- Email: [email protected]
multiwalled carbon nanotube nanocomposites
Chemical Physics Letters 523 (2012) 87–91
Contents lists available at SciVerse ScienceDirect
Chemical Physics Letters journal homepage: www.elsevier.com/locate/cplett
Laser heating effect on Raman spectra of styrene–butadiene rubber/multiwalled carbon nanotube nanocomposites Xinlei Yan a, Yasutaka Kitahama a, Harumi Sato a, Toshiaki Suzuki a, Xiaoxia Han a, Tamitake Itoh b, Liliane Bokobza c, Yukihiro Ozaki a,⇑ a
Department of Chemistry, School of Science and Technology and Research Center of Environmental Friendly Polymer, Kwansei-Gakuin University, Gakuen 2-1, Sanda 669-1337, Japan Nano-Bioanalysis Group, Health Research Center, National Institute of Advanced Industrial Science and Technology (AIST), Takamatsu, Kagawa 761-0395, Japan Université Pierre et Marie Curie-Ecole Supérieure de Physique et Chimie Insdustrielles – Centre National de la Recherche Scientifique, UMR 7615, 10 rue Vauquelin, 75231 Paris Cedex 05, France
b c
a r t i c l e
i n f o
Article history: Received 26 September 2011 In final form 29 November 2011 Available online 19 December 2011
a b s t r a c t The laser heating effect on MWCNTs in styrene–butadiene rubber/multiwalled carbon nanotube (SBR/ MWCNT) composites were studied by Raman spectra. The intensity ratio of the D band to G band (ID/IG) of SBR/MWCNT composites largely decreased with temperature. This indicates the self-rearranging behavior of MWCNTs in the SBR/MWCNTs system during temperature increase. In addition, the temperaturedependent downward shift of the G band of SBR/MWCNT composites was smaller than that of MWCNTs samples. The self-rearrangement of MWCNTs in SBR/MWCNT composites and a mechanical compression were explained as two possible reasons for the different behavior of the G band shift. Ó 2011 Elsevier B.V. All rights reserved.
1. Introduction Recently, much attention has been focused on polymer/carbon nanotubes (CNTs) nanocomposites [1–4]. CNTs are excellent fillers for the fabrication of multifunctional polymer nanocomposites because of their unique mechanical, electrical, and thermal properties, such as a high elastic modulus, high aspect ratio, and high strength-to-weight ratio [5–9]. Some CNTs are even firmer than steel and lighter than aluminum; further, they have higher conductivity than copper [1]. Besides the individual properties of CNTs, numerous potential benefits are expected when they are employed as reinforcing agents for polymers [10–14]. Raman spectroscopy, a nondestructive technique, has been widely used to explore the structural and electronic properties of carbon-based materials [15–19]. Raman spectra of single-walled carbon nanotubes (SWCNTs) and multiwalled carbon nanotubes (MWCNTs) have been studied both experimentally and theoretically. An interface region in a polymer/CNT composite is a critical factor in composite materials because it effectively controls load transfer (a type of force transferred from the polymer matrix to the fiber in composite materials) from a polymer to CNTs [20]. Raman spectroscopy is an excellent tool for interfacial investigations. It has been employed to probe interactions between a polymer and CNTs, such as the reinforcement mechanism of MWCNTsfilled polymers [21], connecting method of MWCNTs and a matrix [22,23], and polymer transition behavior using CNTs [24]. In ⇑ Corresponding author. E-mail address: [email protected] (Y. Ozaki). 0009-2614/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.cplett.2011.11.082
general, the interactions between a polymer and CNTs are reflected as Raman peak shifts or changes in the bandwidth. Raman spectroscopy can also be effectively used to quantify stress and strain transfer from surrounding environments to CNTs [25]. Many research groups have recently been involved in studies on polymer/CNT composites. Bokobza et al. [26–28] successfully synthesized nanocomposites of polydimethylsiloxane (PDMS) and styrene–butadiene rubber (SBR) with MWCNTs (PDMS, SBR/ MWCNTs) with improved mechanical, electrical, and thermal properties. This is an important factor in the improvement of physical properties of (PDMS, SBR)/MWCNTs nanocomposites. To prepare these composites, they simultaneously incorporated CNTs and another type of filler, namely, carbon black (CB) particles. This resulted in a marked improvement in the mechanical and electrical conductivity with a lower percolation threshold than that obtained with samples filled only with CNTs. In addition, they studied polymer/CNTs by Raman and infrared spectroscopy. For example, they investigated the dependence of Raman spectra on the nanotube contents and the interactions on the interface between the polymer and CNTs [23,29]. The purpose of the present study was to explore the laser heating effect on the Raman spectra of SBR/MWCNTs. By measuring the laser-power-induced Raman spectral variations, we investigated the self-rearranging behavior of MWCNTs, changes in the amount of defects and/or degree of disorder in MWCNTs, and mechanical compression from SBR to MWCNTs in the SBR/MWCNT composites. This investigation will open new avenues for further studies on the fine structure and properties of MWCNT-based nanocomposites, such as thermal expansion and thermal conductivity.
