Liquid Phase Exfoliation of Graphite
to Graphene & its Applications in
Polymeric Nanocomposites
By
Khalid Nawaz
School of Chemical and Materials Engineering (SCME)
National University of Sciences and Technology (NUST)
2015
Liquid Phase Exfoliation of Graphite
to Graphene & its Applications in
Polymeric Nanocomposites
Name Reg. No
Khalid Nawaz 2008-NUST-TfrPhD-Em-E-08
This work is submitted as a PhD thesis in partial fulfillment of the
requirement for the degree of
(PhD in Energetic Materials Engineering)
Supervisor Name: Dr. Noaman Ul Haq
School of Chemical and Materials Engineering (SCME)
National University of Sciences and Technology (NUST)
H-12 Islamabad, Pakistan
September, 2015
i
Thesis Submission Certificate It is to certify that work in this thesis has been carried out by Mr. Khalid Nawaz and
completed under my supervision in School of Chemical and Materials Engineering,
National University of Sciences and Technology, H-12, Islamabad, Pakistan.
Supervisor: ______________
Dr. Noaman Ul-Haq
Assistant Professor
National University of Sciences and
Technology, Islamabad
Submitted through
Principal SCME National University of Sciences and Technology, Islamabad
ii
DEDICATION
Dedicated to all those scientists & engineers who sacrificed their
lives for the welfare of humanity
iii
ACKNOWLEDGEMENTS First of all I am greatly thankful to my Almighty Allah, the most Beneficent, the most
Merciful, the Supreme Power and Creator of this Universe and Who enabled me to
complete my research work. Thousands of blessings be upon Hazrat Muhammad
(Peace be upon him), who is the reason for the creation of this universe.
I would like to thank my supervisor Dr. Noaman Ul-Haq for his advisement,
encouragement and guidance during the course of this research. I would also like to
thank Dr. Muhammad Mujahid (Principal, SCME), Dr. Arshad Hussain (HoD,
Chemical engineering department) and Dr. M.B. Khan for their guidance and support
during my research work.
I will never forget the great and out of the way cooperation of Professor Dr. Jonathan.
N.Coleman and Dr. Umar Khan of Trinity College Dublin (TCD), Dublin, Ireland.
They helped me to great extent to complete most of my project work at School of
Physics in TCD, Ireland. I feel delight in the company of Peter May, Harshit Porwal,
Sweta Bansali and Amro (of chemistry department). It has been an immense pleasure
and privilege working with everyone and being part of such a diverse and intelligent
group of talented individuals at TCD.
In particular, I would like to express my gratitude to my colleague, Muhammad Ayub
who helped me a lot to complete the thesis writing, formatting & compilation by
sparing his precious time.
I would also like to thank my family members for their encouragement and support
while I was working on this thesis.
Finally, I cannot conclude these acknowledgments without recognizing the financial
support of my department and Higher Education commission (HEC) of Pakistan
through International Support Initiatives Program (IRSIP) without which I will not be
able to complete my research work at TCD, Ireland and NUST.
iv
Abstract
Nanocomposites are superior to conventional one in terms of its mechanical
performance. Pristine /functionalized graphene sheets (FGS) were incorporated into a
range of model polymers. Solvent aided blending was adopted for better dispersion of
FGS and graphene sheets in these polymers. Graphene was added to selected
polymers like polyurethane (PU), poly (vinyl chloride) (PVC), poly (acrylonitrile)
(PAN), poly (vinyl alcohol) (PVA) and poly (vinyl acetate) (PVAc) in order to
improve the mechanical performance of these materials. Different forms of graphene
nanosheets like pristine/FGS with different lateral dimensions were selected in order
to study its effects on the mechanical performance of selected polymers in terms of
young’s modulus, tensile strength and elongation at break. Graphene nanosheets were
functionalized with octadecylamine and were incorporated in polyurethane and it was
observed that 2.5 vol% is the mechanical percolation level for this polymer as above
this loading there was improvement in the mechanical performance of polyurethane
while at this loading the elongation at break was suffered slightly. Similarly in case of
poly(vinyl chloride) a critical loading(1.5wt%) was observed at which there was
improvement in mechanical properties of these polymers and almost no elongation at
break was observed for this loading and the modulus determined in this case was
superior to calculated from Halpin-Tsai equation. Two type of graphene nanosheets
with different flake size (one and 3.5 micron) were incorporated in poly
(acrylonitrile). Its comparative study was conducted it was observed the big flake
improved the performance of polymers in terms of modulus and UTs while the
response of small flake in terms of elongation at break was better than big flakes.
Large area graphene oxide were synthesized and were introduced to poly(vinyl
alcohol) and the role of these nano fillers were very pronounced in terms of modulus,
UTS and elongation at break was not disturbed but slightly improved.
Graphene flakes were studied through transmission electron microscopy(TEM) and
Raman spectroscopy while dispersion of these flakes were in selected polymers was
confirmed by scanning electron microscopy(SEM) and the mechanical performance of
these nanocomposites were conducted on Zwick-Roell tensile tester. Graphene-based
v
polymer nanocomposites can be a new versatile soft material with numerous
advantages.
Graphite was exfoliated to graphene using NMP and water as solvent as well.
63mg/ml concentration was obtained during tip sonication in NMP while in case of
water as media the maximum concentration obtained was 7mg/ml using sodium
cholate as surfactant. The concentration of graphene nanosheets were studied through
UV-visible spectroscopy while quality of flakes was studied through TEM and Raman
spectroscopy.
vi
Table of Contents
Certificate
i
Dedication ii
Acknowledgement iii
Abstract iv
Table of Contents vi
List of Figures x
List of Tables xv
List of Acronyms xvi
1.1 Overview and history of graphene and graphene-based materials 1
1.2 Graphene-based polymer nanocomposites 5
1.2.1 Overview and historical perspective 7
1.3 Preparation methods of nano-composites 8
1.3.1 Non-covalent dispersion methods: solution and melt mixing 8
1.3.2 Non-covalent/ in- situ polymerization 9
1.3.3 Graphene-based composites with covalent bonds between
matrix and filler
11
1.3.4 Other methods for composite preparation 12
1.4 Graphene synthesis 13
1.4.1 Growth in-situ on a substrate 14
1.4.2 Bottom up methods to synthesize graphene from organic
precursors.
16
1.4.3 Chemical efforts to exfoliate and stabilize graphene sheets in
solution
18
References 23
2.1 Introduction 32
2.2 Materials 32
2.3 Apparatus and Equipments 32
vii
2.4 Characterization Techniques 33
2.4.1 UV-Vis Absorption Spectroscopy 33
2.4.2 Raman Spectroscopy 34
2.4.3 Tensile testing (TT) 36
2.4.4 Scanning Electron Microscopy 39
2.4.5 Transmission Electron Microscopy 42
References 45
3.1 Objective 47
3.2 Introduction 47
3.3 Experimental Details 49
3.3.1 Concentrated dispersion of graphene. 49
3.3.2 Second Method for extremly high concentration of graphene 51
3.4 Charetarization 52
3.4.1 Transmission Electron Microscopy (TEM) Histogram 52
3.5 Size selection of graphene Flakes according to its lateral dimensions 54
3.5.1 Experimental Details 54
3.6 Results and Discussions 55
3.6.1 Concentration study 55
3.6.2 Raman Spectroscopy 57
3.6.3 Transmission Electron Microscopy 58
Conclusion 61
References 61
4.1 Objective 65
4.2 Introduction 65
4.3 Experimental Procedure 67
4.4 Results and Discussion 68
Conclusion 76
References 76
viii
5.1 Objective 80
5.2 Introduction 80
5.3 Experimental Procedure 81
5.4 Results and Discussion 83
Conclusion 90
References 91
6.1 Objective 93
6.2 Introduction 93
6.3 Experimental Procedure 95
6.4 Results and Discussion 96
Conclusion 105
References 105
7.1 Objective 108
7.2 Introduction 108
7.3 Experimental Procedure 110
7.4 Characterization of PVC-Graphene Composites
111
7.5 Results and Discussion
111
Conclusion 116
References 116
8.1 Objective 119
8.2 Introduction 119
8.3 Experimental Part 121
8.3.1 Composites Preparation and Characterization 121
8.4 Results & Discussions 122
Conclusion 129
References 129
9.1 Objective 132
ix
9.2 Introduction 132
9.3 Experimental Section 135
9.4 Characterization 136
9.5 Mechanical characterization 137
9.6 Results and discussion 137
Conclusion 146
References 147
10.1 Summary of main work 153
10.2 Suggestions for future work 156
References 156
x
List of Figures
Fig No. Figure Title Page No.
Fig. 1.1 (A) Schematic of graphene on a SiC substrate (B) E-k
diagram of graphene grown on SiC displaying the band
opening at the Dirac point and (C) An ARPES intensity map,
displaying the band gap.
15
Fig. 1.2 Schematic of the roll based transfer of graphene films grown
on Cu foil
16
Fig. 1.3 (A) Structure of the hexabenzocoronene (HBC) (B) Structure
of the largest polyaromatic hydrocarbons synthesized to data containing 222 C atoms
18
Fig. 1.4 GO structure demonstrating the distorted sp2 atomic
arrangement and attached functionalities (left) and vial of GO brown dispersion (right). Its brown colour is attributed to the
absence of π conjugated structure.
19
Fig. 2.1 Raman Shift 36
Fig. 2.2 Typical stress-strain curves. 38
Fig. 2.3 Schematic of SEM showing the electron path 41
Fig. 2.4 Detector positions in Zeiss Ultra / Supra 42
Fig. 2.5 Schematic of TEM showing the electron path 44
Fig. 2.6 TEM images of a 5 layered graphene flake. These TEM
images demonstrate how the lengths, widths and count the number of layers were measured.
45
Fig. 3.1 Absorbance versus wavelength curve for 1st cycle using (A) sonic
bath (B) sonic tip 49
Fig. 3.2 Change in concentration with time for 1st step using (A) sonic tip
and (B) sonic bath 50
Fig. 3.3 Absorbance versus wavelength curve for 2nd cycle using (A) sonic
tip for 6hrs (B) sonic tip for 10hrs (C) sonic bath 50
Fig. 3.4 Change in concentration with time for 2nd cycle using (A) sonic tip
for 6hrs (B) sonic tip for 10hrs (C) sonic bath
51
Fig. 3.5 (A) Change in concentration with sedimentation of 64mg/ml
(B) sample (B) Comparative study of sedimentation of 20,30 and
45mg/ml graphene concentartion dispersions
52
xi
Fig. 3.6 (A) Comparative study of average length, width and no of layers
for 20mg/mL (B) 30mg/mL (C) 45mg/mL and (D) 63mg/mL
dispersion
53
Fig. 3.7 Schematic diagram of size separation of graphene flakes
55
Fig. 3.8 Concentration of graphene exfoliated by tip sonicator and
centrifuged at various speed (rpm). 56
Fig. 3.9 Concentration of graphene exfoliated by bath sonicator and
centrifuged at various speed (rpm)
56
Fig. 3.10 Change in ID/IG ratio of graphene exfoliated by (A) sonic tip, (B)
sonic bath with speed (rpm) of centrifuge 57
Fig. 3.11 Increase in the Defect peak of normalized Raman spectra with
increasing rpm (A) for sonic tip (B) for sonic bath 59
Fig. 3.12 TEM images of graphene layers observed (a) un-separated
graphene dispersion for sonic tip (b) 500rpm sonic tip (c) 1000rpm
sonic tip (d) 3000rpm sonic tip (e) un separated graphene
dispersion for sonic bath (f) 500rpm sonic bath (g) 1000rpm sonic
bath (h) 3000rpm sonic bath.
60
Fig. 3.13 (A) shows the change in the average values of length, width and
number of layers with change in rpm for sonic tip sample (B) sonic
bath sample
60
Fig. 4.1 Concentration of graphene after centrifugation (500/45) as a
function of sonication time (Cs=5mg/ml). Concentration was calculated using absorption coefficient “α” value equal to 3.62 ml/mg/m
69
Fig. 4.2 Concentration of graphene after centrifugation (500/45) as a
function of sonication time (CSs=10mg/ml). Concentration
was calculated using absorption coefficient “α” value equal to
3.62ml/mg/m.
70
Fig. 4.3 TEM images of graphene flakes deposited from sample having
concentration of 7mg/ml.
71
Fig. 4.4 Histogram showing (A) the number of layers per flake
measured for 96 hours sonication time (B) average length of flakes and (C) the average width of flakes
72
Fig. 4.5 A
& B
(A) SEM images of the flakes present on the interface of the free standing films prepared (B) SEM image showing the fractured interface
73
Fig. 4.6 Increase in Defect peak (D-band) of normalized Raman
spectra as a function of sonication time
74
Fig. 4.7 Change in ID/IG as a function of sonication time 75
Fig. 4.8 Estimated length of graphene flake with D/G values. 76
Fig. 5.1 FTIR spectra of GO (top), ODA (middle) and GO-ODA
(bottom).
83
xii
Fig. 5.2 SEM images of 40 wt% composite film. 84
Fig. 5.3 Representative stress strain curves. Inset: the low strain
regime.
85
Fig. 5.4 Effect of GO-ODA content on mechanical properties of
composites. (A) Young’s modulus, (B) stress at 3% strain, (C) Ultimate tensile strength, (D) strain at break.
87
Fig. 5.5 The same data as in figure 6.4 but plotted as a function of
GO-ODA volume fraction on a log-log plot. The lines in (A)
and (B) illustrate percolation-like behavior while the line in
(C) illustrates linearity. The vertical arrows illustrate the
percolation threshold while the horizontal arrows show that
value of each property displayed by the polymer.
88
Fig. 6.1 (A) Large numbers of multilayer graphene deposited on a
holey carbon TEM grid. (B) Individual graphene multilayer. (C) Photograph of PVAc−graphene films with mass fractions
of 0%, 0.2%, 0.4%, 0.7%, and 1.5% (volume fractions from
0−0.8%). SEM image of (D) a PVAc and (E) a
PVAc/graphene fracture surface
97
Fig. 6.2 Stress−strain curves for the PVAc/graphene composite film
studied in this work. (inset) Stress−strain curves on a log−log scale. The dotted line represents linearity.
98
Fig. 6.3 Mechanical properties of PVAc films. (A) Young’s modulus,
(B) ultimate tensile strength, and (C) strain at break, as a
function of graphene volume fraction.
99
Fig. 6.4 Measurements of adhesive properties of PVAc/graphene glue.
(A) Photograph of samples used for adhesive testing. (left) Two T-shaped wood pieces glued together for tensile testing.
(right) Two wooden bars, glued together along an overlapping
region (dashed line), for use in shear measurements. (B)
Photograph of T-shaped pieces during a tensile test. (C)
Applied stress plotted as a function of displacement in both
tensile and shear modes for samples glued using homemade
PVAc adhesive. (D) Tensile and shear bond strength and (E)
toughness as a function of graphene content for the
homemade PVAc adhesives. (F) Tensile stress−strain curves
for as-bought commercially available glue and same with 0.7
wt % graphene added. (G) Tensile and shear bond strength
and(H) toughness as a function of graphene content for the
adhesives prepared with commercially available PVAc glue
The dotted lines represent the untreated glue. The data points
represent the glue, diluted and re-concentrated during the
process of graphene addition.
103
xiii
Fig. 7.1 TEM images of graphene nano flakes exfoliated in NMP 112
Fig. 7.2 Effect on Young’s Modulus of PVC after using graphene
nano-flakes
114
Fig. 7.3 Effect on UTS of PVC after using graphene nano-flakes 114
Fig. 7.4 Effect on Elongation at Break of PVC after using graphene
nanoflakes
115
Fig. 7.5 Comparison of Theoretical and Experimental values of
Young’s Modulus
115
Fig. 8.1 Ratio of Raman d-g bands measured on films prepared from
size selected dispersion as a function of final centrifugation rates.
123
Fig. 8.2 Raman spectra of graphene thin film, of selected size flakes
prepared after different centrifugation rate (rpm).
123
Fig. 8.3
(A) & (B)
(A)TEM images of graphene flakes separated by
centrifugation at 500 rpm. (B) TEM images of graphene
flakes separated by centrifugation at
5500 rpm.
124
Fig. 8.4
(A) & (B)
(A) Histograms of flakes length of graphene in DMF peparated at 500 rpm. (B) Histograms of flakes length of graphene in DMF separated at 5500 rpm.
125
Fig. 8.5 Effect of 1 micron () and 3.5 micron (♦) nano-fillers
(graphene) incorporated in PAN polymer on Young’s modulus.
126
Fig. 8.6 Effect of 1 micron () and 3.5 micron (♦) nano-fillers
(graphene) incorporated in PAN polymer on Ultimate Tensile Strength (UTS).
127
Fig. 8.7 Effect of 1 micron (♦) and 3.5 micron () nano-fillers
(graphene) incorporated in PAN polymer on Elongation at break.
128
Fig. 9.1 FTIR spectra of Graphene oxide. 137
Fig. 9.2 TEM Image of graphene oxide. 138
Fig. 9.3 Histogram of Length of graphene oxide nano flakes 138
Fig.9.4 SEM images of LAGO dispersion in PVA 139
Fig. 9.5 Effect of LAGO on modulus of PVA 140
xiv
Fig. 9.6 Effect of LAGO on Tensile strength of PVA 141
Fig. 9.7 Effect of LAGO on elongation at break of PVA 142
Fig. 9.8 Comparison of Theoretical and experimental data of
Young’s modulus
144
Fig.9.9 DSC of PVA and LAGO based nanocomposites (0.35 wt %) 145
xv
List of Tables
Table No. Table Caption/Title Page No.
Table 1.1 Mechanical Properties of Graphene/Graphite based Polymer
Nano-Composites
6
xvi
List of Acronyms
2D Two Dimension
BSE Back Scattered Electron
CNTs Carbon Nano Tubes
CVD Chemical Vapour Deposition
DMF Dimethyl Formamide
EG Expanded Graphene
EM Electromagnetic
EVA Ethylene Vinyl Acetate
FGS Functionalized Graphene Sheets
FTIR Fourier Transform Infrared Spectroscopy
FTIR Fourier Transform Infrared Spectroscopy
GNR Graphene Nano Ribbons
GNRs Graphene nanoribbons)
GO Graphite/Graphene Oxide
GO-OD Graphene Oxide Octadecylamine
GPa Giga Pascal
HBC Hexabenzo Coronene
HDPE High Density Polyethylene
HeNe Helium Neon
HOPG Highly Oriented Pyrolytic Graphite
In Indium
LAGO Large area graphene oxide
LDH Low Density Hydrocarbon)
xvii
LPE Liquid Phase Exfoliation
MPa Mega Pascal
MWCNTs Multi Walled Carbon Nano Tubes
NMP N-Methyl Pyrolidinone
ODA Octa Decyl Amine
PA6 Poly Amide
PAHs Polyaromatic Hydrocarbons
PAN Poly(acrylonitrile)
PE-g-MA Polyethylene Grafted Maleic Anhydride)
PET Poly(ethylene terepthalate)
PI Polyimide
PMMA Poly(methyl methacrylate)
PP Poly Propylene
PS Polystyrene
PSS Polystyrene Sulfonate
PVA Polyvinyl Alcohol
PVAc Polyvinyl Acetate
PVC Poly Vinyl Chloride
RPM Round Per Minute
SAED Selected Area electron Diffraction
SDS Sodium Dodecyl Sulfate
SEM Scanning Electron Microscopy
SiC Silicon Carbide
STM Scanning Tunneling Microscopy
SWNT Single Walled Carbon Nano Tubes
TEM Transmission Electron Microscopy
xviii
THF Tetrahydrofuran
TPa Tera Pascal
TPU Thermoplastic polyurethane
UHV Ultra High Vacuum
UTS Ultimate Tensile Strength
UV/Visible Ultra Violet/Visible Spectroscopy
Xe Xenon
1
Chapter 1
Introduction
1.1. Overview and history of graphene and graphene-based
materials
Graphene was once believed to be an academic material [1] a thermodynamically
unstable one that upon isolation would crumple up on itself. However, this didn’t
stop some scientists investigating how thin they could actually make graphite
planes. The search for graphene started and for the last about forty years the
graphene is under study by scientific community [2-3]. Graphene is a “monolayer of
sp2-hybridized carbon atoms arranged,” in “a two-dimensional lattice”, has
attracted “tremendous attention in recent years owing to its exceptional thermal,
mechanical, and electrical properties” [4-6]. The “in-plane elastic modulus of
pristine, defect-free graphene is approximately 1.1 TPa and is the strongest material
that has ever been measured on a micron length scale [5,6].” Graphene also
demonstrates brittleness [7] readily folds and can be stretched up to 20% more than
any other crystal [8] similarly “GO-derived fillers” can exhibit, “high moduli
(reported values ranging from 208 GPa [9] to over 650 GPa [10]) and can be easily
functionalized to tailor their compatibility with the host polymer.” The refrence
“values of stiffness of GO derived filler materials can be higher than those reported
for nano clays” [11] but “generally lower than those reported for single-walled
2
carbon nanotubes (SWNTs)” [12]. However, “the intrinsic mechanical properties of
SWNTs may be comparable to those of pristine graphene” [12,13]. Moreover, “the
two-dimensional platelet geometry” of graphene and graphene based “materials
may offer certain property improvements that SWNTs cannot provide”when
“dispersed in a polymer composite, such as improved gas permeation resistance of
the composite” [14]. The in-plane stiffness for “chemically modified graphene
(CMG) platelets” is lower and decreases with “increasing level of oxidation of the
platelets” [15]. However in a study conducted on CMG platelets using AFM nano
indentation “reported the opposite, the elastic modulus of the platelets evidently
increasing” “with increasing oxidation level, ranging from 250 GPa for reduced
graphene oxide (RGeO) platelets” up to approximately 650 GPa for graphene oxide
(GeO) platelets [10]. These superior mechanical properties with large aspect ratio
of graphene and graphene derived materials made them potential candidate as
reinforcement in polymeric systems [16]. When dispersed in polymer the thin
sheets transform into wavy or wrinkled structure which loses its modulus value, as
wrinkled structure unfold instead of stretching under applied stress [17]. Some time
incomplete exfoliation or restacking of nanosheets also lower modulus due to
decreased aspect ratio [16]. “One of the most promising applications of this
material is in polymer nanocomposites polymer matrix” which incorporates nano
scale fillers as reinforcement. “Nanocomposites with exfoliated silicate fillers have
been investigated in 1950 [17] but significant academic and industrial interest in
nanocomposites came nearly forty years later” when Toyota Motor Corporation
demonstrated significant improvement in mechanical properties of polymer Nylon-
6 by using montmorillonite as filler [18]. Bunnell “proposed the production of
polymer nanocomposites incorporating as thin as possible” GNPs (“derived from
GICs exfoliated either by shear grinding or thermal treatment”) as fillers in a 1991
3
[19], where it has been suggested that with “10 vol% inclusion of graphite flakes in
polyethylene or polypropylene, the stiffness of the finished product will approach
that of aluminum.” A detailed report published in 2000 which explains the
chemistry of nanocomposites based on exfoliated graphite with 10 nm of thickness
which was produced during in-situ polymerization of caprolactam [20].
Tremendous properties improvements have been reported “versus conventional
polymer composites based on micron-scale fillers such as untreated flake graphite
or carbon black” (CB) at low loading [21-27].
Nanofiller usually tends to agglomerate which could become factor for aspect ratio
reduction which ultimately diminishes its reinforcing role [28], while some
researcher reported that large scale aggregate may be beneficial for enhancement of
mechanical performance of system [29,30]. For “effective reinforcement Strong
interfacial adhesion between the platelets and polymer matrix” is also responsible
[31-35]. Apart from this dispersion phenomenon, the two phases (filler and
polymer) should be compatible with each other; otherwise it may also become a
factor for modulus reduction of composite due to low interfacial adhesion matrix-
filler interface.[36] In order to have composite with superior mechanical properties
the matrix-filler interface should be intelligently tailored.
Functionalization of graphene and graphene based materials is selected route to
tailor the interface in order to improve adhesion between filler and polymer either
covalently or non-covalently [14]. “Hydrogen bonding between GO-derived fillers”
and “their matrix has been reported” to be responsible factor for the improvement in
“modulus and strength observed in several polymers that can serve as hydrogen
bond acceptors or donors” [22,36-38]. Stress transfer at interface can be improved
by covalent bonding between graphene oxide and matrix [35] Just at 0.1 wt%
4
loading of graphene oxide to Nylon-6 nanocomposites the modulus improved 100%
as compared to neat Nylon-6 it may be due to in situ step-growth polymerization
between functional groups of polymer and GeO [39]. Likewise improvement in
mechanical properties of, epoxy, polyurethane, Polystyrene (PS), poly methyl
(methacrylate) (PMMA) and PVDF composites “with GNP, GO-derived fillers and
polymer-grafted CMG respectively, has been reported”. The “formation of covalent
bonds between matrix and filler is suggested to be responsible for this improvement
in mechanical properties”[14,40-47].
Improvements in reinforcement might be unequivocally influenced if host polymer
and polymer grafted to nanofiller have same chemical nature and with relative
molecular weight this behavior is specially studied in polyurethane modulus gain
[14,43,45,46,48-51]At 55 wt% GNP Increases in modulus from “approximately 10
MPa to 1.5 GPa” have been reported in polyurethanes, yet ductility was retained to
the level of rigid thermoplastic (e.g., polycarbonate) [49] . It has been shown by
calculations that randomly oriented graphene nanofiller give better mechanical
performance than randomly oriented nanotubes while in case of aligned nanofiller
the result of CNTs is better than graphene nanosheets [52]. Similarly it has also
been studied that the mechanical properties obtained as a result of exfoliated
graphene/graphene derived materials are better than GNP based system it may be
due to high aspect ratio and high modulus of former materials [53-58]. The
improvement in reinforcement may be due to native properties of filler which was
observed during comparing experimental results with theoretical e.g., “Mori-
Tanaka and Halpin-Tsai models”, one [31,59].
The results calculated by these models have been compared with graphene based
nanocomposites and was observed that reinforcement at low loading surpass the
5
predicted values based on these micro-mechanical models [22,60-63]. The apparent
“discrepancy between these results and theory highlights the need to develop
further understanding of the relative” contributions of native “filler properties and
changes in the polymer matrix in regards to the reinforcement of these systems”.
1.2. Graphene-based polymer nanocomposites
The development of a nano-level dispersion of graphene particles in a polymer
matrix has opened a new and interesting area in materials science in recent years
[79]. These Nano-hybrid materials show considerable improvement in properties
that cannot normally be achieved using conventional composites or virgin
polymers. The extent of the improvement is related directly to the degree of
dispersion of the nano-fillers in the polymer matrix. The most important aspect of
these nano-composites is that all these improvements are obtained at very low filler
loadings in the polymer matrix. Different types of nano graphite forms, such as
expanded graphite and exfoliated graphite, have also been used to produce
conducting nano-composites with improved physicochemical properties. There are
many studies on expanded and exfoliated graphite composites based on a range of
polymers, including epoxy, polymethyl methacrylate, polypropylene, low density
polyethylene, high density polyethylene, polystyrene, Nylon, Polyaniline,
phenylethynyl-terminated polyimide, and silicone rubber. Table 1.1 lists the
percentage enhancement in the mechanical characteristics, such as the tensile
strength at break, storage modulus and flexural strength of
6
Matrix
Filler
type
Filler
loading
(Wt.%a,
vol.%b)
Process
%
Increase
%
Increase
TS
% Increase
flexural
strength
Epoxy
EG
EG
EG
EG
1a
1a
1a
0.1a
Sonication
Shear
Sonication and Shear
Solution
8
11
15
-20
-7
-6
87
PMMA
EG
GNP
21a
5 a
Solution
Solution
21
133
PP
EG
xGnP-1
xGnP-15
Graphite
3b
3b
3b
2.5 b
Melt
Melt
Melt
SSSP
60
8
26
8
LLDPE
xGnP
Parrafin
coated
xGnP
15a
30
Solution
Solution
200
22
HDPE
EG
UG
3a
3a
Melt
Melt
100
33
4
PPS
EG
S-EG
4a
4a
Melt
Melt
-20
-30
PVA
GO
Graphene
0.7a
1.8b
Solution
Solution
76
150
TPU
Graphene
Sulfonated
Graphene
5.1b
1a
Solution
Solution
200
75
PETI
EG 5a
10a
In-situ
In-situ
39
42
Table1.1 Mechanical Properties of Graphene/Graphite based Polymer nano-Composites
7
1.2.1. Overview and historical perspective
Graphene reported Young’s modulus is between 0.5 - 1. TPa with ultimate strength
130 GPa, so, it is thought that graphene is an excellent candidate for mechanical
reinforcement of polymer in the area of nanocomposites. To this end, there is
significant research in which graphene has been added into a variety of polymers to
make nanocomposites, with varying level of success [22,56-57,62,64].
Interestingly, besides the mechanical properties of graphene, there are two
additional stiffening mechanisms for graphene and graphene derivative
nanoparticles to stiffen certain polymer matrices. With hydrogen bonding, graphene
oxide (GO) generally interacts with polar polymers and this leads to apparently
superior mechanical reinforcements due to the change in visco-elasticity of the
polymer matrix, graphene can enhance the degree of crystallinity as a nucleating
agent in semi-crystalline polymers, and therefore stiffens the polymer matrix by
increasing the crystallinity.
Strength and elongation at break of graphene polymer nanocomposites changes due
to stiffness change [36,64] so with good dispersion tensile strength increases while
at the same decrease in elongation at break is also observed [36].
The earliest “reports on polymer composites with exfoliated graphite fillers
emerged from studies on the intercalation chemistry of GICs. Alkali metal-GICs
could initiate the polymerization” of ethylene, styrene, methyl methacrylate, and
isoprene [64-67]. Later on it was also observed that the alkali metal-GICs can also
exfoliate the layers of [68, 69]. Numerous preparing methods “have been accounted
for dispersing both GNP and GO-derived fillers into polymer matrices” in recent
years, which are almost same to those used for other nano fillers [70]. The nature of
the bonding between filler and matrix along with other factors, has profound effect
8
on the mechanical performance of nanocomposites. Whereas in some cases
nanocomposites are produced that “are non-covalent assemblies where the polymer
matrix and the filler interact through relatively weak dispersive forces. However,
presently research is focused to develop chemical bonding between graphene” and
“polymer to promote stronger interfacial bonding” for better mechanical
performance of system.
