Reduced graphene oxide derived from used cell graphite and its green fabrication as an eco-friendly supercapacitor

Abstract

Graphite extracted from a used primary cell was converted into reduced graphene oxide (rGO) using calcium carbonate together with rapid and local Joule heating by microwave irradiation. Electrodes were prepared by ultrasonically dispersing rGO in biodegradable poly(vinylpyrrolidone) (PVP) binder and coating this on recyclable poly(ethyleneterephthalate) (PET) sheet using a low cost screen printing technique. The use of the same polymer (PVP) as a binder, in addition to as the solid polymer electrolyte (SPE), enhances the compatibility and ionic conductivity of the hydrophobic rGO electrode in the supercapacitor system. Further, the phosphoric acid (H₃PO₄)-doped biodegradable SPE was screen printed for the first time on the rGO electrodes. Ionic conductivity and dielectric studies of the SPE were carried out at different temperatures and different dopant acid concentrations. The morphology, composition and structure of the graphene electrode components were characterized using Fourier transform infrared spectroscopy (FTIR), Raman spectroscopy, X-ray diffraction (XRD), scanning electron microscopy (SEM) and X-ray photoelectron spectroscopy (XPS) methods. Transmission electron microscopy (TEM) images showed a single layer or a few layers of rGO sheets and selected area electron diffraction showed the presence of slight defects. The fabricated environmentally friendly, industrially favorable and green supercapacitor showed a specific capacitance of 201 F g⁻¹ and cyclic stability with 97% retention of the initial capacitance over 2000 cycles. Furthermore, the performance of this green supercapacitor is comparable to that of those fabricated using rGO synthesized from commercial graphite and in other literature reports.

---

Introduction

The significance of energy devices like supercapacitors, which provide a short load cycle for applications in memory back-up systems, electric vehicles, displays, and energy capture from solar cells, has gained the attention of many researchers in recent decades.¹ Due to this upsurge in supercapacitor research, many materials have been synthesised or modified to achieve the desired properties. Nevertheless, energy devices last for hundreds to millions discharge cycles and then are dumped in landfill. Since the ban on the use of mercury in primary batteries, they are considered to be less harmful to the environment, but many countries are implementing battery directives,² which insist that battery manufacturers collect and recycle used batteries for safety reasons. Primary batteries such as zinc–carbon, alkaline batteries and secondary rechargeable alkaline manganese batteries utilize graphite as the cathode material. Graphite is becoming an important raw material in many applications, such as energy devices, micro-sensors, super-adsorbents, semiconductors^3–6 etc. The fact that graphite has a wider range of both current and emerging uses raises cost and environmental issues around demand for and disposal of graphite. With the increase in research and technology to improve these devices, there will be a hundred times increase in global graphite requirements compared to the present demand. Although graphite is presently available in abundance in nature, sooner or later, to meet the huge demand for graphite, it will be considered similarly to rare earth materials from key supplier countries by profitable businesses. This will impact global demand, resulting in a price hike and a need for purer graphite. Recycling industries will certainly receive a boost and this will decrease the environmental impact of battery disposal. A patent has also been filed for obtaining low cost graphite from recycled tyres for lithium batteries.⁷ We are inspired by the statement that the “greenness” of a battery not only depends on the kind of materials used in the battery, but also on how well the battery is managed throughout its life.^8,9 We have hence made an effort to join the current green revolution, using graphite from worn-out primary cells as an electrode material for supercapacitors. Usually, high purity graphite is used as the cathode material in primary cells, like zinc–carbon cell.¹⁰ When compared to commercial graphite, cell graphite is highly crystalline and contains traces of carbon black. This requires a simple process to achieve an amorphous character and extract pure graphite, and finally convert it into rGO.

Many methods have been used for conversion of graphite to rGO.^11,12 A few green methods have also been implemented, such as that described by Chen¹³ without the use of polymer or surfactant. However, the use of ammonia in this method is a concern when this method is utilised on a large scale. Microwave treatment has also been used for exfoliation of graphite oxide to rGO.¹⁴ The green method wherein ionic liquid-assisted microwave reduction of GO gave a specific capacitance of 135 F g⁻¹ was reported to be rapid and facile.¹⁵ Using Gum Arabic and ultrasonification, graphite was exfoliated, but again the concern is that to bring about 100% pure graphene, 100 h of ultrasonification and acid treatment were required.¹⁶ Similarly, many attempts have been made to develop eco-friendly methods to prepare rGO,^17–22 which are usually associated with complex processes for removal of reducing agents. Using sodium carbonate, rGO was efficiently reduced from GO, yet took 4 h for reduction.²³ Hence, in this study, a new approach to microwave treatment was employed along with the use of environmentally abundant calcium carbonate to produce rGO from graphene oxide solution efficiently within a few minutes.

Graphene has been used as an electrode material in supercapacitors.^1,24–26 A single atomic layer of graphene having high surface area is an ideal electrode material, but complete removal of heteroatoms and functional groups will affect the capacitance.²⁷ Graphene is known to have a theoretical surface area of 2630 m² g⁻¹ and Hantel et al.²⁸ claimed to have achieved 2687 m² g⁻¹ from partially reduced graphene oxide. Fewer layers of graphene gave a surface area of 1400 m² g⁻¹²⁹ and were porous when prepared using a microwave.³⁰ Highly porous 3D graphene produced from biomass showed good conductivity and a specific capacitance of 231 F g⁻¹.³¹ Graphene and rGO are slightly different in structure, since rGO has more lattice defects and a trace amount of functional groups when compared to graphene. A rGO/polymer binder using mesoporous rGO was reported to have good ionic liquid accessibility and hence showed a specific capacitance of 250 F g⁻¹.³² Even macroporous graphene with low BET surface area adsorbed oil significantly better than micro or mesoporous graphene sheets.⁵ Nonetheless, compared to the surface area of carbon nanofibers and activated carbon, rGO has a lower surface area, but still some of these unique properties make rGO useful as an electrode material for supercapacitors, especially in acid doped electrolytes. The surface hydrophobicity and poor dispersibility³³ of graphene make it difficult for liquid electrolytes to gain access and form a double layer at the electrode/electrolyte interface region. Hence, in this study eco-friendly biodegradable PVP was used as a binder. PVP has excellent wetting properties, since it stays sticky even at high humidity conditions, and it is normally used as a binder in printing technology.