88
X. Yan et al. / Chemical Physics Letters 523 (2012) 87–91
Studies have been carried out on the temperature-dependent Raman spectra of CNTs [30–32], but only a few papers have reported studies on CNT-based nanocomposites [19,24]. 2. Experiment 2.1. Materials Multiwalled carbon nanotubes (MWNTs, Nanocyl 7000, 90% purity) were purchased from Nanocyl S.A. (Belgium). The nanotubes were fabricated by the catalytic carbon vapor deposition process without any further purification. Their average diameter and length were approximately 10 nm and 1.5 lm, respectively, and their surface area ranged between 250 and 300 m2 g1. Styrene–butadiene rubber (SBR) (Buna VSL 5025-0, Bayer) was supplied by Formix (Orléans, France). It contained 25 wt% of styrene unit and a butadiene phase with cis (10%), trans (17%), and vinyl (73%) configurations. It was compounded with sulfur (1.1 phr), stearic acid (1.1 phr), cyclohexylbenzothiazole sulfenamide (1.3 phr), diphenylguanidine (1.45 phr), and zinc oxide (1.82 phr). The unit of measurement ‘phr’ stands for parts per hundred parts of rubber by weight. 2.2. Composite processing An appropriate amount of MWNTs was dispersed into cyclohexane at an approximate ratio of 1:10 by weight by sonicating the suspension for 30 min using a Vibra-Cell VCX 500 operating at 40% amplitude with on and off cycles of 4 and 2 s, respectively. The gum containing the SBR rubber and all the ingredients for formulation were mixed separately in cyclohexane under magnetic stirring until complete dissolution and were then mixed with the MWNTs dispersion. The mixture was subjected to further sonication for 30 min if global examination by optical microscopy revealed nanotube agglomeration on a micrometer scale. The sonication process was followed by agitation under magnetic stirring until the solvent was evaporated. Total removal of any remaining solvent was achieved overnight in vacuum at 50 °C before cross-linking process and film formation. The unfilled and filled samples were then cured into plaques at 170 °C for 10 min under a pressure of 150 bar in a standard hot press. The thickness of the resulting films was approximately 300 lm. 2.3. Raman experiments Radiations of 514.5 nm from an Ar ion laser (Spectra Physics) were used for Raman excitation. A laser beam with power ranging from 85 lW to 7 mW was focused onto a spot size with a diameter of 1–2 lm, resulting in a power density of 102–104 W/cm2. A RS2100 Raman spectrophotometer (Photon Design Inc.) equipped with a CCD detector (Princeton Instrument) was employed for Raman measurements. All the Raman spectra reported here were collected with baseline correction. CN4400, an OMEGA thermoelectric device (Boulder, CO), was used as a temperature controller with an accuracy of ±0.1 °C for studying the temperature-dependent Raman spectra. 3. Results and discussion Figure 1 shows a comparison of the 514.5 nm excited Raman spectra of (A) MWCNTs, (B) SBR/MWCNTs (5 phr), and (C) SBR; the laser power was approximately 250 lW. Raman spectra of MWCNTs have been studied extensively, and their band assignments are well established [33]. A band at approximately 1350 cm1 in the Raman spectrum of MWCNTs (see Figure 1A)
Figure 1. Raman spectra of pure MWCNTs (A), SBR/MWCNTs (5 phr) (B), and pure SBR (C) excited with 514.5 nm laser beam. The laser power used was approximately 250 lW.
was assigned to the D band, which originated from the sp2 hybridized disorder in the graphitic structure; this Raman peak was most sensitive to nanotube alignment [34]. The G band at 1580 cm1 was attributed to the in-plane vibrations of the graphitic wall. This band is not very sensitive to the orientation of MWCNTs, but it can be used to investigate the load transfer of nanocomposites. It is also known as the state of filler dispersion [34–36]. Another characteristic Raman band of MWCNTs, named as 2D or G0 , was identified at 2700 cm1 in the Raman spectrum of MWCNTs (see Figure 1A). This band is also not sensitive to the nanotube alignment, but it can reflect polymer transition information and has been used to evaluate the efficiency of stress transfer between the polymer matrix and nanotubes [19,24]. Moreover, the D and G bands yield information about the amount of impurities and degree of disorder in MWCNTs, which helps in the study of crystallinity [37,38]. Most Raman bands of SBR are located in the 1800–900 cm1 (C@C stretching vibration and CAH out-the plane bending vibration) and 3200–2800 cm1 regions (CAH stretching vibrations) (see Figure 1C). In the Raman spectrum of SBR/MWCNTs (5 phr), all Raman bands of pure SBR became almost negligible. The G and D bands of MWCNTs were clearly observed at approximately 1570 and 1350 cm1, respectively, in the Raman spectrum of SBR/MWCNTs (see Figure 1B). A MWCNT is composed of several SWCNTs, which can be represented by a single layer of graphite rolled up into a seamless cylinder with a common axis. The structural properties of CNTs are governed by their special electronic structure and the large-area distribution of electronic state intensities [33]. As a result, resonance Raman scattering occurs more easily on MWCNTs as compared to polymers. Thus, the Raman spectrum in Figure 1B is practically a resonance Raman spectrum of MWCNTs, and it was difficult to obtain information about the SBR structure from this spectrum. Figure 2 shows the laser-power dependence of the Raman spectra of MWCNTs (a) and SBR/MWCNTs (10 phr) (b). The spectra shown in Figure 2 were excited with the 514.5 nm laser beam, and the laser power was varied from 85 lW to 7 mW. Since the thermal conductivity of MWCNTs is much larger than that of SBR (thermal conductivity of nanotubes is more than 103 Wm1 K1, and that of polymeric materials is in general, about 0.1 Wm1 K1 at room temperature [5]), the heating effect to Raman bonds of MWCNTs arises mainly from local temperature (local temperature means the temperature of MWCNTs). Therefore the effect from different absorption efficiency of MWCNTs and SBR for the excitation
X. Yan et al. / Chemical Physics Letters 523 (2012) 87–91
89
Figure 3. Intensity ratio of D and G bands (ID/IG) of MWCNTs (j) and 10 phr (d), 5 phr (N), and 3 phr (.) SBR/MWCNTs versus laser power.
Figure 2. Laser-power dependence of Raman spectra of MWNCTs in 1800– 1100 cm1 region in the sample of MWCNTs (a) and SBR/MWCNTs (10 phr) (b).
laser may be ignored. Figure 2a shows that there was a slight variation in the ID/IG ratio with an increase in the laser power. On the other hand, it was interesting to note that, as shown in Figure 2b, ID/IG reduced with the increase in the laser power. Finally, the intensity of the G band became higher than that of the D band. Further, as seen in Figure 2a, the D and G bands of MWCNTs show a shift to lower wavenumbers with increase in the laser power. A lower wavenumber shift was observed with temperature increase for all the CNT bands. This can be explained on the basis of the elongation of CAC bonds due to thermal expansion [39,40]. Disorders or defects in CNTs materials give enough flexibility to the CNTs structures to accommodate CAC elongation with increase in temperature [30–32,41–43]. In the present study, the G mode frequency shifted by about 15 cm1 as the laser power increased from 85 lW to 7 mW. In a previous study [41], the temperature coefficient of the frequency of the G mode was approximately 0.028 cm1/K. Based on this, the temperature range was calculated to be roughly 530 K, which is reasonable according to the temperature of CNTs estimated as a function of the incident laser power [30]. Figure 2b shows that, there was a slight shift in the D and G bands of MWCNTs in the SBR/MWCNT composites with the increase in the laser power, in contrast to the result for pure MWCNTs. This observation has already been reported by Kao and Young [19] and has been attributed to the state of dispersion of the nanotubes in the polymer matrix.