1.3. Preparation Methods of Nano-composites
1.3.1. “Non-covalent dispersion methods: solution and melt mixing”
This method involves the mixing of colloidal suspensions of “graphene-based
materials with the desired polymer in same solvent to have molecular level”
interaction between nanofiller and polymer [70]. Due to “ease of processing of
graphene nanosheets in aqueous media as well as in organic solvents solution
mixing has been widely reported in the literature [70]. This approach has been used
for incorporating pristine/functionalized graphene” fillers into different polymers,
like: polystyrene (PS) [41,71]. Polycarbonate [72]. Polyacrylamide [73] polyimide
[74] and “poly (methyl methacrylate) (PMMA)” [75,79]. “The facile production of
aqueous GO platelet suspensions via sonication makes this technique particularly
appealing for water-soluble polymers such as poly (vinyl alcohol) (PVA) [36,75,77-
78] and poly (ally amine),” “composites of which can be produced via simple
filtration”[78,80] a broad range of composite films with different loadings of
GO/PVA and GO/PMMA has been prepared [81] having a “layered morphology
comparable” to that of ‘graphene oxide paper’ [82]. The “dispersion of graphene
nanosheets in composite is usually controlled by level of exfoliation” before mixing
or during mixing in solution mixing methods. Thus, solution mixing offers a
9
potentially simple route to dispersing single-layer CMG platelets into a polymer
matrix [83].
In “melt mixing”, “a polymer melt and filler (in a dried powder form) are mixed
under high shear conditions. Relative to solution mixing, melt mixing is often
considered more economical (because no solvent is used) and is more compatible
with many current industrial practices [31].” To date, studies suggest that, such
“methods do not provide the same level of dispersion of the filler as solvent mixing
or in situ polymerization methods” [14]. A thermoplastic polymer is mixed
mechanically “with graphite or graphene or modified graphene at elevated
temperatures using conventional methods, such as extrusion and injection molding.
The polymer chains are then intercalated or exfoliated to form nano-composites.
This is a popular method for preparing thermoplastic nano-composites. Polymers,
which are unsuitable for adsorption or in-situ polymerization,” can be processed
using this technique. A wide range of polymer nanocomposites, such as PP/EG,
HDPE/EG, PPS/EG, PA6/EG, etc., have been prepared using this method. Notably,
“no means of dispersing single- or few-layer GO-derived fillers via melt mixing
without prior exfoliation have been reported akin to layered silicate fillers” [84].
1.3.2. Non-covalent/ in- situ polymerization
This method generally involves “mixing of filler in neat monomer (or multiple
monomers), or a solution of monomer, followed by polymerization in the presence
of the dispersed filler resultantly with precipitation/extraction or solution casting to
generate samples for testing. Covalent linkages between matrix and filler have been
reported in this particular method, however”, “non-covalent composites of a variety
of polymers, such as poly (ethylene) [85] PMMA [86] and poly (pyrrole) [87-88] in
situ polymerization has also been reported”. In situ polymerization high level of
10
dispersion of nano fillers obtained without a prior exfoliation step [84]. The
“intercalation polymerization has been widely investigated for nano clay/polymer
composites-monomer is intercalated between the layers of graphite or GO, followed
by polymerization to separate the layers, [84] which has been also applied to GNP
and GO-derived polymer composites. Graphite, GICs and EG can be exfoliated by
an alkali metal or monomer (e.g., isoprene or styrene), to generate dispersions of
GNPs in the matrix followed by polymerization initiated by the negatively charged
graphene sheets [89]. Anyhow isolation of monolayer graphene sheets yet to be
achieved through this method [24,69,90,91]. In a recent study, an attempt was made
to grow PE chains between the graphitic layers in the presence of graphene
nanosheets via polymerization” of poly (ethylene). Although polymerization
“further exfoliated the GNPs, but monolayer graphene platelets were not observed”
which was confirmed by TEM [85]. Monomers and polymers easily intercalates
into galleries of GO due to larger “interlayer spacing (between about 0.6 and 0.8
nm depending on relative humidity) compared to graphite” (0.34 nm) [92].
Promotion of direct intercalation of hydrophilic molecules takes place due to the
polar nature of graphene oxide, with the enlarged interlayer spacing [93]. “In situ
polymerization has been presented for various GO composite systems, including
poly (vinyl acetate) [94] and poly (aniline) (PANI) [95]”.
Intercalated morphology of these systems was confirmed by X-ray diffraction
studies where in polymer the graphene oxide sheets remain loosely stacked. Study
on GO/PMMA composite confirmed an enlarged interlayer spacing of GO (from
0.64 nm to 0.8 nm) which suggests intercalated morphology of system [86].
11
1.3.3. “Graphene-based composites with covalent bonds between matrix and
filler”
A covalent linkage “between the polymer matrix and pure carbon materials
surfaces (when used as composite filler) is challenging”. So, graphene is oxidized
to graphene oxide in order to produce functional groups on its surface for
introducing chemical bonding between polymer and nanofiller. Both “grafting-from
and grafting-to approaches” have been used for this purpose to have attachment of
nano filler and polymers. “Functionalized GeO platelets were introduced into
different polymers like of surface-attached poly (styrene), poly (methyl
methacrylate), or poly (butylacrylate) and improvement in mechanical and thermal
properties versus the neat matrix polymer were observed” [41,42,96-98]. In
grafting-to approaches azide-terminated poly(styrene) (PS) chains are grafted of to
alkynes-functionalized GeO platelets by using a cupric iodide(CuI-catalyzed) as
catalyst [99]. Similarly PVA is grafted to the surface of GeO platelets via
carbodiimide-activated esterefication [100]. The “grafting density of chains to the
platelet surface” [101] and its affect on dispersion of these polymer-grafted platelets
are the selection criteria for grafting to or grafting from approach if dispersed into a
polymer matrix [29] For certain polymers, prior functionalization is not required
because “covalent bonding between the matrix and GeO platelets may form during
polymerization (on reaction with the functional groups of GeO)”. For example in
case of an GeO/epoxy composite, GeO gets incorporated into cross linked network
when amine hardener is used as curator, [102] polyamide brushes are grafted to
GeO platelets during ring opening polymerization of caprolactam “via condensation
“reactions” between the amine containing monomer and the carboxylic acid groups”
of the GeO platelets [103].
12
1.3.4. Other methods for composite preparation
Several other methods like emulsion polymerizations, lyophilization methods [104]
or phase transfer techniques [105,106] may “offer general approaches to disperse
GeO platelets, CMG platelets and RGeO platelets as filler in a polymer matrix
[107-109] in addition to those mentioned above and which has potential use for
composite fabrication.” Non covalent grafting is one of such approaches of “well-
defined polymers to reduced graphene oxide (RGeO) platelets via pep interactions”.
For instance, the “attachment of pyrene-terminated poly(N-isopropylacrylamide) to
RGeO was recently reported; the composite was stated to retain the thermo-
responsive properties of the neat polymer [110]. This non-covalent grafting opened
a new horizon for the production of graphene based composite [111].” Moreover,
such “non-covalent composites may better preserve the conjugated structure of
graphene-based materials as compared with covalent functionalization or grafting
approaches”. “Attempts to exfoliate graphite directly via conventional melt mixing
techniques have not been successful to date [61]. However, solid state shear
pulverization, which uses a twin screw extruder to blend solid materials using
shear, was reported to exfoliate and disperse unmodified graphite directly into
polypropylene, yielding nanocomposites with platelets having thicknesses of
approximately 10 nm or less [112].” Other “production methods, such as layer-by-
layer assembly of polymer composite films [113] and backfilling of GO platelet
aero gel structures” with polymer may “provide means to produce nanocomposites
with defined morphologies” [114-115].
13
1.4. Graphene synthesis
Graphene was once believed to be an academic material [116], a
thermodynamically unstable material that upon isolation would crumple up on
itself. However, this didn’t stop some scientists investigating how thin they could
actually make graphite planes. The search for graphene started.
Many early attempts to make graphene involved intercalation. The technique
involved wedging the carbon planes apart and inserting various molecules between
them [117]. The end product usually consisted of thin graphitic chucks, or graphene
fragments rather than graphene monolayer. It wasn’t until 2004 when Geim and his
collaborators in Manchester [2], refined the micromechanical cleavage technique to
peel 10 µm sized, two-dimensional graphene from highly oriented pyrolytic
graphite (HOPG). Their research apparatus consisted of HOPG and scotch tape.
This tape was stuck on the graphite and peeled off repeatedly. The graphene were
detached from the scotch tape and pinned (by van der Waals forces) to a Si wafer
with a 300nm oxide layer specifically grown. The graphene was then imaged
optically showing visible contrast on the colourful oxide surface. Next the substrate
was etched to minimize induced effects, and graphene’s novel intrinsic properties
were probed. The results attracted the attentions of scientific community.
This original method to produce graphene is delicate and time consuming and is not
suitable for large scale applications like at industrial level. Developments of many
alternative syntheses have since become known. They can be broken into three
categories, i) growth in-situ on a substrate ii) bottom up methods to synthesize
graphene from organic precursors and iii) top down methods of liquid phase
exfoliation of graphite [118].
14
1.4.1. Growth in-situ on a substrate
Graphene mono- and multi- layers have been grown on single crystal silicon
carbide (SiC). This process involves heating the SiC to temperatures greater than
1000oC, in ultra high vacuum (UHV) conditions. Si desorbs carbon above this
temperature and thus small islands of graphitized carbon form. A significant
advantage of this technique is that SiC substrates offer an insulating supporting
medium. Few layer graphene that is produced this way can be patterned using
standard lithography techniques. However, it is challenging to achieve large
graphene domains with uniform thicknesses. Emtsev et al. have tried to overcome
this issue by an ex-situ graphitization of Si terminated SiC(0001) [119]. This
method produces undisturbed monolayer graphene terraces that are up to 3 µm wide
and ≥50µm in length.
SiC as a supporting medium has also been shown to have appreciable influence on
graphene’s electrical properties and so must not be made comparable to
mechanically cleaved graphene [120]. Zhou et al. found that the interaction between
the substrate and the epitaxially grown graphene results in gaps appearing at the
Dirac points. This can be exploited to induce a band gap, as seen in Figure 1.2 [121-
122]. Band-gap engineering is very encouraging for carbon based electronics. The
growth of mono- and few layer graphene on transition metals is also well
documented and has established itself as a promising means of producing graphene.
The procedure involves exposing the transition metal to a hydrocarbon gas, under
pressure.
15
Fig. 1.1 A) Schematic of graphene on a SiC substrate[122] B) E-k diagram of
graphene grown on SiC displaying the band opening at the Dirac point and C) An
ARPES intensity map, displaying the band gap.
This has been demonstrated on Pt [123], Ir [124], Ru [125], Cu [126] and on both
Ni single [127] and poly- crystalline [128-129] transition metals. There are lot of
requirements for the processing options, for example, high temperatures (~700-
1000oC) and UHV conditions, not to mention variables like cooling rates and gas
phase kinetics. The main issues with chemical vapour deposition (CVD) growth is
grain size limitations, which can result in grain boundaries (i.e. defects) and the
presence of multilayer that are not necessarily AB stacked. CVD growth is favored
for some electronic applications and may eventually lead to integrating graphene
into circuits. Standard lithographic technique can also be employed to pattern the
graphene grown films [129].
16
Another advantage of this processing method is the ability to transfer graphene to a
variety of substrates. CVD grown graphene, transferred onto SiO2, has been shown
to exhibit high electron mobility and even the half integer quantum Hall effect,
indicating that the quality can be as high as mechanically cleaved graphene [129].
At present the largest sheet (30 inch diagonal to diagonal) of CVD grown graphene
has been demonstrated by Bae et al [130]. This unique method involves using a 7.5
inch wide quartz tube wrapped in copper foils that is inserted into the 8 inch wide
furnace. After oven processing the graphene is transferred to an adhesive polymer
support and the copper is etched. The graphene films are then transferred from the
polymer support onto a target substrate by removing the adhesive forces (Figure
1.3). The resulting graphene films have set the bar for transparent conductive
electrodes with a sheet resistance of ~40Ω sq-1 and transparency (550nm) ~90%.
Fig. 1.2 Schematic of the roll based transfer of graphene films grown on Cu foil
[130].
1.4.2. Bottom up methods to synthesize graphene from organic
precursors
Bottom up synthetic approaches for benzene-based macromolecules have been
known for some time [131-132]. They are referred as polyaromatic hydrocarbons
(PAHs) and they lie between molecule and macromolecule structures. The
arrangement of the benzene ring is very similar to the 2-D chicken wire structure of
17
graphene and has thus attracted attention as a possible route for controlled growth
of graphene on substrates. PAHs are also attractive due to their high versatility,
clean processing and the multitude of aliphatic chains that can be attached to
modify their solubility [133]. These routes have been largely explored by Mullen
and co- workers who have produced a number of graphene precursors [133]. The
main disadvantage of increasing the molecular weight of these planar structures is
that their solubility in common solvents decreases, complicating their process
ability [132]. The core molecule in molecular graphene is the hexabenzocoronene
(HBC), which consists of 13 fused hexagon rings (Figure 1.4). This molecule
became the building block along with other hexaphenyl benzene derivatives. The
largest graphene molecules arranged to date has 222 carbon atoms in its core [134].
Further advances came in 2008, when Yang et al [135], demonstrated total
synthesis of graphene nano ribbons (GNRs) with controlled edge configuration. The
electrical properties of these GNRs were characterized by scanning tunneling
microscopy (STM), and thin films were prepared showing liquid crystal properties.
Furthermore, organic synthesis of graphene offers an alternative route to
synthesizing graphene with defined shape, size and edge structure, factors that are
quite important for applications in the field of electronics that require a finite band
gap and edges that allow spin transport.
18
Fig. 1.3 (A) Structure of the hexabenzocoronene (HBC) (B) Structure of the largest
polyaromatic hydrocarbons synthesized to data containing 222 C atoms
1.4.3. Chemical efforts to exfoliate and stabilize graphene sheets in
solution
Dispersing graphene in solution requires overcoming the cohesive energy of the
graphite planes [136]. To overcome this energy barrier, two main methods have
emerged. The first requires the chemical functionalization of graphite which aims to
weaken interlayer interactions [137] and the second involves the sonication of
untreated graphite in solvent [138] or surfactant systems [139].
The first approach results in graphite oxide (GO). This is a product from the
oxidation of graphite which retains the original layered structure of graphite [140].
The principle method to oxidize graphite is the Hummers method [137], and it
involves dispersing graphite in concentrated sulphuric acid, sodium nitrate and
potassium permanganate at 45oC for a few hours. The resulting graphite
intercalation compounds are then rapidly annealed, generating a CO2 over-pressure
that causes the graphite to split. Further ultrasonication results in individual GO
sheets separation. These GO sheets contain large quantities of hydroxyl, carboxyl,
19
carbonyl and epoxides functional groups which are attached to the edge or basal
planes [141]. Undesirably during the oxidation processing, the carbon atom is
transformed from planar sp2 hybridized geometry to distorted sp3 hybridized
geometry, thus losing its electrical properties to become electrically insulating as
shown in Figure 1.5. Hydrazine or hydrogen plasma reduction is used to restore the
electrical conductivity of graphene.
Fig. 1.4 GO structure demonstrating the distorted sp2 atomic arrangement and
attached functionalities (left) and vial of GO brown dispersion (right). Its brown
colour is attributed to the absence of π conjugated structure.
GO is its strongly hydrophilic due to the presence of various functional groups like
hydroxyl, carbonyl etc. and can be readily exfoliated in water to form stable
colloidal dispersions at 2 mg/ml [142]. Further dispersion analysis confirmed that
single layer GO sheets, up to hundreds of microns in size, have been dispersed in a
variety of organic solvents at concentrations higher than ~1.5mg/ml [143-145].
Ang et al. [146] obtained stable dispersions with 90% monolayer yield and mean
sheet areas of 330 ± 10μm2. They explained that intercalated GO sediments formed
after oxidation, via a modified Hummers method, result in oxidized outer layers of
the large sized GO aggregates, but the inner layers consist of mildly oxidized
20
(mainly at the edge planes) graphene sheets. These sediments were then intercalated
using tetra butyl ammonium hydroxide, (40%TBA water solution) under reflux
conditions for two days. After two days the color changes from pale yellow to black
indicating an increase in UV-Vis absorption region due to the presence of extended
conjugate π structure [147]. Then they are dispersed in dimethyl formamide (DMF)
and spin coated onto SiO2. Their XPS data suggests that less that 10% of the carbon
remains oxidized and a conductivity of 15,000 S m-1 was achieved. Dikin et al.
prepared free standing (1 to 30 µm thick) GO paper showing a mean Young’s
modulus of 32GPa and ultimate tensile strength of 60MPa [148]. These results are
greater than most of the reported nanotubes bucky papers [149].
Despite increased process ability of graphene oxide, it retains significant amounts
of oxygen functionalities even after severe reduction processes and can contain
irreversible lattice defects [150]. In comparison to pristine graphene derived from
expanded graphite, it fails to meet the high electrical conductivities due to distorted
sp2 structure and contains many lattice defects [151-153].
The second approach to overcome the forces that bind graphene layers together is
liquid phase exfoliation of pristine graphite in solvent and surfactant systems.
Solvent exfoliation of graphite has been demonstrated by Hernandez et al. [138] at
concentrations of up to 0.01mg/ml. Other groups have demonstrated concentrations
of between 0.05 - 0.1mg/ml [154-156]. Surfactant exfoliated graphene have also
reached concentrations of 0.04mg/ml [139]. Thin films were prepared from these
graphene/surfactant dispersions, shows conductivity of 1.5 x 104S m-1 with a
transparency of ~70% after annealing [157].
21
Liquid phase exfoliation has many advantages including a straightforward approach
that is readily accessible with a low cost. Successful dispersions can be directly
used for mixing or blending with polymers, spin or dip coating, spraying or even
post functionalization. They are also easily analyzed by TEM, can be cast by
filtration or can be made into large thin films by Langmuir–Blodgett assembly in a
layer-by-layer manner [158]. The main drawback, however, is the lack of control
over the exfoliation of the dispersion which can vary considerably from starting
graphite to the method used to exfoliate them [143]. This can result in poly-disperse
dispersions, with flakes of many thicknesses and sizes. It is well understood that
graphene sheets consisting of 10 or less layers, possess electronic structure distinct
from bulk graphite [5,114]. Graphene’s properties also vary as a function of layer
number. As mentioned earlier mono- layer graphene is a zero gap semiconductor,
with linear energy dispersion. Bilayer graphene is also a zero gap semiconductor
but its electrons follow a parabolic energy spectrum [5,158]. Trilayer graphene’s
electronic spectrum becomes even more complicated as several charges appear and
the bands overlap [2, 11, 4, 1, 16]. Thus polydispersity of flakes within dispersions
can result in unpredictable behaviors. To improve this, a post sonication
centrifugation step results in larger graphite pieces sediment to the bottom of the
centrifuge tube. The top percentage of the dispersion is decanted and used for
further analysis. Ultracentrifugation in a density gradient medium has also been
demonstrated [159]. This separates graphene sheets according to their buoyant
density and has produced mono-disperse graphene dispersions.
Solvent exfoliation of graphene is not completely understood. Coleman et al.
explains why solvents exfoliate carbon nanotubes [160-161]. The main factor in
exfoliating nanotubes is the strength of the solvent-nanotubes sidewall interaction.
22
One wants to match the surface energy of the solute (graphene) to the surface
energy of the solvent. This results in minimal energetic stress between the two
species and is the basis of the chemistry rule, “like dissolves like”. Specific solvents
that result in favorable interactions exfoliate and stabilize materials more easily.
Work done by Hernandez et al. investigated if this rule is also true for graphene and
thus the solubility parameters for graphene were determined [162]. The multi
component solubility parameters are numerical values that indicate the relative
solvency behavior of a specific solvent. When graphene concentration was plotted
as a function of these multi component solubility parameters it confirmed that
successful solvents show a sharp dependence on surface tension. The dispersibility
of graphene in 40 solvents, (28 of them previously unreported) was measured. It
was found that good solvents for graphene are characterised by a Hildebrand
solubility parameter T~23MPa1/2. Specific physical interactions between the
solvent and graphene were subsequently investigated using Hansen parameters.
These can be related to the Hildebrand parameter, T, through T2 = 2
D + 2P + 2
H
(where D, P and H refer to the dispersive, polar and H-bonding Hansen
components). The effectiveness of the studied solvents was shown to scale with
proximity and to the calculated Hansen solubility parameters of graphene (
D~18MPa1/2, P~9.3MPa1/2 and H~7.3MPa1/2) [160-162]. TEM analysis was used
to show that the graphene is well exfoliated in all cases. Even in relatively poor
solvents, >63% of observed flakes have <5 layers.
23
Refrences
[1] E. Fradkin, Physical Review B 33 (1986) 3263.
[2] A. Geim and P. Kim, Carbon wonderland. Scientific American 298 (2008) 90.
[3]M.I. Katsnelson and K.S. Novoselov, Solid State Communications 143 (2007) 3
[4] Zhu Y, Murali S, Cai W, Li X, Suk JW, Potts JR, et al. Adv Mater 22 (2010)
3906.
[5] Geim AK, Novoselov KS. Nat Mater 6 (2007) 183.
[6] Compton OC, Nguyen SBT. Small 6 (2010)711.
[7] T. Booth, et al, Nano Letters 8 (2008) 2442.
[8]. C. Lee, et al, Science 321 (2008) 385.
[9] Suk JW, Piner RD, An J, Ruoff RS. ACS Nano; 2010. doi:10.1021/nl101902k.
[10] Gómez-Navarro C, Burghard M, Kern K. Nano Lett 8(2008) 2045.
[11] Alexandre M, Dubois P. Mater Sci Eng R Rep 28(2000)1.
[12] Thostenson ET, Li CY, Chou TW. Compos Sci Technol 65(2005)491.
[13] Li D, Kaner RB. Science 320(2008)1170
[14] Kim H, Miura Y, Macosko CW. Chem Mater 22 (2010)3441
[15] Paci JT, Belytschko T, Schatz GC. J Phys Chem C 111(2007) 18099
[16] Fornes TD, Paul DR. Polymer 44(2003)4993
[17] Carter LW, Hendricks JG, Bolley DS. 2531396, National Lead Company;
1950
[18] Usuki A, Kojima Y, Kawasumi M, Okada A, Fukushima Y, Kurauchi T, et al.
J Mater Res 8 (1993)1179
[19] Bunnell LR. 5186919, Battelle Memorial Institute; 1993.
24
[20] Pan YX, Yu ZZ, Ou YC, Hu GH. J Polym Sci Part B Polym Phys 38
(2000)1626
[21] Winey KI, Vaia RA. MRS Bull 32(2007)314.
[22] Ramanathan T, Abdala AA, Stankovich S, Dikin DA, Herrera-Alonso M,
Piner RD, et al. Nat Nanotechnol 3 (2008)327.
.[23] Chen G, Zhao W,editors. Nano- and biocomposites. CRCPress; 2009. p. 79.
[24] Chen G, Weng W, Wu D, Wu C. Eur Polym J 39 (2003)2329
[25] Chen G, Wu C, Weng W, Wu D, Yan W. Polymer 44 (2003)1781.
[26] Zheng W, Wong S-C. Compos Sci Technol 63 (2003)225
[27] Zheng W, Wong S-C, Sue H-J. Polymer 43: (2002) 6767
[28] Schaefer DW, Justice RS. Macromolecules 40 (2007)8501
[29] Akcora P, Kumar SK, Moll J, Lewis S, Schadler LS, Li Y, et al.
Macromolecules 43 (2010)1003
[30] Akcora P, Liu H, Kumar SK, Moll J, Li Y, Benicewicz BC, et al. Nat Mater
8 (2009)354
[31] Paul DR, Robeson LM. Polymer 49 (2008)3187
[32] Kluppel M, editor. Advances in Polymer Science 164 2003) 1.
[33] C. Harper, Handbook of Plastics, Elastomer, and Composites, McGraw-Hill, Inc, New
York (2002).
[34] Lv C, Xue Q, Xia D, Ma M, Xie J, Chen H. J Phys Chem C 114 (2010)6588.
[35] Wagner HD, Vaia RA. Mater Today 7 (2004)38
[36] Liang JJ, Huang Y, Zhang L, Wang Y, Ma YF, Guo TY, et al. Adv Funct
Mater
19 (2009) 2297
[37] Jiang L, Shen XP, Wu JL, Shen KC. J Appl Polym Sci 118(2010) 275.
[38] Yang XM, Tu YF, Li LA, Shang SM, Tao XM. ACS Appl Mater Interfaces
25
2 (2010); 1707
[39] Xu Z, Gao C. Macromolecules 43 (2010) 6716
[40] Miller SG, Bauer JL, Maryanski MJ, Heimann PJ, Barlow JP, Gosau JM, et al,
Adherent Technologies,Inc. Composite Science and Technology 2010.
[41] Fang M, Wang KG, Lu HB, Yang YL, Nutt S. J Mater Chem 19(2009) 7098
[42] Goncalves G, Marques PAAP, Barros-Timmons A, Bdkin I, Singh MK,
Emami N, et al. J Mater Chem 20 (2010) 9927
[43] Lee YR, Raghu AV, Jeong HM, Kim BK. Macromol Chem Phys 210 (2009);
1247
[44] Pramoda KP, Hussain H, Koh HM, Tan HR, He CB. J Polym Sci Part A Polym
Chem 48 (2010)4262.
[45] Cai DY, Yusoh K, Song M. Nanotechnology 20 (2009) 085712.
[45] Lee HB, Raghu AV, Yoon KS, Jeong HM. J Macromol Sci Part B Phys
49 (2010) 802
[46] Nguyen DA, Lee YR, Raghu AV, Jeong HM, Shin CM, Kim BK. Polym Int
58 (2009)412.
[47] Raghu AV, Lee YR, Jeong HM, Shin CM. Macromol Chem Phys 209 (2008);
2487
[48] Bansal A, Yang H, Li C, Benicewicz BC, Kumar SK, Schadler LS. J Polym
Sci Part B Polym Phys 44 (2006) 2944.
[49] Khan U, May P, O’Neill A, Coleman JN. Carbon 48 (2010)4035.
[50] Quan H, Zhang B-Q, Zhao Q, Yuen RKK, Li RKY. Compos Part A Appl Sci
Manuf 40 (2009)1506
[51] NguyenDA, RaghuAV,Choi JT, JeongHM. PolymPolymCompos18 (2010)351
[52] Liu H, Brinson LC. Compos Sci Technol 68(2008)1502.
[53] Kim S, Drzal LT. J Adhes Sci Technol 23 (2009)1623
26
[54] Kalaitzidou K, Fukushima H, Drzal LT. Carbon 45 (2007)1446
[55] Kim S, Do I, Drzal LT. Polym Compos 31 (2010)755
[56] Kalaitzidou K, Fukushima H, Drzal LT. Compos Part A Appl Sci Manuf 38
(2007)1675
[57] Kalaitzidou K, Fukushima H, Drzal LT. Compos Sci Technol 67(2007)2045
[58] Ramanathan T, Stankovich S, Dikin DA, Liu H, Shen H, Nguyen ST, et al.
J Polym Sci Part B Polym Phys 45 (2007)2097
[59] Hirata M, Gotou T, Horiuchi S, Fujiwara M, Ohba M. Carbon42( 2004)2929
[60] Zhao X, Zhang QH, Chen DJ, Lu P. Macromolecules43 (2010)2357
[61] Kim H, Macosko CW. Macromolecules 41(2008) 3317
[62] Kim H, Macosko CW. Polymer 50 (2009)3797
[63] Li Q, Li ZJ, Chen MR, Fang Y. Nano Lett 9 ( 2009);9:2129
[64] Rafiee MA, Rafiee J, Wang Z, Song HH, Yu ZZ, Koratkar N. ACS Nano 3
(2009) 3884
[65] Podall H, Foster WE, Giraitis AP. J Org Chem 30 (1958)82
[66] Panayotov IM, Rashkov IB. J Polym Sci Part A Polym Chem 11(1973)2615.
[67] Parrod J, Beinert G. J Polym Sci 53 (1961) 99.
[68] Shioyama H. Synth Met 114 (2000)1
[69] Shioyama H. Carbon 35(1997)1664.
[70] Moniruzzaman M, Winey KI. Macromolecules39( 2006)5194.
[71] Stankovich S, Dikin DA, Dommett GHB, Kohlhaas KM, Zimney EJ, Stach
EA, et al. Nature 442(2006)282
[72] Higginbotham AL, Lomeda JR, Morgan AB, Tour JM. ACS Appl Mater
Interfaces 1(2009)2256
[73] Pandey R, Awasthi K, Tiwari RS, Srivastava O.N. Polymer 52 (2011) 5.
[74] Chen D, Zhu H, Liu T. ACS Appl. Mater. Interfaces; 2 (2010) 3702.
27
[75] Das B, Prasad KE, Ramamurty U, Rao CNR. Nanotechnology
20(2009)125705.
[76] Stankovich S, Piner RD, Nguyen ST, Ruoff RS. Carbon 44 (2006)3342
[77] Yang X, Li L, Shang S, Tao X. Polymer 51(2010)3431.
[78] Xu YX, Hong WJ, Bai H, Li C, Shi GQ. Carbon 47(2009)3538
[80] Satti A, Larpent P, Gun’ko Y. Carbon 48(2010):3376
[81] Putz KW, Compton OC, Palmeri MJ, Nguyen SBT, Brinson LC. Adv Funct
Mater 20(2010)3322
[82] Dikin DA, Stankovich S, Zimney EJ, Piner RD, Dommett GHB, Evmenenko
G, et al. Nature448( 2007) 457
[83] Cao Y, Feng J, Wu P. Carbon48( 2010)3834
[84] Sinha Ray S, Okamoto M. Prog Polym Sci 28(2003)1539
[85] Fim FC, Guterres JM, Basso NRS, Galland GB. J Polym Sci Part A Polym
Chem 48 (2010):692
[86] Jang JY, Kim MS, Jeong HM, Shin CM. Compos Sci Technol69( 2009)186.
[87] Gu Z, Zhang L, Li C. J Macromol Sci Part B Phys 48 (2009):1093
[88] Gu ZM, Li CZ, Wang GC, Zhang L, Li XH, Wang WD, et al. J Polym Sci Part
B Polym Phys 48 (2010):1329
[89] Novoselov KS, Geim AK, Morozov SV, Jiang D, Zhang Y, Dubonos SV, et al.
Science 306 (2004)666
[90] Gabriel P, Cipriano LG, Ana JM. Polym Compos 20 (1990)804.
[91] Liu P, Gong K. Carbon 37(1999)701
[92] Jang BZ, Zhamu A. J Mater Sci43 (2008)5092
[93] Matsuo Y, Hatase K, Sugie Y. Chem Mater 10(1998)2266
[94] Liu P, Gong K, Xiao P, Xiao M. J Mater Chem 10(2000)933.
[95] Kyotani T, Moriyama H, Tomita A. Carbon 35(1997)1185
28
[96] Lee SH, Dreyer DR, An JH, Velamakanni A, Piner RD, Park S, et al.
Macromol Rapid Commun 31(2010)281
[97] Layek RK, Samanta S, Chatterjee DP, Nandi AK. Polymer51 (2010)5846.
