Bimetal Modulation Stabilizing a Metallic Heterostructure for Efficient Overall Water Splitting at Large Current Density.

Large current-driven alkaline water splitting for large-scale hydrogen production generally suffers from the sluggish charge transfer kinetics. Commercial noble-metal catalysts are unstable in large-current operation, while most non-noble metal catalysts can only achieve high activity at low current densities <200 mA cm-2 , far lower than industrially-required current densities (>500 mA cm-2 ). Herein, a sulfide-based metallic heterostructure is designed to meet the industrial demand by regulating the electronic structure of phase transition coupling with interfacial defects from Mo and Ni incorporation. The modulation of metallic Mo2 S3 and in situ epitaxial growth of bifunctional Ni-based catalyst to construct metallic heterostructure can facilitate the charge transfer for fast Volmer H and Heyrovsky H2 generation. The Mo2 S3 @NiMo3 S4 electrolyzer requires an ultralow voltage of 1.672 V at a large current density of 1000 mA cm-2 , with ≈100% retention over 100 h, outperforming the commercial RuO2 ||Pt/C, owing to the synergistic effect of the phase and interface electronic modulation. This work sheds light on the design of metallic heterostructure with an optimized interfacial electronic structure and abundant active sites for industrial water splitting.

Introduction

Electrochemical water splitting, including hydrogen evolution reaction (HER) and oxygen evolution reaction (OER), is a sustainable route for the continuous generation of hydrogen.[ , , , , , , , ] The commercial Pt/C and RuO2 have been regarded as typical catalysts for HER and OER, respectively, but the poor stability at large‐current densities and scarcity limit their wide application.[ , ] Despite aplenty reserves, most non‐noble metal electrocatalysts such as MoS2 and NiS2, have attracted extensive attention for alkaline water splitting, owing to their superior electrocatalytic performances at small current densities (<200 mA cm−2).[ , ] However, the limited bifunctional active sites and relatively poor stability result in the large overpotential and the degradation of electrochemical performance at large current densities over 500 mA cm−2,[ ] seriously restricting their application in the industrial high‐output water splitting. Typically, the industrially‐used Raney Ni electrocatalysts operating at 500 mA cm−2 for overall water splitting required a cell voltage of 2.4 V,[ ] which largely exceeds the thermodynamic potential of 1.23 V. To date, the overall water splitting catalyzed by NiMoO @NiMoS ,[ ] FeP@Ni2P[ ] and NiMoN@NiFeN[ ] still required the cell voltages as high as 1.75, 1.73 and 1.72 V, respectively, at 500 mA cm−2. Although tremendous progress by regulation of phase structure, defects, interface, and active sites has been developed to boost activities of electrocatalysts for robust water splitting,[ , , , , , , , , , , ] traditional electronic regulation to construct a semiconductor heterostructure toward overall water splitting is difficult to realize for durable and fast Volmer H* and Heyrovsky H2 generation for large‐current operation, owing to the instinct of relatively low electronic conductivity and high ohmic contact resistance[ , , ] (Figure  ).

Figure 1

Design and characterizations for metallic heterostructure. a) Evolution of water‐splitting catalysts from typical heterostructure with ohmic barrier to the metallic heterostructure in this work with low Mott–Schottky barrier. b) Scheme of MoS2 converting to metallic Mo2S3. c) Schematic illustration of the epitaxial construction of Mo2S3@NiMo3S4. d) Energy band diagrams of Mo2S3 and NiMo3S4 (E vac = vacuum energy, E f = Fermi level, W = work function). e,f) TEM and aberration‐corrected TEM images of Mo2S3 (e) and Mo2S3@NiMo3S4 (f).

