Enhancing electrocatalytic hydrogen evolution via engineering unsaturated electronic structures in MoS₂

Abstract

The search for efficient, earth-abundant electrocatalysts for the hydrogen evolution reaction (HER) has identified unsaturated molybdenum disulfide (MoS₂) as a leading candidate. This review synthesises recent advancements in the engineering of MoS₂ to enhance its electrocatalytic properties. It focuses on strategies for designing an unsaturated electronic structure on metal catalytic centers and their role in boosting the efficiency of the hydrogen evolution reaction (HER). It also considers how to optimize the electronic structures of unsaturated MoS₂ for enhanced catalytic performance. This review commences with an examination of the fundamental crystal structure of MoS₂; it elucidates the classical unsaturated electron configurations and the intrinsic factors that contribute to such electronic structures. Furthermore, it introduces popular strategies for constructing unsaturated electronic structures at the atomic level, such as nanostructure engineering, surface chemical modification and interlayer coupling engineering. It also discusses the challenges and future research directions in the study of MoS₂ electronic structures, with the aim of broadening their application in sustainable hydrogen production.

---

1 Introduction

As global demand for clean energy continues to rise, the development of efficient hydrogen production has emerged as a critical focus. Water electrolysis is a particularly promising approach due to its efficiency and environmentally friendly nature.^1–3 However, its implementation depends heavily on the development of high-performance and durable catalysts for the hydrogen evolution reaction (HER).^4,5 Traditionally, precious metals like platinum (Pt) have been the catalysts of choice due to their exceptional HER activity, which is attributed to their moderate binding interaction with key adsorbates, such as intermediate hydrogen species (H*).^6,7 However, the scarcity and high cost of Pt and other noble metals have driven increasing interest in the search for earth-abundant alternatives.^8–10 This shift is crucial to making hydrogen production more economically viable and scalable for widespread clean energy applications.

Using density functional theory (DFT) calculations, Hinnemann and colleagues identified earth-abundant materials with catalytic properties similar to platinum (i.e., near-zero free energy for H* binding) for HER.¹¹ Among these candidates, molybdenum disulfide (MoS₂) was predicted as a promising HER catalyst, despite practical bulk MoS₂ being regarded as a poor electrocatalyst.^12,13 This apparent contradiction was later clarified by Jaramillo and Ib Chorkendorff, who demonstrated that hydrogen evolution predominantly occurs at the edge sites of layered MoS₂, where unsaturated electronic structures are present.¹⁴ In contrast, the basal plane of MoS₂, with its fully saturated electronic configuration, contributes negligibly to HER due to a lack of active sites.^15,16 These findings underscore the significance of engineering unsaturated electronic structure of MoS₂, supporting Norskov's perspective that the electrocatalytic performance of a material is fundamentally determined by its electronic structure.

Beyond exploring the atomic-level structure–activity relationship, numerous studies have focused on maximizing the intrinsic catalytical performance of MoS₂ while developing additional electric properties to contribute HER progress. For example, Jin and Cui pioneered the use of Li⁺ intercalation to successfully tune the electronic states of MoS₂, transforming semiconducting 2H-MoS₂ into metallic 1T-MoS₂. Their research demonstrated that Li⁺ intercalation introduces single electrons into the Mo d-orbitals, achieving a new stable state through d-orbital re-splitting. This transformation significantly enhances hydrogen species activation and improves charge transfer kinetics.^12,17 In a complementary study, Tsai and colleagues introduced sulfur defects in the basal plane of MoS₂, generating high-energy dangling bonds that increased the mass activity of the catalyst. Their findings suggest that thermodynamically unstable, coordinatively unsaturated sites are more prone to adsorbing H* intermediates, playing a key role in the HER process.¹⁸ Beyond structural modifications to the 2H phase of MoS₂, researchers have explored the introduction of various metal and non-metal dopants to fine-tune the local coordination environment, electron density, and energy level structure of the catalytic centers. For instance, DFT calculations by Cui predicted how different transition metal dopants (e.g., Co, Fe, Ni, and Cu) impact the HER performance of MoS₂.¹⁹ These strategies demonstrate that engineering unsaturated electron configurations in MoS₂ can significantly enhance its electrochemical properties, paving the way for more efficient HER catalysts.

While it has vaguely perceived that unsaturated sites in MoS₂ are somehow linked to its electrocatalytic performance, the precise nature of this relationship remains unclear. To address this, researchers have conducted extensive studies to establish design principles for MoS₂ catalysts with high HER activity, particularly focusing on the interaction between intermediate hydrogen species and catalytic centers.^20–23 Previous reviews have categorized modification methods for MoS₂ catalysts, aiming for advanced technologies that offer high catalytic activity, stability, preparation efficiency, and etc. In contrast, our review provides a systematic summary of the unsaturated electronic structures in modified MoS₂ catalysts from the perspective of band structure and electronic orbital arrangement, which are decisive factors in determining electrocatalytic performance. Additionally, we provide practical guidance for how to construct active catalytic centers with idea unsaturated electronic structures: (1) through nanostructure engineering, single-layer or multi-exposed-edge MoS₂ materials are synthesized without altering their chemical composition. Additionally, stress regulation is employed to introduce or relieve lattice strain, facilitating the formation of active defects in the distorted structure. (2) Surface chemical modification strategies-such as creating vacancy defects, heteroatom substitution, heterogeneous doping, and incorporating heterogeneous crystal domains-are used to modify the local chemical environment of the catalytic centers. These methods regulate the chemical properties of MoS₂ at the micro-regional level, thereby enhancing its HER performance. (3) Semiconductor/metal contact effects have emerged as a promising approach for constructing robust and highly efficient catalytic electrodes. By modulating the van der Waals gap between MoS₂ catalytic layers and metal supports, the orbital overlap and hybridization at the interface can regulate the in-plane electronic structure of MoS₂, thereby promoting the exposure of active centers and enhancing HER efficiency. Finally, we discuss the challenges and future research directions in optimizing the electronic structures of MoS₂ catalysts, with the aim of promoting their broader application in sustainable hydrogen production.

2 Fundamentals of MoS₂ electrocatalyst for HER

2.1 HER mechanism

HER is a two-electron/proton coupling process that occurs on the cathode surface during water electrolysis, with the reaction mechanism dependent on the pH of the solution.^24,25 As outlined in Table 1, the HER involves two main steps: adsorption (Volmer) and desorption (Heyrovsky/Tafel). First, the Volmer step initiates the reaction, where hydrogen ions (in acidic environments) or water molecules (in alkaline environments) are electrochemically adsorbed onto the catalyst surface, forming an intermediate hydrogen species (H_ads). The Heyrovsky and Tafel steps describe two desorption pathways for H_ads. The Heyrovsky step, predominant at low H_ads surface coverage, involves H_ads reacting with another hydrogen ion or water molecule to form H₂. The Tafel step, occurring at high H_ads surface coverage, involves the combination of two H_ads to produce H₂.²⁶ It is clear that an effective HER catalyst is essential for optimizing these adsorption and activation pathways.

Table 1 HER pathways in acid and alkaline solution

	Acid solution	Alkaline solution
Volmer step	M + H^+ + e^− → Hads	M + H2O + e^− → Hads + 2OH^−
Heyrovsky step	M–Hads + H^+ + e^− → H2	M–Hads + H2O + e^− → H2 + OH^−
Tafel step	2M–Hads → H2 + 2M	2M–Hads → H2 + 2M
Overall reaction	2H^+ + 2e^− → H2	2H2O + 2e^− → H2 + 2OH^−

---

2.2 Basics of MoS₂ crystal

2.2.1 Phase and stacking of MoS₂ crystal. A monolayer of MoS₂ has a graphene-like 2D structure, where a layer of molybdenum (Mo) atoms is sandwiched between two layers of sulfur (S) atoms, forming an S–Mo–S stacking arrangement.^27,28 The layers of MoS₂ are held together by weak van der Waals (vdW) interactions, with a repeat distance of 6.5 Å between them, while the in-plane structure is stabilized by strong Mo–S covalent bonds (Fig. 1a).²⁹

Fig. 1 (a) Schematic diagram of the MoS₂ crystal structure, (b) schematic diagram of the structures of 2H phase MoS₂ (hexagonal symmetry, two layers per repeating unit, trigonal prism coordination), 3R phase MoS₂ (rhombohedral symmetry, three layers per repeating unit, trigonal prism coordination) and 1T phase MoS₂ (tetragonal symmetry, one layer per repeating unit, octahedral coordination), the stacking index c represents the number of layers in each stacking sequence, and the interlayer spacing is about 6.5 Å.

As depicted in Fig. 1b, depending on the specific coordination of Mo atoms, MoS₂ cell can exist in either a trigonal prismatic or octahedral configuration.³⁰ Based on the phase structure and stacking sequence, MoS₂ has three standard crystal structures: 2H phase (H stands for hexagonal) with two S–Mo–S molecular units stacking in an AB manner along the vertical (c-axis) direction; 3R (R stands for rhombohedral) with three S–Mo–S units stacked in an ABC manner along the c-axis; 1T phase (T stands for trigonal) with a monolayer S–Mo–S molecular per unit cell.^31,32 The energy level splitting, electron filling, and spin states of the Mo d-orbitals vary with the coordination structure, leading to differences in electrical properties, magnetic properties, and thermodynamic stability.^33,34 Among these, 2H-MoS₂ is the most stable, while the 1T and 3R phases exhibit metastable behavior.³⁵

2.2.2 Band structure &electronic properties of MoS₂. As a typical transition metal dichalcogenides (TMDs) material, the electronic properties of MoS₂ are determined by the interactions between the p, s, and d orbitals of each atom near the Fermi level. Typically, the contribution of the S s orbitals is neglected because of their minimal perturbation on the band structure. In terms of energy, the orbital energy of S atoms is significantly lower than that of the Mo d orbitals. Consequently, the electronic structure and properties of MoS₂ are primarily influenced by the coordination environment and the d orbital energy levels of the Mo atoms.³⁴ The unit cell of MoS₂ contains six p orbitals (two S atoms, each contributing three sub-bands), which can accommodate up to 12 electrons. The S atoms contribute eight electrons, while the remaining four electrons are provided by the Mo atoms. Consequently, only two electrons are available for allocation to the Mo d orbitals (as the Mo atom, with six valence electrons, gives four to the S atoms).

Due to the varying orientations and shapes of the d orbitals in molybdenum ions, the d orbitals of the central metal atom experience electrostatic interactions with ligands that differ in both direction and energy depending on the crystal field environment. To clarify, 1T-phase MoS₂ adopts an octahedral molecular configuration with an O_h point group, where six S ligands are positioned at the vertices of a symmetric octahedron surrounding the Mo atom. The electron clouds of the d_z² and d_x²−y² orbitals are located on the connecting line of Mo–S and directly face the negatively charged ligands, resulting in stronger repulsive interactions and higher energy levels, e_g band.^36,37 On the contrary, the electron clouds of the d_xy, d_yz, and d_xz orbitals with spatial deviation from the Mo–S axis and avoid the ligands, leading to weaker repulsive interactions and lower energy t_2g band (Fig. 2b). These degenerate orbitals remain non-bonding, allowing for quasi-continuous occupation by fermions, thereby giving MoS₂ metallic properties. However, for nonlinear molecules, if their ground electronic state is orbitally degenerate, the molecule becomes unstable. In such cases, the Jahn–Teller distortion effect occurs to relieve the degeneracy.

Fig. 2 The unit cell structure, band structure and Fermi level position of (a) 2H phase MoS₂ and (b) 1T phase MoS₂. (c) HER catalytic behavior on 2H phase MoS₂, 1T phase MoS₂, and the principle to modify MoS₂.

