Fabrication of NiCo₂O₄ and carbon nanotube nanocomposite films as a high-performance flexible electrode of supercapacitors

Abstract

Herein, we demonstrated facile fabrication of NiCo₂O₄ nanoparticles on a CNT film by a chemical deposition method, followed by thermal annealing treatment processing. The synthesized NiCo₂O₄/CNT nanocomposite films were still highly flexible, porous, and conductive. Scanning electrical microscopy (SEM) and transmission electronic microscopy (TEM) observation showed that the web-like structure of the films is still preserved and each carbon nanotube in the film was uniformly wrapped by ball-like nanoparticles (ca. 5 nm). X-Ray Diffraction (XRD) measurement confirmed that spinel NiCo₂O₄ was successfully synthesized during the experiment. The contents of NiCo₂O₄ in the composite films could be controlled by tuning the deposition times. The flexible, porous and conductive nanocomposite films could be employed as high-performance flexible energy saving installments. The three-pole measurements illustrated that the composite films possessed a capacitance of 828 F g⁻¹ at 1 A g⁻¹, and retained over 99% of their capacitance after 3000 cycles of charge/discharge at 5 A g⁻¹, showing high-performance electrochemical properties. Finally, an asymmetric supercapacitor installment was assembled by using the composite film as the positive electrode and commercially available activated carbon (AC) as the negative electrode. The measurement gave an energy density of 28.58 W h kg⁻¹ at a power density of 0.7 kW kg⁻¹.

---

1. Introduction

In order to meet the increasing demands for energy and power in our daily life, electrochemical energy storage, such as lithium ion batteries, fuel cells, and supercapacitors (also named as electrochemical capacitors) have been widely studied in the past few decades. Especially, supercapacitors, with reasonably high power and high energy density, have attracted more and more attention due to various unique advantages, including faster charge/discharge processes, longer lifespans, and better safety assurance, and hold great potential as power sources for applications requiring fast bursts of energy or as back-up power sources in electric vehicles.^1,2 Available high-performance electrode materials obtained from simple synthesis methods at a reasonable cost can be the key to drive the practical application of supercapacitors forward.^3–5 Carbon,^6,7 transition metal oxides^8,9 and conductive polymers¹⁰ are the most widely used active electrode materials. However, the carbon-based materials show lower specific capacitance and limited energy density,¹¹ while conductive polymers suffer from poor electrochemical stability. Therefore, metal oxides have been considered as the most promising materials for high-performance pseudocapacitors.¹²

Among the various metal oxides, spinel nickel cobaltite (NiCo₂O₄) has recently been investigated as a high-performance electrode material for energy storage systems because of its environmental benignity, abundance, low cost and great flexibility in structure and morphology.^13–15 Moreover, NiCo₂O₄ has been reported to possess much better electronic conductivity (at least two orders of magnitude higher) and higher electrochemical activity than those of nickel oxides and cobalt oxides, displaying superior electrochemical performance.¹⁶ However, despite the fact that many studies on NiCo₂O₄ have been extended, the high rate performance is still not very satisfactory,^17,18 because the conductivity of NiCo₂O₄ is too low to support fast electron transport toward high rate capability. Furthermore, increased “dead surface” due to the introduction of polymer binder during the conventional thin film electrode preparation seriously limits the electrochemical performances.⁴ To improve the electrochemical performance of NiCo₂O₄ materials at high rates, it is critical to develop electrodes with large amount of electro-active sites, enhance the transport of ions and electrons in electrodes as well as on the electrode–electrolyte interface. Based on the considerations above, great efforts have been taken to grow electroactive nanostructures on conductive substrates to be directly used as integrated electrodes for supercapacitors.^19–22 Carbon nanotubes are generally regarded as promising candidates for high performance electrodes due to their excellent conductivity, high specific surface areas, high flexibility and high thermal and chemical stability.^2,23,24 CNTs have been introduced to support formation of hybrids with functional materials such as CuS,²⁵ Fe₂O₃,¹⁹ Ni(OH)₂,²⁶ NiO and Co₃O₄,²⁷ polyamine,²⁸ etc. However, only few studies focusing on CNT-supported NiCo₂O₄ have been reported.^29–32 Wang²⁹ and Shakir³⁰ introduced polymer binder to synthesize firm electrodes when modified CNT powder and CNT slurry were used. With troublesome pH controlled and surfactant assisted, Cai³¹ synthesized CNT@NiCo₂O₄ hybrid nanostructure with a specific capacitance of 1038 F g⁻¹ at 0.5 A g⁻¹, and suffering from a capacitance loss of 36% at 10 A g⁻¹. Liu³² prepared spinel NiCo₂O₄ on stainless steel substrate-supported vertically aligned CNTs by co-electrodeposition method followed by annealing in air with a specific capacitance of 695 F g⁻¹ at 1 A g⁻¹ and 576 F g⁻¹ at 20 A g⁻¹ respectively.

