Efficient photocatalytic nitrogen reduction by MoS₂ doped with transition metal and containing sulfur vacancies: enhanced nitrogen activation and inhibition of water decomposition

Abstract

Effectively enhancing nitrogen activation and inhibiting water splitting reactions are key factors in photocatalytic nitrogen reduction. This work integrated defect engineering and metal doping, doping transition metals Sc, Cu, and V into molybdenum sulfide (MoS₂) containing sulfur vacancies (V_S), to prepare TM@V_S-MoS₂ (TM = Sc, Cu, V) and applied it for the photocatalytic nitrogen reduction (pNRR). High angle annular dark field scanning transmission electron microscopy (HAADF-STEM) confirms the existence of S vacancies. The pNRR activity values of Sc@V_S-MoS₂, Cu@V_S-MoS₂ and V@V_S-MoS₂ are 124.26 μmol g⁻¹ h⁻¹, 101.32 μmol g⁻¹ h⁻¹, and 88.64 μmol g⁻¹ h⁻¹, respectively, which are much higher than those of MoS₂, V_S-MoS₂, and TM@MoS₂. This fully demonstrates that the comprehensive use of S vacancies and metal doping can significantly enhance the photocatalytic performance of MoS₂ (with a maximum activity improvement of nearly 5 times). DFT calculations show that transition metal doping of MoS₂ with sulfur vacancies can further promote the activation of N₂ and inhibit the production of H₂ from water decomposition. The above promotion and inhibition effects are consistent with the following order of activity and photoelectronic properties: Sc@V_S-MoS₂ > Cu@V_S-MoS₂ > V@V_S-MoS₂. In situ infrared spectroscopy revealed the generation of –N₂H_y (1 ≤ y ≤ 4) active species during the pNRR process, which is consistent with the most potential pathway (enzymatic mechanism) calculated by DFT.

---

1. Introduction

Ammonia can be decomposed into nitrogen and hydrogen under the action of a catalyst. Compared with hydrogen, ammonia has the advantages of easy storage, convenient transportation, and low cost.¹ In addition, ammonia also serves as an energy storage intermediate, and the product after complete combustion is water, which does not produce harmful gases such as carbon monoxide and the greenhouse gas carbon dioxide.² However, traditionally, the synthesis of ammonia is mostly achieved through the Haber–Bosch process, which requires high temperature (15–25 MPa) and high pressure (673–873 K) reaction conditions, consuming nearly 2% of global energy.³ Therefore, it is highly necessary to find a friendly method that can fix nitrogen into ammonia under environmental conditions. The photocatalytic nitrogen reduction to ammonia reaction (pNRR) using sunlight as the driving force and water as the hydrogen source is a sustainable method for synthesizing ammonia under environmental conditions.

The photocatalytic synthesis of ammonia has attracted widespread attention from researchers due to its advantages such as mild conditions, simple infrastructure, less environmental pollution, and renewable energy from solar energy.⁴ In the photocatalytic synthesis of ammonia, photocatalysts play a crucial role. Researchers have developed various photocatalytic materials such as TiO₂,⁵ g-C₃N₄,⁶ ZnO,⁷ BiVO₄,⁸ and Cu₂O,⁹ but there are still many challenges. The key challenges are to find more efficient photocatalysts yet, adjust the band structure of photocatalysts through various control strategies, broaden the visible light absorption range, and effectively reduce the recombination of photogenerated carriers.¹⁰

Transition metal dichalcogenides (TMDCs) are layered compounds with the chemical formula MX₂, where M is a transition metal atom (Mo, W, Re, Nb, etc.) and X is a chalcogen atom (S, Se, and Te), with Mo, W, S and Se being the most common elemental compositions.¹¹ In the most common TMDCs, the 2H and 3R structures are the semiconductor phase, while the 1T structure is the metallic phase.¹² TMDC materials have attracted much attention due to their low cost, high catalytic activity, high stability, large in-plane carrier mobility, and excellent mechanical properties.¹³ In particular, the 2H phase is highly suitable as a photocatalyst. As a typical TMDC material, MoS₂ has the most thermodynamically stable configuration of 2H phase, while the 1H phase and 3R phase coexist in a metastable state. The bulk MoS₂ is composed of stacked phases with different phases, and the total bandgap of the material is 0.88 eV. Single layer MoS₂ consists of the 2H phase and 3R phase, with a theoretical bandgap of 1.71 eV. In experiments, the bandgap can reach 2.16 eV,¹⁴ making it an excellent photocatalyst. In addition, two-dimensional MoS₂ not only has the general characteristics of two-dimensional materials but also has the following advantages: (1) the electronic structure and physicochemical properties are easy to adjust; (2) the activity of edge sites is high; (3) the inert basal plane can be activated; (4) the existence of Mo–S bonds is similar to the active structure of biological nitrogenase; (5) hydrogen activation is strong.^15,16 Therefore, MoS₂ materials have been applied in photocatalytic nitrogen reduction reactions, photocatalytic hydrogen evolution reactions, and photocatalytic carbon dioxide reduction reactions.^17–19

Although the edges of two-dimensional MoS₂ contain a large number of active sites, the large-area basal plane is still inert, and there is still great room for improvement in the effective activation of nitrogen molecules. Thus, further research is warranted on how to effectively improve the NRR catalytic activity of the MoS₂ substrate. Defect engineering is a method of regulating and modifying the local coordination environment in semiconductor photocatalysts.²⁰ Defect engineering can alter the band structure of semiconductors, accelerate the transfer of photoexcited electrons, and promote the adsorption and activation of N₂ on semiconductor surfaces.²¹ In addition, constructing vacancies/defects (such as oxygen vacancies, sulfur vacancies, etc.) on the MoS₂ substrate is beneficial for regulating the charge distribution of the substrate. These anionic defects can reduce the number of adjacent coordinated non-metallic elements, thereby forming low-priced and electron-rich TM centers.²² When N₂ enters, these TM centers around the defect can serve as N₂ coordination sites, on the one hand, by weakening the HOMO electron density of N₂, and on the other hand by enriching the LUMO electron density of N₂ through excessive electron supply, thereby significantly weakening the N–N triple bond and facilitating subsequent hydrogenation on nitrogen.²³ Zhang et al. synthesized MoS₂ nanosheets (V_S-MoS₂) with different concentrations of S vacancies through experiments for the photocatalytic nitrogen reduction reaction, and the mechanism by which S vacancies enhance the activity of MoS₂ was explored through DFT calculations.²⁴ The exposed Mo atoms generated by S vacancies change the charge layout on the surface of MoS₂ while providing abundant Mo active sites; meanwhile, the introduction of S vacancies narrows the band gap of the material, enhances the adsorption and activation of N₂, and significantly improves the activity of nitrogen reduction.

Another key factor in photocatalytic nitrogen reduction of molybdenum sulfide materials is to suppress water splitting reactions and improve the selectivity of nitrogen reduction. Doping is an effective means of introducing heteroatoms into the target lattice to regulate the fundamental properties of semiconductors, which can change the band structure of materials and expand the range of light response.²⁵ Doping metal elements into semiconductors will generate an impurity state in the bandgap, which can effectively reduce the recombination of photogenerated carriers and increase the surface active site density of the catalyst, potentially changing the reaction pathway of the catalyst.²⁶ Therefore, the strategy of doping transition metal atoms can further regulate the surface electronic structure of MoS₂, suppress hydrogen evolution reactions, and enhance its photocatalytic NRR performance. For example, Fei et al.²⁷ successfully doped P atoms as dopants into Vs-MoS₂ catalysts. P atoms can induce S vacancies as adsorption and activation sites for N₂ molecules. By adjusting the electronic structure of MoS₂, the hydrogen evolution reaction is effectively suppressed, and the NRR activity of the catalyst is enhanced. Zhou et al.²⁸ prepared MoS₂ nanosheets with different Cu doping concentrations. The doping of Cu increased the content of S vacancies on the surface of MoS₂, thereby changing the electronic structure and chemical properties of the surface, and improving the catalytic activity and selectivity of MoS₂ catalysis.

In summary, based on two-dimensional MoS₂, comprehensive defect engineering and metal doping, it is expected to improve the activation of nitrogen molecules and inhibit water decomposition, thereby effectively enhancing the activity of photocatalytic nitrogen reduction. At present, there have been literature reports on the photocatalytic nitrogen reduction of defect engineering or metal-doped modified molybdenum sulfide materials, but there are few reports on the photocatalytic nitrogen reduction of molybdenum sulfide doped with various transition metals containing S defects.^29,30 However, there is no literature report on the intrinsic reasons for the activity and reaction mechanism of various transition metal elements doped and compared in depth.

