Journal of Materials Chemistry A

Transcription

1 Journal of Materials Chemistry A PAPER Cite this: J. Mater. Chem. A, 2013, 1, 884 Fabrication and enhanced dielectric properties of graphene polyvinylidene fluoride functional hybrid films with a polyaniline interlayer Jiwu Shang, ab Yihe Zhang,* a Li Yu, a Xinglong Luan, a Bo Shen, a Zhilei Zhang, a Fengzhu Lv a and Paul K. Chu c Received 8th October 2012 Accepted 26th October 2012 DOI: /c2ta00602b Graphene polyvinylidene fluoride hybrid films (GPNs PVDF) with a polyaniline (PANI) interlayer are fabricated by a facile and effective process. The morphology of the graphene polyaniline nanoflakes (GPNs) is examined by scanning electron microscopy (SEM) and transmission electron microscopy (TEM) and the interaction between graphene and PANI is investigated by Fourier transform infrared spectroscopy (FTIR), UV-visible spectroscopy, X-ray diffraction (XRD), and X-ray photoelectron spectroscopy (XPS). The GPNs have a layered structure resembling a cake with the graphene sheets sandwiched between the PANI layers. The GPNs have a uniform morphology which can be controlled by adjusting the ratio of PANI to graphene. The PANI inter-layer plays an active role in the dielectric properties of the GPNs PVDF composites which have low dielectric loss, high breakdown field, and large energy density. The enhanced dielectric performance originates from the insulating PANI layer which not only ensures good dispersion of graphene sheets in the PVDF but also acts as an inter-particle barrier to prevent direct contact with the graphene sheets. 1 Introduction Advanced energy storage technologies and systems have attracted considerable interest due to continuously increasing demand for more energy and projected exhaustion of fossil fuels. 1 Energy storage capacitors that allow electrical energy to be stored and released under controlled conditions are regarded as one of the solutions 1,2 and development of high power and energy density dielectric materials has become a research focus. According to U e ¼ r E 2 (3 0 ¼ Fm 1 ), the electrical energy density is proportional to the dielectric constant 3 r and square of the electric eld E (ref. 3) and hence, there are two main methods to enhance the electrical energy density, that is, by improving 3 r or E. Because of the high electric breakdown eld (E b ), low dielectric loss (tan d), low cost, and robustness, polymers are the materials of choice for energy storage applications. 1,3 However, insulating polymers have relatively low dielectric constants (3 r < 10). 4 The traditional approach to enhance the dielectric constant is to disperse highdielectric constant ceramic powders and metal particles into the a National Laboratory of Mineral Materials, School of Materials Science and Technology, China University of Geosciences, Beijing , China. zyh@ cugb.edu.cn; Fax: ; Tel: b Bright Crystals Technology Inc., Beijing , China c Department of Physics and Materials Science, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong, China polymer matrix randomly. For instance, conventional high-3 r ceramics such as BaTiO 3, Pb(Zr,Ti)O 3, and Pb(Mg 1/3 Nb 2/3 )O 3 PbTiO 3 are utilized as ceramic llers to fabricate ceramic-polymer 0 3 composites. 5 7 Other ferroelectric ceramics such as CaCu 3 Ti 4 O 12 have also been studied. 8 The ceramic particle polymer composites combine the advantages of ceramics and polymers and possess a relatively high dielectric constant and breakdown strength. However, the concentration of the ceramics must typically be over 50 vol% in order to achieve a high dielectric constant but such a large content may adversely affect the mechanical exibility of the composites. 9 Furthermore, some of the ceramics incorporated into the polymers are lead-based and not environmentally friendly 8 and applications of such composites are thus seriously limited. Another type of materials system based on metal particles may overcome the shortcomings. According to the percolation theory, metal polymer composites exhibit an insulator conductor transition when the metal concentration is increased resulting in a large enhancement of the dielectric constant near the percolation threshold. 