University of Southern California Dissertations and Theses

Multiwall carbon nanotubes reinforced epoxy nanocomposites

(USC Thesis Other)

Multiwall carbon nanotubes reinforced epoxy nanocomposites

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

MULTIWALL CARBON NANOTUBES REINFORCED EPOXY NANOCOMPOSITES

by

Wei Chen

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(MATERIALS SCIENCE)

May 2009

Copyright 2009 Wei Chen

ii
Acknowledgements
This thesis is the account of five and half years of devoted work in the field of carbon
nano-particles reinforced polymer nanocomposites at the M. C. Gill Composite Center of the
University of Southern California, Los Angeles, which would not have been possible without the
help of many.
First of all, I wish to thank my thesis advisor, Dr. S. R. Nutt, for his unfaltering trust and
constant encouragements which have been the guiding light through my life as a graduate student
at USC. I wish to thank him for who he is as a professor, a mentor, a friend, and especially his
efforts in understanding a student’s personality and tailoring his approach accordingly.
In addition, I would like to express my deep honor by the fact that the jury in front of
which I shall defend my PhD thesis will be joined by Prof. Katherine Shing and Prof. Thieo
Hogen Esch, and my gratitude for their time to evaluating my thesis.
My experience at USC would not have been such a pleasurable one without the presence
of all the people working there. I’ve known Prof. Hongbin Lu from Fudan University and Prof.
Maria Lujan Auad from the University of Auburn since 2003. Their guidance and friendship is
important to me and I want to thank them for their invaluable discussion towards my dissertation.
I wish to thank Dr. Changzheng Huang, Dr. Eric Bosze, and Dr. Virginia Alonso for sharing with
me their research ideology and approaches turning ideas into practical solutions. My thanks will

iii
go also to the lab manager, Warren Haby, for his support and advice that greatly facilitated my
experiments, and I sincerely hope he could get well soon from his illness.
Last but not the least, I wish to thank my family for their prayers and their
encouragements throughout my graduate work in USA. I owe my parents, Minghua Zhu and
Baorong Chen much of what I have become. I thank them for their love, their support, and their
confidence throughout the past twenty-eight years. My parents have always put education as a
first priority in my life, and raised me to set high goals for myself. They taught me to value
honesty, courage, and diligence above all other virtues. I have been raised to work hard to achieve
my goals in life and they have always been there for me as an unwavering support. I dedicate this
work to them, to honor their love, patience, and support during these years.

iv
Table of Contents
Acknowledgements ii
List of Tables vii
List of Figures viii
Abstract xii
Chapter 1. Introduction 1
Chapter 1 References 8

Chapter 2. Literature Review on the Mechanical Properties of CNT
Reinforced Polymer Nanocomposites
2.1. Carbon nanotubes – the ideal reinforcement material 10
2.2. CNT polymer nanocomposites – Challenges 18
2.2.1. Purification and selection 20
2.2.2. Dispersion 22
2.2.3. Interface 27
2.2.4 Alignment and assembly 31
2.2.5 Process dependency 33
2.3. Mechanical properties of CNT polymer nanocomposites 34
2.3.1. Interface stress transfer 34
2.3.2. Stiffening effect of CNTs 41
2.3.3. Motivation for the current work 46
Chapter 2 References 51

Chapter 3. Surface assisted dispersion and interfacial stress transfer 65
3.1. Motivation 65
3.2. Experimental 68
3.2.1. Materials 68
3.2.2. Preparation of COOH-MWCNT 68
3.2.3. Preparation of PGE-MWCNT and DGEBA-MWCNT 72
3.2.4. Preparation of the MWCNT- Epoxy Composites 72
3.2.5. Characterization 73
3.3. Results and Discussion 74
3.3.1. Properties of functionalized MWCNT 74
10

v
134
3.3.2. Formation of covalent bonds at interface 76
3.3.3. Microscopic morphology of the composites 80
3.3.4. Elastic and flexural properties 80
3.4. Conclusions 82
Chapter 3 References 85

Chapter 4. Towards interface control 86
4.1. Motivation 86
4.2. Experimental 87
4.3. Results and Discussion 92
4.3.1. Surface Modification Assisted Dispersion 92
4.3.2. Polymerization and Interphase Formation 98
4.3.3. Thermomechanical Properties 100
4.3.4. Microstructural Analysis 105
4.3.5. Model Studies and Refinement 108
4.4. Conclusions 113
Chapter 4 References 115

Chapter 5. Towards hierarchical reinforcement 119
5.1. Motivation 119
5.2. Experimental 121
5.2.1. Materials 121
5.2.2. Chemical Modification of MWCNTs 122
5.2.3. Composite Preparation 123
5.2.4. Characterization 124
5.3. Results and discussion 125
5.3.1. Properties of functionalized MWCNTs. 125
5.3.2. Properties of the Nanotube Reinforced Epoxies (NEs) 128
5.3.3. Properties of the Multi-scale Reinforced Composite
Laminates (MSRs)
5.4. Conclusions 142
Chapter 5 References 144

Chapter 6. Towards multiple functionalities 147
6.1. Motivation 147
6.2. Theory 149
6.3. Formulation 150
6.4. Experiments & model 155

vi
6.5. Results & Discussion 159
6.5.1 Wave speed reduction and density adjustment 159
6.5.2 Reinforcement effect of fillers 160
6.5.3 Transmission Loss 167
6.6. Conclusions 171
Chapter 6 References 172

Chapter 7. Conclusions and Future Work 174
7.1 Summary 174
7.2 NRPs with perfectly embedded CNT array 176
Chapter 7 References 184

Bibliography 185

Appendix: Publications 208

vii
List of Tables

Table 1.1 Comparison and contrast of advanced fiber reinforcements vs.
carbon nanotubes in terms of diameter, density, tensile strength,
and modulus. Credit: Fisher/ Northwestern University.

Table 2.1 Electrical, mechanical, and absorption properties of carbon
nanotubes and the corresponding applications. [3-19]

Table 3.1 Elastic Modulus and Flexural Strength of the Neat Epoxy and
MWCNTs-Epoxy Composites.

Table 4.1 DMA data of the glass transition temperatures (Tg) of MWCNTs
reinforced epoxies vs. MWCNT loadings.

Table 5.1 Glass transition temperatures (DSC) for NEs at different CNT
loadings.

Table 5.2 Experimental Young’s modulus and tensile strength values for
MSR specimens.

Table 6.1 Formulation of Test Samples

Table 6.2 Wave speed and Static Mechanical Properties of the Samples

Table 6.3 Final Sample Properties

4
13
83
103
129
136
151
157
164

viii
List of Figures

Figure 1.1 Schematic views of natural polymer nanocomposites: spider
silk, muscle, wood, mother of pearl.[17-19]

Figure 2.1 a. SEM image of SWCNT bundles made of ~ 100 SWCNTs.
Scale bar 10 nm; Credit: R. E. Smalley/ Rice University. b.
A multi-walled carbon nanotube partly filled with an iron
carbide nanowire. Scale bar 5 nm. Credit: Banhart/Johannes
Gutenberg University.

Figure 2.2 Sword and sheath mechanism in multiwalled carbon
nanotube.[39] (A) A MWCNT having a section length of 6.9
microns under tensile load just before breaking. (B) After
breaking, one fragment of the same MWCNT was attached
on the upper AFM tip and had a length of 6.6 microns. (C)
The other fragment of the same MWCNT was attached on
the lower AFM tip and had a length of 5.9 microns.

Figure 2.3 Nanotube agglomerations and entanglement. Scale bar: 1
micron. Credit: Arkema.com

Figure 2.4 Schematic representations of the mechanism by which
surfactants help to disperse SWCNT. Credit: Yurekli et al.
University of Houston

Figure 2.5 Direct observation of polymer sheathing in carbon
nanotube-polycarbonate composites by Ding et al. [119]

Figure 2.6 Fraction of non-bulk polymer in the interphase region as a
function of volume fraction of fiber inclusion, where t is the
interphase thickness and r
f
is the radius of the nanotube/fiber
inclusion.[114]

Figure 2.7 Direct conversion of nanotube forest aerogel into aligned
CNT sheets and assemblies of those sheets.[127]

5
11
16
24
26
28
30
32

ix
Figure 2.8 TEM images of a MWCNT crossing a hole in an epoxy resin
matrix. A. MWCNT gapping a hole; B.MWCNT partial
pullout by SPM; C. force distance curve.

Figure 2.9 Illustration MD simulations of the cross-linked systems.
Left: crystalline matrix. Right: amorphous matrix. [150]

Figure 2.10 Eshelby's inclusion problem. [170]

Figure 2.11 Finite element modeling of an infinitely long wavy nanotube
embedded in a matrix. [176-177]

Figure 3.1 Main reactions taking place during the anionic
polymerization of an epoxy monomer initiated by a tertiary
amine (R’COOH and R’’OH represent species present in
the reaction medium).

Figure 3.2 TEM image of acidified MWCNTs.

Figure 3.3 Scheme of reactions used to functionalize MWCNTs.

Figure 3.4 TGA thermograms of neat and functionalized MWCNTs.

Figure 3.5 Solubility of neat and functionalized MWCNTs in: a) water
and b) DGEBA (Epon 828).

Figure 3.6 FTIR spectra of neat and functionalized MWCNTs (the peak
at 2350 cm
-1
was caused by adventitious CO
2
in the sample
chamber).

Figure 3.7 SEM micrographs of fracture surfaces obtained under
cryogenic conditions (1wt% CNT): a) neat epoxy, b)
unmodified MWCNTs-epoxy composite, c)
COOH-MWCNTs-epoxy composite, d)
DGEBA-MWCNTs-epoxy composites.

Figure 4.1 Transmission Electronic Microscope image of oxidized
MWCNTs.
35
38
45
47
69
70
71
77
78
79
81
89

x
Figure 4.2 A schematic view of the Zwitterionic reactions between
oxidized MWCNTs and two forms of A-PPO with different
molecular weight.

Figure 4.3 TGA thermal decomposition data of amine functionalized
MWCNTs compared with those of o-MWCNTs and A-PPO.

Figure 4.4 FTIR spectra for functionalized MWCNTs at 900 ~ 1750
and 2800 ~ 3000 wavenumbers.

Figure 4.5 DSC curves for MWCNT reinforced epoxies at 0.75 phr
compared with control epoxy sample.

Figure 4.6 Loss modulus of o-MWCNT and 2000-MWCNT (o-NT and
2000-NT for simplicity) reinforced epoxy nanocomposites.
200-MWCNT composites have similar trend in Tg as o-CNT
composites and therefore data are not shown.

Figure 4.7 Storage modulus (25 °C) of epoxy nanocomposites reinforced
with different MWCNTs and results converge at the datum
point for neat epoxy.

Figure 4.8 (a) – (f) SEM images of the nanocomposite fracture surfaces
(in 0.75 phr). (a). Neat epoxy control sample. (b)
o-MWCNT, arrows indicate CNT debonding and
agglomeration. (c) 200-MWCNT. (d) 2000-MWCNT, fast
fracture surface. (e) 2000-MWCNT, fibril structures. (f)
2000-MWCNT, fibril structures with zoomed in image.

Figure 4.9 Schematic illustration of the CL model with concentric
layers of fiber, matrix, and interface. Double curves for the
interphase region indicate tunable or gradient properties.

Figure 4.10 The full width half maximum (FWHM) values of the
normalized loss modulus curves (E”) vs. nanotube loading
for nanocomposites with varied nanotubes.

Figure 5.1 TEM image of o-CNTs (a) and PGE-CNTs (b).
90
94
95
99
101
102
106
109
110
126

xi
Figure 5.2 SEM images of fracture surface of NEs. All specimens have a
CNT loading of 0.5vol%.

Figure 5.3 a. Experimental tensile stress vs. CNT volume fraction
plotted with theorectical predictions. b. Experimental tensile
strength at break vs. CNT volume fraction.

Figure 5.4 SEM images of basalt fiber fracture surface of MSRs with
0.5vol% CNT inclusions.

Figure 5.5 a. E
c11
vs. CNT volume loading: measured values and
predicted curves. b. CNT redistribution around basalt fiber. c.
CNT alignment along basalt fiber orientation.

Figure 6.1 Pulse-echo method (a) and free-free beam measurement (b) for
longitudinal wave speed test.

Figure 6.2 Test results of neat MWCNT reinforced epoxy composites
with Various CNT Filler Concentration.

Figure 6.3 Flexural Modulus & Flexural Strength of epoxy elastomer
blends.

Figure 6.4 Measured Insertion Loss at Various Angles of Incidence.

Figure 6.5 Computed Transmission Loss of the Window with A)
Measured Shear Loss Factor and B) Assumed Shear Loss
Factor due to Carbon Nanotube Damping.

Figure 7.1 SEM images of CNT forest pillar wetted by SU-8. [2]

Figure 7.2 Ultralow feeding gas flow guiding growth of large-scale
horizontally aligned single-walled carbon nanotube arrays.
[14]

Figure 7.3 MWCNT growth on carbon fiber surface.

127
131
133
141
154
161
162
165
169
178
180
182

xii
Abstract
The emergence of carbon nanotubes (CNTs) has led to myriad possibilities for structural
polymer composites with superior specific modulus, strength, and toughness. While the research
activities in carbon nanotube reinforced polymer composites (NRPs) have made enormous
progress towards fabricating next-generation advanced structural materials with added thermal,
optical, and electrical advantages, questions concerning the filler dispersion, interface, and CNT
alignment in these composites remain partially addressed. In this dissertation, the key technical
challenges related to the synthesis, processing, and reinforcing mechanics governing the effective
mechanical properties of NRPs were introduced and reviewed in the first two chapters.
Subsequently, issues on the dispersion, interface control, hierarchical structure, and
multi-functionality of NRPs were addressed based on functionalized multi-walled carbon
nanotube reinforced DGEBA epoxy systems (NREs). In chapter 3, NREs with enhanced flexural
properties were discussed in the context of improved dispersion and in-situ formation of covalent
bonds at the interface. In chapter 4, NREs with controlled interface and tailored
thermomechanical properties were demonstrated through the judicious choice of surface
functionality and resin chemistry. In chapter 5, processing-condition-induced CNT organization
in hierarchical epoxy nanocomposites was analyzed. In Chapter 6, possibilities were explored for
multi-functional NREs for underwater acoustic structural applications. Finally, the findings of
this dissertation were concluded and future research was proposed for ordered carbon nanotube

xiii
array reinforced nanocomposites in the last chapter. Four journal publications resulted from
this work are listed in Appendix.

1
Chapter 1. Introduction
Mankind’s ability to shape the physical world is evolved with the progress in the use of
signature materials in history -- the Stone Age, the Bronze Age, the Iron Age, and more recently,
the Silicon Ages. With the ever more challenging demands of modern day society, materials with
synergistic properties are chosen to create composite materials with tailored properties, and
therefore, today is the Age of Composites.
By definition, a composite is a multiphase material formed from a combination of
materials which differ in composition or form, retain their own chemical and physical properties,
and maintain an interface between components which act in concert to provide improved specific
or synergistic characteristics not obtainable by any of the original components acting alone [1].
There are two categories of constituent materials in any composite: the matrix and the
reinforcement. The matrix material surrounds and supports the reinforcement materials by
maintaining their relative positions. The reinforcements impart special properties (thermal,
mechanical, optical, and electrical, etc.) to enhance the matrix properties. While the matrix
materials can either be a polymer, ceramic, or metal, composite materials can be divided into four
categories based on the geometry of reinforcement materials: (1) particulate, including particles
or flakes, (2) fiber, (3) laminar, and (4) hybrid (the combinations of any of the above) composites.
Composites have been used throughout history. Even in the earliest ages, Mankind used
composite materials unwittingly. These early experiments are evidenced by the Israelites' use of

2
chopped straw in their brick; the Egyptian sarcophagi from glued wood laminates and their
mummy embalming using cloth tape soaked in resin; the Mongol warriors' archery bows made of
bullock tendon, horn, bamboo strips, silk, and pine resin, which are almost as strong as the
modern fiberglass bows; and the Roman artisans' use of ground marble in their mortar. During the
fabrication of pottery, the ancients also used goat hair, which, after firing, was converted to
carbon, similar to the modern carbon fiber reinforced ceramics. In recent history, one of the key
functions of a materials scientist is to engineer composites to fill in the blanks in materials
selection charts, from which specific materials can be conveniently chosen catering to a wide
variety of needs of our modern society. Advanced composites have been increasingly used in
industries that demand unique engineered materials, including automobile, electronics, wind
power, and aerospace, etc. [2-5] For example, ultra high-modulus but brittle carbon fibers are
combined with low-modulus polymers to create a high-stiffness and light-weight composite with
targeted toughness for aerospace applications [6].
With the improvement of our understanding and fabrication techniques in the past
decades, however, we have reached the design limits of optimizing composites with traditional
micro-meter scale fillers/reinforcements, in which macroscopic defects due to regions of high or
low volume fraction of filler often lead to materials failure and property tradeoffs. [7-8] Recently,
the emergence of nanoscale particles reinforced polymer composites, or polymer nanocomposites,
in which at lease one phase is less than 100 nm in at least one dimension, has led to new hopes to

3
overcome the limitations of traditional micrometer-scale polymer composites. [9] Although, some
nanocomposites (such as carbon black and fumed silica filled polymers [10-11]) have been used a
long time ago, research and development of nanoparticle filled polymers has increased
exponentially since the early 1990s with the ‘discovery’ of carbon nanotubes [12], the properties
of which, particularly modulus, strength, and electrical properties, are significantly different from
those of traditional microscopic fillers and offer exciting possibilities for new generation of
composite materials. (Table 1.1[13]) The unprecedented development in the synthesis of
nanoparticles and the processing of nanocomposites[13-16] has led to our ability to control over
such composites with limitless combinations of properties.
Nature has also shown some of the most amazing forms of nanocomposites. For
example, inside a typical spider silk (as strong as steel on a weight basis), one finds assembled
crystalline particles separated by amorphous protein linkages.[17] Mother of pearl or nacre is
another natural nanocomposite formed by regularly stacked and layered calcium carbonate
platelets bonded together by proteins. [18] Resembling fiber reinforced composites, wood
consists of nanoscale cellulose fibers bonded together with lignin and other carbohydrate
constituents. [19] (Figure 1.1) In all these examples, Nature has demonstrated enormous success
over eons of evolution by utilizing bio-matrix materials to bond and support organized
nano-particles or nano-fibrils to attain superior or specialized properties.

4

Table 1.1 Comparison and contrast of advanced fiber reinforcements vs. carbon nanotubes in
terms of diameter, density, tensile strength, and modulus. Credit: Fisher/ Northwestern
University.

5

Figure 1.1 Schematic views of natural polymer nanocomposites: spider silk, muscle, wood,
mother of pearl.[17-19]

Spider Silk
Structure of Muscle
Nacre (oysters and abalone)
Wood

6
In view of recent development, the engineering of nanocomposites has become
technically feasible. The research community has made enormous advances in the processing of
nanocomposites in terms of manipulating the size, shape, volume fraction, interface, organization,
and degree of dispersion to tailor these materials.[20-22] Although the understanding of the
structure – processing – property relationship of these materials is still in its infantry, we have
already demonstrated exciting possibilities, especially when the combined theoretical and
experimental efforts have generated more information to guide further development.
In this dissertation, the key technical challenges related to the synthesis, processing,
and reinforcing mechanics governing the effective mechanical properties of NRPs are introduced
and reviewed in the first two chapters. Subsequently, issues on the dispersion, interface control,
hierarchical structure, and multi-functionality of NRPs were addressed based on functionalized
multi-walled carbon nanotube reinforced DGEBA epoxy systems (NREs). In chapter 3, NREs
with enhanced flexural properties are discussed in the context of improved dispersion and in-situ
formation of covalent bonds at the interface. In chapter 4, NREs with controlled interface and
tailored thermomechanical properties are demonstrated through the judicious choice of surface
functionality and resin chemistry. In chapter 5, processing condition induced CNT organization in
hierarchical epoxy nanocomposites is analyzed. In Chapter 6, possibilities are explored for
multi-functional NREs for underwater acoustic structural applications. Finally, the findings of

7
this dissertation are concluded and future research was proposed for ordered carbon nanotube
array reinforced nanocomposites in the last chapter.

8
Chapter 1 References
[1] Daniel, Isaac M. Engineering mechanics of composite materials, Oxford University Press,
2006

[2] Reinert, H. S.; Meade, L. E. Roy National SAMPE Technical Conference, v 7, 1975, p
475-487

[3] Malik, B. (Natl. Inst. Foundry Forge Technol.); Talukdar, P.; Mukherjee, S.K. Journal of the
Institution of Engineers (India): Aerospace Engineering Journal, v 82, n 1, May, 2001, p 38-40

[4] Premkumar, M.K. (Alcoa Tech. Center, Alcoa Center, PA, USA); Yun, D.I.; Sawtell, R.R.
Proceedings of the SPIE - The International Society for Optical Engineering, v 2105, 1993, p
399-404

[5] Jiang, Zehui (International Center for Bamboo and Rattan); Sun, Zhengjun; Ren, Haiqing
Acta Materiae Compositae Sinica, v 23, n 3, June, 2006, p 127-129

[6] Fitzer, E.; Terwiesch, B. Carbon, v 10, n 4, Aug, 1972, p 383-390

[7] Theocaris, P.; Paipetis, S.A. Journal of Composite Materials, v 9, n 3, July 1975, p 244-50

[8] Head, Andrew Society of the Plastics Industry, Reinforced Plastics/Composites Institute,
Annual Conference - Proceedings, Feb 1-5, 1988, p 9D.1-9D.3

[9] Misra, Devesh K. The Minerals, Metals and Materials Society, 2006

[10] Johnson, B.L. Industrial and Engineering Chemistry, v 43, n 1, Jan, 1951, p 146-154

[11] Berrod, G.; Vidal, A.; Papirer, E.; Donnet, J. B. Journal of Applied Polymer Science, v 23, n
9, May, 1979, p 2579-2590

[12] Iijima, S. Nature, v 354, n 6348, 7 Nov. 1991, p 56-8

[13] Bakunin, V.N.; Suslov, A.Yu.; Kuzmina, G.N.; Parenago, O.P. Journal of Nanoparticle
Research, v 6, n 2-3, June 2004, p 273-84

[14] Kruis, F. Einar; Fissan, Heinz; Peled, Aaron. Journal of Aerosol Science, v 29, n 5-6,
Jun-Jul, 1998, p 511-535

9
[15] Jordan, J.; Jacob, K.I.; Tannenbaum, R.; Sharaf, M.A.; Jasiuk, L. Materials Science &
Engineering A, v 393, n 1-2, 25 Feb. 2005, p 1-11

[16] Sternitzke, M. Journal of the European Ceramic Society, v 17, n 9, 1997, p 1061-82

[17] Foo, C.W.P.; Patwardhan, S.V.; Belton, D.J.; Kitchel, B.; Anastasiades, D.; Huang, J.; Naik,
R.R.; Perry, C.C.; Kaplan, D.L. Proceedings of the National Academy of Sciences of the United
States of America, v 103, n 25, 20 June 2006, p 9428-33

[18] Sellinger, Alan; Weiss, Pilar M.; Nguyen, Anh; Lu, Yunfeng; Assink, Roger A.; Gong,
Weiliang; Brinker, C. Jeffrey. Nature, v 394, n 6690, Jul 16, 1998, p 256-260

[19] Paris, Oskar; Zollfrank, Cordt; Zickler, Gerald A. Source: Carbon, v 43, n 1, 2005, p 53-66

[20] Gunes, I. Sedat; Jana, Sadhan C. Journal of Nanoscience and Nanotechnology, v 8, n 4,
April, 2008, p 1616-1637

[21] Hussain, F.; Hojjati, M.; Okamoto, M.; Gorga, R.E. Journal of Composite Materials, v 40, n
17, Sept. 2006, p 1511-75

[22] Crainic, N.; Marques, A.T. Source: Key Engineering Materials, v 230-232, 2002, p 656-9

10
Chapter 2. Literature Review on the Mechanical Properties of CNT Reinforced Polymer
Nanocomposites
In this chapter, the basics of carbon nanotubes and CNT reinforced polymer composites
(NRPs) are introduced and the prerequisites for successful mechanical reinforcement are
discussed. Stress transfer and stiffening models are evaluated. The focus will be on the
mechanical properties of these nanocomposites.
2.1 Carbon nanotubes – the ideal reinforcement material
First synthesized as a by-product in arc-discharge method in the synthesis of fullerenes
(Buckyball, or C60, 1985 [1]) by Japanese scientist Iijima 1991[2], carbon nanotubes have
attracted intense interest. Visualized as a single or multiple grapheme layers rolled into cylinders
consisting of one (Single Walled Carbon Nanotube, or SWCNT) or multiple (Multiwalled Carbon
Nanotubes, or MWCNT) concentric layers of hexagonal arrangement of carbons (Figure 2.1),
carbon nanotubes is endued with ultra high modulus[3], strength[4], exceptionally high axial
thermal conductivity[5-6], and unique electrical conductivity (metallic, semi-metallic, or
semi-conducting depending on the rolling direction of the grapheme) [7]. These unique properties
promise a wide range of potential scientific and industrial applications including energy storage,
field emission, probes, anti-static packaging, and nanotube based electronics, etc. [8-10] In
particular, in addition to the superior modulus & strength, the small size, low density, high aspect
ratio, exceptional resilience, and large deformation at break [11-12] makes CNT the ideal

11

(a)

(b)
Figure 2.1 a. SEM image of SWCNT bundles made of ~ 100 SWCNTs. Scale bar 10 nm; Credit:
R. E. Smalley/ Rice University. b. A multi-walled carbon nanotube partly filled with an iron
carbide nanowire. Scale bar 5 nm. Credit: Banhart/Johannes Gutenberg University.

12
nanoscopic fiber reinforcement for polymer nanocomposites with ultra-high specific modulus and
specific strength.
Depending on the angle between nanotube axial direction and vector describing the
hexagonal lattice on CNT surface, SWCNTs with different electronic properties and diameters
(mostly ~ 0.7 -1.4 nm) are described using one of three different types: armchair, zigzag, and
chiral. [13] Due to SWCNTs’ one-dimensional nature, charge carriers can travel through them
without scattering resulting in ballistic transport, in which Joule heating is minimized so that
nanotubes can carry very large current densities of up to 100 MA/cm2. [14] SWCNTs are also
very conductive for phonons. Theory predicts a room temperature thermal conductivity of up to
6000 W/m K. [15] On the other hand, MWCNTs with diameters range from ~ 2 to 100 nm consist
of several layers of coaxial carbon tubes with a constant interlayer separation of 0.34 Angstrom,
between which interwall coupling is relatively weak. [16-17] Electronically, MWCNTs act as
either metals or very small bandgap semi-conductors. Ballistic conduction has been observed in
MWCNTs by a number of groups[18] and thermal conductivities as high as 3000 W/m K have
been measured. [19]
Both SWCNTs and MWCNTs are being prepared by a variety of methods, including
arc-discharge[20], laser ablation[21], high-pressure carbon monoxide (HIPCO) conversion[22],
chemical vapor deposition (CVD) [23], electrolysis[24], and solar energy methods[25]. A length

13

Table 2.1 Electrical, mechanical, and absorption properties of carbon nanotubes and the
corresponding applications. [3-19]

Properties Reported Values Applications
Electrical Properties
Electrical conductivity
1000-200,000 S/cm, 10 e +13
A/m2
Thermal conductivity 2000 W/m/K
Electromagnetic wave shielding

Functional composites
Electron field emission FED
Mechanical Properties
Young’s modulus 0.9 -1.28 TPa
Tensile strength 30-150 GPa
Reinforcement for composite
materials
Meso-pore and large surface area
Absorption
Adsorption 300-1000 m
2
/g
Templates

Hydrogen storage (fuel cells)
Re-chargeable lithium batteries
(improved lifetimes)
Super-capacitor electrode materials
Biosensors

14
of tens or hundreds of microns (or even in millimeters or centimeters [26-27]) is typically
obtained. It should be noted that, while high quality CNTs can be produced, some defects, such as
Stone-Wales defects [28], are always present and the amount of defects are strongly dependent on
the synthesis method used. [29] Well-graphitized, relatively defect free MWCNT can be produced
by the arc discharge method. The presence of defects may significantly affect the physical and
chemical properties of the nanotubes.
Since the emphasis of this dissertation is on the use of carbon nanotubes as a
reinforcing phase within a polymer material, the mechanical properties of carbon nanotubes will
be the focus. Much of the initial work studying the mechanical properties of nanotubes was
consisted of simulation methods based on ab initio models and molecular mechanics/dynamics,
before sufficient quantities of nanotubes were produced to allow mechanical measurements. As
early as 1992-1993, theoretical calculations have predicted a wide range of Young’s moduli for
very small single wall carbon nanotubes ranging from 0.5 to 5.5 TPa. [30-31] Overney et al.
calculated the rigidity of short SWCNT using ab initio local density calculations to determine the
parameters in a Keating potential. The calculated Young’s modulus was 1500 GPa, similar to that
of graphite. [31] This was followed by a range of papers predicting that the Young’s modulus of
nanotubes was close to 1 TPa. Yao et al.’s molecular simulation showed that Young's modulus of
SWCNTs increases significantly with decreasing tube diameter and increases slightly with
decreasing tube helicity. [32] Most of these models assume defect-free nanotubes. Since strength

15
is more closely related to the presence of defects within a material, it has been expected that low
defect SWCNTs formed may approach theoretical limits of SWCNT strength. Reference 33-37
indicated predicted SWCNT strength ranging from 70Gpa to 140.4 GPa with ultimate strain from
11% to as much as 30%.
More recently, a great deal of progress has been realized in the manipulation and testing
of individual nanotubes and nanotube bundles, which have validated the computational &
theoretical predictions. Pristine, isolated SWCNT are rarely available to experimentalists.
Therefore, the first measurement of the Young’s modulus of CNTs came from Treacy and
co-workers’ [38] measurement on MWCNTs. TEM was used to measure the mean-square
vibration amplitudes of arc-grown MWCNTs. The average value of the Young’s modulus derived
from this technique was 1.8TPa (11 tubes measured between 0.40 TPa and 4.15 TPa). The authors
suggest a trend for higher moduli with smaller tube diameters. Poncharal et al. [38] induced
electrostatic and dynamic mechanical deflections for cantilevered MWCNTs in TEM, and
suggested strong dependency of CNT modulus on tube diameter. Novel experimental work
looking at the fracture behavior of nanotubes has been carried out using a nanostressing stage
located within an SEM. Yu et al. [39] in 2000 performed stress–strain measurements on
individual MWCNTs synthesized through arc discharge inside an electron microscope. For a
range of tubes they obtained modulus values of 0.27–0.95 TPa. In addition, they showed fracture
of MWCNT at strains of up to 12% and with strengths in the range 11–63 GPa. This allows the

16

Figure 2.2 Sword and sheath mechanism in multiwalled carbon nanotube.[39] (A) A MWCNT
having a section length of 6.9 microns under tensile load just before breaking. (B) After breaking,
one fragment of the same MWCNT was attached on the upper AFM tip and had a length of 6.6
microns. (C) The other fragment of the same MWCNT was attached on the lower AFM tip and
had a length of 5.9 microns.

17
estimation of nanotube toughness at 1240 J/g. More interestingly failure was observed at the outer
tube with the inner walls telescoping out in a “sword and sheath” mechanism (Figure 2.2) due to
the low intertube interaction. CVD grown MWCNTs are expected to display significantly reduced
values. CNT bending test via AFM by Salvetat et al. [40] reported arc grown CNT has modulus
values ~0.8 -0.16/+0.41 TPa as compared to ~10-50GPa for CVD grown CNTs. Xie et al. [41]
made miniature stress–strain measurements on long CVD-MWCNT. They measured a modulus of
0.45 TPa and values of tensile strength of ~1.5-3.6 GPa. The much larger variation in modulus for
CVD-MWCNT compared to arc-MWCNT strongly suggests that the modulus is very sensitive to
defect concentration and type.
The method of measuring thermal vibration amplitudes by TEM has been extended by
Krishnan et al. [42] to measure SWCNTs at room temperature. The average of 27 tubes yielded a
value of E = 1.3 +0.6/ −0.4 TPa. Salvetat et al. [43] measured the deflection of SWCNTs clamped
between membrane pores and observed a tensile modulus of ~1 TPa for small diameter SWCNT
bundles. According to Yu et al. [44], for SWCNT bundles the maximum tensile strain was
estimated to be 5.3%, with the tensile strength and modulus of the individual SWCNTs estimated
to be 13 to 52 GPa and 320 to 1470 GPa respectively. However, the properties of larger diameter
bundles were dominated by shear slippage of individual nanotubes within the bundle, especially
for tubes at the outer layer of the bundle.

18
While the predicted as well as the measured properties of carbon nanotubes compare
quite favorably to conventional reinforcement fibers, issues related to the effective incorporation
of CNTs within a polymer matrix, must be developed in order to fully utilize the properties of the
nanotubes in structural polymer composites, as will be discussed next.
2.2 CNT polymer nanocomposites – Challenges
The discovery of CNTs has created a high level of activity in materials research
towards the practical realization of extraordinary properties of CNTs with their numerous
possibilities for new composite materials. In particular, composites obtained by dispersing
nanotubes into different polymeric matrices, or carbon nanotube reinforced polymers (NRPs),
have attracted wide attention in order to develop ultra light weight and strong materials. [45-46]
Compared with their macroscopic counterpart of fiber reinforced composites, the most important
difference of NRPs is the unique properties introduced by CNTs and small size of the filler
particles, which not only prevents large stress concentrations and subsequently preserves the
ductility of the matrix polymers but also leads to an exceptionally large interfacial area that
dominates the bulk composite properties.
The incorporation of CNTs into a polymer matrix, thermoset or thermoplastic, has been
facilitated by a variety of processing techniques. These techniques include solution mixing[47],
sonication[48], coagulation[49], melt mixing/compounding[50], micro-emulsion[51],
ball-milling/pulverization[52-53], high speed shearing[54], electrospinning[55], in-situ

19
polymerization[56], chemical functionalization[57], layer by layer assembly[58], surfactants
assisted processing[59], etc. or a combination of the these techniques. Despite of the difference in
these processing techniques, two trends are obvious. One is that the utility of nanoparticles is
becoming increasingly functional, and one expects achieving the combination of different
functionalities such as electronic, optical, magnetic properties, other than the long term challenge
in terms of mechanical performances. The other is the tendency to improve understanding on the
fundamental mechanisms in manipulating nanoparticles and controlling over the microstructure
of NRPs.
The experiment observations have shown promising increases in mechanical, electrical,
and thermal properties for NRPs. Coleman et al. estimated that dY/dVf, or Young’s modulus
reinforcement was as high as ~1000 GPa for a couple of studies. [60]; electrical conductivity in
the range of 2S/m at ultra low percolation threshold of 0.0025wt% were observed by Sandle et al.
[61], and 125% increase in the thermal conductivity at room temperature with 1wt% of unpurified
SWCNT material [62]. The improvement brought about by the incorporation of nanotubes,
although notable, remains minor with regard to the potential based on the properties of the
nanotube itself and theoretical predictions. Many uncertainties, such as purification/selection of
CNTs (diameter, chirality, and length), dispersion, alignment, stress transfer properties, and
processing variables for these nanocomposites, still remain to be very challenging topics.