[98] Fang M, Wang KG, Lu HB, Yang YL, Nutt S. J Mater Chem 20(2010)1982
[99] Sun ST, Cao YW, Feng JC, Wu PY. J Mater Chem 20 (2010) 5605.
[100] Veca LM, Lu FS, Meziani MJ, Cao L, Zhang PY, Qi G, et al. Chem
Commun;(2009):2565
[101] Coleman J, Khan U, Gun’ko Y. Adv Mater18 (2006)689
[102] Yang H, Shan C, Li F, Zhang Q, Han D, Niu L. J Mater Chem 19(2009)8856
[103] Zhang M, Parajuli RR, Mastrogiovanni D, Dai B, Lo P, Cheung W, et al.
Small 6 (2010) 1100
[104] Dikin DA, Stankovich S, Zimney EJ, Piner RD, Dommett GHB, Evmenenko
G, et al. Nature448( 2007) 457
[105] Choi EY, Han TH, Hong J, Kim JE, Lee SH, Kim HW, et al. J Mater Chem
20 (2010) 1907.
[106] Wei T, Luo GL, Fan ZJ, Zheng C, Yan J, Yao CZ, et al. Carbon47(2009)
2296
[107] Hu HT, Wang XB, Wang JC, Wan L, Liu FM, Zheng H, et al. Chem Phys
Lett 484 (2010)247
[108] Zheming G, Ling Z, Chunzhong L. J Macromol Sci Part B Phys48( 2009)226
[109] Tkalya E, Ghislandi M, Alekseev A, Koning C, Loos J. J Mater Chem 20
(2010)3035
[110] Liu JQ, Yang WR, Tao L, Li D, Boyer C, Davis TP. J Polym Sci Part A
Polym Chem 48 (2010)425
[111] Liu JQ, Tao L, Yang WR, Li D, Boyer C, Wuhrer R, et al. Langmuir26
(2010) 10068
29
[112] Vickery JL, Patil AJ, Mann S. Adv Mater 21(2009)2180
[113] Wu JH, Tang QW, Sun H, Lin JM, Ao HY, Huang ML, et al. Langmuir24
(2008) 4800
[114] Wang J, Ellsworth MW. ECS Trans 19(2009)241
[115] B. Partoens and F. Peeters, Physical Review B 74 (2006) 75404.
[116] E. Fradkin, Physical Review B 33 (1986) 3263.
[117] M. Dresselhaus and G. Dresselhaus., Advances in Physics 30 (1981) 139.
[118] M. Allen, V. Tung and R. Kaner, Chemical Reviews 110 (2010) 132.
[119] Emtsev, K.V., et al., Nature Materials 8 (2009) 203.
[120] W. de Heer, et al, Solid State Communications 143 (2007) 92.
[121] K. Novoselov, Nature materials 6 (2007) 720.
[122] S. Zhou, G. Gweon and A. Fedorov, Nature materials 6 (2007) 770.
[123] H. Ueta, et al, Surface Science 560 (2004) 183.
[124] C. Busse, et al, New Journal of Physics 11 (2009) 22.
[125] P. Sutter, J. Flege and E. Sutter, Nature materials 7 (2008) 406.
[126] X. Li, et al, Science 324 (2009) 1312.
[127] S. Kumar, et al, Chemical communications 46 (2010) 1422.
[128] A. Reina, et al, Nano Letters 9 (2008) 30.
[129] K.S. Kim, et al, Nature 457 (2009) 706.
[130] S. Bae, et al, Arxiv preprint arXiv: 0912 (2009) 5485.
[131] I. Gutman and S. Cyvin, Introduction to the theory of benzenoid hydrocarbons.
Springer (1989)
[132] E. Clar, The aromatic sextet. John Wiley & Sons (1972).
[133] J. Wu, W. Pisula and K. Mullen, Chemical Reviews 107 (2007) 718.
[134] C. Simpson, et al, Chemistry -Weinheim-European journal 8 (2002) 1424.
[135] X. Yang, et al, Journal of the American Chemical Society 130 (2008) 4216.
[136] S. Niyogi, et al, Journal of the American Chemical Society 128 (2006) 7720.
30
[137] W. Hummers Jr and R. Offeman., Journal of the American Chemical Society 80
(1958) 1339.
[138] Y. Hernandez, et al, Nature Nanotechnology 3 (2008) 563.
[139] M. Lotya, et al, Journal of the American Chemical Society 131 (2009) 3611.
[140] H. He, et al, Journal of Physical Chemistry 100 (1996) 19954.
[141] Y. Geng, S. Wang and J. Kim, Journal of Colloid And Interface Science 336
(2009) 592.
[142] Y. Si and E. Samulski, Nano Letters 8 (2008) 1679.
[143] A. Green and M. Hersam, The Journal of Physical Chemistry Letters 1 (2010) 544.
[144] V. Tung, et al, Nature Nanotechnology 4 (2009) 25.
[145] S. Park, et al, Nano Letters 9 (2009) 1593.
[146] P. Ang, et al, ACS Nano 3 (2009) 3587.
[147] D. Li, et al, Nature Nanotechnology 3 (2008) 101.
[148] D.A. Dikin, et al, Nature 448 (2007) 457.
[149] M.F. Yu, et al, Physical review letters (Copyright (C) 2010 The American
Physical Society) 84 (2000) 5552.
[150] S. Park and R. Ruoff, Nature Nanotechnology 4 (2009) 217.
[151] S. Stankovich, et al, Carbon 45 (2007) 1558.
[152] S. Stankovich, et al, Journal of Materials Chemistry 16 (2006) 155.
[153] R. Hao, et al, Chemical Communications 48 (2008) 6576.
[154] P. Blake, et al, Nano Letters 8 (2008) 1704.
[155] Bourlinos, A.B., et al., Small 5 (2009) 1841.
[156] S. De, et al, Small 6 (2010) 458.
[157] X. Li, et al, Nature Nanotechnology 3 (2008) 538.
[158] E. McCann, D.S.L. Abergel and V.I. Fal'ko, Solid State Communications 143
(2007) 110.
[159] A.A. Green and M.C. Hersam, Nano Letters 9 (2009) 4031.
31
[160] S. Bergin, et al, ACS Nano 3 (2009) 787.
[161] J.N. Coleman, Advanced Functional Materials 19 (2009) 3680.
[162] Y. Hernandez, et al, Langmuir 26 (2010) 3208.
32
Chapter 2
Materials and Characterization
Techniques
2.1 Introduction
This chapter discusses the materials used and outlines of the characterization
techniques.
2.2 Materials Graphite (Sigma-Aldrich), 1-Methyl-2-Pyrrolidinone (Fluka), N,N,di-methyl
formamide (Sigma-Aldrich), Tetrahydrofuran (Sigma-Aldrich), Poly (vinyl acetate)
(Sigma-Aldrich), Poly (vinyl alcohol) (Sigma-Aldrich), Poly (vinyl chloride)
(Sigma-Aldrich), Poly (acrylonitrile) (Sigma-Aldrich), Poly(urethane) (Hauntsman).
Octdecyl amine (Sigma-Aldrich). Sulfuric acid (Sigma-Aldrich), Hydrochloric acid
(Sigma-Aldrich). Sodium nitrite (Sigma-Aldrich), Potassium permangate (Fluka).
Hydrogen peroxide (Sigma-Aldrich).
These materials were used as received without further purification
2.3 Apparatus and Equipments Ultra Sonication Tip (Make- Vibra Cell- VCX 500, Power- 500 watt, Frequency-
20KHz), Sonication Bath (Make- Bransonic- 1510E MT), Centrifugation Machine
(Make- Hettich Zentrifugen – Mikro 220R, D-78532), Vaccum Pump (Make- Buchi
Switzerland, V-700). Magnetic Stirrer (Make- Heidolph, MR-3002), Oven (Make-
MTI Corporation), Teflon Trays (length x breadth x height, 4x4x1 cm)
33
2.4 Characterization Techniques
2.4.1 UV-vis Absorption Spectroscopy
UV-Vis absorption spectroscopy involves exciting a sample with electromagnetic
radiation (EM radiation) of a certain wavelength and measuring the proportion of
radiation that is absorbed by the material. When EM radiation is pointed on a
material, the radiation excites a bonding electron in an atom or molecule into an
unfilled non-bonding orbital (or promoting the electron from the ground state into an
excited state). The change in energy acquired by the electron relates to a line in the
absorption spectrum which occurs at a characteristic wavelength (or energy). As
each electronic transition has associated rotational and vibrational transitions, the
line is broadened to become a peak centered on the characteristic wavelength. The
intensity of the absorption at the wavelength is related to how much energy is
absorbed by the molecule.
UV-Vis spectrometers generally use a broad excitation source such a xenon (Xe)
lamp along with a mono chromator in order to illuminate the sample across a range
of excitation wavelengths to measure the absorbed light as a function of the
wavelength. In this work a Cary 6000i spectrophotometer was used which can
measure absorbance from 350 nm to 850 nm (i.e. in the UV-Visble region). The
spectrometer is run in dual beam mode which means that the exciting radiation is
split into two equal intensity beams using a half-mirror so that simultaneous
measurement can be made on a sample and a background or reference sample for
accurate background subtraction. The intensities of the reference and sample
radiation are measured as I0 and I respectively. The ratio of I to I0 is called the
transmittance, T. The Beer-Lambert law empirically relates T to the length, l of the
sample and the concentration, c of the absorbing species as follows:
T I 10
cl (2.1)
I 0
Where ε is known as the molar extinction coefficient and is unique for different
materials.
34
In UV-Vis absorption spectroscopy, the Absorbance, A is usually the parameter used
instead of Transmittance. The Absorbance is defined as:
A = log10 Io/I = -log10 T = αCl, (2.2)
Where C is concentration, l is the path length and α is the absorption coefficient.
For liquid samples, if the path length and absorption coefficient of a sample is
known then the concentration can be calculated from the measured absorbance.
2.4.2 Raman Spectroscopy
Raman Spectroscopy is a form of vibrational spectroscopy and is a measure of the
inelastic scattering of light by molecules. The Raman Effect was first observed
experimentally by Raman and Krishnan in 1928 when they used sunlight and a
narrow band filter to pass monochromatic light through a number of liquids. When a
crossed filter was used to block the wavelength of the incident light after passing
through the liquid, light of a different frequency passed through the filter [1]. This
effect had been predicted previously but this paper was the first observations of what
became known as the Raman Effect. Over the next number of years, further
investigations into this effect continued and as the quality of the light sources
available increased from sunlight to mercury arc lamps and then onto lasers, the
Raman effect became more widely used as a spectroscopic tool to help identify
chemical bonds and molecules [2].
When light is shone onto a sample, the photons interact with the molecules of the
sample. The photons may be reflected, absorbed or scattered. The vast majority of
scattered photons are scattered elastically, i.e. the scattered photons have the same
wavelength and hence energy as the incident photons. This type of scattering is
known as Raleigh scattering. However, a small number of photons (generally less
than 1 in 10-7
) is scattered in elastically with a wavelength that differs to that of the
incident photons. This occurs when the incident photons interact with the electron
cloud and bonds of the molecule. In the Raman Effect, the incident photon excites
35
the molecule into a virtual excited state. The molecule then relaxes into a different
rotational or vibrational energy state by emitting a photon of a different energy to
that of the incident photon. This difference in energy between the incident photon
and the emitted photon is the Raman shift and is usually expressed as a change in
frequency in wave numbers [3-4].
As the overall energy of the system must remain constant, if the final energy state of
the molecule is higher than the initial state, then the Raman scattered photons must
have a lower energy (and hence lower frequency) than the incident light. This shift in
frequency is known as a Stokes shift. Similarly if the molecule’s final energy is
lower than its initial energy, then the Raman scattered photons have a higher energy
than the incident photons and the change in frequency is known as an anti-Stokes
shift. At room temperature, the majority of the molecules are likely to be in their
ground state energy levels and as such, Stokes Raman scattering is much larger than
anti-Stokes Raman scattering.
Raman spectroscopy is similar to infrared spectroscopy in that they both probe the
vibrational energies of molecules – in other words, the nature of the bonds between
the atoms of the molecules. However, whereas IR spectroscopy can only measure
vibrations which cause the dipole moment of the molecule to change, for a transition
to be Raman active, the polarisability of the molecule must be changed by the
transition. In this way we can say that IR and Raman spectroscopy are
complementary to each other.
In this research, Raman spectroscopy is used to characterize graphene nanosheets.
Horiba Jobin Yvon Lab Ram HR spectrometer is utilized by using a 633 nm HeNe
laser with laser powers up to 12 mW to excite the samples. A long working distance
100x objective lens and a diffraction grating of 600 lines mm-1
give spatial
resolutions of 3 – 5 cm and a quoted frequency resolution of 0.3 cm-1
. The 633 nm
light interacts well with electronic transitions of the materials investigated (i.e. C-C,
C-H bonds etc.). For this reason, Raman spectroscopy like this is also known as
resonance Raman spectroscopy.
36
The Raman spectrum of graphene is dominated by three main features,G, D, and 2D-
Raman modes each having different physical origins. The peak at 1580 cm-1 (G
band), arising from emission of zone-centre optical phonons, corresponds to the
doubly degenerate E2g mode of graphite related to the vibration of sp2 bonded carbon
atoms. The disorder-induced D (1350 cm-1) band and its symmetry –allowed 2-D
overtone band (2700 cm-1) involve preferential coupling to transverse zone-boundary
optical phonons. The “D band gives evidence of the presence of defects, that is,
either edges or topological defects in the population. We can quantify the defect level
by D-to-G- band intensity ratio, (ID/IG). As shown in the inset of Figure 2.1, ID/IG
increases gradually from the powder value with increasing” “sonication time. In
addition, we found that (ID/IG) increase smoothly with rotation rate”. [5]
Figure. 2.1 Raman shift
2.4.3 Tensile testing (TT)
Tensile testing is the most widely used tool to investigate the mechanical properties
of materials. Mechanical properties were measured using “Zwick Roell tensile tester”
with 100 N load cell. In this thesis we have monitored stress-strain behavior of, PVC
(polyvinylchloride), polyurethane (Morethane), polyvinyl alcohol (PVA), poly
(acrylonitrile) (PAN), polyvinyl acetate (PVAc) and graphene based polymeric
composites. A sample is clamped in the jaws/head of a tensile tester. One of the jaws
moves continuously against a static jaw to stretch the sample (until break). Applied
force/stress is plotted as a function of strain. Such a plot is termed as stress-strain
curve. Strain (ε) is given by equations 2.3 and 2.4.
37
L1 L0
(2.3)
L0
Or
L
(2.4)
L0
Where, L L1 L0
Where L0 is the initial length while L1 is the final length of the sample.
Stress (σ) at a point is defined as the applied force per cross sectional area and given
by equation 2.5:
Force
A where A is cross sectional area and in case of a film sample, is given
as A W Tk
W is width of sample and Tk is thickness of sample.
(2.5)
(2.6)
38
Fig. 2.2 Typical stress-strain curves.
Various regions in the stress-strain plot have been shown in Fig. 2.2 which are the characteristic of various materials.
1. The initial part of stress-strain curve is usually linear (see Fig 2.2). This part
of the curve is generally known as the elastic or proportional region. During
stretching of a sample the inter-atomic/molecular bonding distance slightly
increases elastically. Therefore, the Young’s modulus is a measure of inter
atomic/molecular forces. The linearity in a stress-strain curve also represents
the degree of order in a material which is why crystalline solids have linear
elastic regions compared to amorphous solids. In some cases, the initial
elastic part of a stress-strain curve may not be linear because of lower inter
atomic/molecular attractive forces and lesser or no order. For this behavior
the secant modulus is usually used. A secant is drawn from the origin to some
point of stress-strain curve and the slope is taken as secant modulus. In the
elastic region, stress is increasing proportionally with increased strain or vice
versa. This is because of constant strain of the polymer chain and filler in this
39
part. The deformation in this region is reversible. The slope of the linear
region is called Young’s modulus. This is a parameter for measuring a
material’s stiffness. The area under the elastic region of the stress strain curve
is usually termed as the resilience.
2. The point at which stress-strain curve no longer remains linear and an
increase in strain occurs without an increase in stress is called the yield point
and the stress at that point is called the yield point stress or yield stress [5].
3. As stress is further increased beyond the elastic limit, the material starts to
deform irreversibly. The region can be relatively flat (depending on the
material). This part of the curve is called the viscous or plastic part. It
continues until the material breaks. The strain at which the material breaks is
termed as the strain at break (εB ) and stress at that point is known as the
strength at break and denoted as (σB.). The highest stress value in stress-strain
curve is called the ultimate tensile strength (UTS). Strength at break (εB ) and
UTS may or may not be the same. If a material breaks at the UTS point both
UTS and σB will be the same, while if a material breaks at a lower stress
value than UTS the UTS and σB will be different (Fig 2.1). The area under
the stress-strain curve is called strain energy or material toughness.
2.4.4 Scanning Electron Microscopy Scanning electron microscopy (SEM) is a powerful tool used to help image the
surface of samples using a beam of electrons instead of light. First used in the 1930s
it has since become a well known, well established technique both in scientific
research and in industry. In SEM an accelerated beam of electrons is focused onto
the surface of a sample under vacuum using a series of electromagnetic lenses. The
beam is then rastered across the surface using a series of coils. The surface is imaged
using a range of different detectors depending on how the electron beam interacts
with the sample surface. Figure 2.3 shows a schematic of the electron beam through
the SEM before it interacts with the sample.
When the beam hits the sample, several different interactions take place depending
on the energy of the electron beam and the nature of the sample substrate. When the
40
beam (consisting of primary electrons) hits the surface, secondary and backscattered
electrons are dislodged from the surface of the sample. They are collected by
detectors consisting of positively charged grids, converted to digital signals and
converted to an image. The primary electrons interact with the sample in a teardrop
shape that extends from between 100 nm to 5 nm into the surface depending on the
beam energy and surface state. Secondary electrons are generated by inelastic
scattering of primary electrons on the atomic core or inner core electrons of atoms on
the surface of the sample and used in the most common imaging mode which uses
the In Lens detector whose position in the chamber is shown in Figure 2.3 Secondary
electrons detectable with the In Lens detector have a low penetration depth and
images formed are very surface sensitive. Electrons that are given off with larger
energies are more commonly back scattered electrons or electrons that have
undergone more interactions with the substrate and have undergone several
scattering events with the surface. They have travelled deeper into Backscattered
electrons carry information on chemical composition as materials with higher atomic
number are better scatterer and hence appear brighter images. the sample and are
detected using either backscattered electron (BSE) detector or an SE2 detector as
shown in Figure 2.4. In the research discussed in the course of this thesis, one of
three different SEMs manufactured by Carl Zeiss Ltd have been used: a Zeiss Supra
variable pressure FE SEM; a Zeiss Ultra Plus FE SEM and a Zeiss Auriga Focused
Ion Beam SEM. Each of these SEMs has InLens detectors and SE2 detectors along
with a range of other detectors.
41
Fig. 2.3 Schematic diagram of SEM showing the electron path [6]
42
Fig. 2.4 Detector positions in Zeiss Ultra / Supra [7]
2.4.5 Transmission Electron Microscopy
TEM working is similar to a slide projector. A “projector shines a beam of light
through (transmits) the slide, as the light passes through it is affected” by the
structures and objects on the slide. “These effects result in only certain parts of
the light beam being transmitted through certain parts of the slide”.”This
transmitted beam is then projected onto the viewing screen, forming an enlarged
image of the” slide.
TEMs work the “same way except that they shine a beam of electrons (like the
light) through the specimen (like the slide). Whatever part is transmitted is
projected onto a phosphor screen for the user to see”. Here is another more
scientific explanation of TEM and its working:
43
1. The "Virtual Source" at the top represents the electron gun, producing a
stream of monochromatic electrons. 2. This stream is focused to a small, thin, coherent beam by the use of
condenser lenses 1 and 2. The first lens (usually controlled by the "spot size
knob") largely determines the "spot size"; the general size range of the final
spot that strikes the sample. The second lens (usually controlled by the
"intensity or brightness knob" actually changes the size of the spot on the
sample; changing it from a wide dispersed spot to a pinpoint beam. 3. The “beam is restricted by the condenser aperture (usually user selectable),
knocking out high angle electrons (those far from the optic axis, the dotted
line down the center)”. 4. The “beam strikes the specimen and parts of it are” transmitted. 5. This “transmitted portion is focused by the objective lens into an image”. 6. Optional “Objective and Selected Area metal apertures can restrict the
beam”; the Objective aperture enhancing contrast by blocking out high-angle
diffracted electrons, the Selected Area aperture enabling the user to examine
the periodic diffraction of electrons by ordered arrangements of atoms in the
sample. 7. The “image is passed down the column through the intermediate and
projector lenses, being enlarged all the” way. 8. The “image strikes the phosphor image screen and light is generated”,
allowing the user to see the image. The darker areas of the image represent
those areas of the sample that fewer electrons were transmitted through (they
are thicker or denser). The lighter areas of the image represent those areas of
the sample that more electrons were transmitted through (they are thinner or
less dense) as shown in Figure 2.5.
44
Fig. 2.5 Schematic diagram of TEM showing the electron path TEM of the dispersions was performed on a Jeol 2100, operated at 200kV. Both
bright field and selected area electron diffraction (SAED) imaging modes were used.
Sample preparation involved dropping the graphene dispersions onto a holey carbon
grid (400 mesh). This type of TEM grid allowed flakes to be captured while the
solvent was free to percolate through the membrane. The volume dropped depended
on the concentration of the dispersion and in most cases the dispersion was diluted
by a factor of 10, or even 20. The grid was then either dried in a vacuum oven or in
the lab overnight. The bright field images taken on this TEM were used for
determining the level of exfoliation and for statistical analysis of dimensions and
thicknesses.
45
A) B) C)
Fig. 2.6 TEM images of a 5 layered graphene flake. These TEM images demonstrate
how the lengths, widths and count the number of layers were measured.
As one can see from Figure 2.6 A, Lateral dimensions were measured by
approximating the longest axis as its length, L, and the dimension perpendicular to
the long axis as its width, w. The number of layers, N, was estimated by zooming in
on the edge of a flake and identifying the strata (Figure 2.6 C). Additional high
resolution electron diffraction patterns were obtained on the Titan Zeiss TEM. The
resulting spot diffraction patterns correspond to electrons that have been diffracted
from a specific region in single crystal graphene. Such patterns can be used for
identification of mono- , bi- or multi- layered graphene [8-10].
References [1] C.V. Raman and K.S. Krishnan., Nature 121 (1998) 501.
[2] M.J. Pelletier., Oxford: Blackwell Science vii (1999) 60.
[3] Horiba Scientific. Raman Spectroscopy.
http://www.horiba.com/scientific/products/raman-spectroscopy/
[4] J. Javier. An Introduction to Raman Spectroscopy,
http://www.spectroscopynow.com/coi/cda/detail.cda?id=1882&type=EducationFeat
ure&chId=6&page=1
[5] C. Harper, Handbook of Plastics, Elastomer, and Composites, McGraw-Hill, Inc,
New York (2002).
[6] SEM Schematic. 14 October 2010
46
http://www.mse.iastate.edu/microscopy/path2.html.
[7] J. Ackermann, Manual for the SUPRA (VP) and ULTRA Scanning Electron
Microscopes (SmartSEM V 05.00). 2005, Carl Zeiss Ltd.
[8] J.C. Meyer, et al, Nature 446 (2007) 60.
[9] S. Horiuchi, et al, Applied Physics Letters 84 (2004) 2403.
[10] J. Meyer, et al, Solid State Communications 143 (2007) 101.
47
Chapter 3
Concentrated Dispersion of Graphene&
Size Selection via Centrifugation
3.1 Objective
1.8mg/ml concentration of of graphene dispersion was reported before the
completion of this work which was not enough for use in the preparation of
nanocomposites,because production of nanocomposites requires the concentrated
dispersion of graphene. Highly concentrated dispersion of graphene was obtained in
this work using organic solvent. Similarly different flake sizes have different effects
on the mechanical performance of nanocomposites in terms of modulus, ultimate
tensile strength and elongation at break (dL at break), so graphene sheets were
separated according to its lateral dimensions for use as reinforcement in selected
polymer.
3.2. Introduction
It has been known for some years that graphite can exfoliated in the liquid phase to
give graphene [1]. There are two main ways to do “this oxidation of graphite
followed by exfoliation in water to give graphene oxide” [1-7] One “advantage of
GO based dispersion is that the flakes tend to be predominately monolayer”
However, the “oxidization process tends to introduce large quantities of structural
defects which shift the physical properties away from pristine graphene”[2-5].
While another procedure is exfoliation of graphite in solvents or surfactant solutions
to give dispersed pristine graphene [8-25]. “Solvent or surfactant exfoliated graphene
gives defect-free flakes but with relatively low monolayer content”. “Each method
results in dispersions with concentrations of up to a few mg/ml produced” in up to
litre batches [1].
Although advances in this field have been rapid, a number of outstanding problems
remain. Of these, probably the most important is the relatively low concentration of
48
dispersed graphene. For example, graphene oxide has been “dispersed in some
organic solvents at concentrations of up to 1 mg/mL” [6,26-28] and in water having
concentrations of around 7 mg/mL [4]. Similarly, graphene was initially dispersed in
solvents at extremely low concentrations of ~10-2 mg/mL [16,17]. Recently, it was
shown that this could be increased to 1 mg/mL [18]. In contrast, surfactant-dispersed
graphene has not been achieved at concentrations above 1 mg/mL [20,29,30].
Although these concentrations are now in the appropriate range for a number of
applications,yet they are not high enough for applications like nanocomposite
production. For example, solution-phase polymer/graphene composite formation
[1,30,31] would be much simpler if well exfoliated graphene dispersions were
available at high concentrations. In addition, the deposition of thin films by vacuum
filtration followed by membrane dissolution [12] requires dilution with large
quantities of water before filtration. In addition, graphene flakes can be selected by
size or thickness by chromatography or density gradient centrifugation [13]. In both
cases, the amount separated is limited by the starting concentration. In these and
many other areas, a significant barrier to progress is the lack of high-concentration
dispersions.
Two different approaches were followed by using N-methyl-2-pyrrolidinone (NMP)
as solvent for the extremly high concentration of graphene.
First method gives concentrations of 17 mg/mL and second one about 63mg/mL.
Liquid “exfoliation of graphite is usually considered as a method to produce
graphene in large quantities for applications such as in composites materials.
However, many of these applications require flakes” “which are considerably larger
than those currently available”. Gong et al. recently “showed that in order to produce
effectively reinforced graphene–poly (methyl methacrylate) composites, the flake
length would have to be a few microns” or greater [27]. Currently available
“exfoliated graphene is usually significantly smaller than this which partly explains
why most graphene composite papers describe reinforcement values much lower
than the theoretical limit” [28] of dY/dVf ~1 TPa where Y is the composite modulus
and Vf is the graphene volume fraction [29–38]. Thus, there is a real need to
“increase the size of dispersed flakes. Ideally, we would tune the
dispersion/exfoliation process to give larger flakes”. However, “while some progress
has been made in this area, it is “worth exploring methods to post-treat existing
dispersions to select flakes by size””. While a number of methods have been
49
demonstrated to separate GO flasks by lateral size [39–43], to our knowledge, lateral
size “selection has not been demonstrated for defect free graphene. Here, we describe
a method to take an existing dispersion of graphene in solvent and separate flakes by
size using controlled centrifugation”. We have produced a set of “dispersions with
mean flake lengths varying from 1 to 3.5 microns. This method is versatile and could
easily be applied to surfactant stabilized graphene” [19, 20, 22] or indeed any
exfoliated layered compounds [44].
3.3 Experimental Details
3.3.1 Concentrated dispersion of graphene.
Two samples of graphite dispersion in NMP with 100 mg/mL concentration
(dispersing 10gms graphite in 100 mL NMP) were sonicated by sonic tip and bath
sonicator indenpendetly.The small aliquots were taken from sonication system and
were periodically analyzed by UV-Vis spectroscopy and concentration of the
dispersion was measured at 660 nm peak for all the samples with the alpha
coefficient value to be 3.62 [18] as shown in Fig.3.1
For the 1st step of exfoliation maximum concentration of 0.545mg/mL and 2.01
mg/ml was observed for sonic bath and sonic tip samples respectively as shown Fig
3.2
Fig. 3.1 Absorbance versus wavelength curve for 1st cycle using (A) sonic bath (B) sonic tip
Time (hrs) Time (hrs)
50
Fig. 3.2 Change in concentration with time for 1st step using (A) sonic tip and (B) sonic bath
After drop in concentration, fresh NMP was added to the already sonicated sample
after filteration and was again sonicated for six and ten hours respectivelyand
independently.A difference in the concentration was observed with time for both the
samples as shown in UV-Vis spectra Fig. 3.3
Fig. 3.3 Absorbance versus wavelength curve for 2nd cycle using (A) sonic tip for 6hrs (B)
sonic tip for 10hrs (C) sonic bath
A
A
51
For sonic tip samples which was sonicated for 6 hrs continuously a maximum
concentration of 12 mg/mL was observed.while other sample which was pre-
sonicated for ten hours continiously maximum concentration of 17.8 mg/mL was
observed after 28 hrs sonication for it.(Figure 3.4 A and B).
The concentration obtained in second case is better than first one which suggests
that concentration of dispersion in 2nd cycle depends on the time for which it was
sonicated in 1st cycle.
Fig. 3.4 Change in concentration with time for 2nd cycle using (A) sonic tip for 6hrs (B)
sonic tip for 10hrs (C) sonic bath
3.3.2 Second Method for extremly high concentration of graphene
100 mg/ml dispersion was made by dispersing 10 gms of graphite in 100 mL of NMP
and was bath sonicated continuously for 11days. After 11days dispersion was
filtered, washed two times with fresh NMP and than re- dispersed in fresh NMP and
was again sonicated for 24 hours.Concentration was recorded initially after one
hour,then after every three hours and finally after every 24 hours using UV
spectroscopy.
3.3.2.1 Dispersions having,20 30,45 and 60mg/ml concentration.
52
The prepared dispersion was centrifuged at 500 rpm for 45 minutes the supernatant
was collected and was filtered on filter membranne. Then the collected cake was re-
dispersed in 8 mL of NMP to make 20, 30,45 and 60 mg/mL concentrated
dispersions separately.Then its sedimentation study was conducted in order to study
its settling behavoiur with passage of time and these samples remained under study
for about 192 hrs. Its concentration was recorded after every 24 hrs.It was observed
that concentration remain stable at about 35 mg/mL for 63mg/mL and 45 mg/mL
sample even after the lapse of about 200 hrs.
