Nitrogen and fluorine co-doped holey graphene hydrogel as a binder-free electrode material for flexible solid-state supercapacitors

Yulin Chen a, Yu Li *acd, Fengnan Yao a, Cong Peng a, Chen Cao a, Yiyu Feng acd and Wei Feng *abcd
aSchool of Materials Science and Engineering, Tianjin University, Tianjin 300072, P. R China. E-mail: weifeng@tju.edu.cn; Fax: +86-22-27404724; Tel: +86-22-28578269
bCollaborative Innovation Center of Chemical Science and Engineering (Tianjin), Tianjin 300072, P. R China
cKey Laboratory of Advanced Ceramics and Machining Technology, Ministry of Education, Tianjin 300072, P. R China
dTianjin Key Laboratory of Composite and Functional Materials, Tianjin 300072, P. R China

Received 6th March 2019 , Accepted 6th May 2019

First published on 15th May 2019


Two-dimensional (2D) graphene-based materials are widely applied as electrodes for supercapacitors due to their vast surface area and excellent charge mobility. However, the pursuit of the theoretical capacitance of graphene has lasted besides the improvements of charge/discharge rate and cycling life. In this work, a nitrogen and fluorine co-doped holey graphene hydrogel (NF-HGH) was synthesized by using consecutive solvothermal treatments for creating vacancies on the graphene basal plane and introducing heteroatoms into the graphene framework. The NF-HGH as an electrode for supercapacitors exhibited a remarkable electrochemical performance of 345.4 F g−1 at 1 A g−1 in 6 M KOH electrolyte. In addition, the specific capacitance of the assembled all-solid-state flexible symmetric supercapacitor (SSC) was 67.3 F g−1 at 1 A g−1, with an outstanding long-term cycling stability over 10[thin space (1/6-em)]000 cycles. The delivered maximum energy and power densities were 7.99 W h kg−1 and 10.08 kW kg−1, respectively. The abundant pyrrolic-N doped at the edge of etched vacancies and the formed semi-ionic C–F bonds around pyrrolic-N were critical for electrochemical developments, which provided a sensational method to construct carbonaceous frameworks for energy storage systems.


1. Introduction

Supercapacitors have captured extensive attention because of their unique advantages, such as high-power density, long cycling life, and great potential applications in various fields, like miniaturized and wearable electronic devices.1 Graphene is an ideal electrode material for supercapacitors and the theoretical electrochemical double-layer capacitance (EDLC) of graphene monolayers is as high as 550 F g−1.2,3 However, most of the graphene-based materials for supercapacitors usually deliver specific capacitances in the range of 100–200 F g−1. The huge gap between the theoretical value and the actual value is caused by the tightly stacked graphene nanosheets via the strong π–π interaction, which limits the accessible surface area and blocks the diffusion of ions.4,5 Three-dimensional (3D) graphene assemblies constructed by randomly stacked graphene nanosheets,6–8 such as aerogels, sponges, hydrogels and foams, exhibit massive porous structures, which facilitate ion transfer and enhance the electrolyte wettability significantly.9 However, the lack of micropores (≤2 nm) in these 3D graphene assemblies cannot fulfill the calculated capacitance of graphene completely. Therefore, extra activations are employed to produce abundant micropores in graphene nanosheets so as to enlarge the accessible surface area further. Among the various activation reagents, H2O2 is an environment-friendly and mild reagent that generates plentiful micropores with a narrow size distribution.10,11

On the other hand, heteroatom doping into graphene frameworks is another widely applied method to improve the electrochemical storage capability of graphene through the enhanced quantum capacitance and the additional pseudo-capacitance derived from the faradaic charge transfer of functional groups.12 Nowadays, co-doping of two or more kinds of heteroatoms into graphene frameworks simultaneously has drawn increasing interest because of the synergetic effects from different heteroatoms. For instance, a N and B co-doped graphene aerogel exhibited superior capacitive performances than pristine and N- or B-doped aerogels because the interaction between adjacent N and B atoms facilitated charge transfer with neighboring C atoms.13 Other studies have also demonstrated the superior electrochemical performances of graphene co-doped with N and S,14–16 N and P,17–19 S and P,20 and N and F21 to the counterparts doped with single heteroatoms. Among various heteroatoms, the F element, with the highest electronegativity, effectively enhances the electrochemical activity and stability of graphene in the energy storage field because of the charge redistribution and the high bond energy of C–F bonds.22 For example, Gao et al. prepared N and F co-doped carbon microspheres by a low-temperature solvothermal method, which showed an ultrahigh volumetric capacitance of 521 F cm−3 because F atoms withdrew the charges on N atoms and reduced the band gap significantly.23 Jang et al. also prepared N and F co-doped highly porous carbon nanofibers with outstanding supercapacitance ascribed to the dynamic porous structure and heteroatom co-doping.24 Our group has also synthesized various F-doped graphene electrodes,25–28 and F-doped graphene hydrogels serving as electrodes for supercapacitors exhibited a high-power density owing to the abundant semi-ionic C–F bonds that facilitated electron transfer throughout the electrode.

The configurations of doping species have a great impact on the electrochemical performances of graphene-based electrodes, which are closely dependent on the structure of graphene basal planes. For instance, it has been demonstrated that N-doping tends to take place at the edge of vacancies in defective graphene nanosheets14 and F-doping species are determined by the oxygen containing groups in the graphene basal plane.25 Based on the above perception, a N and F co-doped holey graphene hydrogel (NF-HGH) with abundant micropores was designed in this study. The activation of graphene oxide (GO) with H2O2 was first performed to generate a holey graphene precursor, and then it was subjected to a hydrothermal reaction with pyridine hydrofluoride (C5H5N·HF) as the doping source of both N and F simultaneously, as illustrated in Fig. 1. The great numbers of micropores in graphene nanosheets served as the active sites for N-doping, mainly in the form of pyrrolic-N, which provided extra pseudo-capacitance through the reversible redox reaction; F atoms were likely to bond with C atoms around pyrrolic-N and formed semi-ionic C–F bonds subsequently, which guaranteed the electrical conductivity of the NF-HGH.29 The electrochemical performances of the NF-HGH as an electrode material for supercapacitors in 6 M KOH electrolyte were tested, and it exhibited superb capacitive performances due to the synergetic effects of micropores and co-doped heteroatoms. Particularly, the corresponding fabricated symmetric flexible solid-state supercapacitor by using NF-HGH electrodes and polyvinyl alcohol (PVA)-KOH gel electrolyte delivered a maximum power density as high as 10.08 kW kg−1, associated with a maximum energy density of 7.99 W h kg−1, which provided a constructive strategy to engineer the architecture and components of graphene basal planes to improve electrochemical performances further.


image file: c9se00142e-f1.tif
Fig. 1 Schematic illustration of the preparation of NF-HGH.

