- Email: [email protected]
Anisotropic studies of multi-wall carbon nanotube (MWCNT)-filled natural rubber (NR) and nitrile rubber (NBR) blends
Polymer Testing 32 (2013) 1229–1236
Contents lists available at ScienceDirect
Polymer Testing journal homepage: www.elsevier.com/locate/polytest
Material properties
Anisotropic studies of multi-wall carbon nanotube (MWCNT)filled natural rubber (NR) and nitrile rubber (NBR) blends Pattana Kueseng a, Pongdhorn Sae-oui b, Chakrit Sirisinha a, c, Karl I. Jacob d, e, Nittaya Rattanasom f, g, * a
Department of Chemistry, Faculty of Science, Mahidol University, Bangkok 10400, Thailand National Metal and Materials Technology Center, 114 Thailand Science Park, Klong-Luang, Pathumthani 12120, Thailand c Research and Development Centre for Thai Rubber Industry (RDCTRI), Faculty of Science, Mahidol University, Salaya, Nakhon Pathom 73170, Thailand d School of Materials Science and Engineering, Georgia Institute of Technology, Atlanta, GA 30332, USA e G. W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA 30332, USA f Rubber Technology Research Centre, Faculty of Science, Mahidol University, Salaya, Nakhon Pathom 73170, Thailand g Institute of Molecular Biosciences, Mahidol University, Salaya, Nakhon Pathom 73170, Thailand b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 31 May 2013 Accepted 14 July 2013
50/50 NR/NBR blends with various MWCNT loadings were prepared by mixing with MWCNT/NR masterbatches on a two-roll mill and sheeted off at the smallest nip gap. Then, the effect of milling direction, machine direction (MD) and transverse direction (TD), on the mechanical and electrical properties of the blends was elucidated. Dichroic ratio and SEM results confirmed that most of the MWCNTs were aligned along MD when MWCNT was less than 4 phr, and the number of agglomerates increased when MWCNT was more than 4 phr. Additionally, anisotropic properties were clearly observed when 4 phr MWCNT was loaded. At 4 phr MWCNT, 100% modulus and tensile strength in the MD were about 1.5 and 1.3 times higher than those in the TD, respectively. Moreover, electrical conductivity in the MD was superior to that in the TD by about 3 orders of magnitude. Results from dynamic mechanical tests also showed that the maximum tan d in the MD sample was lower than that in the corresponding TD sample. In addition, the storage modulus at 30 C for the MD sample containing 4 phr MWCNT was 1.15 higher than that of the corresponding TD sample. This stronger reinforcement efficiency resulted from the combination of the greater alignment and dispersion of most MWCNTs in the MD sample. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: MWCNT NR/NBR blends Dichroic ratio Mechanical properties Electrical conductivity
1. Introduction Carbon nanotubes (CNTs) are nanofibers consisting of rolled-up graphene sheet built from sp2 carbon units [1–3]. It is considered to be one of the ideal reinforcing fillers in a wide range of composite systems due to its high aspect ratio, low density, unique mechanical properties, and high * Corresponding author. Rubber Technology Research Centre, Faculty of Science, Mahidol University, Salaya, Nakhon Pathom 73170, Thailand. E-mail addresses: [email protected], [email protected] (N. Rattanasom). 0142-9418/$ – see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.polymertesting.2013.07.005
electrical and thermal conductivities [1–4]. Recently, interest in developing MWCNT-filled rubber composites with superior mechanical properties has grown significantly with potential applications [5–8]. Key requirements to achieve high mechanical properties are uniform dispersion of MWCNT in the matrix, MWCNT alignment and the interface cohesion. Our previous work has shown that the MWCNT-filled 50/50 NR/NBR blends prepared using the predispersing method (P) give better multi-wall carbon nanotube (MWCNT) dispersion materials than the corresponding blends prepared using the conventional method (C), thus resulting in better mechanical properties [9]. Also,
1230
P. Kueseng et al. / Polymer Testing 32 (2013) 1229–1236
the thermal conductivity of the P blend having 4 phr of MWCNT was 1.7 times higher than that of the corresponding C blend [9]. In addition, the alignment of MWCNT can change the composites from isotropic to anisotropic, leading to superior mechanical properties in the oriented direction [10–12]. This is because MWCNTs are also considered as short fibers and MWCNT-filled polymer composites are analogues of short fiber-reinforced composites [10]. Therefore, the theories for the strength and modulus of short fiber-reinforced polymer composites can be extended to the case of MWCNT-reinforced polymer composites [13]. The quantitative evidence to support the alignment of CNTs is defined in terms of the dichroic ratio (R) obtained from Raman spectroscopy results [14,15]. R is the ratio of the intensities of the G band of CNT at 1570 cm1 obtained from the sample arranged in parallel and perpendicular to the polarized light. If R value approaches 1, it can be said that no alignment occurs. On the other hand, if R is higher than 1, it means that the majority of MWCNTs is aligned along the MD or parallel (0 ) to the polarized light direction [14,15]. Several methods have been adopted to control the alignment of CNTs in the polymer matrix [11,12,16,17]. The alignment of MWCNTs in polystyrene (PS) can be achieved by using a twin-screw extruder [12]. It was found that the storage modulus of oriented MWCNT/PS containing 5%wt of MWCNT is 4.9 times higher than that of the corresponding sample having randomly oriented MWCNT [12]. Moreover, when the MWCNT alignment in polycarbonate (PC) is controlled by using a magnetic alignment method, the tensile strength and electrical conductivity of the magnetically aligned MWCNT/PC composite are 1.2 and 7.6 times, respectively, higher than those of the randomly aligned MWCNT/PC composite [11]. In this research, a two-roll mill, which is a simple machine, was used to prepare the 50/50 NR/NBR blends having various amounts of MWCNTs. 10 end-roll passes of the blended compounds were always made in the same direction before sheeting off at the smallest nip gap. Then, the effect of milling direction on the tensile properties, tear strength and electrical conductivity of the blends was investigated. The alignment of MWCNTs was examined using Raman spectroscopy and scanning electron microscope. 2. Experimental
