Aqueous Rechargeable Batteries with High Electrochemical Performance

Home , Daniell cell, Weston cell

Aqueous Rechargeable Batteries with High

Electrochemical Performance

von der Fakultät für Naturwissenschaften der Technischen Universität Chemnitz

genehmigte Dissertation zur Erlangung des akademischen Grades

doctor rerum naturalium

(Dr. rer. nat.)

vorgelegt von

M.Sc. Yu Liu

geboren am 06.05.1988 in Tongyu Town, Jiangsu Province, China

eingereicht am 10 Mai 2017

Gutachter: Prof. Dr. Rudolf Holze

Prof. Dr. Qunting Qu

Tag der Verteidigung: 28 Juli 2017

Bibliographische Beschreibung und Referat

Bibliographische Beschreibung und Referat Y. Liu

Wässrige Akkumulatoren mit hoher elektrochemischerischer Leistung

Mit der Entwicklung der Weltwirtschaft steigt der Energieverbrauch weiterhin stark an.

Darüber hinaus reduzieren sich die nicht erneuerbaren Energiequellen, wie Öl, Erdgas und Kohle und die Umweltverschmutzung wird größer. Daher soll die Energienutzung in eine neue, erneuerbare und umweltfreundliche Richtung gehen. Die Arbeit hat zum Ziel innovative, wässrige Akkumulatoren zu entwickeln.

Im Allgemeinen können wässrige Akkumulatoren gemäß der Elektrolyte in drei verschiedenen Kategorien eingeteilt werden. Es gibt feste, organische und wässrige

Elektrolyte einschließlich saurer, alkalischer und neutraler. In Bezug auf metallbasierte negative Elektroden können sie auch als Lithiumbatterie, Natriumbatterie sowie

Magnesiumbatterie etc. bezeichnet werden. Daher werden im ersten Kapitel einige typische Akkumulatoren, wie die Lithiumionenbatterien, Daniell-Element, Weston-Zelle,

Nickel-Cadmium-Batterie und Bleibatterie vorgestellt.

Im Vergleich zu organischen Elektrolyten wurden wässrige Akkumulatoren aufgrund ihrer billigen, leichten und sicheren Bauweise in den letzten Jahren umfassend untersucht. Zusätzlich dazu ist die ionische Leitfähigkeit von wässrigen Elektrolyten um zwei Größenordnungen höher als die von organischen Elektrolyten. Dies garantiert eine hohe Entladungsrate für wässrige wiederaufladbare Batterien. Somit bieten wiederaufladbare Batterien potentielle Anwendungen in der Energiespeicherung und - umwandlung.

Allerdings verursachen starke Säuren oder Basen, die als Elektrolyte für sekundäre

Batterien eingesetzt werden, eine starke Korrosion. Somit wären neutrale wässrige

Elektrolyten (oder Elektrolytlösungen) mit einem pH-Wert in der Nähe von sieben, wie

2 Bibliographische Beschreibung und Referat

zum Beispiel schwach basisch oder sauer, die beste Wahl für wässrige Akkumulatoren.

Aktive Elektrodenmaterialien der Batterien, die hochgiftige Schwermetalle wie Blei,

Quecksilber und Cadmium enthalten, belasten die Umwelt.

Um die Menge an Schwermetallen und Säure (oder Basen) zu verringern, sowie die spezifische Kapazität von Batterien zu erhöhen, untersucht diese Dissertation vor allem die elektrochemische Leistung der PbSO4/0,5M Li2SO4/LiMn2O4-Zelle, der Cd/0,5M

Li2SO4+10mM Cd(Ac)2/LiCoO2-Zelle und von C/Cu/CNT-Gemischen als negative

Materialien in 0,5 M K2CO3–Elektrolyt-Halbzellen. Die zugehörigen experimentellen

Ergebnisse werden wie folgt zusammengefaßt:

Im Kapitel 3 wurde eine säurefreie Bleibatterie auf Basis des LiMn2O4-Spinells als positive Elektrode, PbSO4 als negativer Elektrode und der wässrigen Lösung von 0,5 M

Li2SO4 als Elektrolyt zusammengesetzt. Die spezifische Kapazität auf Basis von LiMn2O4 beträgt 128 mA·h·g-1 und die durchschnittliche Entladungsspannung beträgt 1,3 V. Die berechnete Energiedichte ist 68 W·h·kg-1, bezogen auf die praktischen Kapazitäten der beiden Elektroden. Diese Ergebnisse zeigen, dass die positive Elektrode der Bleibatterie

(PbO2) vollständig durch umweltfreundliches und billiges LiMn2O4 ersetzt werden kann, wodurch 50 % des Bleis eingespart werden können. Außerdem wird Schwefelsäure nicht benötigt.

Kapitel 4 zeigt eine wässrige wiederaufladbare Lithiumionenbatterie, die metallisches Cadmium als negative Elektrode, LiCoO2-Nanopartikel als positive

Elektrode und eine wässrige, neutrale Lösung von 0,5 M Li2SO4 und 10 mM Cd(Ac)2 als

Elektrolyt enthält. Die durchschnittliche Entladungsspannung beträgt 1,2 V und die

-1 spezifische Entladungskapazität beträgt 107 mA·h·g auf Basis von LiCoO2. Die berechnete Energiedichte beträgt 72 W·h·kg-1, bezogen auf die praktischen Kapazitäten der beiden Elektroden. Wie bereits oben beschrieben demonstrieren die Ergebnisse, dass

100 % von Quecksilber und der alkalischen Elektrolyt im Vergleich zur Weston-Zelle

3 Bibliographische Beschreibung und Referat

bzw. der Ni-Cd-Batterie, eingespart werden können.

Kapitel 5 zeigt einen Verbundwerkstoff von Kupfer, das auf der Oberfläche von

CNTs durch eine Redoxreaktion zwischen Kupferacetat und Ethylenglykol, zur

Verwendung als negative Elektrode bei hohen Strömen in der Energiespeicherung, hergestellt wurde. Der so hergestellte C/Cu/CNT-Verbundwerkstoff zeigt ein besseres

Geschwindigkeitsverhalten und eine höhere Kapazität ebenso wie eine exzellente

Zyklusstabilität in wässrigen 0,5 M K2CO3-Lösungen im Vergleich zu einfachem Kupfer.

Die Kohlenstoffbeschichtung kann die Auflösung von Kupfercarbonatkomplexen verhindern, die Elektrodenleitfähigkeit erhöhen und die Oberflächenchemie des aktiven

Materials verbessern.

Schlüsselwörter: Akkumulator, wässrige Elektrolyte, Interkalationsverbindungen, Blei,

Cadmium, Kupfer, Kohlenstoff-Beschichtung

4 Abstract

Abstract

Y. Liu

Aqueous Rechargeable Batteries with High Electrochemical Performance

With the economic development of the world, energy consumption continues to rise sharply. Moreover, non-renewable energy sources including fossil oil, natural gas and coal are declining gradually and environmental pollution is becoming more severe. Hence, energy usage should go into a new direction of development that is renewable and environmental-friendly. This thesis aims to explore innovative aqueous rechargeable batteries.

Generally, rechargeable batteries could be classified into three categories according to the different electrolytes. There are solid electrolytes, organic electrolytes and aqueous electrolytes including acidic, alkaline and neutral. In terms of metal-based negative electrodes, they also could be named lithium battery, sodium battery as well as magnesium battery etc. Therefore, some typical rechargeable batteries are introduced in

Chapter 1, such as lithium ion batteries, Daniell-type cell, Weston cell, Ni-Cd battery and lead-acid battery.

Compared to organic electrolytes, aqueous rechargeable batteries have been investigated broadly in recent years because they are inexpensive, easy to construct and safe. Additionally, the ionic conductivity of aqueous electrolytes is higher than that of organic electrolytes by about two orders of magnitude. Furthermore, it ensures high rate capability for aqueous rechargeable battery. Consequently, aqueous rechargeable batteries present potential applications in energy storage and conversion.

However, strong acid or alkaline, which is used as the electrolyte for secondary batteries, will cause serious corrosion. Thus, neutral aqueous electrolyte (or pH value of electrolyte solution close to 7 such as weak alkaline and acid) would be the best choice

Abstract

for aqueous rechargeable battery. In addition, the electrode active materials of batteries containing highly toxic heavy metals such as Pb, Hg and Cd, pollute the environment.

As a result, in order to reduce the amount of heavy metals and acid (or alkaline) as well as increase the specific capacity of batteries, this dissertation mainly studies the electrochemical performance of PbSO4/0.5M Li2SO4/LiMn2O4 full battery, Cd/0.5M

Li2SO4+10 mM Cd(Ac)2/LiCoO2 full battery and C/Cu/CNT composites as negative material in 0.5 M K2CO3 electrolyte as half cell. The related experimental results are as follows:

In Chapter 3, an acid-free lead battery was assembled based on spinel LiMn2O4 as the positive electrode, PbSO4 as the negative electrode, and 0.5 M Li2SO4 aqueous

-1 solution as the electrolyte. Its specific capacity based on the LiMn2O4 is 128 mA·h·g and the average discharge voltage is 1.3 V. The calculated energy density is 68 W·h·kg-1 based on the practical capacities of the two electrodes. These results show that the positive electrode of the lead acid battery (PbO2) can be totally replaced by the environmentally friendly and cheap LiMn2O4, which implies that 50 % of Pb can be saved.

In addition, H2SO4 is not needed.

Chapter 4 shows an aqueous rechargeable lithium ion battery using metallic Cd as the negative electrode, LiCoO2 nanoparticles as the positive electrode, and an aqueous neutral solution of 0.5 M Li2SO4 and 10 mM Cd(Ac)2 as the electrolyte. Its average discharge voltage is 1.2 V and the specific discharge capacity is 107 mA·h·g-1 based on the LiCoO2 . In addition, the calculated energy density based on the capacities of the electrodes is 72 W·h·kg-1. As described above, the results demonstrate that 100 % of Hg and alkaline electrolyte can be saved compared with the Weston cell and the Ni-Cd battery, respectively.

The work reported in Chapter 5 deals with a composite of copper grown on the surface of CNTs as prepared by a redox reaction between copper acetate and ethylene

Abstract

glycol for use as negative electrode at high currents in energy storage. The as-prepared

C/Cu/CNTs composite exhibits better rate behavior and higher capacity as well as excellent cycling stability in aqueous 0.5 M K2CO3 solution compared to the unsupported copper. The carbon coating can effectively prevent the dissolution of copper carbonate complexes, increase the electrode conductivity, improve the surface chemistry of the active material and protect the electrode from direct contact with electrolyte solution.

Keywords: rechargeable battery; aqueous electrolyte; intercalation compounds; lead; cadmium; copper; carbon coating

Zeitraum, Ort der Durchführung

Die vorliegende Arbeit wurde in der Zeit von Aug. 2014 bis Jan. 2017 unter Leitung von

Prof. Dr. Rudolf Holze am Lehrstuhl für Physikalische Chemie/Elektrochemie der

Technischen Universität Chemnitz durchgeführt.

Acknowledgements

On the completion of my thesis, I would like to express my deepest acknowledgements to all those whose advice and help have made this research work possible. At first, I wish to give my sincere gratitude to my supervisor Prof. Dr. Rudolf Holze, whose expertise, valuable suggestions, unique insights, correct guidance and extraordinary patience greatly increased my knowledge. I deeply appreciate his erudition in the field of Physical

Chemistry/Electrochemistry.

Secondly, I should thank all the members of the Institute of Chemistry and Institute of Physic, Chemnitz University of Technology, who provided help in experimental measurements and analysis, especially to Prof. M. Mehring, L. Mertens, T. Jagemann, S.

Schulze, N. Rüffer and E. Dietzsch.

Moreover, I also wish to acknowledge the present and previous members of

Electrochemical group for their help, support, encouragement and respect in my study and life, particularly to Mrs. Nora Younadam, Ms. Elisabeth Klinge and Mr. Alexander Wiek.

Additionally, I would like to take this opportunity to thank my parents for their tremendous sacrifices and endless love, which is of great help for me to finish this dissertation successfully.

Lastly, special acknowledgements should go to professors and savants who will act as the reviewers of this thesis.

Dedication

Dedicated to

My Loving Parents and Sisters

Table of Contents

Bibliographische Beschreibung und Referat 2

Abstract 5

Zeitraum, Ort der Durchführung 8

Acknowledgments 9

Dedication 10

Table of Contents 11

List of Abbreviations and Symbols 14

Chapter 1 16

1 Introduction 16

1.1 Lithium ion battery 18

1.1.1 Conventional lithium ion battery 18

1.1.2 First generation of ARLIBs 19

1.1.3 Second generation of ARLIBs 20

1.1.4 Lithium-sulfur (Li-S) battery 22

1.1.5 Lithium-air (Li-O2) battery 24

1.2 Copper metal-based battery 26

1.2.1 Daniell-type cell 26

1.2.2 All-copper redox flow battery 28

1.3 Cadmium metal-based battery 29

1.3.1 Weston cell 29

1.3.2 Nickel-Cadmium (Ni-Cd) battery 30

1.3.3 Cadmium-related redox flow battery 32

1.4 Lead-acid battery 34

1.5 Aims and Tasks of this Study 37

Table of Contents

Chapter 2 38

2 Experimental 38

2.1 Chemicals 38

2.2 Synthesis of materials 38

2.2.1 Synthesis of LiMn2O4 38

2.2.2 Synthesis of LiCoO2 38

2.2.3 Synthesis of C/Cu/CNT composites 39

2.3 Characterization of the as-prepared materials 40

2.4 Electrochemical measurements 41

2.4.1 Preparation of electrodes 41

2.4.2 Characterization of electrochemical performance 42

Chapter 3 44

3 An acid-free rechargeable battery based on PbSO4 and spinel LiMn2O4 44

3.1 Characterization of morphology and structure of LiMn2O4 44

3.2 Cyclic voltammetry (CV) of PbSO4 and LiMn2O4 electrodes 45

3.3 Galvanostatic charge/discharge measurement of PbSO4//LiMn2O4 battery 47

3.4 Cycling performance of PbSO4//LiMn2O4 battery 48

3.5 Conclusions 49

Chapter 4 51

4 An aqueous lithium ion rechargeable battery with high rate capability based on 51

metallic Cd and LiCoO2

4.1 Characterization of morphology and structure of LiCoO2 51

4.2 Cyclic voltammetry (CV) of Cd//Cd and Cd//LiCoO2 battery 52

4.3 Impedance measurement of Cd//LiCoO2 battery 55

4.4 Galvanostatic charge/discharge measurement of Cd//LiCoO2 battery 57

4.5 Cycling performance of Cd//LiCoO2 battery 58

Table of Contents

4.6 Conclusions 60

Chapter 5 61

5 Improved electrochemical behavior of amorphous carbon-coated copper/CNT 61

composites as negative electrode and their energy storage mechanism

5.1 Characterization of morphology and structure of C/Cu/CNT composites 62

5.2 Infrared (IR) spectra of CNTs and C/CNTs composites 65

5.3 Thermogravimetic (TG) analysis of C/Cu/CNT composites 66

5.4 BET surface area measurement of C/Cu/CNT composites 67

5.5 XPS measurements of C/Cu/CNT composites 68

5.6 Energy storage mechanism and CV of C/Cu/CNT composites 69

5.7 Impedance measurement of C/Cu/CNT composites 71

5.8 Galvanostatic charge/discharge measurement of C/Cu/CNT composites 74

5.9 Cycling performance of C/Cu/CNT composites 76

5.10 Conclusions 78

Chapter 6 80

6 Summary and outlook 80

6.1 Summary 80

6.2 Outlook 81

References 83

Selbständigkeitserklärung 92

Curriculum Vitae 93

List of Abbreviations and Symbols

ARLB Aqueous rechargeable lithium battery

ARLIB Aqueous rechargeable lithium ion battery

BET Brunauer-Emmett-Teller

C Carbon

CNT Carbon nanotube

CDL Double layer capacitance

CV Cyclic voltammetry dE/dt Scan rate

DEC Diethyl carbonate

DMC Dimethyl carbonate eV Electron volt

EC Ethylene carbonate

EG Ethylene glycol

EIM Electrochemical impedance measurement

ESCE Potential versus saturated calomel electrode

Fig. Figure

GPE Gel polymer electrolyte

HRTEM High resolution transmission electron micrograph

IR Infrared

LATSP Li1+x+yAlxTi2-xSiyP3-yO12

LISICON Lithium super ionic conductor

M Mol per liter

NLSF Nonlinear least squares fitting

NaSICON Sodium super ionic conductor

OCP Open circuit potential

List of Abbreviations and Symbols

PC Personal computer

PEO Poly(ethylene oxide)

PMMA Polymethyl methacrylate

PPy Polypyrrole

PTFE Poly(tetrafluoroethylene)

PVDF Polyvinylidene fluoride

QCPE Constant phase element

Rsol Solution resistance

Rct Charge transfer resistance

RFB Redox flow battery

RVC Reticulated vitreous carbon

SCE Saturated calomel electrode

SEM Scanning electron micrograph

TEM Transmission electron micrograph

TG Thermogravimetic

V Volt vs. Versus

XPS X-ray photoelectron spectroscopy

XRD X-ray diffraction

Yo Admittance

Zdiff Warburg diffusion

1G First generation

2G Second generation

Chapter 1: Introduction

Chapter 1 Introduction

Energy has always been an important issue in the economic development of the world and lives of human beings. Moreover, supplies of non-renewable energy sources such as crude oil, coal and natural gas are gradually decreasing and the environmental problems are becoming more serious. Therefore, energy should move in a new direction that is renewable and environmentally friendly. Currently, renewable energies including wind, solar and tide energies have received much attention around the world. However, these intermittent renewable energy sources depend on time and season. As a result, energy storage and conversion systems are urgently needed to solve this problem. That is to say, when the energy is surplus, it will be stored. On the contrary, it also can be released [1 -

3].

