Electrode materials based on α-NiCo(OH)₂ and rGO for high performance energy storage devices

Abstract

Described herein is a composite material based on reduced graphene oxide (rGO) and α-NiCo(OH)₂ nanoparticles exhibiting very fast charge/discharge processes, like that in capacitors combined with the high energy density of batteries, suitable for application to high performance energy storage devices. NiCo@rGO showed a superior performance compared to stabilized α-Ni(OH)₂ and α-NiCo(OH)₂ nanoparticles and Ni@rGO, the analogous composite of the former with rGO. In the nanocomposite, each rGO sheet is decorated with stabilized nickel–cobalt hydroxide nanoparticles conferring specific capacitances as high as 1300 F g⁻¹ and discharge rates superior to 100 A g⁻¹. This nanomaterial is also demonstrated to be remarkably stable, retaining up to 96% of its maximum capacitance after 5000 galvanostatic charge/discharge cycles.

---

Introduction

Recently, electrochemical capacitors/supercapacitors (ECs) have become promising candidates for energy storage systems due to their excellent performance in high power and high energy density conditions, associated with their long life cycle.¹ There are two types of supercapacitors: (a) the electric double layer capacitors (EDLCs) and (b) the electrochemical pseudocapacitors, characterized by high capacitive and faradaic charge storage, and classified based on their storage mechanism.² Specific capacitance, rate capability and stability are important parameters to estimate the performance of energy storage materials. For example, carbon-based materials possess high charge storage capacity at the electrode/electrolyte interface, responsible for high power densities and excellent charge–discharge rates but detrimental energy densities.³ In contrast, metal oxides/hydroxides and conducting polymers can achieve high specific capacitances and provide fast surface redox reactions, but their electronic/ionic conductivity limits the rate and cycling performance to a large extent.^3,4

It is known that the dimensions and morphology of electrode materials are quite important for their performance as supercapacitors, especially for rate capability, since the diffusion time of ions (t) is proportional to the square of the diffusion length,^1,5 L, and t ≈ L²/D. Accordingly, great efforts have been focused on the preparation of nanostructured materials exhibiting longer diffusion lengths and higher density of active sites maximizing the electrochemical performance.¹ However, most of the reports were mainly focused on the enhancement of the absolute specific capacitance, and less attention has been paid on attributes such as the rate capability.³

Nickel hydroxide (Ni(OH)₂) has received increasing interest as electrode material because it has high specific capacitance (theoretical specific capacitance of 2082 F g⁻¹) and well-defined redox processes.^6–8 Also, it is known that devices based on α-Ni(OH)₂ will have better performance but its metastable turbostratic structure is more or less easily converted to the thermodynamically stable and crystalline beta phase, spontaneously or after few charge–discharge cycles in alkaline medium,^9–12 precluding its use in electrochemical devices. Nevertheless, the incorporation of cobalt was shown to improve the conductivity, stability and electrochemical properties^13–16 because cobalt can substitute nickel in the mixed hydroxide material forming more conductive CoOOH during the charging process. In fact, various nickel/cobalt hydroxide electrodes have been prepared, and their enhanced electrochemical performance reported.^3,16–18

Graphene and its derivatives are interesting carbon nanostructures that brings new perspectives in many fields, due to its exceedingly high surface area and electrical conductivity^19–21 associated with lightweight and good mechanical properties. Thus, graphene can enhance the electrical conductivity and improve the properties of composite materials.^1,2,22–29 However, there are not many examples combining the properties of those materials in a synergic way maximizing the specific capacitance at high current densities. In fact, the viability of energy storage materials is defined by both, the rate performance and the specific power density, two critical issues that must be addressed simultaneously.³ Accordingly, herein reported are nanocomposites based on stabilized α-NiCo(OH)₂ nanoparticles strongly anchored in reduced graphene oxide sheets (rGO), realizing electrode materials characterized by high specific capacitance and charge/discharge rates for high performance energy storage devices.

Experimental

Synthetic procedures

All reagents and solvents were of analytical grade and used as received. Anhydrous glycerin, isopropyl alcohol, cobalt acetate tetrahydrate, sulfuric acid, sodium nitrate, sodium borohydride, potassium permanganate and potassium hydroxide were purchased from Synth Brasil. Nickel acetate tetrahydrate was purchased from Sigma-Aldrich, and n-butyl alcohol was acquired from Vetec. Graphite was purchased from Vonder.

Sol suspensions of nickel hydroxide and nickel/cobalt mixed hydroxide nanoparticles stabilized in the alpha crystalline phase were prepared according to a previously described method,^9,11,30 by dissolving 4.8 mmol of the metal acetates in 25 mL of glycerin and adding 9.6 mmol of KOH in n-butanol, at room temperature. Samples of α-Ni(OH)₂ and α-Ni_0.6Co_0.4(OH)₂, named NiCo, suspensions were prepared accordingly.

