Spin Manipulation of Co sites in Co₉S₈/Nb₂CT_x Mott–Schottky Heterojunction for Boosting the Electrocatalytic Nitrogen Reduction Reaction

Abstract

Regulating the adsorption of an intermediate on an electrocatalyst by manipulating the electron spin state of the transition metal is of great significance for promoting the activation of inert nitrogen molecules (N₂) during the electrocatalytic nitrogen reduction reaction (eNRR). However, achieving this remains challenging. Herein, a novel 2D/2D Mott–Schottky heterojunction, Co₉S₈/Nb₂CT_x‐P, is developed as an eNRR catalyst. This is achieved through the in situ growth of cobalt sulfide (Co₉S₈) nanosheets over a Nb₂CT_x MXene using a solution plasma modification method. Transformation of the Co spin state from low (t_2g ⁶e_g ¹) to high (t_2g ⁵e_g ²) is achieved by adjusting the interface electronic structure and sulfur vacancy of Co₉S₈/Nb₂CT_x‐P. The adsorption ability of N₂ is optimized through high spin Co(II) with more unpaired electrons, significantly accelerating the *N₂→*NNH kinetic process. The Co₉S₈/Nb₂CT_x‐P exhibits a high NH₃ yield of 62.62 µg h⁻¹ mg_cat. ⁻¹ and a Faradaic efficiency (FE) of 30.33% at −0.40 V versus the reversible hydrogen electrode (RHE) in 0.1 m HCl. Additionally, it achieves an NH₃ yield of 41.47 µg h⁻¹ mg_cat. ⁻¹ and FE of 23.19% at −0.60 V versus RHE in 0.1 m Na₂SO₄. This work demonstrates a promising strategy for constructing heterojunction electrocatalysts for efficient eNRR.

A novel 2D/2D Mott–Schottky heterojunction, Co₉S₈/Nb₂CT_x‐P is preparaed by a solution plasma modification method. The transformation of the Co 3d orbital electron configuration frow low (t_2g ⁶e_g ¹) to high (t_2g ⁵e_g ²) spin is induced by S_v and interface effects. The highly spin‐polarized state of the Co₉S₈/Nb₂CT_x‐P demonstrated superior eNRR performance.

1. Introduction

As an excellent alternative to fossil fuels, ammonia (NH₃) is regarded as a clean energy, offering advantages including low cost, easy availability, volatility, convenient storage, low pollution, high burning value, a high‐octane number, relatively safe operation, and compatibility with common materials.^{[ 1 ]} Despite the high N₂ content in air (78%), converting N₂ from air to NH₃ and related compounds is challenging, owing to the inertness and high dissociation energy of molecular N₂.^{[ 2 ]} Commercial NH₃ is synthesized using the Haber–Bosch process, which requires harsh conditions, thereby leading to substantial energy consumption and significant CO₂ emissions.^{[ 3 ]} In contrast to the traditional high‐energy Haber‐Bosch process, the electrocatalytic nitrogen reduction reaction (eNRR) offers an emerging alternative strategy for NH₃ generation under environmental conditions, using renewable energy sources.^{[ 4 ]} However, the eNRR is typically conducted in acidic media. This is because the presence of abundant hydrogen protons (H⁺) promotes the formation of the *NNH intermediate, which is the rate‐determining step (RDS) in NH₃ synthesis.^{[ 5 ]} Given the growing environmental concerns, there is a high demand for the conversion of N₂ into NH₃ in neutral media. Therefore, the development of advanced NRR electrocatalysts for the production of NH₃ in pH‐universal media is highly desirable.

Transition metal atoms comprise incompletely filled d orbitals that can effectively activate N₂. Consequently, various transition metal materials have been developed as effective catalysts for the eNRR.^{[ 6 ]} Nevertheless, the high electronegativity between the transition metal sites and neighboring N atoms renders the free energy of the intermediate (e.g., *NNH) unsuitable for adsorption.^{[ 7 ]} Electrocatalysts in a highly spin‐polarized state can promote N₂ adsorption in the eNRR and regulate the bonding strength of the adsorption/desorption intermediates on the electrocatalyst surface, affording tunable surface electrochemical reaction kinetics.^{[ 8 ]} Transition metal spin states are closely related to the occupancy of the e_g and t_2g orbitals. Therefore, adjusting the electron configuration of these two orbitals can accelerate the eNRR and improve NH₃ production efficiency. Various strategies for heteroatom doping,^{[ 7} , ^{9 ]} element (e.g., S, O, N, or metal) vacancy engineering,^{[ 10 ]} and interface engineering^{[ 11 ]} have been exploited to regulate the electron configuration of the e_g and t_2g orbitals of active metal sites. For example, oxygen vacancies have been introduced into metal oxides (e.g., oxygen‐rich MoO₂ ^{[ 12 ]} and TiO₂ ^{[ 13 ]}) to regulate the electronic state and adsorption properties of the active sites. This strategy weakens the N─N bond, thereby facilitating NH₃ production.^{[ 14 ]} In addition, NH₃ production using heterojunction electrocatalysts can be enhanced through the accumulation of reactants, formation of specific intermediates,^{[ 15 ]} and inhibition of byproducts.^{[ 16 ]} Current investigations on the spin states of active metal sites in heterojunctions mainly focus on revealing the origin of intrinsic activities.^{[ 16} , ^{17 ]} However, research exploring the working mechanism of the electron spin states in such electrocatalysts and their impact on effective spin‐related electron transfer during the eNRR remains scarce. In particular, regulating the bonding strength of intermediates on electrocatalysts by manipulating the electron spin states of metal atomic active sites remains a significant challenge for promoting the activation of inert N₂ during the eNRR.