Thus, a facile method for binder-free electrodes and controllable mass loading of NiCo₂O₄ on CNTs remain challengeable. To the best of our knowledge, there were no reports about NiCo₂O₄ nanoparticles deposition on flexible CNT films. Herein, we developed a template-free method for in situ growth of ultrafine NiCo₂O₄ nanoparticles on flexible CNT film via an improved chemical bath deposition (CBD) at room temperature followed by a post-annealing treatment. In our work, NiCo₂O₄ could easily and uniformly anchor on CNT matrix and the mass loading of NiCo₂O₄ on the CNT network could be tailored by tuning the chemical deposition times. Furthermore, the structure of CNT film was well-reserved during the deposition of NiCo₂O₄. So the hybrid film could be directly applied for binder-free electrode which could provide more active sites. Combined with its porous features, fast ion and electron transfer, and good strain accommodation, the NiCo₂O₄/CNTs composite exhibited excellent rate capability and cycling stability.

2. Experimental

2.1 Materials

The initial CNT films were fabricated by chemical vapor deposition (CVD) method.^33,34 To get functionalized with carboxylic and hydroxyl groups attached on its surface, the prepared CNT films were immersed in mixed acid with a volume ratio of 3 : 1 between H₂SO₄ (98%) and HNO₃ (65%) for 12 h, followed by washed up to neutral pH and drying, denoted as H-CNTs. All chemical reagents were of analytical grade and used without further purification. NiCl₂·6H₂O, CoCl₂·6H₂O, NH₃·H₂O (25–28%) and ethanol were purchased from Sinopharm chemical reagent Co. Ltd.

2.2 Synthesis

Firstly, H-CNT films were sandwiched between two pieces of glass to prevent the films from curling or excessive precipitates stacking on the surface of CNT films. Then NiCo₂O₄ nanoparticles were chemically deposited onto the CNT films though “layer-by-layer assembly” chemical deposition chemical bath deposition.³⁵ In a typical experiment, 0.1 mL ethanolic solution containing 0.2 M metal ion (Co²⁺ : Ni²⁺ = 2 : 1) was impregnated into the network of CNT films, and then the film was dried at 40 °C first. When the ethanol was volatile completely, 0.1 mL diluted ammonia ethanolic solution (containing 0.02 mL NH₃·H₂O) was added into the films. After reaction, deposited films were washed up with ethanol, and then dried. Variable mass loading of NiCo₂O₄ could be achieved by repeating the steps in the same way. At last, the samples were annealed at 300 °C in air for 2 h with a ramping rate of 1 °C min⁻¹ to transform the precursor into spinel NiCo₂O₄ nanoparticles anchored on CNT film. The mass loading of NiCo₂O₄ nanoparticles on CNTs were calculated by carefully weighting CNT films before deposition and after thermal annealing.

The prepared composite films were denoted as NiCo₂O₄/CNTs-x, where x represented the mass fraction of NiCo₂O₄ (vs. the whole composite film). The pure Ni–Co precipitation was synthesized without CNT film in the same way.

2.3 Materials characterization

The morphologies of the as-synthesized samples were characterized by scanning electron microscope (SEM, Hitachi, S4800) and transmission electron microscope (TEM, Tecnai G2F20). The crystal structure was investigated by X-ray powder diffraction (XRD, Rigaku D/Max 2500 v/pc). The chemical structure of the samples was examined by using the Fourier transform infrared spectrophotometer (FT-IR, Bruke Tensor 27). The Brunauer–Emmett–Teller (BET) specific surface area was obtained from N₂ adsorption–desorption isotherm record at 77 K (Quantachrome, Nova2000), and the pore size distribution was evaluated by using the Barrett–Joyner–Halenda (BJH) model.

2.4 Electrochemical measurements

The electrochemical performance of the as-synthesized composite was conducted with a CHI660D electrochemical workstation in an aqueous KOH solution (2 M) with a three-electrode cell. Working electrode was fabricated by pressing the composite film (which was cut into 1 cm × 1 cm multi-layered pieces) upon a piece of nickel foam under 10 MPa for 3 minutes (seen in Fig. S1†). The reference electrode was a standard calomel electrode (SCE) and a Pt foil served as the counter electrode.

3. Results and discussion

3.1 Structure characterization

The schematic illustration of NiCo₂O₄/CNTs fabrication is shown in Fig. 1 (on the topside of the figure). Fig. 1a shows typical SEM image of Ni–Co hydroxide/CNTs composites which possess an interconnected 3D porous networks with pore size ranging from submicrometer to micrometer. Uniform and continuous cladding of Ni–Co hydroxide along the long and straight CNT nanotube can be clearly seen. Interestingly, the NiCo₂O₄/CNTs product keeps the original morphology after sintering in air as shown in Fig. 1b, indicating the robustness of the structure. And the size of the precipitation anchored on CNTs is hardly distinguished by SEM, which indicates ultrafine nanoparticles are synthesized during the experiment. Importantly, naked part is seen as well along the axis direction of the CNT nanotube, suggesting the stress increased during chemical reaction can be alleviated. Meanwhile, the pure NiCo₂O₄ presents rhombic-like shape with a particle size of 50 nm (see in Fig. S3†), indicating that carbon nanotubes can effectively inhibit the aggregation and growth of NiCo₂O₄ nanoparticles.