Graphite oxide (GrO) was prepared by modified Hummers' method.^31,32 Concentrated H₂SO₄ (115 mL) was added to a mixture of graphite flakes (5.0 g) and NaNO₃ (2.5 g). Then, KMnO₄ (15.0 g) was slowly added, the reaction mixture heated at 50 °C and stirred for 6 h. The mixture was cooled and carefully diluted with 400 mL of water, then cooled in a water bath, and reacted with 20 mL of H₂O₂ 30% v/v. After 30 minutes, the reaction mixture was centrifuged (5000 rpm, 10 min), and the supernatant decanted away. The solid material was then successively washed and centrifuged (5000 rpm for 10 min) with 800 mL of 30% HCl and then with 400 mL of ethanol. The black solid was suspended in 400 mL of ether, filtered with a 0.45 μm PTFE membrane, and the solid dried overnight under vacuum at room temperature. Graphene Oxide (GO) was prepared by dispersing the GrO (100 mg) in 100 mL of alkaline water (pH ≅ 13).

Reduced graphene oxide, rGO, was prepared by refluxing 500 mL of GO (1.0 mg mL⁻¹) dispersion in pH = 13 KOH solution with 2.0 g of NaBH₄ for 3 h. This mixture was concentrated to 75 mL in a rotary evaporator and diluted to 250 mL with glycerol, in order to get a 4 mg mL⁻¹ rGO stock suspension.

The Ni@rGO and NiCo@rGO nanocomposites were prepared by slowly adding respectively the α-Ni(OH)₂ and α-NiCo(OH)₂ nanoparticles sol into a rGO suspension generating a 1 : 2 v/v mixture, and keeping under stirring for 24 h, at room temperature.

Characterization

The samples were characterized by X-ray diffractometry (XRD) in a Bruker D2 Phaser diffractometer equipped with a Cu Kα source (λ = 1.5418 Å, 30 kV, 15 mA, step = 0.05°), in the 2θ range from 5 to 70°. Infrared spectra were recorded in a Bruker ALPHA FTIR spectrophotometer. UV-vis absorption spectra were registered in a Hewlett Packard 8453A diode-array spectrophotometer, in the 190 to 1100 nm range, using 10.0 mm quartz cuvettes. The concentration of α-Ni(OH)₂ and α-NiCo(OH)₂ in the sol was determined by total X-ray fluorescence spectroscopy (TXRF) using a Bruker PICOFOX S2 equipment.

Transmission electron microscopy (TEM) images were obtained in a JEOL JEM-2100 equipment at accelerating voltage of 200 kV. Samples were prepared on copper grids (TedPella), by dispersing 3 μL of suspension diluted in water. Scanning electron microscopy (SEM) images of FTO electrodes modified with nanocomposites were obtained in a JEOL JSM 7401F (FEG) equipment. Scanning Probe Microscopy (SPM) images were obtained in a PicoSPM I equipment, with a PicoScan 2100 controller, using silicon cantilever, under attractive regime with amplitude set point around 6.5 V. The typical tip radius was less than 10 nm. The thickness of the films after heat treatment at 240 °C was measured using a Dektak Stylus Profiler.

Confocal Raman spectra were registered using a WITec 300R Alpha confocal Raman microscope equipped with a 488 nm Ar laser (WITec, 0.37 mW cm⁻²), 20× objective (0.2 NA), 600 lines per mm grating, and integration time of 120 s. Raman images were obtained mapping 50 × 50 μm areas (2500 points) point-by-point using a piezo-driven xyz table, setting the laser power to 4.70 mW cm⁻² and the integration time to 1 s per point. The 499, 522 and 1357 cm⁻¹ bands were respectively selected for Raman imaging of Ni(OH)₂, NiCo(OH)₂ and rGO domains.

Cyclic voltammetry and electrochemical impedance spectroscopy (EIS) measurements were carried out using an EcoChemie Autolab PGSTAT30 potentiostat/galvanostat and a conventional three electrodes cell, constituted by Ag/AgCl (3.0 mol L⁻¹ in KCl) reference, a platinum wire counter and modified FTO working electrodes. Electrochemical impedance spectra were registered from 0.01 to 1 000 000 Hz, modulating the frequency of the sinusoidal potential wave (amplitude = 10 mV) superimposed to a DC potential.

Preparation of electrodes and electrochemical characterization

The modified FTO glass electrodes were prepared defining the area (A = 1 cm²) with a Scotch tape, spin-coating 25 μL of nanomaterial suspension, and heating to 240 °C, for 30 min, in order to produce adherent films. Electrodes modified with α-Ni(OH)₂ nanoparticles (denominated N1, N2 and N3), NiCo (NC1, NC2 and NC3), Ni@rGO (NG1, NG2 and NG3), and NiCo@rGO (NCG1, NCG2 and NCG3) layers of decreasing thickness were prepared by spin-coating respectively at 1500, 2500 and 4000 rpm.

