ELECTRON AND PHONON TRANSPORT IN NANOSTRUCTURED
EXFOLIATED GRAPHENE NANOPLATELETS AND ITS POTENTIAL IN
THERMAL ENERGY CONVERSION

By
Jinglei Xiang

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Materials Science and Engineering
2012

 

 

ABSTRACT
ELECTRON AND PHONON TRANSPORT IN NANOSTRUCTURED
EXFOLIATED GRAPHENE NANOPLATELETS AND ITS POTENTIAL IN
THERMAL ENERGY CONVERSION
By
Jinglei Xiang
Graphene nanoplatelets (GNP), produced from microwave assisted thermal exfoliation of
graphite intercalated compounds, possess exceptional electrical and thermal properties thanks to
the well preserved conjugation in the graphene basal plane as a result of this production method.
The potential of these intriguing physical properties can only be realized if these nanoparticles
are assembled on a macroscopic scale. However, numerous interfaces thus created between
nanoparticles could have a large impact on the physical properties of GNP based nanocomposites.
This research is dedicated to the investigation of the difference of electron and phonon transport
at the interfaces and how to make use of this unique distinction to nanostructure and assemble
GNP for thermoelectric application which requires a high electrical to thermal conductivity ratio
σ/κ as well as high Seebeck coefficient S of the material.
In the first part of the discussion, the effect of incorporating GNP into a polymer to
improve its electrical and thermal conductivity is discussed. It is found that there are distinct
property enhancements around percolation threshold in the nanocomposites originating from the
different mechanisms of electron and phonon transport across the GNP/polymer interface. In
order to take full advantage of the excellent in-plane physical properties of GNP, a highly

ordered, highly flexible binder free GNP paper was prepared by a simple filtration technology
and annealing that shows much enhanced electrical and thermal conductivity.
The second part of this research is dedicated to separation of electron and phonon
transport in GNP paper by various nanostructuring techniques in order to enhance σ/κ ratio. In
the first approach, monodispersed metal nanoparticle spacers were synthesized on the surface of
GNP particles. The Au/GNP hybrid film prepared by filtration shows decoupled electron and
phonon transport. As a result, the in-plane electrical conductivity increased by 70% while a 7%
reduction in thermal conductivity was observed in comparison to a neat GNP paper. In the
second approach, polyaniline (PANi) was synthesized by in-situ chemical oxidative
polymerization in the presence of GNP where GNP served as a template for nucleation and
growth of PANi. Depending on the composition and protonation ratio of PANi, σ/κ increased
from 2.5 for neat GNP paper to 4.7 for the as-made PANi/GNP and further to 11 for the
reprontonated PANi/GNP film. The optimal thermoelectric properties was achieved at a
protonation of 0.2 with approximately 40wt% of PANi in the nanocomposite, reaching an
electrical conductivity of 59 S/cm, a thermal conductivity of 12W/mK and a Seebeck coefficient
of 33 µV/K at 300K. In the third approach, three dimensional folding of a core-shell
nanostructure based on PANI/GNP paper was adopted to take advantage of the high anisotropic
conduction in a GNP paper, its high flexibility and foldability, but more importantly the strong
phonon scattering at the corner of the folds as compared to a much lower disruption in the
electron flow.
Taken together, the ability to produce and nanostructure these inexpensive graphene
nanoplatelets opens up numerous opportunities for the development of plastic thermoelectrics in
the future. 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Copyright by
Jinglei Xiang
2012
 

 

 
 
 
 
 
 
 
 
 
 
To my parents and beloved fiancée Weijun Lu
 

 

v 
 

ACKNOWLEDGEMENTS

First of all, I would like to express my sincere gratitude to my research advisor Prof.
Lawrence. T. Drzal for his continuous and kind support, patience and guidance throughout my
study. It has been a wonderful experience working in Prof. Drzal's research group and I benefited
a lot from his personal suggestions and words of encouragement.
I would also like to thank my committee member Prof. Don Morelli for his support and
guidance during my graduate studies. I learned a great deal from discussing my research with
him. Likewise, I am also grateful for Prof. Neil Wright, Prof. Jeff Sakamoto and Prof. Phillip
Duxbury for their kind support and valuable advice for me as my committee members.
I thank all the staff members in CMSC, Mike Rich, Per Askeland, Brian Rook, Ed Drown,
Wanjun Liu, Hwanman Park, Hazel-Ann Hosein, Saswata Bose, Inhwan Do, Hiroyuki
Fukushima, Xiaobing Li as well as Karen Lillis, Inger Weitlauf for their generous assistance to
my research.
I am also deeply thankful to my officemates and friends who have shared time with me
during my graduate study and shared with me their joys and sorrows in our lives. Special thanks
go to Anchita Monga, Xian Jiang, Deb Saha, Dee Rollins, Yan Li, Pat Aderhold, Tao Wang,
Huang Wu, Sanjib Biswas, Kevin Zhou, Chang Liu, Shaowen Ji, Leyun Wang, Xueling Fei and
Dafei Kang.
Lastly, I would like to thank my parents and my beloved fiancée for their unconditional
love and support in the course of my life. My life is full of hope and joy with you around.
 

 

vi 
 

TABLE OF CONTENTS

LIST OF TABLES ....................................................................................................................... xi
 

LIST OF FIGURES .................................................................................................................... xii
 

Chapter 1 Introduction and Background ................................................................................... 1 
1.1 The rise of graphene: ............................................................................................................... 1
1.2 An alternative material: exfoliated graphene nanoplatelets (GNP) ......................................... 4
1.3 Multifunctional graphene-based polymer nanocomposites (electrical and thermal properties) 6
1. 3. 1 Electron transport: ..................................................................................................... 10
1. 3. 2 Phonon transport:....................................................................................................... 13
1.4 Bi-continuous paper like graphene based nanocomposites ................................................... 16
1.5 Background in thermoelectric energy conversion materials:................................................. 17
1.6 Motivation and objective of research:.................................................................................... 21
References: .................................................................................................................................... 23
 

Chapter 2 Investigation of GNP in Improving the Thermal and Electrical Conductivity of
Paraffin Wax and Its Use in Thermal Energy Storage ........................................................... 32
2.1 Abstract: ................................................................................................................................. 32
2.2 Introduction ............................................................................................................................ 32
2.3 Experimental: .......................................................................................................................... 34
2.4 Characterization: .................................................................................................................... 36
2.5 Results and discussions: ......................................................................................................... 40
2.5.1 Morphology .................................................................................................................. 40
2.5.2 Electrical conductivity of paraffin/GNP nanocomposites: ........................................... 43
2.5.3 Thermal conductivity paraffin/GNP nanocomposites: ................................................. 44
2.5.4 Thermal interface resistance: ........................................................................................ 46
2.5.5 Effect of GNP particles on other thermo physical properties of nanocomposites: ...... 51
2.5.6 Thermal gravimetric analysis: ...................................................................................... 52
vii 
 

2.6 Conclusions:........................................................................................................................... 54
Acknowledgements ....................................................................................................................... 55
References ..................................................................................................................................... 56
 

Chapter 3 Thermal Conductivity of a Monolayer of Exfoliated Graphene Nanoplatelets
Prepared by Liquid-Liquid Interfacial Self-Assembly............................................................ 59
3.1 Abstract: ................................................................................................................................. 59
3.2 Introduction ............................................................................................................................ 60
3.3 Experimental details: .............................................................................................................. 61
3.4 Results and discussions:.......................................................................................................... 63
3.4.1 Scanning electron microscopy: ..................................................................................... 63
3.4.2 Optical Transmission and Absorption of GNP Monolayers:........................................ 66
3.4.3 Confocal Micro-Raman Spectroscopy Analysis: ......................................................... 68
3.4.5 Calculation of the Thermal Conductivity of GNP Monolayers:................................... 70
3.5 Conclusions: ............................................................................................................................ 73
Acknowledgement: ....................................................................................................................... 74
References: .................................................................................................................................... 76
 

Chapter 4 Thermal Conductivity of Exfoliated Graphene Nanoplatelet Paper ................... 78 
4.1 Abstract ................................................................................................................................... 78
4.2 Introduction:............................................................................................................................ 78
4.3 Experimental ........................................................................................................................... 80
4.4 Results and discussions:.......................................................................................................... 83
4.4.1 Morphology .................................................................................................................. 83
4.4.2 Thermogravimetric analysis of GNP paper .................................................................. 88
4.4.3 Thermal diffusivity and thermal conductivity of GNP paper under various treatments
............................................................................................................................................... 89
4.5 Conclusions:........................................................................................................................... 94
References ..................................................................................................................................... 97
 

viii 
 

Chapter 5 Electron and Phonon Transport in Gold Nanoparticle Decorated Graphene
Nanoplatelet Nanostructured Paper ......................................................................................... 99 
5.1 Abstract: .................................................................................................................................. 99
5.2 Introduction:............................................................................................................................ 99
5.3 Experimental: ........................................................................................................................ 102
5.4 Results and discussions:........................................................................................................ 104
5.4.1 Morphology ................................................................................................................ 104
5.4.2 Thermal diffusivity and thermal conductivity of Au/GNP hybrid paper ................... 113
5.4.3 Electrical properties of Au/GNP hybrid paper ........................................................... 119
5.5 Conclusion: ........................................................................................................................... 123
Supporting information ............................................................................................................... 125
References ................................................................................................................................... 130
Chapter 6 Templated Growth of Polyaniline on GNP and Its Thermoelectric Properties 134 
6.1 Abstract: ................................................................................................................................ 134
6.2 Introduction:.......................................................................................................................... 134
6.3 Experimental: ........................................................................................................................ 137
6.4 Results and discussions:........................................................................................................ 143
6.4.1 Synthesis of PANi/GNP and characterizations: ......................................................... 143
6.4.2 Thermoelectric properties of PANi/GNP: .................................................................. 157
6.4.3 Effect of protonation on thermoelectric properties of PANi/GNP nanocomposites: . 164
6.5 Conclusion: ........................................................................................................................... 168
Acknowledgement: ..................................................................................................................... 169
References ................................................................................................................................... 170
Chapter 7 Improving Thermoelectric Properties of Exfoliated Graphene Nanoplatelets
Paper by Folding ....................................................................................................................... 174 
7.1 Abstract: ................................................................................................................................ 174
7.2 Introduction:.......................................................................................................................... 174
7.3 Experiments: ......................................................................................................................... 176
7.4 Results and discussions:........................................................................................................ 181
ix 
 

7.4.1 GNP paper folded ....................................................................................................... 181
7.4.2 Electronic conduction of the folded paper .................................................................. 186
7.4.3 Folded PANi/GNP paper as a p-type thermoelectric material ................................... 188
7.5 Conclusion: ........................................................................................................................... 196
References: .................................................................................................................................. 197
Chapter 8 Summary and Recommended Studies .................................................................. 200 
8.1 Summary: .............................................................................................................................. 200
8.2 Recommended future works: ................................................................................................ 204
8.2.1 Solvent processing of PANi/GNP: ............................................................................. 204
8.2.2 Carrier transport analysis for PANi/GNP: .................................................................. 205
References: .................................................................................................................................. 208

x 
 

LIST OF TABLES
 

Table 2-1: Heating and cooling characteristics of GNP-15 composite PCM (average of three
samples) ........................................................................................................................................ 52 
Table 2-2: Heating and cooling characteristics of GNP-1 composite PCM (average of three
samples) ........................................................................................................................................ 52 
Table 4-1: Density variation of GNP paper under different conditions........................................ 90 
Table 5-1: Average densities of GNP and Au/GNP hybrid papers measured as-made, annealed,
annealed & cold compacted (standard deviation < 0.05g/cm3). ................................................. 113 
Table 6-1: Typical atomic percentage of nitrogen and chlorine species on PANi/GNP samples at
different protonations prepared at 0.05M aniline and estimation of the protonation level......... 152 
Table 6-2: The thermoelectric properties, σ/κ and ZT of neat PANi, neat GNP and PANi/GNP
(50mM aniline, as-made) and PANi/GNP(50mM aniline, reprotonated) at room temperature. 168 
Table 7-1: Electrical resistance measurement on the folded paper (1 fold between voltage probes)
..................................................................................................................................................... 187 
Table 7-2: Electrical resistance measurement on the folded paper (2 folds between voltage
probes)......................................................................................................................................... 187 
Table 7-3: Electrical conductivity of the folded PANi/GNP paper with and without compaction
..................................................................................................................................................... 192 
 

xi 
 

LIST OF FIGURES
 

Figure 1-1: Electrical percolation, as applied to conductive composites. The formation of the first
complete particle linkage results in a significant drop in resistivity at Vc. For interpretation of the
references to color in this and all other figures, the reader is referred to the electronic version of
this dissertation. ............................................................................................................................ 13 
Figure 1-2: Schematic representation of Seebeck effect when a sample is subjected to a
temperature gradient, in this case, holes are the dominant carriers. ............................................. 18 
Figure 1-3: Schematic representation of a thermoelectric unicouple consisting of a p and n type
semiconductor materials connected electrically in series and thermally in parallel operating under
power generation mode (left) or refrigeration mode (right) ......................................................... 19 
Figure 2-1: (a) Photograph of the customized aluminum mold; (b) Photograph of Bolling two roll
mill; (c) Scheme for the sample preparation routes: casting and two roll milling. ....................... 35 
Figure 2-2: (a) Schematics for in plane thermal conductivity measurement of the two roll milled
sample; (b) Photograph of the sample prepared from compressing the two roll milled sheets to a
desired thickness (1 inch) and a circular cross section of that was produced by a punch. ........... 39 
Figure 2-3: a: 4wt% GNP-1/paraffin by casting; b: 4wt% GNP-15/paraffin by casting; c: 4wt%
GNP-15/paraffin by two roll mill; d: 10wt% GNP-15/paraffin by two roll mill; e: enlarged edge
view of the nanoplatelets (scale bar 100 nm); f: enlarged edge view of one platelet (scale bar: 10
nm) ................................................................................................................................................ 41 
Figure 2-4: Electrical resistivity of GNP/paraffin composite PCM.............................................. 44 
Figure 2-5: (a) Through plane thermal conductivity of GNP/paraffin prepared by casting; (b)
Through-plane (left column) and in-plane (middle column) thermal conductivity of GNP15/paraffin prepared by two roll mill and through-plane thermal conductivity by casting (right
column). ........................................................................................................................................ 45 
Figure 2-6: overview of the dominate factors in interfacial thermal conductance in carbon based
nanocomposites as well as possible techniques to reduce interface resistance. Adapted from ref.
15................................................................................................................................................... 47 
Figure 2-7: Thermal conductivity of paraffin/GNP-15 fitted with effective medium theory below
electrical percolation. .................................................................................................................... 50 
Figure 2-8: (a) TGA analysis of GNP-1/paraffin nanocomposites in N2. (b) TGA analysis of
GNP-15/paraffin nanocomposites in N2 ....................................................................................... 53 
Figure 3-1: (a) SEM image of a LSA GNP monolayer film supported on a glass substrate. (b)
SEM image of a HSA GNP monolayer film supported on a glass substrate. ............................... 65 
xii 
 

Figure 3-2: Percent transmission of light for LSA and HSA monolayers scanning from 350 nm to
1000 nm. Inset is the photo of LSA and HSA coated monolayers supported on glass substrates as
well as the blank slide. .................................................................................................................. 66 
Figure 3-3: Raman G band shift of GNP with temperature and linear curve fitting with Microsoft
Excel. Two measurements were conducted indicated by blue and red marks. ............................. 67 
Figure 3-4: (a) Raman G peak shift of the LSA monolayer at different laser power levels. (b)
Raman G peak shift of the HSA monolayer using different laser power levels. .......................... 69 
Figure 3-5: (a) Local temperature of an individual GNP at the center of a LSA monolayer vs. the
absorbed laser power with x50 and x100 objectives. The solid lines represent the linear curve
fittings. (b) Local temperature of an individual GNP at the center of a HSA monolayer vs. the
absorbed laser power with x50 and x100 objectives. The solid lines represent the linear curve
fittings. .......................................................................................................................................... 72 
Figure 4-1: Typical Raman spectrum of GNP particles showing D peak, G peak and 2D peak. . 81 
Figure 4-2: Schematic representation of the GNP paper making process by vacuum assisted self
assembly and actual photos of GNP paper peeled off the ANODISC filter paper after drying. .. 84 
Figure 4-3: (a) SEM micrographs of cross section of the as-made GNP-15 paper (scale bar: 10
µm), (b) annealed and cold pressed GNP-15 paper (scale bar: 10 µm), (c) as-made GNP-1 paper
(scale bar: 5 µm), (d) annealed and cold pressed GNP-1 paper (scale bar: 1 µm). ...................... 85 
Figure 4-4: (a): TGA weight loss curves of GNP-15 paper before and after annealing, (b) TGA
weight loss curves of GNP-1 paper before and after annealing.................................................... 88 
Figure 4-5: (a) Through-plane thermal diffusivity of GNP-1 and GNP-15 papers. (b) Throughplane conductivity of GNP-1 and GNP-15 papers. (c) In-plane diffusivity of GNP-1 and GNP-15
papers. (d) In-plane conductivity of GNP-1 and GNP-15 papers. (note: left column represents
GNP1, right column represents GNP15) ...................................................................................... 92 
Figure 5-1: Schematics of experimental procedures and the interactions between the metal ions
and the active sites on PEI coated GNP and formation of Au nanoparticles .............................. 105 
Figure 5-2 (a): Low magnification SEM image of Au nanoparticle decorated GNP prepared at
0.3wt% PEI concentration (scale bar: 1 µm). b: High magnification SEM image of Au
nanoparticle decorated GNP prepared at 0.3wt% PEI (scale bar: 100 nm) c: Low magnification
SEM image of Au nanoparticle decorated GNP prepared at 0.6wt% PEI concentration (scale bar:
1 µm). d: High magnification SEM image of Au nanoparticle decorated GNP prepared at 0.6wt%
PEI concentration (scale bar: 100nm) ......................................................................................... 107 
Figure 5-3: (a) Size and distribution of Au nanoparticles on GNP prepared at 0.3wt% PEI
concentration. (b) Size and distribution of Au nanoparticles on GNP prepared at 0.6wt% PEI
concentration. .............................................................................................................................. 109 

xiii 
 

Figure 5-4: Schematics of the experiment showing the Au nanoparticle decorated GNP dispersed
in water for subsequent vacuum assisted self assembly. ............................................................ 110 
Figure 5-5: Photos of the as-made GNP paper showing flexibility and mechanical robustness
under bending and the Au/GNP hybrid sample used for thermal diffusivity measurement (1 inch
diameter) ..................................................................................................................................... 111 
Figure 5-6: (a) Representative cross sectional SEM image of “as-made” GNP paper (scale bar: 5
µm). (b) as-made Au/GNP hybrid paper prepared at 0.3wt% PEI (scale bar: 5 µm). ................ 111 
Figure 5-7: (a) Through-plane thermal diffusivity of GNP and Au/GNP paper. b: Through-plane
thermal conductivity of GNP and Au/GNP paper. c: In-plane thermal diffusivity of GNP and
Au/GNP paper. d: In-plane thermal conductivity of GNP and Au/GNP paper. Note: left column
represents GNP15, middle column represents GNP15/Au(0.3wt% PEI), right column represents
GNP15/Au(0.6wt% PEI) ............................................................................................................ 116 
Figure 5-8: (a) In-plane electrical conductivity of anneal & cold compacted GNP and Au/GNP
paper. (b) Through-plane electrical resistivity of anneal & cold compacted GNP and Au/GNP
paper............................................................................................................................................ 121 
Figure 5-9: TGA curve of neat PEI in air ................................................................................... 125 
Figure 5-10: TGA curves of annealed GNP paper, as-made GNP paper and as-made Au/GNP
paper in air. ................................................................................................................................. 126 
Figure 5-11: Example SEM images analyzed by Image-Pro to determine the distribution of
nanoparticle size (scale bar 100nm) ............................................................................................ 127 
Figure 5-12: Particle size distribution obtained from Figure 5-11. ............................................ 128 
Figure 5-13: Schematic representation of through-plane resistivity measurement of thin film
sample ......................................................................................................................................... 129 
Figure 6-1: A schematic representation of replacing the Au nanoparticle spacers in the highly
ordered GNP paper with a layer of polymer. .............................................................................. 135 
Figure 6-2: Schematic representation of the experimental setup for measuring the Seebeck
coefficient and thermal conductivity of thin film sample. Inset is the photo of the actual sample
mounted to the "L-shaped" copper block.................................................................................... 141 
Figure 6-3: Schematic representation of in-situ polymerization of aniline in the presence of GNP
particles and a photo showing the flexibility of the final nanocomposite film. .......................... 145 
Figure 6-4: (a,b) SEM micrographs of neat polyaniline salt showing the granular structure and
small needle-like nanostructure growing out of the surface. (c,d) surface morphology of
PANi/GNP particles (in-situ polymerization of aniline in the presence of GNP particles in 1M
HCl at 0.03M aniline). (e,f) surface morphology of PANi/GNP particles (in-situ polymerization
of aniline in the presence of GNP particles in 1M HCl at 0.05M aniline). ................................ 146 
xiv 
 

Figure 6-5: a: TGA of neat polyaniline salt, neat GNP, and PANI/GNP synthesized at 0.03M and
0.05M initial concentration of aniline. b: Derivative of weight loss vs. temperature for neat PANi,
neat GNP, and PANi/GNP synthesized at 0.03M and 0.05M initial concentration of aniline. .. 150 
Figure 6-6: (a) Regional scans of N1s core-level spectra of as-made polyaniline salt showing the
presence of imine nitrogen(-N=), amine nitrogen (-NH-) and positively charged nitrogen (N+) (b):
regional scan of Cl2p core-level spectra of as-made polyaniline salt showing Cl anions. (c):
regional scans of N1s core-level spectra of as-made PANi/GNP prepared at 0.05M of aniline. (d):
regional scans of Cl2p of PANi/GNP(0.05M). ........................................................................... 153 
Figure 6-7: Raman shift of neat PANi powder, neat GNP film and the as-made PANi/GNP film
prepared at 50mM aniline concentration. ................................................................................... 157 
Figure 6-8: (a) Seebeck coefficient of GNP/PANi nanocomposites prepared at different aniline
concentrations (0.03M and 0.05M) as well as the performance for neat GNP paper and neat
polyaniline salt measured from 200K to 300K. (b). Thermal conductivity of GNP/PANi (0.03M
and 0.05M) as well as neat GNP paper and neat polyaniline salt from 200K to 300K. ............. 160 
Figure 6-9: (a) Proposed polymerization process for polyaniline in the absence of GNP and the
charge transport mechanism by inter and intrachain hopping. (b) Proposed polymerization of
PANi in the presence of GNP and the proposed charge transport mechanism in the composites.
..................................................................................................................................................... 162 
Figure 6-10: Electrical conductivity of GNP/PANi (0.03M and 0.05M) as well as neat GNP
paper and neat polyaniline salt at room temperature. ................................................................. 163 
Figure 6-11: (a) Seebeck coefficient of PANi/GNP prepared at 0.05M aniline with different
protonation levels. (b) Thermal conductivity of PANi/GNP (50mM) with different levels of
protonations................................................................................................................................. 165 
Figure 6-12: Electrical conductivity of PANi/GNP samples prepared at 0.05M aniline with
different protonation ratio. .......................................................................................................... 167 
Figure 7-1: Schematic representation of experimental procedures to prepare PANi coated GNP
paper by rapid mixture polymerization. ...................................................................................... 176 
Figure 7-2: custom made mold for paper folding, and stacking of folded paper strips into desired
thickness for subsequent measurements. .................................................................................... 178 
Figure 7-3: (a) schematic representation of experiment setup for Seebeck coefficient, thermal
conductivity, electrical conductivity measurement for the folded sample. (b) a photo of the
folded PANi/GNP paper mounted on the cryostat before measurement. ................................... 179 
Figure 7-4: schematic representation of electron transport within the folded paper, the air gaps
between individual fold is exaggerated....................................................................................... 182 
Figure 7-5: Photos of flat GNP paper, GNP paper cut to strips, folding the GNP strips on the
custom-made acrylic mold, and the folded paper.(from top left to bottom right) ...................... 183 
xv 
 

Figure 7-6: (a) SEM micrographs of a folded strip at low magnification (scale bar: 1mm), (b) the
convex side of the folding edge (scale bar: 100µm), (c) the concave side of the folding edge
(scale bar: 100 µm), (d) a magnified view of the tip of the folding edge (scale bar: 20 µm) ..... 184 
Figure 7-7: (a) Photos of four-point electrical resistance measurement on the slightly folded GNP
paper and (b) a heavily folded GNP paper with one fold between the voltage probes............... 186 
Figure 7-8: (a) Photo of the as-synthesized PANi/GNP paper (5 cm × 5 cm). (b) An overview of
the surface morphology of PANi/GNP paper by SEM (scale bar: 10 µm). (c) SEM micrograph of
PANi nanofibrils grown on the surface of GNP (scale bar: 100 nm). (d) folding edge of the
PANi/GNP paper showing the interconnected "core" (scale bar: 100 µm). ............................... 189 
Figure 7-9: TGA analysis of the as synthesized PANi/GNP paper and the neat polyaniline in air.
..................................................................................................................................................... 191 
Figure 7-10: :(a) Seebeck coefficient of the folded PANi/GNP paper (pressed vs. unpressed)
from 270K to 320K (b) Thermal conductivity of the folded PANi/GNP paper (pressed vs.
unpressed) from 270K to 320K................................................................................................... 193 
Figure 7-11: schematic representation of the folded paper for estimating the bulk electrical
conductivity and mathematic derivation. .................................................................................... 194 
Figure 8-1: Schematics and actual photo of the quantum design versalab module and wiring,
CH1 measures the Hall voltage, CH2 measures the in-plane resistivity of the sample. z direction:
coming out of the paper. ............................................................................................................. 206 

xvi 
 

Chapter 1 Introduction and Background
 

1.1 The rise of graphene:
Graphene consists of a single layer of atomic carbon arranged in a honeycomb lattice
structure. Although graphene has been known for over one hundred years, the recent attention
given to graphene is largely due to the Nobel Prize winning work by Geim and Novoselov in
2004 [1-5]. The graphene structure is the building block for graphitic carbon materials of all
other dimensionalities (such as 0D buckyball, 1D carbon nanotubes and 3D graphite). Ever since
its discovery, graphene or few-layer graphene (FLG) quickly attracts great interest from research
communities all over the world. It has been shown that graphene has exceptional electrical,
thermal and mechanical properties. For example, it is one of the strongest materials with a
Young's modulus exceeding 1 TPa [6]. An ideal graphene sheet has extremely high electrical
conductivity and it was shown that charge carriers behave like massless Dirac Fermions with
carrier mobility measured between 3000 and 10,000 cm2/Vs at room temperature [7,8]. Thermal
conductivity of graphene, measured by a non-contact optical method, is reported to reach values
of up to (5.3+0.48)×103 W/mK [9]. Research conducted in the past few years has shown great
promise for the use of graphene in multifunctional composites, electrochemical energy storage
and electronics, as well as in dye sensitized solar cells [10-13]. It is expected that this novel
lightweight nanomaterial will have a huge impact on our society given all these possible
technological applications demonstrated so far.
Large-scale, economical production of graphene or FLG is a significant challenge for its
wide spread application. Currently, there are four major routes to synthesize graphene or FLG.
First, graphene has been prepared by micromechanical cleavage from a highly oriented pyrolitic
1 
 

graphite (HOPG). This is also widely known as the "scotch tape" method in which a single sheet
of graphene is lifted from the HOPG because of the weak van der Waals interaction between
adjacent graphene sheets in bulk graphite [2]. Second, graphene has been grown by chemical
vapor deposition of hydrocarbons on a metal substrate (such as Ni and Cu) [14,15]. Third,
epitaxial graphene was grown on single-crystal silicon carbide by vacuum graphitization [16].
Last, a colloidal suspension of graphene sheets can be prepared by converting graphite to
graphite oxide (GO) followed by reduction with strong reducing agents. Among all these
methods, the "scotch tape" method is capable of producing high quality graphene, but the yield is
extremely low. It is not a suitable method to adopt if a large quantity of graphene is needed. The
chemical vapor deposition and vacuum graphitization methods are energy intensive, producing
single or FLG with a large size best suited for the electronics industry [17,18].
Production of graphene through reduction of GO is a popular route because of its high
yield. GO is produced by an oxidative treatment of graphite using a method developed by
Hummers, which is less time consuming and hazardous compared with earlier approaches [19].
In general, oxidation of graphite is accomplished by treating graphite with a water-free mixture
of concentrated sulfuric acid, sodium nitrate and potassium permanganate. It takes less than two
hours to complete the oxidation at a temperature less than 45°C given reasonable safety
precautions during the process. After oxidation, the interlayer distance between adjacent
graphene sheets expands as a result of accommodation of various oxygen functional groups such
as the epoxide, hydroxyl as well as carbonyl and carboxyl groups introduced during oxidation
[20]. Depending on the degree of drying, the interlayer distance varies from 0.63 nm for dried
GO to 1.2 nm for hydrated GO [21,22]. Oxidation transforms the highly hydrophobic graphene
basal plane to a more hydrophilic derivative. Therefore, it is also reasonable to think of GO as a

2 
 

graphite intercalation compound with covalently bound oxygen and non-covalently bound water
molecules between graphene layers. The increased hydrophilicity of GO makes it easy to
disperse it in water with ultrasonication.
Much research has been done in reducing GO chemically to graphene [23]. One of the
most commonly reported methods is chemical reduction of GO in the presence of strong
reducing agents such as hydrazine monohydrate, sodium borohydride (NaBH4) etc. During the
reduction process, the yellow brown color of GO dispersion gradually turned black due to
elimination of oxygen functional groups and partial restoration of sp2 carbon. The basal plane
becomes increasingly hydrophobic and eventually precipitated out forming irreversible
agglomerates as a result of strong van der Waals force between individual graphene. The
resulting graphene agglomerates appear to be insoluble in water and organic solvents, making
further processing rather difficult [24]. The restacking of graphene sheets can be effectively
prevented if a charged polyelectrolyte such as PSS (polystyrenesulfonate) or a small organic
molecule (e.g. pyrenebutyric acid) is present during the reduction process [25,26]. The
adsorption of the polyelectrolyte or small molecules on the basal plane of graphene surface due
to π electron interaction overcomes the strong van der Waals force between graphene sheets.
However, foreign stabilizers are not always desirable in many applications. Recently, Li et al
found that GO reduced by hydrazine can be electrostatically stabilized by the negatively charged
carboxylic acid group without using a polymer or a surfactant and the same functional group
which may not be easily reduced by hydrazine can facilitate the dispersion of reduced graphene
oxide (RGO) given careful control over the pH of the solution as well as the initial concentration
of GO [27].

3 
 

Apart from colloidal reduction of GO to graphene using strong reducing agents, thermal
treatment of GO is another route to produce thin graphene nanosheets. Rapid heating to 1050°C
(>2000 °C/min) resulted in expansion and delamination of GO caused by vaporization of
intercalated water molecules and thermal pyrolysis of oxygen containing functional groups [2831]. The functionalized graphene sheets produced by this method have similar C/O ratio
(C/O:10.3 by elemental analysis) to graphene reduced in the presence of hydrazine. However, it
is expected that structural defects formed during the exfoliation process, leaving behind
numerous vacancies on the RGO sheet which might adversely affect its properties.
After the reduction, the change of color from yellow brown typical of GO to black
indicates partial restoration of electronic conjugation in the reduced material, making it
electrically conductive. Jung et al. measured the electrical conductivity of individually reduced
graphene oxide with three different reduction routes (i.e. vacuum annealing, hydrazine reduction
or a sequential combination of the two). It is found that an additional thermal treatment led to a
further reduction in the degree of functionalization and dramatically increased the electrical
conductivity. The reported range of electrical conductivity of these individually reduced GO is
only 33-85 S/m [32]. One of the reasons for this low level of conductivity is the remaining
oxygen groups clustering into rows and islands, resulting in graphitic regions intermixed with
islands of oxygen-functionalized species [33]. These point defects act as scattering centers,
changing the ballistic transport of electrons to a "hop-like" transport similar to electron tunneling.

