Green reduction and patterning of graphene oxide via photothermal and electrochemical methods

In the next chapter, some backgrounds on graphene, graphene oxide and reduced graphene oxide would be introduced, covering the structure, properties, methods of synthesis and application

DECLARATION

I hereby declare that the thesis is my original work and it has been written by me in its entirety I have duly acknowledged all the sources of information which have been used in the thesis

This thesis has also not been submitted for any degree in any university previously

Tao Ye

21st May 2013

Acknowledgements

It has been an exciting and fulfilling experience for me to have spent the last three years in the Nanomaterials Research Lab at the National University of Singapore Special thanks had to be given to my supervisor, Associated Professor Sow Chorng-Haur for his contagious passion and optimism that encourages me on, valuable ideas

to direct me through the bottle-necks in research, and constant help and guidance in the everyday experiments No matter where I would be in the future, I will be sure to take along with me this enriching and unforgettable experience

Hearfelt gratitude must be given to Dr Binni Varghese and Ms Sharon Lim for interesting suggestions and numerous help, and to Mr Teoh Hao Fatt for all the collaborations I would also thank Prof Tok Eng Soon, Dr Zhang Zheng for helping with XPS measurements; Dr Wang Shuai and Prof Loh Kian Ping for supplying with experimental materials; Dr Cong Chunxiao and Asst Prof Yu Ting for micro Raman mapping Sincere appreciation must also be given to my fellow labmates Mr Zheng Minrui, Mr Lu Junpeng, Mr Xi Yilin, Mr Chang Sheh Lit, Mr Rajesh, Mr Rajiv, Ms Loh Pui Yee, Mr Lee Kian Keat, and all the technicians Ms Foo Eng Tin,

Mr Chen Gen Seng, Mr Ong for all the help I had received over the time and for making the lab a warm and homely place to work in

Last but not least, I would like to thank my family and friends for their endless support and for always being there for me through all difficulties and frustrations I hereby dedicate this piece of work to them

Table of Contents

Acknowledgements ii

Table of Contents iii

Abstract vi

List of Publications and Presentations vii

List of Figures viii

Chapter 1: Introduction 1

References 3

Chapter 2: Theoretical Background 4

2.1 Graphene 4

2.1.1 Structure and Properties 4

2.1.2 Synthesis and Modification 7

2.2 Graphene Oxide 8

2.2.1 Structure 9

2.2.2 Synthesis 10

2.2.3 Properties 12

2.3 Reduced Graphene Oxide 14

2.3.1 Structure 14

2.3.2 Reduction of Graphene Oxide 15

2.3.3 Properties 17

2.4 Applications of Graphene Oxide-based Materials 21

2.4.1 Thin Films of GO or RGO 21

2.4.2 Transparent Conductor 21

2.4.3 Sensing 22

2.4.4 Precursor to Graphitic Nanostructure 22

2.4.5 Precursor to Graphene-based Composites 22

References 24

Chapter 3: Experimental Methods 29

3.1 Sample Preparation 29

3.2 Direct Writing with Focused Laser System 31

3.3 Electrical Measurement 34

3.4 Raman Spectroscopy 35

3.4.1 Basic Principles 35

3.4.2 Vibrational States 37

3.4.3 Raman Spectrum of Graphite-based Materials 38

3.4.4 Raman Spectra for Defective Graphite 41

3.4.5 Substrate Effect 42

3.5 X-ray Photoemission Spectroscopy 43

3.6 AFM 43

3.7 Microscopy and Spectrometer 44

References 45

Chapter 4: Photothermal Reduction 47

4.1 Introduction 47

4.2 Conductivity change 48

4.2.1 Increased Electrical Conductivity 49

4.2.2 Contact Resistance 51

4.2.3 Effect of Channel Dimensions 52

4.2.4 Effect of Repeated Irradiation 53

4.3 Chemical Composition 56

4.3.1 Raman Spectroscopy 56

4.3.2 XPS 58

4.4 Morphological Change 60

4.4.1 Film Thickness 60

4.4.2 Optical Contrast 68

4.4.3 Patterning Ability 72

4.5 Surface Properties 73

4.6 Proposed Mechanism of Reduction 74

4.7 Conclusion 75

References 76

Chapter 5: Visible Electrochemical Reduction 78

5.1 Introduction 78

5.2 Electrochemical Reduction 79

5.2.1 Directional 79

5.2.2 Reversibility 80

5.3 Change of Properties 82

5.3.1 Chemical Composition 82

5.3.2 Morphological Change 84

5.4 Temporal Behavior 86

5.4.1 Area Change with respect to Time 86

5.4.2 Area Change correlated to Conductivity 90

5.5 Mechanism 93

5.5.1 Moist assisted Redox Reaction 93

5.6 Conclusion 96

References 97

Chapter 6: Conclusion 98

List of Publications and Presentations

1 Ye Tao, Binni Varghese, Manu Jaiswal, Shuai Wang, Zheng Zhang, Barbaros

Oezyilmaz, Kian Ping Loh, Eng Soon Tok, Chorng Haur Sow Localized

insulator-conductor transformation of graphene oxide thin films via focused

laser beam irradiation Appl Phys A 106, 523-531 (2012)

2 Hao Fatt Teoh, Ye Tao, Eng Soon Tok, Ghim Wei Ho, Chorng Haur Sow Direct enabled graphene oxide–Reduced graphene oxide layered structures with micropatterning J Appl Phys 112, 064309 (2012)

laser-3 H F Teoh, Y Tao, E S Tok, G W Ho, and C H Sow Electrical current

mediated interconversion between graphene oxide to reduced grapene oxide

Applied Physics Letters 98, 1 (2011)

4 Conference presentation: Recent Advances in Graphene and Related

Materials Localized Insulator-Conductor Transformation of Graphene Oxide

Film via Focused Laser Beam Irradiation (2010)

List of Figures

Figure 3.1A) Schematics of the Laser Beam System to focus laser onto sample

placed on the computer controlled x-y stage; B) Schematic diagram depicting focused laser beam following onto the GO film on substrate with pre-deposited gold electrodes for electrical measurement 33 Figure 3.2 Basic principle of Raman Scattering 36 Figure 4.1 A) Optical micrograph of GO film after pruning a single channel across the gold electrode (insert is the optical image of the as-deposited GO film); B) I-

V curves recorded from the as-deposited (black line) and laser pruned (blue line) GO film 48 Figure 4.2 Mobility measurement from the single channel 50 Figure 4.3 Comparison of I-V curve between different laser irradiated patterns: A) single channel scanned (left optical image in the insert) and large area scanned

at contacts (right optical image in the insert); B) Single channel with contacts scanned (left optical image in the insert) and additional large area scanned in the centre (right optical image in the insert) 52 Figure 4.4 A) i) to v) Optical microscopy image of the channel with increasing width vi) Schematic representation of the laser-scanning sequence to create a

conducting channel with fixed length and increasing width; B) I-V curves

recorded from channels i) to v); C) A plot of conductance of laser created

channel as a function of channel width 53 Figure 4.5 A) Optical micrograph to show the repeatedly irradiated area between the electrodes; B) Change of conductance of reduced GO with repeated laser

irradiation, each time with 6 mW irradiation power 54 Figure 4.6 A) Raman spectrum and B) Raman-mapping of as-deposited and laser-pruned GO 57 Figure 4.7 A) C1s scan of XPS for GO and rGO; B) [pending] O1s scan for GO and rGO 58 Figure 4.8 XPS Spectrum of A) as-deposited and B) laser-irradiated GO film for valence band 59 Figure 4.9 Change in film thickness due to laser irradiation as measured by AFM A) AFM image of single laser irradiated channel across two electrodes B) AFM scan of the area enclosed by dotted line in Figure A and its corresponding height profile; C) AFM image of a scratch on the GO film; D) Height profile near the scratched region (blue) and fitted difference in film thickness (red) 61

