Recent development and challenges in TMD-based 2D materials towards OER/ORR electrocatalysis

Abstract

The search for efficient electrocatalysts to drive the oxygen evolution reaction (OER)/oxygen reduction reaction (ORR) has reached a pivotal juncture with the emergence of transition metal dichalcogenides (TMDs), particularly WS₂, WTe₂ and MoTe₂. These materials, with their unique electronic structures, tunable surface properties, and exceptional stability, have opened new frontiers in electrocatalysis. This review provides a comprehensive exploration of the synergistic interplay between experimental validation and computational modeling in unraveling the electrocatalytic potential of these TMD materials. Advanced experimental techniques, such as in situ spectroscopy and electrochemical microscopy, have unveiled the dynamic structural transformations and active site engineering under operational conditions. Currently, state of the art computational approaches, including density functional theory (DFT) and machine learning (ML)-guided descriptor analysis, have enabled the rational design of TMD-based catalysts by predicting reaction pathways, overpotentials, and selectivity. This review presents a novel integrated approach combining experimental techniques and computational modeling to explore the electrocatalytic potential of TMDs for the OER and ORR. By focusing on defect engineering, heterostructures, and phase transitions, this work provides a comprehensive roadmap for the development of next-generation electrocatalysts for sustainable energy application.

---

Kashif Nawaz Khattak

Kashif Nawaz Khattak is currently a Ph.D. candidate at Harbin Institute of Technology, Shenzhen. He received his master's degree from China West Normal University in June 2024. In September 2024 he joined Prof Jia Zhou's research group to pursue his doctoral studies. His research mainly focuses on computational design and modeling to unravel the catalytic mechanisms of two-dimensional materials and with theoretical simulation insight, he aims to accelerate the discovery of efficient, durable, and sustainable catalysts for clean energy technologies.

Yueyue Shao

Yueyue Shao is currently a Ph.D. candidate at Harbin Institute of Technology, Shenzhen. She received her M.S. degree from Shanghai University in June 2018. In March 2022, she joined Prof. Jia Zhou's research group to pursue her doctoral studies. Her research mainly focuses on exploring the electrocatalytic reduction of carbon dioxide on two-dimensional materials by combining first-principles calculations with machine learning approaches.

Jia Zhou

Dr. Jia Zhou received his B.Eng. degree in 2003 from Shanghai Jiao Tong University and M.Sc. degree in 2006 from Fudan University. In 2011, he received his Ph.D. degree from Wayne State University under the supervision of Prof. H. B. Schlegel. Afterward he worked as a postdoc at Texas A&M University. In 2012, he moved to Oak Ridge National Laboratory. Since 2014, he has worked at Harbin Institute of Technology as an associate professor. His research interests mainly focus on materials simulation and computational chemistry. He has published over 100 peer-reviewed papers in journals including Nat. Commun., J. Am. Chem. Soc., Angew. Chem., Int. Ed., etc.

---

1 Introduction

The global transition to sustainable energy systems prompted the development of efficient electrocatalysts for critical reactions such as the OER/ORR^1–4 which are central to technologies such as water splitting, fuel cells, and metal–air batteries.^5–7 Despite significant progress, the reliance on expensive and scarce noble metal-based catalysts has hindered widespread adoption^8,9 necessitating the exploration of alternative materials with high activity, stability, and scalability.^10,11 In particular, TMDs are a class of layered materials that have recently attracted the attention of the electrocatalysis community.^12–14 Among these, WS₂, WTe₂ and MoTe₂ stand out due to their unique electronic structures, adjustable surface chemistries, and exceptional mechanical and thermal stability.^15–18 These TMD materials offer a rich playground for engineering active sites in which both metal sites (e.g., Mo, W) and anion sites (e.g., sulfur, tellurium) serve as active sites for catalysis, but they play different roles; metal sites are the primary active sites in reactions like the OER and ORR, where they participate in electron transfer and oxidation–reduction processes. Anion sites, especially those with defects or vacancies, can also contribute by adsorbing reactants and stabilizing intermediates during the reaction, optimizing reaction pathways, and enhancing catalytic performance through defect engineering, heterostructure, and doping.^19–22 However, realizing their full potential requires a holistic understanding of their atomic-scale behaviour under operational conditions, a challenge that can only be addressed through the synergistic integration of experimental and computational approaches.^23–25 On the experimental front, advanced techniques such as in situ X-ray absorption spectroscopy (XAS), scanning tunnelling microscopy (STM), and electrochemical impedance spectroscopy (EIS) have provided unprecedented understanding into the dynamic structural transformations, surface reconstructions, and reaction intermediates of WTe₂, WS₂ and MoTe₂ during the OER/ORR.^26–29 Complementing these efforts, computational modelling ranging from DFT to ML guided descriptor analysis has enabled the prediction of catalytic activity, overpotentials, and selectivity, while unravelling the fundamental mechanisms governing these processes.^30–33 Together, these approaches have not only validated the exceptional promise of TMD materials but also guided the rational design of next-generation electrocatalysts.^20,34–36 This review delves into the transformative role of WTe₂, WS₂ and MoTe₂ in revolutionizing the OER/ORR, highlighting the critical interplay between experimental validation and computational modelling.^37–40 By examining the latest evolution in defect engineering, phase transitions, and interfacial effects, we aim to provide a comprehensive roadmap for harnessing the potential of TMD materials.^41–46 Additionally, transition metal dichalcogenides (TMDs), such as WS₂, WTe₂, and MoTe₂, have gained significant attention in the field of electrocatalysis, particularly for the oxygen evolution reaction (OER) and oxygen reduction reaction (ORR), due to their unique electronic structures and tunable surface properties. Compared to traditional metal oxides, TMDs offer several distinct advantages, including high conductivity, layered structures, and the ability to engineer active sites through defect manipulation and doping. These properties enhance the catalytic efficiency, stability, and scalability of TMDs, making them superior candidates for sustainable energy applications. Their monolayer form further increases the number of active sites for reactions, enabling better performance than conventional oxides in both catalytic activity and long-term durability.^47–49 We also address the challenges of scalability, durability, and practical implementation, offering a forward-looking perspective on how these materials can drive the development of sustainable energy technologies. Although previous studies have examined the electrocatalytic properties of TMDs in isolation, this review uniquely integrates cutting-edge experimental validation with computational predictions, offering a more comprehensive understanding of their catalytic potential (Fig. 1).

Fig. 1 Timescale for the recent development of TMD 2D materials.^50–55

2 Transition metal dichalcogenides (TMDs)

2.1 Structure and properties

TMDs such as WTe₂, WS₂ and MoTe₂ are layered materials with distinct crystallographic and electronic properties governed by their unique atomic arrangements.⁵⁶ WS₂ has a layer crystal structure as shown in Fig. 2(a), and adopts a hexagonal packed atom structure as shown in Fig. 2(b), where the a plane of transition metal atom (W) is covalently sandwiched between two chalcogen layers (S), forming an S–W–S trilayer.^57,58 The c-axis, perpendicular to the basal plane, defines the vertical stacking of these monolayers via weak van der Waals interactions. For WS₂, the c-axis lattice parameter is ∼6.18 Å, typical of semiconducting TMDs; WS₂ also has two other coordination modes which are trigonal prismatic and octahedral as shown in Fig. 2(c) and (d). Meanwhile WTe₂ and MoTe₂ exhibit larger interlayer spacings (∼7–7.5 Å) due to the bulkier tellurium atoms and structural distortions.^59,60 The monolayer thicknesses for all three materials range between 4 and 7 Å. The top view, side view and four layers are shown in Fig. 2(e) and (f), corresponding to three atomic layers (chalcogen–metal–chalcogen).^61,62 WS₂ is a prototypical semiconducting TMD, crystallizing in the 2H phase with trigonal prismatic coordination (D_3h symmetry).^63,64 Its monolayer structure retains hexagonal symmetry (space group P6₃/mmc) and exhibits a direct bandgap (∼2.0 eV), enabling strong photoluminescence and applications in optoelectronics. In contrast, WTe₂ adopts a distorted 1T′ structure (orthorhombic symmetry, space group Pmn2₁) due to Peierls distortion, where tungsten atoms form zigzag chains.^65–67 This distortion reduces symmetry and induces metallic behaviour, though monolayer WTe₂ can display semiconducting properties depending on strain or substrate effects. MoTe₂ exhibits polymorphism, existing in the semiconducting 2H, metallic 1T′, or Weyl semi metallic (T_d) phase.^68–70 The 2H-MoTe₂ monolayer has a direct bandgap (∼1.1 eV), while the 1T′ phase features a distorted octahedral coordination with lowered symmetry, enabling unique electronic states like type-II Dirac fermions in its T_d phase under strain. The symmetry of these materials critically influences their electronic and optical responses.^71–74 In WS₂ monolayers, broken inversion symmetry allows valley-selective circular dichroism, a key feature for valleytronics. Fig. 2(g) shows the PL intensity of monolayer WS₂ with and without right-handed plasmonic Archimedes spiral (RHPAS) nanostructures. Under right-handed circularly polarized light (RCP, red), the PL intensity of WS₂–2TPAS was enhanced over ten times compared to the as-grown WS₂ (green). Under left-handed circularly polarized light (LCP, blue), the enhancement was minimal. This strong polarization contrast indicates exciton–plasmon coupling, with the enhancement factor (EF) defined as EF = Iwi/Iwo, where Iwi and Iwo are the intensities with and without the Archimedes spiral, respectively. The EF for RCP emission is greater than 10, while for LCP, it is nearly 1. This shows that coupling WS₂ with PAS nanostructures efficiently tailors exciton–plasmon interaction by controlling light polarization. The inset shows an SEM image of the WS₂–2TPAS array (scale bar: 5 μm) and the PL intensity of WS₂–2TRHPAS increased significantly as the laser power rose from 1.9 to 16.2 μW. The PL peak shifted slightly to blue up to 5.5 μW, after which it remained steady from 5.5 to 16.2 μW. Fig. 2(h) shows a log–log plot of PL intensity vs. laser power for WS₂ with (red, triangle) and without (blue, inverted triangle) plasmonic coupling. The relationship between PL intensity (IPL) and laser power (Plaser) follows the equation IPL = (Plaser)^m, where m ≈ 1.37. The maximum enhancement factor (ε) was greater than 10 at 9.1 μW.^75–78 MoTe₂ phase-dependent symmetry (e.g., hexagonal in 2H vs. monoclinic in 1T′) governs its topological properties, such as protected surface states in the T_d phase. Similar results in both directions were observed by optical calculation of [1 00] and [0 10] for the WTe₂ monolayer. The real and imaginary components of the complex dielectric function of the WTe₂ monolayer show the calculated values of the static refraction index n(X), as in the [0 10] direction shown in Fig. 2(i), indicating different number of layers and different values of 1.90, 2.40 and 2.81 for the monolayer, bilayer and tri-layer respectively. The highest quad layer of all four is 3.18. There is also a change in the extinction coefficient k(x) due to photon energy which clearly indicates that the maximum extinction coefficient occurs at 4.80 eV, and the main hints for all layers can take about 2.13 eV, 7.50 eV and 10.77 eV. Among all TMDs, the energy carrier between 2H and 1T′-MoTe₂ is very small as shown in Fig. 2(j),⁷⁹ and the modules in these phases are also very close;⁸⁰ therefore it is easy to easily change the phases of MoTe₂ from 2H to 1T′ and vice versa. This makes MoTe₂ the perfect platform for building homojunctions with the 2D hetero-phase. Interestingly, the 1T′ and T_d phases of MoTe₂ have attractive properties such as topological insulation, superconductor, and type II Wyle semimetal (WSM) properties.^81–83Fig. 2(k) shows the Raman spectra of the three phases. For 1T′ and T_d phases, the Raman spectra are almost identical because of their similar crystal structures except for the Ag³ peak at 128–134 cm⁻¹, which splits into two peaks, namely Ag³ and Ag^3′, across the transition from 1T′ to T_d phase. The latter peak originates from the out-of-plane vibration mode and cannot be detected in the 1T′ phase because of the existence of inversion symmetry.^84,85

Fig. 2 (a–d) Schematic of WS₂ crystal structure showing the layered arrangement, c-axis view, and unit cells with trigonal prismatic and octahedral coordination.⁸⁶ Copyright 2021, Wiley. (e) WTe₂ monolayer and four-layer top and side views: the inter-layer distances are marked as L′s.⁸⁷ Copyright 2023, Elsevier. (f) The crystal structure of single-layer molybdenum ditelluride (MoTe₂).⁸⁸ (g) PL spectra of WS₂ with/without Au spirals, excited by CPL at 514 nm, 8.8 μW power. Inset: SEM of the WS₂–2TRHPAS array (scale bar: 5 μm). (h) Power-dependent PL intensity with (red) and without (blue) Au spirals, showing Purcell enhancement (dark red).⁸⁹ Copyright 2022, Wiley. (i) WTe₂ monolayer real part of dielectric function, imaginary part of dielectric function, refractive index, extinction coefficient for 1 to 4 layers.⁸⁷ Copyright 2023, Elsevier. (j) Total energies per unit cell of 1T-, 1T′-, and 2H-phase MoTe₂ across varying levels of electron doping. (k) Experimetal analysis of raman spectra for MoTe₂ phase diffenert colors correspond to 2H-, 1T′-, and T_d.⁹⁰ Copyright 2021, MDPI Publishing.

2.2 Synthesis methods

The synthesis of TMDs, particularly WS₂, WTe₂, and MoTe₂, has attracted widespread attention due to their extraordinary two-dimensional properties, making them highly desirable for applications in catalysis, nanoelectronics, and energy conversion.^91,92 Experimentally, these TMDs can be synthesized through a variety of bottom-up and top-down approaches, each offering specific advantages in terms of layer control,^93,94 crystallinity, phase selectivity, and defect engineering.^95–97 Among the bottom-up methods, chemical vapor deposition (CVD) stands out as a highly controllable and scalable technique for producing high-quality, monolayer or few-layer TMD films.^98,99 In this method, volatile precursors such as WO₃ or MoO₃ are used in conjunction with a chalcogen source (such as sulphur or tellurium) under a reducing atmosphere, typically hydrogen/argon, at elevated temperatures ranging from 200 to 1000 °C. However, controlling the growth temperature plays a critical role in determining the phase. At higher temperatures (e.g., 710 °C), the 1T′ phase is the dominant product, whereas at lower temperatures (e.g., 670 °C), the 2H phase predominates. Additionally, tensile strain applied during the growth process can facilitate the transition from 2H to 1T′, as strain reduces the activation energy barrier for this phase change. Electron doping through chemical dopants or external electric fields further stabilizes the 1T′ phase, enhancing its catalytic properties.^42,100 The reaction proceeds on a substrate like SiO₂/Si or sapphire, where the vapor-phase reactants condense and form uniform crystalline layers. For instance, in the synthesis of monolayer WS₂ by the hydrothermal method shown in Fig. 3(a), WO₃ is sulfurized in the presence of sulphur vapor to yield a uniform hexagonal 2H-WS₂ film, while MoTe₂ and WTe₂ require more complex control of temperature and tellurium partial pressure due to the higher volatility and reactivity of Te, often resulting in mixed or metastable phases such as 1T′ or T_d.¹⁰¹ In addition to CVD, hydrothermal and solvothermal methods have been effectively employed to synthesize TMD nanostructures like nanosheets, nanoflowers, and quantum dots, particularly valuable in electrocatalysis.^102,103 These solution-phase techniques involve the reaction of metal salts (e.g., ammonium tetrathiotungstate for WS₂ or MoCl₅ for MoTe₂) with chalcogen sources (like thiourea or Na₂TeO₃) in an autoclave under moderate temperatures and pressures. These methods allow control over morphology, defect concentration, and doping by adjusting reaction parameters such as pH, temperature, and surfactant presence. The crystal structure of monolayer WTe₂ is shown in Fig. 3(b), often favouring the formation of layered or nanorod structures due to the strong tendency of Te to crystallize in lower symmetry phases, and careful tuning of the reaction conditions is necessary to access the desired phase purity and stoichiometry. The crystal structures of monolayer WTe₂ are shown in Fig. 3(b) and (c), comparing to the side view and best view, individually. It can be clearly seen that the octahedron of tellurium molecules is distorted, in which reversal symmetry is broken. The auxiliary inclusion has endowed this material with distinctive properties that set it apart from other TMDs. Fig. 3(b) portrays the schematic outline of the test setup and compares synthesis methodologies utilized for the blend of WTe₂ nanostructures. The controlled growth of 1T′-WTe₂ nanostructures essentially relies on precursor design and substrate engineering strategies in our study. Firstly, W thwart was purposely oxidized in discuss environment shaping uniform WO₃ film on the foil surface. The details of oxidized W foils will be methodically examined afterward. High purity Te powder (500 mg) was utilized as the Te precursor with the help of one drop of water (0.05 ml) as shown in Fig. 3(b). The WO₃ foil and c-plane sapphire substrate were stacked vertically shaping a confined-space as show in Fig. 3(b), which can guarantee a steady environment and accurately control the Te and W concentrations through tuning the confined height. Fig. 3(c) shows the OM picture of as-prepared samples on a sapphire substrate showing NF shape. The structure of synthesized WTe₂ NFs was characterized with the help of Raman spectroscopy. Fig. 3(d) presents the typical Raman range of as-prepared 1T′ WTe₂ NF. It is interesting to note that six vibrational Raman modes were observed, which were denoted as A₂ (88.9 cm⁻¹), A₂₄ (109.6 cm⁻¹), A₁₃ (117.7 cm⁻¹), A₁₄ (135.9 cm⁻¹), A₁₇ (163.8 cm⁻¹) and A₁₉ (214.2 cm⁻¹), individually, in great agreement with past reports.^104,105Fig. 3(e) shows the Raman mapping of A⁹₁ mode, which further affirms the consistency of as-grown samples. As shown in Fig. 3(f), the thickness of few-layer WTe₂ NR was found to be ∼1.8 nm as examined by AFM, comparable to two nuclear layers. Furthermore, KPFM was employed to measure the surface potential of the synthesized WTe₂ nanoribbons under ambient conditions with 30% relative humidity, revealing that the surface potential of WTe₂ is 35 mV higher than that of the sapphire substrate, as shown in Fig. 3(g). The KPFM picture of WTe₂ NF was generally homogeneous revealing the uniform surface potential and charge distributions, which is typical for WTe₂-based nanodevices with high-performance. Alternatively, mechanical and liquid-phase exfoliation methods are widely used to isolate monolayer or few-layer TMD flakes from bulk crystals through sonication or shear mixing in appropriate solvents. While these top-down techniques are limited by low yield and structural inhomogeneity, they remain critical for fundamental studies and for producing defect-rich sheets, which are particularly beneficial for electrocatalytic reactions like the OER/ORR. Computationally, first-principles DFT plays a crucial role in guiding and understanding the synthesis of TMDs. DFT calculations help predict the thermodynamic stability of different phases (e.g., 2H, 1T, and 1T′) under varying chemical potentials and external conditions such as strain, doping, or electric field. The computational phase diagrams of MoTe₂ are shown in Fig. 3(h), where the 1T′ phase is stabilized at higher Te chemical potential or under electron doping, providing a theoretical basis for tailoring synthesis conditions to favour metallic phases suitable for catalysis. The most thermodynamically stable phases were reported to vary at various treatment temperatures.¹⁰⁶ Therefore, by controlling the growth temperature, a particular phase can be selectively obtained. The 1T′ phase was the dominant product at higher temperatures (710 °C), and the 2H phase was reported to be dominated at lower temperatures (670 °C).¹⁰⁷ Due to the continuous growth of 1T′ and 2H-MoTe₂, they formed separate lateral 1T′-2H-MoTe₂-homojunctions as shown in Fig. 3h. When the temperature rises, phase development was observed in Fig. 3(j), namely pure 1T′-phase → 1T′- and 2H-phase → pure 2H-phase → 2H- and 1T′-phase → pure 1T′-phase.¹⁰⁸ Strategies such as extending growth time¹⁰⁹ increasing Te atomic flux by raising the source temperature as mentioned in Fig. 3(i),¹¹⁰ and using molecular sieves and overlapping substrates to increase the control of MoO₃ precursor concentration¹¹¹ provide excessive Te during the growth periods, thus forming a stable stoichiometric-2H phase. The inverse 2H → 1T′ phase transition in MoTe₂ can be realized by applying temporary pressure at the same temperature and then rapidly releasing it.¹⁰⁹ Additionally, level 1D-2D Mo₆Te₆–MoTe₂ heterostructures within the mix dimensions were synthesized directly under medium-scale Te flux.¹¹² Due to the pioneering derivatives of different compositions, 2H and 1T′-MoTe₂,¹¹³ products after differentiation of MoO_2.0–2.5 and MoO₃ thin films were identified as 2H or 1T′-MoTe₂ as shown in Fig. 3(k). The authors suggested that the different densities of the products achieved selective phase synthesis of the products. Derivatives such as 2D-MoTe₂ hetero-phase¹¹⁰ connections, MoTe₂ field effect transistor (FET) arrays^114,115 and integrated MoTe₂ circuits were successfully manufactured.¹¹³

Fig. 3 (a) WS₂ nanosheet fabrication and roxarsone electrochemical detection.¹¹⁶ Copyright 2019, Springer Nature Link. (b) Diagram of the CVD setup used to grow WTe₂ nanoflakes and the synthesis method. (c) Optical microscopy image of WTe₂ nanoflakes on a sapphire substrate. (d) Raman spectrum of the as-grown WTe₂ nanoflakes. (e) Optical image of a single nanoflake with Raman intensity mapping at a specific vibration peak. (f) AFM image showing the thickness of a WTe₂ nanoflake. (g) Surface potential map of the same nanoflake.¹¹⁷ Copyright 2022, Nature. MoTe₂ hetero-phase homojunctions are synthesized directly. (h) Growth temperature control, (i) flux-controlled tellurization, and (j) growth temperature and gas flow rate control. (k) Tellurization of patterned antecedents with varying compositions at the same time.⁹⁰ Copyright 2021, MPDI publishing.