1.2 An alternative material: exfoliated graphene nanoplatelets (GNP)
The unique properties of graphene are likely to be compromised because of the
irreversible structural damages incurred during oxidation and reduction process of GO. It is by
4 
 

far almost impossible to restore the perfect sp2 carbon in the basal plane of graphene once
graphite has been oxidized. Researchers are currently investigating alternative routes to produce
thin sheet of graphene while still maintaining the strong and robust carbon scaffold intact during
the production process. Research led by Drzal et al. has successfully produced exfoliated
graphene nanoplatelets (GNP) of tens of nanometers in thickness and 100-10000 nm in diameter
using a microwave assisted thermal exfoliation followed by pulverization [34,35]. The idea is
similar to splitting graphite oxide to produce functionalized graphene sheet. In this case, the
starting material is graphite intercalated compound (GIC). The intercalates are usually a
combination of nitric and sulfuric acids at certain proportion. The advantage of using microwave
irradiation instead of conventional furnace is that graphite is an excellent microwave absorber
and heats instantly. The rapid heating rate caused vaporization of intercalated acids giving rise to
a huge pressure that overcomes the interlayer van der Waals interaction. The volume of the
exfoliated graphite is several hundred that of the original GIC and is known as "worm". The
accordion-like worms that are several hundred micrometers were then broken down to smaller
nanoplatelets by controlled combination of ultrasonication and ball milling. The diameter of
GNP after ultrasonication for 2 h is reduced to approximately 15 µm known as GNP-15. GNP-15
can be further downsized to approximately 1 µm or less with rotary ball milling for 72 h. The
thickness of individual nanoplatelets is less than 10 nm as measured by TEM. The surface area
of GNP was measured by Brunauer-Emmet-Teller (BET) using N2 adsorption at 77K which is
approximately 20-40 m2/g for GNP-15 and 100-130m2/g for GNP-1. Although these particles
consist of multiple layers of graphene, It still has an extremely large length-to-thickness ratio
which can be very desirable in many applications. Most importantly, the intercalation and
exfoliation process leave the basal plane of graphene intact as evidenced by very high C/O ratio

5 
 

in X-ray photoelectron spectroscopy (XPS) and Raman spectroscopy. It is reasonable to expect
that the intrinsic properties of graphene such as electrical and thermal conductivity persist in
these nanoplatelets.
While the single or few-layer graphene has promising application in electronics industry
due to their unique band structure (zero bandgap) that can be tuned by applying a gate bias
[36,37], the band structure of as-produced GNP particles approaches that of the bulk graphite,
which might restrict its application in electronics. However, the ease of mass production at a
relatively low cost of these GNP particles might find its entry point to the market through
composites application, much earlier than all other graphene nanomaterials. Multifunctional
graphene-based polymer composites is one of the most exciting research topics during the past
decade because of their tremendous potential in various technological applications.

1.3 Multifunctional graphene-based polymer nanocomposites (electrical and
thermal properties)
Polymers and their composites have found wide spread applications in automobiles,
aerospace, consumer products, electronic packaging, etc. due to their low cost, light-weight,
good chemical resistance and high performance [38,39]. The properties of a polymer matrix can
be completely modified by adding micrometer or nanometer sized fillers. For example, the
insulating polymer can be made conductive with the addition of metal particles, carbon blacks,
carbon fibers or carbon nanotubes, which could be very useful for electrostatic charge dissipation,
EMI shielding, and bipolar plates in fuel cells applications [40,41]. On the other hand, the
thermal conductivity of polymers is very low, which poses great limitations in certain
applications that require fast heat dissipation such as in electronic packaging, engine components

6 
 

in automobiles or in airplanes. Addition of thermally conductive metal particles, carbon fibers
and carbon nanotubes have been proved useful in increasing the thermal conductivity of the
resulting composites [42]. In addition, mechanical properties, gas permeability as well as fire
retardancy can also be greatly enhanced by incorporation of the fillers into the polymer host
[43,44].
The use of graphite based fillers can be traced back to Aylsworth [45,46] who first
developed and proposed exfoliated graphite as a reinforcement for polymers in the 1910s while
others proposed the dispersion of intercalated graphite in polymer resins by conventional
composite processing techniques [47]. It was only recently that carbon nanomaterials such as
carbon nanotubes and graphene nanosheets thanks to their exceptional physical properties started
to gain momentum in replacing conventional fillers to introduce multifunctionalities to polymer
hosts. However, the high surface area of these carbon nanomaterials also gives rise to a strong
van der Waals interaction between individual nanoparticles. Therefore, their maximum
contribution to the properties of composites can only be achieved if they remain reasonably
dispersed inside the polymer without significant agglomeration. Typically, thermoplastic
polymer composites were made by one of the three processing routes: solvent blending, in situ
polymerization and melt extrusion.
1. Solvent blending:
In this approach, graphene based fillers are dispersed in a suitable solvent such as water, acetone,
chloroform, tetrahydrofuran (THF), dimethyl formamide (DMF) or toluene. The polymer was
then added and solubilized. Ultrasonication was usually used to homogenize the dispersion of
nanofillers in the polymer host. It is also important to note that long time exposure to
ultrasonication can introduce defects to graphene sheets which might be detrimental to the

7 
 

properties. During solvent blending, graphene sheets were coated with polymers and remained
separate. Solvent was then removed by evaporation at a temperature around or below the boiling
point of the solvent during which graphene sheets reassemble, sandwich the polymer to form
nanocomposites. However, several possible drawbacks associated with this approach is that the
nanofillers tend to agglomerate during slow solvent evaporation, resulting in inhomogeneous
distribution of graphene sheets in the polymer. In addition, residue solvent molecule might
remain on the surface of GNP, which cannot be easily removed. Various polymer composites
such as expanded graphite (EG)-PMMA[48], functionalized graphene-PS[49], graphenePVA[50], graphene-PVC[51], PVDF-reduced graphene [52] have been prepared using this
technique.

2. in situ polymerization
In this approach, graphene based fillers were dispersed together with monomers in the presence
of a solvent. Polymerization reaction took place with the addition of certain amount of initiators
at a controlled temperature for certain amount of time. This technique enables not only the
covalent bonding between functionalized graphene with the polymer matrix but also non
covalent bonding such as PMMA-GO [53], PP-GO nanocomposites [54]. For graphite-based
layered structure in a polymer matrix, this technique further facilitated the exfoliation of
expanded graphite into thinner platelets by intercalation of the monomers, forming welldispersed graphene in the polymer. While this approach has offered a good technique to achieve
a homogeneous dispersion of graphene inside the polymer host and possibly enhanced
interactions, it requires careful control over the conditions used in the synthesis of
nanocomposites such as reaction temperature, concentration of monomer vs. the initiator,

8 
 

reaction time, as well as the pH of the solution. Examples of using in situ polymerization to
fabricate nanocomposites are EG-PMMA[55], graphene-PMMA[56], graphene-PS[57], EGnylon 6[58,59], and EG-PANi [60].

3. Melt extrusion
Melt extrusion uses strong shearing forces to disperse the reinforcement phase at an elevated
temperature that liquefies the polymer. Compared with the previous two approaches, melt
extrusion has its unique advantage because of its environmental friendliness and versatility,
making it more commercially attractive. This strategy usually involves mixing graphene sheets
and melted polymer in a twin-screw extruder. The temperature, screw speed, screw rotation and
mixing time are the major parameters to adjust for optimal dispersion of the nanofiller in the
matrix. Several disadvantages associated with this approach is that it is confined by the viscosity
of the nanocomposites especially at a high filler loading which prevents uniform dispersion of
graphene nanosheets in the polymer, compromising the physical properties of the composites. A
variety of graphene reinforced composites prepared by this methods are EG-polyester[61], EGPPS[62], graphene-PP[63], graphene-PC[64].
The resulting physical properties of graphene reinforced nanocomposites vary widely
depending on the processing methods, concentration of fillers, aspect ratio, presence of
functional groups, inter-particle contact as well as the filler orientation in the polymer matrix.
The next section focuses on the electrical and thermal properties of conductive particle
(especially those filled by graphene) filled nanocomposites in general and discusses the
fundamental difference between the two.

9 
 

1. 3. 1 Electron transport:
The addition of electrically conductive fillers into an insulating polymer matrix results in
a significant drop in its resistivity at a certain filler loading. This mechanism of electronic
transport in composite materials has been discussed previously and is described by a
phenomenon called percolation. When the number of conductive filler is enough to form
continuous linkage that span the entire specimen, the resistivity experiences a significant drop,
turning an insulator into a semiconductor at the percolation threshold [65]. Figure 1-1 illustrates
the percolation phenomenon as applied to conductive composites. In spite of the high electrical
conductivity of the individual filler, it is difficult to obtain composites with resistivity within five
orders of magnitude of the filler itself [66,67]. The resistance of the composite specimen
consists of three independent components: 1) contact resistance (constriction resistance), 2)
quantum mechanical electron tunneling resistance, and 3) intrinsic resistivity of the fillers.
Contact resistance, also known as constriction resistance (Rcr), results from a resistance
associated with the constriction of electron flow through the contact of two conductive spheres.

Rcr derived from the ohmic conduction and electric fields can be expressed as:

Rcr 

i
d

1.1

where ρi is the intrinsic resistivity of the fillers, d is the diameter of the contact spot. Assume the
size of the conductive filler particle is D, when D/d is >10, equation (1.1) begins to diverge and a
very high resistance results even when the conductive particles are in close contact with each
other [68]. On the other hand, if the conductive particles are coated with a thin insulating film
during material processing, such as the residual organic film that may remain on the conductive
10 
 

powder surface after milling, or in a conductor-filled polymer composites, the polymer may
completely cover the surface of the filler. In both cases, tunneling resistance arises that increases
the resistivity of the composites. In particular, the magnitude of tunneling resistance (or the
probability of electrons to tunnel through a potential barrier) is directly related to the work
function of the conductor, the thickness of the insulating film, and the relative dielectric
permittivity of the film. Since tunneling is independent of the resistivity of the film itself, it is
believed that the organics, polymers, oxides with similar work functions, permittivity as well as
film thickness will have similar tunneling resistance. Dietrich has shown mathematically in his
pioneer work that described the tunneling resistivity on insulating film thicknesses and presented
a general tunneling curve based on TiO2 film on Ti, which is considered applicable to all other
materials [69]. Generally speaking, quantum mechanical tunneling can occur for thin films on
the order of 10 nm or less. If the film thickness is known, tunneling resistivity (ρt) can be
estimated, therefore the tunneling resistance

Rt which is inversely proportional to the contact

area A can be described by:

Rt 

t
A

1.2

The resistance of a contact Rc is the sum of the two separate effects given by:

Rc 

i
d



t
A

1.3

However, determination of either the contact diameter d or the contact area A is the major
challenge in estimating the overall contact resistance primarily because both parameters are
dependent on the morphology, dispersion state and more importantly the elastic modulus of the
11 
 

conductive particles due to possible mechanical deformation as a result of the change in the
hydrostatic pressure and temperature at which the sample is kept at [65]. For carbon based
nanocomposites, contact resistance dominates when filler loading is high and physical contacts
are present while tunneling resistance becomes significant when filler loading is low and there is
an insulating polymer film coated on the fillers. Due to the temperature dependent differences
involved in the origin of constriction and tunneling resistances (i.e. contact resistance increases
with temperature while tunneling resistance decreases with temperature [70]), Jovic et al
reported a "phase transition"

from predominately "contact conductivity" to "tunneling

conductivity" by measuring the conductivity temperature coefficient (CTC=dσ/dT) in an
epoxy/expanded graphite composites with intermediate filler contents when heated from room
temperature to 120°C [71]. While this phenomenon depends largely on the starting polymer
matrix, the filler loading as well as the processing history of composites, It did offer a nice
technique to separate the individual component's contribution to overall interfacial contact
resistance and determine which one is dominating from temperature dependence resistivity of the
composites.
In general, the electronic transport of a graphene based composites depends not only on
the intrinsic properties of graphene, but also on the interfacial contact resistance which is directly
related to the concentration, dispersion, and the interaction of the filler with the polymer host.

12 
 

 

Figure 1-1: Electrical percolation, as applied to conductive composites. The formation of the first
complete particle linkage results in a significant drop in resistivity at Vc. For interpretation of the
references to color in this and all other figures, the reader is referred to the electronic version of
this dissertation.

1. 3. 2 Phonon transport:
Heat transport involves movement of energy from one place to another by energy carriers.
For example, molecules in the liquid or gas phase bump into each other due to Brownian motion,
transferring their kinetic energy (thermal energy) at the same time. Energy can also be
transported by an ordered drift of molecules in certain direction such as advection. In polymers,
heat is predominantly transported by phonons, which are quantized modes of lattice vibration
[72]. The theoretical prediction of thermal conductivity (λ) of polymers can be calculated by the
kinetic formula:



C p vl
3
13 

 

1.4

where Cp is the specific heat capacity, v is the average phonon velocity, l is the phonon mean
free path. In particular, l is a very small number (usually a few angstroms) for polymers because
of the numerous structural defects along the chains of the polymers that act as phonon scattering
centers, leading to a very low thermal conductivity. Thermally conductive composites can be
used in a variety of applications such as automobiles, aerospace, electronic circuit boards, heat
sinks, electronic packaging, protective clothing, etc. There has been plenty of research directed at
enhancing thermal conductivity of polymers by adding thermally conductive fillers such as
carbon blacks, carbon fibers, graphite, metal or ceramic particles [73-77]. However, it typically
requires a very high filler loading (30 vol%) to achieve a thermal conductivity suitable for
certain applications, which poses great challenges in composite processing as a result of
increased viscosity of the nanocomposites. Moreover, the mechanical properties and density of
resulting composites are altered after the introduction of a high loading of fillers.
Graphene-based nanomaterials recently have shown promising results in terms of
enhancing the thermal conductivity of composites [78,79]. Taking epoxy as the polymer matrix,
thermal conductivity of graphene filled epoxy has been reported to achieve different levels of
enhancement depending on the composition of the filler, the aspect ratio, the filler size
distribution, as well as its chemical compatibility with the matrix [80-83]. Due to the 2D platelet
morphology of graphene sheets, thermal conductivity of composites is also dependent on the
orientation of the filler in the matrix. Heat flow in the direction perpendicular to the platelet is
significantly restricted as compared to the heat flow in the direction parallel to the platelets
[84,85]. This unique anisotropic heat conduction in graphene sheets filled composites can be
very useful in directional thermal management for electronics and protective clothings.

14 
 

Theoretical prediction and modelling:
Two basic models have been used to represent the upper and lower bound for thermal
conductivity of composites, referred to as the parallel and the series model. In the parallel model,
each phase is considered to contribute to conductivity independently, proportional to its volume
fraction:

 c   f  f   m m

(1.5)

where κc, κf, κm are the thermal conductivity of the composite, filler, matrix, respectively, and

φf and φm are the volume fractions of the filler and the matrix. This model assumes perfect
contact between the fillers in a percolated network and has some relevance to the case of
continuous fiber composites in the direction parallel to the fibers. However, it usually results in a
large overestimation with respect to experimental values. The series model, on the other hand,
assumes no contact between the filler and the matrix, thermal resistance of the composite is
considered the sum of the resistance from the filler and that of the matrix arranged in series.
Depending on the volume fractions, thermal conductivity is predicted by:

1

c 
((

f
f

)(

m
))
m

(1.6)

The experimental values obtained usually falls between these two predictions, with the true value
being closer to the series model prediction. More complicated models have been proposed and
developed that take into account the filler geometry, aspect ratio, alignment, as well as the
interfacial thermal resistance for better description of thermal behaviors of various systems.
15 
 

Among all these parameters that play a role in heat transport in composites, the interfacial
thermal resistance (or Kapitza resistance) seems to be drawing most of the attention. Kapitza
resistance, named after the Russian physicist who discovered a temperature discontinuity at
metal-liquid interface, can be calculated theoretically or by means of computer simulation as a
result of the technical challenges involved in accurately measuring the interfacial heat flow in
nanocomposites directly [86,87]. Monte-Carlo based methods and molecular dynamics offer the
advantage of manipulating the interaction of nanofiller in the matrix on a molecular scale,
simulation results can later be fitted with experimental data to back calculate the parameter of
interest [88,89].
As suggested by the previous discussion on the mechanism of energy transport in
composites, It is believed that no matter electronic or phononic conduction, both are affected by
the existence of interfaces in the composites (i.e. mostly the matrix-filler, filler-filler interfaces).
However, the conduction mechanisms are quite different when it comes to interfacial phases
when electron tunneling occurs while phonon tunneling is still not fully verified or understood .
Therefore, failing to consider the effect of interfacial transport of phonons might explain the
discrepancies observed in the estimation of thermal conductivity by the parallel and series model
described above.

1.4 Bi-continuous paper like graphene based nanocomposites
Very different from the approach taken in the previous section to fabricate bulk
nanocomposites, a highly ordered, highly flexible paper-like nanocomposite can be prepared
using a very inexpensive way such as vacuum assisted self assembly. In this case, graphene
sheets were dispersed in an aqueous or organic solvent and then filtrated on a filter membrane to
16 
 

create a paper. For example, carbon nanotubes mat (buckypaper)[90], graphene oxide (GO)
paper[91,92], and reduced graphene paper [93-95] have been reported by various research groups.
The electrical conductivity can reach as high as 200 S/cm for aligned buckypaper [90] and 350
S/cm for reduced graphene paper [95]. A highly aligned, bi-continuous paper-based
nanocomposite can be prepared by vacuum infiltration of a polymer resin or simply immerse the
porous paper structure in a solution containing the polymer of interest followed by solvent
evaporation, leaving the polymer intercalated in the pores of the paper. Due to the already
established interconnectivity of the filler and its preferential alignment, the thermal and electrical
conductivity in the aligned direction can be much higher than if the fillers are randomly
distributed in the polymer matrix due to the reduced contact resistance [84,85,96,97]. Depending
on the initial dispersion state of graphene based fillers, the paper-like nanocomposites still
possess certain flexibility and mechanical robustness. While most of graphene paper reported in
the literature consist of reduced graphene oxide that has to go through an extra chemical
reduction that cannot fully restore the conjugation of graphene, the electrical and thermal
properties of the resulting paper is compromised. For GNP developed in Drzal's group, the
highly aromatic graphene basal plane remains intact during the intercalation-expansion
fabrication route, offering another ideal starting material for the preparation of flexible, robust
paper based structure that can be used in a variety of applications.

1.5 Background in thermoelectric energy conversion materials:
The multi-functionality that can be introduced to a polymer matrix by incorporation of
graphene nanosheets has opened up numerous opportunities in the field of renewable energy
generation and conversion. One of the promising technologies that could potentially benefit from
17 
 

the development of graphene based polymer nanocomposites is thermoelectric energy conversion.
The thermoelectric effect was first discovered in 1821 by Thomas Seebeck who reported that
heating a junction between two dissimilar metals produced a potential difference. The Seebeck
coefficient (S), is defined as the ratio of voltage generated per unit temperature change:

S

V
T

(1.7)

Figure 1-2 is a schematic representation of the potential difference generated due to the diffusion
of charge carriers from the hot to the cold side of the sample. An equilibrium is reached when the
carriers drifting back due to a built in electric field balances the carrier diffusion induced by the
temperature gradient. Seebeck coefficient is determined by the dominant charge carriers present
in the materials and is positive for holes and negative for electrons.

 

Figure 1-2: Schematic representation of Seebeck effect when a sample is subjected to a
temperature gradient, in this case, holes are the dominant carriers.
 

Another thermoelectric effect is discovered by a French watchmaker Peltier who found that
when a current is passed though a junction of different metals, heat can be either absorbed or
released at the junction depending on the direction of the current. The Seebeck effect and the
18 
 

Peltier effect are the theoretical foundations for the state of the art solid state heating and cooling
technology.
In order to illustrate how a thermoelectric device work, it is essential to understand the
basic configuration of a thermoelectric unicouple composed of a n-type and p-type material
connected in series. If a heat source is applied to one end of the couple, charge carriers become
energized and diffuse down toward the cold side causing a Seebeck voltage to be generated
across the couple, this thermally induced voltage gives rise to a current when a resistive load is
connected to the couple. On the other hand, a current passing through the unicouple causes the
electrons and holes to move away from one end of the couple, effectively cooling it down. A
thermoelectric module can be fabricated by connecting many unicouples electrically in series but
thermally in parallel so as to maintain the effective temperature difference. Figure 1-3 is a
schematic depiction of a thermoelectric unicouple that operates under either a power generation
mode (Seebeck effect) or refrigeration mode (Peltier effect).

Figure 1-3: Schematic representation of a thermoelectric unicouple consisting of a p and n type
semiconductor materials connected electrically in series and thermally in parallel operating under
power generation mode (left) or refrigeration mode (right)
The conversion efficiency η of the thermoelectric device defined as the ratio between power W
delivered to the load to the heat supplied Q from the source can be expressed as:
19 
 





W T


Q Th

1 Z T 1

T
1 Z T  c
Th

(1.8)



where Z is a material dependent parameter,

T , Th and Tc are the average operating temperature,

hot side temperature and cold side temperature respectively. Parameter Z, also referred to as the
thermoelectric figure of merit is the quantity to be maximized for higher conversion efficiency.
The most commonly used dimensionless figure of merit ZT is obtained by multiplying Z by the
operating temperature T which is expressed as:

ZT 

S 2



T
(1.9)

where S is the Seebeck coefficient, σ is the electrical conductivity, κ is the thermal conductivity
of the material. From the definition of Z, a good thermoelectric material should have a high
Seebeck coefficient, a high electrical conductivity and a low thermal conductivity. Apparently, in
addition to having a high Seebeck coefficient, a large electrical to thermal conductivity ratio σ/κ
is highly desired. Current research activities are primarily focused on inorganic semiconductors
and developing complex material structures that behave like phonon glass electron crystals (i.e.
allowing electrons to flow undisrupted while blocking phonon transport in the materials)[98-100].
Certain approaches such as impurity atom substitution [101], introduction of rattler
atoms[102,103] or use of low dimensional materials[104,105] have been shown to be successful
in reducing the thermal conductivity while causing less disruption to electrical conductivity.
However, the delicate nanostructures that are intentionally introduced to the bulk could be very
difficult to maintain due to their susceptibility to stoichiometry, temperature, pressure, sintering
20 
 

time and etc, making it almost impractical in large scale production. In addition, large
consumption of raw materials could also impose added risk and cost when faced with a
dwindling mining and production of rare earth metals.
Recently, research in polymer based thermoelectric materials have been on the rise
because of its intrinsic low thermal conductivity, light-weight, low cost as well as ease of large
scale production. During the course of this research, polymer composites based on carbon
nanotubes have also been proposed to show promising thermoelectric properties [106-109]. The
combination of a polymer matrix having low thermal conductivity and highly electrically
conductive fillers such as graphene nanoplatelets seems to be an interesting and viable route to
explore in terms of increasing the σ/κ ratio so as to achieve a higher figure of merit.

1.6 Motivation and objective of research:
Graphene nanoplatelets are produced by acid intercalation and thermal exfoliation of
graphite intercalated compounds. These nanoplatelets have excellent electrical, thermal and
mechanical properties as a result of preservation of the pristine aromatic character of the
graphene basal plane during production. However, the benefits of GNP can only be realized if
they can be assembled macroscopically or into a bulk. In this case, it is reasonable to assume that
the numerous contacts or the interfaces created play an important role in the performance and
behavior of the resulting nanocomposites. Based on this premise, this dissertation is dedicated to
the investigation of the difference in energy transport (electron and phonon in particular)
occurring at the interfaces and discusses several useful nanostructuring techniques to modify the
behavior of carrier transport in GNP-based nancomposites with a focus on developing an
inexpensive, lightweight, flexible nancomposites for thermoelectric application.
21 
 

In the first part of the research, the effect of incorporating GNP on the electrical and
thermal properties of an insulating polymer is studied and the distinction between conduction of
electrons and phonons across GNP/polymer interfaces is illustrated. A novel method based on
micro-Raman spectroscopy that investigates the effect of particle-particle junctions on phonon
transport is proposed.
With the previous learning in mind, as well as an aim to maximize the potential of
individual GNP particles, the second part of the research is based on a highly ordered, highly
flexible multilayer GNP paper prepared by vacuum assisted self assembly and focuses on the
modification and separation of electron and phonon transport by various nanostructuring
techniques in order to design a nanocomposite thermoelectric material with a high σ/κ ratio and
Seebeck coefficient. In particular, a facile, one-pot synthesis of metal nanoparticles on the
surface of GNP is proposed with microwave assisted heating and the effect of metal
nanoparticles on carrier transport is explored. In another approach, a conductive polymer
(polyaniline) was synthesized on the surface of GNP by in situ polymerization forming a coreshell nanostructure. The effect of polymer interaction with GNP, the concentration as well as
protonation ratio of polyaniline on the thermoelectric properties of PANi/GNP nanocomposites
is discussed. The last part explores a novel approach that completely alters the way electrons and
phonons transport inside a PANi decorated GNP paper by a simple folding, the potential of
creating an electrically connected thermally disconnected structure from this approach for
thermoelectric application is discussed.

22 
 

 

REFERENCES
 

 

23 
 

References
 

1. Geim, A.K.; Graphene: status and prospects, Science, 2009, 324, 1530.
2. Geim, A.K. and Novoselov. K.S. The rise of graphene, Nature Materials, 2007, 6, 183-191.
3. Rao, C. N. R.; Sood, A. K.; Subrahmanyam, K. S.; Govindaraj, A. Graphene: the new twodimensional nanomaterial. Angew. Chem. Int. Ed. 2009, 48, 7752-7777.
4. Allen, M. J.; Tung, V. C.; Kaner, R. B. Honeycomb carbon: a review of graphene. Chem. Rev.
2010, 110, 132-145.
5. Singh, V.; Joung, D.; Zhai, L.; Das, S.; Khondaker, S. I.; Seal, S. Graphene based materials:
Past, present and future. Prog. Mater. Sci, 2011, 56, 1178-1271.
6. Lee, C.; Wei, X.; Kysar, J. W.; Hone, J. Measurement of the elastic properties and intrinsic
strength of monolayer graphene. Science, 2008, 321, 385-388.
7. Novoselov, K. S. Geim, A.K.; Morozov, S.V.; Jiang, D.; Katsnelson, M.I.; Grigorieva, I. V.;
Dubonos, S. V.; Firsov, A. A. Two-dimensional gas of massless Dirac fermions in graphene,
Nature, 2005, 438, 197-200.
8. Novoselov, K.S.; Geim, A.K.; Morozov, S.V.; Jiang, D.; Zhang. Y.; Dubonos, S. V.;
Grigorieva, I. V. and Firsov. A. A. Electric field effect in atomically thin carbon films, Science
2004, 306, 666-669.
9. Balandin, A.A.; Ghosh, S.; Bao, W.; Calizo, I.; Teweldebrhan, D.; Miao, F et al. Superior
thermal conductivity of single-layer graphene. Nano Lett. 2008, 8, 902-907.
10. Kuilla, T.; Bhadra, S.; Yao, D.; Kim, N. H.; Bose, S.; Lee, J. H. Recent advances in graphene
based polymer composites. Prog. Polym. Sci. 2010, 35, 1350-1375.
11. Roy-Mayhew, J.D.; Bozym, D. J.; Punckt, C.; Aksay, I. A. Functionalized graphene as a
catalytic counter electrode in dye-sensitized solar cells. ACS Nano, 2010, 4, 6203-6211.
12. Brownson, D. A. C.; Kampouris, D. K.; Banks, C. E. An overview of graphene in energy
production and storage applications. J Power Sources. 2011, 196, 4873-4885.
13. Wang, S.; Ang, P. K.; Wang, Z.; Tang, A. L. L.; Thong, J. T. L.; Loh, K. P. High mobility,
printable, and solution-processed graphene electronics. Nano Lett. 2010, 10, 92-98.
14. Eizenberg, M.; & Blakely, J. M.; Carbon monolayer phase condensation on Ni (111). Surf.
Sci. 1970, 82, 228-236.
15. Li, X.; Cai, W.; An, J.; Kim, S.; Nah, J.; Yang, D et al. Large-Area synthesis of high-quality
and uniform graphene films on copper foils. Science, 2009, 324, 1312-1314.

24 
 

16. Berger. C.; Song, Z.; Li, X.; Wu, X.; Brown, N.; Naud, C et al. Electronic confinement and
coherence in patterned epitaxial graphene, Science, 2006, 312, 1191-1196.
17. Reina, A.; Jia, X.; Ho, J.; Nezich, D.; Son, H.; Bulovic, V et al. Large area, few-layer
graphene films on arbitrary substrates by chemical vapor deposition. Nano Lett. 2009, 9, 30-35.
18. Chen, J. H.; Ishigami, M.; Jang, C.; Hines, D. R.; Fuhrer, M. S.; Williams, E. D. Printed
graphene circuits. Adv. Mater. 2007, 19, 3623-3627.
19. Hummers W, Offeman R. Preparation of graphite oxide. JACS, 1958, 80,1339.
20. Dreyer, D. R.; Park, S.; Bielawski, C. W.; Ruoff, R. S. The chemistry of graphene oxide.
Chem. Soc. Rev. 2010, 39, 228-240.
21. Szabo, T.; Berkesi O.; Dekany, I.; DRIFT study of deuterium-exchanged graphite oxide.
Carbon, 2005, 43, 3186-3189.
22. de Boer JH,; Van Doorn ABC.; Graphite oxide I: The adsorption of water. Koninkl Ned
Akad Wetenschap. Proc 1958; 61B, 242-252.
23. Park, S.; Ruoff, R. S. Chemical methods for the production of graphenes. Nat. Nanotechnol.
2009, 4, 217-224.
24. Stankovich, S.; Dikin, D. A.; Piner, R. D.; Kohlhaas, K. A.; Kleinhammes, A.; Jia, Y et al.
Synthesis of graphene-based nanosheets via chemical reduction of exfoliated graphite oxide,
Carbon, 2007, 45, 1558-1565.
25. Stankovich. S.; Piner. R.D.; Chen, X.; Wu, N.; Nguyen, S. T.; Ruoff. R.S.; Stable aqueous
dispersions of graphitic nanoplatelets via the reduction of exfoliated graphite oxide in the
presence of poly(sodium 4-styrenesulfonate), J. Mater. Chem. 2006, 16, 155-158.
26. Xu, Y.; Bai, H.; Lu, G.; Li, C.; Shi, G. Flexible graphene films via the filtration of watersoluble noncovalent functionalized graphene sheets. JACS, 2008, 130, 5856-5857.
27. Li, D.; Muller, M. B.; Gilje, S.; Kaner, R. B.; Wallace, G.G.; Processable aqueous
dispersions of graphene nanosheets. Nat. Nanotechnol, 2008, 3, 101-105.
28. Schniepp, H. C.; Li, J. L.; McAllister, M. J.; Sai, H.; Alonso, M. H.; Adamson, D. H et al.
Functionalized single graphene sheets derived from splitting graphite oxide, J. Phys. Chem. B.
2006, 110, 8535-8539.
29. McAllister, M.J.; Li, J. L.; Adamson, D. H.; Schniepp, H. C.; Abdala, A. A.; Liu, J et al.
Single sheet functionalized graphene by oxidation and thermal expansion of graphite. Chem.
Mater. 2007, 19, 4396-4404.
30. Larciprete, R.; Lacovig, P.; Gardonio, S.; Baraldi, A.; Lizzit, S. Atomic oxygen on graphite:
chemical characterization and thermal reduction. J. Phys. Chem. C. 2012, 116, 9900-9908.