2. Experimental

2.1 Activation of GO

GO was prepared from natural graphite powder according to a modified Hummer's method.30 The resulting GO was suspended in water to get a homogeneous GO colloidal solution with a concentration of 2 mg mL−1. Then 3 mL H2O2 (Aladdin, 30 wt%) was added into 30 mL of prepared GO colloidal solution and the mixture was stirred for 10 min. After that, the mixture was sealed in a 50 mL Teflon-lined autoclave and transferred to a drying chamber for heating at 100 °C for 4 h. After naturally cooling down to room temperature, the residual H2O2 was removed by centrifugation at 10[thin space (1/6-em)]000 rpm for 10 min and washed with deionized water several times to obtain a glutinous colloid, which was diluted with deionized water through vigorous ultrasonication for a few minutes to obtain a homogeneous colloidal solution with a concentration of 2 mg mL−1 as the precursor for NF-HGH. Similarly, another colloidal solution was also prepared through the same process without adding H2O2 for the activation.

2.2 Synthesis of the NF-HGH

The NF-HGH was obtained by a conventional hydrothermal reaction. Typically, 10 mL C5H5N·HF (Aladdin, 30 wt% pyridine, 70 wt% HF) was added into 30 mL of prepared homogeneous colloidal solution after H2O2 activation and the mixture was stirred for 10 min. After that, the mixture was sealed in a 50 mL Teflon-lined autoclave and transferred to a drying chamber for heating at 150 °C for 24 h, and then naturally cooled down to room temperature. The resultant hydrogel was immersed in deionized water for enough time to remove the residual C5H5N·HF, and then freeze dried so as to obtain the solid NF-HGH product. In order to investigate the effects of micropores, the NF-GH was also obtained by using the colloid solution without H2O2 activation through the same hydrothermal reaction. On the other hand, in order to investigate the effects of heteroatom co-doping, reduced graphene hydrogels (RGH) and holey graphene hydrogels (HGH) were also prepared through the hydrothermal reaction at 150 °C for 24 h by using the corresponding colloidal solutions directly without adding C5H5N·HF.

2.3 Characterization

The morphology of the prepared hydrogels was investigated by using field-emission scanning electron microscopy (FESEM) (Hitachi S4800) and field-emission transmission electron microscopy (FETEM) (JEOL JEM-2100F). Fourier transform infrared (FT-IR) spectra were recorded on a Bruker Tensor 27 spectrometer. Raman scattering was recorded on a Raman spectrometer (DXR Microscope, Thermo Electron) using a laser excitation of 532 nm. The crystal structure was determined by X-ray diffraction (XRD) at a scan rate of 1.4° min−1 (Rigaku D/max-2500 with Cu Kα radiation). X-ray photoelectron spectroscopy (XPS) analysis was performed by using a PerkinElmer PHI 3056 with an Al anode source operated at 15 kV to analyze the chemical composition of the materials. The N2 adsorption–desorption isotherms of the products were investigated with a pore size analyzer (Autosorb-iQ2-MP, America) at 150 °C and the specific surface area (SSA) was analyzed by the Brunauer–Emmett–Teller (BET) method.

2.4 Electrochemical measurements

The working electrodes were fabricated according to the reported method,31–33 in which the wet hydrogels were cut into a small piece with a thickness of about 2 mm and pressed onto nickel foil (1 cm × 1 cm) at 10 MPa for 30 s before moving to a vacuum oven for heating at 80 °C for 6 h for use as an electrode. The dry weight of the active materials for each electrode was approximately 2–3 mg. The prepared electrode films were soaked in 6 M KOH solution for 6 h before testing. The electrochemical performances of the prepared hydrogels were tested in the three-electrode cell configuration on a CHI 660D (CH Instruments, Inc.) electrochemical workstation at room temperature. The prepared hydrogel electrodes were used as the working electrodes, with a saturated standard Hg/HgO reference electrode and a platinum counter electrode, and 6 M KOH aqueous solution was used as the electrolyte. Cyclic voltammograms (CV) were recorded in the potential window of −1.0 to 0 V at various scan rates of 5, 10, 20, 50, 100, 200, 500 and 1000 mV s−1. Galvanostatic charge–discharge profiles were recorded at various current densities of 1, 2, 5, 10, 20, 50 and 100 A g−1 in the potential range from −1.0 to 0 V as well. The gravimetric specific capacitance (C, F g−1) of the active material was calculated from the galvanostatic discharge profile by using eqn (1):
 
C = (IΔt)/(mΔV)(1)
where I (A) is the constant current, m (g) is the mass of the dried electrode materials, Δt (s) is the discharge time, and ΔV (V) is the range of discharge potentials. Electrochemical impedance spectroscopy (EIS) measurements were performed in the frequency range from 0.01 to 106 Hz at a 5 mV amplitude referring to open circuit potential.

