An exfoliation of organoclay in thermotropic liquid crystalline polyester nanocomposites

Polymer 43 (2002) 2969±2974

www.elsevier.com/locate/polymer

An exfoliation of organoclay in thermotropic liquid crystalline
polyester nanocomposites
Jin-Hae Chang a,*, Bo-Soo Seo a, Do-Hoon Hwang b
a

Department of Polymer Science and Engineering, Kumoh National University of Technology, Kumi 730-701, South Korea
b
Department of Applied Chemistry, Kumoh National University of Technology, Kumi 730-701, South Korea
Dedicated to Prof. Jung-Il Jin of Korea University, Seoul, Korea, on the occasion of his 60th birthday
Received 9 October 2001; received in revised form 7 January 2002; accepted 8 February 2002

Abstract
A thermotropic liquid crystalline polyester (TLCP) with an alkoxy side-group was synthesized from 2-ethoxyhydroquinone and 2bromoterephthalic acid. Nanocomposites of TLCP with Cloisite 25A (C25A) as an organoclay were prepared by the melting intercalation
method above the melt transition temperature (Tm) of the TLCP. Liquid crystallinity, morphology, and thermo-mechanical behaviors were
examined with increasing organoclay content from 0 to 6%. Liquid crystallinity of the C25A/TLCP hybrids was observed when organoclay
content was up to 6%. Regardless of the clay content in the hybrids, the C25A in TLCP was highly dispersed in a nanometer scale. The
hybrids (0±6% C25A/TLCP) were processed for ®ber spinning to examine their tensile properties. Ultimate strength and initial modulus of
the TLCP hybrids increased with increasing clay content and the maximum values of the mechanical properties were obtained from the
hybrid containing 6% of the organoclay. Thermal, morphological and mechanical properties of the nanocomposites were examined by
differential scanning calorimetry (DSC), thermogravimetric analyzer (TGA), polarized optical microscope, electron microscopes (SEM and
TEM), and capillary rheometer. q 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Thermotropic liquid crystalline polyester; Organoclay; Nanocomposite

1. Introduction
Thermotropic liquid crystalline polymers have already
been established as high performance commercial engineering polymers. This is due to their speci®c chemical
structures, high strengths, high moduli, low viscosities,

and other good mechanical properties [1±4]. The structureproperty relationships of thermotropic liquid crystalline
polyesters (TLCPs) have been the subject of much research
[4±7]. In spite of their inferior physical strength when
compared with lyotropic liquid crystalline polyamides,
TLCPs are attracting a great deal of interest based on their
melt processability [8,9].
Although wholly aromatic TLCPs exhibit very attractive
mechanical properties, they generally have high melting
points, thus making them dif®cult to process [10,11]. Inclusion of ¯exible alkyl groups in otherwise wholly aromatic
polyesters not only lowers the melting point, but also
improves solubility and increases mixing entropy. Thus,
despite the predictable reduction in mechanical properties,
* Corresponding author. Tel.: 182-54-467-4292; fax: 182-42-483-6155.
E-mail address: (J.-H. Chang).

these polyesters possess considerable advantages in some
applications and show improved interfacial adhesion
between the two phases [5,12,13].
Nanocomposites possess unique properties, such as stiffness, strength and gas permeability, for their dispersion
structure [14±18]. The methods used for creating nanocomposites include in situ polymerization, solution intercalation, and melting intercalation [19,20]. Of them, melting
intercalation can be used with the most polymers, especially
thermoplastic materials, but it needs a polymer that has
good process properties in the melting state. In recent
years much attention has been paid to layered clay/polymer
nanocomposites, since these represent advanced plastic
materials prepared via the melting intercalation method.
In our previous paper [21], large improvements were
achieved in the thermal stabilities of TLCP nanocomposites
by using organo-montmorillonite. This enhancement of the
thermal stabilities explains reasonably well the dispersed

structure of clay in the nanocomposites caused by the
formation of the large aspect ratio of the clay particles.
For this paper, we synthesized TLCP with an alkoxy side
group base on a nematic liquid crystalline phase. We also
examined the correlation between the thermo-mechanical

0032-3861/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0032-386 1(02)00125-8


2970

J.-H. Chang et al. / Polymer 43 (2002) 2969±2974

earlier described by us [22], as well as by Lenz et al. [23].
The ethoxy side group and Br on the TLCP not only lowers
the melting point, but also improves some applications, as
mentioned in the previous section. The polymer formed was
thoroughly washed with methanol, with dilute HCl, and then
with water prior to drying at 60 8C in a vacuum oven.

