Introduction

During the past few decades, supercapacitors, also known as electrochemical capacitors (ECs), are promising power sources and have been widely used in portable electronics, hybrid electric vehicles, digital communication systems, and uninterrupted power supply of high-power devices [1–3]. Based on the charge-storage mechanism, ECs include pseudocapacitors and electrical double layer capacitors (EDLCs) [4]. Nevertheless, as a precondition for practical application, there are still several pivotal issues for ECs that need to be improved, for example, the rate capability, specific capacitance, cycling stability, cost, etc. [1, 2, 5, 6]. From a materials science point of view, these issues concerning ECs are closely related to the electrode and electrolyte materials used. In order to enhance the property of ECs, most of the research is concentrated on using innovative electrode materials, suitable electrolytes, and tuning the electrolyte/electrode interface performances. The consecutive development of nanometer-sized materials benefits the progress of supercapacitor technologies a lot. Recently, carbon-based nanofillers, such as carbon nanotubes (CNTs), activated carbon, expanded graphite, and graphene (GN), with high surface area have been widely used for ECs [7, 8].

Meanwhile, graphene oxide (GO), a single sheet of graphite oxide, whose edges and basal planes are abundant with various oxygen functional groups (−OH, −COOH, −CHO, epoxy groups), has also attracted great interest. It can be easily prepared in large scales by treating natural graphite with strong aqueous oxidizing agents. Due to its intercalating, superior mechanical, ion-exchange properties and low cost [9–11], it is preferred over other expensive fillers like GN and CNTs. Also, thanks to its abundant oxygen-containing functional groups, GO possesses excellent hydrophilic properties and can therefore be readily dispersed in water. This provides abundant opportunities for the formation of GO-based nanocomposites. So far, several GO-based polymer composites have been developed, and the mechanical, thermal, and electrical properties of these composites were found to be augmented, as reported [12, 13].

Polypyrrole (PPy) is a promising electrode material for ECs with unusual doping/dedoping processes, good electrical conductivity, low cost, environmental stability, and simple synthesis. However, it also exhibits poor stability and rate capability, which limit its wide application [14]. Significantly, composites based on PPy and GO have shown synergetic effects, for example, enhancement in capacitances, stability, electrical conductivity, and electrochemical cyclability [15, 16]. Feng et al. fabricated polypyrrole/modified graphite oxide (PPy/MGO) composites by in situ polymerization and demonstrated that its specific capacitance is 202 F g−1 at a current density of 1 A g−1 and the capacitance retention of PPy/MGO is 169 F g−1 after 1,000 cycles at a scan rate of 1 A g−1 [17]. Konwer et al. fabricated PPy/GO composites via in situ polymerization of pyrrole in the presence of GO in various proportions (5 and 10 %). High specific capacitance of the PPy/GO composite of 421.4 F g−1 was obtained in the potential range from 0 to 0.80 V at 2 mA current compared with 237.2 F g−1 for pure PPy by galvanostatic charge–discharge analysis. Incorporation of GO into the PPy matrix has a pronounced effect on electrochemical capacitance performance of PPy/GO nanocomposites [18]. In most instances, preparation of conducting polymer nanocomposites is conducted by an in situ polymerization technique. But the performance of ECs assembled in this way still has room for improvement as the GO sheets tend to stack together easily in this method, and it has been suggested that these stacks become in fact barriers to electron transport [19, 20].

This work focuses on the combination of the complementary properties of GO and PPy. Herein, careful studies on the intercalation of PPy into layered GO by an in situ emulsion polymerization method in the presence of sodium dodecylbenzene sulfonate (SDBS) have been carried out. First, GO can be delaminated into sheets and dispersed in SDBS aqueous solution to form SDBS-GO (soft-hard) templates. The surfactants intercalated between GO sheets can effectively inhibit aggregation of GO sheets during the process of composition. Then, PPy particles were synthesized via emulsion polymerization and spontaneously assembled onto the GO sheets under the π-π interaction between PPy and the unoxidized domain of GO sheets. To the best of our knowledge, no previous work has been reported for preparing polypyrrole-graphene oxide (PPyGO) composite by using SDBS as a soft template and stabilizer. This method is a more controllable and slower reaction than conventional polymerization techniques. Moreover, it is a facile and novel approach to synthesize bulk quantities of PPyGO nanocomposites. For comparison, PPyGO composites prepared without SDBS and PPy particles alone were prepared using the same method. The obtained PPyGO nanocomposites using SDBS as soft templates exhibited excellent electrochemical and cycling properties.