25 
 

31. Li, J-L.; Kudin, K. N.; McAllister, M. J.; Prud'homme, R.K.; Aksay, I.A.; Car, R.; Oxygendriven unzipping of graphitic materials, Phys. Rev. Lett. 2006, 96, 176101.
32. Jung, I.; Dikin, D.; Park, S.; Cai, W.; Mielke, S. L.; Ruoff, Rodney.; Effect of water vapor on
electrical properties of individual reduced graphene oxide sheets. J. Phys. Chem. C, 2008, 112,
20264-20268.
33. Navarro, C. G.; Weitz, T. R.; Bittner, A. M.; Scolari, M.; Mews, A.; Burghard, M et al.
Electronic transport properties of individual chemically reduced graphene oxide sheets. Nano
Lett. 2007, 7, 3499-3503.
34. Drzal, L.T.; Fukushima, H.; Expanded graphite and products produced therefrom, U.S Patent
Application 11/363,336, 2006.
35. Fukushima, H.; Graphite reinforcements in polymer nanocomposites. Ph.D dissertation,
Michigan State University, East Lansing, Michigan, 2003.
36. Castro, E.V.; Novoselov, K. S.; Morozov, S. V.; Peres, N. M. R.; Lopes dos Santos, J. M. B.;
Nilsson, J et al. Biased bilayer graphene: semiconductor with a gap tunable by the electric field
effect. Phys. Rev. Lett. 2007, 99, 216802.
37. Zhang, Y.; Tang, T. T.; Girit, C.; Hao , Z.; Martin, M. C.; Zellt, A et al. Direct observation of
a widely tunable bandgap in bilayer graphene. Nature 2009, 459, 820-823.
38. Vaia, R. A.; Giannelis E. P. Polymer melt intercalation in organically-modified layered
silicates: model predictions and experiment. Macromolecules, 1997, 30, 8000-8009.
39. Okamoto, M.; Nam, P. H.; Maiti, P.; Kotaka, T.; Nakayama, T.; Takada, M et al. Biaxial
flow-induced alignment of silicate layers in polypropylene/clay nanocomposite foam. Nano Lett,
2001, 1, 503-505.
40. Chung, D. D. L. Electrical applications of carbon materials. J. Mater. Sci. 2004, 39, 26452661.
41. Cho, E. A.; Jeon, U. S.; Ha, H. Y.; Hong, S. A.; Oh, I. H. Characteristics of composite
bipolar plates for polymer electrolyte membrane fuel cells. J Power Sources. 2004, 125, 178-182.
42. Thostenson, E. T.; Ren, Z.; Chou, T. W. Advances in the science and technology of carbon
nanotubes and their composites: a review. Compos Sci Technol, 2001, 61, 1899-1912.
43. Coleman, J. N.; Khan, U.; Blau, W. J.; Gun'ko, Y. K. Small but strong: A review of the
mechanical properties of carbon nanotube-polymer composites. Carbon, 2006, 44, 1624-1652.
44. Wu, Q.; Zhu, W.; Zhang, C.; Liang, Z.; Wang, B. Study of fire retardant behavior of carbon
nanotube membranes and carbon nanofiber paper in carbon fiber reinforced epoxy composites.
Carbon, 2010, 48, 1799-1806.
45. Aylsworth, J. W. Expanded graphite and composition thereof. US Patent 1,137,373, 1915.
26 
 

46. Aylsworth, J. W. Expanded graphite. US Patent 1,191,383; 1916.
47. Lincoln, V. F.; Claude, Z. Inventors, organic matrix composites reinforced with intercalated
graphite. US Patent 4,414,142; 1983.
48. Zheng, W.; Wong, S. C. Electrical conductivity and dielectric properties of PMMA/expanded
graphite composites. Compos Sci Technol. 2003, 63, 225-235.
49. Liu, N.; Luo, F.; Wu, H.; Liu, Y.; Zhang, C.; Chen, J. One step ionic-liquid-assisted
electrochemical synthesis of ionic-liquid functionalized graphene sheets directly from graphite.
Adv Funct. Mater. 2008, 18, 1518-1525.
50. Jiang, L.; Shen, X. P.; Wu, J. L.; Shen, K. C Preparation and characterization of
graphene/poly(vinyl alcohol) nanocomposites. J Appl Polym Sci, 2010, 118, 275-279.
51. Vadukumpully, S.; Paul, J.; Mahanta, N.; Valiyaveettil, S. Flexible conductive
graphene/poly(vinyl chloride) composite thin films with high mechanical strength and thermal
stability. Carbon, 2011, 49, 198-205.
52. Ansari, S.; Giannelis, E. P. Functionalized graphene sheet-poly(vinylidene fluoride)
conductive nanocomposites. J Polym Sci Part B-Polym Phys, 2009, 47, 888-897.
53. Jang, J. Y.; Kim, M. S.; Jeong, H. M.; Shin, C. M. Graphite oxide/poly(methyl methacrylate)
nanocomposites prepared by a novel method utilizing macroazoinitiator. Compos Sci Technol,
2009, 69, 186-191.
54. Huang, Y.; Qin, Y.; Zhou, Y.; Niu, H.; Yu, Z. Z.; Dong, J. Y. Polypropylene/graphene oxide
nanocomposites prepared by in-situ Ziegler-Natta polymerization. Chem Mater, 2010, 22,40964102.
55. Chen, G.; Weng, W.; Wu, D.; Wu, C. PMMA/graphite nanosheets composite and its
conducting properties. Eur. Polym. J. 2003, 39, 2329-2335.
56. Chen, G. H.; Wu, D. J.; Weng, W. G.; Yan, W. L. Dispersion of graphite nanosheets in a
polymer matrix and the conducting property of the nanocomposites. Polym. Engng. Sci. 2001, 41,
2148.
57. Hu, H.; Wang, X.; Wang, J.; Wan, L.; Liu, F.; Zheng, H et al. Preparation and properties of
graphene nanosheets-polystyrene nanocomposites via in situ emulsion polymerization. Chem.
Phys. Lett. 2010, 484, 247-253.
58. Pan, Y. X.; Yu, Z. Z.; Ou, Y.C.; Hu, G. H. A new process of fabricating electrically
conducting nylon 6/graphite nanocomposites via intercalation polymerization. J. Polym. Sci. Part
B, Polym. Phys. 2000, 38, 1626-1633.
59. Weng, W.; Chen, G.; Wu, D.; Chen, X.; Lu, J.; Wang, P.; Fabrication and characterization of
nylon 6/foliated graphite electrically conducting nanocomposites. J Polym Sci Part B, Polym
Phys. 2004, 42, 2844-2856.
27 
 

60. Du, X. S.; Xiao, M.; Meng, Y. Z.; Facile synthesis of highly conductive polyaniline/graphite
nanocomposites. Eur Polym J. 2004, 40, 1489-1493.
61. Kim, H.; Macosko, C. W. Morphology and properties of polyester/exfoliated graphite
nanocomposites. Macromolecules, 2008, 41, 3317-3327.
62. Zhao, Y. F.; Xiao, M.; Wang, S. J.; Ge, X. C.; Meng, Y. Z. Preparation and properties of
electrically conductive PPS/expanded graphite nanocomposites. Compos Sci Technol 2007, 67,
2528-2534.
63. Kalaitzidou, K.; Fukushima, H.; Drzal, L. T. A new compounding method for exfoliated
graphite-polypropylene nanocomposites with enhanced flexural properties and low percolation
threshold. Compos Sci Technol, 2007, 67, 2045-2051.
64. Kim, H.; Macosko C. W. Processing-property relationships of polycarbonate/graphene
nanocomposites. Polymer 2009, 50, 3797-3809.
65. Ruschau, G. R.; Yoshikawa, S.; Newnham, R. E. Resistivities of conductive composites. J.
Appl. Phys. 1992, 72, 953-959.
66. Hu, K. A.; Moffatt, D.; Runt, J.; Safari, A.; Newnham, R. E. V2O3-polymer composite
thermistors. J. Am. Ceram. Soc. 1987, 70, 583-585.
67. Bhattacharya, S. K. Metal-filled polymers (Marcel-Dekker, New York, 1986)
68. R. Holm, Electric contacts. (Springer, Berlin, 1967)
69. Dietrich, L. Z. Phys. 1952, 132, 231.
70. Sheng, P.; Sichel, E. K.; Gittleman, J. I. Fluctuation-induced tunneling conduction in carbonpolyvinylchloride composites. Phys. Rev. Lett. 1978, 40, 1197-1200.
71. Jovic. N.; Dudic, D.; Montone, A.; Antisari, M. V.; Mitric, M.; Djokovic, V. Temperature
dependence of the electrical conductivity of epoxy/expanded graphite nanosheet composites.
Scripta Mater, 2008, 58, 846-849.
72. Majumdar A.; Microscale transport phenomena, In: Rohsenow WM, Hartnett, J. R. Cho, Y. I,
editors. Handbook of heat transfer, 3rd ed. New York; McGraw-Hill; 1998, p.8.1-8.
73. Pierson, H.O.; Handbook of carbon, graphite, diamond and fullerenes: properties. Processing
and applications. New Jersey: Noyes Publications; 1993.
74. Wypych, G. Handbook of fillers: physical properties of fillers and filled materials. Toronto:
ChemTec Publishing; 2000.
75. Wolff, S.; Wang, M. J.; Carbon black science & technology. 2nd ed, New York; Marcel
Dekker; 1993.
76. Kelly, B. T.; Physics of graphite. Barking (UK): Applied science publishers; 1981.
28 
 

77. Kuriger, R. J.; Alam, M. K.; Thermal conductivity of thermoplastic composites with
submicrometer carbon fibers. Exp Heat Trans, 2002, 15, 19-30.
78. Tu, H.; Ye, L. Thermal conductive PS/graphite composites. Polym. Adv. Technol. 2009, 20,
21-27.
79. Fukushima, H.; Drzal, L. T.; Rook, B. P.; Rich, M. J. Thermal conductivity of exfoliated
graphite nanocomposites. J. Therm. Anal. Calorim. 2006, 85, 235-238.
80. Hung, M. T.; Choi, O.; Ju, Y. S.; Hahn, H. T. Heat conduction in graphite-nanoplatelet
reinforced polymer nanocomposites. Appl. Phys. Lett. 2006, 89, 023117.
81. Yu, A.; Ramesh, P.; Itkis, M. E.; Bekyarova, E.; Haddon, R. C. Graphite nanoplatelet-epoxy
composite thermal interface materials. J. Phys. Chem. C. 2007, 111, 7565-7569.
82. Sun, X.; Ramesh, P.; Itkis, M. E.; Bekyarova, E.; Haddon, R. C. Dependence of the thermal
conductivity of two-dimensional graphite nanoplatelet-based composites on the nanoparticle size
distribution. J. Phys.: Condens. Matter, 2010, 22, 334216 (9pp).
83. Ganguli, S.; Roy, A. K.; Anderson, D. P. Improved thermal conductivity for chemically
functionalized exfoliated graphite/epoxy composites. Carbon, 2008, 46, 806-817.
84. Veca, L. M.; Meziani, M. J.; Wang, W.; Wang, X.; Lu, F.; Zhang, P et al. Carbon nanosheets
for polymeric nanocomposites with high thermal conductivity. Adv. Mater. 2009, 21, 2088-2092.
85. Marconnet, A. M.; Yamamoto, N.; Panzer, M. A.; Wardle, B. L.; Goodson, K. E. Thermal
conduction in aligned carbon-nanotube-polymer nanocomposites with high packing density.
ACS Nano, 2011, 5, 4818-4825.
86. Huxtable, S. T.; Cahill, D. G.; Shenogin, S.; Xue, L.; Ozisik R.; Barone, P et al. Interfacial
heat flow in carbon nanotube suspensions. Nat Mater. 2003, 2, 731-734.
87. Shenogin S.; Xue, L.; Ozisik R.; Keblinski, P.; Cahill, D. G.; Role of thermal boundary
resistance on the heat flow in carbon-nanotube composites. J Appl Phys, 2004, 95, 8136-8144.
88. Duong, H. M.; Yamamoto, N.; Papavassilliou, D. V.; Maruyama, S.; Wardle, B. L. Intercarbon nanotube contact in thermal transport of controlled morphology polymer nanocomposites.
contactNanotechnology, 2009, 20, 155702.
89. Duong, H. M.; Papavassiliou, D. V.; Mullen, K. J.; Maruyama, S. Computational modelling
of the thermal conductivity of single-walled carbon nanotube-polymer composites.
Nanotechnology, 2008, 19, 065702.
90. Wang, D.; Song, P.; Liu, C.; Wu, W.; Fan, S. Highly oriented carbon nanotube papers made
of aligned carbon nanotubes. Nanotechnology, 2008, 19, 075609 (1-6)
91. Dmitriy, A. D.; Stankovich, S.; Zimney, E. J.; Piner, R. D.; Dommett, G. H. B.; Evmenenko,
G et al. Preparation and characterization of graphene oxide paper. Nature, 2007, 448, 457-460.
29 
 

92. Park, S.; Lee, K. S.; Bozoklu, G.; Cai, W.; Nguyen, S. T.; Ruoff, R. S. Graphene oxide
papers modified by divalent ions-enhancing mechanical properties via chemical cross-linking.
ACS Nano, 2008, 2, 572-578.
93. Li, D.; Kaner, R. B. Graphene-based materials. Science, 2008, 320, 1170-1171.
94. Xu, Y.; Bai, H.; Lu, G.; Li, C.; Shi, G. Flexible graphene films via the filtration of watersoluble noncovalent functionalized graphene sheets. JACS, 2008, 130, 5856-5857.
95. Chen, H.; Muller, M. B.; Gilmore, K. J.; Wallace, G. G.; Li, D. Mechanically strong,
electrically conductive, and biocompatible graphene paper. Adv. Mater. 2008, 20, 3557-3561.
96. Du, F.; Guthy, C.; Kashiwagi, T.; Fischer, J. E.; Winey, K. I. An infiltration method for
preparing single-walled nanotube/epoxy composites with improved thermal conductivity. J.
Polym Sci: Part B: Polym Phys, 2006, 44, 1513-1519.
97. Liang, Q.; Yao, X.; Wang, W.; Liu, Y.; Wong, C. P. A three-dimensional vertically aligned
functionalized multilayer graphene architecture: an approach for graphene-based thermal
interface materials. ACS Nano, 2011, 5, 2392-2401.
98. Sales. B. C.; Mandrus, D.; Chakoumakos, B. C.; Keppens, V.; Thompson, J. R. Filled
skutterudite antimonides: electron crystals and phonon glasses. Phys. Rev. B, 1997, 56, 1508115089.
99. Nolas, G.S.; Morelli, D.T.; Tritt, T.M.; Skutterudites: A phonon-glass-electron crystal
approach to advanced thermoelectric energy conversion applications. Annu. Rev. Mater. Sci.
1999, 29, 89-116.
100. Nolas, G.S.; Slack, G.A.; Schujman, S. B. Semiconductor clathrates: a phonon glass
electron crystal material with potential for thermoelectric applications. Semiconduct Semimet.
2001, 69, 255-300.
101. Ioffe, A. F. Heat transfer in semiconductors, Can. J. Phys, 1956, 34, 1342-1355.
102. Slack, G. A.; Tsoukala, V. G. Some properties of semiconducting IrSb3. J. Appl. Phys. 1994,
76, 1665-1671.
103. Morelli, D. T.; Meisner, G. P. Low temperature properties of the filled skutterudite
CeFe4Sb12. J. Appl. Phys, 1995, 77, 3777-3781.
104. Hicks, L. D.; Dresselhaus, M. S. Effect of quantum well structures on the thermoelectric
figure of merit. Phys. Rev. B, 1993, 47, 12727-12731.
105. Harmon, T. C.; Walsh, M. P.; Laforge, B. E.; Turner, G. W. Nanostructured thermoelectric
materials. J Electron. Mater. 2005, 34, L19-L22.
106. Meng C.; Liu, C.; Fan, S. A promising approach to enhanced thermoelectric properties
using carbon nanotube networks. Adv. Mater. 2009, 21, 1-5.
30 
 

107. Yao, Q.; Chen, L.; Zhang, W.; Liufu, S.; Chen, X. Enhanced thermoelectric performance of
single-walled carbon nanotubes/polyaniline hybrid nanocomposites. ACS Nano, 2010, 4, 24452451.
108. Kim, D.; Kim.Y.; Choi, K.; Grunlan, J.C.; Yu, C. Improved thermoelectric behavior of
nanotube-filled polymer composites with poly(3,4-ethylenedioxythiophene) polystyrenesulfonate.
ACS Nano, 2010, 4, 513-523.
109. Yu, C.; Choi, K.; Liang, Y.; Grunlan, J.C. Light-weight flexible carbon nanotube based
organic composites with large thermoelectric power factors. ACS Nano, 2011, 5, 7885-7892.
 

 

31 
 

Chapter 2 	Investigation of GNP in Improving the Thermal and
Electrical Conductivity of Paraffin Wax and Its Use in Thermal
Energy Storage

2.1 Abstract:
This chapter serves as a primer to explore the physical properties of GNP and its effect on a
insulating polymer matrix (i.e. paraffin wax). In particular, nanocomposites consisting of GNP
and paraffin wax with high latent heat have great potential for high efficiency phase change
materials (PCM) for thermal energy storage application. Composite PCMs were made by mixing
two different kinds of GNP (GNP-1 and GNP-15) into paraffin wax. Direct casting and two roll
milling were used to prepare samples. The investigation on the thermal and electrical
conductivity of nanocomposites with these two nanofillers was performed. Higher thermal
conductivity of composite PCM can be achieved with nanofillers of larger aspect ratio, higher
particle alignment. The thermal physical properties of the nanocomposites were investigated by
differential scanning calorimetry and thermal gravimetric analysis. It was found that the latent
heat of the nanocomposites was not adversely affected by the presence of GNP and the thermal
stability improved. This chapter is based on a journal publication: Xiang, J., Drzal, L. T. (2011).
Investigation of exfoliated graphite nanoplatelets (xGnP) in improving thermal conductivity of
paraffin wax-based phase change material. Solar Energy Materials and Solar Cells, 95(7), 1811–
1818.

2.2 Introduction
Latent heat storage materials have been reported in the literature as useful for thermal
32 
 

energy storage applications because of their high energy density [1-3]. Heat is stored through the
phase transition within the material, such as solid-solid or solid-liquid transitions. Paraffin wax
has attracted numerous attentions for its low cost, moderate energy densities, low vapor pressure,
negligible supercooling and chemical inertness [4]. However, one of the intrinsic disadvantages
associated with paraffin wax and other organic PCM are their low thermal conductivity, which
severely limited the rate of absorbing and releasing heat from and to the environment. Various
high thermally conductive fillers (e.g. metallic fins [5], ceramic powder fillers, graphitic carbon
fibers [6], carbon nanofibers [7], graphite particles [8] and exfoliated graphite [9]) have been
reported to improve the effective thermal conductivity of phase change materials. However, one
usually needs a very high concentration of these traditional conductive fillers to achieve a
noticeable improvement in thermal conductivity, resulting in increased density of the composites,
processing difficulties and increased cost.
In this work, GNP-1 and GNP-15 were mixed with paraffin wax. The viscosity of
paraffin wax is low and is thus conducive to dispersing conductive particles uniformly in the
polymer matrix. The thermal and electrical conductivities of nanocomposites with nanofillers of
different sizes were investigated. Different samples were prepared to study the effect of particle
orientation on the conductivities of nanocomposites. Furthermore, the experimental results were
compared with effective medium theory first proposed by Nan [10] on carbon nanotubes to
estimate the interfacial thermal resistance. In addition, the effect of adding different GNP
particles on melting/crystallization enthalpies, latent heat and thermal stability of paraffin wax is
also discussed.

33 
 

2.3 Experimental:
Materials:
GNP-15 and GNP-1 were produced by the acid intercalation and microwave-assisted
exfoliation process described in the previous chapter. Paraffin wax (n-docosane) with melting
temperature of 53-57oC, xylenes (reagent grade) were purchased from Sigma-Aldrich and used
directly.
Sample preparation:
Casting:
Paraffin wax was melted at 80°C. GNP were then added to the liquid paraffin under
constant stirring at 1wt%, 2wt%, 4wt%, 6wt%, 8wt%, and 10wt%. At higher loadings
(i.e. >6wt%), xylenes were used as a solvent to dissolve wax first, reducing the melt viscosity
and facilitate the uniform dispersion of GNP. The liquid composite was sonicated for 1 h prior to
solvent evaporation. Samples were poured into a customized aluminum mold (shown in Figure
2-1a) and solidified at room temperature. Disk samples of 1’’ in diameter and 5 mm in average
thickness were used for subsequent experiments.
Two roll milling:
In order to study the effect of particle orientation on thermal conductivity, composite
samples from cast molding was further processed by two roll mill (shown in Figure 2-1b) in
which extensive shear force and compression at the nip point of the rolls allow particles to align
within the plane. In two roll mill processing, the quality of the composite film depends on the
temperature of the rolls, a higher temperature yet still below the melting point of paraffin reduces
the chance of film shrinkage by giving the wax molecules enough thermal energy to undergo
stress relaxation while the film is still wrapped around the rolls. The speed of the roll is adjusted
34 
 

so that the front roll rotates faster than the rear roll. It is worth noting that the speed difference
between the rolls effectively creates enough shear force at the gap of the two rolls, facilitating
the dispersion of GNP and their alignment in the matrix. The composite materials were
processed in the rolls for 10 minutes before they were removed. The final product composite
sheet is around 0.4-0.5 mm in thickness. Both the through plane and in plane thermal
conductivity was measured. Figure 2-1c is the schematic representation of sample preparation
routes:

a

b

Figure 2-1: (a) Photograph of the customized aluminum mold; (b) Photograph of Bolling two roll
mill; (c) Scheme for the sample preparation routes: casting and two roll milling.
35 
 

Figure 2-1 (cont'd)

c

2.4 Characterization:
Scanning electron microscopy (SEM)
To investigate the dispersion of GNP particles in paraffin matrix at different loadings,
disk samples prepared by casting and two roll milling were liquid nitrogen fractured. The
fractured surface was oxygen plasma treated for 5 minutes with a Radio Frequency power at
36 
 

550W. The treated sample was later gold coated with the Denton Vacuum sputter coater to
ensure that the sample is electrically conductive. A JEOL (JSM 6400) scanning electron
microscope with an accelerating voltage of 15kV was used to examine the fractured composite
surface at a working distance of 15 mm.
Thermal conductivity
The thermal conductivity of paraffin/GNP composites was measured with a Unitherm
(TM) Model 2022 (Anter Corporation, Pittsburgh, PA). The test was performed according to
ASTM E1530. For samples prepared by cast molding, the average thickness of the sample is 5
mm, and it can be directly used for through-plane thermal conductivity measurement. However,
for the samples prepared by two roll mill, the through plane thermal resistance is too small to be
measured due to the very small thickness of the composite thin film (0.4-0.5 mm). 4 or 5 sheets
were stacked in parallel so that the resistance falls in the range of the instrument. As for the inplane measurement, the sample was made by compressing the thin sheets released from the roll
mills in a closed rectangular mold to a desired thickness, and used a circular iron punch to
produce a 1’’ disk sample from the cross section of the block (shown in Figure 2-2a,b). All the
samples were tested at 20 oC under an applied load of 20 psi.
Electrical conductivity
The resistivity of paraffin/GNP nanocomposites was measured with a Gamry instrument
under FAS2TM Femtostat plug system and potentiostatic mode. A rectangular portion was
fabricated by a Buehler IsometTM low speed saw from the cast molded disk samples for the
measurements. The resistance of the sample was measured in the frequency range from 0.6Hz to
100kHz and resistivity of samples were calculated by taking the geometry of the sample into
account using equation (2.1):
37 
 

  R A/ L

(2.1)

R is the resistance of the sample at 1Hz, A is the cross section area of the sample and L is the
length of the sample.
Differential scanning calorimetry (DSC)
A sample size of around 5-10 mg was loaded to the DSC sample cell and the data was
nd

collected for the 2

run at a scanning rate of 5°C/min. The solid-solid transformation in paraffin

wax, the melting temperature, the crystallization temperature, and the latent heat of paraffin were
analyzed by Universal Analysis 2000 Version 4.5A.

Thermogravimetric analysis (TGA)
A sample of 10-15 mg of composite samples was loaded to the Pt pan and the
experiments were conducted in nitrogen atmosphere. The temperature range scanned was from
25°C to 500°C with a 25°C/min ramp rate and a 4°C/min resolution upon thermal decomposition
events.

38 
 

a

b

Figure 2-2: (a) Schematics for in plane thermal conductivity measurement of the two roll milled
sample; (b) Photograph of the sample prepared from compressing the two roll milled sheets to a
desired thickness (1 inch) and a circular cross section of that was produced by a punch.

39 
 

2.5 Results and discussions:
2.5.1 Morphology
Figure 2-3(a) shows the fractured cross section of the 4wt% GNP-1/paraffin
nanocomposites prepared by casting. GNP are either embedded in the paraffin matrix evidenced
by the white lines protruding from the background or lying on the surface indicated by their
irregular but sharp edges different from the smooth and soft paraffin wax. At the same nanofiller
loading, GNP-15 particles in the nanocomposites are more easily recognized due to their larger
size shown in Figure 2-3(b). It is also found that unlike the morphology reported by Kalaitzidou
[11] in PP/GNP nanocomposites that most nanoplatelets tend to roll up or fold on itself during
sample preparation (melt extrusion), GNP in paraffin maintained their platelet shape as a result
of the low viscosity of the polymer matrix and thus the high aspect ratio of GNP is not negatively
affected in sample fabrication. In the samples prepared by two roll mill with GNP loading at
4wt%, 10wt% shown in Figure 2-3(c-d). It is very clear that particles assumed a preferential
alignment as indicated by the white lines in SEM. Figure 2-3(e-f) shows the magnified edge
views of the nanoplatelets, it is also evident that the individual nanoplatelet is thin with thickness
less than 10 nm and thus they have a very high aspect ratio. The distance between particles was
increased due to the expansion of material at the rolling nip point and it is the same compressive
and shearing force that contributed to the alignment of the particles.

 

40 
 

 
 

Figure 2-3: a: 4wt% GNP-1/paraffin by casting; b: 4wt% GNP-15/paraffin by casting; c: 4wt%
GNP-15/paraffin by two roll mill; d: 10wt% GNP-15/paraffin by two roll mill; e: enlarged edge
view of the nanoplatelets (scale bar 100 nm); f: enlarged edge view of one platelet (scale bar: 10
nm)

41 
 

Figure 2-3 (cont'd)

42 
 

Figure 2-3 (cont'd)

 

2.5.2 Electrical conductivity of paraffin/GNP nanocomposites:
In order to find out the effect of dispersion of GNP on electrical conductivity of the
resulting nanocomposites, the resistivity of paraffin/GNP-1 and paraffin/GNP-15 samples were
measured and compared in Figure 2-4. For nanocomposites with GNP-15 as fillers, resistivity
43 
 

undergoes a sharp decrease from 1010 (ohm cm) to around 103 (ohm cm) at 1 vol% which
suggests that conductive paths were formed in paraffin. However, GNP-1 nanocomposites did
not percolate even at 2 vol%. Despite the fact that the absolute number of GNP-1 is almost 250
(15×15) times more than GNP-15 in the composites at the same loading level, the aspect ratio of
is much smaller (100 vs. 1000). It can be seen from SEM micrographs that the platelet like
morphology for larger nanoplatelets were maintained better than smaller ones, which suggests
the importance of filler morphology and aspect ratio on the electrical conductivity of
nanocomposites.

Resistivity (ohm cm) Log scale

12
10
8
6

GNP-15 composite
PCM

4

GNP-1 composite
PCM

2
0
0

2
4
GNP loadings (vol%)

6

 

Figure 2-4: Electrical resistivity of GNP/paraffin composite PCM
 

2.5.3 Thermal conductivity paraffin/GNP nanocomposites:
Thermal conductivity of paraffin/GNP nanocomposites prepared by casting at different
filler concentrations was measured and shown in Figure 2-5(a). The thermal conductivity
increased linearly for GNP-15 samples up to 2 vol% beyond which the data exhibits super linear
behavior. In GNP-1 composites, thermal conductivity also increased linearly with GNP loading,
44 
 

but the improvement is not as effective as GNP-15 although the absolute number of nanofiller
particles is much larger at the same loading level. Given the fact that GNP-15 is superior to
GNP-1 in improving the thermal conductivity of paraffin nanocomposites, two samples prepared
by two roll mill containing 2vol% and 5vol% GNP-15 in paraffin matrix were used for the
measurements. Both samples show higher thermal conductivity in the in-plane direction rather
than in the through-plane direction shown in Figure 2-5(b). Aligning the particles in its basal
plane direction effectively utilizes the high in-plane thermal conductivity of the graphite
nanoplatelets. However, thermal conductivity of the two roll milled sample is not as high as the

Thermal conductivity (W/mK)

cast mold samples at the same loading level.

3.5

a

3
2.5
2

GNP-15
nanocomposites

1.5

GNP-1
nanocomposites

1
0.5
0
0

2
4
GNP volume %

6

Figure 2-5: (a) Through plane thermal conductivity of GNP/paraffin prepared by casting; (b)
Through-plane (left column) and in-plane (middle column) thermal conductivity of GNP15/paraffin prepared by two roll mill and through-plane thermal conductivity by casting (right
column).

45 
 

Figure 2-5(cont'd)

Thermal conductivity (W/mK)

3.5

b

3
2.5
2

two roll mill
through plane

1.5

two roll mill in
plane

1

cast through
plane

0.5
0
2
5
GNP-15 volume %

 
 

2.5.4 Thermal interface resistance:
It is obvious that the formation of a continuous chain structure of highly thermally
conductive GNP particles inside the polymer is not making a large contribution to the heat
conduction of the composites. In fact, similar phenomenon where thermal percolation is not
observed or even a slight drop in thermal conductivity in the vicinity of electrical percolation
threshold has been reported by other research groups in carbon nanotubes based nancomposites
[12-14]. This interesting effect can be explained by two counteracting processes: incorporation of
GNP particles into the matrix introduces new interfaces between the filler and the matrix giving
rise to a thermal interfacial resistance (Kapitza resistance). On the other hand, noticeable
increase in the number of GNP particles creates a highly conducting phase in the matrix
contributing to a better heat conduction. The fact that thermal conductivity only shows a slight
46 
 

increase in GNP/paraffin possibly indicates that the latter process still dominates. Many
researchers have attributed the lack of thermal percolation to the presence of a high thermal
interfacial resistance resulting from a weak coupling between the filler and the polymer matrix.
In order to find out different parameters that might result in a large interfacial resistance, Xu et al.
used molecular dynamics simulation to investigate heat conduction in carbon nanotubes based
nanocomposites and the author proposed two major approaches to improve interfacial thermal
conductivity (i.e. phonon spectrum matching and better interfacial coupling) as shown in Figure
2-6 [15].

 

Figure 2-6: overview of the dominate factors in interfacial thermal conductance in carbon based
nanocomposites as well as possible techniques to reduce interface resistance. Adapted from ref.
15.
 

With the above two approaches in mind, we proposed the following discussion for
GNP/paraffin nanocomposites studied here. Heat transport in graphene nanoplatelets takes place
by phonons of varying frequencies. Phonons are transported from one platelet to another via the
polymer in between. Since there is a difference in acoustic properties, the phonon spectrum will
be different in these two materials. In addition, due to the weak van der Waals interaction
between paraffin and GNP, it is believed that only low frequency phonon vibration modes are
47 
 

available to carry heat energy. The high frequency phonons, which also carry heat energy, need
to interact with other phonons before they can be transferred to low energy vibration states to
couple with the polymer matrix according to a theoretical study conducted by Luo et al. [16]. As
a result of this combined effect of phonon mismatch as well as weak coupling with the matrix,
strong phonon scattering occur at the interfaces that corresponds to a high thermal interface
resistance reducing the efficiency of heat transfer in nanocomposites. This process would be
more prevalent in nanocomposites containing finer filler particles due to a greater density of
interfaces because the chance of phonon scattering at the interface is greater as in the case of
GNP-15 and GNP-1 composites.
In terms of nanofiller alignment, random orientations of GNP in paraffin especially at
higher GNP loading for the cast molded samples led to significant overlapping among particles
which increased the geometric overlap of phonon modes due to a larger physical contact area. On
the other hand, two roll mill process not only aligned the particles in the matrix, but also broke
GNP aggregates and separated them, disrupting the percolating networks and introducing more
phonon scattering at GNP/paraffin interfaces.
As already mentioned previously, thermal interface resistance presents a significant
barrier for phonons to transport from one conductive filler particle to another surrounded by a
polymer matrix, it is interesting to predict the thermal interface resistance from experimental data.
Many models have been proposed to predict the thermal conductivity of nanocomposites, but
few have taken the interfacial resistance into account. Nan proposed a model based on carbon
nanotubes and assumed an interfacial barrier layers between the nanotubes and the polymer, a
Maxwell-Garnett type effective medium approach was used to predict the thermal interface
resistance [10]. For the case of isotropically oriented ellipsoidal particles including platelets, the

48 
 

thermal conductivity ratio Ke/Km (defined as effective thermal conductivity vs. matrix thermal
conductivity) can be expressed by Equation (2.2):

K e 3  f ( x   z )

Km
3  f x
g
g
K33
2(K11  Km )
x 
; z 
1
g
K11  Km
Km

(2.2)

where f is the volume fraction of graphite platelets, Assume individual GNP is coated with a
very thin interfacial thermal barrier layer,

g
K11

and

g
K33

are the equivalent thermal

conductivities along transverse and longitudinal axes of a nanoplatelet, and can be expressed in
Equation (2.3):

K 
g
11

1

Kg
g
; K 33 
2a k K g
tK m

1

where t and d are the thickness and diameter of GNP;

Rk

(2.3)

dK m

ak is the Kapitza radius defined by:

ak  Rk K m
where

Kg
2a k K g

(2.4)

is the thermal interface resistance. The model described above assumes that the

nanofiller particles are completely surrounded by the polymer matrix and they do not touch each
other. Since the percolation threshold for GNP/paraffin composites is approximately 1 vol% as
suggested by electrical resistivity measurement, the experimental data used for curve fitting are
49 
 

limited to those below 1 vol%. Assume the diameter and thickness of the GNP-15 particles are
15 µm and 10 nm and the in-plane and through-plane thermal conductivities are 3000 W/m K
-8

2

and 10 W/m K. The thermal interface resistance of approximately7 to 9×10 m K/W produces
the best fit with experimental results. This Rk value is very close to the interfacial resistance
across the carbon nanotube matrix reported by Cahill and Keblinski group which is about

8 . 3  10  8 m 2 K / W

[17].

Figure 2-7 shows the curve fitting result using this model

assuming various interfacial resistances.

Thermal conductivity ratio (ke/km)

Experimental

Rk=1E-8

Rk=3E-8

Rk=5E-8

Rk=7E-8

Rk=9E-8

10

1
0.1

1
Volume fraction (%)
 

Figure 2-7: Thermal conductivity of paraffin/GNP-15 fitted with effective medium theory below
electrical percolation.