Inherent viscosity of the TLCP was 0.64 dL/g which was
measured at 30 8C at a concentration of 0.2 g/dL solutions in
a phenol/1,1,2,2-tetrachloroethane ˆ 50/50 (v/v) mixture.
Fig. 1 shows the thread nematic textures for pure TLCP at
both 197 and 210 8C.
2.3. Preparation of C25A /TLCP nanocomposites

Fig. 1. Optical micrographs of TLCP taken at (a) 197 8C and (b) 210 8C
( £ 250).

properties and the clay content in TLCP nanocomposites
with variances in the dispersed morphology of the clay
particles. The general goal of this work was to use a
minimum amount of clay in the hybrids and still obtain
thermo-mechanical properties signi®cantly superior to
those of matrix polymer.
2. Experimental
2.1. Monomer synthesis
All reagents were purchased from Aldrich Chemical Co.
Commercially available solvents were puri®ed by distillation. The compound 2-ethoxyhydroquinone was synthesized via a multi-step route, and 2-bromoterephthalic acid
was purchased from Aldrich Chemicals.
2.2. Polymer preparation
The TLCP was prepared by direct polycondensation of
equivalent weights of the appropriate 2-ethoxyhydroquinone and 2-bromoterephthalic acid in the presence of
thionyl chloride and pyridine. The detailed procedure was

Cloisite 25A (organically modi®ed MMT; C25A) was
obtained from Southern Clay Product, Co. Since the
synthetic procedures for C25A/TLCP nanocomposites
with different weight percent (wt%) organoclay are very
similar, only a representative example for the preparation
of the C25A/TLCP (2 wt%) is given. 50 g of TLCP and 1 g
of C25A were dry-mixed and melt-blended at 190 8C,
within the nematic region of the polymer, for 30 min
using a mechanical mixer. For simplicity, the hybrids will
be referred to as 0% C25A/TLCP, 2% C25A/TLCP, 4%
C25A/TLCP, and so on, in which C25A and TLCP represent
the organoclay and polymer components used to prepare the
hybrids, respectively, and the number denotes the organoclay weight percent in the hybrid.

2.4. Extrusion
The TLCP hybrids were processed for ®ber spinning to
examine their tensile properties. The dried blends were
pressed at 160 8C, 2500 kg/cm 2 for a few minutes on a hot
press. The ®lm-type blends were dried in a vacuum oven for
24 h prior to being extruded through the die of a capillary
rheometer. From the capillary rheometer, the hot extrudates
were immediately drawn at constant take-up speed to form
extended extrudates having the same diameters. The cylinder
temperature of the extruder was 190 8C and the mean
residence time in the capillary rheometer was about 2±3 min.
To identify chemical reactions such as transesteri®cation
and thermal degradation at the processing temperature,
annealing was conducted for 4% C25A/TLCP hybrid at
190 8C. DSC thermograms of the heat-treated hybrids are
shown in Fig. 2. When heat treatment time increased from
10 to 60 min at 190 8C, there were no signi®cant changes in
the DSC scans. Chemical changes thus do not take place to
any appreciable extent at the extrusion processing temperature 190 8C. It was also con®rmed by 1H- and 13C-NMR
spectroscopy that no detectable transesteri®cation reaction
occurred in TLCP under the processing condition.


J.-H. Chang et al. / Polymer 43 (2002) 2969±2974

Fig. 2. DSC thermograms of 4% C25A in TLCP hybrid annealed at 190 8C
for different times.

2.5. Characterization
The thermal and the thermogravimetric analyses of

hybrids were carried out under N2 atmosphere on Du Pont
910 equipment. The samples were heated and cooled at a
rate of 20 8C /min. Wide-angle X-ray diffraction (XRD)
measurements were performed at room temperature on a
Rigaku (D/Max-IIIB) X-ray Diffractometer, using Ni®ltered Co-Ka radiation. The scanning rate was 28/min
over a range of 2u ˆ 2±308. Tensile properties of the extrudate were determined using an Instron Mechanical Tester
(Model 5564) at a crosshead speed of 2 mm/min. The specimens were prepared by cutting strips 5 by 70 mm long. An
average of at least eight individual determinations was
obtained. The experimental uncertainties in tensile strength
and modulus were ^1 MPa and ^0.05 GPa, respectively.
A polarizing microscope (Leitz, Ortholux) equipped with
a Mettler FP-5 hot stage was used to examine the liquid
crystalline behavior. The morphology of the fractured
surfaces of the extrusion samples was investigated using a
Hitachi S-2400 scanning electron microscope (SEM). The
fractured surfaces were sputter-coated with gold for
enhanced conductivity using an SPI Sputter Coater. TEM
photographs of ultrathin section polymer/organoclay hybrid
samples were taken on an EM 912 OMEGA (CARL ZEISS)
transmission electron microscope using an acceleration
voltage of 120 kV.

3. Results and discussion
3.1. Dispersibility of organoclay in TLCP
The XRD patterns of C25A, pure TLCP, and their TLCP

2971

Fig. 3. XRDs of C25A and C25A/TLCP hybrids.