20
2.2.1 Purification and selection
The as-produced carbon nanotubes consisting of tubes of different chirality, length, and
diameter [63] are mixed with amorphous carbon, bucky onions, spheroidal fullerenes, polyhedron
graphite nano-particles, and metallic catalyst particles. [64] The difference in length, diameter,
chirality, and impurity content can significantly affect the final properties of the nanocomposites.
Both the purification of the as-produced nanotubes and the selection of CNTs over the length,
diameter, and chirality have received enormous attention and attained moderate success. [65-66]
The majority of purification methods for carbon nanotubes are based on oxidation,
which occurs preferentially at the nanotube ends and on nano-particles with topological defects.
Using gas phase oxidation method, SWCNTs are purified by burning away of the tube ends and
amorphous carbon at ~700 degree Celsius in air or oxidative gas. [67] However, the yield is
extremely low (<5%) and metal catalyst particles still remain in the mixture. On the other hand,
liquid phase oxidation methods using oxidative acid prove to be effective in eliminating both
amorphous carbon as well as metallic particles. Liu et al. purified CNTs by a mixture of
concentrated sulfuric acid and nitric acid. [68] Electrochemical cyclic voltammetric (CV)
oxidation behavior of arc discharge derived SWCNTs was investigated by Fang et al. [69] in
KOH solution. However, oxidation methods create defects on CNTs and reduce CNT aspect ratio,
and some require additional steps to remove the metal catalyst particles. Alternatively,
physical-based purification methods including filtration [70], centrifugation [71], solubilization of

21
funtionalized CNTs [72], electrophoresis [73], and annealing [74] were used in the purification of
CNTs. Yamamoto et al. [75] used AC electrophoresis to treat the CNTs dispersed in iso-propyl
alcohol. They found that the separation from impurity particles depended on the frequency of the
applied field. Combinations of abovementioned chemical and physical methods are expected to
achieve optimal results. Tohji et al. [76-77] suggested a hybrid purification method that included
hydrothermally initiated dynamic extraction (HIDE) treatment along with extraction of fullerenes,
thermal oxidation, and dissolution in 6 M hydrochloric acid. Liu et al.[78] developed a
purification method that consisted of refluxing in 2.6 M nitric acid and re-suspending CNTs in pH
10 water with surfactant followed by filtration with a cross-flow filtration system. This is an
efficient method to purify large quantities of CNTs owing to the combined advantages of the
chemical and physical methods. Li et al.[79] also developed a multi-step method to purify
SWCNTs synthesized by CVD, which includes acid washing, ultrasonication and freezing
treatments in liquid nitrogen. After purification, SWCNTs with a purity of 95% (estimated from
EDS and SEM) and a yield of 40 wt% were obtained and the procedure did not destroy the
SWCNT bundles.
The electronic and optical properties of CNTs strongly depend on the chirality[80],
length[81], and diameter[82-83], the selection over which is important for future multifunctional
smart NRPs. McCarthy et al.[84] used a functional organic polymer,
poly(m-phenylene-co-2,5-dioctoxy-pphenylenevinylene) (PmPV), as a filtration system to purify

22
CNTs. They found that the solution of PmPV is capable of suspending nanotubes indefinitely
whilst the accompanying amorphous graphite separates out. Bandow et al. [85] devised a process
that involves the suspension of carbon nanospheres, metal nanoparticles, and SWCNTs in an
aqueous solution using a cationic surfactant and the subsequent trapping of SWCNTs on a
membrane filter. No oxidative treatment is required. Li et al. [86] showed that narrow
diameter/chirality growth combined with chemical separation by ion exchange chromatography
(IEC) greatly facilitates achieving single (m,n) SWCNT samples, as demonstrated by obtaining
highly enriched (8,4) SWNTs with near elimination of metallic SWCNTs existing in the as-grown
material. Duesberg et al. [87] showed that carbon nanospheres, metal particles, and amorphous
carbon could be successfully removed by size exclusion chromatography (SEC) applied to
surfactant stabilized dispersions of SWCNT raw material and length separation of the tubes could
also be achieved. Although suitable to treat small amounts of as produced CNTs, these novel
techniques certainly advance our understanding in the selective separation of CNTs.
2.2.2 Dispersion
In order to achieve efficient load transfer to the nanotube network, nanotubes must be
uniformly dispersed to the level of isolated or individual tubes into the polymer matrix.
Aggregates of nanotubes effectively reduce the reinforcement aspect ratio, introduce structural
defects, and lead to stress concentration in the NRPs. Enhanced dispersion can also assist with the
realization of a network for conductivity of electrical and thermal energy at ultra low CNT

23
content. However, the huge surface area of CNTs makes the Van de Waals force difficult to over
come. [88] In addition, carbon nanotube is neither hydrophilic nor hydrophobic. [89] Thus,
dispersing carbon nanotubes individually in a polymer matrix is a daunting task.
SWCNTs tend to form aligned CNT bundles that randomly entangled together due to
their large surface energy and great flexibility. These bundles contain a number of SWCNTs with
varied chirality arranged in hexagonal arrays and the mechanical properties of which are
generally inferior to those of isolated SWCNT due to the inter-tube slippage, which presents a
serious hurdle in structural composites applications. Yu et al.[44] observed that failure occurred
for the nanotubes on the perimeter of the bundle only with the rest of the tubes slipping apart.
While do not form aligned bundles, MWCNTs are as easy to agglomerate and entangle as
SWCNTs bundles, and thus limit the efficiency of MWCNTs on polymer matrices. (Figure 2.3)
Agglomeration is prominent in CVD grown CNTs where there exists substantial nanoscale
spaghetti-like entanglement of nanotubes. [90] The effects of poor dispersion can be seen in a
number of systems when the nanotube loading level is increased beyond the point where
significant aggregation begins. This is generally accompanied by a decrease in strength and
deformation at failure in NRPs. [91] Salvetat et al. [92] reviewed the effect of dispersion of
MWCNTs on the mechanical properties of polymer/MWCNT composites, and found that poor
dispersion and rope-like entanglement of MWCNTs led to drastic weakening of the composites.

24

Figure 2.3 Nanotube agglomerations and entanglement. Scale bar: 1 micron. Credit:
Arkema.com

25
The dispersion of CNTs in polymeric matrices could be optimized through thermal mechanical
force, non-covalent, or covalent modification methods.
Firstly, the dispersing CNT can be achieved by mechanical means such as high shear
force and ultrasonic cavitation. Li et al. [93] and Obrzut et al. [94 demonstrated that the
dispersion of MWCNTs in a polymer matrix and the associated mechanical and electrical
propperties depend greatly on the shear stress exerted during melt processing. Temperatures and
pressures of up to 15 000 K, 1000 atm, fluid strain rate of up to 107 s-1 [95] can be created in
ultrasonication process, which has been widely employed in CNT dispersion and
functionalization.
Secondly, the nanotube – polymer interaction could be enhanced through the use of
anionic, cationic, and non-ionic surfactants as coupling agents, which help to overcome the
intertube attractive force and allow good dispersion of individual nanotubes within the polymers
matrices. [96-100] In addition, π-conjugated polymers such as PmPV could wrap around CNTs
and form stable suspensions in organic solutions. [101-103] Moreover, long-ranged entropic
repulsion and electrical interaction among polymer-decorated tubes acts as a barrier that prevents
the tubes from approaching. [104-105]
Thirdly, CNT surface modification by covalently introduce compatible molecular species
onto CNT surface could also enhance dispersion through either the match of effective solubility
of CNTs with the targeted polymer systems [106] or barrier effects that come from the steric or

26

Figure 2.4 Schematic representations of the mechanism by which surfactants help to disperse
SWCNT. Credit: Yurekli et al. University of Houston

27
electrical repulsion similar to those non-covalently modified species. Strategies for chemical
functionalization of carbon nanotubes have been extensively reported and two different trends are
observed. One group of methods involves so the called pseudo carbon chemistry by oxidizing the
defect sites on carbon nanotubes (oxidation acid treatment [68], ozone[107], plasma oxidization
[108], ball milling [52]) to introduce functionalities such as carboxyl, carbonyl, and hydroxyl on
CNT surface, which could be used to attach additional molecular species in the secondary
reactions. The other group of methods employs the orthodox sp2 carbon chemistry ，such as
nucleophilic addition, cycloaddition, alkylation, and fluorination similar to that used for fullerene
chemistry. Detailed review work for the chemical modification of CNTs can be found in the
resent articles by Banerjee et al. and Lin et al. [109-110] Despite the abovementioned
achievements in solubilizing CNTs in polymer systems, maintaining separated nanotubes during
the processing of NRPs is still the subject of ongoing work.
2.2.3 Interface
As defined in traditional composites, the interfacial region is the region beginning at the
point in the filler at which the properties differ from those of the bulk filler and ending at the
point in the matrix at which the properties start to become equal to those of the bulk matrix. It can
be a region of altered chemistry, chain mobility, degree of cure, or crystallinity (for crystalline or
semi-crystalline polymers). [111] A number of studies suggest an interfacial region of polymer
with morphology and properties different to the bulk is formed in nanotube reinforced polymer

28

Figure 2.5 Direct observation of polymer sheathing in carbon nanotube-polycarbonate composites
by Ding et al. [119]

29
matrices, such as PMMA, PS, PC, and Epoxy. [112-114] Observations of crystallinity nucleation
at the interface have been made by a number of groups. Assouline et al. [115] observed nucleation
of crystallinity in the presence of nanotubes. The crystallites were observed to be fibular in nature
as compared to spherical in the pure polymer. Coleman et al. [116] observed similar behavior and
linked it to the formation of a high strength crystalline coating. In addition, dendritic polymer
growth has been reported to initiate from defects on CNTs.[117] That these interfacial regions
have different properties was shown by Barber et al.’s pullout experiments[118], which suggest
that interfacial polymer had anomalously high shear strength. Ding et al. [119] observed thick (10
s of nm) layers of polymer coating nanotubes protruding from composite fracture surfaces, an
indication of both a high interfacial shear stress (IFSS) and a layer of high shear strength
polymer.(Figure 2.5)
While there is much debate as to the nature of the polymer morphology and property at
the interface, it is clear that it plays a major role in the mechanical reinforcement process. Fig. 2.6
shows the interface area per unit volume as a functional of particle size for spherical particles that
are ideally dispersed: the interfacial region is essentially a large part of the entire matrix and
increase exponentially with nano-particle volume fractions. Even if the interfacial region is only a
few nanometers in thickness, the entire polymer matrix has a different behavior than the bulk due
to the large surface area of the nano-particles. With more extended interface region, the polymer
matrix behavior can be altered at much smaller filler loadings. A topic of equal importance in

30

Figure 2.6 Fraction of non-bulk polymer in the interphase region as a function of volume fraction
of fiber inclusion, where t is the interphase thickness and r
f
is the radius of the nanotube/fiber
inclusion.[114]

31
terms of mechanical reinforcement is the load transfer at the interface. [120-121] Researchers
have found evidence of promising nanotube polymer stress transfer in NRPs and the detailed
mechanisms of stress transfer will be discussed in section 2.3.1
Although recent molecular dynamics study suggests that polymer morphology, and
specifically the helical wrapping of the polymer around the nanotubes is a key factor influencing
the strength of the interface (Lordi and Yao 2000 [122]), it is unclear whether such processing
agents can be employed to promote nanotube dispersion without compromising the mechanical
performance of NRPs. Tailoring the CNT matrix interaction through chemical methods that
introduce molecular species both solubility matching and chemically active with the matrix
system, would attain not only good dispersion but also strong interface adhesion as is proposed by
Sun et al.[123]
2.2.4 Alignment and assembly
In traditional fiber reinforced composites, macroscopic fibers are aligned or organized in
such a way to maximize the load bearing efficiency and to meet the targeted design criteria.
Orientation in the direction of applied forces allows for greater load transfer and fully utilization
of the reinforcement potential of fibrous fillers, which is a factor of five in composite modulus
with perfect fiber alignment as compared with random orientation.[124] In addition, having all
the reinforcement oriented in the same direction allows for easier transfer of thermal and
electrical energy along the alignment direction. Similar concept can be used for CNT

32

Figure 2.7 Direct conversion of nanotube forest aerogel into aligned CNT sheets and assemblies
of those sheets.[127]

33
nanocomposites. However, the alignment or organization of nano-scopic fibers is constrained by
the small size and efficiency. Despite of these difficulties, moderate success in the manipulation
of carbon nanotubes for macroscopic materials has been reported recently. Alignment of CNTs in
the matrix could be enhanced by ex situ techniques, such as vapor phase infiltration [125-126],
aerogel[127], shear force[128-129], electromagnetic fields[130-133], template assisted
methods[134], electrospinning[135], tensile loading[136], and liquid crystalline phase induced
methods[137], etc. For example, Ajayan et al. [138] found that cutting thin slices (on the order of
100 nm) of a nanotube-reinforced epoxy film introduced preferential orientation via shear flow.
Strobl et al. [139] observed the alignment of multiwalled carbon nanotubes parallel to the wave
vector of surface acoustic waves. An alternative method that may be more suitable for larger
samples is tensile loading of the NRP at temperatures above the glass transition temperature of
the polymer (Bower, Rosen et al. 1999[136]).
2.2.5 Process dependency
Not to mention the ex-situ methods in facilitating the CNT alignment discussed above,
the effective properties of the NRPs strongly depend on the processing technique(s) used. For
example, shortening and mechanical damage such as bending, cutting, and dislocations in the
carbon structures can be induced by the sonication technique. [48] Shortening the nanotubes and
introduction of mechanical damages can be detrimental to the mechanical properties of NRPs
since it reduces either the high aspect ratio or the modulus of the nanotubes. Pegel et al. [140]

34
studied the relationships between melt processing conditions and MWNT dispersion and
distribution polycarbonate samples. In another example, a combination of solvent casting and
melt mixing was also found to produce a high degree of nanotube alignment (Haggenmueller et
al. 2000[141]). In addition, NRP processing techniques are usually used in combination with one
another in order to attain optimum results. Thus, literature reports on the properties are hard to
compare and care must be taken to deconvolute the processing dependency of these materials.
Coleman et al. (2006 [142]) gave a detailed review on the dependency of NRP mechanical
properties over a variety of processing techniques, including melt mixing, solution casting, and
in-situ polymerization.
2.3. Mechanical properties of CNT polymer nanocomposites
2.3.1 Interface stress transfer
If poor load transfer between CNTs and the polymer matrix attributed to the relative
slipping of individual tubes within the MWCNT (see telescopic pullout [17]) and the rope
(relative slippage [44]) is ignored, the nanotubes provide load transfer in the similar way as
non-continuous fibers in conventional composite systems. The most important requirement for a
nanotubes reinforced composite is that external stresses applied to the composite as a whole are
efficiently transferred to the nanotubes through interface shear stress (IFSS), allowing them to
take a disproportionate share of the load. [143] The stress transferred can be ascertained by
Raman spectroscopy through the shifting of D’ band (located around 2662 cm-1), which is

35

Figure 2.8 TEM images of a MWCNT crossing a hole in an epoxy resin matrix. A. MWCNT
gapping a hole; B.MWCNT partial pullout by SPM; C. force distance curve.

36
sensitive to stress in the nanotubes. This peak tends to shift down in frequency when the
nanotubes are under tension. [144] As the external load increases to some critical value, rupture
or debonding through matrix failure (matrix shear strength < local shear stress) or interface failure
(IFSS < local shear stress) will occur under the large shear stresses in the NRPs.
While IFSS is challenging to measure for fillers as small as carbon nanotubes, some
progress has been made. Several tensile tests on nanotube/polymer nanocomposites have been
reported in the literature to study the bonding behavior between the nanotubes and the matrix, in
which the interface shear strength ranging from 20 to 376 MPa was reported. Barber et al. [145]
mounted a MWCNT onto an AFM tip before pushing it into a heated polymer film. On cooling
they measured the force required to pull the tube out, obtaining values between 20 and 140 MPa.
Cooper et al. [146] used an AFM to manipulate nanotubes protruding from holes in an
epoxy/CNT film. They observed IFSS of 300–376 MPa for tubes with short embedded lengths
but values of 30–90 MPa for tubes with longer embedded lengths suggesting that end effects are
important.(Figure 2.8) However, it should be pointed out that all these results are for
non-covalently bonded composites. Much higher values are expected when nanotubes are
covalently attached to the matrix. Wagner et al. [147] experimentally studied the fragmentation of
MWCNTs covalently embedded within thin polymeric films under both compressive and tensile
strains. They found that the polymer/CNT interfacial shear stress is of the order of 500 MPa,
which is an order of magnitude higher than traditional fiber composites.

37
One of the critical issues yet to be solved is the lack of accurate understanding about load
transfer and a couple of mechanisms are proposed for NRPs. The first one is mechanical
interlocking. The curvature of CNTs or their bundles can be very helpful for the formation of
mechanical interlocking, with a trade off in sacrificing their aspect ratio. In some NRPs, the misfit
of thermal expansion coefficients between CNTs and the matrix also helps to build up the
interlocking. The second mechanism is atomic level bonding between CNTs and the matrix. The
formation of interfacial bonding is largely depended on the chemistry between CNTs and the
polymer matrices. Functionalization of CNTs can help to establish chemical bonding between
CNTs and the matrix and significantly enhance stress transfer. A third mechanism is Van der
Waals force and electrostatic force. The Van der Waals force contributes least to IFSS among the
load transfer mechanisms mentioned above. For instance, the interfacial shear stress by covalent
bonding could reach up to 500 MPa while only reach ~2MPa by Van der Waals force [148].
Efforts have been made to assess the ability of stress transfer through the nanotube–matrix
interface through molecular dynamics (MD) and continuum-based simulations.
In MD simulations, the forces acting on particles in a defined cell are calculated through
force fields (empirical, semi-emperical, or ab-initial force fields, usually a Lennard-Jones 6–12
like potential), and the classical Newtonian equations of motion are integrated numerically to
simulate the evolution of the system. In MD without considering the atomic bonding between the
nanotubes and the matrix, the non-bond interactions consists of electrostatic/Van der Waals

38

Figure 2.9 Illustration MD simulations of the cross-linked systems. Left: crystalline matrix.
Right: amorphous matrix. [150]

39
interactions, deformation induced by these forces, as well as stress/deformation arising from
mismatch in the coefficients of thermal expansion. Usually, these non-bonded interactions are
only capable of producing modest cohesive strength and almost null shear strength. The MD
simulation work of Yang et al. [149] point out the specific monomer structure plays a very
important role in determining the strength of interaction between nanotubes and polymers. The
results of their study suggest that polymers with a backbone containing aromatic rings are
promising candidates for the noncovalent binding of carbon nanotubes into composite structures.
The MD simulations of Frankland et al. [150] showed that interfacial shear strengths could be
enhanced by over an order of magnitude through CNT surface functionalization and formation of
MWCNT/matrix covalent bonds at ~ 1% of the surface carbon atoms. They also found that the
tensile strength of the nanotubes at the functionalization level could not have a significant
difference. In their study, the nano-mechanical interlocking due to mismatch of coefficient of
thermal expansion was also observed and such interlocking after curing or consolidation could
substantially increase the friction at the interface and increase the pullout strength of the
nanotubes. Although confirmed experimental observations, MD simulations are computationally
prohibitive and limited to modeling systems containing only a small number of atoms and
undergoing deformations over a small period of time (~ ps, ns).
In contrast, continuum-based micromechanics and computational models are efficient
and, therefore, very desirable for parametric studies. A number of continuum-based models, such

40
as Kelly-Tyson model and shear-lag model, have been used to elucidate the stress transfer
mechanism at the interface. Wagner et al. [151] employed a modified Kelly–Tyson model, which
assumes uniform interfacial shear and axial normal stresses, to study the stress transferability in
their MWCNT fragmentation test. The shear-lag model originally proposed by Cox (1952)
provides a good estimate of the stresses in the fiber transferred from the matrix through the
interface. Based on this model, which considers two concentric cylinders (i.e., a matrix cylinder
and a fiber cylinder), several additional shear-lag type models [152] have been proposed to
incorporate the effect of the surrounding fibers. However, these variations of shear lag models
considered stress transfer across the curved interfaces only and assumed that both fiber-ends
(breaks) are traction-free. In addition, some important morphological features of
nanotube-reinforced composites (including diameter, wall thickness and chirality and
distribution) are not incorporated in existing shear-lag models. More recently, multi-scale
computational models for interfacial stress transfer in nanotube-reinforced polymer composites
has been developed. Namilae et al. [153] studied the atomic scale interface effects on composite
behavior and developed a hierarchical multiscale methodology linking molecular dynamics and
the finite element method through atomically informed cohesive zone model parameters to
represent interfaces and fiber pullout. Li and Chou, who employs the molecular structural
mechanics approach to characterize the nanotube and the finite element method to analyze the

41
deformation of the polymer matrix, examined the stress distributions at the nanotube/polymer
interface under isostrain and isostress loading conditions. [154]
A number of simulation predictions about the interfacial shear strength match
experimental results well. Frankland et al.[150] used molecular dynamics simulations to estimate
the IFSS at 2.5 MPa for both crystalline and amorphous polyethylene matrices. While Liao and Li
[150] calculated a much larger value of 160 MPa for PS/CNT composites using molecular
mechanics, Wong et al.[112] obtained a similar value of 186 MPa for PS/NT composites and 138
MPa for epoxy/CNT composites. In addition, Lordi and Yao calculated maximum frictional
stresses for a range of polymers coating SWCNT. These frictional forces can be associated with
the IFSS. They obtained values between 18 and 135 MPa.[156] Wall et al. applied the
Frenkel–Kontorova model to ordered monolayers of polymer wrapping SWNCT to show that
strain induced templating can occur resulting in extremely large stress transfer.[157]
2.3.2. Stiffening effect of CNTs
Ab-initial and molecular level simulations can be expensive and prohibitively time
consuming in the direct simulation of bulk materials. A great deal of theoretical work for the
stiffening mechanisms of CNTs is based on the continuum-models for the mechanical properties
of fiber reinforced composites developed since the 1950s. [158] These models use the same
basic assumptions: 1. The fibers and the matrix are linearly elastic, the matrix is isotropic, and the
fibers are either isotropic or transversely isotropic. 2. The fibers are axisymmetric, identical in

42
shape and size, and can be characterized by an aspect ratio l/d. 3. The fibers and matrix are well
bonded at their interface, and remain that way during deformation.
Historically, the V oigt and Reuss averages [159] were the first models to be recognized as
providing rigorous upper and lower bounds for composite materials. To derive the Voigt model,
one assumes that the fiber and matrix have the same uniform strain (imaging a dashpot and a
spring working in parallel), and then minimizes the potential energy. Since the potential energy
will have an absolute minimum when the entire composite is in equilibrium, the potential energy
under the uniform strain assumption must be greater than or equal to the exact result, and the
calculated stiffness will be an upper bound on the actual stiffness. The Reuss model is derived by
assuming that the fiber and matrix have the same uniform stress (imaging a dashpot and a spring
working in serial), and then maximizing the complementary energy. Since the complementary
energy must be maximum at equilibrium, the model provides a lower bound on the composite
stiffness. Based on Voigt and Reuss model, in the simplest possible case a fiber composite can be
modeled as an isotropic elastic matrix filled with aligned elastic fibers that are perfectly bonded
to the matrix and span the full length of the specimen, which in essence is the rule of mixtures
[160] (ROM, iso-strain) and inverse rule of mixtures (IROM, iso-stress) [161]. The Voigt and
Reuss bounds provide isotropic results (provided the fiber and matrix are both isotropic); while in
fact we expect aligned-fiber composites to be highly anisotropic. More importantly, when the
fiber and matrix have substantially different stiffness which is especially true for NRPs, the Voigt

43
and Reuss bounds are quite far apart, and provide little useful information about the actual
composite stiffness.
Hashin and Shtrikman [162] constructed tighter bounds employing a variational principle
for heterogeneous materials through the concept of a reference material. For an upper bound the
reference material must be as stiff as or stiffer than any phase in the composite, and for a lower
bound the reference material must have stiffness equal to or less than any phase. The resulting
bounds are tighter than the Voigt and Reuss bounds, which can be derived from the
Hashin-Shtrikman theory by assigning the reference material infinite or zero stiffness respectively.
Walpole [163] extended the Hashin-Shtrikman bounds based on classical energy principles for
anisotropic materials, including infinitely long fibers and thin disks in both 3-D aligned and
random orientations. The Hashin-Shtrikman-Walpole bounds were adopted for short-fiber
composites by Willis and McCullough [164-165].
Other than these bounding models, analytical models that could predict a full set of
elastic constants of the composites were developed. These models include the Halpin-Tsai [166]
equation and its extensions, the dilute model based on Eshelby’s equivalent inclusion[167], the
self-consistent model[168] for finite-length fibers, and Mori-Tanaka type models[169].
Halpin-Tsai model was originally developed for continuous fiber composites and based on the
work of Hill[168] and gave reasonable estimates for stiffness at low volume fractions but to
underestimate stiffness at high volume fraction. More complex micromechanics models employ

44
the conclusion from Eshelby’s equivalent inclusion (1957). Eshelby solved for the elastic stress
field in and around an ellipsoidal particle in an infinite matrix. By letting the particle be a prolate
ellipsoid of revolution, one can use Eshelby’s result to model the stress and strain fields around a
cylindrical fiber. One can use Eshelby’s result to find the stiffness of a composite with ellipsoidal
fibers at dilute concentrations. Based on Eshelby inclusion in infinite medium elasticity solution,
self-consistent scheme assumes that the infinite medium has properties of composite.
Alternatively, Mori-Tanaka method, an equivalent to the generalization of the
Hashin-Shtrikman-Walpole lower bound[170] , extends Eshelby method by a fourth-order tensor
relating average inclusion strain to average matrix strain and approximately accounting for fiber
interaction effects.
More advanced models consider the thickness of the interface layer and different
physics occurs at various length scales in NRPs in order to realistically model the mechanical
response of NRPs. Extended from the work of Z. Hashin and B. Rosen [171] and R. Christensen
and K. Lo [172], Lagoudas et al. [173] studied the effect of an interphase layer as result of
functionalization using a multi-layer composite cylinders approach. Hybrid models have also
appeared gradually in the past few years. Odegard et al. [174-175] developed a method by
substituting discrete molecular structures with equivalent-continuum models and thus serving as a
link between computational chemistry and solid mechanics.

45

Figure 2.10 Eshelby's inclusion problem. [170]

46
All of these analytical or numerical models incorporated some assumptions that led to
highly optimistic predictions of overall composite stiffness compared to experimental results. In
such cases, factors tending to soften the predictions, including MWCNT bundle clustering,
waviness of the MWCNTs, distribution of CNT morphology (length, diameter, and aspect ratio),
CNT buckling/pullout, process dependency, non-perfect bonding at the interface, and reduced
effective aspect ratio, are noted, and their effects parametrically reconsidered within the adopted
modeling framework. Brinson’s group [176-177]modeled a wavy MWCNT segment as an
infinitely long, ideally bonded sinusoidal-shaped fiber, and the dilute strain concentration tensor
for the fiber domain was extracted directly from FE solutions. Their studies showed that the
effective axial strain in the bonded wavy MWCNT increases markedly with increases in the ratio
of sinusoidal amplitude over wavelength, leading to dramatic reductions in stiffness. Buckling of
CNTs within an epoxy matrix has been accounted for by Hadjiev et al. [178]. Cooper [179], Lurie
[180], and Hammerand [181] took inclusion of less than ideal CNT adhesion to the matrix into
consideration in NRP modeling.
2.3.3 Motivation for the current work
A great deal of experimental and theoretical work has been accumulated on the effective
mechanical properties of NRPs in the literature. NRPs have been found to demonstrate lower
mechanical properties as compared with theoretical predictions. Because of the large number of
parameters that can influence these effective properties, including the method of CNT synthesis,

47

Figure 2.11 Finite element modeling of an infinitely long wavy nanotube embedded in a matrix.
[176-177]

48
size and form of the CNTs, NRP processing conditions, CNT-polymer interaction, and the
specifics of the polymer chemistry, generalizations across these studies are difficult. However,
research efforts have identified several key challenges in the engineering NRPs with superior
effective mechanical responses, including adequate dispersion, control over interface,
microscopic filler organization.
In this dissertation, efforts to address the dispersion and interface issues in particular have
lead to the use of judiciously functionalized MWCNTs in both attaining good dispersion and
producing controlled interphase regions in a DGEBA epoxy nanocomposites system. In chapter 3,
multiwalled carbon nanotubes (MWCNTs) were functionalized in a two-step acid-epoxy
functionalization process, in which suitable surface condition and reactivity compatible with the
DGEBA epoxy resin was introduced. The use of (4-dimethylamino)-pyridine as an initiator for
DGEBA homopolymerization produced covalent bonds between the functionalized MWCNTs
and the epoxy matrix through chain transfer reactions involving the secondary hydroxyls. This
process yielded uniform MWCNTs-stiff epoxy composites with significant enhancement in
flexural strength without sacrificing the elastic modulus when compared to the neat resin. In
chapter 4, an ionic bonding scheme was employed to graft epoxy-compatible amine terminated
linear polypropylene oxide (A-PPO) molecules of different molecular weight to oxidized
multi-walled carbon nanotubes (o-MWCNTs), and thereby modify the interphase of
MWCNT-epoxy nanocomposites. The amine-cured epoxy systems reinforced with different

49
functionalized MWCNTs demonstrated divergent thermomechanical properties. Fractography
revealed distinct fracture behavior for nanocomposites with long-chain A-PPO modified
MWCNTs. The results demonstrate the feasibility of interphase control and performance
engineering through the judicious choice of polymer grafts.
In chapter 5, processing condition induced CNT organization is explored. Specifically,
cross-ply laminates reinforced with basalt fibers and functionalized multi-walled carbon
nanotubes (MWCNTs) were fabricated from unidirectional epoxy prepregs. MWCNTs with
varied surface conditions were prepared by oxidization or esterification, and then dispersed into a
DGEBA epoxy system. The dispersion of the MWCNTs in the epoxy was improved by surface
modification, resulting in improved composite mechanical properties as well. Significant
increases in elastic modulus and strength were observed for epoxies with functionalized
MWCNTs, especially for esterified species. When MWCNT–filled epoxies were used as matrices
for basalt fiber/epoxy laminates, however, the reinforcement effects of MWCNTs on the
composite elastic modulus exceeded micromechanics based semi-empirical predictions and were
independent of surface functionalization. SEM morphological observations and the results of the
micromechanical model revealed that nanotube redistribution and orientation during processing
was responsible for the enhancement of fiber-dominated mechanical properties. This work
demonstrated the feasibility of in-situ alignment and dispersion of funtionalized nanotubes in
multi-scale composite laminates.

50
In chapter 6, shear wave mitigation through the incorporation of oxidized multiwalled
carbon nanotubes (o-MWCNTs) into filled epoxy blends for acoustically transparent structural
materials was demonstrated. The longitudinal wave speed, density, flexural mechanical
properties, loss factor, and insertion loss were characterized for different compositions. The
addition of 0.05wt% o-MWCNTs increased the composite flexural modulus and strength by
370% and 90% respectively with no change in the longitudinal acoustic wave speed. Analytical
predictions also indicated that higher shear loss factor provided by carbon nanotubes ameliorated
shear resonance at oblique insertion angles. Carbon nanotube additions afford the ability to
increase mechanical properties and acoustic transparency, both of which are critical for
underwater sonar windows.

51
Chapter 2 References

[1] Kroto, H.W.; Heath, J.R.; O'Brien, S.C.; Curl, R.F.; Smalley, R.E. Nature, v 318, n 6042, 14
Nov. 1985, 162-3

[2] Iijima, S. Nature, v 354, n 6348, 7 Nov. 1991, 56-8

[3]Qian, Dong; Wagner, Gregory J.; Liu, Wing Kam; Yu, Min-Feng; Ruoff, Rodney S. Applied
Mechanics Reviews, v 55, n 6, November, 2002, p 495-532

[4] Yu, Min-Feng, Journal of Engineering Materials and Technology, Transactions of the ASME,
v 126, n 3, July, 2004, p 271-278

[5] Shaikh, S.; Li, L.; Lafdi, K.; Huie, J. Carbon, v 45, n 13, Nov. 2007, 2608-13

[6] Fujii, Motoo; Zhang, Xing; Xie, Huaqing; Ago, Hiroki; Takahashi, Koji; Ikuta, Tatsuya; Abe,
Hidekazu; Shimizu, Tetsuo Physical Review Letters, v 95, n 6, Aug 5, 2005, p 065502

[7] Ebbesen, T.W.; Lezec, H.J.; Hiura, H.; Bennett, J.W.; Ghaemi, H.F.; Thio, T. Nature, v 382, n
6586, Jul 4, 1996, p 54-55

[8]Novak, J.P.; Lay, M.D.; Perkins, F.K.; Snow, E.S. Solid-State Electronics, v 48, n 10-11
SPEC. ISS., October/November, 2004, p 1753-1756

[9]Baxendale, M. IEE Proceedings Nanobiotechnology, v 150, n 1, June, 2003, p 3-8

[10]Young Hee Lee; Kay Hyeok An; Seong Chu Lim; Won Seok Kim; Hee Jin Jeong; Chil-Hoon
Doh; Seong-In Moon. New Diamond and Frontier Carbon Technology, v 12, n 4, 2002, 209-28

[11] Wang, Yu; Fang, Daining; Soh, Ai Kah; Liu, Bin Acta Mechanica Sinica, v 23, n 6,
December, 2007, p 663-671

[12] Chun Tang; Wanlin Guo; Changfeng Chen Physical Review Letters, v 100, n 17, 2 May
2008, 175501-1-4

[13] O'Connell, Michael. Carbon nanotubes : properties and applications Publisher: CRC Taylor
& Francis, Pub date: 2006. ISBN: 0849327482

52
[14] Windle, A.H.; Moisala, A.; Li, Q.; Kinloch, I.A. Composites Science and Technology, v 66,
n 10, Aug. 2006, 1285-8

[15]Haggenmueller, Reto; Guthy, Csaba; Lukes, Jennifer R.; Fischer, John E.; Winey, Karen I.
Macromolecules, v 40, n 7, Apr 3, 2007, p 2417-2421

[16] Saito, R. (Riichiro) Dresselhaus, G. Dresselhaus, M. S. Publisher: Imperial College Press ;
Pub date: 2005, c1998. ISBN: 1860940935

[17] Min-Feng Yu; Yakobson, B.I.; Ruoff, R.S. Journal of Physical Chemistry B, v 104, n 37, 21
Sept. 2000, 8764-7

[18] Chiu, H.-Y.; Deshpande, V.V.; Postma, H.W.Ch.; Lau, C.N.; Miko, C.; Forro, L.; Bockrath,
M. Physical Review Letters, v 95, n 22, 25 Nov. 2005, 226101/1-4

[19] Kim, P.; Shi, L.; Majumdar, A.; McEuen, P.L. Physical Review Letters, v 87, n 21, 19 Nov.
2001, 215502/1-4

[20] Ando, Yoshinori (Meijo Univ); Iijima, Sumio. Japanese Journal of Applied Physics, Part 2:
Letters, v 32, n 1A-B, Jan 15, 1993, p L107-L109

[21] Cheng, Da-Dian; Yu, Rong-Qing; Liu, Zhao-Yang; Zhang, Qiang; Wang, Yu-Huang;
Huang, Rong-Bin; Zhan, Meng-Xiong; Zheng, Lan-Sun Chemical Journal of Chinese
Universities, v 16, n 6, Jun, 1995, p 949

[22]Scott, C.D.; Povitsky, A.; Dateo, C.; Gkcen, T.; Willis, P.A.; Smalley, R.E. Journal of
Nanoscience and Nanotechnology, v 3, n 1-2, Feb.-April 2003, 63-73

[23] Yudasaka, Masako; Kikuchi, Rie; Matsui, Takeo; Ohki, Yoshimasa; Yoshimura, Susumu;
Ota, Etsuro. Applied Physics Letters, v 67, n 17, Oct 23, 1995, p 2477

[24] Hsu, W.K.; Terrones, M.; Hare, J.P.; Reid, D.G.; Walton, D.R.M.; Kroto, H.W. Proceedings
of the International Winter school on Electronic Properties of Novel Materials. Fullerenes and
Fullerene Nanostructures, 1996, 226-31

[25]Anglaret, E.; Bendiab, N.; Guillard, T.; Journet, C.; Flamant, G.; Laplaze, D.; Bernier, P.;
Sauvajol, J.-L. Carbon, v 36, n 12, 1998, 1815-20

53
[26]Shengdong Li; Zhen Yu; Rutherglen, C.; Burke, P.J. Nano Letters, v 4, n 10, Oct. 2004,
2003-7

[27] Ren, Z.F.; Guang-Yong Xiong; Wang, D.Z. Carbon, v 44, n 5, April 2006, 969-73

[28]Tunvir, Kazi; Kim, Amkee; Nahm, Seung Hoon Nanotechnology, v 19, n 6, 13 February,
2008, p 065703

[29]Fisher, F.T. Dissertation 2002, Northwestern University

[30]Robertson, D.H.; Brenner, D.W.; Mintmire, J.W. Physical Review B, v 45, n 21, 1 June 1992,
p 12592-5

[31] Overney, G.; Zhong, W.; Tomanek, D. Zeitschrift fur Physik D, v 27, n 1, 1993, p 93-6

[32] Yao, N.; Lordi, V. Journal of Applied Physics, v 84, n 4, 15 Aug. 1998, 1939-43

[33] Qingzhong Zhao; Marco Buongiorno Nardelli; Bernholc, J. Physical Review B, v 65, n 14, 1
April 2002, p 144105/1-6

[34] M. Meo, a, and M. Rossia Material Research Center, Department of Mechanical
Engineering, University of Bath, Bath, Ba2 7AY, UK

[35] J.R. Xiao, B.A. Gama, J.W. Gillespie jr, Int. J. Solids Struct. 42 (2005) 3075–3092.

[36] T. Natsuki, M. Endo, Carbon 42 (2004) 2147–2151.

[37] S.L. Mielke, D. Troya, S. Zhang, Li Je-Luen, S. Xiao, R. Car, R.S. Ruoff, .C. Schatz, T.
Beliytschko, Chem. Phys. Lett. 390 (2004) 413–420.

