Improvement of Properties of Polyetherimide/Liquid Crystalline ...

Copyright © 2009 American Scientific Publishers All rights reserved Printed in the United States of America

Journal of Nanoscience and Nanotechnology Vol. 9, 1928–1934, 2009

Improvement of Properties of Polyetherimide/Liquid Crystalline Polymer Blends in the Presence of Functionalized Carbon Nanotubes Sunanda Roy1 , Nanda Gopal Sahoo2 , Madhumita Mukherjee1 , Chapal Kumar Das1 ∗ , Siew Hwa Chan2 , and Lin Li2 ∗ 1

RESEARCH ARTICLE

2

Materials Science Centre, Indian Institute of Technology, Kharagpur 721302, India School of Mechanical and Aerospace Engineering, Nanyang Technological University, Singapore 639798

This article reports the effect of functionalized multiwall carbon nanotubes (MWNT-COOH) on Delivered by Ingenta to: the morphological, dynamic mechanical, mechanical and thermal properties of polyetherimide Nanyang Technological University (PEI)/liquid crystalline polymer (LCP) (Vectra A950) blends. The chemical modification of carbon : 155.69.4.4 nanotube enhanced the compatibility andIPthe miscibility between PEI and LCP in the composites. 2012 14:55:46 Addition of functionalized MWNTsTue, into 24 the Jul blend improved the thermal, mechanical and dynamic mechanical properties of the composite due to the presence of strong interfacial interaction between the polymer matrixes and the nanotubes in polymer composites. The glass transition temperature (from tan ) increased from 208 C to 245 C with the addition of 1.8 wt% functionalized MWNTs in the blend of PEI/LCP. The tensile strength of the composite with 1.8 wt% MWNT-COOH was enhanced by 61% and 44% as compared to PEI/LCP blend and pure PEI. The functional groups on the MWNTs surface played an important role in accelerating both the dispersion of MWNTs and the interfacial adhesion in the composites compared to raw MWNTs.

Keywords: Carbon Nanotube, Miscibility, Interfacial Adhesion, Liquid Crystalline Polymer, Mechanical Property.

1. INTRODUCTION Since the discovery of carbon nanotubes (CNTs) in 1991 by Iijima,1 they are considered to be the most challenging candidates as ideal reinforcing fillers in high strength, light weight polymer nanocomposites due to their low density, high aspect ratio and exceptional mechanical properties.2–5 The impressive and unique properties of carbon nanotubes make them promising materials for a wide variety of applications,1 such as nanotechnology, electronics, optics, space sectors, hightech automotive industries and other fields of materials science and engineering. Improvement of mechanical reinforcements of polymer composites by carbon nanotubes is directed by two main factors: (a) the homogeneous dispersion of carbon nanotubes and (b) the interfacial adhesion between the carbon nanotubes and the matrix polymer. However, carbon nanotubes are strongly affected by Van de Waal’s forces due to their small size and high surface area. These forces give rise to the formation of aggregates, which in turn, inhibits the fine dispersion of the filler in the polymer ∗

Authors to whom correspondence should be addressed.

1928

J. Nanosci. Nanotechnol. 2009, Vol. 9, No. 3

matrix. Chemical functionalization appears to be as an effective way to avoid this aforementioned problem leading to a homogeneous dispersion of carbon nanotubes in polymer matrices. Currently various techniques are used to incorporate CNTs into a polymer matrix, e.g., solution casting, melt mixing, electron spinning, and in situ polymerization.6–13 Melt mixing is a common and simple method, which is particularly useful for thermoplastic polymers. In melt processing, carbon nanotubes are mechanically dispersed into a polymer matrix using a high temperature and high shear force mixer or compounder.14 This approach is simple and compatible with current industrial practices. Various polymer matrices are used for composites such as thermoplastics,15 16 thermosetting resin,17 18 liquid crystalline polymers,19 20 water-soluble polymers,21 conjugated polymers,22 and so on. In this research, we have used polyetherimide (PEI), liquid crystalline polymer (LCP) (Vectra A950) and multiwall carbon nanotubes (MWCNT) to prepare high performance polymer composites using conventional mixing methods such as an internal mixer. Generally most of the thermoplastics are immiscible with LCP at a molecular level.23 24 The immiscibility between a 1533-4880/2009/9/1928/007

doi:10.1166/jnn.2009.403

Roy et al.

