Construction of NiCo₂O₄ nanoflake arrays on cellulose-derived carbon nanofibers as a freestanding electrode for high-performance supercapacitors

Abstract

Cellulose has a wide range of applications in many fields due to their naturally degradable and low-cost characteristics, but few studies can achieve cellulose-nanofibers by conventional electrospinning. Herein, we demonstrate that the freestanding cellulose-based carbon nanofibers are successfully obtained by a special design of electrospinning firstly, pre-oxidation and high-temperature carbonization (1600 °C), which display a superior electrical conductivity of 31.2 S·cm^–1 and larger specific surface area of 35.61 m²·g^–1 than that of the polyacrylonitrile-based carbon nanofibers (electrical conductivity of 18.5 S·cm^–1, specific surface area of 12 m²·g^–1). The NiCo₂O₄ nanoflake arrays are grown uniformly on the cellulose-based carbon nanofibers successfully by a facile one-step solvothermal and calcination method. The as-prepared cellulose-based carbon nanofibers/NiCo₂O₄ nanoflake arrays are directly used as electrodes to achieve a high specific capacitance of 1010 F·g^–1 at 1 A·g^–1 and a good cycling stability with 90.84% capacitance retention after 3000 times at 10 A·g^–1. Furthermore, the all-solid-state symmetric supercapacitors assembled from the cellulose-based carbon nanofibers/NiCo₂O₄ deliver a high energy density of 62 W·h·kg^–1 at a power density of 1200 W·kg^–1. Six all-solid-state symmetric supercapacitors in series can also power a ‘DHU’ logo consisted of 36 light emitting diodes, confirming that the cellulose-based carbon nanofiber is a promising carbon matrix material for energy storage devices.

Key words： cellulose; carbon nanofibers; NiCo₂O₄; supercapacitors

1 Introduction

Currently, it is urgent to demand for portable, safe, renewable, eco-friendly energy storage devices due to the limited fossil fuel supply and serious environmental pollution. Supercapacitors have drew much attention because of their high power density, superb cycling stability and short charging time [1]. There are mainly two energy storage mechanisms in supercapacitors: electrical double layers and pseudo-capacitance [2]. The composition, structure and specific surface area of electrode materials have a great effect in electrochemical performance of supercapacitors [1,3–5]. Up to now, many kinds of electrode materials, such as carbon materials [6,7] and transition metal oxides [8,9], sulfides [10,11], hydroxides [12,13] as well as their combination [14–16] have been widely studied. Especially, composites of carbon materials and binary transition metal oxide were extensively used in high energy density supercapacitors, because carbon materials can not only improve the electrical conductivity, but also buffer the volume expansion of transition metal materials during the cycling.

Carbon materials, including MXene, carbon foam, graphene, carbon nanotubes, micro/meso-porous carbon sphere and carbon nanofibers (CNFs) have been reported a lot. Among them, CNFs have attracted the most attention due to large surface area, light-weight as well as unique three-dimensional (3D) networks, which are directly used as current collector without binders and conductive additives. Most of the carbon fiber precursors are petrochemicals, such as polyacrylonitrile, polyvinylpyrrolidone, polybenzimidazole and pitch, which are high-cost products and emit toxic gases during carbonization. However, cellulose is a better precursor for CNFs because it is low-cost, renewable and natural abundant. There are few reports that the cellulosic derivatives (cellulose acetate and methyl cellulose) nanofibers can be prepared by conventional electrospinning [17–19], but the cellulosic derivatives-based carbon fibers are fragile and the carbon yield is low. It is still a difficult task to obtain freestanding cellulose-based CNFs (C-CNFs) with a unique interconnected fiber network structure owing to the fact that the cellulose is a macromolecular polysaccharide composed of glucose with a specific molecular structure. Additionally, the α-cellulose cannot be dissolved into the water, but only in a few organic solvents due to the rigid pyranose ring structure, and its trifluoroacetic acid/methylene dichloride solution has low viscosity. Therefore, it is difficult for α-cellulose to be used in electrospinning by itself. However, the addition of polyethylene oxide (PEO) can effectively reduce surface tension of electrospinning precursor solution, aiding in fiber formation. In this work, the cellulose nanofibers (C-NFs) were successfully prepared through a special design of new electrospinning receiver with PEO as additives in this study. Then, the freestanding interconnected C-CNFs were successfully obtained after pre-oxidation and carbonization. Although the C-CNFs have been used as freestanding electrodes, the lower capacity cannot meet the requirements of energy storage devices. Thus, the freestanding C-CNFs substrates should be loaded with active materials to directly served as binder-free electrodes for high-performance supercapacitors.

A variety of active electrode materials with high specific capacitance, especially pseudo-capacitive materials (transition metal oxides/sulfides) embrace ultrahigh theoretical capacitance and excellent electrochemical activity. Among them, NiCo₂O₄, a binary transition metal oxide, is extensively researched as electrodes for supercapacitors because of the high theoretical capacity (1370 F·g^–1), rich valence state and lower activation energy for ion and electron transport [18,20–22]. However, the single NiCo₂O₄ usually suffers from a serious structural distortion during the charge/discharge storage process because of the low electrical conductivity and poor mechanical/chemical stability, which reduces the energy storage capability and restricts it application in supercapacitors [23,24]. To enhance the electrochemical property of NiCo₂O₄, it is an effective strategy to use the as-prepared C-CNFs as freestanding carbon substrates to support NiCo₂O₄. As a result, the NiCo₂O₄ nanoflake arrays are uniformly coated on C-CNFs by one-step solvothermal method to obtain binder-free C-CNFs/NiCo₂O₄ nanoflake arrays. Benefiting from the 3D interconnected network structure, the formed C-CNFs/NiCo₂O₄ composite can obtain superior electrochemical performance, such as a higher specific capacitance (1010 F·g^–1 at 1 A·g^–1) and better cycling stability (90.84% capacitance retention at 10 A·g^–1 after 3000 times). Moreover, the all-solid-state symmetric supercapacitors (ASSCs) constructed with C-CNFs/NiCo₂O₄ can power the ‘DHU’ logo consisting of 36 light emitting diodes (LEDs) and thus exhibiting a huge potential of the C-CNFs as carbon matrix in energy storage devices.

2 Experimental

2.1 Preparation of C-CNFs

All regents were analytical grade which could be used directly in the present study. Both α-cellulose and PEO (9:1 w/w) were added into a mixed solution of trifluoroacetic acid and methylene dichloride (9:1 w/w) at room temperature by stirring for 72 h until complete dissolution. The α-cellulose/PEO solution with various concentration (3.5–7.0 wt %) was transferred into a 5 mL syringe with a steel needle to be electrospun. It was 13 cm between the needle tip and the collector. The flow rate was 0.75 mL·h^–1 using a syringe pump (LSP01-3A, Longer Pump), the applied positive voltage was 17 kV and applied negative voltage was 3.0 kV. The C-NFs were gathered by the modified new electrospinning receiving devices. Afterwards, the C-NFs were pre-oxidized at 245 °C for 90 min in air and carbonized at 1600 °C at a rate of 10 °C·min^–1 under N₂ atmosphere for 30 min to obtain C-CNFs. The thickness and areal specific mass of C-CNFs were ~0.08 mm and 1.5 g·cm^–2, respectively. To explore the effect of carbonization temperature on the properties of C-CNFs, the carbonization temperature of C-NFs was added to 800 and 1200 °C, respectively. For comparison, the CNFs based on poly(acrylonitrile-co-methylhydrogen itaconate) (P-CNFs) were also prepared in our previous work [25].