2.3 Surface changes/adjustment

Surface treatment and modification of TMDs,^118,119 particularly WS₂, WTe₂, and MoTe₂, represent critical strategies to tailor their physicochemical properties for advanced applications in electrocatalysis, nanoelectronics, and energy conversion.^14,120 These 2D materials possess unique surface chemistry due to their layered structure and exposed active edges,^121,122 but their pristine basal planes, especially in semiconducting 2H phases, are often chemically inert.^121,123 To overcome these limitations, various experimental surface engineering techniques have been developed to enhance their reactivity, conductivity, and surface area.^124,125 One of the most common approaches is plasma treatment, particularly using O₂ or Ar plasma, which introduces surface vacancies, functional groups, and even partial oxidation.^43,126 Oxygen plasma treatment of WS₂ introduces sulphur vacancies and oxygen-containing functionalities (like W–O bonds), leading to improved wettability, enhanced catalytic activity for the OER, and modified electronic structure.^127–129 WTe₂ and MoTe₂, being more chemically reactive due to tellurium's higher polarizability and lower electronegativity, are amenable to surface oxidation and defect formation under mild plasma or thermal treatments, resulting in a partial phase transition or formation of mixed W–O–Te or Mo–O–Te (ref. 130 and 131) surface species that significantly alter their electronic and catalytic characteristics. Another widely applied surface modification method is chemical functionalization, where specific functional groups or molecules are covalently or non-covalently anchored to the TMD surface.^63,132 In MoTe₂ and WTe₂, soft tellurium atoms can readily bond with organic molecules, thiol ligands, or even oxygen-containing groups introduced via solution-based treatments.^43,133,134 This functionalization can modulate charge transfer behaviour and interfacial compatibility in devices or catalysts.¹³⁵ For electrocatalysis, electrochemical anodization or chemical etching with agents like KOH, H₂O₂, or halogens has been used to selectively etch chalcogen atoms and expose more active metal sites. In WS₂, controlled chemical etching can lead to the transformation of the semiconducting 2H phase to the metallic 1T phase at the surface, a transition that is accompanied by significant enhancement in conductivity and catalytic activity. Furthermore, doping and substitutional surface modification, where foreign atoms such as N, O, F, or transition metals (like Fe, Co, or Ni) are incorporated into the TMD lattice or adsorbed onto the surface, have been used to tune the d-band centre, manipulate the local charge environment, and enhance the adsorption of OER/ORR intermediates. These doping processes often begin at the surface and can be confined to surface layers through methods such as wet impregnation, atomic layer deposition, or low-temperature annealing in reactive atmospheres. On the computational side, DFT plays a vital role in predicting and validating surface modification strategies. DFT calculations have been extensively used to model the formation energies of surface vacancies (e.g., sulphur or tellurium vacancies), analyse the adsorption behaviour of dopants or functional molecules, and simulate the effect of surface oxidation or reconstruction on the electronic density of states and band structure. For instance, DFT studies on WS₂ with sulphur vacancies or oxygen substitutions have shown significant shifts in the Fermi level and enhanced adsorption energies for key oxygenated intermediates (*OH, *O, *OOH), critical for OER performance. In the case of WTe₂ and MoTe₂, computational studies reveal that surface Te vacancies or Te–O terminations induce mid-gap states and reduce charge-transfer barriers, consistent with experimental observations of improved electrocatalytic activity after surface modification. DFT is also used to assess the thermodynamic feasibility and electronic impact of heteroatom doping at the surface, predicting optimal dopant concentrations and configurations that lead to catalytic enhancement. Additionally, machine learning models trained on DFT-derived datasets are beginning to predict how specific surface treatments (e.g., plasma dosage, annealing conditions) will affect surface properties, enabling more targeted experimental design (Table 1).

Table 1 Different types TMDs modified with various chemical agents/surfactants

TMDs	Modification agent/surfactant	Type of modification	Achievements	References
WS2	Sodium dodecylbenzene sulfonate (SDBS)	Surfactant-assisted liquid phase exfoliation	Produced few-layer 2H-WS2 nanosheets with improved electrical conductivity and stability	136
WS2	Sodium cholate	Surfactant-assisted hydrothermal synthesis	Controlled morphology leading to uniform nanosheets suitable for photocatalytic applications	137
WS2	CTAB, SDBS, PVP	Surfactant-assisted regulation in WS2/tourmaline composites	Enhanced dispersion and microstructure control, improving photocatalytic performance	138
WTe2	Oleyl amine	Surfactant-assisted colloidal synthesis	Enabled low-temperature synthesis of few-layer WTe2 nanostructures with tuneable morphology	139
WTe2	None (defect engineering)	Introduction of Te vacancies	Improved electrocatalytic hydrogen evolution reaction (HER) activity with lower overpotential	140
WTe2	Te buffer layer	Low-temperature deposition technique	Achieved high-quality monolayer 1T′-WTe2 with nearly complete coverage, beneficial for electronic applications	69
MoTe2	Reduced graphene oxide (rGO)	Hydrothermal fabrication of MoTe2/rGO composite	Enhanced electrocatalytic activity for HER due to synergistic effects between MoTe2 and rGO	141
MoTe2	p-Block element doping	Surface activity enhancement	Increased catalytic activity and electronic properties through doping strategies	142
MoTe2	None (phase engineering)	Mixed-phase synthesis with exposed edges and defects	Improved catalytic performance due to increased active sites from phase boundaries and defects	143
MoS2	Cetyltrimethylammonium bromide (CTAB)	Surfactant-assisted exfoliation	Achieved stable dispersions of monolayer MoS2, enhancing its applicability in various devices	144
MoSe2	Polyvinylpyrrolidone (PVP)	Non-covalent functionalization	Improved dispersion and stability in aqueous solutions, beneficial for biomedical applications	145
WS2/MoTe2	None (heterostructure formation)	Sputtering–CVD approach to create WS2/MoTe2 heterostructure	Enhanced catalytic activity for dye-sensitized solar cells (DSSCs) due to unique structure	146

---

2.4 Characterization techniques

A tremendous array of novel approaches has been set up to look at and think about the physiochemical and auxiliary properties of 2D materials including TMDs, and specific attention has been paid to the atomic structures and morphologies of TMDs due their advantages. We briefly discuss the common procedures that are utilized to characterize and analyse TMD structures (Table 2).^{119,147–150}

Table 2 Different common techniques used for the characterization of 2D TMD materials

Technique	Purpose	Information provided
Atomic force microscopy^151,152	Surface morphology and thickness measurement	Provides topographical mapping at the nanoscale, allowing determination of layer thickness and surface roughness
Scanning electron microscopy^153	Surface morphology and structural analysis	Offers high-resolution images to assess surface features and morphology of 2D materials
Transmission electron microscopy^154,155	Internal structure and crystallography	Enables visualization of internal structures, defects, and crystallographic information at atomic resolution
X-ray diffraction^156,157	Crystalline structure determination	Identifies phase composition and crystallographic structure through diffraction patterns
Raman spectroscopy^158–160	Vibrational mode analysis	Detects characteristic vibrational modes, providing information on material quality, number of layers, and strain
X-ray photoelectron spectroscopy^161,162	Surface chemical composition analysis	Determines elemental composition and chemical states of elements present on the surface
Photoluminescence (PL) spectroscopy^163–165	Optical property evaluation	Assesses electronic structure and bandgap by measuring emitted light from excited states
Ultraviolet–visible (UV–vis) spectroscopy^166	Optical absorption analysis	Measures absorbance to determine optical bandgap and electronic transitions
Fourier transform infrared spectroscopy^167,168	Functional group identification	Identifies chemical bonds and functional groups through infrared absorption spectra
Scanning tunnelling microscopy^169–171	Surface atomic structure imaging	Provides atomic-scale images of surface topology and electronic states

---

2.5 Applications

TMDs (WS₂, WTe₂, and MoTe₂) have emerged as versatile materials with applications spanning electronics, optoelectronics, energy storage, catalysis, and quantum technologies, accounting for 37%, 28%, 16%, 14%, and 11%, respectively, driven by their unique structural, electronic, and chemical properties. WS₂, a semiconducting TMD with a direct bandgap (∼2.0 eV) in its monolayer form, is extensively utilized in optoelectronics.^172,173 Its strong light–matter interaction and high photoluminescence quantum yield enable efficient photodetectors, light-emitting diodes (LEDs), and photovoltaic cells. Monolayer WS₂-based photodetectors exhibit ultrafast response times and broad spectral sensitivity, while heterostructures with graphene or other TMDs enhance charge separation for solar energy conversion.^174,175 Additionally, WS₂ catalytic edge sites, rich in sulphur vacancies, make it effective for the OER/ORR, where its low overpotential and durability outperform traditional platinum-based catalysts in acidic environments. In flexible electronics, WS₂'s mechanical robustness (Young's modulus ∼270 GPa) and tolerance to strain allow integration into wearable sensors and bendable circuits, while its atomic thickness minimizes interfacial scattering in ultra-scaled transistors.^176–178 WTe₂, a semi-metallic TMD with a distorted 1T′ structure, showcases exceptional electronic anisotropy and non-saturating magnetoresistance, positioning it as a candidate for nanoelectronics and spintronics. Its layered orthorhombic symmetry enables gate-tuneable ambipolar transport, making it suitable for FETs with high on/off ratios. WTe₂'s large magnetoresistance, attributed to electron–hole compensation, is exploited in magnetic sensors and memory devices.^179,180 In monolayer form, its predicted quantum spin Hall effect yields a topological insulator phase with potential for dissipationless edge conduction in quantum computing. In energy storage, WTe₂'s high electrical conductivity and interlayer spacing facilitate efficient ion intercalation, enhancing its performance in lithium-ion and sodium-ion battery anodes. Its anisotropic thermal conductivity also suits thermoelectric applications, where engineered nanostructures leverage phonon scattering to improve the figure of merit (ZT).^181–184 MoTe₂, a polymorphic TMD existing in semiconducting 2H, metallic 1T′, and topological Weyl semi-metallic (T_d) phases, offers phase-dependent functionalities. The 2H-MoTe₂ monolayer, with a narrow direct bandgap (∼1.1 eV), is ideal for near-infrared optoelectronics, including photodetectors and lasers,^185–187 while its type-II band alignment in heterojunctions with WSe₂ enables efficient exciton dissociation for solar cells. The metastable 1T′ phase, stabilized via strain or doping, exhibits metallic conductivity and catalytic activity for CO₂ reduction or OER.^188,189 Most notably, the T_d phase of MoTe₂ hosts Weyl fermions massless quasiparticles with chiral surface states enabling topological quantum devices such as ultra-low-power transistors and Majorana fermion-based qubits.¹⁹⁰ Phase transitions between 2H, 1T′, and T_d states, induced electrically or optically, underpin reconfigurable electronics and non-volatile memory applications.^42,191 In energy storage and catalysis, these TMDs exhibit complementary strengths.¹²³ WS₂'s sulphur-rich surfaces enhance polysulfide trapping in lithium–sulphur batteries, mitigating the shuttle effect,^192,193 while MoTe₂'s tuneable interlayer spacing improves ion diffusion kinetics in supercapacitors.^194,195 WTe₂'s high carrier mobility and stability under bias make it a robust electrocatalyst.¹⁹⁶ Environmental sensing leverages their surface reactivity and WS₂'s sulphur vacancies selectively adsorb NO₂ or NH₃ gases,¹⁹⁷ enabling ultrasensitive chemiresistive sensors,¹⁹⁸ while MoTe₂'s phase-dependent work function shifts detect volatile organic compounds (VOCs) at ppm levels.¹⁹⁹ Computational insights further expand their utility. DFT guides defect engineering in WS₂ to optimize OER/ORR activity,^200,201 while ML models predict optimal doping strategies for WTe₂ to enhance thermoelectric efficiency.^202,203 For MoTe₂, ab initio simulations map topological surface states, aiding the design of quantum interference devices. Experimentally, advances in CVD and molecular beam epitaxy (MBE) enable wafer-scale synthesis of these materials, though challenges like phase purity (MoTe₂) and ambient stability (WTe₂) persist (Fig. 4 and Table 3).^204,205

Fig. 4 The percentage of publications on different properties of TMDs including WS₂, WTe₂ and MoTe₂.^209–212

Table 3 TMD 2D materials' application and stability conditions

Material	Applications	Stability conditions	References
WS2	• Very high near-infrared light photocatalytic activity		206
	• A full solar-spectrum photocatalyst/photodegradation
WTe2	• Quantum spin Hall insulator extraordinary magnetoresistance effect, thermoelectric properties		207
	• Spintronic devices
MoTe2	• Small indirect bandgap, rapid electron transport during electro-reduction, long-term durability	Under air semiconducting dichalcogenides	208
	• Electroreduction of CO2 to CH4

---

3 Advancement on the approach for designing the 2D TMD-based OER/ORR electrocatalysts

3.1 Phase transition

The development of 2D TMD-based OER/ORR electrocatalysts has witnessed significant advancements through phase transition strategies aimed at enhancing their electronic conductivity and catalytic activity.^102,213 Most TMDs, especially in the 2H phase, exhibit an intrinsic semiconducting character.²¹⁴ In Liu, X. et al. and Fang, Y. et al. a detailed structural comparison between the 2M and 2H phases of WS₂ reveals the fundamental atomic rearrangements underlying the phase transition. In the 2M phase (also referred to as 1T′), each layer consists of distorted [WS₆]⁸⁻ octahedra sharing edges along the bc-plane, with tungsten atoms forming zigzag W–W chains along the b-axis. This results in nonuniform W–W spacings of approximately d₁ = 2.28 Å and d₂ = 3.43 Å as shown in Fig. 5(a). When viewed from the b direction, the W and S atoms occupy six distinct atomic planes, stacking in a complex A/A′–B/B′–C/C′ configuration, where sulphur atoms in alternating planes share two W atoms as shown in Fig. 5(a). In contrast, the 2H phase (or 1H structure) adopts a more symmetric and energetically stable form. It is composed of [WS₆]⁸⁻ trigonal prisms, where W atoms are uniformly spaced at d₃ = 2.70 Å as shown in Fig. 5(b). The atomic layers follow a simpler A–B–A stacking sequence, and each pair of sulphur atoms positioned in adjacent planes shares three W atoms, reflecting the prismatic coordination typical of 2H WS₂. This structural transition plays a critical role in tuning the electronic and catalytic properties of WS₂-based materials. The phase transition from 2M to 2H WS₂ is governed by a comprehensive reconstruction of its atomic framework, driven by intrinsic structural differences between the two phases which are mentioned in Fig. 5(c). This transformation entails the dissolution of W–W chains and progressive displacement of tungsten atoms along the c-axis, eventually yielding a uniform W–W spacing characteristic of the 2H phase. Concurrently, sulphur atoms shift laterally within the lattice, reorganizing to form new W–S bonds and converting the local coordination geometry from octahedral (in 2M) to trigonal prismatic (in 2H). This process also modifies bond lengths and angles and alters the stacking sequence of [WS₆]⁸⁻ units from a six-plane (2M) to a three-plane (2H) configuration. Structurally, the transition is marked by a slight in-plane lattice contraction and out-of-plane expansion, reflecting the shorter W–S bonds and greater interlayer separation in the 2H phase.^215,216 In Wang et al., and Dong et al., a comparative analysis of the electronic structures and crystal configurations of monolayer 2H- and T_d-phase WTe₂ reveals distinct structural and stability characteristics. After full structural optimization, both top and side views, Fig. 5(d), confirm the atomic arrangements, where tungsten (W) and tellurium (Te) atoms are denoted by blue and orange spheres, respectively, and the red dotted lines indicate unit cell boundaries. The T_d-phase WTe₂, characterized by an orthorhombic lattice (space group Pmn2₁), places both W and Te atoms at Wyckoff positions 2a and is widely recognized as the most thermodynamically stable phase, exhibiting semi metallic behaviour. The 2H-phase WTe₂ adopts a hexagonal crystal structure and is a direct bandgap semiconductor. Structurally, the T_d-phase exhibits lattice constants of 6.33 Å and 3.50 Å in x and y directions, respectively, while the 2H-phase maintains a single in-plane constant of 3.60 Å. Energy calculations reveal that the 2H-phase is 0.09 eV higher in energy per formula unit than the T_d-phase, corroborating prior theoretical studies and confirming the superior thermodynamic stability of the T_d form. The optimized lattice constants and structural parameters for both phases shown in Fig. 5(e) align closely with the existing literature,²¹⁷ providing a solid foundation for exploring their phase-dependent electronic and catalytic properties.^218,219 In Mortazavi et al., studies of the atomic structures of monolayer MoTe₂ reveal distinct structural features across its three polymorphs: 2H, 1T, and 1T′ phases. As shown in Fig. 5(f), all three configurations exhibit lattice orientations analogous to graphene, characterized by armchair and zigzag directions. The 2H phase adopts an ABA stacking sequence, while the 1T and 1T′ phases follow an ABC stacking order. For both 2H and 1T phases, the lattices are well-described by a hexagonal lattice constant (α) and Mo–Te bond length. The predicted lattice constants are 3.549 Å for 2H and 3.495 Å for 1T, with the 2H value aligning closely with the experimental measurement of 3.551 Å, deviating by less than 0.06%. The corresponding Mo–Te bond lengths are 2.731 Å (2H) and 2.756 Å (1T). The 1T′ phase adopts a distorted rectangular unit cell, with calculated lattice dimensions of 3.449 Å and 6.374 Å, which are in excellent agreement with experimental values of 3.452 Å and 6.368 Å reported by Wang et al.²²⁰ Notably, 1T′-MoTe₂ features two distinct Mo–Te bond lengths, measured at 2.718 Å and 2.823 Å, reflecting its anisotropic and lower-symmetry nature.²²¹ By leveraging these phase transition strategies, TMD 2D materials have demonstrated enhanced electrochemical performance, offering promising pathways for next-generation OER/ORR electrocatalysts in energy conversion and storage applications.^135,222

Fig. 5 (a–c) Crystal structures of 1T′-WS₂ (octahedral, A/A′–B/B′–C/C′ stacking) and 1H-WS₂ (trigonal prismatic, A–B–A stacking). Top and side views shown 1T′-WS₂ overlaid on 1H-WS₂. W (blue) and S (orange) atoms are shown. Dashed lines indicate broken W–W and W–S bonds; orange arrows show atomic shifts. Dissociated S atoms relocate (dotted circles) to form new W–S bonds.²¹⁵ Copyright 2024, Nature. (d and e) Top and side views of 2H- and T_d-phase WTe₂ monolayers. Blue and orange spheres represent W and Te atoms, respectively. Schematics illustrate the atomic arrangements in both phases.²¹⁹ Copyright 2022, Elsevier. (f) Top and side views of monolayer MoTe₂ in 2H, 1T, and 1T′ phases. Mechanical properties were analysed along the armchair and zigzag directions.²²¹ Copyright 2018, MDPI Publishing.