Lithium ion energy devices are considered to have high energy density and are widely used, even though they are still associated with safety risks, due to the flammability of the organic electrolytes. The use of the same polymer as the electrolyte, as well as the separator, has been demonstrated to result in much safer, lighter and flexible energy devices.^34,35 Nevertheless, environmental factors must be considered, since the polymer electrolytes used so far are obtained from petroleum sources and the separators used cannot be recycled.^36,37 In response to these environmental concerns, we first prepared solid biodegradable polymer electrolytes for use in supercapacitors.³⁸ Although the use of lithium salts in these electrolytes resulted in good specific energy, they are still harmful when used in large amounts. Even though lithium salts are available in abundance, the large requirements of energy devices, such as batteries, mean that cost and dependence on good supply are likely to increase and this raises further environmental concerns, as lithium cannot yet be recycled with high purity. Phosphoric acid was therefore used as a dopant in the present system. Many stable polymer electrolytes based on H₃PO₄, H₂SO₄ and other strong acids have been reported to have conductivity ranging from 10⁻² to 10⁻⁵ S cm⁻¹.^39–41 Since conductivity depends on the acid and water concentration,⁴² the polymer used must have good proton conduction and water retention properties. SPEs were adapted as thick films, as they have the desired viscosity and adhesive properties that make them suitable for screen printing.

Printing techniques are manufacturing friendly and the uniformity of coating obtained is high, therefore one can easily achieve the desired output for the coated material. Although a good coating of graphene with better properties was achieved using a transfer printing process,⁴³ it still required special tuning for adhesion to substrate. An inkjet printed rGO electrode showed a low voltage window in cyclic voltammetry.⁴⁴ Roll coating, spin coating and screen printing have proved useful in many electrodes, sensors and OLEDs.^28,45–47 Compared to roll and spin coating methods, screen printing can be carried out on a large scale with uniform thickness and within a short period of time. Screen printing inks are also beneficial, since the dry matter content of the ink can be extremely high.⁴⁸ Lu et al. observed that the surface area of the rGO was not harmed by a screen printing method.⁴⁶ Recently, screen printed rGO/polyaniline electrodes achieved a specific capacitance of 236 F g⁻¹.⁴⁹ Hence, this eco-friendly and low cost screen printing method was used to coat rGO on a PET substrate. To avoid leakage due to liquid electrolyte and complex packing of the supercapacitor, we used phosphoric acid (H₃PO₄)-doped biodegradable PVP as the solid polymer electrolyte (SPE). The basic idea is to bring about good adhesion of the PVP electrolyte with the rGO/PVP binder electrode and thus enhance the double layer formation between the electrode/electrolyte interface.

In view of the aforementioned aspects, we report here for the first time a completely environmentally-friendly green supercapacitor, fabricated using rGO derived from used cell graphite as an electrode material, along with a biodegradable PVP binder and a H₃PO₄-doped PVP as a SPE. Characterization of the prepared graphene-like material and PVP was performed using FTIR, Raman spectroscopy, XRD, SEM and XPS, TEM, SAED methods and electrochemical techniques. Differential scanning calorimetry (DSC) and biodegradation studies can be found in Fig. S1 and S4 of the ESI.† Ionic conductivity and dielectric studies of the SPE (PVP/H₃PO₄, different concentration ratios) were carried out at different temperatures and the respective activation energies were calculated. Screen printed silver paste was used as a current collector. The supercapacitor properties, such as cyclic voltammetry (CV), AC impedance and galvanostatic charge–discharge (GCD) studies, were determined. The results have also been compared with supercapacitors fabricated using rGO derived from commercial graphite.

Experimental

Materials

Completely discharged AA or AAA size zinc–carbon primary cells of various companies were collected from a nearby collecting shop. PVP K30 (LobaChemi) for the binder and PVP (mol. wt 40 000, LobaChemi) for the solid electrolyte were used as obtained. H₃PO₄ (85%, Merck) was used as the dopant acid. Screen printable silver ink was purchased from Toyo ink (Japan). Print grade A-PET sheet was used for the study.

Conductivity measurements

Blocking emery polished stainless steel (SS) sheets were used as current collectors for the ionic conductivity measurements. SPEs were prepared by mixing PVP (high molecular weight) and 1.0 M H₃PO₄ in 2 : 0.5, 2 : 1, 2 : 1.5 and 2 : 2 wt% ratios. A calculated amount of Millipore water was added drop wise to these mixtures until a jelly texture developed. The required amounts of the prepared jelly inks were screen printed twice on the separate SS sheets. The ink dispersed slowly and occupied mesh marks during drying in a vacuum oven at 60 °C for 2 h. The thickness of the SPEs was ∼0.1 mm. The screen printed SS sheets were cut into 3 × 2 cm electrodes and fastened on another SS electrode using adhesive by simple hand pressure. The ionic conductivity of the SPE was measured in a cell by a complex impedance method over a frequency range of 1 MHz to 100 mHz using a small amplitude AC signal of 10 mV. The ionic conductivity of the SPE was calculated from the measured bulk resistance, area and thickness of the polymer thin film using the equation: σ = L/R_bA, where L is the thickness of the polymer electrolyte (cm), A is the area of the blocking electrode (cm²) and R_b is the bulk resistance of the polymer electrolyte. From these AC impedance data, dielectric studies were carried out. The formulae and relationships between complex impedance, dielectric permittivity and dielectric modulus can be found elsewhere.⁵⁰ PVP is hygroscopic in nature, hence the influence of ambient humidity must be taken into account while studying it at different temperatures. The ambient humidity in the cell was thus equilibrated with the water content in the SPE during the measurements. The temperature dependence of the ionic conductivity was measured using a PID controlled oven (SES instruments, model PID 200) under a controlled temperature range of 20–80 °C by giving sufficient time at each temperature to ensure that the sample was in equilibrium.

Preparation of rGO from battery graphite

The used primary cells were cut open and the graphite rods were carefully removed. The rods were washed with Millipore water and subjected to polishing using emery paper. Rods were crushed using a pestle and mortar and washed with Millipore water to ensure removal of impurities from the primary cell material. A lustrous powder was obtained after drying. The graphite powder was then sieved using 100 micron mesh. The graphite powder was oxidised to graphite oxide using the modified Hummers method.^51,52 The graphite oxide thus obtained was washed several times with Millipore water until the suspension attained a pH of 7. This suspension was then further subjected to ultrasonification for 30 min to obtain brownish graphene oxide (GO) solution. This was centrifuged at 2000 rpm for 20 min and dried in a vacuum oven to get GO flakes. CaCO₃ was mixed with GO flakes in 1 : 2 wt% ratios and suspended in 20 ml Millipore water. The solution mixture was placed in a domestic microwave oven and exposed to medium microwave for 10 min. Typically, obtained hydrophobic rGO suspension associated with calcium salts was removed by adding 1 M HCl and washed several times with Millipore water. A small amount of the obtained slurry was flame tested to ensure absence of a brick red color, which is due to calcium ions in the suspension. The slurry was dried in a vacuum oven for 2 h at 60 °C.