2.2 Synthesis of C-CNFs/NiCo₂O₄ nanoflake arrays

The C-CNFs/NiCo₂O₄ composite was synthesized via a facile solvothermal method. In detail, 0.1 g Ni(NO₃)₂·6H₂O, 0.2 g Co(NO₃)₂·6H₂O and 0.3 g hexamethylenetetramine were dissolved in 48 mL methanol with stirring for 10 min and then transferred into a 50 mL Teflon-lined stainless steel autoclave. A piece of C-CNFs (2 cm × 4 cm) was immersed into the solution vertically, and heated to 150 °C for 8 h. When the reaction was completed and cooled to room temperature, the C-CNFs/NiCo₂O₄ precursor was taken out and washed multiple times by deionized water and alcohol, then dried at 60 °C overnight in an oven. At last, C-CNFs/NiCo₂O₄ nanoflake arrays were obtained by annealing the precursor at 350 °C for 2 h with heating rate of 2 °C·min^–1. For comparison, P-CNFs/NiCo₂O₄ nanoflake arrays were synthesized by using P-CNFs instead of C-CNFs with a similar procedure of C-CNFs/NiCo₂O₄ nanoflake arrays.

2.3 Materials characterization

The microstructure of the samples was characterized by field emission scanning electron microscope (SEM, Hitachi SU8010, Japan) and high-resolution transmission electron microscopy (HRTEM, FEI Talos F200S, USA). The crystal structure of the samples was analyzed by an X-ray diffraction diffractometer (XRD, Rigaku D/max-2550, Cu Kα, scan rate: 2 °·min^–1). X-ray photoelectron spectroscopy (XPS, Escalab-250Xi) was employed to obtain the surface elements composition of the sample and binding energy of elements. The structure of electrodes was measured by Raman spectrometer (inVia Reflex, 633 nm, China). The content of NiCo₂O₄ in the composites was obtained by thermogravimetric analysis (TGA, 209 F1). The Fourier transform infrared (FTIR) spectra of samples were characterized by using the NEXUS-670 spectrometer. The nitrogen adsorption−desorption isotherms and pore size distribution of the electrodes were achieved by a fully automatic specific surface area and pore size tester (ASAP2460) at 77 K. The C element was determined by the elemental analyzer (VarioEL-III, Germany) and the electrical conductivities of both C-CNFs and P-CNFs were measured using a four-point probe (MCP-T370, China).

2.4 Electrochemical measurements

Electrochemical characterization of the samples was carried out by an electrochemical workstation (Ivium-V01320). The three-electrode cell was composed of reference electrode (saturated calomel electrode), counter electrode (Pt plate) and working electrode, and the electrolyte was 2 mol·L^–1 KOH aqueous, which was used for cyclic voltammetry (CV), galvanostatic charge-discharge (GCD), and electrochemical impedance spectroscopy (EIS). The CV curves were performed in the range of 5–80 mV·s^–1 at a potential window of –0.1–0.6 V. The current densities of GCD were in the range from 1 to 10 A·g^–1 at the potential range of 0 to 0.45 V. The EIS measurements were performed at an open circuit potential in the frequency range of 0.01 Hz to 100 kHz. The specific capacitance was calculated by the following formula:

(1)C=\frac{I\mathrm{\Delta }t}{m\mathrm{\Delta }V}.

Therein, C (F·g^–1) represents the specific capacitance, the I (A) represents charge or discharge current, m (g) is the mass of active materials, ΔV (V) designates the voltage window.

2.5 Preparation of ASSCs

Polyvinyl alcohol (PVA) (3 g) and KOH (0.6 g) were dissolved in the deionized water (40 mL) under 85 °C to obtain PVA-KOH gel electrolyte. Two pieces of electrodes (C-CNFs/NiCo₂O₄, 1 cm × 2.5 cm) and a piece of polyethylene (PE, 2 cm × 4 cm) were put in the gel electrolyte immersing for 10 min and dried at room temperature. Then, the two pieces of electrodes with PE as the separator were assembled to fabricate the ASSC, which was on the PE terephthalate substrate.

In the two-electrode systems, the specific energy density (W·h·kg^–1) and specific power density (W·kg^–1) of the assembled ASSCs were calculated according to the Eqs. (2), (3) and (4), respectively.

(2)C=\frac{I\mathrm{\Delta }t}{2m\mathrm{\Delta }V},

(3)E=\frac{{C\mathrm{\Delta }V}^2}{2\times 3.6},

(4)P=\frac{3600E}{\mathrm{\Delta }t},

where E (W·h·kg^–1) is energy density, C (F·g^–1) is specific capacitance, ∆V (V) is the potential range, P (W·kg^–1) is the power density, and \mathrm{\Delta }t (s) is the discharge time.

3 Results and discussion

3.1 Materials characterization

(a) shows the preparation process of C-NFs, the positive voltage is attached to the syringe needle and electrospinning receiver is applied negative voltage. It could be seen that the receiving device is covered by a thick and uniform C-NFs film. As from (b), the C-CNFs were acquired after stabilization in air and carbonization in N₂, and the as-prepared C-CNFs were immersed into mixed solutions by solvothermal reaction to obtain C-CNFs/NiCo₂O₄ precursor. After annealed, the C-CNFs/NiCo₂O₄ nanoflake arrays electrodes were successfully fabricated.

Scheme1 Schematic illustration of the preparation of C-CNFs/NiCo₂O₄ nanoflake arrays.

Full size|PPT slide

The concentration of electrospinning precursor played a significant role in the stable interconnected C-NFs, and was studied by adjusting to the concentration of α-cellulose/PEO, using same syringe (5 mL) and needles (17 G) in the electrospinning process. SEM images display different morphologies of C-NFs with various concentration of electrospinning precursor (Fig.1 and S1, cf. Electronic Supplementary Material, ESM). The continuous and interconnected C-NFs networks were successfully prepared using 6.0 and 6.5 wt % of electrospinning precursor (Fig.1(a) and inset, Figs. S1(d) and S1(e)). When the concentrations were reduced to lower values (5.5, 4.5 or 3.5 wt %), the surface of the C-NFs gradually emerged some beads and failed to form the uniform structure (Figs. S1(a–c)). The C-NFs obtained at higher solution concentration (7.0 wt %) were bonded together, and continuous and interconnected C-NF disappeared (Fig. S1(f)). In addition, Fig.1(b) and inset revealed that the C-CNFs (α-cellulose/PEO-6.0 wt %) still maintained uniform and interconnected networks after carbonization of C-NFs, which was similar with morphology of C-NFs. However, the C-NF membrane contracted to a large extent, and the diameter of C-CNFs decreased significantly after carbonization of C-NFs (Fig.1(b) and inset). The results of SEM images exhibited the two cases: (1) the C-NFs without beads failed to form a closely interconnected 3D framework at lower concentration of α-cellulose/PEO; (2) the C-NFs could tangle and form bundles when the concentrations of electrospinning precursor solutions were increased to a high value.