3.2 Defect engineering

Defect engineering and doping of TMDs are critical strategies to enhance their electrocatalytic activity for the OER/ORR. These methods tailor the electronic structure, expose active sites, and optimize adsorption energetics for oxygen intermediates (*OH, *O, *OOH) by introducing atomic-scale defects (e.g., chalcogen vacancies, metal vacancies) or substituting host atoms with hetero elements (e.g., transition metals, nonmetals).^223,224 In Song Li et al., recent investigations into point defects in WS₂ have highlighted the electronic behaviour of a carbon-substituted sulphur (C_s) defect, where a carbon atom replaces a sulphur atom in the top layer of the crystal as shown in Fig. 6(a). This defect can be induced experimentally through hydrogen depassivation of a carbon–hydrogen impurity via an STM voltage pulse. Supported on a graphene substrate, the C⁻_s defect adopts a negatively charged state as discussed in Cochrane, K. A. et al.²²⁵ verified by scanning tunnelling spectroscopy (STS), which reveals occupied and unoccupied defect states at −0.4 V and +0.52 V, respectively. First-principles calculations using HSE-corrected DFT, including spin–orbit coupling (SOC), provide deeper insight into the electronic structure. The inclusion of SOC shifts the valence band maximum (VBM) and introduces band splitting, leading to a refined bandgap of 2.76 eV for pristine WS₂. For the C⁻_s defect, the occupied in-gap level appears at 1.09 eV above the VBM shown in Fig. 6(c), in excellent agreement with the STS-derived value (∼1.1 eV), considering spectral broadening. The degenerate defect states (e_x, e_y), located about 0.62 eV below the VBM, exhibit an SOC-induced splitting of 113 meV, consistent with previous measurements on sulphur vacancies explored in Schuler, B. et al.²²⁶ Spatial wavefunction analysis shown in Fig. 6(d) confirms C_3v symmetry with a₁ orbital character, and these states primarily originate from the d orbitals of adjacent W atoms. Notably, the unoccupied defect level lies well above the conduction band and does not correspond to the C⁻_s state, reaffirming the identification of the observed occupied state as stemming from the negatively charged C⁻_s defect.²²⁷ WTe₂ is well known to crystallize in the distorted 1T (T_d) phase in its three-dimensional (3D) bulk form, where it adopts the orthorhombic space group Pmn2₁ and belongs to the C_2v (mm2) point group. However, as the dimensionality is reduced from bulk to two dimensions through few-layer to monolayer thickness both the structural and electronic characteristics of WTe₂ undergo notable transformations. In particular, the structure transitions into the 1T′ phase, which is stabilized at lower dimensions. Consistent with prior studies, bulk WTe₂ maintains C_2v symmetry, while few-layer structures exhibit C_s symmetry, and the monolayer form transitions to C_2h symmetry, reflecting the increasing influence of reduced dimensionality and symmetry breaking. Monolayer and few-layer samples can be experimentally isolated via mechanical exfoliation from bulk crystals.²²⁸ Complementing this, theoretical modelling of these reduced-dimensional structures was performed by systematically removing excess layers from the bulk model and applying a vacuum buffer of 15 Å along the out-of-plane (z) direction to prevent interlayer interactions. The optimized atomic structures of monolayer (1L) and quadrilayer (4L) 1T′-WTe₂ are presented in Fig. 6(e) and (f), illustrating the characteristic distortions and layer-dependent symmetry evolution that play a critical role in defining the physical properties of WTe₂ at the nanoscale. The atomic structure of a monolayer WTe₂ consists of three covalently bonded atomic planes, arranged in a Te–W–Te stacking sequence along the z-axis. Each tungsten (W) atom forms a triangular pyramidal coordination with three nearest-neighbour tellurium (Te) atoms positioned both above and below the W layer. Notably, these pyramids are rotated by 180° relative to one another across the layer, introducing a distinct structural asymmetry.²²⁹ In Fig. 6(g), the characteristic distortion of W atoms primarily due to strong intermetallic interactions that draw the metal atoms closer together has been calculated to be 0.87 Å along the y-axis and 0.21 Å along the z-axis, values that are well-aligned with existing experimental and theoretical data.²³⁰ This distortion also affects the Te sublattice, where the atoms do not lie in a flat plane but instead form a zigzag configuration along the y-direction. The out-of-plane buckling of Te atoms is measured to be approximately 0.6 Å, further corroborating previous structural studies. Together, these geometric features particularly the buckled Te chains and asymmetric W coordination define the intrinsic anisotropy and low symmetry of monolayer WTe₂, which are crucial to its electronic, optical, and topological properties. In multilayer 1T′-WTe₂, the stacking arrangement between adjacent layers resembles a lock-and-key configuration, where the surface undulations (or ripples) of one layer complement the grooves of the next. This leads to an alternating stacking motif in which successive layers are rotated 180° with respect to each other around the z-axis. The layers are held together by weak van der Waals interactions, and the interlayer separation, often referred to as the van der Waals distance (h), defines the spacing between these atomically thin sheets. To capture the full picture of this stacking behaviour, the structural parameters were systematically evaluated. These include in-plane lattice constants, bond lengths between neighbouring W–Te and W–W atoms, interlayer van der Waals distances, and cohesive energies per atom for both monolayer and multilayer 1T′-WTe₂ configurations. This comprehensive structural analysis underscores how interlayer geometry and weak interlayer coupling influence the physical properties and phase stability of the material.²³¹ To gain deeper insight into butanethiol-induced doping and defect passivation in MoTe₂, a detailed X-ray photoelectron spectroscopy (XPS) study was conducted on exfoliated MoTe₂ flakes subjected to three conditions: pristine, annealed at 250 °C in N₂, and post-treated with butanethiol. As illustrated in Fig. 6(i–k), Te vacancies were intentionally introduced via thermal annealing, as weak Mo–Te bonds are known to dissociate around 200–250 °C. These vacancies were subsequently healed by chemisorption of butanethiol, with sulphur atoms from the thiol molecules occupying the vacant Te sites.²³² XPS analysis revealed a blue shift (∼0.2 eV) in both Mo 3d and Te 3d core-level spectra following N₂ annealing, confirming the formation of Te vacancies present in Fig. 6(l) and (m).²³³ In contrast, thiol-functionalized MoTe₂ exhibited red shifts of 0.15 eV (Mo 3d) and 0.3 eV (Te 3d), indicating vacancy healing and hole doping. Similar red shifts were also observed when butanethiol was applied directly to pristine MoTe₂, underscoring its doping capability even without prior defect generation. These binding energy shifts align with known signatures of carrier polarity modulation in 2D TMD blue shifts for electron doping and red shifts for hole doping. Supporting this, MoTe₂ field-effect transistors (FETs) exhibited altered transfer characteristics and band alignments consistent with enhanced hole transport following thiol treatment. Further confirmation was provided by high-resolution STEM, which visually verified the effective filling of Te vacancies. Collectively, these results provide clear evidence that butanethiol chemisorption not only passivates surface defects in MoTe₂ but also acts as an efficient molecular hole dopant, offering a practical route for defect engineering and electronic modulation in 2D materials.^234,235

Fig. 6 (a–d) Side view of the optimized ground-state geometry of the carbon-substituted sulphur vacancy, energy level diagrams without and with spin–orbit coupling showing spin channels and corrected defect levels, along with spatial wave functions highlighting neighbouring tungsten atoms around the carbon defect.²²⁷ Copyright 2022, Nature. Optimized crystal structures of WTe₂: (e) monolayer (1L) side and top views, and (f) quadrilayer (4L) 1T′ phase side view, with blue-shaded regions indicating unit cells (excluding vacuum). (g) Distorted octahedral coordination of W with Te, showing rotated triangle pairs (red dashed and blue solid lines). (h) 3D orthorhombic Brillouin zone with the 2D BZ and high-symmetry points (Γ, X, S, Y) highlighted.²³¹ Copyright 2022, Nature. High-resolution XPS of MoTe₂ under three conditions: pristine (black), changes after thermal annealing in nitrogen (blue), and butanethiol-treated (red). (i–k) 2D schematics highlight Te vacancies (dark circles) and thiol-bound sulphur (red spheres). (l and m) Mo 3d and Te 3d spectra compare surface binding states across the three treatments.²³⁵ Copyright 2024, American Chemical Society.

3.3 Strain engineering

Engineering strain in 2D TMDs offers a powerful approach to optimize their electrocatalytic performance for the OER/ORR. Strain modulation alters not only electronic and structural properties but also directly influences critical parameters such as current density, bond lengths, and dissociation energies, which collectively govern catalytic efficiency.²³⁶ Wang, F. et al. led the strain mapping of monolayer WS₂ under varying levels of applied tensile strain and revealed key insights into its mechanical response and interfacial behaviour as shown in Fig. 7(a) and (b). At 0% applied strain, the flake displays an average internal strain near zero, though minor inhomogeneities are observed. These variations are attributed to the presence of wrinkles or surface irregularities introduced during sample preparation, specifically when pressing the flake onto the substrate a common artifact reported in previous studies.²³⁷ Notably, these wrinkles tend to flatten out as strain is applied. Upon increasing the external strain to 0.35%, a non-uniform strain distribution develops within the flake, the strain originates from the edges and gradually propagates toward the centre, where it eventually plateaus. This trend is clearly illustrated in Fig. 7(c), which plots strain variations along selected cross-sectional lines of the flake. The observed strain behaviour is consistent with that reported for monolayer graphene, and can be effectively described using continuum mechanics frameworks, particularly the shear-lag model.^238,239 This model captures how strain is transferred from the substrate to a 2D material, reinforcing its applicability to the deformation mechanics of atomically thin semiconductors like WS₂.²⁴⁰ Muscher, K. P. et al. demonstrated a solid-state electrochemical platform, utilizing a poly(ethylene oxide) (PEO)-based solid electrolyte, designed for high-vacuum compatibility and precise control at the single-flake level as shown in Fig. 7(d). This setup enabled real-time observation of lithiation-induced structural and symmetry changes in ultrathin WTe₂ flakes (<100 nm) through operando electron diffraction. The platform preserved the same lithiation/delithiation redox couple used in prior studies but offered improved integration for in situ structural probing. The full cycle of lithium intercalation and deintercalation occurred within a narrow voltage window of just ∼150 mV (1.15–1.30 V vs. Li/Li⁺), suggesting the feasibility of low-voltage switching applications. This small voltage hysteresis is particularly promising for symmetric cell configurations, such as a Li_xWTe₂–Li₀._5−xWTe₂ system, analogous to the behaviour previously reported for Li_xTiO₂ symmetric cells using similar solid-state electrolytes.²⁴¹Fig. 7(e) illustrates a novel strain-engineering approach used to study WTe₂ single crystals, where the crystal is bonded to a Si substrate using indium as an adhesive. The process involves melting indium with an underlying heater, then placing the WTe₂ flake on top. Upon cooling, the indium solidifies, and due to the thermal expansion mismatch between WTe₂ and indium, along with non-uniform solidification, significant local strains are introduced. These strains originate from both the WTe₂/indium interface and the mechanical constraints at the flake's edges, often manifesting as surface rippling on the WTe₂ layer. Strain mapping shows that in regions without structural defects, the shear strain component (ε_xγ) remains relatively low, fluctuating within ±1% as mentioned in Fig. 7(f). However, in regions containing twin domains, the ε_xγ values rise significantly averaging between 5% and 9%, and in some cases reaching up to 7–9% with increased variability. This distinct contrast in shear strain behaviour between normal and twinned regions suggests that ε_xγ is strongly correlated with the formation of twin domain boundaries (TDBs). These findings are consistent with previous reports in monolayer WS₂, where shear strain was identified as a key factor in driving grain boundary migration, highlighting the broader relevance of ε_xγ in strain-driven structural transformations in 2D materials.²⁴²Fig. 7(g) presents the atomic structures of 2H and 1T′ phases of MoTe₂, depicted using a ball-and-stick model. In the 2H phase, tellurium (Te) atoms align symmetrically when viewed from the top, preserving the material's characteristic hexagonal symmetry. In contrast, the 1T′ phase exhibits distorted atomic coordination, causing the Te atoms to shift from their ideal positions and resulting in an in-plane lattice expansion of approximately 3% (6.3 Å in 1T′ versus 6.1 Å in 2H). The phase transition between these two polymorphs can be modulated by tensile strain, as schematically illustrated in Fig. 7(i). This transformation is facilitated by a strain-induced reduction in the activation energy barrier, along with a change in the relative cohesive energies of each phase under mechanical loading. The corresponding temperature–force phase diagram in Fig. 7(h) delineates the boundary between 2H and 1T′ phases, showing how both applied force and temperature influence the transition dynamics. While precise quantification of in-plane strain remains challenging due to spatial variability, the estimated strain values are on the order of 0.1%, which is consistent with results from AFM and Raman spectroscopy under vertical compression. These findings underscore the sensitivity of MoTe₂'s phase behaviour to external mechanical stimuli and highlight the feasibility of strain-engineered phase control in two-dimensional materials.²⁴³ To explore the interplay between mechanical strain and electronic transport in two-dimensional materials, Hou, W. et al. designed a device was using single-crystal relax or ferroelectric Pb (Mg₁/₃Nb₂/₃)_0.71Ti_0.29O₃ (PMN-PT) substrates as a gate dielectric, with a thickness of approximately 0.25–0.3 mm. To monitor strain effects, micropatterned directional strain gauges were integrated onto the ferroelectric surface, while the electrical behaviour of the MoTe₂ channel was assessed using conventional transfer curve measurements as shown in Fig. 7(j). For device fabrication, a 13 nm-thick exfoliated 1T′-MoTe₂ flake was brought into contact with 35 nm nickel (Ni) electrodes, which applied a localized in-plane tensile stress of 0.58 GPa at the contact regions. The resulting transfer characteristics, shown in Fig. 7(l), displayed reversible on–off switching behaviour at room temperature. This response closely mirrored the characteristic strain electric field hysteresis (“butterfly curve”) observed in ferroelectric materials shown in Fig. 7(k), confirming the strain-coupled modulation of conductivity. The reproducibility of this strain-sensitive switching behaviour was validated across more than ten MoTe₂ devices of varying thicknesses fabricated on different PMN-PT (011) substrates, all exhibiting a similar bipolar modulation in conductivity with strain, often exceeding one order of magnitude. These results demonstrate the strong potential of strain engineering via ferroelectric substrates to modulate the electronic transport properties in MoTe₂-based devices, offering promising avenues for non-volatile and tuneable electronics.²⁴⁴

Fig. 7 (a) Strain distribution in monolayer WS₂ based on exciton peak positions at (a) 0% and (b) 0.35% applied strain. (c) Extracted strain profile along the dashed line in (a and b), with shear-lag model fits shown for 0% (orange) and 0.35% (green) strain using ns = 10.²⁴⁰ Copyright 2020, IOPScience Publishing. (d) All-solid electron-transparent single-flake electrochemical platform using PEO electrolyte. Left inset: Cyclic voltammogram at 90 °C for a WTe₂ flake. Centre inset: Electron diffraction of pristine (top) and in situ lithiated. Right inset: Optical image of the platform with WTe₂ flake.²⁴¹ Copyright 2021, Wiley. (e) Schematic illustrating sample stacking and strain induced by thermal expansion/contraction of indium used as adhesive between WTe₂ and the Si substrate.²⁴² Copyright 2019, Springer Nature Link. (f) Average shear strain line profiles within white frames and strain measurements at multiple sample locations separated beyond the STM detection window, showing mean values (dots) with standard deviations (error bars).²⁴² Copyright 2019, Springer Nature Link. (g) Atomic structures of 2H and 1T′ MoTe₂, with phase transition controlled by in-plane tensile strain. (h) Schematic showing strain-modulated phase transition barrier; tensile strain lowers activation energy and reduces the transition temperature. (i) Temperature–force phase diagram of semiconducting 2H and metallic 1T′ MoTe₂.²⁴³ Copyright 2016, American Chemical Society. (j) Schematic of the ferroelectric strain field-effect device. (k) Strain evolution from Ni gauges after the 1st and 6th sweeps, and post temperature cycling. (l) Strain-driven transistor operation on the 13 nm MoTe₂ channel.²⁴⁴ Copyright 2019, Nature Publishing.

3.4 Doping engineering

Doping engineering has emerged as a highly effective strategy to enhance the electrocatalytic properties of 2D TMDs for the OER/ORR. The pristine forms of these TMDs, especially in their stable 2H semiconducting phase, often suffer from poor electrical conductivity and limited active sites, which hinder their catalytic performance.³⁵ Arumugam, D. et al. evaluated the electrochemical performance of engineered 2D materials; a graphene/WS₂ heterostructure (GW) and its transition metal (TM)-doped variants (TM@GW) incorporating dopants from the Sc–Zn series were systematically investigated as shown in Fig. 8. Previous studies established that the pristine GW system behaves as a zero-bandgap semiconductor, while the introduction of transition metal dopants induces metallic character, thereby enhancing its electrochemical activity. To further explore synergistic doping effects, a nitrogen atom was substituted at a sulphur site in the WS₂ lattice, creating N-doped GW (NGW) and N-doped TM@GW (TM@NGW) structures as shown in Fig. 8(b) and (c). Upon structural optimization, the W–N bond length was calculated to be ∼2.02 Å, closely aligning with reported values (∼2.03 Å), while only a slight elongation (∼0.02 Å) was observed in nearby W–S bonds, indicating minimal lattice distortion.²⁴⁵ These changes are consistent with earlier computational findings. The thermodynamic stability and feasibility of formation for these doped systems were evaluated through cohesive energy (E_coh) and formation energy (E_f) calculations. The NGW structure exhibited negative cohesive (∼−0.55 eV) and formation (∼−6.90 eV) energies, confirming that the doped configurations are both energetically favourable and structurally stable. These findings suggest that dual doping through transition metals and nitrogen offers a promising strategy for tailoring the electronic and electrochemical properties of graphene–WS₂ heterostructures for catalytic applications.²⁴⁶ Substitutional doping has emerged as a powerful strategy to modulate the electronic properties of transition metal dichalcogenides (TMDs), particularly by tailoring the electronic band structure near the Fermi level.^247,248 Dong, X. et al. studied the impact of substitutional doping on monolayer WTe₂ in both 2H and T_d phases. Using a 2 × 3 supercell model comprising 36 atoms, researchers substituted a central Te atom with non-metal dopants (Al, B, C, N, P, and Si) and a W atom with transition metals (Co and Fe), generating a total of sixteen doped configurations at a doping concentration of approximately 2.78%. The structural models are depicted in Fig. 8(d–g). To evaluate thermodynamic stability, formation energies were computed using the expression:

E_form = E_doped + E_W/Te − E_pristine − E_x (1)

where each term corresponds to the total energies of the doped system, isolated host atom (W or Te), pristine WTe₂, and dopant atom, respectively.²¹⁹ The interaction between Rh dopants and the MoTe₂ monolayer surface has been explored to understand site selectivity and bonding characteristics. Four potential adsorption sites were evaluated: the centre of the hexagonal ring (TH), the top of Mo (TMo), the top of Te (TTe), and the bridge site between two Te atoms (TB). The binding energy (Eb), defined as:

E_b = E_Rh–MoTe2 − E_Rh − E_MoTe2 (2)

was used to determine the thermodynamic stability of each configuration. Among the tested sites, the TMo site was found to be the most stable, with the lowest binding energy of −2.69 eV. In this configuration, the Rh atom forms three equivalent Rh–Te bonds (2.54 Å), which are shorter than the sum of the covalent radii of Rh and Te (2.61 Å), confirming the formation of strong covalent interactions. In comparison, other sites exhibited less favourable binding energies (−2.14 eV at TH and −1.28 eV at TTe), indicating weaker interactions. Rh showed a stronger affinity for Se in MoSe₂ than for Te in MoTe₂, as evidenced by both shorter Rh–Se bond lengths and more negative binding energies. Charge transfer analysis using the Hirshfeld method reveals that Rh acts as an electron acceptor, drawing ∼0.045 electrons from the MoTe₂ surface.²⁴⁹ This observation is consistent with electron deformation density (EDD) analysis which shows significant charge accumulation around the Rh dopant as shown in Fig. 8(h and i).²⁵⁰

Fig. 8 Optimized structures of (a) graphene/WS₂, (b) N-doped graphene/WS₂, and (c) N- and TM-doped graphene/WS₂.²⁴⁶ Copyright 2024, American Chemical Society. (d–g) Top view of X-doped monolayers, where X = Al, B, C, Co, Fe, N, P, or Si. Orange, blue, and purple spheres represent Te, W, and dopant atoms, respectively.²¹⁹ Copyright 2022, Elsevier. (h) MSC and (i) EDD of Rh-doped MoTe₂ monolayer; green and rosy regions indicate electron accumulation and depletion, respectively.²⁵⁰ Copyright 2020, Springer Open.

