Introduction

In recent years, researchers have been exploring various novel energy storage materials to meet the needs of renewable energy resources, and portable electronic devices [1, 2]. Supercapacitor, a kind of promising energy storage device, has been receiving more and more attention due to its high-power density, rapid charging and discharging rate, long cycle life, and wide working temperature range [3,4,5,6,7,8,9,10].

Supercapacitors are usually divided into double-layer capacitors and pseudocapacitors [11]. The double-layer capacitor materials are mainly carbon materials such as graphite, graphene, reduced graphene oxide (rGO), carbon nanotubes (CNTs), carbon nanofibers (CNFs), and porous carbon [5, 12,13,14]. Among various carbon materials, graphene or rGO with good electrical conductivity, large specific surface area (about 2630 m2 g−1), high flexibility as well as chemical and thermal stability has been applied widely for supercapacitors [15,16,17,18,19]. For example, the cross-linked N-doped graphene showed superior capacitive performance [20]. Various preparation methods have been developed to prepare graphene and rGO, such as chemical vapor deposition [21,22,23], epitaxial growth method [24,25,26,27], chemical/thermal reduction method [28, 29], and solvent hot method [30, 31]. However, these methods are very complicated and the specific capacitance of the as-obtained graphene and rGO is low. And its stability and mechanical properties of graphene are relatively poor owing to the weak interaction between the graphenes, which increases contact resistance between nanosheets and substrates. As a result, the cycle performance and the energy density of capacitor are very poor.

Pseudocapacitor electrode materials mainly come from redox-active materials, including conducting polymers [32, 33] and metal oxide nanoparticles (NPs) (Fe3O4 NPs, Co3O4 NPs, etc.) [34,35,36,37,38]. Among various metal oxides, Fe3O4 is an important one for supercapacitors because of its low cost, abundance, environmental compatibility, and high theoretical capacity. Accordingly, various Fe3O4 NPs have been extensively investigated as the electrode materials for a supercapacitor. For example, Wang et al. prepared Fe3O4 NPs by using FeCl3 and ethanolamine and it exhibited remarkable pseudocapacitive activities [39]. However, Fe3O4 NPs just provide a low capacitance (< 300 F g−1) in aqueous electrolyte due to their poor electrical conductivity. As we all know, incorporating Fe3O4 NPs into carbon materials with good electrical conductivity could improve their capacitance performance significantly [34, 35]. For example, Fe3O4-doped graphene showed desired capacitance and high-energy density [35]. The hybrid of Fe3O4-doped porous carbon nanorods supported by three-dimensional (3D) kenaf stem-derived macroporous carbon exhibited 220.5 F g−1 after 5000 cycles at 2.0 A g−1 [40]. MIL-88A-derived Fe3O4-carbon hierarchical nanocomposites were also applied for electrochemical sensing, and it also showed good performance [41]. Thus, the design of novel Fe3O4/carbon nanocomposite is very significant for achieving high performance of supercapacitors.

In this work, the commercial carbon paper which was composed of graphene layers was treated by a two-step electrochemical peeled method to form functional exfoliation graphene (FEG). Some typical layered fold graphenes were successfully peeled off and stood on the carbon paper matrix to form 3D graphene foam. The peeled graphene provided a large surface area and greatly improved the electron transfer ability of pseudocapacitance materials. Then, the FEG/carbon papers were further functionalized by benzene carboxylic acid to form FFEG/carbon papers. Finally, a large number of small Fe3O4 NPs were grown on the surface of FFEG/carbon paper. The as-synthesized Fe3O4 NPs/ FFEG/carbon paper nanocomposites were directly used as supercapacitor electrode. The specific capacitance was about 470 F/g−1 under current density of 1 A g−1, and it still maintained 432 F g−1 after 5000 cycles.

Experimental

Reagents and materials

Commercial carbon papers were obtained from Physical and Chemical Company (Hong Kong, China). FeCl3·6H2O, KNO3, K2CO3, and p-aminobenzoic acid were purchased from Aladdin Chemistry Co., (Ltd Shanghai, China). NaAc, ethylene glycol, Na2SO4, Na2HPO3, and NaH2PO3 were purchased from Sinopharm Chemical Reagent (Shanghai, China). Other reagents of analytical grade were obtained from Shanghai Reagent Co., Ltd. (Shanghai, China). All reagents were of analytical grade and used as received. Platinum sheet electrode holder was obtained from Tianjin Joint Company (Tianjin, China). All solutions were prepared with ultrapure water, purified by a Millipore–Q system (18.2 MΩ cm).