phenylenediamine (IPPD) was purchased from MDR International Co., Ltd. (Bangkok, Thailand). Elemental sulfur (S8) was purchased from the Siam Chemical Public Co., Ltd. (Bangkok, Thailand). Ethyl alcohol and non-ionic surfactant (Nonidet P40) were obtained from RCI Labscan Co., Ltd. (Bangkok, Thailand). 2.2. Preparation of MWCNT/NR masterbatches MWCNT/NR masterbatches containing various MWCNT loadings (from 0 to 12 phr) were prepared by using the predispersing method described in our previous work [9]. The obtained MWCNT/NR masterbatches were subsequently used to prepare 50/50 NR/NBR blends. 2.3. Preparation of MWCNT-filled 50/50 NR/NBR compounds and vulcanizates The formulations of all 50/50 NR/NBR blends are given in Table 1. The number following the letter “B”, standing for blend, indicates the amount of MWCNTs in phr. MWCNT loadings in the blends prepared from MWCNT/NR masterbatches were varied from 0 to 6 phr. The mixing procedure to prepare the blends was the same as our previous work [9]. After adding the curatives, the compound was further mixed for 5 min and was finally passed through the smallest nip gap of the 2-roll mill ten times in the same direction before sheeting off. The cure characteristics of each compound were determined using a moving die rheometer (TechPro rheotech MDþ; Alpha Technologies, Ohio, USA). Then, the vulcanized sheets and test pieces were prepared using compression molding at 150 C according to their cure time (tc100). 2.4. Determination of mechanical and dynamic properties In this study, it was expected that the MWCNTs in the rubber compounds sheeted out using a two-roll mill at a tight nip gap would be aligned in the milling direction. Thus, the test pieces for both tensile and tear tests were punched from the vulcanized sheets having thickness of about 1 mm in two directions, i.e., machine direction (MD) and transverse direction (TD) as depicted in Fig. 1A and B, respectively. Tensile and tear properties of the specimens were determined by using a universal tester (Instron 5566; Instron Corporation, Massachusetts, USA) in accordance
2.1. Materials Natural rubber (STR 5L) and nitrile rubber (KRYNACÒ 3345 F) having acrylonitrile content of 33 wt% were purchased from Thai Rubber Latex Corporation (Thailand) Public Co., Ltd. (Samutprakarn, Thailand) and Caldic (Thailand) Ltd. (Bangkok, Thailand), respectively. MWCNT (NanocylÔ NC7000) was bought from Nanocyl S.A. In. (Sambreville, Belgium). N-tert-butyl-benzothiazole sulfenamide (TBBS) and tetrabenzylthiuram disulfide (TBzTD) were purchased from Reliance Technochem Co., Ltd. (Bangkok, Thailand). The activators, i.e., stearic acid and zinc oxide (ZnO) were obtained from Chemmin Co., Ltd. (Samutprakarn, Thailand). N-isopropyl-N0 -phenyl-p-
Table 1 NR/NBR compound formulations. Ingredient
B0
B2
B4
B5
B6
NBR MWCNT/NR masterbatch Stearic acid Zinc oxide IPPD TBBS TBzTD Sulfur
50 50a 1 3 1 1.2 0.4 2
50 52b 1 3 1 1.2 0.4 2
50 54c 1 3 1 1.2 0.4 2
50 55d 1 3 1 1.2 0.4 2
50 56e 1 3 1 1.2 0.4 2
a,b,c,d,e MWCNT/NR masterbatches having 0, 4, 8, 10 and 12 phr of MWCNT, respectively.
P. Kueseng et al. / Polymer Testing 32 (2013) 1229–1236
1231
Fig. 1. Specimen preparation for investigating the effect of milling directions; (A) MD and (B) TD.
with ISO 37 (type 1) and 34-1 (method C), respectively. The specimens were tested by using a 1 kN load cell and a crosshead speed of 500 mm/min. The values of tensile and tear properties were the average of 4–5 specimens. The dynamic mechanical properties as a function of temperature were investigated by using a dynamic mechanical thermal analyzer (Explexor TM 25 N; Gabo Qualimeter, Ahlden, Germany) in tension mode with static and dynamic strain amplitudes of 1% and 0.1%, respectively. The frequency was set at 10 Hz and the temperature was scanned from 80 C to 40 C at a heating rate of 2 C/min. 2.5. MWCNT dispersion and alignment examination To investigate the degree of MWCNT dispersion and MWCNT alignment in MD and TD, rubber pieces with the dimensions 2 50 1 mm3 were cut from the vulcanized sheets in both directions, as depicted in Fig. 2. Then, they were immersed in liquid nitrogen for at least 3 min prior to fracture at their middle part. Later, they were mounted on the stubs and the newly exposed surfaces of the samples were sputtered with platinum–palladium to prevent charging on the surface before examining under a scanning
electron microscope (SEM: JSM-6301F, JEOL Ltd., Tokyo, Japan). 2.6. Dichroic ratio measurement The degree of CNT alignment in the blends was investigated by determining the dichroic ratio obtained from Raman spectroscopy results. Raman spectrometer (NTEGRA Spectra; NT-MDT, Eindhoven, Netherlands) with a power of 10 mW was operated with 2 min of accumulation time. The excitation source used was a He/Ne laser with the wavelength of 632.8 nm. Each test piece with dimensions of 20 20 1 mm3 cut from the tensile sheet in MD was arranged in parallel and perpendicular to the polarized light and the spectra data were collected. The dichroic ratio can be calculated by using Eq. (1) [14,15].
Dichroic ratioðRÞ ¼ I== =I t
(1)
where, I//and It are intensities of the G-band of CNT at 1570 cm1 when the test piece in MD was arranged in parallel and perpendicular to polarized light, as shown in Fig. 3, respectively. 2.7. Electrical conductivity measurement The electrical conductivity was measured at room temperature using a four-point probe system with a digital multimeter (Keithley model 2400) having two outer current probes and two inner voltage probes. The test piece with the dimensions of 20 20 1 mm3 was placed on the four-point probe apparatus. The current (I) was applied in the range of 0.05–500 mA and the voltage was measured. The electrical conductivity (s) of the blends was evaluated by using Eq. (2) [18].
s¼
Fig. 2. Preparation of fractured surfaces of the blends in MD and TD.
1 ðV=IÞ)C:F:1 C:F:2 T
(2)
where T is thickness of the test piece and V is the voltage between inner probes. C.F.1 and C.F.2 are the resistance
1232
P. Kueseng et al. / Polymer Testing 32 (2013) 1229–1236
Fig. 3. Arrangement of the test piece in MD for measuring the intensity of the G-band using Raman spectrometer; (A) parallel to polarized light and (B) perpendicular to polarized light.
correction factors depending on the width and thickness of test piece, respectively [18]. 3. Results and discussion 3.1. Dispersion and alignment of MWCNT SEM micrographs showing dispersion and alignment of MWCNTs in the 50/50 NR/NBR blends are depicted in Figs. 4 and 5. It is found that, at low MWCNT loadings (4 phr), the MWCNTs are well dispersed and aligned in
Fig. 4. SEM micrographs of the blends at different MWCNT loadings in MD.
the blends. In addition, in the case of low MWCNT loadings (4 phr), it can be seen that the broken tube ends of MWCNTs protruded from the fractured surfaces of the MD samples, indicating the longitudinal alignment of MWCNTs, while the fractured surfaces of the TD samples displayed MWCNTs aligned parallel to their surfaces, as shown in Figs. 4 and 5. Thus, this result indicates that most of the MWCNTs are aligned along MD and good dispersion is observed. However, at higher loadings (>4 phr), the agglomeration of MWCNTs can be observed. The higher the MWCNT loading, the greater the MWCNT agglomerates.