3.5 Engineering grain edges

Engineering grain edges in 2D TMDs has become a promising strategy for enhancing their electrocatalytic performance in both the OER and ORR. Unlike basal planes, which are often catalytically inert in their pristine form, grain edges introduce highly active sites, unsaturated metal coordination, and altered electronic properties, making them highly desirable for electrocatalysis.^102,251 Liu, L. et al. discussed that grain boundaries (GBs) play a pivotal role in defining the properties of polycrystalline 2D materials such as WS₂. The density of GBs defined as the total GB length per unit area is directly influenced by the lateral grain size and its distribution. When grains are reduced to the nanoscale, the GB density increases sharply. This phenomenon is clearly visualized in the low-magnification HAADF-STEM image shown in Fig. 9(a), which reveals densely packed grains under 100 nm in size, consistent with AFM observations. These grains are interconnected via distinct GBs, confirming that the as-synthesized WS₂ film is polycrystalline in nature. Atomic-level analysis shown that the GBs are structurally complex and are mainly composed of 5|7 ring defects with a sulphur-rich composition indicated in Fig. 9(b). To understand the impact of GBs on mercury ion (Hg²⁺) adsorption, DFT calculations were carried out comparing GB regions and defect-free monolayer WS₂. The results, Fig. 9(c and d), reveal that Hg²⁺ ions preferentially adsorb at pentagonal hollow sites along GBs, with significantly lower adsorption energies than on pristine areas. This suggests that GBs act as effective traps for Hg²⁺. Charge density difference analysis further confirms a notable charge transfer (∼0.1 e⁻) from Hg²⁺ to neighbouring S atoms at GBs, indicating the formation of covalent Hg–S bonds. These findings underline the critical role of GBs not only in determining the structural characteristics of 2D WS₂ films but also in enhancing their chemical reactivity for heavy metal ion capture.²⁵² WTe₂ crystallizes in a layered structure characterized by van der Waals (vdW) interactions between adjacent layers. Its low-symmetry 1T′ phase arises from an in-plane distortion of tungsten (W) atoms along the b-axis, deviating from the ideal 1T structure. This distortion forms zigzag W chains along the a-axis, with adjacent chains displaced vertically to different heights as shown in Fig. 9(e). Notably, the direction of distortion alternates between layers, giving rise to top and bottom surfaces with distinct atomic terminations. Upon mechanical exfoliation, both surface types can co-exist on a single flake, often separated by monolayer-high step edges. This is clearly visualized in the scanning tunnelling microscopy (STM) image in Fig. 9(f), where standing wave patterns arising near step edges and atomic defects propagate along the chain direction. These features highlight the anisotropic electronic structure of WTe₂. The observed step height (∼0.72 nm) corresponds to a single WTe₂ layer, confirming that the two terraces represent different crystallographic surfaces. The nature of these surfaces is further supported by defect symmetry, which serves as an indicator of the underlying atomic configuration. To accurately capture the intrinsic topological properties of monolayer 1T′-WTe₂, the study employs finite-width ribbon models rather than infinitely periodic structures. This approach ensures that the edge states (ES) on opposite sides do not hybridize, preserving their localized nature. Both the n-field method and the hybrid Wannier charge centre (Wilson loop) analysis confirm the presence of nontrivial topological edge states as shown in Fig. 9(g), which arise as a direct consequence of the material's bulk topological character. The presence of these edge states is found to be independent of edge orientation, confirming their robustness. Computational results show that topological edge states emerge at both b-edges and various a-edge terminations with no overlapping trivial states, a hallmark of true topological protection. To further assess environmental stability, the edge-state behaviour was examined in a 30 Å-wide monolayer ribbon placed on another WTe₂ layer. Following full structural relaxation using first-principles calculations, only minor atomic displacements were observed along the edges, along with a slight vertical separation from the substrate. Notably, the topological edge state at the step edge mentioned in Fig. 9(i) retained both qualitative and quantitative agreement with that of the freestanding monolayer shown in Fig. 9(h), highlighting the robustness of the topological edge state against substrate-induced perturbations.²⁵³Fig. 9(j), (i–vi) present a mirror-symmetric grain boundary formed between two 1H-phase MoTe₂ grains with an approximate 60° misorientation, a configuration previously observed in related TMD systems such as MoTe₂/MoS₂ heterostructures.²⁵⁴ In this study, the misoriented grains were not originally present; instead, they emerged dynamically during the experiment from an initially single-crystalline 1H region. Time-resolved imaging reveals that molybdenum (Mo) atoms near the grain boundary migrate in a correlated manner, typically shifting by half a lattice vector. For example, three atoms marked in Fig. 9(j-i) are shown to displace in the indicated direction, and their new positions are confirmed in panel j-ii. This atomic rearrangement drives the migration of the grain boundary within the MoTe₂ lattice. Density functional theory (DFT) simulations show that the activation energy for Mo atom migration at the grain boundary is ∼2.3 eV for the first atom and reduces to ∼2.0 eV for the second, suggesting a cooperative mechanism. The process begins with chalcogen (Te) vacancies, which are created by electron irradiation, followed by Mo atom displacement to stabilize the lattice, a behaviour analogous to grain boundary migration observed in MoSe₂. This migration mechanism contrasts with that in WSe₂ and WS₂, where chalcogen atoms (rather than metal atoms) around the dislocation core are primarily responsible for boundary movement. These findings emphasize the element-specific dynamics of grain boundary evolution in different TMD systems and highlight the unique metal-atom-driven reconstruction behaviour in Mo-based dichalcogenides.²⁵⁵

Fig. 9 (a and b) Low-magnification HAADF-STEM image and SAED pattern of as-grown polycrystalline 1L WS₂; inset in (a) shown grain size distribution. (c) DFT analysis of Hg²⁺ adsorption on WS₂ grain boundaries: (a) six adsorption sites and charge density difference with adsorption energies. S (yellow), W (blue), and Hg (red). (d) Charge density difference between Hg²⁺ and WS₂ at the GB, showing charge transfer to nearby S atoms. Yellow, blue, and red spheres represent S, W, and Hg atoms, respectively. Adsorption energies (E_ads) are listed between (c) and (d).²⁵² Copyright 2021, Nature Publishing. Topography of WTe₂: (e) schematic of crystal structure with step edge (W: cyan, Te: orange), (f) pseudo-3D image showing step edge and defects on both terraces. Black line indicates cross-section; red and green arrows highlight mirror-symmetric defects.²⁵³ Copyright 2017, Nature Publishing. (g) Spin polarization of edge states (red/blue: opposite spin z-components), (h) electronic structure of monolayer 1T′-WTe₂ with relaxed edges (ab initio), (i) the same as (h) with an added bottom 1T′-WTe₂ layer (periodic in a and b), light-red regions indicate bulk states; others are edge states.²⁵³ Copyright 2017, Nature Publishing. (j) Grain boundary migration in MoTe₂ monolayer STEM-HAADF sequence showing Mo atom movement (circles) and migration direction (arrows), (k) simulated grain boundary model with migration paths, (l) DFT-calculated migration barriers for two Mo atoms using the nudged elastic band method, and (m) simulated image of the grain boundary, experimental images are raw data in false colour.²⁵⁵ Copyright 2018, American Chemical Society.

3.6 Fabrication of heterostructures

The fabrication of heterostructures in 2D TMD materials has emerged as a highly effective strategy to enhance their electrocatalytic properties for the OER/ORR in both acidic and alkaline media. The primary limitation of pristine TMDs in electrocatalysis stems from their limited conductivity, poor catalytic site availability, and low stability in electrochemical environments.¹⁰² Alahmadi, M. et al. illustrated a one-step hydrothermal synthesis route for fabricating metal-doped WS₂ nano catalysts as shown in Fig. 10(a), highlighting a widely adopted strategy for modifying the properties of transition metal dichalcogenides (TMDs). Among various doping techniques, substitutional doping where transition metal atoms are inserted directly into the TMD lattice is recognized as particularly effective. This approach allows for precise tuning of the electronic structure, significantly enhancing the material's catalytic and electronic performance. Unlike surface charge transfer doping, which is often reversible and unstable, substitutional doping establishes strong covalent bonds between the dopant and host atoms, resulting in chemically robust and thermally stable modifications. The dopants become an integral part of the crystal lattice, resisting diffusion or loss even under demanding operational conditions. As supported by prior studies,^256,257 this intrinsic stability ensures the long-term preservation of the doped material's desired properties, making it a promising route for engineering high-performance WS₂-based electrocatalysts.²⁵⁸ Amirulloieva, N. et al. presented a schematic of the synthesis strategy for fabricating WTe₂ nanowires and nanoribbons as shown in Fig. 10(b), emphasizing the critical role of precursor substrate positioning within the reaction setup. The outcome of the synthesis is highly sensitive to the spatial arrangement of the WO₃ precursor relative to the Te source. To selectively produce WTe₂ nanoribbons, the WO₃ thin film must be placed directly above the Te boat, facing the tellurium powder. This close proximity facilitates directional Te vapor transport and localized reaction kinetics necessary for ribbon growth. In contrast, placing the substrate further downstream from the Te source promotes uniform vapor diffusion, enabling the formation of a continuous WTe₂ thin film. This tuneable configuration highlights a simple yet effective approach for controlling morphology during the synthesis of WTe₂ nanostructures essential for tailoring their properties for nano electronic and electrocatalytic applications.²⁵⁹ The MoTe₂/rGO composite was synthesized via a hydrothermal method, as illustrated in Fig. 10(c), offering a simple and scalable approach for fabricating hybrid 2D materials with enhanced electrochemical performance. To begin, 0.2 g of MoTe₂ powder was dispersed in 30 mL of deionized water under continuous stirring to ensure a homogeneous suspension. Separately, 0.2 g of reduced graphene oxide (rGO) was added to the solution and thoroughly mixed for 10 minutes to facilitate uniform integration of the two components. The resulting mixture was transferred into a Teflon-lined autoclave and subjected to hydrothermal treatment at 180 °C for 12 hours, promoting the interfacial interaction between MoTe₂ and rGO sheets. After naturally cooling to room temperature, the composite was washed multiple times, three times with DI water and once with ethanol using centrifugation to remove impurities. The final product was then dried at 70 °C, yielding the MoTe₂/rGO composite in powder form. This hydrothermal approach enables strong interfacial bonding and improved dispersion of MoTe₂ nanostructures on rGO, which are essential for applications in energy storage and electrocatalysis, particularly due to enhanced conductivity, stability, and active surface area imparted by the rGO support.²⁶⁰ The morphological and microstructural characteristics of both undoped WS₂ and Al-doped WS₂ (0.04 at%) were examined using scanning electron microscopy (SEM). The SEM images revealed flake-like WS₂ structures exhibiting noticeable surface corrugations as shown in Fig. 10(d), suggesting the presence of strain or growth-induced wrinkles.²⁵⁸ Notably, nanoribbon features were observed forming in circular domains as shown in Fig. 10(e), which are believed to originate from defect sites in the WO₃ precursor film. These nanoribbon-containing regions appeared consistently across all 20 thin film samples analysed, with yields ranging from 2% to 10% of the total film area. Interestingly, the yield improved significantly up to 15–20% coverage when the as-sputtered WO₃ films were subjected to a short preheating step at 600 °C in an inert atmosphere for 10 minutes prior to tellurization. This thermal treatment increases the surface roughness, as reflected by an increase in the root mean square (RMS) roughness (Sq) from ∼2 nm in as-sputtered films to ∼5–9 nm in preheated films, promoting nanoribbon formation.²⁵⁹ Additionally, SEM analysis of the MoTe₂/rGO composite revealed a nanoparticle-rich, agglomerated morphology as shown in Fig. 10(f), indicating successful integration of MoTe₂ onto the rGO matrix. The increased surface complexity and structural diversity observed in these materials are crucial for enhancing surface reactivity, which is beneficial for applications in catalysis and energy storage.²⁶⁰ Alahmadim, M. et al. explored the role of aluminium doping in enhancing the oxygen evolution reaction (OER) performance of WS₂; a series of Al-doped WS₂ electrocatalysts were synthesized with varying Al precursor concentrations and evaluated under alkaline conditions (1 M KOH) using screen-printed carbon electrodes (SPCE). The catalytic activity was assessed through linear sweep voltammetry (LSV) at room temperature. As shown in Fig. 10(g), undoped WS₂ (WS₂/SPCE) displayed limited catalytic activity toward the OER. In contrast, doping with aluminium, particularly at 0.04 at% (0.04% Al-WS₂/SPCE), resulted in a marked improvement, achieving a significantly lower overpotential of 0.65 V at a current density of 6 mA cm⁻². This enhancement is attributed to an increase in electroactive sites and improved charge transfer, likely due to aluminium-induced modulation of the WS₂ structure, which inherently possesses limited edge exposure and low active surface area. To gain further insight into the reaction kinetics, Tafel slope analysis was performed. The undoped WS₂ exhibited a high Tafel slope of 363 mV dec⁻¹, in Fig. 10(h), indicating sluggish OER kinetics. Upon Al doping, the Tafel slope decreased to 281 mV dec⁻¹ for 0.02% Al-WS₂ and further to 227 mV dec⁻¹ for 0.04% Al-WS₂, demonstrating a substantial improvement in catalytic efficiency with increasing dopant concentration. The trend clearly suggests that even minimal Al incorporation into the WS₂ matrix significantly accelerates OER kinetics, likely by enhancing electronic conductivity and optimizing adsorption energies of intermediate species. These findings highlight Al doping as a promising strategy to boost the OER activity of WS₂-based electrocatalysts, paving the way for more efficient and cost-effective water-splitting technologies.²⁵⁸ To evaluate the influence of tellurium (Te) vacancies on the electrocatalytic performance of WTe₂, both pristine and vacuum-annealed single crystals were examined under electrochemical conditions by Kwon, H. et al., as depicted in Fig. 10i; linear sweep voltammetry (LSV) revealed a slight improvement in catalytic activity following annealing: the overpotential at 1.0 mA cm⁻² decreased from −0.53 V for the pristine WTe₂ to −0.40 V for the Te-deficient sample, indicating that Te vacancies marginally enhance hydrogen evolution performance. However, the Tafel slope values, shown in Fig. 10(j), remained relatively unchanged, 154 mV dec⁻¹ for the pristine crystal and 159 mV dec⁻¹ for the Te-deficient one, suggesting that the presence of Te vacancies has little to no effect on the intrinsic reaction kinetics. The persistently high Tafel slopes in both cases imply limited hydrogen conversion rates at the catalytically active sites, despite the altered surface composition. Overall, while vacuum annealing introduces Te vacancies and slightly reduces the onset potential, this indicates that Te vacancies alone are not sufficient to substantially enhance the electrocatalytic activity of WTe₂, highlighting the need for additional structural or compositional modifications to achieve meaningful catalytic improvements.²⁶¹ The electrocatalytic performance of the MoTe₂/rGO composite toward the oxygen evolution reaction (OER) was systematically evaluated using cyclic voltammetry (CV) and linear sweep voltammetry (LSV) at a scan rate of 5 mV s⁻¹ under alkaline conditions. These electrochemical techniques were employed to probe the composite's conductivity, overpotential (η), current density (j), onset potential, and overall reaction kinetics. As shown in Fig. 10(k), the MoTe₂/rGO composite exhibited an onset potential of 1.43 V, indicating its catalytic activity initiates at relatively low energy input. The Tafel slope, a key indicator of reaction kinetics shown in Fig. 10(l), was significantly reduced in the composite (36 mV dec⁻¹) compared to pristine MoTe₂ (53 mV dec⁻¹). This reduction implies that the composite enables faster charge transfer and more efficient catalytic turnover during the OER. The enhanced performance can be attributed to the synergistic interaction between MoTe₂ and reduced graphene oxide (rGO), where the covalent bonding between the two components improves electron mobility, increases the density of active sites, and facilitates better electrolyte catalyst interface interactions. The observed linear relationship between the Tafel slope and log (current density) further supports improved reaction kinetics.²⁶⁰

Fig. 10 (a–c) The synthetic method's schematic illustration of WS₂, WTe₂ and MoTe₂. (d–f) SEM images of WS₂, WTe₂ and MoTe₂. (g and h) LSV polarization curves and corresponding Tafel plots of Al-doped WS₂ and undoped WS₂ on SPCE electrodes measured in alkaline solution. (i and j) Electrocatalytic performance of pristine and annealed WTe₂ single crystals: linear sweep voltammetry with overpotentials at 1 mA cm⁻² (inset), Tafel plots, and optical image of the tested crystal. (k and l) LSV curves and Tafel slopes for MoTe₂ and MoTe₂/rGO.^258–261 Copyright 2024, Elsevier. Copyright 2024, AIP Publishing LLC. Copyright 2024, Elsevier. Copyright 2020, Elsevier.

3.7 Improvement of edge site

The improvement of edge sites of 2D TMDs has become a key strategy for enhancing their electrocatalytic performance in the OER/ORR. Unlike the relatively inert basal planes, edge sites in TMDs exhibit unique coordination environments, electronic structures, and defect densities, making them highly active for electrocatalysis.^135,262 Balasubramanyam, S. et al. explained that the morphology of WS₂ thin films evolves significantly with increasing thickness, as revealed by cross-sectional STEM analysis shown in Fig. 11(a). In the early stages of growth (t_App ∼ 32 nm), the WS₂ layers adopt a well-ordered two-dimensional (2D) lateral orientation parallel to the substrate surface. However, as the film grows thicker, a transition occurs: the subsequent layers begin to tilt and stack at oblique angles, forming three-dimensional (3D) features. These tilted, densely packed layers are primarily observed to terminate at the film's top surface highlighting a morphological shift from planar 2D growth to a more vertically inclined 3D architecture. The electrocatalytic performance of WS₂ thin films was systematically evaluated using linear sweep voltammetry (LSV), with a bare glassy carbon substrate serving as the control as shown in Fig. 11(b). The WS₂ films, of comparable thickness, demonstrated a notably lower onset potential (∼−200 mV) than the bare substrate, indicating enhanced catalytic activity. Notably, films synthesized using the H₂S-only process exhibited higher cathodic current densities than those grown under combined H₂ + H₂S atmospheres, suggesting superior catalytic performance when active site density is maximized. This performance enhancement correlates closely with the density of catalytically active sites, which was effectively tuned through plasma processing conditions during synthesis. The extracted Tafel slopes shown in Fig. 11(c) further support this, as the lower slopes observed imply faster reaction kinetics, an important parameter for efficient catalysis at increased overpotentials. The long-term operational stability of WS₂ films (∼8 nm thick) was assessed under continuous electrolysis. The overpotential required to sustain a current density of 5 mA cm⁻² (η₅) remained consistent over 25 hours, confirming the durability and robustness of the H₂S-processed WS₂ electrocatalyst.²⁶³ Peng and his co-workers provided spectroscopic evidence of quantum spin Hall edge states in monolayer WTe₂. To probe the edge electronic structure, spatially resolved scanning tunneling spectroscopy (STS) was performed across step edges, as shown in Fig. 11(d). The dI/dV spectra revealed a notable enhancement in the local density of states (LDOS) at the monolayer edge, a hallmark of quantum spin Hall (QSH) edge states. This spectral feature, with a spatial extent of approximately 1.2 nm, is consistent with previous STS studies of epitaxially grown WTe₂ on graphene substrates, confirming its robustness across different sample preparations.^264,265 Despite the different synthesis route used here, mechanical exfoliation from bulk crystals rather than molecular beam epitaxy and the use of superconducting rather than graphene substrates, the characteristic spectroscopic signature of the QSH state persists. Fig. 11(e) compares the dI/dV spectra taken on the interior of the monolayer (red) and at the edge (orange), clearly illustrating the presence of metallic edge states within the nominal bandgap. Meanwhile a residual dI/dV signal is observed within the bandgap away from the step edge, this is attributed to substrate-induced effects and possibly to defect states, as discussed later in the article. These observations, in agreement with prior ARPES and STS studies, support the existence of a ∼56 meV bandgap and confirm the presence of QSH edge states in monolayer WTe₂, reinforcing its role as a prototypical 2D topological insulator. Lüpke et al. investigated the coexistence of superconductivity and topological edge states in monolayer WTe₂. To study their spatial evolution, particularly near the quantum spin Hall (QSH) edge, scanning tunneling spectroscopy (STS) was carried out at 2.8 K, as shown in Fig. 11(f). The dI/dV spectra taken across a line approaching the physical edge of the WTe₂ monolayer reveal that the superconducting gap remains present throughout the monolayer, with only subtle variations in its width and depth. The spectral analysis was carried out by fitting the experimental spectra to a combination of BCS superconducting models for both WTe₂ and the underlying NbSe₂ substrate as shown in Fig. 11(g). These fits confirm that the superconducting gap persists even in the region associated with the QSH edge state. Notably, the superconducting signature overlaps spatially with the edge state feature, suggesting a coexistence of topological and superconducting phenomena. To determine the contribution of the edge state region to the overall spectrum, a similar fitting approach was applied near the monolayer edge shown in Fig. 11(h), using the same NbSe₂ background determined from the edge as shown in Fig. 11(g). This approach highlights the consistent presence of superconductivity, reinforcing the prospect of topological superconductivity emerging at the WTe₂ edge.²⁶⁶ To further assess the physical characteristics of the synthesized MoTe₂ polymorphs, atomic force microscopy (AFM) was employed, as shown in Fig. 11(i–l). The analysis revealed a notable difference in thickness between the two phases: the 1T′-MoTe₂ nanosheets measured approximately 8 nm as shown in Fig. 11(k), whereas the 2H-MoTe₂ counterparts exhibited a thickness of around 25 nm as shown in Fig. 11(l), making them roughly three times thicker than the 1T′ phase. AFM imaging also highlighted distinctions in surface uniformity. The 1T′-MoTe₂ samples displayed a more uniform and smooth morphology, while the 2H-MoTe₂ exhibited increased surface variation, as evident in Fig. 11(i) and (j). These observations suggest that the 1T′ phase forms with greater structural homogeneity, which may be advantageous for applications where consistent film quality and thickness are critical, such as in optoelectronics or electrocatalysis.²⁶⁷