Figure 4.10 A) AFM image of the laser irradiated channel and a nearby scratch; B) height profile along the line indicated in the left image (blue) and fitted change

in film thickness (red) The film thickness is ~16nm and sunken depth is ~6nm 63 Figure 4.11 A) AFM image of the laser irradiated area, including the electrodes; B) AFM image of the area enclosed by dotted line in Image (A), GO film to the left

of the dotted line was laser-irradiated, to the right was as-deposited; C) AFM image of a nearby scratch; D) Height profile along the line indicated in image C (blue) and fitted difference in film thickness (red) 64

Figure 4.12 The plot of the remaining thickness of laser-irradiated region against the original film thickenss Laser of 532nm, 10mW was focused over a 1μm by 2μm region and scanned over the sample at 10μm/s 65 Figure 4.13 AFM images of 8 line cuts with 1 to 8 times of laser irradiation each (above) and the height profile along the line drawn in AFM image (below), indicating no significant difference between sunken depth for four different sample with sunken depth of A)<20nm; B)~100nm; C) ~200nm; D)~300nm 67 Figure 4.14 A) Optical contrast of GO film of various thickness with respect to the SiO2 wafer with 100nm oxide layer (background); B) Optical micrograph

image of GO film of thickness i)~30nm; ii)~120nm; iii)~230nm respectively Note the alternating color for the thick patch in Biii) 69 Figure 4.15 A) For a ~70nm film the i) optical micrograph, ii) reflection spectra, iii) contrast spectra of as-deposited and laser pruned areas; B) For a ~230nm film the i) optical micrograph, ii) reflection spectra, iii) contrast spectra of as-

deposited and laser pruned areas 71 Figure 4.16: (A) and (B) Complex structures created by focused laser irradiation on a

GO drop-cast film on SiO2/Si substrate 72 Figure 4.17 A) Optical image of the laser irradiated channel across electrodes (top) and after sonication GO film was removed from substrates while rGO remained (bottom); B) I-V curves of the same channel measured before and after

sonication; C) A single stand-alone channel on SiO2/Si substrate, far away from the electrodes, created by laser pruning and followed by sonication 73 Figure 5.1 The optical micrograph of GO film deposited between two gold electrodes

at t=0, 100, 200 and 300s respectively to show darkening of the sample from negative electrode on the right which extend towards and eventually bridges with the positive electrode on the left 80 Figure 5.2 Reversibility of Electrochemical Reduction of GO A) as-deposited GO film on two gold electrodes Bias voltage of 3V is applied across the two

electrodes starting from t=0, with left electrode being positive; B) darkening of sample from right electrode indicating formation of ERG region until t=120s

when ERG region has not reached the left electrode, applied bias was reversed; C) at t=164s, ERG on the right is receding, while ERG region from left is

extending to the right; D) at t=234s ERG region from the right has completely retracted 81 Figure 5.3 A) Raman of as-deposited and electrochemically reduced GO; B) D and G peaks of as-deposited, reduced and reoxidized GO 83 Figure 5.4 C1s spectrum from XPS measurement of the as-deposited, electrically reduced or electrically re-oxidized GO sample The C=O peak at 287eV

decreased upon reduction and increased to beyond its original value during oxidization There is also a lack of formation of C-O peak around 286eV 84 Figure 5.5 AFM for ERGO samples A) ERG region between two electrodes as

re-imaged by i)optical microscope and ii) AFM image, note the dotted line maps the ERG region; B) section line taken from AFM image; C)Height change along the section line taken 85 Figure 5.6 A) Region between electrodes was captured from optical micrograph to track the area of reduced GO; Note that for the area analysis, only the blue and cyan regions were counted as the others were inhomogeneous patches inherent with the as-deposited GO film B) Evolution of reduced GO area over time from frame by frame analysis of video captured, note the additional growing finger after 9.5s when ERG region has bridged the two electrodes The electrodes are not shown in the graph, with anode on the left 86 Figure 5.7 Area over time curve The curve starts with a quardratic phase followed

by a linear phase The decreasing “tail” at the end of the curve was due to

change of contrast at the end of video 87 Figure 5.8 A) Model of extension of ERG area in Phase I; B) Transition state when the ERG region first fully connects the two electrodes forming a conduction bridge; C) Models for the expansion of ERG area in Phase II, Model I assuming uniform rate of expansion throughout the area, Model II assuming non-uniform rate of expansion, resulting in a linear correlation between width of ERG region and the longitudinal position The actual shape of the ERG region is somewhere

in between Red arrows indicate the velocity of evolution of ERG regions 89 Figure 5.9 A) Current Vs Area B) Change of Sheet Resistance over time Model I and II were shown in Figure 5.8C) 91 Figure 5.11 Laser Pruning guided ERG A) GO film with laser irradiated RGO

segments between gold electrodes; B) The same sample after 5V bias voltage over a period of time Note the dendrites extended from RGO segments towards the cathode, while the end nearer to the anode fades Dendritic connections were formed between other electrodes, or between RGO segments and other

electrodes, too 95

Chapter 1: Introduction

Graphite as an allotrope of carbon has been a subject of interest for scientists since

18th centuries It has been long known that graphite consists of layers of shaped carbon atoms stacked one over another Each atomically thin single layer of graphite is called graphene It has been presumed that free standing single layer graphene is unstable until demonstrated otherwise in 2004 by Geim and Novoselov

honeycomb-et al[1] It has shown unique electronic properties right from its discovery and since

then attracted remarkable research interests

A major hurdle in research and application of graphene is to find an efficient method for large-scale synthesis of the high-quality material One of the potential methods explored was the chemical exfoliation of graphite via oxidation-reduction cycle The oxidation of graphite [2] produces graphene oxide (GO), or earlier known as graphite oxide The reduction of GO forms reduced graphene oxide (RGO) The residual defects in RGO [3] lead to the drastic differences in its structure and properties from pristine graphene It is better viewed as a non-stoichiometric material with highly conducting graphitic domains interspersed in an amorphous matrix On the other hand, the presence of defects renders the material miscible with a wide range of solvents [4, 5] and readily available for chemical or physical adsorption [6-9] The physical properties of GO and RGO can also be chemically tuned by varying the oxygen-containing functional groups [10, 11] Therefore, reduction of GO remains

an economical and up-scalable method for producing solution-processable and chemically viable graphene for a wide range of applications that are less demanding

on the band structures of pristine graphene

Reduction for most experiments was carried out with chemical reducing agents, thermal annealing, or a combination of both These methods would reduce the entire sample, and require furnace or hazardous chemicals Our focus in this thesis is to investigate green reduction methods that can be carried out in ambient environment and room temperature, assisted by either a focused laser beam or a direct applied current More importantly, both methods reported here allow the localized reduction

of GO to RGO on a deposited GO film, with the best resolution of ~1μm Therefore

patterning and reduction of GO was carried out simultaneously instead of the traditional methods of patterning GO thin film via oxidative removal before reduction As the electronic properties of GO and RGO differs drastically, these green methods for localized reduction of GO leads us one-step closer to achieving continuous carbon electronics Via investigation of the properties of RGO produced

as well as the reduction process, we strive to better understand the mechanisms of these methods for its future optimizations and applications