All-solid-state flexible symmetric supercapacitors (SSCs) were fabricated using the prepared hydrogel electrodes and PVA-KOH gel electrolyte. In a typical preparation, 3 g PVA and 3 g KOH were dissolved in de-ionized water under stirring at 85 °C until the solution became clear to obtain the PVA-KOH gel electrolyte, which was cooled to room temperature and used for fabricating the SSC. Two NF-HGH slices with almost the same dry weight of ∼2 mg were soaked in 6 M KOH solution for 6 h and then pressed onto nickel foil at 10 MPa for 30 s to obtain the electrodes directly with the area of the active material being about 1 cm2 (1 cm × 1 cm). Then the PVA-KOH gel electrolyte was slowly poured onto the top electrodes, which were air-dried for 12 h at room temperature to evaporate the excess water. Finally, the two electrodes were pressed and held together for 24 h while the PVA-KOH gel electrolyte solidified completely, and hence a robust all-solid-state flexible SSC was fabricated. Galvanostatic charge–discharge profiles were recorded at various current densities of 0.5, 1, 2, 5, 10 and 20 A g−1 in the potential range from 0 to 1.0 V. The gravimetric specific capacitance of the supercapacitor was calculated from the galvanostatic discharge profile according to eqn (2):

 
C = (IΔt)/(mΔV)(2)
where I (A) is the constant current, m (g) is the total mass of the active material on two electrodes, Δt (s) is the discharge time, and ΔV (V) is the range of discharge potentials. The energy density (E, W h kg−1) and power density (P, kW kg−1) of all-solid-state flexible SSCs were calculated using eqn (3) and (4):
 
E = (CΔV2)/2(3)
 
P = E/(Δt)(4)
where C (F g−1) is the gravimetric specific capacitance of the active material, ΔV (V) is the range of discharge potentials, and Δt (s) is the discharge time.

2.5 Computational analysis

The Density Functional Theory (DFT) calculation was performed using a periodically repeated slab with the DMol3 Package to investigate the prepared NF-HGH.34 A structure mode of the monovacancy was proposed, where defective NF-doped graphene was composed of 33 atoms in a 9.76 × 9.76 × 20 Å hexagonal supercell. Full geometry optimizations were performed for all models and all atoms were free to relax. All calculations were performed using the DFT semicore pseudopots approximation to replace the core electrons with a single effective potential. The double numerical plus polarization (DNP) basis sets were used for the valence orbitals and provided accuracy.35 The exchange–correlation contributions were described within the Perdew–Burke–Ernzerhof (PBE) exchange–correlation functional. The simulated binding energy (Eb) between the F atom and N-doped defective graphene was calculated using the following equation:
 
Eb = Etotal − (EN−G + EF)(5)
where Etotal, EN–G and EF represent the total energy of the bonded system, the energy of N-doped graphene and the energy of the F atom in a vacuum, respectively.

3. Results and discussion

The SEM image of the NF-HGH (Fig. 2a) shows an interconnected 3D framework with pore size ranging from hundreds of nanometers to several micrometers. A similar morphology is also observed in the freeze-dried GO and the other prepared hydrogels (Fig. S1, ESI), though there are less stacked graphene nanosheets in N and F co-doped hydrogels due to the weakened π–π interactions by the doping heteroatoms, which is also represented by the different volumes of the prepared hydrogels (Fig. S2, ESI) because the doped hydrogels tend to encapsulate more water in the self-assembly process during the hydrothermal reaction.25 The TEM image of the NF-HGH (Fig. 2b) shows many wrinkles on the graphene nanosheets, unlike the flat and smooth graphene nanosheet of GO (Fig. S3a, ESI). Similarly, these wrinkles can also be observed in the TEM images of the RGH, HGH and NF-GH (Fig. S3b–d, ESI), though there are more winkles in the doped ones because doped heteroatoms generate more defects in the graphene basal planes.26 In addition, some holey structures are found in the TEM image of the NF-HGH (Fig. 2b), suggesting efficient oxidation and etching of graphene nanosheets by H2O2 activation. The corresponding high-resolution TEM (HRTEM) image of the NF-HGH (Fig. 2c) indicates the well-defined edges of the NF-HGH, constructed by a few layers of graphene nanosheets. The ambiguous diffracted ring in the selected area electron diffraction (SAED) pattern of the NF-HGH (inset of Fig. 2c) proves the amorphous structure of the NF-HGH. In addition, the presented homogeneous distributions of N and F signals throughout the entirely framed area from the elemental mapping of NF-HGH (Fig. 2d) and NF-GH (Fig. S4, ESI) TEM images indicate effective N and F doping into graphene frameworks.
image file: c9se00142e-f2.tif
Fig. 2 (a) SEM, (b) TEM and (c) HRTEM images of the NH-HGH (inset shows the corresponding SAED). (d) The representative TEM image of the NF-HGH and the corresponding elemental mapping images of C, O, N and F components in the framed area.

The FT-IR spectra were used to characterize the chemical structure of GO and the prepared hydrogels (Fig. 3a). The carbonyl stretching mode of –COOH at 1728 cm−1, the breathing vibration mode of –CH(O)CH– at 1228 cm−1 and the stretching vibration of C[double bond, length as m-dash]C at 1625 cm−1 are observed in the FT-IR spectrum of GO.36 The broad band appearing at 3400 cm−1 is attributed to the extensive oxidation of graphene.26 In the FT-IR spectra of the prepared hydrogels, the obviously decreased peak intensities at 1728 and 1228 cm−1 demonstrate the large removal of oxygen-containing functional groups during the solvothermal treatments, but the slightly enhanced intensities of these peaks in the FT-IR spectra of HGH and NF-HGH are ascribed to the H2O2 activation. In the FT-IR spectra of the NF-GH and NF-HGH, the peaks appearing at 1452 and 1256 cm−1 correspond to the N–H bending vibration and C–N stretching vibrations, respectively.37 In the FT-IR spectra of the NF-GH and NF-HGH, the distinct peak located at 1080 cm−1 is ascribed to the stretching vibration of the semi-ionic C–F bonds,38 but the stretching vibration peaks of covalent C–F bonds around 1210 cm−1 are overwhelmed by the breathing vibration mode of the –CH(O)CH– bond.22


image file: c9se00142e-f3.tif
Fig. 3 (a) FT-IR spectra, (b) Raman spectra and (c) XRD patterns of the GO, RGH, HGH, NF-GH and NF-HGH.