In this study, a novel S‐vacancy (S_v)‐enriched Co₉S₈ and Nb₂CT_x MXene (T_x = surface terminal group) Mott–Schottky (M–S) heterojunction, Co₉S₈/Nb₂CT_x‐P, was prepared in situ using a solution plasma (SP) modification method for the efficient NH₃ synthesis via the eNRR in acid and neutral electrolyte. During the SP irradiation procedure, a large number of free radicals and high‐energy particles can be generated, which further attack the electrocatalyst surface.^{[ 18 ]} Rich lattice defects and element vacancies thus formed on the electrocatalyst surface thanks to the dissociation of chemical bonds of active component, which enhanced the electron delocalization and spin state transition of the Co active sites. The highly spin‐polarized state stimulated the production of more unpaired electrons in the d orbitals of the Co sites, which effectively promoted the adsorption and activation of N₂. Moreover, owing to the M–S effect, the strong electron coupling between the resulting heterojunctions further reduced the energy barrier for N₂ protonation. Density functional theory (DFT) calculations revealed that the upward shift of the $d_{x^2}-d_{y^2}$ band center of the Co sites in Co₉S₈/Nb₂CT_x‐P energetically favored N₂ adsorption and N─H bond formation in *NNH. The constructed S_v‐enriched Co₉S₈/Nb₂CT_x‐P M–S heterojunction exhibited superior eNRR performances, yielding a high NH₃ yield of 62.62 µg h⁻¹ mg_cat. ⁻¹ at −0.40 V versus the reversible hydrogen electrode (RHE) in 0.1 m HCl, with a Faradaic efficiency (FE) of 30.33%. Additionally, it achieved an NH₃ yield of 41.47 µg h⁻¹ mg_cat. ⁻¹ at −0.60 V versus RHE in 0.1 m Na₂SO₄, with an FE of 23.19%. This work systematically confirms the application of spin engineering to eNRR activity, paving the way for the design of a new type of heterojunction electrocatalyst for the eNRR.

2. Results and Discussion

Figure  1a illustrates the formation process of the Co₉S₈/Nb₂CT_x‐P M–S heterojunction. First the Co₉S₈/Nb₂CT_x M–S heterojunction was constructed through the in situ growth of Co₉S₈ nanosheets (NSs) around Nb₂CT_x MXene nanoflakes using a hydrothermal method. Subsequently, the obtained heterojunction was modified by SP treatment to form the Co₉S₈/Nb₂CT_x‐P M–S heterojunction comprising abundant S_vs. Figure S1a,b (Supporting Information) clearly depicts that in the Co₉S₈/Nb₂CT_x, small‐sized NSs were vertically grown on the multilayered nanoflakes. Further, the edge of Co₉S₈/Nb₂CT_x was identified as Co₉S₈NSs, which is marked with red dashed lines in the TEM image (Figure S1b, Supporting Information). The high‐resolution transmission electron microscopy (HR‐TEM) image of the Co₉S₈/Nb₂CT_x (Figure S1c, Supporting Information) shows the connected structure of Co₉S₈ and Nb₂CT_x, along with a clear lattice spacing of 0.248 nm attributed to the (400) plane of Co₉S₈.

Figure 1.

a) Schematic illustration of the preparation procedure of the Co₉S₈/Nb₂CT_x‐P M–S heterojunction. b) SEM, c) TEM, d) HR‐TEM images of the Co₉S₈/Nb₂CT_x‐P. e) HR‐TEM images and corresponding lattice line scanning of Co₉S₈/Nb₂CT_x, f) HR‐TEM images and corresponding lattice line scanning of Co₉S₈/Nb₂CT_x‐P. g) The EDS mapping of Co₉S₈/Nb₂CT_x‐P (blue: Co, orange: Nb, and red: S).

Similar results were observed for the heterojunction following SP treatment (Figure 1b–d). However, a few S_vs were observed in the HR‐TEM image of the Co₉S₈/Nb₂CT_x‐P (Figure 1f, encircled in red), which were absent in that of the pristine heterojunction (Figure 1e). Additionally, the energy‐dispersive X‐ray spectroscopy mapping images of the Co₉S₈/Nb₂CT_x and Co₉S₈/Nb₂CT_x‐P (Figure S1d, Supporting Information; Figure 1g, respectively) revealed the co‐existence and homogeneous distribution of Nb, Co, S, C, and O throughout the selected region, indicating the full coverage of Nb₂CT_x with Co₉S₈ NSs. For comparison, the individual components, Co₉S₈ and Nb₂CT_x, before and after SP modification were also prepared, and their surface morphologies and nanostructures were analyzed (Figures S2 and S3, respectively). The Co₉S₈/Nb₂CT_x demonstrated a larger Brunauer–Emmett–Teller surface area of 80.1 m² g⁻¹ compared to those of Co₉S₈ (51.9 m² g⁻¹) and Nb₂CT_x (24.7 m² g⁻¹) (Figure S4a and Table S1, Supporting Information). This superior result is conducive to the penetration of the electrolyte and sufficient contact between the reactants and catalyst.

Figure S5 (Supporting Information) depicts the X‐ray diffraction (XRD) patterns of the Co₉S₈/Nb₂CT_x M–S heterojunction before and after SP modification, revealing low crystallinity indicated by weak signals. Specifically, three weak diffraction peaks at 2θ = 29.7°, 47.8°, and 52.0° were observed in the XRD pattern of the heterojunction, corresponding to the (311), (511), and (440) facets of Co₉S₈, respectively. Furthermore, the diffraction peak at 2θ = 7.7°, which was attributed to the (002) plane of Nb₂CT_x, was relatively weak due to coverage by the Co₉S₈ NSs.

The Fourier‐transform infrared (FT‐IR) spectra of the heterojunction before and after the SP treatment were similar (Figure S6, Supporting Information). The characteristic band at 613 cm⁻¹ was assigned to the stretching vibration of Co‐S bond in Co₉S₈. The peaks located at 483, 1107, 1363, and 1734 cm⁻¹ were assigned to the Nb─C stretching, C─F, C─H, and C═O, respectively.

The X‐ray photoelectron spectroscopy (XPS) survey scan spectra of the Co₉S₈/Nb₂CT_x and Co₉S₈/Nb₂CT_x‐P (Figure S7, Supporting Information) displayed clear Nb 3d (207.3 eV), Co 2p (780.7 eV), S 2p (162.7 eV), C 1s (284.1 eV), and O 1s (531.6 eV) signals, further verifying the combination of the elemental signals of Co₉S₈ and Nb₂CT_x. The Co 2p XPS spectrum of Co₉S₈/Nb₂CT_x‐P (Figure  2a) was deconvoluted into two pairs of peaks corresponding to Co 2p _3/2 (781.1 eV) and Co 2p _1/2 (797.1 eV). The pairs comprised peaks attributed to Co⁰ (779.2 and 796.2 eV), Co³⁺ (781.1 and 797.3 eV), and Co²⁺ (782.9 and 798.7 eV), indicating Co⁰/Co²⁺/Co³⁺mixed valence states. Notably, the binding energy (BE) positions of the Co 2p _3/2 (781.3 eV) and Co 2p _1/2 (797.1 eV) peaks of Co₉S₈/Nb₂CT_x‐P shifted negatively, by 0.2 and 0.4 eV, relative to those of the Co₉S₈/Nb₂CT_x (781.1 and 797.5 eV), respectively. The peak positions of Co 2p _3/2 (781.3 eV) and Co 2p _1/2 (797.1 eV) in the Co₉S₈/Nb₂CT_x spectrum (curve ii, Figure 2a) also exhibited a slight negative shift of 0.22 eV, compared to those of pristine Co₉S₈ (curve i, Figure 2a). Specifically, the electrons of the Co atom in Co₉S₈/Nb₂CT_x can be transferred to the surrounding coordination environment.^{[ 19 ]} Moreover, the Co²⁺/Co³⁺ intensity ratio of the Co₉S₈/Nb₂CT_x‐P M–S heterojunction (0.69) was higher than that of pristine Co₉S₈/Nb₂CT_x (0.54), and both were lower than that of Co₉S₈ (0.81) (Figure 2b).

