Exploration of synthetic routes for nano-structured materials: Their applications in catalysis
Thesis submitted to BHARATHIDASAN UNIVERSITY for the award of the degree of DOCTOR OF PHILOSOPHY IN CHEMISTRY
by
S. Harish, M.Sc
(Ref. No: 38344/Ph.D.1./Chemistry/F.T/April 2009)
ELECTRODICS AND ELECTROCATALYSIS DIVISION CSIR-CENTRAL ELECTROCHEMICAL RESEARCH INSTITUTE KARAIKUDI - 630 006, INDIA
(January, 2013)
DECLARATION
I, Mr S. Harish, hereby declare that the thesis entitled “Exploration of synthetic routes for nano-structured materials: Their applications in catalysis” has been originally carried out by me at Central Electrochemical Research Institute (CSIR–CECRI), Karaikudi,
India under the guidance of Dr James Joseph, Principal Scientist, CSIR–CECRI, Karaikudi,
India. The content of this thesis or part thereof has not been submitted to any other university or institute for the award of any Degree, Diploma or other similar title.
Place: Karaikudi (S. Harish)
Date:
i
CERTIFICATE
This is to certify that the thesis entitled “Exploration of synthetic routes for nano- structured materials: Their applications in catalysis” submitted by Mr. S. Harish is a bonafide record of research work done by him for the degree of Doctor of Philosophy under my guidance at CSIR-Central Electrochemical Research Institute (CSIR-CECRI), Karaikudi, India.
The content of this thesis or part thereof has not been submitted to any other university or institute for the award of any Degree, Diploma or other similar title.
Place: Karaikudi (Dr James Joseph)
Date: Research Supervisor
ii ACKNOWLEDGEMENTS
I wish to express my sincere gratitude to many people who brought my dream of successfully completing this thesis. During this period, I have enjoyed my journey in research and learned from each and every individual whomever I met. I would like to thank my mentor Dr James Joseph, for giving me the opportunity to work on a challenging and interesting research project. My doctoral study would have been very difficult without his intelligent guidance, motivation and constant encouragement. I am grateful to Prof. C. Coutanceau for providing me an opportunity to work in his group at University of Poitiers, France for about six months. I appreciate his constant support during that period. Special thanks to Prof. S. Baranton for helping me in all the experiments during my stay. I also thank my lab mates in Poitiers, France for their scientific discussions and assistance in my research. It was a great experience and memorable days in my research career. I express my sincere thanks to Prof. A. K. Shukla and Dr V. Yegnaraman, former Directors of CSIR-CECRI and also to Dr. Vijayamohanan K. Pillai, Director of CSIR-CECRI for their encouragement, motivation and providing all the necessary facilities required for my research work. I wish to thank Dr K. L. N. Phani, with whom I have started my research career and it is his strong motivation, encouragement and realization of excitement in research every now and then helped me to enroll for Ph.D. My special thanks to Dr J. Mathiyarasu who helped me in different aspects from the date I entered CSIR-CECRI. His constant scientific as well as personal support helped me to join and pursue the doctoral work. I would like to thank Dr A. Ilangovan, Bharthidasan University, Trichy and Dr G. Chandramohan, A. V. V. M. Pushpam college, Poondi who are my doctoral committee members, for illuminating discussions, constant support and encouragement. I am thankful to Dr N. Kalaiselvi and Dr A. S. Prakash for their help in carrying out some experiments related to Li ion battery for my materials and the corresponding scientific discussions.
iii My heartfelt thanks are due to Dr C. Sivakumar, Dr S. Senthil kumar and Dr V. Ganesh, Scientists of EEC Division for their friendship and scientific discussions. I am also extending my sincere thanks to all the EEC Division members and my lab mates Mr. C. Jeyabharathi, Mr. T. Nareshkumar, Mr. S. Boopathi, Mr. S. Anandhakumar, Mr. A. V. Narendrakumar, Ms A. Ananthi, Mr. P. Esakki Karthik, Mr I. Maheswaran, Mr T.V. Vineesh, Mr N. Sreekanth, Ms K. Dhanalakshmi and Ms T. Sowmya who have helped me in all possible ways during my research work. Above all, it is my duty to pay sincere thanks to my parents who had not only encouraged me and allowed me to stay away from the home for a longer time which helped me to pursue my research without any hesitation. Importantly, I show my gratitude to almighty for giving a sound health to my parents and also to me. I am also very thankful to all my relatives for their moral support and encouragement throughout. I express sincere thanks to my lecturers Dr K. Subramani and Dr K. Prem Nazir of Islamiah College, Vaniyambadi for their motivation and encouragement. I take this opportunity to thank my good friends Mr. M. S. Chandrasekar, Dr S. Prakash, Mr. K. Firoz Babu, Mr. Gangulibabu, Mr. R. Ravikumar and Dr K. Sundaram for staying with me closely not only during the happy moments but also whenever I am sad, down from my performance and in necessitate time. I also extend heartfelt thanks to all my friends of CSIR- CECRI. Financial support received from the prestigious agencies like Council of Scientific and Industrial Research (CSIR), Department of Science and Technology (DST) and French Embassy in India is greatly acknowledged. It is impossible to thank each and everyone who directly or indirectly helped me during my research career and hence I would like to place a record of gratitude to all those people.
(S.Harish)
iv
Abstract
In this thesis, novel synthetic routes in the preparation of nanostrucutred materials and their application in catalysis were explored. Nanostructured materials were synthesized using sol-gel, microwave activation and solid state decomposition methods. These materials were characterized using microscopy, spectroscopy and electrochemical techniques. Sol-gel method was followed to synthesize Au, Ag and AuAg alloy nano sols and their role on catalytic activity in chemical reduction of 4-nitrophenol was explored. Microwave assisted polyol method was studied for synthesize of carbon supported Pt, Ru and PtRu alloy nanoparticles and their electro- catalytic activity towards methanol oxidation. In-situ FT-IR technique was used to explain the mechanism of methanol oxidation. Solid state decomposition of Co 3[Co(CN) 6]2 was investigated to synthesize phase pure Co 3O4 and possible use of Co 3O4 as an anode material in Li ion battery was studied. The interaction between chloro metalates (of Au, Pt, Pd) and cyano metalates (of Fe, Ru) were investigated and this interaction was found to be a key step in the preparation of carbon supported alloy nanoparticles. Cyanogel processing in presence of carbon followed by thermal decomposition was exploited for the formation of carbon supported bimetallic alloy. Carbon supported bimetallic alloys were used as an electrocatalyst in methanol oxidation and oxygen reduction reactions.
v
ி , ி ீ
ி ி ி .
, - ி, ீ ி
, ீ ,
ி, ி . Au , Ag Au-Ag
, - ி ி , 4
ீ ி ி .
ி Pt, Ru Pt-Ru
ீ ி ,
ி ி . ி
ி , ி
, ி ி .
Co 3[Co(CN) 6]2 ி Co 3O4 Li
ி ி ,
ி ி . ி
(Au, Pt, Pd ) (Fe, Ru) ி
ி Pt-Ru Pd-Fe
ி .
ி .
vi List of Abbreviations
TOA-Br Tetraoctylammonium bromide
MPCs Monolayer- protected clusters
SDS Sodium dodecylsulfate
CMC Critical Micelle Concentration
BPDC 2– 4, 4 ′-biphenyldicarboxylate
UV-Vis Ultraviolet visible
SPR Surface plasmon resonance
FT-IR Fourier Transform Infrared
SNIFTIR Subtractively Normalized Fourier Transform Interfacial Infra Red
TGA Thermo-gravimetric analysis
XRD X-ray diffraction
FWHM Full Width at Half Maximum
XPS X-ray Photo electron Spectroscopy
SAED Selected Area Electron Diffraction
TEM Transmission Electron Microscopy
SEM Scanning Electron Microscopy
EDS Energy Dispersive x-ray Spectroscopy
EIS Electrochemical Impedance Spectroscopy
3–APS 3-AminoPropyl trimethoxy Siloxane
PEMFC Proton Exchange Membrane Fuel Cell
DMFC Direct Methanol Fuel Cell
RHE Reversible Hydrogen Electrode
vii upd Under Potential Deposition
MW Microwave
LIBs Li-ion batteries ppm Parts Per Million
SEI Solid Electrolyte Interface
GC Glassy Carbon
Pd-HCF Palladium hexacyanoferrate
MeOH Methanol fcc face centred cubic
viii Table of contents Page No.
Abstract v
List of abbreviations vii
CHAPTER: I- Introduction 1.1 General Introduction 01
1.2 General synthesis of nanoparticles 02
1.3 Synthesis of metal nanoparticles 06
1.4 Synthesis of bimetallic nanoparticles 10
1.5 Synthesis of metal oxides nanoparticles 13
1.6 Synthesis of carbon supported metal nanoparticles 16
1.7 Scope and objective of the present work 19
References 21
CHAPTER: II - Characterization techniques 2.1 Introduction 31
2.2 UV-Visible spectroscopy 31
2.3 FT-IR spectroscopy 33
2.4 Insitu- Fourier transform infra red spectroscopy 34
2.5 Thermogravimetric analysis 35
2.6 X-ray diffraction 36
2.7 X-ray photo electron spectroscopy 37
2.8 Transmission electron microscopy 37
2.9 Scanning electron microscopy 39
2.10 Cyclic voltammetry 39 2.11 Electrochemical impedence spectroscopy 43
References 44
CHAPTER: III Role of pH in the synthesis of Au, Ag and their alloy nanosols using 3- aminopropyl siloxane as stabilizer and studying the catalytic activity towards 4-nitrophenol reduction 3.1 Introduction 46
3.2 Experimental procedures 50
3.3 Results and discussion 52
3.4 Conclusions 63
References 64
CHAPTER: IV Microwave assisted polyol method for the synthesis of Pt/C, Ru/C and PtRu/C nanoparticles and their application in electro-oxidation of methanol 4.1 Introduction 66
4.2 Experimental procedures 68
4.3 Results and discussion 72
4.4 Conclusions 96 References 96
CHAPTER: V
Porous Co 3O4 derived from a unique precursor Co 3[Co(CN) 6]2 and its application as an anode in Li-ion battery 5.1 Introduction 101
5.2 Experimental procedures 103
5.3 Results and discussion 105
5.4 Conclusions 119 References 120
CHAPTER: VI Investigations on the interaction between chlorometallate and cyanometallate and synthesis of bimetallic alloys from cyanogels for electrocatalytic applications 6.1 Introduction 123
6.2 Experimental procedure 128
6.3 Results and discussions 131
6.4 Conclusions 162 References 164
7.0 Summary and scope for future work 169
8.0 List of publications 175 1.1 General Introduction
“Nanotechnology” today is considered as most promising field because nanotechnology has changed the perspective of the basic science disciplines (physics, chemistry and biology). Now the researchers are working in the inter-disciplinary fields under the roof of nanoscience and nanotechnology. Term ‘Nano’ was derived from Greek origin which means dwarf. On the length scale, nano is one billionth of a meter. The use of nanomaterials was documented much earlier time (Romans period). Some of the famous examples are Lycurgus Cup (shown in fig 1.1A), consisting gold and silver alloy nanoparticles [1], ‘Damascus Sword’ (shown in fig 1.1B) containing the nanoscale carbon particles [2] etc.,
Fig. 1.1 A) The Lycurgus Cup, B) the Damascus Sword (Pictures are reproduced from the ref 1 and 2)
Faraday’s work on “Experimental relations of gold (and other metals) to light” published in Philosophical Transactions in 1857 is remarkable because the birth of modern colloid science taken place. He reported the formation of deep ruby coloration of colloidal
Au by reduction of an aqueous solution of NaAuCl 4 using phosphorus in CS 2 (a two-phase system). He concluded that the gold was dispersed in the liquid in a very finely divided form and also noted that “The state of division of these particles must be extreme; they have not as
- 1 - yet been seen by any power of the microscope” [3, 4]. After the discovery of electron microscopy, researchers investigated gold colloid prepared by Faraday and found that distribution of particle size from 3-30 nm.
Materials in the micrometer scale exhibit properties mostly similar to that of bulk form; however materials in the nanometer scale exhibit properties distinctively different from their bulk counter part. Metal aggregates cannot be simply considered as a part of the block of a metal as implied by the term divided metal [5]. This is explained on the basis of the band theory. The electronic structure of nano-materials undergoes modification results in the development of discrete energy levels in contrast to continuous energy level in bulk materials. It also observed that metal become semi-conductors and semi-conductors turn insulators at nanoscale [6]. In the nanosize regime, the surface to volume ratio increases markedly with the decrease in size. A nanoparticle of 10 nm diameter would have ~ 10% of atoms on the surface, compared to 100% when the diameter is 1 nm [7].
1.2 General synthesis of nanoparticles
The nanoparticle synthesis can be broadly classified into two types
i) Top down approach
ii) Bottom up approach
In the top down approach, the etching or slicing or cutting of the bulk material was carried out to synthesis nano-particles (shown in fig 1.2). High energy ball milling, [9-12] mechnochemcial processing [13, 14], etching [15], laser ablation [16, 17] are some of the well known examples in top down approach.
- 2 -
Figure 1.2 Schematic representations of top down and bottom up approach (Picture is reproduced from the ref.8)
In bottom up approach, nanoparticles are synthesized by atom by atom or molecule by molecule or cluster by cluster. The schematic picture is shown in fig 1.2. The most common methods in bottom up approach are chemical reduction [18-20], microemulsion [21-
24], electrochemical [25-27] etc.
- 3 - Top down approach suffer from the need to remove large amounts of material and difficulty in achieving the narrow size distribution and control over the shape of the particles.
While in bottom up approach, the real spirit of nanoscience exists because the materials are developed from the nature’s smallest building blocks like atoms or molecules. The size of the nanoparticle in bottom up approach is controlled by ligands which also act as a stabilizer.
Stabilization of nanoparticles can occur through two modes:
1. Electrostatic repulsion
2. Steric repulsion.
In the first scenario, subsequent to their reductive preparation, the particles are surrounded by an electric double layer arising due to adsorption of reactant ions on the surface of nanoparticles. This results in two forces acting on nanoparticles,
1. Van der Waals force of attraction between metal cores
2. Electrostatic force of repulsion (potential energy) due to charged ions on the
surface.
Stability of nanoparticles is dependant on the combination of these two forces.
Fig. 1.3 shows graph of potential energy versus distance from the surface of spherical particle. At a distance far from the surface, both van der Waals attraction potential and electrostatic repulsion potential is zero. Near the surface, a minimum is observed in potential energy due to van der Waals attraction. At a distance not very far away from the surface where electric repulsion dominates the van der Waal attraction potential and the combination of these two opposing forces leads to maxima in the energy curve. This maximum is known as repulsive barrier. If the barrier is greater than certain value, two particles cannot overcome the barrier and thus agglomeration is prevented. Electrostatic stabilization is kinetic
- 4 - stabilization process and it is useful only in the case of dilute solutions. Addition of electrolytes screens the double layer charge leading to aggregation. As can be noticed, such stabilization occurring due to electronic repulsion is highly dependent on several factors and the ideal condition for most stable dispersion can be achieved in a very narrow window [28].
A B
Fig.1.3 (A) Electrostatic and (B) Steric stabilization of colloidal nanoparticles (reproduced from ref.28).
Steric stabilization of nanoparticles can be achieved by co-ordination of organic molecules on the surface of nanoparticles, which act as capping ligands. In this way nanoparticle cores are separated from each other and agglomeration is prevented. Here, we have tabulated some of the common ligands used in the synthesis of metal nanoparticles
(shown in table-1). The bottom up approach is found to be more convenient in controlling the particle size compare to top down approach. Hence, we will discuss about the general synthetic methods in bottom up approach for metal, bimetallic and metal oxide nanoparticles.
- 5 - Table -1: Ligands used as a stabilizer in various nanoparticles
Thiols Au [29-31], Ag [32-34] Pt [35], Pd [36] Amines Au [37-39], Ag [40], Pt [41], Pd [42] Carboxylic acid Ag [43-45], magnetic nanoparticles (Co, Ni,Fe) [46, 47] Phosphines Au [48, 49] Silanes Metals [50, 51], Metal oxide [52]
1.3 Synthesis of metal nanoparticles
There are numerous methods available in the literature for the synthesis of metal nanoparticles. Here, we are presenting the well-known methods in the preparation of metal nanoparticles.
1.3.1 Simple reduction method
The reduction of metal ions in the presence of different kind of stabilizers (like citric acid, polyvinyl pyrolidone, cetyl trimethyl ammonium bromide etc.,) has been widely used procedure for the synthesis of metal nanoparticles [6, 18, 53]. The role of stabilizing agent is to prevent aggregation or improve the chemical stability of the formed nanoparticles.
Colloidal methods are advantageous because of the following reasons
i) Simple experimentation and sophisticated instrument is not necessary.
ii) Solution-based processing and assembly can be readily implemented
iii) Large scale synthesis of nanoparticles can be easily carried out [53, 54]
One of the famous methods of this kind is the synthesis of gold nanoparticles by citrate procedure which was reported by Turkevich et al in 1951 [55]. This synthetic method is widely studied [56-58] and this method of synthesis was also named as “Turkevich method”. The method of preparation is very simple because only three components involved
- 6 - in the synthesis namely, auric acid, citric acid and water. Briefly, the synthetic procedure is auric acid is heated on a hot plate and while solution gets boiled citric acid was added which acts as a reducing agent and stabilizer. The colour of the solution changes immediately to deep red color and the hot plate is turned off. The detailed procedure and other information was clearly given in the Ref. [59]
1.3.2 Biphase method (Brust method)
Two-phase method (or popularly called Brust Procedure) has been extensively used for the generation of very small nanoparticles (1–5 nm) with narrow dispersity [60]. The particles are stabilized by a thiol ligand. Samples generated with the two-phase method are stable for long periods of time when dry and can easily be re-dispersed in many organic solvents. In the first report by Brust et al, the gold salt is first transferred to the organic phase using phase transfer catalysts called tetraoctylammonium bromide (TOA-Br) with thiol compounds. Then, sodium borohydride is added to the aqueous phase. The formation of nanoparticles is indicated by the generation of orange to deep brown color in the organic phase. The ratio of the concentration of gold and thiol, and the reaction temperature, determine the particle size and dispersity. This synthetic procedure is often referred to as generation of monolayer- protected clusters (MPCs) due to the monolayer coverage of the sulfur groups and the small size of nanoparticles generated.
Many improvements to this synthesis procedure have been reported to generate small monodisperse gold nanoparticles and also extended to other metal nanoparticles [20, 61-65]
Fig 1.4 refers to the transfer of different metal ions to the organic medium (toluene) and subsequently reduced by using NaBH 4 to form respective metal nanoparticles [65]
- 7 -
Fig 1.4 Photographs showing the successful transfer of Co (II), Os (III), Rh (III),
Ru (III), Au (III) and Ir (III) metal ions from the aqueous phase to toluene
(Reproduced from ref. 65)
1.3.3 Microemulsion method
This is a relatively new technique for the preparation of nanoparticles, and has received considerable interest in recent years. This method has been successfully applied to synthesize metal, alloy, metal suphides and metal oxide nanoparticles [66-69]. A literature survey depicts that the ultrafine nanoparticles in the size range between 5-50 nm can be easily prepared by this method. This technique uses an inorganic phase as water-in-oil microemulsions which are isotropic liquid media with nanosized water droplets that are dispersed in a continuous oil phase. In general, microemulsion consists of, at least, a ternary mixture of water, a surfactant or a mixture of surface-active agents and oil. The classical examples for emulsifiers are sodium dodecylsulfate (SDS) and aerosol bis(2- ethylhexyl)sulfosuccinate (AOT). The surfactant (emulsifier) molecule stabilizes the water droplets, which have polar head and non-polar organic tails. The organic (hydrophobic) portion faces towards the oil phase and the polar (hydrophilic) group towards water. In diluted water (or oil) solutions, the emulsifier dissolves and exists as a monomer, but when
- 8 - its concentration exceeds a certain limit called the critical micelle concentration (CMC), the molecules of emulsifier associate spontaneously to form aggregates called micelles. These micro-water droplets then form nano-reactors for the formation of nanoparticles [69]. The nanoparticles formed usually have mono-disperse properties. One method of formation consists of mixing of two microemulsions or macroemulsions and aqueous solutions carrying the appropriate reactants in order to obtain the desired particles. The interchange of the reactants takes place during the collision of the water droplets in the microemulsions. The interchange of the reactant is very fast so that for the most commonly used microemulsions, it occurs just during the mixing process. The reduction, nucleation, and growth occur inside the droplets, which controls the final particle size. The chemical reaction within the droplet is very fast, so, the rate-determining step will be the initial communication step of the microdroplets with different droplets. The rate of communication has been defined by a second-order communication-controlled rate constant and represents the fastest possible rate constant for the system. The reactant concentration has a greater influence on the reduction rate. The rate of both nucleation and growth are determined by the probabilities of the collisions between several atoms, between one atom and a nucleus, and between two or more nuclei. Once a nucleus forms with the minimum number of atoms, the growth process starts.
For the formation of monodisperse particles, all of the nuclei must form at the same time and grow simultaneously and with the same rate. The method for the preparation of metal nanoparticles within micelles consists of forming two microemulsions, one with the metal salt of interest and the other with the reducing agent and mixing them together. When two different reactants mix, the interchange of the reactants takes place due to the collision of water microdroplets. The reaction (reduction, nucleation and growth) takes place inside the
- 9 - droplet, which controls the final size of the particles (shown in fig 1.5). The interchange of nuclei between two microdroplets does not take place due to the special restrictions from the emulsifier. Once the particle inside the droplets attains its full size, the surfactant molecules attach to the metal surface thus stabilizing and preventing further growth. A diverse range of metal nanoparticles have been prepared by this method including Fe [66, 67] Fe/Au [68], Pt
[70, 71], Ag [72, 73], Pd [74], and Au [75, 76].
Fig. 1.5 Schematic mechanism of metal nanoparticles formation by the microemulsion
approach (Reproduced from ref. 69)
1.4 Synthesis of bimetallic nanoparticles
Bimetallic nanoparticles are synthesized to tune the various properties of the metallic systems which can be in the form of intermetallics or alloys. In general, there will be
- 10 - siginificant enhancement of the properties can be found in bimetallic system compare to their individual metal which is called as synergistic effect. Bimetallic alloys (A xBy) can be synthesized by various methods to yield uniform size and desired composition of metals. The degree of alloying or surface segregation of metal is depends on the synthetic method followed. In alloy nanoparticles, chemical and physical properties may be tuned by varying the composition, atomic ordering and also greatly depends on the size of the nanoparticles
[77]. As the metallic nanoparticle displays the magic number for their stability, alloy nanoparticle displays magic compositions. Surface structures, compositions and segregation properties will determine their chemical reactivity and catalytic activity [77-80]. In this section, we are going to discuss the general methods in the synthesis of bimetallic nanoparticles.
1.4.1 Chemical reduction method
This method is similar to the synthesis of metal nanoparticles (discussed in section in
1.3.1). Bimetallic nanoparticles are obtained by reducing the respective metal salts using different reducing agents like sodium borohydride, hydrazine etc., The formed nanoparticle can exhibit different structures depends on their reduction potential, miscibility gap and stabilizer (ligand) [81]. Bimetallic alloy nanoparticles will be formed, if the miscibility gap is small. The classical example for this type is gold and silver because the miscibility gap is very less. Hence the formation of gold-silver alloy takes place readily during the co-reduction of their respective salts. If the reduction potentials of two metals are different and sufficiently large then there is a possibility in the formation of core-shell structures. The metal species with the highest reduction potential will undergo reduction first and forms as core on which
- 11 - the second metal will be nucleated to form core-shell structure. The order of reduction can be reversed by choosing the stabilizer (ligand) which significantly bonds to the metal with the highest redox potential and hence the inverse core-shell arrangement will be stabilized [82].
In general, co-reduction of Ag and Pd will lead to Pd core and Ag shell nanoparticles because the Pd has the higher redox potential. However, the core-shell structures can be reversed by reducing in the solution containing ammonia because there will be stronger binding of NH 3 to Pd compare to Ag [81, 82].
1.4.2 Thermal decomposition method
The thermal decomposition of organometallic compounds is a classical route to obtain highly monodisperse bimetallic nanoparticles. A typical example is the synthesis of monodisperse FePt nano particles reported by Sun et al [83, 84]. The composition of FePt can be varied by adjusting the molar ratio of Fe(CO) 5 to Pt(acetylacetonate)2. On decomposition, the rate of decomposition of Fe(CO) 5 is slower compare to
Pt(acetylacetonate)2. Hence Fe 48 Pt 52 , Fe 52 Pt 48 , and Fe 70 Pt 30 nanoparticles were obtained when the Fe(CO) 5/Pt(acetylacetonate)2 ratio was 3:2, 2:1, and 4:1 respectively [83]. In the following years, Sun et al. developed this method and achieved size and shape controlled synthesis of FePt alloys [85-87]
Bimetallic nanoparticles can also be synthesized directly by thermal decomposition of two different organometallic compounds to yield bimetallic nanoparticles [88]. On thermal decomposition, metal carbonyls will tend to decompose at high temperature to form carbon monoxide and corresponding metals due to their thermal instability. The decomposition temperature of various metal carbonyls will be different and hence to obtain the expected
- 12 - composition is difficult. If more than one precursor is used then it is difficult to ensure that two kinds of metal atoms contribute equally to the metal-metal bond formation due to their different reaction kinetics. Hence the reaction kinetics will induce the formation of separate monometallic phase. However, these problems can be overcome if bimetallic compounds are adopted as single-source molecular precursors [89, 90]. For example, FePt nanoparticles were prepared by decomposing a single precursor of Pt 3 Fe 3(CO) 15 [89] and Robinson et al
[90] achieved the syntheses of a series of alloyed nanoparticles by the thermal decomposition of bimetallic carbonyl clusters. Instead of neutral bimetallic carbonyl clusters, they used
− 2− molecular bimetallic carbonyl cluster anions such as [FeCo 3(CO) 12 ] , [Fe 3Pt 3(CO) 15 ] ,
2− 2− [FeNi 5(CO) 13 ] and [Fe 4 Pt(CO) 16 ] [90]. Because stable metal-metal bonds already present in the nucleus of the cluster, the development of new metal-metal bonds between two metal elements can be well promoted, leading to the formation of alloys with a composition reflecting that of the precursor. Therefore, this method is very effective in generating bimetallic nanoparticles which can avoid phase separation and achieve composition control.
1.5 Synthesis of metal oxides nanoparticles
Metal oxides are found to have wide variety of applications like catalysts, sensors, microelectronic devices, Li ion batteries, solar and fuel cells [91-97]. In Li ion batteries, metal oxides are used as an anode material instead of graphite because metal oxides possess
-1 -1 high theoretical capacity (~ 700 mAh g ) compare to graphite (372 mA h g ) [98]. Co 3O4 was most studied metal oxides as an anode material in Li ion batteries [98, 99]. Co 3O4 can be synthesized by various methods like oxidative precipitation [100], combustion method [101,
102], thermal decomposition [103, 104], hydrothermal method [105, 106] etc. In the
- 13 - following, synthesis of Co 3O4 using oxidation and thermal decomposition methods are discussed.
1.5.1 Oxidative precipitation method
This method is considered as a one of the simple method in the preparation of metal oxide nanoparticles. In this procedure, precipitation of metal oxide is mainly depends on pH of the solution. To obtain homogenously distributed particles, the critical super-saturation of precipitating reactants is necessary to generate nuclei. Further, the nuclei have to generate at a moderate rate for the formation of homogenous particles.
Sugimoto et al [107] synthesized monodispersed cubic Co 3O4 (~0.1 µm edge length) from a cobalt acetate solution by ageing at 100 °C for few hours and they also reported that similar ageing of cobalt nitrate, cobalt suphate, cobalt chloride at 100 °C does not lead to the precipitation of Co 3O4. Further, Co 3O4 nanoparticle formation was strongly depends on the pH, high yield of Co 3O4 was observed at neutral pH whereas in the acidic medium, hydrolysis process is decreased and in the basic medium, oxidation of Co (II) is inhibited. In addition to this, oxygen atmosphere plays vital role because Co 2+ in the solution needs to be
2+ 3+ oxidized to form Co (Co )2O4. The presence of oxygen in the solution also related to the yield of Co 3O4 nanoparticles.
Xu et al [108] reported the salt mediated formation of Co 3O4 nanocubes of uniform size (47 nm) at 95 °C by tailor made precursors to oxidize in highly concentrated NaNO 3
II solution. In aqueous solution, β-Co(OH) 2 and Co (OH) 2-x(NO 3)x.nH 2O are gradually
II III oxidized to Co 1-xCo x(OH) 2(NO 3)x.nH 2O and Co 3O4 through which the final single phase
Co 3O4 nanocubes can be prepared with a prolonged oxidation. The reason for the formation
- 14 - of perfect nanocubes was due to the lowering of oxygen solubility and creation of salt-
(solvent) n diffusion boundary on the surfaces, which retards the cobalt oxidation and alters the normal interfacial growth under non-salted conditions. This method is further examined for controlling the size of the Co 3O4 nanocubes [109]. They found that size of the cube and reaction time is linear. Co 3O4 nanocubes in the range of 10-50 nm was prepared under controlled temperature and time. In addition to that, separation of Co 3O4 nanocubes from the remaining solid products was achieved by acid washing.
1.5.2 Thermal decomposition method
Thermal decomposition was considered as a widely accepted method for the preparation of bulk metal oxides. Here, reactions are completely controlled by diffusion of the atomic or ionic species at high temperatures. However, particles will exhibit broad size distribution and non-uniform composition of metal oxide nanoparticles. In the recent past, researchers found the preparation of Co 3O4 nanostructures by thermal decomposition of organocobalt and cobalt complexes.
Zhan et al [110] prepared porous Co 3O4 hexagonal nano sheets by simple thermal decomposition of Co (OH) 2 hexagonal sheets in air at 300 °C for 1 hour. The porous structures were formed due to intrinsic crystal contraction and gas evolution during the thermal decomposition of Co(OH) 2.
Jin et al [111 ] prepared shape controlled Co 3O4 by thermal decomposition of
Co-BPDC (BPDC 2– = 4, 4′-biphenyldicarboxylate) which is a co-ordination polymer. They have prepared Co-BPDC with three different morphologies (flower, multilayer and sheets) and decomposed at 400 °C in air for 1 hour to yield Co 3O4.
- 15 - 1.6 Synthesis of carbon supported metal nanoparticles
Metal nanoparticles are supported on carbon and mainly used as the electrocatlyst in fuel cell. In order to harvest major electrocatalytic activity, metal nanoparticle needs to be dispersed in the conducting matrix [112-115]. The support material should provide high dispersion and stability. Support material should have good electrical conductivity, high surface area, morphology, porosity, corrosion resistance are important factors to choose a good catalyst support. The interaction of catalyst and support material can influence the activity of the catalyst [116-118].
Carbon is found to be the best catalyst support material for low temperature fuel cells.
Carbon black, activated carbon and Vulcan carbon 72 are extensively used as catalyst supports in which Vulcan carbon 72 was used widely [119]. In the recent past, new carbon materials like carbon nanotubes, aerogel carbon, and mesocarbon are reported as a good catalyst support [120]. The carbon supported metal nanoparticles can be synthesized by different methods and the most common methods are given schematically in the fig 1.6. In the following, impregnation method and polyol method are discussed.
1.6.1 Impregnation method
In this method, metal precursors are impregnated inside the pores of the carbon before the reduction of metal ions and procedure is schematically shown in fig 1.6. Now, the metal ions can be reduced chemically or electrochemically to form carbon supported metal nanoparticles. Most chemical reduction is followed using hydrazine, sodium borohydride or hydrogen gas. While using hydrogen gas, the temperature has to elevate more than 300 °C under inert atmosphere. The particle size and distribution is depends on many factors. The
- 16 - carbon support plays major role in the nanoparticle growth. The pores present in the carbon and its distribution will be a factor for the penetration and wetting of the precursor. Reaction time, kinetics and mass transfer of the reducing agent will also affect the nucleation and growth of the nanoparticles [119].
Fig. 1.6 Schematic illustration for the synthesis of carbon supported Pt
nanoparticles (reproduced from the ref. 119)
Interaction of metal precursor and carbon support can be enhanced by modifying the surface of the carbon. Carbon surface was activated by reaction with sodium hypochlorite to form carboxylic acid on the surface and followed by treatment with ammonia to form ammonium salt. The ammonium groups can be exchanged with metal salt and then reduced to form metallic nanoparticles [121]. Similar to this procedure, carbon can also be treated
+ with aqua regia to form hydroxyl groups on the surface. On treatment with acid, − OH 2 form
- 17 - on the surface which can be exchanged with metal ions and followed by reduction to form carbon supported metal nanoparticles [122]. The major drawback of the impregnation method is the lack of particle size control which can be overcome by using highly ordered mesoporous carbon [123].
1.6.2 Polyol method
Polyol method was also widely used to synthesis metal nanoparticle with controlled size [124]. This method finds to be advantageous over other methods because the solvent ethylene glycol acts as a both reducing agent and stabilizer. Hence, the addition of protecting agent like surfactant is not necessary in the synthesis of metal nanoparticles (shown in fig.
1.6). In the synthesis of Pt/C, Liu et al [125] explained that metal salt dissolved in ethylene glycol solution was refluxed at 120 °C – 170 °C. Ethylene glycol undergoes decomposition to generate insitu reducing species for the reduction of metal ions. In the traditional synthesis, the reduction reaction is activated by temperature higher than 120 °C for several hours [124,
126-128]. Bimetallic alloy nanoparticles can also be prepared by polyol method. Bock et al
[129] prepared PtRu catalyst by dissolving both the metal precursors in an ethylene glycol solution and then refluxed at 160 °C for several hours. They explained the reaction mechanism involves the oxidation of ethylene glycol to aldehyde which is not very stable and are oxidized to glycolic acid or oxalic acid. These carboxylic acids are further oxidized to form carbon dioxide or carbonate in the alkaline medium. Pt and Ru ions are get reduced by electrons donated from these oxidation reactions and hence the reduction of metal ions and particle size of the resultant metal nanoparticle are found to be highly dependent on pH of the medium. The other pathway was also suggested because platinum is known as an excellent
- 18 - catalyst to abstract hydrogen from carbon atoms [130]. Thus oxidation of ethylene glycol involves the abstraction of hydrogen and results in the adsorption of CO on platinum that may further oxidize to CO 2 and reduction of platinum ion takes place.
1.7 Scope and objective of the present work
The present investigation deals with the different synthetic routes to prepare nanostructured materials and to exploit their potential in various application areas. In general, nanoparticles will have high surface energy and tends to undergo aggregation when two particles come closer. Hence the synthesis of stable nanoparticles with controlled size is a real challenge. In this thesis, three methods are followed to prepare the nanomaterials namely sol-gel method, microwave assisted polyol method and solid state decomposition. It has been divided into four chapters and given as follows
In Chapter-3, we explored the role of pH on the stabilization of gold, silver and gold- silver alloy nano sols using 3-aminopropyl trimethoxy siloxane as stabilizer. It is known from the literature that gold, silver, platinum and palladium nanoparticles were prepared by using the 3-aminopropyl trimethoxy siloxane as a stabilizer. The metal nanoparticles are highly stable and exhibits narrow size distribution of 5–7 nm but silver alone was found to be very less stable and exhibits broad size distribution of 2–30 nm. Hence, one of our objectives of this work is to synthesis stable silver nanoparticles with narrow size distribution. Further,
APS stablilsed Au-Ag nano alloy sols are also synthesized by chemical co-reduction. The catalytic activity of gold, silver and gold-silver alloy is studied by choosing the reduction of
4-nitrophenol as a model reaction.
- 19 - In Chapter-4, we have investigated on the preparation of carbon supported metal nanoparticles by polyol method using microwave heating instead of conventional oil bath heating. The synthetic procedure of polyol method was discussed in the section 1.6.2. In polyol method, if we follow conventional oil bath heating then it requires long time duration to complete the reduction process. Hence, our objective of this work is to reduce the time duration of the reduction process by using microwave heating which has following advantages
i) less energy
ii) less time duration
iii) Uniform heating (no thermal convection)
iv) implementation simplicity
We have prepared carbon supported Pt, Ru and PtRu alloy using microwave assisted polyol method and studied the electro-catalytic activity towards electro-oxidation of methanol.
In Chapter-5, porous cobalt oxide was synthesized by thermal decomposition of
Co 3[Co(CN) 6]2 (Prussian blue like compound). Even though metal oxides can be synthesized by various methods, metal oxides derived from decomposition of Prussian blue like compounds finds was found to be simple route and new route for preparing porous structures.
Hence, our objective is to synthesis phase pure porous Co3O4 by thermal decomposition of
Co 3[Co(CN) 6]2. Moreover, the as synthesized porous Co 3O4 have been used as anode materials for Li ion batteries.
- 20 - In Chapter-6, we have studied the nature of interaction between chloro metalate and cyano metalate ions. It is known from the literature that either Prussian blue or cyanogel will form during their interaction. We have chosen following three systems for investigation
System –I: HAuCl 4 and K 3[Fe(CN) 6]/ K 4[Fe(CN) 6]
System –II: H2PtCl 6 and K 3[Ru(CN) 6]
System –III: PdCl 2 and K 3[Fe(CN) 6]/ K 4[Fe(CN) 6]
Our objective in this chapter is to investigate on the interaction between chloro metalate and cyano metalate ions by using various analytical techniques. In addition to this, we have systematically analyzed the reason for the non existence of gold analogue of
Prussian blue in literature. The carbon supported bimetallic alloys are derived from cyanogel route which is hitherto not studied in literature. Carbon supported PtRu and PdFe prepared through cyanogel route is demonstrated as an electrocatalyst in electro-oxidation of methanol and oxygen reduction reactions respectively.
References:
1. Freestone, N. Meeks, M. Sax, C. Higgitt, Gold Bulletin 40 (2007) 270.
2. J. D. Verhoeven, A. H. Pendray, W. E. Dauksch, JOM-J. Min. Met. Mat. S 50 (1998)
58.
3. M. Faraday, Philos. Trans. R. Soc. London 147 (1857)145.
4. P. P. Edwards, J. M. Thomas, Angew. Chem. Int. Ed. 46 (2007) 5480.
5. W. A. de Heer, Rev. Mod. Phys. 65 (1993) 611.
6. G. Schmid, Nanoparticles: from theory to application; Wiley-VCH Weinheim, 2004.
- 21 - 7. C. N. R. Rao, G. U. Kulkarni, P. J. Thomas, P. P. Edwards, Chem. Soc. Rev. 29
(2000) 27.
8. G. Cao, Nanostructures and nanomaterials: synthesis, properties and applications,
Imperial College Press (2004).
9. F. Peng, J. Wang, G. Ge, T. He, L. Cao, Y. He, H. Ma, S. Sun, Mater. Lett. 92 (2013).
65.
10. A. S. Heintz, M. J. Fink, B. S. Mitchell, Adv Mater 19 (2007) 3984.
11. C. M. Poffo, J. C. Lima, S. M. Souza, D. M. Triches, T.A. Grandib, R. S. Biasic, J.
Raman Spectrosc. 41 (2010) 1606.
12. T. D. Shen, C. C. Koch, T. L. McCormick, R. J. Nemanich, J. Y. Huang, J. G. Huang,
J. Mater. Res. 10 (1995) 139.
13. A. Stankovic, L. J. Veselinovic, S. D. Skapin, S. Markovic, D. Uskokovic, J. Mater.
Sci. 46 (2011) 3716.
14. P. G. McCormick, T. Tsuzuki, J. S. Robinson, J. Ding, Adv. Mater. 13 (2001) 1008.
15. S. Takahashi, K. Suzuki, M. Okano, M. Imada, T. Nakamori, Y. Ota, K. Ishizaki, S.
Noda, Nat. Mater. 8 (2009) 721.
16. M. F. Becker, J. R. Brock, H. Cai, D. E. Henneke, J. W. Keto, J. Y. Lee, W. T.
Nichols, H. D. Glicksman, Nanostruct. Mater. 10 (1998) 853.
17. F. Mafune, J. Y. Kohno, Y. Takeda, T. Kondow, J. Phys. Chem. B 107 (2003) 4218.
18. R. Sardar, A. M. Funston, P. Mulvaney, R. W. Murray, Langmuir 25 (2009) 13840.
19. Y. Tan, X. Dai, Y. Li, D. Zhu, J. Mater. Chem. 13 (2003) 1069.
20. L. M. L-Marzán, Chem. Commun. 49 (2013) 16.
- 22 - 21. M. A. L-Quintela, C. Tojo, M. C. Blanco, L. G. Rio, J. R. Leis, Curr. Opin. Colloid
Interface Sci. 9 (2004) 264.
22. I. Capek, Adv. Colloid Interface Sci. 110 (2004) 49.
23. C. V-Vazquez, M. B-Lopez, A. Mitra, M. A. L-Quintela, J. Rivas, Langmuir 25
(2009) 8208.
24. S. Q. Qiu, J. X. Dong, G. X. Chen, J. Colloid Interface. Sci. 216 (1999) 230.
25. N. Vilar-Vidal, M. C. Blanco, M. A. L-Quintela, J. Rivas, C. Serra, J. Phys. Chem. C
114 (2010) 15924.
26. M. T. Reetz, W. Helbig, J. Am. Chem. Soc. 116 (1994) 7401.
27. M. J. R-Vazquez, M. C. Blanco, R. Lourido, C. V-Vazquez, E. Pastor, G. A. Planes,
J. Rivas, M. A. L-Quintela, Langmuir 24 (2008) 12690.
28. G. Schmid , J. S. Bradley, Clusters and Colloids VCH, Weinheim (1994).
29. M. Brust, J. Fink, D. Bethell, D. J. Schiffrin, C. Kiely, J. Chem. Soc. Chem.
Commun. (1995) 1655.
30. D. V. Leff, P. C. Ohara, J. R. Heath, W. M. Gelbart, J. Phys. Chem. 99 (1995) 7036.
31. C. S. Weisbecker, M. V. Merritt, G. M. Whitesides, Langmuir 12 (1996) 3763.
32. A. B. Smetana, K. J. Klabunde, C. M. Sorensen, J. Colloid Interface Sci. 284 (2005)
521.
33. T. Linnert, P. Mulvaney, A. Henglein, J Phys Chem 97 (1993) 679.
34. R. C. Doty, T. R. Tshikhudo, M. Brust, D. G. Fernig, Chem. Mater. 17 (2005) 4630.
35. K. V. Sarathy, G. Raina, R. T. Yadav, G.U. Kulkarni, C. N. R. Rao, J. Phys. Chem. B
101 (1997) 9876.
36. S. C. Cook, J. D. Padmos, P. Zhang, J. Chem. Phys. 128 (2008) 154705.
- 23 - 37. M. Green, P. O. Brien, Chem. Commun. (2000) 183.
38. A. Kumar, S. Mandal, P. R. Selvakannan, R. Pasricha, A. B. Mandale, M. Sastry,
Langmuir 19 (2003) 6277.
39. S. Datar, M. Chaudhari, M. Sastry, C. V. Dharmadhikari, Appl. Surf. Sci. 253 (2007)
5109.
40. A. Kumar, H. Joshi, R. Pasricha, A. B. Mandale, M. Sastry, J. Colloid Interface Sci.
264 (2003) 396.
41. A. Kumar, H. M. Joshi, A. B. Mandale, R. Srivastava, S. D. Adyanthaya, R. Pasricha,
M. Sastry, J. Chem. Sci. 116 (2004) 293.
42. E. Ramirez, S. Jansat, K. Philippot, P. Lecante, M. Gomez, A. M. M-Bulto, B.
Chaudret, J. Organometallic Chem. 689 (2004) 4601.
43. W. Wang, S. Efrima, O. Regev, Langmuir 14 (1998) 602.
44. W. Wang, X. Chen, S. Efrima, J. Phys. Chem. B 103 (1999) 7238.
45. T. Bala, A. Swami, B. L. V. Prasad, M. Sastry, J. Colloid Interface Sci. 283 (2005)
422.
46. N. Wu, L. Fu, M. Su, M. Aslam, K. C. Wong, V. P. Dravid, Nano Lett. 4 (2004) 383.
47. D. S. Sidhaye, T. Bala, S. Srinath, H. Srikanth, P. Poddar, M. Sastry, B. L. V. Prasad,
J. Phys. Chem. C 113 (2009) 3426.
48. G. Schmid, R. Pfeil, R. Boese, F. Bandermann, S. Meyer, G. H. M. Calis, W. A
Vandervelden, C. B-Recueil, 114 (1981) 3634.
49. G. Schmid, N. Klein, L. Korste, U. Kreibig, D. Schonauer, Polyhedron 7 (1988) 605.
50. T. M. Owens, K. T. Nicholson, M. M. B. Holl, S. Suzer, J. Am. Chem. Soc. 124
(2002) 6800.
- 24 - 51. T. M. Owens, K. T. Nicholson, D. R. Fosnacht, B. G. Orr, M. M. B. Holl, Langmuir
22 (2006) 9619.
52. A. Y. Fadeev, T. J. McCarthy, J. Am. Chem. Soc. 121 (1999) 12184.
53. A. R. Tao, S. Habas, P. Yang, Small 4 (2008) 310.
54. Y. Lu, W. Chen, Chem. Soc. Rev. 41(2012) 3594.
55. J. Turkevich, J. Hillier, P. C. Stevenson, Discuss. Faraday Soc. 11 (1951) 55.
56. J. Kimling , M. Maier , B. Okenve , V. Kotaidis , H. Ballot , A. Plech, J. Phys.
Chem. B 110 (2006) 15700.
57. X. Ji, X. Song, J. Li, Y. Bai, W. Yang, X. Peng, J. Am. Chem. Soc. 129 (2007)
13939.
58. J. Polte, T. T. Ahner, F. Delissen, S. Sokolov, F. Emmerling, A. F. Thünemann, R.
Kraehnert, J. Am. Chem. Soc. 132 (2010) 1296.
59. A. D. McFarland, C. L. Haynes, C. A. Mirkin, R. P. Van Duyne, H. A. Godwin, J.
Chem. Educ. 81 (2004) 544A.
60. M. Brust, M. Walker, D. Bethell, D.J. Schiffrin, R.J. Whyman, J. Chem. Soc. Chem.
Commun (1994) 801.
61. J. M. McMahon, S. R. Emory, Langmuir 23 (2007) 1414.
62. F. Osterloh, H. Hiramatsu, R. Porter, T. Guo, Langmuir 20 (2004) 5553.
63. P. Sandstrom, M. Boncheva, B. Akerman Langmuir 19 (2003) 7537.
64. H. Y-Lelievre, J. Desbiens, A. M. Ritcey, Langmuir , 23 (2007) 2843.
65. J. Yang, E. H. Sargent, S. O. Kelley, J. Y. Ying, Nat. Mater 8 (2009) 683.
66. M. A. L-Quintela, J. Rivas, J. Colloid Interface Sci. 158 (1993)446.
- 25 - 67. A. K. Lodhi, B. H. Robinson, T. Towey, C. Hermann, W. Knoche, U. Thesing, The
structure, dynamics and equilibrium properties of colloidal systems, Kulwer Acad.
publ., Dordrecht, 324 (1990).
68. C. T. Seip, R. J. O'Connor, Nanostruct. Mater. 12 (1999) 183.
69. M. A. Malik, M. Y. Wani, M. A. Hashim, Arab. J. Chem. 5 (2012) 397.
70. P. R. Vanrheenen, M. J. Mckelvy, W. S. Glaunsinger, J. Solid. State Chem. 67 (1987)
151.
71. H. H. Ingelsten, R. Bagwe, A. Palmqvist, M. Skoglundh, C. Svanberg, K. Holmberg.
J. Colloid. Interface Sci.241 (2001) 104.
72. P. Barnickel, A. Wokaun, W. Sager, H. F. Eicke, J. Colloid. Interface Sci. 148 (1992)
80.
73. A. Manna, B. D. Kulkarni, Chem. Mater. 9 (1997) 3032.
74. R. Kosydar, M. Góral, J. Gurgul, A. Drelinkiewicz, Catal. Commun. 22 (2012) 58.
75. J. B. Nagy, Colloid Surf. 35 (1989) 201.
76. K. Lemke, C. Prietzel, J. Koetz, J. Colloid Interface Sci. (2012), http://dx.doi.org
/10.1016/j.jcis.2012.11.057
77. R. Ferrando, J. Jellinek, R. L. Johnston, Chem. Rev. 108 (2008) 846.
78. A. V. Ruban, H. L. Skriver, J. K. Norskov, Phys. Rev. B 59 (1999) 15990.
79. G. Bozzolo, J. Ferrante, R. D. Noebe, B. Good, F. S. Honecy, P. Abel, Comput.
Mater. Sci. 15 (1999) 169.
80. A. M. Molenbroek, S. Haukka, B. S. Clausen, J. Phys. Chem. B 102 (1998) 10680.
81. R. Ferrando, J. Jellinek, R. L. Johnston, Chem. Rev. 108 (2008) 845.
82. D. V. Goia, E. Matijevic, New J. Chem. 22 (1998) 1203.
- 26 - 83. S. H. Sun, C. B. Murray, D. Weller, L. Folks, A. Moser, Science 287 (2000 ) 1989.
84. S. Sun, Adv. Mater. 18 (2006) 393.
85. M. Chen, J. P. Liu, S. Sun, J. Am. Chem. Soc. (2004) 8394.
86. M. Chen, J. Kim, J. P. Liu, H. Fan, S. Sun, J. Am. Chem. Soc. 128 (2006) 7132.
87. C. Wang, Y. Hou, J. Kim, S. Sun, Angew. Chem. Int. Ed. 46 (2007) 6333.
88. H. Bonnemann, R. A. Brand, W. Brijoux, H. W. Hofstadt, M. Frerichs, V. Kempter,
W. M-Friedrichs, N. Matoussevitch, K. S. Nagabhushana, F. Voigts, V. Caps, Appl.
Organometal. Chem. 19 (2005) 790.
89. R. D. Rutledge, W. H. Morris, M. S. Wellons, Z. Gai, J. Shen, J. Bentley, J. E. Wittig,
C. M. Lukehart, J. Am. Chem. Soc.128 (2006) 14210.
90. I. Robinson, S. Zacchini, L. D. Tung, S. Maenosono, N. T. K. Thanh, Chem. Mater.
21 (2009) 3021.
91. K. Nagaveni, G. Sivalingam, M. S. Hegde, G. Madras, Appl. Catal. B 48 (2004) 83.
92. F-M. Garcia, M-A. Arias, J. C. Hanson, J. A. Rodriguez, Chem. Rev. 104 (2004)
4063.
93. M. Gratzel, Inorg. Chem. 44 (2005) 6841.
94. B. S. Archanjo, G.V. Silveira, A-M. B. Goncalves, D. C. B. Alves, A. S. Ferlauto, R.
G. Lacerda, B. R. A. Neves, Langmuir 25 (2008) 602.
95. O. Harnack, C. Pacholski, H. Weller, A. Yasuda, J. M. Wessels, Nano Lett. 3 (2003)
1097.
96. J. W. Long, B. Dunn, D. R. Rolison, H. S. White, Chem. Rev. 104 (2004) 4463.
97. M. V. D. Bossche, S. McIntosh, Chem. Mater. 22 (2010) 5856.
98. P. Poizot, S. Laruelle, S. Grugeon, L. Dupont, J-M. Tarascon, Nature 407 (2000) 496.
- 27 - 99. D. Larcher, D. Bonnin, R. Cortes, I. Rivals,L. Personaz, J. M. Tarascon,
J. Electrochem. Soc. 150 (2003) A1643.
100. G. Binotto, D. Larcher, A. S. Prakash, R. H. Urbina, M. S. Hegde, J-M. Tarascon
Chem. Mater. 19 (2007) 3032.
101. M. C. G. Merino, M. Palermo, R. Belda, M. E. F. De Rapp, G. E. Lascalea, P. G.
Vázquez Procedia Materials Science 1 (2012 ) 588.
102. K. Venkateswara Rao, C. Sunandana, Solid State Commun. 148 (2008) 32.
103. B. Wang, T. Zhu, H.B. Wu, R. Xu, J. S. Chen, X.W. D. Lou, Nanoscale 4 (2012)
2145.
104. S. Xiong, J. S. Chen, X. W. Lou, H. C. Zeng, Adv. Funct. Mater. 22 (2012) 861.
105. Y. Jiang, Y. Wu, B. Xie, Y. Xie, Y.Qian, Mater. Chem. Phys. 74 (2002) 234.
106. X-L. Huang, X. Zhao, Z-L. Wang, L-M. Wang, X-B. Zhang, J. Mater. Chem. 22
(2012) 3764.
107. T. Sugimoto, E. J. Matijevic, Inorg. Nucl. Chem. 41 (1979) 165.
108. R. Xu, H. C. Zeng, J. Phys. Chem. B 107 (2003) 926.
109. J. Feng, H. C. Zeng, Chem. Mater. 15 (2003) 2829.
110. F. Zhan, B. Geng, Y. Guo, Chem. Eur. J. 15 (2009) 6169.
111. L-N. Jin, Q. Liu, W-Y. Sun, CrystEngComm 14 (2012) 7721.
112. K. Kinoshita, J. Electrochem. Soc. 137 (1990) 845.
113. M. Watanabe, H. Sei, P. Stonehart, J. Electroanal. Chem. 261 (1989) 375.
114. K. Yahikozawa, Y. Fujii, Y. Matsuda, K. Nishimura, Y. Takasu, Electrochim. Acta
36 (1991) 973.
- 28 - 115. A. Kabbabi, F. Gloaguen, F. Andolfatto, R. Durand, J. Electroanal. Chem. 373 (1994)
251.
116. R.L. Augustine, Heterogeneous Catalysis for the Synthetic Chemists, M. Dekker,
New York (1996).
117. K.-W. Park, K.-S. Ahn, Y.-C. Nah, J.-H. Choi, Y.-E. Sung, J. Phys. Chem. B 107
(2003) 4352.
118. H.M. Villullas, F.I. M-Costa, L.O.S. Bulhoes, J. Phys. Chem. B 108 (2004) 12898.
119. K-Y. Chan, J. Ding, J. Ren, S. Cheng, K. Y. Tsang, J. Mater. Chem.14 (2004) 505.
120. E. Antolini, Appl.Catal. B- Environ. 88 (2009) 1.
121. C. Coutanceau, S. Brimaud, C. Lamy, J.-M. Leger, L. Dubau, S. Rousseau, F. Vigier,
Electrochim. Acta 53 (2008) 6865.
122. F. Vigier, C. Coutanceau, A. Perrard, E.M. Belgsir, C. Lamy, J. Appl. Electrochem.
34 (2004) 439.
123. S. H. Joo, S. J. Choi, I. Oh, J. Kwak, Z. Liu, O. Terassaki, R. Ryoo, Nature 412
(2001) 169.
124. F. Fievet, J.P. Lagier, B. Blin, B. Beaudoin, M. Figlarz, Solid State Ionics 32-33
(1989) 198.
125. Z. Liu, L.M. Gan, L. Hong, W. Chen, J.Y. Lee, J. Power Sources 139 (2005) 73.
126. H.S. Oh, J.G. Oh, H. Kim, J. Power Sources 183 (2008) 600.
127. C. Grolleau, C. Coutanceau, F. Pierre, J.M. Leger, J. Power Sources 195 (2010) 1569.
128. B.M. Babic, Lj. M. Vracar, V. Radmilovic, N.V. Krstajic, Electrochim. Acta 51
(2006) 3820.
- 29 - 129. C. Bock, C. Paquet, M. Couillard, G.A. Botton, B.R. MacDougall, J. Am. Chem. Soc.
126 (2004) 8028.
130. V. S. Bagotsky, Y. B. Vassiliev, Electrochim. Acta 12 (1967) 1323.
- 30 - 2.1 Introduction
The main emphasis of this chapter is to give information on basic details of the characterization techniques used in our investigation. Chemicals, experiment details and instrument specifications are included in forthcoming chapters.
