CuS–Ni₃S₂ grown in situ from three-dimensional porous bimetallic foam for efficient oxygen evolution

Abstract

The development of efficient and stable electrocatalysts for the oxygen evolution reaction (OER), which is critical for the development of water splitting, has attracted great interest from many researchers. However, there have been few reports exhibiting satisfactory activity and durability under high current density (up to 1000 mA cm⁻²), which are significant for the practical production of water splitting. Herein, a novel electrocatalyst, CuS–Ni₃S₂, in situ grown from porous CuNi alloy supported on nickel foam (CuS–Ni₃S₂/CuNi/NF) was obtained via a facile two-step method, and it has achieved excellent OER performance in alkaline solution even under high current density. The electrocatalyst only needed low overpotentials of 337, 444 and 510 mV to reach the very high current densities of 100, 500, and 1000 mA cm⁻², respectively, and it presented excellent durability at a relatively high current density of 100 mA cm⁻² for 15 h. Despite the current density being normalized by the catalytic electrode's geometrical area, both the mass activity and turnover frequency (TOF) for each active site showed the superior catalytic activity of CuS–Ni₃S₂/CuNi/NF. With the assistance of theoretical calculations, the participation of CuS was demonstrated to contribute to the excellent OER performance, providing a promising composition for designing and synthesizing novel OER catalysts.

---

1. Introduction

The increase in serious environmental concerns coupled with the shortage of fossil fuel has sparked increased research pressure for exploring renewable and eco-friendly energy sources.¹ Electrochemical water splitting is a clean technique for producing high-purity hydrogen, an ideal energy carrier with zero carbon emission and high energy density.² However, its efficiency has been limited by the sluggish oxygen evolution reaction (OER) due to its complex four-electron transition process.³ Simultaneously, the application of the state-of-the-art OER catalysts RuO₂ and IrO₂ has been restricted by their low abundance and exorbitant prices.⁴

Fortunately, non-noble metal-based materials such as transition-metal sulfides,⁵ oxides,⁶ nitrides⁷ and phosphides⁸ have been widely investigated as OER catalysts. Among these catalysts, Ni₃S₂ with attractive electrocatalytic behavior has been widely studied.⁹ Nonetheless, the OER catalytic activity has been further improved by synthesizing composites of Ni₃S₂ and other materials.^10–12 In recent years, Cu has attracted significant interest from researchers for its high earth-abundance, low cost, high conductivity and extensive redox properties. For example, CuO/copper foam,¹³ CuO/FTO,¹⁴ [(TGG⁴⁻)Cu(II)–OH₂]²⁻,¹⁵ and Co-doped CuO nanoarrays¹⁶ have been proved to be efficient electrocatalysts for the OER. As an important Cu-based material, CuS has been confirmed to be an excellent material in various electrochemical applications, such as CuS dendrite biosensor electrodes,¹⁷ hollow CuS nanosphere supercapacitors,¹⁸ and three dimensional (3D) hierarchical CuS microsphere aluminum-ion batteries.¹⁹ Nevertheless, the participation of CuS in OER electrocatalysts has been rarely reported and to the best of our knowledge, the CuS–Ni₃S₂ composite as an OER electrocatalyst has not been reported till date.

Furthermore, the structure and morphology of the electrocatalyst are essential to its OER catalytic activity.²⁰ Generally, designing a self-supporting catalytic electrode with 3D porous structure has the following advantages: (1) the substrate's 3D porous architecture provides a high loading area for the catalyst, which leads to the tremendous exposures of catalytic active sites.^21,22 (2) 3D porous architectures ensure the sufficient infiltration of electrolytes as well as the fast departure of oxygen bubbles, resulting in expedited reaction kinetics.²³ (3) The interconnected conductive 3D skeletons construct plenty of electron transport highways, contributing to the high electron transfer efficiency.²⁴ However, although commercial 3D nickel foam (NF) with low cost has been widely employed for synthesizing various OER catalysts,^7,25,26 the above-mentioned merits can be further enlarged by designing an improved novel NF-based substrate with further increased porosity and loading capacity.

Herein, for the first time, the CuS–Ni₃S₂ composite supported on an improved 3D porous substrate (CuS–Ni₃S₂/CuNi/NF) as a high-performance catalyst for OER was synthesized via a facile two-step method. Compared with pristine NF, the CuNi alloy electrodeposited on NF (CuNi/NF) with the unique hierarchical porous structure and nanodendrite morphology not only allowed a higher mass transfer and a higher loading area but also constructed more delicate electron transport pathway networks. Afterward, the surface of the bimetallic alloy was sulfurized into the CuS–Ni₃S₂ layer via a hydrothermal method. This in situ growth strategy ensures a close contact between the catalytic phase and bimetallic substrate, which is favorable to the fast electron transfer and the stability of the catalytic electrode. Therefore, the novel catalytic electrode CuS–Ni₃S₂/CuNi/NF presented outstanding electrocatalytic performance with a low overpotentials of 337, 444 and 510 mV at very high current densities of 100, 500, 1000 mA cm⁻², respectively, and a strong durability of 15 h at a high current density of 100 mA cm⁻² in 1.0 M KOH solution. As far as we know, most works have focused on the catalytic performance within relative low current density (smaller than 200 mA cm⁻²),^6,13,25 while practical application of water splitting needs electrocatalysts that can steadily achieve high current density at low overpotential. As a consequence, the high activity and durability of CuS–Ni₃S₂/CuNi/NF under high current density is the significant merit of excellent OER catalysts for practical water splitting.

