Preparation of PVDF/graphene ferroelectric composite films by in situ reduction with hydrobromic acids and their properties

Abstract

A new two-step process was developed to prepare PVDF/reduced graphene oxide (PVDF/rGO) composite films: the synthesis of PVDF/GO composite films and immersion of such films in hydrobromic acids for reduction. This method avoided the agglomeration of rGO during reduction in PVDF/GO solutions and efficiently improved the dispersion effect of rGO in the PVDF matrix. Meanwhile, it simplified the preparation process due to no modification of GO being required, and opened a feasible way to scale up the production of PVDF/rGO composites. Experiments showed that PVDF with nearly all β phase was obtained when the content of rGO was 0.1 wt% (PrGO-0.1), and the dielectric constant increased from 10 for the neat PVDF to 41 for PrGO-0.1 at 1 kHz. The ferroelectric, piezoelectric, and dynamic mechanical properties of the PVDF/rGO composites were also comprehensively studied. As the content of the β phase was nearly 100%, the piezoelectric constant and remnant polarization of the PrGO-0.1 film increased by 78.6% and 69.3%, respectively, compared with those of the neat PVDF, and therefore became the highest among all composite films. The rGO also, to a great extent, helped to enhance the mechanical properties of the PVDF composites. As a result, the improved piezoelectric and ferroelectric properties made the PVDF/rGO composite films with 0.1 wt% rGO content much better piezoelectric energy transfer and ferroelectric storage materials than the neat PVDF.

---

1. Introduction

Poly(vinylidene fluoride) (PVDF) is a very attractive polymer having dielectric, piezoelectric, ferroelectric, pyroelectric and other excellent physical properties. We can use one or more properties to develop numerous devices, such as capacitors, piezoelectric sensors, ferroelectric random access memory, and pyroelectric detectors, etc.^1,2

PVDF is able to crystallize in five different forms, namely the nonpolar α phase, the polar β and γ phase, δ and ε phase. The α and β phases are the most common crystalline structures observed in PVDF.^3,4 The β phase has an all-trans conformation (TTTT), comprising fluorine atoms and hydrogen atoms on opposite sides of the polymer backbone, resulting in a net non-zero dipole moment. The presence of this dipole makes PVDF a good piezo- and ferro-electrically active polymer.⁵ Moreover, the β phase PVDF can convert electrical energy into mechanical energy under applied electric field, but it need higher drive voltage (more than 100 kV mm⁻¹), which has seriously constrained its application. The key is to have higher dielectric constant as it helps to reduce drive voltage while maintaining high electromechanical conversion coefficients.^6,7 However, PVDF usually exists in the nonpolar α phase and the polar β phase cannot naturally form,⁸ plus the dielectric constant of the β phase PVDF is only 12 or so.¹ Thus, enhancement of the content of the polar β phase in PVDF and improvement of dielectric constant are both of great importance.⁹

Graphene has attracted a great amount of attention due to its unique properties such as high mechanical properties, electrical conductivity, mobility of charge carries, specific surface area, and fascinating transport phenomena, and so on.^10–12 Graphene is not only ideal as a nano-filler for improving the mechanical, electrical, and thermal properties of polymers^13–15 but also able to serve as the nucleation agent for the generation of polar phase in PVDF, and this is due to the specific interactions between the large π bond in graphene and the higher electron dense fluorine atoms in PVDF, which results in the β phase formation in the PVDF/graphene composites. Furthermore, graphene is conductive, so PVDF/graphene composite has higher dielectric constant than that of the neat PVDF.^16,17 Nevertheless, it is usually difficult to disperse graphene sheets uniformly in polymer matrix because graphene sheets can easily re-agglomerate or re-stack into graphite due to the strong van der Waals interactions between them. However, only promoting the dispersion of graphene in polymer matrix can strengthen the graphene–polymer interactions, and then benefit the formation of β phase in PVDF and improve the properties of PVDF/graphene. So some efforts have carried out to improve dispersion of graphene in PVDF matrix.^18,19 These studies on the preparation of PVDF/graphene are mainly carried out with modification of graphene. These methods include three steps: (1) preparation of functionalized graphene oxide, (2) preparation of the reduced functionalized graphene oxide, and (3) preparation of composite films of PVDF/reduced functionalized graphene oxide. Though functionalized graphene can uniformly disperse in PVDF matrix, the preparation process, particularly the first step, is very complicated, and modification effect is not easy controlled. So it is necessary to prepare uniformly dispersed PVDF/graphene composites through a simple and efficient method.

