCs can be boosted by the sensitizer NP passivation with indium sulfide protective layers. At that, the highest performance was observed for the photoanodes with two In2S3 layers (Fig. 4.30b). The first layer is placed between AIS NPs and the TiO2 scaffold to mediate the electron transfer from AIS to titania and simultaneously to impede the reverse electron transfers. The second In2S3 layer is deposited on the photoanode surface followed by the deposition of an additional protective ZnS layer—to mediate the hole transfer from AIS NPs to sulfide ions in the electrolyte [198]. 214 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

Fig. 4.31 a PL QY (blue bars) of size-selected CIS NPs and light conversion efﬁciency of TiO2/ CIS heterostructures based on corresponding NPs (red bars). Adapted (a) and reprinted (b) with permissions from Ref. [51]. Copyright (2015) American Chemical Society

The light sensitivity range of SSSCs based on the ternary NPs can be further extended by using ZnS-protected Cu-In-Se NPs with Eg = *1.3 eV [175]. The SSSCs sensitized with Ag-In-Se NPs also can harvest light down to *860 nm (1.44 eV, Fig. 4.32a) [54]. By doping Ag-In-Se NPs with Zn a series of (AgIn)xZn2(1−x)Se2 solid solution NPs can be produced with CB and VB levels shifting continuously to lower values as x is increased (Fig. 4.32b). The offset between the CB levels of sensitizer NPs and titania increases with a decrease in x, but simultaneously Eg expands as well, resulting in a loss of the solar light harvesting capability (Fig. 4.32c, curves 1, 2). As a result, the light conversion efﬁciency of Zn-doped AISe NP-based SSSCs shows a

Fig. 4.32 a, b IPCE spectra (AgIn)xZn2(1−x)Se NP-based photoanodes and energy diagram of corresponsding sensitizer NPs; c band gap (curve 1) and ECB (curve 2, versus vacuum level) of (AgIn)xZn2(1−x)Se (ZAIS) NPs as a function of x. Yellow bars indicate the light conversion efficiency η (given in black figures) of corresponding TiO2/ZAISe/CdS photoanodes as well as TiO2/CdS heterostructure. Reprinted (a, b) and adapted (c) with permissions from Ref. [54]. Copyright (2014) American Chemical Society 4.5 Making Progress in SSSCs—Toward More Efficient … 215 volcano-shaped dependence on the Zn content with a maximal value of 1.9% observed at x =0.5[54](Fig. 4.32c, bars). By similar reasons, the SSSCs with mixed CuInSxSe1 −x NPs show the highest light conversion efficiency at an intermediary chalcogenide composition corresponding to x around 0.4 [199]. A progress in the ternary CIS/AIS NPs applications for the SSSCs stimulated a search for other ternary and more complex formulations that revealed new and promising metal chalcogenide sensitizers, in particular Cu2SnS3 [200]. Such NPs can easily be electrodeposited on the metal oxide substrates from aqueous electrolytes containing Cu and Sn citrate complexes and Na2S2O3 and provide efficient light harvesting in the range of hv > 2 eV [200]. The ternary Bi- and Sb-based sensitizers crystallize in a variety of composition revealing a plethora of band gaps favorable for the solar light harvesting *1eV for Cu3SbS4 [201], *1.3 eV for AgBiS2 [99], *1.7 eV for AgSbS2 [101, 157] and Cu12Sb4S13 [201], 2.1 eV for CuBiS2 [128]. The quaternary kesterite Cu2ZnSnS(Se)4 materials that find broad applications in photovoltaics in the form of microcrystalline thin films [202–205] can be produced as NPs [206] that have a great potential for application as the SSSC sensitizers [72, 156]. Design of new architectures of the metal oxide scaffolds for more efficient SSSCs. The principles and approaches of the design of efficient light-scattering and charge-collecting metal oxide scaffolds for the SSSCs are generally the same as those devised for the dye-sensitized solar cells [16, 19, 20, 26, 207, 208]. Since the sensitizer NPs usually have higher extinction coefficients than typical dye sensitizers the high specific surface area of the scaffold is not so critical for the SSSCs as it is for the DSSCs, however, a high porosity of the metal oxide layer is still welcomed for a better contact between the sensitizer NPs and the electrolyte. At the same time, a high contact area enables also a higher recombination rate and so a certain optimal scaffold porosity is always required to achieve the highest SSSC performance. The oxide scaffold accepts the photogenerated electrons from the sensitizer NPs and transfers them further into the circuit and the efficiency of this process depends directly on the lattice perfection and the defect density in the oxide layer. As the electron travels trough the mesoporous metal oxide scaffold it “visits” *102 oxide NPs and in each NP there exists a probability of the recombination with adsorbed electrolyte species [15, 16, 19, 26, 29, 208, 209]. In this view, single-crystalline oxide NRs or NWs provide a better conductance of electrons than the polycrystalline mesoporous frameworks abundant with the interparticle interfaces and defect states. The electron mobility is estimated to be hundreds of times higher for ZnO NWs as compared with the conventional mesoporous ZnO films [15, 19, 20, 26, 209]. As a result, the SSSC designs based on ZnO NWs [19, 20, 43, 55, 94, 95, 124, 154, 190, 208, 209], and NRs [79, 119, 120, 125, 130, 155, 156, 196, 210, 211] (Fig. 4.33a) were successfully realized. The ZnO NRs and NR arrays can be quite conveniently formed by the electrodeposition, the NR length, and diameter controlled by the electrolysis duration [120]. Alternatively, the NW- and NR-based 216 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

Fig. 4.33 SEM images of arrays of ZnO NWs (a), TiO2 NWs (b), ZnO NTs (c), and TiO2 NTs (d). Reprinted with permissions from Refs. [119](a), [123](b), [212](c), and [92](d). Copyright (2011–2013) American Chemical Society (a, c) and The Royal Society of Chemistry (b, d)

ZnO scaffolds can be produced by a hydrothermal treatment using hexamethylenetetramine as a structure-directing agent [55, 125, 130, 154–156, 196, 210] (see synthesis details in Chap. 5). Similarly, titania NRs produced by the HTT were probed as a transport layer in SSSCs (Fig. 4.33b) [123, 144, 209]. Even in the case of mesoporous materials, a considerable breakthrough was achieved for the ordered scaffolds, such, for example, as mesoporous titania NT arrays. The lateral movement of charge carriers is strongly confined in such NTs making the carriers to move along the main NT axis toward OTE. Also, the NTs reveal a strong light scattering, thus increasing the light absorption probability in the sensitizer NPs. Both titania and ZnO polycrystalline NTs are typically produced by the electrochemical etching. The ZnO NTs can be synthesized by the elec- troetching of previously electrodeposited ZnO NRs (Fig. 4.33c) [121, 185, 212], while titania NT arrays (Fig. 4.33d) are typically formed by the etching of titania foils in various fluorine-based electrolytes [92, 106, 213]. The etching procedures can also be applied to produce porous ZnO NTs [62, 158, 214]. 4.5 Making Progress in SSSCs—Toward More Efficient … 217

In the context of the idea of a strong light scattering by the metal oxide scaffold, inverse opal oxide structures were introduced both in the DSSCs and SSSCs, allowing to “capture” the incoming light and subject it to the multiple scattering within the absorbing NP layer [207]. Typically, titania photonic crystals are produced by self-assembling of the polystyrene microspheres into opals, soaking the opalescent assembly with titanium(IV) precursors and calcination producing the hollow-sphere inverse replica of the original opal [81]. Rutile hollow spheres can be produced by the laser-induced melting of titania NPs and applied as a light scattering layer in SSSCs [129].

4.6 Nanocrystalline Semiconductor Counter-Electrodes for SSSCs

A counter electrode is an important functional constituent of the SSSCs as it collects electrons photogenerated in the visible-light-sensitive photoanode and cat- alyzes the reduction of a component of the redox couple that was oxidized by the photogenerated conduction band holes on the photoanode [30, 31]. As the liquid-junction SSSCs operate most often with aqueous (aqueous/methanolic) solutions of sodium sulfide/polysulfide, that is with a S2−/S0 redox couple, the CE must fulfill several basic requirements. It should be catalytically active with respect to the transformations of the redox couple (in this case, to catalyze reduction of sulfur to S2−), it must provide sufficiently high surface area to avoid any diffusion resistances on the electrolyte/CE interface and, finally, the CE should be stable in the polysulfide electrolyte. At the rise of the studies on the SSSCs it was realized that Pt CE, traditional and the “best” one for the dye-sensitized liquid-junction solar cells exhibits a low efficiency and a low stability in the polysulfide electrolyte due poisoning of the Pt surface with sulfur species. At the same time, some metal sulfides were recognized as very promising materials for the CE of SSSCs with the polysulfide electrolytes [30, 31], such as CuxS[33, 215–225], CoSx [33, 215, 218, 226–228], NiS [215, 229], PbS [35, 219, 229, 230], etc. Also, the nanostructured films of Cu3Se2 grown on FTO by CBD were found to reveal superior catalytic properties in SSSCs as compared to copper sulfide-based CEs [231]. Such materials revealed excellent catalytic properties with respect to the S2−/S0 redox couple as well as a long-term stability. Recently, good perspectives for various porous carbon materials as CEs for the polysulfide-based SSSCs as well as for metal sulfide composites with the carbon materials were also recognized. For a deeper analysis of the current state-of-the-art in this area, the reader is referred to recent excellent reviews on the classification and special features of metal-sulfide, carbonaceous and composite CE materials for SSSCs [31, 232]. Table 4.7 summarizes some of the reported SSSCs with metal sulfide and carbon CE and provides an overview of the achieved efficiencies of the light power conversion in such systems. A large part of these studies was performed 218 4 Semiconductor-Based Liquid-Junction Photoelectrochemical … with similar TiO2/CdS/CdSe photoanodes enabling a comparison of various types of the CE materials. As the first several rows of Table 4.7 show, the Pt CEs appear to be much inferior to the copper and cobalt sulfide films produced on the FTO plates by the SILAR deposition, the sulfidation of metal copper, or the electrodeposition. Similarly to the photoanodes, the metal-sulfide-based CEs can easily be produced by the SILAR from aqueous solutions of metal salts and sodium sulfide. This method can be generally applied to form porous copper, cobalt, nickel and lead sulfide films on the nanocrystalline mesoporous ITO [215]. In the cells with a TiO2/ CdS/CdSe photoanode, the CuxS and CoS CEs demonstrate a superior activity as compared to NiS and PbS as well as to the conventional Pt (Fig. 4.34a). The catalytic activity of the best CuxS CEs was found to depend strongly on the SILAR cycle number N most probably due to the incomplete ITO coverage at small Ns. The activity of various CE materials is closely related to the dynamics of the charge transfer on the CE/electrolyte interface. As a measure of the charge transfer efficiency the charge transfer resistance RCT can be adopted, which is typically determined by studying the CE with the electrochemical impedance spectroscopy (EIS). A higher RCT indicates a lower charge transfer rate and, correspondingly, the lower catalytic activity in the sulfur reduction reflected in a lower light conversion efficiency of the entire SSSCs. For example, for a series of Pt, CoS and CuS 2 electrodes RCT measured by EIS is 560, 48, and 6 O/cm , respectively [218] varying in line with the light conversion efficiency of the corresponding cells with a TiO2/CdxZn1−xS/CdSe/ZnS photoanode (Table 4.7). The relative activity of various CE materials can be evaluated from the “dark” current-voltage curves, that is the J–V dependences registered with no illumination applied in a three-electrode scheme with a metal sulfide film as a working electrode, some auxiliary electrode (Pt) and a reference electrode. The higher is the catalytic activity of the CE material toward S2−/S0 couple, the steeper is the J–V dependence, that is an increment/decrement of J with varied V. Figure 4.34b shows such J–V curves for a series of CuxS CE produced by the SILAR on the non-porous ITO glass and mesoporous ITO films with a different number of deposition cycles. As can be seen, the catalytic effect is very small for nonporous substrates as compared to the porous ones and increases with an increased amount of the CuxS catalyst. For the maximal cycle number N = 12 the J–V dependence is closer to Y axis as compared with the curve for Pt complying with the higher efficiency of the porous CuxS film as a CE. The surface area of CuxS-based materials can be increased by coupling copper sulfide particles to conductive/semiconductive substrates with a high surface area. For example, the composites of CuxS NPs with RGO exhibited quite high activity as CE with TiO2/CdS/CdSe photoanodes [217]. Similarly to the photoanodes, the metal-sulfide CEs with a high specific surface area can be produced by using other nanostructured semiconductors as a platform. In particular, ZnO/CuxS heterostructures can be easily prepared by the SILAR deposition of copper sulfide onto ZnO NRs (Fig. 4.35a) [220]. The CuxS forms a uniform layer on the NR surface gaining from the high surface area of the ZnO substrate. Table 4.7 Some examples of SSSCs produced by with various counter electrodes 219 SSSCs for Counter-Electrodes Semiconductor Nanocrystalline 4.6 −2 Photoanode Counter electrode (CE) CE formation method Jsc,mAÂ cm Voc, mV FF, % η, % Reference

TiO2/CdS/CdSe Pt SILAR 6.70 370 22 0.56 [215] CuxS SILAR 9.38 420 37 1.47 CoS 9.52 370 47 1.41

TiO2/CdS/CdSe CuxS/Cu mesh sulﬁdation 11.54 478 59 3.27 [216] TiO2/CdS/CdSe Pt sulﬁdation 11.3 460 31 1.60 [217] CuxS/RGO 18.4 520 46 4.40 TiO2/CdS-ZnS/CdSe/ZnS Pt e/d 9.1 470 35 1.6 [218] CoS e/d 11.2 520 32 1.9 CuS 13.9 550 35 2.7

TiO2/CuInS2/ Pt doctor blade/precipitation 14.2 430 37 2.3 [219] CdS/ZnS CuS 17.0 550 42 4.0 PbS 18.3 580 45 4.7

TiO2/CdS mesoporous carbon matrix carbonization 4.31 540 46.7 1.08 [244] TiO2/CdS/CdSe CoS2 sulﬁdation 14.44 510 56.5 4.16 [227] TiO2/CdS/CdSe CuS/CoS CBD 17.11 n/r 55.4 4.1 [33] TiO2/CdS/CdSe CoS CBD 14.95 454 50.5 3.4 [226] TiO2/CdS/CdSe carbon foam carbonization 6.85 510 50 1.75 [244] TiO2/CdS/CdSe PbS sulﬁdation 9.28 554 59 3.01 [230] ZnO/CdSe/CdS carbon foam carbonization 12.6 685 42 3.60 [245]

TiO2/CdS/CdSe ZnO/PbS CBD 13.28 633 56.6 4.76 [229] ZnO/CdS/CdSe ZnO/CuS SILAR 14.48 740 35 4.18 [220]

TiO2/CdS/CdSe Cu3Se2 CBD 13.10 567 63.4 4.71 [231] TiO2/CdSe/ZnS Ni foam/CuxS sulﬁdation 9.95 581 61.4 3.55 [246] TiO2/CdSe/CdS Ca-doped CuxS/RGO SILAR/electropho-retic deposition 16.26 520 33 2.73 [34] TiO2/CdSe/CdS PbS CBD 12.17 644 59 4.61 [35] (continued) Table 4.7 (continued) Photoelectrochemical Liquid-Junction Semiconductor-Based 4 220 −2 Photoanode Counter electrode (CE) CE formation method Jsc,mAÂ cm Voc, mV FF, % η, % Reference

TiO2/CdSe/CdS Cu1.8 S platelets CBD 19.1 *600 45 5.16 [221] TiO2/CdS WO3−x electrodeposition 7.9 1004 44.5 3.66 [247] carbon/WO3−x 8.86 951 51.5 4.60 TiO2/CdSexTe1−x CuxS electrodeposition 20.61 698 61.2 8.79 [234] carbon/Ti carbonization 20.67 803 68.6 11.39

TiO2/CdSexTe1−x carbon/CuxS sulﬁdation 21.27 655 60 8.40 [222] TiO2/CdS/CdSe CuInS2 ex situ/doctor blade 13.43 518 52 3.63 [238] carbon/CuInS2 14.16 512 60 4.32 ZnO/ZnSe/CdSe Cu1.8 S solvothermal synthesis 10.51 822 42.3 3.65 [223] Cu2SnS3 11.46 810 43.7 4.06 TiO2/CdSe Cu2ZnSnS4 spray deposition/ 10.53 520 40 2.19 [240] Cu2ZnSnSe4 selenization 15.49 540 52 4.35 TiO2/CdS/CdSe Cu2ZnSn(S,Se)4 ex situ/drop casting 12.71 550 43 3.01 [237] Note AM1.5 if not stated otherwise; ZnS layer typically (see refs.) PS redox-couple if not stated otherwise … 4.6 Nanocrystalline Semiconductor Counter-Electrodes for SSSCs 221

Fig. 4.34 a Light conversion efﬁciency in the SSSCs with TiO2/CdS/CdSe photoanodes and different CE; b Current−voltage curves of ITO glass-supported CuS(x), ITO porous ﬁlm-supported CuS(x), and Pt CEs (x—SILAR cycle number). Reprinted with permissions from Ref. [215]. Copyright (2013) American Chemical Society

Additionally, the CuxS layer reveals its own typical sheet-like morphology of separate particles forming the layer. The activity of ZnO/CuxS heterostructure as a CE for a ZnO/CdS/CdSe photoanode-based SSSC depends on the SILAR cycle number [220]. The light conversion efﬁciency increases with N increasing from 2 to 6 and supersedes the activity of individual CuxS (Fig. 4.35b). Then η decreases as N is elevated from 6 to 8 following closely the variation of the charge transfer resistance RCT. The similarity of both trends shows that at N > 6 the copper sulﬁde layer on the

Fig. 4.35 a SEM of ZnO (a) and ZnO/CuS−6 NRs (b) cover produced by the SILAR; b Light conversion efﬁciency in SSSCs with ZnO/CdS/CdSe photoanodes and CuS-based CEs, and charge transfer resistance for CuS and ZnO/CuS counter electrodes (c). Reprinted with permissions from Ref. [220]. Copyright (2016) The Royal Society of Chemistry 222 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

ZnO surface becomes too thick for the efficient electron transfer to sulfur species in the electrolyte, while at a smaller N the coverage of CuxS is insufficient for the fast transfer of all electrons incoming to the counter electrode, while the ZnO alone has a low catalytic activity in the conversion of S0 to S2−. Alternatively to the SILAR, the metal sulfide catalysts—CoS, NiS, CuS, PbS, can be deposited to the surface of ZnO NRs by CBD [229]. Also in this case the ZnO NRs provide a high-surface-area framework for the metal sulfide loading that is easily accessible by the electrolyte. Due to a high electron mobility in zinc oxide and single-crystalline character of ZnO NRs, the NRs offer an efficient electron pathway from the circuit to the metal sulfide catalyst layer. Unlikely the previous studies on the SILAR-deposited sulfides, the highest light conversion efficiency (4.76%) was observed for a ZnO/PbS CE-based SSSC. The relative activity of the ZnO/metal sulfide heterostructures as CEs can be vividly anticipated from a comparison of their “dark” J–V characteristics (Fig. 4.36a). By the angle between the Y axis and corresponding J–V curves the heterostructures form the following row: ZnO/PbS > ZnO/CuS > ZnO/NiS > ZnO/ CoS > Pt. Exactly the same decreasing sequence is observed for the basic PEC activity parameters of the solar cells with a TiO2/CdS/CdSe photoanode and the above-discussed CEs (Fig. 4.36b). The ZnO/PbS electrodes demonstrated excellent stability after multiple (more than 50) cyclic J–V measurements [229] indicating on the robustness of such CE architecture that is necessary for the applications in the SSSCs. Similarly to the preparation of photoanodes, the photocatalytic deposition of an active metal-sulfide phase can be a good alternative to both SILAR and CBD preparations of CEs. As mentioned before, a copper sulfide film can be photocatalytically deposited onto mesoporous titania immersed into ethanol solution of copper perchlorate and S8 and illuminated with the UV light [138]. The CuxS deposit exerts a light-shielding effect and slows the photoprocess after the

Fig. 4.36 a Cyclic voltammetry (CV) measurements of electrodes formed from ZnO NRs with metal sulfides and Pt in a polysulfide solution with a scan rate of 100 mV Â s−1; b Light conversion efficiency, open-circuit voltage and fill factor for SSSCs with different ZnO-based CEs. Reprinted with permissions from Ref. [229]. Copyright (2016) The Royal Society of Chemistry 4.6 Nanocrystalline Semiconductor Counter-Electrodes for SSSCs 223 deposition of first portions of copper sulfide because it absorbs the incident light but cannot participate in the photo–chemical reactions. To avoid the light shielding and to produce thicker and more robust copper sulfide films it was proposed to perform the photodeposition in two stages, with the photocatalytic formation of copper NPs on the surface of TiO2 or ZnO films followed by their transformation into CuxSina reaction with sodium sulfide (or polysulfide) that can take place in situ in the Na2Sx electrolyte [233]. The photodeposited Cu0 NPs can act as a co-catalyst that collected the photogenerated electrons from the metal oxide photocatalyst and accelerate the two-electron reduction of Cu2+ [239] resulting in a much higher copper content as compared to the direct photodeposition of CuxS. Figure 4.37a shows SEM images of a ZnO/Cu0 composite produced by the photocatalytic deposition of copper on the surface of ZnO films. Copper is deposited in the form of NPs with a broad size distribution—from tens to hundreds

0 Fig. 4.37 SEM images of ZnO/Cu (a, b) and ZnO/CuxS(c, d) films produced by the copper photodeposition [233] 224 4 Semiconductor-Based Liquid-Junction Photoelectrochemical … nm attached to ZnO platelets and even growing through the platelets in some places (Fig. 4.37b), Such morphology indicates that primary smaller Cu NPs participate as co-catalysts in the photodeposition resulting in Cu0 deposition exclusively on the primary metal NPs with no additional nuclei forming in the system. When the ZnO/Cu0 films are immersed into aqueous sodium polysulfide solutions the spontaneous sulfidation takes place and the faceted Cu NPs transform into spherical aggregates of the nanometer-thin copper sulfide platelets (Fig. 4.37c, d). According to EDX, the atomic Cu/S ratio in the film was 1.3–1.4 indicating on a non-stoichiometric character of the copper sulfide coating as well as on a possible partial sulfidation of ZnO microplatelets. The platelet-like morphology is quite typical for the products of the sulfidation 0 of both Cu [216] and Cu2O[235]. The secondary aggregation of separate copper sulfide nanoplatelets is driven most probably by a spheroidal shape of starting photodeposited Cu0 NPs, because in a similar system produced by a ions substitution of Zn(II) in ZnO nanoplatelets with Cu(II) the shape of secondary CuxS nanoplatelet aggregates mimicked the shape of starting ZnO microplatetes [236]. The ZnO/CuxS nanostructured films produced by the above-discussed photoassisted deposition were used as CEs with a ZnO/CdS photoanode showing around 25% higher efficiency of the solar light harvesting as compared to similar systems where the ZnO/CuxS counter electrode was produced by the ion exchange [236]. Typically, the metal-sulfide CEs are non-transparent and, therefore, the cell should be illuminated through the semi-transparent photoanodes thus requiring the photoanode to be formed on a transparent conductive substrate. To expand the range of possible conductive electrodes and use, for example, metal-based electrodes, such, for example, as anodized Ti foils with titania NT arrays, transparent or semitransparent CE are required. The solution to this problem can be found in utilizing metal meshes with micro-/nano-layers of catalytically active materials formed on their surface. For example, by contacting a copper mesh with polysulfide solution a Cu/CuxS heterostructrure can be produced (Fig. 4.38a–d) [216]. Such Cu mesh is light-transparent and conductive and supports a *2–3 lm-thick CuxS layer that is composed of separate thin copper sulfide sheets thus providing a large contact area between the CuxS layer and the polysulfide electrolyte. Lead sulfide was found to reveal “dark” and photochemical catalytic properties in the sulfur reduction to S2− [219]. As PbS combines a p-type photoresponse with broad absorption bands extending to the NIR range it can be used as a photocathode in a SSSCs with a TiO2/CuInS2/CdS photoanode and the polysulfide electrolyte. Under the illumination, such cell outperforms similar cells based on Pt and CuxS CE even though copper sulfide has superior electrocatalytic activity toward the S2−/ S0 redox couple. The PbS-based SSSC showed a light conversion efficiency of 4.7% that is *15% higher than the CuxS-based analog [219]. The higher efficiency of the PbS-based cell originates from two factors. The first is in additional photocurrent produced by the photoexcitation of PbS cathode due to the photostimulated reduction of sulfur in polysulfide anions (Fig. 4.38e). The second factor is an increase in the total Voc of the cell due to the contribution of a 4.6 Nanocrystalline Semiconductor Counter-Electrodes for SSSCs 225

Fig. 4.38 a–d SEM images of semitransparent CuxS/Cu mesh CE; e Schematic energy band diagram and charge transfer processes in the SSSC with PbS-based photo-active CE. Reprinted with permissions from Refs. [216](a–d) and [219](e). Copyright (2013) Elsevier (a–d) and The Royal Society of Chemistry (e)

photovoltage produced by the photoanode (Voc 1 on Fig. 4.38e) to the photovoltage generated by the PbS photocathode (Voc 2 on Fig. 4.38e). Naturally, only a small portion of the light incoming the cell penetrates the photoanode and reaches the photocathode. However, the combination of “dark” catalytic activity of PbS with the photoinduced catalytic effect from this residual light allows outperforming the photochemically inert copper sulfide. High power conversion efficiencies in the SSSCs with CuxS, CoS and other sulfides stimulated further searches for alternative metal chalcogenide materials. For example, cobalt disulfide pyrite CoS2 which is quite abundant in the Earth crust was tested as a CE with the polysulfide electrolyte and a TiO2/CdS/CdSe photoanode demonstrating a reasonably high energy conversion efficiency of 4.16% [227]. More complex ternary chalcopyrite and quaternary kesterite metal chalcogenides were also tested as CE materials, simultaneously with probing of their potential as a light-harvesting component of the SSSCs. In particular, both individual and mixed quaternary chalcogenides with a general formula Cu2ZnSn(S1−xSex)4 were found to be catalytically active towards the S2−/S0 redox-couple, their activity depending strongly on the CE composition and morphology [237]. The light conversion efficiency in SSSCs based on the kesterite CE and a TiO2/CdS/CdSe photoanode reaches a peak value of 3.01% at x = 0.5 and then drops as the selenium content is increased (Fig. 4.39a). All the tested kesterites were more active than Pt (1.24%). The charge transfer resistance RCT shows a similar dependence on the CE composition passing through a minimum for x = 9.5 (Fig. 4.39a), however, it is relatively small both for Cu2ZnSnS4 and Cu2ZnSn(S0.5Se0.5)4, while the corresponding SSSCs differ in η almost by 100%. Such a difference in the CE activity at a relatively low RCT was attributed to a much more developed surface area of the sulfoselenide kesterite (Fig. 4.39b) as compared to the individual Cu2ZnSnS4 (panel c). The non-uniform morphology of Cu2ZnSn(S0.5Se0.5)4 arises most probably from the inhomogeneous 226 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

Fig. 4.39 a Light conversion efficiency and charge transfer resistance RCT for TiO2/ CdS/CdSe/ZnS photoanodes coupled with Cu2ZnSn(S,Se)4 CEs with a different S to Se ratio; b, c Morphology of Cu2ZnSn(S1−xSex)4 films with x = 0.8 (b) and x =0(c). Reprinted with permissions from Ref. [237]. Copyright (2013) American Chemical Society nucleation due to a difference in the Se and S ionic radii favoring to a higher catalytic activity of the mixed sulfoselenide kesterite. An abundant source of various CE materials was found by combining metal-sulfide NPs with diverse carbonaceous materials [247], such as soot, mesoporous carbon foams with a developed surface area [222, 234, 238, 248], carbon NTs [31, 247], RGO [31, 34, 228]. Individual carbon nanomaterials can also be applied as cathode materials in the SSSCs with visible-light-sensitive photoanodes and polysulfide or iodine/iodide redox-shuttles, in particular, mesoporous carbons [234, 249], mesocellular carbon foams [245]. Partial sulfidation of flexible Cu/Ni films was shown to yield stretchable counter electrodes for the CdSe-sensitized SSSCs [246]. Such CEs can be combined with flexible photoanodes based on the titania-modified plastics showing a light conversion efficiency of 3.55% as well as a good chemical and mechanical robustness [246]. Concluding the discussion of various aspects of the semiconductor NP-sensitized photoelectrochemical solar cells we should note that this research area seems to be in its very blossom stage, especially if compared with the dye-sensitized solar cells, where a certain saturation is currently observed. The progress in SSSCs occurs simultaneously in many directions, it includes (i) constant emergence of new nano-materials for the light-harvesting photoanodes and catalytic counter electrodes, especially among the Earth-abundant and low-toxic semiconductors and carbonaceous materials; (ii) steady development of design conceptions to orchestrate the photoinduced electron transfers in composite photoelectrodes such as the cascade design, the bandgap and CB/VB level design, the scaffold 4.6 Nanocrystalline Semiconductor Counter-Electrodes for SSSCs 227 morphology design, etc.; (iii) deep investigations into the factors limiting the light conversion efficiency, such as various recombination processes and interfacial barriers. The total light conversion efficiencies, both in absolute values and in the increments achieved in recent years as compared to the earlier studies, inspire a strong optimism and show a good future for this exciting area of the photochemical solar light conversion research.

References

1. Manassen J, Cahen D, Hodes G, Sofer A (1976) Electrochemical, solid state, photochemical and technological aspects of photoelectrochemical energy converters. Nature 263:97–100. doi:10.1038/263097a0 2. Bhubaneswari P, Iniyanb S, Ranko G (2011) A review of solar photovoltaic technologies. Renew Sus En Rev 15:1625–1636. doi:10.1016/j.rser.2010.11.032 3. Kamat PV (2013) Energy outlook for planet earth. J Phys Chem Lett 4:1727–1729. doi:10. 1021/jz400902s 4. Rühle S, Shalom M, Zaban A (2010) Quantum dot sensitized solar cells. ChemPhysChem 11:2290–2304. doi:10.1002/cphc.201000069 5. Hodes G (2008) Comparison of dye- and semiconductor-sensitized porous nanocrystalline liquid junction solar cells. J Phys Chem C 112:17778–17787. doi:10.1021/jp803310s 6. Kamat PV (2008) Quantum dot solar cells. Semiconductor nanocrystals as light harvesters. J Phys Chem C 112:18737–18753. doi:10.1021/jp806791s 7. Kamat PV, Tvrdy K, Baker DR, Radich JG (2010) Beyond photovoltaics: semiconductor nanoarchitectures for liquid-junction solar cells. Chem Rev 110:6664–6688. doi:10.1021/ cr100243p 8. Kamat PV (2007) Meeting the clean energy demand: nanostructure architectures for solar energy conversion. J Phys Chem C 111:2834–2860. doi:10.1021/jp066952u 9. Kamat PV (2013) Quantum dot solar cells. The next big thing in photovoltaics. J Phys Chem Lett 4:908–918. doi:10.1021/jz400052e 10. Nozik AJ (2010) Nanoscience and nanostructures for photovoltaics and solar fuels. Nano Lett 10:2735–2741. doi:10.1021/nl102122x 11. Grätzel M (2001) Photoelectrochemical cells. Nature 414:338–344. doi:10.1038/35104607 12. Mlinar V (2013) Engineered nanomaterials for solar energy conversion. Nanotechnology 24:042011 13. Fan X, Zhang M, Wang X et al (2013) Recent progress in organic–inorganic hybrid solar cells. J Mater Chem A 1:8694–8709. doi:10.1039/C3TA11200D 14. Hodes G, Cahen D (2012) All-solid-state, semiconductor-sensitized nanoporous solar cells. Acc Chem Res 45:705–713. doi:10.1021/ar200219h 15. Huang J, Yinb Z, Zheng Q (2011) Applications of ZnO in organic and hybrid solar cells. Energy Environ Sci 4:3861–3877. doi:10.1039/C1EE01873F 16. Anta JA, Guillen E, Tena-Zaera R (2012) ZnO-based dye-sensitized solar cells. J Phys Chem C 116:11413–11425. doi:10.1021/jp3010025 17. Selinsky RS, Ding Q, Faber MS et al (2013) Quantum dot nanoscale heterostructures for solar energy conversion. Chem Soc Rev 42:2963–2985. doi:10.1039/C2CS35374A 18. Yang Z, Chen CY, Roy P, Chang HT (2011) Quantum dot-sensitized solar cells incorporating nanomaterials. Chem Commun 47:9561–9571. doi:10.1039/C1CC11317H 19. Zhang Q, Dandeneau CS, Zhou X, Cao G (2009) ZnO nanostructures for dye-sensitized solar cells. Adv Mater 21:4087–4108. doi:10.1002/adma.200803827 228 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

20. Weintraub B, Zhou Z, Li Y, Deng Y (2010) Solution synthesis of one-dimensional ZnO nanomaterials and their applications. Nanoscale 2:1573–1587. doi:10.1039/C0NR00047G 21. Kojima A, Teshima K, Shirai Y, Miyasaka T (2009) Organometal halide perovskites as visible-light sensitizers for photovoltaic cells. J Am Chem Soc, 131(17):6050–6051 22. Grätzel M (2014) The light and shade of perovskite solar cells. Nat Mater 13:838–842. doi:10.1038/nmat4065 23. Petrovic M, Chellappan V, Ramakrishna S (2015) Perovskites: solar cells & engineering applications—materials and device developments. Sol Energy 122:678–699. doi:10.1016/j. solener.2015.09.041 24. Chen J, Zhou S, Jin S et al (2016) Crystal organometal halide perovskites with promising optoelectronic applications. J Mater Chem C 4:11–27. doi:10.1039/c5tc03417e 25. Meng L, You J, Guo TF, Yang Y (2016) Recent advances in the inverted planar structure of perovskite solar cells. Acc Chem Res 49:155–165. doi:10.1021/acs.accounts.5b00404 26. Zhao Y, Zhu K (2016) Organic–inorganic hybrid lead halide perovskites for optoelectronic and electronic applications. Chem Soc Rev 45:655–689. doi:10.1039/c4cs00458b 27. Heo JH, Im SH, Kim HJ et al (2012) Sb2S3-sensitized photoelectrochemical cells: open circuit voltage enhancement through the introduction of poly-3-hexylthiophene interlayer. J Phys Chem C 116:20717–20721. doi:10.1021/jp305150s 28. Hodes G, Manassen J (1980) Electrocatalytic electrodes for the polysulfide redox system. J Electrochem Soc 127:544–549. doi:10.1149/1.2129709 29. Chen H, Zhu L, Liu H, Li W (2013) ITO porous film-supported metal sulfide counter electrodes for high-performance quantum-dot-sensitized solar cells. J Phys Chem 117:3739– 3746. doi:10.1021/jp309967w 30. Meng K, Chen G, Ravindranathan Thampi K (2015) Metal chalcogenides as counter electrode materials in quantum dot sensitized solar cells: a perspective. J Mater Chem A 3:23074–23089. doi:10.1039/c5ta05071e 31. Wang S, Tian J (2016) Recent advances in counter electrodes of quantum dot-sensitized solar cells. RSC Adv 6:90082–90099. doi:10.1039/c6ra19226b 32. Eskandari M, Ahmadi V, Ghahary R (2015) Copper sulfide/lead sulfide as a highly catalytic counter electrode for zinc oxide nanorod based quantum dot solar cells. Electrochim Acta 151:393–398. doi:10.1016/j.electacta.2014.11.037 33. Yang Z, Chen CY, Liu CW et al (2011) Quantum dot–sensitized solar cells featuring CuS/ CoS electrodes provide 4.1% efficiency. Adv Energy Mater 1:259–264. doi:10.1002/aenm. 201000029 34. Khalili SS, Dehghani H (2016) Ca-doped CuS/graphene sheet nanocomposite as a highly catalytic counter electrode for improving quantum dot-sensitized solar cell performance. RSC Adv 6:10880–10886. doi:10.1039/c5ra24053k 35. Gopi C, Bae J, Venkata-Haritha M et al (2015) One-step synthesis of solution processed time dependent highly efficient and stable PbS counter electrodes for quantum dot-sensitized solar cells. RSC Adv 5:107522–107532. doi:10.1039/c5ra22715a 36. Nayak PK, Garcia-Belmonte G, Kahn A et al (2012) Photovoltaic efficiency limits and material disorder. Energy Environ Sci 5:6022–6039. doi:10.1039/C2EE03178G 37. Salant A, Shalom M, Tachan Z et al (2012) Quantum rod-sensitized solar cell: nanocrystal shape effect on the photovoltaic properties. Nano Lett 12:2095–2100. doi:10.1021/ nl300356e 38. Abrutis A, Plausinaitiene V, Skapas M et al (2008) Chemical vapor deposition of chalcogenide materials for phase-change memories. Microelectron Eng 85:2338–2341. doi:10.1016/j.mee.2008.09.014 39. Perednis D, Gauckler LJ (2005) Thin film deposition using spray pyrolysis. J Electroceram 14:103–111. doi:10.1007/s10832-005-0870-x 40. Park YM, Andre R, Kasprzak J et al (2007) Molecular beam epitaxy of CdSe epilayers and quantum wells on ZnTe substrate. Appl Surf Sci 253:6946–6950. doi:10.1016/j.apsusc.2007. 02.012 References 229

41. Bang JH, Suslick KS (2010) Applications of ultrasound to the synthesis of nanostructured materials. Adv Mater 22:1039–1059. doi:10.1002/adma.200904093 42. Tada H, Fujishimaa M, Kobayashi H (2011) Photodeposition of metal sulfide quantum dots on titanium(IV) dioxide and the applications to solar energy conversion. Chem Soc Rev, 40 (7):4232–4243. doi:10.1039/C0CS00211A 43. Xu J, Yang X, Wang H et al (2011) Arrays of ZnO/ZnxCd1−xSe nanocables: band gap engineering and photovoltaic applications. Nano Lett 11:4138–4143. doi:10.1021/ nl201934k 44. Kershaw SV, Susha AS, Rogach AL (2013) Narrow bandgap colloidal metal chalcogenide quantum dots: synthetic methods, heterostructures, assemblies, electronic and infrared optical properties. Chem Soc Rev 42:3033–3087. doi:10.1039/c2cs35331h 45. Rivera-González N, Chauhan S, Watson DF (2016) Aminoalkanoic acids as alternatives to mercaptoalkanoic acids for the linker-assisted attachment of quantum dots to TiO2. Langmuir 32:9206–9215. doi:10.1021/acs.langmuir.6b02704 46. Calzada R, Thompson CM, Westmoreland DE et al (2016) Organic-to-aqueous phase transfer of cadmium chalcogenide quantum dots using a sulfur-free ligand for enhanced photoluminescence and oxidative stability. Chem Mater 28:6716–6723. doi:10.1021/acs. chemmater.6b03106 47. Salant A, Shalom M, Hod I et al (2010) Quantum dot sensitized solar cells with improved efficiency prepared using electrophoretic deposition. ACS Nano 4:5962–5968. doi:10.1021/ nn1018208 48. Martinez-Ferrero E, Mora Sero I, Albero J et al (2010) Charge transfer kinetics in CdSe quantum dot sensitized solar cells. Phys Chem Chem Phys 12:2819–2821. doi:10.1039/ b924970b 49. Pan Z, Zhao K, Wang J et al (2013) Near infrared absorption of CdSexTe1−x alloyed quantum dot sensitized solar cells with more than 6% efficiency and high stability. ACS Nano 7:5215–5222. doi:10.1021/nn400947e 50. Penga Z, Liua Y, Wub L et al (2016) Influence of surface states of CuInS2 quantum dots in quantum dots-sensitized photo-electrodes. Appl Surf Sci 388:437–443. doi:10.1016/j.apsusc. 2015.12.004 51. Jara DH, Yoon SJ, Stamplecoskie KG, Kamat PV (2014) Size-Dependent photovoltaic performance of CuInS2 quantum dot-sensitized solar cells. Chem Mater 26:7221–7228. doi:10.1021/cm5040886 52. Santra PK, Nair PV, Thomas KG, Kamat PV (2013) CuInS2–sensitized quantum dot solar cell. Electrophoretic deposition, excited-state dynamics, and photovoltaic performance. J Phys Chem Lett 4:722–729. doi:10.1021/jz400181m 53. Chang CC, Chen JK, Cp Chen et al (2013) Synthesis of eco-friendly CuInS2 quantum dot-sensitized solar cells by a combined ex situ/in situ growth approach. ACS Appl Mater Interfaces 5:11296–11306. doi:10.1021/am403531q 54. Kameyama T, Douke Y, Shibakawa H et al (2014) Widely controllable electronic energy structure of ZnSe − AgInSe2 solid solution nanocrystals for quantum-dot-sensitized solar cells. J Phys Chem C 118:29517–29524. doi:10.1021/jp508769f 55. Leschkies KS, Divakar R, Basu J et al (2007) Photosensitization of ZnO nanowires with CdSe quantum dots for photovoltaic devices. Nano Lett 7:1793–1798. doi:10.1021/ nl070430o 56. Zarazua I, De la Rosa E, Lopez-Luke T et al (2011) Photovoltaic conversion enhancement of CdSe quantum dot-sensitized TiO2 decorated with Au nanoparticles and P3OT. J Phys Chem C 115:23209–23220. doi:10.1021/jp207744n 4− 3− 3− 57. Jin H, Choi S, Xing G et al (2015) SnS4 , SbS4 , and AsS3 metal chalcogenide surface ligands: couplings to quantum dots, electron transfers, and all-inorganic multilayered quantum dot sensitized solar cells. J Am Chem Soc 137:13827–13835. doi:10.1021/jacs. 5b05787 230 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

58. Meng X, Du J, Zhang H, Zhong X (2015) Optimizing the deposition of CdSe colloidal quantum dots on TiO2 film electrode via capping ligand induced self-assembly approach. RSC Adv 5:86023–86030. doi:10.1039/c5ra17412k 59. Jung MH, Kang MG (2011) Enhanced photo-conversion efficiency of CdSe–ZnS core–shell quantum dots with Au nanoparticles on TiO2 electrodes. J Mater Chem 21:2694–2700. doi:10.1039/c0jm03431b 60. Kongkanand A, Tvrdy K, Takechi K et al (2008) Quantum dot solar cells. Tuning photoresponse through size and shape control of CdSe-TiO2 architecture. J Am Chem Soc 130:4007–4015. doi:10.1021/ja0782706 61. Pernik DR, Tvrdy K, Radich JG, Kamat PV (2011) Tracking the adsorption and electron injection rates of CdSe quantum dots on TiO2: linked versus direct attachment. J Phys Chem C 115:13511–13519. doi:10.1021/jp203055d 62. Chouhan N, Yeh CL, Hu SF et al (2011) Photocatalytic CdSe QDs-decorated ZnO nanotubes: an effective photoelectrode for splitting water. Chem Commun 47:3493–3495. doi:10.1039/c0cc05548d 63. Chen J, Lei W, Deng WQ (2011) Reduced charge recombination in a co-sensitized quantum dot solar cell with two different sizes of CdSe quantum dot. Nanoscale 3:674–677. doi:10. 1039/c0nr00591f 64. King LA, Riley DJ (2012) Importance of QD purification procedure on surface absorbance of QDs and performance of QD sensitized photoanodes. J Phys Chem C 116:3349–3355. doi:10.1021/jp210290j 65. Santra PK, Kamat PV (2013) Tandem-layered quantum dot solar cells: tuning the photovoltaic response with luminescent ternary cadmium chalcogenides. J Am Chem Soc 135:877–885. doi:10.1021/ja310737m 66. Itzhakov S, Shen H, Buhbut S et al (2013) Type-II quantum-dot-sensitized solar cell spanning the visible and near-infrared spectrum. J Phys Chem C 117:22203–22210. doi:10. 1021/jp312190x 67. Hyun BR, Zhong YW, Bartnik AC et al (2008) Electron injection from colloidal PbS quantum dots into titanium dioxide nanoparticles. ACS Nano 2:2206–2212. doi:10.1021/ nn800336b 68. Krüger S, Hickey SG, Tscharntke S, Eychmüller A (2011) Study of the attachment of linker molecules and their effects on the charge carrier transfer at lead sulfide nanoparticle sensitized ZnO substrates. J Phys Chem C 115:13047–13055. doi:10.1021/jp200935x 69. Li TL, Lee YL, Teng H (2012) High-performance quantum dot-sensitized solar cells based on sensitization with CuInS2 quantum dots/CdS heterostructure. Energy Environ Sci 5:5315–5324. doi:10.1039/c1ee02253a 70. Cheng KC, Law WC, Yong KT et al (2011) Synthesis of near-infrared silver-indium-sulfide (AgInS2) quantum dots as heavy-metal free photosensitizer for solar cell applications. Chem Phys Lett 515:254–257. doi:10.1016/j.cplett.2011.09.027 71. Pan Z, Mora-Seró I, Shen Q et al (2014) High-efficiency “Green” quantum dot solar cells. J Am Chem Soc 136:9203–9210. doi:10.1021/ja504310w 72. Xu A, Tao H, Chen S et al (2015) A novel approach to utilize thiol reduced graphene oxide as linker molecule for Cu2ZnSnS4 sensitized solar cell. Inter J Hydrogen Energy 40:15933– 15939. doi:10.1016/j.ijhydene.2015.09.068 73. Wang H, Luan C, Xu X et al (2012) In situ versus ex situ assembly of aqueous-based thioacid capped CdSe nanocrystals within mesoporous TiO2 films for quantum dot sensitized solar cells. J Phys Chem C 116:484–489. doi:10.1021/jp209987q 74. King LA, Parkinson BA (2016) Photosensitization of ZnO crystals with iodide-capped PbSe quantum dots. J Phys Chem Lett 7:2844–2848. doi:10.1021/acs.jpclett.6b01133 75. Raevskaya AE, Stroyuk OL, Panasiuk YV et al (2016) A new route to very stable water-soluble ultra-small core/shell CdSe/CdS quantum dots. Nano-Struct Nano-Objects. doi:10.1016/j.nanoso.2015.12.001 References 231

76. Raevskaya AE, Alontseva VV, Kozitskiy AV et al (2016) Photoelectrochemical properties of titania nanoheterostructures with low-dimensional cadmium selenide particles. Ther Exp Chem 52:152–162 77. Coughlin KM, Nevins JS, Watson DF (2013) Aqueous-phase linker-assisted attachment of 2− cysteinate( )-capped CdSe quantum dots to TiO2 for quantum dot-sensitized solar cells. ACS Appl Mater Interfaces 5:8649–8654. doi:10.1021/am402219e 78. Zhou C, Zhou L, Xu J, Gan Y (2016) Controllable synthesis of CdS quantum dots and their photovoltaic application on quantum-dot-sensitized ZnO nanorods. J Solid State Electrochem 20:533–540. doi:10.1007/s10008-015-3075-5 79. Sarkar S, Makhal A, Lakshman K et al (2012) Dual-sensitization via electron and energy harvesting in CdTe quantum dots decorated ZnO nanorod-based dye-sensitized solar cells. J Phys Chem C 116:14248–14256. doi:10.1021/jp3046593 80. Lan GY, Yang Z, Lin YW et al (2009) A simple strategy for improving the energy conversion of multilayered CdTe quantum dot-sensitized solar cells. J Mater Chem 19:2349– 2355. doi:10.1039/b817000b 81. Nehme AS, Haydous F, Halaoui L (2016) Amplification in light energy conversion at 2− Q-CdTe sensitized TiO2 photonic crystal, photoelectrochemical stability in Se electrolyte, and size-dependent type II Q-CdTe/CdSe formation. J Phys Chem C 120:4766–4778. doi:10.1021/acs.jpcc.5b11478 82. Hu X, Zhang Q, Huang X et al (2011) Aqueous colloidal CuInS2 for quantum dot sensitized solar cells. J Mater Chem 21:15903–15905. doi:10.1039/c1jm12629f 83. Chang JY, Li CH, Chiang YH et al (2016) Toward the facile and ecofriendly fabrication of quantum dot-sensitized solar cells via thiol coadsorbent assistance. ACS Appl Mater Interfaces 8:18878–18890. doi:10.1021/acsami.6b05411 84. Raevskaya AE, Rosovik OP, Kozytskiy AV et al (2016) Non-Stoichiometric Cu-In-S@ZnS nanoparticles produced in aqueous solutions as light harvesters for liquid-junction photoelectrochemical solar cells. RSC Adv 6:100145–100157. doi:10.1039/C6RA18313A 85. Yu XY, Lei BX, Kuang DB, Cy Su (2011) Highly efficient CdTe/CdS quantum dot sensitized solar cells fabricated by a one-step linker assisted chemical bath deposition. Chem Sci 2:1396–1400. doi:10.1039/c1sc00144b 86. Song X, Wang M, Deng J et al (2013) One-step preparation and assembly of aqueous colloidal CdSxSe1−x nanocrystals within mesoporous TiO2 films for quantum dot-sensitized solar cells. ACS Appl Mater Interfaces 5:5139–5148. doi:10.1021/am4009924 87. Nakamura R, Makuta S, Tachibana Y (2015) Electron injection dynamics at the SILAR deposited CdS quantum dot/TiO2 interface. J Phys Chem C 119:20357–20362. doi:10.1021/ acs.jpcc.5b06900 88. Lee YS, Gopi CV, Venkata-Haritha M, Kim HJ (2016) Recombination control in high-performance quantum dot-sensitized solar cells with a novel TiO2/ZnS/CdS/ZnS heterostructure. Dalton Trans 45:12914–12923. doi:10.1039/c6dt02531e 89. Esparza D, ZarazúaI,López-Luke T et al (2015) Photovoltaic properties of Bi2S3 and CdS quantum dot sensitized TiO2 solar cells. Electrochim Acta 180:486–492. doi:10.1016/j. electacta.2015.08.102 90. Hossain MA, Jennings JR, Shen C et al (2012) CdSe-sensitized mesoscopic TiO2 solar cells exhibiting >5% efficiency: redundancy of CdS buffer layer. J Mater Chem 22:16235–16242. doi:10.1039/c2jm33211f 91. Malashchonak MV, Mazanik AV, Korolik OV et al (2015) Influence of wide band gap oxide substrates on the photoelectrochemical properties and structural disorder of CdS nanoparticles grown by the successive ionic layer adsorption and reaction (SILAR) method. Beilstein J Nanotechnol 6:2252–2262. doi:10.3762/bjnano.6.231 92. Wang X, Zheng J, Sui X et al (2013) CdS quantum dots sensitized solar cells based on free-standing and through-hole TiO2 nanotube arrays. Dalton Trans 42:14726–14732. doi:10.1039/C3DT51266E 232 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

93. You T, Jiang L, Han KL, Deng WQ (2013) Improving the performance of quantum dot-sensitized solar cells by using TiO2 nanosheets with exposed highly reactive facets. Nanotechnology 24:245401 94. Tak Y, Hong SJ, Leeb JS, Yong K (2009) Fabrication of ZnO/CdS core/shell nanowire arrays for efficient solar energy conversion. J Mater Chem 19:5945–5951. doi:10.1039/ b904993b 95. Hwang I, Yong K (2013) Environmentally benign and efficient Ag2S-ZnO nanowires as photoanodes for solar cells: comparison with CdS-ZnO nanowires. ChemPhysChem 14:364–368. doi:10.1002/cphc.201200876 96. Emin S, Fanetti M, Abdi FF et al (2013) Photoelectrochemical properties of cadmium chalcogenide-sensitized textured porous zinc oxide plate electrodes. ACS Appl Mater Interfaces 5:1113–1121. doi:10.1021/am3027986 97. Rabinovich E, Hodes G (2013) Effective bandgap lowering of CdS deposited by successive ionic layer adsorption and reaction. J Phys Chem C 117:1611–1620. doi:10.1021/jp3105453 98. Malashchonak M, Streltsov EA, Mazanik A et al (2015) Band-gap and sub-band-gap photoelectrochemical processes at nanocrystalline CdS grown on ZnO by successive ionic layer adsorption and reaction method. Thin Solid Films 589:145–152. doi:10.1016/j.tsf. 2015.04.057 99. Huang PC, Yang WC, Lee MW (2013) AgBiS2 semiconductor-sensitized solar cells. J Phys Chem C 117:18308–18314. doi:10.1021/jp4046337 100. Wan Y, Han M, Yu L et al (2015) Fabrication and photoelectrochemical properties of TiO2/ CuInS2/Bi2S3 core/shell/shell nanorods electrodes. RSC Adv 5:78902–78909. doi:10.1039/ c5ra14548a 101. Ho YR, Lee MW (2013) AgSbS2 semiconductor-sensitized solar cells. Electrochem Commun 26:48–51. doi:10.1016/j.elecom.2012.10.003 102. Liu Z, Huang J, Han J et al (2016) CuSbS2: a promising semiconductor photo-absorber material for quantum dot sensitized solar cells. Phys Chem Chem Phys 18:16615–16620. doi:10.1039/c6cp01688j 103. Braga A, Gimenez S, Concina I et al (2011) Panchromatic sensitized solar cells based on metal sulfide quantum dots grown directly on nanostructured TiO2 electrodes. J Phys Chem Lett 2:454–460. doi:10.1021/jz2000112 104. Kim J, Choi H, Nahm C et al (2013) Graded bandgap structure for PbS/CdS/ZnS quantum-dot-sensitized solar cells with a PbxCd1−xS interlayer. Appl Phys Lett 102:183901 105. Tubtimtae A, Wu KL, Tung HY et al (2010) Ag2S quantum dot-sensitized solar cells. Electrochem Commun 12:1158–1160. doi:10.1016/j.elecom.2010.06.006 106. Chen C, Xie Y, Ali G et al (2011) Improved conversion efficiency of Ag2S quantum dot-sensitized solar cells based on TiO2 nanotubes with a ZnO recombination barrier layer. Nanoscale Res Lett 6:462. doi:10.1186/1556-276X-6-462 107. Chang JY, Lin JM, Su LF, Chang CF (2013) Improved performance of CuInS2 quantum dot-sensitized solar cells based on a multilayered architecture. ACS Appl Mater Interfaces 5:8740–8752. doi:10.1021/am402547e 108. Lee JW, Son DY, Ahn TK et al (2012) Quantum-dot-sensitized solar cell with unprecedentedly high photocurrent. Sci Rep 3:1050. doi:10.1038/srep01050 109. Kozitskiy AV, Stroyuk OL, Kuchmiy SY et al (2013) Morphology, optical, and photoelectrochemical properties of electrodeposited nanocrystalline ZnO films sensitized with CdxZn1−xS nanoparticles. J Mater Sci 48:7764–7773. doi:10.1007/s10853-013-7598-9 110. Lee JH, Song WC, Yi JS et al (2003) Growth and properties of the Cd1−xZnxS thin films for solar cell applications. Thin Solid Films 431–432:349–353. doi:10.1016/S0040-6090(03) 00526-1 111. Antoniadou M, Kondarides DI, Dionysiou DD, Lianos P (2012) Quantum dot sensitized titania applicable as photoanode in photoactivated fuel cells. J Phys Chem C 116:16901– 16909. doi:10.1021/jp305098m 112. Henglein A (1984) Catalysis of photochemical reactions by colloidal semiconductors. Pure Appl Chem 56:1215–1224. doi:10.1351/pac198456091215 References 233

113. Zeug N, Bücheler J, Kisch H (1985) Catalytic formation of hydrogen and carbon-carbon bonds on illuminated zinc sulfide generated from zinc dithiolenes. J Am Chem Soc 107:1459–1465. doi:10.1021/ja00292a001 114. Sadhu S, Patra A (2012) Lattice strain controls the carrier relaxation dynamics in CdxZn1−xS alloy quantum dots. J Phys Chem C 116:15167–15171. doi:10.1021/jp304901w 115. Sun XW, Huang JZ, Wang JX, Xu Z (2008) A ZnO nanorod inorganic/organic heterostructure light-emitting diode emitting at 342 nm. Nano Lett 8:1219–1223. doi:10. 1021/nl080340z 116. Zewdu T, Clifford JN, Hernandez JP, Palomares E (2011) Photo-induced charge transfer dynamics in efficient TiO2/CdS/CdSe sensitized solar cells. Energy Environ Sci 4:4633– 4638. doi:10.1039/c1ee02088a 117. Guijarro N, Lana-Villarreal T, Lutz T et al (2012) Sensitization of TiO2 with PbSe quantum dots by SILAR: how mercaptophenol improves charge separation. J Phys Chem Lett 3:3367–3372. doi:10.1021/jz301528a 118. Srathongluan P, Kuhamaneechot R, Sukthao P et al (2016) Photovoltaic performances of 3+ Cu2−xTe sensitizer based on undoped and In -doped TiO2 photoelectrodes and assembled counter electrodes. J Colloid Interface Sci 463:222–228. doi:10.1016/j.jcis.2015.10.052 119. Bu Y, Chen Z, Li W, Yu J (2013) High-efficiency photoelectrochemical properties by a highly crystalline CdS-sensitized ZnO nanorod array. ACS Appl Mater Interfaces 5:5097– 5104. doi:10.1021/am400964c 120. Yao CZ, Wei BH, Meng LX et al (2012) Controllable electrochemical synthesis and photovoltaic performance of ZnO/CdS core–shell nanorod arrays on fluorine-doped tin oxide. J Power Sources 207:222–228. doi:10.1016/j.jpowsour.2012.01.154 121. Qi X, She G, Liu Y et al (2012) Electrochemical synthesis of CdS/ZnO nanotube arrays with excellent photoelectrochemical properties. Chem Commun 48:242–244. doi:10.1039/ c1cc15674h 122. Yu XY, Liao JY, Qiu KQ et al (2011) Dynamic study of highly efficient CdS/CdSe quantum dot-sensitized solar cells fabricated by electrodeposition. ACS Nano 5:9494–9500. doi:10. 1021/nn203375g 123. Ai G, Sun W, Gao X et al (2011) Hybrid CdSe/TiO2 nanowire photoelectrodes: fabrication and photoelectric performance. J Mater Chem 21:8749–8755. doi:10.1039/C0JM03867A 124. Wang X, Zhu H, Xu Y et al (2010) Aligned ZnO/CdTe coreshell nanocable arrays on indium tin oxide: synthesis and photoelectrochemical properties. ACS Nano 4:3302–3308. doi:10. 1021/nn1001547 125. Tang Y, Wang P, Yun JH et al (2015) Frequency-regulated pulsed electrodeposition of CuInS2 on ZnO nanorod arrays as visible light photoanodes. J Mater Chem A 3:15876– 15881. doi:10.1039/c5ta03255e 126. Zainun AR, Tomoya S, Noor UM et al (2012) New approach for generating Cu2O/TiO2 composite films for solar cell applications. Mater Lett 66:254–256. doi:10.1016/j.matlet. 2011.08.032 127. Wei H, Gong H, Wang Y et al (2011) Three kinds of Cu2O/ZnO heterostructure solar cells fabricated with electrochemical deposition and their structure-related photovoltaic properties. CrystEngComm 13:6065–6070. doi:10.1039/c1ce05540b 128. Suriyawong N, Aragaw B, Shi JB, Lee MW (2016) Ternary CuBiS2 nanoparticles as a sensitizer for quantum dot solar cells. J Colloid Interface Sci 473(2016):60–65. doi:10.1016/ j.jcis.2016.03.062 129. Wang H, Miyauchi M, Ishikawa Y et al (2011) Single-crystalline rutile TiO2 hollow spheres: room-temperature synthesis, tailored visible-light-extinction, and effective scattering layer for quantum dot-sensitized solar cells. J Am Chem Soc 133:19102–19109. doi:10.1021/ ja2049463 130. Tian J, Zhang Q, Uchaker E et al (2013) Constructing ZnO nanorod array photoelectrodes for highly efficient quantum dot sensitized solar cells. J Mater Chem A 1:6770–6775. doi:10. 1039/c3ta11056g 234 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

131. Lin KH, Chuang CY, Lee YY et al (2012) Charge transfer in the heterointerfaces of CdS/CdSe cosensitized TiO2 photoelectrode. J Phys Chem C 116:1550–1555. doi:10.1021/ jp209353j 132. Shalom M, Buhbut S, Tirosh S, Zaban A (2012) Design rules for high-efficiency quantum-dot-sensitized solar cells: a multilayer approach. J Phys Chem Lett 3:2436–2441. doi:10.1021/jz3010078 133. Ho W, Yu JC, Lin J et al (2004) Preparation and photocatalytic behavior of MoS2 and WS2 nanocluster sensitized TiO2. Langmuir 20:5865–5869. doi:10.1021/la049838g 134. Kanda S, Akita T, Fujishima M, Tada H (2011) Facile synthesis and catalytic activity of MoS2/TiO2 by a photodeposition-based technique and its oxidized derivative MoO3/TiO2 with a unique photochromism. J Colloid Interface Sci 354:607–610. doi:10.1016/j.jcis.2010. 11.007 135. Shvalagin VV, Stroyuk OL, Kuchmiy SY (2007) Photocatalytic formation of porous CdS/ZnO nanospheres and CdS nanotubes. Theor Exp Chem 43:229–234. doi:10.1007/ s11237-007-0026-y 136. Stroyuk OL, Shvalagin VV, Kotenko TE et al (2010) Photochemical reduction of sulfur in ethanol in the presence of ZnO nanoparticles. Theor Exp Chem 46:218–224. doi:10.1007/ s11237-010-9143-0 137. Ma B, Wang L, Dong H et al (2011) Photocatalysis of PbS quantum dots in a quantum dot-sensitized solar cell: photovoltaic performance and characteristics. Phys Chem Chem Phys 13:2656–2658. doi:10.1039/C0CP02415E 138. Zhukowskiy MA, Stroyuk OL, Shvalagin VV et al (2009) Photocatalytic growth of CdS, PbS, and CuxS nanoparticles on the nanocrystalline TiO2 films. J Photochem Photobiol A: Chem 203:137–144. doi:10.1016/j.jphotochem.2009.01.007 139. Jin-nouchi Y, Akita T, Tada H (2010) Ultrafast photodeposition of size-controlled PbS quantum dots on TiO2. ChemPhysChem 11:2349–2352. doi:10.1002/cphc.201000229 140. Fujii M, Nagasuna K, Fujishima M et al (2009) Photodeposition of CdS quantum dots on TiO2: preparation, characterization, and reaction mechanism. J Phys Chem C 113:16711– 16716. doi:10.1021/jp9056626 141. Jin-nouchi Y, Naya SI, Tada H (2010) Quantum-dot-sensitized solar cell using a photoanode prepared by in situ photodeposition of CdS on nanocrystalline TiO2 films. J Phys Chem C 114:16837–16842. doi:10.1021/jp1062226 142. Fujishima M, Nakabayashi Y, Takayama K et al (2016) High coverage formation of CdS quantum dots on TiO2 by the photocatalytic growth of preformed seeds. J Phys Chem C 120:17365–17371. doi:10.1021/acs.jpcc.6b04091 143. Nagasuna K, Akita T, Fujishima M, Tada H (2011) Photodeposition of Ag2S quantum dots and application to photoelectrochemical cells for hydrogen production under simulated sunlight. Langmuir 27:7294–7300. doi:10.1021/la200587s 144. Hu H, Ding J, Zhang S et al (2013) Photodeposition of Ag2S on TiO2 nanorod arrays for quantum dot-sensitized solar cells. Nanoscale Res Lett 8:10. doi:10.1186/1556-276X-8-10 145. Tada H, Mitsui T, Kiyonaga T et al (2006) All-solid-state Z-scheme in CdS-Au-TiO2 three-component nanojunction system. Nat Mater 5:782–786. doi:10.1038/nmat1734 146. Stroyuk OL, Shvalagin VV, Raevskaya AE et al (2008) Photochemical formation of semiconducting nanostructures. Theor Exp Chem 44:205–231. doi:10.1007/s11237-008- 9037-6 147. Kozytskiy AV, Stroyuk OL, Kuchmiy SY et al (2013) Effect of the method of preparation of ZnO/CdS and TiO2/CdS film nanoheterostructures on their photoelectrochemical properties. Theor Exp Chem 49:165–171. doi:10.1007/s11237-013-9310-1 148. Kozitskiy AV, Stroyuk OL, Kuchmiy SY (2015) Photoelectrochemical properties of a solar cell based on FTO/ZnO/CdS (photoanode) and FTO/ZnO/CuxS (counter-electrode) heterostructures. Theor Exp Chem 51:203–209. doi:10.1007/s11237-015-9417-7 149. Stroyuk OL, Kuchmiy SY, Zhukovskii MA et al (2009) Effect of the method of production of TiO2/CdS film nanoheterostructures on the effectiveness of photoinduced charge separation. Theor Exp Chem 45:302–307. doi:10.1007/s11237-009-9097-2 References 235

