Controllable growth of NiCo₂O₄ nanoarrays on carbon fiber cloth and its anodic performance for lithium-ion batteries

Abstract

Flexible materials are promising materials for wearable devices and have attracted much attention. Among these, carbon fiber cloth (CFC) is a soft carbon substrate on which it is suitable to grow metal oxide nanoarrays. Meanwhile, a high areal mass loading of an electrode is of great significance in practice. Consequently, this work fabricated high areal mass loading NiCo₂O₄ nanoarrays by direct growth onto a CFC without binder and conductive additives. In the experiments, the thickness of the NiCo₂O₄ layer could be controlled to 1.8–4.4 μm by varying the hydrothermal reaction time between 6–18 hours. Since a high areal mass loading of an electrode is of great significance in practice, the thickness of NiCo₂O₄ was controlled to 4.4 μm using an 18 hour hydrothermal reaction time. The electrochemical behavior of the NiCo₂O₄/CFC anode was evaluated with 10 cycles at five different current rates from 100 to 500 mA g⁻¹ in sequence and another 100 cycles at 100 mA g⁻¹. It had a reversible areal capacity of 2.39 mA h cm⁻² (799 mA h g⁻¹) at the 150th cycle when the current density was 0.299 mA cm⁻² (100 mA g⁻¹). The high areal capacity was ascribed to the larger NiCo₂O₄ mass loading on the CFC substrate. The one-dimensional flexible conductive carbon fiber backbone and the free-standing NiCo₂O₄ nanoarrays on it also released the stress generated from the discharge/charge process. The nanostructure of NiCo₂O₄/CFC in this work offers great benefits for high areal capacity anode materials.

---

1. Introduction

Lithium-ion batteries (LIBs) have been widely applied in our daily lives because of their portability and high energy density. Recently, wearable devices have attracted more attention due to the increasing demand for higher portability, and these required fabrication techniques different from traditional LIBs. When a metal foil current collector is twisted and bent repeatedly, metal fatigue would make the collector broken, and the poor electronic contact would lead to the capacity of the LIB vanishing quickly. Consequently, some soft conductive matrices have been chosen as current collectors, such as carbon nanotubes,^1,2 graphene paper,^3,4 carbon fiber cloth (CFC)⁵ and other low-dimension nanostructures.⁶ For instance, metal nitride nanoarrays could grow on carbon cloth without binder and conductive additives,^7–9 which should also be avoided when fabricating flexible electrodes.¹⁰

Carbon-based flexible materials not only act as current collectors, but also have the ability to store electrochemical energy in LIBs. However, a commercial graphite anode only has a theoretical capacity of 372 mA h g⁻¹. Alternative materials like transition metal oxides have been widely studied as anode materials due to their higher theoretical capacity.^11–13 For example, the spinel Co₃O₄ anode has been found to have a rather high theoretical capacity of 890 mA h g⁻¹.^14,15 However, its low electronic conductivity, high cost and low eco-friendliness have limited its application. NiCo₂O₄ is another spinel metal oxide. Although its theoretical capacity is similar to that of Co₃O₄,^16,17 its electronic conductivity is multiple orders greater than that of Co₃O₄.^18,19 Furthermore, NiCo₂O₄ possesses properties such as structural stability, low diffusion resistance to cations and easy electrolyte penetration, which make the ternary metal oxide fantastic in its application for energy storage.^20–22 For example, Zhang et al. prepared an anode for LIBs by fabricating NiSi_x/NiCo₂O₄ core/shell heterostructures on nickel foam.²³

Although many reports have focused on how to improve the gravimetric capacity of the active material by using a low areal mass loading, a high areal mass loading of electrode is of greater significance in practice.²⁴ For instance, an areal capacity of 4 mA h cm⁻² is regarded as a suitable power for an electronic vehicle.²⁵ Nevertheless, the areal mass loading could be increased by raising the thickness of the electrode. When a thicker electrode is fabricated with a long cycle life, a free-standing film on the collector might be feasible because it could effectively release the stress generated during the discharge/charge process.^26–28

In this work, we report a facile strategy to prepare NiCo₂O₄ nanoarrays directly on carbon fiber cloth and the thickness of NiCo₂O₄ grown on the CFC could be controlled by the reaction time. The NiCo₂O₄/CFC composite was a nanohierarchical structure, in which the three-dimension CFC severed as a flexible conductive backbone and one-dimensional NiCo₂O₄ nanoarrays as an active material subunit. The nanostructure facilitated the transport of the charge and ions in the active material. Furthermore, the anode was fabricated without conductive agent and binder, which avoided “dead volume” contributing to a low capacity and electrochemically inactive materials, respectively. NiCo₂O₄/CFC anodes with different thickness of NiCo₂O₄ coating were successfully fabricated in this work, and the role of the CFC substrate was investigated via the electrochemical performance of the anode. The superb performance of the Ni₂O₄/CFC anode indicated that it might be a potential candidate for a flexible electrode, which is an emerging and promising material for wearable devices.

2. Experimental

2.1 Materials

All the chemicals were of analytical grade and were used directly without further purification. Typically, nickel nitrate hexahydrate and cobalt nitrate hexahydrate were purchased from Sinopharm Chemical Reagent Co., Ltd. Urea was obtained from Sangon Biotech Chemicals, and carbon fiber cloth was obtained from Cetech Co., Ltd.