The CVs and EIS data were recorded using a N₂ gas purged 1.0 mol L⁻¹ KOH solution as electrolyte. Galvanostatic charge–discharge curves were registered using larger modified FTO glass electrodes (A = 4 cm²) prepared as above by spin-coating a suitable volume of nanocomposite suspension (containing 4–8 mg) on the surface, and heating to 240 °C.

Results and discussion

Physicochemical characterization

Graphene oxide, constituted by micrometer large platelets was successfully prepared, as confirmed by atomic force microscopy (AFM) images (Fig. S1†) showing large amounts of 1–5 μm large and 0.9 nm thick GO sheets evenly distributed on mica surface. This can be assigned to monolayers of GO laying flat on mica substrate, as expected for a nanocarbon material completely dispersed in aqueous alkaline solution.

The reduction of GO to rGO was monitored spectrophotometrically by following the shift of the GO π → π* transition band from 232 to 261 nm (Fig. 1A), indicating the reduction of keto and hydroxyl groups by borohydride, restoring the C=C bond network of graphene. The absorbance in 1000 nm increased from 0.15 to 0.52 a.u. during the reaction indicating the presence of relatively large particles in the reduced material suspension. Interestingly, more or less stable rGO dispersions in water could be prepared in the concentration range from 0.025 up to 1.0 mg mL⁻¹ (Fig. 1B), thus allowing the reaction with α-Ni(OH)₂ and α-NiCo(OH)₂ nanoparticles in a more controlled way to generate the Ni@rGO and NiCo@rGO nanocomposites.

Fig. 1 (A) UV-vis spectra of 0.05 mg mL⁻¹ suspension of GO and rGO in water. (B) Photo showing a 1.0 and 0.025 mg mL⁻¹ dispersion of rGO in water.

The nanomaterials were characterized by Raman spectroscopy also, but the low scattering cross-section of α-Ni(OH)₂ and α-NiCo(OH)₂ precluded their detection in the nanocomposite materials by that technique. In fact, when the laser power was increased to enhance the signal-to-noise ratio, typically above 3.2 mW cm⁻², a dehydroxilation/dehydration process took place generating the respective oxides, i.e. NiO and Ni_0.6Co_0.4O nanoparticles. This process can be easily confirmed by the rise of an intense Ni–O stretching band at 499 cm⁻¹ concomitantly with the fading of the O–H stretching peak at 3665 cm⁻¹ in the Raman spectra of α-Ni(OH)₂. Similar process was observed for NiCo (Fig. 2B) except that two intense bands assigned to ν(Ni–O) and ν(Co–O) modes rose respectively at 522 and 467 cm⁻¹. This result indicates that a mixed oxide was formed and Co(II) ions were incorporated in the lattice increasing the Ni–O bond strength. Finally, the Raman bands¹¹ at 1075 (νC–O of polyol), 1460 (δCOO⁻), 2872 (νCH₂) and 2930 cm⁻¹ (νCH₃) indicates the presence of intercalated acetate and glycerin in the nanomaterials.

Fig. 2 (A) Raman spectra of (a) α-Ni(OH)₂ (b) Ni@rGO, (c) α-NiCo(OH)₂ and (d) NiCo@rGO; (B) expanded view of α-NiCo(OH)₂ spectrum in the 300–700 cm⁻¹ region; (C) confocal Raman microscopy image of (NiCo)O@rGO; and (D) FTIR spectra of (a) α-Ni(OH)₂ (b) Ni@rGO, (c) α-NiCo(OH)₂ and (d) NiCo@rGO.

Selected areas of the Ni@rGO and NiCo@rGO samples were first scanned setting the laser power to 4.0 mW cm⁻² in order to convert the hydroxide in oxide nanoparticles (Fig. S2A and B,† respectively) and then Raman spectra were acquired in the same area. Typical nickel oxide and nickel–cobalt mixed oxide spectra were obtained. Interestingly, the ν(Ni–O) peak shifted from 499 in pure NiO to 509 cm⁻¹ in NiO@rGO as a consequence of the interaction of the nanoparticles with rGO sheets, whereas remained at 522 cm⁻¹ in (NiCo)O@rGO. A typical confocal Raman microscopy image of (NiCo)O@rGO weighted by the ν(Ni–O) band is shown in Fig. 2C. A significant signal was observed all around but the red-orange areas apparently concentrate more nanoparticles, probably as a consequence of the orientation of the rGO sheets. The standup or tilted sheets expose a thicker layer of nanoparticles at the focal volume of the microscope thus generating stronger signals, as suggested by the generally linear features. As expected, relatively strong rGO D and G bands were found at 1357 and 1595 cm⁻¹.

The infrared spectra of α-Ni(OH)₂ and α-NiCo(OH)₂ exhibited a broad and strong absorption band at 3415 cm⁻¹ characteristic of O–H stretching modes, as well as bands in 2925 and 2864 cm⁻¹ respectively assigned to –CH₂ groups antisymmetric and symmetric C–H stretching modes of intercalated glycerin (Fig. 2D). The coupled methylene and C–OH group deformation bands were found at 1450 cm⁻¹ whereas the antisymmetric ν(Co–O) band was found at 1115 cm⁻¹. The 677 and 444 cm⁻¹ bands were respectively assigned to in plane O–Ni–O angular deformation and Ni–O axial deformation modes. The Ni@rGO and NiCo@rGO nanocomposites exhibited additional bands at 1558 and 993 cm⁻¹ respectively attributed to rGO ν(C=C) and remaining ν(C–O) modes.