Experimental

Chemicals

Graphite and SDBS were supplied by Alfa Aesar and Nanjing Chemical Reagent Co. Ltd., respectively. Pyrrole (Py) (98 %) was purchased from Aldrich and distilled under vacuum prior to use. All other reagents were of analytical grade and used as received without further purification.

General characterization

The microstructure and morphology of the samples were observed by a scanning electron microscope (SEM) (Hitachi S-4800) with an acceleration voltage of 5 kV and a transmission electron microscope (TEM) (JEM-2100) with an acceleration voltage of 200 kV, respectively. The Fourier transform infrared spectroscopy (FTIR) spectra were recorded on a TENSOR 27 FTIR spectrometer using KBr pellets. X-ray diffraction (XRD) patterns were obtained on a D8 ADVANCE (Bruker AXS, Germany) using Cu/Ka radiation (λ = 1.5406 Å) radiation.

Preparation of graphene oxide

The graphite oxide was prepared from natural graphite powder using a modified Hummers method [21]. For purification, the product was washed several times with 5 % of HCl and distilled water. The product was exfoliated by ultrasonication for 2 h. Finally, a homogeneous GO aqueous dispersion was obtained and used for further preparation of PPyGO-SDBS nanocomposites.

Preparation of PPyGO-SDBS nanocomposites

PPyGO-SDBS nanocomposites were prepared using an in situ emulsion polymerization method with Py and GO in the presence of SDBS. The different mass ratios of Py to GO prepared were as follows: 95:5, 90:10, 85:15, 80:20, 75:25, 65:35, and 50:50. The resulting composites were named as PPyGOratio-SDBS. For example, PPyGO80:20-SDBS indicates that the mass ratio of Py to GO is 80:20. In a typical experiment, firstly, 6.5 mmol of SDBS was dissolved in 50 mL of distilled water. The solution was transferred into a three-neck flask and chilled at 0 ºC. Then, 0.45 mL of Py was added into the SDBS solution and stirred for 2 h. Secondly, the required amount of GO solution was subjected to ultrasonic treatment for 2 h and added into the above mixture and then stirred at 0 °C for 48 h. Thirdly, 2 g of FeCl3⋅6H2O was dispersed in 0.12 M HCl (50 mL) and slowly dropped into the above mixture. The reaction was carried out with stirring for 24 h at 0 °C. Finally, the reaction mixture was filtered, washed with distilled water and ethanol, and dried at 40 °C for 12 h to obtain the PPyGO80:20-SDBS composites. For comparison, PPyGO80:20 composites without SDBS and PPy particles with SDBS were prepared by the same method.

Electrochemical performance tests

The three-electrode cell system was used to evaluate the electrochemical performance of the prepared electrode materials in 1 M KCl aqueous electrolyte. The working electrode was made by mixing 80 wt% active material, 10 wt% acetylene black, and 10 wt% poly(vinylidene fluoride) in N-methyl-2-pyrrolidone, and the slurry was coated onto a 1-cm × 1-cm nickel foam current collector and dried at 40 °C for 8 h to evaporate the solvent. A platinum sheet and a standard calomel electrode (SCE) were used as counter electrode and reference electrode, respectively. Cyclic voltammetry (CV), galvanostatic charge–discharge analysis, and electrochemical impedance spectroscopy (EIS) were conducted using a PARSTAT2263 electrochemical workstation. The electrical conductivity was determined by a four-probe instrument (RTS-8) at room temperature. The samples were compacted into pellets of 15 mm diameter and about 0.5 mm thickness under pressure of 20 MPa.