Our previous work has shown that introducing S vacancies can activate nitrogen molecules to a certain extent, thereby enhancing the activity of MoS₂ photocatalytic nitrogen reduction. However, it still has strong H adsorption free energy and *NH₃ desorption energy.²⁴ Thus, this work further doped Sc, Cu, and V elements into MoS₂ containing S vacancies and constructed three transition metal-doped MoS₂ containing S vacancies (TM@VS-MoS₂, where TM = Sc, Cu, V) and applied it for photocatalytic nitrogen reduction. The structure of TM@VS-MoS₂ (TM = Sc, Cu, V) was confirmed by XRD, SEM, TEM, XPS, EPR and other structural methods. Detailed verification and comparison were conducted on differences in the nitrogen reduction activity of three transition metal-doped molybdenum sulfide containing S vacancies under different reaction conditions, and the stability and reusability of the catalysts were also explored. In addition, through theoretical calculations, detailed analysis and comparison of the structural properties, adsorption and activation capabilities or N₂, and the overall reaction mechanism of nitrogen reduction of three types of TM@V_S-MoS₂ materials were conducted from the perspectives of structural stability and adsorption selectivity for N₂ and H.

2. Experimental

Information on material characterization (Fig. S1†), experimental details of nitrogen reduction (Fig. S2†), and setting of some calculation parameters are provided in the ESI.†

2.1 Materials and sources

Na₂MoO₄·2H₂O (AR, Shanghai Aladdin Biochemical Technology Co., Ltd); (NH₂)₂CS (AR, Shanghai Wanwei Biotechnology Co., Ltd); (CH₂OH)₂ (AR, Shanghai Lingfeng Chemical Co., Ltd), VO(AcAc)₂ (AR, Shanghai Mindray Biotechnology Co., Ltd), CuCl₂ (AR, McLean Chemical Reagent Co., Ltd), Sc(NO₃)₃ (AR, Shanghai Mindray Biotechnology Co., Ltd), N₂H₄·H₂O (AR, Shanghai Aladdin Biochemical Technology Co., Ltd), KOH (AR, Shanghai Aladdin Biochemical Technology Co., Ltd), Ar (Hangzhou Minxing Special Gas Co., Ltd), and H₂ (Hangzhou Minxing Special Gas Co., Ltd).

2.2 Synthesis of molybdenum sulfide (MoS₂)

MoS₂ was prepared by a hydrothermal method: weigh 1.7 g Na₂MoO₄·2H₂O and 1.07 g (NH₂)₂CS separately and dissolve in 50 mL of deionized water, sonicate for 5 min, and then stir for 30 min to further mix the solution evenly. Transfer the homogeneous mixture to a high-pressure vessel lined with polytetrafluoroethylene with a volume of 100 mL and maintain it at a temperature of 200 °C for 24 hours. After the reaction is complete, centrifuge the product and wash multiple times with deionized water and anhydrous ethanol to completely remove residual impurities. Finally, dry it in a drying oven to a constant weight. After drying, place the product in a tube furnace under an argon atmosphere, set the temperature to 150 °C, and heat for 60 minutes to further process the product. Finally, after heating is complete, wait for the furnace body to cool to room temperature, collect the dark precipitate and perform three rounds of centrifugation, during which deionized water and ethanol are used for rinsing. Dry the final product at 80 °C for 24 hours and anneal it at 150 °C for 6 hours in an Ar atmosphere. Then, the target product of MoS₂ is obtained.

2.3 Synthesis of MoS₂ with S vacancies (V_S-MoS₂)

Take an appropriate amount of pre-synthesized MoS₂ and place it in the center of a tube furnace. Heat up to 700 °C at a heating rate of 10 °C min⁻¹ in an H₂ atmosphere, with a H₂ gas flow rate set at 100 sccm (cm³ min⁻¹). Wait for the temperature to increase to 700 °C and maintain it for 2 hours. After the insulation is complete, the tube furnace is cooled to room temperature to obtain MoS₂ containing S defects, denoted as V_S-MoS₂.

2.4 Synthesis of transition metal-doped molybdenum sulfide with sulfur vacancies (TM@V_S-MoS₂; TM = Sc, Cu, V)

Evenly disperse 200 mg of V_S-MoS₂ in 40 mL of ethylene glycol. Subsequently, take 50 mL of aqueous solutions containing 10% atomic ratio of VO(AcAc)₂, CuCl₂, and Sc (NO₃)₃ and add them dropwise to different MoS₂ dispersions and thoroughly stir them to ensure uniform mixing. Next, quickly add 200 μL of N₂H₄·H₂O and 300 μL of 1 M KOH solution to the mixture and continue stirring for 2 hours. Afterwards, transfer the mixture to a 200 mL PTFE lined autoclave and react for 3 hours under hydrothermal conditions at 90 °C. After the reaction is complete, centrifuge the product and wash multiple times with deionized water and anhydrous ethanol to completely remove residual impurities. Finally, dry it in a drying oven to a constant weight. The obtained samples are recorded as TM@V_S-MoS₂ (TM = Sc, Cu, V).

2.5 Model of density functional theory (DFT) calculation

The calculation parameters used, including MoS₂ and V_S-MoS₂ structural models (Fig. S3 and S4†), as well as the calculation formulae for substance adsorption and ammonia synthesis activity, are listed in the ESI.†^31,32

The details of stability of the MoS₂ structure after constructing S vacancies and the calculation of formation energy (E_f) are also provided in the ESI.† To evaluate the ease of doping transition metal (TM) into V_S-MoS₂, the formation energy (E_f(TM)) is calculated using the following formula:

E_f(TM) = E_TM@VS-MoS2 − E_TM + E_Mowhere (1)

E_TM@VS-MoS2 represents the energy of TM-doped V_S-MoS₂ structure, while E_TM and E_Mo represent the energy of doped metal atoms and Mo atoms, respectively.

Previous work has shown that molybdenum sulfide with three vacancies is most favorable for nitrogen reduction, and considering that the experimentally synthesized molybdenum sulfide contains more than one vacancy, the calculations in this paper are based on the doping of transition metal (TM) atoms into molybdenum sulfide with three vacancies (referred to as V_S-MoS₂) to investigate the NRR catalytic activity and reaction mechanism of molybdenum sulfide doped with metals and multiple sulfur vacancies.^24,33 Transition metal (TM) atoms were doped at sites 1 and 2 on V_S-MoS₂, and the formation energies of TM atom-doped structures were compared. The results are listed in Table S1.† The data show that three types of TM atoms (Sc, Cu, and V) are doped into the structure, and the formation energy of TM at position 1 is the lowest. The formation energies of Sc@V_S-MoS₂, Cu@V_S-MoS₂ and V@V_S-MoS₂ are 1.240 eV, 0.151 eV, and 0.971 eV, respectively. The lower formation energy facilitates the combination of doped atoms with the original structure, making it easier to form stable doped structures.^34,35

3. Results and discussion

3.1 Structural characterization of TM@V_S-MoS₂ (TM = Sc, Cu, V)

The morphology and structure of TM@V_S-MoS₂ (TM = Cu, V, Sc) was observed using scanning electron microscopy (SEM). It can be seen from Fig. 1a–d that the morphology of the four samples is a spherical cluster structure composed of plate-like crystals, which indicates that transition metal doping did not significantly alter the structure of V_S-MoS₂. The microstructure of TM@V_S-MoS₂ (TM = Cu, V, Sc) was observed using high-resolution transmission electron microscopy (HRTEM) (Fig. 1e–g). Among them, 0.08 nm, 0.24 nm, and 0.18 nm belong to the (002), (006), and (101) crystal planes of MoS₂, respectively.³⁶ It is worth noting that the three transition metal elements Cu, V, and Sc are uniformly dispersed in the crystal structure of V_S-MoS₂, confirming the successful synthesis of TM@V_S-MoS₂ (TM = Cu, V, Sc) catalysts.

Fig. 1 SEM images of (a) V_S-MoS₂, (b) Sc@V_S-MoS₂, (c) Cu@V_S-MoS₂, and (d) V@V_S-MoS₂; HRTEM images and elemental mapping of (e) Sc@V_S-MoS₂, (f) Cu@V_S-MoS₂, and (g) V@V_S-MoS₂.