[38] Treacy, M.M.J.; Ebbesen, T.W.; Gibson, J.M. Nature, v 381, n 6584, 20 June 1996, 678-80

[38] Poncharal, P.; Wang, Z.L.; Ugarte, D.; De Heer, W.A. Science, v 283, n 5407, 5 March
1999, 1513-16

[39] Min-Feng Yu; Lourie, O.; Dyer, M.J.; Moloni, K.; Kelly, T.F.; Ruoff, R.S. Science, v 287, n
5453, 28 Jan. 2000, 637-40

54
[40] Salvetat, J.-P.; Kulik, A.J.; Bonard, J.-M.; Briggs, G.A.D.; Stockli, T.; Metenier, K.;
Bonnamy, S.; Beguin, F.; Burnham, N.A.; Forro, L. Advanced Materials, v 11, n 2, Jan 22, 1999,
p 161-165

[41] Sishen Xie; Wenzhi Li; Zhengwei Pan; Baohe Chang; Lianfeng Sun. Journal of the Physics
and Chemistry of Solids, v 61, n 7, July 2000, 1153-8

[42]Krishnan, A.; Dujardin, E.; Ebbesen, T.W.; Yianilos, P.N.; Treacy, M.M.J. Physical Review
B, v 58, n 20, 15 Nov. 1998, 14013-19

[43] Salvetat, Jean-Paul; Briggs, G. Andrew D.; Bonard, Jean-Marc; Bacsa, Revathi R.; Kulik,
Andrzej J.; Stockli, Thomas; Burnham, Nancy A.; Forro, Laszlo. Physical Review Letters, v 82, n
5, Feb 1, 1999, p 944-947

[44] Min-Feng Yu; Files, B.S.; Arepalli, S.; Ruoff, R.S. Physical Review Letters, v 84, n 24, 12
June 2000, 5552-5

[45] Kin-tak Lau; Chong Gu; Hui, D. Source: Composites Part B (Engineering), v 37, n 6, 2006,
425-36

[46]Breuer, O.; Sundararaj, Uttandaraman. Polymer Composites, v 25, n 6, December, 2004, p
630-645

[47]Hernandez-Lopez, J.L.; Alvizo-Paez, E.R.; Moya, S.E.; Ruiz-Garcia, J. Carbon, v 45, n 12,
Oct. 2007, 2448-50

[48] Weijie Huang; Yi Lin; Taylor, S.; Gaillard, J.; Rao, A.M.; Ya-Ping Sun. Nano Letters, v 2, n
3, March 2002, 231-4

[49] Fangming Du; Fischer, J.E.; Winey, K.I. Journal of Polymer Science, Part B (Polymer
Physics), v 41, n 24, 15 Dec. 2003, 3333-8

[50] Zhang, Zhinan; Zhang, Jun; Chen, Peng; Zhang, Baoqing; He, Jiasong; Hu, Guo-Hua.
Carbon, v 44, n 4, April, 2006, p 692-698

[51] Grady, Brian P.; Ha, Mai. AIChE Annual Meeting, Conference Proceedings, 05AIChE: 2005
AIChE Annual Meeting and Fall Showcase, Conference Proceedings, 2005, p4964

55
[52] Pierard, N.; Fonseca, A.; Colomer, J.-F.; Bossuot, C.; Benoit, J.-M.; Van Tendeloo, G.;
Pirard, J.-P.; Nagy, J.B. Carbon, v 42, n 8-9, 2004, p 1691-1697

[53]Xia, Hesheng; Wang, Qi; Li, Kanshe; Hu, Guo-Hua Source: Journal of Applied Polymer
Science, v 93, n 1, Jul 5, 2004, p 378-386

[54] Guang-Xin Chen; Yongjin Li; Shimizu, H. Source: Carbon, v 45, n 12, Oct. 2007, 2334-40

[55] Dror, Yael; Salalha, Wael; Khalfin, Rafail L.; Cohen, Yachin; Yarin, Alexander L.;
Zussman, Eyal. Langmuir, v 19, n 17, Aug 19, 2003, p 7012-7020

[56]Cheol Park; Ounaies, Z.; Watson, K.A.; Crooks, R.E.; Smith, J., Jr.; Lowther, S.E.; Connell,
J.W.; Siochi, E.J.; Harrison, J.S.; St Clair, T.L. Chemical Physics Letters, v 364, n 3-4, 4 Oct.
2002, 303-8

[57] Satishkumar, B.C.; Govindaraj, A.; Mofokeng, J.; Subbanna, G.N.; Rao, C.N.R. Journal of
Physics B: Atomic, Molecular and Optical Physics, v 29, n 21, Nov 14, 1996, p 4925-4934

[58] Kim, Yeji; Minami, Nobutsugu; Zhu, Weihong; Kazaoui, Said; Azumi, Reiko; Matsumoto,
Mutsuyoshi Source: Japanese Journal of Applied Physics, Part 1: Regular Papers and Short Notes
and Review Papers, v 42, n 12, December, 2003, p 7629-7634

[59] Angelikopoulos, Panagiotis; Bock, Henry Source: Journal of Physical Chemistry B, v 112, n
44, Nov 6, 2008, p 13793-13801

[60] Coleman, J.N.; Khan, U.; Blau, W.J.; Gun'ko, Y.K. Carbon, v 44, n 9, Aug. 2006, 1624-52

[61]Sandler, J.K.W.; Kirk, J.E.; Kinloch, I.A.; Shaffer, M.S.P.; Windle, A.H. Polymer, v 44, n 19,
Aug 21, 2003, p 5893-5899

[62]Biercuk, M.J.; Llaguno, M.C.; Radosavijevic, M.; Hyun, J.K.; Johnson, A.T.; Fischer, J.E.
Applied Physics Letters, v 80, n 15, Apr 15, 2002, p 2767

[63]Rubio, A. (Dept. de Fisica Teorica, Valladolid Univ., Spain) Applied Physics A (Materials
Science Processing), v 68, n 3, March 1999, 275-82

[64]Lin, M.F.; Chuu, D.S. Physical Review B (Condensed Matter), v 56, n 8, 15 Aug. 1997,
4996-5002

56
[65] Bonard, J.-M.; Stora, T.; Salvetat, J.-P.; Maier, F.; Stoeckli, T.; Duschl, C.; Forro, L.; de
Heer, W.A.; Chatelain, A. Advanced Materials, v 9, n 10, 8 Aug. 1997, p 827-31

[66] Ismail, A.F.; Goh, Pei Sean; Tee, Jia Chee; Sanip, Suhaila M.; Aziz, Madzlan. Nano, v 3, n
3, 2008, p 127-143

[67]Gajewski, S.; Maneck, H.-E.; Dorfel, I.; Mach, R.; Strauss, B.; Friedrich, J.F. Diamond and
Related Materials, v 12, n 3-7, March/July, 2003, p 816-820

[68] Liu, Jie; Rinzler, Andrew G.; Dai, Hongjie; Hafner, Jason H.; Bradley, R. Kelley; Boul,
Peter J.; Lu, Adrian; Iverson, Terry; Shelimov, Konstantin; Huffman, Chad B.;
Rodriguez-Macias, Fernando; Shon, Young-Seok; Lee, T. Randall; Colbert, Daniel T.; Smalley,
Richard E. Science, v 280, n 5367, May 22, 1998, p 1253-1256

[69] Fang, Hai-Tao; Liu, Chen-Guang; Liu, Chang; Li, Feng; Liu, Min; Cheng, Hui-Ming.
Chemistry of Materials, v 16, n 26, Dec 28, 2004, p 5744-5750

[70] Bandow, Shunji; Rao, A.M.; Williams, K.A.; Thess, A.; Smalley, R.E.; Eklund, P.C. Journal
of Physical Chemistry B, v 101, n 44, Oct 30, 1997, p 8839-8842

[71]S.Qin, D.Qin, W.T .F ord, J.E.Herrera, D.E.Resasco, S.M. ， Bachilo, and R.B.W eisman,
Macromolecules 37, 3965 (2004).

[72] Yudasaka, M.; Zhang, M.; Jabs, C.; Iijima, S. Applied Physics A: Materials Science &
Processing, v 71, n 4, Oct, 2000, p 449-451

[73] Shim HC, Lee HW, Yeom S, Kwak YK, Lee SS, Kim SH. Nanotechnology 2007;18(11),
Article no. 115602.

[74]Lambert JM, Ajayan PM, Bernier P, Planeix JM. Chem Phys Lett 1994;226(3–4):364–71.

[75]Yamamoto K, Akita S, Nakayama Y. Appl Phys 1998;31(8):L34–6.

[76]Tohji K, Takahashi H, Shinoda Y, Shimizu N, Jeyadevan B, Matsuoka I, et al. J Phys Chem
B 1997;101(11):1974–8.

[77]Tohji K, Goto T, Takahashi H, Shinoda Y, Shimizu N, Jeyadevan B, et al. Nature
1996;383(6602):679–9.

57
[78]Liu YQ, Gao L, Sun J, Zheng S, Jiang LQ,Wang Y. Carbon 2007;45(10):1972–8.

[79]Li F, Cheng HM, Xing YT, Tan PH, Su G. Carbon 2000;38(14):2041–5.

[80]Maehashi, K.; Ohno, Y.; Inoue, K.; Matsumoto, K. Applied Physics Letters, v 85, n 6, 9 Aug.
2004, 858-60

[81]Duesberg, G.S.; Muster, J.; Krstic, V.; Burghard, M.; Roth, S. Applied Physics A: Materials
Science & Processing, v 67, n 1, Jul, 1998, p 117-119

[82] Tasis, D.; Papagelis, K.; Douroumis, D.; Smith, J.R.; Bouropoulos, N.; Fatouros, D.G.
Journal of Nanoscience and Nanotechnology, v 8, n 1, Jan. 2008, 420-3

[83] Kikuchi, Y.; Inoue, S.; Nakajima, T. Carbon, v 44, n 7, June 2006, 1287-91

[84] McCarthy, B.; Coleman, J.N.; Czerw, R.; Dalton, A.B.; Byrne, H.J.; Tekleab, D.; Iyer, P.;
Ajayan, P.M.; Blau, W.J.; Carroll, D.L. Nanotechnology, v 12, n 3, September, 2001, 8th
Foresight Confernce on Molecular Nanotechnology, p 187-190

[85] Bandow, Shunji; Rao, A.M.; Williams, K.A.; Thess, A.; Smalley, R.E.; Eklund, P.C. Journal
of Physical Chemistry B, v 101, n 44, Oct 30, 1997, p 8839-8842

[86]Li, Xiaolin; Tu, Xiaomin; Zaric, Sasa; Welsher, Kevin; Won, Seok Seo; Zhao, Wei; Dai,
Hongjie. Journal of the American Chemical Society, v 129, n 51, Dec 26, 2007, p 15770-15771

[87] Duesberg, G.S.; Muster, J.; Byrne, H.J.; Roth, S.; Burghard, M. Applied Physics A, v A69, n
3, Sept. 1999, 269-74

[88] Kis, A.; Csanyi, G.; Salvetat, J.-P.; Lee, Thien-Nga; Couteau, E.; Kulik, A.J.; Benoit, W.;
Brugger, J.; Forro, L. Nature Materials, v 3, n 3, March, 2004, p 153-157

[89]Ki Chul Park; Mahiko, T.; Morimoto, S.; Takeuchi, K.; Endo, M. Applied Surface Science, v
254, n 22, 15 Sept. 2008, 7438-45

[90]Bronikowski, Michael J. Carbon, v 44, n 13, November, 2006, p 2822-2832

[91]Wei Chen, Maria L. Auad, Roberto J. J. Williams, Steven R. Nutt. European Polymer
Journal, 42 (2006) 2765-2772.

58
[92]Salvetat, J.-P.; Bhattacharyya, S.; Byron Pipes, R. Journal of Nanoscience and
Nanotechnology, v 6, n 7, July 2006, 1857-82

[93] Li, Yongjin; Shimizu, Hiroshi. Polymer, v 48, n 8, Apr 5, 2007, p 2203-2207

[94] Obrzut, J.; Douglas, J.F.; Kharchenko, S.B.; Migler, K.B. Physical Review B, v 76, n 19, 15
Nov. 2007, 195420-1-9

[95] Benedict, Brian; Pehrsson, Pehr E.; Zhao, Wei. Journal of Physical Chemistry B, v 109, n 16,
Apr 28, 2005, p 7778-7780

[96] Ahir, S.V.; Huang, Y.Y.; Terentjev, E.M. Polymer, v 49, n 18, Aug 26, 2008, p 3841-3854

[97] MacKernan, D.; Blau, W.J. Europhysics Letters, v 83, n 6, 2008, 66009 (6 pp.)

[98] Sun, Zhenyu; Nicolosi, Valeria; Rickard, David; Bergin, Shane D.; Aherne, Damian;
Coleman, Jonathan N. Journal of Physical Chemistry C, v 112, n 29, Jul 24, 2008, p 10692-10699

[99] Yamamoto, Tatsuhiro; Miyauchi, Yuhei; Motoyanagi, Jin; Fukushima, Takanori; Aida,
Takuzo; Kato, Masaru; Maruyama, Shigeo. Japanese Journal of Applied Physics, Part 1: Regular
Papers and Short Notes and Review Papers, v 47, n 4 PART 1, Apr 18, 2008, p 2000-2004

[100] Matarredona, Olga; Rhoads, Heather; Li, Zhongrui; Harwell, Jeffrey H.; Balzano, Leandro;
Resasco, Daniel E. Journal of Physical Chemistry B, v 107, n 48, Dec 4, 2003, p 13357-13367

[101] Yi, Wenhui; Malkovskiy, Andrey; Chu, Qinghui; Sokolov, Alexei P.; Colon, Marisabel
Lebron; Meador, Michael; Pang, Yi. Journal of Physical Chemistry B, v 112, n 39, Oct 2, 2008, p
12263-12269

[102] Didenko, V.V.; Moore, V.C.; Baskin, D.S.; Smalley, R.E. Nano Letters, v 5, n 8, Aug.
2005, 1563-7

[103] Strano, Michael S. Nature Materials, v 5, n 6, Jun 10, 2006, p 433-434

[104]Doe, Changwoo; Choi, Sung-Min; Kline, Steven R.; Jang, Hyung-Sik; Kim, Tae-Hwan
Advanced Functional Materials, v 18, n 18, Sep 23, 2008, p 2685-2691

59
[105] Dror, Yael; Pyckhout-Hintzen, Wim; Cohen, Yachin. Macromolecules, v 38, n 18, Sep 6,
2005, p 7828-7836

[106]Wei Chen, Hongbin Lu, Steven R. Nutt. Composites Science and Technology, Volume 68,
Issue 12, 2008, P2535-2542

[107] Simmons, J.M.; Nichols, B.M.; Eriksson, M.A. Journal of Physical Chemistry B, v 110, n
14, Apr 13, 2006, p 7113-7118

[108] Young-Jae Lee; Hyun-Ha Kim; Hatori, H. Carbon, v 42, n 5-6, 2004, 1053-6

[109] Banerjee, Sarbajit; Hemraj-Benny, Tirandai; Wong, Stanislaus S. Advanced Materials, v
17, n 1, Jan 18, 2005, p 17-29

[110]Lin, Tong; Bajpai, Vardhan; Ji, Tao; Dai, Liming. Australian Journal of Chemistry, v 56, n
7, 2003, p 635-651

[111] Ivanov, ´I`Achko P. (´I`Achko Pavlov) Polymer composite materials : interface phenomena
& processes. Kluwer Academic Publishers, 2001.

[112] Wong, M.; Paramsothy, M.; Xu, X.J.; Ren, J.; Li, S.; Liao, K. Polymer, v 44, n 25, Dec.
2003, 7757-64

[113]Gorga, R.E.; Lau, K.K.S.; Gleason, K.K.; Cohen, R.E. Journal of Applied Polymer Science,
v 102, n 2, 15 Oct. 2006, 1413-18

[114]Fisher, F.T.; Eitan, A.; Andrews, R.; Brinson, L.C.; Schadler, L.S. Composites Science and
Technology, v 66, n 9, July 2006, 1159-70

[115]Assouline, E.; Lustiger, A.; Barber, A.H.; Cooper, C.A.; Klein, E.; Wachtel, E.; Wagner,
H.D. Journal of Polymer Science, Part B: Polymer Physics, v 41, n 5, Mar 1, 2003, p 520-527

[116]Coleman, J.N.; Cadek, M.; Dalton, A.B.; Munoz, E.; Razal, J.; Baughman, R.H.; Blau, W.J.
Proceedings of the SPIE - The International Society for Optical Engineering, v 5118, 2003, 271-9

[117] Bifeng Pan; Daxiang Cui; Feng Gao; Rong He Source: Nanotechnology, v 17, n 10, 28
May 2006, 2483-9

60
[118] Barber, A.H.; Cohen, S.R.; Kenig, S.; Daniel Wagner, H. Composites Science and
Technology, v 64, n 15, 2004, 2283-9

[119] Ding, W.; Eitan, A.; Fisher, F.T.; Chen, X.; Dikin, D.A.; Andrews, R.; Brinson, L.C.;
Schadler, L.S.; Ruoff, R.S. Nano Letters, v 3, n 11, Nov. 2003, 1593-7

[120] Barber, A.H.; Cohen, S.R.; Wagner, H.D. Applied Physics Letters, v 82, n 23, 9 June 2003,
p 4140-2

[121] Frankland, S.J.V.; Caglar, A.; Brenner, D.W.; Griebel, M. Journal of Physical Chemistry B,
v 106, n 12, Mar 28, 2002, p 3046-3048

[122] Lordi, V.; Nan Yao. Journal of Materials Research, v 15, n 12, Dec. 2000, 2770-9

[123] Meguid, S.A.; Sun, Y. Source: Materials & Design, v 25, n 4, June 2004, 289-96

[124] Vasiliev, Valery V. Publisher: Elsevier, Pub date: 2007.

[125]Raravikar, Nachiket R.; Schadler, Linda S.; Vijayaraghavan, Aravind; Zhao, Yiping; Wei,
Bingqing; Ajayan, Pulickel M. Chemistry of Materials, v 17, n 5, Mar 8, 2005, p 974-983

[126]Kilpatrick, S.J.; Cao, A.; Li, X.; Renna, N.J.; Ajayan, P.M. 2005 International
Semiconductor Device Research Symposium, 2005, 2 pp.

[127] Mei Zhang; Shaoli Fang; Zakhidov, A.A.; Lee, S.B.; Aliev, A.E.; Williams, C.D.;
Atkinson, K.R.; Baughman, R.H. Science, v 309, n 5738, 19 Aug. 2005, 1215-19

[128]Lanticse, Leslie Joy; Tanabe, Yasuhiro; Matsui, Keitaro; Kaburagi, Yutaka; Suda, Katsumi;
Hoteida, Masayuki; Endo, Morinobu; Yasuda, Eiichi. Carbon, v 44, n 14, November, 2006, p
3078-3086

[129] Yoong Ahm Kim; Hayashi, T.; Endo, M.; Gotoh, Y.; Wada, N.; Seiyama, J. Scripta
Materialia, v 54, n 1, Jan. 2006, 31-5

[130] Park, Cheol; Wilkinson, John; Banda, Sumanth; Ounaies, Zoubeida; Wise, Kristopher E.;
Sauti, Godfrey; Lillehei, Peter T.; Harrison, Joycelyn S. Journal of Polymer Science, Part B:
Polymer Physics, v 44, n 12, Jun 15, 2006, p 1751-1762

61
[131] Martin, C.A.; Sandler, J.K.W.; Windle, A.H.; Schwarz, M.-K.; Bauhofer, W.; Schulte, K.;
Shaffer, M.S.P. Polymer, v 46, n 3, Jan 26, 2005, p 877-886

[132] Kimura, Tohru; Ago, Hiroki; Tobita, Masayuki; Ohshima, Satoshi; Kyotani, Mutsumasa;
Yumura, Motoo Source: Advanced Materials, v 14, n 19, Oct 2, 2002, p 1380-1383

[133] Camponeschi, Erin; Vance, Richard; Al-Haik, Marwan; Garmestani, Hamid; Tannenbaum,
Rina. Carbon, v 45, n 10, September, 2007, p 2037-2046

[134] Xu-Ming Xie; Yi-Tao Liu; Yan-Fang Gao; Qing-Ping Feng; Lin-Rui Guo; Xiao-Hao
Wang; Xiong-Ying Ye. Carbon, v 44, n 3, March 2006, 599-602

[135]Qinghua Zhang; Zhenjun Chang; Meifang Zhu; Xiumei Mo; Dajun Chen. Nanotechnology,
v 18, n 11, 21 March 2007, 6 pp.

[136] Bower, C.; Rosen, R.; Jin, L.; Han, J.; Zhou, O. Applied Physics Letters, v 74, n 22, 1999, p
3317-3319

[137] Bliznyuk, V.N.; Singamaneni, S.; Sanford, R.L.; Chiappetta, D.; Crooker, B.; Shibaev, P.V.
Polymer, v 47, n 11, May 17, 2006, p 3915-3921

[138] Ajayan, P.M.; Stephan, O.; Colliex, C.; Trauth, D. Science, v 265, n 5176, 26 Aug. 1994,
1212-14

[139]Strobl, C.J.; Schaflein, C.; Beierlein, U.; Ebbecke, J.; Wixforth, A. Applied Physics Letters,
v 85, n 8, Aug 23, 2004, p 1427-1429

[140] Pegel, S.; Potschke, P.; Petzold, G.; Alig, I.; Dudkin, S.M.; Lellinger, D. Polymer, v 49, n
4, 18 Feb. 2008, 974-84

[141] Haggenmueller, R.; Gommans, H.H.; Rinzler, A.G.; Fischer, J.E.; Winey, K.I. Chemical
Physics Letters, v 330, n 3-4, 10 Nov. 2000, 219-25

[142] Coleman, J.N.; Khan, U.; Blau, W.J.; Gun'ko, Y.K. Source: Carbon, v 44, n 9, Aug. 2006,
1624-52

[143]Lourie, O.; Wagner, H.D. Composites Science and Technology, v 59, n 6, May, 1999, p
975-977

62
[144] Cooper, C.A.; Young, R.J.; Halsall, M. Source: Composites Part A, v 32A, n 3-4, 2001,
401-11

[145] Barber, Asa H.; Cohen, Sidney R.; Kenig, Shmuel; Wagner, H. Daniel. Composites Science
and Technology, v 64, n 15 SPEC. ISS., November, 2004, p 2283-2289

[146]Cooper, Carole A.; Cohen, Sidney R.; Barber, Asa H.; Wagner, H. Daniel. Applied Physics
Letters, v 81, n 20, Nov 11, 2002, p 3873-3875

[147] Wagner, H.D.; Lourie, O.; Feldman, Y.; Tenne, R. Source: Applied Physics Letters, v 72, n
2, 12 Jan. 1998, 188-90

[148]S. J. V. Frankland, A. Caglar, D. W. Brenner, and M. Griebel, J. Phys. Chem. B 106, 3046
~2002.

[149]Yang, Mingjun; Koutsos, Vasileios; Zaiser, Michael Source: Journal of Physical Chemistry
B, v 109, n 20, May 26, 2005, p 10009-10014

[150] Frankland, S.J.V.; Caglar, A.; Brenner, D.W.; Griebel, M. Journal of Physical Chemistry B,
v 106, n 12, 28 March 2002, 3046-8

[151] Wagner, H.D.; Lourie, O.; Feldman, Y.; Tenne, R. Applied Physics Letters, v 72, n 2, 12
Jan. 1998, 188-90

[152]Gao, X.-L.; Li, K. International Journal of Solids and Structures, v 42, n 5-6, March, 2005, p
1649-1667

[153]Namilae, S.; Chandra, N. Source: Transactions of the ASME. Journal of Engineering
Materials and Technology, v 127, n 2, April 2005, 222-32

[154]Tsu-Wei Chou; Chunyu Li Source: Composites Science and Technology, v 66, n 14, Nov.
2006, 2409-14

[155]Kin Liao; Sean Li Source: Applied Physics Letters, v 79, n 25, 17 Dec. 2001, 4225-7

[156] Lordi, V.; Nan Yao. Journal of Materials Research, v 15, n 12, Dec. 2000, 2770-9

63
[157] Wall, A.; Coleman, J.N.; Ferreira, M.S. Physical Review B, v 71, n 12, 15 March 2005,
125421-1-5

[158] Daniel, Isaac M. Engineering mechanics of composite materials, Oxford University Press,
2006

[159] Rakow, J.F.; Waas, A.M. Mechanics of Advanced Materials and Structures, v 11, n 2,
March-April 2004, 151-73

[160] Lee, C.; Hwang, W. Source: Journal of Materials Science Letters, v 17, n 19, 1 Oct. 1998,
1601-3

[161] Sharma, Rajesh; Ranta, Sohan; Ranjan, Ashish. Journal of Reinforced Plastics and
Composites, v 24, n 7, 2005, p 719-724

[162]Hashin, Z.; Shtrikman, S. Journal of Mechanics and Physics of Solids, v 11, n 2, Mar-Apr,
1963, p 127-140

[163]Walpole, L.J. Math. Proc. of the Cambridge Phil. Soc., v 81, pt.2, 1977, 283-9

[164]Campbell, Malcolm D.; Willis, Donald. Industrial & Engineering Chemistry, Product
Research and Development, v 19, n 4, Dec, 1980, p 596-601

[165] McCullough (R. L.); Univ of Delaware, Cent for Composite Materials, Newark, DE, USA
Composites Science and Technology, v 22, n 1, 1985, p 3-21

[166] Halpin, J. C.; Kardos, J. L. Polymer Engineering and Science, v 16, n 5, May, 1976, p
344-352

[167]Eshelby J.D. Proceedings of the Royal Society of London A 241 (1957), p 376

[168] Hill, R. Journal of Mechanics and Physics of Solids, v 13, n 4, Aug, 1965, p 213-222

[169]Tanaka, K.; Mori, T. Philosophical Magazine, v 23, n 183, March 1971, 737-40

[170]Tucker, C.L., III; Liang, E. Composites Science and Technology, v 59, n 5, 1999, 655-71

[171]Hashin Z, Rosen B. W. Journal of Applied Mechanics,1964. 31, 223-232

64
[172]Christensen, R.M.; Lo, K.H. Journal of the Mechanics and Physics of Solids, v 27, n 4, Aug.
1979, 315-30

[173] Lagoudas, D.C.; Seidel, G.D. Mechanics of Materials, v 38, n 8-10, Aug.-Oct. 2006,
884-907

[174] Odegard, Gregory M.; Gates, Thomas S.; Nicholson, Lee M.; Wise, Kristopher E.
Composites Science and Technology, v 62, n 14, November, 2002, p 1869-1880

[175] Odegard, G.M.; Gates, T.S.; Wise, K.E.; Park, C.; Siochi, E.J. Composites Science and
Technology, v 63, n 11, August, 2003, p 1671-1687

[176] Fisher, F.T.; Bradshaw, R.D.; Brinson, L.C. Composites Science and Technology, v 63, n
11, 2003, 1689-703

[177] Bradshaw, R.D.; Fisher, F.T.; Brinson, L.C. Composites Science and Technology, v 63, n
11, 2003, 1705-22

[178] Hadjiev, V.G.; Lagoudas, D.C.; Oh, E.-S.; Thakre, P.; Davis, D.; Files, B.S.; Yowell, L.;
Arepalli, S.; Bahr, J.L.; Tour, J.M. Composites Science and Technology, v 66, n 1, Jan. 2006,
128-36

[179]Cooper, Carole A.; Cohen, Sidney R.; Barber, Asa H.; Wagner, H. Daniel. Applied Physics
Letters, v 81, n 20, Nov 11, 2002, p 3873-3875

[180] Lurie, S.; Belov, P.; Volkov-Bogorodsky, D.; Tuchkova, N. Journal of Materials Science, v
41, n 20, Oct. 2006, 6693-707

[181]Hammerand, D.; Seidel, G.D.; Lagoudas, D.C. Mechanics of Advanced Materials and
Structures, v 14, n 4, May 2007, 55-72

65
Chapter 3. Surface assisted dispersion and interfacial stress transfer
3.1. Motivation
The synthesis of carbon nanotube (CNT)-epoxy nanocomposites with enhanced
thermal and mechanical properties requires generating uniform dispersion of the nanotubes and
strong interfacial bonding for load transfer.[1] This can be accomplished by chemical
modification of the surface of the CNTs. For example, when the sidewalls of CNTs were
functionalized via free-radical polymerization of methyl methacrylate, an 11 % increase in
strength and a 21 % increase in modulus were observed for an epoxy composite with 1 wt %
CNTs.[2] In contrast, addition of untreated CNTs produced a sharp decrease in both properties. In
another study, single-walled carbon nanotubes (SWNTs) were functionalized using a combination
of open-end oxidation and sidewall fluorination.[3] Epoxy composites containing 1 wt % of these
functionalized nanotubes exhibited a 30 % increase in modulus and 18 % increase in tensile
strength compared to the unfilled epoxy. Similarly, the use of 0.3 wt % of fluorinated single-wall
carbon nanotubes to modify an epoxy-anhydride matrix increased the storage modulus by 20 %
without sacrificing impact strength;[4] On the other hand, the stiffness of the cured epoxy system
exerted a strong effect on the properties of the resulting composites. Drops in reinforcing
properties were observed in epoxy systems with increased stiffness.[5] The interaction between
the polymer chain and the functionalized carbon nanotubes as well as their relative motion
dominates the final strengthening effect of these nano reinforcements.

66
In order to overcome the above mentioned drawbacks for the carbon nanotube - stiff
epoxy systems, grafting a chemical structure to the surface of the multi-walled nanotubes
(MWCNTs) that is identical to that of the epoxy precursor in which the final dispersion will be
prepared is an ideal choice, through which a uniform dispersions of functionalized MWCNTs and
a covalently bonded interface between the matrix and the reinforcement in the epoxy network
could be attained.[6] Therefore, an appropriate surface treatment of MWCNTs for use in epoxy
composites is to attach an epoxy monomer to a previously acid-functionalized nanotube.[7] The
reaction between carboxylic and epoxy groups is catalyzed by different bases, such as tertiary
amines, KOH, and triphenylphosphine.[8,9] Using KOH as a catalyst, a diepoxide monomer
based on diglycidylether of bisphenol A (DGEBA) can be covalently bonded to previously
COOH-functionalized MWCNTs according to Eitan et al..[7] However, they reported that no free
epoxy groups of DGEBA were retained at the surface of the MWCNTs (opening of both epoxy
rings in these DGEBA molecules). Thus, the problem of attaining covalent bonding of
DGEBA-functionalized MWCNTs with the epoxy matrix remains unsolved.
The objective of this chapter is to show that the approach of Eitan et al.[7]for
functionalizing MWCNTs can be coupled with a suitable chemistry for the synthesis of the epoxy
network. Secondly, this combined approach would achieve both uniform dispersions of
MWCNTs and the formation of covalent bonds on the interface. The incorporation of these

67
MWCNTs would lead to improvements for the mechanical properties of a stiff epoxy system, the
brittleness of which has always been a limiting issue for the application of this class of materials.
In this work, multiwalled carbon nanotubes (MWCNTs) were functionalized in a
two-step process. First, carboxyl groups were introduced by a standard oxidation procedure.
Subsequently, these groups were reacted with a monoepoxide (phenyl glycidyl ether, PGE), or
with a diepoxide based on diglycidyl ether of bisphenol A (DGEBA), using triplenylphosphine
(TPP) as catalyst, yielding β-hydroxyester groups attached to the surface of MWCNTs. Thus,
three different types of functionalized MWCNTs were obtained: COOH-MWCNTs (after the first
step), and PGE-MWCNTs and DGEBA-MWCNTs (after the second step). The anionic
homopolymerization of an epoxy monomer initiated by a tertiary amine, 4-dimethylamino
pyridine (DMAP), provides the possibility of covalently bonding COOH-MWCNTs,
PGE-MWCNTs and DGEBA-MWCNTs to the epoxy matrix. The primary reactions taking place
in a medium containing an epoxy monomer (DGEBA), a tertiary amine and species with COOH
and OH groups are indicated in Figure 1.[10] The epoxy-acid reaction leads to formation of
β-hydroxyester bonds. Chain transfer reactions involving a propagating alkoxide anion and the
secondary hydroxyls of the β-hydroxyester groups provide a covalent coupling of
PGE-MWCNTs and DGEBA-MWCNTs with the epoxy matrix (reactions 4 and 5 in Figure 1).

68
3.2. Experimental
3.2.1. Materials
MWCNTs were obtained from a commercial source (Shenzhen Nanotech Port Co. Ltd.,
China). The MWCNTs were produced by chemical vapour deposition (CVD) and contained ~5%
impurities, consisting primarily of amorphous carbon and transition metals, such as lanthanum
and nickel. The nanotubes were 5-15 µm long and 40-60 nm in diameter (Figure 2). Two epoxy
monomers were selected for this study. The monoepoxide was phenyl glycidyl ether (PGE,
Aldrich) and the diepoxide was a commercial monomer based on diglycidylether of bisphenol A
(DGEBA, EPON 828, Miller-Stephenson). The initiator selected for DGEBA
homopolymerization was 4-dimethylamino pyridine (DMAP, Sigma-Aldrich).
Triphenylphosphine (TPP, Aldrich) was employed as a catalyst of the esterification reaction to
attach PGE and DGEBA to the surface of acidified-MWCNTs. The reactions are illustrated in
Figure 3. Anhydrous dimethylformamide (DMF), tetrahydrofuran (THF), and anhydrous
ethanol (VWR Scientific) were used as solvents.
3.2.2. Preparation of COOH-MWCNT
MWCNTs (1 g) were dispersed in 250 ml of concentrated 3:1 H
2
SO
4
/70% HNO
3

following the procedure developed by Liu et al.[11] One-hour sonication and 3-hour stirring at
room temperature was optimal to achieve moderate oxidation and maintain a high aspect ratio.
After the acid treatment and exhaustive washes with deionized water, HCl was added to the acid

69

C HCH
2
O
R
R'COOH +
O
OH
RCH CH
2
OCR' 1)

2
O
C
O
O
RCH CH
2
OCR'
R'
OH
CH R CH
2
OH
+
O
OH
RCH CH
2
OC R' 2)

C H CH
2
O
R
+ R
3
N
O
R N R
3
CH
2
CH
+
-
3)

O
R N R
3
CH
2
CH
+
-
OH R'' +
+
OH
R N R
3
CH
2
CH O R''
-
+ 4)

O R''
-
C HCH
2
O
R
+
R
O O R'' CH
2
CH
-
5)

C HCH
2
O
R +
n-1
R
CH O CH
2
R
O O R'' CH
2
CH
-
n-2
R
CH O CH
2
R
O O R'' CH
2
CH
-6)

Figure 3.1 Main reactions taking place during the anionic polymerization of an epoxy monomer
initiated by a tertiary amine (R’COOH and R’’OH represent species present in the reaction
medium).

70

Figure 3.2 TEM image of acidified MWCNTs.

71

Figure 3.3 Scheme of reactions used to functionalize MWCNTs.