The XRD of α-Ni(OH)₂ powder exhibited the characteristic diffraction pattern with a peak at 2θ 9.35° (9.49 Å) and 19.05° (4.72 Å), respectively assigned to the 003 and 006 reflection planes, as well as low intensity peaks attributed to the 101 and 110 planes (Fig. S3a†). It is interesting to note the significant decrease of the degree of crystallinity in NiCo, as confirmed by the broadening of the 003 peak and the disappearance of the 006, 101 and 110 peaks, reflecting the effect of cobalt ions in the structure of α-Ni(OH)₂ (Table 1).

Table 1 X-Ray diffraction 2θ angles corresponding to the 003, 006, 101 and 110 reflections peaks and basal plane distances calculated from the 003 and 006 peaks for α-Ni(OH)₂, α-NiCo(OH)₂, Ni@rGO and NiCo@rGO. The calculated distances from the 101 and 110 reflections were also included

Sample	2θ(003) – d(Å)	2θ(006) – d(Å)	2θ(101) – d(Å)	2θ(110) d(Å)
α-Ni(OH)2	9.25° – 9.59 Å	19.05° – 4.72 Å	34.00° – 2.75 Å	60.25° – 1.77 Å
Ni@rGO	9.25° – 9.59 Å	—	34.55° – 2.72 Å	60.60° – 1.76 Å
α-NiCo(OH)2	9.35° – 9.49 Å	—	—	—
NiCo@rGO	9.35° – 9.49 Å	—	—	—

---

The nanocomposite materials were characterized by transmission electron microscopy (TEM, Fig. 3) revealing the presence of rGO sheets covered by a more or less homogeneously and densely packed monolayer of electrochemically active Ni(OH)₂ (Fig. 3A) and NiCo nanoparticles (Fig. 3B). The lighter area in the middle corresponds to a crack exposing the underneath layer of the nanocomposite material. A closer inspection revealed the presence of a similar NiCo@rGO structure, i.e. a rGO sheet completely covered by NiCo nanoparticles. Such well defined structure is possible only when the material is prepared from completely exfoliated reduced graphene oxide and completely dispersed NiCo nanoparticles with relatively strong interaction with rGO, or when the interaction in between rGO sheets is weak enough to allow the dispersion of the agglomerates as nickel hydroxide nanoparticles bind on the surface. The last assumption is the most probable from Fig. 1 that indicates the presence of relatively large agglomerates in aqueous media.

Fig. 3 TEM images of (A) Ni@rGO and (B) NiCo@rGO with increasing degree of magnification from left to right. The HRTEM of a typical NiO nanoparticle in Ni@rGO, after dehydroxilation/dehydration reaction induced by the electron beam, is shown in the inset (right top).

The nanoparticles are more or less crystalline as verified by HRTEM (inset Fig. 3) and SAED (selected area electron diffractometry) that showed electron diffraction patterns consistent with the face centred cubic (fcc) structure of NiO and (NiCo)O, with peaks corresponding to the 111, 200, 220, 311 and 222 reflection planes (JCPDS card no. 73-1523). The lattice parameters a = b = c = 4.17 Å for unit cell, determined for NiO nanoparticles, are consistent with the literature (Fig. S4A and B†). The conversion of α-Ni(OH)₂ and α-NiCo(OH)₂ in NiO and (NiCo)O, respectively, was induced by the electron beam as it promotes local heating generating about 4 nm diameter nanoparticles of the corresponding oxide material. The high vacuum in the sample chamber also should be favouring that process. Similar dehydroxilation/dehydration process of the hydroxide materials was observed when the laser power of the confocal Raman microscope was increased, as discussed before.

The NiCo@rGO nanocomposite was also characterized by EDS. In fact, no nickel or cobalt (Fig. S5B†) were found in a region with absence of NPs (area 1, Fig. S5A†) but the signals of those elements were clearly seen in area 2 (Fig. S5C†) confirming the presence of NiCo mixed oxide nanoparticles. The amount of C, O and K was also larger in this region as expected for the presence of rGO and probably of potassium sulfate and carboxilated organic species intercalated in the NPs, in the as prepared material. Si is an impurity of graphite (Vonder).

Electrochemical characterization

The electrochemical behavior of α-Ni(OH)₂, Ni@rGO, NiCo and NiCo@rGO were evaluated by cyclic voltammetry (CV), electrochemical impedance spectroscopy (EIS) and galvanostatic charge–discharge (GCD), using FTO electrodes modified with those nanomaterials. They were deposited by spin-coating varying the rotation speed (1500, 2500 and 4000 rpm), in order to evaluate the influence of thickness (Fig. S6 and Table S1†) and rGO in the electrochemical behavior of α-Ni(OH)₂ and α-NiCo(OH)₂. Now on, when not specified otherwise, the behavior of the electrodes modified at 1500 rpm are presented.