Electrode preparation and fabrication of supercapacitors

The dried rGO powder was mixed well with PVP binder at 4 : 1 ratio. A calculated amount of Millipore water was added drop wise to the mixture until a paste is obtained (∼1 g ml⁻¹). The paste was ultrasonicated for 10 min and allowed to set at ambient temperature for 2 h. Using 200 micron mesh and manual screen printing technique, the rGO/PVP paste was coated on to the PET surface. The coated materials were dried in a vacuum oven at 60 °C for 2 h. Conductive silver ink was screen printed using 100 micron mesh on the edges of the rGO/PVP binder coated area and allowed to dry at ambient temperature. The weight of electrode material was ∼3 mg. Electrical conductivity of the electrodes was studied using four-point probe method (SES instruments, Roorkee).

A PVP/H₃PO₄ polymer electrolyte solution exhibiting a high ionic conductivity was screen printed twice on the prepared rGO electrode and allowed to set at 50 °C for 1 h. The supercapacitor was fabricated as follows: the prepared rGO electrode containing SPE was cut into a 3 cm × 2 cm piece and placed on another rGO electrode without SPE and fastened using non-conducting adhesive by simple hand pressure.

Characterization

FTIR measurements of the graphite, graphene oxide (GO), rGO, rGO/PVP binder, un-doped and acid doped PVP polymer electrolyte were carried out at room temperature using Shimadzu FTIR 8400S (Japan) spectrophotometer. KBr disk method was employed for this study. The micro SEM images were taken using ZEISS EVO 18 Special Edition. TEM and SAED images were obtained from Tecnai 20 G2, (Netherland). XRD was studied using Rigaku’s MiniFlex 600. XPS was determined using non-monochromatic Al Kα X-ray source (1486.6 eV) with pass energy of 50.0 eV for the general scan and 40 eV for the core level spectra of each element. Laser-Raman was analyzed using Renishaw Invia Raman Microscope having source: He–Ne laser 633 nm, 18 mW. DSC measurements were taken on first run using Shimadzu DSC 60 model instrument under nitrogen atmosphere (Fig. S1†). Electrochemical characterization was carried out by CV, AC impedance and GCD studies. All the electrochemical studies were carried out using a BioLogic SP-150 instrument.

Results and discussion

Possible reduction mechanism of GO

CaCO₃ was used to reduce GO. CaCO₃ in aqueous medium produces calcium and carbonate ions. Furthermore, carbonate ions undergo hydrolysis to give hydroxide ions and bicarbonate ions, which again hydrolyse to produce hydroxide ions.²³ These hydroxide ions show high polarizability and efficiently absorb microwave energy and generate heat, which in turn reduces the GO within a few minutes. Moreover, GO undergoes strong deoxygenation in strong alkali solutions.^53,54 The CaCO₃–rGO mixture can be separated by adding dilute HCl followed by repeated centrifuging. The role of CaCO₃ is not only that of a reducing agent, but is also useful in removing the un-reacted H₂SO₄ used during Hummers method. CaCO₃ will readily react with this to form CaSO₄, CO₂ and water, thus extensive cleaning after oxidation of graphite can be easily accomplished.

Conductivity and dielectric studies

The ionic conductivities of the SPEs calculated at different temperatures using R_b values taken from AC impedance spectra are shown in Fig. 1a. The ionic conductivities increased with increase in acid concentration and temperature. The temperature dependence of the conductivity follows the Arrhenius equation σ = σ_o exp(−E_a/k_oT), where E_a is activation energy for conduction, σ_o is the pre-exponential factor and k_o is the Boltzmann constant. It implies that, for higher acid concentration, ionic conductivity occurs by means of Vehicle-type mechanisms in the SPE matrix, while, at lower concentrations of acid, proton transfer takes place via the interstitial water pathway, i.e. a Grotthus-type mechanism.⁵⁵ This is confirmed by E_a values as a function of acid concentration, i.e. ranging from 0.11–0.30 eV. The Nyquist plot of the sample with highest ionic conductivity is shown in Fig. 1b, since similar spectra were obtained for all the SPEs under investigation. The highest ionic conductivity value of 2.2 × 10⁻⁴ S cm⁻¹ was reached at 80 °C. A depressed semicircular arc in the higher frequency region was observed due to the migration of protons through polymer electrolyte. In contrast, in the low frequency region a spike was observed to be angled towards the real part of the impedance spectra, which is attributed to the double layer/polarization at the electrode/electrolyte interface and associated hindrance to diffusion of ions towards electrode. Considering the effect of temperature on the SPEs, the R_b value decreased due to enhancement of the segmental motion of the polymer chains revealing viscous channels (hydrogen bonded water) for proton mobility.

Fig. 1 (a) Variations of conductivities of SPEs at different temperature, (b) Nyquist impedance plots of the SPE at different temperatures.

Fig. 2a depicts dielectric constant (ε_R) as a function of frequency, wherein ε_R decreases with increasing frequency. Since protons are unable to exchange with the blocking electrodes, a non-Debye-type behavior was observed in the low frequency region, whilst, in the high frequency region, high periodic reversal of the electric field occurred at the interface, reducing the contribution of charge carriers towards the electrode. The variation in the dielectric constant could be due to a hopping conduction mechanism, similar to dielectric polarization.⁵⁶ High dielectric constant indicates good dissociation of acid in the polymer system. As the temperature increases, the dielectric constant also increases due to an increase in the orientation of dipoles in the direction of the applied field and hence increase the ionic conductivity of the SPE. Fig. 2b depicts dielectric loss (ε_I) as a function of frequency, wherein ε_I decreases with increasing frequency. In an acid-based SPE, the role of water in the ionic conductivity is important, since free charge carriers significantly affect the polarization at the electrode/electrolyte interface. It was observed that ε_I increases at low frequency and decreases in the higher frequency region. It is worth noting that there are no relaxation peaks due to charges building up at the interface during the free charge motion within the material. The ε_I increased with an increase in the temperature, indicating that water largely influences ionic conductivity in the sample. Fig. 2c and d shows the real, M_R, and imaginary, M_I, parts of electrical modulus as a function of frequency. The electric modulus study helps one to understand the influence of frequency on the bulk of the SPE, since polarization is negligible in this analysis. M_R shows a long tail feature in the low frequency region, indicating the capacitive nature of the SPE.⁵⁷ With an increase in temperature, less variation was observed due to the fast motion of protons in the bulk and this was confirmed in AC impedance studies. M_I show peaks in the higher frequency region, which can be attributed to ionic conduction of protons, which are spatially confined to their potential wells and perform only localized motion. As the temperature increased, the M_I decreased, liberating these ionic wells and thereby confirming the Grotthus-type mechanism in which a proton moves rapidly from H₃O⁺ to a hydrogen-bonded water molecule and is transferred further along a series of hydrogen-bonded water molecules by a rearrangement of hydrogen bonds.⁵⁸

Fig. 2 Plots of dielectric studies versus frequency at different temperatures (a) dielectric constant, (b) dielectric loss, (c) real part of electric modulus, and (d) imaginary part of electric modulus of 1 M H₃PO₄ doped SPE.