Fig. 5. SEM micrographs of the blends at different MWCNT loadings in TD.
P. Kueseng et al. / Polymer Testing 32 (2013) 1229–1236
1233
3.2. Dichroic ratio Fig. 6 shows the G-band at 1570 cm1 of MWCNTs in the blends when the test piece in MD was placed parallel and perpendicular to polarized light. The dichroic ratio calculated using Eq. (1) is also plotted against MWCNT loading and presented in Fig. 7. As can be seen, R of the blends is higher than 1 at all loadings, indicating that most MWCNTs align parallel to polarized light. In other words, it could be said that the majority of MWCNTs align along MD. However, the level of MWCNT alignment is different for each MWCNT loading. It can be observed that R tends to increase with increasing MWCNT loading up to 4 phr and then decreases afterwards. The rise of R at low MWCNT loadings is attributed to the increase of aligned MWCNTs in MD or the direction parallel to the polarized light. On the other hand, the reduction of R at high MWCNT loading may be explained by the increase of MWCNT agglomerates, which is confirmed by SEM results. This leads to the limitation to increase the number of aligned MWCNTs in the machine direction. Therefore, the R values drop at high MWCNT contents. 3.3. Mechanical and dynamic properties The effect of MWCNT alignment on mechanical properties of the blends is illustrated in Figs. 8–11. It can be seen in Fig. 8 that, at a given MWCNT loading, tensile strength of the blends in the MD is greater than that in the TD. A similar result was also observed in NBR composites filled with nylon fiber, as reported by other researchers [19,20]. The greater tensile strength in the MD of the MWCNT-filled
Fig. 7. Plot of dichroic ratio (R) of the blends versus MWCNT loading.
rubber composites could be explained by the basis of fiber orientation. Generally, the tensile strength of the composite depends on the alignment of fibers. When fibers are aligned longitudinally, breakage and pulling out of the fibers take place before rupture [19,20]. It is also found that the tensile strength of the blends in both directions increases with increasing MWCNT loading until it reaches a maximum at 4 phr. The reduction of tensile strength at higher loadings is thought to arise from poor MWCNT dispersion. At high MWCNT loadings, a greater degree of entanglement could be expected. Entanglements of the long tubes make the dispersion become more difficult, and the agglomeration of MWCNTs at high loadings is also unavoidable. Elongation at break of the blends displayed in Fig. 9 decreases continuously with increasing MWCNT
Fig. 6. Raman spectra of the blends measured in parallel and perpendicular to polarized light.
1234
P. Kueseng et al. / Polymer Testing 32 (2013) 1229–1236
Fig. 8. Tensile strength of the blends in MD and TD as a function of MWCNT loading.
Fig. 10. Moduli of the blends in MD and TD at various MWCNT loadings.
loading, regardless of the alignment direction. The dilution effect could be used to explain the results. Unlike tensile strength, the results reveal that the MWCNT alignment only slightly influences the elongation at break of the blends. The effect of MWCNT alignment on moduli (100% and 300%) and tear strength of the blends is depicted in Figs. 10 and 11, respectively. It was found that the moduli and tear strength in the MD are higher than those in the TD. Again, these results confirm the alignment of MWCNTs in the MD. Thus, the higher stress is needed to deform the composites having most MWCNTs aligned parallel to the applied force. Also, the MWCNTs aligned parallel to the applied force are able to obstruct the progress of the fracture front, leading to higher tear strength [19]. It could also be observed that the moduli and tear strength of both MD and TD increase continuously with increasing MWCNT loading. This is attributed to the combination of dilution and reinforcing effects of MWCNTs. The plots of tan d (damping factor) of the blends versus temperature are displayed in Fig. 12. It is evident that there are two glass transition temperatures (Tg) around 56 C and 14 C corresponding to NR and NBR phases, respectively. It is found that MWCNT loading and alignment direction slightly influence the Tgs of both phases. However, the tan dmax of both phases tends to reduce with increasing MWCNT loading, regardless of the alignment direction. The
Plots of electrical conductivity and dichroic ratio of the blends as a function of MWCNT loading are illustrated in Fig. 14. As can be seen, the unfilled NR/NBR blend has the lowest electrical conductivity of about 4.81 1011 S/m. The addition of a small quantity of MWCNTs results in changing the electrical conductivity by orders of magnitude. For example, when the MWCNT loading increases from 2 phr to 4 phr, the electrical conductivities of the blends along the MD and TD increase from 2.35 109 to 4.24 104 S/m, and 5.88 1010 S/m to 2.74 106 S/m, respectively. In other words, the electrical conductivities of the 4 phr MWCNT-filled blends increase by about 7 (MD)
Fig. 9. Elongation at break of the blends in MD and TD at various MWCNT loadings.
Fig. 11. Tear strength of the blends in MD and TD as a function of MWCNT loading.
dilution effect could be used to explain this finding. It is also observed that tan dmax in the MD of both phases is lower than that in the TD. Fig. 13 shows the average storage modulus values at 30 C for the blends. It appears that the storage modulus of all blends in MD is higher than that of the corresponding TD samples. This result may be explained by the higher reinforcement efficiency for MWCNT aligned in the longitudinal direction. These results correspond well with the moduli and tear strength of the blends. 3.4. Electrical properties
P. Kueseng et al. / Polymer Testing 32 (2013) 1229–1236
Fig. 12. Loss tangent (tan d) of various blends in MD and TD as a function of temperature.
and 5 (TD) orders of magnitude with respect to that of the unfilled blend. The increase of electrical conductivity with increasing MWCNT loading is explained by the dilution effect. In addition, the electrical conductivity of the MD blends containing the aligned MWCNTs in the MD is superior to the corresponding TD samples. For instance, the electrical conductivity of MD sample containing 4 phr MWCNTs is higher than that of the TD sample about 3 orders of magnitude. The possible explanation is given by a greater number of aligned MWCNTs in MD sample relative to the TD sample resulting in increasing the formation of conductive networks. Therefore, electrons can flow through the formed MWCNT networks in the MD samples more than the TD samples. The strong dependence of electrical conductivity on dichroic ratio can also be observed in Fig. 14 when MWCNT loading is not more than 4 phr. It is interesting that when the dichroic ratios of both MD and TD samples reduces at higher MWCNT loading (>4 phr), the level of electrical conductivity still remains at similar level to that of the blends containing 4 phr MWCNT. As mentioned previously, the MWCNT agglomerates clearly form in the blends when MWCNT content is more that 4 phr. Interestingly, the formation of MWCNT agglomerates does not cause a reduction of the electrical conductivity. It is possible that, although the increase of MWCNT content
1235
Fig. 14. Electrical conductivity and dichroic ratio of the blends as a function of MWCNT loading.
