Wright State University

CORE Scholar
Browse all Theses and Dissertations Theses and Dissertations

2013

Carbon-Based Nanostructured Materials as

Electrode in Lithium-Ion Batteries and
Supercapacitors
Zhuo Yao
Wright State University

Follow this and additional works at: http://corescholar.libraries.wright.edu/etd_all

Part of the Engineering Science and Materials Commons

Repository Citation
Yao, Zhuo, "Carbon-Based Nanostructured Materials as Electrode in Lithium-Ion Batteries and Supercapacitors" (2013). Browse all
Theses and Dissertations. Paper 792.

This Thesis is brought to you for free and open access by the Theses and Dissertations at CORE Scholar. It has been accepted for inclusion in Browse all
Theses and Dissertations by an authorized administrator of CORE Scholar. For more information, please contact
[email protected].
Carbon-based Nanostructured
Materials as Electrodes in
Lithium-Ion Batteries and
Supercapacitors

A thesis submitted in partial fulfillment

of the requirements for the degree of
Master of Science in Engineering

By

Zhuo Yao
B.E., Dalian Jiaotong University, 2010

2013
Wright State University
 WRIGHT STATE UNIVERSITY

GRADUATE SCHOOL

April 30,2013

I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY

SUPERVISION BY Zhuo Yao ENTITLED Carbon-Based Nanostructured Materials as
Electrodes in Lithium-ion Batteries and Supercapacitors BE ACCEPTED IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF Master of Science
in Engineering.

Hong Huang, Ph.D.

Thesis Director

George Huang, Ph.D., Chair

Committee on
Final Examination Department of Mechanical and Materials Engineering
College of Engineering and Computer Science

Hong Huang, Ph.D.

Dr. Dan Young, Ph.D.

Dr. Sharmila M. Mukhopadhyay,

Ph.D.

R. William Ayres, Ph.D.

Interim Dean, Graduate School
 ABSTRACT

Yao, Zhuo. M.S.Egr, Department of Mechanical and Materials Engineering, Wright

State University, 2013

“Carbon-based Nanostructured Materials as Electrodes in Lithium-ion Batteries and

Supercapacitors”

Carbon-based nancomposites attract much attention as electrode materials in

electrochemical energy storage systems due to their low-cost, extraordinary high

electrical and thermal conductivity, super high surface area etc. In this research,

graphene nanosheets (GNS) and their nanocomposites with manganese oxides

(GNS/MnOx) or silicon nanowires (GNS/SiNWs/CF) are studied for their applications

in Li-ion batteries and supercapacitors. GNS powders are synthesized via a chemical

oxidation of nature graphite followed by appropriate reduction process. GNS/MnOx

composites are synthesized rendering MnOx nanoparticles embedded on the surfaces

of GNS. SiNWs grown in carbon fibers are mechanically mixed with GNS powders.

The impacts of the manganese valence and GNS content in the GNS/MnOx

nanocomposites on the lithium storage characteristics in terms of Coulombic

efficiency, capacity, cycle life, and rate capability are investigated experimentally.

GNS/MnOx hybrid anode material shows a high reversible lithium storage capacity up

to 838 mAh/g with a Coulombic efficiency of 65%, good cyclic performance with 0.8%

fading per cycle, and high rate capability with 67% at the current density of 800 mA/g.

The GNS/SiNWs/CF hybrid exhibits high reversible

iii
1discharge capacity of 975 mAh/g with Coulombic efficiency of 85.7%. Further the

capacitive charge storage characteristics of both GNS and GNS/MnOx are studied in

ionic liquid electrolytes and aqueous electrolytes, respectively. Supercapacitors

consisting of GNS electrodes in EMIMBF4 ionic liquid electrolyte show the high

specific capacitance of 109 F/g, good cyclic life of over 1000 cycles, high power

density of 44.4 kW/kg, and high energy density of 256 Wh/kg etc. Supercapacitors

consisting of GNS/MnO2 nanocomposite electrodes in Na2SO4 aqueous electrolyte

deliver specific capacitance up to 164 F/g. This study shows significantly improved

electrochemical performances in lithium-ion batteries and supercapacitors by using

GNS-based composites.

iv
 Table of Contents
Chapter 1: Introduction .................................................................................................. 1

1.1 Lithium- ion Batteries and Anode Materials ................................................ 1

1.1.1 Carbon Anode Materials: ................................................................... 4

1.1.2 Metal Oxides: ..................................................................................... 8

1.1.3 Alloy-based Materials: ..................................................................... 11

1.2 Supercapacitors and Related materials ....................................................... 14

1.2.1 Electrode Materials of Supercapacitors ........................................... 15

1.2.2 Electrolyte Materials of Supercapacitors ......................................... 18

1.3 Objective and Scope of This Research ....................................................... 21

Chapter 2: Syntheses and Li-Storage Characteristics of Graphene Nanosheets and

Their MnOx Composites .............................................................................................. 22

2.1 Introduction ................................................................................................ 22

2.2 Experimental Aspects ................................................................................. 26

2.2.1 Chemical Synthesis of GNS Powders .............................................. 26

2.2.2 Synthesis of GNS/MnOx Hybrid Powders ....................................... 27

2.2.3 Structure, Morphology and Composition Analyses ......................... 27

2.2.4 Electrode and Electrochemical Characterizations ........................... 28

2.3 Results and Discussion ............................................................................... 29

2.3.1 Structure and Morphological Material Characterizations ................ 29

2.3.2 Electrochemical Characterizations................................................... 33

2.4 Summary .................................................................................................... 40

Chapter 3: Morphological and Li-Storage Characteristics of Silicon Nanowires in

Compositing with Carbon Fibers and Graphene Nanosheets ...................................... 41

3.1 Introduction ................................................................................................ 41

v
 3.1.1 Si - Carbon Matrix Composites ....................................................... 42

3.1.2 Nano-Structured Silicon Anodes...................................................... 45

3.2 Experimental Aspects ................................................................................. 50

3.2.1 Silicon Nanowires Growth ............................................................... 50

3.2.2 Electrode Preparation and Electrochemical Characterizations ........ 51

3.3 Results and Discussion ............................................................................... 51

Electrochemical Characterizations............................................................... 53

3.4 Summary .................................................................................................... 56

Chapter 4: Graphene-based Supercapacitors ............................................................... 58

4.1 Introduction ................................................................................................ 58

4.2 Research Status of Graphene-Based Electrodes ......................................... 60

4.2.1 Graphene Nanosheets: ..................................................................... 60

4.2.2 Graphene-based Composites:........................................................... 62

4.2.3 Graphene Oxide -MnOx Nanocomposites: ...................................... 64

4.3 Experimental Aspects ................................................................................. 66

4.4 Results and Discussion ............................................................................... 71

4.4.1 Cyclic Voltammetric Characteristics of GO and RGO in EMIMBF4

Electrolyte .................................................................................................... 71

4.4.2 Cyclic Voltammetric Characteristics of CRG in EMIMBF4

Electrolyte .................................................................................................... 72

4.4.3 Cyclic Voltammetric Characteristics of CRG in C8mimBF4 ........... 77

4.4.4 Electrochemical Impedance Spectroscopy Analys .......................... 79

4.4.5 Ionic Liquid Electrolytes Mixed at Different Ratio ......................... 81

4.4.6 GNS/MnOx in Aqueous Electrolyte ................................................. 84

4.5 Summary .................................................................................................... 88

vi
Chapter 5: Conclusion.................................................................................................. 91

REFERENCES ............................................................................................................ 94

vii
 LIST OF FIGURES

Figure 1.1 Schematic illustration of charge-discharge process in a lithium ion cell.

...................................................................................................................... 2

Figure 1.2 The first charge-discharge curves of nature flake graphite (NFG), mild

oxidized graphite (MGO), and carbon-coated MGO (C/MGO) . .................. 5

Figure 1.3 First charge-discharge curves for a hard carbon (HC) and two types of

porous carbons (PC-1 and PC-2). .................................................................. 7

Figure 1.4 The discharge-charge curves for graphene sheets. ............................... 8

Figure 1.5 First galvanostatic discharge-charge curves for SnO2. ......................... 9

Figure 1.6 First three galvanostatic discharge-charge curves for rutile TiO2

with different particle sizes . ..................................................................... 10

Figure 1.7 First galvanostatic discharge-charge curves for Co3O4 . .................... 11

Figure 1.8 SEM images of different nanostructured silicon anodes, (a) Si/carbon

composite, (b) silicon nanowires, (c) silicon nanotubes, and (d) silicon

hollow particles . .......................................................................................... 13

Figure 1.9 Principle of a single-cell double-layer capacitor and illustration of the

potential drop at the electrode/electrolyte interface . ................................... 14

Figure 1.10 Cyclic voltammogram profiles of different electrodes in aqueous

electrolyte: (a) active carbon, (b) PPy/VGCF, (c) hydrous RuO2; graphene

based electrode supercapacitors in different electrolytes: (d) aqueous

electrolyte (KOH), (e) organic electrolyte (TEABF4/AN) and (f)

EMIM-NTF2. ............................................................................................... 20

viii
Figure 2.1 Schematic view for synthesizing (a) RGO and (b) MnO2 Nanotubes

layer-by-layer thin films ............................................................................ 25

Figure 2.2 XRD spectra of the GNS/MnOx nanocomposites in comparison with

GNS and GNS+MnAC ................................................................................ 30

Figure 2.3 SEM Images of GNS at low magnification (a) and high magnification

(b). ................................................................................................................ 31

Figure 2.4 SEM images of GNS/MnOx-1 at low magnification (a) and high

magnification (b). ......................................................................................... 32

Figure 2.5 SEM images of GNS/MnOx-2 (a) and (b), and MnAC-400 (c) and (d)

...................................................................................................................... 32

Figure 2.6 Galvanostatic discharge-charge profiles of GNS anode. .................... 34

Figure 2.7 Galvanostatic discharge-charge profiles of GNS and GNS/MnOx

obtained at the current density of 50 mA/g .................................................. 36

Figure 2.8 The discharge-charge profiles of (a) GNS, (b) GNS/MnOx-1, (c)

GNS/MnOx-2 and (d) GNS/MnOx-3 at different discharge rates from 25

mA/g to 50 mA/g (the charge rate is set up at 50 mA/g), and (e) capacity as

a function of the discharge rate. ................................................................... 39

Figure 3.1 Schematics of a) CNT sponge used in CVD; b) Conformal CVD

deposition of amorphous Si onto CNT surface to form Si-CNT coaxial

nanostructure, and c) Pressed nanostrucutred Si-CNT for battery test. ....... 43

Figure 3.2 Top (a, c, e) and cross-sectional views (b, d, f) of as prepared porous

Si sample at different magnifications ......................................................... 46

Figure 3.3 SEM images of pristine SiNWs before (a) and after (b)

electrochemical cycling. .............................................................................. 48

ix
Figure 3.4 SEM images of Si crystalline – amorphous on stainless steel

substrate at 485℃for (a) 10 min, (b) 20 min, and (c) 40 min. ............... 48

Figure 3.5 SEM image of a vertically aligned SiNTs. ......................................... 49

Figure 3.6 Schematic of synthesizing SiNWs through CVD method. ................. 51

Figure 3.7 SEM and EDX mapping images showing morphology and element

distribution of SiNWs/CF. ........................................................................... 53

Figure 3.8 Galvanostatic discharge-charge profiles of (a) carbon nano-fibers (CF)

and (b) SiNWs (28wt%)/CF. ........................................................................ 55

Figure 3.9 Galvanostatic discharge-charge profiles of GO/SiNWs/CF

(67%/13%/17%）composite anode at the current density of 50 µA/cm2. ... 55

Figure 4.1 Schematics of a) graphene sheets and b) nanoparticle-modified

graphene sheets in its dispersion and dry state . .......................................... 62

Figure 4.2 Schematic illustration of fabrication process of RG/MnO2. ............ 68

Figure 4.3 Structure schematics of (a) [EMIM] +, (b) [C8mim] +, (c) [BF4]-. ...... 69

Figure 4.4 Cyclic voltammetric characteristics of GO (a) and RGO (b) electrode

in EMIMBF4 electrolyte at the scan rate of 20 mV/s with the voltage upper

window ranging from 2.5 V to 3.5 V. .......................................................... 71

Figure 4.5 Cyclic voltammetric characteristics of CRG electrode in EMIMBF4

electrolyte at the san rate of (a) 20 mV/s and (b) 50 mV/s with the voltage

upper window from 2.5 V to 4 V.................................................................. 73

Figure 4.6 Cyclic voltammetric characteristics of CRG electrode in EMIMBF4

electrolyte at the different scan rates, i.e. 20 mV/s, 50 mV/s, 100 mV/s, 250

mV/s, and 500 mV/s, with the voltage window of 0 – 4 V. ......................... 74

x
Figure 4.7 Galvanostatic charge-discharge curves of CRG electrode in

EMIMBF4 electrolyte at the current density of 0.5 A/g with the upper

voltage window ranging from 2.5 V – 3.5 V. ............................................... 75

Figure 4.8 Galvanostatic charge-discharge curve of CRG electrode in EMIMBF4

electrolyte at the current densities of 0.5 A/g and 1 A/g between 0 and 3.5V.

...................................................................................................................... 75

Figure 4.9 Coulombic efficiency and cyclic performance of CRG electrode in

EMIMBF4 electrolyte at current density of 1 A/g.. ...................................... 77

Figure 4.10 Cyclic voltammetric characteristics of CRG electrode in C8mimBF4

at the scan rate of (a) 20 mV/s and (b) 50 mV/s with the upper voltage

window ranging from 2.5 V – 4 V ............................................................... 78

Figure 4.11 Galvanostatic charge-discharge curves of CRG electrode in

C8mimBF4 electrolyte at the current density of 0.5 A/g and 1 A/g with the

voltage window of 0 – 3.5 V. ....................................................................... 79

Figure 4.12 Nyquist plots of CRG electrode in EMIMBF4 electrolyte (a) and

C8mimBF4 electrolyte (b) before charge/discharge process. ....................... 80

Figure 4.13 Cyclic voltammetric characteristics of CRG electrode in the mixing

EMIMBF4 and C8mimBF4 electrolyte with different volumetric ratio at the

scan rate of 20 mV/s with the upper voltage window at 2.5 V (a,b), 3.0V(c,d)

and 3.5 V (e,f). ............................................................................................. 82

Figure 4.14 Specific capacitances of CRG electrode changing with the

volumetric ratio of the mixing EMIMBF4 and C8mimBF4 electrolyte. ....... 83

Figure 4.15 Schematic illustration showing the charge distribution in individual

and mixing ionic liquids............................................................................... 84

xi
Figure 4.16 Cyclic voltammetric characterstics of (a) CRG/Mn3O4-50, (b)

CRG/Mn3O4-66, and (c) CRG/Mn3O4-75 electrodes in 1 M Na2SO4

electrolyte at different scan rates from 5 to 50 mV/s ................................... 86

Figure 4.17 Cyclic voltammetric curves of CRG/MnO2 electrode: at the scan rate

of 5 to 50 mV/s. ........................................................................................... 87

Figure 4.18 Cyclic voltammetric curves of RG/MnO2 electrode: at the scan rate

of 5 to 50 mV/s. ........................................................................................... 87

xii
Chapter 1: Introduction

Electrochemical energy conversion systems have many advantages in terms of

high energy conversion efficiency, high energy density, freedom from noise pollution,

and good mobility etc. Lithium ion batteries (LIBs) and supercapacitors are two most

common electric energy storage devices based on electrochemical principles. LIBs,

having the highest energy density among different kinds of batteries, no environment

effect, low self-discharge rate, and excellent long cycle life, etc, are now widely used

in consumer electronics such as laptops, cell phones, and personal digital assistants

(PDAs) [1] and are also very popular in military, hybrid electrical vehicle (HEV), and

aerospace applications. Supercapacitors due to the high power density, and

outstanding long cycle life etc., have tremendous application values and market

potential in electrical vehicles, communication, and national defense.

1.1 Lithium- ion Batteries and Anode Materials

A battery is made up of one or a group of interconnected electrochemical cells,

depending on the specific application it is designed for. An electrochemical cell

consists of a cathode (positive electrode), anode (negative electrode) separate with an

ionic conducting electrolyte. The electrical energy is generated by the redox reactions

between the cathode and anode. Redox is short for reduction-oxidation, and is an

electrochemical processing involving electron transfer to or from a molecule or ion

changing its oxidation state.

1
 LIBs using the insertion compounds as anode are developed from lithium

batteries to tackle the safety problem because of high chemical activity of lithium

anode. The discharge/charge process of a LIB is a reversible process of lithium ions

insertion/extraction into/from the host electrode materials, with no formation of

metallic lithium at the anode side. On charging, lithium ions transport from the

cathode to electrolyte and then intercalate into the carbonaceous anode material.

Electrons are removed from cathode through an external circuit to the anode. On

discharging, the lithium ions which are intercalated in the anode transport back to the

cathode, and oxidation reaction occurs at the anode and reduction at the cathode [2]. A

schematic of a typical Li ion cell is shown in Fig 1.1 [2].

Figure 1.1 Schematic illustration of charge-discharge process in a lithium ion cell.

Depends on different requirements and applications, various types of carbons, e.g.

graphite, coke, and mesocarbonmicrobeads (MCMB), whose electrical potentials are

close to the potential of Li are always chosen as anode materials. Common cathode

materials are lithium transition metal compounds LiMyXz(M represents the

transitional metal like Fe, Co, Ni, Mn, etc, while X represents the anion). Electrolyte
2
can be either liquid or solid. Liquid electrolyte is usually non-aqueous solution of

lithium salt such as LiClO4, LiPF6, and LiBF4 and various solvent including different

kinds of ethers, esters and carbonates. The reactions at both electrodes and the whole

Li ion cell can be shown as the following:

Cathode: Li1-xMyXz ↔ Li1-x-δMyXz + δLi+ +δe (1-1)

Anode: LixCn +δLi+ +δe ↔ Lix+δCn (1-2)

Overall: Li1-xMyXz+ LixCn ↔ Li1-x-δMyXz +δCn (1-3)

Li-ion batteries, compared with the other types of high-performance secondary

batteries, have many significant advantages:

 High work voltage. Due to using carbon as the negative electrode which has a

relative lower electrical potential (0.01-1.5V vs. Li/Li+), the loss of voltage will

be reduced to the minimum level. Li-ion batteries have a high working voltage

around 3.6V, which is equivalent to three Ni-Cd or Ni-MxH batteries.

 High energy density. Because the high working voltage and capacity, the energy

density is high. At present, the specific energy of Li-ion batteries can be as high

as 180Wh/kg which is four times higher than Ni-Cd batteries and two times

higher than Ni-MxH.

 The efficiency of charge-discharge can be as high as 100%.

 Low self-discharge rate. When a Li-ion battery is charging at the first time, there

will be a layer of solid-electrolyte interface (SEI) film forming on the surface of

electrodes. Ions are able to pass through the SEI, but the electrons will not. For

this reason, the self-discharge rate of the Li-ion battery is effectively prevented.

 Environmental friendly. The Li-ion battery doesn’t consist of poisonous

elementals such as Hg, Pb, and Cd.

3
 Theoretically, all the materials which have a relative low electric potential

compared with lithium metal, and the lithium ions are able to insertion and extraction

in them can be used as anode materials. In addition, ideal anode materials require the

following features: 1) good charge/discharge reversibility and cycling perdurability; 2)

low first irreversible capacity; 3) good compatibility with electrolyte solution; 4)

relative high specific capacity; 5) safety and environmental friendly; and 6) the cost is

low, etc. Different types of anode materials are listed in Table 1.1.

Table 1.1 Different Types of Anode Materials Used in Li-ion Batteries

Theoretical Practical
Material capacity capacity Advantages Disadvantages
(mAh/g) (mAh/g)
Graphite 372 250 Low cost, safe. Low charge capacities
Hard Carbon 372 418 High specific capacity High irreversible capacity
and poor cyclic
performance

Carbon 782 High charge-discharge

Nanotubes capacity and cyclic
performance
Graphene 744 350 Good mechanical properties. irreversible capacity is
High theoretical capacity high and voltage hysteresis
Metal oxide 880 700 High theoretical capacities Poor cyclic performance
composites
Tin-based 994 400 High theoretical capacities Poor cyclic performance
Material
Si 4200 2158 High theoretical capacities Poor cyclic performance
Si-Carbon Good cyclic performance,
composite 850 high capacity

1.1.1 Carbon Anode Materials:

Since lithium insertion in carbons was reported in 1980s, carbon materials are

widely used as the anode material. The insertion and extraction of lithium ions in

carbons, during the charge/discharge process, does not cause dramatic volume change.

