Composites: Part A 41 (2010) 1345–1367

Contents lists available at ScienceDirect

Composites: Part A
journal homepage: www.elsevier.com/locate/compositesa

Review

Dispersion and functionalization of carbon nanotubes for polymer-based

nanocomposites: A review
Peng-Cheng Ma a, Naveed A. Siddiqui a, Gad Marom b, Jang-Kyo Kim a,*
a
Department of Mechanical Engineering, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, China
b
Casali Institute of Applied Chemistry, The Institute of Chemistry, The Hebrew University of Jerusalem, Jerusalem 91904, Israel

a r t i c l e i n f o a b s t r a c t

Article history: Carbon nanotubes (CNTs) hold the promise of delivering exceptional mechanical properties and multi-
Received 26 March 2010 functional characteristics. Ever-increasing interest in applying CNTs in many different fields has led to
Received in revised form 28 June 2010 continued efforts to develop dispersion and functionalization techniques. To employ CNTs as effective
Accepted 10 July 2010
reinforcement in polymer nanocomposites, proper dispersion and appropriate interfacial adhesion
between the CNTs and polymer matrix have to be guaranteed. This paper reviews the current under-
standing of CNTs and CNT/polymer nanocomposites with two particular topics: (i) the principles and
Keywords:
techniques for CNT dispersion and functionalization and (ii) the effects of CNT dispersion and function-
A. Polymer–matrix composites (PMCs)
A. Nano-structures
alization on the properties of CNT/polymer nanocomposites. The fabrication techniques and potential
E. Surface treatments applications of CNT/polymer nanocomposites are also highlighted.
Carbon nanotubes Ó 2010 Elsevier Ltd. All rights reserved.

1. Carbon nanotubes (CNTs) and their properties
1.1. Introduction to CNTs
Different from other carbon materials, such as graphite, dia-
mond and fullerene (C60, C70, etc.), CNTs are one dimensional car-
bon materials which can have an aspect ratio greater than 1000.
They can be envisioned as cylinders composed of rolled-up graph-
ite planes with diameters in nanometer scale [1–4]. The cylindrical
nanotube usually has at least one end capped with a hemisphere of
fullerene structure.
Depending on the process for CNT fabrication, there are two
types of CNTs [2–5]: single-walled CNTs (SWCNTs) and multi-
walled CNTs (MWCNTs). SWCNTs consist of a single graphene layer
(Fig. 1A) rolled up into a seamless cylinder whereas MWCNTs con-
sist of two or more concentric cylindrical shells of graphene sheets
(Fig. 1B) coaxially arranged around a central hollow core with van
der Waals forces between adjacent layers.
According to the rolling angle of the graphene sheet, CNTs have
three chiralities: armchair, zigzag and chiral one. The tube chirality
is defined by the chiral vector, Ch = na1 + ma2 (Fig. 2), where the inte-
gers (n, m) are the number of steps along the unit vectors (a1 and a2)
of the hexagonal lattice [3,4]. Using this (n, m) naming scheme, the
three types of orientation of the carbon atoms around the nanotube
circumference are specified. If n = m, the nanotubes are called ‘‘arm-
chair”. If m = 0, the nanotubes are called ‘‘zigzag”. Otherwise, they
are called ‘‘chiral”. The chirality of nanotubes has significant impact
on their transport properties, particularly the electronic properties.
For a given (n, m) nanotube, if (2n + m) is a multiple of 3, then the
nanotube is metallic, otherwise the nanotube is a semiconductor.
Each MWCNT contains a multi-layer of graphene, and each layer
can have different chiralities, so the prediction of its physical prop-
erties is more complicated than that of SWCNT.
1.2. Properties of CNTs
The chemical bonding of CNTs is composed entirely of sp2 car-
bon–carbon bonds. This bonding structure – stronger than the
sp3 bonds found in diamond – provides CNTs with extremely high
mechanical properties. It is well known that the mechanical prop-
erties of CNTs exceed those of any existing materials [5–7].
Although there is no consensus on the exact mechanical properties
of CNTs, theoretical and experimental results have shown unusual
mechanical properties of CNTs with Young’s modulus as high as
1.2 TPa and tensile strength of 50–200 GPa [6]. These make CNTs
the strongest and stiffest materials on earth.
In addition to the exceptional mechanical properties associated
with CNTs, they also possess other useful physical properties and

* Corresponding author. Tel.: +852 2358 7207; fax: +852 2358 1543.
E-mail address: mejkkim@ust.hk (J.-K. Kim).

1359-835X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.compositesa.2010.07.003
1346 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367

Fig. 1. TEM images of different CNTs (A: SWCNTs; B: MWCNTs with different layers of 5, 2 and 7). After [1,2].

Fig. 2. Schematic diagram showing how a hexagonal sheet of graphene is rolled to form a CNT with different chiralities (A: armchair; B: zigzag; C: chiral). After [3,4].

Table 1
Physical properties of different carbon materials.

C material
Property Graphite Diamond Fullerene SWCNT MWCNT
Specific gravity (g/cm3) 1.9–2.3 3.5 1.7 0.8 1.8
Electrical conductivity (S/cm) 4000p, 3.3c 102–1015 105 102–106 103–105
Electron mobility (cm2/ (V s)) 2.0  104 1800 0.5–6 105 104–105
Thermal conductivity (W/(m K)) 298p, 2.2c 900–2320 0.4 6000 2000
Coefficient of thermal expansion (K1) 1  106p (13)106 6.2  105 Negligible Negligible
2.9  105c
Thermal stability in air (°C) 450–650 <600 600 >600 >600

p: in-plane; c: c-axis.

Table 1 summarizes these properties [8–10]. It is clear that CNTs
have many advantages over other carbon materials in terms of
electrical and thermal properties. These properties offer CNTs great
potential for wide applications in field emission, conducting plas-
tics, thermal conductors, energy storage, conductive adhesives,
thermal interface materials, structural materials, fibers, catalyst
supports, biological applications, air and water filtration, ceramics
and so on [4,9,10].
2. Dispersion of CNTs
2.1. Nature of dispersion problem for CNTs
Many research efforts have been directed towards producing
CNT/polymer composites for functional and structural applica-
tions [4,9–10]. However, even after a decade of research, the full
potential of employing CNTs as reinforcements has been severely
 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367
Table 2
Dimension and corresponding number of particles in composites for different fillers.
Description
Filler Average dimension Density Na
of filler (g/cm3)
Al2O3 100 lm in diameter (d) 4.0 1.9
particle
Carbon 5 lm in diameter 2.25 255
fiber (d)  200 lm in length (l)
GNP 45 lm in length (square, 2.2 6.58  104
l), 7.5 nm in thickness (t)
CNT 12 nm in diameter 1.8 4.42  108
(d)  20 lm in length (l)
a area between the Al2O3 and CNT particles: for example, the total
N: number of particles in 1.0 mm3 with 0.1 vol.% filler content.
b
S: surface area of individual particles.
limited because of the difficulties associated with dispersion of
entangled CNT during processing and poor interfacial interaction
between CNTs and polymer matrix. The nature of dispersion
problem for CNTs is rather different from other conventional fill-
ers, such as spherical particles and carbon fibers, because CNTs
are characteristic of small diameter in nanometer scale with high
aspect ratio (>1000) and thus extremely large surface area. In
addition, the commercialized CNTs are supplied in the form of
heavily entangled bundles, resulting in inherent difficulties in
dispersion.
Table 2 compares the dimensions of commonly used fillers,
including Al2O3 particles, carbon fibers, graphite nanoplatelets
(GNPs) and CNTs and the number of particles corresponding to a
uniform filler volume fraction of 0.1% in a composite of 1.0 mm3
cube. Because there are large differences in dimension and geom-
etry of these four different reinforcements, the number of fillers
contained for a given filler volume fraction will also be largely
varying. For example, there are only two pieces of Al2O3 particles
and the number increases to two hundred when carbon fiber is
added in the composite with the same filler volume fraction
whereas there are some 442 million pieces of CNTs. This observa-
tion would shed light on why the dispersion of CNTs in a polymer
matrix is more difficult than the other fillers.
The three-dimensional distribution of these micro- and nano-
scale fillers in a polymer matrix is schematically shown in Fig. 3,
which gives a vivid impression on the different distribution
behavior of particles in the matrix due to the size and geometry
effects. The distribution of micro-scale fillers (A and B in Fig. 3) is
homogeneous throughout the matrix, and the differentiation of
individual particles in a matrix can be done easily. However,
when GNPs and CNTs are filled into the same volume of matrix,
it is hard to disperse individual particles uniformly (C and D in
Fig. 3). Considering the particle agglomeration due to the electro-
static interaction and van der Waals force, the real distribution of
Fig. 3. Distribution of micro- and nano-scale fillers of the same 0.1 vol.% in a reference volume of 1 mm3 (A: Al2O3 particle; B: carbon fiber; C: GNP; D: CNT).
1347
nano-scale fillers should be more complicated than the schemat-
ics shown here.
The large quantity of particles and their size effect will lead to
Sb an exceptionally large surface area of nano-scale fillers in the com-
posite. Fig. 4 shows the theoretical surface area of fillers plotted as
S = pd2
a function of volume fraction of filler under the assumption that
fillers are dispersed uniformly in the composite. The bulk density
S = pdl + and formula used to calculate the surface area of filler are listed
pd2/2 in Table 2. For the filler with size less than 100 nm, such as CNT
S = 4l2 +
and GNP, the surface areas are drastically higher than their coun-
2lt
S = pdl + terparts with size in micrometer scale, such as Al2O3 particles
pd2/2 and carbon fiber. There is about 500 times difference in surface
surface areas are 6.0  108 and 2.8  105 m2, respectively, for
the Al2O3 and CNT particles of the same 0.1 vol.% in a cube of
1.0 mm3. A large surface area of fillers means a large interface or
interphase area present between the filler and matrix. A classic
definition of the ‘‘interface” in composites is a surface formed by
a common boundary of reinforcing fillers and matrix that is in con-
tact and maintains the bond in between for load transfer [11]. In
contrast, the ‘‘interphase” is defined as the region with altered
chemistry, altered polymer chain mobility, altered degree of cure
and altered crystallinity that are unique from those of the filler
or the matrix. The interphase thickness of conventional fiber com-
posites containing carbon or glass fibers is known to be on the
magnitude of one to a few microns [12,13]. Meanwhile, the inter-
phase size of CNT/polymer–matrix composites has been reported
to be as large as about 500 nm according to the size and dimension
of fillers [4]. Even if the interfacial region is only a few nanometers
thick, the CNTs would poise tremendous problems in uniform
dispersion.
Fig. 4. Total surface area of fillers in composites with varying filler contents.
1348 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367
Fig. 5. Electronic microscope images of different CNTs: (A) TEM image of SWCNT bundle (After [14]); (B) SEM image of entangled MWCNT agglomerates.
In addition to the size effect of fillers, the physical nature of par-
ticles also plays an important role in dispersing them into polymer
matrix. As-produced CNTs are held together in bundles or entan-
glements consisting of 50 to a few hundred individual CNTs by
van der Waals force. Fig. 5 shows electronic microscope images
of CNT bundles [14] and entanglements. It has been proved that
these bundles and agglomerates result in diminished mechanical
and electrical properties of composites as compared with theoret-
ical predictions related to individual CNTs [4,10–11]. The challenge
is thus: how to incorporate individual CNTs, or at least relatively
thin CNT bundles or disentangled CNTs, inside a polymer matrix.
In other words, dispersion of CNTs is not only a geometrical prob-
lem, dealing with the length and size of the CNTs, but also relates
to a method on how to separate individual CNTs from CNT agglom-
erates and stabilize them in polymer matrix to avoid secondary
agglomeration.
2.2. Mechanical dispersion of CNTs
As discussed previously, the incorporation of CNTs into a poly-
mer matrix leads to an exceptionally large quantity of particles and
high surface area of fillers, resulting in the difficulties to uniformly
disperse these particles. There is a sizable volume of literature re-
ported on the techniques developed for CNT dispersion in polymer
matrix [15–20]. However, little information has been hitherto
reported on the principles and features of these dispersion
techniques. Therefore, the basis of CNT dispersion methods are
introduced here, some typical results on CNT dispersion in poly-
mers by employing these methods are demonstrated, and a general
Fig. 6. Sonicators with different modes for CNT dispersion (A: water bath sonicator; B: probe/horn sonicator), and the effect of sonication on the structure of CNTs (C: Raman
spectra of CNTs before and after sonication. After [22]).
guideline for selecting proper techniques for CNT dispersion is
provided.
2.2.1. Ultrasonication
Ultrasonication is the act of applying ultrasound energy to agi-
tate particles in a solution for various purposes. In the laboratory, it
is usually achieved using an ultrasonic bath or an ultrasonic probe/
horn (A and B in Fig. 6), also known as a sonicator. It is the most
frequently used method for nanoparticle dispersion. The principle
of this technique is that when ultrasound propagates via a series
of compression, attenuated waves are induced in the molecules
of the medium through which it passes. The production of these
shock waves promotes the ‘‘peeling off” of individual nanoparticles
located at the outer part of the nanoparticle bundles, or agglomer-
ates, and thus results in the separation of individualized nanopar-
ticles from the bundles [21].
Ultrasonication is an effective method to disperse CNTs in liq-
uids having a low viscosity, such as water, acetone and ethanol.
However, most polymers are either in a solid or viscous liquid
state, which requires the polymer to be dissolved or diluted using
a solvent to reduce the viscosity before dispersion of CNTs.
Standard laboratory sonicators (in a water bath) run at 20–
23 kHz with a power less than 100 W. Commercial probe sonica-
tors have an adjustable amplitude ranging from 20% to 70% and a
power of 100–1500 W. The probe is usually made of an inert metal
such as titanium. Most probes are attached with a base unit and
then tapered down to a tip with a diameter from 1.6 to 12.7 mm
[21]. This means that the energy from the wide base is focused
on the tip, thus giving the probe high intensity. The consequence
 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367 1349

A B
Feeding material

Feeding roller Apron roller Material

collection
Fig. 7. Calendering (or three roll mills) machine used for particle dispersion into a polymer matrix (A) and corresponding schematic showing the general configuration and its
working mechanism (B). After [24].