2.2 Synthesis of NiCo₂O₄ nanoarrays on carbon fiber cloth

The schematic synthesis route is presented in Fig. 1. Firstly, carbon fiber cloth was cut into round discs with a diameter of 13.5 mm and mass around 21 mg. Then every disc was cleaned in ethanol solution via sonication and weighed after drying at 60 °C. After pretreatment of the CFC substrate, 0.3635 g of Ni(NO₃)₂·6H₂O, 0.7275 g of Co(NO₃)₂·6H₂O and 0.9000 g of urea were dissolved in 50.00 mL distilled water. After 5 minutes of stirring, the clear reddish-brown solution obtained was transferred into a 100 mL autoclave with the CFC disc hanging in it. The autoclave was kept in an electronic oven at 110 °C for various reaction times and cooled naturally to room temperature. The hydrothermal products were washed with distilled water and ethanol three times, and were sonicated for 30 minutes to remove residual ions and unstable attached materials. After drying at 60 °C overnight, it was pyrolysed at 300 °C for 120 minutes with a ramp of 2 °C min⁻¹ in air. The final product was weighed again to calculate the loading mass of NiCo₂O₄ on the CFC. The NiCo₂O₄/CFC samples were named NCFC_x, where x represents the hydrothermal times of 6, 12, and 18 hours, respectively. NCFC_18 is the default sample of NiCo₂O₄/CFC if not specified.

Fig. 1 Schematic representation of the procedure used to fabricate NiCo₂O₄/CFC.

Similarly, pure NiCo₂O₄ nanoparticles (NPs) were prepared without CFC substrate in the hydrothermal step of the reaction.

2.3 Characterization

The morphologies of the as-prepared products were investigated using scanning electron microscopy (SEM, Hitachi S-3400N) with energy dispersive spectroscopy (EDS, Hitachi E-550) and transmission electron microscopy (TEM, JEM2100). The phase of the products was characterized using powder X-ray diffraction (XRD, Bruker D8 Advanced) and Raman spectroscopy (Jobin-Yvon LabRam-HR). The thermal stability of NiCo₂O₄ was evaluated using thermogravimetric/differential thermal analysis (TG/DTA, Perkin-Elmer Pyris Diamond). The BET surface areas were characterized using N₂ adsorption at 77 K (BELSORP-mini II). The chemical composition of CFC was measured via FTIR spectroscopy (FTIR, Nicolet 6700).

2.4 Electrochemical measurements

The CR2032-type half-coin cells were assembled in an argon-filled glovebox with H₂O and O₂ content less than 1 ppm. Carbon fiber cloth or NiCo₂O₄/CFC was used directly as the working electrode. The loading mass of NiCo₂O₄ on the CFC was calculated according to the difference of the weight between the CFC and NiCo₂O₄/CFC. NiCo₂O₄ NPs pasted on the copper foil were made as in our previous work.²⁹ Metallic lithium foil was used as the counter and reference electrodes. A Celgard 2300 microporous polypropylene film was used as the separator. The electrolyte was made from 1.0 mol L⁻¹ LiPF₆ dissolved in a mixture of ethylene carbonate and dimethyl carbonate (DMC) with a volume ratio of 1 : 1. The batteries were galvanostatically charged and discharged on a Land battery test system at different current densities calculated according to the loading mass of NiCo₂O₄ on the CFC. Cyclic voltammetry (CV) tests were recorded on a CHI760D electrochemical workstation at a scan rate of 0.2 mV s⁻¹ in the voltage range of 0.01–3.00 V. Electrochemical impedance spectroscopy (EIS) was carried out by applying an amplitude voltage of 10 mV in the frequency range of 100 kHz to 10 mHz at an open-circuit potential.

3. Results and discussion

3.1 Characteristics of NiCo₂O₄/CFC

The commercially available carbon fiber cloth was a new carbonaceous substrate with properties like a special porous structure, good compressibility, good elasticity, high electronic conductivity and high mechanical strength. The CFC in this work was woven with carbon fiber (Fig. S1A†), whose diameter was 9.5 ± 0.8 μm (Fig. S1B†). The thickness of the CFC was ca. 310 μm. The wettability³⁰ of the CFC was quantified as hydrophobic with a contact angle of ca. 138° (Fig. S2†) and the ATR-FTIR data (Fig. S2†) showed that there was hydroxyl radical on the CFC.³¹

Before immobilizing NiCo₂O₄ on the CFC, the synthesis conditions were evaluated by synthesizing pure NiCo₂O₄. Co–Ni precursors were prepared via hydrothermal reaction for 12 hours at 90, 110 and 130 °C. After the hydrothermal reaction, the three purple Co–Ni precursors collected were annealed in 300 °C for 2 hours in air, and the black NiCo₂O₄ products were finally obtained. The morphologies of the three NiCo₂O₄ products were compared using SEM on different scales. Fig. S1C† is the image of the pure NiCo₂O₄ NPs prepared via a hydrothermal reaction temperature of 110 °C. The NiCo₂O₄ NPs were urchin-like and dropped on the surface with needle-like nanoarrays, and had a diameter of ∼5 μm (Fig. S1D†). It was shown that the urchin-like NPs prepared at the three different hydrothermal temperatures had similar morphologies and sizes. Hence, a hydrothermal temperature of 110 °C was used to prepare the precursors in the following experiments, including the preparation of NiCo₂O₄/CFC.