The CVs and EIS data of α-Ni(OH)₂ and NiCo are shown in Fig. S7.† The CVs of the first one exhibit a sudden shift and increase of charge of the cathodic wave after the first cycle and a progressive shift of the anodic wave as a function of the number of scans, both preserving the current intensity. The Nyquist and Bode phase plots of α-Ni(OH)₂ after 1, 50, 100, 200 and 300 successive redox cycles in the 0.0 to 0.55 V range, are shown respectively in Fig. S7B and C.† The first activated event in high frequencies was assigned to the Ni(OH)₂/NiOOH process whereas the second one at lower frequencies was associated with an electron transfer process to the electrolyte solution promoting the oxidation of water. This event appears because the applied DC potential was set to 0.6 V, near the onset of that reaction. However, it is evident that the impedance of the Ni²⁺/Ni³⁺ process remained unchanged and only the resistance of the second event decreased as a function of the number of scans, shifting the associated peak in the Bode phase plot to higher frequencies. The simulation of the Nyquist spectra revealed a diffusion controlled region in between those arcs, characterized by a short linear segment with slope of 0.5, that was assigned to electron diffusion in the film.

NiCo showed a broad pair of waves in the first scan that become sharper and increased in intensity as a function of the number of scans. The resistance associated with the NiCo(OH)₂/NiCoOOH process remained constant and of the water oxidation process decreased progressively as in α-Ni(OH)₂ (Fig. S7E†). Interestingly enough, the frequency of this event remained more or less constant at 2–3 Hz but the peak associated with the first one shifted to lower frequencies (3 to 0.8 kHz) in the Bode phase plot (Fig. S7F†), as expected for a capacitance increase associated with the activation of more redox sites. This tend to increase the current intensity in the CVs and decrease the diameter of the arc associated with oxidation of water in the Nyquist plots, since the electrocatalytic activity of the material is proportional to the concentration of active sites and enhanced electrolyte diffusion through the material. Thus, EIS is a convenient technique to monitor in detail the changes in nickel hydroxide materials as a function of the number of redox scans and aging.

The electrodes modified with Ni@rGO showed an anodic wave at 0.4 V with tendency to shift increasing ΔE_p, fading out (Fig. 4A) and increasing the resistance of the Ni²⁺/Ni³⁺ process as a function of the number of successive redox scans (Fig. 4B), clearly indicating a larger tendency of conversion to the beta form than pure α-Ni(OH)₂. This is confirmed by the increase of the phase angle in the Bode phase plots at 10³ to 10⁴ Hz (Fig. 4C), as confirmed by the CVs. This behavior can be related to the partial removal of glycerin, a substance that plays a very important role in the stabilization of nickel hydroxide nanoparticles in the alpha phase, as shown in Fig. S8† where the CVs of as prepared sol suspension and the one washed with water are compared. Note that the current intensity decreases and the peak potential shifts rapidly increasing ΔE_p, clearly indicating the conversion of the alpha phase material to the beta phase.

Fig. 4 Cyclic voltammograms (A and D), Nyquist (B and E) and Bode phase (C and F) plots of EIS data registered for Ni@rGO (A–C) and NiCo@rGO (D–F) along 300 CV cycles, intercalated by EIS measurements at DC potential of 0.6 V superimposing an AC perturbation (ΔE = 10 mV) in the 0.01 to 100 000 Hz range. (G) The equivalent electrical circuit used to fit all EIS data is shown in (G).

In contrast, the waves of NiCo@rGO showed a progressive enhancement (Table S2†) indicating the activation of increasing number of redox sites of the nanomaterial^13,14 (Fig. 4D), but in different chemical environments as suggested by the shift of the peak potential with no significant increase of the charge transfer resistance and rate of the high frequency event in the Nyquist and Bode phase plots (Fig. 4E and F). Furthermore, there is a significant decrease in the resistance for oxidation of water as a function of the density of active redox sites, as confirmed by the decrease of diameter of the charge transfer event in the low frequency region. However, the large increase in capacitance (larger density of redox active sites) seems to be compensated by the high conductivity of rGO and only a very small increase in resistance was observed in contrast with α-NiCo(OH)₂. In fact, the presence of rGO decreased the resistance of Ni²⁺/Ni³⁺ process in 3 ohms generating a high conductivity network in the material. Thus, the electrochemical properties of α-Ni(OH)₂ and α-NiCo(OH)₂ nanoparticles seems to be preserved in the respective rGO nanocomposites. The Nyquist and Bode phase impedance data were simulated using the equivalent circuit shown in Fig. 4G and R_S, R_TC1, R_w, C_μ1, R_TC2 and C_μ2 parameters determined and listed in Table 2.

