English

• The decarbonization of ammonia (NH₃) synthesis is a critical challenge for sustainable transformation of the global chemical and energy industries. As the cornerstone of fertilizer production and a promising carbon-free energy carrier, NH₃ is currently synthesized predominantly via the Haber–Bosch process, which operates under high temperatures and pressures and relies heavily on fossil-derived energy, generating substantial CO₂ emissions [1-5]. Thus, the development of alternative NH₃ synthesis routes compatible with renewable electricity and mild operating conditions is urgently needed. Non-thermal plasma (NTP) catalysis has emerged as a promising strategy to overcome the kinetic limitations associated with N₂ activation under room temperature and atmospheric pressure [6-8]. By generating vibrationally excited molecules, ions, and reactive radicals, plasma enables reaction pathways that are inaccessible under thermal equilibrium [9,10]. Despite extensive research efforts, most plasma-catalytic NH₃ synthesis systems remain constrained by low energy efficiency and limited NH₃ yield. These limitations primarily stem from the lack of tailored catalysts capable of synergistically exploiting plasma-generated species to achieve comprehensive and efficient N₂ activation.

  Existing plasma-catalytic NH₃ synthesis catalysts are predominantly metal-based or metal nitride/oxide materials, which operate through fundamentally distinct N₂ activation mechanisms. For nitride-based catalysts (e.g., TiN and VN), N₂ activation proceeds via a surface Mars–van Krevelen (MvK) mechanism mediated by nitrogen vacancies (NVs) [11,12]. In this pathway, lattice nitrogen atoms are hydrogenated by H₂ plasma to form NH₃, generating NVs that are subsequently replenished by N₂ plasma to complete the catalytic cycle. Although this mechanism exhibits superior compatibility with plasma conditions and reduces the N₂ activation barrier relative to the conventional dissociative/associative pathways of metal-based catalysts, its performance remains inherently constrained by its reliance on surface-active sites and the relatively sluggish hydrogenation of lattice nitrogen, which arises from limited H₂ activation capability of nitride materials. In contrast, metal oxides catalysts (e.g., CeO₂ and In₂O₃) excel at H₂ activation under plasma discharge, efficiently dissociating H₂ and promoting the desorption of surface-adsorbed hydrogen as hydrogen radicals (H^•), which are potent reducing agents and direct hydrogen sources that initiate a radical-dominated gas-phase N₂ reduction pathway with markedly improved kinetics [13-15]. However, oxides lack intrinsic N₂ activation sites and thus function primarily as “radical generators” in plasma-catalytic systems, only indirectly facilitating the reduction of gaseous N₂ molecules. Moreover, rapid H₂ dissociation and efficient H^• production on oxide catalysts typically require high-power plasma discharge, resulting in substantial energy consumption and compromising the sustainability of the process. The independent investigation of nitride and oxide catalysts therefore exposes a fundamental challenge in plasma-catalytic NH₃ synthesis: Monofunctional materials are unable to integrate surface-mediated heterogeneous N₂ activation with gas-phase homogeneous N₂ activation within a single system. Addressing this challenge is essential for enhancing catalytic performance while reducing energy input, as maximizing NH₃ synthesis efficiency inherently necessitates the coexistence and synergistic coupling of both surface and gas-phase N₂ activation pathways in the plasma-catalytic system.

  Metal imide compounds (e.g., Ca₂NH) represent an ideal catalyst platform for realizing this integration, as they combine structural and functional characteristics of both nitrides and hydrides in a single material [16,17]. As a nitride-based compound, metal imide possesses a suitable metal–nitrogen (M–N) bond framework that supports the MvK mechanism, where surface N atoms can be hydrogenated to form NH₃, generating NVs that capture and activate plasma-excited N₂ molecules to regenerate the imide structure, thereby sustaining the surface catalytic cycle. Simultaneously, metal imides exhibit hydride-like hydrogen storage and release properties. Owing to the weak polar covalent nature of their M–H bonds, lattice hydrogen atoms in metal imides can be readily released as H^• radicals under plasma discharge. The resulting hydrogen vacancies (HVs) can be subsequently replenished by H₂ molecules, establishing a self-sustaining H^• release/replenishment cycle that continuously supplies reactive hydrogen species for gas-phase N₂ reduction. This dual functionality endows metal imides with the unique ability to activate N₂ through two parallel and complementary pathways, specifically an MvK-mediated surface pathway and an H^• radical-dominated gas-phase pathway. Specifically, a fraction of N₂ molecules is captured and activated on the imide surface via the MvK cycle, while unabsorbed N₂ species in the plasma phase are hydrogenated by H^• radicals released from the catalyst. The integration of these two mechanisms enables both heterogeneous and homogeneous N₂ activation in the plasma environment, thereby overcoming the limitations of single-site or single-pathway mechanisms and achieving comprehensive N₂ activation across the entire reaction space. Such synergy not only maximizes the utilization efficiency of plasma-derived hydrogen species but also reduces overall energy input by lowering the activation barriers of both surface and gas-phase processes. Despite their significant potential, metal imides such as Ca₂NH have remained largely unexplored in plasma-catalytic NH₃ synthesis. To date, no systematic investigations have elucidated their reaction mechanisms or clarified the synergistic effects between surface MvK mechanism and gas-phase radical pathways.