Characterization

The structure and distribution properties of cell graphite, rGO + binder and rGO were characterized using XRD analysis, as shown in Fig. 3a. For battery graphite, the intense and sharp peaks at 2θ = 25.8° and 2θ = 22.5° indicated the highly organized layered structure, with an interlayer spacing of 0.34 to 0.39 nm, respectively.⁵⁹ Lattice defects were observed due to breakdown of the hexagonal symmetry of graphite induced by mechanical grinding.⁶⁰ After reduction, low intensity peaks were observed and were found to be shifted to 2θ = 15.0° and 2θ = 22.1°, corresponding to an increase in interlayer spacing to 0.58 nm and 0.41 nm. This indicates efficient exfoliation of battery graphite during microwave reduction of GO in the presence of CaCO₃. A broad peak observed after mixing of PVP binder with rGO shows the uniform distribution of rGO in the polymer matrix while maintaining the rigid structure sufficiently to allow easy access to ions during charging and discharging. Although a high peak was observed at 2θ = 21.4°, the interlayer spacing was 0.41 nm which was found to be same as the rGO. This implies that the binder was able to decrease the crystallinity of rGO to some extent but could not penetrate between the two graphene sheets. When compared with the XRD pattern of commercial graphite (Fig. S2a†), the intensity of battery graphite was much higher, indicating its highly crystallinity nature. rGO obtained from commercial graphite showed broader peaks than rGO from battery graphite, indicating that our reduction process was able to produce the amorphous characteristics of rGO sheets with few layers.

Fig. 3 (a) XRD pattern of battery graphite, rGO and rGO + binder, (b) Raman spectrum of GO and rGO, (c) FTIR spectra of graphite, GO and rGO, and (d) FTIR spectra of pure PVP, PVP + H₃PO₄ and PVP + rGO.

Raman spectroscopy was used as a non-destructive tool to characterize structural changes, like disorder and defects. D and G bands are expected in the Raman spectra of graphite and rGO, which are correlated to first order scattering of E_2g phonons of sp²-bonded carbon atoms and j-photons of A_1g symmetry of sp³-bonded carbon atoms of disordered rGO.⁶¹ As shown in Fig. 3b, for GO the D band appeared at 1384 cm⁻¹ and the G band was located at 1627 cm⁻¹, while for rGO the D band shifted and broadened at 1346 cm⁻¹ and the G band at 1598 cm⁻¹, indicating reduction of GO⁵⁹ during microwave treatment. The peaks at ∼2700 cm⁻¹ are referred to as 2D peaks, since they are second order D peaks that also behave similarly in shifts with respect to GO and rGO. Furthermore, a significant change in shape and intensity of the 2D peak of rGO are observed compared to GO and roughly five times more intense than G peak. This results matches those of ref. 62 and gives an indication of a varying but small number of layers in the rGO.

Fig. 3c shows the FTIR spectra of battery graphite, GO and rGO. The graphite spectrum showed skeletal vibrations at 1600 cm⁻¹, while no other significant peak was found. In GO, the presence of different type of oxygen functionalities was confirmed by a peak at 3401 cm⁻¹ due to OH stretching vibrations. Stretching vibrations due to C=O, C–OH and C–O were observed at 1720 cm⁻¹, 1220 cm⁻¹ and 1060 cm⁻¹, respectively.⁶³ The intensity of the peak at 1620 cm⁻¹ was increased, suggesting an increase in adsorbed water molecules. After reduction of GO, the OH stretching at 3401 cm⁻¹ significantly reduced, while a trace quantity of C=O at 1720 cm⁻¹ still remained along with C–O at 1060 cm⁻¹. However, phenol C=C ring stretching at 1587 cm⁻¹ was also observed.⁶⁴ The absence of CO₃²⁻ group, which forms a weak peak at 870 cm⁻¹, and intensive peaks at 1465 and 1530 cm⁻¹ indicates complete removal of carbonate impurities associated with rGO during the reduction process. Fig. 3d shows FTIR spectra of pure PVP and H₃PO₄-doped PVP SPE. The C=O stretching band at 1670 cm⁻¹ for pure PVP sharpened after the addition of acid, indicating strong interactions. The asymmetric stretching at 2954 cm⁻¹ and symmetric stretching at 2890 cm⁻¹ (CH₂) were less intense. Less intense –C–N stretching vibrations and –C–H bending vibrations of the acid-doped PVP polymer were observed at 1285 and 1443 cm⁻¹, respectively, when compared to pure PVP. This indicates that the acid is well-dissociated in the PVP matrix, leading to interactions with the polymer chains. The PO₄³⁻ group formed typical intense peaks at 570–600 cm⁻¹ and at 1100 to 1200 cm⁻¹.⁶⁵ The peaks at 3480 cm⁻¹ were relatively broad due to adsorbed water in both the samples. The spectrum of PVP binder + rGO shows similar characteristics to acid-doped PVP, the only difference being the absence of the PO₄³⁻ group. This also implies that a non-covalent interaction exists between rGO and PVP binder, which would play an important role in stability of the electrode material, whilst maintaining its spongy nature for easy accessibility of ions. FTIR spectra (Fig. S2b†) of commercial derived graphite materials showed similar characteristic peaks to battery graphite, but upon reduction the rGO spectrum showed very low intensity peaks forOH, C=O and C–O. This result for commercial rGO confirms the XRD results implying complete reduction of GO using CaCO₃ and microwave heating.

Fig. 4a shows the C1s XPS spectrum of GO after deconvolution. The peaks show a considerable degree of oxidation of graphite corresponding to carbon atoms in different functional groups, which can be observed at 284.2, 284.8, 286.1, and 288.4 eV originating from C–C, C–O, C=O, and O–C=O groups, respectively. In Fig. 4b the O1s XPS spectrum of GO shows peaks at 529.4 and 532 eV from C=O and C–O, respectively. However, after microwave treatment the peaks of rGO (Fig. 4c and d) reduced slightly and shifted to a higher binding energy region. The peaks at 286.9, 287.4, 288.7 and 290.2 corresponding to C–C, C–O, C=O and O–C=O decreased significantly, indicating efficient deoxygenation of GO and the formation of rGO.⁵⁹ Furthermore, the calculated C/O ratio changed from 1.75 to 4.40. Nevertheless, the intensity in the O1s spectrum of rGO remained the same, but shifted slightly to a higher binding energy and sharpened. This implies trace amount of C–O groups are associated with the rGO.

Fig. 4 (a and b) C1s and O1s XPS spectra of GO, (c and d) C1s and O1s XPS spectra of rGO.