150. Kozytskiy AV, Stroyuk OL, Skoryk MA et al (2015) Photochemical formation and photoelectrochemical properties of TiO2/Sb2S3 heterostructures. J Photochem Photobiol A: Chem 303:8–16. doi:10.1016/j.jphotochem.2015.02.005 151. Kozytskiy AV, Stroyuk OL, Kuchmiy SY (2013) Inorganic photoelectrochemical solar cells based on nanocrystalline ZnO/ZnSe and ZnO/CuSe heterostructures. Catal Today 230:227– 233. doi:10.1016/j.cattod.2013.09.043 152. Zhou SX, Wang YG, Han W, Wang N (2008) I−V characteristics of metal−oxide−ZnSe nanowire structure. J Phys Chem C 112:18644–18650. doi:10.1021/jp8068046 153. Yang L, Xie R, Liu L et al (2011) Synthesis and characterization of ZnSe nanocrystals by W/O reverse microemulsion method: the effect of cosurfactant. J Phys Chem C 115:19507– 19512. doi:10.1021/jp204798y 154. Xu J, Yang X, Yang QD et al (2012) Arrays of CdSe sensitized ZnO/ZnSe nanocables for efficient solar cells with high open-circuit voltage. J Mater Chem 22:13374–13379. doi:10. 1039/c2jm31970e 155. Lu Z, Xu J, Xie X et al (2012) CdS/CdSe double-sensitized ZnO nanocable arrays synthesized by chemical solution method and their photovoltaic applications. J Phys Chem C 116:2656–2661. doi:10.1021/jp208254z 156. Akram MA, Javed S, Islam M et al (2016) Arrays of CZTS sensitized ZnO/ZnS and ZnO/ ZnSe core/shell nanorods for liquid junction nanowire solar cells. Sol Energy Mater Sol Cells 146:121–128. doi:10.1016/j.solmat.2015.11.034 157. Han J, Liu Z, Guo K et al (2015) AgSbS2 modified ZnO nanotube arrays for photoelectrochemical water splitting. Appl Catal B 179:61–68. doi:10.1016/j.apcatb.2015. 05.008 158. Han J, Liu Z, Guo K et al (2014) High-efficiency AgInS2-modified ZnO nanotube array photoelectrodes for all-solid-state hybrid solar cells. ACS Appl Mater Interfaces 6:17119– 17125. doi:10.1021/am5047813 159. Duan J, Zhang H, Tang Q et al (2015) Recent advances in critical materials for quantum dot-sensitized solar cells: a review. J Mater Chem A 3:17497–17510. doi:10.1039/ C5TA03280F 160. Pietryga JM, Park YS, Lim J et al (2016) Spectroscopic and device aspects of nanocrystal quantum dots. Chem Rev 116:10513–10622. doi:10.1021/acs.chemrev.6b00169 161. Sharma D, Jha R, Kumar S (2016) Quantum dot sensitized solar cells: recent advances and future perspectives in photoanode. Sol Energy Mater Sol Cells 155:294–322. doi:10.1016/j. solmat.2016.05.062 162. Thomas SR, Chen CW, Date M et al (2016) Recent developments in the synthesis of nanostructured chalcopyrite materials and their applications: a review. RSC Adv 6:60643– 60656. doi:10.1039/C6RA05502H 163. Manser JS, Saidaminov MI, Christians JA et al (2016) Making and breaking of lead halide perovskites. Acc Chem Res 49:330–338. doi:10.1021/acs.accounts.5b00455 164. Wang D, Wright M, Elumalai NK, Uddin A (2016) Stability of perovskite solar cells. Sol Energy Mater Sol Cells 147:255–275. doi:10.1016/j.solmat.2015.12.025 165. Pedesseau L, Sapori D, Traore B et al (2016) Advances and promises of layered halide hybrid perovskite semiconductors. ACS Nano 10:9776–9786. doi:10.1021/acsnano.6b05944 166. Colella S, Mazzeo M, Rizzo A et al (2016) The bright side of perovskites. J Phys Chem Lett 7:4322–4334. doi:10.1021/acs.jpclett.6b01799 167. Roose B, Ummadisingu A, Correa-Baena JP, Saliba M, Hagfeldt A, Grätzel M, Steiner U, Abate A (2017) Spontaneous crystal coalescence enables highly efficient perovskite solar cells. Nano Energy 39:24–29. doi:10.1016/j.nanoen.2017.06.037 168. Correa-Baena JP, Tress W, Domanski K, Anaraki EH, Turren-Cruz SH, Roose B, Boix PP, Grätzel M, Saliba M, Abate A (2017) Identifying and suppressing interfacial recombination to achieve high open-circuit voltage in perovskite solar cells. Energy Environ Sci 10:1207– 1212. doi:10.1039/C7EE00421D 236 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

169. Bi D, Yi C, Luo J, Decoppet JD, Zhang F, Zakeeruddin SM, Li X, Hagfeldt A, Grätzel M (2016) Polymer-templated nucleation and crystal growth of perovskite films for solar cells with efficiency greater than 21%. Nat Energy 1:16142. doi:10.1038/nenergy.2016.142 170. Divitini G, Cacovich S, Matteocci F, Cina L, Di Carlo A, Ducati C (2016) In situ observation of heat-induced degradation of perovskite solar cells. Nat Energy 1:15012. doi:10.1038/nenergy.2015.12 171. Capasso A, Matteocci F, Najafi L, Prato M, Buha J, Cina L, Pellegrini V, Di Carlo A, Bonaccorso F (2016) Few-Layer MoS2 flakes as active buffer layer for stable perovskite solar cells. Adv Energy Mater 6:1600920. doi:10.1002/aenm.201600920 172. Reiss P, Carriere M, Lincheneau C et al (2016) Synthesis of semiconductor nanocrystals, focusing on nontoxic and earth-abundant materials. Chem Rev 116:10731–10819. doi:10. 1021/acs.chemrev.6b00116 173. Aldakov D, Lefrançois A, Reiss P (2013) Ternary and quaternary metal chalcogenide nanocrystals: synthesis, properties and applications. J Mater Chem C 1:3756–3776. doi:10. 1039/C3TC30273C 174. Sato K, Ono K, Izuishi T et al (2016) The effect of CdS on the charge separation and recombination dynamics in PbS/CdS double-layered quantum dot sensitized solar cells. Chem Phys 478:159–163. doi:10.1016/j.chemphys.2016.03.014 175. Kim JY, Yang J, Yu JH et al (2015) Highly Efficient Copper indium selenide quantum dot solar cells: suppression of carrier recombination by controlled Zns overlayers. ACS Nano 9:11286–11295. doi:10.1021/acsnano.5b04917 176. Dong J, Zhu Y, Jia S, Zhu Z (2016) Blocking the back reaction in quantum dot sensitized solar cells via surface modification with organic molecules. RSC Adv 6:14224–14228. doi:10.1039/c5ra26168f 177. Huang F, Zhang Q, Xu B et al (2016) A comparison of ZnS and ZnSe passivation layers on CdS/CdSe co-sensitized quantum dot solar cells. J Mater Chem A 4:14773–14780. doi:10. 1039/c6ta01590e 178. de la Fuente MS, Sánchez RS, González-Pedro V et al (2013) Effect of organic and inorganic passivation in quantum-dot-sensitized solar cells. J Phys Chem Lett 4:1519–1525. doi:10. 1021/jz400626r 179. Lan X, Voznyy O, García Pelayo, de Arquer F et al (2016) 10.6% certified colloidal quantum dot solar cells via solvent-polarity-engineered halide passivation. Nano Lett 16:4630–4634. doi:10.1021/acs.nanolett.6b01957 180. Lan X, Voznyy O, Kiani A et al (2016) Passivation using molecular halides increases quantum dot solar cell performance. Adv Mater 28:299–304. doi:10.1002/adma.201503657 181. Zhang X, Johansson EMJ (2017) Reduction of charge recombination in PbS colloidal quantum dot solar cells at the quantum dot/ZnO interface by inserting a MgZnO buffer layer. J Mater Chem A 5:303–310. doi:10.1039/c6ta07775g 182. Hwang JY, Lee SA, Lee YH, Seok SI (2010) Improved photovoltaic response of nanocrystalline CdS-sensitized solar cells through interface control. ACS Appl Mater Interfaces 2:1343–1348. doi:10.1021/am900917n 183. Zhua G, Pana L, Xua T et al (2011) Cascade structure of TiO2/ZnO/CdS film for quantum dot sensitized solar cells. J Alloys Comp 509:7814–7818. doi:10.1016/j.jallcom.2011.05. 043 184. Chen Z, Wei C, Li S et al (2016) CdS/CdSe Co-sensitized solar cells based on hierarchically structured SnO2/TiO2 hybrid films. CdS/CdSe co-sensitized solar cells based on hierarchically structured SnO2/TiO2 hybrid films. doi:10.1186/s11671-016-1493-7 185. Kuang PY, Su YZ, Xiao K et al (2015) Double-Shelled CdS- and CdSe-cosensitized ZnO porous nanotube arrays for superior photoelectrocatalytic applications. ACS Appl Mater Interfaces 7:16387–16394. doi:10.1021/acsami.5b03527 186. Chen C, Ali G, Yoo SH et al (2011) Improved conversion efficiency of CdS quantum dot-sensitized TiO2 nanotube-arrays using CuInS2 as a co-sensitizer and an energy barrier layer. J Mater Chem 21:16430–16435. doi:10.1039/C1JM13616J References 237

187. Mahadik MA, Shinde PS, Cho M, Jang JS (2015) Fabrication of a ternary CdS/ZnIn2S4/TiO2 heterojunction for enhancing photoelectrochemical performance: effect of cascading electron–hole transfer. J Mater Chem A 3:23597–23606. doi:10.1039/c5ta07454a 188. Kokal RK, Kumar PN, Deepa M, Srivastavab AK (2015) Lead selenide quantum dots and carbon dots amplify solar conversion capability of a TiO2/CdS photoanode. J Mater Chem A 3:20715–20726. doi:10.1039/c5ta04393j 189. Chen Z, Peng W, Zhang K et al (2012) Surface ion transfer growth of ternary CdS1−xSex quantum dots and their electron transport modulation. Nanoscale 4:7690–7697. doi:10.1039/ C2NR31703F 190. Li H, Cheng C, Li X et al (2012) Composition-graded ZnxCd1−xSe@ZnO core − shell nanowire array electrodes for photoelectrochemical hydrogen generation. J Phys Chem C 116:3802–3807. doi:10.1021/jp204747w 191. Baran MP, Korsunskaya NE, Stara TR et al (2016) Graded ZnS/ZnSxO1−x heterostructures produced by oxidative photolysis of zinc sulfide: structure, optical properties and photocatalytic evolution of molecular hydrogen. J Photochem Photobiol A: Chem 329:213–220. doi:10.1016/j.jphotochem.2016.07.003 192. Wang M, Qin H, Fang Y et al (2015) FeS2-sensitized ZnO/ZnS nanorod arrays for the photoanodes of quantum-dot-sensitized solar cells. RSC Adv 5:105324–105328. doi:10. 1039/c5ra23281c 193. Ning Z, Tian H, Yuan C et al (2010) Solar cells sensitized with type-II ZnSe–CdS core/shell colloidal quantum dots. Chem Commun 47:1536–1538. doi:10.1039/c0cc03401k 194. Kim J, Yang W, Oh Y et al (2017) Self-oriented Sb2Se3 nanoneedle photocathodes for water splitting obtained by a simple spincoating method. J Mater Chem A 5:2180–2187. doi:10. 1039/c6ta09602f 195. Torimoto T, Kameyama T, Kuwabata S (2014) Photofunctional materials fabricated with chalcopyrite-type semiconductor nanoparticles composed of AgInS2 and its solid solutions. J Phys Chem Lett 5:336–347. doi:10.1021/jz402378x 196. Sasamura T, Okazaki K, Kudo A et al (2012) Photosensitization of ZnO rod electrodes with AgInS2 nanoparticles and ZnS-AgInS2 solid solution nanoparticles for solar cell applications. RSC Adv 2:552–559. doi:10.1039/c1ra00423a 197. Kadlag KP, Patil P, Rao MJ et al (2014) Luminescence and solar cell from ligand-free colloidal AgInS2 nanocrystals. CrystEngComm 16:3605–3612. doi:10.1039/c3ce42475h 198. Wang Y, Zhang Q, Li Y, Wang H (2015) Preparation of AgInS2 quantum dot/In2S3 co-sensitized photoelectrodes by a facile aqueous-phase synthesis route and their photovoltaic performance. Nanoscale 7:6185–6192. doi:10.1039/c4nr06458e 199. McDaniel H, Fuke N, Pietryga JM, Klimov VI (2013) Engineered CuInSexS2−x quantum dots for sensitized solar cells. J Phys Chem Lett 4:355–361. doi:10.1021/jz302067r 200. Guo Y, Yin X, Yang Y, Que W (2016) Construction of ZnO/Cu2SnS3 nanorod array films for enhanced photoelectrochemical and photocatalytic activity. RSC Adv 6:104041–104048. doi:10.1039/c6ra22674d 201. van Embden J, Latham K, Duffy NW, Tachibana Y (2013) Near-Infrared Absorbing Cu12Sb4S13 and Cu3SbS4 nanocrystals: synthesis, characterization, and photoelectrochemistry. J Am Chem Soc 135:11562–11571. doi:10.1021/ja402702x 202. Mitzi DB, Gunawan O, Todorov TK et al (2011) The path towards a high-performance solution-processed kesterite solar cell. Sol Energy Mater Sol Cells 95:1421–1436. doi:10. 1016/j.solmat.2010.11.028 203. Siebentritt S, Schorr S (2012) Kesterites—a challenging material for solar cells. Prog Photovolt Res Appl 20:512–519. doi:10.1002/pip.2156 204. Siebentritt S (2013) Why are kesterite solar cells not 20% efficient? Thin Solid Films 535: 1–4. doi:10.1016/j.tsf.2012.12.089 205. Liu X, Feng Y, Cui H et al (2016) The current status and future prospects of kesterite solar cells: a brief review. Prog Photovolt Res Appl 24:879–898. doi:10.1002/pip.2741 206. Guo Q, Hillhouse HW, Agrawal R (2009) Synthesis of Cu2ZnSnS4 nanocrystal ink and its use for solar cells. J Am Chem Soc 131:11672–11673. doi:10.1021/ja904981r 238 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

207. Toyoda T, Shen Q (2012) Quantum-dot-sensitized solar cells: effect of nanostructured TiO2 morphologies on photovoltaic properties. J Phys Chem Lett 3:1885–1893. doi:10.1021/ jz3004602 208. Lin L, Zhain T, Bando Y, Golberg D (2012) Recent progress of one-dimensional ZnO nanostructured solar cells. Nano Energy 1:91–106. doi:10.1016/j.nanoen.2011.10.005 209. Tiwari JN, Tiwari RN, Kim KS (2012) Zero-dimensional, one-dimensional, two-dimensional and three-dimensional nanostructured materials for advanced electrochemical energy devices. Progr Mater Sci 57:724–803. doi:10.1016/j.pmatsci.2011.08.003 210. Chen J, Li C, Eda G et al (2011) Incorporation of graphene in quantum dot sensitized solar cells based on ZnO nanorods. Chem Commun 47:6084–6086. doi:10.1039/c1cc10162e 211. Zhu Z, Qiu J, Yan K, Yang S (2013) Building high-efficiency CdS/CdSe-sensitized solar cells with a hierarchically branched double-layer architecture. ACS Appl Mater Interfaces 5:4000–4005. doi:10.1021/am400235g 212. Sun B, Hao Y, Guo F et al (2012) Fabrication of Poly(3-hexylthiophene)/CdS/ZnO coreshell nanotube array for semiconductor-sensitized solar cell. J Phys Chem C 116:1395–1400. doi:10.1021/jp206067m 213. Mukherjee B, Smith YR, Subramanian V (2012) CdSe nanocrystal assemblies on anodized TiO2 nanotubes: optical, surface, and photoelectrochemical properties. J Phys Chem C 116:15175–15184. doi:10.1021/jp208879f 214. Yang K, Xu C, Huang L et al (2011) Hybrid nanostructure heterojunction solar cells fabrica– ted using vertically aligned ZnO nanotubes grown on reduced graphene oxide. Nanotechnology 22:405401 215. Chen H, Zhu L, Liu H, Li W (2013) ITO porous film-supported metal sulfide counter electrodes for high-performance quantum-dot-sensitized solar cells. J Phys Chem C 117:3739–3746. doi:10.1021/jp309967w 216. Yang YY, Zhang QX, Wang TZ et al (2013) Novel tandem structure employing mesh-structured Cu2S counter electrode for enhanced performance of quantum dot-sensitized solar cells. Electrochim Acta 88:44–50. doi:10.1016/j.electacta.2012.09.094 217. Radich JG, Dwyer R, Kamat PV (2011) Cu2S reduced graphene oxide composite for 2− 2− high-efficiency quantum dot solar cells. overcoming the redox limitations of S /Sn at the counter electrode. J Phys Chem Lett 2:2453–2460. doi:10.1021/jz201064k 218. Balis N, Dracopoulos V, Bourikas K, Lianos P (2013) Quantum dot sensitized solar cells based on an optimized combination of ZnS, CdS and CdSe with CoS and CuS counter electrodes. Electrochim Acta 91:246–252. doi:10.1016/j.electacta.2013.01.004 219. Lin CY, Teng CY, Li TL et al (2013) Photoactive p-type PbS as a counter electrode for quantum dot-sensitized solar cells. J Mater Chem A 1:1155–1162. doi:10.1039/c2ta00251e 220. Eskandari M, Ghahary R, Shokric M, Ahmadi V (2016) Zinc oxide/copper sulfide nanorods as a highly catalytic counter electrode material for quantum dot sensitized solar cells. RSC Adv 6:51894–51899. doi:10.1039/c6ra11034g 221. Savariraj AD, Rajendrakumar G, Selvam S et al (2015) Stacked Cu1.8S nanoplatelets as counter electrode for quantum dot-sensitized solar cell. RSC Adv 5:100560–100567. doi:10. 1039/c5ra20965j 222. Guo W, Du Z, Zhao Q et al (2016) Controlled sulfidation approach for copper sulfide − carbon hybrid as an effective counter electrode in quantum-dot-sensitized solar cells. J Phys Chem C 120:16500–16506. doi:10.1021/acs.jpcc.6b05211 223. Xu J, Yang X, Wong TL, Lee CS (2012) Large-scale synthesis of Cu2SnS3 and Cu1.8S hierarchical microspheres as efficient counter electrode materials for quantum dot sensitized solar cells. Nanoscale 4:6537–6542. doi:10.1039/c2nr31724a 224. Zhang H, Yang C, Du Z et al (2017) Graphene hydrogel-based counter electrode for high efficiency quantum dot-sensitized solar cells. J Mater Chem A 5:1614–1622. doi:10.1039/ c6ta08443e 225. Venkata-Haritha M, Gopi CV, Lee YS, Kim HJ (2016) Phase transformations of novel CuxS nanostructures as highly efficient counter electrodes for stable and reproducible quantum dot-sensitized solar cells. RSC Adv 6:101185–101197. doi:10.1039/c6ra23763k References 239

226. Yang Z, Chen CY, Liu CW, Chang HT (2010) Electrocatalytic sulfur electrodes for CdS/CdSe quantum dot-sensitized solar cells. Chem Commun 46:5485–5487. doi:10.1039/ c0cc00642d 227. Faber MS, Park K, Cabán-Acevedo M et al (2013) Earth-abundant cobalt pyrite (CoS2) thin film on glass as a robust, high-performance counter electrode for quantum dot-sensitized solar cells. J Phys Chem Lett 4:1843–1849. doi:10.1021/jz400642e 228. Khalili SS, Dehghani H, Afrooz M (2017) Composite films of metal doped CoS/carbon allotropes: efficient electrocatalyst counter electrodes for high performance quantum dot-sensitized solar cells. J Colloid Interface Sci 493:32–41. doi:10.1016/j.jcis.2017.01. 005 229. Gopi C, Venkata-Haritha M, Lee Y, Kim H (2016) ZnO nanorods decorated with metal sulfides as stable and efficient counter-electrode materials for high-efficiency quantum dot-sensitized solar cells. J Mater Chem A 4:8161–8171. doi:10.1039/c6ta02415g 230. Tachan Z, Shalom M, Hod I et al (2011) PbS as a highly catalytic counter electrode for polysulfide-based quantum dot solar cells. J Phys Chem C 115:6162–6166. doi:10.1021/ jp112010m 231. Wang S, Shen T, Bai H et al (2016) Cu3Se2 nanostructure as a counter electrode for high efficiency quantum dot-sensitized solar cells. J Mater Chem C 4:8020–8026. doi:10.1039/ c6tc02309f 232. Wu M, Lin X, Wang Y, Ma T (2015) Counter electrode materials combined with redox couples in dye- and quantum dot-sensitized solar cells. J Mater Chem A 3:19638–19656. doi:10.1039/c5ta03682h 233. Kozytskiy AV, Stroyuk OL, Skoryk M, Kuchmiy SY (2015) Photoassisted formation of 2− 2− CuxS-based cathodes for CdS-sensitized solar cells with S /Sx electrolyte. Photochem Photobiol Sci 14:942–947. doi:10.1039/C4PP00314D 234. Du Z, Pan Z, Fabregat-Santiago F et al (2016) Carbon counter-electrode-based quantum-dot-sensitized solar cells with certified efficiency exceeding 11%. J Phys Chem Lett 7:3103–3111. doi:10.1021/acs.jpclett.6b01356 235. Jayathilaka K, Kapaklis V, Siripala W, Jayanetti J (2012) Sulfidation of electrodeposited microcrystalline/nano-crystalline cuprous oxide thin films for solar energy applications. Semicond Sci Technol 27:125019 236. Kozytskiy AV, Stroyuk OL, Breslavskiy AV et al (2013) Nanostructured films of CuxS— counter electrodes for solar cells based on FTO/ZnO/CdS heterostructures and sulfide/poly– sulfide redox couple. Theor Exp Chem 49:213–218. doi:10.1007/s11237-013-9317-7 237. Cao Y, Xiao Y, Jung JY et al (2013) Highly electrocatalytic Cu2ZnSn(S1−xSex)4 counter electrodes for quantum-dot-sensitized solar cells. ACS Appl Mater Interfaces 5:479–484. doi:10.1021/am302522c 238. Zhang X, Huang X, Yang Y et al (2013) Investigation on new CuInS2/Carbon composite counter electrodes for CdS/CdSe cosensitized solar cells. ACS Appl Mater Interfaces 5:5954–5960. doi:10.1021/am400268j 239. Stroyuk OL, Kuchmiy SY, Kryukov AI, Pokhodenko VD (2010) Semiconductor catalysis and photocatalysis on the nanoscale. Nova Science Publishers, Inc., New York 240. Zeng X, Zhang W, Xie Y et al (2013) Low-cost porous Cu2ZnSnSe4 film remarkably superior to noble Pt as counter electrode in quantum dot-sensitized solar cell system. J Power Sources 226:359–362. doi:10.1016/j.jpowsour.2012.11.023 241. Wang X, Liu H, Shen W (2016) Controllable in situ photo-assisted chemical deposition of CdSe quantum dots on ZnO/CdS nanorod arrays and its photovoltaic application. Nanotechnology 27:085605 242. Zhang Y, Pei Q, Liang J et al (2015) Mesoporous TiO2-based photoanode sensitized by BiOI and investigation of its photovoltaic behavior. Langmuir 31:10279–10284. doi:10.1021/acs. langmuir.5b02248 243. Wang R, Shang Y, Kanjanaboos P et al (2016) Colloidal quantum dot ligand engineering for high performance solar cells. Energy Environ Sci 9:1130–1143. doi:10.1039/C5EE03887A 240 4 Semiconductor-Based Liquid-Junction Photoelectrochemical …

244. Paul GS, Kim JH, Kim MS et al (2012) Different hierarchical nanostructured carbons as counter electrodes for CdS quantum dot solar cells. ACS Appl Mater Interfaces 4:375–381. doi:10.1021/am201452s 245. Seol M, Ramasamy E, Lee J, Yong K (2011) Highly efficient and durable quantum dot sensitized ZnO nanowire solar cell using noble-metal-free counter electrode. J Phys Chem C 115:22018–22024. doi:10.1021/jp205844r 246. Du Z, Tong J, Guo W et al (2016) Cuprous sulfide on Ni foam as a counter electrode for flexible quantum dot sensitized solar cells. J Mater Chem A 4:11754–11761. doi:10.1039/ c6ta04934f 247. Kumar PN, Kolay A, Kumar SK et al (2016) Counter electrode impact on quantum dot solar cell efficiencies. ACS Appl Mater Interfaces 8:27688–27700. doi:10.1021/acsami.6b08921 248. Sudhagar P, Ramasamy E, Cho WH et al (2011) Robust mesocellular carbon foam counter electrode for quantum-dot sensitized solar cells. Electrochem Commun 13:34–37. doi:10. 1016/j.elecom.2010.11.006 249. Jiao S, Du J, Du Z et al (2017) Nitrogen-doped mesoporous carbons as counter electrodes in quantum dot sensitized solar cells with a conversion efficiency exceeding 12%. J Phys Chem Lett 8:559–564. doi:10.1021/acs.jpclett.6b02864 Chapter 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

Photochemical activity of a semiconductor substance depends on a variety of different properties, including the spectral sensitivity range, band gap Eg, positions (potentials) of the conduction band ECB and valence band EVB, mobility of photogenerated charge carriers and density of donor and acceptor states, surface morphology, adsorption capability, etc. Only a limited number of the reported semiconductors has a “complete” set of characteristics necessary for the photocatalytic action, mostly from the AIVBVI and AIIBVI groups. The photochemical activity was broadly reported for metal oxides (mostly TiO2, ZnO and rarely— WO3,Fe2O3,SnO2,Bi2O3, etc.), metal chalcogenides (most frequently—CdS, CdxZn1−xS, ZnS and rarely—CdSe, CdTe, In2S3, HgS, MoS2, etc.), and salts of metal based acids—metallates (for example, Na2Ti2O7, NaTaO3, SrTiO3, etc.). The photochemical activity of other semiconductors is reported much scarcely. The situation, when the multiple selection criteria are met by only a limited semiconductor substances led to a dual character of the development in the synthetic aspects of the photo-active nanocrystalline semiconductors. The ﬁrst, relatively minor, direction consists in a search and testing of new semiconductors among more and more complex and exotic substances with reported or yet unreported semiconductor properties. The second, major, direction combines the studies of nanocrystalline materials with new structures but produced from “usual” (discussed above) photoactive semiconductors. This group includes nanocrystalline powders and ﬁlms, mesoporous and layered semiconductors, as well as doped nanodispersed semiconductors and composites of semiconductors with other semiconductors, metals, conjugated polymers, carbon allotropes, and other substances. Each new photocatalyst/photoelectrocatalyst and a new structure exhibit both advantages and drawbacks associated with the issues of activity, stability, technological applicability, etc. In this view, a constant exploration of new synthetic ways to the photo-active materials in both directions is of a high importance for the progress of the semiconductor-based light harvesting systems.

© Springer International Publishing AG 2018 241 O. Stroyuk, Solar Light Harvesting with Nanocrystalline Semiconductors, Lecture Notes in Chemistry 99, https://doi.org/10.1007/978-3-319-68879-4_5 242 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

5.1 Colloidal Semiconductors

Colloidal semiconductors were studied as photocatalysts/photoelectrocatalysts for more than four decades. They are popular owing to a low light scattering enabling extensive spectral characterization, relative simplicity of the synthesis and stabilization, quasi-homogeneity of the reaction systems allowing for detailed kinetic studies as well as a variety of reproducible methods of the semiconductor particle size variation by means of the “classical” colloidal chemistry. Besides, the optical transparency and homogeneity of colloidal semiconductors made them perfect models for the studies of photophysical and primary photochemical events allowing to shed light on the mechanisms of many photochemical/photocatalytic/ photoelectrochemical transformations. This section presents an overview of the syntheses and stabilization of colloidal semiconductors exhibiting photochemical activity—metal oxides, sulfides, selenides as well as some other classes. A special accent is made on the size variation of colloidal semiconductors, typically achieved by a post-synthesis treatment. All the literature sources discussed in the present chapter reported some or other photochemical/photoelectrochemical/photocatalytic process, both of endothermic nature, such as the water splitting and CO2 reduction and of the oxidative nature, in particular, decomposition of inorganic and organic compounds, the water oxidation, etc. Metal chalcogenides. Colloidal metal sulfides are typically produced in reactions between soluble metal salts and hydrogen sulfide or its soluble salts as well as with the substances releasing sulfide/hydrosulfide ions during the hydrolysis, such − − 2− as thiourea (SC(NH2)2 +2H2O + 3OH ! HS + 2NH4OH + CO3 ) or thioac- − − etamide (CH3CSNH2 +OH +2H2O ! HS +CH3COOH + NH4OH). The selenide ions are typically generated by treating selenium (2Se + N2H4 + − − 2− − 2OH ! 2HSe +N2 +2H2O) or selenite salts (4SeO3 + 3BH4 +H2O ! − − − 4HSe +H2BO3 + 4OH ) with strong reducing agents, or via the decomposition 2− − of sodium selenosulfate Na2SeSO3 in alkaline solutions (SeSO3 +OH ! − 2− HSe +SO4 ). Telluride ions can be conveniently produced by the electrochemical reduction of metallic tellurium in alkaline solutions (Te + 2e− ! Te2−). Most of the photoactive metal chalcogenides have a low solubility in water, while the reaction between chalcogenide and metal (or a metal complex) ions is typically very fast. As a result, such reactions yield highly aggregated precipitates when performed without additional substances—stabilizers. The stabilizer stops the metal chalcogenide crystal formation on the stage of nanoparticles (NPs) preventing further growth due to the strong adsorption on the NP surface [1, 2]. By the stabilization mechanism, the stabilizers can be assorted into several types, including potential-depending ions (one of the ions forming the NP lattice), organic sulfur-containing acids and alcohols, organic/inorganic polymers, as well as colloidal nanoparticulate stabilizers. 5.1 Colloidal Semiconductors 243

Stabilization of colloidal metal chalcogenides with ionic agents can be achieved by introducing an excess of the metal cations over the amount needed for the binding of the chalcogenide anions. After the metal chalcogenide formation the excessive metal ions adsorb on the surface of colloidal NPs and impart them with a positive charge. The charge creates an electrostatic barrier preventing the inter-particle interaction [2, 3]. For example, colloidal CdS particles can be stabilized by a Cd2+ excess in N,N-dimethylformamide (DMF) [4]. The stabilization of CdS NPs by excessive Cd2+ ions is broadly reported for aqueous and methanol/ ethanol solutions [5–9]. A similar stabilization effect of a Zn2+ excess was also reported for 2–5 nm ZnS particles in DMF, acetonitrile, and methanol [10, 11]. The metal ion-stabilized colloidal semiconductors typically reveal a relatively low stability that depends strongly on the temperature and solution pH as well as on the presence of other ion admixtures that can neutralize the surface charge. Much more stable colloidal metal chalcogenide NPs can be synthesized in the presence of organic mercapto-acids, mercapto-alcohols, and other bifunctional organic compounds. The mercapto- compounds can form strong covalent bonds with the under-coordinated metal ions on the NP surface. The stabilizer layer forms a steric or electrostatic (in the case of bifunctional mercapto-acids with ionized carboxyl group) barrier preventing the agglomeration of colloidal NPs. In particular, the photo-active CdS NPs were produced via the stabilization with mercaptoacetic acid (MAA) and its salts [12, 13], thiophenol [14], alkyl thiols with C6,C12 and C18 alkyl radicals [15], and mercaptoethanol [16]. Similar approaches were used to stabilize PbS [17], ZnS [18], Bi2S3 [19], In2S3 [16], and CdTe NPs [20]. A pronounced NP stabilization effect was also reported for some amino-compounds capable of forming complexes with the metal ions on the NP surface. The amine-assisted stabilization was reported for CdS [21, 22], PbS [23], and CdTe NPs [24]. Stabilization with polymers. The metal sulfide NPs can be reliably stabilized in aqueous solutions by inorganic polyanions such as sodium polyphosphate (SPP) that can strongly adsorb on the NP surface and create a dense steric/electrostatic barrier against the NP aggregation. The SPP composition can be described as (NaPO3)n with n varying in a broad range with a distribution maximum at n = 6. The polyphos- phates combine mild buffer properties, chemical stability, and inertness toward typical photochemical reactions taking place on the surface of metal-chalcogenide NPs. As a result, SPP was broadly used to produce colloidal NPs of CdS [16, 25–29], CdxZn1−xS[17], Bi2S3 and Sb2S3 [30–33], and Ag2S[34]. The bulky organic polymers such as polyvinyl alcohol (PVA), polyvinylpyrrolidone (PVP), polyethylene glycol (PEG), gelatin, polyacrylamide, peptides, polyethyleneimine (PEI) etc., can also adsorb on the surface of metal chalcogenide NPs providing a thick steric barrier preventing the NPs from contacting each other. These polymers can be used for the stabilization of metal-sulfide NPs (CdS [35–37], CdxZn1−xS[17], PbS [16, 38]) in water and in polar solvents such as acetonitrile, tetrahydrofuran, DMF or methanol. Finally, the metal-sulfide nano-photocatalysts can be stabilized by colloidal particles of inert materials, such as silica, that interact with the NPs via electrostatic forces and shield the 244 5 Synthesis of Nanocrystalline Photo-Active Semiconductors inter-particle attraction. For example, colloidal CdS NPs were stabilized by the colloidal Ludox© silica [39]. Size variation of colloidal metal chalcogenide nano-photocatalysts. As the size of metal-chalcogenide NPs is reduced, a size dependence of the photochemical properties of metal-chalcogenide NPs becomes more and more pronounced. The most distinct size dependences can be observed in the regime of strong spatial exciton confinement, typically for the sizes of 1–10 nm (the so-called quantum-sized NPs). Therefore, the size variation of colloidal NPs, as well as the focusing/defocusing of size distribution, have a paramount importance for the studies of special features of the quantum-sized NPs and for the progress of the light harvesting using nanocrystalline metal chalcogenide semiconductors in general. The synthetic approaches that were briefly discussed above provide two basic ways of NP size variation: (i) through variation of the synthesis conditions; (ii) through a post-synthesis treatment of the previously prepared NPs. The first group of methods is based on the variation of the ratio between the primary nuclei formation rate and the nuclei growth rate. The nuclei formation rate Vn depends on the oversaturation of the solution with respect to the low-soluble −1 substance [3]: Vn = kn Â (C − C0) Â C0 , where kn is the nucleation rate constant, C and C0 are the concentrations of over-saturated and saturated solutions, respectively. For most metal-chalcogenides typically C C0, and, therefore, the nuclei formation rate is constant and very high. The nucleation lowers the over-saturation and, after reaching a certain critical concentration, the formation of new nuclei stops, while the present nuclei continue to grow. The nuclei growth rate Vg can be −1 −1 expressed [3]asVg = kg Â DS(C − C0) Â d DSC Â d , where k2 is a nuclei growth rate constant, D is a diffusion coefficient, S is the NP surface area, d is a diffusion layer length where the concentration changes from C to C0. Therefore, at Vn = const, any influence that lowers Vg results in a decrease of the size of final metal-chalcogenide NPs. The typical factors are the viscosity and temperature of the solution, the concentration of the reactants, the presence of a stabilizer as well as the stabilizer type and content. An increase in the solution viscosity results in a slowing of the metal-chalcogenide monomers diffusion toward the growing NP surface and can be used as an efficient tool for affecting the NP size. For example, the absorption band edge kbe of ZnS NPs synthesized in aqueous solutions can be found at 330 nm corresponding to an average size of dav = 7 nm. As the viscosity of the solution is increased via partial water substitution with glycerol the absorption band shows a blue shift indicating a decrease of the average NP size. The ZnS NPs produced in pure glycerol are characterized by kbe = 297 nm corresponding to dav =4nm (Fig. 5.1a) [40]. A similar dependence of the average NP size on the solvent viscosity was observed for CdS NPs produced in glycerol, ethylene glycol, ethanol, and water. The average size of metal-chalcogenide NPs decreases with a lowering of the solvent temperature. The diffusion coefficient can be expressed as D = kTB−1, where B is a constant depending on the shape of colloidal NPs. As the temperature is decreased the diffusion coefficient of metal-chalcogenide monomers decreases as 5.1 Colloidal Semiconductors 245

Fig. 5.1 Absorption spectra of a colloidal ZnS NPs synthesized in water (curve 1), glycerol (curve 4) and water:glycerol mixtures with a ratio of 1:2 (curve 2) and 2:1 (curve 3); b CdSe NPs synthesized in water at 4 °C (curve 1) and 40 °C (curve 2); c CdS NPs synthesized at [CdCl2]: [Na2S] = 1:2 (curve 1) and 1:1 (curve 2), dashed lines reflect approximations of the first excitonic maxima with Gaussian curves well, resulting in a slowing of the NP growth. For example, the average size of CdS NPs stabilized by thiophenol can be tuned from 3.8 to 5.0 nm via an elevation of the solution temperature from 5 to 25 °C [14]. A temperature increase from 4 to 40 °C during the growth of SPP-stabilized CdSe NPs in aqueous solutions results in a pronounced shift of the absorption band edge from 650 to 585 nm (Fig. 5.1b) indicative of dav increase from 4 to 8 nm [41]. The average size of colloidal NPs grows also with an increase in the reactant concentration. For example, as the CdCl2 and Na2S content is elevated from 1 Â 10−4 Mto1Â 10−3 M the bandgap of CdS NPs forming in the presence of SPP decreases from 2.64 to 2.50 eV indicating a dav increase from 6.5–6.6 to *10 nm [25]. At the same time, for the metal chalcogenides a ratio between the concentrations of metal salt and chalcogenide source is typically a much more important factor than the absolute reactant concentrations. By introducing an excess of the metal or chalcogen one can strongly influence the size and optical properties of colloidal NPs in a broad range. In particular, the presence of a 100% excess of sodium sulfide during the synthesis of aqueous CdS NPs results in a *70 nm blue shift of kbe (Fig. 5.1c) attesting to a decrease of the average NP size from *9to 4 nm. The synthesis performed with a Na2S excess allows also to focus the size distribution of CdS NPs. The width of size distribution can be evaluated from an absorption maximum width. As Fig. 5.1c shows the size distribution of CdS NPs 2+ 2− decreases from dav ± 40% for the stoichiometric Cd :S ratio to dav ± 20% for colloidal solutions produced with a 100% Na2S excess. A size variation can be achieved by varying the nature and concentration of stabilizers. A dependence between the stabilizer content and the average size of metal-chalcogenide NPs was reported for the CdS NPs capped with thioglycerol [42], CdS and In2S3 NPs protected by mercaptoethanol [16], and MPA-stabilized Bi2S3 NPs [43]. In particular, by varying the metal/thioglycerol ratio the average 246 5 Synthesis of Nanocrystalline Photo-Active Semiconductors size of CdS and ZnS NPs can be tuned in the range of 3.8–7.2 nm [42] and 1.8– 3.5 nm [18], respectively. Of special interest for the NP size variation are organized micro/nano-objects, in particular, the lipid vesicles and inverted micelles. The inverted “water in oil” microemulsions can be formed by the surfactant-assisted water solubilization in non-polar organic solvents. The most precise NP size variation is reported for the inverted micellar systems based on the non-ionogenic Triton X-100 and anionic surfactant sodium bis-octadecyl sulfosuccinate (Aerosol OT or AOT, Fig. 5.2a). The size of water drops solubilized inside the AOT micelles (Fig. 5.2b) can be precisely tailored by changing the ratio of molar concentrations of water and the surfactant, w =[H2O]/[AOT]. The average radius r of the water droplets can be 3 −3 −1 calculated as r Â (r − L) =1+V2 Â (wV1) , where L is the linear size of an 3 AOT molecule (1.5 Å), V1 is the volume of a water molecule (30 Å ), V2 is the volume of an AOT molecule (825 Å3)[44]. One can tune the average size of metal chalcogenide NPs by varying the size of water droplets where the interaction between metal salts and chalcogenide sources takes place. Figure 5.2c exemplifies this approach for CdS NPs synthesized in an inverted “water/AOT/heptane” system, where a nearly linear dependence between dav and w was observed [44, 45]. At the same time, a very uniform distribution of the solubilized water droplets —“nanoreactors” by their size allows reaching a very narrow size distribution of semiconductor NPs formed in such media. The size-selected 2–8 nm CdS NPs can be extracted from the inverted micellar media by using the alkyl thiol-grafted silica, yielding visible-light-driven photocatalysts of the hydrogen evolution from water/2-propanol solutions [46]. A broad assortment of MoS2 NPs in a size range of 2–15 nm can also be produced in the AOT-based micellar media [47, 48]. An alternative way of tailoring the average size of metal-chalcogenide nano-photocatalysts is a post-synthesis treatment of raw colloidal solutions where an ensemble of differently sized NPs is present. The most frequent are the size-selective fractionation and the thermal treatment.

Fig. 5.2 Molecular structure of AOT (a), layout of an inverted “water-in-heptane” micelle (b) and dependence between the size of CdS NPs and w (plotted using the data reported in [44, 45]) 5.1 Colloidal Semiconductors 247

The size-selective fractionation can be achieved owing to a different adsorption capability of differently sized NPs (for the case of gel chromatographic separation) or to a different rate of propagation in the gel in an external electric ﬁeld. For example, the CdS NPs can be separated by the electrophoresis of an NP-rich polyacrylamide gel [27, 49]. After the electrophoresis, the gel parts containing differently sized NPs can be separated mechanically and the NPs extracted by water. The thermal treatment of polydisperse metal chalcogenide colloids results in Ostwald ripening, that is in the growth of larger NPs at the expense of the dissolution of smaller NPs [3]. The driving force of the process is a difference in the surface tension of the smaller and larger NPs. The smaller NPs with an excess of surface energy tend to dissolve creating a concentration gradient in the treated solution. The gradient results in a mass transfer from smaller to larger NPs resulting in the growth of larger NPs and complete dissolution of smaller NPs. The Ostwald ripening that focuses the NP size distribution is far from being the sole result of the thermal treatment. In a 2–5 nm particle a large portion of atoms resides on the NP surface in a partly under-coordinated and/or disordered state. These atoms become natural “traps” for the photogenerated charge carriers resulting in the radiative and non-radiative recombination competing with the photochemical reactions. The thermal treatment of colloidal solutions is accompanied by a reconstruction of the NP surface layer and the elimination (partial or sometimes complete) of such defects. The NP surface ordering results thus in an increase of the photochemical activity and aggregative stability of colloidal semiconductors. A vivid example of the thermal treatment effect is provided by aqueous colloidal CdSe NPs stabilized by SPP. Heating of the solutions at the boiling point (around 98 °C) for 2 h results in a large “red” shift of the absorption band edge indicating a considerable increase in the NP size (Fig. 5.3a) [41, 50]. The ﬁgure also shows that the thermal treatment results in a steeper absorption edge indicating a narrower size distribution of CdSe NPs in the treated colloids in accordance with the above-discussed Ostwald ripening mechanism. In the frames of this mechanism the NP volume, or r3 (r is the NP radius), increases linearly with the treatment duration t [3]. Linear dependences presented in Fig. 5.3b indeed show that the Ostwald ripening is a principal NP growth mechanism for CdSe NPs at higher temperatures [41]. At the same time, cooling of the colloidal CdSe solutions down to 4 °C allows to “freeze” the existing size distribution. Metal oxides. Colloidal metal oxides are typically produced by the hydrolysis of metal precursors with a partial or complete dehydration of an intermediary metal hydroxide. The most broadly studied semiconductor photocatalyst—titanium dioxide can be synthesized in the form of colloidal NPs via a sol-gel method based on the hydrolysis of inorganic salts (TiCl4, TiOSO4, TiOCl2, Ti(SO4)2, etc.) or organic ethers of Ti(IV) followed by the polycondensation of the intermediate hydroxy-compounds [3, 51]: 248 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

Fig. 5.3 a Absorption spectra of SPP-stabilized aqueous colloidal CdSe NPs before (curve 1) and after (curve 2) thermal treatment at *98 °C for 2 h. (b) “r3—t” dependences for CdSe NPs produced at 4 °C (curve 1), 20 °C (curve 2), and 40 °C (curve 3) [41]

ðÞ ! ðÞðÞ !ÁÁÁ! ðÞ Ti OR 4 +H2O Ti OR 3 OH + ROH Ti OH 4 ÀÁ ¼ ; ðÞ ðÞ R CH3CH2 CH3 2CH, CH3 CH2 3 ð !ð ð ð 2Ti(OR)x OH)y OH)y OR)xTi-O-Ti(OR)x OH)yÀ1 +H2O

The process starts with the formation of a colloidal TiOx sol, which then transforms into a highly viscous gel. The sol-gel transformation is completed with the gel drying into a xerogel and the xerogel annealing (or hydrothermal/microwave treatment). To produce colloidal TiO2 only two ﬁrst stages of the sol-gel transformation are needed. For example, the TiCl4 hydrolysis in water at *0 °C yields colloidal crystalline 3–5-nm titania NPs [52–60]. In a similar way, 5–10-nm ZrO2 NPs can be produced from ZrCl4 [55–57]. The hydrolysis of Ti(IV) ethers, such as titanium tetraisopropoxide (TTIP) in acidic aqueous solutions yields larger, mostly amorphous titania NPs. The crystallinity can be enhanced by carrying out the TTIP hydrolysis in dry alcohols [49, 61]. The TiO2 NPs can then be extracted by the vacuum evaporation of the solvent and redispersed in water [61, 62]. Similarly to the above-discussed metal chalcogenides, colloidal titania NPs have a partially amorphous structure and a surface abundant with defects [63] that induce inter-particle interaction and agglomeration. The stability of TiO2 colloids can be enhanced by adding amines and some organic polymers, for example, PVA [63–68]. Similar stabilization methods can be also applied to colloidal ZrO2 [66, 67] and SnO2 [69]. A synthesis of TiO2 NPs in micro-capsules of polyelectrolytes—polyallylamine and polystyrene sulfonate in the presence of the PVA stabilizer results in the formation of photocatalytic microreactors capable of the photocatalytic production of urea from CO2 and nitrate ions at the expense of PVA oxidation [70]. 5.1 Colloidal Semiconductors 249

Titania NPs were stabilized by EDTA in anhydrous 2-propanol then extracted and redispersed in aqueous solutions [71]. Dodecylsulfonate anion was found to be an efficient stabilizer of colloidal ZrO2 NPs produced from zirconium tetra-isopropoxide. The sol-gel method can be realized in a two-stage scheme, when amorphous Ti (OH)4 is first produced via the hydrolysis of TiCl4 or TTIP and then a peptizing agent is added to the precipitate, typically, HNO3 or HCl. A prolonged (several days) interaction between the precipitate and the acid at room T results in complete dissolution of Ti(OH)4 and the formation of crystalline TiO2 NPs. The size of titania NPs depends on the acid concentration and the peptization temperature and can be varied in a range of 3–40 nm [72–85]. A sol-gel synthesis in the inverted micellar solutions were used to produce ultra-small titania NPs. For example, TiO2 NPs as small as 0.5 nm were formed in the inverted “water/AOT/heptane” systems [44]. The TiCl4 hydrolysis in an inverted micellar medium formed by water, cetyl dimethyl benzyl ammonium chloride, and benzene was applied to synthesize 0.7–0.9 nm TiO2 NPs [86]. Highly crystalline colloidal titania NPs can be produced by a post-synthesis hydrothermal treatment (HTT) of the as-prepared colloids. In this method, the colloidal solutions were kept in the supercritical conditions at 150–250 °C in steel Teflon-lined autoclaves for 12–48 h. By using colloidal TiO2 produced from TTIP in a water/ethanol mixture in the presence of nitric acid as a raw material, anatase nanocrystals were produced by the HTT with the average size varying from 7 to 25 nm depending on the TTIP concentration (Fig. 5.4a) and the water/alcohol ratio (Fig. 5.4b) [87]. Zinc hydroxide is resistant to the dehydration in aqueous solutions and, therefore, ZnO NPs cannot be produced directly in water by the hydrolysis techniques [88, 89]. In view of this, colloidal ZnO NPs are typically prepared in anhydrous aliphatic alcohols (ethanol, 2-propanol) via the interaction between zinc acetate

Fig. 5.4 Size variation of TiO2 NP produced by HTT as a function of TTIP concentration (a) and water/ethanol volume ratio (b) (plotted using the data reported in [87]) 250 5 Synthesis of Nanocrystalline Photo-Active Semiconductors with sodium, potassium or lithium hydroxide [90–95]. Then the solvent is evaporated and ZnO NPs can be redispersed in water. When the Zn acetate hydrolysis in alcohols starts, very small, <1 nm, ZnO nuclei are ﬁrst formed. As the process continues, the ZnO nuclei grow into 5–6nm crystals [88, 89]. The growth can be clearly observed by a continuous red shift of the absorption band edge of ZnO colloids. To ensure the stability of forming ZnO NPs in alcohols the post-synthesis annealing at 50–60 °C is typically required resulting also in an increased crystallinity of the ZnO NPs [90–96]. By extracting portions of the colloid during the annealing several fractions of ZnO NPs can be produced with an average NP size of 2.9–4.1 nm [90]. The size of ZnO NPs in alcohols can be also varied by changing the concentration of a zinc salt and an alkali. An increase of the zinc acetate and NaOH concentration by an order of magnitude results in a red shift of the absorption band edge of ZnO NPs from 343 to 356 nm indicating an increase of dav from 3.7 to 4.4 nm [97]. Stable colloidal 3–6 nm ZnO NPs can be produced in DMSO in the interaction between zinc acetate and tetraalkyl ammonium hydroxides (Fig. 5.5)[98, 99]. The average size of ZnO nanocrystals does not depend on the nature of the tetraalkyl ammonium cation and can be tailored mainly by varying the duration and temperature of the post-synthesis thermal treatment (Fig. 5.5a). A variation of the treatment duration at a constant T or a T variation at a constant duration allows for the precise adjustment of the average size of ZnO nanocrystals in the range of 3–6nm (Fig. 5.5b).

Fig. 5.5 a Absorption spectra of ZnO/SiO2 NPs produced from ZnO NPs subjected to the thermal treatment at 60 °C for 1 min (curves 1), 5 min (curves 2), 15 min (curves 3), 60 min (curves 4), and 120 min (curves 5). b TEM images of ZnO/SiO2 NPs produced after 1 min aging before the deposition of a SiO2 shell. The scale bar is 50 nm. Insert in (b): size distribution of ZnO/SiO2 NPs 5.1 Colloidal Semiconductors 251

The growth of ZnO nanocrystals can be terminated at any time by freezing of the colloidal solution or by introducing tetraethyl orthosilicate that hydrolyzes resulting in the formation of core–shell ZnO@SiO2 nanocrystals [98, 99]. The core-shell ZnO/SiO2 NPs can be incorporated into polymer ﬁlms, e.g. PVA, and dried to powders where the nanoparticulate character of ZnO cores is preserved [99]. Also, the DMSO-based colloids can be mixed with a variety of solvents, including water and benzene.

5.2 Nanocrystalline Powdered Semiconductors

Metal chalcogenides. The nanocrystalline metal sulfide powders can be synthesized through the interactions of sulfur-containing precursors (Na2S, thiourea, thioacetamide) with metal salts, often in the presence of various structure-directing agents, such as polymers, amino-compounds, surfactants. As a rule, the synthesis is com- plemented with a HTT of metal sulfide in the form of a colloidal solution or a precipitate immersed in the parental solution. The HTT product is then separated from the supernatant, dried and annealed in an inert atmosphere. For example, the hexagonal 5–8 nm CdS crystals can be produced by the HTT of aqueous solutions of cadmium acetate and thiourea at 150 °C for 24 h [100]. By introducing ethylenediamine as a structure-directing agent the nanocrystals can be converted into CdS nanowires (NWs) with a diameter of *50 nm and a length of 3–4 lm (Fig. 5.6a) [101]. In a similar way, diethylenetriamine was used to produce ZnS nanorods (NRs) [102]. In the presence of PVP and the same HTT conditions, the process yields spherical 1.5–2.0 lm agglomerates composed of 12-nm ZnS NPs [103]. By introducing a mixture of two metals, ternary cadmium-indium-sulfide and zinc-indium-sulfide nanopowders can be produced [104]. In the case of CdIn2S4, the HTT in aqueous solutions produces spherical NPs, while in methanol nanotubes (NTs) with a diameter of 25 nm are predominantly formed [105]. The nanocrystalline ZnIn2S4 powders were synthesized by the HTT of aqueous solutions of metal salts, thioacetamide, and a surfactant—cetyl trimethyl ammonium bromide or sodium dodecyl sulfonate [106, 107]. The reaction yields porous hierarchical ZnIn2S4 microspheres with a size of 100–400 nm composed of *20-nm units (Fig. 5.6b). After the photodeposition of Pt NPs, such materials act as excellent photocatalysts of the hydrogen evolution from aqueous sulfide/sulfite solutions [105–109] as discussed in details in Chap. 2. Similar photocatalytic properties were observed for the nanocrystalline Cu2WS4 (Fig. 5.6c) and CdLa2S4 powders produced by the HTT [110, 111]. Nanocrystalline powdered metal selenides were obtained via the reduction of selenites and selenates of various metals, or through the hydrothermal decomposition of metal selenosulfates. For example, the HTT of aqueous solutions containing zinc nitrate, Na2SeSO3,N2H4 and EDTA at 180 °C for 2 h yields nanocrystalline ZnSe powders (Fig. 5.6d) [112]. 252 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

Fig. 5.6 SEM images of CdS nanowires (a), ZnIn2S4 microspheres (b), Cu2WS4 nanoplates (c), and ZnSe nano-“stars” (d). Reprinted with permissions from Refs. [101](a) and [112](d), copyright (2007, 2008) American Chemical Society; [106](b) and [110](c), copyright (2008, 2010) Elsevier

Metal oxides. The photocatalytic properties of nanocrystalline metal oxides depend on a broad variety of factors, in particular, on the NP size and crystal structure, surface area and acidity, porosity, density of surface hydroxyl groups and structural defects, etc. [113]. The synthetic approaches to nanocrystalline metal oxides should, therefore, provide broad and reliable means of varying these parameters in a controllable manner. As discussed above, nanocrystalline titanium dioxide is typically produced by the hydrolysis of salts and alkoxides of Ti(IV), drying of the forming gel and the HTT/annealing of the xerogel. This general scheme is illustrated in Fig. 5.7 and can include a broad variety of additional steps and options. For example, the hydrolysis of TiCl4 in cold diluted aqueous acid solutions followed by the aging at 80 °C for 24 h yields the nanocrystalline mixtures of anatase (3–4 nm) and rutile (13–15 nm) [114]. The oxidative TiCl3 hydrolysis in the presence of ammonia and H2O2 and annealing at 300 °C produces the nanocrystalline anatase powders with an average grain size of 13 nm [115]. Adsorption of TTIP on amorphous carbon followed by the hydrolysis with the air moisture and annealing at 400–600 °C can be used to 5.2 Nanocrystalline Powdered Semiconductors 253

Fig. 5.7 Layout of the sol-gel synthesis of nanocrystalline titania photocatalysts

form 10–20-nm titania crystals for the photocatalytic hydrogen evolution from water/ethanol mixtures [116]. The nanocrystalline titania with a grain size of 15 nm was prepared through the TTIP hydrolysis in the presence of ice acetic and nitric acids followed by the HTT at 220 °C for 12 h [117]. The hydrothermal treatment of aqueous solutions of TiOSO4,H2TiO(C2O4)2 or TiO(NO3)2 results in the nanocrystalline TiO2 powders with a NP size of 20–50 nm [118]. Monodisperse titania nanocrystals were synthesized by a multiple repetition of the TiCl4 addition to hot (80 °C) water and the neutralization of released acetic acid with ammonia. The treatment results in the dissolution of smaller NPs allowing to focus the size distribution of titania NP ensemble. The ﬁnal size of TiO2 NPs grows gradually from 17 to 22 nm as the number of repetitions of the procedure is increased from 7 to 30 [119, 120]. The principal factors affecting the phase composition and grain size of nanocrystalline titania powders formed by the HTT are the temperature and duration of the treatment [121, 122], as well as the parental solution acidity [123, 124]. Figure 5.8a shows a scheme of the formation of three typical titania phases from the individual TiO6 octahedra [125]. As the HTT temperature is increased from 100 to 300 °C the average size of titania nanocrystals grows from 7 to 18 nm (Fig. 5.8b) [121]. 254 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

Fig. 5.8 a Formation of three typical titania phases from the individual TiO6 octahedra. Reprinted with permissions from Ref. [125]. Copyright (2008) Elsevier. b The average size of TiO2 nanocrystals as a function of the HTT temperature (plotted using the data reported in [121])

The phase composition of titania nanocrystals depends on the grain size. The smallest reported TiO2 NPs (5–10 nm) crystallize in the anatase modification, larger NPs reveal a structure of brookite, while even larger crystals can be typically found as a rutile admixture [124, 125]. The Ti(IV) hydrolysis in strongly acidic solutions results mostly in nanocrystalline rutile. The rutile NRs can be also produced via HTT of TiCl4 solutions containing chloroform [126]. A prolonged (around 3 months) aging of titanate NTs in diluted aqueous solutions of mineral acids (HCl, H2SO4) results in the NT conversion into ultrasmall (around 3 nm) rutile nanocrystals [127]. The HTT of protonated titanates, such as H2Ti4O9 Â 1.2H2O in acid solutions yields titania in the form of nanocubes and NRs with a size depending on the HCl concentration [128]. Along with the HTT temperature and duration the structural characteristics of nanocrystalline TiO2 can also be affected by the composition and pressure of gaseous phase present in the autoclave. For example, HTT of a titania gel at 80 °C in deaerated conditions yields aggregated 30-nm titania crystals. However, in the presence of air with other conditions being equal the synthesis results in TiO2 NRs with a length of up to 200 nm [129]. The microwave treatment of titania gels is an alternative to HTT and allows to enhance the crystallinity of the products and to prepare smaller NPs. As reported in [130], the microwave heating of a TiO2 gel applied instead of the conventional HTT results in a decrease of the grain size from 30–40 to 15–20 nm. The growth of TiO2 NPs during the thermal treatment can be retarded by the introduction of compounds capable of the crystallization simultaneously with titania, for example, SiO2. The nanocrystalline TiO2/SiO2 powders can be formed by the simultaneous hydrolysis of titanium tetra-butoxide (TBT) and tetraethylorthosilicate 5.2 Nanocrystalline Powdered Semiconductors 255 or produced by mixing separately prepared TiO2 and SiO2 sols [131, 132]. The silica present in such composites inhibits the growth of TiO2 NPs and suppresses the transformation of anatase into rutile during the annealing [131, 133–136]. As the silica fraction in TiO2/SiO2 powders is elevated to 0.3 the average size of anatase crystals decreases from 17 to 4 nm, while the specific surface area grows from 32 to 240 m2/g (Fig. 5.9a) [131]. Nanocrystalline TiO2/SiO2 powders can also be produced by the hydrolysis of Ti (IV) alkoxides on the surface of 150–160 nm silica nanobeads. The final size of titania NPs in such heterostructures depends on the annealing temperature growing from 4–5to11–12 nm with an increase in the calculation T from 500–600 to 1000 °C [137] (Fig. 5.9b). The growth of TiO2 NPs can also be restricted by performing the Ti(IV) hydrolysis in the presence of surfactants and bulky organic molecules. For example, nanocrystalline titania can be prepared by the hydrolysis of TBT in ethanol in the presence of dodecylamine at 40–80 °C [138]. The amine is then removed during the calcination at 350–500 °C. The introduction of surfactants allows also to influence the shape of the TiO2 nanocrystals. The hydrolysis of TiCl4 or TBT in the presence of sodium dodecyl benzyl sulfonate followed by the calcination yields elongated ellipsoid crystals with a length of up to 500 nm, while by using sodium dodecyl sulfonate or hydrox- ypropyl cellulose TiO2 nanocubes and NRs, respectively, can be synthesized [139]. Nanocrystalline titania is often produced with the triblock copolymers such as Pluronic 123 [140–142]. The polymer composition can be described as (EO)20– (PO)70–(EO)20, where EO and PO are ethylene oxide (–CH2CH2O–) and propylene oxide (–CH2(CH3)2CHO–) fragments. The method yields rutile NRs with a diameter of 10–20 nm and a length of 100–200 nm with the surface-anchored 5– 15-nm anatase NPs [143]. In the case of nanocrystalline titania produced by the oxidative TiCl3 hydrolysis in ethanol in the presence of PEGs the size and phase composition of

Fig. 5.9 Average size of TiO2 nanocrystallites in the TiO2/SiO2 heterostructures (a), curve 1; (b) and the surface are of TiO2/SiO2 composite ((a), curve 2) as functions of the silica fraction (a) and the annealing temperature (b). Plotted using the data reported in [131](a) and [137](b) 256 5 Synthesis of Nanocrystalline Photo-Active Semiconductors nano-dispersed products depend on the molecular mass M of the polymer. In a range of M = 200–4000 g/mole the synthesis yields anatase crystals with a size of 35–75 nm (Fig. 5.10a), while at a higher M brookite nanocrystals as large as 70– 80 nm can be observed [144]. The hydrolysis of Ti(IV) alkoxides in anhydrous alcohols occurs much slower than in water allowing to achieve a high crystallinity of the products without a solvothermal treatment. For example, the sol-gel TTIP transformation in ethanol/ 2-propanol followed by the gel drying at 70 °C and the xerogel annealing at 200– 500 °C resulted in 7–12 nm titania crystals with a preferential (90%) anatase lattice [145]. The size of titania NPs can be controlled when Ti(IV) compounds are subjected to the hydrolysis in micellar solutions. Such approach was applied to produce 15-nm anatase crystals in water droplets solubilized by hexadecyl trimethyl ammonium bromide in cyclohexanol [146]. The simultaneous hydrolysis of TTIP and ammonium vanadyl in a micellar “water/Triton X-100/heptane” system followed by the thermal treatment produced 8–13 nm crystals of mixed vanadium-titanium oxide [147]. Highly dispersed titania powders are frequently produced by the pyrolytic methods. For example, anatase nanocrystals can be synthesized by burning TiCl4 in a hydrogen/oxygen mixture [148]. By varying the TiCl4 feed rate the particle size can be tuned in a range of 15–30 nm, while by adjusting the rate of the gas mixture, the oxygen content and the ﬂame temperature in a range of 1400–1700 °C one can vary the anatase fraction from 40 to 80 wt.%. Nanocrystalline titania is often prepared by using self-igniting mixtures that include a Ti(IV) precursor, oxidant, and a substance-fuel. The ﬁrst two components are often present in a single compound—titanyl nitrate TiO(NO3)2, while the fuel role is typically delegated to various amino-acids and polyalcohols. Such mixtures ignite when heated to 300–350 °C and burn steadily with the formation of highly

Fig. 5.10 a Titania nanocrystal size as a function of the molecular mass (values on the graph) of PEG present at the oxidative TiCl3 hydrolysis; b ZnO NP size as a function of the HTT temperature. Plotted using the data reported in [144](a) and [158](b) 5.2 Nanocrystalline Powdered Semiconductors 257 dispersed TiO2. A popular combination for the self-igniting mixtures is titanyl nitrate —glycine [149–153] that burns as described by the following brutto-equation: 9TiO (NO3)2 +10H2NCH2COOH ! 9TiO2# +14N2 +20CO2 +25H2O. In the recent years, the synthesis of titania nanofibers and nanowires (NWs) has come into the focus as such materials find broad applications for the preparation of photoactive paper, tissues, and reinforced polymers. The TiO2 NWs are typically prepared using PVP fibers [154] or layered protonated titanates [155] as templates. The titania NRs and NWs can also be formed by the electrophoretic deposition of TiO2 NPs into the anodized alumina pores followed by the calcination and final dissolution of the Al2O3 matrix [156]. Similarly to titania, nanocrystalline ZnO is most often produced using the hydrothermal syntheses. For example, the HTT of zinc acetate solution in ethanol at 90 °C for 12 h produces elongated ZnO (40 Â 80 nm) crystals [157]. As the HTT temperature is elevated from 150 to 500 °C the average size of ZnO nanocrystals grows from 10 to 90 nm (Fig. 5.10b) [158]. The HTT of aqueous solutions containing zinc acetate and histidine at 150 °C yields a variety of products depending on the reactants concentration—nanoprisms, hollow microspheres, NR aggregates, etc. [159]. Nanocrystalline ZnO was synthesized by burning a solution containing glycine and Zn(OH)2 at 1500–1800 °C [160], as well as by the thermal decomposition of zinc hydroxycarbonate [161] and Zn[N(SiCH3)2]2 [162]. The mechanochemical treatment of ZnCl2 and Na2CO3 diluted with NaCl yields nanocrystalline ZnO with a grain size of 30–60 nm [163]. The zinc oxide NRs/NWs on substrates are typically produced by a two-stage “seeding” method [164]. On the first stage, a layer of an aqueous solution of zinc nitrate and hexamethylenetetramine is deposited onto a rotating plate and annealed at around 200 °C. The ZnO NPs produced by this method serve as nuclei for the formation of ZnO NRs during the HTT of the substrate plate in the same solution at 95 °C for 10–30 h. The ZnO NRs on substrates can also be conveniently produced by the thermal evaporation/redeposition (Fig. 5.11a) [165].