The molecular configuration of 2H-MoS₂ is such a distorted octahedron. By rotating one triangular face by 60° relative to the opposite face, a trigonal prismatic molecular configuration, point group D_3h, is obtained. In this trigonal prismatic configuration, due to the lack of symmetrical alignment between the S ligands and the Mo 4d orbitals, the d orbitals tend to overlap “sideways” resulting in repulsion from all ligands. This causes the molecular orbitals to split into bonding and antibonding bands, creating an empty band gap. In this case, the d orbitals split into three bands: e_g2, e_g1, and a₁, corresponding to the d_xz, d_yz; d_xy, d_x²−y²; and d_z², respectively (Fig. 2a).

As discussed above, the 2H and 1T crystalline phases of MoS₂, the two most widely applied phases, exhibit fundamentally distinct solid-state physical and electronic properties, resulting in notable differences in HER catalytic performance. Regarding the HER process (Fig. 2c), the semiconducting 2H phase is characterised by a significant band gap and limited electron mobility. Additionally, its strong bonding within the basal plane provides excellent thermodynamic stability but renders it catalytically inert. Consequently, the HER activity of 2H-MoS₂ primarily originates from its exposed edge sites, where unsaturated electronic states facilitate the adsorption and conversion of intermediates. This highlights edge engineering as a critical strategy for enhancing the catalytic performance of the 2H phase. In contrast, the metallic 1T phase offers excellent electronic conductivity, enabling rapid electron transfer during the HER process. Furthermore, the high density of states near the Fermi level strengthens interactions between 1T-MoS₂ and water molecules, promoting efficient adsorption and activation of intermediates. However, the 1T phase suffers from structural instability, often requiring stabilization strategies such as intercalation or heteroatom doping to maintain its catalytic performance under operational conditions.

2.3 Relationship between electron configuration of MoS₂ and its electrocatalytic activity

Generally, the basal plane of MoS₂ exhibits limited catalytic activity due to its saturated electronic configuration.^38,39 However, modifying the coordination environment of the Mo centers and tuning the energy levels of their d orbitals offer significant opportunities for creating high-activity, unsaturated MoS₂ configurations, thereby enhancing its catalytic potential for HER.^40,41 Numerous efforts have been made in this direction, as presented in as presented in Fig. 3 and Table 2. To clarify, in the ground-state 2H-MoS₂ with a trigonal prismatic structure, the crystal field splitting energy is greater than the electron pairing energy. Following the Aufbau principle, electrons first occupy the low-energy d_z² orbital with two electrons of opposite spins, resulting in diamagnetism. Interestingly, when the host Mo atom is substituted by noble metals such as Pt and Rh, electrons are lost from the d_z² orbital due to the higher electronegativity of these metals.^42,43 Consequently, this results in a low-spin unsaturated state that exhibits metallic-like properties.

Fig. 3 Orbital electron unsaturation strategy for MoS₂ electrocatalysts based on crystal field theory.

Table 2 Electronic band structure and characteristics of unsaturated MoS₂

Catalysts	Phase	Modification approach	Description of band structure and electrons state	Conductivity	Ref.
Pristine MoS2	2H	—	Three bands: a1 (dz^2), eg1 (dxy, dx^2−y^2), and eg2 (dyz, dxz)	Semi-conductor	34
			Electron state: paired electrons fully occupy the lowest a1 band
Pt-MoS2	2H	Mo loses electrons through substitution with highly electronegative noble metals	Three bands: a1 (dz^2), eg1 (dxy, dx^2−y^2), and eg2 (dyz, dxz)	Metallic-like	42
			Electron state: single-spin electron half-fill the lowest a1 band
FD-MoS2	2H	Mo losing electrons by making Frenkel disorder	Three bands: a1 (dz^2), eg1 (dxy, dx^2−y^2), and eg2 (dyz, dxz)	Metallic-like	42
			Electron state: single-spin electron half-fill the lowest a1 band
Vs-MoS2	2H	Adding electrons by making S vacancy	Three bands: a1 (dz^2), eg1 (dxy, dx^2−y^2), and eg2 (dyz, dxz)	Metallic-like	43
			Electron state: paired electrons fully occupy the lowest a1 band, and single-spin electron incompletely occupy the higher eg1 band
1T-MoS2	1T	—	Two bands: t2g (dxy, dyz, dxz) and eg (dx^2−y^2, dz^2)	Metallic	34
			Electron state: two parallel-spin electrons incompletely occupy the degenerate t2g band
CoO@1T-MoS2	1T	Adding electrons by co-substituting Mo and S with Co and O atoms	Two bands: t2g (dxy, dyz, dxz) and eg (dx^2−y^2, dz^2)	Metallic	44
			Electron state: three parallel-spin electrons half-fill the degenerate t2g band
1T′-MoS2	Stretch 1T phase	Re-split band structure by intercalating alkine ions	Two major bands: t2g (dxy, dyz, dxz) and eg (dx^2−y^2, dz^2)	Metallic-like	45
			Electron state: two parallel-spin electrons half-fill the lower t2g band
1T′′-MoS2	Stretch 1T phase	Re-split band structure by chemical exfoliation	Two major bands: t2g (dxy, dyz, dxz) and eg (dx^2−y^2, dz^2)	Metallic-like	46
			Electron state: two parallel-spin electrons half-fill the lower t2g band
1T′′′-MoS2	Compress 1T phase	Re-split band structure by de-intercalating of alkine ions	Two major bands: t2g (dxy, dyz, dxz) and eg (dx^2−y^2, dz^2)	Semi-conductor	45
			Electron state: paired electrons fully occupy the lower t2g band

---

Conversely, donating additional electrons to the e_g band is more challenging due to the presence of a band gap. Creating vacancies can be a viable approach, as it destabilizes the lattice. This destabilization may trigger a transition from the 2H to the 1T phase upon electron absorption. However, the behavior of electrons differs significantly in the octahedral case. First, the crystal field splitting energy is lower than the electron pairing energy, allowing electrons to preferentially fill all orbitals of similar energy while avoiding pairing. As a result, the incomplete occupation of the t_2g orbitals in 1T-MoS₂ leads to an intermediate spin state, which demonstrates moderate activity but low stability.

Doping 1T-MoS₂ with electron-donating atoms (e.g., Co, Mn, Fe) can achieve a half-filled t_2g band, promoting a higher spin state and enhanced stability.^44,47 Furthermore, significant distortion occurs when 1T-MoS₂ undergoes external strain, such as compression or stretching. This strain can induce the spontaneous conversion of the 1T phase into 1T′ and 1T′′ phases, thereby lowering its energy through Jahn–Teller deformation effects.^40,46 Typically, these distorted phases feature zigzag Mo–Mo chains and the formation of Mo–Mo bonds, such as dimerized Mo (1T′′) or trimerized Mo (1T′′′) atoms. These structures may act as key catalytic sites during the HER process, although they are sometimes overlooked in catalytic evaluation.

3 Nanostructure engineering

The differences in atomic arrangement and geometry contribute to the diverse physical and chemical properties of MoS₂ catalysts. This includes variations in polymorphic structures (e.g., 2H and 1T phases), as well as band structure and crystal plane exposure, which are influenced by the number of layers and the structure of MoS₂ catalysts.^48,49 In recent years, geometry and strain engineering of nanomaterials has attracted great attention by adjusting the properties and electrocatalytic performance of MoS₂ by regulating the crystal atomic arrangement and deformation.^50,51 This section outlines recent advances in the development of monolayer materials, edge dangling bond engineering, and stress regulation strategies for the design of advanced MoS₂ catalysts.

3.1 Atomically thin structure

Bulk MoS₂, held together by van der Waals (vdW) interactions, exhibits the properties of an indirect bandgap semiconductor.^52,53 Differently, the monolayer MoS₂ obtained by Novoselov et al. through mechanical exfoliation in 2005 is considered to have a direct band gap with metallic properties.⁵⁴ Therefore, it was found that the electrical properties of MoS₂ are highly related to the number of layers, and few-layer or monolayer MoS₂ shows great potential in electrocatalysis due to its unique electrical properties.^55,56 As depicted in Fig. 4a–g, to achieve controllable synthesis of monolayer MoS₂, Zhang's group has focused on exploring the fabrication of atomically thin monolayer MoS₂ films. These efforts include the epitaxial growth of monolayer WS₂ on sapphire and the growth of monolayer MoS₂ of various sizes on Au foil via low-pressure chemical vapor deposition (CVD), both of which can serve as outstanding electrocatalysts for HER.^57–59 Based on this, Zhang and coauthors successfully prepared a four-layer 2H-TaS₂ film on an Au foil, exposing edge active sites by controlling the number of layers. This approach achieved Pt-like hydrogen evolution efficiency, with a Tafel slope of 33–42 mV dec⁻¹ and an exchange current density of ∼179.47 μA cm⁻² (Fig. 4h and i).⁶⁰ Additionally, Ye et al. achieved high-yield monolayer MoS₂ films by hydrolyzing a lithium salt-containing MoS₂ precursor, resulting in a monolayer MoS₂ catalyst with a catalytic activity of 1563.6 μmol h⁻¹.⁶¹

Fig. 4 (a) Schematic representation of the dependence of coverage or flake size on different precursor-substrate distances. (b) Photographs of monolayer MoS₂ films on SiO₂/Si, a bare quartz substrate, and MoS₂ on quartz. (c) A triangular MoS₂ flake transferred onto SiO₂/Si substrate. (d) Atomic force microscopy (AFM) image captured from the dashed box in (c) and corresponding section-view along the dashed arrow showing the monolayer feature of the flake (∼0.9 nm in apparent height). (e) Optical images of submonolayer MoS₂ flakes after transference onto SiO₂/Si and quartz substrates. (f) TEM image of a MoS₂ triangle transferred on a carbon film supported on copper grids. (g) TEM image on a folded edge indicating the monolayer nature of the MoS₂ flake. (h) Schematic illustration of the HER process of 2H-TaS₂/Au foils. (i) Polarization curves (iR-corrected) of as-grown 2H-TaS₂ with different thicknesses, Au foil, and commercial Pt. Band structure for MoS₂ of the bulk (j) and the monolayer (k). The band structure has been projected onto constituting atomic species with the blue color representing S and red representing Mo. The band-edge states relevant to optical transition (K_V, K_C, Γ_V, and T_C) are labeled and the band gap transitions are indicated.

To elucidate the mechanism by which the electronic properties of MoS₂ are regulated by the number of layers, Zunger et al. employed first-principles calculations and revealed that in bulk MoS₂, the valence band maximum and conduction band minimum are located at the T_v and T_c points in momentum space, respectively.⁶² However, in monolayer MoS₂, these points shift to the K point in the Brillouin zone, inducing a bandgap transition, as illustrated in Fig. 4j and k. Secondly, the monolayer structure of MoS₂ stabilizes the edge regions by eliminating constraints from adjacent layers, which facilitates the exposure of edge sites and the introduction of structural defects, thereby accelerating the HER process. Although monolayer or few layer MoS₂ catalysts exhibit excellent electrical properties and a high ratio of exposed active edges, the exfoliation and preparation of such MoS₂ catalysts remain long-standing challenges, making it unsuitable for the large-scale production of cost-effective HER catalysts.