On this account, the cycle life and safety are much better than that of lithium metal

batteries. Lithiated carbon has a potential very close of metallic lithium (potential

difference between lithiated carbon and lithium metal is ≤0.5V) which result in

Li-ion batteries a comparable whole cell voltage to the lithium metal batteries [3].
4
Extensive studies in the past decades revealed that the morphology of different

carbonaceous materials plays an important role in the lithium insertion process.

The theoretical capacity of natural flake graphite (NFG) is 372mAh/g (based on

the formation of LiC6), 10 times lower than that lithium metal (3862 mAh/g).

Practically, the reversible capacity is only about 300-350 mAh/g because of the

impurities and defects existing in graphite powders. NFG is also sensitive to the

electrolyte solution resulting in low value of first Coulombic efficiency and poor

cycle life performance. Several methods were applied to improve the electrochemical

performances of NFG, from mild oxidation[4], polymer coating [5], to metals or

metal oxides deposition [5, 6] . For instance, shown in figure 1.2, the NFG anode

showed the recharge capacity of 330.7 mAh/g with the Coulombic efficiency of

73.4%. After the mild oxidization and carbon coating process, the first charge

capacity increased to 344.7 mAh/g and 359.7 mAh/g with the Coulombic efficiency

of 89.7% and 83.2%, respectively.

Figure 1.2 The first charge-discharge curves of nature flake graphite (NFG), mild oxidized

graphite (MGO), and carbon-coated MGO (C/MGO) [7].

5
 Hard carbon is formed by interlaced single-layer graphite nanosheets. When

lithium ions are intercalated into it, Li ions can be adsorbed on both sides of the

single-layer graphite sheet leading to a higher specific lithium storage capacity.

Different from NFG, there is no obvious voltage platform in the charge/discharge

curves of hard carbons, which is beneficial to estimate the state of charging and

discharging in Li-ion batteries. In addition, it is clarified lithium metal deposition does

not occur up to 110% over-discharge of the reversible capacity of hard carbons [8].

Whereas in the case of graphite, the lithium metal deposition will occur at 105%

over-discharge. So the safety is improved by using the hard carbon as anode material.

High irreversible capacity and poor cyclic performance are the main limitations for

hard carbons. Some methods are applied to improve the electrochemical performance,

such as 1) reducing the impurities of H, so that the specific capacity and cyclic

performance will be enhanced [9] and 2) synthesizing the hard carbons under vacuum

to reduce the initial irreversible capacity [10]. Figure 1.3 shows the first

charge-discharge curves for a hard carbon and two types of porous hard carbons

treated by different chemical process methods [11]. The initial discharge and charge

capacities of the as-prepared specimen were 802.7 mAh/g and 223.9 mAh/g with the

Coulombic efficiency as low as 27.9%. In contrast, the discharge capacities of other

two types of porous carbon material were increased to around 1100 mAh/g with the

Coulombic efficiency about 40%.

6
Figure 1.3 First charge-discharge curves for a hard carbon (HC) and two types of porous carbons

(PC-1 and PC-2).

Figure 1.1 The discharge-charge curves for graphene sheets [13].

Carbon nanotubes (CNTs) are allotropes of carbon with cylindrical nanostructure.

CNTs with nano-sized tube diameter as well as interlaced apertures between the tubes,

provide superior intercalation media for high-capacity lithium storage. Lithium ions

can not only embed inside the tubes, but also in the apertures between the tubes.

7
 Graphene is another unique carbon allotrope which has a two-dimensional

crystal lattice structure. Graphene has outstanding mechanical properties, thermal and

electrical conductivities, and large specific surface area, offering great advantages for

application in Li-ion batteries. It is reported that graphene nanosheets have

theoretical capacity of 744 mAh/g, which corresponds to insertion of one lithium ion

per 3 carbon atoms[12]. Figure 1.4 presents the first discharge-charge profiles for

graphene nanosheets as anode material at the current density of 100 mA/g. The initial

discharge and charge capacities were as high as 2035 and 1264 mAh/g with the

Coulombic efficiency of 62.1%. However, some problems, such as high irreversible

capacity and large voltage hysteresis etc., need to be solved before it can be partially

utilized in Li-ion batteries.

1.1.2 Metal Oxides:

A variety of different metal oxides has been developed as anode materials for

LIBS. These metal oxides anodes have many distinguished features, such as high

capacity, environmental friendly, and good stability, etc. Typically, there are three

sorts of metal oxides, namely tin dioxide (SnO2), titanium dioxide (TiO2), and

transition metal oxides, which are categorized by their own distinct lithium storage

mechanisms and characteristics.

Tin oxide (SnO2) reacts with lithium through a two-step reaction: firstly, the tin

oxide is reversibly reduced to tin which is simultaneously embedded in the matrix of

Li2O; and then tin reversibly alloys with lithium. This procedure can be expressed by

the following equations:

4𝐿𝑖 + + 4𝑒 − + Sn𝑂2 ↔ 2𝐿𝑖2 𝑂 + 𝑆𝑛

Sn + x𝐿𝑖 + + 𝑥𝑒 − → 𝑆𝑛𝐿𝑖𝑥 (0 ≤ 𝑥 ≤ 4.4)

8
 The theoretical lithium storage capacity of SnO2 is as high as ~790mAh g-1,

which is almost twice higher than that of graphite (~372mAh g-1). Figure 1.6 shows

the first discharge-charge curves for a nanocrystalline SnO2 anode material. There is a

significant horizontal plateau in the 1 V region and a monotonous decrease below the

0.5 V region. The first discharge/charge capacity was 1735 and 665 mAh/g with the

Coulombic efficiency of 38.3%. However, the biggest obstacle prevents tin oxide

using in practical application is the large volume change during charge and discharge

process which leads the pulverization of the electrodes [14].

Figure 1.2 First galvanostatic discharge-charge curves for SnO2[15].

TiO2 is another kind of metal oxides anode material with distinct electrochemical

characteristics. There are eight polymorphs TiO2well known asrutile, anatase, spinel,

TiO2-B, TiO2-R, TiO2-H, TiO2-Ⅱ and TiO2-Ⅲ[16]. No matter which types of TiO2

the insertion reaction of Li ions into TiO2 matrix can be generally expressed as the

following equation [17]:

Ti𝑂2 + 𝑥𝐿𝑖 + + 𝑥𝑒 − → 𝐿𝑖𝑥 𝑇𝑖𝑂2

9
 TiO2 has many promising advantages such as good stability, low volume

expansion when lithiation, and ease of structure tailoring. However, TiO2 delivers a

relative low capacity (~170 mAh g-1) which limits its application. Researchers believe

that this problem can be overcome through the nanostructure modifications [18, 19].

In addition, with the decreasing of particle size, the ratio of atoms resides near the

surface layers have been increased hence increased surface Li storage. The transport

length for Li insertion into the bulk of TiOx will be decreased, and the Li insertion

will be more efficiently. As can be seen from figure 1.5, the initial discharge

capacities were increased from110 to 378.4 mAh/g with the particle sizes decreased

from 300 nm to 15 nm. Also, about 0.7 Li (264 mAh/g) per unit rutile TiO2 can be

reversibly extracted from particles with a size of 15 nm [20].

Figure 1.3 First three galvanostatic discharge-charge curves for rutile TiO2 with different particle

sizes [21].

Transition-metal oxides (TMO, 𝑇𝑀 = 𝐶𝑜, 𝐹𝑒, 𝑁𝑖, 𝐶𝑢 𝑒𝑡𝑐) have been reported to

have over 700 mAh/g theoretical capacities. The mechanism of Li reactivity differs

from the classical Li insertion/de-insertion (like TiO2) or Li-alloying processes (like

SnO2). Lithium reacts with such TMO anodes involving reversible formation and

decomposition of Li2O, accompanying the reduction and oxidation of metal

10
nanoparticles (in the range 1-5 nanometers) [22]. This reaction procedure can be

depicted by the following equation [23]:

2𝑒 − + 2𝐿𝑖 + + 𝑇𝑀𝑂 ↔ 𝑥𝐿𝑖2 𝑂 + 𝑀

The electrical conductivity of transition-metal oxides is relative low, and the

volume change during the charge-discharge process is high, so the discharge rate and

the cycling stability are poor. Reducing the material size will promote the secondary

reaction on the surface of the material lead to high irreversible capacity and shorten

the cycling performance. The micromorphology of the material affects the

electrochemical performance greatly. Taking Co3O4 for example, the electrochemical

performance of macroporous Co3O4 plates is superior to that of nanoparticles and

rod-like nanoclusters [24]. Using the composite material to enhance the stability of

the transition-metal oxides also attracts significant research interest. Zhang et al.[25]

have successfully synthesized the carbon coated iron oxide nanospindles which

enhance the conductivity of the electrode and lead to stabilized SEI films.

Figure 1.4 First galvanostatic discharge-charge curves for Co3O4 [26].

1.1.3 Alloy-based Materials:

In 1971, Dey [27] found that lithium can be electrochemically alloyed with some

metals, like Sn, Pb, Al, Cu, Mg, Zn, Ag, Cd and Si etc, at room temperature. The

11
theoretical capacity of tin is 994 mAh/g which is higher than that of graphite.

However, the volume change is great during the alloying/dealloying. So the key factor

to improve the electrochemical performance of metal-based anodes is to increase the

cycling performance. Tin alloys (SnM, M=Cu, Ni, Fe, and Co, ect) materials showed

excellent processing performances, good electric conductivity, and not as sensitive to

the environment as the graphite anode materials. There is a promising prospect to

replace the graphite anode materials. By changing the preparation methods to obtain

the nano-sized tin-based anode materials is a direction for the research in the future.

However, with reducing the size of the material there are some side effects, such as

the oxidation on the surface leading to the increasing of initial irreversible capacity

and the trend of aggregation will increase.

Silicon is an attractive alloy-based anode material for its highest theoretical

capacity (4200 mAh/g), and it is low cost and environment friendly. Similar to tin,

during the lithium ions insertion and extraction process the volume change by 400%

which will cause the electrode crushing and capacity fading. To circumvent this

problem, several research efforts have been carried out to disperse fine Si particles

within a solid, mixed conducting host matrix [28]. Carbonaceous materials are the

most common materials which act as both the electrochemically active materials and

the structural buffer materials. There are many advantages by using carbonaceous

materials as host matrix. First of all, the electric contact can be greatly improved

because of the high electric conductivity of carbonaceous materials. Furthermore,

carbonaceous host matrix can effectively remit the internal stress during the

charge/discharge process so that the volume of electrodes will not change greatly. At

last, carbonaceous materials can also contribute to the reversible capacity. Graphite,

graphene, mesophasemicrobeads (MCMB), etc. have been used as the host matrix for

12
silicon. Preparing the nano-sized and micro-organized Si-based materials as negative

electrode to overcome the pulverization problem and enhance the cycling life

performance is another hot area in recent years. Various types of Si nanomaterial have

been studied in the last four years, including Si nanowires, Si nanotubes, and Si

hollow nanoparticles [29]. A high specific capacity of ~3500mAh g-1 has been

achieved, which almost reached the theoretical capacity of silicon. At the same time,

the energy density and power density are increased significantly, and it allows the

battery working at high operation rate.

(a) (b)

(c)
(d)
Figure 1.5 SEM images of different nanostructured silicon anodes, (a) Si/carbon composite [30],

(b) silicon nanowires [31], (c) silicon nanotubes [32] and (d) silicon hollow particles [33].

13
1.2 Supercapacitors and Related materials

The structure of supercapacitors is similar to that of batteries in design and

manufacturing. As shown in Fig 1.9[34], an supercapacitor consist of two electrodes,

an electrolyte, and a separator [35]. When charging/discharging, opposite charges

accumulate on the surface of each electrode forming the electrical field that allows the

supercapacitor to storage energy. The most important task in supercapacitors is to

fabricate nanoscale electrode materials which have high surface area and/or high

porosity.

Figure 1.6. Principle of a single-cell double-layer capacitor and illustration of the potential drop

at the electrode/electrolyte interface [34].

Generally speaking, supercapacitors can be classified into two categories, based

on the energy storage mechanism [36]. One is called electrical double layer capacitor

(EDLC), and the other is pseudo-capacitor. In the EDLCs, the capacitance comes

from the pure electrostatic charge accumulated at the electrode/electrolyte interface.

Since there is no transfer of charge between the electrode and electrolyte, there are no
14
chemical or composition changes associated with non-Faradic processes. Because of

these, electrical energy storage in EDLCs is highly reversible and hence very long

cycle life. In contrast to EDLCs, the electric energy storage mechanism in

pseudo-capacitors involves charge transfer between electrode and electrolyte, which is

accomplished through electrosorption, reduction-oxidation reactions, and intercalation

processes [37, 38]. Due to this reason, the energy densities and capacitances of

pseudo-capacitors may be greater than EDLCs but cycle life will be shorter [39, 40].

There are three common electrode materials used in supercapacitors: carbons,

metal oxides, and conducting polymers. Not only the electrode material affects the

performance of supercapacitors strongly, but also the electrolyte is an important

component. When operating at high voltage, the supercapacitors provide high power

density and energy density. However, it is limited by the performance of electrolyte

applied, such as electrochemical stability, voltage window, and ionic concentration. In

general, the electrolyte applied in supercapacitors can be classified into three

categories, aqueous electrolyte, organic electrolyte, and ionic liquids (ILs).

1.2.1 Electrode Materials of Supercapacitors

Carbon Materials:

Carbon based materials are seen as prospective electrode materials in this

application. There are many advantages for carbon materials, such as low cost, easy

processing, abundance, environmental friendly, good electronic conductivity, higher

specific surface area, wide operating temperature range, and high chemical stability

[41]. Carbon materials are always used in EDLCs. Lots of porous carbon materials

with well-controlled uniform pore size have been synthesized [42], since the porous

15
carbons were first synthesized by Knox, et al. [43]. Figure 1.10 (a) shows a typical

cyclic voltammogram of an activated carbon in aqueous electrolyte.

Activated carbons have a complex structure composed of different size pores to

achieve high surface areas. Theoretically, the capacitance of supercapacitors strongly

depends on the surface area of electrode materials. However, not all of the high

surface area active carbons have higher capacitances. When the surface areas are

electrochemical inaccessible to conducting ions in electrolyte, such as closed pores or

electrolyte ions are too large to get into the micropores [44], these surface area may be

not fully effectively utilized in charge accumulation. As a consequence, the

capacitance of various materials does not linearly increase with the increasing of the

specific surface area [35]. It was observed that a porous carbon with a surface area as

high as 3000 m2 g-1only had a specific capacitance < 10 μF cm-2, which is much

smaller than the theoretical EDL capacitance (15-25 μF cm-2) [45].

Recently, nanostructured carbon allotropes like carbon nanotubes and graphene

nanosheets have attracted great attention in using as the electrode materials because of

their superior electrical properties and highly surface area. The mesopores in carbon

nanotube electrodes are interconnected, so the utilizing of surface area is more

efficiently to achieve higher capacitances. The electrolyte ions are allowed to transfer

freely inside the nanotubes. Details of graphene-based electrode materials in

supercapacitors will be addressed in Chapter 4.

Conducting Polymers:

Conducting polymers are rendered conductive through a conjugated bond system

along the polymer backbone. Typically, they are from either chemical oxidation or

electrochemical oxidation reaction of the monomer [46]. There are many advantages

for using conducting polymers as electrode materials for supercapacitors such as high

16
voltage window, high storage capacity, low cost, and adjustable redox activity through

chemical modification [47, 48]. It is believed that the n/p-type polymer configuration,

with one negatively charged and one positively charged conducting polymer electrode,

has the greatest power density and potential energy. However, there are also some

disadvantages by using conducting polymers. Firstly, the n-doped conducting polymer

materials prevented the pseudo-capacitors from reaching their potential [40, 49].

Secondly, the stability of the pseudo-capacitors is limited by the mechanical stress on

conducting polymers during the reduction-oxidation reactions [50]. Figure 1.10 (b)

shows a typical cyclic voltammogram of PPy/VCF polymer-carbon fiber composite

electrode in aqueous electrolyte.

Metal Oxides:

Metal oxides are always used as electrode materials for pseudo-capacitors. In

general, metal oxides show higher capacitance than carbon materials, and more stable

than conducting polymers. Ruthenium oxide (RuO2) attracts more attention in

aqueous-electrolyte based supercapacitor (see figure 1.10 c ) than other metal oxide

materials because of its high proton conductivity, long cycle life, good thermal

stability, wide potential window, and high rate capability [51]. Ruthenium oxide

achieves its capacitance by the intercalation of protons into its amorphous structure.

Furthermore, the resistance of ruthenium oxide is lower than that of other electrode

materials. Although it is prospective, the industrialization of it has been limited by its

prohibitive cost. Thus, how to lower the cost while maintain the performance is the

major area of research subject [39]. In recent years, people tried to use RuO2 and

carbon composited materials as electrode materials. By this method, the quantity of

RuO2 used has been significantly reduced, and at the same time, higher capacitance

17
has been maintained. Figure 1.10 (c) shows a typical cyclic voltammogram of an

hydrous RuO2 in aqueous electrolyte.

1.2.2 Electrolyte Materials of Supercapacitors

Aqueous electrolytes:

H2SO4, KOH, Na2SO4 etc. based aqueous solutions can provide relatively low

resistance and high ionic concentration. Supercapacitor which utilizes the aqueous

electrolyte always shows higher capacitances than that with organic electrolyte.

Furthermore, the process of preparing and utilizing aqueous electrolytes is not very

strict. The main disadvantage of aqueous electrolytes is its low voltage window (about

1.2 V) due to the thermodynamic decomposition of water, which limits the power and

energy density.

Organic electrolytes:

Organic electrolytes are lithium salts and quaternary ammonium salt, such as

LiClO4 and TEABF4, dissolved in some common organic solvent such as

polycarbonate (PC), acrylonitrile (ACN), and gamma – butyrolactone (GBL) etc.

Organic electrolyte has relatively higher achievable voltage and hence higher energy

density because the maximum energy stored is determined by the square of the

unit-cell voltage. The voltage of a supercapacitor applied organic electrolyte can be

above 2 V. The float voltage is 2.3 V, and it is possible to increase up to 2.7 V for a

short time [34]. Water content in the organic electrolyte is the main limitation for the

cell voltage. For this reason, extreme purification procedures are necessary before

these electrolytes are applied. Additionally, special surface coating on the electrodes is

necessary to protect water absorption. Besides the water contamination and high cost,

18
other considerations on the use of organic electrolytes are their low conductivity, high

resistance, and hence low power density.

Ionic liquids:

Ionic liquids (ILs) are the molten organic salts at room temperature which are

composed of solely cations and anions. ILs can provide a wide voltage window

ranging from 2 – 6 V and conductivity about 10 mS/cm [52]. With the increasing of

operating voltage, the maximum energy/power density will be greatly increased.

ILs, such as 1-Ethyl-3-methylimidazolium tetrafluoroborate (EMIMBF4),

1-Ethyl-3-methylimidazoliumhexafluorophosphate(EMIMPF6),1-Butyl-3-methylimid

azolium tetrafluoroborate (BMIMBF4) ect, have been widely researched in recent

years and regarded as the next generation electrolytes due to their interesting physical

properties such as high electrochemical stability over a wide potential window (>3 V),

high thermal stability, non-toxicity, and non-flammability, etc. However, because of

the high viscosity liquids and low ionic conductivity, the electrochemical performance

working as electrolyte in supercapacitors is effected [35]. Figure 1.10 (d-f) compares

the typical cyclic voltammograms of graphene-based electrode in the three different

electrolytes. The figure clearly shows that the voltage window extending from 1V to

around 2.4V and up to 3.5V when the electrolyte is changed from KOH aqueous

electrolyte, TEABF4/AN organic electrolyte, to EMIM-NTF2 ionic liquid.

19
 （a） (b) (c)

(d) (e) (f)

Figure 1.7 Cyclic voltammogram profiles of different electrodes in aqueous electrolyte: (a) active carbon, (b) PPy/VGCF [53] and (c) hydrous RuO2 [54];

graphene based electrode supercapacitors in different electrolyte: (d) aqueous electrolyte (KOH), (e) organic electrolyte (TEABF 4/AN) [55] and (f)

EMIM-NTF2 [56].