In order to further determine the presence of S vacancies in the sample, high angle annular dark field scanning transmission electron microscopy (HAADF-STEM) analysis was performed on the original V_S-MoS₂, and the results are shown in Fig. 2a–c. High quality atomic resolution images show that the atomic arrangement of the V_S-MoS₂ sample is very regular (Fig. 2a). The magnified images from the local area (Fig. 2b and c) and the intensity distribution curve (Fig. 2e) along the selected rectangular area in the panel reflection clearly indicate the presence of S atomic vacancies in the lattice growth direction. Electroparamagnetic resonance (EPR) characterization also showed that the original MoS₂ had no obvious EPR signal, while V_S-MoS₂ exhibited a strong EPR response at a field strength of 3505.00 gauss (g = 2.000), corresponding to the presence of S vacancies.³ The above results demonstrate the successful synthesis of MoS₂-based materials containing S vacancies.

Fig. 2 (a) HAADF-STEM image of V_S-MoS₂; (b and c) enlarged areas of the highlighted region in (a); (d) EPR spectra of as-prepared MoS₂ and V_S-MoS₂; (e) the intensity profiles along the selected rectangular regions in panels (a–c), reflecting the missed surface S atoms in V_S-MoS₂.

The crystallographic phase of the material was determined by X-ray diffraction (XRD) analysis. As shown in Fig. 3a, the XRD curve of MoS₂ is consistent with the standard card (PDF no. 37-1492), with clear diffraction peaks of (002), (101), (006), and (110) crystal planes at 14.378°, 33.508°, 44.151°, and 58.334°, indicating the successful preparation of MoS₂.^36,37 Compared with the pristine MoS₂, TM@V_S-MoS₂ (TM = Cu, V, Sc) does not have any additional peaks corresponding to the corresponding sulfides/oxides, indicating that the doped elements are effectively bound in MoS₂.³⁸ In the electron paramagnetic resonance (EPR) spectrum shown in Fig. 3c, S vacancies characterized by g = 2.000 were clearly identified for TM@V_S-MoS₂ (TM = Cu, V, Sc).³⁹

Fig. 3 (a) XRD patterns of MoS₂, V_S-MoS₂ and TM@V_S-MoS₂ (TM = Sc, Cu, V); (b) XPS survey spectra of MoS₂; (c) EPR spectra of MoS₂, V_S-MoS₂ and TM@V_S-MoS₂ (TM = Sc, Cu, V); (d–f) XPS spectra of (d) C 1s, (e) Mo 3d, (f) S 2p, (g) Sc 2p, (h) Cu 2p, and (i) V 2p.

Subsequently, the elemental composition and oxidation state of the catalyst surface were carefully examined using X-ray photoelectron spectroscopy (XPS) (Fig. 3b and d–i). The XPS analysis of the Mo signal (Fig. 3e) revealed two distinct characteristic peaks with binding energies of 233.5 eV and 230.3 eV, respectively, attributed to Mo 3d_3/2 and Mo 3d_5/2.^40,41 As shown in Fig. 3f, the characteristic peaks with binding energies of 163.6 eV and 162.2 eV can be assigned to the S 2p_1/2 and S 2p_3/2 orbitals, respectively.⁴² In the XPS analysis of TM@V_S-MoS₂ (TM = Cu, V, Sc), characteristic peaks of Sc, Cu, and V were detected (corresponding to Fig. 3g–i). The binding energies of Sc 2p orbitals can be located at 407.5 eV and 402.4 eV.⁴³ XPS analysis detected signal responses of Cu 2p at 953.2 eV and 932.2 eV.⁴⁴ The characteristic peaks detected at binding energies of 524.3 eV and 516.7 eV can be attributed to the V 2p orbital.⁴⁵ It is worth noting that compared to the original MoS₂, the XPS binding energies of Mo 3d and S 2p orbitals in the transition metal doped sulfur defect structure (TM@Vs-MoS₂, TM = Sc, Cu, V) exhibit significant negative shifts (Fig. S5 and S6†). This phenomenon indicates that transition metal atoms anchor through sulfur vacancies (V_S) and inject additional electrons into the MoS₂ lattice as electron donors, resulting in an increase in the surface electron density of Mo and S atoms. This type of charge redistribution can effectively regulate the local charge density of Mo active centers, thereby optimizing the σ-donation/π-feedback synergistic adsorption mode of N₂ molecules.^41,42

In addition, the composition and elemental content of five samples were tested by ICP, and the results are listed in Table S2.† It can be seen that the content of elements is basically as expected, and the S content in V_S-MoS₂ has decreased compared to the original MoS₂ due to the construction of vacancies, which indirectly confirms the presence of S vacancies in the sample.

3.2 Reaction activity, stability and recycling as well as photoelectric properties of TM@V_S-MoS₂ (TM = Sc, Cu, V) for the pNRR

A series of characterization experiments were conducted on the photoelectric performance of photocatalysts. As shown in Fig. 4a, UV-vis diffuse reflectance spectroscopy (UV-vis DRS) analysis showed that all catalysts had no obvious absorption edges, indicating good absorption performance for light with wavelengths of 200–700 nm. From the electrochemical impedance spectroscopy (EIS) spectra of the photocatalyst (Fig. 4b), it can be observed that compared to the original MoS₂ and V_S-MoS₂, TM@V_S-MoS₂ (TM = Sc, Cu, V) has a smaller arc radius. In particular, Sc@V_S-MoS₂ exhibits the smallest arc radius, indicating the lowest interfacial charge transfer resistance.⁴⁶ As shown in Fig. 4c, the dynamics of charge carriers within the photocatalyst were studied using photoluminescence spectroscopy (PL). Sc@V_S-MoS₂ has the lowest signal intensity, indicating the lowest degree of electron and hole recombination.⁴⁷

Fig. 4 (a) UV-vis DRS spectra; (b) Nyquist plots; (c) PL spectra of MoS₂, V_S-MoS₂ and TM@V_S-MoS₂ (TM = Sc, Cu, V); (d) photocatalytic ammonia synthesis performance of MoS₂, V_S-MoS₂, TM@MoS₂ and TM@V_S-MoS₂ (TM = Sc, Cu, V); (e) time-dependent photocatalytic ammonia production of MoS₂, V_S-MoS₂ and TM@V_S-MoS₂ (TM = Sc, Cu, V); (f) comparison of production of ammonia, oxygen, and hydrogen catalyzed by Sc@V_S-MoS₂; (g) comparison of ammonia production efficiency under different conditions of light and atmosphere catalyzed by Sc@V_S-MoS₂; (h) comparison of ammonia production efficiency under different solvents catalyzed by Sc@V_S-MoS₂; (i) reuse experiments of photocatalytic nitrogen reduction by TM@V_S-MoS₂ (TM = Sc, Cu, V).

This work uses the Nessler indicator method and indophenol blue method to detect the product ammonia. Comparing with the test results of the previous samples, it was found that the activity detected by the two methods was basically similar. Therefore, the results determined by the Nessler indicator method were used for subsequent measurements.⁴⁸ After 1 hour of illumination, the activity comparison of MoS₂, V_S-MoS₂, TM@MoS₂ and TM@V_S-MoS₂ (TM = Sc, Cu, V) materials in catalyzing nitrogen reduction to synthesize ammonia is shown in Fig. 4d. It can be seen that the activity of pristine MoS₂ was only 20.18 μmol g⁻¹ h⁻¹, while the activity of V_S-MoS₂ increased to 55.72 μmol g⁻¹ h⁻¹, indicating that S vacancies can effectively enhance the photocatalytic performance of MoS₂. The nitrogen reduction activity of TM@MoS₂ (TM = Sc, Cu, V) materials is in the range of 34.38 μmol g⁻¹ h⁻¹ to 46.16 μmol g⁻¹ h⁻¹, which is also higher than that of the original MoS₂, indicating that metal doping can also increase the activity of MoS₂. In particular, the photocatalytic nitrogen reduction activity data of Sc@V_S-MoS₂, Cu@V_S-MoS₂ and V@V_S-MoS₂ are 88.64 μmol g⁻¹ h⁻¹,101.32 μmol g⁻¹ h⁻¹, and 124.26 μmol g⁻¹ h⁻¹, respectively, which are much higher than those of MoS₂, V_S-MoS₂, TM@MoS₂. This fully demonstrates that the comprehensive use of S vacancies and metal doping can significantly enhance the photocatalytic performance of MoS₂ (with a maximum activity improvement of nearly 5 times). To verify the stability of the catalyst, the activity data of MoS₂, V_S-MoS₂, and TM@V_S-MoS₂ (TM = Sc, Cu, V) catalyzing the nitrogen reduction reaction for 12 hours are shown in Fig. 4e. It can be seen that the ammonia production of the five catalysts increases linearly with the extension of light exposure time, and after 12 hours of reaction, the overall ammonia yield decay is less than 9%. The above results fully demonstrate the good reliability and stability of MoS₂-based photocatalytic synthesis of ammonia. In addition, to verify the effect of side reactions on the influence of TM@V_S-MoS₂ photocatalyst in nitrogen reduction was investigated using Sc@V_S-MoS₂ as an example for the hydrogen (HER) and oxygen (OER) production reactions (Fig. 4f). The results showed that the yield of ammonia was significantly higher than that of by-products hydrogen and oxygen (more than 25 times), indicating that the competitive reaction between hydrogen and oxygen production was well suppressed.⁴⁹