The method in this text uses the characteristics of the graphene oxide (GO) to prepare PVDF/graphene composite films. GO are graphene sheets on which oxygen-containing functional groups are thought to be present in the form of carboxyl, carbonyl, and hydroxyl, epoxy groups.¹¹ GO can uniformly disperse in polymer matrix because of hydrogen bond or strong and specific interactions between oxygen-containing functional groups in GO surface and groups in polymer. Parades et al.²⁰ achieved stable suspensions of GO in the organic solvent, such as N,N-dimethylformamide (DMF), N-methyl-2-pyrrolidone (NMP) and tetrahydrofuran (THF). Moreover, PVDF can completely dissolve in DMF or NMP, so the large-scale preparation of GO homogeneous dispersions in PVDF is highly desirable.

In conclusion, the characteristics of GO provide the basis for the preparation of PVDF/graphene composite materials. Thus, the method in this text contains two steps: the first step is dispersing PVDF and GO in DMF respectively, then mixing the above two solutions, a stable mixture is obtained thanks to the strong interactions between oxygen containing functional groups in GO surface and PVDF chains, then drying the mixture and obtaining PVDF/GO film; in the second step, PVDF/GO films are immersed into a chemical reducing agent-hydrobromic acids for reduction and obtain PVDF/rGO films. This method has three advantages: (1) GO does not need to be modified, preparation process is simplified; (2) aggregation of rGO is possibly avoided during direct reduction in PVDF/GO solution, and efficiently improved the dispersion effect of rGO in PVDF matrix; (3) no residual reducing agent remains in the composites. Therefore, this method opens a feasible way to scale up the production of PVDF/rGO composite materials.

In this study, a simpler and more efficient method was used to prepare PVDF/rGO composite films. Then, dispersion of rGO, crystal transformation of PVDF composite films was investigated. Optimizing the rGO concentration in the PVDF matrix was also explored. At last, dielectric, piezoelectric, ferroelectric and dynamic mechanical properties of composite films were comprehensively studied, which provided a theoretical basis for the applications of ferroelectric PVDF/rGO composites.

2. Experimental

2.1 Materials

Graphene oxide was purchased from Nanjing xfnano Co., Ltd. PVDF was obtained from Dongguan Ensi Chemical Co., Ltd. Analytical grade dimethylformamide (DMF) and the 40% hydrobromic acids were purchased from Sinopharm Chemical Reagent Co., Ltd. (SCRC).

2.2 Samples preparation

The GO sheets were dispersed in DMF solvent (0.5 mg ml⁻¹) and ultra-sonicated for 4 h at room temperature, PVDF powder was separately dissolved in DMF (80 mg ml⁻¹) by stirring for 2 h at 60 °C. Then, the GO suspension was added drop wise to the PVDF solutions while stirring, and the mixture was stirred for an additional 30 min at 60 °C, ultra-sonicated for another 1 h at room temperature and stirred again for 1 h at 60 °C, then let it sit for 45 min at room temperature in a vacuum oven in order to remove air bubbles. Finally, this uniform mixture solution was poured into a self-made polishing quartz glass dish to evaporate in the 70 °C oven for 24 h and then placed in a vacuum oven for 48 h at 70 °C to further evaporate the solvent and get equilibrium weight. Films with GO content of 0.05, 0.08, 0.1, 0.3, 0.5, 0.7 wt% were prepared. Samples were denoted as PGO-0.05, PGO-0.08, PGO-0.1, PrGO-0.3, PGO-0.5 and PGO-0.7. The neat PVDF films were prepared in the same conditions.

The PVDF/GO films were then immersed into the HBr acids solution in sealed cuvettes that were placed in an oil bath at 90 °C for 5 h to yield PVDF/rGO composite films. Samples were denoted as PVDF, PrGO-0.05, PrGO-0.08, PrGO-0.1, PrGO-0.3, PrGO-0.5 and PrGO-0.7.

After reduction, the color of the film changed from yellow-brown to black, which indicated a successful reduction of GO within PVDF matrix (Fig. 1). In the following text, the marked contents of rGO was actually the contents of GO. Obviously, the true content of rGO was less than that of GO as part of oxygen-functional groups had been removed during reducing process. However, the effect of rGO could still be demonstrated in the discussion below.

Fig. 1 Pictures of PGO-0.5 (A) and PrGO-0.5 (B) films.