20
1.3 Objective and Scope of This Research

With the rapidly developing of various electronic devices such as electrical

vehicles and hybrid electric vehicles, the demand for high energy storage system has

significantly increased. Both lithium-ion batteries and supercapacitors are the main

electrochemical energy storage devices whose energy density and performance largely

depend on the chemical and physical properties of the electrode materials. In this

research study, several carbon-based nanomaterials are synthesized and employed as

the potential electrode materials in the lithium-ion batteries and supercapacitors

towards performance improvements.

Firstly, graphene nanosheets (GNS) and their manganese oxide composite

materials were synthesized through a new chemical approach. The material

morphological and electrochemical characteristics of the nanocomposite materials

synthesized at different conditions are revealed and discussed in chapter 2.

Secondly, silicon nanowires grown on carbon nano-fibers and their composites

with GNS were structurally characterized and preliminarily studied as the anode

material in lithium-ion batteries, which are presented in chapter 3.

Thirdly, GNS powders, synthesized using Hummers’s approach followed by

different reduction treatments were employed as the electrode materials in

supercapacitors with two types of ionic liquid (C8mimBF4 and EMIMBF4)

electrolytes. In addition, GNS in composites with manganese oxide were tested as the

electrode materials in supercapacitors in aqueous (1 M Na2SO4) electrolyte. Their

electrochemical performances are discussed in chapter 4.

The conclusions of the thesis work are summarized in chapter 5.

21
Chapter 2: Syntheses and Li-Storage Characteristics of

Graphene Nanosheets and Their MnOx Composites

2.1 Introduction

Nanoparticulate transition-metal oxides (TMOx, where M is Cu, Mn, Fe, Co or Ni

etc) have received much attention as alternative anode materials for Li-ion batteries.

The mechanism of TMOx working as anode materials is different from the classical Li

insertion or Li-alloying processes which can be described in equation 2-1 [57, 58]

2𝑒 − + 2𝐿𝑖 + + 𝑇𝑀𝑂𝑥 ↔ 𝑥𝐿𝑖2 𝑂 + 𝑇𝑀

According to equation 2-1, at the initial lithiation stage metal oxides are

converted to metallic state along with Li2O formation and will be reversibly returned

to its initial state upon delithiation. Since the oxidation state is fully utilized and more

than one electrons are involved in the conversion reaction, the transition metal oxide

anode materials always show the high energy densities and reversible capacities.

TMOx shows high reversible electrochemical capacities of 700 - 900 mAh/g, which is

over twice of the gravimetric capacity and six times volumetric capacity compared

with graphitic carbon anode [22, 59].

Manganese oxide (MnOx) has many advantages over other transition metal

oxides, such as high lithiation capacity (theoretical capacity 755.5 mA/g), relative low

electrical motivation force (1.032 V vs. Li/Li+), as well as environmental friendly and

low cost. However, there are still some limits for using manganese oxide-based

anodes in practical Li-ion batteries because of their poor electrical conductivity and

large volume change during cycling. It is widely believed that the nanoparticle size of

22
the MnOx may address the volumetric change problems and improve the

electrochemical performances. Zhong et al.[23] compared different nanoparticle sizes

of MnO powder through mechanical milling, and then coated with 5% carbon. These

two approaches effectively increased the cyclic performances. A sponge-like

nanosized Mn3O4 [58] was synthesized by reducing Mn(OOCCH3) with ammonium

hydroxide and hydrazine at 80℃ and then heated at 300℃ for 5 hours. The first

discharge and charge capacity were 1327 mAh/g and 869 mAh/g, respectively, at the

current rate of 0.25 C. The capacity maintained at ~800 mAh/g after 40 cycles which

indicated the good cyclic performance. In addition, as the current density increased up

to 10 C, this anode material could still deliver a specific capacity as high as 500

mAh/g indicating improved rate capability. Recently, many research groups are

working on synthesizing different types of nanostructured manganese oxide-based

anode materials, such as Mn3O4 nano-fibers, and interconnected MnOx nanowires, etc.

Meanwhile, it is believed that graphene nanosheets (GNS) are able to serve as the

good matrix for the transition metal oxides for its ultra-high surface area, and good

electrical conductivity which will compensate the low conductivity of metal oxides

and improve the electrochemical performance as anode materials. In order to further

increase the surface area of the GNS and the lithium storage performance, MnOx has

been added to form 3 D nanostructured MnOx-GNS hybrids. There are many

approaches to synthesis the graphene/MnOx composites around the world.

Kim et al.[60] used chemical reduction method to disperse Mn3O4 nanoparticles

(20 – 30 nm) GNS. This hybrid material retained a high capacity of ~990 mAh/g after

30 cycles. It was observed that the first discharge curve was quite different from

the following ones, which indicated that an irreversible structural change

occurred. Through XRD and Raman spectroscopy analysis, they reported that at

23
discharging to 1.0 V some degree of solid solution LiyMn3O4 (1≤ y) formed. As

the continuing discharge, MnOx (x ~ 2) formed.

Liu et al.[61] synthesized the manganese oxide/GNS using the hydrothermal

reduction method. The Mn3O4 and MnO nanoparticles were uniformly dispersed on

the surface of graphite nanosheets. This sandwich structure improved the

electrochemical performance of barely Mn3O4 and MnO nanoparticles because of the

buffering, confining and conducting effects of graphite nanosheets. The Mn3O4/GNS

hybrid material maintained 437 mAh/g at the current density of 200 mA/g after 50

cycles. It is needed to mention that the first discharge capacity of this anode material

was 1100 mAh/g, and charge capacity was 750 mAh/g, accordingly the Coulombic

efficiency was 68%.

Wang et al.[62] hydrolyzed Mn(CH3COO)2 in GO suspension and then did the

hydrothermal treatment at 180℃ to obtain reduced graphene oxide (RGO) and

Mn3O4 nanocomposites. The as-prepared anode material showed high capacity of 810

mAh/g (based on the total mass of the hybrid) and ~900 mA/g (based on the mass of

Mn3O4) at the current density of 40 mA/g. Also, the capacity retention was as high as

98% after first five cycles. Even increased the current density to ten times (400 mA/g),

the capacity still maintained as high as ~780% mAh/g after 40 cycles. This hybrid

anode material greatly increased the electrochemical performance compared with free

Mn3O4 nanoparticles whose specific capacity was as low as ~300 mAh/g and

decreased to ~115 mAh/g after 10 cycles.

Yu et al.[63] prepared a layer-by-layer RGO-MnO2 nanotube composite (see

figure 2.1) as anode material. The composite was composed by 20 layers films (10

layers of RGO nanosheets and 10 layers of MnO 2 nanosheets with 1:1 weight ratio).

The researchers reported that this composite film exhibited a high reversible capacity

24
of 686 mAh/g at the current density of 100 mA/g, and 208 mAh/g event at the current

density as high as 1600 mA/g.

Figure 2.1 Schematic view for synthesizing (a) RGO and (b) MnO2 Nanotubes layer-by-layer thin

films [63].

Table 2.1 summarizes the information of MnOx nanomaterials and GNS/MnOx

nanocomposites collected from published literatures including their morphology,

synthesis conditions, manganese component phases, GNS contents, and

electrochemical performances.

In this research study, we developed a low-cost sol-gel approach to disperse

MnOx nanoparticles in graphene GNS matrix and employed the nanocomposites as

anode materials for lithium-ion batteries. The objective is to experimentally clarify the

impacts of the manganese valence and GNS content in the composite on the lithium

storage characteristics in terms of Coulombic efficiency, capacity, cycle life and rate

capability.

25
Table 2.1 Summary of synthesis and electrochemical Information about MnO x–based

nanomaterials collected from published literatures.

Sample Morphology Basic synthesis Info 1st d/c capacities Rate capability Reference
(mAh/g), efficiency (mAh/g at mA/g)
MnO2 MnAC+(NH4)S2O7, 140℃ 1146/627, 55% [23]
rodlike 100–150nm in 2hrs, then 500℃ 10 hrs
Mn2O3 diameter,1-2 Ibid, amorphous MnO2 350℃ 1156/547, 47% No info [23]
µminlength 10hrs
Mn3O4 ibid amorphous MnO2 280℃ 3 1265/528, 42% [23]
hrs in H2/Ar, 700℃ 2 hrs in Ar
MnO ibid amorphous MnO2 400℃ 1728/488, 28% [23]
10 hrs in H2/Ar, 700℃ 2 hrs in
Ar
MnO aggolomerate, 0.5–1.5 MnO (commercial) 1270/690, 53% 300 at 800 [64]
µm Carbon coated-MnO 1080/670, 62%
MnO nanosheets, 0.5–1.5 Ball milled-MnO 1240/750, 61% 300 at 800 [64]
µm 1080/700, 65%
Mn3O4 Spongelike 30 – 80 nm MnAC and ammonium 1327/869, 65% 520 at 500 [65]
hydroxide at 300℃ 5 hrs
Mn3O4/GNS 20 – 30 nm GO + MnCl2 + KMnO4 + 1789/1100, 62% 400 at 1000 [60]
NH4OH, 80 ℃ 8 hrs, 120 ℃
dried
Mn3O4/GNS 30 – 50 nm GO + MnAC + NaOH, 1100/750, 68% 610 at 200 [61]
hydrothermal 180℃ for 12 hrs,
15% GNS
MnO/GNS 30 – 50 nm ibid, then 400℃ 2 hrs in H2/N2, 1320/820, 62% 200 at 600 [61]
12% GNS
Mn3O4/GNS 10 – 20 nm GO + MnAC, hydrolysis, 1320/850, 64% 600 at 800 [62]
hydrothermal at 180 ℃ for 10
hrs, 10% RGO
MnO2/GNS 70 – 80 nm GO + MnO2 nanotube 1232/686, 55.7% 300 at 800 [66]
hydrothermal

2.2 Experimental Aspects

2.2.1 Chemical Synthesis of GNS Powders

Natural graphite powders (HMP 850) were oxidized to graphite oxide (GO) using

the modified hummer’s method. 100 mL 98% sulfuric acid was poured into the

mixture of 2 gram graphite (HMP 850) and 2 gram sodium nitrate. Then 10 gram of

potassium permanganate (KMnO4) is gradually added. The mixture was completely

stirred for overnight, and diluted with deionized water. 12 mL hydrogen peroxide was

added to the solution, and the colour of the mixture was changed to greenish yellow.

The obtained solution was put on the hot plate at 150℃ until it gets dried. The dried

sample was washed with deionized water until its pH value turns to neutral. Then the
26
sample was placed in ultrasonic wave for 2 hours. GNS powder was obtained after the

as prepared sample was dried.

2.2.2 Synthesis of GNS/MnOx Hybrid Powders

22 mg as-prepared GNS powders dissolved in water and were ultrasonicated for 3

hours. Then 78 mg manganese acetate (MnAC) dissolved in water and was gradually

added to the GNS solution then continuously stirred for 2 hours. Ammonium

hydroxide and hydrazine were added to the mixture and stirred for 3 hours at

100℃.The mixture was filtered and dried at 150℃.The different GNS/MnOx nano

composite anode material with their acronyms and description were listed in Table

2.2.

Table 2.2 GNS/MnOx Anode Samples

S.No Acronym Anode Sample

1 GNS/MnOx-1 Nano composite heat treated at 150°C in air.
2 GNS/MnOx-2 Nano composite heat treated at 400°C in air.

Nanocomposite heat treated at 400°C in 5% H2/Ar

3 GNS/MnOx-3
atmosphere.

2.2.3 Structure, Morphology and Composition Analyses

Bruker D8 X-ray diffractometer (XRD) was used to identify the crystal structure

of the manganese component in the hybrid. The morphologies of the GNS/MnOx

sample were visualized by the JEOL scanning electron microscopy (SEM). The

carbon content was identified by analyzing the weight loss of the powder after heat

treatment and energy dispersive spectroscopy (EDX).

27
2.2.4 Electrode and Electrochemical Characterizations

Below are the steps of preparing Swagelok-type battery cells used for testing in

electrochemical working station. The active anode material was mixed with a binder

polyvinylidene fluoride (PVDF) in N-Methyl-Pyrrolidone (NMP) at the weight ratio

of 9:1 to from viscous slurry. The slurry was uniformly coated on Cu foil with a blade.

The as-prepared electrode sheets were dried at 120℃ for 12 hours in vacuum oven.

Afterwards, the electrode sheet is punched into round specimens with a diameter of

3/8 inch for electrochemical characterizations. The electrode weight is the total weight

of the copper, PVDF binder and the active anode material. The net weight of the

active anode material, i.e. GNS or GNS/MnOx, is the total weight minus the weight of

the copper foil and PVDF binder. The actual weight was used for calculating the

performance of the anode material. All the chemicals used in the electrode preparation

are purchased from Sigma-Aldrich.

Swagelok cells were assembled in glove box which controlled the oxygen and

moisture levels less than 0.5 ppm. The electrolyte was 1 M LiPF6 dissolved in a

mixture of ethylene carbonate (EC) and dimethyl carbonate (DEC) at 1:1 volumetric

ratio. Lithium metal foil was used as counter and reference electrode.

The cells were galvanostatic charge-discharge at the preset current density with

the voltage window ranging from 0.01 – 3.0 V on a battery testing station (Land CT).

Electrochemical impedance spectra (EIS) was obtained by applying a sine-wave

signal with an amplitude of 10 mV in the frequency range of 1 MHz to 0.1 Hz on

Camry electrochemical analysis system at the preset capacity interval after relaxation

for 2 hours.

28
2.3 Results and Discussion

2.3.1 Structure and Morphological Material Characterizations

Figure 2.2 shows the XRD profiles of the three samples of GNS/MnOx hybrid

with different processes methods. For comparison, GNS and GNS+MnAC simple

mechanically mixing of GNS + MnAC were also included. In the GNS profile, there

was only a board peak at 24.6°which is the characteristic XRD spectrum of GNS.

This indicated that all graphites were converted into reduced graphene oxide

nanosheets. The spectrum of GNS + MnAC shows many crystalline diffraction peaks

corroborated well with those of crystalline MnAC. It is reported that crystalline

MnAC can only transform to Mn3O4 with the thermal treatment above 400℃ in air.

However, in GNS/MnOx-1 profile, all the crystalline MnAC related peaks disappeared,

which confirms that all in the MnAC had transformed to MnOx. In addition, all the

observed peaks in GNS/MnOx-1 corroborated well with the Mn3O4. The nanoparticles

size of the Mn3O4 was estimated about 30 nm, based on the width of diffraction peaks.

In profiles of GNS/MnOx-2 which was heated at 400℃, the phase of Mn3O4 was not

altered but the width of the diffraction peaks was increased indicating the increase of

particles size. Thermal treatment at 400℃ in the H2/Ar environment, the MnO peaks

can be observed and are labeled with stars. Semi-quantitative analyses, from the XRD

spectrum, indicated that the atomic ratio of MnO to Mn3O4 was 3:2.

29
 * * GNS/MnOx-3

GNS/MnOx-2

GNS/MnOx-1
Intensity

GNS+MnAC

GNS

20 24 28 32 36 40 44 48 52 56 60 64



Figure 2.2 XRD spectra of the GNS/MnOx nanocomposites in comparison with GNS and

GNS+MnAC

EDX micro elemental composition analysis was employed to determine the

carbon and manganese content. Five different regions with different area sizes were

selected on each sample compositional analyses to improve the accuracy. The average

carbon contents in the three specimens were 44 wt%, 20 wt%, and 42%, respectively.

The prepared conditions, manganese component phases, and GNS contents in the

samples prepared at the three different conditions are listed in Table 2.3.

The morphologies of the GNS and GNS/MnOx hybrids were examined by using

SEM. Figures 2.3 (a) and (b) show the SEM images of GNS at low and high

magnification, respectively. From the images it can be observed that the GNS

synthesized by the modified Hummer’s method shows thin wrinkle flakes indicating

high surface area. Atomic force microscopic images also confirmed the existence of

single graphene nanosheets [67].

30
 (a) (b)
Figure 2.3 SEM Images of GNS at low magnification (a) and high magnification (b).

During the synthesis process of GNS by modified Hummer’s method, a number

of different function groups such as epoxyl, carboxyl and hydroxyl, attached on the

surface and edges resulting the wrinkled morphology. These function groups will

coordinate with Mn ions which is driven by electrostatic interactions. Mn ion – GNS

mixture formed when all the Mn ions absorbed onto GNS. Ammonium hydroxideand

hydrazine solution was added to help the precipitation of Mn ion into Mn(OH)2.

Mn(OH)2 is a metastable compound that can be easily oxided to Mn3O4 by oxygen. It

can be observed from figures 2.4 (a) and (b) that manganese oxide nanoparticles

homogeneously bonded onto both sides of the GNS surface. The GNS embedded with

the MnOx nanoparticles, which stacked and cross-linked formed multilayer sandwich

structure leading to relatively thick flaky type morphology than GNS. SEM images of

GNS/MnOx – 3 are similar to those of GNS/MnOx – 1 and hence are not shown

again.

Lithium storage electrochemical reaction is anticipated to be benefited from this

sandwich structure. Firstly, the GNS working as a strain buffer would accommodate

the volume change caused by the insertion/extraction of lithium ions. Secondly, the

GNS network provided a short path for electrical conducting and decreased the

internal impendence of the electrode. At last, MnOx nanoparticles served as “spacer”

31
between the GNS forming the 3-D structure which will effectively prevent the

restacking of individual graphene nanosheets.

(a) (b)
Figure 2.4 SEM images of GNS/MnOx-1 at low magnification (a) and high magnification (b).

（a） (b)

(c) (d)
Figure 2.5 SEM images of GNS/MnOx-2 (a) and (b), and MnAC-400 (c) and (d)

Loose fluffy agglomerates can be observed in the GNS/MnOx-400 SEM images,

shown in figures 2.5 (a) and (b). Some parts of GNS were burnoff after heated at

400℃, hence, some Mn3O4 nanoparticles existed as free-standing agglomerates. This

32
is similar with pure manganese acetate heated at 400℃ in air for 12 hours, which

showed large spherical agglomerates as shown in figure 2.5 (c) and (d). This structure

is unfavorable that will decrease the cycle performance and electrical conduction for

reversible lithium storage.

2.3.2 Electrochemical Characterizations

Electrochemical Characteristics of GNS

Figure 2.6 shows the first, second, fifth and tenth charge-discharge curves of the

GNS anode at the current density of 50 mA/g. The first discharge curve rapidly

decreased to around 1.2 V and then gradually decreased to 0 V with a midpoint of

around 0.5 V. While charging, the curve gradually increased with a midpoint of

around 1.5 V. The charge-discharge curves showed large voltage hysteresis and no

distinguishable plateau, which are the characteristics of lithium storage in GNS. The

first discharge and charge specific capacity were 1248 mAh/g and 843 mAh/g,

respectively, with a Coulombic efficiency of 68%. The irreversible capacity is the

results of SEI formation, irreversible lithium absorption, and electrolyte reaction with

the function groups of GNS. The reversible capacity of GNS was as high as around

800 mAh/g which is almost twice of commercially used graphite (372 mAh/g) anode.

This is because lithium ions are able to embed on the both sides of the graphene

nanosheet and also on some function groups of GNS. In the following ten cycles,

Coulombic efficiency significantly increased and stabilized over 95%. This indicates

the good cycle performance of GNS anode material. The general lithium storage

reaction in GNS can be expressed in the following equation:

GNS + x𝐿𝑖 + + 𝑥𝑒 − ↔ 𝐿𝑖𝑥 𝐺𝑁𝑆 (1)

33
 3.5
GNS
10 5 2 1
3.0

2.5

2.0

Voltage (V)
1.5

1.0

0.5
1st discharge
0.0
0 200 400 600 800 1000 1200 1400

Capacity (mAh/g)

Figure 2.6 Galvanostatic discharge-charge profiles of GNS anode.