Figure 2.

a) The Co 2p XPS spectra, and b) corresponding the different Co valence intensity ratio, c) S 2p XPS spectra of i) Co₉S₈, ii) Co₉S₈/Nb₂CT_x and iii) Co₉S₈/Nb₂CT_x‐P. d) Co K‐edge XANES and e) FT‐EXAFS spectra of Co foil, CoO, Co₉S₈, Co₉S₈/Nb₂CT_x, and Co₉S₈/Nb₂CT_x‐P. f) FT‐EXAFS fitting results of Co₉S₈, Co₉S₈/Nb₂CT_x, and Co₉S₈/Nb₂CT_x‐P. WT‐EXAFS contour map of g) Co₉S₈, h) Co₉S₈/Nb₂CT_x, and I) Co₉S₈/Nb₂CT_x‐P.

During plasma irradiation in water, a high flux of hydrogen radicals, reactive oxygen species, electrons, and ions is produced via water splitting at the open end of the discharge zone.^{[ 20 ]} These active sites can then attack the Co─S bond in Co₉S₈, leading to its dissociation. Consequently, Co₉S₈/Nb₂CT_x‐P remains enriched with S_vs with very low energy at the electrocatalyst surface, without altering the bulk properties. The pre‐occupied S 2p orbital electrons become delocalized around the Co³⁺ ions neighboring the S vacancies. Subsequently, the unpaired electrons resulting from the removal of S atoms can be accommodated in the orbitals of Co³⁺ ions, leading to the partial reduction of Co³⁺ to Co²⁺.^{[ 21 ]} Additionally, the S 2p XPS spectrum of Co₉S₈/Nb₂CT_x‐P (curve iii, Figure 2c) included peaks corresponding to Co─S (163.9 eV), C═S (162.7 eV), and S─O (168.9 eV). The peak positions of S 2p _3/2 (160.7 eV) and S 2p _1/2 (161.9 eV) in the Co₉S₈/Nb₂CT_x spectrum (curve ii, Figure 2c) were negatively shifted by 0.7 and 0.8 eV compared to those of Co₉S₈ (161.7 and 162.8 eV) (curve i, Figure 2c), respectively. The M–S junction interface generated between Co₉S₈ and Nb₂CT_x can modulate the density of Co active sites, resulting in electron transfer from Nb₂CT_x to Co₉S₈.^{[ 22 ]} Moreover, the BE positions of S 2p _3/2 and S 2p _1/2 in the Co₉S₈/Nb₂CT_x‐P spectrum exhibited negative shifts of 0.2 and 0.12 eV, respectively, relative to those of the Co₉S₈/Nb₂CT_x spectrum. This was attributed to the generation of abundant S_vs in Co₉S₈/Nb₂CT_x‐P, which reduces the electron density on the S atom. In addition, the atomic composition, determined through XPS analysis, revealed a smaller S/Co atomic ratio of Co₉S₈/Nb₂CT_x‐P (0.53) compared to those of Co₉S₈ (0.69) and Co₉S₈/Nb₂CT_x (0.62; Figures S8–S10, Supporting Information), indicating the formation of more S_vs.

To further probe the local coordination geometry at the atomic level of Co₉S₈/Nb₂CT_x before and after SP treatment, X‐ray absorption near‐edge structure (XANES) and extended X‐ray absorption fine structure (EXAFS) analyses were conducted on Co₉S₈, Co foil, and CoO for comparison. The Co₉S₈, Co₉S₈/Nb₂CT_x, and Co₉S₈/Nb₂CT_x‐P XANES spectra (Figure 2d) displayed similar peak shapes and adsorption edges, indicating their close valence states. Moreover, the energy absorption threshold of Co₉S₈/Nb₂CT_x‐P fell between those of the standard Co foil and CoO. These results indicated that the valence state of the Co atom in Co₉S₈/Nb₂CT_x‐P is higher than that of Co⁰ but lower than that of Co²⁺. The FT k²‐weighted phase‐uncorrected Co K‐edge EXAFS spectrum of Co₉S₈/Nb₂CT_x‐P (Figure 2e) displayed its main peak at 1.75 Å, attributed to the single scattering path of Co─S. Furthermore, the pristine heterojunction demonstrated a similar single scattering path of Co─S, which, however, was slightly shorter than that of Co₉S₈ (1.79 Å). Notably, the peak intensity of the Co─S bond in Co₉S₈/Nb₂CT_x‐P was lower than those in Co₉S₈/Nb₂CT_x and Co₉S₈. This weaker Co─S peak intensity of Co₉S₈/Nb₂CT_x‐P was primarily attributed to the presence of S_vs.

The quantitative least‐squares EXAFS curves were fitted to analyze the local coordination of Co in the three catalysts (Figures 2f and S11 and Table S2, Supporting Information). Co₉S₈ presented two different sets, namely Co─S1 with bond length of 2.08 Å and Co─S2 with bond length of 2.21 Å, suggesting the coexistence of two different unit cells. The coordination numbers for Co─S1 and Co─S2 were fitted to be 1.53 and 3.19, respectively. For the Co₉S₈/Nb₂CT_x‐P heterojunction, the coordination number for Co─S1 was determined to be 1.26, which is lower than those of Co₉S₈ (1.53) and Co₉S₈/Nb₂CT_x (1.47) owing to the abundant formation of S_vs. The decreased length of the Co─S bond and the low coordination number of Co₉S₈/Nb₂CT_x‐P demonstrate negligible structural deformation attributed to the high population of S_vs. Moreover, the wavelet transform analysis of the three samples (Figure 2g–I) revealed that the Co─S path exhibited maximum intensity at R ≈ 2.15 Å and k ≈ 6.7 Å⁻¹.

