The discovery of 2D TIs represents a significant development in condensed matter physics. These materials possess an insulating bulk interior and conducting topological surface states (TESs). In many 2D TIs, strong spin-orbit coupling (SOC) modifies the electronic band structure, leading to band inversion between the valence and conduction bands and the formation of TESs. This inversion transforms the topological invariant of the entire band structure from "trivial" to non-trivial [Figure 1A and B]^[29,30]. The Z₂ invariant (ν) represents the topological invariant, where ν = 1 indicates topological nontriviality (TI), and ν = 0 denotes topological triviality (ordinary insulator, vacuum, and air). When a material with ν = 1 forms an interface with a topologically trivial material with ν = 0, a conductive metallic edge state inevitably emerges at their interface. This state connects two distinct electronic states with different Z₂ invariants^[2,24]. Additionally, spin-momentum locking occurs at the metallic edge state, where the direction of electron motion and its intrinsic spin are forced to be perpendicular [Figure 1C]. Ordinary non-magnetic impurities cannot alter the spin direction of the electrons. As a result, electrons are prevented from bypassing non-magnetic impurities through backscattering, leading to direct migration without backscattering or energy loss, thereby facilitating electron transport. Moreover, the backscatter-free migration is protected by time-reversal symmetry, ensuring consistently high electron mobility^[31].

A recent study^[32] has shown that a novel 2D magnetic higher-order TI, CrOCl, exhibits fully spin-polarized corner states that remain stable in the presence of SOC, symmetry breaking, and even magnetic disruption. Under the influence of an electric field, CrOCl undergoes a phase transition where the corner states merge with the edge or bulk states. Xu et al.^[33] reported a single-layer 2D TI, ZrTe₅, and discovered a new structural phase distinct from its bulk structure. In the study, a topologically nontrivial band gap was observed at approximately 254 meV, along with metallic edge states. Meanwhile, Yalameha et al.^[34] applied first-principles calculations to predict that monolayer K₂Be₂P₂ could transition from a conventional semiconductor to a TI under strain control. Its symmetrical lattice, strain-driven topological phase transition, and TES make it a promising candidate for novel 2D TIs.

TSM phases can arise through two primary mechanisms: band inversion and crystal symmetry protection [Figure 2]. Most DSMs originate from band inversion. The key distinction between 2D DSMs and 2D TIs lies in their bulk electronic structures. In DSMs, band inversion in the bulk causes the conduction and valence bands to directly touch at the Dirac point, resulting in a vanishing bandgap. Near this point, the energy bands exhibit symmetric conical dispersions, forming a Dirac cone [Figure 3A] with fourfold degeneracy.

Near the Dirac point, the electron energy varies linearly with momentum, resulting in massless Dirac fermions with an effective mass approaching zero. These quasiparticles can experience reduced scattering rates, consequently exhibiting exceptionally high carrier mobility. Moreover, Dirac points protected by time-reversal symmetry and spatial-inversion symmetry contribute to the stability of the electronic structure. Based on the degree of band tilting, 2D DSMs can be classified as type I and type II. Type I DSMs exhibit nearly vertical band dispersions near the Dirac point, whereas type II DSMs exhibit strongly tilted bands, leading to pronounced anisotropy in properties, including electrical conductivity and magnetoresistance. These characteristics make type II DSMs particularly promising for developing superconducting materials^[35].

Zhao et al.^[36] theoretically predicted that the 2D antiferromagnetic DSM, r-MnN₄ monolayer, exhibited ideal 2D Dirac properties. At atomic-layer thickness, the monolayer simultaneously exhibited magnetic properties and a Dirac-crossed band structure, making the integration of magnetic and Dirac states possible. A theoretical study^[37] has shown that combining a single-layer BaCu with appropriate distortion and heterostructures can exhibit 2D Weyl and type-III Dirac point properties, indicating that stable, tunable DSMs can be realized in 2D systems through geometric control. Wang et al.^[38] identified the TCH-SSH-2D, a pure carbon 2D allotrope whose calculated band structure exhibits Dirac nodal lines, providing new insights into 2D carbon-based TSMs.

2D WSMs are derived from 2D DSMs. As shown in Figure 3B, when time-reversal symmetry or crystal-symmetry-protected Dirac point is broken, the fourfold-degenerate Dirac point splits into two doubly degenerate Weyl points with opposite chirality. At the Weyl points, the valence and conduction bands meet, resulting in linear dispersion and the emergence of Weyl fermions. Compared with the solitary Dirac points found in the bulk of 2D DSMs, the two Weyl points split from a single Dirac point are discretely distributed within the bulk^[39]. This characteristic enables Fermi arcs to form at the edges of 2D WSMs. The Fermi arcs connect the projections of the discrete pair of Weyl points at the edge of the material. As Fermi arcs provide conductive channels along the edges of 2D WSMs and the bulk states contribute to conduction, these materials exhibit transport properties that combine both edge and bulk effects^[40].

A study^[41] examining the spin-polarized Weyl points and corresponding edge states using spin-angle-resolved photoemission spectroscopy (spin-ARPES) and scanning tunneling spectroscopy (STS) revealed 2D bismuthene monolayers grown on SnS or SnSe substrates. Another study^[42] revealed that applying an appropriate in-plane Zeeman field and modulating the Fermi level with an electric field in Cd₃As₂ films with controlled thickness induced a 2D WSM phase. This approach provides a theoretical pathway for reducing 3D TEMs to 2D. Xu et al.^[43] theoretically proposed the TiTe monolayer as a novel 2D topological phase. This material integrates magnetic and Weyl semimetallic properties, enabling switching between the Weyl semimetallic phase and the quantum Hall phase by controlling the magnetization direction.