Electrochemical Characteristics of GNS/MnOx

The shapes of GNS charge-discharge curves significantly changed while Mn3O4

nanoparticles are embedded on GNS (shown in Figure 2.7 (a) – (c)). The first

discharge/charge capacities for the three GNS/MnOx hybrid samples were 1430/850,

1504/578, and 1344/838 mAh/g with the Coulombic efficiency of 59%, 40% and 65%,

respectively. The initial discharge plot can be divided into 3 regions. The first region

was from 3.0 V to 0.5 V which is correlated with the insertion of lithium ions into

Mn3O4 forming LiMn3O4 as follows:

Mn3O4 + xLi+ + xe-→LixMn3O4 (2)

The plateau between 0.5 V to 0.3 V is correlated with the formation of SEI and

the phase transformation from LixMn3O4 to MnO as follows:

LiMn3O4 + Li+ + e-→ 3MnO + Li2O (3)

Because the SEI formation and reaction (2) and (3) are irreversible, the first discharge

plot is always different from the following ones. Comparing figure 2.7 (a) and (b), the

GNS/MnOx-2 exhibited insignificant high-voltage slop. This is because the less

content of GNS decreased the electronic conductivity, which resulted in the large

internal impendence and overvoltage.

34
 The reaction of lithium ion with manganese oxide can be observed at the long

plateau at around 0.4V as well as in the region below 0.4 V in the initial discharge

plot and expressed as follows [60, 61]:

MnO + 2Li+ + 2e ↔ Mn + Li2O (4)

Comparing with the initial discharge plot of GNS (see figure 2.6 and equation (1)), it

can be concluded that in the low-voltage region, GNS also contributes a portion of the

capacity.

Among the three reaction equations between lithium and manganese oxides,

reactions (2) and (3) are irreversible and mainly contributes to the irreversible

capacity during the first discharge process while only the reaction (4) is reversible.

According to previous studies [60, 68] based on XRD and Raman analyses, no matter

the phase of the starting material is MnO, Mn3O4, or MnO2, manganese would be

oxidized to nanosized MnO once fully charged to 3.0 V. Hence the high valency

manganese oxide will lead to form extra Li2O which produce extra irreversible

capacity in the initial discharge process. The formation of extra Li2O is in the order of

MnO2 (1·Li2O) > Mn2O3 (1/2·Li2O) > Mn3O4 (1/3·Li2O).

The 2nd, 5th, and 10th cycles charge-discharge profiles are also shown in figures 2.7

(a) to (d). The shapes of the plot from the 2nd cycle didn't change significantly which

indicated the stable cycle performance. Both of GNS/MnOx-1 and GNS/MnOx-3

showed the Coulombic efficiency as high as over 92%. However, the Coulombic

efficiency of GNS/MnOx-2 was only about 80% in the following nine cycles.

35
 3.5
GNS/MnOx-1
3.0 10 5 2 1

2.5

Voltage (V)
2.0

1.5

1.0

0.5
1st discharge

0.0
0 200 400 600 800 1000 1200 1400

(a) Capacity (mAh/g)

3.5
GNS/MnOx-2
10 5 2 1
3.0

2.5

2.0
Voltage (V)

1.5

1.0

0.5
1st discharge

0.0
0 200 400 600 800 1000 1200 1400 1600 1800

(b) Capacity (mAh/g)

3.5
GNS/MnOx-3
10 5 2 1
3.0

2.5

2.0
Voltage (V)

1.5

1.0

0.5
1st discharge

0.0
0 200 400 600 800 1000 1200 1400

(c) Capacity (mAh/g)

Figure 2.7 Galvanostatic discharge-charge profiles of GNS and GNS/MnOx obtained at the current

density of 50 mA/g

36
Table 2.3 Summary of GNS and GNS/MnOx hybrid anode materials with their prepared conditions,
manganese component phases, GNS contents, and electrochemical performances.

Sample GNS GNS/MnOx-1 GNS/MnOx-2 GNS/MnOx-3

Synthesis info As-prepared As-prepared 400℃ 12 h in air 400℃ 12 h in 5% H2/Ar

GNS content 100% 44% 20% 42%
MnOx phase N/A Mn3O4 Mn3O4 Mn3O4 +MnO (40 : 60)
st
1 capacity 1248/843 1430/850 1433/578 1279/838
st
1 efficiency 68% 59% 40% 65%
nd
2 capacity 810 806 494 772
th
10 capacity 790 619 233 719
Capacity fading 0.3% 2.6% 5.9% 0.8%

Table 2.3 summarizes the GNS and GNS/MnOx hybrid anode materials with their

prepared conditions, manganese component phases, GNS contents, and

electrochemical performances. As shown in Table 2.3, the capacity fading of

GNS/MnOx-2 is the highest of 5.9% compared the other two GNS/MnOx hybrid

materials, because the GNS content in the former sample is the lowest of 20%. This

means the presence of sufficient GNS can provide good buffer to accommodate the

volume change during the lithium ions insertion/extraction process, and also good

electronic conductive path. Comparing GNS/MnOx – 1 with GNS/MnOx – 2, it can be

concluded that the more Mn3O4 in the starting materials the less Coulombic efficiency.

Comparing with GNS/MnOx-1 and GNS/MnOx-3, the GNS contents are almost same,

but there is more MnO in GNS/MnOx-3 and the efficiency is higher than

GNS/MnOx-1. The difference of the cyclic performance is due to the presence of

MnO in GNS/MnOx-3. However, the sample with the best cyclic performance of

GNS/MnOx-3 still displayed slightly inferior to pure GNS, indicating the limited

reversibility of the displacement reaction of MnOx.

Figures 2.8 (a) to (d) show the charge-discharge profiles of the four samples at

different discharge current rates from 25 mA/g to 800 mA/g. Also, the capacity as a

37
function of discharge current rate is shown in figure 2.8 (e). GNS/MnOx-3 showed

best rate capability for its content of GNS and MnO is high and even better than that

of GNS. At the current density increased from 50 mA/g to 800 mA/g, the capacity

decreased from 810 mAh/g to 425 mAh/g (52.4%) for GNS. In contrast, for

GNS/MnOx-3, there was still 67% of full capacity maintained at the current density of

800 mA/g.

The metallic lithium formation was avoided by presetting the cutoff potential to

10 mV. So the polarization caused by increasing the discharge current density is

equivalent to shift the cutoff voltage upwards. From figure 2.8 (a) it can be observed

that the main capacity was contributed from the voltage region below 0.3 V. When

increasing the discharge current density, it is equal to up-shifting the voltage baseline

and decreasing the capacity. For GNS anode, when increasing the discharge current

from 25 mA/g to 800 mA/g, the overvoltage potential was around 300 mV. The

capacity read from the discharge curve in 2.8 (a) at 300 mV and 25 mA/g is 410

mAh/g which is consistent with the discharge capacity at 800 mA/g (425 mAh/g).

However, in figure GNS/MnOx-3, the capacity was mainly contributed above 300 mV.

Therefore, at 300 mV polarization, there was less impact on it, and 475 mAh/g was

still maintained.

38
 (a) (b)

(c) (d)

(e)

Figure 2.8 The discharge-charge profiles of (a) GNS, (b) GNS/MnOx-1, (c) GNS/MnOx-2 and (d)

GNS/MnOx-3 at different discharge rates from 25 mA/g to 50 mA/g (the charge rate is set up at

50 mA/g), and (e) capacity as a function of the discharge rate.

39
2.4 Summary

In this study, GNS/MnOx hybrid anode materials with different content of GNS

and MnOx were chemically synthesized. Their structures, morphology as well as the

electrochemical characteristics were analyzed. MnOx nanoparticles worked as “spacer”

between graphene nanosheets and formed a 3-D structure, which greatly improved the

lithium storage performance. In addition, GNS worked as a “buffer” which

accommodated the volume change caused by the manganese oxide displacement

reaction, and shorten the electronic conductivity path.

It was verified that the lower manganese valency would deliver higher reversible

capacity. GNS/MnOx-3, consisting of GNS 42% wt, Mn3O4 23% wt%, and MnO 35%

which is the one with highest GNS and MnO contents delivered better performances

than the other specimens. Specifically, the reversible capacity up to 838 mAh/g has

been obtained, with a Coulombic efficiency of 65%. In addition, the specimen

exhibited good cyclic performance with 0.8% fading per cycle and high rate

capability with 67% at the current density of 800 mA/g. GNS/MnOx has a relatively

high lithium storage potential versus Li (0.5 V) which is helpful to increase the high

rate capability and reduce the risk of metallic lithium formation and safety hazard.

40
Chapter 3: Morphological and Li-Storage Characteristics of

Silicon Nanowires in Compositing with Carbon Fibers and

Graphene Nanosheets

3.1 Introduction

Silicon is an alternative promising anode materials for Li-ion batteries. It was

reported that the alloying process in silicon anodes results in formation of Li12Si7,

Li14Si6, Li13Si4, and Li22Si5, alloys [69]. Table 3.1 shows the crystal structure, unit cell

volume, and volume per silicon atom of different Li-Sn alloys. When the Li22Si5 alloy

forms, each silicon atom accommodates 4.4 lithium atoms corresponding the specific

capacity is 4200 mAh/g, which is the highest value known so far. However, the

progressively large volume expansion occurs simultaneously. The volume of Li4.4Si is

four times higher than that of the parent silicon. Such high volume change will cause

disintegration of the electrode and the dramatically fade of cycling performance.

Table 3.1Crystal structure, unit cell volume and volume per Si atom for the Li-Si system [69]

Compound crystal structure Unit cell Volume ( ̇ 3) Volume per silicon atom ( ̇ 3)

Silicon Cubic 160.2 20.0

Li12Si7, (Li1.71Si) Orthorhombic 243.6 58.0

Li14Si6, (Li1.71Si) Rhombohedral 308.9 51.5

Li13Si4, (Li3.25Si) Orthorhombic 538.4 67.3

Li22Si5, (Li4.4Si) Cubic 659.2 82.4

41
 Various methods have been employed to improve the electronic contact between

particles to mediate the volume change impacts. One is mixing conductive additives,

for instance, graphite flakes or nano-scale carbon black to increase the electrical

contact between components in the electrode[70]. The conductive additives suppress

the Si volume change on some extent, and improve the Li de-alloying kinetics, also

relieving mechanical stresses. Reducing the Si particles size to nano-size and

developing nanostructured Si is another very effective method.

3.1.1 Si - Carbon Matrix Composites

Carbon has been used as the matrix because of its softness, good electronic

conductivity, relative low mas, reasonable Li-insertion ability, and small volume

expansion [71]. Therefore, Si dispersed in carbon and its allotrope tested as anode

material aroused great interest.

Si- Active Carbon Composites:

Z. Luo et al. [72] reported Si-carbon composites by mixing Si nanoparticles

(30nm to 50nm) with active carbon. In order to find the best amount of silicon carbon

composites for the best electrochemical performance, several different samples were

made by different silicon powder content ranged from 1% to 7%. It was found that the

reversible capacity increased and the first efficiency decreased correspondingly. The

highest first discharge capacity of 579 mAh/g was obtained at 7% silicon content. The

reversible capacity increased from 358 mAh/g to 478 mAh/g when the silicon content

ranged from 1% to 7%, and the first Coulombic efficiency decreased from 85.2% to

82.6%.

Q. Si, et al. [73] showed that the electrochemical performance of the Si-carbon

composite depended on the particle size and the content of Si. They found the best

42
result was obtained from the specimen with silicon particle size of 50 nm composite

with 48% carbon. Coulombic efficiency in the first cycle of 69.2% and the second

cycle reversible discharge capacity of 970 mAh/g were obtained. Meanwhile, the

capacity fading was 0.24% per cycle in the first 40 cycles. Also, 10-20 nm nano-Si

deposited on the graphite by thermal vapor deposition of silane [74] were studied. The

initial charge and discharge capacity were 1350 mAh/g and 1000 mAh/g, respectively.

After 100 cycles, the reversible capacity was maintained over 900 mAh/g.

Si- Carbon Nanotube Composites:

L. Hu, et al.[29] prepared Si-carbon nanotube (Si-CNT) made of amorphous Si

on porous embedded in 3D CNT sponges through chemical vapor deposition (CVD)

which is shown as Fig.3.1. CNT functions as both mechanical support and electrical

conductor. Excellent contact between the electrolyte and silicon was obtained because

the porous structure of the CNT provided sufficient channels for the flow of

electrolyte. The discharge and charge capacities in the first cycle were 3200 mAh/g

and 2750 mAh/g, respectively, with a Coulombic efficiency of 86%. They pointed that

the cycling performance could be greatly enhanced with the adjustment of the cutoff

voltage although the capacity will decrease a bit.

Figure 3.1 Schematics of a) CNT sponge used in CVD; b) Conformal CVD deposition of amorphous

Si onto CNT surface to form Si-CNT coaxial nanostructure, and c) Pressed nanostrucutred Si-CNT

for battery test [29].

43
Si- Carbon Nano-fibers Anode:

Ji and Zhang [75] reported a high performance Si and carbon nano-fibers (CNF)

composite with porous structure which was simply synthesized by electrospinning and

subsequent carbonization of the mixture of polyacrylonitrile (PAN), poly-L-lactic acid

(PLLA) and nano-Si. This composite exhibited the initial reversible capacity as high

as 1100 mAh/g, with the Coulombic efficiency of 82.1%. However, the reversible

capacity retention was not that good. After 30 cycles, the capacity decreased from

1100 mAh/g to 600 mAh/g.

Similarly, Q. Si et al. [76] prepared a composite of nano-Si powder with a

pyrolytic carbon and carbon nano-fibers (CNF). The sample was prepared by two

approaches. One was directly mixing of CNF with the nano-Si powder coated with

carbon produced by pyrolysis of polyvinylchloride (PVC). The other is mixing of

CNF, nano-Si powder and PVC with subsequent firing. Both of the samples exhibited

promising cycling performances. A capacity of more than 900 mAh/g was maintained

after 30 cycles for the first sample, and the Coulombic efficiency of the first cycle was

only 53%. The second sample had a higher Coulombic efficiency of 73%. In this

experiment, 20 mg GO, 10 mg SiNWs/CF composite (146% weight increase) 5%

SBR and 5% CMC were mechanically mixed. The mixture was then pressed into

pellet for using as the working electrode.

Si- Graphene Anode:

Si-graphene has attracted a lot interest recently because Si has the highest lithium

insertion capacity and graphene acts as the buffering to accommodate the large

volume change during the insertion and execration of lithium. Compared with carbon

nanotube, carbon black and graphite, graphene can provide better ability to prevent

the agglomeration of Si particles because of its sheet structure [73].

44
 Lee et al. [77] prepared composites of Si nanoparticles dispersed between

graphene sheets, and supported by a 3-D network of graphite. This composite showed

high specific capacity and excellent cycling life. The capacity was as high as 2200

mAh/h after 50 cycles and 1500 mAh/g after 200 cycles, corresponding to the

capacity fade less than 0.5% per cycle. Ji et al. [78] synthetized Si-graphene multiple

layer structured anodes. Initial charge and discharge capacities of 2499 mAh/g and

2217 mAh/g were obtained, respectively, with the Coulombic efficiency of 88.71%.

From the second cycle, this composite delivered a reversible capacity of 2163 mAh/g,

with the capacity retention of 97.6%. The Coulombic efficiency was as high as 96%

after the second cycle.

3.1.2 Nano-Structured Silicon Anodes

Nanoparticle Si Anodes:

It was reported that reducing the particle size could obtain high Li storage

capacity and enhance the cycling performance. For instance, a discharge capacity of

2775 mAh/g and a charge capacity of 2097 mAh/g were obtained by a nano-Si anode

in the first cycle. Also its cycle life was much better than that of bulk Si anode. The

reversible capacity on the first 10th cycle was 1729 mAh/g[79].

Three-Dimensional Porous Si:

Porous Si is a new type of nano-structured material with the three dimensional

(3-D) structure [80]. Because of its 3-D structure, the available volume of the active

material can be much more effectively utilized. Moreover, the energy density will

increase significantly while still stay with high power density. Shin et al. [80]

fabricated porous Si using an electrochemical etching process. The micropores

consisted of one-dimensional channels with pore diameter varying from 1 to 1.5μm

45
and depth extending up to 15 μm [6]. Electrolyte can flow inside of the channels,

facilitating a process producing high energy and power density. Fig 3.2 [81] shows the

morphological of porous Silicon. It was found that with the increasing channel depth

of the porous Si the capacity increased. Electrochemical test on 3-D porous Si

reported by Cho et al [82] showed the capacity of 2820 mAh/g to 2780 mAh/g after

100 cycles at 0.2C rate. Furthermore, the Coulombic efficiency is about 98%. It can

be seen that the capacity retention and initial Coulombic efficiency of the 3-D porous

Si represent a significant improvement over that of the nanosize Si.

Figure 3.2 Top (a, c, e) and cross-sectional views (b, d, f) of as prepared porous Si sample at

different magnifications [6].

46
Si Nanowires:

Si nanowires (SiNWs) are another promising nanostructured anode materials

which can accommodate large strain without pulverization, and provide good

conduction and electronic contact. Firstly, the nanowires allow for better

accommodation of the large volume changes and relax mechanism strain inside [83,

84]. Secondly, all the nanowires can be utilized to contribute to the capacity because

each Si nanowire is electrically connected to the metallic current collector [85].

Thirdly, efficient charge transport is allowed by the one-dimensional (1-D) nanowires

structure.

Chan et al. successfully grew Si nanowires by employing a vapor-liquid-solid

(VLS) method. Initial discharge capacity of 4277 mAh/g, and charge capacity of 3124

mAh/g were reported, corresponding to a Coulombic efficiency of 73%. Figure 3.3 (a)

shows the pristine, unreacted SiNWs have a crystalline structure with the diameter

about 89 nm. After electrochemical cycling reacted with Li, seen figure 3.3 (b), the

diameter of NWs significantly increased to around 141 nm. During the cycling

process, crystalline silicon reacted with Li to form LixSi alloy, and at 10 mV, all of

silicon changed to amorphous Li4.4Si. Later, Cui’s group [86] reported a Si crystalline

– amorphous core – shell with hierarchical structure anode. Figure 3.4 shows the SEM

images of the Si crystalline – amorphous specimen grown on stainless steel substrate

at 485℃. The as- prepared Si nanowires delivered a high reversible discharge

capacity of ~ 1000 mAh/g, good cyclic performance with high Coulombic efficiency

of over 90% after 100 cycles. Also good electrochemical performance at high

discharging – charging rate was observed (6.8 A/g, 20 times of carbon at 1 h rate).

47
 (a) (b)
Figure 3.3 SEM image of pristine SiNWs before (a) and after (b) electrochemical cycling [85].

Figure 3.4 SEM images of Si Crystalline – amorphous on stainless steel substrate at 485℃for (a)

10 min, (b) 20 min and (c) 40 min [86].

Recently, Hu, L et al. [87] deposited Si nanoparticles on the substrate of vertically

grown Si nanowires from an aqueous, binder-free Si nanoparticle ink. The Si

nanowires function as the conductive path to the collector for Si nanoparticles and the

active anode material. Such Si nanoparticles decorated Si nanowires network anode

exhibited excellent electrochemical performance. The reversible discharge capacity

was ~1800 mAh/g, and the areal capacity was greatly increased form ~0.6 mAh/cm2

(pure Si nanowires) to ~2.5 mAh/cm2, and the Coulombic efficiency maintained about

98% after 50 cycles.

Si Nanotubes:

48
 Song et al [88] successfully synthesized silicon nanotubes (SiNTs), shown in

figure 3.5. The special property of the SiNTs is that they have axial void spaces which

provide enough free surfaces to benefit the mechanics, without creating additional

surfaces for accelerated SEI formation. This particular structure makes the electrode

available to endure great mechanics stresses and volume change during the

intercalation of lithium ions. The SiNTs were obtained three-step processes: 1) using

hydrothermal process to grow dense ZnO nanorods on the surface of stainless steel; 2)

chemical vapor deposition (CVD) Si onto the nanorods; and 3) selective removing

ZnO via a high temperature reduction process. The first discharge and charge capacity

is as high as 3860 mAh/g and 3360 mAh/g, respectively, at a rate of 0.05C. The

capacity decreases from 3360 to 1490 when increasing the rate from 0.05C to 2C. It

can be explained by the intrinsic property of Si which is low electron conductivity and

lithium ion diffusivity.

Figure 3.5 SEM image of a vertically aligned SiNTs. [88]

Silicon nanowires (SiNWs) based nanostructured materials, as discussed

previously, are attractive anode candidates for Li-ion batteries because of the high
49
specific surface area and short ion diffusion length. The nanowire structure can be

more adaptable for volume change during charge-discharge. In this part of research

study, in collaboration with University of Dayton, SiNWs were grown on the surface

of carbon nano-fibers utilizing the chemical-vapor-deposition (CVD) method. On the

other hand, graphene can improve the electric contact of the composite material due to

the good electrical conductivity of the carbonaceous materials. Furthermore, graphene

is also a good alternative candidate of anode material, and will contribute to the

overall reversible capacity. In order to further improve the cyclic performance of the

SiNWs/CF composite anode, GO was added working as a buffer to accommodate the

volume change during the lithiation and de-lithiation process. In this study, graphene

is mechanically mixed with SinNW/CF in an attempt to enhance the cycle

performance. The experiments and preliminary characterizations will be presented in

this chapter.