72
mixture to convert the carboxylates into carboxylic acid groups on the defect sites of
MWCNTs.[12]The solution was extensively washed again with deionized water until pH value
reached 5 - 6. The acid-treated nanotubes (denoted COOH-MWCNTs) were collected on a 0.45
µm PTFE membrane by vacuum filtration and dried overnight in a vacuum oven at 90 ºC.
3.2.3. Preparation of PGE-MWCNT and DGEBA-MWCNT
The esterification of COOH groups with PGE or DGEBA was carried out in a DMF
solution, using TPP as a catalyst (0.1 mole TPP per mole of epoxy groups). The reaction was
performed by refluxing at 150 ºC under nitrogen for 36 h. (The model reaction of the
esterification of benzoic acid by PGE suggested that complete conversion could be attained under
these conditions.) After the treatment, PGE-MWCNTs and DGEBA-MWCNTs were thoroughly
washed with ethanol, collected with the PTFE membrane, and dried in a vacuum oven for 24
hours.
3.2.4. Preparation of the MWCNT- Epoxy Composites
Unmodified MWCNTs, COOH-MWCNTs, PGE-MWCNTs, and DGEBA-MWCNTs
were dispersed in THF using bath sonication for 5 min. DGEBA was dissolved in THF (1:1 by
volume) in a dual axis high-speed mixer (Keyence HY501). Next, both mixtures were blended in
the mixer and then bath-sonicated for another 5 minutes. Finally, THF was evaporated in a
vacuum chamber at 100 ºC overnight. DMAP was added to the mixture in a molar ratio of 0.08
moles DMAP per mole of epoxy groups. Further mixing (10 minutes) and degasification (10

73
minutes) were performed with the high-speed mixer. The final blend was cast into an aluminium
mold pretreated with a release agent. The following curing cycle reported in the literature for this
particular initiator was employed:[13] 3 h at 80 ºC, 3 h at 120 ºC, and 30 min at 160 ºC.
Nanocomposites contained 0.5, 1 and 3 wt % MWCNTs (for modified-MWCNTs, the weight %
refers to that of the neat MWCNT devoid of the organic part). The neat epoxy samples were also
prepared following the same procedure for comparison purposes.
3.2.5. Characterization
Thermal gravimetry analysis (TGA 2050, TA Instruments) was employed to quantify
the wt % of organic part in functionalized-MWCNTs. Samples were heated to 800 ºC at 10
ºC/min under nitrogen. Transmission mode Fourier transform infrared spectroscopy (FTIR,
Nicolet 4700) was employed to assess the presence of the organic groups in
functionalized-MWCNTs. Transmission electron microscopy (TEM, Philips EM420, 120 KV),
and scanning electron microscopy (SEM, Cambridge 360) were used to assess the quality of the
dispersion in epoxy composites. Fracture surfaces of specimens pre-chilled by liquid nitrogen
were sputter-coated with gold prior to SEM observation. Dynamic mechanical analysis (DMA
2980, TA Instruments) was carried out to determine the glass transition temperature (Tg) of the
composites, defined as the onset temperature where the storage modulus exhibited a sharp
decrease. Measurements were performed at a heating rate of 10 ºC /min and a load frequency of 1
Hz. Flexural tests were performed on a universal testing machine using a four-point fixture

74
according to the ASTM standard D790M, to obtain elastic modulus and flexural strength.
Beam-shaped specimens were cut and polished for each type of sample for flexural and dynamic
mechanical tests. Five specimens were prepared for each test. TGA (Thermogravimetric
Analyses) provide a measurement of the mass fraction of organic modifier that can be attached to
the MWCNTs (Figure 4).
3.3. Results and Discussion
3.3.1. Properties of functionalized MWCNT
The solubility of neat and functionalized MWCNTs in water and DGEBA is shown in
Figure 5. COOH-MWCNTs exhibit a good dispersibility in water but not in DGEBA. In the latter
case, COOH-MWCNTs agglomerate and settle to the bottom. The functionalization of MWCNTs
with both PGE and DGEBA led to stable and uniform dispersions in epoxy, as expected.
The observed mass loss of the unmodified MWCNTs from TGA is assigned to the
presence of small amounts of amorphous carbon and impurities. Discounting the mass loss of
carbon nanotubes, the mass fraction of organic groups eliminated at 600 ºC was close to 7 wt %
for COOH-MWCNTs, 16 wt % for DGEBA-MWCNTs, and 28 wt % for PGE-MWCNTs. These
values, though approximate, indicate that: a) the two-step surface modification process was
efficient and led to a significant mass fraction of organic groups chemically bonded to MWCNTs;
b) the mass fraction of PGE bonded to COOH groups was significantly larger than that of
DGEBA. Additional insight is provided by comparing experimental mass fractions with those

75
expected from a complete reaction of COOH with epoxy groups. In the case of the reaction
between COOH groups and PGE, the ca. 7 g of COOH attached to 93 g of the neat MWCNT can
incorporate 23.3 g PGE, leading to a theoretical organic mass fraction equal to 25 % at full
conversion, close to the experimental value (28 wt %). This means that esterification of COOH
groups by PGE was carried out to complete conversion (an amount of PGE greater than the
experimental value can be explained by oligomerization of PGE by chain transfer reactions
involving the β-hydroxyester group. However, the epoxy-acid reaction proceeds at a much
faster rate and consequently is complete before the homopolymerization of epoxy groups).[10] In
the case of COOH groups reacting with DGEBA, the theoretical mass fraction of DGEBA
attached to the surface depends on whether one or both epoxy groups participate. The results of
Eitan et al.[7] suggest that both epoxy groups participate in these reactions. Based on this
assertion, 7 g of COOH groups attached to 93 g of MWCNTs should incorporate 29.2 g of
DGEBA. (The equivalent weight per epoxy groups of Epon 828 was 188 g/eq.) This leads to a
theoretical organic mass fraction equal to 28%, which is significantly higher than the
experimental value (16 %). Possibly, steric restrictions do not allow a complete reaction of
DGEBA with COOH groups.
FTIR spectra also corroborate the attachment of PGE and DGEBA to the surface of
MWCNTs, as shown in Figure 6. Peaks located at 2920 cm
-1
, 2850 cm
-1
, 1460 cm
-1
, and 1375
cm
-1
are characteristic C-H stretching and deformation frequencies. The band at 1250 cm
-1
is a

76
characteristic vibration of aryl ethers while the one at 1710 cm
-1
is characteristic of the carbonyl
group (no indication of the acid to ester conversion could be obtained from this peak). Bands in
the region 1100 – 1300 cm
-1
are characteristic of C-O stretching vibrations in esters.[14] To
assess the presence of residual epoxy rings in the DGEBA-MWCNTs requires selection of a
characteristic band that is not overlapped with other bands. A band that is
typical of epoxy rings is located at 917 cm
-1
.[7] The absence of this band in the FTIR spectrum
indicates that most of the epoxy groups of DGEBA were consumed in the reaction, a conclusion
similar to results reported by Eitan et al.[7]
3.3.2. Formation of covalent bonds at interface
The use of 4-dimethylamino pyridine (DMAP) as an initiator for DGEBA
homopolymerization produced covalent bonds of PGE-MWCNTs and DGEBA-MWCNTs with
the epoxy matrix through chain transfer reactions involving the secondary hydroxyls. While, in
situ conversion of COOH-MWCNTs into DGEBA-MWCNTs during cure did not produce
uniform dispersions of nanotubes and was not effective. In the presence of a tertiary amine,
epoxy-acid reactions are much faster than epoxy homopolymerization.[10] However, this
reaction, which produced an in situ transformation of COOH-MWCNTs into DGEBA-MWCNTs,
did not lead to uniform dispersion. In the absence of an adequate mixing stage, converting
agglomerates of MWCNTs to a uniform dispersion should require extremely long times. As
polymerization proceeds and the matrix viscosity increases, migration of nanotubes will be

77

Figure 3.4 TGA thermograms of neat and functionalized MWCNTs.

0 100 200 300 400 500 600
50
55
60
65
70
75
80
85
90
95
100
Weight (%)
Temperature (
o
C)
MWNT
COOH-MWNT
DGEBA-MWNT
PGE-MWNT

78

Figure 3.5 Solubility of neat and functionalized MWCNTs in: a) water and b) DGEBA (Epon
828).

CNT COOH-CNT PGE-CNT DGEBA-CNT
CNT COOH-CNT PGE-CNT DGEBA-CNT

79

Figure 3.6 FTIR spectra of neat and functionalized MWCNTs (the peak at 2350 cm
-1
was caused
by adventitious CO
2
in the sample chamber).

80
increasingly inhibited and ultimately arrested by gelation. Therefore, in situ functionalization
does not appear to be a suitable choice to obtain uniform dispersions of CNTs in epoxy.
3.3.3. Microscopic morphology of the composites
All of the MWCNT-epoxy composites were glassy at room temperature and exhibited a
glass transition temperature of around 130 ºC according to dynamic mechanical analysis. Fracture
surfaces of the cured epoxy and of MWCNTs-epoxy composites with 1 wt % MWCNTs obtained
under cryogenic conditions, were examined by SEM (Figure 7). Composites prepared with
unmodified MWCNTs and with COOH-MWCNTs exhibited isolated regions of agglomerated
MWCNTs and other regions devoid of nanotubes (an interphase between both regions is shown in
Figure 7c). Both PGE-MWCNTs and DGEBA-MWCNTs were uniformly dispersed in the epoxy
matrix as was also demonstrated previously by their respective solubility in the epoxy. The
SEM image, Figure 7d, shows MWCNTs functionalized with DGEBA. For the particular
concentration used to obtain this series of composites, the dispersed phase appears distant from
the percolation threshold.
3.3.4. Elastic and flexural properties
The two-step surface modification and the subsequent anionic homopolymerization
yielded uniform MWCNTs-epoxy composites with significant enhancement in flexural strength
and a modest increase in the elastic modulus when compared to the neat epoxy. The
monofunctional epoxide (PGE) was more efficient than the diepoxide (DGEBA) for

81

Figure 3.7 SEM micrographs of fracture surfaces obtained under cryogenic conditions (1wt%
CNT): a) neat epoxy, b) unmodified MWCNTs-epoxy composite, c) COOH-MWCNTs-epoxy
composite, d) DGEBA-MWCNTs-epoxy composites.Enlarge photos to extend to page margins.
No sense in wasting real estate.

(a) (b)
(c) (d)
2µm
2µm
2µm 2µm

82
functionalizing the nanotubes and led to higher values of flexural strength at the same loading of
MWCNTs. The measured elastic moduli (GPa) and flexural strength values (MPa) for the neat
epoxy and for different MWCNTs-epoxy composites are shown in Table 1. A non-uniform
dispersion of nanotubes, as in the case of unmodified MWCNTs and COOH-MWCNTs, led to a
notable decrease in elastic modulus and flexural strength of the composites compared to values
for the neat epoxy. Composites exhibiting uniform dispersions of MWCNTs showed significant
increases in flexural strength and modest increases in elastic modulus. Compared with other more
flexible epoxy systems, [2-5] in which significant increases in both the flexural strength and the
elastic modulus were observed, the relative small improvement for the elastic modulus in this
study could be explained by the difficulty of the relative motion between MWCNTs and the
polymer chains/chain sectors in this rigid epoxy system at ambient temperature. At the same wt %
of reinforcement, PGE-MWCNTs exhibited a larger increase in flexural strength than
DGEBA-MWCNTs. This can be attributed to the higher efficiency of surface functionalization in
the case of PGE-MWCNT that leads to more OH groups available to form covalent bonds with
the epoxy matrix.
3.4. Conclusions
A two-step acid-epoxy functionalization of MWCNTs coupled with anionic
homopolymerization chemistry to build up the epoxy network was effective in achieving uniform
dispersions of MWCNTs in a stiff epoxy system. The resulting composites exhibited significant

83

Table 3.1 Elastic Modulus and Flexural Strength of the Neat Epoxy and MWCNTs-Epoxy
Composites.
Material
MWCNTs (wt
%)
Elastic Modulus
(MPa)
Flexural Strength
(MPa)
Neat Epoxy 0 3.48 ± 0.30 51 ± 15
0.5 3.06 ± 0.81 38 ± 7
1 3.28 ± 0.52 39 ± 11
Unmodified
MWCNTs-Epoxy
3 3.15 ± 0.34 33 ± 13
0.5 3.11 ± 0.24 50 ± 10
1 3.10 ± 0.13 42 ± 7
COOH-MWCNTs-Epoxy

3 3.06 ± 0.20 26 ± 3
0.5 3.48 ± 0.32 73 ± 12
PGE-MWCNTs-Epoxy
1 3.58 ± 0.29 83 ± 19
0.5 3.52 ± 0.52 55 ± 7
1 3.59 ± 0.41 58 ± 3 DGEBA-MWCNTs-Epoxy
3 4.02 ± 0.57 110 ± 5

84
enhancement in flexural strength without any trade-offs in elastic modulus compared with the
neat epoxy.
The findings underscore the significance of introducing suitable surface compatibility
and reactivity with matrix system onto MWCNTs to construct the nanocomposite network.
Further improvements in mechanical and thermal performance are likely to be possible with
optimized surface treatments tailored to specific polymer matrices. The addition of small
amounts of CNTs affords myriad opportunities to design polymers with a broader range of
enhanced properties for applications as adhesives, components, and composite matrices.

85
Chapter 3 References
[1] (a) Sun, Y.-P.; Fu, K.; Lin, Y.; Huang, W. Acc. Chem. Res. 2002, 35, 1096. (b) Niyogi, S.;
Hamon, M. A.; Itkis, M. E.; Haddon, R. C. Acc.Chem. Res. 2002, 35, 1105. (c) Bahr, J. L.; Tour,
J. M. J. Mater. Chem. 2002, 12, 1952. (d) Hirsch, A. Angew. Chem., Int. Ed. 2002, 41, 1853.

[2] Tiano, T.; Roylance, M.; Gassner, J. 32nd SAMPE Conf. 2000, 192.

[3] Shu, J.; Kim, J. D.; Peng H.; Margrave, J. L.; Khabashesku, V. N.; Barrera, E. V. Nano Lett.
2003, 3, 1107.

[4] Miyagawa, H.; Drzal, L. T. Polymer 2004, 45, 5163.

[5] (a) Ci, L. ;Bai, J. Composites Scince and Technology 2005 Article In Press, Available online
11 July 2005. (b)Liu, L.; Wagner, H.D. Composites Science and Technology 2005, 65,
1861-1868

[6] Lin, Y.; Zhou, B.; Shiral Fernando, K.A.; Liu, P.; Allard, L. F.; Sun, Y. P. Macromolecules
2003, 36, 7199-7204

[7] Eitan, A.; Jiang, K.; Dukes, D.; Andrews, R.; Schadler, L. S. Chem. Mater. 2003,15, 3198.

[8]Schechter, L.; Wynstra, J. Ind. Eng. Chem. 1956, 48, 86.

[9] Bartlet, P.; Pascault, J. P.; Sautereau, H. J. Appl. Polym. Sci. 1985, 30, 2955.

[10] Hoppe, C. E.; Galante, M. J.; Oyanguren, P. A.; Williams, R. J. J. Macromol. Mater. Eng.
2005, 290, 456.

[11] Liu, J.; Rinzler, A. G.; Colbert, D. T.; Smalley, R. E. Science 1998, 280, 1253.

[12] Chen, J.; Hamon, M. A.; Hu, H.; Chen, Y.; Rao, A. M.; Eklund, P. C.; Haddon, R. C.
Science 1998, 282, 95.

[13] dell’Erba, I. E.; Fasce, D. P.; Williams, R. J. J.; Erra-Balsells, R.; Fukuyama, Y.; Nonami, H.
Macromol. Mater. Eng. 2004, 289, 315.

[14] Bellamy, L. J. The Infra-red Spectra of Complex Molecules; John Wiley: London,1966.

86
Chapter 4. Towards interface control
4.1. Motivation
Nanocomposites that feature inorganic nano-fibrils incorporated into a synthetic
polymer offer the promise of ultra-stiff, ultra-tough, low-density multifunctional materials [1].
Among nano-reinforcements, carbon nanotubes (CNTs) stand out because of the low density,
large aspect ratio, and ultra-strong/stiff tube-like structures [2-4]. CNTs have received widespread
attention as reinforcements for polymer nanocomposites, especially for versatile matrix systems
such as epoxy [5, 6].
Despite the promising developments for CNT-epoxy nanocomposites, the inherent
processing difficulties in CNT-epoxy composites, such as dispersion and interphase control, have
caused mechanical performance to fall short of theoretical predictions [7-9]. An effective CNT
surface grafting scheme should address both issues - dispersion and interface stress transfer
[10-15]. “Grafting to” and “grafting from” methods have been employed to introduce a variety
of molecules onto CNTs to modify surface chemistry [16-19].
Recently, epoxy nanocomposites systems based on amine-functionalized CNTs were
investigated to exploit the reactivity of amine-functionalities with epoxies and consequent
formation of cross-linkages at the interface, which facilitate stress transfer. For example, Wang et
al. [20] reported a ~15 °C increase in glass transition temperature and a nearly 100% increase in
impact strength with the inclusion of only ~0.5wt% amine-functionalized MWCNTs. In other

87
work, Shen et al. [21] reported marked increases in the onset decomposition temperature and
asserted that the observed T
g
shift arose from the rigidity of polymer grafts. Ye et al. also reported
a decrease in T
g
combined with significant boosts in composite moduli. [22].
In some of the studies cited above, attempts were made to correlate the MWCNT
surface chemistry with the bulk properties of the nanocomposites through the notion of an
interphase region, yet no direct microscopic observations and limited analyses of
thermomechanical evidence were offered to support the speculations. In addition, little attention
was given to the match of solubility between the surface grafts and the epoxy systems. In this
work, epoxy-compatible amine terminated linear polypropylene oxide (A-PPO) molecules with
varied molecular weight were grafted onto oxidized multi-walled carbon nanotubes
(o-MWCNTs) via a Zwitterionic reaction to modify the interface chemistry. Thermomechanical
measurements and micro-structural observations were carried out and results were analyzed in the
context of a concentric layer model with a non-bulk interphase. The focus was to elucidate the
relationships between polymer grafts, microstructure, and thermomechanical properties.
4.2. Experimental
CVD grown MWCNTs 20-60 nm in diameter (Shenzhen Nanotech Port Co. Ltd.,
China) were selected for the experiments. Three grams of the MWCNTs were dispersed in 250
ml of a concentrated acid mixture (3:1 volume ratio of 98%H
2
SO
4
/70% HNO
3
) following the
procedure used by Liu et al. [23]. Moderate oxidation was achieved by thirty minutes of

88
sonication (Aquasonic 150D, VWR) followed by dilution with 1000 ml deionized water and
agitation at 20°C for one hour. This procedure retained a high nanotube aspect ratio confirmed
by TEM observations (Fig.1). After the acid treatment and exhaustive washes with deionized
water, 36.5% HCl was added to the acid mixture to terminate defect sites on the MWCNTs with
carboxylic acid groups, rather than carboxylates. The solution was again extensively washed with
deionized water until the pH value of the decantate was the same as deionized water. The
oxidized MWCNTs (o-MWCNTs) were collected on a cellulose acetate membrane (0.45 µm pore
size, VWR scientific) by vacuum filtration and were dried in a vacuum oven at 80 °C for 24 hours.
The o-MWCNTs were milled and mixed with two forms of A-PPO molecules from
Huntsman Co. (Fig.2) and subsequently reacted in an oil bath at 140 °C under nitrogen flow for 48
hours. Zwitterionic linkages were formed between the primary amine groups and the carboxylic
groups on o-MWCNTs [24, 25]. The MWCNTs with (PPO)
3
-amine and (PPO)
33
-amine surface
grafts were denoted as 200-MWCNT and 2000-MWCNTs respectively, reflecting the molecular
weight of the grafted polymers (~ 200 and 2000 amu). After the treatment, the nanotubes were
thoroughly washed with anhydrous ethanol, collected with a PTFE membrane (0.22µm pore size,
VWR Scientific), and dried in a vacuum oven at 80 °C for 24 hours.
The morphology of o-MWCNTs was inspected by transmission electron microscopy
(TEM, Philips EM420 at 120 KV) to ensure the retention of aspect ratio after acid attack.
Thermogravimetric analysis (TGA) was employed to measure the weight loss or

89

Figure 4.1 Transmission electronic microscope image of oxidized MWCNTs.

90

X=2.6 (PPO)
3
-diamine Mw ~200
X= 33 (PPO)
33
-diamine Mw ~ 2000

Figure 4.2 A schematic view of the Zwitterionic reactions between oxidized MWCNTs and two
forms of A-PPO with different molecular weight.

+ o-MWCNT Æ
OOC-MWCNT

91
defunctionalization of modified MWCNTs at a temperature ramp rate of 10°C per minute to a
maximum of 1000°C with constant nitrogen flow (TGA 2050, TA Instruments). The formation of
the ionic complexes on the MWCNT surfaces was characterized by FTIR spectroscopy in
transmission mode (Nicolet 4700, Thermo Co.). Spectra were acquired from films of nanotube
suspensions in anhydrous ethanol (ethanol was thoroughly evaporated before testing), supported
by a NaCl crystal. The NaCl crystal support was used instead of KBr pellets to avoid the ion
exchange between KBr and the ionic species on CNT surfaces. Spectral data were collected with
a 256-scan average and 4 cm
-1
step size under nitrogen purge.
The three variants of MWCNTs - o-MWCNTs, 200-MWCNTs, and 2000-MWCNTs -
were dispersed directly in epoxy resin (EPON 828, Miller-Stephenson Chemical) at 0.1, 0.25, and
0.75 phr (parts per hundred resin, or 0.08wt%, 0.38wt%, and 0.57wt% respectively) using bath
sonication for 30 minutes at 80 °C. Nanocomposite specimens were prepared based on the “true”
nanotube content obtained from thermogravimetric data at 600 °C (Section 3.1). The weight
associated with the surface functionalities were ignored when calculating the epoxy
stoichiometric ratio because the weight fractions were negligible. After cooling to 25 °C, the
amine curing agent (D230, Huntsman Co.) was blended with the suspensions at a ratio of 32 phr
(or 24.2 wt %) for 5 minutes and degassed for 3 minutes in a hybrid shear mixer (HM-501,
Keyence). Then, the mixtures were cast into aluminum molds and degassed in a vacuum oven at
80 °C for another 5 minutes. Finally, all of the samples were cured at room temperature to reduce

92
residual stress. To ensure full curing, samples were post-cured for 2 hours at 80 °C followed by 3
hours at 120 °C. Three beam-shaped coupons (60×10×3mm) were cut and polished from each
specimen for dynamic mechanical tests. Following the procedure described above, control
samples from pure epoxy resin were also prepared and tested.
Differential scanning calorimetry (DSC), dynamic mechanical analysis (DMA), and
scanning electron microscopy (SEM) observations were carried out on bulk nanocomposite
samples. DSC tests were performed on composites heated from -60 °C to 250 °C at a 10 °C /min
ramp rate with a temperature modulation of ±1 °C /min (DSC2920, TA Instruments). Transition
temperatures were obtained from the mid-point of the heat-flow steps. Dynamic mechanical
analyses were employed using a dual cantilever beam geometry to monitor the dynamic moduli
(E’, E”) and glass transition temperature (T
g
, from loss modulus peak) change in the composites
from 25 °C to 150 °C. A heating rate of 5 °C per minute was used with a load frequency of 1 Hz
(DMA 2980, TA Instruments). The fracture surfaces of the nanocomposites were examined by
SEM imaging (Cambridge 360 at 20kV) after sputter-coating with gold.
4.3. Results and Discussion
4.3.1. Surface Modification Assisted Dispersion
Surface modification of MWCNTs involved a two-step wet chemistry process in which
acid oxidation was followed by grafting. The acid treatment of the as-received CVD-grown
MWCNTs removed amorphous carbon and residual transition metal catalysts. The oxidation

93
process also created defect sites and introduced surface functional groups of carbonyl, carboxyl,
and phenol onto the MWCNT surface [26]. In the second step - the “grafting to” stage - surface
carboxylic groups formed Zwitterionic linkages with A-PPO molecules (Figure 2).
TGA results (Fig.3) confirmed that organic species were introduced onto the MWCNT
surface. The thermal decomposition curves for both A-PPO functionalized MWCNTs showed
steps at ~340 °C. The steps were within the range of decomposition temperatures of the A-PPO
molecules. The surface organic content was calculated from the TGA data at 600 °C. At this
temperature, the nanotube weight loss was negligible, but the surface grafts had decomposed
completely, resulting in measurable weight loss that was attributed solely to the loss of surface
organic species. Since the value of the weight loss was a dependant of the grafting density, the
molecular weight of the surface grafts, and the properties of the nano-particles, the grafting
density was estimated to be ~ 2-4 chains per square nanometer for the 2000-MWCNT and
200-MWCNT respectively. The relatively high values of surface graft density and the higher
grafting efficiency for grafts with lower molecular weight were consistent with other reports [27,
28].
Additional insights into the interface linkage were provided by interpretation of FTIR
spectra (Fig. 4). A broad band was present at 1085 cm
-1
for both the 200-MWCNTs and the
2000-MWCNTs. The prominent band, which dwarfed the contribution from other functionalities
on MWCNT surface, was attributed to the aliphatic ether band and the primary amine band. In

94

200 400 600 800 1000
0
10
20
30
40
50
60
70
80
90
100
o-MWCNT
200-MWCNT
2000-MWCNT
(PPO)
33
-diamine
Normalized Weight (%)
Temperature (
o
C)

Figure 4.3 TGA thermal decomposition data of amine functionalized MWCNTs compared with
those of o-MWCNTs and A-PPO.

95

Figure 4.4 FTIR spectra for functionalized MWCNTs at 900 ~ 1750 and 2800 ~ 3000
wavenumbers.

96
addition, carboxylic acid peaks and carboxylic anion peaks were observed for the o-MWCNTs
and A-PPO functionalized species at 1733 cm
-1
and 1670 cm
-1
respectively. The downshifting of
the carboxylic peak implied that carboxylic bonds were weakened as a result of the ionic
interaction between the terminal amine and carboxylic acid. The mutually overlapping absorption
bands at ~ 1600 cm
-1
were attributed to multiple functionalities, including the unreacted primary
aliphatic amine, the NH
3+
group of primary amine salt, and the so-called “mystery band”
observed in carbonaceous materials [29]. Unlike the A-PPO functionalized MWCNTs, no peaks
were present at the C-H bond region (2800-3000 cm
-1
) for o-MWCNTs.
Both TGA and FTIR data provided evidence for the existence of a layer of A-PPO
molecules on o-MWCNT surfaces. Considering the unperturbed radius of gyration of these
A-PPO chains [30], the thickness of this molecular layer is estimated to be sub-nanometer. Such
molecular chains with single-end constraints on solid surfaces have been described as polymer
brushes [31-34], and the characteristic length of such tethered molecule layers is reportedly
governed by the molecular weight of the surface grafts, the grafting density, and solvent
conditions [35]. For closely matched solubility conditions and high surface grafting density, the
characteristic length of the grafts is estimated to be in the nanometer range, even for long-chain
A-PPO grafts (2000 amu).
The dispersion of raw MWCNTs in aqueous or organic solvents is dictated by
thermodynamic criteria and Van der Waals forces. For functionalized nanotubes, the process is

97
also governed by the steric effect and the solubility of the polymer grafts. Moreover, if the
functionalities are ionized, electrostatic repulsive forces play an important role. Thus, in the
solvent-free curing scheme described above, judicious selection of the polymer grafts is essential
to achieve stable dispersions. The dispersion of functionalized nanotubes depends on the
solubility of the nanotube surface grafts in the solvent (matrix) system. However, measuring the
solubility parameters of a chain molecule in which one end is tethered (i.e., covalently bonded to
the CNTs) presents a challenge. Instead, a qualitative indication of miscibility at the process
temperature and 1 atmosphere pressure is provided for the solute (untethered A-PPO) and the
solvent (DBEBA epoxy). Because of the complete miscibility of A-PPO and epoxy at the
abovementioned conditions, the solubility parameter of A-PPO is close to that of the DGEBA
epoxy (~ 20 MPa
1/2
, in terms of the Hildebrand solubility parameter) [36].
The dispersion of MWCNTs in the selected epoxy resin was facilitated by the A-PPO
grafts through solubility matching, thereby validating the hydrophobic, or “epoxy-philic” in this
case, design concept. The observed A-PPO/epoxy miscibility indicated that the graft-resin
interaction was energetically favourable relative to the intra-chain interaction of the surface grafts.
Consequently, the A-PPO surface grafts interacted strongly with the resin system and led to a
brush-like molecular layer on MWCNT surfaces in the nanotube-epoxy solid suspension. As a
result, both 200-MWCNTs and 2000-MWCNTs formed stable solid suspensions (<1wt%) in the

98
epoxy system after sonication at 80 °C for 20 minutes. In contrast, the o-MWCNTs precipitated
out of the epoxy system when left for 24 hours after sonication.
4.3.2. Polymerization and Interphase Formation
The matching of solubility parameters between the A-PPOs and the epoxy system
supported the notion of “epoxy compatible” surface grafts, yet the ability for these surface grafts
to form linkages with the epoxy system was equally important. The surface reaction at the
vicinity of functionalized MWCNT surface was distinctive. In particular, the surface functional
groups on the o-MWCNTs reacted directly with the epoxy, resulting in cross-links near the
nanotube surface [37]. For A-PPO functionalized MWCNTs, the grafted A-PPO chains formed
cross-links at one of the terminal amine groups. The other amine group that participated in the
ionic bond was not in an energetically favourable position to form cross-links because of steric
effects. Despite the small weight fractions of the grafts, the thermodynamics, the mass-transport,
and hence the curing process in the vicinity of the functional groups were significantly different
from those of the bulk epoxy. Consequently, the surface grafts led to the formation of a non-bulk
interphase in the nanocomposites.
The assertion posed above was supported by DSC results that showed a notable
secondary transition for the 2000-MWCNT nanocomposites at cryogenic temperatures (Fig. 5),
while no secondary transition step was detected in any of the other samples (all in 0.75 phr). This
secondary transition produced a relatively broad step, indicating the existence of a third phase, the

99

Figure 4.5 DSC curves for MWCNT reinforced epoxies at 0.75 phr compared with control epoxy
sample.

100
interphase region, with low cross-linking density in 2000-MWCNT composites. An analogy can
be drawn between the secondary transition and the T
g
depression observed in thin films [38, 39],
which Mayes and Schadler et al. ascribed to a “softer” non-bulk region at the interface.
The formation of this “softer” region in the 2000-MWCNT system can be ascribed to
two causes. First, the long compliant PPO chains reduced cross-linking density by either
serving as spacers between the cross-linking sites and the MWCNT surface or lowering the
grafting efficiency mentioned above. Second, preferential surface adsorption of curatives onto the
2000-MWCNTs also contributed to the non-bulk layer, a phenomenon reported for similar
systems with long-chain amine-functionalized MWCNT or SWCNT inclusions [20, 21]. As a
result, a slight shift from stoichiometric composition of bulk epoxy was observed in
the2000-MWCNT composites, as indicated by the T
g
drop in Fig.5. Additional evidence for the
existence of a non-bulk interphase was provided by thermomechanical tests and microscopic
observations, described in the following sections.
4.3.3. Thermomechanical Properties
The incorporation of low weight fractions of modified MWCNTs into the epoxy matrix
caused significant changes in the glass transition temperature (T
g
) and the storage modulus (E’)
according to DMA measurements (Figs. 6, 7 and Table 1). The values of both T
g
and E’ increased
monotonically with increasing o-MWCNT and 200-MWCNT loading, albeit at a faster rate for
the o-MWCNT system. In contrast, the nanocomposites reinforced with 2000-MWCNT exhibited

101

60 65 70 75 80 85 90 95 100 105 110 115
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Epoxy Control
o-NT 0.1
o-NT 0.25
o-NT 0.75
2000-NT 0.1
2000-NT 0.25
2000-NT 0.75
Normalized Loss Modulus
Temperature (
o
C)

Figure 4.6 Loss modulus of o-MWCNT and 2000-MWCNT (o-NT and 2000-NT for simplicity)
reinforced epoxy nanocomposites. 200-MWCNT composites have similar trend in Tg as o-CNT
composites and therefore data are not shown.

102

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
2100
2200
2300
2400
2500
2600

o-MWCNT
200-MWCNT
2000-MWCNT
Storage modulus (MPa)
Nanotube Concentration (phr)

Figure 4.7 Storage modulus (25 °C) of epoxy nanocomposites reinforced with different
MWCNTs and results converge at the datum point for neat epoxy.

103

Table 4.1 DMA data of the glass transition temperatures (Tg) of MWCNTs reinforced epoxies vs.
MWCNT loadings.

o-MWCNT 200-MWCNT 2000-MWCNT
0 phr 92.2 °C 92.2 °C 92.2 °C
0.1 phr 95.9 °C 93.5 °C 91.1 °C
0.25 phr 96.7 °C 95.0 °C 89.7 °C
0.75 phr 97.1 °C 96.3 °C 88.6 °C

104
decreases in T
g
and E’. Although the exact T
g
values slightly differed from those determined
from the DSC measurements, the trends were consistent and reproducible.
The disparate trends in glass transition temperatures arose from the distinctive
mechanisms active in these systems. As noted, the decrease in T
g
in the 2000-MWCNT
composites was attributed to the off-stoichiometric composition, which arose because of selective
surface adsorption. On the other hand, the observed increases in T
g
for both o-MWCNT and
200-MWCNT composites were caused by the local confinement effect of CNTs at the interface
[40, 41] and the CNT percolation network at a global level (section 3.5). The confinement effects
dominated the relaxation behaviour in these short-graft systems, a conclusion supported by the
absence of a prominent secondary transition step in the DSC data.
Nancomposites with o-MWCNTs and 200-MWCNTs exhibited increases in storage
modulus (E’), the magnitudes of which increased with MWCNT loadings. A ~10% increase in E’
was observed for 0.57wt% (~0.3v%) o-MWCNT and 200-MWCNT composites respectively.
Variations in raw materials and processes, as well as the limitations of current theoretical models,
limited our ability make a quantitative assessment of the reinforcement efficiency and compare
this with literature reports and predictions [42]. Nevertheless, the reinforcement trends are clear
for these MWCNTs with short surface functionalities.
In contrast, the E’ values in the 2000-MWCNT system decreased with increasing
CNT loading. The decreases in E’ were attributed to the dominant effect of interphase properties

105
in composites with long A-PPO functionalized MWCNTs. Stress transfer was most efficient
when the interphase deformed primarily through molecular bond length change under external
forces [22, 43]. However, when long-chain grafts were used, the chain relaxation process at the
interphase was dominated by conformational entropy similar to rubbery polymers. Thus, the
stress transfer efficiency was diminished by the compliant interphase for nanocomposites with
2000-MWCNTs. In addition, as the volume fraction of the annular interphase regions increased,
the interphase properties began to dominate the bulk properties, particularly the modulus. While
the decline in E’ might be viewed as undesirable, beneficial effects include enhancements in
impact toughness [20] and damping properties [44] similar to those of polymer alloys [45].
4.3.4. Microstructural Analysis
Microscopic observation indicated that the “epoxy-philic” functionalization increased
the uniformity of CNT dispersion in the thermoset network and established linkage at the
interface. Uniform dispersion of the A-PPO functionalized MWCNTs was maintained during
gelation and fixed in the epoxy network upon curing, as evidenced in Fig. 8. c-d. No direct
(clean) nanotube pullout in the A-PPO functionalized species was observed, and this was
attributed to the interfacial bonding at the interface. A mechanical interlocking mechanism,
stemming from the naturally curved configurations of the nanotubes, also inhibited telescopic
separation of concentric tubes. In contrast, CNT agglomerations with pullout/debonding
(indicated by arrows in Fig. 8-b) were observed in the o-MWCNT composites. The existence of

106

Figure 4.8 (a) – (f) SEM images of the nanocomposite fracture surfaces (in 0.75 phr). (a). Neat
epoxy control sample. (b) o-MWCNT, arrows indicate CNT debonding and agglomeration. (c)
200-MWCNT. (d) 2000-MWCNT, fast fracture surface. (e) 2000-MWCNT, fibril structures. (f)
2000-MWCNT, fibril structures with zoomed in image.

107
clusters in the o-MWCNT composites accounted for the slightly lower reinforcement efficiency at
higher CNT concentrations (Fig.7). The composite modulus was insensitive to nanotube
agglomerations at low CNT loadings [46]. Consequently, the storage moduli for o-MWCNT
and 200-MWCNT nanocomposites were comparable at the nanotube observed despite of the
markedly difference in the dispersion state.
The fracture surfaces of the 2000-MWCNT composites exhibited features distinct from
the brittle fracture surfaces of composites with 200-MWCNTs (Fig. 8-c). Overall, the fracture
surfaces of the 2000-MWCNT composites showed predominantly brittle fracture characteristics
(Fig. 8-d, f). However, SEM inspection revealed scattered regions with clusters of long fibrils
that extended from the fracture surface (Fig. 8-e, f). These fibrils were tens of microns in length
and 1-2 µm in diameter, which is ~30 times the mean diameter of the MWCNTs.
The dimensions and shapes of the observed fibrils were consistent with the existence of
nanotubes embedded within the fibrils. Most reported observations of nanotube pullout have
described nanotubes extending from the fracture surface and covered by a layer of matrix material,
as opposed to “clean” pullout of CNTs [47]. For example, CNTs coated with thick interphase
polymer layers (4-6 times the radius of the embedded nanotube, or a fiber-to-interphase volume
ratio of approximately 1:25) were observed on nanocomposite fracture surfaces reinforced with
non-functionalized MWCNTs [48]. In addition, the separation of these fibrils from the matrix was
dominated by shear deformation, as shown in Fig.8-e and indicated by arrows. The appearance of

108
the fibrils on the fracture surface implied the existence of a compliant annular region induced by
the long-chain surface grafts around the 2000-MWCNTs. The diameter variation of these fibrils
also indicated that the adsorption properties and the crosslinking process varied on the surface of
the treated MWCNTs. Although the diameter of the fibrils was roughly an order of magnitude
larger than that of MWCNTs, the exact thickness of the interphase layer was uncertain for these
MWCNTs functionalized with long-chain surface species.
The distribution of these fibrils on the surface shed some light on the fracture
mechanisms for the 2000-MWCNT nanocomposites. As described above, the fibrils were
clustered and situated adjacent to large featureless areas characteristic of fast brittle fracture
(Fig.8-d, f). Therefore, the observed fibrillated regions were likely formed at sites of crazing that
preceded the brittle fracture. During the latter stage of brittle fracture, the crack was wide and
advanced quickly, and fracture was dominated by the brittle epoxy matrix rather than the
nano-scale mechanisms that would facilitate fibrils formation. Within these fast fracture regions,
MWCNTs were observed as bright dots, similar to those reported by Gojny et al. [40].
4.3.5. Model Studies and Refinement
In light of the thermomechanical and micro-morphological evidence presented above, a
concentric layer (CL) model of constituent phases can be employed to elucidate the material
behaviour. This model, illustrated in Figure 9, was initially outlined in the study of CNT
reinforced thermoplastic materials [49, 50] In Fig.9, the modulus of the interphase is denoted by

109

Figure 4.9 Schematic illustration of the CL model with concentric layers of fiber, matrix, and
interface. Double curves for the interphase region indicate tunable or gradient properties.