10 Many metal particles such as Ag, Ni, Cu, Al, Zn, and stainless steel bers have been studied In addition, carbon-based particles, such as carbon nanotubes, exfoliated graphite, and carbon black, have also been utilized to enhance the dielectric constant However, these composites are prone to the compositional variations. Moreover, the dielectric loss can be quite large and the breakdown eld is low due to the insulator conductor transition near the percolation 884 J. Mater. Chem. A, 2013, 1, This journal is ª The Royal Society of Chemistry 2013

2 threshold. 20,21 Insertion of an insulating layer may overcome the limitations of metal polymer composites and for example, Shen has reported a hybrid Ag@C particle with thin organic shells coated on Ag cores. The organic dielectric shells serve as the electrical barriers between the Ag cores to form a continuous interparticle-barrier-layer network leading to large dielectric constant and low dielectric loss. 22,23 Yang has also reported an MWCNTs@PPy PS composite 24 which has a stable and large dielectric constant (44), rather small loss (<0.07), and large energy density (up to 4.95 J cm 3 ). Recently, we have produced a novel graphene polyvinylidene uoride (GNS PVDF) composite. 25 Graphene is a two-dimensional sheet of sp 2 - bonded carbon atoms packed in a honeycomb crystal lattice. 26 Because of its remarkable mechanical, thermal, electrical, optical, and catalytic properties, graphene attracts tremendous interest in the composite elds. 27 Our GNS PVDF nanocomposites exhibit a much lower percolation threshold (1.29 vol%) than materials lled with carbon nanotubes. More importantly, a high dielectric constant of 63 (at 100 Hz) is achieved at a graphene concentration of 1.27 vol% and it is 9 times higher than that of pure PVDF. 25 However, similar to other metal polymer composites, the dielectric loss of the GNS PVDF composite is quite high (>0.1 at 100 Hz) thereby hampering its potential applications. In the work reported in this paper, uniform and layered graphene polyaniline nano akes (GPNs) with graphene sheets as the inner core and polyaniline (PANI) as the outer wall are fabricated by polymerization. The morphology of the composites can be controlled by adjusting the ratio of the aniline monomer to graphene and the conductivity of graphene can be restored by the incorporated PANI. The GPNs are used as llers to prepare GPNs PVDF composites by solution casting and hot pressing. Our results indicate that the PANI layer plays an active role in the dielectric properties of the GPNs PVDF composites. The PANI dielectric shells not only act as inter-particle barriers to prevent direct contact with the graphene sheets, but also produce excellent compatibility between the llers and the polymer matrix ensuring dispersion of llers in the matrix. Compared to GNS PVDF composites, the GPNs PVDF composites boast small dielectric loss, high breakdown eld, and large energy density. The novel hybrid lms can be used to store electrical energy and have potential applications in modern electronics and electric power systems. 2 Experimental details 2.1 Materials Natural ake graphite (300 mesh) was provided by Shuangxing graphite processing plant, China and PVDF (polyvinylidene uoride, density 1.77 g cm 3 ) was purchased from Jinan Shunhua Chemical Co., Ltd. N,N-Dimethylacetamide (DMAc, $99.0%), hydrazine hydrate (N 2 H 4 $H 2 O, 80% water solution), aniline (C 6 H 5 NH, $99.5%), ammonium persulfate (APS, >98.0%), and perchloric acid (HClO 4,70 72%) were obtained from local commercial sources and used as received. 2.2 Preparation of GPN nano akes Graphene was prepared from puri ed natural graphite by chemical reduction of exfoliated Graphene Oxide (GO) 28 and the GPNs were fabricated by dilution polymerization. 