Until now, there have been limited investigations on the changes in ID/IG of CNTs and polymer/CNT composites with varying laser power and temperature. According to a study by Zhang et al. [44], the decrease in ID/IG ratios of CNTs with increasing laser power indicates a decrease in the amount of impurities and/or degree of disorders in the CNTs. Figure 3 shows the changes in ID/IG for both MWCNTs and SBR/MWCNTs (3, 5, and 10 phr) composites with varying laser power. For MWCNTs, the ID/IG ratio showed a small decrease with the initial increase in the laser power, and then remained almost constant with the further increase in the laser power (see Figures 2a and 3). This suggests that with increasing laser power, there was little change in the amount of impurities (i.e., other carbonaceous materials such as C60 and amorphous carbon) or the degree of disorder (such as tangling with each other) in the MWCNTs. However, the ID/IG ratio of SBR/MWCNT composites showed significantly different behavior than the MWCNTs. Figures 2b and 3 show that the ID/IG ratios of SBR/MWCNTs decreased rapidly with the initial increase in the laser power, and the laserpower-dependent variations in the ID/IG ratio were not dependent on the MWCNTs contents in the SBR/MWCNT composites. They remained almost constant with subsequent increase in the laser power. It should be noted that the intensity of the G band became stronger than that of the D band. Our explanation of this interesting phenomenon is that with increasing laser power, the temperature in the SBR system and MWCNTs increased. In air, the stability of CNTs is maintained up to temperatures of approximately 1023 K [45], which is much higher than that of SBR. The differential scanning calorimetry (DSC) result shows that the decompose temperature point of SBR starts at around 520 K (250 °C) and thermogravimetric analysis (TGA) has shown that pyrolysis of SBR occurs at 733 K [26]. During the temperature increase in the SBR/MWCNTs nanocomposites, SBR first began to decompose because of its low stability-threshold temperature. On the other hand, MWCNTs are released from the interaction from the SBR matrix, and assume a new separation state to form a better alignment than pure MWCNTs. This was demonstrated by the rapid decrease in ID/IG and the fact that the intensity of the G band became much higher than that of the D band (see Figure 3). Figure 3 also reveals that the laser-power and MWCNTs-content dependences of the ID/IG ratios of SBR/ MWCNTs were similar. The laser-power-dependent shifts in the G band for both MWCNTs and SBR/MWCNT composites (10, 5, and 3 phr) are shown in Figure 4. The downward shift in the G bands of CNTs with
90
X. Yan et al. / Chemical Physics Letters 523 (2012) 87–91
Figure 4. Wavenumber of G band of MWCNTs (j) and 10 phr (d), 5 phr (N), and 3 phr (.) SBR/MWCNTs versus laser power.
temperature increase has been ascribed to the elongation of CAC bonds due to thermal expansion. These temperature effects become minor in CNT materials that are free of disorder or defects, such as highly oriented pyrolytic graphite [41,46]. Figure 4 shows that with the increase in the laser power, the G band shifts to a lower wavenumber for all the samples. However, as compared to MWCNTs, the shift in the G band reduced for the SBR/MWCNT composites, especially when the density of the MWCNTs in the SBR/MWCNT composites was low. With the increase in the density of MWCNTs in the SBR/MWCNT composites, the shift in the G band increased and finally became very close to that of pure MWCNTs. For SBR/MWCNTs (3 phr) composites, the G band shifted by 9 cm1 as the laser power increased from 85 lW to 7 mW; the corresponding shift for pure MWCNTs samples was about 16 cm1. This interesting phenomenon of different shift behaviors of the G band for MWCNTs and SBR/MWCNT composites under laser heating effect could be explained in two ways One possible reason was that during the increase in the laser power, the MWCNTs rearranged themselves in the SBR/MWCNT composites, leading to a better alignment of MWCNTs in the composites and reduction in the degree of disorders or amount of defects. This phenomenon was also indicated by the change in ID/IG with laser power (see Figure 3). Another possible reason was that, in the SBR/MWCNT composites, when MWCNTs were connected to SBR, the SBR matrix induced a mechanical compression on the MWCNTs fillers. This shrinkage was transferred to the MWCNTs system, blocking the expansion of the MWCNTs with increasing temperature. The mechanical compression from the surrounding polymer to CNTs has been studied in detail [47,48]. According to previous studies, shift in the G band of Raman spectrum of CNTs embedded in a polymer matrix has been attributed to the transfer of shrinkage and thermal stresses from the polymer to the CNTs [36]. Moreover, it has been found that the G band of MWCNTs shifts to a higher wavenumber with hydrostatic pressure from a diamond anvil cell (DAC) [48]. In polymer/CNT composites, because the polymer matrix transfers the hydrostatic stress to the CNTs, it is also very likely that the G band shows a Raman frequency shift because the polymer matrix forms a compressive envelope around the nanotubes. In the present work, a difference of approximately 7 cm1 was caused in the G band shift between MWCNTs and SBR/MWCNTs (3 phr) with the increase in laser power from 85 lW to 7 mW. In a previous study on the frequency shift of the G band in a Raman spectrum of SWCNTs with hydrostatic pressure from a diamond anvil cell [48], the G band shift to a higher wavenumber by about
Figure 5. Raman spectra of SBR/MWCNTs (5 phr) measured at 303, 363, 393, 483, and 723 K using temperature controller. The excitation wavelength was 514.5 nm, and the laser power was 375 lW.
7 cm1 was found to correspond to an increase in pressure by about 200 MPa. The effects of temperature on the Raman spectra of SBR/ MWCNTs were also investigated by using a temperature controller. Figure 5 shows the Raman spectra of SBR/MWCNTs (5 phr) measured at 303, 363, 393, 483, and 723 K. It can be seen that the temperature dependence of Raman spectra of SBR/MWCNT composites investigated by using the temperature controller is consistent with that observed by varying the laser power. The ID/IG ratios of the SBR/MWCNT composites reduced considerably with increasing temperature, indicating better alignment and reduction in the degree of disorder in the MWCNTs of the SBR matrix. At the same time, the G band showed a weak shift. This was explained by the reduced degree of disorder in MWCNTs and the mechanical compression induced by the SBR matrix.
4. Conclusion This article has reported the laser heating effect on the Raman spectra of MWCNTs-based polymer nanocomposites. When the laser power was increased, the ID/IG ratios of the MWCNTs in the SBR/MWCNTs nanocomposites showed a rapid decrease in the initial stages; finally, the intensity of the G band became stronger than that of the D band. On the other hand, the ID/IG ratios of the MWCNTs showed only a small decrease. The results for the nanocomposites indicate that the amount of impurities or degree of disorders in the MWCNTs in the SBR/MWCNTs nanocomposites decreased considerably. The different behavior of ID/IG during laser heating between the SBR/MWCNT composites and MWCNTs suggests that the MWCNTs in the SBR/MWCNT composites rearranged themselves during the increase in the laser power. Moreover, the G band of the SBR/MWCNTs nanocomposites showed a smaller downward shift with temperature increase as compared to that of MWCNTs. This result may be explained in two ways. One possible reason was that, when the laser power was increased, MWCNTs in the SBR/MWCNTs systems underwent rearrangement; the improved alignment reduced the degree of disorder or amount of defects. The other possible reason was that the transfer of mechanical compression from the SBR matrix to the MWCNTs, resulting in shrinkage of the MWCNTs. A further thermal behavior study, particularly on SBR are expected by analyzing both Raman and infrared bands of SBR.