Fig.1 The SEM images of (a) C-NFs and (b) C-CNFs; the comparison of (c) FTIR spectra, (d) XRD pattern, (e) Raman spectra and (f) electrical conductivity of C-CNFs at different carbonization temperatures. The insets of (a) and (b) are the closer observation under a higher magnification factor.

Full size|PPT slide

To observe the effect of different carbonization temperatures on the chemical structure of C-CNFs, the FTIR spectra were displayed in Fig.1(c). With the carbonization temperature increasing, the intensity of the stretching vibration absorption peak of the –OH (3400 cm^–1), the C–H (2900 cm^–1) and C=O stretching vibration peak (1061 cm^–1) [26,27], gradually weakened or even disappeared, indicating that the more complete the removal of hydrogen and oxygen in the molecular chain with the higher carbonization temperature. The XRD pattern of C-CNFs was employed to explore the structural changes at various carbonization temperatures. From Fig.1(d), two broad diffraction peaks around 30° and 42.5° belonged to amorphous carbon [28], and the intensity of diffraction peak at 30° would be strong with a higher carbonization temperature, confirming a good crystallinity of C-CNFs at 1600 °C. In addition, the Raman spectra of C-CNFs (Fig.1(e)) revealed two obvious D (1351 cm^–1) of disordered carbon structure and G peaks (1599 cm^–1) of graphitic carbon structure, respectively. The degree of graphitization is evaluated via intensity ratio of D to G bands (I_D/I_G), and the value of I_D/I_G also increases from 1.01 to 1.48 with the carbonization temperature from 800 to 1600 °C, which was in keeping with the reported literature [29–31]. Derived from the quantitative elemental contents analyzed by elemental analysis, the carbon contents of C-CNFs were 88.38% (800 °C), 90.74% (1200 °C) and 98.76% (1600 °C), respectively (Table S1, cf. ESM). The electrical conductivities of C-CNFs with different carbonization temperature were characterized by Four-probe Tester, the Fig.1(f) showed that the C-CNFs carbonized at 1600 °C had a higher conductivity (31.2 S·cm^–1) than those of other C-CNFs (5.8 S·cm^–1, 800 °C; 20.5 S·cm^–1, 1200 °C). Especially, the conductivity of C-CNFs carbonized at 1600 °C was 1.69 times of P-CNFs carbonized at 1600 °C (18.5 S·cm^–1) (Fig. S2, cf. ESM), further indicating C-CNFs could be used as a potential carbon matrix for electrode materials.

The C-CNFs prepared by 1600 °C were used as carbon substrates to combine with NiCo₂O₄ for supercapacitors. The SEM and (HR)TEM images of C-CNFs/NiCo₂O₄ nanoflake arrays were illustrated in Fig.2. The NiCo₂O₄ nanoflake arrays uniformly covered with the surface of interconnected C-CNFs, forming highly ordered nanoflake arrays with the thickness of the average diameter of 190 nm (Fig.2(a) and 2(b) and inset). Apparently, there was no obvious morphology change of C-CNFs/NiCo₂O₄ nanoflake arrays before and after calcination. The stability of nanoflake arrays can strengthen the contact area between electrolyte and electrodes, accumulate the penetration rate of ions, and boost the process of redox reaction. For compassion, the SEM images of P-CNFs/NiCo₂O₄ nanoflake arrays and bare NiCo₂O₄ were shown in Figs. S3 and S4 (cf. ESM), respectively. The morphology of the P-CNFs/NiCo₂O₄ nanoflake arrays was similar with C-CNFs/NiCo₂O₄ nanoflake arrays, and the ultrathin nanoflake arrays were uniformly grown on the surface of P-CNFs in Figs. S3(a) and S3(b). Obviously, bare NiCo₂O₄ was composed of irregular flakes in Figs. S4(a) and S4(b). The TEM images in Fig.2(d) illustrated as-prepared C-CNFs/NiCo₂O₄ nanoflake arrays with the C-CNFs inside (ca. 150 nm) and the ultrathin nanoflake arrays (ca. 190–200 nm) on the surface. The HRTEM image in Fig.2(e) revealed two obvious lattice fringes with interplane spacing are 0.148 and 0.204 nm, corresponding to the (440) and (400) crystal planes of spinel NiCo₂O₄, respectively. The TEM and energy-dispersive X-ray spectroscopy (EDS) mapping is characterized in Fig.2(c) and Fig.2(f) to show the elements distribution of Co, Ni, O and C. The C was mainly concentrated in the inner part, corresponding to C-CNFs and the Ni, Co and O are distributed on the outer sides, originating from the surface of NiCo₂O₄ nanoflake arrays, which exhibited uniform elements distribution in the C-CNFs/NiCo₂O₄ electrode. Additionally, the TEM image and EDS mapping of bare NiCo₂O₄ were also illustrated in Figs. S4(c) and S4(d), some thin irregular flakes formed bare NiCo₂O₄ and the Ni, Co and O were evenly distributed within the flakes.

Fig.2 SEM images of (a) C-CNFs/NiCo₂O₄ precursors and inset image is the high magnification of (a) and (b) C-CNFs/NiCo₂O₄ nanoflake arrays and inset image is the high magnification of (b) after calcination. (c), (d) TEM images; (e) HRTEM image; (f) elemental mapping (Co, Ni, O and C) of C-CNFs/NiCo₂O₄ nanoflake arrays.

Full size|PPT slide

The crystal features of the C-CNFs and C-CNFs/NiCo₂O₄ electrode were analyzed by XRD in Fig.3. As shown in Fig.3(a), a typical crystalline peak at 2θ = 26° for C-CNFs and C-CNFs/NiCo₂O₄ nanoflake arrays, which was in accordance with amorphous carbon [28,32]. However, a series of relatively weak diffraction peaks at 2θ values of 31.2, 36.7, 44.6, 59.1 and 65.0 could be ascribed to (220), (311), (400), (511), and (440) plane, respectively, which correspond to standard diffraction peaks of the spinel structure NiCo₂O₄ (JCPDS 20-0781). In Fig.3(b), the Raman spectra clearly illustrated the D (1350 cm^–1) of disordered carbon structure and G peaks (1590 cm^–1) of graphitic carbon structure for two materials, respectively [4]. The Raman peaks at 192.4, 476.2, 527.2 and 652.7 cm^–1 were in good agreement with the F_2g, E_g, F_2g and A_1g vibration modes of Ni–O and Co–O in NiCo₂O₄, respectively. These results were consistent with the results of XRD analysis.