Improvement of Properties of PEI/LCP Blends in the Presence of Functionalized Carbon Nanotubes

polymer matrix and the LCP makes the composite thermodynamically less stable due to the weaker interfacial adhesion. LCPs are able to reduce the viscosity of a composite due to their inherent low viscosity feature. The chemically modified carbon nanotubes behave not only as a reinforcing filler but may also act as a bonding agent at the interface between LCP and the matrix polymer (PEI). Therefore, the objective of the present study is to develop PEI/LCP/MWCNT composites with enhanced thermal and mechanical properties. The MWCNT has been chemically modified for better dispersion and a potential as a compatibilizer for PET and LCP. A comprehensive study of the thermal, mechanical, dynamic, rheological and morphological properties of the composites has been carried out. The effect of nanotube content on the properties of the composites has also been examined.

Table I. Compounding formulations for PEI/LCP/MWCNT composites prepared in this study, where amounts of PEI, LCP, and MWNTs are weight portions. Sample code

PEI

LCP

MWNTCOOH

Raw MWNT

Wt% of MWNT

Tensile strength (MPa)

PEI PEI/LCP PLMC0.5 PLRC1 PLMC1 PLMC2 PLMC3

100 100 100 100 100 100 100

— 60 60 60 60 60 60

— — 0.5 — 1 2 3

— — — 1 — — —

0 0 0.3 0.6 0.6 1.2 1.8

930 780 — 1021 1125 1172 1341

J. Nanosci. Nanotechnol. 9, 1928–1934, 2009

1929

RESEARCH ARTICLE

was observed by scanning electron microscopy (SEM), after gold coating. The analysis was done using a JEOL JSM-5800 SEM at an accelerating voltage of 20 KV. Tensile test measurements were carried out with the compression molded Dumb-bell shaped samples in a Uni2. EXPERIMENTAL DETAILS versal Testing Machine i.e., a Hounsfield HS 10 KS (uniDelivered by Ingenta to: The MWNTs used in this study were purchased from Iljin versal testing machine), at room temperature with an Nanyang Technological University Nano Technology, Korea. The diameter, length and aspect extension speed of 5 mm/min and an initial gauge length IP : 155.69.4.4 ratio were 10–20 nm, 20 m and 1000–2000, respectively. of 14:55:46 35 mm. For each composite four measurements were Tue, 24 Jul 2012 MWNT-COOH was prepared by oxidation of raw MWNTs made within an experimental error of ±2%. with H2 SO4 /HNO3 (3:1) at 90 C for 10 min with vigorDynamic mechanical analysis of composites were conous stirring as described where else previously.7 The amorducted using a TA Instruments DMA 2980 model under phous matrix polymer, polyetherimide, used in this study, single cantilever clamp arrangement in a bending mode at was in pellet form (Grade-Ultem 1000, GE, USA), and a frequency of 1 Hz from 25 to 300 C at a heating rate it is a transparent and amber high performance thermoof 10 C/min. The storage modulus (E ) and loss tangent plastic polymer. Its density is about 1.27 g/cm3 at 25 C. (tan ) were measured for each sample in the temperature Thermotropic liquid crystalline polymer (TLCP, or simply range. LCP) was Vectra A950 from Ticona, USA. The TLCP has Thermal stability of the composites was examined under the comonomer composition of 75 mole% of hydroxybendry air using a TGA V 50 IA Dupont 2100 thermo gravizoic acid (HBA) and 25 mole% of hydroxynaphthoic acid metric analyzer from 25 to 700 C at a heating rate of (HNA). 10 C/min. Electrical conductivity measurement was car100 weight fractions of polyetherimide was mixed with ried out using a four-point test fixture combined with a 60 weight fractions of LCP and 0.5, 1.0, 2.0, and 3 weight Keithley electrometer Model 230. fractions of MWNTs respectively using a Sigma-high temperature internal mixer (S. C. Dey Pvt. Ltd, Kolkata, India) at 325 C for 15 min. After taken out from the mixing 3. RESULTS AND DISCUSSION machine the blends were compression molded at 325 C The FTIR and Raman spectra for the functionalized under 10 MPa pressure for 10 min. Then the blends were MWNTs and raw MWNTs are shown in Figures 1 and 2. allowed to cool to room temperature under the same pres The FT-IR spectra of the modified MWNTs showed the sure at the rate of about 2 C/min. The compounding forpeaks at 3458, 1640, and 1185 cm−1 , corresponding to mulations of the blends are tabulated in Table I. OH, C O, and C C O stretching, respectively. These Raman spectroscopy (BRUKER, RFS, 100/s) was used characteristic bands were not observed in the case of to investigate the structural changes of MWNTs by the raw MWNTs. This was attributed to the carboxylic acid acid treatment. A 632.8 nm He-Ne laser was used as the groups generated at the surface of the MWNTs after light source. Transmission electron microscopic (TEM) acid treatment in H2 SO4 /HNO3 . From Figure 2 it was analysis was performed on a JEM-2010F (JEOL Co.) elecclearly evident that the two bands at around 1 579 and tron microscope. 1 325 cm−1 in the spectra were assigned to the tanThe melt rheological measurements were performed on gential mode (G-band) and the disorder mode (D-band), a Rheometric Scientific AR1000 with the compression respectively.7 25 The D-band intensity was increased in the molded samples. The experiments were carried out in an modified MWNTs compared to raw MWNTs. The peak oscillatory shear mode using parallel plate geometry with intensity ratio (ID/IG = 133) at D-band and G-band for a diameter of 40 mm at 325 C under nitrogen atmosphere. The surface morphology of the tensely fractured samples the modified MWNTs exceeded those of raw MWNTs