3.2 Morphology control

The highly reactive edges of MoS₂ are typically difficult to stabilize, leading bulk MoS₂ to terminate with smooth, inert surfaces.⁶³ To overcome this limitation, researchers have exploited the anisotropy and morphological diversity of MoS₂ nanosheets to expose and stabilize active edge sites.⁶⁴ For instance, as depicted in Fig. 5a, Cui et al. grew vertically MoS₂ nanoarray on a Mo substrate via a CVD method, effectively increasing the reactive area and promoting the contact between the active edges and the electrolyte.⁶⁵ This approach enables MoS₂ to achieve an intrinsic turnover frequency (TOF) of 0.013 s⁻¹ without additional energy input, which is close to the experimental fit value (0.016 s⁻¹) for pure edge sites reported by Chorkendorff et al.¹⁴ In addition, Yang and co-workers also prepared a similar Mo-based MoS₂ electrocatalyst, reaching remarkable stability over more than 20 000 cycles in 0.5 M H₂SO₄.^66,67

Fig. 5 (a) (left) TEM images of MoS₂ and MoSe₂ films with edge-terminated structures. (right) Idealized structure of edge-terminated molybdenum chalcogenide films with the layers aligned perpendicular to the substrate. (b) TEM image of the MoS₂ QDs, the inset displays the HRTEM image of the same. (c) The size distribution of the MoS₂ QDs in (b). The red solid line is the simulated Gaussian curve. (d) Schematic illustration of the fabrication of mesoporous MoS₂ foam (mPF-MoS₂). (e and f) SEM and TEM images of mPF-MoS₂. (g) HER polarization curves for mPF-Co-MoS₂ with different Co doping contents in comparison with mPF-MoS₂ and 40% Pt/C. (h) Durability measurement of mPF-Co-MoS₂.

In addition to utilizing metal foils, MoS₂ nanosheets can also be fabricated into mesoporous foams (mPF-MoS₂-SiO₂) through SiO₂ template etching. The resulting catalyst features an average pore size of ∼30 nm, which effectively enhances the transport of H₃O⁺ and H⁺ ions. Moreover, the directional growth of MoS₂ nanosheets around the mesopores contributes to edge stabilization, resulting in a catalyst that demonstrates cycle durability exceeding 5000 cycles (Fig. 5d–h). Preparing MoS₂ as highly dispersed nanoparticles is a feasible approach to obtaining a large number of accessible edge sites.⁶⁸ Jiang and coauthors successfully synthesized highly dispersed MoS₂ nanoparticles on reduced graphene oxide-modified carbon nanotube/polyimide films (PI/CNT-RGO-MoS₂) through an electrochemical method. The study indicates that PI/CNT-RGO facilitates the intimate growth of MoS₂ particles on its surface, while the coupling between RGO and MoS₂ enhances electron transfer between the edge sites and the substrate, enabling the catalyst to exhibit a low overpotential of 90 mV at a current density of 10 mA cm⁻² and a Tafel slope of 61 mV dec⁻¹.⁶⁹

Reducing the dimension of 2D MoS₂ to 0D quantum dots (MoS₂ QDs) is another advanced strategy to achieve an ultra-high specific surface area and a unique edge-defect-rich structure (Fig. 5b and c). Moreover, monolayer MoS₂ QDs can eliminate the interlayer barrier in the vertical direction, allowing more free electrons to transfer along the MoS₂ surfaces. Under acidic conditions, the MoS₂ QDs catalyst developed by Ren exhibits an overpotential of 39 mV at a current density of 10 mA cm⁻², which is 14 times lower than that of bulk MoS₂.⁷⁰ More interestingly, Jaramillo et al. reported a unique double-helical MoS₂ structure (DG) MoS₂ film, which exposed a greater fraction of catalytically active edge sites through high surface curvature and promoted the diffusion process of H₂ bubbles.⁶³ Various morphology designs not only increase the active specific surface area of MoS₂ catalysts but also expand their application potential. On one hand, morphology control allows for the manipulation of the micro-interface environment of electrodes, improving mass transfer behavior of the reactants. Additionally, in Amper-scale HER applications, nanoarray morphologies are expected to facilitate efficient H₂ bubble desorption and diffusion, addressing kinetic challenges posed by massive hydrogen accumulation.

3.3 Strain engineering

Beyond controlling the number of layers and stacking patterns, applying local lattice strain or pressure engineering is a powerful approach for fine-tuning the band structure and electrical properties of TMDs. As depicted in Fig. 6a and b, Nayak et al. investigated the electronic structure and lattice vibrational dynamics of pristine 2H-MoS₂ and distorted monolayer 1T-MoS₂ using diamond anvil cell experiments and DFT calculations.^71,72 Their study revealed that inevitable lattice distortion occurs, leading to an electronic transition from semiconducting to metallic at a pressure of approximately 19 GPa (Fig. 6c and d). The eventual metallization occurs due to the overlap of the valence and conduction bands, driven by S–S interactions as the interlayer gap decreases. Later, Zhao et al. demonstrated the z-axis in the MoS₂ crystal cell is much more compressible than a-axis during the compression process, due to the weak vdW interaction between layers.⁷³ Therefore, the eventual metallization arises from the overlap of the valence and conduction bands owing to the reduction of interlayer gap, which makes the MoS₂ improved conductivity and faster HER kinetic process.⁷⁴

Fig. 6 (a) A 3D illustration of multilayered MoS₂ in a DAC pressure medium for compression experiments. (b) Theoretical charge density isosurfaces of multilayered MoS₂, charge accumulation (green) and depletion (orange) at the in-plane. (c) Theoretical calculation of the pressure-dependent band gap of multilayered MoS₂. Inset: the unit cell of MoS₂. (d) Pressure-dependent electrical resistivity of MoS₂. Three characteristic regions have been identified: semiconducting (SC), intermediate state (IS) and metallic regions. (Inset) Theoretically calculated pressure-dependent electrical resistivity. (e) Schematic models of single-layered MoS₂ with 2H and 1T phases in basal plane and cross-section views. Mo, blue; top S, orange; bottom S′, purple. (f) Transformation of 2H to 1T phase caused by S-plane sliding; the sliding of the Mo plane leads to the transformation of the 2H to the 2H’ phase. (g and h) Atomic movements during 2H to 1T phase transformation in single-layered MoS₂ at T = 600 °C.

In addition to applying crystal pressure, researchers have found that electron beam irradiation can also relieve lattice stress in MoS₂ by disrupting lattice symmetry. This process causes atoms to migrate to lower energy sites, leading to structural reconstruction. As depicted in Fig. 6e and f, Lin and co-workers employed electron beam irradiation to disrupt the local stress equilibrium within the 2H-MoS₂ lattice.⁷⁵ This process induced the migration of Mo and S atoms, leading to the formation of a distorted 1T phase at the reconstructed domains. The resulting phase exhibited a triangular morphology with well-defined 2H–1T phase boundaries (Fig. 6g and h). Moreover, Katagiri and colleagues successfully demonstrated phase transformation patterning of MoS₂ at room temperature using a 160 MeV nm⁻² electron beam. Due to the interaction between the excited electron beam and the lattice, the 1T/2H in-plane Schottky junctions produced by this technique exhibited a lower barrier (0.13–0.18 eV) than the theoretical value (0.47 eV). Furthermore, the MoS₂ crystal domains minimized electron scattering and barriers, thereby enhancing electron mobility while improving the kinetics of electrocatalytic reactions.⁷⁶ Fine-tuning the micro-regional configuration and electronic properties of MoS₂ through intracrystalline stress can promote the formation of unsaturated active sites and induce specific phase structures. However, due to the complexity of the required technology and high equipment costs, this approach is often coupled with theoretical calculations to gain deeper insights into the structural evolution and catalytic activity mechanisms.

4 Surface chemical modification

The redistributed electronic landscape of MoS₂ alters the interactions between the electrocatalytic centers and intermediate hydrogen species when the host catalyst comes into contact with additives (such as atoms, ions, or molecules) or when vacancies are introduced.^77,78 Surface chemical modification is an effective approach for the regulation of the crystal field and ligand field of MoS₂, which can greatly affect the physical–chemical properties and significantly improve its catalytic activity. To date, the ligand structure of MoS₂ can be modulated through vacancy defect, atomic doping, phase transition engineering and heterostructure engineering to achieve significant changes in their electrical nature, magnetic properties, and electrocatalysis performance.^15,40,44,79 This section provides a detailed discussion of these approaches and highlights several representative examples.

4.1 Vacancy defect

Introducing vacancy defects is a straightforward and effective strategy for creating additional active sites and fine-tuning the coordination environment of catalytic centers, making it a valuable method for catalyst modification. Typically, these defects are generated by disrupting the crystal periodic structure without the need for external additives.⁷⁹ As depicted in Fig. 7a–c, Wang et al. precisely controlled the concentration of S-vacancy on the MoS₂ surface by H₂O₂ etching, revealing the fact that the residual carriers left in S-vacancy prefer transferring into adjacent Mo atoms, which leading the Mo sites stronger H* attraction as a result of de delocalized electrons.⁴⁶ Frenkel-defect (FD) is one of the typical vacancy defects among many related works. For example, Liu Song's group used a simple argon gas annealing treatment to exploit this defect in MoS₂ and boost hydrogen evolution. It is revealed by DFT calculations that the FD can raise a unique charge landscape and appropriate sites for H* species. By comparison to the single atomic Pt doped MoS₂ (Pt-MoS₂), the overpotential of FD-MoS₂ decreased from 211 mV to 164 mV at a current density of 10 mA cm⁻².⁴³ Subsequently, they prepared antisite defect structures with a concentration of 4% (where two sulfur atoms or a single sulfur atom occupy the position of a molybdenum atom, denoted as S2_Mo-MoS₂ or S_Mo-MoS₂) within the monolayer MoS₂ basal plane through calcination in an H₂/Ar atmosphere, and investigated the role of these antisite defects in promoting the HER process (Fig. 7d). According to DFT analysis, the formation of antisite defects disrupts the original coordination of catalytically S atoms, resulting in a distinctly different and asymmetric charge density distribution in MoS₂ (Fig. 7e). This change induces a transition from a semiconductor to a metallic state by introducing electronic states at the Fermi level, thereby enhancing H* adsorption.⁸⁰

Fig. 7 (a) Schematic of the chemical etching process to introduce single S-vacancies. (b) STEM images of CVD-grown single-layer MoS₂ thin films after etching. The yellow dashed circles represent S vacancies. (c) Free energy versus the reaction coordinates of MoS₂ with diverse S-vacancy states. Inset: corresponding MoS₂ models in which Mo atoms and the upper S-vacancies are symbolized by the numbers 1–16 and the letters A–P, respectively. (d) Schematic diagram of the formation process of the antisite defect. (e) Total and partial density of states (DOS) of pristine MoS₂, V_Mo-MoS₂, and S2_Mo-MoS₂. The Fermi level is set to zero and denoted by the black dotted line.

It is worth noting that specific types of vacancies can be generated by doping with designated heteroatoms, for example, metal dopants are more likely to expose S vacancies. Therefore, fully leveraging the synergistic effects of metal heteroatoms and vacancy defects can lead to remarkable improvements in HER performance. More examples are reflected in alkali metals that have strong interactions with S atoms, such as Zn, Fe, Co, Pd, and Cu.^78,81–83 The concentration of missing S atoms can be easily controlled by controlling the amount of alkali metals.^82,84 In addition, as low-loaded noble metals are expected to boost HER performance, the synergistic effect of noble metals and vacancy defects is highlighted in reducing the H* adsorption/desorption energy barrier and improving HER performance.

