Enhancing the dielectric properties of highly compatible new polyimide/γ-ray irradiated MWCNT nanocomposites

Abstract

Novel polyimide/γ-ray irradiated MWCNT (PI/γ-MWCNT) nanocomposites with improved dielectric properties were fabricated by casting and curing processes. The interfacial interactions between the two domains, i.e. PI and MWCNTs, were enhanced by hydrogen bonding between the hydroxyl groups present on PI and modified CNTs. A PI matrix having pendant phenolic hydroxyl groups was derived from pyromellitic dianhydride (PMDA) and diamine monomer 4,4′-diamino-4′′-hydroxytriphenylmethane. MWCNTs (5–20 wt%) were dispersed in the synthesized PI matrix. Before addition to PI, the surface of MWCNTs was equipped with hydroxyl and carboxylic groups by irradiating with γ-rays under a dry oxygen environment. Surface examination of PI/γ-MWCNTs composite films by scanning electron microscopy (SEM) revealed that MWCNTs are uniformly dispersed and completely wrapped by the PI matrix, most likely due to the hydrogen bonding. The influence of greater adhesion of MWCNTs with PI matrix on the dielectric, visco-elastic, and mechanical properties of final PI/γ-MWCNTs nanocomposites was explored using appropriate analytical techniques. The composite films exhibited high dielectric constant, a 7.6 fold improvement as compared to pristine PI. The storage modulus (E′) and glass transition temperature (T_g) demonstrated an improvement of 1.4 and 1.2 fold, respectively. Similarly, mechanical and thermal properties were also found to be improved remarkably. We believe that significant property enhancement of PI/γ-MWCNTs nanocomposites is the direct consequence of increased interface compatibility via hydrogen bonding between the polymer matrix and the carbon nano-filler.

---

Introduction

Polyimide–CNTs nanocomposites have attracted the interest of researchers as materials with high dielectric constant.^1–5 These nanocomposites are widely used as dielectric material in actuators,^6–12 capacitors,^13–16 and high power density pulsed power applications.^17–19 Traditionally, ceramic based dielectrics were being used widely. However, due to high density, brittleness, and poor plasticity they cannot make them compatible with the requirements of many applications.^20–23 Polymer-based composites were explored as dielectric materials to solve the problems associated with other conventional dielectrics.^24,25 These polymer-based dielectrics provided an alternative for ceramic based dielectrics due to their low cost, mechanical flexibility, and easy processing.^2,26–28 Among these polymers, polyvinylidene fluoride (PVDF) and its copolymers are the most widely explored materials due to high dielectric constants originating from their high dipole density.^1,27,29 The other polymers chosen as dielectric materials include polyethersulfone,³⁰ polyarylene ether nitriles,³¹ epoxy polymers,³² polyimide,³³ polyurethane,³⁴ poly(methyl methacrylate).³⁵ Considering high mechanical strength, good solvent resistance, excellent thermal stability, and low dielectric loss, PIs gained preference over other polymeric dielectrics.^36–42 PIs have a dielectric constant of 2.8–5.0 which is considered low for meeting the requirements of certain applications.^43–45 The applicability of PIs as dielectric material can be enhanced by substantially increasing their dielectric constant while retaining their high tensile strength, thermal stability, and chemical resistance. Generally, dielectric constant of polymers can be increased by two approaches. One approach is the addition of high-k ceramic fillers including nano-scale TiO₂,³³ Al₂O₃,⁴⁶ CaCu₃–Ti₄O₁₂ (ref. 47) into PI matrix. However, the disadvantage of this method is that high dielectric constant can be achieved only by adding a larger amount of ceramic filler, which can dramatically decrease the tensile properties of PI hindering their tendency as dielectric materials.^1,24 The other approach is the incorporation of conductive filler such as carbon nanotubes, carbon black, and graphene.⁴³ Various researchers reported the dispersion of CNTs in poly(methyl methacrylate),³⁵ PVDF,^{29,37,48–50} cyanate ester,^51–54 and polyurethane.^34,55 All these studies demonstrate that the dielectric constant increased up to a certain level at which the nano-filler was uniformly dispersed in matrix, beyond that level the aggregation of filler was observed which decreased the dielectric constant of polymer composites. The aggregation of filler is due to reduced interfacial interaction between filler and matrix, resulting in small available surface area, absence of stress transfer, and creation of stress concentration defects.⁵⁶ It is well established that interfacial interactions dictate the uniform dispersion of filler into a matrix which influence the physical and mechanical properties of polymer based dielectrics.^57–60 Molecular design of interfacial chemistry is needed for controlling physico-chemical properties of high performance nanocomposites. Several strategies have been reported to improve interface interaction between CNTs and the polymer matrix. For instance, Jia et al.⁶¹ reported PI–CNTs nanocomposites in which the super-long CNTs and PI chains were supposed to be vertically aligned and interface bonding between two domains was expected to improve through van der Waals interactions. The nano-filler was found to be dispersed uniformly below 0.3 wt% contents, as the contents increased above 0.3 wt%, CNTs were aggregated which was reflected by a decrease in tensile strength above 0.3 wt% content of CNTs. In an another study, Wang et al.⁶² added 4,4′-diaminodiphenylether (ODA)-functionalized CNTs in PI matrix. He reported the improvement in thermal and mechanical properties of resulting composites up to 3 wt% addition of CNTs, after which the CNTs were aggregated resulting in a decrease of properties. Similarly, Chen et al.¹ succeeded in uniform dispersion of 10 wt% of NH₂-functionalized MWCTs into PI matrix, however, 15 and 20 wt% incorporation resulted in CNTs aggregation, and thereby, decline in mechanical and dielectric properties. In all the previous studies, the approach adopted to improve interfacial interaction was based only on modification and functionalization of CNTs, however, creation of functional groups in PI matrix was not addressed.