Raman spectra were further employed to investigate the structural regularity of GO and the prepared hydrogels (Fig. 3b), and the typical D and G bands at approximately 1345 and 1590 cm−1 are all observed.39 The G band corresponds to the recovery of the hexagonal network of carbon atoms with defects,40 and the intensity ratios of D and G bands (ID/IG) of the GO, RGH, HGH, NF-GH, and NF-HGH are 0.84, 1.02, 1.06, 1.10, and 1.12, respectively. The reduction of GO by removing oxygen-containing groups during the solvothermal treatments increases the disorder degree.26 Moreover, the NF-HGH exhibits the highest ID/IG ratio because co-doped N and F atoms and the generated porous structure by H2O2 activation both introduce defects in the graphene basal plane.

XRD characterization was applied to investigate the crystal structures of GO and the prepared hydrogels (Fig. 3c). The XRD pattern of GO exhibits a sharp diffraction peak at 10.2°, corresponding to the (002) plane with an interlayer spacing of 0.87 nm. The broad diffraction peaks in the XRD patterns were in agreement with SEAD patterns. The upshift of diffraction peaks in the range of 23–25° confirms the decreasing interlayer distance due to the removal of the oxygen containing functional groups during the solvothermal treatments.41 However, it should be noted that H2O2 activation creates some oxygen-containing groups and F-doping generates C–F bonds vertical to the graphene basal plane, which increase the interlayer spacing simultaneously,10,25 and hence the interlayer spacing of the NF-HGH is 0.37 nm, larger than those of the left hydrogels.

The porous structures of the prepared hydrogels were tested by N2 adsorption/desorption analysis. All the prepared hydrogels exhibit the IV-type isotherm with a pronounced hysteresis loop (Fig. 4a), indicating the presence of a large number of mesopores.5,42 Furthermore, the hysteresis loops in the isotherms of the HGH and NF-HGH at the lower relative pressure also imply the presence of slit-shaped micropores.11 The BET SSA of the RGH, HGH, NF-GH and NF-HGH is 420, 947, 647 and 1465 m2 g−1, respectively. The corresponding pore size distributions calculated from the density functional theory (DFT) model (Fig. 4b) demonstrate the major mesopores in the RGH and NF-GH with a double-peak pore size at 3.5 and 10.5 nm. However, the HGH and NF-HGH not only show abundant mesopores with a predominant peak pore size of 5 nm but also exhibit some micropores less than 2 nm, which is in agreement with their isotherms. The H2O2 activation not only generates massive micropores on graphene basal planes through the oxidization process, but also affects the stacking, entanglement and overlapping of graphene nanosheets, leading to the change of mesopore size distribution.43


image file: c9se00142e-f4.tif
Fig. 4 (a) N2 adsorption–desorption isotherms and (b) DFT pore-size distribution of the freeze-dried RGH, HGH, NF-GH and NF-HGH.

XPS analysis was further carried out to determine the doping levels and doping species of prepared hydrogels. From the XPS survey spectra of GO and the prepared hydrogels (Fig. 5a), the appeared N1s and F1s signals in the NF-GH and NF-HGH verify the successful N and F co-doping into graphene basal planes, which is in agreement with their FT-IR spectra. The contents of N and F in the NF-GH are 1.87 and 0.52 at%, which increase to 2.56 and 1.14 at% in the NF-HGH, respectively (Table S1, ESI). Therefore, it can be deduced that the H2O2 activation is beneficial to improve the doping level because the plentiful micropores in graphene basal planes as the defective sites facilitate N doping at the edge of vacancies14 and the generated oxygen-containing groups promote the substitution reaction with F at the same time (Fig. S5, ESI).44


image file: c9se00142e-f5.tif
Fig. 5 (a) XPS survey spectra of the GO, RGH, HGH, NF-GH and NF-HGH and high-resolution (b) C1s, (c) N1s and (d) F1s spectra of the NF-HGH.

The fitted high-resolution C1s spectrum of the NF-HGH is shown in Fig. 5b. The intensive peak at 284.7 eV is attributed to the non-functionalized sp2-hybridized C atoms, suggesting the conjugated honeycomb lattice in the NF-HGH. The peak at 285.8 eV is assigned to the C atoms connected to the residual O atoms with sp3 hybridization, and the intensity of this peak is significantly lower than that in GO (Fig. S6a, ESI), indicating that most of the oxygen-containing functional groups are removed during the solvothermal treatments. A similar phenomenon is also observed in the C1s spectra of the RGH, HGH and NF-GH (Fig. S6b–d, ESI). The peaks at 287.1 eV in the C1s spectra of all prepared hydrogels are ascribed to the coincident C[double bond, length as m-dash]O or C[double bond, length as m-dash]N bonds. Nevertheless, in the C1s spectra of the NF-HGH and NF-GH, the peaks at 288.6 eV are attributed to the semi-ionic C–F bonds, and the consecutive peaks at 289.5, 291.1 and 292.8 eV correspond to the different configurations of covalent C–F bonds.27 The high-resolution N1s spectrum of the NF-HGH (Fig. 5c) shows the typical three peaks assigned to pyridinic-N (398.0 eV), pyrrolic-N (400.0 eV) and graphitic-N (401.5 eV) with the percentages of 14.3, 68.0 and 17.7 at%, respectively. Compared with the N1s spectrum of the NF-GH (Fig. S7a, ESI), the notably higher pyrrolic-N doping level in the NF-HGH confirms that the vacancies in graphene basal planes etched by H2O2 promote N-doping in the form of pyrrolic-N.14 The high-resolution F1s spectrum of the NF-HGH (Fig. 5d) represents the two peaks ascribed to semi-ionic and covalent C–F bonds at 686.0 and 688.0 eV, respectively, which is consistent with the results of its C1s spectrum. These features are also observed in the F1s spectrum of the NF-HG (Fig. S7b, ESI). In addition, semi-ionic C–F bonds account for more than 80 at% of the total formed fluorinated species in both doped hydrogels which guarantee the superb electrical conductivity of the prepared hydrogels.45