4.2 In-plane substitution

In-plane substitutional doping of MoS₂ can be achieved through two strategies: replacing host Mo and S atoms, which is the main way to achieve in-plane charge rearrangement.⁸⁵ Non-metal elements such as C, N, O, and P. are used as the main in-plane substituents due to their easy operation and low cost.^86–89 Ge and co-authors replaced the S atoms in MoS₂ with O atoms through a simple wet chemical method. The high electronegativity of O induced electron outflow from the adjacent S atoms, thereby constructing an efficient HER 1T-MoS₂ catalyst. Additionally, O atoms tend to form hydrogen bonds with H₂O in the electrolyte, enhancing the adsorption and dissociation of H₂O on the catalyst surface. While the high electronegativity of non-metal elements is an effective strategy for driving charge transfer from host S atoms, an excessive electronegativity gap can lead to localized charge accumulation and slow H* desorption process.⁸⁸ Jiang et al. addressed this issue by Co-doping MoS₂ with both Se and O atoms, whose electronegativities are 2.55 and 3.5, respectively. Studies have shown that Se–O co-doped MoS₂ exhibits moderate energy level splitting due to the synergistic effect of Se–O on the electronic structure. Furthermore, the smaller atomic diameter of Se can induce lattice distortion by replacing partial S atoms, resulting in Se-MoS₂ electrocatalyst exhibiting a low Tafel slope of ∼47 mV dec⁻¹ under acidic conditions.⁹⁰

To avoid the violent host–guest electron interaction, it is feasible to further awaken the inertness of metal sites. The synergistic effect of metal-nonmetal diatomic is developed to stimulate the utilization of MoS₂. Transition metals (TMs), with their abundant d-orbital electron states and spin magnetism, are well-suited for adjusting the Fermi level of MoS₂ by forming s–p hybridized orbitals with S atoms.⁴⁰ As depicted in Fig. 8a, Deng and coworkers synthesized a Co and Se co-doped MoS₂ nanostructured foam electrocatalyst (Co/Se-MoS₂-NF) by confining Co atoms within the sublayer and Se atoms within the surface lattice of MoS₂ using a one-step hydrothermal method. In this configuration, Co atoms serve as substitutional dopants for sublayer Mo, activating the inert S sites, while Se atoms are confined in the surface to stabilize the basal plane. The synergistic effect of Co and Se is evident in the optimized thermodynamics and structural stability of the HER sites (Fig. 8b and c). The near-zero free energy significantly accelerates the formation of H₂ formation, while the balanced electronic structure ensures the long-term stability of the active sites. Tests revealed that Co/Se-MoS₂-NF exhibited a low Tafel slope of 67 mV dec⁻¹ and a long-term performance over 360 h at a high current density of 1000 mA cm⁻² in 0.5 M H₂SO₄.⁹¹

Fig. 8 (a) Atomic resolution HAADF-STEM image of Co/Se-MoS₂-NF. Inset: magnification of the white dashed area in (a). (b) H* adsorption structures at the (b) Mo edge, (c) plane, and (d) S edge positions of Co/Se-MoS₂. (c) H* adsorption free energies (ΔG_H*) at the plane, Mo edge, and S edge positions of pure MoS₂, Co-doped, Se-doped, and Co/Se co-doped MoS₂, denoted as MoS₂, Co-MoS₂, Se-MoS₂, and Co/Se-MoS₂, respectively. The hydrogen coverage at the Mo edge is defined based on the terminated S or Se monomer positions, while the hydrogen coverage at the S edge is defined based on the terminated S or Se dimer positions, which are the active sites for HER. (d) Schematic diagram of Pd and Ru dual-doped MoS_2−xOH_y phase. The blue, yellow, purple, green, and red spheres represent Pd, S, Mo, Ru, and O atoms, respectively. (e) (Left) HER mechanism of Pd, Ru-MoS_2−xOH_y in acidic media; (Right) HER mechanism of Pd, Ru-MoS_2−xOH_y in alkaline media.

More recently, bimetallic atoms pair has also been widely used for the tandem regulation of the HER process. Li et al. developed a Ni, Co bimetallic atoms dual modified MoS₂ catalyst (Ni, Co-MoS₂) to enhance HER activity in alkaline solutions. The study showed that Ni atoms serve as initial activation sites for the H₂O, while Co atoms facilitate the adsorption and desorption kinetics of H intermediates by influencing the electronic structure of the neighboring S. The Ni, Co-MoS₂ catalyst exhibited a lower Tafel slope of 78 mV dec⁻¹.⁹² Similarly, Han and coauthors developed a Cu and Pd dual-doped MoS₂ catalyst (Cu–Pd-MoS₂) using a two-step approach. The study showed that Cu enhanced the conductivity, while Pd facilitated the phase transition from 2H to 1T phase, thereby improving both the electron transfer and the number of active sites.⁹³ Additionally, Luo et al. engineered a di-anionic surface (Pd, Ru-MoS_2−xOH_y) on MoS₂ to modulate the HER mechanism (Fig. 8d). The study demonstrates that by replacing S sites with –OH anions, a strong non-covalent hydrogen interaction was established, effectively attracting hydronium ions and water molecules. Furthermore, the hydroxyl network works synergistically with adjacent metal atoms (M–OH), overcoming the kinetic barriers associated with water dissociation in alkaline media (Fig. 8e). As a result, the Pd, Ru-MoS_2−xOH_y catalyst exhibits superior kinetic performance in both acidic and alkaline conditions, with Tafel slopes of 85 mV dec⁻¹ and 121 mV dec⁻¹, respectively.⁹⁴

4.3 Embedding doping

Driven by the high chemical potential of the reactants, the interstitial spaces within the MoS₂ lattice become occupied by free atoms.^95–97 These structural additives can subsequently modify the catalytic behavior by tuning the Fermi level and the density of states of neighboring atoms. Interestingly, these additives may directly alter the reaction pathway or indirectly influence the chemical process by facilitating the proper adsorption of H* by adjacent atoms.^98,99 However, the instability of these structural additives and their strong dependence on the coordination environment pose challenges to heterogeneous catalysis. Noble metal single atoms are widely considered to be excellent regulators of HER, and the potential candidates for heterotopic doping as well. As depicted in Fig. 9a, Shi and coauthors utilized a photoreduction method to load dispersed Pt single atoms onto a 1T′-MoS₂ catalyst. Structural characterization revealed that the loaded Pt atoms predominantly exist in three forms: substitutional Pt (Pt_sub), Pt adsorbed on top of S (Pt_ads-S), and Pt adsorbed on top of Mo (Pt_ads-Mo). Among these, the Pt_ads-Mo sites exhibited nearly zero free energy for hydrogen evolution (Fig. 9b), resulting in a mass activity of 85 ± 23 A mg_Pt⁻¹ for the s-Pt/1T′-MoS₂ catalyst at an overpotential of −50 mV.¹⁰⁰

Fig. 9 (a) Atomic-resolution HAADF-STEM image of s-Pt/1T′-MoS₂ showing the isolated Pt atomically dispersed on 1T′-MoS₂ NS. Three kinds of single-atomically dispersed Pt, denoted as Pt_sub, Pt_ads-S and Pt_ads-Mo, respectively, are enclosed by the white dotted rectangles. (Inset) Simulated atomic structure diagram and corresponding FFT pattern. (b) Calculated ΔG_H* diagrams for HER. (c) Schematic diagram of the fabrication process for SA Co-D 1T MoS₂. (d) The electron density difference of 3 × 3 case. (e) HER polarization plots of MoS₂, 1T MoS₂ prepared by lithiation, Co NDs/MoS₂ and SA Co-D 1T MoS₂ without and with SCN-ions. The inset shows that the cobalt HER active centers are blocked by SCN-ions.

Utilizing the d-orbital effects of transition metal Co, Qi et al. employed cyclic voltammetry (CV) to deposit Co nanodisks (NDs) onto the MoS₂ surface (Fig. 9c). The incorporated Co atoms were found to occupy positions atop Mo atoms rather than at MoS₂ defects or edge sites, leading to the formation of a distorted 1T phase (Fig. 9d). Research indicates that these single Co atoms act as active centers, directly participating in the binding of H intermediates on the MoS₂. The resulting Co-D 1T MoS₂ demonstrated outstanding HER performance as illustrated in Fig. 9e, with a low Tafel slope of 32 mV dec⁻¹ and long-term stability over 2400 h.¹⁰¹ To develop a highly stable interstitial doping system, Zhou et al. incorporated carbon atoms with an ultra-small atomic radius and high chemical potential into the sublayer interstices of MoS₂. The study focused on the structure–performance relationship of the HER in the interstitial carbon-doped MoS₂ electrocatalyst (C_ia-MoS₂). Through hydrothermal synthesis, the atomic C was strategically embedded into the fcc sites of the MoS₂ cell, resulting in a favorable local charge landscape and a 1T phase that promoted HER activity. Unlike previous studies, the C_ia did not directly accelerate the HER process but instead served as a structural stabilizer for the emerging metallic phase. The synthesized C_ia-MoS₂ catalyst showed an 81% reduction in overpotential at 10 mA cm⁻² compared to pristine MoS₂.⁹⁹ The introduction of heterotopic structural additives is increasingly being developed to enhance the stability of active phases, or provide additional adsorption sites and conductivity. These atypical doping sites can effectively alter reaction pathways and are generally suitable for additives with high dispersion, small size, and large chemical potential differences from the host catalyst.

4.4 Heterostructure construction

When 2D materials are combined through regions with different atomic arrangement orientations, rich physical and chemical properties have emerged, and the transition zone is called domain wall.^102,103 Fansy phenomenon can be seen near the dislocation region, e.g., effective charge transfer, impurity band, intermediate structure and inhomogeneous electronic structure.^104,105 Hong et al. developed a chemical vapor deposition technique to synthesize a single-layer MoS₂ with a 1T–2H heterostructure, in which the 1T phase was as high as 61.5%. DFT calculations revealed that the strong electronic coupling was observed at the heterogeneous interface, and the electrons spontaneously transfer from the 2H phase to the 1T phase. This phenomenon was lead an increased electronic density around 1T phase while a depletion of electrons around the 2H phase, causes the transition region a low HER barrier (−0.156 eV) and fast kinetic process (Tafel slope is 78 mV dec⁻¹).¹⁰⁶ Along this line, more heterogeneous structure can be organized by MoS₂ and other materials. Such as 1T_0.81-MoS₂@Ni₂P HER catalyst studied by Liu et al. Firstly, the high electronegativity of P drives pristine MoS₂ partial shift to 1T phase; then the electrons were transfer from the MoS₂ into Ni₂S or Ni₂P regions, leading a rich electronic density environment around the domain (Fig. 10a and b). According to the rich heterogeneous interfaces, the as synthesized 1T_0.81-MoS₂@Ni₂P catalyst accelerates the H₂O activation, as well as the generation of intermediate H species (Fig. 10c and d), which possesses an extremely low overpotential in acidic (38.9 mV) and alkaline (95 mV) solutions at 10 mA cm⁻², corresponding to a considerable Tafel slope of 41 and 42 mV dec⁻¹, respectively.¹⁰⁷

Fig. 10 (a) The deformation of the electronic density of 2H_phase-MoS₂@NiS₂ and 1T_phase-MoS₂@NiS₂ interface, in which yellow/green isosurfaces correspond to positive/negative spin densities (0.00295308 e Å⁻³). (b) Band structure and density of states (DOS) for 1T_phase-MoS₂@NiS₂. (c) Free-energy diagrams for HER on the 2H_phase-MoS₂, 1T_phase-MoS₂, pure Ni₂P, pure NiS₂, 1T/2H_mix-MoS₂, 2H_phase-MoS₂@Ni₂P, 2H_phase-MoS₂@NiS₂, 1T_phase-MoS₂@NiS₂ and 1T_phase-MoS₂@Ni₂P interface edge. (d) Schematics showing water activation, *H intermediate formation and hydrogen generation on multi-heterojunction electrocatalysts. (e) The HR-TEM images of the Mo₅N₆-MoS₂ heterojunction by in-plane chemical bonding of MoS₂ (100) basal planes and Mo₅N₆ (110) plane. (f) Schematic diagrams of the energy band of Schottky-type Mo₅N₆-MoS₂ heterojunction (E_vac = vacuum energy, E_c = conduction band, E_v = valence band, E_f = Fermi level, E_g = energy gap, W = work function, and Φ = depletion region). (g) Energy profiles for water dissociation and free energy diagrams for HER on 2H-MoS₂ basal plane, 2H-MoS₂ edge plane, Mo₅N₆ (110), and Mo₅N₆-MoS₂ heterointerface.