The normalized secondary electron cutoff energies of the samples were measured using ultraviolet photoelectron spectroscopy (UPS) (Figure  3a), enabling the deduction of the work function (Φ) of the electrocatalyst. As illustrated in Figure 3b, the Φ of Co₉S₈/Nb₂CT_x is ≈4.26 eV, which is marginally smaller than those of Co₉S₈ (4.58 eV) and Nb₂CT_x (4.49 eV). The semiconductor type of Co₉S₈ and Co₉S₈/Nb₂CT_x was probed using M–S curves (Figure S12, Supporting Information), revealing characteristics consistent with n‐type semiconductors. Ultraviolet‐visible (UV–vis) absorption spectra were recorded for all the prepared electrocatalysts (Figure S12c, Supporting Information). Additionally, the forbidden bands of Nb₂CT_x, Co₉S₈, and Co₉S₈/Nb₂CT_x were calculated as 0.84, 0.87, and 0.78 eV, respectively, from Tauc plots (Figure S12d–f, Supporting Information).

Figure 3.

a) UPS spectra in the normalized secondary electron cutoff energy (E _cutoff) regions of i) Co₉S₈, ii) Nb₂CT_x, and iii) Co₉S₈/Nb₂CT_x. b) Calculated Φ stemming from UPS spectra. c) Energy band structures of Co₉S₈, Nb₂CT_x, and Co₉S₈/Nb₂CT_x (E_F: Fermi level, E_VB: valence band level, and E_vac: vacuum level). d) Energy band diagrams of Nb₂CT_x and Co₉S₈ before and after the Mott‐Schottky junction formation. e) Schematic illustration of the interfacial electronic couple effect between Nb₂CT_x and Co₉S₈. f) EPR spectra i) Co₉S₈, ii) Co₉S₈/Nb₂CT_x, iii) Co₉S₈/Nb₂CT_x‐P. The electron localization function of g) Co₉S₈/Nb₂CT_x and h) Co₉S₈/Nb₂CT_x‐P. I) Magnetic susceptibilities and j) The number of unpaired electrons of i) Co₉S₈, ii) Co₉S₈/Nb₂CT_x, iii) Co₉S₈/Nb₂CT_x‐P. k) Illustration of Co²⁺ spin states in Co₉S₈/Nb₂CT_x‐P.

The combination of M–S plots, UPS, and UV–vis results confirmed the formation of an M–S heterojunction between Nb₂CT_x and Co₉S₈. Furthermore, the energy band plots and band structures of Nb₂CT_x and Co₉S₈ (Figure 3c,d) revealed the redistribution of electrons at the interface between Nb₂CT_x and Co₉S₈ owing to the Schottky barrier. When Nb₂CT_x and Co₉S₈ come into contact and form a heterojunction interface, electrons from Nb₂CT_x spontaneously flow to Co₉S₈ until dynamic equilibrium is reached in terms of the Fermi levels. At this moment, the holes in the valence band of Co₉S₈ flow to Nb₂CT_x, facilitating charge transfer at the interface. This induces band bending and generates an internal electric field at the M–S interfaces, thereby promoting the flow of electrons from Nb₂CT_x into Co₉S₈ (Figure 3e). In the presence of the Fermi level and interfacial built‐in electric field, interfacial electron transfer accelerates the catalytic process of the eNRR. This acceleration is attributed to the tendency of the S_v‐enriched Co₉S₈/Nb₂CT_x‐P to donate more electrons to the intermediates in the eNRR process.

Theoretically, the eNRR performance of transition metal‐based electrocatalysts is related to the occupancy rates of their d orbitals.^{[ 23 ]} Therefore, to investigate the electronic configurations of the Co center, we conducted electron paramagnetic resonance (EPR), zero‐field cooling (ZFC) temperature‐dependent magnetization (χ_m). Figure 3f illustrates the signals of the Co─S dangling bonds at g = 2.003 in the EPR spectra of Co₉S₈/Nb₂CT_x before and after SP modification, which are proportional to the concentrations of dangling bonds from the S_vs in Co₉S₈/Nb₂CT_x and Co₉S₈/Nb₂CT_x‐P. The intensity of the g signal of Co₉S₈/Nb₂CT_x‐P was higher than that of the pristine heterojunction, indicating an increase in the number of defects. Specifically, the S_vs in Co₉S₈/Nb₂CT_x‐P result from a region of electron delocalization at the S position, further causing a slight localization of electrons in the surrounding S atom (Figure 3g,h). Consequently, the electronic structure of the surrounding Co active site can be modulated, facilitating electron transfer from the catalyst to N₂ and the intermediates generated in the eNRR process. This enhances the NRR performance of the constructed electrocatalyst.

Given that charge redistribution is accompanied by the transition of the 3d electron spin configuration, temperature‐dependent magnetization in the ZFC mode was conducted at 1000 Oe in the temperature range of 5–300 K to gain deeper insight into the effect of the magnetic moments of Co₉S₈, Co₉S₈/Nb₂CT_x, and Co₉S₈/Nb₂CT_x‐P. From the χ_m versus T plots (Figure 3I), the effective magnetic moment (µ_eff) of Co₉S₈/Nb₂CT_x‐P was estimated to be 4.34 µ_B, which is markedly higher than those of Co₉S₈ (2.03 µ_B) and Co₉S₈/Nb₂CT_x (2.01 µ_B). Therefore, with an increase in the number of S_vs, electrons can fill the high‐energy orbital (e _g), thereby improving the spin state of Co. According to the equation 2.828$\sqrt{{\chi }_m\mathrm{T}}$ = µ_eff = $\sqrt{\mathrm{n}(\mathrm{n}+2)},$ ^{[ 24 ]} the number of unpaired d electrons (n) in Co₉S₈/Nb₂CT_x‐P (Figure 3j) was calculated to be ≈3.45, which is significantly larger than those in Co₉S₈/Nb₂CT_x (≈1.24) and Co₉S₈ (≈1.26). The result indicated that Co₉S₈/Nb₂CT_x‐P had three unpaired electrons, leading to a high spin state (t_2g ⁵e_g ²). However, Co₉S₈/Nb₂CT_x and Co₉S₈ only had one unpaired electron, forming a low spin state (t_2g ⁶e_g ¹). As a result, it determines the electronic structure of Co 3d in Co₉S₈/Nb₂CT_x‐P (Figure 3k).

Owing to quantum spin exchange interactions, a greater number of unpaired electrons occupying the active orbital can enhance the catalytic activity of the electrocatalyst by weakening the binding of the reaction intermediates.^{[ 25 ]} Accordingly, the strong interaction between the S_vs and the formation of the heterojunction effectively reshaped the electronic structure of Co, resulting in a Co 3d electron spin configuration transition from low to high. Thus, Co₉S₈/Nb₂CT_x‐P exhibited more unpaired electrons in the 3d orbitals, which enhanced its catalytic activity compared to those of Co₉S₈/Nb₂CT_x and Co₉S₈ (Figure  4a).