X. Yan et al. / Chemical Physics Letters 523 (2012) 87–91
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23]
M. Moniruzzaman, K.I. Winey, Macromolecules 39 (2006) 5194. J.H. Du, J. Bai, H.M. Cheng, Express Polym. Lett. 1 (2007) 253. R.F. Gibson, E.O. Ayorinde, Y.F. Wen, Compos. Sci. Technol. 67 (2007) 1. S.C. Tjong, Energy Environ. Sci. 4 (2011) 605. A. Loiseau, P. Launois, P. Petit, S. Roche, J.P. Salvetat, Understanding Carbon Nanotubes, Springer, 2006. E.T. Thostenson, Z.F. Ren, T.W. Chou, Compos. Sci. Technol. 61 (2001) 1899. B.I. Yakobson, C.J. Brabec, J. Bernholc, Phys. Rev. Lett. 76 (1996) 2511. P.G. Collins, P. Avouris, Sci. Am. 283 (2000) 62. C.F. Cornwell, L.T. Wille, Chem. Phys. Lett. 278 (1997) 262. B. Tonpheng, J.C. Yu, B.M. Andersson, O. Andersson, Macromolecules 43 (2010) 7680. Y. Zeng, L. Zhao, P.F. Liu, J.H. Du, C. Liu, Chem. J. Chin. Univ.-Chin. 31 (2010) 1263. C.Y. Min, X.Q. Shen, Z. Shi, L. Chen, Z.W. Xu, Polym.-Plast. Technol. Eng. 49 (2010) 1172. J.K. Yuan, W.L. Li, S.H. Yao, Y.Q. Lin, A. Sylvestre, J.B. Bai, Appl. Phys. Lett. 98 (2011). W.J. Choi, R.L. Poweii, D.S. Kim, Polym. Compos. 30 (2009) 415. S. Lebedkin, A. Gromov, S. Giesa, R. Gleiter, B. Renker, H. Rietschel, W. Kratschmer, Chem. Phys. Lett. 285 (1998) 210. R.J. Nemanich, S.A. Solin, Phys. Rev. B 20 (1979) 392. A.M. Rao, A. Jorio, M.A. Pimenta, M.S.S. Dantas, R. Saito, G. Dresselhaus, M.S. Dresselhaus, Phys. Rev. Lett. 85 (2000) 3545. P.H. Tan, Y. Tang, Y.M. Deng, F. Li, Y.L. Wei, H.M. Cheng, Appl. Phys. Lett. 75 (1999) 1524. C.C. Kao, R.J. Young, Compos. Sci. Technol. 64 (2004) 2291. A. Eitan, F.T. Fisher, R. Andrews, L.C. Brinson, L.S. Schadler, Compos. Sci. Technol. 66 (2006) 1162. H.J. Barraza, F. Pompeo, E.A. O’Rear, D.E. Resasco, Nano Lett. 2 (2002) 797. M. Baibarac, I. Baltog, S. Lefrant, Carbon 47 (2009) 1389. L. Bokobza, Polymer 48 (2007) 4907.
91
[24] Q. Zhao, J.R. Wood, H.D. Wagner, J. Polym. Sci. Part B-Polym. Phys. 39 (2001) 1492. [25] L. Bokobza, Macromol. Symp. 305 (2011). [26] L. Bokobza, M. Rahmani, C. Belin, J.L. Bruneel, N.E. El Bounia, J. Polym. Sci., Part B: Polym. Phys. 46 (2008) 1939. [27] L. Bokobza, M. Rahmani, Kgk-Kautschuk Gummi Kunststoffe 62 (2009) 112. [28] L. Bokobza, Kautsch. Gummi Kunstst. 62 (2009) 23. [29] L. Bokobza, Macromol. Symp. 205 (2004) 61. [30] H.D. Li et al., Appl. Phys. Lett. 76 (2000) 2053. [31] S. Chiashi, Y. Murakam, Y. Miyauchi, S. Maruyama, Jpn. J. Appl. Phys. 47 (2008) 2010. [32] L. Song, W.J. Ma, Y. Ren, W.Y. Zhou, S.S. Xie, P.H. Tan, L.F. Sun, Appl. Phys. Lett. 92 (2008). [33] M.S. Dresselhaus, G. Dresselhaus, R. Saito, A. Jorio, Phys. Rep. 409 (2005) 47. [34] S. Abbasi, P.J. Carreau, A. Derdouri, Polymer 51 (2010) 922. [35] D. Tomanek, A. Jorio, M.S. Dresselhaus, G. Dresselhaus, Carbon Nanotubes 111 (2008) 1. [36] S.R. Wang, R. Liang, B. Wang, C. Zhang, Chem. Phys. Lett. 457 (2008) 371. [37] P. Potschke, H. Brunig, A. Janke, D. Fischer, D. Jehnichen, Polymer 46 (2005) 10355. [38] S. Abbasi, P.J. Carreau, A. Derdouri, M. Moan, Rheol Acta 48 (2009) 943. [39] E. Fitzer, F. Rozploch, Carbon 26 (1988) 594. [40] P.V. Huong, R. Cavagnat, P.M. Ajayan, O. Stephan, Phys. Rev. B 51 (1995) 10048. [41] F.M. Huang, K.T. Yue, P.H. Tan, S.L. Zhang, Z.J. Shi, X.H. Zhou, Z.N. Gu, J. Appl. Phys. 84 (1998) 4022. [42] P.H. Tan, Y.M. Deng, Q. Zhao, Phys. Rev. B 58 (1998) 5435. [43] A. Siitonen, H. Kunttu, M. Pettersson, J. Phys. Chem. C 111 (2007) 1888. [44] L. Zhang et al., Phys. Rev. B 65 (2002). [45] E.T. Thostenson, C.Y. Li, T.W. Chou, Compos. Sci. Technol. 65 (2005) 491. [46] M.Z. Atashbar, S. Singamaneni, Appl. Phys. Lett. 86 (2005). [47] O. Lourie, H.D. Wagner, J. Mater. Res. 13 (1998) 2418. [48] J.R. Wood, M.D. Frogley, E.R. Meurs, A.D. Prins, T. Peijs, D.J. Dunstan, H.D. Wagner, J. Phys. Chem. B 103 (1999) 10388.