of this configuration is that sonication can generate substantial Calendering is widely used to mix electronic thick film inks,
heat rapidly. Therefore, for the CNTs dispersed in volatile solvents, high performance ceramics, carbon/graphite, paints, printing inks,
such as ethanol and acetone, the samples should be kept cold (e.g. pharmaceuticals, chemicals, glass coatings, dental composites, pig-
using an ice bath) and the sonication must be done in short ment, coatings, adhesives, sealants, and foods. Typical calendaring
intervals. machine and the principle are shown schematically in Fig. 7. The
If the sonication treatment is too aggressive and/or too long, employment of a calender to disperse CNTs in a polymer matrix
CNTs can be easily and seriously damaged, especially when a probe has become a very promising approach to achieve a relatively good
sonicator is employed. The Raman spectroscopy (Fig. 6C) con- CNT dispersion according to some recent reports [25,26]. A high
firmed that ultrasonication of CNTs for a long time resulted in a shear stress is required to apply to disentangle CNT bundles and
significant increase in the intensity of D band (representing disor- distribute the dispersed CNT into polymer matrix, while a short
dered sp3 carbon on CNTs), suggesting the generation of defects on residence time will likely limit the breakage of individual nano-
CNT surface [22]. In extreme cases, the graphene layers of CNTs are tubes [26].
completely destroyed and the nanotubes are converted into amor- However, there are several concerns in using this technique for
phous carbon nanofibers [23]. The localized damage to CNTs dete- CNT dispersion: for example, the minimum gap between the rollers
riorates both the electrical and mechanical properties of the CNT/ is about 1–5 lm, which is comparable to the length of CNTs, but is
polymer composites. much larger than the diameter of individual CNTs. These dimen-
sional disparities between the roller gap and the CNT dimensions
may suggest that calendaring can better disperse the large agglom-
erated CNTs into small ones at sub-micron level, although some
2.2.2. Calendering process
individual CNTs may be disentangled out from the agglomerates
The calender, or commonly known as three roll mills, is a ma-
(Fig. 8). In addition, the feeding materials should be in the viscous
chine tool that employs the shear force created by rollers to mix,
state when mixing with nanoparticles, thus this tool may not be
disperse or homogenize viscous materials. This method has also
applied for dispersing CNTs into thermoplastic matrices, such as
been used to disperse color pigments for cosmetics and lacquers.
polyethylene, polypropylene and polystyrene. In contrast, CNTs
The general configuration of a calendering machine consists of
three adjacent cylindrical rollers each of which runs at a different
velocity (Fig. 7A) [24–26]. The first and third rollers, called the
feeding and apron rollers (n1 and n3 in Fig. 7B), rotate in the same
direction while the center roller rotates in the opposite direction.
The material to be mixed is fed into the hopper, where it is drawn
between the feed and center rollers. When pre-dispersed, the
material sticks to the bottom of the center roller, which transports
it into the second gap. In this gap, the material is dispersed to the
desired degree of fineness. Upon exiting, the material that remains
on the center roller moves through the second nip between the
center roller and apron roller, which subjects it to even higher
shear force due to the higher speed of the apron roller. A knife
blade then scrapes the processed material off the apron roller
and transfers it to the apron. This milling cycle can be repeated
several times to maximize dispersion. The narrow gaps (controlla-
ble from 500 to about 5 lm) between the rollers, combined with
the mismatch in angular velocity of the adjacent rollers, result in
locally high shear forces with a short residence time. One of the un-
ique advantages of this technique is that the gap width between
the rollers can be mechanically or hydraulically adjusted and
maintained, thus it is easy to get a controllable and narrow size
distribution of particles in viscous materials. In some operations, Fig. 8. TEM images of CNTs dispersed by calendering machine in an epoxy matrix.
the width of gaps can be decreased gradually to achieve the desired Note that dispersed CNTs and the aggregated ones co-existed in the nanocompos-
level of particle dispersion [24]. ites. After [26].
1350 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367
can be conveniently dispersed into the liquid monomer or oligo-
mer of thermosetting matrices, and nanocomposites can be ob-
tained via the in situ polymerization.
2.2.3. Ball milling
Ball milling is a type of grinding method used to grind materials
into extremely fine powder for use in paints, pyrotechnics and
ceramics. During the milling, a high pressure is generated locally
due to the collision between the tiny, rigid balls in a concealed con-
tainer, which is schematically shown by Fig. 9. An internal cascad-
ing effect of balls reduces the material to fine powder. Different
materials, including ceramic, flint pebbles and stainless steel, are
used as balls.
Industrial ball mills can operate continuously to feed materials
at one end and discharge them at the other end. High-quality ball
mills can grind mixture particles to as small as 100 nm, enor-
mously increasing the surface area of the materials. Ball milling
has been successfully applied to carbon materials for a variety of
purposes, e.g. to transform CNTs into nanoparticles [27], to gener-
ate highly curved or closed shell carbon nanostructures from
graphite [28], to enhance the saturation of lithium composition
in SWCNTs [29], to modify the morphologies of cup-stacked CNTs
[30], and to generate different carbon nanoparticles from graphitic
carbon for hydrogen storage applications [31].
Ball milling of CNTs in the presence of chemicals not only en-
hances their dispersibility, but also introduces some functional
groups onto the CNT surface. A simple chemo-mechanical method
was used to achieve in situ amino functionalization of CNTs using
ball milling [32,33]. The results showed that the CNTs milled with
ammonium bicarbonate (NH4HCO3) were more effectively disen-
tangled and shortened than those without the chemical (Fig. 10),
and the CNT length was controlled by choosing an appropriate
milling time. Some useful functional groups, like amine and amide
Fig. 9. Schematics of ball milling technique (A) and container (B).
Fig. 10. TEM images of pristine CNTs and those milled after 9 h (A: pristine; B: milled with NH4HCO3; C: milled without NH4HCO3). CNTs milled with NH4HCO3 were more
effectively disentangled and shortened than those without the chemical. After [32].
terminals, were attached onto the CNT surface after milling, allow-
ing covalent bonding to polymers and biological systems. These
nitrogen atoms introduced on CNT surface function as the electron
donors in the material, resulting in the conversion of the semicon-
ducting behavior of CNTs from p-type to n-type [32]. The electrical
conductivity of the treated CNTs also increased gradually with ball
milling time, presenting a remarkable 250% increase after 9 h of
milling.
2.2.4. Stir and extrusion
Stir is a common technique to disperse particles in liquid sys-
tems and can be used as well to disperse CNTs in a polymer matrix.
Size and shape of the propeller and the mixing speed control the
dispersion results. After intensive stirring of CNTs in polymer ma-
trix, a relatively fine dispersion can be achieved [34]. MWCNTs can
be dispersed more easily than SWCNTs by employing this tech-
nique, although the MWCNTs tend to re-agglomerate. This experi-
mentally observed behavior is mainly caused by frictional contacts
and elastic interlocking mechanisms [35]. Other parameters, like
sliding forces and weak attractive forces have only little effect on
this tendency during stir, however, the agglomeration becomes
spontaneous under static conduction. For some thermosetting
polymers, such as epoxy, obvious CNT re-agglomerations were ob-
served after several hours of curing reaction [36]. In case of se-
verely agglomerated CNTs, higher shear forces are needed to
achieve a fine dispersion in the polymer matrix, this can be accom-
plished by employing high speed shear mixer (Fig. 11A) at a speed
up to 10,000 rpm.
Extrusion is a popular technique to disperse CNTs into solid
polymers like most thermoplastics, where thermoplastic pellets
mixed with CNTs are fed into the extruder hopper. Twin screws ro-
tate at a high speed creates high shear flow (Fig. 11B), resulting in
dispersion of CNT agglomerates and mixing them with the polymer
 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367
Fig. 11. Shear mixer (A) and extruder (B) used for CNT dispersion.
melt. This technique is particularly useful to produce CNT/polymer
nanocomposites with a high filler content [37].
Table 3 compares the characteristics of various CNT dispersion
techniques, which can serve as a general guideline for the selection
of appropriate dispersion technique to prepare CNT/polymer nano-
composites. It should be noted that the techniques for CNT disper-
sion are not limited to those described above. Many of the recent
studies are often based on the use of a combination of aforemen-
tioned techniques, such as ultrasonication plus ball milling [33],
and ultrasonication plus extrusion [38,39]. In addition, researchers
should bear in mind that there is no almighty tool to achieve per-
fect dispersion of different CNTs in different types of polymer
matrices. Many factors, such as physical (solid or liquid) and chem-
ical (thermoplastic or thermoset) states of polymer matrix, dimen-
sions and content of CNTs to be added, availability of techniques
and fabrication processes, should be taken into account when
selecting a proper technique for CNT dispersion.
3. Functionalization of CNTs
The performance of a CNT/polymer nanocomposite depends on
the dispersion of CNTs in the matrix and interfacial interactions be-
tween the CNT and the polymer. However, the carbon atoms on
CNT walls are chemically stable because of the aromatic nature
of the bond. As a result, the reinforcing CNTs are inert and can
interact with the surrounding matrix mainly through van der
Waals interactions, unable to provide an efficient load transfer
across the CNT/matrix interface. Therefore, significant efforts have
been directed towards developing methods to modify surface
properties of CNTs. There are several comprehensive review papers
that describe the chemistry of functionalized CNTs and the reaction
Table 3
Comparison of various techniques for CNT dispersion in polymer composites.
Technique Factor
Damage to CNTs Suitable polymer matrix
Ultrasonication Yes Soluble polymer, low viscous
polymer or oligomer, monomer
Calendering No. CNTs may be Liquid polymer or oligomer,
aligned in matrix monomer
Ball milling Yes Powder (polymer or monomer)
Shear mixing No Soluble polymer, low viscous
polymer or oligomer, monomer
Extrusion No Thermoplastics
1351
mechanisms between the CNTs and functional groups [40–43].
These methods can be conveniently divided into chemical func-
tionalization and physical methods based on the interactions be-
tween the active molecules and carbon atoms on the CNTs.
Major principles of these methods along with the corresponding
advantages and disadvantages are summarized in Table 4. While
many studies have hitherto been initiated to modify the CNT sur-
face characteristics based on the below techniques, there are many
processes and material variables that have not been fully opti-
mized. In addition, the concerns that structural changes resulting
from the chemical functionalization of CNTs and the damaging ef-
fects of ultrasonication and other dispersion and mixing processes
still exist because of inevitable adverse effects.
3.1. Chemical functionalization
Chemical functionalization is based on the covalent linkage of
functional entities onto carbon scaffold of CNTs. It can be per-
formed at the termini of the tubes or at their sidewalls. Direct
covalent sidewall functionalization is associated with a change of
hybridization from sp2 to sp3 and a simultaneous loss of p-conju-
gation system on graphene layer (Fig. 12A). This process can be
made by reaction with some molecules of a high chemical reactiv-
ity, such as fluorine. It was shown that the fluorination of purified
SWCNTs occurred at temperatures up to 325 °C and the process
was reversible using anhydrous hydrazine that could remove the
fluorine [44]. The fluorinated CNTs have C–F bonds that are weaker
than those in alkyl fluorides [45] and thus providing substitution
sites for additional functionalization [46]. Successful replacements
of the fluorine atoms by amino, alkyl and hydroxyl groups have
been achieved [47,48]. Besides sidewall fluorination of CNTs, other
Governing factors Availability
Power and mode of sonicator, Commonly used in lab, easy operation and
sonication time cleaning after use
Rotation speed, distance between Operation training is necessary, hard to clean
adjacent rolls after use
Milling time, rotation speed, size of Easy operation, need to clean after use
balls, balls/CNT ratio
Size and shape of the propeller, Commonly used in lab, easy operation and
mixing speed and time cleaning after use
Temperature, configuration and Large-scale production, operation training is
rotation speed of the screw necessary, hard to clean after use
1352 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367

Table 4
Advantages and disadvantages of various CNT functionalization methods.

Method Principle Possible damage to Easy to Interaction with polymer Re-agglomeration of CNTs in
CNTs use matrixa matrix
p p
Chemical Side wall Hybridization of C atoms from  S
method sp2 to sp3
p p p
Defect Defect transformation S
p
Physical Polymer Van der Waals force, p–p  V 
method wrapping stacking
p
Surfactant Physical adsorption  W 
adsorption
p
Endohedral Capillary effect   W
method
a
S: Strong; W: Weak; V: Variable according to the miscibility between matrix and polymer on CNT.

NH2 Amino F
substitution O

=
NH-R-NH2
RMgBr/RLi Amidation
R
O

=
Fluorination and derivative reactions OR
Esterification
o
T>450 C
H O
Dehydrogenation Metal ions
SWCNT or = OH
O
(Au, Ag, etc)

=
Hydrogenation A B SH-R
MWCNT
Cl Thiolation : Metal
ROOC-N Cl nanoparticles
Nitrene Carbene O

=
EtOOC Si-R Silanization
EtOOC n O

=
Nucleophilic Diels-Alder O-(M-M)n
cyclopropanation Grafting to 1
Cycloaddition
Polymer
grafting (M1)n
R
R: Alkyl, aryl, etc Grafting from2
Radical (R·) attachment Monomer 1

(M1)n -(M2)m
Monomer 2 3
1
Carboxylic CNTs react with reactive polymers;
2
Functionalized CNTs act as initiators to initiate polymerization;
3
Living polymerization, CNT-copolymer can be obtained; n, m: Degree of polymerization.

Fig. 12. Strategies for covalent functionalization of CNTs (A: direct sidewall functionalization; B: defect functionalization).