29 In a typical process, graphene was loaded in a 250 ml round-bottom ask to which water ethanol (3/1, v/v) was added followed by sonication until it became clear with no visible particulate matters. The aqueous graphene solution (0.1 mg ml 1 ) was added to 40 ml of 1 M HClO 4 solution containing a variable amount of the aniline monomer. A er stirring in an ice bath for 1 h, 10 ml of 1 M HClO 4 containing (NH 4 ) 2 S 2 O 8 (molar ratio of aniline/aps is 1.5) was poured rapidly into the solution and the reaction proceeded for 24 h. An emerald product was isolated by ltration and washed with large amounts of 0.1 mol L 1 HClO 4, methanol, and deionized water, respectively. In this study, different ratios of the aniline monomer to graphene, 0 : 1 (GPNs-0), 1 : 1 (GPNs-1), 3 : 1 (GPNs-3), 5 : 1 (GPNs-5), 10 : 1 (GPNs-10) and 12 : 1 (GPNs-12), were used. GPNs were washed with a 0.14 mol L 1 ammonia solution to form the insulating PANI layer. 2.3 Fabrication of GPNs PVDF hybrid lms In our study, the apparent density of the GPNs and PVDF particles was accurately measured using a density bottle method. A measured amount of GPNs was loaded in a 100 ml conical ask and DMAc (40 ml) was added to yield an inhomogeneous dispersion. It was sonicated for 3 h to form a homogeneous suspension. A er mixing with PVDF particles (4 g) and stirring at 80 C for 2 h, the GPNs PVDF solution was drop-cast on a glass plate and kept in an oven at 80 C for 3 h to evaporate the solvent slowly to obtain the GPNs PVDF composite lms. To improve the uniformity, the lms were peeled off in water and clipped into small pieces which were compression-molded into steel boards at 210 C for 20 min at a pressure of 15 MPa. Here, the volume fraction of GPNs is 0.2, 0.5, 1.0, 1.5, 2.0, 3.0 and 5.0 vol%, respectively. The volume fraction is de ned as follows: f GPNs ¼ m G r G = m G þ m P (1) r G r P where f GPNs is the volume fraction, m is the mass and r is the density. The suffixes G and P represent the GPNs and PVDF, respectively. According to eqn (1), the measured amount of GPNs can be calculated by m G ¼ f GPNsm P r G (2) ð1 f GPNs Þr P 2.4 Characterization The morphology of the as-prepared powders was examined by scanning electron microscopy (SEM, Hitachi S-4300) and transmission electron microscopy (TEM. JEOL JEM2100F). The structure was characterized by Fourier transform infrared spectroscopy (FTIR, PerkinElmer Spectrum 100), UV-visible spectrophotometry (PerkinElmer Lambda 900), X-ray diffraction This journal is ª The Royal Society of Chemistry 2013 J. Mater. Chem. A, 2013, 1,

3 Fig. 1 SEM images of (a) graphene, (b) GPNs-1, (c) GPNs-3, (d) GPNs-5, (e) GPNs-10, and (f) GPNs-12. (XRD, Rigaku D/max RA X-ray diffractometer), as well as X-ray photoelectron spectroscopy (XPS, Axis Ultra). The thermal analysis was carried out on a TA Q100 differential scanning calorimeter at a heating rate of 10 C min 1. The dielectric property was determined by an impedance analyzer (Agilent 4294 A) at frequencies ranging from 10 2 Hz to 10 6 Hz. The lm strips were cut accurately from the samples. Prior to the measurement, silver electrodes were fabricated on the sides of these strips using conductive silver paint (Agar no. 0443). The electrical breakdown tests (300 K) were performed on an AC 2000 high-voltage instrument system (Tsing Hua University, China). An increasing voltage was applied to the top 25 mm diameter cylindrical electrode at a rate of 500 V s 1 until electrical failure. 3 Results and discussion 3.1 Morphology and microstructure of GPNs Wang proposed the possible bonding force of graphene PANI composites mainly including p p* stacking and electrostatic interactions. 