Results and discussion

The synthesis mechanism of PPyGO-SDBS nanocomposites

The synthesis mechanism of the PPyGO-SDBS nanocomposites is illustrated in Scheme 1. First of all, SDBS formed a mass of micelles in aqueous solution and the Py molecules entered into the above micelles due to its good oil solubility. Secondly, graphite oxide was delaminated into GO nanosheets in the SDBS solution. GO is reported to be uniquely amphiphilic with negatively charged hydrophilic edges and hydrophobic basal plane [22]. SDBS is thought to assemble on the surface of GO sheets through π-π interactions and hydrophobic interactions [23]. Finally, with the addition of FeCl3, PPy particles were synthesized and assembled onto the GO sheets owing to the π-π interaction between PPy and the unoxidized domains of the GO sheets.

Scheme 1
scheme 1

The synthesis mechanism of PPyGOratio-SDBS nanocomposites

Full size image

Structure and morphology studies

The structure and morphology of GO, PPy particles, PPyGO80:20, and PPyGO80:20-SDBS were investigated by SEM (Fig. 1) and TEM (Fig. 2). The PPy particles prepared by the emulsion polymerization with SDBS are shown in Fig. 1a. The PPy looks like cauliflower agglomerated with spherical particles. Particle sizes are slightly less than 100 nm. As depicted in Fig. 1b, GO exhibits smoother plates and multilayered lamellar flakes with different sizes stacked together which is consistent with the literatures [24, 25]. In Fig. 1c, the PPyGO80:20 agglomerated into big lumps and its layer structure is not obvious as that of the PPyGO80:20-SDBS shown in Fig. 1d. A careful inspection of the Fig. 1d suggested that there are some typical characteristics that are worthy of mention. Firstly, the PPy particles seem to be stitched together, and a flaky, rough morphology can be observed on the surface of the GO. Furthermore, the thickness of the obtained composite nanosheets is much thicker than bare GO owing to the depositing and insertion of PPy. Comparing panel d with c of Fig. 1, we can deduce that SDBS plays a pivotal role to disperse the PPy particles between the interlayers of the GO and obtain the more homogeneous composites compared with that without SDBS.

Fig. 1
figure 1

SEM images of a PPy particles, b pure GO, c PPyGO80:20, and d PPyGO80:20-SDBS

Full size image
Fig. 2
figure 2

TEM images of a PPy particles, b GO, c PPyGO80:20, and d PPyGO80:20-SDBS

Full size image

More details can be observed from the TEM images of the different samples. In Fig. 2a, the PPy particles gather together and many lumps could be observed which is consistent with the result of the SEM. In Fig. 2b, GO shows an overlapped lamellar structure and the obvious wrinkle of an individual GO sheet is shown. Compared with Fig. 2b, PPyGO80:20-SDBS composites in Fig. 2d are fuzzy and not so smooth as the GO due to the polymerization of PPy covering the surface of the GO sheets, which has been confirmed by the SEM image of Fig. 1d. In Fig. 2c, the PPy particles are deposited between the interlayers of the GO and are not well distributed, so it seems that clusters of PPy particles agglomerated with GO which may influence their electrochemical performance.