50 
 

2.5.5 Effect of GNP particles on other thermophysical properties of nanocomposites:
Differential scanning calorimetry was used to investigate the effect of GNP on thermo physical properties of paraffin nanocomposites and its capacity to store thermal energy. Both the
melting and cooling behaviors of paraffin/GNP were recorded, two peaks were observed from
the DSC analyses which correspond to solid-solid transformation and melting/crystallization of
paraffin wax. The relevant enthalpies were calculated by integration the peaks above the base
line given by the software itself.
The temperatures that characterize the melting and crystallization peak are not affected
significantly by addition of GNP particles. The melting and crystallization enthalpies for paraffin
nanocomposites at low GNP concentration did not deviate from neat paraffin very much as
opposed to the samples at high loading levels when the composite’s ability to absorb and release
heat is degraded because of the increasing replacement of paraffin wax with GNP particles.
However, the presence of more GNP in the paraffin matrix did not negatively affect the thermal
physical properties of wax itself. This can be further supported by compensating the presence of
GNP in the composites and recalculating the melting and crystallization enthalpies for paraffin
wax alone in the nanocomposites using Equation (2.7). It is found that for both nanocomposites,
the heat absorbing and releasing capacity during heating and cooling for paraffin did not degrade
but improve slightly with GNP loadings.

H wax 
                         

H comp
1  xG

51 
 

                                                              (2.7)

where

ΔHcomp is latent heat of composite PCM with GNP. ΔHwax is latent heat of wax in

PCM. xG is the weight fraction of xGnP. Table 2-1 and Table 2-2 collected data from paraffin
nanocomposites containing 0wt%, 2wt%, 4wt% and 8wt% GNP-15 or GNP-1.
Table 2-1: Heating and cooling characteristics of GNP-15 composite PCM (average of three
samples)
Sample
names

Melt
peak(°C)

Neat
2wt%
4wt%
8wt%

54.54
54.82
55.13
54.9

Melting
enthalpy of
composite
(J/g)
185.9
188.6
183.5
171.9

Melting
enthalpy of
wax (J/g)

Cryst.
peak(°C)

185.9
192.4
191.1
186.8

51.73
51.27
51.87
51.63

Cryst.
enthalpy of
composite
(J/g)
186.8
189.9
185
172.9

Cryst.
enthalpy of
wax (J/g)
186.8
193.7
192.7
187.9

Table 2-2: Heating and cooling characteristics of GNP-1 composite PCM (average of three
samples)
Sample
names

Melt
peak(°C)

Neat
2wt%
4wt%
8wt%

54.54
54.33
53.97
54.35

Melting
enthalpy of
composite
(J/g)
185.9
186.05
180.4
176.65

Melting
enthalpy of
wax (J/g)

Cryst.
peak(°C)

185.9
189.85
187.9
192

51.73
52.09
52.22
51.94

Cryst.
enthalpy of
composite
(J/g)
186.8
187.15
180.2
178.1

Cryst.
enthalpy of
wax (J/g)
186.8
191
187.7
193.5

2.5.6 Thermal gravimetric analysis:
Since graphite is known for its chemical inertness and resistance to thermal degradation,
thermal decomposition of paraffin nanocomposites upon addition of GNP was analyzed by TGA
in N2 shown in Figure 2-8. It is shown in the analysis that thermal decomposition of paraffin
nanocomposites gradually shifted to higher temperature with increasing GNP. Although the
interaction between GNP and paraffin is weak, they still interact with the surrounding matrix and
52 
 

improve the overall stability of composite PCM. The remaining weight at 500oC, at which
temperature paraffin wax was completely thermally decomposed but GNP were unaffected, was
close to the initial loadings 0wt%, 2wt%, 4wt% and 8wt% respectively.

a

2wt% GNP15/Paraffin

4wt% GNP15/Paraffin

Weight retained()%

Neat Paraffin

8wt% GNP15/Paraffin

1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0

100

200

300

400

500

600

Temperature (C)

Figure 2-8: (a) TGA analysis of GNP-1/paraffin nanocomposites in N2. (b) TGA analysis of
GNP-15/paraffin nanocomposites in N2

53 
 

Figure 2-8(cont'd)

b

2wt% GNP1/Paraffin

4wt% GNP1/Paraffin

Weight retained(%)

Neat Paraffin

8wt% GNP1/Paraffin

1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0

100

200
300
400
Temperature (C)

500

600
 

2.6 Conclusions:
GNP of two different sizes (GNP-1 and GNP-15) were successfully mixed with paraffin
wax to make a composite phase change materials (PCM) of high latent heat and high thermal and
electrical conductivity. In particular, GNP-15 nanocomposites prepared by casting show lower
electrical percolation and higher thermal conductivity than GNP-1 counterparts possibly because
of the large difference in the aspect ratio of the two nanofillers. The distinctive property
enhancements was also observed at the electrical percolation where electrical conductivity
increased by many orders of magnitude while thermal conductivity remained almost the same.
Lack of thermal conductivity in the nanocomposites can be explained by different mechanisms
for electron and phonon transport across the polymer/GNP interface. In particular, phonon

54 
 

transport is dependent on the physical properties (phonon spectra) of the two components as well
as their coupling strength, a high interfacial resistance due to phonon scattering is the major
reason for inefficient heat transfer. In comparison, electron transport across interfaces takes place
by electron tunneling where physical contact between filler particles is not required. The
interfacial contact resistance is believed to account for the much lower than expected
improvement in thermal conductivity of the nanocomposites. The interfacial thermal resistance
estimated by Nan’s model to be 7 to 9×10-8m2K/W, was found to be comparable to that of carbon
nanotubes. For the purpose of thermal energy storage, DSC results show the latent heat of
paraffin nanocomposites were not degraded by adding GNP. Latent heat of paraffin alone and the
thermal stability of the composite PCM measured by TGA actually improved slightly in the
presence of GNP.

Acknowledgements
The author appreciated the generous help from Brian Rook from Composite Materials and
Structures Center on operating the two roll mill. Financial support for this project was provided
by the Michigan Economic Development Corporation through the 21st Century Jobs Fund.

55 
 

 
 
 
 
 
 
 
 
 

REFERENCES
 

 

56 
 

References
1. Hasnain, S.M. Review on sustainable thermal energy storage technologies, Part I: heat storage
materials and techniques. Energy Convers Manage. 1998, 39, 1127-1138.
2. Telkes. M. Thermal storage for solar heating and cooling, in: Proceedings of the Workshop on
Solar Energy Storage Subsystems for the Heating and Cooling of Buildings, Charlottesville,
Virginia, USA, 1975.
3. Pillai, K.K.; Brinkworth, B.J.; The storage of low grade thermal energy using phase change
materials. Appl. Energy. 1976, 2, 205-216.
4. Luyt, A. S. Role of polymers in developing phase change materials for energy storage.
eXpress Polym. Lett. 2008, 2, 147-147.
5. Velraj, R.; Seeniraj, R.V.; Hafner, B.; Faber, C.; Schwarzer, K.; Experimental analysis and
numerical modeling of inward solidification on a finned vertical tube for a latent heat storage
unit. Sol Energy. 1997, 60, 281-290.
6. Fukai, J.; Kanou, M.; Kodama, Y.; Miyatake, O.; Thermal conductivity enhancement of
energy storage media using carbon fibers. Energy Convers Manage. 2000, 41, 1543-1556.
7. Elgafy, A.; Lafdi, K.; Effect of carbon nanofiber additives on thermal behavior of phase
change materials. Carbon. 2005, 43, 3067-3074.
8. Pincemin, S.; Olives, R.; Py, X.; Highly conductive composites made of phase change
materials and graphite for thermal storage. Sol Energy Mater Sol C. 2008, 92, 603-613.
9. Kim, S.; Drzal, L. T.; High latent heat storage and high thermal conductive phase change
materials using exfoliated graphite nanoplatelets. Sol Energy Mater Sol C. 2009, 93, 136-142.
10. Nan, C. W.; Liu, G.; Lin, Y.; Li, M. Interface effect on thermal conductivity of carbon
nanotube composites. App Phys Lett, 2004, 85, 3549-3551.
11. Kalaitzidou, K.; Fukushima, H.; Drzal, L. T.; Multifunctional polypropylene composites
produced by incorporation of exfoliated graphite nanoplatelets. Carbon. 2007, 45, 1446-1452.
12. Mamunya, Y.; Boudenne, A.; Lebovka, N.; Ibos, L.; Candau, Y.; Lisunova, M. Electrical and
thermophysical behavior of PVC-MWCNT nanocomposites. Compos Sci Technol 2008, 68,
1981-1988.

57 
 

13. Grunlan J. C.; Kim, Y. S.; Ziaee, S.; Wei, X.; Abdel-Magid, B.; Tao, K. Thermal and
mechanical behavior of carbon-nanotube-filled latex. Macromol Mater Eng 2006, 291, 10351043.
14. Moisala, A.; Li, Q.; Kinloch, I. A.; Windle, A. H. Thermal and electrical conductivity of
single and multi-walled carbon nanotube-epoxy composites. Compos Sci Technol 2006, 66,
1285-1288.
15. Xu, Z.; Buehler, M.J. Nanoengineering heat transfer performance at carbon nanotube
interfaces. ACS Nano, 2009, 3, 2767-2775.
16. Luo. T.; Lloyd, J. R. Non-equilibrium molecular dynamics study of thermal energy transport
in Au-SAM-Au junctions. , Int. J. Heat Mass Transfer, 2010. 53, 1-11.
17. Huxtable, S.; Cahill, D. G.; Shenogin, S.; Xue L.; Ozisik, R.; Barone, P et al. Interfacial heat
flow in carbon nanotube suspensions. Nat Mater. 2003, 2, 731-734.
 

 

58 
 

Chapter 3 Thermal Conductivity of a Monolayer of Exfoliated
Graphene Nanoplatelets Prepared by Liquid-Liquid Interfacial SelfAssembly	

3.1 Abstract:
A monolayer film composed of exfoliated graphene platelets GNP was extracted from a
chloroform-water interface and supported on a glass substrate. The nanoplatelets are
interconnected at the edges without overlapping forming a very densely packed structure with
uniform thickness. Micro-Raman spectroscopy with a 50 mW 532 nm laser generating heat at the
center of the sample was used to probe the thermal conductivity of the GNP monolayer at
different power levels. The Raman G peak shift of graphite was used to record the local
temperature rise in the monolayer. The cross sectional area of heat conduction is determined by
the thickness of individual platelets. A UV-Vis spectrometer was used to measure the absorption
of light by the monolayer. Depending on the interface density, thermal conductivities are
estimated to be 380 W/m K and 290 W/m K for monolayers with average particle size of 10 µm
and 5 µm, respectively. This chapter is based on a journal publication: Xiang, J., Drzal, L. T
(2010) Thermal Conductivity of a Monolayer of Exfoliated Graphite Nanoplatelets Prepared by
Liquid-Liquid Interfacial Self-Assembly. Journal of Nanomaterials, 2010,1-6.

 

59 
 

3.2 Introduction
This chapter discusses a novel approach to visually inspect the heat conduction in a
percolated network of pure GNP, providing insights into the effect of interfacial resistance on
thermal conductivity.
Earlier work has shown that graphene supports 2D phonon propagation giving rise to a
remarkably high thermal conductivity in its basal plane [1]. It is difficult to use conventional
techniques such as 3ω, laser flash, and thermal bridge to measure thermal conductivity of this
very delicate system such as single layer graphene without causing physical damage during
sample preparation and handling. Confocal micro-Raman microscopy, a non-destructive, optical
technique has been successfully used to extract thermal properties of thin films [2,3]. Balandin et
al. successfully measured thermal conductivity of a mechanically cleaved single layer graphene
ribbon with Raman spectroscopy that uses the Raman G peak of graphene as a probe for local
temperature determination at the laser spot. The measured thermal conductivity of graphene is
4400-5300 W/m K [4]. Similarly with Raman spectroscopy, Geim et al used temperature readout
obtained from the intensity ratio of Stokes to anti-Stokes signal to study the thermal properties of
a large circular shaped graphene membrane also obtained from micromechanical cleavage and
they concluded thermal conductivity to be 600 W/m K [5]. These single layer graphene were all
suspended over a micro trench with their edges connecting to either bulk graphite or gold heat
sinks. Recently, Ruoff et al reported a CVD grown graphene monolayer suspended or supported
over a holey Au/SiN substrate. The thermal conductivity of graphene was measured by
monitoring the graphene G peak shift with laser heating. The measured conductivities for
suspended and supported graphene are 2500+1100/-1050 W/m K and 370+650/-320 W/m K
respectively [6]. It was surmised that the huge difference is due to the phonon leakage and

60 
 

scattering at the substrate-graphene interface. More recently, Balandin et al investigated the
dimensional crossover of thermal transport in few-layer graphene (FLG) using the same
technique, they showed that the room temperature thermal conductivity changes from 2800 W/m
K to 1300 W/m K as the number of atomic planes in FLG increases from 2 to 4 [7].
Despite the numerous literature that focus on the thermal conductivity of graphene and
few-layer graphene, there is no report that investigates the thermal conductivity of a monolayer
film consisting of individual graphene nanoplatelets. Here, a monolayer film of GNP was
extracted from a liquid-liquid interface following the procedures reported by Biswas [8] and
deposited on a substrate for thermal conductivity measurement by confocal Raman spectroscopy
that shed light on the effectiveness of those highly conductive nanoplatelets in heat conduction.
In a monolayer, individual GNP is lying flat on the glass substrate and densely connected to each
other resembling a percolated conductive network in a nanocomposite except that the
nanoparticles are highly ordered with well defined contacts instead of a random distribution of
particles with uncontrolled contacts. Heat spreads in a 2D plane in the monolayer and the
efficiency of heat conduction of such monolayers was studied.

3.3 Experimental details:
Materials:
Two kinds of GNP produced from different graphite sources were measured and their
surface areas were determined by Brunauer-Emmet-Teller (BET) using N2 adsorption at 77K to
be 20-40 m2/g and 200-220 m2/g, and are referred to as low surface area (LSA) GNP and high
surface area (HSA) GNP in later discussion. Chloroform, used for liquid-liquid interface

61 
 

extraction of a monolayer of GNP was purchased from J.T. Baker and used directly without
purification.
Extraction of a Monolayer GNP at the Liquid-Liquid Interface:
A monolayer of densely packed GNP particles was prepared on the glass substrate
following the procedure by Biswas [8]. To summarize the process, GNP were dispersed in
chloroform at 0.1 mg/ml. The supernatant of the suspension was obtained by 10 min
centrifugation at 5000 rpm. Then water was added to the supernatant solution followed by high
power sonication. GNP particles then quickly adsorbed at the chloroform-water interface and a
monolayer was extracted on a glass slide. The samples were dried in air and then annealed at
250 °C for 2 h. The area of GNP coated region is 3 cm × 2.5 cm.
UV/Vis Spectroscopy:
The optical transparency of the monolayer coated glass slide was measured by UV-Vis
spectrometer (Perkin Elmer Lambda 900) to determine the amount of light absorbed by the
monolayer. The amount of light transmitted was measured and the light absorbed by GNP
monolayer was calculated from subtracting the percentage transmitted from the total light source.
Scanning Electron Microscopy
A JEOL JSM-7500F scanning electron microscope was used to examine the morphology
of the monolayer on the glass slide. The sample was not gold coated and the imaging was done at
5kV at working distance of 4.5 mm.
Confocal Raman Spectroscopy:
A LabRAM ARAMIS laser Raman spectrometer (Horiba Scientific) with 532 nm 50mW
DPSS laser was used to probe the G band of GNP particles. To determine the temperature
62 
 

dependence of G peak, an INSTEC (STC 200) heating stage was used to heat the sample and a
very low laser power was used to collect the Raman signal to avoid laser induced heating.
Monolayers of both LSA and HSA GNP were prepared on the microscopic glass slides for
subsequent measurements. The laser was focused on the center of the monolayer and a filter that
controls different amount of laser coming through the objective lens was used to induce different
levels of heating in the samples. A D1, D06, D03, D0 filter was used in each experiment that
corresponds to 10%, 25%, 50% and 100% laser power. Both the 50× and 100× objective lenses
were used for each sample measured. The laser spot size r0 was 0.24 µm and 0.19 µm for the 50×
and 100× respectively which were determined using r0=λ/πNA, where NA is the numerical
aperture value of 0.9 and 0.75 for the ×100 and ×50 objectives.

3.4 Results and discussions:
3.4.1 Scanning electron microscopy:
The formation of a stable monolayer of GNP particles that were closely connected
without overlapping was driven by the gain in interfacial energy that enables the nanoplatelets to
be adsorbed to the oil-water interface [9-11]. Since these graphene nanosheets are highly
hydrophobic on their basal planes while the oxygen functional groups on the edges render them
hydrophilic, it is expected that a meniscus will be generated around the particles and ‘like’
menisci overlap to generate large attractive capillary force to form a close packed structure at the
liquid-liquid interface [11,12]. While the magnitude of this lateral capillary force is negligible for
spherical particles with diameter less than 10 µm [13], Bowden et al showed that, for planar
shaped particles at the liquid-liquid interface, the lateral capillary force is quite significant even
63 
 

for particles with thicknesses of few nanometers [11]. Depending on the starting material and
intercalation and exfoliation conditions, the average thickness of individual nanoplatelets may
vary which will be reflected in the optical transparency measurement.
Figure 3-1(a,b) shows the SEM images of the two glass substrates coated with LSA and
HSA GNP monolayers. From SEM micrographs, individual platelets come into contact with each
other at the edges (indicated by arrows in the figures) so well that there is no noticeable gap left
between them thanks to the more regular geometry of the platelet edges. In some portions of the
glass substrate, gaps can be observed between nanoplatelets but every single one is connected to
the adjacent particle without overlapping for both the LSA and HSA monolayers which proves
both interfacial energy minimization and the strong edge interaction among nanoplatelets. With
multiple inspection of different parts of the monolayers, it is concluded that the LSA monolayer
has an average particle size of around 5-10 µm. In the HSA monolayer, however, more gaps
exist among individual platelets and the average size of the platelet is within 1-5 µm. It is
obvious from the SEM images that the interface density in the HSA monolayer is higher than the
LSA monolayer.

64 
 

Figure 3-1: (a) SEM image of a LSA GNP monolayer film supported on a glass substrate. (b)
SEM image of a HSA GNP monolayer film supported on a glass substrate.

65 
 

3.4.2 Optical Transmission and Absorption of GNP Monolayers:
The optical transparency of those monolayers was determined by scanning a blank glass
slide and a GNP coated slide to eliminate the effect of reflective loss from the slides. Since the
reflection of the graphite nanosheets is negligibly small [14], it is not accounted for in this
calculation. The amount of light absorbed was calculated by subtracting the light transmitted
through the glass slide:

 

Figure 3-2: Percent transmission of light for LSA and HSA monolayers scanning from 350 nm to
1000 nm. Inset is the photo of LSA and HSA coated monolayers supported on glass substrates as
well as the blank slide.
 

66 
 

1580.5
1580
1579.5
G peak (cm-1)

1579
1578.5
1578
1577.5
y = -0.0216x + 1586.4
R² = 0.9702

1577
1576.5
1576
1575.5
250

300

350
400
Temperature (K)

450

500
 

Figure 3-3: Raman G band shift of GNP with temperature and linear curve fitting with Microsoft
Excel. Two measurements were conducted indicated by blue and red marks.

A (absorption) = 1-T (transmission). Figure 3-2 shows the percentage transmission of light for
both the LSA and HSA monolayers from 350 nm to 1000 nm. The optical transparency for the
LSA and HSA monolayer is 20.5 % and 49.6 %, respectively at the 532 nm. The absorption is
then calculated to be 79.5 % and 50.4 % for the LSA and HSA monolayers. The fact that the
HSA GNP went through better intercalation and exfoliation contributes to the production of
much thinner nanoplatelets, which boosts the optical transparency. The thickness of the
individual HSA GNP and LSA GNP were previously measured to be around 10 nm and 20 nm
respectively by AFM [8].

67 
 

3.4.3 Confocal Micro-Raman Spectroscopy Analysis:
The laser induced heating of the graphene nanoplatelets causes a red shift of the G band
because of bond softening. The G band shift is believed to be linearly dependent on the
temperature of the sample [15]. A heating stage was used to heat individual GNP and the Raman
signal was collected at a very low laser power. Two measurements were taken at each
temperature up to 473K. The slope of the temperature dependence of G peak was determined to
be -0.021 cm-1/K shown in Figure 3-3. However, in the monolayer samples, heat absorbed by
GNP propagates to the edge of the platelet and the large thermal interface resistance is expected
to scatter the phonons (lattice vibrations). This is in stark contrast to constructing a graphite or
Au heat sink around an individual nanoplatelet. A phonon being transported along the monolayer
encounters numerous interfaces that reduce the heat conduction efficiency. It is necessary to
focus the laser on the center of the monolayer to ensure that heat spreads uniformly in a circular
geometry. In fact, if the laser beam was focused on the edge of the sample, the local temperature
will increase much higher due to the presence of a boundary that impedes heat dissipation.
During the experiment, the edges of the monolayer sample was kept at ambient temperature at
around 300K, additional heat sinks were not used because of the huge surface area of the
monolayer compared with the laser spot size. The Raman G-peak shift was collected at different
laser power levels (i.e. 100%, 50%, 25% and 10%) with the selection of a proper filter. Figure 34(a,b) shows the G-peak shift at different power levels for both the LSA and HSA monolayers.
The presence of oxygen functional groups on the edges of GNP are likely to alter the carbon
bonding from sp2 to sp3 type and this gives rise to a D band in the Raman signal (1350 cm-1).
Due to the difference in the size of the nanoplatelets in LSA and HSA monolayer, the HSA

68 
 

sample has more edge areas exposed to the laser and this provides a possible explanation to a
higher D band observed in HSA samples.

Figure 3-4: (a) Raman G peak shift of the LSA monolayer at different laser power levels. (b)
Raman G peak shift of the HSA monolayer using different laser power levels.

69 
 

3.4.5 Calculation of the Thermal Conductivity of GNP Monolayer:
The temperature rise at the center of the monolayer caused by the Gaussian-shaped laser
heating is proportional to the absorbed laser power and inversely proportional to the efficiency of
heat conduction. The volumetric laser heating is qv=q0/t exp(-r2/r02) with r0 being the laser spot
size, r being the radial position measured from the center of the laser beam. Here, q0 is the
absorbed laser power per unit area at the laser spot and t is the thickness of monolayer thickness
which is the same as the thickness of individual platelets. The glass substrate (κ=1 W/m K)
interacts with the phonons and causes phonon leakage which reduces the thermal conductivity.
The temperature distribution of the supported monolayer can be obtained by the steady state
form of the heat diffusion equation in cylindrical coordinate following Cai’s approach [6]:

q
1 d dT g
(r )  (T  Ta )  v  0
r dr dr t


(3.1)

where g is the interfacial thermal conductance per unit area between the monolayer and the
substrate as well as the air molecules. κ is the thermal conductivity of the monolayer,

Ta is the

ambient temperature. From the Gaussian-shaped laser heating, the total absorbed power by the
monolayer Q is:

Q  t  q v 2 rdr  q 0 r0
Let

θ= T-Ta

and

(3.2)

z= (g/κt)1/2 r, equation (3.1) can be converted to a nonhomogeneous

Bessel’s equation:

70 
 

2

q0
 2 1 
z2
    exp(  2 )

z 2 z z
g
z0

(3.3)

The solution to the above equation is:

 ( z )  C1 I 0 ( z )  C 2 K 0 ( z )   p ( z )

(3.4)

where I0(z) and K0(z) are the zero-order modified Bessel functions of the first and second kind.
The particular solution θp(z) can be obtained using the variation of parameters method and is
found to depend on q0/g [16]. The temperature rise in the monolayer measured by Raman laser
is:


z2
  ( z ) exp(  z 0 2 ) zdz
m  0 
z2
 exp(  z 0 2 ) zdz
0

71 
 

(3.5)

1000

a

Temperature (K)

900

y = 16.114x + 254.14

800
700
600

x50

500

x100

400

y = 11.175x + 265.04

300
200
0

1200

10

20
30
40
Absorbed power (mW)

b

1100

y = 35.36x + 235.87

1000
Temperature (K)

50

900
800
700

x50

600

x100

500

y = 23.127x + 259.82

400
300
200
0

10
20
Absorbed power (mW)

30
 

Figure 3-5: (a) Local temperature of an individual GNP at the center of a LSA monolayer vs. the
absorbed laser power with x50 and x100 objectives. The solid lines represent the linear curve
fittings. (b) Local temperature of an individual GNP at the center of a HSA monolayer vs. the
absorbed laser power with x50 and x100 objectives. The solid lines represent the linear curve
fittings.

72 
 

The measured thermal resistance defined as

R=θ m/Q

can be calculated from equation

3.2 and 3.5. Given the temperature dependence of Raman G peak, the temperature rise

θ m and

absorbed power Q is found to be linearly related as is shown in Figure 3-5(a,b) and the slope of
which yields RLSA equal to 1.12×104 K/W and 1.61×104 K/W respectively for the LSA
monolayer with 50× and 100× objectives. Similarly, RHSA equals to 2.31×104 K/W and 3.54×104
K/W for the HSA monolayer with 50× and 100× objectives. The ratio of the two R values is
found to be dependent only on g/κ. With one of the R values measured with either objective, g2/κ
can be further obtained. After a curve fitting procedure for g/κ and g2/κ, it is found that g≈50
MW/m2K and κLSA≈380 W/m K for the LSA monolayer while g≈30 MW/m2K and κHSA≈290
W/m K for the HSA monolayer. Due to the difference in GNP size, the HSA monolayer has a
higher interface density than the LSA monolayer which might cause the reduction in overall
efficiency of heat conduction due to strong phonon scattering at the particle interfaces. 
 

3. 5 Conclusions:
A densely packed monolayer of GNP was successfully extracted from a chloroformwater interface driven by interfacial energy minimization and attractive capillary interactions.
Confocal micro-Raman spectroscopy was used as a heater and a thermometer to probe the heat
conduction efficiency of monolayers made from GNP of different sizes and thicknesses. The
presence of interfaces in the monolayers significantly affects thermal conductivity, reducing the
thermal conductivity in this assembly of graphene nanoplatelets by almost an order of magnitude
lower than the highest reported value in the literature measured by the same technique. However,

73 
 

the in-plane thermal conductivities of both monolayers are still high with negligible cross plane
conduction making it a promising candidate for applications that require fast heat dissipation to
reduce hot spots.

Acknowledgement:
The authors wish to acknowledge the support of the State of Michigan 21st Century Jobs Fund
for partial support of this research.

74 
 

REFERENCES

75 
 

References
 

1. Nika, D. L.; Ghosh, S.; Pokatilov, E. P.; Balandin, A. A. Lattice Thermal Conductivity of
Graphene Flakes: Comparison with Bulk Graphite. Appl. Phys. Lett. 2009, 94, 203103.
2. Perichon, S.; Lysenko, L.; Remaki, B.; Barbier, D.; Champagnon, B. Measurement of Porous
Silicon Thermal Conductivity by Micro-Raman Scattering, Appl. Phys. Lett. 1999. 86, 47004702.
3. Lysenko, V.; Perichon, S.; Remaki, B.; Barbier, D.; Champagnon. B. Thermal Conductivity of
Thick Meso-Porous Silicon Layers by Micro-Raman Scattering. Appl. Phys. Lett. 1999, 86,
6841-6846.
4. Balandin, A. A.; Ghosh, S.; Bao, W.; Calizo, I.; Teweldebrhan, D.; Miao, F et al. Superior
Thermal Conductivity of Single-Layer Graphene, Nano Lett. 2008, 8, 902-907.
5. Faugeras, C.; Faugeras, B.; Orlita, M.; Potemski, M.; Nair, R. R.; Geim, A. K. Thermal
Conductivity of Graphene in Corbino Membrane Geometry. ACS Nano, 2010, 4, 1889-1892.
6. Cai, W.; Moore, A. L.; Zhu, Y.; Li, X.; Chen, S.; Shi, L et al. Thermal Transport in Suspended
and Supported Monolayer Graphene Grown by Chemical Vapor Deposition. Nano Lett. 2010, 10,
1645-1651.
7. Ghosh, S.; Bao, W.; Nika, D. L.; Subrina, S.; Pokatilov, E. P.; Lau, C. N et al. Dimensional
Crossover of Thermal Transport in Few-Layer Graphene. Nat Mater, 2010, 9, 555-558.
8. Biswas, S.; Drzal, L. T. A Novel Approach to Create a Highly Ordered Monolayer Film of
Graphene Nanosheets at the Liquid-Liquid Interface, Nano Lett. 2009, 9, 167-172.
9. Lin, Y.; Skaff, H.; Emrick, T.; Dinsmore, A. D.; Russell, T. P. Nanoparticle Assembly and
Transport at Liqiud-Liquid Interfaces. Science, 2003, 299, 226-229.
10. Binder, W. H. Supramolecular Assembly of Nanoparticles at Liquid-Liquid Interfaces.
Angew, Chem. Int. Ed, 2005, 44, 5172-5175.
11. Bowden, N.; Terfort, A.; Carbeck, J.; Whitesides, G. M. Self-Assembly of Mesoscale
Objects into Ordered Two-Dimensional Arrays. Science, 1997, 276, 233-235.
12. Bowden, N.; Oliver, S. R. J.; Whitesides, G. M. Mesoscale Self-Assembly: Capillary Bonds
and Negative Menisci. J. Phys. Chem. B, 2000, 104, 2714-2724.
76 
 

13. Kralchevsky, P. A.; Paunov, V. N.; Denkov, N. D.; Ivanov, I. B.; Nagayama, K. Energetical
and Force Approaches to the Capillary Interactions between Particles Attached to a Liquid-Fluid
Interface, J. Colloidal Interface Sci, 1993, 155, 420-437.
14. Nair, R. R.; Blake, P.; Grigorenko, A. N.; Novoselov, K. S.; Booth, T. J.; Stauber, T et al.
Fine Structure Constant Defines Visual Transparency of Graphene. Science, 2008, 320, 13081308.
15. Calizo, I.; Balandin, A. A.; Bao, W.; Miao, F.; Lau, C. N. Temperature Dependence of the
Raman Spectra of Graphene and Graphene Multilayers. Nano Lett. 2007, 7, 2645-2649.
16. Greenberg M. Advanced Engineering Mathematics, 2nd ed, Prentice Hall, Upper Saddle
River, NJ. 1998.
 

 

77 
 

Chapter 4 Thermal Conductivity of Exfoliated Graphene
Nanoplatelet Paper	
4.1 Abstract
Instead of making a monolayer GNP film with only a few tens of nanometer in thickness, a novel
method was proposed to make multilayer GNP film. In particular, these highly hydrophobic
nanoparticles were dispersed and stabilized in a DI water/polyethyleneimine (PEI, a cationic
polyelectrolyte) solution. A free standing, mechanically robust paper of GNP was prepared by
vacuum filtration. The effect of GNP size, polyelectrolyte coating and paper porosity on thermal
transport properties was investigated. It was found that thermal annealing improved thermal
conductivity by decomposing the PEI molecule that was adsorbed on the GNP particles while
still maintaining the porosity of the paper. Mechanically compressing the sample effectively
reduced the pore volumes within the paper and increased the contact area among individual
platelet. The strong alignment effect and larger contact area was evidenced by a 80% increase in
in-plane thermal conductivity (207  12 W/m·K) and a through-plane conductivity of only 1.7

 0.12 W/m·K. This flexible, lightweight, low-cost, paper material made of GNP particles is a
promising candidate for applications requiring 2D heat conduction. This chapter is based on a
journal publication: Xiang, J., Drzal, L. T. (2011). Thermal conductivity of exfoliated graphite
nanoplatelet paper. Carbon, 49(3), 773-778.