hybrids with 2±6% C25A were represented in the region
from 2u ˆ 2±158 in Fig. 3. The interlayer spacing was
Ê ) for C25A. A peak
observed in 2u ˆ 5.648 (d ˆ 18.14 A
Ê ) for pure TLCP.
was observed in 2u ˆ 4.698 (d ˆ 21.98 A
When the amount of organoclay increased from 2 to 6%,
C25A/TLCP hybrids showed a same peak at the same position (2u ˆ 4.698). In the TLCP hybrids with 2±6% C25A,
no obvious clay peaks appeared in their X-ray diffraction
curves. This indicated that these clay layers were exfoliated
and dispersed homogeneously in the TLCP matrix. This was
also direct evidence that the C25A/TLCP hybrids formed
nanocomposites.
Unfortunately, XRD is unable to detect regular stacking
Ê . One may note that the commonly used
exceeding 88 A
de®nition of an exfoliated nanocompoisite is based on
layer spacing larger than this value. In reality, it was the
electron microscopic analyses that evidenced the formation
of nanocomposites.
Fractured surfaces of the ®lms were viewed under SEM.
A comparative analysis of the SEM photograph for TLCP
hybrids with different clay content exhibiting the ®brous and
platelet orientation distribution morphology including overall projection, as shown in Fig. 4.
More direct evidence of the formation of a true nanocomposite is provided by TEM of an ultramicrotomed section.
TEM micrographs of TLCP with different C25A content
from 2 to 6% are shown in Fig. 5(a)±(c), respectively.
The dark lines are the intersections of the clay layer of 1nm-thickness and the spaces between the dark lines are
interlayer spaces. This TEM photograph proves that most
clay layers of organoclay were exfoliated and dispersed

homogeneously into the TLCP matrix. This is consistent
with the observation of XRD studies shown in Fig. 3. In
conclusion, we were able to successfully synthesize TLCP
nanocomposites using C25A via a melting intercalation


2972

J.-H. Chang et al. / Polymer 43 (2002) 2969±2974

are slightly increased to 150 8C (see Table 1). This increase
in the thermal behavior of the hybrids may result from the
heat insulation effect of the clay layer structure, as well as
the strong interaction between the organoclay and TLCP
molecular chains. The isotropic transition temperatures
(Ti) of pure TLCP was virtually unchanged regardless of
organoclay loading, compared with the TLCP hybrids.
Fig. 7 shows the thread nematic textures for 2 and 6%
C25A/TLCP hybrids, respectively. Regardless of the clay
content in the hybrids, liquid crystallinity of the C25A/
TLCP hybrids was observed when organoclay content was
up to 6%.
In addition to having a higher melting point, thermal degradation properties of TLCP hybrids also show improvement.

Fig. 4. SEM photomicrographs of (a) 0% (pure TLCP), (b) 2%, (c) 4%, and
(d) 6% C25A in TLCP hybrids.

method. Considering the preceding results, the existing state
of clay particles could be determined to affect the thermal
behaviors and the tensile mechanical properties for each

organoclay/polymer hybrid.
3.2. Thermal behaviors
The thermal properties of TLCP hybrids with different
contents of C25A are listed in Table 1. The glass transition
temperatures (Tg) of TLCP hybrids linearly increased from
92 to 98 8C with clay loading from 0 to 4 wt% and leveled
off at the content range of more than 4 wt% of organoclay.
The increase in the Tg of these hybrids could be the result of
two factors. First, the effect of small amounts of dispersed
clay layers on the free volume of TLCP is signi®cant, and
does in¯uence the glass transition temperature of TLCP
hybrids. The second factor is ascribed to the con®nement
of the intercalated polymer chains within the clay galleries,
which prevents segmental motions of the polymer chains.
DSC traces of the pure TLCP and the hybrids are shown
in Fig. 6. The endothermic peak of the pure TLCP appears at
143 8C and corresponds with the melt transition temperature
(Tm). Maximum transition peaks of the TLCP hybrids
containing different clay contents in the DSC thermograms
Table 1
Thermal behavior values of C25A/TLCP hybrids
clay (wt%)

Tg (8C)

Tm (8C)

Ti (8C)

TD i a (8C)

wtR600 b (%)

0 (pure TLCP)
2
4
6

92
95
98
98

143
150
150
149

225
226
225
225

330
352
352
353

37
42

44
47

a
b

Initial weight reduction onset temperature.
Weight percent of residue at 600 8C.

Fig. 5. TEM photomicrographs of (a) 2%, (b) 4%, and (c) 6% C25A in
TLCP hybrids.


J.-H. Chang et al. / Polymer 43 (2002) 2969±2974

2973

Fig. 8. TGA thermograms of C25A and C25A/TLCP hybrids.

Fig. 6. DSC thermograms of C25A and C25A/TLCP hybrids.

A comparative thermal gravimetric analysis (TGA) of pure
TLCP and three nanocomposites with 2±6% C25A is shown
in Table 1 and Fig. 8. TGA curves do not show weight loss
below 100 8C, as shown in Fig. 8, indicating no water

remained in the samples. The weight loss due to the decomposition of TLCP and its hybrids was nearly the same until a
temperature of about 300 8C. After this point, the initial
thermal degradation temperature (TDi ) was in¯uenced by
organoclay loading in hybrids. Table 1 summarizes TDi of

the C25A/TLCP hybrids (at 2% weight loss) increased with
the amount of organoclay. TDi was observed at 352±353 8C
depending on the composition of the clay from 2 to 6 wt% in
the TLCP hybrids, with a maximum increase of 23 8C in the
case of the 6% C25A/TLCP as compared with that of the
pure TLCP. Weight of the residue at 600 8C increased with
clay loading from 0 to 6%, ranging from 37 to 47%. This
enhancement of the char formation is ascribed to the high
heat resistance exerted by the clay itself.
Considering the above results, it is consistently believable
that the introduction of inorganic components into organic
polymers can improve their thermal stability on the basis of
the fact that clays have good thermal stability [24,25].
3.3. Tensile properties
The pure TLCP and the TLCP hybrids were extruded
through a capillary die with draw ratio (DR) ˆ 1 to examine
the tensile strength and modulus of the extrudates. The DR
was calculated from the ratio of the diameter of the drawn
extrudate to that of the extruder die.
The tensile mechanical properties of pure TLCP and its
hybrid ®bers are listed in Table 2. The tensile strength and
initial modulus of C25A/TLCP hybrids increased with
corresponding increases in the amount of organoclay. The
Table 2
Tensile properties of C25A/TLCP hybrid ®bers

Fig. 7. Optical micrographs of (a) 2% and (b) 6% C25A in TLCP hybrids
taken at 200 8C ( £ 250).