To address the above issues for robust overall water splitting, we propose a metallic heterostructure with the following advantages to satisfy the industrial demands: i) superior electrical conductivity for efficient electron transport; ii) sufficient reactive active sites for fast hydroxyl capture at large‐current densities; iii) excellent structural stability in alkaline medium at large‐current densities. In particular, the metallic sulfide support for the epitaxial growth of metallic heterostructure could be prepared by phase modulation of MoS2 with additional Mo implantation. The Mo modulation with the presence of additional Mo–Mo bonding with coordinately unsaturated centers can not only provide active sites for H* and OH* adsorption with reduced water dissociation barrier, that is the kinetically limited step in alkaline medium,[ , , ] but accelerate the desorption kinetics of H* and OH* with tunable local coordination structures in the alkaline medium.[ ] Furthermore, we construct the metallic heterostructure with enhanced bifunctional activities by incorporating Ni into the sulfide framework. On the one hand, the introduction of Ni with asymmetric 3d orbitals could recombine with the vacant 4d orbitals of Mo atoms in the sulfide framework, which can improve the interfacial electron transfer, weaken the proton adsorption and provide abundant active sites[ , ] (Figure 1a). Moreover, the strong interaction of Mo with Ni in constructed metallic heterostructure is promising for conquering the challenge of OH‐induced oxidation in alkaline electrolytes.[ ] On the other hand, the introduction of Ni into the Mo3S4 framework with electronic modulation and structural rearrangement could weaken the proton adsorption and provide abundant active sites to enhance Heyrovsky H2 generation.[ , , , , ]

In this work, a metal sulfide, Mo2S3 with metallic conductivity, transformed from typical MoS2 with additional Mo–Mo bonding is utilized as the metallic support to in situ epitaxially grow NiMo3S4, for the construction of the metallic heterostructured Mo2S3@NiMo3S4, which has metallic conductivity and abundant active sites enabling the water electrolysis at large‐current densities. Meanwhile, the in situ epitaxial growth of NiMo3S4 nanosheets on Mo2S3 nanorods with low interfacial resistance and enhanced interfacial charge transfer properties are favorable for the fast OER dynamics. The rationally designed Mo2S3@NiMo3S4 metallic heterostructure with coupled 2D nanosheets and 1D nanorods shows superior electrocatalytic performance with small overpotentials of 173, 256 and 390 mV for OER, and 32, 124 and 174 mV for HER at 10, 100 and 1000 mA cm−2. When it was integrated into a symmetric two‐electrode electrolyzer, water electrolysis required only 1.639 and 1.672 V at industrial current densities of 500 and 1000 mA cm−2 with outstanding durability (over 100 h), which is among the best non‐noble electrocatalysts for large‐current water splitting.

Results and Discussion

Design and Characterizations for Metallic Heterostructure

Figure 1b shows the schematic illustration of semiconductor MoS2 converting to metallic Mo2S3 by introducing excessive Mo in MoS2 with the appearance of Mo–Mo bonds.[ ] The transformation of typical semiconductor MoS2 to metallic Mo2S3 can be revealed by the following Equation (1):

Compared with MoS2, the conductivity of Mo2S3 was greatly improved due to the formation of Mo–Mo coordinated vertical zigzag chains. The metallic heterostructured Mo2S3@NiMo3S4 was constructed by the in situ generation of NiMo3S4 nanosheets on the Mo2S3 nanorods via a facile hydrothermal method (Figure 1c). The introduction of metallic Mo2S3 with a high Fermi level as support was conducive to interfacial charge transfer. Meanwhile, Mo2S3 possessing shorter Mo—Mo bonds than pure Mo metal could generate delocalized electrons and electronic states for tuning the adsorption/desorption behavior of H* and OH* with enhanced HER dynamics. The energy band diagrams of Mo2S3, NiMo3S4 and Mo2S3@NiMo3S4 revealed the favorable electron transfer from NiMo3S4 to Mo2S3 through the formation of heterointerface (Figure 1d). The ultraviolet photoelectron spectroscopy (UPS) spectra of a secondary edge region of Mo2S3, NiMo3S4, and Mo2S3@NiMo3S4 were further measured to characterize the interfacial charge polarization (Figure S1, Supporting Information). The work function (W) corresponding to the energy difference between Fermi level (E f) and vacuum level was calculated according to the equation of W = hν − E cutoff, where hν is the incident photon energy (21.22 eV) and E cutoff is the normalized secondary electron cutoff.[ , ] The work functions of Mo2S3, NiMo3S4 and Mo2S3@NiMo3S4 are 4.06, 3.63 and 3.78 eV, respectively, which indicates the electron transfer from NiMo3S4 to Mo2S3 through heterointerface, consistent with the energy band diagrams.