3.2 Experimental Aspects

3.2.1 Silicon Nanowires Growth

The growth of SiNWs on carbon fibers was conducted by Q. Jiang at University

of Dayton [89]. Figure 3.6 schematically shows the CVD of directly growing SiNWs

on carbon nano-fiber with the help of nanoparticle Au catalyst. Hydrogen

tetrachloroaurate (III) hydrate was dissolved in ethanol with the ratio of 1:5 to form

Au solution. The catalyst solution was dipped on the carbon nano-fibers and then

dried at 60℃ for 2 hours. The as-prepared sample was placed in the middle of the

quartz tube which was heated by the three-zone furnace. Argon was firstly purged at

the flow rate of 800 sccm for 30 minutes in order to get rid of air. As the temperature

50
of the furnace reached 700℃, high-purity hydrogen gas and silane gas were supplied

into the tube at the ratio of 1:2. The growth time was set as 10 minutes.

Figure 3.6 Schematic of synthesis SiNWs through CVD method.

3.2.2 Electrode Preparation and Electrochemical Characterizations

For electrochemical test, the substrate with the deposited SiNWs was cut into 0.5

× 0.5cm square. For GO/SiNW/CF electrode, GO and SiNW (146wt%) /CF were

mechanically mixed with the binder (5% SBR and 5% CMC) and pressed into pellet.

Swagelok cell was then prepared in the argon-filled glove box, and contain an

SiNWs as anode, a Li foil cathode, a microporous polyethylene separator and an

electrolyte solution of 1M LiPF6 in ethylene carbonate/dimethy1 carbonate (EC/DMC)

(1:1 vol.%). The cells were galvanostatic charge-discharge at the preset current

density with the voltage window ranging from 0.01 – 3.0 V on a battery testing station

(Land CT).

3.3 Results and Discussion

Jiang [89] has reported the growth mechanism and structural characterization

elsewhere. In short, it was founded that 1) the droplets of Au-Si alloy formed on the

surface of carbon nano-fibers at the initial stage with an average droplet diameter of

~100 nm; 2) both length and diameter of SiNWs gradually increase with prolonging

51
the growth time. ( ~300 nm at 10 min growth time); 3) initially the fine crystalline

structure SiNWs formed but with the time increased amorphous Si deposited on the

outside of the crystalline which results the diameter and length growth.

Morphology and Element Distribution

Figures 3.7 (a-e) show the morphology and the element distribution of SiNWs/CF

samples characterized by using scanning electron microscope (SEM) integrated with

electron dispersive spectroscopic (EDS) analyses. Si, O, Au, and C elemental

mapping were performed on the samples.

Figure 3.7 (a) shows the SEM image of SiNWs/CF sample. From which, it can be

observed that the relative thinner SiNWs with a diameter of ~100 nm homogeneously

grown on the surface of carbon nano-fiber. Figure 3.7 (b) shows the distribution of

carbon element in the same region of the SiNW/CF sample. It confirms the carbon

element dominant in fibers. The randomly distributed carbon element signals

distributed outside the carbon fibers in the image is the effect of the carbon sticker

used for holding the samples for SEM imaging. From figure 3.7 (c) it can be observed

that silicon mainly distributes on the surface of carbon nano-fiber and the nanowires

are basically pure silicon. Because gold is used as the catalyst, the gold widely

distribute on the surface of carbon nano-fiber, as can be seen in Fig. 3.7 (e). In

addition, insignificant amount of oxygen was found due to subsequent exposure the

samples in the air after growth (see figure 3.7 (d)).

52
 (a) (b)

(c) (d)

(e)
Figure 3.7 SEM and EDX mapping images showing morphology and element distribution of

SiNWs/CF.

Electrochemical Characterizations

Figures 3.8 (a) and (b) show the charge-discharge profiles of carbon nano-fiber

(CF) and carbon nano-fiber supported silicon nanowires (SiNWs/CF), respectively.

CF was tested as the baseline for the calculation of the net lithium storage in SiNWs.

53
The first discharge capacity on bare CF was close to 390 mAh/g and the first charge

capacity was 191 mAh/g with the Coulombic efficiency of 48.6%. Seen from figure

3.8 (b), the first discharge capacity (including both CF and SiNWs) of SiNWs/CF

specimen significantly increased to 1540 mAh/g and the first charge capacity was 980

mAh/g giving the Coulombic efficiency of 64%.

It is believed that during the first discharge, crystal Si reacted with Li to form

LixSi. Large irreversible capacity occurred mainly above 0.3 V during the first

discharge. Compare figure 3.8 (a) and (b), the irreversible slop centered around 1.3 V

is the consequence of irreversible lithium reaction and electrolyte reaction with the

functional groups of carbon fiber and native silicon oxide layer on SiNWs. The

following dramatically decreased voltage between 1 V to 0.3 V is due to the formation

of SEI layer formation.

521 CF
3.0

2.5

2.0
Voltage (V)

1.5

1.0

0.5

st
1 Discharge
0.0
0 50 100 150 200 250 300 350 400 450

Capacity (mAh/g)

(a)
5 2 1 SiNWs/CF
3.0

2.5

2.0
Voltage (V)

1.5

1.0

0.5

st
1 Discharge
0.0
0 200 400 600 800 1000 1200 1400 1600

Capacity (mAh/g)

(b)

54
Figure 3.8 Galvanostatic discharge-charge profiles of (a) carbon nano-fibers (CF) and (b) SiNWs

(28wt%)/CF

Based on the capacity results, the galvanostatic discharge and charge capacity of

SiNWs can be calculated leading to the value of 4500 mAh/g and 3800 mAh/g for Si,

respectively. Hatchard et al [90] reported that they employed pure SiNWs as anode

and Li4.4Si was formed at room temperature giving the first discharge capacity of

3579 mAh/g. The obtained reversible charge capacity is very close to the reported

theoretical value. From the second cycle, the capacity dramatically dropped which is

because of the brittle property of the sample and the problem of pulverization. The

capacity fading was also caused by the volume change after 1 cycle performance. The

capacity still maintained over 820 mAh/ after 5 cycles.

3.0

2.5

2.0
Voltage (V)

1.5

1.0

0.5

0.0
0 200 400 600 800 1000 1200 1400
Capacity (mAh/g)

Figure 3.9 Galvanostatic discharge-charge profiles of GO/SiNWs/CF (67%/13%/17%）composite

anode at the current density of 50 µA/cm2.

Figure 3.9 shows the discharge-charge profiles of SiNWs/CF in composite with

GO powders. This mixed composite anode showed the first discharge and charge

capacity of 1367 and 780 mAh/g, respectively, with a Coulombic efficiency of 77%.

From the 2nd cycle, the Coulombic efficiency increased up to 86%. The discharge

capacity of SiNWs in this composite can be calculated leading to the value of 2555

55
mAh/g and 2300 mAh/g, which is well consistent with the theoretical capacity value

of Si. However the cycle performance is not as good as that of SiNWs/CF, after

cycling for 5 cycles the capacity decreased from 1367 mAh/g to 600 mAh/g,

correspondingly the capacity fading per cycle was 11%. Apparently, this experiment is

not optimized resulting in unsatisfactory cycling performance. In the future work, the

weight content of GO and SiNWs/CF as well as binder should be finely adjusted to

achieve better performance.

3.4 Summary

Table 3.2 summarized the galvanostatic charge-discharge results of the CF and

SiNWs/CF as anode. The SiNWs (28wt%) /CF specimen can deliver the discharge

capacity 1540 mAh/g and charge capacity 980 mAh/g in the first cycle, corresponding

to values of 4500 mAh/g and 3800 mAh/g for Si, respectively. The capacity still

maintained over 820 mAh/g after 5 cycles. The sample of SiNWs/CF combined with

GO (weight 67%) showed the first discharge and charge capacity of 1367 and 780

mAh/g, respectively, with the first Coulombic efficiency of 77%. However, the

optimal composition in both cases is yet optimized to achieve better cycling

performances.

Table 3.2 Galvanostatic charge-discharge results of CF, SiNWs/CF, and GNS/SiNWs/CF anodes

Sample Weight 1st 2nd 1st D/C 1st Cycle 2nd Cycle
Content D/C D/C Capacity Coulombic Coulombic
(%) Capacity Capacity (mAh/g) Efficiency Efficiency
(mAh/g) for (mAh/g) for for Si
total total
CF 393/191 223/206 48.6% 92.3%

SiNWs/CF 28/72 1540/985 975/836 4200 64% 85.7%

GO/SiNWs/CF 67/20/13 1367/780 898/818 2555 77% 86%

56
 Nano structured silicon is a promising candidate anode material which provide

high reversible capacity and at the same time maintain the good cyclic performance.

In the future work, graphene can be added to the SiNWs forming the

SiNWs/Graphene composite. In this composite graphene can provide short conductive

path and worked as a good “buffer” to accommodate the volume change during the

lithiation and de-lithiation. Furthermore, more elaborate electrochemical impedance

studies are needed to understand the chemical reactions and kinetics of electrode

process occurred on the charge – discharge process.

57
Chapter 4: Graphene-based Supercapacitors

4.1 Introduction

In capacitors, different from batteries, charges will accumulate on the interface of

the electrode and the electrolyte. The capacitance of a capacitor is directly related

with the interface, and can be expressed as the following:

𝐴𝜀
𝐶= 4-1
4𝜋𝑑

where A is the area of the electrode surface, d is the distance between the electrodes.

In the supercapacitors A represents the active surface of the electrode porous layer, 𝜀

is the medium dielectric constant, d is the thickness of the electrical double layer [35].

From this point of view, choosing a material with high surface area and good

electrical conductivity as the electrode material is especially important for a

supercapacitor. Graphene, a two-dimension nanostructure material, in addition to the

amazing properties above, can provide a large active surface area for fast transport of

ions to achieve high double-layer capacitance. On the other hand, MnOx is considered

as one of the promising electrode materials for aqueous electrolyte supercapacitors

owing to is low-cost, abundant on earth, and environmental friendly. However, due to

the low specific surface area of MnO2, the specific capacitance is very low.

Furthermore, although by preparing nanostructured MnOx powders to increase the

specific surface area, the structure is not stable and will be easily changed resulting in

the poor cyclic performance. In addition, the low electrical conductivity will severe

58
affect their specific capacitance [91]. Given this situation, the MnOx is combined with

graphene which will both provide the large specific surface area high electrical

conductivity, and the supercapacitors using graphene-MnOx as electrode material can

be benefited from EDL mechanism and the pseudo-mechanism. In this part of

research study, three different types of graphene, synthesized utilizing different

methods and their composites with MnOx are employed as electrodes of electrical

double-layer supercapacitors (EDLS).

Supercapacitors are suffering the main limitation that the power density is

significantly lower than that of the other energy storage devices. How to increase the

power density of supercapacitors is another important issue should be taken into

consideration. The max power of a supercapacitor can be expressed in equation 4-2:

𝑉2
𝑃= 4-2
4𝑅𝑀

where V is the operating voltage of the supercapacitor; R is the equivalent series

resistance, which is composed from the internal components of the supercapacitor, for

example electrodes, current collector, and electrolytes; M is the total mass of the

electrodes. From equation (4 - 2) it can be easily deduced that the max power of a

supercapacitor is primarily limited by the operating voltage V and ESR. Electrolyte is

the mainly limitation of the operating voltage. Compared with aqueous electrolyte

(voltage window is about 1.2 V), ionic liquids (ILs) can provide a wide voltage

window ranging from 2 – 6 V and conductivity about 10 mS/cm [52] which is an

effective method to overcome its main drawback. In this research study, two types of

ionic liquids 1-ethyl-3-methylimidazolium tetrafluoroborate (EMIMBF4) and

1-methyl-3-octylimidazolium tetrafluoroborate (C8mimBF4) and their mixtures at

various volumetric ratios are tested as electrolytes, and their electrochemical

characteristics are compared.

59
4.2 Research Status of Graphene-Based Electrodes

4.2.1 Graphene Nanosheets:

Graphene oxide suspension reduced by hydrazine hydrate is the simplest but

useful method to prepare reduced graphene materials. Stoller and co-workers [55]

prepared the reduced graphene electrodes materials through this method firstly. After

the reduction procedure, the graphene sheets agglomerated into particles which was

around 15 – 25 μm in diameter, and the specific surface area measured by N2

absorption Brunauer—Emmett—Teller (BET) method was 705 m2/g. The specific

capacitances of 135 F/g and 99 F/g were obtained with KOH and TEABF4 in

acetonitrile as electrolyte, respectively. It is pointed that while the graphene oxide

dispersed in aqueous solution, irreversibly agglomerates will occur [57]. Chen and

co-works [92] employed the gas-based hydrazine to reduce the solid-state graphene

oxide instead of dispersing the graphene oxide in aqueous solutions. The

agglomeration of the graphene sheets was effectively reduced. The accessibility of the

ions into the reduced graphene electrodes was enhanced, not only the outer region of

the solids but also the inner region compared the convention reduced by the hydrazine

solution. Thus the large specific capacitance of 205 F/g with the power density of 10

kW/kg was achieved in an aqueous electrolyte.

Graphene also can be obtained through the thermal reduction of graphene oxide

method. Rao et al. [93] reduced the graphene oxide at 1050℃, and the specific

surface area was as high as 925 m2/g. Accordingly the specific capacitance was 117

F/g in H2SO4 electrolyte and 75 F/g in PYR14TFSI electrolyte. The drawback of this

method is that the high-temperature exfoliation process is energy consuming and

difficult to control.

60
 Yang el al. [94] introduced a high vacuum environment thus lower the reduction

temperature of graphene oxide to around 200℃. The vacuum environment offered an

outward driving force which will promote the fast exfoliation of graphene layers, and

results in an effectively exfoliation of graphene layers. With the discharge current

density of 100 mA/g, the capacitance for aqueous and organic electrolyte system were

264 F/g and 122 F/g, respectively. These values are much higher than that of high

temperature reduced samples [93].

Cao and co-works [95] reported that the graphene oxide was reduced in air

similar as the low-temperature thermal exfoliation approach. At the current density up

to 1 A/g with KOH electrolyte, the specific capacitance was 232 F/g. This high

capacitance was attributed from both the electrical double layer mechanisms from

graphene sheets and pseudo-capacitance from the oxygen groups on the surface of

graphene sheets. According to the authors, although the capacitance can be a little bit

increased, the pseudo-capacitance is still unwanted because this will have a negative

effect on the stability of the capacitors.

Ruoffet al. [56] reported that graphene oxide could be exfoliated and well

dispersed in propylene carbonate (PC) solvent with the sonication treatment. The

suspension was heated at 150℃ to remove oxygen function group on the surface of

graphene oxide. The conductivity of the reduced graphene prepared through this

method was as high as 5230 S/m. Supercapacitor with this kind of reduced graphene

as electrode material and EMIM-NTF2 as electrolyte showed the capacitance of 187

F/g, energy density 6.5 Wh/kg, and power density 2.4 kW/kg with the voltage window

of 3 V.

61
4.2.2 Graphene-based Composites:

After reducing the graphene oxide, due to the loss of oxygen-group between the

graphene sheets, the electrostatic repulsions would decrease which will cause the

reduced graphene sheets to agglomerate. Thus this will greatly limit the accessibility

of ions into the surface of reduced graphene sheets, and lower the electrochemical

performance of supercapacitors. From this point of view, many reports have

introduced the so called “stabilizer” or “spacers” into the graphene layers to inhibit

the agglomeration of reduced graphene sheets.

Si et al.[96, 97] prepared a platinum nanoparticles intercalated graphene electrode

material which effective inhibited the agglomeration of graphene sheets yielding a

high specific surface area about 862 m2/g. While compared with the dried graphene

with the capacitance of 14 F/g, the Pt-graphene hybrid significantly increased the

capacitance up to 269 F/g. Figure 4.1shows graphene sheets and

nanoparticle-modified graphene sheets in its dispersion and dry state.

Figure 4.1 Schematics of a) graphene sheets and b) nanoparticle-modified graphene sheets in its

dispersion and dry state [96].

62
 Fan and co-works [98] employed carbon black as “spacers” to inhibit the

agglomeration of graphene sheets and thus the accessibility of ions into the graphene

layers. The carbon black particles were deposited on the edge of graphene sheets that

help the diffusion and migration of electrolyte ions easier. Correspondingly, the

obtained specific capacitance was 175 F/g at the scan rate of 10 mV/s in 6 M KOH

electrolyte. Even the scan rate increased up to 500 mV/s, the capacitance still

remained at 118 F/g.

Zhang et al. [99] prepared a series of surfactant-stabilized graphene materials by

intercalating graphene oxide with different surfactants including tetrabutylammonium

hydroxide (TBAOH), cetyltrimethylammonium bromide (CTAB) and sodium

dodecylbenzenesulfonate (SDBS), then reduced by hydrazine. The researchers

pointed that the wettability of such graphene materials was greatly enhanced, thus the

performance as electrode of supercapacitor was improved. Among these three

electrode materials, the one intercalated by the TBAOH had the best performance.

The specific capacitance of that was 194 F/g in 2 M H2SO4 electrolyte at the current

density of 1 A/g.

Recently, Wei et al.[92] successfully synthesized graphene through chemical

vapor deposition method. They grew nanomeshes morphology graphene on the

template porous MnO layers. By this method, only one or two layers graphene sheets

was produced and with the specific surface area as high as 1654 m2/g. Unlike many

other methods the graphene was produced by only a small amount at one time, they

could synthesis nanomesh graphene at gram-scale one time. Correspondingly, the

specific capacitance tested by cyclic voltammetry was 255 F/g at the scan rate of 10

mV/s in 6 M KOH electrolyte solution. In addition, with the increasing of scan rate

from 10 mV/s to 500 mV/s, only 21% specific capacitance lose.

63
 Jang and co-works [100] reported that they had successfully synthesized a

mesoporous graphene which has very good accessibility for the ions of ionic liquids

electrolytes. They first synthesized the graphene oxide using the modified Hummer’s

method, and then they injected the suspension into a forced conventional oven

providing a fludized-bed situation to obtain the curved graphene sheets. This method

effectively avoided the agglomeration and prevented the restacking of the graphene

sheets with one another when packed or compressed into an electrode structure. With

this curved-graphene as electrode and ionic liquid as electrolyte, the capacitance of

100 – 250 F/g with a voltage window of 4.0 V as high current density of 1 A/g was

achieved, and the energy density was as high as 85 Wh/kg at room temperature.

4.2.3 Graphene Oxide -MnOx Nanocomposites:

Wang et al. [101] prepared GO – MnO2 nanocomposite via a straightforward

approach to deposit MnO2 nanoparticles on the surface of the graphene oxide sheets

in water/isopropanol solution. The formation mechanism that Mn ions intercalated

and adsorbed onto the surface of GO sheets, followed by the nucleation and growth of

the crystal species in the double solvent system which in turn exfoliated the graphene

oxide sheets. It was confirmed that the -MnO2 could successfully attached to both

sides of the GO sheets due to the function groups, such as hydroxyland epoxy groups

attached on both sides of the GO sheets. The specific capacitance of GO – MnO2

nanocomposite was 216, 197.2, 141.5 and 111.1 F/g calculated at the current density

of 150, 200, 500 and 1000 mA/g. The specific capacitance of nano – MnO2 was 211.2

F/g calculated at a current density of 200 mA/g. However, the specific capacitance

retention of nano – MnO2 was only 69% (145.7 F/g) which was lower than that of GO

– MnO2 of 84.1% (165.9 F/g). Compared GO – MnO2 and nano – MnO2 electrodes, it

64
was confirmed that the specific capacitance of GO – MnO2 was contributed from both

electrical double layer mechanism and pseudo-capacitive mechanism. Although the

specific capacitance of GO – MnO2 was slightly lower than that of nano – MnO2, the

electrochemical stability was obviously enhanced.