Figure 4.

a) The schematic illustration of the spin regulation of Co 3d for the acceleration of the protonation process. b) LSV polarization curves of Co₉S₈/Nb₂CT_x‐P in an N₂‐ and Ar‐saturated 0.1 m HCl electrolyte. Dependence of the NH₃ yield and FE of Co₉S₈/Nb₂CT_x‐P at each applied potential in N₂‐saturated c) 0.1 m HCl and d) 0.1 m Na₂SO₄ with the NRR measurement time of 6000 s. e) The eNRR performances of different electrocatalysts of Co₉S₈, Nb₂CT_x, and Co₉S₈/Nb₂CT_x before and after the SP treatment at −0.40 V versus RHE in 0.1 m HCl. f) Comparison of the NH₃ yield rate of Co₉S₈/Nb₂CT_x‐P with some reported electrocatalysts.

The primary product of eNRR is NH₃, which can be captured by electrolyte solution. Additionally, N₂H₂ as an unexpected by‐product also affects the performance of eNRR. Therefore, the production of NH₃ and N₂H₄ in the cathode compartment in each of the two solutions was measured after each electrolysis procedure, according to the calibration curves standardized by the indophenol blue and Watt–Chrisp method methods (Figures S13–S16, Supporting Information). The eNRR performance of the Co₉S₈/Nb₂CT_x‐P M–S heterojunction was investigated in both 0.1 m HCl and 0.1 m Na₂SO₄, using Co₉S₈ and pristine Co₉S₈/Nb₂CT_x for comparison. Amongst, the influence of the Co₉S₈ content in Co₉S₈/Nb₂CT_x‐P on the NRR performances was systematically probed (Figure S17, Supporting Information). Linear sweep voltammetry polarization curves of Co₉S₈/Nb₂CT_x‐P were recorded separately in Ar‐ and N₂‐saturated 0.1 m HCl (Figure 4b) and 0.1 m Na₂SO₄ (Figure S18, Supporting Information) at a scan rate of 5 mV s⁻¹. A significant current density gap of Co₉S₈/Nb₂CT_x‐P was observed within the potential window of −0.23 to −0.70 V and −0.68 to −1.10 V versus RHE in 0.1 m HCl and 0.1 m Na₂SO₄, respectively. The onset potential of Co₉S₈/Nb₂CT_x‐P in 0.1 m HCl electrolyte is −0.40 V, while it in 0.1 m Na₂SO₄ is −0.68 V, in which the onset potential is determined at the current density of −0.5 mA cm⁻². As results, the current gap of Co₉S₈/Nb₂CT_x‐P between N₂ and Ar saturated electrolyte indicates that enhanced eNRR may occur at the nitrogen/electrocatalyst/electrolyte three‐phase interface. Additionally, cyclic voltammetry curves and Tafel plots of Co₉S₈/Nb₂CT_x‐P were investigated in Ar‐ and N₂‐saturated 0.1 m HCl or Na₂SO₄ electrolyte (Figure S19, Supporting Information). Subsequently, chronoamperometry tests were conducted for 6000 s in 0.1 m HCl and 0.1 m Na₂SO₄ to evaluate the NH₃ yield and FE of Co₉S₈/Nb₂CT_x‐P (Figure S20, Supporting Information). The NH₃ yield was calculated from the UV–vis absorption spectra of the electrolyte (Figure S21, Supporting Information). Notably, no significant N₂H₄ byproduct was detected during the eNRR (Figure S22, Supporting Information). These results indicate that the Co₉S₈/Nb₂CT_x‐P M–S heterojunction exhibits good selectivity for NH₃ production in both acidic and neutral media. As depicted in Figure 4c, Co₉S₈/Nb₂CT_x‐P exhibited a maximum NH₃ yield rate of 62.62 µg h⁻¹ mg_cat. ⁻¹ and a high FE of 30.33% at −0.40 V versus RHE in 0.1 m HCl. However, when a more negative potential was applied, both the NH₃ yield and FE of Co₉S₈/Nb₂CT_x‐P decreased markedly, owing to the competing hydrogen evolution reaction (HER). In addition, the electrocatalyst displayed extraordinary eNRR activity in 0.1 m Na₂SO₄ (Figure 4d), achieving a maximum NH₃ yield of 41.47 µg h⁻¹ mg_cat. ⁻¹ at −0.60 V versus RHE and an FE of 23.19%.

As illustrated in Figure 4e, the NH₃ yield and FE of the Co₉S₈/Nb₂CT_x‐P heterojunction are also higher than those of Co₉S₈ (NH₃ yield, 7.85 µg h⁻¹ mg_cat. ⁻¹; FE, 16.62%) and Co₉S₈/Nb₂CT_x (NH₃ yield, 16.89 µg h⁻¹ mg_cat. ⁻¹; FE, 13.83%) at −0.40 V versus RHE in 0.1 HCl. This performance also surpasses those of other Co‐ or MXene‐based electrocatalysts (Table S3, Supporting Information; Figure 4f), such as Co‐SAs/NC,^{[ 26 ]} CoP₃/CC,^{[ 27 ]} 1T‐MoS₂/g‐C₃N₄,^{[ 28 ]} MnO₂‐Ti₃C₂T_x,^{[ 29 ]} and C@CoS@TiO₂.^{[ 30 ]} Similarly, the NH₃ production rate and FE of Co₉S₈/Nb₂CT_x‐P in a neutral medium exceed those of most state‐of‐art electrocatalysts under the same conditions (Figure 4f; Table S4, Supporting Information), such as ZnO‐CoS QD,^{[ 31 ]} FeCoOOH HNCs,^{[ 32 ]} O‐CoP/CNT@G,^{[ 33 ]} and CoS@S‐Mas.^{[ 34 ]} This highlights the potential application of the constructed Co₉S₈/Nb₂CT_x‐P electrocatalyst for NH₃ production under ambient conditions. The superior eNRR ability of Co₉S₈/Nb₂CT_x‐P is ascribed to the strong electronic coupling of Co₉S₈ with Nb₂CT_x and the abundant S_vs, which significantly adjust the electronic structure of the Co active sites. The presence of more unpaired electrons in the 3d orbital of the Co active sites enhances the activation ability toward N₂ molecules, thereby boosting the eNRR performance.