Constructing in-plane contact between metal and semiconductor is another effective approach to generate a built-in electric field using Schottky barriers. Pi et al. prepared the Mo₅N₆-MoS₂ heterojunction nanosheets (Mo₅N₆-MoS₂/HCNRs) on conductive hollow carbon nanoribbons by a simple hydrothermal method (Fig. 10e). The results showed that there was a Fermi level difference in the metal/semi-contact, which led to the electron transfer from 2H-MoS₂ to Mo₅N₆ (Fig. 10f). In addition, the lattice dislocations led to abundant defective sites, which increased the number of active sites and charge transfer process, giving the catalyst a low overpotential of 53 mV at 10 mA cm⁻² and a Tafel slope of 37.9 mV dec⁻¹ (Fig. 10g).¹⁰⁸ Heterogeneous structures are typically characterized by abundant crystal domains, and fully utilizing these interfaces can significantly enhance the electrical properties of catalysts. In addition, domain walls are often available for H* species adsorption and activation, which play a role in increasing active sites.

5 Interlayer coupling engineering

Layered intercalation is an effective method to modulate interlayer coupling and crystal field in MoS₂.¹⁰⁹ By widening the vdW gap, the crystal field of MoS₂ switches from tetrahedral to octahedral coordination, leading to a re-splitting of energy levels.^23,77 In addition, direct growth of MoS₂ on metal support has been emerging as a novel approach to synthesize catalytic complexes. Taking advantage of the Fermi level difference between metals and semi, a large number of electrons can be transferred to MoS₂ through metal supports, thus promoting the reconstruction of active centers.^22,110–112 This section focuses on two typical interlayer couplings and their applications for enhancing the HER performance of MoS₂ catalysts.

5.1 Modulation of vdW gap

2D layered MoS₂ are attracting each layer by weak physical vdW interaction (shown as Fig. 11c), and the properties of MoS₂ catalysts directly endued by the height of vdW gap.¹¹³ Existing strategies for expanding the interlayer gap are mainly based on electrochemical intercalation techniques, whereby additive ions, molecules, and nanoparticles are driven into the vdW gaps of MoS₂.^34,114

Fig. 11 (a) Zero-valent intercalation of metal nanoparticles by an in situ reduction strategy. (b) LSV curves of metal intercalated MoS₂ catalysts. (c) Intrinsic van der Waals distance and van der Waals gap of MoS₂ (van der Waals distance = 6.2 Å, van der Waals gap = 1.8 Å). (d) Schematic illustrations of the vdW gap engineering achieved by adsorption of water molecules, and the cross-sectional STEM image of the vdW interface formed by vdW gap engineering, showing a clear vdW gap while maintaining atomically sharp and clean interfaces. Scale bars, 5 nm (STEM image).

Alkali metal ions are popular in modifying the vdW gap and chemical properties of MoS₂, as they have extremely small diameter and low ionization potential.¹¹⁵ For example, Cui et al. reported a vertically aligned MoS₂ electrocatalyst intercalated by Li⁺ for rapid HER. First of all, the Li⁺ acts as a gap extender, which directly changes the vdW interaction and alters the electronic properties. Secondly, Li⁺ transfers extra electrons to MoS₂ due to its high reducibility, thus changing the spin–orbit coupling of Mo 3d orbital. Last, it was proved that the 2H phase MoS₂ shifts into 1T phase when the concentration of Li⁺ is over 0.28.¹⁹ In addition, the capillary action of MoS₂ powder and the molten K⁺ was used for the modulation of the vdW gap. It was reported the added K atoms can be driven into the vdW gaps to prepare complex artificial superlattices, due to the difference between the electron affinity of MoS₂ (4.45 eV) and the ionization potential of K⁺ (4.34 eV).²⁶ Based on this, as depicted in Fig. 11a and b, Chen et al. achieved the intercalation of several metal nanoparticles into a Li⁺-intercalated 1T′ Li_xMoS₂ catalyst by in situ reduction of noble metal ion precursors. In this process, the wider vdW gaps provided channels for the diffusion of target noble species, endowing the novel catalyst with a higher redox potential. Unlike the in-plane doping, the intercalation of alkali metal ions species reacts with surface S atoms and forms M–S bonds (M = K, Na, Pt, Au), which suppresses electron scattering and maintains a high electron concentration around the d orbitals of Mo.

However, the introduction of alkali metal ions was realized through a hysteresis diffusion-limited process, leading to a fuzzy crystalline phase and reduced structural stability.¹¹⁶ To overcome these drawbacks, a self-intercalation approach was developed by Loh and his colleagues, where the host metal atoms move into the vdW gaps by exploiting their high chemical potential. It was found that the self-driven intercalant introduces additional spin-split bands, which can modulate the electronic properties of the carrier transport.¹¹⁷ An advanced solution was proposed by Zou et al., who exploited the vdW gaps to confine small pre-adsorbed water molecules. By controlling the saturation vapour pressure of water, the ‘vdW water gap’ was monotonic increasing up to 53.6 Å (Fig. 11d). This gap-engineering approach highlights the potential of vdW water gaps to modulate electronic properties and more utility for diode and electrochemical devices.¹¹³

5.2 Semiconductor/metal contacting

Generally, 2D films are grown on support like metal (e.g., Au, Pd, Cu, and Mo) or oxides (e.g., mica, sapphire, and silica) for robust action.²⁰ Much effort has been put into exploring various supports, and a strong interaction between Au foil and MoS₂ films has been discovered in 2014. It is revealed that the electronic coupling effects between Au and MoS₂ and eliminate the charge transfer resistance, thus promote the catalysts reaching a high exchange current density of 38.1 μA cm⁻².⁵⁹ Obviously, semi/metal contacting effects play a vital role in design Schottky barrier and regulate the conductivity. Voiry et al. pioneered to concern the relationship of semi/metal contact effects on HER activity,²² and Charlie Tsai and coauthors studied how different supports influence the Gibbs free energy of Mo edge sites. Their results show that the HER reactivity of edged-Mo sites can be influenced by tuning the contact effects between semi-MoS₂ films and different metal supports, where the electronic interaction induced Fermi level shift and reduced interface resistance.¹¹⁸ Along this line, as depicted in Fig. 12a, several metals such as In, Ti, Au, Pd, and Mo were used as simulated support to contact with the surface and edges part of monolayer MoS₂. The orbital overlap, Schottky barrier, and tunnel barriers of TMDs and metal contact effects were disclosed by DFT calculations. Interestingly, the separate distance (d) of the metal-edge is much smaller than the distance of metal-surface in all combinations, thus strong covalent bonds can form between the metal and edge site. In addition, an atomic orbital overlap of sulfides and metal can be found in each case, which may lead to a reduction of the Schottky barrier and inner resistance of the monomer catalyst.¹¹⁹

Fig. 12 (a) Optimized geometries of top contacts to MoS₂: Au-MoS₂ (in different views), In-MoS₂, Pd-MoS₂, Ti-MoS₂, Mo-MoS₂ (in different views). d is defined as the physical separation (the z component of the nearest core-to-core distance between the metal atoms and the chalcogenide atoms). (b) Atomic projected electronic band structure of Sb (0001)–MoS₂ contact and charge density near junction EF. (c) Atomic projected electronic band structure of Sb (0112)–MoS₂ contact and charge density near junction EF (red, positive; blue, negative). (d) (Left) Cross-section HAADF-STEM image of the Sb (0112)–MoS₂ contact. Scale bar, 2 nm. (Right) Zoom-in atomic-resolution image from the red box in (c). The Mo, S and Sb atoms are overlayed on the image. The vdW gap of 0.285 nm is marked. Scale bar, 1 nm. (e) A schematic showing the preparation process of CuMo₆S₈/Cu electrode. (f) Cross-sectional high-resolution transmission electron microscopy (HRTEM) zoom-in view of the CuMo₆S₈/Cu electrode. The inset of h is the corresponding fast Fourier transform patterns of CuMo₆S₈ (top) and Cu (bottom). (g) Polarization curves of three electrodes including CuMo₆S₈/Cu, MoS₂, and Pt/C. All tests are done in 1 M KOH at a scan rate of 1 mV s⁻¹ with 85% iR correction.

More recently, Li et al. studied the contact effect between Sb with different plane orientations Sb (0001) and Sb (0112) prepared by electron beam evaporation and monolayer MoS₂ (Fig. 12b–d). It is observed that the Mo d orbital has a large real-space overlap with the Sb (0112) p_z orbital in the vertical direction, resulting in strong Sb (0112)–MoS₂ hybridization and forcing the interface resistance to approach the quantum limit. In contrast, no obvious electron localization was observed between Mo d orbital and Sb (0001) p orbital.¹²⁰ This work emphasizes the significance of crystal plane orientation in the preparation of zero contact resistance composite catalyst. Directly construct carrier pathway between metal support and MoS₂ active component also considered an effective method to tune semi/metal contact interface. For example, Shin and Liu developed a highly mechanically stable novel electrocatalyst by growing MoS₂ nanosheets on a self-support Cu foil, where the MoS₂ are vertically embedded in the Cu (Fig. 12e). Differently, there is a composite region between the MoS₂ and Cu, Cu_xMo₆S₈ (CMS), which is known as the Chevrel phase (Fig. 12f). It was shown that the CMS region serves as an effective charge transfer channel to promote the carriers transfer into the MoS₂ active component, thus exhibiting a low overpotential of ∼334 mV and sustained stability of over 100 h at a current density of 2500 mA cm⁻² (Fig. 12g).^121,122 Similar studies were reported by Liu's group, who achieved in situ growth of a vertically oriented TaS₂ active layer on a metallic Ta by an oriented solid-phase synthesis method. It was confirmed that electrons can directly inject into TaS₂ layers without passing the vdW gap. Additionally, this novel integrated structure demonstrates excellent mechanical stability, with the Ta/TaS₂ catalyst operating for over 200 hours while maintaining a low overpotential of 398 mV at a high current density of 2000 mA cm⁻². These advances pave the way to apply this monomer catalyst to boost HER process.¹²³

6 Challenges and outlooking

Although the unsaturated electronic structure is regarded as the key feature of MoS₂ for achieving efficient HER, the long-term electrocatalytic stability is often compromised by the emergence of active structural configurations. Fig. 13a and b illustrate some common cases associated with HER performance. Based on current electrocatalytic test data, it is evident that TMDs based on Mo element tend to perform better under acidic conditions,^{99,101,107,122,124–128} whereas catalysts containing Ni element exhibit greater durability in alkaline environments.^{92,107,108,129–132} These findings provide valuable guidance for selecting appropriate catalysts for large-scale HER processes. As the study of constructing and regulating the structure–activity relationship of MoS₂-like catalysts at the atomic level is still in its early stages, we propose several research topics to assist researchers to navigate through the scientific opportunities in this area (Fig. 14).

Fig. 13 HER performance of various catalysts under (a) acidic conditions and (b) alkaline conditions.

Fig. 14 The potential development researches from modern theoretical analysis, advanced synthesis and characterization, and possible applications in electrocatalysis for MoS₂.