2. Experimental section

2.1 Synthesis of CuNi /NF

First, in order to remove the oxidation layer on the surface, NF was cleaned in acetone, 1 M HCl and deionized water (DI water) successively for 10 min under ultrasonication, then dried by high purity nitrogen. 3D porous CuNi alloy was electrodeposited via the CHI 660e electrochemical workstation. NF (working area: 0.4 cm × 0.3 cm), graphite rod and Ag/AgCl (filled with saturated KCl solution) were used as the working electrode, counter electrode and reference electrode, respectively. The electrodeposition process was performed in an aqueous electrolyte (saturated with N₂) containing CuSO₄·5H₂O (0.01 M), NiSO₄·6H₂O (0.5 M), and H₂SO₄ (1.5 M) at a constant current density of 1.8 A cm⁻² for 90 s. After deposition, the electrode was thoroughly cleaned with DI water and immediately dried by nitrogen. For comparison, Ni/NF was synthesized by a similar process without CuSO₄·5H₂O.

2.2 Synthesis of CuS–Ni₃S₂/CuNi/NF

The surface of the CuNi alloy was sulfurized to CuS–Ni₃S₂ by the hydrothermal process. The obtained CuNi/NF was submerged into the Teflon-lined stainless autoclave containing 30 mL of 0.01 M thiourea solution. In addition, 0.005 M and 0.02 M concentrations were respectively employed for contrast. The autoclave was sealed and maintained at 150 °C for 10 h. After that, the autoclave was cooled down to room temperature naturally. The synthesized CuS–Ni₃S₂/CuNi/NF was repeatedly washed with DI water and ethanol in sequence, and dried in nitrogen. For comparison, Ni/NF was sulfurized by the same method to obtain Ni₃S₂/Ni/NF. The loading masses of CuS–Ni₃S₂/CuNi/NF (16.7 mg cm⁻²), Ni₃S₂/Ni/NF (12.1 mg cm⁻²) and CuNi/NF (11.7 mg cm⁻²) were measured using an analytical balance.

2.3 Characterization

Scanning electron microscope (SEM) images and their corresponding energy dispersive X-ray (EDX) spectra and EDX mapping images were obtained on a Quanta 200 FEI. Transmission electron microscopy (TEM) and high-resolution transmission electron microscopy (HRTEM) images, and their corresponding EDX spectra and EDX mapping images were obtained on an FEI TENAI G2 F20 at 200 kV. EDX Samples for TEM were prepared by ultrasonic vibration in ethanol, and then the suspension was dropped onto carbon-enhanced copper grids and dried in air. X-ray photoelectron spectroscopy (XPS) measurements were performed using a Thermo Fisher K-Alpha instrument and XPS spectra were referenced to the C 1s peak (284.8 eV). X-ray diffraction (XRD) patterns were recorded in the range of 5–100° (2θ) on an X-ray diffractometer (Bruker, AXS D8) with Cu Kα radiation operated at 40 kV and 30 mA.

2.4 Electrochemical measurements

Electrochemical measurements were performed with a standard three-electrode system on an electrochemical workstation CHI 660e in 1.0 M KOH solution. The Hg/HgO (1.0 M KOH) electrode and graphite rod were respectively used as the reference electrode and the counter electrode. The obtained catalytic electrode was directly used as the working electrode. All reported potentials were calibrated to the RHE, and were converted using the equation E_RHE = E_Hg/HgO + 0.098 + 0.059 × pH (1.0 M KOH, pH = 14). The overpotential (η) was calculated using the following equation: η = E_RHE − 1.23 V.

In order to fabricate the IrO₂ decorated working electrodes (abbreviated as IrO₂), IrO₂ was dispersed in a mixture solution of DI water : ethanol : Nafion = 10 : 9 : 1 via sonication to form a homogeneous ink. Then the IrO₂ catalyst ink was drop-casted onto NF with loading mass of 5 mg cm⁻² and 16.7 mg cm⁻², respectively and dried in air at room temperature.

For OER activity, cyclic voltammetry (CV) curves were recorded at the scan rate of 5 mV s⁻¹. All CV curves were corrected for ohmic losses. The ohmic losses that constitute the series resistance (R_s) can be obtained from the electrochemical impedance spectroscopy (EIS) measurements. The correction was performed using the following equation: E_corrected = E − IR_s, where E is the potential and I is the current. The Nyquist plots for the electron transfer efficiency investigation were obtained by the EIS measurement recorded at the overpotential of 0.376 V in the frequency range of 100 kHz to 10 mHz. Chemical double-layer capacitance (C_dl) was acquired through CV curves conducted at the non-faradaic current region ranging from 0.574 to 0.674 V (vs. RHE) at various scan rates of 30, 50, 70, 90, 110, 130 and 150 mV s⁻¹. Long-term durability tests were conducted by employing chronopotentiometric methods at 100 mA cm⁻² for 15 h and cyclic voltammetry with the scan rate of 100 mV s⁻¹ for 5000 cycles from 1.124 to 1.624 V (vs. RHE).

2.5 Theoretical calculations

The generalized gradient approximation (GGA) with the Perdew–Burke–Ernzerhof exchange–correlation functional correction was used to account for the dispersion interactions, and a 500 eV cutoff for the plane-wave basis set was employed to perform all the density functional theory (DFT) computations of the materials within the frame of the Vienna ab initio simulation package (VASP).^27,28 The projector-augmented plane wave (PAW) method was used to describe the electron-ion interactions. 3 × 3 × 1 k-points were employed for the geometric optimization of the bulk Ni₃S₂ and four slab surfaces, of which two layers were fixed. Usually, the scheme developed by Nørskov et al. can be employed to gain insight into the OER, where the OER is assumed to involve four elementary reaction steps:^29,30

H₂O + * → *OH + H⁺ + e⁻ (A)

*HO → *O + H⁺ + e⁻ (B)

*O + H₂O → *OOH + H⁺ + e⁻ (C)