Fan and co-workers [91]also synthesized graphene – MnO2 nanocomposite by an

easy deposition of nanoscale MnO2 on the surface of graphene sheets under

microwave irradiation. According to their report, the highest specific capacitance of

310 F/g was achieved with 78% wt MnO2 content at 2 mV/s scan rate in 1 M Na2SO4

electrolyte. As the scan rate increased to 100 and 500 mV/s, the specific capacitance

retention ratio still maintained 88% and 74%, respectively. Such promising

electrochemical performance was attributed to the following reasons. First of all, the

coated MnO2 piled up to form porous which significantly improved the Na ions

diffusion rate in the bulk materials. Furthermore, the nanoscale MnO2 reduced the

diffusion length of Na ions and enhanced the utilization of MnO2. In addition,

graphene in the nanocomposite not only acted as the supporters for the deposition of

MnO2 nanoparticles but also provided short electrical conductive channels for Na ions

diffusion.

Li et al. [102] reported that flexible graphene– MnO2 composite papers were

fabricated through a simple chemical method. According to their report, the

supercapacitor with 24% wt MnO2 showed higher specific capacitance (256 F/g, at

current density of 500 mA/g) and better cycle stability (74% after 400 cycles) than the

one with 12% wt MnO2 component (141 F/g at the current density of 500 mA/g and

52% after 400 cycles). This was ascribed to the fact that high content MnO2

contributed more capacitance, the graphene – MnO2 with a wall-packed layered

structure not only prevents the agglomeration of the GS but also provided an easy

65
access of the surface to the electrolyte. Table 4.1 summarized the current research

status on various graphene based electrode materials for supercapacitors.

Table 4.1 Summary of Current Research Status on Various Graphene Based Electrode Materials

for Supercapacitors

Basic Synthesis info Specific Electrolyte Calculate Reference

Sample Capacitance method
(F/g)
GO reduced by 135 KOH GDC 10 mA [55]
hydrazine 99 TEABF4 in acetone CV 20 mV/s
GO reduced by 205 Aqueous electrolyte [92]
gas-based hydrazine
RGO GO reduced at 1050oC 117 H2SO4 CV 5 mV/s [93]
75 PYR14TFSI
GO reduced at 200oCin 264 Aqueous electrolyte GCD 0 – 1V [94]
vacuum 122 Organic electrolyte 100 mA/g
GO reduced at 200oCin 232 KOH CV 5 mV/s [95]
air
GO in PC, reduced at 187 EMIM-NTF2 GCD 0-3V [56]
150oC 1 A/g
Pt-Graphene Platinum intercalate in 269 Aqueous electrolyte CV 5 mV/s [96]
graphene sheet
Carbon Carbon black deposit on 175 6 M KOH CV 10 mV/s [98]
black-Graphene graphene sheets
Surfactant-Stabilized TBAOH intercalate in [99]
Graphene graphene, reduced by 194 2 M H2SO4 GCD 1 A/g
hydrazine
Curved-Graphene Apply compressed 154 EMIMBF4 GCD 1 A/g [100]
airstream
deposit 216 1 M Na2SO4 GCD 0.15 [101]
MnO2nanoparticles in A/g
water
GO-MnOx Deposit MnO2 under 310 1 M Na2SO4 CV 2 mV/s [91]
microwave irradiation
Fabricate GO-MnO2 256 1 M Na2SO4 GCD 0.5 [102]
paper layer by layer A/g

4.3 Experimental Aspects

4.3.1 Synthesis of Active Electrode Materials

GO is synthesized by Modified Hummer’s method, and then RG is obtained by a

series of chemical reduction process. Details have been described in chapter 2.4.1.

Reduced graphene oxide (RGO) is obtained from GO followed by several reduction

steps. Firstly, 2 M sodium hydroxide solution was added into the 1 L GO suspension

(1 g/L) until the pH value reached to 10. Then 0.2 mL hydrazine monohydrate was

66
added, and the mixture was heated at 95℃ for 2 hours. Then the sample was filtered

and dried in the vacuum oven, and RGO was obtained.

Restacking of graphene nanosheets has hampered the performance of

supercapacitors using graphene materials as electrodes. The key issue of increasing

the performance of graphene-based materials electrodes is to prevent the restacking

and agglomeration of graphene sheets during the electrodes fabrication [92, 100].

During the process of synthesis GO, a concentrate air stream has been employed to

obtain curved graphene (CRG) sheets with improved surface area.

In addition, graphene-MnOx nanocomposites with different types of graphene (GO,

RG, CRG) and different phases of MnOx (Mn3O4 and MnO2) were synthesized and

employed as electrode materials for supercapacitors. Synthesis procedure of

GO-Mn3O4 was described in chapter 2. GO-MnO2composites were obtained through

the following steps. Firstly, 40 mg GO (or CRG) were placed in the 2 mL KMnO4

(0.05 M) aqueous solution and added 10 mL deionized water. The suspension solution

was under ultrasonic treatment for 30 min. Then 2 mL citric acid (0.05 M) was added

and continuously stirred for 15 min. The dispersion was refluxed at 80℃ for 7 hours

and dried in air. Finally, the as prepared nanocomposite was annealed at 400℃ for 2

hours in H2/Ar (5%-95%) atmosphere. Figure 4.2 presents the fabrication process of

GNS/MnO2. The MnO2 nanoparticles were homogeneously dispersed on the GNS

surface, and the oxygen function groups on the surface of GO were reduced by the

thermal reduction at 400℃. Hence, a 3-D sandwich-like nanostructure supported by

MnO2 nanoparticles was formed. The sample RG/MnO2 (CRG/MnO2) was obtained

and tested as the electrode material. The electrode samples with their acronym and

description are listed in Table 4.2.

67
 Figure 4.2 Schematic illustration of fabrication process of RG/MnO2..

Table 4.2 List of Graphene and Graphene-MnOx Samples Used in This Study

S. NO Acronym Description

1 GO Graphene Oxide synthesized by modified hummer’s method.

2 RG Reduced Graphene Oxide with hydrazine solution.
3 CRG Dried in fludized-bed to form “Curved” graphene.
4 CRG-Mn3O4 Curved graphene and Mn3O4 nanocomposite
5 CRG-MnO2 Curved graphene and MnO2 nanocomposite
6 GO-Mn3O4 Graphene oxide and Mn3O4 nanocomposite
7 RG-MnO2 Reduced graphene and MnO2 nanocomposite

4.3.2 Electrode Preparation

The mass of the active material and the thickness of the electrode influence

significantly on the testing results [103]. In order to increase the credibility of the

experiment data the electrodes were prepared by pellet pressing instead of coating.

The electrode material was composed of 80% active electrode materials (listed in

table 4.2), 5% styrene butadiene rubber (SBR), 5% carboxymethylcellulose (CMC),

and 10% acetylene black. The SBR and CMC were working together as the binder,

and the acetylene black was used to enhance the conductivity of the electrodes

materials. The electrode materials was completely dried in the oven for 12 hours at

120℃, and pressed at 1000 psi in a special model. The geometrical area of the

68
pelleted electrode is around 0.8 cm2, correspondingly, the specific surface area of the

electrode is 400 cm2/g. However, due to the porous structure existed in the electrode

sample, the active surface area didn’t decrease much compared with the original

specific surface area of the sample about 1200 cm2/g.

4.3.3 Electrolytes

The two ionic liquids 1-ethyl-3-methylimidazolium tetrafluoroborate (EMIMBF4)

and 1-methyl-3-octylimidazolium tetrafluoroborate(C8mimBF4) were selected in this

study. There structure schematics are shown in Fig 4.3, and the physical properties are

summarized in Table 4.3. The two types ILs were also mixed at different volumetric

ratio and their correlation with electrochemical performances were studied. The two

ILs were placed in the glove box, and used two different syringes with the same size

to take the ILs in volume. At last, the two ILs were mixed in one baker and stirred

thoroughly with a glass stick. .

(a) (b)
(c)

Figure 4.3 Figure 4.3 Structure schematics of (a) [EMIM] +, (b) [C8mim] +, (c) [BF4].

Table 4.3 Physical properties of EMIMBF4 and C8mimBF4

Ionic Liquid Cation Diameter Anion Diameter Viscosity Conductivity

(nm) (nm)

[EMIM]+[BF4]- 0.79 0.132 9.78 31.1

[C8mim]+[BF4]- 1.08 0.132 336 10

69
 For aqueous electrolyte, 0.1 mole NaOH was gradually added to 0.05 M (50 mL)

H2SO4 solution, and 1 M Na2SO4 aqueous solution was obtained after the reaction

completed. The as-prepared aqueous electrolyte was used with GO-MnOx electrode

for electrochemical characterizations.

4.3.4 Electrochemical Characterizations

Swagelok cells, consisting of two same electrodes sandwiched a separator soaked

with the electrolyte, were assembled in the argon-filled glove box. After assembling

the cells in the glove box, they were laid up for 24 hours before tested. The main

purpose was to make sure the electrolytes will penetrate through the separator and

completely soak in the electrodes materials.

Cyclic voltammetry was obtained at the preset scan rate ranging from 5 mV/s to

500 mV/s with the voltage window varying with different electrolytes on Camry

electrochemical analysis system. The cells were also galvanostatically

charged/discharged at the preset current density with the voltage window ranging

from 0V up to 4.0 V on a battery testing station (Land CT). Electrochemical

impedance spectra (EIS) was obtained by applying a sine-wave signal with an

amplitude of 10 mV in the frequency range of 500kHz to 0.1 Hz on Camry

electrochemical analysis system after the cell was assembled and stood for 24 hours.

For the CV tests, the specific capacitance values were calculated using the

formula:
𝒇𝒊𝒏𝒂𝒍
𝟒 ∫𝒊𝒏𝒊𝒕𝒊𝒂𝒍 𝒊𝒅𝑽
Csp= 4-3
𝒖𝑽𝑴

where u, V, M are the scan rate (V/s), operating voltage (V), total mass of the active

material in both of the electrodes (g), respectively. For the galvanostatic

charge-discharge tests, the specific capacitance values were calculated from the

formula:
70
 𝟒𝑰𝜟𝒕
Csp=𝒎𝜟𝑽 4-4

Where Csp, I, Δt, ΔV, m are the specific capacitance (F/g), the current loaded (A),

the discharge time (s), the operating voltage (V) and the mass of active material in

the electrodes (g), respectively.

4.4 Results and Discussion

4.4.1 Cyclic Voltammetric Characteristics of GO and RGO in EMIMBF4

Electrolyte

0.08
0-2.5 V 20mV/s 0-2.5 V 50mV/s
0-3 V 20mV/s 0-3 V 50mV/s
0.15
0-3.5 V 20mV/s 0.06 0-3.5 V 50mV/s

0.10
Current Density (A/g)

Current Density (A/g)

0.04

0.05 0.02

0.00 0.00

-0.05 -0.02

-0.10 -0.04

-0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Voltage (V) Voltage (V)

(a) (b)

Figure 4.4 Cyclic voltammetric characteristics of GO (a) and RGO (b) electrode in EMIMBF4

electrolyte at the scan rate of 20 mV/s with the voltage upper window ranging from 2.5 V to 3.5

V.

Figure 4.4 (a) and (b) shows the CV curves for GO and RGO as electrode

material in EMIMBF4 electrolyte, respectively. For GO, as the voltage window ran

from 2.5 V up to 3.5 V, the corresponding capacitance increased from 5 F/g to 13 F/g.

Both of the CV curves show the relative rectangular shape which indicates the EDL

mechanism of the capacitor. While the GO was chemically reduced to RGO it seems

didn’t improve too much. The specific capacitance of RGO was only about 2 F/g to 5

F/g. This phenomenon can be explained from two aspects. Firstly, during the
71
synthesizing process of the GO and RGO, the nanosheets re-stacked dramatically

which caused the specific surface area of these materials was very low. Secondly, in

this experiment, EMIMBF4 was employed as the electrolyte. The diameter of the

cations and anions are much larger than the other types of electrolytes. Also the

relative high viscosity and low conductivity further prohibited the ions from

infiltrating into the electrode materials.

4.4.2 Cyclic Voltammetric Characteristics of CRG in EMIMBF4 Electrolyte

Figure 4.5 (a) and (b) show the CV curves of CRG electrode in EMIMBF4 at the

scan rate of 20 mV/s and 50 mV/s with the potential window ranging from 2.5 V to 4

V. Both of the CV curves exhibits approximately rectangular shape, which indicates

the CRG electrode in EMIMBF4 as electrolyte mainly possess electrical double layer

mechanism [55]. At the scan rate of 20 mV/s, the specific capacitance increased from

67 F/g to 143 F/g as the voltage window increasing from 2.5 V to 4 V. Even when the

scan rate raise up to 50 mV/s, the specific capacitance still increased from 55 F/g to

109 F/g with the voltage window increasing from 2.5 V to 4 V. As the voltage run up

to 4 V, an obvious sharp peak occurs, and also another peak appeared around the

voltage of 1.5 V. This means with the increasing voltage window, the specific

capacitance was not only contributed by the EDL mechanism, but also the pseudo

capacitance.

Although the pseudo capacitance will slightly increase the specific capacitance of

the supercapacitor, it is still unwanted because this may be the “unstable” factor in the

capacitor and will decrease the cycle life.

72
 0-2.5 V 20mV/s
2.0 0-3 V 20mV/s
C = 143 F/g
0-3.5 V 20mV/s sp
0-4 V 20mV/s
1.5
C = 107 F/g
sp

Current Density (A/g)

1.0 C
sp
= 82 F/g

0.5
C = 67 F/g
sp
0.0

-0.5

-1.0

-0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5

Voltage (V)
(a)
4.0 0-2.5 V 50mV/s
3.5 0-3 V 50mV/s
0-3.5 V 50mV/s Csp = 109 F/g
3.0
0-4 V 50mV/s
2.5
C = 87 F/g
Current Density (A/g)

sp
2.0
C = 67 F/g
sp
1.5

1.0
C = 55 F/g
0.5 sp

0.0

-0.5

-1.0

-1.5

-2.0

-0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5

Voltage (V)
(b)

Figure 4.5 Cyclic voltammetric characteristics of CRG electrode in EMIMBF4 electrolyte at the san

rate of (a) 20 mV/s and (b) 50 mV/s with the voltage upper window from 2.5 V to 4 V.

Figure 4.6 shows that even at relative high scanning rate (e.g. 500 mV/s),in the

voltage window of 0 – 4 V, the CV curves still remained relative close to the

rectangular shape suggesting the fast switching behavior of the ions at the CRG

electrode /EMIMBF4 electrolyte interfacial contacts. The specific capacitance

decreased dramatically from 101.3 F/g to 31.6 F/g. It is noteworthy that CV curves

became distorted gradually with increasing the scanning rate to 500 mV/s. This can be

73
attributed to the time constant of the system where the transient response falls off

exponentially with the RC time constant [104]. When increasing the scan rate, the

concentration gradient at the electrode/electrolyte interface will increase and the time

constant will become larger. As a consequence, the transient response is pronounced

leading to a lag between charging and discharging of the capacitor causing distortion

of the CV[105].

10
20mV/s
8 50mV/s
100mV/s
6 250mV/s
500mV/s
Current Density (A/g)

-2

-4

-6

-8
-0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Voltage (V)

Figure 4.6 Cyclic voltammetric characteristics of CRG electrode in EMIMBF4 electrolyte at the

different scan rates, i.e. 20 mV/s, 50 mV/s, 100 mV/s, 250 mV/s, and 500 mV/s, with the voltage

window of 0 – 4 V.

The galvanostatic charge-discharge tests were also performed to further

investigate the supercapacitive performance of the CRG electrode. Figure 4.7 shows

the galvanostatic charge-discharge curves for at the same current density (0.5 A/g) but

with different high voltage window ranging from 2.5 V – 3.5 V. Figure 4.8 shows the

galvanostatic charge-discharge curve for CRG electrode in EMIMBF4 electrolyte

within the same voltage window range (0-3.5V) at two different current densities, i.e.

0.5 A/g and 1 A/g.

74
 EMIMBF4 0.5 A/g 0 - 2.5 V
3.5 EMIMBF4 0.5 A/g 0 - 3 V
EMIMBF4 0.5 A/g 0 - 3.5 V

3.0

2.5

Voltage (V)
2.0

1.5

1.0

0.5

0.0
-350 -300 -250 -200 -150 -100 -50 0 50 100 150 200 250
Time (S)

Figure 4.7 Galvanostatic charge-discharge curves of CRG electrode in EMIMBF4 electrolyte at the

current density of 0.5 A/g with the upper voltage window ranging from 2.5 V – 3.5 V.

4.0
0.5 A/g
3.5 1 A/g

3.0

2.5
Voltage (V)

2.0

1.5

1.0

0.5

0.0

0 100 200 300 400 500

Time (S)

Figure 4.8 Galvanostatic charge-discharge curve of CRG electrode in EMIMBF4 electrolyte at the

current densities of 0.5 A/g and 1 A/g between 0 and 3.5V.

Both the discharge curves and the charge curves exhibit high symmetry which

indicates the good capacitive behavior. The initial portions of the discharge curves

show the “IR drop”, which represents the internal resistance of the supercapacitor. It

is well known that the internal resistance is the sum of the resistance of the electrolyte

75
solution, the contact resistance of the interface active material and current collector,

and the intrinsic resistance of the active material. The resistance of the electrolyte

solution is always the predominate factor [22]. The conductivity and the viscosity are

two important parameters of electrolytes, and will affect the resistance of the

electrolyte a lot. From figure 4.7 it can be seen that with the increasing of voltage

window, the “IR drop” also increased. Thus minimal additional energy can be stored

at higher voltages compared with lower voltages. Seen from figure 4.8, the specific

capacitance of the supercapacitor was 109 F/g at the current density of 0.5 A/g within

the voltage window of 0 – 3.5 V. Even at the high current density of 1 A/g, the

supercapacitor still maintained the specific capacitance as high as 89 F/g.

100
80

95 70

Specific Capacitance (F/g)

60
Coulombic Efficiency (%)

90
50
Coulombic Efficiency (%)
85 Specific Capacitance (F/g) 40

30
80

20

75
10

70 0
0 200 400 600 800 1000
Cycle Number

Figure 4.9 Coulombic efficiency and cyclic performance of CRG electrode in EMIMBF4 electrolyte

at current density of 1 A/g.

Cycle life test experiment of CRG electrode in EMIMBF4 electrolyte was

measured by the galvanostatic charge-discharge at the current density as high as 1 A/g

with the voltage window 0 – 3.5 V for 1000 cycles. After 1000 cycles, the coloumbic

76
efficiency calculated from its charge time and discharge time still remained 95% as

shown in Fig 4.9. At the first 10 cycles the capacity retention increased from 89% to

96%. This might because of the activation of the electrode material, and the electrode

material soaked in the EMIMBF4 electrolyte thoroughly. The specific capacitance of

CRG electrode was found to gradually decrease from 80 F/g to 54 F/g, and the

specific capacitance fade per cycle was only 2.6%. All these above indicate the good

cyclic performance of this supercapacitor.

4.4.3 Cyclic Voltammetric Characteristics of CRG in C8mimBF4

Similar to those in EMIMBF4 electrolyte, CV curves of CRG electrode in

C8mimBF4 are rectangular shape at the scan rate of 20 mV/s and 50 mV/s as shown in

figure 4.10 (a) and (b). This indicates the EDL mechanism. However, it can be clearly

observed that as the voltage window runs up to 3.0 V and above, sharp peaks occur

during charge and a corresponding discharge peak emerges around 0.5 V. This

phenomenon indicates that although the stable potential window with C8mimBF4

electrolyte is claimed to reach up to 4 V, there is still some processes (may be

involved in some redox reactions) occurred at the interface between CRG electrode

and C8mimBF4 electrolyte. Firstly, the “ionic absent effect” is usually observed when

the local voltage is raised too high even though the voltage window still doesn’t reach

the decomposition voltage of the electrolyte [56, 106]. Secondly, the ionic liquid has

strong polarity and it is very difficult to completely remove the water in it which may

also cause the partial decomposition of the ionic liquid electrolyte and affect the upper

voltage window [99]. Since this process is partially reversible, the capacitance of the

capacitor is not only contributed by the electrical double-layer mechanism but also the

77
pseudo mechanism. It is worthy to note that this redox process will reduce the

capacitance upon charge/discharge cycling.