The structure and component of the prepared PPyGO nanocomposites were characterized by Fourier transformed infrared (FTIR) spectra and XRD. Figure 3a shows the FTIR spectra of GO, PPy particles, and PPyGO80:20-SDBS composites. GO exhibits the following characteristic absorptions, for example, the C=O stretching vibration peak at 1,732 cm−1, the vibration and deformation peaks of O–H groups at 3,386 and 1,417 cm−1, and the C–O (alkoxy) stretching peak at 1,057 cm−1 [24, 26]. Compared with GO, several new peaks ascribed to PPy appeared in the spectrum of PPyGO80:20-SDBS. The new peaks at 1,548, 1,458, and 3,437 cm−1 are ascribed to the C–C, C–N, and N–H stretching vibration in the PPy ring. In addition, the bands at 2,925 and 2,851 cm−1 attributed to the asymmetric stretching and symmetric vibrations of CH2 in the SDBS were also observed [27, 28]. It should be remarked that the peak due to the C=O group within the PPyGO80:20-SDBS has been downshifted to 1,700 cm−1 which is maybe due to the π-π interactions and hydrogen bonding between the GO films and aromatic polypyrrole rings [29, 30]. The obtained PPyGO nanocomposites were further studied by powder XRD measurements, as shown in Fig. 3b. PPy particles exhibit a weak and broad diffraction peak at 2θ = 25.9° (d = 0.343 nm), which indicates that the PPy is amorphous. XRD patterns of the GO showed an intense, sharp peak centered at 2θ = 9.98°, corresponding to an average interplanar distance (d) about 0.886 nm of the GO. In the case of the PPyGO80:20-SDBS and PPyGO80:20, because the GO surface has been covered by PPy, the peak at 2θ = 9.98° disappeared and the characteristic peak (2θ = 25.9°) of PPy becomes much weaker and broader. However, in comparison with PPyGO80:20-SDBS (2θ = 25.9°), the diffraction peak of PPyGO80:20 is weaker and shifts to a higher angle (2θ = 27.1°). The interplanar space of PPyGO80:20 is about 0.329 nm, a little less than that of the PPyGO80:20-SDBS, implying that the interplanar space tends to expand with the assistance of SDBS and the PPy particles entering into the interlayer of GO more easily. This result is confirmed by Figs. 1 and 2.

Fig. 3
figure 3

a FTIR spectra of pristine GO, PPy particles, and PPyGO80:20-SDBS. b XRD patterns of pure GO, PPy particles, PPyGO80:20, and PPyGO80:20-SDBS

Full size image

Electrical properties

Conductivity

The room temperature average conductivities of PPy particles and PPyGO composites are summarized in Table 1. The conductivity of PPy particles prepared by utilizing SBDS as the template is about 0.020 S m−1 and that of the obtained PPyGO80:20 composites without SDBS is 0.023 S m−1. The value of the above two materials is similar but several orders of magnitude higher than that of GO (1.28 × 10−6 S cm−1) and pure PPy (1.08 × 10−4 S cm−1) without SDBS. Interestingly, PPyGO80:20-SDBS shows better conductivity (0.280 S m−1) over ten times greater than the other two samples. The conductivity of the PPyGO80:20-SDBS seems to increase with low GO content and upon doping with the SDBS. The doping of PPy with SDBS and composition with GO due to π-π conjugation induces a great increase of the conductivity of the PPyGO80:20-SDBS. More importantly, GO was dispersed well in SDBS solution which is beneficial to PPy polymerization between intercalated GO flakes, and GO sheets may serve as a bridge connecting PPy conducting domains and increase its effective percolation. Furthermore, the π-π electron stacking between the Py ring of PPy and unoxidized domain of GO could divert electrons more effectively [28].

Table 1 The conductivity, C sp, power, and energy densities of different electrode materials
Full size table

EIS analysis

The electrode conductivity and ion-transport kinetics were further characterized by EIS. EIS was recorded in the frequency range from 105 to 0.1 Hz at open circuit potential with alternate current amplitude of 10 mV. The resulting Nyquist plots for PPy particles, PPyGO80:20, and PPyGO80:20-SDBS are shown in Fig. 4. In the region with low frequencies, the slope of the plot for the PPyGO80:20-SDBS composites is steeper than that of PPy particles and PPyGO80:20, indicating that PPyGO80:20-SDBS possesses better capacitive behavior (vertical line for an ideal capacitor) and lower diffusion resistance of ions [33]. In the higher frequency region, the real axis intercept is the equivalent series resistance (R s), and the radius of the semicircle plotted is indicative of electrode conductivity and the charge transfer resistance (R ct) in the electrode materials [34]. Careful inspection of the plots at higher frequencies reveals that the PPyGO80:20-SDBS exhibits a smaller semicircle than PPy particles and PPyGO80:20 which illustrated that the PPyGO80:20-SDBS had a lower R ct and much faster charge transfer rate. SDBS not only assists PPy to insert successfully into GO interlayers and form the interpenetrating conducting structure, but also enhances surface wettability by electrolytes [35] and, hence, greatly improves the conductivity of the nanocomposites, which is favorable for their supercapacitor applications.