causes the increase of MWCNT agglomerates, it does not reduce the amount of MWCNT-to-MWCNT junctions which favors the formation of conductive networks. Moreover, it has been reported that the formed agglomerates may function as junctions between the conductive MWCNT networks instead of obstructing the conductivity [21]. 4. Conclusions The alignment of MWCNTs in the 50/50 NR/NBR blends can be controlled by milling the compounds at a tight nip gap. The dichroic ratio obtained from Raman results is good quantitative evidence to indicate level of MWCNT alignment. The difference of the modulus, storage modulus at 30 C, tensile strength and electrical conductivity in MD and TD was greatest when 4 phr MWCNTs was loaded, corresponding to the highest dichroic ratio. When MWCNT content is more than 4 phr, the increase of MWCNT agglomerates and reduction of dichroic ratio are observed. This leads to the reduction of mechanical properties while the electrical conductivity remains almost constant. It can be stated that the formed MWCNT agglomerates do not obstruct the electrical conductivity if the percolating networks has formed but they may function as junctions between the conductive MWCNT networks instead. Acknowledgments The authors gratefully acknowledge the Thailand Research Fund through the Royal Golden Jubilee Ph.D. Program (Grant No. PHD/0013/2553) for the research grant. Sincere appreciation is also extended to the Research and Development Centre for Thai Rubber Industry, National Metal and Materials Technology Center (MTEC), Georgia Institute of Technology, and Mahidol University for supporting most of the materials and instruments used in this work. References
Fig. 13. Storage modulus at 30 C of the blends in MD and TD at various MWCNT loadings.
[1] X.L. Xie, Y.W. Mai, X. Ping, Dispersion and alignment of carbon nanotubes in polymer matrix: a review, Mater. Sci. Eng. 49 (4) (2005) 89.
1236
P. Kueseng et al. / Polymer Testing 32 (2013) 1229–1236
[2] J.N. Coleman, U. Khan, W.J. Blau, Y.K. Gun’ko, Small but strong: a review of the mechanical properties of carbon nanotube – polymer composites, Carbon 44 (9) (2006) 1624. [3] A.M. Shanmugharaj, J.H. Bae, K.Y. Lee, W.H. Noh, S.H. Lee, S.H. Ryu, Physical and chemical characteristics of multiwalled carbon nanotubes functionalized with aminosilane and its influence on the properties of natural rubber composites, Compos Sci. Technol. 67 (9) (2007) 1813. [4] R.H. Baughman, A.A. Zakhidov, W.A. De Heer, Carbon nanotubes the route toward applications, Science 297 (2002) 787. [5] M. Shaffer, J. Sandler, Carbon nanotube/nanofibre polymer composites, in: S.G. Advani (Ed.), Processing and Properties of Nanocomposites, World Scientific Publishing Company, 2006. [6] T.W. Ebbesen, H.J. Lezec, H. Hiura, J.W. Bennett, H.F. Ghaemie, T. Thio, Electrical conductivity of individual carbon nanotubes, Nature 382 (1996) 54. [7] A. Das, K.W. Stöckelhuber, R. Jurk, M. Saphiannikova, J. Fritzsche, H. Lorenz, M. Klüppel, G. Heinrich, Modified and unmodified multiwalled carbon nanotubes in high performance solution-styrenebutadiene and butadiene rubber blends, Polymer 49 (24) (2008) 5276. [8] L. Lan, Z. Yinghao, Z. Yong, O. Christopher, G. Sharon, Reinforcement of hydrogenated carboxylated nitrile-butadiene rubber by multiwalled carbon nanotubes, Appl. Surf. Sci. 255 (5) (2008) 2162. [9] P. Kueseng, P. Sea-oui, N. Rattanasom, Mechanical and electrical properties of natural rubber and nitrile rubber blends filled with multi-wall carbon nanotube: effect of preparation methods, Polym. Test. 32 (4) (2013) 731. [10] V.A. Buryachenko, A. Roy, Effective elastic moduli of nanocomposites with prescribed random orientation of nanofibers, Compost. B 36 (5) (2005) 405.
[11] A. Sharma, B. Tripathi, Y.K. Vijay, Dramatic improvement in properties of magnetically aligned MWCNT/polymer nanocomposites, J. Membr. Sci. 361 (1–2) (2010) 89. [12] E.T. Thostenson, T.W. Chou, Aligned multi-walled carbon nanotubereinforced composites: processing and mechanical characterization, J. Phys. D: Appl. Phys. 35 (16) (2002) 77. [13] J.N. Coleman, M. Cadek, K.P. Ryan, A. Fonseca, J.B. Nagy, W.J. Blau, Reinforcement of polymers with carbon nanotubes. The role of an ordered polymer interfacial region. Experiment and modeling, Polymer 47 (26) (2006) 8556. [14] Y. Kim, N. Minami, W.H. Zhu, S. Kazaoui, R. Azumi, M. Matsumoto, Langmuir-Blodgett films of single-wall carbon nanotubes: layer-bylayer deposition and in-plane orientation of tubes, Jpn. J. Appl. Phys. 42 (2003) 7629. [15] M.S. Dresselhaus, G. Dresselhaus, R. Saito, A. Jorio, Raman spectroscopy of carbon nanotubes, Phys. Rep. 409 (2005) 47. [16] Y.A. Kim, T. Hayashi, M. Endo, Y. Gotoh, N. Wada, J. Seiyama, Fabrication of aligned carbon nanotube-filled rubber composite, Scripta Mater. 54 (1) (2006) 31. [17] X. Wang, P.D. Bradford, W. Liu, Mechanical and electrical property improvement in CNT/nylon composites through drawing and stretching, Compos Sci. Technol. 71 (14) (2011) 1677. [18] D.K. Schroder, Semiconductor Material and Device Characterization, John Wiley & Sons, Inc., 2006. [19] T.D. Sreeja, S.K.N. Kutty, Acrylonitrile butadiene rubber/reclaimed rubber-nylon fiber composite, Adv. Polym. Tech. 20 (4) (2001) 281. [20] T.D. Sreeja, S.K.N. Kutty, Studies on acrylonitrile butadiene rubbershort nylon fiber composites, J. Elastom. Plast. 34 (2) (2002) 157. [21] J.O. Aguilar, J.R. Bautista-Quijano1, F. Avilés1, Influence of carbon nanotube clustering on the electrical conductivity of polymer composite films, Express Polym. Lett. 4 (5) (2010) 292.