Fig. 5.11 Electron microscopic images of ZnO NRs on Zn foil (a), macroporous WO3 (b), BiOI microspheres (c). Reprinted with permissions from Refs. [165](a), [166](b), and [170](c). Copyright (2008) American Chemical Society (a, c) and The Royal Society of Chemistry (b) 258 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

The synthetic methods described for TiO2 and ZnO can be extended to other oxide and metallate semiconductors. A set of examples is provided below illustrating a general character of the methods of synthesis of thermally stable photo-active metal oxides. The calcination of closely packed polystyrene microspheres soaked with a (NH4)6H2W12O40 solution was used to synthesize macroporous nanocrystalline WO3 (Fig. 5.11b) [166]. The WO3 NRs with a diameter of *8 nm can be formed by the decomposition of ammonium metatungstate in the pores of SBA-15 zeolite [167]. The WO3 NW arrays formed by a solvothermal synthesis on the conductive glass can be used as the visible-light-sensitive photoanodes of solar cells [168]. The dehydration of Bi(OH)3 under the ultrasound irradiation in the presence of PVP yields nanocrystalline Bi2O3 with a grain size of 40–100 nm [169]. The HTT of aqueous solution containing bismuth(III) nitrate, sodium halogenide (NaCl, NaBr, NaI) and ethylene glycol results in platelet-shaped BiOX (X = Cl, Br, I) nanocrystals [170, 171]. The HTT of BiOI nanoplates in aqueous ammonia solutions induces in their agglomeration into microspheres (Fig. 5.11c) with a much higher photoactivity as compared to original nanoplates [172]. Metal titanates and zirconates are typically produced by the co-hydrolysis of Ti (IV) or Zr(IV) with the corresponding metal salts followed by the HTT of gels or precipitates. This method was applied for the preparation of nanocrystalline SrTiO3 with a grain size of 8–10 nm [173]. Of the three syntheses of nanocrystalline strontium titanate—the co-hydrolysis of Ti(IV) and Sr(II) compounds (a final NP size of 30 nm), a solid-state high-temperature reaction (140 nm), and the mechanochemical treatment of a precursor mixture (30 nm), it is the co-hydrolysis that yielded the most efficient photocatalysts of the H2 evolution from water/methanol mixtures [174]. Nanocrystalline potassium niobate K4Nb6O17 was produced by the HTT of zirconium oxide in highly basic solutions (1.0 M KOH) at 150–200 °C (Fig. 5.12) [175–177]. To increase the photo-activity of the layered material it was intercalated with tetrabutyl ammonium hydroxide that expands the interlayer galleries making them more accessible for reactants [177]. Metal tungstates are typically produced from tungstic acid or related alkali metal salts, often introduced as the polynuclear compounds. For example, the interaction between Bi2O3 and (NH4)2O Â 12WO3 Â 5H2O in the presence of diethylenetri- aminepentaacetic acid yields nanocrystalline Bi2WO6 powders [178]. Sometimes, this method is modified by the microwave [179, 180] or solvothermal treatment [180] to produce more active photocatalysts. The HTT of solutions containing Bi(NO3)3 and sodium tungstate results in plate-shaped Bi2WO6 nanocrystals [181, 182]. The nanoplate thickness depends on the HTT temperature and duration (Fig. 5.13a) and increases from 20 to 100 nm as the HTT is prolonged from 4 to 24 h. The HTT of a solution of Na2WO6, Zn(NO3)2 and cetyl trimethyl ammonium bromide in a concentration high enough for the formation of cylindric micelles, at 120–140 °C results in rod-shaped Zn2WO4 nanocrystals [183]. At a higher HTT 5.2 Nanocrystalline Powdered Semiconductors 259

Fig. 5.12 TEM images of niobate nanosheets exfoliated by using propylamine (PA) hydrochloride (a) and niobate nanoscrolls (b) produced with tetrabuthyl ammonium (TBA) hydroxide; (c) scheme of K4Nb6O17 niobate nanosheets and nanoscrolls synthesis. Reprinted with permissions from Ref. [177]. Copyright (2008) Elsevier

temperature the synthesis yields conventional Zn2WO4 nanocrystals showing no shape anisotropy. The size of such nanocrystals, similarly to Bi2WO4, can be tuned by varying the HTT duration (Fig. 5.13b). Nanocrystalline vanadates can be produced using the whole assortment of synthetic approaches discussed above for oxides and other metallates. For example, plate-shaped BiVO4 nanocrystals with a thickness of 10–40 nm can be produced by the HTT of solutions containing ammonium vanadate, bismuth nitrate, and a structure-directing agent—sodium dodecyl benzyl sulfonate [184]. The microwave heating of a mixture of cerium nitrate and NH4VO4 in the presence of PEG-600 yields nanocrystalline cerium orthovanadate with a particle size of 15–20 nm [185]. 260 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

Fig. 5.13 Thickness of Bi2WO4 nanoplates (a) and size of Zn2WO4 nanocrystals (b) as a function of the HTT duration (plotted using the data reported in [182](a) and [183](b)

5.3 Nanocrystalline Films of Photo-Active Semiconductors

One of the most popular methods of the preparation of porous nanocrystalline semiconductor ﬁlms consists in a slow (with a rate of 1–2 mm per min) extraction of a substrate (typically a glass plate, a conductive glass plate—ITO, FTO, a metal foil of polymer ﬁbers) from a colloidal solution of semiconductor NPs that contains simultaneously pore-forming polymeric additives and other functional components, in particular those increasing the solution viscosity (Fig. 5.14a). The method is typically referred to as the “dip coating” [3].

Fig. 5.14 a Synthesis of nanocrystalline TiO2 on glass. b Dependence of the thickness and surface area of the titania ﬁlm on the dip coating repetition number (plotted using the data reported in [195]) 5.3 Nanocrystalline Films of Photo-Active Semiconductors 261

The film thickness depends on the number of dip coating repetition, while the parameters of porous structure can be affected by the type and size of pore-forming additive and the thermal treatment conditions. The drying and calcination of dip-coated films result in nanoporous coatings formed by loosely aggregated semiconductor NPs. For example, titania films were prepared from TTIP or TiCl4 in the presence of Triton X-100 [186–190], Pluronic 123 [191] and F127 [192], non-ionogenic Tween 20 [193], methyl cellulose [194], terpineol with hydrox- ypropyl cellulose [195], PVA [196], and PEG [197]. In the latter case, a linear dependence between the dip coating repetition number and the film thickness was observed—the latter increased from 0.9 to 5.4 lm as the repetition number was elevated from 5 to 30 [195] (Fig. 5.14b). The dip coating can be performed from a titania sol synthesized in an inverted micellar solution «water/Triton X-100/hexane» [128], a suspension of TiO2 NPs produced by the HTT and dispersed in water by the ultrasound treatment [197], and SnO2 sols synthesized via the SnCl4 hydrolysis in 2-propanol [198]. Relatively thick (*10 lm) and optically transparent titania films can be produced from TiO2 sols TM mixed with highly viscous PEG Carbowax [199]. The films of TiO2 [200–202], WO3 [203] and Bi2MoO6 [204] can be dip-coated onto wafers and optical fibers. Negatively charged subnanometer titania sheets can be produced via the exfoliation of alkali metal titanates assisted by the intercalation of tetrabutyl ammonium cations. In the presence of a cationic surfactant—octadecyl ammonium chloride these sheets accumulate on the solution/air interface and can be dip-coated on glass producing ultrathin (0.75 nm) TiO2 Â nH2O films with a lateral size of up to 1 lm [205]. Another popular method of the formation of nanocrystalline films from colloidal semiconductors was christened “spin coating”. In this approach, the colloidal solution/suspension is deposited onto a rapidly rotating substrate (Fig. 5.15). Under a combination of the centrifugal and centripetal forces, the solution is distributed evenly on the substrate surface [3]. The film thickness depends mainly on the rotation speed. In this way, photo-active films are frequently produced from the commercial nanocrystalline TiO2 Evonik P25 [206–208] or titania sols with added viscous agents [82, 209–212]. For example, photo-active porous nanocrystalline TiO2 can be synthesized using carbon nanospheres produced by the HTT of glucose solutions as a pore-forming agent [213] as well as conventional PEGs [214]. Nanocrystalline titania can be formed as films in mild conditions by the liquid phase deposition. In a typical procedure, a glass plate is immersed into a solution, 2− where the hydrolysis of TiF4 [215] or TiF6 [216, 217] occurs in the presence of − boric acid that binds fluoride ion in the form of BF4 . Ordered films of TiO2 on quartz were produced by multiple repetitions of the successive adsorption of negatively charged titania NPs and positively charged polydiallyl dimethyl ammonium cations [218]. Porous films of titania form as a result of the NP deposition in an external electric field. For example, the electric field applied to a suspension of TiO2 (Evonik P25) with PVA results in the opening of channels between the polymer 262 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

Fig. 5.15 Synthesis of nanocrystalline titania films on glass by spin coating globules favoring to the formation of inner branched pores in the final annealed titania films [219]. Photoactive nanocrystalline films can be deposited from the gas phase using exclusively physical processes or their combinations with chemical reactions [3]. In the first group of methods, the growth of nanocrystalline films results from the substance condensation on the surface of a substrate in the form of NPs, while in the second case it occurs via the decomposition (oxidation) of a gaseous precursor and the product deposition on the surface of a heated substrate. Of special interest are methods of mild deposition allowing to form photo-active coatings on thermally unstable substrates. For example, such films can be deposited via the bombardment with TiO2 clusters accelerated by an electron beam [220]. The clusters are produced distantly from the substrate at the titanium vapor oxidation with O2 (Fig. 5.16a). Also, the films can be formed by the gas-phase “cold” deposition. This method was realized for titania aerosol particles caught and transported by a nitrogen stream at a rate of 300–1200 m/s [221, 222]. Photoactive titania films were produced by the thermal decomposition of TiCl4 [206, 207] or titanyl 2,4-penthadionate [223]. The consecutive pyrolysis of SnCl4 and titanyl acetylacetonate yields mixed nanocrystalline SnO2/TiO2 with a thickness up to *800 nm [224]. Thin-film photoelectrodes for the water splitting were synthesized by the flame aerosol deposition [225]. The pyrolysis of aerosol produced from a solution of zinc and indium chlorides with thiourea admixtures was used to form nanocrystalline 5.3 Nanocrystalline Films of Photo-Active Semiconductors 263

Fig. 5.16 Formation of TiO2 ﬁlms on thermally unstable substrates via the bombardment with + charged (TiO2)n clusters [220](a) and via the “cold” deposition (b)[221, 222]

ZnIn2S4 films that were successfully applied as a photocatalyst of the water reduction [226]. The atomic layer deposition is often used for films preparation on porous substrates [3, 227]. In this method, one of the precursors is injected into the reaction chamber, then the chamber is cleaned with a gas carrier, another precursor is injected, and the chamber is flooded with the inert gas again (Fig. 5.16b). The growth of a nanocrystalline film occurs due to alternating surface reactions in the state of adsorption saturation, resulting in uniform and structurally perfect films. Tungsten oxide [228] and iron oxide [229] films produced by the atomic layer deposition were used as electrodes in the photoelectrochemical water-splitting cells. The nanocrystalline titania films formed by the magnetron sputtering on glass and metal substrates can also be applied as photocatalysts of this process [230, 231]. A post-synthesis etching of such films in HF solutions results in an enhancement of the light conversion efficiency due to an increase of the surface area and the density of donor centers. A combined gas-phase sputtering of TiO2 and Pt on different sides of titanium foils yielded photoelectrodes for the water splitting, where the oxidation of water to O2 occurs on the nanocrystalline TiO2 film, while the water reduction to H2 takes place on platinum [232]. A similar method was applied to form nanocrystalline titania films with a gradient O/Ti ratio increasing from 1.93 near the substrate surface to 2.00—on the outer film surface [233]. The presence of Ti(III) in such films makes them an active photocatalyst of the hydrogen evolution from water/methanol mixtures under the Vis-illumination [233]. 264 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

The laser pulse deposition allows forming nanocrystalline semiconductor films at relatively low temperatures. In particular, this method was applied to produce CaFe2O4 films showing photoelectrocatalytic properties in the hydrogen evolution [234]. The electrodeposition is broadly used to produce films on various conductive substrates. For example, electrodeposited hematite a-Fe2O3 films doped with Pt [235], Cr or Mo [236] were used for the water photoelectrolysis. Nanocrystalline Cu2O films can be deposited onto conductive glasses (ITO) from solutions containing copper sulfate and lactic acid [237]. The electrodeposition was also applied to produce ZnIn2S4 films with a grain size of around 20 nm from solutions of zinc and indium chlorides and Na2S2O3 [238]. An anodic treatment of mixed Al/Ti film in a solution of phosphoric and oxalic acids results in the formation of porous nanocrystalline Al2O3/TiO2 films ready for the photocatalytic applications [239]. An electrochemical etching of silicon yields nanocrystalline Si films with a particle size of 1–4 nm that were successfully used for the photocatalytic CO2 reduction [240]. The electrochemical oxidation of a silver layer in chloride-containing electrolytes resulted in the nanocrystalline AgCl films [241].

5.4 Mesoporous Photo-Active Semiconductor Nanomaterials

The smaller are the semiconductor crystals the higher is their surface energy. The nanocrystals are thermodynamically non-stable and tend to lower their surface energy via the agglomeration. This process is typically unfavorable for the synthesis of nano-photocatalysts and is inhibited by introducing various stabilizers that adsorb on the NP surface and lower the surface tension. However, the tendency of semiconductor NPs to agglomeration can be used for the preparation of a special kind of porous photo-active materials. For this aim, special substances are added to the reaction mixtures that do not inhibit the agglomeration completely but prevent the NPs from the conversion into a dense body during the annealing. As a result, the synthesis yields framework structures formed by loosely aggregated semiconductor nanocrystals with nanometer voids between the separate NPs—the mesoporous materials [242]. The pore size can be tailored by performing the synthesis in the presence of surfactant micelles or polymer globules of a given size that, after being destructed during the calcination, leave the voids of a corresponding size. To form a mesoporous structure both the size of semiconductor NPs and the pore size should be in a nanometer range [242]. The mesoporous semiconductor materials reveal a set of features very favorable for the photocatalysis and solar cell applications [243]. They have a high specific surface area (100–300 m2/g) and a developed system of mesopores that can adsorb reactants in the inner pore volume and accumulate them till as long as the capillary 5.4 Mesoporous Photo-Active Semiconductor Nanomaterials 265 condensation occurs. The high pore curvature allows for the strong adsorption and long retention of reactants in the pores allowing for a more efficient contact with the photogenerated charge carriers as compared to the non-porous or macroporous materials. As the mesoporous materials are formed by the NP building blocks each contacting with many neighboring NPs, there exists a unique opportunity for the migration of the photogenerated charge carriers along the NP network decreasing the probability of recombination losses. Finally, in the case of hollow microspheres with mesoporous walls an effect of the multiple light scattering and reflection in the inner voids can emerge favoring to a higher light absorption. As the annealing of mesoporous material precursors is obligatory in almost all cases for the elimination of pore-forming agents, the assortment of reported mesoporous materials is confined mostly to thermally stable oxide semiconductors [242]. For the production of mesoporous metal chalcogenide materials, alternative methods are typically proposed, such as the anchoring of the metal sulfide NPs on the inner walls of zeolites and membranes followed with the host dissolution, the ultrasound treatment of colloidal solutions, the photochemical methods [244], etc. Mesoporous oxide photocatalysts. The mesoporous metal oxides are typically produced by the hydrolysis of corresponding precursors in the presence of various functional additives followed by the HTT, drying and final calcination. A broad variety of surfactants and polymers were reported as pore-forming additives for the synthesis of mesoporous TiO2 and related nano-heterostructures (TiO2/SiO2 [245], TiO2/ZrO2 [246], TiO2/VOx [247], TiO2/In2O3 [248], etc.) including Pluronic family (Fig. 5.17a) [246, 249–255], agarose (Fig. 5.17b) [256]), Triton X-100 [257], Tween surfactants [250], PEGs [250, 258], polyatomic alcohols (glucose, sorbitol [247]), bulky organic molecules such as 4,4′-dioxydiphenylpropane [259], triethanolamine [260]), dodecyl- and tetradecyl amine [261], lauril amine [262], cetyl trimethyl ammonium bromide [245, 263–267], alkyl phosphates (C12–C18) [268], etc.

Fig. 5.17 TEM (a) and SEM (b, c) images of mesoporous titania produced using Pluronic 123 (a) and agarose (b), CdxZn1−xS microspheres (c). Reprinted with permissions from Refs. [249](a), [256](b), and [282](c). Copyright (2008) Elsevier (a) and American Chemical Society (b, c) 266 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

The textural characteristics of mesoporous titania can be varied by changing the nature/size of pore-forming agents, and the conditions of HTT and ﬁnal calcination. For example, an average size of NPs forming the framework of mesoporous TiO2 produced with PEGs decreases from 17 to 13–14 nm with an increase in the average molecular mass of the polymer from 200 to 20,000 g/mole (Fig. 5.18a). As the calcination temperature is elevated from 350 to 650 °C the NP size increases from 10 to 35–37 nm (Fig. 5.18b, curve 1), while the pore size grows from 7–8to 15–16 nm (Fig. 5.18b, curve 2) [258]. The mesoporous titania with a particle size of 18–20 nm and a surface area of 70 m2/g was produced by the TTIP hydrolysis in the presence of HCl followed by the gel aging at 80 °C during 12 h and the calcination at 450 °C [269]. The material with 2–20 nm pores was synthesized from a mixed TiO2/SiO2 oxide via the selective silica dissolution in HF solutions [270]. The TBT hydrolysis in the presence of acetylacetone during the HTT results in 100–200-nm TiO2 NRs. Titania NP frameworks form with a NP size of 5–10 nm and a pore size of 7–12 nm can be observed on the NR surface [271]. A synthesis in water/ethanol solutions yields porous materials with a bimodal pore size distribution—2–8 nm for the intra-aggregate pores and 40–50 nm for the pores between aggregates of 5–10 nm TiO2 particles [272, 273]. The high-intensity ultrasound treatment during the TTIP hydrolysis accelerates the nuclei formation and favors to their condensation and agglomeration into mesoporous 100–200-nm microspheres composed of 10 nm titania crystals separated by 4 nm pores [274]. A mesoporous material composed of nitrogen-doped TiO2 nanocrystals was produced by the calcination of a precipitate forming in a reaction between NH4OH 2− and a complex [TiO(C2O4)2] ion [275]. The size of titania NPs and pores in the mesoporous framework depend distinctly on the calcination temperature.

Fig. 5.18 Particle size (a and b, curve 1) and pore diameter (b, curve 2) of mesoporous TiO2 produced in the presence of PEGs of a different molecular weight (a) and at a different calcination temperature (b) (plotted using the data reported in [258]) 5.4 Mesoporous Photo-Active Semiconductor Nanomaterials 267

Similar methods were used to produce other mesoporous oxides for the photocatalytic and photoelectrochemical applications, for example, Ta2O5 and Nb2O5.In contrast to titania, the porous structure of mesoporous tantalum and niobium oxides is not stable at higher T and can collapse during the annealing. To preserve the mesoporous framework these oxides are annealed in a two-stage regime—at a lower T (350 °C) to eliminate the organic pore-forming agent and at a higher T (500 °C)—to allow for the oxide crystallization [276]. Alternatively, silica NPs can be introduced into the mesoporous oxides enhancing their stability at elevated T. The SiO2 NPs can then be removed by the chemical etching [277]. Such methods yield mesoporous materials with an oxide NP size of 8–12 nm and a pore diameter of 4–5 nm that revealed photocatalytic properties in the hydrogen evolution from water/alcohol mixtures [276–278]. The protonation of K4Nb6O17 niobate followed by the intercalation with tetrabutyl ammonium cations results in the exfoliation of the layered material into ultra-thin niobate sheets. The sheet deposition on MgO followed by the calcination and acidic etching results in mesoporous Nb2O5 revealing photocatalytic properties in the water splitting [279]. Mesoporous InVO4 synthesized in the presence of cetyl trimethyl ammonium bromide was applied as a visible-light-sensitive photocatalyst of the H2 evolution from aqueous solutions of oxalic acid [260]. The annealing of Ni(OH)2 NPs deposited from ethanol onto a conducting glass was used to produce mesoporous nickel oxide photoelectrodes [280]. Mesoporous metal chalcogenide nanophotocatalysts. Metal chalcogenide materials are far less stable during the thermal treatment and, therefore, require mild methods to be produced in the mesoporous form, such, for example, as the ultrasound treatment. This method was successfully used for the preparation of mesoporous CdS with a pore size of 5–6 nm formed by loosely aggregated 4–6 nm NPs [281]. The sonochemical treatment of a hot (80 °C) aqueous solution containing cadmium and zinc acetates, thioacetamide, and sodium dodecyl sulfonate, yields mesoporous CdxZn1−xS formed by 3–5 nm particles associated into porous 60– 100 nm spheres (Fig. 5.17c) [282]. The HTT of precursor solutions in methanol is a relatively mild route to mesoporous metal sulﬁde nano-materials, for example, CdIn2S4 nanotubes that exhibit photocatalytic activity in the water reduction [105]. The thermal decomposition of a complex of zinc sulﬁde with ethylenediamine ZnS(en)0.5 results in porous platelet-like ZnS and ZnO particles with a size of 20 Â 30 nm active as photocatalysts of the H2 evolution from aqueous Na2S/Na2SO3 solutions [283]. Porous ZnS microaggregates formed by 4–10 nm particles were synthesized via the HTT of micellar solutions composed of water, cetyl trimethyl ammonium bromide, cyclohexanone and pentanol and containing zinc acetate and urea [284]. The microsphere size increases from 200 nm to 1.5 lm as the ratio of water to surfactant is elevated from 8 to 32. 268 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

5.5 Spatially Organized Nanocrystalline Photo-Active Semiconductors

Mesoporous hollow microspheres. Recently, a considerable interest was observed for the photochemical properties of hollow metal-oxide mesoporous microspheres [243]. Such hollow spheres (HSs) can be produced using the so-called “sacriﬁcial” pore-forming microsphere-templates (SiO2, organic polymeric globules, carbon microspheres, etc.). In a typical synthesis, the TTIP hydrolysis is performed on the surface of polystyrene latex microparticles (Fig. 5.19a–c) [285]. The latex is formed by the styrene polymerization in a PVP solution in the presence of sodium chloride and is composed of highly monodisperse spherical *200-nm particles. The calcination of hydrolytic products yields monodisperse titania HSs with a size of 250– 260 nm [285]. In a similar manner TiO2/SiO2 [286] and ZnO HSs [287] can be produced (Fig. 5.19d–f).

Fig. 5.19 (a, b, c) Polystyrene microbeads (a), microbeads coated with titania (b) and titania HSs (c); (d, e, f) ZnO HSs produced from the microbeads of a different size—450 nm (d), 700 nm (e), and 1100 nm (f). Reprinted with permissions from Refs. [285](a–c) and [287](d–f). Copyright (2008) Elsevier (a–c) and American Chemical Society (d–f) 5.5 Spatially Organized Nanocrystalline Photo-Active Semiconductors 269

The TBT hydrolysis in a sulfonated polystyrene latex with a globule size of 200 nm followed by calcination results in TiO2 HSs with an inner void diameter of 150 nm and the 15-nm-thick mesoporous walls [288]. A prolonged (around 72 h) HTT of spherical TiO2 NP aggregates converts them into HSs [289]. Similar methods were used also to convert dense cubic agglomerates of ZnSn(OH)6 NPs into hollow micro-cubes [282]. TiO2 HSs can be prepared by the HTT of (NH4)2TiF6 [290] or TBT in 2-propanol in the presence of glucose [291, 292]. The role of a pore-forming agent is played by carbon microspheres formed on the first step of HTT. Titania HSs with a size of 70 nm and a wall thickness of 4–7 nm were synthesized by burning Ti(IV) acetylacetonate in a stream of oxygen [293]. TiO2 HSs with a size of up to 550 nm and pore size of 4–5 nm were reported to form in the presence of suspended NaCl crystals in ethanol [294]. The TiOSO4 hydrolysis in alcohols containing glycerol and diethyl ether allows forming HSs without additional pore-forming templates [295]. Depending on the synthesis conditions, this method can yield dense titania microspheres, HSs, as well as more complex geometries, for example, HSs enclosing porous titania microparticles in the inner voids. The iron(III) oxide HSs were produced by the FeCl3 hydrolysis in ethylene glycol in the presence of sodium dodecyl benzyl sulfonate under the microwave heat treatment [296, 297]. The HSs are formed from Fe2O3 nanoplates appearing on the first step of the synthesis. An alternative method for the preparation of hollow metal oxide spheres is based on the local crystallization induced by the presence of fluoride anions. The crystallization results in the Ostwald ripening of amorphous semiconductor particles that propagates from the microsphere center to the outer surface and results in the formation of inner voids. This effect was used for the preparation of Sn-doped titania HSs via the HTT of Ti(SO4)2 solutions with SnCl4 and NH4F[298]. Nanotubes. Titania and titanate NTs have good perspectives for the utilization in various light-harvesting systems [243]. NTs are typically produced via the HTT of titania powders in strongly alkaline solutions or by the electrochemical etching of titania foils in fluoride or phosphate electrolytes [299–301]. The HTT of nanocrystalline TiO2 in alkaline media results in the formation of titanate nanosheets (NSs) with similar characteristics irrespectively of the phase composition or size of the original titania crystals. When such NSs are treated with acids (HCl, HNO3 or H2SO4) the alkali metal ions are exchanged with protons and the protonated form of titanates spontaneously rolls into the NTs (Fig. 5.20). A broadly used method for the preparation of mesoporous TiO2 NTs consists in the electrochemical etching (anodization) of metallic titanium [302–316] in solutions containing H2O2, NaF/HF, alkali phosphates, ethylene glycol and other additives. The etching results in the formation of closely packed arrays of polycrystalline NTs with a length of 1–6 lm, an inner void diameter of 15–120 nm and a wall thickness of 10–20 nm. The anodization of titanium foils with a surface treated preliminarily with ammonium persulfate results in bilayer nanostructures [316]. An upper 100 nm 270 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

Fig. 5.20 a Synthesis of titanate NTs from TiO2 NPs (a). TEM images of titanate NTs (b). SEM images of anodized titania NT arrays (cross-sectional view (c) and upper view (d)). Reprinted with permissions from Ref. [303](c, d). Copyright (2007) Elsevier layer is formed by the anatase nanorods with a diameter of around 20 nm, while a lower 250 nm layer is composed of the mesoporous NTs with the outer and inner diameters of 120 and 80 nm, respectively. Layered nanophotocatalysts. Layered semiconductor materials such as titanates, niobates, and tantalates of alkali metals have the inter-layer sub-nanometer galleries that can accommodate substrates of the photochemical and photocatalytic transformations thus resembling porous nanomaterials. Typically, such layered photocatalysts are prepared by the sintering of corresponding oxides with cesium, potassium or sodium carbonates. The produced metallates are then converted in a protonated form by the ion exchange in acidic solutions (Fig. 5.21). In some cases, the protonated metallates are additionally treated with tetrabuthyl ammonium salts that can intercalate into the interlayer galleries and expand them or, otherwise, disrupt the layer structure completely resulting in the exfoliation to ultra-thin metallate NSs [317–322]. Such materials are broadly studied as photocatalysts of the water splitting [318, 319, 321]. The ultra-thin metallate NSs can be used as building blocks for the preparation of more complex multi-component nanomaterials. For example, HCa2Nb3O10 perovskite NSs with a thickness of around 1 nm were combined with Pt and IrO2 NPs by a bridge molecule—3-aminopropyl trimethoxysilane producing a photocatalyst of the total water splitting [322]. The intercalation of TiO2 NPs into the interlayer galleries of a partially exfoliated H0.67Ti1.83□0.17O4 Â H2O titanate (□ is an anion vacancy) results in a 4– 5-fold increase of photocatalytic activity of the composite layered material in the water splitting [319]. The RhCl3 hydrolysis on the surface of potassium niobate nanoscrolls produced by the exfoliation of layered niobate with tetrabuthyl ammonium hydroxide results in tubular Rh(OH)3/K4Nb6O17 nanostructures with a 5.5 Spatially Organized Nanocrystalline Photo-Active Semiconductors 271

Fig. 5.21 a Preparation and modiﬁcation of layered metallates (on the example of NaNb3O8); b, c TEM images of exfoliated calcium niobate sheets (b) and Ca2Nb3O10 sheets with photodeposited Pt NPs used as a photocatalyst of the water splitting. Reprinted with permissions from Refs. [321] (b) and [322](c). Copyright (2007–2008) American Chemical Society diameter of 30 nm and a wall thickness of 5 nm. The tubular nanostructures are 100 nm long and decorated with sub-nanometer rhodium hydroxide NPs. The alternating interaction of silica NPs with negatively charged potassium niobate sheets and polydiallyl dimethyl ammonium cations resulted in porous “core/shell” structures that were used as active photocatalysts of the hydrogen evolution from water/methanol mixtures [323].

5.6 Nanocrystalline Photo-Active Semiconductors on Carriers

Porous nano-photocatalysts can be prepared by anchoring semiconductor NPs on the surface of zeolites and molecular sieves, in the interlayer galleries of clays, etc. [243, 324]. Alternatively the semiconductor NPs can be formed directly on the surface of a porous carrier that provides abundant sites for the nucleation and simultaneously restricts the NP growth. TiO2 NPs were formed in the pores of natural silicates (mordenite, bentonite, palygorskite, kaolinite, etc. [325–336]) and ceramics [337]. Porous titania-based nanomaterials were synthesized basing on various zeolites, in particular zeolite Y [338–340], b [340], MCM-41 [341–345], SBA-15 [329, 346, 347], ZSM-5 272 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

[340, 348–350], etc. In a typical synthesis, the zeolite is impregnated with a solution of Ti(IV) alkoxide and annealed. The zeolite carriers were successfully applied also for the preparation of metal sulfide-based porous nanophotocatalysts. The synthesis proceeds via an ion exchange step when the alkali cations are exchanged with cadmium, indium, zinc ions. The following sulfidation of such zeolites yields metal-sulfide NPs with a size smaller than the pore size of the zeolite host [351, 352] allowing in future for the diffusion of reactants of the photocatalytic transformations. For example, the deposition of 2.5–2.6 nm CdS [352]andIn2S3 NPs [351] in the pores of a titanosilicate Ti-MCM-41 yields efficient photocatalysts for the water reduction. The water splitting can also be photocatalyzed by CdS NPs incorporated into the pores of zeolite Y [353–355], b [353], L [355], SBA-15 [355, 356], and ZSM-5 [281, 353], aluminium and magnesium oxides [281], as well as into the interlayer galleries of potassium niobate [354]. Semiconductor NPs are often stabilized in the perfluorinated Nafion membranes. The membrane consists of polytetrafluoroethylene fragments alternated with lateral perfluorovinyl ether fragments with the terminal sulfonic groups. It contains regularly separated 4–5 nm voids that have a hydrophilic character due to the presence of sulfonic groups [357]. This feature, coupled with mechanical and chemical stability, and a low light absorption in UV and visible spectral range make the Nafion a perfect host for the formation of semiconductor NPs. The Nafion-stabilized NPs are photostable and accessible for typical substrates of the photocatalytic and photoelectrochemical reactions [357]. For example, the Nafion membrane impregnation with TTIP followed by the treatment with boiling water resulted in 4 nm TiO2 NPs that were used as a photocatalyst of the CO2 reduction [358, 359]. CdS NPs can be formed in the Nafion by reacting CdCl2 and thioacetamide [357]. In contrast to Na2S, thioacetamide can diffuse through the membrane providing a much more uniform distribution of CdS NPs. The H2 evolution photocatalysts were produced by the deposition of 2–5nm CdS NPs onto the surface of polystyrene micro-globules [360]. For the same purposes CdxZn1–xS NPs were attached to the Whatman paper [361]. Porous photocatalysts were produced by impregnating the mesoporous graphite with a TiO2 sol [362]. The carrier can be produced by the styrene polymerization in the presence of mono-disperse (30 nm) colloidal SiO2 NPs followed by the carbonization in an Ar stream. The deposition of TTIP from the gas phase on the surface of activated carbons with the following hydrolytic transformation yields 10– 50-nm TiO2 crystals dispersed in the host material pores. WO3 NPs deposited on the amorphous carbon were used as a visible-light-sensitive harvesting component of the photoelectrochemical systems [363]. By impregnating porous alumina membranes with TiF4 titania NTs were produced with an inner diameter of 5–100 nm, depending on the synthesis duration [364]. Titania dispersions can be also attached to fibers and tissues using silica sols as binding additives [364]. Also, the photo-active semiconductor NPs can adhere to various thermally stable materials, such as metals or ceramics [365, 366]. 5.7 Doped Semiconductor Nano-Photocatalysts 273

5.7 Doped Semiconductor Nano-Photocatalysts

As discussed in Chaps. 2 and 3, doping of the wide-bandgap semiconductor with moieties that can be excited by the visible light is one of the most broadly used ways of spectral sensitization [51, 367–371]. The metal ion doping results in discrete states in the band gap that can be involved in the electron transitions. Additionally, the dopant atoms can serve as traps for the photogenenared charge carriers helping to avoid the electron-hole recombination. Very often the dopant atoms can also inhibit to some extent the growth of metal oxide semiconductor NPs during the sol-gel synthesis as well as to stabilize some low-temperature phases (anatase in the case of titania) during the annealing. Also, the doping typically strengthens the mesoporous frameworks preventing them from collapse at the high T calcination. Doping with non-metals (N, C, S, P, B, etc.) in the metal oxides results in a partial substitution of the lattice oxygen. At that, the p-orbitals of the dopant lay higher (in terms of the energy) than the p-orbitals of oxygen in the valence band, thus narrowing the band gap and making the semiconductor photo-active under the illumination with the visible light. Doping with metals. The most frequent method of doping metal oxides with metal ions consists of the introduction of dopant salts into the reaction mixtures before the hydrolysis and sol-gel transformation of the precursors. Alternatively, the doped metal oxides can be deposited in the pyrolytic processes from dopant-containing solutions. The metal oxides can also be bombarded with ion beams and subjected to the thermal or mechanochemical treatment in the presence of a dopant. The hydrolysis of Ti(IV) precursors in the presence of dopant salts serves as a general method for the synthesis of doped titania powders and films. The metal-doped TiO2 is often produced from the self-igniting mixtures. For example, Pd-doped [372] and Sm-doped [373] nanocrystalline titania was synthesized by burning a mixture of titanyl nitrate, fuel (glycine [372] or ethylene glycol and citric acid [373]) and a dopant nitrate. The collisions of accelerated ionized dopant beam with the surface of titania result in the incorporation of dopant ions into the oxide lattice. This method was applied to modify TiO2 with vanadium [374, 375] and chromium [376], both dopant penetrating 70–80 nm deep from the crystal surface into the titania crystal bulk without the formation of separate phases or surface compounds. Nanodisperse titania is often doped with alkali earth ions (Mg, Ba) to reduce the crystal size and to enhance the thermal stability of the anatase modification during the annealing [377, 378]. This method allows to decrease the titania crystal size from 20 to 5–10 nm and to produce pure anatase nanomaterials. Metal sulfide nano-photocatalysts can be doped with ions of metals forming metal sulfides with a lower solubility. Such metals (for example, Bi, Sb, Ag, Cu) can substitute the main ions (Cd, Zn, In) in the near-surface layer of the sulfide 274 5 Synthesis of Nanocrystalline Photo-Active Semiconductors crystals forming point defects, islands or separate NPs depending on the dopant concentration [379]. Doping with non-metals. Two basic methods are used to synthesize nonmetal-doped oxide semiconductors: (i) by introducing corresponding additives into the reaction mixtures where the formation and aging of oxide NPs take place, and (ii) by a post-synthesis treatment of metal oxide NPs in a solution or gas stream containing dopant compounds. The most popular doped metal oxide in the photocatalysis and related solar light harvesting technologies is nitrogen-doped titania. TiO2:N samples were produced by introducing nitrogen-containing admixtures into the solutions where the hydrolysis of Ti(IV) of the oxidative hydrolysis of Ti(III) takes place [380, 381]. The most frequent sources of nitrogen are NH4OH [204, 275, 382–389] and ammonium salts [388], hydrazine [388–391], urea [380, 381, 392–397], alky- lamines [398–400], ethylenediamine [401], hexamethylenetetramine [402], bipyridyl [403]. The product of sol-gel transformation is typically subjected to the HTT and annealed at 350–550 °C. In some cases, mixtures of various N sources are used to increase the total nitrogen amount in the final product. For example, by using a mixture of NH4OH and N2H4 up to 0.45 wt.% nitrogen can be introduced into the titania lattice while for the separately taken nitrogen sources this level is only around 0.24 wt.% [389]. The nitrogen dopant can be introduced into the nanocrystalline powders and NTs of TiO2 [404] and TiO2/SiO2 composites [405] by the thermal treatment in the ammonia stream [404–412]. The treatment results in the surface nitridation of the metal oxide but does not alter the crystal structure as a whole. The nitrogen dopant typically shows an absorption “shoulder” at 400–550 nm depending on the N content making titania sensitive to the visible light. TiO2:N materials can be conveniently produced by the mechanochemical treatment of commercial nanocrystalline titania (Evonik P25) with hexamethylenetetramine followed by the annealing at 400 °C to decompose the amine [413]. Ammonium salts were also used as a nitrogen source during the mechanochemical synthesis of TiO2:N [388]. The post-synthesis nitridation can be achieved by applying a high-voltage electric discharge to the nanocrystalline titania in the nitrogen atmosphere [414]. Photochemically active TiO2:N samples were reported to form as a result of the partial oxidation of nanocrystalline TiN [388, 415]. The nitrogen doping was observed at the magnetron sputtering of nanocrystalline titania [416] and ZnO [417] films. Similarly to nitrogen, carbon can be introduced during the sol-gel transformations [418–421]. The carbon atoms can either substitute oxygen in the titania lattice [419] or remain in the interstitial positions [421], in both cases making the metal oxide sensitive to the visible light. The ultrasound-assisted electrochemical etching of titanium foils in solutions of ethylene glycol and NH4F followed by the annealing yields TiO2 NTs doped with carbon [422]. Alternatively, TiO2 NTs can be annealed in a mixture of argon and acetylene producing TiO2:C [423]. 5.7 Doped Semiconductor Nano-Photocatalysts 275

Nanopowders of TiO2:C were produced from the self-igniting mixtures such as a composition of TiCl4, citric acid and ethylene glycol [424]. The hydrolytic and pyrolytic methods were applied to introduce sulfur dopants. For example, the sol-gel transformation of TBT in the presence of thiourea yields TiO2:S samples containing 1–3 wt.% sulfur and demonstrating photocatalytic properties in the hydrogen production under the Vis-illumination [425]. The TBT hydrolysis in the presence of NH4F can be used to synthesize TiO2:F samples [426, 427]. Alternatively, TiO2:F was prepared by the HTT of acidic TiF4 solutions [428]. Nanocrystalline TiO2:F ﬁlms are typically formed by the annealing of titanium foils with a surface layer of sodium ﬂuoride solution [429] as well as by the HTT of titanium foils in HF solutions [430, 431]. The hydrolysis of Ti(IV) tetraethoxide in a hot solution of H3PO4 [432] or HPO3 [433] combined with the sol-gel transformation and annealing results in TiO2:P nano-powders sensitive to the visible light. The sol-gel conversion of TTIP in the presence of a mixture of boric acid and B(CH3CH2O)3 leads to the samples of TiO2: B with a grain size of 10–15 nm [431, 434–436]. The anodization of titanium foils in H3BO3 solutions yields boron-doped TiO2 NTs with a spectral response at k < 540 nm [435].

5.8 Bi- (Multi-) Component Photo-Active Semiconductor Nanostructures

Binary semiconductor nanostructures with “concerted” positions of the conduction and valence bands allowing for the efficient interfacial charge transfer between the components is one of the most potent ways of achieving enhanced spatial separation of the photogenerated charge carriers and boosting of the quantum efficiency of the photocatalytic transformation. Metal chalcogenide/metal oxide composites. Very popular are “metal oxide— metal chalcogenide” nanostructures that can be produced by forming chalcogenide NPs in the presence of colloidal or nanopowdered metal oxide component. For example, TiO2/CdS heterostructures were synthesized by impregnating commercial TiO2 (Evonik P25) with a Cd(II)-thiourea complex solution followed by the thermal treatment in the nitrogen stream [437]. Similar nanostructures were obtained by the deposition of MPA-capped CdS NPs on the TiO2 surface [438, 439]. TiO2/CdS and ZrO2/CdS composites can be easily produced by the consecutive impregnation of nanocrystalline TiO2 and ZrO2 films with solutions of cadmium acetate and sodium sulfide [440]. In the same way, PbS, CdS, Ag2S, Sb2S3, and Bi2S3 NPs were anchored to the surface of TiO2, SnO2,Nb2O5,Ta2O5 films [441]. TiO2/PbS nanostructures with 3-nm PbS NPs were formed on the surface of Evonik P25 titania by the CBD from hot alkaline aqueous solutions of lead acetate and thiourea [442]. This method was also applied to produce CdS NPs on the surface of anodized titania NTs [443], mesoporous TiO2 [444], titania NRs 276 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

[127, 445, 446] and NWs [447], as well as on rod-like ZnO nanocrystals [448, 449]. Cadmium sulfide NPs can be formed on the titanate NTs by exchanging alkali metal ions on the NT surface with Cd2+ followed by the treatment in a hot thiourea solution [450, 451]. By purging H2S into a solution of bismuth iodide and MPA in the presence of nanocrystalline TiO2 films TiO2/Bi2S3 nanoheterostructures were produced [452]. Mesoporous TiO2/CdS composites can be synthesized by the SILAR method [214, 453]. This method can be extended to the preparation of TiO2/CdxZn1−xS [454], TiO2/In2S3 [455], TiO2/PbS [456, 457], and TiO2/Cu2S[458]. Tubular TiO2/CdS nanostructures produced by the ultrasound-assisted electrodeposition of cadmium sulfide NPs (Fig. 5.22a, b) revealed a high photoelectrochemical activity in the water splitting [459]. The photoelectrochemically active ternary ZnO/CdSe/CdS heterostructures can be formed by the successive deposition of CdSe and CdS NPs (Fig. 5.22c, d) [460]. In a similar way, ternary ZnO/ZnS/AgInS2 [461] and ZnO/CdS/PbS [462] film heterostructures were produced as well as nanotubular TiO2/CdS/CdSe composites [463] and SnO2/CdS/CdSe nanostructures [464]. The TTIP hydrolysis in the presence of 40-nm CdS crystals followed by the thermal treatment at 450 °C yields CdS/TiO2 heterostructures comprising 9–10-nm anatase NPs (Fig. 5.23)[465]. In a similar manner, TiO2 NPs were anchored to the surface of CdS NWs [466, 467]. The 5-nm CdS and CdSe NPs produced in tetrahydrofuran in the presence of tributhyl phosphine and 2,3-dimercaptosuccinic acid can be attached covalently to

Fig. 5.22 Array of anodized TiO2 NTs (a) and TiO2/CdS heterostructure (b). c, d Ternary ZnO/CdS/CdSe heterostructure. Reprinted with permission from Refs. [459](a, b) and [460](c, d). Copyright (2010, 2011) Elsevier 5.8 Bi- (Multi-) Component Photo-Active Semiconductor Nanostructures 277

Fig. 5.23 Rod-like CdS/TiO2 heterostructures as a photocatalyst of the hydrogen evolution. Reprinted with permissions from Ref. [466]. Copyright (2008) Elsevier

the titania surface via the mercapto-group. Mercaptocarboxylic acids HS–(CH2)n– COOH (n =1–3) [468–471], as well as mercaptopropyl-trimethoxysilane [472] were also used as “bridges” to bind CdS and TiO2 NPs. The photochemically active HNbWO6/(Pt, Cd0.8Zn0.2S) heterostructures were synthesized by the successive intercalation of an ammonia complex of platinum and a mixture of cadmium and zinc acetate followed by the sulfidation [473]. Using a similar approach, layered K4Ce2M10O30 (M = Ta, Nb) materials were decorated with CdS NPs [474]. The chalcogenide/oxide heterostructures can also be produced by the co-precipitation of both components or by mixing separately prepared NPs. For example, the simultaneous precipi–ta–tion of zinc sulfide from ZnSO4 and the TiOSO4 hydrolysis in the presence of thioacetamide was used to prepare porous 2 ZnS–TiO2 heterostructures with a specific surface area of up to 200 m /g comprising 15–20-nm ZnS NPs and 6–7 nm titania crystals [475]. The binary composites can be produced by mixing colloids of CdS and TiO2 [44, 476–479], ZnO and CdS [480], CdS and AgI [479]. An additional treatment of a mixture of cadmium sulfide and titania NPs with titanium(IV) chloride results in a stronger electric contact between CdS and TiO2 NPs favoring to the spatial separation of the photogenerated charge carriers [481]. TiO2/CdS nanostructures can be formed by the electrochemical sulfur reduction in DMSO on the surface of titania NT arrays in the presence of cadmium salts [482]. In a similar tellurium reduction process TiO2/CdTe NTs were obtained [483]. The nanoheterostructures comprising cadmium selenide NPs are typically synthesized in colloidal selenide solutions synthesized in the high-boiling-point organic solvents [484] or in the aqueous solutions in a reaction between Na2SeSO3 and cadmium salts in the presence of nitrilotriacetic acid [472, 485–487]. 278 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

Alternatively, selenide ions can be generated by reducing Se (IV)/Se (VI) with N2H4 [488]. Then, colloidal CdSe NPs can be coupled with colloidal TiO2 [489]or with titania nanopowders [402] and mesoporous materials [485, 486, 490–495], TiO2 NTs and NWs [487, 496]. Similarly, ZnO/CdSe nanowhiskers were synthesized [497]. Such heterostructures are broadly used for the preparation of light harvesting cell electrodes [485–487, 490–497]. Oxide nanocomposites. Binary oxide nanostructures are typically formed by the impregnation of a major component with salt solutions followed by the annealing, or by the co-hydrolysis, thermal treatment of salt mixtures, pyrolysis, electrodeposition, etc. The photoactivity of nanocrystalline titania was increased by coupling it with ZnO [498], Fe(OH)3 [499, 500], Ni(OH)2 [501], Cu(OH)2 [499], and MnOx [502]. Such heterostructures can be produced by a sequence of the impregnation of Evonik P25 with metal salt solutions, salt hydrolysis, and annealing. The calcination of nanopowdered TiO2 soaked with copper(II) salts yields TiO2/CuxO composites with x close to 1 after the treatment at 350 °C and increasing to 2 at 450 °C [503]. The TiO2/CuO heterostructured photocatalysts of the hydrogen evolution were prepared in a similar way from titania NTs [504]. ZnO–TiO2 nanostructures can be produced by the simultaneous hydrolysis of TBT and zinc nitrate in the presence of sodium dodecyl benzyl sulfonate followed by the annealing [505] or, alternatively, by the HTT of solutions of titanium and zinc chlorides, urea [506] or ammonia [507]. By annealing the coprecipitated oxides at 500 °C the SnO2–ZnO and TiO2–WO3 composites were produced [508–510]. Tungsten oxide NPs affect the crystallization of titania rendering the anatase modification of TiO2 stable up to 800 °C and decreasing the average titania NP size from 7 nm (with no WO3 present) to 2.5 nm for the composites containing 4 wt.% WO3. Binary nanocrystalline TiO2/Cu2O films were produced by a two-step synthesis [511]. In the first step, a suspension TiO2 nanocrystals (Evonik P25) ultrasonically dispersed in a Triton X-100 solution was deposited onto the conductive ITO glass plate, then a layer of copper(I) oxide NPs was electrodeposited as a second step. The co-hydrolysis was applied to produce photo-active Fe2O3–TiO2 [512], SnO2–TiO2 [513, 514], and ZnO–TiO2–SiO2 nanostructures [515]. The HTT of alkaline solutions of tin (IV) chloride and zinc acetate yields mesoporous SnO2– ZnO composites [s1124]. The TTIP hydrolysis in the presence of colloidal SnO2 NPs results in “core/shell” SnO2/TiO2 composites [516]. In a similar method, Fe2O3/TiO2 with an iron oxide core [517] and TiO2/Fe2O3 with a titania core were synthesized [518, 519], as well as InVO4/TiO2 [520], TiO2/ZrO2 [521], and Bi2O3/TiO2 heterostructures [522]. The HTT of potassium titanate results in K0.3Ti4O7.3(OH)1.7 NWs with a length of 0.5–1.5 lm and 10–20-nm anatase crystals attached to the NW surface [523]. The anion exchange followed by the hydrolysis of adsorbed Fe (III) was applied to form HTiNb(Ta)O5/Fe2O3 heterostructures [524]. In a similar way, the photoactive TiO2/MnO2 composites were produced from porous MnO2 [525]. TiO2/SnO2 can be formed via the pyrolysis of tetramethyl tin in the presence 5.8 Bi- (Multi-) Component Photo-Active Semiconductor Nanostructures 279 of TiO2 Evonik P25 [526] as well as by the pyrolytic decomposition of mixed SnCl4 and TiO[C5H7O2]2 solutions [527]. The pyrolysis of solutions containing ammonia complexes of zinc and iron or (NH4)10W12O41 was used for the preparation of Fe2O3–ZnO and WO3–ZnO heterostructures [528]. The electron beam sputtering was applied to form ordered WO3 NR arrays coated with a titania nano-layer [529]. The titania NT arrays were decorated with hematite NPs by the electrodeposition [530]. A relatively short contact time (1 h) between the nanotubes and Fe(III) solution results in a selective precipitation of a-Fe2O3 NPs on the NT tips. A prolonged contact (24 h) yields titania NTs densely covered with hematite NPs both on the inner and outer surface (Fig. 5.24)[530]. Semiconductor heterostructures with carbon nanomaterials. The synthesis of nanostructures containing various carbon forms was partly discussed earlier in the section dedicated to the NP-on-a-carrier photocatalysts. However, the role of carbons is not confined only to the inert support of photoactive semiconductor NPs. The carbon nanomaterials are reported to participate actively in the charge separation processes both in primary photoprocesses, by donating/accepting electrons, and in the secondary steps of the photocatalytic reactions as co-catalysts or adsorbents [531, 532]. The active carbons used in the photocatalytic systems are

Fig. 5.24 Titania NT arrays (a, b) and TiO2/Fe2O3 (c, d) nano–composites produced by the electrodeposition during 1 h (c) and 24 h (d). Reprinted with permissions from Ref. [530]. Copyright (2011) American Chemical Society 280 5 Synthesis of Nanocrystalline Photo-Active Semiconductors carbon NTs, fullerenes, graphite, graphite oxide and graphene oxide as well as reduced graphene oxide (RGO). The carbon NTs can be decorated with titania NPs by the Ti(IV) alkoxide hydrolysis [531–541] or by drying the NT suspensions with nanocrystalline TiO2 [542, 543]. In a similar way, the carbon NTs were modiﬁed by ZnO NPs [544]. The nanocomposites of titania with carbon NRs or fullerene were produced by the TBT hydrolysis in the presence of NTs or C60 that were partially pre-oxidized with m- chloroperbenzoic acid [545]. The microwave treatment of carbon NTs in solutions of zinc acetate and thioacetamide results in NT/ZnS heterostructures (Fig. 5.25a) [546]. Similar methods were applied to synthesize NT/ZnSe [547] and NT/CdSe composites [548]. The methane decomposition at 600 °C on the surface of Fe(III)-doped TiO2 nanocrystals starts with the formation of amorphous carbon islands which then transform into the carbon NTs with the edges attached to the titania crystals (Fig. 5.25b) [549]. Recently the composite semiconductor nano-photocatalysts covered with a layer of amorphous or graphitic carbon were intensely studied. The titania nanocomposites with graphitic carbon can be produced by the thermal annealing of a mixture of Evonik P25 with polyvinyl alcohol [550]. The carbon covering TiO2 nanocrystals prohibits the transformation of anatase into rutile at the temperature of up to 800 °C. The thermal decomposition of glucose adsorbed on the ZnO NR surface under the microwave illumination in an inert atmosphere results in ZnO/C nanostructures with the carbon forming a uniform and dense 3–4-nm shell over the core semiconductor [551]. The HTT of TiO2/C nanocomposite produced by the glucose decomposition allows enhancing the graphitization of the carbon layer [552]. Similar structures based on 10–20-nm ZnO nanocrystals were also produced by the glucose carbonization and graphitization at 800 °C in the N2 stream [553]. The ultrasonic treatment of graphite suspensions with titania yields intercalated

Fig. 5.25 Composites of carbon NTs with ZnS (a) and TiO2 NPs (b). Reprinted with permissions from Refs. [546](a) and [549]. Copyright (2008,2009) American Chemical Society (a) and Elsevier (b) 5.8 Bi- (Multi-) Component Photo-Active Semiconductor Nanostructures 281 graphite/TiO2 composite that can be used as a light-harvesting component of the solar cells [554]. In the case of composite TiO2/graphite ﬁlms deposited from the gas phase, the introduction of graphite was found to decrease the size of TiO2 NPs and to enhance their photoactivity [555]. Carbon/WO3 nanostructures produced by the gas-phase deposition onto the carbon paper exhibited a photoelectrochemical activity under the Vis-illumination [363]. An explosive growth of interest in the chemistry of graphene and its derivatives was stimulated by the mechanical exfoliation of graphite reported in 2004 by Geim and Novoselov [556, 557]. The most popular methods of the production of graphene or, more exactly, partially reduced graphene oxide are based on the reduction of grapehene oxide (GO), that can be obtained by the exfoliation of graphite oxide, while the latter is synthesized by the oxidation of graphite [558, 559]. One of the main challenges is to prevent the RGO aggregation into graphitic structures during the GO reduction. This aim can be achieved by reducing GO in the presence of semiconductor NPs that can arrange along the basal plane of GO and RGO particles interacting with the functional groups on their surface and making impossible the restacking and restoration of the original graphitic layered structure. In the case of photo-active semiconductor NPs, such as CdS, CdSe, TiO2, ZnO, etc. the photochemical and catalytic properties of the corresponding composites with RGO are typically expressed much strongly that those of the individual semiconductor NPs. The fact opens broad possibilities of the synthesis of new nano-catalysts and nano-photocatalysts [560, 561]. The RGO/semiconductor nanocomposites are typically produced by the HTT of a nano-dispered semiconductor (or its precursors) with a suspension (colloidal solution) of GO. Alternatively, the RGO-based composites can be synthesized by the in situ photocatalytic reduction of GO over semiconductor NPs [562]. The nanostructures of C60 fullerene with semiconductor NPs are formed by the hydrolysis of metal alkoxides in the presence of suspended fullerene (TiO2/C60 [563]) or by mixing colloidal semiconductors and fullerenes (CdSe/C60 [564]).