2.3 Characterization

The fractured morphology of the film was examined with field emission scanning electron microscopy (SEM) equipped with an energy-dispersive spectrometer (EDS). Reducing effect was evaluated by EDS, X-ray powder diffraction (XRD) spectra and Fourier transform infrared (FT-IR) spectra. Crystal transformation was studied with XRD, FT-IR and Differential scanning calorimeter (DSC). XRD was recorded on a D8 Advance, Bruker diffractometer in reflection mode, using copper-radiation, Cu Kα₁ radiation (λ = 0.15148 nm) operated at 40 kV and 30 mA. All samples were scanned from 2θ = 5–40° at the scan rate 0.5 s per with a step size of 0.05°. FT-IR was recorded in the mid-infrared region of 4000–400 cm⁻¹ using an IR, Tensor 27, Bruker Corp. The spectra were collected at 2 cm⁻¹ resolution and by averaging 32 scans at ambient condition. Differential scanning calorimeter (DSC) were performed on a TA Instruments Q1000 modulated calorimeter, over a temperature range from 25 to 200 °C at the heating/cooling rates of 10 °C min⁻¹ and a N₂ flow of 50 ml min⁻¹. Piezoelectric coefficient (d₃₃) of the composites was measured by a ZJ-type standard static d₃₃ meter (IACAS ZJ-4AN). For each type of sample, five samples were measured at five different points. The electrodes were deposited using a magnetron sputtering system (JGP-800 SKY) for dielectric property test. The dielectric permittivity and loss tangent of the samples as a function of frequency were carried out using an impedance analyzer (4294A HP). The breakdown experiment was performed on the high voltage test equipment (YD2013). The ferroelectric hysteresis loops (P–E loops) were measured using ferroelectric test system (WS-2000 RADIANT). The polarization of the films was performed with the high voltage polarization instrument (ET2673A). The storage modulus (G′) and loss modulus (G′′) and tan δ values of the composites were measured using a dynamic mechanical analyzer (DMA) (model Q-800, TA instruments). Samples were prepared in the film form (25 mm × 6 mm × 0.04 mm). The films were then installed at the tension clamp of a calibrated instrument and heated from 40 to 170 °C at the heating rate of 3 °C min⁻¹. The G′, G′′ and tan δ values are measured at a constant frequency of 1 Hz with a static force of 0.02 N.

3. Results and discussion

3.1 The reduction of GO and PVDF/GO composite films

The possible reducing mechanism is that hydrobromic acids enter into the interfaces of PVDF and GO in the form of ions, then react with GO: (1) hydrobromic acids can catalyze the ring-opening reaction of epoxy groups and turn them into hydroxyl groups;²¹ (2) there is substitution reaction of the hydroxyl group by the bromine atom. Elias et al.²² indicated the hydrogenation of graphene in which hydrogen atoms can be attached and removed from graphene without destroying the carbon lattice. The binding energy between a carbon and bromine is lower than that of the C–H bond, so that bromine atoms are more easily removed than hydrogen atoms from the carbon basal plane.

XRD was used to evaluate the reduction by examining the variations in interlayer distance of graphene powder.²³ XRD spectrum showed that the interlayer distance of the rGO (Fig. 2) was decreased to 3.58 Å (2θ = 24.6°) from 8.32 Å (2θ = 10.6°) for the original GO sheets due to the elimination of the oxygen containing groups on the GO sheets.

Fig. 2 XRD patterns of GO (A) and rGO (B).

FT-IR spectrum was also used to evaluate the reduction. FT-IR spectra of the GO and rGO were shown in Fig. 3. The GO had four characteristic peaks at 3379 cm⁻¹ (the hydroxyl groups or the adsorbed bound water), 1727 cm⁻¹ (carboxyl or carbonyl stretching vibration), 1042 cm⁻¹ (C–O stretching vibrations in C–O–C epoxide) and 1625 cm⁻¹. For the peak at 1625 cm⁻¹ in the GO and at 1573 cm⁻¹ in the rGO might be attributed to the absorbed water or the skeletal vibration of unoxidized graphitic domains.^24–26 For the rGO, the peak intensity of the oxygen functional groups became very weak compared with that of GO. Additionally, the strong peak at 1042 cm⁻¹ disappeared but at 1727 and 3379 cm⁻¹ were still observed in spectrum of the rGO. That meant the GO was partially reduced.

Fig. 3 FT-IR spectra of GO (A) and rGO (B).

Element analysis method was used to evaluate the reduction of PVDF/rGO composite films. Fig. 4 showed the corresponding EDS element analysis. The results of the EDS elemental analysis for C and O were shown in Table 1. The C content in the GO an rGO was 68.44% and 90.58%, respectively. However, the O content of the rGO decreased from 29.56% in the GO to 8.42%. The mass ratio of C/O was 2.31 in the GO and increased to 10.76 in the rGO. The C content of the PGO-0.5 and PrGO-0.5 was 65.39% and 70.80%, respectively. However, the O content in the PVDF/GO and PVDF/rGO was 1.29% and 0.50%, respectively. The O content in the rGO and PVDF/rGO indicated that some oxygen-containing functional groups remained in the samples.