Clay (wt%)

Ult. Str. (MPa)

Ini. Modu. (GPa)

E.B. a (%)

0
2
4
6

11.03
15.10
16.15
17.28

2.91
4.03
4.38
5.76

2
1
1
1

a

Elongation percent at break.

2974

J.-H. Chang et al. / Polymer 43 (2002) 2969±2974

ultimate tensile strength of TLCP hybrid ®bers increased as
the organoclay contents increased. When the C25A was
increased from 0 to 6% in hybrids, the strength linearly
improved from 11.03 to 17.28 MPa. The ultimate strength of
6% C25A/TLCP was 1.6 times higher than that of pure TLCP.
The same kind of behavior was observed for the initial
moduli. For example, the initial tensile modulus of 2%
C25A was 4.03 GPa, which was about 140% higher than
the modulus of pure TLCP. When the organoclay in TLCP
reaches 6%, the modulus increases about 2.0 fold
(5.76 GPa) over that of the pure TLCP.
This large increase in tensile property of hybrids owing to
the presence of organoclay can be explained as follows: the
amount of the increase of tensile property by clay layers
depends on the interactions between rigid, rod-shaped
TLCP molecules and layered organoclays, as well as on
the rigid nature of the clay layers. Moreover, the clay was
much more rigid than the TLCP molecules, and did not
deform or relax as the TLCP molecules did. This improvement was possible because organoclay layers could be
highly dispersed and exfoliated in the TLCP matrix. This
is consistent with the general observation that the introduction of organoclay into a matrix polymer increases its
strength and modulus [26,27].
The percent elongation at break of all samples, however,
decreases from 2 to 1% and then remains constant with clay

addition.
4. Conclusions
An aromatic thermotropic LCP with ethoxy side group
was synthesized and its optical texture was nematic. The
addition of 2±6% C25A to a TLCP maintains liquid crystallinity. C25A was exfoliated and dispersed homogeneously
in the matrix polymer. This was direct evidence that the
C25A/TLCP hybrids formed nanocomposites. This was
also cross-checked using XRD and TEM.
In general, thermal behaviors (Tg, Tm, and TDi ) of the
hybrids were enhanced with increasing clay content from
0 to 6 wt%. On the other hand, the isotropic transition
temperatures (Ti) of the hybrids were unchanged regardless
of organoclay loading.
Hybrids of different C25A contents were extruded with
DR ˆ 1 from a capillary rheometer to investigate the
mechanical properties of the hybrids. The ultimate strength
and initial modulus of the hybrids increased with increasing

C25A content. When the amount of organoclay in TLCP
reached 6 wt%, a 1.6-fold increase in the ultimate strength
and a 2.0-fold increase in the initial modulus were obtained,
as compared with the strength and modulus of the pure
polymer matrix. In this system, it was found that small
additions of organoclay were enough to improve the properties of the matrix polymer, TLCP.
Acknowledgements
This work was supported by Korea Research Foundation
Grant (KRF-2000-041-E00358).
References
[1] Kiss G. Polym Engng Sci 1987;27:410.
[2] Blizard KG, Baird DG. Polym Engng Sci 1987;27:653.

[3] Dutta D, Fruitwala H, Kohli A, Weiss RA. Polym Engng Sci
1990;30:1005.
[4] Joseph EG, Wilkes GL, Baird DG. Polymeric liquid crystals. New
York: Plenum Press, 1985.
[5] Heitz T, Rohrbach P, Hocker H. Macromol Chem 1989;190:3295.
[6] Jackson Jr. WJ, Kuhfuss HF. J Polym Sci A Polym Chem
1976;14:2043.
[7] Sukhadia AM, Done D, Baird DG. Polym Engng Sci 1980;30:519.
[8] Kenig S. Polym Engng Sci 1987;27:887.
[9] Lenz W. Faraday Disc Chem Soc 1985;79:21.
[10] Isayev AI, Modic M. Polym Compos 1987;8:158.
[11] Chung T. Plast Engng 1987;October:39.
[12] Chang J-H, Farris RJ. Polym J 1995;27:780.
[13] Chang J-H, Jo B-W. J Appl Polym Sci 1996;60:939.
[14] Giannelis EP. Adv Mater 1996;8:29.
[15] Lagaly G. Appl Clay Sci 1999;15:1.
[16] Usuki A, Koiwai A, Kojima Y, Kawasumi M, Okada A, Kurauchi T,
Kamigaito O. J Appl Polym Sci 1995;55:119.
[17] Yano K, Usuki A, Okada A, Kurauchi T, Kamigaito O. J Polym Sci A
Polym Chem 1993;31:2493.
[18] LeBaron PC, Wang Z, Pinnavaia TJ. Appl Clay Sci 1999;15:11.
[19] Yang F, Ou Y, Yu Z. J Appl Polym Sci 1998;69:355.
[20] Kato M, Usuki A. Polymer-clay nanocomposites, Wiley Series in
Polymer Science. New York: John Wiley & Sons, 2000. Chapter 5.
[21] Chang J-H, Park D-K. Polymer (Korea) 2000;24:399.
[22] Chang J-H, Jo B-W, Jin J-I. Korea Polym J 1994;2:140.
[23] Lenz RW, Frukawa A, Bhowmik P, Go RO, Majusz J. Polymer
1991;32:1703.
[24] Petrovic XS, Javni I, Waddong A, Banhegyi G. J Appl Polym Sci
2000;76:133.