5.9 Photo-Active Semiconductor/Metal Nanostructures

Photocatalytic transformations of many substrates, for example, the reduction of water and CO2, CO decomposition, alcohols dehydrogenation, etc., require additional co-catalysts, typically metal NPs that accumulate electrons photogenerated in the nanocrystalline semiconductor photocatalyst and accelerate secondary “dark” redox processes. One of the most popular methods of the synthesis of semiconductor/metal nanostructures is based on the semiconductor impregnation with a metal salt followed by the chemical reduction with a suitable agent (NaBH4, hydrogen, hydrazine, etc.). This method was applied to produce mesoporous TiO2/Ag [261, 565], TiO2/Au [565– 567], TiO2/Cu [568], TiO2/Ni [569] composites as well as plate-like KTiNbO5/Au 282 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

Fig. 5.26 TEM images of selected semiconductor/metal nanocomposites: the sheet-like KtiNbO5/ Au composite (a), core/shell Ag/TiO2 (b) and rod-like ZnO/Ag heterostructure (c). Reprinted with permissions from Refs. [570](a), [571](b), and [580](c). Copyright (2007, 2011) Elsevier (a) and American Chemical Society (b, c) nanostructures (Fig. 5.26a) [570]. The photo-active composites with an “inverted” structure formed by a 10-nm titania layer on 15-nm silver nanocrystals were prepared by the TTIP hydrolysis on the surface of pre-synthesized Ag NPs (Fig. 5.26b) [571]. Nanostructured ITO/Au/TiO2 electrodes produced by the electrodeposition of gold NPs with the following formation of a titania layer on their surface can generate photocurrent when illuminated both by UV and the visible light [572]. TiO2/Pt nanocomposites revealing photoactivity in many reactions can be synthesized by the impregnation of commercial TiO2 Evonik P25 with platinum salts and the reduction with formaldehyde [573] or hydrogen [573–576]. Alternatively, metal NPs can be anchored to the semiconductor surface by the electrophoretic deposition [577–579]. The reduction of silver ions on the surface of ZnO NRs yields photoactive ZnO/Ag nanostructures (Fig. 5.26c) [580]. The NR-shaped ZnO/Ag composites were produced by a two-step synthesis, when the silver NWs are ﬁrst formed by the solvothermal synthesis in PVP, and then the ZnO NRs were deposited on the metal NW surface by the Zn(NO3)2 hydrolysis in the presence of hexamethylenetetramine [581]. The visible-light-sensitive Si/Pt electrodes were produced by reducing H2PtCl6 with silicon in HF solutions [582]. Nanostructured CdS/Au [583] and TiO2/Au ﬁlms [584] can be formed by the deposition of positively charged metal NPs onto the nanocrystalline semiconductors charged negatively by adsorbed mercaptoacetate or citrate anions. The ultrasound treatment of semiconductor suspensions in the presence of metal salts is a relatively mild and convenient method for the preparation of various semiconductor/metal composites. This method was used to form TiO2/M (M = Pt, Au, Pd) [585] and TiO2/Au–Pd heterostructures [586]. The ultrasonic treatment of 3− Evonik P25 suspension in a solution of Au(S2O3)2 also yields TiO2/Au nanostructures with the metal NP size increasing from 1.8 to 3.0 nm as the precursor concentration is elevated from 1 to 8 wt.% [587]. 5.9 Photo-Active Semiconductor/Metal Nanostructures 283

The incubation of a solution containing 4 nm Au NPs and Na2ZnO4 at 90–95 °C results in the formation of photoactive Au/ZnO nanostructures with a 2.0–2.5 nm zinc oxide shell [588]. TiO2/Pt heterostructures were prepared by the microwave treatment of titania NWs in ethylene glycol in the presence of H2PtCl6 [589], as well as by the pyrolysis of a mixture of TiO2 Evonik P25 with hexachloroplatinic acid [590]. The solvothermal treatment of methanol solutions of a H2−2x[Pt(NH3)4]xTi4O9 Â 0.25H2O complex produced from Pt(NH3)4Cl2 and H2Ti4O9 Â 0.25H2O yields fibrous TiO2/Pt nanocomposites [591]. When titania NTs are used as a carrier for Pt NPs the charge of metal precursor affects strongly the size of Pt NPs. For example, the surface protons of titania NTs can be exchanged with the positively charged platinum-ammonia complexes. The following annealing yields highly dispersed Pt NPs with a size of around 2 nm [592], active as a co-catalyst of the photocatalytic water splitting. At the same time, the thermal treatment of titania NTs impregnated with H2PtCl6 results in much larger Pt crystals (20–50 nm) showing a low activity in this photoprocess [592]. The thermal treatment of silver compounds introduced during the synthesis of titanium dioxide or into the porous TiO2 materials results in TiO2/Ag heterostructures [207, 266, 593, 594]. The photoactive ZnO/Ag composites form after the calcination of a product of the co-precipitation of Ag(I) and Zn(II) hydroxides [595], as well as by the HTT of solutions containing zinc acetate, AgNO3 and hexamethylenetetramine [596]. The thermal treatment was also applied to synthesize photoactive TiO2/Cu nanostructures [597, 598]. The flame pyrolysis of precursors is a universal method for the synthesis of semiconductor/metal nanocomposites. It was used, in particular, to synthesize TiO2/M nanostructures (M = Ag, Au, Au–Ag, Pt) [599]. The flame pyrolysis of a mixture of TTIP and gold(III) dimethylacetylacetonate was used to prepare a TiO2/Au photocatalyst of the water splitting [600]. The pulsed electrodeposition of metal NPs is also a convenient method for the decoration of titania NT arrays [601] and mesoporous TiO2 films [602] with gold NPs. Along with the chemical methods for the preparation of semiconductor/metal nanostructures the practice of photocatalysis broadly uses the in situ photochemical/photocatalytic reduction of metal compounds on the surface of micro- and nanocrystalline semiconductors. All the above discussed heterostructures can be produced by the photocatalytic reduction and the assortment of reported composites includes TiO2/Hg [58], TiO2/Cd [603, 604], TiO2/Zn, TiO2/ Mn, TiO2/Tl [604], ZnO/Pt [605–607], ZnO/Au [608], ZnO/Ag [158, 605], ZnO/Cu [605, 609, 610], ZnO/Pd [611], etc. The addition of dye sensitizers allows accomplishing the photocatalytic deposition of metals under the Vis-illumination. In this way, TiO2/Pt, TiO2/Pd, TiO2/Au [s1375], TiO2/Ag [612–614], and ZnO/Ag [96] heterostructures were synthesized. The semiconductor/metal heterostructures produced by the photodeposition are broadly studied in the water splitting processes and other endothermic reactions. For example, the mesoporous TiO2/Ni and TiO2/Cu formed via the photocatalytic 284 5 Synthesis of Nanocrystalline Photo-Active Semiconductors reduction of metal salts were then applied as photocatalysts of the hydrogen evolution from water/ethanol mixtures [615, 616]. The applicability area of the semiconductor/metal composites synthesized by the photodeposition is not confined to the photocatalysis. For example, spectral parameters of the surface plasmon resonance bands of silver NPs photodeposited onto the surface of nanocrystalline titania films depend strongly on the dielectric properties of the solvent. The fact allows applying the TiO2/Ag heterostructures as solvatochromic sensors for the detection of alcohols, urea, glucose and other substrates [617–620]. The nanostructures formed by the photodeposition of Ag NPs on titania nanotubes [621] and the nanocrystalline TiO2 films [617, 618] revealed photochromic properties—the films become colorless under the Vis-illumination and restore the color under the UV illumination. Local fluctuations of the magnetic field created by the Ag NPs deposited photocatalytically on the surface of nanocrystalline titania result in a giant enhancement of the photoluminescence of TiO2/Ag heterostructures [622].

References

1. Henglein A (1989) Small-particle research: Physicochemical properties of extremely small colloidal metal and semiconductor particles. Chem Rev 89:1861–1873. doi:10.1021/ cr00098a010 2. Eychmüller A (2000) Structure and photophysics of semiconductor nanocrystals. J Phys Chem B 104:6514–6528. doi:10.1021/jp9943676 3. Burda C, Chen X, Narayanan R, El-Sayed MA (2005) Chemistry and properties of nanocrystals of different shapes. Chem Rev 105:1025–1102. doi:10.1021/cr030063a 4. Fujiwara H, Hosokawa H, Murakoshi K et al (1997) Effect of surface structures on photocatalytic CO2 reduction using quantized CdS nanocrystallites. J Phys Chem B 101:8270–8278. doi:10.1021/jp971621q 5. Stroyuk OL, Granchak VM, Korzhak AV, Sya Kuchmii (2004) Photoinitiation of buthylmethacrylate polymerization by colloidal semiconductor nanoparticles. J Photochem Photobiol, A 162:339–351. doi:10.1016/S1010-6030(03)00371-X 6. Shiragami T, Ankyu H, Fukami S et al (1992) Semiconductor photocatalysis: visible light induced photoreduction of aromatic ketones and electro-deficient alkenes catalyzed by quantised cadmium sulfide. J Chem Soc, Faraday Trans 88:1055–1061. doi:10.1039/ FT9928801055 7. Stroyuk OL, Shvalagin VV, Raevskaya AE et al (2003) Photocatalysis of Cd2+ ions reduction in 2-propanol by CdS nanoparticles. Theoret Exp Chem 39:331–336 8. Hoffman AJ, Mills G, Yee H, Hoffmann MR (1992) Q-sized CdS: synthesis, characterization, and efficiency of photoinitiation of polymerization of several vinylic monomers. J Phys Chem 96:5546–5552. doi:10.1021/j100192a067 9. Shiragami T, Pac C, Yanagida S (1990) Visible-light-induced two-electron-transfer photoreductions on CdS: Effect of morphology. J Phys Chem 94:504–506. doi:10.1021/ j100365a002 10. Kanemoto M, Hosokawa H, Wada Y et al (1996) Semiconductor photocatalysis. Part 20. Role of surface in the photoreduction of carbon dioxide catalyzed by colloidal ZnS nanocrystallites in organic solvent. J Chem Soc, Faraday Trans 92:2401–2411. doi:10.1039/ FT9969202401 References 285

11. Kanemoto M, Shiragami T, Pac C, Yanagida S (1992) Semiconductor photocatalysis. Effective photoreduction of carbon dioxide catalyzed by ZnS quantum crystallites with low density of surface defects. J Phys Chem 96:3521–3526. doi:10.1021/j100187a062 12. Korgel BA, Monbouquette HG (1997) Quantum confinement effects enable photocatalyzed nitrate reduction at neutral pH using CdS nanocrystals. J Phys Chem B 101:5010–5017. doi:10.1021/jp971016n 13. Zhang JZ, O’Neil RH, Roberti TW (1994) Femtosecond studies of photoinduced electron dynamics at the liquid-solid interface of aqueous CdS colloids. J Phys Chem 98:3859–3864. doi:10.1021/j100065a049 14. Datta A, Priyam A, Bhattacharyya SN et al (2008) Temperature tunability of size in CdS nanoparticles and size dependent photocatalytic degradation of nitroaromatics. J Colloid Interface Sci 322:128–135. doi:10.1016/j.jcis.2008.02.052 15. Fisher CH, Henglein A (1989) Photochemistry of colloidal semiconductors. 31. Preparation and photolysis of CdS sols in organic solvents. J Phys Chem 93:5578–5581. doi:10.1021/ j100351a049 16. Weib H, Fernandez A, Kisch H (2001) Elektronische Halbleiter-träger-wechselwirkung – ein neuer Effekt in der Halbleiterphotokatalyse. Angew Chem 113:3942–3945 17. Nenadović MT, Čomor MI, Vasić V, Mićić OI (1990) Transient bleaching of small PbS colloids. Influence of surface properties. J Phys Chem 94:6390–6396. doi:10.1021/ j100379a044 18. Nanda J, Sapra S, Sarma DD et al (2000) Size-selected zinc sulfide nanocrystallites: synthesis, structure, and optical studies. Chem Mater 12:1018–1024. doi:10.1021/ cm990583f 19. Riley DJ, Waggett JP, Wijayantha KGU (2004) Colloidal bismuth sulfide nanoparticles: a photoelectrochemical study of the relationship between bandgap and particle size. J Mater Chem 14:704–708. doi:10.1039/B311517H 20. Haram SK, Kshirsagar A, Gujarathi YD et al (2011) Quantum confinement in CdTe quantum dots: investigation through cyclic voltammetry supported by density functional theory (DFT). J Phys Chem C 115:6243–6249. doi:10.1021/jp111463f 21. Warrier M, Lo MKF, Monbouquette H, Garcia-Garibay MA (2004) Photocatalytic reduction of aromatic azides to amines using CdS and CdSe nanoparticles. Photochem Photobiol Sci 3:859–863. doi:10.1039/B404268A 22. Kumar A, Mital S (2004) Synthesis and photophysics of 6-dimethylaminopurine-capped Q-CdS nanoparticles—a study of its photocatalytic behavior. Inter J Photoenergy 6:61–68. doi:10.1155/S1110662X04000091 23. Torimoto T, Sakata T, Mori H, Yoneyama H (1994) Effect of surface charge of 4-amino-thiophenol-modified PbS microcrystal photocatalysts on photoinduced charge transfer. J Phys Chem 98:3036–3043. doi:10.1021/j100062a048 24. Nakashima T, Sakashita M, Nonoguchi Y, Kawai T (2007) Sensitized photopolymerization of an ionic liquid-based monomer by using CdTe nanocrystals. Macromol 40:6540–6544. doi:10.1021/ma0707988 25. Raevskaya AE, Stroyuk OL, Sya Kuchmii (2004) Photocatalytic oxidation of hydrosulfide ions by molecular oxygen over cadmium sulfide nanoparticles. J Nanopart Res 6:149–158. doi:10.1023/B:NANO.0000034719.30620.d3 26. Stroyuk OL, Sobran IV, Korzhak AV et al (2008) Photopolymerization of water-soluble acrylic monomers induced by colloidal CdS and CdxZn1−xS nanoparticles. Colloid Polym Sci 286:489–498. doi:10.1007/s00396-007-1824-4 27. Matsumoto H, Uchida H, Matsunaga T et al (1994) Photoinduced reduction of viologens on size-separated CdS nanocrystals. J Phys Chem 98:11549–11556. doi:10.1021/j100095a041 28. Nosaka Y, Miyama H, Terauchi M, Kobayashi T (1988) Photoinduced electron transfer from colloidal cadmium sulfide to methylviologen: a picosecond transient absorption study. J Phys Chem 92:255–256. doi:10.1021/j100313a003 286 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

29. Nosaka Y, Fox MA (1986) Effect of light intensity on the quantum yield of photoinduced electron transfer from colloidal cadmium sulfide to methylviologen. J Phys Chem 90:6521– 6522. doi:10.1021/j100282a021 30. Raevskaya AE, Korzhak AV, Stroyuk OL, Kuchmiy SY (2005) Spectral, optical and photochemical properties of ZnS nanoparticles. Theoret Exp Chem 41:105–109 31. Raevskaya AE, Korzhak AV, Stroyuk OL, Kuchmiy SY (2005) Photocatalytic reduction of (II) with the participation of ZnS nanoparticles. Theoret Exp Chem 41:231–235 32. Raevskaya AE, Korzhak AV, Stroyuk OL, Kuchmiy SY (2005) Photocatalysis of the formation of ZnS/Au heterostructure via reduction of complex gold ions by ZnS nanoparticles. Theoret Exp Chem 41:359–364 33. Stroyuk OL, Raevskaya AE, Korzhak AV, Kuchmiy SY (2007) Zink sulfide nanoparticles: spectral properties and photocatalytic activity in metals reduction reactions. J Nanopart Res 9:1027–1039. doi:10.1007/s11051-006-9183-5 34. Kryukov AI, Zinchuk NN, Korzhak AV et al (2004) Optical and catalytic properties of Ag2S nanoparticles. J Mol Catal A 221:209–221. doi:10.1016/j.molcata.2004.07.009 35. Meahcov L, Sandu I (2004) Colloidal CdS fluorescence quenching by MV2+ under continuous irradiation. J Fluorescence 14:181–185. doi:10.1023/B:JOFL.0000016289. 30498.e9 36. Rossetti R, Brus LE (1986) Picosecond resonance Raman scattering study of methylviologen reduction on the surface of photoexcited colloidal CdS crystallites. J Phys Chem 90:558– 560. doi:10.1021/j100276a014 37. Kanade KG, Baeg JO, Mulik UP et al (2006) Nano-CdS by polymer-inorganic solid-state reaction: visible light pristine photocatalyst for hydrogen generation. Mater Res Bull 41:2219–2225. doi:10.1016/j.materresbull.2006.04.031 38. Kamat PV, Dimitrijević NM, Fessenden RW (1988) Photoelectrochemistry in particulate systems. 7. Electron-transfer reactions of indium sulfide semiconductor colloids. J Phys Chem 92:2324–2329. doi:10.1021/j100319a044 39. Henglein A (1982) Photochemistry of colloidal cadmium sulfide. 2. Effects of adsorbed methyl viologen and of colloidal platinum. J Phys Chem 86:2291–2293. doi:10.1021/ j100210a010 40. Igumenova TI, Vasil’tsova OV, Parmon VN (1996) Formation and photocatalytic properties of Q-sized nanoparticles of various transition metal sulphides in the inner cavities of lecithin vesicles modified with sodium dodecylsulphate. J Photochem Photobiol, A 94:205–213. doi:10.1016/1010-6030(95)04175-3 41. Raevskaya AE, Stroyuk OL, Kuchmiy SY (2006) Preparation of colloidal CdSe and CdS/CdSe nanoparticles from sodium selenosulfate in aqueous polymers solutions. J Colloid Interface Sci 302:133–141. doi:10.1016/j.jcis.2006.06.018 42. Bavykin DV, Savinov EN, Parmon VN (2000) Studies on the kinetics of interfacial electron transfer sensitized by colloidal CdS. J Photochem Photobiol, A 130:57–61. doi:10.1016/ S1010-6030(99)00196-3 43. Variano BF, Hwang DM, Sandroff CJ et al (1987) Quantum effects in anisotropic semiconductor clusters: colloidal suspensions of Bi2S3 and Sb2S3. J Phys Chem 91:6455– 6458. doi:10.1021/j100310a006 44. Sant PA, Kamat PV (2002) Interparticle electron transfer between size-quantized CdS and TiO2 semiconductor nanoclusters. Phys Chem Chem Phys 4:198–203. doi:10.1039/ B107544F 45. Yagi I, Mikami K, Ebina K et al (2006) Size-dependent carrier dynamics in CdS nanoparticles by femtosecond visible-pump/IR-probe measurements. J Phys Chem B 110:14192–14197. doi:10.1021/jp061073t 46. Hirai T, Nanba M, Komasawa I (2002) Dithiol-mediated immobilization of CdS nanoparticles from reverse micellar system onto Zn-doped silica particles and their high photocatalytic activity. J Colloid Interface Sci 252:89–92. doi:10.1006/jcis.2002.8430 47. Thurston TR, Wilcoxon JP (1999) Photooxidation of organic chemicals catalyzed by nanoscale MoS. J Phys Chem B 103:11–17. doi:10.1021/jp982337h References 287

48. Wilcoxon JP, Samara GA (1995) Strong quantum-size effects in a layered semiconductor: MoS2 nanoclusters. Phys Rev B 51:7299–7302. doi:10.1103/PhysRevB.51.7299 49. Matsumoto H, Matsunaga T, Sakata T et al (1995) Size dependent fluorescence quenching of CdS nanocrystals caused by TiO2 colloids as a potential-variable quencher. Langmuir 11:4283–4287. doi:10.1021/la00011a019 50. Raevskaya AE, Stroyuk OL, Kuchmiy SY et al (2006) Growth and spectroscopic characterization of CdSe nanoparticles synthesized from CdCl2 and Na2SeSO3 in aqueous gelatine solutions. Colloids Surfaces A 290:304–309. doi:10.1016/j.colsurfa.2006.05.038 51. Chen X, Mao SS (2007) Titanium dioxide nanomaterials: synthesis, properties, modifications, and applications. Chem Rev 107:2891–2959. doi:10.1021/cr0500535 52. Serpone N, Lawless D, Khairutdinov R (1995) Size effects on the photophysical properties of colloidal anatase TiO2 particles: size quantization or direct transitions in this indirect semiconductor? J Phys Chem 99:16646–16654. doi:10.1021/j100045a026 53. Hykrdova L, Jirkovsky J, Mailhot G, Bolte M (2002) Fe (III) photoinduced and Q-TiO2 photocatalysed degradation of naphthalene: comparison of kinetics and proposal of mechanism. J Photochem Photobiol, A 151:181–193. doi:10.1016/S1010-6030(02)00014-X 54. Ahn WY, Sheeley SA, Rajh T, Cropek DM (2007) Photocatalytic reduction of 4-nitrophenol with arginine-modified titanium dioxide nanoparticles. Appl Catal B 74:103–110. doi:10. 1016/j.apcatb.2007.01.016 55. Pelet S, Grätzel M, Moser JE (2003) Femtosecond dynamics of interfacial and intermolecular electron transfer at eosin-sensitized metal oxide nanoparticles. J Phys Chem B 107:3215–3224. doi:10.1021/jp027358x 56. Huber R, Moser JE, Grätzel M, Wachtveitl J (2002) Real-time observation of photoinduced adiabatic electron transfer in strongly coupled dye/semiconductor colloidal systems with a 6 fs time constant. J Phys Chem B 106:6494–6499. doi:10.1021/jp0155819 57. Huber R, Spörlein S, Moser JE et al (2000) The role of surface states in the ultrafast photoinduced electron transfer from sensitizing dye molecules to semiconductor colloids. J Phys Chem B 104:8995–9003. doi:10.1021/jp9944381 58. Chen LX, Rajh T, Wang Z, Thurnauer MC (1997) XAFS studies of surface strutures of TiO2 nanoparticles and photocatalytic reduction of metal ions. J Phys Chem B 101:10688–10697. doi:10.1021/jp971930g 59. Dimitrijevic NM, Rajh T, Saponjic ZV et al (2004) Light-induced charge separation and redox chemistry at the surface of TiO2/host-guest hybrid nanoparticles. J Phys Chem B 108:9105–9110. doi:10.1021/jp049028d 60. Dimitrijevic NM, Saponjic ZV, Bartels DM et al (2003) Revealing the nature of trapping sites in nanocrystalline titanium dioxide by selective surface modification. J Phys Chem B 107:7368–7375. doi:10.1021/jp034064i 61. Bahnemann D, Henglein A, Lilie J, Spanhel L (1984) Flash photolysis observation of the absorption spectra of trapped positive holes and electrons in colloidal TiO2. J Phys Chem 88:709–711. doi:10.1021/j150648a018 62. Ban T, Kondoh S, Ohya T et al (2003) Photocatalysis of a transparent titanate aqueous sol prepared from titanium tetraisopropoxide and tetramethylammonium hydroxide. J Photochem Photobiol, A 156:219–225. doi:10.1016/S1010-6030(02)00404-5 63. Serpone N, Sharma DK, Moser J, Grätzel M (1987) Reduction of acceptor relay species by conduction band electrons of colloidal titanium dioxide: light-induced charge separation in picosecond time domain. Chem Phys Lett 136:47–51. doi:10.1016/0009-2614(87)87296-2 64. Moser J, Grätzel M (1983) Light-induced electron transfer in colloidal semiconductor dispersions: single vs. dielectronic reduction of acceptors by conduction-band electrons. J Am Chem Soc 105:6547–6555. doi:10.1021/ja00360a003 65. Matylitsky VV, Lenz MO, Wachtveitl J (2006) Observation of pH-dependent back-electron-transfer dynamics in alizarin/TiO2 adsorbates: importance of trap states. J Phys Chem B 110:8372–8379. doi:10.1021/jp060588h 288 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

66. Walters KA, Gaal DA, Hupp JT (2002) Interfacial charge transfer and colloidal semiconductor dye-sensitization: mechanism assessment via stark emission spectroscopy. J Phys Chem B 106:5139–5142. doi:10.1021/jp015540c 67. Ramakrishna G, Ghosh HN (2001) Emission from the charge transfer state of xanthene dye-sensitized TiO2 nanoparticles: a new approach to determining back electron transfer rate and verifying the Marcus inverted regime. J Phys Chem B 105:7000–7008. doi:10.1021/ jp011291g 68. Kölle U, Moser J, Grätzel M (1985) Dynamics of interfacial charge-transfer reactions in semiconductor dispersions. Reduction of cobaltoceniumdicarboxylate in colloidal TiO2. Inorg Chem 24:2253–2258. doi:10.1021/ic00208a026 69. Müller BR, Majoni S, Meissner D, Memming R (2002) Photocatalytic oxidation of ethanol on micrometer and nanometer-sized semiconductor particles. J Photochem Photobiol, A 151:253–265. doi:10.1016/S1010-6030(02)00010-2 70. Shchukin DG, Möhwald H (2005) Photosynthesis inside polyelectrolyte capsules: effect of confined media. Langmuir 21:5582–5587. doi:10.1021/la050429+ 2+ 71. Zang L, Rodgers MAJ (2000) Diffusion-controlled charge transfer from excited Ru(bpy)3 into nanosized TiO2 colloids with EDTA. J Phys Chem B 104:468–474. doi:10.1021/ jp992636m 72. Fessenden RW, Kamat PV (1995) Rate constants for charge injection from excited sensitizer into SnO2, ZnO, and TiO2 semiconductor nanocrystallites. J Phys Chem 99:12902–12906. doi:10.1021/j100034a032 73. Weng YX, Wang YQ, Asbury JB et al (2000) Back electron transfer from TiO2 III 3− nanoparticles to Fe (CN)6 : origin of non-single-exponential and particle size independent dynamics. J Phys Chem B 104:93–104. doi:10.1021/jp992522a 74. Yang X, Tamai N (2001) How fast is interfacial hole transfer? In situ monitoring of carrier dynamics in anatase TiO2 nanoparticles by femtosecond laser spectroscopy. Phys Chem Chem Phys 3:3393–3398. doi:10.1039/B101721G 75. Fernandez A, Caballero A, Gonzalez-Elipe AR et al (1995) In situ EXAFS study of the photocatalytic reduction and deposition of gold on colloidal titania. J Phys Chem 99:3303– 3309. doi:10.1021/j100010a047 76. Rabani J, Yamashita K, Ushida K et al (1998) Fundamental reactions in illuminated titanium dioxide nanocrystallite levels studied by pulsed laser. J Phys Chem B 102:1689–1695. doi:10.1021/jp973411j 77. Yoshihara T, Katoh R, Furube A et al (2004) Identification of reactive species in photoexcited nanocrystalline TiO2 films by wide-wavelength-range (400–2500 nm) transient absorption spectroscopy. J Phys Chem B 108:3817–3823. doi:10.1021/jp031305d 78. Hao YQ, Wang YF, Weng YX (2008) Particle-size-dependent hydrophilicity of TiO2 nanoparticles characterized by Marcus reorganization energy of interfacial charge recombination. J Phys Chem B 112:8995–9000. doi:10.1021/jp802532w 79. Martini I, Hodak JH, Hartland GV (1999) Dynamics of semiconductor-to-dye electron transfer from anthracene dyes bound to different sized TiO2 particles. J Phys Chem B 103:9104–9111. doi:10.1021/jp990302r 80. Ghosh HN (1999) Charge transfer emission in coumarin 343 sensitized TiO2 nanoparticle: a direct measurement of back electron transfer. J Phys Chem B 103:10382–10387. doi:10. 1021/jp9918611 81. Kim J, Lee D (2007) Size-controlled interparticle charge transfer between TiO2 and quantized capacitors. J Am Chem Soc 129:7706–7707. doi:10.1021/ja071329o 82. Behar D, Rabani J (2001) Laser photolysis of TiO2 layers in the presence of aqueous iodide. J Phys Chem B 105:6324–6329. doi:10.1021/jp003176w 83. Skinner DE, Colombo DP, Cavaleri JJ, Bowman RM (1995) Femtosecond investigation of electron trapping in semiconductor nanoclusters. J Phys Chem 99:7853–7856. doi:10.1021/ j100020a003 References 289

84. Grabner G, Quint RM (1991) Pulsed-laser-induced charge-transfer reactions in aqueous TiO2 colloids. A study of the dependence of transient formation on photon ﬂuence. Langmuir 7:1091–1096. doi:10.1021/la00054a012 85. Yonezawa Y, Kometani N, Sakaue T, Yano A (2005) Photoreduction of silver ions in a colloidal titanium dioxide suspension. J Photochem Photobiol, A 171:1–8. doi:10.1016/j. jphotochem.2004.08.020 86. Zhang R, Gao L, Zhang Q (2004) Photodegradation of surfactants on the nanosized TiO2 prepared by hydrolysis of the alkoxide titanium. Chemosphere 54:405–411. doi:10.1016/ S0045-6535(03)00588-5 87. Chae SY, Park MK, Le SK et al (2003) Preparation of size-controlled TiO2 nanoparticles and derivation of optically transparent photocatalytic ﬁlms. Chem Mater 15:3326–3331. doi:10.1021/cm030171d 88. Bahnemann DW, Kormann C, Hoffmann MR (1987) Preparation and characterization of quantum size zinc oxide. A detailed spectroscopic study. J Phys Chem 91:3789–3798. doi:10.1021/j100298a015 89. Bahnemann DW (1993) Ultrasmall metal oxide particles: preparation, photophysical characterization, and photocatalytic properties. Isr J Chem 33:115–136. doi:10.1002/ijch. 199300017 90. Wang CY, Böttcher C, Bahnemann DW, Dohrmann JK (2003) A comparative study of nanometer sized Fe (III)-doped TiO2 photocatalysts: synthesis, characterization and activity. J Mater Chem 13:2322–2329. doi:10.1039/B303716A 91. Kamat PV, Patrick B (1992) Photophysics and photochemistry of quantized ZnO colloids. J Phys Chem 96:6829–6834. doi:10.1021/j100195a055 92. Hoffman AJ, Yee H, Mills G, Hofmann MR (1992) Photoinitiated polymerization of methyl methacrylate using Q-sized ZnO colloids. J Phys Chem 96:5540–5546. doi:10.1021/ j100192a066 93. Henglein A (1993) Physicochemical properties of small metal particles in solution “microelectrode” reactions, chemisorption, composite metal particles, and the atom-to-metal transition. J Phys Chem 97:5457–5471. doi:10.1021/j100123a004 94. Su S, Lu SX, Xu WG (2008) Photocatalytic degradation of reactive brilliant blue X-BR in aqueous solution using quantum-sized ZnO. Mater Res Bull 43:2172–2178. doi:10.1016/j. materresbull.2007.08.029 95. van Dijken A, Meulenkamp EA, Vanmaekelbergh D, Meijerink A (2000) The kinetics of the radiative and nonradiative processes in nanocrystalline ZnO particles upon photoexcitation. J Phys Chem B 104:1715–1723. doi:10.1021/jp993327z 96. Stroyuk OL, Shvalagin VV, Kuchmiy SY (2007) Photochemical synthesis of ZnO/Ag nanocomposites. J Nanopart Res 9:427–440. doi:10.1007/s11051-006-9086-5 97. Stroyuk OL, Dzhagan VM, Shvalagin VV et al (2010) Size-dependent optical properties of colloidal ZnO nanoparticles charged by photoexcitation. J Phys Chem C 114:220–225. doi:10.1021/jp908879h 98. Panasiuk YV, Raevskaya OE, Stroyuk OL et al (2014) Colloidal ZnO nanocrystals in dimethylsulfoxide: a new synthesis, optical, photo- and electroluminescent properties. Nanotechnology 25:075601. doi:10.1088/0957-4484/25/7/075601 99. Raevskaya AE, Panasiuk YV, Stroyuk OL et al (2014) Spectral and luminescent properties of ZnO/SiO2 core/shell nanoparticles with size-selected ZnO cores. RSC Adv 4:63393– 63401. doi:10.1039/C4RA07959K 100. Bao N, Shen L, Takata T et al (2007) Facile Cd-thiourea complex thermolysis synthesis of phase-controlled CdS nanocrystals for photocatalytic hydrogen production under visible light. J Phys Chem C 111:17527–17534. doi:10.1021/jp076566s 101. Jang JS, Joshi UA, Lee JS (2007) Solvothermal synthesis of CdS nanowires for photocatalytic hydrogen and electricity production. J Phys Chem C 111:13280–13287. doi:10.1021/jp072683b 290 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

102. Xiong S, Xi B, Wang C et al (2007) Tunable synthesis of various wurtzite ZnS architectural structures and their photocatalytic properties. Adv Func Mater 17:2728–2738. doi:10.1002/ adfm.200600891 103. Wang X, Wan F, Han K et al (2008) Large-scale synthesis well-dispersed ZnS microspheres and their photoluminescence, photocatalysis properties. Mater Charact 59:1765–1770. doi:10.1016/j.matchar.2008.04.005 104. Song L, Wei H, Xu H, Zhan J (2010) Ethanol-thermal synthesis of Cd1−xZnxS nanoparticles with enhanced photodegradation of 4-chlorophenol. Mater Res Bull 45:1396–1400. doi:10. 1016/j.materresbull.2010.06.044 105. Kale BB, Baeg JO, Lee SM et al (2006) CdIn2S4 nanotubes and marigold nanostructures: a visible-light photocatalyst. Adv Func Mater 16:1349–1354. doi:10.1002/adfm.200500525 106. Shen S, Zhao L, Guo L (2008) Cetyltrimethylammoniumbromide (CTAB)-assisted hydrothermal synthesis of ZnIn2S4 as an efficient visible-light-driven photocatalyst for hydrogen production. Inter J Hydrogen En 33:4501–4510. doi:10.1016/j.ijhydene.2008.05. 043 107. Shen S, Zhao L, Guo L (2009) Crystallite, optical and photocatalytic properties of visible-light-driven ZnIn2S4 photocatalysts synthesized via a surfactant-assisted hydrothermal method. Mater Res Bull 44:100–105. doi:10.1016/j.materresbull.2008.03.027 108. Chai B, Peng T, Zeng P et al (2011) Template-free hydrothermal synthesis of ZnIn2S4 floriated microsphere as an efficient photocatalyst for H2 production under visible-light irradiation. J Phys Chem C 115:6149–6155. doi:10.1021/jp1112729 109. Fan WJ, Zhou ZF, Xu WB et al (2010) Preparation of ZnIn2S4/fluoropolymer fiber composites and its photocatalytic H2 evolution from splitting of water using Xe lamp irradiation. Inter J Hydrogen En 35:6525–6530. doi:10.1016/j.ijhydene.2010.04.036 110. Jing D, Liu M, Chen Q, Guo L (2010) Efficient photocatalytic hydrogen production under visible light over a novel W-based ternary chalcogenide photocatalyst prepared by a hydrothermal process. Inter J Hydrogen En 35:8521–8527. doi:10.1016/j.ijhydene.2010.04. 170 111. Kale BB, Baeg JO, Kong K et al (2011) Self assembled CdLa2S4 hexagon flowers, nanoprisms and nanowires: novel photocatalysts for solar hydrogen production. J Mater Chem 21:2624–2631. doi:10.1039/C0JM02890H 112. Cao F, Shi W, Zhao L et al (2008) Hydrothermal hynthesis and high photocatalytic activity of 3D wurtzite ZnSe hierarchical nanostructures. J Phys Chem C 112:17095–17101. doi:10. 1021/jp8047345 113. Li G, Ciston S, Saponjic ZV et al (2008) Synthesizing mixed-phase TiO2 nanocomposites using a hydrothermal method for photooxidation and photoreduction applications. J Catal 253:105–110. doi:10.1016/j.jcat.2007.10.014 114. Kanna M, Wongnawa S (2008) Mixed amorphous and nanocrystalline TiO2 powders prepared by sol–gel method: Characterization and photocatalytic study. Mater Chem Phys 110:166–175. doi:10.1016/j.matchemphys.2008.01.037 115. Melghit K, Al-Rabaniah SS, Al-Amri I (2008) Low temperature preparation and characterization of nanospherical anatase TiO2 and its photocatalytic activity on Congo red degradation under sunlight. Ceram Internat 34:479–483. doi:10.1016/j.ceramint.2006.11. 011 116. Tang W, Chen Z, Katoh S (2004) Preparation of a nanocrystalline TiO2 photocatalyst using a dry-process with acetylene black. Chem Lett 33:1200–1201. doi:10.1246/cl.2004.1200 117. Tang J, Durrant JR, Klug DR (2008) Mechanism of photocatalytic water splitting in TiO2. Reaction of water with photoholes, importance of charge carrier dynamics, and evidence for four-hole chemistry. J Am Chem Soc 130:13885–13891. doi:10.1021/ja8034637 118. Kolen′ko YV, Churagulov BR, Kunst M et al (2004) Photocatalytic properties of titania powders prepared by hydrothermal method. Appl Catal B 54:51–58. doi:10.1016/j.apcatb. 2004.06.006 References 291

119. Neppolian B, Yamashita H, Okada Y et al (2004) Preparation of TiO2 photo-catalysts by multi-gelation and their photocatalytic reactivity for the degradation of 2-propanol. Chem Lett 33:268–269. doi:10.1246/cl.2004.268 120. Neppolian B, Yamashita H, Okada Y et al (2005) Preparation of unique TiO2 nanoparticle photocatalysts by a multigelation method for control of the physicochemical parameters and reactivity. Catal Lett 105:111–117. doi:10.1007/s10562-005-8013-1 121. Chou TP, Zhang Q, Russo B et al (2007) Titania particle size effect on the overall per– formance of dye-sensitized solar cells. J Phys Chem C 111:6296–6302. doi:10.1021/ jp068939f 122. Nakade S, Saito Y, Kubo W et al (2003) Influence of TiO2 nanoparticle size on electron diffusion and recombination in dye-sensitized TiO2 solar cells. J Phys Chem B 107:8607– 8611. doi:10.1021/jp034773w 123. Li Y, Lee NH, Hwang DS et al (2004) Synthesis and characterization of nano titania powder with high photoactivity for gas-phase photo-oxidation of benzene from TiOCl2 aqueous solution at low temperatures. Langmuir 20:10838–10844. doi:10.1021/la0489716 124. Testino A, Bellobono IR, Buscaglia V et al (2007) Optimizing the photocatalytic properties of hydrothermal TiO2 by the control of phase composition and particle morphology. A systematic approach. J Am Chem Soc 129:3564–2575. doi:10.1021/ja067050+ 125. Di Paola A, Cufalo G, Addamo M et al (2008) Photocatalytic activity of nanocrystalline TiO2 (brookite, rutile and brookite-based) powders prepared by thermohydrolysis of TiCl4 in aqueous chloride solutions. Colloids Surfaces A 17:366–367. doi:10.1016/j.colsurfa.2007. 11.005 126. Jia H, Xu H, Hu Y et al (2007) TiO2@CdS core–shell nanorods films: fabrication and dramatically enhanced photoelectrochemical properties. Electrochem Commun 9:354–360. doi:10.1016/j.elecom.2006.10.010 127. Bavykin DV, Kulak AN, Shvalagin VV et al (2011) Photocatalytic properties of rutile nanoparticles obtained via low temperature route from titanate nanotubes. J Photochem Photobiol, A 218:231–238. doi:10.1016/j.jphotochem.2011.01.006 128. Yu JC, Tang HY, Yu J et al (2002) Bactericidal and photocatalytic activities of TiO2 thin films prepared by sol–gel and reverse micelle methods. J Photochem Photobiol, A 153:211– 219. doi:10.1016/S1010-6030(02)00275-7 129. Xia XH, Liang Y, Wang Z et al (2008) Synthesis and photocatalytic properties of TiO2 nanostructures. Mater Res Bull 43:2187–2195. doi:10.1016/j.materresbull.2007.08.026 130. Jung JM, Wang M, Kim EJ et al (2008) Enhanced photocatalytic activity of Au-buffered TiO2 thin films prepared by radio frequency magnetron sputtering. Appl Catal B 84:389– 392. doi:10.1016/j.apcatb.2008.04.020 131. Hou YD, Wang XC, Wu L et al (2008) N-Doped SiO2/TiO2 mesoporous nanoparticles with enhanced photocatalytic activity under visible-light irradiation. Chemosphere 72:414–421. doi:10.1016/j.chemosphere.2008.02.035 132. Wilhelm P, Stephan D (2007) Photodegradation of rhodamine B in aqueous solution via SiO2@TiO2 nano-spheres. J Photochem Photobiol, A 185:19–25. doi:10.1016/j.jphotochem. 2006.05.003 133. Xu G, Zheng Z, Wu Y, Feng N (2009) Effect of silica on the microstructure and photocatalytic properties of titania. Ceram Internat 35:1–5. doi:10.1016/j.ceramint.2007.09. 008 134. Ismail AA, Ibrahim IA, Ahmed MS et al (2004) Sol–gel synthesis of titania-silica photocatalyst for cyanide photodegradation. J Photochem Photobiol, A 163:445–451. doi:10.1016/j.jphotochem.2004.01.017 135. Anderson C, Bard AJ (1995) An improved photocatalyst of TiO2/SiO2 prepared by a sol–gel synthesis. J Phys Chem 99:9882–9885. doi:10.1021/j100024a033 136. Kominami H, Yukishita K, Kimura T et al (2008) Direct solvothermal formation of nanocrystalline TiO2 on porous SiO2 adsorbent and photocatalytic removal of nitrogen oxides in air over TiO2–SiO2 composites. Topics Catal 47:155–161. doi:10.1007/s11244- 007-9016-5 292 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

137. Lim SH, Phonthammachai N, Pramana SS, White TJ (2008) Simple route to monodispersed silica-titania core-shell photocatalysts. Langmuir 24:6226–6231. doi:10.1021/la703899j 138. Li K, Ding YY, Guo J, Wang D (2008) Surface electron structures and mechanism of nonradiative transitions on crystalline TiO2 nanoparticles. Mater Chem Phys 112:1001– 1007. doi:10.1016/j.matchemphys.2008.07.010 139. Liao DL, Liao BQ (2007) Shape, size and photocatalytic activity control of TiO2 nanoparticles with surfactants. J Photochem Photobiol, A 187:363–369. doi:10.1016/j. jphotochem.2006.11.003 140. Scotti R, Bellobono IR, Canevali C et al (2008) Sol–gel pure and mixed-phase titanium dioxide for photocatalytic purposes: Relations between phase composition, catalytic activity, and charge-trapped sites. Chem Mater 20:4051–4061. doi:10.1021/cm800465n 141. Neppolian B, Celik E, Anpo M, Choi H (2008) Ultrasonic-assisted pH swing method for the synthesis of highly efficient TiO2 nano-size photocatalysts. Catal Lett 125:183–191. doi:10. 1007/s10562-008-9563-9 142. Choi H, Stathatos E, Dionysiou DD (2007) Effect of surfactant in a modified sol on the physicochemical properties and photocatalytic activity of crystalline TiO2 nanoparticles. Topics Catal 44:513–521. doi:10.1007/s11244-006-0099-1 143. Jiu J, Isoda S, Adachi M, Wang F (2007) Preparation of TiO2 nanocrystalline with 3–5nm and application for dye-sensitized solar cell. J Photochem Photobiol, A 189:314–321. doi:10.1016/j.jphotochem.2007.02.021 144. Stengl V, Bakardjieva S, Murafa N et al (2007) Preparation, characterization and photocatalytic activity of optically transparent titanium dioxide nanoparticles. Mater Chem Phys 105:38–46. doi:10.1016/j.matchemphys.2007.04.037 145. Carrera R, Vázquez AL, Arce E et al (2007) Photodecomposition of NO by sol–gel TiO2 catalysts under atmospheric conditions: Effect of the method on the textural and morphologic properties. J Alloys Compounds 434–435:788–791. doi:10.1016/j.jallcom.2006.08.334 146. Zhang S, Yu Q, Chen Z et al (2007) Nano-TiO2 particles with increased photocatalytic activity prepared by the miniemulsion method. Mater Lett 61:4839–4842. doi:10.1016/j. matlet.2007.03.054 147. Kubacka A, Fuerte A, Martinez-Arias A, Fernandez-Garcia M (2007) Nanosized Ti–V mixed oxides: Effect of doping level in the photocatalytic degradation of toluene using sunlight-type excitation. Appl Catal B 74:26–33. doi:10.1016/j.apcatb.2007.01.011 148. Jang HD, Kim SK, Kim SJ (2001) Effect of particle size and phase composition of titanium dioxide nanoparticles on the photocatalytic properties. J Nanopart Res 3:141–147. doi:10. 1023/A:1017948330363 149. Marimuthu A, Madras G (2008) Photocatalytic oxidative degradation of poly(alkylacrylates) with nanoTiO2. Ind Eng Chem Res 47:2182–2190. doi:10.1021/ie0712939 150. Vinu R, Madras G (2008) Photocatalytic degradation of poly(acrylamide-co-acrylic acid). J Phys Chem C 112:8928–8935. doi:10.1021/jp801887t 151. Surolia PK, Lazar MA, Tayade RJ, Jasra RV (2008) Photocatalytic degradation of 3,3′- dimethylbiphenyl-4,4′-diamine (o-tolidine) over nanocrystalline TiO2 synthesized by sol– gel, solution combustion, and hydrothermal methods. Ind Eng Chem Res 47:5847–5855. doi:10.1021/ie800073j 152. Priya MH, Madras G (2006) Photocatalytic degradation of nitrobenzenes with combustion synthesized nano-TiO2. J Photochem Photobiol, A 178:1–7. doi:10.1016/j.jphotochem.2005. 06.012 153. Vinu R, Madras G (2008) Kinetics of simultaneous photocatalytic degradation of phenolic compounds and reduction of metal ions with nano-TiO2. Environ Sci Technol 42:913–919. doi:10.1021/es8025648 154. Alves AK, Berutti FA, Clemens FJ et al (2009) Photocatalytic activity of titania fibers obtained by electrospinning. Mater Res Bull 44:312–317. doi:10.1016/j.materresbull.2008. 06.001 155. Li W, Liu C, Zhou Y et al (2008) Enhanced photocatalytic activity in anatase/TiO2(B) core-shell nanofiber. J Phys Chem C 112:20539–20545. doi:10.1021/jp808183q References 293

156. Lin Y (2008) Photocatalytic activity of TiO2 nanowire arrays. Mater Lett 62:1246–1248. doi:10.1016/j.matlet.2007.08.021 157. Ren X, Han D, Chen D, Tang F (2007) Large-scale synthesis of hexagonal cone-shaped ZnO nanoparticles with a simple route and their application to photocatalytic degradation. Mater Res Bull 42:807–813. doi:10.1016/j.materresbull.2006.08.030 158. Wang H, Xie C (2008) Effect of annealing temperature on the microstructures and photocatalytic property of colloidal ZnO nanoparticles. J Phys Chem Solids 69:2440–2444. doi:10.1016/j.jpcs.2008.04.036 159. Ye C, Bando Y, Shen G, Golberg D (2006) Thickness-dependent photocatalytic performance of ZnO nanoplatelets. J Phys Chem B 110:15146–15151. doi:10.1021/ jp061874w 160. Dodd AC, McKinley AJ, Saunders M, Tsuzuki T (2006) Effect of particle size on the photocatalytic activity of nanoparticulate zinc oxide. J Nanopart Res 8:43–51. doi:10.1007/ s11051-005-5131-z 161. Liqiang J, Baifu X, Fulong Y et al (2004) Deactivation and regeneration of ZnO and TiO2 nanoparticles in the gas phase photocatalytic oxidation of n-C7H16 or SO2. Appl Catal A 275:49–54. doi:10.1016/j.apcata.2004.07.019 162. Wang G, Chen D, Zhang H et al (2008) Tunable photocurrent spectrum in well-oriented zinc oxide nanorod arrays with enhanced photocatalytic activity. J Phys Chem C 112:8850–8855. doi:10.1021/jp800379k 163. Sun T, Qiu J, Liang C (2008) Controllable fabrication and photocatalytic activity of ZnO nanobelt array. J Phys Chem C 112:715–721. doi:10.1021/jp710071f 164. Wan Q, Wang TH, Zhao JC (2005) Enhanced photocatalytic activity of ZnO nanotetrapods. Appl Phys Lett 87:083105. doi:10.1063/1.2034092 165. Wang Y, Li X, Lu G et al (2008) Highly oriented 1-D ZnO nanorod arrays on zinc foil: direct growth from substrate, optical properties and photocatalytic activities. J Phys Chem C 112:7332–7336. doi:10.1021/jp7113175 166. Sadakane M, Sasaki K, Kunioku H et al (2008) Preparation of nano-structured crystalline tun–gsten (VI) oxide and enhanced photocatalytic activity for decomposition of organic compounds under visible light irradiation. Chem Commun 48:6552–6554. doi:10.1039/ B815214D 167. Sadakane M, Tamura N, Kanome N et al (2011) Preparation of crystalline tungsten oxide nanorods with enhanced photocatalytic activity under visible light irradiation. Chem Lett 40:443–444. doi:10.1246/cl.2011.443 168. Su J, Feng X, Sloppy JD et al (2011) Vertically aligned WO3 nanowire arrays grown directly on transparent conducting oxide coated glass: synthesis and photoelectrochemical properties. Nano Lett 11:203–208. doi:10.1021/nl1034573 169. Zhang J, Shi F, Lin J et al (2008) Self-assembled 3-D architectures of BiOBr as a visible light-driven photocatalyst. Chem Mater 20:2937–2941. doi:10.1021/cm7031898 170. Zhang X, Ai Z, Jia F, Zhang L (2008) Generalized one-pot synthesis, characterization, and photocatalytic activity of hierarchical BiOX (X = Cl, Br, I) nanoplate microspheres. J Phys Chem C 112:747–753. doi:10.1021/jp077471t 171. Xiao X, Zhang WD (2010) Facile synthesis of nanostructured BiOI microspheres with high visible light-induced photocatalytic activity. J Mater Chem 20:5866–5870. doi:10.1039/ C0JM00333F 172. Sun M, Li D, Zhang W et al (2010) Rapid microwave hydrothermal synthesis of GaOOH nano-rods with photocatalytic activity toward aromatic compounds. Nanotechnology 21:355601. doi:10.1088/0957-4484/21/35/355601 173. Chang CA, Ray B, Paul DK et al (2008) Photocatalytic reaction of acetaldehyde over SrTiO3 nanoparticles. J Mol Catal A 281:99–106. doi:10.1016/j.molcata.2007.08.014 174. Liu Y, Xie L, Li Y et al (2008) Synthesis and high photocatalytic hydrogen production of SrTiO3 nanoparticles from water splitting under UV irradiation. J Power Sources 193:701– 707. doi:10.1016/j.jpowsour.2008.05.057 294 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

175. Zhang G, He F, Zou X et al (2008) Hydrothermal preparation and photocatalytic properties of sheet-like nanometer niobate K4Nb6O17. J Phys Chem Solids 69:1471–1474. doi:10.1016/ j.jpcs.2007.10.095 176. Zhou C, Chen C, Wang Q (2011) High photocatalytic activity of porous K4Nb6O17 microsphere with large surface area prepared by homogeneous precipitation using urea. J Mol Catal A 339:37–42. doi:10.1016/j.molcata.2011.02.009 177. Sarahan MC, Carrol EC, Allen M et al (2008) K4Nb6O17-derived photocatalysts for hydrogen evolution from water: nanoscrolls versus nanosheets. J Solid State Chem 181:1681–1686. doi:10.1016/j.jssc.2008.06.021 178. Zhang S, Zhang C, Man Y, Zhu Y (2006) Visible-light-driven photocatalyst of Bi2WO6 nanoparticles prepared via amorphous complex precursor and photocatalytic properties. J Solid State Chem 179:62–69. doi:10.1016/j.jssc.2005.09.041 179. Wu L, Bi J, Li Z et al (2008) Rapid preparation of Bi2WO6 photocatalyst with nanosheet morphology via microwave-assisted solvothermal synthesis. Catal Today 131:15–20. doi:10. 1016/j.cattod.2007.10.089 180. Bi J, Wu L, Li J et al (2007) Simple solvothermal routes to synthesize nanocrystalline Bi2MoO6 photocatalysts with different morphologies. Acta Mater 55:4699–4705. doi:10. 1016/j.actamat.2007.04.034 181. Wu J, Duan F, Zheng Y, Xie Y (2007) Synthesis of Bi2WO6 nanoplate-built hierarchical nest-like structures with visible-light-induced photocatalytic activity. J Phys Chem C 111:12866–12871. doi:10.1021/jp073877u 182. Zhang C, Zhu Y (2005) Synthesis of square Bi2WO6 nanoplates as high-activity visible-light-driven photocatalysts. Chem Mater 17:3537–3545. doi:10.1021/cm0501517 183. Lin J, Lin J, Zhu Y (2007) Controlled synthesis of the ZnWO4 nanostructure and effects on the photocatalytic performance. Inorg Chem 46:8372–8378. doi:10.1021/ic701036k 184. Zhang L, Chen D, Jiao X (2006) Monoclinic structured BiVO4 nanosheets: hydrothermal preparation, formation mechanism, and coloristic and photocatalytic properties. J Phys Chem B 110:2668–2673. doi:10.1021/jp056367d 185. Mahapatra S, Nayak SK, Madras G, Guru Row TN (2008) Microwave synthesis and photocatalytic activity of nano lanthanide (Ce, Pr, and Nd) orthovanadates. Ind Eng Chem Res 47:6509–6516. doi:10.1021/ie8003094 186. Bouras P, Lianos P (2005) Photodegradation of dyes in aqueous solutions catalyzed by highly efficient nanocrystalline titania films. J Appl Electrochem 35:831–836. doi:10.1007/ s10800-005-1616-9 187. Bouras P, Stathatos E, Lianos P (2007) Pure versus ion-metal-doped nanocrystalline titania for photocatalysis. Appl Catal B 73:51–59. doi:10.1016/j.apcatb.2006.06.007 188. Strataki N, Bekiari V, Kondarides DI, Lianos P (2007) Hydrogen production by photocatalytic alcohol reforming employing highly efficient nanocrystalline titania films. Appl Catal B 77:184–189. doi:10.1016/j.apcatb.2007.07.015 189. Strataki N, Bekiari V, Stathatos E, Pianos P (2007) Effect of aggregation of dyes adsorbed on nanocrystalline titania films on the efficiency of photodegradation. J Photochem Photobiol, A 191:13–18. doi:10.1016/j.jphotochem.2007.03.023 190. Yu JC, Yu J, Ho W, Zhao J (2002) Light-induced super-hydrophilicity and photocatalytic activity of mesoporous TiO2 thin films. J Photochem Photobiol, A 148:331–339. doi:10. 1016/S1010-6030(02)00060-6 191. Liu Y, Wang X, Yang F, Yang X (2008) Excellent antimicrobial properties of mesoporous anatase TiO2 and Ag/TiO2 composite films. Micropor Mesopor Mater 114:431–439. doi:10. 1016/j.micromeso.2008.01.032 192. Angelome PC, Andrini L, Calvo ME et al (2007) Mesoporous anatase TiO2 films: use of Ti K XANES for the quantification of the nanocrystalline character and substrate effects in the photocatalysis behavior. J Phys Chem C 111:10886–10893. doi:10.1021/jp069020z 193. Chen Y, Dionysiou DD (2008) Bimodal mesoporous TiO2–P25 composite thick films with high photocatalytic activity and improved structural integrity. Appl Catal B 80:147–155. doi:10.1016/j.apcatb.2007.11.010 References 295

194. Habibi MH, Esfahani MN, Egerton TA (2007) Photochemical charactrization and photocatalytic properties of a nanostructure composite TiO2 film. Inter J Photoen 2007:1–8 195. Negishi N, Matsuzawa S, Takeuchi K, Pichat P (2007) Transparent micrometer-thick TiO2 films on SiO2-coated glass prepared by repeated dip-coating/calcination: characteristics and photocatalytic activities for removing acetaldehyde or toluene in air. Chem Mater 19:3808– 3814. doi:10.1021/cm070320i 196. Guan H, Zhu L, Zhou H, Tang H (2008) Rapid probing of photocatalytic activity on titania-based self-cleaning materials using 7-hydroxycoumarin fluorescent probe. Anal Chem Acta 608:73–78. doi:10.1016/j.aca.2007.12.009 197. Yang J, Chen C, Ji H et al (2005) Mechanism of TiO2-assisted photocatalytic degradation of dyes under visible irradiation: Photoelectrocatalytic study by TiO2-film electrodes. J Phys Chem B 109:21900–21907. doi:10.1021/jp0540914 198. Ercan A, Bakir U, Karakas G (2006) Photocatalytic microbal inactivation over Pd doped SnO2 and TiO2 thin films. J Photochem Photobiol, A 184:313–321. doi:10.1016/j. jphotochem.2006.05.001 199. Mills A, Elliott N, Hill G et al (2003) Preparation and characterisation of novel thick sol–gel titania film photocatalysts. Photochem Photobiol Sci 2:591–596. doi:10.1039/B212865A 200. Ao CH, Lee SC, Yu JC (2003) Photocatalyst TiO2 supported on glass fiber for indoor air purification: effect of NO on the photodegradation of CO and NO2. J Photochem Photobiol, A 156:171–177. doi:10.1016/S1010-6030(03)00009-1 201. Sunada K, Watanabe T, Hashimoto K (2003) Studies on photokilling of bacteria on TiO2 thin film. J Photochem Photobiol, A 156:227–233. doi:10.1016/S1010-6030(02)00434-3 202. Xu J, Ao Y, Fu D et al (2008) Photocatalytic activity on TiO2-coated side-glowing optical fiber reactor under solar light. J Photochem Photobiol, A 199:165–169. doi:10.1016/j. jphotochem.2008.05.019 203. Yang B, Barnes PRF, Zhang Y, Luca V (2007) Tungsten trioxide films with controlled morphology and strong photocatalytic activity via a simple sol–gel route. Catal Lett 118:280–284. doi:10.1007/s10562-007-9188-4 204. Nakamura N, Tanaka T, Nakato Y (2004) Mechanism for visible light responses in anodic photocurrents at N-doped TiO2 film electrodes. J Phys Chem B 108:10617–10620. doi:10. 1021/jp048112q 205. Umemura Y, Shinohara E, Koura A et al (2006) Photocatalytic decomposition of an alkylammonium cation in a Langmuir-Blodgett film of a titania nanosheet. Langmuir 22:3870–3877. doi:10.1021/la0530743 206. Mills A, Elliott N, Parkin IP et al (2002) Novel TiO2 CVD films for semiconductor photocatalysis. J Photochem Photobiol, A 151:171–179. doi:10.1016/S1010-6030(02) 00190-9 207. Brook LA, Evans P, Foster HA et al (2007) Highly bioactive silver and silver/titania composite films grown by chemical vapour deposition. J Photochem Photobiol, A 187:53– 63. doi:10.1016/j.jphotochem.2006.09.014 208. Ohko Y, Nakamura Y, Fukuda A et al (2008) Photocatalytic oxidation of nitrogen dioxide with TiO2 thin films under continuous UV-light illumination. J Phys Chem C 112:10502– 10508. doi:10.1021/jp802959c 209. Yoon KH, Noh JS, Kwon CH, Muhammed M (2006) Photocatalytic behavior of TiO2 thin films prepared by sol–gel process. Mater Chem Phys 95:79–83. doi:10.1016/j.matchemphys. 2005.06.001 210. Mishra PR, Shukla PK, Srivastava ON (2007) Study of modular PEC solar cells for photoelectrochemical splitting of water employing nanostructured TiO2 photoelectrodes. Inter J Hydrogen En 32:1680–1685. doi:10.1016/j.ijhydene.2006.10.002 211. Li D, Guo Y, Hu C et al (2004) Preparation, characterization and photocatalytic property of 7− the PW11O39 /TiO2 composite film towards azo-dye degradation. J Mol Catal A 207:183– 193. doi:10.1016/S1381-1169(03)00491-6 296 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

212. Gao R, Stark J, Bahnemann DW, Rabani J (2002) Quantum yields of hydroxyl radicals in illuminated TiO2 nanocrystallite layers. J Photochem Photobiol, A 148:387–391. doi:10. 1016/S1010-6030(02)00066-7 213. Ao Y, Xu J, Fu D, Yuan C (2008) Preparation of porous titania thin film and its photocatalytic activity. Appl Surf Sci 255:3137–3140. doi:10.1016/j.apsusc.2008.08.102 214. Piris J, Ferguson AJ, Blackburn JL et al (2008) Efficient photoinduced charge injection from chemical bath deposited CdS mesoporous TiO2 probed with time-resolved microwave conductivity. J Phys Chem C 112:7742–7749. doi:10.1021/jp800527r 215. Song XM, Wu JM, Yan M (2008) Photocatalytic and photoelectrocatalytic degradation of aqueous Rhodamine B by low-temperature deposited anatase thin films. Mater Chem Phys 112:510–515. doi:10.1016/j.matchemphys.2008.06.009 216. Yu JG, Yu HG, Cheng B et al (2003) The effect of calcination temperature on the surface microstructure and photocatalytic activity of TiO2 thin films prepared by liquid phase deposition. J Phys Chem B 107:13871–13879. doi:10.1021/jp036158y 217. Gu DE, Yang BC, Hu YD (2007) A novel method for preparing V-doped titanium dioxide thin film photocatalysts with high photocatalytic activity under visible light irradiation. Catal Lett 118:254–259. doi:10.1007/s10562-007-9179-5 218. Sagawa T, Kotani M, Nada H et al (2003) Photoinduced reduction of methylviologen with TiO2/polymer films. Chem Lett 32:962–963. doi:10.1246/cl.2003.962 219. Zhao Y, Zhang X, Zhai J et al (2008) Enhanced photocatalytic activity of hierarchically micro-/nano-porous TiO2 films. Appl Catal B 83:24–29. doi:10.1016/j.apcatb.2008.01.035 220. Takeuchi M, Yamashita H, Matsuoka M et al (2000) Photocatalytic decomposition of NO on titanium oxide thin film photocatalysts prepared by an ionized cluster beam technique. Catal Lett 66:185–187. doi:10.1023/A:1019095406121 221. Burlacov I, Jirkovsky J, Kavan L et al (2007) Cold gas dynamic spraying (CGDS) of TiO2 (anatase) powders onto poly(sulfone) substrates: microstructural characterisation and photocata–lytic efficiency. J Photochem Photobiol, A 187:285–292. doi:10.1016/j. jphotochem.2006.10.023 222. Ryu J, Park DS, Hahn BD et al (2008) Photocatalytic TiO2 thin films by aerosol-deposition: from micron-sized particles to nano-grained thin film at room temperature. Appl Catal B 83:1–7. doi:10.1111/j.1551-2916.2009.03391.x 223. Ohko Y, Tatsuma T, Fujishima A (2001) Characterization of TiO2 photocatalysis in the gas phase as a photoelectrochemical system: behavior of salt-modified systems. J Phys Chem B 105:10016–10021. doi:10.1021/jp011459y 224. Shinde PS, Sadale SB, Patil PS et al (2008) Properties of spray deposited titanium dioxide thin films and their application in photoelectrocatalysis. Sol En Mater Sol Cells 92:283–290. doi:10.1016/j.solmat.2007.09.001 225. Thimsen E, Rastgar N, Biswas P (2008) Nanostructured TiO2 films with controlled morphology synthesized in a single step process: performance of dye-sensitized solar cells and photo watersplitting. J Phys Chem C 112:4134–4140. doi:10.1021/jp710422f 226. Li M, Su J, Guo L (2008) Preparation and characterization of ZnIn2S4 thin films deposited by spray pyrolysis for hydrogen production. Inter J Hydrogen En 33:2891–2896. doi:10. 1016/j.ijhydene.2008.04.008 227. Kemell M, Pore V, Tupala J et al (2007) Atomic layer deposition of nanostructured TiO2 photocatalyst via template approach. Chem Mater 19:1816–1820. doi:10.1021/cm062576e 228. Liu R, Lin Y, Chou LY et al (2011) Splitting by tungsten oxide prepared by atomic layer deposition and decorated with an oxygen-evolving catalyst. Angew Chem Int Ed 50:499– 502. doi:10.1002/ange.201004801 229. Klahr BM, Martinson ABF, Hamann TW (2011) Photoelectrochemical investigation of ultrathin film iron oxide solar cells prepared by atomic layer deposition. Langmuir 27:461– 468. doi:10.1021/la103541n 230. Kitano M, Tsujimaru K, Anpo M (2008) Hydrogen production using highly active titanium oxide-based photocatalysts. Topics Catal 49:4–17. doi:10.1007/s11244-008-9059-2 References 297