Fig. 1 Schematic illustration of Ni–Co hydroxide deposition on CNT films in process (1) (2) (3) (on the topside). SEM image (a) and HRTEM image (c) of the corresponding Ni–Co precursor; SEM image (b) and HRTEM image (d) and the SAED pattern (e) and XRD patterns (f) of flexible NiCo₂O₄/CNTs-21% composite film.

To better illustrate the differences between the as-prepared Ni–Co hydroxide/CNTs and the NiCo₂O₄/CNTs composite, representative TEM images are shown in Fig. 1c and d. From Fig. 1c, continuous claddings can be seen outside carbon nanotubes. And the thickness of the outer symmetric Ni–Co hydroxide shell layer is about 12 nm. However, the surface of NiCo₂O₄/CNTs composite in Fig. 1d becomes coarse and porous (marked with red circles) due to release of gas during the annealing process; these small pores are advantageous for the ion in and out during the charge and discharge process.³¹ Moreover, it is evidently observed in Fig. 1d that the highly conductive carbon nanotube is tightly bonded and covered by NiCo₂O₄ nanocrystals with a size around 5 nm in a ball-like shape, forming a core/shell hetero-structured architecture. Thickness of the outer NiCo₂O₄ shell layer is about 12 nm as well, in agreement with the observation in Fig. 1c. The selected-area electron diffraction (SAED) pattern in Fig. 1e reveals well-defined rings, which are indexed to the (111), (220), (311), and (400) planes of NiCo₂O₄ phase (JCPDS no. 20-0781). Two symmetric diffractions are clearly seen in the SAED patter giving a calculated d-spacing of 0.34 nm close to the graphite layers spacing of 0.338 nm, corresponding to the graphitic walls of CNTs.

To further determine crystalline phase of the NiCo₂O₄/CNTs composite, XRD measurement is employed. As shown in Fig. 1e, sharp and narrow diffraction peaks are observed in pure NiCo₂O₄, suggesting fine crystallization. And there are no diffraction peaks from other impurities, indicative of high purity of the synthesized NiCo₂O₄ in the work. Diffraction peaks observed at 19.1°, 31.1°, 36.7°, 44.6°, 59.1° and 65.0° in NiCo₂O₄/CNTs correspond to the (111), (220), (311), (400), (511) and (440) diffraction planes of spinel NiCo₂O₄ (JCPDS no. 20-0781), respectively. The (002) and (101) peaks of CNT can be also observed in the patterns of NiCo₂O₄/CNTs composite. But the (002) reflection of CNTs is weakened.³⁰ Moreover, a tiny but definitive shift in the position of the (002) CNT peaks in NiCo₂O₄/CNTs composite is seen, possibly suggesting strain (and particle size) effects originating the integration between the NiCo₂O₄ nanoparticles and the CNT surface. It also suggests that our process leads to a real integration across the interface rather than the formation of a mere mixture.^27,36 Additionally, the relative intensity of the corresponding diffraction peaks for NiCo₂O₄ significantly decreases compared to those of pure NiCo₂O₄, demonstrating a decrease in grain size of NiCo₂O₄ particles anchored on CNTs, in accordance with the SEM and TEM results.

Fig. 2a shows the N₂ adsorption–desorption isotherms of H-CNTs and NiCo₂O₄/CNTs composite film. Both of the two samples exhibit typical IV isotherms with H1 hysteresis loops, indicating the presence of mesopores. The resulting specific surface area of NiCo₂O₄/CNTs is calculated to be 158.5 m² g⁻¹, which is much higher than that of bare CNT network (95.7 m² g⁻¹). Based on the Barrett–Joyner–Halenda (BJH) method, the pore diameter of NiCo₂O₄/CNTs composite film is mainly distributed in the range of 3–9 nm, revealing striking similarities with bare CNTs (Fig. 2b). But the pore volume of NiCo₂O₄/CNTs composite is much higher than that of bare CNT films. This is probably attributed to the ultrafine NiCo₂O₄ coatings outside carbon nanotubes that barely change the pore structure of CNT's network. Additionally, according to TEM results, it is found out that the surface of the ultrafine coating is coarse and porous and the new pores are within 10 nm. As a result, the surface area and pore volume of NiCo₂O₄/CNTs composite is higher than that of bare CNT films. The larger surface area of the electrode materials enlarges the contact between the electrode and the electrolyte, and thus benefits the utilizing efficiency of electrochemical actives materials. Therefore, NiCo₂O₄/CNTs hybrid film is promising for the application of supercapacitors.

Fig. 2 Adsorption and desorption isotherm curves (a) and the corresponding pore-size distribution (b) of flexible NiCo₂O₄/CNTs-21% composite film.