*OOH → * + O₂ + H⁺ + e⁻ (D)

Herein, * and X* represent an adsorption site and an adsorbed X intermediate on the surface, respectively. The free energy of the OER is computed by the following equation:

ΔG = ΔE + ΔZPE − TΔS

The value of ΔE was obtained by the computation of geometrical structures. The values of ΔZPE and ΔS were determined by employing the computed vibrational frequencies and standard tables for the reactants and products in the gas phase. The dependence of enthalpy on temperature is not considered in the calculation. Moreover, an external bias U was imposed on each step by including the -eU term in the computation of reaction free energy. Consequently, the reaction free energy of each step could be expressed as follows:

ΔG_A = E(*OH) − E(*) – E_H2O + 1/2E_H2 + (ΔZPE − TΔS) − eU (1)

ΔG_B = E(*O) − E(*OH) + 1/2E_H2 + (ΔZPE − TΔS) − eU (2)

ΔG_C = E(*OOH) − E(*O) − E_H2O + 1/2E_H2 + (ΔZPE − TΔS) − eU (3)

ΔG_D = E(*) − E(*OOH) + E_O2 + 1/2E_H2 + (ΔZPE − TΔS) − eU (4)

Herein, E (*), E (*OH), E (*O), and E(*OOH) are the computed DFT energies of the pure surface and the adsorbed surfaces with *OH, *O, and *OOH, respectively. E_H2O, E_H2 and E_O2 are the computed energies for the sole H₂O, H₂ and O₂ molecules, respectively. For the total reaction H₂O → 1/2O₂ + H₂, the free-energy change was fixed at the experimental value of 2.46 eV per H₂O molecule. Therefore, the Gibbs free energy of the reaction is 4.92 eV for each part of the oxygen generated. The difference between the corresponding minimum voltage needed for changing all the free-energy steps into downhill and the equilibrium potential (U = 1.23 V) is called the reaction overpotential.

3. Results and discussion

As shown in Scheme 1, the nanodendrite morphology of CuNi/NF was realized by the hydrogen bubbles generated and released from the cathode (NF) during the growth of deposited metal and functioned as a soft template.^20,21 After the electrodeposition process, the silver NF turned reddish brown (photograph inset Scheme 1), which is attributed to the red color of Cu. Afterward, the as-prepared material underwent a sulfurization process by the hydrothermal method and became dark grey. This suggests that the surface layer of the bimetallic alloy was successfully converted into the CuS–Ni₃S₂ composite. Compared with the smooth surface of the pristine NF (Fig. S1†), Fig. S2A† exhibits the rough surface of the electrodeposited CuNi/NF, which uniformly and densely covered with numerous nanodendrites (shown in Fig. S2D† at higher magnifications). Moreover, as shown in Fig. S2B and C,† macropores of ∼50–100 nm in diameter on the plane surface of NF and larger macropores of ∼6–9 μm on the ridge of NF skeleton were uniquely constructed, which demonstrates the significantly increased porosity of the substrate. The EDX mapping images (Fig. S2E†) of a single nanodendrite revealed the uniform distribution of Ni and Cu in the CuNi nanodentrite, which attests to the bimetallic alloy's successful electrodeposition. Moreover, the SEM images of electrodeposited Ni/NF shown in Fig. S3† exhibit the flat layer of deposited Ni, with neither nanodendrite morphology nor 3D porous structures. Therefore, Cu has played an essential role in constructing the unique 3D structure of CuNi/NF with increased porosity, as well as increased loading capacity for the subsequently synthesized catalyst.

Scheme 1 Schematic illustration of the synthesis process of CuS–Ni₃S₂/CuNi/NF. The inset images are digital photographs of (A) NF, (B) CuNi/NF and (C) CuS–Ni₃S₂/CuNi/NF. The electrode's working area is 0.3 × 0.4 cm².

After the sulfurization process, the 3D structure with increased porosity remained intact (Fig. 1A–D). Fig. 1A presents a homogeneous and dense loading of nanoparticles on the substrate. The piled up nanoparticles uniformly constructed macropores of ∼50–100 nm in diameter on the plane surface of NF (Fig. 1B and C), and the more stereoscopic macropores of ∼6–9 μm on the ridge of NF skeleton (Fig. 1D). However, the surface of sulfurized Ni/NF (Fig. S4A–C†) exhibited compactly and uniformly dispersed Ni₃S₂ nanohills on both the plane surface and the ridge of NF with negligibly increased porosity as compared with NF. This illustrates that the CuNi/NF substrate has successfully constructed the sophisticated and porous CuS–Ni₃S₂/CuNi/NF with extremely abundant active site exposure and sufficient space for electrolyte infusion as well as oxygen bubbles release during the OER process.

Fig. 1 (A–D) SEM images of the 3D porous CuS–Ni₃S₂/CuNi/NF at different magnifications. (E) HRTEM and TEM (inset) images of CuS–Ni₃S₂ and (F) the EDX mapping images of the CuS–Ni₃S₂ nanolayer.

CuS–Ni₃S₂ was peeled off from the CuS–Ni₃S₂/CuNi/NF electrode by the strong ultrasonication process, and the TEM image of CuS–Ni₃S₂ (the inset image of Fig. 1E) revealed its nanolayer structure. In the HRTEM images of CuS–Ni₃S₂ (Fig. 1E), four sets of lattice fringes can be observed. According to previous reports, the lattice fringe spacings of 0.20 and 0.30 nm correspond to the (107) and (102) crystal planes of CuS respectively,^31,32 while 0.28 and 0.40 nm correspond to the (110) and (101) crystal planes of Ni₃S₂, respectively.^33,34 EDX mapping images of the CuS–Ni₃S₂ nanolayer (Fig. 1F) and the CuS–Ni₃S₂/CuNi/NF electrode skeleton (Fig. S5†) indicate the homogeneous dispersion of Cu, Ni and S on both the catalyst film and throughout the catalytic electrode.

