Oxygen-deficient MoO_x/Ni₃S₂ heterostructure grown on nickel foam as efficient and durable self-supported electrocatalysts for hydrogen evolution reaction

Abstract

High-performance and ultra-durable electrocatalysts are vital for hydrogen evolution reaction (HER) during water splitting. Herein, by one-pot solvothermal method, MoO_x/Ni₃S₂ spheres comprising Ni₃S₂ nanoparticles inside and oxygen-deficient amorphous MoO_x outside in situ grow on Ni foam (NF), to assembly the heterostructure composites of MoO_x/Ni₃S₂/NF. By adjusting volume ratio of the solvents of ethanol to water, the optimized MoO_x/Ni₃S₂/NF-11 exhibits the best HER performance, requiring an extremely low overpotential of 76 mV to achieve the current density of 10 mA∙cm^‒2 (η₁₀ = 76 mV) and an ultra-small Tafel slope of 46 mV∙dec^‒1 in 0.5 mol∙L^‒1 H₂SO₄. More importantly, the catalyst shows prominent high catalytic stability for HER (> 100 h). The acid-resistant MoO_x wraps the inside Ni₃S₂/NF to ensure the high stability of the catalyst under acidic conditions. Density functional theory calculations confirm that the existing oxygen vacancy and MoO_x/Ni₃S₂ heterostructure are both beneficial to the reduced Gibbs free energy of hydrogen adsorption (|∆G_H*|) over Mo sites, which act as main active sites. The heterostructure effectively decreases the formation energy of O vacancy, leading to surface reconstruction of the catalyst, further improving HER performance. The MoO_x/Ni₃S₂/NF is promising to serve as a highly effective and durable electrocatalyst toward HER.

Key words： molybdenum oxides; oxygen vacancies; heterostructure; electrocatalysts; hydrogen evolution reaction

1 Introduction

With the fast growth of energy demands and deterioration of environment, it is necessary to develop renewable and clean energy [1]. Hydrogen (H₂), thanks to its environmental friendliness and high energy density, has attracted great attention [2]. Electrocatalytic water splitting has be considered as the most eco-friendly and economical route to produce H₂, because water is an abundant and renewable source [3]. However, the hydrogen evolution reaction (HER) during water electrolysis suffers from large overpotential and slow kinetics [4]. To speed up the HER kinetics, highly active and durable electrocatalysts are applied to lower the dynamic overpotentials [5]. Pt-based compounds have been considered as the most efficient electrocatalysts toward HER, but the preciousness and scarcity of Pt seriously restrict its practical utilization in water electrolysis [6]. Thus, it is vital to explore low-cost while high-efficiency electrocatalysts to replace Pt-based materials.

Nonprecious metal materials, such as Mo-based sulfides [7], oxides [8], phosphides [9], and carbides [10], have been investigated as potential HER catalysts. Compared with molybdenum sulfides, molybdenum oxides are highly stable and readily available in large scale and can also show acceptable HER property [11]. MoO₃ (α-MoO₃), as a low-price and nontoxic material, has been reported to be widely applied in HER, exhibiting an overpotential of 112 mV achieving the current density of 10 mA∙cm^‒2 (η₁₀ = 112 mV) in acid media [12]. However, because of the less active sites and poor conductivity of intrinsic MoO₃, its catalytic performance has been reported to be much lower than that of Pt-based materials [13]. If the MoO₃ is combined with other substances to fabricate certain active interfaces, the active sites of MoO₃ may be exposed as much as possible [14]. Zhang’s group [15] coupled MoO₃ with 1T'-MoS₂ to assembly 1T'-MoS₂/MoO₃ heterostructure nanosheets, in which the MoS₂/MoO₃ interfaces can facilitate electron transfer and surface hydrogen generation, leading to the improvement of electrocatalytic performance. Besides, loading P element onto MoO₃ nanosheets can also improve the HER catalytic performance [16], for which the P component can boost the adsorption/desorption of proton and afterwards increase electrocatalytic performance. A composite electrocatalyst comprising RuO₂ supported on MoO₃ nanosheets was investigated as an HER electrocatalyst, in which the synergetic effect created through interaction between MoO₃ and RuO₂ led to the enhancement of catalytic activity [17]. Nevertheless, their HER performance is still inferior to that of precious metals. In addition, the oxygen vacancy in MoO₃ may not only produce greater conductivity, but also act as HER active site [13]. For example, MoO_3‒x with oxygen vacancies depicted higher HER catalytic activity than com-MoO₃ [13]. Density functional theory (DFT) calculation revealed that the oxygen vacancy prominently reduces the adsorption energy of H₂O and then improves the catalytic activity. Thus, fabricating oxygen vacancies and active interfaces into molybdenum oxides may be effective ways to enhance the HER performance of molybdenum oxides. The known researches on molybdenum oxides with oxygen vacancy are all about crystalline MoO₃, but there are lack of reports focused on amorphous MoO_x with oxygen vacancy.

Meanwhile, Ni-based materials have shown a big potential for using as HER catalysts [18]. Ni₃S₂, because of the existence of Ni–Ni networks throughout its crystal structure, exhibits good metallic conductivity [19]. So far, many studies revealed that Ni₃S₂ owns HER activity [20], while its catalytic activity and stability are still less competitive relative to noble-metal catalysts. On the other hand, self-supported catalysts, compared to powdery ones, avoid the addition of polymer binder. Meanwhile, the introduction of substrates in self-supporting catalysts can raise electrochemical active areas. The three-dimensional nickel foam (NF) is a low-cost template with large surface area, and is normally used as a conductive substrate to host electrocatalyst materials [21]. Using thiourea, Ni(NO₃)₂·6H₂O and NF as reactants, Ni₃S₂/NF was fabricated to depict high HER activity and good durability in basic and neutral conditions [22]. Hybridizing Ni₃S₂ with other electrocatalysts such as MoS₂ [23] and/or Co₉S₈/MoS₂ [24] is also a commonly employed method to augment the catalytic activity. Moreover, Ni₃S₂ is generally combined with other sulfides, while the integration of Ni₃S₂ with amorphous MoO_x is rarely reported. As known, Ni₃S₂ and NF matrix are unstable in acidic solutions, while MoO₃ or MoO_x is acid-resistant. Therefore, if NF and the in situ grown Ni₃S₂ are wrapped by dense layers of MoO_x, it is expected to improve the acid resistance of the integrated elelctrocatalysts.

