The Advanced Progress of MoS₂ and WS₂ for Multi-Catalytic Hydrogen Evolution Reaction Systems

Abstract

Two-dimensional transition-metal dichalcogenides (TMDs) are considered as the next generation of hydrogen evolution electrocatalysts due to their adjustable band gap, near-zero Gibbs free energy, and lower cost compared to noble metal catalysts. However, the electrochemical catalytic hydrogen evolution performance of TMDs with two-dimensional properties is limited by innate sparse catalytic active sites, poor electrical conductivity, and weak electrical contact with the substrate. It remains challenging for the intrinsic activity of TMDs for electrocatalytic and photocatalytic hydrogen evolution reactions (HERs) to compete with the noble metal platinum. In recent years, significant development of transition metal chalcogenides, especially MoS₂ and WS₂, as catalysts for electrocatalytic and photocatalytic HERs has proceeded drastically. It is indispensable to summarize the research progress in this area. This review summarizes recent research results of electrocatalysts and photocatalysts for hydrogen evolution reactions based on two-dimensional materials, mainly including MoS₂, WS₂, and their compounds. The challenges and future development directions of two-dimensional hydrogen evolution reaction electrocatalysts and photocatalysts are summarized and prospected as well.

1. Introduction

For a long time, researchers have been dedicated to exploiting renewable and clean energy sources to pursue sustainable development and avoid catastrophic climate change. Hydrogen has the highest energy density and can be converted to enormous energy without producing harmful by-products. The near-perfect production and recycling path from solar energy to hydrogen energy has always been a great prospect and expectation. It will not only address human daily needs but also enhance the energy security of all countries and reduce international conflicts caused by the fossil energy crisis, thereby promoting harmonious global development, enhancing sustainability and reducing pollution [1,2,3]. Ultrathin MoS₂ nanosheet is a typical material in transition metal sulfide systems attracting extensive attention in recent years, possessing superior structural and photocatalytic properties, and becoming a substitute for platinum in highly active hydrogen production reactions [4,5].

Among TMD materials, MoS₂ has attracted attention due to its excellent physical, chemical, and mechanical properties [6,7], and has been widely used in electrocatalytic hydrogen evolution (HER). Early studies concluded that the catalytic HER activity of MoS₂ was deficient. In 2005, Nørskov et al. calculated ΔGH at the Mo (1010) edge of MoS₂ by using DFT, and the results showed 0.08 eV ΔG_H value at 50% hydrogen coverage, approaching the optimal (0 eV). This value is comparable to platinum, theoretically indicating that MoS₂ has great application potential in HER catalysis [8]. Subsequently, Jaramillo et al. confirmed that the catalytic performance of MoS₂ layers is positively correlated with the length from the active edge and that the inert substrate limits its potential HER performance [9]. To this end, researchers are committed to activating the inert MoS₂ substrate using strategies such as defect engineering, phase engineering [10,11,12,13,14,15,16,17,18,19], strain engineering [20,21,22,23,24], and doping engineering to improve HER catalytic performance [25,26,27,28,29,30,31,32].

In addition to being used in electrocatalytic hydrogen evolution, MoS₂ and WS₂ are also used as cocatalysts to promote photocatalysts. Due to the clean, low-cost and environment-friendly characteristics of solar energy, solar photocatalytic hydrogen production has become an effective strategy to solve the energy problem. It is challenging for a single semiconductor photocatalyst to meet all the requirements of photocatalysis due to its low utilization rate of light, easy recombination of electrons and holes, and lack of sufficient active sites [33,34,35,36,37]. Supported cocatalysts can promote charge separation and improve photocatalytic efficiency [38,39,40]. This review mainly introduces the role of MoS₂ and WS₂ as cocatalysts in photocatalytic hydrogen production, such as enhancing light absorption, promoting charge separation, increasing the active sites, etc. At the same time, the loading method of MoS₂ and WS₂ on semiconductors such as TiO₂, CdS, and ZnIn₂S₄ also are introduced [41,42,43,44,45,46].

Recent research results of electrocatalysts for hydrogen evolution reactions based on two-dimensional materials, mainly including MoS₂ and WS_2, are recapitulated. Subsequently, the research progress of MoS₂ and WS₂ as cocatalysts to promote photocatalysts for HERs are collated and summarized. The challenges and future development tendencies of two-dimensional hydrogen evolution reaction electrocatalysts and photocatalysts are summarized and prospected.

2. Structural Engineering for Electrocatalytic Systems and the EY/TEOA System

One promising approach to improve electrocatalysis and photocatalysis is to increase the density of active sites, activate the intrinsic activity of the original inert site, and adjust the electronic structure of the catalyst by structural engineering. To achieve highly efficient and economical electrocatalysis and photocatalysis, research has been dedicated to fine-tuning the structure of catalysts by defect engineering, doping engineering, phase engineering, and strain engineering.

2.1. Defect Engineering

Defect engineering shows tremendous significance in modulating the electronic properties of catalysts, bringing about surprising optimizations of physical and chemical properties to improve the intrinsic bottleneck, thereby boosting HER performance. Presently, considerable efforts have been made to induce point defects (0-dimensional defects), line defects (1-dimensional defects), surface defects (2-dimensional defects), and body defects (3-dimensional defects) into the inert plane and bulk to increase intrinsic potential activity [47,48,49,50,51,52,53,54]. To achieve efficiency and a low-cost approach to forming defects that boost HER performance, methods including chemical and physical etch, ball milling, interlayer regulation, and finely tuned synthesis steps were applied [47,48,49,50,51].