The composition of CuS–Ni₃S₂/CuNi/NF was further demonstrated by XPS spectra and XRD patterns. The two strong peaks in the Cu 2p region (Fig. 2A) at 932.1 and 952.1 eV, respectively, correspond to Cu 2p_3/2 and Cu 2p_1/2 of Cu²⁺, as previously reported.^17,35 In the Ni 2p spectrum (Fig. 2B), peaks at 855.5 and 873.6 eV correspond to Ni 2p_3/2 and Ni 2p_1/2 of Ni₃S₂, respectively, and peaks at 861.5 and 879.2 eV can be assigned to the shakeup-type peaks.^25,36 It is worth noting that the Ni 2p_3/2 signal of Ni₃S₂ in CuS–Ni₃S₂/CuNi/NF exhibits a positive shift of about 0.1 eV relative to that in Ni₃S₂/Ni/NF, which suggests the existence of electronic interactions between Ni₃S₂ and CuS, and the existence of the coupling interfaces.³⁷ Moreover, the binding energies of S 2p (Fig. 2C) appearing at 161.9 and 163.0 eV can be assigned to S 2p_3/2 and S 2p_1/2, thus indicating the presence of S²⁻.^38,39 The peak at 168.8 eV corresponds to SO₄²⁻, which may have been formed due to the oxidation of the surface of the sulfide composite.⁴⁰

Fig. 2 XPS spectra in (A) Cu 2p, (B) Ni 2p and (C) S 2p regions of CuS–Ni₃S₂/CuNi/NF; (D–F) XRD patterns of CuNi/NF and CuS–Ni₃S₂/CuNi/NF.

As shown in the XRD patterns of CuNi/NF (Fig. 2D), the diffraction peaks at 43.3°, 50.4°, 74.1° and 89.9° can be assigned to Cu (PDF#04-0836) and the diffraction peaks at 44.5°, 51.8°, 76.4°, and 92.9° can be attributed to Ni (PDF#04-0850). In the XRD pattern of CuS–Ni₃S₂/CuNi/NF, the observed diffraction peaks at 27.7°, 31.8°, 32.8°, 59.3° and 88.9° can be assigned to the CuS phase (PDF#06-0464), and the peaks at 21.7°, 31.1°, 37.8°, 50.1°, 55.3°, 86.3° and 90.4° can be attributed to the Ni₃S₂ phase (PDF#44-1418). In comparison, the characteristic peaks of CuNi/NF remain after the sulfurization process, illustrating the existence of the CuNi metal core under the sulfide phase.

In order to explore the effect of thiourea concentration on the catalyst synthesis and the corresponding OER performance, 0.02, 0.01 and 0.005 M thiourea were respectively employed to sulfurize CuNi/NF. The electrochemical OER performances of the different sulfurization products are shown in Fig. S6.† To achieve the current density of 200 mA cm⁻², CuS–Ni₃S₂/CuNi/NF (synthesized by 0.01 M thiourea) needs the lowest overpotential of 380 mV compared with that required for CuS–Ni₃S₂/CuNi/NF-0.02 M (400 mV) and CuS–Ni₃S₂/CuNi/NF-0.005 M (390 mV), suggesting that 0.01 M is the optimal thiourea concentration for the synthesis of the superior OER electrocatalyst. Fig. S7† shows the morphologies of CuS–Ni₃S₂/CuNi/NF-0.02 M and CuS–Ni₃S₂/CuNi/NF-0.005 M. Compared with the above-described morphology of CuS–Ni₃S₂/CuNi/NF (Fig. 1A–D), the pores and cavities on the ridge of the CuS–Ni₃S₂/CuNi/NF-0.02 M skeleton (Fig. S7A–C†) collapsed, while the pores disappeared on the plane surface of CuS–Ni₃S₂/CuNi/NF-0.005 M; both lost plenty of exposed active sites. As a consequence, the superior OER performance of CuS–Ni₃S₂/CuNi/NF has arisen from its unique morphology with high porosity. Hereafter, all CuS–Ni₃S₂/CuNi/NF mentioned throughout the manuscript were obtained via 0.01 M thiourea.

OER activities of NF, CuNi/NF and Ni₃S₂/Ni/NF were measured for comparison. As shown in Fig. 3A, the redox peaks at ∼1.4 V before the oxygen evolution of CuS–Ni₃S₂/CuNi/NF originated from the electron-transfer between Ni²⁺ and Ni³⁺ species.³⁶ CuS–Ni₃S₂/CuNi/NF requires the overpotential of 337 mV to reach the current density of 100 mA cm⁻², which is much lower than that of Ni₃S₂/Ni/NF (465 mV), CuNi/NF (555 mV) and NF (631 mV). It only needs 444 and 510 mV to reach 500 and 1000 mA cm⁻², respectively, which is much smaller than that of the most recently reported non-noble metal-based OER catalysts at such high current densities (Table S1†). Additionally, the reaction kinetics of the above-studied electrodes were revealed by Tafel plots as shown in Fig. 3B. The relatively large Tafel slope, 142 mV dec⁻¹ of CuS–Ni₃S₂/CuNi/NF can be attributed to the high oxidation current peak, which cannot be separated well from the OER catalytic current, as observed in previous reports.^25,41 Nonetheless, the Tafel slope of CuS–Ni₃S₂/CuNi/NF (142 mV dec⁻¹) is significantly smaller than that of Ni₃S₂/Ni/NF (216 mV dec⁻¹), CuNi/NF (210 mV dec⁻¹), and the pure NF (218 mV dec⁻¹), attesting to the much faster OER kinetics of CuS–Ni₃S₂/CuNi/NF.⁴²