The following characterization techniques are used in this thesis
i) Ultraviolet visible spectroscopy
ii) Fourier transform infra red spectroscopy
iii) Insitu- Fourier transform infra red spectroscopy
iv) Thermo gravimetric analysis
v) X-ray diffraction
vi) X-ray photo electron spectroscopy
vii) Transmission electron microscopy
viii) Scanning electron microscopy
ix) Cyclic voltammetry
x) Electrochemical impedence spectroscopy
2.2 Ultraviolet visible (UV-Vis) spectroscopy
UV-Vis spectroscopy is related to the absorption or reflectance spectroscopy carried out in the UV-Vis spectral region of the electromagnetic spectrum. On the absorption of light in the UV-Vis spectral region, the molecules undergo electronic transitions in order to excite the electrons to higher energy levels.
- 31 - The intensity of absorption may be expressed as transmittance (T), defined by
T = I/I 0 (1)
Where I0 is the intensity of the incident light and I is the intensity of the transmitted light.
UV-Vis spectroscopy can be used as a quantitative analysis in determining the concentration of the species. Beer–Lamberts law states that the absorbance of the molecule is directly proportional to the concentration of absorbing species in the solution and the path length which can be expressed by following equation
A = abc (2)
Where A is the measured absorbance, a is absorptivity, b is the cell-path length, and c is the analyte concentration.
In the other way, Mie theory gives a detailed explanation for the light scattering by simple spherical particles [1]. Surface plasmon resonance (SPR) band exhibited noble metal nanoparticle (Au, Ag and Cu) is explained using Mie theory. SPR can be briefly explained as the collective oscillation of electron cloud, induced by the electric vector of an electromagnetic wave [2]. For example, gold nano particles exhibit SPR in the visible region around 514 nm. When a small spherical metallic nanoparticle is irradiated by light, the oscillating electric field causes the conduction electrons to oscillate coherently. This is schematically pictured in Fig 2.1
Fig. 2.1 Polarization of a spherical metal particle by the electrical field vector of the incoming electromagnetic
- 32 - 2.3 Fourier Transform infra red (FT-IR) spectroscopy
Fourier transform infrared spectroscopy (FT-IR) is also a type of absorption spectroscopy that deals with the infrared region of the electromagnetic spectrum. Infrared region is generally divided into three regions namely
i) Near – IR (14000 – 4000 cm -1)
ii) Mid – IR (4000 – 400 cm -1)
iii) Far – IR (400 – 40 cm -1)
In near–IR region overtone or harmonic vibrations of the molecules are observed.
Generally researchers used to study in mid –IR region where the fundamental vibrations of the molecule takes place. Far–IR region is used to study the rotational spectroscopy of the molecules. Chemical bonds have specific frequencies for their vibrations corresponding to the energy levels. The resonant frequencies or vibrational frequencies are determined by the shape of the molecular potential energy surfaces, the masses of the atoms and, eventually by the associated vibronic coupling. In order for a vibrational mode in a molecule to be IR active, it must be associated with change in the dipole moment [3]. The IR spectrum was recorded by passing a beam of infrared radiation through the sample, and the amount of energy absorbed at each wavelength is recorded. This may be done by scanning through the entire spectral region with a monochromatic beam which changes in wavelength over time, or by using a Fourier transform operation, measurement can be done at once in the entire IR region. Performing a mathematical Fourier transform on this signal, results in a spectrum will be identical to that from conventional (dispersive) infrared spectroscopy. From this, a transmittance or absorbance spectrum may be plotted, which shows at which wavelengths the sample absorbs the IR, and allows an interpretation of bonds present.
- 33 - 2.4 In-situ FT-IR spectroscopy
FT-IR is a powerful technique for studying adsorbed molecules at the solid/gas as well as the solid/liquid interfaces. This technique is widely used by electrochemists to examine the electrode-electrolyte interface at the molecular level [4]. FT-IR at electrochemical interfaces can be used to study species adsorbed at the electrode surface and species in a thin layer of solution adjacent to the electrode, yielding information on the chemical nature of the species present at the interface and adsorption sites. The orientation of adsorbed molecules and on the lateral interactions between adsorbates can also be studied using this technique [5].
Electrochemical Stark effect is observed in electrochemical FT-IR, due to the huge electric fields (ca. 10 9 V m -1) typically attained at the electrochemical double layer, is the variation of the absorption frequency of a band with the potential [6]. The Stark effect can be explained either by the potential-induced variations in the electronic density involved in the chemical bond between the surface and the adsorbed molecule, or by the interaction between the electric field at the interface and the dynamic dipole moment associated to the vibrational mode. It can be demonstrated that these two effects are one and the same thing, and cannot be separated from each other.
FT-IR spectra are recorded by potential modulation is called as Subtractively
Normalized Fourier transform interfacial infra red spectroscopy (SNIFTIRS). In SNIFTIRS, the electrode potential is modulated between the two potentials and the spectra of the reflected infrared radiation are measured. The reflection absorption spectrum is obtained by following relationship
- 34 - ∆R R R = Efinal − Einitial − (3) R R Einitial
Where R and R are the reflectance from the thin layer electrolyte (at the working Efinal Einitial electrode) at the final and initial potentials. Due to subtraction, SNIFTIRS are devoid of the common background signal due to the aqueous electrolyte. These experiments are carried out in a specially designed electrochemical cell as shown below in fig 2.2.
Fig. 2.2 Electrochemical cell used in SNIFTIR spectroscopy studies
2.5 Thermo-gravimetric analysis (TGA)
TGA is used to analyze the changes in the physical and chemical properties of a material as a function of increasing temperature. The physical properties like second-order phase transitions, including vaporization, sublimation, absorption, adsorption desorption etc. and chemical properties like dehydration, chemisorption, decomposition etc. can be analyzed from TGA. In general, TGA results are showed weight loss or gain against temperature and it is called as thermogram.
- 35 - 2.6 X-ray diffraction (XRD)
X-ray diffraction is a powerful technique for the structural characterization of crystalline solid [7]. Any solid material with a periodically ordered atomic or molecular structure diffracts electromagnetic radiation of wavelengths of about the same length as the interatomic distances (shown in Fig 2.3).
Fig.2.3 Typical picture of X-ray reflections from a crystal
Where λ is the wavelength of the monochromatic x-ray radiation, θ is the Bragg angle, d is the inter planar distance.
XRD was shown by plotting intensity versus 2 θ values and the peaks observed in the
x-ray diffraction patterns are indexed using standard pattern. The crystallite size (L V) of metal nanoparticles has been estimated using Scherer formula:
λ L = φ − − − − − − − )3( v FWHM .cos θ
Where, φ is the shape factor (0.89 for spherical crystallite), λ the radiation wavelength
(1.5406 Å), FWHM the full width at half maximum, and θ the angle at the maximum intensity
- 36 - 2.7 X-ray photo electron spectroscopy
XPS is a surface sensitive technique of chemical analysis, based on the photoelectric effect. The thin film on a flat substrate is exposed to monochromatic x-rays, causing photoemission from the valence and core levels of atoms in the outermost atomic layers of the sample. Photoemission of valence electrons provides information of the bonding in the solid, and emission of core electrons of its chemical composition [8]. In majority of the studies, XPS is used to study the oxidation state of the element. The binding energy scale of the spectra was aligned through the C (1 s) peak at 285 or 284.5 eV. Measurements were made using photoelectron take-off angles of 30° with respect to the surface normal of the sample. Peak fitting and deconvolution are done using the XPSPEAK version 4.1 program.
2.8 Transmission electron microscopy
Modern transmission electron microscopes can be used for a number of sophisticated analysis techniques, but the two basic ones used most frequently are object imaging and electron diffraction. Fig. 2.4 shows the optical principles behind the two configurations from sample to screen. When the electron beam is transmitted through the sample, the elastically scattered electrons are passed through apertures and electromagnetic lenses before projected onto the screen. When the selected area electron diffraction (SAED) mode is in operation, a selected area aperture is inserted into the beam, while in imaging mode an objective aperture is sometimes used to enhance the contrast [9].
- 37 - In electron diffraction, the scattering angle θ is very small (typically < 2 º) and the d- values can be calculated from
Lλ dhkl = − − − − − − − )4( rhkl
Where L is the camera length in mm, λ is the electron wavelength, and rhkl is the distance between the central electron beam (000) and the diffraction spot hkl in mm.
Figure 2.4 Configurations used in TEM for electron diffraction (left) and sample imaging (right)
- 38 - 2.9 Scanning electron microscopy
The principal differences between TEM and SEM are that in the latter technique the electron beam is scanned over the sample surface, instead of being passed through the sample. The accelerating voltage used is also considerably low (usually 5-20 kV). When the electron beam hits the atoms in the sample, both electrons and photons are emitted. The emitted electrons are collected and used to form a 3D picture of the sample, providing both topographical and structural information. Different kinds of electromagnetic radiation are emitted from the sample, and one of them is x-rays. Since the emitted x-ray photons have energies characteristic for each element, they can be used to determine the chemical composition of the sample both qualitatively and quantitatively by a technique called energy dispersive x-ray spectroscopy (EDS) [9].
2.10 Cyclic voltammetry
If the electrochemical properties of a molecule are studied for the first time then cyclic voltammetry is the apt technique to follow. In cyclic voltammetry, three electrodes were used namely working electrode, counter electrode and reference electrode. The potential is measured between the reference electrode and the working electrode and the current is measured between the working electrode and the counter electrode. The potential of the working electrode was scanned in a triangular mode with respect to time at a constant scan rate and resulting current is plotted versus the applied potential to give cyclic voltammogram. The typical cyclic voltammogram is shown in fig 2.5.
- 39 -
Fig.2.5 Typical cyclic voltammetric response of a reversible system
In the waveform shown in fig.2.5, the forward scan produces a current peak for any analytes that can be reduced (or oxidized) through the range of the potential scan. The current will increase as the potential reaches the reduction (or oxidation) potential of the analyte, but then falls off as the concentration of the analyte is depleted close to the electrode surface. As the applied potential is reversed, it will reach a potential that will reoxidize (or re-reduce) the product formed in the first reduction reaction, and produce a current of reverse polarity from the forward scan. This oxidation (or reduction) peak will usually have a similar shape to the reduction (or oxidation) peak. As a result, information about the redox potential and electrochemical reaction rates of the compounds is obtained. For instance, if the electron transfer at the surface is fast and the current is limited by the diffusion of species to the electrode surface, then the current peak will be proportional to the square root of the scan rate.
The peak shape of the oxidative and reverse current-potential (I-E) curve in Fig.2.5 is typical for an electrode reaction in which the rate is governed by diffusion to a planar electrode surface. That is, the rate of the electron transfer step is relatively fast compared to
- 40 - that of diffusion. In such a case the peak current I p, is governed by the Randles-Ševèík relationship [10, 11]:
5 3/2 1/2 1/2 ip = 2.69 x 10 n A D C0 υ (5)
Where, n is the mole of electrons transferred per mole of electroactive species; A is
2 2 the area of the electrode in cm . D is the diffusion coefficient in cm /s; C 0 is concentration in mole/L; and υ is the scan rate of the potential in V/s. The I p is linearly proportional to the bulk concentration, C, of the electroactive species and the square root of the scan rate, ν1/2 .
1/2 Thus, an important diagnostic is a plot of the I p vs υ . If the plot is linear then the electrode reaction is controlled by diffusion, which is the mass transport of electroactive species to the surface of the electrode across a concentration gradient. The thickness, d, of the “diffusion layer” can be approximated by: δ~ [Dt] 1/2 , where D is the diffusion coefficient and t is time in seconds. A "quiet", i.e., unstirred solution is required. The presence of supporting electrolyte is required to eliminate movement of the charged electroactive species due to migration in the electric field gradient.
For a reversible reaction, the difference between the anodic and cathodic peak values
(∆Ep) is equal to 59 mV/n. This value will be independent of the scan rate. There is a caveat to the analysis for reversibility when there is a notable solution resistance (ohmic) between the working and reference electrode. The measured potential then contains an additional component of potential equal to E (ohmic) = IR. Modern electrochemical workstations include features that can compensate for this IR drop and give the potential values free from this component.
- 41 - Cyclic voltammetry can also be used for evaluating the interfacial behavior of electroactive compound. Both the reactant and the product can be involved in an adsorption – desorption process. A gradual increase of the cathodic and anodic peak current is indicating progressive adsorptive accumulation at the surface. Note that separation between the peak potentials is smaller than expected for solution phase process. Indeed, ideal Nernstian behavior of surface-confined non-reacting species is manifested by symmetrical cyclic voltammetric peaks ( Ep= 0) and peak half–width of 90.06/n mV (fig.2.6).
Fig.2.6: Ideal cyclic voltammetric behavior for a surface layer on an electrode
The peak current is directly proportional to the potential scan rate [10, 11]
2 2 2 ip = n F Γ Aν/ 4RT (6)
Where n is the mole of electrons transferred per mole of electroactive species; A is the area of the electrode in cm 2, F is Faraday constant in C mol -1,R is molar gas constant in JK -1mol -1.
Γ is the surface coverage of the electrode reaction substance in mol cm -2, T is the absolute temperature in Kelvin, ν is the scan rate of the potential in V/s. The I p is linearly proportional
1/2 the scan rate, ν . Thus, an important diagnostic is a plot of the I p vs υ should be linear. The
- 42 - ideal behavior is approached for relatively slow scan rate, and for an adsorbed layer that shows no intermolecular interaction and fast electron transfers.
2.11 Electrochemical impedance spectroscopy
Electrochemical impedance in a general sense is a vector quantity describing the relationship between a potential wave and a current wave, when an electrochemical system is perturbed by an alternating current or potential. Impedance extends the concept of resistance to alternating current. In EIS, the response of current (or potential) is measured when a sinusoidal perturbation of potential (or current) is applied at different frequencies.
Perturbation has to be limited to small amplitude in order to ensure linear (or pseudo-linear) relationship between the current and potential wave. The frequency dependent proportionality between the potential signal and the current response results in the impedance data, which is normally presented in form of a Nyquist plot (-Zimaginary vs. Z real ) or a bode plot
(magnitude vs. frequency and phase vs. frequency). An electrochemical system can be assumed to be a network of various resistive, inductive, and capacitive components. For interpreting the spectra normally an equivalent circuit containing different electrical components is designed. If the equivalent circuit model fits the experimental spectrum then each electrical component in the equivalent circuit is ascribed to physical processes which might be present in the system. Normally either two-electrode or three-electrode cells can be used for EIS measurements. Two-electrode cells measure the total polarization of the cell
(which would be a sum of all the resistive components present); and hence normally used to study bulk or average phenomenon like electrolyte resistance, total resistance of the cell etc.
Normally a three-electrode setup is more helpful because it allows precise measurement of
- 43 - the potential of the working electrode. Potential of the working electrode is precisely measured with respect to a reference electrode. Ideally the reference electrode should be so placed that it minimizes the resistive contribution of the electrolyte. The current is applied between the working and counter electrode, and hence the reference electrode should be placed such that it does not perturb the current flow between the working and counter electrodes. With EIS, processes such as electronic/ionic conduction in electrodes and electrolytes, interfacial charging (on surface or double layer), charge transfer, and mass transfer processes can be qualitatively determined [12, 13]. EIS is widely used in the field of batteries and fuel cells to study formation of passive films on electrode surfaces, estimate state of charge, determine reaction rates, measure conductivity of electrolytes etc.
References:
1. G. Mie, Ann. Phys. 25 (1908) 377.
2. K. L. Kelly, E. Coronado, L. L. Zhao, G. C. Schatz, J. Phys. Chem. B 107 (2003) 668.
3. J. R. Dyer, Applications of absorption spectroscopy of organic compounds, Prentice-
Hall (1965).
4. K. Kunimatsu, J. Electroanal.Chem. 140 (1982) 205.
5. B. Beden, C. Lamy, in R. J. Gale (Ed.), Spectroelectrochemistry – theory and
practice, (Ed.:), Plenum Press., New York (1988)
6. D. K. Lambert, Electrochim. Acta 41 (1996) 623
7. B.D. Cullity Elements of X-ray diffraction Addison-Wesley Publishing Company,
Inc. Massachusetts (1956)
- 44 - 8. D. Briggs, M. P. Seah Practical Surface Analysis - Auger and X-ray Photoelectron
Spectroscopy (2 nd ed.) Wiley Interscience (1990)
9. Z. L. Wang, Characterization of Nanophase Materials, Weinheim, Wiley-Vch Verlag
Gnibh & Co. (2000).
10. A. J. Bard, L. R. Faulkner, Electrochemical Methods:Fundamentals and Applications,
Second Edition, John Wiley and Sons Publishers (2001).
11. J. Wang, Analytical Electrochemistry, VCH Publishers (1994).
12. M. Orazem, B. Tribollet, Electrochemical impedance spectroscopy Hoboken, N.J.
Wiley (2008).
13. E. Barsoukov, J. R. Macdonald, Eds Impedance Spectroscopy: Theory, Experiment,
and Applications, Second Edition Wiley-Interscience (2005).
- 45 - 3.1 Introduction
Nanosized noble metal colloids have generated a great deal of interest among researchers owing to their electronic and optical properties, distinctly different from their bulk counterpart [1–3]. Various general methods in the synthesis of metal nanoparticles are described in the chapter-1. Leff et al. reported the synthesis of alkylamine capped gold nanoparticles similar to the preparation of thiol-stabilized gold sols [4]. These sols were stable only in organic solvents. Aqueous dispersion of stabilized metal nanoparticles in the siloxane network [5–13] will provide a promising way to study their optical and catalytic applications. One advantage of using the siloxane stabilized sol is that the optical properties of the nanocrystalline dispersant are not altered much by the transparent silicate matrix. The porous siloxane network formed by hydrolysis and condensation of silane can facilitate the percolation of reactant and products. Sol–gel monomers containing amine functional group can exhibit dual role such as chelating with metal ion and cap the metal sol. Bharathi et al. studied aminosilicate stabilized Au, Pt, Pd and Ag nanosol [6–13] (Detailed explanation was given in the Sec 3.1.1). They established the functional role played by the amino group and siloxane network in stabilizing the mono-metal sol. It was found that the average particle size of the Au, Pd and Pt nanoparticles were in the range of 5–7 nm, while the size of the Ag nanoparticles prepared by the same method were larger in size c.a.2–30 nm (Shown in Fig.1)
Au–Ag alloy formation takes place easily than the other systems because of the lattice constants of these metals were nearly the same and the miscibility gap between Au and Ag also less [14]. Sampath et al. [13] have recently reported the synthesis of alloy dispersions stabilized by organically modified silicate matrices, wherein the electrocatalytic activity of
- 46 - the dried films of alloy sols immobilized on glassy carbon electrodes for the oxygen reduction reaction was demonstrated.
Fig.1 Particle size distribution of Au, Pt, Pd and Ag nanosols stabilized by N, N ′-[3-
(trimethoxysilyl)propyl]-ethylenediamine (Reproduced from Ref. 12).
In this work, we have prepared 3-aminopropyl trimethoxy siloxane (3–APS) stabilized Au and Ag sols. The pH of the medium was found to have major effects on the stability of the sol. The alloys of Ag and Au of various compositions, stabilized by 3–APS were prepared by co-reduction of the two metal salts. The nano sols were characterized by using UV–visible (UV–Vis) spectroscopy and transmission electron microscopy (TEM).
- 47 - 3.1.1 Importance of 3-Aminopropyltrimethoxy silane
In the stabilization of metal nanoparticles, the presence of amine and siloxane functional groups in the 3–APS were equally responsible for the formation of stable metal nano dispersions. The equal importance of amine and siloxane group in 3–APS was confirmed by choosing propyl amine and tetraethylorthosilicate in the preparation of metal nanosols. As expected, propyl amine and tetraethylorthosilicate did not act as a stabilizer which was observed by the immediate precipitation of metal particles. Hence, this result clearly demonstrated that the amino group caps the metal center and further siloxane network acts as a stabilizer. The interaction of amine and metal ion are in the form of the respective amine-metal complexes or mixed chloride, depending on the chloride and amine concentrations. It was found that Au, Pt and Pd have high stability constants except Ag. So, the stability of this complex slows down the rate of reduction of the metal ions. Upon reduction, the amine groups immediately cap the surface of the nucleating metal surfaces to form stable metal sols. It was observed that the Au, Pt, Pd gave similar sized nanoparticles except Ag. Broad size distribution and very less stable Ag nanoaparticles are formed by the similar procedure. Hence, we explored this method further for the formation of stable Ag nanoparticles and also Au-Ag alloy nanoparticles. [6-14].
3.1.2 Reduction of 4- Nitrophenol
Functionalized aminobenzenes are important intermediates for the manufacture of many agro-chemicals, pharmaceuticals, dyes and pigments. The mechanism of hydrogenation of aromatic nitro compounds was explained in detail by Corma et al. [15] and Blaser [16].
They have reported that reduction of 4- nitrophenol takes place by either direct route or by
- 48 - condensation route. The reaction mechanism was given in scheme-1. Condensation route was favored under basic conditions. Direct pathway follows nitroso and hydroxylamine as intermediates. In nitroso pathway, the reduction pathway is fast but the nitro and nitroso compounds are strongly adsorbed on the catalyst surfaces. In hydroxylamine pathway, the
O-N bond breaking step is slow and the slowness leads to the accumulation of hydroxyl amine intermediates, the formation of side products and further decomposition also proceeds in the course of reaction. Hence, catalyst requirement is indispensable to achieve the complete reduction of nitro to amino group. The reaction is further complicated if an additional group is present which can also undergo reduction or cleavage by hydrogen. This issue can be overcome by selectively reducing nitro group in the presence of other reducible group using chemo-selective catalyst [15, 16].
Scheme .1 Detailed mechanism of functionalized aromatic nitrophenol reduction
(Reproduced from the ref. 16).
In our earlier studies, metal nanoparticles embedded in poly-
(ethylenedioxythiophene) polymer matrix was used as a catalyst to study the reduction of
- 49 - 4-nitrophenol [17, 18] and others have also chosen this reaction as a model reaction to study the catalytic activity [19, 20]. Herein, we have used 3–APS stabilized Au, Ag and Au-Ag alloy sols to study the catalytic activity in 4-nitrophenol reduction.
3.2 Experimental procedures
3.2.1 Materials and methods
All chemicals used were of analytical grade. Solutions were prepared using Milli-Q water (resistivity 18.2 MΩ.cm). All glassware were thoroughly cleaned with freshly prepared aqua-regia and rinsed with Milli-Q water before use. UV–Vis spectra were collected using
Varian Cary 500 spectrophotometer. The TEM images were collected using Philips CM-200 microscope working at 200 kV. The metal sols were diluted and sonicated for uniform dispersion. The samples for TEM were prepared by placing a drop of the sample onto a copper grid coated with carbon film (400 meshes). The grid was allowed to get dried in air overnight at room temperature.
3.2.2 Synthesis of gold/silver nanoparticles
The solutions of 0.1 M 3–APS were prepared in water and initial pH was found to be
10.6. The pH of the solution was adjusted to 8.0 by adding dilute HNO 3. This solution was sonicated for 10 min. Hydrolysis and condensation of 3–APS was initiated on adding HNO 3.
In the case of the preparation of Ag sol, use of HCl and H 2SO 4 were avoided to prevent precipitation of corresponding silver salt. To the above solution, either 0.5 mM HAuCl 4 or
+ AgNO 3 (ratio of 3–APS: M = 200:1 in molar ratio) was added and sonicated for 30 min in order to homogeneously distribute the metal ions in the solution. The chemical reduction was
- 50 - effected by adding a few drops of freshly prepared 50 mM NaBH 4 to the 3–APS stabilized metal ion solution. Gold/silver nanoparticles formed were stabilized by 3–APS. The final pH of the solution was 8.2. The reaction was carried out at room temperature of 25 ° C. A similar procedure was also employed for the preparation of metal sols at pH 5.0 for comparison.
3.2.3 Synthesis of gold–silver alloy
Instead of adding gold or silver salts individually, we have added both salts to the solution containing 3–APS at pH 8.0 and co-reduced by adding a few drops of freshly prepared 50 mM NaBH 4. The molar ratio of gold and silver was varied by taking various ratios of Au and Ag salts, maintaining the total precursor salt concentration at 0.5 mM. The final pH of the alloy sol after reduction was found to be 8.2.
3.2.4 Chemical reduction of 4-Nitrophenol
The reduction of 4-nitrophenol was chosen as a model reaction to show the catalytic behavior of synthesized Au–Ag nanoparticles. The reduction reaction was carried out in a standard quartz cuvette of 1 cm path length and of 3 mL volume. The procedure entailed in mixing excess NaBH 4 (10 mM) with a 4-NP (0.05 mM) solution in water in the quartz cell followed by the addition of aliquots of catalyst particle dispersions (50 µL). The absorption spectra were recorded at time intervals of 50 seconds by scanning the wavelength range from
250 nm to 700 nm at 25 °C.
- 51 - 3.3 Results and discussion
3.3.1 Characterization of gold and silver sols
In the preparation of the Au sol (as described in Section. 3.2.2), 3–APS undergoes slow hydrolysis and condensation catalyzed by acid to form a siloxane network. After the chemical reduction of HAuCl 4 in 3–APS by sodium borohydride, the colour of the solution turned to characteristic wine red. The UV-Vis spectra of the gold sol will show characteristic surface plasmon resonance (SPR) around 520 nm [21]. The 3–APS and N, N′-[3-
(trimethoxysilyl)propyl]-ethylenediamine (EDAS) stabilized Au sols were reported to form very stable dispersions of Au at pH 5.0 [8, 11] which clearly established the role played by the aminosilanes in binding and dispersing gold nanoparticles in aqueous media. The formation of stable Au nano sols stabilized by 3–APS at pH 5.0 was first reported by
− Bharathi et al. [11]. Similar binding of AuCl4 to amino siloxane network was also reported by Henao et al. [22].
In exploration of the role of pH, we have increase the pH of the medium from 5.0 to
8.2 and then the similar methodology is followed in the preparation of Au nanoparticles. The
Au nanoparticles prepared at pH ~ 8.2 also showed the SPR band at 533 nm (fig. 2) which is at the higher wavelength compare to the Au nanoparticles prepared at pH~5.0 [11]. The characteristic wine red color of Au nanoparticles was also observed (inset in fig.2). However,
Au sols prepared at pH ~ 8.2 were found to be less stable compared to pH ~ 5.0 and found to undergo aggregation after two days. TEM image shown in fig. 3A reveals the aggregated structure of Au nanoparticles prepared at pH ~ 8.2. The Au particles are having very broad range of size distribution c.a 5−11 nm (shown in fig. 3B). This observation led us to consider
- 52 - − that the electrostatic attraction between AuCl4 and protonated amine in the siloxane network plays a vital role in stabilization of Au sol.
Fig. 2 UV-Vis spectra of Au and Ag nanoparticles stabilized by 3–APS at pH ~ 8.2.
Similarly, the Ag sol was also prepared by the same procedure as described in Section
3.2.2. SPR of Ag nano sols was observed at 405 nm (Fig. 2) which was in agreement with other reports [21] . The characteristic yellow color of the Ag sol was also observed (inset in
Fig.2). The 3–APS stabilized Ag sols at pH ~ 8.2 were found to be very stable for at least a month in contrast to the observed aggregation of Au sols at pH ~ 8.2. However, the Ag sols were found to be unstable at pH ~ 5.0 and agglomeration of bigger black silver particles takes place instantaneously. A non-uniform particle size distribution of Ag particles was observed in the TEM (Fig. 3A). The TEM images show that Ag particles are well separated from each other. The histogram shows the narrow sized distribution with slight distortion.
Narrow size range of 6–8 nm was found from the histogram shown in Fig. 3B.
- 53 - A) Au Ag
B) Au Ag Counts Counts
5 6 7 8 9 10 11 4 5 6 7 8 9 10 11 12 13 Particle size, nm Particle size, nm Fig. 3 A) TEM images of Au and Ag nanoparticles stabilized by 3−APS at pH ~ 8.2, B) Histogram of Au and Ag nanoparticles.
3.3.2 Role of pH on the stability of sols
+ − It is well known that HAuCl 4 will be dissociated to form H and AuCl4 ions. The amino group in 3–APS will be in the protonated form at pH ~ 5.0. Hence there will be strong electrostatic attraction between anionic tetrachloroaurate and protonated amine in 3–APS
(shown in scheme-2). However, when the pH was increased from 5.0 to 8.2, the amine group will exist in non-protonated form. Hence, there will be weak electrostatic attraction between
- 54 - − tetrachloroaurate anion ( AuCl4 ) and free amine group in 3–APS. Therefore, Au nanoparticles are found to be stable at pH~5 compare to pH~8.2.
− Scheme. 2 Schematic illustration of electrostatic interaction between AuCl4 & 3–APS at pH
5.0 and Ag + & 3–APS at 8.2
+ − In contrast, AgNO 3 dissociates to form Ag and NO 3 ions and hence the stronger electrostatic interaction will exist between Ag + and free amine group in 3–APS at pH~8.2
(shown in scheme-2) compare to protonated amine in 3–APS (pH~5.0). The presence of a free electron pair on nitrogen atom in 3–APS can act as a complexing ligand with pseudo- coordinate interaction with Ag + ion. Hence, the stability or instability of Ag nanoparticle sol depends on the charge of amino group in 3–APS at a given pH. Further, Bharathi et al. [12] have reported the synthesis of Au, Pt, Pd and Ag nanoparticles using 3–APS at pH ~ 5.0.
They observed that the particle size of Au, Pt and Pd were smaller than the size of Ag. They concluded that a weak interaction between the amino group of 3–APS and the Ag is the origin of the formation of larger particle size of Ag at pH 5.0.
In this study, we found that the electrostatic interactions plays vital role in the stabilization of the metal sols. The electrostatic interactions were dictated by the charge
- 55 - present on the amine group in 3–APS (pH dependent) and charge associated with precursor metal ions. The schematic illustration of metal ion entrapped in the 3–APS at pH 5.0 for Au and pH 8.2 for Ag was given in scheme 2. Hence, we found that Au nanoparticles are stable at pH~5.0 but not at pH~8.2 and Ag nanoparticles are stable at pH~8.2 but not at pH~5.0
3.3.3 Formation of Au–Ag alloy sols
We found that physical mixing of Au and Ag nanosols prepared at pH ~ 8.2 did not lead to the formation of alloy sols, as observed from the individual SPR band in UV–Vis spectra (shown in fig. 4). Similar report was observed by Devarajan et al [13] in the physical mixing of Au and Ag nanosols stabilized by N, N ′-[3-(trimethoxysilyl)propyl] diethylenetriamine.
Fig. 4 UV-Vis spectra of physically mixed gold and silver nanoparticles stabilized by
3–APS at pH ~ 8.2, ratio of gold and silver in the nano sols are indicated in graph.
- 56 - Au–Ag alloy sols of various molar ratios were prepared by co-reduction of a mixture of solutions of Au and Ag salts in 3–APS at pH ~ 8.2. Surprisingly, there is only a single
SPR band (Fig. 5A) different from the results observed in physical mixture of Au and Ag sols (Fig. 4). Instead of two SPR band from Au and Ag, single SPR band confirmed the formation of Au-Ag alloy. Further alloy formation was confirmed from the blue shift in the
UV–Vis spectra (Fig. 5A) as a function of increasing composition of Ag in the Au–Ag alloy
[23]. The colour of the Au, Au–Ag and Ag colloids gradually changes from wine red to yellow as shown in the photograph [inset in Fig. 5A]. It was reported that a very close miscibility gap and the similar lattice parameters of Au and Ag are favorable for the formation of alloy [13, 14]. The absorption maximum depends on the mole fraction of the Au in the alloy sol as seen from a continuous shift of the SPR band compared to mono-metal sols which varies linearly with the composition of mono metals (Fig. 5B) [24].
Fig.5 A) UV-Vis spectra of Au-Ag alloy (compositions are mentioned in graph) prepared by co-reduction of HAuCl 4 and AgNO 3 in 3–APS at pH ~ 8.2; Inset:
Photograph of Au, Au–Ag (different compositions) and Ag nano sols from left to right
(a–g), B) Plot of wavelength corresponding to the maximum absorbance for various mole fractions of Au in Au–Ag alloy nano sol.
- 57 - The TEM image of Au–Ag (1:1) showed nearly uniform particles (Fig. 6A) and particles were well separated from each other. The particle size was calculated from the histogram shown in Fig. 6B and it is found to be in the range of 3–5 nm. Au–Ag alloy sol shows smaller particle size when compare to the size of the respective monometallic colloids.
A B
Counts
2 3 4 5 6 7 8 910 Particle size, nm
Fig.6 A) TEM image of Au-Ag (1:1) alloy stabilised by 3–APS at pH 8.2, B) Histogram of Au-Ag (1:1) nanoparticles.
The stability of the Au sol at pH ~ 8.2 was found to increase dramatically on alloying even with 4% of Ag. The increased stability of Au sols alloyed with 4% Ag characterized by the invariance of the spectra observed on standing for a month. However, the alloying of Au at lower concentration levels did not increase the stability of Ag sols at pH~5.0. It may be due to the fact that the Au is nobler metal than Ag. Hence, Au will have a tendency to form the nuclei first, over which the silver forms as a shell. High-resolution TEM studies may elucidate the core–shell nature of the particles.
- 58 -
3.3.4. Reduction of 4-nitrophenol
Fig.7. UV-Visible spectra of A) 4-Nitrophenol and B) 4-Nitrophenol in excess NaBH 4
It was well known that the reduction of 4-nitrophenol to 4- aminophenol requires catalyst along with a reducing agent [Details given in Section 3.1.2]. In fig. 7, UV-visible spectra showed the absorption of 4-Nitrophenol takes place at 313 nm. On addition of NaBH 4 to the 4-nitrophenol, the absorption undergoes red shift and it is observed at 400 nm (colour changes from the light yellow to dark yellow) [17, 18]. This red shift is due to the formation of nitrophenolate anion. NaBH 4 is strong reducing agent but it cannot reduce the nitro group in 4-nitrophenol which is evidenced from the absorption peak at 400 nm which remains unaltered for several days. We have also observed that the reduction doesn’t take place even at higher concentrations of NaBH 4. This proves that the catalyst is indispensable for the reduction of 4-nitrophenol. This feature was observed from the earlier work published from
- 59 - our laboratory in which dispersions of either Au or Pd nanoparticles stabilized in the polymer matrix were used as the catalyst for the conversion of 4–nitrophenol to 4–amino phenol. The rate constants were calculated and found Pd was more catalytic than Au [17, 18]. The catalytic activity of metal nanoparticles embedded in various stabilizers was also explored for the reduction of 4–nitrophenol by others [19, 20, 25, 26]. Herein, Au, Ag and Au–Ag alloy sols stabilized by siloxane network in 3–APS was used as a catalyst in the reduction of 4– nitrophenol. Interestingly, the reduction of 4–nitrophenol to 4–aminophenol was observed on addition of small aliquots of Au, Ag, Au–Ag (0.08–0.20 µg/ µL) colloids to the solution containing 4-nitrophenol in the presence of excess NaBH 4. This reduction reaction can be observed by continuous fading of yellow color. (Schematic diagram was shown in Fig.8). It has been established that the addition of aliquot amounts of Au or Ag or Au–Ag colloidal sols will not interfere in the UV–vis spectra. The rate of the reaction is studied by recording
UV-Vis spectra of the reduction of 4–nitrophenol at regular time intervals. UV–Vis spectra of reduction of 4–nitrophenol using Au, Ag and different composition of Au-Ag alloy sols are shown in fig. 8. The reduction reaction can be monitored by decrease in the intensity of absorption of 4–nitrophenol at 400 nm and the emergence of a new absorption peak at 310 nm corresponds to the formation of 4–aminophenol [17, 18]. More importantly, large excess of NaBH 4 takes in to account the slow but noticeable hydrolysis of this reagent at alkaline pH. In the course of reaction, the concentration of NaBH 4 is in excess and remains constant.
The rate of the reaction is depends only on the concentration of 4–nitrophenol. Hence, order of the reaction is considered as pseudo-first order [17-20, 25, 26]. This observation is confirmed by linear plot between logarithmic value of absorbance of 4–nitrophenol and time
- 60 - (inset fig. 8). Irrespective of alloy composition, the similar reaction pathway and order of reaction is observed. The rate constant is calculated by following relationship in equation 1
Ct k = ------(1) dC( dt/ )
k – Rate constant (s -1)
-1 Ct – Concentration at time ‘t’ seconds (mol.lit )
dc/dt – change in concentration with respect to time (mol.lit -1s-1)
Concentration at different time intervals was calculated from Beer-Lamberts equation given in equation 2
A C = ------(2) ε.l
C – Concentration (mol.lit -1)
A – Absorbance
ε – Molar absortivity (lit. mol -1. cm -1)
l – Path length of the cuvette (cm)
The calculated apparent rate constant for 4-nitrophenol reduction using Au, Ag and Au-Ag alloy sols of different ratios is tabulated in Table 1.
- 61 -
Fig. 8 A) Schematic illustration of conversion of 4-nitrophenol to 4-aminophenol, B-F)
UV-Vis spectra of 4-nitrophenol reduction in the presence of excess NaBH 4 using Au,
Ag and Au-Ag alloy as a catalyst (Inset: linear plot of log A vs. time). Composition of
AuAg alloy is given in the respective graphs.
- 62 - Table 1: Comparison of apparent rate constant values for 4-Nitrophenol reduction using Au, Ag and Au-Ag nanoparticles as catalysts in presence of excess NaBH 4
Rate constant Catalyst ( k × 10 -3 ) s -1 Au 6.133 Au-Ag(96:4) 14.342 Au-Ag(50:50) 27.002 Au-Ag(4:96) 12.168 Ag 19.949
From Table 1, Au sol showed lower catalytic activity compared to the Ag sol. The
Au–Ag (1:1) sol showed higher rate constant, compared to the respective mono metals (Ag and Au sols) and rate constants were found to be comparable to those reported in literature
[18]. The lower rate constant shown by Au may be attributed to the instability of Au nanoparticles at pH ~ 8.2. The average particle size of the Au–Ag sol was less (4.2 nm) compared to Ag (7.2 nm) and Au (7.8 nm). Here again, it may be emphasized that the pH conditions under which the mono-metal or alloy colloids are stable govern the rate of the reduction of nitrophenol. In the catalytic application of these mono-metal and alloy sols, the experimental rate constants for the reduction of 4–nitrophenol to 4–aminophenol revealed that the alloying up to 50% is beneficial for the catalytic properties of Au–Ag alloy.
3.4 Conclusions
This work reports the role of pH in the stabilization of Au, Ag and Au-Ag sols using
3–APS as a stabilizer. We showed that Au nanosols were stable at pH~5.0 but not at pH~ 8.2 and in contrast Ag nanosols were stable at pH~8.2 but not at pH~5.0. We have explained the
- 63 - − plausible mechanism by electrostatic interactions of amino group in 3–APS with AuCl4 and
Ag +. The stability of Au nanoparticles at pH~8.2 was improved by alloying with Ag nanoparticles but the stability of Ag nanoparticles cannot be improved by alloying with Au nano. This may be due to the reduction of Au is faster compare to Ag and hence surface of the nanoparticle may be rich with Ag which was not stable at pH~5.0. Reduction of
4–nitrophenol was studied with Au, Ag and Au–Ag alloy sols and its catalytic activity was compared. The calculated apparent rate constants showed that Au–Ag (1:1) alloy was more catalytic compare to mono metallic Au and Ag nanosols. This method of synthesis of alloy nano sol stabilized by functional siloxane network will be suitable for designing several heterogeneous catalytic systems.
References
1. C.N.R. Rao, G.U. Kulkarni, P.J. Thomas, P.P. Edwards, Chem. Soc. Rev. 29 (2000) 27.
2. J. Brinker, G. Scherer, Sol–Gel Science, Academic Press, San Diego (1989).
3. O. Lev, Z. Wu, S. Bharathi, V. Glezer, A. Modestov, J. Gun, L. Rabinovich, S. Sampath,
Chem. Mater. 9 (1997) 2354.
4. D.V. Leff, L. Brandt, J.R. Heath, Langmuir 12 (1996) 4723.
5. C.A. Morris, M.L. Anderson, R.M. Stroud, C.I. Merzbacher, D.R. Rolison, Science 284
(1999) 622.
6. S. Bharathi, M. Nogami, O. Lev, Langmuir 17 (2001) 2602.
7. S. Bharathi, M. Nogami, S. Ikeda, Langmuir 17 (2001) 1.
8. S. Bharathi, J. Joseph, O. Lev, Electrochem. Solid State Lett. 2 (1999) 284.
9. T. Hayakawa, Y. Usui, S. Bharathi, M. Nogami, Adv. Mater. 16 (2004) 1408.
10. S. Bharathi, O. Lev, Anal. Commun. 35 (1998) 29.
- 64 - 11. S. Bharathi, O. Lev, Chem. Commun. 23 (1997) 2303.
12. S. Bharathi, N. Fishelson, O. Lev, Langmuir 15 (1999) 1929.
13. S. Devarajan, P. Bera, S. Sampath, J. Colloid Interface Sci. 290 (2005) 11.
14. K. Kim, K.L. Kim, S.J. Lee, Chem. Phys. Lett. 403 (2005) 77.
15. A. Corma, P. Serna, Science 313 (2006) 332.
16. H.U. Blaser, Science 313 (2006) 312.
17. S. Senthilkumar, C. Sivakumar, J. Mathiyarasu, K.L.N. Phani, Langmuir 23 (2007) 340.
18. S. Harish, J. Mathiyarasu, K.L.N. Phani, V. Yegnaraman, Catal. Lett. 128 (2009) 197.
19. Y. Lu, Y. Mei, M. Drechsler, M. Ballauff, Angew. Chem. Int. Ed. 45 (2006) 813.
20. N. Pradhan, A. Pal, T. Pal, Colloid Surf. A 196 (2002) 247.
21. P.K. Jain, X. Huang, I.H. El-Sayed, M.A. El-Sayed, Acc. Chem. Res. 41 (2008) 1578.
22. J.D. Henao, Y.W. Suh, J.K. Lee, M.C. Kung, H.H. Kung, J. Am. Chem. Soc. 130 (2008)
16142.
23. S. Link, Z.L. Wang, M.A. El-Sayed, J. Phys. Chem. B 103 (1999) 3529.
24. M.P. Mallin, C.J. Murphy, Nano Lett. 2 (2002) 1235.
25. S. Saha, A. Pal, S. Kundu, S. Basu, T. Pal, Langmuir 26 (2010) 2885.
26. S. Panigrahi, S. Basu, S. Praharaj, S. Pande, S. Jana, A. Pal, S.K. Ghosh, T. Pal, J. Phys.
Chem. C 111 (2007) 4596.
- 65 - 4.1 Introduction
The application of fuel cell is currently limited by inadequacy in materials performance. Sir William Grove (Inventor of fuel cells) had already stated in 1839 [1] that
‘‘the chief difficulty was to obtain anything like a notably surface for action”. For this reason, nanosized particles of platinum and platinum–ruthenium supported on carbon are still the most used electrocatalysts in Proton Exchange Membrane Fuel Cell (PEMFC) and Direct
Methanol Fuel Cell (DMFC) electrodes, respectively.
For obtaining such nanostructured catalysts, numerous synthesis methods were developed, including electrochemical deposition [2, 3], physical vapour deposition [4], colloidal routes [5], impregnation reduction route [6], etc. Amongst the chemical methods, colloidal routes based on the use of organic surfactant as stabilizing agent are not industrially scalable methods. However, the polyol method has shown very promising potential for the preparation of Pt [7] and bimetallic Pt-based nanoparticles [8]. This method, well described by Fievet et al. [9], allows obtaining metal nanoparticles by reduction of metallic salts in ethylene glycol and can be performed without addition of any surfactant. So, the polyol method uses inexpensive solvent (ethylene glycol), does not need the presence of surfactant to achieve well dispersed catalytic particles of small mean sizes and is very easy to implement.
Bock et al. [10] have studied the synthesis of PtRu nanocatalysts via the polyol method. The reaction was performed by refluxing at 160 °C for several hours. They showed that the reaction mechanism involved the oxidation of ethylene glycol to aldehyde and then to glycolic acid or, depending on the pH, to glycolate, while the Pt 4+ and Ru 3+ precursor salts were reduced. In the case of Pt/C catalyst synthesis, Liu et al. [11] explained that in the
- 66 - synthesis process the polyol solution containing the metal salt was refluxed at 120 °C to 170
°C in order to decompose ethylene glycol and to generate in situ reducing species for the reduction of the metal ions to their elemental states. In traditional synthesis method, the reduction reaction is then activated by temperature: the synthesis is carried out by heating the reaction mixture at temperature higher than 120 °C for several hours [7, 9, 12, 13].
Recently, Lebègue et al. [14] developed a method based on the activation reaction by microwave impulsion, which allowed energy and preparation time savings, to synthesize well-dispersed Pt/C catalyst of ca. 2.5 nm mean size, low size distribution, leading to high electrochemical surface area. These authors showed that such method leads to highly active
Pt/C catalyst towards the oxygen reduction reaction and highlighted the effect of microwave activation at low temperature (80 °C and 100 °C) on the structure of Pt/C catalysts, particularly on active surface area and metal loading on carbon.
The microwave dielectric heating leads to thermal and non-thermal effects [15].
Thermal effects arise from different temperature regimes under microwave heating, whereas non-thermal effects result from effects inherent to the microwaves [16]. Tsuji et al. showed that these effects lead to different morphologies and sizes of metallic nanostructures in comparison with those obtained by a conventional oil-bath heating. They also underlined that the detailed mechanism for the preparation of metallic nanostructures under microwave irradiation has not been yet clarified. But, according to the generally accepted metal salt reduction mechanism via the formation of aldehyde intermediate during the polyol synthesis process, and considering that ethylene glycol possesses high dielectric losses and high reduction ability [16], it can be proposed that the role of microwave could be to favour the dehydrogenation of the molecule and hence the reduction of metal ions takes place. Anyway,
- 67 - the main advantages of microwave irradiation were discussed by Tsuji et al [16]. the uniform heating of the solution leading to a more homogeneous nucleation and shorter crystallization time, a very short thermal induction period, the generation of localized high temperatures at the reaction sites resulting in enhancement of reduction rates of metallic ions and superheating of solvents over the boiling points as a consequence of the microwave dissipation over the whole liquid volume. Microwave dielectric loss heating appears then as a better synthesis option in view of its energy efficiency, preparation time saving, uniformity of heating across the whole solvent volume (no thermal convection), and implementation simplicity.
For these reasons, polyol method activated by microwave irradiation was implemented in the present work in order to synthesis Pt/C, Ru/C and PtRu/C (different compositions) catalysts. In the synthesis, parameters like temperature, microwave power and composition of Pt and Ru were taken into account for optimization. The nanosized materials were characterized by TEM and XRD and their electro-catalytic behaviour towards electro- oxidation of carbon monoxide and methanol was evaluated using cyclic voltammetry and in- situ infrared study.
4.2 Experimental procedures
4.2.1 Synthesis of the Pt/C, Ru/C and PtRu/C catalysts
Appropriate amounts of H 2PtCl 6. 6H 2O and/or RuCl 3, (99.9% purity, Alfa Aesar) were dissolved in 100 ml of ethylene glycol (purity. p.a., >99.5% Fluka) in order to reach a concentration of metals of 0.375 g lit –1 (Table 1). Then, pH of the solution was adjusted to
11.0 for the pure Pt sample and 10.0 for Ru containing catalysts by adding a solution of
- 68 - NaOH (1 M) in ethylene glycol drop wise. Carbon Vulcan XC72R (150 mg) thermally treated for 4 h at 400 °C under Nitrogen (U Quality from “Air Liquide”) was then added to the solution in order to obtain a nominal metal loading of 20 wt% on carbon and the mixture was ultrasonically homogenized for 5 min. The reactor equipped with a cooler was put inside a MARS oven from CEM Corporation. Such set up activates the synthesis reaction by microwave irradiation at atmospheric pressure, without evaporation of ethylene glycol and/or water due to temperature increase during microwave irradiation. The synthesis of catalysts was performed under continuous microwave irradiation at a 100% power of 1600 W (unless otherwise mentioned) until reaching the desired reaction temperature, and then microwave pulses were applied to maintain it for 5 min. Fig. 1 shows the scheme representing the microwave sequence and corresponding solvent temperature profile during the synthesis process by pulsed microwaves. The catalytic powders were washed with acetone and ultra pure water (MilliQ, Millipore, 18.2 M Ω cm), and filtrated. Finally, thermal treatment of Pt/C,
Ru/C and PtRu/C catalysts was performed at 160 °C [10] for 1 h under air to remove traces of ethylene glycol.
4.2.2. Physicochemical characterization of the Pt/C, Ru/C and PtRu/C catalysts
Thermogravimetric analyses (TGA) were carried out with a TA Instrument SDT
Q600 apparatus. A few milligrams of catalytic powder was put in an alumina crucible and heated under air from 25 °C to 900 °C with a temperature slope of 5 °C min –1. Transmission electron microscopy (TEM) measurements were carried out with a JEOL JEM 2010 (HR) with a resolution of 0.35 nm. The determination of the nanoparticle size distribution was performed with the ImageJ free software [17] and estimated from the measurement of 200–
300 isolated nanoparticles to have acceptable statistical samples. X-ray diffraction (XRD)
- 69 - patterns were recorded on a Bruker D 5005 Bragg-Brentano ( θ–θ) diffractometer operated with a copper tube powered at 40 kV and 40 mA (CuK α1= 1.5406 Å). Measurements were performed from 2 θ of 15 ° to 90 ° in step mode, with steps of 0.06 ° and a fixed acquisition time of 10 s/ step. UV–visible measurements were carried out using a spectrophotometer
Evolution 100 UV–visible from Thermo Electron Corp.
Table-1: Metal salt weights dissolved in 100 mL of ethylene glycol and corresponding concentration for the synthesis of catalysts loaded with 20 wt. % of metal. The carbon powder mass added is 150 mg.
Metal salt weight Metal Concentration Metal –1 Samples in mg in mg lit Metal + Carbon H PtCl . 6H O RuCl Pt Ru 2 6 2 3 ratio in weight% Pt/C 100 – 37.5 – 20
Ru/C – 77 – 37.5 20
Pt 0.5 Ru 0.5 /C 65.75 26.33 24.8 12.8 20
Pt 0.8 Ru 0.2 /C 88.37 8.85 33.3 4.3 20
Pt 0.34 Ru 0.66 /C 50.0 38.6 18.8 18.8 20
4.2.3. Electrochemical studies of the Pt/C, Ru/C and PtRu/C catalysts
Electrochemical measurements were carried out in a standard three electrode electrochemical cell at room temperature with a reversible hydrogen electrode (RHE) as the reference electrode and a glassy carbon plate as the counter electrode. The support electrolyte was a 0.5 M H 2SO 4 (Suprapur, Merck) solution in ultra pure water. Methanol electro– oxidation experiments were carried out in N 2-saturated supporting electrolyte containing 0.1
- 70 - M methanol (Absolute Puriss. ≥99.8%, Sigma–Aldrich). The working electrode was prepared by deposition of a catalytic ink on a 0.071 cm 2 glassy carbon disk according to a method proposed by Gloagen et al. [18]. Catalytic ink has the composition of 5 mg of catalytic powder dissolved into 0.5 mL of ultra pure water and 0.1 mL of Nafion solution (5 wt% in water and aliphatic alcohol solution from Aldrich). After 30 min homogenization in an ultrasonic bath, a volume of 3 µL of catalytic ink is deposited from a syringe onto a fresh polished glassy carbon substrate, yielding 25 µg of catalytic powder, i.e. 5 µg of metal on the electrode. The solvent is then evaporated in a stream of ultra pure nitrogen at room temperature.
Cyclic voltammograms and CO stripping measurements are carried out using a Model
362 Scanning Potentiostat from Princeton Applied Research.
Infrared spectra were obtained by using the SNIFTIR (Substractively Normalized
Interfacial Fourier Transform Infra Red) spectroscopy method. The working electrode potential was modulated between two potential values (Einitial and Efinal ) according to a square wave signal. The reflectivity was obtained at two electrode potentials (frequency of potential modulation: 0.025 Hz) and resulted from the co-addition of 128 interferograms 30 times at each potential. Final spectra were calculated from the equation
∆∆∆R R R === Ef −−− Ei −−− −−− −−− −−− −−− −−− −−− −−− −−− −−− −−− −−− )1( R R Ei
Where Ei is the initial and Ef is the final potential of the modulation and E = Ef – Ei = 0.2 V is kept constant. R is the reflectance. In this case, a negative peak means the production of species and a positive band indicates the consumption of species at the electrode surface.
- 71 - 4.3 Results and discussion
4.3.1 Studies on Pt and Ru colloid formation
The synthesis of catalysts was performed under continuous microwave irradiation at a power of 1600 W until reaching the desired reaction temperature, and then microwave pulses were applied to maintain it. Fig. 1 gives a scheme of the microwave sequence and of the temperature change as a function of time in course of the synthesis procedure. temperature microwave power
0 5 time / min
Fig. 1 Scheme of the microwave sequence and of the temperature change as a function of time in course of the synthesis procedure
The minimal temperature at which the reduction of metal can occur, as well as the synthesis time for the complete reduction of metal, is not known a priori. In order to determine the reduction temperature of Pt and Ru salts, UV–visible spectroscopy was used, since it has previously been shown that UV–visible spectroscopy is an efficient tool to follow the Pt and PtRu colloidal formation process [19, 20]. For this purpose, the spectra recorded after heating of the H2PtCl 6 or RuCl 3/ethylene glycol/NaOH reaction mixture by microwave irradiation at different temperature for 5 min were compared to those recorded before
- 72 - microwave irradiation. It was found that the reduction of Pt4+ ions by ethylene glycol occurred when the temperature reaches 100 °C and this observation is in agreement with previous works [14, 19]. Indeed, in Fig. 2A the shape of the UV–visible spectrum before microwave irradiation is different from that recorded after microwave irradiation at 100 °C for 5 min over the available wavelength range. The shape of the latter spectrum, displaying strong absorption increasing gradually from ca. 700 nm towards lower wavelengths, is typical of that of a colloid [21], whereas the absorption feature at low wavelengths in the spectrum recorded before the Pt salt reduction reaction corresponds likely to the bottom of
2− the absorption peak related to ligand-to-metal charge transfer in PtCl6 ion (peak centered at ca. 260 nm [22]).
Fig .2 UV-visible spectra recorded before and after 5 min microwave irradiation at
100 °°°C for Pt (A) and at 130 °°°C for Ru (B)
In the case of the ruthenium ion reduction by ethylene glycol under microwave irradiation (Fig. 2B), changes in the UV–visible spectra was related to the reduction of Ru 3+ ions by ethylene glycol into Ru 0, started at 130 °°°C. In the literature, it was proposed that the
- 73 - formation of Ru nanoparticles gives rise to a new peak at 280 nm whose position and shape depends on duration of microwave heating [23]. The higher activation temperature was needed for the reduction of Ru 3+ ions than that for the reduction of Pt 4+ ions can be explained on the basis of the redox potential of Ru 3+/Ru (E0 = 0.84 V) is much lower than that of Pt 4+/Pt
(E0 = 1.41 V).
4.3.2 Effect of temperature
From the UV-Vis studies, we have showed that the reduction of Pt and Ru takes place at lower temperatures using microwave irradiation. Hence the synthesis of carbon supported
Pt and Ru (Pt/C and Ru/C) was found at 100 °C and 130 °C respectively. Alloying of Pt and
Ru may require higher temperature and hence we have synthesized PtRu/C alloy at three different temperatures (130 °C, 160 °C, and 180 °C).
4.3.2.1 TGA and XRD studies of Pt/C and Pt 0.5 Ru 0.5 /C
The TGA measurements indicated that the whole metallic salts were reduced on carbon matrix in 5 minutes of microwave irradiation and TGA results were given in Table.2
Table.2: Metal loading and crystallite size calculation from TGA and XRD
Metal loading (TGA) Crystallite size (XRD) Samples /weight% /nm Pt/C–100 °C 20 3.5 Ru/C–130 °C 18 –
Pt 0.5 Ru 0.5 /C–130 °C 19 3.3
Pt 0.5 Ru 0.5/C–160 °C 20 2.8
Pt 0.5 Ru 0.5 /C–180 °C 20 3.6
XRD patterns of Pt/C, Ru/C and PtRu/C (130 °C, 160 °C and 180 °C) are represented in Fig. 3. The diffraction patterns were analyzed by the method of Levenberg– Marquardt,
- 74 - using Voigt fit by means of a computer refinement program (Fityk free software [24]). All diffraction patterns recorded on Pt-containing catalysts display the typical diffraction peaks of the fcc structure of platinum, whereas, the XRD pattern of the Ru has not displayed well defined broad peaks but showed typical pattern of amorphous material. Small diffraction peaks located at ca. 35 ° and 55 ° corresponds to crystalline RuO 2 (101) and (211) orientations, respectively [25], whereas the diffraction peak at ca. 43 ° corresponds to metallic
Ru hcp (101) orientation.