Fig. 1. FTIR spectra of raw (lower curve) and functionalized (upper one) MWNTs.

Roy et al.

Fig. 3. TEM image of MWNT-COOH.

viscosity of elongated LCP particles which enables easy flow and leads to melt lubrication. Thus, LCP acts as a (ID/IG = 106). This result indicates that some of the sp by Ingenta to: Delivered processing aid for PEI. This result is in good agreement 3 carbon atoms (C C) were converted toNanyang sp carbonTechnological atoms University with the literature.26 It is apparent from Figure 4 that the treat(C C) at the surface of the MWNTs after the acidIP : 155.69.4.4 modified carbon nanotubes had a dramatic effect on the ment in H2 SO4 /HNO3 . This may also be evidenced by Jul the 2012 Tue, 24 14:55:46 viscosity of PEI/LCP blend. The viscosity of the comTEM observation as shown as Figure 3. On the surface of posites dramatically increased with increasing MWNTthe modified CNT, there are some defects which could be COOH loading compared with that of the PEI/LCP blends. caused by the carbon–carbon bonding associated with the At a higher loading of MWNT-COOH, the interaction formation of carboxylic acid groups on the surface. between the MWNTs and the polymers became higher, Figure 4 shows the variation in viscosity with shear rate which restricted the melt flow of the composite. for PEI, PEI/LCP and PEI/LCP/CNT nanocomposites. In The SEM photographs for the tensely fractured surafces the shear rate range studied, the PEI and composites exhibof the PEI/LCP blend and the MWNTs nanocomposites ited the non-Newtonian behavior, which became more are shown in Figure 5. The LCP phase was clearly seen to prominent in the case of PEI/LCP/MWNT-COOH composexist in a globular morphology rather than microfibrills in ites with higher loading of MWNT-COOH. From Figure 4, the PEI matrix as shown in Figure 4(a). These large globit was clearly observed that the viscosity of the PEI/LCP ules and voids thorough out the matrix were formed due to blend was much lower than that of pure PEI. The reduced the lack of interfacial adhesion between the PEI matrix and viscosity of PEI by LCP is considered to be caused by the LCP phase. The deficient interfacial adhesion is probthe interphase slip between the PEI and LCP phases. This ably an indication of the immiscibility between the blend interlayer slip is mainly attributed to two factors: (1) the immiscibility between the two phases and (2) the low D-band MWNT-COOH

G-band

Intensity (arb. unit)

RESEARCH ARTICLE

2

Raw MWNT 600

800

1000

1200

1400

Raman shift Fig. 2.