4.2 Introduction:
Free-standing paper-like graphitic materials have great technological importance to
modern society because of their chemical inertness, good mechanical properties, and
impermeability to gases and fluids. Early work has shown that these exfoliated graphite flakes
78 
 

can be rolled into flexible thin sheets and used as gaskets [1-4], flow field plates in fuel cells,
and electrodes for Li ion batteries [5]. The discovery of carbon nanotubes leads to densely
packed CNT mats prepared by vacuum filtration and their superior strength and electrical
conductivity are useful in many applications [6-8]. More recently, it is shown that graphite oxide
can be exfoliated in water to yield hydrophilic, oxygenated graphene oxide nanosheet that can be
assembled macroscopically to a paper-like structure through flow directed assembly [9]. Ruoff et
al also shows that these graphite oxide nanosheets can be chemically reduced to graphene with
hydrazine in the presence of a strong base such as KOH to restore the outstanding mechanical
and electrical properties of graphene [10].
However, there are very few reports on thermal transport properties of those papers.
Hone reported the thermal conductivity of densely packed CNT paper to be a few tens W/m·K,
[11] much lower than conventional inorganic materials for heat dissipation although individual
carbon nanotube is believed to have very high thermal conductivity [12]. A. Balandin et al.
recently used micro-Raman spectroscopy, a non-contact, optical method to determine the thermal
conductivity of single layer graphene, they concluded κ=4400-5300W/m·K, among the highest in
nature [13]. Yet, it remains elusive how well a paper-like material composed of individual
graphite nanosheet conducts heat and how to tailor the thermal conductivity of these papers for
various purposes.
To make a paper with GNP particles, these highly hydrophobic nanoparticles were
dispersed in a polyethyleneimine/water solution. The cationic polyelectrolyte induces steric and
electrostatic stabilization [14] between the nanoplatelets [15] and a paper was prepared by
                                                            


Grafcell (http://www.graftechaet.com/GRAFCELL/GRAFCELL-Home.aspx)

 

79 
 

subsequent vacuum filtration and drying. The thermal conductivity of the as-made GNP paper
consisting of individual platelets of different sizes and made with various processing steps (e.g.
thermal annealing, cold pressing) was measured and related to the heat conduction in this paperlike structure.

4.3 Experimental
Materials:
GNP were prepared as usual by exfoliating the H2SO4/HNO3 intercalated natural graphite
(GIC, A3772) purchased from Asbury Graphite Mills, Inc, NJ. Microwave exfoliation proves
superior to conventional thermal treatments due to the fast heating process and high energy
density, which vaporizes the acids within the layers of graphite causing a significant and rapid
expansion of the graphite gallery. The measured surface area is 20-40 m2/g for GNP-15 and 100130 m2/g for GNP-1.

80 
 

 

Figure 4-1: Typical Raman spectrum of GNP particles showing D peak, G peak and 2D peak.

The Raman spectrum of the GNP particle is shown in Figure 4-1, the small D band and a sharp G
band confirms the sp2 type bonding of the carbon atoms in the basal plane. It is thus believed that
the intercalation process and subsequent pulverization processes did not oxidize the surface of
the GNP which might degrade the electrical and thermal conductivity of an individual platelet
[16].
GNP paper preparation:
Polyethyleneimine (PEI) was purchased from Aldrich. (branched, Mw=25,000). PEI
solution was prepared containing 0.1wt% of the polymer in DI water. 0.1g GNP was dispersed in
100ml PEI solution and bath sonicated for 30 mins under constant stirring. The suspension was
81 
 

kept stirring for 24 h before the excess polyelectrolyte was washed away by filtering through the
0.2 μm Durapore PVDF filter paper and re-dispersed in water at 1mg/ml. The pH of the
suspension was adjusted to around 5-6 to induce positive charges on the polyelectrolyte [17].
The GNP-15 particles, due to larger effect of gravitational sedimentation, settled in few days,
while the GNP-1 particles could be stabilized in the suspension for weeks.
An ANODISC membrane (0.2 μm, Whatman) was used for paper making by vacuum
filtration. A controlled multi-filtration process (5 ml at a time) was used to minimize the
disordered layering structure caused by water flow. 50 ml of the suspension was filtered to make
one paper. The paper was then suction dried for 1h and then put into a vacuum oven and dried at
60ºC overnight before peeling the GNP paper off the membrane. This paper was then referred to
as “as-made” in later discussions. The free-standing paper was then thermally annealed at a
furnace at 340ºC for 2h and is identified as “annealed”. The annealed paper was then coldpressed at 100 psi at room temperature for 1h and is identified as “annealed and cold-pressed”.
Scanning electron microscopy:
A JEOL JSM-7500F was used to image both the top surface and cross section of the GNP
paper. The as-made, annealed and cold pressed papers prepared from both GNP-1 and GNP-15
were examined. For the cross sectional imaging, a portion of the GNP paper was mounted by
epoxy and maintained at a position perpendicular to the polishing surface by a Buehler SamplKlip. Then the sample surface was polished and O2 plasma treated for 45 mins at a RF power of
225W and gold coated for 2 mins before imaging.

82 
 

Thermogravimetric analysis (TGA):
To investigate the effect of thermal annealing on sample thermal conductivity, a sample
of 10-15 mg of GNP paper before and after annealing was loaded to the Pt pan and the
experiments were conducted in a nitrogen atmosphere. The temperature range scanned was from
25°C to 700°C with a 10°C/min ramp rate and a 4°C/min resolution for thermal decomposition
events.
Flash lamp laser diffusivity (Netzsch 447 NanoFlash)
The thermal diffusivity (α, m2/s) of the 1’’ diameter GNP paper was measured by a LFA
Nanoflash 447 Light flash system. Both the in-plane and through-plane thermal diffusivity were
measured for the as-made, annealed, and anneal and cold pressed samples. To calculate the
thermal conductivity of the samples, the apparent densities (ρ: kg/m3) of the samples were
measured by dividing the mass over the volume, and the specific heat capacity (Cp: J/kg K) was
measured by DSC (Q2000, TA instruments). The thermal conductivity (κ) is calculated
according to the equation: κ=    Cp .

4.4 Results and discussions:
4.4.1 Morphology
To create a more ordered layer structure, a multi-filtration process (5 ml at a time) is
necessary. During each filtration, water easily flows through the pores within the paper and the
highly hydrophobic nanoplatelet floats at the water-air interface forming a very densely packed
structure to minimize the interfacial energy. The first few layers were stacked more orderly
because of the uniformity of water flow rate through the tubular shaped pores in the filter paper.
83 
 

However, as the thickness of the paper builds up, the flow of water became more chaotic since
more pores are blocked by the planar shaped GNP particles. It is thus recommended that the total
filtration volume do not exceed 100 ml (at a concentration of 1 mg/ml) for maintaining the
layered structure within the paper. Figure 4-2 shows schematic representation of filtration and
the photos of the as-made self-standing, and mechanically robust GNP paper after being peeled
off the filter paper.

Figure 4-2: Schematic representation of the GNP paper making process by vacuum assisted self
assembly and actual photos of GNP paper peeled off the ANODISC filter paper after drying.

84 
 

In order to examine the stacking condition of the GNP particles in the paper-like structure,
FESEM images of the cross sections of GNP papers are shown for both the as-made samples and
the annealed and cold-pressed samples shown in Figure 4-3.

Figure 4-3: (a) SEM micrographs of cross section of the as-made GNP-15 paper (scale bar: 10
µm), (b) annealed and cold pressed GNP-15 paper (scale bar: 10 µm), (c) as-made GNP-1 paper
(scale bar: 5 µm), (d) annealed and cold pressed GNP-1 paper (scale bar: 1 µm).

85 
 

Figure 4-3(cont'd)

86 
 

Figure 4-3 (cont'd)

 

It is noted in Figure 4-3a that the bottom side of the paper (the side attached to the filter
paper) has a better layered structure than the top part due to the changing flow conditions of the
water. Figure 4-3(a,c) shows the porosity of the as-made paper, the layering is more obvious in
GNP-15 due to the larger platelet size. The thermal annealing process facilitates the
decomposition of polyelectrolyte adsorbed on the surface of GNP and the same process will lead
to a slight disruption of the layered structure inside the paper as a result of decomposition of PEI
from the interior of the paper. It is necessary to cold-press the sample to restore the structure and
alignment of the paper, eliminate porosity, and increase contact areas among GNP shown in
Figure 4-3(b,d).

87 
 

4.4.2 Thermogravimetric analysis of GNP paper
The TGA curves for GNP paper before and after annealing are shown in Figure 4-4. It is
obvious that there is a slope change at around 300ºC which marks the decomposition of PEI. At
around 600 ºC when GNP is starting to oxidize, the difference in the weight retained before and
after annealing is 1.5wt% and 2.5wt% for GNP-15 and GNP-1 papers respectively which also
corresponds to the amount of PEI coated on the GNP surface. In addition, the thermal annealing
process effectively “cleans” the GNP by removing the water molecules adsorbed on the surface
and the edges as is seen by the initial weight loss below 250 ºC for the as-made samples.

Figure 4-4: (a): TGA weight loss curves of GNP-15 paper before and after annealing, (b) TGA
weight loss curves of GNP-1 paper before and after annealing.

88 
 

Figure 4-4 (cont'd)

 
 

4.4.3 Thermal diffusivity and thermal conductivity of GNP paper under various treatments
A 1’’ disk was punched out from the 36 mm diameter GNP paper for both the throughplane and in-plane thermal diffusivity measurement. Results for through-plane, in-plane thermal
diffusivity and conductivity were compared for GNP-1 and GNP-15 paper for the as-made, the
annealed, and the annealed and cold-pressed samples in Figure 4-5(a-d). The specific heat
capacity of GNP particles in all the experiments was measured to be 710 J/kg K. Table 4-1
shows the densities of different samples.

89 
 

Table 4-1: Density variation of GNP paper under different conditions
Density (ρ: g/cm3)

as-made

annealed

annealed &cold pressed

GNP-15 paper

0.73+0.08

0.73+0.08

1.15+0.2

GNP-1 paper

0.71+0.03

0.71+0.05

1.3+0.2

For the through-plane measurement, the 1’’ GNP paper was placed in the sample holder
covered by a ring shaped mask to block any leakage of light from sample edges. The entire
bottom surface of the sample was irradiated and the heat propagated to the top surface was
collected by the IR camera. The Cowan + pulse model provided by the software package
(Netzsch Proteus) was used to calculate the thermal diffusivity. It is believed that the existence of
thermally insulating polymers (PEI) either on the GNP surface or at the interfaces hinders
phonon transport, increasing the thermal interface resistance which is believed to play a major
role in nanoscale heat transfer. Thus, the heat treatment above the decomposition temperature of
polymer would remove most of the PEI, improving the “thermal contact” among platelets and
evidenced by a 20% improvement in thermal diffusivity and hence the conductivity given the
fact that the density of the sample remains the same before and after annealing for GNP-15 and
GNP-1 paper. However, after the cold pressing process, the density of the sample increases
because most of the pores created during filtration and annealing were eliminated, but the
compressive stress effectively aligns the particles within a plane perpendicular to the force
direction as shown in the SEM images. In addition, the nanointerfaces among individual platelet
scatter phonons and further impede the heat transfer from particle to particle. The dual effects of
both strong alignment and multiple nanointerfaces results in a through plane thermal
conductivity less than 2 W/mK. Also, there is noticeable difference in the heat conducting
90 
 

capability of GNP-15 and GNP-1 papers. The mere fact that the number of nanointerfaces in
GNP-1 paper are a few hundred times greater in number than for the larger GNP-15 illustrates
that the boundary phonon scattering will reduce thermal conductivity significantly.
The in-plane diffusivity and conductivity were measured using a method that employs a
mask to collimate the laser beam and heat an approximately 5 mm diameter spot at the bottom of
a 1’’ disk sample. Heat will flow through the thickness of the sample and spread out in the plane.
Another mask with embedded annulus of 9 to 12 mm in diameter was placed on the sample to
allow heat radiation to be collected by the IR detector. In this case, the large anisotropy of the
graphitic material (GNP) in heat conduction would affect the in-plane measurement, i.e. the time
for the heat to reach the slit in the mask will depend on both the radial diffusion time and the
axial diffusion time. Thus, the through-plane diffusivity of the sample was measured first and
input to the in-plane analysis and used in the axial part of the heat flow calculation. Then an
iterative scheme is used to separate the axial and radial diffusivities. The model that considers all
this is “anisotropic + heat loss”. For both GNP-15 and GNP-1 samples, a modest 10%
improvement in thermal diffusivity and conductivity were observed due to thermal annealing.
For the cold-pressed samples, a dramatic enhancement in thermal conductivity from 71 W/m·K
to 207 W/m·K for GNP-15 paper implies that the strong alignment of particles and solid “thermal
contact” induced by large compressive stress improves thermal conductivity and reduces thermal
contact resistance. The same trend applies to the GNP-1 sample, but the GNP size reduction
from 15 μm to 1 μm introduces multiple interfaces that reduce phonon transport significantly. As
a result, the thermal conductivity is only 10 percent of the GNP-15 for all the samples measured
at three different conditions. Despite the promising results reported here, it is still worth noting

91 
 

that there are some commercial exfoliated graphite products with thermal conductivity as high as
1500W/m·K.

Figure 4-5: (a) Through-plane thermal diffusivity of GNP-1 and GNP-15 papers. (b) Throughplane conductivity of GNP-1 and GNP-15 papers. (c) In-plane diffusivity of GNP-1 and GNP-15
papers. (d) In-plane conductivity of GNP-1 and GNP-15 papers. (note: left column represents
GNP1, right column represents GNP15)

                                                            


Graftech International (http://www.graftech.com/MARKETS/Lighting-Thermal-

Management.aspx)
92 
 

Figure 4-5(cont'd)

93 
 

Figure 4-5(cont'd)

 
 

4.5 Conclusions:
GNP stabilized by PEI were successfully dispersed in water and a paper-like structure
was prepared by flow directed self assembly. Thermal diffusivity and conductivity of both GNP15 and GNP-1 papers were measured at three different conditions. Annealing that helps
decompose the polymer enhances thermal conductivity possibly due to the reduction of thermal
interface resistance. However, the dominant factor in nanoscale heat transfer is the density of
nanointerfaces within the paper which gives rise to phonon boundary scattering. Cold pressing
restores the strong alignment of individual particle, improves the thermal contact by eliminating
the pores in the paper. After annealing and pressing, the GNP papers are very anisotropic in heat

94 
 

conduction, which is very promising for applications that require moving heat in a particular
direction (e.g. 2D heat spreader) or to reduce the maximum temperature at a “hotspot” by
spreading heat laterally for protective purposes.

95 
 

REFERENCES
 

 

96 
 

References
 

1. Dowell, M. B.; Howard, R.A.; Tensile and compression properties of flexible graphite foils.
Carbon 1986, 24, 311-323.
2. Leng, Y.; Gu, J.; Cao, W.; Zhang, T.Y.; Influences of density and flake size on the mechanical
properties of flexible graphite. Carbon 1998, 36, 875-881.
3. Reynolds, R. A.; Greinke, R. A.; Influence of expansion volume of intercalated graphite on
tensile properties of flexible graphite. Carbon 2001, 39, 479-481.
4. Chung, D. Flexible graphite for gasketing, adsorption, electromagnetic interference shielding,
vibration damping, electrochemical applications, and stress sensing. J Mater Eng Perform
2000, 9, 161-163.
5. Yazici, M. S.; Krassowski, D.; Prakash, J. Flexible graphite as battery anode and current
collector. J. Power Sources 2005, 141, 171-176.
6. Baughman, R. H.; Cui, C.; Zakhidov, A. A.; Iqbal, Z.; Barisci, J. N.; Spinks, G. M et al.
Carbon nanotube actuators. Science 1999, 284, 1340-1344.
7. Hennrich, F.; Lebedkin, S.; Malik, S.; Tracy, J.; Barczewski, M.; Rosner, H et al. Preparation,
characterization and applications of free-standing single walled carbon nanotube thin films.
Phys Chem Chem Phys 2002, 4, 2273-2277.
8. Coleman, J. N.; Blau, W. J.; Dalton, A. B.; Munoz, E.; Collins, S.; Kim, B. G et al. Improving
the mechanical properties of single-walled carbon nanotubes sheets by intercalation of
polymeric adhesives. Appl. Phys. Lett 2003, 82, 1682-1684.
9. Dmitriy, A. D.; Stankovich, S.; Zimney, E. J.; Piner, R. D.; Dommett, G. H. B.; Evmenenko,
G et al. Preparation and characterization of graphene oxide paper. Nature 2007, 448, 457-460.
10. Park, S.; An, J.; Piner, R. D.; Jung, I.; Yang, D.; Velamakanni, A. Aqueous suspension and
characterization of chemically modified graphene sheets. Chem. Mater 2008, 20, 6592-6594.
11. Hone, J.; Whitney, M.; Zettl, A. Thermal conductivity of single-walled carbon nanotubes.
Synth Met 1999, 103, 2498-2499.
12. Li, Q.; Liu, C.; Wang, X.; Fan, S.; Measuring the thermal conductivity of individual carbon
nanotubes by the Raman shift method. Nanotechnology 2009, 20, 145702.
13. Balandin, A. A.; Ghosh, S.; Bao, W.; Calizo, I.; Teweldebrhan, D.; Miao, F et al. Superior
thermal conductivity of single-layer graphene. Nano Lett 2008, 8, 902-907.
14. Hiemenz, P. C.; Rajagopalan, R. Principles of Colloid and Surface Chemistry, 3rd edition,
revised and expanded, Marcel Dekker, New York: 1997: 575-614.
97 
 

15. Yu, D.; Dai, L.; Self-assembled graphene/carbon nanotube hybrid films for supercapacitors. J.
Phys. Chem. Lett 2010, 1, 467-470.
16. Padgett, C. W.; Brenner, B. W. Influence of chemisorption on the thermal conductivity of
single-wall carbon nanotubes. Nano Lett 2004, 4, 1051-1053.
17. Zhang, J.; Xu, Q.; Ye, F.; Lin, Q.; Jiang, D.; Iwasa, M. Effect of citric acid on the adsorption
behavior of polyethyleneimine (PEI) and the relevant stability of SiC slurries. Colloid
Surface A 2006, 276, 168-175.

98 
 

Chapter 5 Electron and Phonon Transport in Gold Nanoparticle
Decorated Graphene Nanoplatelet Nanostructured Paper
5.1 Abstract:
Monodispersed Au nanoparticles were synthesized on the surface of GNP in the presence of
polyethyleneimine (PEI) with microwave assisted heating. A highly structured layered Au/GNP
“paper” with good flexibility and mechanical robustness was prepared by vacuum assisted self
assembly. The thermal and electrical conductivity of the hybrid paper with and without the Au
nanoparticles were investigated after different experimental processing conditions including
thermal annealing and cold compaction. Annealing effectively removed the adsorbed PEI
molecules and improved thermal contact between Au/GNP particles while cold compaction
reduced porosity and induced stronger alignment of the Au/GNP within the hybrid paper. Both
approaches led to improvement in electrical and thermal conductivity. It was also found that
adjacent GNP particles were electrically connected by metal nanoparticles but thermally
disconnected. It is believed that phonons are scattered at the Au/GNP interfaces while electrons
tunnel across this interface, resulting in a separation of electron and phonon transport within this
hybrid paper. This chapter is based on a journal publication: Xiang, J., Drzal, L. T. (2011)
Electron and phonon transport in Au nanoparticle decorated graphene nanoplatelet
nanostructured paper. ACS applied materials & interfaces, 3(4), 1325-1332.

5.2 Introduction:
Chapter 4 discussed in detail the electrical and thermal transport of a highly ordered GNP
paper prepared by vacuum assisted self assembly. It was found that both annealing and cold
compaction is beneficial in improving the electron and phonon transport in the paper as a result
99 
 

of reduced contact resistance. This chapter continues to investigate this paper-like structure but
with an aim to separate the carrier transport within the GNP paper for thermoelectric application
which requires a decoupled electron and phonon transport. In particular, metal nanoparticles such
as gold was deposited on the surface of GNP and a hybrid Au/GNP paper was prepared.
Decoration of GNP with metal nanoparticles changes to some extent the nature of GNP-GNP
interfaces from a surface contact to a point to point contact, the question this chapter attempts to
answer is how these nanoparticles would change the electron and phonon transport within this
hybrid paper.
The graphene-metal hybrid systems have already been discussed in the literature to have
various engineering applications such as fuel cell catalysis [1-5], electrochemical sensing [6, 7]
and electrochemical energy storage [8,9]. Methods to synthesize metal nanoparticles on the basal
plane surface of graphene have been explored quite extensively. Among those, solution and
supercritical liquid approaches have the advantage of covering the entire surface area of
graphene nanosheets with nanoparticles. For example, one common method is to chemically
functionalize the graphitic surface in order to introduce anchoring sites for the metal precursors.
The metal precursor is then reduced in the presence of a reducing agent to produce nanoparticles
that covalently attaches to the basal plane of the graphene [10-12]. The disadvantage of this
approach is the disruption of the sp2 bonded carbon atoms in the basal plane which leads to
reduced transport properties of graphene because of additional scattering sites [13]. Another
widely adopted technique involves non-covalent encapsulation of the graphitic surface with a
surfactant or a polymer and subsequent growth of nanoparticles on the surfactant in the presence
of a reducing agent [14, 15]. This technique induces minimum chemical perturbation of the basal
planes, preserving the conjugated system for undisrupted carrier transport. Other techniques of

100 
 

metal nanoparticle decoration on graphitic nanostructure include electro-deposition [16],
evaporation [17], and solventless bulk synthesis [18]. While these methods have some processing
advantages over solution-phase techniques, they are usually quite expensive and energy
intensive. In order to synthesize nanoparticles using the solution-phase approach with a narrow
size distribution, several factors need to be controlled carefully which include the concentration
of metal salt and reducing agent [19,20], the presence of a protecting agent [21], the reaction
time [22] and reaction temperature [23].
Here, monodispersed gold nanoparticles were synthesized on individual GNP in the
presence of polyethyleneimine with microwave assisted heating. Previous publications have
reported on the deposition of Au nanoparticles on graphene nanosheets through simultaneous
reduction of graphene oxide and Au(III) by microwave heating and sonolysis [24,25]. This
chapter demonstrates a fast, one-pot, aqueous synthesis of Au nanoparticles directly on pristine
graphene nanosheets. Polyethyleneimine, a hydrophilic polymer with primary, secondary and
tertiary amino groups and an overall positive charge in the neutral aqueous solution, adsorbs on
the highly hydrophobic GNP surface and stabilizes the GNP particles in water. The adsorption of
PEI on GNP effectively induces both electrostatic and steric stabilization. In addition, PEI
contains one of the highest densities of amino groups among all polymers, donating electrons
that help reduce metal ions [26-28]. Gold nanoparticles were chosen here because of their strong
resistance to oxidation and excellent thermal and electrical properties. Gold nanoparticle
decorated GNP was prepared into a paper form by vacuum assisted self-assembly. The electrical
and thermal conductivity of the Au/GNP paper was characterized under different conditions such
as annealing and compaction similar to Chapter 4. The impact of gold nanoparticles on both
electron and phonon transport in this hybrid paper is discussed.

101 
 

5.3 Experimental:
Preparation of exfoliated graphene nanoplatelets (GNP)
GNPs were produced from a commercial graphite intercalated compound (Asbury Mills,
NJ). Thermal exfoliation was carried out in a microwave oven followed by mechanical size
reduction with a combination of ultra-sonication and ball milling to produce nanoplatelets with
the desired aspect ratios. GNP-15 particles are used throughout the experiment and are simply
referred as GNP.

Decoration of GNP with Au nanoparticles
Polyethyleneimine (PEI) was purchased from Sigma Aldrich. (Branched, Mw=25,000).
The GNP particles were dispersed in a PEI/water solution containing 0.3wt% and 0.6wt% PEI.
The suspensions were bath sonicated for 1 h and kept stirred overnight to ensure sufficient
interaction between PEI and GNP. Tetrachloroauric acid (HAuCl4·3H2O) purchased from Sigma
Aldrich (> 99.9% trace metal basis) was then added into the suspension and mixed thoroughly.
The suspension was irradiated with microwave radiation for a short period of time and then
centrifuged at 6000 rpm and washed with copious amount of water to produce Au nanoparticle
decorated GNP. The resulting Au/GNP product was dried under vacuum at 70 ̊C overnight.

Preparations of Au/GNP hybrid paper (as made, annealed, anneal and cold compacted):
The Au/GNP hybrid paper was prepared by vacuum filtration with an ANODISC 47 mm
filter membrane. (0.2μm, Whatman) The hybrid paper was then washed with water and dried at
60 ̊C under vacuum overnight before being removed from the filter. These were identified as
“as-made” samples. The paper identified as “annealed” sample was placed in a furnace and

102 
 

annealed at 340 ̊C in air for 2 h to decompose the PEI. The “annealed and cold compacted” paper
was mechanically compressed in a hydraulic press at 100 psi at room temperature for 1 h. For
comparison, a neat GNP paper without Au nanoparticles was prepared in the same way.

Flash lamp thermal diffusivity measurement by Nanoflash 447 (Netzsch Instruments)
For through-plane thermal diffusivity measurement, the 1 inch disk sample was placed in
the sample holder covered by a ring shaped mask to block any leakage of light from sample
edges. The entire bottom surface of the sample was irradiated and the heat propagating to the top
surface was collected by the IR camera. The “cowan + pulse” model provided by the software
package (Netzsch Proteous) was used to calculate thermal diffusivity. The in-plane diffusivity
was measured using a method that employs a mask that collimates the laser beam at a 5 mm
diameter spot at the bottom of the 1 inch disk sample. Another mask with embedded annulus of 9
to 12 mm was placed on the sample allowing heat propagating radially to be collected. The “inplane, anisotropic + heat loss” model incorporated in the software was used for subsequent
diffusivity calculation.

Electrical conductivity characterizations of hybrid papers:
In-plane electrical conductivity was measured with a Keithley 2400 SourceMeter in a
four point configuration with a probe spacing of 1 cm on thin rectangular films. Electric current
was scanned from 20 mA to 100 mA with 20 mA intervals. Voltage drops on the sample were
recorded and plotted against the current to obtain the resistance. The through-plane conductivity
of thin film samples was measured by using 2 instead of 4 leads. Electrical resistance of the
sample was obtained by calculating the slope of the voltage vs. current data. Taking the

103 
 

geometry of the samples into account, the in-plane electrical conductivity and through-plane
resistivity (see supporting information) were calculated.

5.4 Results and discussions:
5.4.1 Morphology
Dispersion of GNP particles in water proves difficult and results in an immediate floating
and aggregation of the particles at the air-water interface because of the highly hydrophobic
nature of the basal plane. Polyethyleneimine, a cationic polyelectrolyte, adsorbs on the GNP
surface due to a thermodynamic driving force to reduce the hydrophobic interface between GNP
and the polar solvent and minimize the interfacial energy. At pH <10, the positively charged
polymer chain also induced electrostatic repulsion that contributed to good dispersion of the
nanoplatelets. As the tetra-chloroauric acid dissolved in water was mixed with the GNP
suspensions, the AuCl4- ions complex with the positive functional groups of PEI adsorbed on the
GNP surface (chloride ligands were replaced by amine groups of PEI). The adsorbed PEI chains,
in this case, serve as templates for subsequent Au nanoparticle growth on the surface of GNP.
The addition of H+ ions also lowered the pH of the solution to around 6, inducing more positive
charges on the chain of PEI and more electrostatic attraction with AuCl4- [29]. The reduction of
Au3+ to Au0 was then carried out in a microwave oven where the heating rate was fast enough to
cause a rapid temperature increase in the solution due to the high polarity of water molecules,
creating a high concentration of radicals which facilitated the electron transfer from the radicals
of PEI to the metal precursor. As soon as the solution was supersaturated with metal atoms, Au
atoms formed nuclei. The critical size of the nucleus as well as nucleation activation energy
control the nucleation rate and depend on the surface tension of the nuclei-solvent interface when

104 
 

the radius of nuclei is small [30, 31]. Therefore, a higher surface tension corresponds to a larger
critical nucleus size and a lower nucleation rate, which should be avoided in synthesizing
monodispersed small nanoparticles. Upon formation of the nucleus, polyethyleneimine, with a
lower surface tension than Au atoms, adsorbed on the surface of the metal atoms reducing the
nucleation energy barrier and critical nucleus size. As a result, more primary particles with low
surface energy were formed leading to generation of smaller secondary particles whose growth
and coarsening were also affected by surface tension [32]. Figure 5-1 represents the typical
procedures and the interactions between the metal ions and the active sites on GNP particles and
subsequent formation of nanoparticles on GNP.

Figure 5-1: Schematics of experimental procedures and the interactions between the metal ions
and the active sites on PEI coated GNP and formation of Au nanoparticles

From SEM micrographs shown in Figure 5-2, it was found that the size and loading of
the Au nanoparticles synthesized by this technique correlated with the concentration of PEI in
the solution. The size of the nanoparticles was reduced while the loading increased with higher
105 
 

concentration of PEI in the solution. With a higher concentration of PEI in the solution, the
number of nuclei with a lower surface energy increased which leads to formation of more
primary and secondary particles. In addition, more PEI chains would adsorb on to the surface of
GNP acting as templates for Au nanoparticle growth. There is no obvious agglomeration of Au
nanoparticles on the surface of GNP, which is also believed to be the result of PEI encapsulation
that stabilizes the nanoparticles in the solution electrostatically. Figure 5.3 shows the size and
distribution of Au nanoparticles on the GNP surface at 0.3wt% and 0.6wt% PEI concentrations.
This is obtained by analyzing representative SEM images of magnifications at 80k or higher in
Image-Pro, a commercially available software. (See Supporting Information). It is clear that the
whole spectrum shifts to a smaller size range while the number of nanoparticles increased for
samples prepared in 0.6wt% PEI. The size of the Au nanoparticles prepared at 0.3wt% PEI is
10.8+2.4 nm while they are reduced in size to 8.3+2.1 nm at 0.6wt% PEI.

106 
 

Figure 5-2 (a): Low magnification SEM image of Au nanoparticle decorated GNP prepared at
0.3wt% PEI concentration (scale bar: 1 µm). b: High magnification SEM image of Au
nanoparticle decorated GNP prepared at 0.3wt% PEI (scale bar: 100 nm) c: Low magnification
SEM image of Au nanoparticle decorated GNP prepared at 0.6wt% PEI concentration (scale bar:
1 µm). d: High magnification SEM image of Au nanoparticle decorated GNP prepared at 0.6wt%
PEI concentration (scale bar: 100nm)
107 
 

Figure 5-2(cont'd)

 

108 
 

 
 

 

Figure 5-3: (a) Size and distribution of Au nanoparticles on GNP prepared at 0.3wt% PEI
concentration. (b) Size and distribution of Au nanoparticles on GNP prepared at 0.6wt% PEI
concentration.
109 
 

 

Figure 5-4: Schematics of the experiment showing the Au nanoparticle decorated GNP dispersed
in water for subsequent vacuum assisted self assembly.
 

Figure 5-4 shows the vacuum assisted self assembly process to prepare Au/GNP hybrid
paper. To create a more ordered multilayer structure within the paper, it is important to do multifiltration (5 ml at a time) instead of filtration all at once. After the filtration process, it was found
that the bottom few layers in contact with the filter membrane is more ordered than the top layers
as a result of a non-uniform pressure difference in the paper during filtration as thickness builds
up. Therefore, 30 – 40 ml of Au/GNP suspension at 1 mg/ml concentration was filtered to
achieve a better aligned structure within the hybrid paper. Figure 5-5 shows the photos of the asmade papers after being removed from the filter membrane. The as-made paper is very flexible
and shows certain mechanical robustness under bending.

110 
 

Figure 5-5: Photos of the as-made GNP paper showing flexibility and mechanical robustness
under bending and the Au/GNP hybrid sample used for thermal diffusivity measurement (1 inch
diameter)
 
 

 

Figure 5-6: (a) Representative cross sectional SEM image of “as-made” GNP paper (scale bar: 5
µm). (b) as-made Au/GNP hybrid paper prepared at 0.3wt% PEI (scale bar: 5 µm).
 
 
 

111 
 

Figure 5-6(cont'd)

 

The cross sections of both as-made GNP and Au/GNP papers for SEM observation were
prepared by epoxy embedding shown in Figure 5-6. The neat as-made GNP paper shows a highly
aligned structure while preserving certain porosity introduced during filtration. However,
porosity is higher in the Au/GNP hybrid paper. The observed difference in morphology is
attributed to the presence of Au nanoparticles on GNP. The Au nanoparticles act as spacers,
creating more gaps between individual nanoplatelets when they were assembled into a paper.
Due to the more open structure of the hybrid ‘paper’, epoxy can easily penetrate into the gaps as
shown in Figure 5-6b. However, a highly ordered and highly condensed structure was formed
after annealing and cold compaction. The increase in porosity for Au/GNP hybrid sample can
also be supported by a reduction in the apparent density of the paper which is shown in Table
5.1.