In the present study, the interface compatibility between MWCNTs and PI matrix was enhanced by hydrogen bonding between two domains. In this regard, a PI containing pendant phenyl hydroxyl group was prepared by reacting 4,4′-diamino-4′′-hydroxytriphenylmethane with pyromellitic dianhydride (PMDA). In order to ensure uniform dispersion, MWCNTs were modified by γ-ray irradiation under a dry O₂ atmosphere,⁶³ which furnished the MWCNTs surface with hydroxyl and carboxylic groups. As a consequence, flexible PI/γ-irradiated MWCNTs nanocomposite films were prepared in which interfacial interactions were remarkably enhanced due to hydrogen bonding between hydroxyl groups present at MWCNTs surface and PI chains simultaneously. Furthermore, the carbonyl groups of both components might have increased interfacial interactions through secondary bond forces. To best of our knowledge, this is the first report that demonstrates a uniform dispersion of 20 wt% MWCNTs into PI matrix through improved compatibility, which resulted in a remarkable increase in the dielectric properties of PI matrix along with improvement in tensile strength and thermo-mechanical properties polyimide/MWCNTs nanocomposites.

Experimental

Materials

MWCNTs, 90% pure (length 1.5 μm, average diameter 9.5 nm) were purchased from Nanocyl™ Co. Pyromellitic dianhydride (PMDA, 97%), diamine monomer, 4-(bis(4-aminophenyl)methyl)phenol was synthesized according to previously published method,⁶⁴ dimethylacetamide (DMAc, 99.8%, water content <0.005%), hydrochloric acid from Sigma-Aldrich, 4-hydroxybenzaldehyde from Merck, whereas dimethylformamide (DMF) and ethanol were purchased from the Lab-Scan.

Surface modification of MWCNTs

The surface of MWCNTs was modified in order to develop hydrogen bonding between PI and MWCNTs. Hydroxyl and carboxylic groups were induced on the surface of MWCNTs by direct irradiation with a ⁶⁰Co γ-ray (1.33 MeV per an atom) under a dry O₂ atmosphere. Two different irradiation rates, i.e. 5 and 15 kGy h⁻¹ were used for a period of two hours on the surface of MWCNTs for giving a total irradiation doses of 10 and 30 kGy, respectively. The irradiation process was carried out at the Korea Atomic Energy Research Institute using high-level γ-ray irradiation equipment.⁶³

Synthesis of poly(amic acid) (PAA)

High molecular weight PAA was prepared by a conventional two step method. The diamine monomer 4-(bis(4-aminophenyl)methyl)phenol⁶⁴ (23.74 mmol) having pendant hydroxyl group was dissolved in DMAc (50 mL). A stoichiometric amount of PMDA (23.74 mmol) was added to the solution of diamine monomer. An increase in the viscosity was observed at the advent of polymerization reaction. In order to dilute solution and facilitate stirring, DMAc (66 mL) was added further. The polymerization was carried out at 0–5 °C in a sealed glove box for 18 hours, which resulted in the preparation of poly(amic acid) (PAA). PAA solution was preserved at 0–5 °C for further use.