The hierarchical porous structure and N and F co-doping make the NF-HGH a promising electrode material for supercapacitors. Therefore, the capacitive performances of the prepared hydrogels were evaluated using a three-electrode system in aqueous 6 M KOH solution at first. The quasi-rectangular shape of their CV curves at a scan rate of 10 mV s−1 in the potential range from −1.0 to 0 V (Fig. 6a) indicates the energy storage mechanism mainly based on EDLC, and this shape of CV curves is maintained even at high scan rates (Fig. S8, ESI). Moreover, the highest current density in the CV curve of the NF-HGH (Fig. 6a) demonstrates its best energy storage capability. Note that two pairs of redox peaks are identified in the CV curves of the NF-GH and NF-HGH due to the extra pseudo-capacitance derived from co-doping heteroatoms.12 The first pair of redox peaks at −0.27 V in the cathodic process and −0.25 V in the anodic process is ascribed to the N-doping species,14,46 and the second pair of redox peaks at −0.70 V in the cathodic process and −0.65 V in the anodic process is generated by the F-doping species because of its higher electronegativity.25 The slightly asymmetric triangle shapes of their galvanostatic charge–discharge profiles at 1 A g −1 (Fig. 6b) indicate the existence of pseudo-capacitance besides the major EDLC further. The longer discharge time than the charge time in NF-GH and NF-HGH electrodes is attributed to the pseudocapacitive interactions between positive K+ ions and doping heteroatoms in the graphene framework, which are indicated by the sloping shape of the cathodic branch in the CV curves (Fig. 6a),47 and this behavior has been found in many heteroatom doped graphene hydrogel electrodes.48–50 The specific capacitances of the prepared hydrogels at different current densities are shown in Fig. 6c, calculated from the corresponding charge–discharge profiles (Fig. S9, ESI). The NF-HGH shows the highest specific capacitance of 345.4 F g−1 at a current density of 1 A g−1. This is mainly ascribed to the high specific area and the pseudo-capacitance produced by heteroatom doping that significantly increase the specific capacitance of doped graphene. As a result, the specific charge–discharge profiles of the NF-HGH after 10[thin space (1/6-em)]000 cycles at a current density of 2 A g−1 (inset of Fig. 6c) demonstrate that the capacitance of the NF-GH is a little higher than that of the HGH despite the latter showing a higher pore volume. In Fig. 6c, the NF-HGH exhibits the best rate capability among the prepared hydrogels and its specific capacitance is still maintained as high as 197.9 F g−1 at a current density even up to 100 A g−1. The capacitance retention of 96% obtained from the galvanostatic charge–discharge profiles of the NF-HGH after 10[thin space (1/6-em)]000 cycles at a current density of 2 A g−1 (inset of Fig. 6c) demonstrates the excellent cycling stability of the NF-HGH as well. Furthermore, the more perpendicular line in the Nyquist plot of the NF-HGH than those of the other three hydrogels (Fig. 6d) indicates the most rapid ion diffusion coefficient with the ideal capacitive behavior,14,23 and the decreased values of bulk resistances (Rb) as well as charge-transfer resistance (Rct) of the prepared hydrogels after H2O2 activation and heteroatom co-doping reflect the effectiveness of these methods in the optimization of electrochemical performances. However, because of the wide pore size distributions of the prepared hydrogel electrodes, the different penetrabilities associated with pore size variations lead to a small inclined Nyquist curve in the real part (inset of Fig. 6d).51 Considering the highest specific capacitance of the NF-HGH, its electrochemical performances in organic electrolyte (1 M LiPF6 dissolved in EC/DMC) were also tested. However, the specific capacitance of the NF-HGH was 107.8 F g−1 at a current density of 0.05 A g−1, associated with the inferior cycling stability (Fig. S10, ESI) due to poor wettability with organic electrolyte and the inappropriate pore sizes for anions with the large radius of lithium salts. The optimization of the porous structure of graphene-based electrodes for organic electrolytes has been carried out.


image file: c9se00142e-f6.tif
Fig. 6 (a) Electrochemical performances of the RGH, HGH, NF-GH and NF-HGH in the potential window from −1.0 to 0 V in 6 M KOH solution: (a) CV curves at a scan rate of 10 mV s−1; (b) the galvanostatic charge–discharge profiles at a current density of 1 A g−1; (c) specific capacitances at different current densities (inset shows the charged-discharge profiles of the NF-HGH in the first and the 10[thin space (1/6-em)]000th cycles at a current density of 2 A g−1); (d) Nyquist plots from 0.01 to 105 Hz (inset shows the close-up view of the high-to-medium frequency regime).

To provide further information about the reversible faradaic redox reaction of the NF-HGH, first-principles DFT calculations were employed. Based on the XPS results, pyrrolic-N was doped independently in the vicinity of defects within a single carbon layer at first and the optimized structure (Fig. S11a, ESI) shows a strong N–C bond with a distance of 1.40 Å. Subsequently, F can be introduced into pyrrolic-N doped graphene at various possible sites that were labeled as α, β and γ in Fig. S11b, ESI, respectively. The calculated free energies for F doped in the β and γ sites with respect to the α site are 582 and 261 meV, respectively. Hence, F prefers to bond with the carbon atom beside the pyrrolic-N atom and the optimized structure of N,F-codoped graphene (Fig. S11b, ESI) shows a C–F bond with a distance of 1.52 Å that is considerably comparable with the semi-ionic C–F bond as revealed from the F1s XPS spectrum of the NF-HGH. In addition, the C atom bonded with the F atom is slightly buckled from the graphene basal plane due to the semi-ionic nature of the formed bond.52 Furthermore, deformation charge densities corresponding to pyrrolic-N doped graphene (Fig. 7a) and N,F-codoped graphene (Fig. 7b) were calculated to investigate the donation and acceptance of the electrons around the doped heteroatoms, in which the increase and decrease of total electron density with the density of isolated atoms subtracted are denoted as blue and red, respectively. It can be demonstrated that F-doping leads to the charge redistribution of the adjacent pyrrolic-N atom significantly. Because of the strongest electronegativity of the F atom, the electrons slightly move from the C–N bond to the C–F bond, associated with the prolonged C–N bond length to 1.51 Å. As a result, the electron-donor of pyrrolic-N with an extra pair of electrons converts to an electron-acceptor type represented by the drastically reduced electron density around this atom. The partial density of states (PDOS) of 2p orbitals of pyrrolic-N doped graphene and N,F-codoped graphene models (Fig. 7c) demonstrates that the unoccupied band of the latter is larger than that of the former one. The number of states in the unoccupied band may be correlated with the capacitance of graphene-based materials,53 and the enhanced electronic acceptance of N,F-codoped graphene is attributed to the mixed valence states of N atoms after F-doping. Furthermore, the confirmed semi-ionic C–F bonds ensure the electrical conductivity of graphene basal planes, increase the wettability and accessibility of electrolyte and enlarge the interlayer distance of stacked graphene nanosheets, not to mention the hierarchical porous structure that provides a huge accessible surface and a continuous path for ion contact and diffusion, respectively. Therefore, the NF-HGH exhibits enhanced specific capacitance and superior rate capability due to the synergetic effects of the engineered framework structure by abundant micropores and configurations of doped heteroatoms.