The Co–Ni precursor of the urchin-like NPs was analyzed using TG analysis under an air atmosphere (Fig. S3†). It indicated that its weight dropped from 92.0% to 71.7% within the temperature range of 230 to 360 °C, which should have been the result of the phase change from the Co–Ni precursor to NiCo₂O₄. The calculated weight loss induced by the phase change was 22.1%, which was closer to the theoretical weight loss of the Co–Ni carbonate hydroxyl salt (24.2%),³² instead of that of the Co–Ni carbonate salt (32.5%)³³ or a mixture of Co–Ni hydroxyl salts (13.6%).³⁴ The gradual weight drop to 66.0% at 770 °C partially resulted from the decomposition of NiCo₂O₄. When the temperature rose higher above 860 °C, the little change in weight suggested that NiCo₂O₄ had decomposed completely. Hence, the ternary metal oxide was obtained by pyrolyzing the precursor at 300 °C under an air atmosphere in this study.

Optical photos of CFC and NiCo₂O₄/CFC showed that the electrodes were mechanically robust and highly flexible even after being folded and twisted (Fig. S4†). The morphologies of NiCo₂O₄/CFC were examined using SEM and (HR)TEM measurements, which are shown in Fig. 2. The surface of the CFC substrate was covered with a layer of NiCo₂O₄ NPs (Fig. 2A), and the layer of NiCo₂O₄ contained needle-like nanoarrays (Fig. 2B). The NiCo₂O₄ stripped off from the NiCo₂O₄/CFC (Fig. 2C) showed that the as-prepared NiCo₂O₄ NPs consisted of nanoneedles with a maximum diameter of ca. 80 nm and length up to the micrometer scale. The high magnification TEM (Fig. 2D) illustrated that the NiCo₂O₄ crystals were mesoporous structures. HRTEM observation (Fig. 2E) showed that the measured lattice spacings of 0.247 and 0.205 nm matched well to the (311) and (400) planes of NiCo₂O₄, and the XRD data further confirmed this.

Fig. 2 (A and B) SEM images of NiCo₂O₄/CFC at different magnifications, (C and D) TEM and (E) HRTEM images of NiCo₂O₄/CFC.

EDS (Fig. 3A) was used to determine the Ni/Co atomic ratio on the carbon fiber cloth. It showed strong signals of C, O, Co, and Ni elements. Calculated from the EDS result, the ratio of Ni/Co was about 1 : 1.984, in close agreement with the expected stoichiometric ratio of the NiCo₂O₄ phase.³⁵ Fig. 3B shows the XRD patterns of the NiCo₂O₄ NPs and NiCo₂O₄/CFC composite, which were compared with those of the standard PDF card (JCPDS, no. 20-0781). The diffraction peaks of the former at 2θ = 29.9°, 36.7°, 44.9°, 58.8° and 65.0° could be indexed to the (220), (311), (400), (511) and (440) crystal planes of NiCo₂O₄, respectively.³⁶ When NiCo₂O₄ nanoarrays were loaded on the CFC, the above five planes could be found at similar places on the 2θ axis and the broad diffraction peak at 2θ = 25.7° assigned to the (002) planes of the carbon material indicated the existence of the CFC substrate.^35,37 It was demonstrated that NiCo₂O₄ can be successively supported on the surface of the CFC substrate.

Fig. 3 (A) EDS spectrum of NiCo₂O₄/CFC. Inset: the element ratio in NiCo₂O₄/CFC; (B) XRD patterns of NiCo₂O₄ and NiCo₂O₄/CFC; (C) Raman spectra of NiCo₂O₄, CFC and NiCo₂O₄/CFC; (D) nitrogen adsorption/desorption isotherm and pore size distribution (inset) of NiCo₂O₄/CFC.

Fig. 3C shows a further comparison of the Raman spectra of the NiCo₂O₄ NPs, naked CFC and NiCo₂O₄/CFC composite within the range between 200–2000 cm⁻¹. For the NiCo₂O₄ NPs, the peaks at 454, 491 and 644 cm⁻¹ corresponded to the E_g, F_2g and A_1g vibrational modes of NiCo₂O₄, respectively.^35,38 As for the naked CFC, the peaks at 1333 and 1604 cm⁻¹ were assigned to the D and G bands of the carbon material.³⁹ As for the NiCo₂O₄/CFC composite, both the D and G bands of the carbon material were suppressed, while the peaks of NiCo₂O₄ was observed at 456, 501 and 648 cm⁻¹, which suggested that the surface of the CFC substrate was covered by the NiCo₂O₄ nanoarrays.

The N₂ adsorption isotherm of NiCo₂O₄/CFC showed a type IV curve with an H3 hysteresis loop (Fig. 3D), which indicated that there was mesoporous structure to the sample. The Brunauer–Emmett–Teller (BET) specific surface area and the total pore volume of NiCo₂O₄/CFC were 31.2 m² g⁻¹ and 0.07 m³ g⁻¹, respectively. The Barrett–Joyner–Halenda (BJH) pore size distribution analysis, shown in the inset of Fig. 3D, showed that the sizes of the mesopores were approximately 2 nm. Together with the interspace between the NiCo₂O₄ nanoneedles and that between the carbon fibers, the hierarchical porous structure of NiCo₂O₄/CFC could contribute to the infiltration of the electrolyte, the reduction of the lithium diffusion distance, and the improvement of the electrochemical properties.