Fig.3 (a) XRD patterns and (b) Raman spectrums of C-CNFs and C-CNFs/NiCo₂O₄ nanoflake arrays; (c) N₂ adsorption–desorption isotherms and pore size distribution curves of C-CNFs/NiCo₂O₄ nanoflake arrays; (d) TG patterns of C-CNFs/NiCo₂O₄ nanoflake arrays, C-CNFs and pure NiCo₂O₄.

Full size|PPT slide

The N₂ adsorption–desorption isotherms and the pore size distribution curves were analyzed in Fig.3(c) and S(5–7) (cf. ESM). The N₂ adsorption–desorption isotherms revealed a typical type IV isotherm with H2-H3 hysteresis loops, confirming that the mesoporous structure of all the as-prepared materials. For carbon matrix comparison, according to the above analysis results, Fig. S5 illustrated the Brunner–Emmett–Teller specific surface area of C-CNFs was 35.61 m²·g^–1, larger than the P-CNFs (12.0 m²·g^–1), indicating that C-CNFs were more suitable as carbon substrate material to combine with NiCo₂O₄. In addition, the specific surface area (85.61 m²·g^–1) of C-CNFs/NiCo₂O₄ nanoflake arrays was 2.86 times that of P-CNFs/NiCo₂O₄ nanoflake arrays (29.98 m²·g^–1 in Fig. S6). From the inset of Fig.3(c), the mesopore size distribution of C-CNFs/NiCo₂O₄ nanoflake arrays mainly concentrated at 3.6 nm, and the main mesopore diameter of P-CNFs/NiCo₂O₄ nanoflake arrays was about 2 nm. Additionally, for the bare NiCo₂O₄ in Fig. S7, the specific surface area was only 18.7 m²·g^–1, which was much lower than that of C-CNFs/NiCo₂O₄ nanoflake arrays. And the mesopore diameter of bare NiCo₂O₄ primarily lied at 2.6 nm. The interconnected meso-porosity of C-CNFs/NiCo₂O₄ nanoflake arrays with higher specific surface area could buffer the volume change and offer more diffusion channels for the fast ion transfer during charge/discharge. The contents of NiCo₂O₄ and carbon were calculated by TG curves under air condition (Fig.3(d) and S8, cf. ESM). And the contents of NiCo₂O₄ in C-CNFs/NiCo₂O₄ and P-CNFs/NiCo₂O₄ nanoflake arrays were 44.8% and 38.6%, respectively. The content of NiCo₂O₄ in C-CNFs/NiCo₂O₄ confirmed that C-CNFs were beneficial to improve the mass loading of active materials and increase the specific capacitance of supercapacitors.

To further analyze the atomic composition and binding energy of C-CNFs/NiCo₂O₄ electrode materials, the XPS spectra was measured in Fig.4. The full XPS survey scan spectra (Fig.4(a)) revealed the existence of Ni, Co, O and C, which was in accordance with the TEM mapping. The high-resolution spectrum of Ni 2p was displayed in Fig.4(b), peaks at 873.3 and 855.8 eV were attributed to the Ni³⁺, while the ones at 871.7 and 853.9 eV were assigned to Ni²⁺ [33]. Two obvious shakeup satellite peaks were located at 872.3 and 855.1 eV. For the Co 2p high-resolution spectrum (Fig.4(c)), there were two doublet peaks for fitting peaks at 796.8 and 781.5 eV, which corresponded to the Co²⁺, and the peaks at 795.3 and 779.4 eV are indexed to Co³⁺ [21]. And the peaks located at 802.9 and 785.7 eV were ascribed to satellite peaks. The O 1s spectra in Fig.4(d) had three characteristic peaks, the binding energies of 529.2, 530.6 and 531.4 eV were assigned to the characteristic peaks of the metal–oxygen bond M–O–M (M = Ni or Co, O₁), the defects with low oxygen coordination in materials (O₂) and the oxygen in the hydroxyl groups within the surface of materials (O₃), respectively [20,34]. In addition, there were also three fitted peaks in the spectrum of C 1s (Fig. S9, cf. ESM), C=C–C (284.0 eV), C–O (284.7 eV) and O–C=O (287.7 eV). The results confirmed that the elements of Ni and Co possessed rich valence states (Ni²⁺/Ni³⁺and Co²⁺/Co³⁺), which might be conducive to enhance the electrochemical performance of supercapacitors.

Fig.4 XPS spectra of C-CNFs/NiCo₂O₄ nanoflake arrays: (a) the survey scan, (b) Ni 2p, (c) Co 2p, (d) O 1s.

Full size|PPT slide

3.2 Electrochemical performance

The CV measurements of C-CNFs/NiCo₂O₄ and P-CNFs/NiCo₂O₄ nanoflake arrays were tested in a three-electrodes using 2 mol·L^–1 KOH aqueous electrodes at a potential window of –0.1–0.6 V from 5 to 50 mV·s^–1 in Fig.5(a) and S10 (cf. ESM). As illustrated in Fig.5(a) and S10, a pair of well-defined redox peaks within –0.1 to 0.6 V can be clearly observed, which is mainly ascribed to the faradaic reactions. And the shape of CV curves maintained unchanged and the peaks of CV kept similar position with the scanning rates increasing, indicating the rapid redox reactions and good reversibility due to the 3D fiber networks facilitating the fast transfer of electron [4,35,36]. In addition, the integrated area of the CV curves (C-CNFs/NiCo₂O₄) was higher that of the P-CNFs/NiCo₂O₄ at 5 mV·s^–1, which showed the C-CNFs/NiCo₂O₄ possessed a specific capacitance. The reason may be that higher contact surface between electrodes and electrolytes, providing more active sites owing to the larger specific surface area of C-CNFs/NiCo₂O₄ (Fig.5(b)). And the peaks at 0.3 and 0.43 V were in keep with the revisable reactions of Co³⁺/Co⁴⁺ and Ni²⁺/Ni³⁺ [36]. The redox reactions were described by the following equations [4,37]:

Fig.5 (a) The CV curves of C-CNFs/NiCo₂O₄ nanoflake arrays at various scan rates. (b) The comparison of CV curves at 5 mV·s^–1 and (c) logarithm relationship between anodic/cathodic peaks current and scan rate of C-CNFs/NiCo₂O₄ and P-CNFs/NiCo₂O₄ nanoflake arrays. (d) The capacitive contribution at 5 mV·s^–1 and (e) contribution comparison at different scan rates of C-CNFs/NiCo₂O₄ nanoflake arrays. (f) Linear relationship of the anodic/cathodic peaks current (i_p) and the square root of the scan rate (v^1/2) for C-CNFs/NiCo₂O₄ and P-CNFs/NiCo₂O₄ nanoflake arrays.