Fig. 3 (A) CV curves and the corresponding (B) Tafel plots of CuS–Ni₃S₂/CuNi/NF, Ni₃S₂/Ni/NF, CuNi/NF and NF. (C) The mass activities of CuS–Ni₃S₂/CuNi/NF, Ni₃S₂/Ni/NF and CuNi/NF. (D) Linear fitting of the capacitive currents of the catalysts versus scan rates of CuS–Ni₃S₂/CuNi/NF and Ni₃S₂/Ni/NF. (E) Nyquist plots recorded at an overpotential of 376 mV for CuS–Ni₃S₂/CuNi/NF and Ni₃S₂/Ni/NF. (F) Calculated TOFs of CuS–Ni₃S₂/CuNi/NF and Ni₃S₂/Ni/NF.

To compare the OER activity of CuS–Ni₃S₂/CuNi/NF with the benchmark OER catalyst IrO₂, two IrO₂ decorated electrodes with different loading masses have been further investigated as shown in Fig. S8.† When the loading mass of IrO₂ (16.7 mg cm⁻²) and CuS–Ni₃S₂/CuNi/NF were the same, CuS–Ni₃S₂/CuNi/NF achieved a comparable OER activity with IrO₂, and the catalytic activity of CuS–Ni₃S₂/CuNi/NF even exceeded that of IrO₂ at the high potential range. However, it is too expensive to employ such a high loading mass of the noble metal-based benchmark catalyst IrO₂ in real applications. Thus, the loading mass of IrO₂ was decreased to 5 mg cm⁻², which is still higher than that of IrO₂ in other reports.^43,44 In this case, CuS–Ni₃S₂/CuNi/NF obviously outperforms IrO₂, which confirms its outstanding electrocatalytic behavior. On the other hand, it is worth mentioning that compared with both IrO₂ decorated electrodes, CuS–Ni₃S₂/CuNi/NF can reach very high current densities of 500 and 1000 mA cm⁻² with a steady CV curve, without the serious shaking caused by oxygen release. This may have resulted from the unique 3D structure with increased porosity, which led to a fast oxygen release process.

In order to comprehensively evaluate the intrinsic OER performances of CuS–Ni₃S₂/CuNi/NF, Ni₃S₂/Ni/NF, and CuNi/NF,⁴⁵ their mass activities were examined and are shown in Fig. 3C. To reach the current density of 10 mA mg⁻¹, CuS–Ni₃S₂/CuNi/NF needs a lower overpotential of 368 mV than that of Ni₃S₂/Ni/NF (481 mV) and CuNi/NF (568 mV). Therefore, CuS–Ni₃S₂/CuNi/NF exhibited a superior catalytic activity measured by the current density normalized by both the geometrical area and the loading mass, and the excellent performance of CuS–Ni₃S₂/CuNi/NF may be ascribed to the unique 3D porous structure of the catalytic electrode and the synergetic effect of CuS and Ni₃S₂ in the sulfide composite.

To further attest to the above estimation, the C_dl values of CuS–Ni₃S₂/CuNi/NF and Ni₃S₂/Ni/NF were then evaluated to uncover the actual quantities of catalytic sites of electrocatalysts as the electrochemical surface area (ECSA) is proportional to the C_dl.⁴⁶ CV curves were collected in the non-faradaic current region of 0.574 to 0.674 V in 1.0 M KOH of the scan rate from 30 to 150 mV s⁻¹ (Fig. S9†).⁴⁷ The corresponding C_dl values of CuS–Ni₃S₂/CuNi/NF and Ni₃S₂/Ni/NF electrodes can be calculated by the slope extracted from the linear relationship of the current density against the increasing scan rates. As shown in Fig. 3D, C_dl of CuS–Ni₃S₂/CuNi/NF is 10.99 mF cm⁻², about 7 times that of the C_dl of Ni₃S₂/Ni/NF (1.45 mF cm⁻²), which indicates the much higher ECSA of CuS–Ni₃S₂/CuNi/NF as compared to Ni₃S₂/Ni/NF; this can be ascribed to the following. Firstly, compared with the much flatter morphology of Ni₃S₂/Ni/NF, which may conceal many active sites, the 3D porous morphologies constructed by the piled up nanoparticles of CuS–Ni₃S₂/CuNi/NF have realized the exposure of the numerous active sites. Secondly, the numerous pores of CuS–Ni₃S₂/CuNi/NF helped the electrolyte to infuse into the catalytic electrode and sufficiently contact the exposed active sites, which also contributed to the larger ECSA. Thirdly, coupling the interfaces of the CuS phase and Ni₃S₂ phase may boost more new active sites. Furthermore, the electrode kinetics of the OER was investigated by the Nyquist plots. As shown in Fig. 3E, the charge transfer resistance (R_ct) of CuS–Ni₃S₂/CuNi/NF (3.2 Ω) is much smaller than that of Ni₃S₂/Ni/NF (14.8 Ω), which indicates the fast electron transfer between the active sites on the surface of CuS–Ni₃S₂/CuNi/NF and the electrolyte. This may be attributed to tremendous delicate high-efficiency electronic transmission pathways constructed by the 3D interconnected porous CuNi skeleton of CuS–Ni₃S₂/CuNi/NF.