A method using a ball mill to cut MoS₂ nanosheets vertically and horizontally to form S-vacancy point defects was presented by Luo et al. [55], resulting in a significant change in the surface tension of the original MoS₂, followed by an ultrasonic irradiation process to shear the defective MoS₂ to form a single-layer MoS₂ nanomesh, as shown in Figure 1A. Approximately seven layers of the intact MoS₂ nanosheets pile up, forming a wavy edge, as shown in Figure 1B. As can be observed in Figure 1C,D, the ball-milled and ultrasonically treated multilayer MoS₂ nanosheets turn into many small pieces with holes. As shown in Figure 1E, the MoS₂ nanomesh exhibits the best catalytic activity with a low overpotential of 160 mV (V vs. RHE), compared to the overpotential of 230 mV for MoS₂ nanosheets and 208 mV for defective MoS₂ nanosheets at 10 mA cm⁻². It indicates that the defective atomic-sized pores in the basal planes of MoS₂ nanomesh can significantly increase the electrochemical HER activity. The corresponding Tafel slopes derived from the polarization curves are shown in Figure 1F. A lower Tafel slope means more efficient kinesis of H₂ evolution at a constant increase in overpotential. The MoS₂ nanomesh displays a Tafel slope of 46 mV·dec⁻¹, which is smaller than those of other catalysts except for the state-of-the-art Pt/C catalyst. Figure 1G shows that in visible light, using eosin Y (EY) as photosensitizer and triethanolamine (TEOA) as the sacrificial reagent, MoS₂ nanomesh exhibits the highest hydrogen production efficiency and considerable stability in the hydrogen evolution time process.

By adjusting and improving the nanostructure of the material, the performance of the material in catalysis can be significantly optimized [52,53,54]. Through theoretical and experimental calculations, it was found that the preparation of an unsaturated and coordinated 1D metal Mo chain on the structure of MoS₂ leads to a significant improvement of HER activity [57,58,59]. Figure 2A shows the formation of 1D Mo chains by removing S atoms from the MoS₂ atomic layer. Such 1D Mo chains act as HER catalytic active sites on the surface. In the wake of the first formed S-vacancy, the adjacent S atoms present an unstable state, and thus tend to fall out of the lattice to form more S-vacancies. As a result, a line defect can be induced with the accumulation of the S-vacancies. Also, the longer the line defect, the higher the stability. S atoms are disengaged from MoS₂ to form a quasi-one-dimensional Mo chain by inserting electrochemical sodium. The Mo chain can be observed in Figure 2B after phase correction and monochromatic processing of the Mo channel in the HRTEM image of MoS₂. It can be observed that a single Mo chain, two parallel Mo chains, and three parallel Mo chains emerged after S-vacancies formed.

The HER free energy diagram through Volmer–Heyrovsky pathways was constructed to unravel its reaction mechanism. To gain a deeper understanding of the Tafel mechanism and kinetics of H atom recombination and release of H₂, the quasi-1D Mo atomic chains were used as a representative to locate the minimum energy paths and the transition states via the nudged elastic band (NEB) method. The corresponding initial, intermediate or transition, and final states are displayed in Figure 2C. As the NEB method predicted, in Figure 2D, the quasi-one-dimensional Mo chain shows a low overpotential of 45 mV to achieve a current density of 10 mA cm⁻², outperforming commercial Pt/C near the practically meaningful high-current region. The corresponding Tafel plots in Figure 2E showed that 1D unsaturated Mo chains in Mo/MoS₂ reduce the Tafel slope from 90 mV dec⁻¹ to 40 mV dec⁻¹. The influence of Mo chains on HER can be inferred from the band structure of 2H-MoS₂ with an increasing number of quasi-1D Mo atomic chains in Figure 2F. In the presence of Mo chains, new bands appear in the gap near the Fermi level, responsible for hydrogen adsorption on unsaturated Mo atomic chains. The corresponding Kohn–Sham orbitals for the band directly below the Fermi level indicate that these states are localized around the Mo chains.

The structure of MoS₂ with the quasi-one-dimensional Mo chain is more conducive to the Tafel reaction’s rapid progression, making the H adsorption and release process more accessible. As a result, the hydrogen production rate has been dramatically improved.

A method is introduced to improve the catalytic hydrogen evolution performance of MoS₂ nanosheets by using dislocation as a line defect [61]. The synthesis diagram of a high-density misaligned flexible MoS₂ nanosheet is shown in Figure 3A; with ammonium molybdate tetrahydrate and high-concentration thiourea as precursors, MoS₂ nanowire with high dislocation density can be prepared by hydrothermal reaction at low temperature for ten days. The formation mechanism of D-S-MoS₂ is shown in Figure 3B, where [Mo₇O₂₄]²⁻ decomposed by (NH₄)₆Mo₇O₂₄·4H₂O produces two different sulfides under the action of a high concentration source of sulfur. After ten days of hydrothermal reaction without splicing, the two kinds of sulfide clusters gradually form MoS₂ nanosheets with a weak crystalline structure, which show short-range order and relatively low crystallinity. It has a porous structure and abundant dislocations (in-plane and out-of-plane). The (100) crystal phase image with a large number of dislocations obtained by FFT processing is shown in Figure 3C due to the presence of plentiful dislocations (these dislocations are represented by “T”); in-plane strain appears at the corresponding position. In addition, dislocations in one crystal phase may lead to strain effects in another crystal phase, and dislocations are not included in the same region and are marked with white squares. The synthesized D-S-MoS₂ not only has an in-plane dislocation structure but also shows an out-of-plane dislocation structure. The formation mechanism of out-of-plane dislocations is shown in Figure 3D. The strain induced by the dislocations can cause the surface of MoS₂ nanosheets to be raised or sunken, resulting in interlayer lattice distortion and a more significant Burger vector. The sharp edge dislocations indicate that they are caused by atomic slippage.