112 
 

Table 5-1: Average densities of GNP and Au/GNP hybrid papers measured as-made, annealed,
annealed & cold compacted (standard deviation < 0.05g/cm3).
Density (ρ: g/cm3)

As-made

Annealed

GNP paper
Au/GNP (0.3wt%
PEI)
Au/GNP (0.6wt%
PEI)

0.62

0.62

Annealed & cold
compacted
1.5

0.58

0.57

1.5

0.58

0.58

1.5

5.4.2 Thermal diffusivity and thermal conductivity of Au/GNP hybrid paper
Thermal diffusivity of the samples prepared at different experimental conditions (asmade, annealed, annealed and cold compacted) were measured. Given the density of the sample,
and the specific heat, which was measured by differential scanning calorimetry, thermal
conductivity of the sample can be obtained: =α ρ Cp where  is thermal conductivity with a
unit of W/moK, α is thermal diffusivity with a unit of mm2/s, and Cp is specific heat with a unit
of J/goK.
As shown in Figure 5-7a, the through-plane thermal diffusivity of all the samples
followed the same trend with a slight increase after thermal annealing and a reduction after cold
compaction. Previous study (Chapter 4) showed that the enhancement brought by annealing was
due to the thermal decomposition of polyethyleneimine at 300 ̊C [33 and Supporting
Information]. It is believed PEI adsorbed either on the GNP surface or on the Au nanoparticles
scatter the phonons in this paper-like structure, contributing to a larger thermal interface
resistance which is a major impediment to phonon transport. A thermal annealing treatment
removed most of the PEI in the paper, presumably reducing the interfacial resistance as
suggested by a 20% improvement in through-plane diffusivity. However, upon annealing and
cold compaction, thermal diffusivity decreased for all the samples. Compaction of the sample

113 
 

effectively eliminated large pores within the paper created during filtration and the nanoplatelets
were more aligned due to the compressive stress. The better orientation of the nanoplatelets
reduced the through-plane phonon transport due to the intrinsically low thermal conductivity of
GNP with multiple layers of graphene held together with weak Van der Waals forces and the
high probability of interface scattering between individual nanoplatelets. It is also found that the
through-plane diffusivity of the Au/GNP hybrid paper is lower than the neat GNP paper. Gold
nanoparticles on GNP are likely to interfere with GNP from making contact with each other and
the paper becomes more porous. As a result, thermal diffusivity reduced with an increasing
concentration of Au nanoparticles on the GNP surfaces.
Cold compaction densifies the “loose” porous nanostructure. Despite the fact that macro
sized pores collapse upon compaction in neat GNP paper, there are still numerous slit pores
inside it that would interfere with the transport of electrons and phonons. With the hybrid paper,
Au nanoparticles on the surface of GNP now have a much higher probability of contact with
each other, serving as bridging agents or mini columns which might facilitate propagation of
phonons within the paper structure. Nonetheless, the experimental results for “annealed and cold
compacted” samples indicated a reduction in the through-plane diffusivity for the Au/GNP
hybrid paper. In this case, a phonon is forced to go through a Au nanoparticle to go from one
GNP to another, which slows down the heat transfer, thus the thermal diffusivity. The same trend
can be observed for thermal conductivity of the samples. In particular, the cross sectional area of
heat conduction is reduced due to the presence of Au nanoparticles. As a result, the throughplane thermal conductivity of the Au/GNP hybrid sample is lower than the neat GNP paper after
annealing and cold compaction. The result is shown in Figure 5-7b. The reported through-plane

114 
 

thermal conductivity for the highly oriented pyrolytic graphite measured by pump-probe
thermoreflectance technique is 6.1 W/moK [34].
The in-plane thermal diffusivity and conductivity are shown in Figure 5-7(c,d). Thermal
annealing caused decomposition and removal of PEI from within the sample, improving the
“thermal contact area” by reducing interface resistance. A moderate increase of around 10% was
observed for samples measured after annealing. Cold-compaction of the samples restores the
orientation of the nanoplatelets, making the paper more anisotropic in heat conduction. It is
worth noting that both thermal diffusivity and conductivity of the Au/GNP hybrid samples are
lower than the neat GNP paper, which is in accord with the trend observed from through-plane
measurements. It is believed that those Au nanoparticles are more likely to serve as phonon
scattering centers than mini heat conduction channels within the hybrid samples. Recent work by
Wang et al used a thermal bridge configuration to measure the in-plane thermal conductivity of
supported few layer graphene nanosheets and concluded that the conductivity is 1250 W/moK at
room temperature is comparable to bulk graphite [35]. The value reported here is lower because
of the strong interface scattering which is absent in either individual graphene nanosheets or bulk
graphite.

115 
 

Figure 5-7: (a) Through-plane thermal diffusivity of GNP and Au/GNP paper. b: Through-plane
thermal conductivity of GNP and Au/GNP paper. c: In-plane thermal diffusivity of GNP and
Au/GNP paper. d: In-plane thermal conductivity of GNP and Au/GNP paper. Note: left column
represents GNP15, middle column represents GNP15/Au(0.3wt% PEI), right column represents
GNP15/Au(0.6wt% PEI)

116 
 

Figure 5-7(cont'd)

117 
 

Figure 5-7(cont'd)

118 
 

Figure 5-7(cont'd)

5.4.3 Electrical properties of Au/GNP hybrid paper
The self assembly of GNP with and without Au nanoparticles did have an impact on the
thermal conductivity of the samples. In particular, the gold nanoparticles that were expected to
provide additional heat channels through the plane of the paper structure to improve the heat
conduction turned out to be ineffective. To investigate if the same phenomenon observed in
phonon transport also applies to electrons, both the through-plane and in-plane electrical
conductivity of these papers were measured and reported in Figure 5-8.
From Figure 5-8a, the in-plane electrical conductivity of the papers increased with
annealing and cold compaction. The improvement can be explained by the effective
119 
 

decomposition of electrically insulating PEI from the samples through thermal annealing and
significant reduction of pore volumes through cold compaction, both contributing to decreasing
electron tunneling resistance. In particular, the GNP paper decorated with Au nanoparticles of
different size and distribution density showed higher conductivity than the neat GNP paper.
Hybrid GNP papers with individual nanoplatelets decorated with smaller and higher coverage of
Au nanoparticles exhibited higher in-plane electrical conductivity. Upon cold compaction, the
collapse of macro pores in the sample brings more Au nanoparticles into immediate contact,
effectively bridging previously isolated nanoplatelets. It is worth noting that the in-plane
conductivity reaches a value as high as 1500 S/cm for Au/GNP (0.6wt% PEI) hybrid paper,
which is almost 70% higher than the neat GNP paper (880 S/cm) at the same condition. It is
worth noting that the in-plane electrical conductivity is measured to be 2×104 S/cm for bulk
graphite [36], one order of magnitude higher than the value reported here for Au/GNP hybrid
sample.
The through-plane electrical resistivity of the papers was measured by a two-point
method (see Supporting Information) on the samples that have been annealed and cold
compacted. The result is shown in Figure 5-8b. The results from as-made and annealed samples
are not reported because the measured electrical resistance fluctuates a large amount possibly
due to the open porous structure inside the sample. The measurements on the annealed and cold
compacted samples were stable enough thanks to the dense and compact structure of the paper.
The Au nanoparticles effectively bridged adjacent GNPs, creating mini channels for the cross
plane electron transport as evidenced by the drop in the through-plane resistivity. It is thus
evident that although the through-plane conductivity is small due to a greater anisotropic
electronic conduction induced by cold compaction, the Au nanoparticles did facilitate the

120 
 

electron transport within the paper structure, making individual nanoplatelets more electrically
connected. As a comparison, the through-plane resistivity for bulk graphite is 0.1~1 ohm cm, a
few orders of magnitude lower than the highly ordered layer structure reported here [36].

Figure 5-8: (a) In-plane electrical conductivity of anneal & cold compacted GNP and Au/GNP
paper. (b) Through-plane electrical resistivity of anneal & cold compacted GNP and Au/GNP
paper.

121 
 

Figure 5-8(cont'd)

The findings reported here indicate that electron and phonon transport are affected
differently by the presence of Au nanoparticles within the hybrid sample. Phonons travel at
different frequencies in two materials of different densities (e.g. Au and graphite). Nonequilibrium molecular dynamic studies show that when a large mismatch in the phonon
vibrational density of states (VDOS) exists in two distinctive mediums, direct coupling and
propagation of phonons across the interface will be reduced. As a result, the thermal energy can
only be transferred across the interface indirectly through anharmonic scattering which enables
phonon-phonon interactions and gives rise to a high thermal interface resistance [37]. Despite the
fact that Au nanoparticles could bridge the small gaps between individual nanoplatelets, the
anharmonicity in VDOS in gold and carbon result in strong interface scattering that reduces the
122 
 

efficiency of heat transfer. However, the improvement observed in electrical conductivity
suggests that the Au nanoparticles allow more electrons to tunnel through, making the hybrid
paper structure more electrically connected as compared with the neat GNP paper.
A few literature citations point out that few-layer graphene shows improved electrical
conductivity when Au precursor were spontaneously reduced to Au nanoparticles driven by the
redox potential difference between the Au3+ and the graphene nanosheets [38-40]. As a result,
the few-layer graphene was p-type doped with electrons being transferred away from it to the
metal precursor. The mere increase of the carrier concentration (i.e. holes) contributes to an
increase in the electrical conductivity. However, the microwave assisted synthesis of
nanoparticles in PEI matrix reported here did not involve transfer of electrons from individual
graphite nanoplatelets to Au3+ ions. The fact that electrons mainly come from polyethyleneimine
instead of GNP eliminates the possibility that the individual nanoplatelets would be p-type
doped. Therefore, the enhancement in electrical conductivity observed here originates from the
reduction in the interface resistance instead of doping, i.e. more electrons can find routes with
least resistive pathways within the hybrid paper structure because of the highly electrically
conductive Au nanoparticles.

5.5 Conclusion:
Gold nanoparticles of a uniform size distribution were successfully synthesized on
exfoliated graphene nanoplatelets by microwave assisted heating in a polyethyleneimine/water
solution. It is believed that fast super-saturation of metal atoms with low surface tension is
desired for producing monodispersed small nanoparticles. A self-standing, mechanically robust
Au/GNP nanostructured multilayered ‘paper’ was prepared by vacuum assisted self assembly.

123 
 

Thermal annealing to decompose and remove PEI improves both electrical and thermal
conductivities. Paper porosity was significantly reduced by cold compaction which produces
closer contact between individual nanoplatelets. Au nanoparticles on the surface of GNP reduce
phonon transport by introducing more scattering sites at the interface due to a vast difference in
the density of states for phonons in gold and GNP and by reducing the total cross sectional area
for heat conduction. However, the Au nanoparticles act as electron super highways that make
nanoplatelets more electrically connected. The decoupled transport phenomenon observed here
provided valuable insights into designing GNP paper based thermoelectric materials.

Acknowledgement
The authors wish to acknowledge the Army Research Office for financial support of this
research, Grant # W911NF-09-1-0451, Dr. Robert Mantz, Technical Program Manager.
 
 
 
 
 
 
 

124 
 

Supporting information

TGA (decomposition of PEI)
TGA curve of the neat polyethyleneimine decomposed in air is shown in Figure 5-9. It is noted
that PEI begins a major decomposition event at around 300˚C. In addition, the TGA curves of the
as-made GNP paper, annealed GNP paper and as-made Au/GNP hybrid prepared at 0.6wt% PEI
are shown in Figure 5-10. PEI decomposition can be observed in both the as-made GNP and
Au/GNP hybrid samples. While for the annealed GNP sample, major decomposition event
occurs at the oxidation temperature of GNP itself, indicating that PEI is removed from the
sample. 

 

Figure 5-9: TGA curve of neat PEI in air
125 
 

 

Figure 5-10: TGA curves of annealed GNP paper, as-made GNP paper and as-made Au/GNP
paper in air.
 

Size distribution of Au nanoparticles on GNP by Image-Pro
The size distribution data shown in Figure 5-3 was obtained by analyzing the SEM images in
Image-Pro, which is a software commercially available. The general procedures are described
below: select a SEM image with a magnification of at least 80k, adjust the contrast until the Au
nanoparticles stand out from the background. Let the software pick out the Au nanoparticles
according to the pre-determined selection criterions (e.g. diameter > 4nm, 1<aspect ratio<1.5).
The average diameter of each nanoparticle on the analyzed SEM is shown and the histogram is
plotted according to this information. Figure 5-11 is a representative SEM that was analyzed by
the software. Figure 5-12 is the statistics of the nanoparticle distribution.

126 
 

 

Figure 5-11: Example SEM images analyzed by Image-Pro to determine the distribution of
nanoparticle size (scale bar 100nm)

127 
 

 

Figure 5-12: Particle size distribution obtained from Figure 5-11.
 

Two point through-plane electrical resistivity measurements of thin film (<100μm)
The reported through-plane electrical resistivity is measured by the two point probe method
because of the inherent geometric constraints of working with a thin film. However, the
configuration is different from the in-plane four point probe measurement where a tip probe is in
contact with the sample, giving rise to a very large contact resistance. In order to eliminate the
effect of contact resistance, strips of adhesive copper foils were used to cover the top and bottom
surfaces of the sample to spread the electrical current. Figure 5-13 shows cross sectional
schematics of the measurement:

128 
 

 

Figure 5-13: Schematic representation of through-plane resistivity measurement of thin film
sample

It is noted that the resistance measured include the resistances of the copper tapes, the contact
and the resistance of the sample. To account for the effect of copper and the contact, a gold foil
with similar geometry to the samples (100μm in thickness, 99.99%, Sigma Aldrich) was used to
estimate copper and contact resistance because the resistance of the gold foil in this case is
negligibly small. The measured resistance is 0.26 ohm which is around 10% of the measured
sample resistance. Therefore, the data reported is modified by subtracting 0.26 ohm from the
measured sample resistance to obtain the electrical resistivity.

129 
 

REFERENCES

130 
 

References
1. Xu, C.; Wang, X.; Zhu, J.; Graphene-metal particle nanocomposites. J. Phys. Chem. C.
2008, 112, 19841-19845.
2. Yoo, E.; Okata, T.; Akita, T.; Kohyama, M.; Nakamura, J.; Honma, I. Enhanced
electrocatalytic activity of Pt subnanoclusters on graphene nanosheets surface. Nano Lett. 2009,
9, 2255-2259.
3. Li, Y. M.; Tang, L. H.; Li, J. H.; Preparation and electrochemical performance for methanol
oxidation of Pt/graphene nanocomposites. Electrochem. Commun. 2009, 11, 846-849.
4. Serger, B.; Kamat, P. V.; Electrocatalytically active graphene-platinum nanocomposites.
Role of 2-D carbon support in PEM fuel cells. J. Phys. Chem. C. 2009, 113, 7990-7995.
5. Zhu, C.; Guo, S.; Zhai, Y.; Dong, S.; Layer-by-layer assembly for constructing a
graphene/platinum nanoparticle three-dimensional hybrid nanostructure using ionic liquid as a
linker. Langmuir. 2010, 26, 7614-7618.
6. Lu, J.; Do, I.; Drzal, L. T.; Worden, R. M.; Lee, I.; Nanometal-decorated exfoliated graphite
nanoplatelet based glucose biosensors with high sensitivity and fast response. ACS Nano. 2008,
2, 1825-1832.
7. Guo, S.; Wen, D.; Zhai, Y.; Dong, S.; Wang, E.; Platinum nanoparticle ensemble-ongraphene hybrid nanosheet: One-pot, rapid synthesis, and used as new electrode material for
electrochemical sensing. ACS Nano. 2010, 4, 3959-3968.
8. Si, Y.; Samulski, E.; Exfoliated graphene separated by platinum nanoparticles. Chem. Mater.
2008, 20, 6792-6797.
9. Li, Y.; Lv, X.; Lu, J.; Li, J.; Preparation of SnO2-Nanocrystal/Graphene-nanosheets
composites and their lithium storage ability. J. Phys. Chem. C. 2010, 114, 21770-21774.
10. Kamat, P. V.; Graphene-based nanoarchitectures, Anchoring semiconductor and metal
nanoparticles on a two-dimensional carbon support. J. Phys. Chem. Lett. 2010, 1, 520-527.
11. Muszynski, R.; Seger, B.; Kamat, P. V.; Decorating graphene sheets with gold
nanoparticles. J. Phys. Chem. C. 2008, 112, 5263-5266.
12. Goncalves, G.; Marques, P. A. A. P.; Granadeiro, C. M.; Nogueira, H. I. S.; Singh, M. K.;
Gracio, J. Surface modification of graphene nanosheets with gold nanoparticles: the role of
oxygen moieties at graphene surface on gold nucleation and growth. J. Chem. Mater. 2009, 21,
4796-4802.
13. Padgett, C. W.; Brenner, B. W.; Influence of chemisorption on the thermal conductivity of
single-walled carbon nanotubes. Nano Lett. 2004, 4, 1051-1053.
131 
 

14. Kongkanand, A.; Vinodgopal, K.; Kuwabata, S.; Kamat, P. V.; Highly dispersed Pt catalyst
on single walled carbon nanotubes and their role in methanol oxidation. J. Phys. Chem. B, 2006,
110, 16185-16188.
15. Ostojic, G. N.; Ireland, J. R.; Hersam, M. C.; Noncovalent functionalization of DNAwrapped single walled carbon nanotubes with platinum-based DNA cross-linkers. Langmuir,
2008, 24, 9784-9789.
16. Bayati, M.; Abad, J. M.; Nichols, R. J.; Schiffrin, D. J.; Substrate structural effects on the
synthesis and electrochemical properties of platinum nanoparticles on highly oriented pyrolytic
graphite. J. Phys. Chem. C. 2010, 114, 18439-18448.
17. Ren, G.; Xing, Y.; Deposition of metallic nanoparticles on carbon nanotubes via a fast
evaporation process. Nanotechnology. 2006, 17, 5596-5601.
18. Lin, Y.; Watson, K. A.; Fallbach, M. J.; Ghose, S.; Smith, J. G. Jr; Delozier, D. M et al.
Rapid, solventless, bulk preparation of metal nanoparticle-decorated carbon nanotubes. ACS
Nano, 2009, 3, 871-884.
19. Teranishi, T.; Miyake, M.; Size control of palladium nanoparticles and their crystal
structures. Chem. Mater. 1998, 10, 594-600.
20. Leff, D. V.; Thermodynamic control of gold nanocrystal size: experiment and theory. J.
Phys. Chem, 1995, 99, 7036-7041.
21. Do. I.; Metal decoration of exfoliated graphite nanoplatelets for fuel cell applications, Ph.D
dissertation, Michigan State University, East Lansing, Michigan, 2006.
22. Meguro, K.; Torizuka, M.; Esumi, K.; The preparation of organo colloidal precious metal
particles. Bull. Chem. Soc. Jpn, 1988, 61, 341-345.
23. Mohamed, M. B.; Wang, Z. L.; El-Sayed, M. A.; Temperature-dependent size-controlled
nucleation and growth of gold nanoclusters. J. Phys. Chem. A. 1999, 103, 10255-10259.
24. Jasuja, K.; Linn, J.; Melton, S.; Berry, V.; Microwave-reduced uncapped metal
nanoparticles on graphene: tuning catalytic, electrical, and Raman properties. J. Phys. Chem.
Lett. 2010, 1, 1853-1860.
25. Vinodgopal, K.; Neppolian, B.; Lightcap, I. V.; Grieser, F.; Ashokkumar, M.; Kamat, P. V.
Sonolytic design of graphene-Au nanocomposites. Simultaneous and sequential reduction of
graphene oxide and Au(III). J. Phys. Chem. Lett. 2010, 1, 1987-1993.
26. Bai, L.; Zhu, H.; Thrasher, J. S.; Street, S. C.; Synthesis and electrocatalytic activity of
photoreduced platinum nanoparticles in a poly(ethylenimine) matrix. ACS App. Mater.
Interfaces. 2009, 1, 2304-2311.

132 
 

27. Kuo, P. L.; Chen, W. F.; Huang, H. Y.; Chang, I. C.; Dai, S. A.; Stabilizing effect of
pseudo-dendritic polyethyleneimine on platinum nanoparticles supported on carbon. J. Phys.
Chem. B. 2006, 110, 3071-3077.
28. Sun, X.; Dong, S.; Wang, E.; One-step synthesis and characterization of polyelectrolyteprotected gold nanoparticles through a thermal process. Polymer. 2004, 45, 2181-2184.
29. Zhang, J.; Xu, Q.; Ye, F.; Lin, Q.; Jiang, D.; Iwasa, M.; Effect of citric acid on the
adsorption behavior of polyethyleneimine (PEI) and the relevant stability of SiC slurries.
Colloids Surf A. 2006, 276, 168-175.
30. Oxtoby, D. W.; Nucleation of first order phase transitions. Acc. Chem. Res. 1998, 31, 9197.
31. Furedi-Millhofer, H.; Spontaneous precipitation from electrolyte solutions. Pure Appl.
Chem, 1981, 53, 2041-2055.
32. Peng, X. G.; Wickham, J.; Alivisatos, A. P.; Kinetics of II-VI and III-V colloidal
semiconductor nanocrystal growth: “Focusing” of size distribution. J. Am. Chem. Soc. 1998,
120, 5343-5344.
33. Xiang, J.; Drzal, L. T.; Thermal conductivity of exfoliated graphite nanoplatelet paper
Carbon, 2011, 49, 773-778.
34. Schmidt, A. J.; Chen, X.; Chen, G.; Pulse accumulation, radial heat conduction, and
isotropic thermal conductivity in pump-probe transient thermoreflectance. Rev. Sci. Instrum.
2008, 79, 114902(1-9).
35. Wang, Z.; Xie, R.; Bui, C. T.; Liu, D.; Ni, X.; Li, B et al. Thermal transport in suspended
and supported few-layer graphene. Nano Lett. 2011, 11, 113-118.
36. Morgan, G. J.; Uher, C.; The C-axis electrical resistivity of highly oriented pyrolytic
graphite. Philos. Mag. B. 1981, 44, 427-430.
37. Luo, T.; Lloyd, J. R.; Non-equilibrium molecular dynamics study of thermal energy
transport in Au-SAM-Au junctions. Int. J. Heat Mass Transfer, 2010. 53, 1-11.
38. Kong, B. S.; Geng, J.; Jung, H. T.; Layer-by-layer assembly of graphene and gold
nanoparticles by vacuum filtration and spontaneous reduction of gold ions. Chem Commun,
2009, 16. 2174-2176.
39. Shi, Y.; Kim, K. K.; Reina, A.; Hofmann, M.; Li, L. J.; Kong, J.; Work function
Engineering of graphene electrode via chemical doping. ACS Nano, 2010, 4, 2689-2694.
40. Kim, K. K.; Reina, A.; Shi, Y.; Park, H.; Li, L. J.; Lee, Y. H et al. Enhancing the
conductivity of transparent graphene films via doping. Nanotechnology, 2010, 21, 285205(1-6).

133 
 

Chapter 6 Templated Growth of Polyaniline on GNP and Its
Thermoelectric Properties	

6.1 Abstract:
Polyaniline (PANi)/exfoliated graphene nanoplatelets (GNP) nanocomposites were prepared by
in situ polymerization of aniline monomer in the presence of GNP for thermoelectric
applications. PANi has a strong affinity for GNP due to π electron interactions, forming a
uniform nanofibril coating. A paper-like nanocomposite was prepared by controlled vacuum
filtration of an aqueous dispersion of PANi decorated GNP. The Seebeck coefficient of the
resulting nanocomposite changes with initial concentration of aniline in the solution as well as
the protonation of PANi, reaching as high as 33 µV/K for nanocomposites containing
approximately 40wt% of PANi and with a protonation ratio of 0.2. The presence of GNP
improved the electrical conductivity of the nanocomposites to 59 S/cm. As a result,
thermoelectric figure of merit ZT of the nanocomposites is 2 orders of magnitude higher than
either of the constituents, exhibiting a significant synergistic effect. This chapter is based on a
journal publication: Xiang, J., Drzal, L. T.(2012). Templated growth of polyaniline on exfoliated
graphene nanoplatelets (GNP) and its thermoelectric properties. Polymer, 2012 (in press)
 

6. 2 Introduction:
Previous research (Chapter 5) has shown that it is possible to separate the electron and
phonon transport in gold nanoparticle modified GNP paper nanocomposites. Given the high
price of gold nowadays, this method proves the concept but is not practical for real application.
Instead of using metal nanoparticles to serve as spacers, a conductive polymer can be similarly

134 
 

deposited at the interfaces of GNP to preferentially reduce thermal conductivity. Figure 6-1 is a
schematic representation of replacing Au with a polymer. This chapter is dedicated to the
investigation of thermoelectric properties of this novel nanocomposite.

 

Figure 6-1: A schematic representation of replacing the Au nanoparticle spacers in the highly
ordered GNP paper with a layer of polymer.

Thermoelectric materials have great potential in applications such as power generation
and solid state heating/cooling, realizing direct conversion between thermal energy and electrical
energy without involving moving mechanical components or hazardous working fluids [1,2]. The
efficiency of thermoelectric materials is dependent on a dimensionless figure of merit ZT,
defined by (S2σ/)T (S, σ, , T are the Seebeck coefficient, electrical conductivity, thermal
conductivity and absolute temperature, respectively and S2σ is the power factor) The major
challenge in designing a good thermoelectric material is to separate the interdependence of
parameters in ZT. For example, improving electrical conductivity by increasing carrier
concentration usually results in a reduction in Seebeck coefficient, while improving the carrier
mobility is generally considered as the most effective way to increase both electrical
135 
 

conductivity and Seebeck coefficient [3,4]. Until now, inorganic semiconductors and their
compounds have received the most attention because of their high power factor [5,6]. Recently,
there has been research underway directed at developing low dimensional nanomaterials or their
composites to replace bulk components to achieve a higher ZT [7]. However, the relative high
cost of raw materials and lack of scalability of production has created a great need for
inexpensive, easy to process materials for emerging novel applications.
Polymer based thermoelectrics are gaining ever increasing interest because of their low
cost and low thermal conductivity [8-10]. In particular, conductive polymers such as polyaniline
has been extensively studied and are known for their environmental stability and facile acid base
doping and de-doping characteristics [11-13]. The thermoelectric properties of polyaniline have
also been extensively studied [14-17]. However, the performance is usually limited by its low
power factor which presents a significant challenge for polymers to compete with inorganic
semiconductors. To further improve the thermoelectric properties of conductive polymers,
Grunlan et al. first proposed a segregated-network polymer composite where highly conducting
CNTs were introduced to the emulsion polymer matrix to create electrically connected but
thermally disconnected pathways in an effort to decouple the electron and phonon transport [18].
This composite approach was then adopted in conductive polymer systems and a synergistic
effect was observed [19-23], counterintuitive to what was commonly believed that a simple
mixture of two components will not lead to an enhancement in performance [24,25].
Given the fact that the performance of carbon nanotubes based thermoelectric materials is
still low with a highest reported room temperature ZT in the range of 10-3-10-2, a very large area
of active material is needed to generate the same amount of power as the conventional bismuth
telluride module. Therefore, the high cost of producing high quality carbon nanotubes in a

136 
 

commercial scale might still be the bottleneck for wide spread application of CNT based
nanocomposites.
While PANi can be successfully grown on the surface of carbon nanotubes due to their
structural similarity [26,27], the actual interaction area between individual CNT and PANi could
be very small due to the extremely small diameter of the nanotubes. Rapid synthesis of PANi in
the presence of GNP by in situ polymerization was carried out at various concentrations of
aniline. The 2D platelet morphology provides a much larger "interaction surface area" with PANi
than CNT and possibly enables a templating effect for ordered growth of individual PANi
nanofiber. Due to the strong affinity of PANi for the basal plane of GNP, a uniform coating of
PANi can be formed. It is also mentioned in several literature citations that the Seebeck
coefficient of conjugated polymers is related to its electrical conductivity, which is primarily
determined by the dopant used and doping strength [28,29]. Protonation (deprotonation) of PANi
on the surface of GNP was altered by exposing the as-made PANi/GNP to different acid/base
concentrations. A highly ordered composite film was prepared on a filter membrane which
showed good flexibility and mechanical robustness upon drying and compaction. The
thermoelectric properties (Seebeck coefficient, thermal and electrical conductivity) of the
PANi/GNP nanocomposites with various composition of polymer and various degree of
protonation are discussed.

6.3 Experimental:
Production of exfoliated graphene nanoplatelets
GNP are produced by exfoliating the graphite intercalated compounds (Asbury Inc.) in a
conventional kitchen microwave oven following the procedures described previously. The
137 
 

intercalation and exfoliation does not damage the sp2 carbon scaffold in the basal plane, therefore
maintaining the undisrupted carrier transport. The GNP particles used in this experiment are
GNP200 with an average diameter of 5µm and BET surface area of about 200 m2/g.
Synthesis of PANi/GNP core-shell nanostructure (as-made)
120 mg of GNP200 was dispersed in 60 ml 1M HCl with slight sonication. Certain
weight of aniline (Mw: 93.13g/mol from Sigma Aldrich) was added to the suspension and
sonicated for 60 mins at 5°C. Ammonium persulfate (Mw: 228 g/mol from Sigma Aldrich) as the
oxidant was pre-dissolved in 5 ml 1M HCl at 5°C and added within 1 min to the aniline chloride
solution with GNP particles. The solution was further bath sonicated for 15 mins and the reaction
was kept for 24 h under magnetic stirring. The reaction product was washed with centrifugation
with 1M HCl and ethyl alcohol many times. A polyaniline/GNP thin film structure was then
formed on a PVDF (Durapore, 0.22µm) filter membrane by vacuum assisted self assembly. The
film was then dried in a vacuum oven at 60°C for 48h.
Synthesis of neat emeraldine salt polyaniline
Ammonium persulfate (APS, 0.05 mole, 11.5g) was dissolved in 200 ml 1M HCl at 05°C. Aniline (20 ml, 0.219 mole) was dissolved in 300 ml 1M HCl at 0-5°C. APS was added to
the solution of aniline hydrochloride within 1 min. The solutions were mixed by strong stirring
for 2 mins followed by continuous stirring at around 0-5°C for 24 h. The reaction product was
filtered and washed with several portions of 1M HCl first and ethanol to get rid of un-reacted
products and oligomer. The dark green sample cake was vacuum dried at 60°C for 48 h and is
known as the emeraldine salt form of polyaniline.

138 
 

Achieving polyaniline with different levels of protonations:
To achieve a higher protonation of PANi, the as-made PANi coated particles were
suspended in 0.1M HCl for 2 h with stirring. The re-protonated powder was then filtered on a
PVDF membrane to form a thin film and subsequently dried in dynamic vacuum at 60°C. A
deprotonated PANi/GNP was prepared by suspending the particles in 0.1M NH4OH solution for
2 h. The resulting powder was filtered on PVDF membrane to form a thin film and subsequently
dried in dynamic vacuum at 60°C.
Seebeck coefficient and thermal conductivity measurement:
Thermal conductivity, Seebeck coefficient measurements were conducted on the thin film
nanocomposite in a Janis model ST-100 cryostat cooled with flowing liquid nitrogen. Due to the
small thickness of the sample (usually around 100 µm), the metal-film resistor cannot be
attached to heat the sample, therefore a polyimide backed strain gauge (350 Ω) was used as a
resistive heater. A silver-filled epoxy was used to fix the gauge to the side of the film, the other
end to a "L-shaped" copper block, and two thin copper strips along the length of the sample. (See
Figure 6-2) Two thermocouples made from copper and constantan wires were soldered directly
to the copper strips. The Seebeck coefficient (S) and thermal conductivity (κ) were measured
simultaneously by passing a current through the metal film resistor (Ih), thereby generating a
temperature gradient (∆T) in the sample. The Seebeck coefficient was then calculated according
to equation (6.1) when the system reached steady-state.

S 

V
T

139 
 

(6.1)

where ∆V is the thermoelectric voltage measured using the copper leads of the thermocouples.
Thermal conductivity was calculated according to equation (6.2):

Vh I h l
 
T s

(6.2)

where Vh, l and s are the measured voltage across the resistor, distance between thermocouples
and the cross sectional area of the sample. Measurements were conducted under vacuum (<10-5
torr) to minimize heat conduction and convection through the surrounding medium. A fused
quartz sample was used as a calibration standard for thermal conductivity and the data reported
here have been corrected for parasitic and radiation heat losses. Labview software was used for
automated data acquisition.

140 
 

 

Figure 6-2: Schematic representation of the experimental setup for measuring the Seebeck
coefficient and thermal conductivity of thin film sample. Inset is the photo of the actual sample
mounted to the "L-shaped" copper block.

Electrical conductivity measurement (room temperature):
In-plane electrical conductivity was measured with a Keithley 2400 sourcemeter in a four
point configuration with a probe spacing d of 1 cm on thin rectangular film. Electrical resistance
was measured and the conductivity of the rectangular film was calculated by equation (6.3):

141 
 

 ( S / cm ) 

d i
v  w t

(6.3)

where d is the probe spacing in cm, i is the current in mA, v is the voltage in mV, w is the width
of the film in cm, and t is the film thickness in cm.
Scanning electron microscopy:
A scanning electron microscope (Auriga 39, Carl Zeiss Microscopy, LLC ) was used to
image PANi/GNP nanocomposites with an accelerating voltage of 2 kV and a working distance
around 2 mm. The sample was placed on a aluminum holder.
X-ray photoelectron spectroscopy:
XPS results of PANi/GNP with different levels of protonation were obtained from a
Physical Electronics PHI 5400 ESCA system with a Mg Ka x-ray source (1253.5eV photons).
All samples were mounted on standard sample holders by means of double-sided adhesive tapes.
Regional scans of N 1s and Cl 2p spectra were collected to investigate the presence of various
nitrogen species and chlorine species in order to estimate the protonation ratio of polyaniline.
Raman spectroscopy:
A LabRAM ARAMIS laser Raman spectrometer (Horiba Scientific) with 532 nm 50mW
DPSS laser was used to inspect the samples. Neat PANi powder, the as-made GNP film as well
as PANi/GNP film were used for the measurement.