30 The morphology of the composite is highly affected by the ratio of the aniline monomer to graphene and the typical morphology of the graphene and GPNs is presented in Fig. 1. The morphology differs from individual graphene sheets and changes with the increased feeding ratio of two raw materials. Fig. 1a reveals severe wrinkled structures of graphene and the sheet surface is smooth. When the aniline/graphene ratio is 1, there is no apparent change on the graphene sheets surface (Fig. 1b). According to the EDS result (inset in Fig. 1b), a speci ed amount of nitrogen (comes from NH and N]) is recorded, which indicates that there is a thin PANI layer on the graphene surface. Moreover, chlorine is also obtained which can affect the electrical properties of the lms and is never desirable for electrical devices. As is known to all, PANI has a simple acid/base doping/dedoping chemistry, and chlorine is from the incomplete base dedoping of emeraldine salt form. EDS results show that chlorine hardly depends on the GPNs content. When the ratio is increased to 3, there are many nanodots on the surface (Fig. 1c) and more nanodots appear as the ratio is increased to 5 (Fig. 1d). According to our study, the best aniline/graphene ratio is 10. As shown in Fig. 1e, many PANI nanodots, with an average diameter of 35 nm exist on the graphene surface, but PANI bers are obtained when the aniline/graphene ratio is increased to 12 (Fig. 1f). The layered GPNs-10 sample (Fig. 2b) is further con rmed by TEM. The graphene crystal lattice with a spacing of 0.35 nm can be observed from the high-resolution image (Fig. 2c). The structure of the GPNs is determined by FTIR and UV-vis spectroscopy, XPS, and XRD. As shown in Fig. 3a, all the characteristic bands of the PANI chains such as C]C stretching of quinoid at 1576 cm 1, benzene rings at 1497 cm 1, and C N stretching of secondary aromatic amine at 1302 cm 1 can be observed from the GPNs. 31 In the UV spectra in Fig. 3b, the GO peak centered at nm disappears a er chemical reduction. Fig. 2 TEM images of GPNs-10 at low (a, b) and high resolution (c). 886 J. Mater. Chem. A, 2013, 1, This journal is ª The Royal Society of Chemistry 2013

4 Fig. 3 Structural characterization of graphene, PANI, and GPNs-10 by (a) FTIR, (b) UV-vis, (c) XPS, and (d) XRD. In the PANI sample, the absorption band at nm can be assigned to the p p* transition of the benzenoid ring, 32 whereas in the GPNs, the absorption peak at nm is attributed to the n p* transition of C]O in graphene and that at nm is due to the p p* interaction between PANI and graphene. 33 Fig. 3c shows the XPS results of the layered GPNs which are mainly composed of C, O, and N. A small amount of chlorine can be detected, which is in agreement with the discussion in Fig. 1b. The deconvoluted N (1s) core level spectrum in the inset of Fig. 3c consists of three components with binding energies of ev (]N ), ev ( NH ), and ev (N + ), respectively. 33 The doping level of PANI calculated by the N + /N ratio is about 0.11, which is smaller than that previously reported. 34 This is probably due to the low concentrations of positively charged nitrogen atoms. According to the XRD spectra (Fig. 3d), two new peaks of GPNs-10 centered at 2q ¼ 19.2 and 25.1 are almost the same as that of pure PANI, which are the characteristic Bragg diffraction peaks of PANI. 31 The observed similar peaks also indicate that the PANI layer is successfully deposited onto the graphene surface. Moreover, the typical peak at 2q ¼ 23.7 of graphene completely disappears in graphene PANI composites, indicating that graphene has no aggregation and is fully exfoliated from the nano akes. Therefore, the characterization results con rm that PANI has been successfully and uniformly deposited onto the graphene surface. GPNs-10 possess perfect morphology and structure, and the property is also the best. In the following experiments, GPNs-10 are used to prepare the GPNs PVDF composites. It is noted that DMAc was used to prepare GPNs PVDF composites in this work. The PANI layer on the graphene surface may be dissolved in the DMAc because it is a good solvent for organic compounds. To con rm the possibility, an additional experiment was implemented to verify the solubility of PANI in DMAc. A measured amount of GPNs was loaded in a 100 ml conical ask and DMAc (40 ml) was added, yielding an inhomogeneous dispersion. The dispersion was sonicated and then stirred at 80 C for 2 h. A er ltration and drying, tiny mass loss was obtained. It is shown that the PANI layer on the graphene surface is not dissolved in the solvent. 3.2 Fractured surface of GPNs PVDF hybrid lms The dispersion state of GPNs-10 in the PVDF matrix is investigated. Fig. 4 shows the SEM images of the cross-section of PVDF and composites with a loading of 5.0 vol% GPNs-10. The image Fig. 4 SEM images for the fractured surface of (a) PVDF and (b d) GPNs PVDF composites with 5.0 vol% loading of GPNs-10. The row (d) labels the GPNs-10 dispersed in the PVDF matrix. This journal is ª The Royal Society of Chemistry 2013 J. Mater. Chem. A, 2013, 1,