image file: c9se00142e-f7.tif
Fig. 7 The deformation charge densities of (a) pyrrolic-N doped graphene and (b) N,F-codoped graphene; (c) the partial density of states (PDOS) of 2p orbitals of pyrrolic-N doped graphene and N,F-codoped graphene models based on DFT. The grey, blue and green balls represent carbon, nitrogen and fluorine atoms, respectively.

Moreover, because of the extra pseudo-capacitance from the higher heteroatom doping level, especially pyrrolic-N, and the enhanced EDLC from the enlarged SSA, the NF-HGH exhibits high specific capacity. The superior rate capability of the NF-HGH is attributed to the abundant semi-ionic C–F bonds and highly porous structure, which ensure electron transfer throughout the electrode and facilitate ion diffusion from the electrolyte to the active material, respectively,23,54 because semi-ionic C–F bonds improve the wettability and accessibility to the electrolyte, and the hierarchical porous structure provides a continuous and fluent path for ion transfer.10,25 As described above, the improved electrochemical performances of the NF-HGH are ascribed to the co-doped N and F species as well as the hierarchical porous structure. At first, the micropores in the graphene nanosheets of the NF-HGH not only enlarge the accessible surface area for ion contact, but also promote the formation of pyrrolic N around these defective sites, which provide mainly extra pseudo-capacitance by improving the mobility of negative charges on the carbon surface.46,55 In addition, the abundant semi-ionic C–F bonds enhance the rate capability of the NF-HGH by preserving the conjugated structure of graphene as well as increasing the wettability to aqueous electrolyte for electron and ion transfer.23,56 Moreover, the semi-ionic C–F bonds lying between graphene nanosheets tend to repel each other and increase the interlayer distance of the NF-HGH, as revealed by the XRD patterns, which lower the barrier for ion diffusion to the active sites. Last but not least, the hierarchical porous structure of the NF-HGH containing abundant micro- and mesopores provides a continuous path for ion diffusion to the active material.

Because of the best electrochemical performances of the NF-HGH, it was selected as the typical one to fabricate an all-solid-state flexible symmetric supercapacitor (SSC) with PVA-KOH electrolyte to investigate its practical application. The triangular shapes of the charge–discharge profiles of the fabricated SSC at different current densities from 0.5 to 20 A g−1 in the potential window of 0–1.0 V (Fig. 8a) imply the retained excellent capacitive behavior, but the slight IR drops in these charge–discharge profiles are caused by the increased internal resistance, compared to the three-electrode system. The estimated specific capacitances of the NF-HGH at different current densities are depicted in Fig. 8b, with a capacitance retention of 76.08% as the current densities increase from 0.5 to 20 A g−1. The Ragone plot of the fabricated SSC (Fig. 8c) delivers a maximum energy density of 7.99 W h kg−1 with a power density of 250 W kg−1 at a current density of 0.5 A g−1. As the current density increases to 20 A g−1, the energy density decreases to 5.89 W h kg−1 due to the drop of specific capacitance, associated with a power density of 10.08 kW kg−1. Compared with the electrochemical performances of other all-solid-state flexible SSCs fabricated using other carbonaceous electrodes,13,14,57–60 the superior rate capability of the NF-HGH SSC is ascribed to the enlarged SSA by formed micropores and co-doping heteroatoms doped into graphene basal planes that improve the physical and chemical properties of SSC devices significantly. In addition, this symmetric supercapacitor shows an excellent cycling stability with a capacitance retention of 87.8% after 10[thin space (1/6-em)]000 cycles at a current density of 2 A g−1 (Fig. 8d). Three fabricated SSCs were connected in series and charged at 3 V for 20 s, and then a red light-emitting diode can be lit for several seconds without apparent change of luminescence (inset of Fig. 8d and Movie S1, ESI), which demonstrates the practicability of the NF-HGH as an electrode material for ultrafast supercapacitors.


image file: c9se00142e-f8.tif
Fig. 8 The electrochemical performances of the symmetric NF-HGH supercapacitor in the potential window from 0 to 1.0 V in PVA-KOH electrolyte: (a) charge–discharge profiles at different current densities; (b) specific capacitances at different current densities; (c) Ragone plots with comparison with other symmetric flexible solid-state supercapacitors; (d) the cycle stability after 10[thin space (1/6-em)]000 cycles at a current density of 2 A g−1 (inset shows the LED illuminated by the devices connected in series).

4. Conclusions

In summary, a NF-HGH as a supercapacitor electrode with high specific capacitance, excellent rate capability and long-term cycling stability has been synthesized through consecutive solvothermal treatments with H2O2 and C5H5N·HF. The H2O2 activation generated abundant micropores in graphene basal planes which not only increased the accessible surface area for ions but also induced N-doping at the edge of vacancies in the form of pyrrolic-N. Consequently, F-doping tended to occur adjacent to pyrrolic-N and form semi-ionic C–F bonds, which improved the electrical conductivity of the NF-HGH and the wettability of the electrolyte. Furthermore, the heteroatom co-doped configuration facilitates electron transfer and enhanced faradaic pseudocapacitance. As a result, the NF-HGH exhibited a high specific capacitance of 345.4 F g−1 at 1 A g−1. The capacitance of the all-solid-state flexible SSC fabricated using the NF-HGH achieved 67.7 F g−1 at 1 A g−1 with outstanding long-term cycling stability over 10[thin space (1/6-em)]000 cycles. Remarkably, the SSC delivered a superior power density of 10.05 kW kg−1 associated with a maximum energy density of 7.99 W h kg−1. Therefore, this study provides an effective technique to improve graphene-based electrode performances for energy storage devices through configuration optimization.