This thesis is organized as follows In the next chapter, some backgrounds on graphene, graphene oxide and reduced graphene oxide would be introduced, covering the structure, properties, methods of synthesis and applications The current methods for reduction and patterning of GO is also presented, as well as some theories on the electrical conduction in RGO

In Chapter 3, the experimental methods for the synthesis, reduction and characterization of GO film is described Some theories on Raman spectroscopy of graphitic material were also detailed In Chapter 4, methods of photothermal reduction and patterning of GO is detailed as well as characterization of the properties of GO and RGO In Chapter 5, electrochemical reduction of GO is detailed with a focus on the reduction process Finally, the concluding Chapter 6 summarizes the results of the previous two chapters

References

[1] K S Novoselov et al., Science 306, 666 (2004)

[2] D C Marcano et al., Acs Nano 4, 4806 (2010)

[3] C Gomez-Navarro et al., Nano Letters 10, 1144 (2010)

[4] S Park et al., Nano Letters 9, 1593 (2009)

[5] D Li et al., Nature Nanotechnology 3, 101 (2008)

[6] D R Dreyer et al., Chem Soc Rev 39, 228 (2010)

[7] S Park, and R S Ruoff, Nature Nanotechnology 5, 309 (2010)

[8] K P Loh et al., J Mater Chem 20, 2277 (2010)

[9] G Eda, and M Chhowalla, Advanced Materials 22, 2392 (2010)

[10] U Kurum et al., Applied Physics Letters 98 (2011)

[11] J Yan et al., Physical Review Letters 98 (2007)

Chapter 2: Theoretical Background

2.1 Graphene

Graphene is a single layer of graphite discovered as the first atomically thin film being metallic and continuous under ambient conditions[1] Its unique band structure has drawn great research attention on its properties and applications

2.1.1 Structure and Properties

Graphene has the same atomic structure as a single layer of graphite, with carbon atoms in hexagonal arrangements Each atom is connected to three neighboring carbon atoms via a single bond All carbon atoms, except those on the edge, are in sp2 hybridization The honeycomb lattice of carbon atoms has been confirmed by transmission electron microscopy Rippling of the flat graphene monolayer is present

in both suspended graphene or graphene on a substrate, which is believed to compensate for the instability of 2D crystals [2]

2.1.1.1 Band Structure

The calculation of energy band of graphene is the same as that of a single layer of graphite, ignoring inter-layer interaction Each carbon has four valence electrons, three of which form tight bonds with neighboring atoms 120° apart in the same plane Their wave functions are of the form [3]

where is the 2s wave function for carbon and are the 2p wave functions whose axes are in the directions towards its three neigbours in the plane The fourth electron is in the 2pz orbital perpendicular to the plane

Each unit cell of graphene contains two atoms, with unit vectors

where is the unit cell size of 1.42Å The corresponding reciprocal space is defined

by reciprocal vectors

The reciprocal unit cell or Brillouin zone is therefore also a hexognal cell rotated π/6 from the unit cell The distance between centre of Brillouin zone Γ to the midpoint of one side M is

The points of the hexagon are termed K and K’ alternatively Tight-binding calculation generates the energy dispersion relationship of graphene as

,

where t=2.75 eV is the nearest neighour transfer integral The dispersion relationship

in the Γ-K or Γ-K’ directions can be obtained as

,

which upon first-order approximation gives a linear dispersion relationship

As the effective mass of fermions is defined as

, at Dirac points graphene can be viewed as having massless fermions[4], which leads to a range of interesting phenomenon such as Quantum Hall Effect[5] that gave rise to the significance of graphene in both fundamental research and applications

2.1.1.2 Optical Contrast

The unique optical property of graphene has been noted since its discovery for its potential as a transparent conductor [6] It has also been noted that graphene layers has a distinctive optical contrast on an oxidized Si wafer As a good conductor, the transparency of graphene is sensitively dependent on the thickness so much so that difference of a single layer can be identified with optical contrast under the right condition This property allows optical micrograph to be the most practical fast method to identify monolayer graphene as compared to the traditional methods like Atomic Force Microscopy (AFM) or Scanning Electron Microscopy (SEM) that are much slower and damaging to the sample

In the earliest works, the single sheet graphene can be sufficiently visible for optical detection under a microscope only on 300nm SiO2[7] Small changes in substrate property, such as thickness of the oxide layer would lead to a significant reduction of contrast As expected the optical contrast also varies from group to group due to different preparation method and sample quality

Various follow-up works have been carried out to explain the origin of the optical contrast [8-11] and therefore enhance optical detection of single sheet graphene[12] Calculations are mostly done with Fresnel’s conditions with frequency-dependent dielectric function of silicon and silicon dioxide [13] while for graphene, some studies approximate it as the real part of the complex dielectric function of graphite that is dispersionless for the visible range [6, 11], while others use experimentally measured frequency-dependent conductivity of graphene[8] or refractive index of graphite [10] Thin film interference alone does not account for the optical contrast

of single layers of graphene merely 0.34nm in thickness Another important factor is the opacity of graphene that modulates the relative amplitude of the interfering paths giving red shift of interferences in the reflection spectrum [11] Standing wave resonances in the oxide layer lead to resonance cancellation and therefore reflection zeroes for specific matching conditions [8, 11], consequently the most suitable wavelength for human eye sensitivity occurs for oxide thickness of 300nm

2.1.2 Synthesis and Modification

The first graphene film was prepared by mechanical exfoliation of pyrolytic graphite [1], commonly referred to as the scotch tape method The method is labor-intensive and film size is limited to 10μm, but it produces the graphene film with the best integrity Graphene produced can be suspended to be considered “free-standing”, or transferred to other substrates such as silicon (Si), silicon dioxide (SiO2) or boron nitride (BN)

Ever since then there has been much effort devoted to synthesize graphene faster in larger quantities with higher quality The vast number of different methods can be classified into five approaches [14]: mechanical exfoliation of bulk graphite [1]; chemical exfoliation of bulk graphite, usually via graphene oxide or graphene fluoride[15]; expitaxial growth of graphene films from silicon carbide (SiC) [16-18]; chemical vapor deposition (CVD) of graphene monolayers[19, 20]; and longitudinal unzipping of carbon nanotubes (CNTs)[21-23]

Chemical exfoliation of graphite via oxidation-reduction cycle was, in the early stage, one of the potential methods for producing graphene in a cost-effective and up-scalable manner This advantage was less significant with the development of CVD methods to produce large-scale graphene rapidly However, chemical exfoliation is still the only method to produce solution-processable graphene, and maintains its competitive edge for functionalization which is essential for certain applications of graphene

2.2 Graphene Oxide

Graphene oxide (GO) or earlier known as graphite oxide has been the subject of study for a long time[24-26] The initial research attention concentrated on it after the discovery of graphene was for its potential as a precursor for solution-based synthesis of graphene[27] Subsequent studies, however, found that although reduction of GO can remove the oxygen functional groups; it is difficult to restore the defect sites introduced during oxidation The non-stoichiometric nature of its chemical structure also presents difficulty in its understanding, as the band structure

of GO as well as RGO varies between samples The presence of residual defect sites renders reduced graphene oxide (RGO) ineffective of demonstrating the fundamental two-dimensional condensed-matter effects unique to graphene such as quantum hall effects or ballistic transport