Full size|PPT slide

(5)\mathrm{N}\mathrm{i}\mathrm{C}{\mathrm{o}}_2{\mathrm{O}}_4+\mathrm{O}{\mathrm{H}}^{-}+{\mathrm{H}}_2\mathrm{O}=\mathrm{N}\mathrm{i}\mathrm{O}\mathrm{O}\mathrm{H}+2\mathrm{C}\mathrm{o}\mathrm{O}\mathrm{O}\mathrm{H}+{\mathrm{e}}^{-}

(6)2\mathrm{C}\mathrm{o}\mathrm{O}\mathrm{O}\mathrm{H}+\mathrm{O}{\mathrm{H}}^{-}=2\mathrm{C}\mathrm{o}{\mathrm{O}}_2+{\mathrm{H}}_2\mathrm{O}+{\mathrm{e}}^{-}

To further analyze the electrochemical kinetics of electrode materials. The relationship between the redox peaks current (i) and various scan rates (v) was measured by the following formula [38]:

(7)i=a{v}^b\rightrightarrows \mathrm{l}\mathrm{o}\mathrm{g}i=\mathrm{l}\mathrm{o}\mathrm{g}a+b\mathrm{l}\mathrm{o}\mathrm{g}v.

Here the values of a and b were constants. The b value is between 0.5 and 1, which was calculated by the slop of logi versus logv plot. As we all know, the value of b approaches 0.5, indicating the diffusion-controlled process. And the b value is closed to 1, which reveals the capacitance-controlled mechanism. Fig.5(c) depicted that the calculated b values of C-CNFs/NiCo₂O₄ nanoflake arrays for anodic and cathodic peaks were 0.97 and 0.90, which were higher than that of 0.57 (anodic peak) and 0.67 (cathodic peak) for P-CNFs/NiCo₂O₄ nanoflake arrays, respectively. And the total capacitances of C-CNFs/NiCo₂O₄ electrode were predominant by the capacitive-controlled reactions, revealing that the ultrathin nanoflake arrays and C-CNFs allowed penetration ions of electrolytes more easily, accumulating fast kinetics process of ions transfers, which contributed to the charge/discharge process at high current densities.

Moreover, to obtain the ratios of capacitive contribution in the total capacitance, the current response i(V) at constant potential was divided into capacitance-controlled effects (k₁v) and diffusion-dominated process (k₂v^1/2) based on the following equation:

(8)i\left(V\right)=k_{1}v+k_2{v}^{1/2}\rightrightarrows i\left(V\right)/v^{1/2}=k_1{v}^{1/2}+k_2,

where i(V) and v were current at constant potential and scan rates, respectively. And the values of k₁ and k₂ could be calculated by the slop and intercept of the i(V)/v^1/2 vs. v^1/2 plot, respectively. By fitting the CV curves at various sweep rates at a fixed potential, the contribution of the capacitance-controlled to the entire current can be obtained. Fig.5(d) exhibited the capacitive contribution of C-CNFs/NiCo₂O₄ nanoflake arrays accounted for 62.2% at 5 mV·s^–1, confirming that the electrochemical reaction was indeed controlled by the capacitive process. To compare the changes in capacitive contribution at various scan rates, Fig.5(e) and S11 (cf. ESM) revealed capacitive process at higher scan rate. It could be seen that the proportion of the capacitive effect was positively related with the scan rates, and it even reached as high as 86.7% at a high scan rate of 50 mV·s^–1. Compared with C-CNFs/NiCo₂O₄ nanoflake arrays, all the capacitive contribution ratios of P-CNFs/NiCo₂O₄ nanoflake arrays at various scan rates were less than that of them (Fig. S12, cf. ESM), suggesting that C-CNFs/NiCo₂O₄ nanoflake arrays possessed fast reactions kinetics process. Such fast charge/discharge mechanisms were ascribed to ultrathin nanosheets of NiCo₂O₄ and continuous conductive network structure of C-CNFs with high specific surface areas, which provided more electrochemical active contact area with electrolyte. Additionally, the capacitive contributions, CV curves and b-value at various scan rates ranging from 1 to 5 mV·s^–1 were also displayed in Fig. S13 (cf. ESM). A pair of well-defined redox peaks within –0.1 to 0.6 V was clearly observed (Fig. S13(a)), which was consistent with the results of Fig.5(a). The calculated b-values were 0.98 (anodic peaks) and 0.92 (cathodic peaks), and the proportion of the capacitive effect was positively related with the scan rates and it can reach as high as 68.1% at 5 mV·s^–1 (Figs. S13(b) and S13(c)), further confirming that zinc energy storage mechanism in C-CNFs/NiCo₂O₄ nanoflake arrays was capacitance-controlled during the charging/discharging process.

To further confirm the fast ion transport kinetics, the linear relationship of peak current (i_p) and square root of scan rate (v^1/2) is shown in Fig.5(f), and the ion diffusion coefficient (D) could be calculated based on the Randles-Sevcik equation [39,40]:

(9)i_{\mathrm{p}}=\left(2.69\times {10}^5\right)n^{3/2}\times S\times D^{1/2}\times C\times v^{1/2}.

Therein, n is the number of electrons per molecule attending the electronic transfer reaction, S is the area of the electrodes, C is the concentration of ion in the electrodes. Obviously, the ion diffusion coefficient of C-CNFs/NiCo₂O₄ electrodes (1.33 × 10^–9 cm²·s^–1 for anode and 9.64 × 10^–10 cm²·s^–1 for cathode) is much higher than the P-CNFs/NiCo₂O₄ (2.44 × 10^–11 cm²·s^–1 for anode and 4.35 × 10^–11 cm²·s^–1 for cathode), proving that the higher ion diffusion kinetics of C-CNFs/NiCo₂O₄ nanoflake arrays, which may benefit from the higher effective contact area between high-density ultrathin NiCo₂O₄ nanoflake arrays anchored on C-CNFs and electrolyte. And this result is beneficial to the superior high-rate cycling stability.

To quantify the capacitive performance of various electrodes, the GCD curves were carried out at 1–10 A·g^–1 in Fig.6, S14 and S15 (cf. ESM). Obviously, all the GCD curves had obvious charging and discharging platform in Fig.6(a) and Fig.6(b), S14(a) and S15(a), meaning that the Faradaic redox reactions, which corresponded to CV results. The C-CNFs/NiCo₂O₄ nanoflake arrays showed a long discharge time (Fig.6(b)) and a high specific capacitance of 1010 F·g^–1 at 1 A·g^–1, which was larger than those capacitances of P-CNFs/NiCo₂O₄ nanoflake arrays (976 F·g^–1, Fig.6(c)) and bare NiCo₂O₄ (966.7 F·g^–1, Fig. S15(b)). Meanwhile, even at a high current density of 10 A·g^–1, the specific capacitance of C-CNFs/NiCo₂O₄ nanoflake arrays still maintained 666.7 F·g^–1. Moreover, the electrochemical performance of bare C-CNFs was also illustrated in Fig. S16 (cf. ESM), it showed a small specific capacitance of 174.4 F·g^–1 at 1 A·g^–1, only accounting for 17.3% of the specific capacitance of the C-CNFs/NiCo₂O₄ nanoflake arrays. Consequently, the NiCo₂O₄ contributed most of the specific capacity in composite electrodes instead of C-CNFs as a good conductive carbon matrix. From Fig.6(d), the long-term stability of C-CNFs/NiCo₂O₄ and P-CNFs/NiCo₂O₄ nanoflake arrays was established, and the retention of capacitance was 90.84% and 76.33% at 10 A·g^–1 after 3000 cycles, respectively. However, the capacitance retention of bare NiCo₂O₄ was only 56.4% at 10 A·g^–1 after 1000 cycles in Fig. S15(c). The results revealed that the C-CNFs/NiCo₂O₄ electrodes had best cycling stability in all electrode materials. The reason of results might be that both high conductivity and specific surface area of C-CNFs were larger than that value of P-CNFs. On one hand, compared with P-CNFs, such high capacity was due to a larger mass of NiCo₂O₄ nanoflake arrays anchoring to the C-CNFs; on the other hand, high conductivity of electrodes could accelerate the rapid transfer of electrons and specific surface area of electrodes with multiple mesopores also could promote the permeation and reservoir of the electrolyte ions, which could improve capacity and ensure stable cycles during the fast charging/discharging process [22].