Fig. 3 XRD diffractograms in 2 θθθ range from 30 °°° to 90 °°° obtained with Pt/C, Ru/C and
Pt 0.5 Ru 0.5 /C (three different temperatures) catalysts synthesized by microwave assisted polyol method.
In addition to the typical diffraction peaks of the fcc structure of platinum, the XRD pattern of the Pt 0.5 Ru 0.5 /C catalyst also display the RuO 2 peaks at ca. 35 ° and ca. 55 ° as well
- 75 - as that of hcp structure of Ru at ca 43 °. According to the Vegard’s law for a true PtRu alloy, the value of the cell parameter should decrease when the ruthenium content increases. In other words, the diffraction peaks are expected to shift towards higher 2 θ value when ruthenium is alloyed with platinum [26]. The diffraction peaks of the Pt 0.5 Ru 0.5 /C catalyst are slightly shifted towards higher 2 θ values in comparison to those recorded for the Pt/C material. Hence, it is possible to determine the alloy composition by XRD [27]. The amount of ruthenium present in PtRu/C synthesized at three different temperatures was found to be
10 % in 130 °C, 21 % in 160 °C and 28 % in 180 °C. Therefore, it can be proposed that the fcc structure of the Pt 0.5 Ru 0.5 /C catalyst synthesized at different temperatures as Pt 0.9 Ru 0.1 for
130 °C, Pt 0.8 Ru 0.2 for 160 °C, Pt 0.72 Ru 0.28 for 180 °C (these compositions were calculated from XRD reflections). In addition to this, it also contains rutile RuO 2 nanoclusters and may be also of hcp Ru nanoclusters. This result clearly suggests that the inclusion of ruthenium in platinum fcc structure has increased with respect to temperature. Irrespective of temperature, the position of XRD reflections in Pt 0.5 Ru 0.5 /C doesn’t change but there is considerable change in the broadness of the peak. Hence, crystallite size is calculated from the Sherrer equation [28] presented in Eq. (2)
λ Lv = φ − − − − − − − )2( FWHM .cos θ
Where Lv is the volume-weighted column length, φ is the shape factor (0.89 for spherical crystallite), λ the radiation wavelength (1.5406 Å), FWHM the full width at half maximum, and θ the angle at the maximum intensity. In Table-2, crystallite sizes of Pt/C and
Pt 0.5 Ru 0.5 /C (prepared at different temperatures) were tabulated. It is found to be that Pt/C
- 76 - and Pt 0.5 Ru 0.5 /C prepared at 130 °C & 180 °C have similar crystallite size but Pt 0.5 Ru 0.5 /C prepared at 160 °C gives smaller crystallite size of 2.8 nm.
4.3.2.2 TEM studies of Pt/C and Pt 0.5 Ru 0.5 /C
Fig. 4 TEM Photographs of (A) Pt/C, (B) Ru/C and (C) Pt 0.5 Ru 0.5 /C-160 °°°C catalysts synthesized by the polyol method activated by microwave irradiation.
Fig. 4 shows the TEM images of Pt/C, Ru/C and Pt 0.5 Ru 0.5 /C-160 °C. All samples display a homogeneous repartition of metallic particles on the carbon support. Assuming a spherical shape of the metallic clusters, the mean particle sizes could be evaluated according to the following relation:
n ∑ n d d = i=1 i i − − − − − − − )3( n
Where ni, di, and n are the number of particles of diameter di, the diameter of particles and the total number of particles, respectively. The determination of the mean catalyst particle size has been made for each catalytic powder. In the case of the pure ruthenium catalyst (Fig.
4B), the contrast of Ru on carbon are low and the particles have very small diameters, lower
- 77 - than 1.5 nm. The determination of the mean size was difficult to perform. Higher mean particle size was found for the pure Pt/C ca. 3.0 nm (Fig. 4B). The Pt 0.5 Ru 0.5 /C sample displayed an intermediate mean particle size of ca. 2.5 nm (Fig. 4C) between those of both pure metals. Crystallite size calculated from XRD is also in very close agreement with TEM results.
4.3.2.3 Voltammetric studies of Pt/C and Pt 0.5 Ru 0.5 /C
Fig. 5 Cyclic voltammograms of Pt /C and Pt 0.5 Ru 0.5 /C (synthesized at three different
–1 temperatures) recorded at a scan rate of 20 mV. s in a N 2-saturated 0.5 M H 2SO 4 electrolyte at 20 °°°C.
The active surface areas of Pt/C and Pt 0.5 Ru 0.5 /C (prepared at three different temperatures) catalysts were determined from cyclic voltammograms (Fig. 5) by integrating the charge in the hydrogen desorption region after correcting double layer capacity contribution [29, 30], considering a charge of 210 µC per square centimeter for the
- 78 - desorption of a hydrogen monolayer from a smooth platinum surface [31, 32]. Measurements
–1 were carried out in a N 2-saturated electrolyte at a scan rate of 20 mV s between 0.05 V and
1.2 V vs. RHE for Pt/C and 0.05 V and 0.8 V vs. RHE for Pt 0.5 Ru 0.5 /C. The latter upper limit of potential was used in order to avoid dissolution of the ruthenium from the surface. Typical voltammogram of surface clean Pt/C and Pt 0.5 Ru 0.5 /C were recorded from which active surface areas were calculated and tabulated in table-3.
Table-3: Active surface area calculation from the integration the charge in hydrogen desorption and CO oxidation regions
Active surface area Samples 2 2 Hupd , m /g CO Stripping , m /g 20% Pt/C 80.0 80.0
20% Pt 0.5 Ru 0.5 /C -130 °C 26.2 58.8
20% Pt 0.5 Ru 0.5 /C -160 °C 29.5 57.0
20% Pt 0.5 Ru 0.5 /C -180 °C 21.0 60.5
CO stripping measurements were used in order to obtain information on structure, and activity of nanoparticles [33–36]. For this purpose, platinum catalyst surface was saturated with CO at 0.1 V vs. RHE for 5 min. Before CO stripping measurements recorded
–1 at a scan rate of 20 mVs (Fig. 6), CO was removed from the electrolyte bulk by N 2 bubbling for 15 min, under potential control at 0.1 V. This potential was controlled to maintain CO adsorbtion on the electrode surface. For the Pt/C catalyst (Fig. 6A), the complete disappearance of the current peaks in the hydrogen desorption region (from 0.05 V to 0.4 V vs. RHE) in the first voltammetric cycle shows that the platinum surface is completely blocked by the presence of adsorbed CO. Then, the oxidation of adsorbed CO into CO 2 is responsible for the positive current peaks in the potential range from 0.65 V to
- 79 - 1.0 V vs. RHE. In the negative going potential scan, current peaks in the hydrogen adsorption region appeared and the second cyclic voltammogram is typical of a clean Pt/C catalyst, which indicates that adsorbed CO was totally removed from the platinum surface during the first volatmmetric cycle. From the determination of the charge involved for the oxidation of
2 –1 the adsorbed CO, the same active surface area value as from the H upd region of ca. 80 m g could be determined considering a charge of 420 µC per square centimeter for the desorption of a CO monolayer from a smooth platinum surface, which is an indication of the cleanness of the catalyst surface.
Fig. 6 CO stripping voltammograms of (A) Pt /C and (B) Pt 0.5 Ru 0.5 /C (synthesized at
–1 three different temperatures) recorded at a scan rate of 20 mV s in a N 2-saturated
0.5 M H 2SO 4 electrolyte at 20 °°°C.
With the Pt 0.5 Ru 0.5 /C catalysts synthesized at different temperature using microwave assisted polyol method (Fig. 6B), the onset potential of CO ads oxidation is shifted by ca. 0.2 V
- 80 - towards lower potentials compared with that of Pt/C catalyst, as a result of the presence of ruthenium at the material surface [37]. The active surface area calculated from the charge involved for the oxidation of the CO saturating layer is shown in table-3. The active surface area caluculated from CO stripping were found to be higher than that calculated from the
Hupd region. This discrepancy in the active surface area values determined from H upd region and from CO stripping experiment can be explained by the ability of ruthenium atoms to adsorb CO [38]. By comparing the active surface area values calculated from H upd region and
CO stripping measurement, and assuming that hydrogen does not adsorb on Ru sites and that
CO is linearly adsorbed on both metal sites, it is possible to propose the Pt/Ru atomic surface
2 -1 composition. The Pt 0.5 Ru 0.5 /C synthesized at 160 °C showed active surface area of 57 m g from CO stripping method and this value is almost twice than that calculated from the H upd region. Hence, we propose that the amount of Pt and Ru on the surface of the electrode is nearly equal to 1:1 ratio. In the case of Pt 0.5Ru 0.5 /C synthesized at 130 °C and 180 °C, the active surface area calculated from CO stripping method is higher than the twice the value calculated from H upd . This shows that the amount of ruthenium is higher compare to platinum on the surface. The methanol oxidation is surface sensitive reaction and thus these atomic ratio variations of platinum and ruthenium atoms on the surface will definitely play an important role. Now, let us focus on methanol oxidation at Pt/C and Pt 0.5 Ru 0.5 /C catalysts.
4.3.2.4 Electro-oxidation of methanol on Pt/C and Pt 0.5 Ru 0.5 /C
The polarisation curves recorded in presence of 1.0 M methanol in the supporting electrolyte is given in Fig. 7. The onset potential of methanol oxidation is ca. 0.35 V vs RHE at Pt 0.5 Ru 0.5 /C, i.e. 0.2 V less positive than that at Pt/C catalyst, which is agreement with
- 81 - result obtained with PtRu catalysts prepared by other colloidal methods [39] under the same experimental conditions. The Pt 0.5 Ru 0.5 /C catalyst displayed higher catalytic activity towards methanol oxidation from 0.35 V vs. RHE to ca. 0.65 V vs. RHE. For higher potentials, the pure platinum supported catalyst becomes the more active one. From this potential, platinum is able to activate water and is no longer poisoned by adsorbed CO species; therefore this material becomes more active for methanol electrooxidation, than platinum–ruthenium catalyst [40].
Fig. 7 polarization curves recorded on Pt /C and Pt 0.5 Ru 0.5 /C (synthesized at three
–1 different temperatures) recorded at a scan rate of 5 mVs in a N 2-saturated
0.5 M H 2SO 4 electrolyte containing 1.0 M methanol at 20 °°°C.
- 82 - Both CO stripping and methanol electrooxidation experiments emphasize the well- known role of ruthenium, i.e. allowing the bifunctional mechanism at lower potential than on pure platinum [41, 42]. The presence of ruthenium brings down the water discharging at lower potentials than platinum which is necessary to complete the oxidation of adsorbed CO into CO 2, making the bimetallic catalyst more active than Pt/C from 0.35 V to 0.7 V vs.
RHE. In addition to this, particle size of Pt-Ru also plays vital role that increase in particle size decreases the active surface area of the catalyst and consequently electrocatalytic activity was lowered [43-45]. Ren et al observed higher catalytic activity for methanol oxidation with
PtRu having a particle size of 2–3 nm [46].
In our studies, we found that Pt 0.5 Ru 0.5 /C synthesized at 160 °C gives maximum mass
−1 current density of ca. 200 A gmetal at ca. 0.7 V higher compare to the Pt 0.5 Ru 0.5 /C synthesized at 130 °C and 180 °C. This may be due to the smaller particle size of Pt 0.5 Ru 0.5 /C synthesized at 160 °C compare to 130 °C and 180 °C (shown in table – 2). In the methanol oxidation reaction, it is very difficult to conclude that particle size alone plays an important factor in the catalytic activity because composition of Pt and Ru should also be studied [46]. Before focusing on the composition of platinum and ruthenium, we have studied the effect of microwave power in the synthesis of Pt 0.5 Ru 0.5 /C at 160 °C.
4.3.3 Effect of microwave power
Microwave (MW) power needs to be optimized in the synthesis of Pt 0.5 Ru 0.5 /C because the increase in temperature is directly proportional to the applied MW power. Hence, this will have major impact on the nucleation rate and resulting particle size of PtRu. We
- 83 - have chosen three different MW powers (10%, 50% and 100 % of 1600 W) to reach the temperature of 160 °C in the synthesis of Pt 0.5 Ru 0.5 /C.
Table-4: Metal loading and crystallite size calculation from TGA and XRD
Metal loading (TGA) Crystallite size (XRD) Samples / weight% /nm
Pt 0.5 Ru 0.5 /C–10 MW 18 4.3
Pt 0.5 Ru 0.5 /C–50 MW 19 4.4
Pt 0.5 Ru 0.5 /C–100 MW 20 2.8
4.3.3.1 TGA and XRD studies of Pt/C and Pt 0.5 Ru 0.5 /C
Metal loading on carbon is calculated using TGA measurements (table-4) which indicates the almost whole metal salts are reduced by irradiation with the microwave for 5 minutes after reaching 160 °C. XRD reflections of Pt 0.5 Ru 0.5 /C synthesized at 160 °C with three different MW powers were given in Fig. 8. As seen earlier in fig. 3, the Pt 0.5 Ru 0.5 /C displayed the diffraction peaks of platinum fcc structure and also reflections for RuO 2 at 35 ° and 55 °. XRD reflections for Pt 0.5 Ru 0.5 /C observed at higher angles compare to platinum which confirms the formation of alloy. The degree of alloying can be found using XRD [27].
The amount of ruthenium present in PtRu/C synthesized at three different MW power was found to be 0.07 % in 10% MW, 14 % in 50% MW and 21 % in 100% MW. Therefore, it can be proposed that the fcc structure of the Pt 0.5 Ru 0.5 /C catalyst in different MW power as
Pt 0.93Ru 0.07 for 10% MW, Pt 0.84 Ru 0.16 for 50% MW, Pt 0.8 Ru 0.2 for 100% MW and also rutile
RuO 2 and hcp Ru nanoclusters.
- 84 -
Fig. 8 XRD diffractograms in 2 θθθ range from 30 °°° to 90 °°° obtained with Pt 0.5 Ru 0.5 /C synthesized at three different microwave powers at 160 °°°C using microwave assisted polyol method.
When XRD reflections of Pt 0.5 Ru 0.5 /C in fig. 8 are keenly observed, the change in the
FWHM of the peak can be identified with MW power. The crystallite size is inversely proportional to FWHM and hence crystallite size is calculated using Scherrer equation [28] given in equation 2. We found that 10% and 50% MW gave crystallite size of 4.3 and 4.4 nm respectively whereas the crystallite size for 100% MW is 2.8 nm (given in table-4).
Crystallite size obtained in 10% and 50% MW is significantly higher for studying the catalytic activity towards methanol electro-oxidation and hence we focused our study on the composition of Pt and Ru.
- 85 - 4.3.4 Effect of composition
From the above studies, we came to the conclusion that PtRu/C with three different compositions can be synthesized by microwave assisted polyol method at 160 °C with 100%
MW power.
4.3.4.1 TGA and XRD studies of Pt/C and PtRu/C
TGA studies showed that complete reduction of metal salt in ethylene glycol takes place during the microwave irradiation and these values are very close to the nominal loading of metal on carbon (given in table-5).
Table-5 Metal loading and crystallite size calculation from TGA and XRD
Samples Metal loading (TGA) / weight% Crystallite size (XRD)/nm
Pt 0.34 Ru 0.66 /C 18 3.3
Pt 0.50Ru 0.50/C 20 2.8
Pt 0.80 Ru 0.20 /C 20 3.7
XRD reflections of Pt 0.34 Ru 0.66 /C, Pt 0.5 Ru 0.5 /C, and Pt 0.8 Ru 0.2 /C are given in fig. 9.
Similar to previous results, diffraction patterns were matched with platinum fcc structure and also from RuO 2 at 35 ° and 55 °. XRD reflections for all the compositions of Pt and Ru found to be at higher angles compare to platinum which confirms the formation of alloy. The degree of alloying can be found using XRD [27]. The ruthenium atomic ratio present in each composition is 0.18 % in Pt 0.34 Ru 0.66 /C, 0.21% in Pt 0.5Ru 0.5 /C, 0.17% in Pt 0.8Ru 0.2 /C (these compositions were calculated from XRD reflections). Therefore, it can be proposed that the structure of Pt 0.34 Ru 0.66 /C contains Pt 0.82 Ru 0.18 , Pt 0.5 Ru 0.5 /C contains Pt 0.79 Ru 0.21 and Pt 0.8 Ru 0.2 contains Pt 0.83Ru 0.17 . In addition to this all the compositions of Pt and Ru also consist of rutile
RuO 2 and hcp Ru nanoclusters. The crystallite sizes were calculated from Scherrer equation
- 86 - [28] given in equation 2 and tabulated in table-5. The Pt 0.5Ru 0.5 /C showed lower crystallite size compare to Pt 0.34 Ru 0.66 /C and Pt 0.8Ru 0.2 /C.
Fig. 9 XRD diffractograms in 2 θθθ range from 30 °°° to 90 °°° obtained with Pt 0.34 Ru 0.66 /C,
Pt 0.5 Ru 0.5 /C and Pt 0.8Ru 0.2/C synthesized at 160 °°°C and 100% microwave power synthesized using microwave assisted polyol method.
- 87 -
4.3.4.2 Voltammetric studies on Pt 0.34 Ru 0.66 /C, Pt 0.5 Ru 0.5 /C and Pt 0.8Ru 0.2/C
Fig. 10 Cyclic voltammogram of Pt 0.34 Ru 0.66 /C, Pt 0.5 Ru 0.5 /C and Pt 0.8Ru 0.2 /C recorded
–1 at scan rate of 20mVs in a N 2-saturated 0.5 M H 2SO 4 electrolyte at 20 °°°C.
The active surface area is calculated in a similar procedure followed in section
4.3.2.3 . Cyclic volammogram shown in fig. 10 was recorded in a N 2-saturated electrolyte at a scan rate of 20 mV s–1 between 0.05 V and 0.8 V vs. RHE. The active surface area values were calculated and tabulated in table-6.
The active surface area is calculated in a similar procedure followed in section
4.3.2.3 . PtRu/C catalyst surface was saturated with CO at 0.1 V vs. RHE for 5 min. Before
CO stripping measurements recorded by cyclic voltammetry at a scan rate of 20 mVs –1
- 88 - (Fig. 11), CO was removed from the electrolyte bulk by N 2 bubbling for 15 min, under potential control at 0.1 V.
Table -6: Active surface area calculation from the integration the charge in hydrogen desorption and CO oxidation regions
Active surface area Samples 2 2 Hupd , m /g CO Stripping , m /g 20% Pt/C 80.0 80.0
20% Pt 0.34Ru0.66 /C 42.0 80.5
20% Pt 0.50 Ru 0.50 /C 29.5 57.0
20% Pt 0.80 Ru 0.20 /C 42.8 61.0
It is known from the literature [37, 41, 42] and also from our earlier discussion in the section 4.3.2.4 that the enhanced catalytic activity in PtRu/C is due to the bifunctional mechanism shown by Pt and Ru. The presence of Ru in the electrode surface lowers the potential for the oxidation of CO. Hence, the amount of ruthenium present in the PtRu will show a marked effect in the catalytic activity of CO stripping. In fig.11, the first voltammetric cycle H upd region is completely blocked by the adsorbed CO. After the oxidation of adsorbed CO to CO 2, the H upd is observed during the second voltammetric cycle.
We have measured onset potential for CO oxidation in all the three compositions and values are found to be 0.521 V, 0.506 V and 0.457 V for Pt0.34 Ru 0.66 /C, Pt 0.5Ru 0.5/C and Pt 0.8Ru 0.2/C respectively. This clearly shows that the onset potential of CO stripping slightly shifted towards lower potentials on decreasing the ruthenium content (showed by dotted line in fig.
11). This result indicates that Pt 0.8Ru 0.2/C showed better catalytic activity towards CO oxidation compare to Pt 0.5Ru 0.5/C and Pt 0.34 Ru 0.66 /C.
- 89 -
Fig. 11 CO stripping voltammograms of Pt 0.34 Ru 0.66 /C, Pt 0.5 Ru 0.5 /C and Pt 0.8Ru 0.2 /C
–1 recorded at a scan rate of 20 mV s in a N 2-saturated 0.5 M H 2SO 4 electrolyte at 20 °°°C.
- 90 - From the table-6, active surface area calculated from CO stripping is found to be higher than that calculated from the H upd region. This discrepancy in the active surface area values is due to the fact that the H upd take place only on Pt but the CO stripping take place on both Pt and Ru. From the active surface area values, it is possible to calculate the surface atomic ratio of Pt and Ru. This feature was discussed in detailed manner in the section
4.3.2.3 and also showed Pt 0.5 Ru 0.5 /C exhibits 1:1 atomic ratio of Pt and Ru on the electrode surface. Hence, now we will discuss about Pt 0.8 Ru 0.2 /C and Pt 0.34 Ru 0.66 /C. The active surface area value for Pt 0.8 Ru 0.2 /C and Pt 0.34 Ru 0.66 /C are given in table-6. These values suggest that the atomic ratio in Pt 0.8 Ru 0.2 /C on the electrode surface is very close to the expected 4:1 ratio.
In contrast Pt 0.34Ru 0.66 /C shows very high values than expected and at present we are unable to explain the appropriate reason for this result.
4.3.4.3 Electro-oxidation of methanol on Pt 0.34 Ru 0.66 /C, Pt 0.5 Ru 0.5 /C and Pt 0.8Ru 0.2/C
The polarisation curves recorded for methanol oxidation on three different compositions namely Pt 0.34 Ru 0.66 /C, Pt 0.5Ru 0.5/C and Pt 0.8Ru 0.2/C is given in fig. 13. The onset potential of methanol oxidation is ca. 0.35 V vs RHE at all the three compositions of PtRu/C which is 0.2 V lower than Pt/C. The reason for methanol oxidation at lower potential is explained in section. 4.3.2.4. We found that Pt 0.8Ru 0.2/C showed maximum mass current
−1 density of ca. 290 A gmetal at ca. 0.73 V higher compare to the Pt 0.5 Ru 0.5 /C and Pt 0.34 Ru 0.66 /C.
Similar results were reported in the literature [3, 47, 48] and they suggests that the electrocatalytic activity is depends on composition, temperature and oxidation potential.
Methanol oxidation below 0.5 V vs. RHE at 25 °C requires high content of platinum catalyst to show better catalytic activity compare to Pt 0.5 Ru 0.5 /C whereas at the temperature above
- 91 - 45 °C, Pt 0.5 Ru 0.5 /C shows higher catalytic activity compare to PtRu with higher content of Pt.
The temperature dependant enhancement is due to ruthenium gains more ability to adsorb and dehydrogenate methanol at temperatures above 40 °C [3]. In our study, we have carried out the methanol oxidation experiment at 20 °C and hence Pt rich catalyst (Pt 0.8Ru 0.2 /C) shows higher catalytic activity compare to Pt 0.5 Ru 0.5 /C and Pt 0.34 Ru 0.66 /C.
Fig. 12 Polarization curves recorded on Pt 0.34 Ru 0.66 /C, Pt 0.5 Ru 0.5 /C and Pt 0.8Ru 0.2 /C
–1 recorded at a scan rate of 20 mV s in a N 2-saturated 0.5 M H 2SO 4 electrolyte containing 1.0 M methanol at 20 °°°C.
- 92 - The chronoamperometry curves recorded at 0.6 V on Pt 0.8Ru 0.2 /C, Pt 0.5Ru 0.5 /C and
Pt 0.34 Ru 0.66 /C in the presence of 1.0 M MeOH in the supporting electrolyte is given in
Fig. 13. This analysis will give information on the long term activity of the electrocatalyst.
The potential 0.6 V was chosen slightly higher than the onset of CO and methanol oxidation potential. The main factors affecting the resulting current are methanol adsorption/desorption, direct oxidation of methanol, oxidation of adsorbed CO, poisoning of catalyst by CO and anion adsorption etc., and hence this analysis results in the estimation of the long term activity of the catalyst [44].
Fig. 13 Chronoamperometry curve recorded on Pt 0.34 Ru 0.66 /C, Pt 0.5 Ru 0.5 /C and
Pt 0.8 Ru 0.2 /C in N 2-saturated 0.5 M H 2SO 4 electrolyte containing 1.0 M methanol
(T = 20 °°°C) at 0.6 V vs. RHE.
- 93 - In fig .13, the initial current drop is observed in all the catalysts which are due to the decay of non-faradic current, sulphate anion adsorption and also due to poisoning of the catalyst surface [49-51]. This current decay slows down and reaches almost steady state after
−1 a short duration. The mass current density values at 3600 seconds are 26 Agmetal ,
−1 −1 70 A gmetal and 74 A gmetal for Pt 0.8Ru 0.2 /C, Pt 0.5Ru 0.5 /C and Pt 0.34 Ru 0.66 /C respectively. These values suggest that Pt 0.8Ru 0.2 /C and Pt 0.5Ru 0.5 /C show higher activity than Pt 0.34 Ru 0.66 /C.
Eventhough polarization studies showed Pt 0.8Ru 0.2 /C as the best catalyst but the steady state mass current density of Pt 0.8Ru 0.2 /C is nearly equivalent to Pt 0.5Ru 0.5 /C. This can be due to poisoning of the catalyst due to CO might be more in Pt 0.8Ru 0.2 /C compare to Pt 0.5Ru 0.5 /C.
4.3.5 Substractively Normalized Interfacial Fourier Transform Infra Red Spectroscopy study on Pt 0.5 Ru 0.5 /C
Fig. 14 shows SNIFTIR spectra recorded between 1900 cm –1 and 2400 cm –1 at
Pt 0.5 Ru 0.5 /C catalytic surface with 1.0 M methanol. The absorption bands located between
–1 2030 and 2055 cm are assigned to linearly bonded CO (CO L) [52]. No additional typical
CO absorption band around 1950 cm –1 is visible, which could be assigned to bridge-bonded
CO (CO B) species [53].
- 94 -
0.1 – 0.3 V vs. RHE 0.2 – 0.4 V vs. RHE
0.3 – 0.5 V vs. RHE
0.4 – 0.6 V vs. RHE
0.5 – 0.7 V vs. RHE
ΔR 4 0.6 – 0.8 V vs. RHE === 5×××10−−− R
1900 2000 2100 2200 2300 2400 wavenumber / cm -1
Fig. 14 SNIFTIR spectra of recorded during the electrooxidation of 1.0 M methanol in
0.5 M H 2SO 4 on Pt 0.5 Ru 0.5 /C.
The CO L band appears in the first potential modulation between 0.1 and 0.3 V vs.
RHE, indicating that methanol chemisorption and dehydrogenation to form adsorbed CO takes place at very low potentials on Pt 0.5 Ru 0.5 /C catalyst. The intensity of these bands first increases with the increase of potential to reach a maximum in the second (0.2 and 0.4 V) and third (0.3 and 0.5 V) modulations. Then, it starts to decrease from the fourth modulation
(0.4 and 0.6 V), for which an absorption band located at 2345 cm –1 corresponding to interfacial CO 2 [52, 53] starts to be observed. In addition, the CO L absorption band first undergo a red shift as the average potential modulation increases, and then, from the third potential modulation (0.3 and 0.5 V), a blue shift, which are due to a combination of the
Stark effect (red shift) [54] with a coverage-dependent shift of the CO infrared band (blue shift) [55].
- 95 - 4.4 Conclusions
In this work, we have synthesized Pt/C and PtRu/C by microwave assisted polyol method which is considered as an industrially scalable method. First, the method is based on a polyol route which uses inexpensive solvent (ethylene glycol), does not need the presence of surfactant and is very easy to implement. Second, instead of using traditional thermal activation, the activation by microwave irradiation allows fast synthesis, lower energy consumption and uniform heating. Temperature, microwave power and composition of Pt and Ru were optimized to harvest the better electrocatalytic activity towards methanol electro-oxidation. PtRu/C catalysts were composed of a Pt/Ru alloy in interaction with Ru and RuO 2 clusters. The optimized temperature was 160 °C and 100 % microwave power for this synthesis. Pt 0.5Ru 0.5 /C and Pt 0.8Ru 0.2 /C showed good catalytic activity towards CO and methanol electro-oxidation. From the SNIFTIR spectroscopy study, we have shown that the dehydrogenation of methanol takes place at very low potentials and methanol oxidation was followed by linearly bonded CO was confirmed. Pt 0.5Ru 0.5 /C showed long term stability of
−1 the catalyst with constant mass current density of 70 A gmetal during methanol electro- oxidation.
References
1. W.R. Grove, Phil. Mag. 21 (1842) 417.
2. K.H. Choi, H.S. Kim, T.H. Lee, J. Power Sources 75 (1998) 230.
3. C. Coutanceau, A. Rakotondrainibe, A. Lima, E. Garnier, S. Pronier, J.M. Léger, C.
Lamy, J. Appl. Electrochem. 34 (2004) 61.
- 96 - 4. A. Caillard, C. Coutanceau, P. Brault, J. Mathias, J.-M. Léger, J. Power Sources 162
(2006) 66.
5. L. Dubau, C. Coutanceau, E. Garnier, J.-M. Léger, C. Lamy, J. Appl. Electrochem. 33
(2003) 419.
6. F. Vigier, C. Coutanceau, A. Perrard, E.M. Belgsir, C. Lamy, J. Appl. Electrochem.
34 (2004) 439.
7. H.S. Oh, J.G. Oh, H. Kim, J. Power Sources 183 (2008) 600.
8. J. Guo, G. Sun, S. Shiguo, Y. Shiyou, Y. Weiqian, Q. Jing, Y. Yushan, X. Qin,
J. Power Sources 168 (2007) 299.
9. F. Fievet, J.P. Lagier, B. Blin, B. Beaudoin, M. Figlarz, Solid State Ionics 32–33
(1989) 198.
10. C. Bock, C. Paquet, M. Couillard, G.A. Botton, B.R. MacDougall, J. Am. Chem. Soc.
126 (2004) 8028.
11. Z. Liu, L.M. Gan, L. Hong, W. Chen, J.Y. Lee, J. Power Sources 139 (2005) 73.
12. C. Grolleau, C. Coutanceau, F. Pierre, J.M. Leger, J. Power Sources 195 (2010) 1569.
13. B.M. Babic, Lj. M. Vracar, V. Radmilovic, N.V. Krstajic, Electrochim. Acta 51
(2006) 3820.
14. E. Lebègue, S. Baranton, C. Coutanceau, J. Power Sources 196 (2011) 920.
15. S.A. Galema, Chem. Soc. Rev. 26 (1997) 233.
16. M. Tsuji, M. Hashimoto, Y. Nishizawa, M. Kubokawa, T. Tsuji, Chem. Eur. J. 11
(2005) 440.
17. W.S. Rasband, Image J, U S. National Institutes of Health, Bethesda, Maryland,
USA, http://imagej.nih.gov/ij/, 1997e2011.
- 97 - 18. F. Gloaguen, N. Andolfatto, R. Durand, P. Ozil, J. Appl. Electrochem. 24 (1994) 863.
19. Z. Zhou, W. Zhou, S. Wang, G. Wang, L. Jiang, H. Li, G. Sun, Q. Xin, Catal. Today
93-95 (2004) 523.
20. Z. Liu, X.Y. Ling, X. Su, J.Y. Lee, J. Phys. Chem. B 108 (2004) 8234.
21. D.N. Furlong, A. Launikonis, W.H.F. Sesse, L.V. Sanders, J. Chem. Soc. Faraday
Trans. 180 (1984) 571.
22. T. Teranishi, M. Hosoe, T. Tanaka, M. Miyake, J. Phys. Chem. B 103 (1999) 3818.
23. R. Harpeness, Z. Peng, X. Liu, G. V-Pol, Y. Koltypin, A. Gedanken, J. Colloid.
Interface Sci. 287 (2005) 678.
24. M. Wojdyr, J. Appl. Cryst 43 (2010) 1126.
25. A. Devadas, S. Baranton, T.W. Napporn, C. Coutanceau, J. Power Sources 196
(2011) 4044.
26. P. Vogel, H. Britz Bönnemann, J. Rothe, J. Hormes, J. Phys. Chem. B 101 (1997)
11029.
27. C.A. Angelucci, M. D’Villa Silva, F.C. Nart, Electrochim. Acta 52 (2007) 7293.
28. P. Scherrer, Nachr. Ges. Wiss. Göttingen, Math.-Phys. Klasse 26 (1918) 98.
29. C. Coutanceau, M.J. Croissant, T. Napporn, C. Lamy, Electrochim. Acta 46 (2000)
579.
30. C. Grolleau, C. Coutanceau, F. Pierre, J.M. Léger, Electrochim. Acta 53 (2008) 7157.
31. V.S. Bakotzky, Y.B. Vassilyev, Electrochim. Acta 12 (1967) 1323.
32. N.M. Markovic, B.N. Grgur, P.N. Ross, J. Phys. Chem. B 101 (1997) 5405.
33. F. Maillard, S. Schreier, M. Hanzlik, E.R. Savinova, S. Weinkauf, U. Stimming,
Phys. Chem. Chem. Phys. 7 (2005) 385.
- 98 - 34. M. Arenz, K.J.J. Mayrhofer, V. Stamenkovic, B.B. Blizanac, T. Tomoyuki, P.N.
Ross, N.M. Markovic, J. Am. Chem. Soc. 127 (2005) 6819.
35. A. Cuesta, A. Couto, A. Rincón, M. Pérez, A. López Cudero, C. Gutíerrez, J.
Electroanal. Chem. 586 (2006) 184.
36. A. L. Cudero, J. S. Gullón, E. Herrero, A. Aldaz, J.M. Feliu, J. Electroanal. Chem.
644 (2010) 117.
37. M. Watanabe, S. Motoo, J. Electroanal. Chem. 60 (1975) 275.
38. L. Dubau, F. Hahn, C. Coutanceau, J.M. Léger, C. Lamy, J. Electroanal. Chem. 554–
555 (2003) 407.
39. C. Coutanceau, S. Brimaud, L. Dubau, C. Lamy, J.-M. Léger, S. Rousseau, F. Vigier,
Electrochim. Acta 53 (2008) 6865.
40. H.A. Gasteiger, N. Markovic, P.N. Ross, E.J. Cairns, J. Electrochem. Soc. 141 (1994)
1795.
41. M. Watanabe, S. Motoo, J. Electroanal. Chem. 75 (1960) 267.
42. M. Watanabe, S. Motoo, J. Electroanal. Chem. 75 (1960) 275.
43. Y. Takasu, H. Itaya, T. Iwazaki, R. Miyoshi, T. Ohnuma, W. Sugimoto, Y.
Murakami, Chem. Commun. (2001) 341.
44. X. Li, X. Qiu, H. Yuan, L. Chen, W. Zhu, J. Power Sources 184 (2008) 353,
45. S. Aricò, S. Srinivasan, V. Antonucci, Fuel cells 1 (2001) 133
46. X. Ren, P. Zelenay, S. Thomas, J. Davey, and S. Gottesfeld, J. Power Sources 2000,
86, 111
47. M. Neergat, D. Leveratto, U. Stimming, Fuel cells 2 (2002) 25
48. C. Xu, L. Wang, X. Mu, Y. Ding, Langmuir 26 (2010) 7437
- 99 - 49. L.H. Jiang, G. Q.Sun, X. S. Zhao, Z. H. Zhou, S. Y. Yan, S. H. Tang, G. X. Wang, B.
Zhou, Q. Xin, Electrochim. Acta 50 (2005) 2371
50. J. W. Guo, T. S. Zhao, J. Prabhuram, R. Chen, C. W. Wong, Electrochim. Acta 51
(2005) 754
51. J. Jiang, A. Kucernak, J. Electroanal. Chem. 543 (2003) 187
52. K. Kunimatsu, J. Electroanal. Chem. 140 (1982) 205
53. B. Beden, F. Hahn, S. Juanto, C. Lamy, J.-M. Léger, J. Electroanal. Chem. 225
(1987) 215
54. B. Beden, C. Lamy, A. Bewick, K. Kunimatsu, J. Electroanal. Chem. 121 (1981) 343
55. C. Rice, Y.Y. Tong, E. Oldfield, A. Wieckowski, F. Hahn, F. Gloaguen, J.-M. Léger,
C. Lamy, J. Phys. Chem. B 104 (2000) 5803.
- 100 - 5.1. Introduction
Co 3O4 attracts the attention of the materials researchers because of its promising applications in various fields viz., anode material for Li-ion batteries (LIBs) [1], super capacitors [2, 3], gas sensors [4], catalytic processes [5] etc. Highly porous nanostructured materials of high surface area are required for these applications. Co 3O4 nanoparticles were synthesized through various routes that include the commonly employed oxidative precipitation [6], thermal decomposition [7-9], hydrothermal synthesis [10-13] etc.
In the recent past, porous Co 3O4 was synthesized from different single source precursors by simple thermal decomposition. For instance, single source precursors include
(NH 4)2Co 8(CO 3)6(OH) 6.4H 2O [14], Co(CO 3)0.5 (OH)0.11H 2O [8,9,15], Co 4(CO) 12 [16]. The main advantage of using single source precursor was the formation of porous structure because of the expulsion of large volume of gaseous product during thermal decomposition.
Porous Co 3O4 has received special interest as anode materials in LIBs due to high theoretical capacity (890 mAh g -1) compared to commercially used graphite anode (372 mAhg-1) [1].
Even though, Co 3O4 exhibits high discharge capacity, the practical usage in LIBs was mostly hindered by poor cycling performance [15]. During several charge-discharge cycles, severe volume variation occurs due to conversion reaction and this leads to decrease in cycle life of the LIB. One possible way to alleviate this problem is to design a porous nanostructured material. Porous material endows the sufficient space for the volume change during charging and discharging of cell and it also gives the facile transport of lithium and electron.
In the present work, we have explored Prussian blue type compound as a new precursor for the synthesis of porous Co 3O4 and used as an anode material in LIB. Few groups have studied the formation of metal alloys/oxides by thermal decomposition of
- 101 - Prussian blue type compounds [17-23]. Bocarsly’s group has studied extensively on Prussian blue like compounds generally called as “cyanogels”. Cyanogels are inorganic polymer system that results from the polymerization of chlorometalate and cyanometalate ions. On polymerization, two of the chloride ligands on the chlorometalate were replaced by the nitrogen end of the cyanide ligand on the cyanometalate creating the cyanide bridge.
(Chapter-6 has more details about cyanogel chemistry) [18-20]. In a typical synthesis,
Na 2PdCl 4 and K 3[Co(CN) 6] were mixed together to form PdCo “cyanogel”, which was decomposed at elevated temperatures to form PdCo alloy. It was inferred from the available experimental results that CoPd(CN) 4 and Pd(CN) 2 are formed as intermediates which on further heating results in the formation of PdCo alloy [19]. Epple’s group found that the crystal structure of the precursor dictates the structure and morphology of the resulting products when thermolysis was carried out under moderate temperatures [22-23]. In one of the studies, Cu/ZnO catalyst was synthesized from metal complexes containing cyanide as a ligand in one case and ethylenediamine and cyanide as ligands in the other [23]. The catalytic activity towards the formation of methanol from synthetic gas (CO/CO 2/H 2) was studied using Cu/ZnO catalyst. Surprisingly, catalytic activity was not observed on Cu/ZnO catalysts synthesized from Cu[Zn(CN) 3] whereas the Cu/ZnO synthesized from ethylenediamine and cyanide as ligands showed 20-30 % catalytic activity [23]. These observations highlight the role of the precursor in determining the crystal structure of metal oxides and their catalytic properties. In this work, we present the formation of porous Co 3O4 from Co 3 [Co(CN) 6]2 which has similar crystal structure. Both the compounds are cubic structures in which Co 2+ ,
Co 3+ are positioned in tetrahedral and octahedral co-ordination, respectively [24]. This is one
- 102 - of the main motivations for choosing this unique precursor. Other motivation is that facile synthesis of porous Co 3O4 from Co 3[Co(CN) 6]2 by thermal decomposition.
5.2 Experimental procedures
5. 2.1 Synthesis of Co 3[Co(CN) 6]2.12 H 2O
Co 3[Co(CN) 6]2.12 H 2O was synthesized by dropwise addition of 0.15 M cobalt acetate (E-Merck) solution to 0.1 M potassium hexacyanocobaltate(II) (Sigma-Aldrich) in the molar ratio of 3:2 under vigorous stirring for 20 minutes at room temperature (shown in scheme 1). The solution was left aside for an hour and the precipitate was separated by centrifugation at 8000 rpm. The precipitate was washed twice with double distilled water and then with acetone. The product was dried overnight under vacuum at 50 °C.
Scheme 1: Schematic representation of Co 3O4 synthesis
5.2.2 Synthesis of Co 3O4
Co 3O4 was prepared by thermal decomposition of Co 3[Co(CN) 6]2.12H 2O at various temperatures for two hours under mixed argon/oxygen atmosphere in a tubular furnace
(shown in scheme 1). Mixed atmosphere was used to slow down the decomposition reaction.
- 103 - The toxic gaseous products evolved during the decomposition of Co 3[Co(CN) 6]2. 12H 2O were trapped in alkali. The temperature was optimized to obtain pure Co 3O4 phase.
Co 3[Co(CN) 6]2.12 H 2O was also subjected to microwave irradiation by using domestic kitchen microwave oven (IFB model 25SC1).
5.2.3 Instrumentation
Centrifugation was done using Hettich universal 320R. X-ray diffraction (XRD) patterns were recorded using PANalytical diffractometer (Model PW3040/60 X’pert PRO) operated with Cu K α radiation ( λKα = 0.154 nm) generated at 40 kV and 20mA. Fourier transform-infra red (FT-IR) spectra were recorded using Nexus (670 model) with DTGS detector. Decomposition temperature of single source precursor was examined using PL
Thermal Sciences Instrument (Model STA 1500) in O2 flow from room temperature to 1000
°C, at a heating rate of 10 °C minute -1.
Scanning electron microscope (SEM) studies were made using Hitachi (Model S-
3000H) with 10 kV (acceleration voltage) . Transmission electron microscope (TEM) analysis was made by placing a drop of sample dispersed in acetone onto a copper grid coated with carbon film (400 meshes) and then dried. TEM images were collected from Philips CM200 microscope working at 200 kV.
5.2.4 Electrochemical studies
The as-obtained material was tested as working electrode against lithium using 2032 type coin cells. Slurry containing Co 3O4 (prepared by the above method), SP carbon and binder (Polyvinylidene difluoride) in the weight ratio of 85:10:5 was prepared and was then cast as a thin film (of 9 µm thick) over a Cu foil (current collector) using K-control coater
(with 60µm rod). The film was air dried for 2 to 3 hours and then transferred to a vacuum
- 104 - oven maintained at 100 °C and dried overnight. The film was calendared in a press at 2 Pa for a minute and cut into 13 mm diameter discs using a precision disc cutter (DANVEC). The sealing and assembling of cells were carried out inside an argon filled glove box (O2, H 2O
≤ 0.5 ppm). The half cells composed of Co 3O4 coated Cu disc was the working electrode and metallic lithium pressed onto a stainless steel current collector was used as the counter/reference electrode. The two electrodes were separated by Whatman glass fiber WD-
40 discs and 1 M LiPF 6 dissolved in EC/DMC (1:1 by vol.) was used as the electrolyte.
Cyclic voltammetery and electrochemical impedence spectroscopy were carried out using
Autolab electro-chemical work station. In cyclic voltammetry, the electrode was scanned in the potential region of 0.0 V to 3.0 V at the scan rate of 0.05 mV s -1. Electrochemical impedence measurement was studied in the frequency range from 100 MHz to 100 mHz at open circuit potential with an alternating current perturbation of 10 mV. Charge/discharge cycles were carried out on Li half cells in a voltage range from 3.0 V to 0 V using VMP3
(Bio-Logic) multichannel potentiostat/galvanostat. The charge-discharge studies were carried out at C/2 rate.
5.3. Results and discussion
5.3.1 Synthesis and Characterization of Co 3[Co(CN) 6]2.12H 2O
Co 3[Co(CN) 6]2.12H 2O was precipitated by mixing the solutions of cobalt acetate and potassium hexacyano cobaltate (III). The product can easily be identified by its color transitions in hydrated and dehydrated forms. The color of the synthesized compound was pink, while the color changed to dark blue when dehydrated by heating (shown in fig. 1). The color transition was reversible in nature and it turned to pink immediately on absorbing
- 105 - moisture from atmosphere. These physical changes also reported in the literature and were attributed to the transformation of octahedral to tetrahedral co-ordination of Co 2+ site [24,
25]. The compound was further confirmed by using FT-IR and XRD.
Fig.1 Reversible color transition of Co 3[Co(CN) 6]2.12 H 2O (Pink colour) and
Co 3[Co(CN) 6]2 (blue colour) (Snapshot from camera).
Fig.2 FT-IR spectra of K 3[Co(CN) 6] and Co 3[Co(CN) 6]2. 12 H 2O.
- 106 - From the FT-IR spectra (Fig.2), we can observe a positive shift in its stretching vibration of cyanide group compared to the parent molecule K 3[Co(CN) 6]. This observation was quite expected because only the terminal cyanide ligands exist in K 3[Co(CN) 6] while in the case of Co 3[Co(CN) 6]2.12H 2O, two type of cyanide ligands exist, viz (i) terminal cyanide and (ii) bridged cyanide. The synthesized compound, Co 3[Co(CN) 6]2.12 H 2O showed a broad cyanide stretching band at 2174 cm -1 which is in concurrence with the observation made by
Brown et al [25 ] and the cyanide stretching vibration of corresponding parent molecule is
2129 cm -1.
Fig.3 shows the XRD pattern of Co 3[Co(CN) 6]2.12H 2O. The (hkl) planes of the compound can be indexed as face centered cubic phase that is in good agreement with the standard pattern (JCPDS reference code 01-089-3737).
Fig.3 XRD of Co 3[Co(CN) 6]2.12 H 2O.
- 107 - In fig. 4, thermogram of Co 3[Co(CN) 6]2.12 H 2O showed three regions of weight losses in fig. 3. (i) Weight loss of 25.5% upto 120° C could be due to loss of water molecules; (ii) 27.15% weight loss at 302° C corresponds to the loss of cyanide ligands and formation of mixed cobalt oxides and (iii) 3.65% loss at > 900° C is due to the conversion of
Co 3O4 to CoO.
o Fig.4 TGA and DTA of Co 3[Co(CN) 6]2. 12 H 2O in air, heating rate 10 C.
5.3.2 Phase composition analysis
From TGA, it was observed that the decomposition of cyanide ligands was completed
° below 350 C and hence we have synthesized cobalt oxide by fixing the temperature range
- 108 - ° ° from 400 C to 650 C in mixed Ar/O 2 atmosphere. In fig. 5, XRD analysis of cobalt oxide
° prepared at 400 C was found to have a mixture of CoO and Co 3O4 phases which did not match with the XRD pattern of Co 3[Co(CN) 6]2 precursor (shown in fig.3), thereby confirming the decomposition of Co 3[Co(CN) 6]2.
Fig.5 XRD of cobalt oxide prepared at different temperature. (#) CoO phase and
(+) Co 3O4 phase
The existence of the CoO phase at a lower temperature may be due to the initial conversion of Co 3[Co(CN) 6]2 partly to cobalt metal and then to metal oxides, viz. CoO and
Co 3O4. This mechanism was suggested because during thermal decomposition, cyanide oxidizes to cyanogens and reduction of metal center takes place [19]. XRDs were recorded for the samples prepared at different temperatures and revealed the formation of mixed cobalt oxides at temperature below 650 °C (Fig. 5). Reflections for CoO phase was found at the 2θ
- 109 - positions: 36.4, 42.4, 61.4 which can be matched with the standard (JCPDS No: 01-078-
0431) while other reflections at 19.0, 31.2, 36.8, 38.5, 44.8, 55.6, 59.3, and 65.2 can be indexed to Co 3O4 phase (JCPDS No. 01-078-1969 ). Also, the peak intensity of CoO phase gradually decreases with increasing temperature and the pure Co 3O4 phase was found at
650 °C.
o Fig. 6 shows the FT-IR spectrum of Co 3O4 prepared at 650 C. Two well defined sharp peaks of the characteristic metal-oxygen stretching were found at 574 and 663 cm -1 [26]. The former stretching was attributed to Co 3+ ion positioned at octahedral sites while the latter
2+ corresponds to Co in the tetrahedral position of Co 3O4 crystal structure.
o Fig.6 FT-IR spectrum of Co 3O4 obtained at 650 C
- 110 - 5.3.3 Morphology and particle size studies
400 °°°C 500 °°°C
5 µµµm 5 µµµm
600 °°°C 65 0 °°°C
5 µµµm 5 µµµm
Fig. 7 SEM images of cobalt oxide prepared at different temperatures
Decomposition of Co 3[Co(CN) 6]2, results in a significant change in the mass of the sample. The possible decomposition products in the mixed Ar/O 2 atmosphere are H 2O, CO 2,
(CN) 2 and N xOy [20]. The release of these gaseous products could produce highly porous metal oxide materials and these were confirmed by SEM images of porous Co 3O4 prepared at different temperatures (shown in Fig.7). We can observe that porous structure in Co 3O4 was retained irrespective of the temperature but the pore volume increased with temperature.
- 111 - The shape and particle size of the ripened Co 3O4 was analyzed from TEM studies.
These images show (Fig. 8) a large void space in the material. The regular distribution of these void spaces in the matrix could have been generated due to release of gaseous products
(as discusses above) decomposition of the cyano complex followed by its crystallization to polygon-shaped Co 3O4 crystals of ~ 20 nm.
o Fig. 8 TEM image of Co 3O4 prepared at 650 C
5.3.4 Decomposition of Co 3[Co(CN) 6]2 by microwave heating
It was interesting to know that Co 3[Co(CN) 6]2 was active in the microwave region and hence microwave irradiation can also be followed to decompose Co 3[Co(CN) 6]2. Initial attempts were made to synthesize cobalt oxide by microwave irradiation in the domestic kitchen microwave oven. We found that within few minutes compound undergoes decomposition and obtained products were characterized using SEM (for morphology) and
XRD (for phase composition). Surprisingly, the microwave assisted synthesis also results in the porous cobalt oxide (Fig.9A) which is similar to thermal decomposition method. XRD
- 112 - reflections in Fig. 9B show the existence of cobalt oxide in two different phase’s c.a Co 3O4 and CoO.
B
Fig.9 A) SEM image of cobalt oxide B) XRD of cobalt oxide synthesized from microwave irradiation of Co 3[Co(CN) 6]2. 12 H 2O, (#) CoO phase (+) Co 3O4 phase.
Phase pure Co 3O4 can be synthesized by optimizing the microwave power and irradiation time. The initial results suggests that Co 3[Co(CN) 6]2 is a microwave susceptor and hence microwave irradiation can be used to decompose metal cyano compounds into metal
- 113 - oxides on a very short time scale compare to conventional heating. However, we know that further studies are required to understand the mechanism of the reaction and hence, the studies are underway in our laboratory.
5.3.5 Electrochemical studies against Li +/Li
Discovery of “conversion reaction” [1] using 3d metal oxides opened new materials as negative electrodes in the dominating classical material chemistry of LIBs. The typical conversion reaction between Li and the metal oxide is given in equation 1:
Discharging −n + xM 0 +++ yLi O ------(1) MxOy + 2ye + 2yLi 2 Charging
The mechanism involves the formation of Li 2O matrix over which nanocrystalline metal particles are embedded upon reduction. Highly reacting nature of metal nanoparticles helps in the partial decomposition of Li 2O when reverse polarization is applied [1, 7, 26-28]. As reported in the literature, porous metal oxides help better percolation of electrolyte [29] and further improve cycling stability due to the accommodation of a large volume change in the voids present in the structure [30]. In this work, we have explained the formation of the porous structure of Co 3O4 from SEM and TEM studies (Fig. 7 & 8) and this porous structure is composed of small polyhedral Co 3O4 particles that allow the facile transport of Li ion and electron.
Fig. 10 shows the cyclic voltammogram of electrodes made from porous Co 3O4 nanoparticles cycled at a scan rate of 0.5 mV.s-1 between 3.0 and 0.0 V. In the first cycle, the broad reduction peak observed at 0.64 V corresponds to multi step electrochemical reduction
(lithiation process) of cobalt ions because cobalt exists in two oxidation states in Co 3O4
(Co 2+ , Co 3+ ) and also partly due to electrolyte decomposition leads to the formation of solid
- 114 - electrolyte interface (SEI) [Organic solvent undergoes decomposition easily on the electrode surface during cycling which forms the SEI. SEI is electrically insulating but it is ionically conducting. This interphase prevents the decomposition of electrolyte on further cycling].
The anodic peak observed at 2.03 V was due to oxidation of cobalt. In the third cycle, the reduction peak shifted to 0.89 V and further cycling did not alter the peak intensity (current) and integral areas (charge). However, there was a considerable variation in the charge compare to first cycle and the same behavior was reflected as irreversible capacity loss in the charge-discharge studies. These changes were due to electrolyte decomposition which takes place only in the first cycle. Similar cyclic voltammograms were observed by Wu et al [30].
As discussed earlier, electrochemical reaction of Li with Co 3O4 differs from the classical Li insertion/de-insertion reaction in layered materials and the mechanism followed was the reduction and oxidation of Co which was accompanied with the formation and decomposition of Li 2O [1] as given in the equation 1.
-1 Fig.10 Cyclic voltammogram of porous Co 3O4 electrode at a scan rate of 0.5 mV s and at the temperature of 30 °°°C
- 115 - We have also examined the electrochemical behavior of the pure Co 3O4 as an anode material for LIBs. Fig. 11A shows the plot of specific capacity (mAhg -1) vs. voltage (V vs
Li +/Li) obtained for the charge-discharge profiles of the first, fifth and tenth cycles.
Fabricated coin cells were cycled at a rate of C/2 from 3 to 0 V. On discharging from OCP, well defined conversion plateaus at 1.13 V and followed at 0.79 V to 0.62 V and thereon slopes falls to 0.02 V. The discharge capacity of the first cycle was found to be as high as
1131 mAhg -1 (corresponding to x = 11.3). These are generally attributed to the conversion of
Co 3O4 to an intermediate-phase CoO (or Li xCo 3O4) and then to metallic Co, respectively [1,
7, 9]. The sloping region may be due to the formation of a solid electrolyte interface (SEI) which leads to an irreversible capacity loss [4]. On charging, a distinct plateau was seen at
1.95 V which corresponds to the formation of a less lithiated phase (Li nCo 3O4, n < x) with a delivering capacity of 838 mAhg -1 (corresponding to x = 8.4) and 74% coulombic efficiency.
Hence, an irreversible loss of 3.26 (x) Li + occurs during the first cycle. Theoretically, only 8
+ + Li ions should react with Co 3O4 (given in equ. 1) but in this work 11.3 Li ions participate in the conversion reaction. The observed higher intake of Li was due to the formation of SEI.
Irreversible capacity loss experienced in the first cycle was may be due to incomplete decomposition of Li 2O and difficult in dissolution of SEI [9].
- 116 -
Fig. 11 A) First, Fifth, Tenth discharge and charge cycle of porous Co 3O4 at C/2 rate. B)
Discharge capacity versus cycle number for porous Co 3O4 at C/2 rate
Fig.11B shows cycle life study of porous Co 3O4 electrodes at C/2 rate. Except in the first cycle, subsequent cycles delivered almost a constant discharge capacity of 850 mAh g -1
- 117 - and the coulombic efficiency observed was about 96%. This value was found to be very
-1 close to the theoretical capacity of Co 3O4 (890 mAh g ) and suggests that Co 3O4 derived from the Co 3[Co(CN) 6]2 complex shows promising anode material in the Li ion battery.