On the other hand, however, its heterogeneous chemical and electronic structures have led to unique properties and potential of GO itself GO can be synthesized conveniently from oxidation of graphite[28] The various oxygen-functional groups allow GO to interact with a wide range of organic and inorganic materials It is therefore miscible with a variety of solvents [29, 30] and can be deposited with controlled thickness onto various substrates, and also readily complex with many organic and inorganic systems for the synthesis of functional hybrids and composites [31-34] Furthermore, GO is an electronically hybrid material with conducting π-states from sp2 hybridized sites and a large carrier transport gap between the σ-states

of the sp3 hybridized defect sites The fraction of sp2 and sp3 sites, and in turn the band gap as well as conductivity can be chemically tuned over the range from insulator to semi-metal GO was therefore studied as a promising candidate for number of applications like plastic electronics, solar cells, biosensors as well as super-capacitors

2.2.1 Structure

Graphene oxide (GO) is chemically similar, if not identical, to its precursor graphite oxide Structurally, however, GO is referring to the monolayers exfoliated from the stacked structure of graphite oxide Thickness of GO monolayer was determined as

~1nm from atomic force microscopy studies[35-37] The significant increase of layer thickness from that of single-layer graphene, which is 0.34nm [38], is attributed to the oxygen-containing functional groups and adsorbed water above and below the carbon basal plane The intrinsic thickness from dehydrated samples is measured to

be ~0.6nm [39] Lateral dimensions of GO can vary from a few nanometers to hundreds of micrometers depending on various synthesis routes [40, 41]

The oxidative mechanisms as well as the precise chemical structure of GO has been debated over the years Before 1996 all the proposed structure of GO consists of regular lattices with discrete repeat units with variation in the distribution and type of the oxygen functional groups [42, 43] Through solid-state magnetic resonance

(NMR) studies [44] Lerf et a.l characterized a series of GO derivatives and proposed

the widely accepted structural model of GO Lerf’s model proposed that GO is of non-stoichiometric and amorphous nature which explains the complexity and the sample-to-sample variation that presents the primary challenge in elucidation of its structure They have demonstrated that the fundamental features of GO are mainly tertiary alcohols and epoxides present on the surface with the double bonds being either aromatic or conjugate [45] They are ambiguous about the presence or absence

of carboxylic acid groups, if in very low quantities, at the periphery of GO[46] Other slight modifications have been proposed over the years, including esters of tertiary alcohols, with five – and six-membered lactol rings decorating the edge [47, 48], but the essence of model has not changed

As-synthesized GO is primarily a covalent material with ~60% of carbon atoms in the basal plane being sp3 hybridized through σ-bonding with oxygen in the form of epoxy and hydroxyl groups Yet, the atomic structure of GO is unique in that the graphene basal plane is retained despite of the large strain An ideal graphene sheet consists entirely of sp2 hybridized carbon atoms, GO on the other hand is a two-

dimensional network consisting of variable sp2 and sp3 concentrations Therefore

GO is also commonly viewed as a graphene-like material consisting of ordered small

sp2 clusters isolated within sp3 C-O matrix[49] The oxygen-containing functional groups responsible for the sp3 matrix have a wide range of variability in terms of type and coverage, primary due to the difference in starting materials, i.e quality of graphite, and the oxidation protocols The ordered-cluster-in-amorphous-matrix model can be applied to explain various experimental observations including Raman spectroscopy[50, 51], scanning tunneling microscopy[52], high resolution transmission electron microscopy[53, 54] and transport studies[27, 55]

2.2.2 Synthesis

The oxidation of graphite was dated back to some of the earliest studies on chemistry

of graphite As early as 1859 graphite was oxidized with potassium chlorate (KCoO3) and nitric acid(HNO3) in the effort to determine the molecular formula of graphite [31] Nearly a century later, Hummers and Offeman developed an alternative oxidation method[56] using a combination of potassium permanganate (KMnO4) and concentrated sulfuric acid (H2SO4) Diamanganese heptoxide (Mn2O7) formed from KMnO4 in the presence of strong acid is the main oxidizing agent

Subsequent methods of synthesis are mainly modified from these two primary reaction routes The reaction products show strong variance, depending not only on the particular oxidation agents used, but also on the reaction conditions, and the graphite sources, as the localized defects in its π-structure serve as seed points for the oxidation process [31] The Hummers method, with its relatively shorter reaction time and absence of hazardous ClO2 has seen more use in current research, such as the Modified Hummer’s methods used in our experiment [57] One drawback of the Hummers method is potential contamination by excess permanganate ions, which could be removed by an additional treatment with hydrogen peroxide[58]

To obtain graphene oxide (GO), it is necessary to exfoliate the stacked graphite oxide into monolayer or few-layered stacks The most common exfoliation method is simple sonication or stirring of GO in water or polar organic media Compared to mechanical stirring, sonication is faster, but causes substantial damage[59] to the platelets leading to smaller size and a larger distribution of sizes [60] In addition, the oxidation process itself also causes breaking of the graphitic structure into smaller fragments [61] Exfoliated GO is then dispersed in a basic media solution so that the surface functionality weakens the platelet-platelet interactions and prevents agglomerations

Other than differences in starting materials or oxidation protocols, the extent of oxidation also leads to the variability of the structure and properties of graphene oxide Theoretical calculations predict that partial oxidation is thermodynamically favored over complete oxidation; the exact identity and distribution of oxide functional groups also depend strongly on the extent coverage, for example the epoxy to alcohol ratio increases with increasing oxidation [62] The fluidity of the

GO structure presents great challenges in its characterization and understanding, but also great potential in its application as controlled and careful modification of the oxygen-containing functional groups would allow tuning of the sp2 fraction and tailoring of the electrical, optical and chemical properties of GO [49]

2.2.3 Properties

2.2.3.1 Solubility

Graphene oxide (GO) can be dispersed in a number of different organic and inorganic solvents, and therefore readily available for solution-based reactions and depositions Such dispersions are the precursors from which other graphene-based derivatives are prepared

Graphene oxide is intrinsically hydrophilic and readily disperses in water by mild sonication In fact GO is mostly synthesized in an aqueous solution The maximum dispersibility of graphene oxide in solution, which is important for processing and further reactions, depends both on the solvent and the extent of surface functionalization imparted during oxidation At slightly basic pH, negatively charged, hydrophilic oxygen-containing functional groups on the graphene oxide surface can stabilize dispersions of exfoliated sheets and prevent agglomerations in aqueous media The greater the polarity of the surface, the greater the dispersability will be, reported values typically range from 1 to a few mg mL-1

The preparation of graphene oxide dispersions in the organic solvents is highly desirable too, because it may significantly facilitate the practical use of this material

in forming graphene-polymer composites [63] or graphene-based hybrid materials [64] Suspending GO in organic solvent is not so easily accomplished GO was first dispersed in organic solvents via covalent functionalization with different molecules and polymers[65] However the presence of such stabilizers is not desirable as surface modification can complicate the subsequent processing of materials and affect both the mechanical and electronic properties As-prepared GO can form stable dispersion in several organic solvents[59] Suspension of unmodified GO in organic media was achieved by prolonged sonication of fine GO powder[59] or by serial dilution of an aqueous dispersion of aqueous graphene oxide with appropriate organic solvent into a primarily organic media[29]