Controllable mass loading of NiCo₂O₄ on CNTs can be achieved by tuning the chemical deposition times. When the mass fraction of NiCo₂O₄ (vs. the whole composite film) is 11% (deposition only once), only slight precipitation embedded on carbon nanotubes is seen, as shown in Fig. 3a. With the deposition times increases, more excessive nanoparticles and more compacting structures can be obtained in NiCo₂O₄/CNTs-34% (Fig. 3b). Fig. 3c shows TEM image of NiCo₂O₄/CNTs-34% in which the nanotube is surrounded by more homogeneous nanoparticle of NiCo₂O₄ and small pores between them; and the thickness of the outer shell layer doubles in NiCo₂O₄/CNTs-34% compared with NiCo₂O₄/CNTs-21%, in accord with the result of SEM results. However, when repeated for five times or more, excessive deposition occurs on the surface of CNT film; the outline of the carbon nanotube can't be seen anymore, as shown in ESI (Fig. S4†). Excessive deposition leads to aggregation of NiCo₂O₄ nanoparticles which will inhibit the permeation of electrolyte into the composite structure. So, a rational mass loading design on CNT film is essential.

Fig. 3 SEM images of NiCo₂O₄/CNTs-11% (a) and NiCo₂O₄/CNTs-34% (b) and HRTEM image of NiCo₂O₄/CNTs-34%.

On the other hand, it is worth noting that a drying process before adding ammonia is essential to get uniform precipitation on CNTs. As shown in Fig. S5,† disordered precipitation resulted from self-nucleation between CNTs can be seen when there is no drying process before ammonia added, even though the mass fraction of NiCo₂O₄ is only 10%. Because there is Ni–Co solution residual between the porous networks (a reaction container) tending to self-nucleate. However, as the ethanol volatilizes, there is power driving Ni–Co chlorate moving outward till they are attached to nanotubes (solute migration). When ammonia impregnated, the oxygen functional groups on CNTs act as nuclei to trigger the precipitation of Ni–Co precursor (see in Fig. 1). In other words, all Ni–Co chlorate heterogeneous nucleates on CNT surface with drying process plugged in. When repeated, the previous deposition functions as new heterogeneous nucleation and triggers more nanoparticles anchoring along CNT nanotubes. As a result, no precipitation between the networks can be seen in Fig. 1b and 2. In the experiment, owning to the affinity for CNT film and its volatilization, the ethanolic solution could thoroughly permeate the CNT film and the film would dry up quickly. Furthermore, Ni–Co co-precipitation will not dissolve in ethanolic solution even though ammonia is excessive. Therefore, no troublesome pH controlled process is needed in our work, demonstrating a facile and low-cost method.

3.2 Electrochemical performance

Fig. 4a–c show digital photographs of free standing NiCo₂O₄/CNTs-21% nanocomposite films. The consistent black film indicates that the film structure is well preserved during the deposition of NiCo₂O₄. Moreover, the hybrid film can always return to its origin flat shape after deformations, showing its feature of high flexibility. To investigate the flexibility of the composite electrode (Fig. 4d), the device is tested at a scan rate of 20 mV s⁻¹ with various bending angles from 0–150° (Fig. 4e and f). Interestingly, the size and the shape of the CV loops almost remained the same at all angles (Fig. 4e), indicating that the as-prepared electrode is able to withstand stress at different curvatures with no drastic changes in its electrochemical performance. Detailed measurement and comparison of the electrochemical performance under normal bending condition are presented in Fig. 5 to further evaluate the electrochemical properties of self-supported NiCo₂O₄/CNTs composites electrodes. Firstly, the specific capacitance of NiCo₂O₄/CNTs-21% is determined by CV and CD measurements (Fig. 5a and b).

Fig. 4 Digital images (a), (b) and (c) of flexible NiCo₂O₄/CNTs-21% composite film: it can always return to original flat shape after different deformations, such as twisting and folding, indicating its feature of high flexibility. Fabrication of the wording electrode: the composite film is cut into 1 cm × 1 cm pieces and directly pressed onto Ni foam (d). Cyclic voltammetry curves for NiCo₂O₄/CNT-21% composite electrode at 20 mV s⁻¹ with various bending angles from 0–150° (e). Digital images of the electrode bent at different angles: 30°, 60°, 90°, 120°, 150°.

Fig. 5 CV curves at various scan rates (a) and galvanostatic charge–discharge at different current densities (b) curves of NiCo₂O₄/CNTs-21%; comparison of CV curves of different electrodes at a scan rate of 20 mV s⁻¹ (c) and the specific capacitance change of different electrodes as a function of current density (d).

The CV is measured in a potential window of 0–0.45 V at scanning rates ranging from 5 to 70 mV s⁻¹. The shape of the CV curves clearly reveals the pseudocapacitive characteristics derived from faradic reactions. Two pairs of redox peaks can be obviously observed in Fig. 5a, corresponding to the reversible reactions of Co²⁺/Co³⁺ and Ni²⁺/Ni³⁺ which do not occur at the same potentials.²⁰ The redox reactions in the alkaline electrolyte are based on the following equations:

NiCo₂O₄ + OH⁻ + H₂O ↔ NiOOH + 2CoOOH + e⁻ (1)

CoOOH + OH⁻ ↔ CoO₂ + H₂O + e⁻ (2)

With the 20-fold increase in the sweep rate from 5 to 100 mV s⁻¹, the position of the cathodic peaks shifts slightly from 0.22 to 0.16 V, indicating a relatively low resistance of the electrode and good electrochemical reversibility.