Fig. 4 Energy dispersive spectroscopic analysis of GO, rGO, PGO-0.5 and PrGO-0.5.

Table 1 Results of the EDS elemental analysis for GO, rGO and PVDF composite films

Elements (wt%)	GO	rGO	PGO-0.5	PrGO-0.5
C	68.44	90.58	65.39	70.80
O	29.56	8.42	1.29	0.50
C/O ratio	2.31	10.76	50.69	141.60

---

Through the above analysis, we demonstrated that not all of the functional group in GO were decomposed when it was reduced with HBr acids, PVDF/rGO composites were prepared with partially reduced graphene, which could ensure the interactions of rGO and PVDF chains.

3.2 Dispersion of rGO

The morphology of the films was observed by the SEM characterization. Fig. 5 showed the SEM images of fracture surface of the neat PVDF, PrGO-0.1, PrGO-0.3 and PrGO-0.7. These micrographs displayed a wrinkled topography for the composite films comparing with the relatively smooth fracture surface for the neat PVDF matrix. Such a wrinkled topography indicated that the rGO sheets were encapsulated in the PVDF matrix and the presence of interactions between rGO sheets and PVDF chains.

Fig. 5 SEM images of fracture surface of the neat PVDF, PrGO-0.1, PrGO-0.3 and PrGO-0.7.

3.3 The β phase formation of the composite films

FT-IR spectra of the neat PVDF and PVDF/rGO composites were shown in Fig. 6. They revealed typical bands in the neat PVDF and PVDF/rGO composites (the peak at 3020 cm⁻¹ and 2980 cm⁻¹ are attributed to C–H stretching vibration; 1402 cm⁻¹ is deformation vibration of CH₂; 1072 cm⁻¹ is CF₂ stretching vibration; 1180 cm⁻¹ and 880 cm⁻¹ are C–C skeletal vibration). It indicated hydrobromic acids would not react with the PVDF matrix.

Fig. 6 FT-IR spectra of the neat PVDF and PVDF/rGO composites.

To reveal the crystal phase in the composite films, X-ray diffraction experiments of PVDF/GO and PVDF/rGO composites were preformed and the results were given in Fig. 7. The neat PVDF exhibited major crystalline peaks at 2θ = 17.9°, 18.5° and 26.5°, corresponding to the (1 0 0), (0 2 0) and (1 1 0) reflections of α phase, respectively. The peak appeared at 2θ = 20.2 ± 0.2° corresponding to the (1 1 0) and (2 0 0) reflections of β phase.²⁷ So the neat PVDF was the coexistence of α and β phase. The formation of the β phase in the neat PVDF was due to the dipole moment of DMF.²⁸ Compared with the neat PVDF, the peaks of all composites films at 2θ = 26.5°, 17.9° nearly disappeared and 2θ = 18.5° became rather weak, which meant that the content of α phase decreased. That was, adding only 0.05 wt% of GO can partially transform the α phase into β phase. The peaks of α phase almost all disappeared when the contents of GO were 0.1, 0.3, 0.5 and 0.7 wt%, respectively, which indicated that only 0.1 wt% of GO was sufficient to nucleate PVDF chains into nearly all β phase. The Fig. 7A showed α phase had already been transformed into β phase during PVDF/GO composite films preparation. The structure of β phase remained unchanged after the PVDF/GO composite films were reduced by hydrobromic acids, as showed in Fig. 7B.

Fig. 7 XRD patterns of the neat PVDF, PVDF/GO (A) and PVDF/rGO (B) composites.

To evaluate the relative quantity of the different phases in the neat PVDF and PVDF/rGO composites, the ratio between α and β phase was calculated with the following eqn (1):

where S_α, S_β are the integral area of α and β phases in XRD patterns after the peak-fitting of superposition of the α and β phases, respectively. The results were shown in Fig. 8. It suggested that the content of β phase of PVDF/rGO composites was greatly improved, as a comparison with that of the neat PVDF, the β phase content can reach to nearly 100% when rGO's content was greater than or equal to 0.1 wt%.

Fig. 8 F(β) derived from XRD.