1930

1600

1800

(cm–1)

Raman spectra of raw and functionalized MWNTs.

2000

Fig. 4. Log–log plots of viscosity versus shear rate for PEI, LCP, PEI/LCP and PEI/LCP/MWNT-COOH composites from bottom to top: (a) LCP, (b) PEI/LCP, (c) PEI, (d) PEI/LCP/0.6 wt% MWNT-COOH, and (e) PEI/LCP/1.2 wt% MWNT-COOH.


Roy et al.

Improvement of Properties of PEI/LCP Blends in the Presence of Functionalized Carbon Nanotubes (a)

(b)

(c)

(d)

Delivered by Ingenta to: Nanyang Technological University IP : 155.69.4.4 Tue, 24 Jul 2012 14:55:46

components, which is in good agreement with the results found for other PEI blends with different LCP’s.26 27 As can be seen in Figure 5(b), the globules and voids size were reduced by the addition of 0.6 wt% MWNT-COOH into the PEI/LCP blend. This indicated a better miscibility between PEI and LCP in the presence of MWNTCOOH. At higher loading of MWNT-COOH (Fig. 5(c)), the microfibril formation was observed instead of the globular morphology, which demonstrated an enhanced miscibility between PEI and LCP to result in an improved dispersion of LCP in the matrix in the presence of modified MWNT. Figure 4(d) manifested a different morphology. There was no void formation to be observed and the rather finer dispersion of LCP microfribilles indicates a better miscibility and interfacial adhesion between PEI and LCP in the presence of MWNT-COOH even at the higher loading, which is expected to lead to the superior mechanical, dynamic and thermal properties of the concerned composites. From these results, we can conclude that the functionalized MWNTs could act as a compatilizer for PEI/LCP blends. J. Nanosci. Nanotechnol. 9, 1928–1934, 2009

The tensile strength for pure PEI, PEI/LCP and PEI/LCP/MWNT composites are illustrated in Figure 6 and Table I. As can be seen, the tensile strength of the PEI/LCP blend was lower than that of pure PEI. This is due to the lack of interfacial adhesion between the matrix and the LCP phase, which may cause pulling out of the LCP domain from the matrix. The addition of MWNTs into the PEI/LCP blend improved the tensile strength of the blend. The tensile strength of the composite with 0.6 wt% MWNT-COOH was enhanced by 44% and 21% as compared to the PEI/LCP and pure PEI, respectively, while an increase of 31% and 9.6% was achieved by incorporating the same amount of raw MWNTs into the PEI/LCP blend. Finally, the tensile strength of the nanocomposites increased from 78 MPa in the PEI/LCP to 134.1 MPa when the content of the functionalized MWNTs reached 1.8 wt% in the composites. Because the functionalized MWNTs prepared by the acid treatment contain COOH groups, which could be helpful for improving the interaction among oxygen (O) groups both in PEI and LCP chains. The possible interaction between PEI, LCP and MWNT-COOH is schematically proposed in Figure 7. This mechanism 1931

RESEARCH ARTICLE

Fig. 5. SEM images for the tensely fractured surfaces of (a) PEI/LCP blend and (b) to (d) PEI/LCP/0.6, 1.2 and 1.8 wt% of MWNT-COOH respectively.


Roy et al.

160

Tensile strength (MPa)

140 120 100 80 60 40 20 0 PEI

RESEARCH ARTICLE

Fig. 6.

PEI/LCP PLRC1 PLMC1 PLMC2 PLMC3

Tensile strength of various samples as listed in Table I.