Fig. 12 Electrochemical impedence spectra of Co 3O4 electrode measured at open circuit potential for as fabricated coin cell and after every five cycles.
Electrochemical impedance spectroscopy (EIS) was used to understand the total electrochemical impedances of the coin cell. The characteristic EIS curves (Nyquist plot) of porous Co 3O4 nanoparticles were shown in fig. 12. In Nyquist plot, two well defined regions were observed i) small semicircle ii) inclined line. The diameter of the semicircle represents charge transfer resistance and intercept of real part axis at the high frequency region corresponds to the solution resistance. The inclined line at an approximately 45 ° angle to the
- 118 - real axis in the low frequency region corresponds to the mass transfer by diffusion process
[31-33]. The coin cell fabricated from porous Co 3O4 nanoparticles shows the high charge transfer resistance (Fig. 12). The diameter of the semi circle was reduced to the greater extend after five electrochemical cycles and it implies that the charge transfer resistance was decreased. This feature may be attributed as due to the formation of SEI on the electrode surface which helps in the facile diffusion for Li + ion and charge transfer process [32]. In further cycling, there is no considerable change in the diameter of the semi circle was observed (enlarged view in the inset of fig. 12) but slowly inclined line moves parallel to y axis (imaginary axis of Nyquist plot). This behavior may be due to pseudo capacitive character of the electrode [31].
5.4. Conclusions
In summary, we have shown a facile route for the synthesis of porous Co 3O4 from
Prussian blue analogue, Co 3[Co(CN) 6]2. This method of preparation of a nanostructured oxide was found to be novel route in the preparation of electrode materials for LIBs.
Co 3[Co(CN) 6]2 was prepared by simple precipitation from the solution of cobalt acetate and potassium hexacyanocobaltate. On thermal decomposition of Co 3[Co(CN) 6]2, formation of cobalt oxide takes place and hence the temperature has to be optimized for the synthesis of phase pure Co 3O4. During the thermal decomposition of Co 3[Co(CN) 6]2, a large change in mass and expulsion of gases take place resulting in the formation of a porous cobalt oxide as evident from SEM and TEM images. We have evaluated the performance of porous Co 3O4 as an anode material in LIBs. It showed good capacity value of 850 mAh g-1 which is very close
-1 to the theoretical capacity of Co 3O4 (890 mAh g ). In addition to these, this method allows
- 119 - one to synthesize different Prussian blue analogues which can be further decomposed to give corresponding mixed metal oxides.
References:
1. P. Poizot, S. Laruelle, S. Grugeon, L. Dupont, J-M. Tarascon, Nature 407 (2000) 496.
2. L. Hou, C. Yuan, L. Yang, L. Shen, F. Zhang, X. Zhang, RSC. Adv. 1 (2011) 1521.
3. X. Wang, A. Sumboja, E. Khoo, C. Yan, P. S. Lee, J. Phys. Chem. C 116 (2012)
4930.
4. W-Y. Li, L-N. Xu, J. Chen, Adv. Funct. Mater. 15 (2005) 851.
5. M. C-Cabanas, G. Binotto, D. Larcher, A. Lecup, V. Giordani, J.-M. Tarascon, Chem.
Mater. 21(2009) 1939.
6. T. Sugimoto, E. Matijevic, J. Inorg. Nucl. Chem. 1 (1979) 165.
7. D. Larcher, G. Sudant, J.-B. Leriche, Y. Chabre, J.-M.Tarascon, J. Electrochem. Soc.
149 (2002) A234 .
8. B. Wang, T. Zhu, H.B. Wu, R. Xu, J. S. Chen, X.W. D. Lou, Nanoscale 4 (2012)
2145.
9. S. Xiong, J. S. Chen, X. W. Lou, H. C. Zeng, Adv. Funct. Mater. 22 (2012) 861.
10. M. M. Rahman, J-Z. Wang, X-L. Deng, Y. Li, H-K. Liu, Electrochim. Acta 55 (2009)
504.
11. Y. Jiang, Y. Wu, B. Xie, Y. Xie, Y.Qian, Mater. Chem. Phys. 74 (2002) 234.
12. F. Cao, D. Wang, R. Deng, J. Tang, S. Song, Y. Lei, S. Wang, S. Su, X. Yang, H.
Zhang, Cryst. Eng. Comm 13 (2011) 2123.
- 120 - 13. X-L. Huang, X. Zhao, Z-L. Wang, L-M. Wang, X-B. Zhang, J. Mater. Chem. 22
(2012) 3764.
14. Y. Wang, H.J. Zhang, J. Wei, C.C. Wong, J. Lin, A. Borgna, Energy Environ. Sci.
4 (2011) 1845.
15. J. Ma, A. Manthiram, Rsc. Adv. 2 (2012) 3187 .
16. N. Du, H. Zhang, B. Chen, J. Wu, X. Ma, Z. Liu, Y. Zhang, D. Yang, X. Huang,
J. Tu, Adv. Mater. 19 (2007) 4505.
17. D. H. M. Buchold, C. Feldmann, Chem. Mater.19 (2007) 3376.
18. M. Vondrova, C. M. Burgess, A. B. Bocarsly, Chem. Mater. 19 (2007) 2203.
19. M. Vondrova, T. M. McQueen, C. M. Burgess, D. M. Ho, A. B. Bocarsly, J. Am.
Chem. Soc. 130 (2008) 5563.
20. M. Heibel, G. Kumar, C. Wyse, P. Bukovec, A. B. Bocarsly, Chem. Mater. 8 (1996)
1504.
21. C. Kappenstein, J. Cernak, R. Brahmi, D. Duprez, J. Chomic, Thermochim. Acta 279
(1996) 65.
22. Y. Guo, R. Weiss, M. Epple, Eur. J. Inorg. Chem. (2005) 3072.
23. R. Weiss, Y. Guo, S. Vukojevic, L. Khodeir, R. Boese, F. Schüth, M. Muhler, M.
Epple, Eur. J. Inorg. Chem. (2006) 1796.
24. G. W. Beall, D. F. Mullica, W. O. Milligan, Inorg. Chem. 19 (1980) 2876.
25. D. B. Brown, D. F. Shriver, Inorg. Chem 8 (1969) 37.
26. G. Binotto, D. Larcher, A. S. Prakash, R. Herrera Urbina, M. S. Hegde, J-M.
Tarascon, Chem. Mater. 19 (2007) 3032.
- 121 - 27. F. Badway, I. Plitz, S. Grugeon, S. Laruelle, M. Dolle, A. S. Gozdz, J.-M. Tarascon,
Electrochem. Solid-State Lett. 5 (2002) A115.
28. A. S. Prakash, P. Manikandan, K. Ramesha, M. Sathiya, J.-M. Tarascon, A. K.
Shukla, Chem. Mater. 22 (2010) 2857.
29. J. Liu, H.Xia, L. Lu, D. Xue, J. Mater. Chem. 20 (2010) 1506.
30. Z-S. Wu, W. Ren, L. Wen, L. Gao, J. Zhao, Z. Chen, G. Zhou, F. Li, H-M. Cheng,
ACS nano 4 (2010) 3187.
31. E. Barsoukov, J. R. Macdonald, Eds Impedance Spectroscopy: Theory, Experiment,
and Applications, Second Edition Wiley-Interscience (2005)
32. Y. Liu, X. Zhang, Electrochim. Acta 54 (2009) 4180.
33. J. Zhu, Y. K. Sharma, Z. Zeng, X. Zhang, M. Srinivasan, S. Mhaisalkar, H. Zhang, H.
H. Hng, Q. Yan, J. Phys. Chem. C 115 (2011) 8400.
- 122 - 6.1 Introduction
Ferric ferrocyanide (Fe 4[Fe(CN) 6]3) complex is known as Prussian blue (PB) which was the first synthetic co-ordination compound, having a history for more than three centuries. It is one of the most extensively studied co-ordination compound and still considered as ‘hot’ in research because of its interesting ionic, electronic and electrochemical properties [1-4]. PB and its analogues were prepared by both chemical and electrochemcial procedures. In the chemical synthesis, the PB analogues can be prepared by simple mixing of chlorometalate and cyanometalate ions in aqueous solutions. Control of size and shape of PB analogues is not possible by simple mixing. However, PB nanoparticles with controlled size and shape can be controlled by different synthetic procedures like addition of reactants under vigorous stirring [5], in presence of surfactants [6], microemulsion medium [7] etc., PB and its analogues were also formed as thin films by potential cycling of electrodes in a medium containing potassium ferricyanide and corresponding metal salts with potassium chloride or nitrate [8].
Recently, Bocarsly’s group have developed a method to synthesis ‘cyanogels’ which are inorganic polymer system that results from the polymerization of chlorometalate and cyanometalate ions in aqueous solutions [9]. They proposed a mechanism that two of the chloride ligands on chlorometalate were replaced by the nitrogen end of cyanide ligand on the cyanometalate creating the cyanide bridge which results in the formation of cyanide linkage between two metal centers and repeated units forms the cyanogel. The polymer initially forms as small sol particles (cyanosol) that undergo a sol-gel transition that ultimately forms the bulk hydrogel (cyanogel). The cyanogel consists of two separate phases, one is inorganic polymer phase and the second one is solvent (water) phase. This solvent
- 123 - phase can be removed via slow evaporation, leaving behind a highly porous xerogel. It has been shown that by introducing a large positively charged counter-ion will prevent the agglomeration of soluble cyanosol nanoparticles. Metal compositions of various ratios can be isolated with a narrow size distribution [10]. The cyanogels can be further processed to form homogenous alloys at much lower temperature than that of traditional synthesis and the main advantages are as follows.
i) Position of metal atoms in the complex gels was fixed and hence the resulting
alloy will be homogeneous in the nanoscale.
ii) Reduction of metal ion takes place in-situ during the decomposition of cyanogels
in inert atmosphere. The decomposition product was only cyanogen gas, escapes
leaving behind the metal.
iii) Cyanogel compositions can be varied by changing the ratio of metal complexes in
the gelation mixture and hence the metal composition in resulting bimetallic alloy
also will be varied.
iv) Cyanogels can be converted into either alloys or mixed oxides, depending on the
heat treatment ambience.
As described above, either PB or cyanogel can be formed while mixing chlorometalate and cyanometalate ions in aqueous solutions. The salient differences between
PB and Cyanogel were tabulated in Table-1. Hence the investigations on interaction of chlorometalate and cyanometalate will decipher the pathway of reaction mechanism. We have chosen three systems for a detailed study in what follows
- 124 - System – I HAuCl 4 and K 3[Fe(CN) 6]/ K 4[Fe(CN) 6]
System – II H2PtCl 6 and K 3[Ru(CN) 6]
System – III PdCl 2 and K 3[Fe(CN) 6]/ K 4[Fe(CN) 6]
Table-1- Differences in Prussian blue and cyanogels
Prussian blue Cyanogel Crystalline [11] Amorphous [12] On mixing chlorometalate and On mixing chlorometalate and cyanometalate, precipitation of Prussian blue cyanometalate, gelation takes place. takes place. Heating in inert atmosphere leads to the Heating in inert atmosphere leads to formation of metal carbides [13]. the formation of alloys [12].
6.1.1 System – I: HAuCl 4 and K 3[Fe(CN) 6]/ K 4[Fe(CN) 6]
Electrochemical synthesis and characterization of various PB analogues of Co(II),
Ni(II), Zn(II), Cd (II), Cu(II), Ag(I) etc., were studied in detail by several research groups [2,
3, 14-16]. In group IB of periodic table, copper, silver and gold elements are present. Copper and silver were reported to form PB analogues both chemically and electrochemically [17,
18] but, there is no report found for the PB analogue of gold in the literature.
Senthil et al [19] showed that the electrochemical cycling in medium containing potassium ferricyanide and gold chloride leads to the formation of Au-PB nanocomposite and the mechanism was studied in great detail. In brief, during electrochemical cycling, the first step was found to be the electro deposition of gold on the GC electrode surface which aids the decomposition of ferricyanide ion to free ferric ion and further leads to the formation of
PB on gold nuceli. Hence, this study confirms that PB analogue of Au could not be formed on glassy carbon (GC) electrode by electrochemical cycling in a potassium nitrate medium
- 125 - containing potassium ferricyanide and gold chloride. Au nanoparticles were used as catalyst in the electron transfer reaction between potassium ferrocyanide and thiosulphate [20] or borohydride [21]. However to the best of our knowledge, interaction between gold (III) chloride and potassium ferrocyanide or ferricyanide has not been studied in detail.
In this work, we have adopted the similar experimental procedure of Liu et al [22].
They reported the formation of Pt analogue on GC electrode by reacting potassium ferricyanide with chloroplatinic acid. Pt analogue formed on GC was characterized by the reversible surface redox process occuring at 0.77 V (vs saturated calomel electrode) due to redox transition of low spin Fe (II/III). Similarly, we have studied the interaction between gold (III) chloride and potassium ferrocyanide or ferricyanide in the absence of electrochemical cycling. We present here, the possible mechanistic pathways of the chemical reaction between gold (III) chloride and potassium ferrocyanide or ferricyanide in detail. The two pathways of reaction were deciphered at using Ultraviolet–visible (UV-Vis), X-ray photo electron spectroscopy (XPS) and cyclic voltammetry.
6.1.2 H2PtCl 6 & K3[Ru(CN) 6] and PdCl 2 & K3 [Fe (CN) 6]/ K 4[Fe(CN) 6]
Formation of cyanogel was first reported by Bocarsly’s group using PdCl 2 and
K3[Fe(CN) 6] [9]. They have extensively studied these cyanogel systems utilizing a variety of metals (Pd, Pt, Fe, Co, Ru, Ir) and it has been shown that heating the cyanogel under an inert atmosphere causes auto-reduction of the metal centers, releasing hydrogen cyanide and cyanogen, leaving behind metal alloys [23]. The alloy formation via cyanogel procedure has several advantages (listed in section 6.1) over other conventional methods like impregnation, polyol etc. The main disadvantages with other methods are
- 126 - i) Alloy formation needs very high thermal activation.
ii) Complete removal of by-products and stabilizer is difficult.
iii) Heterogeneity in the alloy composition.
Cyanogel method of synthesis can also be tailored to form supported alloys. In order to study the magnetic property, it is desirable to limit the magnetic interactions between the particles and hence dispersing them in a non-magnetic matrix is essential. Bocarsly et al have synthesized PdCo alloy supported on silica by heating a two phase co-gel material of silica geland PdCo cyanogel. PdCo alloy supported on silica exhibited the super paramagnetic transition between 50 and 75 K [24]. Support materials like silica, carbon were used to isolate the small amounts of the metal precursors from each other and after reduction interactions of metal nanoparticles with the support will prevent the agglomeration of nanoparticles.
Even though several combinations of alloys [12] can be synthesized using the cyanogel method, their electrochemical characterization and application have not yet been explored. In order to study the electrochemical applications, Cyanogel method has to be suitably modified for the synthesis of carbon supported alloys. In this work, we have attempted to study the formation of cyanogels by electrochemical methods and followed by preparation and characterization of carbon supported alloys derived from cyanogel route which is hitherto unexplored in literature.
- 127 - 6.2 Experimental procedures
6.2.1. Material and methods:
HAuCl 4 (Sigma-Aldrich), K 3[Fe(CN) 6] (E-Merck), K 4[Fe(CN) 6] (E-Merck), NaBH 4
(Sigma-Aldrich), H 2PtCl 6 (Sigma-Aldrich), K 4[Ru(CN) 6] (Sigma-Aldrich), PdCl 2 (Sigma-
Aldrich), acetone (E-Merck), KNO 3 (E-Merck), H 2SO 4 (E-Merck), Nafion (Sigma-Aldrich).
The aqueous solutions were freshly prepared every time using Milli-Q water (18.2 M cm)
(Millipore).
UV–Vis absorption spectra were collected using Varian Cary 500 scan UV–Vis spectrophotometer with incident light normal to the 1 cm path length quartz cell. XPS was done by using the Multilab 2000 (Thermoscientific, UK) photoelectron spectrometer fitted with a twin anode X-ray source. The Au 4f core-level photoemission spectra were recorded using the Mg Kα (1253.6 eV) source. Deconvolution of XPS peaks was done using XPSpeak
4.1 software. Thermo-Nexus 670 model spectrometer was used for Fourier transform infra red (FT-IR) measurements. FT-IR spectra were recorded using attenuated total reflectance
(ATR) attachment for thin film and using KBr pellet method for powder samples. In the electrochemical experiments, a Glassy carbon (GC) electrode (diameter of 3 mm, BAS, Inc.) and a platinum foil were used as the working and auxiliary electrodes respectively, in a standard 3-electrode configuration. All the potential values were reported against the
Ag/AgCl reference electrode unless otherwise mentioned. GC electrode was polished using
4/0 emery sheet and sonicated in Milli Q water for 5 minutes before each experiment. Cyclic voltammetric experiments were done on a BAS-100B electrochemical system. Thermal analysis was performed through SDT Q600 V8.3 Build 101 (Universal V4.3A TA
Instruments) at a heating rate of 10 °C per minute under Nitrogen atmosphere. XRD patterns
- 128 - were recorded using the PANalytical diffractometer Model PW3040/60 X’pert PRO operating with x-ray source Cu K α radiation (k = 0.15406 nm) generated at 40 kV and 20 mA. TEM examination was made by placing a drop of the sample (dispersed in acetone) onto a copper grid coated with carbon film (400 meshes) using Philips CM200 microscope working at 200 kV.
6.2.2. System – I: Synthesis and characterization of KFe x[Au(CN) 2]y
The synthesis of KFe x[Au(CN) 2]y was carried out by simple mixing of HAuCl 4 and
K4[Fe(CN) 6] or K3[Fe(CN) 6]. While mixing HAuCl 4 and K3[Fe(CN) 6] there was no change in color of the solution or precipitation was not observed. On mixing HAuCl 4 and
K4[Fe(CN) 6], immediately a color change was observed but there is no precipitation. Hence,
UV-Vis analysis alone was carried out using mixture of solutions and other characterizations were carried by adopting the experimental procedure reported by Liu et al [22]. Special care was taken for each analysis and hence details of the sample preparation in each analysis were described in the following. For cyclic voltammetry analysis, 2.5 µL of 10 mM HAuCl 4 and
2.5 µL of 10 mM K4[Fe(CN) 6]/K 3 [Fe(CN) 6] were placed on a GC electrode and on drying the mixture of the reactants forms the charge transfer complex. The reactant mixture was allowed to dryness on the surface of the GC electrode for about 10 hours and then washed with Millipore water to remove unreacted reactants, if any. Here, 5 µL of 0.1 M NaBH 4 was dropped on the formed thin film on the GC electrode for complete reduction of gold ions in the complex. FT-IR spectra was recorded by forming a thin film of charge transfer complex by drop casting a mixture of 10 mM of HAuCl 4 and 10 mM of K 4 [Fe(CN) 6] on 1 cm×1 cm non-conducting glass slide. XPS was recorded before and after the reduction of
- 129 - KFe x[Au(CN) 2]y. Before reduction, sample preparation was similar to that of FT-IR analysis and gold ions in the complex were reduced using 10 µL of 0.1 M NaBH 4.
6.2.3 System – II: Synthesis and characterization of Pt-Ru cyanogel and PtRu/C
Pt-Ru cyanogel was synthesised by heating the mixture of of H2PtCl 6 and
o K4[Ru(CN) 6] solutions in the appropriate ratios at ~75 C for about 1 hour. Cyanogel formation was observed by change in the color of the solution from yellow to green and then cyanogel was separated from the solution by centrifugation. Cyanogel was washed with water, acetone and dried in vacuum oven at 50 oC for 24 hours. To synthesis carbon supported alloy catalyst, Vulcan carbon (Cabot) was added to the gelling mixture prior to heating. The dried cyanogels with carbon were heated at 1000 oC for 3 hours in a tubular furnace under Ar atmosphere to form carbon supported PtRu 40% (by weight) loadings of
Pt 70 Ru 30 and Pt 50 Ru 50 on carbon were prepared by taking appropriate amount of H2PtCl 6 and
K4 Ru(CN) 6 in the gelling mixture. GC electrode was modified by casting 3 µL of catalyst ink (Prepared by mixing the 2.5 mg of the PtRu/C in 125 µL of 5% Nafion, 625 µL of water) and electrode was dried in air.
6.2.4. System – III: Synthesis and characterization of Pd-Fe cyanogel and PdFe/C
Pd-Fe cyanogel was formed by mixing PdCl 2 and K4[Fe(CN) 6]/K 3 [Fe(CN) 6]. The cyanogel formation takes place immediately which was observed by immediate color change.
Cyanogel was separated by centrifugation, washed with water followed by acetone and dried in vacuum oven at 50 oC for 24 hours. To synthesis carbon supported Pd-Fe (PdFe/C) catalyst, Vulcan carbon (Cabot) was added to the gelling mixture. The dried cyanogels with
- 130 - carbon were heated at 900 oC for 3 hours to form PdFe/C 40% (by weight) loadings of PdFe,
Pd 2Fe, Pd 3Fe on carbon were prepared by taking appropriate amount of PdCl 2 and
K4[Fe(CN) 6] in the gelling mixture. GC electrode was modified by casting 3 µL of catalyst ink (Prepared by mixing the 2.5 mg of the PdFe/C in 125 µL of 5% Nafion, 625 µL of water) and drying in air for 30 minutes.
6.3 Results and Discussions
6.3.1. System – I:
Investigation on interaction between HAuCl 4 and K 3[Fe(CN) 6]/ K 4[Fe(CN) 6]:
In this section, we have investigated the interaction between the HAuCl 4 and
K4[Fe(CN) 6]/K 3[Fe(CN) 6]. When HAuCl 4 and K3[Fe(CN) 6] solutions were mixed together , there was no considerable change in color of the solution but there was instantaneous color change when HAuCl 4 and K4[Fe(CN) 6] were mixed each other. Hence, we investigated the reason for different reaction mechanism and analyzed using UV-Vis, XPS and cyclic voltammetry techniques.
6.3.1.1 UV-Visible and FT-IR spectral studies
The electronic spectra of 1 mM gold chloride, potassium ferrocyanide, potassium ferricyanide, a mixture of potassium ferricyanide and gold chloride were shown in
Fig. 1. The spectrum recorded in a mixture of potassium ferrocyanide and gold chloride solution was given in the inset of Fig. 1 for comparison. The absorption wavelength ( λmax ) values for the solutions were tabulated in Table-1.
- 131 -
Fig.1 UV-Vis spectra of 1 mM each of HAuCl 4, K 4[Fe(CN) 6], K 3[Fe(CN) 6] and
K3[Fe(CN) 6] + HAuCl 4. Inset: K 4 [Fe (CN) 6] + HAuCl 4.
Table -1 Observed λmax value for each reactant from UV-Vis spectra.
-3 -3 Concentration of Reactants/ 1x 10 mol dm λmax / nm
HAuCl 4 306
K4[Fe(CN) 6] 332
K3[Fe(CN) 6] 260, 303, 325, 421
HAuCl 4 + K 3[Fe(CN) 6] 306, 414
HAuCl 4 + K 4[Fe(CN) 6] 305, 414, 692
It is clear from Fig. 1 and Table-1 that there is no significant change in the absorption wavelength when potassium ferr icyanide and gold chloride solutions were mixed together.
- 132 - Even after keeping the mixture for a few hours, the solution did not show any visible color change indicating the absence of interaction between two anionic reactants. This observation is consistent with the results of Zhai et al [25] who observed no change in the absorption band after the addition of ferricyanide ions to gold chloride solution, indicating that there was no bond formed between gold chloride and ferricyanide. Freund et al [20] have studied the redox reaction between ferricyanide and thiosulfate catalysed by colloidal gold whereas
Romero et al [21] have demonstrated that the gold nanoparticles can act as an efficient catalyst in the reduction of ferricyanide ion to ferrocyanide ion by sodium borohydride.
Fig.2. FT-IR spectra of Au-Fe complex formed by 10 mM of HAuCl 4 and 10 mM of
K4 [Fe(CN) 6] on 1 cm×1 cm non-conducting glass slide.
Interestingly, in contrast to ferricyanide, we could observe a redox reaction with intense green color formation when solutions of potassium ferrocyanide and gold chloride
- 133 - were mixed in equi-molar ratio. The UV-vis spectra (Fig .1) were marked by the absence of absorption peak corresponding to ferrocyanide ion. Instead, we have noticed absorption peak corresponding to ferricyanide ion at 305 and 414 nm. When the oxidation state of Fe changes from (II) to (III), it is natural to expect concomitant changes in the oxidation state of Au in gold chloride but there was no absorption corresponding to the characteristic surface plasmon resonance band for Au nano . Our attempts to reduce gold chloride in the presence of a stabiliser like sodium dodecylsulfate also did not indicate the formation of Au nano . These observations indicate that Au 3+ in gold chloride may undergo reduction to Au +. Incidentially, the formation of Au + from Au 3+ on the addition of potassium ferrocyanide was known from
Vrublevskaya et al [26]. In additon, a charge transfer band was also observed at 690 nm which is common in PB-like compounds [27]. Fig.2 shows the FT-IR spectrum of Au-Fe complex, stretching vibration at 2168 cm -1 confirms the presence of cyanide ligand. Cyanide stretching vibration shifts to higher wave numbers, when compared to the that in KAu(CN) 2
(2140 cm -1). A similar shift to longer wave number compared to free cyano complex was also observed in PB type compounds [28].
6.3.1.2 XPS studies
We have formed films of the charge transfer complex on a plain glass plate by drop casting a mixture of 10 mM potassium ferrocyanide and gold chloride and subsequent drying as described in the experimental section 6.2.2. XPS of this film showed Au 4f 5/2 and 4f 7/2
+ – peaks at 85.92, 89.50, 93.07 eV (Fig. 3A) which corresponds to Au in [Au(CN) 2] [29-33].
Relative shift in the binding energy values of Au + with reference to those of Au 0 was found to vary between 1.2-2.3 eV [31] and the similar shifts were found by others and attributed
- 134 - either as due to the substrate interaction or to the influence of the chemical environment [29-
34]. In a control experiment, the complex was reduced using NaBH 4 and analysed by XPS. It showed the peak shifts to lower binding energy values of 84.17 and 87.84 eV corresponding to metallic gold Au 0 [35] (Fig.3B)
Fig.3. XPS of Au-Fe Complex A) before and B) after NaBH 4 reduction (XPS peaks were deconvulated using xpspeak 4.1 and solid lines indicate the deconvulated peak).
6.3.1.3. Cyclic voltammetric studies
To characterize the interaction between potassium ferrocyanide and gold chloride, we have followed an unconventional approach of forming film on GC by drop casting the mixture as described in the experimental section. This electrode showed a voltammetric response (Fig. 4A) that clearly exhibits two redox processes; one at 0.18 V and the other at
0.63 V. The peak currents of both the processes decrease with increasing cycle number.
However we have obtained a stable and reversible redox peak at 0.15 V when the range of potential cycling is limited to 0.0 V to 0.6 V (Fig. 4B). Similar improvement in the stability was earlier achieved in the case of silver hexacyanoferrate modified electrode when the potential cycling was limited to 0.9 V [36]. The redox peak obeys characteristics of an ideal
- 135 - surface reaction as seen from the linear relation between the plot of peak current and scan rate [inset in Fig. 4B].
Fig.4. A) Cyclic voltammogram of Au-Fe complex during cycling in 0.1 M KNO 3 at the
-1 scan rate of 50 mV s ; B) Cyclic voltammogram of Au-Fe complex in 0.1 M KNO 3 at the scan rate of 20 mV s -1. Inset Fig. Plot of peak current vs. scan rate; C) Cyclic voltammogram of Au-Fe complex after reduction with NaBH 4 in 0.5 M H 2SO 4 at the scan rate of 50 mV s -1.
Gold ions in Au-Fe complex (formed on GC electrode) was reduced using NaBH 4 (as followed in XPS analysis) to form metallic gold. NaBH 4-treated electrode showed voltammetric peaks characteristic of gold oxide formation at 750 mV and gold oxide reduction at 430 mV versus Hg/Hg 2SO 4 reference electrode in 0.5 M H 2SO 4 (Fig. 4C).
- 136 - A similar modification of GC electrode surface by drop casting a mixture of potassium ferricyanide and gold chloride results in the formation of PB on the surface of electrode as seen from the two sets of redox process at 0.16 V and 0.837 V (Fig. 5). Both the redox couples remain stable on potential cycling. The origin of these redox processes in the case of films formed on the GC surface from mixtures of gold chloride with ferricyanide/ferrocyanide is discussed below.
Fig.5. Cyclic voltammogram of thin film formed on the GC electrode by drop casting of
2.5 µµµL of 10 mM of K 3[Fe(CN) 6] and 2.5 µµµL of 10mM HAuCl 4 in 0.1 M KNO 3 at the scan rate of 50 mV s -1.
6.3.1.4. Discussion on the interaction of HAuCl 4 and K 4[Fe(CN) 6]/ K3[Fe(CN) 6]
Various PB analogues can be electrochemically prepared as thin films on conducting substrates (like GC, Indium tin oxide, Pt) by potential cycling the electrode in a medium
- 137 - containing the metal salt and potassium ferricyanide. However, there is no report on the formation of gold hexacyanoferrate by either electrochemical or chemical means.
Electrochemical cycling of the GC electrode surface in a solution containing gold chloride and potassium ferricyanide in KNO 3 solution lead to the formation of Au-PB nanocomposite
[19]. Our present experiments suggest that on similar cycling of GC electrode in a solution containing potassium ferrocyanide and gold chloride in KNO 3 solution also resulted in the formation Au @ PB nanocomposite (Fig. 6).
Fig.6 Cyclic voltammogram of gold-Prussian blue nanocomposite formed from 1mM
-1 K3[Fe(CN) 6] and 1 mM of HAuCl 4 in 0.1 M KNO 3 at the scan rate of 50 mV s .
The UV-visible spectra [Fig. 1] do not present any features characteristic of a reaction between ferricyanide and gold chloride. However, when the same mixture is drop cast on
GC and dried to form a film, the modified electrode showed two sets of redox peaks; one at
0.136 V and the other at 0.837 V typical of PB formation. It can now be stated that on
- 138 - mixing gold (III) chloride and potassium hexacyanoferrate (II/III) on GC, the following processes take place:
Step 1: increase in acidity due to an increase in [H +] during solvent evaporation (the initial pH of the mixture of 1mM gold chloride and potassium ferricyanide or ferrocyanide was ~3.2);
Step 2: decomposition of potassium ferricyanide ion to free ferric ion under low pH conditions [37] [Similar decomposition of potassium ferricyanide to ferric ion during gold hydroxide formation caused by the interfacial acidity was reported by Senthil et al [19];
Step 3: formation of PB takes place during electrochemical cycling through a complex formation between ferricyanide ion with free ferric ion.
On the other hand, a redox reaction takes place when potassium ferrocyanide was mixed with gold chloride. The standard oxidation potential of potassium ferrocyanide is
– – –0.37 V [38] and the reduction potential of [AuCl 4] to [AuCl 2] is 0.926 V [39]. Hence the reaction between the potassium ferrocyanide and gold chloride was spontaneous whereas this was not in the case of potassium ferricyanide. Spontaneous reaction between potassium ferrocyanide and gold chloride was observed in the UV-visible spectra showing the conversion of ferrocyanide ion to ferricyanide ion and Au3+ , in turn undergoing reduction to either Au + or Au 0. Although the oxidation potential of ferrocyanide appear to be sufficient to
3+ 0 3+ 0 0 reduce Au to Au (E Au / Au =1.002V), we have not observed Au formation from the UV-
Vis spectral and XPS analysis. The difference between the reaction of gold chloride with potassium ferricyanide or ferrocyanide was due to the associated redox reaction in the later
- 139 - case. It can be summarized that the following steps are involved in the reaction between the potassium ferrocyanide and gold chloride
Step 1: redox reaction between potassium ferrocyanide and gold chloride leading to the formation of ferricyanide ion and Au +.
Step 2: concomitant decomposition of ferrocyanide / ferricyanide ion to free ferrous/ferric ion at high acidity conditions.
The XPS spectra of the film prepared by drop casting the mixture on the plain glass
– plate indicate the formation of [Au(CN) 2] . Au (I), Ag (I) and Cu (I) were known to form stable cyano-complexes [37] in which Au has more affinity towards CN – ion and hence the extraction of Au was perfomed by cyanidation, known as ‘MacArthur-Forrest process’ [40].
– In addition to this, the formation constant of [Au(CN) 2] in the cyanide medium was also very high (ca.10 38 ) compared to other metals [37]. Hence, the gold ions might extract ligands from the potassium ferrocyanide to form the cyano-complex of gold. Similar ligand exchange isomerism between iron and chromium in chromium hexacynanoferrate during electrochemical cycling resulted in the formation of iron hexacyanochromate as reported by
Dostal et al [41]. To explain the stable voltammetric response shown in fig.4B, we propose
3+ 2+ – the formation of a complex between free Fe / Fe and [Au(CN) 2] . Similar dicyanoaurate complex with metal ion/complexes were sparsely reported [42-46]. Dong et al [46]
– synthesized {Mn[Au(CN) 2]2(H 2O) 2}n and {KFe[Au(CN) 2]3}n from [Au(CN) 2] . The voltammetric response of the GC modified with 1:1 molar ratio of gold chloride and potassium ferrocyanide showed a stable redox couple at 0.17 V. The redox peak potential matches with the high spin Fe 2+/3+ transitions in the complex. Based on our results and those
- 140 - of Dong et al [46], we propose that the probable complex formed was of the general formula
KFe x[Au(CN) 2]y. As expected from this formula, the cyanide stretching vibration in FT-IR spectra was comparable with that of Dong et al [46], modified electrode was unstable when potential exceeds 0.6 V and the redox peak depends on K + ion concentration in the electrolyte.
The cyanide stretching vibration observed from the FT-IR spectra was 2168 cm -1 and the reported value for the same type of complex was 2154 cm -1 [46]. This small variation observed in the wave number may be due to the difference in the stoichiometry of the complex. The instability of the cyclic voltammetric response when the potential cycling is
– – extended beyond 0.6V is probably due to the conversion of [Au(CN) 2] / [Au(CN) 4] couple that occurs between 0.53 and 0.59 V [47], requiring additional cyanide ligands from the medium. The absence of the cyanide ligand in the supporting electrolyte 0.1 M KNO 3 might cause destabilization of the film during cycling.
The K + ion dependence was known for various PB analogues and the voltammetric mid-peak potential would decrease with decreasing potassium ion concentration for both the oxidation and reduction reactions [2] . The K + ion dependence of the formed complex was shown in Fig. 7 where redox peaks shifts in the positive direction when K + ion concentration increases from 0.01 M to 1M and the mid-peak potential shows a near-Nernstian shift.
- 141 -
Fig.7 Cyclic voltammograms of K+ ion dependence of Au-Fe complex at the scan rate of
10 mV s -1
6.3.2 System – II: Investigation on interaction between H 2PtCl 6 and K4 [Ru(CN) 6]
In this section we have investigated the interaction between H 2PtCl 6 and
K4 [Ru(CN) 6]. When H 2PtCl 6 and K 4 [Ru(CN) 6] were mixed each other, there was no color change but when the same solution mixture was heated in a water bath for 75 °C the color of solution turns from yellow to green (shown in photograph.1). Further, viscosity of the solution was increased on heating and remained stable even after cooling the gelation mixture to room temperature. We have studied the Pt-Ru cyanogel using cyclic voltammetry and UV-Vis techniques. Thermo gravimetric analysis was carried out on dried cyanogel to identify the decomposition temperature of cyanide ligands. We have also attempted to synthesis carbon supported PtRu (PtRu/C) via cyanogel route and we characterized using
XRD and cyclic voltammetry studies.
- 142 - 75 0 C
Photograph 1: PtRu cyanogel formation observed by the color change from yellow to green at 75 °°°C
6.3.2.1 Cyclic voltammetric studies on Pt-Ru cyanogel formation
Cyclic voltammetry was used as a tool to study the formation of Pt-Ru cyanogel. The
Fig 8 A depicts the voltammogram of 0.5 mM K 4 [Ru(CN) 6] in 0.1 M KNO 3 which shows a reversible redox couple at 0.65 V. The reversible redox reaction is given below
3– –n 4– [Ru(CN) 6] + e [Ru(CN) 6] ------(1)
Bocarsly et al have reported the formation of cyanogels on mixing chloro and cyano complex of transition metals [9] and they synthesized a number of cyanogels with metal centers of Pd-Co, Pd-Fe, Pt-Ru etc., For e.g. In the case of Pd-Co complexes, the gelation take place instantaneously on mixing the corresponding chloro and cyano complex of Pd and
Co respectively [12, 23]. In contrast, while mixing H2PtCl 6 and K 4[Ru(CN) 6] solutions gel formation was not observed under normal experimental conditions. This feature was observed with no considerable change in the redox peak of K4[Ru(CN) 6] even after the
o addition of H 2PtCl 6 (Fig. 8A) However, when the above complexes are heated to 75 C, the mixture shows a colour change from yellow to green. The associated increase in viscosity also indicated the formation of cyanogels. We have monitored the change in the cyclic voltammogram on heating the mixture of H2PtCl 6 and K 4[Ru(CN) 6] (Fig. 8B).
- 143 -
Fig.8 A) Cyclic voltammogram of 0.5 mM K 4 [Ru(CN) 6] and 0.5 mM K 4[Ru(CN) 6] +
X mM (X=0.25, 0.5, 0.75, 1 mM) H 2PtCl 6 in 0.1 M KNO 3 at the scan rate of 50mV/s
(before heating). B) Cyclic voltammogram of (a) 0.5 mM K 4[Ru(CN) 6], (b) 0.25 mM
H2PtCl 6 + 0.5 mM K 4[Ru(CN) 6], (c) 0.5 mM H 2PtCl 6 + 0.5 mM K 4[Ru(CN) 6], (d)
0.75 mM H 2PtCl 6 + 0.5 mM K 4[Ru(CN) 6], (e) 1 mM H 2PtCl 6 + 0.5 mM K 4[Ru(CN) 6] in
o 0.1 M KNO 3 at the scan rate of 50mV/s (after heating the samples at 80 C for 2 hours).
These experiments lead to following conclusions.
i) The redox peak current of K4[Ru(CN) 6] decreases with increasing amount of
H2PtCl 6 in the bath. The decrease in peak current observed was due to the
formation of PtRu cyanogel. This result confirms the atomic ratio of Pt and Ru
can be tuned with the respective amount of reactants.
ii) Current increase beyond 700 mV may be probably due to the electro-oxidation of
the adsorbed species on the electrode surface. (Fig. 8B)
iii) The mixture became completely electro inactive, when the molar ratio of
K4[Ru(CN) 6] and H2PtCl 6 was 1:2 .
- 144 - Cl
N Pt Cl N 2 4 N - - C Cl Cl CN C C Cl Cl NC CN Ru 2 Pt + Ru C C Cl Cl Cl CN NC C N Cl CN N Pt N Cl Cl
Scheme-1: Schematic representation for the formation of Pt-Ru cyanogel
The gelation is due to the formation of cyanide bridging between Pt (IV) and Ru (II) with the elimination of halide ions. The formation of Pt-Ru gels are schematically represented in the scheme-1.
6.3.2.2 UV-Visible studies on Pt-Ru cyanogel formation
Fig.9 shows the UV-Vis spectra of Pt-Ru cyanogel formation at fixed concentration of K4[Ru(CN) 6] (6 mM) on addition of various concentrations of H2PtCl 6 (2 mM-10 mM).
The broad absorption (at 1000 nm) in the near infrared region observed was attributed as due to the formation of inter valance charge transfer (IVCT) complex between Pt (IV) and Ru
(II). This type of IVCT absorption band was well known in the Prussian blue class of compounds [27]. The intensity of this absorption band in the NIR region increases with increasing content of H 2PtCl 6 in the gelling mixture. From the UV-Vis spectra, it was clear that the IVCT band show maximum intensity when the Pt (IV) and Ru (II) complex ratio was
2:1.
- 145 -
Fig.9 UV-Vis spectra of PtRu cyanogel with different compositions a) 2 mM and 6 mM b) 4 mM and 6 mM c) 6 mM and 6 mM d) 8 mM and 6 mM e) 10 mM and 6 mM of
H2PtCl 6 and K 4[Ru(CN)] 6 respectively
6.3.2.3 Thermogravimetric analysis
After the preparation Pt-Ru cyanogel, the cyanogel was washed thrice with millipore water and then twice with acetone. Now, the sample was slowly dried in vacuum at 50 °C for
24 hours. The dried sample was termed as Pt-Ru xerogel. Fig. 10 shows thermogram of Pt-
Ru xerogel done in nitrogen atmosphere. Pt-Ru xerogel decomposition started around 327 °C followed by a sudden fall at 400 °C signifying that all cyanide ligands were decomposed in this temperature range. The decomposition was completed around 600 °C and no more appreciable weight change was observed till 1000 °C. We report that loss of cyanide ligand
- 146 - from cyanogel takes place as cyanogens gas and similar observation was reported in the literature [12].
Fig. 10 TGA of Pt-Ru Xerogel in N 2 atmosphere
6.3.2.3 Fourier Transform infrared studies
Fig. 11 show the FT-IR spectra of PtRu xerogel, presence of characteristic -CN stretching vibration was found at 2080 cm -1 and it was observed at a higher wave number compare to the vibration of terminal cyanide stretching which found in K4[Ru(CN) 6] at 2048 cm -1. The shift to higher wave number was due to the formation of cyanide network in the gel
[12]. The complete removal of cyanide on heating was also confirmed by the absence of cyanide stretching vibrations in the FT-IR spectra of the Pt-Ru alloy.
- 147 -
Fig.11 FT-IR spectra of K 4 [Ru(CN) 6], Pt-Ru Xerogel and Pt-Ru alloy
6.3.2.4 X-ray diffraction studies
Carbon supported PtRu (PtRu/C) catalysts were synthesized by heating carbon supported Pt-Ru xerogel for 3 hours at 1000 °C under Argon atmosphere. The complete removal of cyanide on heating was confirmed by the absence of cyanide stretching vibrations in the FT-IR spectra shown in fig. 11. XRD in fig. 12, show the shift in the ‘2 θ’ to higher values clearly indicating the formation of an alloy with contraction in fcc phase. The lattice parameters were found to be 0.3864 and 0.3850 nm for Pt 70 Ru 30 /C and Pt 50Ru 50/C respectively which was less compared to bulk Pt (ca.0.3923 nm). A separate peak for the phase-separated Ru exists at 44 o either in the form of native metal or electron- or as proton- conducting hydrous oxide (RuO xHy or RuO 2.xH 2O) in both PtRu/C alloys was observed.
Presence of the Ru peak will play a vital role in the oxidation of methanol [48]. Similar phase separated Ru peaks were also reported earlier in the commercial samples of PtRu/C. [49, 50]
- 148 -
Fig.12 XRD pattern of Pt 70 Ru 30 /C and Pt 50 Ru 50 /C in comparison with bulk Pt. PtRu phase (*) and Ru phase (#)
6.3.2.5 Transmission electron microscopy studies
The alloy particles were homogeneously distributed on carbon matrix without any agglomeration as shown in the TEM pictures of Pt 70 -Ru 30 /C alloy at lower and higher magnifications (fig. 13A) and the average particle size was found to be approx. 7 nm as calculated from the histogram fig. 13B. Therefore, this method offers a direct route for the synthesis of carbon supported bimetallic nano-alloys that will find extensive electrocatalytic applications.
- 149 - A B
Fig. 13 A) TEM image of Pt 70 Ru 30 /C alloy (Scale Bar 50nm). Inset Fig with higher magnification (Scale Bar 20nm), B) Histogram of Pt 70 Ru 30 /C alloy
6.3.2.6 Cyclic voltammetric studies of Pt 70Ru 30/C and Pt 50Ru 50/C
The modification of the GC with the carbon supported catalysts was described in the experimental section 6.2.3. Voltammetric behaviour of Pt 70 Ru 30 /C and Pt 50 Ru 50 /C alloys show that the hydrogen adsorption-desorption region was prominent for Pt 70 -Ru 30 /C compared with that of Pt 50 Ru 50 /C alloy (fig.14A). This observation clearly indicates that the composition of PtRu/C can be varied by changing the composition of the cyanogel. In our studies, Pt 70 Ru 30 /C was found to be more catalytic than Pt 50 Ru50 /C towards electro-oxidation of methanol and the possible reasons of this behavior was discussed below. Even though good methanol oxidation was reported for Pt 50 Ru 50 /C by some groups [49, 50], many studies show that a higher content of Pt than Ru [51-53] was required to exhibit better methanol electro-oxidation. It was also reported that Pt 50 Ru 50 /C was more active only at temperature above 60 oC while at temperatures below 60 oC, a higher Pt content is required for increased electrocatalytic activity [54-55]
- 150 -
Fig.14 A) Cyclic voltammogram of Pt 70 Ru 30 /C and Pt 50 Ru 50 /C in 0.5M H 2SO 4 at the scan rate of 50mv/s B)Cyclic voltammogram of 0.5M methanol oxidation at Pt/C and
Pt 70 Ru 30 /C in 0.5 M H 2SO 4 at the scan rate of 50mV/s
Fig. 14B shows the comparison of methanol oxidation on Pt/C (E-tek), and Pt 70 -
Ru 30 /C which clearly indicates that the oxidation of methanol on Pt 70 Ru 30 /C alloy takes place at 0.75 V vs RHE, 100 mV less anodic than that of Pt/C. It was well known that Pt/C was not a good catalyst for methanol oxidation due to more CO poisoning on the Pt surface. This is evidenced in fig. 14B from the CO stripping peak during the electro-oxidation of methanol on Pt/C. Interestingly, methanol oxidation on Pt 70 Ru 30 /C does not suffer much on poisoning which was evidenced by the absence of CO stripping peak during electro oxidation of methanol (shown in fig. 14B) [56, 57].
Fig.15 shows the linearity of the oxidation current versus the concentration of methanol up to 1 M further confirms the absence of any significant poisoning effects. The catalyst particle was surrounded by the Ru hydrous oxides and it was reported as an ideal mixed proton-electron conducting medium which favours the oxidation of methanol [49, 50].
It was one of the requirements for a typical bifunctional catalyst like PtRu/C for electro- oxidation of alcohols. A certain amount of the oxidized form of Ru was also present in the
- 151 - PtRu/C samples and amount of Ru hydrous oxide present in Pt 50 Ru 50 /C appears to be more than that in Pt 70 Ru 30 /C, as inferred from the XRD reflections in fig. 12. Pt surface was likely covered predominantly with Ru-hydrous oxide in Pt 50 Ru 50 /C and hence the adsorption of methanol (a key step in the bifunctional mechanism) on Pt sites will be difficult. However, in Pt 70 Ru 30 /C, the content of Ru hydrous oxide was low and as a result, there should be three adjacent Pt sites available for adsorption of methanol [57]. Our aim is limited to the demonstration of the usefulness of the methodology of synthesizing Pt-Ru/C through cyanogel route for the electrocatalysis of methanol.
Fig. 15 Linear plot of oxidation current of methanol on Pt 70 Ru 30 /C vs. concentration of methanol.
- 152 - 6.3.3 System – III:
Investigation on interaction between PdCl 2 and K4[Fe(CN) 6]/K 3[Fe(CN) 6]
In this section, we have investigated the interaction between the PdCl 2 and
K4[Fe(CN) 6]/K 3[Fe(CN) 6]. When PdCl 2 and K4[Fe(CN) 6] solutions are mixed together, the resulting mixture changes its color immediately which depends on the ratio of both reactants
(The observed colors are tabulated in table-2 and also shown in photograph 2). Over a period of time, viscosity of the solution slowly increases and forms a complete polymeric gel. After the formation of gel, it slowly loses its water content with respect to time.
Photograph 2: Three different volumetric ratios of PdCl 2 and K4[Fe(CN) 6] displays
different colors
Table-2 Preparation of three different ratios of PdCl 2and K4[Fe(CN) 6]
(Volumetric ratios) * Total volume is 20 mL
Ratio of Pd-Fe 56 mM of PdCl 2 56 mM of K 4[Fe(CN) 6] Water* Cyanogel & Resultant colour 5 mL 5 mL 10 mL 1:1 & Green 10 mL 5 mL 5 mL 2:1 & Brown 15 mL 5 mL – 3:1 & Reddish orange
- 153 - In contrast, when PdCl 2 and K3[Fe(CN) 6] solutions are mixed together, the resulting mixture changes to red colour irrespective of the concentration of two reactants. Similar feature was also observed in Bocarsly’s group [9]. At this point of time, we are not able to arrive the conclusion that why when PdCl 2 and K 4[Fe(CN) 6] gives different colours while mixing in various ratios but not in PdCl 2 and K 3[Fe(CN) 6]. On drying the different coloured cyanogels formed from PdCl 2 and K4[Fe(CN) 6], all turned into bluish green colour. Hence, we speculate that the reason for origin of different colour could be due to different hydration of the complexes in each ratio of Pd-Fe cyanogel. Apart from these general observations, we have characterized the Pd-Fe cyanogel using cyclic voltammetry, FT-IR and TGA. We have also attempted to synthesis carbon supported PdFe (PdFe/C) via cyanogel route. Synthetic procedure of PdFe/C was given in experimental section 6.2.4. PdFe/C was characterized using XRD and voltammetry analysis.
6.3.3.1 Cyclic voltmammetric studies on Pd-Fe cyanogel formation
Cyclic voltammetry was used as a tool for studying Pd-Fe cyanogel formation. It was well known that K 4[Fe(CN) 6] is a reversible redox couple and hence changes in the redox couple after the gel formation was monitored using the cyclic voltammogram. In fig.16, cyclic voltammogram of 2 mM K 4[Fe(CN) 6] in 0.1 M KNO 3 showed a reversible redox couple at 201 mV. Cyclic voltammogram of 0.5 mM PdCl 2 and 2 mM K 4[Fe(CN) 6] showed the 24 mV anodic shift in the redox peak and also decrease in the peak current of
K4[Fe(CN) 6]. Similar anodic shift in the redox couple was observed while using the 1 mM
PdCl 2 and 2 mM K 4[Fe(CN) 6] in 0.1 M KNO 3. This kind of positive shift in the redox potentials were common for cyanometalate species, since coordination of the bridging
- 154 - cyanide to a second metal facilitates metal cyanide back bonding and thereby making the metal center more difficult to oxidize [58].
Fig.16 Cyclic voltammogram of 2 mM K 4[Fe(CN) 6] and 2 mM K 4[Fe(CN) 6] + X mM
-1 (X = 0.5 and 1 mM ) of PdCl 2 in 0.1 M KNO 3 at the scan rate of 50 mV s .
While using equimolar concentration of PdCl 2 and K 4[Fe(CN) 6] in 0.1 M KNO 3, the cyclic voltammogram showed two peaks (in fig. 17A) instead of single peak which we have observed when concentration of PdCl 2 was lower than K 4[Fe(CN) 6] (in fig.16). In fig. 16A, the first peak potential at 312 mV was due to the redox reactions of iron present in the solution and the second peak potential at 607 mV was due to the redox reactions of iron present on the electrode surface. On cycling, the peak current at 312 mV decreases but the peak current at 607 mV increases. This phenomenon was due to the formation of palladium
- 155 - hexacyanoferrate (Pd-HCF) film on the electrode surface. Similar cyclic voltammogram response for Pd-HCF was observed by Jiang et al. [59].
Fig. 17 A) Cyclic voltammogram of Pd-HCF formation on GC electrode from 2 mM
-1 PdCl 2 and 2 mM K 4[Fe(CN) 6] in 0.1 M KNO 3 at 50 mV s . B) Scan rate effect on Pd-
HCF modified GC electrode. C) Linear plot of anodic peak current and scan rate.
In fig 17B, cyclic voltammogram of Pd-HCF on the GC electrode surface in 0.1 M
KNO 3 at various scan rates were recorded. As expected, we observed a peak only for the surface bound Pd-HCF film which further confirms that the peak observed in the lower
- 156 - potential in fig. 17A was due to solution species. The redox peak in fig. 17B obeys the characteristics of an ideal surface reaction as seen from the linear relation between the plot of peak current and scan rate (fig. 17C).
6.3.3.2 FT-IR studies on Pd-Fe cyanogel
In Fig. 18, FT-IR spectra of Pd-Fe cyanogel prepared from different ratios of PdCl 2 and K 4 [Fe(CN) 6] was recorded. Here, our interest was to observe the changes in cyanide stretching vibration. Terminal and bridging cyanide stretching was observed at 2084 cm -1 and
2230 cm -1 respectively. The ratio of terminal and bridged cyanide stretching depends on the molar ratio of reactants. When equimolar ratio of PdCl2 and K4[Fe(CN) 6] was used, the intensity of terminal cyanide was more compared to bridged cyanide and when the PdCl 2 concentration was increased, the intensity of terminal cyanide stertching decreased and bridging cyanide stretching increased. Based on the above result, it was clear that the ratio of bridged and terminal cyanides was depends on the ratio of cyano and chloro metalates.
Fig. 18 FT-IR spectra of dried Pd-Fe cyanogel in 1:1, 2:1 and 3:1 ratio
- 157 - 6.3.3.3 Thermogravimetric analysis of Pd-Fe cyanogel
In fig. 19, thermogram shows the initial weight loss below 100 °C clearly due to loss of water. Major decomposition of cyanide ligand can be observed from the significant change in the weight loss at 410 °C for 3:1 ratio of Pd-Fe cyanogel. In 1:1 and 2:1 ratio of Pd-Fe cyanogel, the decomposition of cyanide ligand starts at very low temperature at 230 °C. In all the ratios the decomposition found to take place in a stepwise manner. Hence, there was weight losses occurred at different temperatures and beyond 850 °C further weight losses were not observed.
Fig. 19 TGA of three different ratios of Pd-Fe cyanogel in nitrogen atmosphere
6.3.3.4 X-ray diffraction analysis of Pd-Fe cyanogel
PdFe/C, Pd 2Fe/C and Pd 3Fe were synthesized via cyanogel route by heating the respective ratio of Pd-Fe cyanogel in Ar atmosphere for 3 hours (details given in
- 158 - experimental section 6.2.4). XRD reflections in fig. 20 show the reflections of Pd and reflections corresponding to Fe were not observed. The particle size was calculated from
Scherer formula and the particle sizes were tabulated in Table-3. There was no significant difference in the 2 theta values when compare to the Pd bulk. Hence, we speculate that the alloying of PdFe was not significant and the iron species may be remaining amorphous in the matrix. This feature can also be confirmed from the lattice parameter calculation of PdFe/C,
Pd 2Fe/C and Pd 3Fe/C because lattice parameter will change when there was a contraction or expansion of the crystal structure takes place [60]. Here, we have calculated the lattice parameter of PdFe/C, Pd 2Fe/C, Pd 3Fe/C and compared with standard Pd bulk. The calculated values were tabulated in table-3 and it showed that there were no significant change in the lattice parameter compare to Pd bulk. As we speculated from the 2 theta positions, it was confirmed that significant alloying of PdFe did not take place. More studies were needed to study the effect of temperature on the formation of PdFe/C which is beyond the scope of the present investigation.
Fig. 20 XRD of PdFe/C, Pd 2Fe/C and Pd 3Fe/C prepared from Pd-Fe cyanogel heated in
Ar atmosphere at 900 °°°C for 3 hours.
- 159 - Table-3 Particle size and lattice parameter values calculated from XRD reflections.
* In reference to JCPDS file. No 894897
Sample Average crystallite size, nm Lattice parameter, nm
Pd bulk - 0.3889 * PdFe/C 58.93 0.3891
Pd 2Fe/C 38.62 0.3892
Pd 3Fe/C 46.20 0.3889
6.3.3.5 Voltammetric analysis of PdFe/C, Pd 2Fe/C and Pd 3Fe/C
-1 Fig. 21 Cyclic voltammogram of PdFe/C in 0.5 M H 2SO 4 at the scan rate of 50 mV s
Cyclic voltammetry studies were performed on PdFe/C nanoparticles immobilized on a GC electrode. From Fig. 21, we found that the leaching of non-noble metal takes place during electrochemical cycling and it was marked by the decrease in peak current at 0.72 V.