Fig. 11 (a) Cross-sectional images of WS₂ grown via the H₂S process highlight edge terminations, (b) with electrocatalytic performance assessed by IR-corrected LSV curves and (c) corresponding Tafel plots.²⁶³ Copyright 2024, American Chemical Society. (d–h) Spatially resolved dI/dV spectra across a WTe₂ step edge, along with the corresponding height profile, reveal an enhanced local density of states (LDOS) at the monolayer edge, attributed to the quantum spin Hall (QSH) edge state. Spectra collected perpendicular to the edge confirm a bandgap in the monolayer region. Tunnelling measurements on monolayer WTe₂ atop NbSe₂ were analysed using a two-channel model, separating the spectral contributions of each material. The induced superconducting gap in the WTe₂ monolayer and at its QSH edge was extracted, with edge regions exhibiting stronger tunnelling signals, indicating a more pronounced proximity effect.²⁶⁶ Copyright 2020, Nature Publishing. (i–l) AFM analysis of the as-grown multilayer MoTe₂ films shows topographical images and corresponding height profiles for both 1T′-MoTe₂ and 2H-MoTe₂ phases, highlighting their surface morphology and thickness variations.²⁶⁷ Copyright 2019, Elsevier.

3.8 Superlattice engineering

Superlattice engineering has emerged as a transformative approach for enhancing the electrocatalytic properties of 2D TMDs for the OER/ORR. By structuring these materials into periodic superlattices, the electronic structure, ion transport behaviour, and charge transfer efficiency can be precisely modulated to achieve optimal catalytic performance.^268,269Fig. 12(a–c) illustrates a representative laterally stacked 2D superlattice, formed by the alternate in-plane epitaxial growth of material A surrounding material B within the same monolayer plane. This strategy effectively preserves material integrity and offers a robust, generalizable method for growing diverse heterostructures, multi heterostructures, and superlattices.²⁷⁰ The electrochemical performance of NiO@Ni/WS₂/CC was evaluated alongside WS₂/CC, NiO@Ni/CC, and commercial Pt/C (20 wt% Pt/XC-72) on carbon cloth in 1.0 M KOH at a scan rate of 5 mV s⁻¹ as shown in Fig. 12(d). To ensure accurate comparison, iR-correction was applied to all data based on EIS measurements. Tafel slope analysis shown in Fig. 12(e) revealed that NiO@Ni/WS₂/CC exhibited a much lower slope (83.1 mV dec⁻¹) compared to WS₂/CC (234.9 mV dec⁻¹) and NiO@Ni/CC (136.4 mV dec⁻¹), indicating markedly improved electrocatalytic kinetics. For reference, Pt/C/CC showed a slope of 43 mV dec⁻¹. The electrochemical stability of NiO@Ni/WS₂/CC was further verified through cyclic voltammetry cycling.²⁷¹ Kang, K., et al. introduced moiré potentials into bilayer WTe₂ by stacking a monolayer of WSe₂ in the H-phase structure, as illustrated in Fig. 12(f). In this configuration, the tungsten atoms in WSe₂ are coordinated trigonal prismatically between selenium atoms. The WSe₂ zigzag crystal axis was aligned with the a-axis of WTe₂ to create a nearly cantered-rectangular moiré superlattice. Due to lattice mismatches of approximately 5% and 9% along the a and b axes, periodicities of ∼6.9 nm and ∼6.4 nm are expected, respectively, yielding a moiré density (n_m) of ∼4.5 × 10¹² cm⁻² as shown in Fig. 12(g). These theoretical estimates are supported by experimental observations using piezo response force microscopy (PFM), which confirm the moiré pattern and the ∼84° angle between the moiré lattice vectors. Unlike the triangular or honeycomb moiré lattices commonly reported, this system forms a robust, twist-angle-insensitive rectangular moiré lattice, an effect attributed to the large lattice mismatch. The WTe₂/WSe₂ heterostructure exhibits type-I band alignment, with both the conduction band minimum and valence band maximum residing in bilayer WTe₂, while the WSe₂ layer remains charge-neutral, acting solely as a periodic moiré potential that flattens the WTe₂ bands. As shown in Fig. 12(h), zone folding occurs in the band structure, and the top WTe₂ layer directly interfaced with WSe₂ experiences a stronger moiré potential, leading to significantly flatter bands compared to the bottom layer. This layer-specific moiré modulation, coupled with the intrinsic ferroelectricity of WTe₂, underpins the ability to electrically tune the moiré potential via ferroelectric switching. The heterostructure is encapsulated between boron nitride layers and top/bottom graphite gates, enabling independent control over both the charge carrier density (ν, in units of n_m) and the perpendicular electric field (E), where positive E is defined from WTe₂ toward WSe₂.²⁷² The MoS₂/WTe₂ hybrid catalyst demonstrates significantly enhanced electrocatalytic performance compared to individual MoS₂ nanosheets and WTe₂ flakes shown in Fig. 12(j). It exhibits a low onset potential of −50 mV and an overpotential of −140 mV at 10 mA cm⁻², while carbon fibre paper used as a control shows negligible activity within this potential window. This improvement is attributed to facilitated charge transfer from the WTe₂ substrate into the MoS₂ nanosheets, enhancing the intrinsic activity of MoS₂, which has already been shown to outperform WTe₂ alone. Tafel analysis shown in Fig. 12(k) further supports this enhancement: the MoS₂/WTe₂ hybrid displays a Tafel slope of ∼40 mV dec⁻¹, notably lower than WTe₂ flakes (66 mV dec⁻¹) and MoS₂ nanosheets (112 mV dec⁻¹). This slope aligns well with the Volmer–Heyrovsky mechanism, indicating that the rate-determining step likely involves electrochemical desorption.²⁷³ Saruta, Y. et al. observed that () R30° superstructure is most plausibly attributed to moiré potentials arising from the stacking of monolayer MoTe₂ on bilayer graphene (BLG), as illustrated in Fig. 12. While moiré patterns are commonly known from twisted bilayer graphene systems where a small twist angle (≈1°) between layers of identical lattice constants leads to band folding and emergent phenomena such as superconductivity here, the effect emerges in a hetero-bilayer system with lattice mismatch.²⁷⁴ Specifically, the lattice constants of graphene (∼2.46 Å) and MoTe₂ (∼3.4 Å) differ significantly. Despite this incommensurability, a clear moiré pattern forms when MoTe₂ is stacked at a 30° rotation relative to the graphene lattice.²⁷⁵ As shown in Fig. 12(m), no moiré pattern is observed at 0° due to the mismatch, whereas a distinct () R30° periodicity emerges at 30°, matching well with the STM observations. This near-commensurate stacking condition enables the formation of a moiré superlattice even between lattice-mismatched materials. Angle-resolved photoemission spectroscopy (ARPES) measurements confirm that the monolayer MoTe₂ is indeed rotated by 30° with respect to the underlying bilayer graphene. Together, these findings strongly support the conclusion that the () R30° superstructure originates from a moiré potential induced by this specific rotational stacking.²⁷⁶ The electrocatalytic activity of MoTe₂ nanostructures was systematically evaluated and benchmarked against commercial Pt/C using linear sweep voltammetry (LSV) in 0.5 M H₂SO₄ shown in Fig. 12(n). Among the tested morphologies, MoTe₂ nanotubes exhibited superior HER performance, showing notably low overpotentials of −317 mV and −349 mV at current densities of −10 mA cm⁻² and −100 mA cm⁻², respectively. These values were significantly lower than those of MoTe₂ nanoparticles and nanowires, as previously reported by Mao, J. et al.²⁷⁷ The nanotubes were synthesized with varying amounts of VC, and their LSV profiles demonstrated a sharp rise in current density with increasing applied potential highlighting improved catalytic efficiency. Tafel slope analysis shown in Fig. 12(o) further confirmed the enhanced reaction kinetics of the MoTe₂ nanotubes, which displayed a slope of 54 mV dec⁻¹ outperforming even the commercial Pt/C catalyst (61 mV dec⁻¹). In comparison, MoTe₂ nanoparticles exhibited a higher slope of 67 mV dec⁻¹. These results suggest that the catalytic activity on MoTe₂ nanotubes likely follows the Volmer–Heyrovsky mechanism, consistent with earlier mechanistic interpretations (Watzele et al.).^278,279

Fig. 12 (a) Schematic of 2D superlattice growth and (b) optical image of a WS₂–WSe₂–WS₂–WSe₂ superlattice on a SiO₂/Si substrate. (c) Raman mapping at 250 cm⁻¹ (WS₂) and 350 cm⁻¹ (WSe₂) distinctly reveals the alternating layered structure of the superlattice.^268,270 Copyright 2022, Wiley. Copyright 2017, Science Publishing. (d) LSV curves and (e) Tafel plots comparing catalytic performance of NiO@Ni/WS₂/CC, WS₂/CC, Pt/C/CC, and NiO@Ni/CC.²⁷¹ Copyright 2018, American Chemical Society. Moiré superlattice in bilayer T_d-WTe₂/monolayer H-WSe₂ heterostructure: (f) lattice model showing high symmetry MM sites; (g) PFM image and Fourier transform confirming moiré periodicity; (h) schematic band structure with moiré induced modulation; (i) polarization-dependent hole distributions and internal electric field in bilayer WTe₂.²⁷² Copyright 2023, Nature Publishing. Electrocatalytic performance of MoS₂/WTe₂ hybrid: (j) polarization curves and (k) Tafel plots compared with individual MoS₂ and WTe₂ components.²⁷³ Copyright 2019, Wiley. (l) Moiré superstructure of 1T-MoTe₂/graphene heterostructure at 0° and (m) 30° rotation showing () R30° periodicity. Catalytic performance of MoTe₂ nanocrystals in 0.5 M H₂SO₄.²⁷⁶ Copyright 2023, Wiley. (n) LSV curves and (o) Tafel plots with CV scan rates (inset).²⁷⁹ Copyright 2022, Frontiers Publishing.

3.9 Acquiring machine learning

The acquisition of ML techniques for the study and design of 2D TMDs has marked a significant shift in the field of electrocatalysis, particularly for the OER/ORR.^280,281 The integration of two-dimensional (2D) materials in electrochemical energy storage systems including fuel cells, batteries, and supercapacitors has garnered significant attention in recent studies. The KDD (knowledge discovery in databases) approach has been widely applied to optimize these materials for energy applications. Fig. 13(a) provides a comprehensive overview highlighting the roles, advantages, and key parameters of 2D materials when used as electrode materials, along with the focus areas where machine learning techniques have been employed to accelerate material discovery and performance optimization.²⁸² Han, B. et al. conducted the quantitative assessment of material properties, and two key predictors linked to the bandgap and crystal structure were selected to evaluate the output of 16 independently trained 2DMOINet models. As shown in Fig. 13(b) (1) and (b) (2), the projected mean values and their standard deviations for each material were plotted as histograms with error bars, providing statistical insight into how well the model captures underlying physical characteristics. The projected values exhibited a clear correlation with actual experimental parameters: Fig. 13(b) (1) reveals alignment with bandgap values, while Fig. 13(b) (2) demonstrates agreement with crystal structure classifications. These projections can therefore serve as reliable probabilistic indicators for predicting the bandgap or structure class of unknown 2D materials. While a few misclassifications were noted, their presence suggests opportunities for improvement through either more advanced model architectures or larger, more diverse training datasets including more images and material types. Moreover, this framework offers the potential for systematic investigation of other material-influencing factors, such as mechanical exfoliation techniques and bulk crystal quality, further enhancing its utility in accelerated materials discovery and property prediction.²⁸³ To overcome the high computational cost of screening all transition metal single-atom catalysts (SACs) for OER activity using pure DFT, a topological information-based machine learning (ML) model was developed to predict the OER overpotentials of TM-SACs efficiently shown in Fig. 13(c). This ML approach consists of three main steps: step 1: a training dataset of 15 TM-based single-atom catalysts (SACs) was built using DFT-calculated OER overpotentials. Each catalyst was represented using atomic-scale descriptors including atomic mass, radius, d-electron count, electronegativity, electron affinity, and ionization energy. These served as input features for the ML model.²⁸⁴ Step 2: a graph-based ML model was developed, encoding catalyst structures as atomic networks. The model architecture included six convolutional layers and three fully connected layers. It was trained and tested on randomly split datasets (26 training, 4 testing) using PyTorch. The final model achieved high accuracy, predicting overpotentials with only ∼6.5–6.7% deviation from DFT, see in Fig. 13(e). Step 3: the trained ML model was applied to predict overpotentials for a broader set of SACs. It delivered results comparable to DFT but at a much lower computational cost, enabling fast and scalable screening of OER catalysts.^285,286Fig. 13(d) demonstrates the strong agreement between machine learning (ML) predictions and density functional theory (DFT) results for oxygen evolution reaction (OER) overpotentials across 14 single-atom catalyst (SAC) systems. Specifically, the ML model achieved a mean relative error of just 6.70% on the training set and 6.49% on the test set, underscoring the accuracy of the topological information-based algorithm. To further validate the model, additional SACs (Rh-SV, Rh-DV, W-DV, W-SV, Hf-SV, and Hf-DV) were evaluated, with ML prediction errors ranging from 0.54% to 12.84%, all remaining within an acceptable range for catalyst screening. A critical advantage of the ML model lies in its massive acceleration of prediction speed. While a single DFT calculation of OER overpotential requires ∼36 CPU hours on a 40-core machine (∼1445 hours on a single core), the ML model completes the same prediction in ∼40 seconds on a single core making it over 130 000 times faster, see in Fig. 13(e). This efficiency enables rapid screening of large catalyst libraries, overcoming the bottleneck of conventional DFT-based methods.²⁸⁷ In Sun, Y. et al., as shown in Fig. 13(f), to complement DFT results, machine learning (ML) methods were employed to predict structural and catalytic properties, using Python 3.0 and the Scikit-learn package.²⁸⁸ Five supervised learning algorithms were explored: Gradient Boosting Regression (GBR),²⁸⁹ random forest regression (RFR),²⁹⁰ support vector regression (SVR),²⁹¹K-nearest neighbour regression (KNR),²⁹² and Gaussian process regression (GPR)²⁹³ which are well-established for analysing OER and ORR activities in 2D materials.^294,295

Fig. 13 (a) Infographic showing the roles, mechanisms, benefits, and ML optimization targets of 2D materials in electrochemical energy storage.²⁸² Copyright 2025, American Chemical Society. Ensemble prediction of material properties using 16 trained 2DMOINets: (b1) projected bandgap values and (b2) crystal structure classifications for training and unseen materials, with error bars indicating model uncertainty.²⁸³ Copyright 2020, Wiley. (c) Schematic of ML training strategy: (c1) data generation, (c2) topology-based model training/testing, and (c3) overpotential prediction workflow. (d) ML strategy for OER overpotential prediction: DFT vs. ML-predicted values, (e) comparison of average computational costs for predicting catalytic activity of transition metal single-atom catalysts (SACs) using pure DFT calculations versus machine learning (ML) models.²⁸⁷ Copyright 2021, Cell Press Publishing. (f) Overview map highlighting the catalytic performance and reaction mechanisms of TM–N₄–graphene/XS₂ catalysts for the ORR and OER. Studies using DFT are marked in green, while those employing machine learning (ML) approaches are shown in pink.²⁹⁵ Copyright 2021, American Chemical Society.

4 Conclusions

By combining experimental techniques and computational predictions, this review highlights the transformative potential of TMDs (WS₂, WTe₂, MoTe₂) as electrocatalysts for the OER/ORR due to their sustainable energy applications and provides a roadmap for future research aimed at overcoming existing challenges. However, challenges like scalability, phase stability, long-term durability, and defect control remain significant hurdles in their commercial application. Defect engineering, heterostructure creation, and doping are promising approaches for enhancing catalytic performance, but further optimization is required for industrial application.

4.1 Comparative discussion with emerging 2D electrocatalysts

(i) Emerging 2D electrocatalysts such as MXenes, MBenes, and silica-based materials show comparable electrocatalytic performance. However, TMDs are more tunable and have greater flexibility in electronic properties, making them advantageous for optoelectronic applications. (ii) MXenes are known for their excellent conductivity and hydrophilicity, making them efficient in OER/ORR electrocatalysis. Their ability to undergo surface functionalization is a potential advantage over TMDs in some applications. (iii) MBenes offer high chemical stability, making them competitive for energy storage and electrocatalysis under harsh conditions. (iv) Silica-based electrocatalysts offer a high surface area but may lack the electronic tunability of TMDs, which are vital for specific electrocatalytic applications such as the OER/ORR.^296–298

4.2 Research gaps

(a) Phase control and stabilization: more research is needed to stabilize metastable phases, such as 1T′ MoTe₂ and T_d-WTe₂, under practical reaction conditions to ensure high performance and reproducibility: (b) defect engineering: defect-induced electrocatalytic activity needs further investigation, especially in terms of dopants and vacancies that can tune reaction pathways for the OER/ORR. (c) Understanding atomic-level mechanisms: atomic-scale simulations (e.g., DFT) can help unravel the reaction intermediates and rate-limiting steps for electrocatalysis to optimize material design for enhanced efficiency. (d) Surface modifications and doping: doping strategies, especially heteroatom doping in TMDs, need to be optimized for enhanced active site density and improved catalytic activity. (e) Computational insights: further use of machine learning (ML) and DFT to design new electrocatalysts, predict optimal synthesis conditions, and tailor properties for real-world applications is essential.

4.3 Proposed strategies for industrial-scale implementation

(i) Enhanced synthesis: CVD and hydrothermal synthesis techniques should be optimized to produce high-quality, defect-engineered TMDs on a large scale. (ii) Stabilization of active phases: strategies like composite formation, surface coatings, and heterostructures should be explored to stabilize the active phases of TMDs under operational conditions. (iii) Surface passivation and functionalization: techniques such as plasma treatment and surface functionalization can improve TMD stability and electrocatalytic performance under real-world conditions. (iv) Cost-effective production: more affordable solution-phase synthesis methods could reduce the cost of TMD-based electrocatalysts and make them more viable for large-scale industrial applications.

4.4 Current challenges

1. Scalability: large-scale synthesis of monolayer or few-layer TMDs remains a major challenge, especially with maintaining high-quality monolayers and uniform defect densities across large batches.

2. Phase stability: TMDs undergo phase transitions (e.g., from 2H to 1T′ and T_d), which may affect electrocatalytic activity and stability during real-world applications.

3. Long-term durability: degradation of TMDs under high potential or acidic conditions during OER/ORR processes is a critical issue for real-world applicability.