Fig.6 (a) The GCD curves of C-CNFs/NiCo₂O₄ nanoflake arrays at various current densities; (b) the comparison of GCD curves at a current density of 1 A·g^–1; (c) specific capacitance at various current densities; (d) long-cycling performances and (e) Nyquist plots of C-CNFs/NiCo₂O₄ and P-CNFs/NiCo₂O₄ nanoflake arrays (inset were the Nyquist plots fitted and the equivalent circuit diagram); (f) the schematic of ion and electron transfer of C-CNFs/NiCo₂O₄ nanoflake arrays.

Full size|PPT slide

To further study the transport kinetics of the electrochemical reaction process of electrode materials, the EIS test was conducted at open circuit potential in the frequency range between 0.01 Hz and 100 kHz. As shown in Fig.6(e), the curves were divided into two parts: a semicircle in the high-frequency region and a straight line in the low-frequency area [41]. From the inset of Fig.6(e), Nyquist plots fitted were carried out by the equivalent circuit (Fig. S17, cf. ESM) for C-CNFs/NiCo₂O₄ and P-CNFs/NiCo₂O₄ nanoflake arrays. The R_s was regarded as internal resistance and R_ct represented the charge transfer resistance at high-frequency region. Compared with the P-CNFs/NiCo₂O₄ nanoflake arrays (R_s: 3.48 Ω and R_ct: 0.2 Ω), the C-CNFs/NiCo₂O₄ nanoflake arrays showed smaller values of R_s (2.93 Ω) and R_ct (0.14 Ω), which was beneficial with electrochemical performance owing to fast ion diffusion under low resistance. Moreover, the W₀ was in accordance with Warburg impedance, depending on the kinetics of ion diffusion. Therefore, the value of ion diffusion coefficient was calculated based on the Warburg equation [42]. The ion diffusion coefficient value was 3.45 × 10^–10 cm²·s^–1 (C-CNFs/NiCo₂O₄ nanoflake arrays), which was also higher than the value of 8.75 × 10^–11 cm²·s^–1 (P-CNFs/NiCo₂O₄ nanoflake arrays), ensuring the fast transfer and diffusion of ion in C-CNFs/NiCo₂O₄ nanoflake arrays. In addition, to analyze the ion and electron transfer process of C-CNFs/NiCo₂O₄ nanoflake arrays, the schematic diagram of mechanism was illustrated in Fig.6(f). The C-CNFs was used as current collector to transfer electrons, the effective contact between nanoflake arrays NiCo₂O₄ and electrolyte could be used to transport ions (OH^–). The unique 3D nanofiber networks structure of C-CNFs/NiCo₂O₄ nanoflake arrays could effectively improve the electric contact between the C-CNFs and the NiCo₂O₄, enhancing the electrochemical performance of C-CNFs/NiCo₂O₄ nanoflake arrays.

To explore the practical application of C-CNFs/NiCo₂O₄ nanoflake arrays electrodes, the ASSCs were fabricated by using two pieces of C-CNFs/NiCo₂O₄ electrodes (2 cm × 1 cm) and all the electrochemical performances were displayed in Fig.7. As shown in Fig.7(a), the CV curves of ASSCs were performed at 50 mV·s^–1 between 0.6 and 1.35 V, while the CV shape of ASSCs deformed due to the serious polarization at a voltage widow of 1.35 V. Thus, the optimized operating voltage of ASSCs was 1.2 V. In addition, the CV curves of ASSCs were analyzed at various scan rates (from 5 to 80 mV·s^–1) within 0–1.2 V (Fig.7(b)). The symmetry and rectangular shape of CV curves still could be maintained even at a high scanning rate of 80 mV·s^–1, confirming good rate performance and reversibility of the assembled ASSCs. From Fig.7(c), the GCD profiles of ASSCs under various current densities from 1 to 10 A·g^–1, showed that the specific capacitances were 310.2, 232, 217, 177.5 and 146.7 F·g^–1 according to Eq. (2), respectively. The cycling performance of ASSCs was displayed in Fig.7(d), and it maintained the retention of 78.8% at 5 A·g^–1 after 1000 cycles. Fig.7(e) represented the Ragone plot of ASSCs device in comparison with previous works based on Eqs. (3) and (4) from the GCD curves. The obtained ASSCs device exhibited a maximum energy density of 62 W·h·kg^–1 at a power density of 1200 W·kg^–1, which were much larger than previously reported NiCo₂O₄ materials [4,43–49]. Moreover, the ‘DHU’ logo consisted of 36 LEDs could be lighted from six ASSCs devices in series (Fig.7(f)), suggesting the application potential of electrodes in energy storage.

Fig.7 (a) CV curves of ASSCs with different operating voltages at 50 mV·s^–1; (b) CV curves of ASSCs at different scanning rates from 5 to 80 mV·s^–1; (c) GCD profiles of ASSCs at various current densities from 1 to 10 A·g^–1; (d) cycling performance of ASSCs at 5 A·g^–1; (e) Ragone plot of ASSCs our work in comparison with other reported ones; (f) the photographs LED of “DHU” logo on a breadboard powered by ASSCs devices.

Full size|PPT slide

4 Conclusions

In summary, a novel matrix material of C-CNFs is successfully prepared by a special design of electrospinning, subsequent pre-oxidation and carbonization process. Then both C-CNFs/NiCo₂O₄ and P-CNFs/NiCo₂O₄ nanoflake arrays are synthesized via a simple one-step solvothermal and calcination method. The electrochemical test results depict that C-CNFs/NiCo₂O₄ possesses a better cycling stability (90.84% capacitance retention) than that of P-CNFs/NiCo₂O₄ electrodes (76.33% capacitance retention) at a high current density of 10 A·g^–1 after 3000 cycles, suggesting the superiority of C-CNFs as a substrate in comparison with P-CNFs. The as-prepared ASSCs devices assembled from C-CNFs/NiCo₂O₄ can imply a large energy density of 62 W·h·kg^–1 at a power density of 1200 W·kg^–1, and six ASSCs in series can power a ‘DHU’ logo consisting of 36 LEDs, demonstrating the potential application of as-fabricated C-CNFs/NiCo₂O₄ electrodes in energy storage systems.