Conflicts of interest

There are no conflicts of interest to declare.

Acknowledgements

This work was financially supported by the National Key R&D Program of China (No. 2016YFA0202302), National Natural Science Funds for Distinguished Young Scholars (No. 51425306), the State Key Program of National Natural Science Foundation of China (No. 51633007), and the National Natural Science Foundation of China (No. 51573125 and 51773147).

References

  1. Y. Shao, M. F. El-Kady, L. J. Wang, Q. Zhang and R. B. Kaner, Chem. Soc. Rev., 2015, 44, 3639 RSC.
  2. Y. W. Zhu, M. Shanthi, W. W. Cai, X. S. Li, S. J. Won, J. R. Potts and R. S. Ruoff, ChemInform, 2010, 22, 3906 CAS.
  3. Y. Sun, Q. Wu and G. Shi, Energy Environ. Sci., 2011, 4, 1113 RSC.
  4. M. D. Stoller, S. Park, Y. Zhu, J. An and R. S. Ruoff, Nano Lett., 2008, 8, 3498 CrossRef CAS PubMed.
  5. Y. X. Xu, Z. Y. Lin, X. Q. Huang, Y. Wang, Y. Huang and X. F. Duan, Adv. Mater., 2013, 25, 5779 CrossRef CAS PubMed.
  6. V. Chabot, D. Higgins, A. Yu, X. Xiao, Z. Chen and J. Zhang, Energy Environ. Sci., 2014, 7, 1564 RSC.
  7. N. Stefania, C. Daniel, F. M. Luisa, M. C. Gutiérrez and D. M. Francisco, Chem. Soc. Rev., 2013, 42, 794 RSC.
  8. H. P. Cong, X. C. Ren, P. Wang and S. H. Yu, ACS Nano, 2012, 6, 2693 CrossRef CAS PubMed.
  9. P. Hao, Z. Zhao, J. Tian, H. Li, Y. Sang, G. Yu, H. Cai, H. Liu, C. P. Wong and A. Umar, Nanoscale, 2014, 6, 12120 RSC.
  10. Y. Xu, Z. Lin, X. Zhong, X. Huang, N. O. Weiss, Y. Huang and X. Duan, Nat. Commun., 2014, 5, 4554 CrossRef CAS PubMed.
  11. Y. Bai, X. Yang, Y. He, J. Zhang, L. Kang, X. Hua, S. Feng, Z. Lei and Z. H. Liu, Electrochim. Acta, 2015, 187, 543 CrossRef.
  12. X. W. Wang, G. Z. Sun, R. Parimal, K. Dong-Hwan, W. Huang and P. Chen, Chem. Soc. Rev., 2014, 43, 7067 RSC.
  13. Z. S. Wu, W. Andreas, L. Chen, Y. Sun, T. Andrey, X. L. Feng and M. Klaus, Adv. Mater., 2012, 24, 5130 CrossRef CAS PubMed.
  14. N. Q. Tran, B. K. Kang, M. H. Woo and D. H. Yoon, ChemSusChem, 2016, 9, 2261 CrossRef CAS PubMed.
  15. J. Li, G. Zhang, C. Fu, L. Deng, S. Rong and C. P. Wong, J. Power Sources, 2017, 345, 146 CrossRef CAS.
  16. L. Cheng, Y. Hu, D. Qiao, Y. Zhu, H. Wang and Z. Jiao, Electrochim. Acta, 2018, 259, 587 CrossRef CAS.
  17. X. Sun, C. Ping, H. Wang, X. Hua, L. Dang, Z. Liu and Z. Lei, Carbon, 2015, 92, 1 CrossRef CAS.
  18. X. Wang, Y. Liu and P. Wu, Chem. Eng. J., 2017, 328, 417 CrossRef CAS.
  19. Y. Wen, T. E. Rufford, D. Hulicova-Jurcakova and L. Wang, ChemSusChem, 2016, 9, 513 CrossRef CAS PubMed.
  20. Y. Xu, Y. Kang and H. S. Park, Carbon, 2016, 101, 49 CrossRef.
  21. X. Qiao, S. Liao, G. Wang, R. Zheng, H. Song and X. Li, Carbon, 2016, 99, 272 CrossRef CAS.
  22. C. B. Sun, Y. Y. Feng, Y. Li, C. Q. Qing, Q. Q. Zhang and W. Feng, Nanoscale, 2014, 6, 2634 RSC.
  23. J. Zhou, J. Lian, L. Hou, J. Zhang, H. Gou, M. Xia, Y. Zhao, T. A. Strobel, L. Tao and F. Gao, Nat. Commun., 2015, 6, 8503 CrossRef CAS PubMed.
  24. W. Na, J. Jun, W. P. Jin, G. Lee and J. Jang, J. Mater. Chem. A, 2017, 5, 17379 RSC.
  25. H. An, L. Yu, L. Peng, G. Yi, C. Qin, C. Chen, Y. Feng and F. Wei, J. Power Sources, 2016, 312, 146 CrossRef CAS.
  26. S. Z. Huang, Y. Li, Y. Y. Feng, H. R. An, P. Long, C. Q. Qin and W. Feng, J. Mater. Chem. A, 2015, 3, 23095 RSC.
  27. H. R. An, Y. Li, Y. Gao, C. Cao, J. K. Han, Y. Y. Feng and W. Feng, Carbon, 2017, 116, 338 CrossRef CAS.
  28. H. R. An, Y. Li, Y. Y. Feng, Y. Cao, C. Cao, P. Long, S. Li and W. Feng, Chem. Commun., 2018, 54, 2727 RSC.
  29. S. G. Peera, A. K. Sahu, A. Arunchander, S. D. Bhat, J. Karthikeyan and P. Murugan, Carbon, 2015, 93, 130 CrossRef CAS.