FT-IR is another useful method for distinguishing different crystalline forms of PVDF. FT-IR of the neat PVDF and PVDF/rGO composites were shown in Fig. 9. The neat PVDF was characterized by transmittance bands at 615, 764, 796, and 972 cm⁻¹ correspond to the α phase; while the bands at 839 cm⁻¹ and 1274 cm⁻¹ were assigned to the β phase structure.^29–31 FT-IR results confirmed the presence of a mixture of α and β phase in the PVDF, PrGO-0.05 and PrGO-0.08 films. All α phase transmittance bands were missing with only the β phase bands found at PrGO-0.1, PrGO-0.3, PrGO-0.5 and PrGO-0.7 films. That was, nearly all β phase was formed when rGO content was only 0.1 wt%. The FT-IR results were in agreement with those obtained by XRD analysis.

Fig. 9 FT-IR spectra of the neat PVDF and PVDF/rGO composites.

A surprising consistence could be also verified from the differential scanning calorimetry (DSC) measurement. DSC heating and cooling curves of the neat PVDF and PVDF/rGO composites were shown in Fig. 10, and the crystallinity of all films and the peak melting temperatures of the crystals (T_m) were detailed in Table 2. The crystallinity was calculated according to the following equation:

where ΔH_m is the melting heat of samples. ΔH⁰_m is the melting heat of 100% crystalline PVDF, which is 103.40 J g⁻¹. ϕ is the weight percentage of rGO in the polymer matrix.^32,33

Fig. 10 DSC first-melting (A) and crystallization (B) traces of the neat PVDF and PVDF/rGO composites.

Table 2 DSC parameters and x_c of the neat PVDF and PVDF/rGO composites

Samples	Tm (°C)	ΔHm (J g^−1)	Tc (°C)	xc (%)
PVDF	160.6	43.91	130.9	42
PrGO-0.05	164.6	50.43	132.6	49
PrGO-0.08	164.8	50.34	133.3	49
PrGO-0.1	164.8	52.06	133.7	50
PrGO-0.3	165.9	50.23	133.9	49
PrGO-0.5	164.9	50.97	132.9	49
PrGO-0.7	164.3	51.86	132.5	50

---

It can be concluded that the crystallinity of PVDF/rGO composites promoted from Table 2. The neat PVDF exhibited a melting peak at 160.6 °C, the corresponding melting for the PVDF/rGO composites appeared at higher temperature about 164.3–165.9 °C, the β phase melting peak of the neat PVDF and the α phase melting peaks of the PVDF/rGO composites might be covered up because of their small numbers. As a more accepted viewpoint, the β phase of PVDF has higher melting temperature than the α phase,^34,35 as the β phase has all-trans conformation and chains can pack in the crystal more compactly causing a higher melting point. Thus, the neat PVDF was mainly α phase, and the PVDF/rGO composites became the β phase predominant, which was in good agreement with the above XRD and FT-IR results. The cooling DSC curve showed that all composites crystallized at a higher crystallization temperature (T_c) than the neat PVDF. This was because rGO tend to induce heterogeneous nucleation by providing its very large surface area for adsorption of the PVDF chain and thereby causing easier nucleation, crystallization temperature increased.

Based on the above analysis, we concluded that GO could induce crystal transformation. Absorption energy calculation of α and β polymorph by Yu et al. is used³⁶ in order to understand the cause of the β phase formation. The adsorption energy between α phase and β phase is quite different, this will increase the energy barrier between TGTG and TT structure, and results in great difficulty in the transformation from TGTG to TT structure in the process of polymer crystallization. If the energy barrier can be overcome, reduce the adsorption energy difference between the two kinds of structures, the transformation from α phase to β phase will become easier. So the β phase formation in PVDF/GO films can be explained by the complete adsorption of chains of PVDF on the GO surface with the help of the shock waves which are generated during sonication of the mixture.¹⁸ The possibility of adsorption has close relationship with the interactions between the PVDF's chains and the oxygen functional groups on the GO. This adsorption resulted in completely extension of the TGTG chains into TT conformation. The adsorption of TT conformation on the GO's surface serves as nucleating agent for the β phase formation.

Experiments showed that there were nearly all β phase when rGO's content was only 0.1 wt%, and crystallinity was 50%, which were compared to some functionalized graphene/PVDF and CNTS/PVDF composites,^18,37 but filler's content was lower. So the piezoelectric and ferroelectric properties of PrGO-0.1 film could be improved due to high β phase content, and low content of rGO did not destroy excellent performance of the PVDF.

3.4 Dielectric properties of the composite films

The dielectric constant and dielectric loss of the neat PVDF and PVDF/rGO composites were compared as function of frequency at 25 °C as shown in Fig. 11. It can be seen that a higher concentration of rGO corresponded to a larger dielectric constant and greater dielectric loss. This increase in the dielectric constant at lower frequency was ascribed to the Maxwell–Wagner–Sillars (MWS) polarization mechanism and the exchange coupling at the surface and interface.³⁸ According to the MWS polarization, charges could be accumulated at the interface when an external electric field was acted on composites due to the heterogeneous conductively of rGO and PVDF. The huge interfacial area of composites provided numerous sites for the reinforced MWS polarization. This phenomenon often appeared in low frequency because it took a long time to complete MWS polarization.