Below 0.3 V, the peaks observed were due to the hydrogen adsorption and desorption on the surface of Pd [61]. Similar voltammetric peaks were obtained for Pd 2Fe/C and Pd 3Fe/C after prolonged voltammetric cycling. This may be due to the dissolution of iron atoms from the
- 160 - electrode surface due to poor alloying of Fe with Pd. This poor electrochemical stability is reflected in the electrocatalysis of PdFe/C catalysts. Hence, more work is needed to improve the electrochemical stability of PdFe/C, Pd 2Fe/C and Pd 3Fe/C by increasing the alloying temperature and heating duration etc. However, we have attempted to show the electrocatalytic activity of Pd-Fe/C (which are not precycled) by choosing oxygen reduction reaction (ORR) as a model reaction.
Fig. 22 Polarization curves for the ORR on PdFe, Pd 2Fe/C and Pd 3Fe/C in 0.5 M H 2SO 4 saturated with oxygen. Scan rate 5 mV s -1
Fig. 22 displays the ORR polarization curves for PdFe, Pd2Fe/C and Pd 3Fe/C in oxygen saturated 0.5 M H 2SO 4 solution. The electrocatalytic activity can be measured from its half-wave potentials and we found a gradual anodic shift in the half-wave potential from
PdFe/C to Pd 3Fe. Although ORR was occurred at less anodic potential compared to the
- 161 - reported values [62], there is a possibility to improve the electrocatalytic activity by enhancing both alloying and electrochemical stability of PdFe.
6.4 Conclusions
In system - I, the interaction between gold chloride and potassium ferricyanide or ferrocyanide was studied using UV–Vis, FT-IR, XPS and cyclic voltammetery. The studies showed that there is no reaction between potassium ferricyanide and gold chloride but a redox reaction takes place when potassium ferrocyanide reacts with gold chloride. This redox reaction involves the conversion of ferrocyanide ion to ferricyanide ion and Au (III) to Au (I) resulting in the formation of KFe x[Au(CN) 2]y. Due to high affinity of gold to attract cyanide ligand, the ligand exchange isomerism takes place from iron to gold. The Au (I) complex formation was confirmed from the XPS analysis and the presence of high spin iron in the outer sphere was also shown in cyclic voltammogram. The results show that the formation of
“gold hexacyanoferrate” is not possible either by chemical or electrochemical reactions of gold (III) chloride and potassium hexacyanoferrate (II/III) in contrast to other PB analogues.
In system - II, the interaction of H2PtCl 6 and K 4[Ru(CN) 6] at room temperature does not form cyanogel but while heating the solution at 75 °C leads to the formation of Pt-Ru cyanogel. After heating the K 4 [Ru(CN) 6] and H 2PtCl 6 solution mixture, the decrease in the redox peak current of K 4 [Ru(CN) 6] in cyclic voltammetry and the formation of IVCT band at
NIR region in UV-Vis spectra was observed which confirmed the formation of cyanogel.
PtRu/C alloy was formed by heating the PtRu cyanogel in the inert atmosphere. From the
XRD reflections, lattice parameter was calculated and the PtRu alloy formation was confirmed. The particle size was analyzed from TEM and found to be 6-10 nm. Electro-
- 162 - oxidation of methanol on Pt 70 Ru 30 /C and Pt 50 Ru 50 /C was studied and we found that the
Pt 70 Ru 30 /C shows more electro-catalytic activity compare to Pt 50 Ru 50 /C. The poisoning of the catalyst due to CO formation during methanol oxidation was also found to be minimal when compare to the commercial Pt 50 Ru 50 /C catalyst. More studies on optimization of composition, loading and alloying degree of PtRu are required to explore the electrocatlytic activity towards methanol oxidation.
In system - III, the interaction of PdCl 2 and K3[Fe(CN) 6]/ K 4[Fe(CN) 6] leads to the formation of cyanogel. When PdCl 2 and K4[Fe(CN) 6] reacts in three different ratios, formation of three distinctly different colors were noticed but only one color was observed when K3[Fe(CN) 6] was used instead of K4[Fe(CN) 6]. In cyclic voltammogram, redox peaks of K4[Fe(CN) 6] undergoes positive shift ca. 24mV during the cyanogel formation of PdCl 2 and K 4[Fe(CN) 6] which confirms the formation of cyanogel. PdFe/C was synthesized by mixing the carbon during the formation of cyanogel and followed by heating in inert atmosphere. From XRD reflections, it was confirmed that the significant alloying of PdFe did not take place. Cyclic voltammograms showed that PdFe/C was not stable and this may be due to poor alloying of PdFe which causes iron to leach out in the electrolyte during electrochemical cycling. Oxygen reduction reaction was carried out to study the elctrocatalytic activity of Pd 3Fe, Pd 2Fe and PdFe. The order of elctrocatalytic activity was found to be Pd 3Fe>Pd 2Fe>PdFe. More studies are required to improve the alloying and electrochemical stability of PdFe.
- 163 - In general, synthesis of carbon supported catalysts via cyanogel route possess following advantages
i) This method was found to be generic and opens up new possibilities to synthesize
carbon supported bimetallic alloys which have extensive applications in
electrocatalysis
ii) This method overcomes the limitations in the traditional synthetic routes and also
offers the preparation of homogenous nano-alloys which was a major challenge in
the alloy preparation
iii) Gel composition can be tailored by varying the ratios of the metal complex and
the corresponding hexacyanometallate and is amenable to compositional
modulation in the electrocatalyst.
iv) Direct synthesis of alloy nanoparticles on carbon support.
References
1. S. K. Ritter, Chem. Eng. News 83 (2005) 32.
2. K. Itaya, I. Uchida, V. D. Neff, Acc. Chem. Res. 19 (1986) 162.
3. N.R.D. Tacconi, K. Rajeshwar, R. O. Lezna, Chem. Mater. 15 (2003) 3046.
4. K. R. Dunbar, R. A. Heintz, Chemistry of Transition Metal Cyanide Compounds:
Modern Perspectives, Progress in Inorganic Chemistry, 45 (2007), John Wiley &
Sons, Inc., Hoboken, NJ, USA.
5. A. Gotoh, H. Uchida, M. Ishizaki, T. Satoh, S. Kaga, S. Okamoto, M. Ohta, M.
Sakamoto, T. Kawamoto, H. Tanaka, M. Tokumoto, S. Hara, H. Shiozaki, M.
Yamada, M. Miyake, M. Kurihara, Nanotechnology 18 (2007) 345609.
- 164 - 6. A. Johansson, E. Widenkvist, J. Lu, M. Boman, U. Jansson, Nano Lett. 5 (2005)
1603.
7. S. Vaucher, M. Li, S. Mann, Angew. Chem. Int. Ed. 39 (2000) 1793.
8. V. D. Neff, J. Electrochem.Soc 125 (1978) 886.
9. B. W. Pfennig, A. B. Bocarsly, R. K. P. Homme J. Am. Chem. Soc. 115 (1993) 2661.
10. C. M. Burgess, N. Yao, A. B. Bocarsly, J. Mater. Chem. 19 (2009) 8846.
11. H. J. Buser , D. Schwarzenbach , W. Petter , A. Ludi Inorg. Chem. 16 (1977) 2704.
12. M. Vondrova, C. M. Burgess, A. B. Bocarsly, Chem. Mater. 19 (2007) 2203.
13. C. Aparicio, L. Machala, Z. Marusak, J. Therm. Anal. Calorim. 110 (2012) 661.
14. J. Joseph, H. Gomathi, G. P. Rao, Electrochim. Acta 36 (1991) 153.
15. J. Joseph, H. Gomathi, G.P. Rao J. Electroanal. Chem. 431 (1997) 231.
16. S. Boopathi, S. Senthil Kumar , J. Joseph, K. L.N. Phani, Electrochem. Commun. 13
(2011) 294.
17. L. M. Siperko, T. J. Kuwana, J. Electrochem. Soc. 130 (1983) 396.
18. S. B. Moon, A. Xidis, V. D. Neff, J. Phys. Chem. 1993, 97, 1634.
19. S. Senthil Kumar, J. Joseph, K. L. N Phani, Chem. Mater. 19 (2007) 4722.
20. P.L. Freund , M. Spiro, J. Phys. Chem. 89 (1985) 1074.
21. S. C. Romero, J. P. Juste, P. Hervées, L.M.L. Marzáan, P.Mulvaney, Langmuir 26
(2010) 1271.
22. S. Liu, H. Li, M. Jiang, P. Li, J. Electroanal. Chem. 426 (1997) 27.
23. M. Vondrova, T. M. Mcqueen, C. M. Burgess, D. M. Ho, A. B. Bocarsly, J. Am.
Chem. Soc. 130 (2008) 5563.
- 165 - 24. M. Vondrova, T. Klimczuk, V. L. Miller, B. W. Kirby, N. Yao, R. J. Cava, A. B.
Bocarsly, Chem. Mater. 17 (2005) 6216.
25. J. Zhai, Y. Zhai, S. Dong, Colloid Surface A 335 (2009) 207.
26. O.N. Vrublevskaya, T. N. Vorobyova, H.K. Lee, S. B. Koo, Trans. Inst. Met.
Finish., 85 (2007) 254.
27. K. Itaya , I. Uchida, Inorg. Chem. 25 (1986) 389.
28. S. F. A. Kettle, E. Diana, E. M. C. Marchese, E. Boccaleri,G. Croce, T. Sheng, P. L.
Stanghellini, Eur. J. Inorg. Chem. (2010) 3920.
29. R. Cook, E.A. Crathorne, A.J. Monhemius, D.L. Perry, Hydrometallurgy, 22 (1989)
171.
30. A. Warshawsky, N. Kahana, V. Kampel, I. Rogachev, R. M. Kautzmann, J. L.
Cortina , C. H. Sampaio, Macromol. Mater. Eng. 286 (2001) 285.
31. G. J. McDougall, R. D. Hancock, M. J. Nicol, O. L. Wellington, R. G. Copperthwaite,
J.S. Afr. Inst. Min. Metall. 80 (1980) 344.
32. C. Klauber, Langmuir 7 (1991) 2153.
33. R. Cervini, R. J. Fleming, B. J. Kennedy, K.S. Murray, J. Mater. Chem. 4 (1994) 87.
34. N. SëSuèzer, S. Ertasë, O.Y.Kumser, O. Y.Ataman Appl. Spectrosc. 51 (1997) 1537.
35. D. Briggs, M. P. Seah, Practical Surface Analysis - Auger and X-ray Photoelectron
Spectroscopy (2nd ed.), Wiley Interscience (1990).
36. U. Schroöder , F. Scholz, Inorg. Chem. 39 (2000) 1006.
37. A. G. Sharpe, The chemistry of cyano complexes of the transition metals; Academic
Press: New York (1976).
38. P. A. Rock, J. Phys. Chem. 70 (1966) 576.
- 166 - 39. Z. Guo, Y. Zhang, A. Xu, M. Wang, L. Huang, K. Xu , N. Gu, J. Phys. Chem. C 112
(2008) 12638.
40. J. O. Marsden, C. I. House, Chemistry of Gold Extraction (2 nd ed.), Society for
Mining, Metallurgy, and Exploration, (2006).
41. A. Dostal, U. Schroeder , F. Scholz, Inorg. Chem. 34 (1995) 1711.
42. J. Lefebvre, D. Chartrand , D. B. Leznoff, Polyhedron 26 (2007) 2189.
43. J. Lefebvre, F. Callaghan, M. J. Katz, J. E. Sonier, D. B. Leznoff, Chem.Eur. J, 12
(2006) 6748.
44. B. F. Hoskins, R. Robson, N. V. Y. Scarlett, Angew. Chem., Int. Ed. 34 (1995) 1203.
45. M. J. Katz, T. Ramnial, H. Yu, D. B. Leznoff, J. Am. Chem. Soc. 130 (2008) 10663.
46. W. Dong, L.N. Zhu, Y.Q. Sun, M. Liang, Z. Q. Liu, D. Z. Liao, Z.H. Jiang, S. P. Yan,
P. Cheng, Chem. Commun. (2003) 2544.
47. A.E. Remick, J. Am. Chem. Soc. 69 (1947) 94.
48. J. W. Long, R. M. Stroud, K. E. S. Lyons, D. R. Rolison, J. Phys. Chem. B. 104
(2000) 9772.
49. E. S. Steigerwalt, G. A. Deluga, D. E. Cliffel, C. M. Lukehart, J. Phys. Chem. B 105
(2001) 8097.
50. D. R. Rolison, P. L. Hagans, K. E. Swider, J. W. Long, Langmuir 15 (1999) 774.
51. Z. Liu, X. Y. Ling, X. Su, J.Y. Lee, J. Phys. Chem. B 108 (2004) 8234.
52. Y. Ando, K. Sasaki, R. Adzic, Electrochem. Commun. 11(2009) 1135.
53. L. Dubau, C. Coutanceau, E. Garnier, J.M. Leger, C. Lamy, J.Appl. Electrochem 33
(2003) 419.
54. M. Neergat, D. Leveratto, U. Stimming, Fuel cells 2 (2002) 25.
- 167 - 55. C. Xu, L. Wang, X. Mu, Y. Ding, Langmuir, 26 (2010) 7437.
56. A.Wieskoski, E.R. Savinova, C.G.Vayenas, Catalysis and electrocatalysis at
nanoparticle surfaces, Marcel Dekker, Inc. New york (2003).
57. A. S. Arico, S. Srinivasan, V. Antonucci, Fuel cells 1 (2001) 133.
58. M. Zhou, B. W. Pfennig, J. Steiger, D. V. Engen, A. B. Bocarsly, Inorg. Chem. 29
(1990) 2456.
59. M. Jiang, Z. Zhao J. Electroanal. Chem. 292 (1990) 281.
60. B.D. Cullity Elements of X-ray diffraction Addison-Wesley Publishing Company,
Inc. Massachusetts (1956).
61. M. Neergat, V. Gunasekar, R. Rahul, J. Electroanal. Chem. 658 (2011) 25.
62. M -H. Shao, K. Sasaki, R. R. Adzic, J. Am. Chem. Soc. 128 (2006) 3526.
- 168 - In this thesis, we have investigated the different synthetic routes to prepare nanostructured materials. The important conclusions arrived at each chapter with special mention on the plausible ways for the extension of this promising research is given below.
This will help us to highlight some of the new unexplored directions arrived in this thesis for future work.
Chapter – III
Summary
1. The stable gold, silver and their alloy nano sols stabilized by 3-APS were
synthesized successfully and used as catalysts for the chemical reduction of 4-
nitrophenol.
2. Amino group in 3–APS undergoes protonation in acidic medium but not in the
alkaline medium. The precursors used in the synthesis of gold and silver
− + nanoparticles have different charge ( AuCl4 and Ag ). Hence the stability of metal
− + sols depends on the electrostatic interaction of AuCl4 and Ag with amino group in
3–APS at given pH.
3. Au nanosols were stable at pH~5.0 but not at pH~ 8.2 and in contrast, Ag nanosols
were stable at pH~8.2 but not at pH~5.0.
4. The stability of Au nanoparticles at pH~8.2 was improved by alloying with Ag
nanoparticles but the stability of Ag nanoparticles was not improved by alloying
with Au nano at pH~5.0. This is due to the fact that the Au undergoes reduction
faster than Ag and hence surface of the alloy particle will be rich with Ag
nanoparticles.
- 169 - 5. The Au, Ag and their alloy nanoparticles were confirmed from the surface plasmon
resonance band in UV-Visible spectroscopy and the transmission electron
microscopy studies showed that the particle size of Au-Ag is less than that of
individual metals.
6. The calculated apparent rate constants of the 4-nitrophenol reduction showed that
the Au-Ag (1:1) alloy was more catalytic compare to their mono metallic Au and Ag
nanosols.
Future work
1. Core–shell (Au@Ag or Ag@Au) nanoparticles can be synthesized by following the
sequential reduction of Au and Ag salts in presence of 3–APS.
2. Au, Ag and their alloy nanoparticles stabilized by 3–APS can be explored for the
preparation of substrates for surface enhanced Raman spectroscopy (SERS) studies.
Chapter–IV
Summary
1. Carbon supported Pt, Ru and PtRu (Pt/C, Ru/C and PtRu/C) were successfully
synthesized using microwave assisted polyol method. In this synthesis, inexpensive
solvent ethylene glycol was used as a reducing agent as well as stabilizer. The
electrocatalytic properties of the Pt/C, Ru/C and PtRu/C towards electro-oxidation of
methanol were studied in detail.
2. The reduction temperature of ruthenium ion was found to be higher than that of
platinum ion because of the difference in the redox potential.
- 170 - 3. PtRu alloy was formed by systematic variation of temperature, microwave power and
composition. XRD results showed that Pt 0.5 Ru 0.5 /C (synthesized at 160 °C with 100%
microwave power) had lower particle size (2.8 nm) compared to other compositions.
4. Pt 0.8 Ru 0.2 /C (synthesized at 160 °C with 100% microwave power) has shown superior
catalytic activity towards electro–oxidation of methanol compared to other
compositions. Due to poisoning, long term stability of Pt 0.8 Ru 0.2 /C towards methanol
oxidation is less compared to Pt 0.5 Ru 0.5 /C.
5. SNIFTIR spectroscopy studies revealed that the dehydrogenation of methanol takes
place at lower potentials and methanol oxidation was followed by linear bonding with
CO.
Future work
1. Platinum decorated on the ruthenium nanoparticle surface can be synthesized by
microwave assisted polyol method and for enhanced catalytic activity towards
electro–oxidation of methanol. This approach will undoubtedly reduce the amount of
noble metal (Pt) used in the bimetallic catalyst.
2. Series of bi-metallic alloys based on platinum can be synthesized by microwave
assisted polyol method.
3. The real time analysis can be carried out to test the activity of synthesized
electrocatalysts in the fuel cell performance.
- 171 - Chapter–V
Summary
1. Simple thermal decomposition procedure for the synthesis of porous Co 3O4 from a
single precursor is reported.
2. Thermal decomposition of Co 3[Co(CN) 6]2 results in porous Co 3O4 and this porous
structure is due to release of gaseous products like NO x, (CN) 2, CO and CO 2
3. The decomposition of Co 3[Co(CN) 6]2 was carried out at different temperatures to
analyze the formation of phase pure Co 3O4 and the required temperature is found to
be 650 °C.
4. In Li ion battery, Co 3O4 synthesized in this procedure was used as an anode material.
It showed reversible discharge capacity of 850 mAh g -1 which is very close to the
-1 theoretical capacity of Co 3O4 (890 mAh g ).
Future work
1. This method allows one to synthesize different Prussian blue analogues which can be
further decomposed to give corresponding metal oxides.
2. We have demonstrated that the Prussian blue analogues are active towards microwave
irradiation and these analogues can be decomposed to yield porous metal oxides in
few minutes for electrochemical applications. Hence, this will be focus for our future
work.
- 172 - Chapter–VI
Summary
1. The interaction between chloro metalates (of Au, Pt, Pd) and cyano metalates (of Fe,
Ru) are investigated for the formation of either Prussian blue analogue or cyanogel.
2. Interaction of HAuCl 4 and K 4[Fe(CN) 6] leads to the formation of KFe x[Au(CN) 2]y
3+ + 4– 3– through reduction of Au to Au and oxidation of [Fe(CN) 6] to [Fe(CN) 6]
confirmed from XPS and UV-vis studies , whereas there is no redox reaction observed
with K 3[Fe(CN) 6].
3. Gold has high affinity towards cyanide and hence it leads to the formation of
KFe x[Au(CN) 2]y instead of gold analogue of Prussian blue.
4. On interaction of H 2PtCl 6 and K 4[Ru(CN) 6], the gelation doesn’t takes at room
temperature and it requires thermal activation to form cyanogel. Gelation was
observed at 75 °C which was confirmed from UV-Vis and cyclic voltammetry
analysis.
5. In order to prepare carbon supported PtRu alloy, gelation was performed in the
presence of carbon and the composite gel was decomposed in an inert atmosphere at
1000 °C. XRD results confirmed the formation of PtRu alloy. Pt 70 Ru 30 /C alloy
showed good catalytic activity towards methanol electro-oxidation with minimal
poisoning due to CO.
6. Interaction of PdCl 2 and K4[Fe(CN) 6] leads to the formation of cyanogel at room
temperature (~25 °C) and interestingly the color of the cyanogel depends on the ratio
of the reactants. Similar formation of cyanogel was observed when K 3[Fe(CN) 6] was
- 173 - used instead of K 4[Fe(CN) 6] but the color of the cyanogel does not change with ratio
of the reactants.
7. Gel formed in presence of carbon was decomposed in the inert atmosphere at 900 °C
to form carbon supported PdFe alloy but XRD study showed the alloy formation does
not take place and thus electrochemical cycling stability is poor. However, it showed
promising activity towards oxygen reduction reaction.
Future work
1. Gold is present in +1 oxidation state in KFe x[Au(CN) 2]y and hence the study on the
formation of gold nanoparticles from Au +1 will be interesting.
2. In PtRu/C and PdFe/C, more studies are required to optimize the composition,
loading, extent of alloying etc.,
3. Cyanogels are active towards microwave and hence bimetallic alloys can also be
synthesized by microwave irradiation which will reduce the duration of time from
hours to minutes.
- 174 - List of Publications
Related to docotoral work
1. Role of pH on aminosilane stabilized colloidal gold/silver and their alloy sols – application in catalysis S. Harish, R.Sabarinathan, James Joseph, K.L.N. Phani Mat.Chem.Phys 127 (2011) 203-207 DOI:10.1016/j.matchemphys.2011.01.060
2. Interaction between gold chloride and hexacyanoferrate (II/III) ion - Does it lead to gold analogue of Prussian blue? S. Harish , James Joseph, K.L.N. Phani Electrochim. Acta – 56 (2011) 5717 - 5721 DOI : 10.1016/j.electacta.2011.04.044
3. Microwave assisted polyol method for the preparation of Pt/C, Ru/C and PtRu/C nanoparticles and its application in electro-oxidation of methanol S. Harish , S. Baranton, C. Coutanceau, James Joseph J. Power Sources – 214(2012) 33-39 DOI: 10.1016/j.jpowsour.2012.04.045
4. Synthesis of nanoporous Co 3O4 obtained from a unique precursor Co 3[Co(CN)6] 2 - Application in Li-ion battery S. Harish, K. Hemalatha, James joseph Mat. Chem. Phys – Submitted
5. Effect of temperature, microwave power and composition in microwave assisted polyol method for the preparation of PtRu/C nanoparticles and their application in electro-oxidation of methanol S. Harish , S. Baranton, C. Coutanceau, James Joseph Under preparation
- 175 - Other publications
1. PEDOT/Palladium composite material: synthesis, characterization and application to simultaneous determination of dopamine and uric acid. S. Harish, J. Mathiyarasu, K. L. N. Phani, V. Yegnaraman J. Appl. Electrochem 38 (2008) 1583 – 1588. DOI 10.1007/s10800-008-9609-0
2. Synthesis of Conducting Polymer Supported Pd Nanoparticles in aqueous medium and catalytic activity towards 4-Nitrophenol Reduction. S. Harish, J. Mathiyarasu, K. L. N. Phani, V. Yegnaraman Catal. Lett 128(2009) 197 – 202. DOI 10.1007/s10562-008-9732-x
3. Barrier films to control loss of 9, 10-anthraquinone-2-sulphonate dopant from PEDOT films during electrochemical transitions S. Harish , D. Sridharan, S. Senthil Kumar, James Joseph, K.L.N. Phani Electrochim. Acta 54 (2009) 3618-3622 DOI:10.1016/j.electacta.2009.01.032
4. Generation of gold–PEDOT nanostructures at an interface between two immiscible solvents. S. Harish, J. Mathiyarasu, K. L. N. Phani Mater. Res. Bull 44 (2009) 1828 -1833 DOI:10.1016/j.materresbull.2009.05.022
5. Ni x-Fe (1-x) Fe(CN) 6 hybrid thin films electrodeposited on glassy carbon: Effect of tuning of redox potentials on the electrocatalysis of hydrogen peroxide
A. V. NarendraKumar, S. Harish , James Joseph, K.L.N. Phani J. Electroanal. Chem – 659 (2011) 128 - 133 DOI : 10.1016/j.jelechem.2011.05.006
- 176 - This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright Author's personal copy
Materials Chemistry and Physics 127 (2011) 203–207
Contents lists available at ScienceDirect
Materials Chemistry and Physics
journal homepage: www.elsevier.com/locate/matchemphys
Role of pH in the synthesis of 3-aminopropyl trimethoxysilane stabilized colloidal gold/silver and their alloy sols and their application to catalysis
S. Harish, R. Sabarinathan, James Joseph ∗, K.L.N. Phani
Electrodics and Electrocatalysis Division, CSIR-Central Electrochemical Research Institute, Karaikudi 630 006, India article info abstract
Article history: A general method for the synthesis of nano-alloy dispersions of noble metals like Au and Ag in organically Received 24 August 2010 modified aminosilicate sol is presented with an emphasis on the influence of pH on the sol stability. The Received in revised form alloys of Au–Ag were synthesized by co-reduction of solutions of gold and silver salts using borohydride 14 November 2010 as the reducing agent at a suitable pH in the medium containing 3-aminopropyltrimethoxysilane (APS) Accepted 24 January 2011 stabilizer. Organosilanes with amine functional group have multiple roles in the formation of stable sols such as the formation of siloxane network structures by hydrolysis and condensation and the interaction Keywords: of amino group with metal ions in solution electrostatically. Mono metal and alloy sols were characterized Sol–gel growth Nanostructures by transmission electron microscopy (TEM), ultraviolet–visible (UV–vis) spectroscopy. The stable alloy Alloys sols were found to catalyze the chemical reduction of 4-nitrophenol to 4-aminophenol and the catalytic rate constants were evaluated from UV–vis spectroscopy. © 2011 Elsevier B.V. All rights reserved.
1. Introduction et al. have recently reported the synthesis of alloy dispersions sta- bilized by organically modified silicate matrices [13], wherein the Nanosized noble metal colloids have generated a great deal of electrocatalytic activity of the dried films of alloy sols immobilized interest among researchers owing to their electronic and opti- on glassy carbon electrodes for the oxygen reduction reaction was cal properties, distinctly different from their bulk counterpart demonstrated. [1–3]. Leff et al. reported the synthesis of alkylamine-capped gold In the work described in this paper, we have prepared APS- nanoparticles similar to the preparation of thiol-stabilized gold sols stabilized Au and Ag sols. The pH of the medium was found to [4]. These sols were stable only in organic solvents. The stabilization have major effects on the stability of the sol. The alloys of Ag and of metal nanoparticles in the silane stabilized sols [5–13] provides a Au of various compositions, stabilized by APS were prepared by promising way to study this dispersion in aqueous solution for their co-reduction of the two metal salts. The nano-sols were charac- optical and catalytic applications. One advantage of using the silane terized by UV–vis spectroscopy and TEM. The application of finely stabilized sol is that the optical properties of the nanocrystalline dispersed noble metal sols in siloxane networks for catalysis is dispersant are not altered much by the transparent silicate matrix. not exploited much. Functionalised aminobenzenes are important The porous siloxane network formed by hydrolysis and condensa- intermediates for the manufacture of many agro-chemicals, phar- tion of silane can facilitate the percolation of reactant and products. maceuticals, dyes and pigments. The mechanism of hydrogenation Sol–gel monomers containing amine functional group can exhibit of aromatic nitro compounds was explained in detail by Corma dual role such as chelating with metal ion and cap the metal sol. et al. [15] and Blaser [16]. The reduction of 4-nitrophenol was Bharathi et al. studied aminosilicate stabilized Au, Pt, Pd and Ag used as a model reaction to demonstrate the catalytic behaviour of nanosol [6–13]. They established the functional role played by the metal nanoparticles embedded in PEDOT colloids from this labora- amino group in stabilizing the mono-metal sol. It was found that tory [17,18] and also by others [19,20]. Hence, application of these the average particle size of the Au, Pd and Pt particles to be in the sols in the catalysis of 4-nitrophenol reduction is chosen as model range of 5–7 nm, while the size of the Ag nanoparticles prepared by reaction. Change in the catalytic rate constant of 4-nitrophenol the same method are larger in size (2–20 nm). reduction was studied as a function of alloy composition. Au–Ag alloy formation takes place easily than the other systems because of the lattice constants of these metals are nearly the same 2. Experimental and the miscibility gap between Au and Ag also less [14]. Sampath 2.1. Materials and methods
All chemicals used were of analytical grade. Solutions were prepared using Milli- Q water (resistivity 18.2 M). All glasswares were thoroughly cleaned with freshly ∗ Corresponding author. Tel.: +91 4565 227550; fax: +91 4565 227779. prepared aqua-regia and rinsed with Milli-Q water before use. UV–vis spectra were E-mail address: [email protected] (J. Joseph). collected using Varian Cary 500 spectrophotometer. The TEM images were collected
0254-0584/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.matchemphys.2011.01.060 Author's personal copy
204 S. Harish et al. / Materials Chemistry and Physics 127 (2011) 203–207
using Philips CM-200 microscope working at 200 kV. The metal sols were diluted and sonicated for uniform dispersion. The samples for TEM were prepared by placing a drop of the sample onto a copper grid coated with carbon film (400 meshes). The grid was allowed to get dried in air for overnight at room temperature.
2.2. Synthesis of gold/silver nanoparticles
The solutions of 0.1 M APS were prepared in water and initial pH was found
to be 10.6. The pH of the solution was adjusted to 8.0 by adding dilute HNO3. This solution was sonicated for 10 min. Hydrolysis and condensation of APS was initi-
ated on adding HNO3. In the case of the preparation of Ag sol, use of HCl and H2SO4 were avoided to prevent precipitation of corresponding silver salt. To the above n+ solution, either 0.5 mM HAuCl4 or AgNO3 (ratio of APS:M = 200:1 in molar ratio) was added and sonicated for 30 min in order to homogeneously distribute the metal ions in the solution. The chemical reduction was effected by adding a few drops of
freshly prepared 50 mM NaBH4 to the APS stabilized metal ion solution. The for- mation gold/silver nanoparticles formed were stabilized by the APS. The final pH of the solution was 8.2. The reaction was carried out at room temperature of 25 ◦C. A similar procedure was employed for the preparation of nano-colloids at pH 5.0 also.
2.3. Synthesis of gold–silver alloy
Instead of adding gold or silver salts individually, we have added both salts to a Fig. 1. . The UV–vis spectra of Au and Ag nanoparticles stabilized by APS at pH ∼8.2. medium containing APS at pH 8.0 and co-reduced by adding a few drops of freshly prepared 50 mM NaBH4. This resulted in the formation of gold–silver alloy sols. The molar ratio of gold and silver was varied by taking various ratios of Au and Ag salts, maintaining the total precursor salt concentration at 0.5 mM. The final pH of the alloy sol after reduction was found to be 8.2.
Fig. 2. (A) TEM images of Au and Ag nanoparticles stabilized by APS at pH ∼8.2. (B) Histogram of Au and Ag nanoparticles. Author's personal copy
S. Harish et al. / Materials Chemistry and Physics 127 (2011) 203–207 205
2.4. Chemical reduction of nitrophenol
The reduction of 4-nitrophenol (4-NP) was chosen as a model reaction to show the catalytic behaviour of synthesized Au–Ag nanoparticles. The reduction reaction was carried out in a standard quartz cuvette of 1 cm path length and of 3 mL volume.
The procedure entailed in mixing excess NaBH4 (10 mM) with a 4-NP (0.05 mM) solution in water in the quartz cell followed by the addition of aliquots of catalyst particle dispersions (50 L). The absorption spectra were recorded at time intervals of 50 s by scanning the wavelength range from 250 nm to 700 nm at 25 ◦C.
3. Results and discussion
3.1. Characterization of gold and silver sols
In the preparation of the Au sol (described in Section. 2), APS undergoes slow hydrolysis and condensation catalysed by acid or base to form a siloxane network. After the chemical reduction of APS containing HAuCl4 by borohydride, the colour of the solution turned to characteristic wine red. It is observed that the molar ratio of APS:HAuCl4 is above 20 for the formation of a stable Au sol. The absorption spectrum of the gold sol shows the characteristic surface plasmon resonance (SPR) at 533 nm [21] (Fig. 1). The APS and N-[3-(trimethoxysilyl)propyl]-ethylenediamine (EDAS) stabi- Fig. 3. UV–vis spectra of physically mixed gold and silver nanoparticles stabilized ∼ lized Au sols were also reported to form very stable dispersions by APS at pH 8.2. of Au at pH 5.0 [8,11] which clearly establishes the role played by the aminosilanes in binding and dispersing gold nanoparticles in group in APS at a given pH. Further, Bharathi et al. [12] have aqueous media. The formation of stable APS stabilized Au sols at reported the synthesis of Au, Pt, Pd and Ag nanoparticles using APS pH 5.0 was first reported by Bharathi et al. [11]. This observation at pH ∼5.0. They observed that the particle size of Au, Pt and Pd − led us to consider that the electrostatic attraction between AuCl4 were smaller when compared to the size of Ag. They concluded that and protonated amine in the siloxane network plays a vital role in a weak interaction between the amino group of APS and the Ag is − stabilizing the Au sol. Similar binding of AuCl4 to aminosiloxane the origin of the formation of larger particle size of Ag at pH 5.0. networks was also reported by Henao et al. [22]. However, the gold In this study, we found that by increasing the pH to 8.2, the inter- sols prepared at pH ∼8.2 were found to be less stable compared to actions between the Ag+ and the APS can be increased and hence those at pH ∼5.0 as the aggregation of gold particles was observed the increased stability. Thus, pH plays a vital role in the preparation on standing for more than two days. TEM image shown in Fig. 2a and stability of the noble metal colloids. reveals the aggregated structure of Au nanoparticles prepared at pH ∼8.2. The Au particles with average size of 7.8 nm were observed 3.3. Formation of Au–Ag alloy sols from Fig. 2b. Similarly, the Ag sol was prepared by the procedure described We found that physical mixing of Au and Ag nanosols prepared in Section 2. Fig. 1 shows a band at 405 nm [21] which charac- at pH ∼8.2 did not lead to the formation of alloy sols, as observed terises SPR of Ag nanoparticles, the protecting agent being APS. in UV–vis spectroscopy (Fig. 3). The marked presence of individ- The colour of the Ag sol is yellow. The APS-stabilized Ag sols at pH ual SPR peaks for both metals confirmed this observation. This is in ∼8.2 were found to be very stable for at least a month in contrast accordance with an earlier report in Ref. [13]. In this study, Au–Ag to the observed aggregation of Au sols at pH ∼8.2. A non-uniform sols of various molar ratios were prepared by co-reduction of a mix- particle size distribution of Ag particles was observed in the TEM ture of solutions of Au and Ag salts in APS at pH ∼8.2. Surprisingly, (Fig. 2a). The TEM images show that Ag particles are well separated there was only a single SPR peak indicating possible alloying of Au from each other and the average size of the particles was 7.2 nm as with Ag. The alloy formation was confirmed from the blue shift in calculated from the histogram Fig. 2b. However, the Ag sols were the UV–vis spectra (Fig. 4A) as a function of increasing composi- found to be unstable at pH ∼5.0 and agglomeration of bigger black tion of Ag in the Au–Ag alloy [23]. The colour of the Au, Au–Ag and silver particles takes place instantaneously. Ag colloids gradually changes from wine red to yellow as shown in the photograph [inset in Fig. 4A]. It is reported that a very close 3.2. Stability of the sols miscibility gap and the similar lattice parameters of Au and Ag are favourable for the alloy formation [13,14]. The absorption maxi- + − Hydrogen tetrachloroaurate (III) dissociates into H and AuCl4 mum depends on the mole fraction of the Au in the alloy sol as seen in aqueous solutions. At pH ∼5.0, the amino group in APS is pro- from a continuous shift of the SPR band compared to mono-metal tonated and hence there will be strong electrostatic attraction sols and it varies linearly with the composition (Fig. 4B) [24]. The between tetrachloroaurate anion and APS. However, the interac- TEM image of nearly uniform Au–Ag(1:1) particles (Fig. 5a) shows − tion between tetrachloroaurate anion (AuCl4 ) and free amine in that the particles are well separated from each other. The average APS will not be effective at pH ∼8.2 (the amino group in APS particle size as calculated from the histogram shown in Fig. 5bof is not protonated in the alkaline pH range). By contrast, the Ag the Au–Ag alloy sol is about 4.2 nm which are small when compared nanoparticles are more stable at pH ∼8.2. Here, the precursor to the size of the respective monometallic colloids. The stability of + − AgNO3 dissociates into Ag and NO3 , and hence the electrostatic the Au sol at pH ∼8.2 is found to increase dramatically on alloying attraction between Ag+ and free amino functionality in APS will even with 4% of Ag. The increased stability of Au sols alloyed with be the least. Nevertheless, due to the presence of a free electron 4% Ag characterized by the invariance of the spectra observed on pair in the amino-group, APS can act as a complexing ligand with standing for a month. However, the alloying of Au at lower concen- pseudo-coordinate interaction with Ag+ ion. Hence, the stability or tration levels did not increase the stability of Ag sols at acidic pH. instability of Ag nanoparticle sol depends on the charge of amino It may be due to the fact that the Au being the nobler metal has Author's personal copy
206 S. Harish et al. / Materials Chemistry and Physics 127 (2011) 203–207
Fig. 4. (A) UV–vis spectra of Au–Ag alloy prepared by co-reduction in APS at pH ∼8.2; inset: photograph shows the colour change of the samples (a–g) – from left to right. (B) Plot of wavelength corresponding to the maximum absorbance for various mole fractions of alloy nanoparticles. a tendency to form the nuclei first, over which the silver forms as Fig. 5. (A) TEM image of Au–Ag(1:1) alloy stabilized by APS at pH 8.2. (B) Histogram a shell. High-resolution TEM studies may elucidate the core–shell of Au–Ag(1:1) nanoparticles. nature of the particles formed at pH 5.0. the presence of excess NaBH4, continuous fading of yellow colour 3.4. Reduction of 4-nitrophenol was noticed. This indicates the conversion of 4-nitrophenol to 4- aminophenol. In order to follow the kinetics of this reaction, UV–vis It is well known [15,16] that the reduction of 4-nitrophenol to 4- spectrophotometric studies were carried out to determine the rate aminophenol requires the presence of a metal catalyst along with constant of the reaction. It has been established that the addition a reducing agent. 4-Nitrophenol absorbs in the visible region at of aliquot amounts of Au or Ag or Au–Ag colloidal sols will not 313 nm which undergoes a red shift to 400 nm on the addition of interfere in the UV–vis spectra. NaBH4 [not shown in figure] (colour changes from light yellow to The kinetics of the catalytic conversion of 4-nitrophenol to 4- dark yellow). The reason for the red shift is associated with the for- aminophenol was monitored in UV–vis spectroscopy at regular mation of nitrophenolate anion. The absorption at 400 nm remains time intervals. As a representative case, the spectral change with unaltered even after the addition of excess NaBH4 which shows that the addition of the Au–Ag(1:1) alloy catalyst to the reactant solution reduction of nitro group is not possible in the absence of a metal is shown Fig. 6. The reduction reaction can be monitored from the catalyst. This feature is observed from the work published from our decrease in the intensity of absorption of 4-nitrophenol at 400 nm laboratories in which dispersions of Au and Pd nanoparticles sta- and the emergence of a new absorption peak at 310 nm corre- bilized in the PEDOT-PSS matrix were used as the catalyst for the sponding to the formation of 4-aminophenol [17,18]. In the course conversion of 4-nitrophenol to 4-amino phenol. The calculated rate of reaction, the concentration of NaBH4 is in excess and remains constant values showed Pd to be more catalytic than Au [17,18]. The constant and hence the order of the reaction can be considered catalytic activity of metal nanoparticles embedded in various stabi- pseudo-first order. More importantly, large excess of NaBH4 takes lizers was also explored in the reduction of 4-nitrophenol by others in to account the slow but noticeable hydrolysis of this reagent at [19,20,25,26]. alkaline pH. The plot between the logarithmic value of absorbance Interestingly, on addition of small aliquots of Au, Ag, Au–Ag and time is found to be linear (inset Fig. 6). A similar observation is (0.08–0.20 g/l) colloids to the solution containing nitrophenol in made for the chemical reductions using other alloy compositions Author's personal copy
S. Harish et al. / Materials Chemistry and Physics 127 (2011) 203–207 207
ing up to 50% is beneficial for the catalytic properties of Au–Ag alloy.
4. Conclusions
This work reports the synthesis of compositionally different Au–Ag alloy sols stabilized by APS. The stability of the colloidal solution of Au and Ag sols stabilized with APS was found to be dependent on the pH of the medium. Alloying with Ag was found to increase the stability of Au sol at pH ∼8.2. In the catalytic application of these mono-metal and alloy sols, the experimental rate con- stants for the reduction of 4-nitrophenol to aminophenol revealed that the alloying up to 50% is beneficial for the catalytic properties of Au–Ag alloy. This method of synthesis of alloy sol stabilized by functional siloxane network may be suitable for designing several heterogeneous catalytic systems.
Acknowledgements
Fig. 6. UV–vis spectra showing the formation of 4-aminophenol in the presence of excess NaBH4 using Au–Ag(1:1) alloy as a catalyst (inset: linear plot of log A vs time). The Department of Science and Technology, New Delhi, India is thanked for the financial assistance (SR/S1/PC-22/2007). Authors also thank the Central Instruments Facility for using UV–vis spec- Table 1 Comparison of values of apparent rate constant for Au, Ag and Au–Ag nanoparticles trometry and TEM analysis. S. Harish thanks CSIR, India for the grant as catalysts for the reduction of 4-nitrophenol in the presence of excess NaBH4. of a senior research fellowship.
Catalyst Rate constant, k (×10−3 s−1) References Au 6.133 Au–Ag(96:4) 14.342 [1] C.N.R. Rao, G.U. Kulkarni, P.J. Thomas, P.P. Edwards, Chem. Soc. Rev. 29 (2000) Au–Ag(50:50) 27.002 27. Au–Ag(4:96) 12.168 [2] J. Brinker, G. Scherer, Sol–Gel Science, Academic Press, San Diego, 1989. Ag 19.949 [3] O. Lev, Z. Wu, S. Bharathi, V. Glezer, A. Modestov, J. Gun, L. Rabinovich, S. Sampath, Chem. Mater. 9 (1997) 2354. [4] D.V. Leff, L. Brandt, J.R. Heath, Langmuir 12 (1996) 4723. [5] C.A. Morris, M.L. Anderson, R.M. Stroud, C.I. Merzbacher, D.R. Rolison, Science also. The calculated values of the apparent rate constant using the 284 (1999) 622. [6] S. Bharathi, M. Nogami, O. Lev, Langmuir 17 (2001) 2602. methodology adopted by Panigrahi et al. [26] for 4-nitrophenol [7] S. Bharathi, M. Nogami, S. Ikeda, Langmuir 17 (2001) 1. reduction using mono-metal and alloys of different ratios is tab- [8] S. Bharathi, J. Joseph, O. Lev, Electrochem. Solid State Lett. 2 (1999) 284. ulated in Table 1. From Table 1, it is seen that the Au sol shows [9] T. Hayakawa, Y. Usui, S. Bharathi, M. Nogami, Adv. Mater. 16 (2004) 1408. [10] S. Bharathi, O. Lev, Anal. Commun. 35 (1998) 29. a lower catalytic activity compared to the Ag sol. The Au–Ag(1:1) [11] S. Bharathi, O. Lev, Chem. Commun. 23 (1997) 2303. sol shows a higher rate constant, compared to the Ag and Au sols [12] S. Bharathi, N. Fishelson, O. Lev, Langmuir 15 (1999) 1929. alone. The lowest rate constant shown by Au may be attributed [13] S. Devarajan, P. Bera, S. Sampath, J. Colloid Interface Sci. 290 (2005) 11. to the instability of Au nanoparticles at pH ∼8.2. The rate con- [14] K. Kim, K.L. Kim, S.J. Lee, Chem. Phys. Lett. 403 (2005) 77. [15] A. Corma, P. Serna, Science 313 (2006) 332. stant for the 4-nitrophenol reduction in the presence of Au–Ag(1:1) [16] H.U. Blaser, Science 313 (2006) 312. alloy is of higher value than the respective metal nanoparticles [17] S. Senthilkumar, C. Sivakumar, J. Mathiyarasu, K.L.N. Phani, Langmuir 23 (2007) and this can be attributed to the higher exposed area of catalysts 340. [18] S. Harish, J. Mathiyarasu, K.L.N. Phani, V. Yegnaraman, Catal. Lett. 128 (2009) available due to lower particle size distribution (higher surface-to- 197. volume ratio). The average size of the Au–Ag sol is less (4.2 nm) [19] Y. Lu, Y. Mei, M. Drechsler, M. Ballauff, Angew. Chem. Int. Ed. 45 (2006) 813. when compared to Ag (7.2 nm) and Au (7.8 nm). The rate constants [20] N. Pradhan, A. Pal, T. Pal, Colloid Surf. A 196 (2002) 247. [21] P.K. Jain, X. Huang, I.H. El-Sayed, M.A. El-Sayed, Acc. Chem. Res. 41 (2008) 1578. were found to be comparable to those reported in literature [18]. [22] J.D. Henao, Y.W. Suh, J.K. Lee, M.C. Kung, H.H. Kung, J. Am. Chem. Soc. 130 (2008) Here again, it may be emphasized that the pH conditions under 16142. which the mono-metal or alloy colloids are stable govern the rate [23] S. Link, Z.L. Wang, M.A. El-Sayed, J. Phys. Chem. B 103 (1999) 3529. [24] M.P. Mallin, C.J. Murphy, Nano Lett. 2 (2002) 1235. of the reduction of nitrophenol. In the catalytic application of these [25] S. Saha, A. Pal, S. Kundu, S. Basu, T. Pal, Langmuir 26 (2010) 2885. mono-metal and alloy sols, the experimental rate constants for the [26] S. Panigrahi, S. Basu, S. Praharaj, S. Pande, S. Jana, A. Pal, S.K. Ghosh, T. Pal, J. reduction of 4-nitrophenol to aminophenol revealed that the alloy- Phys. Chem. C 111 (2007) 4596. (This is a sample cover image for this issue. The actual cover is not yet available at this time.)
This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright Author's personal copy
Journal of Power Sources 214 (2012) 33e39
Contents lists available at SciVerse ScienceDirect
Journal of Power Sources
journal homepage: www.elsevier.com/locate/jpowsour
Microwave assisted polyol method for the preparation of Pt/C, Ru/C and PtRu/C nanoparticles and its application in electrooxidation of methanol
Srinivasan Harish a,b, Stève Baranton a, Christophe Coutanceau a,*, James Joseph b a Université de Poitiers, IC2MP, UMR CNRS 7285, 4 rue Michel Brunet B27, 86022 Poitiers cedex, France b Electrodics and Electrocatalysis Division, CSIR e Central Electrochemical Research Institute, Karaikudi, Tamilnadu, India article info abstract
Article history: A polyol process activated by microwave irradiation was implemented to prepare efficient Pt/C, Ru/C and Received 16 January 2012 Pt1Ru1/C electrocatalysts for methanol oxidation with reducing synthesis cost and time. Study of the Received in revised form post-synthesis mixture by UVevisible spectroscopy led to determine the minimum batch temperature 16 March 2012 and synthesis time necessary for the complete reduction of metal salts. It was shown that disappearance Accepted 1 April 2012 of metal salts and colloid formation took place after 5 min at 100 C for Pt, 5 min at 130 C for Ru and Available online 25 April 2012 5 min at 160 C for Pt1Ru1. The PtRu catalyst characterizations by TGA, TEM and XRD indicated that the Key words: nominal loading and nominal composition were achieved, and that the structure of this material con- e CO stripping sisted of a mixture of Pt0.8Ru0.2 monocrystalline particles of ca. 2.5 3.0 nm, RuO2 nanoclusters and Methanol oxidation probably Ru nanoclusters. Pt1Ru1/C catalyst displayed a high activity towards CO and methanol elec- Microwave activation trooxidation, with onset potentials of ca. 0.2 V lower than those obtained on Pt/C catalysts, and a low Platinum surface poisoning at 0.6 V vs. as demonstrated by chronoamperometry measurement RHE and in situ Polyol method infrared reflectance spectroscopy. Ruthenium Ó 2012 Elsevier B.V. All rights reserved.
1. Introduction method uses inexpensive solvent (ethylene glycol), does not need the presence of surfactant to achieve well-dispersed catalytic The application of fuel cell is currently limited by inadequacy in particles of small mean sizes and is very easy to implement. materials performance. Sir William Grove had already stated in Bock et al. [10] have studied the synthesis of PtRu nanocatalysts 1839 [1] that ‘‘the chief difficulty was to obtain anything like via the polyol method. The reaction was performed at reflux at a notably surface for action”. For this reason, nanosized particles of 160 C for several hours. They showed that the reaction mechanism platinum and platinumeruthenium supported on carbon are still involved the oxidation of ethylene glycol to aldehyde and then to þ the most used electrocatalysts in Proton Exchange Membrane Fuel glycolic acid or, depending on the pH, to glycolate, while the Pt4 þ Cell (PEMFC) and Direct Methanol Fuel Cell (DMFC) electrodes, and Ru3 precursor salts were reduced. In the case of Pt/C catalyst respectively. synthesis, Liu et al. [11] explained that in the synthesis process the For obtaining such nanostructured catalysts, numerous polyol solution containing the metal salt was refluxed at synthesis methods were developed, including electrochemical 120 Ce170 C in order to decompose ethylene glycol and to deposition [2,3], physical vapour deposition [4], colloidal routes [5], generate in situ reducing species for the reduction of the metal ions impregnation reduction route [6], etc. Amongst the chemical to their elemental states. In traditional synthesis method, the methods, colloidal routes based on the use of organic surfactant as reduction reaction is then activated by temperature: the synthesis stabilizing agent are not industrially scalable methods. However, is carried out by heating the reaction mixture at temperature the polyol method has shown very promising potency for the higher than 120 C for several hours [7,9,12,13]. preparation of Pt [7] and bimetallic Pt-based nanoparticles [8]. This Recently, Lebègue et al. [14] developed a method based on the method, well described by Fievet et al. [9], allows obtaining metal activation reaction by microwave impulsion, which allowed energy nanoparticles by reduction of metallic salts in ethylene glycol and and preparation time savings, to synthesize well-dispersed Pt/C can be performed without addition of any surfactant. So, the polyol catalyst of ca. 2.5 nm mean size, low size distribution, leading to high electrochemical surface area. These authors showed that such method leads to highly active Pt/C catalyst towards the oxygen * Corresponding author. Tel.: þ33 5 49 45 48 95; fax: þ33 5 49 45 35 80. reduction reaction and highlighted the effect of microwave acti- E-mail address: [email protected] (C. Coutanceau). vation at low temperature (80 C and 100 C) on the structure of Pt/
0378-7753/$ e see front matter Ó 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.jpowsour.2012.04.045 Author's personal copy
34 S. Harish et al. / Journal of Power Sources 214 (2012) 33e39
C catalysts, particularly on active surface area and metal loading on temperature increase during microwave irradiation. The synthesis carbon. of catalysts was performed under continuous microwave irradia- The microwave dielectric heating leads to thermal and non- tion at a power of 1600 W until reaching the desired reaction thermal effects [15]. Thermal effects arise from different tempera- temperature, and then microwave pulses were applied to maintain ture regimes under microwave heating, whereas non-thermal it. Fig. 1 shows the schema representing the microwave sequence effects result from effects inherent to the microwaves [16]. These and corresponding solvent temperature profile during the last authors showed that these effects lead to different morphol- synthesis process by pulsed microwaves. ogies and sizes of metallic nanostructures in comparison with those The catalytic powders were washed with acetone and ultra pure obtained by a conventional oil-bath heating. They also underlined water (MilliQ, Millipore, 18.2 MU cm), and filtrated. Finally, thermal that detailed mechanism for the preparation of metallic nano- treatment of Pt/C catalysts is performed at 160 C [10] for 1 h under structures under microwave irradiation has not been yet clarified. air to remove traces of ethylene glycol. But, according to the generally admitted metal salt reduction mechanism via the formation of aldehyde intermediate during the 2.2. Physicochemical characterization of the Pt/C, Ru/C and Pt1Ru1/ polyol synthesis process, and considering that ethylene glycol C catalysts possesses high dielectric losses and high reduction ability [16],it can be proposed that the role of microwave could be to favour the Thermogravimetric analyses were carried out with a TA dehydrogenation of the molecule and hence the reduction of Instrument SDT Q600 apparatus. A few milligrams of catalytic metallic salts. Anyway, the main advantages of microwave irradi- powder was put in an alumina crucible and heated under air from ation were discussed by Tsuji et al. [16]: the uniform heating of the 25 Cto900 C with a temperature slope of 5 C min 1. solution leading to a more homogeneous nucleation and shorter Transmission electron microscopy (TEM) measurements were crystallization time, a very short thermal induction period, the carried out with a JEOL JEM 2010 (HR) with a resolution of 0.35 nm. generation of localized high temperatures at the reaction sites The determination of the nanoparticle size distribution was per- resulting in enhancement of reduction rates of metallic ions and formed with the ImageJ free software [17] and estimated from the superheating of solvents over the boiling points as a consequence of measurement of 200e300 isolated nanoparticles to have accept- the microwave dissipation over the whole liquid volume. Micro- able statistical samples. wave dielectric loss heating appears then as a better synthesis X-ray diffraction (XRD) patterns were recorded on a Bruker D option in view of its energy efficiency, preparation time saving, 5005 Bragg-Brentano (qeq) diffractometer operated with a copper uniformity of heating across the whole solvent volume (no thermal tube powered at 40 kV and 40 mA (CuKa1 ¼1.5406 Å). Measure- convection), and implementation simplicity. ments were performed from 2q ¼ 15 to 2q ¼ 90 in step mode, For these reasons, a polyol method activated by microwave with steps of 0.06 and a fixed acquisition time of 10 s/step. irradiation was implemented in the present contribution in order to UVevisible measurements were carried out using a spectro- synthesis Pt/C and Pt1Ru1/C catalysts. The nanosized materials were photometer Evolution 100 UVevisible from Thermo Electron Corp. characterized by TEM and XRD and their electro-catalytic behav- iour towards electrooxidation of carbon monoxide and methanol 2.3. Electrochemical studies of the Pt/C, Ru/C and PtRu/C catalysts was evaluated using cyclic voltammetry and in situ infrared study. Electrochemical measurements were carried out in a standard 2. Experimental three electrode electrochemical cell at room temperature with a reversible hydrogen electrode (RHE) as the reference electrode 2.1. Synthesis of the Pt/C, Ru/C and Pt Ru /C catalysts 1 1 and a glassy carbon plate as the counter electrode. The support electrolyte was a 0.5 M H SO (Suprapur, Merck) solution in ultra $ 2 4 Appropriate amounts of H2PtCl6 6H2O and/or RuCl3, (99.9% pure water. Methanol electrooxidation experiments were carried purity, Alfa Aesar) were dissolved in 100 mL of ethylene glycol out in N -saturated support electrolyte containing 0.1 M methanol 2 (puriss. p.a., 99.5% Fluka) in order to reach a concentration of (Absolute Puriss. 99.8%, SigmaeAldrich). 1 metals of 0.375 g L (Table 1). Then, pH of the solution was The working electrode is prepared by deposition of a catalytic adjusted at 11 for the pure Pt sample and 10 for Ru containing ink on a 0.071 cm2 Glassy carbon disk according to a method catalysts by adding a solution of NaOH (1 M) in ethylene glycol drop wise. Carbon Vulcan XC72R (150 mg) thermally treated for 4 h at 400 C under Nitrogen (U Quality from “Air Liquide”) was then added to the solution in order to obtain a nominal metal loading of 20 wt% on carbon and the mixture was ultrasonically homogenized for 5 min. The reactor equipped with a cooler was put inside a MARS oven from CEM Corporation. Such set up activates the synthesis reaction by microwave irradiation at atmospheric pres- sure, without evaporation of ethylene glycol and/or water due to erature p
Table 1 tem Metal salt weights dissolved in 100 mL of ethylene glycol and corresponding concentration for the synthesis of catalysts loaded with 20 wt.% of metal. The carbon microwave power powder mass added is 150 mg.