Compared with the discrete Dirac or Weyl points in the bulk electronic structures of 2D DSMs and 2D WSMs, respectively, the crossing of the conduction and valence bands in 2D NLSMs does not occur at a discrete point but rather forms a continuous line or loop^[44]. Although electronic states at Dirac and Weyl points are zero-dimensional and localized, those at nodal lines are one-dimensional and extended. This difference can result in anisotropic electron dispersion near the nodal lines, characterized by approximately linear band dispersion perpendicular to the nodal line and parabolic dispersion along the tangential direction, with the latter corresponding to a large effective mass. Moreover, due to the bulk-edge correspondence principle, topologically protected edge states exist along the boundaries. Based on the dispersion characteristics of the nodal lines, they are classified into three types: I, II, and hybrid nodal lines. Type I nodal lines are formed through the intersection of one electron-type band and one hole-type band and are the most common in 2D NLSMs. Type II nodal lines are formed through the intersection of two electron-type bands or two hole-type bands. Hybrid nodal lines simultaneously exhibit both type I and type II characteristics^[45-48].

He et al.^[49] reported the existence of mixed nodal lines in CrN monolayers [Figure 4], where two types of nodal lines coexist. Chen et al.^[50] achieved an ideal type II NLSM in bilayer AlN through a van der Waals interlayer intercalation strategy, in which the nodal lines were stabilized near the Fermi level by regulating interlayer hybridization. A recent study^[51] identified C-Me-graphene as an ideal 2D NLSM. Its nodal lines lie at the Fermi level without interference from extraneous bands, resulting in a well-defined band structure, while also exhibiting excellent mechanical properties.

2D topological materials have attracted considerable attention from researchers owing to their diverse electronic structures. Additionally, the combination of different topological materials or topological materials and other functional materials with unique properties provides new opportunities for discovering and designing novel topological phases. For example, combining materials with topological properties and those with superconductivity results in topological superconductors—novel materials with the zero-resistance characteristics of superconductivity. Examples of 2D topological superconductors include Ag₄H^[52], monolayer h-V₂N₃^[53], and 2D MXC₃ materials (InAsC₃, SeAsC₃, and InTeC₃)^[54]. Similarly, integrating topological materials with magnetic materials results in magnetic TIs, topological axion insulators, and antiferromagnetic TIs. Common examples in this category include the MnBi₂Te₄ family^[55], V₂WS₄^[56], and Eu₃In₂As₄^[57], which exhibit remarkable magnetoelectric effects. In addition to these categories, numerous novel materials with coupled properties exhibit multiple functional properties within a single crystal. These materials provide numerous research opportunities and have considerable potential for technological applications. However, in the context of electrocatalytic water splitting—particularly the HER—the catalytic performance of the coupled-property materials is yet to be fully explored. This presents vast opportunities for developing electrocatalysts with enhanced activity and improved stability.

MBE is an epitaxial growth technique for growing high-quality thin films layer by layer on a substrate by directing molecular beams onto the substrate in a vacuum environment. The main factors influencing MBE include substrate selection, temperature, surface passivation, and beam current ratio. Substrate selection significantly influences film quality, with Si, SiO₂/Si, SrTiO₃, and GaAs substrates being suitable for growing various 2D topological materials^[25]. Liu et al.^[58] successfully fabricated MnBi₂Te₄ films on Si(111) substrates. He et al.^[59] successfully grew monolayer 2H-MoTe₂ films on SiO₂/Si substrates. Figure 5A and B illustrate the schematic apparatus and process control for MoTe₂ preparation through MBE. Ip et al.^[60] fabricated Bi₂Se₃ films on SrTiO₃ substrates, and the process flow is shown in Figure 5C-H. A further study^[61] successfully grew NbAs films on both GaAs(100) and GaAs(111) substrates.

Substrate temperature influences growth kinetics, film quality, and surface morphology. Different substrate temperatures have different effects on film growth. Liu et al.^[62] grew BiSbTe₃ films on Al₂O₃ (0001) substrates and compared the surface morphologies of the films prepared at different temperatures [Figure 6A-C]. The images reveal triangular terraced step surfaces at all temperatures, which are closely related to the symmetry of the trigonal (hexagonal) crystal structure of the substrate. As the temperature increases, the step dimensions progressively enlarge, indicating improved crystal quality. Furthermore, all five terraced layers of BiSbTe₃ exhibit excellent epitaxial relationships with the substrate.

Crystal defects significantly affect thin-film growth quality. Therefore, effectively controlling and eliminating defects are essential for producing high-quality films. Common defects include anti-site atoms, vacancies, and twin domains. Surface passivation of the substrate is a widely used method for suppressing these defects. Wickramasinghe et al.^[63] introduced a novel selenium (Se) passivation technique in which Se passivation of the surface of an InP(111)B substrate forms a virtual In₂Se₃ substrate layer, as shown in Figure 6D. This virtual substrate provides an atomically flat and highly smooth surface. Bi₂Se₃ and Sb₂Te₃ films grown on this substrate completely suppress crystal defects, particularly twin domains.

The beam ratio is another critical factor influencing film quality. Nelson et al.^[64] grew TaAs films on GaAs (001) substrates [Figure 6E] by varying the substrate temperature from 400 to 650 °C while adjusting the As/Ta ratio from 1 to 60. Their results revealed that below 590 °C, both excessively high and low As/Ta ratios adversely affected the phase composition of the film. However, between 640 and 650 °C, the As/Ta ratio minimally impacted the phase composition, demonstrating a close correlation between the beam ratio and substrate temperature.