110

Figure 4.10 The full width half maximum (FWHM) values of the normalized loss modulus curves
(E”) vs. nanotube loading for nanocomposites with varied nanotubes.

0.00.2 0.40.6 0.8
6.5
7.0
7.5
8.0
8.5
9.0
9.5

Control Sample
2000-MWCNT
200-MWCNT
o-MWCNT
FWHM (
o
C)
Nanotube Concentration (phr)

111
two different curves representing distinct interphase compliance gradient. If the interphase is
designed to impart “added benefits”, such as electrical conductivity or UV absorption, the CL
model can also be adjusted to provide an adequate description of the multifunctional system.
Although factors such as nanotube waviness, agglomeration, and misorientation are
ignored [51, 52] in the CL model, when suitable constitutive homogenization techniques are
incorporated, the nature of the interphase dominates the material behaviour. Despite a lack of
quantitative information about the thickness and properties of the interphase, a qualitative
explanation for the storage modulus and glass transition trends reported in this work can be
offered by postulating either a compliant or a rigid interphase. As pointed out by Eitan et al. [49],
the local nanotube confinement effect can be approximated by a restricted or “rigid” interphase.
Therefore, a rigid interphase, as in the 200-MWCNT and o-MWCNT systems, results in
increased values of modulus and T
g
, while a compliant interphase, as in composites with
2000-MWCNTs, leads to a decrease in modulus and secondary glass transition when the
compliant interphase becomes large enough to dominate the thermomechanical responses.
Although the non-bulk molecular layer of the CL model was essential for explaining
some of the thermomechanical and micro-morphological properties observed in our study,
additional mechanism needed to be introduced in order to fully elucidate the behaviour of these
materials. Reports on thermoplastic composite systems [49, 53] indicate that the thermal
relaxation behavior of the bulk material is altered by the relaxatory nature of the interphase,

112
which results from either a rigid or compliant non-bulk layer outlined in the CL model. A good
indication of the uniformity of thermal relaxation is provided by the full width half maximum
(FWHM) values of loss modulus curves [37] shown in Fig. 10. If the relaxation behaviour of the
non-bulk layer is close to that of the bulk matrix, a merged loss modulus curve, broadened and
shifted, will result. However, no significant change in the shape of the loss modulus curve was
observed for the 2000-MWCNT system, and peak narrowing similar to previous reports [37, 40]
was observed for both the o-MWCNT and the 200-MWCNT systems. For the 2000-MWCNT
system, the interphase properties differed significantly from those of matrix, and thus two
separate peaks were expected in the loss modulus curve, as supported by the two transition steps
from the DSC heat flux measurement (Fig. 5). On the other hand, for those systems with short
surface grafts, the CL model alone is insufficient to explain the observed peak narrowing, and a
global confinement mechanism must be invoked as a complement o the CL model. Global
confinement suppresses the overall segmental motion at all temperatures and frequency ranges, as
explained below.
Since the T
g
values were insensitive to nanotube loading at higher concentrations
(Table 1), the narrowed loss modulus curves were attributed to the overall confinement effect of a
3-D nanotube percolation network. The existence of a percolation threshold at low CNT
concentrations (< 0.1wt %) [54, 55] and the formation of the percolation network at higher
concentrations contributed to the observed increase in glass transition temperature. This effect

113
was also strengthened by the local CNT confinement effect for systems with short surface grafts
described in Section 3.3. The CNT percolation network functioned collectively with the
cross-linked matrix and resulted in loss modulus curve up-shifting (T
g
increase) and narrowing, in
a manner similar to epoxy systems with higher cross-linking density [37]. While the global
confinement effect was alleviated in the 2000-MWCNT system because of the compliant
interphase, the concentric layer model for nanotube reinforced composites will undoubtedly
benefit from refinements that incorporate the global confinement effect of a 3-D CNT percolation
network.
4.4. Conclusions
Oxidized MWCNTs were grafted with epoxy-compatible A-PPO chains of varied
lengths via an ionic bonding scheme and dispersed in epoxy. After polymerization, the dispersion
was preserved, and covalent bonds were formed at the reinforcement-matrix interface.
Nanocomposites reinforced with long or short A-PPO functionalized MWCNTs exhibited
opposite trends in thermomechanical behaviour due to the formation of an either rigid or
compliant interphase. The compliant interphase in the nanocomposites with long-chain A-PPO
grafted MWCNTs also resulted in a distinctive fracture surface morphology.
The findings indicate that CNT functionalization can be used to introduce either a stiff
or a flexible interphase in thermoset nanocomposites, depending on the nature of surface grafts
and their interaction with the matrix system. A concentric layer model captures the salient

114
features of the nanocomposite behaviour, although a more realistic model that accounts for the
percolation network is required to fully illuminate the structure-property relationships. More
broadly, the present findings demonstrate the ability to engineer interphases in inorganic
nano-fibril-reinforced polymer composites. This capability is essential to the design and synthesis
of biomimetic or hybrid materials that may one day approach the performance levels of natural
nanocomposites. Through the judicious selection of solubility parameters, molecular weight,
flexibility, and reactivity of the grafts with those of the host matrix, the thermomechanical
properties of the nanocomposites can be tailored. When combined with nano-atomic scale
assembly techniques, the ability to engineer interphase properties opens myriad possibilities for
designing high-modulus, high-toughness, and multifunctional materials.

115
Chapter 4 References
[1] Chou TW, Thostenson ET. Carbon 2006; 44 (14):3022-29.

[2] Iijima S. Nature 1991; 354: 56 – 58.

[3] Ruoff RS, Lorents DC. Carbon 1995; 33 (7): 925-30.

[4] Yu M, Lourie O, Dyer MJ, Moloni K, Kelly TF, Ruoff RS. Science 2000; 287:637-40.

[5] Wagner HD, Liu LQ. Compos Sci Technol 2005; 65 (11-12): 1861-68.

[6]Gojny FH, Wichmann, MHG, Kopke U, Fiedler B, Schulte K. Compos Sci Technol 2004;
64(15):2363-71.

[7] Miyagawa H, Misra M, Mohanty AK. J Nanosci Nanotechnol 2005; 5(10):1593-615.

[8] Thostenson ET, Ren ZF, Chou TW. Compos Sci Technol 2001; 61(13):1899-912.

[9] Jonathan N, Coleman, UK, Yurii K, Gun K. Adv Mater 2006; 18:689-706.

[10 ] Zeng LL, Zhang L , Barron AR. Nano Lett 2005; 5(10):2001-04.

[11] Paiva MC, Zhou B, Fernando KAS, Lin Y, Kennedy JM, Sun YP. Carbon 2004;
42(14):2849.

[12] Paiva MC, Zhou B, Fernando KAS, Lin Y, Lopes PE, Pennington WT, Kennedy JM, Sun
YP. Compos Inter 2005; 12(8-9):757-68.

[13] Frankland SJV, Caglar A, Brenner DW, Griebel M. J Phys Chem B 2002;106: 3046-48.

[14] Schadler LS, Giannaris SC, Ajayan PM, Load transfer in carbon nanotube epoxy
composites. Appl Phys Lett 1998;73(26):3842-44.

[15] Chen W, Auad ML., Williams RJJ, Nutt SR. Euro Polym J 2006; 42: 2765-72.

[16] Tasis D, Tagmatarchis N, Bianco A, Prato M, 2006;106:1105-36.

[17] Banerjee S, Hemraj-Benny T, Wong SS. Adv Mater 2005;17(1):17-29.

116
[18] Wu HX, Tong R, Qiu XQ, Yang HF, Lin YH, Cai RF, Qian SX. Carbon 2007; 45:152-59.

[19] Cui J, Wang WP, You YZ, Liu CH, Wan PH. Polymer 2004;45:8717-21.

[20] Wang JG, Fang ZP, Gu AJ, Xu LH, Liu F. Journal of Appl Polym Sci 2006;100:97-104.

[21] Shen JF, Huang WS, Wu LP, Hu YZ, Ye MX. Compois Sci Technol 2007; 67:3041-50.

[22] Shen JF, Huang WS, Wu LP, Hu YZ, Ye MX. Composites: A 2007;38:1331-36.

[23] Liu J, Rinzler AG, Dai H, Hafner JH, Bradley RK, Boul PJ, Lu A, Iverson T, Shelimov K,
Huffman CB, Rodriguez-Macias F, Shon Y, Lee TR, Colbert DT, Smalley RE. Science
1998;280:1253.

[24] Wang M, Pramoda KP, Goh SH, Carbon 2006;44: 613-17.

[25] Hamon MA, Chen J, Hu H, Chen YS, Itkis ME, Rao AM, Eklund PC, Haddon RC. Adv
Mater 1999; 11(10):834-40.

[26] Zhang J, Zou HL, Qing Q, Yang YL, Li QW, Liu ZF, Guo XY, Du ZL. J Phys Chem B
2003;107:3712-18.

[27] Xie L, Xu F, Qiu F, Lu HB, Yang YL. Macromolecules 2007; 40(9): 3296-305

[28] Chen GX, Kim HS, Park BH, Yoon JS, J of Phys Chem B 2005; 109(47): 22237-43

[29] Zawadzki J. Chemistry and physics of carbon, v21, Torun, Poland: Institute of chemistry,
Nicolaus Copernicus University;1989.

[30] Bazuaye A, Huglin, MB. Polymer 1979;20(1):44-50

[31] Netz RR, Schick M. Europhys Lett 1997; 38(1): 37-42

[32] Shull KR. J Chem Phys 1991;94(8): 5723-38

[33] Grest GS. J Chem Phys 1996; 105(13): 5532-5541

[34] Mezzenga R, Boogh L, Manson JAE. J Polym Sci B 2000; 38(14):1883-92

117
[35] Kilbey II, SM, Watanabe H, Tirrell M. Macromolecules 2001;34(15):5249-59

[36] Barton AFM. CRC, CRC Press; 1990.

[37] Ellis B. Glasgow UK: Blackie Academic & Professional; 1993, Ch.3.

[38] Mayes AM. Nature Mater 2005; 4:651-52.

[39] Bansal A, Yang H, Li CZ, Cho K, Benicewicz BC, Kumar SK, Schadler LS. Nature Mater
2005;4:693-98.

[40] Gojny, FH, Schulte K. Compos Sci Technol 2004; 64(15):2303-08.

[41] Wang JG, Fang ZP, Gu AJ, Xu LH, Li, F. J Appl Polym Sci 2006; 100(1):97-104.

[42] Ajayan PM, Schadler LS, Braun PV. Nanocomposite Science and Technology. John Wiley,
Sons Inc.; 2003 Ch.1, 2.

[43] Wang S, Liang ZY, Liu T, Wang B, Zhang C. Nanotechnology 2006; 17:1551-57.

[44] Ajayan PM, Suhr J, Koratkar N. J Mater Sci 2006; 41(23):7824-29.

[45] Hu R, Dimonie VL, El-Aasser MS, Pearson RA, Hiltner A, Mylonakis SG, Sperling LH. J
Polym Sci, B 1997;35(10):1501-14.

[46] Shi DL, Feng XQ, Huang YG, Hwang KC, Gao HJ. Trans. ASME 2004; 126: 250-57.

[47] Wong M, Paramsothy M, Xu XJ, Ren Y, Li S, Liao K. Polymer 2003;44(25):7757-64.

[48] Ding W, Eitan A, Fisher FT, Chen X, Dikin DA, Andrews R, Brinson LC, Schadler LS,
Ruoff RS. Nano Lett 2003; 3(11):1593-97.

[49] Eitan A, Fisher FT, Andrews R, Brinson L, Schadler LS. Compos Sci Technol 2006;
66:1162-73.

[50] Fisher, FT and LC Brinson (2001). Compos Sci Technol, 61:731-48.

[51] Fisher FT, Bradshaw RD, Brinson LC. Appl Phys Lett 2002;80(24):4647-49.

118
[52] Anumandla V, Gibson RF. Composites Part A: Appl Sci Manuf 2006;37: 2178-85.

[53] Yang J, Lin YH, Wang JF, Lai MF, Li J, Liu JJ, Tong X, Cheng HM, J Appl Polym Sci
2005; 98(3):1087-91.

[54] Wang L, Dang ZM. Appl Phys Lett 2005;87(4):42903-13.

[55] Zhang QH, Rastogi S, Chen DJ, Lippits D, Lemstra PJ. Carbon 2006;44(4):778-85.

119
Chapter 5. Towards hierarchical reinforcement
5.1. Motivation
Several approaches have been employed to improve the properties of fiber-reinforced
epoxies. One is to select fiber reinforcements which optimize specific thermal, mechanical,
chemical, electrical, or optical properties [1-3]. A second approach is to toughen the matrix and
overcome the inherent brittleness of epoxy systems through techniques such as
rubber/thermoplastic toughening or epoxy composition modulation [4, 5]. A third alternative is to
optimize the fiber-matrix interface to enhance the stress transfer properties [6, 7]. More recently,
multi-scale approaches have been explored to design optimized microstructures at multiple
reinforcement length scales [8-11].
Fiber-reinforced epoxy composites filled with nano-particles afford the opportunity to
improve the bulk composite properties with minimal sacrifice of other properties of the
composites. Among nano reinforcements, multi-walled carbon nanotubes (MWCNTs) stand out
because of the ultra-high strength/stiffness [12], large aspect ratio, and relative affordability.
Examples of multi-scale composites utilizing MWCNT-filled epoxies in conventional
fiber-reinforced composites (FRCs) have been reported. For example, Thostenson et al. [13, 14]
demonstrated that the anchorage of MWCNTs onto carbon fibers through either direct CVD
growth or electrophoresis selectively tailored the “interface properties”. In other work, Qiu et
al. [15] and Gojny et al. [16] reported that MWCNTs enhanced the electrical properties and the

120
interlaminar shear strength respectively, while preserving or enhancing tensile properties, thus
improving “matrix properties”.
Despite the temptation to assemble or organize ordered nanotube arrays into
composites [17, 18], direct dispersion of MWCNTs with tailored surface properties, offers the
most practical route because of the simplicity and compatibility with existing composite
processing methods. Three key processing challenges are associated with direct CNT
incorporation - nanotube dispersion, interfacial bonding, and alignment. [19, 20] The dispersion
and interfacial bonding issues can be addressed by identifying a suitable nanotube surface
modification, which not only disperses the MWCNTs in the resin system but also facilitates stress
transfer at the interface through the formation of covalent bonds [21, 22]. In addition, high shear
force, electric field, and magnetic field have been used successfully to align CNTs in polymer
matrices [23-25]. However, few has mentioned the influence of processing conditions and CNT
surface modification to CNT dispersion and alignment in a complex composite system with
multiple reinforcement length scale and processing stages.
In the present study, multi-walled carbon nanotubes were grafted with
epoxy-compatible surface modifiers [26] before being dispersed directly into the epoxide. The
resultant nanotube-filled epoxies were used as matrices in composite laminates reinforced with
continuous basalt fibers, selected for the unique thermal, mechanical, and chemical
characteristics. The morphology and mechanical behaviors of the nanotube reinforced epoxies

121
(NEs) were analyzed and related to those of NE/basalt laminates in order to elucidate the
composition, processing, structure, and property relationship in these hierarchically reinforced
composites
5.2. Experimental
5.2.1. Materials
Raw MWCNTs (raw-CNTs) were obtained from a commercial source (Nanotech Port
Co. Ltd., Shenzhen., China). The nanotubes were produced by chemical vapor deposition (CVD),
and contained ~5wt% impurities, consisting primarily of amorphous carbon and transition metals.
The nanotubes were 5 - 15 µm in length and 20 - 60 nm in diameter. A di-glycidyl ether of
bisphenol A (DGEBA) type epoxy resin (EPON 828, Miller-Stephenson Chemical Co. Inc.) was
selected as the matrix material. The epoxy was used in combination with a curing agent and an
accelerator (DICY and Omicure, respectively, Air Products & Chemicals, Inc.) to prepare the
composites.
H
2
SO
4
, HCl, and HNO
3
were generic chemicals (VWR Scientific) used during the
oxidation. Triphenylphosphine (TPP, from Aldrich) was employed as a catalyst to facilitate the
esterification reaction that attached phenyl glycidyl ether (PGE, Aldrich) onto the acidified
carbon nanotubes. Cellulose acetate and PTFE membranes (VWR and Fisher Scientific) were
used during the filtration of MWCNTs. DMF, THF, and anhydrous ethanol were used as solvents
(VWR Scientific).

122
The basalt fibers used (Rov 68-680/10/Int/Ext., BASALTEX, a division of
MASUREEL Holding nv, Belgium) exhibited the following properties: a melting point
=1350
o
C±100, diameter = 10 µm±1.5, tensile modulus= 84 GPa±3, and the sizing type was
silane.
5.2.2. Chemical Modification of MWCNTs
MWCNTs (5 grams) were dispersed in 500 ml of concentrated 3:1 H
2
SO
4
/70% HNO
3
following the procedure used by Liu et al.[27] Moderate oxidation while maintaining a high
MWCNT aspect ratio was achieved by 30 minutes sonication and 30 minutes stirring at room
temperature. After the acid treatment and exhaustive washes with deionized water, HCl was
added to the acid mixture to facilitate the termination of CNT surface defects with carboxylic acid
groups, rather than carboxylate. The solution was again extensively washed with deionized water
until the pH value of the decantate became the same as deionized water. The acid-treated
MWCNTs (denoted as o-CNTs) were collected on a 0.45 µm cellulose acetate membrane by
vacuum filtration and were dried in a vacuum oven at 80
o
C for 24 hours.
Some of the o-CNTs were esterified through a procedure similar to the one reported by
Chen et al. [26] The esterification was carried out by mixing milled o-CNTs and PGE in a DMF
solution with TPP as the catalyst. The mixtures were refluxed in an oil bath heated to 150
o
C
under inert gas (N
2
) atmosphere for 36 hours. These esterified nanotubes with phenyl and
beta-hydroxyl functionalities on the surface were denoted as the PGE-CNTs. After the treatment,

123
the nanotubes were thoroughly washed with anhydrous ethanol, collected with the PTFE
membrane, and dried in a vacuum oven at 80
o
C for 24 hours.
5.2.3. Composite Preparation
Nano-epoxy coupons (NE coupons)
The raw-CNTs, o-CNTs, and PGE-CNTs were milled and dispersed in THF using a
bath sonication for 10 min. The epoxy was mixed with THF (1:1 by volume) in a dual-axis
high-speed shear mixer (Keyence HY501). The two mixtures were blended for another 10
minutes and degassed for 5 minutes in the mixer. The resultant suspension was denoted as
nanotube-filled epoxy (NE). Afterwards, THF was evaporated at 80
o
C in a vacuum chamber for
48 hours. Finally, the curing agent and the accelerator were added to the NE at 4.21 phr and 1.05
phr respectively. The epoxy system was then mixed and degassed again in the mixer prior to
casting. The off-stoichimetry epoxy composition was optimized for filament winding/hot press
processes and was used throughout this work. The blend was cast into an aluminum mold
pretreated with epoxy mold release to produce the NE specimens. The curing cycle was 1 hour at
120
o
C and post curing at 160
o
C for three hours. NE specimens were prepared at 0.25, 0.5, and 1.5
vol% MWCNT loadings. (See reference 28 for the estimation of vol%. The percentage of CNTs
refers to the fraction of CNT in epoxy resin throughout this work unless otherwise stated.) Five
beam-shaped NE coupons were cut and polished from each specimen for tensile test. Residual
samples were milled into powder for Differential Scanning Calorimetry (DSC) tests. Following

124
the procedure described above, control samples from pure epoxy resin were also prepared and
tested for comparison.
Multi-scale reinforced laminates (MSRs)
The NEs were also used to fabricate the MSR specimens. Unidirectional prepreg was
made on a lab-scale drum-winding machine, where basalt fibers were pulled through a resin bath
before being wound onto a spinning drum at preset rates. The resin bath contained NE (0vol%,
0.5vol% or 1.5vol% MWCNT loading), the curative and the initiator, which were prepared
beforehand using the same mixing and degas procedure mentioned above except that THF was
used as diluent at this stage. The unidirectional prepreg was cut from the drum and heated at 60
o
C
for 20 min. Four plies of prepreg were cut and stacked in a (0/90/90/0)

sequence. Lamination was
performed using a hot press at 120
o
C and 100 psi (~0.69 MPa) pressure for 1 hour, followed by a
post-cure at 160
o
C for 3 hours. Finally, laminated MSR samples were cut into five tensile
coupons (2.54 cm x 20.32 cm x 0.05 cm) with 3.81 cm tabs glued to both ends of each coupon.
5.2.4. Characterization
The o-CNTs and PGE-CNTs were observed by transmission electron microscopy
(TEM, Philips EM420, 120 KV). The weight loss, or defunctionalization of modified MWCNTs,
was recorded by thermogravimetric analyses (TGA) at a temperature ramp of 10
o
C per minute to
a maximum of 800
o
C under constant nitrogen flow. The attachment of functional groups on the

125
surface of MWCNTs was verified by FTIR spectroscopy in transmission mode (Nicolet 4700).
TGA and FTIR data were reported in reference 26.
DSC tests were performed (DSC2920, TA Instruments) using milled NE powder. The
samples were heated from 25 °C to 250 °C at 10 °C /min ramp rate with a temperature
modulation of ±1 °C /min. The glass transition temperatures were obtained at mid-point of the
transition step. The microscopic morphology of the NE and MR composites was both investigated
by SEM (Cambridge 360 at 20kV). Brittle fracture surfaces (pre-chilled by liquid nitrogen) were
sputter-coated with gold prior to observation. Tensile tests were performed on NE coupons and
MSR composites using a universal testing machine according to ASTMD638 (specimen type V)
and ASTM D3039 respectively.
5.3. Results and discussion
5.3.1. Properties of functionalized MWCNTs.
As described above, MWCNTs were functionalized through either an oxidation or an
esterification scheme, details of which have been reported elsewhere [26]. The functionalized
nanotubes (o-CNT and PGE-CNT) are shown in TEM images in Fig 1. Inspection of the entire
sample area revealed that structural integrity of the graphite layers was preserved and CNT aspect
ratios of over 100 were maintained. The observed nanotube diameters range from ~ 20 to 60 nm,
which confirmed the information from the supplier. Through the introduction of epoxy
compatible surface functionalities (such as carboxyls and hyroxyls [29]), solubility and reactivity

126

(a) (b)
Figure 5.1 TEM image of o-CNTs (a) and PGE-CNTs (b)

100 nm
100 nm

127

(a) Neat Epoxy (b). raw–CNT epoxy

(c). O-CNT epoxy (d). PGE-CNT epoxy

Figure 5.2 SEM images of fracture surface of NEs. All specimens have a CNT loading of
0.5vol%

5 micron
5 micron
5 X
5 micron
5 micron

128
compatibilization was attained between the MWCNTs and the matrix precursor [21]. CNT
surface modification improved nano-particle dispersion and facilitated the interface stress transfer
through the formation of covalent bonds, leading to the morphological and mechanical properties
discussed next.
5.3.2. Properties of the Nanotube Reinforced Epoxies (NEs)
SEM inspection of the NE composite fracture surface revealed that chemical
functionalization improved nanotube dispersion and interfacial bonding. (Fig.2). For epoxy with
raw-CNTs, CNT agglomerations at the length scale of micrometers adjacent to extended
CNT-devoid regions were observed (Fig.2b). Debonded raw-CNTs were also seen in the inset of
Fig.2b. Nanotube agglomerates, debonding, and telescopic pullout [30, 31] were absent
throughout the o-CNTs and PGE-CNTs fracture surface. MWCNTs appeared as separated white
dots, or high contrast spots with an average tube to tube distance of ~ 1-2 microns under SEM
(Fig.2 c, d). The inclusion of functionalized nanotubes altered the fracture surface features of neat
epoxy dramatically. Prominent parabolic features and “ductile” bands surrounding MWCNTs
(indicated by arrows in Fig.2c, d.) were observed in the o-CNT and PGE-CNT systems
respectively, in contrast to the fast fracture surface of the neat epoxy. These fracture surface
characteristics also suggested the activation of toughening mechanisms, such as crazing
retardation, crack deflection, and pinning of brittle fracture propagation [32].

129

Table 5.1 Glass transition temperatures (DSC) for NEs at different CNT loadings
0 vol%
T
g
(
o
C), 0.25vol%
MWCNTs
T
g
(
o
C), 0.5vol%
MWCNTs
T
g
(
o
C), 1.5vol%
MWCNTs
NE
Control
115.2 - - -
NE
RAW-CNT
- 114.3 113.7 113.5
NE
O-CNT
- 119.5 122.7 119.2
NE
PGE-CNT
- 120.8 126.6 122.1

130
Incorporation of less than 1.5vol% of MWCNTs into the epoxy matrix triggered
divergent changes in the glass transition temperature (T
g
) of the NEs. DSC measurements
revealed that T
g
increased with increasing MWCNT loading for both o-CNT and PGE-CNT
reinforced composites, although T
g
showed a decrease at 1.5vol% compared with 0.5vol% (Table
1). On the other hand, T
g
consistently decreased with nanotube concentration in raw CNT filled
NEs.
The trends observed in the measured T
g
values were attributed to the mobility of the
molecular chains or chain sectors in the NEs and were consistent with SEM observations. The
increase in glass transition temperature in o-CNT and PGE-CNT composites originated from
multiple causes. First, the incorporation of MWCNTs into polymer matrix confined the mobility
of near-interface moleculars, as reported for similar systems [33, 34]. Secondly, carboxyls and
beta-hydroxyls on the functionalized MWCNT surface increased the local cross-linking density
[35]. In contrast, the larger free volume resulting from CNT agglomeration and poor interface
affinity accounted for the decrease in T
g
observed for composites reinforced with raw CNTs.
Similarly, the nano-epoxies with 1.5vol% o-CNTs and PGE-CNTs both exhibited slightly
decreased T
g
values compared with lower loadings, and this was also attributed to the presence of
agglomerations at high CNT loading. The tensile modulus and fracture strength of the NE
specimens are illustrated in Figure 3 (each point represents the average of 5 test values). Linear

131
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
3
4
Tensile Modulus (GPa)
MWCNT Volume Fraction (vol%)
Raw -CNT
O-CNT
PGE -CNT

EP
EO
E
xx
(HT
xx
)
E
yy
(HT
yy
)
H-P

(a)
0.0 0.5 1.0 1.5
40
60
80
100
120
140
Tensile Strength at Break (MPa)
MWCNT Volume Fraction (vol%)
Raw-CNT
O-CNT
PGE-CNT

(b)
Figure 5.3 a. Experimental tensile stress vs. CNT volume fraction plotted with theorectical
predictions. b. Experimental tensile strength at break vs. CNT volume fraction

132
increases in composite tensile modulus with respect to nanotube volume fractions were observed
for all three systems. Although the composites with raw-CNTs and o-CNTs exhibited similar
increases in modulus, the elastic strength of the NEs with raw-CNTs decreased, an indication of
CNT agglomerations and poor CNT/epoxy interface stress transfer. The linearity of strength data
for both o-CNT and PGE-CNT was present for low CNT loadings and failed to hold at 1.5vol%.
Notably, PGE-CNT composites with 1.5vol% CNTs exhibited modulus and strength increments
of ~30% and ~ 60% respectively. With judiciously functionalized nano-fillers [35], PGE-CNT
reinforced epoxy presented the highest reinforcement efficiency of the samples tested.
The predicted composite moduli based on Halpin-Tsai (HT) micromechanics [36, 37]
were compared with the experimental values of the NEs. (Fig.3a ) A modulus of 700GPa [12] and
an aspect ratio of 100 for MWCNTs was assumed [27]; and the elastic modulus of the epoxy was
taken as 3.1 GPa. The observed modulus values for the NEs ranged between the values predicted
by Halpin and Pagano (H-P model for randomly oriented fibers, 1968) and those given by
Halpin-Tsai, which constitute a lower bond (E
yy
). Over-estimation of modulus values of
nanocomposites is not uncommon among recent reports [38, 39], and this variance may be
associated with the inherent nanotube waviness and agglomeration of CNTs, which was not
included in this model [40, 41]. The relevant equations are given below.
Halpin-Tsai model (E
xx
and E
yy
) E
xx or yy
=(1+ ζηV
CNT
)E
m
/(1- ηV
CNT
)
η = (E
CNT
/E
m
-1)/ (E
CNT
/E
m
+ ζ)

133

Figure 5.4 SEM images of basalt fiber fracture surface of MSRs with 0.5vol% CNT inclusions.

134
ζ = 2l/t + 40V
10
CNT
(for E
xx
) and 2w/t+40V
10
CNT
(for E
yy
)
Halpin-Pagano model (H-P) E
m’
=3E
xx
/8 + 5E
yy
/8
In the equations, V and E refer to the volume percentage and the modulus of the
individual phase under discussion. X and y refer to the directions parallel and perpendicular to
nanotube orientation respectively, l is the CNT length, w, t: width and height of the filler or CNT
diameter here, and CNT, m, and m subscripts’

denote the property for nanotubes, epoxy matrix,
and the NE respectively. Because predictions based on the HT model and the HP model fail to
provide an accurate estimate for the elastic moduli of NEs, the experimental data were linearly
fitted with the following empirical formulae.
Empirical formula for o-CNT (EO) E
m’
= V
m
E
m
+ 1/14.5V
CNT
E
CNT

Empirical formula for PGE-CNT (EP) E
m’
= V
m
E
m
+ 1/8.5V
CNT
E
CNT

5.3.3. Properties of the Multi-scale Reinforced Composite Laminates (MSRs)
SEM images of fracture surfaces revealed that the inclusion of 0.5vol% MWCNTs in
NE enhanced fiber-matrix adhesion and increased surface irregularity. (Fig.4) The incorporation
of MWCNTs resulted in fracture surfaces with a more “ductile” appearance compared to the
clean debonded surface characteristic of the control sample. For samples reinforced with
raw-CNTs, debonded regions were observed on basalt fiber surfaces (indicated by arrows in
Fig.4b), a result of the CNT-rich region and CNT-free regions in the raw-CNT filled NE (Fig.2b).

135
In contrast, basalt fibers were covered with an interfacial NE layer without any debonded region
for both o-CNT and PGE-CNT reinforced MSRs in all the fracture surfaces observed.
The o-CNTs and PGE-CNTs on the basalt fiber fracture surface revealed functionalized
CNT re-distribution and re-orientation during multiple processing stages employed during the
fabrication of MSR, including resin impregnation, drum-winding, and hot pressing. For example,
o-CNT nanocomposites exhibited a uniform NE layer with individually dispersed CNTs at the
basalt fiber fracture surface (Fig.4e). Based on the average 2-D inter-tube distance, the local
o-CNT loading at the interface was estimated to be as high as 5-6vol% (as opposed to CNT
loading of 0.5vol%). The morphological observations revealed the preferential distribution of
o-CNTs around basalt fiber and the formation of an interphase layer characterized by high CNT
loading. The high CNT loading in the interphase layer was induced by the CNT surface
modification and the processing conditions. In contrast, few PGE-CNTs were observed on the
basalt fiber fracture surface (Fig.4f).
Because of strong interfacial bonding between the PGE-CNT and the epoxy matrix,
CNT debonding along the sidewall at the CNT-matrix interface is not observed. The scarcity of
the PGE-CNTs observed on basalt fiber surface was explained by the PGE-CNT alignment along
the basalt fiber direction. Although the exact physics of the alignment is not well understood and
awaits more detailed investigation, CNT surface functionalities and shear flow within
micron-sized inter-fiber channels during fabrication must play important roles in aligning and

136

Table 5.2 Experimental Young’s modulus and tensile strength values for MSR specimens.
Epoxy/Basalt Laminates
Young’s Modulus (GPa)
% Improvement
Tensile Strength (MPa)
% Improvement
Control Laminate 0vol%
27.65±0.41 0 584.7±10.3 0
Raw CNT 0.5vol% 27.44±0.76 -0.76 564.0±31.0 -3.54
o-CNT 0.5vol % 30.41±1.31 +9.98 635.7±33.8 +8.72
PGE-CNT 0.5vol% 29.92±0.83 +8.20 608.8±20.0 +3.98
Raw-CNT 1.5vol% 28.53±1.18 +3.73 504.0±42.0 -13.8
o-CNT 1.5vol% 36.4±0.97 +31.65 627.7±25.5 +7.35
PGE-CNT 1.5vol% 34.90±0.77 +26.2 615.1±19.7 +5.20
ATT -Clay 1.25vol% 29.85±1.03 +7.96 530.9±43.4 -9.20
ATT -Clay 2.5vol% 28.06±2.62 +1.48 465.4±94.5 -20.4

137
dispersing the PGE-CNTs. Additional evidence for the alignment of PGE-CNTs comes from the
mechanical property measurements, described next.
The results of tensile tests for the multi-scale composites laminates are reported in
Table 2. Attapulgite, a rod like clay nano-particle (ATT-Clay), reinforced epoxies are also
included in the table for comparison. Increases in the elastic modulus of approximately 10% were
measured for the laminates with o-CNT or PGE-CNT modified epoxy at 0.225vol% overall
nanotube volume fraction (CNTs account for 0.5vol% of the NE, and the volume fraction of the
NE in the MSR is 45vol%.). The o-CNT and PGE-CNT reinforced laminates exhibited similar
linear increases in elastic modulus relative to CNT loading, irrespective of the chemical
modification. In contrast, no proportional strength increment was observed at 1.5vol% CNT
loading for both species, a result of the sensitivity of tensile strength to CNT dispersion. In
addition, the tensile strength of the o-CNTs reinforced laminates was consistently greater than
those with PGE-CNTs. This finding was attributed to superior fiber-matrix stress transfer that
resulted from the high CNT loading layer at the interface.
In order to predict the elastic moduli of the MSRs and evaluate the experimental values
reported above, a conventional micromechanics approach was employed. In this approach, the
elastic moduli for the NE matrices were adopted from the effective properties from the EO, EP,
and HP predictions. Subsequently, the Rule of Mixtures (ROM) and Inverse Rule of Mixtures

138
(IROM) was used to determine the elastic moduli of a single ply at different CNT loadings (vol%
of CNT in NEs).
E'
11
= V
f
E
f
+ V
m’
E
m’
(ROM)
E'
22
=( V
f
/E
f
+V
m’
/E
m’
)
-1
(IROM)
In the expression, 1 and 2 refer to the fiber direction and the direction perpendicular to
fiber orientation respectively. The basalt fiber loading for all samples was calculated to be
~55vol%; the elastic modulus for basalt fiber was taken as 85 GPa; The slash ', the subscripts C,
and

subscripts

f

denote the property of a single ply, composite laminate, and fiber respectively.
The elastic properties of the multi-scale composite laminates (denoted E
c
) are
calculated using a micromechanics formulation and assuming plane strain. A composite
reinforced with continuous unidirectional fibers in which all fibers are aligned in the 1 direction is
treated as a transversely isotropic material with three planes of symmetry.
For layers with fiber direction parallel to 1 (layer No.1, 4 in a 0/90/90/0 composite
laminate):
11 12 12
1
12 22 23
2
12 23 22
3
22 23
23 23
13 13
66
12 12
66
00 0
00 0
00 0
00 0 0 0
2
00 0 0 0
0 0 000
QQ Q
QQ Q
QQ Q
QQ
Q
Q
1
2
3
⎡⎤
σε
⎧⎫ ⎧ ⎫
⎢⎥
⎪⎪ ⎪ ⎪
⎢⎥
σε
⎪⎪ ⎪ ⎪
⎢⎥
⎪⎪ ⎪ ⎪ σε
⎢⎥
=
⎨⎬ ⎨ ⎬ −
⎢⎥
τγ
⎪⎪ ⎪ ⎪
⎢⎥
⎪⎪ ⎪ ⎪
τγ
⎢⎥
⎪⎪ ⎪ ⎪
⎢⎥
τγ
⎩⎭ ⎩ ⎭
⎢⎥
⎣⎦

139
At plane strain conditions:
3
23
12
0
0
0
σ=
τ=
τ=

We have:
111 12 1
212 22 2
12 66 12
0
0
00
QQ
QQ
Q
σε
⎧⎫ ⎡ ⎤⎧ ⎫
⎪⎪ ⎪ ⎪
⎢⎥
σ= ε
⎨⎬ ⎨ ⎬
⎢⎥
⎪⎪ ⎪ ⎪
⎢⎥ τγ
⎩⎭ ⎣ ⎦⎩ ⎭

Filling in the elastic constants for transversely isotropic materials:
11 12 22
11
12 22 22
22
12 12
12
''
0
''
0
00
EE
DD
EE
DD
G
υ
⎡⎤
⎢⎥
σε
⎧⎫ ⎧ ⎫
⎢⎥
υ ⎪⎪ ⎪ ⎪
⎢⎥
σ= ε
⎨⎬ ⎨ ⎬
⎢⎥
⎪⎪ ⎪ ⎪
τγ ⎢⎥
⎩⎭ ⎩ ⎭
⎢⎥
⎢⎥
⎣⎦

Similarly, for layers with fiber direction perpendicular to 1 (or layer No. 2, 3):
22 12 22
11
12 22 11
22
12 12
12
''
0
''
0
00
EE
DD
EE
DD
G
υ
⎡⎤
⎢⎥
σε
⎧⎫ ⎧ ⎫
⎢⎥
υ ⎪⎪ ⎪ ⎪
⎢⎥
σ= ε
⎨⎬ ⎨ ⎬
⎢⎥
⎪⎪ ⎪ ⎪
τγ ⎢⎥
⎩⎭ ⎩ ⎭
⎢⎥
⎢⎥
⎣⎦

12 1 D 21 =−υ υ = ~ 0.945 ( υ
12,
υ
21
were estimated based on reference 42, 43)
Because of the symmetry of the multi-scale reinforcement composite laminae
(symmetric balanced cross-ply), the calculation of the laminate elastic moduli at longitudinal

140
direction (or direction 1) is reduced to the following relationship when the strain in the 2 direction
is small. A similar result can be deduced for the transverse direction:
E
c 11
= E
c 22
= (E’
11
+E’
22
)/2D
The E
c
vs. CNT loading relationship is illustrated by the measured values plotted in
Fig.5a. The curves plotted are the predictions from the micromechanics model, which
significantly under-estimates the elastic properties for fiber-dominated directions (both E
c11
and
E
c22
for cross-ply laminates). Conventional wisdom dictates that MWCNT reinforcements should
be most effective in boosting matrix-dominated properties, such as interlaminar shear modulus,
G
IC
, and G
IIC,
particularly when used as nano-scopic additives in fiber-reinforced composites [44,
45]. However, current predictions vastly underestimate the ~ 20% per vol% (vol% of CNT in NE)
increase observed in the elastic modulus of the composite laminates. Therefore, additional
reinforcing mechanisms must be considered.
The o-CNT re-distribution and the PGE-CNT alignment observed under SEM (Fig 5: b,
c) provided insight into the underlying causes of he discrepancies between the measured modulus
values and predictions from the micromechanics model. The annular interphase regions aroung
basalt fibers (with high o-CNT loadings) provides a high modulus layer around basalt fibers, the
elastic modulus of which can be as high as 10~20 GPa (H-P, Exx) based on the estimation of
5vol% CNT concentration on basalt fiber surface. A higher interfacial shear modulus (~1/3 of the
elastic modulus) also improves the interfacial stress transfer properties between the NE and basalt

141
-0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
26
28
30
32
34
36
38
Control
EP
EO
HT
O-CNT
PGE-CNT
Raw-CNT
E
c11
or E
c22
(GPa)
MWCNT loading (vol%)
EO Prediction
EP Prediction
H-P Prediction

Figure 5.5 a. Ec11 vs. CNT volume loading: measured values and predicted curves. b. CNT
redistribution around basalt fiber. c. CNT alignment along basalt fiber orientation

(a)
(b)
(c)

142
fibers and, consequently, enhances the elastic modulus and strength. On the other hand,
micromechanical predictions are consistent with the alignment of PGE-CNTs along the direction
of basalt fibers indicated by the SEM observations discussed previously. An effective matrix
elastic modulus of ~4.7 GPa can be deduced based on 0.5vol% of aligned and straight CNTs
(E
xx
). Using the same micromechanics model employed above, the estimated E
c11
(at 0.5vol%) of
~29.3 GPa is within the experimentally measured range of longitudinal elastic modulus values for
PGE-CNT reinforced MSR (Fig.5a). Thus, the unusually high reinforcement efficiency of the
functionalized MWCNTs for fiber dominated properties of multi-scale reinforced composite
laminates is attributed to the re-distribution or re-orientation of CNTs during processing
conditions.
5.4. Conclusions
Appropriate choice of surface groups led to uniform dispersion and covalent integration
of MWCNTs into a DGEBA epoxy system. Tensile tests of the bulk NEs revealed increases in
both elastic modulus and strength, albeit lower than theoretical predictions due to natural CNT
waviness and agglomeration.
Direct incorporation of functionalized MWCNTs into epoxy/basalt composite
laminates improved the elastic properties of the composites, particularly in the fiber direction, and
this increase was not predicted by classical micromechanics prediction based on the equivalent
matrix properties obtained from NE specimens. SEM observations identified CNT re-distribution

143
and re-orientation resulting from processing. When properly positioned and oriented within the
matrix, carbon nanotube reinforcement markedly influences properties that are usually dominated
by fiber reinforcements. In-situ alignment of carbon nanotubes and formation of a distinct,
hardened, annular interphase around basalt fibers were instrumental in attaining the high
reinforcement efficiency in the fiber dominated direction of these hierarchical composites.