Table 2 Resistance and capacitance values (R_S, R_TC1, R_w, C_μ1, R_TC2 and C_μ2) determined by using the equivalent circuit shown in Fig. 4G

Electrode	RS (Ω)	RTC1 (Ω)	Rw (Ω)	Cμ1 (μF)	RTC2 (Ω)	Cμ2 (mF)
N1	15.2	8.40	1.55	5.89	6.93	3.26
N2	20.8	5.96	1.59	6.37	7.89	1.85
N3	18.8	5.83	1.39	6.09	7.64	1.17
NG1	12.9	9.21	2.04	7.12	6.48	1.65
NG2	16.1	9.40	1.45	6.10	6.84	1.08
NG3	17.1	10.8	0.943	6.54	6.76	0.976
NC1	21.9	8.22	1.77	27.2	10.2	5.52
NC2	17.5	7.67	2.22	5.60	10.6	2.95
NC3	15.2	6.73	1.69	8.08	10.5	1.29
NCG1	18.6	5.22	2.78	51.1	8.41	5.45
NCG2	18.6	6.09	2.64	7.13	8.68	3.05
NCG3	19.4	11.1	2.95	6.99	10.3	1.64

---

The amount of material deposited on the electrode surface is proportional to the intensity of the anodic and cathodic waves (Table S1†), as expected, and tend to increase the resistance of the Ni²⁺/Ni³⁺ process (N3 = 5.8, N2 = 6.0, N1 = 8.4 Ω and NC3 = 6.7, NC2 = 7.7 NC1 = 8.2 Ω) in α-Ni(OH)₂ and α-NiCo(OH)₂ (Fig. S9†). Interestingly enough, an opposite tendency was observed for the Ni@rGO and NiCo@rGO composites, as confirmed in the Nyquist plots (Fig. 5B–E), where the resistance changed as follows: NG3 = 10.8, NG2 = 9.4, NG1 = 9.2 Ω, and NCG3 = 11.1, NCG2 = 6.1, NCG1 = 5.2 Ω. In order to confirm these results galvanostatic charge–discharge experiments were carried out as described below.

Fig. 5 Cyclic voltammograms at v = 50 mV s⁻¹ (A and D), Nyquist (B and E) and Bode (C and F) plots of EIS data registered for Ni@rGO (A–C) and NiCo@rGO (D–F) as a function of thickness, after 300 CV cycles, in 1.0 mol L⁻¹ KOH solution. The EIS measurements were carried out at DC potential of 0.6 V, superimposing an AC perturbation (ΔE = 10 mV) in the range from 0.01 to 100 000 Hz.

The modified electrodes were submitted to GCD (galvanostatic charge–discharge) processes during 5000 consecutive cycles (Fig. 6A and B). α-Ni(OH)₂ presented the maximum capacitance C_s of 427 F g⁻¹ after about 70 successive charge–discharge cycles, that decreased to 380 F g⁻¹ as a consequence of its conversion to β-Ni(OH)₂. Ni@rGO exhibited a similar profile and C_s reached the maximum of 670 F g⁻¹ after about 180 cycles but unfortunately the phase transformation took place more quickly stabilizing at 50 F g⁻¹, suggesting that the pure α-Ni(OH)₂ probably was not completely converted to β-Ni(OH)₂. The capacitance of α-NiCo(OH)₂ and NiCo@rGO showed similar profile increasing more or less rapidly as a function of scan cycles. The capacitance of NiCo reached up to 1016.4 F g⁻¹ whereas the composite achieved 1298 F g⁻¹ after 200 cycles, a larger value reflecting the effect of rGO in the material. This larger specific capacitance is proportional to the galvanostatic charge–discharge time, as shown in the chronopotentiograms shown in Fig. 6A. The high capacitance of NiCo@rGO can be assigned to synergistic contributions of the electrochemical double layer capacitance of rGO and the pseudocapacitance of α-NiCo(OH)₂, as well as the higher conductivity of the nanocomposite material. To achieve such a high synergic effect, the NiCo nanoparticles should be establishing a good electric contact with rGO sheets and the resultant nanostructures with themselves.

Fig. 6 (A) Chronopotentiograms showing the last ten charge–discharge cycles of a total of 5000 successive cycles; (B) plots of specific capacitance vs. number of discharge cycles and (C) plots of specific capacitance as a function of the current density (5, 15, 25, 35, 45, 55, 65, 75, 85 and 100 A g⁻¹).