The scanning electron microscopy (SEM) and transition electron microscope (TEM) images exhibited that the as‐synthesized Mo2S3 was nanorods with a smooth surface (Figure 1e and Figure S2, Supporting Information), and the characteristic lattice fringes of 0.272 nm matched the (−202) plane of Mo2S3. The NiMo3S4 nanosheets were observed to be well distributed onto Mo2S3 nanorods (Figure 1f and Figures S3–S5, Supporting Information). The spherical aberration‐corrected TEM image with atomic resolution showed the lattice fringes with interplanar spacings of 0.321 nm, attributable to the (002) plane of NiMo3S4. The corresponding elemental mappings displayed the uniform distribution of Ni, Mo, and S (Figure S6, Supporting Information).

The crystal structure of the as‐synthesized samples was determined by the X‐ray diffraction (XRD) patterns. The representative peaks attributed to the Mo2S3 phase (JCPDS No. 40–0972) confirmed the successful formation of Mo2S3 without other impurities (Figure  ). After epitaxial growth by hydrothermal reaction, new peaks attributed to trigonal NiMo3S4 (JCPDS No. 30–0847) were observed, revealing the successful formation of Mo2S3@NiMo3S4 composites. The temperature dependent resistivity was performed to investigate the electrical conductivities (Figure 2b). The electrical resistivity of Mo2S3@NiMo3S4 was 0.014 Ω mm at 298 K, which was ahead of that of Mo2S3 (0.032 Ω mm) and NiMo3S4 (0.047 Ω mm), illustrating its excellent electrical conductivity. Remarkably, the electrical resistivities of Mo2S3 and Mo2S3@NiMo3S4 decreased with reduced temperature, indicating that both of them are metallic.[ , ] Further characterization by UPS spectra at low energy onset regions of Mo2S3, NiMo3S4, and Mo2S3@NiMo3S4 confirmed the metallic properties of Mo2S3 and Mo2S3@NiMo3S4 (Figure 2c).

Figure 2

Structural characterizations of metallic heterostructure. a–c) XRD patterns (a), temperature dependence of the resistivity (b), and UPS spectra for low energy onset regions (c) of Mo2S3, NiMo3S4 and Mo2S3@NiMo3S4. d) Ni 2p XPS spectra of NiMo3S4 and Mo2S3@NiMo3S4. e,f) Mo 3d and S 2p XPS spectra of Mo2S3 and Mo2S3@NiMo3S4. g,h) Band structure and density of states of Mo3S4 (g) and NiMo3S4 (h).

The electronic properties and valence states of the as‐synthesized Mo2S3, NiMo3S4, and Mo2S3@NiMo3S4 were analyzed by X‐ray photoelectron spectroscopy (XPS). The high‐resolution Ni 2p XPS spectra of NiMo3S4 showed that the peaks centered at 856.1 and 873.9 eV with two Ni satellite peaks could be assigned to characteristic Ni2+ spin‐orbit signals (Figure 2d). Compared with NiMo3S4, these peaks of Mo2S3@NiMo3S4 shifted slightly toward higher binding energy, indicating the partial oxidation of Ni2+. The peaks in Mo 3d XPS spectra of Mo2S3@NiMo3S4 located at 228.1/231.5 and 232.7/235.8 eV could be indexed to Mo3+ and Mo6+ components, respectively[ ] (Figure 2e), while the main peaks in S 2p XPS spectra at 161.6/163.4 eV were related to the S 2p3/2 and 2p1/2 orbitals of S2− in Mo2S3 (Figure 2f). Therefore, these results implied that the selective epitaxial growth of sulfide species can change the electronic structures and coordination environments of the metal active sites with the asymmetric electron distribution and a reduced electronic loss ability in an electrocatalytic reaction, which is conducive to the enhanced catalytic activity.[ ] DFT calculation revealed the reduced bandgap upon introduction of Ni into the typical semi‐conductor Mo3S4 crystal to form NiMo3S4, and the intensity of partial density of states of NiMo3S4 was higher than that of Mo3S4 at the Fermi level, suggesting the favorable electron mobility[ ] of the NiMo3S4 catalysts (Figure 2g,h).