Instead of removing the functional groups, it is also possible to add other groups to

GO platelets using various chemical reactions via either covalent or non-covalent bonding, resulting in chemically modified graphene (CMG) The chemically reactive oxygen-containing functional groups on GO includes carboxylic acid groups at their edges and epoxy and hydroxyl groups on the basal plane For the carboxylic groups, introduction of substituted amines is one of the most common methods of covalent functionalization and the final products have been investigated for various applications in optoelectronics [66], biodevices[67], drug-delivery [68] and polymer composites [60] [69] The epoxy groups can be easily modified through ring-opening reactions under various conditions, such as nucleophilic attack by the amine Reaction of epoxy group can be also used to cross-link GO platelets via the epoxy groups and strengthen the graphene paper [70] Polymers have also been attached to the surface of GO, typically by grafting-onto or grafting-from approaches [69, 71]

An ideal approach would utilize orthogonal reactions to selectively functionalize one site over another However, demonstration of the selectivity of these chemical transformations remains challenging Reaction with multiple functionalities is possible, and the wide range of variability in the chemical composition of GO presents immense difficulties in isolation and rigorous characterization of the products Despite of these challenges, GO is regarded as a versatile precursor for a wide range of applications as would be discussed later

2.3 Reduced Graphene Oxide

As one can tell from the name, reduced graphene oxide (RGO) is the product from the reduction of graphene oxide (GO) Reduction is one of the most important reactions of graphene oxide as it restores the disrupted sp2 bonding network of GO,

in order to recover graphene-like electrical properties due to the π-network The product of reduction is also called highly-reduced graphene oxide (HRG), and chemically derived graphene (CDG) in literaures However the product has significant structural difference from graphene that would be made apparent For clarity, we will use the term reduced graphene oxide (RGO) from here onwards

2.3.1 Structure

The structure of RGO is similar to GO except for the removal of oxygen-containing functional groups As the reduction produced CO and CO2 instead of O2 [72], it was expected that the removal of the functional groups would not readily lead to restoration of the sp2 bonding Rather, a high concentration of residual defects would remain on the π-network, such as remnant oxygen atoms[73], Stone-Wales defects[50, 53] (pentagon-heptagon pairs) and holes[53] due to loss of carbon in the form of CO and CO2 Improvement in high resolution imaging has allowed direct visualization of the defective nature of RGO as shown in Figure 2 of Reference [53] The high percentage of defect sites limits the electronic quality of RGO and therefore

it is not as effective as graphene in fundamental research of two-dimensional materials

In contrast with GO, the work on proposing calculated models for RGO has been

limited [49] The work by Bagri et al demonstrated the evolution of the atomic

structure of GO as a function of the degree of reduction [74] They observed that RGO is disordered, consisting of holes within the basal plane due to the evolution of

CO and CO2 in agreement with the microscopy observations They also found that residual oxygen in fully reduced GO is a consequence of the formation of highly stable carbonyl and ether groups that cannot be removed without destroying the graphene basal plane These calculations confirm and explain the difficulties in restoring sp2 structures of RGO

2.3.2 Reduction of Graphene Oxide

Reduction of GO to graphene can be carried out via a number of different approaches such as thermal[75], chemical[76] and electrochemical methods[77]

When dispersed in solvents, a variety of chemical means may be used to reduce graphene oxide The most commonly used and one of the first reducing agent to be reported was hydrazine monohydrate[76] The strongly reducing chemical does not react with water, making it perfect for GO reduction One of the disadvantages of using chemical methods of reduction, hydrazine in particular, is the introduction of heteroatomic impurities While effective at removing oxygen functional groups, nitrogen tends to remain covalently bonded to the surface of graphene oxide and affect the electronic structure of the graphene Later sodium borohydride (NaBH4) [78] was found to be more effective The principal impurities are additional alcohols and as an indication of effective reduction, sheet resistance of the product was lowered to 59000Ωsq-1

Other reductants used include gaseous hydrogen[79] and strongly alkaline solutions[80] The use of multiple chemical reductants has also been demonstrated as a route to rigorously reduce graphene [81]

Rather than using a chemical to strip the oxygen-containing functional groups from the surface, it is also possible to create thermodynamically stable carbon oxide species by directly heating graphite oxide in a furnace [82] Exfoliation of the stacked structure occurs through the extrusion of carbon dioxide generated by heating GO whereby the high temperature gas creates enormous pressure within the stacked layers[83] A notable effect of thermal exfoliation is the structural damage caused to the platelets by the releases of carbon dioxide[51] Approximately 30% of the mass of the GO is lost during the exfoliation process, leaving behind vacancies and topological defects throughout the plane of the RGO platelet [84] Defects inevitably affect the electronic properties of the product and may also have an effect

on the mechanical properties of the product, compared to a chemically-reduced sample [70, 85, 86] An alternative heat source for thermal reduction other than furnaces is strong light Photothermal reduction was carried out using camera flash light[87], pulsed femtosecond laser[88], or focused laser beam[89], with the added

advantage of localized reduction and therefore formation of a continuous RGO region in a GO matrix A more detailed account would be discussed in Chapter 4 Another method that shows promise for reduction of GO relies on the electrochemical removal of the oxygen functional groups Electrochemical reduction has been started rather late [90] After depositing thin films of GO on a variety of substrates, electrodes were placed at opposite ends of the film and linear sweep voltammetry was run in a sodium phosphate buffer Reduction began at -0.60V and reached a maximum at -0.87V Rapid reduction was observed during the first 300s, followed by a reduced rate of reduction up to 2000s and finally a decrease to background current levels up to 5000s The conductivity of the film was measured to

be approximately 8500 Sm-1 The reduction mechanism remains unclear, though a reaction pathway was proposed, highlighting the crucial role of hydrogen ions in the buffer solution Electrochemical reduction is effective yet mild at removal of the oxygen functional groups and it precludes the use of harzadous chemicals and the need to dispose of the by-products Up-scalability, however, would be an issue as it would be limited by configuration of the electrodes and substrates

2.3.3 Properties

2.3.3.1 Electrical Conduction:

As an indication of the degree of reduction[91], conductivity of reduced GO has been demonstrated to increase significantly from the nearly insulating GO to 0.1~50 S/cm for reduced GO (RGO) [76, 92, 93] Thermal annealing combined with chemical vapor deposition repair of defects has been able to increase conductivity further to

~100 S/cm[94] Mattevi et al presented a method that combined thermal and

chemical treatment to yield RGO with an impressive conductivity of ~1000 S/cm[50], though still quite far from that of pristine graphene[95]

The difference in the electronic properties of RGO from graphene is due to the incomplete recovery of the π-network due to the presence of residual defects discussed in Section 2.3.1.The reduction of GO removes oxygen functional groups, generates CO2 and CO [31] and creates new sp2 clusters The domain size does not increase, but rather the newly formed domains percolate with present ones to form new conduction pathways The newly created isolated domains can be viewed electronically as the creation of isolated molecular states that aid transport by hopping[27, 55] as explained below

2.3.3.1.1 Conduction in reduced GO

The mechanism of conduction in RGO was revealed from studying the temperature dependence of conductivity[55] The RGO samples showed a significant conductivity decrease by more than 3 orders of magnitude upon cooling from 298 to