The mechanism to generate the Co–Ni precursor on the CFC in the hydrothermal reaction can be expressed using the following chemical equations:

CO(NH₂)₂ + H₂O → 2NH₃ + CO₂ (1)

CO₂ + H₂O → CO₃²⁻ + 2H⁺ (2)

NH₃ + H₂O → NH₄⁺ + OH⁻ (3)

4Co²⁺ + 2Ni²⁺ + 6OH⁻ + 3CO₃²⁻ → 3Ni_2/3Co_4/3(OH)₂CO₃ (4)

At the beginning of the hydrothermal reaction, urea might be absorbed on the CFC due to the large electronegativity of the heteroatoms in urea (N, O). During the process of the hydrothermal reaction, NH₃ and CO₂ are slowly released via the hydrolysis of urea on the CFC and generate the CO₃²⁻ anion and NH₄⁺ cation (eqn (1)–(3)). Meanwhile, bimetallic carbonate hydroxide salts might be generated and anchor on the CFC substrate according to the reaction between the newborn anions around the carbon fiber and metal cations in the solution (eqn (4)). What's more, it was observed that part of the carbon fiber on the CFC substrate was not completely covered by the NiCo₂O₄ nanoarrays. Therefore, both the growth properties and the thickness of NiCo₂O₄ on the CFC could be estimated using the junction area, as shown in Fig. 4. When the hydrothermal reaction time was 6 hours, the thickness of the NiCo₂O₄ layer on the CFC was 1.8 μm (Fig. 4A), and there was a narrow multi-directional NiCo₂O₄ layer between the NiCo₂O₄ nanoneedles and CFC (Fig. 4B). When the time increased to 12 hours, the thickness of the NiCo₂O₄ layer increased to 3.3 μm (Fig. 4C), the multi-directional NiCo₂O₄ layer became thicker and the nanoneedles seemed to grow on the transition metal layer (Fig. 4D). When the time increased to 18 hours, the thickness of the NiCo₂O₄ layer rose to 4.4 μm (Fig. 4E). An obviously thicker layer was at the bottom of the nanoneedles and some nanoneedles seemed to aggregate (Fig. 4F). TEM images showed that the diameters of the NiCo₂O₄ nanoneedles scratched off from the above three samples were close (Fig. 1 and S5†). It was supposed that, during the growth of the nanoneedles, some clusters might anchor onto previously generated nanoneedles and grow as new nanoneedles. Consequently, the nanoneedles at the bottom seemed to self-aggregate and the transition metal layer seemed to become thicker. Furthermore, mesoporous structures and the (400) and (311) planes of the NiCo₂O₄ crystal in the three different reaction times were observed from the (HR)TEM images. The morphology evolution and growth mechanism of the nanoarrays on the CFC was similar with the proposal of Co₃O₄ grown on nickel foam by Kong et al.;⁴⁰ the morphology of the NiCo₂O₄ nanoarrays on the CFC were affected by the varied hydrothermal reaction time, which resulted in a varied concentration of CO₃²⁻ and NH₄⁺ during the reaction. Therefore, it might be deduced that a longer reaction time would lead to a thicker nanoarray layer and that the nanoneedles growing on the tip were growing along with the self-aggregation at the bottom of the nanoarrays.

Fig. 4 SEM images of (A and B) NCFC_6, (C and D) NCFC_12 and (E and F) NCFC_18, and their junction areas between NiCo₂O₄ and CFC.

Finally, NiCo₂O₄ nanoarrays were generated on the CFC during the low temperature calcination and the release of CO₂ and H₂O gave the NiCo₂O₄ nanoarrays a mesoporous characteristic (eqn (5)).

3.2 Electrochemical performance of the NiCo₂O₄/CFC anode

The electrochemical performance of NiCo₂O₄/CFC was evaluated in a 2032 coin cell with Li foil as the counter electrode. The lithium storage mechanism of NiCo₂O₄ was studied in previous reports.^33,41 Generally, the electrochemical reaction involved in the charge–discharge process can be expressed as follows:

NiCo₂O₄ + 8Li⁺ + 8e⁻ → Ni + 2Co + 4Li₂O (6)

Ni + Li₂O ⇌ NiO + 2Li⁺ + 2e⁻ (7)

Co + Li₂O ⇌ CoO + 2Li⁺ + 2e⁻ (8)

3CoO + Li₂O ⇌ Co₃O₄ + 2Li⁺ + 2e⁻ (9)

The first three CV curves of NiCo₂O₄/CFC are displayed in Fig. 5A. Compared with the CV plots of the NiCo₂O₄ NPs and CFC (Fig. S6A and C†), during the first cathodic sweep, the peak centered at 0.70 V was ascribed to the decomposition of NiCo₂O₄, which was reduced to metallic Co and Ni. Because the surface of the carbon fibers was partially exposed to the electrolyte without the coverage of NiCo₂O₄, the obvious redox peaks below ∼0.5 V in the figure were ascribed to the intercalation/deintercalation process of Li⁺ in the carbon.^39,42 In the following sweeps, the oxidation peaks located at 1.45 and 2.16 V corresponded to the oxidation of metallic Ni and Co to their metal oxides, respectively, while the reduction peak shifted to 0.84 V corresponded to the reduction process of the metal oxides to the metal. It was notable that the larger integration area contributed by the CFC would make it impossible to underestimate the contribution of the CFC to the total capacity of the whole electrode.