Herein, via a facile solvothermal method, we fabricate MoO_x/Ni₃S₂ spheres which consist of dispersed Ni₃S₂ nanoparticles and amorphous MoO_x wrapped outside. By adjusting ratios of water/ethanol solvents, the as-prepared MoO_x/Ni₃S₂/NF-11 (volume ratio of ethanol:water is 1:1) exhibits optimum HER performance, requiring an extremely low overpotential (η₁₀ = 76 mV) and a much low Tafel slope of 46 mV·dec^‒1 in 0.5 mol·L^‒1 H₂SO₄. Benefiting from the existing oxygen vacancy and formed heterojunction, Mo atoms are exposed as much as possible to exhibit reduced Gibbs free energy of hydrogen adsorption (∆G_H*) and act as active-sites. Additionally, the stability of the catalyst under acidic conditions is greatly improved because the acid-resistant amorphous MoO_x layer is wrapped around Ni₃S₂/NF. The MoO_x/Ni₃S₂/NF-11 displays extremely high stability (≥ 100 h at 23.5 mA∙cm^‒2). This study would give a novel perspective on the fabrication of amorphous molybdenum oxides or oxygen-deficient materials with excellent HER performance.

2 Experimental

2.1 Synthesis of MoO_x/Ni₃S₂/NF composites

A series of MoO_x/Ni₃S₂/NF composites (labeled as MoO_x/Ni₃S₂/NF-mn, where m and n are volume ratios of ethanol:water) were synthesized via a one-pot solvothermal reaction. Typically, two pieces of commercial NF (1 cm × 2 cm) were first immersed in an acid solution (1 mol·L⁻¹ HCl) for 20 min, to remove the oxides on the surface of NF. Then the NF was washed using acetone, water and ethanol in turn, and then was dried in vacuum at 45 °C, followed by weighing. Meanwhile, 2 mmol of Na₂MoO₄·2H₂O was dissolved in 30 mL of ethanol:water mixture solvents. After stirred for 30 min, 10 mmol of thiourea was dissolved in the solutions. The abovementioned solutions and pre-treated NF were simultaneously transferred to a Teflon-lined autoclave (50 mL) and reacted at 140 °C for 18 h. After cooled, the obtained MoO_x/Ni₃S₂/NF-mn products were washed with excess deionized water and ethanol several times, and then dried in vacuum at 45 °C overnight.

2.2 Electrochemical measurements

The electrochemical tests were conducted in a three-electrode cell in a 0.5 mol·L⁻¹ H₂SO₄ solution. Ag/AgCl (saturated KCl) electrode and a graphite rod electrode worked as reference and counter electrodes, respectively. The as-prepared samples were used directly as working electrodes. For a reliable comparison, we loaded Pt−C and commercial MoO₃ (com-MoO₃) on NF with the same mass loading as MoO_x/Ni₃S₂/NF-11. Table S1 (cf. Electronic Supplementary Material, ESM) listed the mass loading of all catalysts. The details of electrochemical measurements are shown in ESM.

3 Results and discussion

3.1 Material characterization

The fabrication procedure of the MoO_x/Ni₃S₂/NF composites is described in . In the solvothermal reaction, the thiourea may break down to release HS⁻ ions (Eq. (1)); meantime, the NF can release Ni²⁺ into the solution, and the Ni²⁺ would react with the active HS⁻ to generate the Ni₃S₂ particles (Eq. (2)) [24]:

Scheme1 Schematic illustration of formation of oxygen-deficient MoO_x/Ni₃S₂/NF composites.

Full size|PPT slide

(1)\mathrm{N}{\mathrm{H}}_2\mathrm{C}\mathrm{S}\mathrm{N}{\mathrm{H}}_2+3{\mathrm{H}}_2\mathrm{O}\to 2\mathrm{N}{{\mathrm{H}}_4}^{+}+\mathrm{H}{\mathrm{S}}^{-}+\mathrm{H}\mathrm{C}{{\mathrm{O}}_3}^{-}

(2)3\mathrm{N}\mathrm{i}+2\mathrm{H}{\mathrm{S}}^{-}+2{\mathrm{H}}_2\mathrm{O}\to \mathrm{N}{\mathrm{i}}_3{\mathrm{S}}_2+2\mathrm{O}{\mathrm{H}}^{-}+2{\mathrm{H}}_2

The structures of the MoO_x/Ni₃S₂/NF-mn were studied firstly by X-ray diffraction (XRD) measurements. As shown in Fig.1, for all samples, there appeared diffraction peaks at 21.7°, 31.1°, 37.8°, 49.7° and 55.2° respectively assigned to the (101), (110), (003), (113) and (122) planes of hexagonal Ni₃S₂ (JCPDS no. 44-1418) [25]. This indicates that the partial NF surface was converted to Ni₃S₂ under the solvothermal conditions. Three strong diffraction peaks of Ni were derived from the NF substrate. However, no distinguishable diffractions related to MoO_x were found, suggesting that if there is certain MoO_x phase, it would exist in amorphous form. The transmission electron microscope (TEM) characterization discussed below with the absence of lattice fringes would provide important evidence of the amorphous characteristic.

Fig.1 XRD patterns of (a) MoO_x/Ni₃S₂/NF-01, (b) MoO_x/Ni₃S₂/NF-11 and (c) MoO_x/Ni₃S₂/NF-21.

Full size|PPT slide

The morphologies of NF (Fig. S1, cf. ESM) and MoO_x/Ni₃S₂/NF-mn (Fig.2) were characterized by scanning electron microscopy (SEM). Figure S1(a) shows the 3D porous characteristic of NF, and at higher magnification, the surface of NF can be observed to be very smooth (Fig. S1(a′)). The special skeleton of NF provides large surface area and ideal mechanical stability. As seen in Fig.2, the entire NF surface is tightly covered by the MoO_x/Ni₃S₂. When ethanol solvent was absent in the reaction (no ethanol), the resulting product of MoO_x/Ni₃S₂/NF-01 (Fig.2(a) and Fig.2(a′)) has some cracks on its surface, which may reduce the conductivity of the catalyst [26]. At the ethanol:water ratio of 1:1, the as-formed MoO_x/Ni₃S₂/NF-11 presents irregular small spheres grown on the substrate (Fig.2(b) and Fig.2(b′)). In addition, it can be seen from Fig.2(b) that there is no crack on the surface of the sample, which means that the acid-resistant NF matrix is completely coated by MoO_x/Ni₃S₂, so that NF cannot be corroded by acid, paving the way for the excellent stability of the catalyst under acidic conditions. As the ethanol:water ratio is increased to 2:1 (with more ethanol), as found in MoO_x/Ni₃S₂/NF-21, the formed MoO_x/Ni₃S₂ spheres on the NF surface become larger (Fig.2(c) and Fig.2(c′)).

Fig.2 SEM images of (a, a′) MoO_x/Ni₃S₂/NF-01, (b, b′) MoO_x/Ni₃S₂/NF-11 and (c, c′) MoO_x/Ni₃S₂/NF-21.