Taking the Sc@V_S-MoS₂ catalyst as an example, two sets of comparative experiments were conducted to verify the true activity of ammonia synthesis. The results indicate that the yield of ammonia synthesis can be ignored by introducing nitrogen gas under non-light conditions and argon gas under light conditions, confirming that the nitrogen in the photocatalytic nitrogen reduction of ammonia synthesis comes from the introduced nitrogen gas (Fig. 4g). After replacing the solvent with organic solvents (DMF and methylamine), the yield of ammonia synthesis decreased from 124.26 μmol g⁻¹ h⁻¹ to <5.0 μmol g⁻¹ h⁻¹ when the solvent was water, fully indicating that the hydrogen in the photocatalytic nitrogen reduction of ammonia synthesis comes from water in the solvent (Fig. 4h).⁵⁰

After one hour of reaction, nitrate (NO₃⁻) and nitrite (NO₂⁻) byproducts in the nitrogen reduction process were detected. The results are shown in Fig. S7,† and it was found that they were basically not produced. The yields of hydrazine (N₂H₄) and hydrogen (H₂) were also measured, and it was found that their yields were basically negligible compared to ammonia (Fig. S8†). In addition, the reusability study of TM@V_S-MoS₂ (TM = Sc, Ti, V) in catalyzing nitrogen reduction was conducted, and the results are shown in Fig. 4i. After three cycles, the activity of the three catalysts decreased by less than 10%, indicating that TM@V_S-MoS₂ has not only excellent stability but also certain reusability.

3.3 Analysis of the nitrogen reduction mechanism based on DFT and in situ FTIR

In order to further investigate the activity of TM doping in enhancing the catalytic nitrogen reduction synthesis of ammonia by the molybdenum sulfide vacancies and compare the inherent differences between different metal doping, detailed theoretical calculations and analysis were conducted.

3.3.1 Structural and property analyses of TM@V_S-MoS₂ (TM = Sc, Ti, V). As shown in Fig. 5a–c, when three types of TM atoms (Sc, Cu, and V) were doped into V_S-MoS₂, it can be observed from the top and side views that the structure has not undergone significant deformation, indicating that the influence of doped atoms on the structure is relatively small. In the side view, it can be seen that the positions of Sc, Cu, and V are similar to the previously exposed Mo atoms and do not protrude from the surface. From the perspective of dynamic stability, the energy fluctuations of three types of TM@V_S-MoS₂ (TM = Sc, Ti, and V) structure are all within the range of 0.02 eV at a temperature of 400 K and a time of 5 ps. Very small fluctuations and no significant deformation were observed in the simulated structure, which show that the structure has good thermal stability (Fig. 5d–f).⁵¹

Fig. 5 Top and side views of (a) Sc@V_S-MoS₂, (b) Cu@V_S-MoS₂, and (c) V@V_S-MoS₂; dynamic stability of (d) Sc@V_S-MoS₂, (e) Cu@V_S-MoS₂, and (f) V@V_S-MoS₂ at 400 K (1 fs step and 5000 fs), with the beginning and the end of the structural model in the middle; deformation charge density diagrams for (g) Sc@V_S-MoS₂, (h) Cu@V_S-MoS₂, and (i) V@V_S-MoS₂ (isovalue = 0.060 e Å⁻³, blue indicates charge accumulation, red indicates charge departure, and the numbers are the number of charges transferred).

The differential charge density can further aid understanding of the charge changes before and after TM atom doping in the structure, and there are some differences in the charge transfer exhibited by different TM atoms. From Fig. 5g–i, it can be seen that the Sc atom doped in the structure can transfer the most electrons to the surrounding atoms, at 0.321e, followed by Cu atoms at 0.080e. V atoms exhibit different electronic properties, −0.093e indicates attraction to the surrounding electrons. The departure of charges facilitates the formation of positive charge centers, which can better accommodate lone pair electrons of N₂ and thus have a stronger adsorption effect. Negative charge centers may have weaker adsorption of N₂, but more electrons can feedback to the antibonding orbitals of N≡N bonds, weakening the bond strength. Sc and Cu form positive charge centers, while V forms negative charge centers. Changes in the surface charge layout can affect the adsorption and activation of N₂ molecules.

3.3.2 Adsorption selectivity of N₂ and H on TM@V_S-MoS₂ (TM = Sc, Ti, V). A comparison is made of the adsorption free energy of N₂ molecules and H atoms for V_S-MoS₂ and three types of TM@V_S-MoS₂ (TM = Sc, Ti, V) (Table 1), which is used to understand the adsorption selectivity of different structures for N₂ and H, in order to determine their reaction selectivity for the NRR and HER. It can be seen from the table that the ability of V_S-MoS₂ to adsorb N₂ molecules laterally and end wise is similar, both stronger than the adsorption of H. The adsorption energy of TM@V_S-MoS₂ (TM = Sc, Ti, V) for H has decreased from −0.659 eV to −0.459 eV, −0.381 eV and −0.423 eV, significantly inhibiting the progression of the HER and facilitated the NRR. On the other hand, its ability to adsorb N₂ molecules towards the end is greatly reduced, while its ability to adsorb N₂ molecules towards the side is still strong, with the latter being much greater than the former, indicating that nitrogen tends to adsorb laterally on TM@V_S-MoS₂ (TM = Sc, Ti, V).

Table 1 Free energies of adsorption of N₂ molecules and H atoms on V_S-MoS₂ and TM@V_S-MoS₂ (TM = Sc, Cu, V)

Structural	ΔGH* (eV)			ΔN≡N (Å)
VS-MoS2	−0.659	−0.907	−0.888	+0.096
Sc@VS-MoS2	−0.459	−0.152	−0.866	+0.192
Cu@VS-MoS2	−0.381	−0.126	−0.841	+0.149
V@VS-MoS2	−0.423	0.099	−0.695	+0.129

---

When N₂ is adsorbed on TM@V_S-MoS₂ (TM = Sc, Ti, V), its N≡N bond length will be elongated, which can preliminarily determine whether N₂ molecules are adsorbed and activated.⁵² As shown in Table 1, compared with the N≡N bond length of 1.108 Å before adsorption, the bond lengths of N₂ laterally adsorbed on V_S-MoS₂ and TM@V_S-MoS₂ (TM = Sc, Ti, V) were stretched by 0.096 Å, 0.192 Å, 0.149 Å, and 0.129 Å, respectively, with significant elongation, indicating that these structures can not only adsorb nitrogen gas well but also have certain activation ability, which is conducive to the subsequent catalytic NRR. Thus, the following discussion focuses on the structural properties and NRR activity of three types of TM@V_S-MoS₂.

3.3.3 Adsorption and activation of N₂ on TM@V_S-MoS₂ (TM = Sc, Ti, V). The differential charge density of N₂ molecules adsorbed on the catalyst surface can provide information on the degree of N₂ activation and the differences in the activation of N₂ by different structures. Differential charge density analysis diagrams of N₂ molecules adsorbed laterally on TM@V_S-MoS₂ (TM = Sc, Ti, V) are shown in Fig. 6a–c. From the adsorption configuration of the molecule, one end of N₂ forms a bond with TM atoms, and the other end forms a bond with Mo atoms. The effect of bimetallic atoms not only enhances the adsorption of nitrogen but also better activates nitrogen. When N₂ is adsorbed on the surface of the catalyst, the surface charge will undergo transfer, and the amount of charge transferred from Sc@V_S-MoS₂ to N₂ molecules is the highest at 0.533e. The amount of charge transferred from Cu@V_S-MoS₂ and V@V_S-MoS₂ to N₂ is 0.352e and 0.272e, respectively, both of which are higher than 0.264e of the original V_S-MoS₂ (Fig. S9†). These indicate that the doping of TM atoms can regulate the surface charge, effectively transferring the charge to N₂ molecules, stretching the bond length of N≡N bonds, thereby weakening the strength of N≡N bonds and activating N₂ molecules.