Fig. 5 (A) First three CV curves of NiCo₂O₄/CFC at a scan rate of 0.5 mV s⁻¹ between 0.01 and 3.00 V vs. Li/Li⁺; (B) discharge/charge curves for the 1st, 2nd and 3rd cycles in the voltage range 0.01–3.00 V at a current rate of 100 mA g⁻¹; (C) gravimetric capacity vs. cycle number at a current rate of 100 mA g⁻¹ for NiCo₂O₄, CFC and NiCo₂O₄ in three NiCo₂O₄/CFC samples with different mass loadings; (D) areal capacity vs. cycle number at a current rate of 100 mA g⁻¹ for the above five anodes; (E) specific gravimetric capacity at different current rates of 100, 200, 300, 400, 500 and 100 mA g⁻¹ for NiCo₂O₄/CFC, NiCo₂O₄ and CFC; (F) Nyquist plots of NiCo₂O₄/CFC before cycling and after cycling at various current rates for a total of 150 cycles. Inset: the equivalent circuit used to fit the impedance spectra.

The discharge/charge curves of NiCo₂O₄/CFC for the 1st, 2nd and 3rd cycles at 100 mA g⁻¹ in the voltage range of 0.01–3.00 V are presented in Fig. 5B. There is a long plateau at 1.0 V in the first discharge curve, which might be a result of the irreversible decomposition of NiCo₂O₄ and the formation of a solid-electrolyte interphase (SEI) film. The plateaus at 1.3–1.6 V and 1.9–2.3 V in the first charge curve should be due to the oxidation of Ni and Co. The coulombic efficiency (CE) of the first cycle was 78.1% corresponding to a discharge capacity of 1529 mA h g⁻¹ and a charge capacity of 1194 mA h g⁻¹. After the 1st cycle, the plateau at 1.0 V in the discharge curve was replaced by a plateau at 1.1–0.7 V, which meant that NiCo₂O₄ had thoroughly decomposed.

Fig. 5C and D compare the cycle performance of the CFC, NiCo₂O₄, and three different NiCo₂O₄/CFC samples in two different ways. The three NiCo₂O₄/CFC samples were named NCFC_6, NCFC_12 and NCFC_18, in which the loading mass of NiCo₂O₄ on the CFC was 1.09, 2.25 and 5.09 mg cm⁻², respectively. Fig. 5C shows the actual gravimetric capacity of NiCo₂O₄ in the above three NiCo₂O₄/CFC samples, and the gravimetric capacity of CFC and NiCo₂O₄. The actual gravimetric capacity of NiCo₂O₄ in NiCo₂O₄/CFC was calculated based on eqn (10).

C_NiCo2O4= (Q_total − C_CFCm_CFC)/m_NiCo2O4 (10)

Here C_NiCo2O4 is the gravimetric capacity of NiCo₂O₄, Q_total is the measured Coulomb charge, m_NiCo2O4 is the mass of NiCo₂O₄, and C_CFC is the gravimetric capacity of CFC. When the current rate was 100 mA g⁻¹, the gravimetric capacity of the CFC was about 95 mA h g⁻¹, while that of NiCo₂O₄ decayed quickly and then was a similar level to the CFC. The actual gravimetric capacity of NCFC_6 was higher than the theoretical capacity of NiCo₂O₄ (817 mA h g⁻¹). The higher actual capacity of NiCo₂O₄ in NiCo₂O₄/CFC might be ascribed to the synergistic effect between NiCo₂O₄ and CFC. Furthermore, the mass of CFC was ten times higher than that of NiCo₂O₄, and variations in the CFC capacity in the control experiment would result in larger fluctuations of the calculated actual capacity. This might be another reason causing the higher actual gravimetric capacity of NiCo₂O₄. The actual gravimetric capacities of NCFC_12 and NCFC_18 decayed at first and then were 754 and 363 mA h g⁻¹ at the 100th cycle, respectively. The actual gravimetric capacity of NiCo₂O₄ in NiCo₂O₄/CFC after cycling decreased when the loading mass increased. The fact that a lower loading mass led to a higher capacity of NiCo₂O₄ indicated that a lower loading mass of NiCo₂O₄ on CFC, or a thinner NiCo₂O₄ layer on the carbon fiber was beneficial to improving the capacity of NiCo₂O₄.

Along with further developing research on LIBs, it was important to provide the loading mass of the active materials to evaluate their practical use.⁴³ Fig. 5D compares the five anodes in Fig. 5C in the form of the areal capacity of the whole electrode. Although the gravimetric capacities of the NiCo₂O₄ anode and CFC anode were at a similar level, the lower loading mass of NiCo₂O₄ (1.07 mg cm⁻²) made its areal capacity negligible, while the larger loading mass of CFC (14.68 mg cm⁻²) made that of the CFC 1.16 mA h cm⁻². The areal capacities of the three NiCo₂O₄/CFC samples, NCFC_6, NCFC_12 and NCFC_18 at the 80th cycle, were 2.59, 3.04 and 3.10 mA h cm⁻², respectively. The areal capacity of the whole anode was provided by both the CFC and the NiCo₂O₄ layer covering it. When the loading mass of NiCo₂O₄ in NiCo₂O₄/CFC was low, the contribution of the CFC was almost close to that of NiCo₂O₄ in NiCo₂O₄/CFC. However, it was observed that the areal capacity of the NiCo₂O₄/CFC anode could be improved slightly by sacrificing the dosage of NiCo₂O₄ in this work.

