Effect of cross-linking degree of EPDM phase on the electrical properties and formation of dual networks of thermoplastic vulcanizate composites based on isotactic polypropylene (iPP)/ethylene–propylene–diene rubber (EPDM) blends

Abstract

Thermoplastic vulcanizates (TPVs), as a special class of high-performance thermoplastic elastomers, have been widely used in the automotive industry, building, and electronics due to their good processability and recyclability. Here, the electrical performance and the formation of dual networks of TPV composites filled with carbon black (CB) based on isotactic polypropylene/ethylene–propylene–diene rubber (iPP/EPDM) blends was investigated by varying the content of the curing agent, phenolic resin (PF). With the incorporation of 6 wt% PF, the crosslinking degree of the EPDM phase reaches a high value of 47.4 wt% and the domains reach the smallest size. The electrical percolation threshold of TPV/CB composites decreases as the cross-linking degree increases and at last maintains a steady value of 13.9 wt%. The morphological structure, dynamic rheology behaviors and crystallization behaviors of TPV/CB composites were characterized to explain the selective dispersion of CB particles and the microstructure evolution of TPV/CB composites. With an increase in the curing degree of the EPDM phase, a denser dual network including the CB conductive network and the network of EPDM particles is formed in the iPP matrix. This study provides an effective strategy to realize the control of electrical properties and the formation of dual networks of TPV composites, and can be easily introduced into industrial applications.

---

Introduction

Conductive polymer composites (CPCs) have received extensive technological and academic interest because of their good thermal stability, chemical stability, and dimensional stability.^1,2 Through incorporation of conductive fillers into the polymer matrix including rubbers and thermoplastics, the characteristics of the polymer matrix can be combined with the mechanical and functional properties of the fillers. Among various fillers for polymer composites, carbon black (CB) is one of the most common fillers widely employed to improve the electrical conductivity and mechanical properties of polymers due to its mechanical strength, chemical stability and abundant source.^3–6

As is well-known, only when a continuous conducting path is formed above the critical concentration of the conductive fillers, i.e., the percolation threshold, an electrical conductive polymer material can be obtained.^7,8 In fact, the formation of conductive network in an electrical conductive composite is affected not only by the filler volume fraction, but also by mixing methods, inherent physical properties of polymer matrix loaded with conductive particles (e.g. crystalline behaviour and viscosity ratio of the polymer blends), utilization of a third component (e.g. surfactant, compatibilizer, etc.), and other means (e.g. in situ polymerization, mixing with polymer latex, etc.).^9–12 Pötschke and co-workers¹⁰ studied the effect of viscosity ratio of polycarbonate (PC)/poly(styrene-acrylonitrile) (SAN) composites, with multi-walled carbon nanotubes (MWCNTs) selectively localized in PC component, on the morphological and electrical properties, and found that the MWCNT dispersion was improved with decreasing PC viscosity. The highest electrical conductivity was achieved at the lowest PC/SAN viscosity ratio. Liu et al.¹² studied the effect of crystalline behaviour of isotactic polypropylene (iPP) on the formation of CB conductive network and found that CB particles were rejected to the amorphous region or the interlamellar region of spherulites followed by the completion of crystallization process, and formed an intact conductive network. Also the electrical property can be enhanced due to the improved crystalline behaviour during the isothermally crystallization process at a certain temperature.

Thermoplastic vulcanizates (TPVs), as a group of high performance thermoplastic elastomers prepared by dynamic vulcanization, consists of a high content of crosslinked rubber as the dispersed phase and a low content of thermoplastics as the continuous phase.^13–15 TPVs combine the excellent resilience of conventional vulcanized elastomers and the good processability and recyclability of thermoplastics. Therefore, TPVs have become one of the fastest growing elastomers because of the requirements of environmental protection and resource saving. Recently, to expand practical applications of TPVs, much effort has been paid to the development of conductive TPVs materials. However, to date, much attention has been devoted to the effect of content of conductive fillers on the morphology and electrical properties of TPV composites,^16–19 and little attention has been paid to the evolution of electrical properties, dispersion of CB particles, and conductive network formation of TPV/CB composites at a constant CB loading but different curing degree of the rubber phase.