Fig. 3.5 (A) Change in concentration with sedimentation of 63mg/mL sample (B)
Comparative study of sedimentation of 20,30 and 45mg/mL graphene concentartion
dispersions
3.4 Charetarization
3.4.1 Transmission Electron Microscopy (TEM) Histogram
Re--centrifuged
53
Average flake size and number of layers in graphene flakes were found decreased
with sedimentation process of high concentrated dispersions. This result suggests that
bigger and heavier flakes settle with the sedimentation process and we are left with
thinner and lighter flakes after sedimentation (Figure 3.6,A,B,C,D). This study was
confirmed by conducting TEM statistics on about eighty flakes for high concentrated
dispersions of graphene
Fig 3.6 (A) Comparative study of average length, width and no of layers for 20mg/mL (B)
30mg/mL (C) 45mg/mL and (D) 63mg/mL dispersion
54
3.5 Size selection of graphene Flakes according to its lateral
dimensions.
3.5.1 Experimental Details
100 mg/mL dispersions were made and were sonicated by sonic tip and sonic bath.
Tip sonic sample was sonicated for ten hours while that of sonic bath sample was
sonicated for ten days later on both sample were centrifuged at 500 rpm for 45
minutes in order to separate the exfoliated graphene from un-exfoliated graphite
flakes.
The supernatant were collected and filtered through nylon membrane. The filtered
cake was again re-dispersed in fresh NMP and both samples were centrifuged
according to given schematic diagram. For example sample was centrifuged at 5500
rpm for 45 minutes the supernatant was kept aside then again to the same sediments
fresh 1bout 20 mL NMP was added and again centrifuged. The process was
continued till 500 rpm and all the supernatants were kept aside for further studies.
55
Fig 3.7. Schematic diagram of size separation of graphene flakes [45]
3.6 Results and Discussion
3.6.1 Concentration study
UV-Visible Measurements were carried out on the samples to understand the change
in concentration of sample for various rpms.
A very interesting phenomenon was observed for both sonic bath and sonic tip
exfoliation process. The concentration of graphene contents was gradually decreased
with decreasing the speed of centrifuge in terms of revolutions per minute (rpm)
while for sonic tip process the concentration of graphene increases with decreasing
centrifuge speed (rpm). It is very clear from Fig.3.8 & 3.9. The concentration at
various rpms tells us that we may have different concentration of different flake size
distribution in our initial dispersion.
56
Fig. 3.8 Concentration of graphene exfoliated by tip sonicator and centrifuged at
various speed (rpm).
Fig. 3.9 Concentration of graphene exfoliated by bath sonicator and centrifuged at various
speed (rpm)
0 1000 2000 3000 4000 5000 6000
0.0
0.5
1.0
1.5
2.0
2.5
3.0
co
nce
ntr
atio
n a
fte
r d
ilutio
n (
mg
/ml)
rpm
0 500 1000 1500 2000 2500 3000 3500 4000 4500
0.04
0.06
0.08
0.10
0.12
0.14
0.16
0.18
con
cen
tra
tion
(m
g/m
l)
rpm
57
3.6.2 Raman Spectroscopy
Free standing films were prepared of the samples, based on alumina membranes,
(Whatman Anodisc 47mm with pore size of 0.02µm) for Raman spectroscopy.
Raman spectroscopy gives us information about the size of graphene flakes. If size of
flake is small then it will have high ratio of ID/ IG (“D” stands for Defect and”G”
graphite peak) and vice versa. An increase in the ID/IG ratio was observed with
increasing speed of centrifuge (rpm) indicating the increase in defect peak. It is
believed that the increasing ID/IG is dominated by new edge formation rather than by
basal plane defects (Figure 3.10). So it is expected that for small size graphene flakes
we will have more number of edges thus higher Defect peak as a result high value of
ID/ IG. This is confirmed by Raman spectra however, Raman spectroscopy is used in
tandem with TEM for better understanding of flake size.
Fig. 3.10 Change in ID/IG ratio of graphene exfoliated by (A) sonic tip, (B) sonic bath with
speed (rpm) of centrifuge
ID/IG ratio increases as number of defects increases in the sample as shown Figure
3.11 which is not due to the structural defects but due to the defects that are formed
58
with more number of edges which means small flake size separated at high speed
(rpm) [18].
3.6.3 Transmission Electron Microscopy
As we discussed for the study of graphene flake size TEM is also used in
combination with Raman spectroscopy. TEM analysis performed on selected
samples which were also analyzed by Raman spectroscopy. The result of TEM
analysis confirmed that those flakes which were separated on high rpm having high
value of ID/IG ratio has small sizes as compared to those which were separated at low
rpm. In short flake size of graphene decreases with increasing speed (rpm) of
centrifuge and vice versa as shown in Figs.3.12 and 3.13.
59
Fig. 3.11 Increase in the Defect peak of normalized Raman spectra with increasing rpm (A)
for sonic tip (B) for sonic bath
60
Fig. 3.12 TEM images of graphene layers observed (a) un-separated graphene dispersion for
sonic tip (b) 500rpm sonic tip (c) 1000rpm sonic tip (d) 3000rpm sonic tip (e) un separated
graphene dispersion for sonic bath (f) 500rpm sonic bath (g) 1000rpm sonic bath (h)
3000rpm sonic bath.
Fig.3.13 (A) shows the change in the average values of length, width and number of
layers with change in rpm for sonic tip sample (B) sonic bath sample
61
The above results become clearer when we plot the average of length, width and no
of layers with various rpms. It was found that there was a big difference in the flakes
size from the start to the end rpm in sonic bath sample but for sonic tip sample size
distribution of 500 and 1000rpm were almost same (Fig 3.10). As there was a good
difference in ID /IG ratio for sonic tip sample in Raman spectroscopy we assume that
there is a chance of error in calculating flake size distribution from TEM because in
this method we are just taking the average of 80 flakes so there is a big chance of
error which can justify the error in our graph.
Conclusion
Through this study we not only got highly concentrated dispersions of graphene via
two approaches but also separated it according to its lateral dimensions by controlled
centrifugation in order to study the effects of different flake size on the mechanical
properties of polymer based nanocomposites.
References
[1] S. Park, R.S. Ruoff, Nat Nanotechnol 4 (2009) 217.
[2] S. Park, J.H. An, I.W. Jung, R.D. Piner, S.J. An, X.S. Li, et al. Nano Lett. 9 (2009)
1593.
[3] S. Park, J.H. An, R.D. Piner, I. Jung, D.X. Yang, A. Velamakanni, et al, Chem.
Mater. 20 (2008) 6592.
[4] S. Stankovich, D.A. Dikin, G.H.B. Dommett, K.M. Kohlhaas, E.J. Zimney, E.A.
Stach, et al, Nature 442 (2006) 282.
[5] S. Stankovich, D.A. Dikin, R.D. Piner, K.A. Kohlhaas, A. Kleinhammes, Y. Jia, et
al, Carbon 45 (2007) 1558.
[6] S. Stankovich, R.D. Piner, X.Q. Chen, N.Q. Wu, S.T. Nguyen, R.S. Ruoff, J Mater
Chem. 16 (2006) 155.
[7] P. Blake, P.D. Brimicombe, R.R. Nair, T.J. Booth, D. Jiang, F. Schedin, et al, Nano
Lett. 8 (2008) 1704.
62
[8] A.B. Bourlinos, V. Georgakilas, R. Zboril, T.A. Steriotis, A.K. Stubos. Small 5
(2009) 1841.
[9] A.B. Bourlinos, V. Georgakilas, R. Zboril, T.A. Steriotis, A.K. Stubos, C. Trapalis,
Solid State Communication 149 (2009) 2172.
[10] J.N. Colemanm, Adv Funct Mater. 19 (2009) 3680.
[11] S. De, P.J. King, M. Lotya, A. O’Neill, E.M. Doherty, Y. Hernandez, et al, Small 6
(2009) 458.
[12] A.A. Green, M.C. Hersam, Nano Lett. 9 (2009) 4031.
[13] C.E. Hamilton, J.R Lomeda, Z.Z. Sun, Nano Lett. 9 (2009) 3460.
[14] R. Hao, W. Qian, L.H. Zhang, Y.L. Hou, Chem Commun. 48 (2008) 6576.
[15] T. Hasan, F. Torrisi, Z. Sun, D. Popa, V. Nicolosi, G. Privitera, et al, Phys Status
Solidi B-Basic Solid State Phy. 247 (2010) 2953.
[16] Y. Hernandez, M. Lotya, D. Rickard, S.D. Bergin, J.N. Coleman, Langmuir 26
(2010) 3208.
[17] Y. Hernandez, V. Nicolosi, M. Lotya, F.M. Blighe, Z.Y. Sun, S. De, et al, Nat
Nanotechnol. 3 (2008) 563.
[18] U. Khan, A. O’Neill, M. Lotya, S. De, J.N. Coleman. Small 6 (2010) 864.
[19] M. Lotya, Y. Hernandez, P.J. King, R.J. Smith, V. Nicolosi, L.S. Karlsson, et al, J
Am Chem Soc. 131 (2009) 3611.
[20] M. Lotya, P.J. King, U. Khan, S. De, J.N. Coleman, ACS Nano 4 (2010) 3155.
[21] A. O’Neill, U. Khan, P.N. Nirmalraj, J.J. Boland, J.N. Coleman, J Phys Chem C.
115 (2011) 5422.
[22] R.J. Smith, M. Lotya, J.N. Coleman, New J Phys. 12 (2010) 125008.
[23] S. Vadukumpully, J. Paul, S. Valiyaveettil, Carbon 47 (2009) 3288.
[24] V. Alzari, D. Nuvoli, S. Scognamillo, M. Piccinini, E. Gioffredi, G. Malucelli, et al,
J Mater Chem. 21 (2011) 8727.
[25] A. Catheline, C. Valles, C. Drummond, L. Ortolani, V. Morandi, M. Marcaccio, et
al, Chem Commun. 47 (2011) 5470.
63
[26] D. Nuvoli, L. Valentini, V. Alzari, S. Scognamillo, S.B. Bon, M. Piccinini, et al, J
Mater Chem. 21 (2011) 3428.
[27] L. Gong, I.A. Kinloch, R.J. Young, I. Riaz, R. Jalil, K.S. Novoselov, Adv Mater. 22
(2010) 2694.
[28] G.E. Padawer, N. Beecher, Polym Eng Sci. 10 (1970) 185.
[29] H.W. Hu, G.H. Chen, Polym Compos. 31(2010):1770.
[30] L. Jiang, X.P. Shen, J.L. Wu, K.C. Shen, J Appl Polym Sci. 118 (2010) 275.
[31] I.H. Kim, Y.G. Jeong, J Polym Sci Pol Phys. 48 (2010) 850.
[32] J.J. Liang, Y. Huang, L. Zhang, Y. Wang, Y.F. Ma, T.Y. Guo, et al, Adv Funct
Mater, 19 (2009) 2297.
[33] S.G. Miller, J.L. Bauer, M.J. Maryanski, P.J. Heimann, J.P. Barlow, J.M. Gosau, et
al, Compos Sci Technol. 70 (2010) 1120.
[34] K.W. Putz, O.C. Compton, M.J. Palmeri, S.T. Nguyen, L.C. Brinson, Adv Funct
Mater, 20 (2010) 3322.
[35] T. Ramanathan, A.A. Abdala, S. Stankovich, D.A. Dikin, M. Herrera- Alonso, R.D.
Piner, et al, Nat Nanotechnol, 3 (2008) 327.
[36] P. Steurer, R. Wissert, R. Thomann, R. Mulhaupt, Macromol Rapid Commun. 30
(2009) 316.
[37] X.M. Yang, L.A. Li, S.M. Shang, X.M. Tao, Polymer 51 (2010) 3431.
[38] X. Zhao, Q.H. Zhang, D.J. Chen, P. Lu, Macromolecules 43 (2010) 2357.
[39] P.N. Nirmalraj, T. Lutz, S. Kumar, G.S. Duesberg, J.J. Boland, Nano Lett. 11 (2011)
16.
[40] P.E. Lyons, S. De, F. Blighe, V. Nicolosi, L.F.C. Pereira, M.S. Ferreira, et al, J Appl
Phys. 104 (2008) 044302.
[41] G. Eda, M. Chhowalla, Nano Lett. 9 (2009) 814.
[42] A.A. Green, M.C. Hersam, J Phys Chem Lett. 1 (2010) 544.
[43] X.M. Sun, Z. Liu, K. Welsher, J.T. Robinson, A. Goodwin, S. Zaric, et al, Nano Res.
1 (2008) 203.
64
[44] J.N. Coleman, M. Lotya, A. O’Neill, S.D. Bergin, P.J. King, U. Khan, et al, Science
331 (2011) 568.
[45] U. Khan, A. O'Neill, H. Porwal, P. May, K. Nawaz, J.N. Coleman, Carbon 50
(2012) 470.
65
Chapter 4
Effect of Surfactant Concentration on
the Exfoliation of Graphite to
Graphene in Aqueous Media
4.1 Objective Graphite was exfoliated to graphene by tip sonic using sodium cholate as surfactant
in the presence of Millipore water as medium. Use of water as solvent in this study
for exfoliation purpose is very important due to its environment friendly nature and
almost no cost contrary to organic media. Two different concentration ratios of
surfactants are used. Graphene dispersions with two different concentrations of
5mg/ml and about 7 mg/mL respectively were obtained in aqueous media. It was
observed that optimum concentration of surfactant has effective role on exfoliation of
graphite to graphene. Concentrations of graphene dispersions were studied through
UV spectroscopy while, Raman spectroscopy, Scanning electron Microscopy (SEM)
and Transmission Electron Microscopy (TEM) were used to study the quality of
exfoliated graphene flakes.
4.2 Introduction Graphene is a nearly transparent, two-dimensional semimetal consisting of a single
atomic lattice of hexagonally arranged sp2 hybridized carbon atoms [1]. Since the
isolation of graphene and the discovery of its unique properties there have been
unprecedented levels of research on and related to the remarkable material. The work
carried out by Geim and Novoselov in 2004 was a simple exfoliation method in
which protrusions of highly-oriented pyrolytic graphite (HOPG) were embedded in
photo resist and adhesive tape was used to successively peel off layers of graphene
66
[2]. Although this method is tedious and cannot be scaled up to industrial level yet
open new horizons of research in this specific field. This so called scotched tape
method is simple and does not require any modification to environmental parameters
such as temperature and pressure. In addition this method provides high quality (high
mobility and low defect) single and few layer graphene sheets with large areas as
high as 100μm [3]. Usually strong acids are used for the oxidation of graphite to
graphene oxide (GO) which results in stable aqueous solution of GO [4]. Then this
dispersion of GO can be reduced by aqueous hydrazine as reducing agent [5,6] or by
thermal reduction under a reducing atmosphere [7-9]. Graphene growth by chemical
vapor deposition (CVD) is typically carried out under ultra-high vacuum and at high
temperatures [10]. In this process volatile or gas phase carbon precursor is flowed
over a metallic substrate which acts as a catalyst and nucleation site for graphene
growth [11]. Graphene produced by CVD was first reported by Somani and co-
workers in 2006 using nickel foil and camphor for the metallic substrate and carbon
precursor respectively [12].
Liquid exfoliation of graphite to graphene, also referred to as solution based
graphene exfoliation, was first carried out by the Coleman group [13] in 2008 via
sonication of graphite flakes in organic solvents such as N-methyl-Pyrrolidinone
(NMP) and dimethyl formamide (DMF). Coleman’s work stemmed from previous
research involving dispersion of carbon nanotubes (CNTs) in organic solvents which
was concerned with matching the surface energies associated with CNTs and the
solvent [14]. The use of surfactants in liquid exfoliation is also carried out to create
aqueous dispersion of graphene help mitigate colloidal aggregation of graphene in
solution [15]. Other less common but noteworthy liquid exfoliation methods include
intercalation of graphite with alkaline [16] or halogen salts [17] to form graphite
intercalation compounds (GICs). The GIC’s can be either directly dispersed or exfoliated in solution by
sonication [16]. Likewise these can also be thermally expanded at high temperatures
in which the intercalating compounds volatilize to form expanded graphite (EG)
[18]. In next step these expanded graphite is subsequently exfoliated in solution via
sonication [17]. Liquid exfoliation of graphite to graphene is advantageous method
as compared to other methods such as CVD growth and mechanical exfoliation due
to the simplicity of the process [13]. This does not require high vacuum and high
temperatures as well as the low cost of the starting materials. There are presently
number of commercially available surfactants that have been used in the literature for
67
solution processing of graphene by various methods and solvents [16,17]. Surfactant
assisted exfoliation has permitted the use of water as a solvent for solution
processing which is attractive from an environmental standpoint as well as for
applications which cannot tolerate organic solvents. The three main classes of
surfactants include cationic, anionic and nonionic surfactants. This also includes
small molecular surfactants such as sodium dodecyl sulfate (SDS) and sodium
cholate which consists of a hydrophobic tail and a polar head group. Similarly,
cationic, anionic, and non-ionic Pluronic and Tetronic block copolymer surfactants
have been used to form aqueous dispersions of graphene [19]. Likewise it has been
shown that graphene oxide can be dispersed in some organic solvents at
concentration up to 1mg/ml [6,20-22] and in water this concentration raised to
7mg/ml [23]. Similarly graphene concentration increased to about 1mg/ml in organic
solvents [24]. While surfactant based graphene dispersion in aqueous medium above
1mg/ml is not reported [25-27].
In this work we used two different concentrations of surfactants in order to study its
effects on exfoliation of graphite to graphene. Very reasonable concentrations of
graphene dispersions in water are obtained which is not previously reported.
Moreover, it is very interesting to see that low concentration of sodium cholate as
surfactant has very promising role on exfoliation and got high concentration of
graphene dispersion in water for low concentration of surfactant.
4.3 Experimental Procedure Graphite powder and Sodium cholate (surfactant) were purchased from Sigma –
Aldrich and were used as supplied. Sonication was performed by using sonic tip
(GEX600, 48W, 24kHz, flat head probe) running at 25% of maximum power and
sonic bath (Branson 1510E-MT). Centrifugation was performed using a Hettich
Mikro22R typically at 500 rpm for 45 minutes. After centrifugation the 70 % of top
portion of dispersed solution was removed and concentration was determined by
UV-Vis-IR absorption spectroscopy Varian Cary 6000i (with 1mm cuvettes). TEM
was done using a Joel 2100 and holey carbon grids (400 mesh). Thin film was made
using porous alumina membrane (“Whatman Anodisc 47 mm, pore size = 0.02
micron”). “Raman spectra (633 nm) were recorded on a Horiba Jobin Yvon
LabRAM-HR. Scanning Electron Microscopy” (SEM) was performed in a Hitachi S-
4300 field emission.
68
4.4 Results and Discussion Graphite was exfoliated to graphene by using sonic tip and Millipore water (water
purified by Millipore technology) was used as solvent. During exfoliation sodium
cholate was used as surfactant to ease the exfoliation of graphite to graphene. We
used two different concentrations of the surfactant, i.e. 5 mg/ml and 10mg/ml
(concentration of surfactant CS=5mg/ml and CS=10mg/ml). Initially 10 grams of
graphite was added to Millipore water. Then surfactant was added in different
amounts to the graphite dispersions. Both of these dispersions were sonicated under
same conditions and small samples were taken from these dispersions hourly. The
samples were bath sonicated for 15 minutes using sonic bath. Finally these bath
sonicated samples were centrifuged at 500 rpm for 45 minutes. The 70 % of the top
portion of centrifuged sample is taken for absorbance study using UV
spectrophotometer, to know the concentration of graphene in these samples. This
process was continued for 96 hours and samples were taken after required interval of
time. The concentration was studied through UV spectroscopy using Lambert-Beer
law eq.(1) [24]
A= α Cl (4.1)
Where the absorption coefficient α is related to the absorbance A, C is concentration
and is l” the path length. We have selected α value equal to 3.62 ml/mg/mm [24].
The concentration was measured by recording the absorbance at 660 nm and
transformed this into the concentration using eq. (4.1) [24]. It is reported that the
exfoliation was conducted using organic solvent N-methyl Pyrrolidinone (NMP). The
NMP used to spoil after 6 hours which might be due to the oxidative degradation
[28]. But this phenomenon was not observed in our study and graphene concentration
tends to rise after every hour. Although the rate of exfoliation was not as fast and
high, as it was observed in NMP [28] but after 96 hours we got 5 mg/mL and 7
mg/mL concentrations of graphene for CS =10 mg/mL and CS = 5 mg/mL of
surfactant respectively.
In this work it has been observed that optimum concentration of surfactant has an
69
effective role on the exfoliation of graphite to graphene. After 96 hours the
concentration of graphene exfoliated was 5 mg/mL in the presence of CS=10 mg/mL
of surfactant. While in case of CS= 5 mg/mL (concentration of sodium cholate) the
graphite was exfoliated to 7mg/ml under same conditions and time. It is very clear
from UV graph (Figure 4.1) that there is rapid increase in graphene concentration at
low concentration of surfactant (CS=5 mg/mL) compared to high (CS=10 mg/mL).
Latoya et. al. exfoliated graphite to graphene and obtained very low concentration in
presence of surfactant in aqueous medium [29].
Fig. 4.1 Concentration of graphene after centrifugation (500/45) as a function of sonication
time (Cs=5mg/ml). Concentration was calculated using absorption coefficient ―α value
equal to 3.62 mL/mg/mm
The TEM analysis were performed on flakes of graphene obtained after 96 hours of
sonication using surfactant with the concentration value CS=5 mg/mL deposited on
holey carbon grid. It is apparent that the exfoliated graphene flakes were in few
layers of graphene.
70
Fig. 4.2 Concentration of graphene after centrifugation (500/45) as a function of sonication
time (CS=10mg/ml). Concentration was calculated using absorption coefficient ―α value
equal to 3.62mL/mg/mm.
TEM images are presented in Figure 4.3. It was revealed that large numbers of
graphene flakes of various types with different size are present as shown Figure 4.4
histogram-- showing number of layers of graphene after 96 hours of sonication along
with its length and thickness.
71
Fig. 4.3 TEM images of graphene flakes deposited from sample having concentration of
7mg/mL
72
Fig. 4.4 Histogram showing (Bottom) the number of layers per flake measured for 96
hours sonication time (centre) average length of flakes and (top) the average width of
flakes
It is very clear from this histogram that most of the population consists of few layer
graphene (less than five layers) and the length of flake is about 1.0 micron while the
width of the maximum population of graphene flakes is about 0.6 micron. It confirms
that graphite is fully exfoliated. Likewise Scanning Electron Microscopy (SEM) of
thin film (the segment of the film used for SEM was coated with 10-20 nm of
gold/palladium) also reveals that the exfoliated graphene consist of few layer as
shown in (Figure 4.5 A and B).
This study revealed about size of graphene flakes and defects in it. Spectrum of
graphite materials can be characterized by certain and specific bands like D-band
(1350cm-1
), G-band (1582cm-1
), and 2D band (2700cm-1
) [30]. D-band shows the
evidence of the presence of topological defects in sheets or edges of nanosheets. [31].
73
Fig. 4.5 (A) SEM images of the flakes present on the interface of the free standing films
prepared
Fig. 4.5 (B) SEM image showing the fractured interface
A typical Raman spectrum was measured on film prepared from the sample having
7mg/ml concentration. The film was deposited on alumina membrane by filtering the
aqueous graphene dispersion under vacuum. This film was thoroughly washed with
plenty of de-ionized water to make it free from surfactant. For solvent exfoliated
74
graphene the D band is associated with presence of flakes edges and can be linked to
flake length by relation in eq. 2.
ID/IG-(ID/IG) powder = k/L (4.2)
Where k is constant [34, 35], so increase in ID/IG value shows decrease in flake size
and vice versa.
The value of k is reported to be =0.26, this also gave (ID/IG) powder = 0.037 [24,32].
It is clear from Figure 4.6, ID/IG value of exfoliated graphene as function of
sonication time increases. The ID/IG of graphene sonicated for 30 hours has low
value of ID/IG while as sonication proceeds on the ID/IG increases and reaches its
maximum value at 96 hours.
Fig. 4.6 Increase in Defect peak (D-band) of normalized Raman spectra as a function of
sonication time
The data is also presented in Figure 4.7 just to show the values of ID/IG with passage
1600 2400 3200 0.00 0.31 0.62 0.93 0.00 0.38 0.76 1.14 0.00 0.38 0.76 1.14 0.00 0.38 0.76 1.14 0.00 0.38 0.76 1.14 0.00 0.38 0.76 1.14
1000 2000 3000
Wave number (cm-1)
24 hrs
36 hrs
48 hrs
64 hrs
24 hrs
80 hrs
96 hrs
Norm
aliz
ed
avera
ge
75
of time. The ID/IG value continuously increases from 0.113 at 30hrs to 0.317 at 96
hrs, while (ID/IG) powder for graphite powder is 0.037 [24,32]. This suggests that
with sonication time there is increase in ID/IG value suggesting decrease in flake
size. Which suggests that sonication creates number of defects in graphene flakes by
cutting the size of graphite creating new edges [33-34]. Equation (2) can be used to
estimate the flake length. By putting obtained different values of ID/IG in eq. (4.2).
Fig. 4.7 Change in ID/IG as a function of sonication time
It is shown in Figure 4.8 that flake length decreases from 3.42 micron to 1 micron
after sonication for 96 hours. Similarly the Raman data of estimated flake length of
graphene nanosheets on the basis of ID/IG value is in close agreements with TEM
study. The histogram shown in Figure 4.4 indicates that most of the flakes consist of
less than five layers and having flake length between 1-1.5 microns with average
width of 0.6 micron.
76
Fig. 4.8 Estimated length of graphene flake with D/G values.
Conclusion Although high concentration of exfoliated graphene is earlier reported in organic
solvent, yet due to the use of organic solvent environmental issues can be raised. So
in this work water was used as solvent. It has been observed that reasonable
concentration of graphene in aqueous media in short time can also be obtained by
using sonic tip in the presence of surfactant. The flakes length obtained through this
procedure is about one micron with few layer thicknesses. Likewise the remaining
part of graphite crystallites/un-exfoliated graphite can further be used for further
exfoliation after filtration for better results in terms of high concentration.
Concentration of graphene dispersion can be increased from 7mg/mL by just finding
some other suitable surfactant for this purpose.
References
[1] R. Ruoff, Nature Nanotechnology 3 (2008) 10.
[2] K.S. Novoselov, A.K. Geim, S.V. Morozov, D. Jiang, Y. Zhang, S.V. Dubonos, I.V.
Grigorieva, A.A. Firsov, Science 306 (2004) 666.
[3] A.H. Castro Neto, F. Guinea, N.M.R. Peres, K.S. Novoselov, A.K. Geim, Rev. Mod.
0
0.5
1
1.5
2
2.5
3
3.5
4
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
D/G Values of Raman Spectrum
Fla
ke length
of
Gra
phene (
mic
ron)
77
Phys. 81 (2009) 109.
[4] S. Some, Y. Kim, E. Hwang, H. Yoo, H. Lee, Chem. Commun. 48 (2012) 7732.
[5] D. Li, M.B. Muller, S. Gilje, R.B. Kaner, G.G. Wallace, Nature Nanotechnology 3
(2008) 101.
[6] V.C. Tung, M.J. Allen, Y. Yang, R.B. Kaner, Nature Nanotechnology 4 (2009) 25.
[7] X.F. Gao, J. Jang, S. Nagase, J. Phys. Chem. C 114 (2010) 832.
[8] N.W. Pu, C.A. Wang, Y.M. Liu, Y. Sung, D.S. Wang, M.D. Ger, J. Taiwan Inst.
Chem. Eng. 43 (2012) 140.
[9] D. Yang, A. Velamakanni, G. Bozoklu, S. Park, M. Stoller, R.D. Piner, S.
Stankovich, I. Jung, D.A. Field, C.A. Ventrice, R.S. Ruoff, Carbon 47 (2009) 145.
[10] W. Choi, I. Lahiri, R. Seelaboyina, Y.S. Kang, Critical Reviews in Solid State and
Materials Sciences 35 (2010) 52.
[11] V. Singh, D. Joung, L. Zhai, S. Das, S.I. Khondaker, S. Seal, Progress in Materials
Science 56 (2011) 1178.
[12] P.R. Somani, S.P. Somani, M. Umeno, Chemical Physics Letters 430 (2006) 56.
[13] Y. Hernandez, V. Nicolosi, M. Lotya, F.M. Blighe, Z. Sun, S. De, I.T. McGovern, B.
Holland, M. Byrne, Y.K. Gun'ko, J.J. Boland, P. Niraj, G. Duesberg, S.
Krishnamurthy, R. Goodhue, J. Hutchison, V. Scardaci, A.C. Ferrari, J.N. Coleman,
Nature Nanotechnology 3 (2008) 563.
[14] S.D. Bergin, V. Nicolosi, P.V. Streich, S. Giordani, Z. Sun, A.H. Windle, P. Ryan,
N.P.P. Niraj, Z.-T.T. Wang, L. Carpenter, W.J. Blau, J.J. Boland, J.P. Hamilton, J.N.
Coleman, Advanced Materials 20 (2008) 1876.
[15] A.A. Green, M.C. Hersam, Nano Letters 9 (2009) 4031.
[16] K.H. Park, B.H. Kim, S.H. Song, J. Kwon, B.S. Kong, K. Kang, S. Jeon, Nano
Letters 12 (2012) 2871.
[17] S. Lin, C.-J. Shih, M.S. Strano, D. Blankschtein, Journal of the American Chemical
Society 133 (2011) 12810.
[18] Y.A. Nikitin, M.L. Pyatkovskii, Powder Metallurgy and Metal Ceramics 36 (1997)
78
41.
[19] J.-W.T. Seo, A.A. Green, A.L. Antaris, M.C. Hersam, Journal of Physical Chemistry
Letters 2 (2011) 1004.
[20] S. Stankovich, D.A. Dikin, R.D. Piner, K.A. Kohlhaas, A. Kleinhammes, Y. Jia, Y.
Wu, S.T. Nguyen, R.S. Ruoff, Carbon 45 (2007) 1558.
[21] J.R. Lomeda, C.D. Doyle, D.V. Kosynkin, W.-F. Hwang, J.M. Tour, Journal of the
American Chemical Society 130 (2008) 16201.
[22] G. Williams, B. Seger, P.V. Kamat, Acs Nano 2 (2008) 1487.
[24] S. Park, J. An, R.D. Piner, I. Jung, D. Yang, A. Velamakanni, S.T. Nguyen, R.S.
Ruoff, Chemistry of Materials 20 (2008) 6592.
[25] U. Khan, A. O'Neill, M. Lotya, S. De, J.N. Coleman, Small 6 (2010) 864.
[26] M. Lotya, P.J. King, U. Khan, S. De, J.N. Coleman, Acs Nano 4 (2010) 3155.
[27] L. Guardia, M.J. Fernandez-Merino, J.I. Paredes, P. Solis-Fernandez, S. Villar-
Rodil, A. Martinez-Alonso, J.M.D. Tascon, Carbon 49 (2011) 1653.