4. Lack of reproducibility: there is limited reproducibility in the synthesis of TMDs with consistent properties at a large scale, which can hinder their mass production.

5. Limited active sites: the basal planes of TMDs are often chemically inert, with active sites typically found at defects or edges, which limits their electrocatalytic efficiency.

6. Complex synthesis: the synthesis of TMDs requires precise control over factors such as temperature, precursor concentration, and substrate to produce materials with the desired properties.

7. Tuning electronic properties: the electronic properties of TMDs need to be fine-tuned for specific applications (e.g., OER/ORR), but control over doping and defect density is still challenging.

8. Phase control: achieving phase control during synthesis is difficult, and metastable phases (such as 1T′) often degrade into more stable phases that may not exhibit the desired catalytic performance.

4.5 Future perspectives

Several key areas present the advancements in TMD 2D materials promising breakthroughs in electrocatalysis, energy storage, and quantum technologies, driven by enhanced synthesis methods, defect engineering, and scalable production.

1. Advanced synthesis techniques: plasma-enhanced CVD and solution-based synthesis methods should be explored for mass production and high-quality material synthesis.

2. Defect engineering: investigating defect engineering and dopant optimization will play a pivotal role in improving the electrocatalytic efficiency of TMDs.

3. Surface modifications: surface passivation and functionalization methods (e.g., doping with transition metals, nonmetals) can help tune electronic structures and enhance electrocatalytic activity.

4. Machine learning for catalyst design: machine learning can accelerate the design of TMD electrocatalysts by predicting optimal dopant configurations, defect distributions, and reaction pathways for the OER/ORR.

5. Industrial-scale production: developing scalable production methods, such as wet-chemical methods and roll-to-roll fabrication, will be essential for large-scale deployment.

6. Thermal and structural stability: new strategies to improve thermal stability and phase consistency of TMDs under operating conditions should be prioritized.

7. Catalyst-heterostructures: heterostructure fabrication involving TMDs and other materials (e.g., graphene, MXenes) can create synergistic effects, improving electrocatalytic properties.

8. Quantum and optoelectronic applications: beyond electrocatalysis, TMDs should be explored for quantum applications and optoelectronics, utilizing their light–matter interaction and unique electronic properties for next-generation technologies.

9. Optimization of electrolyzers: it remains a crucial avenue for improving the efficiency of OER/ORR electrocatalysis. Future efforts should focus on enhancing the catalytic activity and stability of materials, including the use of TMDs, defect engineering, and doping strategies. These advancements can potentially reduce overpotentials, improve energy efficiency, and contribute to the scalability of electrolyzers for widespread energy applications. Additionally, the integration of computational models with experimental findings can guide the design of next-generation electrocatalysts, addressing the current challenges in commercializing electrolyzer technology.

Conflicts of interest

There are no conflicts to declare.

Data availability

Data will be made available on request.

Acknowledgements

This work acknowledges the funding support from the State Key Laboratory of Urban-rural Water Resources & Environment (Harbin Institute of Technology) (No. 2025TS36), Guangdong Basic and Applied Basic Research Foundation (2023A1515011332), and Shenzhen Science and Technology Program (GXWD20231130100010001).