Preparation of PI/γ-MWCNTs nanocomposites

A measured amount of γ-ray irradiated MWCNTs was added in DMAc (1 mg/0.5 mL), followed by sonication for 1 h with high power ultrasonicator. Afterwards, a calculated amount of PAA solution (7.3 mL) was added into MWCNTs suspension. The PAA–MWCNTs mixture was stirred for 12 h at room temperature and then sonicated for another 0.5 h. Flexible free standing films were casted from PAA–MWCNTs mixture using the solvent elution technique. The solution was transferred to Teflon mold and heated at 70 °C for 12 h, resulting PAA–MWCNTs films were thermally imidized to fabricate PI/γ-MWCNTs nanocomposites by successive heating at 100 °C, 200 °C, 300 °C each for 1 h. A series of PAA–MWCNTs mixtures containing different γ-ray irradiated MWCNTs contents (5 wt%, 10 wt%, 15 wt%, 20 wt%) were prepared using the above mentioned method. Recipe and route for synthesis of PI/γ-MWCNTs is furnished in Scheme 1 and Table 1, respectively.

Scheme 1 Synthesis of PI/γ-MWCNTs nanocomposites.

Table 1 Compositions of PI/γ-MWCNTs nanocomposite films

Nanocomposites	Codes	γ-ray irradiate MWCNTs (%)
Pristine PI	PIneat	0
PI/γ-MWCNTs	PI–γCNT5	5
	PI–γCNT10	10
	PI–γCNT15	15
	PI–γCNT20	20

---

Characterization and instruments

The dispersion state of MWCNTs in PI matrix was observed by Hitachi S-4800 scanning electron microscope. Prior to analysis, the specimen were cryo-fractured and sputter-coated with 10 nm platinum coating. SEM images were captured with a beam voltage of 10 keV using In-lens detector. Thermogravimetric analysis (TGA) of composites was carried out from 30 °C to 1000 °C on TG 209 F3 instrument with a heating rate of 10 °C min⁻¹ under the nitrogen atmosphere. A thermo-mechanical analyzer (TMA-Q400, TA Inc.) was used to measure visco-elastic properties of the composite films. The measurements were performed with tension mode in the temperature range of 30–500 °C using 5 °C min⁻¹ scan rate and 1 Hz oscillation frequency. The vacuum dried samples of 1 mm thickness, 2–3 mm width, and 24 mm length were used for the analysis. Dielectric constant and dielectric loss of PI/γ-MWCNTs nanocomposite films were measured using Agilent 16451B dielectric text fixture. A Shenzen sans tensile testing machine was used to measure tensile properties. The samples (3–4 mm width, 20 mm length) were dried under vacuum at 100 °C for 24 h. Analysis was performed at extension speed of 5 mm min⁻¹ with 10 kN load cell using ASTM D638 standard.

Results and discussion

Microstructural analysis of PI–MWCNs nanocomposites

The microstructure of the fractured surfaces of PI/γ-MWCNTs nanocomposites reinforced with 10 wt% (Fig. 1a and b) and 20 wt% (Fig. 1c and d) is shown in Fig. 1a–d. FE-SEM micrographs (Fig. 1a and c, PI–γCNT₁₀ and PI–γCNT₂₀ at 2 μm, respectively) reveal that the γ-ray irradiated MWCNTs are uniformly dispersed in PI matrix. Furthermore, it is also evident from Fig. 1b and d (PI–γCNT₁₀ and PI–γCNT₂₀ at 500 nm, respectively) that every single nanotube is completely wrapped by PI matrix. Uniform dispersion of MWCNTs and their lamination by PI matrix can be attributed to increased interfacial interaction via hydrogen bonding between the matrix and filler. The hydrogen bonding of MWCNTs with PI matrix isolated them from each other and prevented their aggregation. In a number of previously published reports,^1,43,61 the uniform dispersion of MWCNTs can be achieved until a critical value of filler contents after which the forces of attraction between MWCNTs dominated over interfacial interactions with PI matrix resulting in the aggregation of nanotubes. Whereas, in our work hydrogen bonding between pendant phenyl hydroxyl groups and hydroxyl groups of γ-ray irradiated MWCNTs dominated their cohesive forces impeding their aggregation.