4 K, in contrast to unmodified graphene whose conductivity was reduced by less than

1 order of magnitude[1] [96] Best linear fits of the temperature-dependent data were obtained by plotting ln (I/A) versus T-1/3 [55], which serves as an evidence of variable range hopping (VRH) as explained in Section 2.3.3.1.2 VRH involves consecutive inelastic tunnelling processes between two localized states and has been frequently observed in disordered systems, including amorphous carbon [97] The two-dimensional character reflected by the observed T1/3 dependence is consistent with the two-dimensional structure of the sheets In view of the structure of RGO, hopping presumably occurs between the pristine graphene domains which are separated by clusters of defects

Despite their different temperature dependence of electrical conductivity, RGO monolayers showed ambipolar behaviour in the gate dependence of resistance similar

to that of pristine graphene[1] Due to the defective nature of the reduced layers, the mobility of RGO is approximately two orders of magnitude lower than that of graphene, for which values between 3000 and 10000 cm2/Vs have been reported [1, 98] The presence of the intact graphene domains thus accounts for the principal similarity between the gate dependence of the RGO and graphene The resistance maxima occur close to zero gate voltage in low pressure Helium, but shift towards positive gate voltages with prolonged exposure of the samples to the ambient environment The shift of the maxima could be reversed in vacuum Comparable shifts observed for graphene and single-walled carbon nanotubes have been attributed to doping by oxygen and/or water absorption [1, 99]

For multilayered GO sheets, there is sizable conductivity variation between different samples[100] Low bias conductivity measured for bilayer reproducibly exceeded that of the first layer by more than a factor of 2 Assuming that charge negligible vertical conductions, and that the resistance of the first layer remains identical to the monolayer case, it is an indication that the second layer is more conducting than the first one by factors between 2 and 5 No further increase in volume conductivity could be detected for thicker sheets Evidently conduction in the first layer is limited

by the interaction with the substrate Such substrate interactions may account for the difference observed from sheet to sheet It is also interesting to note that thermal annealing of the RGO sheets decreases the resistance of bilayers but not the monolayers, suggesting that the substrate interaction also restricts healing processes requiring atomic diffusion, which may be responsible for the resistance decrease

2.3.3.1.2 Variable Range Hopping Theory

Variable range hopping as explained by Mott [101] is the most commonly proposed theory for electron conduction in RGO

The hopping theory concerns the conduction in a degenerate electron gas in a highly disordered medium Consider what happens to the direct current conductivity when the Fermi energy lies in the range of energies where states are localized, two

mechanisms are possible One is excitation of electrons to , the contribution to the conductivity is where is the value of at and This form of conduction is normally predominant at high temperatures or when is small

The other is thermally activated hopping conduction by electrons in states near the Fermi energy The rate determining process is the hopping of an electron from a state below the Fermi energy to one above The probability of hopping event per unit time

is in turn affected by various factors: Boltzmann factor where W is the difference between the energies of the two states, a factor depending on the phonon spectrum and a factor depending on the overlap of the wavefunctions If localization is very strong, an electron will normally jump to the state nearest in space because the term falls off rapidly with distance The number of electrons jumping a distance R in the direction of the field will be made up of the number of electrons per unit volume within a range kT of the Fermi energy and the difference of the hopping probabilities in the two directions

Nearest-neighbour hopping with an exponential factor can be observed only if states are Anderson localized throughout the whole band If is comparable with

or less than unity, or in all cases at sufficiently low temperatures, the phenomenon of variable-range hopping is to be expected The hopping distance R increases with decreasing temperature This gives a conductivity that depends on T in the limit of low T as in three-dimensions or in two dimensions

2.3.3.2 Solubility

Same as GO, solution dispersions of RGO serve as the key precursor for many graphene-based materials For the conversion from hydrophilic GO to hydrophobic RGO to be successful, both species must be stabilized in a single medium so as to prevent aggregation during the reduction process [44] The first stable RGO dispersion was produced by reducing an aqueous GO dispersion with hydrazine

hydrate in the presence of the amphiphilic surfactant poly (sodium styrenesulfonate) (PSS) [102] Reduction in the absence of PSS surfactant, however, gave graphene aggregates that quickly precipitated from the aqueous medium Subsequently graphene dispersions in THF, CCl4 were obtained by converting the edge carboxylic acid groups into octadecylamides [103] A variety of modifiers and surfactants including isocyanates, ionic liquids, and single-stranded DNA have also been employed to stabilize either aqueous or organic dispersions of graphene

4-Again a stable stabilizer-free dispersion of RGO is always preferred It was achieved

in primarily organic solvents via serial dilutions[29] Hydrazine reduction of a GO suspension in DMF/water (9:1 v/v) solvent produced a stable dispersion that could be further diluted with DMF, DMSO, THF or NMP to give suspensions with as little as

is also development in preparative methods that exclusively generate defect-free graphene sheets in the absence of either oxidative preparation steps or stabilizing groups [67] Graphite powder is sonicated in solvents whose surface energy matches well with that of graphite and allow strong interaction of the solvent with the graphite and minimize the energy required for exfoliation After the sonication, the exfoliated materials are predominantly single sheets, bi- and tri-layer stacks

2.4 Applications of Graphene Oxide-based Materials 2.4.1 Thin Films of GO or RGO

GO can be assembled into free-standing paper-like film by a commercially available flow-directed filtration of the aqueous GO solution[105] The thickness of the paper can be tuned by modifying the amount of graphene oxide in the filtered dispersion, the thinnest one being ~ 2nm and nearly transparent It is non-conductive but exhibits impressive mechanical properties with Young’s modulus of ~30 GPa These mechanical properties can be further improved via modification of the functional groups[70] The modified GO paper can withstand thermal annealing at 300°C without complete loss of mechanical integrity

Thin films of GO ranging from single layer to a few layers were also prepared and investigated as components of field-effect transistors [106] or as efficient electrical conductors after reduction Typically GO thin films were synthesized via drop-casting, spin-coating or layer-by-layer assembly The details would be discussed in Chapter 3

2.4.2 Transparent Conductor

One of the most promising applications of graphene based materials is their use as transparent conductive thin films Such films can be used as electrodes in solar cells, transistors in next-generation displays and highly sensitive detectors for explosives Fabrication from RGO and reduction of GO thin films affords an easy means to prepare large films though they have conductivity limiting defects that are intrinsic to

the oxidation-reduction procedures Li et al compared graphene thin films prepared

from thermally annealed GO thin films against graphene thin films prepared directly from graphene dispersions [107] The former had a sheet resistance nearly three orders of magnitude higher than that of the latter due to the residual oxygen-containing functional groups Increasing film thickness from ~1 to ~3 nm resulted in more pathways for electron transport but at the expense of transparency

As such intense efforts were devoted to produce large-area films using this route that can compete with pristine graphene in conductivity [76, 84] The most convenient method is to reduce pre-fabricated GO thin films via thermal annealing, exposure to

hydrazine gas or a combination of both Thermal treatment of GO thin films produces highly reduced films with sheet resistance values nearly four orders of magnitude lower than those only exposed to hydrazine gas [108] High temperatures were necessary to produce films with low sheet resistance Higher annealing temperatures were also found to increase film transparency in the spectral region

>300nm in wavelength, accounted by increases in the extent of graphitization

2.4.3 Sensing

GO can also exhibit non-covalent binding on the sp2 networks that are not oxidized

or engaged in hydrogen bonding Lu et al reported a DNA sensor that utilized a

non-covalent binding interaction between DNA or proteins and GO platelets [109] demonstrating that this material holds promise as a platform for sensitive and selective detection of DNA and proteins The authors suggested π-π interactions and hydrophobic interactions between GO and doxorubicin hydrochloride (DXR) were the primary interactions that linked the two units together