Among various energy storage systems, electrochemical energy storage and conversion systems such as batteries, supercapacitors and fuel cells have attracted considerable attention in recent years because they can improve efficient use of electric energy, and thus reduce emission of greenhouse gases and promote the use of renewable energy [4,5]. Furthermore, rechargeable batteries surpass in energy density and power density supercapacitors and fuel cells, respectively, and serve as the key components for portable devices, electric vehicles and even large-scale energy storage [6, 7]. Among rechargeable batteries, lithium batteries possess high energy density, but the disadvantages for portable applications should be tackled, including high cost and inherent safety risks [8, 9]. Moreover, lithium ions have low conductivity in organic electrolytes, leading to inadequate power density for lithium ion batteries [ 10 ].

Consequently, aqueous rechargeable batteries have been researched widely in recent years, which are inherently safe by avoiding flammable organic electrolyte solutions. Moreover, aqueous electrolytes are inexpensive and easy to assemble. In addition, the ionic conductivity of aqueous electrolytes is about two orders of magnitude higher than that of

Chapter 1: Introduction

organic electrolytes, which ensures high rate capability for aqueous rechargeable battery

[ 11 , 12 ]. As mentioned above, aqueous rechargeable batteries exhibit potential applications in energy storage and conversion.

In general, rechargeable batteries could be classified into three categories according to the different electrolytes. There are solid electrolytes (e.g. LISICON and NASICON)

[13], organic electrolytes (e.g. LiPF6 dissolved in EC:DEC (1:1 wt) organic solvents, polymer electrolytes and ionic liquids) [14, 15, 16] and aqueous electrolytes (e.g. acidic, alkaline and neutral) [17, 18]. In terms of metal-based negative electrodes, they also could be called lithium battery [19], sodium battery [20], magnesium battery [21], aluminum battery [22], iron battery [23], Ni-Cd battery [24], zinc battery [11] and lead-acid battery

[25] etc. Some characteristics of these batteries are listed in Table 1, and much more details will be introduced below.

Table 1 Some characteristics of various rechargeable batteries

Average Capacity Energy Density Type Negative Positive Electrolytes Ref. Voltage (V) (mA·h/g) (W·h/kg)

LiPF6/

Li-ion Graphite LiCoO2 EC/DEC 3.6-3.7 140 130 26, 27

Saturated ARLIB PPy LiCoO2 0.85 47.7 ---- 28 Li2SO4

PPy/ 0.5M ARLIB LiMn2O4 1.22 110 45 29 MoO3 Li2SO4

0.5M ARLB Li LiCoO2 3.7 130 465 19 Li2SO4

0.5M ARLB Li LiMn2O4 4.0 115 446 30 Li2SO4

Li-I2 Li Carbon I2/LiI/KI 3.5 207 330 31

Glass Li-Br Li LiBr/Br 3.96 335 1220 32 Carbon 2

Sulfur/ LiCF3SO3/ Li-S Li 2.1 450 300 33 Carbon PEO

MnO2/ Li-O2 Li LiBF4 2.96 2707 8400 34 La2O3/Pt/C

Chapter 1: Introduction

2M LiNO / Daniell cell Zn Cu 3 0.8 843 68.3 35 Zn(NO3)2

Ni-Cd Cd(OH)2 Ni(OH)2 KOH 1.2 ---- 45 36

Lead-acid Pb PbO2 H2SO4 2 ---- 35 37, 38

1.1 Lithium ion battery

1.1.1 Conventional lithium ion battery

Lithium batteries and subsequently Li-ion batteries have been the subject of intensive research since 1976 when Exxon reported the first phenomenon of reversible lithium intercalation in LixTiS2 [39]. However, the commercialization of the organic Li-based technology was stopped due to the safety reasons related to the dendritic growth of lithium. Attempts to solve this problem led to the emergence of the Li-ion concept in the

1980s and the commercialization of the C//LiCoO2 battery by Sony in 1991 [40]. In general, lithium ion battery exhibits a reversible intercalation/de-intercalation of lithium ions into/from a host matrix during the charge/discharge process as shown in Fig. 1.1. To be specific, lithium ion battery use mostly graphite as the anode host and the layered

LiCoO2 as the cathode host. An inorganic salt such as LiPF6 dissolved in a mixture of organic solvents including ethylene carbonate (EC) and diethyl carbonate (DEC) is used as the electrolyte [27]. In this case, lithium metal was replaced by carbon, in which lithium forms graphite intercalation compounds, and the highest level can be LiC6. The related electrode reactions during charge/discharge cycling are shown as follows [26]:

Cathode reaction: (1)

Anode reaction: (2)

Total reaction: (3)

Chapter 1: Introduction

Fig. 1.1 Illustration of lithium ion battery consisting of graphite as an anode and layered

LiCoO2 as a cathode during charge/discharge process [27].

1.1.2 First generation of ARLIBs

Aqueous rechargeable lithium ion batteries (ARLIBs) were invented in 1994 [2].

Compared to the combustible organic electrolytes of lithium ion batteries, operating in environmentally friendly aqueous electrolyte solutions with high ionic conductivity are attractive for energy storage in terms of cost and safety [41]. Since 2007, the development of them has been very fast and a few breakthroughs have been achieved. For instance,

LiCoO2, LiMn2O4, Li[Ni1/3Co1/3Mn1/3]O2 and LiFePO4 as positive materials show much better electrochemical performance for ARLBs than in organic electrolytes [42]. To be specific, an aqueous rechargeable lithium ion battery (ARLIB) based on doping (PPy as negative electrode) and intercalation (LiCoO2 as positive electrode) mechanisms was shown in Fig. 1.2. However, the potential of n-type doping in organic electrolyte is much lower than that of Li+/Li. Thus, only p-type doping and un-doping can take place in aqueous solution, which is presented as follow [28]:

(4) where Py is the monomer repeat unit of the electro-active polymer and is the 19

Chapter 1: Introduction

electrolyte anion.

Fig. 1.2 Principle of the ARLIB system based on doping/un-doping at negative and

deintercalation/intercalation at positive electrodes [28].

1.1.3 Second generation of ARLIBs

The main disadvantage of the above mentioned ARLIBs, which could be called the first generation of ARLIBs (1G ARLIBs), is that their average discharge voltage and energy density are still much less than those of traditional lithium ion batteries owing to the limit of the electrochemically stability window of water (1.23 V) [41]. Therefore, one kind of new ARLBs, also called the second generation (2G) ARLBs, has been further developed in recent years. The main characteristic of these 2G ARLBs is that they use coated lithium metal as the negative electrode [42]. Specifically, the coated lithium metal is schematically shown in Fig. 1.3. The coating consists of a home-developed gel polymer electrolyte (GPE) and a LISICON film (refers to a solid film consisting of Li2O-Al2O3-

SiO2-P2O5-TiO2-GeO2). This film is a solid electrolyte which can also serve as a separator.

Moreover, protons, hydrated, solvated ions and water cannot pass through [30, 43]. The

GPE consists of a sandwiched structured PVDF/PMMA/PVDF soaked with an organic electrolyte (1 M LiClO4 solution in ethylene carbonate, diethyl carbonate and dimethyl

Chapter 1: Introduction

carbonate, volume ratio is 1:1:1). When LISICON film contacts lithium metal directly, some metal oxides such as GeO2 in the LISICON film will be reduced by lithium metal resulting in poor ionic conductivity. Consequently, the GPE ensures the good electrochemical stability of the LISICON film, and only lithium ions can transfer between the GPE and the LISICON film. Finally, the coated lithium metal is really stable in aqueous electrolytes [30].

Fig. 1.3 A schematic illustration of the coated lithium metal [30].

As for the application of the coated lithium metal, the schematic structure of the assembled ARLB using the coated lithium metal as the negative electrode and LiMn2O4 as the positive electrode is exhibited in Fig. 1.4. The related electrode reactions are expressed as follows [30]:

Negative reaction: (5)

Positive reaction: (6)

Total reaction: (7)

During the charging process, lithium ions from the aqueous electrolyte pass through the coating layer, are reduced on the surface of Li metal and deposited as lithium metal. In terms of the LiMn2O4 positive electrode, lithium ions de-intercalate from the tetrahedral

8a- and octahedral 16c-sites, causing two pairs of redox peaks, which are similar to the behavior in organic electrolytes [44]. During discharging, the opposite processes take

Chapter 1: Introduction

place.

Fig. 1.4 Schematic illustration of ARLB using the coated lithium metal as negative

electrode, LiMn2O4 as positive electrode and 0.5M Li2SO4 aqueous solution as electrolyte

[30].

1.1.4 Lithium-sulfur (Li-S) battery

Lithium-sulfur (Li-S) battery was invented in 1960s [45]. It represents a promising energy storage system due to its higher theoretical specific capacity (1675 mA·h·g-1) and energy density (2600 W·h·kg-1) compared with conventional lithium ion batteries [46].

The main difference between them depends on their mechanism of energy storage.

Lithium ion batteries work based on intercalation of lithium ions into layered electrode materials, which limit the energy density (~ 420 W·h·kg-1). However, lithium-sulfur batteries are based on metal plating and stripping on the lithium negative electrode and a conversion reaction on the sulfur positive electrode. Therefore, the non-topotactic nature of these reactions contributes to sulfur cathodes and lithium anodes with high specific capacities [47]. Moreover, sulfur is inexpensive, abundant on earth and environmentally friendly, which makes Li-S battery a strong candidate for future low-cost energy storage technology [48].

Generally, a Li-sulfur battery consists of a stable sulfur composite as positive

Chapter 1: Introduction

electrode, a lithium metal as negative electrode, an organic electrolyte and a separator between the two electrodes, which is shown in Fig. 1.5 (a) [49]. During the ideal discharge process, S8 first experiences the ring-opening reaction with the generation of high-order lithium polysulfides Li2Sx (4

1.5 (b)) [51]. Fig. 1.5 (c) shows that two discharge platforms at 2.3 and 2.1 V can be obtained in ether-based liquid electrolytes, which correspond to the reactions of S8 to

Li2S4 and Li2S4 to Li2S, respectively [52]. In principle, the reaction kinetics of discharge plateau at 2.1 V is slower than that of another discharge plateau at 2.3 V owing to the extra energy needed for nucleation of the solid phase and sluggish solid-state diffusion to change Li2S2 to Li2S [53]. However, during charging, Li2S transforms back to S8 by a reversible phase transition, which often shows only one overlapping charge voltage plateau [51, 54 ]. Apparently, these unique features inevitably cause several serious problems of Li-S battery as shown in Fig. 1.5 (d): (1) the poor ionic and electronic conductivities of sulfur and its discharge product Li2S; (2) the volume change of active materials during charge/discharge; (3) the diffusion and dissolution of lithium polysulfides in electrolytes; (4) safety problem caused by lithium dendrites on Li metal anode [48, 49]. As a result, much more investigations should be done to solve these issues in future.

Chapter 1: Introduction

Fig. 1.5 Schematic illustration of the Li-sulfur battery: (a) cell structure; (b)

charge/discharge reaction processes; (c) electrochemical features; (d) major issues [49].

1.1.5 Lithium-air (Li-O2) battery

Lithium-air batteries, also called Li-O2 batteries (since most laboratory studies were conducted under pure oxygen environment in order to avoid the influence of water and

CO2 in air [55 ]), are considered as one of the most attractive energy storage and conversion systems in recent years owing to their high theoretical energy density (11140

W·h·kg-1, excluding oxygen), which is rivaling that of gasoline (13000 W·h·kg-1) [56]. As early as 1976, the concept of Li-air chemistry in an aqueous system was proposed [57].

Afterwards, the first study of Li-air battery system based on organic polymer electrolyte was reported in 1996 [58]. Recently, four types of Li-air batteries are distinguished by the type of electrolyte used, as shown in Fig. 1.6: aqueous, aprotic, solid-state and hybrid aqueous/aprotic [59]. For all types of Li-air batteries, an open system is required to breathe oxygen from the air since O2 is the active material of the positive electrodes.

Lithium metal also must be used as negative electrodes to offer the lithium source at the present stage. In the case of aqueous and hybrid systems, a coating layer on lithium metal is necessary to prevent the severe reaction between water and Li metal [60]. 24

Chapter 1: Introduction

Fig. 1.6 Schematics of the four different configurations of Li-air batteries [59].

In terms of energy storage mechanisms, aqueous and hybrid aqueous/aprotic systems possess the same reaction mechanisms because the air electrodes for both of them are exposed to aqueous electrolyte. Generally, several electrochemical reactions in aqueous electrolytes could occur at the air electrode according to the different pH-values of the electrolyte. In a basic aqueous solution, the overall cell reaction of Li-air battery is shown in the following equation [60]:

(8)

Another example is in an acidic aqueous solution, the total reaction of lithium with O2 is described as below [61]:

(9)

As for the solid-state Li-air battery, which may be similar to the aprotic system in reaction mechanisms (Eqs. 10-13), even though the solid-state Li-air battery was not investigated broadly in academic research owing to the lack of a solid-state electrolyte 25

Chapter 1: Introduction

with high lithium ion conductivity. As a result, aprotic systems have dominated the research effort on Li-air batteries in the past decade [60]. In the case of aprotic system, the discharge product at the air electrode is Li2O2 with a small portion of Li2O. The associated electrode reactions of Li-air battery with Li2O2 as the product can be expressed as follows [59]:

Negative reaction: (ESHE = -3.05V) (10)

Positive reaction: (ESHE = -0.09V) (11)

Total reaction: (E = 2.96V) (12)

For discharge product of Li2O, the overall reaction can be simply described by the following equation:

(E = 2.91V) (13)

Comparing to lithium peroxide, equation (13) is not desirable since it is not reversible completely. That is to say, it cannot be totally discharged back to O2 and Li [59, 62]. Thus,

Li2O2 is regarded as the ideal discharge product at the air electrode.

1.2 Copper metal-based battery

1.2.1 Daniell-type cell Zinc-Copper (Zn-Cu) Daniell cell was invented by John Frederic Daniell in 1836 [63, 64].

In a typical configuration of Daniell cell as shown in Fig. 1.7, zinc and copper electrodes are placed in aqueous solution of ZnSO4 and CuSO4, respectively, connected by a salt bridge with saturated KCl aqueous electrolyte to conduct ions. The associated redox reactions are described as the following equations [65]:

Negative reaction: (ESHE = -0.7618 V) (14)

Positive reaction: (ESHE = +0.340 V) (15)

Total reaction: (E = +1.1018 V) (16)

Chapter 1: Introduction

Fig. 1.7 Schematic illustration of conventional Daniell cell in chemistry curricula [35].

However, Daniell cell has been replaced by more modern battery systems including lead-acid battery and Ni-Cd battery since 1860s due to copper ion crossover-induced self- discharge [35, 65]. Therefore, the new structure of the Zn-Cu rechargeable battery is demonstrated in Fig. 1.8. It consists of a Zn anode in 1M Zn(NO3)2 electrolyte solution, a

Cu cathode in 2M LiNO3 electrolyte solution and a ceramic LATSP (Li1+x+yAlxTi2-xSiyP3-

2+ 2+ yO12) thin film. During the charge process, metallic copper is oxidized into Cu and Zn is reduced into metallic zinc on the surface of Zn plate, whereas lithium ions are transported from cathodic region to anodic region across the LATSP ceramic film to balance the charges. During discharging, the reverse reaction occurs. Thus, the total reaction is similar to conventional Daniell cell for the new Zn-Cu rechargeable battery, which can be expressed as below [35]:

(17)

Chapter 1: Introduction

Fig. 1.8 Schematic illustration of Zn-Cu rechargeable battery with a lithium ion exchange

membrane [35].

1.2.2 All-copper redox flow battery Since the first emergence of the Fe-Cr flow cell in 1973, many types of redox flow batteries (RFB) have been investigated widely [ 66 ]. However, the most developed technology in the field of RFB is the all-vanadium redox flow battery in recent years.

Other systems including the Fe/Cr, the Zn/Br and the polysulfide bromide have also come close to full scale commercialization [67]. Regarding the cost, a facile strategy is the partial or total substitution of the electrolytes of RFB systems by alternative redox couples. But, these alternative redox couples should be non-toxic, abundant and highly soluble in water. In addition, they should provide a redox potential close to the limits of the potential window of the supporting electrolyte, they should be simple to recycle and highly conductive [68]. As a result, all-copper redox flow battery appeared in recent years, which could reduce the problem of cross-contamination through the membrane, allowing the use of inexpensive and simple microporous separators.

In the case of all-copper RFB, the energy storage mechanism is shown in Equations

(18), (19) and (20). The fresh electrolyte consisted of cuprous chloride complexes at the

Chapter 1: Introduction

beginning. During charging, Cu+ is converted to Cu2+ in the positive region and deposited as metallic copper on the surface of Cu negative electrode. The opposite reactions happen during the discharge process [68].

Negative reaction: (18)

Positive reaction: (19)

Total reaction: (20)

1.3 Cadmium metal-based battery 1.3.1 Weston cell Weston cell is a classical electrochemical system that was invented by Edward Weston in

1893 [69]. In a typical configuration as illustrated in Fig. 1.9, Weston cell is set up in a H- shaped glass vessel with the cadmium amalgam as negative electrode in one side and pure mercury in the other side. Electrical connections to the mercury and the cadmium amalgam are made by platinum wires fused through the lower ends of the glass vessel.