1.5
0-2.5 V 20mV/s Csp = 37
0-3 V 20mV/s 47
0-3.5 V 20mV/s 64
1.0 0-4 V 20mV/s 74 F/g
Current Density (A/g)

0.5

0.0

-0.5

-0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5

Voltage (V)

(a)
3.0
0-2.5 V 50mV/s Csp = 28
2.5 0-3 V 50mV/s 32
0-3.5 V 50mV/s 44
2.0 0-4 V 50mV/s 60 F/g
Current Density (A/g)

1.5

1.0

0.5

0.0

-0.5

-1.0

-1.5
-0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5

Voltage (V)

(b)

Figure 4.10 Cyclic voltammetric characteristics of CRG electrode in C8mimBF4 at the scan rate of

(a) 20 mV/s and (b) 50 mV/s with the upper voltage window ranging from 2.5 V – 4 V.

Figure 4.11 shows the galvanostatic charge-discharge curves for CRG electrode

in C8mimBF4 electrolyte at the current density of 0.5 A/g and 1 A/g with the voltage

78
window ranging from 0 – 3.5 V. Compared with figure 4.8, the discharge and charge

profile becomes asymmetrical and the “IR” drop is much higher.

4.0
0.5 A/g
3.5 1 A/g

3.0

2.5
Voltage (V)

2.0

1.5

1.0

0.5

0.0

0 100 200 300 400 500 600

Time (S)

Figure 4.11 Galvanostatic charge-discharge curves of CRG electrode in C8mimBF4 electrolyte at

the current density of 0.5 A/g and 1 A/g with the voltage window of 0 – 3.5 V.

4.4.4 Electrochemical Impedance Spectroscopy Analys

The EIS Nyquist plots were utilized to analyze the electrolyte resistance,

electrode impedance and the diffusion procedure of the electrolyte ions. The more

vertical of the tail at low frequency the more cell responds closely to an ideal

capacitor [55].Figure 4.12 (a) and (b) shows the Nyquist plot of CRG electrode in

EMIMBF4 electrolyte and C8mimBF4 electrolyte, respectively. It can be seen that the

tail at low frequency region exhibits a 45°slope line instead of ideal capacitive

vertical line. This 45°portion of curve is referred to Warburg resistance correlating

with the ion diffusion in the electrode [55]. The dominant Warburg resistance is

because of the relative low conductivity and high viscosity of the ionic liquid

electrolytes.

79
 At the high frequency region, electrolyte resistance can be determined from the

x-intercept of the spectra. The equivalent series resistances of the capacitor in

EMIMBF4 electrolyte and C8mimBF4 electrolyte are 15 Ω and 25 Ω, respectively.

This analysis corroborated well with the IR drop observed in the charge/discharge

profiles. The conductivity and the viscosity are two important parameters of

electrolytes and will affect the resistance of the electrolyte. Because the [C 8mim]+

cation has a much longer molecular chain than [EMIM]+ cation, the former one offers

the higher electrostatic attraction and hampered the mobility of ions. As a result, the

resistance and the “IR drop” of the supercapacitor with C8mimBF4 electrolyte is much

higher than that with EMIMBF4.

（a）

(b)
Figure 4.12 Nyquist plots of CRG electrode in EMIMBF4 electrolyte (a) and C8mimBF4 electrolyte

(b) before charge/discharge process.

80
4.4.5 Ionic Liquid Electrolytes Mixed at Different Ratio

The two types of ionic liquids, EMIMBF4 and C8mimBF4, are mixed together to

test the effects of different ratio of two ionic liquids on the electrochemical

performance of the capacitor. Figure 4.13 shows the CV curves of the CRG in mixed

with a different ratio at the scan rate of 20 mV/s with the upper voltage windows set

to 2.5 V, 3.0Vor 3.5 V.

0% 2.5 V 60% 2.5 V

0.8 10% 2.5 V 0.8 70% 2.5 V
20% 2.5 V 80% 2.5 V
30% 2.5 V 90% 2.5 V
0.6 0.6
40% 2.5 V 100% 2.5 V
50% 2.5 V
0.4 0.4

Current Density (A/g)

Current Density (A/g)

0.2 0.2

0.0 0.0

-0.2 -0.2

-0.4 -0.4

-0.6 -0.6

0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.5 1.0 1.5 2.0 2.5
Voltage (V) Voltage (V)

(a) (b)
0% 3 V 60% 3 V
1.0 1.0
10% 3 V 70% 3 V
20% 3 V 80% 3 V
0.8 30% 3 V
0.8 90% 3 V
40% 3 V 100% 3 V
0.6 50% 3 V 0.6
Current Density (A/g)

Current Density (A/g)

0.4 0.4

0.2 0.2

0.0 0.0

-0.2 -0.2

-0.4 -0.4

-0.6 -0.6

0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0
Voltage (V) Voltage (V)

(c) (d)
2.0 2.0
0% 3.5 V 60% 3.5 V
20% 3.5 V 70% 3.5 V
1.5 30% 3.5 V 1.5 80% 3.5 V
40% 3.5 V 90% 3.5 V
50% 3.5 V 100% 3.5 V
Current Density (A/g)

Current Density (A/g)

1.0 1.0

0.5 0.5

0.0 0.0

-0.5 -0.5

-1.0 -1.0
-0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Voltage (V) Voltage (V)

(e) (f)

81
Figure 4.13 Cyclic voltammetric characteristics of CRG electrode in the mixing EMIMBF4 and

C8mimBF4 electrolyte with different volumetric ratio at the scan rate of 20 mV/s with the upper

voltage window at 2.5 V (a,b), 3.0V(c,d) and 3.5 V (e,f).

Figure 4.13 compiles the specific capacitances of supercapacitors made up of

electrolyte with different volumetric ratio of the two ionic liquids. When the capacitor

was totally in EMIMBF4 as electrolyte, the capacitance is the highest (116 F/g)

between the voltage window of 0 – 3.5 V. Then the specific capacitance gradually

decrease with the increasing volumetric ratio of C8mimBF4 to 40% (80 F/g).

Subsequently with the increasing of C8mimBF4, the specific capacitance changes less

significantly within the range of 80 F/g to 92 F/g with a second peak at 80%

C8mimBF4 and 20% EMIMBF4.

120
(0-2.5V)
105 (0-3V)
(0-3.5V)
90
Specific Capacitance (F/g)

75

60

45

30

15

0
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Volumetric ratio of C8mimBF4

Figure 4.14 Specific capacitances of CRG electrode changing with the volumetric ratio of the

mixing EMIMBF4 and C8mimBF4 electrolyte.

82
 We attempt to propose a model to explain the observed phenomenon. Because the

specific capacitance of supercapacitor is directly correlated with the charges

accumulated on the surface of active electrode materials, the more charges on the

surface the higher specific capacitance it will be. If there were only one size of ions

on the surface, the utilization of the ions would not be sufficient. Mixing the different

sizes of ions together is helpful to increase the charges accumulating. Figure 4.15 is

the schematic illustration of this model.

In the current research, two types of ionic liquids electrolytes, C8mimBF4 and

EMIMBF4 are tested. It is an effective method to increase the power density and

energy density which are also the main limits for the applications of supercapacitors.

In the future work, more types of ionic liquids with different anion and cation

structures should be experimentally tested. Also mixing two or more types of ionic

liquids may be able to decrease the viscosity and increase the conductivity of the

mixture.

Figure 4.15 Schematic of the specific capacitance verses size of anions.

83
4.4.6 GNS/MnOx in Aqueous Electrolyte

4.4.6.1 Effects of GNS content:

Figure 4.16 shows the CV curves of CRG/Mn3O4 nanocomposite electrode with

different CRG content in 1 M Na2SO4 electrolyte. It can be observed that all the CV

curves show relative rectangular shape indicating its electric double layer capacitance.

With the increasing of scan rate from 5 to 50 mV/s, the curves exhibit slightly

distortion. At the scan rate of 5 mV/s, the specific capacitance of the supercapacitor

was 29 F/g, as the scan rate increasing up to 50 mV/s, a specific capacitance of 13 F/g

maintains. The decrease in capacitance at higher scan rates is assigned to the presence

of inner active sites that cannot sustain the redox transitions, which is possibly due to

the diffusion effect of ions within the electrode. The decreasing trend of the

capacitance indicates that parts of the surface of the electrode are inaccessible at high

charging–discharging rates[107].

Obviously, the shape of the CV curves, shown in figures 4.16 (a) – (c), changed

from ideal rectangular to oval with the increasing of scan rate. According to Li et al.

[108], this phenomenon can be caused by the following reasons: firstly, the internal

resistance of the electrode inhibited the charge collection; secondly, the relative low

conductivity of the 1 M Na2SO4 electrolyte solution; finally, the diffusion limitation of

Na+ in the electrode.

By comparing figure 4.16 (a) – (c), it can be easily concluded that with the

increasing of Mn3O4 weight content ratio in the hybrid electrodes the specific

capacities are gradually decreasing. For example, at the Mn3O4 weight ratio of 50%,

66% and 75%, the specific capacities are 31, 23, and 19 F/g, respectively. This

phenomenon can be contributed by the decreasing of GNS content in the hybrid

material which reduces the electrical conductivity.

84
4.4.6.2 Effects of different MnOx phase

CV and galvanostatic charge-discharge electrochemical tests were also performed

to investigate the electrochemical performance of the CRG/MnO2 electrode sample

(see figure 4.17). The specific capacitance of CRG/MnO2 was calculated to be 109.6,

74, and 62 F/g at the scan rate of 5, 20, 50 mV/s, respectively, Compare CRG/MnO2

(figure 4.17) and CRG/Mn3O4 (figure 4.16), the specific capacitance of the former

one (109.6 F/g) is much higher than the latter one (30 F/g). This implies that the

MnOx phase affects the electrochemical properties significantly and MnO2 is more

suitable than Mn3O4 in working as electrode material for supercapacitors.

4.4.6.3 Effect of different types of GNS

Since the RG/MnO2 and CRG/MnO2 were all synthesized with the same

procedure, both of the samples have the same weight ratio of MnO2 and the same

phase of manganese oxide, the only difference is the graphene nanosheets. The CV

profiles of RG/MnO2 were presented in figure 4.18. The specific capacitance of

RG/MnO2 hybrid material calculated by the CV profiles was 164, 121 and 80 F/g at

the scan rate of 5, 20, 50 mV/s, respectively.

Compare figure 4.18 to figure 4.17, the former one shows more ideal rectangular

shape even at high scan rate of 50 mV/s, which indicates the CRG/MnO2 has a lower

internal resistance. However, the specific capacitance of CRG/MnO2 was slightly

lower than that of RG/MnO2. This might because the specific surface area of CRG is

much higher than that of RG, and during the preparing procedure, not all the MnO2

was able to attach on the surface of graphene nanosheets.

85
 0.6
5 mV/s Csp = 38
0.5
20 mV/s 31
0.4 50 mV/s 23 F/g

0.3

Current Density (A/g)

0.2

0.1

0.0

-0.1

-0.2

-0.3

-0.4

-0.5
0.0 0.2 0.4 0.6 0.8 1.0

Voltage (V)

(a)
CRG+Mn3O4 1:2 Csp MnOx = 23.5 F/g
0.6
5 mV/s Csp = 20
0.5 20 mV/s 23
50 mV/s 22 F/g
0.4

0.3
Current Density (A/g)

0.2

0.1

0.0

-0.1

-0.2

-0.3

-0.4

-0.5
0.0 0.2 0.4 0.6 0.8 1.0

Voltage (V)

(b)
Csp MnOx = 18 F/g
CRG+Mn3O4 1:3
0.6
5 mV/s Csp = 20
0.5 20 mV/s 19
0.4
50 mV/s 17 F/g

0.3
Current Density (A/g)

0.2

0.1

0.0

-0.1

-0.2

-0.3

-0.4

-0.5
0.0 0.2 0.4 0.6 0.8 1.0

Voltage (V)

(c)
Figure 4.16 Cyclic voltammetric characterstics of (a) CRG/Mn3O4-50, (b) CRG/Mn3O4-66, and (c)

CRG/Mn3O4-75 electrodes in 1 M Na2SO4 electrolyte at different scan rates from 5 to 50 mV/s

86
 5 mV/s Csp = 109.6
2.0
20 mV/s 74
50 mV/s 62 F/g
1.5

Current Density (A/g)

1.0

0.5

0.0

-0.5

-1.0

0.0 0.2 0.4 0.6 0.8 1.0

Voltage (V)

Figure 4.17 Cyclic voltammetric curves of CRG/MnO2 electrode: at the scan rate of 5 to 50 mV/s

2.5
5 mV/s Csp = 164
2.0 20 mV/s 121
50 mV/s 80 F/g
1.5

1.0
Current Density (A/g)

0.5

0.0

-0.5

-1.0

-1.5

-2.0
0.0 0.2 0.4 0.6 0.8 1.0
Voltage (V)

Figure 4.18 Cyclic voltammetric curves of RG/MnO2 electrode: at the scan rate of 5 to 50 mV/s

87
4.5 Summary

The GO and RG made by the modified Hummer’s method had relative low

surface area. When working as electrode materials in EMIMBF4 electrolyte, the

specific capacitance were as low as 13 F/g and 5 F/g, respectively, due to the “re-stack”

during the preparation of the samples. The CRG electrode in EMIMBF4 electrolyte

shows a high specific capacitance of 143 F/g at the scan rate of 20 mV/s with the

voltage window of 0 – 4 V, and good cycle performance with 2.6% decreasing per

cycle after 1000 cycles. The curved shape graphene sheets (CRG) effectively inhibit

the agglomeration and highly increase the specific surface area. This makes it an

extremely promising candidate electrode material for the electrical double-layer

capacitors.

The electrochemical characteristics of CRG in two different types of ionic liquid

electrolytes, i.e. EMIMBF4 and C8mimBF4, are compared and summarized in Table

4.4. This study shows that EMIMBF4 can provide a higher specific capacitance, a

wider stable voltage window, leading to higher energy and higher power density.

Since the two ionic liquids has the same anion, [BF4] +, and the same structure of

cations. The only different is the length of chain group on the imidazole ring.

C8mimBF4 with the relative longer chain group has higher van der Walls’s force than

that of EMIMBF4. Therefore, C8mimBF4 has a lower conductivity and higher

viscosity, and the electrochemical performance of the supercapacitor in C8mimBF4 is

not as good as the one in EMIMBF4.

Table 4.4 Properties of CRG electrode in EMIMBF4 and C8mimBF4 as electrolyte

Electrode Electrolyte Voltage Specific Power Density Energy Density Resistance

Window (V) Capacitance (kW/kg) (Wh/kg) (Ω)
(F/g)
CRG EMIMBF4 0 – 3.5 116 34 197 15
CRG C8mimBF4 0–3 66 13 82.5 25

88
 Two types of ionic liquids, EMIMBF4 and C8mimBF4, which have the same

anions and different cationchains were mixed together, in order to find out the effects

of the length of cation chains to the electrochemical performance of ionic liquids.

When the electrolyte was solely EMIMBF4, the supercapacitor had the highest

specific capacitance of 116 F/g (0 – 3.5 V, 20 mV/s). With the increasing of

C8mimBF4 to 40% mass ratio, the specific capacitance continued decreasing, and

roughly kept the same as the mass ratio of C8mimBF4 from 60% - 100%.

Different types of GNS/MnOx hybrid nanocomposites were synthesized and

compared as electrode materials for supercapacitors. The specific capacitance and

weight ratio of MnOx content for each sample are summarized in Table 4.5. It was

found that the phase of GNS and MnOx both affect the electrochemical performance

significantly. In the nanocomposites RG is more suitable than CRG, and MnO2 is

better than Mn3O4 obviously

Table 4.5 The specific capacitance and weight ratio of MnOx content for each sample

Sample Specific Capacitance of Specific Weight Percentage of

Total (F/g) Capacitance of MnOx
MnOx (F/g)

PureRG 2.9 -

RG/Mn3O4 29 32.5 89%

102 637 16%

RG/MnO2 96 342 28.8%
164 482 34%
71 173 41%
50 106 47%
38 76 50%
CRG/Mn3O4 23 35 66%
20 27 75%

CRG/MnO2 109.6 363 30%

The RG/MnO2 with 34% MnO2 content reached the highest specific capacities of

164 F/g at the scan rate of 20 mV/s. In the hybrid materials, MnOx contributes a lot on

89
the specific capacitance by the pseudo-capacitive mechanism. However, when the

weight ratio of MnOx in the hybrid material is too much the specific capacitance will

be decreased. This is because MnOx will reduce the conductivity of the electrode

material..

90
Chapter 5: Conclusion

Carbon-based electrode materials have showed great potential in alternative

energy technologies. Electrochemical characteristics of carbon-based electrode

materials graphene nanosheets (GNS), reduced graphene (RG), GNS/MnOx hybrid

materials employed in both lithium-ion batteries and supercapacitors were studied. In

addition, 1-ethyl-3-methylimidazolium tetrafluoroborate (EMIMBF4) and

1-methyl-3-octylimidazolium tetrafluoroborate (C8mimBF4) are tested as electrolytes

for supercapacitors, and their electrochemical characteristics are compared.

Furthermore, silicon nanowires (SiNWs) were directly deposited on the substrate of

carbon nano-fibers (CF) through the chemical vapor deposition with Au catalyst, and

the SiNWs/CF composite were tested as anode material for lithium-ion batteries.

GO/MnOx is a priority anode in Li-ion battery, and the lower manganese valency

would deliver higher reversible capacity. GNS/MnOx– 150 (synthesized at 150℃ in

air) delivered a reversible discharge capacity of 619 mAh/g with an average

Coulombic efficiency of 92% and low capacity fade per cycle of 2.6%. GNS/MnOx–

400 (synthesized at 400℃ in air) delivered a reversible discharge capacity of 213

mAh/g with an Coulombic efficiency of 80% and capacity fade per cycle of 5.9%.

The decreased electrochemical performance is caused by the higher thermal treatment

temperature which greatly increased the particle size of MnOx. GNS/MnOx– 400 inert

(synthesized at 400℃ in 5% H2/Ar atmosphere) showed a reversible discharge

91
capacity of 719 mAh/g with the Coulombic efficiency of 92% and low capacity fade

per cycle of 0.8%. On the one hand, MnOx nanoparticles working as “spacer” between

the GNS formed a 3 – D structure which greatly increased the accessibility of lithium

ions. On the other hand, GNS worked as a “buffer”accommodated the great volume

change caused by the manganese oxide displacement reaction, and provided short

conductivity path for lithium ions.

SiNWs gown on carbon fibers (28wt% of SiNWs) delivered a high reversible

discharge capacityof 975 mAh/g with an average Coulombic efficiency of 85.7% and

capacity fade per cycle of 9.3%. The sample combined with GO and SiNWs/CF

showed a reversible discharge capacity of 890 mAh/g equivalent to the capacity of

2490 mAh/g based on Si with the Coulombic efficiency of 77%, and the capacity fade

per cycle was 11% after 5 cycles. This promising result indicates that utilizing an

effective buffer (GO) to accommodate will greatly promote the cyclic performance.

For supercapacitors three different types of GNS were tested as the electrode

material in two types of ionic liquids electrolytes, C8mimBF4 and EMIMBF4. Both of

the ionic liquids have the same anions [BF4]-, and the only difference is the length of

the brunch chains of the cations. It was experimentally verified that the longer brunch

chain will decreased the conductivity and charge accumulated on electrode surface

resulting in low capacitance. The curved GNS, which has the highest specific surface

area, delivered a high specific capacitance of 116 F/g in EMIMBF4 electrolyte at the

scan rate of 20 mV/s with the power density of 34 kW/kg and energy density of 197

Wh/kg.

GNS/MnOx (CRG/Mn3O4, RG/Mn3O4, CRG/MnO2 and RG/MnO2) specimens

were tested as the electrode material for supercapacitor in 1 M Na2SO4 electrolyte.

RG/ MnO2 composite showed highest specific capacities of 164 F/g at the scan rate of

92
20 mV/s with the voltage window ranging from 0 – 4 V. In the hybrid material, MnOx

contribute a lot by the pseudo – capacitive mechanism.

93
 REFERENCES

[1] G. P. Gholam-Abbas Nazari, Lithium Batteries Science and Technology, 2009.