[28] Y.T. Liang, M.C. Hersam, Journal of the American Chemical Society 132 (2010)
17661.
[29] U. Khan, H. Porwal, A. O'Neill, K. Nawaz, P. May, J.N. Coleman, Langmuir 27
(2011) 9077.
[30] M. Lotya, Y. Hernandez, P.J. King, R.J. Smith, V. Nicolosi, L.S. Karlsson, F.M.
Blighe, S. De, Z. Wang, I.T. McGovern, G.S. Duesberg, J.N. Coleman, Journal of
the American Chemical Society 131 (2009) 3611.
[31] A.C. Ferrari, J.C. Meyer, V. Scardaci, C. Casiraghi, M. Lazzeri, F. Mauri, S.
Piscanec, D. Jiang, K.S. Novoselov, S. Roth, A.K. Geim, Physical Review Letters
97 (2006).
[32] C. Casiraghi, A. Hartschuh, H. Qian, S. Piscanec, C. Georgi, A. Fasoli, K.S.
Novoselov, D.M. Basko, A.C. Ferrari, Nano Letters 9 (2009) 1433.
[33] A. O'Neill, U. Khan, P.N. Nirmalraj, J. Boland, J.N. Coleman, J. Phys. Chem. C 115
(2011) 5422.
79
[34] U. Khan, A. O'Neill, H. Porwal, P. May, K. Nawaz, J.N. Coleman, Carbon 50
(2012) 470.
80
Chapter 5
Observation of mechanical percolation
in functionalized graphene oxide –
elastomer composites
5.1 Objective
We have covalently functionalized graphene oxide (GO) with octadecylamine
(ODA) to form GO-ODA. This material can be dispersed in tetrahydrofuran (THF)
and subsequently formed into composites with polymers such as polyurethane.
Prominent rise in stiffness and low-strain stress were observed at loading levels up to
50wt%. However, most interestingly we found no increase in these properties at
loading levels below 2.5vol%. Reinforcement appeared to turn on sharply at this
volume fraction and subsequently increase as a power law with volume fraction.
This behavior is typical of percolation and shows that the low-strain stress cannot be
enhanced until the functionalized graphene flakes form a percolating network.
Slightly different behavior is observed for properties related to material failure. The
ultimate tensile strength increased linearly with graphene content up to the
percolation threshold before subsequently falling off. Similarly the ductility was
constant below the percolation threshold but fell off dramatically above it. This work
shows the importance of network formation in the reinforcement of elastomeric
materials.
5.2. Introduction
Graphene oxide (GO) [1] is an exceptional material which has shown great promise
in a number of application areas, [2-5]. One area where GO is expected to make a big
impact is as a filler in composites. Incorporation of GO into polymers such as
polyvinyl alcohol can result in significant increases in mechanical properties, [6],
81
while composite formation with reduced GO can give large increases in electrical
properties [1]. However, due to its polar nature, as-produced GO is not a natural
filler material for non-polar polymers. This can be addressed by the covalent
functionalization of GO with non-polar polymers or molecules. These groups alter
the surface chemistry thus compatiblising the functionalized GO with both non-polar
solvents and polymer matrices.
Of particular interest is the functionalization of graphene oxide with octadecylamine
(ODA). The reason for this is that the amine group of the ODA should reach readily
with epoxides or carboxylic acid groups in the GO [7-9]. Critically, this reaction can
be followed with FTIR [10] allowing researchers to be confident that they have
produced GO-ODA. Because of the presence of covalently attached non-polar
groups, GO-ODA should disperse well in non-polar polymers. This makes it a good
model system for the exploration of the properties of polymer-graphene composites
processed using organic solvents.
One interesting class of composite consists of thermoplastic elastomers filled with
nano-fillers such as carbon nanotubes or graphene [11-14]. These are of great
interest as addition of the nano-filler can result in the scaling of ductility and low-
strain stress all the way from those of an elastomers to a rigid thermoplastic [15, 16].
However, the reinforcement mechanism at work in these systems is unclear. We
suggest that GO-ODA is an ideal model system to investigate the nature of this
mechanism. In this work, we functionalize GO with ODA. FTIR shows the
attachment to be covalent while SEM analysis shows the GO-ODA to be well
dispersed, even at very high loading levels. Mechanical measurements show no
reinforcement at volume fractions below 2.5%. Above this filler content, both
stiffness and low-strain stress increases significantly in line with Percolation theory.
However, we find that once a percolating network is first formed, the ultimate tensile
strength and ductility begin to fall with increasing graphene content.
5.3 Experimental Procedure Graphite powder (Sigma Aldrich) was oxidized using a modified version of the
Hummers method [17]. In brief, “graphite powder was oxidized using NaNO3,
H2SO4 and KMnO4 in an ice bath [17]. The material obtained was centrifuged at
2000 rpm for 10 minutes. The supernatant, containing inorganic salts”, was decanted
and the deposit washed several times with de-ionized water until the pH was neutral.
82
This sediment was dried overnight in a vacuum oven at 80°C.
The dried GO was treated with ODA as previously described [8] with slight
modification.Briefly,300 mg of “GO was dispersed in 30 ml of water” while 300 mg
of ODA was dissolved in 30 ml of dimethyl formamide (DMF). Both the dispersions
were mixed together in round bottom flask and refluxed for 48 hours at 100oC under
continuous stirring. The resultant material was filtered through a “nylon membrane
of pore size 0.45 µm” (Sterlitech). The filtered material was thoroughly washed with
THF to remove any un-reacted, free ODA.
In next step the ODA functionalized graphene (GO-ODA) was dispersed in THF by
sonication for 30 minutes in a sonic bath (Branson 1510-MT). The dispersion was
centrifuged at 500 rpm for 45 minutes using a Hettich Mikro 22R to remove large
aggregates or un-functionalized GO. The supernatant was collected and filtered
through a nylon membrane of pore size 0.45 µm (Sterlitech). The membrane
supported GO-ODA was dried in a vacuum oven overnight at 60 oC to facilitate
accurate mass determination. The thermoplastic polyurethane (TPU) used in this
work was Morthane which was obtained from Huntsman Polyurethanes (Morthane
PS455-203 - an aromatic polyester based thermoplastic polyurethane). This polymer
(6 g) was dissolved in 100ml of THF (60 mg/ml) by stirring at 40oC for 24 hours.
GO-ODAwas added to the TPU solution at the required concentration and sonicated
for 30 minutes.
Many composite dispersions were prepared containing various mass fractions of
ODA-GO from 1 to 50 wt %. These dispersions were sonicated for four hours in a
sonic bath and were cast in Teflon trays of dimensions 4×4×2 cm. These cast
samples were dried at room temperature and 900 mbar in a vacuum oven for 24 hrs.
The dried films were then placed in an oven at 60oC for 48 hours to remove trace
THF. All samples were of constant mass (~150 mg). The film thicknesses were in
the range of 50 to 60 microns. A reference sample of TPU was also prepared.
Mechanical “testing was performed using a Zwick-Roell tensile tester with a 100N
load cell at a strain rate of 50mm/min. Scanning electron microscopy (SEM) was
performed using a Carl Zeiss Ultra Plus Field Emission Scanning Electron
Microscope”. FTIR was measured on crushed powder on a glass slide in
transmittance mode using Nexus Nicolet FTIR.
83
5.4 Results and Discussion Initially, it is important to ascertain whether the ODA has been covalently
attached to the GO. We expect the reaction to be of the form [8].
R-CO-OH + R-NH2 RCO-NH-R + H2O (5.1) Where the Nitrogen of the ODA bonds to the carbonyl carbon of the GO, although
reaction with the basal plane epoxides has also been suggested,[7-9].To test this
hypothesis, we performed FTIR spectroscopy on the GO, neat ODA and GO-ODA
samples (Figure 5.1). The GO exhibits the characteristic bands at 1720, 1630, 1390,
1220 and 1050 cm-1
, which can be associated with C=O, C=C, C-O (carboxy), COH
and C-O (alkoxy) groups, and a broad absorption band between 3000 and 3500
cm−1
associated with the hydroxyl groups [18-21].
C-O
GO-ODA
ODA
GO
C=O
C-OH
AlkylC-H
N-H
Amide C=O
(alkoxy)
C-O (alkoxy)
C-O (carboxy)
Tra
nsm
issio
n
4000 3000 2000 1000
Wavenumber (cm-1)
Fig. 5.1 FTIR spectra of GO (top), ODA (middle) and GO-ODA (bottom).
84
The FTIR spectrum of the neat ODA shows a peak characteristic of alkyl groups at
2850 cm-1
and a peak at 2918 cm-1
due to asymmetric C-H stretching. The peaks
around 1500 cm-1
are due to methylene scissoring deflection and other effects
associated with the alkyl chain [22]. Most importantly, we observe a band at 3333
cm-1
associated with the amine group (N–H) [4, 17]. In the GO-ODA FTIR spectra,
this peak has totally disappeared, strongly suggesting that the nitrogen of the ODA
has covalently bonded to the GO as described above [8]. Further evidence of
functionalization comes from the fact that carbonyl C=O band at 1720 cm-1
in GO
sample is obscured in the GO-ODA spectra and new band has appear at around 1640
cm-1
. This is consistent with amide C=O stretching as would be expected from the
reaction above [18]. Thus, we believe that “FTIR spectra clearly confirm that the
ODA molecules were attached to the graphene nanosheets through chemical
modification”.
Unlike GO, the GO-ODA was readily dispersible in THF. This allows the formation
of composites by blending dispersions of GO-ODA with solutions of TPU in THF.
These composite dispersions can then be formed into composite films. In all cases,
including 50wt% GO-ODA, composites were very uniform to the naked eye with no
appearance of aggregates. Figure 5.2 shows an SEM image of the fracture surface of
a 40 wt% GO-ODA composite. The inset shows a high magnification image showing
a homogeneous dispersion of GO-ODA within the polymer.
Fig. 5.2 SEM images of 40 wt% composite film.
We performed mechanical characterization of GO-ODA/TPU composites for a range
85
of GO-ODA mass fractions. Major stress strain curves for the composites are shown
in Figure 5.3.
40
10 4%
8
(MP
a) 30
6
4 1%
2
0 PU
20
0 2 4 6 8 10
10%
Str
ess,
50%
40%
10
0
0 200 400 600 800 1000
Strain, (%)
Fig. 5.3 Representative stress strain curves. Inset: the low strain regime.
It is clear from this graph that the presence of GO-ODA has a significant effect on
the polymer. For low mass fractions the stress at all strains appears to increase while
for higher mass fractions, both stress and ductility fall dramatically. From the stress
strain curves, one can extract four main mechanical properties; the Young’s modulus
or stiffness, E, the stress at low (3%) strain, 3% , the ultimate tensile strength,
B , and the strain at break, B . The average values of these quantities are shown as a
function of the filler mass fraction in Figure 5.3. The Young’s modulus appeared to
increase almost linearly with increasing GO-ODA mass fraction from 9.6 MPa for
the polymer to 335 MPa for the 50wt% composite. Similarly, the stress at 3% strain
increased in an almost linear fashion from 0.3 MPa for the polymer to ~10 MPa for
the 50wt% composite. In contrast, the ultimate tensile strength initially increased
from 27 MPa for the polymer to 38 MPa for the 3 wt% composite before falling
steadily, reaching 10 MPa for the 50wt% composite. Interestingly, the initial increase
was linear with a slope of 330 MPa. Although a linear increase is “predicted by the
rule of mixtures”, this slope is low compared with typical values of GPa usually
86
found for composites of thermoplastic polymers filled with nanotubes or graphene
[23, 24]. The strain at break decreased steadily with increasing GO-ODA mass
fraction from ~1000% for the polymer to ~10% for the 50 wt% sample.
Broadly speaking, these results are typical of what is observed for elastomers filled
with high volume fractions of nanotubes [11, 13, 16] or graphene [15]. These results
appear similar to those of Khan et al who mixed pristine solvent-exfoliated graphene
with thermoplastic polyurethane [15]. The starting polyurethane had very similar
properties while their 50wt% sample had values of E~1 GPa, 3% ~10 MPa,
B~30 MPa and B~10%, values almost identical to those found here. This implies
that the ODA functional groups (or indeed the oxides) play no significant role in the
reinforcement but ensure good dispersion.
We can also compare these results with other work on composites of thermoplastic
elastomers filled with other nano materials. Liff et. al. [25] increased the modulus of
polyurethane by a factor of 20 by addition of 20wt% nanoclay. By the inclusion of
functionalized nanotubes in TPU, Koerner et. al. [26] reported a thirteen-fold
increase in modulus and a significant enhancement of the strength. Sahoo et. al. [27],
fabricated functionalized SWNT/ PU composites and achieved an increment of 3
times in modulus at 20% loading. While a seven fold increase in modulus was
reported by Cheng et. al. in PU composites reinforced by functionalized MWCNTs
[28], the ductility of the composites was reduced significantly.
It is worth looking at the dependence of mechanical properties on graphene content
in more detail. To do this we plot the same data as in Figure 5.4 A and B but in
double logarithmic format.
87
0.0
0.1
0.2
0.3
0.4
0.5
E (
GP
a)
A
0
2
4
6
8
10
=
3% (
MP
a) B
0
10
20
30
40
50
B (
MP
a)
C
0 10 20 30 40 50
0
400
800
1200
B (
%)
Graphene content (wt%)
D
Fig. 5.4 Effect of GO-ODA content on mechanical properties of composites. (A) Young’s
modulus, (B) stress at 3% strain, (C) Ultimate tensile strength, (D) strain at break.
In addition, we have transformed the mass fraction to volume fraction assuming a
GO-ODA density of 1800 kg/m3 and a TPU density of 1100 kg/m
3. This data is
shown in Figure 5.5 A and B and immediately illustrates some behavior not
88
apparent from the linear graphs. It is clear from figures 5.4 A and B that the modulus
and stress at low strain hardly increase at all for graphene contents below ~2.5vol%.
A
E (
GP
a)
0.1
0.01
(MP
a) 10 B
=3
%
1
50
(MP
a) 40
30
20
B
10 C
0
1000
(%) 800
600
B
400
200
D
0
0.01 0.1 0.6
Graphene volume fraction,
Fig. 5.5 The same data as in figure 6.4 but plotted as a function of GO-ODA volume fraction
on a log-log plot. The lines in (A) and (B) illustrate percolation-like behavior while the line
in (C) illustrates linearity. The vertical arrows illustrate the percolation threshold while the
horizontal arrows show that value of each property displayed by the polymer.
However, above this threshold, both properties increase with graphene content as a
power law. Such non-linear increases in mechanical properties of elastomeric
composites are generally described as the Payne effect. This is usually attributed to
either the effects of strong matrix-filler interaction or the formation of a filler
network [29]. Here the fact that essentially no increase in modulus (or 3% ) are
observed below 2.5 vol% suggests the latter mechanism to be at work. Indeed, this
lack of reinforcement at low volume fraction implies that neither polymer-filler stress
transfer nor hydrodynamic effects are important [29]. The increase in E and 3%
89
above 2.5 vol% implies that network formation begins at this critical volume
fraction. Network formation and growth can be described by percolation theory [30]
Where the formation of networks of the filler controls the mechanical properties of
composites, it has been shown that certain mechanical properties are described by a
percolation-like scaling law [21, 25]
A AP A0 c t
(5.2)
Here A can represent either E or 3% , is the “filler volume fraction, c is the
percolation threshold, i.e. the critical filler” volume fraction when a network first
forms and AP is the value of the relevant mechanical property at the percolation
threshold. A0 and t are constants. As can be seen from figure 5.5 A and B, both E and
3% are described very well by this expression. In both cases, the percolation
threshold was c=2.5 vol% while the percolation exponent was t=0.8. This can be
compared to percolation thresholds of 1.4vol% and 6wt% and exponents of 1.5 and 2
as reported by Ramorino and Liff respectively using nanoclays as fillers [21, 25].
These results imply that that reinforcement of TPU is not due to standard
mechanisms which rely on stress transfer at the polymer-filler interface or the effect
of the filler particles on polymer flow under stress. Rather, the critical factor is the
formation of a network which mechanically stiffens the material. We interpret this by
assuming that once the network forms, the GO-ODA flakes form a jammed system.
It has been shown that for high concentration carbon black dispersions, such a
system can be described by a shear modulus which obeys a percolation scaling law
[31]. We believe that in the case of a solid state composite, this behavior manifests
itself as the tensile modulus (and low-strain tensile stress) obeying a percolation
scaling law.
The formation of such a network is likely to impact on other mechanical properties.
We note that E and 3% are low strain properties. In figure 5.5 C and D we
consider high strain properties i.e. those associated with fracture; B and B . As
described above, the ultimate tensile strength, B , increases linearly (solid line) at
low strain. However, it is clear from figure 5.5 C that the point where linearity ceases
and B begins to fall off coincides with the percolation threshold. In addition, as
demonstrated in figure 5.5 D, the strain at break is reasonably constant up to the
90
percolation threshold, after which it falls dramatically with GO-ODA volume
fraction. It is not clear why this should be the case. However, it may be that polymer
chains in the vicinity of the graphene surface tend to have reduced mobility. This
may reduce the scope of large strain deformation resulting in premature failure.
With this in mind, it is worth considering the nature of the interfacial region.
Previously, it has been shown that ODA functionalized SWNTs selectively interact
with the hard segments of thermoplastic elastomers as indicated by differential
scanning Calorimetry and other techniques [13]. Interestingly, these composites
showed very similar mechanical properties to those studied here; an apparent linear
increase in modulus and a peak in strength close to 2% volume fraction. Also they
observed a slow decrease in strain at break as ODA-SWNTs were added, reaching
~600% for a nanotubes content of 10%, very similar to what was observed here.
Thus it is likely that the ODA functionalized graphene sheets interact predominately
with the PU hard segments. Anchoring of these hard segments at the GO-ODA
surface may immobilize adjacent soft segments, limiting polymer flow and
ultimately causing premature failure.
Conclusion
We have used graphene oxide covalently functionalized with octadecylamine as model
filler in polyurethane based composites. We find no appreciable increases in either
stiffness or low-strain stress for loading levels below 2.5vol%. However above this
threshold, both mechanical quantities increase as a power law. This behavior is
consistent with mechanical percolation. This implies that the graphene oxide platelets
are effectively isolated at low volume fractions but begin to form a network at a volume
fraction of 2.5vol%. This loading level can be thought of as a percolation threshold. As
the loading level is increased, the network becomes more extensive and the stiffness and
low-strain stress increase as described by the percolation scaling law. Interesting the
ultimate tensile strength initially increases but reaches a maximum at the percolation
threshold. Similarly the ductility is invariant with graphene content up to the
percolation threshold, after which it falls steadily. This work shows that the
mechanical properties of elastomers reinforced with graphene can depend on
parameters other than interfacial stress transfer. For example, the formation of a
network of filler particles which acts like a jammed system can dominate the
mechanical properties of the system.
91
References
[1] S. Stankovich, D.A. Dikin, G.H.B. Dommett, et al, Nature. 442 (2006) 282.
[2] M.J. Allen, V.C. Tung, R.B. Kaner. Chemical Reviews. 110 (2010) 132.
[3] D.R. Dreyer, S. Park, C.W. Bielawski, R.S. Ruoff, Chemical Society Reviews. 39
(2010) 228.
[4] P. Capkova, M. Pospisil, M. Valaskova, et al, Journal of Colloid and Interface
Science. 300 (2006) 264.
[5] Y.W. Zhu, S. Murali, W.W. Cai, et al, Advanced Materials. 22 (2010) 3906.
[6] K.W. Putz, O.C. Compton, M.J. Palmeri, et al, Advanced Functional Materials. 20
(2010) 3322.
[7] A.B. Bourlinos, D. Gournis, D. Petridis, et al, Langmuir. 19 (2003) 6050.
[8] G.X. Wang, X.P. Shen, B. Wang, et al, Carbon. 47 (2009) 1359..
[9] O.C. Compton, D.A. Dikin, K.W. Putz, et al, Advanced Materials. 22 (2008) 892.
[10] S. Niyogi, E. Bekyarova, M.E. Itkis, et al, Journal of the American Chemical
Society. 128 (2006) 7720.
[11] F.M. Blighe, W.J. Blau, J.N. Coleman, Nanotechnology. 19 (2008) 415709.
[12] D. Cai, K. Yusoh, M. Song, Nanotechnology. 20 (2009) 085712.
[13] U. Khan, F.M. Blighe, J.N. Coleman, The Journal of Physical Chemistry C. 114
(2010) 11401.
[14] A.V. Raghu, Y.R. Lee, H.M. Jeong, C.M. Shin, Macromolecular Chemistry and
Physics. 209 (2008) 2487.
[15] U. Khan, P. May, A. O'Neill, J.N. Coleman, Carbon. 48 (2010) 4035.
92
[16] U. Khan, P. May, A. O'Neill, J.J. Vilatela, A.H. Windle, J.N. Coleman. Small. 7
(2011) 1579.
[17] W.S. Hummers, R.E. Offeman, Journal of the American Chemical Society. 80
(1958) 1339.
[18] Y.S. Yun, Y. Bae, D.H. Kim, et al, Carbon. 49 (2011) 3553.
[19] S. Park, D.A. Dikin, S.T. Nguyen, R.S. Ruoff, Journal of Physical Chemistry C. 113
(2009) 15801.
[20] J.I. Paredes, S. Villar-Rodil, A. Martinez-Alonso, J.M.D. Tascon, Langmuir. 24
(2008) 10560.
[21] S. Stankovich, R.D. Piner, S.T. Nguyen, R.S. Ruoff. Carbon. 44 (2006) 3342.
[22] J.J. Benitez, M.A. San-Miguel, S. Dominguez-Meister, et al, Journal of Physical
Chemistry C. 115 (2011) 19716.
[23] J.N. Coleman, U. Khan, W.J. Blau, et al, Carbon. 44 (2006) 1624.
[24] T. Kuilla, S. Bhadra, D.H. Yao, et al, Progress in Polymer Science. 35 (2010) 1350.
[25] S.M. Liff, N. Kumar, G.H. McKinley, Nature materials. 6 (2007) 76.
[26] H. Koerner, W. Liu, M. Alexander, et al, Polymer. 46 (2005) 4405.
[27] N.G. Sahoo, Y.C. Jung, H.J. Yoo, J.W. Cho. Macromolecular Chemistry and
Physics. 207 (2006) 1773.
[28] H.K.F. Cheng, N.G. Sahoo, Y. Pan, et al, Journal of Polymer Science Part B:
Polymer Physics. 48 (2010) 1203.
[29] G. Ramorino, F. Bignotti, S. Pandini, et al, Composites Science and Technology. 69
(2009) 1206.
[30] D. Stauffer, A. Aharony, Introduction to Percolation Theory. 2nd ed. London:
Taylor & Francis 1985.
[31] V. Trappe, V. Prasad, L. Cipelletti, et al, Nature. 411 (2001) 772.
93
Chapter 6
Improved Adhesive Strength and
Toughness of Polyvinyl Acetate Glue
on Addition of Small Quantities of
Graphene
6.1 Objective Composites of polyvinyl acetate (PVAc) reinforced with solution exfoliated
graphene were prepared. A 50% increase in stiffness and a 100% increase in tensile
strength on addition of 0.1 vol % graphene compared to the pristine polymer was
observed. As “PVAc is commonly used commercially as glue, we have tested such
composites as adhesives”. The “adhesive strength and toughness of the composites
were up to 4 and 7 times higher”, respectively, than the pristine polymer.
6.2 Introduction Adhesives play a critical role of modern manufacturing and are essential in a wide
range of areas from packaging to electronics [1] to aerospace technology [2,3]. While
they come in many forms, possibly the simplest are synthetic thermoplastic
adhesives. Essentially, these are high concentration polymer solutions which can be
spread on the surfaces to be bonded. After the surfaces are brought into contact, the
solvent slowly evaporates to give a solid polymer which forms an effective bond.
In general, adhesives can fail cohesively or adhesively, that is within the bulk of the
adhesive or at the adhesive−surface interface. Many synthetic thermoplastic
adhesives form relatively strong interfacial bonds. In addition, when a porous
material such as wood is bonded, the adhesive can permeate into the pores, resulting
94
in mechanical interlocking and an increase in the bonded area [4]. This means that
the limitations of synthetic thermoplastic adhesives can sometimes be associated
with the mechanical properties of the polymer. Amorphous polymers tend to have
limited mechanical strengths which are generally below 50 Mpa [5]. In addition,
many of the thermoplastics commonly used as adhesives have a glass transition
temperature which is close to room temperature [6], resulting in limited thermal
stability of the bond.4 It is common practice to modify the properties of the adhesive
by the addition of additives. While such additives are usually included to alter the
adhesive properties [7−9], some researchers have used additives to improve the
mechanical properties of the adhesive [10,11]. In addition, it is worth noting that in
the last few years a small number of researchers have begun to explore using nano
materials as additives in adhesives [9−12]. One of the most commonly used
thermoplastic adhesives is polyvinyl acetate (PVAc) [4,10,13,14]. We note that this
material is not to be confused with polyvinyl alcohol (PVA), a polymer that has been
much studied as a nano composite matrix [15,16]. Generally found as a water-based
emulsion, PVAc is most often used as an adhesive for porous materials such as wood
and paper. As such, it generally forms a strong adhesive bond, and so, the adhesive
strength tends to be limited by the mechanical properties of the polymer. A number
of papers have described reinforcement [17], of PVAc with nano materials such as
carbon nanotubes [18] cellulose nano fibers [19] or nano clays [20]. Adhesives based
on PVAc loaded with small quantities of nano clays have even exhibited small but
significant increases in adhesive strength [10].
However, the adhesives studied all display some negative aspects. For example,
carbon nanotubes, while very promising as filler due to their extremely high strength
and stiffness [17], are ultimately impractical due to their high cost. At the other
extreme, nano clays are extremely cheap but do not have the superlative mechanical
properties displayed by nanotubes [21,22]. However, recently a new nano material
has become available which combines the high strength of carbon nanotubes with the
low cost of clays. Graphene is a two-dimensional sheet of sp2 bonded carbon which
has become renowned for its superlative properties [23], For example, “pristine
graphene has a modulus and strength” of 1 TPa and 130 GPa, respectively [24].
Originally produced in very small quantities [25], “graphene can now be produced in
large quantities by exfoliation [26], of graphite in solvents” [27], aqueous surfactant
solutions [28], or polymer solutions [29,30]. Already, graphene has displayed
95
significant success in reinforcing [31−34], both thermoplastics [35−37] and
elastomers [38,39] in some cases at very low loading level [36,40,41]. With this in
mind, graphene appears to be a promising additive for thermoplastic adhesives.
However, to the best of our knowledge, no work has been done in this area. In this
report, we use solution processing to prepare composites of PVAc and solvent
exfoliated graphene. We show that the addition of <1% graphene can result in a
doubling of the composite strength and stiffness without significant reduction in
ductility. In addition, we find the adhesive properties of the composite to be
significantly better than the neat polymer.
6.3. Experimental Procedure
Graphite powder (10 g, Sigma Aldrich) was exfoliated by sonicating (GEX600, 24
kHz, flat head probe, 25% amplitude) in 100 mL N-methyl pyrrolidinone (NMP)
(100 mg/mL) for 6 h. “The resulting dispersion was centrifuged at 1000 rpm for 45
min” (Hettich Mikro 22R). This results in the sedimentation of un-exfoliated graphite
and large graphene flakes. The sediment was collected and re-dispersed in fresh
NMP by sonicating in a sonic bath (Branson 1510E-MT) for 15 min. This dispersion
was centrifuged at 500 rpm for 45 min to remove the un-exfoliated graphite. The
supernatant, which is expected to contain reasonably large graphene flakes [42], was
retained. This supernatant was filtered through a nylon 0.45 micron membrane and
washed with 200 mL tetrahydrofuran (THF), resulting in a re-aggregated graphene
filter cake. Previous studies have shown that such materials tend to be free of defects
and oxides and consist of flakes of good quality graphene [27,43]. In addition, such
cakes are known to be easily re-dispersed in appropriate solvents [44, 45]. During
this work, it was found that re-aggregated graphene filter cakes could be effectively
re-dispersed, even in poor solvents such as THF. Such a dispersion (5 mg/mL),
prepared by bath sonication (Branson 1510E-MT, for 4 h) was used as a graphene
stock dispersion. While such dispersions are unstable, they can be stabilized by
subsequent addition of a polymer such as PVAc. If carefully chosen, the polymer can
partially bind to the graphene sheets stabilizing them against re-aggregation by the
steric mechanism [30]. Polyvinyl acetate (Sigma Aldrich, Mw = 100 000 g/mol) was
dissolved in THF at two concentrations, 30 and 200 mg/mL. These solutions were
96
blended with graphene/THF dispersion (5 mg/mL) in the required ratio to give the
desired graphene/PVAc mass fraction. The resulting mixtures were further bath
sonicated for 4 h to homogenize. These dispersions were stable with no visible
evidence of aggregation in the liquid phase. Dispersions were characterized by
depositing a drop of liquid onto a holey carbon transmission electron microscopy
(TEM) grid and analyzed using a Jeol 2100.
The composite dispersions with PVAc concentration of 30 mg/mL were poured into
Teflon trays and dried at room temperature for 24 h and then at 60 °C for 8 h. They
were cut into strips of thickness ~50 micron and lateral dimensions 2.5 mm × 20 mm
using a die cutter. “Tensile testing was performed with a Zwick Z100 at a strain rate
of 15 mm/ min. The fracture surfaces were imaged using a Zeiss Ultra scanning”
electron microscope (SEM) operating at 2 kV. The mass fractions were converted to
volume fraction assuming mass densities of ńG = 2100 kg/m3 and ńP = 1180 kg/m
3.
The composite dispersions with PVAc concentration of 200 mg/mL were used for
adhesive testing. In all cases, equal masses of the high concentration dispersion were
spread on a wood surface over a well-defined area. An identical piece of wood was
then pressed onto the glue. These assemblies were then placed in a custom built
holder and 0.042 MPa applied for three days at room temperature and further dried
over night at 60 °C. Both tensile and shear adhesive testing was performed. For
tensile tests, the wood pieces were in the shape of the letter T with the glue applied
to the top of the T over an area of 2.5 mm × 27 mm. During testing, the applied
stress was in a direction perpendicular to the glued surface. For shear tests, the wood
was in the shape of a bar with the glue applied to the side of the bar over an area of
10 mm × 14 mm. During testing, the applied stress was in a direction parallel to the
glued surface. In each case the strain rate was 0.1 mm/min. For both shear and
tensile measurements, 3−5 assemblies were tested for both polymer and composite
adhesives
6.4 Results and Discussion High concentration dispersions of graphene in THF (5 mg/ mL) were mixed with
solutions of PVAc in THF (30 mg/mL) to yield hybrid polymer-graphene
97
dispersions with graphene volume fractions in the range 0−0.85%. The “exfoliation
state of the graphene in these hybrid dispersions” can be assessed by TEM. Shown in
Figure 6.1 A and B are TEM images of typical exfoliated graphene flakes. They
appear to be of good quality, with no holes or other obvious defects.