Characterization

Scanning electron microscopy (SEM) images were obtained with a Hitachi S3400N at an accelerating voltage of 20 kV equipped with an energy dispersive spectrometer (EDS). Raman spectra were recorded on a LabRAM HR (Horiba Jobin Yvon) with a 633-nm laser. Cyclic voltammograms (CVs), electrochemical impedance spectroscopy (EIS), and constant current charging/discharging tests (CP) were performed using a CHI760e electrochemical workstation (CH Instruments Inc., China) at room temperature. CVs and CP were carried out in 1 M Na2SO3 with a three-electrode system by using Fe3O4 NPs/FFEG/carbon paper as the working electrode, a saturated calomel electrode (SCE) as the reference electrode, and a platinum electrode as the counter electrode. And the binder was not used in the electrode preparation. EIS measurements were performed in 0.1 M KCl containing 5.0 mM Fe(CN)6 3−/4- in the frequency range from 100 kHz to 0.1 Hz at open circuit potential by applying a 5-mV signal. CP tests were tested at different current density at the potential window between − 0.6 V and 0.1 V. The specific capacitance of single electrode was calculated according to the following Eq. (1) (galvanostatic charge/discharge curves):

$$ {C}_s=\frac{I\Delta t}{\mathrm{m}\left({V}_a-{V}_c\right)} $$
(1)

where C s is the specific capacitance (F g−1), m is the mass of Fe3O4 (g), (V a  − V c ) represents the potential window (V), I is the discharge current (A), and Δt is the discharge time (s).

Preparation of FEG/carbon paper

Firstly, the commercial carbon papers were pre-treated by ultrasound in acetone, ethanol, and distilled water, respectively, and then dried at 60 °C. After that, the carbon paper was cut into small pieces with 1 × 1.5 cm size. The two-step peeled method was carried out by using carbon paper as the working electrode whose working area was 1 × 1 cm2 (Fig. S1, Supporting Information). The carbon paper electrode was firstly treated by CVs scan in 0.5 M K2CO3 solution at the scanning potential from 0.5 to 1.8 V at the scan rate of 20 mV s−1 for 6 cycles. Next, it was further scanned in 1.0 M KNO3 solution at the scanning potential of − 0.9–1.9 V at the scan rate of 20 mV s−1 for 20 cycles. Finally, the carbon paper was removed out and rising by ultrapure water to remove off inorganic residues. The carbon paper electrode was dried in the oven to obtain the FEG/carbon paper for next time. The preparation process of FEG/carbon paper was illustrated in Fig. 1.

Fig. 1
figure 1

Schematic illustration of fabrication process of Fe3O4 NPs/FFEG/carbon paper

Full size image

Benzene carboxylic acid functionalized FEG/carbon paper

The 4-benzoate diazonium salt was prepared as follows. The p-aminobenzoic acid (480 mg) was firstly dissolved in 40 mL ultrapure water solution which contained NaOH (140 mg). Then, 236 mg NaNO2 was slowly added into the above solution in the condition of ice bath. After that, 6.4 M HCl (3 mL) was quickly added. After 45 min, the light yellow 4-benzoate diazonium salt was obtained. The FEG/carbon paper electrode was immersed in the above 4-benzoate diazonium salt and scanned at the working potential window between − 0.60 and 0.3 V for scanning 20 cycles at the scan rate of 100 mV s−1. After 20 cycles, the 4-benzene acid was grafted to FEG/carbon paper by electrochemical reduction and growth to form the FFEG/carbon paper. The preparation process of FFEG/carbon paper was illustrated in Fig. 1.