2.4.4 Precursor to Graphitic Nanostructure

The presence of defects due to oxygen groups creates chemically reactive sites that allow GO to be cleaved into smaller sheets [110] by chemical or physical means, generating nanosized GO or nanoribbons that have markedly different properties from the micrometer-sized counterpart GO can be disintegrated into small fragments and poly-aromatics by sonochemical treatment in acids [111], which can be reconstituted into fullerenes and carbon wires Other methods of fabrication of nanosized GO include hydrothermal cleave of GO to graphene quantum dots in suspension [112]

2.4.5 Precursor to Graphene-based Composites

GO can be utilized as a minor filler component embedded within either a polymer or

an inorganic matrix due to the rich oxygen content on its surface A variety of based nanocomposites have been prepared as thin films, and commonly reduced to RGO composites for conductivity studies Spin-cast films of silica containing up to 11wt% of GO have been prepared and reduced to give transparent conductive layers [113] Paperlike thin films with up to 1.4 wt% of polystyrene have been made by co-

GO-filtration of an aqueous solution containing both GO and polystyrene nanoparticles [87] Drop-cast polyurethane films containing 4.4 wt% of GO were prepared and found to increased Young’s modulus by an impressive ~900% in comparison to films

of the pristine polymer with only minimal decrease in tensile strength [114]

Chemical or thermal reduction of GO, on the other hand, partially restore the graphitic network in the basal plane of RGO, as was discussed previously Consequently, reduced GO have been frequently modified by non-covalent physisorption of both polymers [115] and small molecules[116] onto their basal planes via π-π stacking or van der Waals interactions Few examples of covalent functionalization of RGO exist [117] Most covalent chemical modifications of GO occurred at one or more of the various oxygen-containing functional groups present

in GO Hence the reactivity observed in RGO could be caused by residual functional groups left intact after incomplete reduction

GO can also form hybrid material with other inorganic matrix For example, ions are intercalated between two layers to form a graphite intercalation compound [118] which is an important material for secondary batteries [119]

References

[1] K S Novoselov et al., Science 306, 666 (2004)

[2] J C Meyer et al., Nature 446, 60 (2007)

[3] P R Wallace, Physical Review 71, 622 (1947)

[4] K S Novoselov et al., Nature 438, 197 (2005)

[5] Y B Zhang et al., Nature 438, 201 (2005)

[6] P Blake et al., Applied Physics Letters 91 (2007)

[7] A K Geim, and K S Novoselov, Nature Materials 6, 183 (2007)

[8] D S L Abergel, A Russell, and V I Fal'ko, Applied Physics Letters 91

(2007)

[9] Z H Ni et al., Nano Lett 7, 2758 (2007)

[10] C Casiraghi et al., Nano Lett 7, 2711 (2007)

[11] S Roddaro et al., Nano Lett 7, 2707 (2007)

[12] M Dorn et al., Journal of Applied Physics 108 (2010)

[13] J Henrie et al., Optics Express 12, 1464 (2004)

[14] O C Compton, and S T Nguyen, Small 6, 711 (2010)

[15] V G Makotchenko et al., Carbon 49, 3233 (2011)

[16] T A Land et al., Surface Science 264, 261 (1992)

[17] C Berger et al., Science 312, 1191 (2006)

[18] A Tzalenchuk et al., Nature Nanotechnology 5, 186 (2010)

[19] K S Kim et al., Nature 457, 706 (2009)

[20] A Reina et al., Nano Lett 9, 30 (2009)

[21] L Y Jiao et al., Nature 458, 877 (2009)

[22] D V Kosynkin et al., Nature 458, 872 (2009)

[23] H Santos, L Chico, and L Brey, Physical Review Letters 103 (2009)

[24] R J Beckett, and R C Croft, Journal of Physical Chemistry 56, 929 (1952)

[25] G Charpy, Comptes Rendus Hebdomadaires Des Seances De L Academie

[28] D C Marcano et al., Acs Nano 4, 4806 (2010)

[29] S Park et al., Nano Lett 9, 1593 (2009)

[30] D Li et al., Nature Nanotechnology 3, 101 (2008)

[31] D R Dreyer et al., Chem Soc Rev 39, 228 (2010)

[32] S Park, and R S Ruoff, Nature Nanotechnology 5, 309 (2010)

[33] K P Loh et al., J Mater Chem 20, 2277 (2010)

[34] G Eda, and M Chhowalla, Advanced Materials 22, 2392 (2010)

[35] J I Paredes et al., Langmuir 25, 5957 (2009)

[36] I Jung et al., Journal of Physical Chemistry C 112, 8499 (2008)

[40] G Eda, and M Chhowalla, Nano Lett 9, 814 (2009)

[41] S A Wang et al., Nano Lett 10, 92 (2010)

[42] U Hofmann, and R Holst, Berichte Der Deutschen Chemischen Gesellschaft

72, 754 (1939)

[43] T Nakajima, and Y Matsuo, Carbon 32, 469 (1994)

[44] H Y He et al., Journal of Physical Chemistry 100, 19954 (1996)

[45] A Lerf et al., Journal of Physical Chemistry B 102, 4477 (1998)

[46] H Y He et al., Chemical Physics Letters 287, 53 (1998)

[47] W Gao et al., Nat Chem 1, 403 (2009)

[48] J M Cai et al., Nature 466, 470 (2010)

[49] K P Loh et al., Nat Chem 2, 1015 (2010)

[50] C Mattevi et al., Advanced Functional Materials 19, 2577 (2009)

[51] K N Kudin et al., Nano Lett 8, 36 (2008)

[52] M Ishigami et al., Nano Lett 7, 1643 (2007)

[53] C Gomez-Navarro et al., Nano Lett 10, 1144 (2010)

[54] N R Wilson et al., Acs Nano 3, 2547 (2009)

[55] A B Kaiser et al., Nano Lett 9, 1787 (2009)

[56] W S Hummers, and R E Offeman, J Am Chem Soc 80, 1339 (1958)

[57] G Eda, G Fanchini, and M Chhowalla, Nature Nanotechnology 3, 270

(2008)

[58] J A Johnson et al., Carbon 47, 2239 (2009)

[59] J I Paredes et al., Langmuir 24, 10560 (2008)

[60] S Stankovich et al., Nature 442, 282 (2006)

[61] L Zhang et al., Carbon 47, 3365 (2009)

[62] D W Boukhvalov, and M I Katsnelson, J Am Chem Soc 130, 10697

(2008)

[63] D Cai, and M Song, J Mater Chem 17, 3678 (2007)

[64] D Cai, M Song, and C Xu, Advanced Materials 20, 1706 (2008)

[65] S Stankovich et al., Carbon 44, 3342 (2006)

[66] Y F Xu et al., Advanced Materials 21, 1275 (2009)

[67] N Mohanty, and V Berry, Nano Lett 8, 4469 (2008)

[68] X M Sun et al., Nano Research 1, 203 (2008)

[69] L M Veca et al., Chemical Communications, 2565 (2009)

[70] S Park et al., Acs Nano 2, 572 (2008)

[71] Y F Yang et al., Langmuir 25, 11808 (2009)

[72] I Jung et al., Journal of Physical Chemistry C 113, 18480 (2009)

[73] K A Mkhoyan et al., Nano Lett 9, 1058 (2009)