A control experiment was conducted under the same conditions using bare carbon paper (CP) at an open‐circuit voltage in N₂‐saturated 0.1 m HCl (Figure S23a, Supporting Information) and 0.1 m Na₂SO₄ (Figure S23b Supporting Information). However, no significant NH₃ production was observed with bare CP, indicating that NH₃ is generated entirely from the eNRR catalyzed by Co₉S₈/Nb₂CT_x‐P. The origin of the N source was further determined using isotopic labeling measurements. As illustrated in the NMR spectra in Figures 5a and S24–S27 (Supporting Information), a distinct triplet (or doublet) coupling of ¹⁴NH₄ ⁺ (or ¹⁵NH₄ ⁺) was observed when ¹⁴N₂ (or ¹⁵N₂) was used as the feeding gas. In contrast, no significant ¹⁴NH₄ ⁺ (or ¹⁵NH₄ ⁺) signals were detected when the feeding gas was Ar (Figure S28, Supporting Information). The corresponding yields of ¹⁴N‐ and ¹⁵N‐labeled NH₃ obtained using ¹H NMR show that the amounts of NH₃ produced by Co₉S₈/Nb₂CT_x‐P in 0.1 m HCl were 65.75 and 59.39 µg h⁻¹ mg_cat. ⁻¹, respectively. In 0.1 m Na₂SO₄, these values decreased to 44.74 and 38.29 µg h⁻¹ mg_cat. ⁻¹, respectively. These results are similar to those obtained using the indophenol blue method (Figure 5b). The introduced gas and experimental environment were carefully controlled throughout the experiments. Therefore, we supposed that the difference between the indophenol blue and isotopic labeling methods originated from experimental errors.

Figure 5.

¹⁵N₂, ¹⁴N₂ isotope labeling and indophenol blue experiments in 0.1 m HCl and 0.1 m Na₂SO₄. a) Baseline‐subtracted ¹H NMR spectra of the post‐electrolyte with ¹⁵N₂ and ¹⁴N₂ compared with the reference ¹⁴NH₄Cl and ¹⁵NH₄Cl. b) Comparison of the NH₃ yields calculated by NMR and indophenol blue tests. c) Cycling stability of Co₉S₈/Nb₂CT_x‐P at −0.40 V versus RHE in 0.1 m HCl. d) The NH₃ yield and FE of original and post‐eNRR electrolysis in 0.1 M HCl and 0.1 M Na₂SO₄. e) The corresponding electrochemical double layer capacitances and f) the EIS Nyquist plots of Co₉S₈, Co₉S₈/Nb₂CT_x, and Co₉S₈/Nb₂CT_x ‐P at −0.4 V versus RHE in 0.1 m HCl. Inset: equivalent circuit model, where R _s is the electrolytic resistance, R _ct is charge‐transfer resistance, and CPE (constant phase angle element) is the double‐layer capacitance.

The stability of Co₉S₈/Nb₂CT_x‐P was assessed in 0.1 m HCl (Figure 5c) and 0.1 m Na₂SO₄ (Figure S29, Supporting Information). No substantial variations were observed in either the NH₃ yields or FEs during five cycles of chronoamperometric runs, indicating the excellent stability of Co₉S₈/Nb₂CT_x‐P. Moreover, the chronoamperometry curves indicated that the current density of Co₉S₈/Nb₂CT_x‐P toward the eNRR remained stable without significant fluctuation, for 24 h, in both 0.1 m HCl (Figure S30a, Supporting Information) and 0.1 m Na₂SO₄ (Figure S30b, Supporting Information). Consequently, no significant changes were observed in the NH₃ production rate or FE before and after 24 h (Figure 5d), verifying the high stability of Co₉S₈/Nb₂CT_x‐P toward the eNRR. In addition, the pH value of the post‐electrolyte after eNRR was studied, revealing a negligible difference in pH between the 0.1 m HCl and 0.1 m Na₂SO₄ electrolytes (Figure S31, Supporting Information). The SEM and TEM images and XRD patterns of Co₉S₈/Nb₂CT_x‐P after the long‐term tests in 0.1 m HCl (Figure S32, Supporting Information) and 0.1 m Na₂SO₄ (Figure S33, Supporting Information) also displayed no substantial changes, further confirming its good structural stability. The XPS characterizations of the spent electrocatalyst showed that the BE positions of the Co 2p (Figure S34a, Supporting Information) and S 2p XPS (Figure S34b, Supporting Information) spectra displayed positive shifts in both 0.1 m HCl and 0.1 m Na₂SO₄. This indicated that the electron density of the Co active sites in Co₉S₈ increased after use in the eNRR. In contrast, the Nb 3d XPS spectrum (Figure S34c, Supporting Information) displayed a negative shift, revealing a decrease in the electron density of the Nb sites. Collectively, these observations suggest that electron transfer occurred between Co₉S₈ and Nb₂CT_x, indicating charge reconstruction during the eNRR process.

To elucidate the catalytic mechanism of the eNRR, we conducted electrochemical active surface area and electrochemical impedance spectroscopy (EIS) measurements for all the catalysts in 0.1 m HCl and 0.1 m Na₂SO₄. Cyclic voltammetry curves of all samples (Figure S35, Supporting Information), obtained within the narrow potential window of 0.10–0.20 V versus Ag/AgCl (saturated KCl solution), were constructed to evaluate the double‐layer capacitance (C _dl), which represents the electrochemical surface area. Co₉S₈/Nb₂CT_x‐P displayed a C _dl value of 2.14 mF cm⁻², which was higher than that of Co₉S₈ (1.34 mF cm⁻²) and lower than that of Co₉S₈/Nb₂CT_x (3.52 mF cm⁻²) (Figure 5e). Moreover, the EIS Nyquist plots (Figure 5f) revealed that Co₉S₈/Nb₂CT_x‐P exhibited a smaller R _ct (21.3 Ω) compared to those of Co₉S₈ (41.2 Ω) and Co₉S₈/Nb₂CT_x (60.0 Ω). In 0.1 m Na₂SO₄, the C _dl of Co₉S₈/Nb₂CT_x‐P (4.20 mF cm⁻²) was slightly higher than that of Co₉S₈/Nb₂CT_x (3.60 mF cm⁻²) (Figure S36, Supporting Information). Similarly, the EIS Nyquist plot of Co₉S₈/Nb₂CT_x‐P in 0.1 m Na₂SO₄ (Figure S37, Supporting Information) indicated a smaller R _ct (54.6 Ω) compared to those of Co₉S₈ (110.0 Ω) and Co₉S₈/Nb₂CT_x (71.2 Ω). These results suggest that Co₉S₈/Nb₂CT_x‐P facilitates faster electron transfer and possesses more catalytically active sites than the other catalysts, thereby enhancing the eNRR performance at a relatively low overpotential.