Fig. 1 FE-SEM micrographs (a) and (b) PI–γCNT₁₀, (c) and (d) PI–γCNT₂₀.

Dielectric properties of PI–MWCNTs nanocomposites

The dielectric constant of prepared PI/γ-MWCNTs nanocomposites was evaluated as a function of MWCNTs contents (5–20 wt%) and frequency (1–1000 kHz). Variations in the values of dielectric constant as a function of MWCNTs contents are depicted in Fig. 2. It is obvious from Fig. 2 that integration of γ-ray irradiated MWCNTs improved dielectric constant from 4.72 for neat PI to 36.25 at 20 wt% MWCNTs loading. The linear increase of dielectric constant of composites with increasing MWCNTs contents can be explained on the basis of a micro-capacitor network model.^14,35,65 The adjacent nanotubes, which are completely wrapped by PI matrix, may act as electrodes while the PI bed between them may be considered as the dielectric layer, furnishing various micro-capacitor in the PI–MWCNTs nanocomposites. With an increasing percentage of filler the number of micro-receptor increases, whereas the dielectric layer thickness decreases, leading to enhanced capacitance and an increased dielectric constant. Similarly, the dielectric loss of PI/γ-MWCNTs varies between 0.009–0.29. With increasing frequency, the dielectric loss increases, however, it still remained low as compared to traditional polymer based composites.^34,37 Our PI–MWCNTs nanocomposites have practical applications up to 15% loading of modified MWCNTs where the value of dielectric loss is under 0.1. However, at 20% loading of filler dielectric loss has a value of 0.29. Zhou et al.⁵⁰ reported that the value of dielectric loss can be controlled by coating MWCNTs with non-conducting emeraldine base prior to dispersion in the polymer matrix. In future, we have plan to use this core–shell strategy for treating MWNTs so that the practical applications of PI–MWCNTs can be extended up to 20% loading of filler.

Fig. 2 Variations in dielectric constant and dielectric loss of PI/γ-MWCNTs nanocomposite films as a function of MWCNTs contents (%), measured at 1 kHz and room temperature.

The plot of dielectric constant at different frequencies (1, 10, 100, and 1000 kHz) versus MWCNTs contents is shown in Fig. 3. The values of dielectric constant for pristine PI vary as 4.72–4.41 in the applied frequency range. It is obvious from Fig. 3 that for all nanocomposite films the dielectric constant increase with further MWCNTs addition, however, for all MWCNTs concentrations, the dielectric constant decreases with increasing frequency. This later phenomenon becomes more prominent at high MWCNTs. It can be concluded from these results that the dielectric properties of PI/γ-MWCNTs nanocomposites can be controlled by adjusting the PI/γ-ray irradiated MWCNTs ratio, which indicates the potential of these nanocomposites as promising dielectric material.

Fig. 3 Variations in dielectric constant with MWCNTs (%) at different frequencies.

Visco-elastic properties of PI–MWCNTs nanocomposites

Hydroxyl groups were furnished on the surface of MWCNTs and PI backbone aiming for development of hydrogen bonding, and thereby, for maximum load transfer between two components. The temperature variations of the storage modulus (E′) for PI/γ-MWCNTs nanocomposites having 5–20 wt% γ-ray irradiated MWCNTs are displayed in Fig. 4, whereas, Table 2 presents E′ values at two different temperature i.e. at 30 °C and at T_g of respective composite film, similarly, glass transition temperatures of all nanocomposites are also shown in Table 2. It can be interpreted from Fig. 4 that incorporation of carbon nano-filler enhanced the value of E′ for composites as compared to neat PI in TMA measurement range (30–500 °C). For instance, the value that was 3170 MPa at 30 °C for PI_neat raised to 3549 MPa at 5 wt% loading of MWCNTs, while at 20 wt% loading, i.e. for PI–γCNT₂₀, it reaches to 4687 MPa. The value of E′ for all the films decreased rapidly near the T_g. Again, in the rubbery plateau, i.e. the region above the T_g, the nanocomposites demonstrated higher values of E′ as compared to pristine PI. Neat PI has E′ of 327 MPa at T_g, this value increases with increase in MWCNTs contents and PI–γCNT₂₀ exhibited maximum value of 1917 MPa. Thus, it can be concluded that the significantly enhanced interfacial interaction of γ-ray irradiated MWCNTs with PI matrix through hydrogen bonding hindered the segmental motion of matrix material, consequently, raising the E′ profile both in glassy and rubbery plateau.⁵⁶ The variations in tan δ as a function of temperature are shown in Fig. 5. These tan δ curves illustrate the effect of improved interfacial interaction between PI and γ-ray irradiated MWCNTs on the cooperative movement of PI chains and damping behavior during glass transition. It can be noted from Fig. 5 that with the addition of CNTs the curves are broadened and their maxima move consistently to higher temperature. As the filler contents increase from 0–20 wt%, T_g shifts from 354 °C for neat PI to 437 °C for PI–γCNT₂₀. The drift in the T_g to higher temperature can be attributed to the role of hydrogen bonding in the chain segmental motions. Due to enhanced interfacial bonding between matrix and filler the three dimensional network needs a higher temperature for commencement of chain segmental motion as compared to the pristine PI, consequently, shifting the T_g to higher value.