similar methods, including cycloaddition, such as Diels–Alder reac-
tion, carbene and nitrene addition [49–51], chlorination, bromina-
tion [52], hydrogenation [53], azomethine ylides [54], have also
been successfully employed in recent years. All these methods
can be regarded as the derivative of sidewall functionalization.
Defect functionalization is another method for covalent func-
tionalization of CNTs. This process takes advantage of chemical
transformation of defect sites on CNTs (Fig. 12B). Defect sites can
be the open ends and/or holes in the sidewalls, pentagon or hepta-
gon irregularities in the hexagon graphene framework. Oxygenated
sites can also to be considered as defects. Defects can be created on
the sidewalls as well as at the open ends of CNTs by an oxidative
process with strong acids such as HNO3, H2SO4 or a mixture of
them [55], or with strong oxidants such as KMnO4 [56], ozone
[57,58], reactive plasma [59,60]. The defects on CNTs created by
oxidants are stabilized by bonding with carboxylic acid (–COOH)
or hydroxyl (–OH) groups. These functional groups have rich
chemistry and the CNTs can be used as precursors for further
chemical reactions, such as silanation [58], polymer grafting [61],
esterification [62], thiolation [63], alkylation and arylation [64]
and even some biomolecules [10]. The CNTs functionalized in this
way are soluble in many organic solvents because the hydrophobic
nature of CNTs is changed to hydrophilic one due to the attach-
ment of polar groups. The chemically functionalized CNTs can pro-
duce strong interfacial bonds with many polymers, allowing CNT-
based nanocomposites to possess high mechanical and functional
properties.
3.2. Physical functionalization
Functionalization of CNTs using covalent method can provide
useful functional groups onto the CNT surface. However, these
methods have two major drawbacks: firstly, during the functional-
ization reaction, especially along with damaging ultrasonication
process, a large number of defects are inevitably created on the
CNT sidewalls, and in some extreme cases, CNTs are fragmented
into smaller pieces. These damaging effects result in severe degra-
dation in mechanical properties of CNTs as well as disruption of p
electron system in nanotubes. The disruption of p electrons is det-
rimental to transport properties of CNTs because defect sites scat-
 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367
Fig. 13. Schematics of CNT functionalization using non-covalent methods (A: polymer wrapping; B: surfactant adsorption; C: endohedral method). After [40,70].
ter electrons and phonons that are responsible for the electrical
and thermal conductions of CNTs, respectively. Secondly, concen-
trated acids or strong oxidants are often used for CNT functionali-
zation, which are environmentally unfriendly. Therefore, many
efforts have been put forward to developing methods that are con-
venient to use, of low cost and less damage to CNT structure.
Non-covalent functionalization is an alternative method for
tuning the interfacial properties of nanotubes. The suspension of
CNTs in the presence of polymers, such as poly(phenylene vinyl-
ene) [65] or polystyrene [66], lead to the wrapping of polymer
around the CNTs to form supermolecular complexes of CNTs. This
is a typical example of non-covalent functionalization of CNTs
(Fig. 13A). The polymer wrapping process is achieved through
the van der Waals interactions and p–p stacking between CNTs
and polymer chains containing aromatic rings.
Besides polymers, surfactants have also been employed to func-
tionalize CNTs (Fig. 13B). Several studies have contributed to
studying the effects of surfactant on dispersibility and other prop-
erty of CNTs [67–77]. The surfactants studied previously include:
(i) non-ionic surfactants, such as polyoxyethylene 8 lauryl
(CH3(CH2)11(OCH2–CH2)7OCH2CH3) [67], nonylphenol ethoxylate
(Tergitol NP-7) [68], polyoxyethylene octylphenylether (Triton X-
100) [69,70]; (ii) anionic surfactants, such as sodium dodecylsul-
fate (SDS), sodium dodecylbenzenesulfonate (SDBS), poly(styrene
sulfate) (PSS) [71–73]; (iii) cationic surfactants, such as dodecyl
tri-methyl ammoniumbromide (DTAB) [74], cetyltrimethylam-
mounium 4-vinylbenzoate (CTVB) [75]. In particular, a recent
study provides a comprehensive review of the mechanisms behind
the improved dispersibility of CNTs [76]. The physical adsorption
of surfactant on the CNT surface lowered the surface tension of
CNT, effectively preventing the formation of aggregates. Further-
more, the surfactant-treated CNTs overcome the van der Waals
attraction by electrostatic/steric repulsive forces. The efficiency of
this method depended strongly on the properties of surfactants,
medium chemistry and polymer matrix. It was concluded that in
water-soluble polymers such as polyethylene glycol, cationic
surfactants had some advantages, whereas in water-insoluble
polymers like polypropylene, CNT dispersion was promoted by a
non-ionic surfactant [76,77]. The treatment of non-ionic surfac-
tants is based on a strong hydrophobic attraction between the solid
surface and the tail group of surfactant. Once the surfactant is ad-
sorbed onto the filler surface, the surfactant molecules are self-
assembled into micelles above a critical micelle concentration
(CMC) [70].
Another non-covalent method for CNT functionalization is the
so-called endohedral method (Fig. 13C). In this method, guest
atoms or molecules are stored in the inner cavity of CNTs through
the capillary effect. The insertion often takes place at defect sites
localized at the ends or on the sidewalls. The insertion of inorganic
nanoparticles into the tubes [78], such as C60, Ag, Au and Pt, is a
typical example of endohedral functionalization. Small biomole-
cules, such as proteins, DNA, can also be entrapped in the inner
1353
hollow channel of nanotubes by simple adsorption, forming natu-
ral nano-test tubes [40,79]. The combination of these two materials
(i.e. CNTs and guest molecules) is particularly useful to integrate
the properties of the two components in hybrid materials for use
in catalysis, energy storage, nanotechnology and molecular scale
devices [78].
4. CNT/polymer nanocomposites
4.1. Classification of CNT/polymer nanocomposites
Polymer composites, consisting of additives and polymer matri-
ces, including thermoplastics, thermosets and elastomers, are con-
sidered to be an important group of relatively inexpensive
materials for many engineering applications. Two or more materi-
als are combined to produce composites that possess properties
that are unique and cannot be obtained each material acting alone.
For example, high modulus carbon fibers or silica particles are
added into a polymer to produce reinforced polymer composites
that exhibit significantly enhanced mechanical properties includ-
ing strength, modulus and fracture toughness. However, there
are some bottlenecks in optimizing the properties of polymer com-
posites by employing traditional micron-scale fillers. The conven-
tional filler content in polymer composites is generally in the
range of 10–70 wt.%, which in turn results in a composite with a
high density and high material cost. In addition, the modulus and
strength of composites are often traded for high fracture toughness
[9].
Unlike traditional polymer composites containing micron-scale
fillers, the incorporation of nanoscale CNTs into a polymer system
results in very short distance between the fillers, thus the proper-
ties of composites can be largely modified even at an extremely
low content of filler. For example, the electrical conductivity of
CNT/epoxy nanocomposites can be enhanced several orders of
magnitude with less than 0.5 wt.% of CNTs [36]. As described pre-
viously, CNTs are amongst the strongest and stiffest fibers ever
known. These excellent mechanical properties combined with
other physical properties of CNTs exemplify huge potential appli-
cations of CNT/polymer nanocomposites. Ongoing experimental
works in this area have shown some exciting results, although
the much-anticipated commercial success has yet to be realized
in the years ahead. In addition, CNT/polymer nanocomposites are
one of the most studied systems because of the fact that polymer
matrix can be easily fabricated without damaging CNTs based on
conventional manufacturing techniques, a potential advantage of
reduced cost for mass production of nanocomposites in the future.
Following the first report on the preparation of a CNT/polymer
nanocomposite in 1994 [80], many research efforts have been
made to understand their structure–property relationship and find
useful applications in different fields, and these efforts have be-
come more pronounced after the realization of CNT fabrication in
1354 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367
15000
12000
Number of Papers
9000
CNT
6000
CNT/Polymer
Nanocomposites
3000
0 0
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010
Academic Year
Fig. 14. Number of published papers related to CNT and CNT/polymer nanocom-
posites as a function of academic year. After [81].
industrial scale with lower costs in the beginning of the 21st cen-
tury, as shown in Fig. 14 [81]. According to the specific application,
CNT/polymer nanocomposites can be classified as structural or
functional composites [82]. For the structural composites, the un-
ique mechanical properties of CNTs, such as the high modulus, ten-
sile strength and strain to fracture, are explored to obtain
structural materials with much improved mechanical properties.
As for CNT/polymer functional composites, many other unique
properties of CNTs, such as electrical, thermal, optical and damping
properties along with their excellent mechanical properties, are
utilized to develop multi-functional composites for applications
in the fields of heat resistance, chemical sensing, electrical and
thermal management, photoemission, electromagnetic absorbing
and energy storage performances, etc.
4.2. Fabrication for CNT/polymer nanocomposites
4.2.1. Solution mixing
Solution mixing is the most common method for the fabrication
of CNT/polymer nanocomposites because it is amenable to small
sample sizes [17,77,82]. Typically, solution blending involves three
major steps: dispersion of CNTs in a suitable solvent by mechanical
mixing, magnetic agitation or sonication. The solvent can also dis-
solve polymer resins. Subsequently, the dispersed CNTs are mixed
with polymer matrix at room or elevated temperatures. The nano-
composite is finally obtained by precipitating or casting the
mixture. This method is often used to prepare composite films.
4.2.2. Melt blending
Melt blending is another commonly used method to fabricate
CNT/polymer nanocomposites. Thermoplastic polymers, such as
polypropylene [48], polystyrene [66], poly(ethylene 2,6-naphtha-
late) [83], can be processed as matrix materials in this method.
Epoxy monomer
Mixing and
re-dispersion
Dispersed CNTs
in solvent
Vacuum oven: to remove entrapped
air bubbles and remained solvent
Curing agent
Fig. 15. Schematics of CNT/epoxy nanocomposites fabrication. After [88].
2500 The major advantage of this method is that no solvent is employed
to disperse CNTs. Melt blending uses a high temperature and a high
2000
shear force to disperse CNTs in a polymer matrix, and is most com-
patible with current industrial practices. Special equipments, such
as extruder, injection machine, which are capable of being oper-
1500 ated at an elevated temperature and generating high shear forces,
are employed to disperse CNTs. Melt blending or variants of this
1000 technique are frequently used to produce CNT/polymer composite
fibers. Compared with the solution mixing methods, this technique
is generally considered less effective to disperse CNTs in polymers
500
than that of solution mixing, and its application is also limited to
low filler concentrations in thermoplastic matrices [17].
4.2.3. In situ polymerization
In situ polymerization is an efficient method to realize uniform
dispersion of CNTs in a thermosetting polymer. In this method,
CNTs are mixed with monomers, either in the presence or absence
of a solvent, and then these monomers are polymerized via addi-
tion or condensation reactions with a hardener or curing agents
at an elevated temperature. One of the major advantages of this
method is that covalent bonding can be formed between the func-
tionalized CNTs and polymer matrix, resulting in much improved
mechanical properties of composites through strong interfacial
bonds. Epoxy-based nanocomposites comprise the majority of re-
ports using in situ polymerization methods [25,84–88], and
Fig. 15 illustrates a flowchart of typical processes used to produce
CNT/epoxy nanocomposites.
4.2.4. Latex technology
A relatively new approach to incorporate CNTs into a polymer
matrix is based on the use of latex technology [17,77,82]. Latex is
a colloidal dispersion of discrete polymer particles, usually in an
aqueous medium. By using this technology, it is possible to dis-
perse single and multi-walled CNTs in most of polymers that are
produced by emulsion polymerization, or that can be brought into
the form of an emulsion. Contrary to the in situ polymerization
system, the addition of CNTs in this technique takes place after
the polymer has been synthesized. The first step of the process
consists of exfoliation (for SWCNT bundles) or dispersion/stabiliza-
tion (for MWCNT entanglements) of CNTs in an aqueous surfactant
solution. This is followed by mixing the stable dispersion of surfac-
tant-treated CNTs with a polymer latex. After freeze-drying and
subsequent melt-processing, a nanocomposite consisting of dis-
persed CNTs in a polymer matrix can be obtained.
The advantages of this technique are obvious [77,82]: the whole
process is easy (because it basically consists of a simple mixing of
two aqueous components), versatile, reproducible, and reliable,
and allows incorporation of individual CNTs into a highly viscous
polymer matrix. The solvent used for CNT dispersion is water, thus
the process is a safe, environmentally friendly and low-cost meth-
od. Nowadays, polymer latex is industrially produced on a large
CNT/monomer Distillation: to
mixture remove solvent
Curing and post curing CNT/epoxy
nanocomposites
(Hardener)
 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367
scale and this industry is matured. Since the process is relatively
easy, the prospect for scale-up fabrication of CNT/polymer nano-
composites using this technique is very bright.
4.2.5. Other methods
To obtain CNT/polymer nanocomposites with a very high CNT
content or for some specific applications, new methods have been
developed in recent years: they include densification, spinning of
coagulant, layer-by-layer deposition and pulverization. Table 5
summarizes these methods and their features are briefly described.
It should be noted that as nanocomposite materials are an emerg-
ing field, many studies are being made to devise new processing
methods that can produce nanocomposites with unique structures
and properties for specialty end applications.
5. Effects of CNT dispersion and functionalization on the
properties of CNT/polymer nanocomposites
5.1. Mechanical properties
CNTs exhibit excellent mechanical properties with Young’s
modulus as high as 1.2 TPa and tensile strength of 50–200 GPa.
The combination of these exceptional mechanical properties along
with the low density, high aspect ratio and high surface area make
CNTs an ideal candidate for reinforcement in composite materials.
Both SWCNTs and MWCNTs have been utilized for reinforcing
thermosetting polymers, such as epoxy, polyurethane, and phe-
nol–formaldehyde resins, as well as thermoplastic polymers,
including polyethylene, polypropylene, polystyrene, nylon, and so
on. The CNT reinforced nanocomposites can be considered as a
kind of particulate composites with the filler dimensions on the
nanometer scale and a high aspect ratio. Unlike the microscopic
particulate composites, the mechanical properties of nanocompos-
ites depend strongly on the dispersion state of nanofillers, apart
from the properties of fillers and matrix material. In addition to
dispersion, there are other major requirements that need to be sat-
isfied for effective reinforcement of CNTs in composites [11]: they
include a high aspect ratio, alignment and interfacial interactions
between CNT and polymer matrix.
The aspect ratio must be sufficiently large to maximize the load
transfer between the CNTs and matrix material and thus to achieve
enhanced mechanical properties. For example, polystyrene nano-
composites reinforced with well-dispersed 1.0 wt.% CNTs of a high
aspect ratio had more than 35% and 25% increases in elastic modu-
lus and tensile strength, respectively [94]. Similar encouraging
results have also been reported [95,96], but other reports demon-
strated only modest improvements in modulus and strength. For
example, the mechanical properties of CNT/epoxy nanocomposites
Table 5
New techniques for the fabrication of CNT/polymer nanocomposites.
Method CNT
Densification As-grown CNT forest
Spinning of coagulant Pre-dispersion of CNT using a surfactant
solution
Layer-by-layer Pre-dispersion of CNT in solvent is
deposition necessary
Pulverization As-received CNTs
1355
reinforced by two types of MWCNTs with different aspect ratios,
and found that the impact resistance and fracture toughness were
significantly improved only for those containing CNTs of a higher
aspect ratio [97]. However, the corresponding tensile modulus
and strength showed very limited improvements of less than 5%,
due probably to weak bonds between the CNTs and polymer matrix
and agglomeration of CNTs.
Indeed, dispersion is the foremost important issue in producing
CNT/polymer nanocomposites. Many different techniques have
been employed for CNT dispersion, as discussed in Section 2.2. A
good dispersion not only makes more filler surface area available
for bonding with polymer matrix, but also prevents the aggregated
filler from acting as stress concentrator, which are detrimental to
mechanical performance of composites [98]. However, some chal-
lenges, such as CNT content in composites, length and entangle-
ment of CNTs as well as viscosity of matrix have to be overcome
to obtain uniform CNT dispersion in nanocomposites. There were
many reports [85–88] showing that there is a critical CNT content
in the matrix below which the strengthening effect for CNT/poly-
mer composites increases with increasing CNT content. Above this
critical CNT content, however, the mechanical properties of CNT/
polymer composites decrease, and in some cases, they decrease be-
low those of the neat matrix material. These observations can be
attributed to (i) the difficulties associated with uniform dispersion
of CNTs at high CNT contents; (ii) and lack of polymerization reac-
tions that are adversely affected by the high CNT content. The lat-
ter effect becomes more pronounced when functionalized CNTs are
employed to produce CNT/polymer nanocomposites.
The technique employed for CNT dispersion can influence, to a
large extent, on the mechanical properties of CNT/polymer nano-
composites. Table 6 [99] summarizes the flexural properties of
CNT/epoxy composites fabricated from the same CNT type and
content (NK-50, NanoKarbon) and the same epoxy matrix using
different techniques for CNT dispersion. Examinations were made
on polished thin nanocomposite specimens [99] as shown in
Fig. 16. The best CNT dispersion was achieved by employing calen-
dering machine (Fig. 16D), and the nanocomposites fabricated
Table 6
Effect of CNT dispersion on the flexural properties of CNT/epoxy nanocomposites.
After [99].
CNT dispersion Flexural modulus Flexural strength
technique (GPa) (MPa)
Neat epoxy 3.43 ± 0.01 (+0.00%) 140.0 ± 2.3 (+0.00%)
Sonication in water bath 3.41 ± 0.04 (0.58%) 144.1 ± 1.8 (+2.93%)
Shear mixing 3.36 ± 0.09 (2.08%) 140.4 ± 2.4 (+0.29%)
Probe sonication 3.47 ± 0.06 (+1.17%) 142.7 ± 1.7 (+1.93%)
Calendering 3.65 ± 0.02 (+6.41%) 145.2 ± 0.82 (+3.71%)
Fabrication process Advantages Ref.
CNT forest was prepared and CNT vol.% can be controlled (5–20 vol.%) by [89]
transferred to a pool of uncured varying the densification of CNT forest, aligned
epoxy. The matrix was infused into CNTs in nanocomposites
the CNT forest and then cured
Coagulation of CNT into a mesh by Employed to fabricate CNT/polymer fiber [90]
wet spinning into a polymer solution,
and converting the mesh into a solid
fiber by a slow draw process
Dipping of a solid substrate (glass Structural defects originated from the phase [91]
slides, silicon wafers) into the CNT/ segregation of the polymer/CNT can be
polymer solutions and following minimized, high loading of CNT (up to 50 wt.%)
curing
Polymer and CNTs were mixed and Possible grafting of polymers on CNTs, easy [92,93]
pulverized by pan mill or twin screw scale-up, solventless process
1356 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367

Fig. 16. Dispersion of CNTs in nanocomposites fabricated using different dispersion techniques (A: sonication in water bath; B: shear mixing; C: probe sonicator; D:
Calendering; Scale bar = 20 lm). After [99].

using this technique exhibited the best flexural properties amongst
the four techniques used (Table 6).
It should be noted that the dispersion states of CNTs in the poly-
mer matrix can be totally different depending on the magnification
or scale. Using the term ‘‘uniform” dispersion or ‘‘good” dispersion
to evaluate the CNT dispersion without any distinctive description
may simply be misleading or inaccurate, according to a recent
study by Li et al. [36]. For conventional composites, ‘‘dispersion”
is generally defined as even distribution of fillers in a matrix med-
ium without aggregates. In contrast, for CNT/polymer nanocom-
posites, dispersion has two aspects: (i) disentanglement of CNT
bundles or agglomerates, which is nanoscopic dispersion and (ii)
uniform distribution of individual CNTs or their agglomerates
throughout the nanocomposites, which is more of a micro- and
macroscopic dispersion.
From geometric consideration, the difference between random
orientation and alignment of CNTs can result in significant changes
in various properties of composites [11]. The storage modulus of
the polystyrene composite films containing random and oriented
CNTs were 10% and 49% higher than the unreinforced bulk poly-
mer, respectively [100]. The alignment can be regarded as a special
case for CNT dispersion. A few techniques, including mechanical
stretching [101], melt-spinning [102], dielectrophoresis [103],
application of electrical or magnetic field [104–106], have been
employed during the composite fabrication to align CNTs in a poly-
mer matrix. The degree of CNT alignment in the composite can be
governed by two factors: (i) diameter of CNTs and (ii) CNT content.
A smaller diameter of CNT can enhance the degree of CNT align-
ment due to the greater extensional flow; and a higher CNT content
decreases their alignment because of the CNT agglomeration and
restrictions in motion from neighboring CNTs [107]. While align-
ment is necessary to maximize the strength and modulus, it is
not always beneficial because the aligned composites have very
anisotropic mechanical properties, i.e., the mechanical properties
along the alignment direction can be enhanced, whereas these
properties are sacrificed along the direction perpendicular to this
orientation.
As for fiber-reinforced polymer composites, the external load
applied to CNT/polymer composites have to be transferred to the
CNTs, allowing them to take the major share of the load
[11,13,85]. The efficiency of load transfer depends on the interfa-
cial bond between the filler and the matrix material. A strong inter-
facial adhesion corresponds to high mechanical properties of
composites through enhanced load transfer from matrix to CNT.
Chemical and physical functionalization of CNTs has proven to en-
hance the interfacial adhesion. Table 7 summarizes recent studies
on the effects of CNT functionalization on the mechanical proper-
ties of CNT nanocomposites made from thermosets and thermo-
plastics. These results indicate clearly that functionalization of
CNTs can enhance the modulus, strength as well as fracture resis-
tance of nanocomposites.
5.2. Electrical properties
Electrically conducting composites with a volume conductivity
higher than 1010 S/cm, are considered to be an important group

Table 7
Effect of CNT functionalization on the mechanical properties of CNT/polymer nanocomposites.