[2] R. L. Hirsch, Energy policy vol. 36, 2008.
[3] S. Hossain, Ed., Handbook of Batteries. New York: McGraw-Hill Inc, 1995,
p.^pp. Pages.
[4] G. E. Blomgren, "Liquid electrolytes for lithium and lithium-ion batteries,"
Journal of Power Sources, vol. 119–121, pp. 326-329, 2003.
[5] Y. T. Lee, et al., "Improved electrochemical performance of Li-doped natural
graphite anode for lithium secondary batteries," Journal of Power Sources, vol.
139, pp. 230-234, 2005.
[6] Y. Wang, et al., "Tin Nanoparticle Loaded Graphite Anodes for Li-Ion Battery
Applications," ChemInform, vol. 36, pp. no-no, 2005.
[7] Z. Jian, et al., "Natural flake graphite modified by mild oxidation and carbon
coating treatment as anode material for lithium ion batteries," Procedia
Engineering, vol. 27, pp. 55-62, 2012.
[8] H. Fujimoto, et al., "7Li nuclear magnetic resonance studies of hard carbon
and graphite/hard carbon hybrid anode for Li ion battery," Journal of Power
Sources, vol. 196, pp. 1365-1370, 2011.
[9] J. Hu, et al., "Electrochemical behavior and microstructure variation of hard
carbon nano-spherules as anode material for Li-ion batteries," Solid State
Ionics, vol. 178, pp. 265-271, 2007.
[10] T. Liu, et al., "Effect of vacuum carbonization treatment on the irreversible
capacity of hard carbon prepared from biomass material," Materials Letters,
vol. 64, pp. 74-76, 2010.
[11] J. Yang, et al., "Study of nano-porous hard carbons as anode materials for
lithium ion batteries," Materials Chemistry and Physics, vol. 135, pp. 445-450,
2012.
[12] S. Stankovich, et al., "Graphene-based composite materials," Nature, vol. 442,
pp. 282-286, 2006.
[13] P. Lian, et al., "Large reversible capacity of high quality graphene sheets as an
anode material for lithium-ion batteries," Electrochimica Acta, vol. 55, pp.
3909-3914, 2010.
[14] C.-M. Park, et al., "Li-alloy based anode materials for Li secondary batteries,"
Chemical Society Reviews, vol. 39, pp. 3115-3141, 2010.
[15] V. Subramanian, et al., "Nanocrystalline SnO2 and In-doped SnO2 as anode
materials for lithium batteries," Solid State Ionics, vol. 175, pp. 181-184,
2004.
[16] L. Kavan, et al., "Electrochemical and Photoelectrochemical Investigation of
Single-Crystal Anatase," Journal of the American Chemical Society, vol. 118,
pp. 6716-6723, 1996/01/01 1996.
[17] X. Su, et al., "Advanced titania nanostructures and composites for lithium ion
battery," Journal of Materials Science, vol. 47, pp. 2519-2534, 2012/03/01
2012.

94
[18] J. S. Chen, et al., "Constructing Hierarchical Spheres from Large Ultrathin
Anatase TiO2 Nanosheets with Nearly 100% Exposed (001) Facets for Fast
Reversible Lithium Storage," Journal of the American Chemical Society, vol.
132, pp. 6124-6130, 2010/05/05 2010.
[19] S. Ding, et al., "Graphene-supported anatase TiO2 nanosheets for fast lithium
storage," Chemical Communications, vol. 47, pp. 5780-5782, 2011.
[20] D. Deng, et al., "Green energy storage materials: Nanostructured TiO2 and
Sn-based anodes for lithium-ion batteries," Energy & Environmental Science,
vol. 2, pp. 818-837, 2009.
[21] C. Jiang, et al., "Nanocrystalline Rutile TiO2 Electrode for High-Capacity and
High-Rate Lithium Storage," Electrochemical and Solid-State Letters, vol. 10,
pp. A127-A129, January 1, 2007 2007.
[22] P. Poizot, et al., "Nano-sized transition-metal oxides as negative-electrode
materials for lithium-ion batteries," Nature, vol. 407, pp. 496-499, 2000.
[23] K. Zhong, et al., "MnO powder as anode active materials for lithium ion
batteries," Journal of Power Sources, vol. 195, pp. 3300-3308, 2010.
[24] Y. Lu, et al., "Macroporous Co3O4 platelets with excellent rate capability as
anodes for lithium ion batteries," Electrochemistry Communications, vol. 12,
pp. 101-105, 2010.
[25] N. Liu, et al., "One-Step Ionic-Liquid-Assisted Electrochemical Synthesis of
Ionic-Liquid-Functionalized Graphene Sheets Directly from Graphite,"
Advanced Functional Materials, vol. 18, pp. 1518-1525, 2008.
[26] X. Yao, et al., "Co3O4 nanowires as high capacity anode materials for lithium
ion batteries," Journal of Alloys and Compounds, vol. 521, pp. 95-100, 2012.
[27] A. N. Dey, "Electrochemical Alloying of Lithium in Organic Electrolytes,"
Journal of The Electrochemical Society, vol. 118, pp. 1547-1549, 1971.
[28] H. W. Masaki Yoshio, Kenji Fukuda, Tatsuo Umeno, Nikolay Dimov and
Zempachi Ogumi, "Carbon-Coated Si as a Lithium-Ion Battery Anode
Material," Journal of The Electrochemical Society vol. 149, pp. A1598-A1603,
2002.
[29] L. Hu, et al., "Silicon–Carbon Nanotube Coaxial Sponge as Li-Ion Anodes
with High Areal Capacity," Advanced Energy Materials, vol. 1, pp. 523-527,
2011.
[30] J. Guo, et al., "Cyclability study of silicon–carbon composite anodes for
lithium-ion batteries using electrochemical impedance spectroscopy,"
Electrochimica Acta, vol. 56, pp. 3981-3987, 2011.
[31] R. Ruffo, et al., "Impedance Analysis of Silicon Nanowire Lithium Ion Battery
Anodes," The Journal of Physical Chemistry C, vol. 113, pp. 11390-11398,
2009/07/02 2009.
[32] M.-H. Park, et al., "Silicon Nanotube Battery Anodes," Nano Letters, vol. 9,
pp. 3844-3847, 2009/11/11 2009.
[33] Y. Yao, et al., "Interconnected Silicon Hollow Nanospheres for Lithium-Ion
Battery Anodes with Long Cycle Life," Nano Letters, vol. 11, pp. 2949-2954,
2011/07/13 2011.
[34] R. Kötz and M. Carlen, "Principles and applications of electrochemical
capacitors," Electrochimica Acta, vol. 45, pp. 2483-2498, 2000.
[35] G. Wang, et al., "A review of electrode materials for electrochemical
supercapacitors," Chemical Society Reviews, vol. 41, pp. 797-828, 2012.
[36] L. L. Zhang and X. S. Zhao, "Carbon-based materials as supercapacitor
electrodes," Chemical Society Reviews, vol. 38, pp. 2520-2531, 2009.
95
[37] B. E. Conway, "Transition from “Supercapacitor” to “Battery” Behavior in
Electrochemical Energy Storage," Journal of The Electrochemical Society, vol.
138, pp. 1539-1548, January 1, 1991 1991.
[38] B. E. Conway, et al., "The role and utilization of pseudocapacitance for energy
storage by supercapacitors," Journal of Power Sources, vol. 66, pp. 1-14,
1997.
[39] I.-H. Kim and K.-B. Kim, "Ruthenium Oxide Thin Film Electrodes for
Supercapacitors," Electrochemical and Solid-State Letters, vol. 4, pp.
A62-A64, January 1, 2001 2001.
[40] M. Mastragostino, et al., "Polymer-based supercapacitors," Journal of Power
Sources, vol. 97–98, pp. 812-815, 2001.
[41] Y. Zhang, et al., "Progress of electrochemical capacitor electrode materials: A
review," International Journal of Hydrogen Energy, vol. 34, pp. 4889-4899,
2009.
[42] J. Lee, et al., "Recent Progress in the Synthesis of Porous Carbon Materials,"
Advanced Materials, vol. 18, pp. 2073-2094, 2006.
[43] J. H. Knox, et al., "Structure and performance of porous graphitic carbon in
liquid chromatography," Journal of Chromatography A, vol. 352, pp. 3-25,
1986.
[44] D. Qu and H. Shi, "Studies of activated carbons used in double-layer
capacitors," Journal of Power Sources, vol. 74, pp. 99-107, 1998.
[45] B. E. Conway, Electrochemical Supercapacitors. New York:
Academic/Plenum Press, 1999.
[46] K. Lota, et al., "Capacitance properties of
poly(3,4-ethylenedioxythiophene)/carbon nanotubes composites," Journal of
Physics and Chemistry of Solids, vol. 65, pp. 295-301, 2004.
[47] L.-Z. Fan and J. Maier, "High-performance polypyrrole electrode materials for
redox supercapacitors," Electrochemistry Communications, vol. 8, pp. 937-940,
2006.
[48] V. Gupta and N. Miura, "High performance electrochemical supercapacitor
from electrochemically synthesized nanostructured polyaniline," Materials
Letters, vol. 60, pp. 1466-1469, 2006.
[49] C. Arbizzani, et al., "Polymer-based redox supercapacitors: A comparative
study," Electrochimica Acta, vol. 41, pp. 21-26, 1996.
[50] E. Frackowiak, et al., "Supercapacitors based on conducting
polymers/nanotubes composites," Journal of Power Sources, vol. 153, pp.
413-418, 2006.
[51] H. Lee, et al., "RuOx/polypyrrole nanocomposite electrode for
electrochemical capacitors," Synthetic Metals, vol. 160, pp. 1055-1059, 2010.
[52] M. Galiński, et al., "Ionic liquids as electrolytes," Electrochimica Acta, vol. 51,
pp. 5567-5580, 2006.
[53] J.-H. Kim, et al., "Polypyrrole/carbon composite electrode for high-power
electrochemical capacitors," Electrochimica Acta, vol. 52, pp. 1727-1732,
2006.
[54] U. M. Patil, et al., "Chemically synthesized hydrous RuO2 thin films for
supercapacitor application," Journal of Alloys and Compounds, vol. 509, pp.
1677-1682, 2011.
[55] M. D. Stoller, et al., "Graphene-Based Ultracapacitors," Nano Letters, vol. 8,
pp. 3498-3502, 2008/10/08 2008.

96
[56] T. Y. Kim, et al., "High-Performance Supercapacitors Based on Poly(ionic
liquid)-Modified Graphene Electrodes," ACS Nano, vol. 5, pp. 436-442,
2011/01/25 2010.
[57] M. M. Thackeray, et al., "Lithium insertion into manganese spinels,"
Materials Research Bulletin, vol. 18, pp. 461-472, 1983.
[58] M. M. Thackeray, et al., "Structural characterization of the lithiated iron
oxides LixFe3O4 and LixFe2O3 (0&lt;x&lt;2)," Materials Research Bulletin,
vol. 17, pp. 785-793, 1982.
[59] S. L. Cheekati, et al., "The Impacts of Graphene Nanosheets and Manganese
Valency on Lithium Storage Characteristics in Graphene/Manganese Oxide
Hybrid Anode," Journal of Nanomaterials, vol. 2012, p. 10, 2012.
[60] H. Kim, "Electrochemical and ex-situ analysis on manganese oxide/graphene
hybrid anode for lithium rechargeable batteries," Materials Research vol. 26, p.
6, 2011.
[61] S.-Y. Liu, et al., "Nanocrystal manganese oxide (Mn3O4, MnO) anchored on
graphite nanosheet with improved electrochemical Li-storage properties,"
Electrochimica Acta, vol. 66, pp. 271-278, 2012.
[62] A. Yu, et al., "Free-Standing Layer-By-Layer Hybrid Thin Film of
Graphene-MnO2 Nanotube as Anode for Lithium Ion Batteries," The Journal
of Physical Chemistry Letters, vol. 2, pp. 1855-1860, 2011/08/04 2011.
[63] H. Wang, et al., "Mn3O4−Graphene Hybrid as a High-Capacity Anode
Material for Lithium Ion Batteries," Journal of the American Chemical Society,
vol. 132, pp. 13978-13980, 2010/10/13 2010.
[64] K. Zhong, et al., "Investigation on porous MnO microsphere anode for lithium
ion batteries," Journal of Power Sources, vol. 196, pp. 6802-6808, 2011.
[65] J. Gao, et al., "Spongelike Nanosized Mn3O4 as a High-Capacity Anode
Material for Rechargeable Lithium Batteries," Chemistry of Materials, vol. 23,
pp. 3223-3227, 2011/07/12 2011.
[66] D. Pasero, et al., "Co-doped Mn3O4: a possible anode material for lithium
batteries," Journal of Power Sources, vol. 141, pp. 156-158, 2005.
[67] S. Cheekati, et al., "Nanographene Platelets and Their Composites for Li-Ion
Batteries," Meeting Abstracts, vol. MA2010-02, p. 197, July 8, 2010 2010.
[68] X. Fang, et al., "Electrode reactions of manganese oxides for secondary
lithium batteries," Electrochemistry Communications, vol. 12, pp. 1520-1523,
2010.
[69] B. A. Boukamp, et al., "All-Solid Lithium Electrodes with Mixed-Conductor
Matrix," Journal of The Electrochemical Society, vol. 128, pp. 725-729, 1981.
[70] N. Ding, et al., "Improvement of cyclability of Si as anode for Li-ion
batteries," Journal of Power Sources, vol. 192, pp. 644-651, 2009.
[71] U. Kasavajjula, et al., "Nano- and bulk-silicon-based insertion anodes for
lithium-ion secondary cells," Journal of Power Sources, vol. 163, pp.
1003-1039, 2007.
[72] Z. Luo, et al., "High performance silicon carbon composite anode materials
for lithium ion batteries," Journal of Power Sources, vol. 189, pp. 16-21,
2009.
[73] Q. Si, et al., "Highly reversible carbon–nano-silicon composite anodes for
lithium rechargeable batteries," Journal of Power Sources, vol. 189, pp.
761-765, 2009.
[74] A. Anani and R. A. Huggins, "Multinary alloy electrodes for solid state
batteries I. A phase diagram approach for the selection and storage properties
97
 determination of candidate electrode materials," Journal of Power Sources,
vol. 38, pp. 351-362, 1992.
[75] Y. Zhang, et al., "Composite anode material of silicon/graphite/carbon
nanotubes for Li-ion batteries," Electrochimica Acta, vol. 51, pp. 4994-5000,
2006.
[76] Q. Si, et al., "A high performance silicon/carbon composite anode with carbon
nanofiber for lithium-ion batteries," Journal of Power Sources, vol. 195, pp.
1720-1725, 2010.
[77] J. K. K. B. C. M. H. H. LeeSmithHaynerKung, "Silicon nanoparticles–
graphene paper composites for Li ion battery anodes," Chemical
Communications, vol. 46, p. 2025, 2010.
[78] L. Ji, et al., "Graphene/Si multilayer structure anodes for advanced half and
full lithium-ion cells," Nano Energy.
[79] H. Li, et al., "A High Capacity Nano ­ Si Composite Anode Material for
Lithium Rechargeable Batteries," Electrochemical and Solid-State Letters, vol.
2, pp. 547-549, November 1, 1999 1999.
[80] H.-C. Shin, et al., "Porous silicon negative electrodes for rechargeable lithium
batteries," Journal of Power Sources, vol. 139, pp. 314-320, 2005.
[81] D.-K. Kang, et al., "Microstructured Nanopore-Walled Porous Silicon as an
Anode Material for Rechargeable Lithium Batteries," Journal of The
Electrochemical Society, vol. 155, pp. A276-A281, 2008.
[82] J. Cho, "Porous Si anode materials for lithium rechargeable batteries," Journal
of Materials Chemistry, vol. 20, pp. 4009-4014, 2010.
[83] J. Yang, et al., "Small particle size multiphase Li-alloy anodes for
lithium-ionbatteries," Solid State Ionics, vol. 90, pp. 281-287, 1996.
[84] R. Huggins and W. Nix, "Decrepitation model for capacity loss during cycling
of alloys in rechargeable electrochemical systems," Ionics, vol. 6, pp. 57-63,
2000.
[85] C. K. Chan, et al., "High-performance lithium battery anodes using silicon
nanowires," Nat Nano, vol. 3, pp. 31-35, 2008.
[86] L.-F. Cui, et al., "Crystalline-Amorphous Core−Shell Silicon Nanowires for
High Capacity and High Current Battery Electrodes," Nano Letters, vol. 9, pp.
491-495, 2009/01/14 2008.
[87] L. Hu, et al., "Si nanoparticle-decorated Si nanowire networks for Li-ion
battery anodes," Chemical Communications, vol. 47, pp. 367-369, 2011.
[88] T. Song, et al., "Arrays of Sealed Silicon Nanotubes As Anodes for Lithium
Ion Batteries," Nano Letters, vol. 10, pp. 1710-1716, 2010/05/12 2010.
[89] Q. Jiang, "Atmospheric Growth of Silicon Nanowires by Chemical Vapor
Deposition," 2011.
[90] T. D. Hatchard and J. R. Dahn, "In Situ XRD and Electrochemical Study of
the Reaction of Lithium with Amorphous Silicon," Journal of The
Electrochemical Society, vol. 151, pp. A838-A842, 2004.
[91] J. Yan, et al., "Fast and reversible surface redox reaction of graphene–MnO2
composites as supercapacitor electrodes," Carbon, vol. 48, pp. 3825-3833,
2010.
[92] G. Ning, et al., "Gram-scale synthesis of nanomesh graphene with high
surface area and its application in supercapacitor electrodes," Chemical
Communications, vol. 47, pp. 5976-5978, 2011.
[93] S. Vivekchand, et al., "Graphene-based electrochemical supercapacitors,"
Journal of Chemical Sciences, vol. 120, pp. 9-13, 2008.
98
[94] W. Lv, et al., "Low-Temperature Exfoliated Graphenes: Vacuum-Promoted
Exfoliation and Electrochemical Energy Storage," ACS Nano, vol. 3, pp.
3730-3736, 2009/11/24 2009.
[95] Q. Du, et al., "Preparation of functionalized graphene sheets by a
low-temperature thermal exfoliation approach and their electrochemical
supercapacitive behaviors," Electrochimica Acta, vol. 55, pp. 3897-3903,
2010.
[96] Y. Si and E. T. Samulski, "Exfoliated Graphene Separated by Platinum
Nanoparticles," Chemistry of Materials, vol. 20, pp. 6792-6797, 2008/11/11
2008.
[97] M. Alamgir and K. M. Abraham, "Studies to improve the
Li/SO&lt;sub&gt;2&lt;/sub&gt;Cl&lt;sub&gt;2&lt;/sub&gt; cell," Journal of
Applied Electrochemistry, vol. 24, pp. 288-297, 1994.
[98] J. Yan, et al., "Electrochemical properties of graphene nanosheet/carbon black
composites as electrodes for supercapacitors," Carbon, vol. 48, pp. 1731-1737,
2010.
[99] K. Zhang, et al., "Surfactant-intercalated, chemically reduced graphene oxide
for high performance supercapacitor electrodes," Journal of Materials
Chemistry, vol. 21, pp. 7302-7307, 2011.
[100] C. Liu, et al., "Graphene-Based Supercapacitor with an Ultrahigh Energy
Density," Nano Letters, vol. 10, pp. 4863-4868, 2010/12/08 2010.
[101] S. Chen, et al., "Graphene Oxide−MnO2 Nanocomposites for
Supercapacitors," ACS Nano, vol. 4, pp. 2822-2830, 2010/05/25 2010.
[102] Z. Li, et al., "Flexible graphene/MnO2 composite papers for supercapacitor
electrodes," Journal of Materials Chemistry, vol. 21, pp. 14706-14711, 2011.
[103] M. D. Stoller and R. S. Ruoff, "Best practice methods for determining an
electrode material's performance for ultracapacitors," Energy & Environmental
Science, vol. 3, pp. 1294-1301, 2010.
[104] R. Kötz, et al., "Temperature behavior and impedance fundamentals of
supercapacitors," Journal of Power Sources, vol. 154, pp. 550-555, 2006.
[105] W. Lu, et al., "High performance electrochemical capacitors from aligned
carbon nanotube electrodes and ionic liquid electrolytes," Journal of Power
Sources, vol. 189, pp. 1270-1277, 2009.
[106] A. V. Rozhkov, et al., "Electronic properties of mesoscopic graphene
structures: Charge confinement and control of spin and charge transport,"
Physics Reports, vol. 503, pp. 77-114, 2011.
[107] G. S. Gund, et al., "Enhanced activity of chemically synthesized hybrid
graphene oxide/Mn3O4 composite for high performance supercapacitors,"
Electrochimica Acta, vol. 92, pp. 205-215, 2013.
[108] J. Li, et al., "Nickel foam-based manganese dioxide–carbon nanotube
composite electrodes for electrochemical supercapacitors," Journal of Power
Sources, vol. 185, pp. 1569-1574, 2008.

99