Fig. S3† is a representative SEM image of battery graphite powder, showing hexagonal rods along with defects due to crushed graphite grains having a size ranging from 2–4 μm. Fig. 5a is a SEM image showing a particle size ranging from 500 to 800 nm for rGO powder obtained by CaCO₃ and microwave treatment of GO. An almost regular grain size, broken from the hexagonal rod structures of graphite with agglomeration and forming mesoporous structures, was obtained. The Fig. 5b (front view) and c (cross section view) shows SEM images of screen printed rGO and PVP binder. These images suggest that rGO particles are less aggregated with few clumped sheets and are stuffed in PVP binder. This PVP prevents rGO from re-stacking during the pressure applied for fabrication of supercapacitor. Due to this morphology, the ions can readily access not only the 2D structure of rGO, but also the 3D structure wherein well-wetted PVP chains provides channels allowing the ions easy access through these stuffed rGO grains, hence contributing to capacitance.

Fig. 5 (a) rGO powder and (b and c) screen printed rGO + PVP binder top and cross section view.

Fig. 6a depicts TEM images of graphene sheets ranging from single to few layers which were found to be similar to rGO reported by Stoller et al.⁶⁶ Rippling structures, folding and scrolling nature of the rGO sheets can be observed. Agglomerations are also observed, illustrating how both sides of the graphene sheets are difficult for the electrolyte to access. The use of the same polymer as binder and SPE therefore creates passage of ions between these graphene sheets. However, the number of layers was not confirmed by TEM and an uncertain number of layers was even evident from Raman spectra. In Fig. 6b, the selected area electron diffraction (SAED) showed multiple hexagonal rings of different intensity and spot size on the same circle confirming the random orientation of graphene layers within the graphene sheets.⁶⁷ The existence of few sheets and the occurrence of spots in different positions, other than in a circle having a hexagonal pattern, shows the presence of oxygen and hydroxyl groups in a few graphene sheets, but it is worth mentioning that almost all oxygen and hydroxyl groups are expelled using rapid, facile microwave treatment in the presence of CaCO₃.

Fig. 6 (a) TEM image of rGO powder and (b) SAED of rGO sheets.

The eco-friendly aspects of SPEs were measured using a soil burial biodegradation method (Fig. S4†). SPEs having higher acid concentrations showed higher degradation, indicating that a higher concentration of acid is able to break the intermolecular bonding of polymer chains and create more sites for degradation.

Supercapacitor studies

Fig. 7a shows the CV of the fabricated supercapacitor as a function of scan rate. All the curves are nearly rectangular, which is characteristic of an ideal capacitor. As the scan rate increased, there was less distortion in the curves and high rate properties. This suggests that a high charge transfer kinetic exists at the electrode–electrolyte interface. The specific capacitance values of the supercapacitors have been calculated using the equation;⁶⁸where I is the average current, ΔV is the scan rate and m is the mass of materials at each electrode. A factor of 2 is used because the series capacitance is formed in a two-electrode system. The maximum specific capacitance value is 201 F g⁻¹ at 2 mV s⁻¹. This specific capacitance value is almost similar to the value of specific capacitance obtained from rGO prepared from commercial graphite (Fig. S5†). Although the battery graphite was more crystalline than commercial graphite, the reduction of GO using CaCO₃ and a microwave resulted in a uniform grain size of rGO, thus obtaining similar specific capacitance. The rGO obtained from used battery graphite can hence be effectively used as electrode material.

Fig. 7 (a) CVs of supercapacitor at different scan rates, (b) AC impedance plot of supercapacitor, (c) plots of normalized reactive power |Q|/|S|% and active power |P|/|S|% versus frequency (Hz), (d) GCD plots at different current densities, and (e) variation of specific capacitance of the supercapacitor during long-term cycling.

Normally, an electrochemical double layer is formed between electrode and electrolyte, rather than storing in the bulk of the SPE. A Nyquist plot (Fig. 7b) of this green supercapacitor can demostrate the importance of using the same polymer as binder and SPE. A small semicircle arc can be seen in the high frequency region, which represents the intrinsic internal resistance of the electrode material and electrolyte, followed by an almost 45° Warburg impedance in the low frequency region, which is the trend of ideal capacitive behaviour, and agrees with the CV results. The small semicircle in the high frequency region implies low intrinsic internal resistance due to improved hydrophilicity of the rGO material/PVP electrode materials, resulting in enhanced wettability and facilitating rapid electrolyte ion transport within the spacing of the graphene layers and polymer chains.⁶⁹ In addition, the ionic diffusion in the middle frequency region is achieved very fast, and a more vertical line seen in thr low frequency region indicates that the supercapacitor has higher ionic conductivity and electrochemical performance.⁷⁰ Introducing surface PVP as a binder to rGO, along with a SPE based on the same PVP, appears to be an effective way of improving the capacitance of an electrode. A similar pattern was observed for rGO derived from commercial graphite, but with a lower resistivity value (Fig. S5†). The values of fitted circuit R₁ + Q₂/(R₂ + W₂) can be found in Table S1.†

Fig. 7c shows normalized imaginary part (reactive power) |Q|/|S|% and real part (active power) |P|/|S|% of the complex power versus frequency plot of the supercapacitor at 303 K. As observed, the normalized imaginary part of power |Q|/|S|% increases as frequency decreases and a maximum is reached, wherein the supercapacitor behaves like a pure capacitor. At the same time, when |P|/|S|% is 100% at high frequencies, the supercapacitor behaves like a pure resistor, i.e. power is dispersed into the system and then |P|/|S|% decreases as frequency increases. The crossing of two plots appears when |P| = |Q| at the time constant τ₀. The theoretical details of this plotting technique can be found elsewhere.⁷¹ The calculated time constant was found to be equal to 3 ms, and hence indicates that the present system can be efficiently used at low frequencies.

In GCD studies, curves with triangular symmetrical distribution indicate good capacitive properties in a supercapacitor. Fig. 7d show the GCD curves of the supercapacitor at current densities of 0.2 A g⁻¹, 0.4 A g⁻¹ and 0.6 A g⁻¹. The specific capacitance was derived from the charge–discharge curve according to the following equation⁷²

where I is the applied discharge current, Δt is the discharged time after IR drop, ΔV is the discharge potential window after IR drop, and m is the mass of the single electrode materials. C_s values decreased with increasing current density 208, 184 and 162 F g⁻¹. Comparatively, the specific capacitance was almost the same as that for 3D graphene-based bulk materials,³¹ i.e. 231 F g⁻¹, and even that of hydrazine reduced GO.²⁶ Fig. 7d shows a slight voltage drop due to ohmic resistance and quite good stability was exhibited at low current density. This implies that if the graphene layers with the binder chains are adjusted to the size of ions, the pore volume may become saturated by electrolyte species, limiting the maximum operating voltage and the energy stored in the supercapacitor.⁷³ Even the presence of hygroscopic PVP would largely interfere by promoting proton transfer within the spacing of rGO materials, thereby causing a small deviation from ideal capacitor behaviour. However, with an increase in current density, the voltage drop decreased, indicating that a sufficient amount of ions was accommodated and, during discharge, intercalation/insertion and trapping of ions were not observed. Supercapacitor performance was therefore optimized with respect to wettability, based to a certain extent on the texture/structure of the electrode material and the SPE. Further intensive research is required to understand exact performance based on the texture of carbon materials and optimized ionic conductors, rather than just reaching a conclusionbased on maximum capacitance.^64,74 The energy (E) and power (P) densities and equivalent series resistance (ESR) were calculated from the following equations,