144
Chapter 5 References
[1] Sgriccia N, Hawley MC. Compos Sci Technol 2007; 67 (9): 1986-91.

[2] Silva C, Marotta E, Schuller M, Peel L, O'Neill M. J Thermophys Heat Transf 2007; 21
(3):460-7.

[3] Ogihara S, Uehara T. J Mate Sci Lett 2003; 22 (9): 683-5.

[4] Park JS, Park SS, Lee S. Adv Polym Emerg Technol 2007: 568-72.

[5] Huang L, Wang C, Lu, YF. J Reinf Plast Compos 2008; 27(7): 725-38.

[6] Dilandro L, Dibenedetto AT, Groeger J. Polym Compos 1988; 9(3):209-21.

[7] Boogh LCN, Meier, RJ, Kausch HH, Kip BJ. J Polym Sci B 1992; 30(4):325-33.

[8] Mathur RB, Chatterjee S, Singh BP. Compos Sci Technol 2008; 68(7-8): 1608-15.

[9] Bekyarova E, Thostenson ET, Yu AP, Itkis ME, Fakhrutdinov D, Chou TW, Haddon RC. J
Phys Chem C 2007; 111(48): 17865-71.

[10] Dean D, Obore AM, Richmond S, Nyairo E. Compos Sci Technol 2006; 66(13): 2135-42.

[11] Kim YA, Kamio S, Tajiri T, Hayashi T, Song SM, Endo M, Terrones M, Dresselhaus, M.S.
Appl Phys Lett 2007; 90(9):93125-1-3.

[12] Yu MF, Lourie O, Dyer MJ, Moloni K, Kelly TF, Ruoff RS. Science 2000; 287(5453):
637-40.

[13] Bekyarova E, Thostenson ET, Yu A, Kim H, Gao J, Tang J, Hahn HT, Chou TW, Itkis ME,
Haddon RC. Langmuir 2007; 23(7):3970-4.

[14] Thostenson ET, Li WZ, Wang DZ, Ren ZF, Chou TW. J Appl Phys 2002; 91(9):6034-7.

[15] Qiu JJ, Zhang C, Wang B, Liang R. Nanotechnology 2007; 18(27):275708.

[16] Gojny FH, Wichmann MHG, Fiedler B, Bauhofer W, Schulte K. Composites: Part A 2005;
36(11):1525-35.

145
[17] Veedu VP, Cao AY, Li XS, Ma KG, Soldano C, Kar S, Ajayan PM, Ghasemi-Nejhad MN.
Nat Mate 2006; 5(6): 457-62.

[18] Garcia EJ, Wardle BL, Hart AJ. Composites: Part A 2008; 39:1065-70.

[19]Thostenson ET, Ren ZF, Chou TW. Compos Sci Technol 2001; 61(13):1899-912.

[20]Jonathan N, Coleman, UK, Yurii K, Gun K. Adv Mater 2006; 18:689-706.

[21] Paiva MC, Zhou B, Fernando KAS, Lin Y, Lopes PE, Pennington WT, Kennedy JM, Sun
YP. Compos Inter 2005; 12(8-9):757-68.

[22] Frankland SJV, Caglar A, Brenner DW, Griebel M. J Phys Chem B 2002;106: 3046-48.

[23] Miaudet P, Badaire S, Maugey M, Derre A, Pichot V, Launois P, Poulin P, Zakri C. Nano
Letters 2005; 5(11): 2212-15.

[24] Camponeschi E, Vance R, Al-Haik M, Garmestani H, Tannenbaum R. Carbon 2007; 45(10):
2037-46

[25] Hilding JM, Hong MPL, Grullke EA. J Nanosci Nanotechnol 2005; 5(5):742-6.

[26] Chen W, Auad ML., Williams RJJ, Nutt SR. Euro Polym J 2006; 42: 2765-72.

[27] Liu J, Rinzler AG, Dai H, Hafner JH, Bradley RK, Boul PJ, Lu A, Iverson T, Shelimov K,
Huffman CB, Rodriguez-Macias F, Shon Y, Lee TR, Colbert DT, Smalley RE. Science 1998;280:
1253-1256

[28] Fisher FT, Eitan A, Andrews R, Brinson LC, Schadler LS. Compos Sci Technol
2006;66(9):1159-70.

[29] Ellis B. Glasgow UK: Blackie Academic & Professional; 1993, Ch.3.

[30] Kuzumaki T, Mitsuda Y. Carbon 2005; 43(10): 2227-9.

[31] Cumings J, Zettl A. Phys Rev Lett 2004;93 (8):086801-1-4.

[32] Xia Z, Riester L, Curtin WA, Li H, Sheldon BW, Liang J, Chang B, Xu JM. Acta Mater
2004; 52(4): 931-44.

146
[33] Gojny, FH, Schulte K. Compos Sci Technol 2004; 64(15):2303-08.

[34] Wang JG, Fang ZP, Gu AJ, Xu LH, Li, F. J Appl Polym Sci 2006; 100(1):97-104.

[35]Chen W, Lu HB, Nutt SR. Compos Sci Technol 2008; Article in press. DOI information:
10.1016/j.compscitech.2008.05.011.

[36]Gojny FH, Wichmann MHG, Kopke U, Fiedler B, Schulte K. Compos Sci Technol
2004;64(15): 2363-2371.

[37] Li XD, Gao HS, Scrivens WA, Fei DL, Xu XY, Sutton MA, Reynolds AP, Myrick ML. J
Nanosci Nanotechnol 2007; 7(7): 2309-17.

[38] Garcia EJ, Hart AJ, Wardle BL, Slocum AH. Adv Mater 2007; 9(16):2151-56.
[39] Thostenson ET, Chou TW. J Phys: D 2003; 6(5):573-82.

[40] Fisher FT, Bradshaw RD, Brinson LC. Appl Phys Lett 2002;80(24):4647-49.

[41] Anumandla V, Gibson RF. Compos Part A: Appl Sci Manuf 2006;37: 2178-85.

[42] Shan HZ, Chou TW. Compos Sci Technol 1995; 53(4):383-91.

[43] Mead DJ, Joannides RJ. Composites 1991;22(1):15-29.

[44] Khabashesku VN, Barrera EV, Zhu J, Imam A, Crane R, Lozano K. Compos Sci Technol
2007; 67(7-8):1509-17.

[45] Wichmann MHG, Sumfleth J, Gojny FH, Quaresimin M, Fiedler B, Schulte K. Engi Fract
Mech 2006;73(16):2346-59.

147
Chapter 6. Towards multiple functionalities
6.1. Motivation
Underwater sonar windows are designed for acoustic transparency and structural
integrity, and desirable material characteristics include durability, ease of processing, and
consistency. However, the primary design criterion for window materials is sound speed and
density that closely match those of seawater [1]. In addition, the material design should
minimize dynamic shear modulus, dilatational loss factor, insertion loss, and maximize the shear
loss factor, static shear modulus, tensile modulus, tensile strength, and toughness at the relevant
temperature and frequency range (0-50
o
C, 1-100KHz).
The available designs of sonar windows have various limitations. For example,
sandwich structures are often considered for underwater acoustic applications [2]. However, the
acoustic transparency of a sandwich construction can be tuned only to a limited range of
frequencies and is accompanied by high insertion loss. This in turn requires a unique design for
each application with trade-offs to accommodate the pertinent critical angles of incidence.
Consequently, a macroscopic uniform and castable material presents processing and economic
advantages over sandwich designs. Thus, non-structural synthetic elastomers, such as
polychloroprene rubber, EPDM rubber, and various polyurethanes have been employed in
underwater acoustics with sound speed and density that matches water (~1500 m/s, 1000kg/m
3
).
Such elastomers, however, are compliant and have poor structural rigidity. Alternatively,

148
structural thermosets, such as epoxies, have also been explored as the base material for
underwater sonar windows [3-5]. For example, DGEBA epoxies (~2400 m/s) or fluoroepoxies
(~1500-2000 m/s) have been compositionally modified with liquid rubber additives, glass
microspheres, polymeric micro-balloons, and compliant epoxy pre-polymers to adjust for acoustic
and structural properties. These epoxy based composites either suffer from low rigidity and/or
high shear resonances, which significantly hamper the direct application for sonar windows. Such
limitation highlights the need for an acoustically transparent material or structure that is inclusive
of all frequency ranges and insertion angles, while providing structural performance.
Recent reports have shown that carbon nanotubes (CNTs) are both efficient
reinforcement materials [6-8] and excellent dampers for shear resonance modes in polymer
matrices [9, 10], and they may thus offer a means to address both of the issues outlined above. To
pursue this possibility, we have produced thermoplastic toughened epoxy mixtures filled with
milled glass fibers, polymeric micro-balloons, and oxidized multiwalled carbon nanotubes
(o-MWCNTs) in an attempt to design a castable, macroscopically uniform, inherently damped,
and acoustically transparent structural material. The resulting composite exploits the
multi-functionality of each ingredient and the flexibility it offers to achieve conflicting design
criteria.

149
6.2. Theory
In acoustically transparent materials, or structures, acoustic energy can enter and pass
through the materials, or structures, without significant losses due to reflection and attenuation.
For underwater applications in which incident sound waves impinge perpendicular to the surface,
formulations for acoustic transparency in water require that the product of the density and
longitudinal wave speed, or acoustic impedance Z
0
(Eqn. 1), matches that of water. However, for
applications that require an insertion angle other than 0
o
, density and sound speed should be
matched with water individually, a requirement that is expressed as
Z
0
= ρ×c Equation 1
Z
0
: acoustic impedance
ρ: density
c: wave speed
For acoustically transparent structures, inherent damping that minimizes radiated
energy without affecting sound transmission is also essential. A high dilatational loss factor,
increases transmission loss (TL) and thus must be minimized. However, increasing the shear
modulus loss factor dramatically reduces the amplitude of the TL peaks. Hence, the real
challenge in the design of sonar windows is to maximize the shear modulus loss factor while
maintaining a low dilatational loss factor. This allows the material to damp out the shear waves
generated and also minimize destructive interference between shear waves and longitudinal
waves.

150
CNTs exhibit high aspect ratios and superb mechanical properties [11], and are thus
considered as ideal reinforcement materials [12, 13]. On the other hand, for acoustic
considerations, conventional fillers such as glass fibers and phenolic micro-balloons can either
increase modulus at the cost of increasing wave speed or decrease the modulus while also
reducing the wave speed. Because of their small size, CNTs can exert relatively small effects on
acoustic waves at the frequency ranges of interest. Furthermore, carbon nanotubes additions
inherently dampen the shear modes of the material by improving the shear loss factor without
altering the dilatational loss factor through a stick-slip mechanism at carbon nanotube-polymer
interface [9, 10]. Suhr and coworkers [9] pointed out that multi-walled carbon nanotubes
(MWCNTs) increase the shear modulus loss factor up to 1400% at low frequencies without
sacrificing the storage modulus. Carbon nanotubes and carbon nanofibers were also reported to
enhance damping effectively at higher frequencies and in stiffer matrices [14, 15].
6.3. Formulation
Acoustic transparency can be engineered by inclusion of CTBN [3], aliphatic epoxy
pre-polymer [4], and phenolic micro-balloons [1, 4] to reduce wave speed and density, and
subsequently adjusting the structural properties by the addition of reinforcing fillers such as glass
fibers and o-MWCNTs. While the addition of milled glass fibers will increase the sound speed,
density, and modulus of the composites, the addition of compliant components, specifically,

151
Table 6.1 Formulation of Test Samples

EPON828
(phr)
DER732
(phr)
2,4 -EMI
1
(phr)
CTBN
(phr)
MWCNT
2
(wt%)
Phenoset

(wt%)
Glass
Fiber

(wt%)
Nano-cla
y

(wt%)
N
0
100 0 3 -- -- -- -- --
A 100 0 3 20 -- -- -- --
B 100 0 3 20 0.05 -- -- --
C 75 25 3 20 -- -- -- --
D 75 25 3 20 0.05 -- -- --
E 75 25 3 20 0.50 -- -- --
F 60 40 3 20 -- -- -- --
G 60 40 3 20 0.05 -- --
H 60 40 3 20 0.05 2 -- --
I 60 40 3 20 0.05 2 -- 2.5
J 60 40 3 20 0.05 2 2 2.5
Note:
1
2, 4 -EMI was used as curing agent for all system listed.
2
MWCNTs were oxidized before use.

152
rubber based additives (CTBN), and the phenolic micro-balloons, will have a compensating
effect.
In the present work, formulations were based on the above principles and the key
samples are listed in Table 1 as A through H, with control sample N
0
. Three approaches were
employed to control the wave speed of the DGEBA epoxy system. First, the base epoxy backbone
was varied from rigid to compliant by varying the proportions of the aromatic diglycidyl ether of
bisphenol-A (DGEBA) epoxy monomer, EPON828, and the aliphatic monomer, DER 732. The
rigidity of the epoxy sample was varied by blending the compliant DER732 aliphatic epoxy at 0,
25, and 40 parts per hundred resin (phr) to DGEBA. The rigid systems used 100 phr EPON828,
while semi-rigid systems used 75 phr EPON828 and 25 phr DER 732. The compliant systems
used 60 phr EPON828 and 40 phr DER 732. Each of the formulations also included an
elastomeric modifier (carboxyl terminated butadiene acrylonitrile, CTBN) to reduce the wave
speed and increase toughness of the epoxy systems [3]. CTBN dissolved in the epoxy resin when
mixed at elevated temperature, although during subsequent curing, it phase-segregated into
micron-sized spherical rubbery domains, which enhanced the toughness and elongation of the
resulting material. Finally, phenolic micron-balloons with an average density of 314 kg/m
3

(Phenoset, supplier), were added to adjust the density and meet the design criteria. According to
Thompson [1], the addition of polymeric micro-balloons will decrease composite density
substantially and decrease wave speed moderately.

153
Milled glass fibers (1/16”) and o-MWCNTs were added to increase the strength and
modulus of the polymer composites. The reinforcing behavior of short-fiber-reinforced polymer
composites can be estimated by the Halpin-Tsai or Mori-Tanaka model [16, 17]. Glass fibers
increase the modulus of the composites at the cost of higher acoustic wave speed, similar to that
observed for glass micro-spheres [1]. Oxidized MWCNTs were added in samples B, D, E, G, H, I,
and J. The oxidization process introduced carboxylic, carbonyl, and hydroxyl groups on to
MWCNT surface. These surface functionalities provided cross-linking sites that enhanced the
interfacial stress transfer properties at the MWCNT/matrix interface, while simultansiouly
enhancing the dispersion of the nano-fibrils. In a parallel study conducted in this work, the role of
o-MWCNTs as a non-intrusive reinforcement for underwater acoustics was investigated utilizing
the base epoxy formulation (N
0
) filled with o-MWCNT of 0.01, 0.025, 0.05, 0.075, 0.1, 0.5, 1,
and 2 wt% respectively.
The surface-modified nano-clay garamite was used as an anti-settling agent (ASA) to
prevent phase separation and microsphere flotation or migration by increasing the viscosity or
enhancing shear-hardening properties of the mixture. The nano-clay also served as a
reinforcement component in the formulation.
All samples were prepared in a dual-axis, high-speed mixer (Keyence, HM501) to
provide high shear force and attain uniform materials dispersion. Three-minute degassing was
performed in the same mixer after each mixing step to remove air bubbles generated during

154

(a) (b)
Figure 6.1 Pulse-echo method (a) and free-free beam measurement (b) for longitudinal wave
speed test.

Path 1
Path 2

155
mixing. The following general sample preparation procedures were employed: First, DER732
was mixed with EPON828 for 5 min, and then CTBN was introduced at 150
o
C and mixed for
minimum 1 hr to attain a uniform mixture.The mixture was then cooled to room temperature.
Afterwards, milled oxidized MWCNTs were added and the resultant mixture was sonicated for 30
minutes (Aquasonic, 150D, VWR Scientific) and mixed for 5 min. Subsequently, nano-clay,
phenolic micro-balloons, glass fibers, and the curative were mixed (5 min) and degassed (3 min)
separately in that order. Finally the mixture was cast into an aluminum mold pretreated with
releasing agent and cured. All samples were cured at 65
o
C for 24 hours and post-cured at 120
o
C
for 3 hours. Small specimens were cut for subsequent measurements. The final scaled-up
sample (45.72cm x 45.72 cm x 1.27 cm) was made in sub-batches, and each batch weighed
approximately 170g due to the sample size limitation of the hybrid mixer.
6.4. Experiments & model
Pulse-echo measurements (Fig.1a) were performed in a standard NDT ultrasonic
immersion tank at 5 MHz to estimate the longitudinal wave speed of the material at 25
o
C in
fresh water. For samples with phenolic microspheres, pulse-echo tests were not viable due to the
absorption of frequencies ~5 MHz associated with the microspheres. Therefore, a free-free
excitation set-up was also used for longitudinal wave speed measurement, as shown in Fig. 1b.
The free-free beam measurements were used to excite column resonances for determination of the

156
compressional wave speed and agreement between these two methods were established for
samples tested with both approaches.
Clamped-free slender beam excitation was used to determine values of elastic and
shear modulus. The complex moduli were based on the relations:

The average for the lowest 4 to 5 modes was used to determine the loss factors. The
elastic and shear loss factors were determined by measuring the bandwidths of the individual
bending and torsional beam resonances respectively. The Young’s modulus and shear modulus
used in Eqn. 2, 3 were estimated using the following relationships:
Young’ s modulus:
Equation 4
Shear modulus:
Equation 5

()
2
2
2
4
12
⎥
⎥
⎦
⎤
⎢
⎢
⎣
⎡
⎟
⎠
⎞
⎜
⎝
⎛
=
n
n
l
a
l
E
β
ω
ρ
()
()
1
2
2
2
2 2 2 2
1 2 3 k b n
b a l
G
n
−
+
=
π
ω ρ
()
() 281 . 0
2
1 2
Factor Shape k
n
l
Density
Width b
Thickness a
Length l
n
=
⎟
⎠
⎞
⎜
⎝
⎛ −
=
=
=
=
=
π β
ρ
()
()
G
E
i G G
i E E
η
η
+ =
+ =
1 '
1 '
factor loss shear
factor loss elastic
G
E
−
−
η
η
Equation 2
Equation 3

157

Table 6.2 Wave speed and Static Mechanical Properties of the Samples
Flexural Properties
(MPa) Sample Wave Speed (m/s)
Modulus Strength
Density (kg/m
3
)
N
0

2480 2800 78 1158
A
2200 734 56 1112
B
2130 1126 45 1109
C
1910 92 8 1099
D
1935 430 15 1101
E
1922 97 12.5 1117
F
1799 34 2 1093
G
1783 325 6 1095
H
-- -- -- --
I 1702 462 9.5 920
J
1750 600 14 980

158
The
n
ω values used in Eqn. 4, 5 were typically for the first mode. The expressions
shown for E and G do not account for rotary inertia and shear-deformation effects. These effects
become increasingly noticeable as the mode number increases.
The flexural tests were performed on a universal testing machine using a three-point
bending fixture to obtain the flexural modulus and flexural strength values according to the
ASTM D790M at a strain rate of 1 s
-1
. Five beam-shaped specimens were cut and polished for
each test. Densities of the samples were measured with a pyconometer.
Dynamic mechanical analyses (DMA, TA Instruments) were performed to monitor the
changes of the glass transition temperature (T
g
), storage moduli (E’), and loss modulus (E”) in the
composites from 25 °C to 150 °C at a heating rate of 5 °C per minute and load frequency of 1 Hz.
The acoustic performance of the scaled-up panel was tested at the underwater insertion
loss facility of the US Navy’s NUWC Division at Newport, RI (APTF). The insertion loss of
the panel was measured at a distance of 2 meters at the centerline of the test tank in 20
o
C fresh
water. Insertion loss was scanned between 10 and 100 KHz at normal incidence (0
o
) and at both,
15
o
, and 30
o
off normal incidence.
The acoustic performance for sonar windows was defined in terms of a plane wave
Transmission Loss (TL) and was predicted using a finite element model of a submerged, planar,
isotropic, viscoelastic layer of infinite extent. The following input parameters were used for the
window material based on measurements of the final sample (J). The window thickness was taken

159
as 1.27 cm, the same as the final scaled-up panel. The nominal density and sound speed of water
was taken to be 1000 kg/m
3
and 1500 m/s respectively.
6.5. Results & Discussion
6.5.1 Wave speed reduction and density adjustment
The measured wave speed, density, and flexural properties are shown in Table 2.
Because of the buoyant migration of phenolic micro-spheres, test results for sample H were not
recorded. The incorporation of liquid rubber additive (CTBN), compliant aliphatic epoxy
pre-polymer, and phenolic micro-spheres reduced the wave speed (and density) to values
comparable to fresh water (~1500m/s).
The CTBN-toughened specimen B (20 phr) exhibited a 4% decrease in density and an
11.3% decrease in wave speed compared with the control specimen N
0
. Ramotowski [3] reported
that CTBN additions also increase both toughness and elongation of the composite. Although
flexural properties decreased, the resulting sample exhibited acoustic properties superior to other
castable polymers of similar modulus at room temperature, a factor attributed to the acoustic
impedance matching with water.
Blends containing higher concentrations of DER732 were more flexible. A reduction in
wave speed of 100-120 m/s for every 10 phr of flexible epoxy inclusions was observed for all
samples affected. Sample density reduction was also negligible. Blends rich in DER732 may be
suitable as potting materials but lack the mechanical strength and rigidity required for sonar

160
window applications. However, an efficient non-intrusive reinforcing material would make these
formulations suitable for acoustic window applications.
The acoustic impedance can be more closely matched by reducing density through the
addition of phenolic micro-balloons. While the addition of micro-balloons altered the density
match with water and slightly decreased the wave speed and mechanical properties [1, 4], the
combined effect of phenolic micro-balloon and nano-clay effectively reduced composite density
while enhancing the mechanical properties by 30-50%.
6.5.2 Reinforcement effect of fillers
Although the composite wave speed and density were effectively reduced by the
additions of CTBN, aliphatic epoxy, and phenolic microspheres, the mechanical properties were
degraded. Thus, milled glass fiber, nano-clay, and o-MWCNTs were employed to enhance the
mechanical properties, resulting in different reinforcing efficiencies. Interactions between fillers
were assumed to be negligible in terms of the reinforcing effects.
Garamite nano-clay was included in formulations I and J as an anti-settling agent, and
caused an increase strength and stiffness, an effect opposite to the effect of adding phenolic
micro-balloons. Both the flexural modulus and strength increased 30-50% for specimen I
(compared with sample G), a phenomenon attributed to the inclusion of 2wt% of surface
modified nano-clays. However, the longitudinal wave speed does not conform to the square root
rule of the wave speed – modulus correlation (Eqn. 6) based on specimens G and I, and this effect

161

0.0 0.5 1.0 1.5 2.0
1600
1800
2000
2200
2400
2600
2800
3000
82
84
86
88
90
92
94
96
98
Storage Modulus

Storage Modulus (MPa)
o-MWCNT weight percenrage (wt%)
Glass Transition Temperature
Glass Transition Temperature (
o
C)
0.0 0.5 1.0 1.5 2.0
0
500
1000
1500
2000
2500
3000
3500
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Longitudinal wave speed

Longitudinal wave speed (m/s)
o-MWCNT weight percentage (wt%)
Density

Density (g/cm
3
)

Figure 6.2 Test results of neat MWCNT reinforced epoxy composites with Various CNT Filler
Concentration.

162

CD E
50
100
150
200
250
300
350
400
450
500
0
100
200
300
Flexural Modulus
Flexural Strength

Flexural Modulus (%)
Sample Denominations
Flexural Strength (%)

Figure 6.3 Flexural Modulus & Flexural Strength of epoxy elastomer blends.

163
was attributed to the nano-scopic clay reinforcement used. The classical relationship described in
Eqn. 6 is not suitable for heterogeneous materials with nano-scale reinforcements.
c = (E/ ρ)
1/2
Equation 6
E: modulus
ρ: density
Milled glass fibers (1/16”) were also used to increase the mechanical properties in
sample J. The increase in stiffness observed for J compared to I is attributed to the addition of 2%
glass fibers. The wave speed of the composite also increased by 20-25 m/s per wt%, an effect
similar to that observed for glass micro-spheres [1, 18]. In addition, the density of sample I was
also increased slightly to provide a closer match to fresh water.
The longitudinal wave speed, density, glass transition temperature (T
g
), and storage
modulus were measured and compared for each sample to determine the role of o-MWCNTs. As
shown in Fig. 2, o-MWCNT additions provided additional reinforcement without affecting the
longitudinal wave speed. The storage modulus increased ~20%, while the longitudinal wave
speed was largely unaffected. The behavior of composites with o-MWCNTs deviated from Eqn. 6
(just as with nano-clay systems), and the deviation was attributed to vast difference in length
scales for MWCNTs compared with the acoustic wavelengths. Both the storage modulus and
the glass transition temperature peaked at 0.05wt% for oxidized MWCNT (Fig. 2a). At higher
nanotube loadings, the decrease in both storage modulus and glass transition temperature resulted

164
Table 6.3 Final Sample Properties
Wave Speed 1750 m/s
(estimate from free-free beam excitation))
Density 980 kg/m3
Tg Below room temperature (25
o
C)
Static Modulus 600 MPa (3-point bending)
Strength 14 MPa (3-point bending)
Failure Strain 11% (3-point bending)
Elastic Modulus 1290 MPa
(Clamp-free slender beam excitation)
Shear Modulus 440 MPa
(Clamp-free slender beam excitation)
Dilational Loss Factor 0.048
(Clamp-free slender beam excitation)
Shear Loss Factor 0.049
(Clamp-free slender beam excitation)
Poisson’s Ratio 0.45
(Clamp-free slender beam excitation)

165

Figure 6.4 Measured Insertion Loss at Various Angles of Incidence.

166
from o-MWCNT agglomeration and the exponential increase in viscosity during processing [6, 7].
The dispersion of o-MWCNTs, and thus the reinforcement efficiency, was partly limited by the
processing methods. With optimized dispersion techniques, carbon nanotube additions are
expected to cause the modulus and strength of the composites to increase by orders of magnitude
[8] without affecting the longitudinal wave speed.
The results of samples A through J showed that o-MWCNTs imparted greater
reinforcement efficiency for matrices with lower modulus. For example, in the stiff epoxy
systems (sample A, B), addition of 0.05wt% o-MWCNT increased the flexural modulus and
strength by 53% and 19%. On the other hand, in the semi-rigid systems (C, D, and E), the
addition of 0.05 wt% o-MWCNT caused the modulus to increase by 370% and the strength to
increase by 90% (Fig. 3). However, composites with higher CNT content (0.5wt% o-MWCNT,
sample E) exhibited a reduction in flexural strength and modulus relative to specimens with lower
nanotube loadings. Finally, the greatest reinforcement efficiency was observed for the most
compliant matrices (G, I, J) where both strength and modulus values increased several-fold after
additions of 0.05wt% o-MWCNT.
The properties of the scaled-up sample are listed in Table 3. The wave speed of the
final composite was reduced by 30% from the original DGEBA epoxy composition, while the
density was 980 kg/m
3
, nearly matching that of fresh water. The mechanical properties of the final
sample increased significantly compared with sample F. Further improvements in dispersion,

167
alignment, and interface control of CNT-reinforced composites will be key elements in future
research in order to realize the full reinforcing effects of CNTs.
6.5.3 Transmission Loss
The measured insertion loss data on the final scaled-up panel at 0, 15, and 30 degrees
of incidence are illustrated in Fig. 4. Insertion loss at 0
o
and 15
o
incidence increased linearly with
frequency and was less than 3 dB over the frequency range 10-100 KHz for 0 (normal), 15, and
30 degree incidence. The drop in insertion loss between 45-50 KHz is a testing artifact. The
results show an insertion loss peak at 25 KHz that increases as the angle of incidence deviated
from the normal incidence. At 30 degree incidence, the peak occurs at 25 KHz, which is the
frequency at which the panel thickness measures one-half the shear wavelength. The frequency of
the insertion loss peaks is dependent on the panel thickness, the density match to the water
medium, the boundary conditions of the panel and the lateral dimensions of the panel. The
magnitude of this shear resonance peak is controlled by the shear modulus of the composite, the
panel thickness and the angle of incidence. The increase in peak magnitude with angle of
incidence is expected, because for any given panel, the effective thickness of the panel increases
with incidence angle. Consequently, more incident energy is converted into shear modes when the
effective thickness of the panel is higher. This results in more prominent insertion loss peaks at
higher angles of incidence.

168
The insertion loss (IL) of the panel was simulated as a function of frequency at
different angles of incidence. Because the trends indicated by IL were qualitatively equivalent to
the TL trends and TL analytical tools were readily available, TL predictions were used to study
the role of CNT damping in enhancing acoustic transparency. The analytical model for predicting
the Transmission Loss (TL) of the panel was two-dimensional and planar. The window material
was embedded in an infinite acoustic medium and represented as a homogeneous and isotropic
viscoelastic layer of given thickness and infinite extent. It was thus defined by its thickness, its
mass density and two elastic (Lame's) constants, taken to be complex to model dissipation in the
material. A plane, harmonic, acoustic wave, pi(q;w), was assumed to be incident on the panel at
an arbitrary angle (q) to the panel normal, and the portion of the wave transmitted across the
panel, pt(q;w), was computed. In turn, TL values were defined as:
TL=-20log|pt(q;w)/pi(q;w)| Equation 7
Results of TL predictions from the analytical model are plotted in Fig. 5 from 1-40 kHz, from
normal incidence ( 0 = θ
o
) to near grazing (
o
90 → θ ).The TL values are less than 1 dB for
frequencies less than 10 KHz for all incidence angles below 60
o
. The TL also peaked at 25
KHz, which coincides with the measured IL data. Thus, the predicted TL qualitatively agrees with
the insertion loss measurements of the fabricated panel. The qualitative similarity serves as partial
validation of the model and establishes the utility for predicting the TL trends of
newer/hypothetical formulations.

169

Figure 6.5 Computed Transmission Loss of the Window with A) Measured Shear Loss Factor
and B) Assumed Shear Loss Factor due to Carbon Nanotube Damping.

170
The shear resonance associated with the panel thickness can be ameliorated by
increasing the shear loss factor of the material. (For materials with Poisson’s ratio one, this
increase in damping will somewhat degrade performance at other frequencies that are influenced
by shape factors other than panel thickness.) The introduction of carbon nanotubes can directly
affect the shear loss factor, thus offering a possible solution to this design dilemma. Current
measurements of the shear loss factor of the final sample are close to the dilatational loss factor, ~
0.048, and this effect is attributed to the additions of CTBN and phenolic micro-balloons, not
CNTs. The reason is two-fold: First, shear damping occurs only selectively under a stick-slip
mechanism, and the energy dissipated is dependant on the surface area of the nano-filler involved.
Thus, the selective damping at shear modes may be realized only at higher CNT concentrations.
Second, the functionalization of CNT surfaces promotes cross-linking at CNT surfaces and
enhances uniform CNT dispersion in the epoxy. Furthermore, it increases the interface shear
modulus, and thus suppresses the stick-slip mechanism necessary for shear mode damping. With
improved dispersion methods and controlled CNT surface condition, CNT additions could reduce
shear resonance effectively even at higher CNT loadings. To illustrate this hypothesis, the shear
loss factor was increased from 049 . 0 = =
d s
η η to 3 . 0 =
s
η (500% increase as a conservative
estimate) and the TL was predicted (see Fig. 5b). The predicted TL trends show a significant
suppression of the TL peaks even at higher angles of incidence for the shear loss factor of 0.3.

171
This demonstrates that selective CNT damping can improve the acoustic transparency of
monolithic panels for a wide range of incident angles. .
6.6. Conclusions
Epoxy pre-polymers were copolymerized to synthesize matrices for acoustically
transparent composite materials for underwater sonar windows. CTBN, phenolic micro-balloons,
glass fibers, nano-clay, and o-MWCNTs were incorporated into the epoxy system, and the
acoustic behavior and mechanical properties of each composition was measured and analyzed.
The addition of o-MWCNT significantly increased the strength and modulus of the
composite without affecting the longitudinal wave speed. Thus, carbon nanotubes show promise
as an efficient non-intruding reinforcement for acoustically transparent materials. In addition, the
selective damping of shear modes resulting from CNT additions will enhance the acoustic
transparency of monolithic composites at a wide range of incidence angles. Optimized processing
methods for CNT dispersion will more fully exploit the reinforcing efficiency and selective
damping potential of CNT composites for underwater sonar structural materials.