To determine the valance states and local coordination structures of Mo2S3@NiMo3S4, the X‐ray absorption near‐edge structure (XANES) and extended X‐ray absorption fine structure (EXAFS) were performed. The pre‐edge feature of Ni in the Ni K‐edge XANES spectra of Mo2S3@NiMo3S4 shifted slightly to a higher energy than that of NiMo3S4 (Figure  ), confirming the oxidation of Ni,[ ] consistent with the XPS results. The k3‐weighted EXAFS spectra in Figure 3b showed that the peaks of Mo2S3@NiMo3S4 (1.9 Å) and NiMo3S4 (2.0 Å) shift significantly to the shorter radial distance, compared with the Ni foil (2.3 Å, Ni–Ni coordination), confirming the presence of Ni–S coordination.[ ] The decrease in oscillation intensity oscillation curves of Ni K‐edge for Mo2S3@NiMo3S4 of Mo2S3@NiMo3S4 revealed an increase in disorder due to the existence of two crystal phases (Figure 3c).

Figure 3

Electronic structures of metallic heterostructure. a–c) Ni K‐edge XANES spectra (a), R‐space EXAFS spectra (b) and corresponding oscillations (c) of Ni foil, NiMo3S4 and Mo2S3@NiMo3S4. d–f) Mo K‐edge XANES spectra (d), R‐space EXAFS spectra (e) and corresponding oscillations (f) of Mo foil, Mo2S3 and Mo2S3@NiMo3S4. g,h) Wavelet transforms for the k3‐weighted Ni K‐edge EXAFS (g), and Mo K‐edge EXAFS (h).

The Mo K‐edge XANES spectra displayed that the pre‐edge feature of Mo in Mo2S3@NiMo3S4 shifted slightly to lower energy than that of Mo2S3 (Figure 3d), indicating the electron transfer from NiMo3S4 to Mo2S3. The R‐space shows the coordination peaks at ≈1.90 and ≈2.50 Å, which could be assigned to the Mo–S and Mo–Mo peaks (Figure 3e). Compared with the Mo–S peak of Mo2S3 (1.85 Å), the characteristic Mo–S peak of Mo2S3@NiMo3S4 shifted to a longer radial distance (1.90 Å), suggesting the electron transfer from Ni to Mo. The oscillation curves of Mo K‐edge for Mo2S3@NiMo3S4 and Mo2S3 also revealed the decreased oscillation intensity of Mo2S3@NiMo3S4 (Figure 3f). The wavelet transformed (WT) Ni K‐edge EXAFS oscillation was further performed to reveal the atomic dispersion.[ , , , , ] Compared with the maximum intensity of Ni foil (Ni–Ni contribution) at ≈2.3 Å, the maximum intensities of Mo2S3@NiMo3S4 and NiMo3S4 at ≈1.9 and ≈2.0 Å can be attributed to the Ni–S contributions (Figure 3g). Meanwhile, the maximum intensities of Mo2S3@NiMo3S4 and Mo2S3 at ≈2.30 and ≈2.35 Å, smaller than that of Mo foil (2.45 Å) suggested the presence of the Mo–S contributions (Figure 3h).

Electrocatalytic OER and HER Performance

The electrocatalytic performances of Mo2S3, NiMo3S4, and Mo2S3@NiMo3S4 supported on nickel foam were evaluated by the linear scan voltammogram (LSV) in O2‐saturated 1.0 m KOH solution at room temperature (Figure  ). The Mo2S3@NiMo3S4 required the overpotentials (η) of 173, 256 and 390 mV to deliver 10, 100 and 1000 mA cm−2, respectively, which were considerably lower than that of Mo2S3 (286, 370, and 620 mV), NiMo3S4 (270, 342, and 546 mV), and commercial RuO2 (306, 475, and >800 mV). It might be induced by the weak proton adsorption capacity of Mo2S3@NiMo3S4. To investigate the OER kinetic mechanism, the Tafel plots of Mo2S3, NiMo3S4 and Mo2S3@NiMo3S4 were performed (Figure 4b). Compared with Mo2S3 (66.6 mV dec−1) and NiMo3S4 (55.6 mV dec−1), Mo2S3@NiMo3S4 achieved the lowest Tafel slope of 33.7 mV dec−1, indicating the fast OER kinetics. The chronoamperometry curves of Mo2S3@NiMo3S4 were tested to evaluate the durability. No obvious change at the current densities of 100 and 500 mA cm−2 over 48 h exhibited superior long‐term stability (Figure 4c).