A maximum E of 11 W h kg⁻¹ (with P of 2 kW kg⁻¹) at 0.2 A g⁻¹ and P of 5 kW kg⁻¹ (with E of 8 W h kg⁻¹) at 0.6 A g⁻¹ were obtained. The ESR value was 4.7 ohm. E is found to be lower than that of other reported supercapacitors. Nevertheless, the high value of the P is well-suited for surge-power delivery applications.⁷⁵ Table 1 and 2 shows the comparison of specific capacitance, energy density and power density of supercapacitors with different electrode material prepared using various methods.

Table 1 Comparison of rGO derived from different methods and their specific capacitance

Electrode material and electrolyte	Reducing method	Specific capacitance	Ref.
Porous activated rGO, TEABF4/AN	KOH + 800 °C	GCD: 120 F g^−1 at 10 A g^−1	76
rGO + MnO2, 1 M Na2SO4	Hydrothermally	CV: 211.5 F g^−1 at 2 mV s^−1	77
rGO sheets, 6 M KOH	Thermal	GCD: 260.5 F g^−1 at 0.4 A g^−1	78
rGO, 0.1 M Na2SO4	Electrochemical	CV: 164.8 F g^−1 at 20 mV s^−1	79
		GCD: 150.4 F g^−1 at 5 A g^−1
rGO, 2 M H2SO4	Surfactants	GCD: 194 F g^−1 at 1 A g^−1	80
rGO, KOH	Chemically using hydrazine	205 F g^−1 at 0.1 A g^−1	26
rGO, 1 M H2SO4	Hydrobromic acid, 110 °C	348 F g^−1 at 0.2 A g^−1	81
rGO + metal oxide	Spray pyrolysis	CV: 687 F g^−1 at 5 mV s^−1	82
rGO	Ionic liquid assisted microwave reduction	135 F g^−1 at 2 A g^−1	15
Exfoliated rGO	Microwave	191 F g^−1 at 150 mA g^−1	83
rGO/PVP binder, PVP + H3PO4	CaCO3 and microwave	CV: 201 F g^−1 at 2 mV s^−1	Present work
		GCD: 208 F g^−1 at 0.2 A g^−1

---

Table 2 Comparison of power density and energy density of various electrode materials

Power density (W kg^−1)	Energy density (W h kg^−1)	Electrode material and electrolyte	Ref.
9838 at 8 A g^−1	90 at 0.5 A g^−1	Graphene, EMIMBF4	32
6.8k at 10 A g^−1	19.7 at 0.5 A g^−1	rGO + RuO2/rGO, PVA + H2SO4	40
246k at 8 A g^−1	58 at 2 A g^−1	rGO, EMIM·NTf2	15
32k	15	Activated carbon, TEMABF3	84
170	20.1	rGO + TiO2	85
200	11	rGO + Fe3O4	86
500	26	rGO, TEABF3	76
8.1k		rGO + ZnO	87
0.22k	30	MgO + carbon nanotude, 1 M LiPF6	88
2k at 0.2 A g^−1	11 at 0.2 A g^−1	rGO/PVP binder, PVP + H3PO4	Present work

---

Cyclic durability is an important electrochemical property of a supercapacitor for practical application. The C_s variation at 0.2 A g⁻¹ with the cycle number is shown in Fig. 7e. The specific capacitance was reversibly maintained with 97% retention of the initial capacitance over 2000 cycles. This superior performance was attributed to its excellent mechanical robustness and the intimate interfacial contact between the multiple components.³⁸ Since the initial decrease in capacitance was small, the use of same polymer as binder and SPE had only a minor effect when compared to other internal resistance in the device. This supercapacitor thus possesses good stability and a very high degree of reversibility during repetitive charge–discharge cycling.

Scheme 1 shows the preparation of the supercapacitor by the screen printing method and the possible mechanism for ion transport in the supercapacitor. Considering that PVP has a strong water adsorption capability,⁸⁹ the dried rGO/PVP coating has an affinity for water. When a SPE containing acid aqueous electrolyte was coated on the surface, the tendency of the electrode material for water absorption results in it readily accommodating some amount of the acid electrolyte and immediately creates channels for ionic movement by forming interactions with the polymer chains of PVP and the rGO layers. The addition of PVP binder avoids agglomeration of rGO, which normally blocks the accessible surface of the rGO. At the point of contact of the SPE and the electrode material, a high degree of continuous and interconnected ionic transport channels could be formed. This forms a kind of adaptive rGO/PVP gel film allowing easy access of ions. As discussed earlier, the transport of protons takes place via Grotthus-type mechanism in the SPE and they easily get transferred to the electrode/electrolyte interface region, thereby forming a double layer very rapidly. Charging–discharging thus results in higher power density and stability.

Scheme 1 Screen printing of rGO + binder and SPE on PET and fabrication of supercapacitor with possible mechanism of ions transportation.

Conclusion

In summary, we have successfully demonstrated the extraction of graphite from used cells and its conversion to rGO, using a green microwave reduction method in the presence of CaCO₃, for the first time. This rGO was mixed with eco-friendly PVP binder and screen printed on PET substrate, and supercapacitors were fabricated using these novel rGO electrodes and SPE. These fabricated supercapacitors showed a specific capacitance of 201 F g⁻¹ at 2 mV s⁻¹. Ionic conductivity of the SPE was not significantly affected by the variation in temperature. The dielectric studies showed good capacitive behavior of the material and the influence of water on the transport of protons. The characterization studies showed good interaction between rGO, PVP and SPE. High power density was exhibited with low ESR value. The high performance of the supercapacitor is attributed to the enhanced interaction between the adaptive electrode material and SPE. The obtained specific capacitance was comparable to that of a supercapacitor fabricated using rGO derived from commercial graphite.

Acknowledgements

The authors acknowledge with thanks the financial support received from the Defence and Research Development Organization (DRDO), Govt. of India, New Delhi. The facilities provided for printing by Manipal Technologies Limited, Manipal are greatly acknowledged.