142 
 

TGA analysis:
A thermogravimetric analyzer (TA instruments Q500) was used. 10-15 mg of PANi/GNP
sample was loaded into a Pt pan. The sample was heated at 10°C/min at a resolution of 4°C/min
to 700°C in air.

6.4 Results and discussions:
6.4.1 Synthesis of PANi/GNP and characterizations:
Direct dispersion of GNP particles in 1M HCl is difficult because of the highly
hydrophobic basal planes. Bath sonication is essential prior to the addition of the oxidizing agent
to help break down the agglomerates of GNP in the aqueous solution so that more GNP surface
area is accessible for PANi nucleation. A few minutes into the reaction, nucleation started and
the color of the solution turned dark at which point sonication can be turned off. The nuclei of
PANi grew on the basal plane of GNP due to strong π-π interaction, which helped overcome the
van der Waals attraction and facilitated dispersion of GNP in the aqueous solution. PANi
continued to grow on the nuclei during synthesis until all monomer dissolved in the solution was
reacted. The amount of PANi coated on the surface of GNP is directly related to the initial
concentration of aniline in the solution. In particular, a higher concentration of aniline in the
vicinity of GNP led to the formation of more PANi nuclei and a longer growth period. In all the
experiments, the as-synthesized PANi/GNP hybrid particles after the reaction dispersed well in
the aqueous solution possibly due to the slight positive charges carried by the polymer backbone
as a result of proton doping on the imine nitrogen [30]. The film prepared at 50mM aniline
exhibited a dark brown color to it after vacuum drying while it was much lighter in color for film

143 
 

prepared at 30mM aniline. Figure 6-3 depicts the synthesis of PANi/GNP hybrid particle by insitu polymerization and the resulting paper-making process by vacuum assisted self-assembly.
Detailed morphology of the PANi/GNP samples as well as PANi synthesized in the
absence of GNP were revealed by scanning electron microscopy shown in Figure 6-4(a-f). Neat
polyaniline has an interconnected granular feature with PANi nanofibers growing from the
surface. This distinctive morphology with the formation of needle-like nanostructure could result
from a heterogeneous nucleation and growth since the later formed PANi are more likely to grow
onto the "seeds" that were already formed in the solution. In comparison, PANi grew onto the
surface of GNP and formed a uniform coating. Depending on the initial concentration of
monomer, the as-formed PANi grew longer and formed a high density, highly ordered PANi
fibril forest in the nanocomposite prepared at 50mM aniline solution. The structural similarity
between PANi and GNP plays an important role in creating this highly ordered nanostructure.
Unlike the bulk synthesis of PANi, GNP provided a strong and flat template for adhesion of assynthesized PANi. It is expected that π electron interaction is the primary interaction to hold
PANi and GNP together. Further increasing the concentration of monomer in the solution is
found to be ineffective in "loading" more nanofibrils on GNP surface because only the aniline in
close vicinity of GNP particles would be attracted to the surface while the rest of the monomer
undergoes homogeneous nucleation and growth. Polyaniline nanofibers synthesized in the bulk
did not bind to the surface of GNP and remained dispersed in the aqueous solution. The
excessive unbound PANi can be easily washed off by repeated centrifugation and washing. In
order to make a film from the PANi decorated GNP particles, multi-filtration was carried out to
ensure that a higher particle alignment can be achieved in the final nanocomposite film.

144 
 

 

Figure 6-3: Schematic representation of in-situ polymerization of aniline in the presence of GNP
particles and a photo showing the flexibility of the final nanocomposite film.
 
 
 
 
 
 
 
 
 
 
 
 

145 
 

 

 

Figure 6-4: (a,b) SEM micrographs of neat polyaniline salt showing the granular structure and
small needle-like nanostructure growing out of the surface. (c,d) surface morphology of
PANi/GNP particles (in-situ polymerization of aniline in the presence of GNP particles in 1M
HCl at 0.03M aniline). (e,f) surface morphology of PANi/GNP particles (in-situ polymerization
of aniline in the presence of GNP particles in 1M HCl at 0.05M aniline).
146 
 

Figure 6-4(cont'd)

 

 
 

147 
 

Figure 6-4(cont'd)

 

 
 

148 
 

The amount of polyaniline coated on the surface of GNP can be estimated from TGA weight loss
curve and the derivative of weight loss vs. temperature shown in Figure 6-5(a-b). In particular,
the weight loss before 100°C is due to some volatile species and the moisture adsorbed on PANi
while the weight loss within 200 to 250°C could originate from the loss of bound water
molecules acting as secondary dopant in PANi since synthesis was carried out in an aqueous
medium. It is observed in Figure 6-5b that neat PANi has a much sharper peak at 233°C as
compared to the PANi/GNP composites, suggesting a lower amount of bound water molecules
on the composites possibly due to a more hydrophobic surface of PANi coated GNP than the
neat PANi surface. The major weight loss started from 300°C with a maximum thermal
decomposition at approximately 461°C for neat PANi. For PANi/GNP samples, the major
weight loss did not start until 350°C with maximum decomposition occurring at 480°C which is
20°C higher than the neat PANi. These results provide additional evidence that PANi is
interacting with the surface of GNP, leading to an enhanced thermal stability. While GNP itself
decomposes at 650-700°C, it can be estimated from the curve that the amount of PANi
synthesized on GNP is around 30wt% for PANi/GNP (0.03M aniline) and 40wt% for PANi/GNP
(0.05M)

149 
 

 

Figure 6-5: a: TGA of neat polyaniline salt, neat GNP, and PANI/GNP synthesized at 0.03M and
0.05M initial concentration of aniline. b: Derivative of weight loss vs. temperature for neat PANi,
neat GNP, and PANi/GNP synthesized at 0.03M and 0.05M initial concentration of aniline. 
 
 
 
 
 
 
 

150 
 

Figure 6-5(cont'd)

 

Polyaniline synthesized by current oxidative polymerization is in its emeraldine oxidation state
and it has also been established in the literature that the imine nitrogens are preferentially
protonated by hydrochloric acid [12]. Different levels of protonation are believed to alter the
electronic structure of the polyaniline as well as the thermoelectric properties of the
nanocomposite. Figure 6-6(a-b) shows the N1s core level spectrum of PANi/GNP (@0.05M
aniline) showing there are primarily three distinctive nitrogen species in the sample. In particular,
the imine nitrogen (-N=) with binding energy at 399.1eV, amine nitrogen (-NH-) with binding
energy at approximately 400eV and the positively charged nitrogen (N+) between 401eV and
402eV due to protonation in close vicinity of the Cl anions [31,32]. The Cl2p core-level
151 
 

spectrum suggested the presence of two major components at about 197.2 and 200.1eV which
can be attributed to the chloride anion and the covalently bonded chlorine shown in Figure 6-6(cd). While it is difficult to directly quantify protons by XPS, the amount of counter ion (i.e.
chloride anion) can be readily estimated from XPS. Therefore, a good indicator of the level of
protonation achieved is to quantify the negative Cl species with protonation expressed by the
ratio between Cl anion and total N atoms (i.e. [Cl-]/[N]). Theoretically, the maximum
protonation that can be achieved is 0.5 which means all the imine nitrogens in the polymer are
protonated [33]. Table 6-1 shows the atomic percentage of different nitrogen and chlorine
species in the samples with different protonation levels. It is found from the table that the asmade GNP/PANi has a relatively low protonation, similar to the neat polyaniline salt which is
around 0.2.
Table 6-1: Typical atomic percentage of nitrogen and chlorine species on PANi/GNP samples
prepared at 0.05M aniline and estimation of the protonation level

-N= (≈399.1eV)
neat PANi
PANi/GNP as
made
PANi/GNP
reprotonated
PANi/GNP
deprotonated

N+(>401eV)

Cl total

Cl-

N total

Protonation:
[Cl-]/[N]

3.65

0.79

1.47

1.41

6.88

0.20

5.65

1.5

2.75

2.27

11.76

0.19

4.2

2.8

4.37

3.9

9.85

0.4

8.88

0.8

0.5

0.18

10.81

0.02

 
 
 
152 
 

 

 
Figure 6-6: (a) Regional scans of N1s core-level spectra of as-made polyaniline salt showing the
presence of imine nitrogen(-N=), amine nitrogen (-NH-) and positively charged nitrogen (N+) (b):
regional scan of Cl2p core-level spectra of as-made polyaniline salt showing Cl anions. (c):
regional scans of N1s core-level spectra of as-made PANi/GNP prepared at 0.05M of aniline. (d):
regional scans of Cl2p of PANi/GNP(0.05M).
 
 

153 
 

 
Figure 6-6 (cont'd)

 
 
 
 

154 
 

 

Figure 6-6 (cont'd)
 

 
 
 

155 
 

 

Figure 6-6 (cont'd)
 

 

Raman spectroscopy also verifies the presence of polyaniline on the surface of GNP as shown in
Figure 6-7. In particular, several characteristic peaks of polyaniline were observed at 1162 cm-1
(C-H bending of the quinoid and benzenoid rings), 1218 cm-1 (weak C-N stretching), 1485 cm-1
156 
 

(C=N stretching of the quinoid ring), 1580 cm-1 (C-C stretching) as well as the peaks
characteristics of GNP itself such as 1350 cm-1 (D peak), overlapping 1580 cm-1 (G peak) and
the 2D peak at 2700 cm-1.

 

Figure 6-7: Raman shift of neat PANi powder, neat GNP film and the as-made PANi/GNP film
prepared at 50mM aniline concentration.

6.4.2 Thermoelectric properties of PANi/GNP:
The Seebeck coefficient of a material is dependent on the concentration as well as the
mobility of the charge carriers (i.e. holes and electrons). It is generally believed that bi-carrier
material systems are not preferred for thermoelectric applications because electrons and holes
157 
 

have opposite Seebeck coefficients that offset each other according to the equation (6.4) shown
below:

S e e  S h h
S
e h

(6.4)

where Se, Sh, σh and σe are electron Seebeck coefficient, hole Seebeck coefficient, hole
conductivity and electronic conductivity. Since graphite is known to be a bi-carrier system with
an overlapping band structure, both electrons and holes contribute to the establishment of
Seebeck voltage under a temperature gradient which unfortunately counteract each other,
resulting in a low overall Seebeck coefficient [34]. Figure 6-8a shows the Seebeck coefficient (S)
of various GNP/PANi samples from 200K to 300K. It is observed that neat GNP paper has a low
S of about 5µV/K at room temperature as a result of bi-carrier transport. However, the
GNP/PANi nanocomposite prepared at 0.03M of initial aniline concentration reached 19µV/K at
300K while it further increased to 33µV/K for samples prepared at 0.05M of aniline.
Loading of more PANi on the GNP surface is contributing to the Seebeck voltage that
can be generated. The positive value also suggested very clearly that holes are the dominating
charge carriers in the nanocomposites because protons carrying positive charges are being
incorporated to the polymer during the protonation process. On the other hand, the Seebeck
coefficient of the as-made polyaniline is only 7µV/K at room temperature. While a similar
improvement in the Seebeck coefficient has been observed on PANi coated CNT network
[19,20], the CNT network itself has been p-type doped by oxygen during synthesis, therefore
leading to a higher base line value. The highest enhancement ratio reported (∆S/Scnt) is about 80%
[19]. However, the highest enhancement ratio reported here with respect to neat GNP paper is
158 
 

(33-5)/5=560%. One of the possible reasons that could contribute to the large improvement in
Seebeck coefficient for GNP samples compared with CNT is that GNP provides a much larger
and flatter surface for polyaniline to grow on as opposed to a surface with very high curvature in
the case of CNT. Therefore, the polymer chains synthesized in the presence of GNP will be more
ordered and aligned as a result of the templating effect. While the charge transport in a polymer
is dependent on the interchain and intrachain hopping, a more aligned conformation of the
molecular chains is likely to reduce the carrier hopping resistance, therefore leading to higher
carrier mobility and a higher Seebeck coefficient. The schematic representation of the difference
in charge carrier transport in PANi synthesized with and without the presence of GNP was
proposed and shown in Figure 6-9. Another perspective to look at it is that PANi on GNP is
chemically stretched so that the chains are more ordered. In fact, similar improvement in
Seebeck coefficient was reported in which the solution cast PANi film was mechanically
stretched in order to improve the orientation of the polymer chains, resulting in an enhanced
carrier mobility as well as a higher Seebeck coefficient [35,36].
Thermal conductivity of the nanocomposites are presented in Figure 6-8(b) with neat
GNP having the highest conductivity of about 72W/mK at 300K. The conductivity reduces to
39W/mK for samples prepared at 0.03M aniline (corresponding to about 30wt% PANi in the
composite) and further reduces to 13W/mK for samples prepared at 0.05M aniline
(corresponding to about 40wt% PANi). The fine feature of PANi fibrils on GNP surface is
believed to strongly scatter heat carrying phonons, leading to a lower interfacial thermal
conductance. Although there is a significant reduction in thermal conductivity for the
nanocomposites as a result of polymer encapsulation, they are still one order of magnitude higher
than the thermal conductivity of neat PANi measured to be 0.6W/mK at room temperature.

159 
 

Despite the absolute value of thermal conductivity is still too high for a good thermoelectric
material, nanoengineering the GNP surface by a uniform coating of polymer proved effective in
reducing interfacial thermal conductance, therefore decreasing the overall thermal conductivity.
It is also worth noting that nanostructuring techniques such as the paper making process and
polymer encapsulation could potentially completely alter the heat conduction capability of a
highly thermally conductive particle like GNP that has an estimated intrinsic thermal
conductivity greater than 1000W/mK.

 

Figure 6-8: (a) Seebeck coefficient of GNP/PANi nanocomposites prepared at different aniline
concentrations (0.03M and 0.05M) as well as the performance for neat GNP paper and neat
polyaniline salt measured from 200K to 300K. (b). Thermal conductivity of GNP/PANi (0.03M
and 0.05M) as well as neat GNP paper and neat polyaniline salt from 200K to 300K.
 

160 
 

Figure 6-8 (cont'd)

 
 

161 
 

 

Figure 6-9: (a) Proposed polymerization process for polyaniline in the absence of GNP and the
charge transport mechanism by inter and intrachain hopping. (b) Proposed polymerization of
PANi in the presence of GNP and the proposed charge transport mechanism in the composites.
 

The electrical conductivity of the nanocomposites is reported in Figure 6-10. It is
observed that the encapsulation of GNP with PANi led to a decrease in the conductivity. For the
nanocomposites with the highest amount of PANi (≈40wt%), conductivity reduced from 200
S/cm to 59 S/cm. The same surface PANi fibrils that grew out of GNP also scatter the electrons
in addition to phonons. Although PANi fibrils are electrically conductive, they are far from being
162 
 

fully doped to achieve the maximum conductivity as evidenced by its relatively low protonation
level determined by XPS. Given that PANi can be thought of as conducting (protonated) islands
intermixed in a sea of non-conducting regions, electron transport took place primarily by
hopping both interchain and intrachain, significantly lowering the mobility of the carriers [37]. In
particular, if the PANi chain is less ordered, the electron scattering is likely to further reduce
conductivity as in the case of PANi synthesized without the templating effect of GNP particles.
For comparison, the electrical conductivity of neat PANi in compressed pellets form is 1.5 S/cm
at room temperature.

 

Figure 6-10: Electrical conductivity of GNP/PANi (0.03M and 0.05M) as well as neat GNP
paper and neat polyaniline salt at room temperature.
 

163 
 

6.4.3 Effect of protonation on thermoelectric properties of PANi/GNP nanocomposites:
It has been reported that there is an inverse relationship between Seebeck coefficient and
electrical conductivity for solvent cast PANi film (i.e. higher protonation correlates to a lower
Seebeck coefficient). One purpose of this article is to further investigate if the trend still holds
for PANi synthesized in an aqueous environment that cannot be dissolved in the solvent (i.e.
PANI in its aggregated form). The maximum protonation that can be achieved theoretically on
PANi is 0.5 with two imine nitrogens being protonated every four nitrogen atoms on PANi
chains. Figure 6-11a shows the Seebeck coefficient of PANi/GNP prepared at 0.05M aniline at
different protonation levels. It is found that Seebeck coefficient reached a maximum for the asmade samples (protonation: 0.15-0.2) while further reprotonation or deprotonation decreased the
Seebeck coefficient. This is in analogy to a recently published work that claims highest
thermoelectric performance of PEDOT can be achieved through controlling the oxidation level
[38]. No matter it is the oxidation level or the protonation level, one possible explanation for this
phenomenon is that different protonation ratio or oxidation level in a conductive polymer results
in a shift of the Fermi level to regions with different curvature in electronic density of states.
Park et al. reported the highest Seebeck coefficient of neat PANi salt in compressed pellets form
to be approximately 35µV/K at room temperature but the conductivity is only 0.03 S/cm [16].
While the GNP/PANi nanocomposite achieved a similar Seebeck coefficient at 300K, the
electrical conductivity is 3 orders of magnitude higher because of the presence of GNP particles.
In essence, GNP particles effectively make up the loss in electrical conductivity caused by
insufficient protonation of PANi in an effort to achieve the highest thermoelectric performance.
Thermal conductivity of PANi/GNP, however, is not directly affected by the protonation as
shown in Figure 6-11(b). Phonon transport is dependent on the interfacial thermal conductance

164 
 

which is determined by bonding strength, phonon spectrum matching and geometrical contact
area. Protonation is not likely to alter any of the above mentioned properties, therefore the
thermal conductivity at 300K remained at approximately 15 W/mK.
 

a

 

Figure 6-11: (a) Seebeck coefficient of PANi/GNP prepared at 0.05M aniline with different
protonation levels. (b) Thermal conductivity of PANi/GNP (50mM) with different levels of
protonations.
 
 
 

165 
 

Figure 6-11 (cont'd)

b

 
 

Electrical conductivity of the nanocomposites is immediately affected by protonation as
shown in Figure 6-12. A higher protonation of PANi suggests a higher concentration of charge
carriers which leads to a higher electrical conductivity. It is also worth noting that re-protonation
of PANi/GNP nanocomposites can only be carried out in a very dilute acid concentration to
avoid possible hydrolysis of PANi which would alter the molecular structure of PANi, making it
non-conducting [39]. For the nanocomposite that was re-protonated in 0.1M HCl, electrical
conductivity reached as high as 174 S/cm.

166 
 

 

Figure 6-12: Electrical conductivity of PANi/GNP samples prepared at 0.05M aniline with
different protonation ratio.

The thermoelectric properties, the σ/κ ratio as well as the thermoelectric figure of merit
ZT at room temperature are compared in Table 6-2 for neat GNP, neat PANi, PANi/GNP(50mM,
as-made) as well as the PANi/GNP(50mM, re-protonated). In particular, σ/κ ratio increases
significantly from 2.5 to around 12 suggesting a successful separation of electrical and thermal
conductivity of the nanocomposites. At the same time, ZT remains basically the same for neat
GNP and neat PANi. The highest ZT was obtained for the nanocomposite prepared at 0.05M
aniline with a protonation of around 0.2 reaching 1.51×10-4. The fact that both ZT and Seebeck
coefficient of the composites are almost 2 orders of magnitude larger than either of its
constituents calls into question the previously established viewpoint that thermoelectric
properties cannot be enhanced through simple mixture of two components. Although ZT is still
167 
 

low, the unique flexibility and mechanical robustness of this nanocomposite makes it possible for
them to be installed to the surface of irregular geometry such as the roof and exterior of the
house, potentially complementing electricity generated by solar cells.
Table 6-2: The thermoelectric properties, σ/κ and ZT of neat PANi, neat GNP and PANi/GNP
(50mM aniline, as-made) and PANi/GNP(50mM aniline, reprotonated) at room temperature.
Samples
Neat PANi
Neat GNP
PANi/GNP
(50mM, as made)
PANi/GNP(50mM,
reprotonated)
 

σ (S/cm)
1.5
200

κ (W/m K)
0.6
74

σ/κ ratio
2.5
2.7

S (µV/K)
7
5

ZT(300K)
3.68×10-6
3.04×10-6

59

13

4.6

33

1.51×10-4

174

15

11.6

19

1.26×10-4

6.5 Conclusion:
Polyaniline was synthesized in the presence of exfoliated graphene nanoplatelets to make
lightweight, low cost, flexible nanocomposite for thermoelectric applications. Depending on the
initial concentration of aniline, highly ordered, highly densified PANi nanofibril were grown on
the surface of GNP as a result of a strong π electron interaction. The Seebeck coefficient of the
nanocomposites depended on the weight composition of PANi with the best performance
achieved at 50mM as the starting concentration of monomer. The degree of protonation of PANi
determined by the dopant strength was estimated by the ratio of chloride anion to the total
nitrogen atoms in the composites and was found to be closely related to the thermoelectric
properties. A higher protonation did not necessarily lead to better performance because of the
inverse relationship between electrical conductivity and Seebeck coefficient. As a result of this
trend, a composite approach by using GNP not only preserved the high Seebeck coefficient of
PANi but also significantly increased the electrical conductivity of the nanocomposites. In

168 
 

addition, it appears that GNP served as a "template" for PANi to grow, potentially stretching the
molecular chains of PANi chemically to enhance carrier mobility. The combined improvement
of power factor overcomes the increase of thermal conductivity, resulting in a net gain in
performance. This flexible, lightweight, inexpensive nanocomposite based on exfoliated
graphene is an alternative organic thermoelectrics that are still primarily dominated by carbon
nanotubes composites. The combination of low cost and high availability of GNP/PANi
nanocomposites make them attractive for solar thermal electric conversion generating extra
electricity from the sun.

Acknowledgement:
I would like to thank Prof. Don Morelli and Dr. Kevin Zhou from the department of Chemical
Engineering and Materials Science for their assistance in this project.

169 
 

REFERENCES

170 
 

References
 

1. Snyder, G.J.; Toberer, E.S. Complex thermoelectric materials, Nat. Mater. 2008, 7, 105-114.
2. Tritt, T.M.; Bottner, H.; Chen, L.D. Thermoelectrics: Direct solar thermal energy conversion.
MRS Bull. 2008, 33, 366-367.
3. Makala, R.S.; Jagannadham, K.; Sales, B.C.J. Pulsed laser deposition of Bi2Te3-based
thermoelectric thin films. J. Appl. Phys. 2003, 94,3907-3918.
4. Liufu, S.C.; Chen, L.D.; Yao, Q.; Wang, C.F. Assembly of one-dimensional nanorods into
Bi2S3 films with enhanced thermoelectric transport properties. Appl. Phys. Lett. 2007, 90,
112106.
5. Yan, X.; Poudel, B.; Ma, Y.; Liu, W. S.; Joshi, G.; Wang, H et al. Experimental studies on
anisotropic thermoelectric properties and structures of n-type Bi2Te2.7Se0.3. Nano Lett. 2010, 10,
3373-3378.
6. Martin, J.; Wang, L.; Chen, L.; Nolas, G. S. Enhanced Seebeck coefficient through energybarrier scattering in PbTe nanocomposites. Phys. Rev. B. 2009, 79, 115311(1-5).
7. Dresselhaus, M.S.; Chen, G.; Tang, M. Y.; Yang, R. G.; Lee, H.; Wang, D. Z et al. New
directions for low-dimensional thermoelectric materials. Adv. Mater. 2007, 19, 1043-1053.
8. Dubey, N.; Leclerc, M. Conducting polymers: efficient thermoelectric materials. J. Polym. Sci.
Part B. Polym. Phys. 2011, 49, 467-475.
9. Toshima. N. Conductive polymers as a new type of thermoelectric material. Macromol. Symp.
2002, 186, 81-86.
10. Chang, K.C.; Jeng, M. S.; Yang, C. C.; Chou, Y. W.; Wu, S. K.; Thomas, M. A et al. The
thermoelectric performance of poly(3,4-ethylenedioxythiophene) /poly(4-styrenesulfonate) thin
films, J. Electron. Mater. 2009, 38, 1182-1188.
11. MacDiarmid, A.G.; Chiang, J.C.; Richter, A.F.; Epstein, A.J. Polyaniline: A new concept in
conducting polymers. Synth. Met. 1987, 18, 285-290.
12. Chiang, J.C.; and MacDiarmid, A.G. 'Polyaniline': Protonic acid doping of the emeraldine
form to the metallic regime. Synth. Met. 1986, 13, 193-205.
13. Epstein, A.J.; Ginder, J. M.; Zuo, F.; Bigelow, R. W.; Woo, H. S.; Tanner D. B et al.
Insulator to metal transition in polyaniline. Synth. Met. 1987, 18, 303-309.
171 
 

14. Li, J.; Tang, X.; Li, H.; Yan, Y.; Zhang, Q. Synthesis and thermoelectric properties of
hydrochloric acid-doped polyaniline. Synth. Met. 2010, 160, 1153-1158.
15. Sun, Y.; Wei, Z.; Xu, W.; Zhu, D. A three-in-one improvement in thermoelectric properties
of polyaniline brought by nanostructures. Synth. Met. 2010, 160, 2371-2376.
16. Park, Y. W, Lee, Y. S, Park, C.; Shacklettee, L.W.; Baughman, R. H. Thermopower and
conductivity of metallic polyaniline, Solid State Commun, 1987, 63, 1063-1066.
17. Yao, Q.; Chen, L.; Xu, X.; Wang, C. The high thermoelectric properties of conducting
polyaniline with special submicron-fiber structure. Chem. Lett. 2005, 34, 522-523.
18. Yu, C.; Kim, Y.S.; Kim, D.; Grunlan, J.C. Thermoelectric behavior of segregated-network
polymer nanocomposites. Nano Lett. 2008, 8, 4428-4432.
19. Meng C.; Liu, C.; Fan, S. A promising approach to enhanced thermoelectric properties using
carbon nanotube networks. Adv. Mater. 2009, 21, 1-5.
20. Yao, Q.; Chen, L.; Zhang, W.; Liufu, S.; Chen, X. Enhanced thermoelectric performance of
single-walled carbon nanotubes/polyaniline hybrid nanocomposites. ACS Nano, 2010, 4, 24452451.
21. Kim, D.; Kim. Y.; Choi, K.; Grunlan, J.C.; Yu, C. Improved thermoelectric behavior of
nanotube-filled polymer composites with poly(3,4-ethylenedioxythiophene) polystyrenesulfonate.
ACS Nano, 2010, 4, 513-523.
22. Yu, C.; Choi, K.; Liang, Y.; Grunlan, J.C. Light-weight flexible carbon nanotube based
organic composites with large thermoelectric power factors. ACS Nano, 2011, 5, 7885-7892.
23. Chen, J.; Gui, X.; Wang, Z.; Li, Z.; Xiang, R.; Wang, K et al. Superlow thermal conductivity
3D carbon nanotube network for thermoelectric applications, ACS Appl. Mater. Interfaces, 2012,
4, 81-86.
24. Bergman, D. J.; Levy, O. Thermoelectric properties of a composite medium. J. Appl. Phys.
1991, 70, 6821-6833.
25. Lin-Chung, P. J.; Reinecke, T. L. Thermoelectric figure of merit of composite superlattice
systems, Phys. Rev. B. 1995, 51, 13244-13248.
26. Zengin, H.; Zhou, W.; Jin,J.; Czerw, R.; Smith, D.W. Jr. Carbon nanotube doped polyaniline.
Adv. Mater. 2002, 14, 1480-1483.
27. Sainz, R.; Benito, A. M.; Martinez, M. T.; Galindo, J.F. Soluble self-aligned carbon
nanotube/polyaniline composites. Adv. Mater. 2005, 17, 278-281.
172 
 

28. Mateeva,N.; Niculescu, H.; Schlenoff, J.; Testardi, L.R.; Correlation of Seebeck coefficient
and electric conductivity in polyaniline and polypyrrole. J. Appl. Phys. 1998, 83, 3111-3117.
29. Maddison, D.S.; Unsworth, J.; Roberts, R.B. Electrical conductivity and thermoelectric
power of polypyrrole with different doping levels. Synth. Met. 1988, 26, 99-108.
30. Li, D.; Kaner, R. B.; Processable stabilizer-free polyaniline nanofiber aqueous colloids.
Chem. Commun. 2005, 26, 3286-3288.
31. Tan, K.L.; Tan, B.T.G.; Kang, E.T.; and Neoh, K.G. X-ray photoelectron spectroscopy
studies of the chemical structure of polyaniline. Phys. Rev. B, 1989, 39, 8070-8073.
32. Yue, J and Epstein, A.J. XPS study of self-doped conducting polyaniline and parent systems.
Macromolecules, 1991, 24, 4441-4445.
33. Zuo.F.; Angelopoulos, M.; MacDiarmid, A.G.; Epstein, A.J. Transport studies of protonated
emeraldine polymer: A granular polymeric metal system. Phys. Rev. B. 1987, 36, 3475-3478.
34. Uher.C. Thermopower of exfoliated graphites between 1.7 and 300K. Phys. Rev. B. 1982, 25,
4167-4172.
35. Yan, L.; Ohta, N.; Toshima, N. Stretched polyaniline films doped by (+)-10Camphorsulfonic acid: Anisotropy and improvement of thermoelectric properties. Macromol.
Mater. Eng. 2001, 286, 139-142.
36. Holland, E.R.; Monkman, A.P. Thermoelectric power measurements in highly conductive
stretch-oriented polyaniline films. Synth. Met. 1995, 74, 75-79.
37. Lin, Y.F.; Chen, C. H.; Xie, W. J.; Yang, S. H.; Hsu, C. S.; Lin, M. T et al. Nano approach
investigation of the conduction mechanism in polyaniline nanofibers. ACS Nano, 2011, 5, 15411548.
38. Bubnova, O.; Khan, Z. U.; Malti, A.; Braun, S.; Fahlman, M.; Berggren, M et al.
Optimization of the thermoelectric figure of merit in the conducting polymer poly(3,4ethylenedioxythiophene). Nat. Mater. 2011, 10, 429-433.
39. Gospodinova, N.; Mokreva, P.; Terlemezyan, L. Influence of hydrolysis on the chemical
polymerization of aniline. Polymer, 1994, 35, 3102-3106.

173 
 

Chapter 7 Improving Thermoelectric Properties of Exfoliated
Graphene Nanoplatelet Paper by Folding 	

7.1 Abstract:
A highly ordered, highly flexible paper-like structure consisting of individual exfoliated
graphene nanoplatelets (GNP) was folded, turning a 2D paper into a 3D structure with improved
thermoelectric properties. As a result of this nano-structuring of the graphene nanoplatelets
through folding, an electrically connected, thermally disconnected structure was created due to
the interconnected edges within the folded paper that allowed electrons to conduct freely while
increasing phonon scattering. In addition, the folded sample can be made into a p-type
thermoelectric material by coating the surfaces of the GNP paper with a thin layer of polyaniline
by rapid mixture polymerization. Mechanical compaction on the folded sample was found to be
essential because a net gain in power factor achieved outweighed the increase in thermal
conductivity, leading to an overall improvement in ZT at room temperature.  
 

7.2 Introduction:
Solid state heating/cooling using thermoelectrics are gaining ever increasing interest and
are projected to be one of the promising technologies that could potentially revolutionize the way
people produce energy [1]. As discussed previously, a thin film nanocomposite consisting of
PANi polymerized on the surface of GNP not only reduced thermal conductivity of GNP paper,
but also improved the Seebeck coefficient without significantly reducing the electrical
conductivity. The research reported here describes the ability to nano-structure the GNP by three
dimensional folding in order to decouple the electron and phonon transport. Electron-crystal
174 
 

phonon-glass material systems such as skutterudites, clathrates, etc. have been extensively
studied in the past for their promising thermoelectric properties [2-4]. Research has also
identified inorganic materials having delicate nanostructures to scatter more phonons than
electrons [5,6]. However, the high cost of raw material, complex material design, sensitivity of
the desired nanostructure to processing conditions are major hurdles for commercial applications
of these nanocomposites [7,8]. In addition, there are very few choices apart from bismuth
telluride alloys when it comes to room temperature application [9,10]. The development of
inexpensive, easy to produce alternatives for low grade heat conversion is a potential significant
opportunity for clean, solar energy generation.
While various nano-structuring techniques have been demonstrated to successfully
separate the electron and phonon transport in previous publications, this report tackles this
challenge from a different perspective. In particular, simple folding of flexible GNP paper was
performed that completely altered the way electron and phonon transport takes place in this
paper. A GNP paper consisting of polyelectrolyte stabilized graphene nanoplatelets in water can
be prepared by vacuum assisted self-assembly following the same procedures described
elsewhere. When folded, electrons are still able to pass through the folded GNP paper due to the
interconnected edges but phonon transport is significantly reduced due to the high anisotropy of
the individual fold. On the other hand, the Seebeck coefficient, which is approximately 5µV/K
for neat unfolded GNP paper can be greatly enhanced by synthesizing a thin layer of polyaniline
by rapid mix polymerization [11] on the surface of GNP paper. The thermoelectric properties of
the folded PANi/GNP are improved and the effect of consolidation on folded paper is discussed.
 