The CVD technique is used to generate high-quality thin films through chemical reactions between gaseous precursors and substrate surfaces^[65,66]. Depending on the number of precursors involved in the reaction, CVD can be categorized into two methods: single-precursor and multi-precursor methods. Due to the complex multi-component chemical compositions and strict atomic ratios of 2D topological materials, achieving precise elemental ratios with a single precursor is challenging. Consequently, the multi-precursor method is generally used for preparing 2D topological materials.

The selection and ratio of precursors mainly influence film quality. Li et al.^[67] successfully grew few-layer 1T'-WTe₂ films on SiO₂/Si substrates using CVD. In their experiment, a mixed precursor with a molar ratio of Te:WCl₆ ≈ 4.15:1 was used, significantly exceeding the stoichiometric 2:1 ratio of WTe₂. This difference indicates that an excess of Te was used experimentally. As shown in Figure 7A, additional Te powder was placed upstream to maintain a Te-rich atmosphere within the reaction tube, ensuring that the reaction consistently progressed toward WTe₂ formation. The morphological structure of the 1T'-WTe₂ films was further analyzed using AFM and a high-resolution transmission electron microscope (HRTEM) in Figure 7B-D.

Different substrates, substrate temperatures, and growth durations significantly influence the quality of the resulting films. Poplinger et al.^[68] grew Bi₂Se_3-xS_x films on three substrates: Si/SiO₂, sapphire, and Si/SiO₂ covered with a monolayer of graphene. They compared the growth morphologies and properties of these films [Figure 8]. When the films were grown at 570 °C for 10 min on a Si/SiO₂ substrate, the crystals grew vertically and formed isolated structures. When the films were grown on the sapphire substrate at 580 °C for 180 min, crystals grew both laterally and vertically, resulting in aggregation. When the films were separately grown on the Si/SiO₂ substrate covered with a monolayer of graphene at 580 °C for 60, 120, and 180 min, the crystals gradually transformed from isolated hexagonal crystals into continuous films.

Owing to its excellent film coverage, high uniformity, and relatively rapid growth rate, multi-precursor CVD has emerged as an important synthesis method for high-quality 2D topological materials. Other common examples include PtTe₂^[69], Mo_xW_1-xTe₂^[70], Bi_2-xSb_xTe_3-ySe_y^[71], and (Bi_xSb_1-x)₂Te₃^[72]. Furthermore, with improved control over film composition and structure, the CVD technique holds significant potential for future breakthroughs.

The mechanical exfoliation method applies physical force to peel atomic layers from crystalline materials, thereby producing high-quality 2D materials. Common mechanical exfoliation techniques include dry grinding, dry ball milling, and tape exfoliation. Dry grinding and dry ball milling are generally unsuitable for producing high-quality 2D topological materials because the intense mechanical stresses can damage the crystal structure and provide limited control over the exfoliation process, often yielding small-sized and defective nanosheets. In recent years, more efficient tape exfoliation techniques, such as polymer-assisted exfoliation^[73], metal-assisted exfoliation^[74], and oxide-assisted exfoliation have been developed^[75].

A study^[73] prepared graphene using polyethyleneimine-assisted exfoliation, as shown in Figure 9A. The bonding energy between polyethyleneimine and graphene exceeded the interlayer exfoliation energy of graphite, enabling efficient exfoliation of layered graphite into graphene films. As shown in Figure 9B and C, analysis of the morphology and film dimensions revealed that the product exhibits good performance in terms of size, morphology, quality, and yield.

Another study^[76] has revealed a method for preparing monolayer and few-layer MoTe₂ and WTe₂ films by introducing an Au film as an adhesion layer to assist in exfoliation. The morphology and structure of films were further analyzed. As shown in Figure 10A and B, the lighter-colored large areas were identified as monolayers, whereas slightly darker smaller areas were identified as bilayers. The effectiveness of an Au film in assisted exfoliation is attributed to the stronger Au-Te bond, which exceeds the interlayer van der Waals forces, enabling more efficient separation. Additionally, the Au film facilitates selective exfoliation, improving process controllability.

The oxide-assisted exfoliation method is similar to the aforementioned approaches, with the key difference being the use of an oxide layer as the stripping aid. Deng et al.^[77] fabricated Fe₃GeTe₂ films using an Al₂O₃-assisted exfoliation method, as shown in Figure 10C. Similarly, Deng et al.^[78] prepared MnBi₂Te₄ films through the Al₂O₃-assisted exfoliation technique.

Liquid-phase exfoliation involves dispersing layered bulk materials in suitable solvents and applying external energy, typically through ultrasonication, to overcome interlayer interactions and obtain single- or few-layer nanosheets. Surface energy matching is crucial during exfoliation, making solvent selection critical. Common solvents include N-methylpyrrolidone (NMP), isopropyl alcohol (IPA), and ethanol.

Sun et al.^[79] successfully synthesized few-layer Bi₂Se₃ using NMP as the solvent through liquid-phase ultrasonic exfoliation, as shown in Figure 11A. In organic solvents, NMP has appropriate surface tension and plays a crucial role in detaching and stably dispersing bulk Bi₂Se₃ into nanosheets while preventing their reaggregation.