References

1. L. L. Zhang, R. Zhou and X. S. Zhao, J. Mater. Chem., 2010, 20, 5983.
2. http://pt.farnell.com/images/en_UK/rohs/pdf/battery_directive_nov08.pdf, accessed July 2014.
3. F. Zhang, T. Zhang, X. Yang, L. Zhang, K. Leng, Y. Huang and Y. Chen, Energy Environ. Sci., 2013, 6, 1623.
4. L. Huang, Y. Huang, J. Liang, X. Wan and Y. Chen, Nano Res., 2011, 4, 675.
5. S. J. Yang, J. H. Kang, H. Jung, T. Kim and C. R. Park, J. Mater. Chem. A, 2013, 1, 9427.
6. L. Yan, Y. B. Zheng, F. Zhao, S. Li, X. Gao, B. Xu, P. S. Weiss and Y. Zhao, Chem. Soc. Rev., 2012, 41, 97.
7. A. K. Naskar, M. P. Paranthaman, Z. Bi, US Patent Application Filed: UTB 2980.0, Pyrolytic Carbon Black Composite and Method of Making Same, Ref: 6321–434 [19079.0380], 2013.
8. R. Lankey and F. McMichael, Rechargeable Battery Management and Recycling: A Green Design Educational Module, Green Design Initiative Technical Report, Carnegie Mellon University, 1999, http://www.ce.cmu.edu/GreenDesign/gd/education/Battery.pdf, accessed July 2014.
9. P. Poizot and F. Dolhem, Energy Environ. Sci., 2011, 4, 2003.
10. http://www.baystreet.ca/articles/research_reports/fundamental_research/FMS-Update-Nov2012.pdf, accessed July 2014.
11. S. Eigler, M. Enzelberger-Heim, S. Grimm, P. Hofmann, W. Kroener, A. Geworski, C. Dotzer, M. Röckert, J. Xiao, C. Papp, O. Lytken, H.-P. Steinrück, P. Müller and A. Hirsch, Adv. Mater., 2013, 25, 3583.
12. L. Tang, Y. Wang, Y. Li, H. Feng, J. Lu and J. Li, Adv. Funct. Mater., 2009, 19, 2782.
13. H. Chen, M. B. Müller, K. J. Gilmore, G. G. Wallace and D. Li, Adv. Mater., 2008, 20, 3557–3561.
14. M. Cai, D. Thorpe, D. H. Adamson and H. C. Schniepp, J. Mater. Chem., 2012, 22, 24992.
15. T. Kim, H. C. Kang, T. T. Tung, J. D. Lee, H. Kim, W. S. Yang, H. G. Yoon and K. S. Suh, RSC Adv., 2012, 2, 8808.
16. V. Chabot, B. Kim, B. Sloper, C. Tzoganakis and A. Yu, Sci. Rep., 2013, 3, 1378.
17. L. J. Cote, R. Cruz-Silva and J. Huang, J. Am. Chem. Soc., 2009, 131, 11027.
18. G. Williams, B. Seger and P. V. Kamat, ACS Nano, 2008, 2, 1487.
19. Y. Wang, Z. Shi and J. Yin, ACS Appl. Mater. Interfaces, 2011, 3, 1127.
20. K. H. Chang, Y. F. Lee, C. C. Hu, C. I. Chang, C. L. Liu and Y. L. Yang, Chem. Commun., 2010, 46, 7957.
21. W. Chen, L. Yan and P. Bangal, J. Phys. Chem. C, 2010, 114, 19885.
22. D. Zhang, X. Zhang, Y. Chen, C. Wang and Y. Ma, Electrochim. Acta, 2012, 69, 364.
23. Y. Jin, S. Huang, M. Zhang, M. Jia and D. Hu, Appl. Surf. Sci., 2013, 268, 541.
24. Y. Sun, Q. Wu and G. Shi, Energy Environ. Sci., 2011, 4, 1113.
25. M. D. Stoller, S. Park, Y. Zhu, J. An and R. S. Ruoff, Nano Lett., 2008, 8, 3498.
26. Y. Wang, Z. Shi, Y. Huang, Y. Ma, C. Wang, M. Chen and Y. Chen, J. Phys. Chem. C, 2009, 113, 13103.
27. D. Hulicova, M. Kodama and H. Hatori, Chem. Mater., 2006, 18, 2318.
28. M. M. Hantel, T. Kaspar, R. Nesper, A. Wokaun and R. Kötz, Electrochem. Commun., 2011, 13, 90.
29. Z. J. Li, B. C. Yang, S. R. Zhang and C. M. Zhao, Appl. Surf. Sci., 2012, 258, 3726.
30. V. Murugan, T. Muraliganth and A. Manthiram, Chem. Mater., 2009, 21, 5004–5006.
31. L. Zhang, F. Zhang, X. Yang, G. Long, Y. Wu, T. Zhang, K. Leng, Y. Huang, Y. Ma, A. Yu and Y. Chen, Sci. Rep., 2013, 3, 1408.
32. C. Liu, Z. Yu, D. Neff, A. Zhamu and B. Z. Jang, Nano Lett., 2010, 10, 4863.
33. H. K. Chae, D. Y. Siberio-Perez, J. Kim, Y. Go, M. Eddaoudi, A. J. Matzger, M. O’Keeffe and O. M. Yaghi, Nature, 2004, 427, 523.
34. A. S. Aricò, P. Bruce, B. Scrosati, J.-M. Tarascon and W. V. Schalkwijk, Nat. Mater., 2005, 4, 366.
35. Y. G. Sun, Nano Energy, 2013, 2, 811.
36. L. Nyholm, G. Nyström, A. Mihranyan and M. Strømme, Adv. Mater., 2011, 23, 3751.
37. R. C. Agrawal and G. P. Pandey, J. Phys. D: Appl. Phys., 2008, 41, 223001.
38. M. Selvakumar and D. K. Bhat, J. Appl. Polym. Sci., 2008, 110, 594.
39. J. Przyluski, Z. Poitarzewski and W. Wieczorek, Polymer, 1998, 39, 4343.
40. B. G. Choi, S.-J. Chang, H.-W. Kang, C. P. Park, H. J. Kim, W. H. Hong, S. Lee and Y. S. Huh, Nanoscale, 2012, 4, 4983.
41. C. Meng, C. Liu, L. Chen, C. Hu and S. Fan, Nano Lett., 2010, 10, 4025.
42. W. Wieczorek, Z. Florjaficzyk and J. R. Stevens, Electrochim. Acta, 1995, 40, 2327.
43. J. H. Chen, M. Ishigami, C. Jang, D. R. Hines, M. S. Fuhrer and E. D. Williams, Adv. Mater., 2007, 19, 3623.
44. L. T. Le, M. H. Ervin, H. Qiu, B. E. Fuchs and W. Y. Lee, Electrochem. Commun., 2011, 13, 355.