The rate capability of the NiCo₂O₄/CFC anode is demonstrated in Fig. 5E and S6E.† The loading mass of NiCo₂O₄ on the CFC in the NiCo₂O₄/CFC sample was 2.99 mg cm⁻². The anode ran 10 cycles at five current rates and then another 100 cycles at the first current again. The experimental results showed that the discharge capacity of the NiCo₂O₄/CFC anode decreased from 4.57 mA h cm⁻² (1529 mA h g⁻¹) to 2.68 mA h cm⁻² (898 mA h g⁻¹) at a current rate of 100 mA g⁻¹, and it was about 2.17, 1.98, 1.81 and 1.69 mA h cm⁻² at current rates of 200, 300, 400 and 500 mA g⁻¹, respectively. Then, when the current rate returned to 100 mA g⁻¹, the anode of NiCo₂O₄/CFC continued to run for another 100 discharge/charge cycles (Fig. S6E†). Its discharge capacity was found to be 2.42 mA h cm⁻² at the 51st cycle, and then slowly decreased to 2.39 mA h cm⁻² at the 150th cycle, during which the CE of most cycles was above 99.4%. Therefore, it could be concluded that the anode had a stable capacity over 100 cycles at a higher areal mass loading.

Electrochemical impedance spectroscopy (EIS) was used to compare the electrochemical kinetics difference of the NiCo₂O₄/CFC anodes before and after cycling. The Nyquist plots (Fig. 5F and S7†) consist of a semicircle in the middle frequency region and an inclined line in the low frequency region. The diameter of the semicircle and the slope of the inclined line can be decoded as the charge transfer resistance (R_ct) and the Warburg resistance (W), which represent the charge transfer at the interface and the diffusion of the redox species in the electrode, respectively. EIS data was fitted to the equivalent circuit^44,45 inset in the figure, and the fitting results are listed in Table S1.† The R_ct value of NiCo₂O₄/CFC with different loading mass was obviously lower than that of NiCo₂O₄, which should be a result of the superior conductivity of CFC. The small difference in the R_ct values of NiCo₂O₄/CFC with different loading mass should also be ascribed to the influence of CFC. The thickness of the NiCo₂O₄ layer on the CFC increased when the NiCo₂O₄/CFC sample was prepared over a long time. Correspondingly, the W value increased when the diffusion path length for lithium ions became longer. Moreover, after cycling at various current rates for 150 cycles, both R_ct and W values of NCFC_18 in Fig. S6E† increased, which indicated that the diffusion of electrons and Li⁺ turned out to be slower in the LIB after cycling. The difference would be a result of the structural evolution of NiCo₂O₄/CFC during the cycling.

The structural stability of the NiCo₂O₄/CFC anodes was then evaluated via SEM images of anodes after cycling (Fig. 6). The batteries were disassembled inside a glovebox, and the anodes were washed with DMC to remove the lithium salts. Although the NiCo₂O₄ nanoneedles were not easy to distinguish from each other, the NiCo₂O₄ layer could still cover the carbon fiber (Fig. 6A). This indicated that the NiCo₂O₄ needle nanoarrays and 1-D carbon fiber could efficiently release the stress from the volumetric change during the discharge/charge process. However, the slits between adjacent NiCo₂O₄/carbon fibers (Fig. 6B) indicated that the NiCo₂O₄ layers on adjacent carbon fibers might squeeze each other during the discharge/charge process, and result in the stripping of the NiCo₂O₄ layer off the CFC substrate and a capacity decay of the active materials. Hence, it might be helpful to improve the capacity of NiCo₂O₄/CFC by preparing a thinner NiCo₂O₄ layer on the CFC or weaving a CFC substrate with a larger grid to accommodate the thickness change of the coating.

Fig. 6 SEM images of NiCo₂O₄/CFC after cycling at 100 mA g⁻¹ for 100 cycles at different magnifications.

The electrochemical performance of NiCo₂O₄/CFC in this work was compared with those of some other NiCo₂O₄ electrodes reported in previous studies. The calculated results are listed in Table 1, which show that the areal capacity of the NiCo₂O₄/CFC anode in this work was higher than those of other works. Presumably, the high areal capacity of NiCo₂O₄/CFC should be ascribed to the following reasons. First, the areal capacity of the whole anode was provided by both NiCo₂O₄ and the CFC. Second, the electronic conductivity of the CFC was good because it was a carbon material, and that of NiCo₂O₄ is relatively greater than that of many other transition metal oxides. Third, the mesoporous structure of NiCo₂O₄ could facilitate the transfer of the charge and the ions. Fourth, the robust layer of NiCo₂O₄ was directly covered onto the CFC without binder and conductive additives, which kept the active material from stripping the current collector. Finally, both the flexible CFC substrate and the free-standing NiCo₂O₄ nanoarrays could release the stress from the volumetric change during the discharge/charge process to facilitate the stability of the capacity.