The calculation of the turnover frequency (TOF) for each active site makes it possible to compare the intrinsic activity of CuS–Ni₃S₂/CuNi/NF with other different catalytic electrodes. The number of active sites was quantified by the electrochemical method, referring to the previous literature.^48,49 Fig. S10† shows the CV curves of CuS–Ni₃S₂/CuNi/NF in the region of 0.25 to 0.85 V at pH 7, and the integrated charge over the whole potential range should be proportional to the total number of active sites. Detailed descriptions of the calculations of the quantity of active sites and the corresponding TOF are presented in Fig. S10.†Fig. 3F presents the CV curves normalized by the quantity of active sites, and expressed in terms of TOF. To obtain the TOF value of 5 s⁻¹, Ni₃S₂/Ni/NF required the overpotential of 534 mV, while CuS–Ni₃S₂/CuNi/NF only needed a much lower overpotential of 425 mV. This demonstrated the superior intrinsic activity of the catalytic active phase, CuS–Ni₃S₂, as compared to the individual Ni₃S₂ phase. The intrinsically enhanced activity may be attributed to the participation of CuS in the CuS–Ni₃S₂ composite.

In order to prove the above experimental phenomenon, theoretical calculations focused on the influence of CuS on the OER catalytic activity of CuS–Ni₃S₂ were conducted, and the functional theory (DFT) was used to calculate the Gibbs free energy (ΔG) of the OER. In the coupling interfaces between CuS and Ni₃S₂, Cu atoms substituting for Ni atoms in the Ni₃S₂ phase were rationally constructed and the top view and side view of the CuS–Ni₃S₂ interface theoretical model are shown in Fig. 4A and B. According to the lowest chemisorption free energy, Ni sites acted as the adsorption active sites of *OH, *O and *OOH intermediates during the OER in both the Ni₃S₂ surface and CuS–Ni₃S₂ interfaces, and the adsorption configurations on the CuS–Ni₃S₂ interface are shown in Fig. 4C–E. Thus, all the reaction steps discussed below took place on Ni sites, and we focused on the effect of the participation of CuS on Ni sites in CuS–Ni₃S₂ interfaces. As shown in Fig. 4F and G, the free energy diagrams of OER for the Ni₃S₂ surface and the CuS–Ni₃S₂ interface exhibited a four-step reaction involving three intermediates of *OH, *O and *OOH. Firstly, when no bias was applied (U = 0 V), all of the steps on the Ni₃S₂ surface and CuS–Ni₃S₂ interface were found to move uphill (ΔG > 0). Until the equilibrium potential (U = 1.23 V) for OER was applied, ΔG_{(*OOH→*+O2)} of Ni₃S₂ became negative, while on the CuS–Ni₃S₂ interface, ΔG_{(*+H2O→*OH)}, ΔG_(*O→*OOH) and ΔG_{(*OOH→*+O2)} simultaneously became negative values. These three spontaneous reaction steps could lead to an easier OER process, and according to the only positive ΔG value for CuS–Ni₃S₂, *OH + OH⁻ → *O + H₂O should be the rate-determining step of CuS–Ni₃S₂ for the OER. Moreover, compared with 0 eV ΔG_{(*+H2O→*OH)} of Ni₃S₂, the negative ΔG_{(*+H2O→*OH)} value −0.02 eV of CuS–Ni₃S₂ implies the more effective adsorption in the first catalytic step, which could easily trigger the subsequent reaction steps. Finally, we employed the minimum bias to change all the steps into downhill (ΔG < 0), and the difference between the above-mentioned minimum bias and the equilibrium potential (U = 1.23 V) for the OER is exactly the onset overpotential of the catalyst for the OER.⁹ With the participation of CuS, the required overpotential value decreased from 380 mV of Ni₃S₂ to 320 mV of CuS–Ni₃S₂, which is consistent with the experimental data shown in Fig. 3A. Moreover, the ΔG_{(*+H2O→*OH)} value is −0.34 eV for CuS–Ni₃S₂, which is much lower than 0 eV ΔG_{(*+H2O→*OH)} of Ni₃S₂, demonstrating the more obviously thermodynamically favored catalytic reaction step with the assistance of CuS. Therefore, the participation of CuS effectively enhanced the Ni sites’ OER catalytic activity of Ni₃S₂ by decreasing both the ΔG of the intermediates and the overpotential needed to drive the OER. Additionally, in order to further explore the influence of the participation of CuS in the electron conductivity of the catalytic material, the density of states (DOS) for CuS–Ni₃S₂ and pristine Ni₃S₂ were calculated. As shown in Fig. S11,† DOS of CuS–Ni₃S₂ and pristine Ni₃S₂ catalyst are continuous around the Fermi level, suggesting their intrinsic metallic characters. Remarkably, the DOS of CuS–Ni₃S₂ near the Fermi level is stronger than that of pristine Ni₃S₂, indicating that the electrical behavior of CuS–Ni₃S₂, especially the concentration of charge carrier and the electronic conductivity, have been further enhanced by the participation of CuS. The improvement of the electron transfer efficiency proved by DOS theoretical calculation is consistent with the experimental data shown in the Nyquist plots in Fig. 3E.

Fig. 4 (A) The top view and (B) side view of the CuS–Ni₃S₂ interface. Side views of the adsorption configurations of (C) *OH, (D) *O and (E) *OOH, respectively. Free-energy diagram for the four steps of the OER on the (F) Ni₃S₂ surface and (G) CuS–Ni₃S₂ interface at the different applied potentials.