Metal salt weight/mg Metal concentration/g L 1 Me/(Me þ C) ratio/wt.% 0 5 H2PtCl6$6H2O RuCl3 Pt Ru Pt/C 100 0.375 20 time / min Ru/C 77 0.375 20 Fig. 1. Schema of the microwave sequence and of the temperature change as a func- Pt Ru /C 66 26 0.248 0.128 20 1 1 tion of time in course of the synthesis procedure. Author's personal copy
S. Harish et al. / Journal of Power Sources 214 (2012) 33e39 35 proposed by Gloagen et al. [18]. Catalytic ink is composed of 5 mg of 3.0 catalytic powder dissolved into 0.5 mL of ultra pure water and A fi Ò 0.1 mL of Na on solution (5 wt% in water and aliphatic alcohol 2.5 Before microwave irradiation solution from Aldrich). After 30 min homogenization in an ultra- sonic bath, a volume of 3 mL of catalytic ink is deposited from 2.0 a syringe onto a fresh polished glassy carbon substrate, yielding 25 mg of catalytic powder, i.e. 5 mg of metal on the electrode. The After microwave irradiation solvent is then evaporated in a stream of ultra pure nitrogen at 1.5 room temperature. Cyclic voltammograms and CO stripping measurements are 1.0 carried out using a Model 362 Scanning Potentiostat from Princeton
Applied Research. Absorbance / u.a. 0.5 Infrared spectra were obtained by using the SNIFTIRS (Sub- stractively Normalized Interfacial Fourier Transform Infra Red 0.0 Spectroscopy) method. The working electrode potential was E E modulated between two potential values ( i and f) according to 300 400 500 600 700 800 a square wave signal. The reflectivity was obtained at two electrode potentials (frequency of potential modulation: 0.025 Hz) and Wavelength / nm resulted from the co-addition of 128 interferograms 30 times at 3.0 each potential. Final spectra were calculated as: B 2.5 DR RE RE ¼ f i (1) R RE i 2.0 where Ei is the initial and Ef the final potential of the modulation and DE ¼ Ef Ei ¼ 0.2 V is kept constant. In this case, a negative 1.5 after microwave irradiation peak means the production of species and a positive band indicates the consumption of species at the electrode surface. 1.0 For the in situ IR measurements, the working electrode was before microwave irradiation prepared in the same way as for cyclic voltammetry measurements. Abosrbance / a.u. Abosrbance / 0.5
3. Results and discussion 0.0
The synthesis of catalysts was performed under continuous 300 400 500 600 700 800 microwave irradiation at a power of 1600 W until reaching the Wavelength / nm desired reaction temperature, and then microwave pulses were applied to maintain it. Fig. 1 gives a schema of the microwave Fig. 2. UVevisible spectra recorded before and after 5 min microwave irradiation at sequence and of the temperature change as a function of time in 100 C for Pt and at 130 C for Ru. course of the synthesis procedure. However, the minimal temper- ature at which the metal salt reduction can occur, as well as the synthesis time for the complete reduction of metal salt, is not gives rise to a new peak at 280 nm, which position and shape known a priori. In order to determine the reduction temperature of depends on duration of microwave heating [23]. The higher acti- 3þ Pt and Ru salts, UVevisible spectroscopy was used, since it has vation temperature needed for the reduction of Ru ions than that 4þ previously been shown that UVevisible spectrometry was an effi- for the reduction of Pt ions can be explained by the fact that the 3þ 0 cient tool to follow the Pt and PtRu colloidal formation process redox potential of Ru /Ru (E ¼ 0.84 V) is much lower than that of 4þ E0 ¼ [19,20]. For this purpose, the spectra recorded after the heating of Pt /Pt ( 1.41 V). However, to achieve the coreduction of plat- inum and ruthenium salts for synthesizing PtRu/C catalyst with aH2PtCl6 or RuCl3/ethylene glycol/NaOH reaction mixture by microwave irradiation at different temperature for 5 min were a nominal Pt/Ru ratio of 1/1, it has been necessary to applied at compared to those recorded before microwave irradiation. It was temperature of at least 160 C. It appeared then that the activation þ found that the reduction of P4 ions by ethylene glycol occurred as energy for the formation of the bimetallic material was higher than soon as 100 C, in agreement with previous works [14,19]. Indeed, that for the formation of the pure metal particles. e in Fig. 2A the shape of the UVevisible spectrum before microwave The combination of the UV visible results with TGA measure- irradiation is different to that recorded after microwave irradiation ments (Table 2) indicated that the whole metallic salts were at 100 C for 5 min over the available wavelength range. The shape reduced after 5 min microwave irradiation under the considered of this latter spectrum, displaying strong absorption increasing synthesis conditions. This fast reduction allows also proposing that gradually from ca. 700 nm towards lower wavelengths, is typical of the Pt/Ru ratio in the bimetallic catalysts is likely close to the that of a colloid [21], whereas the absorption feature at low nominal one. wavelengths in the spectrum recorded before the Pt salt reduction reaction corresponds likely to the bottom of the absorption peak Table 2 2 Metal loading on carbon support, particles size and Scherrer length as determined related to ligand-to-metal charge transfer in PtCl6 ion (peak cen- by TGA, TEM and XRD measurements, respectively. tred at ca. 260 nm [22]). In the case of the ruthenium salt reduction d L by ethylene glycol under microwave irradiation (Fig. 2B), changes in Metal loading (TGA)/wt.% (TEM)/nm v (XRD)/nm þ the UVevisible spectra related to the reduction of Ru3 ions by Pt/C 20 3.0 0.5 3.5 ethylene glycol into Ru0 started only to be observed at 130 C. In the Ru/C 18 (<1.5) e Pt Ru /C 20 2.5 0.5 2.8 literature, it is proposed that the formation of Ru nanoparticles 1 1 Author's personal copy
36 S. Harish et al. / Journal of Power Sources 214 (2012) 33e39
The synthesized carbon supported catalysts were examined by (111) TEM. Fig. 3 shows TEM images of Pt (20 wt%)/C (A), Ru (20 wt%)/C Pt/C (B) and Pt1Ru1 (20wt%)/C (C) samples. All samples display (200) (311) (331) a homogeneous repartition of metallic particles on the carbon (222) support. Assuming a spherical shape of the metallic clusters, the mean particle sizes could be evaluated according to the following relation: Pt1Ru1/C P n n d d ¼ i ¼ 1 i i n (2) where ni, di, and n are the number of particles of diameter di, the diameter of particles and the total number of particles, respectively. Ru/C RuOO Ru hcp The determination of the mean catalyst particle size has been made 2 RuO2 (101) (101) for each catalytic powder and is given in Table 2. In the case of the (211) pure ruthenium catalyst (Fig. 3B), the contrast of Ru on carbon is low and the particles have very small diameters, lower than 1.5 nm. The determination of the mean size was difficult to perform. Higher 30 40 50 60 70 80 90 mean particle size was found for the pure Pt/C (ca. 3.0 nm). The Pt1Ru1/C sample displayed an intermediate mean particle size of ca. 2 / ° 2.5 nm between those of both pure metals. q XRD patterns of Pt/C, Ru/C and Pt1Ru1/C are represented in Fig. 4. Fig. 4. XRD diffractograms in 2 range from 30 to 90 obtained with Pt/C, Ru/C and The diffraction pattern was analysed by the method of Pt1Ru1/C catalysts synthesized by the polyol method activated by microwave irradiation. LevenbergeMarquardt, using a Voigt fit by means of a computer refinement program (Fityk free software [24]). All diffraction patterns recorded on Pt-containing catalysts display the typical diffraction peak shift. Then it was possible to determine the alloy diffraction peaks of the fcc structure of platinum, whereas, the XRD composition by XRD [27]; the ruthenium atomic ratio was found to pattern of the Ru displays not well defined broad peaks typical of be ca. 21%. Therefore, it can be proposed that the structure of the amorphous material. The low crystallinity of the Ru/C material is Pt1Ru1/C catalyst consists in a mixture of fcc Pt0.8Ru0.2 nano- convenient with the low particle size as determined by TEM. Small crystallites, of rutile RuO2 nanoclusters and maybe also of hcp Ru diffraction peaks located at ca. 35 and 55 corresponds to crys- nanoclusters. talline RuO (101) and (211) orientations, respectively [25], whereas 2 However, from the determination of the full width at half the diffraction peak at ca. 43 corresponds to metallic Ru hcp (101) maximum (FWHM) and using the Sherrer equation [28] presented orientation. in Eq. (3), the mean size Lv for the Pt/C and PtRu/C fcc crystallites In addition to the typical diffraction peaks of the fcc structure of could be evaluated (Table 2). platinum, the XRD pattern of the Pt1Ru1/C catalyst also display the RuO2 peaks at ca. 35 and ca. 55 as well as that of hcp structure of l ’ L ¼ f (3) Ru at ca 43 . According to the Vegard s law for a true PtRu alloy, the v FWHMcos q value of the cell parameter should decrease when the ruthenium L f content increases. In other words, the diffraction peaks shift where v is the volume-weighted column length, is the shape l towards higher 2q value when ruthenium is alloyed with platinum factor (0.89 for spherical crystallite), the radiation wavelength q [26]. The diffraction peaks of the PtRu/C catalyst are slightly shifted (1.5406 Å), FWHM the full width at half maximum, and the angle towards higher 2q values in comparison to those recorded for the at the maximum intensity. Crystallite sizes of 3.5 nm and 2.8 nm Pt/C material. However, this fact can be related either to PtRu alloy were determined for the Pt/C and Pt1Ru1/C catalysts, respectively. formation or to size effect. But, the apparent particles sizes deter- These values are in very good agreement with the mean particle mined by TEM for both Pt/C and PtRu/C catalysts are very close to size determined by TEM, which seems to indicate that the metallic each other, so that size effect is probably not significant in the slight particles are monocrystalline.
Fig. 3. TEM Photographs of (A) Pt/C, (B) Ru/C and (C) Pt1Ru1/C catalysts synthesized by the polyol method activated by microwave irradiation. Author's personal copy
S. Harish et al. / Journal of Power Sources 214 (2012) 33e39 37
monolayer from a smooth platinum surface [31,32]. Measurements are carried out in a N2-saturated electrolyte at a scan rate of 20 20 mV s 1 between 0.05 V and 1.2 V vs. RHE for Pt/C and 0.05 V and 10 0.8 V vs. RHE for PtRu/C. The latter upper limit potential was used in order to avoid dissolution of the ruthenium from the surface. Typical voltammogram of surface clean Pt/C and PtRu/C were 0 recorded from which active surface areas of ca. 80 m2 g 1 and 2 1 -1 30 m g were found for Pt/C and PtRu/C, respectively. -10 CO stripping measurements were used in order to obtain information on structure, morphology and activity of nanoparticles j / A g / A j -20 [33e36]. For this purpose, CO surface saturation of platinum cata- lysts was performed at 0.1 V vs. RHE for 5 min. Before CO stripping -30 measurements recorded by cyclic voltammetry at a potential scan rate 20 mV s 1 (Fig. 6), CO was removed from the electrolyte bulk -40 by N2 bubbling for 15 min, under potential control at 0.1 V. For the 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Pt/C catalyst (Fig. 6A), the complete disappearance of the current peaks in the hydrogen desorption region (from 0.05 V to 0.4 V vs. E vs. RHE / V RHE) of the first voltammetric cycle shows that the platinum surface is completely blocked by the presence of adsorbed CO. Fig. 5. Cyclic voltammograms of Pt(20 wt.%)/C (solid line) and Pt1Ru1(20 wt.%)/C (dashed line) synthesized by the microwave assisted polyol method, recorded at Then, the oxidation of adsorbed CO into CO2 is responsible for the 1 20 mV s ,20 C, in a N2-saturated 0.5 M H2SO4 electrolyte (T ¼ 20 C). positive current peaks in the potential range from 0.65 V to 1.0 V vs. RHE. In the negative going potential scan, current peaks in the
The active surface areas (ASA) of Pt/C and PtRu/C catalysts were determined from cyclic voltammograms (Fig. 5) by integrating the charge in the hydrogen desorption region corrected from the 350 double layer capacity contribution [29,30], considering a charge of A 210 mC per square centimetre for the desorption of a hydrogen 300 250
200 -1 150
j / A g 100
50
0
-50 0.0 0.2 0.4 0.6 0.8 E vs. RHE / V 140 B 120
100
-1 80
j / A g 60
40
20
0 0 500 1000 1500 2000 2500 3000 3500 time / s
Fig. 7. (A) j(E) polarization curves recorded on Pt(20 wt.%)/C (dashed line) and on 1 Pt1Ru1(20 wt.%)/C (solid line) at 5 mV s in N2-saturated 0.5 M H2SO4 electrolyte containing 1.0 M methanol (T ¼ 20 C). (B) Chronoamperometry curve recorded on
Fig. 6. CO stripping voltammograms of (A) Pt(20 wt.%)/C and (B) Pt1Ru1(20 wt.%)/C Pt1Ru1(20 wt.%)/C in N2-saturated 0.5 M H2SO4 electrolyte containing 1.0 M methanol 1 recorded at 20 mV s ,20 C, in N2-saturated 0.5 M H2SO4 electrolyte. (T ¼ 20 C) at 0.6 V vs. RHE. Author's personal copy
38 S. Harish et al. / Journal of Power Sources 214 (2012) 33e39
0.1 – 0.3 V vs. RHE 0.2 – 0.4 V vs. RHE
0.3 – 0.5 V vs. RHE
0.4 – 0.6 V vs. RHE
0.5 – 0.7 V vs. RHE
0.6 – 0.8 V vs. RHE
1900 2000 2100 2200 2300 2400 wavenumber / cm-1
Fig. 8. SNIFTIR spectra recorded during the electrooxidation of 1.0 M methanol in 0.5 M H2SO4 on Pt1Ru1(20 wt.%)/C. hydrogen adsorption region appeared and the second cyclic vol- C catalyst. The chronoamperometry curve recorded at 0.6 V at tammogram is typical of a clean Pt/C catalyst, which indicates that aPt1Ru1/C catalyst in the presence of 1.0 M MeOH in the support adsorbed CO was totally removed from the platinum surface during electrolyte is given in Fig. 7B. Initially a current drop is observed, the first volatmmetric cycle. From the determination of the charge but it stabilizes to a constant mass current density value (ca. 1 involved for the oxidation of the adsorbed CO, the same active 70 A gmetal) after a short time. 2 1 surface area value as from the Hupd region of ca. 80 m g could be Both CO stripping and methanol electrooxidation experiments determined considering a charge of 420 mC per square centimetre emphasize the well-known role of ruthenium, i.e. allowing the for the desorption of a CO monolayer from a smooth platinum bifunctional mechanism at lower potential than on pure platinum surface, which is an indication of the cleanness of the catalyst [41,42]. Ruthenium brings at lower potentials than platinum the OH surface. With the PtRu catalyst (Fig. 6B), the onset potential of COads species necessary to complete the oxidation of adsorbed CO into oxidation is shifted by ca. 0.2 V towards lower potentials compared CO2, making the bimetallic catalyst more active than Pt/C catalyst with that of Pt/C catalyst, as the result of the presence of ruthenium from 0.35 V to 0.7 V vs. RHE. at the material surface [37]. The active surface area calculated from Fig. 8 shows SNIFTIR spectra recorded between 1900 cm 1 and 1 the charge involved for the oxidation of the CO saturating layer is 2400 cm at the Pt1Ru1/C catalytic surface with 1.0 M methanol. ca. 57 m2 g 1. This value is almost twice higher than that calculated The absorption bands located between 2030 and 2055 cm 1 are from the Hupd region. The discrepancy in the ASA values deter- assigned to linearly bonded CO (COL) [43]. No additional typical CO 1 mined from Hupd region and from CO stripping experiment can be absorption band around 1950 cm is visible, which could be explained by the ability of ruthenium atoms to adsorb CO [38].By assigned to bridge-bonded CO (COB) species [44]. The COL band comparing the active surface area values calculated from Hupd appears in the first potential modulation between 0.1 and 0.3 V vs. region and CO stripping measurement, and assuming that RHE, indicating that methanol chemisorption and dehydrogenation hydrogen does not adsorb on Ru sites and that CO is linearly to form adsorbed CO takes place at very low potentials on Pt1Ru1/C adsorbed on both metal sites, it can be proposed that the Pt/Ru catalyst. The intensity of this band first increases with the increase surface atomic surface composition could be close to the catalyst of potential to reach a maximum in the second and third modula- nominal Pt/Ru atomic ratio (1/1). tions. Then, it starts to decrease from the fourth modulation, for 1 Now, let us focus on methanol oxidation at Pt/C and Pt1Ru1/C which an absorption band located at 2345 cm corresponding to catalysts. The polarisation curves recorded in presence of 1.0 M interfacial CO2 [43,44] starts to be observed. In addition, the COL methanol in the support electrolyte is given in Fig. 7A. The onset absorption band first undergo a red shift as the average potential potential of methanol oxidation is ca. 0.35 V vs RHE at Pt1Ru1/C, i.e. modulation increases, and then, from the third potential modula- 0.2 V before that at platinum catalyst, which is agreement with tion, a blue shift, which are due to a combination of the Stark effect result obtained with PtRu catalysts prepared by other colloidal (red shift) [45] with a coverage-dependent shift of the CO infrared methods [39] under the same experimental conditions. The Pt1Ru1/ band (blue shift) [46]. C catalyst displayed higher catalytic activity towards methanol oxidation from 0.35 V vs. RHE to ca. 0.65 V vs. RHE. For higher 4. Conclusion potentials, the pure platinum supported catalyst becomes the more active one. From this potential, platinum is able to activate water In this contribution, we developed an industrially scalable and is no longer poisoned by adsorbed CO species; therefore this method for the synthesis of Pt/C and PtRu/C catalysts for DMFC material becomes more active for methanol electrooxidation, than application. First, the method is based on a polyol route which uses platinumeruthenium catalyst [40]. The maximum mass current inexpensive solvent (ethylene glycol), does not need the presence 1 density of ca. 200 A gmetal at ca. 0.7 V was achieved with the Pt1Ru1/ of surfactant and is very easy to implement. Second, instead of Author's personal copy
S. Harish et al. / Journal of Power Sources 214 (2012) 33e39 39 using traditional thermal activation, the activation by microwave [16] M. Tsuji, M. Hashimoto, Y. Nishizawa, M. Kubokawa, T. Tsuji, Chem. Eur. J. 11 irradiation allows fast synthesis, lower energy consumption and (2005) 440. [17] W.S. Rasband, Image J, U S. National Institutes of Health, Bethesda, Maryland, uniform heating. The PtRu/C catalyst displayed the nominal metal USA, http://imagej.nih.gov/ij/, 1997e2011. loading, particle size of ca. 2.5e3.0 nm and was composed of a Pt/ [18] F. Gloaguen, N. Andolfatto, R. Durand, P. Ozil, J. Appl. Electrochem. 24 (1994) 863. Ru alloy in interaction with Ru and RuO2 clusters. Such structure [19] Z. Zhou, W. Zhou, S. Wang, G. Wang, L. Jiang, H. Li, G. Sun, Q. Xin, Catal. Today gives a high activity towards CO and methanol electrooxidation to 93-95 (2004) 523. the Pt1Ru1/C catalyst prepared by this method. In order to improve [20] Z. Liu, X.Y. Ling, X. Su, J.Y. Lee, J. Phys. Chem. B 108 (2004) 8234. the synthesis procedure and the catalytic performance of PtRu/C [21] D.N. Furlong, A. Launikonis, W.H.F. Sesse, L.V. Sanders, J. Chem. Soc., Faraday Trans. 1 80 (1984) 571. catalysts prepared by microwave assisted polyol method, the future [22] T. Teranishi, M. Hosoe, T. Tanaka, M. Miyake, J. Phys. Chem. B 103 (1999) 3818. study will focus on the effect of batch temperature and of Pt/Ru [23] R. Harpeness, Z. Peng, X. Liu, G. V-Pol, Y. Koltypin, A. Gedanken, J. Colloid. nominal composition. Interf. Sci. 287 (2005) 678. [24] M. Wojdyr, J. Appl. Cryst 43 (2010) 1126. [25] A. Devadas, S. Baranton, T.W. Napporn, C. Coutanceau, J. Power Sources 196 Acknowledgement (2011) 4044. [26] P. Vogel, H. Britz Bönnemann, J. Rothe, J. Hormes, J. Phys. Chem. B 101 (1997) This work could be realized thanks to the Sandwich PhD 11029. [27] C.A. Angelucci, M. D’Villa Silva, F.C. Nart, Electrochim. Acta 52 (2007) 7293. Fellowship Programme 2011 of the French Embassy in India. [28] P. Scherrer, Nachr. Ges. Wiss. Göttingen, Math.-Phys. Klasse 26 (1918) 98. This work was presented in 6th Asian conference on Electro- [29] C. Coutanceau, M.J. Croissant, T. Napporn, C. Lamy, Electrochim. Acta 46 chemical Power Sources held on Jan 5e8, 2012 at Chennai, Tamil- (2000) 579. nadu, India. [30] C. Grolleau, C. Coutanceau, F. Pierre, J.M. Léger, Electrochim. Acta 53 (2008) 7157. [31] V.S. Bakotzky, Y.B. Vassilyev, Electrochim. Acta 12 (1967) 1323. References [32] N.M. Markovic, B.N. Grgur, P.N. Ross, J. Phys. Chem. B 101 (1997) 5405. [33] F. Maillard, S. Schreier, M. Hanzlik, E.R. Savinova, S. Weinkauf, U. Stimming, [1] W.R. Grove, Phil. Mag. 21 (1842) 417. Phys. Chem. Chem. Phys. 7 (2005) 385. [2] K.H. Choi, H.S. Kim, T.H. Lee, J. Power Sources 75 (1998) 230. [34] M. Arenz, K.J.J. Mayrhofer, V. Stamenkovic, B.B. Blizanac, T. Tomoyuki, [3] C. Coutanceau, A. Rakotondrainibe, A. Lima, E. Garnier, S. Pronier, J.M. Léger, P.N. Ross, N.M. Markovic, J. Am. Chem. Soc. 127 (2005) 6819. C. Lamy, J. Appl. Electrochem. 34 (2004) 61. [35] A. Cuesta, A. Couto, A. Rincón, M. Pérez, A. López Cudero, C. Gutíerrez, [4] A. Caillard, C. Coutanceau, P. Brault, J. Mathias, J.-M. Léger, J. Power Sources J. Electroanal. Chem. 586 (2006) 184. 162 (2006) 66. [36] A. López Cudero, J. Solla Gullón, E. Herrero, A. Aldaz, J.M. Feliu, J. Electroanal. [5] L. Dubau, C. Coutanceau, E. Garnier, J.-M. Léger, C. Lamy, J. Appl. Electrochem. Chem. 644 (2010) 117. 33 (2003) 419. [37] M. Watanabe, S. Motoo, J. Electroanal. Chem. 60 (1975) 275. [6] F. Vigier, C. Coutanceau, A. Perrard, E.M. Belgsir, C. Lamy, J. Appl. Electrochem. [38] L. Dubau, F. Hahn, C. Coutanceau, J.M. Léger, C. Lamy, J. Electroanal. Chem. 34 (2004) 439. 554e555 (2003) 407. [7] H.S. Oh, J.G. Oh, H. Kim, J. Power Sources 183 (2008) 600. [39] C. Coutanceau, S. Brimaud, L. Dubau, C. Lamy, J.-M. Léger, S. Rousseau, [8] J. Guo, G. Sun, S. Shiguo, Y. Shiyou, Y. Weiqian, Q. Jing, Y. Yushan, X. Qin, F. Vigier, Electrochim. Acta 53 (2008) 6865. J. Power Sources 168 (2007) 299. [40] H.A. Gasteiger, N. Markovic, P.N. Ross, E.J. Cairns, J. Electrochem. Soc. 141 [9] F. Fievet, J.P. Lagier, B. Blin, B. Beaudoin, M. Figlarz, Solid State Ionics 32e33 (1994) 1795. (1989) 198. [41] M. Watanabe, S. Motoo, J. Electroanal. Chem. 75 (1960) 267. [10] C. Bock, C. Paquet, M. Couillard, G.A. Botton, B.R. MacDougall, J. Am. Chem. Soc. [42] M. Watanabe, S. Motoo, J. Electroanal. Chem. 75 (1960) 275. 126 (2004) 8028. [43] K. Kunimatsu, J. Electroanal. Chem. 140 (1982) 205. [11] Z. Liu, L.M. Gan, L. Hong, W. Chen, J.Y. Lee, J. Power Sources 139 (2005) 73. [44] B. Beden, F. Hahn, S. Juanto, C. Lamy, J.-M. Léger, J. Electroanal. Chem. 225 [12] C. Grolleau, C. Coutanceau, F. Pierre, J.M. Leger, J. Power Sources 195 (2010) 1569. (1987) 215. [13] B.M. Babi c, Lj. M. Vra car, V. Radmilovi c, N.V. Krstaji c, Electrochim. Acta 51 [45] B. Beden, C. Lamy, A. Bewick, K. Kunimatsu, J. Electroanal. Chem. 121 (1981) (2006) 3820. 343. [14] E. Lebègue, S. Baranton, C. Coutanceau, J. Power Sources 196 (2011) 920. [46] C. Rice, Y.Y. Tong, E. Oldfield, A. Wieckowski, F. Hahn, F. Gloaguen, J.-M. Léger, [15] S.A. Galema, Chem. Soc. Rev. 26 (1997) 233. C. Lamy, J. Phys. Chem. B 104 (2000) 5803. This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright Author's personal copy
Electrochimica Acta 56 (2011) 5717–5721
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Interaction between gold (III) chloride and potassium hexacyanoferrate (II/III)—Does it lead to gold analogue of Prussian blue?
S. Harish, James Joseph ∗, K.L.N. Phani
Electrodics and Electrocatalysis Division, CSIR-Central Electrochemical Research Institute, Karaikudi 630006, Tamilnadu, India article info abstract
Article history: Prussian blue analogues are a class of compounds formed by the reaction between metal salt and potas- Received 29 December 2010 sium hexacyanoferrate (II/III). In our earlier report, the formation of Au@Prussian blue nano-composite Received in revised form 12 April 2011 was noticed on potential cycling the glassy carbon electrode in a medium containing gold (III) chloride Accepted 13 April 2011 and potassium hexacyanoferrate (III). Hence in this work, the formation of gold hexacyanoferrate was Available online 27 April 2011 attempted by a simple chemical reaction. The reaction of gold (III) chloride with potassium hexacyano- ferrate (II/III) was examined by UV–Vis spectroscopy and found that there is no redox reaction between Keywords: gold (III) chloride and potassium hexacyanoferrate (III). However, the redox reaction occurs between Prussian blue Interaction gold (III) chloride and potassium hexacyanoferrate (II) leading to the formation of charge transfer band Potassium ferrocyanide and the conversion of hexacyanoferrate (II) to hexacyanoferrate (III) was evidenced by the emergence of Potassium ferricyanide new absorption peaks in UV–Vis spectra. The oxidation state of gold in Au–Fe complex was found to be Gold (III) chloride +1 from X-ray photoelectron spectroscopy. The stability of the Au–Fe complex was also studied by cyclic voltammetry. Cyclic voltammetric results indicated the presence of high spin iron in Au–Fe complex.
Hence ‘as formed’ Au complex may be KFex[Au(CN)2]y. The results revealed that the formation of gold hexacyanoferrate was not feasible by simple chemical or electrochemical reaction in contrast to other Prussian blue analogues. © 2011 Elsevier Ltd. All rights reserved.
1. Introduction However to the best of our knowledge, interaction between gold (III) chloride and either potassium ferrocyanide or ferricyanide has Prussian blue (PB) was the first synthetic co-ordination com- not been studied in detail. Liu et al. [12] reported the formation pound having history for more than three centuries. This compound of Pt analogue on GC electrode by reacting potassium ferricyanide is still considered ‘hot’ in research because of its interesting ionic, with chloroplatinic acid. Pt analogue formed on GC was character- electronic and electrochemical properties [1–4]. Various PB ana- ized by the surface redox process occuring at 0.77 V (vs saturated logues are formed by precipitation of ferrocyanide ion with the calomel electrode) due to redox transition of low spin Fe (II/III). In transition metal ions like Co(II), Ni(II), Zn(II), Cd (II), Cu(II), Ag(I) this work, we have studied the interaction between the gold (III) etc.,[5]. These PB analogues were also formed as thin films by elec- chloride and potassium ferrocyanide or ferricyanide in the absence trochemical potential cycling of electrodes in a medium containing of electrochemical cycling. We found the reaction of gold (III) chlo- potassium ferricyanide and corresponding metal salts [6]. Among ride with either potassium ferricyanide or ferrocyanide on GC led group IB metals, Cu and Ag analogues could be formed electrochem- to the formation of PB in the former and a different charge trans- ically on solid electrodes [7,8]. However Au analogue of PB could fer complex involving Fe (II)/(III) and Au(I) in the later case. We not be formed on glassy carbon (GC) electrode by electrochemi- present here, the different mechanistic pathways of the chemical cal cycling in medium containing potassium ferricyanide and gold reaction between gold (III) chloride and potassium ferrocyanide chloride. Instead it leads to the formation of Au-PB nanocomposite. or ferricyanide in detail. The two pathways of reaction were deci- The mechanism leading to the formation of Au-PB nanocompos- phered at using Ultraviolet–Visible (UV–Vis), X-ray photo electron ite was studied in great detail by Senthil et al. [9]. Au nanoparticles spectroscopy (XPS) and cyclic voltammetry. have been used as catalyst in the electron transfer reaction between potassium ferrocyanide and thiosulfate [10] or borohydride [11]. 2. Experimental
HAuCl4 (Sigma-Aldrich), K3 [Fe (CN) 6], K4 [Fe (CN) 6] (E-Merck) ∗ Corresponding author. Tel.: +91 4565 227555; fax: +91 4565 227779. and NaBH4 (Sigma-Aldrich) Analar grade were purchased. UV–Vis E-mail address: [email protected] (J. Joseph). absorption spectra were collected using Cary 500 scan UV–Vis
0013-4686/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.electacta.2011.04.044 Author's personal copy
5718 S. Harish et al. / Electrochimica Acta 56 (2011) 5717–5721
Table 1
Observed max value for each reactant from UV–Vis spectra.
−3 −3 Reactants (1 × 10 mol dm ) max (nm)
HAuCl4 306 K4[Fe(CN)6] 332 K3[Fe(CN)6] 260, 303, 325, 421 HAuCl4 +K3[Fe(CN)6] 306, 414 AuCl4 +K4[Fe(CN)6] 305, 414, 692
vation is consistent with the results of Zhai et al. [13] in that they observed no change in the absorption band after the addi- tion of ferricyanide to gold chloride, indicating that no bond was formed between gold chloride and ferricyanide. Freund et al. [10] studied the redox reaction between ferricyanide and thiosulfate catalysed by colloidal gold whereas Romero et al. [11] demon- strated that the gold nanoparticles act as an efficient catalyst in the reduction of ferricyanide ion to ferrocyanide ion by sodium Fig. 1. UV–Vis spectra of 1 mM HAuCl4,K4[Fe(CN)6], K3[Fe(CN)6], K3[Fe(CN)6] borohydride. + HAuCl4. Inset: K4 [Fe (CN) 6] + HAuCl4. Interestingly, in contrast to ferricyanide, we observed a redox reaction with intense green color formation when potassium fer- spectrophotometer with incident light normal to the 1 cm path rocyanide and gold chloride were mixed in equi-molar ratio. The length quartz cell. XPS was done by forming a thin film of charge UV–Vis spectra were marked by the absence of absorption cor- transfer complex by drop casting 10 l of 10 mM each of potassium responding to ferrocyanide ion. Instead, we noticed absorption ferrocyanide and gold chloride on 1 cm × 1 cm non-conducting corresponding to ferricyanide ion. When the oxidation state of Fe glass slide and analysed using the Multilab 2000 (Thermoscientific, changes from 2+ to 3+, it is natural to expect concomitant changes UK) photoelectron spectrometer fitted with a twin anode X-ray in the oxidation state of Au in gold chloride but there was no source. 10 l of 0.1 M NaBH4 was dropped on the formed thin film absorption corresponding to the characteristic surface plasmon for reducing gold ions. The Au 4f core-level photoemission spectra resonance band for Aunano. Our attempts to reduce gold chlo- were recorded using the MgK␣ (1253.6 eV) source. Deconvolution ride in the presence of a stabiliser like sodium dodecylsulfate also of XPS was done using XPSpeak 4.1 software. Thermo-Nexus 670 did not indicate the formation of Aunano. These observations indi- model spectrometer with ATR attachment was used for fourier cate that Au3+ in gold chloride may undergo reduction to Au+. transform infra red (FT-IR) measurements and spectral range was Incidentially, the formation of Au+ from Au3+ on the addition limited to 2300–1800 cm−1. Sample preparation for FT-IR mea- of potassium ferrocyanide is known to occur [14]. In addition, a surement was similar to that of XPS analysis. Glassy carbon (GC) charge transfer band was also observed at 690 nm which is com- electrode of 3 mm diameter was used in this study and GC elec- mon in PB-like compounds [15]. Fig. 2 shows the FT-IR spectrum − trode was polished using 4/0 emery sheet and sonicated for 5 min of Au–Fe complex, stretching vibration at 2168 cm 1 confirms the before each experiment. 2.5 l of 10 mM HAuCl4 and 10 mM K4 presence of cyanide ligand. Cyanide stretching vibration shifts to [Fe(CN)6]/K3 [Fe(CN)6] were placed on a GC electrode, the mixture higher wave numbers, when compared to the that in KAu(CN)2 − of the reactants forms the charge transfer complex. The reactant (2140 cm 1). A similar shift was observed in PB type compounds mixture was allowed to dryness on the surface of the GC electrode [16]. for about 10 h and then washed with Millipore water to remove unreacted reactants, if any. An aliquot of 5 l of 0.1 M NaBH4 was dropped on the formed thin film on the GC electrode for reducing gold ions. Cyclic voltammetric experiments were done on a BAS- 100B electrochemical system and all the potentials mentioned in this text are against Ag/AgCl (3 M NaCl) reference electrode, unless otherwise stated. The solutions were freshly prepared every time using Millipore water (18.2 M cm).
3. Results
3.1. UV–Vis and FT-IR spectral studies
The electronic spectra of 1 mM gold chloride, potassium ferrocyanide, potassium ferricyanide, a mixture of potassium fer- ricyanide and gold chloride are shown in Fig. 1. The spectrum recorded in a mixture of potassium ferrocyanide and gold chloride is given in the inset of Fig. 1. The absorption wavelength (max) values for the solutions are tabulated in Table 1. It is clear from Fig. 1 and Table 1 that there is no significant change in the absorp- tion wavelength when potassium ferricyanide and gold chloride are mixed together. Even after keeping the mixture for a few hours, the solution did not show any visible color change indicating the absence of interaction between two anionic reactants. This obser- Fig. 2. FT-IR spectra of Au–Fe complex. Author's personal copy
S. Harish et al. / Electrochimica Acta 56 (2011) 5717–5721 5719
Fig. 3. XPS of Au–Fe complex (A) before and (B) after NaBH4 reduction.
3.2. XPS studies
We have formed films of the charge transfer complex on a plain glass plate by drop casting a mixture of 10 mM potassium ferro- cyanide and gold chloride and subsequent drying. XPS of this film showed Au 4f5/2 and 4f7/2 peaks at 85.92, 89.50, 93.07 eV (Fig. 3A) + − which correspond to Au in [Au (CN)2] [17–21]. Relative shift in the binding energy values of Au+ with reference to those of Au0 was found to vary between 1.2 and 2.3 eV [19] and the shift was found to be consistent with the substrate interaction [22]. This observed shift [17–21] may be attributed to the above reasons and also to the influence of the chemical environment. In a control experiment, the complex is reduced using NaBH4 and analysed for XPS. It showed a peak shift to lower binding energy values of 84.17 and 87.84 eV corresponding to metallic gold [23] (Fig. 3B).
3.3. Cyclic voltammetric studies
To characterize the interaction between potassium ferrocyanide Fig. 4. Cyclic voltammogram shows the redox behavior of Au–Fe complex in 0.1 M −1 and gold chloride, we have followed an unconventional approach KNO3 at the scan rate of 50 mV s . of forming a film by drop casting the mixture on GC as described in the experimental section. This electrode showed a voltammetric response (Fig. 4) that clearly exhibits two redox processes; one at 0.29 V and the other at 0.63 V. The peak currents of both the pro- cesses decrease with increasing cycle number. However we obtain a stable redox at 0.15 V when the potential range is limited to 0.0–0.6 V (Fig. 5). Similar improvement in the stability was earlier achieved in the case of silver hexacyanoferrate modified electrode when the potential cycling was limited to 0.9 V [24]. The redox peak obeys characteristics of an ideal surface reaction as seen from the linear relation between the plot of peak current and scan rate [inset in Fig. 5]. Gold ion in Au–Fe complex formed on GC electrode was reduced using NaBH4 (as followed by XPS analysis) to metallic gold. NaBH4-treated electrode in 0.5 M H2SO4 showed voltammet- ric peaks characteristic of gold oxide formation at 750 mV and gold oxide reduction at 430 mV versus Hg/Hg2SO4 reference electrode (Fig. 6). A similar modification of GC surface by drop casting a mixture of potassium ferricyanide and gold chloride result in the forma- tion of PB on the surface of electrode as seen from the two sets of redox process at 0.16 V and 0.837 V (Fig. 7). Both the redox couples remain stable on potential cycling. The origin of these Fig. 5. Cyclic voltammogram shows the redox behavior of Au–Fe complex in 0.1 M −1 redox processes in the case of films formed on the GC surface from KNO3 at the scan rate of 20 mV s . Inset figure: plot of peak current and scan rate. Author's personal copy
5720 S. Harish et al. / Electrochimica Acta 56 (2011) 5717–5721
+ Fig. 6. Cyclic voltammogram of Au–Fe complex after reduction with NaBH4 in 0.5 M Fig. 8. Cyclic voltammogram shows the K ion dependence of the Au–Fe complex −1 −1 H2SO4 at the scan rate of 50 mV s . at the scan rate of 10 mV s .
mixtures of gold chloride with ferricyanide/ferrocyanide is dis- ified electrode showed two sets of redox peaks; one at 0.136 V and cussed below. the other at 0.837 V typical of PB formation. It can now be stated that on mixing gold (III) chloride and potassium hexacyanoferrate (II/III) 4. Discussion on GC, the following processes take place: step 1: increase in acidity due to an increase in [H+] during solvent evaporation (the initial pH Various PB analogues can be electrochemically prepared as thin of the mixture of 1 mM gold chloride and potassium ferricyanide ∼ films on conducting substrates (like GC, indium tin oxide, Pt) by or ferrocyanide is 3.2); step 2: decomposition of potassium ferri- potential cycling the electrode in a medium containing the metal cyanide ion to free ferric ion under low pH conditions [5] [a similar salt and potassium ferricyanide. However, there is no report on decomposition of potassium ferricyanide to ferric ion during gold the formation of gold hexacyanoferrate by either electrochemical hydroxide formation caused by the interfacial acidity was reported or chemical means. Electrochemical cycling of the GC surface in a by Senthil et al. [9]]; step 3: formation of PB takes place dur- mixture containing gold chloride and potassium ferricyanide was ing electrochemical cycling through a complex formation between ferricyanide ion with free ferric ion. On the other hand, a redox found to lead to the formation of Aunano incorporated PB [9]. Our present experiments suggest that on similar cycling in a mixture reaction takes place when potassium ferrocyanide is mixed with gold chloride. The standard oxidation potential of potassium ferro- of potassium ferrocyanide and gold chloride also resulted in the − cyanide is −0.37 V [25] and the reduction potential of [AuCl4] to formation Au @ PB nanocomposite (figure not shown). The UV–Vis − spectra [Fig. 1] do not present any features characteristic of a reac- [AuCl2] is 0.926 V [26]. Hence the reaction between the potassium spontaneous tion between ferricyanide and gold chloride. However, when the ferrocyanide and gold chloride is whereas this is not in same mixture is drop cast on GC and dried to form a film, the mod- the case of potassium ferricyanide. Spontaneous reaction between potassium ferrocyanide and gold chloride is observed in the UV–Vis spectra showing the conversion of ferrocyanide ion to ferricyanide ion and Au3+, in turn undergoing reduction to either Au+ or Au0. Although the oxidation potential of ferrocyanide is sufficient to 3+ 0 E 3+ 0 0 reduce Au to Au ( Au /Au = 1.002 V), we have not observed Au formation from the UV–Vis spectra and XPS analyses. The difference between the reaction of gold chloride with potassium ferricyanide or ferrocyanide is due to the associated redox reaction in the later case. It may be summarized that the following steps are involved in the reaction between the potassium ferrocyanide and gold chloride, step 1: redox reaction between potassium ferrocyanide and gold chloride leading to the formation of ferricyanide ion and Au+. Step 2: concomitant decomposition of ferrocyanide/ferricyanide ion to free ferrous/ferric ion at high acidity conditions. The XPS spectra of the film prepared by drop casting the mixture on the plain glass − plate indicate the formation of [Au(CN)2] . Au (I), Ag (I) and Cu (I) are known to form stable cyano-complexes [5] in which Au has more affinity towards CN−ion and hence the extraction of Au was performed by cyanidation, known as ‘MacArthur–Forrest process’ − [27]. In addition to this, the formation constant of [Au(CN)2] in the cyanide medium is also very high (ca.1038) compared to other metals [5]. Hence, the gold ions might extract ligands from Fig. 7. Cyclic voltammogram of Prussian blue formed from ferricyanide and gold the potassium ferrocyanide forming the cyano-complex of gold. −1 chloride in 0.1 M KNO3 at the scan rate of 50 mV s . Similar ligand exchange isomerism between iron and chromium Author's personal copy
S. Harish et al. / Electrochimica Acta 56 (2011) 5717–5721 5721 in chromium hexacynanoferrate during electrochemical cycling chloride and potassium hexacyanoferrate (II/III) in contrast to other resulted in the formation of iron hexacyanochromate as reported PB analogues. by Dostal et al. [28]. To explain the stable voltammetric response shown in Fig. 5, we propose the formation of a complex between Acknowledgements 3+ 2+ − free Fe /Fe and [Au(CN)2] . Similar dicyanoaurate complex with metal ion/complexes were sparsely reported [29–32]. Dong et al. We thank the Department of Science and Technology, India for [33] synthesized Mn[Au(CN)2]2(H2O)2 n and KFe[Au(CN)2]3 n financial assistance through a grant-in-aid programme [SR/S1/PC- − { } { } from [Au(CN)2] . The voltammetric response of the GC modified 22/2007]. S. Harish thanks Council of Scientific and Industrial with 1:1 molar ratio of gold chloride and potassium ferrocyanide Research, New Delhi for the award of a senior research fellowship. showed a stable redox couple at 0.17 V. The redox peak potential Authors also thank the Central Instrumentation Facility for using matches with the high spin Fe2+/3+ transitions in the complex. Based XPS and FT-IR analysis. on our results and those of Dong et al. [33], we propose that the probable complex formed is of the general formula KFex[Au(CN)2]y. References As expected from this formula, the cyanide stretching vibration is comparable with that of Dong et al. [33], modified electrode is [1] S.K. Ritter, Chem. Eng. News 83 (2005) 32. unstable when potential exceeds 0.6 V and the redox peak depends [2] K. Itaya, I. Uchida, V.D. Neff, Acc. Chem. Res. 19 (1986) 162. [3] N.R.D. Tacconi, K. Rajeshwar, R.O. Lezna, Chem. Mater. 15 (2003) 3046. + on K ion concentration in the electrolyte. The cyanide stretching [4] K.R. Dunbar, R.A. Heintz, Chemistry of Transition Metal Cyanide Compounds: vibration observed from the FT-IR spectra is 2168 cm−1 and the Modern Perspectives, Progress in Inorganic Chemistry, 45, John Wiley & Sons, reported value for the same type of complex is 2154 cm−1 [33]. Inc., Hoboken, NJ, USA, 2007. [5] A.G. Sharpe, The Chemistry of Cyano Complexes of the Transition Metals, Aca- This small variation observed in the wave number may be due to demic Press, New York, 1976. the difference in the stoichiometry of the complex. The instabil- [6] V.D. Neff, J. Electrochem. Soc. 125 (1978) 886. ity of the cyclic voltammetric response when the potential cycling [7] L.M. Siperko, T.J. Kuwana, J. Electrochem. Soc. 130 (1983) 396. [8] S.B. Moon, A. Xidis, V.D. Neff, J. Phys. Chem. 97 (1993) 1634. is extended beyond 0.6 V is probably due to the conversion of [9] S. Senthil Kumar, J. Joseph, K.L.N. Phani, Chem. Mater. 19 (2007) 4722. − − [Au(CN)2] /[Au(CN)4] couple that occurs between 0.53 and 0.59 V [10] P.L. Freund, M. Spiro, J. Phys. Chem. 89 (1985) 1074. [34], requiring additional cyanide ligands from the medium. The [11] S.C. Romero, J.P. Juste, P. Hervées, L.M.L. Marzáan, P. Mulvaney, Langmuir 26 (2010) 1271. absence of the cyanide ligand in the supporting electrolyte 0.1 M [12] S. Liu, H. Li, M. Jiang, P. Li, J. Electroanal. Chem. 426 (1997) 27. KNO3 might cause destabilization of the film during cycling. The [13] J. Zhai, Y. Zhai, S. Dong, Colloids Surfaces A: Physicochem. Eng. Aspects 335 K+ ion dependence is known for various PB analogues and was (2009) 207. [14] O.N. Vrublevskaya, T.N. Vorobyova, H.K. Lee, S.B. Koo, Trans. Inst. Met. Finish. found that the voltammetric mid-peak potential would decrease 85 (2007) 254. with decreasing potassium ion concentration for both the oxidation [15] K. Itaya, I. Uchida, Inorg. Chem. 25 (1986) 389. and reduction reactions [2]. The K+ ion dependence of the formed [16] S.F.A. Kettle, E. Diana, E.M.C. Marchese, E. Boccaleri, G. Croce, T. Sheng, P.L. complex is shown in Fig. 8 where redox peaks shifts in the positive Stanghellini, Eur. J. Inorg. Chem. (2010) 3920. [17] R. Cook, E.A. Crathorne, A.J. Monhemius, D.L. Perry, Hydrometallurgy 22 (1989) + direction when K ion concentration increases from 0.01 M to 1 M 171. and the mid-peak potential shows a near-Nernstian shift. These [18] A. Warshawsky, N. Kahana, V. Kampel, I. Rogachev, R.M. Kautzmann, J.L. Cortina, are the initial evidences in for supporting the proposed formula C.H. Sampaio, Macromol. Mater. Eng. 286 (2001) 285. [19] G.J. McDougall, R.D. Hancock, M.J. Nicol, O.L. Wellington, R.G. Copperthwaite, and further structural investigations of the complex are underway J.S. Afr, Inst. Min. Metall. 80 (1980) 344. in our laboratory. [20] C. Klauber, Langmuir 7 (1991) 2153. [21] R. Cervini, R.J. Fleming, B.J. Kennedy, K.S. Murray, J. Mater. Chem. 4 (1994) 87. [22] N. SëSuèzer, S. Ertasë, O.Y. Kumser, O.Y. Ataman, Appl. Spectrosc. 51 (1997) 5. Conclusions 1537. [23] D. Briggs, M.P. Seah, Practical Surface Analysis – Auger and X-Ray Photoelectron The interaction between gold chloride and potassium ferri- Spectroscopy, 2nd ed., Wiley Interscience, 1990. [24] U. Schroöder, F. Scholz, Inorg. Chem. 39 (2000) 1006. cyanide or ferrocyanide was studied using UV–Vis spectroscopy, [25] P.A. Rock, J. Phys. Chem. 70 (1966) 576. FT-IR, XPS and cyclic voltammeteric analysis. The studies showed [26] Z. Guo, Y. Zhang, A. Xu, M. Wang, L. Huang, K. Xu, N. Gu, J. Phys. Chem. C 112 no reaction between potassium ferricyanide and gold chloride but (2008) 12638. [27] J.O. Marsden, C.I. House, Chemistry of Gold Extraction, in: Society for Mining a redox reaction taking place when potassium ferrocyanide reacts Metallurgy, and Exploration, 2nd ed., 2006. with gold chloride. This redox reaction involves the conversion of [28] A. Dostal, U. Schroeder, F. Scholz, Inorg. Chem. 34 (1995) 1711. ferrocyanide ion to ferricyanide ion and Au (III) to Au (I) result- [29] J. Lefebvre, D. Chartrand, D.B. Leznoff, Polyhedron 26 (2007) 2189. [30] J. Lefebvre, F. Callaghan, M.J. Katz, J.E. Sonier, D.B. Leznoff, Chem. Eur. J. 12 (2006) x y ing in the formation of KFe [Au(CN)2] , gold having more affinity 6748. to attract cyanide ligand. The Au (I) complex formation was con- [31] B.F. Hoskins, R. Robson, N.V.Y. Scarlett, Angew. Chem., Int. Ed. 34 (1995) firmed from the XPS analysis and the presence of high spin iron 1203. in the outer sphere was also shown in cyclic voltammogram. The [32] M.J. Katz, T. Ramnial, H. Yu, D.B. Leznoff, J. Am. Chem. Soc. 130 (2008) 10663. [33] W. Dong, L.N. Zhu, Y.Q. Sun, M. Liang, Z.Q. Liu, D.Z. Liao, Z.H. Jiang, S.P. Yan, P. results show that the formation of “gold hexacyanoferrate” is not Cheng, Chem. Commun. (2003) 2544. possible either by chemical or electrochemical reactions of gold (III) [34] A.E. Remick, J. Am. Chem. Soc. 69 (1947) 94. J Appl Electrochem (2008) 38:1583–1588 DOI 10.1007/s10800-008-9609-0
ORIGINAL PAPER
PEDOT/Palladium composite material: synthesis, characterization and application to simultaneous determination of dopamine and uric acid
S. Harish Æ J. Mathiyarasu Æ K. L. N. Phani Æ V. Yegnaraman
Received: 17 December 2007 / Accepted: 16 May 2008 / Published online: 3 June 2008 Ó Springer Science+Business Media B.V. 2008
Abstract Palladium (Pd) incorporated poly (3,4-ethyl- materials have a wide range of applications in the field of enedioxythiophene) (PEDOT) films were synthesized optical, electronic, electro-chromic devices, and sensors through an electrochemical route and characterized using etc. Among the reported conducting polymer materials, scanning electron microscopy (SEM) and atomic force poly (3,4-ethylenedioxythiophene) (PEDOT) is considered microscopy (AFM). The electrochemical study showed to be a promising candidate for its regioregular polymeri- catalytic oxidation of dopamine (DA) with optimum zation, low bandgap, stability and optical transparency loading of Pd. DA and uric acid (UA) were detected using [2–4]. The recent technological interests are in the syn- differential pulse voltammetry (DPV). In the presence of thesis of conducting polymers incorporated with metal ascorbic acid (AA), DA-AA showed peak potential sepa- nanoparticles for varied applications. Conducting polymers ration of 0.19 V while 0.32 V between UA-AA on Pd- are widely employed as support materials for dispersing the incorporated PEDOT. These peak separations are large metal particles and the resultant composite materials pos- enough for sensing DA and UA in the presence of AA. DA sess improved catalytic efficiency [5, 6]. The latter were and UA exhibited linear calibration plots and the minimum widely used in the applications for oxidation of small detection limits are 0.5 and 7 lM respectively. On Pd- organic molecules, dioxygen reduction etc [7–11]. PEDOT, the reversibility of DA oxidation was found to Metal nanoparticle-incorporated PEDOT composite increase compared to bare glassy carbon electrode (GCE) materials have been widely reported. Tsakova et al studied and PEDOT modified GCE. Fouling effects were also the crystallization of copper (Cu), Pd and bimetallic found to be minimal making Pd-PEDOT composite suit- (Cu–Pd) in PEDOT matrix which was used for electrore- able for electroanalysis. duction of nitrate ions in neutral solutions [12–16]. Current-sensing atomic force microscopy showed that Keywords Voltammetry Dopamine Uric acid PEDOT with its sulfur atoms (having better aligned sur- Palladium PEDOT face) form strong bonding with platinum (Pt), gold (Au) and silver (Ag) nanoparticles [17]. In lithium ion batteries,
PEDOT incorporated with LiCoO2/VS2 composite material 1 Introduction was used as a cathode material. These were reported to have good thermal and chemical stability, fast electro- Conducting polymers are emerging as intelligent materials chemical switching and high electrical conductivity in the [1] and they are in the forefront of research in the synthetic p-doped state [18, 19]. chemistry of functionalized pi-conjugated systems. These In our recent communications, we reported the utility of Au incorporated PEDOT in the sensing of DA and UA in the presence of excess AA [20–24] wherein PEDOT showed S. Harish (&) J. Mathiyarasu K. L. N. Phani larger voltammetric peak separations. In continuation of this V. Yegnaraman effort, we have attempted the synthesis of Pd incorporated Electrodics and Electrocatalysis Division, Central Electrochemical Research Institute, Karaikudi 630 006, India PEDOT composite material for electroanalysis. Further, Pd e-mail: [email protected] nanoparticles are reported to be catalytic for several organic 123 1584 J Appl Electrochem (2008) 38:1583–1588 reactions such as Heck, Suzuki, Kumuda etc., [25–27] and 12 used as a material for detection of gaseous hydrogen [28], 10
cholesterol [29], hydrazine [30] and dissolved oxygen [31]. -2 The electrochemically synthesized Pd incorporated 8
PEDOT matrix is characterized using surface analytical 6 techniques. Besides, optimization of Pd loading into the polymeric films, their application to sensing of DA and UA 4 in the presence of excess AA are demonstrated. 2 > 0 Current density / mA cm < 2 Experimental -2
-4 2.1 Materials -0.5 0.0 0.5 1.0 1.5 Potential Vs Ag wire / V
3,4-Ethylenedioxythiophene (EDOT, Baytron M) was a gift Fig. 1 Cyclic voltammogram showing the polymerization of 0.01 M sample provided by Bayer AG (Germany). PdCl2 (E-Merck), EDOT in 0.1 M TBAPC on GCE from acetonitrile solution; Scan -1 Dopamine (Acros), Ascorbic acid (E-Merck), Uric acid rate: 100 mV s (E-Merck), potassium dihydrogen phosphate (E-Merck), 1.6 V vs. Ag wire (reference electrode) and the scan rate sodium hydroxide (E-Merck) were used as received. used was 0.1 V s-1. From the cyclic voltammogram, it was The aqueous solutions were prepared using Milli-Q observed that EDOT oxidation starts at *1.4 V and further water (18.3 MX) (Millipore). A standard 3-electrode con- cycling facilitates PEDOT film formation on the electrode figuration consisting of GCE (/ 3 mm, BAS, Inc.) as surface. In order to obtain a thin film, PEDOT was allowed working electrode and Pt foil as auxiliary electrode and Ag to grow on the GCE surface for three successive scans. wire/Ag-AgCl reference electrodes were employed in This was seen from the increasing anodic and cathodic electrochemical experiments. peak current densities at *0.118 V and *-0.18 V during the forward and reverse scan, respectively. 2.2 Instrumentation Pd incorporated PEDOT composite films were synthe- sized as follows: PEDOT films initially synthesized from Electrochemical experiments were carried out using a po- acetonitrile solution were kept at a potential of -0.1 V vs. tentiostat/galvanostat Autolab PGSTAT-30 (Eco-Chemie MSE (Mercury/Mercurous sulfate electrode) for about 360 s B.V., The Netherlands) at room temperature 25 ± 1 °C. To in 0.5 M sulphuric acid containing palladium chloride. At record the DPV, the following input parameters were used: this potential, the Pd2+ ions were reduced to Pdo and incor- scan rate: 0.012 mV s-1, sample-width: 17 ms, pulse- porated into the polymer matrix. Figure 2 shows the redox amplitude: 25 mV, pulse-width (modulation time): 50 ms, pulse-period (interval): 500 ms and rest-time: 5 s. Peak 20 18 16 currents were determined after subtraction of a manually 14 -2 12 added baseline. PEDOT/Pd-PEDOT films coated on 15 10 -2 Indium doped Tin Oxide (ITO) glass substrates (Donnelly 8 Corp., USA) were characterized by AFM (Molecular 6 10 4 Imaging, USA) using Au coated SiN3 cantilevers (Force 2 Current density / mA cm constant 3 N/W). SEM measurements were made using 0 -2 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 Hitachi Model S-3000H with 10 kV (acceleration voltage). 5 Potential Vs MSE / V
Current density/ mAcm 0 3 Results and discussions
-5 3.1 Synthesis of PEDOT/Pd-PEDOT composites -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 Potential Vs MSE/ V Figure 1 shows the electrochemical synthesis of PEDOT Fig. 2 Cyclic voltammogram showing the redox behaviour of Pd/ from acetonitrile solution containing 0.01 M EDOT and -1 PEDOT on GCE in 0.5 M H2SO4 solution; Scan rate: 100 mV s 0.1 M tetrabutylammonium perchlorate (TBAPC) as sup- (inset: Cyclic voltammogram shows the redox behaviour of PEDOT porting electrolyte. The potential window was -0.9 to on GCE in 0.5 M H2SO4 solution) 123 J Appl Electrochem (2008) 38:1583–1588 1585 behaviour of Pd incorporated PEDOT composite film in polymer network has a highly porous structure that can 0.5 M sulphuric acid and the inset shows the redox behaviour easily entrap the foreign material yielding a composite. The of PEDOT alone. The voltammetric features clearly show the image of PEDOT after Pd incorporation showed that the redox characteristic of Pd that confirms its incorporation in structure and morphology of the polymer film changed the polymer matrix whereas it is absent in PEDOT alone. The significantly. The regions where Pd is deposited clearly hydrogen adsorption–desorption regions were also observed showed discrete areas of high contrast, suggesting the in this potential region. Further, the increase in Pd content in presence of Pd due to its conductivity difference. As can be the polymer film is reflected by an increase in the reduction seen from Fig. 3b, the fibrils containing Pd deposit retain a current as shown in the voltammograms. high porosity fibrillar network structure.