Fig. 4 Variation in storage modulus with temperature for PI/γ-MWCNTs nanocomposites.

Table 2 Visco-elastic properties of PI–MWCNTs nanocomposites

Nanocomposites	Tg	E′ at 30 °C	E′ at Tg
PIneat	354	3170	327
PI–γCNT5	372	3549	400
PI–γCNT10	381	4106	470
PI–γCNT15	412	4244	706
PI–γCNT20	437	4687	1917

---

Fig. 5 Dependence of tan δ on temperature for PI/γ-MWCNTs nanocomposites.

Thermal properties of PI–MWCNTs nanocomposites

The thermal properties of nanocomposite films were analyzed using thermogravimetric analysis (TGA) to evaluate the effect of hydrogen bonding between γ-ray irradiated MWCNTs and PI matrix on thermal properties of resulting nanocomposite films. The thermograms of PI/γ-MWCNTs nanocomposites reinforced with 5–20 wt% γ-ray irradiated MWCNTs are shown in Fig. 6 and some extracted data is provided in Table 3. Thermal stability was investigated in term of temperature at 10% weight loss (T₁₀) and residual mass at 800 °C (R₈₀₀). In case of neat PI significant weight loss was not noted before 530 °C, however, after that temperature a rapid degradation was observed. As obvious from Fig. 6, the addition of γ-ray irradiated MWCNTs improved the thermal stability and shifted the degradation to higher temperature. All the nanocomposite films have higher value of T₁₀ as compared to neat PI. The value of T₁₀ is 514 °C in the case of neat PI, with the addition of γ-ray irradiated MWCNTs, this value keeps on increasing and reaches to 564 °C at 20 wt% loading of γ-ray irradiated MWCNTs. The residual mass (R₈₀₀) exhibited the same trend, its value increases linearly with the incorporation of γ-ray irradiated MWCNTs. Thus, enhanced thermal properties of nanocomposite films may be attributed to hydrogen bonding between PI matrix and γ-ray irradiated MWCNTs. The significantly improved interfacial bonding resulted in a uniform dispersion of γ-ray irradiated MWCNT, subsequently, a linear increase in thermal stability was observed.

Fig. 6 TGA thermograms of pristine PI and different types of PI/γ-MWCNTs nanocomposites.

Table 3 Thermal properties of PI–MWCNTs nanocomposites

Nanocomposites	T10 (°C)	Residual mass (R800)
PIneat	514	46
PI–γCNT5	538	51
PI–γCNT10	553	56
PI–γCNT15	559	61
PI–γCNT20	564	67

---

Mechanical properties

In order to effectively evaluate the multifunctional applications of nanocomposite films, their mechanical properties were investigated. Jia et al.⁶¹ reported an increase in the mechanical properties of the PI matrix with an increase in CNTs contents. However, interfacial interactions, and hence, control of CNTs aggregation plays a key role in improving the tensile strength.^66,67 Fig. 7 furnishes the typical stress–strain curves for the prepared PI/γ-MWCNTs nanocomposites. As the γ-ray irradiated MWCNTs content increases, the tensile strength consistently increases from 91.88 MPa for the neat PI to 122.47 MPa for PI–γCNT₂₀. The gradual increase in tensile strength can be explained on the bases of improved interface interaction between PI matrix and carbon nano-filler. As the strain shifts from lower to higher value, PI coils stretch to accommodate the deformation. This stretching of PI coils creates greater space for hydrogen bonding with nanotubes, resultantly, external load is shifted from continuous low modulus PI phase to the discontinuous high modulus CNTs phase through a shear stress at the CNTs/PI interface.⁵⁶ Thus, composite films get more and more stiffer with higher load bearing tendency in axial direction.^61,68 Hence, the increase in nanotubes contents increases the extent of interface bonding at higher deformations, which ultimately enhances the strength of PI/γ-MWCNTs nanocomposites. However, the presence of MWCNTs decreases the plastic flow of matrix,⁶⁹ as a result elongation at break decreases with increasing filler contents. In spite of decrease in % elongation of the prepared nanocomposites, their mechanical properties can meet the requirements of various practical applications.