  In this work, we report Ca₂NH, a typical metal imide, as a dual-functional plasma catalyst for ultra-efficient NH₃ synthesis under room temperature and atmospheric pressure. Through kinetic analysis, isotopic labeling, radical trapping, and in situ spectroscopic characterization, we demonstrate that the integrated nitride-hydride nature of Ca₂NH enables simultaneous and synergistic surface and gas-phase N₂ activation. The synergistic interplay between these two pathways establishes an ultra-efficient N₂ activation mode, which accounts for the exceptional catalytic performance of Ca₂NH. This study not only introduces a novel catalyst design strategy for next-generation plasma catalysts but also provides fundamental insights into the coupling of surface catalytic reactions and gas-phase radical reactions, paving the way for the development of high-performance plasma-catalytic systems for sustainable N₂ fixation.

  The Ca₂NH catalyst was synthesized through a series of facile preparation steps. For comparative studies, the Ca₃N₂ and Ca₂N precursors used during the synthesis were also evaluated as catalyst samples. The Ca₃N₂ precursor was prepared from commercial calcium nitride powders via a typical liquid-phase exfoliation method [18]. Subsequently, the Ca₂N precursor was synthesized through a conventional solid-state reaction using Ca₃N₂ powder and Ca metal at a molar ratio of 1:1 [19]. The Ca₂NH catalyst was then obtained by heating Ca₂N powder at 400 ℃ under a flowing N₂/H₂ mixture. Powder X-ray diffraction (XRD) patterns (Figs. 1a–c) indicate that Ca₃N₂, Ca₂N and Ca₂NH exhibit characteristic diffraction peaks corresponding to the cubic Ca₃N₂ phase (JCPDS No. 22–0152), hexagonal Ca₂N phase (JCPDS No. 21–0837), and cubic Ca₂NH phase (JCPDS No. 26–0308), respectively. Transmission electron microscopy (TEM) images (Figs. S1a–c in Supporting information) reveal that all three catalysts possess a nanosphere morphology composed of aggregated nanoparticles. High-resolution TEM images (Figs. S1d–f in Supporting information) further display well-defined lattice fringes with spacings of 0.331, 0.182 and 0.293 nm, which can be indexed to the (222) plane of Ca₃N₂, the (110) plane of Ca₂N, and the (222) plane of Ca₂NH, respectively. The valence state of Ca in the catalysts were investigated using Ca K-edge X-ray absorption near-edge structure (XANES) spectroscopy. As shown in Fig. 1d, the absorption edge positions of Ca₃N₂, Ca₂N, and Ca₂NH fall within the range of 4036−4040 eV, indicating a formal Ca(Ⅱ) oxidation state in all samples [20]. The slightly lower absorption edge position observed for Ca₂N can be attributed to its electride nature. Overall, these results demonstrate that the three catalysts possess highly similar crystal structures, morphologies and element valence states; therefore, structural differences are not expected to influence the comparative analysis of their catalytic performance in this study.