Preparation of Fe3O4 NPs/FFEG/carbon paper

The growth of Fe3O4 NPs on FFEG/carbon paper was carried out as follows. Firstly, 0.34 g FeCl3 was mixed with 10 mL ethylene glycol under stirring for 10 min. The solution was then added into 0.25 g NaAc under ultrasound. After that, the hybrid solution was transformed into 25 mL Teflon-lined stainless steel and then the as-prepared FFEG/carbon paper (250 mg) was immersed in the Teflon-lined stainless to make the solution cover FFEG/carbon paper absolutely. Subsequently, the Teflon-lined stainless steel was put in the oven and reacted at 200 °C. After 10 h, the FFEG/carbon paper was took out and rinsed by ethanol and water several times to obtain the Fe3O4 NPs/FFEG/carbon paper. The weight of Fe3O4 NPs loaded on the FFEG/carbon paper was obtained according to the quality of FFEG/carbon paper before and after loading. As a comparison, the Fe3O4 NPs/FEG/carbon paper was also prepared by the same procedure. The preparation process of Fe3O4 NPs/FFEG/carbon paper was illustrated in Fig. 1.

Results and discussion

The characterization of FFEG/carbon paper

The preparation of FEG/carbon paper was monitored by CVs. As shown in Fig. S2A (Supporting Information), the current started to increase at about 1.1 V and rapidly raised near 1.2 V, which was attributed to the oxygen evolution reaction. Therefore, it produced lots of oxygen around the carbon paper surface. Due to many defects around the periphery of carbon paper, the small bubbles of oxygen made graphene partly to be peeled off. Finally, 3D graphene foam was formed on carbon paper after six CVs. Then, the electrode was transferred in 1.0 M KNO3 and was canned in the scanning potential of − 0.9–1.9 V at the scanning rate of 20 mV−1. As shown in Fig. S2B (Supporting Information), the current rapidly raised near 1.2 V, indicating the graphene continued to be peeled off. In the reverse scan, several peaks were found and they could be ascribed to the reduction of oxygen-containing groups on carbon paper. Figure S3 (Supporting Information) showed the CVs of obtained FEG/carbon paper electrode in 4-benzoate diazonium salt solution scanned in the working potential window between − 0.60 and 0.3 V for 20 cycles. As shown in Fig. S3 (Supporting Information), a reduction peak was found at − 0.1 V which was attributed to the reduction of 4- benzene carboxylic diazonium salt cations. In the reverse scan, an oxidation peak appeared at 0.2 V which was attributed to the oxidation of benzene carboxylic. With the increase of scanning cycles, cathodic peak current decreased and the cathode potential was gradually moving in the positive direction, indicating 4-benzene carboxylic functional group was successfully modified on the surface of FEG to form FFEG/carbon paper.

As shown in Fig. 2a, the surface of carbon paper was very smooth and no typical graphene layer was observed. After the first step to peel off, some typical graphene layers were peeled off by the production of oxygen bubble (Fig. 2b). The large hole might originate from the oxygen bubble. The parts of graphene layers still stood on the surface of carbon paper. After the secondary oxidation stripping, more graphene layers were produced and they formed 3D graphene foam (Fig. 2c). Such 3D porous structure could promote diffusion of ions and accordingly it was helpful to the transfer of electrochemical active materials and provided a large specific surface area. Figure 2d showed the Raman spectra of three materials. Three peaks belonged to carbon materials was found and they corresponded to D bond, G bond and 2D bond, respectively [42]. The D bond was attributed to the rotation of plane carbon atom, which showed the edge flaw of graphene. The higher intensity of D peak came from the greater disordering degree of the material. It was obvious that the intensity of D bond was enhanced after two-step stripping, indicating more and more defect structures and the successful formation of graphene.

Fig. 2
figure 2

SEM images of carbon paper (a), FEG/carbon paper prepared by the first step (b), and FEG/carbon paper prepared by two steps (c). Raman spectra of carbon paper (curve a) (d), FEG/carbon paper prepared by the first step (curve b), and FEG/carbon paper prepared by two steps (curve c)