Contents lists available at SciVerse ScienceDirect
Chemical Physics Letters journal homepage: www.elsevier.com/locate/cplett
Laser heating effect on Raman spectra of styrene–butadiene rubber/multiwalled carbon nanotube nanocomposites Xinlei Yan a, Yasutaka Kitahama a, Harumi Sato a, Toshiaki Suzuki a, Xiaoxia Han a, Tamitake Itoh b, Liliane Bokobza c, Yukihiro Ozaki a,⇑ a
Department of Chemistry, School of Science and Technology and Research Center of Environmental Friendly Polymer, Kwansei-Gakuin University, Gakuen 2-1, Sanda 669-1337, Japan Nano-Bioanalysis Group, Health Research Center, National Institute of Advanced Industrial Science and Technology (AIST), Takamatsu, Kagawa 761-0395, Japan Université Pierre et Marie Curie-Ecole Supérieure de Physique et Chimie Insdustrielles – Centre National de la Recherche Scientifique, UMR 7615, 10 rue Vauquelin, 75231 Paris Cedex 05, France
b c
a r t i c l e
i n f o
Article history: Received 26 September 2011 In final form 29 November 2011 Available online 19 December 2011
a b s t r a c t The laser heating effect on MWCNTs in styrene–butadiene rubber/multiwalled carbon nanotube (SBR/ MWCNT) composites were studied by Raman spectra. The intensity ratio of the D band to G band (ID/IG) of SBR/MWCNT composites largely decreased with temperature. This indicates the self-rearranging behavior of MWCNTs in the SBR/MWCNTs system during temperature increase. In addition, the temperaturedependent downward shift of the G band of SBR/MWCNT composites was smaller than that of MWCNTs samples. The self-rearrangement of MWCNTs in SBR/MWCNT composites and a mechanical compression were explained as two possible reasons for the different behavior of the G band shift. Ó 2011 Elsevier B.V. All rights reserved.
1. Introduction Recently, much attention has been focused on polymer/carbon nanotubes (CNTs) nanocomposites [1–4]. CNTs are excellent fillers for the fabrication of multifunctional polymer nanocomposites because of their unique mechanical, electrical, and thermal properties, such as a high elastic modulus, high aspect ratio, and high strength-to-weight ratio [5–9]. Some CNTs are even firmer than steel and lighter than aluminum; further, they have higher conductivity than copper [1]. Besides the individual properties of CNTs, numerous potential benefits are expected when they are employed as reinforcing agents for polymers [10–14]. Raman spectroscopy, a nondestructive technique, has been widely used to explore the structural and electronic properties of carbon-based materials [15–19]. Raman spectra of single-walled carbon nanotubes (SWCNTs) and multiwalled carbon nanotubes (MWCNTs) have been studied both experimentally and theoretically. An interface region in a polymer/CNT composite is a critical factor in composite materials because it effectively controls load transfer (a type of force transferred from the polymer matrix to the fiber in composite materials) from a polymer to CNTs [20]. Raman spectroscopy is an excellent tool for interfacial investigations. It has been employed to probe interactions between a polymer and CNTs, such as the reinforcement mechanism of MWCNTsfilled polymers [21], connecting method of MWCNTs and a matrix [22,23], and polymer transition behavior using CNTs [24]. In ⇑ Corresponding author. E-mail address: [email protected] (Y. Ozaki). 0009-2614/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.cplett.2011.11.082
general, the interactions between a polymer and CNTs are reflected as Raman peak shifts or changes in the bandwidth. Raman spectroscopy can also be effectively used to quantify stress and strain transfer from surrounding environments to CNTs [25]. Many research groups have recently been involved in studies on polymer/CNT composites. Bokobza et al. [26–28] successfully synthesized nanocomposites of polydimethylsiloxane (PDMS) and styrene–butadiene rubber (SBR) with MWCNTs (PDMS, SBR/ MWCNTs) with improved mechanical, electrical, and thermal properties. This is an important factor in the improvement of physical properties of (PDMS, SBR)/MWCNTs nanocomposites. To prepare these composites, they simultaneously incorporated CNTs and another type of filler, namely, carbon black (CB) particles. This resulted in a marked improvement in the mechanical and electrical conductivity with a lower percolation threshold than that obtained with samples filled only with CNTs. In addition, they studied polymer/CNTs by Raman and infrared spectroscopy. For example, they investigated the dependence of Raman spectra on the nanotube contents and the interactions on the interface between the polymer and CNTs [23,29]. The purpose of the present study was to explore the laser heating effect on the Raman spectra of SBR/MWCNTs. By measuring the laser-power-induced Raman spectral variations, we investigated the self-rearranging behavior of MWCNTs, changes in the amount of defects and/or degree of disorder in MWCNTs, and mechanical compression from SBR to MWCNTs in the SBR/MWCNT composites. This investigation will open new avenues for further studies on the fine structure and properties of MWCNT-based nanocomposites, such as thermal expansion and thermal conductivity.