The filled electrolyte is a saturated solution of cadmium sulfate. Therefore, the associated redox reactions can be expressed as follows:

Negative reaction: (ESHE = -0.403V) (21)

Positive reaction: (ESHE = +0.613V) (22)

Total reaction: (E = +1.016V) (23)

To the best of our knowledge, Weston cell is only used for calibration of voltmeters in laboratory since it has a low temperature coefficient, which leads to a highly stable voltage. However, it has never been used in commercial applications like lead-acid rechargeable batteries, perhaps due to the low working voltage and highly toxic heavy metal inventory (mercury and cadmium).

Chapter 1: Introduction

Fig. 1.9 Schematic illustration (top) and photographs (left bottom: internal structure and

right bottom: appearance) of Weston cell.

1.3.2 Nickel-Cadmium (Ni-Cd) battery Nickel-Cadmium (Ni-Cd) battery was proposed by Waldemar Jungner in 1899 [70].

Currently, Ni-Cd batteries have various designs and shapes. However, the cylindrical configuration will be taken as an example since it is the most commonly encountered case. 30

Chapter 1: Introduction

Specifically, a schematic illustration of the main parts of a sealed cylindrical Ni-Cd battery is shown in Fig. 1.10. The material of the exterior case is Ni-plated steel or carbon steel normally covered with a plastic label. The electrode is jelly-roll type, which is separated by a polymer membrane filled with KOH electrolyte. Usually, the anode and the cathode rolled plaques are shifted slightly. The anodic plate has contact with the negative pole at the bottom while the cathodic plate is enclosed at the upper level to connect with the positive pole (see Fig. 1.10). The plates are composed of a perforated steel plate which serves as a support grid, where the electrode materials are attached [71].

Fig. 1.10 Schematic illustration of the Ni-Cd battery construction [71].

Additionally, a schematic diagram of a dismantled waste Ni-Cd battery is presented in Fig. 1.11. In terms of energy storage mechanism, the associated redox reactions are described by the following equations [72]:

Negative reaction: (24)

Positive reaction: (25)

Total reaction: (26)

During the charge process, Ni(OH)2 is oxidized to NiOOH at the positive electrode and cadmium hydroxide is converted to cadmium metal at the negative electrode. On the

Chapter 1: Introduction

contrary, both reactions (24) and (25) are inverted during the discharge. Although the oxidation states of the active materials are changed during charge/discharge processes, their physical states keep the same. For similar reasons, the concentration of KOH electrolyte does not change.

Oxygen produced at the positive electrode is reduced by the negative electrode to generate Cd(OH)2 by oxidation of cadmium metal as shown in equation (27):

(27)

Moreover, a separator is used between the two porous electrodes, which is permeable to

O2 and allows O2 formed at the positive electrode to be transported to the negative electrode [72].

Fig. 1.11 Schematic diagram of a dismantled waste Ni-Cd battery [73].

1.3.3 Cadmium-based redox flow battery

As discussed above in section 1.2.2, the flow battery is regarded to be the large scale power supplier and has attracted considerable attention. Recently, cadmium related redox flow batteries are reported [74, 75]. To be specific, a novel single flow Cd-PbO2 battery was proposed for enhancing the energy efficiency and specific power. It is composed of

PbO2 as positive electrode, deposited cadmium as negative electrode and a mixed solution

-1 of 2 M H2SO4/1M CdSO4 with 0.6 g·L DPE-3 (A kind of short carbon chain condensate 32

Chapter 1: Introduction

of ethylenediamine polycondensate with a mixture of dimethylaminopropylamine and epichlorohydrin) as the electrolyte to control the morphology of deposited cadmium electrode and reduce the self-discharge [74]. Furthermore, the schematic diagram of Cd-

PbO2 single flow battery is shown in Fig. 1.12. The experimental results exhibit that the battery provides the highest capacity efficiency of 98.1 % and energy efficiency of 90.3 % in the eleventh cycle at the current density of 0.4 C. After 300 cycles, the battery still retains 82.3 % energy efficiency and 93.5 % capacity efficiency, and presents excellent cycling performance [74].

Fig. 1.12 Schematic diagram of the Cd-PbO2 single flow battery (1-tank, 2-pump, 3-liquid

separator valve, 4-electrodeposited cadmium negative electrode, 5- PbO2 positive

electrode, 6-overflow valve) [74].

Another example is the iron-cadmium (Fe-Cd) redox flow battery, which is shown in

Fig. 1.13. It relies on the redox couples Cd2+/Cd and Fe3+/Fe2+ in an acidic solution as the negative and positive electrolytes, respectively, which are separated via a separator or an ion-exchange membrane. The redox reactions of the Fe-Cd RFB are described as the following equations [75]: 33

Chapter 1: Introduction

Negative reaction: (ESHE = -0.40V) (28)

Positive reaction: (ESHE = +0.77V) (29)

Total reaction: (E = 1.17V) (30)

During the cycling test, the Fe-Cd RFB gives good capacity retention of 99.87 % per cycle and stable energy efficiency of 80.2 %, which ensures good cyclic stability for practical application [75].

Fig. 1.13 Schematic of Fe-Cd redox flow battery with a premixed cadmium and iron

solution [75].

1.4 Lead-acid battery Lead-acid battery has become one of the most important secondary batteries, it was invented by Raymond Gaston Planté in 1859 [76]. It is composed of porous lead as negative electrode, lead dioxide as positive electrode and diluted sulfuric acid as electrolyte. During the discharge process, the lead dioxide positive electrode and the lead negative electrode will transform into PbSO4 and consume sulfate ions. The related redox reactions are described by the following equations [77]:

Negative reaction: (31) 34

Chapter 1: Introduction

Positive reaction: (32)

Total reaction: (33)

Generally, in the industrial production of lead-acid batteries, leady oxide (various valencies of lead in oxide) is used as original active material of both cathode and anode of the batteries. In formation process, leady oxide can be transformed to lead dioxide in positive electrode and spongy lead in negative electrode [78]. Fig. 1.14 shows the schematic diagram of the assembly process for fabricating lead-acid battery. After mixing and pasting, the positive plates are cured in a curing vessel at a temperature of 80 °C and a relative humidity of 95 %. Subsequently, the cured pasted plates were dried under 70 °C for 20 h. After the formation process the plates were treated in H2SO4 solution with a specific gravity of 1.05 g·cm-3, including two processes. In the first step, the cured plate was soaked in H2SO4 solution at open circuit for 2 h, leading to formation of small amounts of PbSO4. In the second step, the formation of positive plate was conducted under a constant current for 48 h. Finally, each dried positive plate was matched with two commercial negative plates immersed in H2SO4 solution, which implies lead-acid battery was assembled successfully [79].

Chapter 1: Introduction

Fig. 1.14 Schematic diagram of the assembly of lead-acid battery [79].

To the best of our knowledge, the advantages of lead-acid battery are high safety, low cost, easy construction and low self-discharge. However, its biggest drawback is the low energy density of the entire battery system [76]. To enhance energy density, an effective way is to use carbon current collectors such as reticulated vitreous carbon (RVC) instead of metal alloys. Actually, the idea of lead-acid battery construction with RVC, demonstrated in Fig. 1.15 was first proposed by Czerwinski in 1995 [80, 81]

Chapter 1: Introduction

Fig. 1.15 Electrode reactions in prospective lead-acid battery with modified RVC grids

[81].

1.5 Aims and Tasks of this Study

Regarding low cost, safety, easy construction and environment compatibility, aqueous rechargeable batteries have been investigated widely in recent years. However, strong acid or alkaline, which is used as the electrolyte for secondary battery, will cause serious corrosion to equipment. As a result, neutral aqueous electrolyte (or pH value of electrolyte solution close to 7 such as weak alkaline and acid) would be the best choice for aqueous rechargeable battery.

As mentioned above, the electrode active materials of lead-acid battery and Ni-Cd battery contain highly toxic heavy metals such as Pb and Cd. In order to increase the specific capacity of batteries and reduce the amount of heavy metals and acid (or alkaline), the positive electrode materials of lithium ion batteries including LiCoO2 and LiMn2O4 can be used to replace heavy metals. Moreover, a neutral electrolyte Li2SO4 solution can be selected to match the positive and negative electrode materials. In addition, copper nanoparticles as negative electrode are studied in aqueous potassium carbonate electrolyte solution due to its high environmental compatibility.

In conclusion, according to the previous discussion the main tasks in this dissertation are as follows:

(1) Investigation of electrochemical performance of an acid-free rechargeable battery

based on PbSO4 and LiMn2O4;

(2) Investigation of electrochemical performance of an aqueous lithium ion rechargeable

battery with high rate capability based on metallic Cd and LiCoO2;

(3) Study of the electrochemical behavior of amorphous carbon-coated copper/CNT

composites as negative electrode material and their energy storage mechanism.

Chapter 2: Experimental

Chapter 2 Experimental

2.1 Chemicals

Manganese sulfate (MnSO4), ammonium peroxydisulfate ((NH4)2S2O8), ammonium sulfate ((NH4)2SO4), lithium hydroxide (LiOH·H2O), lead sulfate powder (99.9% PbSO4) were used as received.

Lithium nitrate (LiNO3), cobalt nitrate (Co(NO3)2·6H2O), starch ((C6H10O5)n), acetic acid (99% CH3COOH), cadmium acetate (Cd(Ac)2·2H2O), and lithium sulfate (Li2SO4) were used as received without further purification.

Copper acetate (Cu(Ac)2·H2O), ethylene glycol (EG), nitric acid (65% HNO3), D(+) glucose (C6H12O6), potassium carbonate (K2CO3), and absolute ethanol (99% C2H5OH) were used as received without further purification. Aqueous solutions were prepared with deionized ultrapure water (Seralpur Pro 90 C).

2.2 Synthesis of Materials

2.2.1 Synthesis of LiMn2O4

At first, birnessite-MnO2 ( -MnO2) nanowires were prepared via a hydrothermal reaction

◦ from a solution of MnSO4, (NH4)2S2O8 and (NH4)2SO4 in a molar ratio of 1:1:4 at 140 C as described previously [82]. Subsequently, the as-prepared -MnO2 nanowires were milled with LiOH·H2O in ethanol; then the mixture was sonicated for 5 h. After the ethanol was evaporated, the mixture was transferred into a furnace and heat-treated at

700 °C for 8 h to get LiMn2O4 nanocubes.

2.2.2 Synthesis of LiCoO2

LiCoO2 nanoparticles were prepared by a starch-assisted sol-gel method [83]. Briefly, 0.4 g starch was placed in a round-bottom flask and 25 mL deionized ultrapure water was added into the flask. Subsequently, the mixture was heated at 110 ◦C until the solution became transparent under stirring (named as solution A). Secondly, 5 mmol lithium

Chapter 2: Experimental

nitrate and 5 mmol cobalt nitrate were dissolved into 5 mL ultrapure water to get a homogeneous solution (named as solution B). Thirdly, solution B was dropped into solution A under stirring and then this mixture was kept at 110 °C for 2 h. Thirdly, the mixture was dried at 110 °C to get the precursor in the form of a foam. Lastly, the precursor was further calcined at 700 °C for 36 h with heating rate of 2 °C/min to get

LiCoO2 nanoparticles.

2.2.3 Synthesis of C/Cu/CNT composites

Commercial CNTs were treated in 6M HNO3 by refluxing at 60 °C for 2 h to remove impurities and endow the surface with hydrophilic groups such as –OH and –COOH.

After washing with ultrapure water several times, the as-treated CNTs were filtered and collected. Copper nanoparticles were grown on the CNTs through a direct redox reaction between Cu2+ and ethylene glycol (EG) solution by a solvothermal method. This is a facile process, in which only cupric acetate monohydrate and CNTs are used as initiating materials, ethylene glycol as the solvent and reducing agent, without employing any additional surfactant. When cupric acetate monohydrate was dissolved in ethylene glycol, copper ions were solvated in ethylene glycol micelles, which are selectively connected to

CNTs with terminal hydroxyl groups through hydrogen bond interaction. The whole synthesis process is sketched in Fig. 2.1. Briefly, 0.5 g of Cu(Ac)2·H2O and 60 mg of

CNTs were dissolved in 175 mL of EG under ultra-sonication for 30 min. The mixture was subsequently transferred into a 250 mL PTFE-lined stainless steel autoclave, and heated at 180 °C for 48 h. Finally, the as-synthesized product was collected and washed with water and absolute ethanol several times, and then dried at 70 °C for 12 h. For comparison, unsupported copper material was also prepared under the same conditions without CNTs, and no carbon coating was applied.

For carbon coating 0.2 g Cu/CNTs composite was dispersed in distilled water and ethanol (1:3 v/v; 20 mL), and a glucose solution (0.35 g glucose/10 mL ultrapure water)

Chapter 2: Experimental

was added. This solution was sonicated for 30 min and then concentrated to dryness.

Finally, the dried powder was calcined at 600 °C for 10 min and cooled naturally to room temperature. Carbon-coated CNT-nanocomposites without copper were also prepared under the same conditions for comparison.

Fig. 2.1 Preparation process for the C/Cu/CNTs composite.

2.3 Characterization of the as-prepared materials

The spinel LiMn2O4 nanocubes were analysed using X-ray diffraction on a Bruker D4 X- ray diffractometer with Ni-filtered Cu K radiation. Scanning electron micrographs (SEM) were obtained on a Nova NanoSEM NPE207 scanning electron microscope.

Crystal structure of the prepared LiCoO2 nanoparticles was characterized by X-ray powder diffraction (XRD) using Rigaku Rotaflex RU-200B diffractometer with CuKα radiation filtered by a nickel thin plate at 40 kV and 40 mA at a scan rate of 0.02 s-1.

Scanning electron micrographs (SEM) were obtained on a JEOL JSM-7500F scanning electron microscope.

Crystal structures of the prepared Cu and C/Cu/CNTs composite were characterized by X-ray powder diffraction (XRD) using a Rigaku Rotaflex RU-200B diffractometer with CuKα radiation filtered by a nickel thin plate at 40 kV and 40 mA at a scan rate of

Chapter 2: Experimental

0.02 s-1. The XPS measurements were performed with ESCALAB 250Xi XPS

Microprobe (Thermo Scientific) equipped with a monochromatized Al Kα X-ray source

(hν = 1486.6 eV). The pass energy was 200 eV for survey spectra and 20 eV for high- resolution spectra. The built-in charge compensation option was used during the measurement in order to avoid charging effects in the powder samples. Scanning electron micrographs (SEM) and transmission electron micrographs (TEM) were obtained on a

JEOL JSM-7500F scanning electron microscope and a JEOL JEM-2100 transmission electron microscope, respectively. Infrared spectra were obtained on a FTIR spectrometer

Bruker IFS66. Thermogravimetric analysis was performed on a Mettler Toledo 1600 system with an MX1 balance.

2.4 Electrochemical measurements

2.4.1 Preparation of electrodes

The LiMn2O4 nanocubes were mixed with acetylene black and poly(tetrafluoroethylene)

(PTFE) in a weight ratio of 7.5:1.5:1 and dispersed in ethanol. After drying, the mixture was pressed into a film with an active mass loading of about 3.75 mg cm-2; the film was cut into a disk of about 2 mg that was pressed onto Ni-grid at a pressure of 10 MPa, and

◦ finally dried at 120 C for 12 h to serve as the positive electrode. The PbSO4 negative electrode was prepared in the same way as the LiMn2O4 electrode.

The LiCoO2 positive electrode was prepared by pressing a powdered mixture of the prepared LiCoO2 nanoparticles, acetylene black, and poly(tetrafluoroethylene) (PTFE) in a mass ratio of 8:1:1. The obtained film was punched into a small disk of about 2 mg mass, 0.25 cm2 area and 0.3 mm thickness. Finally, these disks were pressed onto stainless steel-grids at a pressure of 10 MPa and then dried at 120 ◦C for 12 h. The cadmium plate (1.5 cm width, 2.5 cm length, 0.1 cm thickness) was used as negative electrode. In order to remove the oxide film on the surface, it was cleaned in 50 % acetic acid using ultrasonic device for 2 minutes before testing. The main reason for choosing acetic acid as

Chapter 2: Experimental

chemical polishing agent is to avoid introducing other ions, such as nitrate and phosphate ions. Additionally, sulfuric acid also cannot be selected as etchant since cadmium plate can react fast in strong acid [84].

The negative electrode was prepared by pressing a powdered mixture of the sample

(C/Cu/CNTs or unsupported Cu), acetylene black, and poly(tetrafluoroethylene) (PTFE) in a weight ratio of 80:10:10. The obtained sheet was punched into small disks of about 2 mg mass, 0.36 cm2 area and 0.4 mm thickness. Finally, these disks were pressed onto stainless steel-grids (Stainless steel 1.4401, mesh 181 with 0.09 mm width and 0.05 mm thickness, F. Carl Schroeter, Hamburg, Germany) at a pressure of 10 MPa and then dried at 120 ◦C for 12 h.

2.4.2 Characterization of electrochemical performance

The cyclic voltammetric (CV) of LiMn2O4 and PbSO4 electrodes was performed in 0.5 M

Li2SO4 aqueous solution in a three-electrode cell with a nickel grid and a saturated calomel electrode (SCE) as counter and reference electrodes, respectively. The CV data were collected on an electrochemical work station CHI660C. A two-electrode cell consisting of the above LiMn2O4 positive electrode and the PbSO4 negative electrode with a distance of about 1 cm was used to test the charge-discharge and cycling behavior in 0.5

M Li2SO4 solution on a Land 2001A cell tester.