Fig. 6 Deformation charge density diagrams of N₂ side-on adsorption on (a) Sc@V_S-MoS₂, (b) Cu@V_S-MoS₂, and (c) V@V_S-MoS₂ (isovalue = 0.060 e Å⁻³, blue indicates charge accumulation, red indicates charge departure, and the numbers are the number of charges transferred); PDOS diagrams of N₂ lateral adsorption on (d) Sc@V_S-MoS₂, (e) Cu@V_S-MoS₂, and (f) V@V_S-MoS₂; Gibbs free energy plots of enzymatic, consecutive, and mixed reaction paths of the NRR on (g) Sc@V_S-MoS₂, (h) Cu@V_S-MoS₂, and (i) V@V_S-MoS₂.

As shown in Fig. 6d–f, the PDOS diagram of N₂ laterally adsorbed on three types of TM@V_S-MoS₂ (TM = Sc, Cu, V) clearly shows the degree of energy overlap between the d orbitals of TM and Mo atoms and the p orbitals of N atoms (the PDOS diagram and related explanations of N₂ lateral adsorption on V_S-MoS₂ are listed in Fig. S10 in the ESI†). The empty d orbitals in TMs can accommodate lone pair electrons of N₂ molecules, while the d orbitals with electrons can feedback to the antibonding orbitals in N₂. This interaction can weaken the strength of the N≡N bond. Sc 3d orbitals in Sc@V_S-MoS₂ overlap significantly with the N 2p orbitals at −4.8–0.6 eV, and there is more overlap between the Mo 4d orbitals and the N 2p orbitals, indicating strong hybridization between them and their joint action on N₂ molecules. The 3d orbitals of Cu and V have more electrons than Sc, resulting in a significant increase in the overlap between Cu 3d orbitals, V 3d orbitals, and N 2p orbitals. Because N₂ molecules have one end adsorbed on Mo atoms, there is a significant hybridization between Mo 4d orbitals and N 2p orbitals in all three structures. The d orbitals of TM and Mo atoms can jointly act on N₂ molecules, thereby activating N₂ molecules, which is beneficial for the subsequent NRR.

3.3.4 reaction pathway of TM@V_S-MoS₂ (TM = Sc, Ti, V)-catalyzed nitrogen reduction to ammonia by combined DFT and in situ FTIR. The different adsorption configurations of nitrogen on catalysts can lead to the NRR proceeding along different reaction pathways. According to Table 1, TM@V_S-MoS₂ preferentially adsorbs N₂ molecules laterally, and the end adsorption will be unstable and less than H adsorption. When the two ends of the N₂ molecule are adsorbed on the structure, that is, laterally adsorbed, the two N atoms are equivalent in the first hydrogenation reaction, and the NRR will proceed along enzymatic, continuous, and mixed pathways (Fig. S11†).^53,54 The hydrogenation mode of the enzymatic pathway (red) is the same as the alternating pathway, which involves hydrogenation at both ends sequentially. The continuous path (purple) is due to the fact that the N atom after the first hydrogenation step may be more easily reacted with the subsequent H atom, so the hydrogenation mode of the continuous path is consistent with that of the distal path. The mixed path (green) takes into account the possibility of crossover among the first four paths. The nitrogen reduction reaction on TM@V_S-MoS₂ (TM = Sc, Ti, V) proceeds along enzymatic, continuous, and mixed pathways. The relevant Gibbs free energy data and free energy plots are shown in Table 2 and Fig. 6g–i.

Table 2 Gibbs free energy of the reaction path of the NRR on V_S-MoS₂ and TM@V_S-MoS₂ (TM = Sc, Cu, V)

Reaction steps	Gibbs free energies (ΔG, eV)
	VS-MoS2	Sc@VS-MoS2	Cu@VS-MoS2	V@VS-MoS2
Enzymatic
R1: N2 + * → *N2	−0.551	−0.866	−0.841	−0.695
R2: *N2 + H^+ + e^− → *NNH	0.014	−0.557	−0.493	−0.151
R3: *NNH + H^+ + e^− → *NHNH	−0.310	−1.232	−0.792	−0.985
R4: *NHNH + H^+ + e^− → *NHNH2	−0.712	−1.345	−0.835	−1.124
R5: *NHNH2 + H^+ + e^− → *NH2NH2	−0.076	−1.194	−0.299	−0.780
R6: *NH2NH2 + H^+ + e^− → *NH2 + NH3	−1.802	−1.263	−1.205	−1.667
R7: *NH2 + H^+ + e^− → *NH3	−1.284	−0.870	−0.788	−0.986
Consecutive
R1: N2 + * → *N2	−0.551	−0.866	−0.841	−0.695
R2: *N2 + H^+ + e^− → N*NH	0.014	−0.557	−0.493	−0.151
R3: *NH + H^+ + e^− → *NH2	−1.106	−0.657	−0.804	−0.598
R4: NH2 + H^+ + e^− → *N + NH3	−1.859	−1.193	−1.028	−0.862
R5: *N + H^+ + e^− → *NH	−2.864	−1.732	−1.768	−1.514
R6: *NH + H^+ + e^− → *NH2	−1.802	−1.263	−1.205	−1.667
R7: *NH2 + H^+ + e^− → *NH3	−1.284	−0.870	−0.788	−0.986
Total: N2 + 6H^+ + 6e^− → 2NH3	−0.460	−0.460	−0.460	−0.460

---

The usual reason for restricting the NRR is that the potential determination step (PDS) is too high, which prevents the reaction from proceeding normally. According to previous reports, the first hydrogenation reaction (*N₂ + H⁺ + e⁻ → *NNH) and the final hydrogenation reaction (*NH₂ + H⁺ + e⁻ → NH₃) of the NRR are generally the PDS of the entire reaction.⁵⁵ By comparing the PDS of the three NRR pathways on V_S-MoS₂, it was found that the potential barrier was the highest when generating *NH₂NH₂ and *NH₂ intermediates, and these two hydrogenation reactions were the PDS of the entire reaction (Fig. S12 and S13†). The free energy for generating *NH₂NH₂ in the enzymatic pathway is 0.635 eV, while the maximum free energy (ΔG_max) for generating *NH₂ is 1.062 eV. The excessively high PDS hinders the continuous pathway. The U_L of V_S-MoS₂ is −0.635 V, which does not dominate the competition with the HER. The weak adsorption of N₂ on V_S-MoS₂, strong competitive adsorption of H, and high PDS are some of the reasons for its low catalytic NRR activity.

Sc@V_S-MoS₂ requires an energy input of 0.310 eV for the first hydrogenation reaction of the catalyst, which is slightly lower than the 0.322 eV of the original V_S-MoS₂ (Fig. 6g). In the enzymatic pathway, the second and third steps of hydrogenation are exothermic, with energy released at 0.675 eV and 0.113 eV, until the fourth step of hydrogenation (step 5) in the enzymatic pathway generates a *NH₂NH₂ intermediate with an energy increase, requiring crossing an energy barrier of 0.151 eV. The subsequent fifth step of hydrogenation is exothermic, and the energy of the hydrogenation step that generates *NH₃ increases by 0.393 eV (ΔG_max), which is the PDS of the entire path. The highest free energy of the continuous path occurs in the process of generating *NH₂ from *NH, requiring an energy input of 0.469 eV, which is the PDS of this path. Sc@V_S-MoS₂ will preferentially undergo the NRR along the enzymatic pathway, with an U_L of −0.393 V, and mixed pathways can also occur.

Cu@V_S-MoS₂ has a free energy of 0.348 eV for the first step of NRR hydrogenation, which is higher than that of Sc@V_S-MoS₂ (Fig. 6h). In the enzymatic pathway, there is an energy increase in the generation of *NH₂–NH₂ from *NH–NH₂, with a ΔG_max of 0.536 eV, which is the PDS of this pathway, indicating that a large energy input is required for the generation of *NH₂–NH₂. When generating the last NH₃, it is necessary to cross an energy barrier of 0.417 eV, and other hydrogenation steps exhibit exothermic behavior. The PDS of the continuous pathway appears in the generation of *NH₂ intermediate, with a ΔG_max of 0.564 eV. Except for the first and last steps of hydrogenation, all other steps exhibit exothermic behavior. The most exothermic process is the generation of *NH from *NH–NH₂ in the mixed pathway at 1.368 eV. The enzymatic pathway is likely to transition into a continuous pathway at this step. Cu@V_S-MoS₂ is more likely to propagate along a continuous path, with an U_L of −0.564 V, which is also lower than the −0.635 V of V_S-MoS₂.