Table 1 Electrochemical performances of different NiCo₂O₄ electrodes

Morphologies	Areal capacity (mA h cm^−2)	Current rate (mA g^−1)	Loading mass (mg cm^−2)	Substrate	Year	Reference
Nanoneedles	2.59	100	1.09	CFC	—	This work
Nanoflakes	1.64	500	2.5	Ni foam	2015	46
Nanowires	0.86	100	2.1	Ni foam	2015	47
Nanosheets	0.77	100	1	Cu foil	2015	48
Nanoflakes	0.88	500	1	Cu foil	2014	49
Nanobelts/nanowires	0.98/0.52	500	0.8	Carbon textiles	2014	19

---

4. Conclusions

In summary, we had fabricated a robust layer of free-standing NiCo₂O₄ nanoarrays covered on carbon fiber cloth for LIB anodes. The CFC substrate provided a flexible electronic conductive substrate to load NiCo₂O₄. In addition, the composite synthesis and electrode fabrication processes are simple, and free of binder and conductive reagent. As a result, the NiCo₂O₄/CFC presented here showed great promise for a high areal capacity anode.

Acknowledgements

This work was financially supported by National Natural Science Foundation of China (21165010, 21465014 and 21465015), Science and Technology Support Program of Jiangxi Province (20123BBE50104 and 20133BBE50008), Natural Science Foundation of Jiangxi Province (20142BAB203101), The Ministry of Education by the Specialized Research Fund for the Doctoral Program of Higher Education (20133604110002) and the Ground Plan of Science and Technology Projects of Jiangxi Educational Committee (KJLD14023), Innovation Foundation for graduate student of Jiangxi Province and Jiangxi Normal University (YC2013-B026) and Scientific Research Funds of Jiangxi Normal University (No. 5487).