For electrochemical impedance measurements (EIM) of Cd//LiCoO2 battery a potentiostat Solartron SI 1287 connected to a frequency response analyzer SI 1255 interfaced to a PC was used. The measurements were carried out at the spontaneously established open circuit potential (OCP) with a modulation amplitude of 5 mV in a frequency range from 105 to 0.1 Hz. Evaluation of the impedance data was performed with Boukamp software version 2.4. Galvanostatic charge and discharge between 0.4 and

1.8 V (vs. Cd2+/Cd) was executed with a two-electrode electrochemical cell, on an

ATLAS 0961 Multichannel Battery Tester and potentiostat-galvanostat. Cyclic

Chapter 2: Experimental

voltammetry (CV) was performed on an IVIUMSTAT Electrochemical Interface with a two-electrode electrochemical cell using 0.5 M Li2SO4 and 10mM Cd(Ac)2 solution as electrolyte.

In the case of C/Cu/CNTs composite electrode, cyclic voltammetry (CV) was executed on a custom built potentiostat connected to a computer with an AD/DA- converter interface using home-developed software with a three-electrode electrochemical cell using 0.5 M K2CO3 electrolyte solution. For electrochemical impedance measurements (EIM) a potentiostat Solartron SI 1287 connected to a frequency response analyzer SI 1255 interfaced to a PC was used. The measurements were carried out at the spontaneously established open circuit potential (OCP) with a modulation amplitude of 5 mV in a frequency range from 105 to 10-3 Hz. Evaluation of the impedance data was performed with Boukamp software version 2.4. Galvanostatic charge and discharge between −0.8 < ESCE < 0 V were performed with a three-electrode electrochemical cell using a saturated calomel electrode (SCE) and a stainless steel mesh as reference and counter electrodes, respectively, on an ATLAS 0961 Multichannel Battery Tester and potentiostat-galvanostat. All electrochemical measurements were performed at ambient temperature.

Chapter 3: An acid-free rechargeable battery based on PbSO4 and spinel LiMn2O4

Chapter 3 An acid-free rechargeable battery based on PbSO4 and

spinel LiMn2O4

As mentioned above, the lead-acid battery is commercially one of the most successful electrochemical systems, and no other battery has yet been able to replace it in the field of energy storage. Its market share has been growing in the last few years, although batteries based on other chemistries are fast catching up [85, 86, 87]. The main reasons for this are its high reliability, easy construction and low cost. Additionally, its raw materials are practically unlimited, and about 95 % are recycled. However, sulphuric acid (H2SO4), which is used as the electrolyte for the lead acid battery, causes serious damage to devices, and Pb leads to serious environmental pollution. Consequently, it is urgently necessary to reduce the amounts of acid and lead being used.

Recently, most research in this field has focused on electrolytes and electrodes of batteries [88, 89]. For example a capacitor-battery system based on a activated carbon negative electrode and PbO2 positive electrode was reported [89]. However, it still uses acidic electrolyte and the amount of Pb only can be reduced by 25 wt% at the most [90].

Additionally, its energy density is below 50% of that of the lead acid battery, lower than

25 W·h·kg-1, albeit its cycling life and rate capability are much improved.

In this chapter, in order to reduce the amount of Pb and acid as well as increase the specific capacity of lead rechargeable battery, we introduce an acid-free lead battery based on spinel LiMn2O4 as the positive electrode and PbSO4 as the negative electrode in

0.5 M Li2SO4 neutral aqueous electrolyte.

3.1 Characterization of morphology and structure of LiMn2O4

The FESEM micrograph of the prepared LiMn2O4 (Fig. 3.1a) shows clearly that this material exists in nanocubes. This is different from the NaxMnO2 nanorods prepared via the heat-treatment of Na2CO3 with MnO2 nanowires [82]. The main reasons are presumably the differences in the crystallization processes. The X-ray diffraction pattern 44

Chapter 3: An acid-free rechargeable battery based on PbSO4 and spinel LiMn2O4

(Fig. 3.1b) shows that the as-prepared LiMn2O4 is highly crystalline with the spinel structure; the characteristic peaks of planes can be clearly identified. This high degree of ordering is caused by the high temperature (700 °C) during the synthesis process.

(b) 111 LiMn O 2 4

311

400

Intensity / a.u. 440

222

531

444

511

551

622

331 533

10 20 30 40 50 60 70 80 90 2Theta / o

Fig. 3.1 (a) Field emission scanning electron micrograph (FESEM) and (b) X-ray

diffraction pattern of the as-prepared LiMn2O4 nanocubes.

3.2 Cyclic voltammetry of PbSO4 and LiMn2O4 electrodes

9 -1 LiMn O 5 mV s 2 4 6 PbSO 4 3

-3 Current / mA / Current

-6

-9 -0.9 -0.6 -0.3 0.0 0.3 0.6 0.9 1.2

Potential / V vs. SCE

Fig. 3.2 Cyclic voltammogramms of PbSO4 and LiMn2O4 electrodes in a 0.5 M Li2SO4

aqueous electrolyte solution at the scan rate of 5 mV·s-1, which was tested using nickel-

grid and SCE as the counter and reference electrodes, respectively, and each electrode is

about 2 mg.

Chapter 3: An acid-free rechargeable battery based on PbSO4 and spinel LiMn2O4

Cyclic voltammogramms (CVs) of PbSO4 and LiMn2O4 electrodes in a 0.5 M Li2SO4 aqueous electrolyte at the scan rate of 5 mV·s-1 are shown in Fig. 3.2. In the case of

PbSO4, there is one set of redox peaks located at -0.6/-0.36 V (vs. SCE), which is due to the redox reaction of PbSO4/Pb. This implies that PbSO4 is stable in neutral aqueous solution and its redox reactions can take place reversibly. Two sets of redox peaks related to LiMn2O4 are situated at 0.62/0.86 V and 0.78/1.0 V (vs. SCE), respectively, which

+ correspond to the intercalation/de-intercalation of Li ions into/from LiMn2O4 in the aqueous electrolyte solution. This is similar to former reports [91, 92, 93]. Since there is a potential difference between LiMn2O4 and PbSO4, they can be assembled into a battery system whose redox reactions during charge/discharge processes are schematically shown in Fig. 3.3.

Fig. 3.3 Schematic illustration of the redox reactions for acid-free PbSO4//LiMn2O4

aqueous rechargeable battery during the charge/discharge processes.

Their electrode and cell reactions are shown simplified in the following equations:

Negative electrode: (34)

Positive electrode: (35)

Total reaction:

(36)

Chapter 3: An acid-free rechargeable battery based on PbSO4 and spinel LiMn2O4

During charging PbSO4 is turned into Pb releasing sulphate anions into the neutral aqueous electrolyte. In the meanwhile, LiMn2O4 is turned into Li1-xMn2O4 by releasing

Li+ cations into the electrolyte. As a result, the released sulphate anions and lithium cations keep the electrolyte always electrically neutral. During the discharge process, Pb gets sulphate anion from the neutral electrolyte to turn into PbSO4 together with the gain

+ of Li ions from the electrolyte into Li1-xMn2O4 to become LiMn2O4. This means that this battery does not need acid solution since the positive electrode is LiMn2O4 instead of

PbO2. The latter needs an acid medium to realize its fast redox reaction to form PbSO4.

According to the definition of aqueous rechargeable lithium batteries (ARLBs) [41,

94], it can be assigned as a 1st generation ARLBs type. During the charge or the discharge process, both electrodes exist as solid phases, and the total volume change is less than those of Zn//LiMn2O4, Zn//LiFePO4 or Zn//NaxMnO2 [82, 95, 96] since Pb or

PbSO4 could not dissolve into the neutral solution as Zn.

3.3 Galvanostatic charge/discharge measurement of PbSO4//LiMn2O4 battery

2.0 140 Charge (b) 1.8 (a) PbSO //LiMn O 120 4 2 4 1.6 -1 0.2 A g 100 1.4 -1 0.4 A g 1.2 0.6 A g-1 Discharge current density -1 80 0.8 A g 0.2 0.4 0.6 0.8 1.0 1.5 2.0 2.5

1.0 -1

1.0 A g -1 60 Rate: A/g

0.8 1.5 A g 3.0 4.0 5.0 Voltage / V 2.0 A g-1 0.6 -1 40 Discharge 3.0 A g 0.4 Capacity / mAh/g 20 0.2 0.0 0 0 20 40 60 80 100 120 140 160 180 0 5 10 15 20 25 30 35 40 45 50 55 Capacity / mAh g-1 Cycle number

Fig. 3.4 (a) Charge and discharge curves and (b) the discharge specific capacity of the

PbSO4/Li2SO4/LiMn2O4 battery at different current densities between 0 and 1.8 V. The

data were calculated based on the mass of the LiMn2O4.

The electrochemical performance of the PbSO4//LiMn2O4 battery at different current

Chapter 3: An acid-free rechargeable battery based on PbSO4 and spinel LiMn2O4

densities between 0 and 1.8 V is shown in Fig. 3.4 (a). At the lower current densities two discharge-charge voltage platforms at about 1.27/1.35 V and 1.44/1.23 V, respectively, can be identified, which are consistent with the CV results. This is due to the intercalation/deintercalation of Li+ ions into/from different sites of the spinel host. Its average discharge voltage is 1.30 V. The reversible capacity of the battery is 128 mA·h·g-

1 -1 at 200 mA·g based on the LiMn2O4 positive electrode, which is similar to the reported value with aqueous electrolytes [91, 92, 93]. When the charge-discharge current increases from 0.4, 0.6, 0.8, 1.0, 1.5, 2.0 to 3.0 A·g-1, the discharge capacity decreases from 117.5,

114.6, 110.9, 108.8, 97.5, 94.6, to 78.4 mA·h·g-1, respectively. This suggests that the acid-free lead battery can be charged and discharged fast, and can meet the rapid change of power supply and demand from the grid (Fig. 3.4b). However, the rate capability of

+ LiMn2O4 (vs. Pb/PbSO4) cannot be compared with those of LiNi0.5Mn1.5O4 (vs. Li /Li) and Li[Ni1/3Co1/3Mn1/3]O2 (vs. SCE) [97, 98]. The main reason for this is the slow redox kinetics of Pb/PbSO4; this is also a key problem for the lead acid rechargeable batteries

[99].

In the case of the lead acid battery, about half of the theoretical capacity of the negative electrode can be obtained practically. Applying this to the present battery system

-1 the practical capacity of PbSO4 would be about 88 mA·h·g . Combining the practical

-1 capacity of LiMn2O4 (128 mA·h·g ) and the average discharge voltage (1.3 V), the calculated energy density based on the two electrodes will be 68 W·h·kg-1. This value is comparable with that of lead acid batteries.

3.4 Cycling performance of PbSO4//LiMn2O4 battery

The cycling performance of the PbSO4//LiMn2O4 at the full discharge capacity of 128

-1 -1 mA·h·g based on LiMn2O4 at the current density of 400 mA·g (Fig. 3.5) shows that the

Coulombic efficiency is almost 100 % except in the initial cycles; this again is similar to lithium ion batteries. This means that the acid-free lead battery shows good charge and

Chapter 3: An acid-free rechargeable battery based on PbSO4 and spinel LiMn2O4

discharge reversibility. The spinel LiMn2O4 is very stable, its redox potential is below that of oxygen evolution and it can be cycled up to 10,000 full cycles [91]. For the redox couple Pb/PbSO4, Fig. 3.2 shows clearly that its redox potential is above that for hydrogen evolution due to the large hydrogen overpotential of Pb. In the case of LiMn2O4, its redox potentials are also below that of oxygen evolution, which has been well illustrated in ARLBs [30, 91, 93]. As a result, there are basically no side reactions

(gassing) resulting in 100 % Coulombic efficiency which is evidently higher than that of the lead acid battery. Another reason is that the charge cut-off voltage is 1.8 V, which is much lower than 2.4 V for the lead acid battery. If it is used for energy storage, this means that it is superior to the lead acid battery since its self-discharge rate will also be lower due to the lower working voltage and less irreversible reactions. After 110 full cycles, the capacity loss based on LiMn2O4 is less than 1 % compared with the initial cycles. In the case of the Pb/PbSO4 electrode, it is known that there is no phase change and dendrite problem like with Zn, and it will be stable during cycling. As a result, the cycling behavior of this battery system will be superior to that reported for Zn//LiMn2O4 and Zn//NaxMnO2 [95, 96], which can be above 1000 cycles. Of course, further research on the extended cycling is needed.

160 160 PbSO //LiMn O 4 2 4 140 140% / efficiency Coulombic

-1 120 120

100 100

80 80

60 60

Capacity / mAh g 40 40

20 20 0 10 20 30 40 50 60 70 80 90 100 110 Cycle number

-1 Fig. 3.5 Cycling performance of the battery at 400 mA·g based on LiMn2O4.

Chapter 3: An acid-free rechargeable battery based on PbSO4 and spinel LiMn2O4

3.5 Conclusions

In conclusion, an acid-free lead battery has been assembled using spinel LiMn2O4 nanocubes as the positive electrode, PbSO4 as the negative electrode, and an aqueous solution of 0.5 M Li2SO4 as the electrolyte. Its average discharge voltage is 1.3 V and the

-1 specific capacity is stable at 128 mA·h·g based on the LiMn2O4. The calculated energy density is 68 W·h·kg-1 based on the practical capacities of the two electrodes. The battery shows less than 1 % capacity loss after 110 full cycles at the current density of 400 mA g-

1 between 0 and 1.8 V and 100 % Coulombic efficiency except for the initial several cycles. The above results exhibit that the positive electrode of the lead acid battery (PbO2) can be completely substituted by the environmentally friendly and cheap LiMn2O4, which implies that 50 % of Pb can be saved. In addition, H2SO4 is not needed. It will be more userfriendly to the operators. This battery system can be a promising candidate as a power supply for electric bicycles or for energy storage on a large scale.

Chapter 4: An aqueous lithium ion rechargeable battery with high rate capability based on metallic Cd and LiCoO2

Chapter 4: An aqueous lithium ion rechargeable battery with high

rate capability based on metallic Cd and LiCoO2

In the past decades, most research related to cadmium has focused on Ni-Cd rechargeable batteries and some redox flow batteries, which have been discussed in section 1.3.

However, they also met the same problem as the lead-acid battery. That is, acid or alkaline cause serious corrosion to device, and heavy metals lead to serious environmental pollution. In addition, the energy density of Ni-Cd battery is less than that of lithium ion battery [100]. Consequently, these drawbacks should be solved urgently.

In this chapter, a similar strategy will be adopted to tackle these problems. That is to combine the advantages of Ni-Cd battery and lithium ion battery. To be specific, an aqueous lithium ion battery based on metallic Cd as the negative electrode and LiCoO2 as the positive electrode in 0.5 M Li2SO4 and 10mM Cd(Ac)2 neutral aqueous electrolyte will be investigated below.

4.1 Characterization of morphology and structure of LiCoO2

° SEM micrograph and X-ray diffraction pattern of the prepared LiCoO2 at 700 C are shown in Fig. 4.1. As we can see from Fig. 4.1a, the LiCoO2 nanoparticles are about 50-

100 nm, which are agglomerated to become submicron-structure for bulk. In Fig.4.1b, all diffraction peaks can be indexed to LiCoO2 in accordance with literature values (JCPDS

50-0653) [101]. The peaks at 19.0 °, 37.4 ° and 45.0 ° can be ascribed to the (003), (101) and (104) planes of LiCoO2, respectively, which shows that it has a rhombohedral lattice

(R3m space group) and a hexagonal layered α-NaFeO2 crystal structure. This layered structure enables a facile intercalation/deintercalation of lithium ions in the inner space, which in return determines the charge and discharge efficiency of batteries [102].

Chapter 4: An aqueous lithium ion rechargeable battery with high rate capability based on metallic Cd and LiCoO2

(a) (b) (003)

(104)

(101) (110)

(018) (113) Intensity / a.u. (009)

10 20 30 40 50 60 70 80 90 2 Theta / o

Fig. 4.1 (a) SEM micrograph and (b) XRD patterns of LiCoO2 nanoparticles.

4.2 Cyclic voltammetry (CVs) of Cd//Cd and Cd//LiCoO2 battery

Fig. 4.2a shows the cyclic voltammograms of metallic Cd electrode at different scan rates

-1 from 1 to 100 mV·s in 0.5M Li2SO4 and 10mM Cd(AC)2 aqueous electrolyte solution in

2+ the range -0.5 < E (vs.Cd /Cd) < 0.5 V. It can be seen clearly that the CV curves reflect the good reversibility of redox reaction of Cd2+/Cd due to the existence of highly symmetric redox peaks even at high scan rate of 100 mV·s-1. It also means that metallic cadmium electrode has high rate capability. The reduction peak potential decreases slightly and the reduction peak current increases with the scan rate from 1 to 100 mV·s-1, indicating the fast kinetics of Cd2+/Cd. It also can be proved by analyzing the CV data using the Randles-Sevcik equation as below [32]:

3 1/2 3/2 1/2 1/2 Ip = 0.4463 (F /RT) n AD cv

= 2.69×105 n3/2AD1/2cv1/2 (37) where slope b is 2.69×105 n3/2AD1/2c, n is the number of electrons transferred in the redox process (here n is 2), A is the electrode area (geometric surface area of the electrode is used instead, here 3.75 cm2), F is the Faraday constant (in C·mol-1), c is the concentration of Cd2+ ions (here 0.01 mol·L-1) and D is diffusion coefficient (in cm2·s-1). According to equation (37), the plot of the reduction peak current (Ip) versus the square root of scan rate

Chapter 4: An aqueous lithium ion rechargeable battery with high rate capability based on metallic Cd and LiCoO2

(v1/2) should be a straight line, which is presented in Fig. 4.2b. It shows a function y = a + b·x and the slope b is obtained directly from linear fitting. Based on above-mentioned parameters, the cadmium ion diffusion coefficient was further calculated as 1.85×10-5 cm2·s-1. These results also mean that the Cd2+/Cd reduction reaction is diffusion- controlled. Apparently, the diffusion coefficient is greater than that of lithium ions in

-10 2 −1 LiCoO2 which is generally about 10 cm ·s [102]. In addition, this is also comparable to the oxygen diffusion coefficient in aqueous and organic electrolyte solutions for Li//air batteries, which is about 7×10-6 cm2·s-1 and 1.67×10-5 cm2·s-1, respectively [103, 104].