Acknowledgements

This work was supported by Key Support Project of State Key Laboratory for Modification of Chemical Fibers and Polymer Materials (Grant No. 21M1060212), Open Project of Shanghai Key Laboratory of Lightweight Structural Composite Materials (Grant No. 2232019A4-02), National Natural Science Foundation of China (Grant No. 51503086) and the Fundamental Research Funds for the Central Universities and Graduate Student Innovation Fund of Donghua University (Grant No. CUSF-DH-D-2022013).

Electronic Supplementary Material

Supplementary material is available in the online version of this article at https://dx.doi.org/10.1007/s11705-022-2268-6 and is accessible for authorized users.

References

1 Sun M, Tie J, Cheng G, Lin T, Peng S, Deng F, Ye F, Yu L. In situ growth of burl-like nickel cobalt sulfide on carbon fibers as high-performance supercapacitors. Journal of Materials Chemistry A: Materials for Energy and Sustainability, 2015, 3(4): 1730–1736

2 Han X, Chen Q, Zhang H, Ni Y, Zhang L. Template synthesis of NiCo₂S₄/Co₉S₈ hollow spheres for high-performance asymmetric supercapacitors. Chemical Engineering Journal, 2019, 368: 513–524

3 Wu X, Han Z, Zheng X, Yao S, Yang X, Zhai T. Core-shell structured Co₃O₄@NiCo₂O₄ electrodes grown on flexible carbon fibers with superior electrochemical properties. Nano Energy, 2017, 31: 410–417

4 Zhao D, Hu F, Umar A, Wu X. NiCo₂O₄ nanowire based flexible electrode materials for asymmetric supercapacitors. New Journal of Chemistry, 2018, 42(9): 7399–7406

5 Deng L, Young R J, Kinloch I A, Abdelkader A M, Holmes S M, De Haro-Del Rio D A, Eichhorn S J. Supercapacitance from cellulose and carbon nanotube nanocomposite fibers. ACS Applied Materials & Interfaces, 2013, 5(20): 9983–9990

6 El-Kady M F, Shao Y, Kaner R B. Graphene for batteries, supercapacitors and beyond. Nature Reviews Materials, 2016, 1(7): 1–14

7 Yin J, Zhang W, Alhebshi N A, Salah N, Alshareef H N. Synthesis strategies of porous carbon for supercapacitor applications. Small Methods, 2020, 4(3): 1900853

8 Zhang Z, Bao F, Zhang Y, Feng L, Ji Y, Zhang H, Sun Q, Feng S, Zhao X, Liu X. Formation of hierarchical CoMoO₄@MnO₂ core–shell nanosheet arrays on nickel foam with markedly enhanced pseudocapacitive properties. Journal of Power Sources, 2015, 296: 162–168

9 Zhang P, Zhou J, Chen W, Zhao Y, Mu X, Zhang Z, Pan X, Xie E. Constructing highly-efficient electron transport channels in the 3D electrode materials for high-rate supercapacitors: the case of NiCo₂O₄@NiMoO₄ hierarchical nanostructures. Chemical Engineering Journal, 2017, 307: 687–695

10 Dang A, Sun Y, Liu Y, Xia Y, Liu X, Gao Y, Wu S, Li T, Zada A, Ye F. Flexible Ti₃C₂T_x/carbon nanotubes/CuS film electrodes based on a dual-structural design for high-performance all-solid-state supercapacitors. ACS Applied Energy Materials, 2022, 5(7): 9158–9172

11 Li Y, An F, Wu H, Zhu S, Lin C, Xia M, Xue K, Zhang D, Lian K A. NiCo₂S₄/hierarchical porous carbon for high performance asymmetrical supercapacitor. Journal of Power Sources, 2019, 427: 138–144

12 Kumar K S, Choudhary N, Jung Y, Thomas J. Recent advances in two-dimensional nanomaterials for supercapacitor electrode applications. ACS Energy Letters, 2018, 3(2): 482–495

13 Wan L, Wang Y, Zhang Y, Du C, Chen J, Tian Z, Xie M. FeCoP nanosheets@Ni-Co carbonate hydroxide nanoneedles as free-standing electrode material for hybrid supercapacitors. Chemical Engineering Journal, 2021, 415: 128995

14 Zhou Q, Fan T, Li Y, Chen D, Liu S, Li X. Hollow-structure NiCo hydroxide/carbon nanotube composite for high-performance supercapacitors. Journal of Power Sources, 2019, 426: 111–115

15 Shi Z, Shen X, Zhang Z, Wang X, Gao N, Xu Z, Chen X, Liu X. Hierarchically urchin-like hollow NiCo₂S₄ prepared by a facile template-free method for high-performance supercapacitors. Journal of Colloid and Interface Science, 2021, 604: 292–300

16 Kunwar R, Krishnan S G, Misnon I I, Zabihi F, Yang S, Yang C C, Jose R. Transformation of supercapacitive charge storage behaviour in a multi-elemental spinel CuMn₂O₄ nanofibers with alkaline and neutral electrolytes. Advanced Fiber Materials, 2021, 3(4): 265–274

17 Cai J, Niu H, Li Z, Du Y, Cizek P, Xie Z, Xiong H, Lin T. High-performance supercapacitor electrode materials from cellulose-derived carbon nanofibers. ACS Applied Materials & Interfaces, 2015, 7(27): 14946–14953

18 Cai J, Niu H, Wang H, Shao H, Fang J, He J, Xiong H, Ma C, Lin T. High-performance supercapacitor electrode from cellulose-derived, inter-bonded carbon nanofibers. Journal of Power Sources, 2016, 324: 302–308

19 Frenot A, Henriksson M W, Walkenström P. Electrospinning of cellulose-based nanofibers. Journal of Applied Polymer Science, 2007, 103(3): 1473–1482

20 Kuzmenko V, Naboka O, Gatenholm P, Enoksson P. Ammonium chloride promoted synthesis of carbon nanofibers from electrospun cellulose acetate. Carbon, 2014, 67: 694–703

21 Bhagwan J, Nagaraju G, Ramulu B, Sekhar S C, Yu J S. Rapid synthesis of hexagonal NiCo₂O₄ nanostructures for high-performance asymmetric supercapacitors. Electrochimica Acta, 2019, 299: 509–517

22 BoopathiRaja R, Parthibavarman M. Desert rose like heterostructure of NiCo₂O₄/NF@PPy composite has high stability and excellent electrochemical performance for asymmetric super capacitor application. Electrochimica Acta, 2020, 346: 136270

23 Wei S, Wan C, Zhang L, Liu X, Tian W, Su J, Cheng W, Wu Y. N-doped and oxygen vacancy-rich NiCo₂O₄ nanograss for supercapacitor electrode. Chemical Engineering Journal, 2022, 429: 132242

24 Wang S, Li L, He W, Shao Y, Li Y, Wu Y, Hao X. Oxygen vacancy modulation of bimetallic oxynitride anodes toward advanced Li-ion capacitors. Advanced Functional Materials, 2020, 30(27): 2000350