Contents lists available at ScienceDirect
Polymer Testing journal homepage: www.elsevier.com/locate/polytest
Material properties
Anisotropic studies of multi-wall carbon nanotube (MWCNT)filled natural rubber (NR) and nitrile rubber (NBR) blends Pattana Kueseng a, Pongdhorn Sae-oui b, Chakrit Sirisinha a, c, Karl I. Jacob d, e, Nittaya Rattanasom f, g, * a
Department of Chemistry, Faculty of Science, Mahidol University, Bangkok 10400, Thailand National Metal and Materials Technology Center, 114 Thailand Science Park, Klong-Luang, Pathumthani 12120, Thailand c Research and Development Centre for Thai Rubber Industry (RDCTRI), Faculty of Science, Mahidol University, Salaya, Nakhon Pathom 73170, Thailand d School of Materials Science and Engineering, Georgia Institute of Technology, Atlanta, GA 30332, USA e G. W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA 30332, USA f Rubber Technology Research Centre, Faculty of Science, Mahidol University, Salaya, Nakhon Pathom 73170, Thailand g Institute of Molecular Biosciences, Mahidol University, Salaya, Nakhon Pathom 73170, Thailand b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 31 May 2013 Accepted 14 July 2013
50/50 NR/NBR blends with various MWCNT loadings were prepared by mixing with MWCNT/NR masterbatches on a two-roll mill and sheeted off at the smallest nip gap. Then, the effect of milling direction, machine direction (MD) and transverse direction (TD), on the mechanical and electrical properties of the blends was elucidated. Dichroic ratio and SEM results confirmed that most of the MWCNTs were aligned along MD when MWCNT was less than 4 phr, and the number of agglomerates increased when MWCNT was more than 4 phr. Additionally, anisotropic properties were clearly observed when 4 phr MWCNT was loaded. At 4 phr MWCNT, 100% modulus and tensile strength in the MD were about 1.5 and 1.3 times higher than those in the TD, respectively. Moreover, electrical conductivity in the MD was superior to that in the TD by about 3 orders of magnitude. Results from dynamic mechanical tests also showed that the maximum tan d in the MD sample was lower than that in the corresponding TD sample. In addition, the storage modulus at 30 C for the MD sample containing 4 phr MWCNT was 1.15 higher than that of the corresponding TD sample. This stronger reinforcement efficiency resulted from the combination of the greater alignment and dispersion of most MWCNTs in the MD sample. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: MWCNT NR/NBR blends Dichroic ratio Mechanical properties Electrical conductivity
1. Introduction Carbon nanotubes (CNTs) are nanofibers consisting of rolled-up graphene sheet built from sp2 carbon units [1–3]. It is considered to be one of the ideal reinforcing fillers in a wide range of composite systems due to its high aspect ratio, low density, unique mechanical properties, and high * Corresponding author. Rubber Technology Research Centre, Faculty of Science, Mahidol University, Salaya, Nakhon Pathom 73170, Thailand. E-mail addresses: [email protected], [email protected] (N. Rattanasom). 0142-9418/$ – see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.polymertesting.2013.07.005
electrical and thermal conductivities [1–4]. Recently, interest in developing MWCNT-filled rubber composites with superior mechanical properties has grown significantly with potential applications [5–8]. Key requirements to achieve high mechanical properties are uniform dispersion of MWCNT in the matrix, MWCNT alignment and the interface cohesion. Our previous work has shown that the MWCNT-filled 50/50 NR/NBR blends prepared using the predispersing method (P) give better multi-wall carbon nanotube (MWCNT) dispersion materials than the corresponding blends prepared using the conventional method (C), thus resulting in better mechanical properties [9]. Also,
1230
P. Kueseng et al. / Polymer Testing 32 (2013) 1229–1236
the thermal conductivity of the P blend having 4 phr of MWCNT was 1.7 times higher than that of the corresponding C blend [9]. In addition, the alignment of MWCNT can change the composites from isotropic to anisotropic, leading to superior mechanical properties in the oriented direction [10–12]. This is because MWCNTs are also considered as short fibers and MWCNT-filled polymer composites are analogues of short fiber-reinforced composites [10]. Therefore, the theories for the strength and modulus of short fiber-reinforced polymer composites can be extended to the case of MWCNT-reinforced polymer composites [13]. The quantitative evidence to support the alignment of CNTs is defined in terms of the dichroic ratio (R) obtained from Raman spectroscopy results [14,15]. R is the ratio of the intensities of the G band of CNT at 1570 cm1 obtained from the sample arranged in parallel and perpendicular to the polarized light. If R value approaches 1, it can be said that no alignment occurs. On the other hand, if R is higher than 1, it means that the majority of MWCNTs is aligned along the MD or parallel (0 ) to the polarized light direction [14,15]. Several methods have been adopted to control the alignment of CNTs in the polymer matrix [11,12,16,17]. The alignment of MWCNTs in polystyrene (PS) can be achieved by using a twin-screw extruder [12]. It was found that the storage modulus of oriented MWCNT/PS containing 5%wt of MWCNT is 4.9 times higher than that of the corresponding sample having randomly oriented MWCNT [12]. Moreover, when the MWCNT alignment in polycarbonate (PC) is controlled by using a magnetic alignment method, the tensile strength and electrical conductivity of the magnetically aligned MWCNT/PC composite are 1.2 and 7.6 times, respectively, higher than those of the randomly aligned MWCNT/PC composite [11]. In this research, a two-roll mill, which is a simple machine, was used to prepare the 50/50 NR/NBR blends having various amounts of MWCNTs. 10 end-roll passes of the blended compounds were always made in the same direction before sheeting off at the smallest nip gap. Then, the effect of milling direction on the tensile properties, tear strength and electrical conductivity of the blends was investigated. The alignment of MWCNTs was examined using Raman spectroscopy and scanning electron microscope. 2. Experimental