CD measurement is considered to be a more accurate technique for pseudo-capacitances.³⁷ Therefore, to calculate the specific capacitance as well as understand the rate capability of NiCo₂O₄/CNTs-21%, the charge/discharge measurements are performed in Fig. 5b. The nonlinear discharge curves and voltage plateaus further verifies the pseudocapacitance behavior of the electrodes. According to the formula: C = IΔt/mΔV, where I is the discharge current, Δt is the discharge time, m is the mass of the active material (i.e., the total mass of the hybrid nanostructure), and ΔV is the potential drop during discharge. The specific capacitance of NiCo₂O₄/CNTs-21% is calculated to be 828, 801, 715, 698 and 656 F g⁻¹ at a current density of 1, 2, 5, 10, 20 A g⁻¹. It is imperative to note that the capacity retention rate (compared to 1 A g⁻¹) is 79% at a current density of 20 A g⁻¹ which is much higher than the reported NiCo₂O₄ grown on carbon nanofibers (only 48%)²² or on carbon nanotubes.^29,31 The excellent capacity retention performance of NiCo₂O₄/CNTs-21% is attributed to its ultrafine and porous features which enable faster permeation process of the KOH electrolyte by significantly reducing the diffusion time of OH⁻ ions. Meanwhile, the naked part helps to accommodate the strain arising from the high rate of insertion and extraction of OH⁻ ions. Furthermore, the highly conductive and interconnected scaffolds significantly improve the kinetics and electrochemical utilization of NiCo₂O₄ nanoparticles due to super highway for shortened electro- and iron-transport path-ways. Thus, it can be concluded that the rate capability will drop if thicker stacking wrapped outside of CNTs. NiCo₂O₄/CNTs-34% confirms the conclusion with a 71% capacity retention rate at a current density of 20 A g⁻¹ when compared to 1 A g⁻¹ (Fig. 5d). This is attributed to a compact structure outside CNTs which hinders the electrolyte ion diffusion. As a result, the electrode of NiCo₂O₄/CNTs-34% is easily polarized at large current densities and then the specific capacitance is reduced. Fig. 5c shows the CV curves for four different electrodes at a scan rate of 20 mV s⁻¹. As the specific capacitance is proportional to the area of the CV curve, the order of the specific capacitance is NiCo₂O₄/CNTs-21% > NiCo₂O₄/CNTs-34% > NiCo₂O₄/CNTs-11% > H-CNT film. When the mass loading of NiCo₂O₄ is up to be 50%, the specific capacitance drops to 667 F g⁻¹ at 1 A g⁻¹ and 561 F g⁻¹ at 5 A g⁻¹ (seen in Fig. S6a†).

Since the long-term electrochemical stability is an important requirement for supercapacitor electrodes, NiCo₂O₄/CNTs-21% and NiCo₂O₄/CNTs-34% electrodes are further tested at a current density of 5 A g⁻¹ for 3000 cycles respectively. As shown in Fig. 6, it is noted that the specific capacitance of NiCo₂O₄/CNTs-21% and NiCo₂O₄/CNTs-34% electrodes increase with cycles at first. This phenomenon is caused by the active materials' activation process, ascribed as the cycling-induced improvement in the surface wetting of the electrodes, leading to more electroactive areas.^17,38 The specific capacitance of NiCo₂O₄/CNTs-34% reaches the maximum value of 767 F g⁻¹ after 800 cycles as a result of more NiCo₂O₄ loadings on CNTs. But then the specific capacitance drops to a lower stage and reaches 621 F g⁻¹ after 3000 cycles (remaining 81% of its activated maximum capacitance and 101% of its initial capacitance). The specific capacitance of NiCo₂O₄/CNTs-50% drops to 560 F g⁻¹ after 1500 cycles as shown in Fig. S6b.† For NiCo₂O₄/CNTs-21%, the specific capacitance reaches a maximum value of 714 F g⁻¹ after 600 cycles. Significantly, more than 99% of the specific capacitance is maintained (comparing to its activated maximum capacitance and 116% to its initial capacitance) after 3000 cycles even at a high current density of 5 A g⁻¹. The capacitance retention is higher than those in previous reported works, concluding a highly competitive excellent long cycle life for NiCo₂O₄/H-CNTs composite.^17,29,32 Furthermore, such NiCo₂O₄/CNTs-21% composite electrode has much better rate performance and cycling life compared with other reported NiCo₂O₄-base materials, such as NiCo₂O₄–reduced graphene oxide composite, NiCo₂O₄–graphene composite, NiCo₂O₄–carbon nanofiber composite and so on (the corresponding results are listed in Table 1).^{20,22,39–47} The insert in Fig. 6 gives a sketch map of the composite's structure to give a better understanding of charge transfer mechanism. The ultrafine NiCo₂O₄ nanoparticles coating forming a porous structure outside the carbon nanotubes is in favour of electrolyte permeating into the whole structure. This will guarantee the reaction fast and successful completion in eqn (1) and (2). On the other hand, the carbon nanotubes provide highway for electron transportation which makes for increased charge rate for high rate performance and nearly no specific capacitance decrease after 3000 cycles. As a result, NiCo₂O₄/CNTs-21% electrode demonstrates excellent rate performance and long cycle life. However, thicker and more compact NiCo₂O₄ coating outside the carbon nanotube hinders the electrolyte from permeating into the structure. So, it needs more cycles for NiCo₂O₄/CNTs-34% to reach its maximum specific capacitance value. Additionally, thicker and more compact coating also hinders electron transport path-ways which is against for good rate performance and long cycle life.