5 Table 1 Dielectric permittivity of GPNs PVDF composites with differentgpns-10 contents at 10 4 Hz f GPNs (vol%) Dielectric permittivity Fig. 5 Dependence of the (a) AC conductivity, (b) dielectric constant, and (c) loss tangent on frequency at room temperature. demonstrates that the fractured morphology of PVDF is smooth and clean (Fig. 4a). In contrast, the SEM images of the GPNs PVDFD lms clearly reveal a layered-structure (Fig. 4c). It is well known that the interface is a basic component of a composite and has a decisive effect on the properties of the composite. Our previous work reveals that the graphene sheets are well dispersed and highly aligned in the PVDF matrix. 25 It should be noted that GPNs are hard to locate compared to GNS PVDF composites (Fig. 4d). This is because of the improved interfacial interaction between the PVDF matrix and graphene sheets coated by a PANI layer. 3.3 Enhanced dielectric properties of GPNs PVDF hybrid lms The dependence of the AC conductivity, the real part of dielectric constant (3 ¼ 3 0 i3 00 ), and dielectric loss (tan d ¼ 3 00 /3 0 )of the composites with different GPNs-10 on frequencies is shown in Fig. 5. As shown in Fig. 5a, the conductivity of the composites exhibits strong frequency dependence and increases almost linearly with frequencies. It is well known that the doping concentration is one of the most important factors affecting the electrical conductivity and other properties of PANI, 35 and the GPNs conductivity can be tuned from being insulating to conducting by varying the degree of protonation. The GPNs composites have an AC conductivity of 10 1 Scm 1 a er doping with HClO 4. However, a er dedoping by NH 3 $H 2 O, the GPNs composites exhibit much smaller conductivity (10 6 Scm 1,10 2 Hz). Although according to previous reports, 12 the conductivity of metal polymer composites increases rapidly due to the insulator conductor transition near the percolation threshold, the increase in conductivity is not observed for the GPNs PVDF hybrid lms because of the coated insulating PANI interlayer (Fig. 5a). Fig. 5b represents the frequency dependence of the dielectric constant of the PVDF and its composites. Interestingly, the dielectric permittivity decreased and then increased with the increase of the amount of GPNs (Table 1). This phenomenon is attributed to the composition and structure of functional llers used. Functional ller GPNs are made up of two parts comprising conductive graphene and an insulative PANI layer. The permittivity usually increases with the increase of the amount of conductive llers because micro-capacitors are formed in nanocomposites. 36 However, when the content of functional ller GPNs is low, few micro-capacitors formed in the nanocomposite and this effect is weak. At this time, the insulative part of functional llers with lower permittivity than PVDF plays a key role in in uencing the permittivity of nanocomposites. Thus, the permittivity decreases with the increase of the amount of GPNs due to the lower permittivity of PANI. When the GPNs content is high, many micro-capacitors are formed, and the permittivity increases with the increase of the amount of GPNs just as observed in other reports. 