The time course of photocatalytic hydrogen evolution on D-S-MoS₂ is shown in Figure 3E, and hydrogen production linearly increases with irradiation time. The results show that it has not only good hydrogen evolution performance but also has specific stability. The Gibbs free energy of D-S-MoS₂ at different lattice compressibility is shown in Figure 3F. Dislocations usually consist of different structural motifs, of which 5/7 nuclei are more accessible to obtain due to their lower formation energy. This article uses a 5/7 dislocation core as a model. From a thermodynamic point of view, the ideal Gibbs free energy value for hydrogen adsorption is 0 eV [63]. The strain effect caused by dislocations causes the lattice to stretch or contract. In contrast, the presence of high-density dislocations causes the lattice to compress by 0.1~4%, as can be seen from the diagram when the lattice compression ratio is 0.1%, the ΔG_H of D-S-MoS₂ is about 0.42 eV and can be optimized to −0.18 eV when the lattice compression ratio is 4%. The results show that the dislocation–strain interaction is a crucial factor in improving the activity of the catalyst.

The discovery of moiré superlattices (MSLs), caused by the twisting of van der Waals force-stacked two-dimensional layered materials, [64,65]. Research on the relationship between this unique structure and electrical properties has become a hot topic in electrocatalysis [66,67,68]. Zhao et al. proposed a new method to synthesize two-dimensional MSLs [69]. A one-step solvothermal method was used to prepare helical WS₂ nanobelts with local torsion deformation, and moiré stripes were formed due to the torsion angle between the layers. A single conical tube curled from one end of the WS₂ nanobelt can be observed through a scanning transmission electron microscope, as shown in Figure 4A. WS₂ nanobelts are easily twisted by force imbalances during synthesis [55,70], forming a nano-cone at one end. The relative stress distribution can be known through finite element analysis, as shown in Figure 4B. Strain generation may activate the substrate of the nanobelts [62,71], primarily by altering the electronic structure of the active catalytic site and facilitating mass transfer. The strain generated by the distortion weakens the van der Waals forces between the corresponding multilayer WS₂ nanobelts, causing a high probability of layer slip formation. As shown in Figure 4C, honeycomb moiré patterns due to layer slip are observed through HRTEM. Further analysis of the production of moiré patterns in Figure 4D shows a simulation diagram of the atomic structure of WS₂ MSLs. Twisting can be observed between nanofilm layers at an angle of 14°. Twisting and folding can easily cause S atoms to slide in the WS₂ MSLs to produce 1T′-WS₂, for which the ΔG_H of 14° twisted bilayer 1T′-WS₂ NSs is calculated using density functional theory (DFT). The experimental data showed that the ΔG_H of 1T′-WS₂ MSLs’ basal plane site (−0.24 eV) was close to zero, indicating high activity of the HER. As an HER catalyst, WS₂ MSLs have a unique structure due to strain and phase engineering, which promotes the generation of active sites on the basal plane and dramatically reduces the ΔG_H of sites. At the same time, it also allows other TMDs to be used for MSL engineering. This research on MSLs dramatically expands the research direction of topophysical engineering based on mechanical flexibility. It provides a broad idea for designing catalysts for more diversified structures in the future.

2.2. Doping Engineering

Heteroatom doping is considered to be an effective method to enhance catalytic activity. Ni atoms are doped to replace Mo, and the induced S_V formation and phase transition will transform into a stable distorted 1T′ phase [72,73]. 1T′-MoS₂ S_V is a catalyst with a high electrical conductivity that integrates all multifunctional active sites, which is expected to enhance HER activity. Ni atomic doping can introduce S_V and trigger phase transition to control the electronic structure of MoS₂ [74,75,76], as shown in Figure 5A, including three aspects of defect engineering, phase transition, and electronic features. This method of doping low-valent metal atoms ensures more exposure to active sites at the edge of MoS₂ nanosheets. It is another important way to improve the catalytic performance of HERs.

The atomic structure of Ni-doped 1T′-MoS₂ is shown in Figure 5B. In general, a regular octahedron with Ni doping as the center and S as vertices is formed, in which S₀–S₅ represents the S atoms connected to the doping atoms. Compared with the undoped 1T′-MoS₂ structure, DFT can calculate the respective S_V formation energies. The analysis shows that the S_V formation energies around Ni atoms in Ni-doped 1T′-MoS₂ are much lower, indicating a high concentration of S_Vs is generated in the structure after doping with Ni atoms and that the conversion of semiconducting MoS₂ to metallic MoS₂ can be accelerated due to the existence of S_Vs. The primary process for the preparation of Ni-doped 1T′-MoS₂ is illustrated in Figure 5C. The preparation mainly used three crystalline substances, (NH₄)₆Mo₇O₂₄·4H₂O, NiSO₄·6H₂O, and CH₄N₂S. Firstly, a certain amount of (NH4)₆Mo₇O₂₄·4H₂O as a reducing agent and an excess amount of CH₄N₂S as a source of S are completely dissolved in a solution of NiSO₄·6H₂O as a source of Ni. Moreover, the two crystals are evenly distributed in the solution, and the desired solution can be obtained when thoroughly combined. Then the solution was dried in a vacuum oven at 200 °C for 24h, and the final drying was completed to obtain Ni-doped 1T′-MoS₂.

Transmission electron microscopy (TEM) scanning was carried out in the experiment. The Ni-doped 1T′-MoS₂ has a sheet-like structure in the nanoscale, as illustrated in Figure 5D. In order to further understand the electronic structure of Ni-doped 1T′-MoS₂, a high-precision observation instrument is used in Figure 5E. Two parallel rows of bright spots (Mo atoms) were close to each other to form a one-dimensional zigzag Mo atom chain, which also illustrates the asymmetric feature of the distorted 1T′ structure. The two red dashed lines b and c in the figure can also show in further experiments that there is a sizeable Mo-Mo gap in the structure of 1T′-MoS₂. As shown in Figure 5F, the experimental characterization of Ni-1T′-MoS₂ by high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) shows that the structural phase matches well with 1T′-MoS₂. The effect of Ni-doped 1T′-MoS₂ catalysts on the hydrogen production index under different catalyst dosages is shown in Figure 5G. The black curve in the figure represents the cumulative hydrogen yield under the influence of only solution eosin Y (EY) (without Ni-doped 1T′-MoS₂). It can be seen that the cumulative hydrogen yield is almost unchanged after 90 min, which indicates that water splitting has stopped after this. Furthermore, compared with other samples with linear function curves of Ni-doped 1T′-MoS₂ catalysts, this also means that their hydrogen production almost maintains the original growth rate throughout the experimental time. It can also be seen from the figure that the optimal sample value is 10 mg, and the hydrogen evolution value is 91.625 μmol·h⁻¹.