Matrix Dispersion technique Functionalization technique CNT content Enhancement on mechanical propertiesa Ref.
Modulus (%) Strength (%) Toughness (%)
Thermo-plasticb PA Twin-screw extruder Diamine treatment 1.0 wt.% 6.1 (42) 5.3 (18) – [108]
PB Shear mixing Polymer grafting 1.5 wt.% 18 (91) 27 (61) 40 (67) [109]
PE Shear mixing Maleic anhydride and amine treatment 1.5 wt.% 22 (75) 17 (33) 69 (61) [110]
PI Ultrasonication Acid treatment 7.0 wt.% 39 (61) 19 (31) – [111]
PP Ultrasonication and stir Undecyl radicals attachment 1.5 wt.% 55 (84) 10 (13) – [112]
PS Probe ultrasonication Butyl attachment 0.25 wt.% 8.3 (25) 2.1 (50) – [113]
PVA Ultrasonication and stir Polymer grafting 2.5 wt.% 35 (40) 4.8 (17) – [114]
PMMA Ultrasonication Polymer grafting 0.10 wt.% 57 (104) 2.7 (86) – [115]
Thermo-setc EP Calendering Amino treatment 0.10 wt.% 2.1 (6.7) 2.2 (3.1) 17 (19) [25]
Probe ultrasonication Surfactant treatment 0.25 wt.% 8.6 (24) 6.8 (20) 35 (60) [70]
Ultrasonication Organic silane 0.25 wt.% 8.7 (22) 3.6 (18) 22 (8.5) [85]
PU Ball mill Polymer grafting 0.7 wt.% 48 (178) 27 (23) – [116]
Ultrasonication Acid treatment 10 wt.% 340 (500) 51 (111) – [117]
VR Ball mill Acid treatment 25 phr 444 (594) 175 (244) – [118]
Roll mill Organic silane 1 phr 35 (28) 7.8 (25) – [119]
a
Data in brackets indicate the enhancement on mechanical properties by employing functionalized CNTs.
b
PA (Nylon): polyamide; PB: polybutylene; PE: polyethylene; PI: polyimide; PP: polypropylene; PS: polystyrene; PVA: poly (vinyl acetate); PMMA: poly(methyl
methacrylate).
c
EP: epoxy; PU: polyurethane; VR: vulcanized rubber.
 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367 1357

of relatively inexpensive materials for many engineering applica- SWCNT/polymer nanocomposites

tions (Fig. 17A) [77,120,121], such as electrically conducting adhe- MWCNT/polymer nanocomposites

Percolation threshold (wt%)

sives, antistatic coatings and films, electromagnetic interference
shielding materials for electronic devices, thermal interface mate-
rials, etc.
The percolation theory can be applied to explain the electrically
conducting behavior of composites consisting of conducting fillers
and insulating matrices. When the conducting filler content is
gradually increased, the composite undergoes an insulator-to-con-
ductor transition. The critical filler content is referred to as the per-
colation threshold where the measured electrical conductivity of
the composite sharply jumps up several orders of magnitude due
to the formation of continuous electron paths or conducting net-
works. Below the percolation transition range, electron paths do
not exist and the electrical properties are dominated by the matrix
material. Above the percolation transition range, multiple electron
paths exist in the matrix so that the electrical conductivity of the
composite often shows a saturation plateau. This behavior is
graphically shown in Fig. 17B [122–124].
The concentration of the conducting filler must be above the
percolation threshold in order to achieve conducting networks in
the composite. Conventional conducting fillers are usually
micrometer-scale metal powder or carbonaceous materials, such
as carbon black (CB), exfoliated graphite and carbon fibers. In order
to achieve the percolation threshold with these fillers, the filler
content needs to be as high as 10–50 wt.% [121–125], resulting
in a composite with poor mechanical properties and a high density.
The incorporation of CNTs into polymer offers an attractive
route to minimize aforementioned problems. Compared with tra-
ditional conducting fillers, CNTs have unique advantages, i.e. high
aspect ratio and excellent electrical conductivity, which in turn
facilitate the formation of conducting networks and thus transform
the insulating polymer to a conducting composite at a low CNT
content, often as low as 0.5 wt.%.
Fig. 18 shows the percolation threshold of nanocomposites
filled with CNTs and different polymers [126]. For most polymer
matrices, they transit from an insulator to a conductor when CNT
content is below 5 wt.%. However, there is no apparent consensus
on percolation threshold values of CNTs: for example, the values
reported for typical CNT/epoxy nanocomposites varied from
0.002 to over 7 wt.% [36,84,85,127,128], depending on the type of
CNTs (SWCNTs or MWCNTs) and the processing techniques used
to produce the nanocomposites. This large variation indicates that
the dispersion state and surface characteristics of CNTs affected by
functionalization as well as processing conditions are important in
determining the electrical properties of CNT/polymer nanocom-
posites. Bauhofer et al. [126] summarized the effect of experimen-
tal parameters on the percolation threshold of CNT/polymer
PA PANI PC PE PI PMMA PP PS PVA PU EP
Polymer matrix
Fig. 18. Percolation threshold of CNT/polymer nanocomposites. After [126].
nanocomposites, and concluded that the well-dispersed CNTs gave
rise to 50 times higher conductivities than the entangled ones. The
CNT dispersion played a more important role than the type and
production method of CNTs. Li et al. [36] produced CNT/epoxy
nanocomposites with percolation thresholds varying in the range
from 0.1 to >1 wt.% by controlling the dispersion state and aspect
ratio of CNTs. They noted that the critical factors determining the
percolation threshold were: (i) aspect ratio of CNTs; (ii) disentan-
glement of CNT agglomerates on the nanoscopic scale; and (iii)
uniform distribution of individual CNTs or CNT agglomerates on
microscopic scale. There is a critical value of CNT aspect ratio,
above which the percolation threshold was sensitive to dispersion
state, while below which the percolation threshold increased rap-
idly with decreasing aspect ratio.
Many studies have shown that CNT functionalization has tre-
mendous influences on the electrical conductivities of nanocom-
posites: proper functionalization facilitates the CNT dispersion
and the formation of conducting networks in composites, resulting
in lowered percolation threshold of nanocomposites. However,
excessive functionalization introduces too many heterogeneous
atoms onto the CNT surface, resulting in the perturbation of p elec-
trons and thus degrading the intrinsic electrical properties of CNTs.
In addition, functionalization of CNTs using acids of a high concen-
tration can severely damage and fragment CNTs into smaller pieces
with decreased aspect ratios. Both of these are detrimental to the
formation of conducting networks in nanocomposites. Fig. 19
shows the effect of CNT functionalization on the electrical conduc-
tivities of CNT/epoxy nanocomposites. The CNTs functionalized
with an organic silane led to wrapping of insulating material onto
CNTs [85], the wrapping being more serious for well-dispersed

A 1.00E+02 B
Conducting adhesive and coating
Electrical conductivity (S/cm)

1.00E+00 Thermal interface and

electromagnetic shielding materials
1.00E-02
Conducting thin film, elastic electrode
1.00E-04

Antistatic coatings and films

1.00E-06

1.00E-08 Semiconducting materials, papers and

textiles, household appliances shell
1.00E-10
Insulating material

Fig. 17. Typical applications of conducting composites (A) and a schematic of percolation phenomenon and conducting network in conducting composites (B).
1358
1.00E+00
Electrical conductivity (S/cm)
1.00E-02
1.00E-04
1.00E-06
1.00E-08
1.00E-10
1.00E-12
1.00E-14
0 0.1 0.2 0.3
CNT content (wt%)
Fig. 19. The effect of CNT functionalization on the electrical conductivities of CNT/
epoxy nanocomposites. After [85,86].
individual CNTs, thus exhibiting no percolation behavior for the
nanocomposites containing these CNTs. When amino-functional-
ized CNTs were employed [86], the nanocomposites showed a typ-
ical percolation behavior, although the conductivities were in
general lower than those containing pristine or untreated CNTs.
The measured electrical conductivities of CNT/polymer compos-
ites typically ranged from 105 to 103 S/cm above the percolation
threshold [84,85,127,128]. While further increase in CNT content
above the percolation threshold can enhance marginally the electri-
cal conductivity of composites, the solution viscosity becomes too
high to produce void-free composites when the CNT content is
higher than 1.0 wt.%. This limits the use of CNT–polymer nanocom-
posites for applications where high CNT contents are necessary.
Therefore, processing techniques for improving the electrical con-
ductivity of nanocomposites below or near the percolation thresh-
old became crucial to producing highly conducting composites.
Simulation results indicated that the contact resistance of CNTs in
composites played an important role in enhancing the conductivity
of nanocomposites [129,130]. This was confirmed by a recent study
[86] in which when silver decorated CNTs were employed as con-
ducting filler, the nanocomposites exhibited a significantly higher
conductivity above the percolation threshold than those containing
pristine CNTs (Fig. 19). A high conductivity value of 0.81 S/cm was
achieved with 0.5 wt.% of silver decorated CNTs. This observation
was attributed to the silver nanoparticles that were tightly attached
onto the defect sites of CNT surface, which compensated the nega-
tive effect of damaged CNTs (due to the amino functionalization) by
enhancing the inherent conductivity of CNTs and reducing the con-
tact resistance of CNT junctions in the matrix.
The high cost of CNTs, especially SWCNTs, compared with other
fillers like graphite, CB and carbon fibers, limits the widespread
applications of CNT-based conducting composites. Therefore,
nanocomposites containing hybrid fillers of CNTs and other low-
cost particles were developed in recent years [88,131–136]. In
these composites, the CNT networks serve as skeleton and the
other conducting fillers facilitate the transportation of electrons
along the CNT skeleton, thus enhancing the overall electrical prop-
erties of nanocomposites. Li et al. [131] reported the preparation of
conducting nanocomposites filled with hybrid fillers of GNPs and
CNTs: they showed that the nanocomposites containing 1% GNP
and 1% CNT achieved the highest electrical conductivity of
4.7  103 S/cm, which was more than two orders of magnitude
higher than that of the nanocomposites with 2 wt.% GNP alone.
In another study [88], nanocomposites reinforced with hybrid fill-
ers of CNTs and CB were prepared, aiming at enhancing the electri-
cal conductivity of composites with balanced mechanical
properties while lowering the cost of the final product. The results
showed that the addition of CNTs in CB composites not only re-
P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367
duced the percolation threshold, but also enhanced the ductility
and fracture toughness of nanocomposites, confirming the syner-
gistic effect arising from the hybrid fillers. Hybrid fillers with un-
ique geometric shapes and different dispersion characteristics
may offer a new way to lower the final cost of CNT-based nano-
composites with multi-functional properties.
Ag@CNTs
Amino-CNTs
Silane-CNTs
5.3. Thermal properties
Pristine CNTs
The thermal conductivity of a material is dominated by atomic
vibrations or phonons, and the conduction by electrons is generally
negligible for insulating materials [137]. For CNT/polymer nano-
0.4 0.5 composites, the thermal conductivity depends several factors,
including the content, aspect ratio, dispersion of CNTs and their
interfacial interactions with polymer matrix. The excellent thermal
properties of CNTs, such as high thermal conductivity and good
thermal stability (Table 1), led to the expectation that CNTs could
make useful functional fillers to rectify the thermal properties of
polymers. Contrary to this expectation, the enhancements of ther-
mal properties due to the incorporation of CNTs in polymer com-
posites have not been quite remarkable, with the exception of
few isolated cases.
Many efforts have been devoted to employing CNTs as thermal
conducting filler in polymer composites and some enhancements
have indeed been observed. For example, Biercuk et al. [138] used
SWCNTs to augment the thermal transport properties of epoxy,
showing exhibited 70% and 125% increases in thermal conductivity
at 40 K and room temperature, respectively, with 1 wt.% SWCNTs.
Evseeva et al. [139] found that the introduction of 0.1–1.0 wt.%
MWCNTs enhanced the thermal conductivity of epoxy resin by
about 40%. However, comparison of the data obtained from the
above reports revealed that the thermal conductivities of CNT/
polymer composites showed only a marginal improvement com-
pared with the electrical conductivities with the same CNT content.
Taking into account the different transport mechanisms between
the thermal and electrical conductivities of composites, Moniruzz-
aman et al. [17] proposed that the phonons, major carriers for ther-
mal conduction, were much more likely to travel through the
matrix rather than through the CNT networks because the differ-
ence in thermal conductivity between the CNT and polymer, i.e.
about 104 W/(m K) with KCNT  103 and Kpolymer  101 W/(m K),
is much smaller than that of the electrical conductivity, i.e. in the
range of 1015–1019 with rCNT  102–106 and rpolymer < 1013 S/cm.
The aspect ratio of CNTs also has a significant effect on the ther-
mal conductivity of composites. This can be understood from two
viewpoints: (i) for a given volume fraction, the number of CNT–
polymer–CNT contacts increases with decreasing aspect ratio,
which in turn reduces the influence of high interfacial thermal
resistance between CNTs and polymer matrix [140]; (ii) increasing
the aspect ratio of CNTs shifts the phonon dispersion towards low-
er frequencies, resulting in a better CNT-liquid thermal coupling
[141]. Cai’s work [142] on CNT/polyurethane composites based
on the latex technology has proven the effect of aspect ratio: the
thermal conductivity of the composites increased by more than
200% with 3 wt.% CNTs whose aspect ratio was maintained over
3000 (50 lm in length, and 8–15 nm in diameter).
There is another important factor that may affect the thermal
conductivity, namely the interfacial adhesion between CNTs and
polymer matrix. The improved CNT–polymer interfacial interac-
tions can inhibit the phonon transportation along CNTs, and in-
creased the interfacial thermal resistance by affecting the
damping behavior of the phonons’ vibration [143]. Ma et al. [86]
compared the thermal conductivities of composites filled with
CNTs of different surface functionalities (Fig. 20), showing that
the thermal conductivity was in general higher for those contain-
ing pristine CNTs than those with amino-functionalized CNTs (con-