Full size|PPT slide

For the MoO_x/Ni₃S₂/NF-11 with the most regular and perfect appearance, detailed morphology and structure were further explored by TEM (Fig.3 and S3, cf. ESM) via scratching the samples from the NF substrate. TEM images (Fig.3(a) and Fig.3(b)) show that the MoO_x/Ni₃S₂ spheres are comprised of nanoparticles. Fig.3(c) reveals the interface region of the Ni₃S₂ nanoparticles, indicating that there are two clearly distinguishable phases between the inside and outside of the particle, and the Ni₃S₂ nanoparticle is closely connected to the amorphous MoO_x. For clarity, the original image (without scaling) of Fig.3(c) is shown in Fig. S2 (cf. ESM). The TEM investigations of MoO_x/Ni₃S₂-11 (see Fig.3(d) and Fig.3(e)) further exhibit the tight connection between the Ni₃S₂ particles and MoO_x. The lattices of 0.24 (Fig.3(d)) and 0.29 nm (Fig.3(e)) are ascribed to (003) and (110) crystal planes of Ni₃S₂ [27], further proving its presence. The above results can be supported by the SAED pattern (Fig.3(f)), which displays characteristic crystal planes of Ni₃S₂ and diffuse rings of the amorphous MoO_x. The amorphous characteristic of MoO_x leads to the absence of XRD diffractions (see Fig.1). The MoO_x surrounding for Ni₃S₂ further ensures the acid resistance of the electrocatalyst. The compositional distributions of the MoO_x/Ni₃S₂-11 sphere are confirmed by elemental mapping analyses (Fig.3(g)), high angle annular dark field scanning transmission electron microscope (HAADF-STEM) image (Fig. S3(a)) and line scans (Fig. S3(b)). In Fig.3(g), uneven distribution of Mo/O and Ni/S is clearly observed throughout the whole sphere. The elements of Mo and O appear in the same region, while Ni and S elements locate in other regions, indicating that the two phases of MoO_x and Ni₂S₃ have distinct interfaces. From Fig.3(g), the atom ratio of Mo:O is 1:2.94, which is close to MoO₃, with the oxygen content slightly reduced. All these results prove the existence of Ni₃S₂ crystalline phase along with amorphous MoO_x.

Fig.3 (a, b) TEM images of MoO_x/Ni₃S₂-11, (c, d, e) HR-TEM images of MoO_x/Ni₃S₂-11, and the visible lattice fringes images obtained from the blue and yellow square regions; (f) SAED pattern of MoO_x/Ni₃S₂-11; (g) HAADF-STEM and energy dispersive spectroscopy elemental mappings of Mo, O and Ni of MoO_x/Ni₃S₂-11; (h) electron paramagnetic resonance (EPR) spectrum of MoO_x/Ni₃S₂/NF-11 before and after 60 h I–t test, and the control sample of com-MoO₃.

Full size|PPT slide

To clarify the detailed structure of MoO_x/Ni₃S₂/NF-11, EPR spectroscopy and X-ray photoelectron spectroscopy (XPS) were carried out and the results were shown in Fig.4, with com-MoO₃ as a control sample. In Fig.3(h), the EPR spectra of com-MoO₃ (black line) shows no signal, while MoO_x/Ni₃S₂/NF-11 (red line) displays a signal at about 3512 Guass (g = 2.0036), indicating the existence of oxygen vacancy. For XPS spectra of MoO_x/Ni₃S₂/NF-11 (Fig.4(a)), the binding energies of 235.6 and 232.4 eV are related to Mo 3d_3/2 and 3d_5/2 of Mo⁶⁺, the same as found in com-MoO₃ [28]. In contrast to com-MoO₃, MoO_x/Ni₃S₂/NF-11 shows a pair of peaks of Mo⁵⁺ (234.0 eV for Mo 3d_3/2 and 230.8 eV for Mo 3d_5/2) [29]. Comparison of O 1s spectra of MoO_x/Ni₃S₂/NF-11 and com-MoO₃ is presented in Fig.4(b)_. As shown, for com-MoO₃, the binding energy of 531.9 eV is assigned to lattice oxygen (O^2‒) [30], while in MoO_x/Ni₃S₂/NF-11, the energy of O^2‒ is reduced to a lower level (531.5 eV). Also, the half-peak width of O 1s core-level spectra of the MoO_x/Ni₃S₂/NF-11 is widened in comparison with that of the com-MoO₃. The difference of half-peak width of O 1s core-level spectra in the materials with and without oxygen vacancies has also been observed in other literature [31]. For MoO_x/Ni₃S₂/NF-11, the appearance of Mo⁵⁺ and the shift of O 1s simultaneously prove the presence of oxygen vacancy, which is consistent with EPR results. All these indicate a changed coordination configuration between Mo and O, as discussed in literatures [13,16]. It was reported that the shift of O 1s to a lower energy level means the electron transfer to the neighboring oxygen vacancies [32]. Meanwhile, one weak peak detected at 533.4 eV can be assigned to O 1s of surface adsorbed species (here is OH^‒) [33]. In Fig.4(c), for Ni 2p spectra of MoO_x/Ni₃S₂/NF-11, the peaks at 873.8 and 856.0 eV are related to Ni 2p_1/2 and Ni 2p_3/2 of Ni²⁺, accompanied by two satellite peaks at 879.8 and 861.7 eV [34]. The peak at 853.4 eV is assigned to Ni⁰, which belongs to Ni₃S₂ or NF [35]. In the spectra of S 2p (Fig.4(d)), the two signals at 162.9 and 161.7 eV belong to 2p_1/2 and 2p_3/2 of S^2‒, and the other two signals at 164.4 and 163.2 eV are attributed to 2p_1/2 and 2p_3/2 of S₂^2‒ [36], suggesting the presence of terminal unsaturated S of Ni–S bonds [37]. The peak at 168.2 eV is assigned to oxidized sulfur species (here is SO₄^2‒) owing to surface oxidation [38]. In Raman spectra of MoO_x/Ni₃S₂/NF-11 (Fig. S4, cf. ESM), the band at 324 cm^‒1 is associated with A₁ vibration mode of the Ni₃S₂ phase [39], and the peaks ranging from 800 to 1000 cm^–1 are related to Mo=O modes [16]. We also measured the infrared spectrum of MoO_x/Ni₃S₂/NF-11 (Fig. S5, cf. ESM). The bands at 975 and 914 cm⁻¹ correspond to the stretching vibration of Mo=O, and the bands at 615 and 514 cm⁻¹ are assigned to stretching and bending vibrations of Mo–O, respectively [40]. No Mo–S related spectral peaks can be observed, which proves that the sample does not contain Mo–S bond.

Fig.4 X-ray photoelectron spectra with deconvolution of (a) Mo 3d, (b) O 1s for MoO_x/Ni₃S₂/NF-11 and com-MoO₃, (c) Ni 2p, and (d) S 2p for MoO_x/Ni₃S₂/NF-11.