References

1. Y. Tian, L. Xu, J. Qiu, X. Liu and S. Zhang, Sustainable Mater. Technol., 2020, 25, e00204–e00217.
2. A. Vazhayil, L. Vazhayal, J. Thomas, C. S. Ashok and N. Thomas, Appl. Surf. Sci. Adv., 2021, 6, 100184.
3. H. Araújo, B. Šljukić, S. Gago and D. M. F. Santos, Front. Energy Res., 2024, 12, 1373522–1373551.
4. S. Piracha, Y. Zhang, A. Raza and G. Li, Chem. Commun., 2024, 60, 9918–9929.
5. M. Yang, J. Ding, Z. Wang, J. Zhang, Z. Peng and X. Liu, Chin. Chem. Lett., 2025, 36, 1108661–1108667.
6. R. Yin, Z. Wang, J. Zhang, W. Liu, J. He, G. Hu and X. Liu, Small Methods, 2025, 9, 2401976–2401983.
7. H. Li, L. Zhang, W. Liu, T. Gui, L. Xiao, Q. Liu, J. Ding, G. Hu and X. Liu, Mater. Chem. Front., 2025, 9, 2658–2667.
8. S. Aralekallu, K. S. Lokesh and V. Singh, Fuel, 2024, 357, 129753–129775.
9. R. D. A. Bostami, A. A. Othman, M. Tawalbeh and A. G. Olabi, Energy Nexus, 2025, 17, 100387–100413.
10. K. J. Jenewein, G. D. Akkoc, A. Kormányos and S. Cherevko, Chem Catal., 2022, 2, 2778–2794.
11. A. Merchant, S. Batzner, S. S. Schoenholz, M. Aykol, G. Cheon and E. D. Cubuk, Nature, 2023, 624, 80–85.
12. Y. Sim, Y. Chae and S. Y. Kwon, iScience, 2022, 25, 105098–105126.
13. B. Saeloo, T. Saisopa, P. Chavalekvirat, P. Iamprasertkun, K. Jitapunkul, W. Sirisaksoontorn, T. R. Lee and W. Hirunpinyopas, Inorg. Chem., 2024, 63, 18750–18762.
14. J. H. Lim, K. Kim, J. H. Kang, K. C. Kwon and P. H. W. Jang, ChemElectroChem, 2024, 11, e202300614–e202300640.
15. A. Eftekhari, J. Mater. Chem. A, 2017, 5, 18299–18325.
16. Z. Wang, J. Sun, H. Wang, Y. Lei, Y. Xie, G. Wang, Y. Zhao, X. Li, H. Xu, X. Yang, L. Feng and X. Ma, Appl. Surf. Sci., 2020, 504, 144371–144379.
17. C. R. Fernandez, A. Nieminen, F. Ahmed, J. Pietila, H. Lipsanen and H. Caglayan, ACS Appl. Mater. Interfaces, 2023, 15, 35692–35700.
18. E. M. Tari, R. S. Mamoory and A. Yourdkhani, Phys. B, 2023, 650, 414568.
19. M. Lunardon, T. Kosmala, M. G. Asl, A. V. Krasheninnikov, S. Kolekar, C. Durante, M. Batzill, S. Agnoli and G. Granozzi, ACS Energy Lett., 2023, 8, 972–980.
20. Q. Fu, J. Han, X. Wang, P. Xu, T. Yao, J. Zhong, W. Zhong, S. Liu, T. Gao, Z. Zhang, L. Xu and B. Song, Adv. Mater., 2021, 33, 1907818–1907841.
21. J. Yu, G. Peng, L. Peng, Q. Chen, C. Su, L. Shang and T. Zhang, Green Energy Environ., 2024, 25, 105098–105127.
22. Y. Sim, Y. Chae and S. Y. Kwon, iScience, 2022, 25, 105098–105126.
23. Y. Lei, T. Zhang, Y. C. Lin, T. G. Nakajima, G. Bepete, D. A. Kowalczyk, Z. Lin, D. Zhou, T. F. Schranghamer, A. Dodda, A. Sebastian, Y. Chen, Y. Liu, G. Pourtois, T. J. Kempa, B. Schuler, M. T. Edmonds, S. Y. Quek, U. Wurstbauer, S. M. Wu, N. R. Glavin, S. Das, S. P. Dash, J. M. Redwing, J. A. Robinson and M. Terrones, ACS Nanosci. Au, 2022, 2, 450–485.
24. R. Ding, J. Chen, Y. Chen, J. Liu, Y. Bando and X. Wang, Chem. Soc. Rev., 2024, 53, 11390–11461.
25. B. Wang, D. Liu, Y. Wu, A. S. Vasenko and O. V. Prezhdo, J. Phys. Chem. Lett., 2024, 15, 10384–10391.
26. B. Rotonnelli, M. S. D. Fernandes, F. Bournel, J. J. Gallet and B. L. Kaiser, Curr. Opin. Electrochem., 2023, 40, 101314–101324.
27. J. Wang, C. S. Hsu, T. S. Wu, T. S. Chan, N. T. Suen, J. F. Lee and H. M. Chen, Nat. Commun., 2023, 14, 6576–6598.
28. Y. Gorlin, B. L. Kaiser, J. D. Benck, S. Gulb, S. M. Webb, V. K. Yachandra, J. Yano and T. F. Jaramillo, J. Am. Chem. Soc., 2013, 135, 8525–8534.
29. M. Wang, L. Árnadóttir, Z. J. Xu and Z. Feng, Nano-Micro Lett., 2019, 11, 47–64.
30. X. Jiang, G. Liu, L. Zhang and Z. Hu, Catalysts, 2025, 15, 309–322.
31. Z. Fang, S. Li, Y. Zhang, Y. Wang, K. Meng, C. Huang and S. Sun, J. Phys. Chem. Lett., 2024, 15, 281–289.
32. L. H. Mou, T. Han, P. E. S. Smith, E. Sharman and J. Jiang, Adv. Sci., 2023, 10, 2301020–2301039.
33. Z. Yang, W. Gao and Q. Jiang, J. Mater. Chem. A, 2020, 8, 17507–17515.
34. S. Mehta, R. Thakur, S. Rani, B. M. Nagaraja, S. Mehla and I. Kainthla, Int. J. Hydrogen Energy, 2024, 82, 1061–1080.
35. K. B. Ibrahim, T. A. Shifa, S. Zorzi, M. G. Sendeku, E. Moretti and A. Vomiero, Prog. Mater. Sci., 2024, 144, 101287–101314.
36. S. N. Upadhyay, J. A. K. Satrughna and S. Pakhira, Emergent Mater., 2021, 4, 951–970.
37. H. Hosseini, C. J. Herring, C. F. Nwaokorie, G. A. Sulley and M. M. Montemore, J. Phys. Chem. C, 2024, 128, 18144–18157.
38. L. Ge, H. Yuan, Y. Min, L. Li, S. Chen, L. Xu and W. A. Goddard III, J. Phys. Chem. Lett., 2020, 11, 869–876.
39. M. A. Zaeem, S. Thomas, S. Kavousi, N. Zhang, T. Mukhopadhyay and A. Mahata, 2D Mater., 2024, 11, 042004–042064.
40. Z. Shi, J. Zhang, W. Zeng and Q. Zhou, Langmuir, 2023, 39, 4125–4139.
41. Z. Tan, X. Y. Kong, B. J. Ng, H. S. Soo, A. R. Mohamed and S.-P. Chai, ACS Omega, 2023, 8, 1851–1863.
42. J. H. Kim, H. Sung and G. H. Lee, Small Sci., 2024, 4, 2300093–2300157.
43. S. Sovizi, S. Angizi, S. A. A. Alem, R. Goodarzi, M. R. R. T. Boyuk, H. Ghanbari, R. Szoszkiewicz, A. Simchi and P. Kruse, Chem. Rev., 2023, 123, 13869–13951.
44. J. Wang, F. Cheng, Y. Sun, H. Xu and L. Cao, Phys. Chem. Chem. Phys., 2024, 26, 7899–8012.
45. J. G. DiStefano, A. A. Murthy, H. J. Jung, R. D. Reis and V. P. Dravid, Appl. Phys. Lett., 2021, 118, 223103–223109.
46. H. Chen, J. Zhang, D. Kan, J. He, M. Song, J. Pang, S. Wei and K. Chen, Crystals, 2022, 12, 1381–1400.
47. J. Zeng, Y. Liu, Z. Huang, H. Qiao and X. Qi, Catalysts, 2024, 14, 689–726.
48. X. Yin, Y. Hua and Z. Gao, Renewables, 2023, 1, 190–226.
49. C. Das, N. Sinha and P. Roy, Small, 2022, 18, 2202033–2202100.
50. V. Yadav, S. Roy, P. Singh, Z. Khan and A. Jaiswal, Small, 2018, 15, 1803706–1803738.
51. M. Sedki, Y. Chen and A. Mulchandani, Sensors, 2020, 20, 4811–4831.
52. Y. Wang, J. Xiao, H. Zhu, Y. Li, Y. Alsaid, K. Y. Fong, Y. Zhou, S. Wang, W. Shi, Y. Wang, A. Zettl, E. J. Reed and X. Zhang, Nature, 2017, 550, 487–491.
53. S. Acosta and M. Quintana, Sensors, 2024, 24, 1817–1838.
54. Y. Hu, L. Rogée, W. Wang, L. Zhuang, F. Shi, H. Dong, S. Cai, B. K. Tay and S. P. Lau, Nat. Commun., 2023, 14, 8470–8481.
55. B. Lu, Y. Xia, Y. Ren, M. Xie, L. Zhou, G. Vinai, S. A. Morton, A. T. S. Wee, W. G. V. Wiel, W. Zhang and P. K. J. Wong, Adv. Sci., 2024, 11, 2305277–2305316.
56. X. Duan, C. Wang, A. Pan, R. Yu and X. Duan, Chem. Soc. Rev., 2015, 44, 8859–8876.
57. Y. Jing, B. Liu, X. Zhu, F. Ouyang, J. Sun and Y. Zhou, Nanophotonics, 2020, 9, 1675–1694.
58. K. Ghatak, K. N. Kang, E. H. Yang and D. Datta, Sci. Rep., 2020, 10, 1648–1658.
59. L. S. Banu, K. Balakrishnan, V. Veerapandy, N. Vajeeston and P. Vajeeston, ACS Omega, 2025, 10, 13515–13528.
60. J. L. Hart, L. Bhatt, Y. Zhu, M. G. Han, E. Bianco, S. Li, D. J. Hynek, J. A. Schneeloch, Y. Tao, D. Louca, P. Guo, Y. Zhu, F. Jornada, E. J. Reed, L. F. Kourkoutis and J. J. Cha, Nat. Commun., 2023, 14, 4803–4811.
61. S. D. Lawaniya, S. Kumar, Y. Yu, H. G. Rubahn, Y. K. Mishra and K. Awasthi, Mater. Today Chem., 2023, 29, 101428–101468.
62. M. M. S. Villamayor, S. Husain, R. O. Nuñez, F. O. L. Johansson, R. Lindblad, P. Lourenço, R. Bernard, N. Witkowski, G. Prévot, N. L. A. N. Sorgenfrei, E. Giangrisostomi, A. Föhlisch, P. Svedlindh, A. Lindblad and T. Nyberg, Nanoscale, 2022, 14, 6331–6338.
63. C. Stangel, E. Nikoli and N. Tagmatarchis, Adv. Energy Sustainability Res., 2022, 13, 2200097–2200112.
64. B. A. Hanedar and M. C. Onbaşlı, Phys. Chem. Chem. Phys., 2025, 27, 1809–1818.
65. Z. Lin, X. Yang, J. He, N. Dong and B. Li, Appl. Phys. Rev., 2025, 12, 011301.
66. P. V. Shinde, M. Hussain, E. Moretti and A. Vomiero, SusMat, 2024, 4, e236–e274.
67. Y. Wang and X. Zhang, Micromachines, 2024, 15, 761–770.
68. C. Zhao, M. Hu, J. Qin, B. Xia, C. Liu, S. Wang, D. Guan, Y. Li, H. Zheng, J. Liu and J. Jia, Phys. Rev. Lett., 2020, 125, 046801–046806.
69. J. Yao, H. Zhang, D. Song, J. Zhang and Y. P. Jiang, Appl. Phys. Lett., 2025, 126, 073101.
70. I. Zulfiqar, M. A. Khan, S. Gul, N. Ul Hassan, M. A. Rehman, M. A. Basit, H. M. W. Khalil, M. Ouladsmane, S. Rehman and M. F. Khan, Phys. B, 2023, 669, 415313–415320.
71. J. Shi, Y. Q. Bie, A. Zong, S. Fang, W. Chen, J. Han, Z. Cao, Y. Zhang, T. Taniguchi, K. Watanabe, X. Fu, V. Bulović, E. Kaxiras, E. Baldini, P. J. Herrero and K. A. Nelson, Nat. Commun., 2023, 14, 5905–5912.
72. T. Fukuda, U. Ozaki, S. Jeong, Y. Arashida, K. En-ya, S. Yoshida, P. J. Fons, J. I. Fujita, K. Ueno, M. Hase and M. Hada, J. Phys. Chem. C, 2023, 127, 13149–13156.
73. A. Bera, M. Kundu, B. Das, S. Kalimuddin, S. Bera, D. S. Roy, S. K. Pradhan, S. Naskar, S. K. De, B. Das and M. Mondal, Appl. Surf. Sci., 2023, 610, 155409–155416.
74. C. Kim, S. Issarapanacheewin, I. Moon, K. Y. Lee, C. Ra, S. Lee, Z. Yang and W. J. Yoo, Adv. Electron. Mater., 2020, 6, 1900964–1900970.
75. F. Lorén, C. Genet and L. M. Moreno, Opt. Mater. Express, 2023, 13, 3366–3375.
76. F. Caruso, M. Schebek, Y. Pan, C. Vona and C. Draxl, J. Phys. Chem. Lett., 2022, 13, 5894–5899.
77. C. Robert, S. Park, F. Cadiz, L. Lombez, L. Ren, H. Tornatzky, A. Rowe, D. Paget, F. Sirotti, M. Yang, D. V. Tuan, T. Taniguchi, B. Urbaszek, K. Watanabe, T. Amand, H. Dery and X. Marie, Nat. Commun., 2021, 12, 5455–5461.
78. K. Buchkov, R. Todorov, P. Terziyska, M. Gospodinov, V. Strijkova, D. Dimitrov and V. Marinova, Nanomaterials, 2021, 11, 2262–2274.
79. D. H. Keum, S. Cho, J. H. Kim, D. H. Choe, H. J. Sung, M. Kan, H. Kang, J. Y. Hwang, S. W. Kim, H. Yang, K. J. Chang and Y. H. Lee, Nat. Phys., 2015, 11, 482–486.
80. Y. Sun, J. Pan, Z. Zhang, K. Zhang, J. Liang, W. Wang, Z. Yuan, Y. Hao, B. Wang, J. Wang, Y. Wu, J. Zheng, L. Jiao, S. Zhou, K. Liu, C. Cheng, W. Duan, Y. Xu, Q. Yan and K. Liu, Nano Lett., 2019, 19, 761–769.
81. L. Fu, C. L. Kane and E. J. Mele, Phys. Rev. Lett., 2007, 98, 106803–106807.
82. Y. Qi, P. G. Naumov, M. N. Ali, C. R. Rajamathi, W. Schnelle, O. Barkalov, M. Hanfland, S. C. Wu, C. Shekhar, Y. Sun, V. Süβ, M. Schmidt, U. Schwarz, E. Pippel, P. Werner, R. Hillebrand, T. Förster, E. Kampert, S. Parkin, R. J. Cava, C. Felser, B. Yan and S. A. Medvedev, Nat. Commun., 2016, 7, 11038–11044.
83. A. A. Soluyanov, D. Gresch, Z. Wang, Q. Wu, M. Troyer, X. Dai and B. A. Bernevig, Nature, 2015, 527, 495–498.
84. K. Zhang, C. Bao, Q. Gu, X. Ren, H. Zhang, K. Deng, Y. Wu, Y. Li, J. Feng and S. Zhou, Nat. Commun., 2016, 7, 13552–13557.
85. S. Y. Chen, T. Goldstein, D. Venkataraman, A. Ramasubramaniam and J. Yan, Nano Lett., 2016, 16, 5852–5860.
86. C. Lan, C. Li, J. C. Ho and Y. Liu, Adv. Electron. Mater., 2021, 7, 2000688–2000723.
87. J. B. Barot, S. K. Gupta and P. N. Gajjar, Mater. Today: Proc., 2023 DOI:10.1016/j.matpr.2023.01.191.
88. M. C. Yuan, Y. L. Che, F. L. Yen and T. Kazuhito, Jpn. J. Appl. Phys., 2016, 55, 1102A1.
89. W. H. Lin, P. C. Wu, H. Akbari, G. R. Rossman, N. C. Yeh and H. A. Atwater, Adv. Mater., 2022, 34, 2104863–2104873.
90. J. Guo and K. Liu, Nanomaterials, 2021, 12, 110–123.
91. J. You, M. D. Hossain and Z. Luo, Nano Convergence, 2018, 5, 26–38.
92. M. S. Sokolikova and C. Mattevi, Chem. Soc. Rev., 2020, 49, 3952–3980.
93. S. Rani, M. Sharma, D. Verma, A. Ghanghass, R. Bhatia and I. Sameera, Mater. Sci. Semicond. Process., 2022, 139, 106313–106341.
94. Á. Coogan and Y. K. Gun'ko, Mater. Adv., 2021, 2, 146–164.
95. J. Wang, X. Zhao, G. Zou, L. Zhang, S. Han, Y. Li, D. Liu, C. Fernandez, L. Li, L. Ren and Q. Peng, Mater. Today Nano, 2023, 22, 100336–100358.
96. A. Jangir, D. T. Ho and U. Schwingenschlögl, J. Phys. Chem. Lett., 2025, 16, 811–817.
97. S. Ippolito and P. Samorì, Small Sci., 2022, 2, 2100122–2100132.
98. A. Turchanin and A. George, Small, 2024, 21, 2403089–2403116.
99. S. Bhowmik and A. G. Rajan, iScience, 2022, 25, 103832–103863.
100. K. H. Ruiz, Z. Wang, M. Ciprian, M. Zhu, R. Tu, L. Zhang, W. Luo, Y. Fan and W. Jiang, Small Sci., 2022, 2, 2100047–2100075.
101. W. H. Lin, C. S. Li, C. I. Wu, G. R. Rossman, H. A. Atwater and N. C. Yeh, Adv. Sci., 2024, 11, 2304890–2304904.
102. A. Raza, J. Hassan, U. Qumar, A. Zaheer, Z. D. Babar, V. Iannotti and A. Cassinese, Mater. Today Adv., 2024, 22, 100488–100526.
103. Y. Zhang, K. Nie, L. Yi, B. Li, Y. Yuan, Z. Liu and W. Huang, Adv. Sci., 2023, 10, 2302301–2302345.
104. J. Li, S. Cheng, Z. Liu, W. Zhang and H. Chang, J. Phys. Chem. C, 2018, 122, 7005–7012.
105. Q. Zhang, R. Zhang, J. Chen, W. Shen, C. An, X. Hu, M. Dong, J. Liu and L. Zhu, Beilstein J. Nanotechnol., 2019, 10, 1745–1753.
106. T. A. Empante, Y. Zhou, V. Klee, A. E. Nguyen, I. H. Lu, M. D. Valentin, S. A. N. Alvillar, E. Preciado, A. J. Berges, C. S. Merida, M. Gomez, S. Bobek, M. Isarraraz, E. J. Reed and L. Bartels, ACS Nano, 2016, 11, 900–905.
107. J. H. Sung, H. Heo, S. Si, Y. H. Kim, H. R. Noh, K. Song, J. Kim, C. S. Lee, S. Y. Seo, D. H. Kim, H. K. Kim, H. W. Yeom, T. H. Kim, S. Y. Choi, J. S. Kim and M. H. Jo, Nat. Nanotechnol., 2017, 12, 1064–1070.
108. L. Yang, W. Zhang, J. Li, S. Cheng, Z. Xie and H. Chang, ACS Nano, 2017, 11, 1964–1972.
109. J. C. Park, S. J. Yun, H. Kim, J. H. Park, S. H. Chae, S. J. An, J. G. Kim, S. M. Kim, K. K. Kim and Y. H. Lee, ACS Nano, 2015, 9, 6548–6554.
110. Y. Yoo, Z. P. DeGregorio, Y. Su, S. J. Koester and J. E. Johns, Adv. Mater., 2017, 29, 1605461–1605465.
111. L. Zhou, K. Xu, A. Zubair, X. Zhang, F. Ouyang, T. Palacios, M. S. Dresselhaus, Y. Li and J. Kong, Adv. Funct. Mater., 2016, 27, 1603491–1603498.
112. H. Kim, J. E. Johns and Y. Yoo, Small, 2020, 16, 2002849–2002857.
113. Q. Zhang, X. F. Wang, S. H. Shen, Q. Lu, X. Liu, H. Li, J. Zheng, C. Ping Yu, X. Zhong, L. Gu, T. Ling Ren and L. Jiao, Nat. Electron., 2019, 2, 164–170.
114. X. Xu, S. Liu, B. Han, Y. Han, K. Yuan, W. Xu, X. Yao, P. Li, S. Yang, W. Gong, D. A. Muller, P. Gao, Y. Ye and L. Dai, Nano Lett., 2019, 19, 6845–6852.
115. S. Yang, X. Xu, W. Xu, B. Han, Z. Ding, P. Gu, P. Gao and Y. Ye, ACS Appl. Nano Mater., 2020, 3, 10411–10417.
116. M. Govindasamy, S. F. Wang, R. Jothiramalingam, S. N. Ibrahim and H. A. Al-lohedan, Microchim. Acta, 2019, 186, 420–429.
117. G. Zhou, H. Gao, J. Li, X. He, Y. He, Y. Li and G. Hao, Nanotechnology, 2022, 33, 175602–175613.
118. C. R. Ratwani, S. Zhao, Y. Huang, M. Hadfield, A. R. Kamali and A. M. Abdelkader, Small, 2023, 19, 2207081–2207093.
119. K. Lasek, J. Li, S. Kolekar, P. M. Coelho, L. Guo, M. Zhang, Z. Wang and M. Batzill, Surf. Sci. Rep., 2021, 76, 100523–100574.
120. M. Singh, D. R. Paudel, H. Kim, T. H. Kim, J. Park and S. Lee, Energy Adv., 2025, 4, 716–742.
121. M. A. Akhound, K. W. Jacobsen and K. S. Thygesen, J. Am. Chem. Soc., 2025, 147, 5743–5754.
122. E. Chen, W. Xu, J. Chen and J. H. Warner, Mater. Today Adv., 2020, 7, 100076–100110.
123. N. Choudhary, M. A. Islam, J. H. Kim, T. J. Ko, A. Schropp, L. Hurtado, D. Weitzman, L. Zhai and Y. Jung, Nano Today, 2018, 19, 16–40.
124. B. Ruqia, M. K. Kabiraz, J. W. Hong and S. I. Choi, J. Energy Chem., 2022, 17, 217–240.
125. A. Chaves, J. G. Azadani, H. Alsalman, D. R. da Costa, R. Frisenda, A. J. Chaves, S. H. Song, Y. D. Kim, D. He, J. Zhou, A. C. Gomez, F. M. Peeters, Z. Liu, C. L. Hinkle, S. H. Oh, P. D. Ye, S. J. Koester, Y. H. Lee, P. Avouris, X. Wang and T. Low, npj 2D Mater. Appl., 2020, 4, 29–49.
126. B. S. Y. Kim, T. D. Ngo, Y. Hassan, S. H. Chae, S. G. Yoon and M. S. Choi, Adv. Sci., 2024, 11, 2407175–2407202.
127. N. Ahmed, W. Luo, R. Zhao, Q. Zhang, D. Wu, X. Zheng, H. Wang and C. Cui, Precis. Chem., 2025, 3, 110–127.
128. J. Park, I. Cho, H. Jeon, Y. Lee, J. Zhang, D. Lee, M. K. Cho, D. J. Preston, B. Shong, I. S. Kim and W. K. Lee, Adv. Mater., 2024, 36, 2314031–2314043.
129. T. Singh, D. Deepak, A. Panghal and S. S. Roy, Int. J. Hydrogen Energy, 2024, 84, 939–950.
130. H. Yang, K. Synnatschke, J. Yoon, H. Mirhosseini, I. M. Hermes, X. Li, C. Neumann, A. Morag, A. Turchanin, T. D. Kühne, S. S. P. Parkin, S. Yang, A. S. Nia and X. Feng, ACS Nano, 2025, 19, 14309–14317.
131. H. J. Kim, S. H. Kang, I. Hamada and Y. W. Son, Phys. Rev. B, 2017, 95, 180101–180105.
132. J. Azadmanjiri, P. Kumar, V. K. Srivastava and Z. Sofer, ACS Appl. Nano Mater., 2020, 3, 3116–3143.
133. Y. Jeong, B. Han, A. Tamayo, N. Claes, S. Bals and P. Samorì, ACS Nano, 2024, 18, 18334–18343.
134. F. Shirvani and Z. Razavifar, Phys. B, 2025, 696, 416609–416615.
135. H. Hu, Z. Wang, M. Pan, Y. Chen, Y. Han and J. Wang, Adv. Sci., 2025, 12, 2500226–2500248.
136. A. Roy, P. Kalita and B. Mondal, J. Mater. Sci.: Mater. Electron., 2022, 34, 224–237.
137. G. Tang, H. Tang, C. Li, W. Li and X. Ji, Mater. Lett., 2011, 65, 3457–3460.
138. X. Wang, K. Cui, Y. Zhao, M. Hao, L. Bian, M. Wang and F. Wang, Molecules, 2024, 29, 4555–4567.
139. S. Mourdikoudis, M. Menelaou, N. F. Maneiro, G. Zheng, S. Wei, J. P. Juste, L. Polavarapu and Z. K. Sofer, Nanoscale Horiz., 2022, 7, 941–1015.
140. W. Lin, B. Zhang, J. Jiang and L. Ma, Int. J. Electrochem. Sci., 2022, 17, 220738–220748.
141. Y. Feng, N. Feng, G. Zhang and G. Du, CrystEngComm, 2013, 16, 214–222.
142. D. M. Florjan and M. J. Szary, Acta Mater., 2024, 272, 119951–119964.
143. X. Zhang, Z. Zhou, D. Zhang, J. Chen, J. Zhang and Z. Wang, Electrochim. Acta, 2023, 457, 142455–142462.
144. A. Gupta and S. Vasudevan, J. Phys. Chem. C, 2018, 122, 19243–19250.
145. V. Georgakilas, J. N. Tiwari, K. C. Kemp, J. A. Perman, A. B. Bourlinos, K. S. Kim and R. Zboril, Chem. Rev., 2016, 116, 5464–5519.
146. L. Song, M. Song, Z. Lu, G. Yu, Z. Liang, W. Hou, Q. Liao and Y. Song, Coatings, 2022, 12, 1152–1172.
147. R. C. Vitoria, T. Hotta, M. Xue, S. Zhang and R. Kitaura, JACS Au, 2023, 3, 775–784.
148. A. Spustaka, A. Terehovs, A. Jansons and G. Kunakova, APL Mater., 2025, 13, 041110–041117.
149. J. Bradford, M. Shafiei, J. MacLeod and N. Motta, Sci. Rep., 2020, 10, 17334–17343.
150. S. Pace, L. Martini, D. Convertino, D. H. Keum, S. Forti, S. Pezzini, F. Fabbri, V. Miseikis and C. Coletti, ACS Nano, 2021, 15, 4213–4225.
151. S. Su, J. Zhao and T. H. Ly, Small Methods, 2024, 8, 2400211–2400231.
152. M. A. R. Laskar, S. Chakrabarti, S. Ahmed, S. A. Ghoreishi, P. Tummala, V. Afanas'ev, A. Molle, A. Lamperti and U. Celano, Mater. Sci. Semicond. Process., 2025, 188, 109247–109254.
153. E. Tillotson, J. G. McHugh, J. Howarth, T. Hashimoto, N. J. Clark, A. Weston, V. Enaldiev, S. S. Allsop, W. Thornley, W. Wang, M. Lindley, A. J. Pollard, V. I. Fal'ko, R. V. Gorbachev and S. J. Haigh, ACS Nano, 2024, 18, 34023–34033.
154. O. Bubnova, Nat. Nanotechnol., 2021, 16, 1179.
155. P. Li, Z. Kang, Z. Zhang, Q. Liao, F. Rao, Y. Lu and Y. Zhang, Mater. Today, 2021, 51, 247–272.
156. G. A. Naikoo, F. Arshad, M. Almas, I. U. Hassan, M. Z. Pedram, A. A. A. Aljabali, V. Mishra, A. S. Aroca, M. Birkett, N. B. Charbe, R. Goyal, P. Negi, M. E. Tanani and M. M. Tambuwala, Chem.-Biol. Interact., 2022, 365, 110081–110096.