25 Ni X, Chen H, Liu C, Zeng F, Yu H, Ju A. A freestanding nitrogen-doped carbon nanofiber/MoS₂ nanoflowers with expanded interlayer for long cycle-life lithium-ion batteries. Journal of Alloys and Compounds, 2020, 818: 152835

26 Altin Y, Celik Bedeloglu A. Polyacrylonitrile/polyvinyl alcohol-based porous carbon nanofiber electrodes for supercapacitor applications. International Journal of Energy Research, 2021, 45(11): 16497–16510

27 Hanna S, Yehia A, Ismail M, Khalaf A. Preparation and characterization of carbon fibers from polyacrylonitrile precursors. Journal of Applied Polymer Science, 2012, 123(4): 2074–2083

28 Ni X, Jiang Y, Chen H, Li K, Chen H, Wu Q, Ju A. Fabrication of 3D ordered needle-like polyaniline@hollow carbon nanofibers composites for flexible supercapacitors. Chinese Chemical Letters, 2021, 32(8): 2448–2452

29 Zhang Z, Li L, Qing Y, Lu X, Wu Y, Yan N, Yang W. Manipulation of nanoplate structures in carbonized cellulose nanofibril aerogel for high-performance supercapacitor. Journal of Physical Chemistry C, 2019, 123(38): 23374–23381

30 Wang Y, Ren H, Cheng C, Xu C, Fan G, Liu Y. Tailorable negative permittivity of carbon materials derived from microcrystalline cellulose at different carbonizing temperature. ECS Journal of Solid State Science and Technology, 2020, 9(8): 083001

31 Qiu B, Wang Y, Sun D, Wang Q, Zhang X, Weeks B L, O’Connor R, Huang X, Wei S, Guo Z. Cr(VI) removal by magnetic carbon nanocomposites derived from cellulose at different carbonization temperatures. Journal of Materials Chemistry A: Materials for Energy and Sustainability, 2015, 3(18): 9817–9825

32 Ni X, Cui Z, Luo H, Chen H, Liu C, Wu Q, Ju A. Hollow multi-nanochannel carbon nanofibers@MoSe₂ nanosheets composite as flexible anodes for high performance lithium-ion batteries. Chemical Engineering Journal, 2021, 404: 126249

33 Zhang X, Yang F, Chen H, Wang K, Chen J, Wang Y, Song S. In situ growth of 2D ultrathin NiCo₂O₄ nanosheet arrays on Ni foam for high performance and flexible solid-state supercapacitors. Small, 2020, 16(44): 2004188

34 Yin X, Li H, Yuan R, Lu J. Metal-organic framework derived hierarchical NiCo₂O₄ triangle nanosheet arrays@SiC nanowires network/carbon cloth for flexible hybrid supercapacitors. Journal of Materials Science and Technology, 2021, 81: 162–174

35 Lei D, Li X D, Seo M K, Khil M S, Kim H Y, Kim B S. NiCo₂O₄ nanostructure-decorated PAN/lignin based carbon nanofiber electrodes with excellent cyclability for flexible hybrid supercapacitors. Polymer, 2017, 132: 31–40

36 Deng F, Yu L, Cheng G, Lin T, Sun M, Ye F, Li Y. Synthesis of ultrathin mesoporous NiCo₂O₄ nanosheets on carbon fiber paper as integrated high-performance electrodes for supercapacitors. Journal of Power Sources, 2014, 251: 202–207

37 Liu C, Jiang W, Hu F, Wu X, Xue D. Mesoporous NiCo₂O₄ nanoneedle arrays as supercapacitor electrode materials with excellent cycling stabilities. Inorganic Chemistry Frontiers, 2018, 5(4): 835–843

38 Ni X, Cui Z, Jiang N, Chen H, Wu Q, Ju A, Zhu M. Hollow multi-nanochannel carbon nanofiber/MoS₂ nanoflower composites as binder-free lithium-ion battery anodes with high capacity and ultralong-cycle life at large current density. Journal of Materials Science and Technology, 2021, 77: 169–177

39 Tang S B, Lai M O, Lu L. Li-ion diffusion in highly (003) oriented LiCoO₂ thin film cathode prepared by pulsed laser deposition. Journal of Alloys and Compounds, 2008, 449(1): 300–303

40 Rui X H, Yesibolati N, Li S R, Yuan C C, Chen C H. Determination of the chemical diffusion coefficient of Li⁺ in intercalation-type Li₃V₂(PO₄)₃ anode material. Solid State Ionics, 2011, 187(1): 58–63

41 Khan A S, Pan L, Farid A, Javid M, Huang H, Zhao Y. Carbon nanocoils decorated with a porous NiCo₂O₄ nanosheet array as a highly efficient electrode for supercapacitors. Nanoscale, 2021, 13(27): 11943–11952

42 Zhao X, Jia W, Wu X, Lv Y, Qiu J, Guo J, Wang X, Jia D, Yan J, Wu D. Ultrafine MoO₃ anchored in coal-based carbon nanofibers as anode for advanced lithium-ion batteries. Carbon, 2020, 156: 445–452

43 Zhang J N, Liu P, Jin C, Jin L N, Bian S W, Zhu Q, Wang B. Flexible three-dimensional carbon cloth/carbon fibers/NiCo₂O₄ composite electrode materials for high-performance all-solid-state electrochemical capacitors. Electrochimica Acta, 2017, 256: 90–99

44 Cheng J P, Wang B Q, Gong S H, Wang X C, Sun Q S, Liu F. Conformal coatings of NiCo₂O₄ nanoparticles and nanosheets on carbon nanotubes for supercapacitor electrodes. Ceramics International, 2021, 47(23): 32727–32735

45 Wang W D, Li X F, Zhang P P, Wang B Q, Gong S H, Wang X C, Liu F, Cheng J P. Preparation of NiCo₂O₄@CoS heterojunction composite as electrodes for high-performance supercapacitors. Journal of Electroanalytical Chemistry, 2021, 891: 115257

46 Yin X, Li H, Yuan R, Lu J. Hierarchical self-supporting sugar gourd-shape MOF-derived NiCo₂O₄ hollow nanocages@SiC nanowires for high-performance flexible hybrid supercapacitors. Journal of Colloid and Interface Science, 2021, 586: 219–232

47 Ko T H, Lei D, Balasubramaniam S, Seo M K, Chung Y S, Kim H Y, Kim B S. Polypyrrole-decorated hierarchical NiCo₂O₄ nanoneedles/carbon fiber papers for flexible high-performance supercapacitor applications. Electrochimica Acta, 2017, 247: 524–534

48 Wang L, Jiao X Y, Liu P, Ouyang Y, Xia X F, Lei W, Hao Q L. Self-template synthesis of yolk-shelled NiCo₂O₄ spheres for enhanced hybrid supercapacitors. Applied Surface Science, 2018, 427: 174–181

49 Kumar D R, Prakasha K R, Prakash A S, Shim J J. Direct growth of honeycomb-like NiCo₂O₄@Ni foam electrode for pouch-type high-performance asymmetric supercapacitor. Journal of Alloys and Compounds, 2020, 836: 155370