1 5.4. Optical properties
Thermal conductivity (W/(m-K)
0.9
0.8 (63%)
0.7 (43%)
(36%)
(33%)
0.6 (24%)
(14%)
0.5
(0.0%)
0.4
0 0.1
CNT content (wt%)
Fig. 20. Comparison of thermal conductivities of composites containing CNTs with
different functionalities (data in bracket indicate% increase in thermal conductivity
against the polymer matrix). After [86].
taining amine and amide functional groups on the surface). They
also proposed a method for the recovery of thermal conductivity
by decorating silver nanoparticles onto the CNT surface (see
Fig. 20): silver is an excellent thermal conductor (KAg = 429 W/
(m K)) and the Ag–CNT interface facilitated phonon conduction
by reducing the boundary scattering losses and interfacial resis-
tance to heat flow between the CNTs and polymer matrix.
In addition to thermal conductivity, the incorporation of CNTs
in a polymer can also rectify the thermal stability (glass transi-
tion temperature, melting and thermal decomposition tempera-
tures) as well as flame-retardant properties through their
constraint effect on the polymer segments and chains. Again,
these thermal properties are governed by many factors such as
CNT type, aspect ratio and dispersion, its interfacial interactions
with polymer matrix, etc. For example, adding 0.25 wt.% silane
functionalized MWCNTs into epoxy resulted in a sharp rise in
the glass transition temperature from 147 to 161 °C, which was
even higher than that of the nanocomposites containing CNTs
without functionalization [85]. This capability becomes more
pronounced when CNTs are incorporated into some thermoplas-
tic matrices because CNTs can act as nucleation sites for crystal-
lization of polymers. 1.0 wt.% well-dispersed SWCNTs added into
polymethyl methacrylate (PMMA) gave rise to a 40 °C increase in
glass transition temperature of PMMA [144], while the thermal
decomposition temperature of polypropylene at peak weight loss
was increased by 12 °C with 2 vol.% MWCNTs [145]. Kashiwagi et
al. [146] studied the flammability of CNT/PMMA nanocomposites,
showing that CNTs can outperform nanoclays as effective flame-
retardant additives if they form a jammed network structure in
the polymer matrix. The mechanisms behind this enhancement
are two-folds [146–148]: (i) the networks containing CNTs act
as a heat shield for the neat polymer below the layer, thus signif-
icantly reducing the heat release rate of nanocomposites and (ii)
the incorporation of CNTs into a polymer leads to an enhanced
thermal stability of nanocomposites, thus effectively inhibits
the formation of cracks or openings that compromise the
flame-retardant effectiveness of nanocomposites during burning.
Poorly dispersed CNTs or very low concentration of CNTs re-
sulted in the formation of a discontinuous layer consisting of
fragmented islands with sizes from 1 to 10 mm [146]. The
flame-retardant performance of the nanocomposites containing
these island structures was much poorer than that of the nano-
composites with a continuous protective network layer. It is ex-
pected that the functionalization of CNTs will further reduce the
flammability of nanocomposites as it will not only enhance the
dispersion of CNTs in the polymer matrix, but also enhance the
thermal stability of nanocomposites.
P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367 1359
(94%)
The optical properties of CNT/polymer nanocomposites consist
(84%)
(72%)
(70%) (71%) of optical non-linearity and limitation, photoluminescence, light
(60%) emission and photonic properties.
The non-linear (NL) property of a material describes the behav-
ior of light in non-linear media in which the dielectric polarization
responds nonlinearly to the electric field of the light. The non-lin-
Ag@CNTs
Functionalized CNTs
earity is typically observed at very high light intensities, such as
those provided by pulsed lasers. It has been demonstrated that sus-
Pristine CNTs
pensions of both SWCNTs and MWCNTs have non-linear optical
properties [149,150], thus they can be used to rectify the optical
0.2 0.3 0.4 0.5
properties of polymer composites to fabricate NL optical wave-
guide devices [151]. Tang et al. [152] pioneered this research field:
wrapping of CNTs with poly(phenylacetylenes) (PPA) showed CNTs
with a strong photostabilization effect and the PPA chains being
protected from photodegradation under harsh laser irradiation
with incident fluence. The CNT/PPA solutions effectively limited in-
tense optical pulses, with the saturation fluence tunable by varying
the CNT content. The NL properties of CNT/polymer nanocompos-
ites are governed not only by the dispersion state of CNTs, but also
by the polymer structure. Wang and Blau [153] studied the NL
properties of SWCNTs in N-methyl-2-pyrrolidone (NMP), N,N-
dimethylformamide (DMF) and N,N-dimethylacetamide (DMA),
and found that the NL extinction coefficients of SWCNTs markedly
increased with increasing SWCNT concentration. CNTs dissolved in
NMP exhibited a much better dispersion, although their optical
limiting properties were inferior to those of DMF and DMA. The
average bundle size of CNTs and physical properties of solvent
dominated the NL extinction and optical limiting properties of
CNT dispersions. Recent studies also revealed similar results: a
good dispersion of CNTs did not guarantee a high performance of
NL and optical limiting properties of CNT/polymer nanocomposites
[154,155]. A possible reason for this observation is the complicated
optical limiting mechanisms of composite materials. Besides the
contribution by non-linear scattering, other factors, such as non-
linear absorption and refraction and electronic absorption, also
contributed to the optical limiting properties [156]. In addition of
CNT dispersion, the backbones of polymers have a considerable ef-
fect on the NL properties of CNT/polymer nanocomposites [154]:
e.g. functionalization of CNTs using conjugated polymers that have
high affinity with CNTs resulted in significant improvement in NL
absorption and scattering due to the enhanced p–p interactions
between the CNTs and polymer matrix.
The black color of CNTs combined with their high surface area
and high electron affinity enable CNTs to be an ideal candidate to
improve the photoluminescence of CNT/polymer nanocomposites:
photoluminescence is a process in which a substance absorbs pho-
tons (i.e. electromagnetic radiation) and re-radiates photons. In
CNT/polymer nanocomposites, excitons formed on the polymer
by the absorption of light which can be dissociated by electron
transfer from the photo-excited polymer to CNTs. This process
quenches the photoluminescence of polymers [157]. For example,
Singh et al. [158] studied CNT/poly(3-hexylthiophene) nanocom-
posites, showing that the intensity of photoluminescence emission
from the composite decreased with increasing CNT content and the
absorption coefficient of the composites was insensitive to CNTs of
up to 5 wt.%. CNTs functionalized using phthalocyanines also
showed a considerable quenching effect of the fluorescence inten-
sity in the photoluminescence spectrum [159]. In contrast, earlier
papers reported a strong luminescence from the MWCNTs func-
tionalized with a variety of moieties. The reason for these inconsis-
tencies may be attributed to the different defect states mediating
luminescence by trapping excited electrons, which subsequently
return to the valence band with the emission of light [160,161].
A recent study [157] revealed that visible photoluminescence from
1360 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367
Fig. 21. (A) Photoluminescence spectra of MWCNT/nylon composites for 488-nm excitation at varying MWCNT contents. A broad luminescence band centered around
590 nm is observed (Inset: Photoluminescence spectra of a pure nylon film for 488-nm excitation); (B) integrated PL intensity as a function of MWCNT content. After [157].
emissive impurities in nylon can be drastically enhanced when
acid-treated MWCNTs were added (Fig. 21). The origin of the lumi-
nescence was through the absorption by the acid-treated MWCNTs
and energy transfer to emissive impurities intrinsic to the organic
host matrix. Acid-treated MWCNTs were also shown to enhance
the resistance of organic chromophores to photobleaching. These
results provide compelling evidence that MWCNTs have an enor-
mous potential as a versatile material in future optoelectronic de-
vices and can be considered as a future platform for both
electronics and photonics.
Other potentially optical applications of CNTs are polymer-
based light-emitting and photonic devices. The advantages of using
organic light-emitting diodes (OLEDs) based on conjugated poly-
mers include low cost, low operating voltage, excellent process-
ability and flexibility. However, their low quantum efficiency and
stability have limited their applications and developments
[16,162]. Kim et al. [163] observed that the device qualities, such
as external quantum efficiency, were improved by 2–3 times with
the addition of up to 0.2 wt.% SWCNTs in OLEDs. They ascribed this
observation to the facile hole injection and the polymer–CNT inter-
actions that stiffened the polymer chains. Subsequently, the better
hole transport in the metallic SWCNT–polymer medium induced
more efficient single exciton formation at or near the interface re-
gion. The single-layer LEDs fabricated using MWCNT doped poly-p-
phenylene benzobisoxazole decreased the threshold voltage by
2 V. Up to 0.1 wt.% of MWCNTs, the diode emission current in-
creased by about two orders of magnitude over those of the LEDs
without MWCNTs. However, a further increase of CNT content
caused a successive decrease in light emission intensity due to
the poor dispersion of CNTs [164].
Besides light-emitting applications, CNT/polymer nanocompos-
ites are also widely used in photonic devices [162]. Arranz-Andrés
et al. [165] investigated the effects of incorporating different types
of CNTs, including single-walled, double-walled and multi-walled
CNTs, on the photovoltaic behavior of CNT/poly(3-hexiltiophene)
nanocomposites. The results showed that the power conversion
efficiency of the device made using this composite increased by
three orders of magnitude compared with that without CNTs, but
there was no clear correlation with the number of walls or diameter
of CNTs. The highest open-circuit voltage and the best performance
were achieved at a critical concentration of 5 wt.% of CNT for the
majority of the CNTs studied. Notice that the CNT content used in
this study was about 5 wt.%, which was much higher than the gen-
eral CNT content of less than 1 wt.% for polymer nanocomposites. It
is expected that improved CNT dispersion can further enhance the
photovoltaic properties of nanocomposites. In addition, charge
transfer between the CNT and polymer matrix plays an important
role in converting photovoltaic properties of nanocomposite-based
devices. Therefore, techniques for CNT functionalization using elec-
tron donors have also been developed recently [166–168], and
some interesting results have been generated by employing these
functionalized CNTs [166–170].
Another interesting photonic application of CNT/polymer nano-
composites is the photomechanical actuator. The photomechanical
effect is the change in the shape of a material when it is exposed
to light [171]. Zhang and Ijima [172] demonstrated that SWCNT fi-
bers can bend under light. This movement was sensitive to the inten-
sity of the light, and was attributed to the localized electrostatic
effects due to uneven distributions of photogenerated charges in
CNTs. Actuators constructed using CNTs dispersed in several poly-
mer matrices, such as Nafion, a sulfonated tetrafluoroethylene based
fluoropolymer–copolymer [173], and polydimethylsiloxane (PDMS)
[174], have been reported. Fig. 22A shows a schematic for the char-
acterization of photomechanical actuation behavior of CNT/polymer
nanocomposites. The introduction of CNTs in PDMS resulted in the
composite actuator exhibiting better performance than those con-
taining CB due to the high electrical conductivity of nanotubes
(Fig. 22B). Further studies revealed that the photomechanical re-
sponse of CNT/polymer nanocomposites was an intrinsic behavior
of CNTs, depending on the CNT alignment in the polymer matrix
[174]. The incorporation of aligned CNT thin films in an acrylic elas-
tomer allowed stable expansive actuation of 0.3% at a light illumina-
tion below 120 mW with actuation stroke large enough to
manipulate small objects [175]. Clearly, polymer/CNT nanocompos-
ites represent a new class of multi-functional material that can be
employed in photonic devices with improved performance [16,162].
5.5. Rheological properties
Rheology is the study of the flow behavior of a material under
conditions in which they flow rather than elastic or plastic defor-
mation. It is also concerned with establishing predictions for
mechanical behavior (on the continuum mechanical scale) based
on the micro- or nanostructure of the materials [176]. The study
of rheological response of CNT/polymer nanocomposites have both
practical importance related to composite processing and scientific
importance as a probe of the composite dynamics and microstruc-
ture [17].
The rheological properties of CNT/polymer nanocomposites de-
pend on factors such as the characteristics of the filler loading, as-
pect ratio and dispersion, polymer molecular weight, and the
interfacial interaction between the polymer and filler [177–181].
The variations of viscosity and storage modulus of composites as
a function of frequency are two commonly used techniques to
characterize the rheological properties of CNT/polymer nanocom-
posites. Fig. 23 shows the complex viscosity (Fig. 23A) and storage
 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367 1361

Fig. 22. (A) Schematic of experimental setup for characterizing the photomechanical actuation; (B) comparison of photomechanical actuation in SWCNT/PDMS multilayer
actuators (solid line) with thermal actuation of the same structure (j) and the photomechanical actuation of a CB/PDMS multilayer actuator (N). After [171].

Fig. 23. Rheological response (A: complex viscosity; B: storage modulus) of CNT/polycarbonate (PC) nanocomposites as a function of frequency at 230 °C and corresponding
rheological percolation threshold (C) at frequency of 1 rad/s. After [181].