Fig. 8. Tan as a function of temperature at 1 Hz for pure PEI, PEI/LCP and PEI/LCP/MWNTs composites (a) PEI, (b) PEI/LCP, (c) PLRC1, (d) PLMC1 and (e) PLMC3.

explains how the modified MWNTs acted as a bridge across the interface between the PEI and LCP to increase the interfacial adhesion between PEI and LCP and thus enhance by Ingenta to: Delivered side with the addition of MWNTs. These results suggest the dispersion of LCP in the PEI matrix. As a result, the Nanyang Technological University the inhibition of the molecular motion of the PEI matrix overall mechanical performance of the composites could IP :be155.69.4.4 due to the constraint effect of MWNTs.30 The glass tranimproved. Tue, 24 Jul 2012 14:55:46 sition temperature increased from 218 C for the neat PEI The loss tangent (tan ) and the storage modulus verto 245 C for the PEI/LCP/MWNT-COOH composite with sus temperature curves are represented in Figures 8 and 9. 1.8 wt% MWNT-COOH. It is well known that the temperature corresponding to From Figure 9, it is clearly seen that the storage modthe maxima of a tan peak is normally associated with ulus sharply decreased in the temperature range from the glass transition temperature Tg . From the variation of C to 245 C, which was also correlated with the 208 tan with temperature (Fig. 8), Tg of pure PEI appeared glass transition. The storage modulus of the nanocomposat approximately 218 C, which slightly shifted to the ites dramatically increased when MWNT-COOH or raw lower temperature side with the addition of LCP to the MWNTs were incorporated into the PEI/LCP blend. The PEI/LCP blend. This is due to a more active movement 28 effect of MWNT-COOH on the increase in the storage of LCP chains in the blend. The magnitude of the loss modulus was more dominant than the unmodified MWNT. tangent delta peak decreased with an addition of LCP in The storage modulus for the nanocomposite containing the blend (the curve b in Fig. 8). The loss tangent peaks 1.8 wt% MWNT-COOH showed an increase by 68% and were indicative of the capacity of a material in dissipat107% at room temperature (25 C) as compared to the pure ing mechanical energy. The lower magnitude of loss tanPEI and the PEI + LCP blend, respectively. The enhanced gent delta peak indicated a lower capacity of mechanical energy dissipation.29 The curves c, d and e in Figure 8 revealed that Tg of PEI shifted to the higher temperature

Fig. 7. Possible formation of hydrogen bonding among the PEI, LCP chain and MWNT-COOH.

1932

Fig. 9. Storage modulus as a function of temperature at of pure PEI, PEI/LCP and PEI/LCP/MWNTs composites: (a) PEI/LCP, (b) PEI, (c) PLRC1, (d) PLMC1, (e) PLMC2 and (f) PLMC3.


Roy et al.

Improvement of Properties of PEI/LCP Blends in the Presence of Functionalized Carbon Nanotubes 4e – 4

Conductivity (S/cm)

3e – 4

2e – 4

1e – 4

0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

MWNTs (wt%) Fig. 10. TGA thermograms for (a) PEI/LCP, (b) PEI, (c) PLRC1, (d) PLMC1, (e) PLMC2 and (f) PLMC3.

Fig. 11. Dependence of electrical conductivity on the MWNT content for PEI/LCP/MWNT-COOH nanocomposites.

PLMC1 PLMC2 PLMC3

510 522 544


607 618 629

Acknowledgments: This work was supported by IIT Kharagpur and the A∗ STAR SERC Grant (0721010018). 1933