Full size|PPT slide

3.2 Electrocatalytic performance for HER

HER performance of MoO_x/Ni₃S₂/NF-mn synthesized at different ethanol:H₂O ratios and control samples of com-MoO₃/NF, Ni₃S₂/NF and Pt–C/NF was measured (Fig.5 and Fig. S6, cf. ESM). As observed (Fig. S6), among the three samples, the MoO_x/Ni₃S₂/NF-11 at an ethanol:H₂O ratio of 1:1 exhibits an extremely low η₁₀ value of 76 mV, demonstrating HER performance much better than MoO_x/Ni₃S₂/NF-01 (η₁₀ = 127 mV) without ethanol added during synthesis and MoO_x/Ni₃S₂/NF-21 (η₁₀ = 114 mV) prepared at an ethanol:H₂O ratio of 2:1. From the SEM images mentioned above, the MoO_x/Ni₃S₂/NF-11 has a relatively small particle size and no crack on the surface, which is conducive to increasing the contact area between the catalyst and the electrolyte and accelerating the electron transmission. Therefore, the MoO_x/Ni₃S₂/NF-11 exhibits the optimal HER performance. Furthermore, the η₁₀ of 76 mV of MoO_x/Ni₃S₂/NF-11 is extremely lower than that (235 mV) of com-MoO₃/NF, and is close to that (37 mV) of Pt–C/NF, showing performance exceeding com-MoO₃ and comparable to Pt–C. As listed in Tab.1 [16,17,41–46], the η₁₀ of MoO_x/Ni₃S₂/NF-11 is obviously lower compared with the reported Mo-related electrocatalysts. The excellent catalytic performance of MoO_x/Ni₃S₂/NF-11 may be mainly attributed to the oxygen vacancy. For the P-MoO₃ [16], though it also contains oxygen vacancies coming from P doping, it has a larger η₁₀ of 166 mV, meaning an inferior HER activity than MoO_x/Ni₃S₂/NF-11. The outstanding HER activity of MoO_x/Ni₃S₂/NF-11 suggests that there may be heterojunctions formed in the two integrated phases of MoO₃ and Ni₃S₂. The heterojunction and oxygen vacancy would offer rapid electron transfer favorable to the HER activity.

Tab.1 HER performance of MoO_x/Ni₃S₂/NF-11 and some reported electrocatalysts in 0.5 mol·L⁻¹ H₂SO₄

Catalysts	η10/mV	Tafel slope/(mV∙dec‒1)	Ref.
MoOx/Ni3S2/NF-11	76	46	This work
P-MoO3‒xa)	166	42	[11]
MoO3@RuO2b)	110	62	[17]
Pd NDs/DR MoS2c)	103	41	[41]
UDSL-MoS2-rGOd)	ca. 210	35	[42]
Mo/Mo2Ce)	89	62	[43]
(1T/2H) MoS2/α-MoO3f)	232	81	[44]
MoP/NGg)	94	50	[45]
MoCxh)	142	53	[46]

a) P doped MoO3−x, prepared via two-step intercalation method by using dodecylamine (DDA) and 4-bromine benzyl phosphoric acid; b) RuO2 nanoparticles supported on MoO3 nanosheets, prepared by a sonochemical method followed by calcination in air; c) Pd nanodisks (NDs) assembled on the basal plane of defect-rich MoS2 nanosheets (DR-MoS2); d) ultradispersed and single-layered MoS2 nanoflakes coupled with reduced graphene oxide sheets (UDSL-MoS2-rGO), synthesized by a hydrothermal method using (NH4)6Mo7O24·4H2O, L-cysteine and GO under pH = 1; e) Mo/Mo2C heteronanosheets, obtained by a NaCl template method followed by the reduction and carbonization under H2 and CH4, respectively; f) (1T/2H) MoS2/α-MoO3 prepared by hydrothermal method using thiourea, MoO3 and N2H4·H2O; g) MoP nanoflakes intercalated nitrogen-doped graphene nanobelts (MoP/NG), synthesized by inserting DDA into MoO3 nanobelts followed by phosphorization; h) MoCx obtained by MOFs-assisted strategy followed by annealing under N2 flow.

Fig.5 (a) Polarization curves of MoO_x/Ni₃S₂/NF-11, com-MoO₃/NF, Ni₃S₂/NF and Pt–C/NF; (b) Tafel slopes of MoO_x/Ni₃S₂/NF-11, com-MoO₃/NF and Pt–C/NF derived from polarization curves; (c) Nyquist plots of MoO_x/Ni₃S₂/NF-11, com-MoO₃/NF and Pt–C/NF at ‒200 mV versus RHE measured from electrochemical impedance spectroscopy (EIS) in the frequency range from 10⁵ to 0.01 Hz; (d) plots of current density as a function of scan rates for MoO_x/Ni₃S₂/NF-11, com-MoO₃/NF and Pt–C/NF; (e) chronoamperometric curve of MoO_x/Ni₃S₂/NF-11 at a constant applied potential of ‒180 mV versus RHE; (f) polarization curves of before and after 100 h I–t test of MoO_x/Ni₃S₂/NF-11. All the measurements were performed in a 0.5 mol·L⁻¹ H₂SO₄ electrolyte.

Full size|PPT slide

To explore the catalytic mechanism of the electrocatalysts, Tafel plots were calculated (Fig.5(b)). The Tafel slope of MoO_x/Ni₃S₂/NF-11 is 46 mV∙dec^‒1, which is close to the value (30 mV∙dec^‒1) of Pt–C/NF and much lower than that (115 mV∙dec^‒1) of com-MoO₃/NF. Tab.1 listed the Tafel slopes of reported Mo-based HER catalysts. The low Tafel slope of MoO_x/Ni₃S₂/NF-11 demonstrates its effectively enhanced kinetics during the H₂O dissociation [17]. On the basis of kinetic models, the Tafel slopes of 120, 40 and 30 mV∙dec^‒1 are assigned to Volmer (Eq. (3)), Heyrovsky (Eq. (4)) and Tafel (Eq. (5)) reactions, respectively:

(3){\mathrm{H}}_3{\mathrm{O}}^{+}+\text{*}+{\mathrm{e}}^{-}\to {\mathrm{H}}_{\mathrm{a}\mathrm{d}\mathrm{s}}\text{*}+{\mathrm{H}}_2\mathrm{O}

(4){\mathrm{H}}_{\mathrm{a}\mathrm{d}\mathrm{s}}\text{*}+{\mathrm{H}}_3{\mathrm{O}}^{+}+{\mathrm{e}}^{-}\to \text{*}+{\mathrm{H}}_2+{\mathrm{H}}_2\mathrm{O}

(5){\mathrm{H}}_{\mathrm{a}\mathrm{d}\mathrm{s}}\text{*}+{\mathrm{H}}_{\mathrm{a}\mathrm{d}\mathrm{s}}\text{*}\to 2\text{*}+{\mathrm{H}}_2\uparrow