DFT calculations are further performed to explore the effect of Ni-1T′-MoS₂ S_V on the catalytic activity. The Gibbs free energy (ΔG_H) of hydrogen adsorption for 1T′-MoS₂ S_V and Ni-doped 1T′-MoS₂ S_V at three different active sites H₀-H₂ is shown in Figure 5H and Figure 5I, respectively. By comparing the two figures, it can be seen that the ΔG_H of the corresponding active sites decreases sharply after Ni doping. This clearly shows that Ni-doped 1T′-MoS₂ has an optimized ΔG_H and confirms that Ni-doping enhances the H adsorption capacity and thus activates the S_V. This synthetic method of doping low-valent metal atoms opens a new route to synthesize diverse active sites and enables the study of doping-dependent properties and applications.

The catalytic activity of MoS₂-based catalysts would be affected by the existing oxygen in the photocatalytic hydrogen evolution reaction [78,79,80,81]. The mechanism of the atomic oxygen doped in intrinsic flexible MoS₂ is further explored through theoretical analysis and experiments [82,83,84,85]. A Bader analysis based on the experimental data is performed at the start, as shown in Figure 6A. The oxygen substitution site has excellent electron affinity, which is beneficial for proton attraction and thus improves HER efficiency. In addition, strain engineering basis on O-doping is implemented. The synergistic effect of oxygen doping and strain engineering adjusts the catalytic activity of MoS₂-based catalysts. Figure 6B shows the influence of O-doping and strain engineering on the band gap of MoS₂. Single-layer MoS₂ is a direct band gap semiconductor, and after oxygen doping, it becomes an indirect band gap semiconductor. It can be found that oxygen doping and strain can reduce the band gap of MoS₂, and the strain could evidently change the band gap.

In order to explore the effect of oxygen doping on the catalytic hydrogen evolution activity of MoS₂ in the actual process, the change of hydrogen production rate per hour in the hydrogen evolution process is elaborately recorded through experiments, as shown in Figure 6C. The maximum hydrogen production rate of oxygen-doped MoS₂ can reach 1.6 mmol h⁻¹ g⁻¹. The catalytic activity of the catalyst tends to be stable with the increase in catalytic time, which confirms that oxygen doping improves the actual catalytic effect of MoS₂.

The internal structure and properties of MoS₂ doped with oxygen were analyzed by XPS, as shown in Figure 6D. It can be found that the formation of Mo-O bonds and O with more electron contribution ability to reduce the electron density around the substitution site, resulting in the decrease in binding energy.

Figure 6E is the formation mechanism of MoS_2−xO_x in actual photo-catalytic reactions. The sensitizer EY is irradiated by sunlight to produce excited state EY^1*, which can easily become triplet EY^3* under the action of ISC. Through the reduction quenching of TEOA, electrons and free radicals EY⁻ are produced, and electrons are easily transferred from EY⁻ to O₂ to produce highly active O₂⁻, which reacts with MoS₂ to produce MoS_2−xO_x. The top schematic of MoS₂ doped with O on the base surface is shown in Figure 6F. It is easy to observe the oxygen substitution site as the active site of the hydrogen evolution reaction. Figure 6G shows HAADF-STEM images of the lattice distortion of MoS₂ with O-doping. The formation of O-Mo bonds may be the reason for the strain in the internal structure [86,87,88]. ΔG_H is an important indicator for the hydrogen evolution reaction activity of the catalyst. In order to study the synergistic effect of oxygen doping and strain on HER performance, the DFT calculation method was used. The calculation results clearly show in Figure 6H that the ΔG_H of oxygen-doped strain MoS₂ is the lowest, only 0.3 eV. Overall, the synergistic effect of oxygen doping and strain on MoS₂ could greatly improve its catalytic hydrogen evolution rate, which provides new insights and new ideas for preparing efficient photocatalysts in the future.

2.3. Phase Engineering

As shown in Figure 7A, vertical 1T-WS₂ nanosheets were prepared on a flat Ti substrate to improve HER catalytic efficiency. The nanosheets are cross-linked, facilitating the exposure of WS₂ edge sites and the rapid transport of protons throughout the catalyst, thus enhancing H₂ evolution. Figure 7B shows that the 1T-WS₂ nanosheets are very transparent, implying that the nanosheets are ultrathin [89]. One of the white luminescent lines represents its good conductivity, which is crucial for H₂ evolution. X-ray photoelectron spectroscopy (XPS) can confirm the chemical state and composition of 1T-WS₂ and distinguish the different metallic phases of WS₂. Figure 7C indicates that the WS₂ nanosheets are a mixture of 1T and 2H phases [90,91]. The Tafel plots in Figure 7D were measured in a 0.5 M H₂SO₄ solution using a typical three-electrode cell setup. The Tafel slope is an important parameter to describe the HER activity of the catalyst. The black linear part of the figure is the Tafel plot of a vertical 1T-WS₂ nanosheet at a small overpotential, which has a Tafel slope of 43 mV dec⁻¹. The Tafel slopes of Pt/C and WS₂ nanosheets are 52 mV dec⁻¹ and 30 mV dec⁻¹, respectively. As shown in Figure 7E, the polarization curves of the 1T-WS₂ nanosheets after 5000 cycles almost overlapped with those of the initial cycles. This demonstrates that the vertical 1T-WS₂ nanosheets are very stable in the acidic environment and remain intact during repeated cycles. Figure 7F examines the ability associated with the continuous catalytic generation of H₂ from vertical 1T-WS₂ nanosheets using time measurements (j–t). This quasi-electrolytic process was carried out in 0.5 M H₂SO₄ at a constant voltage of 160 mV. Apparently, the current density curve presents almost a horizontal line in the 30 h test at a continuous current density, which indicates the ultra-high stability of the vertical 1T-WS₂ nanosheets.