phenylenediamine (IPPD) was purchased from MDR International Co., Ltd. (Bangkok, Thailand). Elemental sulfur (S8) was purchased from the Siam Chemical Public Co., Ltd. (Bangkok, Thailand). Ethyl alcohol and non-ionic surfactant (Nonidet P40) were obtained from RCI Labscan Co., Ltd. (Bangkok, Thailand). 2.2. Preparation of MWCNT/NR masterbatches MWCNT/NR masterbatches containing various MWCNT loadings (from 0 to 12 phr) were prepared by using the predispersing method described in our previous work [9]. The obtained MWCNT/NR masterbatches were subsequently used to prepare 50/50 NR/NBR blends. 2.3. Preparation of MWCNT-filled 50/50 NR/NBR compounds and vulcanizates The formulations of all 50/50 NR/NBR blends are given in Table 1. The number following the letter “B”, standing for blend, indicates the amount of MWCNTs in phr. MWCNT loadings in the blends prepared from MWCNT/NR masterbatches were varied from 0 to 6 phr. The mixing procedure to prepare the blends was the same as our previous work [9]. After adding the curatives, the compound was further mixed for 5 min and was finally passed through the smallest nip gap of the 2-roll mill ten times in the same direction before sheeting off. The cure characteristics of each compound were determined using a moving die rheometer (TechPro rheotech MDþ; Alpha Technologies, Ohio, USA). Then, the vulcanized sheets and test pieces were prepared using compression molding at 150 C according to their cure time (tc100). 2.4. Determination of mechanical and dynamic properties In this study, it was expected that the MWCNTs in the rubber compounds sheeted out using a two-roll mill at a tight nip gap would be aligned in the milling direction. Thus, the test pieces for both tensile and tear tests were punched from the vulcanized sheets having thickness of about 1 mm in two directions, i.e., machine direction (MD) and transverse direction (TD) as depicted in Fig. 1A and B, respectively. Tensile and tear properties of the specimens were determined by using a universal tester (Instron 5566; Instron Corporation, Massachusetts, USA) in accordance
2.1. Materials Natural rubber (STR 5L) and nitrile rubber (KRYNACÒ 3345 F) having acrylonitrile content of 33 wt% were purchased from Thai Rubber Latex Corporation (Thailand) Public Co., Ltd. (Samutprakarn, Thailand) and Caldic (Thailand) Ltd. (Bangkok, Thailand), respectively. MWCNT (NanocylÔ NC7000) was bought from Nanocyl S.A. In. (Sambreville, Belgium). N-tert-butyl-benzothiazole sulfenamide (TBBS) and tetrabenzylthiuram disulfide (TBzTD) were purchased from Reliance Technochem Co., Ltd. (Bangkok, Thailand). The activators, i.e., stearic acid and zinc oxide (ZnO) were obtained from Chemmin Co., Ltd. (Samutprakarn, Thailand). N-isopropyl-N0 -phenyl-p-
Table 1 NR/NBR compound formulations. Ingredient
B0
B2
B4
B5
B6
NBR MWCNT/NR masterbatch Stearic acid Zinc oxide IPPD TBBS TBzTD Sulfur
50 50a 1 3 1 1.2 0.4 2
50 52b 1 3 1 1.2 0.4 2
50 54c 1 3 1 1.2 0.4 2
50 55d 1 3 1 1.2 0.4 2
50 56e 1 3 1 1.2 0.4 2
a,b,c,d,e MWCNT/NR masterbatches having 0, 4, 8, 10 and 12 phr of MWCNT, respectively.
P. Kueseng et al. / Polymer Testing 32 (2013) 1229–1236
1231
Fig. 1. Specimen preparation for investigating the effect of milling directions; (A) MD and (B) TD.
with ISO 37 (type 1) and 34-1 (method C), respectively. The specimens were tested by using a 1 kN load cell and a crosshead speed of 500 mm/min. The values of tensile and tear properties were the average of 4–5 specimens. The dynamic mechanical properties as a function of temperature were investigated by using a dynamic mechanical thermal analyzer (Explexor TM 25 N; Gabo Qualimeter, Ahlden, Germany) in tension mode with static and dynamic strain amplitudes of 1% and 0.1%, respectively. The frequency was set at 10 Hz and the temperature was scanned from 80 C to 40 C at a heating rate of 2 C/min. 2.5. MWCNT dispersion and alignment examination To investigate the degree of MWCNT dispersion and MWCNT alignment in MD and TD, rubber pieces with the dimensions 2 50 1 mm3 were cut from the vulcanized sheets in both directions, as depicted in Fig. 2. Then, they were immersed in liquid nitrogen for at least 3 min prior to fracture at their middle part. Later, they were mounted on the stubs and the newly exposed surfaces of the samples were sputtered with platinum–palladium to prevent charging on the surface before examining under a scanning
electron microscope (SEM: JSM-6301F, JEOL Ltd., Tokyo, Japan). 2.6. Dichroic ratio measurement The degree of CNT alignment in the blends was investigated by determining the dichroic ratio obtained from Raman spectroscopy results. Raman spectrometer (NTEGRA Spectra; NT-MDT, Eindhoven, Netherlands) with a power of 10 mW was operated with 2 min of accumulation time. The excitation source used was a He/Ne laser with the wavelength of 632.8 nm. Each test piece with dimensions of 20 20 1 mm3 cut from the tensile sheet in MD was arranged in parallel and perpendicular to the polarized light and the spectra data were collected. The dichroic ratio can be calculated by using Eq. (1) [14,15].
Dichroic ratioðRÞ ¼ I== =I t
(1)
where, I//and It are intensities of the G-band of CNT at 1570 cm1 when the test piece in MD was arranged in parallel and perpendicular to polarized light, as shown in Fig. 3, respectively. 2.7. Electrical conductivity measurement The electrical conductivity was measured at room temperature using a four-point probe system with a digital multimeter (Keithley model 2400) having two outer current probes and two inner voltage probes. The test piece with the dimensions of 20 20 1 mm3 was placed on the four-point probe apparatus. The current (I) was applied in the range of 0.05–500 mA and the voltage was measured. The electrical conductivity (s) of the blends was evaluated by using Eq. (2) [18].
s¼
Fig. 2. Preparation of fractured surfaces of the blends in MD and TD.
1 ðV=IÞ)C:F:1 C:F:2 T
(2)
where T is thickness of the test piece and V is the voltage between inner probes. C.F.1 and C.F.2 are the resistance
1232
P. Kueseng et al. / Polymer Testing 32 (2013) 1229–1236
Fig. 3. Arrangement of the test piece in MD for measuring the intensity of the G-band using Raman spectrometer; (A) parallel to polarized light and (B) perpendicular to polarized light.
correction factors depending on the width and thickness of test piece, respectively [18]. 3. Results and discussion 3.1. Dispersion and alignment of MWCNT SEM micrographs showing dispersion and alignment of MWCNTs in the 50/50 NR/NBR blends are depicted in Figs. 4 and 5. It is found that, at low MWCNT loadings (4 phr), the MWCNTs are well dispersed and aligned in
Fig. 4. SEM micrographs of the blends at different MWCNT loadings in MD.
the blends. In addition, in the case of low MWCNT loadings (4 phr), it can be seen that the broken tube ends of MWCNTs protruded from the fractured surfaces of the MD samples, indicating the longitudinal alignment of MWCNTs, while the fractured surfaces of the TD samples displayed MWCNTs aligned parallel to their surfaces, as shown in Figs. 4 and 5. Thus, this result indicates that most of the MWCNTs are aligned along MD and good dispersion is observed. However, at higher loadings (>4 phr), the agglomeration of MWCNTs can be observed. The higher the MWCNT loading, the greater the MWCNT agglomerates.