Full size image

Figure 3 showed SEM images of Fe3O4 NPs/FFEG/carbon paper, Fe3O4 NPs/FEG/carbon paper, and Fe3O4 NPs/carbon paper. As shown in Fig. 3a, d, a large number of small Fe3O4 NPs was formed and uniformly dispersed on FFEG/carbon paper. XRD pattern of Fe3O4 NPs/FEG/carbon paper showed diffraction peaks at 30.1°, 35.4°, 43.1°, 56.9°, 62.5°, and 74.9° (Fig. S4, Supporting Information). The diffraction peaks were ascribed to the (220), (311), (400), (511), (440), and (533) crystalline facets of Fe3O4 (JCPDS No. 19-0629) [43], and the (002), (004) crystalline facets of FFEG/carbon paper. The result proved that Fe3O4 NPs were synthesized successfully on the FFEG/carbon paper. While, there also were lots of Fe3O4 NPs uniformly dispersed on the surface of FEG/carbon paper (Fig. 3b, e). As compared with that on FFEG/carbon paper, the amounts of Fe3O4 NPs on FEG/carbon paper were fewer and the size of Fe3O4 NPs was larger. The result confirmed that the –COOH groups of FFEG/carbon paper was benefit for the growth and distribution of Fe3O4 NPs. However, only a few large Fe3O4 NPs formed on carbon paper and they were prone to aggregated (Fig. 3c, f). The result indicated that many flaw sites were also formed on graphene when they were peeled off. The defect sites might be used as anchored sites to guide the growth and distribution of small Fe3O4 NPs. Accordingly, many relatively small Fe3O4 NPs were also formed on Fe3O4 NPs/FEG/carbon paper.

Fig. 3
figure 3

SEM images of (a, d) Fe3O4 NPs/FFEG/carbon paper, (b, e) Fe3O4 NPs/FEG/carbon paper, and (c, f) Fe3O4 NPs/carbon paper

Full size image

Electrochemical behaviors of Fe3O4 NPs/FFEG/carbon paper

CV and EIS were firstly used to explore electrochemical performance of Fe3O4 NPs/FFEG/carbon paper as supercapacitor electrode. As shown in Fig. 4a, a pair of obvious redox peaks was found at Fe3O4 NPs/FFEG carbon paper (curve a) and they might be ascribed to the redox of Fe3O4 NPs since no peak was observed on FFEG/carbon paper (curve c) and carbon paper (curve e). Although the pair of redox peaks was also observed on Fe3O4 NPs/FEG/carbon paper (curve b), it was smaller than that of Fe3O4 NPs/FFEG/carbon paper. It was very difficult to observe the pair redox peaks on the Fe3O4 NPs/carbon paper because only a few large Fe3O4 NPs on carbon paper (curve d). The results clearly indicated that all the materials have the large capacitance but the capacitance of Fe3O4 NPs/FFEG/carbon paper was biggest. Figure 4b showed Nyquist plots of various materials. All the materials showed small impedance because no semicircle was found on these curves but the ion diffusion rate of Fe3O4 NPs/FFEG/carbon paper was the biggest. The rapid ion diffusion rate was a benefit for the performance of supercapacitors.

Fig. 4
figure 4

a CVs of Fe3O4 NPs/FFEG/carbon paper (curve a), Fe3O4 NPs/FEG/carbon paper (curve b), FFEG/carbon paper (curve c), Fe3O4 NPs/carbon paper (curve d), and carbon paper (curve e) in 1 M Na2SO3 at the scanning rate of 10 mV s−1. b The Nyquist plots of Fe3O4 NPs/FFEG/carbon paper (curve a), Fe3O4 NPs/FEG/carbon paper (curve b), FEG/carbon paper (curve c), and carbon paper (curve d) in 5.0 mM Fe(CN)6 3−/4− at the scanning frequency range from 0.01 Hz to 10.0 KHz

Full size image

Figure 5 showed that the constant current charge/discharge curves of Fe3O4 NPs/FFEG carbon paper (Fig. 5a), Fe3O4 NPs/FEG/carbon paper (Fig. 5b), and Fe3O4 NPs/carbon paper (Fig. 5c) electrodes in the potential range of − 0.6–0.1 V at various specific current densities. All the materials gave the significant constraint capacitance platform. The specific capacitance of Fe3O4 NPs/FFEG carbon paper was 470, 394.64, 299.64, 264.29, and 246.43 F g−1 at the current density of 1, 1.25, 2.5, 5, and 6.25 A g−1, respectively (Fig. 5a). The capacitance of Fe3O4 NPs/FEG/carbon paper under the different current densities was 335, 301.79, 225, 142.85, 125 F g−1 (Fig. 5b) and the capacitance of Fe3O4 NPs/carbon paper in various current densities was 130.71, 119.29 , 85.35, 62.28, and 57.32 F g−1 (Fig. 5c). As shown in Fig. 5d and Fig. S5 (Supporting Information), the Fe3O4 NPs/FFEG/carbon paper showed the largest capacitance, which could be ascribed to the 3D porous structures and loading largest amount of small Fe3O4 NPs. Figure S6 (Supporting Information) showed that the reaction time of 10 h was optimal.