Fig. 6 Comparison of the cycling stability of NiCo₂O₄/CNTs-21% and NiCo₂O₄/CNTs-34% electrodes at a current density of 5 A g⁻¹ (insert gives a sketch map of the composite's structure).

Table 1 Comparison of the electrochemical performances of the as-prepared NiCo₂O₄/CNTs-21% composite with the reported ones

Materials	Preparation method	Specific capacitance (F g^−1)	Rate performance	Capacity retention	Reference
NiCo2O4–G	Solution method	618 (5 mV s^−1)	53.6% (200 mV s^−1)	85% (10 000 cycles)	20
NiCo2O4–CNF	Solution method	1024 (1 A g^−1)	48% (20 A g^−1)	96.4% (2000 cycles)	22
NiCo2O4–(Ni–Co) (OH)2	Solution method	1132 (2 mA cm^−2)	61.8% (50 mA cm^−2)	90% (2000 cycles)	39
NiCo2O4@NiCo2O4	Hydrothermal and chemical deposition	900 (1 A g^−1)	75% (20 A g^−1)	98.6% (4000 cycles)	40
NiCo2O4–RGO	Self-assembly	835 (1 A g^−1)	74% (16 A g^−1)	86% (4000 cycles)	41
NiCo2O4–RGO	Hydrothermal	882 (1 A g^−1)	69% (10 A g^−1)	93.8% (3000 cycles)	42
NiCo2O4/RGO	Solution method	1186.3 (0.5 A g^−1)	52.6% (10 A g^−1)	97% (100 cycles)	43
NiCo2O4/RGO	Hydrothermal	947.4 (0.5 A g^−1)	76.6% (10 A g^−1)	97.9% (3000 cycles)	44
RGO/NiCo2O4	Self-assembly	1693 (1 A g^−1)	67.6% (16 A g^−1)	90.8% (2000 cycles)	45
NiCo2O4@NiO	Hydrothermal	2220 (1 A g^−1)	74% (20 A g^−1)	93.1% (3000 cycles)	46
NiCo2O4@RGO	Hydrothermal	737 (1 A g^−1)	50% (10 A g^−1)	94% (2000 cycles)	47
NiCo2O4/CNT	CBD method	828 (1 A g^−1)	79% (20 A g^−1)	99% (3000 cycles)	This work

---

According to the results discussed above, it is known that a rational mass loading design on CNT film is essential and NiCo₂O₄/CNTs-21% behaves as the best electrochemical active material. Moreover, Fig. 7 shows the morphologies of NiCo₂O₄/CNTs-21% after the cyclic test. Well-maintained hierarchical structure is seen, indicating little volume change from charge–discharge redox reaction of NiCo₂O₄/CNTs composite, which further demonstrates the structural stability of the porous composite as an electrode for supercapacitors. To further evaluate the NiCo₂O₄/CNTs-21% composite film for real device application, we fabricated asymmetric capacitors in two-electrode test system by using NiCo₂O₄/CNTs-21% and commercially available activated carbon as positive and negative electrodes (NiCo₂O₄/CNTs//AC), respectively, with mass ratio of 1 : 2. Fig. 8a shows the CV data for the asymmetric capacitors at various scan rates from 2 to 20 mV s⁻¹. Unlike the three-electrode electrochemical feature, NiCo₂O₄/CNTs//AC exhibits reasonable capacitive behavior with a quasi-rectangle CV curve. The nearly symmetric galvanostatic charge and discharge curves with very small voltage drops at different current density are indicative of good supercapacitor behaviors (Fig. 8b). The cell capacitance of NiCo₂O₄/CNTs//AC based on the total masses of the two electrodes is calculated to be 105, 100, 89, 72 and 55 F g⁻¹ at a current density of 1, 2, 5, 10 and 20 A g⁻¹ respectively, manifesting higher cell capacitance and better rate performance than the reported one.³¹ To facilitate the performance assessment of our asymmetric capacitors, Ragone plots relating the corresponding power and energy density are calculated from the galvanostatic discharge curves (Fig. 8c). Working in a potential range of 1.4 V, NiCo₂O₄/CNTs//AC exhibits a high energy density of 28.58 W h kg⁻¹ at a power density of 0.7 kW kg⁻¹, and the energy density maintains 14.18 W h kg⁻¹ at a power density of 14.18 kW kg⁻¹. Furthermore, NiCo₂O₄/CNTs//AC exhibits excellent cycling stability as well. After 3000 charge–discharge cycles, the specific capacitance maintains almost 100% of its maximum value as shown in Fig. 8d.

Fig. 7 SEM images of NiCo₂O₄/CNTs-21% (a) and magnified SEM images of NiCo₂O₄/CNTs-21% after cyclic performance tests.