[25] Frischer HR, Gielgens LH, Koster TPM. Acta Polym 1999;50:122.
[26] Wang Z, Pinnavaia TJ. Chem Mater 1998;10:3769.
[27] Li J-X, Wu J, Chan C-M. Polymer 2000;41:6935.


EUROPEAN
POLYMER
JOURNAL

European Polymer Journal 43 (2007) 374–379

www.elsevier.com/locate/europolj

Macromolecular Nanotechnology – Short communication

Critical aspects related to processing of carbon
nanotube/unsaturated thermoset polyester nanocomposites
A. Tug˘rul Seyhan a, Florian H. Gojny b, Metin Tanog˘lu

a,*

, Karl Schulte

b

_
_
Izmir
Institute of Technology (IZTECH), Mechanical Engineering Department, 35437 Izmir,
Turkey

Polymer Composites, Technische Universita¨t Hamburg-Harburg (TUHH), Denickestrasse 15, 21073 Hamburg, Germany
a

MACROMOLECULAR NANOTECHNOLOGY

b

Received 17 June 2006; received in revised form 16 August 2006; accepted 14 November 2006

Abstract
Carbon nanotubes (CNTs) have outstanding mechanical, thermal and electrical properties. As a result, particular interest has been recently given in exploiting these properties by incorporating carbon nanotubes into some form of matrix.
Although unsaturated polyesters with styrene have widespread use in the industrial applications, surprisingly there is
no study in the literature about CNT/thermoset polyester nanocomposite systems. In the present paper, we underline some
important issues and limitations during the processing of unsaturated polyester resins with different types of carbon nanotubes. In that manner, 3-roll mill and sonication techniques were comparatively evaluated to process nanocomposites
made of CNTs with and without amine (NH2) functional groups and polyesters. It was found that styrene evaporation
from the polyester resin system was a critical issue for nanocomposite processing. Rheological behaviour of the suspensions containing CNTs and tensile strengths of their resulting nanocomposites were characterized. CNT/polyester suspensions exhibited a shear thinning behaviour, while polyester resin blends act as a Newtonian fluid. It was also found that
nanotubes with amine functional groups have better tensile strength, as compared to those with untreated CNTs. Transmission electron microscopy (TEM) was also employed to reveal the degree of dispersion of CNTs in the matrix.
Ó 2006 Elsevier Ltd. All rights reserved.
Keywords: Carbon nanotubes; Thermosetting resin; Mechanical properties; Viscosity

1. Introduction
Scientific and industrial efforts have been recently
focused on nanotechnology and nanomaterials.
Nanomaterials are exhibiting some superior properties, as compared to their micro or macro size counterparts. Carbon nanotubes (CNTs) are composed
*
Corresponding author. Tel.: +90 232 750 7806; fax: +90 232
750 7890.
E-mail address: (M. Tanog˘lu).

of thin tubes with diameters of only a few nanometers, but a length of few microns. They exhibit higher

aspect ratio, extraordinary mechanical, thermal and
electrical properties, which make them prime candidates as reinforcing constituents in various polymers
for the production of nanocomposites. Although
there is a number of work published [1–5] on CNT
reinforced polymer composites, realization of the
expected enhancement in the properties of the composites, such as mechanical properties has not
entirely been established so far. This is because of

0014-3057/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.eurpolymj.2006.11.018


A.T. Seyhan et al. / European Polymer Journal 43 (2007) 374–379

no reported work in the literature on the processing
and properties of CNT/polyester systems. Thus,
CNTs have a great potential to improve the properties of a low cost resin like polyester at very low filler
content and to induce new characteristics such as
electrical conductivity. In this paper, we address
some critical aspects on the processing of CNT/
polyester nanocomposites prepared with the use of
3-roll-milling and also sonication techniques. Transmission electron microscopy (TEM) was employed
to reveal the degree of dispersion of carbon nanotubes with and without functional groups in the
involved resin. Some rheological and mechanical
properties of the composites are also discussed.
2. Experimental details
An isophtalic commercial unsaturated polyester
resin Cam Elyaf 266 with 35 wt.% of styrene was
obtained from CAM ELYAF Inc, Turkey. Also, special polyester resin blends, composed of an allylic
based polyester resin Poliya 240 with negligible