25,37 The dielectric constant of the PVDF is almost independent of the frequency in the measured range, and the frequency dependence becomes gradually stronger as the content of the GPNs-10 increases. The drop in the dielectric constant at high frequencies may be attributed to the leakage current of the composites. It is noteworthy that the dielectric constant does not rise dramatically as the concentration of GPNs increases. This result is in contrast to that of composites lled with conductive llers in which the dielectric constant rises dramatically when the ller content approaches the percolation threshold. 38 The phenomenon can be attributed mainly to the dielectric PANI layer. In these composites, the PANI shell serves as the electrical barrier with small loss similar to PVDF. The GPNs PVDF lms exhibit small dielectric loss (<0.05) in the frequency range of Hz. At higher frequencies, the increase in the dielectric loss becomes dominant. As shown in Fig. 5, a change of AC conductivity slope is clearly visible at 10 5 Hz. In addition, a decrease of the dielectric permittivity and simultaneously an increase of the dielectric 888 J. Mater. Chem. A, 2013, 1, This journal is ª The Royal Society of Chemistry 2013

6 Fig. 6 Dependence of the (a) dielectric constant, and (b) loss tangent on frequency for PVDF, 0.5 vol% GNS PVDF, and 0.5 vol% GPNs PVDF. Fig. 7 Dependence of electrical breakdown strength and energy density of the GPNs PVDF hybrid films on GPNs-10 content. loss are also visible at the same frequency of measurement. As is well known that conductivity, dielectric permittivity, dielectric loss and electric breakdown strength are four essential indices for dielectric materials and are interconnected. The PVDF and GPNs composites have a characteristic relaxation drop in the dielectric constant at about 10 5 Hz. 39 The dielectric loss increases dramatically because of the rising leakage currents. According to s 0 (u) ¼ u (u), where u ¼ 2pv and 3 0 is the vacuum dielectric constant, 4 the AC conductivity increases correspondingly at around 10 5 Hz. Fig. 6 shows the positive effect of the PANI interlayer on the GPNs PVDF composites. The dielectric PANI layer not only acts as an isolated barrier to prevent direct contact with the graphene sheets, but also improves the compatibility between the llers and the polymer matrix. In the graphene PVDF composites, the dielectric constants drop more rapidly and depend evidently on frequencies. 25 In the presence of the PANI interlayer, the dielectric constants of the GPNs PVDF composites depend slightly on frequencies and the trend is similar to the frequency dispersion of the PVDF matrix (Fig. 6a). It should be emphasized that the dielectric loss of the GPNs PVDF composites remains at a low level (0.04) at 10 2 to 10 5 Hz, which is much smaller than those of many other conductive llers polymer composites (Fig. 6b) ,40 This result can be due to the modi cation of the graphene sheets inducing the effect of electron tunnelling because of the good interface between the graphene and the binary polymer. 3.4 Energy storage study of GPNs PVDF hybrid lms The energy storage density (U e ) is another important characteristic of dielectric materials. U e ¼ r E b, where 3 0 is the vacuum dielectric constant and E b is the dielectric breakdown eld. 41 Fig. 7 shows the E b and U e variations of the GPNs PVDF hybrid lms as a function of f GPNs-10 at room temperature. Our previous study reveals a large decrease in the E b in the GNS PVDF composites due to the occurrence of large leakage current. 25 As shown in Fig. 7, E b of the GPNs PVDF hybrid lms increases to a maximum and then decreases with increasing concentrations of GPNs-10. At f GPNs-10 ¼ 0.5 vol%, the corresponding composites exhibit an E b of MV m 1, which is improved by about 18.0% in comparison with the PVDF matrix. The trend in U e is similar to that of E b. As shown in Fig. 7, the composites with about 0.5 vol% GPNs-10 have a large energy density of up to 3.10 J cm 3, which is about two times higher than that of the PVDF. 4 Conclusions Uniform and layered graphene polyaniline nano akes were synthesized by a dilution polymerization method, and SEM and TEM indicated that the morphology can be controlled by adjusting the ratio of the aniline