Fig. 8 CV curves at different scan rates (a) and galvanostatic charge and discharge at different current densities (b) and Ragone plot (c) and cycling stability (d) of NiCo₂O₄/CNTs//AC.

The excellent electrochemical performance of the NiCo₂O₄/CNTs-21% electrode benefits from the following facts. First, a rational mass loading of NiCo₂O₄ nanoparticles are successfully synthesized to provide a minimizing diffusion distance. Second, a porous structure with large surface area which is essential to the facile diffusion of electrolyte into the inner part of the electrode leads to the maximum utilization of electrochemical active materials. Third, ultrafine and uniform NiCo₂O₄ nanoparticles directly contact with the conductive CNT film forming an integrated binder-free electrode. Forth, the highly conductive and interconnected scaffolds not only provide super highway for shortened electro- and iron-transport path-ways but also accommodates the volume change arising from the high rate of insertion and extraction of OH-ions, resulting in increased charge rate for high rate performance and stability. Because of these intriguing advantages, NiCo₂O₄/CNTs-21% exhibits a dramatically enhanced pseudocapacitive performance.

4. Conclusion

In summary, ultrafine NiCo₂O₄ nanoparticles with tailored mass ratio on CNT film were successfully synthesized by an improved facile CBD method coupled with post-annealing treatment. Orderly growth of NiCo₂O₄ on intact conductive network had a large specific surface area, exhibiting high capacitance and excellent rate capability. And it is imperative to notice that the specific capacitance of the optimized hybrid structure maintains more than 99% of its activated maximum capacitance (116% to its initial capacitance) after 3000 cycles at a current density of 5 A g⁻¹. This enhanced performance is attributed to several important structural factors of the NiCo₂O₄/CNTs hybrid structure, including ultrafine nanoparticles, carbon nanotube support and porous hierarchical structure. Moreover, the composite film is of high flexibility. In view those advantages, this hybrid nanostructure might hold great promise as electrode materials of advanced devices for energy storage.

Acknowledgements

This work was supported by funding from the National Natural Science Fund Program of China (Grant No. 51072130).