In our previous study,^20,21 we revealed that for iPP/ethylene–propylene–diene rubber (EPDM) blends, CB particles are dispersed in EPDM phase, while for TPV/CB composites, CB particles are selectively dispersed in iPP matrix and dual networks of CB and EPDM particles owing to dynamic vulcanization are constructed. In this study, inspired by the morphology evolution and selectively dispersion of CB particles caused by dynamic vulcanization, we aimed to reveal the effect of cross-linking degree of EPDM phase on the electrical properties and the dual network formation of TPV/CB composites with iPP as the thermoplastic matrix.

Experimental

Materials

A commercial iPP (T30s, MFR = 2.3 g/10 min at 230 °C and 2.16 kg load), was purchased from Lanzhou Petroleum Chemical Co, Ltd., China. Amorphous EPDM (Nordel IP 4725P) was from Dow Elastomers Co, Ltd, Wilmington, DE, with 4.9 wt% ethylidene norbornene (ENB). CB particles (VXC 68, Cabot) have an average primary particle size of 25 nm and a dibutyl phthalate volume of 1.23 ml g⁻¹. PF, a kind of brominated phenolic formaldehyde resin with the trademark TXL-201, used as the curing agent, were obtained from Yuantai Biochemistry Industry Company, China.

Sample preparation

The melt-reactive blending process for preparing TPV samples (iPP : EDPM = 50 : 50 wt%) was carried out in an XSS-300 torque rheometer (Shanghai Kechuang Rubber Plastics Machinery Set Ltd, China) with a rotor speed of 75 rpm and a set temperature of 200 °C. iPP and EPDM were first added in the mixer, and after 3 min, PF (0 wt%, 2 wt%, 4 wt% and 6 wt% to the total weight of the blends) was added and melt-reactive blending was continued for another 5 min. At last, CB was introduced and mixed for another 8 min, after which the mixture was taken out and cut up. For the sake of brevity, iPP/EPDM blends filled with CB particles were named as TPE-x, where x stands for CB content by weight and TPV composites with varying CB loading, were labelled as TPVa-b, where a represents the PF dosage by weight, and b stands for the CB content by weight. The mixtures were compression-molded into sheets with a thickness of about 1.0 mm at 200 °C for 10 min under a pressure of 10 MPa.

Gel content

The gel contents of samples were determined by extraction of 0.3 g of powdered sample through a 120 mesh stainless steel pouch in boiling xylene, in accordance with ASTM D-2765.

Fourier transformed infrared spectroscopy (FTIR) analysis

Several grams of the residue of TPV samples for gel content tests and TPE sample were compression molded into thin films between aluminum sheets at 200 °C under 10 MPa. FTIR spectra were determined on a Nicolet 6700 FTIR spectrometer (Nicolet Instrument Company, USA) and FTIR-attenuated total reflection spectra were recorded from 650 to 4000 cm⁻¹ by averaging 32 scans at a resolution of 2 cm⁻¹.

Electrical conductivity measurements

Electrical conductivity was measured by a two-probed method using a digital multi-meters (6517B, Keithley Instruments, Inc, Ohio, USA) when the volume resistivity was below 10⁸ Ω cm. Copper paste was used to ensure good contact between the sample surface with the electrodes. A high resistivity meter (ZC36, Shanghai Precision Instruments Co., Ltd., China) was used when the volume resistivity of the samples was beyond 10⁸ Ω cm. The sample dimensions for high and low volume resistivity measurements were 3 × 50 × 50 mm³ and 3 × 12 × 120 mm³, respectively.