6.1 Electronic structure evolution of central metal during electrocatalysis and decay

The evolution of the electron states of d-orbital of the central metal site in transition metal sulfide catalysts during electrocatalysis remains a significant challenge.¹³³ Some studies have reported that utilize low-temperature or in situ characterization approaches can allow for the semi-quantitative analysis of the local electronic structure and unsaturated states of transition metals. In a reaction process involving electrons transfer, electron transitions within the central metal, electron donation from dopants to the central metal, and the transfer of delocalized electrons between the central metal and ligands can alter the redox properties of the metal sites, thus, the formation, adsorption, or desorption of intermediates and products. Unlike other catalytic mechanisms, the presence of additional energy from magnetocaloric effects associated with intermediate or high-spin electrons could enhance reaction kinetics or intermediate mobility. This could lead to the appearance of “mirages” in electrocatalytic reactions that thermodynamic theory would deem impossible. Hence, researches on the influence of spin effects in transition metal catalysts on reaction activity and selectivity may become a critical focus in future catalysis research.

6.2 Role of unsaturated electronic structure in the interaction between reactants/products and catalyst centers

Unsaturated electronic structures are typically accompanied by unpaired electrons, which significantly impact the magnetic properties and polarity of local catalytic centers. As a result, investigating the role of electron deficiency in spin states, magnetism, polarity, and conductivity in MoS₂ could offer valuable insights into reaction mechanisms and Gibbs free energy pathways, especially in the context of MoS₂ catalysts applied to more complex reactions. For instance, when products are present in a non-polar bubble state, adjusting the polarity of catalytic centers may be crucial for enhancing the desorption and diffusion of bubble-phase products. Despite bubble trouble becomes a growing focus on this topic in large-scale HER processes, research from this perspective remains limited.

6.3 Advantages of metal/semiconductor contacts in constructing highly active electronic structures

In heterogeneous electrocatalysis, harnessing the interfacial contact effects between supported metals and semiconductor catalysts to modulate band positions, charge transfer, and structural stability of electrocatalysts remain a significant challenge. Recent studies suggest that as the van der Waals gap between 2D layered catalysts narrows, it can promote the formation of new chemical bonds or delocalized electrons, thereby creating internal pathways for electron transfer. Thus, exploring and harnessing the interfacial contact effects between supported metals and semiconductor catalysts may offers a promising “two birds with one stone” strategy for monomer catalysts in electrocatalysis-simultaneously improving catalyst stability and optimizing the active electronic structure.

6.4 Advanced characterization methods to study unsaturated electronic structures

In the exploration of unsaturated electronic structures in 2D materials (such as MoS₂), advanced characterization methods are crucial. Commonly used techniques include Raman spectroscopy, X-ray absorption spectroscopy (XAS), X-ray photoelectron spectroscopy (XPS), and electron paramagnetic resonance (EPR). These approaches provide valuable information about electronic structure, chemical environment, and unpaired electrons, but they face challenges when applied to light-sensitive 2D materials. For instance, while Raman spectroscopy is sensitive to unsaturated electronic structures, the signals are often weak at low-energy excitation source and susceptible to interference from fluorescence. XAS and XPS can offer semi-quantitative data on electronic states, but they mainly capture average structures, making it difficult to precisely characterize local defects or active sites. Future advancements may involve combining high-angle annular dark field scanning transmission electron microscopy (HAADF-STEM) with electron energy loss spectroscopy (EELS) to enhance spatial resolution, along with computational simulations (such as DFT calculations) to predict electron deficiency impact on the electrocatalytic performance.

6.5 Utilizing unsaturated sites to couple hydrogen reduction with other important reactions

By employing various strategies, both electrophilic and nucleophilic active sites on MoS₂ can be designed to meet the adsorption and activation requirements of different chemically active species. Introducing sulfur vacancies or other defects in the 2H phase MoS₂ creates metallic-like sites with unpaired electrons, which effectively adsorb and activate electrophilic species such as CO₂ and CO. Additionally, by doping high-electronegativity noble metals (such as Pt and Au), positively charged nucleophilic Mo sites (Mo^δ+) can be engineered on the MoS₂ surface to interact with electron-rich reactants. Under appropriate stoichiometric conditions, these active sites facilitate in situ hydrogenation reactions, yielding high-value products such as methanol, formic acid, or other hydrocarbons. This approach demonstrates that MoS₂ not only exhibits excellent catalytic performance in HER but also holds promise for applications in hydrogenation, CO₂ reduction reactions, and other significant processes.

Data availability

No primary research results, software or code have been included and no new data were generated or analyzed as part of this review.

Author contributions

All of the authors contributed to the manuscript preparation. Q. Z., X. R., and D. M. conceived the outline of the manuscript. Q. Z., and H. H. wrote the original draft of the manuscript. Z. C., X. R., and D. M. discussed and helped revise the manuscript.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors thank the financial support from the National Natural Science Foundation of China (52202202 and 22472005) and the Provincial Key Research and Development Project of Sichuan (2024YFHZ0040).