amount of styrene and Poliya 420 without any sty_
rene were obtained from POLIYA
POLYESTER
Corp., Turkey. Double-wall carbon nanotubes
(DWCNT) and multi-walled carbon nanotubes
(MWCNT) with and without amine functional group
(NH2) produced by chemical vapor deposition
(CVD) were obtained from Nanocyl (Namur/Belgium) and used as additives in the involved resin systems. DWCNTs and MWCNTs have average
diameters of 2.8 and 15 nm, respectively, with a
length of approximately 50 lm. Cobalt naphthanate
(CoNAP) and methyl ethyl ketone peroxide (MEKP)
were used as an accelerator and initiator, respectively, to polymerize the resin suspensions that contain various amounts of CNTs.
To prepare CNT/polyester nanocomposites, the
first approach was the utilization of the 3-roll-milling process, successfully employed to process epoxy
resins, employing to a commercial unsaturated
polyester resin. In this manner, the samples were
prepared under excessive shear forces for the dispersion of 0.1, 0.3, and 0.5 wt.% of carbon nanotubes
in Cam Elyaf 266 resin, setting the dwell time of
CNTs/polyester suspension on the rolls for about
2 min. The resin suspensions were polymerized with
the addition of 0.3 wt.% of CoNAP and 1 wt.% of
MEKP into the system. During the application of
this technique, we have experienced some difficulties. The major concern was the styrene evaporation
from the polyester resin during the processes, which

MACROMOLECULAR NANOTECHNOLOGY

the fact that nanotubes have strong tendency to
exist in agglomerated form via their huge surface
area, which leads to non-homogeneous dispersion

and random distribution of the nanotubes inside
the resin. Therefore, homogeneous dispersion of
CNTs in the polymer matrix is one of the key factors to enhance mechanical properties of the composites [1–9]. The common dispersion techniques
for processing CNT/polymer composites have been
direct mixing and sonication [1–15]. In addition,
Gojny et al. [6] showed that the utilization of 3roll-milling, which applies intensive shear forces
on the processed compounds, is an appropriate
technique to exfoliate and disperse carbon
nanotubes in an epoxy resin. They also concluded
that 3-roll-milling technique provided a better dispersion of CNTs in the epoxy resin resulting in
higher mechanical properties, as compared to those
prepared by sonication. Furthermore, besides the
physical approaches for the CNT dispersion, some
other attempts including the use of surfactants and
chemical functionalization of the CNT-surfaces
have been made in order to alter the degree of
dispersion and to tailor the interface between the
matrix and carbon nanotubes. In the near future,
the further development in chemical functionalization of nanotubes may be the key challenge for
advanced nanocomposites with the desired properties. Consequently, it is obvious that a better
understanding of the relationship between processing, interfacial optimization, surface chemistry and
composite properties is necessary for the potential
future applications of CNTs in polymer matrices.
Unsaturated polyesters (UP) with good cost/
property relation have been the most commonly
employed matrix materials for glass fiber reinforced
polymer composite parts. UP based materials have
been utilized in many applications including automotive, construction, transportation, storage tanks
and piping industry. Unsaturated polyesters become
insoluble and infusible by crosslinking with a monomer, which is usually styrene. The miscibility of the

resin and the styrene depends on the resin composition. Commercial polyester resins contain about
30–40% by mass of styrene. Polyester resins are versatile, quick curing, and have a long shelf life at
room temperature. The disadvantages of these thermoset resins are self-polymerization at higher temperatures and significantly higher cure shrinkage,
as compared to epoxy. Despite the fact that polyester resins have been commonly employed in many
industrial applications, to our knowledge, there is

375


MACROMOLECULAR NANOTECHNOLOGY

376

A.T. Seyhan et al. / European Polymer Journal 43 (2007) 374–379

caused a dramatic increase of the viscosity. Styrene
evaporation was accelerated due to heat occurred
on the rolling mills due to higher shear effect. The
polyester resin with high viscosity stacked on the
rolls and it caused some difficulties for the collection
of the resin, due to the uncontrolled styrene evaporation, and thus the final styrene compositions
within the resin blends were unknown. Alternatively, the sonication method was employed with
the same CNT/resin systems. Some problems with
the sonication method similar to 3-roll-milling process were observed. Even though the sonication
bath was cooled by water, the local heating due to
energy created within the resin system, caused styrene evaporation from the polymer suspension,
leading to a more viscous resin. In addition, it was
observed that nanotubes were agglomerated in the
volumes closer to the tip of the sonicator. Van der
Waals attractive forces between the CNT-surfaces