172
Chapter 6 References
[1] Thomson, C.M., J. Acoust. Soc. Am., 1990. 87 (3): 1138-1143.

[2] “Acoustic Window”, United States Patent 6,831,876

[3]Thomas S. Ramotowski, Master’s Thesis, University of Rhode Island (2003)

[4] Robert E. Montgomery, Fred J. Weber, and David F. White, J. Acoust. Soc. Am., 1982. 71(3):
p. 735-741.

[5] (a) Twardowski, T.E., and Geil, P.H., J. Acoust. Soc. Am., 1990. 41: p. 1047-1054. (b)
Twarsowski, T.E., and Geil, P.H., J. Acoust. Soc. Am., 1991. 42: p. 69-74. (c) Twardowski, T.E.,
and Geil, P.H., J. Acoust. Soc. Am., 1991. 42: p. 1721-1726.

[6] Wei Chen, Maria L Auad, William R.J.J., Steven R. Nutt, European Polymer Journal, 42
(2006): 2765-2772.

[7] Wei Chen, Hongbin Lu, Steven R. Nutt, Composite Science and Technology, DOI:
10.1016/j.compscitech.2008.05.011, In press, available online 21 May 2008

[8] Wei Chen, Hongbin Shen, Maria L. Auad, Steven R. Nutt, submitted to Composites Part B,
(2008), under review.

[9] Jonghwan Suhr, Nikhil Koratkar, Pawel Keblinski, and Pulickel Ajayan, Nature Materials, 4,
134-137. (2005)

[10]Koratkar, N.A. et al, Adv. Mater. 2002, 14: p. 997-1000.

[11]Salvetat, J. P. et al, Appl Phys A: Mater Sci Process 1999. 68: p. 287-92.

[12]Koratkar, N.A. et al, Compos Sci and Tech, 2003. 63: p. 1525-31.

[13]Liu, L., Wagner, H. D., Composites Science and Technology, 2005. 65: p. 1861-1868.

[14]Jihua Gou, Scott O’Braint, Haichang Gu, and Gangbin Song, Journal of Nanomaterials,
32803, 1-7, (2006)

173
[15]Ioana C. Finegan, Gary G. Gibbetts, and Ronald F. Gibson, Composites Science and
Technology, 63, 1629-1635 (2003).

[16]Dasgupta, A.; Bhandarkar, S.M. Source: Mechanics of Materials, v 14, n 1, Nov. 1992, p
67-82

[17]Chatterjee, A.P. Journal of Applied Physics, v 100, n 5, 1 Sept. 2006, p 54302-1-8

[18]D’Almeida, J.R.M., Compos Sci and Tech, 1999. 59: p. 2087-2091.

174
Chapter 7. Conclusions and Future Work
7.1 Summary
The mechanical properties of carbon nanotubes make them attractive candidates as a
reinforcement filler material in polymer based structural composites. The possibility of
multifunctional composite materials with controllable thermal and electrical properties, in
addition to order-of-magnitude enhancements in the mechanical behavior, has led to a
tremendous amount of work dedicated to these material systems. Experimental work has
demonstrated success in enhancing the effective properties and manipulating the microstructures
of NRPs within the last decade. However, understanding the mechanisms of NRPs is made
difficult because of complexities related to the size/ shape of the nanotubes, the
dispersion/orientation of the CNTs, the nature of the interface, the CNT-polymer load transfer,
and processing dependency. Accurate models of how these issues influence the properties of the
NRPs will be necessary in order to optimize the fabrication and effective properties of these
materials. Based on the study of MWCNT epoxy NRP systems, we conclude the following:
Firstly, suitable MWCNT surface condition and reactivity compatible with the DGEBA
epoxy resin could be introduced through a two-step acid-epoxy functionalization process. This
process yielded uniform MWCNTs-stiff epoxy composites with significant enhancement in
flexural strength without sacrificing the elastic modulus compared to the neat resin.

175
Secondly, an ionic bonding scheme could be employed to graft epoxy-compatible
amine terminated linear polypropylene oxide (A-PPO) molecules of different molecular weight to
oxidized multi-walled carbon nanotubes (o-MWCNTs), and thereby modify the interphase of
MWCNT-epoxy nanocomposites. Through the judicious selection of solubility parameters,
molecular weight, flexibility, and reactivity of the grafts with those of the host matrix, the
thermomechanical properties of the nanocomposites can be tailored by introducing either a stiff
or a flexible interphase in these NREs.
Thirdly, direct incorporation of functionalized MWCNTs into epoxy/basalt composite
laminates improved the elastic properties of the composites, particularly in the fiber direction, and
this increase was accounted for by CNT re-distribution and re-orientation resulting during
processing. In-situ alignment of carbon nanotubes and formation of a distinct, hardened, annular
interphase around basalt fibers were instrumental in attaining the high reinforcement efficiency in
the fiber dominated direction of these hierarchical composites.
Fourthly, the findings suggest MWCNTs show promise as an efficient reinforcement
for acoustically transparent materials. The selective damping of shear modes and non-intrusive
nature of MWCNTs will enhance the acoustic transparency of monolithic composites at a wide
range of incidence angles.
In summary, this dissertation demonstrated dispersion enhancement and the ability to
control MWCNT-epoxy interphase through judicious selection of CNT surface functionalization

176
and matrix chemistry, and consequently, the bulk thermal and mechanical properties of NRPs can
be tailored. In addition, processing condition induced CNT organization presents to be a key
element in determining the effective properties in hierarchical composite structures. Moreover,
the unique shear wave mitigation properties of MWCNTs are promising in structural damping
and underwater acoustic applications. When combined with nano-atomic scale assembly
techniques, the ability to engineer interphase properties and attain additional functionalities is
essential to the design and synthesis of biomimetic, hybrid, multifunctional, hierarchical materials
that may one day approach the performance levels of natural nanocomposites mentioned in
Chapter 1. Multifuntional NRPs with perfect CNT dispersion, alignment, and bonding may not be
a remote dream when theories and experiments have accumulated enough understanding and push
further advancement towards the fine control of NRPs. In the following section, some future work
is proposed as a continuation of this dissertation.
7.2 NRPs with perfectly embedded CNT arrays
Polymer composites with perfectly aligned and bonded CNTs are the ultimate form of
NRPs, which will greatly facilitate experimental and theoretical work in the field. The ideal NRPs
will exhibit extraordinary anisotropic properties and can be applied as ultra-high performance
structural materials, flexible field-emission devices, thermal conducting pads,
microelectromechanical devices, chemical vapor sensors, and switchers for ion transportation.
While direct manipulation of individual CNT for bulk materials still remains to be an enormous

177
technical challenge, few available techniques are capable of assembling macroscopic CNT
performs or arrays in reasonable time frame and cost. In an effort to develop these NRPs, carbon
nanotube array reinforced polymers (NARPs) and carbon nanotube array reinforced fiber
composites (NARCs) are proposed recently.
In the NARP scheme, CNT arrays were synthesized and incorporated into polymer
matrix through a variety of methods. Although polymer assisted assembly of CNT arrays was
reported by Xie and colleagues, who successfully fabricated ordered carbon nanotube arrays via a
template (porous anodic aluminum oxide) based method [1], the majority of NARP schemes
involved CNT arrays synthesized using CVD method. In addition, many of these methods
consisted of capillarity driven wetting of aligned carbon nanotube arrays with polymers. [2]
Huang et al. fabricated composite films for thermal management by impregnating CVD grown
CNT arrays with a thermally conductive low viscosity elastomer. [3] Peng reported flexible,
transparent, conductive composites by spin coating pre-stabilized CNT sheets produced from
CVD method. [4] Alternative methods such as in-situ polymerization, electrochemical deposition,
and direct drawing of CNT forest were also explored recently. Feng et al. demonstrated
nanocomposites films with controlled organization through in-situ polymerization of polyaniline
films around MWCNT arrays. [5] Chen et al. electrochemically deposited polypyrrole over
nanotube arrays and formed composite films. [6] Strong, transparent, and multifunctional CNT

178

Figure 7.1 SEM images of CNT forest pillar wetted by SU-8. [2]

179
sheets capable of serving as CNT fabrics in NRPs were directly drawn from CVD nanotube forest
by Zhang et al. [7]
In the NARC scheme, CNT arrays are either transferred or grown directly onto
macroscopic fiber reinforcement or prepreg. Garcia et al. proposed the direct transfer of CNT
forest between composite laminates in an effort to enhance the mode I and Mode II fracture
toughness. [8] Alternatively, CNT arrays could also be assembled directly on/in fiber
reinforcement, and the CNT-fiber multi-reinforcement subsequently could be infiltrated with low
viscosity resin through suitable processing conditions. Qian et al. [9] grafted CNT forest on
carbon fibers using a CVD setup by pre-depositing ion nanoparticles using an incipient wetness
technique. A dramatic improvement in IFSS over the grafted carbon fiber/epoxy composites was
observed in the single fiber pull-out tests, but no significant change was shown in the push-out
tests. Garcia et al. [10] performed direct growth of aligned CNTs on the surface of advanced
fibers in a woven fabric and demonstrated a 69% increase in interlaminar shear strength and 10
6

(in-plane) and 10
8
(through-thickness) increases in laminate-level electrical conductivity.
Moreover, NARPs can be used as a resin film and subsequently incorporated with fiber
reinforcement with the maturity of NARP technology, albeit few have explored the possibilities.
However the abovementioned efforts are mostly limited to carbon nanotube forest
aligned in the vertical direction. Interpenetrating nanotube network consisting of both vertical and
horizontal CNTs are important in order to attain the so called nanotube fabrics. A couple of

180

Figure 7.2 Ultralow feeding gas flow guiding growth of large-scale horizontally aligned
single-walled carbon nanotube arrays. [14]

181
approaches have been identified to assist with the horizontal growth of nanotube arrays. Zhang et
al. [11] employed a template of one dimentional anodic aluminum oxide nanopore array to
fabricate horizontally aligned carbon nanotubes. The study of Kocabas et al. [12] shown that
Y-cut single-crystal quartz could be used to generate well-aligned, densely packed, horizontal
arrays of pristine SWCNTs over large areas, similar to those found on miscut sapphire[13].
Zhong et al. [14] prepared large-scale horizontally aligned ultralong SWCNT arrays using an
ultralow gas flow chemical vapor deposition strategy. These examples here are not trying to be
exhaustive, and present experimental capabilities are still a long way from realistic materials
application, however, they demonstrated the feasibility of nano-scopic fabrics that mimic
conventional fiber reinforced composite materials. With the maturity of the technology, these
materials will find unprecedented applications for on-chip micro-electric-mechanical systems and
high performance aerospace structural materials.
Recently, research activities towards the NARC have also been conducted at M C Gill
Composite Center of USC. Preliminary experiments effort on the CVD growth of carbon
nanotube forest with controlled CNT diameter, density, and length on carbon fibers through a
freeze-dry catalyst deposition method are underway. Preheating temperature and catalytic
solution concentration are considered to be the key parameters in determining CNT diameter and
forest density respectively. These hierarchical nano-structures will be incorporated into epoxy
matrix through the infiltration of low viscosity epoxy in a vacuum bagging process. The

182

Figure 7.3 MWCNT growth on carbon fiber surface.

2 μ
20 μ

183
thermomechanical properties, tensile properties on a micro tensile stage, and morphology will be
studied and compared with the control composites in the upcoming research papers.

184
Chapter 7 References
[1] Xu-Ming Xie; Yi-Tao Liu; Yan-Fang Gao; Qing-Ping Feng; Lin-Rui Guo; Xiao-Hao Wang;
Xiong-Ying Ye. Carbon, v 44, n 3, March 2006, 599-602

[2] Garcia, E.J.; Wardle, B.L.; Slocum, A.H. Nanotechnology, v 18, n 16, 25 April 2007, 11 pp.

[3]Huang, Hua; Fan, Shoushan. Advanced Materials, v 17, n 13, Jul 4, 2005, p 1652-1656

[4] Peng, Huisheng. Journal of the American Chemical Society, v 130, n 1, Jan 9, 2008, p 42-43

[5] Feng, W.; Bai, X.D.; Lian, Y.Q.; Liang, J.; Yoshino, K. Carbon, v 41, n 8, 2003, 1551-7

[6] Chen, J.H.; Huang, Z.P.; Wang, D.Z.; Yang, S.X.; Wen, J.G.; Ren, Z.F. Applied Physics A, v
A73, n 2, Aug. 2001, 129-31

[7] Mei Zhang; Shaoli Fang; Zakhidov, A.A.; Lee, S.B.; Aliev, A.E.; Williams, C.D.; Atkinson,
K.R.; Baughman, R.H. Science, v 309, n 5738, 19 Aug. 2005, 1215-19

[8] Garcia, E.J.; Wardle, B.L.; Hart, A.J. Composites Part A: Applied Science and
Manufacturing, v 39, n 6, June 2008, 1065-70

[9]Qian, Hui; Bismarck, Alexander; Greenhalgh, Emile S.; Kalinka, Gerhard; Shaffer, Milo S.P.
Chemistry of Materials, v 20, n 5, Mar 11, 2008, p 1862-1869

[10]Garcia, Enrique J.; Wardle, Brian L.; John Hart, A.; Yamamoto, Namiko Composites Science
and Technology, v 68, n 9, July, 2008, p 2034-2041

[11] Hongguo Zhang; Zhi Chen; Tianxiang Li; Fengjuan Wang; Saito, K. Journal of the
Electrochemical Society, v 154, n 2, Feb. 2007, H124-6

[12] Kocabas, Coskun; Hur, Seung-Hyun; Gaur, Anshu; Meitl, Matthew A.; Shim, Moonsub;
Rogers, John A. Small, v 1, n 11, November, 2005, p 1110-1116

[13] A. Ismach, L. Segev, E. Wachtel, E. Joselevich, Angew. Chem. 2004, 116, 6266 –6269;
Angew. Chem. Int. Ed. 2004, 43, 6140 –6143.

[14] Zhong Jin; Haibin Chu; Wenchang Tan; Yan Li. Nano Letters, v 7, n 7, July 2007, 2073-9

185
Bibliography
Ahir, S.V.; Huang, Y.Y.; Terentjev, E.M. Polymer, v 49, n 18, Aug 26, 2008, p 3841-3854

Ajayan PM, Schadler LS, Braun PV. Nanocomposite Science and Technology. John Wiley, Sons
Inc.; 2003 Ch.1, 2.

Ajayan PM, Suhr J, Koratkar N. J Mater Sci 2006; 41(23):7824-29.

Ajayan, P.M.; Stephan, O.; Colliex, C.; Trauth, D. Science, v 265, n 5176, 26 Aug. 1994,
1212-14

Ando, Yoshinori (Meijo Univ); Iijima, Sumio. Japanese Journal of Applied Physics, Part 2:
Letters, v 32, n 1A-B, Jan 15, 1993, p L107-L109

Angelikopoulos, Panagiotis; Bock, Henry Source: Journal of Physical Chemistry B, v 112, n 44,
Nov 6, 2008, p 13793-13801

Anglaret, E.; Bendiab, N.; Guillard, T.; Journet, C.; Flamant, G.; Laplaze, D.; Bernier, P.;
Sauvajol, J.-L. Carbon, v 36, n 12, 1998, 1815-20

Anumandla V, Gibson RF. Compos Part A: Appl Sci Manuf 2006;37: 2178-85.

Assouline, E.; Lustiger, A.; Barber, A.H.; Cooper, C.A.; Klein, E.; Wachtel, E.; Wagner, H.D.
Journal of Polymer Science, Part B: Polymer Physics, v 41, n 5, Mar 1, 2003, p 520-527

A. Ismach, L. Segev, E. Wachtel, E. Joselevich, Angew. Chem. 2004, 116, 6266–6269; Angew.
Chem. Int. Ed. 2004, 43, 6140–6143.

Bahr, J. L.; Tour, J. M. J. Mater. Chem. 2002, 12, 1952. (d) Hirsch, A. Angew. Chem., Int. Ed.
2002, 41, 1853.

Bakunin, V.N.; Suslov, A.Yu.; Kuzmina, G.N.; Parenago, O.P. Journal of Nanoparticle Research,
v 6, n 2-3, June 2004, p 273-84

Bandow, Shunji; Rao, A.M.; Williams, K.A.; Thess, A.; Smalley, R.E.; Eklund, P.C. Journal of
Physical Chemistry B, v 101, n 44, Oct 30, 1997, p 8839-8842

186
Banerjee, Sarbajit; Hemraj-Benny, Tirandai; Wong, Stanislaus S. Advanced Materials, v 17, n 1,
Jan 18, 2005, p 17-29

Bansal A, Yang H, Li CZ, Cho K, Benicewicz BC, Kumar SK, Schadler LS. Nature Mater
2005;4:693-98.

Barber, A.H.; Cohen, S.R.; Kenig, S.; Daniel Wagner, H. Composites Science and Technology, v
64, n 15, 2004, 2283-9

Barber, A.H.; Cohen, S.R.; Wagner, H.D. Applied Physics Letters, v 82, n 23, 9 June 2003, p
4140-2

Bartlet, P.; Pascault, J. P.; Sautereau, H. J. Appl. Polym. Sci. 1985, 30, 2955.

Barton AFM. CRC, CRC Press; 1990.

Baxendale, M. IEE Proceedings Nanobiotechnology, v 150, n 1, June, 2003, p 3-8

Bazuaye A, Huglin, MB. Polymer 1979;20(1):44-50

Bekyarova E, Thostenson ET, Yu A, Kim H, Gao J, Tang J, Hahn HT, Chou TW, Itkis ME,
Haddon RC. Langmuir 2007; 23(7):3970-4.

Bekyarova E, Thostenson ET, Yu AP, Itkis ME, Fakhrutdinov D, Chou TW, Haddon
RC.Bellamy, L. J. The Infra-red Spectra of Complex Molecules; John Wiley: London,1966.

Benedict, Brian; Pehrsson, Pehr E.; Zhao, Wei. Journal of Physical Chemistry B, v 109, n 16, Apr
28, 2005, p 7778-7780

Berrod, G.; Vidal, A.; Papirer, E.; Donnet, J. B. Journal of Applied Polymer Science, v 23, n 9,
May, 1979, p 2579-2590

Biercuk, M.J.; Llaguno, M.C.; Radosavijevic, M.; Hyun, J.K.; Johnson, A.T.; Fischer, J.E.
Applied Physics Letters, v 80, n 15, Apr 15, 2002, p 2767

Bifeng Pan; Daxiang Cui; Feng Gao; Rong He Source: Nanotechnology, v 17, n 10, 28 May
2006, 2483-9

187
Bliznyuk, V.N.; Singamaneni, S.; Sanford, R.L.; Chiappetta, D.; Crooker, B.; Shibaev, P.V.
Polymer, v 47, n 11, May 17, 2006, p 3915-3921

Bonard, J.-M.; Stora, T.; Salvetat, J.-P.; Maier, F.; Stoeckli, T.; Duschl, C.; Forro, L.; de Heer,
W.A.; Chatelain, A. Advanced Materials, v 9, n 10, 8 Aug. 1997, p 827-31

Boogh LCN, Meier, RJ, Kausch HH, Kip BJ. J Polym Sci B 1992; 30(4):325-33.

Bower, C.; Rosen, R.; Jin, L.; Han, J.; Zhou, O. Applied Physics Letters, v 74, n 22, 1999, p
3317-3319

Bradshaw, R.D.; Fisher, F.T.; Brinson, L.C. Composites Science and Technology, v 63, n 11,
2003, 1705-22

Breuer, O.; Sundararaj, Uttandaraman. Polymer Composites, v 25, n 6, December, 2004, p
630-645

Bronikowski, Michael J. Carbon, v 44, n 13, November, 2006, p 2822-2832

Campbell, Malcolm D.; Couttas, Robert S. P.; Michell, Anthony J.; Willis, Donald. Industrial &
Engineering Chemistry, Product Research and Development, v 19, n 4, Dec, 1980, p 596-601

Camponeschi E, Vance R, Al-Haik M, Garmestani H, Tannenbaum R. Carbon 2007; 45(10):
2037-46

Camponeschi, Erin; Vance, Richard; Al-Haik, Marwan; Garmestani, Hamid; Tannenbaum, Rina.
Carbon, v 45, n 10, September, 2007, p 2037-2046

Chatterjee, A.P. Journal of Applied Physics, v 100, n 5, 1 Sept. 2006, p 54302-1-8

Chen W, Auad ML., Williams RJJ, Nutt SR. Euro Polym J 2006; 42: 2765-72.

Chen W, Lu HB, Nutt SR. Compos Sci Technol 2008; Article in press. DOI information:
10.1016/j.compscitech.2008.05.011.

Chen, J.; Hamon, M. A.; Hu, H.; Chen, Y.; Rao, A. M.; Eklund, P. C.; Haddon, R. C. Science
1998, 282, 95.

188
Chen, J.H.; Huang, Z.P.; Wang, D.Z.; Yang, S.X.; Wen, J.G.; Ren, Z.F. Applied Physics A, v
A73, n 2, Aug. 2001, 129-31

Cheng, Da-Dian; Yu, Rong-Qing; Liu, Zhao-Yang; Zhang, Qiang; Wang, Yu-Huang; Huang,
Rong-Bin; Zhan, Meng-Xiong; Zheng, Lan-Sun Chemical Journal of Chinese Universities, v 16,
n 6, Jun, 1995, p 949

Cheol Park; Ounaies, Z.; Watson, K.A.; Crooks, R.E.; Smith, J., Jr.; Lowther, S.E.; Connell,
J.W.; Siochi, E.J.; Harrison, J.S.; St Clair, T.L. Chemical Physics Letters, v 364, n 3-4, 4 Oct.
2002, 303-8

Chiu, H.-Y.; Deshpande, V.V.; Postma, H.W.Ch.; Lau, C.N.; Miko, C.; Forro, L.; Bockrath, M.
Physical Review Letters, v 95, n 22, 25 Nov. 2005, 226101/1-4

Chou TW, Thostenson ET. Carbon 2006; 44 (14):3022-29.

Christensen, R.M.; Lo, K.H. Journal of the Mechanics and Physics of Solids, v 27, n 4, Aug.
1979, 315-30

Chun Tang; Wanlin Guo; Changfeng Chen Physical Review Letters, v 100, n 17, 2 May 2008,
175501-1-4

Ci, L. ;Bai, J. Composites Scince and Technology 2005 Article In Press, Available online 11 July
2005.

Coleman, J.N.; Cadek, M.; Dalton, A.B.; Munoz, E.; Razal, J.; Baughman, R.H.; Blau, W.J.
Proceedings of the SPIE - The International Society for Optical Engineering, v 5118, 2003, 271-9

Coleman, J.N.; Khan, U.; Blau, W.J.; Gun'ko, Y.K. Carbon, v 44, n 9, Aug. 2006, 1624-52

Coleman, J.N.; Khan, U.; Blau, W.J.; Gun'ko, Y.K. Source: Carbon, v 44, n 9, Aug. 2006,
1624-52

Compos Sci Technol 2007; 67(7-8):1509-17.

Cooper, C.A.; Young, R.J.; Halsall, M. Source: Composites Part A, v 32A, n 3-4, 2001, 401-11

Cooper, Carole A.; Cohen, Sidney R.; Barber, Asa H.; Wagner, H. Daniel. Applied Physics
Letters, v 81, n 20, Nov 11, 2002, p 3873-3875

189
Crainic, N.; Marques, A.T. Source: Key Engineering Materials, v 230-232, 2002, p 656-9

Cui J, Wang WP, You YZ, Liu CH, Wan PH. Polymer 2004;45:8717-21.

Cumings J, Zettl A. Phys Rev Lett 2004;93 (8):086801-1-4.

D’Almeida, J.R.M., Compos Sci and Tech, 1999. 59: p. 2087-2091.

Daniel, Isaac M. Engineering mechanics of composite materials, Oxford University Press, 2006

Dasgupta, A.; Bhandarkar, S.M. Source: Mechanics of Materials, v 14, n 1, Nov. 1992, p 67-82

Dean D, Obore AM, Richmond S, Nyairo E. Compos Sci Technol 2006; 66(13): 2135-42.

dell’Erba, I. E.; Fasce, D. P.; Williams, R. J. J.; Erra-Balsells, R.; Fukuyama, Y.; Nonami, H.
Macromol. Mater. Eng. 2004, 289, 315.

Didenko, V.V.; Moore, V.C.; Baskin, D.S.; Smalley, R.E. Nano Letters, v 5, n 8, Aug. 2005,
1563-7

Dilandro L, Dibenedetto AT, Groeger J. Polym Compos 1988; 9(3):209-21.

Ding W, Eitan A, Fisher FT, Chen X, Dikin DA, Andrews R, Brinson LC, Schadler LS, Ruoff
RS.

Ding, W.; Eitan, A.; Fisher, F.T.; Chen, X.; Dikin, D.A.; Andrews, R.; Brinson, L.C.; Schadler,
L.S.; Ruoff, R.S. Nano Letters, v 3, n 11, Nov. 2003, 1593-7

Doe, Changwoo; Choi, Sung-Min; Kline, Steven R.; Jang, Hyung-Sik; Kim, Tae-Hwan Advanced
Functional Materials, v 18, n 18, Sep 23, 2008, p 2685-2691

Dror, Yael; Pyckhout-Hintzen, Wim; Cohen, Yachin. Macromolecules, v 38, n 18, Sep 6, 2005, p
7828-7836

Dror, Yael; Salalha, Wael; Khalfin, Rafail L.; Cohen, Yachin; Yarin, Alexander L.; Zussman,
Eyal. Langmuir, v 19, n 17, Aug 19, 2003, p 7012-7020

Duesberg, G.S.; Muster, J.; Byrne, H.J.; Roth, S.; Burghard, M. Applied Physics A, v A69, n 3,
Sept. 1999, 269-74

190
Duesberg, G.S.; Muster, J.; Krstic, V.; Burghard, M.; Roth, S. Applied Physics A: Materials
Science & Processing, v 67, n 1, Jul, 1998, p 117-119

Ebbesen, T.W.; Lezec, H.J.; Hiura, H.; Bennett, J.W.; Ghaemi, H.F.; Thio, T. Nature, v 382, n
6586, Jul 4, 1996, p 54-55

Eitan A, Fisher FT, Andrews R, Brinson L, Schadler LS. Compos Sci Technol 2006; 66:1162-73.

Eitan, A.; Jiang, K.; Dukes, D.; Andrews, R.; Schadler, L. S. Chem. Mater. 2003,15, 3198.

Ellis B. Glasgow UK: Blackie Academic & Professional; 1993, Ch.3.

Eshelby J.D. Proceedings of the Royal Society of London A 241 (1957), p 376

Fang, Hai-Tao; Liu, Chen-Guang; Liu, Chang; Li, Feng; Liu, Min; Cheng, Hui-Ming.
Chemistry of Materials, v 16, n 26, Dec 28, 2004, p 5744-5750

Fangming Du; Fischer, J.E.; Winey, K.I. Journal of Polymer Science, Part B (Polymer Physics), v
41, n 24, 15 Dec. 2003, 3333-8

Feng, W.; Bai, X.D.; Lian, Y.Q.; Liang, J.; Wang, X.G.; Yoshino, K. Carbon, v 41, n 8, 2003,
1551-7

Fisher FT, Bradshaw RD, Brinson LC. Appl Phys Lett 2002;80(24):4647-49.

Fisher FT, Eitan A, Andrews R, Brinson LC, Schadler LS. Compos Sci Technol
2006;66(9):1159-70.

Fisher, F.T. Dissertation 2002, Northwestern University

Fisher, F.T.; Bradshaw, R.D.; Brinson, L.C. Composites Science and Technology, v 63, n 11,
2003, 1689-703

Fisher, F.T.; Eitan, A.; Andrews, R.; Brinson, L.C.; Schadler, L.S. Composites Science and
Technology, v 66, n 9, July 2006, 1159-70

Fisher, FT and LC Brinson (2001). Compos Sci Technol, 61:731-48.

Fitzer, E.; Terwiesch, B. Carbon, v 10, n 4, Aug, 1972, p 383-390

191
Foo, C.W.P.; Patwardhan, S.V.; Belton, D.J.; Kitchel, B.; Anastasiades, D.; Huang, J.; Naik,
R.R.; Perry, C.C.; Kaplan, D.L. Proceedings of the National Academy of Sciences of the United
States of America, v 103, n 25, 20 June 2006, p 9428-33

Frankland SJV, Caglar A, Brenner DW, Griebel M. J Phys Chem B 2002;106: 3046-48.

Frankland, S.J.V.; Caglar, A.; Brenner, D.W.; Griebel, M. Journal of Physical Chemistry B, v
106, n 12, Mar 28, 2002, p 3046-3048

Fujii, Motoo; Zhang, Xing; Xie, Huaqing; Ago, Hiroki; Takahashi, Koji; Ikuta, Tatsuya; Abe,
Hidekazu; Shimizu, Tetsuo Physical Review Letters, v 95, n 6, Aug 5, 2005, p 065502

Gajewski, S.; Maneck, H.-E.; Knoll, U.; Neubert, D.; Dorfel, I.; Mach, R.; Strauss, B.; Friedrich,
J.F. Diamond and Related Materials, v 12, n 3-7, March/July, 2003, p 816-820

Gao, X.-L.; Li, K. International Journal of Solids and Structures, v 42, n 5-6, March, 2005, p
1649-1667

Garcia EJ, Hart AJ, Wardle BL, Slocum AH. Adv Mater 2007; 9(16):2151-56.

Garcia EJ, Wardle BL, Hart AJ. Composites: Part A 2008; 39:1065-70.

Garcia, E.J.; Hart, A.J.; Wardle, B.L.; Slocum, A.H. Nanotechnology, v 18, n 16, 25 April 2007,
11 pp.

Garcia, E.J.; Wardle, B.L.; Hart, A.J. Composites Part A: Applied Science and Manufacturing, v
39, n 6, June 2008, 1065-70

Garcia, Enrique J.; Wardle, Brian L.; John Hart, A.; Yamamoto, Namiko Composites Science and
Technology, v 68, n 9, July, 2008, p 2034-2041

Gojny FH, Wichmann MHG, Fiedler B, Bauhofer W, Schulte K. Composites: Part A 2005;
36(11):1525-35.

Gojny FH, Wichmann MHG, Kopke U, Fiedler B, Schulte K. Compos Sci Technol 2004;64(15):
2363-2371.

Gojny FH, Wichmann, MHG, Kopke U, Fiedler B, Schulte K. Compos Sci Technol 2004;
64(15):2363-71.

192
Gojny, FH, Schulte K. Compos Sci Technol 2004; 64(15):2303-08.

Gorga, R.E.; Lau, K.K.S.; Gleason, K.K.; Cohen, R.E. Journal of Applied Polymer Science, v
102, n 2, 15 Oct. 2006, 1413-18

Grady, Brian P.; Ha, Mai. AIChE Annual Meeting, Conference Proceedings, 05AIChE: 2005
AIChE Annual Meeting and Fall Showcase, Conference Proceedings, 2005, p4964

Grest GS. J Chem Phys 1996; 105(13): 5532-5541

Guang-Xin Chen; Yongjin Li; Shimizu, H. Source: Carbon, v 45, n 12, Oct. 2007, 2334-40

Gunes, I. Sedat; Jana, Sadhan C. Journal of Nanoscience and Nanotechnology, v 8, n 4, April,
2008, p 1616-1637

Hadjiev, V.G.; Lagoudas, D.C.; Oh, E.-S.; Thakre, P.; Davis, D.; Files, B.S.; Yowell, L.; Arepalli,
S.; Bahr, J.L.; Tour, J.M. Composites Science and Technology, v 66, n 1, Jan. 2006, 128-36

Haggenmueller, R.; Gommans, H.H.; Rinzler, A.G.; Fischer, J.E.; Winey, K.I. Chemical Physics
Letters, v 330, n 3-4, 10 Nov. 2000, 219-25

Haggenmueller, Reto; Guthy, Csaba; Lukes, Jennifer R.; Fischer, John E.; Winey, Karen I.
Macromolecules, v 40, n 7, Apr 3, 2007, p 2417-2421

Halpin, J. C.; Kardos, J. L. Polymer Engineering and Science, v 16, n 5, May, 1976, p 344-352

Hammerand, D.; Seidel, G.D.; Lagoudas, D.C. Mechanics of Advanced Materials and Structures,
v 14, n 4, May 2007, 55-72

Hamon MA, Chen J, Hu H, Chen YS, Itkis ME, Rao AM, Eklund PC, Haddon RC. Adv Mater
1999; 11(10):834-40.

Hashin Z, Rosen B. W. Journal of Applied Mechanics,1964. 31, 223-232
Hashin, Z.; Shtrikman, S. Journal of Mechanics and Physics of Solids, v 11, n 2, Mar-Apr, 1963,
p 127-140

Head, Andrew Society of the Plastics Industry, Reinforced Plastics/Composites Institute, Annual
Conference - Proceedings, Feb 1-5, 1988, p 9D.1-9D.3

193
Hernandez-Lopez, J.L.; Alvizo-Paez, E.R.; Moya, S.E.; Ruiz-Garcia, J. Carbon, v 45, n 12, Oct.
2007, 2448-50

Hilding JM, Hong MPL, Grullke EA.Hill, R. Journal of Mechanics and Physics of Solids, v 13, n
4, Aug, 1965, p 213-222

Hongguo Zhang; Zhi Chen; Tianxiang Li; Fengjuan Wang; Saito, K. Journal of the
Electrochemical Society, v 154, n 2, Feb. 2007, H124-6

Hoppe, C. E.; Galante, M. J.; Oyanguren, P. A.; Williams, R. J. J. Macromol. Mater. Eng. 2005,
290, 456.

Hsu, W.K.; Terrones, M.; Hare, J.P.; Reid, D.G.; Walton, D.R.M.; Kroto, H.W. Proceedings of
the International Winter school on Electronic Properties of Novel Materials. Fullerenes and
Fullerene Nanostructures, 1996, 226-31

Hu R, Dimonie VL, El-Aasser MS, Pearson RA, Hiltner A, Mylonakis SG, Sperling LH. J Polym
Sci, B 1997;35(10):1501-14.

Huang L, Wang C, Lu, YF. J Reinf Plast Compos 2008; 27(7): 725-38.

Huang, Hua; Liu, Changhong; Wu, Yang; Fan, Shoushan. Advanced Materials, v 17, n 13, Jul 4,
2005, p 1652-1656

Hussain, F.; Hojjati, M.; Okamoto, M.; Gorga, R.E. Journal of Composite Materials, v 40, n 17,
Sept. 2006, p 1511-75

Iijima S. Nature 1991; 354: 56 – 58.

Ioana C. Finegan, Gary G. Gibbetts, and Ronald F. Gibson, Composites Science and Technology,
63, 1629-1635 (2003).

Ismail, A.F.; Goh, Pei Sean; Tee, Jia Chee; Sanip, Suhaila M.; Aziz, Madzlan. Nano, v 3, n 3,
2008, p 127-143

Ivanov, ´I`Achko P. (´I`Achko Pavlov) Polymer composite materials : interface phenomena &
processes. Kluwer Academic Publishers, 2001.

J.R. Xiao, B.A. Gama, J.W. Gillespie jr, Int. J. Solids Struct. 42 (2005) 3075–3092.

194
Jiang, Zehui (International Center for Bamboo and Rattan); Sun, Zhengjun; Ren, Haiqing Acta
Materiae Compositae Sinica, v 23, n 3, June, 2006, p 127-129

Jihua Gou, Scott O’Braint, Haichang Gu, and Gangbin Song, Journal of Nanomaterials, 32803,
1-7, (2006)

Johnson, B.L. Industrial and Engineering Chemistry, v 43, n 1, Jan, 1951, p 146-154

Jonathan N, Coleman, UK, Yurii K, Gun K. Adv Mater 2006; 18:689-706.

Jonghwan Suhr, Nikhil Koratkar, Pawel Keblinski, and Pulickel Ajayan, Nature Materials, 4,
134-137. (2005)

Jordan, J.; Jacob, K.I.; Tannenbaum, R.; Sharaf, M.A.; Jasiuk, L. Materials Science &
Engineering A, v 393, n 1-2, 25 Feb. 2005, p 1-11

Khabashesku VN, Barrera EV, Zhu J, Imam A, Crane R, Lozano K.

Ki Chul Park; Mahiko, T.; Morimoto, S.; Takeuchi, K.; Endo, M. Applied Surface Science, v
254, n 22, 15 Sept. 2008, 7438-45

Kikuchi, Y.; Inoue, S.; Nakajima, T. Carbon, v 44, n 7, June 2006, 1287-91

Kilbey II, SM, Watanabe H, Tirrell M. Macromolecules 2001;34(15):5249-59

Kilpatrick, S.J.; Cao, A.; Li, X.; Renna, N.J.; Ajayan, P.M. 2005 International Semiconductor
Device Research Symposium, 2005, 2 pp.