Considering the application, for example, in asymmetric supercapacitors (ASCs), the materials should possess high specific capacitance preferentially in a broad range of charge/discharge rates. The plots shown in Fig. 6C indicates that the capacitance decreased significantly for α-Ni(OH)₂ and Ni@rGO as a function of the current density but more drastically for Ni@rGO, from 1167 F g⁻¹ at 5 A g⁻¹ to 445 F g⁻¹ indicating the retention of only 38% of capacitance at 100 A g⁻¹. Unexpectedly, the capacitance increased rapidly as a function of current density, reaching a maximum around 25 A g⁻¹ for α-NiCo(OH)₂ and NiCo@rGO, indicating that they are promising materials for use in pseudo supercapacitors due to rapid charge–discharge process. In fact C_s remained more or less constant from 25 to 100 A g⁻¹ (1309 and 1301 F g⁻¹). This means that those materials can be charged and discharged like a capacitor and has energy density comparable to lithium batteries, as well as high enough stability for most applications, even the most demanding ones. The remarkable rate capability is a consequence of the high conductivity of NiCo mixed hydroxide nanoparticles associated with the high conductivity of rGO and the presence of channels allowing fast ion diffusion. Those three requirements should be addressed simultaneously to allow very fast charge/discharge processes as observed for the NiCo@rGO composite. It is also relevant mentioning that the 3–4 nm diameter NiCo nanoparticles favour fast redox processes, and electron and electrolyte diffusion in the NPs. On the other hand, they are quite strongly bond on rGO surface in aqueous media because of their low solubility, probably increasing the electron transfer to rGO sheets. Finally, the short distances between the rGO sheets defined by the minute electrochemically active NPs, associated with the tendency of formation of islands (up to 3–4 μm high stacks, Fig. S6 and S10†) should be favouring the electron transfer, thus explaining the high conductivity of that nanocomposite material.

At this point, it is important to note that in most of recent works describing high capacitance materials, only C_s values measured at low current densities were reported (Table 3). One exception is the work by Liu et al.¹ describing the electrochemical characterization of nanocomposites based on Ni(OH)₂ and rGO at high current densities. However, in contrast with the materials described in the present work, a 89.4% loss in C_s was observed when the current density was increased from 0.5 to 100 A g⁻¹. In fact, all materials presented in Table 3, except for the ones described in this work, show capacitance retention inferior to 86.3% even considering lower current densities, indicating that the NiCo@rGO nanocomposite is very promising for application in high efficiency supercapacitors.

Table 3 Comparison of the specific capacitances and percentage of capacitance retention as a function of the current density for nickel hydroxide based materials

Sample	Specific capacitance (F g^−1)	% retention (Δ current density A g^−1)	Reference
rGO/Ni(OH)2/PANI	514	86.3% (from 2 to 5)	22
FGN	820	63.9% (from 4 to 11.2)	23
Ni(OH)2 GS	1335	71.4% (from 2.8 to 45.7)	24
Ni(OH)2-RGOCNT	1235	63.1% (from 1 to 20)	26
rGO&Ni(OH)2	1576	63.9% (from 6 to 42)	2
rGO-Ni(OH)2	1717	10.6% (from 0.5 to 100)	1
NiCo-LDH/RGO	1911.1	76.9% (from 2 to 20)	28
Co0.5Ni0.5(OH)2/graphene	1650	86% (from 0.5 to 20)	17
NiCo hydroxide/CNTs	1151.2	61% (from 1 to 70)	3
NiCo@rGO	1348	96.9% (from 35 to 100)	This work
α-NiCo(OH)2	1036	96.2% (from 35 to 100)	This work

---

Conclusions

The mixed hydroxide NiCo(OH)₂ nanoparticles and the corresponding nanocomposite NiCo@rGO, where each rGO sheet is decorated with stabilized nickel–cobalt hydroxide nanoparticles, exhibit high specific capacitance at high current densities, capacitance retention capability superior to 96% up to 100 A g⁻¹, and stability to withstand more than 5000 galvanostatic charge/discharge cycles with no significant changes in the electrochemical properties. These characteristics assure very fast charge/discharge processes like in capacitors combined with the high energy densities of batteries, thus being suitable for application in high performance energy storage devices. The structure and morphology of nanomaterials were revealed by TEM, XRD, FTIR and TXRF; and the electrochemical properties characterized by CV, EIS and galvanostatic charge–discharge measurements for determination of specific capacitance, internal resistance and capacitance retention. NiCo@rGO showed superior performance as compared to stabilized α-Ni(OH)₂ and α-NiCo(OH)₂ nanoparticles and Ni@rGO, the analogous composite of the first one with rGO.

Acknowledgements

The authors are grateful to the Brazilian agency “Conselho Nacional de Desenvolvimento Científico e Tecnológico”, CNPq, for financial support (CNPq 402281/2013-6) and fellowships (JMG CNPq 141853/2015-8, RFG CNPq 141386/2011-8, BBNSB CNPq 153523/2014-0). We are also grateful to Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP 2013/24725-4) and to Petrobras (confocal Raman microscope). We also thanks Prof. Antonio C. Seabra (Laboratório de Sistemas Integrados, POLI-USP) for the profilometry measurements.