Fig. 11 Dielectric constants (A) and dielectric losses (B) as a function of frequency of the neat PVDF and PVDF/rGO composites.

The dielectric constant and dielectric loss increased slowly at lower rGO loading values, which was a typical characteristic of insulating materials. When the rGO's content was 0.5 wt% and above. The dielectric constant and dielectric loss showed a marked increase. For example, at 0.3 wt%, the values were 46 and 0.12 (1 kHz), but the corresponding values for 0.5 wt% were 122 and 0.41. Therefore, an insulator–conductor transition occurred at 0.5 wt%, as rGO was conductive filler and it uniformly dispersed in the PVDF matrix, two neighboring rGO could be regarded as electrodes with a thin insulating PVDF layer as the dielectric, together comprising a microcapacitor model,^39,40 these microcapacitors interactions had effect on improving of dielectric constant on macro basis. So the PVDF/rGO composites that were prepared by this method were also ideal candidates as high dielectric constant materials in high-charge storage capacitors. As rGO contents increased beyond the percolation threshold, the dielectric constants continued to rise, such as PrGO-0.7, although the dielectric losses were also high, these high-dielectric-constant and high-dielectric-loss composites would be used in electromagnetic-wave absorption fields.⁴¹

The dielectric constant of PrGO-0.1 was 41 (1 kHz), approximately 4 times higher than that of the neat PVDF (≈10), the dielectric loss was 0.1, which was still acceptable for applications.³⁹ The drive voltage of PrGO-0.1 decreased due to the high dielectric constant when it was used for energy conversion materials. Furthermore, the storage charge also increased during polarization of PrGO-0.1, so piezoelectric and ferroelectric properties of PrGO-0.1 were improved to a certain degree.

The breakdown field strength was an important parameter measuring the dielectric materials. Fig. 12 showed the breakdown field strength of the neat PVDF and PVDF/rGO composites. The breakdown field strength of the neat PVDF was 148 kV mm⁻¹, it decreased with the increasing of the rGO content and was 84 kV mm⁻¹ for 0.1 wt% rGO content. The reduction in the breakdown field strength indicated the presence of conductive rGO, defects and impurities in the films.⁴²

Fig. 12 Breakdown field strength of the neat PVDF and PVDF/rGO composites.

3.5 Piezoelectric properties

Fig. 13 showed the d₃₃ coefficient of the neat PVDF and PVDF/rGO composites with different rGO contents. It can be observed from the figure that d₃₃ coefficient of PVDF/rGO composites increased compared to the neat PVDF. The value of d₃₃ for PVDF was 22 pC N⁻¹, but PrGO-0.1 increased to 39.3 pC N⁻¹ due to the fact that the nearly all β phase, which were comparable to some piezoelectric ceramics/PVDF and CNTS/PVDF composites.^43,44 Though PrGO-0.3, PrGO-0.5 and PrGO-0.7 films also had nearly all β phase, their d₃₃ coefficients were lower than that of PrGO-0.1 film due to their high rGO contents, that was, the rGO was conductive filler, a part of electric energy was consumed in the form of leakage current during the polarization of the PVDF/rGO films and resulted in reduction of polarization degree. The higher rGO's content was, the more leakage current was, so the piezoelectric constant of PrGO-0.1 was higher than that of PrGO-0.3, PrGO-0.5 and PrGO-0.7.

Fig. 13 Piezoelectric constant d₃₃ of the neat PVDF and PVDF/rGO composites.

3.6 Ferroelectric properties

Fig. 14 showed the hysteresis loops for the neat PVDF and PrGO-0.1. Conductive silver slurry was used as the upper and lower electrodes. The area of the electrode was 0.1256 cm², and the thickness of the film was 50 μm. It can be observed that the remnant polarization increased for PrGO-0.1. The values of Pr of neat PVDF were 1.79 μC cm⁻², and were 3.03 μC cm⁻² for PrGO-0.1. Only the ferroelectric property of the PrGO-0.1 was compared with that of the neat PVDF due to its nearly all β phase content and low rGO contents.

Fig. 14 P–E hysteresis loop of the neat PVDF (A) and PrGO-0.1 (B).

From the studies of piezoelectric and ferroelectric properties can be seen, PrGO-0.1 composite film prepared by this method can achieve the higher piezoelectric constant and remnant polarization than those of the neat PVDF because it contained almost all β phase and the relatively low rGO contents.