0.045 0.04 (a) (b) 0.040 0.03 0.035 CV data 0.02 Linear fit 0.030 0.01 0.025

0.00

1mV/s 0.020 2mV/s -0.01 5mV/s 0.015

CurrentA / 10mV/s -0.02 Equation y = a + b*x 20mV/s 0.010 Adj. R-Square 0.9978

-0.03 40mV/s Peakcurrent A / Value Standard Error 60mV/s 0.005 CV data Intercept 7.45101E-4 3.83377E-4 -0.04 80mV/s CV data Slope 0.00388 6.44961E-5 100mV/s 0.000 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0 1 2 3 4 5 6 7 8 9 10 11 1/2 1/2 -1/2 Voltage / V vs.Cd2+/Cd v / mV s

Fig. 4.2 (a) two-electrode CV curves of Cd//Cd in 0.5M Li2SO4 and 10mM Cd(Ac)2

aqueous electrolyte with different scan rates and (b) plot of the cathodic peak-current Ip

versus v1/2 the square root of different scan rates.

Cyclic voltammogramms of two-electrode Cd//Cd-setup and Cd//LiCoO2 full batteries in 0.5 M Li2SO4 and 10 mM Cd(Ac)2 aqueous electrolyte at the scan rate of 5 mV·s-1 are shown in Fig. 4.3. In the case of Cd//Cd two-electrode, there is one couple of redox peaks situated at -0.13/0.12 V (vs. Cd2+/Cd), which is due to the plating/dissolution of Cd or the redox reaction of Cd2+/Cd. This suggests that the metallic cadmium electrode has good reversibility in this neutral aqueous electrolyte. One set of redox peaks related to

2+ Cd//LiCoO2 full battery is located at 1.15/1.61 V (vs. Cd /Cd), which corresponds to the

Chapter 4: An aqueous lithium ion rechargeable battery with high rate capability based on metallic Cd and LiCoO2

+ intercalation/de-intercalation of Li ions into/from LiCoO2 in the aqueous electrolyte.

These data are lower than those for Cd//PbO2 battery [74] and cadmium-iron RFB [75] owing to the different redox couples for positive active materials. In the case of Cd2+-ion, its ionic radius is 97 pm, which is larger than that of Li+ (60 pm). Moreover, its charge is

+ +2 and cannot easily go into the crystal site of the Li ion in LiCoO2. Because there is a potential difference between Cd and LiCoO2, they can be assembled into a rechargeable battery. Meanwhile, the mechanism of redox reactions during charge/discharge processes is schematically shown in Fig. 4.4.

5 5 -1 De-intercalation

4 5mVs 4 -2 mAmg / density Current 3 Dissolution 3 2 2 Cd//Cd Cd//LiCoO2 1 1

0 0

-1 -1

-2 -2

-3 Plating -3 -4 -4 Currentdensity / mAcm Intercalation

-1 -5 -5 -0.6 -0.3 0.0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 2+ Voltage / V vs.Cd /Cd

Fig. 4.3 CV curves of Cd//Cd two-electrode and Cd//LiCoO2 full battery in 0.5M Li2SO4

-1 and 10mM Cd(Ac)2 aqueous electrolyte at the scan rate of 5 mV·s .

Their electrode and total cell reactions are expressed simplified by the following equations:

Negative electrode: (38)

Positive electrode: (39)

Cell reaction: (40)

During the charge process Cd2+ is deposited onto Cd electrode by gaining two electrons in the neutral aqueous electrolyte. At the same time, LiCoO2 is turned into Li1-xCoO2 by

Chapter 4: An aqueous lithium ion rechargeable battery with high rate capability based on metallic Cd and LiCoO2

liberating Li+ ions into the electrolyte solution. Consequently, the added Cd2+-ions and the released Li+ ions keep the charge always balanced. During the discharge process, Cd

2+ electrode loses electrons and liberates Cd cations into neutral electrolyte, whereas Li1-

+ xCoO2 gains Li ions from the electrolyte to become LiCoO2. This implies that this battery does not need alkaline solution and mercury heavy metal compared to Ni-Cd battery and Weston cell, respectively. The main reason for this is that the positive electrode is replaced by LiCoO2. Depending on the definition of aqueous rechargeable lithium ion batteries (ARLIBs), it can be classified as a first generation ARLIB type [41].

Fig. 4.4 Schematic illustration of the redox reactions in a Cd//LiCoO2 aqueous

rechargeable battery during the charge/discharge processes.

4.3 Impedance measurement of Cd//LiCoO2 battery

The impedance of Cd//LiCoO2 full battery was measured before and after several CV cycles at open circuit potential, results are shown in Fig. 4.5. The obtained impedance data were further analyzed by the complex nonlinear least squares fitting (NLSF) method

[5, 105 ] according to the equivalent circuit given in Fig. 4.5c. The simulated and experimental impedance data are displayed in Figs. 4.5a and 4.5b. They are similar in form with an arc in the high frequency range and a line in the low frequency range.

Chapter 4: An aqueous lithium ion rechargeable battery with high rate capability based on metallic Cd and LiCoO2

900 180 (a) (b) 800 160 Experimental data Experimental data 700 140 Simulated data Simulated data 600 120 500

100

400 80 300

-Z'' / Ohm -Z'' / Ohm 200 40 20 100

0 0

0 10 20 30 40 50 60 70 80 0 100 200 300 400 500 600

Z' / Ohm Z' / Ohm

W CPE

Rsol Rct Zdiff Q

Fig. 4.5 Nyquist plots of experimental data and simulated data for Cd//LiCoO2 battery (a)

before and (b) after 50 cycles, (c) equivalent circuit used for fitting impedance spectra.

Table 2: Results of impedance evaluation

Rsol/Ω Rct/Ω CDL/F Yo(=1/Q, n=1)/S Yo(=1/Zdiff)/S

Before cycling 8.15 11.27 1.34×10-5 1.33×10-3 5.60×10-3

After cycling 8.55 62.23 4.88×10-5 4.99×10-3 1.58×10-3

Chapter 4: An aqueous lithium ion rechargeable battery with high rate capability based on metallic Cd and LiCoO2

The fitted values of the elements in the equivalent circuit are listed in Table 2. The intercept at highest frequency at the real part (Z’) axis represents the solution resistance

(Rsol). Obviously, Rsol is almost the same before and after several cycles for rechargeable battery. However, a small difference is observed for the semicircle in the high frequency region, corresponding to the double layer capacitance (Cdl) formed on the active materials interface and the charge transfer resistance (Rct) caused by Faradaic reactions of positive and negative electrode. After cycling, the Rct = 62.23 Ω is greater than that before cycling

(Rct =11.27 Ω), which might be attributed to the low conductivity of the by-product

Cd(OH)2 produced by overcharge with hydrogen evolution or the Cd dendrites deposited by cadmium ions [106, 107]. At low frequency, the linear part is ascribed to the frequency dependence of ion diffusion/transport in the electrolyte solution, which is so-called

Warburg diffusion [105]. The constant phase element (QCPE) represents the non-uniform charge distribution at the grain interfaces and describes the deformed nature of the semicircles in Nyquist plots, which is expressed as follow [108]:

n -1 Q(CPE) =(Yo(jω) ) (41) where Yo is determined by combination of some properties of both electro-active part and surface. The exponent n is frequency power, when n = 0, 1 and -1, it represents a resistor, capacitor and inductor, respectively; when n = 0.5, it describes Warburg diffusion.

4.4 Galvanostatic charge/discharge measurement of Cd//LiCoO2 battery

The electrochemical performance of Cd//LiCoO2 battery at different current densities between 0.4 and 1.8 V is presented in Fig. 4.6. The reversible capacity of the battery is

-1 -1 122.3 mA·h·g at current density of 0.2 A·g based on the LiCoO2 positive electrode, which is similar to previously reported data with aqueous electrolyte [19, 109]. In this case, the average discharge/charge voltages are 1.2 and 1.6 V, respectively, which is consistent with the CV results. When the charge/discharge current densities increase from

0.4, 0.6, 0.8, 1.0, 1.5, 2.0 to 3.0 A·g-1, the discharge capacity decreases from 118.5, 111.5,

Chapter 4: An aqueous lithium ion rechargeable battery with high rate capability based on metallic Cd and LiCoO2

-1 103.6, 100.1, 92.9, 82.3 to 69.4 mA·h·g , respectively. This implies that the Cd//LiCoO2 battery can be charged and discharged quickly, and can meet the fast change of power supply and demand from portable and mobile devices. Moreover, this high rate capability is comparable with those of Zn//LiNi1/3Co1/3Mn1/3O2 and Zn//Na0.95MnO2 batteries [11,

82]. The main reason for this is the fast redox kinetics of Cd2+/Cd, which is in agreement with CV results (see Fig. 4.2a). In terms of negative Cd electrode, the theoretical capacity is 477 mA·h·g-1. Applying about a quarter of this to the present battery system, the practical capacity of Cd metal would be about 119 mA·h·g-1. Combining the capacity of

-1 LiCoO2 (122 mA·h·g ) and the average discharge voltage (1.2 V), the further calculated energy density will be 72 W·h·kg-1 based on the two electrodes. This value rivals that of

Ni-Cd batteries.

2.0 Charge 1.8

1.6 0.2 Ag-1 0.4 Ag-1 1.4 0.6 Ag-1 0.8 Ag-1 1.0 Ag-1 1.2 1.5 Ag-1 2.0 Ag-1

1.0 3.0 Ag-1 VoltageV / 0.8 Discharge 0.6

0.4 0 20 40 60 80 100 120 140 160

Capacity / mAh g-1

Fig. 4.6 Charge-discharge curves of Cd//LiCoO2 battery at different current densities

based on the LiCoO2 positive electrode.

4.5 Cycling performance of Cd//LiCoO2 battery

The cycling performance of the Cd//LiCoO2 battery at the full discharge capacity of 107

-1 -1 mA·h·g based on LiCoO2 at the current density of 700 mA·g (Fig. 4.7) exhibits that the

Chapter 4: An aqueous lithium ion rechargeable battery with high rate capability based on metallic Cd and LiCoO2

Coulombic efficiency is about 100 % except for the initial several cycles, which is also similar to lithium ion batteries. Meanwhile, it also means that this battery system displays excellent charge/discharge reversibility. After 100 full cycles, the capacity retention is

75.8 % compared to the first cycles. The possible reasons for this are discussed as follows.

In the case of the LiCoO2 positive electrode, the cyclic stability and high rate capability are affected by nanosize of particles. Specifically, nanocrystalline LiCoO2 has an excellent rate capability compared to bulk LiCoO2. Conversely, the smaller LiCoO2 nanocrystals cause the lower cycling stability compared with bulk LiCoO2 [110]. As a result, both of them contribute to this battery system due to the existence of nanocrystalline and bulk material in the LiCoO2 positive electrode (see Fig. 4.1a).

Another reason is that water splitting is slightly occurring around the potential of the redox reaction of Cd2+/Cd according to the Pourbaix diagram [109]. Furthermore, the formation of Cd(OH)2 by-product promotes the depletion of metallic cadmium, which leads to poor cycling number. Of course, much more efforts should be made to improve cycling performance in future.

200 120

180 100 160 % / efficiency Coulombic

140 Cd//LiCoO 80 120 2

100 60

80 40 60

40 20 Discharge Capacity / (mAh/g) 20 0 0 0 10 20 30 40 50 60 70 80 90 100 Cycle number

-1 Fig. 4.7 Cycling behavior of the battery at 700 mA·g based on the LiCoO2.

Chapter 4: An aqueous lithium ion rechargeable battery with high rate capability based on metallic Cd and LiCoO2

4.6 Conclusions

In conclusion, an aqueous rechargeable lithium ion battery has been assembled using metallic Cd as the negative electrode, LiCoO2 nanoparticles as the positive electrode, and an aqueous neutral electrolyte solution of 0.5M Li2SO4 and 10mM Cd(Ac)2. Its average discharge voltage is 1.2 V and the specific discharge capacity is 107 mA·h·g-1 based on

-1 the LiCoO2 at the current density of 700 mA·g . After 100 cycles, the capacity retention is 75.8% compared to the first cycles between 0.4 and 1.8 V. The calculated energy density is 72 W·h·kg-1 based on the practical capacity of the two electrodes. As mentioned above, the results demonstrate that the positive electrode (Hg) of Weston cell can be totally replaced by the LiCoO2, which implies that 100 % of Hg can be saved.

Additionally, alkaline electrolyte is not needed compared with Ni-Cd batteries.

Consequently, this battery system can be a good candidate as a power supply for portable devices or for large scale energy storage.

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

Chapter 5 Improved electrochemical behavior of amorphous carbon- coated copper/CNT composites as negative electrode and their energy

storage mechanism

To date, a variety of elementary metals are attractive alternatives to lithium- or carbon- based negative electrodes in batteries, including Na, Mg, Al, Fe, Pb, Cd and Zn etc, which have been mentioned in the introduction part. All batteries based on pure metals as negative electrode have much higher theoretical capacities because of the smaller atomic weight of elements compared to their corresponding oxides, according to the equation C =

(F·n)/(3.6·M) (C, F, n and M refer to capacity, Faraday constant, number of transferred electrons per unit and atomic/molecular mass, respectively) [10]. However, less reports focused on copper in batteries except for Daniell-type and all-copper flow batteries as described in section 1.2. In addition, few reports on copper-infiltrated carbonized wood monoliths [111], copper-coating of activated carbon used as electrode material in a Li-ion capacitor [ 112 ], copper-doped metal oxides [ 113 ] and copper as substrate [ 114 ] demonstrate electrochemical activity of copper in alkaline electrolyte solution. Therefore, copper is a promising candidate electrode material for high power energy storage due to its high conductivity and environmental compatibility [112, 113].

Generally, composite electrode materials which contain metal elements in nanostructures on high specific surface area as conductive support, such as carbon nanotubes (CNTs) [115], carbon fiber paper [116], carbon nanofoams [117], graphene

[118] or templated mesoporous carbon [119], are assumed to be a perfect and practical approach to optimize the electrochemical performance both as electrode in secondary batteries and in supercapacitors. Furthermore, CNTs reveal ideal properties for the use as template material due to their extreme aspect ratio and the fact that they are synthesized of abundant, cheap and nontoxic carbon. However, the realization of CNT-based applications requires the formation of reliable contacts to metals, either on the walls or on 61

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

the tips of the CNTs [120].

Coating with carbon is the most commonly used technique to improve the cycling performance of electrode materials. Carbon coating could increase the electronic conductivity of the nanocomposite and prevent chemical surface reactions of the material inside with the electrolyte solution. Moreover, it may also act as a structural buffering matrix to cushion the mechanical stress caused by large volume change during the charge/discharge process [121].

In this chapter, we prepare a hybrid composite as high current electrode by inserting

CNTs into copper nanoparticles finally coated with carbon. CNTs possess high surface area and serve as a conducting additive enabling better utilization of the active material without providing too much weight contribution of inert material in the electrode.

Moreover, CNTs can be modified by acid treatment to remove impurities and to endow the surface with hydrophilic groups such as –OH and –COOH. The electrochemical performance was studied with an aqueous potassium carbonate electrolyte solution because of its high environmental compatibility, results are discussed and compared with those obtained with unsupported copper material. Full characterization of all materials was performed before and after contact with this solution.

5.1 Characterization of morphology and structure of C/Cu/CNT composites

XRD patterns of unsupported copper, C/CNT and C/Cu/CNT composites are shown in

Fig. 5.1. All diffraction peaks can be indexed to copper in good agreement with literature values (JCPDS 4-0836) [122]. The peaks at 43.4°, 50.4° and 74.1° can be ascribed to the

(111), (200) and (220) planes of metallic copper, respectively. Moreover, peak positions of C/Cu/CNT composites are the same as those of unsupported copper. The diffraction peak at about 2θ = 26° of the composite assigned to the (002) plane of carbon is weaker than that of other diffraction peaks proving nevertheless the existence of carbon in C/CNT.

Meanwhile, this broad carbon peak is different from the other sharp diffraction peaks

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

indicating the amorphous microstructure of carbon [123].

(111) C@CNT Unsupported Cu

C@Cu@CNT

Intensity / a.u. / Intensity (002)

(200) 20 21 22 23 24 25 26 27 28 29 30 o

2 Theta / (220)

Intensity / a.u.

10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 2 Theta / o

Fig. 5.1 XRD patterns of C/CNTs, unsupported copper and the C/Cu/CNTs composite.

(111)

CNT (002)

Intensity / a.u. / Intensity

(200) Intensity / a.u. 20 21 22 23 24 25 26 27 28 29 30 2 Theta / o (220)

10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 2 Theta / o

Fig. 5.2 XRD pattern of the Cu/CNTs composites.

Diffraction peak positions of Cu/CNT composites (see Fig. 5.2) are the same as those of unsupported copper. The diffraction peak at about 2θ = 26° (see insert) of the composite assigned to the (002) plane of carbon is weaker than that of other diffraction peaks proving the existence of carbon from CNTs, which present the same peak position compared to carbonized carbon coating of the C/Cu/CNT composite (see Fig. 5.1).

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

Fig. 5.3 (a) and (b) SEM micrographs, (c) and (d) TEM micrographs, (e) and (f) HRTEM

micrographs of unsupported copper and the C/Cu/CNTs composite, respectively.