References

1. Y. Liu, R. Wang and X. Yan, Sci. Rep., 2015, 5, 11095.
2. J. L. Zhang, H. D. Liu, L. H. Huang, S. Z. Tan, W. J. Mai and X. Cai, J. Solid State Electrochem., 2015, 19, 229–239.
3. M. Li, K. Y. Ma, J. P. Cheng, D. Lv and X. B. Zhang, J. Power Sources, 2015, 286, 438–444.
4. G. Wang, L. Zhang and J. Zhang, Chem. Soc. Rev., 2012, 41, 797–828.
5. S. Yang, Y. Gong, Z. Liu, L. Zhan, D. P. Hashim, L. Ma, R. Vajtai and P. M. Ajayan, Nano Lett., 2013, 13, 1596–1601.
6. J. Ji, L. L. Zhang, H. Ji, Y. Li, X. Zhao, X. Bai, X. Fan, F. Zhang and R. S. Ruoff, ACS Nano, 2013, 7, 6237–6243.
7. S. Min, C. Zhao, G. Chen and X. Qian, Electrochim. Acta, 2014, 115, 155–164.
8. J. Yan, Z. Fan, W. Sun, G. Ning, T. Wei, Q. Zhang, R. Zhang, L. Zhi and F. Wei, Adv. Funct. Mater., 2012, 22, 2632–2641.
9. M. A. Rocha, F. J. Anaissi, H. E. Toma, K. Araki and H. Winnischofer, Mater. Res. Bull., 2009, 44, 970–976.
10. M. A. Rocha, H. Winnischofer, K. Araki, F. J. Anaissi and H. E. Toma, J. Nanosci. Nanotechnol., 2011, 11, 3985–3996.
11. M. Danczuk, C. Nunes Jr, K. Araki and F. Anaissi, J. Solid State Electrochem., 2014, 18, 2279–2287.
12. C. V. Nunes Jr, M. Danczuk, A. A. Bortoti, J. M. Gonçalves, K. Araki and F. J. Anaissi, J. Power Sources, 2015, 297, 408–412.
13. P. R. Martins, A. L. A. Parussulo, S. H. Toma, M. A. Rocha, H. E. Toma and K. Araki, J. Power Sources, 2012, 218, 1–4.
14. P. R. Martins, S. H. Toma, M. Nakamura, H. E. Toma and K. Araki, RSC Adv., 2013, 3, 20261–20266.
15. K. Provazi, M. J. Giz, L. H. Dall'Antonia and S. I. Córdoba de Torresi, J. Power Sources, 2001, 102, 224–232.
16. H.-Y. Hsu, K.-H. Chang, R. R. Salunkhe, C.-T. Hsu and C.-C. Hu, Electrochim. Acta, 2013, 94, 104–112.
17. Y. Cheng, H. Zhang, C. V. Varanasi and J. Liu, Energy Environ. Sci., 2013, 6, 3314–3321.
18. L. Huang, D. Chen, Y. Ding, S. Feng, Z. L. Wang and M. Liu, Nano Lett., 2013, 13, 3135–3139.
19. K. S. Novoselov, V. I. Falko, L. Colombo, P. R. Gellert, M. G. Schwab and K. Kim, Nature, 2012, 490, 192–200.
20. A. K. Geim and K. S. Novoselov, Nat. Mater., 2007, 6, 183–191.
21. C. N. R. Rao, K. Biswas, K. S. Subrahmanyam and A. Govindaraj, J. Mater. Chem., 2009, 19, 2457–2469.
22. D. Ghosh, S. Giri, M. Mandal and C. K. Das, RSC Adv., 2014, 4, 26094–26101.
23. Z. Sun and X. Lu, Ind. Eng. Chem. Res., 2012, 51, 9973–9979.
24. H. Wang, H. S. Casalongue, Y. Liang and H. Dai, J. Am. Chem. Soc., 2010, 132, 7472–7477.
25. H. Yan, J. Bai, B. Wang, L. Yu, L. Zhao, J. Wang, Q. Liu, J. Liu and Z. Li, Electrochim. Acta, 2015, 154, 9–16.
26. L. L. Zhang, Z. Xiong and X. S. Zhao, J. Power Sources, 2013, 222, 326–332.
27. X. Zhu, Y. Zhong, H. Zhai, Z. Yan and D. Li, Electrochim. Acta, 2014, 132, 364–369.
28. X. Cai, X. Shen, L. Ma, Z. Ji, C. Xu and A. Yuan, Chem. Eng. J., 2015, 268, 251–259.
29. Y. Tang, Y. Liu, W. Guo, T. Chen, H. Wang, S. Yu and F. Gao, J. Phys. Chem. C, 2014, 118, 24866–24876.
30. P. R. Martins, M. A. Rocha, L. Angnes, H. E. Toma and K. Araki, Electroanalysis, 2011, 23, 2541–2548.
31. W. S. Hummers and R. E. Offeman, J. Am. Chem. Soc., 1958, 80, 1339.
32. D. C. Marcano, D. V. Kosynkin, J. M. Berlin, A. Sinitskii, Z. Sun, A. Slesarev, L. B. Alemany, W. Lu and J. M. Tour, ACS Nano, 2010, 4, 4806–4814.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra20317e

---