2.4. Strain Engineering

Strain is one of the most common means of changing the structure of a catalyst to improve its HER performance. The electronic, magnetic, and catalytic properties of materials can be enhanced by topology [92]. The aqueous solution electrochemical activation method can promote the topological transformation of WS₂ from the 2H phase to the 1T phase [93]. The phase transition caused by this strain greatly improves the catalytic performance of the spiral-shaped 1T phase of WS₂ [94]. Figure 8A shows the relationship between the number of potential cycles and the cathode current density. It can be seen that the early current density increases significantly with the increase in the number of cycles and the current density stabilizes after 10,000 cycles. The high stability of WS₂ is maintained even after 20,000 cycles. It can be seen that WS₂ nanosheets are the ideal catalysts for long-term electrochemical processes. To explore the main reason for the above outstanding performance of WS₂, in situ Raman spectroscopy of WS₂ nanosheets was implemented, as shown in Figure 8B. It can be seen that the peaks of the Raman activity pattern corresponding to the 1T phase become sharper and more prominent. By calculating the area under the peak, it is found that the proportion of 1T is 70% and the proportion of 2H is 30% due to the strain that causes the WS₂ nanosheets to transition from the 2H phase to the 1T phase. The electrochemical activity of the WS₂ nanosheets is improved because the transition phase reduces the charge transfer resistance. As shown in Figure 8C, the morphology of the WS₂ nanosheets also evolves during the cycle. As the number of cycles increases, the WS₂ nanosheets gradually twist and rotate from the original straight to helical.

The driving force of the process from straight to helical is the strain in the WS₂ nanohelices caused by the H₂ bubbles. In Figure 8D, we introduced Gibbs free energy for comparison to understand further the effect of different strains and phases on the catalytic performance of WS₂ nanohelices. The closer the value of Gibbs free energy is to 0, the better the balance between the absorption and removal of hydrogen atoms at the active site is. The 1T phase WS₂ nanohelices at 3% strain have the best catalytic performance because the Gibbs free energy is closest to 0. For comparison, the Gibbs free energy of strain on WS₂ nanobelts in the 2H phase was also calculated, and it was found that the strain had no effect on its catalytic performance. From these data, we can draw a conclusion that the strain has caused the transformation of WS₂ nanosheets from the 2H phase to the 1T phase. The synergy of strain and phase transition leads to decreasing charge transfer resistance, increases in activity at each active site, and increases in the number of spiral-shaped active sites of WS₂ nanosheets. All these changes have a positive effect on the HER of WS₂ nanosheets.

Structural engineering offers a promising strategy to artificially regulate and control catalysts’ inherent activity in electrocatalysis and photocatalysis, thus achieving catalytic targets for specific scenarios.

Table 1. Comparison of HER performance on different catalysts modified by structural engineering.

Catalyst	Engineering Strategy	Electrolyte	η@10 mA cm−2 (mV)	Tafel Slope (mV dec−1)	Ref.
Monolayer MoS2	Defect/strain	0.5 M H2SO4	170	60	[95]
Monolayer MoS2	Defect	0.5 M H2SO4	362.4	135	[96]
MoS2 thin-film	Defect	0.5 M H2SO4	350 (η10.97)	105	[97]
MSORx	Doping	1 M KOH	63	43	[98]
MoS2/NiS2	Doping	1M KOH	62	50	[99]
Ru/np-MoS2	Strain/doping	1M KOH	30	31	[100]
F-MoS2	Strain	0.5 M H2SO4	-	59	[101]
1T′-MoS2	Phase	0.5 M H2SO4	65	100	[102]
Ni-1T-MoS2	Phase/doping	1M KOH	199	103	[103]

summarizes the performance of different catalyst modified by structural engineering in recent advances.

3. As Co-Catalyst to Compound with Various Semiconductors

The number of exposed active edges of MoS₂ nanosheets is the key to the catalytic activity of MoS₂ for HER [104,105,106]. Making single- or few-layer MoS₂ nanosheets vertically stand on a substrate with maximally exposed edge sites is a reasonable way to enhance HER catalytic activity [107]. Porous TiO₂ nanofibers with abundant defects are appropriate for MoS₂ nanosheets to grow as nucleation centers. Single- or few-layer MoS₂ nanosheets standing vertically on porous TiO₂ nanofibers (TiO₂@MoS₂) were successfully prepared by a simple hydrothermal method, as illustrated in Figure 9A. Firstly, the Na₂MoO₄ and C₂H₅NS were dissolved in deionized water to form a transparent solution. Then TiO₂ nanofibers were immersed into the above solution and stirred to obtain a suspension. The extremely tough and porous TiO₂ nanofibers with vast numbers of defects can easily interact with metallic precursors and then act as nucleation for the growth of MoS₂ nanosheets. MoS₂ embryos grew from inside to outside of the pores of TiO₂ nanofibers. The dark field FESEM-STEM images of the TiO₂@MoS₂ heterostructures are shown in Figure 9B. It can be seen that the transparent, flexible, and curly MoS₂ sheets with a lateral size of about 100 nm vertically stand on the surface of TiO₂ nanofibers. The shapes of MoS₂ nanosheets that grow vertically in the dioxide pores of TiO₂ nanofibers are similar to tender shoots growing on the surface of bamboo. The photocatalytic HER mechanism of the TiO₂@MoS₂ heterostructure in Figure 9C indicated that the growing method ensures the 2D MoS₂ nanosheets are firmly attached to the TiO₂ nanofibers. The mechanism of TiO₂@MoS₂ heterostructure can not only strengthen the contacts between TiO₂ and MoS₂ but also accelerate electron transfer in the TiO₂@MoS₂ heterostructures. To further improve the HER catalytic activity of the TiO₂@MoS₂ heterostructure, the MoS₂ nanosheets in the TiO₂@MoS₂ heterostructure were exfoliated by Li intercalation. Then, the 1T phase of MoS₂ could be obtained, as shown in Figure 9D. Figure 9E indicated that the excited S vacancies and strain were the main reason for the stability of the 1T phase by reducing the surface energy.