3.7 Dynamic mechanical property

Analysis of storage modulus and tan δ curves had proven to be an effective tool to assess the reinforcing efficiency of rGO fillers under stress and temperature. Fig. 15 showed the storage modulus and tan δ as a function of temperature for the neat PVDF and PVDF/rGO composites. The storage modulus (G′) decreased with the increase in temperature, indicating a progressive loss at the elastic property of the films. It was noteworthy that the curve of PrGO-0.7 film had a convex plate, which resulted in increasing of the storage modulus due to the cold crystallization. The storage modulus (G′) of the neat PVDF at 40 °C was 733 MPa, which increased with increasing the loading of rGO. This increasing was rather small for the composites containing 0.05 and 0.08 wt% rGO and became significant for the high loading composites. For example, PrGO-0.1 and PrGO-0.7 composites showed a storage modulus of 1102 and 1496 MPa, which was almost 1.5–2 times that of the neat PVDF. That was due to the reinforcement by the rGO because of its high aspect ratio and good miscibility in the PVDF matrix. Furthermore, the uniform dispersion and unidirectional distribution of rGO may be the cause of the highest reinforcement.

Fig. 15 Mechanical property–temperature curves of the neat PVDF and PVDF/rGO composites: (A) storage modulus; (B) tan δ.

The loss factor tan δ was very sensitive to solid structural transformation in materials. The neat PVDF and PVDF/rGO was semi-crystallization polymer which can be viewed as two interpenetrating networks built up by the amorphous entanglement phase and the crystalline phase. Because the experimental temperature range was 40–170 °C that was well above T_g (≈−35 °C) and the experimental frequency was 1 Hz, the tan δ curves of all samples only showed a peak which was α_∐ peaks corresponded to the diffusion-like motion of chain segments in the crystallites.⁴⁵ The position of the α_∐ peak of the neat PVDF and the PVDF composites varied greatly. The α_∐ peak of PVDF was observed at around 115 °C, which was in agreement with Lovinger and Wang.⁴⁶ The α_∐ peak of PVDF composites shifted towards higher temperature (around 150 °C), where the β phase was prevalent. The better packing of β phase chains than that of α phase chains might bring additional constraints to the diffusion process.

In conclusion, PrGO-0.1 composite film prepared by this method demonstrated higher storage modulus and smaller mechanical loss, which ensured that PrGO-0.1 composite films could withstand greater stress when it was used as piezoelectric materials.

4. Conclusions

In summary, a new method was explored in this paper for the preparation of high performance PVDF/rGO composites based on the reduction of as-prepared PVDF/rGO films. Such a method not only avoided the agglomeration of GO during reduction in PVDF/GO solutions and efficiently improved the dispersion effect of rGO in PVDF matrix, but also simplified the preparation process due to no modification of GO required. As a result, it opened a feasible way to scale up the production of PVDF/rGO composite materials.

Since nearly all β phase was achieved when rGO's content was only 0.1 wt% (PrGO-0.1), compared with those of the neat PVDF, the dielectric constant increased to 41 at 1 kHz, and a high increase in piezoelectric constant and remnant polarization was also observed, being 78.6% and 69.3%, respectively. Furthermore, there was an obvious rise in the storage modulus.

The PVDF/rGO composites prepared by this method proved to have much better properties and could one day become preferable piezoelectric energy transfer and ferroelectric storage materials to the neat PVDF.