The NRR occurring on V@V_S-MoS₂ is similar to the previous two structures (Fig. 6i). The free energy of the first hydrogenation step is 0.544 eV, which is 0.222 eV higher than the original V_S-MoS₂. The higher initial hydrogenation step makes the reaction difficult to proceed. In the subsequent hydrogenation steps, the energy of the enzymatic pathway increased by 0.345 eV during the generation of the *NH₂–NH₂ intermediate, followed by the largest decrease in energy during the generation of *NH₂ (0.887 eV) and an increase in energy during the generation of the last *NH₃, with a ΔG_max of 0.680 eV, which is the PDS of this pathway. The energy of continuous paths generating N–*NH₂, *N, *NH and *NH₂ is decreased by 0.447 eV, 0.264 eV, 0.652 eV, and 0.153 eV, respectively. Unlike the previous two structures, the generation of *NH₂ is exothermic. The PDS of V@V_S-MoS₂ appears in the final hydrogenation reaction, with a U_L of −0.680 V. Due to the greater decrease in energy for the generation of NH–*NH in the second hydrogenation reaction, the NRR tends to proceed along the enzymatic pathway.

To verify the correctness of theoretical calculations for the NRR pathway, in situ infrared spectroscopy of the Sc@V_S-MoS₂ photocatalytic nitrogen reduction process was performed, and the results are shown in Fig. 7a. Two distinct absorption signals can be observed at 1118 cm⁻¹ and 1295 cm⁻¹, attributed to N–N stretching and –NH₂ oscillation, respectively.⁵⁶ The peak intensity increases with the prolongation of reaction time, indicating that more active sites gradually participate in the catalytic cleavage of N–N triple bonds and the production of –N₂H_y (1 ≤ y ≤ 4) species on the catalyst surface.⁵⁷ This is highly consistent with the intermediates in steps 2–5 of the enzymatic mechanism, further proving that the calculated enzymatic mechanism is the most likely reaction pathway for the Sc@V_S-MoS₂ photocatalytic nitrogen reduction. In addition, the adsorption band located at 1420 cm⁻¹ corresponds to the asymmetric deformation vibration of NH₄⁺ generated during the nitrogen reduction reaction.⁵⁶ The absorption peak at 1638 cm⁻¹ is due to the O–H stretching and H–O–H bending vibration of water. In particular, the frequency of the 2780 cm⁻¹ absorption band is attributed to the chemical adsorption of N₂, indicating that Sc@V_S-MoS₂ can effectively adsorb and activate N₂ molecules.⁵⁷

Fig. 7 (a) In situ infrared spectroscopy of the Sc@V_S-MoS₂ photocatalytic nitrogen reduction process; (b) the schematic diagram for TM@V_S-MoS₂ catalyzing the pNRR: metal doping and S vacancies synergistically activate N₂ and inhibit water decomposition, thereby achieving efficient nitrogen reduction.

In summary, DFT calculations show that Sc@V_S-MoS₂ and V@V_S-MoS₂ follow enzymatic pathways, while Cu@V_S-MoS₂ tends to follow a continuous pathway. And taking Sc@V_S-MoS₂ as an example, it was verified through in situ infrared spectroscopy, and the results were suppressed. In addition, Sc@V_S-MoS₂ exhibits the best reaction activity, with a U_L of only −0.393 V, which is superior to the original V_S-MoS₂, Cu@V_S-MoS₂ and V@V_S-MoS₂. This is related to Sc@V_S-MoS₂ and exhibits excellent adsorption and activation of N₂, with a charge transfer of 0.533e to N₂, resulting in a lower free energy for the first hydrogenation reaction. The gradual decrease in subsequent free energy also indicates a stronger ability for electron transfer. In addition, the adsorption free energy of H by TM@V_S-MoS₂ (TM = Sc, Cu, V) is lower than that of the original V_S-MoS₂, which is −0.659 eV (Tables 1 and 2), effectively suppressing the hydrogen evolution reaction. From the Gibbs free energy data, it can also be concluded that, the energies of the final step of *NH₃ dissociation into NH₃ on TM@V_S-MoS₂ (TM = Sc, Cu, V) are 0.410 eV, 0.328 eV, and 0.526 eV, respectively, which are much lower than the 1.599 eV of the original V_S-MoS₂. This can effectively reduce the desorption energy of NH₃ and facilitate the continuous progress of the reaction. The schematic diagram for TM@V_S-MoS₂ catalyzing the pNRR is listed in Fig. 7b, in which metal doping and S vacancies synergistically activate N₂ and inhibit water decomposition, thereby achieving efficient nitrogen reduction.

4. Conclusions

This paper is based on MoS₂ containing S vacancies and doping with three types of transition metals, Sc, Cu, and V, to prepare transition metal-doped MoS₂ with S vacancies (TM@V_S-MoS₂, where TM = Sc, Ti, V), which is applied for photocatalytic nitrogen reduction. Through a series of characterization methods such as HAADF-STEM, XRD, SEM, TEM, XPS, EPR, etc., it has been demonstrated that TM@V_S-MoS₂ (TM = Sc, Ti, V) has good crystal structure, morphology, and defect structure. The nitrogen reduction experiment showed that the activity of TM@V_S-MoS₂ (TM = Sc, Ti, V) ranges from 88.64 μmol g⁻¹ h⁻¹ to 124.26 μmol g⁻¹ h⁻¹, with the order of activity being Sc@V_S-MoS₂ > Cu@V_S-MoS₂ > V@V_S-MoS₂, which is consistent with the order of the photoelectric properties of the material. Moreover, there is a maximum improvement of activation up to 5 times than MoS₂, V_S-MoS₂, and TM@MoS₂, which fully demonstrates that the comprehensive use of S vacancies and metal doping can significantly enhance the photocatalytic performance of MoS₂. In addition, the production of ammonia is significantly higher than the yields of by-products hydrogen and oxygen (more than 25 times), indicating that the competitive reaction between hydrogen and oxygen production has been well suppressed. DFT calculations show that the adsorption free energy of TM@V_S-MoS₂ (TM = Sc, Cu, V) is lower than that of the original V_S-MoS₂ (−0.659 eV), which can effectively suppress the hydrogen evolution reaction. The energies of the final step of *NH₃ dissociation into NH₃ on TM@V_S-MoS₂ (TM = Sc, Cu, V) are 0.410 eV, 0.328 eV, and 0.526 eV, respectively, which are much lower than the 1.599 eV of original V_S-MoS₂. This can effectively reduce the desorption energy of NH₃ and facilitate the continuous progress of the reaction. Among them, the main reason why Sc@V_S-MoS₂ exhibits the best reaction activity is that (1) for its rate control step (PDS) ability, U_L is only −0.393 V, which is superior to the original V_S-MoS₂, Cu@V_S-MoS₂ and V@V_S-MoS₂; (2) the charge transferred from Sc@V_S-MoS₂ to N₂ is as high as 0.533e, resulting in a lower free energy for the first hydrogenation reaction and strong electron transfer ability, thus exhibiting excellent adsorption and activation performance of N₂.

Data availability

The authors confirm that the data supporting the findings of this study are available within the article and/or its ESI.† And the data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (22278371 and 22478349).