As shown in Figure 9F, the amount of H₂ evolution of TiO₂@MoS₂ heterostructure after exfoliation showed a nonlinear increase compared with the TiO₂@MoS₂ before exfoliation. The phenomena indicate that the active sites of the TiO₂@MoS₂ after exfoliation had been optimized during the process of PHER. To uncover the self-optimization mechanism, an enlarged HRTEM image of MoS₂ in exfoliated TiO₂@MoS₂ after seven hours of HER was obtained as shown in Figure 9G. The distorted 1T′ phase with the super-lattice structure from Mo atom clustering into zigzag chains can be clearly observed. Figure 9H shows eight S sites with different H adsorption coverages. The relationship between Gibbs free energy and strain of 1T′ phase was conducted in Figure 9I. It is easy to find that 1T′ phase is highly active for the HER and the strain has huge influences on the catalytic activity of the MoS₂ basal plane.

To achieve idealized photocatalytic H₂ evolution performance under visible light irradiation, the TiO₂ nanofibers of TiO₂@MoS₂ could be replaced by narrow bandgap semiconductor CdS nanowires to design a CdS/MoS₂ heterojunction. The CdS/MoS₂ heterojunction was produced by a simple hydrothermal reaction with or without adding glucose. With the help of glucose, the MoS₂ nanosheets grow nearly vertically on the CdS nanowires, which allows a high exposure of the active edge sites and improves the charge separation and transfer rates. In addition, glucose can make MoS₂ thinner and can act as a binder to help MoS₂ nanosheets grow along the longitudinal axis, simultaneously forming 1D/2D nearly vertical CdS/MoS₂ heterojunctions for efficient photocatalytic hydrogen evolution. Figure 10A shows the synthesis process of CdS/MoS₂(G) and CdS/MoS₂(W) heterojunctions by a two-step solvothermal approach. As shown in Figure 10B,C, the MoS₂ nanosheets are grown nearly perpendicularly to the CdS nanowires with an acid of glucose, while MoS₂ nanosheets are discrete connected to the CdS nanowires without glucose. The TEM image of CdS/MoS₂ heterojunctions in Figure 10D further indicates that MoS₂ nanosheets grow almost vertically on the outer surface of CdS nanowires. Figure 10E shows that the layer number of MoS₂ nanosheets is almost 1–2 layers, indicating a reduced tendency of MoS₂ nanosheet agglomeration.

Figure 10F shows the UV diffuse reflectance spectra (DRS) of CdS, MoS₂, CdS/MoS₂(W) and CdS/MoS₂(G) samples. The light-harvesting ability of the CdS/MoS₂(G) sample is higher than that of the CdS/MoS₂(W) sample in both the UV and visible regions, which indicates that the MoS₂ nanosheets have better adhesion on the CdS nanowires and the open near-vertical structure of the CdS/MoS₂(G) sample has almost no effect on the light absorption of the CdS core. Figure 10G shows the photocatalytic hydrogen production activity of CdS/MoS₂(G) catalysts with different MoS₂ loadings under visible light irradiation. With the increasing MoS₂ content in the heterojunction, the H₂ evolution rate of CdS/MoS₂(G) increases. Even if the loading of MoS₂ on the CdS/MoS₂(G) heterojunction is 1 wt%, the hydrogen production rate is higher than that of pure CdS. The hydrogen production performance achieves the best maximum value of 9.73 mmol h⁻¹ g⁻¹ when the MoS₂ loading contents reaches 10 wt%., It is 36.04 times higher than that of pure CdS nanowires. However, the degradation of photocatalytic performance appeared when the MoS₂ content exceeded 10 wt%, the decline being due to the excessive MoS₂ hindering the absorption of visible light by CdS. In addition, the high overlap of MoS₂ also reduces the exposure of active sites, thus weakening photocatalytic H₂ evolution. Figure 10H shows that the H₂ evolution rate of CdS/MoS₂(G) is the highest and much higher than that of CdS/MoS₂(W). This indicates that the almost vertical MoS₂ nanosheets on CdS/MoS₂(G) maximally expose the HER active edge, thus enhancing H₂ evolution. In contrast, the MoS₂ nanosheets on CdS/MoS₂(W) with random orientation and severe aggregation inhibit H₂ evolution. In addition, the physical mixture of CdS nanowires and MoS₂ nanosheets showed poor HER activity, indicating that the close contact between MoS₂ and CdS is an important factor for H₂ evolution. The CdS/MoS₂(G) catalyst exhibits the best hydrogen production rate due to the near-vertical nanostructure and the tight heterogeneous interface. Figure 10I show that the CdS/MoS₂(G) heterojunction was tested for five reaction cycles under visible light irradiation. The test data indicate that the introduction of MoS₂ has the function of inhibiting the photocorrosion of CdS, which leads to the good stability of CdS/MoS₂(G). The nearly vertical and open structure can provide an optimal electron transport pathway and a high density of active sites; thus, the electron-hole pair separation and photocatalytic performance are improved.