References

1. W. Heywang, K. Lubitz and W. Wersing, Piezoelectricity: Evolution and Future of a Technology (photoengraving), Peking University Press, 2013.
2. J. M. Ha, H. O. Lim and N. J. Jo, Adv. Mater. Res., 2007, 29, 363.
3. A. J. Lovinger, Macromolecules, 1982, 15, 40.
4. H.-J. Ye, L. Yang, W.-Z. Shao, S.-B. Sun and L. Zhen, RSC Adv., 2013, 3, 23730.
5. T. Lei, X. Cai, X. Wang and L. Yu, et al., RSC Adv., 2013, 3, 24952.
6. D. Y. Tang, J. S. Zhang, D. R. Zhao and L. C. Zhao, J. Mater. Sci., 2005, 40, 3339.
7. A. T. Hori, Y. Ohira and S. Yano, Composites, Part A, 2001, 32, 287.
8. Y. Pang and P. Y. Wu, J. Polym. Sci., Part B: Polym. Phys., 2004, 45, 5295.
9. P. Martins, A. C. Lopes and S. Lanceros-Mendez, Prog. Polym. Sci., 2014, 39, 683.
10. Y. Zhu, S. Murali and W. Cai, et al., Adv. Mater., 2010, 22, 3906.
11. M. Terrones, A. R. Botello-mendez and J. Campos-Delgado, et al., Nanotoday, 2010, 5, 351.
12. C. Lee, X. Wei, J. W. Kysar and J. Hone, Science, 2008, 321, 385.
13. M. D. Stoller, S. Park, Y. Zhu, J. An and R. S. Ruoff, Nano Lett., 2008, 8, 3498.
14. T. Kuilla, S. Bhadra and D. Yao, et al., Prog. Polym. Sci., 2010, 35, 1350.
15. S. C. Tjong, Nanocrystalline Materials, 2nd edn, 2014, p. 335.
16. J.-K. Yuan and S.-H. Yao, et al., J. Phys. Chem., 2011, 115, 5515.
17. S. Ansari and E. P. Giannelis, J. Polym. Sci., Part B: Polym. Phys., 2009, 47, 888.
18. R. K. Layek, S. Samanta, D. P. Chatterjee and A. K. Nandi, Polymer, 2010, 51, 5846.
19. J. c. Wang, J. Wu, W. Xu, Q. Zhang and Q. Fu, Compos. Sci. Technol., 2014, 91, 1.
20. J. I. Parades, S. Villar-Rodil, A. Martinez-Alonso and J. M. D. Tascon, Langmuir, 2008, 24, 10560.
21. J. M. Coxon and M. A. E. Townsend, Tetrahedron, 2007, 63, 5665.
22. D. C. Elias, R. R. Nair, T. M. G. Mohiuddin, S. V. Morozov, P. Blake and M. P. Halsall, et al., Science, 2009, 323, 610.
23. S. Pei and J. Zhao, et al., Carbon, 2010, 48, 4466.
24. J. Wang, S. J. Wang and J. Kim, J. Colloid Interface Sci., 2009, 336, 592.
25. M. Acik, C. Mattevi, C. Gong and G. Lee, et al., ACS Nano, 2010, 10, 5861.
26. J. Wu, X. Shen, L. Jiang, K. Wang and K. Chen, Appl. Surf. Sci., 2010, 256, 2826.
27. Yuan, Z. Dang, S. Yao, J. Zha, T. Zhou, S. Li and J. Bai, J. Mater. Chem., 2010, 20, 2441.
28. M. Benz, B. William and O. J. Gregory, Macromolecules, 2002, 35, 2682.
29. A. Salimi and A. A. Yousefi, Polym. Test., 2003, 22, 699.
30. P. Sajkiewicz, A. Wasiak and Z. Golclowski, Eur. Polym. J., 1999, 35, 423.
31. T. Boccaccio, A. Bottino, G. Capannelli and P. Piaggio, J. Membr. Sci., 2002, 210, 315.
32. X. Li, Y. Wang, X. Lu and C. Xiao, J. Membr. Sci., 2008, 320, 477.
33. L. He, Q. Xu, C. Hua and R. Song, Polym. Compos., 2010, 921.
34. J. S. Andrew and D. R. Clarke, Langmuir, 2008, 24, 670.
35. R. Song, D. B. Yang and L. H. J. He, J. Mater. Sci., 2007, 42, 8408.
36. S. S. Yu, W. T. Zheng, W. X. Yu, Y. J. Zhang, Q. Jiang and Z. D. Zhao, Macromolecules, 2009, 42, 8870.
37. L. He, G. Xia, J. Sun, Q. Zhao, R. Song and Z. Ma, J. Colloid Interface Sci., 2013, 393, 97.
38. H. B. Lu, X. Y. Zhang and H. Zhang, J. Appl. Phys., 2006, 100, 054104.
39. F. He, S. Lau, H. L. Chan and J. Fan, Adv. Mater., 2009, 21, 710.
40. J.-K. Yuan, Z.-M. Dang, S.-H. Yao, J.-W. Zha, T. Zhou, S.-T. Li and J. Bai, J. Mater. Chem., 2010, 20, 2441.
41. D. D. L. Chung, Carbon, 2012, 50, 3342.
42. H. Stoyanov, M. Kollosche, S. Risse, D. N. McCarthy and G. Kofod, Soft Matter, 2011, 7, 194.
43. X. Guan, Y. Zhang, H. Li and J. Ou, Sens. Actuators, A, 2013, 194, 228.
44. N. Levi, R. Czerw and S. Xing, et al., Nano Lett., 2004, 4, 1267.
45. Y. Men and J. Rieger, Macromolecules, 2003, 36, 4689.
46. A. J. Lovinger and T. T. Wang, Polymer, 1979, 20, 725.

---