The first step is Volmer reaction (Eq. (3)), for which one proton (H₃O⁺) first adsorbs on the surface of the catalyst and accepts one electron to generate one hydrogen atom (H_ads*). The second step involves two kinds of reactions: one is Heyrovsky reaction and the other is Tafel reaction. For the Heyrovsky reaction (Eq. (4)), which is an electrochemical desorption, another proton (H₃O⁺) accepts one electron and desorbs from catalyst surface and react with the former H atom to generate one H₂. For the Tafel step (Eq. (5)), two H_ads desorb from catalyst surface and combine directly and then release one H₂. The slope of 46 mV∙dec^‒1 for MoO_x/Ni₃S₂/NF-11 indicates it follows the Volmer–Heyrovsky mechanism, while Pt–C/NF with the much lower slope of 30 mV∙dec^‒1 follows the Volmer–Tafel mechanism. Nyquist plots were obtained by measuring EIS and the charge-transfer resistance (R_ct) values were fitted by the equivalent circuit model (Fig.5(c)). It can be seen that the MoO_x/Ni₃S₂/NF-11 shows an extremely small R_ct of 6.51 Ω, which is much lower than the com-MoO₃/NF (21.51 Ω), indicating that the MoO_x/Ni₃S₂/NF-11 has faster charge transfer dynamics in HER process (see Table S2, cf. ESM). The fast charge transfer is probably attributed to synergetic effect of the amorphous MoO_x and the Ni₃S₂ nanoparticles. The abundant interfaces of the two phases make the hydrogen adsorption and desorption more effective [17].

In general, good electrocatalytic activity is accompanied by a large electrochemically active surface area, which is normally proportional to the electrochemical double-layer capacitance (C_dl). From the cyclic voltammetry curves carried out at varied scan rates (Fig. S7, cf. ESM), the C_dl of MoO_x/Ni₃S₂/NF-11 is calculated as 42 mF∙cm^‒2 (Fig.5(d)). This C_dl value is much larger than those many reported catalysts, such as MoO₃ (13.58 mF∙cm^‒2) [14], UDSL-MoS₂-rGO (24 mF∙cm^‒2) [42], MoP/NG (9.1 mF∙cm^‒2) [45] and MoO₂/MoSe₂ (18.68 µF∙cm^‒2) [47]. This large C_dl unveils the large amount of electrochemical active sites in the MoO_x/Ni₃S₂/NF-11. These results reveal that the MoO_x/Ni₃S₂/NF-11 with oxygen vacancies generates more accessible active sites, thereby promoting the H₂ evolution process.

3.3 DFT calculations

DFT calculations (Fig.6) were employed to well understand the influence of oxygen vacancy and Ni₃S₂ component on the catalytic activity. According to TEM characterization, MoO_x was found to be wrapped outside the Ni₃S₂. Therefore, when building the optimized models, the possibility of H^* adsorption on MoO₃ is given priority, and Ni₃S₂ is regarded as the “substrate” of MoO₃. As shown in Fig.6(A), we calculate the ΔG_H* of MoO₃ (on O site), oxygen-deficient MoO₃ (noted as O_v-MoO₃, on Mo site), MoO₃/Ni₃S₂ (on O site), and O_v-MoO₃/Ni₃S₂ (on O and Mo sites). The MoO₃ (Fig.6(A-a)) shows a ΔG_H* of ‒2.37 eV, for which the negative value indicates easy H adsorption (Volmer reaction, Eq. (3) discussed above) but difficult H desorption (Heyrovsky reaction, Eq. (4) discussed above). Because the surface of MoO₃ is all O atoms and the O–H^* bond is strong, the H^* is difficult to be desorbed. At this time, desorption step becomes the rate-determining step. When oxygen vacancy is present, as in O_v-MoO₃ (Fig.6(A-b)), Mo sites are exposed to exhibit a smaller |∆G_H*| of 2.30 eV. When MoO₃ and Ni₃S₂ form the heterojunction (Fig.6(A-c), the MoO₃/Ni₃S₂ has a much smaller |∆G_H*| of 1.82 eV, which means weakened binding between O and H, thus facilitating the H desorption. While for the heterojunction of O_v-MoO₃/Ni₃S₂ containing oxygen vacancy, |∆G_H*| on the O site adjacent to oxygen vacancy is slightly increased to 1.98 eV (Fig.6(A-d)), indicating strengthened O–H bonding (difficult H desorption). More importantly and interestingly, the Mo site of O_v-MoO₃/Ni₃S₂ has a much lower |∆G_H*| of 1.54 eV (Fig.6(A-e)), which is the most close to the value of 0. As known an electrocatalyst with |∆G_H*| approaching to zero would have high HER activity. So the exposed Mo due to oxygen defect serves as active sites. By comparing (a) and (c), or (b) and (d) in Fig.6(A), it can be seen that the heterojunction greatly improves the catalytic activity of both O and Mo sites, while comparison of (c), (d) and (e) in Fig.6(A) shows that although oxygen deficiency slightly reduces the activity of O sites, the exposed Mo sites result in much better catalytic activity.

Fig.6 (A) Optimized structures of (a) MoO₃, (b) O_v-MoO₃, (c) MoO₃/Ni₃S₂, (d) O_v-MoO₃/Ni₃S₂ with H^* on O site and (e) O_v-MoO₃/Ni₃S₂ with H^* on Mo site and the corresponding H adsorption free energy (ΔG_H*) at a potential U = 0 V relative to standard hydrogen electrode at pH = 0; (B) differential charge density distribution of O_v-MoO₃/Ni₃S₂; (C) vacancy formation energy of O_v-MoO₃/Ni₃S₂ and O_v-MoO₃ and corresponding models. Symbols for atoms: Mo is cyan in A and C while purple in B, S is yellow, Ni is blue in A and C while grey in B, O is red and H^* is white. The dotted box shows the position of oxygen vacancy.

Full size|PPT slide

Differential charge density distribution of the O_v-MoO₃/Ni₃S₂ has been calculated, with results shown in Fig.6(B). It is found that there is a strong charge transfer between Ni₃S₂ and MoO₃. Compared with MoO₃/Ni₃S₂ (without oxygen vacancy, Fig. S8, cf. ESM), the O atoms in O_v-MoO₃/Ni₃S₂ (Fig.6(B)) gain more electrons. The increase of surface charge of oxygen will strengthen the binding of O–H and make the desorption more difficult. This is the reason why the |∆G_H*| on O site of O_v-MoO₃/Ni₃S₂ is larger than that of MoO₃/Ni₃S₂. In addition, Fig.6(B) depicts electron deficiency of exposed Mo site in O_v-MoO₃/Ni₃S₂. During the Heyrovsky reaction, Mo–H^* bond needs to be dissociated prior to H₂ release. In this process, electrophilic Mo will pull the electrons between Mo and H, which will weaken the Mo–H^* bond thus promoting hydrogen evolution via the cleavage of Mo–H^* [48].