157. M. Chubarov, T. H. Choudhury, X. Zhang and J. M. Redwing, Nanotechnology, 2018, 29, 055706–055714.
158. X. Cong, X. L. Liu, M. L. Lin and P. H. Tan, npj 2D Mater. Appl., 2020, 4, 13–24.
159. X. Zhao, Z. Li, S. Wu, M. Lu, X. Xie, D. Zhan and J. Yan, Adv. Electron. Mater., 2023, 10, 2300610–2300629.
160. H. Jeong, H. C. Suh, G. H. Cho, R. S. Montiel, H. Ko, K. K. Kim and M. S. Jeon, Appl. Surf. Sci., 2024, 646, 158823–158829.
161. R. Addou and R. M. Wallace, J. Electron Spectrosc. Relat. Phenom., 2019, 231, 94–103.
162. E. Nolota, S. Cadota, F. Martina, P. Hönicke, C. Zech and B. Beckhof, Spectrochim. Acta, Part B, 2020, 166, 105788–105794.
163. M. Tebyetekerwa, J. Zhang, Z. Xu, T. N. Truong, Z. Yin, Y. Lu, S. Ramakrishna, D. Macdonald and H. T. Nguyen, ACS Nano, 2020, 14, 14579–14604.
164. M. Gao, L. Yu, Q. Lv, F. Kang, Z. H. Huang and R. Lv, J. Materiomics, 2023, 9, 768–786.
165. J. Y. Kim, Ł. Gelczuk, M. P. Polak, D. Hlushchenko, D. Morgan, R. Kudrawiec and I. Szlufarska, npj 2D Mater. Appl., 2022, 6, 75–86.
166. D. R. Hang, Y. Q. Pan, K. H. Sharma, M. M. C. Chou, S. E. Islam, H. F. Wu and C. T. Liang, Nanomaterials, 2020, 10, 2045–2063.
167. U. Celano, D. Schmidt, C. Beitia, G. Orji, A. V. Davydov and Y. Obeng, Nanoscale Adv., 2024, 6, 2260–2269.
168. M. R. Charapale, T. D. Dongale, O. A. Patil, A. M. Teli, S. A. Beknalkar, S. B. Mullani, J. S. M. Mane, J. Lee and S. A. Mast, J. Solid State Electrochem., 2023, 28, 181–195.
169. F. X. R. Chen, C. Y. Lin, H. Y. Siao, C. Y. Jian, Y. C. Yang and C. L. Lin, Sci. Data, 2023, 10, 91–99.
170. R. G. Rivera, F. J. G. Garcia, T. Hayashi, Z. Wang and J. O. Medina, Comput. Mater. Sci., 2024, 242, 113076–113083.
171. H. Y. Chen, H. C. Hsu, J. Y. Liang, B. H. Wu, Y. F. Chen, C. C. Huang, M. Y. Li, I. P. Radu and Y. P. Chiu, ACS Nano, 2024, 18, 17622–17629.
172. C. Cong, J. Shang, Y. Wang and T. Yu, Adv. Opt. Mater., 2018, 6, 1700767–1700781.
173. Y. Sheng, T. Chen, Y. Lu, R. J. Chang, S. Sinha and J. H. Warner, ACS Nano, 2019, 13, 4530–4537.
174. H. Song, S. Chen, X. Sun, Y. Cui, T. Yildirim, J. Kang, S. Yang, F. Yang, Y. Lu and L. Zhang, Adv. Sci., 2024, 11, 2403176–2403199.
175. E. Wu, D. Wu, C. Jia, Y. Wang, H. Yuan, L. Zeng, T. Xu, Z. Shi, Y. Tian and X. Li, ACS Photonics, 2019, 6, 565–572.
176. L. Liu, X. Zhou, X. Wang, N. Liu, C. Dai, R. Xu, G. Yu, N. Wang and B. Chen, Energy Fuels, 2024, 38, 23847–23856.
177. D. Akinwande, N. Petrone and J. Hone, Nat. Commun., 2014, 5, 5678–5689.
178. J. Wu, W. Zhong, C. Yang, W. Xu, R. Zhao, H. Xiang, Q. Zhang, X. Li and N. Yang, Appl. Catal., B, 2022, 310, 121332–121343.
179. N. D. Canicoba, N. Zagni, F. Liu, G. McCuistian, K. Fernando, H. Bellezza, B. Traoré, R. Rogel, H. Tsai, L. L. Brizoual, W. Nie, J. J. Crochet, S. Tretiak, C. Katan, J. Even, M. G. Kanatzidis, B. W. Alphenaar, J. C. Blancon, M. A. Alam and A. D. Mohite, ACS Mater. Lett., 2019, 1, 633–640.
180. I. Zulfiqar, M. A. Khan, S. Gul, N. Ul Hassan, M. A. Rehman, M. A. Basit, H. M. W. Khalil, M. Ouladsmane, S. Rehman and M. F. Khan, Phys. B, 2023, 669, 415313–415320.
181. G. A. Asamoah, M. Korsah, P. G. S. A. Jeyasundar, M. Ahmed, S. Y. Lau and M. K. Danquah, Sustainability, 2024, 16, 9231–9276.
182. R. Hu, Z. Zhou, C. Sheng, S. Han, H. Yuan and H. Liu, Phys. Lett. A, 2022, 430, 127986–127990.
183. J. Ma, Y. Chen, Z. Han and W. Li, 2D Mater., 2016, 3, 045010–045017.
184. Y. Shao, Y. Zhang, N. Jiang, Y. Hao, K. Qu, Y. Zeng, Z. Liu, X. Lu, Y. Li, Q. Yang and J. Qiu, iScience, 2023, 26, 108470–108491.
185. R. Sankar, G. N. Rao, I. P. Muthuselvam, C. Butler, N. Kumar, G. S. Murugan, C. Shekhar, T. R. Chang, C. Y. Wen, C. W. Chen, W. L. Lee, M. T. Lin, H. T. Jeng, C. Felser and F. C. Chou, Chem. Mater., 2017, 29, 699–707.
186. Y. Deng, X. Zhao, C. Zhu, P. Li, R. Duan, G. Liu and Z. Liu, ACS Nano, 2021, 15, 12465–12474.
187. D. H. Kim, R. Lee, S. Kim and T. W. Kim, J. Alloys Compd., 2019, 789, 960–965.
188. L. Wang, Z. Wen and G. Luo, J. Phys. Chem. Lett., 2025, 16, 969–974.
189. H. Jin, T. Song, U. Paik and S. Z. Qiao, Acc. Mater. Res., 2021, 2, 559–573.
190. S. Singh, J. Kim, K. M. Rabe and D. Vanderbilt, Phys. Rev. Lett., 2020, 125, 046402–046407.
191. S. S. Awate, K. Xu, J. Liang, B. Katz, R. Muzzio, V. H. Crespi, J. Katoch and S. K. F. Shirey, ACS Nano, 2023, 17, 22388–22398.
192. T. Lei, W. Chen, J. Huang, C. Yan, H. Sun, C. Wang, W. Zhang, Y. Li and J. Xiong, Adv. Energy Mater., 2016, 7, 1601843–1601850.
193. Y. Zhu, S. Yang, Z. Wei, H. He, Y. Zhu and J. Zhu, Carbon, 2025, 232, 119790–119797.
194. M. Liu, Z. Wang, J. Liu, G. Wei, J. Du, Y. Li, C. An and J. Zhang, J. Mater. Chem. A, 2017, 5, 1035–1042.
195. M. R. Panda, S. Sau, Q. Bao, M. Majumder and S. Mitra, Mater. Today: Proc., 2022, 50, 113–116.
196. H. Yang, Y. Zhao, Q. Wen, R. Yang, Y. Liu, H. Li and T. Zhai, ACS Appl. Mater. Interfaces, 2021, 13, 14302–14311.
197. T. Järvinen, G. S. Lorite, J. Peräntie, G. Toth, S. Saarakkala, V. K. Virtanen and K. Kordas, Nanotechnology, 2019, 30, 405501–405510.
198. M. Scardamaglia, J. C. Cháfer, R. Temperton, F. E. Annanouch, A. Mohammadpour, G. Malandra, A. Das, A. Alagh, I. Arbouch, L. Montoisy, D. Cornil, J. Cornil, E. Llobet and C. Bittencourt, ACS Sens., 2024, 9, 4079–4088.
199. Y. R. Kumar, K. Deshmukh, T. Kovářík and S. K. K. Pasha, Coord. Chem. Rev., 2022, 461, 214502–214559.
200. A. Das, D. Roy, B. K. Das, M. I. Ansari, K. K. Chattopadhyay and S. Sarkar, Catal. Today, 2023, 423, 113921–113931.
201. S. Lu, H. L. Huynh, F. Lou, K. Guo and Z. Yu, Nanoscale, 2021, 13, 12885–12895.
202. B. Lu, Y. Xia, Y. Ren, M. Xie, L. Zhou, G. Vinai, S. A. Morton, A. T. S. Wee, W. G. van der Wiel, W. Zhang and P. K. J. Wong, Adv. Sci., 2024, 11, 2305277–2305316.
203. A. Siddiqui and N. D. M. Hine, npj Comput. Mater., 2024, 10, 169–179.
204. K. Momeni, Y. Ji, Y. Wang, S. Paul, S. Neshani, D. E. Yilmaz, Y. K. Shin, D. Zhang, J. W. Jiang, H. S. Park, S. Sinnott, A. V. Duin, V. Crespi and L. Q. Chen, npj Comput. Mater., 2020, 6, 22–39.
205. K. Ludwiczak, A. K. Da̧browska, J. Binder, M. Tokarczyk, J. Iwański, B. Kurowska, J. Turczyński, G. Kowalski, R. Bożek, R. Stȩpniewski, W. Pacuski and A. Wysmołek, ACS Appl. Mater. Interfaces, 2021, 13, 47904–47911.
206. Y. Sang, Z. Zhao, M. Zhao, P. Hao, Y. Leng and H. Liu, Adv. Mater., 2014, 27, 363–369.
207. J. Yang, Y. Jin, W. Xu, B. Zheng, R. Wang and H. Xu, J. Phys. Chem. Lett., 2018, 9, 4783–4788.
208. X. Liu, H. Yang, J. He, H. Liu, L. Song, L. Li and J. Luo, Small, 2018, 14, 1704049–1704056.
209. R. Zhou, J. Wu, Y. Chen and L. Xie, Chin. J. Chem., 2022, 40, 989–1004.
210. F. M. Altincicek, Master of Science (MS), University of North Dakota, 2019.
211. H. Zeng, Y. Wen, L. Yin, R. Cheng, H. Wang, C. Liu and J. He, Front. Phys., 2023, 18, 53603–53650.
212. I. Shahbaz, M. Tahir, L. Li and Y. Song, Mater. Today, 2024, 77, 142–184.
213. Z. A. C. Ramli, J. Pasupuleti, S. Samidin, T. S. T. Saharuddin, W. N. R. W. Isahak, A. G. N. Sofiah, S. P. Koh and S. T. Kiong, Fuel, 2025, 387, 134386–134418.
214. J. Lim, J. I. Lee, Y. Wang, N. Gauriot, E. Sebastian, M. Chhowalla, C. Schnedermann and A. Rao, Nature, 2024, 633, 83–89.
215. X. Liu, P. Zhang, S. Wang, Y. Fang, P. Wu, Y. Xiang, J. Chen, C. Zhao, X. Zhang, W. Zhao, J. Wang, F. Huang and C. Guan, Nat. Commun., 2024, 15, 1263–1272.
216. Y. Fang, J. Pan, D. Zhang, D. Wang, H. T. Hirose, T. Terashima, S. Uji, Y. Yuan, W. Li, Z. Tian, J. Xue, Y. Ma, W. Zhao, Q. Xue, G. Mu, H. Zhang and F. Huang, Adv. Mater., 2019, 31, e1901942–e1901948.
217. M. N. Ali, L. Schoop, J. Xiong, S. Flynn, Q. Gibson, M. Hirschberger, N. P. Ong and R. J. Cava, Europhys. Lett., 2015, 110, 67002–67006.
218. Y. Wang, L. Wang, X. Liu, H. Wu, P. Wang, D. Yan, B. Cheng, Y. Shi, K. Watanabe, T. Taniguchi, S.-J. Liang and F. Miao, Nano Lett., 2019, 19, 3969–3975.
219. X. Dong, H. Li, T. Chen, L. Xu and G. Zhou, Mater. Sci. Semicond. Process., 2022, 143, 106537–106543.
220. Y. Wang, J. Xiao, H. Zhu, Y. Li, Y. Alsaid, K. Y. Fong, Y. Zhou, S. Wang, W. Shi, Y. Wang, A. Zettl, E. J. Reed and X. Zhang, Nature, 2017, 550, 487–491.
221. B. Mortazavi, G. R. Berdiyorov, M. Makaremi and T. Rabczuk, Extreme Mech. Lett., 2018, 20, 65–72.
222. S. Che, Y. Jia and Y. Li, Catalysts, 2025, 15, 205–231.
223. J. Zeng, Y. Liu, Z. Huang, H. Qiao and X. Qi, Catalysts, 2024, 14, 689–725.
224. K. Sathiyan, T. Mondal, P. Mukherjee, S. Gopal Patra, I. Pitussi, H. Kornweitz, R. B. Ziv and T. Zidki, Nanoscale, 2022, 14, 16148–16155.
225. K. A. Cochrane, J. H. Lee, C. Kastl, J. B. Haber, T. Zhang, A. Kozhakhmetov, J. A. Robinson, M. Terrones, J. Repp, J. B. Neaton, A. W. Bargioni and B. Schuler, Nat. Commun., 2021, 12, 7287–7296.
226. B. Schuler, D. Y. Qiu, S. Refaely-Abramson, C. Kastl, C. T. Chen, S. Barja, R. J. Koch, D. F. Ogletree, S. Aloni, A. M. Schwartzberg, J. B. Neaton, S. G. Louie and A. W. Bargioni, Phys. Rev. Lett., 2019, 123, 076801–076807.
227. S. Li, G. Thiering, P. Udvarhelyi, V. Ivády and A. Gali, Nat. Commun., 2022, 13, 1210–1216.
228. Y. Huang, Y. H. Pan, R. Yang, L. H. Bao, L. Meng, H. L. Luo, Y. Q. Cai, G. D. Liu, W. J. Zhao, Z. Zhou, L. M. Wu, Z. L. Zhu, M. Huang, L. W. Liu, L. Liu, P. Cheng, K. H. Wu, S. B. Tian, C. Z. Gu, Y. G. Shi, Y. F. Guo, Z. G. Cheng, J. P. Hu, L. Zhao, G. H. Yang, E. Sutter, P. Sutter, Y. L. Wang, W. Ji, X. J. Zhou and H. J. Gao, Nat. Commun., 2020, 11, 2453–2461.
229. F. Ye, J. Lee, J. Hu, Z. Mao, J. Wei and P. X. L. Feng, Small, 2016, 12, 5802–5808.
230. C. H. Lee, E. C. Silva, L. Calderin, M. A. T. Nguyen, M. J. Hollander, B. Bersch, T. E. Mallouk and J. A. Robinson, Sci. Rep., 2015, 5, 10013–10020.
231. I. Ozdemir, A. W. Holleitner, C. Kastl and O. Ü. Aktürk, Sci. Rep., 2022, 12, 12756–12771.
232. F. Ahmed, A. M. Shafi, D. M. A. Mackenzie, M. A. Qureshi, H. A. Fernandez, H. H. Yoon, M. G. Uddin, M. Kuittinen, Z. Sun and H. Lipsanen, Adv. Mater. Interfaces, 2021, 8, 2100950–2100957.
233. B. Han, S. M. Gali, S. Dai, D. Beljonne and P. Samor, ACS Nano, 2023, 17, 17956–17965.
234. D. H. Lee, M. Rabeel, Y. Han, H. Kim, M. F. Khan, D. K. Kim and H. Yoo, ACS Appl. Mater. Interfaces, 2023, 15, 51518–51526.
235. Y. Jeong, B. Han, A. Tamayo, N. Claes, S. Bals and P. Samorì, ACS Nano, 2024, 18, 18334–18343.
236. X. Xu, T. Liang, D. Kong, B. Wang and L. Zhi, Mater. Today Nano, 2021, 14, 100111–100127.
237. A. C. Manikas, M. G. Pastore Carbone, C. R. Woods, Y. Wang, I. Souli, G. Anagnostopoulos, M. Hadjinicolaou, K. S. Novoselov and C. Galiotis, Nanoscale, 2019, 11, 14354–14361.
238. F. Colangelo, A. Morandi, S. Forti, F. Fabbri, C. Coletti, F. V. D. Girolamo, A. D. Lieto, M. Tonelli, A. Tredicucci, A. Pitanti and S. Roddaro, Appl. Phys. Lett., 2019, 115, 183101–183106.
239. Z. Li, I. A. Kinloch and R. J. Young, Philos. Trans. R. Soc., A, 2016, 374, 20150283–20150296.
240. F. Wang, S. Li, M. A. Bissett, I. A. Kinloch, Z. Li and R. J. Young, 2D Mater., 2020, 7, 045022–045035.
241. P. K. Muscher, D. A. Rehn, A. Sood, K. Lim, D. Luo, X. Shen, M. Zajac, F. Lu, A. Mehta, Y. Li, X. Wang, E. J. Reed, W. C. Chueh and A. M. Lindenberg, Adv. Mater., 2021, 33, 2101875–2101883.
242. G. Y. Wang, W. Xie, D. Xu, H. Yang Ma, H. Yang, H. Lu, H. Hua Sun, Y. Y. Li, S. Jia, L. Fu, S. Zhang and J. F. Jia, Nano Res., 2019, 12, 569–573.
243. S. Song, D. H. Keum, S. Cho, D. Perello, Y. Kim and Y. H. Lee, Nano Lett., 2016, 16, 188–193.
244. W. Hou, A. Azizimanesh, A. Sewaket, T. Peña, C. Watson, M. Liu, H. Askari and S. M. Wu, Nat. Nanotechnol., 2019, 14, 668–673.
245. Y. Singh, S. Back and P. Y. Jung, ChemElectroChem, 2018, 5, 4029–4035.
246. D. Arumugam, D. Subramani, A. Muralidharan and S. Ramasamy, ACS Appl. Mater. Interfaces, 2024, 16, 64916–64928.
247. R. Kumar, A. K. Shringi, H. J. Wood, I. M. Asuo, S. Oturak, D. E. Sanchez, T. S. K. Sharma, R. Chaurasiya, A. Mishra, W. M. Choi, N. Y. Doumon, I. Dabo, M. Terrones and F. Yan, Mater. Sci. Eng., R, 2025, 163, 100946–100979.
248. Y. C. Lin, R. Torsi, D. B. Geohegan, J. A. Robinson and K. Xiao, Adv. Sci., 2021, 8, 2004249–2004274.
249. H. Cui, G. Zhang, X. Zhang and J. Tang, Nanoscale Adv., 2019, 1, 772–780.
250. H. Zhu, H. Cui, D. He, Z. Cui and X. Wang, Nanoscale Res. Lett., 2020, 15, 129–139.
251. L. Zhou, M. Li, W. Wang, C. Wang, H. Yang and Y. Cao, Nano Res., 2022, 15, 9890–9905.
252. L. Liu, K. Ye, C. Lin, Z. Jia, T. Xue, A. Nie, Y. Cheng, J. Xiang, C. Mu, B. Wang, F. Wen, K. Zhai, Z. Zhao, Y. Gong, Z. Liu and Y. Tian, Nat. Commun., 2021, 12, 3870–3877.
253. L. Peng, Y. Yuan, G. Li, X. Yang, J. J. Xian, C. Jiang Yi, Y. G. Shi and Y. S. Fu, Nat. Commun., 2017, 8, 659–665.
254. H. C. Diaz, Y. Ma, R. Chaghi and M. Batzill, Appl. Phys. Lett., 2016, 108, 191606–191611.
255. K. Elibol, T. Susi, G. Argentero, M. R. A. Monazam, T. J. Pennycookm, J. C. Meyer and J. Kotakoski, Chem. Mater., 2018, 30, 1230–1238.
256. T. Zhang, K. Fujisawa, F. Zhang, M. Liu, M. C. Lucking, R. Nunes Gontijo, Y. L. Liu, K. Crust, T. G. Nakajima, H. Terrones, A. L. Elías and M. Terrone, ACS Nano, 2020, 14, 4326–4335.
257. L. G. Bulusheva, G. I. Semushkina and A. D. Fedorenko, Nanometerials, 2020, 13, 2182–2222.
258. M. Alahmadi and S. B. Aoun, J. Saudi Chem. Soc., 2024, 28, 101845–101853.
259. N. Amirulloieva, J. Sušinska, A. T. Deksne, R. Ignatans, M. Zubkins, J. Gabrusenoks, A. Sarakovskis and G. Kunakova, APL Mater., 2025, 13, 041128–041134.
260. M. Ramzan, A. S. Alqarni, K. Ahmad, A. G. Al-Sehemi, A. M. A. Henaish, M. Aslam, M. J. Khan and M. Saleem, Diamond Relat. Mater., 2024, 148, 111364–111375.
261. H. Kwon, B. Ji, D. Bae, J. H. Lee, H. Jung Park, D. H. Kim, Y. M. Kim, Y. W. Son, H. Yang and S. Cho, Appl. Surf. Sci., 2020, 515, 145972–145977.
262. X. Yin, Y. Hua and Z. Gao, Renewables, 2023, 1, 190–226.
263. S. Balasubramanyam, M. Shirazi, M. A. Bloodgood, L. Wu, M. A. Verheijen, V. Vandalon, W. s. M. M. Kessels, J. P. Hofmann and A. A. Bol, Chem. Mater., 2019, 31, 5104–5115.
264. S. Tang, C. Zhang, D. Wong, Z. Pedramrazi, H. Z. Tsai, C. Jia, B. Moritz, M. Claassen, H. Ryu, S. Kahn, J. Jiang, H. Yan, M. Hashimoto, D. Lu, R. G. Moore, C. C. Hwang, C. Hwang, Z. Hussain, Y. Chen, M. M. Ugeda, Z. Liu, X. Xie, T. P. Devereaux, M. F. Crommie, S. Kwan Mo and Z. X. Shen, Nat. Phys., 2017, 13, 683–687.
265. L. Peng, Y. Yuan, G. Li, X. Yang, J. J. Xian, C. J. Yi, Y. G. Shi and Y. S. Fu, Nat. Commun., 2017, 8, 659–665.
266. F. Lüpke, D. Waters, S. C. de la Barrera, M. Widom, D. G. Mandrus, J. Yan, R. M. Feenstra and B. M. Hunt, Nat. Phys., 2020, 16, 526–530.
267. L. Sun, M. Ding, J. Li, L. Yang, X. Lou, Z. Xie, W. Zhang and H. Chang, Appl. Surf. Sci., 2019, 496, 143687–143694.
268. Z. Zhang, P. Liu, Y. Song, Y. Hou, B. Xu, T. Liao, H. Zhang, J. Guo and Z. Sun, Adv. Sci., 2022, 9, 2204297–2204322.
269. C. Sekhar Rout, Engineered 2D Materials for Electrocatalysis Applications, IOP, 2024,  DOI:10.1088/978-0-7503-5719-7.
270. Z. Zhang, P. Chen, X. Duan, K. Zang, J. Luo and X. Duan, Science, 2017, 357, 788–792.
271. D. Wang, Q. Li, C. Han, Z. Xing and X. Yang, ACS Cent. Sci., 2018, 4, 112–119.
272. K. Kang, W. Zhao, Y. Zeng, K. Watanabe, T. Taniguchi, J. Shan and K. F. Mak, Nat. Nanotechnol., 2023, 18, 861–866.
273. Y. Zhou, J. V. Pondick, J. L. Silva, J. M. Woods, D. J. Hynek, G. Matthews, X. Shen, Q. Feng, W. Liu, Z. Lu, Z. Liang, B. Brena, Z. Cai, M. Wu, L. Jiao, S. Hu, H. Wang, C. M. Araujo and J. J. Cha, Small, 2019, 15, 1900078–1900089.
274. Y. Cao, V. Fatemi, S. Fang, K. Watanabe, T. Taniguchi, E. Kaxiras and P. J. Herrero, Nature, 2018, 556, 43–50.
275. S. Ulstrup, R. J. Koch, S. Singh, K. M. McCreary, B. T. Jonker, J. T. Robinson, C. Jozwiak, E. Rotenberg, A. Bostwick, J. Katoch and J. A. Miwa, Sci. Adv., 2020, 6, 6.
276. Y. Saruta, K. Sugawara, H. Oka, T. Kawakami, T. Kato, K. Nakayama, S. Souma, T. Takahashi, T. Fukumura and T. Sato, Adv. Sci., 2023, 10, 2304461–2304467.
277. J. Mao, L. Zhou, Y. Li, Y. Tao, K. Chai, Y. Shi and W. Xu, Mater. Lett., 2021, 290, 129471.
278. S. Watzele, J. Fichtner, B. Garlyyev, J. N. Schwämmlein and A. S. Bandarenka, ACS Catal., 2018, 8, 9456–9462.
279. Y. Lei, X. Xiao, T. Ma, W. Li, H. Zhang and C. Ma, Front. Chem., 2022, 10, 1005782–1005790.
280. R. Ding, J. Chen, Y. Chen, J. Liu, Y. Bando and X. Wang, Chem. Soc. Rev., 2024, 53, 11390–11461.
281. S. C. Cho, J. H. Seok, H. N. Manh, J. H. Seol, C. H. Lee and S. U. Lee, Nano Convergence, 2025, 12, 4–31.
282. K. Deshsorn, P. Chavalekvirat, S. Deepaisarn, H. C. Chuang and P. Iamprasertkun, ACS Mater. Au, 2025 DOI:10.1021/acsmaterialsau.5c00030.
283. B. Han, Y. Lin, Y. Yang, N. Mao, W. Li, H. Wang, K. Yasuda, X. Wang, V. Fatemi, L. Zhou, J. I.-J. Wang, Q. Ma, Y. Cao, D. R. Legrain, Y. Q. Bie, E. Navarro Moratalla, D. Klein, D. MacNeill, S. Wu, H. Kitadai, X. Ling, P. Jarillo Herrero, J. Kong, J. Yin and T. Palacios, Adv. Mater., 2020, 32, 2000953–2000962.
284. S. Nayak, S. Bhattacharjee, J. H. Choi and S. C. Lee, J. Phys. Chem. A, 2020, 124, 247–254.
285. A. S. Adamson and A. Smith, JAMA Dermatol., 2018, 154, 1247–1248.
286. A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga, A. Desmaison, A. Köpf, E. Yang, Z. DeVito, M. Raison, A. Tejani, S. Chilamkurthy, B. Steiner, L. Fang, J. Bai and S. Chintala, presented in part at the arXiv, 2019, preprint.
287. L. Wu, T. Guo and T. Li, iScience, 2021, 24, 102398–102425.
288. Y. Chen, H. Cui, Q. Jiang, X. Bai, P. Shan, Z. Jia, S. Lu, P. Song, R. Feng, Q. Kang, Z. Liang and H. Yuan, ACS Appl. Nano Mater., 2023, 6, 7694–7703.
289. D. A. Otchere, T. O. A. Ganat, J. O. Ojero, B. N. T. Otoo and M. Y. Taki, J. Pet. Sci. Eng., 2022, 208, 109244–109244.
290. Q. Cheng, Z. Chunhong and L. Qianglin, Sci. Rep., 2023, 13, 9149–9163.
291. T. Yuxin, Y. Hairong, G. Hang, S. Yuying and L. Gang, J. Ind. Inf. Integr., 2019, 15, 247–251.
292. M. Durbin, M. A. Wonders, M. Flaska and A. T. Lintereur, Nucl. Instrum. Methods Phys. Res., Sect. A, 2021, 987, 164826–164834.
293. V. L. Deringer, A. P. Bartók, N. Bernstein, D. M. Wilkins, M. Ceriotti and G. Csányi, Chem. Rev., 2021, 121, 10073–10141.
294. X. Wan, W. Yu, H. Niu, X. Wang, Z. Zhang and Y. Guo, Chem. Eng. J., 2022, 440, 135946–135955.
295. X. Wan, Z. Zhang, H. Niu, Y. Yin, C. Kuai, J. Wang, C. Shao and Y. Guo, J. Phys. Chem. Lett., 2021, 12, 6111–6118.
296. P. Aghamohammadi, B. Hüner, O. C. Altıncı, E. T. Akgul, B. Teymur, U. B. Simsek and M. Demir, Int. J. Hydrogen Energy, 2024, 87, 179–198.
297. S. Onajah, R. Sarkar, M. S. Islam, M. Lalley, K. Khan, M. Demir, H. N. Abdelhamid and D. A. A. Farghaly, Chem. Rec., 2024, 2024, e202300234–e202300282.
298. S. Wu, J. Cao, Y. Zhou, J. He, J. Zhou, M. Demir, H. Yang, Z. Tong and P. Ma, Int. J. Hydrogen Energy, 2025, 127, 275–285.

---

Footnote

† These authors contributed equally to this work.

---