Isopropanol is a suitable solvent for preparing 2D topological materials through the liquid-phase exfoliation method. Goswami et al.^[80] used isopropanol to synthesize few-layer NiTe₂. As shown in Figure 11B-D, the dimensional statistics conducted on the width and thickness of flakes demonstrate the successful exfoliation of 2D NiTe₂ flakes.

Yeon et al.^[81] reported a liquid-phase ultrasonic exfoliation method for synthesizing graphene using sodium dodecyl sulfate (SDS) as a surfactant and a mixed ethanol-methanol solvent as the co-solvent. This approach enabled the production of high-quality graphene under mild and environmentally friendly conditions.

Electrochemical exfoliation applies an external electric field to drive ions into layered bulk materials, weakening the interlayer forces and enabling efficient exfoliation. This method is mainly categorized into anodic and cathodic exfoliation. Anodic exfoliation generally promotes oxidation and introduces oxygen-containing functional groups, whereas cathodic exfoliation may induce defects. Due to the inherent stability of 2D topological materials, they exhibit low sensitivity to defects. However, surface oxidation and the introduction of oxygen-containing functional groups during anodic exfoliation can affect the quality of the exfoliated film, making cathodic exfoliation the preferred method.

WTe₂ films have been successfully prepared through cathodic electrochemical exfoliation using a mixed cation system of diammonium salts (TPA⁺ and TMA⁺)^[82]. As shown in Figure 12A, the mixed cations intercalate between the WTe₂ layers, increasing the interlayer spacing and weakening the van der Waals forces between the layers. Subsequent ultrasonic treatment accelerates the exfoliation of WTe₂ films. Similarly, Ma et al.^[83] successfully prepared bilayer PtSe₂ and PtTe₂ films by intercalating tetraalkylammonium (TAA) cations into the interlayer space using organic electrolytes [Figure 12B and C].

Cathodic electrochemical exfoliation can also be used to prepare other 2D topological materials, including Bi₂Se₃^[84], Bi₂Te₃^[85], and graphene^[86]. Due to its efficient and gentle exfoliation mechanism, this method is widely used for preparing 2D topological materials. It offers excellent controllability and holds significant potential for functionalization and large-scale production.

As shown in Table 1, we summarize and compare the five methods discussed above. Each method has its advantages and limitations, and the most appropriate approach is selected based on the specific requirements of the 2D topological materials being prepared. Although numerous thin-film preparation techniques have been developed, new methods should be developed, and existing techniques should be improved. These advancements will improve product efficiency and quality and reduce costs.

2D TIs have attracted considerable attention in electrocatalysis owing to their distinctive combination of bulk insulating properties and edge-conducting states. In recent years, researchers have focused on developing a range of 2D TIs as electrocatalysts for water splitting^[103,104].

Doping 2D TIs with other 2D materials can enhance the catalytic performance of the resulting catalyst^[105,106]. For pure 2D TIs, active sites are mainly confined to the edges. Doping with additional materials exposes more active sites, thereby increasing catalytic activity. Furthermore, this approach regulates the hydrogen adsorption free energy. Moreover, heterostructures formed by integrating 2D topological materials and other functional materials induce interfacial electronic reconstruction and charge redistribution, resulting in modified electronic structures with unique properties. This reconfigured electronic structure optimizes the adsorption and desorption energies of H^* on the material surface, bringing the Gibbs free energy closer to zero.

Sujita et al.^[105] investigated the enhanced performance of Bi₂Te₃ composites doped with bismuthene in the HER. They prepared Bi₂Te₃/bismuthene composites through liquid-phase exfoliation and solvothermal synthesis. As shown in Figure 13A and B, the microstructures reveal that the layered structure and high surface area of bismuthene facilitate the enhanced adsorption of hydrogen intermediates, while Bi₂Te₃ ensures excellent conductivity. Figure 13C presents the polarization curves of different materials at a scan rate of 5 mV/s. The calculated overpotentials revealed that the composite containing 10% bismuthene exhibited the lowest overpotential of 119 mV vs. SHE. This result confirms that the addition of bismuthene effectively enhances the catalytic active sites in Bi₂Te₃.

The kinetic process of water splitting can be characterized using the Tafel slope. Figure 13D shows the Tafel curves for different materials, along with the corresponding Tafel slopes. The curves reveal that the composite material containing 10% bismuthene exhibits the lowest Tafel slope of 95.4 mV/dec. As shown in Figure 13E, the 10% bismuthene composite exhibits the lowest charge transfer resistance (R_ct)^[105]. Doping bismuthene into Bi₂Te₃ increases active sites at the material edges while optimizing the adsorption and desorption of H* at the surface. The synergistic effect of these two factors improves the catalytic performance.

2D TIs can also serve as additives or substrates for other electrocatalysts. For example, Yang et al.^[106] reported that Bi₂Se₃, acting as a conductive substrate, exhibited exceptional electronic dynamics and stability by uniformly growing MoSe₂ nanosheets onto monolayer Bi₂Se₃ hexagonal flakes^[107,108]. They further analyzed the morphology and electrochemical performance of the composite material. The results showed that Bi₂Se₃ doping significantly enhanced the conductivity and charge transfer capacity of the material, promoting synergistic interactions between the two components. These interactions increased the catalytic activity of the relatively inactive MoSe₂ surface, improving the adsorption and desorption of H^* and water-splitting efficiency. Furthermore, MoSe₂ grew vertically in an ultrathin and small-sized form on the Bi₂Se₃ substrate, exposing more active sites on the MoSe₂ surface. The topological stability of Bi₂Se₃ prevents the collapse of the structure, thereby significantly extending its lifetime.