Given the superior eNRR ability demonstrated by the constructed Co₉S₈/Nb₂CT_x‐P M–S junction, assembly of a Zn–N₂ aqueous battery for NH₃ synthesis and simultaneous electricity generation is envisioned.^{[ 35 ]} To explore this possibility, we conceptually manufactured a home‐made Zn–N₂ battery composed of a cathode of Co₉S₈/Nb₂CT_x‐P coated on the CP and an anode of Zn foil. A photograph of the reaction cell is shown in Figure  6a. N₂ was continuously bubbled into the cathode within an asymmetric electrolyte system comprising an acid catholyte and alkaline anolyte separated by a bipolar membrane.

Figure 6.

a) Schematic diagram of a rechargeable Zn–N₂ battery. b) Discharging polarization curves and the power density plot of the Co₉S₈/Nb₂CT_x‐P‐assembled Zn–N₂ battery. c) Comparison of the power densities of the Zn–N₂ battery assembled with different catalysts. d) The discharge curve of the constructed Zn–N₂ battery with Co₉S₈/Nb₂CT_x‐P at 0.01 mA cm⁻² for 7200 s. e) The NH₃ yields produced by the Zn–N₂ battery with Co₉S₈/Nb₂CT_x‐P for five repeated tests. f) The discharge curve of the manufactured Zn–N₂ battery at the current density of 0.01 mA cm⁻².

The rate–discharge curve of Co₉S₈/Nb₂CT_x‐P in Figure 6b displays the voltage changes at different current densities. As intended, the Zn–N₂ battery assembled with Co₉S₈/Nb₂CT_x‐P delivered a high discharge current density, peaking at 48.8 mA cm⁻² at 0.10 V versus Zn²⁺/Zn. Furthermore, Figure 6b shows that the Co₉S₈/Nb₂CT_x‐P‐based Zn–N₂ cell can deliver a power density of 11.17 mW cm⁻², significantly outperforming some reported Zn–N₂ batteries (Figure 6c), including Co_xNi_3‐x(HITP)₂/BNSs‐P (2.5 mW cm⁻²),^{[ 36 ]} CoPi/NPCS (0.49 mW cm⁻²),^{[ 37 ]} and CoPi/HSNPC (0.31 mW cm⁻²).^{[ 38 ]} After the constructed cell was discharged at 0.01 mA cm⁻² for 7200 s (Figure 6d), an average NH₃ yield of 11.8 µg h⁻¹ cm⁻² (Figure 6e) was determined based on five repeated tests. Additionally, a high energy density of 124 mA h g⁻¹ was achieved at 0.01 mA cm⁻² (Figure 6f).

To thoroughly elucidate the electrocatalytic pathway of the eNRR on Co₉S₈/Nb₂CT_x‐P, time‐dependent in situ ATR‐FTIR spectra were recorded in 0.1 m N₂‐saturated HCl at −0.40 V versus RHE. As depicted in Figure  7a, the appearance of the absorption peak at 1104 cm⁻¹, corresponding to the N─N tensile bond, suggests the dissociation of the N≡N bond of the adsorbed N₂ molecules on the working electrode surface.^{[ 39 ]} Additionally, the three adsorption peaks at 1272, 1442, and 3278 cm⁻¹ were assigned to ─NH₂ vibration, H─N─H bending, and N─H stretching vibrations, respectively.^{[ 40 ]} Notably, the intensities of these peaks increased with increasing reaction time. Consequently, we supposed that the reaction intermediates accumulate during the eNRR process. Most importantly, the absorption peak at 1637 cm⁻¹, attributed to N₂ molecules adsorbed on the catalyst surface, gradually decreases in intensity, indicating the consumption of N₂ molecules.^{[ 36 ]} These results suggest that the eNRR over the Co₉S₈/Nb₂CT_x‐P surface may follow an alternative pathway. Similar results were observed in the in situ ATR‐FTIR spectra of the electrocatalyst recorded in 0.1 m Na₂SO₄ (Figure S38, Supporting Information), indicating the same catalytic pathway.

Figure 7.

a) In situ electrochemical FT‐IR spectra and b) in situ Raman spectra on Co₉S₈/Nb₂CT_x‐P during the eNRR in 0.1 m HCl. c) The N₂ adsorption energy of Nb₂CT_x, Co₉S₈, Co₉S₈‐P, Co₉S₈/Nb₂CT_x, and Co₉S₈/Nb₂CT_x‐P. d) The projected density of states of Co 3$d_{x^2}-d_{y^2}$ orbitals for Co₉S₈, Co₉S₈/Nb₂CT_x, and Co₉S₈/Nb₂CT_x‐P. e) The schematic illustration of the shift of $d_{x^2}-d_{y^2}$ orbital toward E _F on the facilitation of the protonation of Co₉S₈/Nb₂CT_x‐P, E _H, and E _BC represent the $d_{x^2}-d_{y^2}$ highest occupied crystal orbital and band center, respectively. f) Partial density of states of N₂ adsorption on Co₉S₈, Co₉S₈‐P, Co₉S₈/Nb₂CT_x, and Co₉S₈/Nb₂CT_x‐P. g) Charge density difference of N₂ adsorption on Co₉S₈/Nb₂CT_x‐P (red sphere: S, blue sphere: Co, brown sphere: Nb, and gray sphere: C. Yellow and cyan‐colored iso‐surfaces show electron gain and loss, respectively). h) Calculated Gibbs free energies of the eNRR on Co₉S₈‐P, Co9S8, Co₉S₈/Nb₂CT_x, and Co₉S₈/Nb₂CT_x‐P along alternating pathways.

To identify the operative active sites of Co₉S₈/Nb₂CT_x‐P, the eNRR was further monitored via electrochemical in situ Raman spectroscopy at various potentials (Figure 7b). Notable Co─N bond formation was observed at 263 cm⁻¹ with increasing applied potential,^{[ 41 ]} indicating that N₂ molecules were predominately captured by the Co site (Figure 7b). Furthermore, the significant peak at ≈1578.5 cm⁻¹ was assigned to H─N─H tensile vibration (Figure 7b).^{[ 42 ]} Notably, the intensity of this peak decreased with increasing potential, possibly due to NH₃ accumulation. The strongest Co─N bond was observed at −0.40 V versus RHE in the potential‐dependent in situ Raman spectra, consistent with experimental findings.