Fig. 5
figure 5

Constant current charge/discharge curves of a Fe3O4 NPs/FFEG/carbon paper, b Fe3O4 NPs/FEG/carbon paper, and c Fe3O4 NPs/carbon paper materials in 1 M Na2SO3 solution at a current density of 1.0, 1.25, 2.5, 5.0, and 6.25 A g−1. d The plots of specific capacitance of various materials versus different current densities

Full size image

Next, the cycle life of Fe3O4 NPs/FFEG/carbon paper was tested (Fig. 6). The constant current charge/discharge test was carried out under the current density of 1 A g−1 for 5000 cycles. The capacitance slightly decreased to 432 F g−1. The good stability might be ascribed to the firm immobilization of Fe3O4 NPs on –COOH groups of FFEG, the firm graphene foam on carbon paper as well as uniform distribution of small Fe3O4 NPs. Figure S7 displayed CVs of Fe3O4 NPs/FFEG /carbon paper at various scan rates. With the increasing scan rates, the anodic peaks shift toward positive potential and the cathodic peaks shift toward negative potential due to the electrode polarization at larger scan rates. The current linearly increased as the scan rate increased, indicating a good rate capability. To exhibit the advantage of proposed materials, the performance of different FeOx nanocomposites based supercapacitors were compared in Table 1. By comparing, the Fe3O4 NPs/FFEG carbon paper showed the largest specific capacitance because both 3D porous structure and small and uniformly distributed Fe3O4 NPs promoted the electron and mass transfer together. What’s more, the 3D graphene foam and –COOH groups of FFEG also provided a larger surface area to load a large number of small Fe3O4 NPs effectively, which improved the utilization efficiency of the active material largely and avoided the aggregation of Fe3O4 NPs in the charge/discharge process.

Fig. 6
figure 6

Cycle performance of Fe3O4 NPs/FFEG/carbon paper at a current density of 1.0 A g−1. The electrolyte was 1.0 M Na2SO3

Full size image
Table 1 Comparison of the performance of various based-Fe3O4 NPs supercapacitors
Full size table

Conclusion

In summary, a simple two-step electrochemical peeled method to form FFEG/carbon paper was developed to load Fe3O4 NPs as supercapacitor electrode. Some typical layered fold graphenes were successfully peeled off and stood on the carbon paper matrix to form 3D graphene foam. The Fe3O4 NPs/FFEG/carbon paper as a free-binder electrode for supercapacitors exhibited the following advantages. Firstly, both the porous structure and the good conductivity of 3D graphene foam were a benefit for electron and mass transfer. Secondly, the 3D graphene foam could effectively alleviate the agglomeration of Fe3O4 NPs and accordingly improve the performance of the supercapacitors. Thirdly, the 3D graphene foam and a large number of –COOH groups of FFEG provided a larger surface area to load a large number of small Fe3O4 NPs effectively, which improved the utilization efficiency of the active material largely. Fourthly, the firm immobilization of Fe3O4 NPs on –COOH groups of FFEG, the firm graphene foam on carbon paper as well as uniform distribution of small Fe3O4 NPs resulted in good stability. Finally, the 3D graphene foam served not only as the supporting of Fe3O4 NPs but also as the current collector of the supercapacitors, which would reduce the costs. As a result, the Fe3O4 NPs/FFEG/carbon paper exhibited a high-specific capacitance of 470 F g−1 at the current density of 1 A g−1, which was much higher than other carbon materials. Moreover, the good cycle performance indicated that the newly developed supercapacitors based on the Fe3O4 NPs/FFEG/carbon paper might be potentially useful for future energy storage applications.