The durability is another important criterion for an electrocatalyst for the OER. The long-term stability of the catalytic electrode was firstly explored by 5000 continuous CV cycles. As shown in Fig. 5A, the polarization plot after continuously working for 5000 cycles nearly overlaps with the initial polarization plot, demonstrating the excellent stability of CuS–Ni₃S₂/CuNi/NF. Besides, chronopotentiometry at a relatively high constant current density of 100 mA cm⁻² was carried out and is shown in Fig. 5B. Over a period of 15 h, the potential to attain a high current density of 100 mA cm⁻² exhibited only slight decay. The robustness may be due to the in situ growth and close contact between the sulfide composite and bimetallic substrate, which leads to the stable binding strength between them. During the long-term stability test, steady tiny bubbles were generated and released from the surface of the catalytic electrode during the entire chronopotentiometry process and can be observed from the inset photograph in Fig. 5B and Video S1.† As shown in Fig. S12A and B,† the SEM images of CuS–Ni₃S₂/CuNi/NF after the 15 h duration test illustrate that the morphology of CuS–Ni₃S₂/CuNi/NF remained almost the same after the vigorous OER process in the harsh reaction environment, implying the robustness of this catalytic electrode.

Fig. 5 (A) Linear sweep voltammetry (LSV) curves before and after 5000 continuous CV cycles of CuS–Ni₃S₂/CuNi/NF. (B) Chronopotentiometric measurement of CuS–Ni₃S₂/CuNi/NF at 100 mA cm⁻² for 15 h (inset: a digital photograph of oxygen bubbles’ evolution during this process).

4. Conclusion

In summary, we have successfully synthesized a novel OER catalyst, CuS–Ni₃S₂/CuNi/NF, and explored the positive role CuS has played in this Ni₃S₂-based electrocatalyst during the OER process. This 3D OER catalytic electrode preserved its CuNi/NF porous skeleton both as the high-efficiency electron transport pathway networks and the template for CuS–Ni₃S₂/CuNi/NF. Consequently, the 3D porous morphology of CuS–Ni₃S₂/CuNi/NF realized the maximum exposure of active sites and fast mass transfer during the OER process. Moreover, the in situ growth of the sulfide composite nanolayer from the bimetallic alloy guaranteed the fast electron transfer and excellent stability of the catalytic electrode. Based on the above aspects, the catalytic electrode has achieved excellent OER activity and stability at high current density, even greater than 1000 mA cm⁻². Despite the activity measured by the current density normalized by the geometrical area, CuS–Ni₃S₂/CuNi/NF also exhibited superior mass activity. Moreover, the higher TOF of CuS–Ni₃S₂/CuNi/NF compared with that of Ni₃S₂/Ni/NF indicated the improved intrinsic activity of the catalytic active phase by the participation of CuS. Theoretical calculations have demonstrated that CuS can effectively enhance the Ni sites’ OER catalytic activity of Ni₃S₂, and also improve the electronic conductivity of Ni₃S₂. As a consequence, although CuS has been rarely studied as an OER catalyst, this work has proved that it may be promising for use in the design of novel OER catalysts for further improved catalytic performance.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors are most grateful to the talents start scientific research funds of China University of Petroleum, Beijing (No. 2462016YJRC022).