SEM and TEM micrographs of unsupported copper and C/Cu/CNT composite are shown in Fig. 5.3. Evidently, copper (Figs. 5.3a and 5.3c) exists as irregular nano- and sub-microsized particles, about 10–300 nm in diameter. Fig. 5.3b shows a typical SEM image of the C/Cu/CNT composite, which exhibits a 1D CNT structure with diameters ranging from 20 to 50 nm and coarse surfaces. Moreover, carbon layer and less CNT doped into sub-microstructure copper particles can be clearly seen in Figs. 5.3b and 5.3d due to the agglomeration of copper/EG nanoclusters. Such hybrid structures increase the

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

conductivity of the C/Cu/CNT composite, promoting subsequently the surface redox reaction as confirmed by the substantially diminished charge transfer resistance Rct as determined with impedance measurements (see below). HRTEM was employed to further characterize the structure of the unsupported copper (Fig. 5.3e) and C/Cu/CNT composite

(Fig. 5.3f). Fig. 5.3e shows that there are two types of lattice fringes with lattice spacings of approximately 0.2 and 0.18 nm, corresponding to the (111) planes and the (200) planes of the metallic copper particles, respectively. In Fig. 5.3f, lattice fringes of the metallic copper nanoparticles and an amorphous carbon layer, which is produced by the carbonization of the glucose during the annealing processes, are also observed on the surface of CNT.

In Fig. 5.4 (a) substantial sub-micrometer sized copper particles are agglomerated and supported on CNTs. However, this kind of fluffy structure can easily be saturated with electrolyte, contrary to the behavior of the compact structure of unsupported copper.

Moreover, this structure is helpful for the subsequent coating process. Fig. 5.4 (b) shows the morphology of CNTs with 20-40 nm diameter.

(a) (b)

Fig. 5.4 SEM micrographs of (a) the Cu/CNTs composites and (b) CNTs.

5.2 Infrared spectra of CNTs and C/CNTs composites

The IR spectra of CNTs and C/CNTs composite are shown in Fig. 5.5. The C/CNTs

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

composite shows a strong absorbance at 3400 cm-1 which can be attributed to the O-H stretching vibration. A weak absorbance at 2930 cm-1 is assigned to the C-H stretching vibration. The peaks at 1700 cm-1 and 1100 cm-1 can be ascribed to the C=O and C-O stretching vibration, respectively. However, there is only one weak peak that can be seen at 3400 cm-1 for the CNTs, which is also found with the C/CNTs composite. This implies that glucose was not carbonized completely.

CNTs

C@CNTs C-H C-O C=O

O-H Transmittance / %

4000 3500 3000 2500 2000 1500 1000 500 -1 Wavenumber / cm

Fig. 5.5 IR spectra of the CNTs and the C/CNTs composite.

5.3 Thermogravimetic analysis of C/Cu/CNT composites

130 120 110 24.27 % 100 11.89 % 90 80 35.60 % 70

60 Unsupported Cu 50 Cu@CNT C@Cu@CNT 40 C@CNT 30 98.90 % MassRetained (wt.%) 20 10 0 0 100 200 300 400 500 600 700 800 Temperature (oC)

Fig. 5.6 Thermogravimetic analysis of the unsupported copper, the C/CNTs, the Cu/CNTs

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

and the C/Cu/CNTs composite in air.

The TG curves of unsupported copper, Cu/CNT, C/Cu/CNT and C/CNT are shown in Fig.

5.6. The TG curve of copper shows a weight increase between 200 and 450 °C, which can be attributed to the oxidation of Cu to CuO. Cu/CNT and C/Cu/CNT composite show behavior similar to that of copper below 450 °C. However, the slope of the curves for the composites is smaller than for copper. This is mainly due to the presence of carbon and

CNT reducing the percentage of copper in the composites. The weight loss related to the combustion of amorphous carbon and CNT continues between 450 and 600 °C. The

C/CNT composite can be combusted almost totally. The mass percentages of Cu in the

Cu/CNT and C/Cu/CNT composite were estimated to be about 70.0 % and 50.9 %, respectively.

5.4 BET surface area measurement of C/Cu/CNT composites

400

(a) 0.005 (b) 350

300 0.004

Unsupported Cu C/Cu/CNTs

) ) 250 Cu/CNTs -1 Cu/CNTs Å C/Cu/CNTs -1 0.003

g Unsupported Cu

200 3

/g STP

3 cm

(

cm 150 0.002

(

ads 100

V dV/dD 0.001 50

0 0.000 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 10 100 1000 P/P Diameter (Å) O

Fig. 5.7 (a) Nitrogen adsorption/desorption isotherms and (b) Pore size distribution of

unsupported Cu, Cu/CNTs and C/Cu/CNTs composites.

The specific surface areas and the porosities of the unsupported Cu, Cu/CNTs and

C/Cu/CNTs composites are determined from nitrogen adsorption isotherms as shown in

Fig. 5.7. The BET surface area of the C/Cu/CNTs composite is up to 164.1 m2·g-1 due to the substantial contribution of C/CNTs. It is greater than that of the unsupported Cu (8.1

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

m2·g-1) and Cu/CNTs composites (73.4 m2·g-1). The pore diameter of the C/Cu/CNTs composites is smaller than that of the other materials (see size distribution in Fig. 5.7 (b)).

The main reason is that coating with carbonized carbon blocks or shrinks the channels.

However, the mesoporous structure of the Cu/CNTs composites could be very beneficial

2- for the access of bigger CO3 ions, which corresponds to the fluffy structure in Fig. 5.4 (a), confirming the electrochemical results below.

5.5 XPS measurements of C/Cu/CNT composites

Experimental Data C/Cu/CNTs 2 (b) Cu(0) (a) SP Carbon Cu/CNTs C-C O-C=O Unsupported Cu Background Envelope Residuals

Intensity (arb. units) Intensity (arb. units)

960 950 940 930 300 295 290 285 280

Binding energy (eV) Binding energy (eV)

Experimental Data Experimental Data 2 2 (d) SP Carbon (c) SP Carbon 3 SP3 Carbon SP Carbon C-O-C C-O-C O-C=O Carbon Ring Carbon Ring O-C=O Background Background Envelope Envelope

Residuals Residuals

Intensity (arb. units)

300 295 290 285 280 300 295 290 285 280 Binding energy (eV) Binding energy (eV)

Fig. 5.8 (a) Cu 2p spectra of unsupported copper, Cu/CNTs and C/Cu/CNTs composites and C 1s spectra of unsupported copper (b), Cu/CNTs (c) and C/Cu/CNTs composites (d).

XPS measurements were performed and the corresponding spectra are presented in Fig.

5.8. The Cu 2p3/2 and Cu 2p1/2 peaks with binding energies of 932.63 and 952.56 eV, respectively, correspond to metallic copper [124]. For the C 1s spectra, a spectrum typical 68

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

of a sample with high concentration of sp2 carbon, which is known to have the main peak at ~284 eV, can be observed with all materials. The spectrum of the unsupported copper

(Fig. 5.8 b) shows only a relatively weak peak of carbon, which could be a residue from

- both ethylene glycol and CH3COO . Apart from the main peak in Figure 5.8 c and d, one or more satellite features, several eV from the main C1s peak, are also observed. This can be ascribed to the carbon coating and CNTs.

5.6 Energy storage mechanism and cyclic voltammetry of C/Cu/CNT composites

0.4 (a) (b) (111) # : Cu # * : Cu O 0.3 1 mV / s 2 -1 0.2 (111) (200) * # (200) (220) After Cycling * (220) 0.1 * #

0.0

-0.1 Intensity / a.u. -0.2

Current Density / A g C@Cu@CNT -0.3 Unsupported Cu Before Cycling -0.4 C@CNT

-0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 Potential / V vs. SCE 2 Theta / o 3.0 3.5 (c) 3.0 (d) 2.5 1 mV/s 2.5 2 mV/s 2.0

2.0 5 mV/s -1 -1 10 mV/s 1.5 1.5 20 mV/s 1.0

1.0 30 mV/s 0.5

50 mV/s 0.0

0.5 -0.5 1 mV/s -1.0 0.0 2 mV/s -1.5 5 mV/s

-0.5 -2.0 10 mV/s Current Density / A g Current Density / A g 20 mV/s -2.5 -1.0 30 mV/s -3.0 50 mV/s -1.5 -3.5 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 Potential / V vs. SCE Potential / V vs. SCE

Fig. 5.9 (a) CVs of the C/CNTs, the unsupported copper, and the C/Cu/CNTs composite

electrodes at 1 mV·s-1 scan rate, (b) XRD patterns of the unsupported copper electrode

before and after CVs, (c) and (d) CV curves of the unsupported copper and the

C/Cu/CNTs composite electrodes at different scan rates in 0.5 M K2CO3 aqueous

electrolyte solution, respectively.

Fig. 5.9a shows the CVs of C/CNTs, copper and the C/Cu/CNTs composite in 0.5 M aqueous K2CO3 electrolyte solution in the range −0.8 < ESCE < 0 V. At the scan rate of 1 mV·s-1 the unsupported copper electrode exhibits two pairs of redox peaks, one rather 69

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

weak, at ESCE = −0.6/−0.2 V and −0.17/−0.1 V, respectively. According to the results of

XRD (Fig. 5.9b), XPS data (Fig. 5.10) and a previous report [125], the associated electrode reactions can be expressed as follows:

(42)

(43)

According to Fig. 5.9b, the XRD pattern of the copper electrode after CV shows weak peaks of Cu2O and strong peaks of Cu. However, no peaks of Cu2(OH)2CO3 are observed, which may be attributed to overlapping with other, perhaps stronger peaks because of its small amount. With the C/Cu/CNTs composite redox peaks are situated at ESCE =

−0.62/−0.27 V and −0.14/−0.09 V, respectively. Peak positions are very close to those observed with the copper electrode implying the same redox processes are occurring. The peak potential difference of the composite is smaller than with the unsupported copper electrode, which implies that the electrode reaction of the composite is faster (i.e. more reversible) than with the copper electrode due to the presence of C/CNTs. The oxidation peak at ESCE = −0.45 V observed for the C/Cu/CNTs composite is ascribed to the oxidation of –O-H or –C = O functional groups on the C/CNTs composite (Fig. 5.5a). The peak can be observed for the C/CNTs composite at the same position.

CVs of the unsupported copper electrode and the C/Cu/CNTs composites at different

-1 scan rates from 1 to 50 mV·s in a 0.5 M K2CO3 aqueous electrolyte solution are shown in Figs. 5.9c and 5.9d. The unsupported copper electrode shows asymmetric redox peaks growing with an increasing scan rate (Fig. 5.9c). Moreover, the reduction peak disappears gradually at dE/dt ≥ = 50 mV·s-1, which suggests that the unsupported copper electrode has poor rate capability due to the poor electronic conductivity of the formed Cu2O [126].

The C/Cu/CNTs composites (Fig. 5.9d) show relatively symmetric CVs even at the high scan rate of 50 mV·s-1. This is mainly due to the high conductivity of C/CNTs composite.

Both electrodes exhibit only one pair of redox peaks in the studied range of scan rates

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

from 5 to 50 mV·s-1. In agreement with the results of impedance measurements discussed below the electrode reaction equation (42) is diffusion-controlled (Fig. 5.11d).

In order to further understand the mechanism of redox reaction of the unsupported copper electrode, XPS results for electrode before and after CVs are shown in Fig. 5.10.

In Fig. 5.10 (a), a strong peak at 292 eV is due to C-F bond, which comes from the polytetrafluoroethylene binder. Other weak satellite features are attributed to acetylene black added as conductive agent. In addition to carbon peaks described above in the range of 281-298 eV, the K 2p1/2 and K 2p3/2 peaks with binding energies of 295.4 and 292.6 eV, respectively, are exhibited in Figure 5.10 (b) [127], which apparently stem from K2CO3 electrolyte. Most importantly, a peak at 289 eV corresponding to C=O bond for cycled electrode is stronger than that of uncycled electrode, which proves the formation of

Cu2(OH)2CO3 species during the charge/discharge process.

Experimental Data Experimental Data Metal Carbide (a) Metal Carbide (b) 2 sp2 Carbon sp Carbon C-C C-C C-O-C C-O-C O-C=O C=O F-C-F F-C-F Background K 2p3/2 Envelope K 2p1/2 Residuals Background

Envelope

Residuals Intensity (arb. units) Intensity (arb. units)

300 295 290 285 280 300 295 290 285 280 Binding energy (eV) Binding energy (eV)

Fig. 5.10 C 1s spectra of the unsupported copper electrode (a) before and (b) after CVs.

5.7 Impedance measurement of C/Cu/CNT composites

The impedance of the unsupported copper and the C/Cu/CNT composite electrode measured after several cycles at open circuit potential is displayed in Fig. 5.11. The measured impedance data were analyzed using the complex nonlinear least squares fitting

(NLSF) method [105, 128] on the basis of the equivalent circuits given in Figs. 5.11c and

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

5.11d. The experimental and simulated impedance data are shown in Figs. 5.11a and

5.11b, values of the elements in the equivalent circuit are collected in Table 3. The intercept at the real part (Z’) axis gives the magnitude of the solution resistance (Rsol) and additional electronic resistances (presumably negligibly small) of wires, current collector and contacts. Rsol for the two electrodes are almost the same. The Faraday impedance contains all processes (i.e. elements in the equivalent circuit) associated with the electrode process. Combined with the double layer capacity (CDL assuming ideal behavior) the interfacial impedance is obtained [129]. For the unsupported copper electrode a simplified circuit was sufficient because in the studied frequency range contributions from diffusion were not observed. The unsupported copper electrode shows Rct = 2241 Ω,

1400 1200 (a) Simulated Data Simulated Data (b) 1200 Experimental Data 1000 Experimental Data

1000 800 800

600

600 -Z'' / Ohm -Z'' / Ohm 400 400

200 200

0 0 0 250 500 750 1000 1250 1500 1750 2000 2250 2500 0 250 500 750 1000 1250 1500 1750 2000 Z' / Ohm Z' / Ohm

C C (c) DL (d) DL

Rsol Rct Rsol Rct Zdiff

Fig. 5.11 (a) and (b) Nyquist plots of experimental data and simulated data, (c) and (d)

equivalent circuit used for fitting impedance spectra for the unsupported copper and the

C/Cu/CNTs composite electrodes after 30 cycles, respectively.

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

Table 3: Results of impedance evaluation

Rsol/Ω Rct/Ω CDL/F Yo(=1/Zdiff)/S Yo(=1/Q2)/S

unsupported Cu 23.9 2240.9 0.00617 - -

Cu/CNTs 24.5 40.7 0.00155 0.01382 0.00711 (n=0.55) (n = 1)

C/Cu/CNTs 20.4 540.2 0.00185 0.01026 -

the C/Cu/CNT composite has a lower value of Rct = 540 Ω implying electrode kinetics are substantially faster. In the absence of further information in particular about the electrochemically active surface area the influence of the beneficial effect of CNTs in particular when poorly conducting Cu2O produced during charging/discharging process is present can only be assumed. This acceleration is also one reason for the good CV behavior of the C/Cu/CNT composite at the high scan rate of 50 mV·s-1. In the intermediate frequency range impedance data of both electrodes are scattered, this may be attributed to the self-discharge behavior causing perturbation and surface-breaking flaws

[130] when metallic copper reacted with alkaline electrolyte solution. At low frequencies, the linear region of the C/Cu/CNT composite electrode impedance with an angle of about

45 ° corresponds to the Warburg diffusion impedance, indicating that the electrode process is diffusion-influenced. This is not observed with the unsupported copper electrode implying that diffusion does not contribute to the electrode impedance in the selected frequency range.

In Fig. 5.12 the Cu/CNTs composites electrode shows Rct = 41 Ω, which imply electrode kinetics are faster than those of the unsupported copper and the C/Cu/CNTs composites electrodes. It also means the Cu/CNTs composites have excellent rate capability, which can be seen already in Fig. 5.14 (b). However, no scattered data are observed in

Fig. 5.12. This is due to the narrowed frequency range of 105-1 Hz.

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

80 (a) 70

-Z'' / Ohm / -Z'' 30

10 Experimental Data Simulated Data 0 20 30 40 50 60 70 80 90 100 Z' / Ohm

(b) Q1 CPE

W CPE

Q Rsol Rct Zdiff 2

Fig. 5.12 (a) Nyquist plots of experimental data and simulated data and (b) equivalent

circuit used for fitting impedance spectra for the Cu/CNTs composite after 30 cycles.

5.8 Galvanostatic charge/discharge measurement of C/Cu/CNT composites Fig. 5.13a shows the galvanostatic charge/discharge curves of the unsupported copper electrode from −0.8 to 0 V (vs. SCE) at various current densities between 0.2 A·g-1 and 1

A·g-1. The charge/discharge curves are highly asymmetric strongly suggesting substantial faradaic contributions and redox or battery electrode behavior. At lower current densities one well-defined charge-discharge voltage plateau around −0.2/−0.6 V can be identified, which is consistent with the CV results (Fig. 5.9a). When the charge/discharge current increases from 0.2, 0.4, 0.6, 0.8 to 1.0 A·g-1, the discharge time decreases from 670, 106,

30, 13 to 7 s, respectively. For the C/Cu/CNT composite (Fig. 5.13b), the shape of curves

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

is similar to that found with the copper electrode. However, the discharge times are longer than those of the copper electrode at higher current densities except for 0.2 A·g-1. This means that the specific discharge capacity of the composite is smaller than that of the unsupported copper electrode at the lower current density, which can be seen in Fig. 5.13c.

This is mainly due to the substantial fraction of carbon (49.1 wt%) in the composite (Fig.

5.6) not providing faradaic contributions but only double layer capacitance (see Fig. 5.9a).