Figure 4

Electrocatalytic OER and HER performance. a,b) OER polarization curves (a) and corresponding Tafel plots (b) of Mo2S3, NiMo3S4, and Mo2S3@NiMo3S4 in 1 m KOH. c) OER chronoamperometry curves of Mo2S3@NiMo3S4 at 100 and 500 mA cm−2 in 1 m KOH. d,e) HER polarization curves (d) and corresponding Tafel plots (e) of Mo2S3, NiMo3S4, and Mo2S3@NiMo3S4 in 1 m KOH. f) HER chronoamperometry curves of Mo2S3@NiMo3S4 at 100 and 500 mA cm−2 in 1 m KOH. g) The overpotentials required at 10, 100 and 1000 mA cm−2 of Mo2S3, NiMo3S4 and Mo2S3@NiMo3S4 for OER and HER. h) Comparisons of kinetics (Tafel slope) and activities (the overpotential at 10 mA cm−2) for OER and HER.

To investigate the electron transfer during OER, the in situ electrochemical impedance spectra (EIS) of Mo2S3, NiMo3S4, and Mo2S3@NiMo3S4 at different potentials were measured (Figure S7, Supporting Information). When the applied voltage increased to 1.35 V (versus RHE), an obvious mutation occurred in the Nyquist plots, indicating that OER began to occur. The lower charge transfer resistance (R ct) of Mo2S3@NiMo3S4 than that of Mo2S3 and NiMo3S4 demonstrated an enhanced charge transfer property of metallic heterostructure. In addition, the electrochemical double‐layer capacitance (C dl) was calculated through cyclic voltammetry (CV) curves to estimate the permittivity and the electrochemically active surface area (ECSA)[ , , ] (Figures S8 and S9, Supporting Information). Compared with Mo2S3 (17.1 mF cm−2) and NiMo3S4 (20.3 mF cm−2), Mo2S3@NiMo3S4 has a C dl of 24.8 mF cm−2, even superior to commercial RuO2 (Figures S10 and S11, Supporting Information), suggesting the superior kinetics of metallic heterostructure with exposed abundant active sites for large‐current operation.

The HER performance of Mo2S3@NiMo3S4 was obtained by LSV in N2‐saturated 1.0 m KOH solution at room temperature, using Mo2S3, NiMo3S4, and commercial Pt/C as the reference samples (Figure 4d). The Mo2S3@NiMo3S4 displayed a very low overpotential of 32, 124 and 174 mV at 10, 100 and 1000 mA cm−2, which was superior to Mo2S3 (77, 251, and 594 mV), NiMo3S4 (54, 209, and 415 mV) and commercial Pt/C (41, 140, and 254 mV). A small Tafel slope, 41.4 mV dec−1 of Mo2S3@NiMo3S4 was achieved in comparison with Mo2S3 (83.1 mV dec−1) and NiMo3S4 (72.7 mV dec−1), demonstrating the rapid HER kinetics (Figure 4e). Moreover, the Mo2S3@NiMo3S4 can maintain the HER activities at 100 and 500 mA cm−2 after stability measurements over 48 h, showing outstanding durability (Figure 4f).