monomer to graphene. FTIR, UV-vis, XRD, and XPS con rmed the incorporation of the PANI layer into the graphene sheets. The PANI layer played an active role in the dielectric properties of GPNs PVDF composites. Compared to GNS PVDF composites, the GPNs PVDF composites showed a small dielectric loss (0.04), high breakdown eld (275.0 MV m 1 ), and large energy density (3.10 J cm 3 ). The hybrid lms can be used to store electrical energy and have potential applications in modern electronics and electric power systems. Acknowledgements This study was jointly supported by the Key Project of Chinese Ministry of Education (no ), the Fundamental Research Funds for the Central Universities (nos. 2011PY0181, 2011PY0180, 2010ZY46, 2010ZD08), Special fund of Coconstruction of Beijing Education Committee, and City University of Hong Kong Research Grant (no ). This journal is ª The Royal Society of Chemistry 2013 J. Mater. Chem. A, 2013, 1,

7 References 1 Z. C. Zhang, Q. J. Meng and T. C. M. Chung, Polymer, 2009, 50, R. Schroeder, L. A. Majewski and M. Grell, Adv. Mater., 2005, 17, B. J. Chu, X. Zhou and Q. M. Zhang, Science, 2006, 313, H. Stoyanov, D. M. Carthy, M. Kollosche and G. Kofod, Appl. Phys. Lett., 2009, 94, Z. M. Dang, Y. Q. Lin, H. P. Xu, C. Y. Shi, S. T. Li and J. Bai, Adv. Funct. Mater., 2008, 18, M. Dietze and M. Es-Souni, Sens. Actuators, A, 2008, 143, Y. Bai, Z. Y. Cheng, V. Bharti, H. S. Xu and Q. M. Zhang, Appl. Phys. Lett., 2000, 76, M. Arbatti, X. Shan and Z. Cheng, Adv. Mater., 2007, 19, J. W. Wang, Q. D. Shen and C. Z. Yang, Macromolecules, 2004, 37, L. Qi, B. I. Lee, S. Chen, W. D. Samuels and G. J. Exarhos, Adv. Mater., 2005, 17, X. Y. Huang, P. K. Jiang and L. Y. Xie, Appl. Phys. Lett., 2009, 95, M. Panda, V. Srinivas and A. K. Thakur, Appl. Phys. Lett., 2008, 92, J. K. Yuan, W. L. Li, S. H. Yao, Y. Q. Lin, A. Sylvestre and J. Bai, Appl. Phys. Lett., 2011, 98, J. W. Xu and C. P. Wong, Appl. Phys. Lett., 2005, 87, Y. Zhang, Y. Wang, Y. Deng, M. Li and J. Bai, Mater. Lett., 2012, 72, Y. J. Li, M. Xu, J. Q. Feng and Z. M. Dang, Appl. Phys. Lett., 2006, 89, Z. M. Dang, L. Wang, Y. Yin, Q. Zhang and Q. Q. Lei, Adv. Mater., 2007, 19, F. He, S. Lau, H. L. Chan and J. Fan, Adv. Mater., 2009, 21, H. P. Xu, Z. M. Dang, M. J. Jiang, S. H. Yao and J. Bai, J. Mater. Chem., 2008, 18, C. Huang and Q. M. Zhang, Appl. Phys. Lett., 2004, 84, J. W. Wang, Y. Wang, F. Wang, S. Q. Li, J. Xiao and Q. D. Shen, Polymer, 2009, 50, Y. Shen, Y. Lin, M. Li and C. W. Nan, Adv. Mater., 2007, 19, Y. Shen, Y. Lin, M. Li and C. W. Nan, Adv. Funct. Mater., 2007, 17, C. Yang, Y. Lin and C. W. Nan, Carbon, 2009, 447, J. W. Shang, Y. H. Zhang, L. Yu, B. Shen, F. Z. Lv and P. K. Chu, Mater. Chem. Phys., 2012, 134, A. K. Geim and K. S. Novoselov, Nat. Mater., 2007, 6, H. Kim, A. A. Abdala and C. W. Macosko, Macromolecules, 2010, 43, S. Stankovich, D. A. Dikin, R. D. Piner, K. A. Kohlhaas, A. Kleinhammes, Y. Jia, Y. Wu, S. T. Nguyen and R. S. Ruoff, Carbon, 2007, 45, N. R. Chiou, C. Lu, J. Guan, L. J. Lee and A. J. Epstein, Nat. Nanotechnol., 2007, 2, H. Wang, Q. Hao, X. Yang, L. Lu and X. Wang, ACS Appl. Mater. Interfaces, 2010, 2, J. Xu, K. Wang, S. Z. Zu, B. H. Han and Z. Wei, ACS Nano, 2010, 4, X. Zhou, T. Wu, B. Hu, G. Yang and B. Han, Chem. Commun., 2010, 46, S. Goswami, U. N. Maiti, S. Maiti, S. Nandy, M. K. Mitra and K. K. Chattopadhyay, Carbon, 2011, 49, Y. Zhu, D. Hu, M. X. Wan, L. Jiang and Y. Wei, Adv. Mater., 2007, 19, X. R. Zeng and T. M. Ko, Polymer, 1998, 39, J. K. Yuan, Z. M. Dang, S. H. Yao, J. W. Zha, T. Zhou, S. T. Li and J. Bai, J. Mater. Chem., 2010, 20, C. Wu, X. Huang, L. Xie, X. Wu, J. Yu and P. Jiang, J. Mater. Chem., 2011, 21, C. Huang and Q. M. Zhang, Adv. Funct. Mater., 2004, 14, Z. M. Dang, J. B. Wu, L. Z. Fan and C. W. Nan, Chem. Phys. Lett., 2003, 376, L. Qi, B. I. Lee, S. Chen, W. D. Samuels and G. J. Exarhos, Adv. Mater., 2005, 17, S. Zhang, N. Zhang, C. Huang, K. Ren and Q. M. Zhang, Adv. Mater., 2005, 17, J. Mater. Chem. A, 2013, 1, This journal is ª The Royal Society of Chemistry 2013