References

1. M. Wagner, B. Meyer, M. Setvin, M. Schmid and U. Diebold, Nature, 2021, 592, 722–725.
2. W. Huang, Q. He, Y. P. Hu and Y. G. Li, Angew. Chem., Int. Ed., 2019, 131, 8768–8772.
3. T. Binninger and M. L. Doublet, Energy Environ. Sci., 2022, 15, 2519–2528.
4. K. L. Hu, T. Ohto, Y. Nagata, M. Wakisaka, Y. Aoki, J. Fujita and Y. Ito, Nat. Commun., 2021, 12, 203.
5. P. Garrido-Barros, J. Derosa, M. J. Chalkley and J. C. Peters, Nature, 2022, 609, 71–76.
6. J. Greeley, T. F. Jaramillo, J. Bonde, I. B. Chorkendorff and J. K. Nørskov, Nat. Mater., 2006, 5, 909–913.
7. Z. W. Seh, J. Kibsgaard, C. F. Dickens, I. Chorkendorff, J. K. Nørskov and T. F. Jaramillo, Science, 2017, 355, eaad4998.
8. H. Y. Kim, J. M. Kim, Y. Ha, J. Woo, A. Byun, T. J. Shin, K. H. Park, H. Y. Jeong, H. Kim, J. Y. Kim and S. H. Joo, ACS Catal., 2019, 9, 11242–11254.
9. A. S. Dudnik, V. L. Weidner, A. Motta, M. Delferro and T. J. Marks, Nat. Chem., 2014, 6, 1100–1107.
10. Q. Q. Zhou, X. L. Zhou, R. H. Zheng, Z. F. Liu and J. D. Wang, Sci. Total Environ., 2022, 806, 150088.
11. B. Hinnemann, P. G. Moses, J. Bonde, K. P. Jørgensen, J. H. Nielsen, S. Horch, I. Chorkendorff and J. K. Nørskov, J. Am. Chem. Soc., 2005, 127, 5308–5309.
12. M. A. Lukowski, A. S. Daniel, F. Meng, A. Forticaux, L. Li and S. Jin, J. Am. Chem. Soc., 2013, 135, 10274–10277.
13. M. Fang, Y. Q. Peng, P. W. Wu, H. Wang, L. X. Xing, N. Wang, C. M. Tang, L. Meng, Y. K. Zhou, L. Du and S. Y. Ye, Front. Energy, 2024, 18, 308–329.
14. T. F. Jaramillo, K. P. Jørgensen, J. Bonde, J. H. Nielsen, S. Horch and I. Chorkendorff, Science, 2007, 317, 100–102.
15. Y. H. Jia, Y. C. Zhang, H. Q. Xu, J. Li, M. Gao and X. T. Yang, ACS Catal., 2024, 14, 4601–4637.
16. H. S. Meng, Z. J. Chen, J. L. Zhu, B. You, T. Y. Ma, W. Wei, S. Vernuccio, J. Xu and B. J. Ni, Adv. Funct. Mater., 2024, 34, 2405270.
17. H. T. Wang, Z. Y. Lu, S. C. Xu, D. S. Kong, J. J. Cha, G. Y. Zheng, P. C. Hsu, K. Yan, D. Bradshaw, F. B. Prinz and Y. Cui, Proc. Natl. Acad. Sci. U.S.A., 2013, 110, 19701–19706.
18. C. Tsai, H. Li, S. Park, J. Park, H. S. Han, J. K. Nørskov, X. L. Zheng and F. Abild-Pedersen, Nat. Commun., 2017, 8, 15113.
19. H. T. Wang, C. Tsai, D. S. Kong, K. Chan, F. Abild-Pedersen, J. K. Nørskov and Y. Cui, Nano Res., 2015, 8, 566–575.
20. J. Zhang, Q. Y. Zhang and X. L. Feng, Adv. Mater., 2019, 31, 1808167.
21. K. Fujisawa, B. R. Carvalho, T. Zhang, N. Perea-Lopez, Z. Lin, V. Carozo, S. L. L. M. Ramos, E. Kahn, A. Bolotsky, H. Liu, A. L. Elias and M. Terrones, ACS Nano, 2021, 15, 9658–9669.
22. D. Voiry, R. Fullon, J. U. Yang, C. C. C. Silva, R. Kappera, I. Bozkurt, D. Kaplan, M. J. Lagos, P. E. Batson, G. Gupta, A. D. Mohite, L. Dong, D. Er, V. B. Shenoy, T. Asefa and M. Chhowalla, Nat. Mater., 2016, 15, 1003–1009.
23. S. Park, C. Kim, S. O. Park, N. K. Oh, U. Kim, J. Lee, J. Seo, Y. Yang, H. Y. Lim, S. K. Kwak, G. Kim and H. Park, Adv. Mater., 2020, 32, 2001889.
24. Z. J. Chen, S. N. Yun, L. Wu, J. Q. Zhang, X. D. Shi, W. Wei, Y. W. Liu, R. J. Zheng, N. Han and B. J. Ni, Nano-Micro Lett., 2023, 15, 4.
25. Z. J. Chen, X. G. Duan, W. Wei, S. B. Wang and B. J. Ni, Nano Energy, 2020, 78, 105270.
26. W. Yang and S. W. Chen, Chem. Eng. J., 2020, 393, 124726.
27. Q. H. Wang, K. Kalantar-Zadeh, A. Kis, J. N. Coleman and M. S. Strano, Nat. Nanotechnol., 2012, 7, 699–712.
28. B. Radisavljevic, A. Radenovic, J. Brivio, V. Giacometti and A. Kis, Nat. Nanotechnol., 2011, 6, 147–150.
29. Y. Li, K. A. N. Duerloo, K. Wauson and E. J. Reed, Nat. Commun., 2016, 7, 10671.
30. H. M. Wang, C. H. Li, P. F. Fang, Z. L. Zhang and J. Z. Zhang, Chem. Soc. Rev., 2018, 47, 6101.
31. Y. R. Lim, J. K. Han, Y. Yoon, J. B. Lee, C. Jeon, M. Choi, H. Chang, N. Park, J. H. Kim, Z. Lee, W. Song, S. Myung, S. S. Lee, K. S. An, J. H. Ahn and J. S. Lim, Adv. Mater., 2019, 31, 1901405.
32. D. Zhu, J. Liu, Y. Zhao, Y. Zheng and S. Z. Qiao, Small, 2019, 15, 1805511.
33. H. Yang, S. Kim, W. M. Chhowalla and Y. H. Lee, Nat. Phys., 2017, 13, 931–937.
34. M. Chhowalla, H. S. Shin, G. Eda, L.-J. Li, K. P. Loh and H. Zhang, Nat. Chem., 2013, 5, 263–275.
35. R. M. A. Khalil, F. Hussain, A. M. Rana, M. Imran and G. Murtaza, Phys. E Low-dimens. Syst. Nanostruct., 2019, 106, 338–345.
36. Q. Yue, J. Kang, Z. Z. Shao, X. A. Zhang, S. L. Chang, G. Wang, S. Q. Qin and J. B. Li, Phys. Lett. A, 2012, 376, 1166–1170.
37. Q. Z. Huang, J. L. Shen, Y. Lu, R. D. Ye and S. Gong, J. Phys. Chem. C, 2023, 127, 17406–17414.
38. L. A. Zavala, K. Kumar, V. Martin, F. Maillard, F. Maugé, X. Portier, L. Oliviero and L. Dubau, ACS Catal., 2023, 13, 1221–1229.
39. F. Yang, P. Hu, F. F. Yang, B. Chen, F. Yin, K. Hao, R. Sun, L. Gao, Z. Sun, K. Wang and Z. Yin, Small, 2023, 19, 2301468.
40. Y. Cao, ACS Nano, 2021, 15, 11014–11039.
41. Y. W. Zhang, Q. Wu, J. Z. Y. Seow, Y. J. Jia, X. Ren and Z. C. J. Xu, Chem. Soc. Rev., 2024, 53, 8123–8136.
42. Y. Y. Linghu, N. Li, Y. P. Du and C. Wu, Phys. Chem. Chem. Phys., 2019, 21, 9391.
43. J. Xu, G. L. Shao, X. Tang, F. Lv, H. Y. Xiang, C. F. Jing, S. Liu, S. Dai, Y. G. Li, J. Luo and Z. Zhou, Nat. Commun., 2022, 13, 2193.
44. Y. C. Huang, Y. H. Sun, X. L. Zheng, T. Aoki, B. Pattengale, J. Huang, X. He, W. Bian, S. Younan, N. Williams, J. Hu, J. X. Ge, N. Pu, X. X. Yan, X. Q. Pan, L. J. Zhang, Y. G. Wei and J. Gu, Nat. Commun., 2019, 10, 982.
45. L. Y. Zhang, M. Li, A. Q. Zou, S. H. Yu, T. Xiong, L. Wang, J. J. He, Q. Fu, K. Sun, D. H. C. Chua and J. M. Xue, ACS Appl. Energy Mater., 2019, 2, 493–502.
46. X. Wang, Y. W. Zhang, H. N. Si, Q. H. Zhang, J. Wu, L. Gao, X. F. Wei, Y. Sun, Q. L. Liao, Z. Zhang, K. Ammarah, L. Gu, Z. Kang and Y. Zhang, J. Am. Chem. Soc., 2020, 142, 4298–4308.
47. Y. Q. Fang, X. Z. Hu, W. Zhao, J. Pan, D. Wang, K. J. Bu, Y. L. Mao, S. F. Chu, P. Liu, T. Y. Zhai and F. Q. Huang, J. Am. Chem. Soc., 2019, 141, 790–793.
48. T. Zhang, Y. P. Liu, J. Yu, Q. T. Ye, L. Yang, Y. Li and H. J. Fan, Adv. Mater., 2022, 34, 2202195.
49. G. L. Shao, M. Q. Yang, H. Y. Xiang, S. Luo, X. X. Xue, H. M. Li, X. Zhang, S. Liu and Z. Zhou, Nano Res., 2023, 16, 1670–1678.
50. T. Zhang, Q. T. Ye, Z. Y. Han, Q. Y. Liu, Y. P. Liu, D. S. Wu and H. J. Fan, Nat. Commun., 2024, 15, 6508.
51. J. W. Wang, L. Q. He, Y. H. Zhang, H. Y. Nong, S. N. Li, Q. K. Wu, J. Y. Tan and B. L. Liu, Adv. Mater., 2024, 36, 2314145.
52. V. I. Klimov, A. A. Mikhailovsky, D. W. McBranch, C. A. Leatherdale and M. G. Bawendi, Phys. Rev. B:Condens. Matter Mater. Phys., 2000, 61, R13349.
53. F. Wang, T. A. Shifa, X. Zhan, Y. Huang, K. Liu, Z. Cheng, C. Jiang and J. He, Nanoscale, 2015, 7, 19764–19788.
54. K. S. Novoselov, D. Jiang, F. Schedin, T. J. Booth, V. V. Khotkevich, S. V. Morozov and A. K. Geim, Proc. Natl. Acad. Sci. U.S.A., 2005, 102, 10451–10453.
55. Q. Q. Ji, C. Li, J. L. Wang, J. J. Niu, Y. Gong, Z. P. Zhang, Q. Y. Fang, Y. Zhang, J. P. Shi, L. Liao, X. S. Wu, L. Gu, Z. F. Liu and Y. F. Zhang, Nano Lett., 2017, 17, 4908.
56. Z. P. Zhang, J. J. Niu, P. F. Yang, Y. Gong, Q. Q. Ji, J. P. Shi, Q. Y. Fang, S. L. Jiang, H. Li, X. B. Zhou, L. Gu, X. S. Wu and Y. F. Zhang, Adv. Mater., 2017, 29, 1702359.
57. Y. Zhang, Y. F. Zhang, Q. Q. Ji, J. Ju, H. T. Yuan, J. P. Shi, T. Gao, D. L. Ma, M. X. Liu, Y. B. Chen, X. J. Song, H. Y. Hwang, Y. Cui and Z. F. Liu, ACS Nano, 2013, 7, 8963–8971.
58. J. G. Song, J. S. Park, W. Lee, T. Choi, H. Jung, C. W. Lee, S. H. Hwang, J. M. Myoung, J. H. Jung, S. H. Kim, C. L. Matras and H. Kim, ACS Nano, 2014, 8, 8617.
59. J. P. Shi, D. L. Ma, G. F. Han, Y. Zhang, Q. Q. Ji, T. Gao, J. Y. Sun, X. J. Song, C. Li, Y. S. Zhang, X. Y. Lang, Y. F. Zhang and Z. F. Li, ACS Nano, 2014, 8, 10196–10204.
60. J. P. Shi, X. N. Wang, S. Zhang, L. F. Xiao, Y. H. Huan, Y. Gong, Z. P. Zhang, Y. C. Li, X. B. Zhou, M. Hong, Q. Y. Fang, Q. Zhang, X. F. Liu, L. Gu, Z. F. Liu and Y. F. Zhang, Nat. Commun., 2017, 8, 958.
61. K. Chang, X. Hai, H. Pang, H. B. Zhang, L. Shi, G. G. Liu, H. M. Liu, G. X. Zhao, M. Li and J. H. Ye, Adv. Mater., 2016, 28, 10033–10041.
62. L. J. Zhang and A. Zunger, Nano Lett., 2015, 15, 949–957.
63. J. Kibsgaard, Z. Chen, B. N. Reinecke and T. F. Jaramillo, Nat. Mater., 2012, 11, 963–969.
64. Z. Y. Huang, X. Qi, H. Yang, C. Y. He, X. L. Wei, X. Y. Peng and J. X. Zhong, J. Phys. D, 2015, 48, 205302.
65. D. S. Kong, H. T. Wang, J. J. Cha, M. Pasta, K. J. Koski, J. Yao and Y. Cui, Nano Lett., 2013, 13, 1341–1347.
66. X. Geng, W. Wu, N. Li, W. Sun, J. Armstrong, A. Al-hilo, M. Brozak, J. Cui and T. P. Chen, Adv. Funct. Mater., 2014, 24, 6123.
67. Y. Yang, H. Fei, G. Ruan, C. Xiang and J. M. Tour, Adv. Mater., 2014, 26, 8163.
68. J. Deng, H. B. Li, S. H. Wang, D. Ding, M. S. Chen, C. Liu, Z. Q. Tian, K. S. Novoselov, C. Ma, D. H. Deng and X. H. Bao, Nat. Commun., 2017, 8, 14430.
69. Y. M. Jiang, X. Li, S. J. Yu, L. P. Jia, X. J. Zhao and C. M. Wang, Adv. Funct. Mater., 2015, 25, 2693–2700.
70. X. P. Ren, L. Q. Pang, Y. X. Zhang, X. D. Ren, H. B. Fan and S. Z. Liu, J. Mater. Chem. A, 2015, 20, 10693–10697.