Fig. 4
figure 4

Nyquist plots of the PPy particles, PPyGO80:20, and PPyGO80:20-SDBS in 1 M KCl solution measured at open circuit potential

Full size image

The CV characteristics and galvanostatic charge–discharge analysis

PPy particles, PPyGO80:20, PPyGO50:50-SDBS, PPyGO65:35-SDBS, PPyGO80:20-SDBS, and PPyGO95:5-SDBS were examined by CV in 1 M KCl aqueous solution in order to test the potential feasibility as supercapacitors. The cyclic voltammetry curves (CVs) with a potential range from −0.2 to 0.6 V at a scan rate of 5 mV s−1 are shown in Fig. 5a. All of the samples indicate reasonable symmetrical characteristic except PPyGO80:20. In general, compared to PPy particles, the five PPyGOratio-SDBS nanocomposites with different mass ratios have larger electrochemical response currents. It is clear that the response current of PPyGO80:20-SDBS is the largest, illustrating that its specific capacitance is the highest, which is consistent with the results of the galvanostatic charge–discharge analysis as shown in Fig. 5d. Moreover, Fig. 5b exhibits the CVs of the PPyGO80:20-SDBS composite electrode at different scan rates from 5 to 100 mV s−1. The shapes of CVs show symmetric current-potential characteristics, meaning a high reversibility and efficiency [36]. The increase of current with the scan rates means good rate ability for the PPyGO80:20-SDBS electrodes [37]. But the rate ability of PPyGO80:20-SDBS still requires improvement, something which will be explored in our further work.

Fig. 5
figure 5

a CVs of PPy particles, PPyGO80:20, PPyGO50:50-SDBS, PPyGO65:35-SDBS, PPyGO80:20-SDBS, and PPyGO95:5-SDBS at a scan rate of 5 mV s−1. b CVs for PPyGO80:20-SDBS measured at various scan rates. c Galvanostatic charge–discharge curves of PPy particles, PPyGO80:20, and PPyGO80:20-SDBS at a current density of 0.5 A g−1. d The specific capacitances of PPy particles, PPyGO80:20, and PPyGOratio-SDBS electrodes at different current densities. The mass of the sample is 3 mg

Full size image

Figure 5c exhibits the galvanostatic charge–discharge curves (GCD) of PPy particles, PPyGO80:20, and PPyGO80:20-SDBS electrodes at current densities of 0.5 A g−1. The near triangular shape of the curves indicates that the materials have good reversibility. The specific capacitance of the PPyGO80:20-SDBS is much higher than that of PPy particles and PPyGO80:20 under the same current density, which can be clearly found in Fig. 5c, d. The specific capacitance (C sp) is calculated from the discharge process according to the following Eq. (1) [38]:

$$ {C}_{\mathrm{sp}}=\left( I\varDelta t\right)/\left( m\varDelta V\right) $$
(1)

where C sp is the specific capacitance (F g−1), I is the current (A), Δt is the discharge time (s), m is the mass of active materials in the electrode (g), and ΔV is the potential window (V).