are known to be sensitive to heat, so increasing
agglomeration occurred [2,6]. To overcome the difficulties associated with styrene evaporation, we
switched to a resin system, containing negligible
amount of styrene (Poliya 240) and non styrene
(Poliya 420).
With the corresponding novel polyester resin systems nanocomposites were prepared by setting the
appropriate gelation time and viscosity for the
3-roll-milling processing. The styrene was added to
the system after 3-roll-milling. After some experimental trials, polyester resin blends were formulated
based on 45 wt.% of Poliya 420, 30 wt.% of Poliya
240 and 25 wt.% of styrene with the presence of
0.2 wt.% of CoNAP and 1.5 wt.% of MEKP. The
polymer mixture to be used during 3-roll-milling
process was prepared by hand-mixing of two types
of the polyester resin at the given ratio for 10 min.
Nanocomposite samples were prepared by the dispersion of the 0.1, 0.3, and 0.5 wt.% of the carbon
nanotubes within the polyester resin blend. After
collecting the CNT containing polyester suspension
by spatula from the 3-roll-milling, 25 wt.% of styrene was added to the involved resin system. The
whole system was then subjected to the intensive
mixing for half an hour using magnetic stirrer and
finally poured down into an aluminum mold and
cured at room temperature followed by post curing
in an oven at 110 °C for 2 h. Although Poliya 240
with a lower amount of styrene was introduced to
3-roll-milling, comparable viscosity increase of the
resin, and problems due to the stacking of the high
viscosity resin on the rolls were observed. However,

note that in this second approach, we diminished

difficulties with styrene evaporation and unknown
styrene content in the final product by using low styrene containing resin. For that reason, in our further experimental investigations, we focused just
on investigating the properties of the nanomaterials
prepared by the second approach.
The dispersion of the CNTs within the composites
was characterized by transmission electron microscopy (TEM) using a Philips EM 400 at 120 kV
acceleration voltages. The ultra thin TEM samples
with a thickness of 50 nm were prepared by ultramicrotome cutting at room temperature.
TA Instruments RDA III with parallel plate rheometer geometry (500 lm gap, and 50 mm plate
diameter) was used to analyze the rheological
behaviour of the polyester suspensions with different carbon nanotubes loadings. Tests were performed in steady state modes at room temperature
in order to avoid styrene evaporation during the
measurements. For that reason, liquid samples were
taken from the collected resin suspension from the
3-roll-mill. Steady shear rates (SSS) were used to
investigate the flow properties of the polyester suspensions by considering the viscosity as a function
of increasing shear rates.
Mechanical tensile properties of the composites
were determined according to DIN EN ISO 527.1.
Dog bone specimens were prepared by countersinking using a Mutronic Dear Drive 2000. The tensile
samples were tested using a Zwick Z010 Universal
tensile testing machine at a cross head speed of
1 mm/min. The elongation of the specimens during
the test was also measured.
3. Results and discussion
The 3-roll-milling process via intensive shear
forces seems to be more convenient technique than
traditional ones such as sonication and direct mixing for the dispersion of carbon nanotubes within
a liquid polymer resin. Fig. 1 shows the TEM micrographs of MWCNTs and DWCNTs with and without functional groups in the polyester resin blend
for 0.3 wt.% of loading. MWCNTs with functional

groups exhibited better local dispersion in the polyester matrix, as compared to DWCNTs with and
without treatment. In general, DWCNTs were
observed to be more agglomerated form caused by
their pronounced higher surface area. In the literature, rheological behaviour of the polymer suspension was associated with the prediction state of


377

Fig. 1. TEM micrographs of nanocomposites prepared from POLIYATM polyester at 0.3 wt.% loading of (a) MWCNT, (b) MWCNT–
NH2, (c) DWCNT (d) DWCNT–NH2.

2

10

Neat polyester resin
0.1 wt.% MWCNT-NH2
0.3 wt.% MWCNT-NH2
0.5 wt.% MWCNT-NH2

1

10

0

10

-1

10

Neat polyester resin
0.1wt.% MWCNT
0.3wt.% MWCNT
0.5wt.% MWCNT

1

10

Viscosity [Pa.s]

2

10

Viscosity [Pa.s]

CNTs dispersion within the corresponding resin
[11]. Figs. 2 and 3 give the viscosity as a function
of shear rate for the Poliya polyester based suspensions containing MWCNTs, MWCNT–NH2, respectively at different loading rates. As seen in
the figures, shear thinning behavior was observed
for the samples containing either MWCNT or

-2

10

-1

10

0

10

1

10

2

-1

10

3

10

Shear rate [s ]

0

10

Fig. 3. Viscosity of the polyester suspension with MWCNT–NH2
as a function of shear rate.

-1

10

-2

10

-1

10

0

10

1

2

10

10

3

10

-1

Shear rate [s ]
Fig. 2. Viscosity of the polyester suspension with MWCNTs as
a function of shear rate.

MWCNT–NH2, such that viscosity is reducing with
the increase of shear rates. The viscosity of polyester
suspensions with MWCNT decreases sharply at
0.1 wt%, but MWCNT–NH2 has not the same
behavior. This might be due to the fact that nanotubes with amine functional groups reveal better
compatibility or chemical interaction with the polyester chains within the system. Carbon nanotubes

MACROMOLECULAR NANOTECHNOLOGY

A.T. Seyhan et al. / European Polymer Journal 43 (2007) 374–379


378

A.T. Seyhan et al. / European Polymer Journal 43 (2007) 374–379

Ultimate tensile strength [MPa]

28
MWCNT
MWCNT-NH

2

26

24
N
E
A
T

22

20

R
E
S
I
N

0

0.1
0.3
0.5
CNT filler ratio [wt.%]

Fig. 4. Ultimate tensile strength (UTS) of the CNT/polyester
nanocomposites with MWCNT and MWCNT–NH2 as a function of CNT filler ratio.