modulus (Fig. 23B) of CNT/polycarbonate nanocomposites with
different CNT contents [181]. At low frequencies, the fully relaxed
polymer chains exhibit the typical Newtonian viscosity plateau.
With the addition of CNTs, the low-frequency complex viscosity
significantly increases, whereas the whole trend of viscosity drops
with an increase of frequency at a given CNT content (Fig. 23A),
indicating that the relaxation of polymer chains in the nanocom-
posites is effectively restrained by the presence of CNTs. The stor-
age modulus of nanocomposites gradually increases with
increasing frequency and CNT content (Fig. 23B), indicating a tran-
sition from viscous liquid to solid-like behavior. Similar to the elec-
trical percolation behavior, the rheology of a CNT/polymer
nanocomposite also shows a transition from a rheological state
(where the viscosity or storage modulus changes significantly with
increasing filler content) to a solid-like behavior (where the viscos-
ity or storage modulus is insensitive or has only a slight variation
with increasing filler content). The concentration at which this
transition occurs is known as the rheological percolation threshold
[178–181], which is shown in Fig. 23C.
Different from the electrical percolation threshold, the rheolog-
ical percolation threshold of CNT/polymer nanocomposites is
strongly dependent on temperature. In MWCNT/PC nanocompos-
ites, the percolation threshold decreased from 0.8 wt.% to around
0.3 wt.% upon increasing temperature from 210 to 300 °C [181].
The reason for this observation is that the rheological response of
nanocomposites is dominated by the networks formed by polymer
chains and CNT–polymer interactions, and the mobility of these
networks can be significantly enhanced at elevated temperature.
In contrast, the electrical percolation threshold requires contacts
of CNTs to form solid conducting networks where the temperature
effect becomes marginal. Since the polymer chains contribute
much to the rheological response of nanocomposites, it is easy to
understand that the content of CNTs required to form a rheological
percolation threshold should be much lower than that for electrical
percolation. For example, Bose et al. [182] found that the rheolog-
ical percolation threshold in MWCNT/PA6-ABS nanocomposites
was around 1–2 wt.%, which was significantly lower than the cor-
responding electrical percolation of around 3–4 wt.%.
The rheological response of a nanocomposite containing a fixed
CNT content below the percolation threshold can serve as an indi-
rect measurement to show the degree of dispersion of CNTs in the
polymer matrix. The better the dispersion of CNTs, the lower the
viscosity generally is. For example, Siddiqui et al. [183] successfully
reduced the viscosity of MWCNT/epoxy suspension by functionali-
zing the CNTs using a non-ionic surfactant, Triton-X100. The im-
proved dispersion of CNTs in the epoxy matrix was mainly
responsible for the lowered viscosity, which originated from the
hydrophilic and hydrophobic segments of surfactant molecules:
the hydrophobic octyl group of the surfactant interacted with CNTs
through adsorption while the hydrophilic segment linked with
epoxy through hydrogen bonding [70].
Another consequence of CNT functionalization is that the rheo-
logical percolation threshold of nanocomposites can be lowered
due to the improved CNT dispersion. Mitchell et al. [184] also
showed improved SWCNT dispersion in PS by organic diazonium
functionalization along with a concomitant drop in rheological
1362 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367
percolation threshold of CNT/PS composites from 3.0 to 1.5 wt.%.
The storage moduli at low frequencies were also higher for the
nanocomposites containing functionalized CNTs, indicating better
load transfer between the CNT network and the polymer. Ma et
al. [185] studied the effect of functionalization on re-agglomera-
tion of CNTs in an epoxy matrix that was affected by rheological
changes. The pristine CNTs without any functionalization started
re-agglomeration upon application of curing temperature, whereas
the amino-functionalized CNTs remained uniformly dispersed over
the whole curing process, indicating a beneficial effect of function-
alization on the stability of CNT dispersion even at a high temper-
ature. It is postulated that the amino-functional groups attached
on CNT affected the rheological behavior of epoxy, which in turn
facilitated a faster cure and wrapping of a protective polymer layer
onto the CNT surface, preventing them from attracting one
another.
5.6. Damping properties
In contrast to the enormous research on the elastic properties of
CNT/polymer nanocomposites, relatively less attention has been
paid to their damping mechanisms and ability. Damping describes
the trend of materials to reduce the amplitude of oscillations in an
oscillatory system. As such, the damping properties of nanocom-
posites, such as loss factor and damping ratio, are an essential de-
sign parameter for many engineering applications. High damping
properties can be achieved in nanocomposites by taking advantage
of the interfacial friction between the nanofillers and the polymer
matrix [186–188]. The combination of extremely large surface
area, weak interfacial bonding with polymer along with low mass
density of CNTs implies that frictional sliding of nanoscale tubes
with polymer matrix can cause significant dissipation of energy
with a minimal weight penalty, providing a high damping capabil-
ity of CNT-filled nanocomposites. Recent studies have shown that
the damping properties of polymer nanocomposites were en-
hanced by more than 200% by incorporating a very small amount
of CNTs [189–193]. The mechanism behind this enhancement
can be explained by the ‘‘stick–slip” theory [186,193]: when the
nanocomposite is subject to an external load, shear stresses are
generated between the CNT walls and the surrounding matrix resin
due to the difference in their elastic properties according to the
shear lag theory. CNTs are deformed together with the matrix if
they are well bonded. However, when the external load exceeds
a critical value leading to debonding of CNTs from the matrix,
the CNTs will stop elongating together with the matrix and a fur-
ther increase of the load can only result in the deformation of
the matrix. Thus, the polymer starts to flow over the surface of
CNTs and deformation energy will be dissipated through the slip-
page between the CNTs and matrix. This phenomenon leads to
the enhanced damping properties of nanocomposites compared
to the neat polymer.
There are several factors governing the damping properties of
CNT/polymer nanocomposites. Suhr et al. [194] studied the damp-
ing properties of PC-based nanocomposites filled with SWCNTs
and MWCNTs. While both of the two CNTs gave rise to higher
damping than the neat PC, the loss modulus of the nanocomposite
containing SWCNTs was significantly higher than that filled with
MWCNTs, suggesting that the inner layers of MWCNTs did not con-
tribute to interfacial frictional sliding with the matrix for energy
dissipation. The effect of CNT dispersion on the damping properties
of nanocomposites shows a similar trend to their elastic mechani-
cal properties: i.e. improved CNT dispersion can enhance the
damping properties of nanocomposites. As shown in Fig. 24A,
sonication for a longer time of 15 min resulted in a better CNT dis-
persion than sonication for 2 min, and the corresponding compos-
ites exhibited much enhanced loss modulus compared to the
counterpart with poor CNT dispersion under the same testing con-
dition [191].
As described above, a weak interfacial adhesion plays an impor-
tant role in enhancing the damping properties of CNT nanocompos-
ites, although it is detrimental to the mechanical properties of CNT/
polymer nanocomposites. When the interfacial bond between the
pristine CNTs and epoxy was weak with its shear strength of about
0.5 MPa, up to 1400% increase in damping ratio was obtained for
the nanocomposites compared to neat polymer. In contrast, a stron-
ger interfacial bond arising from functionalization inhibited the fil-
ler-matrix friction, thereby lowering the damping response, as
shown in Fig. 24B [191]. In other words, covalent bonding of CNTs
with polymer matrix prevented the activation of interfacial sliding
leading to enhanced storage modulus and reduced loss modulus.
As discussed in Section 5.1, CNT alignment can augment the
strength and modulus of CNT/polymer nanocomposites. However,
it may not be beneficial for the enhancement of damping proper-
ties of nanocomposites because randomly-dispersed CNTs can be
subject to interfacial shear stresses from all directions due to the
same load from the initial vibration source [195,196]. Thus, it is ex-
pected that the higher the load applied to the nanocomposite con-
taining random-dispersed CNTs, the more CNT–matrix interfacial
slippage is to be initiated, and thus the higher will be the energy
dissipation.
6. Conclusions and perspectives
The excellent mechanical properties of CNTs combined with un-
ique transport properties along with other multi-functional char-
acteristics render a huge potential for structural and functional
applications of CNT/polymer nanocomposites. Although numerous
studies have dedicated to the development of CNT/polymer nano-
composites for various purposes, their applications in real products
are still in their early stage of realization. It is identified that there
are two major interrelated issues that must be addressed before
this new class of material can be widely applied in real products
and systems: namely (i) lack of solubility and dispersion when
mixed with polymer resins and (ii) poor interfacial adhesion be-
tween CNTs and various polymers.
In this review, we provide an overview of the research in CNT/
polymer nanocomposites with emphases on the principles of CNT
dispersion and functionalization as well as the effects of CNT dis-
persion and functionalization on the properties of CNT/polymer
nanocomposites. It is demonstrated that controlling these two fac-
tors amongst many material and processing parameters are of ut-
most importance as they govern the resulting properties of CNT/
polymer nanocomposites, and uncertainties of these two issues
are often problematic at different stages of nanocomposite fabrica-
tion. Uniform dispersion can be achieved using various types of
mechanical methods, including ultrasonication, shear mixing, cal-
endaring, ball milling, stirring and extrusion. Selection of a proper
method or a combination of several methods as well as their pro-
cessing conditions has to be based on the desired properties of end
products because use of these methods causes, to largely different
degrees, mechanical damage to CNTs and their breakage into smal-
ler pieces.
Along with the development of dispersion techniques, many
techniques have been attempted with varying success to function-
alize the inherently inert nature of CNT surface. There are two ma-
jor approaches, namely chemical and physical functionalization.
Fluorination, cycloaddition, carbene and nitrene addition, chlorina-
tion, bromination, hydrogenation and silanization belong to the
chemical methods that can provide covalent functional groups
onto the CNT surface. The physical methods include wrapping of
polymer around the CNTs, use of surfactants of various ionic nature
 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367
Fig. 24. Effects of CNT dispersion (A) and functionalization (B) on loss modulus of CNT/PC nanocomposites. After [191].
and the endohedral method. These physical methods are known to
cause no structural damage to CNTs. The benefits of functionaliza-
tion are twofold: it not only facilitates the dispersion of CNTs, but
also improves the interfacial interactions between the CNTs and
polymer matrix, both of which are prerequisites to improved
mechanical and some physical properties of CNT/polymer nano-
composites. Similarly to dispersion techniques, however, excessive
use of acids and chemicals at high concentration in the chemical
functionalization techniques may result in damage to CNT
structure.
Even with the well established understanding of the behavior of
CNTs and the properties of CNT/polymer nanocomposites under
different dispersion conditions and functionalization, there are
many fundamental issues that have not been fully explored/under-
stood and need to be further addressed. They include, but are not
limited to, the following areas:
(i) Techniques based on pre-dispersion of CNTs in a solvent, fol-
lowed by mixing with a monomer or a polymer, have been
proven to be an effective way to achieve relatively uniform
CNT dispersion in nanocomposites. However, the evaluation
of CNT dispersion in a solvent is not a trivial matter due to
the non-transparency even at a very low CNT concentration
and in the absence of proven methods to rate the dispersion
quality. It is necessary to develop tools and techniques for
quantitative analysis of degree of dispersion or agglomera-
tion in a solvent or a polymer, which can directly correlate
with a number of material parameters, including the particle
size or aspect ratio, CNT surface functionalities and surface
energies.
(ii) While many studies have hitherto been devised to modify
the CNT surface characteristics, there are many process vari-
ables that have not been optimized to achieve the best out-
come. Additionally, there is concern that structural changes
resulting from the dispersion and functionalization of CNTs
could be detrimental to their desirable properties. Previous
studies have not been able to provide consensus on the fac-
tors governing the dispersion and functionalization of CNTs,
which in turn determine the mechanical and various func-
tional properties of nanocomposites. Overall, systematic
studies are warranted to establish the correlations between
the various surface treatment, degree of CNT functionaliza-
tion, processing and fabrication conditions for CNT/polymer
nanocomposites and their mechanical/functional properties,
including electrical and thermal properties, failure mecha-
nisms and fracture performance of composites in practical
service environments.
1363
(iii) The concept of stress transfer between the matrix and CNTs
in nanocomposites is central to understanding their
mechanical behavior and transportation properties. Stress
transfer depends on the interfacial adhesion between the
CNTs and polymer matrix. A strong interfacial bond in gen-
eral corresponds an effective transfer of the applied load to
the CNTs, resulting in the improved mechanical properties
of nanocomposites. In spite of the large number of publica-
tions devoted to measuring the mechanical properties of
nanocomposites, very little attention has so far been paid
to fundamental understanding of the stress transfer mecha-
nisms and measuring the interfacial bond strength between
the CNTs and polymer matrix. Fundamental studies in this
field, such as modeling, prediction and measurement of
interfacial adhesion, will pave a way for more widespread
applications of CNT/polymer nanocomposites.
(iv) The prohibitively high cost of producing CNTs compared to
other fillers, such as clay, exfoliated graphite, carbon black
and carbon fiber, has limited the potential of employing
CNTs as a filler in nanocomposites. Fabricating nanocompos-
ites using hybrid fillers consisting of CNTs and other low-
cost fillers may offer a promising alternative to lower the
material cost of CNT-based nanocomposites. However,
proper dispersion of these hybrid fillers into a polymer
matrix to achieve the above synergistic effects is a challeng-
ing task because these hybrid fillers have their own unique
geometric shapes and different dispersion characteristics.
(v) Graphene, a single-atom-thick sheet of sp2-bonded carbon
atoms, has generated much interest in recent years due to
its high specific area and exceptional mechanical, electrical,
and thermal properties comparable to CNTs. Bulk production
of graphene sheets from graphite oxide based on a chemical
method may prove to be cheaper than CNTs [197–199],
graphene may become the latest competitor to CNTs in poly-
mer-based nanocomposites. Some recent results showed
that polymer nanocomposites containing a low content of
graphene sheets outperformed CNT counterparts in terms
of enhancing a variety of mechanical properties [200–203].
Extensive research on the use of highly conducting, trans-
parent thin films made from graphene sheets [204] to
replace expensive indium tin oxide films is also underway.
Nevertheless, there are a number of fundamental issues to
be resolved before these materials can find widespread
applications and commercialization. They include lack of
understanding of its reinforcing mechanisms, problems
associated with dispersion of high graphene contents and
the associated poor mechanical performance, difficulty to
1364 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367