RESEARCH ARTICLE

59 C as compared to the PEI/LCP blend. It is imporstorage modulus and glass transition temperature are tant to point Delivered to: out that the extent of interactions between considered to be due to the effect of the fine dispers- by Ingenta the functionalized Nanyang Technological UniversityMWNTs and the PEI/LCP matrix could ing ability of the functionalized carbon nanotubes into be responsible for the higher thermal stability of the IP : 155.69.4.4 the PEI/LCP blend and the enhanced interaction between composites. 24 Jul 14:55:46 PEI and LCP by the functionalized carbon Tue, nanotubes as 2012 Carbon nanotubes exhibit the high aspect ratio and high schematically illustrated in Figure 7. conductivity, which makes CNTs excellent candidates for In order to investigate the thermal stability of the fabrication of conducting composites. The electrical conPEI/LCP/MWNT composites, TGA measurements were ductivity of the PEI/LCP/MWNT-COOH nanocomposite carried out, and the results are shown in Figure 10 and as a function of MWNT-COOH is shown in Figure 11. The Table II. From Figure 10, it is clearly observed that the conductivity of the composites increased with increasing degradation occurred in a single step pattern for pure PEI, MWNT-COOH content. A sharp increase of the conducwhere as a two step-degradation occurred in all the comtivity was observed between 0.6 to 1.2 wt% MWNTposites. For the composite, the first step of degradation COOH in the composites, suggesting that the percolation corresponded to the degradation of PEI and the second threshold for the conductivity could be around 1.2 wt%. step corresponded to the degradation of LCP. The onsets The percolation threshold for the electrical conductivity in of degradation of pure PEI and PEI/LCP blends are 493 C polymer-CNT nanocomposites depends on dispersion,31 32 and 485 C respectively. It was observed that the incorpoalignment33 and aspect ratio32 34 of carbon nanotubes.32 For our case, it was easier for the modified MWNTs to ration of MWNT significantly increased the thermal stabe well dispersed in the polymer matrix to form electribility of PEI/LCP/CNT composites. For example, from cal pathways due to their improved miscibility with the the onset degradation and 50% degradation temperatures polymer matrix. as shown in Table II, the composite (PLRC1) containing 0.6 wt% MWNT-COOH showed the more delayed degradation as compared to pure PEI and the PEI/LCP with 4. CONCLUSIONS the same loading of raw MWNT. In the case of the comCarboxylic functionalized MWNTs have been incorpoposite with 1.8 wt% MWNT-COOH, the onset degradarated into PEI/LCP blends to demonstrate their effects tion temperature shifted to the higher temperature side by on the rheological, thermal, morphological, mechanical and electrical properties of MWNT-reinforced composites. Table II. Thermal degradation temperatures for PEI/LCP/MWCNT SEM observation showed that the functionalized MWNTs composites. improved the miscibility and adhesion between PEI and Sample Onset 50% degradation LCP. The strong interaction among the PEI, LCP and the code temperature ( C) temperature ( C) functionalized MWNTs greatly enhanced the dispersion of LCP and MWNTs in the polymer matrix. As a result, the PEI 493 565 PEI/LCP 485 544 overall mechanical, thermal and electrical properties of the PLRC1 498 597 composites could be significantly improved.


Roy et al.

References and Notes 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

RESEARCH ARTICLE

16.