30
Ultimate tensile strength [MPa]

MACROMOLECULAR NANOTECHNOLOGY

have a high aspect ratio, which alters significantly
the flow characteristics of involved polymer suspension.
In order to investigate the nanofiller effect with
and without chemical functional groups on the
mechanical properties of the composites, tensile
tests were conducted. The tensile properties of Poliya polyester blend were much lower, as compared
to a common commercial polyester resin in the market. Note that both of the components of Poliya
were specially snythesized and their individual
mechanical properties were lower than those of a
commercial polyester resin. The Figs. 4 and 5 show

DWCNT
DWCNT-NH

28

2

26
24

N
E
A
T

22

R

E
S
I
N

20
0

0.1
0.3
CNT filler ratio [wt.%]

0.5

Fig. 5. Ultimate tensile strength (UTS) of the CNT/polyester
nanocomposites with DWCNT and DWCNT–NH2 as a function
of filler ratio.

the tensile strength of the resulting nanomaterials.
As it can be seen in the figures, there are some differences between the MWCNT reinforced nanocomposites with and without amine functional
groups, as compared to neat polyester resin. Moreover, at each loading rate, composite specimens
containing MWCNTs with amine functional group
have higher values than those with MWCNTs without any functional group. For instance, the nanocomposites with 0.5 wt.% of MWCNT–NH2 have
about 6% and 15% higher strength than those with
the same loading rate of MWCNTs and the neat
resin, respectively. The same findings were also valid
for the composites with DWCNTs and DWCNT–
NH2. The nanocomposites with DWCNT–NH2 at
0.5 wt.% loading ratio have about 17% and 5% of
higher strength values than the neat resin and the

ones with DWCNTs. Note that nanocomposites
with DWCNTs with either functional group or
not have higher strengths than those with
MWCNTs or MWCNT–NH2 at each loading rate.
This can be explained by the higher surface area
of the double wall carbon nanotubes, which may
result in a better load transfer efficiency at the interface region as well as amine functional groups over
CNTs which is supposed to promote the dispersion
and pronounced covalent bonding to some extent.
4. Conclusion
In this paper, we have focused principally on
investigating three common key issues (i) availability of blending of polyester resin with very low
amount of carbon nanotubes, highlighting some
critical aspects and some limitations for the process,
(ii) dispersion state of carbon nanotubes within the
corresponding resin, (iii) interfacial adhesion/interactions of carbon nanotubes with the polyester resin
system. We have concluded that the styrene evaporation and self-polymerization via too much heat
occurred are the two major issues to be considered,
when a thermoset polyester resin is blended with
carbon nanotubes by employing 3-roll-milling and
sonication techniques. We also revealed that
3-roll-milling method is more adequate technique
for the dispersion of CNTs within a thermoset polyester resin blend, as compared to methods such as
sonication and direct mixing. Furthermore, the fact
that CNTs with amine functional groups exhibited
relatively enhanced dispersion state within the polyester resin blend, resulting in better tensile mechanical properties is evidence for that appropriate


chemical functionalization of carbon nanotubes
would be the key for the potential future applications. In the further studies, we are going to concentrate on developing CNT/polyester master batches

by means of different types of functional groups
and without styrene in order to obtain desired
microstructure and mechanical properties of the
nanocomposites.
Acknowledgement
Authors acknowledge the financial support from
_
TUBITAK-JULICH
5 Project.
References
[1] Rosen R, Zheng B, Liu J. Adv Mater 2002;19:14.
[2] Thostenson ET, Li C, Chou TW. Compos Sci Technol
2005;65:491.

379

[3] Gojny FH, Wichmann MHG, Fiedler B, Schulte K. Compos
Sci Technol 2005;65:2300.
[4] Gojny FH, Nastalczyk J, Roszlanic Z, Schulte K. Chem
Phys Lett 2003;370:820.
[5] Gojny FH, Schulte K. Compos Sci Technol 2004;64:2303.
[6] Gojny FH, Wichmann MHG, Ko¨pke U, Fiedler B, Schulte
K. Compos Sci Technol 2005;64:2363.
[7] Yasmin A, Abot JL, Daniel IM. Scripta Mater 2003;49:81.
[8] Ai A, Bai S, Cheng HM, Bai JB. Compos Sci Technol
2002;62:1993.
[9] Andrews R, Jacques D, Minot M, Rantell T. Macromol
Mater Eng 2002;287:395.
[10] Weg UD, Benoit JM, Lebedkin S. Curr Appl Phys 2002;21:
497.

[11] Po¨tschke P, Fornes TD, Paul DR. Polymer 2004;45:8863.
[12] Aizawa M, Shaffer Milo SP. Chem Phys Lett 2003;368:121.
[13] Kymakis E, Alexandou I, Amaratunga GAJ. Synthetic Met
2002;127:59.
[14] Sandler JKW, Kirk JE. Polymer 2003;44:5893.
[15] Lau Kin-Tak, Hui David. Carbon 2002;40:1597.

MACROMOLECULAR NANOTECHNOLOGY

A.T. Seyhan et al. / European Polymer Journal 43 (2007) 374–379