improve the interfacial interactions with polymers, optimi- [30] Kim YA, Hayashi T, Fukai Y, et al. Effect of ball milling on morphology of cup-
stacked carbon nanotubes. Chem Phys Lett 2002;355:279–84.
zation of fabrication processes, etc. Further studies are also
[31] Awasthi K, Kamalakaran R, Singh AK, Srivastava ON. Ball-milled carbon and
needed to compare the relative mechanical and functional hydrogen storage. Int J Hydrogen Energy 2002;27:425–32.
performances between graphene and CNT-based nanocom- [32] Ma PC, Tang BZ, Kim JK. Converting semiconducting behavior of carbon
posites made from a range of polymer matrices and filler nanotubes using ball milling. Chem Phys Lett 2008;458:166–9.
[33] Ma PC, Wang SQ, Tang BZ, Kim JK. In-situ amino functionalization of carbon
weight fractions. nanotubes using ball milling. J Nanosci Nanotechnol 2009;9:749–53.
[34] Sandler JKW, Shaffer MSP, Prasse T, Bauhofer W, Schulte K, Windle AH.
Development of a dispersion process for carbon nanotubes in an epoxy
matrix and the resulting electrical properties. Polymer 1999;40:5967–71.
[35] Schmid CF, Klingenberg DJ. Mechanical flocculation in flowing fiber
Acknowledgements suspensions. Phys Rev Lett 2000;84:290–3.
[36] Li J, Ma PC, Chow WS, To CK, Tang BZ, Kim JK. Correlations between
This project was supported by the Research Grants Council of percolation threshold, dispersion state and aspect ratio of carbon nanotube.
Adv Funct Mater 2007;17:3207–15.
Hong Kong SAR (Project No. 614505; 614607) and Finetex-HKUST
[37] Villmow T, Potschke P, Pegel S, Haussler L, Kretzschmar B. Influence of twin-
R&D Center (Project No: FTG001-MECH.07/08). The paper was pre- screw extrusion conditions on the dispersion of multi-walled carbon
sented as the plenary lecture at 10th International Symposium on nanotubes in a poly(lactic acid) matrix. Polymer 2008;49:3500–9.
Polymers for Advanced Technologies (PAT2009), 11–14 October [38] Moniruzzaman M, Du FM, Romero N, Winey KI. Increased flexural modulus
and strength in SWNT/epoxy composites by a new fabrication method.
2009, Jerusalem, Israel. Polymer 2006;47:293–8.
[39] Isayev AI, Kumar R, Lewis TM. Ultrasound assisted twin screw extrusion of
polymer-nanocomposites containing carbon nanotubes. Polymer
References 2009;50:250–60.
[40] Hirsch A. Functionalization of single-walled carbon nanotubes. Angew Chem
Int Ed 2002;41:1853–9.
[1] Iijima S. Helical microtubules of graphitic carbon. Nature 1991;354:56–8.
[41] Balasubramanian K, Burghard M. Chemically functionalized carbon
[2] Bethune DS, Klang CH, De Vries MS, et al. Cobalt-catalysed growth of carbon
nanotubes. Small 2005;1:180–92.
nanotubes with single-atomic-layer walls. Nature 1993;363:605–7.
[42] Hirsch A, Vostrowsky O. Functionalization of carbon nanotubes. Top Curr
[3] Dresselhaus MS, Dresselhaus G, Saito R. Physics of carbon nanotubes. Carbon
Chem 2005;245:93–237.
1995;33:883–91.
[43] Tasis D, Tagmatarchis N, Bianco A, Prato M. Chemistry of carbon nanotubes.
[4] Thostenson ET, Ren ZF, Chou TW. Advances in the science and technology of
Chem Rev 2006;106:1105–36.
CNTs and their composites: a review. Compos Sci Technol 2001;61:1899–912.
[44] Mickelson ET, Huffman CB, Rinzler AG, et al. Fluorination of single-wall
[5] Yakobson BI, Avouris P. Mechanical properties of carbon nanotubes. Top Appl
carbon nanotubes. Chem Phys Lett 1998;296:188–94.
Phys 2001;80:287–327.
[45] Kelly KF, Chiang IW, Mickelson ET, et al. Insight into the mechanism of
[6] Qian D, Wagner GJ, Liu WK, Yu MF, Ruoff RS. Mechanics of carbon nanotubes.
sidewall functionalization of single-walled nanotubes: an STM study. Chem
Appl Mech Rev 2002;55:495–533.
Phys Lett 1999;313:445–50.
[7] Reich S, Thomsen C, Maultzsch J. Carbon nanotubes: basic concepts and
[46] Touhara H, Inahara J, Mizuno T, et al. Fluorination of cup-stacked carbon
physical properties. New York: Wiley-VCH; 2004. p. 31–40.
nanotubes, structure and properties. Fluorine Chem 2002;114:181–8.
[8] Koo JH. Polymer nanocomposites: processing, characterization, and
[47] Stevens JL, Huang AY, Peng H, et al. Sidewall amino-functionalization of
applications. New York: McGraw-Hill; 2006. p. 1–8.
SWNTs through fluorination and subsequent reactions with terminal
[9] Ajayan PM, Schadler LS, Braun PV. Nanocomposite science and
diamines. Nano Lett 2003;3:331–6.
technology. Weinheim: Wiley-VCH; 2003. p. 77–80.
[48] Zhang QH, Chen DJ. Percolation threshold and morphology of composites of
[10] Coleman JN, Khan U, Gunko YK. Mechanical reinforcement of polymers using
conducting carbon black/polypropylene/EVA. J Mater Sci 2004;39:1751–7.
carbon nanotubes. Adv Mater 2006;18:689–706.
[49] Chen J, Hamon MA, Hu H, et al. Solution properties of single-walled carbon
[11] Kim JK, Mai YW. Engineered interfaces in fiber reinforced
nanotubes. Science 1998;282:95–8.
composites. Oxford: Elsevier; 1998. p. 1–100.
[50] Hu H, Zhao B, Hamon MA, et al. Sidewall functionalization of single-walled
[12] Hodzic A, Stachurski ZH, Kim JK. Nano-indentation of polymer–glass
carbon nanotubes by addition of dichlorocarbene. J Am Chem Soc
interfaces part I: experimental and mechanical analysis. Polymer
2003;125:14893–900.
2000;41:6895–905.
[51] Holzinger M, Steinmetz J, Samaille D, et al. [2 + 1] Cycloaddition for cross-
[13] Kim JK, Sham ML, Wu JS. Nanoscale characterisation of interphase in silane
linking SWCNTs. Carbon 2004;42:941–7.
treated glass fibre composites. Composites Part A 2001;32:607–18.
[52] Unger E, Graham A, Kreupl F, et al. Electrochemical functionalization of multi-
[14] Thess A, Lee R, Nikolaev P, et al. Crystalline ropes of metallic carbon
walled carbon nanotubes for solvation and purification. Curr Appl Phys
nanotubes. Science 1996;273:483–8.
2002;2:107–11.
[15] Breuer O, Sundararaj U. Big returns from small fibers: a review of polymer/
[53] Kim KS, Bae DJ, Kim JR, et al. Modification of electronic structures of a carbon
carbon nanotube composites. Polym Compos 2004;25:630–45.
nanotube by hydrogen functionalization. Adv Mater 2002;14:1818–21.
[16] Xie XL, Mai YW, Zhou XP. Dispersion and alignment of carbon nanotubes in
[54] Tagmatarchis N, Prato MJ. Functionalization of carbon nanotubes via 1,3-
polymer matrix: a review. Mater Sci Eng R 2005;49:89–112.
dipolar cycloadditions. J Mater Chem 2004;14:437–9.
[17] Moniruzzaman M, Winey KI. Polymer nanocomposites containing carbon
[55] Esumi K, Ishigami M, Nakajima A, et al. Chemical treatment of carbon
nanotubes. Macromolecules 2006;39:5194–205.
nanotubes. Carbon 1996;34:279–81.
[18] Fiedler B, Gojny FH. Fundamental aspects of nano-reinforced composites.
[56] Yu R, Chen L, Liu Q, Lin J, Tan KL, Ng SC, et al. Platinum deposition on carbon
Compos Sci Technol 2006;66:3115–25.
nanotubes via chemical modification. Chem Mater 1998;10:718–22.
[19] Gibson RF, Ayorinde EO, Wen YF. Vibrations of carbon nanotubes and their
[57] Sham ML, Kim JK. Surface functionalities of multi-wall carbon nanotubes
composites: a review. Compos Sci Technol 2007;67:1–28.
after UV/ozone and TETA treatments. Carbon 2006;44:768–77.
[20] Bal S, Samal SS. Carbon nanotube reinforced polymer composites – a state of
[58] Ma PC, Kim JK, Tang BZ. Functionalization of carbon nanotubes using a silane
the art. Bull Mater Sci 2007;30:379–86.
coupling agent. Carbon 2006;44:3232–8.
[21] <http://www.sonifier.com>, sonifier products, Branson ultrasonics corp
[59] Wang SC, Chang KS, Yuan CJ. Enhancement of electrochemical properties of
(accessed June 2010).
screen-printed carbon electrodes by oxygen plasma treatment. Electrochim
[22] Lu KL, Lago RM, Chen YK, Green MLH, Harris PJF, Tsang SC. Mechanical
Acta 2009;54:4937–43.
damage of carbon nanotubes by ultrasound. Carbon 1996;34:814–6.
[60] Ávila-Orta CA, Cruz-Delgado VJ, Neira-Velázquez MG, et al. Surface
[23] Mukhopadhyay K, Dwivedi CD, Mathur GN. Conversion of carbon nanotubes
modification of carbon nanotubes with ethylene glycol plasma. Carbon
to carbon nanofibers by sonication. Carbon 2002;40:1373–6.
2009;47:1916–21.
[24] <http://www.exakt.com>, three roll mills, EXAKT company (accessed June
[61] Liu P. Modifications of carbon nanotubes with polymers. Eur Polym J
2010).
2005;41:2693–703.
[25] Gojny FH, Wichmann MHG, Köpke U, Fiedler B, Schulte K. Carbon nanotube
[62] Hamon MA, Hui H, Bhowmik P. Ester-functionalized soluble single-walled
reinforced epoxy-composites: enhanced stiffness and fracture toughness at
carbon nanotubes. Appl Phys A 2002;74:333–8.
low nanotube content. Compos Sci Technol 2004;64:2363–71.
[63] Liu J, Rinzler AG, Dai HJ, et al. Fullerene pipes. Science 1998;280:1253–6.
[26] Thostenson ET, Chou TW. Processing-structure-multi-functional property
[64] Stephenson JJ, Sadana AK, Higginbotham AL, Tour JM. Highly functionalized
relationship in carbon nanotube/epoxy composites. Carbon 2006;44:3022–9.
and soluble multiwalled carbon nanotubes by reductive alkylation and
[27] Li YB, Wei BQ, Liang J, et al. Transformation of carbon nanotubes to
arylation: the billups reaction. Chem Mater 2006;18:4658–61.
nanoparticles by ball milling process. Carbon 1999;37:493–7.
[65] McCarthy B, Coleman JN, Czerw R, Dalton AB, Carroll DL, Blau WJ. Microscopy
[28] Gao B, Bower C, Lorentzen JD, et al. Enhanced saturation lithium composition
studies of nanotube-conjugated polymer interactions. Synth Met
in ball-milled single-walled carbon nanotubes. Chem Phys Lett
2001;121:1225–6.
2000;327:69–75.
[66] Hill DE, Lin Y, Rao AM, Allard LF, Sun YP. Functionalization of carbon
[29] Huang JY, Yasuda H, Mori H. Highly curved carbon nanostructures produced
nanotubes with polystyrene. Macromolecules 2002;35:9466–71.
by ball-milling. Chem Phys Lett 1999;303:130–4.
 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367
[67] Gong X, Liu J, Baskaran S, Voise RD, Young JS. Surfactant-assisted processing
of carbon nanotube/polymer composites. Chem Mater 2000;12:1049–52.
[68] Cui S, Canet R, Derre A, et al. Characterization of multiwall carbon nanotubes
and influence of surfactant in the nanocomposite processing. Carbon
2003;41:797–809.
[69] Vaisman L, Marom G, Wagner HD. Dispersions of surface-modified carbon
nanotubes in water-soluble and water-insoluble polymers. Adv Funct Mater
2006;16:357–63.
[70] Geng Y, Liu MY, Li J, Shi XM, Kim JK. Effects of surfactant treatment on
mechanical and electrical properties of CNT/epoxy nanocomposites.
Composites Part A 2008;39:1876–83.
[71] Strano MS, Moore VC, Miller MK, et al. The role of surfactant adsorption
during ultrasonication in the dispersion of single-walled carbon nanotubes. J
Nanosci Nanotechnol 2003;3:81–6.
[72] Islam MF, Rojas E, Bergey DM, et al. High weight fraction surfactant
solubilization of single-wall carbon nanotubes in water. Nano Lett
2003;3:269–73.
[73] Yu J, Grossiord N, Koning CE, Loos J. Controlling the dispersion of multi-wall
carbon nanotubes, in aqueous surfactant solution. Carbon 2007;45:618–23.
[74] Whitsitt EA, Barron AR. Silica coated single walled carbon nanotubes. Nano
Lett 2003;3:775–8.
[75] Kim TH, Doe C, Kline SR, Choi S. Water-redispersible isolated single-walled
carbon nanotubes fabricated by in situ polymerization of micelles. Adv Mater
2007;19:929–33.
[76] Vaisman L, Wagner HD, Marom G. The role of surfactants in dispersion of
carbon nanotubes. Adv Colloid Interface Sci 2006;128–130:37–46.
[77] Grossiord N, Loos J, Regev O, Koning CE. Toolbox for dispersing carbon
nanotubes into polymers to get conductive nanocomposites. Chem Mater
2006;18:1089–99.
[78] Georgakilas V, Gournis D, Tzitzios V, et al. Decorating carbon nanotubes with
metal or semiconductor nanoparticles. J Mater Chem 2007;17:2679–94.
[79] Tsang SC, Guo Z, Chen YK, et al. Immobilization of platinated and iodinated
oligonucleotides on carbon nanotubes. Angew Chem Int Ed
1997;36:2198–200.
[80] Ajayan PM, Stephan O, Colliex C, Trauth D. Aligned carbon nanotube arrays
formed by cutting a polymer resin-nanotube composite. Science
1994;265:1212–4.
[81] Summary of Searching Results. <http://www.scopus.com> (accessed January
2010).
[82] Du JH, Bai J, Cheng HM. The present status and key problems of carbon
nanotube based polymer composites. Express Polym Lett 2007;1:253–73.
[83] Kim JY, Kim SH. Influence of multiwall carbon nanotube on physical
properties of poly(ethylene 2,6-naphthalate) nanocomposites. J Polym Sci B
2006;44:1062–71.
[84] Moisala A, Li Q, Kinloch IA, Windle AH. Thermal and electrical conductivity of
single- and multi-walled carbon nanotube-epoxy composites. Compos Sci
Technol 2006;66:1285–8.
[85] Ma PC, Kim JK, Tang BZ. Effects of silane functionalization on the properties of
carbon nanotubes/epoxy nanocomposites. Compos Sci Technol
2007;67:2965–72.
[86] Ma PC, Tang BZ, Kim JK. Effect of CNT decoration with silver nanoparticles on
electrical conductivity of CNT–polymer composites. Carbon
2008;46:1497–505.
[87] Kosmidou TV, Vatalis AS, Delides CG, et al. Structural, mechanical and
electrical characterization of epoxy-amine/carbon black nanocomposites.
Express Polym Lett 2008;2:364–72.
[88] Ma PC, Liu MY, Kim JK, Tang BZ, et al. Development of electrically conducting
nanocomposites by employing hybrid fillers of carbon nanotubes and carbon
black. ACS Appl Mater Interfaces 2009;1:1090–6.
[89] Wardle BL, Saito DS, Garcia EJ, Hart AJ, Villoria RG, Verploegen EA. Fabrication
and characterization of ultrahigh-volume fraction aligned carbon nanotube-
polymer composites. Adv Mater 2008;20:2707–14.
[90] Vigolo B, Penicaud A, Coulon C, et al. Macroscopic fibers and ribbons of
oriented carbon nanotubes. Science 2000;290:1331–4.
[91] Mamedov AA, Kotov NA, Prato M, et al. Molecular design of strong single-wall
carbon nanotube/polyelectrolyte multilayer composites. Nat Mater
2002;1:190–4.
[92] Xia H, Wang Q, Li K, Hu GH. Preparation of polypropylene/carbon nanotube
composite powder with a solid-state mechanochemical pulverization
process. J Appl Polym Sci 2004;93:378–86.
[93] Masuda J, Torkelson JM. Dispersion and major property enhancements in
polymer/multiwall carbon nanotube nanocomposites via solid-state
shear pulverization followed by melt mixing. Macromolecules 2008;41:
5974–7.
[94] Qian D, Dickey EC. Load transfer and deformation mechanisms in carbon
nanotube polystyrene composites. Appl Phys Lett 2000;76:2868–70.
[95] Coleman JN, Khan U, Blau WJ, Gun’ko YK. Small but strong: a review of the
mechanical properties of carbon nanotubes-polymer composites. Carbon
2006;44:1624–52.
[96] Jiang B, Liu C, Zhang C, Wang B, Wang Z. The effect of non-symmetric
distribution of fiber orientation and aspect ratio on elastic properties of
composites. Composites Part B 2007;38:24–34.
[97] Hernández-Pérez A, Avilés F, May-Pat A, Valadez-González A, Herrera-Franco
PJ, Bartolo-Pérez P. Effective properties of multiwalled carbon nanotube/
epoxy composites using two different tubes. Compos Sci Technol
2008;68:1422–14231.
1365
[98] Liu LQ, Wagner HD. Rubbery and glassy epoxy resins reinforced with carbon
nanotubes. Compos Sci Technol 2005;65:1861–8.
[99] Ma PC, Siddiqui NA, Cao SG, Tang BZ, Kim JK. Effect of dispersion techniques
and functionalization of CNTs on the properties of CNT/epoxy composites. In:
Proceedings of the 2nd int conference on advanced materials and structures,
Nanjing, China, October; 2008.
[100] Thostenson ET, Chou TW. Aligned multi-walled carbon nanotube-reinforced
composites: processing and mechanical characterization. J Phys D
2002;35:77–80.
[101] Jin L, Bower C, Zhou O. Alignment of carbon nanotubes in a polymer matrix
by mechanical stretching. Appl Phys Lett 1998;73:1197–9.
[102] Fornes TD, Baur JW, Sabba Y, Thomas EL. Morphology and properties of melt-
spun polycarbonate fibers containing single- and multi-wall carbon
nanotubes. Polymer 2006;47:1704–14.
[103] Wang MW. Alignment of multiwall carbon nanotubes in polymer composites
by dielectrophoresis. Jpn J Appl Phys 2009;48:035002.
[104] Park C, Wilkinson J, Banda S, Ounaies Z, Wise KE, Sauti G, et al. Aligned single-
wall carbon nanotube polymer composites using an electric field. J Polym Sci
B 2006;44:1751–62.
[105] Steinert BW, Dean DR. Magnetic field alignment and electrical properties of
solution cast PET-carbon nanotube composite films. Polymer
2009;50:898–904.
[106] Zhu YF, Ma C, Zhang W, Zhang RP, Koratkar N, Liang J. Alignment of
multiwalled carbon nanotubes in bulk epoxy composites via electric field. J
Appl Phys 2009;105:054319–26.
[107] Desai AV, Haque MA. Mechanics of the interface for carbon nanotube-