Morphological observation

The phase morphologies were characterized with a JEOL JSM-5900LV scanning electron microscope (SEM, Japan) at an accelerating voltage of 20 kV. The compression-molded samples were frozen in liquid nitrogen for 30 min and then cryogenically fractured in liquid nitrogen. Furthermore, transmission electron microscopy (TEM, Philips CM120) was also used to characterize the morphologies and distribution of CB particles. Before observation, the samples were cryomicrotomed into 100 nm-thick sections at −130 °C and then vapor stained with ruthenium tetroxide for 30 min.

Quantitative analysis of the morphology was performed using image analysis of Image-Pro Plus 6. At least 400 dispersed domains were measured by manually tracing the phase boundaries to estimate the number-average diameter for each sample.

Rheological characterization

The rheological measurements were performed using an AR2000ex stress-controlled dynamic rheometer (TA Corporation, USA) equipped with a parallel-plate geometry. Dynamic strain sweep was carried out in the strain range of 0.001–100% with a frequency of 1 Hz at 200 °C.

Differential scanning calorimetry (DSC)

The crystallization and melting behaviours of the composites were measured using a differential scanning calorimeter (DSC, TA Q20). 3–5 mg samples were heated up to 200 °C rapidly under a nitrogen atmosphere and held at 200 °C for 5 min to eliminate the thermal history. Afterwards, the samples were cooled to 40 °C at a rate of 10 °C min⁻¹ to record the crystallization behaviour, and then heated again to 200 °C at a heating rate of 10 °C min⁻¹ to record the melting behaviour.

Results and discussion

Dynamic vulcanization

The FTIR spectra of TPE and TPV samples without the loading of CB particles after extraction using xylene steam in Soxhlet extractor for 48 h are shown in Fig. 1(a). The characteristic absorption bands at 2965 cm⁻¹, and 2916, 2849, 1465 and 1374 cm⁻¹ and 1712 cm⁻¹ can be assigned to the –CH₃, –CH₂– for PP and EPDM molecular chains and C=C bands for ethylidene norbornene (ENB) for EPDM.²² For sample TPE, the absorption bands at 1650 cm⁻¹ assigned to C=C bond appears in the form of a tiny broad peak, which should be ascribed to the ultra-low content of ENB (only 4.9 wt%) in EPDM phase. Compared with TPE sample, the absorption bands attributed to C=C bond in TPVs samples disappears for all the samples with different content of curing agent, indicating the complete dissolution of iPP and unvulcanized EPDM. Subsequently, the residual samples after extraction using Soxhlet extractor were used to further characterize the gel content and to elucidate the degree of cross-linking, as shown in Fig. 1(b). It can be observed that the gel content increases with PF content increasing from 0 to 6 wt%. The gel content for TPE is only 0.18 wt%, while for TPV samples, with the PF content increasing, the gel content sharply increases to 28.9 wt% for TPV2, and then gradually increases to 47.4 wt% for TPV6. Obviously, EPDM phase can be effectively cured with the addition of PF resin and with increasing PF loading, the cross-linking degree increases gradually. It is well known that EPDM with ENB as the diene monomer is more reactive toward resole cross-linking than that with dicyclopentadiene (DCPD), 1,4-hexa-diene (HD), or vinylidene norbornene (VNB).^23,24 The detailed curing mechanism has been reported in previous work.²¹ The resole degrades into benzyl cations followed by the diastasis of bromide ion, which eventually reacts with the unsaturated bonds of EPDM and connects EPDM chains via chroman or methylene-bridged structures.

Fig. 1 FTIR spectra of TPE and residual TPV samples after extraction using xylene steam in a Soxhlet extractor (a) and the effect of PF content on the gel content of TPVs (b).