88
X. Yan et al. / Chemical Physics Letters 523 (2012) 87–91
Studies have been carried out on the temperature-dependent Raman spectra of CNTs [30–32], but only a few papers have reported studies on CNT-based nanocomposites [19,24]. 2. Experiment 2.1. Materials Multiwalled carbon nanotubes (MWNTs, Nanocyl 7000, 90% purity) were purchased from Nanocyl S.A. (Belgium). The nanotubes were fabricated by the catalytic carbon vapor deposition process without any further purification. Their average diameter and length were approximately 10 nm and 1.5 lm, respectively, and their surface area ranged between 250 and 300 m2 g1. Styrene–butadiene rubber (SBR) (Buna VSL 5025-0, Bayer) was supplied by Formix (Orléans, France). It contained 25 wt% of styrene unit and a butadiene phase with cis (10%), trans (17%), and vinyl (73%) configurations. It was compounded with sulfur (1.1 phr), stearic acid (1.1 phr), cyclohexylbenzothiazole sulfenamide (1.3 phr), diphenylguanidine (1.45 phr), and zinc oxide (1.82 phr). The unit of measurement ‘phr’ stands for parts per hundred parts of rubber by weight. 2.2. Composite processing An appropriate amount of MWNTs was dispersed into cyclohexane at an approximate ratio of 1:10 by weight by sonicating the suspension for 30 min using a Vibra-Cell VCX 500 operating at 40% amplitude with on and off cycles of 4 and 2 s, respectively. The gum containing the SBR rubber and all the ingredients for formulation were mixed separately in cyclohexane under magnetic stirring until complete dissolution and were then mixed with the MWNTs dispersion. The mixture was subjected to further sonication for 30 min if global examination by optical microscopy revealed nanotube agglomeration on a micrometer scale. The sonication process was followed by agitation under magnetic stirring until the solvent was evaporated. Total removal of any remaining solvent was achieved overnight in vacuum at 50 °C before cross-linking process and film formation. The unfilled and filled samples were then cured into plaques at 170 °C for 10 min under a pressure of 150 bar in a standard hot press. The thickness of the resulting films was approximately 300 lm. 2.3. Raman experiments Radiations of 514.5 nm from an Ar ion laser (Spectra Physics) were used for Raman excitation. A laser beam with power ranging from 85 lW to 7 mW was focused onto a spot size with a diameter of 1–2 lm, resulting in a power density of 102–104 W/cm2. A RS2100 Raman spectrophotometer (Photon Design Inc.) equipped with a CCD detector (Princeton Instrument) was employed for Raman measurements. All the Raman spectra reported here were collected with baseline correction. CN4400, an OMEGA thermoelectric device (Boulder, CO), was used as a temperature controller with an accuracy of ±0.1 °C for studying the temperature-dependent Raman spectra. 3. Results and discussion Figure 1 shows a comparison of the 514.5 nm excited Raman spectra of (A) MWCNTs, (B) SBR/MWCNTs (5 phr), and (C) SBR; the laser power was approximately 250 lW. Raman spectra of MWCNTs have been studied extensively, and their band assignments are well established [33]. A band at approximately 1350 cm1 in the Raman spectrum of MWCNTs (see Figure 1A)
Figure 1. Raman spectra of pure MWCNTs (A), SBR/MWCNTs (5 phr) (B), and pure SBR (C) excited with 514.5 nm laser beam. The laser power used was approximately 250 lW.
was assigned to the D band, which originated from the sp2 hybridized disorder in the graphitic structure; this Raman peak was most sensitive to nanotube alignment [34]. The G band at 1580 cm1 was attributed to the in-plane vibrations of the graphitic wall. This band is not very sensitive to the orientation of MWCNTs, but it can be used to investigate the load transfer of nanocomposites. It is also known as the state of filler dispersion [34–36]. Another characteristic Raman band of MWCNTs, named as 2D or G0 , was identified at 2700 cm1 in the Raman spectrum of MWCNTs (see Figure 1A). This band is also not sensitive to the nanotube alignment, but it can reflect polymer transition information and has been used to evaluate the efficiency of stress transfer between the polymer matrix and nanotubes [19,24]. Moreover, the D and G bands yield information about the amount of impurities and degree of disorder in MWCNTs, which helps in the study of crystallinity [37,38]. Most Raman bands of SBR are located in the 1800–900 cm1 (C@C stretching vibration and CAH out-the plane bending vibration) and 3200–2800 cm1 regions (CAH stretching vibrations) (see Figure 1C). In the Raman spectrum of SBR/MWCNTs (5 phr), all Raman bands of pure SBR became almost negligible. The G and D bands of MWCNTs were clearly observed at approximately 1570 and 1350 cm1, respectively, in the Raman spectrum of SBR/MWCNTs (see Figure 1B). A MWCNT is composed of several SWCNTs, which can be represented by a single layer of graphite rolled up into a seamless cylinder with a common axis. The structural properties of CNTs are governed by their special electronic structure and the large-area distribution of electronic state intensities [33]. As a result, resonance Raman scattering occurs more easily on MWCNTs as compared to polymers. Thus, the Raman spectrum in Figure 1B is practically a resonance Raman spectrum of MWCNTs, and it was difficult to obtain information about the SBR structure from this spectrum. Figure 2 shows the laser-power dependence of the Raman spectra of MWCNTs (a) and SBR/MWCNTs (10 phr) (b). The spectra shown in Figure 2 were excited with the 514.5 nm laser beam, and the laser power was varied from 85 lW to 7 mW. Since the thermal conductivity of MWCNTs is much larger than that of SBR (thermal conductivity of nanotubes is more than 103 Wm1 K1, and that of polymeric materials is in general, about 0.1 Wm1 K1 at room temperature [5]), the heating effect to Raman bonds of MWCNTs arises mainly from local temperature (local temperature means the temperature of MWCNTs). Therefore the effect from different absorption efficiency of MWCNTs and SBR for the excitation
X. Yan et al. / Chemical Physics Letters 523 (2012) 87–91
89
Figure 3. Intensity ratio of D and G bands (ID/IG) of MWCNTs (j) and 10 phr (d), 5 phr (N), and 3 phr (.) SBR/MWCNTs versus laser power.
Figure 2. Laser-power dependence of Raman spectra of MWNCTs in 1800– 1100 cm1 region in the sample of MWCNTs (a) and SBR/MWCNTs (10 phr) (b).
laser may be ignored. Figure 2a shows that there was a slight variation in the ID/IG ratio with an increase in the laser power. On the other hand, it was interesting to note that, as shown in Figure 2b, ID/IG reduced with the increase in the laser power. Finally, the intensity of the G band became higher than that of the D band. Further, as seen in Figure 2a, the D and G bands of MWCNTs show a shift to lower wavenumbers with increase in the laser power. A lower wavenumber shift was observed with temperature increase for all the CNT bands. This can be explained on the basis of the elongation of CAC bonds due to thermal expansion [39,40]. Disorders or defects in CNTs materials give enough flexibility to the CNTs structures to accommodate CAC elongation with increase in temperature [30–32,41–43]. In the present study, the G mode frequency shifted by about 15 cm1 as the laser power increased from 85 lW to 7 mW. In a previous study [41], the temperature coefficient of the frequency of the G mode was approximately 0.028 cm1/K. Based on this, the temperature range was calculated to be roughly 530 K, which is reasonable according to the temperature of CNTs estimated as a function of the incident laser power [30]. Figure 2b shows that, there was a slight shift in the D and G bands of MWCNTs in the SBR/MWCNT composites with the increase in the laser power, in contrast to the result for pure MWCNTs. This observation has already been reported by Kao and Young [19] and has been attributed to the state of dispersion of the nanotubes in the polymer matrix.