Fig. 5. SEM micrographs of the blends at different MWCNT loadings in TD.
P. Kueseng et al. / Polymer Testing 32 (2013) 1229–1236
1233
3.2. Dichroic ratio Fig. 6 shows the G-band at 1570 cm1 of MWCNTs in the blends when the test piece in MD was placed parallel and perpendicular to polarized light. The dichroic ratio calculated using Eq. (1) is also plotted against MWCNT loading and presented in Fig. 7. As can be seen, R of the blends is higher than 1 at all loadings, indicating that most MWCNTs align parallel to polarized light. In other words, it could be said that the majority of MWCNTs align along MD. However, the level of MWCNT alignment is different for each MWCNT loading. It can be observed that R tends to increase with increasing MWCNT loading up to 4 phr and then decreases afterwards. The rise of R at low MWCNT loadings is attributed to the increase of aligned MWCNTs in MD or the direction parallel to the polarized light. On the other hand, the reduction of R at high MWCNT loading may be explained by the increase of MWCNT agglomerates, which is confirmed by SEM results. This leads to the limitation to increase the number of aligned MWCNTs in the machine direction. Therefore, the R values drop at high MWCNT contents. 3.3. Mechanical and dynamic properties The effect of MWCNT alignment on mechanical properties of the blends is illustrated in Figs. 8–11. It can be seen in Fig. 8 that, at a given MWCNT loading, tensile strength of the blends in the MD is greater than that in the TD. A similar result was also observed in NBR composites filled with nylon fiber, as reported by other researchers [19,20]. The greater tensile strength in the MD of the MWCNT-filled
Fig. 7. Plot of dichroic ratio (R) of the blends versus MWCNT loading.
rubber composites could be explained by the basis of fiber orientation. Generally, the tensile strength of the composite depends on the alignment of fibers. When fibers are aligned longitudinally, breakage and pulling out of the fibers take place before rupture [19,20]. It is also found that the tensile strength of the blends in both directions increases with increasing MWCNT loading until it reaches a maximum at 4 phr. The reduction of tensile strength at higher loadings is thought to arise from poor MWCNT dispersion. At high MWCNT loadings, a greater degree of entanglement could be expected. Entanglements of the long tubes make the dispersion become more difficult, and the agglomeration of MWCNTs at high loadings is also unavoidable. Elongation at break of the blends displayed in Fig. 9 decreases continuously with increasing MWCNT
Fig. 6. Raman spectra of the blends measured in parallel and perpendicular to polarized light.
1234
P. Kueseng et al. / Polymer Testing 32 (2013) 1229–1236
Fig. 8. Tensile strength of the blends in MD and TD as a function of MWCNT loading.
Fig. 10. Moduli of the blends in MD and TD at various MWCNT loadings.
loading, regardless of the alignment direction. The dilution effect could be used to explain the results. Unlike tensile strength, the results reveal that the MWCNT alignment only slightly influences the elongation at break of the blends. The effect of MWCNT alignment on moduli (100% and 300%) and tear strength of the blends is depicted in Figs. 10 and 11, respectively. It was found that the moduli and tear strength in the MD are higher than those in the TD. Again, these results confirm the alignment of MWCNTs in the MD. Thus, the higher stress is needed to deform the composites having most MWCNTs aligned parallel to the applied force. Also, the MWCNTs aligned parallel to the applied force are able to obstruct the progress of the fracture front, leading to higher tear strength [19]. It could also be observed that the moduli and tear strength of both MD and TD increase continuously with increasing MWCNT loading. This is attributed to the combination of dilution and reinforcing effects of MWCNTs. The plots of tan d (damping factor) of the blends versus temperature are displayed in Fig. 12. It is evident that there are two glass transition temperatures (Tg) around 56 C and 14 C corresponding to NR and NBR phases, respectively. It is found that MWCNT loading and alignment direction slightly influence the Tgs of both phases. However, the tan dmax of both phases tends to reduce with increasing MWCNT loading, regardless of the alignment direction. The
Plots of electrical conductivity and dichroic ratio of the blends as a function of MWCNT loading are illustrated in Fig. 14. As can be seen, the unfilled NR/NBR blend has the lowest electrical conductivity of about 4.81 1011 S/m. The addition of a small quantity of MWCNTs results in changing the electrical conductivity by orders of magnitude. For example, when the MWCNT loading increases from 2 phr to 4 phr, the electrical conductivities of the blends along the MD and TD increase from 2.35 109 to 4.24 104 S/m, and 5.88 1010 S/m to 2.74 106 S/m, respectively. In other words, the electrical conductivities of the 4 phr MWCNT-filled blends increase by about 7 (MD)
Fig. 9. Elongation at break of the blends in MD and TD at various MWCNT loadings.
Fig. 11. Tear strength of the blends in MD and TD as a function of MWCNT loading.
dilution effect could be used to explain this finding. It is also observed that tan dmax in the MD of both phases is lower than that in the TD. Fig. 13 shows the average storage modulus values at 30 C for the blends. It appears that the storage modulus of all blends in MD is higher than that of the corresponding TD samples. This result may be explained by the higher reinforcement efficiency for MWCNT aligned in the longitudinal direction. These results correspond well with the moduli and tear strength of the blends. 3.4. Electrical properties
P. Kueseng et al. / Polymer Testing 32 (2013) 1229–1236
Fig. 12. Loss tangent (tan d) of various blends in MD and TD as a function of temperature.
and 5 (TD) orders of magnitude with respect to that of the unfilled blend. The increase of electrical conductivity with increasing MWCNT loading is explained by the dilution effect. In addition, the electrical conductivity of the MD blends containing the aligned MWCNTs in the MD is superior to the corresponding TD samples. For instance, the electrical conductivity of MD sample containing 4 phr MWCNTs is higher than that of the TD sample about 3 orders of magnitude. The possible explanation is given by a greater number of aligned MWCNTs in MD sample relative to the TD sample resulting in increasing the formation of conductive networks. Therefore, electrons can flow through the formed MWCNT networks in the MD samples more than the TD samples. The strong dependence of electrical conductivity on dichroic ratio can also be observed in Fig. 14 when MWCNT loading is not more than 4 phr. It is interesting that when the dichroic ratios of both MD and TD samples reduces at higher MWCNT loading (>4 phr), the level of electrical conductivity still remains at similar level to that of the blends containing 4 phr MWCNT. As mentioned previously, the MWCNT agglomerates clearly form in the blends when MWCNT content is more that 4 phr. Interestingly, the formation of MWCNT agglomerates does not cause a reduction of the electrical conductivity. It is possible that, although the increase of MWCNT content
1235
Fig. 14. Electrical conductivity and dichroic ratio of the blends as a function of MWCNT loading.