45. J. Ping, J. Wu, Y. Wang and Y. Ying, Biosens. Bioelectron., 2012, 34, 70.
46. T. Lu, Y. Zhang, H. Li, L. Pan, Y. Li and Z. Sun, Electrochim. Acta, 2010, 55, 4170.
47. H. Chang, G. Wang, A. Yang, X. Tao, X. Liu, Y. Shen and Z. Zheng, Adv. Funct. Mater., 2010, 20, 2893.
48. M. Valkiainen, S. Tuurala, M. Smolander and O. T. Kaukoniemi, RSC Energy and Environment Series No. 2, Innovations in Fuel Cell Technologies, printed enzymatic current sources, ch. 1, p. 16, 2010.
49. Y. Xu, M. G. Schwab, A. J. Strudwick and I. Hennig, Adv. Energy Mater., 2013, 3, 1035.
50. Y. N. Sudhakar and M. Selvakumar, Electrochim. Acta, 2012, 78, 398.
51. W. S. Hummers and R. E. Offeman, J. Am. Chem. Soc., 1958, 80, 1339.
52. N. I. Kovtyukhova, Chem. Mater., 1999, 11, 771.
53. D. R. Dreyer, S. Park, C. W. Bielawski and R. S. Ruoff, Chem. Soc. Rev., 2010, 39, 228.
54. X. Fan, W. Peng, Y. Li, X. Li, S. Wang, G. Zhang and F. Zhang, Adv. Mater., 2008, 20, 4490.
55. Y. N. Sudhakar and M. Selvakumar, J. Appl. Electrochem., 2012, 43, 21.
56. N. Rezlescu and E. Rezlescu, Phys. Status Solidi A, 1974, 23, 575.
57. M. Forsyth, D. R. MacFarlane, P. Meakin, M. E. Smith and T. J. Bastow, Electrochim. Acta, 1995, 40, 2131.
58. M. Selva Kumar and D. K. Bhat, Phys. Rev. B: Condens. Matter Mater. Phys., 2009, 404, 1143.
59. W. Chen and L. Yan, Nanoscale, 2010, 2, 559.
60. C. S. Wang, G. T. Wu and W. Z. Li, J. Power Sources, 1998, 76, 1.
61. M. Kumar, K. Singh, S. K. Dhawan, K. Tharanikkarasu, J. S. Chung, B.-S. Kong, E. J. Kim and S. H. Hur, Chem. Eng. J., 2013, 231, 397.
62. A. C. Ferrari, Solid State Commun., 2007, 143, 47–57.
63. Y. Xu, H. Bai, G. Lu, C. Li and G. Shi, J. Am. Chem. Soc., 2008, 130, 5856.
64. A. Bagri, C. Mattevi, M. Acik, Y. J. Chabal, M. Chhowalla and V. B. Shenoy, Nat. Chem., 2010, 2, 581.
65. http://www.intechopen.com/books/infrared-spectroscopy-materials-science-engineering-andtechnology/research-of-calcium-phosphates-using-fourier-transformation-infrared-spectroscopy.
66. M. D. Stoller, S. Park, Y. Zhu, J. An and R. S. Ruoff, Nano Lett., 2008, 8, 3498.
67. K. S. Hazra, J. Rafiee, M. A. Rafiee, A. Mathur, S. S. Roy, J. McLauhglin, N. Koratkar and D. S. Misra, Nanotechnology, 2011, 22, 025704.
68. G. Wang, L. Zhang and J. Zhang, Chem. Soc. Rev., 2012, 41, 797.
69. A. K. Mishra and S. Ramaprabhu, J. Phys. Chem. C, 2011, 115, 14006–14013.
70. X. Yang, F. Zhang, L. Zhang, T. Zhang, Y. Huang and Y. Chen, Adv. Funct. Mater., 2013, 23, 3353.
71. Y. N. Sudhakar, M. Selvakumar and D. K. Bhat, Ionics, 2012, 19, 277.
72. H. Yu, J. Wu, L. Fan, Y. Lin, K. Xu, Z. Tang, C. Cheng, S. Tang, J. Lin, M. Huang and Z. Lan, J. Power Sources, 2012, 198, 402.
73. F. Béguin, V. Presser, A. Balducci and E. Frackowiak, Adv. Mater., 2014, 26, 2219.
74. C. Largeot, C. Portet, J. Chmiola, P. Taberna, Y. Gogotsi and P. Simon, J. Am. Chem. Soc., 2008, 130, 2730.
75. A. Burke, J. Power Sources, 2000, 91, 37.
76. L. L. Zhang, X. Zhao, M. D. Stoller, Y. Zhu, H. Ji, S. Murali, Y. Wu, S. Perales, B. Clevenger and R. S. Ruoff, Nano Lett., 2012, 12, 1806–1812.
77. Z. Li, J. Wang, S. Liu, X. Liu and S. Yang, J. Power Sources, 2011, 196, 8160–8165.
78. B. Zhao, P. Liu, Y. Jiang, D. Pan, H. Tao, J. Song, T. Fang and W. Xu, J. Power Sources, 2012, 198, 423–427.
79. Y. Shao, J. Wang, M. Engelhard, C. Wang and Y. Lin, J. Mater. Chem., 2010, 20, 743.
80. K. Zhang, L. Mao, L. L. Zhang, H. S. On Chan, X. S. Zhao and J. Wu, J. Mater. Chem., 2011, 21, 7302.
81. Y. Chen, X. Zhang, D. Zhang, P. Yu and Y. Ma, Carbon, 2011, 49, 573–580.
82. A. Chidembo, S. H. Aboutalebi, K. Konstantinov, M. Salari, B. Winton, S. A. Yamini, I. P. Nevirkovets and H. K. Liu, Energy Environ. Sci., 2012, 5, 5236.
83. Y. Zhu, S. Murali, M. D. Stoller, A. Velamakanni, R. D. Piner and R. S. Ruoff, Carbon, 2010, 48, 2118–2122.
84. M. Yamagata, S. Ikebe, K. Soeda and M. Ishikawa, RSC Adv., 2013, 3, 1037.
85. C. Xiang, M. Li, M. Zhi, A. Manivannan and N. Wu, J. Mater. Chem., 2012, 22, 19161.
86. K. Karthikeyan, D. Kalpana, S. Amaresh and Y. S. Lee, RSC Adv., 2012, 2, 12322.
87. J. Wang, Z. Gao, Z. Li, B. Wang, Y. Yan, Q. Liu, T. Mann, M. Zhang and Z. Jiang, J. Solid State Chem., 2011, 184, 1421–1427.
88. K. Karthikeyan, S. Amaresh, V. Aravindan and Y. S. Lee, J. Mater. Chem. A, 2013, 1, 4105.
89. J. Zhang, G. Shen, W. Wang, X. Zhou and S. Guo, J. Mater. Chem., 2010, 20, 10824.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra08347d

---