Until now, there have been limited investigations on the changes in ID/IG of CNTs and polymer/CNT composites with varying laser power and temperature. According to a study by Zhang et al. [44], the decrease in ID/IG ratios of CNTs with increasing laser power indicates a decrease in the amount of impurities and/or degree of disorders in the CNTs. Figure 3 shows the changes in ID/IG for both MWCNTs and SBR/MWCNTs (3, 5, and 10 phr) composites with varying laser power. For MWCNTs, the ID/IG ratio showed a small decrease with the initial increase in the laser power, and then remained almost constant with the further increase in the laser power (see Figures 2a and 3). This suggests that with increasing laser power, there was little change in the amount of impurities (i.e., other carbonaceous materials such as C60 and amorphous carbon) or the degree of disorder (such as tangling with each other) in the MWCNTs. However, the ID/IG ratio of SBR/MWCNT composites showed significantly different behavior than the MWCNTs. Figures 2b and 3 show that the ID/IG ratios of SBR/MWCNTs decreased rapidly with the initial increase in the laser power, and the laserpower-dependent variations in the ID/IG ratio were not dependent on the MWCNTs contents in the SBR/MWCNT composites. They remained almost constant with subsequent increase in the laser power. It should be noted that the intensity of the G band became stronger than that of the D band. Our explanation of this interesting phenomenon is that with increasing laser power, the temperature in the SBR system and MWCNTs increased. In air, the stability of CNTs is maintained up to temperatures of approximately 1023 K [45], which is much higher than that of SBR. The differential scanning calorimetry (DSC) result shows that the decompose temperature point of SBR starts at around 520 K (250 °C) and thermogravimetric analysis (TGA) has shown that pyrolysis of SBR occurs at 733 K [26]. During the temperature increase in the SBR/MWCNTs nanocomposites, SBR first began to decompose because of its low stability-threshold temperature. On the other hand, MWCNTs are released from the interaction from the SBR matrix, and assume a new separation state to form a better alignment than pure MWCNTs. This was demonstrated by the rapid decrease in ID/IG and the fact that the intensity of the G band became much higher than that of the D band (see Figure 3). Figure 3 also reveals that the laser-power and MWCNTs-content dependences of the ID/IG ratios of SBR/ MWCNTs were similar. The laser-power-dependent shifts in the G band for both MWCNTs and SBR/MWCNT composites (10, 5, and 3 phr) are shown in Figure 4. The downward shift in the G bands of CNTs with
90
X. Yan et al. / Chemical Physics Letters 523 (2012) 87–91
Figure 4. Wavenumber of G band of MWCNTs (j) and 10 phr (d), 5 phr (N), and 3 phr (.) SBR/MWCNTs versus laser power.
temperature increase has been ascribed to the elongation of CAC bonds due to thermal expansion. These temperature effects become minor in CNT materials that are free of disorder or defects, such as highly oriented pyrolytic graphite [41,46]. Figure 4 shows that with the increase in the laser power, the G band shifts to a lower wavenumber for all the samples. However, as compared to MWCNTs, the shift in the G band reduced for the SBR/MWCNT composites, especially when the density of the MWCNTs in the SBR/MWCNT composites was low. With the increase in the density of MWCNTs in the SBR/MWCNT composites, the shift in the G band increased and finally became very close to that of pure MWCNTs. For SBR/MWCNTs (3 phr) composites, the G band shifted by 9 cm1 as the laser power increased from 85 lW to 7 mW; the corresponding shift for pure MWCNTs samples was about 16 cm1. This interesting phenomenon of different shift behaviors of the G band for MWCNTs and SBR/MWCNT composites under laser heating effect could be explained in two ways One possible reason was that during the increase in the laser power, the MWCNTs rearranged themselves in the SBR/MWCNT composites, leading to a better alignment of MWCNTs in the composites and reduction in the degree of disorders or amount of defects. This phenomenon was also indicated by the change in ID/IG with laser power (see Figure 3). Another possible reason was that, in the SBR/MWCNT composites, when MWCNTs were connected to SBR, the SBR matrix induced a mechanical compression on the MWCNTs fillers. This shrinkage was transferred to the MWCNTs system, blocking the expansion of the MWCNTs with increasing temperature. The mechanical compression from the surrounding polymer to CNTs has been studied in detail [47,48]. According to previous studies, shift in the G band of Raman spectrum of CNTs embedded in a polymer matrix has been attributed to the transfer of shrinkage and thermal stresses from the polymer to the CNTs [36]. Moreover, it has been found that the G band of MWCNTs shifts to a higher wavenumber with hydrostatic pressure from a diamond anvil cell (DAC) [48]. In polymer/CNT composites, because the polymer matrix transfers the hydrostatic stress to the CNTs, it is also very likely that the G band shows a Raman frequency shift because the polymer matrix forms a compressive envelope around the nanotubes. In the present work, a difference of approximately 7 cm1 was caused in the G band shift between MWCNTs and SBR/MWCNTs (3 phr) with the increase in laser power from 85 lW to 7 mW. In a previous study on the frequency shift of the G band in a Raman spectrum of SWCNTs with hydrostatic pressure from a diamond anvil cell [48], the G band shift to a higher wavenumber by about
Figure 5. Raman spectra of SBR/MWCNTs (5 phr) measured at 303, 363, 393, 483, and 723 K using temperature controller. The excitation wavelength was 514.5 nm, and the laser power was 375 lW.
7 cm1 was found to correspond to an increase in pressure by about 200 MPa. The effects of temperature on the Raman spectra of SBR/ MWCNTs were also investigated by using a temperature controller. Figure 5 shows the Raman spectra of SBR/MWCNTs (5 phr) measured at 303, 363, 393, 483, and 723 K. It can be seen that the temperature dependence of Raman spectra of SBR/MWCNT composites investigated by using the temperature controller is consistent with that observed by varying the laser power. The ID/IG ratios of the SBR/MWCNT composites reduced considerably with increasing temperature, indicating better alignment and reduction in the degree of disorder in the MWCNTs of the SBR matrix. At the same time, the G band showed a weak shift. This was explained by the reduced degree of disorder in MWCNTs and the mechanical compression induced by the SBR matrix.
4. Conclusion This article has reported the laser heating effect on the Raman spectra of MWCNTs-based polymer nanocomposites. When the laser power was increased, the ID/IG ratios of the MWCNTs in the SBR/MWCNTs nanocomposites showed a rapid decrease in the initial stages; finally, the intensity of the G band became stronger than that of the D band. On the other hand, the ID/IG ratios of the MWCNTs showed only a small decrease. The results for the nanocomposites indicate that the amount of impurities or degree of disorders in the MWCNTs in the SBR/MWCNTs nanocomposites decreased considerably. The different behavior of ID/IG during laser heating between the SBR/MWCNT composites and MWCNTs suggests that the MWCNTs in the SBR/MWCNT composites rearranged themselves during the increase in the laser power. Moreover, the G band of the SBR/MWCNTs nanocomposites showed a smaller downward shift with temperature increase as compared to that of MWCNTs. This result may be explained in two ways. One possible reason was that, when the laser power was increased, MWCNTs in the SBR/MWCNTs systems underwent rearrangement; the improved alignment reduced the degree of disorder or amount of defects. The other possible reason was that the transfer of mechanical compression from the SBR matrix to the MWCNTs, resulting in shrinkage of the MWCNTs. A further thermal behavior study, particularly on SBR are expected by analyzing both Raman and infrared bands of SBR.