Long-term durability is another crucial requirement for the catalysts in real applications. To measure the durability of MoO_x/Ni₃S₂/NF-11, the time-dependent current density was measured at a constant potential of ‒0.18 V for 100 h. The excellent stability is due to that the acid-resistant amorphous MoO_x encases the Ni₃S₂ and NF substrate which are unstable under acidic conditions. As seen in Fig.5(e), the MoO_x/Ni₃S₂/NF-11 was continuously operated for 100 h with no significant degradation observed, confirming the ultra-durable stability. Fig.5(f) shows the linear sweep voltammetry before and after 100 h I–t HER test. As observed, the catalytic activity of the MoO_x/Ni₃S₂/NF-11 was even increased after the stability test. This may be thankful to the surface reconfiguration of the catalyst [49]. In order to deeply learn about the reaction mechanism in the HER process, we have carried out SEM and XPS for MoO_x/Ni₃S₂/NF-11 after stability test. From the SEM images shown in Fig. S9 (cf. ESM), we see after the HER test, the microstructure of MoO_x/Ni₃S₂/NF-11 has not changed significantly, and many irregular small spheres are still attached to the surface of NF. From XPS (Fig.7), after long-term stability test (100 h), the valence states of Ni and S do not change [50]. For Mo, the mixed valence state of Mo⁵⁺ and Mo⁶⁺ is still reflected after the stability test. The peaks at 235.6 and 232.4 eV correspond to Mo⁶⁺ [28], and the binding energies of 234.0 and 230.8 eV are related to Mo⁵⁺ [29]. At the same time, it can be seen from the integral area of the curve that the content of Mo⁵⁺ increases compared with that before. In addition, the O 1s peaks shift towards lower binding energy by 0.2 eV after stability test. The increase of Mo⁵⁺ content and the shift of O 1s with lower binding energy also prove that the oxygen vacancy content increased after long-term stability. The in situ surface reconfiguration may result in more oxygen vacancies on catalyst surface, which exposes more Mo active sites. EPR spectroscopy (Fig.3(h)) shows that the MoO_x/Ni₃S₂/NF-11 after 60 h I–t test displays a much stronger signal, proving increased content of oxygen vacancy in the catalyst after HER testing, which is consistent with the XPS characterization. To further confirm the in situ surface reconfiguration of the catalyst, we calculated the formation energy of oxygen vacancy in MoO₃ and MoO₃/Ni₃S₂. As shown in Fig.6(C), when MoO₃ and Ni₃S₂ form heterojunction, the formation energy of oxygen vacancy is significantly reduced, and thus decreases the energy barrier for structural reconfiguration. This will enable MoO₃ to have more oxygen vacancies, thus exposing more active sites of Mo. The low |∆G_H*| of exposed Mo sites is beneficial to the improvement of the HER activity. With the continuation of the electrolysis reactions, the performance of the catalyst will gradually improve. From the long-term durability test (Fig.5(e)), the initial current density is 18.2 mA∙cm^‒2, while until 100 h, the current density increases to 23.5 mA∙cm^‒2. This phenomenon confirms the rationality of the above conjecture. The higher durability of the MoO_x/Ni₃S₂/NF-11 than that of the existing catalysts endows it a broad application.

Fig.7 XPS spectra of (a) Ni 2p, (b) O 1s, and (c) S 2p for MoO_x/Ni₃S₂/NF-11 before and after HER testing, and (d) Mo 3d for MoO_x/Ni₃S₂/NF-11 after HER testing.

Full size|PPT slide

Overall, from the above results we see clearly that the electrocatalytic performance of the MoO_x/Ni₃S₂/NF-11 material exceeds most known transition metal sulfides and oxides catalysts. Benefiting from the oxygen vacancy and heterostructure, the exposed Mo sites exhibit much reduced |∆G_H*|, which endows the function as active sites, leading to the excellent HER performance. The outstanding catalytic activity and wonderful durability of MoO_x/Ni₃S₂/NF-11 demonstrate it is a promising electrocatalyst for HER.

4 Conclusions

In summary, by a facile and cost-effective solvothermal method, amorphous and oxygen-deficient MoO_x integrated with Ni₃S₂ nanoparticles are embedded in the Ni foam to fabricate the new electrocatalysts of MoO_x/Ni₃S₂/NF-mn. The optimum oxygen-deficient MoO_x/Ni₃S₂/NF-11 exhibits remarkable HER activity, requiring a much low η₁₀ value of 76 mV and an extremely small Tafel slope of 46 mV∙dec^‒1 in acidic media. The oxygen vacancy leaves the Mo sites exposed, and the reduced |∆G_H*| of Mo sites ensures the enhancement of HER activity. Moreover, the formed MoO_x/Ni₃S₂ heterostructure further reduces the |∆G_H*| of Mo and O sites, which is another key factor in improving the catalytic activity. The high activity of MoO_x/Ni₃S₂/NF-11 can be maintained for more than 100 h at constant potential of 180 mV. The outside acid-resistant MoO_x layer encases the inside Ni₃S₂ and NF which are unstable under acidic conditions, thus increasing the acid resistance of the overall catalyst. In addition, the heterostructure can effectively reduce the formation energy of oxygen vacancy, which leads to the in situ reconstruction of the catalysts during the electrocatalytic process, thus promoting HER performance. As a high-performance and ultra-durable electrocatalyst, MoO_x/Ni₃S₂/NF-11 might be regarded as a cost-effective candidate to take the place of noble metals catalysts used in HER. This work could provide new ideas for the design of other oxygen-deficient materials, therefore starting new opportunities to develop high-performance materials for HER or other applications.

Acknowledgements

Experimental work is supported by the National Natural Science Foundation of China (Grant No. 22176017), Scientific Research Project of the Ningxia Higher Education Department of China (Grant No. NGY2020034) and CAS “Light of West China Program (Grant No. XAB2020YW16)”.

Electronic Supplementary Material

Supplementary material is available in the online version of this article at https://dx.doi.org/10.1007/s11705-022-2228-1 and is accessible for authorized users.