To further enhance the photocatalytic hydrogen production activity of TiO₂@MoS₂ heterostructure, zero-dimensional CdS nanocrystals acting as the light-harvesting centers could be loaded onto the MoS₂ nanosheets in TiO₂@MoS₂ heterostructure to broaden the light absorption range of the catalyst. The fabrication process of the TiO₂@MoS₂@CdS heterostructure with a “cauline-leaf” structure is shown in Figure 11A. This delicate structural design helps to increase the specific surface area and enhance light capture. The TEM image of TiO₂@MoS₂@CdS is shown in Figure 11B. The MoS₂ nanosheets grow vertically on TiO₂ nanofibers, and the CdS nanoparticles are uniformly distributed on the surface of the MoS₂ nanosheets. The presence of a single or few layers of MoS₂ nanosheets in Figure 11C indicates that the MoS₂ nanosheet is ultrathin. The good crystallinity of MoS₂ and CdS could be confirmed by clear lattice stripes in Figure 11D. The apparent lattice distortion at the interface can indicate that the perfect heterostructure is formed between MoS₂ and CdS.

The photocatalytic hydrogen production activity of TiO₂@MoS₂@CdS with different CdS content under simulated sunlight (320–780 nm) or visible light (>420 nm) irradiation is shown in Figure 11E. When the content of CdS reached 60 wt%, the highest H₂ yield was 12.3 or 6.2 mmol h⁻¹ g⁻¹. In the simulated sunlight irradiation, the photocatalytic HER mechanism of TiO₂@MoS₂@CdS is shown in Figure 11F. The multistage biomimetic cauline-leaf structure facilitates light capture and absorption by photon scattering and multiple absorption effects. The gradient of energy levels between different semiconductors promotes the transfer of photogenerated carriers. The photogenerated electrons are eventually transferred to the enriched edge-exposed MoS₂ to produce hydrogen. This delicate structure design is crucial for the development of photocatalysis. This perfect “cauline-leaf” design is essential for developing photocatalytic materials.

The key to photocatalytic efficiency for hydrogen evolution reactions lies in the development and utilization of catalytic materials [108,109,110,111]. Among many semiconductor photocatalysts, ZnIn₂S₄ has attracted the attention of researchers due to its stable chemical properties, narrowband gap, and excellent photoelectric properties [112,113,114]. However, ZnIn₂S₄ has a high recombination rate of photogenerated carriers in the hydrogen evolution reaction, which could be solved by loading MoS₂ as cocatalysts [115,116]. The main active sites of MoS₂ are located at the edge, while this sandwich-layered structure formed by van der Waals force leads to its very inert basal plane. Ni atoms doped in MoS₂ would activate its inert basal plane. Firstly, ZnIn₂S₄ and Ni-Mo-S were synthesized by a simple solvothermal and hydrothermal method, and the preparation process is shown in Figure 12A. The prepared ZnIn₂S₄ and Ni-Mo-S powders were uniformly mixed by grinding in a specific proportion in a mortar. Finally, ZnIn₂S₄/Ni-Mo-S composite materials were formed.

In order to meet the needs of photocatalytic hydrogen production under different media environments, three hydrogen evolution experimental systems were designed under neutral, alkaline and acidic conditions. To evaluate the HER catalytic activity of Ni-Mo-S nanosheets in the neutral medium environment, the prepared Ni-Mo-S nanosheets were used as the photocatalyst, EY was used as the sensitizer, and TEOA was used as the hole scavenger. Figure 12B shows the photocatalytic hydrogen evolution process of EY/Ni-Mo-S in TEOA aqueous solution under visible light irradiation. As a sensitizer with excellent photocatalytic hydrogen production performance, EY can absorb visible light from the ground state to single-line excited state EY^1* after being excited by light with energy greater than the band gap. The single-line excited state EY^1* has a short lifetime, which is easy to compound and causes fluorescence quenching. However, the single-line excited state electrons are easy to transform into triple excited state EY^3* with a long lifetime through inter-system crossing (ISC) through gap hopping. Subsequently, electrons were extracted from TEOA by reductive quenching, resulting in free radicals EY⁻ and TEOA⁺. It is easy to transfer electrons from highly reducible EY⁻ to the conduction band of Ni-Mo-S nanosheets, resulting in the spatial separation of photogenerated charges and inducing H⁺ reduction from H₂O decomposition to generate excited state H*. Two excited states H* react with each other to produce H₂. At the same time, the holes generated on the valence band of Ni−Mo−S nanosheets will be consumed by TEOA to generate the oxidation donor (TEOA⁺).

In terms of acidic and alkaline media environments, we studied the hydrogen evolution reaction mechanism under the conditions of Na₂S/Na₂SO₃ and lactic acid as hole sacrificial agents by using visible light irradiation on ZnIn₂S₄/Ni-Mo-S composites. We can observe the carrier migration in different media environments in the photocatalytic process in Figure 12C,D, respectively. Under visible light irradiation, ZnIn₂S₄ generates photogenerated carriers from the valence band (VB) to the conduction band (CB). However, because the conduction band of Ni-Mo-S is higher than that of ZnIn₂S₄, the photogenerated electrons of ZnIn₂S₄ transfer easily to Ni-Mo-S. In ZnIn₂S₄/Ni-Mo-S composites, which are closely bound, the photogenerated carriers can be effectively separated, and the recombination and other consumption of photogenerated carriers can be avoided. The H⁺ decomposed by H₂O is reduced to the excited state H^* under the action of electrons, and then the H^* of the two excited states are combined to produce H₂.