causes the increase of MWCNT agglomerates, it does not reduce the amount of MWCNT-to-MWCNT junctions which favors the formation of conductive networks. Moreover, it has been reported that the formed agglomerates may function as junctions between the conductive MWCNT networks instead of obstructing the conductivity [21]. 4. Conclusions The alignment of MWCNTs in the 50/50 NR/NBR blends can be controlled by milling the compounds at a tight nip gap. The dichroic ratio obtained from Raman results is good quantitative evidence to indicate level of MWCNT alignment. The difference of the modulus, storage modulus at 30 C, tensile strength and electrical conductivity in MD and TD was greatest when 4 phr MWCNTs was loaded, corresponding to the highest dichroic ratio. When MWCNT content is more than 4 phr, the increase of MWCNT agglomerates and reduction of dichroic ratio are observed. This leads to the reduction of mechanical properties while the electrical conductivity remains almost constant. It can be stated that the formed MWCNT agglomerates do not obstruct the electrical conductivity if the percolating networks has formed but they may function as junctions between the conductive MWCNT networks instead. Acknowledgments The authors gratefully acknowledge the Thailand Research Fund through the Royal Golden Jubilee Ph.D. Program (Grant No. PHD/0013/2553) for the research grant. Sincere appreciation is also extended to the Research and Development Centre for Thai Rubber Industry, National Metal and Materials Technology Center (MTEC), Georgia Institute of Technology, and Mahidol University for supporting most of the materials and instruments used in this work. References
Fig. 13. Storage modulus at 30 C of the blends in MD and TD at various MWCNT loadings.
[1] X.L. Xie, Y.W. Mai, X. Ping, Dispersion and alignment of carbon nanotubes in polymer matrix: a review, Mater. Sci. Eng. 49 (4) (2005) 89.
1236
P. Kueseng et al. / Polymer Testing 32 (2013) 1229–1236
[2] J.N. Coleman, U. Khan, W.J. Blau, Y.K. Gun’ko, Small but strong: a review of the mechanical properties of carbon nanotube – polymer composites, Carbon 44 (9) (2006) 1624. [3] A.M. Shanmugharaj, J.H. Bae, K.Y. Lee, W.H. Noh, S.H. Lee, S.H. Ryu, Physical and chemical characteristics of multiwalled carbon nanotubes functionalized with aminosilane and its influence on the properties of natural rubber composites, Compos Sci. Technol. 67 (9) (2007) 1813. [4] R.H. Baughman, A.A. Zakhidov, W.A. De Heer, Carbon nanotubes the route toward applications, Science 297 (2002) 787. [5] M. Shaffer, J. Sandler, Carbon nanotube/nanofibre polymer composites, in: S.G. Advani (Ed.), Processing and Properties of Nanocomposites, World Scientific Publishing Company, 2006. [6] T.W. Ebbesen, H.J. Lezec, H. Hiura, J.W. Bennett, H.F. Ghaemie, T. Thio, Electrical conductivity of individual carbon nanotubes, Nature 382 (1996) 54. [7] A. Das, K.W. Stöckelhuber, R. Jurk, M. Saphiannikova, J. Fritzsche, H. Lorenz, M. Klüppel, G. Heinrich, Modified and unmodified multiwalled carbon nanotubes in high performance solution-styrenebutadiene and butadiene rubber blends, Polymer 49 (24) (2008) 5276. [8] L. Lan, Z. Yinghao, Z. Yong, O. Christopher, G. Sharon, Reinforcement of hydrogenated carboxylated nitrile-butadiene rubber by multiwalled carbon nanotubes, Appl. Surf. Sci. 255 (5) (2008) 2162. [9] P. Kueseng, P. Sea-oui, N. Rattanasom, Mechanical and electrical properties of natural rubber and nitrile rubber blends filled with multi-wall carbon nanotube: effect of preparation methods, Polym. Test. 32 (4) (2013) 731. [10] V.A. Buryachenko, A. Roy, Effective elastic moduli of nanocomposites with prescribed random orientation of nanofibers, Compost. B 36 (5) (2005) 405.
[11] A. Sharma, B. Tripathi, Y.K. Vijay, Dramatic improvement in properties of magnetically aligned MWCNT/polymer nanocomposites, J. Membr. Sci. 361 (1–2) (2010) 89. [12] E.T. Thostenson, T.W. Chou, Aligned multi-walled carbon nanotubereinforced composites: processing and mechanical characterization, J. Phys. D: Appl. Phys. 35 (16) (2002) 77. [13] J.N. Coleman, M. Cadek, K.P. Ryan, A. Fonseca, J.B. Nagy, W.J. Blau, Reinforcement of polymers with carbon nanotubes. The role of an ordered polymer interfacial region. Experiment and modeling, Polymer 47 (26) (2006) 8556. [14] Y. Kim, N. Minami, W.H. Zhu, S. Kazaoui, R. Azumi, M. Matsumoto, Langmuir-Blodgett films of single-wall carbon nanotubes: layer-bylayer deposition and in-plane orientation of tubes, Jpn. J. Appl. Phys. 42 (2003) 7629. [15] M.S. Dresselhaus, G. Dresselhaus, R. Saito, A. Jorio, Raman spectroscopy of carbon nanotubes, Phys. Rep. 409 (2005) 47. [16] Y.A. Kim, T. Hayashi, M. Endo, Y. Gotoh, N. Wada, J. Seiyama, Fabrication of aligned carbon nanotube-filled rubber composite, Scripta Mater. 54 (1) (2006) 31. [17] X. Wang, P.D. Bradford, W. Liu, Mechanical and electrical property improvement in CNT/nylon composites through drawing and stretching, Compos Sci. Technol. 71 (14) (2011) 1677. [18] D.K. Schroder, Semiconductor Material and Device Characterization, John Wiley & Sons, Inc., 2006. [19] T.D. Sreeja, S.K.N. Kutty, Acrylonitrile butadiene rubber/reclaimed rubber-nylon fiber composite, Adv. Polym. Tech. 20 (4) (2001) 281. [20] T.D. Sreeja, S.K.N. Kutty, Studies on acrylonitrile butadiene rubbershort nylon fiber composites, J. Elastom. Plast. 34 (2) (2002) 157. [21] J.O. Aguilar, J.R. Bautista-Quijano1, F. Avilés1, Influence of carbon nanotube clustering on the electrical conductivity of polymer composite films, Express Polym. Lett. 4 (5) (2010) 292.