71. A. P. Nayak, T. Pandey, D. Voiry, J. Liu, S. T. Moran, A. Sharma, C. Tan, C. H. Chen, L. J. Li, M. Chhowalla, J. F. Lin, A. K. Singh and D. Akinwande, Nano Lett., 2015, 15, 346–353.
72. A. P. Nayak, S. Bhattacharyya, J. Zhu, J. Liu, X. Wu, T. Pandey, C. Q. Jin, A. K. Singh, D. Akinwande and J. F. Lin, Nat. Commun., 2014, 5, 3731.
73. Z. Zhao, H. J. Zhang, H. T. Yuan, S. B. Wang, Y. Lin, Q. S. Zeng, G. Xu, Z. X. Liu, G. K. Solanki, K. D. Patel, Y. Cui, H. Y. Hwang and W. L. Mao, Nat. Commun., 2015, 6, 7312.
74. H. Y. Yu, G. B. Liu, J. J. Tang, X. D. Xu and W. Yao, Sci. Adv., 2017, 3, e1701696.
75. Y. C. Lin, D. O. Dumcenco, Y. S. Huang and K. Suenaga, Nat. Nanotechnol., 2014, 9, 391–396.
76. Y. Katagiri, T. Nakamura, A. Ishii, C. Ohata, M. Hasegawa, S. Katsumoto, T. Cusati, A. Fortunelli, G. Iannaccone and G. Fiori, Nano Lett., 2016, 16, 3788–3794.
77. X. W. Guo, E. H. Song, W. Zhao, S. M. Xu, W. L. Zhao, Y. J. Lei, Y. Q. Fang, J. J. Liu and F. Q. Huang, Nat. Commun., 2022, 13, 5954.
78. W. Z. Wu, C. Y. Niu, C. Wei, P. Y. Jia, C. Li and P. Q. Xu, Angew. Chem., Int. Ed., 2019, 131, 2051–2055.
79. C. Y. Feng, Z. P. Wu, K. W. Huang, J. H. Ye and H. B. Zhang, Adv. Mater., 2022, 34, 2200180.
80. J. Xu, X. X. Xue, G. L. Shao, C. F. Jing, S. Dai, K. He, P. P. Jia, S. Wang, Y. F. Yuan, J. Luo and J. Lu, Nat. Commun., 2023, 14, 7849.
81. Y. Zhou, J. Zhang, E. H. Song, J. H. Lin, J. D. Zhou, K. Suenaga, W. Zhou, Z. Liu, J. J. Liu, J. Lou and H. J. Fan, Nat. Commun., 2020, 11, 2253.
82. Z. Y. Luo, Y. X. Ouyang, H. Zhang, M. L. Xiao, J. J. Ge, Z. Jiang, J. L. Wang, D. M. Tang, X. Z. Cao, C. P. Liu and W. Xing, Nat. Commun., 2018, 9, 2120.
83. Y. G. Liu, S. J. Guan, X. S. Du, Y. R. Chen, Y. Yang, K. L. Chen, Z. W. Zheng, X. Wang, X. Q. Shen, C. L. Hu and X. B. Li, Energy Fuels, 2023, 37, 5370–5377.
84. J. M. Zhang, X. P. Xu, L. Yang, D. J. Cheng and D. P. Cao, Small Methods, 2019, 3, 1900653.
85. B. Liu, Y. Cheng, B. Cao, M. Hu, P. Jing, R. Gao, Y. Du, J. Zhang and J. Liu, Appl. Catal., B, 2021, 298, 120630.
86. Y. P. Zang, S. W. Niu, Y. S. Wu, X. S. Zheng, J. Y. Cai, J. Ye, Y. F. Xie, Y. Liu, J. B. Zhou, J. F. Zhu, X. J. Liu, G. M. Wang and Y. Qian, Nat. Commun., 2019, 10, 1217.
87. Q. Yang, Z. Wang, L. Dong, W. Zhao, Y. Jin, L. Fang, B. Hu and M. Dong, J. Phys. Chem. C, 2019, 123, 10917–10925.
88. J. Ge, D. Zhang, J. Jin, X. Han, Y. Wang, F. Zhang and X. Lei, Mater. Today Energy, 2021, 22, 100854.
89. Y. Shi, D. Zhang, H. Miao, X. Wu, Z. Wang, T. Zhan, J. Lai and L. Wang, Sci. China: Chem., 2022, 65, 1829–1837.
90. L. Jiang, Y. J. Zhang, X. H. Luo, L. Yu, H. X. Li and Y. J. Li, Chem. Eng. J., 2021, 425, 130611.
91. Z. L. Zheng, L. Yu, M. Gao, X. Y. Chen, W. Zhou, C. Ma, L. H. Wu, J. F. Zhu, X. Y. Meng, J. T. Hu, Y. C. Tu, S. S. Wu, J. Mao, Z. Q. Tian and D. H. Deng, Nat. Commun., 2020, 11, 3315.
92. H. Li, F. Yu, X. Ling, H. Wan, M. Zhang, Y. Zhou, J. Wei, F. Lu, X. Zhang, X. Zeng and M. Zhou, Nanotechnology, 2021, 32, 445703.
93. D. Han, Z. Luo, Y. Li, N. Gao, J. Ge, C. Liu and W. Xing, Appl. Surf. Sci., 2020, 529, 147117.
94. Z. Y. Luo, H. Zhang, Y. Q. Yang, X. Wang, Y. Li, Z. Jin, Z. Jiang, C. P. Liu, W. Xing and J. J. Ge, Nat. Commun., 2020, 11, 1116.
95. Y. T. Luo, S. Q. Zhang, H. Y. Pan, S. J. Xiao, Z. L. Guo, L. Tang, U. Khan, B. F. Ding, M. Li, Z. Y. Cai, Y. Zhao, W. Lv, Q. L. Feng, X. L. Zou, J. H. Lin, H. M. Cheng and B. L. Liu, ACS Nano, 2020, 14, 767–776.
96. J. Wang, C. Cheng, Q. Yuan, H. Yang, F. Q. Meng, Q. H. Zhang, L. Gu, J. L. Cao, L. G. Li, S. C. Haw, Q. Shao, L. Zhang, T. Cheng, F. Jiao and X. Q. Huang, Chem, 2022, 8, 1673–1687.
97. J. Yang, Y. F. Cao, S. Y. Zhang, Q. W. Shi, S. Y. Chen, S. C. Zhu, Y. S. Li and J. F. Huang, Small, 2023, 19, 2207295.
98. V. P. Pham and G. Y. Yeom, Adv. Mater., 2016, 28, 9024–9059.
99. Q. Q. Zhou, Z. Y. Wang, H. D. Yuan, J. D. Wang and H. Hu, Appl. Catal., B, 2023, 332, 122750.
100. Z. Y. Shi, X. Zhang, X. Q. Lin, G. G. Liu, C. Y. Ling, S. B. Xi, B. Chen, Y. Y. Ge, C. L. Tan, Z. C. Lai, Z. Q. Huang, X. Y. Ruan, L. Zhai, L. J. Li, Z. J. Li, X. X. Wang, G. H. Nam, J. W. Liu, Q. Y. He, Z. Q. Guan, J. L. Wang, C. S. Lee, A. R. J. Kucernak and H. Zhang, Nature, 2023, 621, 300–305.
101. K. Qi, X. Q. Cui, L. Gu, S. S. Yu, X. F. Fan, M. C. Luo, S. Xu, N. B. Li, L. R. Zheng, Q. H. Zhang, J. Y. Ma, Y. Gong, F. Lv, K. Wang, H. H. Huang, W. Zhang, S. J. Guo, W. T. Zheng and P. Liu, Nat. Commun., 2019, 10, 5231.
102. H. Hu, Q. Q. Zhou, Z. Y. Wang, J. D. Wang, Y. M. Chen and Y. X. Han, Chin. Sci. Bull., 2024, 69, 4299–4310.
103. H. Xiong, C. F. Du, Z. L. Ma, R. C. Zhi, S. S. Hao, X. Y. Zhao, Z. Liu, F. Xu and H. Q. Wang, Adv. Funct. Mater., 2024, 34, 2402298.
104. W. S. Chen, J. J. Gu, Y. P. Du, F. Song, F. X. Bu, J. H. Li, Y. Yuan, R. C. Luo, Q. L. Liu and D. Zhang, Adv. Funct. Mater., 2020, 30, 2000551.
105. Z. M. Sun, L. Lin, M. W. Yuan, H. Y. Yao, Y. J. Deng, B. H. Huang, H. F. Li, G. B. Sun and J. Zhu, Nano Energy, 2022, 101, 107563.
106. Z. A. Hong, W. T. Hong, B. C. Wang, Q. Cai, X. He and W. Liu, Chem. Eng. J., 2023, 460, 141858.
107. M. Q. Liu, J. A. Wang, W. Klysubun, G. G. Wang, S. Sattayaporn, F. Li, Y. W. Cai, F. C. Zhang, J. Yu and Y. Yang, Nat. Commun., 2021, 12, 5260.
108. C. R. Pi, X. X. Li, X. M. Zhang, H. Song, Y. Zheng, B. Gao, A. Kızılaslan, P. K. Chu and K. F. Huo, Small, 2022, 18, 2201137.
109. Y. H. Xue, Q. Zhang, W. J. Wang, H. Cao, Q. H. Yang and L. Fu, Adv. Energy Mater., 2017, 7, 1602684.
110. Y. C. Wu, D. f. Li, C. L. Wu, H. Y. Hwang and Y. Cui, Nat. Rev. Mater., 2023, 8, 41–53.
111. D. Xu, S. N. Zhang, J. S. Chen and X. H. Li, Chem. Rev., 2023, 123, 1–30.
112. Z. J. Chen, T. Y. Ma, W. Wei, W. Y. Wong, C. Zhao and B. J. Ni, Adv. Mater., 2024, 36, 2401568.
113. C. Liu, X. M. Zou, Y. W. Lv, X. Q. Liu, C. Ma, K. L. Li, Y. Liu, Y. Chai, L. Liao and J. He, Nat. Nanotechnol., 2024, 19, 448–454.
114. H. T. Yuan, H. T. Wang and Y. Cui, Acc. Chem. Res., 2015, 48(1), 81–90.
115. X. Huang, Z. Y. Zeng and H. Zhang, Chem. Soc. Rev., 2013, 42, 1934–1946.
116. Z. X. Chen, K. Leng, X. X. Zhao, S. Malkhandi, W. Tang, B. B. Tian, L. Dong, L. R. Zheng, M. Lin, B. S. Yeo and K. P. Loh, Nat. Commun., 2017, 8, 14548.
117. X. X. Zhao, P. Song, C. C. Wang, A. C. Riis-Jensen, W. Fu, Y. Deng, D. Y. Wan, L. X. Kang, S. C. Ning, J. D. Dan, T. Venkatesan, Z. Liu, W. Zhou, K. S. Thygesen, X. Luo, S. J. Pennycook and K. P. Loh, Nature, 2020, 581, 171–177.
118. C. Tsai, F. A. Pedersen and J. K. Nørskov, Nano Lett., 2014, 14, 1381–1387.
119. J. H. Kang, W. Liu, D. Sarkar, D. Jena and K. Banerjee, Phys. Rev. X, 2014, 4, 031005.
120. W. S. Li, X. S. Gong, Z. H. Yu, L. Ma, W. J. Sun, S. Gao, Ç. Köroğlu, W. F. Wang, L. Liu, T. T. Li, H. K. Ning, D. X. Fan, Y. F. Xu, X. C. Tu, T. Xu, L. T. Sun, W. H. Wang, J. P. Lu, Z. H. Ni, J. Li, X. D. Duan, P. Wang, Y. F. Nie, H. Qiu and X. R. Wang, Nature, 2023, 613, 274–279.
121. C. Bae, T. A. Ho, H. Kim, S. Lee, S. Lim, M. Kim, H. Yoo, J. M. Montero-Moreno, J. H. Park and H. Shin, Sci. Adv., 2017, 3, e1602215.
122. H. M. Liu, R. K. Xie, Y. T. Luo, Z. C. Cui, Q. M. Yu, Z. Q. Gao, Z. Y. Zhang, F. N. Yang, X. Kang, S. Y. Ge, S. H. Li, X. F. Gao, G. L. Chai, L. Liu and B. L. Liu, Nat. Commun., 2022, 13, 6382.
123. Q. M. Yu, Z. Y. Zhang, S. Y. Qiu, Y. T. Luo, Z. B. Liu, F. N. Yang, H. M. Liu, S. Y. Ge, X. L. Zou, B. F. Ding, W. C. Ren, H. M. Cheng, C. H. Sun and B. L. Liu, Nat. Commun., 2021, 12, 6051.
124. M. Cheng, N. Cao, Z. Wang, K. Wang, T. C. Pu, Y. K. Li, T. L. Sun, X. Y. Yue, W. K. Ni, W. X. Dai, Y. He, Y. Shi, P. Zhang, Y. H. Zhu and P. F. Xie, ACS Nano, 2024, 18, 10582–10595.
125. T. Ma, H. Cao, S. Li, S. J. Cao, Z. Y. Zhao, Z. H. Wu, R. Yan, C. D. Yang, Y. Wang, P. A. van Aken, L. Qiu, Y. G. Wang and C. Cheng, Adv. Mater., 2022, 34, 2206368.
126. L. N. Liu, J. X. Wu, L. Y. Wu, M. Ye, X. Z. Liu, Q. Wang, S. Y. Hou, P. F. Lu, L. F. Sun, J. Y. Zheng, L. Xing, L. Gu, X. W. Jiang, L. M. Xie and L. Y. Jiao, Nat. Mater., 2018, 17, 1108–1114.
127. E. E. Benson, H. Y. Zhang, S. A. Schuman, S. U. Nanayakkara, N. D. Bronstein, S. Ferrere, J. L. Blackburn and E. M. Miller, J. Am. Chem. Soc., 2018, 140, 441–450.
128. I. S. Kwon, I. H. Kwak, T. T. Debela, H. G. Abbas, Y. C. Park, J. P. Ahn, J. Park and H. S. Kang, ACS Nano, 2020, 14, 6295–6304.
129. H. B. Wu, B. Y. Xia, L. Yu, X. Y. Yu and X. W. Lou, Nat. Commun., 2015, 6, 6512.
130. H. J. Liu, Q. He, H. L. Jiang, Y. X. Lin, Y. K. Zhang, M. Habi, S. G. Chen and L. Song, ACS Nano, 2017, 11, 11574–11583.
131. G. F. Chen, T. Y. Ma, Z. Q. Liu, N. Li, Y. Z. Su, K. Davey and S. Z. Qiao, Adv. Funct. Mater., 2016, 26, 3314.
132. Y. Q. Xue, X. J. Bai, Y. Y. Xu, Q. Yan, M. Zhu, K. Zhu, K. Ye, J. Yan, D. X. Cao and G. L. Wang, Composites, Part B, 2021, 224, 109229.
133. C. H. An, W. Kang, Q. B. Deng and N. Hu, Rare Met., 2022, 41, 378–384.

---

Footnote

† These authors contributed equally.

---