[74] A Bagri et al., Nat Chem 2, 581 (2010)

[75] H L Wang et al., J Am Chem Soc 131, 9910 (2009)

[76] S Stankovich et al., Carbon 45, 1558 (2007)

[77] Y Y Shao et al., J Mater Chem 20, 743 (2010)

[78] H J Shin et al., Advanced Functional Materials 19, 1987 (2009)

[79] Z S Wu et al., Carbon 47, 493 (2009)

[80] X B Fan et al., Advanced Materials 20, 4490 (2008)

[81] Y Si, and E T Samulski, Nano Lett 8, 1679 (2008)

[82] H P Boehm et al., Zeitschrift Fur Anorganische Und Allgemeine Chemie

316, 119 (1962)

[83] M J McAllister et al., Chemistry of Materials 19, 4396 (2007)

[84] H C Schniepp et al., Journal of Physical Chemistry B 110, 8535 (2006)

[85] S Park et al., Journal of Physical Chemistry C 113, 15801 (2009)

[86] J T Robinson et al., Nano Lett 8, 3441 (2008)

[87] L J Cote, R Cruz-Silva, and J X Huang, J Am Chem Soc 131, 11027

(2009)

[88] Y L Zhang et al., Nano Today 5, 15 (2010)

[89] Y Zhou et al., Advanced Materials 22, 67 (2010)

[90] M Zhou et al., Chemistry-a European Journal 15, 6116 (2009)

[91] I Jung et al., Nano Lett 8, 4283 (2008)

[92] I Jung et al., Journal of Physical Chemistry C 112, 20264 (2008)

[93] S P Pang et al., Advanced Materials 21, 3488 (2009)

[94] V Lopez et al., Advanced Materials 21, 4683 (2009)

[95] J H Chen et al., Nature Nanotechnology 3, 206 (2008)

[96] Y W Tan et al., European Physical Journal-Special Topics 148, 15 (2007)

[97] J Robertson, Adv Phys 35, 317 (1986)

[98] M C Lemme et al., Ieee Electron Device Letters 28, 282 (2007)

[99] R Martel et al., Physical Review Letters 87 (2001)

[100] C Gomez-Navarro et al., Nano Lett 7, 3499 (2007)

[101] N F Mott, and E A Davis, Electronic Processes in Non-crystalline

materials (Oxford, Clarendon Press, 1971)

[102] S Wang et al., Advanced Materials 20, 3440 (2008)

[103] H F Yang et al., Chemical Communications, 3880 (2009)

[104] S H Lee et al., Macromolecular Rapid Communications 31, 281 (2010) [105] D A Dikin et al., Nature 448, 457 (2007)

[106] M Jin et al., Journal of Physics D-Applied Physics 42 (2009)

[107] X L Li et al., Nature Nanotechnology 3, 538 (2008)

[108] H A Becerril et al., Acs Nano 2, 463 (2008)

[109] C H Lu et al., Angewandte Chemie-International Edition 48, 4785 (2009) [110] J L Li et al., Physical Review Letters 96 (2006)

[111] S Wang et al., J Am Chem Soc 131, 16832 (2009)

[112] D Y Pan et al., Advanced Materials 22, 734 (2010)

[113] S Watcharotone et al., Nano Lett 7, 1888 (2007)

[114] D Y Cai, K Yusoh, and M Song, Nanotechnology 20 (2009)

[115] S Stankovich et al., J Mater Chem 16, 155 (2006)

[116] Y X Xu et al., J Am Chem Soc 130, 5856 (2008)

[117] J R Lomeda et al., J Am Chem Soc 130, 16201 (2008)

[118] M S Dresselhaus, and G Dresselhaus, Adv Phys 30, 139 (1981)

[119] M Winter et al., Advanced Materials 10, 725 (1998)

Chapter 3: Experimental Methods

3.1 Sample Preparation

As explained in Section 1.2, suspension of graphene oxide (GO) in various solvents can be readily synthesized via oxidation of commercial grade graphite For application and characterization, it is necessary to fabricate a uniform, single graphene oxide film from the suspensions Such efforts can be traced back before the discovery of graphene, to works on conductive composite films of alternating GO and polyelectrolyte layers [1] Thin films with spotty coverage were prepared by drop casting GO suspensions on a silica substrate or briefly immersing the substrate within the suspensions While films with near monolayer thickness of GO were claimed, coverage >60% was only attained in multilayered structures [1-3] Such films were found to be effective hole conductors but unstable and readily reduced during electrochemical analysis

Recently spin-coating provides an easy means for producing continuous films composed of GO sheets Spin-coating involves depositing a small puddle of excess fluid onto the centre of a substrate and then spinning the substrate at high speed Centrifugal force will spread the fluid to and eventually off the edge of the substrate, forming a thin film on the surface The film thickness depends on nature of the fluid and substrate such as viscosity, rate of evaporation, percent of suspension, surface tension; as well as experimental parameters such as angular velocity and acceleration From earlier works, the thickness of deposited layer can be tuned by varying the concentration of the GO dispersion with 2mg mL-1 giving ~3nm thick films and higher concentrations 12-15 mg mL-1 yielding films ~20nm thick [4] Most spin-coated films are unlikely true monolayer films as they are thicker than individual GO sheets Patterning of GO sheets on the surface of a substrate can be achieved via additional step of templating [5], etching [6]or direct-writing [7]

The fabrication of monolayer GO thin film with good coverage was also reported using Langmuir-Blodgett assembly techniques from basic dispersions of GO [8] The formation of multilayer structures was discouraged due to the repulsive negative charges from deprotonated carboxylic acid groups along the edges of adjacent GO

sheets At high compression, the folded edge layers would partially overlap and interlock to create a continuous film with a relatively uniform monolayer thickness Monolayer was ineffective in our experiments and this technique was therefore not used

To fabricate the GO thin film used in our experiments, 1.5 g of graphite, 1.5 g of NaNO3 and 69 ml of H2SO4 were first mixed and stirred in an ice bath Next, 9 g of KMnO4 was slowly added The solution was kept in room temperature and stirred for 1 h After which, 100 ml of water was added and the temperature was increased

to 90 °C After 30 mins, 300 ml of water followed by 10 ml of H2O2 were slowly added The resultant mixture was filtered and washed by water until the solution pH was about 6 The synthesized GO sheets were dispersed in water: methanol (1:5) mixture and purified with 3 repeated centrifugation steps at 12,000 rpm for 30 mins The purified sample was centrifuged at 8,000 rpm for 30 min to remove the smaller sized GO sheets Then the remaining solution was redispersed in water/methanol mixture with the ratio of 1:5 and centrifuged at 2,500 rpm to recover the GO sheets The GO sheet aqueous solutions were made by centrifugation of the obtained GO solution and followed by re-dispersion into water 1 mg/ml, 4 mg/ml and 6mg/ml

GO aqueous solution was drop-cast or spin coated onto SiO2/Si substrate unless otherwise stated Various thin films with different thicknesses ranging from a few nm

to ~1μm were prepared As an example, spin-coating with a 6mg/ml GO solution on

a SiO2/Si substrate was carried out at 800rpm for 5 minutes to give a film 30~50 nm

in thickness for electrochemical studies All thin films used were thicker than a few nanometers and therefore not monolayers Similar to previous works from literature, monolayer could be deposited by high speed spin-coating, but deposited thin films would have poor coverage and therefore would not be suitable for investigation with electrical methods