Fig. 6.1 (A) Large numbers of multilayer graphene deposited on a holey carbon TEM grid.
(B) Individual graphene multilayer. (C) Photograph of PVAc−graphene films with mass
fractions of 0%, 0.2%, 0.4%, 0.7%, and 1.5% (volume fractions from 0−0.8%). SEM image
of (D) a PVAc and (E) a PVAc/graphene fracture surface
Shown in Figure 6.1 C are free standing films of PVAc and PVAc−graphene
composites (volume fractions of 0−0.84%). It can be seen that while the dispersion
is reasonably good, some aggregation cannot be avoided, even at low volume
fractions. This aggregation probably occurs during film drying due to the increasing
graphene/THF concentration. Figure 6.1 D and E show SEM images of the fracture
surfaces of PVAc and PVAc/ graphene films respectively. While the polymer film
98
shows a relatively featureless surface, the presence of graphene greatly alters the
film morphology with numerous graphene sheets observable.
We performed tensile tests on films with a range of mass fractions (Figure 6. 2). For
the polymer, the stress initially increases nonlinearly with strain.
The polymer yields at approximately 5% strain above which the stress falls on. This
behavior is in line with previous reports of the tensile response of PVAc [46],
although it is important to stress that the mechanical response of PVAc at room
temperature is very sensitive to strain rate [19]. The composites stress strain curves
show greater linearity at low strain but otherwise have broadly similar shapes to the
polymer.
Fig. 6.2 Stress−strain curves for the PVAc/graphene composite film studied in this work.
(inset) Stress−strain curves on a log−log scale. The dotted line represents linearity.
From these stress strain curves, we can obtain a number of mechanical parameters.
Shown in Figure 6.3 A is the Young’s modulus, Y, plotted as a function of graphene
volume fraction. The modulus increases linearly with graphene content from 0.75
GPa for the polymer to 1.5 GPa for the 0.1 vol % composite. The initial rate of
increase was dY/dVf = 530 GPa, reasonably close to the maximum value of 1 TPa
set by the graphene sheet modulus and the rule of mixtures [24,47]. It is likely that
this value is lower than 1 TPa because of the finite length of the flakes used in this
study, [36]. This result agrees well with the value of 680 GPa measured for
99
graphene/poly (vinyl alcohol) composites [36]. At higher volume fractions, the
modulus falls on before rising again albeit at a slower rate. This behavior may be
indicative of aggregation. We note that the initial increase is competitive with
published data (expressed in terms of filler mass fraction, Mf) for PVAc reinforced
Fig. 6.3 Mechanical properties of PVAc films. (A) Young’s modulus, (B) ultimate tensile
strength, and (C) strain at break, as a function of graphene volume fraction.
with cellulose nano fibers (dY/dMf ≈ 80 GPa) [19] carbon nanotubes (dY/dMf ≈ 200
GPa),[18], and nano clays (dY/dMf ≈ 340 GPa) [20] (The last value was calculated
for only two data points so must be treated with caution. The vast majority of
clay−polymer composites show much lower reinforcement [21].
Very similar behavior was observed for the ultimate tensile strength, óB, which
increased linearly from 21 MPa for the polymer to 38 MPa for the 0.1 vol %
composite with a slope of dóB/dVf = 15 GPa (Fig. 6.3B). Such a large increase at
such a low loading level is impressive and is generally only found for high
100
performance nano fillers. For example, this result compares well to the value of
dóB/dVf = 22 GPa measured for graphene/ polyvinyl alcohol composites, [36].
Again, this value is also similar to published data for PVAc reinforced with carbon
nanotubes (dóB/dMf ≈ 10 GPa) [18], but much higher than equivalent data for
cellulose nano fibers (dóB/dMf ≈ 0.2 GPa) [19]. However, the slope is much less
than the value of 130 GPa predicted by the graphene sheet strength and the rule of
mixtures, [24,47]. However, this probably means that the flake length is below the
critical length, [48] (expected to be of order of many micrometers [29,49]). Under
such circumstances, material fracture generally involves failure of the polymer
graphene interface rather than breaking of the flakes [5,36,48]. Under these
circumstances, we can write:
dóB/dVf ≈ ôB[ L + w ]/4 t (6.1)
where ôB is the interfacial strength [36]. Using the flake dimensions given above,
this means ôB ≈ 27 MPa, similar to the value of 29 MPa recently measured for
graphene/PVA composites [36]. Indeed, given the structural similarities between
PVAc and PVA, it is hardly surprising that their interfaces with graphene have
similar shear strength.
We note that both dY/dVf and dóB/dVf values we have measured for PVAc−graphene composites are quite high as discussed above. That the value of
dY/dVf is high implies that the polymer-graphene interfacial stress-transfer is very
effective while the relatively large value of dóB/dVf implies a strong polymer-
graphene interface. Taken together, this suggests a strong interaction between PVAc
and graphene. As described above, a similarly strong interaction is observed for PVA−graphene composites [36]. The detailed nature of these interactions is not well-
understood. However, we suggest that the results described above are consistent with
the hydrogenated parts of the polymer chain binding strongly to the graphene by
dispersive interactions. It is likely that the polar acetate group (or hydroxyl group in
the case of PVA) protrudes outward and so is available to interact with other
polymer chains. However, molecular dynamics simulations are required to test this
hypothesis.
101
The strain at break appeared to increase slightly from -100% for the polymer to -
175% for the 0.23vol% composite sample before subsequently falling. This is
slightly unusual as ductile polymers usually display a decrease on strain at break on
the addition of nano fillers such as nanotubes or graphene [39,50−52]. Indeed
previous work on PVAc filled with nanotubes or nano clays showed a reduction in
ductility for all filler contents [18,19]. It is not clear why this should be the case.
However, for polymers which fail by craze formation, if the fibular bridges were
reinforced by the presence of the nanofiller, this might result in an increase in
ductility in the composite.
Because one of the most common applications of PVAc is as an adhesive [10,13,14].
We tested the effect of adding graphene on the adhesive properties of PVAc. We
prepared very high concentration solutions of PVAc in THF (200 mg/mL) both with
and without the presence of various amounts of graphene from 0.2 to 3 wt %. These
viscous liquids were then coated on pieces of wood over a well-defined area as an
adhesive. Identical pieces of wood were then pressed onto the adhesive in geometries
designed to test both the tensile and shear properties of the adhesive (Figure 6.4A).
The glued assemblies were then pulled apart using a tensile tester (Figure 6.4B).
Typical stress-strain curves for polymer and composite adhesives, tested in both
tensile and shear geometries, are shown in Figure 6.4C For both shear and tensile
measurements, the stress strain curve looked very different to the tensile stress strain
curves of the PVAc and PVAc/graphene composites shown in Figure 6.2. Indeed,
this suggests that the mechanical properties of the bond are not controlled solely by
the mechanical properties of the adhesive.
The tensile adhesive strength increased sub linearly from 0.3 MPa for the pure
polymer to 0.75 MPa for the 3 wt % composite. The shear strength increased linearly
from 0.5 MPa for the PVAc to 2.2 MPa for the 4 wt % sample. Interestingly, the
initial rate of increase of both shear and tensile adhesive strength is similar at ~50
MPa. This is considerably lower than the rate of increase of composite tensile
strength with graphene mass fraction again indicating that the bond strength is not
solely limited by the strength of the composite. This suggests that failure may be
adhesive rather than cohesive. We can compare this with Kaboorani et al. [10], who
tested PVAc filled with 4% nano clay. They achieved 25% increase in adhesive
102
strength, albeit from a much higher base (the shear strength of their commercial
PVAc adhesive was 19 MPa).
We also measured the area under the stress-displacement curve for each test. This
parameter is equal to the energy cost per unit area of breaking the bond between the
wood pieces and can be considered the adhesive toughness. This data is shown in
Figure 6.4E. For both tensile and shear tests, the toughness increases dramatically
with graphene addition up to 1.5 wt % with some fall on observed for the tensile case
at higher graphene content. However, the tensile adhesive toughness increased by
more than 3-fold for 0.7 wt % graphene addition while the shear toughness had
increased by almost 4-fold for the 3 wt % sample. This is an important result as it
shows that graphene-containing adhesives can absorb significantly more energy
before failure than the polymer adhesive alone. We note that the adhesive strength in
both tensile and shear modes was less than 3 MPa. Commercially available PVAc
glues can have strengths of up to 7 MPa for a range of woods [4,13,14].
However, such glues tend to be complex mixtures of PVAc and a range of additives,
which have been developed over decades. In comparison, our PVAc adhesives were
deposited from simple PVAc solutions. It is important to assess the efficiency of
graphene addition to commercially available PVAc wood glue. To test this, we
purchased Tonic Studio Craft Glue PVAc wood glue. The concentration of solids
(mainly PVAc) in the glue was measured by drying a known volume of glue (1 mL)
at 60°C for 3 days to remove the solvent (water) followed by weighing.
The commercial glue was then mixed with a 5 mg/ mL Graphene/THF stock
solution. Excess solvent was evaporated to bring the glue back to its original
concentration (although now dissolved in a THF/water mixture rather than pure
water). Shear and tensile tests were carried out as before both on samples bonded
with as-purchased glue and those bonded with commercial glue with graphene added
(During the graphene addition process, one sample was prepared with processing
identical to the composites but with no added graphene. This sample is included in
the composite glue data set but with graphene content = 0). Representative stress-
displacement curves are shown in Figure 6.4F and were found to be considerably
different to those measured before, possibly due to the presence of additives in the
commercial glue.
103
Fig. 6.4 Measurements of adhesive properties of PVAc/graphene glue. (A) Photograph of
samples used for adhesive testing. (left) Two T-shaped wood pieces glued together for
tensile testing. (right) Two wooden bars, glued together along an overlapping region (dashed
line), for use in shear measurements. (B) Photograph of T-shaped pieces during a tensile test.
(C) Applied stress plotted as a function of displacement in both tensile and shear modes for
samples glued using homemade PVAc adhesive. (D) Tensile and shear bond strength and
(E) toughness as a function of graphene content for the homemade PVAc adhesives. (F)
Tensile stress−strain curves for as-bought commercially available glue and same with 0.7 wt
% graphene added. (G) Tensile and shear bond strength and (H) toughness as a function of
graphene content for the adhesives prepared with commercially available PVAc glue. The
dotted lines represent the untreated glue. The data points represent the glue, diluted and re-
concentrated during the process of graphene addition.
104
We found no significant improvement in the adhesive shear strength on addition of
graphene. However, small but significant changes were observed for the tensile
adhesive strength. On addition of graphene, the tensile adhesive strength increased
linearly from 1.25 MPa for the glue reference sample to 1.75 MPa for the sample
containing 0.7 wt % graphene before falling at higher loading levels. Importantly, we
found that the dilution/re-concentration procedure used to add the graphene had no
effect on the tensile adhesive strength of the graphene-free glue; identical values
were found for the pristine PVAc glue and PVAc glue that had been treated
identically to the composites but with no graphene added. This shows that graphene
addition can have a positive effect on commercial PVAc glue.
We also calculated the adhesive toughness for all glues based on the commercial
adhesive. This data is shown in Figure 6.4H. Increases in both tensile and shear
toughness were observed. The tensile adhesive toughness increased from 0.2 kJ/m2
for the as-purchased glue to 1.5 kJ/m2 for the 0.7 wt % sample, a >7-fold increase. It
is worth noting that this increase in toughness is mostly due to increases in
displacement at failure (see Figure 6.4B) on addition of graphene. A much smaller
but still significant increase in the shear toughness was observed.
It is worth considering the mechanism of failure. Under stress, it is known that
cavities begin to form in the adhesive.11
When failure is cohesive these cavities tend
to be wholly contained within the adhesive. Cavity formation tends to first occur
close to the yield stress (i.e., the maximum stress observed in the stress strain curves
in Figures 6.3 and 6.4B) [11]. Once the cavities have formed, the stress is maintained
by fibrils in a manner similar to crazing in polymers [5]. As the displacement is
increased the cavities expand and the fibrils become extended. This process
dissipates considerable amounts of energy, often resulting in high adhesive
toughness. Failure occurs when the last fibril breaks. Such fibrils can be observed in
Figure 6.4B just before failure. The addition of graphene results in increases in
adhesive stress because graphene both stiffens and strengthens the polymer resulting
in cavity formation at higher stress and the fibrils resisting deformation with greater
stress. The increased work of adhesion is largely due to failure occurring at higher
displacements and is due to the reinforcement of the fibrils which delays failure to
higher displacements.
105
Conclusion
In conclusion, we have shown that the polymer PVAc can be mechanically
reinforced by addition of solvent exfoliated graphene. Addition of 0.1 vol %
graphene results in the doubling of modulus, strength, and ductility. When used as an
adhesive, addition of 0.7% graphene results in increases in both adhesive strength
and toughness. We believe graphene shows great promise as an additive for
adhesives. It is produced from a precursor, graphite, which is very cheap making it
economically plausible. In addition, the results presented here represent only the first
tentative steps in this area. Further work is likely to see further advances in both
strength and toughness of graphene reinforced adhesives.
References [1] B.G.Yacobi, S. Martin, K. Davis, A.Hudson, M.Hubert, J. Appl. Phys., 91
(2002) 6227.
[2] J. Gegner, Materialwiss. Werkstofftech. 39 (2008) 33.
[3] S. Park, Y.W.J. Choi, H.S. Choi, et al, J. Adhes. 86 (2010) 192.
[4] Y. Hatano, B. Tomita, H. Mizumachi, Holzforschung 40 (1986) 255.
[5] W.D. Callister, Materials Science and Engineering an Introduction, 7th ed.; Wiley:
New York, 2007; p 721.
[6] W.W. Lim, H. Mizumachi, J. Appl. Polym. Sci. 66 (1997) 525.
[7] N. Pastor-Sempere, J.C. Fernandez-Garcia, A.C. Orgiles-Barcelo, et al, J.
M. J. Adhes. 59 (1996) 225.
[8] A. Torro-Palau, J.C. Fernandez-Garcia, A.C. Orgiles-Barcelo, et al,
J. Adhes. 57 (1996) 203.
[9] L. L. Zhai, G. P. Ling, J. Li, Y.W. Wang, Mater. Lett. 60 (2006) 3031.
[10] A. Kaboorani, B. Riedl, Compos. Part A 42 (2011) 1031.
[11] T. Wang, C.H. Lei, A.B. Dalton, C. Creton, et al, Adv. Mater. 18 (2006)
2730.
[12] E.N. Gilbert, B.S. Hayes, J.C. Seferis, Polym. Eng. Sci. 43 (2003) 1096.
106
[13] E. Burdurlu, Y. Kilic, G.C. Eli’bol, M. Kilic, J. Appl. Polym. Sci. 100 (2006) 4856.
[14] O. Dajbych, D. Herak, A. Sedla ek, G. G rdil, es. Agric. Eng. 56 (2010)
159.
[15] A.B. Dalton, S. Collins, E. Munoz, J.M. Razal, et al, Nature 423 (2003)
703.
[16] B. Vigolo, A. Penicaud, C. Coulon, C. Sauder, et al, Science 290 (2000)
1331.
[17] J.N. Coleman, U. Khan, W.J. Blau, Y.K. Gun’ko, Carbon 44 (2006) 1624.
[18] R.D. Maksimov, J. Bitenieks, E. Plume, et al, Mech.Compos.Mater. 46 (2010) 237.
[19] G. Gong, J. Pyo, A.P. Mathew, K. Oksman, Compos. Part A 42 (2011) 1275.
[20] Y. Mansoori, A. Akhtarparast, M.R. Zamanloo, et al,Polym.Compos. 32
(2011)1225.
[21] S. Pavlidou,C.D. Papaspyrides, Prog. Polym. Sci. 33 (2008) 1119.
[22] L.J. Zhu, K.A. Narh, J. Polym. Sci. Part B: Polym. Phys. 42 (2004) 2391.
[23] A.K. Geim, Science 324 (2009) 1530.
[24] C. Lee, X. D. Wei, J.W. Kysar, J. Hone, Science 321 (2008) 385.
[25] K.S. Novoselov, D. Jiang, F. Schedin, et al, Proc. Natl. Acad. Sci. U.S.A. 102 (2005)
10451.
[26] J.N. Coleman, Acc. Chem. Res. 46 (2013) 14.
[27] Y. Hernandez, V. Nicolosi, M.Lotya, et al, Nat.Nanotechnol. 3 (2008) 563.
[28] M. Lotya, Y. Hernandez, P.J. King, et al, J. Am.Chem. Soc. 131 (2009) 3611.
[29] U. Khan, P. May, A. O’Neill, A.P. Bell, et al, Nanoscale 5 (2012) 581.
[30] P. May, U. Khan, J.M. Hughes, J.N. Coleman, J. Phys. Chem. C 116 (2012) 11393.
[31] H. Kim, A.A. Abdala, C.W. Macosko, Macromolecules 43 (2010) 6515.
[32] T. Kuilla, S. Bhadra, D.H. Yao, N.H. Kim, et al, Prog. Polym. Sci. 35 (2010) 1350.
[33] R. J. Young, I.A. Kinloch, L. Gong, K.S. Novoselov, Compos. Sci.
Technol. 72 (2012) 1459.
107
[34] J.R. Potts, D.R. Dreyer, C.W. Bielawski, R.S. Ruoff, Polymer 52 (2011) 5.
[35] U. Khan, K. Young, A. O’Neill, J.N. Coleman, J. Mater. Chem. 22 (2012) 12907.
[36] P. May, U. Khan, A. O’Neill, J N. Coleman, J. Mater. Chem. 22 (2011) 1278.
[37] T. Ramanathan, A.A. Abdala, S. Stankovich, D.A. Dikin, et al, Nat.
Nanotechnol. 3 (2008) 327.
[38] O. Menes, M. Cano, A. Benedito, E. Gimenez, P. Castell, et al, Compos.
Sci. Technol. 72 (2012) 1595.
[39] U. Khan, P. May, A. O’Neill, J.N. Coleman, Carbon 48 (2010) 4035.
[40] M.A. Rafiee, J. Rafiee, Z. Wang, et al, ACS Nano 3 (2009) 3884.
[41] S. Morimune, T. Nishino, T. Goto, Polym. J. 44 (2012) 1056.
[42] U. Khan, A. O’Neill, H. Porwal, P. May, K. Nawaz, J.N. Coleman, Carbon 50
(2010) 470.
[43] U. Khan, A. O’Neill, M. Lotya, S. De, J. N. Coleman, Small 6 (2010) 864.
[44] U. Khan, H. Porwal, A. O’Neill, K. Nawaz, P. May, J.N. Coleman, Langmuir 27
(2011) 9077.
[45] A. O’Neill, U. Khan,; P.N. Nirmalraj, J.J. Boland, J.N. Coleman, J. Phys.
Chem. C 115 (2011) 5422.
[46] S. S. Ochigbo, A. S. Luyt, W.W. Focke, J. Mater. Sci. 44 (2009) 3248.
[47] G.E. Padawer, N. Beecher, Polym. Eng. Sci. 10 (1970) 185.
[48] D. Hull, T.W. Clyne, Cambridge University Press: New York, (1996) 344.
[49] L. Gong, I.A. Kinloch, R.J. Young, I. Riaz, et al, Adv. Mater. 22 (2010)
2694.
[50] F.M. Blighe, W.J. Blau, J.N. Coleman, Nanotechnology 19 (2008) 415709.
[51] U. Khan, F.M. Blighe, J.N. Coleman, J. Phys. Chem. C 114 (2010) 11401.
[52] U. Khan, P. May, A. O’Neill, J.J. Vilatela, A.H. Windle, J.N. Coleman,
Small 7 (2011) 1579.
108
Chapter 7
The Effect of Graphene nanosheets on
the Mechanical Properties of
Polyvinylchloride
7.1 Objective
Graphite was exfoliated to graphene using sonic tip and N-methyl-2- Pyrrolidinone
(NMP) as solvent and after specific time the exfoliated graphite (unexfoliated
graphite) was centrifuged at 500 rpm for 30 min. The supernatant of centrifuged
material was stored and filtered. Then the filtered graphene nanoflakes were re-
dispersed in tetrahydrofuran (THF) and sonicated in bath sonicator to make
homogenous dispersion. Then this dispersion was used as nanofiller in PVC as
reinforcement. 03 mg/ml concentration of graphene in THF was used as nanofiller to
PVC matrix. Thin film of composites were prepared by using drop casting and
annealing procedure. An excellent improvement in mechanical properties was
observed. At 1.5% loading tremendous improvement in mechanical properties was
noted. Modulus improved from 1.31Gpa to 2.14Gpa (75 % increment) and UTS
improved from70 Mpa to 83.2Mpa. There was no fall in elongation at break along
with these improvements at this loading.
7.2 Introduction
Recently “discovered planar 2D form of carbon known as graphene has become one
of the most exciting materials today because of its unique properties” [1]. “Individual
graphene sheets show high values of thermal conductivity [2], Young’s modulus [3],
large surface area” [4], ballistic transport on submicron scales and mass less “Dirac-
fermion charge carrier abilities [5,6]. These properties make graphene a promising
material for using in many applications such as photovoltaic devices, sensors”,
109
transparent electrodes, “super capacitors conducting composites [4,7–12]. At present,
carbon-based reinforcing materials employed in polymer composites are dominated
by carbon nanotubes (CNT). But difficulty in dispersing CNT and high cost of
production limits its widespread use”. Challenge is to find an alternative for CNT for
which graphene can prove itself as a suitable candidate due to remarakable and
outstanding mechanical “properties and ultra large interfacial surface area [3,4].
Incorporation of graphitic nano flakes into elastomeric polymer matrix generates
high performance composites with improved mechanical and functional properties
[12–15]. Other interesting properties such as high dielectric permittivity” and “low
percolation threshold have also been observed in graphene incorporated composites
of poly (vinylidene fluoride) and polystyrene, respectively” [16,17]. Recently,
“graphene oxide (GO) has also been used as a filler in various polymer matrices, due
to its hydrophilicity and ease of formation of stable colloidal suspensions” [13,14].
Besides, “functionalized graphene sheets (FGS) are also employed as they provide
better interactions with the host polymers compared to unmodified CNT or
traditional expanded graphite (EG) [12]. In our current investigation”, “poly(vinyl
chloride) (PVC) is chosen as the host polymer matrix, because of its wide range of
applications, low cost, chemical stability, biocompatibility” [18]. However, “PVC
has low thermal stability, which hinders some of its applications [19]. The present
day challenge is to introduce thermal stability along with high mechanical strength
for PVC with the use of minimum amount of fillers”. “Substantial amount of work
has been carried out in the past few decades towards this goal [20–22]. Fillers such
as clay [23,24], “wood flour” [21], wood fibers [25], agricultural residues [26],
cellulose whiskers [27] and calcium carbonate [28] were used to improve the thermal
and mechanical stability of PVC. In recent times, CNT have also been identified as a
suitable filler material for PVC” [29]. “Kevlar coated CNT used as additives to PVC
resulted in composites with improved mechanical properties demonstrating” up to 50
and 70% increase in “tensile strength and Young’s modulus respectively at very low
CNT loading [30]. Similarly, CNT grafted with styrene-maleic anhydride
copolymers (SMA) was found to enhance the interaction with PVC matrix and both
thermal and mechanical stabilities improved considerably” [31]. But “dispersion of
CNT in organic solvents is a challenge, which is very critical for the preparation of
polymer composites” [13]. In our previous study we have used the soluble graphene
110
nano flakes [32] as reinforcing “filler for PVC at a very low loading level. In order to
have efficient reinforcement in polymer composites, it is important to have
molecular level dispersion in the polymer matrix”. In this case, both PVC and
graphene sheets can be readily dispersed in Tetrahydrofuran (THF) for “solution
blending, which will help to achieve molecular level dispersion”. In present study we
have introduced exfoliated graphene nano flakes at different wt % to study its effect
on the mechanical strength of PVC polymer. The loading was from 0.1 wt% to 10
wt%. And the neat PVC sample was taken as reference material.
7.3 Experimental section 100 mg/ml dispersion of graphite flakes (Sigma –Aldrich) were taken in N-Methyl
Pyrrolidinone (NMP) and was sonicated using sonic tip (GEX600,48W,24khz,flat
head probe) continuously for 24 hours keeping temperature between 5oC—10
oC by
using ice bath, so that the heat generated during sonication may not spoil the solvent
(NMP). As a result the obtained exfoliated graphite was separated from un-exfoliated
by centrifugation at 500/45 (rpm/minutes) and supernatant was collected, filtered
through a nylon membrane of pore size 0.45 micron (Sterlitech). The membrane
supported graphene nano flakes were dried at 60oC for accurate mass. Then its
dispersion was made in tetrahydrofuran (THF) and used in thin films casting.
Similarly 60 mg/ml solution of PVC was made in THF by dissolving 6 gm of PVC,
in 100 ml of solvent for about 24 hours.
The composites samples were prepared using different weight percent of graphene to
PVC solution between 0.1 to 13 wt. % and a reference sample was also prepared
using same recipe. The total weight of sample was 150 mg and with constant volume
of 13 ml. These samples were sonicated for four hours in bath sonicatior and were
cast in Teflon trays of dimension 4x4x2. These cast samples were dried in vacuum
oven at room temperature at 900 mbar. Then these samples were shifted to another
oven and kept there for eight hours to remove traces of THF, if any present.
7.4. Characterization of PVC-Graphene Composites
The graphene nano flakes were studied using TEM (Joel 2100, Japan) and hole
carbon grids (400mesh). Mechanical “testing was performed using a Zwick Roell
tensile tester with 100N load cell at a strain rate” of 50mm/min.
111
7.5. Results and Discussion
TEM images of graphene nano flakes exfoliated in NMP as solvent under tip
sonicator were taken to study its flakes size and number of layers. It is very clear
from TEM images that most of the graphene flakes are consist of less than three
sheets with a thickness of 1.4 nm, which indicates that it consists of less than three
layers as shown in Figure 7.1.
Fig. 7.1 TEM images of graphene nano flakes exfoliated in NMP
It was expected that the mechanical properties of PVC would be enhanced to a great
extent, although it is improved but not to very high extent. Figures 8.2-4 shows the
effect of graphene nanoflakes on mechanical properties of composites for a range of
loading levels. i.e. Young’s Modulus, UTS and elongation. For blank PVC the
Young’s Modulus is 131Mpa and its Ultimate Tensile Strength (UTS) is 70Mpa. Just
at 1.5% loading, Modulus improved from 1.31 to 2.14 GPa (61% improvement) as
shown Figure 7.2.
112
Fig. 7.2 Effect on Young’s Modulus of PVC after using graphene nanoflakes
A comparison is of theoretical and experimental data is also shown in Figure 7.5. At
the same time UTS reached to 83.2 MPa (18% improvement) from 70 MPa for 1.5%
graphene loading Figure 7.3.
Improvement in UTS may be due to good dispersion/interaction of polymer and
graphene nano sheets [33] which is expectable at this low level addition of nanofiller.
While Young’s Modulus enhanced to good extent (2.29 GPa for 8 wt.% loading) and
its Ultimate Tensile Strength (UTS) dropped from 83.2 MPa (for 1.5 wt % loading)
to 72 MPa. This decrease in UTS may be due to agglomeration of nanofiller and/or
improper dispersion in polymeric matrix. Likewise increase in mechanical resistance
properties (Modulus and UTS) results in decrease in ductility in terms of elongation
at break. But in our case at 1.5% loading, we not only get improvement in
mechanical properties in terms of UTS and modulus but there was negligible
decrease in elongation at break from 15.3% to 13% Figure 7.4. But at 8% loading the
elongation at break affected disastrously and dropped from 15.3% to 6.8%.
This fall in elongation at break may be due to interaction of graphene with polymer
chain which restricts the movement of polymeric chain [34]. A comparison is shown
in From these results it is concluded that 1.5% loading is critical loading on which
we get very good mechanical properties in terms of Modulus, UTS and elongation at
break. While above this loading there is downtrend in mechanical properties
113
especially in UTS and elongation at break.
The Halpin-Tsai model was used [35-38] in our work for random distribution of
graphene nanofiller in polymer to simulate our graphene/PVC nano composites and
its equation for randomly distribution is given as follow:
In equation 7.1, Ec is the modulus of the nano composites with randomly distributed
graphene nano flakes, where as Vc is volumetric fraction of graphene in polymer.
Likewise Eg and Em are the modulus of graphene nano flakes and polymer (PVC)
which are 1 Tpa [39] and 1.31 Gpa respectively. ξ, 1g and tg are the aspect ratio,
length and thickness of graphene nano flakes While the density of PVC & and of
graphene is1.4gm/mL and 2200kg/m3 respectively` [40]. The statistical average of
the length and thickness of graphene nano flakes were about 1µm and 1.2nm
respectively as determined by TEM. [32]. This gives aspect ratio value of 556.
Putting all these values in above equations we can easily deduce the theoretical
modulus for randomly distributed graphene nano flakes.
114
Fig. 7.3 Effect on UTS of PVC after using graphene nanoflakes
Fig. 7.4 Effect on Elongation at Break of PVC after using graphene nanoflakes It is worth mentioning that two tendencies have been observed in our experiments.
One is at low loading up to 1.5% having close resemblance between theoretical data
Halpin-Tsai data and second is experimental one which is clear from Figure 7.5.
By increasing the contents of filler in polymer, there is clear difference between
theoretical and experimental data. It may be due to the dispersion of graphene nano
flakes.
115
Fig. 7.5 Comparision of Theoratical and Experimental values of Young’s Modulus
At low level of loading, nano fillers are homogenously dispersed while at high
loading it may form agglomerate. So we can say that solution blending is efficient
for low loading but it may not work for high loading (beyond 1.5 wt.%). It is clear
from data that small addition up to 1.5% gives excellent improvement in tensile
strength and modulus.
Conclusion
Here we used in our experiments graphene nano flakes as nanofiller in Polyvinyl
chloride as polymeric matrix. An interesting phenomenon was observed at 1.5 wt. %
loading both mechanical properties UTS and Young Modulus was improved to a
good extent while elongation at break was slightly affected, and this trend was
prevalent up to 1.5wt % loading but beyond this loading up to 8 wt % no
improvement was observed in terms of UTS and elongation at break. Elongation at
break was disastrously affected. The effect on elongation at break may be due to the
restacking of graphene flakes in polymer which restricts the movement of polymer
chain.
116
References [1] K.S. Novoselov, A.K. Geim, S.V. Morozov, D. Jiang, Y. Zhang, S.V. Dubonos, et
al, Science 306 (2004) 666.