231. Dholam R, Patel N, Adami M, Miotello A (2008) Physically and chemically synthesized TiO2 composite thin films for hydrogen production by photocatalytic water splitting. Inter J Hydrogen En 33:6896–6903. doi:10.1016/j.ijhydene.2008.08.061 232. Selli E, Chiarello GL, Quartarone E et al (2007) A photocatalytic water splitting device for separate hydrogen and oxygen evolution. Chem Commun 5022–5024. doi:10.1039/ B711747G 233. Kitano M, Takeuchi M, Matsuoka M et al (2005) Preparation of visible light-responsive TiO2 thin film photocatalysts by an RF magnetron sputtering deposition method and their photocatalytic reactivity. Chem Lett 34:616–617. doi:10.1246/cl.2005.616 234. Cao J, Kako T, Li P et al (2011) Fabrication of p-type CaFe2O4 nanofilms for photoelectrochemical hydrogen generation. Electrochem Commun 13:275–278. doi:10. 1016/j.elecom.2011.01.002 235. Hu YS, Kleiman-Shwarsctein A, Forman AJ et al (2008) Pt-doped a-Fe2O3 thin films active for photoelectrochemical water splitting. Chem Mater 20:3803–3805. doi:10.1021/ cm800144q 236. Kleiman-Shwarsctein A, Hu YS, Forman AJ et al (2008) Electrodeposition of a-Fe2O3 doped with Mo or Cr as photoanodes for photocatalytic water splitting. J Phys Chem C 112:15900–15907. doi:10.1021/jp803775j 237. Nian JN, Hu CC, Teng H (2008) Electrodeposited p-type Cu2O for H2 evolution from photoelectrolysis of water under visible light illumination. Inter J Hydrogen En 33:2897– 2903. doi:10.1016/j.ijhydene.2008.03.052 238. Yu H, Quan X, Zhang Y et al (2008) Electrochemically assisted photocatalytic inactivation of Escherichia coli under visible light using a ZnIn2S4 film electrode. Langmuir 24:7599– 7604. doi:10.1021/la800835k 239. Chu SZ, Inoue S, Wada K et al (2003) Highly porous (TiO2-SiO2-TeO2)/Al2O3/TiO2 composite nanostructures on glass with enhanced photocatalysis fabricated by anodization and sol–gel process. J Phys Chem B 107:6586–6589. doi:10.1021/jp0349684 240. Kang Z, Tsang CHA, Wong NB et al (2007) Silicon quantum dots: a general photocatalyst for reduction, decomposition, and selective oxidation reactions. J Am Chem Soc 129:12090– 12091. doi:10.1021/ja075184x 241. Schürch D, Currao A, Sarkar S et al (2002) The silver chloride photoanode in photoelectrochemical water splitting. J Phys Chem B 106:12764–12775. doi:10.1021/ jp0265081 242. Soller-Illia G, Sanchez C, Lebeau B, Patarin J (2002) Chemical strategies to design textured materials: from microporous and mesoporous oxides to nanonetworks and hierarchical structures. Chem Rev 102:4093–4138. doi:10.1021/cr0200062 243. Shchukin DG, Sviridov DV (2006) Photocatalytic processes in spatially confined micro- and nanoreactors. J Photochem Photobiol, C 7:23–39. doi:10.1016/j.jphotochemrev.2006.03.002 244. Stroyuk OL, Shvalagin VV, Raevskaya AE et al (2008) Photochemical formation of semiconducting nanostructures. Theoret Experim Chem 44:205–231 245. Zhang X, Yang H, Zhang F, Chan KY (3007) Preparation and characterization of Pt–TiO2– SiO2 mesoporous materials and visible-light photocatalytic performance. Mater Lett 61:2231–2234. doi:10.1016/j.matlet.2006.08.053 246. Smirnova N, Gnatyuk Y, Eremenko A et al (2006) Photoelectrochemical characterization and photocatalytic properties of mesoporous TiO2/ZrO2 films. Inter J Photoen 1–6. doi:10. 1155/IJP/2006/85469 247. Mohamed MM, Bayoumy WA, Khairy M, Mousa MA (2008) Structural features and photocatalytic behavior of titania and titania supported vanadia synthesized by polyol functionalized materials. Mircopor Mesopor Mater 109:445–457. doi:10.1016/j.micromeso. 2007.05.055 248. Deshpande AS, Shchukin DG, Ustinovich E et al (2005) Titania and mixed titania/aluminum, gallium, or indium oxide spheres: sol–gel/template synthesis and photocatalytic properties. Adv Func Mater 15:239–245. doi:10.1002/adfm.200400220 298 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

249. Yuan S, Sheng Q, Zhang J et al (2008) Synthesis of thermally stable mesoporous TiO2 and investigation of its photocatalytic activity. Micropor Mesopor Mater 110:501–507. doi:10. 1016/j.micromeso.2007.06.039 250. Oh ST, Choi JS, Lee HS et al (2007) H2O-controlled synthesis of TiO2 with nanosized channel structure through in situ esterification and its application to photocatalytic oxidation. J Mol Catal A 267:112–119. doi:10.1016/j.molcata.2006.11.028 251. Chen L, Yao B, Cao Y, Fan K (2007) Synthesis of well-ordered mesoporous titania with tunable phase content and high photoactivity. J Phys Chem C 111:11849–11853. doi:10. 1021/jp072070z 252. Fan X, Yu T, Wang Y et al (2008) Role of phosphorus in synthesis of phosphated mesoporous TiO2 photocatalytic materials by EISA method. Appl Surf Sci 254:5191–5198. doi:10.1016/j.apsusc.2008.02.038 253. Huang J, Wang X, Hou Y et al (2008) Synthesis of functionalized mesoporous TiO2 molecular sieves and their application in photocatalysis. Micropor Mesopor Mater 110:543– 552. doi:10.1016/j.micromeso.2007.06.055 254. Zhang L, Yu JC (2003) A sonochemical approach to hierarchical porous titania spheres with enhanced photocatalytic activity. Chem Commun 2078–2079. doi:10.1039/B306013F 255. Ismail AA, Bahnemann DW (2011) Mesostructured Pt/TiO2 nanocomposites as highly active photocatalysts for the photooxidation of dichloroacetic acid. J Phys Chem C 115:5784–5791. doi:10.1021/jp110959b 256. Wang X, Mitchell DRG, Prince K et al (2008) Gold nanoparticle incorporation into porous titania networks using an agarose gel templating technique for photocatalytic applications. Chem Mater 20:3917–3926. doi:10.1021/cm703509f 257. van der Meulen T, Mattson A, Österlund L (2007) A comparative study of the photocatalytic oxidation of propane on anatase, rutile, and mixed-phase anatase–rutile TiO2 nanoparticles: role of surface intermediates. J Catal 251:131–144. doi:10.1016/j.jcat.2007.07.002 258. An T, Liu J, Li G et al (2008) Structural and photocatalytic degradation characteristics of hydrothermally treated mesoporous TiO2. Appl Catal A 350:237–243. doi:10.1016/j.apcata. 2008.08.022 259. Shen Q, Yang H, Gao J, Yang J (2007) Low-temperature fabrication of porous anatase TiO2 film with tiny slots and its photocatalytic activity. Mater Lett 61:4160–4162. doi:10.1016/j. matlet.2007.01.045 260. Sang L, Dai H, Sun J et al (2010) Fabrication of the hydrogen-evolving photocatalyst with mesoporous structure. Inter J Hydrogen En 35:7098–7103. doi:10.1016/j.ijhydene.2010.01. 019 261. Wang HW, Lin HC, Kuo CH et al (2008) Synthesis and photocatalysis of mesoporous anatase TiO2 powders incorporated Ag nanoparticles. J Phys Chem Solids 69:633–636. doi:10.1016/j.jpcs.2007.07.052 262. Sreethawong T, Suzuki Y, Yoshikawa S (2006) Platinum-loaded mesoporous titania by single-step sol–gel process with surfactant template: photocatalytic activity for hydrogen evolution. Compt Rend Chim 9:307–314. doi:10.1016/j.crci.2005.05.015 263. Kozlova EA, Vorontsov AV (2007) Influence of mesoporous and platinum-modified titanium dioxide preparation methods on photocatalytic activity in liquid and gas phase. Appl Catal B 77:35–45. doi:10.1016/j.apcatb.2007.07.001 264. Tian G, Fu H, Jing L et al (2008) Preparation and characterization of stable biphase TiO2 photocatalyst with high crystallinity, large surface area, and enhanced photoactivity. J Phys ChemC 112:3083–3089. doi:10.1021/jp710283p 265. De Witte K, Ribbens S, Meynen V et al (2008) Photoatalytic study of P25 and mesoporous titania in aqueous and gaseous environment. Catal Commun 9:1787–1792. doi:10.1016/j. catcom.2008.02.013 266. Ao Y, Xu J, Fu D, Yuan C (2008) Preparation of Ag-doped mesoporous titania and its enhanced photocatalytic activity under UV light irradiation. J Phys Chem Solids 69:2660– 2664. doi:10.1016/j.jpcs.2008.06.100 References 299

267. Song C, Wang D, Xu Y, Hu Z (2011) Preparation of Ag-TiO2 hollow structures with enhanced photocatalytic activity. Mater Lett 65:908–910. doi:10.1016/j.matlet.2010.12.015 268. Dai Q, Shi LY, Luo YG et al (2002) Effect of templates on the structure, stability and photocatalytic activity mesostructured TiO2. J Photochem Photobiol, A 148:295–301. doi:10.1016/S1010-6030(02)00056-4 269. Lakshminarasimhan N, Bae E, Choi W (2007) Enhanced photocatalytic production of H2 on mesoporous TiO2 prepared by template-free method: role of interparticle charge transfer. J Phys Chem C 111:15244–15250. doi:10.1021/jp0752724 270. Matsunaga T, Yamaoka H, Ohtani S et al (2008) High photocatalytic activity of palladium-deposited mesoporous TiO2/SiO2 ﬁbers. Appl Catal A 351:231–238. doi:10. 1016/j.apcata.2008.09.020 271. Pavasupree S, Ngamsinlapasathian S, Nakajima M et al (2006) Synthesis, characterization, photocatalytic activity and dye-sensitized solar cell performance of nanorods/nanoparticles TiO2 with mesoporous structure. J Photochem Photobiol, A 184:163–169. doi:10.1016/j. jphotochem.2006.04.010 272. Wang G (2007) Hydrothermal synthesis and photocatalytic activity of nanocrystalline TiO2 powders in ethanol–water mixed solutions. J Mol Catal A 274:185–191. doi:10.1016/j. molcata.2007.05.009 273. Yu J, Zhang L, Cheng B, Su Y (2007) Hydrothermal preparation and photocatalytic activity of hierarchically sponge-like macro-/mesoporous titania. J Phys Chem C 111:10582–10589. doi:10.1021/jp0707889 274. Yu JC, Zhang L, Yu J (2002) Rapid synthesis of mesoporous TiO2 with high photocatalytic activity by ultrasound-induced agglomeration. New J Chem 26:416–420. doi:10.1039/ B109173E 275. Fang J, Wang F, Qian K et al (2008) Bifunctional N-doped mesoporous TiO2 photocatalysts. J Phys ChemC 112:18150–18156. doi:10.1021/jp805926b 276. Sreethawong T, Ngamsinlapasathian S, Suzuki Y, Yoshikawa S (2005) Nanocrystalline meso–porous Ta2O5-based photocatalysts prepared by surfactant-assisted templating sol–gel process for photocatalytic H2 evolution. J Mol Catal A 235:1–11. doi:10.1016/j.molcata. 2005.03.021 277. Noda Y, Lee B, Domen K, Kondo JN (2008) Synthesis of crystallized mesoporous tantalum oxide and its photocatalytic activity for overall water splitting under ultraviolet light irradiation. Chem Mater 20:5361–5367. doi:10.1021/cm703202n 278. Chen X, Yu T, Fan X et al (2007) Enhanced activity of mesoporous Nb2O5 for photocatalytic hydrogen production. Appl Surface Sci 253:8500–8506. doi:10.1016/j. apsusc.2007.04.035 279. Abe R, Shinohara K, Tanaka A et al (1997) Preparation of porous niobium oxides by soft-chemical process and their photocatalytic activity. Chem Mater 9:2179–2184. doi:10. 1021/cm970284v 280. Zhu H, Hagfeldt A, Boschloo G (2007) Photoelectrochemistry of mesoporous NiO electrodes in iodide/triiodide electrolytes. J Phys Chem C 111:17455–17458. doi:10.1021/ jp077134k 281. Sathish M, Viswanath RP (2007) Photocatalytic generation of hydrogen over mesoporous CdS nanoparticle: effect of particle size, noble metal and support. Catal Today 129:421–427. doi:10.1016/j.cattod.2006.12.008 282. Wang L, Tang K, Liu Z et al (2011) Single-crystalline ZnSn(OH)6 hollow cubes via self-templated synthesis at room temperature and their photocatalytic properties. J Mater Chem 21:4352–4357. doi:10.1039/C0JM03734F 283. Jang JS, Yu CJ, Choi SH et al (2008) Topotactic synthesis of mesoporous ZnO and ZnS nanoplates and their photocatalytic activity. J Catal 254:144–155. doi:10.1016/j.jcat.2007. 12.010 284. Li Y, He X, Cao M (2008) Micro-emulsion-assisted synthesis of ZnS nanospheres and their photocatalytic activity. Mater Res Bull 43:3100–3110. doi:10.1016/j.materresbull.2007.11. 016 300 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

285. Kartsonakis IA, Liatsi P, Danilidis I et al (2008) Synthesis, characterization and antibacterial action of hollow titania spheres. J Phys Chem Sol 69:214–221. doi:10.1016/j.jpcs.2007.08. 071 286. Song X, Gao L (2007) Fabrication of hollow hybrid microspheres coated with silica/titania via sol–gel process and enhanced photocatalytic activities. J Phys Chem C 111:8180–8187. doi:10.1021/jp071142j 287. Deng Z, Chen M, Gu G, Wu L (2008) A facile method to fabricate ZnO hollow spheres and their photocatalytic property. J Phys Chem C 112:16–22. doi:10.1021/jp077662w 288. Syouﬁan A, Nakashima K (2008) Degradation of methylene blue in aqueous dispersion of hollow titania photocatalyst: study of reaction enhancement by various electron scavengers. J Colloid Interface Sci 317:507–512. doi:10.1016/j.jcis.2007.04.027 289. Cao FL, Wang JG, Lv FJ et al (2011) Photocatalytic oxidation of toluene to benzaldehyde over anatase TiO2 hollow spheres with exposed 001 facets. Catal Commun 12:946–950. doi:10.1016/j.catcom.2011.03.007 290. Yu J, Wang G (2008) Hydrothermal synthesis and photocatalytic activity of mesoporous titania hollow spheres. J Phys Chem Sol 69:1147–1151. doi:10.1016/j.jpcs.2007.09.024 291. Ao Y, Xu J, Fu D, Yuan C (2008) Photoelectrochemical application of hollow titania ﬁlm. Electrochem Commun 10:1812–1814. doi:10.1016/j.elecom.2008.09.015 292. Ao Y, Xu J, Fu D, Yuan D (2009) Visible-light responsive C, N-codoped titania hollow spheres for X-3B dye photodegradation. Micropor Mesopor Mater 118:382–386. doi:10. 1016/j.micromeso.2008.09.010 293. Lee SC, Lee CW, Lee SC, Lee JS (2008) Optical and photocatalytic properties of TiO2 with hollow nanostructure. Mater Lett 62:564–566. doi:10.1016/j.matlet.2007.09.008 294. Yang L, Zhu L, Liu C et al (2008) Synthesis and photocatalytic property of porous TiO2 microspheres. Mater Res Bull 43:806–810. doi:10.1016/j.materresbull.2007.06.044 295. Li H, Bian Z, Zhu J et al (2007) Mesoporous titania spheres with tunable chamber stucture and enhanced photocatalytic activity. J Am Chem Soc 129:8406–8407. doi:10.1021/ ja072191c 296. Cao SW, Zhu YJ (2008) Hierarchically nanostructured a-Fe2O3 hollow spheres: preparation, growth mechanism, photocatalytic property, and application in water treatment. J Phys Chem C 112:6253–6257. doi:10.1021/jp8000465 297. Cao SW, Zhu YJ, Cheng GF, Huang YH (2010) Preparation and photocatalytic property of a-Fe2O3 hollow core/shell hierarchical nanostructures. J Phys Chem Sol 71:1680–1683. doi:10.1016/j.jpcs.2010.09.006 298. Yu J, Liu S, Zhou M (2008) Enhanced photocatalytic activity of hollow anatase microspheres by Sn4+ incorporation. J Phys Chem C 112:2050–2057. doi:10.1021/ jp0770007 299. Grimes CA (2007) Synthesis and application of hyghly ordered arrays of TiO2 nanotubes. J Mater Chem 17:1451–1457. doi:10.1039/B701168G 300. Bavykin DV, Friedrich JM, Walsh FC (2006) Protonated titanates and TiO2 nanostructured materials: synthesis, properties, and applications. Adv Mater 18:2807–2824. doi:10.1002/ adma.200502696 301. Bavykin DV, Walsch FC (2010) Titanate and titania nanotubes: synthesis, properties and applications. RSC Nanoscience and Nanotechnology, N 12. R.S.C., London 302. Zhuang HF, Lin CJ, Lai YK et al (2007) Some critical structure factors of titanium dioxide nanotube array in its photocatalytic activity. Environ Sci Technol 41:4735–4740. doi:10. 1021/es0702723 303. Zhang G, Huang H, Zhang Y et al (2007) Highly ordered nanoporous TiO2 and its photo– catalytic properties. Electrochem Commun 9:2854–2858. doi:10.1016/j.elecom.2007.10.014 304. Zhang Z, Yuan Y, Shi G et al (2007) Photoelectrocatalytic activity of highly ordered TiO2 nanotube arrays electrode for azo dye degradation. Environ Sci Technol 41:6259–6263. doi:10.1021/es070212x References 301

305. Bae S, Shim E, Yoon J, Joo H (2008) Enzymatic hydrogen production by light-sensitized anodized tubular TiO2 photoanode. Sol En Mater Sol Cells 92:402–409. doi:10.1016/j. solmat.2007.09.019 306. Wu X, Ling Y, Sun J et al (2008) Photoelectrocatalytic degradation of methylene blue dye on titania nanotube array film. Chem Lett 37:416–417. doi:10.1246/cl.2008.416 307. Quan X, Ruan X, Zhao H et al (2007) Photoelectrocatalytic degraation of pentachlorophenol in aqueous solution using a TiO2 nanotube film electrode. Environ Pollution 147:409–414. doi:10.1016/j.envpol.2006.05.023 308. Bae S, Shim E, Yoon J, Joo H (2008) Photoanodic and cathodic role of anodized tubular titania in light-sensitized enzymatic hydrogen production. J Power Sources 185:439–444. doi:10.1016/j.jpowsour.2008.06.094 309. Allam NK, Shankar K, Grimes CA (2008) Photoelectrochemical and water photoelectrolysis properties of ordered TiO2 nanotubes fabricated by Ti anodization in fluoride-free HCl electrolytes. J Mater Chem 18:2341–2348. doi:10.1039/B718580D 310. Paulose M, Mor GK, Varghese OK et al (2006) Visible light photoelectrochemical and water-photoelectrolysis properties of titania nanotube arrays. J Photochem Photobiol, A 178:8–15. doi:10.1016/j.jphotochem.2005.06.013 311. Zlamal M, Macak JM, Schmuki P, Krysa J (2007) Electrochemically assisted photocatalysis on self-organized TiO2 nanotubes. Electrochem Commun 9:2822–2826. doi:10.1016/j. elecom.2007.10.002 312. Mohapatra SK, Raja KS, Mahajan VK, Misra M (2008) Efficient photoelectrolysis of water using TiO2 nanotube arrays by minimizing recombination losses with organic additives. J Phys Chem C 112:11007–11012. doi:10.1021/jp7100539 313. Sohn YS, Smith YR, Misra M, Subramanian VR (2008) Electrochemically assisted photocatalytic degradation of methyl orange using anodized titanium dioxide nanotubes. Appl Catal B 84:372–378. doi:10.1016/j.apcatb.2008.04.021 314. In S, Nielsen MG, Vesborg PCK et al (2011) Photocatalytic methane decomposition over vertically aligned transparent TiO2 nanotube arrays. Chem Commun 47:2613–2615. doi:10. 1039/C0CC02570D 315. He X, Cai Y, Zhang H, Liang C (2011) Photocatalytic degradation of organic pollutants with Ag decorated free-standing TiO2 nanotube arrays and interface electrochemical response. J Mater Chem 21:475–480. doi:10.1039/C0JM02404J 316. Zhang H, Liu P, Liu X (2010) Fabrication of highly ordered TiO2 nanorod/nanotube adjacent arrays for photoelectrochemical applications. Langmuir 26:11226–11232. doi:10. 1021/la1005314 317. Paek MJ, Kim TW, Hwang SJ (2008) Effects of hydronium intercalation and cation substitution on the photocatalytic performance of layered titanium oxide. J Phys Chem Sol 69:1444–1446. doi:10.1016/j.jpcs.2007.09.012 318. Ohtani B, Ikeda S, Nakayama H, Nishimoto S (2000) Shape- and size-selective photocatalytic reactions by layered titanic acid powder suspended in deaerated aqueous alcohol solutions. Phys Chem Chem Phys 2:5308–5313. doi:10.1039/B006497L 319. Choy JH, Lee HC, Jung H, Hwang SJ (2011) A novel synthetic route to TiO2-pillared titanate with enhanced photocatalytic activity. J Mater Chem 11:2232–2234. doi:10.1039/ B104551M 320. Sakai N, Ebina Y, Takada K, Sasaki T (2004) Electronic band structure of titania semiconductor nanosheets revealed by electrochemical and photoelectrochemical studies. J Am Chem Soc 126:5851–5858. doi:10.1021/ja0394582 321. Compton OC, Carroll EC, Kim JY et al (2007) Calcium niobate semiconductor nanosheets as catalysts for photochemical hydrogen evolution from water. J Phys Chem C 111:14589– 14592. doi:10.1021/jp0751155 322. Compton OC, Mullet CH, Chiang S, Osterloh FE (2008) A building block approach to photochemical water-splitting catalysts based on layered niobate nanosheets. J Phys Chem C 112:6202–6208. doi:10.1021/jp711589z 302 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

323. Zhou H, Sabio EM, Townsend TK et al (2010) Assembly of core-shell structures for photocatalytic hydrogen evolution from aqueous methanol. Chem Mater 22:3362–3368. doi:10.1021/cm903839t 324. Kuwahara Y, Yamashita H (2011) Efficient photocatalytic degradation of organics diluted in water and air using TiO2 designed with zeolites and mesoporous silica materials. J Mater Chem 21:2407–2416. doi:10.1039/C0JM02741C 325. Yang X, Zhu H, Liu J et al (2008) A mesoporous structure for efficient photocatalysts: anatase nanocrystals attached to leached clay layers. Micropor Mesopor Mater 112:32–44. doi:10.1016/j.micromeso.2007.09.017 326. Nikazar M, Golivand M, Mahanpur K (2007) Clinoptylolite-supported titania as a photoatalyst of disperse yellow 23 decomposition in water. Kin Catal 48:230–236 327. Miao S, Liu Z, Han B et al (2006) Synthesis and characterization of TiO2–montmorillonite nanocomposites and their application for removal of methylene blue. J Mater Chem 16:579– 584. doi:10.1039/B511426H 328. Menesi J, Körösi L, Bazso E et al (2008) Photocatalytic oxidation of organic pollutants on titania–clay composites. Chemosphere 70:538–542. doi:10.1016/j.chemosphere.2007.06. 049 329. Liu S, Yang JH, Choi JH (2006) Microporous SiO2–TiO2 nanosols pillared montmorillonite for photocatalytic decomposition of methyl orange. J Photochem Photobiol, A 179:75–80. doi:10.1016/j.jphotochem.2005.07.016 330. Zhang G, Ding X, Hu Y et al (2008) Photocatalytic degradation of 4BS dye by N, S-codoped TiO2 pillared montmorillonite photocatalysts under visible-light irradiation. J Phys Chem C 112:17994–17997. doi:10.1021/jp803939z 331. Chen J, Liu X, Li G et al (2011) Synthesis and characterization of novel SiO2 and TiO2 co-pil–lared montmorillonite composite for adsorption and photocatalytic degradation of hydrophobic organic pollutants in water. Catal Today 164:364–369. doi:10.1016/j.cattod. 2010.11.014 332. Salama TM, Ali IO, Mohamed MM (2007) Synthesis and characterization of mordenites encapsulated titania nanoparticles: photocatalytic degradation of meta-chlorophenol. J Mol Catal A 273:198–210. doi:10.1016/j.molcata.2007.03.071 333. Sun Z, Chen Y, Ke Q et al (2002) Photocatalytic degradation of cationic azo dye by TiO2/ bentonite nanocomposite. J Photochem Photobiol, A 149:169–174. doi:10.1016/S1010-6030 (01)00649-9 334. Zhang GK, Ding XM, He FS et al (2008) Low-temperature synthesis and photocatalytic activity of TiO2 pillared montmorillonite. Langmuir 24:1026–1030. doi:10.1021/la702649v 335. Zhao D, Zhou J, Liu N (2007) Surface characteristics and photoactivity of silver-modified palygorskite clays coated with nanosized titanium dioxide nanoparticles. Mater Character 58:249–255. doi:10.1016/j.matchar.2006.04.024 336. Koči K, Matejka V, Kovař P et al (2011) Comparison of the pure TiO2 and kaoli-nite/TiO2 composite as catalyst for CO2 photocatalytic reduction. Catal Today 161:105–109. doi:10. 1016/j.cattod.2010.08.026 337. Wang WY, Irawan A, Ku Y (2008) Photocatalytic degradation of Acid Red 4 using a titanium dioxide membrane supported on a porous ceramic tube. Water Res 42:4725–4732. doi:10.1016/j.watres.2008.08.021 338. Dubey N, Labhsetwar NK, Devotta S, Rayalu SS (2007) Hydrogen evolution by water splitting using novel composite zeolite-based photocatalyst. Catal Today 129:428–434. doi:10.1016/j.cattod.2006.09.041 339. Zhang J, Hu Y, Matsuoka M et al (2001) Relationship between the local structures of titanium oxide photocatalysts and their reactivities in the decomposition of NO. J Phys Chem B 105:8395–8398. doi:10.1021/jp012080e 340. Cosa G, Galletero MS, Fernandez L et al (2002) Tuning the photocatalytic activity of titanium dioxide by encapsulation inside zeolites exemplified by the cases of thianthrene photooxygenation and horseradish peroxidase photodeactivation. New J Chem 26:1448– 1455. doi:10.1039/B201397E References 303

341. Zheng S, Gao L, Zhang Q, Guo J (2000) Synthesis, characterization and photocatalytic properties of titania-modifed mesoporous silicate MCM-41. J Mater Chem 10:723–727. doi:10.1039/A908799K 342. Shankar H, Rajasudha G, Karthikeyan A et al (2008) Synthesis, characterization and photocatalytic activity of nanotitania loaded W-MCM-41. Nanotechnology 19:315711 343. Bouazza N, Lillo-Rodenas MA, Linares-Solano A (2008) Photocatalytic activity of TiO2- based materials for the oxidation of propene and benzene at low concentration in presence of humidity. Appl Catal B 84:691–698. doi:10.1016/j.apcatb.2008.06.002 344. Zheng S, Gao L (2002) Synthesis and characterization of Pt, Au or Pd clusters deposited titania-modified mesoporous silicate MCM-41. Mater Chem Phys 78:512–517. doi:10.1016/ S0254-0584(02)00353-X 345. Reddy EP, Sun B, Smirniotis PG (2004) Transition metal modified TiO2-loaded MCM-41 catalysts for visible- and UV-light driven photodegradation of aqueous organic pollutants. J Phys Chem B 108:17198–17205. doi:10.1021/jp047419m 346. De Witte K, Busuioc AM, Meyen V et al (2008) Influence of the synthesis parameters of TiO2–SBA-15 materials on the adsorption and photodegradation of rhodamine-6G. Micropor Mesopor Mater 110:100–110. doi:10.1016/j.micromeso.2007.09.035 347. Phanikrishna Sharma MV, Durga Kumari V, Subrahmanyam M (2008) TiO2 supported over SBA-15: an efficient photocatalyst for the pesticide degradation using solar light. Chemosphere 73:1562–1569. doi:10.1016/j.chemosphere.2008.07.081 348. Panpa W, Sujaridworakun P, Jinawath S (2008) Photocatalytic activity of TiO2/ZSM-5 2− composites in the presence of SO4 ion. Appl Catal B 80:271–276. doi:10.1016/j.apcatb. 2007.11.029 349. Neren Ökte A, Yilmaz Ö (2008) Photodecolorization of methyl orange by yttrium incorporated TiO2 supported ZSM-5. Appl Catal B 85:92–102. doi:10.1016/j.apcatb.2008. 07.025 350. Kitano M, Matsuoka M, Ueshima M, Anpo M (2007) Recent developments in titanium oxide-based photocatalysts. Appl Catal A 325:1–14. doi:10.1016/j.apcata.2007.03.013 351. Shen S, Guo L (2006) Structural, textural and photocatalytic properties of quantum-sized In2S3-sensitized Ti-MCM-41 prepared by ion-exchange and sulfidation methods. J Sol State Chem 179:2629–2635. doi:10.1016/j.jssc.2006.05.010 352. Shen S, Guo L (2008) Growth of quantum-confined CdS nanoparticles inside Ti-MCM-41 as a visible light photocatalyst. Mater Res Bull 43:437–446. doi:10.1016/j.materresbull.2007. 02.034 353. Sathish M, Viswanathan B, Viswanath RP (2006) Alternate synthetic strategy for the preparation of CdS nanoparticles and its exploitation for water splitting. Inter J Hydrogen En 31:891–898. doi:10.1016/j.ijhydene.2005.08.002 354. Ryu SY, Choi J, Balcerski W et al (2007) Photocatalytic production of H2 on nanocomposite Photocatalysts. Ind Eng Chem Res 46:7476–7488. doi:10.1021/ie0703033 355. Ryu SY, Balcerski W, Lee TK, Hoffmann MR (2007) Photocatalytic production of hydrogen from water with visible light using hybrid catalysts of CdS attached to microporous and mesoporous silicas. J Phys Chem C 111:18195–18203. doi:10.1021/jp074860e 356. Hirai T, Nanba M, Komasawa I (2003) Dithiol-mediated incorporation of CdS nanoparticles from reverse micellar system into Zn-doped SBA-15 mesoporous silica and their photocatalytic properties. J Colloid Interface Sci 268:394–399. doi:10.1016/j.jcis.2003.09. 011 357. Wang S, Liu P, Wang X, Fu X (2005) Homogeneously distributed CdS nanoparticles in Nafion membranes: preparation, characterization, and photocatalytic properties. Langmuir 21:11969–11973. doi:10.1021/la051072c 358. Pathak P, Meziani MJ, Li Y et al (2004) Improving photoreduction of CO2 with homogene– ously dispersed nanoscale TiO2 catalysts. Chem Commun 1234–1235. doi:10.1039/ B400326H 359. Pathak P, Meziani MJ, Castillo L, Sun YP (2005) Metal-coated nanoscale TiO2 catalysts for enhanced CO2 photoreduction. Green Chem 7:667–670. doi:10.1039/B507103H 304 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

360. Lunawat PS, Senapati S, Kumar R, Gupta NM (2007) Visible light-induced splitting of water using CdS nanocrystallites immobilized overwater-repellent polymeric surface. Inter J Hydrogen En 32:2784–2790. doi:10.1016/j.ijhydene.2007.04.001 361. Roy AM, De GC (2003) Immobilisation of CdS, ZnS and mixed ZnS–CdS on filter paper. Effect of hydrogen production from alkaline Na2S/Na2S2O3 solution. J Photochem Photobiol, A 157:87–92. doi:10.1016/S1010-6030(02)00430-6 362. Yi X, Liqin D, Lizhen A et al (2008) Photocatalytic degradation of rhodamine B and phenol by TiO2 loaded on mesoporous graphitic carbon. Chin J Catal 29:31–36. doi:10.1016/ S1872-2067(08)60014-5 363. Zhang X, Lu X, Shen Y et al (2011) Three-dimentional WO3 nanostructures on carbon paper: photoelectrochemical property and visible light driven photocatalysis. Chem Commun 47:5804–5806. doi:10.1039/C1CC10389J 364. Zhang X, Du AJ, Lee P et al (2008) Grafted multifunctional titanium dioxide nanotube membrane: Separation and photodegradation of aquatic pollutant. Appl Catal B 84:262–267. doi:10.1016/j.apcatb.2008.04.009 365. Plesch G, Gorbar M, Vogt UF et al (2009) Reticulated macroporous ceramic foam supported TiO2 for photocatalytic applications. Mater Lett 63:461–463. doi:10.1016/j.matlet.2008.11. 008 366. Skorb EV, Ustinovich EA, Kulak AI, Sviridov DV (2008) Photocatalytic activity of TiO2: In2O3 nanocomposite films towards the degradation of arylmethane and azo dyes. J Photochem Photobiol, A 193:97–102. doi:10.1016/j.jphotochem.2007.06.012 367. Fujishima A, Rao TN, Tryk DA (2000) Titanium dioxide photocatalysis. J Photochem Photobiol, C 1:1–21. doi:10.1016/S1389-5567(00)00002-2 368. Linsebigler AL, Lu G, Yates JT (1995) Photocatalysis on TiO2 Surfaces: principles, mechanisms, and selected results. Chem Rev 95:735–758. doi:10.1021/cr00035a013 369. Nowotny MK, Sheppard LR, Bak T, Nowotny J (2008) Defect chemistry of titanium dioxide. Application of defect engineering in processing of TiO2-based photocatalysts. J Phys Chem C 112:5275–5300. doi:10.1021/jp077275m 370. Anpo M (2004) Preparation, characterization, and reactivities of highly functional titanium oxide-based photocatalysts able to operate under UV–visible light irradiation: approaches in realizing high efficiency in the use of visible light. Bull Chem Soc Jpn 77:1427–1442. doi:10.1246/bcsj.77.1427 371. Liu G, Wang L, Yang HG et al (2010) Titania-based photocatalysts—crystal growth, doping and heterostructuring. J Mater Chem 20:831–843. doi:10.1039/B909930A 372. Vinu R, Madras G (2008) Synthesis and photoactivity of Pd substituted nano-TiO2. J Mol Catal A 291:5–11. doi:10.1016/j.molcata.2008.07.005 373. Xiao Q, Si Z, Yu Z, Qiu G (2008) Characterization and photocatalytic activity of Sm3+- doped TiO2 nanocrystalline prepared by low temperature combustion method. J Alloys Comp 450:426–431. doi:10.1016/j.jallcom.2006.10.137 374. Zhou J, Takeuchi M, Zhao XS et al (2006) Photocatalytic decomposition of formic acid under visible light irradiation over V-ion-implanted TiO2 thin film photocatalysts prepared on quartz substrate by ionized cluster beam (ICB) deposition method. Catal Lett 106:67–70. doi:10.1007/s10562-005-9192-5 375. Zhou J, Takeuchi M, Ray AK et al (2007) Enhancement of photocatalytic activity of P25 TiO2 by vanadium-ion implantation under visible light irradiation. J Colloid Interface Sci 311:497–501. doi:10.1016/j.jcis.2007.03.007 376. Takeuchi M, Yamashita H, Matsuoka M et al (2000) Photocatalytic decomposition of NO under visible light irradiation on the Cr-ion-implanted TiO2 thin film photocatalyst. Catal Lett 67:135–137. doi:10.1023/A:1019065521567 377. Ventkatachalam N, Palanichamy M, Arabindoo B, Murugesan V (2007) Alkaline earth metal doped nanoporous TiO2 for enhanced photocatalytic mineralisation of bisphenol-A. Catal Commun 8:1088–1093. doi:10.1016/j.catcom.2006.10.025 378. Venkatachalam N, Palanichamy M, Murugesan V (2007) Sol–gel preparation and characterization of alkaline earth metal doped nano TiO2:Efficient photocatalytic References 305

degradation of 4-chlorophenol. J Mol Catal A 273:177–185. doi:10.1016/j.molcata.2007.03. 077 379. Kryukov AI, Kuchmiy SY, Pokhodenko VD (1997) Nanostructured composite photocatalysts based on polycrystalline cadmium sulfide. Theoret Exp Chem 33:306–321 380. Yin S, Aita Y, Komatsu M, Sato T (2006) Visible-light-induced photocatalytic activity of TiO2−xNy prepared by solvothermal process in urea–alcohol system. J Eur Ceram Soc 26:2735–2742. doi:10.1016/j.jeurceramsoc.2005.05.012 381. Gandhe AR, Fernandes JB (2005) A simple method to synthesize N-doped rutile titania with enhanced photocatalytic activity in sunlight. J Sol State Chem 178:2953–2957. doi:10.1016/ j.jssc.2005.06.034 382. Xu JH, Dai WL, Li J et al (2008) Simple fabrication of thermally stable apertured N-doped TiO2 microtubes as a highly efficient photocatalyst under visible light irradiation. Catal Commun 9:146–152. doi:10.1016/j.catcom.2007.05.043 383. Matsumoto T, Iyi N, Kaneko Y et al (2007) High visible-light photocatalytic activity of nitrogen-doped titania prepared from layered titania/isostearate nanocomposite. Catal Today 120:226–232. doi:10.1016/j.cattod.2006.07.047 384. Li Y, Xie C, Peng S et al (2008) Eosin Y-sensitized nitrogen-doped TiO2 for efficient visible light photocatalytic hydrogen evolution. J Mol Catal A 282:117–123. doi:10.1016/j.molcata. 2007.12.005 385. Gao B, Ma Y, Cao Y et al (2006) Great enhancement of photocatalytic activity of nitrogen-doped titania by coupling with tungsten oxide. J Phys Chem B 110:14391–14397. doi:10.1021/jp0624606 386. Li H, Li J, Huo Y (2006) Highly active TiO2N photocatalysts prepared by treating TiO2 precursors in NH3/ethanol fluid under supercritical conditions. J Phys Chem B 110:1559– 1565. doi:10.1021/jp055830j 387. Rhee CH, Bae SW, Lee JS (2005) Template-free hydrothermal synthesis of high surface area nitrogen-doped titania photocatalyst active under visible light. Chem Lett 34:660–661. doi:10.1246/cl.2005.660 388. Livraghi S, Chierotti MR, Giamello E et al (2008) Nitrogen-doped titanium dioxide active in photocatalytic reactions with visible light: a multi-technique characterization of differently prepared materials. J Phys Chem C 112:17244–17252. doi:10.1021/jp803806s 389. Sun H, Bai Y, Liu H et al (2009) Photocatalytic decomposition of 4-chlorophenol over an efficient N-doped TiO2 under sunlight irradiation. J Photochem Photobiol, A 201:15–22. doi:10.1016/j.jphotochem.2008.08.021 390. Sun H, Bai Y, Jin W, Xu N (2008) Visible-light-driven TiO2 catalysts doped with low-con– centration nitrogen species. Sol En Mater Sol Cells 92:76–83. doi:10.1016/j.solmat.2007.09. 003 391. Li D, Huang H, Chen X et al (2007) New synthesis of excellent visible-light TiO2 − xNx photocatalyst using a very simple method. J Sol State Chem 180:2630–2634. doi:10.1016/j. jssc.2007.07.009 392. Cheng P, Deng C, Gu M, Dai X (2008) Effect of urea on the photoactivity of titania powder prepared by sol–gel method. Mater Chem Phys 107:77–81. doi:10.1016/j.matchemphys. 2007.06.051 393. Yuan J, Chen M, Shi J, Shanguang W (2006) Preparations and photocatalytic hydrogen evolution of N-doped TiO2 from urea and titanium tetrachloride. Inter J Hydrogen En 31:1326–1331. doi:10.1016/j.ijhydene.2005.11.016 394. Bacsa R, Kiwi J, Ohno T et al (2005) Preparation, testing and characterization of doped TiO2 active in the peroxidation of biomolecules under visible light. J Phys Chem B 109:5994– 6003. doi:10.1021/jp044979c 395. Sreethawong T, Laehsalee S, Chavadej S (2008) Comparative investigation of mesoporous- and non-mesoporous-assembled TiO2 nanocrystals for photocatalytic H2 production over N-doped TiO2 under visible light irradiation. Inter J Hydrogen En 33:5947–5957. doi:10. 1016/j.ijhydene.2008.08.007 306 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

396. Wang Y, Zhou G, Li T et al (2009) Catalytic activity of meso-porous TiO2 − xNx photocatalysts for the decomposition of methyl orange under solar simulated light. Catal Commun 10:412–415. doi:10.1016/j.catcom.2008.10.007 397. Chen D, Yang D, Geng J et al (2008) Improving visible-light photocatalytic activity of N-doped TiO2 nanoparticles via sensitization by Zn porphyrin. Appl Surf Sci 255:2879– 2884. doi:10.1016/j.apsusc.2008.08.032 398. Gole JL, Stout JD, Burda C et al (2004) Highly efficient formation of visible light tunable TiO2−xNx photocatalysts and their transformation at the nanoscale. J Phys Chem B 108:1230–1240. doi:10.1021/jp030843n 399. Jagadale TC, Takale SP, Sonawane RS et al (2008) N-doped TiO2 nanoparticle based visible light photocatalyst by modified peroxide sol–gel method. J Phys Chem C 112:14595–14602. doi:10.1021/jp803567f 400. Xu J, Ao Y, Fu D, Yuan C (2008) A simple route to synthesize highly crystalline N-doped TiO2 particles under low temperature. J Cryst Growth 310:4319–4324. doi:10.1016/j. jcrysgro.2008.07.045 401. Zhao Y, Qiu X, Burda C (2008) The Effects of sintering on the photocatalytic activity of N-doped TiO2 nanoparticles. Chem Mater 20:2629–2636. doi:10.1021/cm703043j 402. Lopez-Luke T, Wolcott A, Xu L et al (2008) Nitrogen-doped and CdSe quantum-dot-sensitized nanocrystalline TiO2 films for solar energy conversion applications. J Phys Chem C 112:1282–1292. doi:10.1021/jp077345p 403. Sano T, Negishi N, Koike K et al (2004) Preparation of a visible light-responsive photocatalyst from a complex of Ti4+ with a nitrogen-containing ligand. J Mater Chem 14:380–384. doi:10.1039/B311444A 404. Feng C, Wang Y, Jin Z et al (2008) Photoactive centers responsible for visible-light photoactivity of N-doped TiO2. New J Chem 32:1038–1047. doi:10.1039/B719498F 405. Ozaki H, Iwamoto S, Inoue M (2007) Marked promotive effect of iron on visible-light-induced photocatalytic activities of nitrogen- and silicon-codoped titanias. J Phys Chem C 111:17061–17066. doi:10.1021/jp0751211 406. Yang S, Gao L (2008) Photocatalytic activity of nitrogen doped rutile TiO2 nanoparticles under visible light irradiation. Mater Res Bull 43:1872–1876. doi:10.1016/j.materresbull. 2007.06.058 407. Morikawa T, Ohwaki T, Suzuki K et al (2008) Visible-light-induced photocatalytic oxidation of carboxylic acids and aldehydes over N-doped TiO2 loaded with Fe, Cu or Pt. Appl Catal B 83:56–62. doi:10.1016/j.apcatb.2008.01.034 408. Yin S, Liu B, Zhang P et al (2008) Photocatalytic oxidation of NOx under visible LED light irradiation over nitrogen-doped titania particles with iron or platinum loading. J Phys Chem C 112:12425–12431. doi:10.1021/jp803371s 409. Shouxin L, Xiaoyun C, Xi C (2006) Preparation of N-doped visible-light response nanosize TiO2 photocatalyst using the acid-catalyzed hydrolysis method. Chin J Catal 27:697–702. doi:10.1016/S1872-2067(06)60037-5 410. Balcerski W, Ryu SY, Hoffmann MR (2007) Visible-light photoactivity of nitrogen-doped TiO2: photo-oxidation of HCO2HtoCO2 and H2O. J Phys Chem C 111:15357–15362. doi:10.1021/jp074989o 411. Abe H, Kimitani T, Naito M (2006) Influence of NH3/Ar plasma irradiation on physical and photocatalytic properties of TiO2 nanopowder. J Photochem Photobiol, A 183:171–175. doi:10.1016/j.jphotochem.2006.03.013 412. Sano T, Puzenat E, Guillard C et al (2008) Degradation of C2H2 with modified-TiO2 photocatalysts under visible light irradiation. J Mol Catal A 284:127–133. doi:10.1016/j. molcata.2008.01.014 413. Yin S, Yamaki H, Komatsu M et al (2003) Preparation of nitrogen-doped titania with high visible light induced photocatalytic activity by mechanochemical reaction of titania and hexamethylenetetramine. J Mater Chem 13:2996–3001. doi:10.1039/B309217H References 307

414. Shen H, Mi L, Xu P et al (2007) Visible-light photocatalysis of nitrogen-doped TiO2 nanoparticulate films prepared by low-energy ion implantation. Appl Surf Sci 253:7024– 7028. doi:10.1016/j.apsusc.2007.02.023 415. Cui X, Ma M, Zhang W et al (2008) Nitrogen-doped TiO2 from TiN and its visible light photoelectrochemical properties. Electrochem Commun 10:367–371. doi:10.1016/j.elecom. 2007.12.037 416. Liu B, Wen L, Zhao X (2008) The structure and photocatalytic studies of N-doped TiO2 films prepared by radio frequency reactive magnetron sputtering. Sol En Mater Sol Cells 92:1–10. doi:10.1016/j.solmat.2007.07.009 417. Lin HF, Liao SC, Hung SW (2005) The dc thermal plasma synthesis of ZnO nanoparticles for visible-light photocatalyst. J Photochem Photobiol, A 174:82–87. doi:10.1016/j. jphotochem.2005.02.015 418. Wang H, Zhang Z, Chang W (2008) A novel one-step photocatalytic synthesis of benzo[d] oxazol-2(3H)-one with C-doped TiO2 nanoparticle. Chem Lett 37:1156–1157. doi:10.1246/ cl.2008.1156 419. Huang Y, Ho W, Lee S et al (2008) Effect of carbon doping on the mesoporous structure of nanocrystalline titanium dioxide and its solar-light-driven photocatalytic degradation of NOx. Langmuir 24:3510–3516. doi:10.1021/la703333z 420. Treschev SY, Chou PW, Tseng YH et al (2008) Photoactivities of the visible-light-activated mixed-phase carbon-containing titanium dioxide: the effect of carbon incorporation. Appl Catal B 79:8–16. doi:10.1016/j.apcatb.2007.09.046 421. Wang X, Meng S, Zhang X et al (2007) Multi-type carbon doping of TiO2 photocatalyst. Chem Phys Lett 444:292–296. doi:10.1016/j.cplett.2007.07.026 422. Mohapatra SK, Misra M, Mahajan VK, Raja KS (2007) Design of a highly efficient photoelectrolytic cell for hydrogen generation by water splitting: application of TiO2 − xCx nanotubes as a photoanode and Pt/TiO2 nanotubes as a cathode. J Phys Chem C 111:8677– 8685. doi:10.1021/jp071906v 423. Hu X, Zhang T, Jin Z et al (2008) Fabrication of carbon-modified TiO2 nanotube arrays and their photocatalytic activity. Mater Lett 62:4579–4581. doi:10.1016/j.matlet.2008.08.051 424. Xiao Q, Zhang J, Xiao C et al (2008) Solar photocatalytic degradation of methylene blue in carbon-doped TiO2 nanoparticles suspension. Sol En 82:706–713. doi:10.1016/j.solener. 2008.02.006 425. Randeniya LK, Murphy AB, Plumb IC (2008) A study of S-doped TiO2 for photoelectrochemical hydrogen generation from water. J Mater Sci 43:1389–1399. doi:10.1007/ s10853-007-2309-z 426. Xu J, Ao Y, Fu D, Yuan C (2008) Low-temperature preparation of F-doped TiO2 film and its photocatalytic activity under solar light. Appl Surf Sci 254:3033–3038. doi:10.1016/j. apsusc.2007.10.065 427. Xu J, Ao Y, Fu D, Yuan C (2008) Synthesis of fluorine-doped titania-coated activated carbon under low temperature with high photocatalytic activity under visible light. J Phys Chem Sol 69:2366–2370. doi:10.1016/j.jpcs.2008.03.017 428. Zhou JK, Lv L, Yu J et al (2008) Synthesis of self-organized polycrystalline F-doped TiO2 hollow microspheres and their photocatalytic activity under visible light. J Phys Chem C 112:5316–5321. doi:10.1021/jp709615x 429. Wu G, Chen A (2008) Direct growth of F-doped TiO2 particulate thin films with high photocatalytic activity for environmental applications. J Photochem Photobiol, A 195:47–53. doi:10.1016/j.jphotochem.2007.09.005 430. Wu G, Wang J, Thomas DF, Chen A (2008) Synthesis of F-doped flower-like TiO2 nanostructures with high photoelectrochemical activity. Langmuir 24:3503–3509. doi:10. 1021/la703098g 431. Khan R, Kim SW, Kim TJ, Nam CM (2008) Comparative study of the photocatalytic performance of boron-iron Co-doped and boron-doped TiO2 nanoparticles. Mater Chem Phys 112:167–172. doi:10.1016/j.matchemphys.2008.05.030 308 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

432. Yu HF (2007) Photocatalytic abilities of gel-derived P-doped TiO2. J Phys Chem Sol 68:600–607. doi:10.1016/j.jpcs.2007.01.050 433. Lin L, Lin W, Zhu Y et al (2005) Phosphor-doped titania—a novel photocatalyst active in visible light. Chem Lett 34:284–285. doi:10.1246/cl.2005.284 434. Zaleska A, Sobczak JW, Grabowska E, Hupka J (2008) Preparation and photocatalytic activity of boron-modified TiO2 under UV and visible light. Appl Catal B 78:92–100. doi:10.1016/j.apcatb.2007.09.005 435. Li J, Lu N, Quan X et al (2008) Facile method for fabricating boron-doped TiO2 nanotube array with enhanced photoelectrocatalytic properties. Ind Eng Chem Res 47:3804–3808. doi:10.1021/ie0712028 436. In S, Orlov A, Berg R et al (2007) Effective visible light-activated B-doped and B, N-codoped TiO2 photocatalysts. J Am Chem Soc 129:13790–13791. doi:10.1021/ja0749237 437. Tristao JC, Magalhaes F, Corio P, Sansiviero MTC (2006) Electronic characterization and photocatalytic properties of CdS/TiO2 semiconductor composite. J Photochem Photobiol, A 181:152–157. doi:10.1016/j.jphotochem.2005.11.018 438. Demeestere K, Dewulf J, Ohno T et al (2005) Visible light mediated photocatalytic degradation of gaseous trichloroethylene and dimethyl sulfide on modified titanium dioxide. Appl Catal B 61:140–149. doi:10.1016/j.apcatb.2005.04.017 439. Yin H, Wada Y, Kitamura T et al (2001) Enhanced photocatalytic dechlorination of 1,2,3,4-tetrachlorobenzene using nanosized CdS/TiO2 hybrid photocatalyst under visible light irradiation. Chem Lett 30:334–335. doi:10.1246/cl.2001.334 440. Blackburn JL, Selmarten DC, Nozik AJ (2003) Electron transfer dynamics in quantum dot/titanium dioxide composites formed by in situ chemical bath deposition. J Phys Chem B 107:14154–14157. doi:10.1021/jp0366771 441. Vogel R, Hoyer P, Weller H (1994) Quantum-sized PbS, CdS, Ag2S, Sb2S3, and Bi2S3 particles as sensitizers for various nanoporous wide-bandgap semiconductors. J Phys Chem 98:3183–3188. doi:10.1021/j100063a022 442. Hong JS, Choi DS, Kang MG et al (2001) Photocurrent instability of PbS-sensitized TiO2 2− 2− electrodes in S and SO3 solution. J Photochem Photobiol, A 143:87–92. doi:10.1016/ S1010-6030(01)00455-5 443. Sun WT, Yu T, Pan HY et al (2008) CdS quantum dots sensitized TiO2 nanotube-array photoelectrodes. J Am Chem Soc 130:1124–1125. doi:10.1021/ja0777741 444. Shalom M, Dor S, Rühle S et al (2009) Core/CdS quantum dot/shell mesoporous solar cells with improved stability and efficiency using an amorphous TiO2 coating. J Phys Chem C 113:3895–3898. doi:10.1021/jp8108682 445. Das K, De SK (2009) Optical properties of the type-II core-shell TiO2@CdS nanorods for photovoltaic applications. J Phys Chem C 113:3494–3501. doi:10.1021/jp8083953 446. Chen H, Fu W, Yang H et al (2010) Photosensitization of TiO2 nanorods with CdS quantum dots for photovoltaic devices. Electrochim Acta 56:919–924. doi:10.1016/j.electacta.2010. 10.003 447. Xia M, Wang F, Wang Y et al (2010) TiO2 nanowires sensitized with CdS quantum dots and the surface photovoltage properties. Mater Lett 64:1688–1690. doi:10.1016/j.matlet.2010. 05.003 448. Nayak J, Sahu SN, Kasuya J, Nozaki S (2008) CdS–ZnO composite nanorods: synthesis, characterization and application for photocatalytic degradation of 3,4-dihydroxy benzoic acid. Appl Surf Sci 254:7215–7218. doi:10.1016/j.apsusc.2008.05.268 449. Lee W, Min SK, Dhas V et al (2009) Chemical bath deposition of CdS quantum dots on vertically aligned ZnO nanorods for quantum dots-sensitized solar cells. Electrochem Commun 11:103–106. doi:10.1016/j.elecom.2008.10.042 450. Zhang YJ, Yan W, Wu YP, Wang ZH (2008) Synthesis of TiO2 nanotubes coupled with CdS nanoparticles and production of hydrogen by photocatalytic water decomposition. Mater Lett 62:3846–3848. doi:10.1016/j.matlet.2008.04.084 References 309

451. Xing C, Jing D, Liu M, Guo L (2009) Photocatalytic hydrogen production over Na2Ti2O4(OH)2 nanotube sensitized by CdS nanoparticles. Mater Res Bull 44:442–445. doi:10.1016/j.materresbull.2008.04.016 452. Peter LM, Upul Wijayantha KG, Jason Riley D, Waggett JP (2003) Band-edge tuning in self-assembled layers of Bi2S3 nanoparticles used to photosensitize nanocrystalline TiO2. J Phys Chem B 107:8378–8381. doi:10.1021/jp030334l 453. Smith YR, Subramanian VR (2011) Heterostructural composites of TiO2 mesh-TiO2 nanoparticles photosensitized with CdS: a new flexible photoanode for solar cells. J Phys Chem B 115:8376–8385. doi:10.1021/jp110200s 454. Lin ZQ, Lai YK, Hu RG et al (2010) A highly efficient ZnS/CdS@TiO2 photo-electrode for photogenerated cathodic protection of metals. Electrochim Acta 55:8717–8723. doi:10.1016/ j.electacta.2010.08.017 455. Gao C, Li J, Shan Z et al (2010) Preparation ans visible-light photocatalytic activity of In2S3/ TiO2 composite. Mater Chem Phys 122:183–187. doi:10.1016/j.matchemphys.2010.02.030 456. Patil RS, Lokhande CD, Mane RS et al (2007) Room temperature PbS nanoparticle growth, incubation in porous TiO2 electrode for photosensitization application. J Non-Cryst Sol 353:1645–1649. doi:10.1016/j.jnoncrysol.2007.01.014 457. Lee W, Lee J, Min SK et al (2009) Effect of single-walled carbon nanotube in PbS/TiO2 quantum dots-sensitized solar cells. Mater Sci Eng, B 156:48–51. doi:10.1016/j.mseb.2008. 11.014 458. Ratanatawanate C, Bui A, Vu K, Balkus KJ (2011) Low-temperature synthesis of copper (II) sulfide quantum dot decorated TiO2 nanotubes and their photocatalytic properties. J Phys Chem C 115:6175–6180. doi:10.1021/jp109716q 459. Liu Y, Zhou H, Zhou B et al (2011) Highly stable CdS-modified short TiO2 nanotube array electrode for efficient visible-light hydrogen generation. Inter J Hydrogen En 36:167–174. doi:10.1016/j.ijhydene.2010.09.089 460. Seoul M, Kim H, Kim W, Yong K (2010) Highly efficient photoelectrochemical hydrogen generation using a ZnO nanowire array and a CdSe/CdS co-sensitizer. Electrochem Commun 12:1416–1418. doi:10.1016/j.elecom.2010.07.035 461. Sasamura T, Okazaki K, Tsunoda R et al (2010) Immobilization of ZnS-AgInS2 solid solition nanoparticles on ZnO rod array electrodes and their photoresponse with visible light irradiation. Chem Lett 39:619–620. doi:10.1246/cl.2010.619 462. Liu C, Liu Z, Li Y et al (2011) CdS/PbS co-sensitized ZnO nanorods and its photocatalytic properties. Appl Surf Sci 257:7041–7046. doi:10.1016/j.apsusc.2011.02.133 463. Huang S, Zhang Q, Huang X et al (2010) Fibrous CdS/CdSe quantum dot co-sensitized solar cells based on ordered TiO2 nanotube arrays. Nanotechnology 21:375201 464. Anower Hossain M, Jennings JR, Koh ZY, Wang Q (2011) Carrier generation and collection in CdS/CdSe-sensitized SnO2 solar cells exhibiting unprecedented photocurrent densities. ACS Nano 5:3172–3181. doi:10.1021/nn200315b 465. Bessekhouad Y, Chaoui N, Trzpit M et al (2006) UV-vis versus visible degradation of Acid Red II in a coupled CdS/TiO2 semiconductors suspension. J Photochem Photobiol, A 183:218–224. doi:10.1016/j.jphotochem.2006.03.025 466. Jang JS, Kim HG, Joshi UA et al (2008) Fabrication of CdS nanowires decorated with TiO2 nanoparticles for photocatalytic hydrogen production under visible light irradiation. Inter J Hydrogen En 33:5975–5980. doi:10.1016/j.ijhydene.2008.07.105 467. Jang JS, Choi SH, Kim HG, Lee JS (2008) Location and state of Pt in platinized CdS/TiO2 photocatalysts for hydrogen production from water under visible light. J Phys Chem C 112:17200–17205. doi:10.1021/jp804699c 468. Lawless D, Kapoor S, Meisel D (1995) Bifunctional capping of CdS nanoparticles and bridging to TiO2. J Phys Chem 99:10329–10335. doi:10.1021/j100025a040 469. Dibbell RS, Watson DF (2009) Distance-dependent electron transfer in tethered assemblies of CdS quantum dots and TiO2 nanoparticles. J Phys Chem C 113:3139–3149. doi:10.1021/ jp809269m 310 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

470. Hirai T, Suzuki K, Komasawa I (2001) Preparation and photocatalytic properties of composite CdS nanoparticles—titanium dioxide particles. J Colloid Interface Sci 244:262– 265. doi:10.1006/jcis.2001.7982 471. Peter LM, Jason Riley D, Tull EJ, Upul Wijayantha KG (2002) Photosensitization of nanocrystalline TiO2 by self-assembled layers of CdS quantum dots. Chem Commun 1030– 1031. doi:10.1039/B201661C 472. Lee YL, Huang BM, Chien HT (2008) Highly efficient CdSe-sensitized TiO2 photoelectrode for quantum-dot-sensitized solar cell applications. Chem Mater 20:6903–6905. doi:10.1021/ cm802254u 473. Wu J, Lin J, Yin S, Sato T (2001) Synthesis and photocatalytic properties of layered HNbWO6/(Pt, Cd0.8Zn0.2S) nanocomposites. J Mater Chem 11:3343–3347. doi:10.1039/ B103838A 474. Shangguan W (2007) Hydrogen evolution from water splitting on nanocomposite photocatalysts. Sci Technol Adv Mater 8:76–81. doi:10.1016/j.stam.2006.09.007 475. Stengl V, Bakarjieva S, Murafa N et al (2008) Visible-light photocatalytic activity of TiO2/ ZnS nanocomposites prepared by homogeneous hydrolysis. Micropor Mesopor Mater 110:370–378. doi:10.1016/j.micromeso.2007.06.052 476. Kumar A, Jain AK (2003) Photophysics and photocatalytic properties of Ag+-activated sandwich Q-CdS–TiO2. J Photochem Photobiol, A 156:207–218. doi:10.1016/S1010-6030 (02)00396-9 477. Yu JC, Wu L, Lin J et al (2003) Microemulsion-mediated solvothermal synthesis of nanosized CdS-sensitized TiO2 crystalline photocatalyst. Chem Commun 1552–1553. doi:10.1039/B302418K 478. Evans JE, Springer KW, Zhang JZ (1994) Femtosecond studies of interparticle electron transfer in a coupled CdS-TiO2 colloidal system. J Chem Phys 101:6222–6225. doi:10.1063/ 1.468376 479. Gopidas KR, Bohorquez M, Kamat PV (1990) Photophysical and photochemical aspects of coupled semiconductors. Charge-transfer processes in colloidal CdS-TiO2 and CdS-AgI systems. J Phys Chem 94:6435–6440. doi:10.1021/j100379a051 480. Hotchandani S, Kamat PV (1992) Charge-transfer processes in coupled semiconductor systems. Photochemistry and photoelectrochemistry of the colloidal CdS-ZnO system. J Phys Chem 96:6834–6839. doi:10.1021/j100195a056 481. So WW, Kim KJ, Moon SJ (2004) Photoproduction of hydrogen over the CdS–TiO2 nano-composite particulate flms treated with TiCl4. Inter J Hydrogen En 29:229–234. doi:10. 1016/S0360-3199(03)00211-8 482. Chen S, Paulose M, Ruan C et al (2006) Electrochemically synthesized CdS nanoparticle-modified TiO2 nanotube-array photoelectrodes: Preparation, characterization, and application to photoelectrochemical cells. J Photochem Photobiol, A 177:177–184. doi:10.1016/j.jphotochem.2005.05.023 483. Seabold JA, Shankar K, Wilke RHT et al (2008) Photoelectrochemical properties of heterojunction CdTe/TiO2 electrodes constructed using highly ordered TiO2 nanotube arrays. Chem Mater 20:5266–5273. doi:10.1021/cm8010666 484. Lu ZX, Zhang ZL, Zhang MX et al (2005) Core/shell quantum-dot-photosensitized nano-TiO2 films: Fabrication and application to the damage of cells and DNA. J Phys Chem B 109:22663–22666. doi:10.1021/jp054472h 485. Diguna LJ, Shen Q, Sato A et al (2007) Optical absorption and ultrafast carrier dynamics characterization of CdSe quantum dots deposited on different morphologies of nanostructured TiO2 films. Mater Sci Eng, C 27:1514–1520. doi:10.1016/j.msec.2006.06.036 486. Shen Q, Katayama K, Yamaghuchi M et al (2005) Study of ultrafast carrier dynamics of nanostructured TiO2 films with and without CdSe quantum dot deposition using lens-free heterodyne detection transient grating technique. Thin Solid Films 486:15–19. doi:10.1016/j. tsf.2004.11.238 References 311