Kim YA, Kamio S, Tajiri T, Hayashi T, Song SM, Endo M, Terrones M, Dresselhaus, M.S. Appl
Phys Lett 2007; 90(9):93125-1-3.

Kim, P.; Shi, L.; Majumdar, A.; McEuen, P.L. Physical Review Letters, v 87, n 21, 19 Nov. 2001,
215502/1-4

Kim, Yeji; Minami, Nobutsugu; Zhu, Weihong; Kazaoui, Said; Azumi, Reiko; Matsumoto,
Mutsuyoshi Source: Japanese Journal of Applied Physics, Part 1: Regular Papers and Short Notes
and Review Papers, v 42, n 12, December, 2003, p 7629-7634

195
Kimura, Tohru; Ago, Hiroki; Tobita, Masayuki; Ohshima, Satoshi; Kyotani, Mutsumasa;
Yumura, Motoo Source: Advanced Materials, v 14, n 19, Oct 2, 2002, p 1380-1383

Kin Liao; Sean Li Source: Applied Physics Letters, v 79, n 25, 17 Dec. 2001, 4225-7

Kin-tak Lau; Chong Gu; Hui, D. Source: Composites Part B (Engineering), v 37, n 6, 2006,
425-36

Kis, A.; Csanyi, G.; Salvetat, J.-P.; Lee, Thien-Nga; Couteau, E.; Kulik, A.J.; Benoit, W.;
Brugger, J.; Forro, L. Nature Materials, v 3, n 3, March, 2004, p 153-157

Kocabas, Coskun; Hur, Seung-Hyun; Gaur, Anshu; Meitl, Matthew A.; Shim, Moonsub; Rogers,
John A. Small, v 1, n 11, November, 2005, p 1110-1116

Koratkar, N.A. et al, Adv. Mater. 2002, 14: p. 997-1000.

Koratkar, N.A. et al, Compos Sci and Tech, 2003. 63: p. 1525-31.

Krishnan, A.; Dujardin, E.; Ebbesen, T.W.; Yianilos, P.N.; Treacy, M.M.J. Physical Review B, v
58, n 20, 15 Nov. 1998, 14013-19

Kroto, H.W.; Heath, J.R.; O'Brien, S.C.; Curl, R.F.; Smalley, R.E. Nature, v 318, n 6042, 14 Nov.
1985, 162-3

Kruis, F. Einar; Fissan, Heinz; Peled, Aaron. Journal of Aerosol Science, v 29, n 5-6, Jun-Jul,
1998, p 511-535

Kuzumaki T, Mitsuda Y. Carbon 2005; 43(10): 2227-9.

Lagoudas, D.C.; Seidel, G.D. Mechanics of Materials, v 38, n 8-10, Aug.-Oct. 2006, 884-907

Lambert JM, Ajayan PM, Bernier P, Planeix JM. Chem Phys Lett 1994;226(3–4):364–71.

Lanticse, Leslie Joy; Tanabe, Yasuhiro; Matsui, Keitaro; Kaburagi, Yutaka; Suda, Katsumi;
Hoteida, Masayuki; Endo, Morinobu; Yasuda, Eiichi. Carbon, v 44, n 14, November, 2006, p
3078-3086

Lee, C.; Hwang, W. Source: Journal of Materials Science Letters, v 17, n 19, 1 Oct. 1998, 1601-3

196
Li F, Cheng HM, Xing YT, Tan PH, Su G. Carbon 2000;38(14):2041–5.

Li XD, Gao HS, Scrivens WA, Fei DL, Xu XY, Sutton MA, Reynolds AP, Myrick ML. J
Nanosci Nanotechnol 2007; 7(7): 2309-17.

Li, Xiaolin; Tu, Xiaomin; Zaric, Sasa; Welsher, Kevin; Won, Seok Seo; Zhao, Wei; Dai, Hongjie.
Journal of the American Chemical Society, v 129, n 51, Dec 26, 2007, p 15770-15771

Li, Yongjin; Shimizu, Hiroshi. Polymer, v 48, n 8, Apr 5, 2007, p 2203-2207

Lin, M.F.; Chuu, D.S. Physical Review B (Condensed Matter), v 56, n 8, 15 Aug. 1997,
4996-5002

Lin, Tong; Bajpai, Vardhan; Ji, Tao; Dai, Liming. Australian Journal of Chemistry, v 56, n 7,
2003, p 635-651

Lin, Y.; Zhou, B.; Shiral Fernando, K.A.; Liu, P.; Allard, L. F.; Sun, Y. P. Macromolecules 2003,
36, 7199-7204

Liu J, Rinzler AG, Dai H, Hafner JH, Bradley RK, Boul PJ, Lu A, Iverson T, Shelimov K,
Huffman CB, Rodriguez-Macias F, Shon Y, Lee TR, Colbert DT, Smalley RE. Science 1998;280:
1253-1256

Liu J, Rinzler AG, Dai H, Hafner JH, Bradley RK, Boul PJ, Lu A, Iverson T, Shelimov K,
Huffman CB, Rodriguez-Macias F, Shon Y, Lee TR, Colbert DT, Smalley RE. Science
1998;280:1253.

Liu YQ, Gao L, Sun J, Zheng S, Jiang LQ,Wang Y. Carbon 2007;45(10):1972–8.

Liu, J.; Rinzler, A. G.; Dai, H.; Hafner, J. H; Bradley, R. K.; Boul, P. J.; Lu, A.;Iverson, T.;
Shelimov, K.; Huffman, C. B.; Rodriguez-Macias, F.; Shon, Y.;Lee, T. R.; Colbert, D. T.;
Smalley, R. E. Science 1998, 280, 1253.

Liu, Jie; Rinzler, Andrew G.; Dai, Hongjie; Hafner, Jason H.; Bradley, R. Kelley; Boul, Peter J.;
Lu, Adrian; Iverson, Terry; Shelimov, Konstantin; Huffman, Chad B.; Rodriguez-Macias,
Fernando; Shon, Young-Seok; Lee, T. Randall; Colbert, Daniel T.; Smalley, Richard E. Science,
v 280, n 5367, May 22, 1998, p 1253-1256

Liu, L., Wagner, H. D., Composites Science and Technology, 2005. 65: p. 1861-1868.

197
Liu, L.; Wagner, H.D. Composites Science and Technology 2005, 65, 1861-1868

Load transfer in carbon nanotube epoxy composites. Appl Phys Lett 1998;73(26):3842-44.

Lordi, V.; Nan Yao. Journal of Materials Research, v 15, n 12, Dec. 2000, 2770-9

Lourie, O.; Wagner, H.D. Composites Science and Technology, v 59, n 6, May, 1999, p 975-977

Lurie, S.; Belov, P.; Volkov-Bogorodsky, D.; Tuchkova, N. Journal of Materials Science, v 41, n
20, Oct. 2006, 6693-707

M. Meo, a, and M. Rossia Material Research Center, Department of Mechanical Engineering,
University of Bath, Bath, Ba2 7AY, UK

MacKernan, D.; Blau, W.J. Europhysics Letters, v 83, n 6, Sept. 2008, 66009 (6 pp.)

Maehashi, K.; Ohno, Y.; Inoue, K.; Matsumoto, K. Applied Physics Letters, v 85, n 6, 9 Aug.
2004, 858-60

Malik, B. (Natl. Inst. Foundry Forge Technol.); Talukdar, P.; Mukherjee, S.K. Journal of the
Institution of Engineers (India): Aerospace Engineering Journal, v 82, n 1, May, 2001, p 38-40

Martin, C.A.; Sandler, J.K.W.; Windle, A.H.; Schwarz, M.-K.; Bauhofer, W.; Schulte, K.;
Shaffer, M.S.P. Polymer, v 46, n 3, Jan 26, 2005, p 877-886

Matarredona, Olga; Rhoads, Heather; Li, Zhongrui; Harwell, Jeffrey H.; Balzano, Leandro;
Resasco, Daniel E. Journal of Physical Chemistry B, v 107, n 48, Dec 4, 2003, p 13357-13367

Mathur RB, Chatterjee S, Singh BP. Compos Sci Technol 2008; 68(7-8): 1608-15.

Mayes AM. Nature Mater 2005; 4:651-52.

McCarthy, B.; Coleman, J.N.; Czerw, R.; Dalton, A.B.; Byrne, H.J.; Tekleab, D.; Iyer, P.;
Ajayan, P.M.; Blau, W.J.; Carroll, D.L. Nanotechnology, v 12, n 3, September, 2001, 8th
Foresight Confernce on Molecular Nanotechnology, p 187-190

McCullough (R. L.); Univ of Delaware, Cent for Composite Materials, Newark, DE, USA
Composites Science and Technology, v 22, n 1, 1985, p 3-21

198
Mead DJ, Joannides RJ. Composites 1991;22(1):15-29.

Meguid, S.A.; Sun, Y. Source: Materials & Design, v 25, n 4, June 2004, 289-96

Mei Zhang; Shaoli Fang; Zakhidov, A.A.; Lee, S.B.; Aliev, A.E.; Williams, C.D.; Atkinson,
K.R.; Baughman, R.H. Science, v 309, n 5738, 19 Aug. 2005, 1215-19

Mezzenga R, Boogh L, Manson JAE. J Polym Sci B 2000; 38(14):1883-92

Miaudet P, Badaire S, Maugey M, Derre A, Pichot V, Launois P, Poulin P, Zakri C. Nano
Letters 2005; 5(11): 2212-15.

Min-Feng Yu; Files, B.S.; Arepalli, S.; Ruoff, R.S. Physical Review Letters, v 84, n 24, 12 June
2000, 5552-5

Min-Feng Yu; Lourie, O.; Dyer, M.J.; Moloni, K.; Kelly, T.F.; Ruoff, R.S. Science, v 287, n
5453, 28 Jan. 2000, 637-40

Min-Feng Yu; Yakobson, B.I.; Ruoff, R.S. Journal of Physical Chemistry B, v 104, n 37, 21 Sept.
2000, 8764-7

Misra, Devesh K. The Minerals, Metals and Materials Society, 2006

Miyagawa H, Misra M, Mohanty AK. J Nanosci Nanotechnol 2005; 5(10):1593-615.

Miyagawa, H.; Drzal, L. T. Polymer 2004, 45, 5163.

Namilae, S.; Chandra, N. Source: Transactions of the ASME. Journal of Engineering Materials
and Technology, v 127, n 2, April 2005, 222-32

Nano Lett 2003; 3(11):1593-97.

Netz RR, Schick M. Europhys Lett 1997; 38(1): 37-42

Niyogi, S.; Hamon, M. A.; Hu, H.; Zhao, B.; Bhowmik, P.; Sen, R.; Itkis, M. E.; Haddon, R. C.
Acc.Chem. Res. 2002, 35, 1105.

Novak, J.P.; Lay, M.D.; Perkins, F.K.; Snow, E.S. Solid-State Electronics, v 48, n 10-11 SPEC.
ISS., October/November, 2004, p 1753-1756

199
Obrzut, J.; Douglas, J.F.; Kharchenko, S.B.; Migler, K.B. Physical Review B, v 76, n 19, 15 Nov.
2007, 195420-1-9

O'Connell, Michael. Carbon nanotubes : properties and applications Publisher: CRC Taylor &
Francis, Pub date: 2006. ISBN: 0849327482

Odegard, G.M.; Gates, T.S.; Wise, K.E.; Park, C.; Siochi, E.J. Composites Science and
Technology, v 63, n 11, August, 2003, p 1671-1687

Odegard, Gregory M.; Gates, Thomas S.; Nicholson, Lee M.; Wise, Kristopher E. Composites
Science and Technology, v 62, n 14, November, 2002, p 1869-1880

Ogihara S, Uehara T. J Mate Sci Lett 2003; 22 (9): 683-5.
Overney, G.; Zhong, W.; Tomanek, D. Zeitschrift fur Physik D, v 27, n 1, 1993, p 93-6

Paiva MC, Zhou B, Fernando KAS, Lin Y, Kennedy JM, Sun YP. Carbon 2004; 42(14):2849.

Paiva MC, Zhou B, Fernando KAS, Lin Y, Lopes PE, Pennington WT, Kennedy JM, Sun YP.
Compos Inter 2005; 12(8-9):757-68.

Paiva MC, Zhou B, Fernando KAS, Lin Y, Lopes PE, Pennington WT, Kennedy JM, Sun YP.
Compos Inter 2005; 12(8-9):757-68.

Paris, Oskar; Zollfrank, Cordt; Zickler, Gerald A. Source: Carbon, v 43, n 1, 2005, p 53-66

Park JS, Park SS, Lee S. Adv Polym Emerg Technol 2007: 568-72.

Park, Cheol; Wilkinson, John; Banda, Sumanth; Ounaies, Zoubeida; Wise, Kristopher E.; Sauti,
Godfrey; Lillehei, Peter T.; Harrison, Joycelyn S. Journal of Polymer Science, Part B: Polymer
Physics, v 44, n 12, Jun 15, 2006, p 1751-1762

Pegel, S.; Potschke, P.; Petzold, G.; Alig, I.; Dudkin, S.M.; Lellinger, D. Polymer, v 49, n 4, 18
Feb. 2008, 974-84

Peng, Huisheng. Journal of the American Chemical Society, v 130, n 1, Jan 9, 2008, p 42-43

Pierard, N.; Fonseca, A.; Colomer, J.-F.; Bossuot, C.; Benoit, J.-M.; Van Tendeloo, G.; Pirard,
J.-P.; Nagy, J.B. Carbon, v 42, n 8-9, 2004, p 1691-1697

200
Poncharal, P.; Wang, Z.L.; Ugarte, D.; De Heer, W.A. Science, v 283, n 5407, 5 March 1999,
1513-16

Premkumar, M.K. (Alcoa Tech. Center, Alcoa Center, PA, USA); Yun, D.I.; Sawtell, R.R.
Proceedings of the SPIE - The International Society for Optical Engineering, v 2105, 1993, p
399-404

Qian, Dong; Wagner, Gregory J.; Liu, Wing Kam; Yu, Min-Feng; Ruoff, Rodney S. Applied
Mechanics Reviews, v 55, n 6, November, 2002, p 495-532

Qian, Hui; Bismarck, Alexander; Greenhalgh, Emile S.; Kalinka, Gerhard; Shaffer, Milo S.P.
Chemistry of Materials, v 20, n 5, Mar 11, 2008, p 1862-1869

Qinghua Zhang; Zhenjun Chang; Meifang Zhu; Xiumei Mo; Dajun Chen. Nanotechnology, v 18,
n 11, 21 March 2007, 6 pp.

Qingzhong Zhao; Marco Buongiorno Nardelli; Bernholc, J. Physical Review B, v 65, n 14, 1
April 2002, p 144105/1-6

Qiu JJ, Zhang C, Wang B, Liang R. Nanotechnology 2007; 18(27):275708.

Rakow, J.F.; Waas, A.M. Mechanics of Advanced Materials and Structures, v 11, n 2,
March-April 2004, 151-73

Raravikar, Nachiket R.; Schadler, Linda S.; Vijayaraghavan, Aravind; Zhao, Yiping; Wei,
Bingqing; Ajayan, Pulickel M. Chemistry of Materials, v 17, n 5, Mar 8, 2005, p 974-983

Reinert, H. S.; Meade, L. E. Roy National SAMPE Technical Conference, v 7, 1975, p 475-487

Ren, Z.F.; Guang-Yong Xiong; Wang, D.Z. Carbon, v 44, n 5, April 2006, 969-73

Robert E. Montgomery, Fred J. Weber, and David F. White, J. Acoust. Soc. Am., 1982. 71(3): p.
735-741.
Robertson, D.H.; Brenner, D.W.; Mintmire, J.W. Physical Review B, v 45, n 21, 1 June 1992, p
12592-5

Rubio, A. (Dept. de Fisica Teorica, Valladolid Univ., Spain) Applied Physics A (Materials
Science Processing), v 68, n 3, March 1999, 275-82

201
Ruoff RS, Lorents DC. Carbon 1995; 33 (7): 925-30.

S. J. V. Frankland, A. Caglar, D. W. Brenner, and M. Griebel, J. Phys. Chem. B 106, 3046 ~2002.

S.L. Mielke, D. Troya, S. Zhang, Li Je-Luen, S. Xiao, R. Car, R.S. Ruoff, .C. Schatz, T.
Beliytschko, Chem. Phys. Lett. 390 (2004) 413–420.

S.Qin, D.Qin, W.T .F ord, J.E.Herrera, D.E.Resasco, S.M. ， Bachilo, and R.B.W eisman,
Macromolecules 37, 3965 (2004).

Saito, R. (Riichiro) Dresselhaus, G. Dresselhaus, M. S. Publisher: Imperial College Press ; Pub
date: 2005, c1998. ISBN: 1860940935

Salvetat, J. P. et al, Appl Phys A: Mater Sci Process 1999. 68: p. 287-92.

Salvetat, J.-P.; Bhattacharyya, S.; Byron Pipes, R. Journal of Nanoscience and Nanotechnology, v
6, n 7, July 2006, 1857-82

Salvetat, J.-P.; Kulik, A.J.; Bonard, J.-M.; Briggs, G.A.D.; Stockli, T.; Metenier, K.; Bonnamy,
S.; Beguin, F.; Burnham, N.A.; Forro, L. Advanced Materials, v 11, n 2, Jan 22, 1999, p 161-165

Salvetat, Jean-Paul; Briggs, G. Andrew D.; Bonard, Jean-Marc; Bacsa, Revathi R.; Kulik,
Andrzej J.; Stockli, Thomas; Burnham, Nancy A.; Forro, Laszlo. Physical Review Letters, v 82, n
5, Feb 1, 1999, p 944-947

Sandler, J.K.W.; Kirk, J.E.; Kinloch, I.A.; Shaffer, M.S.P.; Windle, A.H. Polymer, v 44, n 19,
Aug 21, 2003, p 5893-5899

Satishkumar, B.C.; Govindaraj, A.; Mofokeng, J.; Subbanna, G.N.; Rao, C.N.R. Journal of
Physics B: Atomic, Molecular and Optical Physics, v 29, n 21, Nov 14, 1996, p 4925-4934

Schechter, L.; Wynstra, J. Ind. Eng. Chem. 1956, 48, 86.

Scott, C.D.; Povitsky, A.; Dateo, C.; Gkcen, T.; Willis, P.A.; Smalley, R.E. Journal of
Nanoscience and Nanotechnology, v 3, n 1-2, Feb.-April 2003, 63-73

Sellinger, Alan; Weiss, Pilar M.; Nguyen, Anh; Lu, Yunfeng; Assink, Roger A.; Gong, Weiliang;
Brinker, C. Jeffrey. Nature, v 394, n 6690, Jul 16, 1998, p 256-260

202
Sgriccia N, Hawley MC. Compos Sci Technol 2007; 67 (9): 1986-91.

Shaikh, S.; Li, L.; Lafdi, K.; Huie, J. Carbon, v 45, n 13, Nov. 2007, 2608-13

Shan HZ, Chou TW. Compos Sci Technol 1995; 53(4):383-91.

Sharma, Rajesh; Ranta, Sohan; Ranjan, Ashish. Journal of Reinforced Plastics and Composites, v
24, n 7, 2005, p 719-724

Shen JF, Huang WS, Wu LP, Hu YZ, Ye MX. Composites: A 2007;38:1331-36.

Shen JF, Huang WS, Wu LP, Hu YZ, Ye MX. Compois Sci Technol 2007; 67:3041-50.

Shengdong Li; Zhen Yu; Rutherglen, C.; Burke, P.J. Nano Letters, v 4, n 10, Oct. 2004, 2003-7

Shi DL, Feng XQ, Huang YG, Hwang KC, Gao HJ. Trans. ASME 2004; 126: 250-57.

Shim HC, Lee HW, Yeom S, Kwak YK, Lee SS, Kim SH. Nanotechnology 2007;18(11), Article
no. 115602.

Shu, J.; Kim, J. D.; Peng H.; Margrave, J. L.; Khabashesku, V. N.; Barrera, E. V. Nano Lett.
2003, 3, 1107.

Shull KR. J Chem Phys 1991;94(8): 5723-38

Silva C, Marotta E, Schuller M, Peel L, O'Neill M. J Thermophys Heat Transf 2007; 21
(3):460-7.

Simmons, J.M.; Nichols, B.M.; Baker, S.E.; Marcus, Matthew S.; Castellini, O.M.; Lee, C.-S.;
Hamers, R.J.; Eriksson, M.A. Journal of Physical Chemistry B, v 110, n 14, Apr 13, 2006, p
7113-7118

Sishen Xie; Wenzhi Li; Zhengwei Pan; Baohe Chang; Lianfeng Sun. Journal of the Physics and
Chemistry of Solids, v 61, n 7, July 2000, 1153-8

Sternitzke, M. Journal of the European Ceramic Society, v 17, n 9, 1997, p 1061-82

Strano, Michael S. Nature Materials, v 5, n 6, Jun 10, 2006, p 433-434

203
Strobl, C.J.; Schaflein, C.; Beierlein, U.; Ebbecke, J.; Wixforth, A. Applied Physics Letters, v 85,
n 8, Aug 23, 2004, p 1427-1429

Sun, Y.-P.; Fu, K.; Lin, Y.; Huang, W. Acc. Chem. Res. 2002, 35, 1096.

Sun, Zhenyu; Nicolosi, Valeria; Rickard, David; Bergin, Shane D.; Aherne, Damian; Coleman,
Jonathan N. Journal of Physical Chemistry C, v 112, n 29, Jul 24, 2008, p 10692-10699

T. Natsuki, M. Endo, Carbon 42 (2004) 2147–2151.

Tanaka, K.; Mori, T. Philosophical Magazine, v 23, n 183, March 1971, 737-40

Tasis D, Tagmatarchis N, Bianco A, Prato M, 2006;106:1105-36.

Tasis, D.; Papagelis, K.; Douroumis, D.; Smith, J.R.; Bouropoulos, N.; Fatouros, D.G. Journal of
Nanoscience and Nanotechnology, v 8, n 1, Jan. 2008, 420-3

Theocaris, P.; Paipetis, S.A. Journal of Composite Materials, v 9, n 3, July 1975, p 244-50

Thomas S. Ramotowski, Master’s Thesis, University of Rhode Island (2003)

Thomson, C.M., J. Acoust. Soc. Am., 1990. 87 (3): 1138-1143.

Thostenson ET, Chou TW. J Phys: D 2003; 6(5):573-82.

Thostenson ET, Li WZ, Wang DZ, Ren ZF, Chou TW. J Appl Phys 2002; 91(9):6034-7.

Thostenson ET, Ren ZF, Chou TW. Compos Sci Technol 2001; 61(13):1899-912.

Tiano, T.; Roylance, M.; Gassner, J. 32nd SAMPE Conf. 2000, 192.

Tohji K, Goto T, Takahashi H, Shinoda Y, Shimizu N, Jeyadevan B, et al. Nature
1996;383(6602):679–9.

Tohji K, Takahashi H, Shinoda Y, Shimizu N, Jeyadevan B, Matsuoka I, et al. J Phys Chem B
1997;101(11):1974–8.

Treacy, M.M.J.; Ebbesen, T.W.; Gibson, J.M. Nature, v 381, n 6584, 20 June 1996, 678-80

204
Tsu-Wei Chou; Chunyu Li Source: Composites Science and Technology, v 66, n 14, Nov. 2006,
2409-14

Tucker, C.L., III; Liang, E. Composites Science and Technology, v 59, n 5, 1999, 655-71

Tunvir, Kazi; Kim, Amkee; Nahm, Seung Hoon Nanotechnology, v 19, n 6, 13 February, 2008, p
065703

Twardowski, T.E., and Geil, P.H., J. Acoust. Soc. Am., 1990. 41: p. 1047-1054.

Twardowski, T.E., and Geil, P.H., J. Acoust. Soc. Am., 1991. 42: p. 1721-1726.

Twarsowski, T.E., and Geil, P.H., J. Acoust. Soc. Am., 1991. 42: p. 69-74.

Vasiliev, Valery V. Publisher: Elsevier, Pub date: 2007.

Veedu VP, Cao AY, Li XS, Ma KG, Soldano C, Kar S, Ajayan PM, Ghasemi-Nejhad MN. Nat
Mate 2006; 5(6): 457-62.

Wagner HD, Liu LQ. Compos Sci Technol 2005; 65 (11-12): 1861-68.

Wagner, H.D.; Lourie, O.; Feldman, Y.; Tenne, R. Applied Physics Letters, v 72, n 2, 12 Jan.
1998, 188-90

Wagner, H.D.; Lourie, O.; Feldman, Y.; Tenne, R. Source: Applied Physics Letters, v 72, n 2, 12
Jan. 1998, 188-90

Wall, A.; Coleman, J.N.; Ferreira, M.S. Physical Review B, v 71, n 12, 15 March 2005,
125421-1-5

Walpole, L.J. Mathematical Proceedings of the Cambridge Philosophical Society, v 81, pt.2,
March 1977, 283-9

Wang JG, Fang ZP, Gu AJ, Xu LH, Li, F. J Appl Polym Sci 2006; 100(1):97-104.

Wang L, Dang ZM. Appl Phys Lett 2005;87(4):42903-13.

Wang M, Pramoda KP, Goh SH, Carbon 2006;44: 613-17.

205
Wang S, Liang ZY, Liu T, Wang B, Zhang C. Nanotechnology 2006; 17:1551-57.

Wang, Yu; Fang, Daining; Soh, Ai Kah; Liu, Bin Acta Mechanica Sinica, v 23, n 6, December,
2007, p 663-671

Wei Chen, Hongbin Lu, Steven R. Nutt, Composite Science and Technology, DOI:
10.1016/j.compscitech.2008.05.011, In press, available online 21 May 2008

Wei Chen, Hongbin Lu, Steven R. Nutt. Composites Science and Technology, Volume 68, Issue
12, 2008, P2535-2542

Wei Chen, Hongbin Shen, Maria L. Auad, Steven R. Nutt, submitted to Composites Part B,
(2008), under review.

Wei Chen, Maria L Auad, William R.J.J., Steven R. Nutt, European Polymer Journal, 42 (2006):
2765-2772.

Wei Chen, Maria L. Auad, Roberto J. J. Williams, Steven R. Nutt. European Polymer Journal, 42
(2006) 2765-2772.

Weijie Huang; Yi Lin; Taylor, S.; Gaillard, J.; Rao, A.M.; Ya-Ping Sun. Nano Letters, v 2, n 3,
March 2002, 231-4

Wichmann MHG, Sumfleth J, Gojny FH, Quaresimin M, Fiedler B, Schulte K. Engi Fract Mech
2006;73(16):2346-59.

Windle, A.H.; Moisala, A.; Li, Q.; Kinloch, I.A. Composites Science and Technology, v 66, n 10,
Aug. 2006, 1285-8

Wong M, Paramsothy M, Xu XJ, Ren Y, Li S, Liao K. Polymer 2003;44(25):7757-64.

Wong, M.; Paramsothy, M.; Xu, X.J.; Ren, J.; Li, S.; Liao, K. Polymer, v 44, n 25, Dec. 2003,
7757-64
Wu HX, Tong R, Qiu XQ, Yang HF, Lin YH, Cai RF, Qian SX. Carbon 2007; 45:152-59.

Xia Z, Riester L, Curtin WA, Li H, Sheldon BW, Liang J, Chang B, Xu JM. Acta Mater 2004;
52(4): 931-44.

206
Xia, Hesheng; Wang, Qi; Li, Kanshe; Hu, Guo-Hua Source: Journal of Applied Polymer Science,
v 93, n 1, Jul 5, 2004, p 378-386

Xie L, Xu F, Qiu F, Lu HB, Yang YL. Macromolecules 2007; 40(9): 3296-305

Xu-Ming Xie; Yi-Tao Liu; Yan-Fang Gao; Qing-Ping Feng; Lin-Rui Guo; Xiao-Hao Wang;
Xiong-Ying Ye. Carbon, v 44, n 3, March 2006, 599-602

Xu-Ming Xie; Yi-Tao Liu; Yan-Fang Gao; Qing-Ping Feng; Lin-Rui Guo; Xiao-Hao Wang;
Xiong-Ying Ye. Carbon, v 44, n 3, March 2006, 599-602

Yamamoto K, Akita S, Nakayama Y. Appl Phys 1998;31(8):L34–6.

Yamamoto, Tatsuhiro; Miyauchi, Yuhei; Motoyanagi, Jin; Fukushima, Takanori; Aida, Takuzo;
Kato, Masaru; Maruyama, Shigeo. Japanese Journal of Applied Physics, Part 1: Regular Papers
and Short Notes and Review Papers, v 47, n 4 PART 1, Apr 18, 2008, p 2000-2004

Yang J, Lin YH, Wang JF, Lai MF, Li J, Liu JJ, Tong X, Cheng HM, J Appl Polym Sci 2005;
98(3):1087-91.

Yang, Mingjun; Koutsos, Vasileios; Zaiser, Michael Source: Journal of Physical Chemistry B, v
109, n 20, May 26, 2005, p 10009-10014

Yao, N.; Lordi, V. Journal of Applied Physics, v 84, n 4, 15 Aug. 1998, 1939-43

Yi, Wenhui; Malkovskiy, Andrey; Chu, Qinghui; Sokolov, Alexei P.; Colon, Marisabel Lebron;
Meador, Michael; Pang, Yi. Journal of Physical Chemistry B, v 112, n 39, Oct 2, 2008, p
12263-12269

Yoong Ahm Kim; Hayashi, T.; Endo, M.; Gotoh, Y.; Wada, N.; Seiyama, J. Scripta Materialia, v
54, n 1, Jan. 2006, 31-5

Young Hee Lee; Kay Hyeok An; Seong Chu Lim; Won Seok Kim; Hee Jin Jeong; Chil-Hoon
Doh; Seong-In Moon. New Diamond and Frontier Carbon Technology, v 12, n 4, 2002, 209-28

Young-Jae Lee; Hyun-Ha Kim; Hatori, H. Carbon, v 42, n 5-6, 2004, 1053-6

Yu M, Lourie O, Dyer MJ, Moloni K, Kelly TF, Ruoff RS. Science 2000; 287:637-40.

207
Yu MF, Lourie O, Dyer MJ, Moloni K, Kelly TF, Ruoff RS. Science 2000; 287(5453): 637-40.

Yu, Min-Feng, Journal of Engineering Materials and Technology, Transactions of the ASME, v
126, n 3, July, 2004, p 271-278

Yudasaka, M.; Zhang, M.; Jabs, C.; Iijima, S. Applied Physics A: Materials Science &
Processing, v 71, n 4, Oct, 2000, p 449-451

Yudasaka, Masako; Kikuchi, Rie; Matsui, Takeo; Ohki, Yoshimasa; Yoshimura, Susumu; Ota,
Etsuro. Applied Physics Letters, v 67, n 17, Oct 23, 1995, p 2477

Zawadzki J. Chemistry and physics of carbon, v21, Torun, Poland: Institute of chemistry,
Nicolaus Copernicus University;1989.

Zeng LL, Zhang L , Barron AR. Nano Lett 2005; 5(10):2001-04.

Zhang J, Zou HL, Qing Q, Yang YL, Li QW, Liu ZF, Guo XY, Du ZL. J Phys Chem B
2003;107:3712-18.

Zhang QH, Rastogi S, Chen DJ, Lippits D, Lemstra PJ. Carbon 2006;44(4):778-85.

Zhang, Zhinan; Zhang, Jun; Chen, Peng; Zhang, Baoqing; He, Jiasong; Hu, Guo-Hua. Carbon, v
44, n 4, April, 2006, p 692-698

Zhong Jin; Haibin Chu; Jinyong Wang; Jinxing Hong; Wenchang Tan; Yan Li. Nano Letters, v 7,
n 7, July 2007, 2073-9

208
Appendix: Publications

[1] Wei Chen*, Shankar Rajaram, Shad Thomas, Steven R. Nutt. Carbon nanotube reinforced
acoustically transparent materials for underwater sonar windows, Journal of Advanced Materials,
accepted, Scheduled for publication in May 2009.

[2] Wei Chen*, Hongbin Shen, Maria L. Auad, Changzheng Huang, Steven Nutt. Basalt Fiber –
Epoxy Laminates with Functionalized Multi-walled Carbon Nanotubes, submitted to Composites
Part A, under review, 2009.

[3] Wei Chen*, Hongbin Lu, Steven R. Nutt. The influence of functionalized MWCNT
reinforcement on the thermomechanical properties and morphology of epoxy nanocomposites,
Composites Science and Technology, Volume 68, Issue 12, 2008, P2535-2542.

[4] Wei Chen*, Maria L. Auad, Roberto J. J. Williams, Steven R. Nutt. Improving the dispersion
and flexural strength of multiwalled carbon nanotubes–stiff epoxy composites through
β-hydroxyester surface functionalization coupled with the anionic homopolymerization of the
epoxy matrix, European Polymer Journal, 42 (2006) 2765-2772.

[5] Wei Chen*, Shankar Rajaram, Steven Nutt. Epoxy Nanocomposites with ultra-low
multiwalled carbon nanotube concentrations for underwater acoustic applications, SAMPE 2008,
Technical Paper (peer reviewed), Long Beach, CA, May 18 - 22, 2008.

Abstract (if available)

Abstract The emergence of carbon nanotubes (CNTs) has led to myriad possibilities for structural polymer composites with superior specific modulus, strength, and toughness. While the research activities in carbon nanotube reinforced polymer composites (NRPs) have made enormous progress towards fabricating next-generation advanced structural materials with added thermal,optical, and electrical advantages, questions concerning the filler dispersion, interface, and CNT alignment in these composites remain partially addressed. In this dissertation, the key technical challenges related to the synthesis, processing, and reinforcing mechanics governing the effective mechanical properties of NRPs were introduced and reviewed in the first two chapters.Subsequently, issues on the dispersion, interface control, hierarchical structure, and multi-functionality of NRPs were addressed based on functionalized multi-walled carbon nanotube reinforced DGEBA epoxy systems (NREs). In chapter 3, NREs with enhanced flexuralproperties were discussed in the context of improved dispersion and in-situ formation of covalent bonds at the interface. In chapter 4, NREs with controlled interface and tailored thermomechanical properties were demonstrated through the judicious choice of surface functionality and resin chemistry. In chapter 5, processing-condition-induced CNT organization in hierarchical epoxy nanocomposites was analyzed. In Chapter 6, possibilities were explored for multi-functional NREs for underwater acoustic structural applications. Finally, the findings of this dissertation were concluded and future research was proposed for ordered carbon nanotube array reinforced nanocomposites in the last chapter. Four journal publications resulted from this work are listed in Appendix.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Printed and flexible carbon nanotube macroelectronics

PDF

Nanoclay-reinforced polyurethane

PDF

Chemical recycling of amine/epoxy composites at atmospheric pressure

PDF

Fiber-reinforced syntactic foams

PDF

Printed electronics based on carbon nanotubes and two-dimensional transition metal dichalcogenides

PDF

Additive manufacturing of piezoelectric and composite for biomedical application

PDF

Fiber reinforced hybrid phenolic foam

PDF

Theory and applications of ultrasonic soft materials and surface analyzer (USMASA)

PDF

Carbon nanotube nanoelectronics and macroelectronics

PDF

Light Emission from Carbon Nanotubes and Two-Dimensional Materials

PDF

Hydrogen storage in carbon and silicon carbide nanotubes

PDF

Electronic and optoelectronic devices based on quasi-metallic carbon nanotubes

PDF

In situ process monitoring for modeling and defect mitigation in composites processing

PDF

Carbon nanotube macroelectronics

PDF

Nanomaterials for macroelectronics and energy storage device

PDF

Analysis of carbon nanotubes using nanoelectromechanical oscillators

PDF

GaAs nanowire optoelectronic and carbon nanotube electronic device applications

PDF

Perfluorocarbon mediated self-assembly of polymers

PDF

Studies of nanosecond pulsed power for modifications of biomaterials and nanomaterials (SWCNT)

PDF

Plasmonic enhancement of catalysis and solar energy conversion

Asset Metadata

Creator Chen, Wei (author)

Core Title Multiwall carbon nanotubes reinforced epoxy nanocomposites

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Materials Science

Publication Date 04/14/2009

Defense Date 01/23/2009

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag carbon nanotube,epoxy,interface,nanocomposites,OAI-PMH Harvest,polymer,reinforcement

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Nutt, Steven R. (committee chair), Shing, Katherine S. (committee member), Thieo, Hogen Esch (committee member)

Creator Email nuaa215@hotmail.com,weichen@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m2087

Unique identifier UC1450036

Identifier etd-Chen-2687 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-230545 (legacy record id),usctheses-m2087 (legacy record id)

Legacy Identifier etd-Chen-2687.pdf

Dmrecord 230545

Document Type Dissertation

Rights Chen, Wei

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu