Enhanced dielectric properties of acrylic resin elastomer based nanocomposite with thermally reduced graphene nanosheets

Abstract

Acrylic resin elastomer (ARE) has drawn considerable attention in recent years due to its excellent film-forming properties, flexibility and elasticity. However, its low dielectric constant restricts its application. In this work, an ARE based composite material with enhanced dielectric properties was obtained via adding graphene nanosheets prepared by thermal reduction (TrGN) to the ARE (ARE/TrGN). Furthermore, polyvinylpyrrolidone (PVP) has been used to improve the dispersibility of TrGN in the ARE matrix (the resulting composites are hereby referred to as ARE/PVP-TrGN). It was found that the addition of PVP to the ARE composite material could increase the percolation threshold and dielectric constant of the composite. The results demonstrated that ARE/PVP-TrGN had a percolation threshold of 1.65 wt%, and when the loading of the filler was 1.6 wt%, at 100 Hz, it had a dielectric constant higher than 100, representing a 30 fold increase with respect to the neat ARE, and a low dielectric loss of 0.2. In addition, the thermogravimetric (TG) analysis indicated that the composite possessed better thermal stability than that of neat ARE. The elastic modulus of ARE/2% PVP-TrGN was lower than 6 MPa.

---

Introduction

Recently, an increasing amount of attention has been paid to high permittivity composite materials for their role in the development of high performance electronic devices, such as supercapacitors, artificial muscle, bio-sensors, intelligent robots, etc.¹ Composite materials with ceramic or polymer as the matrix respectively have been extensively researched to achieve high permittivity.² However, modern society places higher and higher demands on materials, and materials that only have a single application cannot meet the demands of industry, so only high dielectric composite materials with enhanced comprehensive properties have a wide array of applications. Hence, it is of importance to develop dielectric materials with enhanced comprehensive properties.³ In the existing composite systems, polymeric composites have received more and more attention because of their good processibility, elasticity, high breakdown voltage and superior corrosion resistance.⁴

ARE is one type of excellent dielectric elastomer, which has the advantages of a low elastic modulus, large electric field induced strain and excellent film-forming properties compared to other counterparts. Therefore ARE could be a suitable elastomer for use in MEMS devices and in the supercapacitor field.⁵ In recent years, composites with ARE as a matrix have been garnering increasing attention, however, the dielectric constant of ARE is relatively low (<7). At present, the methods for increasing the dielectric constant of a polymer can be classified into three categories: (1) the loading of fillers with a high dielectric constant,⁶ (2) the introduction of organic dipole groups to the polymer matrix,⁷ and (3) the addition of conductive fillers. Copper phthalocyanine (CuPc), for instance, a kind of effective high dielectric filler, was commonly used in the first category. As a typical example of the second category, the push–pull dipole N-allyl-N-methyl-p-nitroaniline was grafted onto the poly(dimethyl siloxane) (PDMS) to enhance the dielectric constant of the polymer. In the third category researchers make use of the percolation mechanism of conductive fillers,⁸ such as metal powder, carbon nanotubes, carbon black and so on. However, for the first category, only when the content of the filler is very large can the dielectric constant become very high. As for the second category, the polar group will result in an apparent decline in the polymers’ breakdown voltage and an increase in humidity sensitivity. By contrast, the third is an efficient way to obtain a high dielectric constant while maintaining excellent performance in other aspects.

A wide window of applications have been opening for nanomaterials due to their structural features and special properties on a nanometer scale in recent years.⁹ As a typical representative, graphene, a two dimensional sheet of sp²-hybridized carbon, is a new kind of carbon nanomaterial. Since its discovery in 2004, graphene has emerged as a rapidly rising star in the field of materials science and has gained tremendous attention due to its extraordinary characteristics, for instance, its large aspect ratio, ultrahigh thermal conductivity and electrical conductivity.¹⁰ Thanks to its exceptional physical properties, incorporating a small amount of graphene into the polymer matrix can dramatically enhance the electrical and mechanical properties of the polymer, so polymeric composites filled with graphene have drawn the unwavering interest of many researchers.¹¹ Moreover, compared with zero and one dimensional nanomaterials, graphene has a larger aspect ratio, which makes it easier to overlap to form a percolation network. As a result, nanocomposites with a high dielectric constant can be prepared by the loading of less graphene. Thus graphene is regarded as a desirable selection of nanofillers for preparing polymeric nanocomposites with excellent properties.¹²

Nowadays, there are some methods for fabricating graphene nanosheets, such as the micromechanical cleavage of graphite,¹³ epitaxial growth method,¹⁴ chemical vapor deposition (CVD), alkali metal intercalation and expansion,¹⁵ the chemical and thermal reduction of graphite oxide (GO),¹⁶ and the exfoliation of GO.¹⁷ Among them the oxidation–reduction method takes the dominant position, because the method is suitable for large industrial scale production. As for preparing graphene by chemical reduction that involves complicated procedures, hydrazine, sodium hydroxide and other compounds are commonly selected as reducing agents. However, most of these reducing agents are highly toxic, which will cause pollution to the environment; moreover, the treatment of waste liquid increases the costs of production. By contrast, fabricating graphene nanosheets by thermal reduction is facile and efficient, and more importantly, it is pollution-free. The thermally reduced graphene is thought to be driven by thermal energy-induced bond cleavage.¹⁸ Therefore, the thermal reduction method has become a preferred route.

The electric and mechanical properties of a polymeric nanocomposite with graphene as filler are largely determined by the dispersion property of the filler in the matrix as well as the adhesion property between the filler and the matrix.¹⁹ But graphene has a high tendency toward agglomerating together via van der Waals attraction because of the special planar sp²-carbon,²⁰ which will result in an increase of the dielectric loss. Thus the crucial task in obtaining a high dielectric constant while maintaining satisfactory dielectric loss is to prevent the agglomeration and overlapping of conductive fillers in the polymer matrix.²¹ The strategies to overcome the strong van der Waals forces can be divided into two categories: (1) covalent modification, in which the graphene is modified by grafting it onto the polymer molecular chain,²² but there are no active groups on the graphene molecule; and (2) noncovalent modification. In this category the graphene is modified by using a surface active agent,²³ which can protect the structure of graphite lattice.

In a recent study, PVP, a type of commonly used low-cost, non-toxic and biocompatible surface active agent that has good solubility in water and many organic solvents, was utilized to prevent the aggregation of graphene in solvents,²⁴ which greatly improved the performance of the composite materials.

In this investigation, a handy strategy was put forward to improve the dielectric properties of ARE. PVP was employed as the surface active agent of the graphene nanosheets prepared by thermal reduction to improve the dielectric properties of the composite, which was also never seen before. The TrGN modified by PVP was blended with ARE, and indeed the composite films obtained by a solution casting method exhibited a high dielectric constant, low dielectric loss as well as good flexibility.

Experimental section

Reagents

Natural graphite, potassium permanganate (KMnO₄) and benzoperoxide (BPO) were purchased from American Aladdin Reagent Co., Ltd. Concentrated sulfuric acid (H₂SO₄, 98%), sodium hydroxide (NaOH), concentrated hydrochloric acid (HCl, 33%) and dimethylformamide (DMF) purchased from Nanjing Chemical Reagent Co., Ltd were of analytical grade, and DMF was dried with CaH₂ and distilled under vacuum before use. Hydrogen peroxide (H₂O₂, 30%) was obtained from Sigma-Aldrich, Inc. Distilled water was prepared in the laboratory. Styrene (St) and butyl acrylate (BA) purchased from Shanghai Ling Feng Chemical Reagent Co., Ltd were all of analytical grade, and inhibitors were removed by NaOH solution followed by distillation before use. Analytical grade hydroxyethyl methylacrylate (HEMA) was purchased from American Aladdin Reagent Co., Ltd and purified with activated carbons. PVP was bought from Chinese Medicine Group Chemical Reagent Co., Ltd.

Preparation of TrGN

The preparation process referred to in ref. 25 was slightly altered. The detailed route is as follows: graphite oxide (GO) was synthesized from natural graphite by a modified Hummers’ method.²⁶ The preparation procedure consisted of three steps. (1) At the low temperature stage, natural graphite powder (2 g) was added into a 0 °C concentrated sulfuric acid solution (46 mL), and homogeneous mixing of natural graphite with concentrated sulfuric acid was achieved by mechanical stirring. Successively, KMnO₄ (6 g) was added gradually to the solution for 1 h under stirring, after which the solution was stirred for another 2 h below 5 °C. (2) The container was transferred to a lukewarm bath to keep the temperature of the solution at about 35 °C, and the solution was stirred for 2 h. (3) Distilled water (92 mL) was slowly added to the system to keep the temperature between 90 °C and 100 °C, and the solution was stirred for 1 h. Later, distilled water (280 mL) and H₂O₂ (30%, 20 mL) were slowly added to the solution in succession under stirring, which was followed by 30 min stirring, and the color of the solution gradually turned yellow. After that, the crude product was obtained by filtration and 10% dilute hydrochloric acid solution was used to wash the product three times to remove metal ions completely. Then the resulting product was washed thoroughly with distilled water followed by centrifugation until the pH of the supernatant was 7. The graphite oxide (GO) was dried at 40 °C for 48 h. Finally, the graphite oxide (GO) was put into a tubular furnace and heated from ambient temperature to 900 °C at a heating rate of 5 °C min⁻¹ under the protection of nitrogen, which was then maintained for 10 min.

Synthesis of ARE

The ARE was synthesized via radical polymerization. The synthetic route referred to in ref. 27 was subject to minor modification. The synthesis procedure is as follows: firstly, St (0.35 g), BA (0.60 g) and HEMA (0.05 g) were successively added into a three-necked flask (100 mL) equipped with an Allihn condenser, a magnetic stirrer and a thermometer. Then, DMF (1.5 mL) was also added into the flask as the solvent and BPO (0.005 g) was added as the initiator (BPO here was firstly dissolved into DMF and then the solution was dropped into the flask instead of adding BPO to the flask directly, which was the minor modification that was made). The reaction was carried out at about 100 °C for 5 h under nitrogen protection. Finally, the reaction device was converted into a vacuum distillation unit to remove unreacted monomers and solvent.

Dispersion performance test of TrGN in DMF

Firstly, the solution of PVP was prepared by dissolving 10 mg PVP in 5 mL DMF and stirring for 10 min. Then 10 mg TrGN was added into the solution, which was subjected to ultrasonic treatment for 1 h at room temperature,²⁸ and the resulting solution was labeled as PVP-TrGN/DMF. As a comparison, the DMF dispersion of TrGN without PVP modification was fabricated under the same conditions and labeled as TrGN/DMF. The two solutions were centrifuged under the same conditions for 3 min, and then left overnight.

Preparation of the composite films

The composite films here were fabricated by a solution casting method. In a typical experiment, ARE (1 g) was dissolved in DMF under stirring at room temperature, then a desired amount of PVP-TrGN dispersed in DMF by ultrasonic vibration was added to the solution under magnetic stirring. The resulting solution was stirred for 1 h, then the solution was cast onto a polytetrafluoroethylene mold and dried at atmospheric pressure at 80 °C for 8 h. Finally the mixture was dried under vacuum at 80 °C for 8 h to remove the DMF completely, and thus the nanocomposite film was obtained. The weight percentage of modified TrGN was designed to be 0.3%, 0.5%, 0.7%, 1.0%, 1.5%, 1.7%, 2.0%, 4.0%, 7.0%, and the corresponding composite films were marked as ARE/x PVP-TrGN (where x is the weight percentage of filler). For comparison, a blend of ARE and TrGN was prepared under the same conditions, and the composite films were marked as ARE/x TrGN.

Characterization

Raman spectra were recorded with a inVia Raman spectrometer (HORIBA, Kyoto, Japan) with 488 nm diode laser excitation on a 300 lines per mm grating at ambient temperature. X-ray diffraction (XRD) patterns of samples were monitored at room temperature with a XRD-6100 diffractometer (SHIMADZU, Kyoto, Japan) equipped with a Cu Kα radiation source (λ = 1.540 Å). The scanning range was from 5° to 40° and the scanning interval was 5°. Fourier transform infrared (FTIR) spectra were collected with a Nicolet Nexus-670 spectrometer (Madison, America) by incorporating the sample in a KBr disk. X-ray photoelectron spectra (XPS) were measured with a PHI Quantera II spectrometer (Ulvac-Phi, Chigasaki, Japan). Transmission electron microscopy (TEM) imaging was performed with a Tecnai-12 electron microscope (Philips company, Amsterdam, Holland) operating at an accelerating voltage of 120 kV, and the TEM samples were prepared by dispersing a small amount of dry powder into ethanol followed by ultrasonic vibration and placing onto copper grids before observation. Nuclear magnetic resonance (¹H-NMR) spectra were measured by a Bruker DRX-500 spectrometer (Bruker, Germany) with chloroform-d as the solvent. The number-average molecular weight of ARE was measured by a PL-GPC 200 (Shanghai, China) instrument operated at 40 °C using standard polyethylene as reference and DMF as solvent at a flow rate of 0.35 mL min⁻¹. The fracture surfaces morphologies of the film samples fractured in liquid nitrogen were observed with a Hitachi S-4800 (Tokyo, Japan) scanning electron microscope (SEM). The electric resistance measurements were accomplished by a standard four-point probe (kdy-1) produced by Guangzhou Kund Technology Co. Ltd. Dielectric properties were measured by a HP 4294A precision impedance analyzer (Dongguan, China) in the frequency range of 40 Hz to 1 MHz at room temperature and the permittivity values (ε) of the samples were calculated by the following equation:where C is the measured capacitance of the sample (in parallel mode), ε₀ is the vacuum dielectric constant (∼8.85 × 10⁻¹² F m⁻¹), A represents the area of the electrode smeared on both sides of the sample before measurement, and h represents the thickness of the sample. The electric breakdown strength was measured by a dielectric withstand voltage test (Beijing Electromechanical Research Institute, China). Thermogravimetric (TG) analysis curves were obtained by a Simultaneous Thermal Analyzer (STA409PC, NETZSCH company, Bavarian State, Germany). The TG film samples with the weight of about 10 mg were sealed in aluminium oxide pans and the temperature was set from 25 °C to 600 °C at a rate of 10 °C min⁻¹. The elastic performance test of the film samples was operated using a DMTA-V dynamic mechanical thermal analyzer (Rheometric Scientific, Inc.) at room temperature and 1 Hz using a strain dependent mode.

Results and discussion

Raman spectroscopy is a widely used tool for characterizing graphene. As shown in Fig. 1, the Raman spectrum of natural graphite displays a well-known strong G peak (the E_2g phonon mode of sp² carbon atoms) at 1582 cm⁻¹, and a weak D peak (the breathing mode of the k-point mode of A_1g symmetry) at 1338 cm⁻¹. The generation of the D peak caused by defects was due to the existence of pentagon, heptagon or other local defects. In the Raman spectrum of TrGN, the G peak is broadened and shifted upward to 1598 cm⁻¹, in addition, the strength of the G peak decreases. Meanwhile, the strength of the D peak at 1338 cm⁻¹ clearly increases, which might be due to the significant decrease of the size of the in-plane sp² domains because of oxidation and thermal exfoliation, and the partially ordered graphite crystal structure of graphene nanosheets. Compared with TrGN, the I_D/I_G ratio (the intensity ratio of the D band relative to the G band) of PVP-TrGN slightly increases, which might indicate that the grain size of graphene continues to decrease and the lattice defects increase due to the presence of PVP.

Fig. 1 Raman spectra of natural graphite, TrGN and PVP-TrGN.

XRD was further utilized to characterize the structures of natural graphite, GO, TrGN and PVP-TrGN. As presented in Fig. 2, a strong peak at 26.4° appeared in the spectrum of graphite, which corresponded to (002) lattice planes. As a result of the intercalation of functional groups that resulted from graphite oxidation into the interlayers of the graphite sheets, GO exhibited a sharp diffraction peak at 9.8° with a wider inter-lamellar spacing (d-spacing) than graphite. A broad swelling diffraction peak between 20° and 25° was observed in the spectra of both TrGN and PVP-TrGN, which was attributed to the (002) lattice planes of the graphene. Moreover, very weak characteristic peaks at about 10° were also observed in the spectra of TrGN and PVP-TrGN separately, which indicated that the degree of reduction of GO was not complete. The diffraction angle of TrGN was slightly larger than that of PVP-TrGN, which was because the lamellar spacing of PVP-TrGN was larger than that of TrGN as a result of the adsorption of PVP on graphene nanosheets. The FTIR spectra of natural graphite, GO, TrGN and PVP-TrGN (Fig. S1 in ESI†) further proved that GO was reduced successfully and PVP was adsorbed on the graphene nanosheets effectively. The results were consistent with the XRD data.

Fig. 2 XRD patterns of natural graphite, GO, TrGN and PVP-TrGN.

Fig. 3a showed the XPS general spectra of graphite, GO, TrGN and PVP-TrGN, and the C/O ratios were 32.10, 2.39, 29.21 and 9.58 respectively. It should also be mentioned that the decrease in the C/O ratio of GO indicated the introduction of oxygen-containing groups in the oxidation process, and the increase in the C/O ratio of TrGN indicated that most of the oxygen-containing groups were removed after thermal reduction at 900 °C. The peak of C–N (286.0 eV) corresponded to the N in the C–N bonds of PVP. XPS high-resolution C 1s spectra of GO and TrGN are shown in Fig. 3b and c separately. There were four peaks in the C 1s spectra of GO: the peak at 284.7 eV could be ascribed to C–C/C=C, and the other peaks arose from C–O (hydroxyl and epoxy, 286.6 eV), C=O (carbonyl, 287.2 eV) and O–C=O (carboxyl, 288.8 eV) groups separately. After thermal reduction, as is shown in Fig. 3c, the oxygen peaks in the C 1s spectra of TrGN were significantly weakened, which indicated that most of the oxygen groups had been removed after reduction.

Fig. 3 XPS general spectra of natural graphite, GO, TrGN and PVP-TrGN (a) and XPS high-resolution C 1s spectra of GO (b) and TrGN (c).

Digital pictures of PVP-TrGN and TrGN dispersed in DMF with the aid of an ultrasonication treatment for 1 h (Fig. S2 in ESI†) indicated that PVP can prevent the graphene nanosheets from aggregating sterically by being adsorbed on the surface of the graphene, and the resulting dispersion was stable against centrifugation. The explanation for the adsorption mechanism is as follows: the unpaired electrons on the carbonyl groups (C=O) and C–N bonds of the PVP molecules could form strong π–π interactions with the electron cloud on the surface of graphene; furthermore, PVP’s long chain could fully extend to the solvent and form a thick adsorption layer, resulting in a steric hindrance effect that prevents the stacking of graphene layers.²⁹ Thus the graphene in the solvent can obtain better stability and dispersion.

TEM images of TrGN and PVP-TrGN are exhibited in Fig. 4. As can be seen, due to the absence of PVP, TrGN showed obvious aggregation and a stacked sheet structure (Fig. 4a); by contrast, PVP-TrGN presented fewer layers because of the presence of PVP (Fig. 4b). This phenomenon further proved the effect of PVP on the stability of graphene.

Fig. 4 TEM images of TrGN (a) and PVP-TrGN (b).

FTIR and ¹H-NMR spectra of ARE (Fig. S3 and S4 in ESI†) were utilized to characterize the structure, and the spectra showed that ARE was synthesized successfully. In order to further investigate the quality of the polymer synthesized, the number-average molecular weight of ARE was measured and the result was 67 549 g mol⁻¹.

The dispersion state of the filler in the matrix was observed clearly with SEM (Fig. S4 in ESI†). The micrographs indicated that TrGN without the modification of PVP tended to aggregate and stack; by contrast, PVP-TrGN exhibited a very homogeneous dispersion in the matrix.

The frequency dependence of the dielectric constant and dielectric loss of the ARE/PVP-TrGN composites are depicted in Fig. 5a and 6a respectively. As demonstrated in Fig. 5a, when the content of the filler was low, the dielectric constant varied little with frequency, showing a nearly constant value of permittivity. When the content of the filler reached a critical value, the dielectric constant decreased with increasing frequency. The explanation for this phenomenon is as follows: certain polarization produced by the orientation arrangement of some dipoles can contribute to the formation of the dielectric constant. Due to the variation of frequency, the dipoles reverse with the change in the external electric field, and when the frequency becomes high, the inversion of dipoles cannot keep up with the electric speed because of the internal resistance of the material.³⁰ Under the condition of high frequency, some dipoles stop inversion, which results in some contributions to the dielectric constant disappearing. The almost independence of the frequency on the dielectric constant when the content of the filler was low indicated that there was no plentiful accumulation of interfacial charges inside the nanocomposites. The variation tendency of the dielectric loss with frequency was also represented in Fig. 6a. Generally speaking, the variation of dielectric loss with frequency was consistent with the variation of the dielectric constant. As a comparison, the frequency dependence of the dielectric constant and the dielectric loss of the ARE/TrGN composites are depicted in Fig. 5b and 6b respectively.

Fig. 5 Dielectric constant of ARE/PVP-TrGN (a) and ARE/TrGN (b) with different weight fractions of filler measured at room temperature as a function of frequency.

Fig. 6 Dielectric loss of ARE/PVP-TrGN (a) and ARE/TrGN (b) with different weight fractions of filler measured at room temperature as a function of frequency.

The content dependence of DC conductivity and dielectric constant (at 100 Hz) of the ARE/PVP-TrGN composites are demonstrated in Fig. 7a and b respectively. Generally speaking, the dielectric constant of the composites increased with the increasing loading of the filler. The increase in dielectric constant is attributed to two reasons: (1) the Maxwell–Wagner–Sillars (MWS) polarization, which is related to the entrapment of nomadic charges between the insulator and conductor interface;³¹ that is, the interfacial polarization. MWS polarization will occur as long as the charge carriers at an interface between the matrix and the filler accumulate. For the composite mentioned above, the charges in conductive TrGN were easily delocalized, which accumulated at the interface of TrGN and the ARE insulator layer when they were driven by the applied electric field, and the accumulation of the charges contributed to a large polarization. With the increase of filler content, the MWS polarization was significantly improved. As a result, the dielectric constant of the composite was quickly enhanced. (2) The formation of “microcapacitors” in the ARE matrix; that is, two parallel graphene sheets are isolated by a thin insulating ARE interlayer.³² In addition, it was also interesting to observe the figures in detail. As is demonstrated in Fig. 7a and b, when the content of filler was low, the direct current (DC) conductivity and dielectric constant increased slowly with the increase of the filler content. When the content reached a critical value, there was a huge increase in the DC conductivity and dielectric constant. These phenomena obviously indicated that the transition from insulator to conductor occurred in the ARE/PVP-TrGN composites. The following interpretative statement was made to explain the above transition. The dielectric constant of the composite could be improved by adding a small amount of graphene to the polymer matrix, which was because conductive graphene, with its two-dimensional structure, and the insulating polymer could constitute microcapacitors and these microcapacitors improved the charge storage ability of the composite. When the content of graphene was low, there were few microcapacitors, so the DC conductivity and the dielectric constant showed a slight increase. With the increase of the filler content, there were more and more microcapacitors, and the DC conductivity and dielectric constant of the composite continued to increase. These two dimensional nanomaterials mutually overlapped in the polymer matrix and gradually built up a conductive network. When the content of graphene was beyond a critical value, the conductive network happened to be able to form, and the transition of the composite from insulator to conductor was realized, which resulted in a significant increase of the DC conductivity and dielectric constant. This phenomenon was called percolation, and the critical value was called the percolation threshold. The percolation threshold of composites can be deduced from the following power laws:³³

σ_dc ∝ (p_c − p)^t for p < p_c (2)

where σ_dc is the DC conductivity of composites, p_c is the percolation threshold and p is the fraction of the conductive filler, and t is a scaling constant. By fitting the DC conductivity figures for p < p_c, as is presented in inset of Fig. 7a, a percolation threshold value of 1.65 wt% was obtained. As is shown in Fig. 7b, the dielectric constant increased remarkably and reached more than 180 when the loading of filler was 1.8 wt%, which was about 60 times larger than that of neat ARE (∼3.2). As a comparison, the dependence of DC conductivity and dielectric constant on the content of the ARE/TrGN composites are also shown in Fig. 7a and b respectively. The percolation threshold of 1.20 wt% was obtained by the same method.

Fig. 7 DC conductivity of ARE/PVP-TrGN and ARE/TrGN nanocomposites as a function of the filler content measured at room temperature (a), and dielectric constant of ARE/PVP-TrGN and ARE/TrGN nanocomposites as a function of the filler content measured at 100 Hz and room temperature (b).

Compared with the ARE/TrGN composite, the ARE/PVP-TrGN composite had a higher dielectric constant and percolation threshold, which was because the addition of PVP improved the dispersion performance of graphene in the matrix. The improved dispersion delayed the formation of the conductive network of filler in the matrix, which helped the production of more microcapacitors, and these microcapacitors in turn contributed to the higher dielectric constant. At the same time, the latency of the formation of the conductive network of filler directly raised the percolation threshold.

The electric breakdown strength of ARE/1.0% PVP-TrGN and ARE/1.0% TrGN was measured by a dielectric withstand voltage test separately. The facts showed that the electric breakdown strength of ARE/1.0% PVP-TrGN (45 V μm⁻¹) was higher than that of ARE/1.0% TrGN (37 V μm⁻¹), which was because TrGN had a poor dispersion in the matrix and it was easier to form a conductive network.

TG curves for neat ARE and ARE/PVP-TrGN composites (Fig. S5 in ESI†) were utilized to investigate the thermal stability of the composites. The TG results indicated that the thermal degradation temperature of neat ARE was about 367 °C, and the thermal degradation temperature of the ARE/PVP-TrGN composite was higher than that of neat ARE. Moreover, the thermal degradation temperature of the ARE/PVP-TrGN composite increased with the increase of the filler content. The curves also demonstrated that the addition of the filler could slow down the weight loss during thermal degradation. The TG curve of the ARE/2 wt% PVP-TrGN composite showed that the amount of weight residue increased by 6.2% in comparison with neat ARE at 600 °C. These phenomena can be explained as follows: PVP has a long nonpolar polyethylene chain, which has a good adsorption effect upon thermally reduced graphene nanosheets. Moreover, because the polymer that we synthesized has high viscosity, there is an inevitable interaction between PVP-TrGN and ARE, which reduces the thermal amplitude of the molecular chains. So the thermally unstable groups of ARE were transformed into more stable structures, which contributed to a reduction of the possibility of thermal degradation.

DSC curves of neat ARE and ARE/PVP-TrGN composites (Fig. S6 in ESI†) showed that the decomposition enthalpy of neat ARE was about 206.8 J g⁻¹, and the decomposition enthalpy of the composites with filler contents of 0.5%, 1.5%, 2.0% were about 236.9 J g⁻¹, 252.6 J g⁻¹, 256.4 J g⁻¹ respectively. Removing the weight of TrGN, the decomposition enthalpy of neat AE were about 238.1 J g⁻¹, 256.4 J g⁻¹, 261.6 J g⁻¹ respectively, which further proved that the addition of graphene improved the thermal stability of the composite.

The elastic deformation stage of representative stress–strain curves of the ARE/PVP-TrGN and ARE/TrGN nanocomposites is depicted in Fig. 8. As can be seen, the addition of PVP-TrGN and TrGN increased the initial modulus of the polymer (1.19 MPa). Moreover, PVP-TrGN showed a better reinforcement effect than that of ARE/TrGN, which might be because the presence of PVP increased the dispersion property of the filler in the matrix. A large surface area of graphene nanosheets was in contact with the ARE matrix and the better interfacial interaction led to an efficient transfer of the load from the ARE matrix to the graphene nanosheets, contributing to an increase in the mechanical properties of nanocomposites. However, regardless of the kind of composite, the elastic modulus of the composites was not very large and could still be accepted in practical applications. Therefore, the preparation of this kind of composite material is of practical significance.

Fig. 8 Elastic deformation stage of stress–strain curves of neat ARE, ARE/PVP-TrGN and ARE/TrGN nanocomposites.

Conclusion

In summary, ARE was synthesized by free radical polymerization. Graphene nanosheets were prepared by the chemical oxidation of natural graphite followed by the thermal exfoliation and reduction of GO. Dielectric elastomer composites, ARE/PVP-TrGN and ARE/TrGN were prepared using a solution casting method. Compared with ARE/TrGN, the ARE/PVP-TrGN composite not only possessed better dielectric properties, but also had enhanced thermal stability and elasticity because of the improved dispersion property of the filler. The ARE/PVP-TrGN composite with a filler content of 1.6% had a dielectric constant higher than 100 and a dielectric loss of 0.2; meanwhile, it also had an elastic modulus lower than 6 MPa at room temperature. However, PVP’s ability to enhance the dispersion of graphene in solvents is limited. Further improvement in properties can be expected through the amelioration in the composite fabrication process, especially by grafting the graphene oxide onto the acrylic resin to increase the distribution uniformity of the filler in the polymer matrix.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 21174063), the Natural Science Foundation of Jiangsu Province (No. BK20131358), the Aeronautical Science Foundation of China (No. 2011ZF52063 and No. 2014ZF52069), and A Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD).

References

1. J. H. Yu, X. Y. Huang, C. Wu and P. K. Jiang, IEEE Trans. Dielectr. Electr. Insul., 2011, 18, 478–484.
2. E. Lee and Y. G. Jeong, Fibers Polym., 2015, 16, 2021–2027.
3. X. Y. Huang, P. K. Jiang and L. Y. Xie, Appl. Phys. Lett., 2009, 95, 2901–2903.
4. 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.
5. Q. M. Zhang, H. F. Li, M. Poh, F. Xia, Z.-Y. Cheng, H. S. Xu and C. Huang, Nature, 2002, 419, 284–287.
6. K. S. Lam, Y. Zhou, Y. W. Wong and F. G. Shin, J. Appl. Phys., 2005 DOI:10.1063/1.1906285.
7. B. Kussmaul, S. Risse, G. Kofod, R. Wache, M. Wegener, D. N. McCarthy, H. Kruger and R. Gerhard, Adv. Funct. Mater., 2011, 21, 4589–4594.
8. S. Zhang, N. Zhang, C. Huang, K. Ren and Q. M. Zhang, Adv. Mater., 2005, 17, 1897–1901.
9. H. K. F. Cheng, N. G. Sahoo, Y. P. Tan, Y. Z. Pan, H. Q. Bao, L. Li, S. H. Chan and J. H. Zhao, ACS Appl. Mater. Interfaces, 2012, 4, 2387–2394.
10. A. K. Geim and K. S. Novoselov, Nat. Mater., 2007, 6, 183–191.
11. C. Soldano, A. Mahmood and E. Dujardin, Carbon, 2010, 48, 2127–2150.
12. R. Sengupta, M. Bhattacharya, S. Bandyopadhyay and A. K. Bhowmick, Prog. Polym. Sci., 2011, 36, 638–670.
13. K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos, I. V. Grigorieva and A. A. Firsov, Science, 2004, 306, 666–669.
14. A. N. Obraztsov, Nat. Nanotechnol., 2009, 4, 212–213.
15. F. T. Thema, M. J. Moloto, E. D. Dikio, N. N. Nyangiwe, L. Kotsedi, M. Maaza and M. Khenfouch, J. Chem., 2013 DOI:10.1155/2013/150536.
16. X. Liu, L. Yi, X. Wang, J. Su, Y. Song and J. Liu, Int. J. Hydrogen Energy, 2012, 37, 17984–17991.
17. G. Moula, M. A. Mumin and P. A. Charpentier, J. Nanomater. Mol. Nanotechnol, 2013, S1–S002,  DOI:10.4172/2324-8777.
18. K. Kalaitzidou, Exfoliated graphite nanoplates as reinforcement for multifunctional polypropylene nanocomposites, Michigan State University, Michigan State, 2006.
19. H. Wang, P. G. Ren, J. Z. Xu, D. X. Yan, Z. M. Li and L. Xu, Compos. Interfaces, 2014, 21, 203–215.
20. N. Ruecha, R. Rangkupan, N. Rodthongkum and O. Chailapakul, Biosensors and Bioelectronics, 2014, 52, 13–19.
21. P. Xu, H. G. Gui, X. X. Wang, Y. D. Hu and Y. S. Ding, Compos. Sci. Technol., 2015, 117, 282–288.
22. F. Wen, Z. Xu, S. B. Tan, W. M. Xia, X. Y. Wei and Z. C. Zhang, ACS Appl. Mater. Interfaces, 2013, 5, 9411–9420.
23. M. Saha, P. Tambe, S. Pal, P. Kubade, G. Manivasagam, M. A. Xavior and V. Umashankar, Compos. Interfaces, 2015, 22, 611–627.
24. S. Liu, X. R. Xing, J. H. Yu, W. J. Lian, J. Li, M. Cui and J. D. Huang, Biosensors and Bioelectronics, 2012, 36, 186–191.
25. H. L. Chen, C. W. Huang, W. Yu and C. X. Zhou, Polymer, 2013, 54, 1603–1611.
26. F. Yavari, H. R. Fard, K. Pashayi, M. A. Rafiee, A. Zamiri, Z. Z. Yu, R. Ozisik, T. B. Tasciuc and N. Koratkar, J. Phys. Chem. C, 2011, 115, 8753–8758.
27. R. N. Liu, J. W. Wang, Q. Li, S. Q. Li, S. Zhang and X. J. Ding, J. Appl. Polym. Sci., 2014, 131,  DOI:10.1002/app.39975.
28. A. S. Wajid, S. Das, F. Irin, H. S. T. Ahmed, J. L. Shelburne, D. Parviz, R. J. Fullerton, A. F. Jankowshi, R. C. Hedden and M. J. Green, Carbon, 2012, 50, 526–534.
29. A. B. Bourlinos, V. Georgakilas and R. Zboril, et al., Aqueous-phase exfoliation of graphite in the presence of polyvinylpyrrolidone for the production of water-soluble graphenes, Solid State Commun., 2009, 149, 2172–2176.
30. B. K. Dash, P. G. R. Achary and N. C. Nayak, J. Mater. Sci.: Mater. Electron., 2015, 26, 7244–7254.
31. Q. Su, S. Pang, V. Alijani, C. Li and X. Feng, Adv. Mater., 2009, 21, 3191–3195.
32. P. Fan, L. Wang, J. T. Yang, F. Chen and M. Q. Zhong, Nanotechnology, 2012 DOI:10.1088/0957-4484/23/36/365702.
33. Z. M. Dang, Y. H. Lin and C. W. Nan, Adv. Mater., 2003, 15, 1625–1629.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra16932e

---