Moreover, it can be seen clearly from Fig. 5.13c that the composite has excellent rate behavior superior to unsupported copper. Obviously, this good rate behavior is due to the existence of CNTs and the carbon coating, which is consistent with the CVs discussed above. In addition, its rate capability is also superior to that of the previously reported copper-infiltrated carbonized wood monoliths in KOH aqueous electrolyte solution [111].

The main reason is an accelerating effect of carbonate ions on the metal electrode reaction rate in K2CO3 solution [131].

0.0 (a) 0.0 (b)

-0.1 -0.1 0.2 A/g 0.2 A/g 0.4 A/g 0.4 A/g -0.2 -0.2 0.6 A/g 0.6 A/g 0.8 A/g 0.8 A/g -0.3 -0.3 1.0 A/g

1.0 A/g 1.5 A/g

2.0 A/g -0.4 -0.4 -0.5 -0.5

Potential / V vs. SCE -0.6 Potential / V vs. SCE -0.6

-0.7 -0.7

-0.8 -0.8 0 100 200 300 400 500 600 700 800 900 1000 1100 1200 0 100 200 300 400 500 600 700 800 900 Time / s Time / s

140 Unsupported Cu (c) 120 C@Cu@CNT

-1 100

80 0.2 0.4 0.6 0.8 1.0 1.5 2.0 -1 Rate : A g 60

Specific Capacity / C g 20

0 0 3 6 9 12 15 18 21 Cycle Number

Fig. 5.13 (a) and (b) Charge–discharge curves and (c) rate capability of the unsupported

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

copper and the C/Cu/CNTs composite at different current densities.

From Fig. 5.14 (a), it can be seen clearly that the Cu/CNTs composites electrode shows two pairs of redox peaks, similar to observations with both unsupported copper and the C/Cu/CNTs composites electrodes. However, current response of the couple of redox peaks at more positive electrode potential for the Cu/CNTs composites electrode is stronger than that of unsupported copper and the C/Cu/CNTs composites, which can be attributed to the fluffy structure with substantial mesopores and high specific surface area promoting sufficient contact between electrolyte and the Cu/CNTs composites. The galvanostatic charge/discharge curves of the Cu/CNTs composites in Fig. 5.14 (b) show two well-defined charge-discharge voltage plateaus at the current density of 0.2 A·g-1, which is consistent with the CV results. When the charge/discharge current increases from 0.2, 0.4, 0.6, 0.8, 1.0, 1.5 to 2.0 A·g-1, the discharge time decreases from 428, 190,

113, 88, 57, 11 to 9 s, respectively. All these results are better than those with the other materials at higher current densities. This can be ascribed to the higher conductivity of the

Cu/CNTs composites. This also can be proved by impedance results shown in Fig. 5.11 and Fig. 5.12.

0.5 (a) 0.0 (b) 0.4 -0.1

-1 0.3 1 mV / s -0.2 0.2 -0.3 0.1

Cu@CNTs 0.2 A/g

0.0 -0.4 0.4 A/g 0.6 A/g -0.1 -0.5 0.8 A/g 1.0 A/g -0.2

Potential / V vs. SCE -0.6 1.5 A/g Current Density / A g -0.3 2.0 A/g -0.7 -0.4 -0.8 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0 100 200 300 400 500 600 700 800 900 1000 1100 Potential / V vs. SCE Time / s

Fig. 5.14 (a) CVs and (b) Charge-discharge curves of the Cu/CNTs composites in 0.5 M

K2CO3 solution.

5.9 Cycling performance of C/Cu/CNT composites

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

100 100 90 90

80 80 -1 70 70

60 60

50 C@Cu@CNT 50

40 C@CNT 40 Unsupported Cu 30 30

20 20 Coulombic Efficiency / % Specific Capacity / C g C / Capacity Specific 10 10

0 0 0 100 200 300 400 500 600 700 800 900 1000 1100 1200 Cycle Number

Fig. 5.15 Cycling behavior of the C/CNTs, the unsupported copper and the C/Cu/CNTs

composite electrodes at 700 mA·g-1.

The cycling behavior of the C/Cu/CNT composite at 700 mA·g-1 (based on the weight of the active material in the electrode) is shown in Fig. 5.15. The composite evidently exhibits improved cycling behavior compared with unsupported copper. The specific capacity of the C/Cu/CNT composite reaches a highest value of 66 C·g-1, higher than that of unsupported copper. This can be attributed to the synergistic effects between metallic copper, CNTs and the carbon coating. After 1200 cycles, when compared to the first few cycles, there is no evident capacity loss. However, the values of specific capacity show some fluctuation. Similar results were also found for the redox reaction on the metal surface [132, 133]. In contrast, the copper electrode exhibits a rapid capacity fading in the aqueous K2CO3 solution. After 50 cycles, the capacity retention is about 30 %. It is believed that copper ions dissolve into the electrolyte solution during the charge/discharge process possibly yielding carbonate deposits not participating in the electrode reaction anymore [125, 134]. The coating apparently inhibits dissolution of copper carbonate complexes and also buffers possible volume changes during the cycling process [135,

136]. Consequently, better cycling behavior is achieved for the composites of CNT/Cu coated with amorphous carbon. The carbon-coated CNTs composite without copper (see

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

Fig. 5.15) is very stable and presents good cycling stability although with lower capacity

(Fig. 5.15).

The cycling behavior of the Cu/CNT composite is worse than that of the C/Cu/CNT composite in Fig. 5.16. The Cu/CNT composite electrode exhibits a rapid capacity fading in the aqueous K2CO3 solution, which is similar to that observed wih the unsupported copper electrode. After 50 cycles, the capacity retention of the latter is about 32 %. The main reason is that the Cu/CNT composite electrode is exposed to the alkaline electrolyte solution directly; promoting copper carbonate complexes dissolve into the electrolyte solution during the charge/discharge process.

120 120

110 110

100 100

-1 90 90

80 80

70 Cu/CNTs 70 60 60

Coulombic efficiency 50 50

40 40

30 30

20 20 Coulombic efficiency / % Specific Capacity / C g C / Capacity Specific 10 10

0 0 0 100 200 300 400 500 600 700 800 900 1000 1100 1200 Cycle Number

Fig. 5.16 Cycling behavior of the Cu/CNTs composite electrode at 700 mA·g-1.

5.10 Conclusions A composite of copper nanograins grown on the surface of CNTs by a redox reaction between Cu(Ac)2 and EG was prepared. The composite material consists of metallic copper, CNTs and carbon coating. The as-prepared C/Cu/CNTs composite exhibits higher capacity and better rate behavior as well as excellent cycling stability in aqueous 0.5 M

K2CO3 solution compared to the unsupported copper negative. The carbon coating can effectively prevent the dissolution of copper carbonate complexes, increase the electrode conductivity, improve the surface chemistry of the active material and protect the

Chapter 5: Improved electrochemical behavior of amorphous carbon-coated copper/CNT composites as negative electrode and their energy storage mechanism

electrode from direct contact with electrolyte solution.

Chapter 6: Summary and Outlook

Chapter 6 Summary and Outlook

6.1 Summary

In this thesis, the first chapter briefly introduces some characteristics, energy storage mechanism, electrochemical performance and history of various rechargeable batteries, including conventional lithium ion batteries, ARLIBs, Ni-Cd battery and lead-acid battery.

As described in introduction part, aqueous rechargeable batteries have some advantages, which are low cost, easy construction and inherent safety by avoiding flammable organic electrolyte solutions. Moreover, the ionic conductivity of aqueous electrolytes is greater than that of organic electrolytes by about two orders of magnitude, which makes sure high rate capability of aqueous rechargeable battery. However, strong acid or alkaline, which is used as the electrolyte for rechargeable battery, will cause serious corrosion to equipment.

In addition, the active electrode materials of Ni-Cd battery and lead-acid battery contain highly toxic heavy metals such as Cd and Pb. Therefore, combining the features of neutral aqueous electrolyte (or pH value of electrolyte solution close to 7 such as weakly alkaline and acid) and environmentally friendly electrode materials, this thesis mainly studied the electrochemical performance of PbSO4/0.5 M Li2SO4/LiMn2O4 full battery, Cd/0.5 M

Li2SO4+10mM Cd(Ac)2/LiCoO2 full battery and C/Cu/CNT composites as negative material in 0.5 M K2CO3 electrolyte for half-cell. The associated conclusions are collected as follows:

(1) In terms of PbSO4/0.5M LiSO4/LiMn2O4 full battery, 50 % of Pb can be saved compared to lead-acid battery. Moreover, sulphuric acid is not needed. The experimental results show that this battery system has 100 % Coulombic efficiency except for the initial several cycles and less than 1 % capacity loss after 110 full cycles at the current density of 400 mA·g-1 between 0 and 1.8 V. In addition, the calculated energy density

Chapter 6: Summary and Outlook

based on the practical capacities of the two electrodes is 68 W·h·kg-1.

(2) In the case of Cd/0.5 M Li2SO4+10 mM Cd(Ac)2/LiCoO2 full battery, 100 % of

Hg can be replaced by LiCoO2 in Weston cell. What’s more, alkaline electrolyte is not necessary in this new battery system if compared with Ni-Cd batteries. The electrochemical performance is worse than that of PbSO4//LiMn2O4 full battery. After 100 cycles, the capacity retention is 75.8 % compared to the first cycles between 0.4 and 1.8

V. Additionally, the calculated energy density is 72 W·h·kg-1, which is similar to that of

PbSO4//LiMn2O4 full battery.

(3) C/Cu/CNT composites as negative electrode in 0.5M K2CO3 electrolyte for half- cell exhibit higher capacity and better rate behavior as well as excellent cycling stability in aqueous electrolyte compared to the unsupported copper negative. However, the discharge capacity is lower than that of other two full batteries, even though copper is much more environmentally friendly.

(4) The carbon coating can effectively prevent the dissolution of copper carbonate complexes, increase the electrode conductivity, improve the surface chemistry of the active material and protect the electrode from direct contact with electrolyte solution.

6.2 Outlook

In the past decades, the commercial application of aqueous rechargeable batteries, including lead-acid battery and Ni-Cd battery, has made great progress. However, challenges still exist in present technology conditions. Specifically, the main problem of aqueous rechargeable batteries is lower energy density compared to organic electrolyte- based lithium batteries, which seriously restricts their application in large-scale energy storage. In order to solve this problem, most research in recent years focused on novel electrode materials with high capacity and broad potential window. Moreover, solid state

Chapter 6: Summary and Outlook

electrolyte as separator can further enhance energy density of ARLIBs, which also provide a new direction for exploring aqueous rechargeable batteries with excellent electrochemical performance.

It is worth noting that Li-S battery and Li-air battery in hybrid liquid electrolyte would be a promising candidate for large-scale energy storage due to their high theoretical energy density. Moreover, probing of new and high efficiency catalyst for oxygen reduction/evolution reactions (ORR/OER) is also of importance in the next decades. In addition, flexible and wearable aqueous rechargeable batteries will attract much more attention in future.

As for the research work in this thesis, some good results have been obtained.

However, few problems should be further researched in the future:

(1) The cycling performance of PbSO4//LiMn2O4 full battery and Cd//LiCoO2 full battery should be improved;

(2) Commercial materials for PbSO4 and cadmium should be replaced by nanomaterials to get much better electrochemical performance;

(3) Optimize the amount of carbon coating to find the best coating thickness, in order to get higher capacity and better rate capability;

(4) Search for suitable positive materials in K2CO3 electrolyte to assemble a full battery with C/Cu/CNT composites.

References

1 Y.P. Wu, X.B. Dai, J.Q. Ma and Y.J. Cheng, Lithium Ion Batteries: Practice and

Applications, Chemical Industry Press, Beijing, 2004.

2 W. Li, J.R. Dahn and D.S. Wainwright, Science 264 (1994) 1115.

3 W. Tang, Y.S. Zhu, Y.Y. Hou, L.L. Liu, Y.P. Wu, K.P. Loh, H.P. Zhang and K. Zhu,

Energy Environ. Sci. 6 (2013) 2093.

4 F.X. Wang, S.Y. Xiao, Y.S. Zhu, Z. Chang, C.L. Hu, Y.P. Wu and R. Holze, J. Power

Sources 246 (2014) 19.

5 Y. Liu, A. Wiek, V. Dzhagan and R. Holze, J. Electrochem. Soc. 163 (2016) A1247.

6 M. Xu, D.G. Ivey, Z. Xie and W. Qu, J. Power Sources 283 (2015) 358.

7 S. Li, C. Cheng and A. Thomas, Adv. Mater. 29 (2017) 1602547.

8 X.W. Wang, F.X. Wang, L. Wang, M. Li, Y. Wang, B. Chen, Y. Zhu, L. Fu, L. Zha, L.

Zhang, Y.P. Wu and W. Huang, Adv. Mater. 28 (2016) 4904.

9 N.S. Choi, Z. Chen, S.A. Freunberger, X. Ji, Y.K. Sun, K. Amine, G. Yushin, L.F.

Nazar, J. Cho and P.G. Bruce, Angew. Chem. Int. Ed. 51 (2012) 9994.

10 X.W. Wang, M.X. Li, Y.F. Wang, B.W. Chen, Y.S. Zhu and Y.P. Wu, J. Mater. Chem.

A 3 (2015) 8280.

11 F.X. Wang, Y. Liu, X.W. Wang, Z. Chang, Y.P. Wu and R. Holze, ChemElectroChem

2 (2015) 1024.

12 F.X. Wang, F. Yu, X. Wang, Z. Chang, L.J. Fu, Y.S. Zhu, Z.B. Wen, Y.P. Wu and W.

Huang, ACS Appl. Mater. Interfaces 8 (2016) 9022.

13 R.J. Chen, W.J. Qu, X. Guo, L. Li and F. Wu, Mater. Horiz. 3 (2016) 487.

14 X.L. Li, J. Huang and A. Faghri, J. Power Sources 332 (2016) 420.

15 Y. Shi, J. Zhang, L.J. Pan, Y. Shi and G.H. Yu, Nanotoday 11 (2016) 738.

16 A. Eftekhari, Y. Liu and P. Chen, J. Power Sources 334 (2016) 221.

References

17 Y. Wang and W.H. Zhong, ChemElectroChem 2 (2015) 22.

18 N. Alias and A.A. Mohamad, J. Power Sources 274 (2015) 237.

19 X.J. Wang, Q.T. Qu, Y.Y. Hou, F.X. Wang and Y.P. Wu, Chem. Commun. 49 (2013)

6179.

20 J.M.D. Iiarduya, L. Otaegui, J.M.L.D. Amo, M. Armand and G. Singh, J. Power

Sources 337 (2017) 197.

21 Z. Chang, Y. Yang, X.W. Wang, M.X. Li, Z.W. Fu, Y.P. Wu and R. Holze, Sci. Rep. 5

(2015) 11931.

22 M.C. Lin, M. Gong, B. Lu, Y.P. Wu, D.Y. Wang, M.Y. Guan, M. Angell, C.X. Chen,

J. Yang, B.J. Hwang and H.J. Dai, Nature 520 (2015) 324.

23 D. Lei, D.C. Lee, A. Magasinski, E. Zhao, D. Steingart and G. Yushin, ACS Appl.

Mater. Interfaces 8 (2016) 2088.

24 J.X. Zhang, J.X. Yu and C. Cha, Int. J. Electrochem. Sci. 7 (2012) 10233.

25 K. Maeda, N. Arakawa, K. Fukui and H. Kuramochi, Phys. Chem. Chem. Phys. 16

(2014) 4911.

26 L.J. Fu, H. Liu, C. Li, Y.P. Wu, E. Rahm, R. Holze, H.Q. Wu, Solid State Sci. 8 (2006)

113.

27 A. Manthiram, A.V. Murugan, A. Sarkar and T. Muraliganth, Energy Environ. Sci. 1

(2008) 621.

28 G.J. Wang, L.C. Yang, Q.T. Qu, B.Wang, Y.P. Wu and R. Holze, J. Solid State

Electrochem. 14 (2010) 865.

29 W. Tang, L.L. Liu, Y.S. Zhu, H. Sun, Y.P. Wu and K. Zhu, Energy Environ. Sci. 5

(2012) 6909.

30 X.J. Wang, Y.Y. Hou, Y.S. Zhu, Y.P. Wu and R. Holze, Sci. Rep. 3 (2013) 1401.

31 Y. Zhao, L. Wang and H.R. Byon, Nature Commun. 4 (2013) 1896.

References

32 Z. Chang, X.J. Wang, Y.Q. Yang, J. Gao, M.X. Li, L.L. Liu and Y.P. Wu, J. Mater.

Chem. A 2 (2014) 19444.

33 L. Carbone and J. Hassoun, Ionics 22 (2016) 2341.

34 K. Cai, R. Yang, X.S. Lang, Q. Zhang, Z. Wang and T.S. He, Sci. Rep. 6 (2016) 34950.

35 X.L. Dong, Y.G. Wang and Y.Y. Xia, Sci. Rep. 4 (2014) 6916.

36 D. Guyomard and J.M. Tarascon, Adv. Mater. 6 (1994) 408.

37 J. Garche, Phys. Chem. Chem. Phys. 3 (2001) 356.

38 N. Opiyo, J. Energy Storage 7 (2016) 1.

39 M.S. Whittingham, Science 192 (1976) 1126.

40 P. Rozier and J.M. Tarascon, J. Electrochem. Soc. 162 (2015) A2490.

41 Z. Chang, Y.Q. Yang, M.X. Li, X.W. Wang and Y.P. Wu, J. Mater. Chem. A 2 (2014)

10739.

42 Y.Y. Hou, X.J. Wang, Y.S. Zhu, C.L. Hu, Z. Chang, Y.P. Wu and R. Holze, J. Mater.

Chem. A 1 (2013) 14713.