X. Yan et al. / Chemical Physics Letters 523 (2012) 87–91
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23]
M. Moniruzzaman, K.I. Winey, Macromolecules 39 (2006) 5194. J.H. Du, J. Bai, H.M. Cheng, Express Polym. Lett. 1 (2007) 253. R.F. Gibson, E.O. Ayorinde, Y.F. Wen, Compos. Sci. Technol. 67 (2007) 1. S.C. Tjong, Energy Environ. Sci. 4 (2011) 605. A. Loiseau, P. Launois, P. Petit, S. Roche, J.P. Salvetat, Understanding Carbon Nanotubes, Springer, 2006. E.T. Thostenson, Z.F. Ren, T.W. Chou, Compos. Sci. Technol. 61 (2001) 1899. B.I. Yakobson, C.J. Brabec, J. Bernholc, Phys. Rev. Lett. 76 (1996) 2511. P.G. Collins, P. Avouris, Sci. Am. 283 (2000) 62. C.F. Cornwell, L.T. Wille, Chem. Phys. Lett. 278 (1997) 262. B. Tonpheng, J.C. Yu, B.M. Andersson, O. Andersson, Macromolecules 43 (2010) 7680. Y. Zeng, L. Zhao, P.F. Liu, J.H. Du, C. Liu, Chem. J. Chin. Univ.-Chin. 31 (2010) 1263. C.Y. Min, X.Q. Shen, Z. Shi, L. Chen, Z.W. Xu, Polym.-Plast. Technol. Eng. 49 (2010) 1172. J.K. Yuan, W.L. Li, S.H. Yao, Y.Q. Lin, A. Sylvestre, J.B. Bai, Appl. Phys. Lett. 98 (2011). W.J. Choi, R.L. Poweii, D.S. Kim, Polym. Compos. 30 (2009) 415. S. Lebedkin, A. Gromov, S. Giesa, R. Gleiter, B. Renker, H. Rietschel, W. Kratschmer, Chem. Phys. Lett. 285 (1998) 210. R.J. Nemanich, S.A. Solin, Phys. Rev. B 20 (1979) 392. A.M. Rao, A. Jorio, M.A. Pimenta, M.S.S. Dantas, R. Saito, G. Dresselhaus, M.S. Dresselhaus, Phys. Rev. Lett. 85 (2000) 3545. P.H. Tan, Y. Tang, Y.M. Deng, F. Li, Y.L. Wei, H.M. Cheng, Appl. Phys. Lett. 75 (1999) 1524. C.C. Kao, R.J. Young, Compos. Sci. Technol. 64 (2004) 2291. A. Eitan, F.T. Fisher, R. Andrews, L.C. Brinson, L.S. Schadler, Compos. Sci. Technol. 66 (2006) 1162. H.J. Barraza, F. Pompeo, E.A. O’Rear, D.E. Resasco, Nano Lett. 2 (2002) 797. M. Baibarac, I. Baltog, S. Lefrant, Carbon 47 (2009) 1389. L. Bokobza, Polymer 48 (2007) 4907.
91
[24] Q. Zhao, J.R. Wood, H.D. Wagner, J. Polym. Sci. Part B-Polym. Phys. 39 (2001) 1492. [25] L. Bokobza, Macromol. Symp. 305 (2011). [26] L. Bokobza, M. Rahmani, C. Belin, J.L. Bruneel, N.E. El Bounia, J. Polym. Sci., Part B: Polym. Phys. 46 (2008) 1939. [27] L. Bokobza, M. Rahmani, Kgk-Kautschuk Gummi Kunststoffe 62 (2009) 112. [28] L. Bokobza, Kautsch. Gummi Kunstst. 62 (2009) 23. [29] L. Bokobza, Macromol. Symp. 205 (2004) 61. [30] H.D. Li et al., Appl. Phys. Lett. 76 (2000) 2053. [31] S. Chiashi, Y. Murakam, Y. Miyauchi, S. Maruyama, Jpn. J. Appl. Phys. 47 (2008) 2010. [32] L. Song, W.J. Ma, Y. Ren, W.Y. Zhou, S.S. Xie, P.H. Tan, L.F. Sun, Appl. Phys. Lett. 92 (2008). [33] M.S. Dresselhaus, G. Dresselhaus, R. Saito, A. Jorio, Phys. Rep. 409 (2005) 47. [34] S. Abbasi, P.J. Carreau, A. Derdouri, Polymer 51 (2010) 922. [35] D. Tomanek, A. Jorio, M.S. Dresselhaus, G. Dresselhaus, Carbon Nanotubes 111 (2008) 1. [36] S.R. Wang, R. Liang, B. Wang, C. Zhang, Chem. Phys. Lett. 457 (2008) 371. [37] P. Potschke, H. Brunig, A. Janke, D. Fischer, D. Jehnichen, Polymer 46 (2005) 10355. [38] S. Abbasi, P.J. Carreau, A. Derdouri, M. Moan, Rheol Acta 48 (2009) 943. [39] E. Fitzer, F. Rozploch, Carbon 26 (1988) 594. [40] P.V. Huong, R. Cavagnat, P.M. Ajayan, O. Stephan, Phys. Rev. B 51 (1995) 10048. [41] F.M. Huang, K.T. Yue, P.H. Tan, S.L. Zhang, Z.J. Shi, X.H. Zhou, Z.N. Gu, J. Appl. Phys. 84 (1998) 4022. [42] P.H. Tan, Y.M. Deng, Q. Zhao, Phys. Rev. B 58 (1998) 5435. [43] A. Siitonen, H. Kunttu, M. Pettersson, J. Phys. Chem. C 111 (2007) 1888. [44] L. Zhang et al., Phys. Rev. B 65 (2002). [45] E.T. Thostenson, C.Y. Li, T.W. Chou, Compos. Sci. Technol. 65 (2005) 491. [46] M.Z. Atashbar, S. Singamaneni, Appl. Phys. Lett. 86 (2005). [47] O. Lourie, H.D. Wagner, J. Mater. Res. 13 (1998) 2418. [48] J.R. Wood, M.D. Frogley, E.R. Meurs, A.D. Prins, T. Peijs, D.J. Dunstan, H.D. Wagner, J. Phys. Chem. B 103 (1999) 10388.