polymer composites. Thin-Walled Struct 2005;43:1787–803.
[108] Meng H, Sui GX, Fang PF, Yang R. Effects of acid- and diamine-modified
MWNTs on the mechanical properties and crystallization behavior of
polyamide 6. Polymer 2008;49:610–20.
[109] Yang BX, Shi JH, Li X, Pramoda KP, Goh SH. Mechanical reinforcement of
poly(1-butene) using polypropylene-grafted multiwalled carbon nanotubes. J
Appl Polym Sci 2009;113:165–72.
[110] Yang BX, Pramoda KP, Xu GQ, Goh SH. Mechanical reinforcement of
polyethylene using polyethylene-grafted multiwalled carbon nanotubes.
Adv Funct Mater 2007;17:2062–9.
[111] Yuen SM, Ma CCM, Lin YY, Kuan HC. Preparation, morphology and properties
of acid and amine modified multiwalled carbon nanotube/polyimide
composite. Compos Sci Technol 2007;67:2564–73.
[112] Koval’chuk AA, Shevchenko VG, Shchegolikhin AN, Nedorezova PM,
Klyamkina AN, Aladyshev AM. Effect of carbon nanotube functionalization
on the structural and mechanical properties of polypropylene/MWCNT
composites. Macromolecules 2008;41:7536–42.
[113] Byrne MT, McNamee WP, Gun’ko YK. Chemical functionalization of carbon
nanotubes for the mechanical reinforcement of polystyrene composites.
Nanotechnology 2008;19:415707.
[114] Paiva MC, Zhou B, Fernando KAS, Lin Y, Kennedy JM, Sun YP. Mechanical and
morphological characterization of polymer-carbon nanocomposites from
functionalized carbon nanotubes. Carbon 2004;42:2849–54.
[115] Kim KH, Jo WH. Improvement of tensile properties of poly(methyl
methacrylate) by dispersing multi-walled carbon nanotubes functionalized
with poly(3-hexylthiophene)-graft-poly(methyl methacrylate). Compos Sci
Technol 2008;68:2120–4.
[116] Xia H, Song M. Preparation and characterisation of polyurethane grafted
single-walled carbon nanotubes and derived polyurethane nanocomposites. J
Mater Chem 2006;16:1843–51.
[117] Sahoo NG, Jung YC, Yoo HJ, Cho JW. Effect of functionalized carbon nanotubes
on molecular interaction and properties of polyurethane composites.
Macromol Chem Phys 2006;207:1773–80.
[118] Sui G, Zhong WH, Yang XP, Yu YH. Curing kinetics and mechanical behavior of
natural rubber reinforced with pretreated carbon nanotubes. Mater Sci Eng A
2008;485:524–31.
[119] Shanmugharaj AM, Bae JH, Lee KY, Noh WH, Lee SH, Ryu SH. Physical and
chemical characteristics of multiwalled carbon nanotubes functionalized
with amino silane and its influence on the properties of natural rubber
composites. Compos Sci Technol 2007;67:1813–22.
[120] Ping ZY, Wang CC. Polymer world. Shanghai: Fudan University Press; 2001. p.
183.
[121] Zhang W, Dehghani-Sanij AA, Blackburn RS. Carbon based conductive
polymer composites. J Mater Sci 2007;42:3408–18.
[122] Aneli JN, Khananasvili LM, Zaikov GE. Structuring and conductivity of
polymer composites. New York: Nova Science Publishers; 1998. p. 1–100.
[123] Li J, Kim JK, Sham ML, Marom G. Morphology and properties of UV/ozone
treated graphite nanoplatelet/epoxy nanocomposites. Compos Sci Technol
2007;67:296–305.
[124] Sham ML, Li J, Ma PC, Kim JK. Cleaning and functionalizaton of polymer
surfaces and nanoscale carbon fillers by UV/ozone treatment: a review. J
Compos Mater 2009;43:1537–64.
[125] Gul’ VE. Structure and properties of conducting polymer
composites. Netherlands: VSP; 1996. p. 27–123.
[126] Bauhofer W, Kovacs JZ. A review and analysis of electrical percolation in
carbon nanotube polymer composites. Compos Sci Technol
2009;69:1486–98.
[127] Sandler JKW, Kirk JE, Kinloch IA, Shaffer MSP, Windle AH. Ultra-low electrical
percolation threshold in carbon nanotube-epoxy composites. Polymer
2003;44:5893–9.
1366 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367
[128] Martina CA, Sandler JKW, Windle AH, et al. Electric field-induced aligned
multi-wall carbon nanotube networks in epoxy composites. Polymer
2005;46:877–86.
[129] Buldum A, Lu JP. Contact resistance between carbon nanotubes. Phys Rev B
2001;63:161403-1–3-4.
[130] Stadermann M, Papadakis SJ, Falvo MR, et al. Nanoscale study of conduction
through carbon nanotube networks. Phys Rev B 2004;69:201402-1–2-3.
[131] Li J, Wong PS, Kim JK. Hybrid nanocomposites containing carbon nanotubes
and graphite nanoplatelets. Mater Sci Eng A 2008;483–484:660–3.
[132] Peng H, Jain M, Peterson DE, Zhu Y, Jia Q. Composite carbon nanotube/silica
fibers with improved mechanical strengths and electrical conductivities.
Small 2008;4:1964–7.
[133] Bokobza L, Rahmani M, Belin C, Bruneel J, Bounia NEE. Blends of carbon
blacks and multiwall carbon nanotubes as reinforcing fillers for hydrocarbon
rubbers. J Polym Sci B 2008;46:1939–51.
[134] Fan ZJ, Wang Y, Luo GH, Li ZF, Wei F. The synergetic effect of carbon
nanotubes and carbon black in a rubber system. New Carbon Mater
2008;23:149–53.
[135] Sumfleth J, Adroher XC, Schulte K. Synergistic effects in network formation
and electrical properties of hybrid epoxy nanocomposites containing multi-
wall carbon nanotubes and carbon black. J Mater Sci 2009;44:3241–7.
[136] Sun Y, Bao HD, Guo ZX, Yu J. Modeling of the electrical percolation of mixed
carbon fillers in polymer-based composites. Macromolecules
2009;42:459–63.
[137] Phillpot SR, McGaughey AJH. Introduction to thermal transport. Mater Today
2005;8:18–20.
[138] Biercuk MJ, Llaguno MC, Radosavljevic M, Hyun JK, Johnson AT, Fischer JE.
Carbon nanotube composites for thermal management. Appl Phys Lett
2002;80:2767–9.
[139] Evseeva LE, Tanaeva SA. Thermal conductivity of micro-and nanostructural
epoxy composites at low temperatures. Mech Compos Mater 2008;44:87–92.
[140] Guthy C, Du F, Brand S, Winey KI, Fischer JE. Thermal conductivity of single-
walled carbon nanotube/PMMA composites. J Heat Trans 2007;129:1096–9.
[141] Shenogin S, Xue LP, Ozisik R, Keblinski P, Cahill DG. Role of thermal boundary
resistance on the heat flow in carbon-nanotube composites. J Appl Phys
2004;95:8136–44.
[142] Cai D, Song M. Latex technology as a simple route to improve the thermal
conductivity of a carbon nanotube/polymer composite. Carbon
2008;46:2107–12.
[143] Gojny FH, Wichmann MHG, Fiedler B, Kinloch IA, Bauhofer W, Windle AH,
et al. Evaluation and identification of electrical and thermal conduction
mechanisms in carbon nanotube/epoxy composites. Polymer
2006;47:2036–45.
[144] Velasco-Santos C, Martínez-Hernández AL, Fisher FT, Ruoff R, Castaño VM.
Improvement of thermal and mechanical properties of carbon nanotube
composites through chemical functionalization. Chem Mater
2003;15:4470–5.
[145] Kashiwagi T, Grulke E, Hilding J, Harris R, Awad W, Douglas J. Thermal
degradation and flammability properties of poly(propylene)/carbon
nanotube composites. Macromol Rapid Commun 2002;232:761–5.
[146] Kashiwagi T, Du F, Douglas JF, Winey KI, Harris RH, Shields JR. Nanoparticle
networks reduce the flammability of polymer nanocomposites. Nat Mater
2005;4:928–33.
[147] Kashiwagi T, Du F, Winey KI, Groth KM, Shields JR, Bellayer SP, et al.
Flammability properties of polymer nanocomposites with single-walled
carbon nanotubes: effects of nanotube dispersion and concentration.
Polymer 2005;46:471–81.
[148] Schartel B, Pötschke P, Knoll PU, Abdel-Goad M. Fire behavior of
polyamide 6/multiwall carbon nanotube nanocomposites. Eur Polym J
2005;41:1061–70.
[149] Riggs JE, Walker DB, Carroll DL, Sun YP. Optical limiting properties of
suspended and solubilized carbon nanotubes. J Phys Chem B
2000;104:7071–6.
[150] O’Flaherty SM, Hold SV, Brennan ME, Cadek M, Drury A, Coleman JN, et al.
Nonlinear optical response of multiwalled carbon-nanotube dispersions. J
Opt Soc Am B 2003;20:49–58.
[151] Sakakibara Y, Itatani T, Matsuzaki S, et al. Carbon nanotube polymer
nanocomposite and its nonlinear optical applications. In: European
conference on optical communications proceedings, Cannes, France,
September; 2006 [Paper ID: 4801138].
[152] Tang BZ, Xu HY. Preparation, alignment, and optical properties of soluble
poly(phenylacetylene) wrapped carbon nanotubes. Macromolecules
1999;32:2569–76.
[153] Wang J, Blau WJ. Solvent effect on optical limiting properties of single-walled
carbon nanotubes dispersions. J Phys Chem C 2008;112:2298–303.
[154] Zhang L, Allen SD, Woelfle C, Zhang F. Influence of polymer structures on
optical power limiting performance of single-walled carbon nanotubes. J
Phys Chem C 2009;113:13979–84.
[155] Qiu XQ, Wu HX, Tong R, Qian SX, Lin YH, Cai RF. Optical limiting properties of
two soluble polymer/multi-walled carbon nanotube composites. Chinese
Phys Lett 2008;25:536–9.
[156] Chen Y, Lin Y, Liu Y, et al. Carbon nanotube-based functional materials for
optical limiting. J Nanosci Nanotechnol 2007;7:1268–83.
[157] Henley SJ, Hatton RA, Chen GY, Gao C, Zeng H, Kroto HW, et al. Enhancement
of polymer luminescence by excitation-energy transfer from multi-walled
carbon nanotubes. Small 2007;3:1927–33.
[158] Singh I, Bhatnagar PK, Mathur PC, et al. Optical and electrical characterization
of conducting polymer single walled carbon nanotube composite films.
Carbon 2008;46:1141–4.
[159] He N, Chen Y, Bai J, Wang J, Blau WJ, Zhu J. Preparation and optical limiting
properties of multiwalled carbon nanotubes with p-conjugated metal-free
phthalocyanine moieties. J Phys Chem C 2009;113:13029–35.
[160] Lin Y, Zhou B, Martin RB, et al. Visible luminescence of carbon nanotubes and
dependence on functionalization. J Phys Chem A 2005;109:14779–82.
[161] Guldi DM, Holzinger M, Hirsch A, Georgakilas V, Prato V. First comparative
emission assay of single-wall carbon nanotubes-solutions and dispersions.
Chem Commun 2003:1130–1.
[162] Baibarac M, Gómez-Romero P. Nanocomposites based on conducting
polymers and carbon nanotubes from fancy materials to functional
applications. J Nanosci Nanotechnol 2006;6:1–14.
[163] Kim JY, Kim M, Kim H, Joo J, Choi JH. Electrical and optical studies of organic
light emitting devices using SWCNTs-polymer nanocomposites. Opt Mater
2003;21:147–51.
[164] Huang JW, Bai SJ. Light emitting diodes of fully conjugated heterocyclic
aromatic rigid-rod polymer doped with multi-wall carbon nanotubes.
Nanotechnology 2005;16:1406–10.
[165] Arranz-Andrés J, Blau WJ. Enhanced device performance using different
carbon nanotube types in polymer photovoltaic devices. Carbon
2008;46:2067–75.
[166] Valentini L, Armentano I, Mengoni F, et al. Chemical gating and
photoconductivity of CF4 plasma-functionalized single-walled carbon
nanotubes with adsorbed butylamine. J Appl Phys 2005;97:114320.
[167] Valentini L, Mengoni F, Armentano I, et al. Enhancement of photoelectrical
properties in polymer nanocomposites containing modified single-walled
carbon nanotubes by conducting dendrimer. J Appl Phys 2006;99:114305.
[168] Xu WJ, Chen HZ, Shi MM, Huang YG, Wang M. Poly(triphenylamine) related
copolymer noncovalently coated MWCNT nanohybrid: fabrication and
observation of enhanced photoconductivity. Nanotechnology
2006;17:728–33.
[169] Yun D, Feng W, Wu H, et al. Controllable functionalization of single-wall
carbon nanotubes by in situ polymerization method for organic photovoltaic
devices. Synth Met 2008;158:977–83.
[170] Kalita G, Adhikari S, Aryal HR, et al. Functionalization of multi-walled carbon
nanotubes (MWCNTs) with nitrogen plasma for photovoltaic device
application. Curr Appl Phys 2009;9:346–51.
[171] Lu S, Ahir S, Velasco V, King B, Xu P, Terentjev EM, et al. Photo-mechanical
actuation of carbon nanotubes: mechanisms and applications in micro and
nano-devices. J Micro-Nano Mech 2009;5:29–41.
[172] Zhang Y, Ijima S. Elastic response of carbon nanotubes bundles to visible
light. Phys Rev Lett 1999;82:3472–5.
[173] Landi BJ, Raffaelle RP, Heben MJ, Alleman JL, VanDerveer W, Gennett T. Single
wall carbon nanotubes-nafion composite actuators. Nano Lett
2002;2:1329–32.
[174] Ahir SV, Terentjev EM. Photomechanical actuation in polymer-nanotube
composites. Nat Mater 2005;4:491–5.
[175] Lu S, Panchapakesan B. Optically driven nanotubes actuators.
Nanotechnology 2005;16:2548–54.
[176] Prentice P. Rheology and its role in plastics processing (Rapra review report).
Rapra Technology Ltd., vol. 7; 1995. p. 3.
[177] Liu C, Zhang J, He J, Hu G. Gelation in carbon nanotube/polymer composites.
Polymer 2003;44:7529–32.
[178] Du F, Scogna RC, Zhou W, Brand S, Fischer JE, Winey KI. Nanotube networks in
polymer nanocomposites: rheology and electrical conductivity.
Macromolecules 2004;37:9048–55.
[179] Huang YY, Ahir SV, Terentjev EM. Dispersion rheology of carbon nanotubes in
a polymer matrix. Phys Rev B 2006;73:125422.
[180] Fan Z, Advani SG. Rheology of multiwall carbon nanotube suspensions. J
Rheol 2007;51:585–604.
[181] Abbasi S, Carreau PJ, Derdouri A, Moan M. Rheological properties and
percolation in suspensions of multiwalled carbon nanotubes in
polycarbonate. Rheol Acta 2009;48:943–59.
[182] Bose S, Bhattacharyya AR, Kulkarni AR, Pötschke P. Electrical, rheological and
morphological studies in co-continuous blends of polyamide 6 and
acrylonitrile–butadiene–styrene with multiwall carbon nanotubes prepared
by melt blending. Compos Sci Technol 2009;69:365–72.
[183] Siddiqui NA, Ma PC, Li CY, Yu YS, Kim JK. Manufacturing and characterization
of CFRP prepregs containing carbon nanotubes. Compos Sci Technol
submitted for publication.
[184] Mitchell CA, Bahr JL, Arepalli S, Tour JM, Krishnamoorti R. Dispersion of
functionalized carbon nanotubes in polystyrene. Macromolecules
2002;35:8825–30.
[185] Ma PC, Mo SY, Tang BZ, Kim JK. Dispersion, interfacial interaction and
reagglomeration of functionalized carbon nanotubes in epoxy composites.
Carbon 2010;48:1824–34.
[186] Buldum A, Lu JP. Atomic scale sliding and rolling of carbon nanotubes. Phys
Rev Lett 1999;83:5050–3.
[187] Koratkar NA, Wei BQ, Ajayan PM. Carbon nanotube films for damping
applications. Adv Mater 2002;14:997–1000.
[188] Koratkar NA, Wei BQ, Ajayan PM. Multifunctional structural reinforcement
featuring carbon nanotube films. Compos Sci Technol 2003;63:1525–31.
[189] Suhr J, Koratkar N, Keblinski P, Ajayan P. Viscoelasticity in carbon nanotube
composites. Nat Mater 2005;4:134–7.
 P.-C. Ma et al. / Composites: Part A 41 (2010) 1345–1367
[190] Salehi-Khojin A, Jalili N. A comprehensive model for load transfer in nanotube
reinforced piezoelectric polymeric composites subjected to electro-thermo-
mechanical loadings. Composites Part B 2008;39:986–98.
[191] Suhr J, Koratkar NA. Energy dissipation in carbon nanotube composites: a
review. J Mater Sci 2008;43:4370–82.
[192] Auad ML, Mosiewicki MA, Uzunpinar C, Williams RJJ. Single-wall carbon
nanotubes/epoxy elastomers exhibiting high damping capacity in an
extended temperature range. Compos Sci Technol 2009;69:1088–92.
[193] Koratkar NA, Suhr J, Joshi A, et al. Characterizing energy dissipation in single-
walled carbon nanotube polycarbonate composites. Appl Phys Lett
2005;87:063102.
[194] Suhr J, Schadler L, Ajayan P, Koratkar N. Comparing damping properties of
singlewalled and multiwalled carbon nanotube polymer composites. In: The
46th AIAA/ASME/ASCE/AHS/ASC structures, structural dynamics and
materials conference. Austin, April; 2005. p. 1550–7.
[195] Zhou X, Shin E, Wang KW, Bakis CE. Interfacial damping characteristics of
carbon nanotube composites. Compos Sci Technol 2004;64:2425–37.
[196] Rajoria H, Jalili N. Passive vibration damping enhancement using carbon
nanotube-epoxy reinforced composites. Compos Sci Technol 2005;65:2079–93.
1367
[197] Li D, Müller MB, Gilje S, Kaner RB, Wallace GG. Processable aqueous
dispersions of graphene nanosheets. Nat Nanotechnol 2008;3:101–5.
[198] Geng Y, Wang SJ, Kim JK. Preparation of graphite nanoplatelets and graphene
sheets. J Colloid Interface Sci 2009;336:592–8.
[199] Park S, Ruoff RS. Chemical methods for the production of graphenes. Nat
Nanotechnol 2009;4:217–24.
[200] Ramanathan T, Abdala AA, Stankovich S, Dikin DA, Herrera-Alonso M, Piner
RD, et al. Functionalized graphene sheets for polymer nanocomposites. Nat
Nanotechnol 2008;3:327–31.
[201] Rafiee MA, Rafiee J, Wang Z, Song H, Yu ZZ, Koratkar N. Enhanced mechanical
properties of nanocomposites at low graphene content. ACS Nano
2009;3:3884–90.
[202] Rafiee MA, Rafiee J, Yu ZZ, Koratkar N. Buckling resistant graphene
nanocomposites. Appl Phys Lett 2009;95:223103.
[203] Rafiee MA, Rafiee J, Srivastava I, Wang Z, Song H, Yu ZZ, et al. Fracture and
fatigue in graphene nanocomposites. Small 2010;6:179–83.
[204] Wang SJ, Geng Y, Zheng QB, Kim JK. Fabrication of highly conducting and
transparent graphene films. Carbon 2010;48:1815–23.