References

1 Zou X, Zhang Y. Noble metal-free hydrogen evolution catalysts for water splitting. Chemical Society Reviews, 2015, 44(15): 5148–5180

2 Turner J A. Sustainable hydrogen production. Science, 2004, 305(5686): 972–974

3 Holladay J D, Hu J, King D L, Wang Y. An overview of hydrogen production technologies. Catalysis Today, 2009, 139(4): 244–260

4 Chen W F, Muckerman J T, Fujita E. Recent developments in transition metal carbides and nitrides as hydrogen evolution electrocatalysts. Chemical Communications, 2013, 49(79): 8896–8909

5 Zhang J, Wang T, Liu P, Liao Z, Liu S, Zhuang X, Chen M, Zschech E, Feng X. Efficient hydrogen production on MoNi₄ electrocatalysts with fast water dissociation kinetics. Nature Communications, 2017, 8(1): 15437

6 Stamenkovic V R, Mun B S, Arenz M, Mayrhofer K J J, Lucas C A, Wang G F, Ross P N, Markovic N M. Trends in electrocatalysis on extended and nanoscale Pt-bimetallic alloy surfaces. Nature Materials, 2007, 6(3): 241–247

7 Wang K W, She X L, Chen S, Liu H L, Li D H, Wang Y, Zhang H W, Yang D J, Yao X D. Boosting hydrogen evolution via optimized hydrogen adsorption at the interface of CoP₃ and Ni₂P. Journal of Materials Chemistry A: Materials for Energy and Sustainability, 2018, 6(14): 5560–5565

8 Lu W, Song Y, Dou M, Ji J, Wang F. Ni₃S₂@MoO₃ core/shell arrays on Ni foam modified with ultrathin CdS layer as a superior electrocatalyst for hydrogen evolution reaction. Chemical Communications, 2018, 54(6): 646–649

9 Wang X, Ma W, Ding C, Xu Z, Wang H, Zong X, Li C. Amorphous multi-elements electrocatalysts with tunable bifunctionality toward overall water splitting. ACS Catalysis, 2018, 8(11): 9926–9935

10 Liang Q, Jin H, Wang Z, Xiong Y, Yuan S, Zeng X, He D, Mu S. Metal–organic frameworks derived reverse-encapsulation Co-NC@Mo₂C complex for efficient overall water splitting. Nano Energy, 2019, 57: 746–752

11 Li L, Zhang T, Yan J, Cai X, Liu S. P doped MoO_3−x nanosheets as efficient and stable electrocatalysts for hydrogen evolution. Small, 2017, 13(25): 1700441

12 Sinaim H, Ham D J, Lee J S, Phuruangrat A, Thongtem S, Thongtem T. Free-polymer controlling morphology of α-MoO₃ nanobelts by a facile hydrothermal synthesis, their electrochemistry for hydrogen evolution reactions and optical properties. Journal of Alloys and Compounds, 2012, 516: 172–178

13 Zhu Y H, Yao Y, Luo Z, Pan C Q, Yang J, Fang Y R, Deng H T, Liu C X, Tan Q, Liu F D, Guo Y. Nanostructured MoO₃ for efficient energy and environmental catalysis. Molecules, 2020, 25(1): 26

14 Li J, Cheng Y, Zhang J, Fu J, Yan W, Xu Q. Confining Pd nanoparticles and atomically dispersed Pd into defective MoO₃ nanosheet for enhancing electro- and photocatalytic hydrogen evolution performances. ACS Applied Materials & Interfaces, 2019, 11(31): 27798–27804

15 Xue X, Zhang J, Saana I A, Sun J, Xu Q, Mu S. Rational inert-basal-plane activating design of ultrathin 1T′ phase MoS₂ with a MoO₃ heterostructure for enhancing hydrogen evolution performances. Nanoscale, 2018, 10(35): 16531–16538

16 Liu P T, Zhu J Y, Zhang J Y, Xi P X, Tao K, Gao D Q, Xue D S. P dopants triggered new basal plane active sites and enlarged interlayer spacing in MoS₂ nanosheets toward electrocatalytic hydrogen evolution. ACS Energy Letters, 2017, 2(4): 745–752

17 Sadhanala H K, Harika V K, Penki T R, Aurbach D, Gedanken A. Ultrafine ruthenium oxide nanoparticles supported on molybdenum oxide nanosheets as highly efficient electrocatalyst for hydrogen evolution in acidic medium. ChemCatChem, 2019, 11(5): 1495–1502

18 Lim K J H, Yilmaz G, Lim Y F, Ho G W. Multi-compositional hierarchical nanostructured Ni₃S₂@MoS_x/NiO electrodes for enhanced electrocatalytic hydrogen generation and energy storage. Journal of Materials Chemistry A: Materials for Energy and Sustainability, 2018, 6(41): 20491–20499

19 Kou T, Smart T, Yao B, Chen I, Thota D, Ping Y, Li Y. Theoretical and experimental insight into the effect of nitrogen doping on hydrogen evolution activity of Ni₃S₂ in alkaline medium. Advanced Energy Materials, 2018, 8(19): 1703538

20 Feng L L, Yu G, Wu Y, Li G D, Li H, Sun Y, Asefa T, Chen W, Zou X. High-index faceted Ni₃S₂ nanosheet arrays as highly active and ultrastable electrocatalysts for water splitting. Journal of the American Chemical Society, 2015, 137(44): 14023–14026

21 Chang Y H, Lin C T, Chen T Y, Hsu C L, Lee Y H, Zhang W, Wei K H, Li L J. Highly efficient electrocatalytic hydrogen production by MoS_x grown on graphene-protected 3D Ni foams. Advanced Materials, 2013, 25(5): 756–760

22 Tang C, Pu Z, Liu Q, Asiri A M, Luo Y, Sun X. Ni₃S₂ nanosheets array supported on Ni foam: a novel efficient three-dimensional hydrogen-evolving electrocatalyst in both neutral and basic solutions. International Journal of Hydrogen Energy, 2015, 40(14): 4727–4732

23 Cao J, Zhou J, Zhang Y, Wang Y, Liu X. Dominating role of aligned MoS₂/Ni₃S₂ nanoarrays supported on three-dimensional Ni foam with hydrophilic interface for highly enhanced hydrogen evolution reaction. ACS Applied Materials & Interfaces, 2018, 10(2): 1752–1760

24 Yang Y, Yao H, Yu Z, Islam S M, He H, Yuan M, Yue Y, Xu K, Hao W, Sun G, Li H, Ma S, Zapol P, Kanatzidis M G. Hierarchical nanoassembly of MoS₂/Co₉S₈/Ni₃S₂/Ni as a highly efficient electrocatalyst for overall water splitting in a wide pH range. Journal of the American Chemical Society, 2019, 141(26): 10417–10430

25 Li T T, Zuo Y P, Lei X M, Li N, Liu J W, Han H Y. Regulating the oxidation degree of nickel foam: a smart strategy to controllably synthesize active Ni₃S₂ nanorod/nanowire arrays for high-performance supercapacitors. Journal of Materials Chemistry A: Materials for Energy and Sustainability, 2016, 4(21): 8029–8040

26 Tang T, Jiang W J, Niu S, Liu N, Luo H, Chen Y Y, Jin S F, Gao F, Wan L J, Hu J S. Electronic and morphological dual modulation of cobalt carbonate hydroxides by Mn doping toward highly efficient and stable bifunctional electrocatalysts for overall water splitting. Journal of the American Chemical Society, 2017, 139(24): 8320–8328

27 He W, Wang C, Li H, Deng X, Xu X, Zhai T. Ultrathin and porous Ni₃S₂/CoNi₂S₄ 3D-network structure for superhigh energy density asymmetric supercapacitors. Advanced Energy Materials, 2017, 7(21): 1700983