References

1. B. Chang, S. Hao, Z. X. Ye and Y. C. Yang, Chem. Commun., 2018, 54, 2393–2396.
2. J. X. Feng, S. H. Ye, H. Xu, Y. X. Tong and G. R. Li, Adv. Mater., 2016, 28, 4698–4703.
3. J. X. Feng, H. Xu, Y. T. Dong, S. H. Ye, Y. X. Tong and G. R. Li, Angew. Chem., Int. Ed., 2016, 55, 3694–3698.
4. H. Zhong, X. Cheng, H. Xu, L. Li, D. Li, P. Tang, N. Alonso-Vante and Y. Feng, Electrochim. Acta, 2017, 258, 554–560.
5. S. Fu, J. Song, C. Zhu, G.-L. Xu, K. Amine, C. Sun, X. Li, M. H. Engelhard, D. Du and Y. Lin, Nano Energy, 2018, 44, 319–326.
6. X. Li, X. Du, X. Ma, Z. Wang, X. Hao, A. Abudula, A. Yoshida and G. Guan, Electrochim. Acta, 2017, 250, 77–83.
7. Y. Wang, B. Zhang, W. Pan, H. Ma and J. Zhang, ChemSusChem, 2017, 10, 4170–4177.
8. J. H. Hao, W. S. Yang, Z. P. Huang and C. Zhang, Adv. Mater. Interfaces, 2016, 3, 8.
9. L. L. Feng, G. Yu, Y. Wu, G. D. Li, H. Li, Y. Sun, T. Asefa, W. Chen and X. Zou, J. Am. Chem. Soc., 2015, 137, 14023–14026.
10. X. Zhong, J. Tang, J. Wang, M. Shao, J. Chai, S. Wang, M. Yang, Y. Yang, N. Wang, S. Wang, B. Xu and H. Pan, Electrochim. Acta, 2018, 269, 55–61.
11. X. Du, Z. Yang, Y. Li, Y. Gong and M. Zhao, J. Mater. Chem. A, 2018, 6, 6938–6946.
12. N. Cheng, Q. Liu, A. M. Asiri, W. Xing and X. Sun, J. Mater. Chem. A, 2015, 3, 23207–23212.
13. C. Lu, J. Wang, S. Czioska, H. Dong and Z. Chen, J. Phys. Chem. C, 2017, 121, 25875–25881.
14. X. Liu, S. Cui, Z. Sun, Y. Ren, X. Zhang and P. Du, J. Phys. Chem. C, 2016, 120, 831–840.
15. M. T. Zhang, Z. Chen, P. Kang and T. J. Meyer, J. Am. Chem. Soc., 2013, 135, 2048–2051.
16. X. L. Xiong, C. You, Z. Liu, A. M. Asiri and X. P. Sun, ACS Sustainable Chem. Eng., 2018, 6, 2883–2887.
17. W. B. Kim, S. H. Lee, M. Cho and Y. Lee, Sens. Actuators, B, 2017, 249, 161–167.
18. H. Heydari, S. E. Moosavifard, S. Elyasi and M. Shahraki, Appl. Surf. Sci., 2017, 394, 425–430.
19. S. Wang, S. Jiao, J. Wang, H. S. Chen, D. Tian, H. Lei and D. N. Fang, ACS Nano, 2017, 11, 469–477.
20. B. J. Plowman, L. A. Jones and S. K. Bhargava, Chem. Commun., 2015, 51, 4331–4346.
21. H. Xu, J.-X. Feng, Y.-X. Tong and G.-R. Li, ACS Catal., 2016, 7, 986–991.
22. W. Wang, S. You, X. Gong, D. Qi, B. K. Chandran, L. Bi, F. Cui and X. Chen, Adv. Mater., 2016, 28, 270–275.
23. H. Ashassi-Sorkhabi and P. La'le Badakhshan, Appl. Surf. Sci., 2017, 419, 165–176.
24. L. Wei, K. Goh, O. Birer, H. E. Karahan, J. Chang, S. Zhai, X. Chen and Y. Chen, Nanoscale, 2017, 9, 4401–4408.
25. Y. Gong, Z. Xu, H. Pan, Y. Lin, Z. Yang and X. Du, J. Mater. Chem. A, 2018, 6, 5098–5106.
26. W. J. Zhou, X. J. Wu, X. H. Cao, X. Huang, C. L. Tan, J. Tian, H. Liu, J. Y. Wang and H. Zhang, Energy Environ. Sci., 2013, 6, 2921–2924.
27. A. J. Misquitta, B. Jeziorski and K. Szalewicz, Phys. Rev. Lett., 2003, 91, 033201.
28. S. Grimme, J. Comput. Chem., 2006, 27, 1787–1799.
29. I. C. Man, H.-Y. Su, F. Calle-Vallejo, H. A. Hansen, J. I. Martínez, N. G. Inoglu, J. Kitchin, T. F. Jaramillo, J. K. Nørskov and J. Rossmeisl, ChemCatChem, 2011, 3, 1159–1165.
30. J. Rossmeisl, Z. W. Qu, H. Zhu, G. J. Kroes and J. K. Nørskov, J. Electroanal. Chem., 2007, 607, 83–89.
31. L. Zhang, Y. Guo, A. Iqbal, B. Li, D. Gong, W. Liu, K. Iqbal, W. Liu and W. Qin, Int. J. Hydrogen Energy, 2018, 43, 1251–1260.
32. F. Li, J. Wu, Q. Qin, Z. Li and X. Huang, Powder Technol., 2010, 198, 267–274.
33. H. Liu, X. Ma, Y. Rao, Y. Liu, J. Liu, L. Wang and M. Wu, ACS Appl. Mater. Interfaces, 2018, 10, 10890–10897.
34. N. Tronganh, Y. Gao, W. Jiang, H. Tao, S. Wang, B. Zhao, Y. Jiang, Z. Chen and Z. Jiao, Appl. Surf. Sci., 2018, 439, 386–393.
35. K.-J. Huang, J.-Z. Zhang, Y. Liu and Y.-M. Liu, Int. J. Hydrogen Energy, 2015, 40, 10158–10167.
36. M. Al-Mamun, H. J. Yin, P. R. Liu, X. T. Su, H. M. Zhang, H. G. Yang, D. Wang, Z. Y. Tang, Y. Wang and H. J. Zhao, Nano Res., 2017, 10, 3522–3533.
37. J. Zhang, T. Wang, D. Pohl, B. Rellinghaus, R. H. Dong, S. H. Liu, X. D. Zhuang and X. L. Feng, Angew. Chem., Int. Ed., 2016, 55, 6702–6707.
38. Q. Liu, L. S. Xie, Z. Liu, G. Du, A. M. Asiri and X. P. Sun, Chem. Commun., 2017, 53, 12446–12449.
39. Z. Li and Z. Zhang, Nano Res., 2018, 11, 1530–1540.
40. Y. Yang, K. Zhang, H. Lin, X. Li, H. C. Chan, L. Yang and Q. Gao, ACS Catal., 2017, 7, 2357–2366.
41. S. Qu, J. Huang, J. Yu, G. Chen, W. Hu, M. Yin, R. Zhang, S. Chu and C. Li, ACS Appl. Mater. Interfaces, 2017, 9, 29660–29668.
42. S. Zhao, M. Li, M. Han, D. Xu, J. Yang, Y. Lin, N.-E. Shi, Y. Lu, R. Yang, B. Liu, Z. Dai and J. Bao, Adv. Funct. Mater., 2018, 28, 1706018.
43. S. Anantharaj, K. Karthick, M. Venkatesh, TVSV. Simha, AS. Salunke, L. Ma, H. Liang and S. Kundu, Nano Energy, 2017, 39, 30–43.
44. H. Wang, T. Zhou, P. Li, Z. Cao, W. Xi, Y. Zhao and Y. Ding, ACS Sustainable Chem. Eng., 2017, 6, 380–388.
45. S. Anantharaj, S. R. Ede, K. Sakthikumar, K. Karthick, S. Mishra and S. Kundu, ACS Catal., 2016, 6, 8069–8097.
46. C. Luan, G. Liu, Y. Liu, L. Yu, Y. Wang, Y. Xiao, H. Qiao, X. Dai and X. Zhang, ACS Nano, 2018, 12, 3875–3885.
47. D. Merki, H. Vrubel, L. Rovelli, S. Fierro and X. L. Hu, Chem. Sci., 2012, 3, 2515–2525.
48. D. Merki, S. Fierro, H. Vrubel and X. Hu, Chem. Sci., 2011, 2, 1262–1267.
49. J. Tian, Q. Liu, A. M. Asiri and X. Sun, J. Am. Chem. Soc., 2014, 136, 7587–7590.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c8qi01148f

---