Notes and references

1. J. R. Miller and P. Simon, Science, 2008, 321, 651–652.
2. P. Simon and Y. Gogotsi, Nat. Mater., 2008, 7, 845–854.
3. C. O. Ania, V. Khomenko, E. Raymundo-Pinero, J. B. Parra and F. Beguin, Adv. Funct. Mater., 2007, 17, 1828–1836.
4. G. Zhang and X. W. Lou, Adv. Mater., 2013, 25, 976–979.
5. R. J. Zou, K. B. Xu, T. Wang, G. J. He, Q. Liu, X. J. Liu, Z. Y. Zhang and J. Q. Hu, J. Mater. Chem. A, 2013, 1, 8560–8566.
6. Y. P. Zhai, Y. Q. Dou, D. Y. Zhao, P. F. Fulvio, R. T. Mayes and S. Dai, Adv. Mater., 2011, 23, 4828–4850.
7. G. H. Xu, C. Zheng, Q. Zhang, J. Q. Huang, M. Q. Zhao, J. Q. Nie, X. H. Wang and F. Wei, Nano Res., 2011, 4, 870–881.
8. J. Jiang, Y. Y. Li, J. P. Liu, X. T. Huang, C. Z. Yuan and X. W. Lou, Adv. Mater., 2012, 24, 5166–5180.
9. C. Y. Lee and A. M. Bond, Langmuir, 2010, 26, 16155–16162.
10. H. L. Cao, X. F. Zhou, Y. M. Zhang, L. Chen and Z. P. Liu, J. Power Sources, 2013, 243, 715–720.
11. G. P. Wang, L. Zhang and J. J. Zhang, Chem. Soc. Rev., 2012, 41, 797–828.
12. M. C. Liu, L. B. Kong, C. Lu, X. M. Li, Y. C. Luo and L. Kang, ACS Appl. Mater. Interfaces, 2012, 4, 4631–4636.
13. J. Zhu, Z. Xu and B. Lu, Nano Energy, 2014, 7, 114–123.
14. Z. B. Wu, Y. R. Zhu and X. B. Ji, J. Mater. Chem. A, 2014, 2, 14759–14772.
15. L. X. Zhang, S. L. Zhang, K. J. Zhang, G. J. Xu, X. He, S. M. Dong, Z. H. Liu, C. S. Huang, L. Gu and G. L. Cui, Chem. Commun., 2013, 49, 3540–3542.
16. T. Y. Wei, C. H. Chen, H. C. Chien, S. Y. Lu and C. C. Hu, Adv. Mater., 2010, 22, 347–351.
17. Y. R. Zhu, X. B. Ji, Z. P. Wu, W. X. Song, H. S. Hou, Z. B. Wu, X. He, Q. Y. Chen and C. E. Banks, J. Power Sources, 2014, 267, 888–900.
18. Y. F. Zhang, M. Z. Ma, J. Yang, H. Q. Su, W. Huang and X. C. Dong, Nanoscale, 2014, 6, 4303–4308.
19. J. Liu, M. Chen, L. Zhang, J. Jiang, J. Yan, Y. Huang, J. Lin, H. J. Fan and Z. X. Shen, Nano Lett., 2014, 14, 7180–7187.
20. H. L. Wang, C. M. B. Holt, Z. Li, X. H. Tan, B. S. Amirkhiz, Z. W. Xu, B. C. Olsen, T. Stephenson and D. Mitlin, Nano Res., 2012, 5, 605–617.
21. Y.-M. Wang, X. Zhang, C.-Y. Guo, Y.-Q. Zhao, C.-L. Xu and H.-L. Li, J. Mater. Chem. A, 2013, 1, 13290–13300.
22. G. Zhang and X. W. Lou, Sci. Rep., 2013, 3, 1–5.
23. W. D. Zhang, B. Xu and L. C. Jiang, J. Mater. Chem., 2010, 20, 6383–6391.
24. M. T. Byrne and Y. K. Gun'ko, Adv. Mater., 2010, 22, 1672–1688.
25. T. Zhu, B. Xia, L. Zhou and X. W. Lou, J. Mater. Chem., 2012, 22, 7851–7855.
26. H. Chen, F. Cai, Y. Kang, S. Zeng, M. Chen and Q. Li, ACS Appl. Mater. Interfaces, 2014, 6, 19630–19637.
27. R. F. Zhou, C. Z. Meng, F. Zhu, Q. Q. Li, C. H. Liu, S. S. Fan and K. L. Jiang, Nanotechnology, 2010, 21, 1–7.
28. C. Peng, S. Zhang, D. Jewell and G. Z. Chen, Prog. Nat. Sci., 2008, 18, 777–788.
29. X. Wang, X. D. Han, M. Lim, N. Singh, C. L. Gan, M. Jan and P. S. Lee, J. Phys. Chem. C, 2012, 116, 12448–12454.
30. I. Shakir, Electrochim. Acta, 2014, 132, 490–495.
31. F. Cai, Y. R. Kang, H. Y. Chen, M. H. Chen and Q. W. Li, J. Mater. Chem. A, 2014, 2, 11509–11515.
32. W.-W. Liu, C. Lu, K. Liang and B. K. Tay, J. Mater. Chem. A, 2014, 2, 5100–5107.
33. H. Gao, F. Hou, X. Zheng, J. Liu, A. Guo, D. Yang and Y. Gong, Vacuum, 2015, 112, 1–4.
34. H. Gao, F. Hou, Z. Wa, S. Zhao, D. Yang, J. Liu, A. Guo and Y. Gong, Electrochim. Acta, 2015, 154, 321–328.
35. J. W. Lang, L. B. Kong, M. Liu, Y. C. Luo and L. Kang, J. Electrochem. Soc., 2010, 157, A1341–A1346.
36. S. Muduli, W. Lee, V. Dhas, S. Mujawar, M. Dubey, K. Vijayamohanan, S.-H. Han and S. Ogale, ACS Appl. Mater. Interfaces, 2009, 1, 2030–2035.
37. L.-F. Chen, X.-D. Zhang, H.-W. Liang, M. Kong, Q.-F. Guan, P. Chen, Z.-Y. Wu and S.-H. Yu, ACS Nano, 2012, 6, 7092–7102.
38. C. Wang, X. Zhang, D. Zhang, C. Yao and Y. Ma, Electrochim. Acta, 2012, 63, 220–227.
39. W. W. Zhou, D. Z. Kong, X. T. Jia, C. Y. Ding, C. W. Cheng and G. W. Wen, J. Mater. Chem. A, 2014, 2, 6310–6315.
40. X. Y. Liu, S. J. Shi, Q. Q. Xiong, L. Li, Y. J. Zhang, H. Tang, C. D. Gu, X. L. Wang and J. P. Tu, ACS Appl. Mater. Interfaces, 2013, 5, 8790–8795.
41. H. W. Wang, Z. A. Hu, Y. Q. Chang, Y. L. Chen, H. Y. Wu, Z. Y. Zhang and Y. Y. Yang, J. Mater. Chem., 2011, 21, 10504–10511.
42. J. P. Wang, S. L. Wang, Z. C. Huang, Y. M. Yu and J. L. Liu, Ceram. Int., 2014, 40, 12751–12758.
43. L. B. Ma, X. P. Shen, Z. Y. Ji, X. Q. Cai, G. X. Zhu and K. M. Chen, J. Colloid Interface Sci., 2015, 440, 211–218.
44. L. B. Ma, X. P. Shen, H. Zhou, Z. Y. Ji, K. M. Chen and G. X. Zhu, Chem. Eng. J., 2015, 262, 980–988.
45. L. Wang, X. H. Wang, X. P. Xiao, F. G. Xu, Y. J. Sun and Z. Li, Electrochim. Acta, 2013, 111, 937–945.
46. X. J. Liu, J. F. Liu and X. M. Sun, J. Mater. Chem. A, 2015, 3, 13900–13905.
47. G. He, L. Wang, H. Chen, X. Sun and X. Wang, Mater. Lett., 2013, 98, 164–167.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra12855b

---