Fig. 7 Stress strain curves for PI/γ-ray MWCNTs nanocomposites films.

Conclusions

In conclusion, the γ-ray irradiated MWCNTs (having –OH, and –COOH) were incorporated at different loading level (5, 10, 15 and 20 wt%) in a PI matrix which also possessed pendant hydroxyl groups. The modified MWCNTs and PI matrix were connected with each other due to hydrogen bonding which resulted from presence of –OH groups in both domains. The improved interfacial interactions between filler and matrix resulted in improvement of dielectric, thermal, mechanical, and dynamic mechanical properties. These PI–MWCNTs nanocomposites have a value of dielectric loss lower than 0.1 at 5–15% MWCNTs loading demonstrating their potential for practical application, for example in the area of embedded capacitors. However, at 20% loading of modified MWCNTs the value of dielectric loss raised to 0.29 limiting their practical application at this level of filler loading.

Acknowledgements

This research was supported by a grant (10037689) from the Fundamental R&D Program for Technology of World Premier Materials (WPM) funded by the Ministry of Trade, Industry and Energy (MOTIE, Korea) and Higher Education Commission of Pakistan.

References

1. Y. Chen, B. Lin, X. Zhang, J. Wang, C. Lai, Y. Sun, Y. Liu and H. Yang, J. Mater. Chem. A, 2014, 2, 14118–14126.
2. Z.-M. Dang, J.-K. Yuan, S.-H. Yao and R.-J. Liao, Adv. Mater., 2013, 25, 6334–6365.
3. Z.-M. Dang, J.-K. Yuan, J.-W. Zha, T. Zhou, S.-T. Li and G.-H. Hu, Prog. Mater. Sci., 2012, 57, 660–723.
4. W. Xu, Y. Ding, S. Jiang, J. Zhu, W. Ye, Y. Shen and H. Hou, Eur. Polym. J., 2014, 59, 129–135.
5. Y. Zhang, Y. Wang, Y. Deng, M. Li and J. Bai, ACS Appl. Mater. Interfaces, 2012, 4, 65–68.
6. R. Pelrine, R. Kornbluh, Q. Pei and J. Joseph, Science, 2000, 287, 836–839.
7. Q. M. Zhang, H. Li, M. Poh, F. Xia, Z. Y. Cheng, H. Xu and C. Huang, Nature, 2002, 419, 284–287.
8. W. Xu, Y. Ding, S. Jiang, L. Chen, X. Liao and H. Hou, Mater. Lett., 2014, 135, 158–161.
9. R. P. Pandey, A. K. Thakur and V. K. Shahi, ACS Appl. Mater. Interfaces, 2014, 6, 16993–17002.
10. T. Sheng, H. Chen, S. Xiong, X. Chen and Y. Wang, AIChE J., 2014, 60, 3614–3622.
11. W.-H. Liao, S.-Y. Yang, S.-T. Hsiao, Y.-S. Wang, S.-M. Li, C.-C. M. Ma, H.-W. Tien and S.-J. Zeng, ACS Appl. Mater. Interfaces, 2014, 6, 15802–15812.
12. H. J. Kim, K. Choi, Y. Baek, D.-G. Kim, J. Shim, J. Yoon and J.-C. Lee, ACS Appl. Mater. Interfaces, 2014, 6, 2819–2829.
13. X. Wang, W. Li, L. Luo, Z. Fang, J. Zhang and Y. Zhu, J. Appl. Polym. Sci., 2012, 125, 2711–2715.
14. Z. M. Dang, L. Wang, Y. Yin, Q. Zhang and Q. Q. Lei, Adv. Mater., 2007, 19, 852–857.
15. Z. Liu, G. Jiao, X. Chao and Z. Yang, Mater. Res. Bull., 2013, 48, 4877–4883.