Notes and references

1. A. Marschilok, C.-Y. Lee, A. Subramanian, K. J. Takeuchi and E. S. Takeuchi, Energy Environ. Sci., 2011, 4, 2943–2951.
2. S. Lee, D. Cho and Y. Jeong, Fibers Polym., 2015, 16, 1600–1604.
3. L. S. Zhang, W. Fan and T. X. Liu, RSC Adv., 2015, 5, 43130–43140.
4. C. Wang, D. Li, C. O. Too and G. G. Wallace, Chem. Mater., 2009, 21, 2604–2606.
5. B. Liu, J. Zhang, X. Wang, G. Chen, D. Chen, C. Zhou and G. Shen, Nano Lett., 2012, 12, 3005–3011.
6. L. Yue, H. Zhong and L. Zhang, Electrochim. Acta, 2012, 76, 326–332.
7. M. S. Balogun, M. H. Yu, C. Li, T. Zhai, Y. Liu, X. H. Lu and Y. X. Tong, J. Mater. Chem. A, 2014, 2, 10825–10829.
8. M. S. Balogun, C. Li, Y. X. Zeng, M. H. Yu, Q. L. Wu, M. M. Wu, X. H. Lu and Y. X. Tong, J. Power Sources, 2014, 272, 946–953.
9. M. S. Balogun, M. H. Yu, Y. C. Huan, C. Li, P. P. Fang, Y. Li, X. H. Lu and Y. X. Tong, Nano Energy, 2015, 11, 348–355.
10. G. Zhou, F. Li and H.-M. Cheng, Energy Environ. Sci., 2014, 7, 1307–1338.
11. Y. Z. Wan, Z. W. Yang, G. Y. Xiong, R. S. Guo, Z. Liu and H. L. Luo, J. Power Sources, 2015, 294, 414–419.
12. T. V. Thi, A. K. Rai, J. Gim and J. Kim, J. Power Sources, 2015, 292, 23–30.
13. H. B. Geng, X. Q. Cao, Y. Zhang, K. M. Geng, G. L. Qu, M. H. Tang, J. W. Zheng, Y. G. Yang and H. W. Gu, J. Power Sources, 2015, 294, 465–472.
14. H. Fang, S. C. Zhang, W. B. Liu, Z. J. Du, X. M. Wu and Y. L. Xing, Electrochim. Acta, 2013, 108, 651–659.
15. Q. T. Qu, T. Gao, H. Y. Zheng, X. X. Li, H. M. Liu, M. Shen, J. Shao and H. H. Zheng, Carbon, 2015, 92, 119–125.
16. J. Xu, L. He, W. Xu, H. Tang, H. Liu, T. Han, C. Zhang and Y. Zhang, Electrochim. Acta, 2014, 145, 185–192.
17. C. C. Sun, J. Yang, X. H. Rui, W. N. Zhang, Q. Y. Yan, P. Chen, F. W. Huo, W. Huang and X. C. Dong, J. Mater. Chem. A, 2015, 3, 8483–8488.
18. H. S. Jadhav, R. S. Kalubarme, C.-N. Park, J. Kim and C.-J. Park, Nanoscale, 2014, 6, 10071–10076.
19. L. Shen, Q. Che, H. Li and X. Zhang, Adv. Funct. Mater., 2014, 24, 2630–2637.
20. L. Zhang, S. Zhang, K. Zhang, G. Xu, X. He, S. Dong, Z. Liu, C. Huang, L. Gu and G. Cui, Chem. Commun., 2013, 49, 3540–3542.
21. W.-M. Liu, T.-T. Gao, Y. Yang, Q. Sun and Z.-W. Fu, Phys. Chem. Chem. Phys., 2013, 15, 15806–15810.
22. Y. Chen, B. Qu, L. Hu, Z. Xu, Q. Li and T. Wang, Nanoscale, 2013, 5, 9812–9820.
23. Q. Zhang, H. Chen, J. Wang, D. Xu, X. Li, Y. Yang and K. Zhang, ChemSusChem, 2014, 7, 2325–2334.
24. Z. Lu, N. Liu, H.-W. Lee, J. Zhao, W. Li, Y. Li and Y. Cui, ACS Nano, 2015, 9, 2540–2547.
25. E. Peled, F. Patolsky, D. Golodnitsky, K. Freedman, G. Davidi and D. Schneier, Nano Lett., 2015, 15, 3907–3916.
26. J.-S. Kim, D. Ko, D.-J. Yoo, D. S. Jung, C. T. Yavuz, N.-I. Kim, I.-S. Choi, J. Y. Song and J. W. Choi, Nano Lett., 2015, 15, 2350–2357.
27. B. Wang, X. Li, B. Luo, Y. Jia and L. Zhi, Nanoscale, 2013, 5, 1470–1474.
28. K. H. Seng, J. Liu, Z. P. Guo, Z. X. Chen, D. Jia and H. K. Liu, Electrochem. Commun., 2011, 13, 383–386.
29. S. Chen, J. Wu, R. Zhou, L. Zuo, P. Li, Y. Song and L. Wang, Electrochim. Acta, 2015, 180, 78–85.
30. Y. D. Zhang, Z. G. Hu, Y. R. Liang, Y. Y. Yang, N. An, Z. M. Li and H. Y. Wu, J. Mater. Chem. A, 2015, 3, 15057–15067.
31. H. L. Yu, C. L. Zhu, K. Zhang, Y. J. Chen, C. Y. Li, P. Gao, P. P. Yang and Q. Y. Ouyang, J. Mater. Chem. A, 2014, 2, 4551–4557.
32. J. Xiao and S. Yang, RSC Adv., 2011, 1, 588–595.
33. J. Li, S. Xiong, Y. Liu, Z. Ju and Y. Qian, ACS Appl. Mater. Interfaces, 2013, 5, 981–988.
34. N. Garg, M. Basu and A. K. Ganguli, J. Phys. Chem. C, 2014, 118, 17332–17341.
35. N. Padmanathan and S. Selladurai, RSC Adv., 2014, 4, 8341–8349.
36. H. Shi and G. Zhao, J. Phys. Chem. C, 2014, 118, 25939–25946.
37. M. Dirican, O. Yildiz, Y. Lu, X. Fang, H. Jiang, H. Kizil and X. Zhang, Electrochim. Acta, 2015, 169, 52–60.
38. W. Xiong, Y. Gao, X. Wu, X. Hu, D. Lan, Y. Chen, X. Pu, Y. Zeng, J. Su and Z. Zhu, ACS Appl. Mater. Interfaces, 2014, 6, 19416–19423.
39. Y. Luo, M. S. Balogun, W. T. Qiu, R. R. Zhao, P. Liu and Y. X. Tong, Chem. Commun., 2015, 51, 13016–13019.
40. D. Kong, J. Luo, Y. Wang, W. Ren, T. Yu, Y. Luo, Y. Yang and C. Cheng, Adv. Funct. Mater., 2014, 24, 3815–3826.
41. L. Li, Y. Cheah, Y. Ko, P. Teh, G. Wee, C. Wong, S. Peng and M. Srinivasan, J. Mater. Chem. A, 2013, 1, 10935–10941.
42. M.-S. Balogun, M. Yu, C. Li, T. Zhai, Y. Liu, X. Lu and Y. Tong, J. Mater. Chem. A, 2014, 2, 10825–10829.
43. M. Zhou, X. Li, B. Wang, Y. Zhang, J. Ning, Z. Xiao, X. Zhang, Y. Chang and L. Zhi, Nano Lett., 2015, 15, 6222–6228.
44. L. N. Gao, X. F. Wang, Z. Xie, W. F. Song, L. J. Wang, X. Wu, F. Y. Qu, D. Chen and G. Z. Shen, J. Mater. Chem. A, 2013, 1, 7167–7173.
45. C. Wang, W. Wan, Y. H. Huang, J. T. Chen, H. H. Zhou and X. X. Zhang, Nanoscale, 2014, 6, 5351–5358.
46. Q. Zheng, X. Zhang and Y. Shen, Mater. Res. Bull., 2015, 63, 211–215.
47. G. Chen, J. Yang, J. Tang and X. Zhou, RSC Adv., 2015, 5, 23067–23072.
48. A. K. Mondal, D. Su, S. Chen, K. Kretschmer, X. Xie, H.-J. Ahn and G. Wang, ChemPhysChem, 2015, 16, 169–175.
49. A. K. Mondal, D. Su, S. Chen, X. Xie and G. Wang, ACS Appl. Mater. Interfaces, 2014, 6, 14827–14835.

---

Footnote

† Electronic supplementary information (ESI) available: Electrochemical performances of carbon fiber cloth and NiCo₂O₄ NPs. See DOI: 10.1039/c5ra19600k

---