The development of novel 2D TIs for electrocatalytic water splitting has received considerable attention. 2D metal-organic frameworks (MOFs) have emerged as a new class of hybrid organic materials. Studies have shown that 2D MOFs exhibit characteristics similar to TIs, including strong spin coupling, TESs, and a substantial topological energy gap^[109-111]. Owing to their high porosity, tunable metallic nodes, and multifunctional organic ligands, 2D MOFs have gained widespread application in electrocatalytic water splitting. Their catalytic performance can be further enhanced through various modification strategies, including 2D MOF nanosheet fabrication^[112,113], defect engineering^[114,115], heteroatom doping^[116,117], and ligand engineering^[118,119].

Sattigeri et al.^[120] converted 2D LiMgAs from an ordinary insulator into a TI through biaxial compressive stress. As shown in Figure 14A and B, they constructed structures with zig-zag and planar edges to investigate their edge states and catalytic properties. The most active sites on the zig-zag edge were identified as the atomic As top site and Li-As bridge site, while the active sites on the planar edge were identified as the Li top site and As top site. The Gibbs free energies of these sites under SOC were analyzed. The results revealed ΔG_H* values of -0.024 and -0.02 eV for the zig-zag As and Li-As sites, respectively, which are very close to the ideal value of 0 eV. Similarly, the planar Li and As sites exhibited ΔG_H* values of 0.026 and 0.027 eV, respectively, both approaching the ideal value. These findings reveal that the edge sites of 2D LiMgAs are situated near the optimal region [Figure 14C and D], indicating that 2D LiMgAs exhibits promising H^* adsorption and desorption capabilities and hydrogen evolution performance.

Wang et al.^[121] reported a novel 2D organometallic material, Ni₃(CNH)₁₂, which exhibited both TI properties and highly efficient bifunctional electrocatalytic performance. This material features four potential adsorption sites at its edges: Ni-top, N-top, C-top, and hole-top. As shown in Figure 15A-C, the C-top site is optimal for OER with an overpotential of 0.95 V, whereas the N-top site is optimal for HER with an overpotential of 0.31 V. Furthermore, the effect of hydrogen coverage on hydrogen evolution efficiency was investigated. By varying hydrogen coverage on the N-top sites within the range of α = 1/12 to α = 1, it was found that the hydrogen adsorption free energy reached 0.1 eV at α = 1/3 [Figure 15D], indicating performance comparable to Pt catalysts.

Although 2D TIs have demonstrated considerable potential for electrocatalytic water splitting, current studies mainly focus on improving catalytic performance through structural engineering rather than developing intrinsic catalytic systems. Studies have shown that heterostructure construction and exfoliation methods can effectively increase exposed active sites and promote charge transfer, as demonstrated in Bi₂Te₃-based and Bi₂Se₃-based materials^[105,106]. However, the catalytic performance of 2D TIs remains highly dependent on surface modification and interface regulation. Future research should focus on precise control of surface structures, optimization of interfacial interactions, and development of scalable preparation methods to further improve their practical applicability.

In electrocatalytic water splitting, catalytically active sites are crucial because they strongly influence the adsorption of water molecules and the formation and binding of H^* intermediates. 2D TIs, with their conductive edges, exclusively confine active sites to these edges. Although incorporating other materials^[122,123] and optimizing edge structures^[120,124] to expose additional sites have been explored, these methods still significantly limit catalytic activity. In contrast, 2D TSMs exhibit excellent conductivity both in the bulk and along the edges, thus maximizing the availability of active sites^[125]. Additionally, the massless, dissipation-free fermions in 2D TSMs provide exceptional mobility, thereby attracting significant attention to 2D TSMs.

2D DSMs, as prototypical examples of 2D TSMs, have been extensively used in electrocatalytic water splitting. In recent years, researchers have enhanced catalytic performance by optimizing the electronic structure of these materials^[126,127] and developing novel material compositions^[128,129].

Wang et al.^[130] proposed a 2D WB₄ lattice material exhibiting multiple Dirac cones near the Fermi level [Figure 16A and B]. These Dirac cones were generated from d-p-π and d-p-σ hybridization between the d orbitals of tungsten (W) and the p orbitals of boron (B). Electrons near the Dirac cones exhibited high Fermi velocities (0.72 × 10⁶ m/s). Experiments revealed the B-B bridging site as the most stable hydrogen adsorption site [Figure 16C]. Additionally, the study revealed that stretching the material increased the number of active sites. The catalytic activity of WB₄ is closely related to the position of the d-band center of W. Consequently, the tensile strain modulated the d-band center by altering the orbital coupling between B and W, shifting the d-band center downward, and weakening the adsorption strength between the active site and H^*. As shown in Figure 16D, at 3% tensile strain, the Gibbs free energy approached the ideal value (0 eV). The effect of hydrogen coverage on Gibbs free energy was also investigated [Figure 16E]. The results showed that the optimal Gibbs free energy (0.02 eV) occurred at a coverage of 1/3.

The growth mode of materials can significantly influence the exposure of catalytic active sites. Catalysts grown on planar substrates mainly confine most active sites to their edges. In contrast, catalysts grown in a 3D mode can provide numerous exposed active sites^[131,132]. Fu et al.^[133] reported the 3D growth of PtTe₂ nanosheets on conductive carbon cloth through atmospheric pressure CVD. The PtTe₂ nanosheets increased the surface area, improved active site utilization, optimized the electron transport pathways, and significantly enhanced the electron mobility. In addition, they revealed the thickness-dependent effects on catalytic activity. By adjusting the loading of the PtCl₂ precursor, the thickness of PtTe₂ nanosheets was controllably varied from 3 nm to bulk-like (> 12 nm) [Figure 17A-C]. When the thickness of PtTe₂ nanosheets increased from 3.0 to 7.5 nm, the overpotential decreased from 65.6 to 38.8 mV, and the Tafel slope decreased from 95.8 to 59.2 mV/dec [Figure 17D and E]. Beyond a thickness of 7.5 nm, further increases in thickness slightly affected catalytic activity, indicating a slight increase in active sites.

The pH value influences catalytic performance, and reaction processes vary under acidic, alkaline, and neutral conditions^[134,135]. The adsorption and desorption behaviors of catalysts toward different ions or molecules vary with the solution pH due to changes in the surface charge state and interfacial interactions. Therefore, achieving efficient catalysis across the entire pH range is challenging. Shi et al.^[134] synthesized 1T-NiTe₂ through CVD and evaluated its catalytic performance in the HER. The results showed that 1T-NiTe₂ exhibited high activity under both acidic and alkaline conditions. Under acidic conditions (0.5 M H₂SO₄), the catalyst achieved a minimum overpotential of 353 mV and Tafel slopes of approximately 33-41 mV/dec, approaching those of commercial Pt-on-carbon catalysts (31 mV/dec). The charge transfer resistance was also low, ranging from 12 to 16 Ω. Under alkaline conditions (1 M KOH), the overpotential remained low (353-390 mV), and the Tafel slope (62-82 mV/dec) remained modest, with the charge transfer resistance at an extremely low level of 15-22 Ω.

The effects of synthesis methods on the electrocatalytic properties of 2D DSMs have been investigated in PtTe₂ and NiTe₂ systems. For PtTe₂, CVD growth enables the formation of high-quality thin films with well-defined crystalline structures and excellent electrical conductivity, facilitating efficient charge transport and providing a stable platform for electrocatalytic water splitting^[133]. In contrast, electrochemical exfoliation can produce ultrathin PtTe₂ layers with reduced thickness and increased surface accessibility, thus exposing more active sites for catalytic reactions^[83]. Similarly, CVD-prepared NiTe₂ generally exhibits uniform morphology, high crystallinity, and strong electrical connectivity, contributing to enhanced electron transfer and structural stability during electrochemical operation^[134]. Liquid-phase exfoliation generates thinner NiTe₂ nanosheets with larger surface areas and abundant exposed edge sites than CVD-grown NiTe₂. These larger surface areas and abundant exposed edge sites enhance electrolyte accessibility and catalytic active site utilization^[80]. These examples demonstrate that the selection of synthesis strategies, including CVD growth and different exfoliation approaches, plays a pivotal role in regulating the structural characteristics, surface properties, and electrochemical behavior of PtTe₂ and NiTe₂, thereby influencing their performance in water splitting applications.

In a separate study, Wang et al.^[135] successfully synthesized high-purity MoP₂ nanosheets and evaluated their HER catalytic performance under neutral and acidic conditions. As shown in Figure 18A-D, under neutral conditions (1.0 M phosphate buffer solution), the onset overpotential and Tafel slope were 193 mV and 82.2 mV/dec, respectively. Under acidic conditions (0.5 M H₂SO₄), the onset overpotential and Tafel slope were 170 mV and 60 mV/dec, respectively. They also conducted stability tests, and the results showed that MoP₂ performed better than Pt [Figure 18E and F].

The studies on 2D DSMs reveal that these materials possess considerable advantages in electrocatalysis owing to their high electrical conductivity and flexible electronic structures. Various strategies, including phase engineering, defect regulation, and composition optimization, have been applied to improve catalytic activity, as observed in PtTe₂, NiTe₂, and MoP₂ systems. However, current studies mainly focus on HER performance, while their application in overall water splitting and OER remains relatively limited. Future studies should expand the application range of DSMs by exploring multifunctional catalysts and rationally designing electrode architectures for efficient and durable water electrolysis.

The edge states of 2D WSMs feature Fermi arcs that connect Weyl points in momentum space. These edge Fermi arcs expose additional active sites at the boundary. Moreover, the unique nature of edge Fermi arcs provides enhanced tunability of electronic structures and improved flexibility in material design^[136,137].

Recent studies have shown that magnetic interactions significantly influence the HER rates^[138,139]. Applying an external magnetic field can accelerate reaction kinetics by regulating charge transfer and spin-related effects. However, its practical implementation for improving electrocatalytic performance remains limited. Notably, when non-trivial band topology coexists with 2D magnetic materials, the resulting 2D magnetic topological materials exhibit unique properties^[140,141]. The coexistence of intrinsic magnetism and non-trivial topological energy structures makes the resulting materials highly suitable for electrocatalysis applications.

Liu et al.^[142] reported monolayer Fe₂Sn as a 2D ferromagnetic Weyl semimetal and systematically investigated the effect of spin polarization on HER activity through density functional theory (DFT) calculations. As shown in Figure 19A, the Fe₂Sn monolayer with spin polarization exhibited a ΔG_H* of -0.06 eV at the Fe-Fe bridge site, while the layer without spin polarization exhibited a ΔG_H* of -1.27 eV. This difference indicates excessively strong hydrogen adsorption and poor activity in the absence of spin polarization. Spin polarization enhances the electron transfer into the antibonding orbital of the hydrogen atom, substantially filling the antibonding state. This effect weakens the Fe-H chemical bond, thereby moderating hydrogen adsorption strength and significantly improving HER activity [Figure 19B].

Wang et al.^[143] investigated the effects of different magnetization directions on the topological states and HER catalytic performance of single-layer TiTe. As shown in Figure 19C-E, without SOC, the single-layer TiTe material exhibits Weyl semimetallic behavior. When SOC is considered, the material remains a Weyl semimetal with magnetization along the x-axis but transforms into a semi-Chern insulator with magnetization along the z-axis. They further investigated the effect of strain on the HER activity for the TiTe monolayer [Figure 19F]. Under -3% to 3% biaxial strain, ∆G_H∗ varied almost linearly, indicating possible manipulation of HER activity through strain engineering. Moreover, the magnetization direction along the x or z axis slightly affects ∆G_H∗, indicating the excellent HER performance of TiTe regardless of magnetization direction, although it can induce a topological phase transition in TiTe between semimetal and insulator states.

WTe₂ is an important type II Weyl semimetal, but conventional preparation methods face challenges in scaling the growth of 2D WTe₂ films. Li et al.^[144] developed a combined hydrothermal and CVD method to prepare highly crystalline 1T'-WTe₂ nanoribbons. Additionally, they investigated the HER catalytic performance of the prepared nanoribbons. As shown in Figure 20A and B, 1T'-WTe₂ exhibits an overpotential of 430 mV and a Tafel slope of 57 mV/dec. As depicted in Figure 20C and D, the polarization curves of 1T'-WTe₂ nanoribbons showed slight decay after 5,000 cyclic voltammetry cycles and a 20-h constant-current test, indicating exceptional stability.

Different synthesis strategies can induce distinct structural features and surface properties, thereby influencing the catalytic performance of WTe₂. For example, CVD-prepared WTe₂ films^[67] generally exhibit high crystallinity, uniform thickness, and strong adhesion to substrates, thereby enhancing charge transfer and long-term structural stability during electrochemical processes. In contrast, WTe₂ nanosheets prepared through electrochemical exfoliation^[82] exhibit reduced thickness, larger specific surface area, and abundant exposed edge sites, providing more accessible catalytic sites for water-splitting reactions. However, their structural stability may be relatively limited. Furthermore, combining hydrothermal-assisted synthesis with CVD enables the fabrication of multilayer WTe₂ with optimized morphology and enhanced electrode compatibility^[144]. These results show that rational selection of synthesis methods is crucial for balancing active site exposure, charge-transfer capability, and operational stability, which ultimately determines the applicability of WTe₂ in practical electrocatalytic water-splitting applications.

MoTe₂ is a transition-metal dichalcogenide material with strongly phase-dependent electronic properties. 2H-MoTe₂^[145,146] exhibits semiconductor behavior, 1T′-MoTe₂^[147,148] exhibits Weyl semimetallic properties, and T_d-MoTe₂^[149,150] exhibits superconductivity. In a recent study^[151], various MoTe₂ phases, including F-1T′-MoTe₂, F-1T′/2H-MoTe₂, P-1T′-MoTe₂, SG-1T′-MoTe₂, and LG-1T′-MoTe₂, were grown on carbon cloth through CVD, and their electrocatalytic performance was compared [Figure 21A]. The results showed that 1T′-MoTe₂ exhibited the lowest overpotential, the lowest Tafel slope, and the highest stability. As shown in Figure 21B and C, DFT calculations show that the hydrogen adsorption free energy of Mo atomic sites at the edges of the 1T′ phase is closer to zero than that in the 2H phase. Additionally, as illustrated in Figure 21D-F, the strain and defects increase the number of active sites, thereby improving the catalytic performance.

Compared with 2D TIs and 2D DSMs, the electrocatalytic research on 2D WSMs remains in an early stage. Existing studies mainly focus on representative materials, such as 1T′-WTe₂ and 1T′-MoTe₂^[144,151], in which high carrier mobility and structural flexibility improve catalytic performance. However, the limited availability of experimentally realizable Weyl semimetal materials and the challenge of achieving high-quality thin-film growth restrict their further development. Future research should prioritize the exploration of new experimentally realizable Weyl systems, optimization of synthesis processes, and integration of WSMs into practical electrode configurations.

Electronic states near Dirac or Weyl points can exhibit approximately isotropic dispersion. In contrast, the closed nodal line structure exhibits anisotropic electron properties near the nodal lines, which are influenced by direction. This anisotropy may cause different crystal orientations or positions at the material edges to exhibit distinct catalytic activities, thereby enhancing catalytic performance^[152,153]. Furthermore, this unique nodal line structure can provide the material with greater tunability^[154,155].

Studies have shown a correlation between the position of the nodal line relative to the Fermi level (E_F) and catalytic activity. Wang et al.^[156] conducted DFT calculations to investigate the electrocatalytic properties of the 2D NLSM Cu₂Si monolayer. The results showed that the ΔG_H* value of the Cu₂Si monolayer was 0.195 eV, and the difference between E_F and the HER standard potential was approximately 0.33 eV [Figure 22A and B]. Additionally, by modulating the energy position of the nodal line through electron/hole doping [Figure 22C-G], a nodal line closer to the Fermi level corresponding to a lower ΔG_H* value was observed. The study confirms that the position of the nodal line relative to the Fermi level mainly influences HER catalytic activity. One possible mechanism is that a topologically protected nodal-line electronic state near the Fermi level modifies the electronic structure, thereby improving both energy-level alignment and hydrogen adsorption strength^[157,158].