To further elucidate the activation process of N₂ at the Co site and the evolution of intermediates during the eNRR process, DFT calculations were performed to provide deeper insight into the electrocatalytic mechanism of the prepared catalyst toward the NRR. As depicted in Figure 7c, the adsorption energy of N₂ molecules on the Co₉S₈/Nb₂CT_x‐P surface was higher compared to that of the plasma‐modified Co₉S₈ (Co₉S₈‐P) and markedly higher than those of pristine Co₉S₈, Nb₂CT_x, and Co₉S₈/Nb₂CT_x. This finding reveals that the formation of a microinterface between the diverse components and the generation of rich S vacancies through plasma modification can enhance N₂ adsorption, thereby improving the eNRR performance. In terms of the density of states (DOS) (Figure S39, Supporting Information), Co₉S₈/Nb₂CT_x‐P displayed an increased integral area of the net spin‐up compared to both Co₉S₈ and Co₉S₈/Nb₂CT_x. This observation confirms that the S_vs and interface effects lead to an increase in the number of spin‐up electrons, which is consistent with the ZFC temperature‐dependent magnetic susceptibility experimental results. In addition, the broader electronic states of the e _g orbitals close to the Fermi level can improve the electron transfer mobility and provide a lower adsorption energy between the active site and reaction intermediate, boosting the eNRR performance.^{[ 7} , ^{43 ]} As illustrated in Figure 7d,e, the broader $d_{x^2}-d_{y^2}$ electron state enabled more efficient electron transfer between N₂ and the electrocatalyst, thereby facilitating the N₂ protonation step. This was verified by examining the projected DOS before and after N₂ adsorption (Figure 7f). The projected DOS (Figures 7f and S39, Supporting Information) revealed that the Co‐3d orbital of Co₉S₈/Nb₂CT_x‐P and the *N₂‐2p orbital can overlap more effectively compared to those of Co₉S₈ (Figure 7f), Co₉S₈‐P (Figure 7f), and Co₉S₈/Nb₂CT_x (Figure 7f). Therefore, the Co₉S₈/Nb₂CT_x‐P M–S heterojunction is more conducive to the adsorption and activation of N₂. Charge difference analysis (Figure S40, Supporting Information) demonstrated that electrons can accumulate at the edge of the interface between Co₉S₈ and Nb₂CT_x. Both negative (charge accumulation) and positive (charge depletion) charges are simultaneously present around *N₂ after N₂ adsorption (Figure 7g).^{[ 44 ]} Thus, the S_vs and their neighboring Co return their 3d electrons to the π* orbital of *N₂, further activating and polarizing N₂ molecules through an “acceptance–donation” mechanism.

We further calculated the Gibbs free energies for the eNRR catalyzed by Co₉S₈‐P, Co₉S₈/Nb₂CT_x‐P, Co₉S₈, and Co₉S₈/Nb₂CT_x (Figure  7h ). Remarkably, the high adsorption ability of N₂ molecules onto Co₉S₈/Nb₂CT_x‐P led to a low energy barrier (0.27 eV) for the first protonation RDS of the *N₂→*NNH formation process. In addition, Co₉S₈/Nb₂CT_x and Co₉S₈ exhibited the energy barriers of 2.19 and 2.27 eV in the first protonation step, respectively, which were markedly higher than those of Co₉S₈/Nb₂CT_x‐P (0.27 eV) and Co₉S₈‐P (0.34 eV) (Figure 7h). Moreover, the RDS barrier of Co₉S₈/Nb₂CT_x‐P is lower than those of most previously reported catalysts.^{[ 45 ]} The competing HER kinetics in the eNRR procedure driven by Co₉S₈/Nb₂CT_x and Co₉S₈/Nb₂CT_x‐P was calculated using DFT, as depicted in Figure S41 (Supporting Information). The result manifested that the Gibbs free energy of hydrogen (*H) adsorption on the Co₉S₈/Nb₂CT_x‐P (−0.26 eV) was remarkably lower than that of Co₉S₈/Nb₂CT_x (0.80 eV). Thereby, the *H species can be spontaneously adsorbed on the Co₉S₈/Nb₂CT_x‐P surface, finally forming the 1/2H product. These results hinted that the generated rich sulfur vacancies remarkably accelerated the formation of the transition intermediate of active *H species, thus boosting the rate‐determining step of the eNRR, i.e., the production of *NNH. In addition, Co₉S₈/Nb₂CT_x‐P showed lower energy of N₂ adsorption (Figure 7c, −0.87 eV) than that of hydrogen adsorption (−0.26 eV). It suggested the efficient HER suppression, thereby improving Faradaic efficiency and NH₃ yield of Co₉S₈/Nb₂CT_x‐P. These results indicate that the S_v abundance together with the synergistic effect of Co₉S₈ with Nb₂CT_x can promote the highly spin‐polarized state of the Co 3d orbital electrons. This activates N₂ and promotes the formation of *NNH, greatly enhancing the eNRR activity.

3. Conclusion

In summary, a new Co₉S₈/Nb₂CT_x‐P hybrid was synthesized by coupling Co₉S₈ with Nb₂CT_x, in which the spin state of the Co²⁺ catalytic center was successfully regulated to improve the efficiency and selectivity of the eNRR. The transformation of the Co 3d orbital electron configuration from low (t_2g ⁶e_g ¹) to high (t_2g ⁵e_g ²) spin was induced by S_v and interface effects. The resulting highly spin‐polarized state of the Co₉S₈/Nb₂CT_x‐P hybrid demonstrated superior eNRR performance in acid and neutral electrolyte. Compared to the pristine Co₉S₈ and Co₉S₈/Nb₂CT_x hybrids, Co₉S₈/Nb₂CT_x‐P exhibited high NH₃ yields of 62.62, and 41.47 µg h⁻¹ mg_cat. ⁻¹ coupled with FEs of 30.33% and 23.19% in 0.1 m HCl and 0.1 m Na₂SO₄, respectively, surpassing the performances of most reported NRR electrocatalysts. DFT calculations indicated that this improvement in catalytic performance is related to the increase and upward shift of the Co $d_{x^2}-d_{y^2}$ orbitals, which significantly promotes the adsorption and activation of N₂ and reduces its protonation energy barrier. This study demonstrates the potential of heterostructured materials in designing high‐performance spin‐based catalysts, highlighting the importance of optimizing the spin states for the eNRR.

Conflict of Interest

The authors declare no conflict of interest.

Supporting information

Supporting Information

Zhang S., Zhao W., Liu J., Tao Z., Zhang Y., Zhao S., Zhang Z., Du M., Spin Manipulation of Co sites in Co₉S₈/Nb₂CT_x Mott–Schottky Heterojunction for Boosting the Electrocatalytic Nitrogen Reduction Reaction. Adv. Sci. 2024, 11, 2407301. 10.1002/advs.202407301

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.