16. Z. Hu, B. Cui, M. Li and L. Li, J. Mater. Sci.: Mater. Electron., 2013, 24, 3850–3855.
17. C. Huang and Q. Zhang, Adv. Funct. Mater., 2004, 14, 501–506.
18. F. He, S. Lau, H. L. Chan and J. Fan, Adv. Mater., 2009, 21, 710–715.
19. M. Rahaman, T. K. Chaki and D. Khastgir, Eur. Polym. J., 2012, 48, 1241–1248.
20. J. Chen, X. Wang, W. Jo and J. Rödel, J. Am. Ceram. Soc., 2010, 93, 1692–1696.
21. C. C. Lin, Y. H. Wu, R. S. Jiang and M. T. Yu, IEEE Electron Device Lett., 2013, 34, 1418–1420.
22. X. Wang, M. Li, C. Zhou, M. Wang and W. Lu, Ceram. Int., 2014, 40, 39–45.
23. W.-Q. Luo, Z.-Y. Shen, Y.-M. Li, Z.-M. Wang, R.-H. Liao and X.-Y. Gu, J. Electroceram., 2013, 31, 117–123.
24. X. Peng, Q. Wu, S. Jiang, M. Hanif, S. Chen and H. Hou, J. Appl. Polym. Sci., 2014, 131, 10,  DOI:10.1002/app.40828.
25. M. Arbatti, X. Shan and Z. Y. Cheng, Adv. Mater., 2007, 19, 1369–1372.
26. B. Chu, X. Zhou, K. Ren, B. Neese, M. Lin, Q. Wang, F. Bauer and Q. M. Zhang, Science, 2006, 21, 334–336.
27. Y. Wang, X. Zhou, Q. Chen, B. Chu and Q. Zhang, IEEE Trans. Dielectr. Electr. Insul., 2010, 17, 1036–1042.
28. E. Tuncer, I. Sauers, D. R. James, A. R. Ellis, M. P. Paranthaman, A. Goyal and K. L. More, Nanotechnology, 2007, 18, 325704.
29. Y. Li, X. Huang, Z. Hu, P. Jiang, S. Li and T. Tanaka, ACS Appl. Mater. Interfaces, 2011, 3, 4396–4403.
30. C. L. Dube, S. C. Kashyap, D. K. Pandya and D. C. Dube, Phys. Status Solidi A, 2009, 206, 2627–2631.
31. X. Huang, Z. Pu, L. Tong, Z. Wang and X. Liu, J. Mater. Sci.: Mater. Electron., 2012, 23, 2089–2097.
32. C. Min, D. Yu, J. Cao, G. Wang and L. Feng, Carbon, 2013, 55, 116–125.
33. X. Liu, J. Yin, Y. Kong, M. Chen, Y. Feng, Z. Wu, B. Su and Q. Lei, Thin Solid Films, 2013, 544, 54–58.
34. C. Wu, X. Huang, X. Wu, L. Xie, K. Yang and P. Jiang, Nanoscale, 2013, 5, 3847–3855.
35. M. Li, X. Huang, C. Wu, H. Xu, P. Jiang and T. Tanaka, J. Mater. Chem., 2012, 22, 23477–23484.
36. Y. Zhang, L. Li, B. Wang, J. Zhang and E. Wang, J. Mater. Sci.: Mater. Electron., 2014, 25, 805–810.
37. L. L. Sun, B. Li, Y. Zhao, G. Mitchell and W. H. Zhong, Nanotechnology, 2010, 21, 305702.
38. L. Yu, Y.-H. Zhang, J. Shang, S.-M. Ke, W.-s. Tong, B. Shen and H.-T. Huang, J. Electron. Mater., 2012, 41, 2439–2446.
39. J.-W. Ha, Y. Kim, J. Roh, F. Xu, J. Il Park, J. Kwak, C. Lee and D.-H. Hwang, J. Polym. Sci., Part A: Polym. Chem., 2014, 52, 3260–3268.
40. B. W. Jo, K. H. Ahn and S. J. Lee, Polymer, 2014, 55, 5829–5836.
41. X. Wang, Y. Dai, W. Wang, M. Ren, B. Li, C. Fan and X. Liu, ACS Appl. Mater. Interfaces, 2014, 6, 16182–16188.
42. C. Yuan, J. Wang, K. Jin, S. Diao, J. Sun, J. Tong and Q. Fang, Macromolecules, 2014, 47, 6311–6315.