The EIS spectra of Mo2S3, NiMo3S4, and Mo2S3@NiMo3S4 were also collected during HER at different overpotentials (Figure S12, Supporting Information). The R ct of Mo2S3@NiMo3S4 was much lower than those of Mo2S3 and NiMo3S4. A large C dl, 36.9 mF cm−2 was achieved for Mo2S3@NiMo3S4, much higher than that of Mo2S3 (13.9 mF cm−2), NiMo3S4 (18.4 mF cm−2) and commercial Pt/C (24.6 mF cm−2) (Figures S13–S16, Supporting Information). Accordingly, the Mo2S3@NiMo3S4 had a highly electrocatalytic performance for both OER and HER in alkaline media, and can even be operated in a large‐current density of 1000 mA cm−2 with a low overpotential of 390 mV for OER and 174 mV for HER (Figure 4g), much lower than those of Mo2S3 (620 and 594 mV) and NiMo3S4 (546 and 415 mV). The fabricated Mo2S3@NiMo3S4 catalyst had comprehensive advantages in low overpotential for both OER (Table S1, Supporting Information) and HER (Table S2, Supporting Information), which was superior to commercial RuO2, Pt/C, and the most reported recent electrocatalysts (Figure 4h). Based on the above results, Mo2S3@NiMo3S4 showed the remarkably high OER and HER electrocatalytic performance in alkaline media, which was mainly attributed to the interfacial electronic engineering in metallic heterostructure with high activity and conductivity.

Overall Water Splitting with Enhanced Reaction Dynamics

To investigate the reaction dynamics, the Bode phase plots of the in situ EIS measurements were performed on Mo2S3, NiMo3S4, and Mo2S3@NiMo3S4. An apparent transition peak was observed on Mo2S3@NiMo3S4 at the potential of 1.40 V (versus RHE) for OER, which was significantly lower than that of Mo2S3 (1.55 V) and NiMo3S4 (1.50 V), suggesting the faster reaction dynamics (Figure  ). Meanwhile, compared with Mo2S3 (40 mV) and NiMo3S4 (30 mV), the Mo2S3@NiMo3S4 exhibited an earlier transition peak at the overpotential of 5 mV for HER, demonstrating the rapid electron transfer (Figure 5b). A symmetric electrolyzer of Mo2S3@NiMo3S4 was further assembled for overall water splitting in 1.0 m KOH solution. The Mo2S3@NiMo3S4 (+,−) couple exhibits a low cell voltage of 1.56, 1.64 and 1.67 V at 100, 500 and 1000 mA cm−2, respectively, far superior to that of commercial electrocatalyst RuO2||Pt/C (Figure 5c). The Mo2S3@NiMo3S4 (+,−) electrolyzer can retain outstanding overall water splitting performance with no noticeable degradation at current densities of 100, 500 and 1000 mA cm−2 over 100 h with the low cell voltage (Figure 5d), which is currently one of the best bifunctional electrolyzers among non‐noble materials[ , , , , , , , , , , , ] (Figure 5e and Table S3, Supporting Information).

Figure 5

The in situ EIS measurements and overall water splitting performance. a,b) Bode phase plots of the in situ electrochemical impedance spectra of Mo2S3, NiMo3S4, and Mo2S3@NiMo3S4 for OER (a) and HER (b). c) Polarization curves by a two‐electrode system of Mo2S3@NiMo3S4||Mo2S3@NiMo3S4 and RuO2||Pt/C. d) Chronoamperometric tests of Mo2S3@NiMo3S4||Mo2S3@NiMo3S4 at 1.56, 1.64, and 1.67 V in 1 m KOH. e) Comparisons of the cell voltages at 500 and 1000 mA cm−2 of Mo2S3@NiMo3S4 with reported bifunctional electrocatalysts.

To explore the effect of metallic heterostructure on the durability of electrocatalytic reaction in alkaline medium, we fabricated the NiS2 nanoparticles for comparison. During the OER process, NiOOH would also be formed on the surface of NiS2 catalyst. However, after several cycles, most of the pristine NiS2 particles would convert into NiOOH, and S was dissolved into the electrolyte (Figure S17, Supporting Information). The XRD patterns of Mo2S3@NiMo3S4 before and after OER show no significant change, indicating that the structure is stable during water splitting (Figure S18, Supporting Information). The construction of metallic heterostructure with a strong interaction of Mo with highly active Ni atoms could improve the structural stability for large‐current operations. Moreover, the introduction of the metallic Mo2S3 could not only increase the electrical conductivity but also provide abundant defective sites at the heterointerface, thus promoting HER activities. Meanwhile, the strong interaction between Ni and Mo in the heterointerface of the catalyst was capable of increasing the interfacial electronic densities and reducing the protons‐adsorption energy, favorable for the enhanced catalytic activities.