43 Y.H. Lu, J.B. Goodenough and Y. Kim, J. Am. Chem. Soc. 133 (2011) 5756.

44 Y.P. Wu, X.Y. Yuan, C. Dong and J.Y. Duan, Lithium ion batteries: Applications and

Practice (2nd Edition), Chemical Industry Press, Beijing, 2012.

45 D. Herbert and J. Ulam, US Patent 3043896 (1962).

46 W. Kang, N. Deng, J. Ju, Q. Li, D. Wu, X. Ma, L. Li, M. Naebe and B.W. Cheng,

Nanoscale 8 (2016) 16541.

47 Z.W. Seh, Y.M. Sun, Q. Zhang and Y. Cui, Chem. Soc. Rev. 45 (2016) 5605.

48 M.N. Liu, F. Ye, W.F. Li, H.F. Li and Y.G. Zhang, Nano Res. 9 (2016) 94.

49 B. Yan, X. Li, Z. Bai, X. Song, D. Xiong, M. Zhao, D. Li and S.G. Lu, J. Power

Sources 338 (2017) 34.

50 R. Cao, W. Xu, D. Lv, J. Xiao and J.G. Zhang, Adv. Energy Mater. 5 (2015) 1402273.

References

51 A. Manthiram, Y. Fu, S.H. Chung, C. Zu and Y.S. Su, Chem. Rev. 114 (2014) 11751.

52 R. Chen, T. Zhao and F. Wu, Chem. Commun. 51 (2015) 18.

53 X. Fang and H. Peng, Small 11 (2015) 1488.

54 A. Rosenman, E. Markevich, G. Salitra, D. Aurbach, A. Garsuch and F.F. Chesneau,

Adv. Energy Mater. 5 (2015) 1500212.

55 X.L. Li, J. Huang and A. Faghri, J. Power Sources 332 (2016) 420.

56 T. Zhang, N. Imanishi, Y. Takeda and O. Yamamoto, Chem. Lett. 40 (2011) 668.

57 E.L. Littauer and K.C. Tsai, J. Electrochem. Soc. 123 (1976) 771.

58 K.M. Abraham and Z. Jiang, J. Electrochem. Soc. 143 (1996) 1.

59 M. Zhong, X.X. Yuan, L. Li, Z.F. Ma, D.P. Wilkinson, L. Zhang and J.J. Zhang,

Energy Environ. Sci. 8 (2015) 2144.

60 J. Lu, L. Li, J.B. Park, Y.K. Sun, F. Wu and K. Amine, Chem. Rev. 114 (2014) 5611.

61 P.G. Bruce, S.A. Freunberger, L.J. Hardwick and J.M. Tarascon, Nature Mater. 11

(2012) 19.

62 A. Kraytsberg and Y. Ein-Eli, J. Power Sources 196 (2011) 886.

63 A. Boulabiar, K. Bouraoui, M. Chastrette and M. Abderrabba, J. Chem. Educ. 81

(2004) 754.

64 G.F. Martins, J. Chem. Educ. 67 (1990) 482.

65 H.P. Zhang, T. Yang, X. Wu, Y. Zhou, C. Yang, T. Zhu and R. Dong, Chem. Commun.

51 (2015) 7294.

66 M. Bartolozzi, J. Power Sources 27 (1989) 219.

67 L. Sanz, D. Lloyd, E. Magdalena, J. Palma, M. Anderson and K. Kontturi, J. Power

Sources 278 (2015) 175.

68 L. Sanz, D. Lloyd, E. Magdalena, J. Palma and K. Kontturi, J. Power Sources 268

(2014) 121.

References

69 E. Weston, Voltaic cell, US Patent 494827, Apr. 4, 1893.

70 A. Fernandes, J.C. Afonso and A.J.B. Dutra, J. Power Sources 220 (2012) 286.

71 C.A. Nogueira and F. Margarido, Waste Management 27 (2007) 1570.

72 C. Hazotte, N. Leclerc, S. Diliberto, E. Meux and F. Lapicque, Environ. Technol. 36

(2015) 796.

73 K. Huang, J. Li and Z.M. Xu, Environ. Sci. Technol. 43 (2009) 8974.

74 J.Q. Pan, M. Yang, X. Jia and Y.Z. Sun, J. Electrochem. Soc. 160 (2013) A1146.

75 Y.K. Zeng, T.S. Zhao, X.L. Zhou, L. Wei and H.R. Jiang, J. Power Sources 330 (2016)

55.

76 T. Chen, H. Huang, H.Y. Ma and D.L. Kong, Electrochim. Acta 88 (2013) 79.

77 Y. Gong, P. Zhang, P.G. Jiang and J.H. Lin, CrystEngComm. 17 (2015) 1056.

78 H. Karami and M. Ghamooshi-Ramandi, Int. J. Electrochem. Sci. 8 (2013) 7553.

79 D.N. Yang, J.W. Liu, Q. Wang, X.Q. Yuan, X.F. Zhu, L. Li, W. Zhang, Y.C. Hu, X.J.

Sun, R.V. Kumar and J.K. Yang, J. Power Sources 257 (2014) 27.

80 A. Czerwinski, Patent RP, No. 167796.

81 A. Czerwinski, Z. Rogulski, S. Obrebowski, H. Siwek, I. Paleska, M. Chotkowski and

M. Lukaszewski, J. Appl. Electrochem. 39 (2009) 559.

82 B.H. Zhang, Y. Liu, X.W. Wu, Y.Q. Yang, Z. Chang, Z.B. Wen and Y.P. Wu, Chem.

Commun. 50 (2014) 1209.

83 W. Tang, S. Tian, L.L. Liu, L. Li, H.P. Zhang, Y.B. Yue, Y. Bai, Y.P. Wu and K. Zhu,

Electrochem. Commun. 13 (2011) 205.

84 W.J. McG. Tegart, The Electrolytic and Chemical Polishing of Metals in Research and

Industry, Pergamon Press, London (1959).

85 M. Armand and J.M. Tarascon, Nature 451 (2008) 652.

86 M. Saravanan, M. Ganesan and S. Ambalavanan, J. Power Sources, 251 (2014) 20.

References

87 B. Hong, L.X. Jiang, K.T. Hao, F.Y. Liu, X.Y. Yu, H.T. Xue, J. Li and Y.X. Liu, J.

Power Sources, 256 (2014) 294.

88 K. Maeda, N. Arakawa, K. Fukui and H. Kuramochi, Phys. Chem. Chem. Phys. 16

(2014) 4911.

89 A. Banerjee, M.K. Ravikumar, A. Jalajakshi, S.A. Gaffoor and A.K. Shukla, in Lead-

acid Batteries and Capacitors, New Designs, And New Applications, ed. K. Bullock,

P.T. Moseley and B.Y. Liaw, ECS Transactions, 2012, vol. 41(13), pp. 101-113.

90 L.T. Lam and R. Louey, J. Power Sources 158 (2006) 1140.

91 Q.T. Qu, L.J. Fu, X.Y. Zhan, D. Samuelis, J. Maier, L. Li, S. Tian, Z.H. Li and Y.P.

Wu, Energy Environ. Sci. 4 (2011) 3985.

92 P.R. Gao, W.X. Lv, R. Zhang, Y. Liu, G.H. Li, X.F. Bu and L.X. Lei, J. Power

Sources 248 (2014) 363.

93 W. Tang, Y.Y. Hou, F.X. Wang, L.L. Liu, Y.P. Wu and K. Zhu, Nano Lett. 13 (2013)

2036.

94 W. Tang, Y.S. Zhu, Y.Y. Hou, L.L. Liu, Y.P. Wu, K.P. Loh, H.P. Zhang and K. Zhu,

Energy Environ. Sci. 6 (2013) 2093.

95 H.P. Zhang, X. Wu, T. Yang, S.S. Liang and X.J. Yang, Chem. Commun. 49 (2013)

9977.

96 J. Yan, J. Wang, H. Liu, Z. Bakenov, D. Gosselink and P. Chen, J. Power Sources 216

(2012) 222.

97 L. Zhang, D.Y. Zhao and X.W. Lou, Angew. Chem. Int. Ed. 51 (2012) 239.

98 F.X. Wang, S.Y. Xiao, Z. Chang, Y.Q. Yang and Y.P. Wu, Chem. Commun. 49 (2013)

9209.

99 M.P. Vinod and K. Vijayamohanan, J. Power Sources 50 (1994) 67.

References

100 P. Gao, Y. Wang, Q. Zhang, Y. Chen, D. Bao, L.Q. Wang, Y.Z. Sun, G.B. Li and

M.L. Zhang, J. Mater. Chem. 22 (2012) 13922.

101 X.L. Xiao, L.M. Yang, H. Zhao, Z.B. Hu and Y.D. Li, Nano Res. 5 (2012) 27.

102 G.J. Wang, Q.T. Qu, B. Wang, Y. Shi, S. Tian, Y.P. Wu and R. Holze, Electrochim.

Acta 54 (2009) 1199.

103 P. Andrei, J.P. Zheng, M. Hendrickson and E.J. Plichtac, J. Electrochem. Soc. 159

(2012) A770.

104 M.J. Trahan, S. Mukerjee, E.J. Plichta, M.A. Hendrickson and K.M. Abrahama, J.

Electrochem. Soc. 160 (2013) A259.

105 J. Yan, Z.J. Fan, T. Wei, W.Z. Qian, M.L. Zhang and F. Wei, Carbon 48 (2010) 3825.

106 U. Palmqvist and R. Sjovall, J. Power Sources 79 (1999) 212.

107 S. Rodrigues, N. Munichandraiah and A.K. Shukla, J. Power Sources 87 (2000) 12.

108 M.M. Burashnikova, I.A. Kazarinov and I.V. Zotova, J. Power Sources 207 (2012)

19.

109 R. Ruffo, C. Wessells, R.A. Huggins and Y. Cui, Electrochem. Commun. 11 (2009)

247.

110 M. Okubo, E. Hosono, J. Kim, M. Enomoto, N. Kojima, T. Kudo, H.S. Zhou and I.

Honma, J. Am. Chem. Soc. 129 (2007) 7444.

111 S. Teng, G. Siegel, M.C. Prestgard, W. Wang and A. Tiwari, Electrochim. Acta 161

(2015) 343.

112 S.O. Kim, H.S. Kim and J.K. Lee, Mater. Chem. Phys. 133 (2012) 38.

113 G. Yuan, Y. Liu, M. Yue, H.J. Li, E. Liu, Y.Y. Huang and D.L. Kong, Ceram. Inter.

40 (2014) 9101.

114 M.M. Momeni, Z. Nazari, A. Kazempour, M. Hakimiyan and S.M. Mirhoseini, Surf.

Engin. 30 (2014) 775.

References

115 L.B. Hu, W. Chen, X. Xie, N.A. Liu, Y. Yang, H. Wu, Y. Yao, M. Pasta, H.N.

Alshareef and Y. Cui, ACS Nano 5 (2011) 8904.

116 H.J. Huang and X. Wang, Nanoscale 3 (2011) 3185.

117 E. Raymundo-Pinero, V. Khomenko, E. Frackowiak and F. Béguin, J. Electrochem.

Soc. 152 (2005) A229.

118 H.L. Wang, H.S. Casalongue, Y.Y. Liang and H.J. Dai, J. Am. Chem. Soc. 132 (2010)

7472.

119 K.R. Prasad and N. Miura, J. Power Sources 135 (2004) 354.

120 G.M. Ilari, N. Kranzlin, R. Longtin, J.R. Sanchez-Valencia, S. Schneider, M.D.

Rossell, M. Niederberger and R. Erni, Carbon 73 (2014) 146.

121 J. Cheng, B. Wang, C.M. Park, Y.P. Wu, H. Huang and F. Nie, Chem. Eur. J. 19

(2013) 9866.

122 H.H. Huo, C.Y. Guo, G.L. Li, H. Xu and C.L. Xu, RSC Adv. 4 (2012) 20459.

123 L.J. Ci, B.Q. Wei, C.L. Xu, J. Liang, D.H. Wu, S.S. Xie, W.Y. Zhou, Y.B. Li, Z.Q.

Liu and D.S. Tang, J. Cryst. Growth 233 (2001) 823.

124 M.C. Biesinger, L.W.M. Lau, A.R. Gerson and R.St.C. Smart, Appl. Surf. Sci. 257

(2010) 887.

125 M. Drogowska, L. Brossar and H. Menard, J. Electrochem. Soc. 139 (1992) 39.

126 T.J. Richardson, J.L. Slack and M.D. Rubin, Electrochim. Acta 46 (2001) 2281.

127 A. V. Naumkin, A. Kraut-Vass, S.W. Gaarenstroom and C.J. Powell, NIST X-ray

Photoelectron Spectroscopy Database, NIST.

128 W. Sugimoto, H. Iwata, K. Yokoshima, Y. Murakami and Y. Takasu, J. Phys. Chem.

B 109 (2005) 7330.

129 R. Holze and W. Vielstich, Electrochim. Acta 29 (1984) 607.

130 N. Harfield, Y. Yoshida and J.R. Bowler, J. Appl. Phys. 80 (1996) 4090.

References

131 Y. Sato, H. Niki and T. Takamura, J. Electrochem. Soc. 118 (1971) 1269.

132 Y. Liu, Z.B. Wen, X. Wu, X.W. Wang, Y.P. Wu and R. Holze, Chem. Commun. 50

(2014) 13714.

133 L.L. Liu, Y.Y. Hou, X.W. Wu, S.Y. Xiao, Z. Chang, Y.Q. Yang and Y.P. Wu, Chem.

Commun. 49 (2013) 11515.

134 Y. Feng, K.S. Siow, W.K. Teo, K.L. Tan and A.K. Hsieh, Corrosion 53 (1997) 389.

135 W.M. Zhang, J.S. Hu, Y.G. Guo, S.F. Zheng, L.S. Zhong, W.G. Song and L.J. Wan,

Adv. Mater. 20 (2008) 1160.

136 H.Q. Li and H.S. Zhou, Chem. Commun. 48 (2012) 1201.

Selbständigkeitserklärung

Hiermit erkläre ich an Eides statt, die vorliegende Arbeit selbständig und ohne unerlaubte

Hilfsmittel durchgeführt zu haben.

Chemnitz, den 10.05.2017 (Yu Liu)

Curriculum Vitae

Personal Information

Name: Yu Liu

Date of Birth: May 6th, 1988

Place of Birth: Jiangsu, China

Nationality: Chinese

Current Address: Vettersstrasse 72/402

09126 Chemnitz, Germany

E-mail: [email protected]

[email protected]

Education

2014-To date: Ph.D. student, Physical Chemistry/Electrochemistry,

TU Chemnitz, Germany. Supervisor: Prof. Dr. Rudolf Holze

2011-2014: M.Sc. (Physical Chemistry)

Jiangxi Normal University, Jiangxi, China

2006-2010: B.Eng. (Applied Chemistry)

College of Xinglin, Nantong University, Jiangsu, China

Publications

1. Y. Liu, A. Wiek, V. Dzhagan and R. Holze. Improved electrochemical behavior of

amorphous carbon-coated copper/CNT composites as negative electrode material

Curriculum Vitae

and their energy storage mechanism. J. Electrochem. Soc. 163 (2016) A1247.

2. Y. Liu, Z.B. Wen, X. Wu, X. Wang, Y.P. Wu and R. Holze. An acid-free

rechargeable battery based on PbSO4 and spinel LiMn2O4. Chem. Commun. 50

(2014) 13714.

3. Y. Liu, S.N. Gao and R. Holze. An aqueous lithium ion rechargeable battery with

high rate capability based on metallic Cd and LiCoO2. (unsubmitted).

4. F.X. Wang, Y. Liu, X.W. Wang, Z. Chang, Y.P. Wu and R. Holze. Aqueous

rechargeable battery based on zinc and a composite of LiNi1/3Co1/3Mn1/3O2.

ChemElectroChem. 2 (2015) 1024.

5. Y. Liu, B.H. Zhang, S.Y. Xiao, L.L. Liu, Z.B. Wen and Y.P.Wu. A nanocomposite

of MoO3 coated with PPy as an anode material for aqueous sodium rechargeable

batteries with excellent electrochemical performance. Electrochim. Acta. 116 (2014)

512.

6. Y. Liu, B.H. Zhang, F.X. Wang, Z.B. Wen and Y.P. Wu. Nanostructured

intercalation compounds as cathode materials for supercapacitors. Pure Appl. Chem.

86 (2014) 593.

7. Y. Liu, B.H. Zhang, Y.Q. Yang, Z. Chang, Z.B. Wen and Y.P. Wu. Polypyrrole-

coated -MoO3 nanobelts with good electrochemical performance as anode

materials for aqueous supercapacitors. J. Mater. Chem. A 1 (2013) 13582.

8. B.H. Zhang, Y. Liu, X.W. Wu, Y.Q. Yang, Z. Chang, Z.B. Wen and Y.P. Wu. An

aqueous rechargeable battery based on zinc anode and Na0.95MnO2. Chem.

Commun. 50 (2014) 1209.

9. B.H. Zhang, Y. Liu, Z. Chang, Y.Q. Yang, Z.B. Wen, Y.P. Wu and R. Holze.

Nanowire Na0.35MnO2 from a hydrothermal method as a cathode material for

aqueous asymmetric supercapacitors. J. Power Sources 253 (2014) 98.

Curriculum Vitae

10. B.H. Zhang, Y. Liu, Y.Q. Yang, Z. Chang, Z.B. Wen and Y.P. Wu. Nanowire

K0.19MnO2 from hydrothermal method as cathode material for aqueous

supercapacitors of high energy density. Electrochim. Acta 130 (2014) 693.