487. Shen Q, Sato T, Hashimoto M et al (2006) Photoacoustic and photoelectrochemical characterization of CdSe-sensitized TiO2 electrodes composed of nanotubes and nanowires. Thin Solid Films 499:299–305. doi:10.1016/j.tsf.2005.07.019 488. Ho W, Yu JC (2006) Sonochemical synthesis and visible light photocatalytic behavior of CdSe and CdSe/TiO2 nanoparticles. J Mol Catal A 247:268–274. doi:10.1016/j.molcata. 2005.11.057 489. Robel I, Kuno M, Kamat PV (2007) Size-dependent electron injection from excited CdSe quantum dots into TiO2 nanoparticles. J Am Chem Soc 129:4136–4137. doi:10.1021/ ja070099a 490. Shen Y, Bao J, Dai N et al (2009) Speedy photoelectric exchange of CdSe quantum dots/mesoporous titania composite system. Appl Surf Sci 255:3908–3911. doi:10.1016/j. apsusc.2008.10.004 491. Loef R, Houtepen AJ, Talgorn E et al (2009) Study of electronic defects in CdSe quantum dots and their involvement in quantum dot solar cells. Nano Lett 9:856–859. doi:10.1021/ nl803738q 492. Im SH, Lee YH, Seok SI et al (2010) Quantum-dot-sensitized solar cells fabricated by the combined process of the direct attachment of colloidal CdSe quantum dots having a ZnS glue layer and spray pyrolysis deposition. Langmuir 26:18576–18580. doi:10.1021/ la1034382 493. Salant A, Shalom M, Hod I et al (2010) Quantum dot sensitized solar cells with improved efficiency prepared using electrophoretic deposition. ACS Nano 4:5962–5968. doi:10.1021/ nn1018208 494. Guijarro S, Lana-Villarreal T, Shen Q et al (2010) Sensitization of titanium dioxide photoanodes with cadmium selenide quantum dots prepared by SILAR: photoelectrochemical and carrier dynamics study. J Phys Chem C 114:21928–21937. doi:10.1021/jp105890x 495. Fuke N, Hoch LB, Koposov AY et al (2010) CdSe quantum-dot-sensitized solar cell with *100% internal efficiency. ACS Nano 4:6377–6386. doi:10.1021/nn101319x 496. Hossain MF, Biswas S, Zhang ZH, Takahashi T (2011) Bubble-like CdSe nanoclusters sensitized TiO2 nanotube arrays for improvement in solar cell. J Photochem Photobiol, A 217:68–75. doi:10.1016/j.jphotochem.2010.09.020 497. Lin CJ, Chen S, Liou YH (2010) Wire-shaped electrode of CdSe-sensitized ZnO nanowire arrays for photoelectrochemical hydrogen generation. Electrochem Commun 12:1513–1516. doi:10.1016/j.elecom.2010.08.021 498. Wang H, Wu Z, Liu Y, Sheng Z (2008) The characterization of ZnO–anatase–rutile three-component semiconductor and enhanced photocatalytic activity of nitrogen oxides. J Mol Catal 287:176–181. doi:10.1016/j.molcata.2008.03.010 499. Wang N, Li J, Zhu L et al (2008) Highly photocatalytic activity of metallic hydroxide/titanium oxide nanoparticles prepared via a modified wet precipitation process. J Photochem Photobiol, A 198:282–287. doi:10.1016/j.jphotochem.2008.03.021 500. Wang N, Zhu LH, Li J, Tang HQ (2007) A novel Fe(OH)3/TiO2 nanoparticles and its high photocatalytic activity. Chin Chem Lett 18:1261–1264. doi:10.1016/j.cclet.2007.08.020 501. Yu J, Hai Y, Cheng B (2011) Enhanced photocatalytic H2-production activity of TiO2 by Ni (OH)2 cluster modification. J Phys Chem C 115:4953–4958. doi:10.1021/jp111562d 502. Mohamed MM, Othman I, Mohamed RM (2007) Synthesis and characterization of MnOx/ TiO2 nanoparticles for photocatalytic oxidation of indigo carmine dye. J Photochem Photobiol, A 191:153–161. doi:10.1016/j.jphotochem.2007.04.017 503. Lalitha K, Sadanandam G, Durga V et al (2010) Highly stabilized and finely dispersed Cu2O/TiO2: a promising visible sensitive photocatalyst for continuous production of hydrogen from glycerol:water mixtures. J Phys Chem C 114:22181–22189. doi:10.1021/ jp107405u 504. Xu S, Du AJ, Liu J et al (2011) Highly efficient CuO incorporated TiO2 nanotube photocatalyst for hydrogen production from water. Inter J Hydrogen En 36:6560–6568. doi:10.1016/j.ijhydene.2011.02.103 312 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

505. Liao DL, Badour CA, Liao BQ (2008) Preparation of nanosized TiO2/ZnO composite catalyst and its photocatalytic activity for degradation of methyl orange. J Photochem Photobiol, A 194:11–19. doi:10.1016/j.jphotochem.2007.07.008 506. Chen D, Zhang H, Hu S, Li J (2008) Preparation and enhanced photo-electrochemical performance of coupled bicomponent ZnO–TiO2 nanocomposites. J Phys Chem C 112:117– 122. doi:10.1021/jp077236a 507. Abdel Aal A, Barakat MA, Mohamed RM (2008) Electrophoreted Zn–TiO2–ZnO nanocomposite coating films for photocatalytic degradation of 2-chlorophenol. Appl Surf Sci 254:4577–4583. doi:10.1016/j.apsusc.2008.01.049 508. Song H, Jiang H, Liu X, Meng G (2006) Efficient degradation of organic pollutant with WOx modified nano TiO2 under visible irradiation. J Photochem Photobiol, A 181:421–428. doi:10.1016/j.jphotochem.2006.01.001 509. Ke D, Liu H, Peng T et al (2008) Preparation and photocatalytic activity of WO3/TiO2 nanocomposite particles. Mater Lett 62:447–450. doi:10.1016/j.matlet.2007.05.060 510. Li XZ, Li FB, Yang CL, Ge WK (2001) Photocatalytic activity of Wox–TiO2 under visible light irradiation. J Photochem Photobiol, A 141:209–217. doi:10.1016/S1010-6030(01) 00446-4 511. Yasomanee JP, Bandara J (2008) Multi-electron storage of photoenergy using Cu2O–TiO2 thin film photocatalyst. Sol En Mater Sol Cells 92:348–352. doi:10.1016/j.solmat.2007.09. 016 512. Celik E, Yildiz AY, Ak Azem NF et al (2006) Preparation and characterization of Fe2O3– TiO2 thin films on glass substrate for photocatalytic applications. Mater Sci Eng, B 129:193– 199. doi:10.1016/j.mseb.2006.01.013 513. Seftel EM, Popovici E, Mertens M et al (2008) SnIV-containing layered double hydroxides as precursors for nano-sized ZnO/SnO2 photocatalysts. Appl Catal B 84:699–705. doi:10. 1016/j.apcatb.2008.06.006 514. Chen LC, Tsai FR, Fang SH, Ho YC (2009) Properties of sol–gel SnO2/TiO2 electrodes and their photoelectrocatalytic activities under UV and visible light illumination. Electrochim Acta 54:1304–1311. doi:10.1016/j.electacta.2008.09.009 515. Ismail AA (2008) Single-step synthesis of a highly active photocatalyst for oxidation of trichloroethylene. Appl Catal B 85:33–39. doi:10.1016/j.apcatb.2008.06.025 516. Bedja I, Kamat PV (1995) Capped semiconductor colloids. Synthesis and photoelectrochemical behavior of TiO2-capped SnO2 nanocrystallites. J Phys Chem 99:9182–9188. doi:10.1021/j100022a035 517. Penpolcharoen M, Amal R, Brungs M (2001) Degradation of sucrose and nitrate over titania coated nano-hematite photocatalysts. J Nanopart Res 3:289–302. doi:10.1023/A: 1017929204380 518. Nishijima K, Ohtani B, Yan X et al (2007) Incident light dependence for photocatalytic degradation of acetaldehyde and acetic acid on S-doped and N-doped TiO2 photocatalysts. Chem Phys 339:64–72. doi:10.1016/j.chemphys.2007.06.014 519. Nishijima K, Fujisawa Y, Murakami N et al (2008) Development of an S-doped titania nanotube (TNT) site-selectively loaded with iron (III) oxide and its photocatalytic activities. Appl Catal B 84:584–590. doi:10.1016/j.apcatb.2008.05.014 520. Xiao G, Wang X, Li D, Fu X (2008) InVO4-sensitized TiO2 photocatalysts for efficient air purification with visible light. J Photochem Photobiol, A 193:213–221. doi:10.1016/j. jphotochem.2007.06.027 521. Lee BT, Han JK, Gain AK et al (2006) TEM microstructure characterization of nano TiO2 coated on nano ZrO2 powders and their photocatalytic activity. Mater Lett 60:2101–2104. doi:10.1016/j.matlet.2005.12.102 522. Xu J, Ao Y, Fu D, Yuan C (2008) Synthesis of Bi2O3–TiO2 composite film with high-photocatalytic activity under sunlight irradiation. Appl Surf Sci 255:2365–2369. doi:10.1016/j.apsusc.2008.07.095 References 313

523. Tawkaew S, Chareonpanich M, Supothina S (2008) Preparation and photocatalytic study of fibrous K0.3Ti4O7.3(OH)1.7—anatase TiO2 nanocomposite photocatalyst. Mater Chem Phys 111:232–237. doi:10.1016/j.matchemphys.2008.03.031 524. Jang JS, Kim HG, Reddy VR et al (2005) Photocatalytic water splitting over iron oxide nanoparticles intercalated in HTiNb(Ta)O5 layered compounds. J Catal 231:213–222. doi:10.1016/j.jcat.2005.01.026 525. Jothiramalingam R, Wang MK (2007) Synthesis, characterization and photocatalytic activity of porous manganese oxide doped titania for toluene decomposition. J Hazard Mater 147:562–569. doi:10.1016/j.jhazmat.2007.01.069 526. Cho J, Denes FS, Timmons RB (2006) Plasma processing approach to molecular surface tailoring of nanoparticles: improved photocatalytic activity of TiO2. Chem Mater 18:2989– 2996. doi:10.1021/cm060212g 527. Martinez AI, Acosta DR, Cedillo G et al (2005) Effect of SnO2 on the photocatalytical properties of TiO2 films. Thin Solid Films 490:118–123. doi:10.1016/j.tsf.2005.04.060 528. Li D, Haneda H (2003) Photocatalysis of sprayed nitrogen-containing Fe2O3–ZnO and WO3–ZnO composite powders in gas-phase acetaldehyde decomposition. J Photochem Photobiol, A 160:203–212. doi:10.1016/S1010-6030(03)00212-0 529. Smith W, Zhao YP (2009) Superior photocatalytic performance by vertically aligned core– shell TiO2/WO3 nanorod arrays. Catal Commun 10:1117–1121. doi:10.1016/j.catcom.2009. 01.010 530. Jeon TH, Choi W, Park H (2011) Photoelectrochemical and photocatalytic behaviors of hematite-decorated titania nanotube arrays: energy level mismatch versus surface specific activity. J Phys Chem C 115:7134–7142. doi:10.1021/jp201215t 531. Woan K, Pyrgiotakis G, Sigmund W (2009) Photocatalytic carbon-nanotube–TiO2 composites. Adv Mater 21:2233–2239. doi:10.1002/adma.200802738 532. Eder D (201) Carbon nanotube-inorganic hybrids. Chem Rev 110:1348–1385. doi:10.1021/ cr800433k 533. Luo Y, Liu J, Xia X et al (2007) Fabrication and characterization of TiO2/short MWNTs with enhanced photocatalytic activity. Mater Lett 61:2467–2472. doi:10.1016/j.matlet.2006. 09.051 534. Gao B, Peng C, Chen GZ, Li Puma G (2008) Photo-electro-catalysis enhancement on carbon nanotubes/titanium dioxide (CNTs/TiO2) composite prepared by a novel surfactant wrapping sol–gel method. Appl Catal B 85:17–23. doi:10.1016/j.apcatb.2008.06.027 535. Yu H, Quan X, Chen S et al (2008) TiO2–carbon nanotube heterojunction arrays with a controllable thickness of TiO2 layer and their first application in photocatalysis. J Photochem Photobiol, A 200:301–306. doi:10.1016/j.jphotochem.2008.08.007 536. Wang W, Serp P, Kalck P et al (2008) Preparation and characterization of nanostructured MWCNT-TiO2 composite materials for photocatalytic water treatment applications. Mater Res Bull 43:958–967. doi:10.1016/j.materresbull.2007.04.032 537. Chen LC, Ho YC, Guo WS et al (2009) Enhanced visible light-induced photoelectrocatalytic degradation of phenol by carbon nanotube-doped TiO2 electrodes. Electrochim Acta 54:3884–3891. doi:10.1016/j.electacta.2009.02.001 538. Yen CY, Lin YF, Hung CH et al (2008) The effects of synthesis procedures on the morphology and photocatalytic activity of multi-walled carbon nanotubes/TiO2 nanocomposites. Nanotechnology 19:045604 539. Xia XH, Jia ZJ, Yu Y et al (2007) Preparation of multi-walled carbon nanotube supported TiO2 and its photocatalytic activity in the reduction of CO2 with H2O. Carbon 45:717–721. doi:10.1016/j.carbon.2006.11.028 540. Wang H, Wang HL, Jiang WF, Li ZQ (2009) Photocatalytic degradation of 2,4-dinitrophenol (DNP) by multi-walled carbon nanotubes (MWCNTs)/TiO2 composite in aqueous solution under solar irradiation. Water Res 43:204–210. doi:10.1016/j.watres. 2008.10.003 314 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

541. Wang H, Wang HL, Jiang WF (2009) Solar photocatalytic degradation of 2,6-dinitro-p-cresol (DNPC) using multi-walled carbon nanotubes (MWCNTs)–TiO2 composite photocatalysts. Chemosphere 75:1105–1111. doi:10.1016/j.chemosphere.2009. 01.014 542. Yao Y, Li G, Ciston S et al (2008) Photoreactive TiO2/carbon nanotube composites: synthesis and reactivity. Environ Sci Technol 42:4952–4957. doi:10.1021/es800191n 543. Tang ZR, Li F, Zhang Y et al (2011) Composites of titanate nanotube and carbon nanotube with high mineralization ratio for gap-phase degradation of volatile aromatic pollutant. J Phys Chem C 115:7880–7886. doi:10.1021/jp1115838 544. Zhu LP, Liao GH, Huang WY et al (2009) Preparation, characterization and photocatalytic properties of ZnO-coated multi-walled carbon nanotubes. Mater Sci Eng, B 163:194–198. doi:10.1016/j.mseb.2009.05.021 545. Oh WC, Jung AR, Ko WB (2009) Characterization and relative photonic efficiencies of a new nanocarbon/TiO2 composite photocatalyst designed for organic dye decomposition and bactericidal activity. Mater Sci Eng, C 29:1338–1347. doi:10.1016/j.msec.2008.10.034 546. Wu H, Wang Q, Yao Y et al (2008) Microwave-assisted synthesis and photocatalytic properties of carbon nanotube/zinc sulfide heterostructures. J Phys Chem C 112:16779– 16783. doi:10.1021/jp8069018 547. Wu HQ, Wang Q, Yao YZ et al (2009) Microwave-assisted synthesis and highly photocatalytic activity of MWCNT/ZnSe heterostructures. Mater Chem Phys 113:539–543. doi:10.1016/j.matchemphys.2008.08.004 548. Maeda K, Eguchi M, Youngblood WJ, Mallouk TE (2008) Niobium oxide nanoscrolls as building blocks for dye-sensitized hydrogen production from water under visible light irradiation. Chem Mater 20:6770–6778. doi:10.1021/cm801807b 549. Tsubota T, Ono A, Murakami N, Ohno T (2009) Characterization and photocatalytic performance of carbon nanotube material-modified TiO2 synthesized by using the hot CVD process. Appl Catal B 91:533–538. doi:10.1016/j.apcatb.2009.06.024 550. Kim TW, Lee MJ, Shim WG et al (2008) Adsorption and photocatalytic decomposition of organic molecules on carbon-coated TiO2. J Mater Sci 43:6486–6494. doi:10.1007/s10853- 008-2978-2 551. Guo Y, Wang H, He C et al (2009) Uniform carbon-coated ZnO nanorods: microwave-assisted preparation, cytotoxicity, and photocatalytic activity. Langmuir 25:4678–4684. doi:10.1021/la803530h 552. Zhong J, Chen F, Zhang J (2010) Carbon-deposited TiO2: Synthesis, characterization, and visible photocatalytic performance. J Phys Chem C 114:933–939. doi:10.1021/jp909835m 553. Zhang L, Cheng H, Zong R, Zhu Y (2009) Photocorrosion suppression of ZnO nanoparticles via hybridization with graphite-like carbon and enhanced photocatalytic activity. J Phys Chem C 113:2368–2374. doi:10.1021/jp807778r 554. Hu X, Huang K, Fang D, Liu S (2011) Enhanced performances of dye-sensitized solar cells based on graphite-TiO2 composites. Mater Sci Eng, B 176:431–435. doi:10.1016/j.mseb. 2010.12.016 555. Sellappan R, Zhu J, Fredriksson H et al (2011) Preparation and characterization of TiO2/carbon composite thin films with enhanced photocataloytic activity. J Mol Catal A 335:136–144. doi:10.1016/j.molcata.2010.11.025 556. Novoselov KS, Geim AK, Morozov SV et al (2004) Electric field effect in atomically thin carbon films. Science 306:666–669. doi:10.1126/science.1102896 557. Meyer JC, Geim AK, Katsnelson MI et al (2007) The structure of suspended graphene sheets. Nature 446:60–63. doi:10.1038/nature05545 558. Dreyer DR, Park S, Bielawski CW, Ruoff RS (2010) The chemistry of graphene oxide. Chem Soc Rev 39:228–240. doi:10.1039/B917103G 559. Park S, Ruoff RS (2009) Chemical methods for the production of graphenes. Nat Nanotechnol 4:217–224. doi:10.1038/nnano.2009.58 References 315

560. Kamat PV (2010) Graphene-based nanoarchitectures. Anchoring semiconductor and metal nanoparticles on a two-dimentional carbon support. J Phys Chem Lett 1:520–527. doi:10. 1021/jz900265j 561. Zhu H, Wei J, Wang K, Wu D (2009) Applications of carbon materials in photovoltaic solar cells. Sol En Mater Sol Cells 93:1461–1470. doi:10.1016/j.solmat.2009.04.006 562. Stroyuk OL, Andryushina NS, Kuchmiy SY et al (2015) Photochemical processes involving graphene oxide. Theoret Exp Chem 51:1–29. doi:10.1007/s11237-015-9393-y 563. Kamat PV, Bedja I, Hotchandani S (1994) Photoinduced charge transfer between carbon and semiconductor clusters. One-electron reduction of C60 in colloidal TiO2 semiconductor suspensions. J Phys Chem 98:9137–9142. doi:10.1021/j100088a008 564. Brown P, Kamat PV (2008) Quantum dot solar cells. Electrophoretic deposition of CdSe-C60 composite films and capture of photogenerated electrons with nC60 cluster shell. J Am Chem Soc 130:8890–8891. doi:10.1021/ja802810c 565. Kathiravan KA, Sathish Kumar P, Renganathan R, Anandan S (2009) Photoinduced electron transfer reactions between meso-tetrakis(4-sulfonatophenyl)porphyrin and colloidal metal-semiconductor nanoparticles. Colloids Surf A 333:175–181. doi:10.1016/j.colsurfa. 2008.09.042 566. Dai W, Zheng X, Yang H et al (2009) The promoted effect of UV irradiation on preferential oxidation of CO in an H2-rich stream over Au/TiO2. J Power Sources 188:507–514. doi:10. 1016/j.jpowsour.2008.12.028 567. Bannat I, Wessels K, Oekermann T et al (2009) Improving the photocatalytic performance of mesoporous titania films by modification with gold nanostructures. Chem Mater 21:1645– 1653. doi:10.1021/cm803455k 568. Abdulla-Al-Mamun M, Kusumoto Y, Ahmmad B, Shariful Islam M (2010) Photocatalytic cancer (HeLa) cell-killing enhanced with Cu–TiO2 nanocomposite. Topics Catal 53:571– 577. doi:10.1007/s11244-010-9489-5 569. Yoshinaga M, Yamamoto K, Sato N et al (2009) Remarkably enhanced photocatalytic activity by nickel nanoparticle deposition on sulfur-doped titanium dioxide thin film. Appl Catal B 87:239–244. doi:10.1016/j.apcatb.2008.08.023 570. Lin HY, Chang YS (2010) Photocatalytic water splitting for hydrogen production on Au/KTiNbO5. Inter J Hydrogen En 35:8463–8471. doi:10.1016/j.ijhydene.2010.06.006 571. Wang W, Zhang J, Chen F et al (2008) Preparation and photocatalytic properties of Fe3+- doped Ag@TiO2 core–shell nanoparticles. J Colloid Interface Sci 323:182–186. doi:10. 1016/j.jcis.2008.03.043 572. Sakai N, Fujiwara Y, Takahashi Y, Tatsuma T (2009) Plasmon-resonance-based generation of cathodic photocurrent at electrodeposited gold nanoparticles coated with TiO2 films. ChemPhysChem 10:766–769. doi:10.1002/cphc.200800704 573. Ikuma Y, Bessho H (2007) Effect of Pt concentration on the production of hydrogen by a TiO2 photocatalyst. Inter J Hydrogen En 32:2689–2692. doi:10.1016/j.ijhydene.2006.09.024 574. Miyao T, Suzuki Y, Naito S (2000) Hydrogen formation by the photodecomposition of water over Pt/TiO2 suspended in a water in oil emulsion. Catal Lett 66:197–200. doi:10. 1023/A:1019032614298 575. Ohtani B, Iwai K, Nishimoto S, Sato S (1997) Role of platinum deposits on titanium (IV) oxide particles: structural and kinetic analyses of photocatalytic reaction in aqueous alcohol and amino acid solutions. J Phys Chem B 101:3349–3359. doi:10.1021/jp962060q 576. Sun B, Vorontsov AV, Smirniotis PG (2003) Role of platinum deposited on TiO2 in phenol photocatalytic oxidation. Langmuir 19:3151–3156. doi:10.1021/la0264670 577. Subramanian V, Wolf EE, Kamat PV (2001) Semiconductor-metal composite nanostructures: to what extent do metal nanoparticles improve the photocatalytic activity of TiO2 films. J Phys Chem B 105:11439–11446. doi:10.1021/jp011118k 578. Dawson A, Kamat PV (2001) Semiconductor-metal nanocomposites. Photoinduced fusion and photocatalysis of gold-capped TiO2 (TiO2/gold) nanoparticles. J Phys Chem B 105:960– 966. doi:10.1021/jp0033263 316 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

579. Subramanian V, Wolf EE, Kamat PV (2003) Influence of metal/metal ion concentration on the photocatalytic activity of TiO2-Au composite nanoparticles. Langmuir 19:469–474. doi:10.1021/la026478t 580. Zheng Y, Zheng L, Zhan Y et al (2007) Ag/ZnO heterostructure nanocrystals: synthesis, characterization, and photocatalysis. Inorg Chem 46:6980–6986. doi:10.1021/ic700688f 581. Gu C, Cheng C, Huang H et al (2009) Growth and photocatalytic activity of dendrite-like ZnO@Ag heterostructure nanocrystals. Cryst Growth Design 9:3278–3285. doi:10.1021/ cg900043k 582. Lombardi I, Marchionna S, Zangari G, Pizzini S (2007) Effect of Pt particle size and distribution on photoelectrochemical hydrogen evolution by p-Si photocathodes. Langmuir 23:12413–12420. doi:10.1021/la7016165 583. Sheeney-Haj-Ichia L, Pogorelova S, Gofer Y, Willner I (2004) Enhanced photoelectrochemistry in CdS/Au nanoparticle bilayers. Adv Func Mater 14:416–424. doi:10.1002/adfm. 200305430 584. Lana-Villarreal T, Gomez R (2005) Interfacial electron transfer at TiO2 nanostructured electrodes modified with capped gold nanoparticles: the photoelectrochemistry of water oxidation. Electrochem Commun 7:1218–1224. doi:10.1016/j.elecom.2005.08.031 585. Mizukoshi Y, Makise Y, Shuto T et al (2007) Immobilization of noble metal nanoparticles on the surface of TiO2 by the sonochemical method: photocatalytic production of hydrogen from an aqueous solution of ethanol. Ultrason Sonochem 14:387–392. doi:10.1016/j. ultsonch.2006.08.001 586. Mizukoshi Y, Sato K, Konno TJ, Masahashi N (2010) Dependence of photocatalytic activities upon the structures of Au/Pd bimetallic nanoparticles immobilized on TiO2 surface. Appl Catal B 94:248–253. doi:10.1016/j.apcatb.2009.11.015 587. Tian B, Zhang J, Tong T, Chen F (2008) Preparation of Au/TiO2 catalysts from Au (I)– thiosulfate complex and study of their photocatalytic activity for the degradation of methyl orange. Appl Catal B 79:394–401. doi:10.1016/j.apcatb.2007.11.001 588. Haldar KK, Sen T, Patra A (2008) Au@ZnO core-shell nanoparticles are efficient energy acceptors with organic dye donors. J Phys Chem C 112:11650–11656. doi:10.1021/ jp8031308 589. Liu Z, Guo B, Hong L, Jiang H (2005) Physicochemical and photocatalytic characterizations of TiO2/Pt nanocomposites. J Photochem Photobiol, A 172:81–88. doi:10.1016/j. jphotochem.2004.11.008 590. Yi Z, Wei W, Lee S, Jianhua G (2007) Photocatalytic performance of plasma sprayed Pt-modified TiO2 coatings under visible light irradiation. Catal Commun 8:906–912. doi:10. 1016/j.catcom.2006.09.023 591. Yin S, Sato T (2005) Photocatalytic activity of platinum loaded fibrous titania prepared by sol–vothermal process. J Photochem Photobiol, A 169:89–94. doi:10.1016/j.jphotochem. 2004.05.038 592. Alam Khan M, Shaheer Akhtar M, Woo SI, Yang O (2008) Enhanced photoresponse under visible light in Pt ionized TiO2 nanotube for the photocatalytic splitting of water. Catal Commun 10:1–5. doi:10.1016/j.catcom.2008.01.018 593. Shan Z, Wu J, Xu F et al (2008) Highly effective silver/semiconductor photocatalytic composites prepared by a silver mirror reaction. J Phys Chem C 112:15423–15428. doi:10. 1021/jp804482k 594. Moonsiri M, Rangsunvigit P, Chavadej S, Gulari E (2004) Effects of Pt and Ag on the photocatalytic degradation of 4-chlorophenol and its by-products. Chem Eng J 97:241–248. doi:10.1016/j.cej.2003.05.003 595. Zhou G, Deng J (2007) Preparation and photocatalytic performance of Ag/ZnO nano-composites. Mater Sci Semicon Proc 10:90–96. doi:10.1016/j.mssp.2007.05.003 596. Zhang Y, Mu J (2007) One-pot synthesis, photoluminescence, and photocatalysis of Ag/ZnO composites. J Colloid Interface Sci 309:478–484. doi:10.1016/j.jcis.2007.01.011 References 317

597. ArañaJ,Doña-Rodriguez JM, Herrera Melian JA et al (2005) Role of Pd and Cu in gas-phase alcohols photocatalytic degradation with doped TiO2. J Photochem Photobiol, A 174:7–14. doi:10.1016/j.jphotochem.2005.03.003 598. Wu NL, Lee MS (2004) Enhanced TiO2 photocatalysis by Cu in hydrogen production from aqueous methanol solution. Inter J Hydrogen En 29:1601–1605. doi:10.1016/j.ijhydene. 2004.02.013 599. Chiarello GL, Aguirre MH, Selli E (2010) Hydrogen production by photocatalytic steam reforming of methanol on noble metal-modified TiO2. J Catal 273:182–190. doi:10.1016/j. jcat.2010.05.012 600. Chiarello GL, Forni L, Selli E (2009) Photocatalytic hydrogen production by liquid- and gas-phase reforming of CH3OH over flame-made TiO2 and Au/TiO2. Catal Today 144:69– 74. doi:10.1016/j.cattod.2009.01.023 601. Lai Y, Zhuang H, Xie K et al (2010) Fabrication of uniform Ag/TiO2 nanotube array structures with enhanced photoelectrochemical performance. New J Chem 34:1335–1340. doi:10.1039/B9NJ00780F 602. Pang H, Li Y, Guan L et al (2011) TiO2/Ni nanocomposites: biocompatible and recyclable magnetic photocatalysts. Catal Commun 12:611–615. doi:10.1016/j.catcom.2010.12.015 603. Ruvarac-Bugarčić IA, Šaponjić ZV, Zec S et al (2005) Photocatalytic reduction of cadmium on TiO2 nanoparticles modified with amino acids. Chem Phys Lett 407:110–113. doi:10. 1016/j.cplett.2005.03.058 604. Somasundaram S, Ming Y, Chenthamarakshan CR et al (2004) Free radical-mediated hetero–geneous photocatalytic reduction of metal ions in UV-irradiated titanium dioxide suspensions. J Phys Chem B 108:4784–4788. doi:10.1021/jp036729m 605. Wood A, Giersing M, Mulvaney P (2001) Fermi level equilibration in quantum dot—metal nanojunctions. J Phys Chem B 105:8810–8815. doi:10.1021/jp011576t 606. Siemon U, Bahnemann D, Testa JJ et al (2002) Heterogeneous photocatalytic reactions comparing TiO2 and Pt/TiO2. J Photochem Photobiol, A 148:247–255. doi:10.1016/S1010- 6030(02)00050-3 607. Rabani J, Behar D (1989) Quenching of aqueous colloidal zinc oxide fluorescence by electron and hole scavengers: effect of a positive polyelectrolyte. J Phys Chem 93:2559– 2563. doi:10.1021/j100343a063 608. Subramanian V, Wolf EE, Kamat PV (2003) Green emission to probe photoinduced charging events in ZnO–Au nanoparticles. Charge distribution and Fermi level equilibration. J Phys Chem B 107:7479–7485. doi:10.1021/jp0275037 609. Vigil E, Gonzales B, Zumeta I et al (2005) Preparation of photoelectrodes with spectral res– ponse in the visible without applied bias based on photochemically deposited copper oxide inside a porous titanium dioxide film. Thin Solid Films 489:50–55. doi:10.1016/j.tsf.2005. 04.098 610. Lu L, Hu S, Lee HI et al (2007) Photoinduced growth of Cu nanoparticles on ZnO from CuCl2 in methanol. J Nanopart Res 9:491–496. doi:10.1007/s11051-006-9087-4 611. Liqiang J, Baiqi W, Baifu X et al (2004) Investigations on the surface modification of ZnO nanoparticle photocatalyst by depositing Pd. J Sol State Chem 177:4221–4227. doi:10.1016/ j.jssc.2004.08.016 612. Bae E, Choi W (2003) Highly enhanced photoreductive degradation of perchlorinated compounds on dye-sensitized metal/TiO2 under visible light. Environ Sci Technol 37:147– 152. doi:10.1021/es025617q 613. Liu Y, Liu C, Zhang Z, Wang C (2001) The surface enhanced Raman scattering effects of composite nanocrystals of Ag–TiO2. Spectrochim Acta A 57:35–39. doi:10.1016/S1386- 1425(00)00326-7 614. He J, Yang P, Sato H et al (2004) Effects of Ag-photodeposition on photocurrent of an ITO electrode modified by a hybrid film of TiO2 nanosheets. J Electroanal Chem 566:227–233. doi:10.1016/j.jelechem.2003.11.031 318 5 Synthesis of Nanocrystalline Photo-Active Semiconductors

615. Korzhak AV, Ermokhina NI, Stroyuk OL et al (2008) Photocatalytic hydrogen evolution over mesoporous TiO2/Metal nanocomposites. J Photochem Photobiol, A 198:126–134. doi:10.1016/j.jphotochem.2008.02.026 616. Ng YH, Ikeda S, Harada T et al (2008) Photocatalytic route for synthesis of hollow porous carbon/Pt nanocomposites with controllable density and porosity. Chem Mater 20:1154– 1160. doi:10.1021/cm702034w 617. Naoi K, Ohko Y, Tatsuma T (2004) TiO2 films loaded with silver nanoparticles: control of multicolor photochromic behavior. J Am Chem Soc 126:3664–3668. doi:10.1021/ja039474z 618. Uddin MJ, Cesano F, Bertarione S et al (2008) Tailoring the activity of Ti-based photocatalysts by playing with surface morphology and silver doping. J Photochem Photobiol, A 196:165–173. doi:10.1016/j.jphotochem.2007.07.037 619. Tanahashi I, Iwagishi H, Chang G (2008) Localized surface plasmon resonance sensing properties of photocatalytically prepared Au/TiO2 films. Mater Lett 62:2714–2716. doi:10. 1016/j.matlet.2008.01.023 620. Tanahashi I (2007) Photocatalytic preparation of Ag/TiO2 films and their localized surface plasmon resonance sensing properties. Bull Chem Soc Jpn 80:2019–2023. doi:10.1246/bcsj. 80.2019 621. Paramasivam I, Macak JM, Ghicov A, Schmuki P (2007) Enhanced photochromism of Ag loaded self-organized TiO2 nanotube layers. Chem Phys Lett 445:233–237. doi:10.1016/j. cplett.2007.07.107 622. Zolotavin P, Permenova E, Sarkosov O et al (2008) Two-photon luminescence enhancement of silver nanoclusters photodeposited onto mesoporous TiO2 film. Chem Phys Lett 457:342– 346. doi:10.1016/j.cplett.2008.04.034 Chapter 6 Probing with Light—Optical Methods in Studies of Nanocrystalline Semiconductors

The studies of photochemical and photocatalytic processes involving the semiconductor NPs are performed using a broad range of modern physical and chemical methods applied to determine the NP size and structure, their spectral, photophysical and other properties. The most frequently used are electron spectroscopy in the absorption, transmission and reflection modes, photoluminescence spectroscopy, electron microscopy (in the scanning and transmission modes), X-rays diffraction, etc. Nuances of the photochemical properties of semiconductor NPs can be revealed using the lamp and laser flash photolysis, the electron paramagnetic resonance and the Raman spectroscopy. The voltammetry is often used to determine the potentials of conduction and valence bands of semiconductor NPs, the adsorption/desorption methods—for the determination of the specific surface area and pore size of semiconductor nanophotocatalysts. A detailed description of these methods is far beyond the scope of the present book. This chapter is confined to the methods using light to probe the properties of nanocrystalline semiconductors and thus shedding light on the structure and properties of these fascinating objects as a result of their interaction with the probing irradiation.

6.1 A Brief Characterization of the Spectral Studies of Nano-Semiconductors

The spectral studies of optical properties of the nanocrystalline semiconductors provide the most ample information when applied to optically transparent colloidal systems and ﬁlms. The family of spectral methods includes the electron absorption spectroscopy, stationary and time-resolved luminescence spectroscopy, Raman spectroscopy, as well as the time-resolved pulse spectroscopy (the ﬂash photolysis) in the femto-microsecond time domains (Fig. 6.1).

© Springer International Publishing AG 2018 319 O. Stroyuk, Solar Light Harvesting with Nanocrystalline Semiconductors, Lecture Notes in Chemistry 99, https://doi.org/10.1007/978-3-319-68879-4_6 320 6 Probing with Light—Optical Methods in Studies …

Fig. 6.1 Family of spectral methods for the studies of nanocrystalline semiconductors

The studies of semiconductor NPs by the electron absorption spectroscopy (in a transmission mode—for colloidal NPs and films, in a reflection mode—for nanocrystalline powders and thick films) allow to determine the type and energy of interband electron transitions in the nano-objects and to calculate the bandgap Eg or the energy of the first exciton transitions E1 for the case of strong confinement effects. For the quantum-sized semiconductor NPs, Eg depends on the NP dimensions and the NP size d can be calculated by using reported empirical Eg(d) dependences, while the size distribution of NPs can be evaluated from the spectral width of the excitonic absorption band. Basing on the optical bandgap Eg and the reported values of bulk CB and VB potentials of a given semiconductor the ECB and EVB potentials can be evaluated for the NPs of a given size. In the case of solid-solution nanophotocatalysts, such as CdxZn1−xS or CdSxSe1−x, the reported bandgap dependences on the composition (x value) can be used to determine the NP composition (for the NPs with weak or no confinement effects). The analysis of a longer-wavelength spectral range of the absorption spectra of semi– conductor NPs corresponding to hv < Eg and originating from the absorbance on the lattice defects provides information on the energy and density of local defect-related electron states residing in the forbidden band of semiconductor NPs. Finally, the analysis of the surface plasmon resonance bands of metal NPs (Au, Ag, Cu) in 6.1 A Brief Characterization of the Spectral Studies … 321 semiconductor/metal nanocomposites allows evaluating the size of metal NPs, the metal content, electron gas density, conductivity and other characteristics. The photoluminescence (PL) spectroscopy is used to determine the energy of photons emitted as a result of the electron-hole recombination and the energy and density of local defect-related states acting as the electron/hole trapping sites. In the case of direct (excitonic) PL, the size and size distribution of semiconductor NPs can be derived from the reported empirical EPL(d) dependences. The observations of PL decay dynamics, as well as the PL quenching by various substrates, provide ample information on the rate of the radiative recombination and mechanisms of the interaction between the semiconductor NPs and other components of the photocatalytic and photoelectrochemical systems. The Raman spectroscopy is used to study the structure of nanocrystalline semiconductors, especially the “core/shell” composites. The method can also reveal the phase composition of NPs, their size and a degree of the NP lattice disorder. The time-resolved (flash) photolysis, both in the conventional transmission mode and in the mode of diffuse reflectance, is a powerful tool for probing the primary photochemical processes that occur with the participation of semiconductor NPs and nanoheterostructures. The observations of the spectral parameters and decay kinetics of the absorption bands in the non-stationary differential spectra registered under the pulsed illumination provide information on the structure and reactivity of the short-lived intermediates. Complementary, the intensity and relaxation rate of the “negative” non-stationary bleaching bands bear valuable information on the capability of semiconductor NPs to the accumulation of an excessive charge during the light pulse as well as the dynamics of following charge transfers to the accepting components of the photochemical system. The above spectral methods can be used both separately and in a complex way. For some systems, combinations of the spectral methods (for example, the PL spectroscopy and flash photolysis) can provide quite unique information on the structure and properties of nano-semiconductors and the products of photochemical reactions on their surfaces.

6.2 Studies of Nano-Photocatalysts by the Electron Absorption Spectroscopy

The basic features of the interaction between the electromagnetic irradiation and semiconducting materials, the nature of absorption bands in electron spectra of the semiconductors, as well as the inﬂuence of spatial exciton conﬁnement on the optical properties of semiconductor NPs were discussed in Chap. 1. Here we focus on the methodology of extracting the information on the type and energy of electron transitions, on the electrophysical parameters of semiconductor NPs, as well on as the NP size and size distribution from the electron absorption spectra. 322 6 Probing with Light—Optical Methods in Studies …

Determination of basic electrophysical parameters of nanocrystalline semiconductors. The character of light absorption by semiconductor NPs depends on many factors, in particular on the light wavelength, th semiconductor composition, the density of “alien” atoms and inherent structural defects, the free charge carrier density, etc. The fundamental light absorption by a semiconductor results in the generation of À þ a couple of a conduction band electron (eCB) and a valence band hole (hVB) bound by the Coulomb force. The energy of electron transition from VB to CB depends on the bandgap Eg (forbidden band between VB and CB). The bandgap value sets the range of spectral sensitivity of the semiconductor which is also a function of the NP size in the case of quantum-sized semiconductor NPs. The bandgap can be estimated using the absorption threshold, that is, from the wavelength corresponding to the fundamental absorption band edge (kbe)as Eg = 1241/kbe, where Eg is expressed in eV, kbe—in nm. It can be calculated as an intersection point between the abscissae axis and a tangent to the linear section of the absorption band edge—the “tangent” method (Fig. 6.2a). As the longer-wavelength band edge of an ensemble of the quantum-sized semiconductor NPs is formed mostly by a fraction of larger NPs, the bandgap derived from the tangent method should be regarded as a lower limit of the bandgaps in the ensemble. A more precise determination of the average band gap can be realized in the cases where a distinct excitonic absorption maximum is present in the absorption spectra corresponding to a transition between quantized (discrete) VB/CB levels (Fig. 6.2b). Typically, the colloidal semiconductor NPs are produced in the form of rather polydisperse NP ensembles with a 15–20% size distribution and the ﬁrst excitonic maximum becomes smeared for such systems making possible only the tangent-based bandgap determination. Both methods for the Eg determination are valid only for the direct interband electron transitions. In cases where the type of transition is unknown, it can also be deduced from the absorption spectrum.

Fig. 6.2 Examples of the determination of the bandgap of colloidal ZnO NPs (a) and CdS NPs (b) by the tangent method (a) and from the position of the ﬁrst absorption maximum (b)[90, 124, 125] 6.2 Studies of Nano-Photocatalysts by the Electron … 323

Determination of the type of electron transition. In general, the fundamental light absorption by a semiconductor can originate from the direct and indirect interband electron transitions (see Chap. 1). The absorption bands corresponding to the indirect transitions are typically featureless and do not have a distinct edge. In the case of a direct electron transition, the fundamental absorption band typically reveals a distinct edge at hv = Eg (Fig. 6.3a) and a low light absorption at hv < Eg tha latter originating both from the size distribution of semiconductor NPs and from the electron transitions with the participation of mid-bandgap defect-related levels. This section of the spectrum appears linear in the coordinates of Urbach Eq. (1) that describes the dependence of the light absorption coefﬁcient a on the light quantum energy

hv ln a ¼ b ; ð1Þ kT where b is a constant characterizing the structure disordering, k is the Boltzmann constant, T is temperature. The fundamental absorption spectrum, that is, the dependence of the fundamental absorption coefﬁcient a on the light quantum energy is described by a general Eq. (2):

ðhv À E Þn a ¼ A g ; ð2Þ hv where A is a constant, while n depends on the transition type and can be equal to ½ 3 and 2 for the allowed direct and indirect transitions, respectively, and to /2 and 3— for the forbidden direct and indirect transitions. The n constant can be determined from a tangent of the absorption spectrum linearized in the coordinates ln(ahv) versus ln(hv − Eg). For example, for colloidal ZnS NPs n was found to be 0.50 ± 0.02 (Fig. 6.3a, insert) indicating that the

Fig. 6.3 a Absorption spectrum of colloidal ZnS solution. Insert: the spectrum linearized in the coordinates ln(ahv) versus ln(hv − Eg); b absorption spectrum of ZnS NPs presented as d{ln(ahv)}/ d{hv} versus hv. The dashed line indicates a discontinuity point [126] 324 6 Probing with Light—Optical Methods in Studies … fundamental absorption band edge of ZnS NPs corresponds to an allowed direct interband electron transition. The ﬁrst derivative of the Eq. (2)—function (3) reveals a discontinuity point at hv = Eg (Fig. 6.3b) allowing for a more precise determination of the bandgap:

dfglnðahvÞ n ¼ : ð3Þ dðhvÞ hv À Eg

For a given n the bandgap Eg can be calculated as a cross point between the linear section of “(ahv)1/n – hv” dependence and the abscissae (hv) axis. Example—the determination of electrophysical parameters of solid-solution CdxZn1−xS NPs. The nanocrystalline materials based on mixed cadmium-zinc sulfide are quite broadly used as the photocatalysts of various redox processes including the reduction of metal ions and methylviologen, evolution of hydrogen from aqueous solutions of sacrificial donors, photopolymerization, etc. The analysis of CdxZn1−xS-based photocatalytic systems and possible mechanisms of the photocatalytic reactions requires information on the electrophysical properties of CdxZn1−xS NPs of any given composition. The methodology of determination of the composition-dependent Eg, ECB, and EVB parameters of cadmium-zinc sulfide solid-solutions proposed in [1–3] is of a general character and can be applied to other solid solutions of semiconductor materials for which the bulk properties of separate components are known. The SPP-stabilized aqueous colloidal CdxZn1−xS NPs with x varied from 0 to 1 can be easily produced via the interaction between a mixture of cadmium(II) and zinc(II) chlorides and sodium sulfide [2]. An XRD study of the dried residuals from such colloids showed that the reaction yields cubic CdxZn1−xS solid solution NPs with an average size of 6 nm, irrespectively of the composition [2]. For such NPs, the quantum size effects are expressed only weakly and therefore the position of the fundamental

Fig. 6.4 Absorption spectra of colloidal CdxZn1−xS NPs produced at x = 0 (curve 1), 0.25 (curve 2), 0.50 (curve 3), 0.75 (curve 4), and 1.00 (curve 5) [1–3] 6.2 Studies of Nano-Photocatalysts by the Electron … 325 absorption band edge is determined mainly by the NP composition, not their size (Fig. 6.4)[1–3]. In the family of bulk (microcrystalline) CdxZn1−xS solid solutions the composition variation from pure cadmium sulﬁde (x = 1.0) to pure zinc sulﬁde (x =0) results in a change of the valence band (VB) potential from 1.6 V versus normal hydrogen electrode (NHE) for CdS [4] to 1.8 V (NHE) for ZnS [4]. This VB variation is small comparatively to a large conduction band (CB) variation from − − bulk 0.8 V for CdS up to 1.8 V for ZnS. Therefore, the EVB of composition-varied CdxZn1−xS can be approximately presented by a linear combination (4)[1, 2]:

bulk bulk bulk : EVB ðÞ¼x xEVB ðÞþCdS ðÞ1 À x EVB ðÞZnS ð4Þ

bulk The bandgap Eg of bulk CdxZn1−xS crystals can be calculated using the reported empirical expression (5)[5, 6]:

bulk : : : 2: Eg ¼ 3 6 À 1 78x þ 0 61x ð5Þ

bulk bulk Basing on Eg and EVB calculated for a given x the bulk CB potential can be expressed as

ECBðÞ¼x EVBðÞÀx EgðÞx : ð6Þ

Table 6.1 presents the Eg, ECB, and EVB parameters of CdxZn1−xS NPs (with no QSEs observable for such NPs) calculated using Eqs. (4–6). For the colloidal CdxZn1−xS NPs, for which the QSEs become possible, the bandgap can be calculated by using the above discussed “tangent” method from the position of kbe. The QSE inﬂuence on the positions of ECB and EVB is typically accounted for by using the effective mass approximation and the values of effective electron and hole masses determined as linear combinations of the corresponding parameters of the individual cadmium and zinc sulﬁde [5]:

Table 6.1 Bandgap and xEg,eV EVB, V (NHE) ECB, V (NHE) conduction and valence band − potentials of bulk cubic 1.0 2.40 1.60 0.80 − CdxZn1−xS solid solutions 0.9 2.47 1.62 0.85 crystals of a varied 0.8 2.54 1.64 −0.90 composition calculated using 0.7 2.63 1.66 −0.97 Eqs. (4–6) 0.6 2.73 1.68 −1.05 0.5 2.85 1.70 −1.15 0.4 2.97 1.72 −1.25 0.3 3.11 1.74 −1.37 0.2 3.26 1.76 −1.50 0.1 3.43 1.78 −1.65 0.0 3.60 1.80 −1.80 326 6 Probing with Light—Optical Methods in Studies …

Ã Ã Ã ; Ã Ã Ã me ðÞ¼x xme ðÞþCdS ðÞ1Àx me ðÞZnS and mhðÞ¼x xmhðÞþCdS ðÞ1Àx mhðÞZnS

Ã Ã Ã Ã where me (CdS) = 0.20m0, me (ZnS) = 0.27m0, mh(CdS) = 0.70m0, mh(ZnS) = 0.58m0 [7], m0 is the free electron rest mass. Then, the potentials of allowed bands of CdxZn1−xS NPs of any given compo- Ã Ã sition and size can be estimated using Eqs. (7) and (8), the me (x) and mh(x) parameters calculated from the bulk effective masses and the bandgap Eg determined from the absorption spectra of colloidal solutions: ÀÁ bulk Ã Ã Ã À1 bulk ; ECBðÞ¼x ECB ðÞÀx mh me þ mh ðEg À Eg Þ ð7Þ ÀÁ bulk Ã Ã Ã À1 bulk EVBðÞ¼x EVB ðÞþx me me þ mh ðEg À Eg Þð8Þ

Determination of the average size d and size distribution of semiconductor NPs using empirical Eg(d) calibration curves. The effective mass approximation (EMA) describes the size scaling of the bandgap of semiconductor NPs with a size d falling into the regime of weak/moderate exciton confinement (see Chap. 1). For the NPs with medium-to-strong QSEs, the basic assumptions of EMA become invalid and the calculational results do not correspond to the experimental measurements by the direct methods (such as TEM). In this view, for a number of semiconductor NPs empirical calibration curves were proposed that combine the data reported in various papers (in some cases, tens of separate papers, as in the case of cadmium selenide). The calibration curves correlate the optical bandgap of size-selected semiconductor NPs with the results of direct size measurements using electron microscopy as well as indirect measurements by the light scattering and the wide angle/small angle X-Ray scattering. Figures 6.5, 6.6, 6.7, 6.8 and 6.9 illustrate some of the reported empirical data on the “bandgap (first excitonic maximum)—NP size” dependences for a series of semiconductor NPs that are typcially used in the nano-photocatalysis and the photoelectrochemical systems, in particular for CdS (Fig. 6.5), ZnO (Fig. 6.6), ZnS (Fig. 6.7), PbS (Fig. 6.8), and CdSe (Fig. 6.9). The computational curves presented in the figures are derived from the “classical” EMA as well as some other reported modeling approaches, while the experimental data were collected by TEM and the wide angle/small angle X-ray scattering. The figures show that EMA agrees well with the experimental data only in the case of CdS NPs. For other presented semiconductor NPs this approximation results in a strong size over-estimation already at d =4–6 nm. The data show that the EMA, though being comparatively simple and universal, should be applied with a caution. In the cases of strong differences between the results of EMA and the direct size measurements, such, for example, as CdSe NPs revealed, alternative and more sophisticated theoretical models or empirical calibration curves built on the 6.2 Studies of Nano-Photocatalysts by the Electron … 327

Fig. 6.5 Calibration Eg(d) data for CdS NPs. Curve 1 is derived from the EMA with Ã Ã me = 0.204m0 and mh = 0.70m0, curve 2 results from the pseudo-potential calculations [18], curve fi Ã Ã 3 is produced by a nite-depth potential well model (3.6 eV, me = 0.18m0, mh = 0.53m0)[18, 127], curves 4–7 were reported as results of various modeling in [128] (curve 4), [129, 130] (curve 5), [131] (curve 6), and [132] (curve 7). The results of TEM/XRD are presented by the hollow circles [133], hollow squares [134], hollow triangles [135], filled areas [136] and filled triangles [137]

Fig. 6.6 Calibration Eg − d data for ZnO NPs. Curve 1 corresponds to the Ã EMA with me = 0.27m0, Ã – mh = 0.50m0 [138 140], curves 2 and 3 are the results of modeling presented in [141] (curve 2) and [9] (curve 3).The results of TEM are given as the hollow squares [138], the results of the small angle X-Ray scattering are presented by the hollow circles [138, 142, 143]

experimental results of TEM and small-angle X-ray scattering measurements should preferably be used to evaluate the NP size. The empirical size calibration curves for the determination of the size of CdSe and CdTe NPs based on tens of separate reports can be found in [8]. Such curves actually accumulate the long history of studies of the NPs produced by different methods and, therefore, the determination of the NP size from optical absorption 328 6 Probing with Light—Optical Methods in Studies …

Fig. 6.7 Calibration Eg − d data for ZnS NPs. Curve 1 corresponds to the bulk EMA with Eg = 3.6 eV, Ã Ã me = 0.27m0, mh = 0.58m0 [144, 145], curve 2 was derived by modeling in [132]. The TEM measurements are given by the hollow circles [146], squares [147], and triangles [148]

Fig. 6.8 Calibration Eg − d data for PbS NPs. Curve 1 corresponds to the bulk EMA with Eg = 0.41 eV, Ã Ã me = 0.08m0, mh = 0.075m0, curves 2–4reﬂect results of modelling reported in [149] (curve 2), [150] (curve 3), and [151] (curve 4). TEM results are presented by the hollow squares [151] and circles [152]; the ﬁlled triangles present results of the small angle X-ray scattering [152]

Fig. 6.9 Calibration E1 − d data for CdSe NPs, where E1 is the position of the first excitonic maximum in absorption spectra. Curve 1 corresponds to the EMA, curve 2–4 is the result of modeling reported in [153, 154] (curve 2), [134, 155] (curve 3), and [156, 157] (curve 4). The TEM data are presented by the hollow squares [158], circles [159], and triangles [19], the filled triangles present the data of small angle X-ray scattering [24] 6.2 Studies of Nano-Photocatalysts by the Electron … 329 spectra using such well-elaborated calibration curves is expected to be one of the most reliable and universal methods [8]. The graphical calibration dependences can be presented in an analytical form to make the size determination even more convenient. Such analytical expressions associate the size of NPs with one of the parameters that can be derived from the absorption spectra—the wavelength of band edge/furst excitonic maximum kbe/k1, the bandgap Eg or the first excitonic maximum energy E1. Several expressions of the kind are presented below for CdS NPs (9)[9], CdSe NPs (10–12)[8–10], CdTe NPs (13, 14)[8, 9], and ZnO NPs (15)[11].

: À8k3 : À4k2 : À2k : dCdS ¼À6 6521 Â 10 be þ 1 9557 Â 10 beÀ9 2352 Â 10 be þ 13 2 ð9Þ

: À9k4 : À6k3 : À3k2 : k : dCdSe ¼ 1 6122 Â 10 1À2 6575 Â 10 1 þ 1 624210 À0 4277 1 þ 41 57 ð10Þ

: : k : À3k2 : À6k3 : dCdSe ¼ 59 60816À0 54736 1 þ 1 8873 Â 10 1À2 85743 Â 10 1 þ 1 62974 À9k4 Â 10 1 ð11Þ ÀÁ 2 À1 EgðÞ¼CdSe 1:858 þ 0:220d þ 0:008d þ 0:373 ð12Þ

: À7k3 : À3k2 : k : dCdTe ¼ 9 8127 Â 10 1À1 7147 Â 10 1 þ 1 0064 1À194 84 ð13Þ ÀÁ 2 À1 EgðÞ¼CdTe 1:596 þ 0:137d þ 0:206 ð14Þ

À1:83 EgðÞ¼ZnO 3:41 þ 3:87 Â d ð15Þ

Such expressions describe quite correctly the dependences of electron and optical properties of the semiconductor NPs on their size in a broad range, typically for d =2–10 nm.

6.3 Luminescence Spectroscopy as a Tool for the Studies of Nanocrystalline Semiconductors

The photoluminescence (PL) spectroscopy, both in the stationary and dynamic regimes, is one of the richest and most versatile methods for the studies of the excited states of nanocrystalline semiconductors. Here, the potential of these methods is discussed on the example of solid solution cadmium-zinc sulfide NPs [1–3]. The optical, photochemical and photocatalytic properties of CdxZn1−xS NPs can be tuned in a broad range by varying both the NP size and composition. The studies 330 6 Probing with Light—Optical Methods in Studies … of composition-selected CdxZn1−xS NPs synthesized in identical conditions can be very useful to gain an insight into the nature and mechanisms of the defect-related PL in individual nanocrystalline cadmium and zinc sulfides [1, 3]. Despite the numerous efforts on both nanosized and bulk crystals of these semiconductors, the mechanisms of defect-related PL in CdS and ZnS still remain a subject of vivid discussions. The present chapter also provides an example of the application of the results of PL spectroscopy for the interpretation of special kinetic features of the photocatalytic processes, in particular, the photocatalytic reduction of 4,4/-dimethylbipyridyl (methylviologen) cation, with the participation of CdxZn1−xS and CdSe NPs. We note also, that a detailed account of the structure of PL spectra of semiconductor NPs, PL mechanisms and decay dynamics depending on the NP composition and size, PL quenching by various substrates, etc. is beyond the scope of the present book. A deep insight into this subject can be found in specialized comprehensive monographs [12, 13]. PL spectra of CdxZn1−xS NPs. The position of fundamental band edge of CdxZn1−xS NPs synthesized in aqueous solutions depends only slightly on the nature of a stabilizer used and remains unchanged upon the NP introduction into the polymer films [2, 3]. This allows to compare the results of PL spectroscopic studies of both colloidal CdxZn1−xS and the polymer films with incorporated cadmium-zinc sulfide NPs which are incomparably more stable than the original colloidal solutions. The PL spectra of such films reveal broad emission bands (Fig. 6.10a, left part) shifted considerably to lower energies (by 0.5–1.2 eV) as compared to the corresponding absorption bands indicating unambiguously on the defect-related nature of such emission [14, 15]. A comparatively large characteristic PL decay time, around 100 ns, also attests to the participation of “deep” (in terms of energy relative to the edges of CB and VB) charge traps in the radiative electron-hole recombination [16, 17]. The trap

Fig. 6.10 a Absorption (curves 1a–5a) and PL spectra (curves 1b–5b) of gelatin films with CdxZn1−xS NPs. x = 0 (curves 1), 0.25 (2), 0.50 (3), 0.75 (4), and 1 (5), PL was excited by hvexc = 2.81 eV; b PL energy in the maxima of D1 and D2 bands registered under the illumination with hvexc = 2.81 eV (filled circles) and 3.82 eV (hollow circles) 6.3 Luminescence Spectroscopy as a Tool for the Studies … 331 states reside deep in the bandgap and typically arise from the point defects on the NP lattice and the undercoordinated surface atoms [16, 17]. The PL band shape of CdxZn1−xS can be approximated by a combination of two Gaussian profiles with the maxima at 2.16–2.36 eV (denoted as D1, Table 6.2) and 1.85–2.00 eV (D2). The distance between D1 and D2 almost does not depend on x indicating that PL is of the same origin for all the NP compositions. As the molar fraction of zinc is increased in CdxZn1−xS NPs the PL bands show a blue shift, however, the shift magnitude is much lower than the corresponding shift of the absorption band edge. Simultaneously, the gap between the PL band maximum and the absorption band edge (the Stokes shift) increases by a factor of 2 and more as the NP composition is varied from CdS to ZnS. At least two alternative mechanisms of the defect-related PL in CdS and ZnS NPs [15, 18–21], and CdxZn1−xS NPs [22, 23] are proposed incurring the radiative À recombination either between a deeply trapped electron (etr ) with a free or shal- + þ lowly trapped hole h (Fig. 6.11, route 1), or between a deeply trapped hole (htr ) and a free or shallowly trapped electron e− (Fig. 6.11, route 2). In the latter case, a considerable increase of the PL energy hvem should be expected with an increase in the molar Zn(II) fraction in CdxZn1−xS NPs because the ECB increases by around 1.0 eV with x varied from 1 (CdS) to zero (ZnS). Alternatively, in the case of the radiative recombination with the participation of free or shallowly trapped holes (route 1) hvem is expected to depend only slightly on the NP composition, because EVB increases by mere 0.2 eV from CdS to ZnS. Table 6.2 shows that the PL maxima shifts do not exceed 0.2 eV with a variation of x from 1.0 to 0, indicating that the PL results from the radiative recombination with the participation of deeply À trapped electrons etr . À Additional arguments in favor of the involvement of etr in the PL emission can be derived from the size-dependence of the PL band position [3]. The effective hole mass is much higher than the effective electron mass for both ZnS and CdS and, therefore, the CB level should be affected by the quantum size effects in a much stronger way than the VB level. In other words, ECB would increase faster than EVB as the NP size is decreased and we can expect a much stronger size dependence for the PL route 2 with the participation of free e− than for the route 2 with the participation of free holes.

Table 6.2 Bandgap (Eg), PL xEg,eV PL FWHM, eV I2/I1 s,ns bands maxima (D1, D2), maximum, FWHM, relative intensity (I2/ eV I1), and the characteristic PL D1 D2 D1 D2 decay time (s)ofCdxZn1−xS NPs incorporated into gelatin 1.00 2.6 2.16 1.85 0.34 0.51 4.57 115 ﬁlms [3] 0.75 2.7 2.18 1.88 0.35 0.52 3.89 130 0.50 2.9 2.22 1.92 0.38 0.55 2.26 140 0.25 3.2 2.26 1.97 0.39 0.59 2.13 155 0.00 3.6 2.36 2.00 0.49 0.64 0.27 – Note accuracy of s determination is ±10 ns 332 6 Probing with Light—Optical Methods in Studies …

Fig. 6.11 Schematic of possible routes of defect-related radiative electron-hole recombination in semiconductor NPs

The positions of D1 and D2 bands were found to be unchanged as the average size of CdS NPs was decreased from 5 to 4 nm. Additionally, the size-dependent PL data can be obtained by probing the NP ensemble with a light of different energy, that is, by selectively exciting smaller or larger NPs within the NP ensemble. Figure 6.10b shows the results of such measurement for two different excitation energies hvexc = 2.81 eV (441.7 nm) and 3.82 eV (325.0 nm) for composition-selected 6–8nmCdxZn1−xS NPs. It is clear, that the variation in D1 position does not exceed 50 meV for the case of the whole NP ensemble excitation at 3.82 eV and for the case of the selective excitation of a fraction of larger NPs at 2.81 eV. This variation decreases as x is decreased and the exciton conﬁnement becomes less strong. The position of D2 does not depend on hvexc in the whole compositional range allowing to assign this PL band to surface defect states because the I2/I1 intensity ratio is higher for hvexc = 3.82 eV when smaller NPs are involved in the photoexcitation and PL emission. The above-discussed observations show that the PL of CdxZn1−xS NPs is, most probably, the result of the radiative recombination between a deeply trapped electron and a free/shallowly trapped hole [3]. This interpretation is corroborated by the reported depths of electron trap levels in cadmium and zinc sulﬁdes, that vary from 0.7–1.0 eV for ZnS [18, 20, 21, 24] to 0.4–0.7 eV for CdS [15, 21, 24]. Figure 6.12 illustrates the primary photophysical processes for individual CdS and ZnS NPs that can occur for any mixed CdxZn1−xS NPs in the frames of the proposed mechanistic interpretation. The defect-related states corresponding to D1 and D2 are presented as narrow bands with a width equal to the FWHM (full width on half-maximum) of the respective PL bands (see Table 6.2). This scheme is valid for a low density of the photogenerated charge when a possible change of the double electric layer on the NP surface due to the electrons accumulation can be neglected. Table 6.2 shows FWHMs of the D1 and D2 band for the composition-selected CdxZn1−xS NPs. The bandwidth depends mostly on an energy distribution of the trap states because a variation of the average NP size from 5 to 8 nm was not 6.3 Luminescence Spectroscopy as a Tool for the Studies … 333

Fig. 6.12 Possible routes of the defect-related radiative recombination in CdS and ZnS NPs with the energies of corresponding electron transitions [3]

accompanied by any appreciable variation in FWHM. The spectral width of D2 band is almost twice as high as that of the D1 band indicating that the latter is, most probably, associated with the radiative recombination on the NP surface defects having a broader energy distribution as compared to the volume NP defects. An Ã increase in FWHM can also be expected for the electron transition D2 between the deep electron and hole traps (Fig. 6.12). As the molar Cd(II) fraction is increased both total PL intensity and the relative contribution of the D2 band (I2/I1 ratio in Table 6.2) increase as well. The following interpretation of this tendency was suggested in [3]. The spatial exciton confinement becomes stronger as x is increased (due to an increase in aB from ZnS to CdS) resulting in an acceleration of the radiative recombination. This tendency is also favored by a decrease of the electron trap depth making stronger the overlapping between the wavefunctions of trapped electrons and holes. At that, the PL decay time is reduced (see below) indicating an acceleration of the non-radiative recombination as well. The latter factor limits the increase of PL intensity. Both the probability of the non-radiative recombination and the radiative process D2 increase 334 6 Probing with Light—Optical Methods in Studies … with x and, correspondingly, with aB, as a result of more probable migration of the photogenerated charge carriers onto the NP surface [16, 17]. The kinetic PL decay curves of the composition-selected CdxZn1−xS NPs almost coincide with the PL decay curve for the pure gelatin in the range of 0–20 ns (Fig. 6.13). The longer components of PL decay can be extracted by using a linear combination of several single-exponential functions (16)[16, 17]: X IðtÞ¼ Ai expðÀt=siÞ; ð16Þ where i =1–4, the amplitude Ai and the lifetime si are the fitting parameters. The approximation revealed three “fast” components common for all the studied films and related to the polymer (s1 = 1.0 ns, s2 = 5.0 ns, s3 = 10 ns). The presence of a fourth component s4 = 120–150 ns (Table 6.2) is typical only for the NP-containing films and is, therefore, related to the radiative recombination in semiconductor NPs. The life time value varies only slightly in a broad range of registration wavelengths (550–750 nm) indicating the recombination occurs via the D1 and D2 channels with a comparable rate. PL spectroscopy in the interpretation of the photocatalytic MV2+ reduction kinetics. The photocatalytic reduction of viologens—4,4/-derivatives of bipyridyl occupies a special place in the semiconductor photocatalysis. The most renowned of viologens—bication of 4,4/-dimethylbipyridyl or methylviologen (MV2+) (Fig. 6.14a) is soluble in water and can strongly adsorb on the surface of various semiconductors serving as an excellent electron relay between the semiconductor crystals and other components of the photochemical system. The one-electron reduction of MV2+ results in the formation of intensely colored cation-radical MV•+ that is ideal for the spectrophotometric detection even at concentrations as low as 10−7–10−8 M. The standard redox potential of the MV2+/MV•+ couple is independent of the solution pH and has a very favorable position (E0 = −0.44 V vs.