3.2 Scanning electron microscopy 3.3 Atomic force microscopy
Figure 3 shows the SEM images of PEDOT and Pd-PE- Figure 4 shows AFM images of PEDOT and Pd-PEDOT DOT composite films electrodeposited on an ITO glass films electrochemically deposited on an ITO glass. Since the substrate. PEDOT (Fig. 3a) deposited from the non-aque- SEM figures do not show high contrast images, AFM anal- ous medium showed a uniform by sized fibrillar network ysis of the composite material was undertaken in this study. structure with a fiber dimension of *20 nm. These struc- The corresponding 3D analyses of the films are also shown in tures were not obtained when synthezised from a Fig. 4. The PEDOT image (Fig. 4a) of 2 9 2 lm shows a microemulsion or aqueous medium [14–16, 33]. The fibrillar network as seen in the SEM images. The PEDOT polymer is arranged in an array type structure and resulting in a network fibrillar structure. From the 3D analysis, it can be observed that the porous structure of the polymer film and the polymer fibril sizes are uniform for the area analyzed. Esti- mation of surface roughness of the polymer film was carried
out and the mean value of roughness (Ra) was calculated as the deviations in height from the profile mean value XN Ra ¼ 1=N jjZi Z ð1Þ i¼1 where Z, is the sum of all height values divided by the number of data points (N) in the profile. The mean roughness value, estimated from these images using Eq. 1 is 4.6 nm. Pd-incorporated composite polymer film (Fig. 4b) clearly shows morphological feature different from that of the polymer film alone. Pd incorporation results in the formation of a composite as is evidenced by the 3D image. Further, it is interesting to note that the mean surface roughness calculated for the composite film shows a value of 4.9 nm. This indicates that the surface porosity of the film does not change much upon incorporation of Pd.
3.4 Electrocatalytic oxidation of composite film
Figure 5 shows the cyclic voltammograms of 0.5 mM DA oxidation in a phosphate buffer solution (PBS 7.4) on bare GCE, GCE | PEDOT and GCE | Pd-PEDOT electrodes. On bare GCE, DA was oxidized at around 0.205 V yielding a peak current of 16 ± 2 lA, while a reduction peak at around 0.112 V was observed on the reverse scan. The
peak separation (EaDA - EcDA, DEp = 0.093 ± 0.002 V) Fig. 3 SEM images of (a) PEDOT & (b) Pd/PEDOT coated on ITO suggests quasi-reversible nature of the oxidation process. glass substrate On GCE | PEDOT, DA undergoes oxidation at around 123 1586 J Appl Electrochem (2008) 38:1583–1588
Fig. 4 AFM images of (a) PEDOT & (b) Pd/PEDOT coated on ITO glass substrate
catalytic oxidation of DA on the modified electrode. Dur-
ing the reverse scan, a cathodic peak and DEp of 0.112 V and 0.068 ± 0.002 V, respectively, were observed which corresponds to a near-reversible process. Upon Pd incor- poration within the polymer matrix, the DA oxidation occurs at 0.174 V with an oxidation current of 51 ± 2 lA showing the improved catalytic behaviour of the composite film compared to PEDOT modified and bare GCE. During
the reverse scan it also showed a peak at 0.1 V and the DEp was found to be 0.074 V indicating the reversibility of the DA oxidation process. The reversible process indicates that DA oxidation on the composite electrodes does not cause electrode fouling. This is a serious problem faced in the analytical determination which in turn affects the precision.
3.5 Effect of Pd loading Fig. 5 Cyclic voltammogram of DA oxidation on (a) bare GCE (b) PEDOT/GCE (c) Pd/PEDOT/ GCE in neutral PBS 7.4 To find the optimum loading of Pd in PEDOT that exhibits 0.180 V, yielding a peak current of 29 ± 2 lA. When higher catalytic activity, polymer films containing various compared to bare GCE, appreciable potential shift for the loadings of Pd were synthesized. In order to vary the Pd DA oxidation on GCE | PEDOT was observed. The loading in the polymer film, the precursor i.e., palladium oxidation current increased by 1.8 times, indicating chloride concentration in the deposition solution was 123 J Appl Electrochem (2008) 38:1583–1588 1587
60
50 A µ
40
30 DA oxidation current /
20 0 100 200 300 400 500 600 Pd loading / ng Fig. 7 Differential pulse Voltammogram of DA and UA oxidation in neutral PBS 7.4 Fig. 6 Effect of palladium loading on PEDOT and its catalytic activity determination whereas it was negligible in DPV mode [37]. varied. The amount of Pd in the film was calculated based Figure 7 shows the DPV of DA and UA in the presence of on the integration of the hydrogen adsorption region. Fig- excess AA in PBS 7.4 on Pd-PEDOT modified GCE. The ure 6 shows the DA oxidation current against amount of Pd peak potential separations observed between DA-AA and loading in the film. The DA oxidation current increases UA-AA were 0.19 V and 0.32 V, respectively. This sepa- with Pd loading up to *250 ng and decreases after the ration was well suited for DA and UA sensing in the critical loading. Pd loading corresponding to higher cata- presence of excess AA. Peak current was found to increase lytic oxidation of DA was taken as the threshold Pd with concentration of the individual analyte molecule. loading and used for DPV analysis. Figure 8 shows calibration curves obtained using DPV. These exhibit a linear relationship between catalytic peak 3.6 Simultaneous determination of DA and UA current and analyte concentration over a range of 0.5–1.0 and 7–11 lM for DA and UA, respectively. The calibration Detection of DA and UA concentration in biological plots were found to be linear. For DA, the corelation samples in the presence of other constituents is an impor- coefficient (R2) was 0.99 with 0.0019 lA/nM slope while tant task in clinical research [33–36]. DA and UA co-exist for UA, R2 was 0.99 with 0.1979 lA/lM slope. Thus, the in biological samples along with other constituents in polymer film showed improved properties for sensing which AA is the excess amount compared to other mole- applications in terms of selectivity and sensitivity com- cules. Therefore, the detection of DA and UA in the pared to the individual materials respectively. presence of excess AA is of great importance. Cyclic voltammetric studies of DA and UA oxidation on bare GCE in the presence of excess AA showed a broad 4 Conclusions and overlapping anodic peak (Figures not included). In contrast, PEDOT-GCE showed well-defined voltammetric Pd-PEDOT films were prepared electrochemically and peaks for each analyte. In order to improve the catalytic characterized using SEM and AFM techniques. PEDOT function, metal was entrapped into the polymer matrix and alone has a fibrillar network structure and Pd was incor- for this reason Pd was incorporated into the PEDOT film. porated within the porous polymer network structure. The separation among the peaks was quite large which Pd-PEDOT composite material was used for sensing DA paved the way for simultaneous/selective determination of and UA in the presence of excess AA. Compared to bare these biologically important molecules. GCE and PEDOT alone, Pd incorporated PEDOT showed The electrooxidation of DA and UA in the presence of excellent catalytic activity towards DA oxidation. The excess AA was investigated simultaneously by varying the reversibility of DA oxidation on Pd-PEDOT film was concentration of the individual analyte species.The DPV considerably improved and hence the composite-modified technique was adopted because it has a much higher cur- electrode does not suffer from fouling due to the products rent sensitivity and better resolution than cyclic of DA oxidation. Pd-PEDOT composite film enables voltammetry. Moreover, the charging current contribution detection up to 0.5 lM of DA and 7.0 lM of UA. Good to the background current is the limiting factor in analytical linear correlations between oxidation current and DA and 123 1588 J Appl Electrochem (2008) 38:1583–1588
1.4 2. Gronendaal LB, Jonas F, Freitag D, Pielartzik H, Reynolds JR (2000) Adv Mater 12:481 y = 1.8849x - 0.7185 3. Roncali J, Blanchard P, Fre`re P (2005) J Mater Chem 15:1589 1.2 2 R = 0.9909 4. Kumar A, Buyukmumcu Z, Sotzing GA (2006) Macromolecules 39:2723 1 5. Biallozor S, Kupnieska A, Jasulajtene V (2004) B Electrochem 20:231 A
µ 0.8 6. Niu L, Li Q, Wei F, Chen X, Wang H (2003) Synth Met 139:271 7. LakshmiKantam M, Roy M, Roy S, Sreedhar B, Madhavendra SS, Choudary BM, De RL (2007) Tetrahedron 63:8002 0.6 8. Dodouche I, Epron F (2007) Appl Catal B: Environ 76:291 Current / 9. Serov AA, Cho SY, Han S, Min M, Chai G, Nam KH, Kwak C 0.4 (2007) Electrochem Commun 9:2041 10. Tarasevich MR, Zhutaeva GV, Bogdanovskaya VA, Radina MV, 0.2 Ehrenburg MR, Chalykh AE (2007) Electrochim Acta 52:5108 11. Bashyam R, Zelenay P (2006) Nature 443:6 12. Tsakova V, Winkels S, Schultze JW (2001) J Electroanal Chem 0 500:574 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 13. Ilieva M, Tsakova V (2004) Synth Met 141:287 µ -3 Concentrations of DA / mol dm 14. Ilieva M, Tsakova V (2004) Synth Met 141:281 2 15. Ilieva M, Tsakova V (2005) Electrochim Acta 50:1669 16. Ilieva M, Tsakova V, Erfurth W (2006) Electrochim Acta 52:816 y = 0.1979x - 0.5505 17. Cho SH, Park SM (2006) J Phys Chem B 110:25656 1.8 R2 = 0.9921 18. Murugan AV, Gopinath CS, Vijayamohanan K (2005) Electro- chem Commun 7:213 1.6 19. Her LJ, Hong JL, Chang CC (2006) J Power Sources 157:457 20. Mathiyarasu J, Kumar SS, Phani KLN, Yegnaraman V (2007) J Nanosci Nanotechnol 7:2206 A µ 1.4 21. Kumar SS, Mathiyarasu J, Phani KLN, Jain YK, Yegnaraman V (2005) Electroanalysis 17:2281 1.2 22. Radhajeyalakshmi S, Kumar SS, Mathiyarasu J, Phani KLN,
Current / Yegnaraman V (2007) Ind J Chem 46:957 23. Kumar SS, Mathiyarasu J, Phani KLN (2005) J Electroanal Chem 1 578:95 24. Kumar SS, Mathiyarasu J, Phani KLN, Yegnaraman V (2006) J 0.8 Solid State Electrochem 10:905 25. Liu WJ, Xie YX, Liang Y, Li JH (2006) Synthesis 8:860 26. Karimi B, Enders D (2006) Org Lett 8:1237 0.6 27. Ackermann L, Althammer A (2006) Org Lett 8:3457 6 7 8 9 10 11 12 13 28. Yu S, Welp U, Hua LZ, Rydh A, Kwok WK, Wang HH (2007) Concentrations of UA / µmol dm-3 Chem Mater 17:3445 29. Dong S, Deng Q, Cheng G (1993) Anal Chim Acta 279:235 Fig. 8 Calibration curves of DA and UA 30. Yang C, Senthil Kumar A, Kuo M, Chien S, Zen J (2005) Anal Chim Acta 554:66 UA concentrations were obtained. As a result Pd-PEDOT 31. Yang C, Senthil Kumar A, Zen J (2006) Electroanalysis 18:64 shows more sensitivity and selectivity towards DA and UA 32. Han D, Yang G, Song J, Niu L, Ivaska A (2007) J Electroanal Chem 602:24 in the presence of excess AA when compared to PEDOT. 33. Li Y, Lin X (2006) Sensor Actuat B-Chem 115:134 34. Aguilar R, Da´vila MM, Elizalde MP, Mattusch J, Wennrich R Acknowledgments S. Harish thanks C.S.I.R., New Delhi for the (2004) Electrochim Acta 49:851 award of Research Internship. 35. Zen JM, Hsu CT, Hsu YL, Sue JW, Conte ED (2004) Anal Chem 76:4251 36. Zare HR, Rajabzadeh N, Nasirizadeh N, Ardakani MM (2006) J References Electroanal Chem 589:60 37. Wang J (2000) Analytical electrochemistry. (Second Ed.), Wiley- VCH, New York 1. Lyons MEG (1994) In: Lyons MEG (ed) Electroactive polymer electrochemistry, Part I. Plenum Press, New York
123 Catal Lett (2009) 128:197–202 DOI 10.1007/s10562-008-9732-x
Synthesis of Conducting Polymer Supported Pd Nanoparticles in Aqueous Medium and Catalytic Activity Towards 4-Nitrophenol Reduction
S. Harish Æ J. Mathiyarasu Æ K. L. N. Phani Æ V. Yegnaraman
Received: 10 July 2008 / Accepted: 14 October 2008 / Published online: 1 November 2008 Ó Springer Science+Business Media, LLC 2008
Abstract We report here the synthesis of palladium (Pd) 1 Introduction nanoparticles incorporated poly-(3,4)ethylenedioxythio- phene (PEDOT) matrix in aqueous medium and its Metal nanoparticles continue to generate intense interest in catalytic performance towards 4-nitrophenol reduction. various fundamental and applied areas of chemistry due to This simple one-pot synthesis involving a redox reaction their properties distinctly different from their bulk coun- between 3,4-ethylenedioxythiophene and palladium chlo- terparts. They are extremely reactive by virtue of their ride (PdCl2) precursor, leads to the formation of Pd intrinsic electronic properties and high surface-to-volume nanoparticles supported on particulate PEDOT. Pd nano- ratio. It is now well recognized that the particle stability particles of size 1–9 nm were found to distribute uniformly against agglomeration can be achieved by using stabilizers over the PEDOT matrix. Morphology of the Pd–PEDOT (solution-phase) and porous solid matrices (heterogeneous nanocomposite was characterized by field emission-scan- catalysis), electronically conductive or otherwise. Most ning electron microscopy and transmission electron commonly used stabilizers and supports are self-assembled microscopy and the crystallographic details obtained using monolayers [1], polymers [2], dendrimers [3–5] functional X-ray diffraction. The chemical nature of the PEDOT organic polymers [6], etc. These novel materials find support matrix was analyzed using Fourier transform-infra applications in the design of sensors [7], catalysts [8] and red (FT-IR) spectroscopy. The catalytic activity of the electrochromic devices [9]. composite was demonstrated using a model reaction, i.e., The synthesis of metal nanoparticles involves reduction reduction of 4-nitrophenol to 4-aminophenol. The value of the metal precursor, i.e., metal salts by various reducing of the apparent rate constant, ca. 65.8 9 10-3 s-1 obtained agents. Facile oxidation of monomers like pyrrole [10], using UV visible spectroscopy of the reduction of aniline [11, 12] and thiophene [13] in presence of a few 4-nitrophenol at the Pd–PEDOT nanocomposite is metal salts forms the basis for producing nanocomposites comparable to those reported for other catalytic systems. of metal nanoparticles contained in a porous conducting polymer matrix, generated by the oxidation of the mono- Keywords Heterogeneous catalysis Palladium mer. Thiophene-class of monomers, for example, 3,4- PEDOT 4-Nitrophenol Reduction Nanoparticle ethylenedioxythiophene (EDOT) can function as reducing agents leading to the formation of metal-conducting polymer composites. Such polymer-supported metal nanoparticles have been of current interest due to their properties relevant to (micro-) heterogeneous catalysis, particularly for hydrogenation reactions [14, 15]. In the S. Harish J. Mathiyarasu (&) K. L. N. Phani class of nitro-compounds, the reduction of 4-nitrophenol V. Yegnaraman has been a model reaction to demonstrate the catalytic Electrodics and Electrocatalysis Division, Central activity of the metal nanoparticles [16]. We recently have Electrochemical Research Institute, Karaikudi 63006, Tamilnadu, India shown that stable dispersions of the Au–poly-(3,4)ethyl- e-mail: [email protected] enedioxythiophene (PEDOT)–PSS nanocomposite [13] can 123 198 S. Harish et al. catalyze this reaction with an apparent rate constant of 2 Experimental 43.9 9 10-3 s-1. The premise of the present investigation is that palladium (Pd), with its catalytic ability to effect 2.1 Chemicals hydrogenation reactions [17–19] can be a suitable candi- date for carrying out reduction of nitrophenol and hence be 3,4-Ethylenedioxythiophene (EDOT) (Aldrich), sodium of significance to the synthesis of nanocomposites. Inter- 4-polystyrenesulfonate (Aldrich), palladium chloride estingly, we found that the salts of Pd can oxidize the (PdCl2) (Merck), hydrochloric acid (Ranbaxy), 4-nitro- thiophene moiety to produce its conducting polymer while phenol (Ranbaxy), and sodium borohydride (Merck) all of also undergoing reduction to Pd nanoparticles, in this analytical grade were used as received. Aqueous solutions process. EDOT functions as a reducing agent since its were prepared using Milli-Q water of 18 MX. oxidation potential is sufficient enough to promote the formation of metal (M = Au, Ag, Pt) nanoparticles. These 2.2 Methods simultaneous processes lead to the formation of metal– polymer nanocomposites that can be obtained as insoluble Scanning electron microscopy (SEM) measurements were powders, porous thin films and can be made water-soluble made using Hitachi Field emission-Scanning electron by including a polymeric stabilizer (and dopant for microscopy (FE-SEM) (Model S4700) with an acceleration PEDOT) like, for example, sodium polystyrenesulfonate, voltage of 10 kV in normal mode. X-ray diffraction (XRD) as different heterogeneous reactions require conditions patterns were recorded in a PANalytical diffractometer specific to the environment, i.e., solid or solution-phase. Model PW3040/60 X’pert PRO operating with Cu Ka It may be noted that a strong reducing agent like boro- radiation (k = 0.15406 nm) generated at 40 kV and hydride needs a metal catalyst to hydrogenate the nitro- 20 mA. Scans were done at 3° min-1 for 2h values compounds [20, 21]. Since Pd itself is a good candidate as between 20° and 90°. Transmission electron microscopy a catalyst for hydrogenation reactions, it is of our current (TEM) examination was made by placing a drop of the interest to generate its nanocomposites with conducting sample (dispersed in acetone) onto a copper grid coated polymers. Such composites can serve as catalysts not only with carbon film (400 meshes) and kept aside for drying in in heterogeneous catalysis but also in electrochemical air for several hours at room temperature. The TEM images processes. Recently, Ilieva et al. [22] reported that Cu–Pd were collected from Philips CM200 microscope working at modified PEDOT shows a marked electroactivity towards 200 kV. For infrared spectroscopic measurements, a nitrate reduction and the enhancement in the reduction is Thermo-Electron Corp., USA, Nexus 670 model Fourier due to the adsorption of hydrogen by the deposited Pd transform-infra red (FT-IR) spectrometer (DTGS detector) within the polymer film. was used. UV–vis spectra were collected using Cary 500 As part of our current efforts in synthesizing and char- spectrophotometer. The catalytic reduction reaction was acterizing conducting polymer–metal nanocomposites, carried out in a standard quartz cell of 1 cm path length
Aunano–PEDOT/PSS system [13] was reported from this with about 3 mL volume and spectra were recorded with a laboratory. EDOT was used to reduce HAuCl4 in solutions time interval of 60 s in a scanning wavelength range of containing sodium polystyrenesulfonate. In the process of 200–600 nm at 25 °C. reducing HAuCl4, EDOT is oxidized to form a polymer showing characteristics of PEDOT. It exhibited a very high 2.3 Synthesis of Pd–PEDOT/Pd–PEDOT–PSS stability in strong salt solutions, pH-sensitivity and cata- lytic activity to the reduction of p-nitrophenol. In a similar Synthesis of the Pd-incorporated PEDOT composite -3 vein, considering the hydrogen adsorption/absorption involves addition of 1 mL of PdCl2 (5 9 10 M) to characteristics of Pd metal, we attempted synthesizing the 10 mL of EDOT (1 9 10-2 M) in aqueous solution.
Pd–PEDOT nanocomposites. Interestingly, when PdCl2 is Within a few minutes, the color of the solution changes to added to EDOT solution, by an instantaneous reaction, Pd black and then slowly black particles are precipitated out nanoparticles are formed along with the PEDOT polymer from the solution. This observation pertains to the forma- from the added EDOT monomer. In addition, Pd is tion of Pd nanoparticles. exploited as a catalyst in various coupling reactions like As mentioned earlier, EDOT functions as a reducing Heck coupling [23], Suzuki coupling [24], hydrogenation agent for the synthesis of Pd nanoparticles since its oxi- of allyl alcohols [25], etc. dation potential is sufficient enough to drive the reduction In the present study, we report the methodology for the of Pd2? while also undergoing oxidation to PEDOT. The synthesis of Pd–PEDOT nanocomposite as solid powders PEDOT formed is insoluble in aqueous medium and hence and dispersions, structural characteristics, morphology and is precipitated. The precipitate was thoroughly washed catalytic activity for p-nitrophenol reduction. several times with Milli-Q water and acetone and dried 123 Conducting Polymer Supported Pd Nanoparticles 199 under vacuum. The dried sample is characterized by XRD, 180 (111) TEM, FT-IR and FE-SEM. 160
The synthesis of Pd–PEDOT–PSS aqueous dispersion 140 involves the reduction of PdCl using EDOT in presence of 2 120 poly-4-styrene sulfonate sodium (Na–PSS). In brief, 10 mL of EDOT (1 9 10-2 M) was dissolved in water along with 100 80 1%Na–PSS (under continuous stirring) which also increa- Counts (200) ses the solubility of EDOT due to the formation of a 60 (311) (220) pseudo-complex [26]. Subsequently, 1 mL of aqueous 40 -3 PdCl2 (5 9 10 M) was added slowly to the initial solu- 20 (222) tion under stirring. Color changes instantaneously from 0 colorless to black, indicating the formation of Pd nano- 20 30 40 50 60 70 80 90 particles, whereas the precipitation is avoided by the 2θ presence of PSS stabilizer. Fig. 2 XRD diffraction pattern of Pd–PEDOT composite
3 Results and Discussion The size of the Pd nanoparticles determined using TEM analysis is more reliable than that determined by using 3.1 Characterization of Pd–PEDOT Scherrer formula in XRD analysis. FE-SEM gives the size of the globules consisting of a composite of Pd particles Figure 1 shows the FE-SEM image of Pd incorporated and PEDOT, whereas, TEM clearly shows the distinction PEDOT composite. The image shows that particles are between the PEDOT matrix and Pd nanoparticles. Figure 3 uniformly distributed throughout the porous structure of shows the TEM images of the Pd–PEDOT composite PEDOT and appears as globules of average size of approx. where the black spots correspond to Pd nanoparticles 80 nm. Since FE-SEM imaging does not distinguish embedded in the polymer matrix. The particles are dis- between Pd and PEDOT, XRD analysis was carried out to persed uniformly on the polymer matrix and size appears to confirm the presence of Pd in the composite. be ranging from 1 to 9 nm. XRD analysis of the Pd–PEDOT composite (Fig. 2) The FT-IR spectra of EDOT and Pd–PEDOT in Fig. 4 shows the reflections at 39.92°, 46.54°, 67.90°, 81.84° and show that EDOT is completely oxidized by Pd2? to a 86.54°, and these peaks correspond to (111), (200), (220), polymer PEDOT and in turn Pd2? undergoes reduction to (311), and (222) lattice planes of Pd. Planes were assigned Pd0. The peaks at 1186 and 892 cm-1 (EDOT) correspond by comparing with the standard Pd and these planes cor- to =C–H in-plane and out-of-plane deformation vibrations, respond to the fcc crystal lattice structure of Pd. No peaks respectively. Those peaks (marked*) are not present in the corresponding to PEDOT were observed. The crystallite spectrum of Pd–PEDOT. A broad peak at 1432 cm-1 is size of Pd particles was evaluated using Scherrer equation attributed to aromatic C=C stretching. These results prove for the (220) peak and is found to be approx. 24 nm in size. that PEDOT is in the oxidized form with a–a0 coupling. The other peak at 1352 cm-1 is due to C–C and C=C stretching of quinoidal structure originating from the thi- ophene ring. The peaks at 907 and 673 cm-1 are assigned to stretching of C–S bond in the thiophene ring and the peaks at 1243 and 1057 cm-1 are due to the –C–O–C-bond [27, 28].
3.2 Catalytic Reduction of 4-Nitrophenol
As Pd is known to be an excellent catalyst for hydroge-
nation reactions, Pdnano–PEDOT can be considered for the reduction of 4-nitrophenol to 4-aminophenol using boro- hydride. In order to perform this reaction, the use of a water-dispersible catalyst is necessitated. For this, Pd nanocomposite was synthesized using sodium polysty- renesulfonate (Na–PSS) as solubilizer and stabilizer. In Fig. 1 FE-SEM image of Pd–PEDOT (scale bar 1.00 lm) addition, Na–PSS also serves as a good dopant in the 123 200 S. Harish et al.
inferred in our earlier work [13]. In the present case, Pd–PEDOT–PSS aqueous dispersion was synthesized by
chemically reducing PdCl2 using EDOT as the reductant in presence of Na–PSS which effectively stabilizes the oxi- dized form of PEDOT compared to other stabilizers during the oxidative polymerization of EDOT [13]. The catalytic application of Pd–PEDOT–PSS nano- composite is demonstrated by 4-nitrophenol reduction. Initially, the reduction of 4-nitrophenol was demonstrated by several groups using sodium borohydride with a metal catalyst and observed that without a catalyst, sodium borohydride is not an effective reducing agent. In this
present work, it is observed that the presence of Pdnano– PEDOT–PSS in aliquot amounts is sufficient for the pro- motion of nitrophenol reduction with a high value of the apparent rate constant. Absorption of 4-nitrophenol (1 9 10-6 M) occurs at
313 nm and after the addition of NaBH4, the absorption peak undergoes red shift to 400 nm with a color change from light yellow to dark yellow (corresponding to the generation of nitrophenolate anion) [29]. The absorption peak at 400 nm remains unaltered even after the addition of
excess of NaBH4. It confirms that reduction is not achievable in the presence of NaBH4 alone. Interestingly, after the addition of Pd–PEDOT–PSS dispersion (Pd loading: 2 lg/ll) to the initial solution, there was a continuous fading of color leading to discoloration of the solution. In order to follow the kinetics of the reduction reaction, the change in the intensity of absorption of nitrophenolate was monitored using UV–visible spectro- Fig. 3 TEM images of Pd–PEDOT composite (a, b) and histogram photometry at regular time intervals. of the particle distribution (c) UV–vis spectral studies reveal that (Fig. 5), the only
peak due to the nitro group in the presence of NaBH4 at 400 nm corresponds to the formation of nitrophenolate
2.0 Pd-PEDOT -0.2 1.8 -0.4 1.6 -0.6
1.4 log A -0.8 1.2 EDOT -1.0 1.0
% Transmission % 0 50 100 150 200 250 0.8 Time in seconds
* 0.6 Absorbance in a.u * 0.4
1600 1400 1200 1000 800 600 400 0.2 Wave number (cm-1 ) 0.0 200 300 400 500 600 Fig. 4 FT-IR pattern of Pd–PEDOT composite Wavelength in nm oxidative polymerization of EDOT and for PEDOT. A Fig. 5 UV–visible spectrum of 4-nitrophenol reduction in the presence of excess NaBH4 and 40 ll of catalyst (loading: 2 lg/ll) porous nanocomposite in the form of metal nanoparticles in taken at regular intervals; inset: plot between log A vs. time in the core and the matrix PEDOT being the shell was initially seconds for the disappearance of 4-nitrophenol absorption at 400 nm 123 Conducting Polymer Supported Pd Nanoparticles 201
Table 1 Comparison of apparent rate constants for 4-nitrophenol be concluded that the reduction takes place only due to Pd reduction at Pd–PEDOT–PSS with different reported catalyst systems nanoparticles. The reaction mechanism can be reasoned by Composition Apparent rate constant Reference the inherent hydrogen adsorption/desorption characteristics (s-1) of Pd [19, 20]. Here, the Pd nanoparticles shuttle the hydrogen transport between the NaBH and 4-nitrophenol. PNIPA gels (Ag) 3.50 9 10-3 [30] 4 -3 The shuttling behavior can be reasoned that the Pd nano- b-D-glucose network (Au) 6.54 9 10 [31] particle adsorbs hydrogen from the NaBH and efficiently PPI–Dendrimers (Au) 13.2 9 10-3 [32] 4 releases during the reduction reaction (Scheme 1) and hence PAMAM–Dendrimers (Au) 3.70 9 10-3 Pd acts as a hydrogen carrier in this reduction reaction. Pure Pt 0.12 9 10-3 [33] Raney Ni 0.15 9 10-3 -3 Ni–Pt (64:36) 0.48 9 10 4 Conclusions Ni–Pt (80:20) 0.90 9 10-3 -3 Ni–Pt (96:4) 1.93 9 10 A new route for the synthesis of Pd nanopartilces in an -3 FGME–Cu 9.00 9 10 [34] aqueous medium at room temperature involving one step -3 FGME–Ag 9.50 9 10 and one-pot process is reported. The nanocomposite was -3 FGME–Au 17.5 9 10 characterized using FE-SEM, XRD, TEM and FT-IR. Pd - -3 Aunano–PEDOT/PSS 43.9 9 10 [13] nanoparticles were found to be in fcc phase and the particle - -3 Pdnano–PEDOT/PSS 65.8 9 10 This work size was found to be 1–9 nm range from TEM measure- -3 Microgel-Pd 1.50 9 10 [35] ments. Catalytic activity of Pd nanoparticles was exploited SPB–Pd 4.41 9 10-3 in the case of 4-nitrophenol reduction. Controlled reaction PAMAM–Pd 3.59 9 10-3 [36] kinetics allows for calculating the apparent rate constant of PPI–Pd 407 9 10-3 the reaction and the value is comparable to the reported ones.
Acknowledgments S. H. thanks to CSIR, New Delhi for the award anion. By adding aliquot amounts (40 ll) of Pd–PEDOT– of Research Internship. The authors thank Dr. A. S. Prakash, CECRI PSS to the initial solution, there is a concomitant emer- for help in TEM measurements and N. Alagirisamy of Hindustan gence of a peak at 310 and 230 nm that corresponds to the Unilever Ltd., Bangalore for FE-SEM measurements. formation of 4-aminophenol. Continuous reduction in the intensity of the peak at 400 nm shows the consumption of 4-nitrophenol. However, there is no proportional increase References in the aminophenol peak intensity; probably due to the difference in the molar extinction co-efficient of 4-nitro- 1. Ulman A (1996) Chem Rev 96:1533 2. Cole DH, Shull KR, Baldo P, Rehn L (1999) Macromolecules phenol and 4-aminophenol. In addition, the plot between 32:771 the logarithmic value of absorbance and time is found to be 3. Wang R, Yang J, Zheng Z, Carducci MD, Jiao J, Seraphin S linear (Fig. 5 inset). The apparent rate constant is calcu- (2001) Angew Chem Int Ed 40:549 lated from the falling of the peak intensity of 4-nitrophenol 4. Zheng J, Stevenson MS, Hikida RS, Van Patten PG (2002) J Phys -3 -1 Chem B 106:1252 at 400 nm and is found to be 65.8 9 10 s . The 5. Oh SK, Niu Y, Crooks RM (2005) Langmuir 21:10209 apparent rate constant obtained for 4-nitrophenol reduction 6. Kra´lik M, Biffis A (2001) J Mol Catal A Chem 177:113 are rather high in comparison to almost all previous find- 7. Huang XJ, Choi YK (2007) Sens Act B 122:659 ings in the literature including those obtained with Pd 8. Astruc D, Lu F, Aranzaes JR (2005) Angew Chem Int Ed 44:7852 9. Namboothiry MAG, Zimmerman T, Coldren FM, Liu J, Kim K, (Table 1). Carroll DL (2007) Synth Met 157:580 In our previous report [13], we reported that PEDOT–PSS 10. Park JE, Atobe M, Fuchigami T (2005) Electrochim Acta 51:849 dispersion alone cannot reduce 4-nitrophenol. Hence, it can 11. Tseng RJ, Huang J, Ouyang J, Kaner RB, Yang Y (2005) Nano Lett 5:1077 12. Li W, Jia QX, Wang HL (2006) Polymer 47:23 13. Kumar SS, Sivakumar C, Mathiyarasu J, Phani KLN (2007) Langmuir 23:3401 14. Drelinkiewicza A, Waksmundzka A, Makowski W, Sobczak JW, Krol A, Zieba A (2004) Catal Lett 94:143 15. Gelder Elaine A, David Jackson S, Martin Lok C (2002) Catal Lett 84:205 16. Kuroda K, Ishida T, Haruta M (2008) J Mol Catal A: Chem. doi: 10.1016/j.molcata.2008.09.009 Scheme 1 Schematic representation of the conversion of 4-nitrophe- 17. Yamauchi M, Ikeda R, Kitagawa H, Takata M (2008) J Phys nol to 4-aminophenol in the presence of NaBH4 and Pd–PEDOT–PSS Chem C 112:3294 123 202 S. Harish et al.
18. Yoswathananont N, Nitta K, Nishiuchi Y, Satogive M (2005) 28. Kvarnstro¨m C, Neugebauer H, Blomquist S, Ahonen HJ, Kankare Chem Commun 40–42 J, Ivaska A (1999) Electrochim Acta 44:2739 19. Kishorea S, Nelsonb JA, Adairb JH, Eklund PC (2005) J Alloys 29. Panigrahi S, Basu S, Praharaj S, Pande S, Jana S, Pal A, Ghosh Compd 389:234 SK, Pal T (2007) J Phys Chem C 111:4596 20. Corma A, Serna P (2006) Science 313:332 30. Lu Y, Mei Y, Drechsler M, Ballauff M (2006) Angew Chem Int 21. Blaser HU (2006) Science 313:312 Ed 45:813 22. Ilieva M, Tsakova V, Erfurth W (2006) Electrochim Acta 52:816 31. Liu J, Qin G, Raveendran P, Ikushima Y (2006) Chem Eur J 23. Karimi B, Enders D (2006) Org Lett 8:1237 12:2131 24. Klingensmith LM, Leadbeater NE (2003) Tetrahedron Lett 32. Hayakawa K, Yoshimura T, Esumi K (2003) Langmuir 19:5517 44:765 33. Ghosh SK, Mandal M, Kundu S, Nath S, Pal T (2004) Appl Catal 25. Wilson OM, Knecht MR, Garcia-Martinez JC, Crooks RM (2006) A: Gen 268:61 J Am Chem Soc 128:4510 34. Pradhan N, Pal A, Pal T (2001) Langmuir 17:1800 26. Sakmeche N, Aeiyach S, Aaron JJ, Jouini M, Lacroix JC, Lacaze 35. Mei Y, Lu Y, Polzer F, Ballauff M (2007) Chem Mater 19:1062 PC (1999) Langmuir 15:2566 36. Esumi K, Isono R, Yoshimura T (2004) Langmuir 20:237 27. Li X, Li Y, Tan Y, Yang C, Li Y (2004) J Phys Chem B 108:5192
123 Author's personal copy
Electrochimica Acta 54 (2009) 3618–3622
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Barrier films to control loss of 9,10-anthraquinone-2-sulphonate dopant from PEDOT films during electrochemical transitions
S. Harish, D. Sridharan, S. Senthil Kumar, James Joseph ∗, K.L.N. Phani
Electrodics and Electrocatalysis Division, Central Electrochemical Research Institute, Karaikudi 630 006, India article info abstract
Article history: We describe a simple approach for the synthesis of stable electroactive poly[3,4-ethylene dioxythiophene] Received 12 September 2008 (PEDOT) in an acid medium, by incorporating a redox active dopant like 9,10-anthraquinone-2-sodium Received in revised form 6 January 2009 sulphonate (AQS) on a glassy carbon (GC) electrode. The modified electrode is responsive up to a pH of 7. Accepted 12 January 2009 The stability of the modified electrode during continuous electrochemical cycling is poor, due to leaching Available online 20 January 2009 of the dopant from the PEDOT film. Efforts are made to improve the stability of the modified electrode by forming an anionic barrier film on the PEDOT-AQS interface either physically or electrochemically. Keywords: The modified electrodes were monitored by cyclic voltammetry and Fourier transform-infrared (FT-IR) PEDOT Anthraquinone sulphonate spectroscopy for the presence of AQS in the film. Nafion © 2009 Elsevier Ltd. All rights reserved. Polystyrene sulphonate Barrier film Stability
1. Introduction provides an easy way to fabricate thin films of conducting polymer matrices with entrapped redox mediators. Conducting polymers are essential components in electrochem- There have been attempts to incorporate polyoxometallate ical devices because of their electronic conductivity and doping electroactive species in conducting polymer thin films. These properties. The latter characteristic allows chemists to individ- compounds are multi-centered redox anionic species. The electro- ually tune the properties of conducting polymers [1]. Various catalytic mediating properties of silicotungstic acid doped PEDOT, redox-active dopants, like viologens, quinones, tetrathiafulvalene, for the reduction of molecular oxygen, was reported recently by ferrocenes and polyoxometallates are used as dopants in the con- Kulesza et al. [10]. One advantage of using PEDOT as a matrix ducting polymer matrix to improve the interfacial properties of for entrapping redox-active species is that, unlike polyaniline, it electrocatalysis, ion sensing and electrochromism [2–6]. The con- remains electroinactive in highly acid medium. In other words, the ducting polymers of polyaniline, polypyrrole, and polythiophene electroactivity of the PEDOT film does not interfere with the redox derivatives are used as electrode modifiers or as matrices for trap- chemistry of the doped electroactive species. Doping of PEDOT ping redox species [5–7]. Among the conducting polymers, PEDOT with high-molecular weight anionic polymeric species such as PSS− has received special attention for various applications, including (polysyrene sulphonate anion) has been well established [11]. Once use as a matrix for entrapping enzymes in biosensors, use in elec- incorporated, it is difficult to remove this bulky dopant because it is trochromic displays, and use as a hole conductor in solid state polymeric. This feature is also responsible for the pseudo-n-doping devices [8]. There is intense interest in this polymer because of its behavior of electronic conducting polymers [12,13].However,ifthe high-electronic conductivity (ca. 300 S cm−1), regio-regular struc- dopant is not big enough (for example, aryl sulphonates), it is likely ture and good chemical stability. The electrocatalytic properties of that, during potential cycling, the dopant will leach out into the the modified electrodes can be altered by modification of interfaces solution, which contains smaller anions such as perchlorate. with thin films of redox species. Electrochemical modification of In this work, we report electropolymerization of EDOT in an glassy carbon/ITO electrodes with redox electron transfer media- acid medium containing AQS. The sulphonate group is highly tors is common practice in electrochemistry [9]. Electrodeposition anionic in nature. Therefore, it can easily be incorporated into the PEDOT film during electropolymerization. The incorporation of an anthraquinone moiety yields a highly reversible symmetric voltam- ∗ Corresponding author. metric peak due to the redox reactions. PEDOT-like films can be E-mail address: [email protected] (J. Joseph). used to incorporate anionic redox mediators through counterion
0013-4686/$ – see front matter © 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.electacta.2009.01.032 Author's personal copy
S. Harish et al. / Electrochimica Acta 54 (2009) 3618–3622 3619 doping. Initially, it was thought that the molecular/ionic size of AQS was sufficiently bulky to become entrapped. However, it is now realized that leaching through ion exchange with the anions in solu- tion leads to the loss of the redox dopant. After successful doping into PEDOT films, stability studies indicated that AQS− leaches out slowly into solution, leaving PEDOT without AQS. In this process, the solution anion replaced the AQS. In order to control the loss of electroactivity of the entrapped AQS−, a macromolecular dopant such as PSS− was employed. This strategy was based on the electro- static interactions and the relative sizes of AQS− and PSS−.Alayer of PSS is one of the constituents of the well-known layer-by-layer assembly approach [11]. The polypyrrole-arylsulphonate system is an appropriate example. The electropolymerization of pyrrole in the presence of various aryl sulphonate anions, including AQS−, was reported by Wang and MacDiarmid [14], Ahmed et al. [15] and Kuwabata et al. [16]. The properties of the doped films were also studied [14]. Loss of AQS− into the solution was not observed with the use of non-aqueous solvents. These solvents are employed in the examination of the electroactivity of the modified electrodes, Fig. 1. Cyclic voltammogram showing polymerization of 5 mM EDOT in 10 mM AQS −1 as they do not allow anion exchange. The use of non-aqueous sol- solution, scan rate 50 mV s . vents facilitated understanding of the two consecutive one-electron transfers taking place in AQS [15]. This finding is in contrast to the concentration in the deposition bath. In contrast to the redox film results obtained in aqueous solutions. The polyaniline-polyvinyl growth in solutions containing AQS, a featureless flat cyclic voltam- alcohol-aryl sulphonate systems studied by Nagaoka et al. [17] are mogram with an increase in only the non-faradaic current was interesting. The doped aryl sulphonates were not ejected from the observed for PEDOT film growth in solution containing LiClO4. polymer thin film during the chemical oxidation-reduction process. The films were prepared by varying the cycle numbers. Different The redox of the composite films is an interdependent action of thicknesses were observed with an increase in coulombic charge redox changes in polyaniline and aryl sulphonates. The redox of the under the redox peaks. The cyclic voltammetric peaks around PEDOT-AQS described in this work is only due to anthraquinone in −0.3 V are due to two-electron transfer in anthraquinone. The peak acidic solutions. This study presents a detailed method to circum- current increase in both directions can be attributed to the increase vent the loss of redox-active dopants from the PEDOT thin films in film thickness. This finding suggests an increase in the content − during electrochemical cycling. of the AQS dopant in the film. Along with electropolymerization of EDOT, there is doping of a charge-compensating anion from the electrolyte. The PEDOT-AQS modified electrode exhibited a highly 2. Experimental reversible redox peak at −0.15 V in 0.5 M H2SO4 medium containing no other electroactive species in the background. The redox group in Voltammetric experiments were carried out in the three- doped AQS (quinol/quinone) is responsible for the highly reversible electrode cell, using a BAS100B work station. The cell contained redox peak seen in the acidic medium (Fig. 2). The reversible redox −2 glassy carbon (GC) (surface area 0.07 cm ) as a working electrode, peak, occurring around 0.15 V, can be ascribed to the reversible a platinum (Pt) sheet as a counter electrode and a normal calomel redox reaction of the quinone/hydroquinone system [18] in AQS− electrode (NCE) as a reference electrode. All reported potentials (Scheme 1). The calculated charge corresponds to 52 mC cm−2 for were referenced to NCE. First, the working electrode was cleaned by films grown for 10 cycles at scan rate of 20 mV s−1. polishing with grade 4/0 emery paper. This cleaning was followed The redox peak at approximately −0.15 V is highly reversible. by de-greasing with acetone and sonication in distilled water. A This is demonstrated by the small peak-to-peak separation Thermo-Nexus 670 model spectrometer was used for FT-IR mea- −1 ( Ep = 20 mV or less at sweep rates lower than 0.01 V s ). The peak surements. All the chemicals used were of Analar grade and the solutions were prepared in fresh MilliQ water (18.2 M ). Buffer solutions were prepared using a 0.2 M sodium acetate/0.2 M acetic acid mixture. The pH was adjusted by adding appropriate amounts of HCl or NaOH.
3. Results and discussion
AQS− doped PEDOT films were synthesized electrochemically by potential cycling of a glassy carbon electrode in a solution con- taining 5 mM EDOT (monomer) and 10 mM AQS (shown in Fig. 1). Potential cycling was carried out in the range of −0.6 to 1.2 V vs. NCE at a scan rate of 50 mV s−1. Electropolymerization of EDOT in the medium containing AQS occurred through electro-oxidation of EDOT at approximately 840 mV vs. NCE. This led to a steep increase in the voltammetric current. From Fig. 1, it is clear that the cyclic voltammetric peak currents increased in both directions, centered around −0.35 V. This increase occurred with an increasing number of cycles, indicating growth of a redox-active film on the surface of Fig. 2. Cyclic voltammogram showing the redox behavior of AQS doped PEDOT in the GC. The thickness of the PEDOT-AQS film can be varied by con- −1 0.5 M H2SO4 medium at a scan rate of 20 mV s . (Inset) Plot of peak current vs. scan trolling the number of potential cycles or by changing the monomer rate. Author's personal copy
3620 S. Harish et al. / Electrochimica Acta 54 (2009) 3618–3622
Scheme 1. Redox reactions of AQS doped in PEDOT. also fulfils another criterion for surface anchored redox species, described in the following equation [19]:
[n2F2A ] Ip = (1) [4RT] In this equation, n is the number of electrons involved, F is Faraday’s constant, A is the geometrical area of the electrode, is the scan rate, R is the gas constant, is surface coverage and T is the absolute temperature. The voltammetric peak cur- rent shows a linear dependence with scan rate up to 0.1 V s−1, confirming ideal surface confined redox entrapped in the PEDOT film as seen from Fig. 2 (inset). However, the peak-to-peak sepa- Fig. 4. Cyclic voltammogram showing the instability of PEDOT-AQS up to 100 poten- ration of the redox film is found to increase with increasing scan −1 tial cycles in 0.5 M H2SO4 at scan rate 20 mV s . rate. This occurs because the AQS− anion leaches out of the film during potential cycling. The anion fails to ingress the interior of tial values with decreasing pH (Fig. 3). The voltammetric peak PEDOT film completely during the reverse cycle at higher scan rates. potential exhibited a near-Nernstian shift of about 52 ± 2 mV/pH The electroinactive sulphate ion may also participate in the dop- units in the acidic range up to pH 6 (Fig. 3 inset). This finding ing/dedoping process during potential cycling. demonstrates the utility of these films as pH sensors in the acidic range. 3.1. Effect of pH 3.2. Stability studies Platinum is classically used for sensing the pH of solutions in combination with the quinone-hydroquinone redox system Fig. 4 represents the voltammetric response of the PEDOT- [20,21]. The Scholz group instead reports using a pH-sensitive AQS-modified GC during continuous electrochemical cycling in a graphite/quinhydrone composite electrode, flow injection poten- sulphuric acid solution. A potential range of 0.0–1.0 V was used at tiometry, solid-composite pH sensors, quinhydrone, solid paraffin, a scan rate of 50 mV s−1. The redox peak current decreased in the and surface-modified graphite powder composites [22,23].We anodic and cathodic directions upon electrochemical cycling. This is attempted to use the GC/PEDOT-AQS interface for its voltammet- likely due to the fact that the anionic dopant (AQS−) slowly leached ric response to pH in a series of buffer solutions in the pH range out of the film into the electrolyte solution. The GC modified with of 1–7. Fig. 3 shows the voltammetric response of the PEDOT-AQS PEDOT-AQS exhibited a featureless voltammetric response after modified electrode in various pH buffers. In a neutral medium cycling of the modified electrode in H2SO4 for more than 100 cycles. (pH 7), the PEDOT-AQS film was redox inactive and turned redox- This is probably due to the loss of AQS− into the solution. active only at lower pH values. The mid-peak potential of the In an effort to stabilize the GC-modified PEDOT-AQS film against E E redox couple ( pa + pc)/2 was found to shift to anodic poten- loss of AQS−, we attempted to form an anionic barrier film over the PEDOT-AQS film. The anionic barrier was formed either by physical casting of an ionomer film or by electrochemical incorporation.
(a) Physical method: Coating the PEDOT-AQS with a layer of anionic Nafion solution. (b) Electrochemical method: Potential cycling of the PEDOT-AQS film in PSS− electrolyte solution.
The efficacy of the above methods in providing an anionic barrier to the loss of AQS− from the film is discussed below.
3.3. Physical method
The barrier film can be formed by using Nafion® (tetrafluoroethylene-perfluoro-3,6-dioxa-4-methyl-7- octenesulphonic acid copolymer dissolved in ethanol), which is a polyelectrolyte with sulphonic acid groups. A thin film of Nafion can be cast on the modified electrode by adding a few microliters of dilute Nafion solution onto the modified electrode and allowing it to dry for two hours in air. This thin film may act as a barrier for the egress of AQS− from the PEDOT-AQS film. From Fig. 5, it is clear that the modified electrode retains about Fig. 3. Cyclic voltammograms showing the pH response of the PEDOT-AQS film at a scan rate of 50 mV s−1 (voltammetric curves are in the order of decreasing pH from 70% of its electroactivity, after continuous potential cycling for − left to right). (Inset) Plot of mid-peak potential vs. pH of the buffer electrolyte. more than 100 cycles at a scan rate of 50 mV s 1. Nafion coating Author's personal copy
S. Harish et al. / Electrochimica Acta 54 (2009) 3618–3622 3621
Fig. 6. Cyclic voltammogram showing the stability of PEDOT-AQS up to 100 cycles
after electrochemical treatment in Na-PSS electrolyte in 0.5 M H2SO4 (scan rate − Fig. 5. Cyclic voltammogram showing the stability of PEDOT-AQS up to 100 potential 20 mV s 1). −1 cycles after casting Nafion on the electrode in 0.5 M H2SO4 at scan rate 20 mV s .
coating of Nafion. This may explain the improved retention of AQS− is known to remove the interference of ascorbates, urates, and in PEDOT in the electrochemically treated films. other compounds in glucose sensors by preventing anionic species A FTIR spectrum of the PEDOT-AQS coated with PSS is shown from getting to the film [24]. The anionic interferants are screened with the spectra of AQS and Na-PSS in Fig. 8. The absorption due to out by the charge-selective nature of the Nafion membranes. In –O–C–O– was observed around 1200 and 1068 cm−1. The aromatic our study, we utilized the charge-selectivity of Nafion to prevent C–H stretching appears at 3050 cm−1. The vibration present at the anionic dopant from leaching out of the modified electrode. <1000 cm−1 is due to C–H out-of-plane bending vibration in EDOT. This type of “reverse approach” has not yet been attempted in The stretching bands around 1486, 1598 and 1682 cm−1 are due to the literature of modified electrodes. The effectiveness of this CH and C C of PEDOT. The peak corresponding to aliphatic C–H approach in improving the stability of the modified electrode stretching, present at 2877 cm−1, indicates that PEDOT is doped towards electrochemical cycling is demonstrated in this simple with PSS−. The absorption peak at 1276 cm−1 can be assigned to experiment. The cation-exchange polymer (Nafion) that forms a the vibration of the sulphonyl group (S–O). The peak at 1750 cm−1 layer on the surface of the electrode works as a stabilizing layer is assigned to the C O of AQS. The peaks observed around 974, for the charged conducting polymer via electrostatic interactions 835 and 684 cm−1 show the C–S bond in the thiophene molecule. [25]. Hence, its macromolecular nature provides a barrier which Thus, FT-IR studies revealed the presence of both PSS and AQS in prevents leaching of the anionic dopant from the polymer film. the polymer film [26,27]. The increased stability of PEDOT-AQS films shown in this work 3.4. Electrochemical method could be useful in several applications. One application is the pro- duction of H O through an anthraquinone route [28–30]. Work − 2 2 The conducting polymer films are doped with PSS (i) during toward achieving this goal is currently underway in our laboratory. polymerization, via electro-oxidation of the monomer in the pres- ence of PSS− or (ii) by potential cycling in a solution. In our study, we treated the PEDOT-AQS-modified GC by electrochemical cycling in the range of 0.0 V to 1.0 V in a medium containing bulky PSS− (MW 70,000), for five cycles at a scan rate of 50 mV s−1. This resulted in a few outer layers of the modified electrode being doped with PSS−. This thin PSS− layer is expected to act as an anionic barrier layer. The treated electrode showed a significant increase in sta- − bility when cycled in 0.5 M H2SO4. The PSS treated film retained more than 85% of its redox activity after 100 cycles in 0.5 M H2SO4 (Fig. 6). To the best of our knowledge, the use of a thin anionic barrier film to retain the electroactivity of a dopant has not been previously reported in the literature. The two methods were compared as described below. The PEDOT-AQS-modified electrode retained maximum redox activ- ity after 100 potential cycles, using electrochemical treatment in a solution of PSS− (Fig. 7). This comparison suggests that PSS− solutions are more electrochemically effective in providing stable PEDOT-AQS films. It is interesting to note that even when PEDOT was made neutral, AQS− anions continued to stay in the film when PSS− was an outer layer. It is likely that a few outer layers of PEDOT undergo PSS− doping in exchange for AQS−. It appears that incor- − Fig. 7. Stability comparison of PEDOT-AQS modified by Nafion casting (physical) and poration of macromolecular PSS dopant into the electro-oxidized − PSS treatment (electrochemical) after 100 cycles in 0.5 M H2SO4 at the scan rate PEDOT is more effective in bringing about a barrier than a physical 20 mV s−1. Author's personal copy
3622 S. Harish et al. / Electrochimica Acta 54 (2009) 3618–3622
[3] M. Rafiee, D. Nematollahi, Electroanalysis 19 (2007) 1382. [4] D. DeLongchamp, P.T. Hammond, Adv. Mater. 13 (2001) 1455. [5] K. Yamamoto, M. Yamada, T. Nishiumi, Polym. Adv. Technol. 11 (2000) 710. [6] R. Mazeikiene,ˇ A. Malinauskas, Eur. Polym. J. 36 (2000) 1347. [7] H.K. Song, G. Tayhas, R. Palmore, Adv. Mater. 18 (2006) 1764. [8] T. Kitamura, M. Maitani, Y. Wada, S. Yanagida, Chem. Lett. 10 (2001) 1054. [9] R.W. Murray, in: A.J. Bard (Ed.), Electroanalytical Chemistry, vol. 13, Marcel Dekker Inc., NY, 1984. [10] L. Adamczyk, P.J. Kulesza, K. Miecznikowski, B. Palys, M. Chojak, D. Krawczyk, J. Electrochem. Soc. 152 (2005) E98. [11] D. Wakizaka, T. Fushimi, H. Ohkita, S. Ito, Polymer 45 (2004) 8561. [12] S. Ghosh, O. Inganäs, Synth. Met. 126 (2002) 311. [13] X. Crispin, S. Marciniak, W. Osikowicz, G. Zotti, A.W. Denier van der Gon, F. Louwet, M. Fahlman, L. Groenendaal, F. De Schryver, W.R. Salaneck, J. Polym. Sci. B: Polym. Phys. 41 (2003) 2561. [14] P.C. Wang, A.G. MacDiarmid, Synth. Met. 119 (2001) 367. [15] S.M. Ahmed, T. Nagaoka, K. Ogra, Anal. Sci. 14 (1998) 535. [16] S. Kuwabata, K.I. Okamoto, O. Ikeda, H. Yoneyama, Synth. Met. 18 (1987) 101. [17] T. Nagaoka, H. Nakao, T. Suyama, K. Ogura, Analyst 122 (1997) 1399. [18] N. Mogharrab, H. Ghourchian, Electrochem. Commun. 7 (2005) 466. [19] A.J. Bard, L.R. Faulkner, Electrochemical Methods. Fundamentals and Applica- − tions, 2nd ed., John Wiley & Sons Inc., New York, 2001. Fig. 8. FTIR spectra of (a) Na-PSS; (b) AQS in KBr and (c) PSS modified PEDOT- [20] E. Biilmann, Ann. Chim. 15 (1921) 103. AQS on an ITO glass plate for 20 cycles followed by cycling for five cycles in Na-PSS [21] R.J. Gostowski, Chem. Ed. 76 (1996) 1103. containing electrolyte. [22] H. Düssel, S.ˇ Komorsky-Lovric,´ F. Scholz, Electroanalysis 7 (1995) 889. [23] H. Kahlert, J.R. Pörksen, I. Isildak, M. Andac, M. Yolcu, J. Behnert, F. Scholz, Electroanalysis 17 (2005) 1085. Acknowledgement [24] J. Rishpon, S. Gottesfeld, C. Campbell, J. Davey, T.A. Zawodzinski Jr, Electroanal- ysis6(1994)17. [25] M. Yasuzawa, A. Kunugi, Electrochem. Commun. 1 (1999) 459. S. Harish thanks CSIR (India) for the award of Research Intern- [26] C.P.L. Rubinger, C.R. Martins, M.A. De Paoli, R.M. Rubinger, Sens. Actuat. B 123 ship. (2007) 42. [27] S.S. Kumar, C.S. Kumar, J. Mathiyarasu, K.L.N. Phani, Langmuir 23 (2007) 3401. [28] J.M. Campos-Martin, G.B. Brieva, J.L.G. Fierro, Angew. Chem. Int. Ed. 45 (2006) References 6962. [29] K. Vaik, A. Sarapuu, K. Tammeveski, F. Mirkhalaf, D.J. Schiffrin, J. Electroanal. [1] L.B. Groenendaal, F. Jonas, D. Freitag, H. Pielartzik, J.R. Reynolds, Adv. Mater. 12 Chem. 564 (2004) 159. (2000) 481. [30] K. Tammeveski, K. Kontturi, R.J. Nichols, R.J. Potter, D.J. Schiffrin, J. Electroanal. [2] P.D. Beer, P.A. Gale, G.Z. Chen, Coord. Chem. Rev. 3 (1999) 185. Chem. 515 (2001) 101. This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright Author's personal copy
Materials Research Bulletin 44 (2009) 1828–1833
Contents lists available at ScienceDirect
Materials Research Bulletin
journal homepage: www.elsevier.com/locate/matresbu
Generation of gold–PEDOT nanostructures at an interface between two immiscible solvents
S. Harish, J. Mathiyarasu *, K.L.N. Phani
Central Electrochemical Research Institute, Karaikudi 630006, Tamilnadu, India
ARTICLE INFO ABSTRACT
Article history: Gold–poly(3,4-)ethylenedioxythiophene (Au–PEDOT) composite with variant morphology was synthe- Received 3 February 2009 sized through interfacial polymerization at room temperature in the absence of a template, phase Received in revised form 28 May 2009 transfer catalyst or surfactant. A systematic variation of the relative concentration of the reactants yields Accepted 29 May 2009 morphologies typical of nanorods of diameter approximately 20–30 nm and length of few 0.5–1 mm. Available online 6 June 2009 Transmission electron microscopy reveals the structure of the nanowire to be the one in which the core is Au and the outer shell is made of PEDOT. Oligomer formation, speculated during the interfacial reaction Keywords: was confirmed by the analysis of the organic phase using ultraviolet–visible (UV–vis) and nuclear A. Composites magnetic resonance (NMR) spectroscopy techniques. A. Nanostructures B. Chemical synthesis ß 2009 Elsevier Ltd. All rights reserved. C. Electron microscopy D. Microstructure
1. Introduction Synthesis of various nanomaterials has been reported using this technique [19–28]. In the recent years, conducting polymer nanocomoposites have Several researchers have addressed the synthesis of nanos- received a great deal of attention due to their unique physical and tructured conducting polymer materials [29,30] and recently, Lu chemical properties and potential applications such as catalysis, et al. [31] have reported the synthesis of Au–PEDOT nanocables. nanoelectronic devices, magnetic devices, sensors and biomaterial Apart from conductivity analysis of the nanocables, they reported separation membranes [1–7]. Myriad methods have been devel- that some products arising from the oxidation of EDOT by HAuCl4 oped and reported for the synthesis of nanocomposite materials. dissolve in the aqueous medium initially to finally get accumulated The widely used template techniques, such as organic templates at the interface. Though the chemical nature of this substance was [8], ionic surfactants [9–10], hard template [11], polyelectrolytes conjectured to be an oligomer of EDOT, detailed analysis was not [12], seeding [5], etc. are commonly employed in the synthesis of reported. In the present study, the reaction between EDOT in nanostructured polymer composites, but such techniques have the organic phase and HAuCl4 in the aqueous phase not only leads to disadvantages of having subsequent removal of the templates to the formation of Au–PEDOT nanostructures at the interface but obtain the nanostructure. Techniques that avoid the use of a also enrichment of the organic phase with oligomer-like species, template are primarily electrochemical based [13,14] and electron when the organic phase is dichloromethane. Chemical analysis of beam lithography or electrospinning techniques [15–17]. These this organic-soluble species showed evidence for the formation of non-templating techniques have been of interest for a number of neutral EDOT oligomers. However when the organic phase is years, with their own merits and demerits [18]. Interfacial toluene, only the interfacial product is formed with the organic polymerization (IP), though a classical technique initially applied phase unaffected. In what follows, we present the results of the for the synthesis of condensation polymers, has not emerged as the work involving the interfacial reaction between the reactants in one that belongs to the class of non-templating techniques. IP has dichloromethane or toluene in the organic phase and the aqueous traditionally been performed using a organic phase—chloroform, phase. toluene, benzene, dichloromethane, etc. and an aqueous phase. In this work, we present our results on the synthesis of Au– PEDOT by a one-step interfacial reaction between EDOT and
HAuCl4 in dichloromethane/water biphasic system and character- ization of the interfacial product. In addition, interestingly, species * Corresponding author. Tel.: +91 4565 227550; fax: +91 4565 227779. arising out of the oxidation of EDOT are also formed in the organic E-mail address: [email protected] (J. Mathiyarasu). phase, when the latter is dichloromethane.