28 Yang J, Zhang F J, Wang X, He D S, Wu G, Yang Q H, Hong X, Wu Y, Li Y D. Porous molybdenum phosphide nano-octahedrons derived from confined phosphorization in UIO-66 for efficient hydrogen evolution. Angewandte Chemie International Edition, 2016, 55(41): 12854–12858

29 Weber T, Muijsers J C, Niemantsverdriet J W. Structure of amorphous MoS₃. Journal of Physical Chemistry, 1995, 99(22): 9194–9200

30 Li L D, Yan J Q, Wang T, Zhao Z J, Zhang J, Gong J L, Guan N J. Sub-10 nm rutile titanium dioxide nanoparticles for efficient visible-light-driven photocatalytic hydrogen production. Nature Communications, 2015, 6(1): 10

31 Li J, Cheng Y, Zhang J, Fu J, Yan W, Xu Q. Confining Pd Nanoparticles and atomically dispersed Pd into defective MoO₃ nanosheet for enhancing electro- and photocatalytic hydrogen evolution performances. ACS Applied Materials & Interfaces, 2019, 11(31): 27798–27804

32 Kang Q, Cao J Y, Zhang Y J, Liu L Q, Xu H, Ye J H. Reduced TiO₂ nanotube arrays for photoelectrochemical water splitting. Journal of Materials Chemistry A: Materials for Energy and Sustainability, 2013, 1(18): 5766–5774

33 Yan J Q, Zhang Y X, Liu S Z, Wu G J, Li L D, Guan N J. Facile synthesis of an iron doped rutile TiO₂ photocatalyst for enhanced visible-light-driven water oxidation. Journal of Materials Chemistry A: Materials for Energy and Sustainability, 2015, 3(43): 21434–21438

34 Xie F, Wu H, Mou J, Lin D, Xu C, Wu C, Sun X. Ni₃N@Ni-Ci nanoarray as a highly active and durable non-noble-metal electrocatalyst for water oxidation at near-neutral pH. Journal of Catalysis, 2017, 356: 165–172

35 Zhou W, Wu X J, Cao X, Huang X, Tan C, Tian J, Liu H, Wang J, Zhang H. Ni₃S₂ nanorods/Ni foam composite electrode with low overpotential for electrocatalytic oxygen evolution. Energy & Environmental Science, 2013, 6(10): 2921

36 Chia X, Sutrisnoh N A A, Pumera M. Tunable Pt-MoS_x hybrid catalysts for hydrogen evolution. ACS Applied Materials & Interfaces, 2018, 10(10): 8702–8711

37 Kuang P Y, Tong T, Fan K, Yu J G. In situ fabrication of Ni−Mo bimetal sulfide hybrid as an efficient electrocatalyst for hydrogen evolution over a wide pH range. ACS Catalysis, 2017, 7(9): 6179–6187

38 Wang B, Huang H, Sun T, Yan P, Isimjan T T, Tian J, Yang X. Dissolution reconstruction of electron-transfer enhanced hierarchical NiS_x−MoO₂ nanosponges as a promising industrialized hydrogen evolution catalyst beyond Pt/C. Journal of Colloid and Interface Science, 2020, 567: 339–346

39 Cheng Z, Abernathy H, Liu M L. Raman spectroscopy of nickel sulfide Ni₃S₂. Journal of Physical Chemistry C, 2007, 111(49): 17997–18000

40 Li Z, Ma J, Zhang B, Song C, Wang D. Crystal phase- and morphology-controlled synthesis of MoO₃ materials. CrystEngComm, 2017, 19(11): 1479–1485

41 Qi K, Yu S S, Wang Q Y, Zhang W, Fan J C, Zheng W T, Cui X Q. Decoration of the inert basal plane of defect-rich MoS₂ with Pd atoms for achieving Pt-similar HER activity. Journal of Materials Chemistry A: Materials for Energy and Sustainability, 2016, 4(11): 4025–4031

42 Huang H L, Huang J Y, Liu W P, Fang Y P, Liu Y. Ultradispersed and single-layered MoS₂ nanoflakes strongly coupled with graphene: an optimized structure with high kinetics for the hydrogen evolution reaction. ACS Applied Materials & Interfaces, 2017, 9(45): 39380–39390

43 Xiong J, Li J, Shi J W, Zhang X L, Suen N T, Liu Z, Huang Y J, Xu G X, Cai W W, Lei X R, Feng L, Yang Z, Huang L, Cheng H. In situ engineering of double-phase interface in Mo/Mo₂C heteronanosheets for boosted hydrogen evolution reaction. ACS Energy Letters, 2018, 3(2): 341–348

44 Manikandan A, Ilango P R, Chen C W, Wang Y C, Shih Y C, Lee L, Wang Z M M, Ko H, Chueh Y L. A superior dye adsorbent towards the hydrogen evolution reaction combining active sites and phase-engineering of (1T/2H) MoS₂/MoO₃ hybrid heterostructured nanoflowers. Journal of Materials Chemistry A: Materials for Energy and Sustainability, 2018, 6(31): 15320–15329

45 Huang C, Pi C R, Zhang X M, Ding K, Qin P, Fu J J, Peng X, Gao B, Chu P K, Huo K F. In situ synthesis of MoP nanoflakes intercalated N-doped graphene nanobelts from MoO₃-amine hybrid for high-efficient hydrogen evolution reaction. Small, 2018, 14(25): 7

46 Wu H B, Xia B Y, Yu L, Yu X Y, Lou X W. Porous molybdenum carbide nano-octahedrons synthesized via confined carburization in metal−organic frameworks for efficient hydrogen production. Nature Communications, 2015, 6(1): 8

47 Chen X, Liu G, Zheng W, Feng W, Cao W, Hu W, Hu P. Vertical 2D MoO₂/MoSe₂ core-shell nanosheet arrays as high-performance electrocatalysts for hydrogen evolution reaction. Advanced Functional Materials, 2016, 26(46): 8537–8544

48 Zhu L F, Liu L J, Huang G M, Zhao Q. Hydrogen evolution over N-doped CoS₂ nanosheets enhanced by superaerophobicity and electronic modulation. Applied Surface Science, 2020, 504: 144490

49 He L, Zhang W, Mo Q, Huang W, Yang L, Gao Q. Molybdenum carbide-oxide heterostructures: in situ surface reconfiguration toward efficient electrocatalytic hydrogen evolution. Angewandte Chemie International Edition, 2020, 59(9): 3544–3548

50 Yilmaz G, Yang T, Du Y, Yu X, Feng Y, Shen L, Ho G. Stimulated electrocatalytic hydrogen evolution activity of MOF-derived MoS₂ basal domains via charge injection through surface functionalization and heteroatom doping. Advancement of Science, 2019, 6(15): 1900140