Active Sites for Oxygen Evolution

To further investigate the catalytic active sites of the Mo2S3@NiMo3S4, the characterizations of structure, surface information, valance states and local coordination were studied after OER. The high‐resolution Ni 2p XPS spectrum displays two peaks located at 856.3 (Ni–S) and 862.0 eV (Sat.), positively shifting to higher energy attributable to the formation of Ni3+ oxyhydroxides species (Figure  ). The S 2p XPS results showed no distinct change as shown in Figure 6b. Furthermore, the Ni K‐edge XANES spectrum of Mo2S3@NiMo3S4 exhibited the distinct shift of Ni pre‐edge to higher energy after OER, demonstrating the oxidation of Ni to a higher valence state (Figure 6c). Compared to the original Mo2S3@NiMo3S4 (1.9 Å), the main peak of Mo2S3@NiMo3S4 after OER in k3‐weighted EXAFS spectra shifts significantly to the shorter radial distance (1.7 Å) attributed to the existence of Ni–O coordination, suggesting the formation of Ni oxyhydroxides (Figure 6d).

Figure 6

Active sites for oxygen evolution. a,b) High‐resolution XPS of Ni 2p (a) and S 2p (b) spectra of Mo2S3@NiMo3S4 before and after OER. c,d) Ni K‐edge XANES spectra (c) and R‐space EXAFS spectra (d) of Mo2S3@NiMo3S4 before and after OER. e) STEM and corresponding elemental mapping images of the Mo2S3@NiMo3S4 after OER. f) In situ Raman spectra of the Mo2S3@NiMo3S4 at various OER potentials.

The STEM image, energy dispersive X‐ray spectroscopy (EDS) line scan, and corresponding elemental mappings of the Mo2S3@NiMo3S4 after OER displayed that an amorphous layer of Ni oxyhydroxides was formed on the surface (Figure 6e and Figure S19, Supporting Information). To further confirm the structure transformation during water splitting, the in situ Raman measurement was carried out to clarify the real‐time evolution of the Mo2S3@NiMo3S4 catalyst during OER (Figure 6f). From 0 to 1.2 V, Raman spectra showed only two peaks at 320 and 401 cm−1, which belong to Mo–S. Another two new peaks at 474 and 551 cm−1 at 1.3 V, corresponding to Ni3+–O bending peak and Ni3+–O stretching peak, respectively revealed the transformation into NiOOH from 1.3 V. When the potential exceeded 1.4 V, the peak intensities of NiOOH increased significantly. Combined with the above results, a thin amorphous layer of NiOOH was verified to form on the surface of Mo2S3@NiMo3S4 during OER, which facilitates the overall water splitting.

Conclusion

In summary, we developed a metallic heterostructure, Mo2S3@NiMo3S4 with enhanced electron transfer properties, fast reaction dynamics and superior structural stability to boost the water electrolysis at large‐current densities by synergistically modulating phase structure via Mo and Ni incorporation. The constructed heterostructured Mo2S3@NiMo3S4 achieves an extraordinarily low overpotential of 173 mV for OER and 32 mV for HER at 10 mA cm−2. The cell voltage of Mo2S3@NiMo3S4 couples’ electrolyzer at a current density of 1000 mA cm−2 is 1.672 V, and the peak current density remains stable after chronoamperometric test over 100 h, which is among the best non‐noble metal‐based overall water splitting electrocatalysts as yet. Based on the XAFS, in situ Raman, and in situ Bode phase plots, the excellent catalytic performance can be attributed to the following aspects: i) the in situ epitaxial growth of NiMo3S4 nanosheets on Mo2S3 nanorods provides abundant active sites; and ii) the metallic conductivity of Mo2S3@NiMo3S4 improves electron transport and reaction dynamics. Therefore, the designed Mo2S3@NiMo3S4 with much‐boosted electrocatalytic performance and excellent durability is a promising electrocatalyst for overall water splitting even at large‐current densities, which opens up new opportunities for developing efficient and stable electrocatalysts to meet the industrial demand in future.