17. O. Gryshchuk, J. Karger-Kocsis, R. Thomann, Z. Konya, and I. Kiricsi, Compos. Part A. 37, 1252 (2006). S. Iijima, Nature 354, 56 (1991). 18. S. Ghose, K. A. Watson, D. M. Delozier, D. C. Working, E. J. Siochi, M. M. Treacy, T. W. Ebessen, and J. M. Gibson, Nature 381, 678 and J. W. Connell, Compos. Part A 37, 465 (2006). (1996). 19. R. A. Mrozek, B. S. Kim, V. C. Holmberg, and T. A. Taton, Nano R. Saito, G. Dresselhaus, and M. S. Dresselhaus, Physical Properties Lett. 3, 1665 (2003). of Carbon Nanotubes, Imperical College Press, London (1998). 20. V. N. Bliznyuk, S. Singamaneni, R. L. Sanford, D. Chiappetta, H. G. Chae, J. Liu, and S. Kumar, Carbon Nanotubes Properties and B. Crooker, and P. V. Shibaev, Polymer 47, 3915 (2006). Applications, edited by M. J. O’Connell, Taylor & Francis Group, 21. B. Zhao, H. Hu, and R. C. Haddon, Adv. Funct. Mater. 14, 71 (2004). LLC, Boca Raton (2006). 22. H. Ago, K. Petritsch, M. S. P. Shaffer, A. H. Windle, and R. H. X. L. Xie, Y. W. Mai, and X. P. Zhou, Mater. Sci. Eng. R 49, 89 Friend, Adv. Mater. 11, 1281 (1999). (2005). 23. L. Incarnato, M. R. Nobile, M. Frigione, O. Motta, and D. Acierno, B. Safadi, R. Andrews, and E. A. Grulke, J. Appl. Polym. Sci. 84, Intern. Polym. Process. VIII, 191 (1993). 2660 (2002). 24. M. R. Nobile, D. Acierno, L. Incarnato, E. Amendola, L. Nicolais, N. G. Sahoo, Y. C. Jung, H. J. Yoo, and J. W. Cho, Macromol. Chem. and C. Carfagna, J. Appl. Polym. Sci. 41, 2723 (1990). Phys. 207, 1773 (2006). 25. H. C. Kuan, C. H. M. Ma, W. P. Chang, S. M. Yuen, H. H. Wu, and M. A. Lopez Manchado, L. Valenti, J. Biagiotti, and J. M. Kenny, T. M. Lee, Compos. Sci. Technol. 65, 1703 (2005). Carbon 43, 1499 (2005). 26. S. Bastida, J. I. Eguiazabal, and J. Nazabal, J. Mat. Sci. 35, 153 W. D. Zhang, L. Shen, I. Y. Phang, and T. Liu, Macromolecules 37, (2000). 256 (2004). 27. S. Bastida, J. I. Eguiazabal, and J. Nazabal, Polym. Comp. 17, 919 R. Sen, B. Zhao, D. Perea, M. E. Itkis, H. Hu, J. Love, E. Bekyarova, (1996). and R. C. Haddon, Nano Lett. 4, 459 (2004). 28. S. Lee, S. M. Hong, Y. Seo, T. S. Park, S. S. Hwang, K. U. Kim, H. Q. Hou, J. J. Ge, J. Zeng, Q. Li, D. H. Reneker, A. Greiner, and by Ingenta Delivered to:Lee, Polymer 35, 519 (1994). and J. W. S. Z. D. Cheng, Chem. Mater. 17, 967 (2005). 29. A. University I. Isayev and P. R. Subramanian, Polym. Eng. Sci. 32, 85 Nanyang Technological H. H. So, J. W. Cho, and N. G. Sahoo, Eur. Polym. J. 43, 3750 (1992). IP : 155.69.4.4 (2007). 30. T. Liu, Y. Tong, and W. D. Zhang, Compos. Sci. Technol. 67, 406 Tue, 24 Jul 2012 14:55:46 H. J. Yoo, Y. C. Jung, N. G. Sahoo, and J. W. Cho, J. Macromol. (2007). Sci. Part B: Phys. 45, 441 (2006). 31. J. K. W. Sandler, J. E. Kirk, I. A. Kinloch, M. S. P. Shaffer, and R. Andrews, D. Jacques, M. Minot, and T. Rantell, Micromol. Mater. A. H. Windle, Polymer 44, 5893 (2003). Eng. 287, 395 (2002). 32. J. Li, P. C. Ma, W. S. Chow, C. K. To, B. Z. Tang, and J. K. Kim, L. Y. Li, C. Y. Li, C. Y. Ni, L. X. Rong, and B. Hsiao, Polymer 48, Adv. Funct. Mater. 17, 3207 (2007). 33. F. Du, J. E. Fisher, and K. I. Winey, J. Polym. Sci. Part B: Polym. 3452 (2007). Phys. 41, 3333 (2003). O. Morales-Teyssier, S. Sanchez-Valdes, and L. F. Ramos-de Valle, 34. J. B. Bai and A. Allaoui, Composites, Part A 34, 689 (2003). Macromol. Mater. Eng. 291, 1547 (2006).

Received: 8 March 2008. Revised/Accepted: 4 April 2008.

1934