For the photocatalytic properties of Ni-Mo-S/EY with different Ni doping amounts, the experimental results can be obtained from Figure 12E. Without any semiconductor as a catalyst, photogenerated carrier recombination rate is high, and pure EY dye shows low photocatalytic hydrogen evolution activity. When pure MoS₂ (Ni:Mo = 0) was used as a catalyst, the hydrogen evolution rate of EY/MoS₂ under visible light was 1.25 mmol h⁻¹ g⁻¹. When the ratio of Ni:Mo in Ni-Mo-S was increased, the amount of hydrogen evolution increased. Through experiments, it was found that the catalytic performance was better when the ratio of Ni:Mo in Ni-Mo-S was 1:1. The maximum production rate of H₂ reached 2.26 mmol h⁻¹ g⁻¹, 2.72 times that of pure EY dye photocatalytic hydrogen evolution, indicating the superiority of EY-sensitized Ni-Mo-S in photocatalytic hydrogen evolution.

The research results of the hydrogen evolution reaction of ZnIn₂S₄/Ni-Mo-S composites in different media environments can be obtained from Figure 12F,G. Pure ZnIn₂S₄ has the lowest hydrogen production performance, which is attributed to its high photogenerated carrier recombination on the surface. When MoS₂ is added as a cocatalyst, the hydrogen production performance is only slightly improved [117,118,119]. The reason is that the inert basal plane of MoS₂ exists, and the active sites still need to be more numerous, so the experimental hydrogen evolution performance is still limited. When Ni atoms doped in MoS₂ were used as cocatalysts, the hydrogen production performance was significantly improved. Through experimental comparison, it is found that when the value of Ni:Mo in Ni−Mo−S nanosheets is 1:2, the hydrogen evolution performance is the best, and the maximum hydrogen evolution rate reaches 3.63 mmol h⁻¹ g⁻¹, which is 2.33 times that of pure ZnIn₂S₄. Compared with the lactic acid system, the hydrogen production rate of pure ZnIn₂S₄ in the Na₂S/Na₂SO₃ system was higher (1.56 mmol h⁻¹ g⁻¹), indicating that Na₂S/Na₂SO₃ was more suitable to be used as hole scavenger under the ZnIn₂S₄ system. The maximum hydrogen evolution rate of ZnIn₂S₄/Ni-Mo-S is 6.07 mmol h⁻¹ g⁻¹, which is 5.28 times the hydrogen production rate of pure ZnIn₂S₄. It is proved that the precise design of ZnIn₂S₄/Ni−Mo−S composite system in an acidic medium can significantly improve the performance of photocatalytic hydrogen evolution. ZnIn₂S₄ and MoS₂ with two-dimensional properties have great potential in the field of photocatalysis due to their large surface area and many active sites. In general, the ZnIn₂S₄/Ni-Mo-S composite materials prepared through experiments showed excellent performance in different media environments, which provided new ideas for the further development of new photocatalytic materials based on ZnIn₂S₄ and MoS₂.

Various strategies of photocatalysis using transition metal sulfide as a cocatalyst have achieved profound development and great progress. MoS₂ as co-catalyst to compound with various semiconductors had been widely applied in boosting photocatalytic hydrogen evolution performance, by exposing the reactive facets or edge sites, designing heterophase/heterofacet junctions and engineering the interfacial coupling and confinement effect. MoS₂ compounds with different semiconductors to strengthen photocatalytic performance are summarized in

Table 2. Comparison of photocatalytic performance on MoS₂ compound with semiconductors.

Combine with Semiconductor	Light Source	Reaction Condition	Photocatalytic Performance	Ref.
MoS2 with ZbIn2S4	300 W xenon lamp (>420 nm)	0.25 M Na2SO3 and 0.35 M Na2S solution	~117 μmol h−1 g−1	[120]
MoS2 with Fe3O4	100 W halogen lamp	15% v/v triethanolamine solution	2480 μmol h−1 g−1	[121]
MoS2 with CdS	300 W xenon lamp (>400 nm)	30 vol% lactic acid solution	29 mmol h−1 g−1	[122]
MoS2 with CdS	300 W xenon lamp (>400 nm)	30 vol% lactic acid solution	31 mmol h−1 g−1	[123]
MoS2 with CdS	300 W xenon lamp (>300 nm)	20 vol% lactic acid solution	1.695 mmol h−1 g−1	[124]
MoS2 with CdSe	300 W xenon lamp (>400 nm)	0.1 M Na2SO3 and 0.1 M Na2S solution	890 μmol h−1 g−1	[125]
MoS2 with AgInZnS	300 W xenon lamp (>420 nm)	10 vol% lactic acid solution	944 mmol h−1 g−1	[126]
MoS2 with TiO2	300 W xenon lamp	20% methanol solution	4300 μmol h−1 g−1	[127]
MoS2 with CdS	300 W xenon lamp (350–800 nm)	20% methanol solution	580 mmol h−1 g−1	[128]

.

4. Conclusions and Outlook

So far, the application of photocatalysis technology is still only in laboratory-scale research, and the application of the actual industrial equipment is faced with significant technical barriers, which limits the application of photocatalysis technology [129,130,131,132]. Therefore, in the following research, it is urgent to carry out research continuously strengthening the understanding of the theories related to photocatalysis, such as the development of more advanced material preparation methods and characterization technology [133,134,135]; to strengthen interdisciplinary ideas and further explore photocatalytic technology in the process of integrating with different disciplines; develop new catalysts that are efficient, stable, safe, economical, and green, and suitable for industrial hydrogen production [136,137].

In the follow-up work, the photocatalysis mechanism under multi-field coupling of light, heat, electricity, and force, such as the principle of carrier separation, will be further studied. The dynamic mechanism of energy capture, efficient conversion, and storage under multi-field coupling should be revealed. The traditional mode of single matching of thermal kinetic energy and energy potential of heat flux, independent use of energy carriers, and the simple thermal cycle will be transformed into the new mode of overall matching of multiple physical and chemical energy and energy potential of energy-carrying substances (such as the overall matching and step utilization of light, heat, and chemical energy potential), the complementarity of multiple energy-carrying forms of photoacoustic-thermal coupling, and the synchronous exploitation of a material cycle and thermal cycle.