[2] A.A. Balandin, S. Ghosh, W. Bao, I. Calizo, D. Teweldebrhan, F. Miao, et al, Nano
Lett. 8 (2008) 902.
[3] C. Lee, X.D. Wei, J.W. Kysar, J. Hone, Science 321 (2008) 385.
[4] M.D. Stoller, S.J. Park, Y.W. Zhu, J.H. An, R.S. Ruoff, Nano Lett. 8 (2008) 3498.
[5] Y.B. Zhang, Y.W. Tan, H.L. Stormer, P. Kim, Nature 438 (2005) 201.
[6] K.S. Novoselov, A.K. Geim, S.V. Morozov, D. Jiang, M.I. Katsnelson,
I.V. Grigorieva, et al, Nature 438 (2005) 197.
117
[7] Z.F. Liu, Q. Liu, Y. Huang, Y.F. Ma, S.G. Yin, X.Y. Zhang. Adv. Mater. 20 (2008)
3924.
[8] J.T. Robinson, F.K. Perkins, E.S. Snow, Z.Q. Wei, P.E. Sheehan. Nano Lett. 8
(2008) 3137.
[9] P.K. Ang, W. Chen, A.T.S. Wee, K.P. Loh. J Am. Chem. Soc. 130 (2008) 14392.
[10] J.B. Wu, H.A. Becerril, Z.N. Bao, Z.F. Liu, Y.S. Chen. Appl Phys Lett. 92 (2008)
263302.
[11] S. Stankovich, D.A. Dikin, G.H.B. Dommett, K.M. Kohlhaas, E.J. Zimney, E.A.
Stach, et al, Nature 442 (2006) 282.
[12] T. Ramanathan, A.A. Abdala, S. Stankovich, et al, Nat Nanotechnol. 3 (2008) 327.
[13] X. Zhao, Q. Zhang, D. Chen, Macromolecules 43 (2010) 2357.
[14] J. Liang, Y. Huang, L. Zhang, Y. Wang, Y.F. Ma, T. Guo, et al, Adv. Funct. Mater.
19 (2009) 2297.
[15] K. Kalaitzidou, H. Fukushima, L.T. Drzal, Compos Part A – Appl Sci. 38 (2007)
1675.
[16] S. Ansari, E.P. Giannelis. J Polym Sci. Part B 47 (2009) 888.
[17] N. Liu, F. Luo, H. Wu, Y. Liu, C. Zhang, J Chen. Adv. Funct. Mater. 18 (2008)
1518.
[18] C.E. Wilkes, J.W. Summers, C.A. Daniels, M.T. Berard, PVC handbook. 1st ed.
Germany: Hanser Verlag; (2005) 414.
[19] B. Iva´n, T. Kelen, F. Tu¨do¨s, Amsterdam, Netherlands: Elsevier Science (1989)
483.
[20] S. Mayeda, N. Tanimoto, H. Niwa, M. Nagata, J Anal. Appl. Pyrol. 33 (1995) 243.
[21] H. Djidjelli, J.J. Martinez-Vega, J. Farenc, D. Benachour, Macromol. Mater. Eng.
287 (2002) 611.
[22] S.Y. Tawfik, J.N. Asaad, M.W. Sabaa. Polym. Degrad. Stabil. 91 (2006) 385.
[23] T. Peprnicek, A. Kalendova, E. Pavlova, et al, Polym. Degrad. Stabil. 91 (2006)
3322.
118
[24] M.E. Romero-Guzman, A. Romo-Uribe, E. Ovalle-Garcia, R. Olayo, et al, Polym.
Adv. Technol. 19 (2008) 1168.
[25] L.M. Matuana, C.B. Park, J.J. Balatinecz. Polym. Eng. Sci. 38 (1998) 1862.
[26] S.T. Georgopoulos, P.A. Tarantili, E. Avgerinos, et al, Polym. Degrad. Stabil. 90
(2005) 303.
[27] L. Chazeau, J.Y. Cavaille, G. Canova, et al, J Appl. Polym. Sci. 71 (1999) 797.
[28] S.S. Sun, C.Z. Li, L. Zhang, H.L. Du, et al, Polym. Int. 55 (2006) 158.
[29] Y. Mamunya, A. Boudenne, N. Lebovka, L. Ibos, et al, Compos. Sci. Technol. 68
(2008) 1981.
[30] I O’Connor, H Hayden, S O’Connor, JN Coleman, et al, J Mater. Chem. 18 (2008)
5585.
[31] G.J. Wang, Z.H. Qu, L. Liu, Q. Shi, et al, Mat. Sci. Eng: A 472 (2008) 136.
[32] U. Khan, H. Porwal, A. O‖Neill, K. Nwaz, P. May, J.N. Coleman, Langmuir 27
(2011) 9077.
[33] J. Pascual, F. Peris, T. Bronat, O. Fenollar, R. Balart, Polym. Eng. Sci. 52 (2012)
733.
[34] S. Vadukumpully, J. Paul, N. Mahanta, S. Valiyaveettil, Carbon 49 ( 2011 ) 1 9 8.
[35] R.R. Tiwari, K.C. Khilar,U.J. Natarajan, J. Appl. Polym. Sci. 108 (2008) 1818
[36] K. Kalaitidou, H.Fukushima,H. Miyagawa, L.T. Drazal, Polym. Eng. Sci. 47
(2007) 1796
[37] J.C. Halpin,J.L. Kardos, Polym. Eng. Sci. 16 (1976) 344.
[38] F.T. Cerezo,C.M.L. Preston, R.A. Shanks, Macromol. Mater. Eng. 292 (2007) 155
[39] U. Khan, P. May , A. O´Neill, J. N. Coleman, Carbon 48 (2010) 4035.
[40] U. Khan, A. O´Neill, M. Lotya, S de, J.N. Coleman. Small 6 (2010) 864.
119
Chapter 8
Effects of selected size of graphene
nanosheets on the mechanical
properties of Polyacrylonitrile (PAN)
Polymer
8.1 Objective Mechanical properties of Poly (acrylonitrile) (PAN) polymer can be remarkably
improved by incorporation of grapheme nanosheets of different sizes. For this
purpose Graphite was exfoliated to graphene using sonic tip in the presence of N-
methyl pyrrolidinone (NMP) as solvent. Exfoliated graphene was separated from un-
exfoliated graphitic crystallites using selected speed (rpm) of centrifuge for specific
time. Then these exfoliated graphene nanosheets were further classified into two
different categories on the basis of flake size, i.e. 1 µm and 3.5 µm on the basis of
different speed of centrifuge. Later on these graphene sheets were incorporated into
PAN to study the effects of these different flake sizes on mechanical properties.
Different mechanical properties such as Young’s modulus, ultimate tensile strength
(UTS) and elongation at break were studied. It was observed that the Young’s
modulus and UTS were improved about 45% & 25% respectively for 3.5 µm
graphene flake size and 40% & 21% for 1 µm graphene sheet.
8.2 Introduction The use of nanofiller/nanoparticles has attracted the attention of scientific
community during the last decade. Due to their large surface area and high aspect
ratio small quantity of these nanofiller may give remarkable changes in desired
120
properties (such as mechanical performance, thermal stability, electrical conductivity of polymers) which are highly attractive for industrial applications [1-6]. Carbon
based reinforcing materials used in polymeric matrix are mostly dominated by
carbon nanotubes (CNTs). But its non-homogenous dispersion and tedious methods
of production has restricted its widespread applications. Due to these factors a new
material (graphene) was realized which may have mechanical performance like
carbon nanotubes but can easily be produced at large scale. So, graphene emerged as
promising candidate for CNTs [7-8]. Theoretical and experimental results show that
the single layered two dimensional (2D) graphene sheets are the mechanically
strongest material developed so far [9-10].
Liquid phase exfoliation (LPE) of graphite to pristine graphene [11-13] is an
attractive approach for many applications as filler in composites or hybrid materials
[14]. Graphene produced through LPE is defect free but it produces small flakes
(average size of about 1 micron) [15]. Through this method dispersions of graphene
in solvents with concentration of few mg/ml can be produced [16]. Some
applications require larger flake size for better results [17]. Gong et. al. demonstrated
that, for better mechanical properties the flake length should be of few microns or
greater [17]. Currently available graphene is significantly smaller in size that is why
it does not impart the desired mechanical properties to polymeric matrices. It is
reported that the rotation rate and flake size affects the mechanical properties
particularly young’s modulus [18]. The reinforcement values obtained by the
addition of graphene are much lower than the theoretical values [19] for example
dy/dVf ~1TPa, where Y is the composite modulus and Vf is the graphene volume
fraction [20−29].
PAN is promising polymer having extensive applications in various fields that’s why
PAN was selected in present study. In this work we used two types of graphene
nanosheets having different flake sizes (i.e. 1 micron and 3.5 micron) to study the
effects of flake sizes on the mechanical performance of PAN in terms of young
modulus, tensile strength and elongation at break. These nano-flakes were used in
different weight percentage (wt. %) from 0.25 wt. % to 12 wt. % and pure PAN was
used as reference. Dimethyl formamide (DMF) was used as solvent for solution
blending in order to have molecular level interaction of nanofiller and matrix.
121
8.3 Experimental Part
PAN polymer, N-methyl pyrrolidinone (NMP) and Dimethyl formamide (DMF),
Graphite flakes were purchased from sigma-Aldrich. Graphite was exfoliated to
graphene using tip sonicator (GEX600, 48W, 24kHz, flat head probe) in the presence
of NMP as solvent and highly concentrated dispersion of graphene was obtained
[30]. Then these exfoliated graphene sheets were separated according to their lateral
sizes using a centrifuge machine (Hettich Mikro 22R) with 5500 and 500 rpm for 45
minutes in both cases respectively. Two different size of graphene nanosheets were
separated using this procedure.
8.3.1 Composites Preparation and Characterization
PAN based composites thin films were prepared by solution casting method. PAN
was dissolved in DMF and the dispersions of two type of graphene nanosheets based
on their flake sizes were also prepared in DMF. Same solvent was used in order to
have uniform and homogenous mixing of polymer and nanofiller. Different wt. % of
nano-composites were prepared using pure PAN as reference materials. PAN
polymer and graphene nano-sheets as nano-fillers were mixed and sonicated for 30
minutes. Then drop cast into Teflon trays (of 4 x 4 x 2 cm dimensions). These trays
were placed into vacuum oven at 900 mbar for the removal of solvent (i.e. DMF) at
80 oC for eight hours. Then these trays were placed at room temperature to get
constant weight of thin films for 24 hours. The film thickness was in the range of
50—60 micron with lateral dimensions of 2.5 x 20 mm. Raman analysis was
performed on 633 nm, Horiba Jobin Yvon Lab RAM-HR. The peak intensity ratio
was calculated using following equation.
IR = ID / IG .(8.1)
Where IR is peak intensity ratio, ID is peak defect and IG is graphitic peak. TEM
analyses were conducted using a Joel 2100. Mechanical properties were measured
using Zwick-Roell tensile tester using 100N load cell at strain rate of 15 mm/min.
122
8.4 Results and discussion Traditional composites structure usually contains high volume (60%) filler in
polymeric materials. While in nanocomposites a significant change in properties can
be obtained by the addition of small quantity of nanofiller. These well-dispersed
nano-fillers at very low loading create the vast interfacial area which can affect the
behavior of surrounding polymer matrix that change its mechanical, thermal and
electrical properties [31,32].
The various flakes were separated at various speeds from 5500 rpm to 500 rpm
separating the supernatant for each cycle, like 5500, 4500, 3000, 2000, 1000 and 500
rpm and 45 minutes were given to each centrifugation cycle [15]. These dispersions
were vacuum filtered using porous alumina membrane (whatman Anodisc 47 mm
with pore size of 0.02 micron) to make thin films for Raman study. These separated
samples were studied using Raman spectroscopy and their peak intensity ratio was
studied. It was observed that at high rpm the flakes separated have high value of IR.
While those separated at low rpm has low value of IR. This indicates that the flakes
separated at high speed has large number of defects (edges) and are free from basal
defects. It was observed that as the rpm of centrifuge is lowered, the IR value
decreases which indicates that the flake size changes with the speed of centrifuge. That’s why we selected two types of flakes (separated at 5500 rpm and 500 rpm) in
this study. We are interested to know the effect of different flake sizes on the
mechanical performance of composites materials in terms of Young’s modulus, Ultimate Tensile Strength (UTS) and elongation at break (dL at break). It is clear from Figure 8.1 – 8.4, that at high centrifuge speed the graphene flakes
have high value of IR (equation. 9.1). This indicates that the graphene flakes
separated at high rates have small flake sizes, while flakes separated at low rpm
(such as 500) has low IR value which indicates that the flakes have large size
compared to previous one (i.e. separated at 5500 rpm). Raman spectroscopy also
reveals that the flakes are free from basal defects and this high value of IR is due to
123
the large number of edges which may be due to the small flake size of graphene nanosheets and vice versa [13,33,34]. Fig. 8.1 Ratio of Raman d-g bands measured on films prepared from size selected dispersion
as a function of final centrifugation rates.
1000 1500 2000 2500 3000
0.99 5000-45
0.66
0.33
0.00
0.99 4000-45
0.66
0.33
Avera
ge
0.00
0.99 3000-45
0.66
0.33
Norm
aliz
ed
0.00
0.96 2000-45
0.64
0.32
0.00
0.9 1000-45
0.6
0.3
0.0
0.96 500-45
0.64
0.32
0.00
1000 1500 2000 2500 3000
wavenumber (cm(-1)) Fig. 8.2 Raman spectra of graphene thin film, of selected size flakes prepared after different
centrifugation rate (rpm).
TEM analysis was performed on the selected samples of two types as shown in Figure 8.3 with different flake size separated at 5500 and 500 rpm.
124
Fig. 8.3 (A) TEM images of graphene flakes separated by centrifugation at 500 rpm.
Fig. 8.3 (B) TEM images of graphene flakes separated by centrifugation at 5500 rpm.
About 80 flakes were examined using TEM and similar observation was found in TEM
histograms (Figure 8.4 A & B). It is also clear that flakes separated at high rpm having small
size which is about 1 micron while those separated at low speed has flakes size about 3.5 micron.
125
Fig. 8.4 (A) Histograms of flakes length of graphene in DMF separated at 500 rpm.
Fig. 8.4 (B) Histograms of flakes length of graphene in DMF separated at 5500 rpm. Earlier it has been shown by some groups that the flake/particle size plays vital role in
improving mechanical properties of composites [17, 35]. In this study an interesting
phenomena was observed, that the selected big flake of graphene nanosheets (about
3.5 micron) has promising effects on the mechanical properties of PAN composites in
terms of Young modulus, and ultimate tensile strength (UTS).
126
Different wt. % samples of nanocomposites were prepared starting from 0.25% to
12%. It is shown in Figure 8.5 that the 50% improvement in mechanical properties in
terms of young modulus has been observed for large flake size, it means that the large
flake size has made the composite more stiff as compared to small flake size because
in case of small flake size there is improvement in modulus but not to the level of big
flake size.
Fig. 8.5 Effect of 1 micron () and 3.5 micron (♦) nano-fillers (graphene) incorporated in
PAN polymer on modulus.
It is very much clear from Figure 8.5 that the modulus was increased from 0.558 GPa
for pure PAN to 0.837 GPa for 12 wt % loading (50% increase). In case of small
flake size (1 micron) the modulus improved from 0.558 GPa (pure PAN) to 0.789
GPa (41% increase) and this is also a great improvement. However it was observed
that, at low loading from 0.25% to 1.5% the improvement in mechanical properties in
terms of modulus is more prominent for small flakes size graphene nanosheets. This
may be due to the well dispersion of nano fillers at low loading.
While at high loading, it may agglomerate and improvement is not prominent as
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0 2 4 6 8 10 12 14
E [
GP
a]
Weight %Graphene weight %
127
observed in large flake size of graphene. The presence of graphene flakes in PAN
matrix may offer resistance to the segmental movement of polymeric chains which
results in enhancement of modulus [36].
Fig. 8.6 Effect of 1micron () and 3.5 micron (♦) nano-fillers (graphene) incorporated in
PAN polymer on UTS.
Likewise it is apparent from Figure 8.6 that UTS also follow the same trend as
observed in case of modulus and also remarkable strength has been observed in
composites for both types of nanofillers of different flakes sizes. But in case of large
flake size this contributions is a little bit more prominent. The UTS improved from
37.8 MPa to 48.1 MPa for 12 wt.% loading (Figure 8.6) which is about 27%
improvement for big flake sizes. While UTS improved from 37.8 MPa (pure PAN) to
46.6 MPa for 12 wt.% loading for small flakes which is more than 23% increase.
Similar results were observed in modulus study. UTS improvement is more
prominent for low level loading of small size graphene nanoflakes; it may be due to
well dispersion of these flakes at low loading.
UTS
(M
Pa)
Graphene weight %
128
This significant enhancement in case of big flake size may be due to the proper
dispersion and adhesion of polymeric matrix. While slight inferior response of small
flake may be due to the poor dispersion of small size flakes which may agglomerate
during incorporation to polymer matrix. Similarly in Figure 8.7 an opposite response
has been noticed. The enhancement in mechanical properties, in terms of modulus and UTS of nanocomposites for both type of nanofillers was observed. While
decrease in dL at break was observed for both type of flakes but this trend is very
prominent in big flake size. However, in small flake size this situation comparatively
less than big flake size. At 12 wt% loading dL at break dropped from 18.8% to 5.3%
for big flake size, while, dL at break was affected greatly, and dropped from 18.8% to
8.6% for small flake size. This decrease in dL at break may be attributed to the
interaction of graphene with polymeric material that restricts the movement of
polymeric chains (i.e. elasticity).
Fig. 8.7 Effect of 1 micron (♦) and 3.5 micron () nano-fillers (graphene) incorporated in
PAN polymer on Elongation at break.
The poor mechanical property in terms of young’s modulus and UTS may be due to
the poor dispersion of small flakes that may agglomerate during incorporation to
polymer in the formation of thin films.
Graphene weight %
Elo
nga
tio
n a
t b
reak
(%
)
129
Conclusion PAN polymeric thin films were prepared with two different types of graphene
nanosheets as reinforcing agent by drop casting method using DMF as solvent. The
mechanical properties of both type of graphene nanosheets contributed well in
enhancement of mechanical properties to great extent. The large size flake
contributions are more prominent in enhancement of young modulus and UTS. While
in terms of dL at break the small flake size role is better than big flake size. In big
flake size Young’s modulus and UTS improved more than 45% and 25%
respectively. While in small flakes these enhancements in Young’s modulus and UTS are about 40% and 21% respectively.
References
[1] D.L. Burris, B. Boesl, G.R. Bourne and W.G. Sawyer, Macromol. Matter. Eng. 292
(2007) 387.
[2] J.Y. Kim, D.K. Kim, S.H. Kim, Polym. Compos. 30 (2009) 1779.
[3] T. Mahrholz, J. Stangle, M. Sinapius, Compos. A. 40 (2009) 235.
[4] M. Maryniak, N. Guskos, J. Typek, D. Petridis, A. Szymezyk, Polimery 54 (2009)
546.
[5] P. Mavinakuli, S.Y. Wei, Q. Wang, A.B. Karki, S. Dhage, Z. Wang, D.P. Young,
Z.H. Guo, J. Phy. Chem. C, 114 (2010) 3874.
[6] W.H. Ruan, Y.L. Mai, X.H. Wang, M.Z. Rong, M.Q. Zhang, Compos. Sci.
Technol. 67 (2007) 2747.
[7] Lee C, Wei XD, Kysar JW, Hone J. Science 321 (2008) 5887.
[8] M.D. Stoller, S.J. Park, Y.W. Zhu, J.H. An, R.S. Ruoff, Nano Letters 8 (2008)
3498. [9] M. J. McAllister, J. L. Li, D. H. Adamson, H. C. Schniepp, A. A. Abdala, et al,
Chem. Mater. 19 (2007) 4396. [10] C.G. Lee, X.D. Wei, J.W. Kysar, J. Hone. science 321 (2008) 385.
130
[11] Y. Hernandez, M. Lotya, D. Rickard, S.D. Bergin, J.N. Coleman, Langmuir, 26
(2010) 3208.
[12] Y. Hernandez, V. Nicolosi, M. Lotya, F.M. Blighe, Z.Y. Sun, et al, Nat.
Nanotechnol. 3 (2008) 563.
[13] U. Khan, A. O’Neill, M. Lotya, S. De, J.N. Coleman, Small 6 (2010) 864.
[14] S. Stankovich, D.A. Dikin, G. Dommett, et al, Nature, 442 (2006) 282.
[15] U. Khan, A. O’ Neill, H.T. Porwal, P. May, K. Nawaz, J. N.Coleman, Carbon, 50
(2011) 470.
[16] S. Park, R.S. Ruoff, Nat. Nanotechnol. b (2009) 217.
[17] L. Gong, I.A. Kinloch, R.J. Young, I. Riaz, R. Jalil, K.S. Novoselov, Adv. Mater. 22
(2010) 2694.
[18] U. Khan, P. May, A. O’Neill, J.N. Coleman, Carbon 48 (2010) 4035.
[19] G.E. Padawer, N. Beecher, Polym. Eng. Sci. b (1970) 185.
[20] H.W. Hu, G.H. Chen, Polym. Compos. 31 (2010) 1770.
[21] L. Jiang, X.P. Shen, J.L. Wu, K.C. Shen, J. Appl. Polym. Sci. 118 (2010)
275. [22] I.H. Kim, Y.G. Jeong. J. Polym. Sci. Pol. Phys. 48 (2010) 850.
[23] J. J. Liang, Y. Huang, L. Zhang, Y. Wang, Y. Ma, et al, Adv. Funct.Mater. 19
(2009) 2297.
[24] S.G. Miller, J.L. Bauer, M.J. Maryanski, P.J. Heimann, et al, Compos. Sci. Technol.
70 (2010) 1120.
[25] K.W. Putz, O.C. Compton, M.J. Palmeri, S.T.Nguyen, L.C.Brinson, Adv.
Funct. Mater. 20 (2010) 3322.
[26] T. Ramanathan, A.A. Abdala, S. Stankovich, D.A. et al, Nanotechnol 3 (2008)
327.
131
[27] P. Steurer, R. Wissert, R. Thomann, R. Mulhaupt, Macromol Rapid Commun.
30 (2009) 316.
[28] X.M. Yang, L.A. Li, S.M. Shang, X.M. Tao, Polymer 51 (2010) 3431.
[29] X. Zhao, Q.H. Zhang, D.J. Chen, P. Lu, Macromolecules 43 (2010) 2357.
[30] U. Khan, H. Porwal, A. O’Neill, K. Nawaz, P. May, J.N. Coleman, Langmuir
27 (2011) 9077. [31] T. Ramanthan, H. Liu, L.C. Brinson, J. Polym. Phys. 43 (2005) 2269.
[32] A. Bansal, H. Yang, C. Li, K. Cho, B. C. Benicewicz, et al, Nature Materials 4
(2005) 693. [33] M. Lotya, P.J. King, U. Khan, S. De, J.N. Coleman, ACS Nano 4 (2010) 3155.
[34] A. O’Neill, U. Khan, P.N. Nirmalraj, J.J. Boland, J.N. Coleman, J. Physical
Chemistry C 115 (2011) 5422.
[35] M. Conradi, M. Zorko, I. Jerman, B. Orel, I. Verpoest, Polymer Engineering &
Science 53 (2012) 1448. [36] S. Vadukumpully, J. Paul, N. Mahanta, S.Valiyaveettil, Carbon 49 (2011)
198.
132
Chapter 9
The Effect of Large Area Graphene
Oxide (LAGO) nanosheets on the
Mechanical Properties of Polyvinyl
Alcohol
9.1 Objective
Large area graphene oxide (LAGO) sheets were synthesized, dispersed in water and
used as nanofiller for mechanical improvement in terms of Young’s Modulus and
Ultimate Tensile Strength (UTS) of polyvinyl alcohol (PVA) at low loading. The
molecular level dispersion and interfacial interactions between the graphene oxides
(GO) and polymeric matrix PVA was the real challenges. An excellent improvement
in mechanical properties at 0.35wt% loading was observed. Modulus improved from
1.58 GPa to 2.72 GPa (~71 % improvement), UTS improved from 120 MPa to 197
MPa (~65% improvement) and in spite of these improvements, interestingly, there
was no fall in elongation at break at this loading.
9.2. Introduction
Planar “2D form of carbon, known as graphene has become one of the most exciting
materials today because of its unique properties [1]. Individual graphene sheets show
high values of thermal conductivity” [1], Young’s modulus [2], large surface area [1],
133
“ballistic transport on submicron scales and mass less Dirac fermions charge carrier
abilities [3]. These properties make graphene a promising material in many
applications such as photovoltaic devices”, “sensors, transparent electrodes, super
capacitors conducting composites” [4, 7–12]. “At present, carbon-based reinforcing
materials employed in polymer composites are dominated by carbon nanotubes
(CNT), but difficulty in dispersing CNT and high cost of production, limits its
widespread use”. Challenge of today is to find any alternative to “CNT and for this
purpose graphene proved itself to be a suitable candidate because of its outstanding
mechanical properties and ultra large interfacial surface area” [3, 4]. However, a
detailed comparative study of pristine graphene and multi walled carbon nanotubes
(MWCNTs) vs. silane functionalized graphene and MWCNTs in Poly (L-lactic acid)
polymer has been reported and they observed almost same improvement in
mechanical properties of polymer in terms of tensile strength, elongation at break and
young’s modulus for pristine graphene and MWCNTs [13].
In fact a lot of work has been done [14-16] showing that graphene and graphene
oxide [11, 17] are potentially effective reinforcement materials [18-26].
“Incorporation of graphitic nano flakes into elastomeric polymer matrix generates
high performance composites with improved mechanical and functional properties”
[12, 19, 27]. It has been shown that the fracture of graphene-polymer composites is
due to failure of the polymer graphene interface [26]. This failure may be due to the
short flake length of graphene than required critical length [28, 29]. It means the flake
length should be greater than critical length, for better interfacial strength, resulting
better load transfer to reinforcement [26]. Recently, “graphene oxide (GO) has also
been used as a filler in various polymeric matrix, due to its hydrophilicity and ease of
formation of stable colloidal suspensions” [19,12]. Functionalized graphene sheets
134
(FGS) are also employed because they provide much better interactions with the “host
polymers compared to unmodified CNT or traditional expanded graphite (EG)” [12].
Likewise in our previous study of incorporating functionalized graphene in
polyurethane, improvement in mechanical properties in terms of modulus and
ultimate tensile strength were observed [30]. Similarly, we observed interesting
phenomena during incorporation of selected size (big and small flakes) of graphene
nanosheets improvement in mechanical properties in terms of modulus and ultimate
tensile strength were observed for big flakes as compared to smaller ones [31].
Today’s challenge is to attain high mechanical strength with the use of minimum
amount of fillers. In this study, we selected LAGO for improving mechanical
properties of selected polymer at low loading. In our current investigation, “PVA is
chosen as the host polymer matrix, because of its wide range of applications, low
cost, chemical stability, and bio- compatibility” [32].
Oxygen containing functional groups affects the van der Waals forces between the
layers of graphene oxide and imparts desired property of water solubility. These
functional groups have been found to be effective means for the improvement of
dispersion of graphene [18,34-36]. Likewise, additional functional groups improve
its solubility / dispersibility in specific solvents [18, 19, 30]. PVA functionalized
graphene oxide sheets have been used as reinforcement in PVA matrix and 60%
improvement in terms of modulus has been observed [36]. Graphene oxide can be
dispersed at the individual sheet level in water so molecular level dispersion of GO
can be made in water. In present study improvement in mechanical properties
specially in terms of Young’s Modulus, ultimate tensile strength (UTS) has been
studied at very low loading i.e.0.35 wt%, while elongation at break was almost
undisturbed at this loading. The observed improvement in mechanical properties may
135
be due to proper dispersion, strong H-bonding of GO and PVA and the interfacial
interaction between the filler and wrapped polymer matrix [33,37-39]. Similarly,
excellent agreement between experimental nanocomposites modulus and theoretical
modulus based on Halpin-Tsai equation has been observed [37,40-42].
9.3. Experimental
A reported procedure for the synthesis of LAGO was followed [43]. In brief, a 2gm
portion of natural graphite (Aldrich), 2gm of NaNO3 (Aldrich) and 96 ml of
concentrated sulfuric acid (Sigma –Aldrich) were mixed at 0 oC. The mixture
obtained was first stirred at 0 oC for 90 minutes and then at 35 oC for 2 hrs. Millipore
water (80ml) was slowly added into the resulting solution in about half hour to dilute
the mixture. Then 200ml of water was added followed by 10ml of hydrogen peroxide
(Aldrich) (H2O2, 30%) and the stirring continued for 10 mins to obtain a graphite
oxide suspension. During this final step, H2O2 reduced the residual permanganate and
manganese dioxide to colorless soluble manganese sulfate. The graphite oxide deposit
was collected from the graphite oxide suspension by high speed centrifugation at
16000 rpm for 10 min, and repeatedly washed with distilled until its pH=7. Then a
mild bath sonication was used to exfoliate the graphite oxide to obtain a graphene
oxide (GO) suspension. Later on a low speed centrifugation at 3000 rpm (3-5 min)
was used to remove thick layer of graphite oxide from exfoliated large area graphene
oxide sheets. The supernatant was further centrifuged at 5000 rpm for 5 min to
separate large flakes (precipitate) from small one (supernatant). Finally the
precipitate was re-dispersed in water to get LAGO sheets suspension and filtered
through a nylon membrane of pore size 0.45 micron (Sterlitech). The membrane
supported graphene nano flakes were dried at 60 oC for accurate mass. Then its
dispersion was made in Millipore water and used in thin films casting. Similarly 60
136
mg/ml solution of PVA was made in Millipore water by dissolving 6 gm of PVA, in
100 ml of water for about 24 hours to get PVA solution.
The composites samples were prepared using different weight percent of LAGO to
PVA (average Mw 89,000-98,000, Aldrich) solution ranging between 0.15 wt% to 3.0
wt. % and a reference sample of PVA was also prepared. These samples were
sonicated for one hour in bath sonicatior for homogenous mixing and were drop cast
in Teflon trays of dimension 4x4x2 cm. These cast samples were dried in vacuum
oven at 900 mbar. Then these samples were shifted to another oven and kept there for
eight hours to remove traces of water, if any present. The weight of each sample was
~ 150 mg. The film thickness of each film was in the range of ~ 80-90 microns.
9.4. Characterization
FT-IR was conducted on crushed powder on a glass slide in transmittance mode using
Nexus Nicolet FTIR. Transmission electron microscopy (TEM) was performed by
dropping small quantity of LAGO containing dispersions on holey carbon grids using
a Jeol 2100, operated at 200 kV. “Scanning Electron Microscopy (SEM) was
performed using a Carl Zeiss Ultra Plus Field Emission Scanning Electron
Microscope”.