References

1. L. Q. Li, C. Tang, H. Y. Jin, K. Davey and S. Z. Qiao, Main-group elements boost electrochemical nitrogen fixation, Chem, 2021, 7, 3232–3255.
2. Z. Y. Chen, Z. L. Huang, J. Y. Yang, Y. Meng, B. Xie, Z. M. Ni and S. J. Xia, Experimental and theoretical investigation on facet-dependent MoO₂/BiOBr Z-scheme heterojunction photocatalytic nitrogen reduction: modulation of bulk charge separation efficiency by built-in electric field intensity, J. Mater. Chem. A, 2024, 12, 16877–16891.
3. X. R. Zhang, Y. H. Lyu, H. J. Zhou, J. Y. Zheng, A. B. Huang, J. J. Ding, C. Xie, R. D. Marco, N. Tsud, V. Kalinovych, S. P. Jiang, L. M. Dai and S. Y. Wang, Photoelectrochemical N₂-to-NH₃ Fixation with High Efficiency and Rates via Optimized Si-Based System at Positive Potential versus Li⁰/⁺, Adv. Mater., 2023, 35, 2211894.
4. R. Shi, X. R. Zhang, G. I. N. Waterhouse, Y. X. Zhao and T. R. Zhang, The Journey toward Low Temperature, Low Pressure Catalytic Nitrogen Fixation, Adv. Energy Mater., 2020, 10, 2000659.
5. H. W. Liao, Y. L. Zhou, Z. Chen, S. Sharshir, S. M. Osman, C. Wang, M. An, Y. Yamauchi, Y. Asakura and Z. H. Yuan, Harnessing the synergistic power of Ce₂S₃/TiO₂ S-scheme heterojunctions for profound C–O bond cleavage in lignin model compounds, ACS Catal., 2024, 14, 5539–5549.
6. M. F. Wang, S. S. Liu, T. Qian, J. Liu, J. Q. Zhou, H. Q. Ji, J. Xiong, J. Zhong and C. L. Yan, Over 56.55% Faradaic Efficiency of Ambient Ammonia Synthesis Enabled by Positively Shifting the Reaction Potential, Nat. Commun., 2019, 10, 341.
7. C. V. S. Kumar and V. Subramanian, Can Boron Antisites of BNNTs Be an Efficient Metal-Free Catalyst for Nitrogen Fixation? – A DFT Investigation, Phys. Chem. Chem. Phys., 2017, 19, 15377–15387.
8. D. N. Liu, J. H. Wang, S. Bian, Q. Liu, Y. H. Gao, X. Wang, P. K. Chu and X. F. Yu, Photoelectrochemical Synthesis of Ammonia with Black Phosphorus, Adv. Funct. Mater., 2020, 30, 2002731.
9. Z. Liu, M. Z. Zhang, H. N. Wang, D. Y. Cang, X. Q. Ji, B. P. Liu, W. R. Yang, D. Li and J. Q. Liu, Defective Carbon-Doped Boron Nitride Nanosheets for Highly Efficient Electrocatalytic Conversion of N₂ to NH₃, ACS Sustainable Chem. Eng., 2020, 8, 5278–5286.
10. A. B. Abdeta, Q. H. Wu, D. H. Kuo, P. Li, H. Y. Zhang, J. B. Zhang, Z. H. Yuan, J. G. Lin and X. Y. Chen, Activated carbon-supported AgMoOS bimetallic oxysulfide as a catalyst for the photocatalytic hydrogen evolution and pollutants reduction, J. Alloys Compd., 2022, 913, 165287.
11. S. Manzeli, D. Ovchinnikov, D. Pasquier, O. V. Yazyev and A. Kis, 2D Transition Metal Dichalcogenides, Nat. Rev. Mater., 2017, 2, 17033.
12. Z. D. Li, N. H. Attanayake, J. L. Blackburn and E. M. Miller, Carbon Dioxide and Nitrogen Reduction Reactions Using 2D Transition Metal Dichalcogenide (TMDC) and Carbide/Nitride (MXene) Catalysts, Energy Environ. Sci., 2021, 14, 6242–6286.
13. Q. H. Wu, X. R. Wang, J. W. Fu, O. A. Zelekew, A. B. Abdeta, D. H. Kuo, J. B. Zhang, Z. H. Yuan, J. G. Lin and X. Y. Chen, Wool-coiled bimetallic oxysulfide MoSrOS catalyst synthesis for catalytic reduction of toxic organic pollutants and heavy metal ions, J. Sci.:Adv. Mater. Devices, 2021, 6, 578–586.
14. H. M. Hill, A. F. Rigosi, K. T. Rim, G. W. Flynn and T. F. Heinz, Band Alignment in MoS₂/WS₂ Transition Metal Dichalcogenide Heterostructures Probed by Scanning Tunneling Microscopy and Spectroscopy, Nano Lett., 2016, 16, 4831–4837.
15. J. Zhang, X. Y. Tian, M. J. Liu, H. Guo, J. D. Zhou, Q. Y. Fang, Z. Liu, Q. Wu and J. Lou, Cobalt-Modulated Molybdenum–Dinitrogen Interaction in MoS₂ for Catalyzing Ammonia Synthesis, J. Am. Chem. Soc., 2019, 141, 19269–19275.
16. B. H. R. Suryanto, D. Wang, L. M. Azofra, M. Harb, L. Cavallo, R. Jalili, D. R. G. Mitchell, M. Chatti and D. R. MacFarlane, MoS₂ Polymorphic Engineering Enhances Selectivity in the Electrochemical Reduction of Nitrogen to Ammonia, ACS Energy Lett., 2019, 4, 430–435.
17. Z. Y. Lu, S. E. Saji, J. Langley, Y. X. Lin, Z. R. Xie, K. Yang, L. Bao, Y. Y. Sun, S. B. Zhang, Y. H. Ng, L. Song, N. Cox and Z. Y. Yin, Selective N₂/H₂O adsorption onto 2D amphiphilic amorphous photocatalysts for ambient gas-phase nitrogen fixation, Appl. Catal., B, 2021, 294, 120240.
18. C. Wu, W. X. Huang, H. M. Liu, K. L. Lv and Q. Li, Insight into synergistic effect of Ti₃C₂ MXene and MoS₂ on anti-photocorrosion and photocatalytic performance of CdS for hydrogen production, Appl. Catal., B, 2023, 330, 122653.
19. S. Singh, R. Punia, K. K. Pant and P. Biswas, Effect of work-function and morphology of heterostructure components on CO₂ reduction photocatalytic activity of MoS₂-Cu₂O heterostructure, Chem. Eng. J., 2022, 433, 132709.
20. X. W. Kong, K. L. Chang, X. M. Tao and L. Huang, Oxygen vacancy and Z-scheme heterojunction cooperate to promote visible light nitrogen photofixation, Chem. Eng. J., 2023, 475, 146159.
21. M. Y. Yu, J. B. Wang, G. J. Li, S. L. Zhang and Q. Zhong, Construction of 3D/2D indium vanadate/graphite carbon nitride with nitrogen defects Z-scheme heterojunction for improving photocatalytic carbon dioxide reduction, J. Mater. Sci. Technol., 2023, 154, 129–139.
22. G. Wang, Y. P. Wang, S. G. Li, Q. S. Yang, D. Y. Li, S. T. Pantelides and J. H. Lin, Engineering the Crack Structure and Fracture Behavior in Monolayer MoS₂ by Selective Creation of Point Defects, Adv. Sci., 2022, 9, 2200700.
23. Y. H. Wang, W. Y. Yu, C. Y. Wang, F. Chen, T. Y. Ma and H. W. Huang, Defects in photoreduction reactions: Fundamentals, classification, and catalytic energy conversion, eScience, 2024, 4, 100228.
24. L. Y. Zhang, Z. L. Huang, B. Xie and S. J. Xia, Experimental and Theoretical Research on Photocatalytic Nitrogen Reduction Using MoS₂ Nanosheets with Polysulfide Vacancies, Inorg. Chem., 2024, 63, 10871–10880.
25. X. F. Li, Q. K. Li, J. Cheng, L. L. Liu, Q. Yan, Y. C. Wu, X. H. Zhang, Z. Y. Wang, Q. Qiu and Y. Luo, Conversion of Dinitrogen to Ammonia by FeN₃-Embedded Graphene, J. Am. Chem. Soc., 2016, 138, 8706–8709.
26. K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos, I. V. Grigorieva and A. A. Firsov, Electric Field Effect in Atomically Thin Carbon Films, Science, 2004, 306, 666–669.
27. H. Fei, T. Guo, Y. Xin, R. Q. Liu, D. Z. Wang, F. Y. Liu and Z. Z. Wu, Sulfur Vacancy Engineering of MoS₂ via Phosphorus Incorporation for Improved Electrocatalytic N₂ Reduction to NH₃, Appl. Catal., B, 2022, 300, 120733.
28. Y. Zhou, F. Liu, S. Geng, M. Q. Yao, J. Ma and J. X. Cao, Tuning the content of S vacancies in MoS₂ by Cu doping for enhancing catalytic hydrogenation of CO₂ to methanol, Mol. Catal., 2023, 547, 113288.
29. X. Z. Zheng, H. J. Han, J. F. Liu, Y. Yang, L. L. Pan, S. J. Zhang, S. G. Meng and S. F. Chen, Sulfur Vacancy-Mediated Electron–Hole Separation at MoS₂/CdS Heterojunctions for Boosting Photocatalytic N₂ Reduction, ACS Appl. Energy Mater., 2022, 5, 4475–4485.