Morphology evolution

To explore the phase structure and dispersion of CB particles of TPE and TPV composites with different cross-linking degree of EPMD phase, SEM and TEM images of these composites filled with 20 wt% CB particles are shown in Fig. 2. According to our previous work,^20,21 it is reasonable to believe that region A and B represent iPP and EPDM phase, respectively. For TPE20 composites, co-continuous phase morphology can be observed and CB particles are almost completely located in EPDM phase and are fused together into CB aggregates. However, for TPV20 composites, EPDM phase was stained with ruthenium tetroxide and is shown as the gray area, while iPP phase is shown as the white area. Magnified images for these composites show that there is almost no CB particles dispersed in EPDM phase for all the TPV composites with 20 wt% CB particles. Further observation shows that there is an obvious phase incorporative phenomenon for the dispersed EPDM phase in TPV2-20 composite. Also, CB particles exist in the form of small aggregates on the surface EPDM domains and the integrated CB particle network is not totally distributed in iPP matrix. It is known that owing to the strong interaction between EPDM and CB particles, CB particles prefer to disperse in EPDM phase for TPE composites.²⁰ When PF content is 2 wt%, the gel content for TPV2-20 composite is only 28.9 wt%, indicating not all EPDM phase has been cured, so the distribution state of CB particles and the conductive network formed should be attributed to the low cross-linking degree and high viscosity of EPDM phase caused by dynamic vulcanization.²⁵ In this case, CB particles still tend to be dispersed in EPDM phase which is driven by the strong interaction between EPDM and CB particles, while the high viscosity of EPDM restricts the migration of CB particles into iPP phase. All these factors result in the dispersion of CB aggregates on the surface of partially cured EPDM particles. However, with increasing PF loading from 4 wt% to 6 wt%, CB particles for TPV4-20 and TPV6-20 composites are all located in iPP phase and form the conductive network in iPP matrix and the CB particle network for TPV6-20 composite is more integrated than that for TPV4-20 composite and the CB aggregates can effectively contact with each other to form a more integrated network seen in enlarged images shown in Fig. 2. To explore the detailed difference for the phase structure between both series of TPV composites, quantitative analyses of average particle diameter and particle size distribution (seen in Fig. 3 and Table 1) have been characterized and shows that the average particle diameter for EPDM phase in TPV2-20 is about 430 ± 39 nm, while that for TPV4-20 and TPV6-20 are 346 ± 18 and 328 ± 33 nm, respectively, indicating that with increased cross-linking degree, the diameter of EPDM phase domains reaches a steady value for TPV6-20 composites, which is accordance to the previous work.^26,27 Further observation shows that compared with that of TPV4-20 composites, there exists a greater number of EPDM particles with a diameter larger than 650 nm for TPV6-20 composites. According to Tian and Nando et al.,^16,17 during the final stage of dynamic vulcanization process or with a highly cross-linking degree of rubber phase, the dispersed rubber phase is formed in agglomerates of rubber particles with a diameter of about 40–100 nm. Combined with the morphological observation of these series TPV composites, it is reasonable to think that, for TPV6-20 composites, although a smaller EPDM phase in diameter is formed due to a high cross-linking degree,^27,28 the high percentage of EPDM phase with a diameter larger than 650 nm should be attributed to the agglomerates of nanoscale EPDM particles.

Fig. 2 Typical SEM and TEM images of TPE20, TPV2-20, TPV4-20 and TPV6-20 composites.

Fig. 3 Particle size distribution of TPV2-20, TPV4-20 and TPV6-20 composites.

Table 1 The average particle size and the detailed values of particle distribution of the series of TPV composites

Code	Average particle size (nm)	Percentage for particles between 0 and 450 nm	Percentage for particles between 450 to 650 nm	Percentage for particles bigger than 650 nm
TPV2-CB	430 ± 39	59.2%	20.1%	20.7%
TPV4-CB	346 ± 18	53.1%	32.6%	14.3%
TPV6-CB	328 ± 33	56.8%	24.3%	18.9%