Experimental Section

Synthesis of Mo2S3

The Mo2S3 was synthesized via the high temperature molten salt method. Mo powder, S powder and NaCl powder were mixed in the molar ratio of 2:3:100, then they were ground thoroughly in a mortar. The mixed powders were pressed into a pellet and sealed in an evacuated quartz tube, which was transferred into a muffle furnace and annealed at 950 °C for 2000 min with a heating rate of 5 °C min−1. After calcination, the quartz tube was quickly taken out and cooled in cold water. The sample was collected and soaked in water to remove NaCl to obtain the Mo2S3 sample. The as‐obtained samples were dried in a vacuum oven at room temperature.

Synthesis of Mo2S3@NiMo3S4

100 mg of Ni(NO3)2·6H2O and 50 mg of Mo2S3 were added to 20 mL of distilled water in a beaker under magnetic stirring. After ≈5 min, 2 mL of oleylamine and 10 mL of ethanol were quickly added and the stirring continued for 0.5 h to produce a homogeneous solution, which was then transferred into a 50 mL Teflon‐lined autoclave. The autoclave was sealed and maintained at 180 °C for 15 h in a convection oven and then naturally cooled to room temperature. The sample was collected and washed with cyclohexane, distilled water and ethanol to remove organics, ions and possible remnants, and dried in a vacuum oven at room temperature. The freshly obtained sample and 270 mg of (NH4)2MoS4 were added to 20 mL of DMF followed by stirring for 15 min under ambient conditions. Then, 0.1 mL of N2H4·H2O was added to the suspension. After stirring for another 15 min to dissolve completely, the mixed solution was transferred to a 50 mL Teflon‐lined autoclave. The autoclave was sealed and maintained at 200 °C for 15 h in a convection oven and then cooled to room temperature naturally. The resulting black sample was collected and washed with ethanol four times and dried in a vacuum oven at room temperature.

Material Characterization

Scanning electron microscope (SEM) images were obtained by JEOL JSM6510. TEM images were obtained by JEOL JEM‐ARM300F. The spherical aberration‐corrected atomic resolution TEM images were obtained by Hitachi‐HF5000. X‐ray diffraction (XRD) characterization was carried out by Bruker D8 advanced diffractometer operating with Cu Kα radiation. The electrical properties measured in the temperature range of 230–300 K were performed using a Quantum Design physical properties measurement system (PPMS). XPS measurements were run on a Thermo Scientific Escalab 250Xi. UPS measurements were performed on a Thermo Scientific Escalab 250Xi. The X‐ray absorption spectra (XAS) including XANES and EXAFS of the sample were collected at the Beamline of TPS44A1 in National Synchrotron Radiation Research Center (NSRRC), Taiwan. Raman spectra were obtained using a thermal dispersive spectrometer with laser excitation at 633 nm.

Electrochemical Experiments

The electrochemical experiments were carried out in a three‐electrode cell using a CHI760E electrochemical workstation at room temperature. The ink was prepared by dispersing 5 mg of the catalyst and 100 µL of 5 wt% Nafion solution in ethanol (900 µL) by ultrasonication for 30 min to form a homogeneous dispersion. Then, 100 µL of the dispersion was drop‐casted onto a nickel foam (NF), leading to a catalyst loading of 2.5 mg cm−2. During the electrochemical measurements, carbon rod and Hg/HgO electrode were used as the counter and reference electrodes, respectively. The potentials reported in this work were calibrated to the RHE. In 1.0 m KOH, E (RHE) = E (Hg/HgO) + 0.059 × pH + 0.098 V. The overpotential (η) was calculated according to the formula: η = E (RHE) − 1.23 V. The KOH electrolyte was bubbled with N2 and O2 for 0.5 h before HER and OER test. Polarization curves were recorded by LSV at a scan rate of 5 mV s−1 in the range of 1.0 to 1.8 V versus RHE for OER, 0.0 to −0.7 V versus RHE for HER. To estimate the electrochemically active surface area (ECSA) of the catalyst, CV was tested by measuring C dl between 1.10 to 1.30 V and between 0.0 to 0.2 V at various scan rates of 20–120 mV s−1 for OER and HER, respectively. The in situ Nyquist plots were measured with frequencies ranging from 100 kHz to 1 Hz, and the amplitude of 5 mV at a certain potential.

Conflict of Interest

The authors declare no conflict of interest.

Supporting information

Click here for additional data file.