  30. J. P. Zhao, S. F. Pei, W. C. Ren, L. B. Gao and H. M. Cheng, ACS Nano, 2010, 4, 5245 CrossRef CAS PubMed.
  31. H. Banda, D. Aradilla, A. Benayad, Y. Chenavier, B. Daffos, L. Dubois and F. Duclairoir, J. Power Sources, 2017, 360, 538–547 CrossRef CAS.
  32. R. Wang, C. Xu and J. M. Lee, Nano Energy, 2016, 19, 210–221 CrossRef CAS.
  33. J. Li, G. Zhang, C. Fu, L. Deng, R. Sun and C. P. Wong, J. Power Sources, 2017, 345, 146–155 CrossRef CAS.
  34. B. Delley, J. Chem. Phys., 2000, 113, 7756 CrossRef CAS.
  35. S. S. Yu, W. T. Zheng, Q. B. Wen and Q. Jiang, Carbon, 2008, 46, 537 CrossRef CAS.
  36. D. Li, Y. Li, Y. Feng, W. Hu and W. Feng, J. Mater. Chem. A, 2015, 3, 2135 RSC.
  37. Z. Hang, T. Kuila, N. H. Kim, S. Y. Dong and J. H. Lee, Carbon, 2014, 69, 66 CrossRef.
  38. Z. J. Li, C. G. D. Del, W. F. Yan, C. D. Liang and S. Dai, J. Am. Chem. Soc., 2004, 126, 12782 CrossRef CAS PubMed.
  39. Z. S. Wu, W. C. Ren, L. B. Gao, J. P. Zhao, Z. P. Chen, B. L. Liu, D. M. Tang, B. Yu, C. B. Jiang and H. M. Cheng, ACS Nano, 2009, 3, 411 CrossRef CAS PubMed.
  40. I. K. Moon, J. Lee, R. S. Ruoff and H. Lee, Nat. Commun., 2010, 1, 73 CrossRef PubMed.
  41. A. K. Krishnamoorthy, B. M. Veerapandian, B. K. Yun and A. S. J. Kim, Carbon, 2013, 53, 38 CrossRef.
  42. S. Chowdhury and R. Balasubramanian, Sci. Rep., 2016, 6, 21537 CrossRef CAS.
  43. B. Xu, S. Yue, Z. Sui, X. Zhang, S. Hou, G. Cao and Y. Yang, Energy Environ. Sci., 2011, 4, 2826 RSC.
  44. T. Chen, X. Wang, Y. Liu, B. Li, Z. Cheng, Z. Wang, W. Lai and X. Liu, Phys. Chem. Chem. Phys., 2017, 19, 5504 RSC.
  45. Z. Si, S. D. Sherpa, D. W. Hess and A. Bongiorno, J. Phys. Chem. C, 2014, 118, 26402 CrossRef.
  46. Y. Liao, Y. Huang, S. Dong, Y. Zhong, J. Hao, C. He, Z. Jie and X. Song, Electrochim. Acta, 2016, 194, 136 CrossRef CAS.
  47. W. Zhang, Z. Chen, X. Guo, K. Jin, Y. Wang, L. Li, Y. Zhang, Z. Wang, L. Sun and T. Zhang, Electrochim. Acta, 2018, 278, 51–60 CrossRef CAS.
  48. J. W. Lee, J. M. Ko and J.-D. Kim, Electrochim. Acta, 2012, 85, 459–466 CrossRef CAS.
  49. X. Zhao, H. Dong, Y. Xiao, H. Hu, Y. Cai, Y. Liang, L. Sun, Y. Liu and M. Zheng, Electrochim. Acta, 2016, 218, 32–40 CrossRef CAS.
  50. Y. Liao, Y. Huang, D. Shu, Y. Zhong, J. Hao, C. He, J. Zhong and X. Song, Electrochim. Acta, 2016, 194, 136–142 CrossRef CAS.
  51. H. K. Song, H. Y. Hwang, K. H. Lee and H. D. Le, Electrochim. Acta, 2000, 45, 2241–2257 CrossRef CAS.
  52. Y. Sato, K. Itoh, R. Hagiwara, T. Fukunaga and Y. Ito, Carbon, 2004, 42, 3243 CrossRef CAS.
  53. L. Z. Li, Z. Xin, H. Ji, M. D. Stoller, L. Lai, S. Murali, S. Mcdonnell, B. Cleveger, R. M. Wallace and R. S. Ruoff, Energy Environ. Sci., 2012, 5, 9618 RSC.
  54. Y. Xu, C. Y. Chen, Z. Zhao, Z. Lin, C. Lee, X. Xu, C. Wang, Y. Huang, M. I. Shakir and X. Duan, Nano Lett., 2015, 15, 4605 CrossRef CAS PubMed.
  55. H. L. Guo, P. Su, X. Kang and S. K. Ning, J. Mater. Chem. A, 2013, 1, 2248 RSC.
  56. F. Withers, Nanoscale Res. Lett., 2011, 6, 526 CrossRef.
  57. F. Miao, C. Shao, X. Li, K. Wang, N. Lu and Y. Liu, J. Mater. Chem. A, 2016, 4, 5623 RSC.
  58. Z. Chen, K. Liu, S. Liu, L. Xia, J. Fu, X. Zhang, C. Zhang and B. Gao, Electrochim. Acta, 2017, 237, 102 CrossRef CAS.
  59. C. Chen, G. Xu, X. Wei and L. Yang, J. Mater. Chem. A, 2016, 4, 9900 RSC.
  60. R. Zhang, X. X. Jing, Y. T. Chu, L. Wang, W. J. Kang, D. H. Wei, H. B. Li and S. L. Xiong, J. Mater. Chem. A, 2018, 6, 17730 RSC.

Footnote

Electronic supplementary information (ESI) available: TEM and SEM images, digital photographs, CV curves at different scan rates, and rate capabilities of the RGH, HGH, NF-GH and NF-HGH. See DOI: 10.1039/c9se00142e

This journal is © The Royal Society of Chemistry 2019