---

Electrical properties

Based on the above results, it can be observed that the phase structure of TPV composites and the selective dispersion of CB particles change obviously due to the different degree of cross-linking of EPDM phase. Also, it is well-known that the electrical conductivity is closely related with the formation of conductive network induced by the selective dispersion of CB particles for polymer blends, so the electrical property of these TPV composites with different cross-linking degree of EPDM was further studied. Fig. 4(a) shows the room-temperature resistivity of all the composites as a function of CB content. It can be seen that with the incorporation of CB particles, the volume resistivity decreased sharply at a certain concentration of CB particles, indicating the formation of a conductive network. This concentration is usually called the percolation threshold.^29–32 Further observation shows that the conductive performance of TPVs composites is much better than that of TPEs with different loading of CB. Detailed results of the electrical resistance for TPV composites as a function of the gel content are shown in Fig. 4(b). At a constant CB loading, with the cross-linking degree of EPDM phase increasing, the volume resistivity of the TPV composites reduces first and then reaches to a steady value. A power law relation was used to describe the threshold of electrical conductivity percolation:³³

ρ = ρ₀(ω − ω_c)^t (1)

where ρ is the electrical resistivity of conductive composites, ρ₀ is the electrical resistivity of conductive fillers, ω is the mass fraction of CB, ω_c is the threshold of the electrical conductivity percolation, and t is the critical exponent. The conductive network in the composites greatly influences the threshold behaviour of the composites, as shown in Table 2.

Fig. 4 The volume resistivity as a function of the CNTs content (a) and the PF loading (b).

Table 2 The fitted parameters of TPE and TPV conductive composites according to eqn (1)

Code	ωc (wt%)	t	ρ0	R^2
TPE-CB	18.40	3.97	0.11	0.9995
TPV2-CB	15.05	3.44	1.66	0.9997
TPV4-CB	13.74	3.45	1.83	0.9999
TPV6-CB	13.98	3.46	2.94	0.9993

---

The fitted values of t are all around 3.4–4, higher than the universal critical exponents (1.1–2.0) derived from the classical conduction model. The microstructural characteristics (compact three-dimensional conductive network and high specific surface area) of CB are thought to be responsible for the high critical exponent.³⁴ TPE/CB composite shows the highest percolation threshold (ω_c = 18.4 wt%), while the conductive percolation threshold for TPV composites decreases gradually with the incorporation of PF from 0 wt% to 4 wt%. Seen in Table 2, the percolation threshold of TPV4 composites is 13.74 wt%, and further increase the content of PF does not largely influence on the percolation value, and a value of 13.98 wt% is obtained for TPV6 composites. These results indicate that on the one hand, a stable and integrate CB conductive network is constructed when the content of PF increases to 4 wt%. On the other hand, it also means that this kind of conductive network in TPV system is very stable even the PF content further increases.

Evolution of the dual network of TPV/CB composites

In our previous work,²⁰ we used dynamic rheometer to characterize the frequency dependence of tan δ of TPE and TPV samples without the loading of CB particles, and found that there exists a peak of tan δ owing to the EPDM particles, which indicated that the EPDM particles in the micro–nano size has formed an effective particle network. This result is in good agreement with Tian's work.¹⁶ So with the loading of CB particles increasing, when CB loading is larger than the percolation threshold, the CB conductive network is also formed, leading to the dual network for TPV composites.

Fig. 5(a) shows the typical dynamic strain sweep curves to explore the dual network evolution for TPE and TPV composites with increasing cross-linking degree of EPDM phase, where the elastic moduli, G′ are plotted as a function of the strain amplitude at a frequency of 1 Hz. The linear region with a solid-like response at low strains is followed by a decrease of G′ with increasing strain amplitude as the sample starts to yield under oscillatory shear.^35,36 It can be observed that the plateau modulus of G′ increases for TPV composites containing 20 wt% CB with increasing cross-linking degree of EPDM phase. It could be attributed to: (1) a higher elasticity of the network of EPDM phase with the higher curing degree; (2) more compact network of EPDM phase caused by the diameter decrease of EPDM phase with a consistent EPDM content as cross-linking degree of EPDM phase increases; (3) much integrated conductive network with increasing cross-linking degree of EPDM phase.

Fig. 5 (a) Storage modulus (G′), (b) storage modulus (G′)/storage modulus (G′(0)) versus strain for TPE and TPVs composites.

Strain dependence of normalized storage modulus (G′/G′(0)) of the TPV composites are shown in Fig. 5(b). The transition point at which deviation from the linear to non-linear viscoelastic behaviour occurs is defined as the critical strain and it varies with the filler content, dispersion quality and evolution of filler network etc.^37,38 The occurrence of critical strain is due to the breakup of some crucial elastic links in the nanoplatelet network.³⁹ It can be observed from Fig. 5(b) that the critical strain decreases with increasing cross-linking degree of EPDM phase, indicating that there exists a much denser and more fragile dual network of the conductive CB particle network and the network of EPDM phase with cross-linking degree of EPDM phase increasing.⁴⁰

Crystallization behaviors

From the above discussion, it is clear that with increasing cross-linking degree of EPDM phase, the dispersion of CB particles undergoes a series of changes from being on the surface of EPDM phase to be distributed in iPP matrix. As is well known, the dispersed quality of nano fillers notably influences the behaviours of polymer crystallization.⁴¹ Thus, the crystallization behaviours of iPP matrix in these TPV composites with increasing cross-linking degree of EPDM phase are attractive and worth to be investigated. Here, the crystallization exotherms of TPV composites at a cooling rate of 10 °C min⁻¹ were investigated and are shown in Fig. 6. With increasing cross-linking degree of EPDM phase, the crystallization peak temperature (T_p) of iPP causes an augment of ca. 3 °C from TPV2-20 to TPV6-20 composite. Combined with the results seen in Fig. 2, the difference of T_p can be mainly attributed to the distinguishing distribution of CB particles of these TPV composites. For TPV2-20, almost all the CB particles are surrounded on the EPDM particles, showing a lower nucleating efficiency for iPP matrix which leads to a lower T_p. While with the cross-linking degree of EPDM phase increasing, CB particles tend to migrate to iPP matrix and CB particles network is formed for TPV4-20 and TPV6-20 composites (seen in Fig. 2). The more uniform CB particles distributed in iPP phase for TPV6-20 composites play a role of heterogeneous nucleation for iPP matrix, and lead to the elevated crystallization peak temperature.

Fig. 6 DSC cooling curves of samples TPV2-20, TPV4-20 and TPV 6-20 at a cooling rate of 10 °C min⁻¹.

To further confirm the effect of CB particles on the crystallization behaviours of iPP matrix, the isothermal crystallization behaviours of TPV composites were monitored at various crystallization temperatures, and the development of heat flow and relative degree of crystallinity X_c of the crystallites as a function of time are displayed in Fig. 7 and 8, respectively. For brevity, the data of isothermal crystallization at 128 °C and 130 °C are not shown here. X_c is a relative parameter defined as⁴²

where dH/dt is the rate of heat evolution, ΔH_t is the heat generated at time t, and ΔH_∞ is the total heat. As shown in Fig. 7, with increasing cross-linking degree of EPDM phase, the time to reach the maximum peak shifts to a shorter time for TPV2-20 and TPV6-20 composite, which attested the heterogeneous nucleation effect of CB particles on the overall crystallization rate of iPP matrix. In addition, the special concentration dependence can be also observed in Fig. 7 and 8.

Fig. 7 DSC heat flow as a function of time during isothermal crystallization at 126 °C (a) and 132 °C (b).

Fig. 8 Relative crystallinity versus time for TPV composites isothermally crystallized at different temperatures at 126 °C (a) and 132 °C (b).

At the same time, the half-crystallization time, t_0.5, often chosen as a key parameter to characterize the overall crystallization kinetics, was obtained from the curves of X_c(t) at X_c(t) = 50% (seen in Fig. 8) and displayed in Table 3. One can see that with increasing cross-linking degree of EPDM phase, t_0.5 of iPP gradually decreases at each isothermal crystallization temperature, confirming that the nucleating effect of CB particles can be enhanced by the migration of CB particles from the surface of EPDM phase to iPP matrix owing to the increased cross-linking degree of EPDM phase.

Table 3 The half crystallization time (t_0.5) of TPVs composites at various isothermal crystallization temperatures

Tc (°C)	t0.5 (s)
	TPV2-20	TPV4-20	TPV6-20
126 °C	2.5	2.3	1.8
132 °C	10.7	9.2	8.2

---

Conclusions

In this study, the effect of cross-linking degree of EPDM phase on the electrical performance and the formation of dual network of TPV/CB composites were investigated. With the incorporation of PF up to 6 wt%, a smaller EPDM particle is obtained and the curing degree of EPDM increases generally to 47.4 wt%. With increasing cross-linking degree of EPDM phase, CB particle migrates from the surface of EPDM domains to iPP matrix, resulting in the decrease of electrical percolation threshold and at last a steady value of 13.9 wt% CB loading is obtained. Finally, a denser dual network constructed by CB and EPDM particles is formed for TPV6 composites. An enhanced nucleation effect also appears due to the migration of CB particles from the surface of EPDM phase to iPP matrix.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (NNSFC Grants 51422305 and 51421061), the Innovation Team Program of Science & Technology Department of Sichuan Province (Grant 2014TD0002) and the Sichuan Provincial Science Fund for Distinguished Young Scholars (2015JQO003). Lifeng Ma's study abroad at the University of Delaware is supported by the State Scholarship Fund of the China Scholarship Council.

References

1. I. Balberg, Carbon, 2002, 40, 139.
2. Z. Spitalskya, D. Tasisb, K. Papagelisb and C. Galiotisa, Prog. Polym. Sci., 2010, 35, 357.
3. A. Malas and C. K. Das, J. Mater. Sci., 2012, 47, 2016.
4. C. Sandler, M. S. P. Shaffer, T. Prasse, W. Bauhofer, K. Schulte and A. H. Whindle, Polymer, 1999, 40, 5967–5971.
5. X. D. Wu, C. H. Lu, Y. Y. Han, Z. H. Zhou, G. P. Yuan and X. X. Zhang, Compos. Sci. Technol., 2016, 124, 44.
6. H. F. Yang, J. Gong, X. Wen, J. Xue, Q. Chen, Z. W. Jiang, N. N. Tian and T. Tang, Compos. Sci. Technol., 2015, 113, 31.
7. X. F. Sánchez-Romate, A. Jiménez-Suárez, M. Sánchez, A. Güemesb and A. Ureñaa, RSC Adv., 2016, 6, 43418.
8. N. K. Shrivastava, S. Suin, S. Maitia and B. B. Khatua, RSC Adv., 2014, 4, 24584.
9. T. Li, L. F. Ma, R. Y. Bao, G. Q. Qi, W. Yang, B. H. Xie and M. B. Yang, J. Mater. Chem. A, 2015, 3, 5482.
10. M. Liebscher, L. Tzounis, P. Pötschke and G. Heinrich, Polymer, 2013, 54, 6801.
11. R. Dou, Y. Shao, S. L. Li, B. Yin and M. B. Yang, Polymer, 2016, 83, 34.
12. J. Deng, Z. Y. Liu, S. D. Zheng, Y. Z. Huang, W. Yang and M. B. Yang, J. Polym. Res., 2016, 23, 47.
13. Y. K. Chen, Y. H. Wang, C. H. Xu, Y. P. Yang and C. Y. Jiang, Ind. Eng. Chem. Res., 2016, 55, 1701.
14. Y. K. Chen, D. S. Yuan and C. H. Xu, ACS Appl. Mater. Interfaces, 2014, 6, 3811.
15. C. F. Antunes, M. van Duin and A. V. Machado, Mater. Chem. Phys., 2012, 133, 410.
16. H. G. Wu, P. J. Yao, N. Y. Ning, L. Q. Zhang, H. C. Tian, Y. P. Wu and M. Tian, RSC Adv., 2016, 6, 32932.
17. P. Deya, K. Naskar, B. Dashb, S. Nairb, G. Unnikrishnanb and G. B. Nando, RSC Adv., 2015, 5, 31886.
18. B. Ranjbar, H. Mirzazadeh, A. A. Katbab and A. N. Hrymak, J. Appl. Polym. Sci., 2012, 123, 32.
19. K. T. Min and G. H. Kim, Polym. Adv. Technol., 2011, 22, 2273.
20. L. F. Ma, R. Y. Bao, S. L. Huang, Z. Y. Liu, W. Yang, B. H. Xie and M. B. Yang, J. Mater. Sci., 2013, 48, 4942.
21. L. F. Ma, R. Y. Bao, R. Dou, S. D. Zheng, Z. Y. Liu, W. Yang and M. B. Yang, Compos. Sci. Technol., 2016, 128, 176.
22. M. van Duin and A. Souphantong, Rubber Chem. Technol., 1996, 68, 717.
23. Y. K. Chen, C. H. Xu, X. Q. Liang and M. L. Cao, J. Phys. Chem. B, 2013, 117, 10619.
24. A. V. Machado and M. van Duin, Polymer, 2005, 46, 6575.
25. H. G. Wu, M. Tian, L. Q. Zhang, H. C. Tian, Y. P. Wu, N. Y. Ning and T. W. Chan, ACS Sustainable Chem. Eng., 2015, 3, 26.
26. X. G. Tang, W. Yang, R. Y. Bao, G. F. Shan, B. H. Xie and M. B. Yang, Polymer, 2009, 50, 4122.
27. M. D. Ellul, A. H. Tsou and W. G. Hu, Polymer, 2004, 45, 3351.
28. P. Dey, K. Naskar, B. Dash and S. Nair, Materials Today Communications, 2014, 1, 19.
29. M. Faghihi, A. Shojaei and R. Bagheri, Composites, Part B, 2015, 78, 50.
30. R. Hashemi and G. J. Weng, Carbon, 2016, 96, 474.
31. A. Kasgoz, D. Akin and A. Durmus, Composites, Part B, 2014, 62, 113.
32. C. Gao, S. L. Zhang, B. Han, H. J. Sun, G. B. Wang and Z. H. Jiang, RSC Adv., 2014, 4, 42175.
33. J. R. Huang, C. Mao, Y. T. Zhu, W. Jiang and X. D. Yang, Carbon, 2014, 73, 267.
34. M. B. Heaney, Phys. Rev. B: Condens. Matter Mater. Phys., 1995, 52, 12477.
35. N. Koumakis, A. Pamvouxoglou, A. S. Poulos and G. Petekidis, Soft Matter, 2012, 8, 4271.
36. G. S. Shibulal and K. Naskar, J. Appl. Polym. Sci., 2013, 128, 4151.
37. M. I. Aranguren, E. Moral, J. V. DeGroot Jr and C. W. Macosko, J. Rheol., 1992, 36, 1165.
38. H. Kim and C. W. Macosko, Polymer, 2009, 50, 3797.
39. S. Aoyama, Y. T. Park, T. Ougizawa and C. W. Macosko, Polymer, 2014, 55, 2077.
40. M. Sabzi, L. Jiang, F. Liu, I. Ghasemi and M. Atai, J. Mater. Chem. A, 2013, 1, 8253.
41. C. L. Liang, Z. H. Mai, Q. Xie, R. Y. Bao, W. Yang and M. B. Yang, J. Phys. Chem. B, 2014, 118, 9104.
42. P. Cebe and S. D. Hong, Polymer, 1986, 27, 1183.

---