In another study, Gao et al.^[159] investigated the theoretical performance of PtB₂ monolayers as HER catalysts. The results showed that ΔG_H* at the B site was as low as 0.057 eV in the unstrained pristine state [Figure 23A and B]. After applying strain, ΔG_H* reached an optimal value of -0.003 eV under -1% strain. Additionally, the study examined the catalytic performance of bilayer PtB₂, revealing that both AA and AB stacking configurations further modulated catalytic activity. Under AA stacking, the nodal line position shifted above the Fermi level, altering carrier properties and DOS at the Fermi surface, thereby enhancing catalytic activity [Figure 23C]. In the AB stacking configuration, the topological properties of the material were completely disrupted, but the activity of its surface Pt sites became active while maintaining B-site activity [Figure 23D]. However, the loss of topological properties caused the material to lose its ultra-high carrier mobility and exceptional stability.

Nodal lines are confined within the 2D Brillouin zone, forming closed-loop structures. However, in certain materials, nodal lines extend along the Brillouin zone boundary, traversing the zone to form open loops^[160,161]. Wang et al.^[162] proposed a strategy for designing 2D materials with open nodal lines and identified a Cu₂C₂N₄ monolayer that satisfied the design requirements. This strategy indicates that open nodal-line states extending across the Brillouin zone can generate extended Fermi arc states at the material boundaries, which may facilitate charge transfer and enhance HER catalytic performance. The Cu₂C₂N₄ monolayer was confirmed to exhibit an open nodal line along the S-Y path of the Brillouin zone, and the theoretical location of its long Fermi arc along the edge is shown in Figure 24A and B. Additionally, the ΔG_H* for the HER was 0.01 eV, indicating its high catalytic performance.

Although the topological properties of most topological materials are mainly determined by s or p orbitals, d orbital electrons typically play a central role in catalytic reactions. Consequently, topological materials whose properties are governed by d orbitals may exhibit superior catalytic performance. Kong et al. designed a 2D topological NLSM SrPd/BaPd monolayer material with topological properties derived from Pd’s d orbitals^[163]. Calculations revealed that the d-origin TESs were partially unoccupied. These unoccupied d-orbital edge states act as catalytic active sites. As shown in Figure 24C and D, the results show extremely low hydrogen adsorption free energies at the edge Pd sites, indicating exceptional catalytic performance.

2D NLSMs represent an emerging class of 2D topological catalysts, and recent studies have highlighted their potential through unique structural designs, such as open nodal lines and d-orbital-derived catalytic sites. For example, studies on Cu₂C₂N₄, SrPd, and BaPd monolayers have shown that rational regulation of electronic structures provides favorable reaction sites and optimizes adsorption behavior^[162,163]. However, compared with other 2D topological materials, experimental verification and practical catalytic studies of 2D NLSMs remain scarce. Future studies should focus on expanding experimentally accessible nodal-line materials, developing reliable synthesis strategies, and establishing systematic relationships between structural characteristics and catalytic performance.

Although different classes of 2D topological materials have shown considerable potential for electrocatalytic water splitting, their performance is closely related to their intrinsic electronic structures and material characteristics. Therefore, no single class exhibits universally superior electrocatalytic performance; instead, different classes offer distinct advantages depending on the specific reaction mechanism and catalytic requirements. Materials for 2D TIs, such as Bi₂Se₃ and Bi₂Te₃, have attracted significant attention owing to their excellent surface states and tunable surface structures. For example, studies have shown that exfoliation, interface construction, and surface regulation enhance the catalytic activity of Bi₂Se₃ nanosheets and Bi₂Te₃-based heterostructures, making them promising for designing functional catalytic interfaces^[105,106].

Representative 2D DSMs, including PtTe₂, NiTe₂, and MoP₂^[133-135] exhibit high carrier mobility, excellent electrical conductivity, and flexible electronic structures, accelerating charge transfer during electrocatalytic reactions. Notably, PtTe₂ and NiTe₂-based systems exhibit excellent HER performance, while MoP₂ can regulate electronic structure and provide abundant active sites. These features make 2D DSMs particularly promising for constructing highly conductive electrocatalytic platforms.

2D WSMs, such as 1T′-WTe₂ and 1T′-MoTe₂, are characterized by their highly anisotropic electronic structures and excellent charge-transport capability. The successful growth of few-layer WTe₂ and MoTe₂ films through advanced synthesis methods has facilitated their exploration as efficient HER catalysts and flexible catalytic electrodes. However, compared with other material classes, the development of 2D WSMs remains limited by challenges in achieving large-scale synthesis and maintaining structural stability under electrochemical conditions.

2D NLSMs represent another promising class of topological materials with unique electronic structures and highly tunable properties. For example, theoretical studies on Cu₂C₂N₄ and SrPd/BaPd monolayers have shown that rational structural design and regulation of electronic states can provide favorable catalytic sites and optimize reaction intermediates^[162,163]. Although experimental studies on 2D NLSMs remain relatively limited, the structural diversity of the semimetals provides new opportunities for developing next-generation electrocatalysts.

Overall, 2D TIs, 2D DSMs, 2D WSMs, and 2D NLSMs each offer distinct advantages for electrocatalytic water splitting owing to their characteristic topological, electronic, and structural properties. Therefore, future studies should focus on selecting appropriate material systems according to specific catalytic requirements and integrating their intrinsic advantages with advanced structural engineering strategies to achieve efficient and durable water electrolysis. To comprehensively compare different 2D topological materials for electrocatalytic water splitting, Table 2 summarizes their synthesis methods, representative catalytic performances, advantages, and limitations.