0025-5408/$ – see front matter ß 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.materresbull.2009.05.022 Author's personal copy
S. Harish et al. / Materials Research Bulletin 44 (2009) 1828–1833 1829
2. Materials and methods reaction because all the monomer would have been converted to the polymer at the interface or oligomers in the organic phase. 2.1. Synthesis of Au–PEDOT Similar observations have also been reported for polyaniline by Kaner and co-workers [32]. Formation of the interfacial product All chemicals were of analytical grade and used as received without any change in the colour of the bulk organic phase points from Sigma–Aldrich. One pot biphasic reaction was carried out in a to the fact that only the solid polymer is formed in this case. That 30-ml cylindrical glass container. About 10 mM of EDOT was some of the products or by-products are soluble in the organic dissolved in 10 ml of dichloromethane or toluene and equal phase leads us to speculate that these are of oligomeric nature. It concentration (10 mM) of HAuCl4 was dissolved in 10 ml of water. In a typical synthesis, above solutions were transferred to the reaction vessel one by one carefully. The course of the reaction was followed as a function of time at regular time intervals. The product formed at the interface and the oligomers in the organic phase were collected with respect to time. Interfacial product PEDOT and oligomers in the organic phase was subjected to various characterizations like transmission electron microscopy (TEM), scanning electron microscopy (SEM), X-ray diffraction (XRD), UV– vis and NMR. The effect of reductant concentration on the formation of Au–PEDOT was studied by increasing the concentra- tion of EDOT to 20 mM and 30 mM, whereas the concentration of
HAuCl4 (10 mM) was kept constant.
2.2. Microscopy and spectral experiments
SEM examination was made using Hitachi FE-SEM (Model S4700) with an acceleration voltage of 10 kV in normal mode and analysis done by placing the sample on the copper grid coated with carbon tape. TEM analysis was made by placing a drop of sample dispersed in acetone onto a copper grid coated with carbon film (400 meshes) and dried in vacuum. TEM images were collected from Philips CM200 microscope working at 200 kV. XRD patterns were recorded on a PANalytical diffractometer Model PW3040/60 X’pert PRO operating with Cu Ka radiation (l = 0.15406 nm) generated at 40 kV and 20 mA. Scans were done at 38 min 1 for 2u values between 208 and 908. For infrared and Raman spectroscopic measurements, a Thermo-Electron Corporation make FT-Raman module (InGaAs detector andNd:YVO4 laser operating at 1064 nm) coupled with a Nexus 670 model FT-IR spectrometer (DTGS detector) was used. NMR spectra were recorded in 400 MHz Bruker
NMR Spectrometer with CDCl3. Ultraviolet–visible (UV–vis) absorption spectra of the oligomer in dichloromethane were collected on a Cary 500 scan UV–vis–NIR spectrophotometer with incident light normal to the 1-cm path length quartz cell. Spectra were collected in the wavelength range of 200–800 nm.
3. Results and discussion
3.1. Synthesis
Aqueous/organic biphasic interface is formed with EDOT in the organic phase and auric chloride in the aqueous phase. A range of organic solvents can be used in the formation of biphasic systems. By choosing a denser organic solvent, it is possible to have the aqueous phase over the organic phase. In a typical synthesis, EDOT is dissolved in dichloromethane and auric chloride in the aqueous phase in equal concentration (10 mM). Equal volume of the aqueous phase is carefully transferred to the organic phase in one- pot to form the aqueous/organic interface. In a short period of time, blue colouration occurs at the interface, due to the formation of PEDOT and the colour intensity increases with respect to time. As the reaction proceeds, the colour of the organic phase also gradually changes from colourless to light blue and then becomes dark blue. The blue colouration occurs throughout the organic phase indicating accumulation of some species arising from the oxidation of EDOT. Also, the invariance of the colour at the Fig. 1. SEM pictures of Au–PEDOT powder (scale bar of 1 mm); concentration ratio interface and in the organic phase indicates completion of the of EDOT and HAuCl4: (A) 1:1, (B) 2:1, and (C) 3:1. Author's personal copy
1830 S. Harish et al. / Materials Research Bulletin 44 (2009) 1828–1833 may be appropriate to recall from evidence in literature that such head and tail and this is understandable considering the ‘‘drag’’ by-products soluble in organic media can be oligomeric in nature of aqueous HAuCl4 into the organic phase. As small amounts of [33,34]. HAuCl4 are sufficient to effect oxidation and polymerization of EDOT, higher amounts only lead to more and more EDOT 3.2. Characterization of interfacial product undergoing oxidative polymerization. As can be seen, the diameter/shape of the nanostructures are affected by the reductant Fig. 1 shows the FE-SEM images of the interfacial product concentration, i.e., EDOT. With an increase in the [EDOT]/[Au] ratio formed during the course of interfacial polymerization reaction. from 1:1 to 2:1 and 3:1, a remarkable influence on the morphology The images in Fig. 1 show that the nanostructures of Au–PEDOT of the obtained Au–PEDOT products is observed. At a high become denser as the ratio of EDOT:HAuCl4 increases. Fig. 1A concentration of EDOT, we believe that the aggregation of PEDOT particularly shows that the product, i.e., nano Au–PEDOT has a results in simple nanoparticle fusion or growth, leading to larger gold–PEDOT nanocomposites; while at low concentration, there may not be sufficient polymer to stabilize the NPs. The formation of gold nanoparticles was confirmed by TEM. The TEM images show that an increase in the concentration of the reductant (i.e., EDOT) leads to a network structure of rod- or fibrillar shapes. EDOT:Au (1:1) ratio yields a network–like structure within which AuNPs are entrapped. The dark spots in the images show Au embedded in the polymer matrix. Reaction at a ratio of EDOT:Au = 2:1 leads to the formation of nanostructured rods with diameter ranging from 20 to 30 nm; the rod length varying from 0.5 to 1 mm. Further increase in the EDOT content, i.e., 3:1 ratio, the number of fibrils formed increases with the resultant nanofibers of 10–20 nm diameter and several micrometers of length (Fig. 2C). In Fig. 3A, the image of a single nanofiber is shown at higher magnification, which reveals the polymer shell and Au core in the
Fig. 2. TEM picture of Au–PEDOT powder (scale bar of 200 nm) (A) 1:1, (B) 2:1, and Fig. 3. (A) TEM picture of single nanofiber of Au–PEDOT formed in EDOT/Au (3:1
(C) 3:1 concentration ratio of EDOT and HAuCl4. ratio) (scale bar of 50 nm) and (B) SAED pattern of Au–PEDOT powder. Author's personal copy
S. Harish et al. / Materials Research Bulletin 44 (2009) 1828–1833 1831
Fig. 6. FT-Raman spectra of EDOT monomer and Au–PEDOT nanocomposite.
of EDOT at a,a0 position. Vibrations at 1520, 1335 cm 1 are attributed to the stretching modes of C55C and C–C bonds in the quinoidal structure of thiophene ring. The vibration modes of the C–S bond in the thiophene ring can be seen at 978, 840 and 692 cm 1. The bands at 1207, 1091 cm 1 and 927 cm 1 are assigned to the stretching modes of the ethylenedioxy group, and its deformation respectively [35,36]. The FT-Raman spectrum of Au–PEDOT nanocomposite and EDOT monomer are presented in Fig. 6. As expected the EDOT monomer shows strong bands at 1485, 1424, 1185, 891, 834, and 763 cm 1. In the Raman spectrum of Au–PEDOT, the strong peak at 1420 cm 1 is attributed to the symmetric stretching vibration of 1 Ca55Cb and weak peak at 1512 cm for asymmetric stretching vibration of Ca55Cb. Other weak peaks observed at 1361, 1262, 1085, 994, 700 are due to Cb55Cb stretching, C–C inner ring stretching, C–O–C deformation and symmetric C–S–C deformation, Fig. 4. (A) XRD pattern of Au–PEDOT powder and (B) energy dispersive analysis of respectively. This spectral information confirms that PEDOT in the Au–PEDOT composite. Au–PEDOT is in the highly doped state [37–39].
fiber of diameter ranging from 8 to 10 nm. Upon focussing on the 3.3. Characterization of organic phase-soluble species nanofiber for selected area electron diffraction (SAED) (Fig. 3B), the polycrystalline nature of the Au nanostructures becomes evident. Au–PEDOT formation takes place during the redox reaction Further, X-ray diffraction analysis of Au–PEDOT (Fig. 4A) also between the Au3+ and EDOT. This reaction will occur only at the confirms the polycrystalline nature of Au with intense peaks at interface between the two phases. Dual solubility (organic and 2u = 38.178, 44.378, 64.568, 77.568 and 81.718 corresponding to aqueous) behaviour of EDOT will be more helpful for partial reflections: (1 1 1), (2 0 0), (2 2 0), (3 1 1) and (2 2 2) of fcc phase of diffusion to the aqueous phase and after equilibration EDOT in the Au. The energy dispersive X-ray analysis of Au–PEDOT (Fig. 4B) interfacial region undergoes a redox reaction with Au3+, resulting shows the presence of Au, S and O. in the formation of Au–PEDOT. The Au–PEDOT formation increases Fig. 5 shows the FT-IR spectra of the Au–PEDOT film and EDOT as time increases and subsequently oligomer formation will occur monomer respectively. It is clear that the strongly intense band to when a polar organic solvent is employed. Completely soluble the C–H bending mode (in-plane and out-of-plane) at 891 cm 1 oligomer forms only in the organic medium because solubility of and the band at 1186 cm 1 disappear in the polymer spectrum in EDOT is poor in aqueous medium compared to that in the organic comparison with that of the monomer, demonstrating the coupling medium [40]. In addition, the oligomers do not form if we employ a non-polar solvent like toluene. This can be reasoned on the basis of the dielectric constant of the medium because mix-up at the interface depends upon the dielectric constant of the solvent. It is well known that if the dielectric constant is high for the solvent then the mix up at the interface will be more and vice versa [41]. Formation of by-products soluble in organic phase can be seen easily by a colour change because the oxidized form of EDOT exhibits blue colour [42]. Hence, in a typical synthesis carried out (as given in Section 2) using dichloromethane whose dielectric constant and solubility in water (g/100 g) are given as 9.08 and 1.32, respectively. It is observed that in a short period of time blue coloured PEDOT is formed at the interface and PEDOT formation increases with respect to time. As the reaction proceeds, the colour of the organic phase also gradually changes to light blue to become Fig. 5. FT-IR spectra of EDOT monomer and Au–PEDOT nanocomposite. dark-blue coloured finally. The blue coloured products remain in Author's personal copy
1832 S. Harish et al. / Materials Research Bulletin 44 (2009) 1828–1833
Photograph 1. Biphasic systems showing the colour change in dichloromethane while no colour changes occur in Toluene. (A: dichloromethane/water system; B: toluene/ water system). the organic phase showing the formation of some kind of calculation of the protons in EDOT, i.e., four protons in the ethylene oligomeric species in the organic phase and also confirms the bridge and two protons in the aromatic ring. Oligomer/polymer completion of the reaction because all the monomers would have formation takes place by coupling at the a,a0 positions in the been converted either to a polymer at the interface or oligomers in monomer. In the oligomer, protons are present only in the ethylene the organic phase. Similar experiment was carried out using bridge and no protons in the aromatic ring. Hence, NMR spectra for toluene whose dielectric constant and solubility in water (g/100 g) the oligomers show a single peak at 5.3. The environment are given as 2.38 and 0.08, respectively. Since the miscibility or experienced by the protons in the ethylene bridge in the monomer solubility of toluene in water is extremely small, the mix-up of and oligomer is different. In oligomers, protons are in a more the two phases at the interface is not expected. It is interesting deshielded environment than monomer. Hence, the peak in the to note that when toluene is employed, there is only the formation oligomer shifts to the lower field when compared to the monomer. of interfacial product and the colour of the organic phase The NMR spectra confirm the oligomeric nature of the products remains unchanged indicating that oligomers are not formed soluble in the organic phase.
(Photograph 1). As suggested above, the oligomer forms only in The UV–visible spectra of the EDOT oligomers in CH2Cl2 (Fig. 8) the high dielectric solvent and does not form in the low dielectric show two absorption peaks at 571 and 617 nm indicating that solvent. This effect can be reasoned as follows: in high dielectric there are two different species present in the organic phase [43]. solvents, mix-up at the interface is more and so interfacial reaction In addition, the broadness of the peak around 600 nm suggests a extends to the formation of oligomers in the organic phase. p–p* transition and the absence of the free tail in the higher Monomer (EDOT) contains two different proton environments, wavelength shows that the species are in the neutral state [38,44]. one in the aromatic ring and the other in the ethylene bridge. It is reported that neutral polymers/oligomers are soluble in the Hence NMR spectra of EDOT (Fig. 7) shows two peaks at 4.2 and 6.4 organic solvents [45]. With an objective of exploring further on which are due to the ethylene bridge protons and aromatic protons this important finding, we are generating different organic- respectively. The peak values are referred from Bruker table. By soluble species in a range of solvents (organic phase), character- integrating the area under the peak, the number of protons can be izing the chain length and other issues using in situ spectral calculated which works out to be in the ratio of 2:1. It matches the studies.
Fig. 7. NMR spectra of EDOT and EDOT oligomers in CH2Cl2/CDCl3. Author's personal copy
S. Harish et al. / Materials Research Bulletin 44 (2009) 1828–1833 1833
[3] A.N. Aleshin, Adv. Mater. 18 (2006) 17. [4] R. Xiao, S.I. Cho, R. Liu, S.B. Lee, J. Am. Chem. Soc. 129 (2007) 4483. [5] X. Zhang, S.K. Manohar, J. Am. Chem. Soc. 127 (2005) 14156. [6] S. Ko, J. Jang, Angew. Chem. Int. Ed. 45 (2006) 7564. [7] M. Steinhart, Z. Jia, A.K. Schaper, R.B. Wehrspohn, U. Go¨sele, J.H. Wendorff, Adv. Mater. 15 (2003) 706. [8] W. Zhong, J. Deng, Y. Yang, W. Yang, Macromol. Rapid Commun. 26 (2005) 395. [9] Z. Wei, Z. Zhang, M. Wan, Langmuir 18 (2002) 917. [10] M.G. Han, S.H. Foulger, Small 10 (2006) 1164. [11] M. Ikegame, K. Tajima, T. Aida, Angew. Chem. Int. Ed. 42 (2003) 2154. [12] K. Mu¨ ller, M.-K. Park, M. Klapper, W. Knoll, K. Mu¨ llen, Macromol. Chem. Phys. 208 (2007) 1394. [13] M. Fujii, S. Abe, H. Ihori, Synth. Met. 152 (2005) 41. [14] M. Woodson, J. Liu, J. Am. Chem. Soc. 128 (2006) 3760. [15] D.H. Reneker, I. Chun, Nanotechnology 7 (1996) 216. [16] I.D. Norris, M.M. Shaker, F.K. Ko, A.G. MacDiarmid, Synth. Met. 114 (2000) 109. [17] X. Zhang, A.G. MacDiarmid, S.K. Manohar, Chem. Commun. (2005) 5328. [18] M. Wan, Adv. Mater. 20 (2008) 2926. [19] U. Sree, Y. Yamamoto, B. Deore, H. Shiigi, T. Nagaoka, Synth. Met. 131 (2002) 161. [20] M. Nakata, Y. Shiraishi, M. Taga, H. Kise, Makromol. Chem. 193 (1992) 765. [21] J. Huang, R.B. Kaner, J. Am. Chem. Soc. 126 (2004) 851. Fig. 8. UV–vis spectrum of EDOT oligomer in CH2Cl2. [22] J. Jang, J. Bae, E. Park, Adv. Mater. 18 (2006) 354. [23] N. Nuraje, K. Su, N.-L. Yang, H. Matsui, ACS Nano 2 (2008) 502. [24] J.M. Pringle, O. Ngamna, C. Lynam, G.G. Wallace, M. Forsyth, D.R. MacFarlane, Macromolecules 40 (2007) 2702. 4. Conclusion [25] X. Feng, H. Huang, Q. Ye, J.J. Zhu, W. Hou, J. Phys. Chem. C 111 (2007) 8463. [26] D.D. Sawall, R.M. Villahermosa, R.A. Liepeles, A.R. Hopkins, Chem. Mater. 16 In summary, we reported here the generation of various (2004) 1606. morphologies of Au–PEDOT nanocomposite at an interface [27] R. Knake, A.W. Fahmi, S.A.M. Tofail, J. Clohessy, M. Mihov, V.J. Cunnane, Langmuir 21 (2005) 1001. between aqueous and organic phases. The effect of EDOT:Au ratio [28] C.R.K. Rao, D.C. Trivedi, Synth. Met. 157 (2007) 432. on the morphology is discussed using the microscopic and XRD [29] B.C. Sih, M.O. Wolf, Chem. Commun. (2005) 3375. data. The Au–PEDOT nanocomposite was characterized by Raman [30] R. Gangopadhyay, A. De, Chem. Mater. 12 (2000) 608. [31] G. Lu, C. Li, J. Shen, Z. Chen, G. Shi, J. Phys. Chem. C 111 (2007) 5926. spectroscopy to be in highly oxidized state. While the formation of [32] J. Huang, S. Virji, B.H. Weiller, R.B. Kaner, J. Am. Chem. Soc. 125 (2003) 314. AuNPs embedded in PEDOT matrix is observed to be straightfor- [33] J. Heinze, in: H. Lund, O. Hammerich (Eds.), Organic Electrochemistry, 4th ed., Dekker, New York, 2001. ward, a byproduct of HAuCl4–EDOT reaction appears to form in the [34] L. Groenendaal, F. Jonas, D. Freitag, H. Pielartzik, J.R. Reynolds, Adv. Mater. 12 organic phase (dichloromethane) of relatively higher dielectric (2000) 481. constant. Formation of such a species is not observed when the [35] X. Li, Y. Li, Y. Tan, C. Yang, Y. Li, J. Phys. Chem. B 108 (2004) 5192. organic phase is toluene, a low dielectric solvent. The dichlor- [36] C. Kvarnstro¨m, H. Neugebauer, S. Blomquist, H.J. Ahonen, J. Kankare, A. Ivaska, Electrochim. Acta 44 (1999) 2739. omethane-soluble species was characterized using UV–vis spec- [37] N. Sakmeche, S. Aeiyach, J.-J. Aaron, M. Jouini, J.C. Lacroix, P.-C. Lacaze, Langmuir troscopy and proton NMR technique. The latter suggests that the 15 (1999) 2566. byproducts are short chain neutral oligomers of EDOT. [38] S. Garreau, G. Louam, S. Lefrant, J.P. Buisson, G. Froyer, Synth. Met. 101 (1999) 312. [39] W.W. Chiu, J. Travasˇ-Sejdic´, R.P. Cooney, G.A. Bowmaker, J. Raman Spectrosc. 37 Acknowledgement (2006) 1354. [40] S.S. Kumar, C. Sivakumar, J. Mathiyarasu, K.L.N. Phani, Langmuir 23 (2007) 3401. S.H. thanks CSIR, New Delhi, for the award of Research [41] A.L. Horvath, Halogenated Hydrocarbons: Solubility–Miscibility with Water, CRC Internship. Press, 1982. [42] M.G. Han, S.H. Foulger, Chem. Commun. (2004) 2154. [43] F. Tran-Van, S. Garreau, G. Louarn, G. Froyer, C. Chevrot, J. Mater. Chem. 11 (2001) References 1378. [44] S. Garreau, G. Louarn, J.P. Buisson, G. Froyer, S. Lefrant, Macromolecules 32 (1999) [1] C.R. Martin, Acc. Chem. Res. 28 (1995) 61. 6807. [2] J.C. Hulteen, C.R. Martin, J. Mater. Chem. 7 (1997) 1075. [45] M. Sato, S. Tanaka, K. Kaeriyama, J. Chem. Soc., Chem. Commun. (1986) 873. Journal of Electroanalytical Chemistry 659 (2011) 128–133
Contents lists available at ScienceDirect
Journal of Electroanalytical Chemistry
journal homepage: www.elsevier.com/locate/jelechem
Nix–Fe(1 x)Fe(CN)6 hybrid thin films electrodeposited on glassy carbon: Effect of tuning of redox potentials on the electrocatalysis of hydrogen peroxide ⇑ Alam Venugopal Narendra Kumar, Srinivasan Harish, James Joseph , Kanala Lakshminarasimha Phani
Electrodics and Electrocatalysis Division, CSIR-Central Electrochemical Research Institute, Karaikudi 630 006, India article info abstract
Article history: The electrochemical deposition of mixed Nix–Fe(1 x)Fe(CN)6 thin films on a glassy carbon (GC) electrode Received 8 February 2011 surface were carried out by electro-deposition under continuous potential cycling condition for the first Received in revised form 6 May 2011 time. The redox potential of the low spin iron in the hybrid film was found to depend on the ratio of metal Accepted 11 May 2011 centres. The improvement in the electrocatalytic property of the nickel hexacyanoferrate film with incor- Available online 19 May 2011 poration of iron ions in the film towards the electrocatalysis of hydrogen peroxide was studied in elec- + trolytes like KNO3 and NaNO3. The stability of hybrid film in Na containing supporting electrolytes Keywords: was examined. Detection limit of H O was about 1 lM as determined by amperometry with the sensi- Chemically modified electrodes 2 2 tivity of 192 nA/ M. The chemically modified glassy carbon electrodes with Ni Fe HCF hybrid films MHCF hybrids l x (1 x) were characterized by cyclic voltammetry, impedance spectroscopy, energy dispersive X-ray analysis H2O2 sensor Electrocatalysis etc. These films may find extensive applications as redox mediators in biosensors which employ oxidase enzymes. Ó 2011 Elsevier B.V. All rights reserved.
1. Introduction metal hexacyanoferrates with combinations such as Fe–Ni, Ni–Pd, and Ni–Tl were reported [8–10]. Dostal et al. demonstrated that The ‘metal hexacyanoferrate’ (MHCF) is a class of compounds the Prussian blue modified electrode can undergo lattice substitu- used for modification of electrochemical interfaces for the last tion during cycling in Cd2+ containing solutions [11]. James et al. 20 years due to their interesting ion transport, electrocatalytic, showed that the Prussian blue (PB) modified GC electrode can be electrochromic and photoelectrochemical properties [1,2]. Possi- converted to CuHCF modified GC completely on cycling in Cu2+ bility of using simple electrochemical control for preparing thin solution [12]. Electrocatalytic hydrogen peroxide reduction on films of these compounds from a bath containing potassium ferri- pure and mixed hexacyanoferrates of Cu, Ce and Pd were reported cyanide and corresponding metal salt is exciting. In prototype tran- [13,14]. Bharathi et al. prepared mixed analogues of PB-NiHCF and sition metal hexacyanoferrate, ferric ferrocyanide (FeHCF) the iron PB-MnHCF mixed metal hexacyanoferrate electrodes [15]. They coordinated with –N of cyanide ligand is termed as high spin iron formed the mixed analogues of Prussian blue by potential cycling and the other which is co-ordinated to –C of the cyanide is low spin the glassy carbon electrode in a neutral medium containing metal iron, established from Mossbauer and infrared studies [2]. salt and potassium ferricyanide to a very high anodic potentials The intense blue colour in FeHCF is due to the charge transfer where the ferricyanide decomposes to form ferric ions. The substi- from Fe(II) in a carbon hole to Fe(III) in a nitrogen hole [3]. Modi- tution of PB lattice by hetero atom is expected to alter the proper- fication of glassy carbon substrates with metal hexacyanoferrates ties of the resulting mixed MHCF. However, there are not much were reported from our laboratory recently [4,5]. The modifica- systematic studies on the effect of substitution of PB lattice with tions of the surfaces with metal hexacyanoferrates (MHCF) gain another metal on the electrochemical properties. Scholz and Reddy further importance owing to the fact that the interface can be mod- prepared the mixed crystals of Fe–NiHCF with various ratios of Ni2+ ified with two MHCFs simultaneously. Electrochemical modifica- and Fe3+ by chemical precipitation and incorporated these precip- tion of electrode with NiHCF prepared in presence of Ag+ make itate onto electrode by mechanical abrasion. Though the authors the electrode less catalytic to Fe2+ oxidation [6]. Cataldi et al. re- were able to show correlation between formal potential of the ported that the CoHCF modified electrode on cycling in RuCl3 solu- mixed crystal on composition, their voltammetric response show tion gains extra stability without alteration in the electrochemical features of the existence of a small fraction of individual MHCF re- properties [7]. Modification of electrodes with mixed analogues of sponse also. In this work, we have demonstrated a simple electro- chemical approach for the electrochemical formation of thin films of hybrid films of Ni–Fe–hexacyanoferrate directly onto the sub- ⇑ Corresponding author. Tel.: +91 4565 227550x441; fax: +91 4565 227779. E-mail address: [email protected] (J. Joseph). strates. This film has applications in catalyzing the hydrogen
1572-6657/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.jelechem.2011.05.006 A.V. Narendra Kumar et al. / Journal of Electroanalytical Chemistry 659 (2011) 128–133 129 peroxide oxidation/reduction electrochemically. Recently Karyakin The low spin iron (peak-II) coordinated with –C of the ligand et al. have demonstrated the record performance of Prussian blue undergo redox reaction at higher potential is given below. modified electrodes in the detection of nanomolar level of hydro- II II Anodic III II fgK2Fe ½ Fe ðÞCN 6 fgKFe ½ Fe ðÞCN 6 þ gen peroxide by electroreduction [16]. The PB type compounds HS LS ðsÞ HS LS ðsÞ þK þ e ð2Þ Prussian blue Cathodic Prussian blue are termed as ‘artificial peroxidase’ [17]. The redox potential of the ferro/ferricyanide couple was found Similarly GC electrodes can be modified with thin films of to be dependent on the composition of metal ions in the hybrid NiHCF by electrochemical cycling of electrode potential between films. Tuning of redox potential can be advantageously used for 0.0 and 1.0 V in the medium containing 1 mM of K3[Fe(CN)6] and creating new electrocatalytic interfaces. The ferric ferrocyanide 1 mM NiCl2. NiHCF modified electrode yield two sets of redox pro- has two electron transport channels namely high spin iron and cesses [19]. Fig. 1A shows the response of the NiHCF modified elec- 3+ 2+ low spin iron [18]. The introduction of heteroatoms Fe on the trode in KNO3. The Ni does not undergo redox reaction under NiHCF network is expected to open up a new electron transport these experimental conditions. Hence the two redox processes fer- channel in the hybrid film. The ion transport properties of the hy- ro/ferricyanide are arising due to the presence of NiHCF with two brid film can be altered by the incorporation of selected metal ion different stoichiometric ratios between Ni and Fe [7,8]. The redox in the MHCFs lattice. Considerable improvement of hybrid’s elec- potential of the ferro/ferricyanide couple showed an anodic shift trochemical stability in sodium containing (Na+) electrolytes com- as a function of the concentration of iron ion in the modification pared to that of PB is reported. Further, the effect of increasing iron mixture as seen from Fig. 1B–E. Linear variation of voltammetric content in the hybrid on the creation of electrocatalytic interface peak potential of low spin iron as a function of composition of iron towards oxidation/reduction of hydrogen peroxide is described is observed as shown in Table 1. (Eqs. (5) and (6)). When the GC electrode is cycled in the modification bath con- taining ferricyanide ions and mixture of NiCl2 and FeCl3, we would 2. Experimental expect the formation of NiHCF and FeHCF independently on the electrode surface. To our surprise, we observed two redox centres, The electrochemical experiments were performed in a three one corresponding to redox process due to high spin iron and the electrode cell using a potentiostat, Autolab Model PGSTAT 30, other corresponding to low spin iron when the GC electrode was Netherlands. The working electrode was a glassy carbon (GC) disc cycled in 0.1 M KNO3 medium containing 1 mM K3[Fe(CN)6] and electrode (Bio Analytical Systems, USA, area 0.07 cm2). The work- 0.5 mM of FeCl3 and 0.5 mM NiCl2 between the potential limit of ing electrode was polished with 4/0 grade alumina coated emery 1.0–0.0 V at 0.05 V/s as seen from Fig. 1C. There was no broadening paper. The polished electrode was subsequently cleaned by ultra or splitting of the redox process due to low spin iron centre. sonication in Millipore water for 2–3 min. A large area platinum Occurrence of a single peak for the redox reaction correspond- foil fused to a glass tube served as counter electrode. All the poten- ing to low spin iron is surprising. The redox peak potential corre- tials were measured with respect to 1 M calomel electrode (NCE). sponding to low spin iron in NiHCF films is shifted to more All the solutions were prepared fresh in Millipore water of resistiv- anodic values with the amount of ferric ion in the modification ity 18.2 M X cm). The electrodes were modified by cycling the mixture. The possibility of merging two redox processes of NiHCF working electrode potential between 1.0 V and 0.0 V at scan rate and FeHCF is ruled out as we did not observe any splitting of redox peaks in the voltammetric response of the hybrids even at sweep of 0.05 V/s for 20 cycles, in 0.1 M KNO3 medium containing 0.01 M HCl, 1 mM potassium ferricyanide and 1 mM of metal chlo- rates as low as 1.0 mV/s. The full width at half maxima (FWHM) rides (total concentrations of ferric chloride and nickel chloride in is a parameter indicating the multiplicity of redox process or the the medium is 1 mM for corresponding hybrid synthesis). The films attractive or repulsive interaction present in the film between re- deposited on the electrodes were characterized for their electro- dox sites [20]. The reversible surface redox process is expected to catalysis towards hydrogen peroxide by cyclic voltammetry. exhibit FWHM of 90.6/n mV. We observed FWHM of 100– Energy-dispersive X-ray analysis (EDX) for the hybrids were con- 130 mV for the hybrid films which may be due to some repulsive ducted on films electrodeposited on glassy carbon (GC) electrode interaction present in the hybrid films. embedded in Teflon. AC impedance experiments for GC modified The chemical precipitation of hybrids containing FeHCF–NiHCF hybrid is performed by Reddy et al. The X-ray diffraction studies of with films of FeHCF and Nix–Fe(1 x)Fe(CN)6 were performed using an impedance analyser (Model IM6-Bio Analytical Systems, USA) these hybrids clearly proved the incorporation of second metal in in the frequency ranges from 1 mHz to 10 Hz. The amplitude was the lattice of the primary metal hexacyanoferrate [21]. The forma- fixed at 10 mV (Number of data points per decade: 4). The imped- tion of these hybrids by electrodeposition is complex due to the ance spectra were recorded at applied bias of mid-peak potential fact that the MHCFs are formed on the electrode surface by elec- i.e. (Epa + Epc)/2. tro-reduction. To rule out the possibility of the formation of MHCF deposition at the initial potential, we have started the initial poten- tial of cycling from 1.0 V in our experiment which ensures that 3. Results and discussion there is no film formation on the electrode initially. Here we wish to state that there was no film formation noticed on GC surface Fig. 1A–F shows the cyclic voltammetric responses of NiHCF and when we potentiostated the electrode at 1.0 V vs. NCE. During FeHCF and their hybrid modified GC electrodes by the procedure reductive cycling, the ferricyanide ion in the medium gets reduced described in the ‘experimental section’ in 0.1 M KNO . The origin 3 to ferrocyanide ion. The solubility product of many transition me- of redox processes of the modified electrodes are elaborately dis- tal ions with ferrocyanide ion is less than the corresponding ferri- cussed in the literature [1,2]. FeHCF has two Fe redox sites one cyanide ion [22]. Hence both the metal ions simultaneously get coordinated with carbon and the other coordination to the nitro- precipitated on the electrode electrochemically as follows.At po- gen of the cyanide ligand. The high spin iron (peak-I) coordinated tential below 0.3 V, with –N of the ligand undergo redox reaction at around 0.15 V as follows. 3 4 ½FeðCNÞ6 þ e !½FeðCNÞ6 ð3Þ II II Anodic III II fgK2FeHS½ FeLSðÞCN 6 fgKFeHS½ FeLSðÞCN 6 þ ðsÞ ðsÞ þK þ e ð1Þ 4 3þ 2þ Everitt0s salt Cathodic Prussian blue ½FeðCNÞ6 þð1 xÞFe þ x Ni ! NixFeð1 xÞFeðCNÞ6 ð4Þ 130 A.V. Narendra Kumar et al. / Journal of Electroanalytical Chemistry 659 (2011) 128–133
20 30 A D 10 20
0 10
-10 0
-10 -20 -20 -30 30 40 E B 30 20 20
A 10 10 μ 0 0 -10 -10
Current, -20 -30 -20 -40 -30
30 80 C 60 F 20 40 10 20
0 0 -20 -10 -40 -20 -60 -80 -30 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 Potential, V vs NCE Potential, V vs NCE
Fig. 1. Voltammetric response of modified GC electrodes in 0.1 M KNO3 medium at the scan rate of 0.02 V/s. Modified mixture containing 1 mM of K3[Fe(CN)6] + 0.1 M
KNO3 + 0.01 M HCl and (A) 1 mM NiCl2 (B) 0.7 mM NiCl2 + 0.3 mM FeCl3 (C) 0.5 mM NiCl2 + 0.5 mM FeCl3 (D) 0.3 mM NiCl2 + 0.7 mM FeCl3 (E) 0.1 mM NiCl2 + 0.9 mM FeCl3 (F)
1 mM FeCl3.
Table 1 In most of the papers dealing with mixed MHCF modification, 2+ Ep values of peak-I and peak-II and potential shift data as a function of Ni % in the redox transition of ferro/ferricyanide takes place at similar poten- modification mixture. tials [8–10]. In the case of Ni (Fe) HCF modified electrodes, intro- 2+ Ni DEp (peak-I) DEp (peak-II) DEp (peak-I) DEp (peak-II) duction of Fe opens up a new electron transport channel in the (%) mV mV mV mV lattice of the hybrid. On introducing the iron ions in the NiHCF lat- 0 40 100 90 810 tice, the following changes in response of the modified electrode 10 40 50 100 760 are noticed. (i) Anodic shift of the redox peak potential correspond- 30 40 20 105 690 ing to ferro/ferricyanide redox centre to more positive direction. 50 35 10 115 655 70 – 50 – 540 (ii) There exists only one redox process for the low spin iron unlike 90 – 75 – 500 the two redox processes observed in NiHCF modified GC. (iii) 100 – 25 – 520 Appearance of a set of new redox peak at 0.15 V corresponding to introduction of new electron transfer channel due to incorpora- tion of high spin iron (above 30% of iron). Table 1 shows the peak to peak separation and the redox peak shift of redox processes of the Bharathi et al. have formed the mixed PB analogues by subject- modified electrode as a function of Ni2+%. When the modification ing the GC in a mixture containing NiCl +K Fe(CN) to extreme 2 3 6 mixture for NiHCF contain iron ion in it we observed a regular ano- positive potential of 1.5 V vs. NCE which decompose the ferricya- dic shift in the redox potentials of ferro/ferricyanide redox centre nide ions to free ferric ion [15]. Decomposition of ferricyanide to in the hybrid MHCF (see Table 1). This kind of tuning of redox po- free ferric ion at extreme positive potential is first reported by tential as a function of composition of the film is not reported Gomathi and Rao [23]. The formation of mixed analogue takes elsewhere. place in two steps. In this work, it is possible to observe signatures of individual M-HCFs independent of others in the work. The charge trapping effects were reported for MHCF films formed as 3.1. Electrocatalysis of Nix-Fe(1 x)Fe(CN)6 modified electrodes bilayers [12,24]. In our case the Prussian blue and NiHCF are formed as hybrids by simultaneous co-electrodeposition and the The Prussian blue modified electrode is known to electrocata- charge trapping effects were not observed for our films. The ratio lyze hydrogen peroxide oxidation or reduction very effectively in of the metal ions in the film was assumed to be the same as that comparison with biocatalyst. So the Prussian blue can be used in present in the modification mixture. This assumption was found biosensor application which indirectly acts as ‘artificial peroxidase’ to be true from EDX studies of the hybrid films. in the detection of hydrogen peroxide [15,25]. Our group has A.V. Narendra Kumar et al. / Journal of Electroanalytical Chemistry 659 (2011) 128–133 131
Fig. 2. (A, B, C-thick lines) Response for Fe–HCF, Ni–HCF and Fe0.5–Ni0.5–HCF in 0.1 M KNO3 medium. (A, B, C-dotted line) shows voltammogram of Fe–HCF, Ni–HCF and Ni0.5–
Fe0.5 HCF in presence of 1 mM hydrogen peroxide.
-12 14 2 12
cm 10 -9 Ω
2 8 |Z| 6 A cm -6 100 1k 10k 100k 1M
Ω ln f 10.3kHz
Z’’, -3
0 Fe-HCF
-12 14
2 12 cm
2 10
-9 Ω
|Z| 8 cm 6 Ω -6 B 100 1k 10k 100k 1M 6.54kHz ln f Z’’, -3 Fe -Ni - HCF 0 0.5 0.5 46810121416 Z’, Ω cm2
Fig. 3. Amperometric response of hybrid Ni0.5–Fe0.5 HCF modified GC electrode at applied potential of 0.09 V vs. NCE. Inset shows a linear plot of current vs. Fig. 5. Nyquist plot for (A) Fe–HCF and (B) Ni0.5–Fe0.5HCF modified GC electrode in the frequency range 1 mHz–10 Hz. Inset figure Bode plot (|Z| vs. ln f) are shown for concentration of H2O2 (from amperometry). both Fe–HCF and Ni0.5–Fe0.5HCF modified GC electrodes . reported an electrochemical preparation of PB-Au nanocomposite tion at FeHCF modified electrode. However, the NiHCF modified for electrocatalysis of hydrogen peroxide in the nano molar levels electrode does not show significant catalytic current due to hydro- [26]. The application of MHCF modified electrode to catalysis is gen peroxide reduction nor oxidation. limited due to the fact that the electrochemical stability of the The Ni Fe HCF modified electrode showed almost similar modified film highly depends on the supporting electrolyte cation. 0.5 0.5 hydrogen peroxide oxidation currents at slightly lower potential. Fig. 2 dotted lines depict the electrocatalytic oxidation/reduction of The peak to peak separation for the low spin iron centre in Fe 1mMH O using FeHCF, NiHCF and Fe Ni HCF modified GC in 2 2 0.5 0.5 HCF is 90 mV compared to 20 mV for Fe Ni HCF which clearly KNO medium. From the voltammetric response, it is clear that the 0.5 0.5 3 indicate improved reversibility of redox process by the Ni2+ incor- hydrogen peroxide can undergo either electro-oxidation or reduc- poration (see Table 1). From the above, it is evident that the elec- trocatalytic properties for the NiHCF modified electrode can be significantly altered by the incorporation of second metal ion 15 which shifts the redox potentials to more favourable values. Our 10 experiments show that the hydrogen peroxide oxidation/reduction 3+ A 5 started taking place after 30% incorporation of Fe in NiHCF mod- μ 0 ified electrode. The hydrogen peroxide could be electrocatalytically detected at the modified electrode at levels with a sensitivity of -5 192 nA/lM very easily. Current,
-10 The electro-oxidation of H2O2 take place at potential above -15 0.64 V as follows: -20 1 H O ! H O þ O þ 2e ð5Þ -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 2 2 2 2 2 Potential, V Vs NCE The electro-reduction of H2O2 takes place below 0.38 V Fig. 4. Cyclic voltammogram of Ni –Fe HCF in 0.1 M NaNO (thick lines). 0.5 0.5 3 H2O2 þ 2e ! 2OH ð6Þ Response after addition of 1 mM of hydrogen peroxide to the cell (dotted line). 132 A.V. Narendra Kumar et al. / Journal of Electroanalytical Chemistry 659 (2011) 128–133
Fig. 6. (thick lines) Voltammetric response of Fe–HCF, Ni–HCF and Ni0.5–Fe0.5 HCF in 0.1 M KNO3 medium before cycling in 0.1 M NaNO3 supporting electrolyte (dotted line). + Voltammetric response of Fe–HCF, Ni–HCF and Ni0.5–Fe0.5 HCF in KNO3 medium after cycling in Na ion containing electrolyte.
Detection of hydrogen peroxide by electro-reduction enables the Rct values from the electrochemical impedance spectra. The j0 3 2 one to avoid interference from routinely occurring oxidizable com- values for FeHCF and Fe0.5 Ni0.5 HCF are 2.657 10 A/cm and 3 2 pounds in bio-fluids. Fig. 3 shows the amperogram for the detec- 3.054 10 A/cm respectively. Higher value of j0 is a sign of bet- tion of hydrogen peroxide using the Fe0.5 Ni0.5 HCF modified ter reversibility of the redox processes. This observation is in line electrode at applied potential of 0.09 V vs. NCE. The linear calibra- with the lower peak to peak separation values for the Fe0.5 Ni0.5 tion plot for the detection of hydrogen peroxide (192 nA/lM) in HCF modified films as given in Table 1. the micromolar level is also given in Fig. 3 (inset). The response time for hydrogen peroxide reduction is very fast as seen from 3.3. Electrochemical stability of Fe –Ni HCF film the response time less than 30 s. Detection of hydrogen peroxide 0.5 0.5 by electro-reduction generates high hydroxyl ion concentration The electrochemical reversibility/stability of redox processes in near the interface (Eq. (6)) which causes to destabilization of MHCF MHCF film depend on the nature of supporting electrolyte cation. at higher analyte concentration [27]. The sensitivity of hydrogen The electrochemical response of Prussian blue electrodes is not sta- peroxide detection can be improved by optimizing the film charac- ble when cycled in Na+ containing electrolytes. The effect of vari- teristics like film thickness, medium pH, etc. [28]. Detailed investi- ous supporting electrolytes on the NiHCF and FeHCF and CuHCF gation aimed at improving the sensitivity of hydrogen peroxide etc. of modified electrodes were studied in detail [31,6,7]. The sta- detection is currently underway in our laboratory. Electroreduc- bility of these modified films in Na+ containing electrolyte is stud- tion of H2O2 in NaNO3 supporting electrolyte is shown in Fig. 4 ied by examining the charge retention in KNO electrolyte after which reveals the usefulness of these hybrids in sodium containing 3 cycling the modified electrode in 0.1 M NaNO for 100 cycles in buffers. The buffers which are used to maintain the physiological 3 the potential range of 0.0–1.0 V (Fig. 6). The FeHCF modified film pH conditions generally contain sodium ions. The increased stabil- show only 35% retention of charge whereas NiHCF modified elec- ity of Fe0.5–Ni0.5HCF hybrid film in sodium containing electrolytes trodes exhibit more than 90% retention. The Fe –Ni HCF modi- towards electrochemical cycling assume great significance in the 0.5 0.5 fied electrode retains almost 80% of charge after cycling in field of biosensing. sodium containing electrolytes. The improvement in the electro- chemical stability in Na+ without losing the electrocatalytic prop- 3.2. Impedance spectroscopy erties can enlarge the scope of these hybrid modified electrodes in sensing applications in sodium ion containing buffer media. Fig. 5 shows the impedance plane plot (imaginary component, 00 0 Z vs. real component Z ) for both FeHCF and Fe0.5Ni0.5HCF in KNO3 medium. The charge transfer resistance (Rct) values derived 4. Conclusions from the Nyquist plot (Fig. 5) of FeHCF modified GC and Fe0.5 Ni HCF at bias potential of Ep=(Epa + Epc)/2 are around 0.5 Simple electrochemical modification of GC with hybrids Nix– 9.66 X cm2 and 8.4 X cm2 respectively. The lower R values ob- ct Fe(1 x)Fe(CN)6 thin film is reported. The tuning of redox potential 2+ served for the modified electrode with Ni incorporation is consis- of FeHCF film by controlled incorporation Ni2+ in the lattice by tent with the increased reversibility observed in terms of lower electrochemical means was successfully demonstrated. These hy- peak separation observed from the voltammetric response for brid film exhibited single redox process for the redox reaction of Fe Ni HCF modified electrodes. From Fig. 5, Bode plot (|Z| vs. 0.5 0.5 low spin iron. Incorporation of Iron ion in the Nix–Fe(1 x)Fe(CN)6 ln f) clearly show the line bend at higher frequency region. Similar lattice has significantly improved the electrocatalytic properties line bend at higher frequency region is observed for CuHCF films of the interfaces towards hydrogen peroxide oxidation/reduction and is attributed to the porous nature of the film [29]. (192 nA/lM). The Ni2+ ions substitution has improved hybrids’ + RT electrochemical cycling stability in Na containing supporting elec- j0 ¼ ð7Þ trolytes. The modification of M-HCF hybrids opens up possibilities nFARct of tailoring the interfacial properties by simple electrochemical
Exchange current density (j0) a kinetic parameter which have more control of composition. This hybrid modified electrode possesses influence on the hydrogen peroxide oxidation and reduction pro- improved electrocatalytic properties towards hydrogen peroxide cess by electrochemically modified electrodes FeHCF and Fe0.5- reduction/oxidation than the individual MHCF modified electrode. Ni0.5HCF, which can be directly calculated from Eq. (7) [30] all the These modified electrodes are expected to play an important role constants in the equation (R, Gas constant; T, Temperature; F, Fara- in the development of oxidase-enzyme based amperometric bio- day; A, area of GC electrode) have its corresponding values, by using sensors which produce H2O2 as intermediate. A.V. Narendra Kumar et al. / Journal of Electroanalytical Chemistry 659 (2011) 128–133 133
Acknowledgements [12] James Joseph, S. Bharathi, H. Gomathi, G.P. Rao, Bull. Electrochem. 18 (2002) 267–271. [13] L. Guadagnini, M. Giorgetti, F. Tarterini, D. Tonelli, Electroanalysis 22 (2010) This work was supported by Department of Science & Technol- 1695–1701. ogy, India through a grant-in-aid project [SR/S1/PC-22/2007]. S. [14] X.Z. Bian, H.Q. Luo, N.B. Li, Electroanalysis 22 (2010) 1364–1368. Harish thanks Council of Scientific and Industrial Research, New [15] S. Bharathi, James Joseph, D. Jeyakumar, G.P. Rao, J. Electroanal. chem. 319 (1991) 341–345. Delhi for the award of senior research fellowship. [16] A.A. Karyakin, E.A. Puganov, I.A. Bolshakov, E.E. Karyakin, Angew. Chem. Int. 46 (2007) 7678–7680. [17] A.A. Karyakin, E.E. Karyakina, Sens. Actuat. B 57 (1999) 268–273. References [18] K. Itaya, I. Uchida, S. Toshima, J. Phys. Chem. 87 (1983) 105–112. [19] J. Joseph, H. Gomathi, G.P. Rao, Electrochim. Acta 36 (1991) 1537–1541. [20] R.W. Murray, Chemically Modified Electrodes, in: A.J. Bard (Ed.), [1] K. Itaya, I. Uchida, V.D. Neff, Acc. Chem. Res. 19 (1986) 162–168. Electroanalytical Chemistry, vol. 13, M. Dekker, New York, 1984. [2] N.R.D. Tacconi, K. Rajeshwar, Chem. Mater. 15 (2003) 3046–3062. [21] S.J. Reddy, A. Dostal, F. Scholz, J. Electroanal. Chem. 403 (1996) 209–212. [3] R.M. Izatt, G.D. Watt, C.H. Bartholomew, J.J. Christensen, Inorg. Chem. 9 (1970) [22] A.G. Sharpe, The chemistry of cyano complexes of the transition metals, 2019. Academic Press, New York, 1976. [4] R. Vittal, K.J. Kim, H. Gomathi, V. Yegnaraman, J. Phys. Chem. B 112 (2008) [23] H. Gomathi, G.P. Rao, J. Appl. Electrochem. 20 (1989) 454–456. 1149–1156. [24] K. Miecznikowski, M. Chojak, W. Steplowska, M.A. Malik, P.J. Kulesza, J. Solid [5] R. Vittal, H. Gomathi, J. Phys. Chem. B 106 (2002) 10135–10143. State Electrochem. 8 (2004) 868–875. [6] P.J. Kulesza, T. Jedrald, Z. Galus, Electrochim. Acta 34 (1989) 851–853. [25] A.A. Karyakin, E.E. Karyakin, L. Gorton, Anal. Chem. 72 (2000) 1720–1723. [7] T.R.I. Cataldi, G.D. Benedetto, A. Bianchini, J. Electroanal. Chem. 471 (1999) 42– [26] S. Senthil Kumar, J. Joseph, K.L.N. Phani, Chem. Mater. 19 (2007) 4722–4730. 47. [27] A.A. Karyakin, E.E. Karyakina, L. Gorton, Electrochem. Commun. 1 (1999) 78– [8] J.K. Walkiewicz, J. Stroka, M.A. Malik, P.J. Kulesza, Z. Galus, Electrochim. Acta 82. 46 (2001) 4057–4063. [28] R. Araminaite, R. Garjonyte, A. Malinauskas, J. Solid State Electrochem. 14 [9] P.J. Kulesza, M.A. Malik, R. Schmidt, A. Smolinska, K. Miecznikowski, S. (2008) 149–155. Zamponi, A. Czerwinski, M. Berrettoni, R. Marassi, J. Electroanal. Chem. 487 [29] L. M Siperko, T. Kuwana, J. Electochem. Soc. 130 (1983) 396–402. (2000) 57–65. [30] A.J. Bard, L.R. Faulkner, Electrochemical Methods. Fundamentals and [10] A. Dostal, G. Kauschka, S.J. Reddy, F. Scholz, J. Electroanal. Chem. 406 (1996) Applications, John Wiley & Sons, Inc., New York, 2001. 155–163. [31] A.P. Baioni, M. Vidotti, P.A. Fiorito, S.I. Córdoba de Torresi, J. Electroanal. Chem. [11] A. Dostal, B. Meyer, F. Scholz, U. Schroeder, A.M. Bond, F. Marken, S.J. Shaw, J. 622 (2008) 219–224. Phys. Chem. 99 (1995) 2096–2103.