Perkin-Elmer DSC 7 was used for Differential Scanning Calorimetry (DSC) under
inert atmosphere with 10 Co/min.
9.4.1. Mechanical characterization
Zwick Roell tensile tester was use for mechanical testing with 100N load cell at a
strain rate of 15mm/min.
9.5. Results and discussion.
137
Graphite oxide (supernatant) was separated from un-oxidized graphite (precipitant)
which is not dispersible in aqueous media by adding water and centrifuged at 500 rpm
for 45 minutes to separate these both entities from each other. Then the suspension
was filtered and dried and once again its aqueous suspension was made from dried
powder and separated its large flakes from smaller one at different centrifuge speed.
The synthesis of GO was confirmed through FTIR spectroscopy as shown in
Figure.9.1.
The GO exhibits the specific peaks at 1720,1630,1390,1220, and 1050 cm-1 which
can be linked with the presence of C=O, C=C, C-O (carboxy), COH, and C-O (alkoxy
group), and broad band absorption between 3000 and 3500 cm-1 associated with
hydroxyl group [30,44-47].
Figure. 9.1 FTIR spectra of Graphene oxide
Tran
smit
tan
ce
Wave number cm-1
138
Figure.9.2 TEM images of LAGO flakes deposited from dispersion
Figure.9.3 Data for LAGO in aqueous media (A) Number of layers per flake of
LAGO (B) Length of LAGO nanosheets ~5.0 micron (C). Width of LAGO
nanosheets ~0.8 micron.
Figure 9.2 and 9.3 shows the TEM image and histogram of dimensions of the flakes
of LAGO. These dimensions in terms of number of layers per flake, length and width
have been measured [48-50]. It is clear from histogram (Figure.3) that the LAGO
flakes consist of about 3 layers in average with width of 0.8 µm and length of ~5.0
139
µm. These flakes were incorporated in PVA at different wt% for study of mechanical
properties in terms of Modulus, UTS and elongation at break.
Fig.9.4 SEM images of LAGO in PVA. (A) 3% (B) 0.35%
Figure.9.4 represents SEM images of dispersion of LAGO in polymer for different %
weight showing homogenous distribution of LAGO at low concentration. Similarly,
dispersion of LAGO in polymer at various loading can be seen in Fig.4 showing
homogenous distribution of LAGO at low concentration. Similarly, mechanical
properties in terms of modulus,UTS , elongation at break and comparative study of
predicted and observed modulus are presented in Figures 9.5- 9.8 respectively.
A B
140
Figure.9.5 Effect of LAGO on modulus of PVA
It was observed that the mechanical properties of PVA (at 0.35 wt%), in terms of
modulus and UTS improved about 71% and 65% as shown in Figures 9.5 and 9.6
respectively. The Young’s Modulus is 1.58 GPa and the Ultimate Tensile Strength
(UTS) is 120 MPa for neat sample of PVA. At 0.35 wt% loading Modulus improved
from 1.58 to 2.72 GPa (71% improvement) as shown Figure.9.5
% weight of LAGO
141
Figure. 9.6 Effect of LAGO on Tensile strength of PVA
while UTS increased from 120 MPa to 197 MPa (~ 65% increase) as shown in
Figure 9.6. This improvement in modulus and UTS may be due to good
dispersion/interaction of polymer and graphene nano sheets at this loading [48].
Beyond this loading (0.35 wt %) improvement in modulus was not linear and the
value of UTS dropped drastically. This non linear behavior in modulus and decrease
in UTS may be due to agglomeration of nanofiller / not proper dispersion in
polymeric matrix [51, 52].
142
Figure. 9.7 Effect of LAGO on elongation at break of PVA
Likewise increase in mechanical resistance properties (Modulus and UTS) results in
decrease in ductility in terms of elongation at break [52]. But in our case at 0.35 wt%
loading, we get not only improvement in mechanical properties in terms of UTS and
modulus but elongation at break is not affected but improved slightly, e.g. from
32.1% to 32.6%, but at 3.0 wt% loading the elongation at break affected disastrously,
dropping from 32.1% to 8.5% as shown in Figure.9.7. The fall in elongation at break
may be due to interaction of graphene with polymer chain which restricts the
movement of polymeric chains [52].
From these results we understand that 0.35 wt% loading is critical loading on which
very good mechanical properties in terms of Modulus, UTS and elongation at break
can be obtained, while above this loading there is downtrend in mechanical properties
143
specially in UTS and elongation at break. The non-linear increases in mechanical
properties of elastomeric composites are generally due to either the effects of strong
polymer-filler interaction or the formation of filler network [53]. In the present study
for all loadings, increase in modulus was observed, so it is believed that there exist
the phenomena of matrix-filler interaction [53]
The Halpin-Tsai model [41-43, 55] for random distribution of filler in polymeric
matrix was used to simulate our obtained results in terms of modulus based on
LAGO/PVA nano composites. Its equation for randomly distribution is given as
follows,
In this equation No. (9.1) Ec shows the modulus of the nano composites with
randomly distributed LAGO nano flakes. Similarly Vc is the volumetric fraction of
LAGO in polymer. Likewise Eg and Em are the modulus of graphene oxide nano
flakes and polymer (PVA) in eq.9.2 and 9.3 which are 0.25 TPa (Tera Pascal) [55]
(9.1)
(9.2)
(9.4)
(9.3)
144
and 1.88 GPa respectively. ξ, 1g, ,tg are the aspect ratio, length and thickness of
graphene nano flakes respectively as presented in eq.9.4. The density of PVA and
graphene oxide is 1.3 g/cm-3 and 2.200 g/cm3respectively [19]. The statistical average
of the length and thickness of LAGO nano flakes were about 5.0µm and 0.8nm
respectively as determined by TEM histogram in Figure.9.3. Now we look in detail
the dependence of mechanical properties on the content of graphene oxide. Putting all
these values in Eq. (9.1) it becomes apparent from Figure.9.8, that the
experimental results are better than theoretical one for low loading, i.e. below 0.7
wt%. This may be explained on the basis of proper dispersion of LAGO nano flakes
in polymer at low concentration, effective load transfer due to H-bonding between the
oxygen containing groups of LAGO sheets and PVA chains, and high aspect ratio of
LAGO [54].
Figure. 9.8 Comparison of Theoretical and experimental data of Young’s modulus
0
0.5
1
1.5
2
2.5
3
3.5
4
0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 2.4 2.7 3
LAGO Content (wt%)
Mo
du
lus
(G
Pa
)
Modulus(experimental)
Modulus(Theoratical)
145
These results can be compared with already published work [19] on molecular level
dispersion of graphene into PVA in which for 0.7% loading 62% improvement in
modulus and 76% increase in UTS was reported, while elongation at break was
drastically decreased as compared to neat PVA. However in our study for 0.35 wt%
loading the improvement in terms of modulus is ~ 71% and enhancement in the value
of UTS is about ~ 65% while elongation at break was not affected but improved
slightly. The elongation at break appeared to increase slightly from ~32.1% for the
polymer to ~32.6% for the 0.35wt% composite sample before subsequently falling
off. This is slightly unusual as ductile polymers usually display a decrease on strain at
break on the addition of nanofiller such as nanotubes or graphene [56-59]. Indeed
previous work on PVA filled with functionalized and pristine graphene showed a
reduction in ductility for all filler contents [19, 36]. However, in our previous study of
polyvinyl acetate as polymeric matrix the mechanical properties in terms of modulus
and UTS were improved and elongation at break was not affected [60].At low level of
loading, nano fillers are homogenously dispersed while at high loading it may form
agglomerates. So, we can say that solution blending is efficient for low loading but it
may not work for high loading.
Figure.9.9 DSC of PVA and LAGO based nano composites (0.35 wt%)
146
PVA is semi crystalline polymer and its mechanical properties depends on its degree
of crystallinity.DSC was conducted in order to see whether these improvement in
mechanical performance is due to change in crystallinity. Enthalpy of pure PVA and
sample containing 0.35 wt% graphene nanofiller was conducted. Both melt curves
has been shown in Figure.9.The melting peaks has similar pattern and both are in the
range of 160-220C0.This indicates that both samples have same crystallinity.
Crystallinity () can be determined as
Xc=ΔHm/ΔH0 (9.5)
In equation (9.5) Hm and H0 are enthalpy measured by DSC and enthalpy of pure
PVA crystallization respectively, which is 138.6 J/gm [59,60]. As there is no distinct
difference between these two melt curves as shown in Fig.9.9 so we can say that the
improvement in mechanical performance cannot be linked with change in
crystallinity.
Conclusion
LAGO flakes as nanofiller have been used in polyvinyl alcohol as polymeric matrix.
An interesting phenomenon was observed at 0.35 wt. % loading both mechanical
properties UTS and Young Modulus were improved to a good extent while elongation
at break was slightly improved. Improvement in terms of modulus is very prominent.
But beyond this loading mechanical properties in term of elongation at break and
UTS affected disastrously. The effect on elongation at break may be due to the
restacking of graphene flakes in polymer which restricts the movement of polymer
chain. Simple method of drop casting was followed using water as processing solvent
from environment stand point. The obtained results were better than as predicted by
Halpin –Tsai equation. It may be due to strong interaction of hydrogen bonding of
147
GO and PVA polymer and homogenous dispersion of GO at low loading i.e. less than
1%.
References
[1] K.S. Novoselov, A.K. Geim, S.V. Morozov, D. Jiang, Y. Zhang, S.V.
Dubonos, I.V. Grigorieva, A.A. Firsov, Science, 306, 666 (2004).
[2] AA Balandin, S Ghosh, W Bao, I Calizo, D Teweldebrhan, F Miao, CN Lau.
Nano Lett., 8, 902 (2008).
[3] C. Lee, X.D. Wei, J.W. Kysar and J. Hone, Science, 321, 385 (2008).
[4] M.D. Stoller, S.J. Park, Y.W. Zhu, J.H. An, R.S. Ruoff, Nano Lett., 8, 3498
(2008).
[5] Y. Zhang, Y.W. Tan, H. L. Stormer and P. Kim, Nature, 438, 201 (2005).
[6] K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, M. I. Katsnelson, I.
V. Grigorieva, S. V. Dubonos & A. A. Firsov, Nature, 438, 197 (2005).
[7] Z.F. Liu, Q. Liu, Y. Huang, Y.F. Ma, S.G. Yin, X.Y. Zhang, J. Adv. Mater.,
20 , 3924 (2008).
[8] J.T.Robinson, F.K. Perkins, E.S. Snow, Z.Q. Wei, P.E. Sheehan, Nano
Letters, 8, 3137 (2008).
[9] P.K. Ang, W. Chen, A.T.S. Wee, K.P. Loh, J. Am. Chem. Soc., 130, 14392
(2008).
[10] J.B. Wu, H.A. Becerril, Z.N. Bao, Z.F. Liu, Y.S. Chen, Appl. Phys. Lett., 92,
263302 (2008).
148
[11] S. Stankovich, D.A. Dikin, G.H.B. Dommett, K.M. Kohlhaas, E.J. Zimney,
E.A. Stach, R.D. Piner, S.T. Nguyen, R.S. Ruoff, Nature, 442, 282 (2006).
[12] T. Ramanathan, A.A. Abdala, S. Stankovich, D.A. Dikin, M. Herrera-Alonso,
R.D. Piner, D.H. Adamson, H.C.Schniepp, X.Chen, R.S. Ruoff, S.T. Nguyen ,
I.A. Aksay, R.K. Prud'Homme , L.C. Brinson, Nat. Nanotechnol., 3 , 327
(2008).
[13] L. Wenxiao, S. Chengbo, S. Mingling, G. Qiwei, X. Zhiwei,W. Zhen, Y.
Caiyun, M. Wei , N. Jiarong, J. Appl. Polym. Sci., 130, 1194 (2013).
[14] Y.G. Yang, C.M. Chen, W. YF, Q.H. Yang, M.Z. Wang, New Carbon
Mater., 23, 345 (2008).
[15] R. Verdejo, M.M. Bernal, L.J. Romasanta, M.A. Lopez-Manchado, J. Mater.
Chem., 21, 3301 (2011).
[16] T. Kuilla, S. Bhadra, D. Yao, N.H. Kim, S. Bose, J.H. Lee, Prog. Polym. Sci.,
35, 1350 (2010).
[17] S. Stankovich, D. A. Dikin, R. D. Piner, K. A. Kohlhaas, A. Kleinhammes, Y.
Jia, Y.Wu, S. T. Nguyen and R. S. Ruoff, Carbon , 45, 1558 (2007).
[18] J.J. Liang, Y. Huang, L. Zhang, Y. Wang, Y.F. Ma, T.Y. Guo, Y.S. Chen,
Adv. Funct. Mater., 19, 2297 (2009).
[19] K.W. Putz, O.C. Compton, M.J Palmeri, S.T. Nguyen, L.C. Brinson, Adv.
Funct. Mater., 20, 3322 (2010).
[20] S. G. Miller, J. L. Bauer, M. J. Maryanski, P. J. Heimann, J. P. Barlow, J. M.
Gosau and R. E. Allred, Compos. Sci. Technol., 70, 1120 (2010).
149
[21] L. Jiang, X-P. Shen, J-L. Wu, K-C. Shen. J. Appl. Polym. Sci., 118, 275
(2010).
[22] I-H. Kim, Y.G. Jeong, J. Polym. Sci. Pol. Phys., 48, 850 (2010).
[23] P. Steurer, R. Wissert, R. Thomann, R. Mu¨lhaupt, Macromol. Rapid Comm.,
30, 316 (2009).
[24] X. Zhao, Q. Zhang, D. Chen, P. Lu, Macromolecules, 43, 2357 (2010).
[25] X. Yang, L. Li, S. Shang, X-M. Tao, Polymer, 51, 3431 (2010).
[26] P. May, U. Khan, A. O’Neill, J.N. Coleman, J. Mater. Chem., 22, 1278
(2012).
[27] K. Kalaitzidou, H. Fukushima, L.T Drzal, Compos. Part A–Appl. S., 38, 1675
(2007).
[28] J.N. Coleman, U. Khan, W.J. Blau, Y.K. Gun’ko, Carbon, 44, 1624 (2006).
[29] G.E. Padawer, N. Beecher, Polym. Eng. Sci., 10, 185 (1970).
[30] K. Nawaz, U. Khan, N. Ul-Haq, P. May, A. O’Neill, J. N. Coleman. Carbon,
50, 4489 (2012).
[31] K. Nawaz, M. Ayub, N. Ul- Haq, M.B.K. Niazi, A. Hussain, Fibers and
Polymers, 15, 2040 (2014).
[32] L.Q.Liu, A.H.Barber, S.Nuriel, H.D.Wagner, Adv. Funct. Mater., 15, 975
(2005).
[33] M.J.McAllister, J.L.Li, D.H.adamson, H.C.schniepp, A.A.Abdala, J.Liu,
M.Herrera Alonso, D.L.Milius, R.Car, R.K.Prud’homme, I.A.Aksay,
ChemMatters – ACS, 19, 4396 (2007).
150
[34] H.C.Schniepp, J.L.Li, .J.McAllister, H.Sai, M.Herrera-Alonso, D.H.adamson,
R.K.Prud’homme, R.Car, D.A.Saville, I.A. Aksay, J. Phy. Chem. B., 110,
8535 (2006).
[35] Y.C.Si, E.T.Samulski, Nano Lett., 8, 1679 (2008).
[36] M. Cano, U. Khan, T.Sainsbury, A. O’Neil, Z. Wang, I.T.Mgovern,
W.K.Maser, A.M. Benito, J.N. Coleman. Carbon, 52, 363 (2013).
[37] D.Qian, E.C.Dickey, R.Andrews,T.Rantell, Appl. Phys. Lett., 76, 2868
(2000).
[38] J.B.Gao, M.E.Itkis, A.P.Yu, E.Bekyarova, B.Zhao, R.C.Haddon, J. Am.
Chem. Soc., 127, 3847 (2005).
[39] S.Jeong, J.S.Moon, S.Y.Jeon, J.H.Park, P.S.Alegaonkar, J.B.Yoo, Thin Solid
Films, 515, 5136 (2007).
[40] R.R.Tiwari, K.C.khilar, U.J.Natarajan, J. Appl. Polym. Sci., 108, 1818 (2008).
[41] K. Kalaitzidou, H. Fukushima, H. Miyagawa, L.T. Drzal, Polym. Eng. Sci.,
47, 1796 (2007).
[42] D.W.Schaefer, R.S.Justice, Macromolecules, 40, 8501 (2007).
[43] J. Zhao, S. Pei, W. Ren, L. Gao and H-M. Cheng, ACS NANO, 4, 5245
(2010).
[44] Y.S. Yun, Y. Bae, D.H. Kim, J.Y. Lee, I.J. Chin, H.J. Jin, Carbon, 49, 2942
(2011).
[45] S. Park, D.A. Dikin, S.T. Nguyen, R.S.Ruoff, J. Phys. Chem. C, 113, 15801
(2009).
151
[46] J.I. Paredes, S. Villar-Rodil, A. Martinez-Alonso, Langmuir, 24, 10560
(2008).
[47] S. Stankovich, R.D. Piner, S.T. Nguyen, R.S. Ruoff, Carbon, 44, 3342 (2006).
[48] U. Khan, A. O’Neill, M, Lotya, S. De, J.N. Coleman, Small, 6, 864 (2010)
[49] U. Khan, H, Porwal, K, Nawaz, P, May, J.N. Coleman, Langmuir, 27, 9077
(2011).
[50] U. Khan, A. O'Neill, H. Porwal, P. May, K. Nawaz and J. N. Coleman,
Carbon, 50, 470 (2012).
[51] J.pascual, F.peris, T.Bronat, O.Fenollar, R.Balart, Polym. Eng. Sci., 52, 733
(2012).
[52] S. Vadukumpully, J. Paul, N. Mahanta, S. Valiyaveettil, Carbon, 49, 198
(2011).
[53] D. Stauffer and A. Aharony, Introduction to Percolation Theory. Taylor &
Francis, 2nd ed. London, 1992.
[54] J.C.Halpin, J.L.Kardos, Polym. Eng. Sci., 16, 344 (1976).
[55] C.Gomez-Navarro, M.Burghad, K.Kern, Nano Lett., 8, 2045 (2008).
[56] U. Khan, P. May, A, O’Neill, J. N. Coleman, Carbon, 48, 4035 (2010).
[57] F. M. Blighe, W. J. Blau, J. N. Coleman, Journal of Nanotechnology, 19,
415709 (2008).
[58] U. Khan, F. M. Blighe, J. N. Coleman, J. Phys. Chem. C., 114, 11395
(2010).
152
[59] U. Khan, P. May, A. O’Neill, J. J. Vilatela, A. H. Windle, J. N. Coleman,
Small, 7, 1579 (2011).
[60] U. Khan, P. May, H. Porwal, K. Nawaz, J.N. Coleman, ACS-Applied
Materials & Interfaces, 5, 1423 (2013).
B
A C B A
153
Chapter 10
Summary and Future Suggestions
10.1. Summary of main work
Nanocomposites are superior to conventional composites in the sense that in former
case small amount (less than 5%) of nano-fillers are added to polymeric matrix which
ultimately produces big impact in the mechanical performance of these selected
polymers. Superior mechanical, electrical and thermal properties and potential high
aspect ratio make graphene versatile polymer reinforcement [1-6]. Graphene can be
modified with organic functional groups via liquid-phase reaction for better
interaction with polymeric materials to have improved mechanical properties [7].
In this study various polymers were selected for the study of mechanical response in
terms of modulus, UTS and dL at break. Graphene nanosheets in various forms and
sizes were selected as nano-fillers for this study.
Graphene was functionalized with organic entity like octadecylamine (ODA) and its
functionalization was confirmed through FT-IR spectroscopy. The functionalized
graphene and poly (urethane) were readily dispersible/soluble in THF solvent which
helped in molecular level interaction of these both entities. It was observed that no
appreciable increases in either stiffness or low-strain stress for loading levels below
2.5vol%. However above this threshold, both mechanical quantities increase as a
power law. This behavior is consistent with mechanical percolation. This implies that
the graphene oxide platelets are effectively isolated at low volume fractions but begin
to form a network at a volume fraction of 2.5vol%. This loading level can be thought
of as a percolation threshold. As the loading level is increased, the network becomes
more extensive and the stiffness and low-strain stress increase as described by the
percolation scaling law, interestingly the ultimate tensile strength initially increases
but reaches a maximum at the percolation threshold. Similarly the ductility is invariant
with graphene content up to the percolation threshold, after which it falls steadily.
The Young’s modulus appeared to increase almost linearly with increasing GO-ODA
mass fraction from 9.6 MPa for the polymer to 335 MPa for the 50wt% composite.
the ultimate tensile strength initially increased from 27 MPa for the polymer to 38
154
MPa for the 3 wt% composite before falling steadily, reaching 10 MPa for the 50wt%
composite . The strain at break decreased steadily with increasing GO-ODA mass
fraction from ~1000% for the polymer to ~10% for the 50 wt% sample.
This work shows that the mechanical properties of elastomers reinforced with
graphene can depend on parameters other than interfacial stress transfer.
Similarly in case of PVC polymer a critical loading (1.5 wt %) was observed at which
the mechanical performance was enhanced and interestingly there was very slight fall
in dL at break. At 1.5wt% loading, modulus improved 63% (from 1.31 to 2.14 GPa),
UTS improved about 19% from (70 MPa to 83.2 MPa), and negligible effect on
elongation at break of PVC (from 15.3% to 13 %) was observed. On other hand, at 10
wt% loading, modulus enhanced 75 % (1.31 to 2.29 GPa) UTS dropped about 22%
(from 70 MPa to 55 MPa), dL at break was disastrously affected (15.3% to 4.4 %).
Modulus determined in this study was compared to that calculated from Halpin-Tsai
equation. During this comparative study it was observed that the response of modulus
was superior to theoretical one based on Halpin-Tsai model at low loading( at or
below 1.5 wt %).I understand that at low loading of nanofiller the solution blending
works effectively and homogenous distribution can be obtained which results in better
mechanical performance of nanocomposites.
Likewise I also incorporated two different flake sizes of graphene nanosheets
(1 µm and 3.5 µm) to PAN polymer. At 12 wt. % loading the modulus increased 50%
(from 0.558 GPa to 0.837 GPa) while the UTS improved about 27% (from 37.8 MPa
to 48.1 MPa). However, in case of small flake size (1 µm) the modulus improved 41%
(from 0.558 GPa to 0.789 GPa). While improvement in UTS is 23 % (from 37.8 MPa
to 46.6 MPa) for 12 wt. % loading for small flakes was observed.
155
dL at break dropped from 18.8% to 5.3% and 8.6% for big and small flake size at 12
wt% loading respectively.
Similarly large area graphene oxides (LAGO) were synthesized and were introduced
in PVA. Water was used as solvent for both polymer and nanofiller in order to have
molecular level interaction. The performance of LAGO as nanofiller was very
impressive in terms of modulus, UTS and dL at break. At 0.35% loading Modulus
improved from 1.88 to 2.64 GPa (71% improvement) UTS increased from 130 MPa to
195 MPa (~ 67% increase) and elongation at break is not affected but improved
slightly, e.g. from 32.1% to 32.6%. But beyond this loading, improvement in modulus
was persistent but UTS and the elongation at break affected disastrously. dL at break
dropped from 32.1% to 8.5% drastically.
I understand that 0.35% loading is critical loading on which we get very good
mechanical properties in terms of modulus, UTS and elongation at break.
While above this loading there is downtrend in mechanical properties especially in
UTS and elongation at break. In this study, for all loading increase in modulus was
observed so it is believed that there exist the phenomena of matrix-filler interaction.
Similarly in case of PVAc 50% increase in stiffness, 100% increase in “tensile
strength on addition of 0.1 vol % graphene compared to the pristine polymer”. The
adhesive “strength and toughness of the composites were up to 4 and 7 times higher,
than the pristine polymer”.
Graphite was exfoliated to graphene in organic and aqueous media, having 63 and 7
mg/mL concentrations in these both media respectively.
156
In case of water as exfoliating media sodium cholate was used as surfactant in order to
ease exfoliation. In the process of this exfoliation sonic tip was used as exfoliating
instrument. Later on the exfoliated graphene was separated via controlled
centrifugation according to its lateral dimensions.
10.2 Future suggestions
Although graphene based nanocomposites are technologically prominent
development to emerge from the interface between graphene and polymeric materials
[8,9]. But there are some certain challenges which must be resolved in order to fully
exploit the mechanical performance of graphene. For example, graphite and GO is
usually exfoliated through sonication which ultimately reduces its size in terms of
lateral dimension as result it negatively affect the mechanical properties of
composites [10-12]. Likewise, Defects and wrinkles in platelets are also one of the
reasons for poor reinforcing capabilities. Composites properties can further be
improved if alignment and spatial organization of graphene are addressed [13-15].
In spite of these problems nanocomposites has very bright future in commercial
activities. Graphene –based composites materials can become commercial reality due
to its extraordinary mechanical properties and low cast raw material like graphite if
its production methods are improved.
References
[1] K.S. Novoselov, A.K. Geim, S.V. Morozov, D. Jiang, Y. Zhang, S.V.
Dubonos, et al, Science 306 (2004) 666.
[2] A.A. Balandin, S. Ghosh, W. Bao, I. Calizo, D. Teweldebrhan, F. Miao, et
al, Nano Lett. 8 (2008) 902.
[3] C. Lee, X.D. Wei, J.W. Kysar, J. Hone, Science 321 (2008) 385.
[4] M.D. Stoller, S.J. Park, Y.W. Zhu, J.H. An, R.S. Ruoff, Nano Lett. 8 (2008)
3498.
157
[5] Y.B. Zhang, Y.W. Tan, H.L. Stormer, P. Kim, Nature 438 (2005) 201.
[6] K.S. Novoselov, A.K. Geim, S.V. Morozov, D. Jiang, M.I. Katsnelson,
I.V. Grigorieva, et al, Nature 438 (2005).
[7] Stankovich, S.; Piner, R. D.; Nguyen, S. T.; Ruoff, R. S. Carbon , 44
(2006) 3342.
[8] Dreyer DR, Park S, Bielawski CW, Ruoff RS. Chem Soc Rev 39
(2010) 228.
[9] Schniepp HC, Li JL, McAllister MJ, Sai H, Herrera-Alonso M,
Adamson DH, et al. J Phys Chem B 110 (2006) 8535.
[10] U. Khan, A. O’ Neill, H.T. Porwal, P. May, K. Nawaz, J. N.Coleman,
Carbon, 50 (2011) 470
[11] Kalaitzidou, H. Fukushima, L.T. Drzal, Journal of Composites Part A –
Applied Science Manufacturing 38 (2007) 7.
[12] U. Khan, H. Porwal, A. O’Neill, K. Nawaz, P. May, J.N. Coleman,
Langmuir 27 (2011) 9077
[13] Xie XL, Mai YW, Zhou XP. Mater Sci Eng R Rep 2005;49:89-112.
[14] Hussain F, Hojjati M,Okamoto M, Gorga RE. J. Compos Mater 40
(2006) 1511.
[15] Marcano DC, Kosynkin DV, Berlin JM, Sinitskii A, Sun Z, Slesarev
A, et al. ACS Nano, 4 (2010) 4806.
158
Author’s List of Publications (as Ist author)
[1] K. Nawaz, U. Khan, N. Ul-Haq, P. May, A. O’Neill, J.N. Coleman.
“Observation of mechanical percolation in functionalized graphene oxide
/elastomers composites.” Carbon 50 (2012) 12.
[2] Khalid Nawaz, Muhammad Ayub , Noaman Ul haq ,M.B.Khan,
Muhammad Bilal khan Niazi,Arshad Hussain “Effects of selected size of
graphene on the mechanical properties of poly (acrylonitrile) (PAN).”
Fibers and Polymers. 15 (2014) 2040.
[3] Khalid Nawaz, Muhammad Ayub, Noaman Ul haq, M.B.Khan,
Muhammad Bilal Khan Niazi, Arshad Hussain,: “Effects of
graphene nanosheets on the mechanical properties of polyvinyl chloride
(PVC)” published on line in Journal of Polymer composites.
[4] Khalid Nawaz, Noaman Ul Haq,Muhammad Ayub, M.B.Khan,
Muhammad Bilal Khan Niazi,Arshad Hussain. “Effects of Large
Area Graphene Oxide (LAGO) on the mechanical properties of Poly
(vinyl alcohol) (PVA).” Published on line in Journal of Polymer
Engineering and Science.
[5] Khalid Nawaz,Muhammad Ayub,M.B.Khan,Arshad Hussain,Abdul
Qadeer Malik, Muhammad Bilal khan Niazi, Noaman Ul Haq “Effect of
surfactant concentration on the exfoliation of graphite to graphene in
159
aqueous media.” Under review in J. NanoMaterials and Nano
Technologies, Manuscript number.2015.0108R1.
[6] Khalid Nawaz, ,Muhammad Ayub, M.B.Khan,Arshad Hussain, Noaman Ul
Haq “Effect of Thermal Treatment on the Exfoliation of Graphite to Graphene
in Acetonitrile (ACN).” Submitted in the journal of “Asia-Pacific Journal of
Chemical Engineering”.
Publications of author as Co-author
[1] U. Khan, H. Porwal, A. O'Neill, K. Nawaz, P. May, J.N. Coleman, “Solvent-
Exfoliated Graphene at Extremely High Concentration.” Langmuir 27 (2011)
9077.
[2] U. Khan, A. O’ Neill, H.T. Porwal, P. May, K. Nawaz, J. N.Coleman, “Size
selection of dispersed, exfoliated graphene flakes by controlled
centrifugation.” Carbon, 50 (2012) 470.
[3] Umar Khan, Peter May, Harshit Porwal, Khalid Nawaz, and Jonathan N.
Coleman. “Improved Adhesive Strength and Toughness of Polyvinyl acetate
Glue on Addition of Small Quantities of Graphene.” ACS Applied Materials
and Interfaces 2013.
160
Papers presented in International conferences.
[1] Khalid Nawaz, Noaman Ul Haq, Muhammad Ayub, M.B.Khan. Paper
presented in 10th International Bhurban Conference of Science and
technology (IBCAST) January 2013. Titled “Effects of graphene
nanosheets on the mechanical properties of polyvinyl chloride (PVC)”
[2] Khalid Nawaz, Noaman Ul Haq,Muhammad Ayub, M.B.Khan Poster
presented in 11th International Bhurban Conference of Science and
technology (IBCAST) January 2014.titled “Effect of Large Area
graphene oxide on the mechanical properties of Polyvinyl alcohol
(PVA).
[3] Khalid Nawaz, Noaman Ul Haq, Muhammad Ayub, M.B.Khan.
“Effect of Surfactant concentration on the exfoliation of graphite to
graphene in aqueous media”. Paper presented in 12th IBCAST 2015.