43. M. Ding, Prog. Polym. Sci., 2007, 32, 623–668.
44. G. Maier, Prog. Polym. Sci., 2001, 26, 3–65.
45. C. Lee, Y. Shul and H. Han, J. Polym. Sci., Part B: Polym. Phys., 2002, 40, 2190–2198.
46. A. Alias, Z. Ahmad and A. B. Ismail, Mater. Sci. Eng., B, 2011, 176, 799–804.
47. Q. Chi, J. Sun, C. Zhang, G. Liu, J. Lin, Y. Wang, X. Wang and Q. Lei, J. Mater. Chem. C, 2014, 2, 172–177.
48. L. Chu, Q. Xue, J. Sun, F. Xia, W. Xing, D. Xia and M. Dong, Compos. Sci. Technol., 2013, 86, 70–75.
49. N. Ning, X. Bai, D. Yang, L. Zhang, Y. Lu, T. Nishi and M. Tian, RSC Adv., 2014, 4, 4543–4551.
50. T. Zhou, J.-W. Zha, Y. Hou, D. Wang, J. Zhao and Z.-M. Dang, ACS Appl. Mater. Interfaces, 2011, 3, 4557–4560.
51. Y.-n. Mi, G. Liang, A. Gu, F. Zhao and L. Yuan, Ind. Eng. Chem. Res., 2013, 52, 3342–3353.
52. B. Wang, D. Qin, G. Liang, A. Gu, L. Liu and L. Yuan, J. Phys. Chem. C, 2013, 117, 15487–15495.
53. H. Wu, A. Gu, G. Liang and L. Yuan, J. Mater. Chem., 2011, 21, 14838–14848.
54. X. Zhang, G. Liang, J. Chang, A. Gu, L. Yuan and W. Zhang, Carbon, 2012, 50, 4995–5007.
55. G. Fabia, B. Sabrina, C. Valter and G. Giuseppe, Smart Mater. Struct., 2013, 22, 055025.
56. M. Yoonessi, M. Lebrón-Colón, D. Scheiman and M. A. Meador, ACS Appl. Mater. Interfaces, 2014, 6, 16621–16630.
57. J. Ma and R. M. Larsen, ACS Appl. Mater. Interfaces, 2013, 5, 1287–1293.
58. C.-H. Tseng, C.-C. Wang and C.-Y. Chen, Chem. Mater., 2006, 19, 308–315.
59. J. Nanda, C. Maranville, S. C. Bollin, D. Sawall, H. Ohtani, J. T. Remillard and J. M. Ginder, J. Phys. Chem. C, 2007, 112, 654–658.
60. S. G. Miller, T. S. Williams, J. S. Baker, F. Solá, M. Lebron-Colon, L. S. McCorkle, N. G. Wilmoth, J. Gaier, M. Chen and M. A. Meador, ACS Appl. Mater. Interfaces, 2014, 6, 6120–6126.
61. X. Jia, Q. Zhang, M.-Q. Zhao, G.-H. Xu, J.-Q. Huang, W. Qian, Y. Lu and F. Wei, J. Mater. Chem., 2012, 22, 7050–7056.
62. J. Wang, Q. Jiao, H. Li, Y. Zhao and B. Guo, Polym. Compos., 2014, 35, 1952–1959.
63. M. Kim, S. C. Mun, C. S. Lee, M. H. Lee, Y. Son and O. O. Park, Carbon, 2011, 49, 4024–4030.
64. T. Akhter, H. Siddiqi, S. Saeed, O. O. Park and S. Mun, J. Polym. Res., 2014, 21, 1–10.
65. X. Huang, P. Jiang and Y. Yin, Appl. Phys. Lett., 2009, 95, 242905.
66. A. B. Dalton, S. Collins, E. Munoz, J. M. Razal, V. H. Ebron, J. P. Ferraris, J. N. Coleman, B. G. Kim and R. H. Baughman, Nature, 2003, 423, 703.
67. Y. C. Jung, D. Shimamoto, H. Muramatsu, Y. A. Kim, T. Hayashi, M. Terrones and M. Endo, Adv. Mater., 2008, 20, 4509–4512.
68. Y. Li and H. Shimizu, Macromolecules, 2009, 42, 2587–2593.
69. S.-M. Yuen, C.-C. M. Ma, C.-L. Chiang, C.-C. Teng and Y.-H. Yu, J. Polym. Sci., Part A: Polym. Chem., 2008, 46, 803–816.

---