30. C. L. Sun, Z. Q. Shao, Y. Hu, Y. Y. Peng and Q. J. Xie, Photoelectrocatalysis Synthesis of Ammonia Based on a Ni-Doped MoS₂/Si Nanowires Photocathode and Porous Water with High N₂ Solubility, ACS Appl. Mater. Interfaces, 2023, 15, 23085–23092.
31. B. Delley, An all-electron numerical method for solving the local density functional for polyatomic molecules, J. Chem. Phys., 1990, 92, 508–517.
32. Y. M. Yang, S. Feng, J. N. Su, Y. Y. Gong and J. J. Wang, Mo-Doped Metal–Organic Frameworks for Efficient Nitrogen Reduction Reaction: A Density Functional Theory Study, ACS Sustainable Chem. Eng., 2022, 10, 14064–14072.
33. Y. Y. Qin, Y. Li, W. S. Zhao, S. H. Chen, T. T. Wu and Y. Q. Su, Computational study of transition metal single-atom catalysts supported on nitrogenated carbon nanotubes for electrocatalytic nitrogen reduction, Nano Res., 2023, 16, 325–333.
34. J. X. Zhao and Z. F. Chen, Single Mo Atom Supported on Defective Boron Nitride Monolayer as an Efficient Electrocatalyst for Nitrogen Fixation: A Computational Study, J. Am. Chem. Soc., 2017, 139, 12480–12487.
35. S. Han, X. M. Wei, Y. H. Huang, J. M. Zhang, J. Yang and Z. D. Wang, Tuning the activity and selectivity of nitrogen reduction reaction on double-atom catalysts by B doping: A density functional theory study, Nano Energy, 2022, 99, 107363.
36. T. Y. Wang, Z. Y. Guo, H. Oka, A. Kumatani, C. W. Liu and H. Li, Origin of electrocatalytic nitrogen reduction activity over transition metal disulfides: critical role of in situ generation of S vacancy, J. Mater. Chem. A, 2024, 12, 8438–8446.
37. R. Q. Liu, H. Fei, J. Wang, T. Guo, F. Y. Liu, Z. Z. Wu and D. Z. Wang, Unveiling the synergistic effect between the metallic phase and bridging S species over MoS₂ for highly efficient nitrogen fixation, Appl. Catal., B, 2024, 343, 123469.
38. H. Y. Xu, F. P. Zhang, L. R. Fang, Y. Q. Xu, Z. W. Yu, L. Ma, D. Q. Guan and Z. P. Shao, Deciphering the Nitrogen Activation Mechanisms on Group VIII Single Atoms at MoS₂, Inorg. Chem., 2024, 63, 19570–19581.
39. X. Y. Chen, Y. R. Wang and X. C. Meng, Fabrication of NiS_x/MoS₂ interface for accelerated charge transfer with greatly improved electrocatalytic activity in nitrogen reduction to produce ammonia, Chem. Eng. J., 2024, 479, 147701.
40. N. Zhang, Y. B. Si, Q. Fu and X. Chen, A bioinspired sulfur-surrounded iron catalyst for ammonia synthesis, J. Mater. Chem. A, 2024, 12, 11602–11611.
41. G. M. Zhan, F. J. Quan, Y. C. Yao, S. X. Zhao, X. F. Liu, H. Y. Gu, Y. Huang, X. Liu, F. L. Jia and L. Z. Zhang, Single-layer MoS₂ with adjacent Mo sites for efficient electrocatalytic nitrogen fixation via spin-delocalized electrons effect, Appl. Catal., B, 2023, 323, 122186.
42. X. H. Xu, H. R. Wang, K. Y. Chen, Q. Mei, J. J. Li and Q. Z. Wang, Novel ternary composite catalyst 2H/1T-MoS₂/Co₃O₄-Ru for photoelectrocatalytic nitrogen reduction, Chem. Eng. J., 2024, 485, 149922.
43. P. Basumatary, D. Konwar and Y. S. Yoon, Conductivity-tailored PtNi/MoS₂ 3D nanoflower catalyst via Sc doping as a hybrid anode for a variety of hydrocarbon fuels in proton exchange membrane fuel cells, Appl. Catal., B, 2020, 267, 118724.
44. L. Cai, J. Yang, X. M. Shao, X. Y. Zhu, Y. Liu, C. X. Wan, F. Wang, Q. Liu and J. Chen, Copper-doped nanostructured MoS₂ with distinct nanoscale morphology for efficient bacteria inactivation and catalytic degradation of antibiotics, J. Environ. Chem. Eng., 2024, 12, 111537.
45. M. D. Sharma, C. Mahala, B. Modak, S. Pande and M. Basu, Doping of MoS₂ by “Cu” and “V”: An Efficient Strategy for the Enhancement of Hydrogen Evolution Activity, Langmuir, 2021, 37, 4847–4858.
46. D. Zabelin, A. Tulupova, A. Zabelina, A. Tosovska, R. Valiev, R. Ramazanov, D. Mares, V. Jerabek, V. Burtsev, M. Erzina, A. Michalcová, A. Skvortsova, V. Svorcik and O. Lyutakov, Enhanced photoelectrochemical nitrogen reduction to ammonia by a plasmon-active Au grating decorated with the gC₃N₄@MoS₂ heterosystem and plasmon-active nanoparticles, J. Mater. Chem. A, 2024, 12, 21310–21320.
47. R. Q. Liu, H. Fei, J. Wang, T. Guo, F. Y. Liu, J. Y. Wang, Z. Z. Wu and D. Z. Wang, Insights of active sites separation mechanism for highly efficient electrocatalytic N₂ reduction to ammonia over glucose-induced metallic MoS₂, Appl. Catal., B, 2023, 337, 122997.
48. T. Ye, K. Ba, X. Y. Yang, T. S. Xiao, Y. Y. Sun, H. Q. Liu, C. Tang, B. H. Ge, P. Zhang, T. Duan and Z. Z. Sun, Valence engineering at the interface of MoS₂/Mo₂C heterostructure for bionic nitrogen reduction, Chem. Eng. J., 2023, 452, 139515.
49. T. X. Ma, R. Q. Li, Y. C. Huang, Y. X. Lu, L. Guo, M. M. Niu, X. Huang, R. A. Soomro, J. Y. Ren, Q. Wang, B. Xu, C. M. Yang, F. Fu and D. J. Wang, Interfacial Chemical-Bonded MoS₂/In–Bi₂MoO₆ Heterostructure for Enhanced Photocatalytic Nitrogen-to-Ammonia Conversion, ACS Catal., 2024, 14, 6292–6304.
50. N. Kumar, R. P. Mishra, B. Dash, S. Bastia and Y. S. Chaudhary, The synergistic chemical coupling of nanostructured MoS₂ with nitrogen-deficient 2-D triazine-based polymeric m-C₃N_x for efficient and selective CO₂ photocatalytic conversion to CO, J. Mater. Chem. A, 2023, 11, 20839–20853.
51. X. Y. Wang, S. F. Zhao, W. Shi, J. L. Sun, Z. M. Ni and S. J. Xia, Photocatalytic degradation of norfloxacin by WO₃@LDHs Z-scheme heterojunction: performance, kinetic and mechanistic studies, Chem. Phys. Lett., 2025, 865, 141904.
52. T. J. Aubry, J. M. Clary, E. M. Miller, D. Vigil-Fowler and J. van de Lagemaat, Activating Nitrogen for Electrochemical Ammonia Synthesis via an Electrified Transition-Metal Dichalcogenide Catalyst, J. Phys. Chem. C, 2024, 128, 7063–7072.
53. P. W. Zhao, H. Z. Wang, Z. Huang, B. B. Fan, Y. Liu, Z. X. Yin, Y. Wang, X. R. Zheng, Y. D. Deng and X. B. Fan, High-Throughput Screening of Nitrogen Reduction Reaction on Single Atom@1T'-MoS₂, Appl. Surf. Sci., 2023, 631, 157480.
54. S. J. Xia, Z. Y. Yuan, Y. Meng, C. Zhang, X. L. Li, Z. M. Ni and X. Q. Zhang, Fabrication of site activated and synergistic dual vacancies ZnIn₂S₄ for highly efficient bifunctional photocatalysis: nitrogen reduction and oxidative degradation, J. Mater. Chem. A, 2024, 12, 2294–2308.
55. A. Jain, M. B. Sadan and A. Ramasubramaniam, Identifying a New Pathway for Nitrogen Reduction Reaction on Fe-Doped MoS₂ by the Coadsorption of Hydrogen and N₂, J. Phys. Chem. C, 2021, 125, 19980–19990.
56. X. Li, G. T. Hai, G. F. Chen, J. Liu, F. L. Zhao, J. S. Zhou, M. K. H. Leung and H. H. Wang, Unique two-electron transfer pathway of Bismuth nanocrystal for enhanced N₂ electroreduction revealed by in situ infrared spectroscopy, Appl. Catal., B, 2024, 342, 123365.
57. Y. Meng, Z. L. Huang, D. Liu, B. Xie and S. J. Xia, Enriched oxygen vacancies Bi₂WO₆-Ov@LDHs photoactivated persulfate degradation of tetracycline: synergistic effect of Z-scheme heterojunction and variable valence state, J. Environ. Chem. Eng., 2025, 13, 115942.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5ta01298h

---