Fig. 6.13 Kinetic PL decay curves of composition-selected CdxZn1 −xS NPs in gelatin ﬁlms, PL excitation—at 3.82 eV, PL registration—at 1.81 eV 6.3 Luminescence Spectroscopy as a Tool for the Studies … 335

Fig. 6.14 Structure of methylviologen cation (a) and kinetic curves (b) of the photocatalytic reduction of MV2+ by sodium sulﬁte in the presence of CdS NPs (curve 1), Cd0.25Zn0.75S NPs (curve 2), and CdSe NPs (curve 3) [25]

NHE) between ECB of many photo-active semiconductors and typical substrates of the photocatalytic reactions (such as water or typical sacrificial electron donors). The photocatalytic reduction of MV2+ was one of pioneer reactions studied for the colloidal semiconductor NPs. The methylviologen appeared to be a very convenient substrate for the studies of quantum size effects in the nano-photocatalysts because the quantum yield of MV2+ photoreduction typically can change quite strongly with small variations of the CB level position induced by the spatial exciton confinement in the semiconductor NPs. The kinetics of MV2+ photoreduction and the character of PL quenching of semiconductor NPs by methylviologen ions often bear important information on the structure, lattice imperfection and surface charge of nano-photocatalysts as well as on the dynamics of interfacial charge transfers with the participation of semiconductor NPs. In this section, we focus on special features of the kinetics of MV2+ reduction on the surface of CdxZn1−xS and CdSe NPs that were interpreted using the PL spectroscopy [25, 26]. In a similar way, the PL spectroscopy can be applied to the studies of other nanocrystalline semiconductors and systems. Figure 6.14b shows the kinetic curves of MV•+ accumulation during the photocatalytic reduction of MV2+ by sodium sulfite (a sacrificial donor) in the presence of CdS, Cd0.25Zn0.75S, and CdSe NPs. In the case of CdSe NPs, the process rate is much higher on an initial stage as compared to CdS and Cd0.25Zn0.75S NPs, however, it decreases gradually as the reaction proceeds and finally MV•+ starts to disappear in some secondary reaction. At the same time, the MV•+ concentration remains constant in the “dark” conditions and in the absence of CdSe NPs, the facts indicating that the consumption MV•+ is the result of a photocatalytic process, similarly to the main reaction of the MV2+ photoreduction. The photochemical reduction of MV2+ by sodium sulfite without any photocatalysts also reveals a “plateau” on the kinetic curves [25] due to the secondary one-electron reduction of MV•+ to the colorless MV0 as a result of the direct photoexcitation of the dimerized MV•+ [26]: 336 6 Probing with Light—Optical Methods in Studies …

þ þ 2MV , (MV Þ2

þ 2 þ 0 ðMV Þ2 þ hv ! MV þ MV

The similar conversion of methylviologen cation-radical into the neutral MV0 was reported for the photocatalytic reduction of MV2+ with the participation of Ru 3+ •+ (bpy)3 complex [27]. These data allow concluding that MV disparition in the photocatalytic system based on CdSe NPs also results from the photocatalytic methylviologen cation-radical reduction to MV0. This conclusion leads to a question—why MV•+ is not reduced to MV0 in similar photocatalytic systems based on CdS and Cd0.25Zn0.75S NPs? A diagram presented in Fig. 6.15 shows that the photogenerated CB electrons of both CdS and CdSe NPs have a potential high enough to reduce MV2+ to MV•+. However, the second process of MV•+ reduction to MV0 can occur only in the case of CdSe NPs with ECB = −1.40 V (NHE). Similar kinetic differences of MV•+ accumulation can be observed for another couple of photocatalysts—CdSe and Cd0.25Zn0.75S (Fig. 6.14b, curves 2, 3) that have close ECB, respectively, −1.40 and −1.45 V. Here the CB potential of both nano-photocatalysts is high enough to reduce MV•+ to MV0, however, this process occurs only in the case of cadmium selenide NPs. To interpret these kinetic differences, a conventional energy scheme of the photocatalytic system that accounts only for the potentials of free CB electrons and VB holes, should be re-considered and alternative ways of the electron transfer can be taken into account, in particular, with the participation of deep electron traps on the NP surface associated with the corresponding sub-bandgap levels (Fig. 6.15) [25]. The exact position of such sub-bandgap states can be derived from the PL spectra of CdSe and Cd0.25Zn0.75S NPs. Such a scheme describes in a much more comprehensive way possible electron transfers with the participation of nano-photocatalysts abundant with various surface defects. The photogenerated electrons which have escaped the radiationless recombination and the interfacial transfer to MV2+ have the option to be trapped by the surface states. The trapping results in a lowering of the electron energy (chemical potential) by around 0.2 eV in the case of CdSe NPs, but still, the trapped electron can be transferred to the adsorbed methylviologen cation-radical. For cadmium-zinc sulﬁde the trapping-associated loss is much higher reaching up to 1.2 eV and making impossible the following electron transfer to MV•+. The presented example shows that the interpretation and prediction of the photocatalytic properties of nanocrystalline semiconductors require a combined analysis of the thermodynamic characteristics of a semiconductor, in particular, the ECB and EVB levels, together with the spectral data on the nature, concentration and energy of the surface defects, which act as the traps of the photogenerated charge carriers and contribute greatly to the photochemical behavior of semiconductor NPs. 6.3 Luminescence Spectroscopy as a Tool for the Studies … 337

Fig. 6.15 Scheme of possible electron transitions in the photocatalytic systems comprising CdS, 2+ CdSe, or Cd0.25Zn0.75S NPs, MV , and Na2SO3. Reprinted with permissions from Ref. [25]. Copyright (2010) Elsevier. hvPL is a defect-related PL quantum. The sub-bandgap levels are presented as the bands with a center and width corresponding to the maximum and FWHM of the PL bands of semiconductor NPs

PL spectroscopy in the studies of Cu–In–SNP-sensitized PEC solar cells. The chalcopyrite copper indium sulfide (CIS) NPs can be prepared with large variations in the stoichiometry while preserving the chalcopyrite structure with no additional binary phases [28–31]. Both stoichiometric CuInS2 and non-stoichiometric Cu–In– S NPs are photochemically stable and reveal broad absorption bands extending to around 800 nm (for the bulk CuInS2 Eg = 1.5 eV) and quite high absorption coefficients exceeding 105 cm−1 [28–30]. A combination of these features makes CIS NPs a good candidate as a visible light harvester both for the solid-state photovoltaic solar cells and for the liquid-junction photoelectrochemical (PEC) solar cells [28–34]. The CIS and core/shell CIS/ZnS NPs used as spectral sensitizers of the TiO2- based photoanodes in the PEC solar cells showed a remarkable correspondence between the PEC activity and PL properties [35]. In particular, at a constant Zn:Cu ratio and at a variation of the copper and indium content the PL intensity of core/shell CIS/ZnS NPs in solutions follows the PEC activity of the TiO2/CIS/ZnS heterostructures produced from such colloids. The data reported in [35]isan example of the application of the PL spectroscopy as a diagnostic tool for the investigations of light-harvesting PEC systems. Similarly to the broadly studied cadmium chalcogenide NPs, the formation of a ZnS shell on the surface of CIS NPs results in a drastic enhancement of the PL intensity as well as in a considerable growth of the PEC activity of NPs coupled to 338 6 Probing with Light—Optical Methods in Studies … titanium dioxide. In particular, an increase in the molar Zn:Cu ratio from 0 (no zinc) to 20 resulted in a more than the 30-fold increment in the PL efficiency (Fig. 6.16a, curve 1). The photocurrent density measured for the TiO2/CIS/ZnS photoanodes based on such NPs was found to increase from around 2.0 mA/cm2 for the uncovered “core” Cu–In–S NPs to 2.5 mA/cm2 for the core/shell CIS/ZnS NPs with a Zn:Cu ratio of 1:1 (Fig. 6.16a, curve 2). As opposite to the PL properties, a further increase in the Zn content results in the deterioration of PEC activity of CIS/ZnS NPs and for the thickest ZnS shell and the highest PL efficiency, at Zn:Cu = 20:1, the photocurrent density is three times lower than for the original core CIS NPs. Apart from the PL enhancement, the doping of Cu–In–S NPs with zinc(II) occurs resulting in a blue shift of the absorption band edge and a con- comitant shift of the PL band maximum. These phenomena are well reported and caused by the penetration of Zn2+ ions into the Cu–In–S core resulting in the lattice reconstruction and band gap widening [30, 34, 36]. The dependencies presented in Fig. 6.16a can be understood by assuming that the deposition of a ZnS shell results in the elimination of surface structural defects acting as the non-radiative recombination sites. At that, the competing processes of the radiative electron-hole recombination in CIS/ZnS NPs and the interfacial electron transfer in the TiO2/CIS/ZnS heterostructures become much more efficient resulting in the PL enhancement and the photocurrent increase. However, as the ZnS shell becomes thicker it impedes the interfacial electron transfer and promotes at the same time the electron-hole recombination in the CIS core. Therefore, at a higher ZnS content, we observe a sharp decrease of the PEC activity of the TiO2/ CIS/ZnS heterostructures with a steady growth of the PL intensity of core/shell CIS/ZnS NPs used for the preparation of the photoanodes.

Fig. 6.16 a Photocurrent density (curve 1) obtained for the TiO2/CIS/ZnS photoanodes and PL intensity of CIS/ZnS NPs (curve 2) as a function of the molar Zn:Cu ratio during the ZnS shell deposition. b Photocurrent density (curves 1, 2) for the TiO2/CIS (curve 1) and TiO2/CISZnS (curve 2) photoanodes and PL intensity of CIS/ZnS NPs (curve 3) as a function of xCu [35] 6.3 Luminescence Spectroscopy as a Tool for the Studies … 339

As can be seen from Fig. 6.16a, the PEC measurements showed that the peak photocurrent density is achieved at a comparatively low Zn(II) content, that is, for a relatively thin ZnS shell on the surface of CIS NPs. At a Zn:Cu ratio of 1:1 (optimal for the solar cell performance) the absorption spectra of uncovered and core/shell NPs are essentially identical) indicating that the Zn(II) doping effect is small to negligible and can be ignored in such conditions. An increase of copper content in the CIS NPs results in a gradual transformation of the structure of the fundamental absorption band of colloidal NPs and a general growth of the absorbance in the UV and visible spectral range, but the absorption spectra of uncovered CIS and core/shell CIS/ZnS NPs were almost identical for each given xCu. Both types of NPs, the core CIS and core/shell CIS/ZnS, revealed also similar dependences of the PEC activity on the composition (Fig. 6.16b). The photocurrent density generated by the TiO2/CIS/ZnS photoanodes increases as the xCu is changed from 0.2 to 1.0, then it peaks at a NP composition corresponding to Cu:In:S = 1:5:10 and decreases at xCu = 2.5 almost to the same value as for the NPs with the minimal copper content. For the most active absorber NPs with xCu = 1.0, the increment of the PEC activity introduced by the ZnS shell deposition reaches around 25% (Fig. 6.16b, curves 1, 2). The dependence of the integral PL intensity of Cu–In–S/ZnS NPs on the xCu follows closely the relationship between the PEC activity and the copper content and reaches the maximal value at xCu = 1.0 as well, dropping quite considerably at a higher copper content (Fig. 6.16b, curve 3). The PL intensity is, therefore, a reliable indicator allowing to anticipate the PEC activity of colloidal CIS NPs basing on the spectral PL data.

6.4 Studies of Nanocrystalline Semiconductors by the Time-Resolved Photolysis Techniques

The flash photolysis allows generating a relatively high amount of photoexcited molecules and short-lived intermediates including radicals during a short light pulse, allowing for their spectral identification and studies of the decay kinetics. This feature is a basic distinction between the flash photolysis and the conventional stationary methods used to study the photochemical processes. The present section aims to exemplify the potential of the lamp and laser flash photolysis on several systems with the semiconductor metal sulfide NPs and more complex nanoheterostructures. It shows the versatility of the pulse photolysis in the studies on the accumulation of the excessive charge in the semiconductor NPs, the dynamics of interfacial transfer of the excessive electrons, and the photochemical transformations of NPs under the powerful photoexcitation. The metal sulfide NPs are unstable in the charged state and prone to the reductive photocorrosion and therefore, the photocharging events cannot be studied for such NPs in the regime of stationary illumination. 340 6 Probing with Light—Optical Methods in Studies …

Pulse photoexcitation of CdxZn1−xS NPs and discharging dynamics: a case of relatively weak lamp pulse photoexcitation. The pulse photoexcitation of aqueous colloidal SPP-stabilized CdxZn1−xS NPs induces a reversible blue shift of the fundamental absorption band edge that can be conveniently observed as a non-stationary bleaching (NB) band in the differential absorption spectra (Fig. 6.17a), as discussed earlier in Chap. 1. The shift is induced by the accumulation of excessive electrons near the CB edge resulting in a complete filling of the available states and an increase in the energy necessary for the photogeneration of new CB electrons—the so-called dynamic Burstein-Moss effect. The NB band intensity of cadmium and cadmium-zinc sulfide NPs is proportional to the excessive charge density and decreases by a factor of *3 in the first 1 ms after the light pulse (Fig. 6.17b) reaching zero in 3–5 ms after the photoexcitation. In aerated aqueous solutions with no other donor/acceptor present the relaxation of the NB band indicates the capture of the excessive electrons by oxygen and water molecules. Figure 6.17a shows that the NB band intensity of the mixed CdxZn1−xS NPs is 5–6 times higher than for CdS NPs indicating that the mixed NPs have a more pronounced capability of accumulating the excessive charge. Such a difference can originate from multiple reasons. As the molar Cd(II) fraction is decreased the PL intensity of CdxZn1−xS drops as well resulting in an increase of the number of long-lived electrons capable of participating in the chemical reactions on the microsecond time scale. Additionally, the FWHM of D1 and D2 PL bands of CdxZn1−xS NPs (that is, the spectrum of possible mid-bandgap electron states) as well as the trap depth increase as the Zn content is elevated (Table 6.2) favoring to the accumulation and retaining of a higher excessive negative charge under the pulse photoexcitation. The kinetic NB decay curves can be fitted with linear combinations of two single-exponential functions (Fig. 6.17b, solid line), similarly to the

Fig. 6.17 a Non-stationary bleaching spectra of colloidal Cd0.67Zn0.33S NPs (curve 1) and CdS NPs (curve 2); b Kinetic curve of the NB decay registered for Cd0.67Zn0.33S NPs at kreg = 440 nm. The hollow circles are experimental data, while the solid line represents a two-exponential fitting 6.4 Studies of Nanocrystalline Semiconductors by the Time-Resolved … 341 above-discussed PL bands of CdxZn1−xS NPs. The fact may be in indication on at least two independent routes of the excessive electrons consumption. The characteristic lifetime of the first channel is 100–150 ls, while for the second one it is higher than 650–700 ls. A comparison of the results of the flash photolysis and the PL spectroscopy allowed to conclude [3] that the two processes under discussion involve the electrons trapped by D1- and D2-type states, respectively. The faster process with t1 = 100–150 ls can be associated with the consumption of electrons trapped by the surface D2 traps more easily accessible for the acceptors, while the slower process with t2 > 650 ls can be assigned to the acceptor reactions with electrons trapped by the volume states D1. Table 6.3 shows the characteristics NB decay lifetimes t1 and t2 at a different registration wavelength (kreg) for CdxZn1−xS with x = 0.67 and 0.40. In both cases, t2 increases at longer kreg, while t1 remains more or less constant. As a variation of the NB registration wavelength allows to probe selectively the CdxZn1−xSNP fractions of a different size, the observed “t2 − kreg” dependence can be interpreted À as a result of the size-dependence of the rate of etr migration to the NP surface. In the presence of electron-donating Na2SO3, the NB signal intensity of CdxZn1−xS NPs increases in a linear manner with the donor concentration (Fig. 6.18a, curve 1). At the same time, the concentrational dependence for another donor—Na2S revealed a maximum (Fig. 6.18a, curve 2) caused by the continuous photogeneration of poly- fi 2À sul de anions Sn capable of capturing the excessive electrons of CdS NPs [37, 38]. In the presence of intentionally introduced polysulfide, even at a comparatively small 2À * −4 Sn concentration ( 1 Â 10 M) a considerable quenching of the NB signal is fi 2À − 2− observed. The possibility of disul de reduction (S2 +2e =2S ) is additionally confirmed by the comparison of the standard potential of this reaction (−0.52 V vs. NHE) with the conduction band potential of CdS NPs (−0.8 V, NHE). 2À The systems, where Sx is used to capture and shuttle the photogenerated charge carriers, occupy an important place in the solar light harvesting with nanocrystalline semiconductors. As shown in Chap. 4, the sulfide/polysulfide redox shuttle is one of the most efficient and popular in the electrolytes of the semiconductor NPs-sensitized solar cells. Despite the broad studies into the factors affecting the light conversion efficiency and photocurrent generation mechanisms in such sys- 2À tems, the primary light-induced charge transfer between CdS NPs and Sx species, as well as the nature and fate of short-lived intermediates of these processes still require a deeper understanding. As mentioned earlier, the introduction of polysulfide into the colloidal CdS NP solutions results in the quenching of NB bands, and a new broad transient absorption band can be observed with a smeared maximum at 580–590 nm (Fig. 6.18b).

Table 6.3 Characteristic TB kreg, nm 425 430 435 440 decay lifetimes t1 and t2 for t1, ls x = 0.40 140 130 130 140 CdxZn1−xS NPs at different kreg [3] x = 0.67 110 110 100 120 t2, ls x = 0.40 910 940 970 1780 x = 0.67 640 720 940 1570 342 6 Probing with Light—Optical Methods in Studies …

Fig. 6.18 a Dependences of the NB signal intensity for 6–8 nm CdS NPs on the concentration of Na2SO3 (curve 1) and Na2S (curve 2), b Transient differential absorption spectra of colloidal CdS −3 solution containing 1 Â 10 MNa2S registered consecutively one after the other (curves 1–3). A total number of pulses in the registration range of 410–690 nm (10 pulses per point in average) is 280 (1), 560 (2), and 840 (3). Curve 4 obtained for CdS solution with a sodium polysulﬁde − addition. [CdS] = 1 Â 10 3 M[37, 160]

2À It is well reported that a relative content of various Sx species (with a different x) depends primarily on the ratio of S2− and S0 during the polysulﬁde formation, as well – 2À – as on the solution pH [39 41]. The Sx distribution diagrams presented in [39 41] show that for [S2−]:[S0] = 1 (the conditions of the experiments under discussion 2À 2À here) the solutions contain predominantly S3 and S4 anions in a ratio close to unity. The analysis of reported literature data [42–46] allowed to assign the positive k – À transient band peaked at max = 580 590 nm to S3 anion-radical. Other radicals that can potentially form in given experimental conditions absorb in a different À k – •—k – spectral range, for example, O2 at max =240 245 nm [47], HS max = 290 •− •− 330 nm [48], free S —kmax = 260 nm [47], S bound on the CdS surface and •− depending on the NP size—kmax = 450–500 nm [48–50], H2S2 —kmax = 380 nm À—k À—k [47], S2 max = 400 nm [42, 43, 46], SO2 max = 365 nm [47], À—k S4 max = 513 nm [46]. À The most probable way of S3 generation in the presence of CdS NPs is the oxidation of polysulﬁde species by the photogenerated CdS valence band holes:

þ 2À À hVB þ S3 ! S3

As the sulfide ions get oxidized during the pulse photolysis to the elemental 2− fi 2À sulfur, the latter interacts with the excess of S producing polysul de and the Sx 6.4 Studies of Nanocrystalline Semiconductors by the Time-Resolved … 343 concentration gradually increases during the flash photolysis experiment as evidenced by a gradual increase of the transient band intensity (Fig. 6.18b). 2À The decay kinetics of the Sx pulse photolysis products cannot be described either by the first-order or second-order models indicating a complex decay À mechanism with several possible reaction of S3 , for example the recombination (second-order reaction) [45] or a reaction with the molecular oxygen [47]. As oxygen is present in a concentration at least by 4 orders higher than the intermediate concentration this reaction is of the pseudo-first kinetic order.

À 0 À S3 +O2 ! S3 +O2

À À 2À S3 +S3 ! S6

À – The characteristic lifetime of S3 produced in the presence of CdS NPs is 2 3 times higher than in homogeneous solutions by the direct photoexcitation of fi À polysul de, most probably due to the S3 adsorption on the NP surface and the inhibition of both decay processes. Pulse photoexcitation of CdxZn1−xS NPs and discharging dynamics: a case of relatively strong laser pulse photoexcitation. Under the photoexcitation with laser pulses with a photon density at least by an order of magnitude higher than in the case of lamp flash photolysis, CdxZn1−xS NPs show some special features due to the probability of the formation of several electron-hole couples in a sole NPs within the time scope of a laser pulse. Illumination of colloidal CdxZn1−xS with the laser pulses at k = 355 nm results in spectral changes similar to those observed in the case of lamp photolysis—a Burstein blue shift of the absorption band edge and a rise of the NB band in the differential absorption spectra indicating the accumulation of an excessive charge. The NB band relaxes during tens-hundreds microseconds (Fig. 6.19a) owing to the reactions between excessive electrons and the solution components. As x is reduced from 1.0 to 0.2 the NB band maximum shifts to shorter wavelengths from 460 to 375 nm (Fig. 6.19b) mimicking the corresponding shift of the fundamental absorption band edge of CdxZn1−xS NPs [51, 52]. When oxygen is bubbled through the colloidal solution the intensity of NB band decreases indicating on the electron scavenging by oxygen molecules [52]: À À e þ O2 ! O2 (Fig. 6.19a). The oxygen exerts only a partial NB quenching and, therefore, the above reaction cannot be the sole or even the main fate of the excessive electrons generated by the laser pulse [52]. Taking into account the photocatalytic properties of CdxZn1−xS in the reduction of water to hydrogen (discussed in Chap. 2), the reduction of water by excessive electrons can be regarded as a principal mechanism of the NB decay in the case of cadmium-zinc − − sulfide NPs [52]: 2e +2H2O ! H2 + 2OH . Therefore, by studying the NB decay dynamics we can derive important information on the mechanism and lim- itating factors of the photocatalytic water reduction. 344 6 Probing with Light—Optical Methods in Studies …

Fig. 6.19 a Kinetic curves of NB decay registered in the NB band maximum in N2-saturated (curve 1) and O2-saturated (curve 2) colloidal CdS solutions. b, c Positions of the NB band maximum (b) and the NB band intensity (c) of colloidal CdxZn1−xS of a different composition [52]

Another special feature of the pulse-excited CdxZn1–xS NPs is a volcano-shaped dependence between the NP composition (x) and the NB band intensity (Fig. 6.19c). As x is decreased from 1.0 to around 0.6 the NB band intensity increases, reaches a peak value at x = 0.6–0.7, and decreases at a further increase of the Zn(II) fraction in CdxZn1−xS NPs down to zero at x = 0.2. The interpretation of the dome-shaped dependence presented in Fig. 6.19c was performed [52] basing on a model introduced for the oxygen one-electron reduction process with the participation of the radiolytically-charged Ag NPs [53]. According to the model, the NP discharging rate (decay of a charge Q) in this reaction depends on the oxygen reduction over-voltage DE and can be expressed as

dQ ÀaFDE À ¼ ke RT ; ð17Þ dt where a is a constant, F is the Faraday number, R is the universal gas constant, T is temperature (К). The over-voltage can be expressed as DE = –Q/C, where C is the electric capacitance of semiconductor NPs, Q is a charge per a NP at a moment t.By integrating Eq. (17) and a logarithmic transformation, Eq. (18) can be derived [52]: RTC a Q ¼ lnð ÞÀln t : ð18Þ aF RTCk

A dependence between the excessive voltage Q and the optical density of NB band can be expressed as Q = bDD, where b is a constant [54]. Therefore,

DD ¼ A À B ln t; ð19Þ 6.4 Studies of Nanocrystalline Semiconductors by the Time-Resolved … 345 where

RTC a RTC A ¼ lnð Þ and B ¼ : ð20Þ aFb RTCk aFb

Figure 6.20a shows that there indeed exists a linear dependence between DD and ln(t), which is in accordance with the experimental results. The tangent of such linear dependence and, therefore, the NB decay curve shape appear to be determined mostly by the electric capacitance of the semiconductor NPs. By making reasonable assumptions on the values of a and b [52] the capacitance of a CdxZn1−xS NP of a given composition can be calculated using Eq. (20). In particular, the a = 0.3 can be taken by analogy with the photocatalytic reduction of MV2+ in the presence of SPP-stabilized CdS NPs reported in [55], while b = 0.3 was calculated in [52] from a dependence of the NB band intensity on the laser − pulse intensity. The NP capacitance C was found to be 0.014 F Â L 1 for CdS NPs, −1 −1 0.043 F Â L for CdxZn1−xS NPs with x = 0.8 and 0.068 F Â L —with x = 0.5. The laser ﬂash photolysis experiments reported in [52] were performed using CdxZn1−xS NPs with an equal average size of 6.0 ± 0.5 nm regardless of their composition. The electric capacitance per NP estimated using the size value was − − − found to be 0.2 Â 10 18 F(x = 1.0), 0.7 Â 10 18 F(x = 0.8), and 1.1 Â 10 18 F (x = 0.5), which is close to the lower limit of the reported range of the capacitance − of colloidal 10 nm CdS NPs determined by other methods, (6–60) Â 10 18 F[54]. An increase in the electric capacitance of CdxZn1−xS NPs with increasing x originates, most probably, from a disordering of the mixed sulﬁde lattice caused by the difference in the ionic radii of Cd2+ and Zn2+ and a different rate of the CdS and ZnS formation during the co-precipitation. The lattice defects generated at the synthesis can act as charge traps prohibiting free migration of the photogenerated

Fig. 6.20 a Kinetic NB decay curves registered in the NB band maxima of CdS NPs (curve 1), Cd0.8Zn0.2S NPs (curve 2), and Cd0.5Zn0.5S (curve 3) presented as DD(t) versus –ln(t)[52]. b, c NB decay curves of CdS NPs (b) and Cd0.8Zn0.2S NPs (c) in the corresponding NB band maxima (480 nm in (a) and 460 nm in (b)) registered without additions (curves 1) and in the presence of ZnSO4 (curves 2) and Na2S (curves 3). Insert in (b) presents kinetic curves as DD(t) vs. –ln(t), solid lines correspond to the linear fits of the experimental data. Insert in (c) presents PL spectra of CdS NPs (curve 1), Cd0.5Zn0.5S NPs (curve 2), and Cd0.2Zn0.8S NPs (curve 3) [52] 346 6 Probing with Light—Optical Methods in Studies … charge carriers and their recombination. Another possible reason for the x-dependence of the capacitance of CdxZn1−xS NPs can be a deviation from the feed stoichiometry set at the synthesis (a ratio of Cd(II) and Zn(II) concentrations) [2] due to the formation of various hydroxo-complexes of zinc that react much slowly with sulfide anions. These forms of Zn(II) can adsorb on the NP surface and also act as traps capturing the photogenerated electrons and thus increasing the overall NP capacitance. The above assumptions are corroborated by the results of the flash photolysis of CdS and Cd0.8Zn0.2S NPs in the presence of intentional additions of 2+ over-stoichiometric ZnSO4 and Na2S[52]. Figure 6.20b shows that a Zn addition only slightly affects the NB band intensity DD0, but results in an increase of the tangent of the DD—ln(t) dependence (insert) indicating an increase of the NP capacitance (see Eq. 19). At the same time, an addition of Na2S results in a considerable reduction of DD0 and the NP capacitance. A similar effect of both additions was observed for the Cd0.8Zn0.2S NPs (Fig. 6.20c), but in this case the introduction of Zn(II) decreases the NB band intensity and this effect becomes more pronounced with a further decrease of x. A deviation from the non-stoichiometry of CdxZn1−xS NPs results also in a drastic increase in the PL intensity as x is decreased from 1.0 to 0.2 (Fig. 6.20c, insert), because the Zn(II) species adsorbed on the NP surface can participate both in the electron accumulation and the radiative electron-hole recombination. An increase in the charge accumulation rate will inevitably cause a decrease in the PL efficiency. In the case of polymer-incorporated CdxZn1−xS NPs (see discussion above) the polymer passivates efficiently the surface under-coordinated Zn(II) and an inverse tendency is observed [3]. The discussed results of the flash photolysis coupled with the PL spectroscopy data allowed to conclude [52] that a volcano-shaped dependence between the NB band intensity and the composition of CdxZn1−xS NPs (Fig. 6.19c) originates from the overlap of two tendencies—(i) an increase of the NP capability of accumulating the excessive charge with a decrease of x and (ii) a increase of the probability of the radiative electron-hole recombination with a decrease of x. Laser flash photolysis of TiO2/CdS film nanoheterostructures. As discussed in Chaps. 2–4, TiO2/CdS is one of the most broadly studied photo-active nanoheterostructures. The application of the flash photolysis to the studies of charge separation in this composite allowed to derive important and quite unique information on the dynamics of the photogenerated charge carriers in the TiO2/CdS heterojunction with a CdS layer produced by different methods. The fundamental band edge of nanocrystalline titania deposited on glass [56]is around 360–370 nm (Fig. 6.21a, curve 1) and, therefore, the laser pulses with k = 355 nm (3.5 eV) can excite interband electron VB–CB transitions. A differential absorption spectrum of the TiO2 films reveal a broad band in the range of 670–710 nm with a peak at k = 680–690 nm (Fig. 6.21a, insert) [57]. The band can be assigned to a long-lived intermediate as the transient signal shows no signs of decay during hundreds of ls after the pulse extinction (Fig. 6.21b). Similar bands are typically observed [58, 59] in the case of the CB electron capture by deep 6.4 Studies of Nanocrystalline Semiconductors by the Time-Resolved … 347

Fig. 6.21 a Absorption spectra of the nanocrystalline TiO2 film (curve 1) and TiO2/CdS nanoheterostructures (curves 2, 3) produced by the photocatalytic CdS deposition (2) and the CBD (3) [56]. Insert in (a): transient absorption spectrum of TiO2 film excited by the laser pulses with k = 355 nm. b Kinetic curves of the transient signal decay at k = 680 nm registered for the CBD-deposited TiO2/CdS films (curves 1, 2) and the photodeposited TiO2/CdS films (curves 3, 4). Curves 2 and 4 registered after the deposition of a sucrose layer [57]

4+ 3+ 4 þ À 3 þ traps (Ti ions) resulting in the Ti generation (Ti þ eCB ! Ti ). The VB holes are also rapidly captured by the hole traps—typically surface hydroxide þ À anions (hVB þ OH ! OH )[58] or interact with the donor compounds adsorbed on the TiO2 NP surface. A large portion of the photogenerated charge carriers decays in the recombination processes, that have a predominantly non-radiative character for the nanocrystalline titania [58]. To increase the transient Ti3+-related band intensity a thin transparent layer of sucrose was applied on top of the TiO2 films that does not interfere with the light absorption but supplies additional electrons as a sacrificial donor [60]. When nanocrystalline cadmium sulfide is deposited on the titania surface by the chemical bath deposition (CBD) a new absorption band appears with an edge at 510–520 nm (Fig. 6.21a, curve 3) while the optical density of the TiO2/CdS film at the laser wavelength (3550 nm) increases to 1.75 indicating a complete light absorption by cadmium sulfide. Such TiO2/CdS nanohetero–structure showed almost zero intensity of the transient signal at 670–710 nm (Fig. 6.21b, curve 1). The fact indicates that the efficiency of interfacial electron transfer from the pho- 3+ toexcited CdS to TiO2 followed by the electron capture and formation of Ti is very low for this TiO2/CdS composite, despite the favorable thermodynamic conditions (ECB(CdS) = −0.8 V vs. NHE [61], ECB(TiO2)=−0.3 V at pH 7 [61]). The electron transfer can be hindered by an interfacial barrier between titania and CdS, because the CBD of cadmium sulfide typically yields hexagonal CdS NPs [62], while TiO2 is crystallized in a cubic anatase modification [63]. Also, the CDB-deposited CdS NPs show drastic recombinational losses of the photogenerated charge carriers as evidenced by a large increase of the transient signal intensity 348 6 Probing with Light—Optical Methods in Studies … in the presence of sucrose capable of capturing of the photogenerated holes (Fig. 6.21b, curves 1, 2) and interfering with the recombinative processes. The flash photolysis of similar TiO2/CdS nanoheterostructures produced by the photocatalytic deposition of cadmium sulfide NPs [56] results in the same tran- sients, however, the signal intensity is much higher in the presence of sucrose in this case (Fig. 6.21b, curves 2, 4). After a correction on the light absorbance, the difference in the transient signal intensity between both nanoheterostructures increases additionally by a factor of 3. The observations indicate that the efficiency of the photoinduced interfacial electron transfer from CdS to TiO2 and the formation of Ti3+ species is by an order of magnitude higher for the photocatalytically produced TiO2/CdS nanoheterostructure as compared with the analog synthesized by the conventional CBD [56]. As opposite to the bare titania films, the TiO2/CdS nanocomposites revealed a second quite intense transient signal in the range of 420–570 nm peaked at 470– 500 nm (Fig. 6.22a). This band can be assigned to surface-adsorbed S•− radicals formed via the photogenerated VB hole capture by the deep traps (lattice S2− anions). The transient band observed for the CDB-deposited TiO2/CdS nanocomposite seems to be composed of two spectral components. However, the kinetic decay curves registered for this band on different wavelengths (460, 490, and 520 nm) are the same and, therefore, describe the decay of a single short-lived intermediate. The

Fig. 6.22 a Transient differential absorption spectra of the TiO2/CdS nanohetero–structures produced by the CBD (curve 1) and the photocatalytic CdS deposition (curve 2); b Normalized kinetic curves of the transient signal decay for the CBD-produced TiO2/CdS registered at 500 nm (curve 1) and for the photodeposited composite registered at 470 nm (curve 2) [57] 6.4 Studies of Nanocrystalline Semiconductors by the Time-Resolved … 349 band distortion can be caused by an overlap of the positive transient signal with a NB band of CdS NPs bearing excessing negative charge (as discussed in the previous section). The NB band maximum for the given TiO2/CdS nanoheterostructure is expected to be at 480–500 nm thus indeed overlapping with the absorption band of S•− anion-radical. The NB band presence indicates that the photogenerated CdS CB electrons accumulate on CdS NPs as a result of a low efficiency of the interfacial •− transfer to TiO2 NPs. A reconstruction of the S absorption band (Fig. 6.22a, dashed line) shows that the band maximum should be observed at 485–490 nm. •− The intensity of S related band of the photocatalytically-produced TiO2/CdS nanocomposites is twice as high as for the CBD-produced analog, its peak blue-shifting to 465 nm. The first observation is in accordance with an increase in the Ti3+ signal intensity at 670–710 nm, while the second one illustrates a well-reported blue shift of the adsorbed S•− band maximum with a decrease of the CdS NP size [64, 65]. The decay dynamics of the sulfur anion-radical is also different for the CBD-deposited and the photodeposited TiO2/CdS nanoheterostructures (Fig. 6.22b). The kinetic curves have a complex shape that cannot be fitted with a simple first-order or second-order kinetic model. The complexity attests to several simultaneous reactions with the participation of S•−. Also, it can arise from a size distribution of CdS NPs. Figure 6.22b shows that the CBD-deposited CdS NPs show a sharp decrease of the transient signal intensity in the first 3–5 ls after the laser pulse followed by a slower signal relaxation till the zero level (at t >50ls). At the same time, no fast component can be observed in the decay curves of the photodeposited TiO2/CdS nanocomposites (Fig. 6.22b, curve 2). The decay is generally slower and more than a half of the photogenerated S•− anion-radical survives as long as 50–100 ls after the exciting pulse. The differences in the decay curve shape cannot be explained solely •− •− 2− by possible differences in the rate of radical recombination (S +S ! S2 )or •− 0 •− the interaction with oxygen (S +O2 ! S +O2 ). Taking into account the presence of excessive electrons on CdS NPs in the CBD-deposited TiO2/CdS nanoheterostructure the different decay dynamics can be assigned to the recombination of the excessive electrons and S•− anion-radicals (Cd •− À (II)S + etr ! CdS) [56]. In the case of photoproduced TiO2/CdS nanocomposites, this process is blocked by the efficient electron transfer from CdS NPs to the titania scaffold. Formally, the latter reaction corresponds to the electron recombination with the deeply trapped hole. The recombination of free charge carriers in CdS NPs is typically over in 1–10 ns after the photoexcitation [64], but the process can be extended to 200–300 ns if one of the carriers gets captured by the deep trap [3, 64]. It can be safely assumed that the recombination between the deeply trapped electron and deeply trapped hole will occur by 1–2 orders of magnitude slower, thus corresponding to the discussed time scale. 350 6 Probing with Light—Optical Methods in Studies …

6.5 Studies of Nanocrystalline Semiconductors Using Raman Scattering Spectroscopy

The Raman spectroscopy is used much rarer for the studies of semiconductor nano-photocatalysts and photo-electrodes as compared to the electron absorption and PL spectroscopies. However, in many cases, the Raman spectroscopy provides a unique information on the nano-photocatalyst structure, especially in the studies of mixed compounds and multi-component nanostructures. For colloidal and thin-film nanocrystalline semiconductors comprising a small amount of a target phase, the resonant Raman spectroscopy appeared to be the most productive, when the samples are excited by the wavelength corresponding to the spectral range of maximal absorbance, for example, into the excitonic band maximum. For example, the Raman spectrum of CdSe NPs incorporated into the gelatin films shows no distinct semiconductor-related features under the illumination with k = 647.1 nm which is not absorbed by the NPs. At the same time, the resonant excitation at k < 550 nm, that is, into the absorption band of cadmium selenide NPs, allows to register the characteristic CdSe phonon bands. The peak positions depend on the excitation wavelength as a result of the selective photoexcitation of differently sized fractions of CdSe NPs in the incorporated NP ensemble [66]. The Raman spectrum of the nanocrystalline sample can be used for the determination of the phase composition because many photochemically active semiconductors have characteristic vibrational frequencies. For example, the main phonon mode (longitudinal optical phonon—LO) of CdS, CdSe, and CdTe can be observed at 305 (Fig. 6.23a), 210, and 170 cm−1 [67]. The characteristic LO frequencies of ZnO, ZnSe, and ZnSe are 350, 580, and 250 cm−1 (Fig. 6.23c), respectively [67]. The anatase modification of titanium dioxide has six active vibrational modes, with the most intense in the Raman spectra being at 144 cm−1

Fig. 6.23 Raman spectra of CdS NPs deposited onto ZnO surface (a)[161], nanocrystalline TiO2 ﬁlm (b, curve 1) and TiO2/Sb2S3 nanoheterostructures with amorphous (curve 2) and crystalline (curve 3) antimony sulﬁde NPs [162], and ZnSe NPs (c)[163] 6.5 Studies of Nanocrystalline Semiconductors … 351

(Fig. 6.23b, curve 1) allowing for a secure spectral distinction between the anatase and rutile [68]. The presence of LO overtones at a double (2LO) and a triple (3LO) main frequency (Fig. 6.23a) is a sign of high crystallinity and high structural order of the NP lattice. In the case of Sb2S3 NPs deposited on titania as a spectral sensitizer, the Raman spectroscopy can provide a definite proofs of amorphous (Fig. 6.23b, curve 2) of crystalline (curve 3) character of antimony sulfide, the latter revealing a much more resolved picture of possible vibrational modes of the crystalline stibnite lattice. The Raman spectra of small NPs with a developed surface area often reveal additional spectral “shoulders” shifted to lower frequencies as compared to the main LO peak (Fig. 6.23c). Such peaks are typically assigned to surface optical (SO) phonons of NPs and a ratio of LO and SO phonon peaks can be used as a measure of the surface area and/or structural disorder of the semiconductor NPs. In some cases, the Raman spectroscopy can be used to evaluate the size of semiconductor NPs. Similarly to the size determination from the optical absorption spectra, this procedure is only possible for the semiconductors with reported empirical correlations (or calculated ones) between the NP size and some of the spectral parameters of Raman spectra, such as the position or FWHM of the LO peak. For example, the size of CdSe NPs can be estimated from the spectral position of the LO peak for the size range of d <6–7 nm (Fig. 6.24a). The dependence originates from the spatial phonon confinement resulting in a gradual decrease of the LO peak frequency as the NP size is reduced [69–72]. A correlation between the FWHM of the most intense Raman peak of TiO2 NPs and the average NP size is reported (Fig. 6.24b) allowing the NP size to be estimated in the range of d <15– 20 nm.

Fig. 6.24 Correlations between the size and PL frequency of CdSe NPs (a) and between FWHM of Eg(1) phonon peak and the size of TiO2 NPs (b). Adapted with permissions from [69](a) and [164]. Copyright (1998, 2005) The American Physical Society 352 6 Probing with Light—Optical Methods in Studies …

The Raman spectroscopy, especially in the combination with the X-ray diffraction and the energy-dispersive X-ray spectroscopies, is a powerful method for the determination of the composition of mixed solid-solution semiconductor nanomaterials, such as CdxZn1−xS and CdSxSe1−x. The phonon peak positions in the Raman spectra of these compounds change in a monotonous way with the NP composition between the positions of the spectral signals of separate components. Figure 6.25 exemplifies this approach for the CdSxSe1−x solid-solution NPs [73]. By using the cadmium sulfoselenide NPs of a known composition, a calibration curve can be plotted for the composition-dependent positions of phonon frequencies (Fig. 6.25b) that can be used for the determination of the composition of CdSxSe1−x samples with unknown x. The latter approach was implemented for the determination of real composition of ZnO/CdxZn1−xS photoanodes produced by the SILAR [74]. The SILAR deposition from aqueous mixed solutions of Cd(II) and Zn(II) nitrate results in the formation of CdxZn1−xS on the ZnO film surface evidenced by a new absorption band with the edge shifting to shorter wavelength as the molar cadmium fraction, x0 = [Cd(II)]/([Cd(II)] + [Zn(II)]), was decreased (Fig. 6.26a). In particular, a decrease of x0 from 0.9 to 0.1 results in a blue shift of the absorption band edge of CdxZn1−xS from 500–505 nm (Eg =2.46–2.48 eV) to 460–465 nm (Eg = 2.67– 2.70 eV). As discussed earlier in this chapter, the size of cadmium-zinc sulfide NPs does not depend considerably on their composition and, therefore, the shift of the absorption band edge of the SILAR-deposited CdxZn1−xS NPs can be assigned exclusively to a variation of the NP composition. The estimations performed using the above-described approach and the empirical Eq. (5) showed that the real molar Cd(II) fraction in CdxZn1−xS NPs derived from the absorption spectra, xabs, is much

Fig. 6.25 a Raman spectra of colloidal CdSxSe1−x NPs at x = 0 (curve 1), 0.2 (curve 2), 0.5 (curve 3), 0.8 (curve 4), and 1.0 (curve 5). Insert: compositional dependences of the LOCdS (squares) and LOCdSe (circles) peaks of CdSxSe1−x NPs [165]. Solid lines represent similar dependences reported in [166]. Reprinted with permission from Ref. [73]. Copyright (2010) Springer. b Dependence of the LO peak position on the composition of CdxZn1−xS NPs. Plotted using data reported in [67] 6.5 Studies of Nanocrystalline Semiconductors … 353 higher than the nominal Cd(II) fraction x0 and varies from 0.94 to 0.67 as x0 is decreased from 0.9 to 0.1. Such a strong difference between x0 and xabs originates, most probably, from a different adsorption of Cd2+ and Zn2+ ions on the ZnO surface, the partial hydrolysis of Zn(II), as well as the lower solubility of CdS. Figure 6.26b–d shows the Raman spectra of ITO/ZnO/CdxZn1−xS ﬁlms excited at kexc = 325 nm (3.82 eV) and 514.5 nm (2.42 eV). The spectra reveal a peak at −1 300–320 cm that was assigned to the LO mode of CdxZn1−xS NPs. The low-frequency wing of the peak at 250–270 cm−1 may be ascribed to the light scattering on the SO phonons as discussed earlier for ZnSe NPs. A large signal-to-noise ratio in the Raman spectra of ZnO/CdxZn1−xS ﬁlms prepared at x0 = 1.00, 0.75, and 0.50 registered at kexc = 514.5 nm attests to the resonant character of the scattering as a result of close energies of the exciting quanta and the bandgaps of CdxZn1−xS NPs. The composition of CdxZn1−xS solid solution can be determined with a good accuracy either from the LO phonon frequency mLO using the well-known empirical

Fig. 6.26 a Normalized absorption spectra of the ITO/ZnO/CdxZn1−xS films synthesized at x0 = 1.0 (curve 1), 0.9 (curve 2), 0.6 (curve 3), 0.4 (curve 4), and 0.1 (curve 5). b Raman spectra of the ITO/ZnO/CdxZn1−xS films produced at x0 = 0.25 (curves 1), 0.50 (curves 2), 0.75 (curves 3), and 1.0 (curves 4). kexc = 514.5 nm (b) and 325 nm (c, d)[74] 354 6 Probing with Light—Optical Methods in Studies … equation x = 0.013 Â (mLO − 303) [75] or directly from the LO peak position using the reported calibration curves [76–78]. The real molar Cd(II) ratios in CdxZn1−xS NP deposited on the ZnO surface determined in this way, xRRS, are presented in Fig. 6.27. Despite the fact that xabs and xRRS were estimated for two sets of ITO/ZnO/CdxZn1−xS samples with different x0, they perfectly match and comple- ment each other thus producing a reliable calibration curve allowing to determine the real composition of SILAR-deposited CdxZn1−xS NPs with x0 varied from 1.0 to zero. Complementary to the optical data, the surface of ITO/ZnO/CdxZn1−xS films was studied by EDX allowing to quantify the atomic composition of the films. It showed that the atomic Cd:S ratio is very close to 1:1 at x0 = 1 and decreases with decreasing x0 indicating the formation of mixed cadmium-zinc sulfide. Similarly to the optical data, the EDX shows a strong deviation of the real composition of the films relative to the Cd:Zn ratio set at the SILAR procedure. The values of real composition of CdxZn1−xS NPs determined by EDX, xEDX, appeared to be in a perfect accordance with the results of the optical absorption and Raman spectroscopies (Fig. 6.27) attesting to the high accuracy and reliability of the above-discussed optical methods. The Raman spectroscopy of nanoheterostructures comprising two and more components provides ample information on their structure and allows to distinguish between the core/shell NPs with a continuous shell and an island-like shell. Also, its can reveal an effect of the interdiffusion of the components having close parameters of the crystal lattice. For example, the resonant Raman spectra of core/shell CdSe/ZnS NPs show a distinct phonon peak at 300 cm−1 typical for cadmium sulfide (Fig. 6.28) indicating the interdiffusion of the materials of the shell (zinc sulfide) and the core (cadmium sulfide) [79–81].

Fig. 6.27 Calibration dependence between the molar Cd(II) fraction x0 in a mixed Cd(II)-Zn(II) solution used for the SILAR deposition of CdxZn1−xS and the real molar Cd(II) fraction in ITO/ZnO/CdxZn1−xS nanocomposites. The values were determined using the optical absorption spectroscopy (xabs, squares), the resonant Raman spectroscopy (xRRS, circles), and the energy-dispersive X-ray spectroscopy (xEDX, diamonds) [74] 6.5 Studies of Nanocrystalline Semiconductors … 355

Fig. 6.28 Resonant Raman spectra of (a) CdSe NPs (curve 1) and the core/shell CdSe/ZnS NPs (curve 2) incorporated into the gelatin ﬁlms (kexc = 457.9 nm) [79, 80]; b CdS NPs (curve 1) and the photocatalytically formed CdS/CdSe nanoheterostructures (curves 2–5) in the gelatin ﬁlms (excitation at 441.7 nm, CdSe contant is 4 mol% (curve 2), 7 mol% (curve 3), 10 mol% (curve 4), and 13 mol% (curve 6) [82]; c CdS/CdSe nanoheterostructure with 13 mol% CdSe (relative to CdS content) registered at a different excitation wavelength [82]

The Raman spectroscopy was fruitfully used in the studies of binary CdS/CdSe nanoheterostructures synthesized by the photocatalytic reduction of Na2SeSO3 on the surface of cadmium sulfide NPs [82]. The resonant Raman spectra of CdS/CdSe showed the LO peaks of both components with the CdSe LO peak growing with an increase in the photodeposited CdSe content (Fig. 6.28b). The main phonon peak of 6–7-nm CdS NPs used as a photocatalyst can be found at 302 cm−1 shifting slightly to lower frequencies as compared to the bulk cadmium selenide (305 cm−1 [67]). The small shift magnitude attests to a weak phonon confinement in such NPs. However, the CdS NPs revealed a much higher FWHM of the LO band (70– 80 cm−1) as compared to both the photodeposited CdSe (40 cm−1) and the CdSe NPs prepared in “dark” conditions via the interaction between Na2SeSO3 and −1 CdCl2 (10–20 cm )[79, 83–85]. Most probably, the fact is associated with a high density of the bulk and surface lattice defects in CdS NPs. The interdiffusion produces a weak signal at 500 cm−1 which is a combination of the second-order vibrations LOCdS +LOCdSe [79] (Fig. 6.28b). The strong interdiffusion is also typical for the core/shell CdSe/CdS NPs produced in a non-catalytic way [79, 83]. As the lattice constants of CdSe and CdS are slightly different, the contact between the two semiconductors results in the diffusion of sulfur atoms into the bulk of cadmium selenide and the formation of a mixed CdSxSe1−x layer. The process results in a larger surface disordering in CdS NPs manifesting as an increased SOCdS intensity for the photocatalytically produced CdS/CdSe NPs. The LO phonon peak of CdSe can be observed as a low-intensity shoulder at 185 cm−1 for low CdSe contents but appears as a well-resolved peak at 200 cm−1 for the highest cadmium selenide content (13 mol%). The peak is shifted by around 10 cm−1 to lower frequencies as compared to the bulk CdSe (210 cm−1 [67]) 356 6 Probing with Light—Optical Methods in Studies … indicating a considerable phonon confinement, that is, a small size of the photodeposited CdSe NPs. The spectral width and position of the LO bands do not reflect directly the size and size distribution of CdSe NPs because the spectra were registered under the resonant conditions (kexc = 441.7 nm) when a spectral contribution of the NPs with the bandgap energy closest to the excitation energy is the highest in the ensemble. At the same time, an increase in the phonon peak intensity with a growing CdSe content indicates that a fraction of the resonantly excited CdSe NPs becomes larger. As the excitation wavelength is increased the resonance conditions for the selective excitation of CdSe NPs become more and more favorable. As a result, the ratio of LOCdSe and LOCdS peaks distinctly grows as kexc is increased from 441.7 to 514.5 nm (Fig. 6.28c) indicating that the laser energy (2.41 eV) is close to the bandgap of the photodeposited CdSe NPs. The fact can be taken as an indication that cadmium selenide is indeed photodeposited as separate NPs, not as sub-nanometer 2D islands, for which the resonance energy is expected to be much higher. The driving force for the formation of 3D NPs can be supplied by the lattice constant mismatch of CdS and CdSe resulting in a compressive stress that can be relaxed via the transformation of primary 2D CdSe islands into the 3D NPs similarly to the well-reported Stransky-Krastanov transformation of epitaxial AIIIBV semiconductor nano-islands [86]. The ratio of the main phonon mode and its overtone 2LO depends on the electron-phonon interaction in the semiconductor lattice amounting to I2LO/ ILO = 0.3–0.4 similarly to 2–5 nm CdSe NPs produced by a non-catalytic method [87], the fact additionally proving the formation of separate 3D CdSe NPs as a product of the photocatalytic deposition. This example of CdS/CdSe nanoheterostructures demonstrates quite clearly the capabilities of the Raman spectroscopy in the studies of composite semiconductor nanoheterostructures that can potentially be used in the photocatalytic and photoelectrochemical light-harvesting systems.

6.6 Studies of Colloidal Semiconductor-Based Systems Using Dynamic Light Scattering

The dynamic light scattering (DLS) or the laser photon correlation spectroscopy can be used as an alternative or a complementary method to the transmission electron microscopy (TEM) in the studies of colloidal systems with semiconductor NPs and other components dispersed in a liquid medium. The method is based on the detection of fluctuations of the elastic (Reighley) light scattering by the colloidal particles changing their position chaotically in the Brownian movement [88]. The method allows determining the diffusion coefficient of NPs (polymer globules, large molecules, etc.) or a distribution of the diffusion coefficients—for the polydisperse 6.6 Studies of Colloidal Semiconductor-Based Systems … 357 colloidal systems. Finally, the size/size distribution of the colloidal species can be calculated using the well-known Einstein-Stokes equation. One of the undoubtful advantages of DLS is the possibility to probe “live” colloids, while the preparation of samples for TEM, scanning electron microscopy (SEM) and X-ray diffraction (XRD) requires the destruction of a colloidal system and the extraction of a dispersed nano-phase that can possibly be accompanied by the phase transitions and particle size changes. Simultaneously with the determination of the diffusion coefficient, the DLS can provide important information on the structure of a double electric layer of the NP surface and the surface charge. Modern DLS setups allow determining the size of colloidal NPs down to 1 nm with an accuracy of ±0.1 nm [89]. On the other hand, the application area of the DLS method is confined to colloidal systems stabilized by adsorbed ions or relatively small ligands. The Einstein-Stokes equation gives the hydrodynamic size of NPs, that is, the size of a colloidal micelle composed of the NP “core”, a layer of stabilizer molecules and a solvation shell moving as a whole entity in the Brownian movement. As a result, DLS cannot typically be applied to the polymer-stabilized NPs because the polymers are present in the form of globules as large as several hundred nm thus masking the target NPs. In the case of colloidal systems with no bulky polymers present, the DLS method can provide quite precise determination of the size of colloidal semiconductor NPs. For example, this method can distinguish colloidal ZnO NPs differing only slightly by the average size [90]. As the starting reagent concentration is − − increased from 2 Â 10 3 to 2 Â 10 2 M the absorption band edge of resulting colloidal zinc oxide NPs shifts from *345 to 355 nm (Fig. 6.29a) corresponding to a variation of the average NP size from 3.7 to 4.4 nm. The results of DLS presented in Fig. 6.29 show the feasibility of the reliable determination of a minute difference in the ZnO NP size that can be observed in the absorption spectra owing to a strong size dependence of the bandgap. The positions of size distribution maxima determined by the DLS (Fig. 6.29b, c) and the average NP size derived

Fig. 6.29 a Normalized absorption spectra (a) and size distributions obtained from the DLS measurements (b, c) of colloidal ZnO NPs in ethanol synthesized at a starting reactant − − concentration of 2 Â 10 3 M(a, curve 1; b) and 2 Â 10 2 M(a, curve 2; c)[90] 358 6 Probing with Light—Optical Methods in Studies … from the optical absorption spectra (the values presented in Fig. 6.29b, c) almost coincide. The DLS size determination was successfully applied to track the growth of ultra-small core/shell CdSe/CdS NPs stabilized in aqueous solutions by mixed Cd (II) complexes with ammonia and mercaptoacetate ions [91]. The growth occurs during the thermal treatment of a starting solution at 90–95 °C and the NP size can be varied by adjusting the treatment duration. As the heating proceeds the bandgap of CdSe QDs determined from the electron absorption band edge decreases gradually from 2.89 to 2.70 eV after 45 min heating (Table 6.4). These values are strongly shifted to higher energies as compared with the band gap of bulk CdSe, 1.75 eV, due to the strong spatial conﬁnement of the photogenerated charge carriers in very small CdSe NPs. The average size d of the CdSe NPs was estimated from Eg using the well-known empirical calibration curve presented in Fig. 6.9 to be as small as 1.9 nm (Table 6.4) for the colloid produced by the 2-min heating and growing to 2.2 nm after the 45-min heating. A TEM study of the CdSe colloid produced by the 2-min thermal treatment (Fig. 6.30a) showed the CdSe NPs to be 1.8–2.0 nm in size revealing a high degree of the size uniformity. The interparticle aggregation was minimized due to an electrostatic barrier of charged Cd(II)-NH3-mercaptoacetate complexes adsorbed on the NP surface. Despite such a small size, the NPs showed a good crystallinity with a well-resolved interplanar distance of 3.5 ± 0.1 Å typical for the cubic CdSe (Fig. 6.30b, c). A scatter of CdSe NP size around the average value did not exceed 0.5 nm (Fig. 6.30d). The DLS spectroscopy conﬁrmed the presence of individual ultra-small CdSe NPs in the colloidal solution (Fig. 6.30e, Table 6.4). The species in the starting solution are characterized by an average hydrodynamic size of 1.8 nm which increases upon the thermal treatment to 2.4 nm for the CdSe NPs produced by the 2-min heating and grows up to 3.5 nm for the CdSe NPs formed after the thermal treatment for 45 min (Fig. 6.30e, curve 6; Table 6.4). Therefore, for the colloidal CdSe solution heated for 2 min the NP size estimated from the spectral curve corresponds to the TEM data and agrees with the DLS results, the difference between d and dDLS indicating the existence of a half-nm-thick surface stabilizer layer on the NP surface. A larger discrepancy between d and dDLS for the colloidal solution produced at the 45-min heating was ascribed to the formation of a shell on the surface of CdSe

Table 6.4 Band gap Eg, size t, min Eg,eV d,nm dDLS,nm d, and hydrodynamic size 0 ––1.8 dDLS of CdSe NPs subjected to thermal treatment during 2 2.89 1.9 2.1 time t [91] 4 2.83 2.0 2.5 5 2.80 2.1 2.7 15 2.73 2.2 2.8 30 2.72 2.2 3.0 45 2.70 2.2 3.5 6.6 Studies of Colloidal Semiconductor-Based Systems … 359

Fig. 6.30 TEM (a), high-resolution TEM (b, c) images and size distribution (d) of CdSe/CdS NPs (2 min thermal treatment). e Hydrodynamic size distribution in starting CdSe nuclei solution (curve 1) and after the thermal treatment for 1 min (curve 2), 3 min (curve 3), 5 min (curve 4), 17 min (curve 5), and 45 min (curve 6) [91]

NPs, that is clearly visible for the DLS spectroscopy but does not contribute substantially to the position of the band edge in the electron absorption spectra. It was assumed [91] that the discrepancy between the size of CdSe NPs derived from the optical absorption (and TEM) data and the hydrodynamic NP size that increases during the thermal treatment originate from the formation of a protecting CdS layer on the surface of CdSe NPs as a result of the partial hydrolysis of mercaptoacetate anions in strongly alkaline solutions at 90–95 °C. This assumption found con- vincing proofs in the results of the Raman and X-ray photoelectron spectroscopy of such core/shell CdSe/CdS NPs [91]. Thus, the DLS spectroscopy combined with optical absorption spectroscopy and TEM can be a powerful tool for probing the structure of semiconductor NPs and nanocomposites even at the low size scale. The application area of the DLS spectroscopy is not limited to “rigid” inorganic NPs having an invariable size and shape. This method can also be applied to probe “soft” systems, in particular those containing ultra-thin layers of various photoactive materials, such as molybdenum or tungsten dichalcogenides. Of special interest are the DLS studies of colloidal graphene oxide (GO) and reduced graphene oxide (RGO) used very often as co-catalysts of various light processes and as components of the photoelectrochemical solar cells [92–101]. The GO sheets feature a random alternation of the aromatic graphene areas of the sp2-hybridized carbon and the oxidized regions, where the sp3-hybridized carbon atoms are bound to various oxygen-containing functional groups—the epoxy and hydroxyl groups, carboxyl groups, etc. [92, 93, 102, 103]. As a result, the GO sheets are flexible and can attain a scrolled or crumpled shape [104–106]. The transformation of GO particles into a nonplanar conformation is also favored by the 360 6 Probing with Light—Optical Methods in Studies … formation of hydrogen bonds between the functional groups from isolated fragments of the same GO sheet, either directly or via the bridge water molecules [92, 93, 103, 105, 107]. In the case of RGO sheets, the intraparticle bonds can also originate from the pp stacking interactions between the isolated aromatic areas of RGO sheets [105, 107, 108]. Therefore, the properties of GO(RGO)-based colloidal systems depend strongly on the shape and aggregation of the RGO sheets that can change with the variations of solvent properties, in particular, pH, ionic strength, and temperature [103, 105–107, 109, 110]. The rare reports on the effects of pH and ionic strength on the shape of colloidal GO (RGO) discuss mostly indirect TEM observations of the sheet aggregates produced by the solvent evaporation/extraction [105–107] or the mathematical modeling of the shape changes [105]. At the same time, direct observations of the shape evolution of GO (RGO) particles in colloidal solutions can be made by the nondestructive DLS method [106, 110]. Such direct studies of the shape evolution and aggregation are of special interest for RGO, which is the most frequently used 2D material in the light-harvesting systems and solar cells. However, the studies are typically obstructed by instability of the colloidal particles caused by the presence of ionic residuals from decomposition of a reducing agent used to convert GO into RGO [103, 107, 109]. To avoid the introduction of chemical reductants, the photochemical reduction of colloidal GO can be applied, as it does not require any additional reagents except for water. Also, the photoreduction does not change pH and ionic strength of the solution and allows to vary smoothly the photoreduction “depth” by adjusting the light intensity and/or the illumination duration (exposure) [111–113]. A DLS study of aqueous GO colloids showed that the average hydrodynamic size of colloidal particles varies from *150 to * 550 nm with a distribution maximum at dDLS = 320 nm (Fig. 6.31a, curve 1) [114]. The absence of planar GO/RGO particles larger than a half-micron typically observed in the atomic force microscopic images indicates that the GO (RGO) particles are crumpled as a result of the pp stacking and the H-bond formations between various functional groups. The GO/RGO particles deposited from colloidal solutions onto hydrophobic substrates, such as carbon films and conductive FTO glass, preserve this partially or strongly crumpled shape which can be visualized by the TEM/SEM measurements (Fig. 6.31b, c). The photoreduction of GO, even at a starting stage (first 30 min illumination), results in considerable changes of the sheet properties manifesting as a drastic growth of dDLS up to 520 nm (Fig. 6.31a, curve 2). According to the absorption, Raman, and infrared absorption spectroscopy this time range is also characterized by the most vivid changes in the structure and bandgap energy of RGO [115, 116], in particular, in a considerable increase of the fraction of aromatic carbon in the RGO sheets. These changes become only deeper at further illumination, however, the hydrodynamic size of colloidal RGO starts to change in a reverse direction decreasing to 360 and 300 nm for 60 and 90-min light exposure, respectively (Fig. 6.31a, curves 3, 4). When the photoreduction is finished at the 180-min exposure, the RGO particles are 6.6 Studies of Colloidal Semiconductor-Based Systems … 361

Fig. 6.31 a Hydrodynamic size distribution of colloidal GO at pH 6 before the photoreduction (curve 1) and after the illumination with the UV light for 30 min (curve 2), 60 min (curve 3), 90 min (curve 4), and 180 min (curve 5). b, c SEM (b) and TEM (c) images of a crumpled GO particle. d Scheme of the photoinduced changes in the shape and structure of colloidal GO particles [114] characterized by an average hydrodynamic size of 260 nm (Fig. 6.31a, curve 5), which is smaller than dDLS of the starting GO particles. The hydrodynamic size of planar GO (RGO) particles is an effective value depending primarily on the way the sheets are crumpled. The DLS studies of colloidal dispersions of the single-layer graphene, MoS2, and WS2 [117] showed that a dependence between the the lateral size L of these planar particles and their b hydrodynamic size dDLS can be expressed as dDLS = aL , where a = 5.9 ± 2.2, b = 0.66 ± 0.06. Therefore, the observed photoinduced size evolution of colloidal GO/RGO particles indicates the changes of the sheet crumpling character as a result of the intra-sheet interactions. The mechanism of such shape evolution can be illustrated by a scheme presented in Fig. 6.31d. The crumpling of starting GO is caused by the formation of intra-sheet hydrogen bonds between the functional groups in different fragments of the GO sheets including the “bridge” water molecules (the case I in Fig. 6.31d) [118, 119]. The feasibility of spontaneous folding of GO sheets and the formation of 0.42-nm thick folds between the sheet fragments was confirmed by the molecular modeling [120]. The GO photoreduction results the primary stage in abrupt changes in the sheet composition, in particular, in a partial elimination of epoxide and hydroxy groups from the basal GO plane and the restoration of its aromatic character. At that, most of the H-bonds interconnecting the folds disappear and the RGO sheet becomes more unfolded, the fact mirrored by an increase of the hydrodynamic size (the case II in Fig. 6.31d). As the RGO becomes photoreduced deeper and deeper, the aromatic character of the basal plane becomes more expressed and new folds start to form between the sheet fragments via the pp stacking of the sp2-hybridized RGO fragments. The formation of new folds is also favored by increasingly hydrophobic character of the photoreduced GO, as the RGO sheets tend to minimize their contact 362 6 Probing with Light—Optical Methods in Studies … with the polar medium. As calculations showed [119], a distance between the fold-forming fragments of the RGO sheets is around 0.39 nm, which is close to the interlayer distance in the bulk graphite (0.34–0.38 nm). A pronounced tendency of RGO to the folding and crumpling was broadly reported [121–123]. A decrease in the interactions between RGO and water and a more compact crumpling of the RGO sheets as a result of the pp stacking results in the fact that the deeply photoreduced RGO has a lower dDLS as compared to the starting GO. The above-discussed model finds support in the results of pH-dependent DLS spectroscopy of the photoreduced GO. As pH of GO/RGO colloids is lowered to 2, the sheet aggregation is observed in all cases. At that, the RGO aggregate size depends on the reduction depth and varies from around 400 nm for the original GO to 2.5–3 lm for the most deeply reduced RGO. A pH increase to 11, on the contrary, results in a decrease in dDLS—down to 100–120 nm for the RGO with the highest reduction depth (Fig. 6.32a, curves 1, 2). The latter observation shows the tendency of RGO sheets to crumple and to minimize the surface contacting with the polar medium with an increased ionic strength. The dialysis purification of alkaline RGO colloids removes the alkali and the hydrodynamic size of RGO sheets returns to a starting value of 280 nm (Fig. 6.32a, curves 2, 3). Then, as NaOH is added for the second time, the hydrodynamic size of RGO particles decreases again (Fig. 6.32a, curves 3, 4). These observations indicate that the pH-induced shape changes of colloidal RGO have a dynamic character, and the conformation of colloidal RGO sheets can “adapt” to the polarity and ionic strength of the dispersive medium. The reversible character of the conformational changes of RGO sheets is also evidenced by changes of the hydrodynamic size induced by the interactions of the

Fig. 6.32 Hydrodynamic size distribution of the photoreduced colloidal RGO sheets pre-illuminated for 180 min at pH 6 (curve 1) and treated in two consequences: a after elevating pH to 10 with NaOH (curve 2), reducing pH to 7 by the dialysis (curve 3) and, again, elevating pH to 10 with NaOH (curve 4); b after elevating pH to 10 with NaOH (curve 2), and adding methylene blue (curve 3) or sodium salt of pyrene sulfonic acid (curve 4) [114, 116] 6.6 Studies of Colloidal Semiconductor-Based Systems … 363

RGO sheets with aromatic molecules. The latter are capable of the strong adsorption on the RGO basal plane via pp stacking. The absorption can provoke the unfolding of the crumpled RGO particles. For example, introduction of methylene blue dye or sodium salt of the pyrene sulfonic acid into the colloidal RGO (with the highest photoreduction depth) solution results in a dDLS increase from 120 to 290– 300 nm (Fig. 6.32b). The methylene blue and pyrene derivatives have a well-known affinity to the aromatic fragments of RGO sheets [93]. An additional electrostatic repulsion between the neighboring adsorbed molecules that bear a positive (methylene blue) or negative charge (pyrene sulfonic acid anion) evidently overwhelms the intra-sheet pp stacking interactions between the fragments of crumpled RGO resulting in the sheet unfolding and an increase of the hydrodynamic size. It should be noted that the results discussed here, along with the data of [110], where the DLS spectroscopy was successfully applied to study reversible interactions between the colloidal GO sheets and DNA molecules at variations of the solution temperature, demonstrate a high potential and a unique character of this method in studying subtle effects, such as the conformational changes of single layer 2D sheets under the external stimuli directly in colloidal solutions. Concluding the discussion of experimental methods that exploit the interaction of nanocrystalline semiconductors with light to probe the structure and properties of such nanomaterials we note that this discussion has an introductory character and does not pretend on a comprehensive characterization of the whole variety of optical and spectroscopic methods applied nowadays for the investigations of nanocrystalline materials. We aimed to provide a general notion on the possibilities of using the light to study semiconductor NPs and nanoheterostructures, high- lighting only some the most useful or unique capabilities of such methods. Also, the optical methods constitute only a small fraction of the versatility of modern arsenal of techniques used to get an insight into the intimate details of the structure and properties of nanocrystalline objects. Recently some disbalance can be observed in the appreciation and application of “classical” optical characterization methods as compared to other modern characterization methods, most of them very demanding and sophisticated in the instrumental sense. The researchers strive to characterize their nanomaterials with the largest possible array of structural methods, such as the XRD, TEM, SEM, atomic force microscopy, X-ray photoelectron and UV photoelectron spectroscopy, nuclear magnetic resonance, etc., trying to collect as much information as possible on the structure of such light-harvesting materials. From the other hand, the utilization of such a versatile array of techniques is, at least partially, caused by more and more rigid standards and demands to the instrumental level of studies of semiconductor nanomaterials put forth by the authoritative scientific journals. However, very often the presented results are analyzed only superficially. At the same time, the optical methods that gained deserved esteem from the early stages of the studies of nanocrystalline semiconductors and nanoheterostructures are applied in a less and less habitual way and retreat undeservedly to the background of modern methodology. We hope that this chapter will help at improving this disbalance by 364 6 Probing with Light—Optical Methods in Studies … delivering a proper impression on the capabilities of the optical methods in collecting the most versatile information on the electron, photophysical and structural characteristics of the nanocrystalline semiconductor light-harvesting materials.