In addition, as shown in Fig. 5d, the C sp for PPyGO80:20-SDBS composites is the highest (483 F g−1) among all the samples, which exceeds that of PPy particles (270 F g−1) and PPyGO80:20 (364 F g−1), at a discharge current density of 0.2 A g−1. The C sp of PPyGO80:20-SDBS composites still remained as high as 315 F g−1 even at a high discharge current density of 5 A g−1, while those of PPy particles and PPyGO80:20 sharply decrease to 25 and 164 F g−1 (Table 1), respectively. It is noted that the PPyGO80:20-SDBS composites show not only high specific capacitance but also better rate capability, demonstrating that the 80:20 ratio is the most optimal, of all the samples tested. This result is important and has not been involved in the previous reports concerning the PPyGO nanocomposites [17, 31]. Based on the above discussion, the better rate capability and improved capacitance of PPyGO80:20-SDBS might be mainly attributed to the reaction of the SDBS with GO which assisted the complete intercalation of PPy into GO layers to form interpenetrating conducting structure. The obtained PPyGO80:20-SDBS had a more homogeneous layer structure compared to that of PPyGO80:20, and SDBS is also known to enhance surface wettability [35] and facilitate electron transport and ion insertion/extraction in the electrode material during the quick charge–discharge processes. So out of all the different ratios of PPy/GO tested, PPyGO80:20-SDBS nanocomposites exhibited the best electrochemical characteristics such as specific capacitance, rate capability, and cycling stability.

The cycling stability

Long cycling life is another important property for supercapacitors. The electrochemical stability of PPy particles, PPyGO80:20, and PPyGO80:20-SDBS electrodes is investigated at a current density of 2 A g−1 , and the results are shown in Fig. 6a. It is found that there are no significant differences among these three samples at the beginning of 400 cycles. After 1,000 charge–discharge cycles, the specific capacitance of PPyGO80:20-SDBS composites still remains 80 % of the initial capacitance, which was much higher than that of PPy particles (50 %) and PPyGO80:20 (66 %), indicating good cycling stability of PPyGO80:20-SDBS as electrode material. The great performance of the PPyGO80:20-SDBS could be attributed to the following two reasons: (1) The addition of SDBS is favorable for GO to delaminate and PPy to deposit on the surface of GO. The homogeneous PPyGO-SDBS nanocomposites reduce the strain associated with the volume change of PPy during the charge–discharge processes and, hence, avoid the destruction of the electrode material. (2) SDBS enhanced the surface wettability and accelerated the electron transport and ion insertion/extraction in the electrode material during the quick charge–discharge processes [35].

Fig. 6
figure 6

a Cycle stability of PPy particles, PPyGO80:20, and PPyGO80:20-SDBS electrodes from the 1st to the 1,000th cycle at a current density of 2 A g−1. b Ragone plots of PPy particles, PPyGO80:20, and PPyGO80:20-SDBS

Full size image

The Ragone plot analysis

A Ragone plot is used to compare the performance of various energy-storing devices [39]. The energy density and power density of PPyGO80:20-SDBS, PPyGO80:20, and PPy in the three-electrode system are estimated from galvanostatic discharge curves at different currents and marked in the Ragone plot, as shown in Fig. 6b. The energy density values of PPyGO80:20-SDBS electrode are in the range of 42.93 ~ 22.88 Wh kg−1, while the power density values are in the range of 80 ~ 6,000 W kg−1, revealing that there is a decrease in energy density of the electrode material with an increase of power density. Compared with PPyGO80:20-SDBS, PPyGO80:20, and PPy electrodes, the energy density of PPyGO80:20-SDBS is the highest under the same power density. Moreover, we provided power and energy densities of similar structured electrode materials (Table 1), to compare with our sample (PPyGO80:20-SDBS). Obviously, power and energy densities of the PPyGO80:20-SDBS are much higher than that of the similar electrode materials prepared by in situ polymerization. Based on the excellent electrochemical properties, the PPyGO-SDBS nanocomposites could be applied as electrode materials for electrochemical supercapacitors.

Conclusions

PPyGO-SDBS composites have been successfully prepared which exhibit excellent electrochemical characteristics including high specific capacitance, great rate capability, and cycling stability. SDBS acting as stabilizer and dopant played an important role in facilitating PPy particles to enter into interlayers of GO. The specific capacity of the obtained PPyGO80:20-SDBS composites was up to 483 F g−1 at a current density of 0.2 A g−1 and retained above 80 % after 1,000 charge–discharge processes. Further optimization and control of the structures to develop better electrochemical properties of GO-based composites are under investigation in our lab.