175 
 

7.3 Experiments:
Preparation of GNP paper
GNP was produced as usual. The detailed procedures can be found in Chapter 4. In brief,
PEI solution was prepared containing 0.1 wt% of the polyelectrolyte in DI water. 0.1 g GNP was
dispersed in 100 ml PEI solution and bath sonicated for 30 min under constant stirring. The
suspension was kept stirred for 24 h and filtered through a 0.2 μm Durapore PVDF filter paper to
make the GNP paper.
Preparation of PANi coated GNP paper
GNP paper was prepared using the procedures described above. However, an extra
thermal annealing was carried out at 340°C for 2h to "burn off" residual polyethyleneimine.
Upon annealing, the paper was further compacted in the press with a pressure of 1000psi. A
solution containing 0.6g aniline dissolved in 110 ml 1M HCl (0.05M) and another solution
containing 1g ammonium persulfate (APS) dissolved in 20 ml 1M HCl were prepared. Two
solutions were then mixed rapidly with magnetic stirring for 3 mins before the GNP paper was
immersed. The reaction took place for 12 h at 0-5°C. The PANi coated GNP paper was then
washed with DI water and isopropanol repeatedly and dried at 60°C for 24h. The schematic
representation was shown in Figure 7-1:

 

Figure 7-1: Schematic representation of experimental procedures to prepare PANi coated GNP
paper by rapid mixture polymerization. 
176 
 

Folding by the mold, (pressed and unpressed)
Folding of the paper was carried on a custom made mold shown in Figure 7-2. In brief, a
5mm × 5mm acrylic rod of 6' was cut into bars of 3'' in length, these bars were then arranged in
the diamond fashion with one side fixed to a 12'' long acrylic rod side rail, and the bottom fixed
to an acrylic substrate. The GNP paper was laid on top of the mold, and individual folds were
introduced to the paper with two additional rods pressed gently down on the cavity of the mold.
Individual folded paper strips prepared from the mold was then arranged in another mold and
compressed to make a sample of desired thickness. To prevent the sample from springing back,
silicone caulk was applied at the open edges of folded sample (referred to as the unpressed
sample). To further consolidate the folded paper, the sample was further compressed to 50% of
its original length in a custom made mold (not shown here) and was referred to as the pressed
sample. It is worth noting that ideally, the sample should consist of only one folded paper.
Multiple strips were used here because a minimum sample length is required for attaching the
thermocouples to the side of the folded paper. (note: thermocouples were attached to the
connected edge of the folds instead of the open edges)

177 
 

 

Figure 7-2: custom made mold for paper folding, and stacking of folded paper strips into desired
thickness for subsequent measurements.

Thermoelectric property measurements (Seebeck coefficient and thermal conductivity)
Near room temperature (270K-320K) thermal conductivity, Seebeck coefficient
measurements were conducted on the folded GNP sample (either pressed or unpressed) in a Janis
model ST-100 cryostat cooled with flowing liquid nitrogen. A silver-filled epoxy was used to fix
one end of the sample to a copper block, the other end to a 1000 Ω metal-film resistor, and two
thin copper strips along the length of the sample. (See Figure 7-3) Two thermocouples made
from copper and constantan wires were soldered directly to the copper strips. The Seebeck
coefficient (S) and thermal conductivity (κ) were measured simultaneously by passing a current
(7mA) through the metal film resistor (Ih), thereby generating a temperature gradient (∆T) in the
sample. (The experimental section in Chapter 6 has detailed calculations of the Seebeck
coefficient and thermal conductivity.)
178 
 

                                   

 

Figure 7-3: (a) schematic representation of experiment setup for Seebeck coefficient, thermal
conductivity, electrical conductivity measurement for the folded sample. (b) a photo of the
folded PANi/GNP paper mounted on the cryostat before measurement.
 

179 
 

Electrical conductivity measurement
Electrical conductivity of the folded paper was measured by the standard 4-point method
where two thin adhesive copper strips were attached to the "interconnected edge" as the voltage
probes, while a current passed through the entire sample. The resistance was measured and the
conductivity (σ) taking into account the geometric factor was obtained by equation (7.1):



l
R A

Equation (7.1)

where l is the distance between voltage probes, R is the measured resistance, A is the cross
sectional area of the folded paper.
Scanning electron microscopy:
A scanning electron microscope (Auriga, Carl Zeiss Microscopy) was used to image
polyaniline nanofiber on the surface of GNP paper with an accelerating voltage of 2kV and a
working distance around 2 mm.
TGA analysis:
A thermogravimetric analyzer (TA instruments Q500) was used. 10-15mg of PANi/GNP,
neat GNP and neat polyaniline was loaded into a Pt pan. The sample was heated at 10°C/min at a
resolution of 4°C/min to 700°C in air.  
 

180 
 

7.4 Results and discussions:
7.4.1 GNP paper folded
Micropores are introduced during the paper making process by vacuum filtration, which
negatively affects the mechanical robustness of the paper. Therefore, it is essential to compress
the paper first to get rid of the pores and create a highly aligned structure inside the paper. The
enhanced alignment of the nanoplatelets results in a large improvement in its flexibility and
"foldability". It was reported previously that the through-plane electrical conductivity of
compressed GNP paper is very low (in the order of 10-2 S/cm) while the in-plane electrical
conductivity exceeds 1000 S/cm because of high anisotropic conduction in the plane of the paper
[12]. In the folded paper, electrons are now able to utilize the high in-plane conduction paths
established by the interconnected edges rather than rely on the limited cross plane conduction.
The ability of electrons to tunnel across conductive particles that are not in contact with each
other offers another advantage. When the edges are partially damaged due to large folding stress,
electrons can still travel between particles without being scattered. On the other hand, the
through-plane thermal conductivity has been reported to be very low in comparison to the high
in-plane value. In a nanostructured GNP paper, heat transport takes place through phonons (i.e.
lattice vibration) and is primarily determined by the interfacial thermal conductance. In particular,
the matching of phonon density of states, the type of interaction or bonding strength, and the
geometrical overlapping area are the three major factors in interfacial heat conduction. It is
expected that the strong bending at the folds significantly reduced the geometrical overlapping
areas of GNP, leading to a reduced contribution from the high in-plane thermal conductivity due
to strong phonon scattering.  Figure 7-4 depicts the concept of the electrically connected,

181 
 

thermally disconnected paper after folding. Figure 7-5 depicts the process of cutting the paper
into strips and folding them on the custom-made mold.
Although the GNP paper did not fracture completely at the edge after experiencing severe
deformation at the folds introduced during the folding process, SEM micrographs did identify
some breakage and discontinuity at the fold areas for the folded paper as shown in Figure 7-6.
However, it seems that the exterior (convex side) of the folded GNP was fractured while the
interior (concave side) experienced much less damage due to the different tensile and
compressive stresses incurred in the folding process. Despite partial damage at the outer layers
of the edge, the "inner core" of the paper was still connected, contributing some mechanical
integrity to the folded structure.  

 

Figure 7-4: schematic representation of electron transport within the folded paper, the air gaps
between individual fold is exaggerated.

182 
 

 

Figure 7-5: Photos of flat GNP paper, GNP paper cut to strips, folding the GNP strips on the
custom-made acrylic mold, and the folded paper.(from top left to bottom right)

183 
 

a 

 

b 

 

Figure 7-6: (a) SEM micrographs of a folded strip at low magnification (scale bar: 1mm), (b) the
convex side of the folding edge (scale bar: 100µm), (c) the concave side of the folding edge
(scale bar: 100 µm), (d) a magnified view of the tip of the folding edge (scale bar: 20 µm)
 

184 
 

Figure 7-6(cont'd)

c 

 

d 

 

185 
 

7.4.2 Electronic conduction of the folded paper
From the SEM micrographs shown above, folding did partially damage the structural
integrity of the paper at the fold. Depending on the intrinsic "folding strength" of the GNP paper
and its thickness, the damage can be controlled. However, the paper remained as an entire
coherent piece without breaking apart at the fold no matter how much they were folded. The
change in resistance of the folded GNP paper and the mechanical integrity was expected to
change with the density of folds. The electrical resistance of a flat strip of paper, a slightly folded
paper strip (Figure 7-7a) and a highly folded paper (Figure 7-7b) with either one fold or two
folds between the probes were measured and tabulated in Table 7-1 and 7-2 below:

a 

 

b 

 

Figure 7-7: (a) Photos of four-point electrical resistance measurement on the slightly folded GNP
paper and (b) a heavily folded GNP paper with one fold between the voltage probes.
 
 
 

186 
 

Table 7-1: Electrical resistance measurement on the folded paper (1 fold between voltage probes)
Sample (1 fold)

Resistance (Ω)

Flat

0.27

Slightly folded

0.28

Heavily folded(pressed)

0.28

 

Table 7-2: Electrical resistance measurement on the folded paper (2 folds between voltage
probes)
Sample (2 folds)

Resistance (Ω)

Flat

0.54

Slightly folded

0.56

Heavily folded(pressed)

0.4

The resistance basically remained the same value after folding. For the measurement conducted
on the folded paper with 2 folds in between the voltage probes, the electrical resistance dropped
when a pressure was applied on the heavily folded sample. It is possible that in this case
electrons not only traveled along the plane of the paper, but also took advantage of the throughplane transport because adjacent folds were now in close contact with each other. Therefore,
folding the paper with slight compression can improve the contact to maintain and even reduce
the electrical resistance of the folded sample. It is therefore believed that partial breakage at the
edge did not increase the resistance of the paper because the "core" of the paper is still connected,
sufficient for electronic conduction.

187 
 

7.4.3 Folded PANi/GNP paper as a p-type thermoelectric material
Thermoelectric properties of neat polyaniline as well as their carbon nanotube composites
have been extensively documented in the literature [13,14]. Due to the structural similarity
between PANi and GNP, PANi has also been successfully synthesized on individual GNP
particles by π electron interaction. A hybrid PANi/GNP paper in which every single GNP
particle was coated with a layer of PANi has been prepared by vacuum assisted self-assembly
[Chapter 6]. However, the flexibility and foldability of such paper is reduced and breakage at the
edge is more likely to occur as compared to the neat GNP paper, failing the purpose of creating a
continuous pathway inside this folded structure. Rapid mix polymerization provided a solution to
this problem by creating a thin layer of conductive polymer coating only on the surfaces of GNP
paper while leaving the interior intact, preserving the foldability and introducing multifunctionality to GNP paper.
The PANi coated paper appeared dark purple in color after washing and drying (Figure 78a). The surface morphology of PANi grown on GNP paper were shown in Figure 7-8(b,c). A
highly ordered and dense structure of PANi nanofibrils were observed, forming a homogeneous
layer encapsulating the top and bottom surfaces of the GNP paper. It is worth noting that the
inner core of the paper was still connected by individual GNP although the surface layer was
disconnected under folding. (Figure 7-8d)
The amount of PANi synthesized on GNP was verified by TGA shown in Figure 7-9. A
significant increase of the decomposition temperature of PANi was observed, possibly due to the
strong interaction between PANi and GNP and a possible templating effect of GNP during the
course of polymerization. It was found that a major weight loss began at about 350°C, indicating
the onset of PANi decomposition until a second major weight loss occurred at around 600°C,

188 
 

characteristic of oxidation of GNP itself. Therefore, the total weight of PANi in the composite
was estimated to be about 10wt%.

a 

 

 

Figure 7-8: (a) Photo of the as-synthesized PANi/GNP paper (5 cm × 5 cm). (b) An overview of
the surface morphology of PANi/GNP paper by SEM (scale bar: 10 µm). (c) SEM micrograph of
PANi nanofibrils grown on the surface of GNP (scale bar: 100 nm). (d) folding edge of the
PANi/GNP paper showing the interconnected "core" (scale bar: 100 µm).
 
 

189 
 

Figure 7-8(cont'd)
 

 
 

 

190 
 

 

Figure 7-9: TGA analysis of the as synthesized PANi/GNP paper and the neat polyaniline in air.

The thermoelectric properties of both the pressed and un-pressed folded PANi/GNP
papers were measured on the liquid nitrogen cooled cryostat platform. For the un-pressed sample,
it is expected that gaps existed between the individual folds, leading to a possible "air-paper"
composite. The existence of air gaps could be detrimental to the performance of the folded paper.
Upon compaction in the press, the air gaps were effectively eliminated, creating solid contact
between the individual folds. The presence of the nanofibril structure could be acting as
"interlocking nanohooks" that helped achieve a more condensed structure. As a result, the
Seebeck coefficient increased from 8µV/K to 21µV/K at room temperature despite the fact that
the thermal conductivity went up from 0.09W/mK to 0.7W/mK due to increased density. The
positive Seebeck coefficient also indicates holes are the dominant charge carriers in the sample,
191 
 

in agreement with the fact that protons were incorporated into polymer during synthesis. As a
result of the compaction, electrons could find more conduction paths inside the folded sample as
the porosity was significantly reduced compared with the un-pressed sample. The conductivity
shown in Table 7-3 also increased by almost an order of magnitude to 0.7 S/cm. While the
absolute value of conductivity is still low due to the geometric design of the sample (i.e. a 5mm
×5mm cross section area, constrained by the side length of the acrylic rod in the mold), a model
for estimating the conductivity of the folded paper as well as ways to enhance conductivity is
discussed below.

Table 7-3: Electrical conductivity of the folded PANi/GNP paper with and without compaction
Probe distance (mm)

Resistance (Ω)

Conductivity (S/cm)

Unpressed

2

17

0.04

Pressed

2

1.2

0.7

 

192 
 

 

 

Figure 7-10: :(a) Seebeck coefficient of the folded PANi/GNP paper (pressed vs. unpressed)
from 270K to 320K (b) Thermal conductivity of the folded PANi/GNP paper (pressed vs.
unpressed) from 270K to 320K. 
193 
 

The electrical conductivity of the folded GNP/PANi paper is directly related to how the paper
was being folded. The folding width is dependent on the side length of the acrylic rod used to
construct the mold for folding. A simple mathematical model was built to estimate the electrical
conductivity of the folded paper. It is known from earlier discussion that the high curvature
introduced in the folding process did not increase the resistance of the paper compared to when it
was not folded. Take an individual folded strip as an example shown in Figure 7-11: the
important parameters relevant to the calculation are the thickness of the unfolded paper (t), the
number of folded segments (n), the folding period (fp, dependent on the mold), as well as the
width of the strip (w). The resistance of this folded strip (R) is assumed to be the same as the
unfolded paper. A consolidated sample is obtained by compaction so that all the gaps inside the
folded paper are eliminated. The length of the bulk sample (l) now becomes the thickness of the
paper times the number of folding segments, with the cross sectional area to be 25mm2 (5mm ×
5mm). The electrical conductivity of the folded paper (σ) can be obtained by the following
equation given that the resistivity (ρ) of a flat GNP paper is known.

 

Figure 7-11: schematic representation of the folded paper for estimating the bulk electrical
conductivity and mathematic derivation.
 

194 
 

 

l

A R

tn
( w  f p )  (  paper 

n fp
wt

(
)

t 2
1
) 
fp
 paper

From the above estimation, electrical conductivity of the folded paper is proportional to (t/fp)2,
given that the thickness of the paper is around 50µm, folding period is 5mm, the resistivity of the
unfolded paper is >1000S/cm, the estimated electrical conductivity of the folded paper is at least
0.1 S/cm, in good agreement with the measured value of about 0.7 S/cm. The reason why the
measured value is higher is because the model described above does not take the cross-plane
tunneling of electrons into account, thus leading to the lower limit of conductivity. In terms of
thermoelectric figure of merit ZT at room temperature, the unfolded GNP paper has an estimated
ZT around 3.7×10-6 while the folded GNP/PANi paper achieves a ZT of 1.3×10-5.
To further improve electrical conductivity, one can increase the ratio between t and fp by
making a thicker paper (e.g. 100µm) and decrease the folding width to 1mm (which may require
a custom made interlocking gears to make the folds instead of making it by hand), the projected
electrical conductivity could be improved to 10 to 20 S/cm, which leads to a room temperature
ZT in the range of 10-4. Another advantage brought by folding is that the original thin film
material is transformed back to the bulk material, which is conducive to establishing a
temperature difference perpendicular to the surface of the thermoelectric module. For thin films,
the maximum temperature difference attainable is limited by the thickness of the film, creating
new challenges as to how to heat the film in the plane of the film because lateral temperature
difference is not easily established. (e.g. think about the wrist watch example, the temperature
difference between body heat and ambient is more easily achieved if heat gradient is established

195 
 

perpendicular to the surface of the module, which requires a bulk configuration). Therefore, this
approach by simple mechanical folding opens up plenty of opportunities for flexible GNP paper
to be used for ambient temperature thermoelectric applications.

7.5 Conclusion:
An electrically connected, thermally disconnected structure was achieved by
mechanically folding a highly ordered, highly flexible GNP paper. The high anisotropic features
of the GNP paper allowed electrons to travel undisrupted in the folded paper, yet significantly
discouraged the transport of phonons thanks to the highly curved but still interconnected folding
edges that scattered most of the phonons. Polyaniline was successfully synthesized on the surface
of GNP paper that resulted in a significant improvement in the Seebeck coefficient of the folded
sample after sample compaction without compromising the "foldability" of the paper. This is the
first article that discusses the potential of decoupling electron and phonon transport of a flexible
thin film material by simple mechanical manipulation. In addition, this approach overcomes the
difficulty of establishing a temperature difference across thin films by converting them to a
highly ordered bulk material, making the folded structure compatible with current thermoelectric
module design.
 

 

196 
 

REFERENCES

197 
 

References
1. Tritt, T. M. Thermoelectrics run hot and cold. Science, 1996, 272,1276-1277.
2. Sales. B. C.; Mandrus, D.; Chakoumakos, B. C.; Keppens, V.; Thompson, J. R. Filled
skutterudite antimonides: electron crystals and phonon glasses. Phys. Rev. B, 1997, 56, 1508115089.
3. Nolas, G.S.; Morelli, D.T.; Tritt, T.M.; Skutterudites: A phonon-glass-electron crystal
approach to advanced thermoelectric energy conversion applications. Annu. Rev. Mater. Sci.
1999, 29, 89-116.
4. Nolas, G.S.; Slack, G.A.; Schujman, S. B. Semiconductor clathrates: a phonon glass electron
crystal material with potential for thermoelectric applications. Semiconduct Semimet. 2001, 69,
255-300.
5. Zebarjadi, M.; Esfarjani, K.; Shakouri, A.; Bahk, J. H.; Bian, Z.; Zeng, J et al. A. Effect of
nanoparticle scattering on thermoelectric power factor. Appl. Phys. Lett. 2009, 94, 202105(1-3).
6. Kim, W.; Zide, J.; Gossard, A.; Klenov, S.; Stemmer, A.; Shakouri, A et al. Thermal
conductivity reduction and thermoelectric figure of merit increase by embedding nanoparticles in
crystalline semiconductors. Phys. Rev. Lett. 2006, 96, 045901(1-4)
7. Snyder, G. J and Toberer. E. S. Complex thermoelectric materials. Nat. Mater. 2008, 7, 105114.
8. Minnich, A. J.; Dresselhaus, M.S.; Ren, Z. F.; Chen, G. Bulk nanostructured thermoelectric
materials: current research and future prospects. Energy Environ. Sci. 2009, 2, 466-479.
9. Yamashita, O.; Tomiyoshi, S.; Makita, K. Bismuth telluride compounds with high
thermoelectric figures of merit. J. Appl. Phys. 2003, 93, 368 (1-7).
10. Poudel, B.; Hao, Q.; Ma, Y.; Lan, A.; Minnich, B.; Yu, X et al. High-thermoelectric
performance of nanostructured bismuth antimony telluride bulk alloys. Science. 2008, 320, 634638
11.Yan, X.; Chen, J.; Yang, J.; Xue, Q.; Miele, P. Fabrication of free-standing, electrochemically
active, and biocompatible graphene oxide-polyaniline and graphene-polyaniline hybrid papers.
ACS Appl. Mater. Interfaces, 2010, 2, 2521-2529.
12. Xiang, J.; Drzal, L. T. Electron and phonon transport in Au nanoparticle decorated graphene
nanoplatelet nanostructured paper. ACS Appl. Mater. Interfaces, 2011, 3, 1325-1332.
13. Li, J.; Tang, X.; Li, H.; Yan, Y.; Zhang, Q. Synthesis and thermoelectric properties of
hydrochloric acid-doped polyaniline. Synth. Met. 2010, 160, 1153-1158.

198 
 

14. Sun, Y.; Wei, Z.; Xu, W.; Zhu, D. A three-in-one improvement in thermoelectric properties
of polyaniline brought by nanostructures. Synth. Met. 2010, 160, 2371-2376.

199 
 

Chapter 8 Summary and Recommended Studies	

8.1 Summary:
Graphene nanoplatelets (GNP) produced from thermal expansion of graphite intercalated
compounds with microwave assisted heating have excellent electrical, thermal and mechanical
properties as a result of the preserved conjugation in the basal plane of graphene. To investigate
how these unique physical properties can be transferred to a polymer matrix, these GNP particles
were incorporated into paraffin wax and the resulting electrical and thermal conductivity of the
nanocomposites prepared by casting were measured. It is found that nanocomposites with GNP
of larger aspect ratio achieved electronic percolation at a lower loading. Interestingly, the
thermal conductivity at the percolation threshold of the nanocomposites remained almost the
same although the conductive network has already been established inside the composite. This
distinction in the properties of the composites is attributed to the different conduction
mechanisms of electrons and phonons across the polymer/GNP interfaces. For electronic
conduction, direct physical contacts between conductive particles is not necessary because of the
ability of electrons to tunnel across a non-conductive barrier of a finite thickness independent of
the intrinsic resistivity of the film. On the other hand, heat is conducted by phonons of different
vibration frequencies. The efficiency of heat transport in composites is largely dependent on the
properties of the matrix-filler interfaces. In the GNP/polymer case, a large thermal interface
resistance is likely to offset the contribution from the highly conductive GNP particles, thus
leading to a marginal or even a decrease in the thermal conductivity especially at a filler loading
close or slightly above the percolation threshold of the nanocomposites. However, the significant

200 
 

increase of electrical to thermal conductivity ratio (σ/κ) at percolation threshold provides insight
to developing GNP/polymer based nanocomposites for thermoelectric energy conversion.
From SEM micrographs, it is found that the GNP particles in the composite prepared by
casting assumed random orientation with some particle folding, wrinkling, fracture incurred
during sample processing. To avoid this, a GNP scaffold was pre-established by dispersing
individual GNP particles in a polyelectrolyte/water solution followed by vacuum assisted selfassembly, forming a highly ordered, highly condensed paper structure with excellent flexibility
as well as mechanical robustness. As a result, the electrical conductivity of this GNP paper is
significantly improved from 150 S/cm to 880 S/cm after annealing and compaction due to a
combined reduction of tunneling and contact resistances as a result of the decomposition of the
polyelectrolyte, increased contact area between GNP particles and a better alignment of the
particles. Thermal conductivity also increased significantly to as high as 200W/mK after
annealing and compaction because of a reduction in interfacial thermal resistance.
In order to take advantage of the high electrical conductivity of the paper for
thermoelectric application, thermal conductivity needs to be reduced. One approach to modify
and separate the electronic and phononic conduction in the GNP paper was achieved by
depositing metal nanoparticles on the surfaces of GNP to act as spacers that prevented immediate
physical contact between adjacent GNP particles. After annealing and compaction, the electrical
conductivity of GNP/Au hybrid paper increased by 80% from 880 S/cm to 1500 S/cm while the
thermal conductivity reduced by almost 10%. Metal nanoparticles created more tunneling
channels for electrons inside the hybrid paper while the phonons are unable to tunnel across but
scattered due to the difference in phonon spectra in metals and carbon. In addition, the weak van
der Waals interaction between metals and GNP also scattered phonons without disrupting

201 
 

electron flow. The separation of carriers in Au nanoparticle decorated GNP paper further
illustrates the fundamental difference in electronic and phononic conduction that can be taken
advantage of in optimizing nanocomposites based on GNP for thermoelectric energy conversion.
Built on this understanding on separation of electron and phonon transport, a conductive
polymer polyaniline was used to replace the metal nanoparticles in an effort to further reduce
thermal conductivity of the nanocomposites. Polyaniline was successfully synthesized on the
surface of GNP forming a homogenous coating layer. Depending on the initial concentration of
monomer prior to polymerization, the amount of PANi synthesized on GNP can be readily
controlled with a highest composition of 40wt% PANi in the composite prepared at 50mM of
aniline starting concentration. SEM micrographs indicated a highly ordered, highly condensed
PANI nanofibrils growing from the surface of GNP. In contrast, PANi synthesized in the absence
of GNP has a random granular structure very distinctive from PANi/GNP nanocomposite. The
difference in morphology also suggests that GNP served as a template during PANi
polymerization and contributed to a more ordered and aligned growth of molecular chains on the
surface of GNP rather than a random coil-like conformation without GNP. The encapsulation
effect of PANi on GNP successfully reduced the thermal conductivity of the nanocomposite film
prepared by vacuum filtration from more than 70W/mK to around 13W/mK with 40wt% PANi
in the composites. The ratio of electrical and thermal conductivity (σ/κ) increased from 2.85 for
neat as made GNP paper to 4.6 for (40wt%PANI)/GNP nanocomposite.
From XPS, the protonation ratio of PANi on GNP can be readily estimated [Cl-]/[N]
which is around 0.2 for the as-made nanocomposites. The as-made samples can be further doped
to a higher protonation ratio by a facile dilute protonic acid treatment (e.g. 0.1M HCl solution for
2h). The electrical conductivity of the re-protonated sample increased from 60S/cm to 174S/cm

202 
 

while the thermal conductivity, which relies much less on the incorporation or ejection of
hydrogen proton, remained almost the same at around 15W/mK. This offers another interesting
route to further separate the electrical and thermal conductivity, effectively increasing the σ/κ
ratio from 4.6 for the as-made samples to 11.6 for the re-protonated samples.
The Seebeck coefficient (S), also benefited greatly by incorporation of PANi into the
GNP paper. In particular, the as-made 40wt%PANi/GNP has a Seebeck coefficient around
33µV/K at room temperature while it is only 5µV/K and 7µV/K for the neat GNP paper and neat
PANi. Since the Seebeck coefficient S is dependent on the mobility of the charge carriers, in the
case of conductive polymers such as polyaniline, inter-chain and intra-chain hopping have a
strong impact on carrier mobility. A highly oriented polymer chain can reduce the barriers of
inter-chain and intra-chain hopping and allow the carriers to move easily because of the high
conjugation in PANi/GNP. The templating effect of GNP is analogous to a chemical stretching
applied to the chains of the conductive polymer, enhancing the carrier mobility. Compared with a
mechanical stretching reported previously to enhance Seebeck coefficient [1], the chemical
stretching could be more effective because it is in situ: every single chain is "stretched" if
polymerization occurs in close vicinity of the template (i.e. GNP particle). One thing to note that
in spite of the increased electrical conductivity achieved by re-protonation, the Seebeck
coefficient dropped to around 20µV/K at room temperature for the highly protonated samples.
One possible explanation is that re-protonation results in a shift of the Fermi level in the
conductive polymer to regions with different curvature in the density of states function, which
still needs to be verified in later studies.
This research has demonstrated with evidence that proper nano-structuring work wonders.
The ability to separate the electron and phonon transport in GNP paper by engineering particle

203 
 

interfaces significantly altered the properties of neat GNP. In the case of PANi decorated GNP,
neither neat GNP nor neat polyaniline is considered as good thermoelectric materials possibly
because of their certain undesired physical properties and a possible bi-carrier transport nature.
Nevertheless, PANi/GNP nanocomposite possess much enhanced thermoelectric properties with
a room temperature ZT of 1.5×10-4 almost 2 orders of magnitude higher than either of its
constituents. These inexpensive graphene nanoplatelets and the ability to control their properties
by nano-engineering will certainly open up numerous opportunities for graphene based
nanocomposites for novel energy generation and conversion application.

8.2 Recommended future works:
 

8.2.1 Solvent processing of PANi/GNP:
Another possible way to prepare PANi/GNP nanocomposites is by solvent processing
and film casting. Instead of synthesizing PANi by in-situ polymerization in the presence of GNP,
neat PANi emeraldine salt can be prepared first. PANi can be dissolved in a proper solvent
followed by addition of GNP, film casting and removal of solvent by evaporation. However, it is
worth noting that the emeraldine salt (ES) form of polyaniline cannot readily dissolve in most of
the solvents while the emeraldine base (EB) form of polyaniline has been found to dissolve well
in solvents such as N-methyl pyrrolidinone (NMP), m-cresol, etc. Therefore, it is necessary to
convert ES PANi to EB PANi first by suspending the as-made PANi in dilute base (e.g. 0.1M
NH4OH) for several hours to deprotonate. At the same time, NMP would be a nice solvent for
GNP too because a recent paper by Hernandez et al. reported successful liquid phase exfoliation
of graphite in NMP to produce thin layers of graphene because of the comparable surface energy
between the solvent and graphene [2].
204 
 

It is expected that PANi dissolved in NMP maintains a more extended and relaxed
molecular conformation as compared to neat PANi powder whose molecular chains are more
coil-like. The extended chains will still interact with the surface of GNP as a result of π electron
stacking forming a coating layer on the surface. A film of PANi coated GNP can be prepared by
film casting, vacuum assisted self-assembly, or doctor blading.
The as-made EB PANi/GNP film after elimination of solvent can be further protonated
by simply dipping the self-standing film in a solution of protonic acids such as HCl, H2SO4,
acetic acid, and etc. Depending on the concentration and dipping time, different protonations can
be achieved. It would be interesting to compare the thermoelectric properties (electrical, thermal
conductivity, Seebeck coefficient) with the samples made by in-situ polymerization reported in
the previous chapters.

8.2.2 Carrier transport analysis for PANi/GNP:
Previous study showed PANi synthesized in the absence of GNP has granular
morphology as opposed to a more ordered structure if synthesized in the presence of GNP
particles. The enhanced alignment of polymer chains is likely to reduce the inter-chain and intrachain carrier hopping barrier, leading to an increased carrier mobility that contributed to a higher
Seebeck coefficient. A possible way to prove that there is indeed an increase in the carrier
mobility is to perform a Hall measurement. This can be done in the Quantum design Versalab
system (module and the schematics shown in Figure 8-1)

205 
 

Figure 8-1: Schematics and actual photo of the quantum design versalab module and wiring,
CH1 measures the Hall voltage, CH2 measures the in-plane resistivity of the sample. z direction:
coming out of the paper.

By measuring the Hall voltage, it is possible to determine the carrier transport properties and
estimate the carrier mobility from a known carrier concentration. The carrier concentration can
be calculated from Hall coefficient RH assuming the samples studied are considered heavily
doped semiconductors and only one type of charge carriers (either electrons or holes) are present
in the system. RH and the carrier concentration n is expressed as:

RH 

Ey
j x Bz



Vy t
I x Bz



Rxy t
Bz



1
ne

(8.1)

where Ey is the induced electric field in the y direction, jx is the current density in the x direction,
Bz is the magnetic field perpendicular to the sample (coming out of the paper), Vy is the Hall
voltage, Ix is the current in the x direction, Rxy is the Hall resistance, t is the sample thickness and
e is the unit charge. Therefore the slope of Hall resistivity (Rxy•t) versus Bz is the Hall coefficient.
However, preliminary results on the PANi/GNP samples suggest a non linear relationship
between Hall resistivity and magnetic field, which provides evidence that the nanocomposite
sample is still a bi-carrier system with both electrons and holes contributing to the Seebeck
206 
 

coefficient. In order to calculate the carrier concentration and mobility of both electrons and
holes, a more complicated bi-carrier transport model must be used to fit the experimental results
[3]. It would be necessary to compare the transport properties of neat PANi with PANi/GNP
samples combined with a bi-carrier model analysis before any solid conclusion can be drawn.

207 
 

REFERENCES

208 
 

References
1. Yan, H.; Ohta, T.; Toshima, N. Stretched polyaniline films doped by (+) 10-camphorsulfonic
acid: anisotropy and improvement of thermoelectric properties. Macromol. Mater. Eng. 2001,
286, 139-142.
2. Hernandez, Y.; Nicolosi, V.; Lotya, M.; Blighe, F. M.; Sun, Z.; De, S et al. High-yield
production of graphene by liquid phase exfoliation of graphite. Nat. Nanotechnol. 2008, 3, 563568.
3. Morelli, D. T.; Canfield, P. C.; Drymiotis, P. Low-temperature transport properties of NdBiPt.
Phys. Rev. B. 1996, 53, 12896-12901.

209