References

1. Stroyuk OL, Sobran IV, Korzhak AV et al (2008) Photopolymerization of water-soluble acrylic monomers induced by colloidal CdS and CdxZn1−xS nanoparticles. Colloid Polym Sci 286:489–498. doi:10.1007/s00396-007-1824-4 2. Raevskaya AE, Stroyuk OL, Kryukov AI et al (2006) Structural and optical characteristics of CdxZn1−xS nanoparticles stabilized in aqueous solutions of polymers. Theor Exp Chem 42:181–185. doi:10.1007/s11237-006-0035-2 3. Stroyuk OL, Dzhagan VM, Kuchmii SY et al (2007) Nanosecond and microsecond decay of photogenerated charges in CdxZn1−xS nanoparticles. Theor Exp Chem 43:297–305. doi:10. 1007/s11237-007-0037-8 4. Grätzel M (ed) (1983) Energy resources through photochemistry and catalysis. Academic Press, Inc., New York 5. Korgel BA, Monbouquette HG (2000) Controlled synthesis of mixed core and layered (Zn, Cd)S and (Hg, Cd)S nanocrystals within phosphatidylcholine vesicles. Langmuir 16:3588– 3594. doi:10.1021/la990139r 6. Petrov DV, Santos BS, Pereira GA, de Mello Donegá C (2002) Size and band-gap dependences of the first hyperpolarizability of CdxZn1−xS nanocrystals. J Phys Chem B 106:5325–5334. doi:10.1021/jp010617i 7. Kasap S, Capper P (eds) (2006) Springer handbook of electronic and photonic materials. Springer Science + Business Media, Inc, Berlin 8. de Mello Donega C, Koole R (2009) Size dependence of the spontaneous emission rate and absorption cross section of CdSe and CdTe quantum dots. J Phys Chem C 113:6511–6520. doi:10.1021/jp811329r 9. Monticone S, Tufeu R, Kanaev AV (1998) Complex nature of the UV and visible fluorescence of colloidal ZnO nanoparticles. J Phys Chem B 102:2854–2862. doi:10.1021/ jp973425p 10. Jasieniak J, Smith L, Embden J, Mulvaney P (2009) Re-examination of the size-dependent absorption properties of CdSe quantum dots. J Phys Chem C 113:19468–19474. doi:10. 1021/jp906827m 11. Schoenhalz AL, Arantes JT, Fazzio A, Dalpian GM (2010) Surface and quantum confinement effects in ZnO nanocrystals. J Phys Chem C 114:18293–18297. doi:10.1021/ jp103768v 12. Gaponenko SV (1998) Optical properties of semiconductor nanocrystals. University Press, Cambridge 13. Rogach A (ed) (2008) Semiconductor nanocrystal quantum dots: synthesis, assembly, spectroscopy and applications. Springer, Vienna 14. Wang Y, Herron N (1991) Nanometer-sized semiconductor clusters: materials synthesis, quantum size effects, and photophysical properties. J Phys Chem 95:525–532. doi:10.1021/ j100155a009 15. Ye C, Fang X, Wang M, Zhang L (2006) Temperature-dependent photoluminescence from elemental sulfur species on ZnS nanobelts. J Appl Phys 99:063504. doi:10.1063/1.2181311 16. Eychmüller A (2000) Structure and photophysics of semiconductor nanocrystals. J Phys Chem B 104:6514–6528. doi:10.1021/jp9943676 17. Burda C, Chen X, Narayanan R, El-Sayed MA (2005) Chemistry and properties of nanocrystals of different shapes. Chem Rev 105:1025–1102. doi:10.1021/cr030063a References 365

18. Nanda K, Kruis F, Fissan H (2001) Energy levels in embedded semiconductor nanoparticles and nanowires. Nano Lett 1:605–611. doi:10.1021/nl0100318 19. Qu L, Peng X (2002) Control of photoluminescence properties of CdSe nanocrystals in growth. J Am Chem Soc 124:2049–2055. doi:10.1021/ja017002j 20. Chen W, Wang Z, Lin Z, Lin L (1997) Absorption and luminescence of the surface states in ZnS nanoparticles. J Appl Phys 82:3111–3115. doi:10.1063/1.366152 21. Tang H, Xu G, Weng L et al (2004) Luminescence and photophysical properties of colloidal ZnS nanoparticles. Acta Mater 52:1489–1494. doi:10.1016/j.actamat.2003.11.030 22. Cizeron J, Pileni MP (1995) Solid Solution of CdyZn1−yS nanosize particles made in reverse micelles. J Phys Chem 99:17410–17416. doi:10.1021/j100048a016 23. Cizeron J, Pileni MP (1997) Solid solution of CdyZn1−yS nanosized particles: photophysical properties. J Phys Chem B 101:8887–8891. doi:10.1021/jp9713571 24. Melamed NT (1957) Sulfur vacancy emission in ZnS phosphors. Phys Rev 107:1727–1728. doi:10.1103/PhysRev.107.1727.2 25. Rayevska AE, Stroyuk OL, Kozytskiy AV, Kuchmiy SY (2010) Electron energy factors in photocatalytic methylviologen reduction in the presence of semiconductor nanocrystals. J Photochem Photobiol A 210:209–214. doi:10.1016/j.jphotochem.2009.11.019 26. Feilchenfeld H, Chumanov H, Cotton TM (1996) Photoreduction of methylviologen adsorbed on silver. J Phys Chem 100:4937–4943. doi:10.1021/jp952329q 27. Ferrer B, Llabrés i Xamena FX, García H (2007) Hollow organosilica spheres as hosts: 2− Photoinduced electron transfer between Ru(bpy)3 and methylviologen. Inorg Chim Acta 360:1017–1022. doi:0.1016/j.ica.2006.07.107 28. Zhong H, Bai Z, Zou B (2012) Tuning the luminescence properties of colloidal I-III–VI semiconductor nanocrystals for optoelectronics and biotechnology applications. J Phys Chem Lett 3:3167–3175. doi:10.1021/jz301345x 29. Liu S, Su X (2014) The synthesis and application of I-III–VI type quantum dots. RSC Adv 4:43415–43428. doi:10.1039/C4RA05677A 30. Leach ADP, Macdonald JE (2016) Optoelectronic properties of CuInS2 nanocrystals and their origin. J Phys Chem Lett 7:572–583. doi:10.1021/acs.jpclett.5b02211 31. Kolny-Olesyak J, Weller H (2013) Synthesis and application of colloidal CuInS2 semicon– ductor nanocrystals. ACS Appl Mater Interfaces 5:12221–12237. doi:10.1021/am404084d 32. Tang Y, Wang P, Yun JH et al (2015) Frequency-regulated pulsed electrodeposition of CuInS2 on ZnO nanorod arrays as visible light photoanodes. J Mater Chem A 3:15876– 15881. doi:10.1039/C5TA03255E 33. Wang S, Yang X, Zhu Y et al (2014) Solar-assisted dual chamber microbial fuel cell with a CuInS2 photocathode. RSC Adv 4:23790–23796. doi:10.1039/C4RA02488E 34. Pan Z, Mora-Sero I, Shen Q et al (2014) High-efficiency “Green” quantum dot solar cells. J Am Chem Soc 136:9203–9210. doi:10.1021/ja504310w 35. Raevskaya AE, Rosovik OP, Kozytskiy AV et al (2016) Non-Stoichiometric Cu-In-S@ZnS nanoparticles produced in aqueous solutions as light harvesters for liquid-junction photoelectrochemical solar cells. RSC Adv 6:100145–100157. doi:10.1039/C6RA18313A 36. Zhang B, Wang Y, Yang C et al (2015) The composition effect on the optical properties of aqueous synthesized Cu–In–S and Zn–Cu–In–S quantum dot nanocrystals. Phys Chem Chem Phys 17:25133–25141. doi:10.1039/C5CP03312H 37. Stroyuk OL, Raevskaya AE, Kuchmii SY (2004) Oxidation of polysulfide ions induced by CdS nanoparticles under pulsed photolysis conditions. Theor Exp Chem 40:130–135. doi:10.1023/B:THEC.0000028910.49933.67 38. Raevskaya AE, Stroyuk OL, Kuchmiy SY (2004) Photocatalytic oxidation of hydrosulfide-ions by molecular oxygen over cadmium sulfide nanoparticles. J Nanopart Res 6:149–158. doi:10.1023/B:NANO.0000034719.30620.d3 39. Vorobets VS, Kovach SK, Kolbasov GY (2001) Ionic equilibria in sulfide/polysulfide solutions with an account for the effect of ionic coupling. Ukr Chem J (in Russian) 67:12–16 40. Giggenbach W (1972) Optical spectra and equilibrium distribution of polysulfide ions in aqueous solution at 20°. Inorg Chem 11:1201–1207. doi:10.1021/ic50112a009 366 6 Probing with Light—Optical Methods in Studies …

41. Licht S, Hodes G, Manassen J (1986) Numerical analysis of aqueous polysulfide solutions and its application to cadmium chalcogenide/polysulfide photoelectrochemical solar cells. Inorg Chem 25:2486–2489. doi:10.1021/ic00235a003 42. Meyer B (1976) Elemental sulfur. Chem Rev 76:367–388. doi:10.1021/cr60301a003 − 43. Chivers T, Drummond I (1972) Characterization of the trisulfur radical anion S3 in blue solutions of alkali polysulfides in hexamethylphosphoramide. Inorg Chem 11:2525–2527. doi:10.1021/ic50116a047 44. Gruen DM, McBeth RL, Zielen AJ (1971) Nature of sulfur species in fused salt solutions. J Am Chem Soc 93:6691–6693. doi:10.1021/ja00753a070 45. Martin RP, Doub WH, Roberts JL, Sawyer DT (1973) Electrochemical reduction of sulfur in aprotic solvents. Inorg Chem 12:1921–1925. doi:10.1021/ic50126a047 46. Annenkova VS, Antonik LM, Haliullin AK et al (1983) Formation of sodium sulfide anion radicals in dipolar aprotic solvents. J General Chem (in Russian) 53:2409–2410 47. Mills G, Schmidt KH, Matheson MS, Meisel D (1987) Thermal and photochemical reactions of sulfhydryl radicals. Implications for colloid photocorrosion. J Phys Chem 91:1590–1596. doi:10.1021/j100290a060 48. Melnikov MY (1994) Photochemistry of organic radicals. Moscow University Publishing, Moscow (in Russian) 49. Vossmeyer T, Katsikas L, Giersig M et al (1994) CdS nanoclusters: synthesis, characterization, size dependent oscillator strength, temperature shift of the excitonic transition energy, and reversible absorbance shift. J Phys Chem 98:7665–7673. doi:10.1021/ j100082a044 50. Kovach SK, Vorobets VS, Vasko AT (1992) Potentiometric titration using a sulfide-selective electrode. Ukr Chem J 58:491–494 51. Dzhagan VM, Stroyuk OL, Rayevska AE et al (2010) Spectroscopic and photochemical + 2+ 2+ 3+ study of Ag -, Cu -, Hg -, and Bi -doped CdxZn1−xS nanoparticles. J Colloid Interface Sci 345:515–523. doi:10.1016/j.jcis.2010.02.001 52. Stroyuk OL, Raevskaya AE, Korzhak AV et al (2009) Photocatalytic production of 0 hydrogen in systems based on CdxZn1−xS/Ni nanostructures. Theor Exp Chem 45:12–22. doi:10.1007/s11237-009-9057-x 53. Henglein A, Lilie J (1981) Storage of electrons in aqueous solution: the rates of chemical charging and discharging the colloidal silver microelectrode. J Am Chem Soc 103:1059– 1066. doi:10.1021/ja00395a011 54. Savinov EN, Nagornyi VE, Parmon VN (1994) Influence of excessive charge of colloidal cadmium sulfide particles on interfacial electron transfer rate. Chem Phys (in Russian) 13:56–65 55. Matsumoto H, Uchida H, Matsunaga T et al (1994) Photoinduced reduction of viologens on size-separated CdS nanocrystals. J Phys Chem 98:11549–11556. doi:10.1021/j100095a041 56. Zhukowskiy MA, Stroyuk OL, Shvalagin VV et al (2009) Photocatalytic growth of CdS, PbS, and CuxS nanoparticles on the nanocrystalline TiO2 films. J Photochem Photobiol, A 203:137–144. doi:10.1016/j.jphotochem.2009.01.007 57. Stroyuk OL, Kuchmii SY, Zhukovskii MA et al (2009) Effect of the method of production of TiO2/CdS film nanoheterostructures on the effectiveness of photoinduced charge separation. Theor Exp Chem 45:302–307. doi:10.1007/s11237-009-9097-2 58. Serpone N, Lawless D, Khairutdinov R, Pelizzetti E (1995) Subnanosecond relaxation dynamics in TiO2 colloidal sols (particle sizes Rp = 1.0–13.4 nm). Relevance to heterogeneous photocatalysis. J Phys Chem 99:16655–16661. doi:10.1021/j100045a027 59. Bahnemann D, Henglein A, Lilie J, Spanhel L (1984) Flash photolysis observation of the absorption spectra of trapped positive holes and electrons in colloidal TiO2. J Phys Chem 88:709–711. doi:10.1021/j150648a018 60. Teoh WY, Mädler L, Beydoun D et al (2005) Direct (one-step) synthesis of TiO2 and Pt/TiO2 nanoparticles for photocatalytic mineralisation of sucrose. Chem Eng Sci 60:5852– 5861. doi:10.1016/j.ces.2005.05.037 References 367

61. Duonghong D, Ramsden J, Grätzel M (1982) Dynamics of interfacial electron transfer processes in colloidal semiconductor systems. J Am Chem Soc 104:2977–2985. doi:10. 1021/ja00375a006 62. Bessekhouad Y, Chaoui N, Trzpit M et al (2006) UV-vis versus visible degradation of Acid Red II in a coupled CdS/TiO2 semiconductors suspension. J Photochem Photobiol A 183:218–224. doi:10.1016/j.jphotochem.2006.03.025 63. Liu Y, Wang X, Yang F, Yang X (2008) Excellent antimicrobial properties of mesoporous anatase TiO2 and Ag/TiO2 composite films. Micropor Mesopor Mater 114:431–439. doi:10. 1016/j.micromeso.2008.01.032 64. Rajh T, Micic OI, Lawless D, Serpone N (1992) Semiconductor photophysics. 7. Photoluminescence and picosecond charge carrier dynamics in cadmium sulfide quantum dots confined in a silicate glass. J Phys Chem 96:4633–4641. doi:10.1021/j100190a090 65. Kamat PV, Dimitrijević NM, Fessenden RW (1987) Photoelectrochemistry in particulate systems. 6. Electron-transfer reactions of small CdS colloids in acetonitrile. J Phys Chem 91:396–401. doi:10.1021/j100286a029 66. Dzhagan VM, Valakh MY, Raevskaya AE et al (2008) Size effects on Raman spectra of small CdSe nanoparticles in polymer films. Nanotechnology 19:305707 67. Landolt-Börnstein (1982) Numerical data and functional relationships in science and technology. Group III, Vol. 17b—Semiconductors, Sect. 3.10.1. Springer, Berlin 68. Chen X, Mao SS (2007) Titanium dioxide nanomaterials: synthesis, properties, modifications, and applications. Chem Rev 107:2891–2959. doi:10.1021/cr0500535 69. Trallero-Giner C, Debernardi A, Cardona M et al (1997) Optical vibrons in CdSe dots and dispersion relation of the bulk material. Phys Rev B 57:4664. doi:10.1103/PhysRevB.57. 4664 70. Zhang JY, Wang XY, Xiao M et al (2002) Lattice contraction in free-standing CdSe nanocrystals. Appl Phys Lett 81:2076–2078. doi:10.1063/1.1507613 71. Hwang YN, Shin S, Park HL et al (1996) Effect of lattice contraction on the Raman shifts of CdSe quantum dots in glass matrices. Phys Rev B 54:15120. doi:10.1103/PhysRevB.54. 15120 72. Rolo AG, Vasilevskiy MI (2007) Raman spectroscopy of optical phonons confined in semi– conductor quantum dots and nanocrystals. J Raman Spectr 38:618–633. doi:10.1002/jrs. 1746 73. Raevskaya AE, Grodzyuk GY, Korzhak AV et al (2011) Preparation and optical properties of polyethyleneimine-stabilized colloidal CdSe and CdSxSe1−x quantum dots. Theor Exp Chem 46:416–421. doi:10.1007/s11237-011-9173-2 74. Kozitskiy AV, Stroyuk OL, Kuchmiy SY et al (2013) Morphology, optical, and photoelectro–chemical properties of electrodeposited nanocrystalline ZnO films sensitized with CdxZn1−xS nanoparticles. J Mater Sci 48:7764–7773. doi:10.1007/s10853-013-7598-9 75. Yükselici H, Persans PD, Hayes TM (1995) Optical studies of the growth of Cd1−xZnxS nanocrystals in borosilicate glass. Phys Rev B 52:11763. doi:10.1103/PhysRevB.52.11763 76. Azhniuk YM, Gomonnai AV, Lopushansky VV et al (2007) Morphology, optical, and photoelectrochemical properties of electrodeposited nanocrystalline ZnO films sensitized with CdxZn1−xS nanoparticles. J Phys 92:012044. doi:10.1007/s10853-013-7598-9 77. Arora AK, Rajalakshmi M (2000) Resonance Raman scattering from Cd1−xZnxS nanoparticles dispersed in oxide glass. J Appl Phys 88:5653. doi:10.1063/1.1321025 78. Sahoo S, Dhara S, Sivasubramanian V, Kalavathi S, Arora AK (2009) Phonon confinement and substitutional disorder in Cd1−xZnxS nanocrystals. J Raman Spectrosc 40:1050. doi:10. 1002/jrs.2232 79. Dzhagan VM, Valakh MY, Raevskaya AE et al (2007) Resonant Raman scattering study of CdSe nanocrystals passivated with CdS and ZnS. Nanotechnology 18:285701 80. Dzhagan VM, Valakh MY, Raevska AE et al (2009) The influence of shell parameters on phonons in core–shell nanoparticles: a resonant Raman study. Nanotechnology 20:365704 81. Raevskaya AE, Stroyuk OL, Kuchmiy SY et al (2007) Optical study of CdS- and ZnS-passivated CdSe nanocrystals in gelatin films. J Phys: Condens Matter 19:386237 368 6 Probing with Light—Optical Methods in Studies …

82. Dzhagan VM, Raevskaya AE, Stroyuk OL et al (2009) Resonant Raman spectroscopy of conﬁned and surface phonons in CdSe-capped CdS nanoparticles. Phys Stat Sol C 6:2043– 2046. doi:10.1002/pssc.200881755 83. Dzhagan VM, Valakh MY, Raevskaya AE et al (2007) Temperature-dependent resonant Raman scattering study of core/shell nanocrystals. J Phys Conf Series 92:012045 84. Dzhagan VM, Valakh MY, Raevskaya AE et al (2008) Characterization of semiconductor core-shell nanoparticles by resonant Raman scattering and photoluminescence spectroscopy. Appl Surf Sci 255:725–727. doi:10.1016/j.apsusc.2008.07.018 85. Raevskaya AE, Stroyuk AL, Kuchmiy SY (2006) Preparation of colloidal CdSe and CdS/CdSe nanoparticles from sodium selenosulfate in aqueous polymers solutions. J Colloid Interface Sci 302:133–141. doi:10.1016/j.jcis.2006.06.018 86. Krasil’nik ZK, Lytvyn P, Lobanov DN et al (2002) Microscopic and optical investigation of Ge nanoislands on silicon substrates. Nanotechnology 13:81–85 87. Raevskaya AE, Stroyuk OL, Kuchmiy SY et al (2006) Growth and spectroscopic characterization of CdSe nanoparticles synthesized from CdCl2 and Na2SeSO3 in aqueous gelatine solutions. Colloids Surfaces A 290:304–309. doi:10.1016/j.colsurfa.2006.05.038 88. Berne BJ, Pecora R (1976) Dynamic light scattering. Wiley, New York 89. www.malvern.com/ZetasizerNanoZS 90. Stroyuk OL, Dzhagan VM, Shvalagin VV et al (2010) Size-dependent optical properties of colloidal ZnO nanoparticles charged by photoexcitation. J Phys Chem C 114:220–225. doi:10.1021/jp908879h 91. Raevskaya AE, Stroyuk OL, Panasiuk YV et al (2016) A new route to very stable water-soluble ultra-small core/shell CdSe/CdS quantum dots. Nano-Objects, Nano-Struct. doi:10.1016/j.nanoso.2015.12.001 92. Zhu Y, Murali S, Cai W et al (2010) Graphene and graphene oxide: synthesis, properties, and applications. Adv Mater 22:3906–3924. doi:10.1002/adma.201001068 93. Dreyer DR, Park S, Bielawski CW, Ruoff RS (2010) The chemistry of graphene oxide. Chem Soc Rev 39:228–240. doi:10.1039/B917103G 94. Bai S, Shen X (2012) Graphene-inorganic composites. RSC Adv 2:64–98. doi:10.1039/ C1RA00260K 95. An X, Yu JC (2011) Graphene-based photocatalytic composites. RSC Adv 1:1426–1434. doi:10.1039/C1RA00382H 96. Shen J, Zhu Y, Yang X, Li C (2012) Graphene quantum dots: emergent nanolights for bioimaging, sensors, catalysis and photovoltaic devices. Chem Commun 48:3686–3699. doi:10.1039/C2CC00110A 97. Singh V, Joung D, Zhai L et al (2011) Graphene based materials: past, present and future. Progr Mater Sci 56:1178–1271. doi:10.1016/j.pmatsci.2011.03.003 98. Zhang Y, Tang ZR, Fu X, Xu YJ (2011) Engineering the unique 2D mat of graphene to achieve graphene-TiO2 nanocomposite for photocatalytic transformation: what advantage does graphene have over its forebear carbon nanotube? ACS Nano 9:7426–7435. doi:10. 1021/nn202519j 99. Xiang Q, Yu J, Jaroniec M (2012) Graphene-based semiconductor photocatalysts. Chem Soc Rev 41:782–796. doi:10.1039/C1CS15172J 100. Yao J, Sun Y, Yang M, Duan Y (2012) Chemistry, physics and biology of graphene-based nanomaterials: new horizons for sensing, imaging and medicine. J Mater Chem 22:14313– 14329. doi:10.1039/C2JM31632C 101. Huang X, Zeng Z, Fan Z, Liu J, Zhang H (2012) Graphene-based electrodes. Adv Mater 24:5979–6004. doi:10.1002/adma.201201587 102. Erickson K, Erni R, Lee Z et al (2010) Determination of the local chemical structure of graphene oxide and reduced graphene oxide. Adv Mater 22:4467–4472. doi:10.1002/adma. 201000732 103. Konkena B, Vasudevan S (2012) Understanding aqueous dispersibility of graphene oxide and reduced graphene oxide through pKa measurements. J Phys Chem Lett 3:867–872. doi:10.1021/jz300236w References 369

104. Wen X, Garland CW, Hwa T et al (1992) Crumpled and collapsed conformations of graphite oxide membranes. Nature 355:426–428. doi:10.1038/355426a0 105. Whitby RLD, Gun’ko VM, Korobeynik A et al (2012) Driving forces of conformational changes in single-layer graphene oxide. ACS Nano 6:3967–3973. doi:10.1021/nn3002278 106. Zangmeister CD, Ma X, Zachariah MR (2012) Restructuring of graphene oxide sheets into monodisperse nanospheres. Chem Mater 24:2554–2557. doi:10.1021/cm301112j 107. Whitby RLD, Korobeynik A, Gun’ko VM et al (2011) pH-driven physicochemical conformational changes of single-layer graphene oxide. Chem Commun 47:9645–9647. doi:10.1039/C1CC13725E 108. Kim K, Lee Z, Malone BD et al (2011) Multiply folded graphene. Phys Rev B 83:245433. doi:10.1103/PhysRevB.83.245433 109. Park S, An J, Piner RD et al (2008) Aqueous suspension and characterization of chemically modiﬁed graphene sheets. Chem Mater 20:6592–6594. doi:10.1021/cm801932u 110. Tang L, Wang Y, Liu Y, Li J (2011) DNA-directed self-assembly of graphene oxide with applications to ultrasensitive oligonucleotide assay. ACS Nano 5:3817–3822. doi:10.1021/ nn200147n 111. Stroyuk OL, Andryushina NS, Shcherban ND et al (2012) Photochemical reduction of graphene oxide in colloidal solution. Theoret Exp Chem 48:1–11. doi:10.1007/s11237-012- 9235-0 112. Shulga YM, Martynenko VM, Muradyan VE et al (2010) Gaseous products of thermo- and photo-reduction of graphite oxide. Chem Phys Lett 498:287–291. doi:10.1016/j.cplett.2010.08. 056 113. Matsumoto Y, Koinuma M, Kim SY et al (2010) Simple photoreduction of graphene oxide nanosheet under mild conditions. ACS Appl Mater Interfaces 2:3461–3466. doi:10.1021/ am100900q 114. Andryushina NS, Stroyuk OL, Yanchuk IB, Yefanov AV (2014) A dynamic light scattering study of photochemically reduced colloidal graphene oxide. Colloids Polym Sci 292:539– 546. doi:10.1007/s00396-013-3134-3 115. Prezioso S, Perrozzi F, Donarelli M et al (2012) Large area extreme-UV lithography of graphene oxide via spatially resolved photoreduction. Langmuir 28:5489–5495. doi:10. 1021/la204637a 116. Stroyuk OL, Andryushina NS, Kuchmy SY et al (2015) Photochemical processes involving graphene oxide. Theor Exp Chem 51:1–29. doi:10.1007/s11237-015-9393-y 117. Lotya M, Rakovich A, Donegan JF, Coleman JN (2013) Measuring the lateral size of liquid-exfoliated nanosheets with dynamic light scattering. Nanotechnology 24:265703 118. Dreyer DR, Park S, Bielawski CW, Ruoff RD (2010) The chemistry of graphene oxide. Chem Soc Rev 39:228–240. doi:10.1039/B917103G 119. Shul’ga YM, Vasilets VN, Baskakov SA et al (2012) Photoreduction of graphite oxide nanosheets with vacuum ultraviolet radiation. High En Chem 46:117–121. doi:10.1134/ S0018143912020099 120. Feng L, Wu L, Qu X (2013) New horizons for diagnostics and therapeutic applications of graphene and graphene oxide. Adv Mater 25:168–186. doi:10.1002/adma.201203229 121. Zhang B, Li L, Wang Z et al (2012) Radiation induced reduction: an effective and clean route to synthesize functionalized graphene. J Mater Chem 22:7775–7781. doi:10.1039/ C2JM16722K 122. Bittolo BS, Piccinini M, Mariani A et al (2011) Wettability and switching of electrical conductivity in UV irradiated graphene oxide ﬁlms. Diamond Rel Mater 20:871–874. doi:10.1016/j.diamond.2011.04.013 123. Zhang HH, Liu Q, Feng K et al (2012) Facile photoreduction of graphene oxide by an NAD (P)H model: Hantzsch 1,4-Dihydropyridine. Langmuir 28:8224–8229. doi:10.1021/ la301429g 124. Rayevska AE, Grodzyuk GY, Dzhagan VM et al (2010) Synthesis and characterization of white-emitting CdS quantum dots stabilized with polyethyleneimine. J Phys Chem C 114:22478–22486. doi:10.1021/jp108561u 370 6 Probing with Light—Optical Methods in Studies …

125. Raevskaya AE, Grodzyuk GY, Stroyuk OL et al (2010) Preparation and spectral properties of high-efficiency luminescent polyethyleneimine-stabilized CdS quantum dots. Theor Exp Chem 46:233–238. doi:10.1007/s11237-010-9145-y 126. Stroyuk OL, Raevskaya AE, Korzhak AV, Kuchmiy SY (2007) Zink sulfide nanoparticles: spectral properties and photocatalytic activity in metals reduction reactions. J Nanopart Res 9:1027–1039. doi:10.1007/s11051-006-9183-5 127. Matsumoto H, Sakata T, Mori H, Yoneyama H (1996) Preparation of monodisperse CdS nanocrystals by size selective photocorrosion. J Phys Chem 100:13781–13785. doi:10.1021/ jp960834x 128. Li J, Xia JB (2000) Hole levels and exciton states in CdS nanocrystals. Phys Rev B 62:12613–12616. doi:10.1103/PhysRevB.62.12613 129. Lippens P, Lannoo M (1989) Calculation of the band gap for small CdS and ZnS crystallites. Phys Rev B 39:10935–10942. doi:10.1103/PhysRevB.39.10935 130. Wang Y, Herron N (1990) Quantum size effects on the exciton energy of CdS clusters. Phys Rev B 42:7253–7255. doi:10.1103/PhysRevB.42.7253 131. Yang CC, Jiang Q (2006) Size effect on the bandgap of II–VI semiconductor nanocrystals. Mater Sci Eng B 131:191–194. doi:10.1016/j.mseb.2006.04.016 132. Sapra S, Sarma D (2004) Evolution of the electronic structure with size in II-VI semiconductor nanocrystals. Phys Rev B 69:125304. doi:10.1103/PhysRevB.69.125304 133. Gorer S, Ganske J, Hemminger J, Penner R (1998) Size-selective and epitaxial electrochemical/chemical synthesis of sulfur-passivated cadmium sulfide nanocrystals on graphite. J Am Chem Soc 120:9584–9593. doi:10.1021/ja981676l 134. Baskoutas S, Terzis A (2006) Size-dependent band gap of colloidal quantum dots. J Appl Phys 99:013708. doi:10.1063/1.2158502 135. Murakoshi K, Hosokawa H, Saitoh M et al (1998) Preparation of size-controlled hexagonal CdS nanocrystallites and the characteristics of their surface structures. J Chem Soc, Faraday Trans 94:579–586. doi:10.1039/A707192B 136. Nandakumar P, Vijayan C, Murti Y (2002) Optical absorption and photoluminescence studies on CdS quantum dots in Nafion. J Appl Phys 91:1509–1514. doi:10.1063/1.1425077 137. Leistikow MD, Johansen J, Kettelarij AJ et al (2009) Size-dependent oscillator strength and quantum efficiency of CdSe quantum dots controlled via the local density of states. Phys Rev B 79:045301. doi:10.1103/PhysRevB.79.045301 138. Wood A, Giersig M, Hilgendorff M et al (2003) Size effects in ZnO: the cluster to quantum dot transition. Austr J Chem 56:1051–1057. doi:10.1071/CH03120 139. Wong EM, Hörtz PG, Liang CJ et al (2001) Influence of organic capping ligands on the growth kinetics of ZnO nanoparticles. Langmuir 17:8362–8367. doi:10.1021/la010944h 140. Brus LE (1984) Electron-electron and electron-hole interactions in small semiconductor crystallites: the size dependence of the lowest excited electronic state. J Chem Phys 80:4403–4444. doi:10.1063/1.447218 141. Fonoberov VA, Balandin AA (2004) Radiative lifetime of excitons in ZnO nanocrystals: The dead-layer effect. Phys Rev B 70:195410. doi:10.1103/PhysRevB.70.195410 142. Wang YS, Thomas PJ, O’Brien P (2006) Nanocrystalline ZnO with ultraviolet luminescence. J Phys Chem B 110:4099–4104. doi:10.1021/jp0566313 143. Kwak H, Tiago ML, Chelikowsky JR (2008) Quantum confinement and strong coulomb corre– lation in ZnO nanocrystals. Sol State Commun 145:227–230. doi:10.1016/j.ssc.2007.11.004 144. van Dijken A, Janssen AH, Smitsmans MHP et al (1998) Size-selective photoetching of nano–crystalline semiconductor particles. Chem Mater 10:3513–3522. doi:10.1021/ cm980715p 145. Kumbhojkar N, Nikesh VV, Kshirsagar A, Mahamuni S (2000) Photophysical properties of ZnS nanoclusters. J Appl Phys 88:6260–6264. doi:10.1063/1.1321027 146. Souici AH, Keghouche N, Delaire JA et al (2006) Radiolytic synthesis and optical properties of ultra-small stabilized ZnS nanoparticles. Chem Phys Lett 422:25–29. doi:10.1016/j.cplett. 2006.02.013 References 371

147. Ghosh PK, Jana S, Nandy S, Chattopadhyay KK (2007) Size-dependent optical and dielectric properties of nanocrystalline ZnS thin films synthesized via rf-magnetron sputtering technique. Mater Res Bull 42:505–514. doi:10.1016/j.materresbull.2006.06.019 148. Nanda J, Sapra S, Sarma DD, Chandrasekharan N, Hodes G (2000) Size-selected zinc sulfide nanocrystallites: synthesis, structure, and optical studies. Chem Mater 12:1018–1024. doi:10.1021/cm990583f 149. Kane RS, Cohen RE, Silbey R (1996) Theoretical study of the electronic structure of PbS nanoclusters. J Phys Chem 100:7928–7932. doi:10.1021/jp952869n 150. Kang I, Wise W (1997) Electronic structure and optical properties of PbS and PbSe quantum dots. J Opt Soc Am B 14:1632–1646. doi:10.1364/JOSAB.14.001632 151. Wang Y, Suna A, Mahler W, Kasowski R (1987) PbS in polymers. From molecules to bulk solids. J Chem Phys 87:7315–7322. doi:10.1063/1.453325 152. Miyoshi H, Yamachika M, Yoneyama H, Mori H (1990) Photochemical properties of PbS Microcrystallites prepared in Nafion. J Chem Soc, Faraday Trans 86:815–818. doi:10.1039/ FT9908600815 153. Sapra S, Shanthi N, Sarma DD (2002) Realistic tight-binding model for the electronic structure of II-VI semiconductors. Phys Rev B 66:205202. doi:10.1103/PhysRevB.66.205202 154. Lippens PE, Lannoo M (1991) Electronic structure of II–VI semiconductor nanocrystals. Mater Sci Eng B 9:485–487 155. Wang LW, Zunger A (1996) Pseudopotential calculations of nanoscale CdSe quantum dots. Phys Rev B 53:9579–9582. doi:10.1103/PhysRevB.53.9579 156. Yu WW, Qu L, Guo W, Peng X (2003) Experimental determination of the extinction coefficient of CdTe, CdSe, and CdS nanocrystals. Chem Mater 15:2854–2860. doi:10.1021/ cm034081k 157. Rogach AL, Kornowski A, Gao M et al (1999) Synthesis and characterization of a size series of extremely small thiol-stabilized CdSe nanocrystals. J Phys Chem B 103:3065–3069. doi:10.1021/jp984833b 158. Shiang JJ, Kadavanich AV, Grubbs RK, Alivisatos AP (1995) Symmetry of annealed wurtzite CdSe nanocrystals: assignment to the C3v point group. J Phys Chem 99:17417– 17422. doi:10.1021/j100048a017 159. Murray CB, Norris DJ, Bawendi MG (1993) Synthesis and characterization of nearly monodisperse CdE (E = sulfur, selenium, tellurium) semiconductor nanocrystallites. J Am Chem Soc 115:8706–8715. doi:10.1021/ja00072a025 160. Raevskaya AE, Stroyuk OL, Kuchmii SY (2003) Optical characteristics of colloidal nanoparticles of CdS stabilized with sodium polyphosphate and their behavior during pulse photoexcitation. Theor Exp Chem 39:158–165. doi:10.1023/A:1024933023783 161. Kozytskiy AV, Stroyuk OL, Kuchmiy SY et al (2014) Photoelectrochemical and Raman characterization of nanocrystalline CdS Grown on ZnO by successive ionic layer adsorption and reaction. Thin Solid Films 562:56–62. doi:10.1016/j.tsf.2014.03.070 162. Kozytskiy AV, Stroyuk OL, Skoryk MA et al (2015) Photochemical formation and photoelectrochemical properties of TiO2/Sb2S3 heterostructures. J Photochem Photobiol A 303–304:8–16. doi:10.1016/j.jphotochem.2015.02.005 163. Kozytskiy AV, Stroyuk OL, Kuchmiy SY (2014) Inorganic photoelectrochemical solar cells based on nanocrystalline ZnO/ZnSe and ZnO/CuSe heterostructures. Catal Today 230:227– 233. doi:10.1016/j.cattod.2013.09.043 164. Swamy V, Kuznetsov A, Dubrovinsky LS et al (2005) Finite-size and pressure effects on the Raman spectrum of nanocrystalline anatase TiO2. Phys Rev B 71:184302. doi:10.1103/ PhysRevB.71.184302 165. Azhniuk YM, Hutych YI, Lopushansky VV et al (2007) Interplay of factors affecting Raman scattering in cadmium chalcogenide nanocrystals in dielectric media. J Phys: Conf Ser 79:012017 166. Bersani D, Loticci PP (1992) Confinement effects on the LO-phonons in CdSexS1−x doped glasses. Phys stat sol (b) 174:575–582. doi:10.1002/pssb.2221740227 Index

A C Absorption band edge/threshold, 4, 244, 322 Cadmium selenide, 18, 21, 51, 72, 145, 178, Absorption spectrum, 167, 321, 357 190, 328, 335, 350 Aerosol AT, 246 Cadmium sulfide, 6, 43, 130, 145, 184, 196, Alloyed nanoparticles, 210 275, 347 Alumina membrane, 273 Cadmium sulfo-selenide, 210, 352 Aluminium oxide, 272 Cadmium telluride, 26, 329 AM1.5 light flux, 58, 167 Cadmium zinc sulfide, 18, 65, 146, 188, 324, Anatase, 67, 131, 141, 254, 273, 278, 350 327, 330, 340, 352 Anistropic shape, 71, 147, 191 Calibration curve, 326, 351 Anodization, 61, 142, 269 Carbon-doped titania, 61 Antimony sulfide, 199, 350 Carbon materials, 54, 217, 226, 279 Artificial photosynthesis, 87, 128 Carbon microspheres, 269 Atomic force microscopy, 198 Carbon nanoparticles, 54, 78, 209 Atomic layer deposition, 263 Carbon nanotubes, 54, 79, 82, 280 Average size, 326 Carbon vacancy, 83 Cascade charge transfer, 52, 190, 203 B Cascade conduction band levels, 208 Band alignment, 209 Cascade design, 208, 211 Band bending, 15 Catalytic activity, 218, 226 Band design, 58 Cathodic polarization, 26, 28 Barrier layer, 207 Ceramics, 272 Bequerel, xxiv Cerium oxide, 144 Bifunctional molecules, 171, 243 Chalcopyrite, 52, 175 Binary heterostructures/nanocomposites, 48, Charge carrier migration, 15 143, 275 Charge collection efficiency, 206 Biomass, 71, 92 Charge compensation, 56 Bio-mimicking, 152 Charge leakage, 203, 206 Bipyridyl, 91, 135, 137, 140 Charge migration, 265 Bismuth oxide, 56, 148, 258 Charge separation, 15, 68, 196, 199 Bismuth oxyhalogenide, 147, 258 Charge transfer, 16 Bismuth sulfide, 53, 65, 145 Charge transfer complex, 43, 139 “Black” titania, 57 Charge transfer rate constant, 178 Blocking layer, 203, 206 Charge transfer resistance, 175, 218, 221, 225 Boron carbide, 144 Charge trapping, 13, 141, 330, 336 Brookite, 131, 254 Chemical bath deposition, 192, 222, 347 Bulk heterojunction solar cells, 162 Chemical vapor deposition, 170 Burstein-Moss effect, 23, 340 Chromium oxide, 60

© Springer International Publishing AG 2018 373 O. Stroyuk, Solar Light Harvesting with Nanocrystalline Semiconductors, Lecture Notes in Chemistry 99, https://doi.org/10.1007/978-3-319-68879-4 374 Index

Ciamician, xvii, xviii Electrochemical reduction, 191, 242, 277 Clays, 271 Electrodeposition, 191, 215, 264 Clusters, 6, 10, 131, 262 Electron absorption spectroscopy, 321 CO2conversion, 127, 273 Electron acceptor, 3 Cobalt phosphate, 72, 77 Electron affinity, 46 Cobalt sulfide, 218 Electron beam sputtering, 279 Co-catalyst, 41, 81, 86, 132, 135, 151, 224, 281 Electron donor, 3 Colloidal semiconductors, 242, 319, 356 Electron-hole recombination, 14, 203, 330, 337 Colloidal titania, 248 Electron mobility, 215, 222 Colloidal zinc oxide, 250 Electron paramagnetic resonance, 15, 51, 131 Conduction band potential, 2, 188 Electron transition, 321, 323 Conjugated polymer, 131, 149, 205 Electron trap, 18, 332 Copper antimony sulfide, 191 Electrophoretic deposition, 182, 257 Copper indium selenide, 205 Electrophysical parameters, 321 Copper indium sulfide, 52, 76, 213, 337 Energy diagram, 214, 225, 336 Copper oxide, 53, 75, 143, 192 Energy-dispersive X-ray spectroscopy, 188, Copper phosphide, 53 200, 354 Copper sulfide, 86, 196, 218, 223 Environmental photocatalysis, 79 Copper telluride, 191 Eosin, 42, 58, 82 Copper tin sulfide, 215 Erythrosin, 43, 82 Core/shell, 144, 195, 251, 337, 354, 358 Evonik P25, 67, 261, 274 Coulomb interaction, 6, 10, 322 Excessive negative charge, 65, 340, 349 coumarine, 42 Exciton absorption, 8, 322 Counter electrode, 217 Exciton binding energy, 11 Cysteine, 145, 176 Exciton Bohr radius, 1, 8 Excitonic photoluminescence, 16 D Exfoliation, 44, 54, 64, 73, 82, 86, 135, 261, de-Broglie wavelength, 2 267 Defect-related photoluminescence, 16, 330 Exposed facets, 50, 131, 133, 142, 147 Defects, 8 Extinction spectrum, 201 Density functional theory, 150 Design of photocatalysts, 129 F Diffusion coefficient, 244 Fermi level, 45, 167 Dip coating, 260 Ferrites, 73 Dipole moment, 63 Fill factor, 167 Direct aqueous synthesis, 179 Finite-depth potential well, 327 Direct electron transition, 7, 323 First-principles calculations, 64 Discharging rate, 344 Flash photolysis, 65, 198 Dodecanthiol, 171 Flexible electrode, 226 Doping, 55, 58, 140, 149, 273 Free Gibbs energy, 3, 128 Double charged layer, 14, 26, 357 Fujishima and Honda, xx, xxi, xxii, xxvi Drop-casting, 180 Fullerene, 49, 54, 76, 281 Dye sensitized solar cell, 137, 161, 164, 217 Fundamental absorption, 7, 322 Dynamic light scattering, 356 G E Gallium nitride, 77 Effective mass, 9, 326 Gallium oxide, 60, 134 Effective mass approximation, 9, 326 Gold nanoparticles, 195 Einstein-Stokes equation, 357 Gradient composition, 67, 69, 210 Electric capacitance, 65, 206, 344 Graphene, 59, 78, 144, 281 Electrochemical deposition, 170 Graphene oxide, 177, 281, 359 Electrochemical etching, 79, 216, 264, 269 Graphite, 273, 281 Electrochemical impedance spectroscopy, 218 Graphitic carbon, 280 Index 375

Graphitic carbon nitride, 43, 54, 79, 92, 135, Lead sulfide, 21, 187, 199, 222, 224, 328 142, 147 Life-time, 330, 334, 341 Ligand exchange, 174 H Light absorption, 6 Heating up synthesis, 171 Light conversion efficiency, 162, 166, 218 Heterojunction, 144, 346 Light harvesting cycle, 162 Hodes G., 186 Light harvesting system, xvii, xix, xx, 92, 161, Hollow sphere, 55, 81, 132, 143, 217, 265, 268 167, 241, 269, 356, 360 Hot electron, 45, 76 Light scattering layer, 194, 216 Hot-injection synthesis, 171 Light-shielding effect, 201 Hydrazine, 67, 71, 147 Light-to-current conversion efficiency, 53 Hydrodynamic size, 357, 360 Linear absorption coefficient, 7 Hydrolysis, 247, 265, 274 Liquid-junction solar cells, 162 Hydrothermal treatment, 215, 249, 257 Liquid phase deposition, 261 Loosely aggregated nanoparticles, 261, 264 I Luminescence/photoluminescence Impedance spectrum, 206 spectroscopy, 329, 335 Incident-photon-to-current efficiency, 47, 168, 214 M Indirect electron transition, 7, 323 Magic-size clusters, 179 Indium oxide, 60, 144 Magnetron sputtering, 57, 90, 263 Indium phosphide, 140 Mechanochemical treatment, 61, 257, 274 Indium selenide, 72 Mercaptopropionic acid, 51, 171 Indium sulfide, 21 Merocyanine, 42 Industrial wastes, 71 Mesoporous cadmium sulfide, 67, 267 Ink, 180 Mesoporous framework, 266, 273 Inorganic complex ligands, 177 Mesoporous materials, 264 In situ deposition, 183 Mesoporous metal chalcogenide, 267 Intercalation, 50, 82, 90, 270, 277 Mesoporous silica, 132, 141 Interdiffusion, 354 Mesoporous titania, 44, 59, 141, 165, 266 Interfacial charge transfer, 3, 275, 337, 347 Metal complex dye, 43 Intermediate, 131, 198, 337, 342, 346 Metallate, 73, 258, 270 Internal electric field, 68 Metal mesh, 224 Inverted (inverse) opal, 50, 217 Metal-organic framework, 83, 135 Inverted micelles, 246 Metal sulfide photocatalyst, 61, 245 Ion exchange, 170, 201, 224, 272 Methylviologen, 16, 21, 22, 334 Iron oxide, 144, 150, 269, 279 Micellar solution, 267 Iron silicide, 78 Microemulsion, 246 Isotopic studies, 134, 148, 150 Microsphere, 131, 134, 144, 258, 266 Microwave treatment, 254, 259 K Mid-bandgap states, 147 Kamat P., xxix Molar absorption coefficient, 7 Kesterite, 5, 52, 77, 215, 225 Molecular orbital, 140 Molecular photocatalyst, 136 L Molybdenium disulfide, 20, 62, 86, 151, 194, Lambert-Beer equation, 6 359 Laser photocorrelation spectroscopy, 356 Monolith reactor, 144 Laser pulse deposition, 264 Mott-Schottky, 133, 135 Lattice defects, 13, 82, 147, 345, 355 Multi-electron process, 28, 130, 133, 135 Layered material, 50, 64, 151, 258, 270 Multi-exciton generation, 23, 165 Layered metal chalcogenide, 64 Multi-layer structures, 182 Lead selenide, 20, 176, 191 376 Index

N Photochromic properties, 284 Nafion membrane, 137, 150, 272 Photocorrosion, 28, 67, 70 Nanocrystalline films, 260 Photocurrent density, 166 Nanocrystalline powders, 252 Photocurrent generation efficiency, 203 Nanorod, 194, 215, 222 Photocurrent spectrum, 186 Nanoscroll, 44, 259 Photoelectric effect, xxiv, xxvi Nanosheet, 46, 73, 133, 149, 259, 269 Photoelectrocatalytic system, 86 Nanotube, 42, 51, 269, 283 Photoelectron spectroscopy, 177 Nanotube array, 60, 142, 216, 224, 269 Photoetching, 66 Nanowire, 47, 52, 175, 201, 215, 257 Photoinduced charge accumulation, 165 N-doped titania, 59, 142, 266, 274 Photoinduced charging, 13, 23 Nickel oxide, 76, 138, 267 Photoinduced electron transfer, 22, 150 Nickel sulfide, 86 Photoinduced polarization, 65 Niobate, 44, 50, 73, 133, 267 Photoluminescence quenching, 181 Nitridation, 59, 274 Photoluminescence spectroscopy, 213 Nitrogen fixation, 127, 146 Photolysis, 70 Nitrogen vacancy, 135, 147 Photonic crystal, 46, 144, 217 Noble metal, 86 Photopolymerization, 21 Non-stationary bleaching, 24, 340, 349 Photovoltage, 188 Nyquist plot, 206 Phthalocyanine, 42, 76, 82, 139, 145 Plasmonic photocatalyst, 44, 81 O Platinum group metals, 52 Oleylamine, 171 p/n Heterojunction, 53 Open-circuit photovoltage, 167, 188 Polycondensation, 247 Optical fiber, 261 Polyelectrolyte, 180 Optical phonon, 350 Polymer films, 330, 350, 355 Oscillator strength, 11 polystyrene latex/microparticles, 268 Ostwald ripening, 247, 269 Pore size distribution, 266 Oxidative photocorrosion, 143 Porphyrin, 84, 135, 140 Oxide/chalcogenide heterostructures, 187, 196, Post-synthesis treatment, 246, 274 201, 275, 346 Power conversion efficiency, 167 Oxygen evolution, 57 Prebiotic photosynthesis, 127 Oxygen vacancy, 58, 131, 134, 147, 148 Primary nuclei, 181, 244 Oxysulfide, 61, 70 Protective shell, 169, 174, 179, 208, 359 Pulse photoexcitation, 25 P Pulse photolysis, 26 Paper, 273, 281 Pyrolysis, 262, 283 Paris Climate Conference, 129 Passivating ligand, 171 Q Perovskite, 5, 60, 73, 91, 133, 166, 270 Quantum confinement, 176 Phase composition, 253, 350 Quantum-sized nanoparticles, 20, 244, 322 Phase size effect, 5 Quantum size effects, 2, 5, 133, 203, 325 Phonon, 7 Quaternary metal chalcogenide, 52, 70, 225 Phonon confinement, 351, 356 Photoaction spectrum, 45, 47, 135, 140, 168 R Photocatalyst, 39 Radiative recombination, 18 Photocatalytic microreactor, 248 “Rainbow-cell” design, 211 Photocatalytic system, xx, xxi, xxii, xxiii, 39, Raman spectroscopy, 186, 350, 353, 359 41–43, 49, 54–56, 58, 59, 61, 62, 62, 67, Reactor geometry, 130 71, 81, 85, 87, 88, 90–92, 127, 129, 130, Redox potential, 19 152, 162 Red phosphorus, 77 Photocathode, 74, 138, 224 Reduced graphene oxide, 64, 76, 84, 87, 132, Photochemical/photocatalytic deposition, 49, 142, 149, 218, 281, 359 52, 59, 81, 170, 194, 222, 283, 348 Resonant excitation, 350, 356 Index 377

Reverse electron transfer, 207, 213 Successive ionic layer adsorption and reaction, Rhodamine, 85 169, 183, 218, 275, 352 Ruthenium bipyridyl, 42, 44 Sulfidation, 224, 226 Rutile, 67, 254 Sulfur-doped titania, 61 Rydberg energy, 9 Sulfur vacancy, 63, 130, 147 Surface defects, 5, 67, 333, 336, 355 S Surface optical phonon, 351 Sacrificial donor, 40, 92 Surface plasmon resonance, 44, 136, 284 Scaffold, 162, 215 Surface states, 14, 341 Schottky barrier, 45 Surface-to-volume ratio, 204 Selection rules, 7 Self-igniting mixtures, 256, 273 T Semiconductor-metal nanostructures, 281 Tafel equation, 22, 189 Semiconductor-sensitized solar cells, 161 Tandem, 138, 140 Sensitization, 84, 273 Tantalates, 50 Sensitizer, 42, 137, 139, 163, 283 Tantalum nitride, 59 Short-circuit photocurrent density, 167 Tantalum oxide, 59 Silica nanoparticles, 267, 271 Tantalum oxynitride, 59 Silicon carbide, 77 Template, 63, 81, 144 Silicon nanoparticles, 76, 135, 151 Ternary metal chalcogenide, 52, 76, 186, 193, Silicon solar cell, 138, 161 225 Siloxene, 77 Thioglycolic acid, 176 Silver antimony sulfide, 191 Third-generation solar cells, 161 Silver bismuth sulfide, 191 Time-resolved laser photolysis, 198, 339 Silver indium sulfide, 52, 213 Time-resolved photoluminescence, 177, 334 Silver sulfide, 187, 199 Tin sulfide, 63 Single-layer sheets, 64, 361 Titanate, 42, 50, 258, 269 Singlet excited state, 85 Titania nanotubes, 145 Single-wall carbon nanotube, 91 Titania-silica composites, 255 Size dependence, 72, 213, 331, 341 Titanium oxyfluoride, 90 Size distribution, 244, 253, 326, 349, 356 Titanosilicate, 272 Size-selected fractionation, 247 Total water splitting, 57, 270 Size-selected nanoparticles, 62, 173, 211, 326, Trapped electron/hole, 198, 332 357 Tungstate, 258 Size variation, 244 Tungsten disulfide, 194 Solar cell market, 161 Tungsten oxide, 50, 133, 258 Solar light simulator, 167 Type II heterojunction, 186 Solar spectrum, 58 Sol-gel method, 247, 274 U Solid solution, 64, 68, 91, 187, 210, 324, 352 Ultra-small nanoparticles, 10, 179, 249, 358 Solvatochromic sensor, 284 Ultrasound treatment, 267, 282 Spatial confinement, 9 Urbach Equation, 323 Spatial design, 152 Spatial organization, 49, 89, 196 V Spatial separation, 49, 52, 144, 198, 209, 275 Valence band potential, 2 Spectral methods, 319 Volcano-shaped dependence, 187, 213, 344 Spin coating, 261 Voltage-current curve, 167 Stokes shift, 331 Structural defects, 204 W Structural disorder, 186, 351 Water oxidation, 59 Structure-directing agent, 251 Water splitting, 89 Sub-bandgap states, 186 Work function, 46 Wurtzite, 68 378 Index

X Zinc indium sulﬁde, 63 X-ray photoelectron spectroscopy, 359 Zinc oxide, 60, 142, 165, 175, 201, 249, 327, X-Ray scattering, 326 352, 357 Zinc selenide, 201, 206, 350 Z Zinc sulﬁde, 18, 130, 208, 323, 328 Zeolite, 50, 62, 81, 132, 143, 271 Zirconate, 258 Zinc blende, 68 Z-scheme, 89, 129, 145