  Figure 1

  

  Figure 1.  XRD patterns of (a) Ca₃N₂, (b) Ca₂N, and (c) Ca₂NH. (d) Ca K–edge XANES spectra of the catalysts. (e) NH₃ yield of the catalysts and catalyst-free conditions at varying input powers (gas flow rate = 50 mL/min, N₂:H₂ = 2:3, catalyst loading = 1 g, reaction temperature = 25 ℃, pressure = 0.1 MPa, sampling time = 1.0 h). (f) Energy efficiency of the catalysts and catalyst-free conditions at varying specific energy inputs (SEI) (gas flow rate = 50 mL/min, N₂:H₂ = 2:3, catalyst loading = 1 g, reaction temperature = 25 ℃, pressure = 0.1 MPa, sampling time = 1.0 h). (g) The modified Arrhenius plots for the catalytic systems; (h) Reported energy efficiency and NH₃ concentration in the plasma-catalytic systems over different catalysts. (i) Comparison of energy efficiency and NH₃ concentration in state-of-the-art sustainable ammonia synthesis technologies.

  DownLoad: Full-Size Img PowerPoint

  To evaluate the plasma-catalytic performance of the prepared catalysts, their NH₃ synthesis activity was first examined in a dielectric barrier discharge (DBD) reactor (Fig. S2 in Supporting information) using a fixed N₂/H₂ mixture ratio of 2:3 and a catalyst loading of 1 g. As shown in Figs. 1e and f, the Ca₂NH catalyst achieved an ultrahigh NH₃ yield of 371.8 mg h^–1 g^–1 and an exceptional energy efficiency of 9.30 g/kWh at an input power of 40 W, representing 1.35- and 1.74-fold enhancements compared with the Ca₂N and Ca₃N₂ catalysts, respectively. Notably, all three catalysts exhibited significantly higher performance than the catalyst-free system, demonstrating the critical role of catalysts in enhancing plasma-assisted ammonia synthesis. Electrical diagnostics (Fig. S3 and Table S1 in Supporting information) reveal that catalyst packing within the discharge zone slightly improved the discharge properties, further confirming the synergistic advantages of catalyst-plasma coupling. Optimization tests conducted under varying gas flow rates and reactant gas ratios (Fig. S4 in Supporting information) ensured that all measurements were performed under optimal reaction conditions while avoiding excessive hydrogen consumption. Furthermore, rigorous control experiments were conducted to confirm that NH₃ production originated from plasma catalysis rather than possible experimental contamination (Figs. S5a and b in Supporting information), directly verifying the reliability and validity of the experimental data. The apparent activation energy (E_a) for plasma-catalytic NH₃ synthesis was determined using a modified Arrhenius plot (Fig. 1g). Instead of using conventional thermal temperature, this method adopts specific energy input (SEI) as the governing parameter, which correlates the plasma equivalent temperature with discharge power [10]. Compared to the Ca₂N (2.60 kJ/mol) and Ca₃N₂ (3.14 kJ/mol) catalysts, Ca₂NH significantly reduced the activation energy to 2.03 kJ/mol, indicating its superior catalytic activity. To preliminarily assess the practical applicability of the Ca₂NH catalyst for scalable ammonia production, its catalytic performance was further evaluated under varying catalyst loadings. As illustrated in Fig. S6a (Supporting information), both the NH₃ productivity and the energy efficiency increased steadily with increasing catalyst loading under constant plasma discharge conditions (Fig. S6b in Supporting information), highlighting the potential scalability of the metal imide-enhanced plasma system. A comprehensive comparison with previously reported plasma-based NH₃ synthesis systems (Fig. 1h and Table S2 in Supporting information) further underscores the advantages of the prepared Ca₂NH catalyst, which outperforms most existing plasma catalysts in terms of both NH₃ productivity and energy utilization. Notably, the Ca₂NH catalyst not only achieved the highest NH₃ concentration of 1.63 × 10⁵ ppm among recent reports, but also exceeded the concentration threshold required for small-scale NH₃ synthesis units designed to minimize separation and recycling energy costs under atmospheric-pressure operation [6,21]. A summary of NH₃ yields and energy efficiencies across various state-of-the-art sustainable technologies is presented in Fig. 1i and Tables S3–S5 (Supporting information). Remarkably, the NH₃ synthesis performance of the Ca₂NH-plasma system not only surpasses other emerging alternatives such as electrochemical N₂ reduction (eNRR), Li-mediated eNRR, and hybrid processes integrating plasma N₂ oxidation with electrocatalytic reduction (pNOR-eNO_x^–RR), but also approaches the production efficiency of the industrial Haber–Bosch process [22]. These results highlight the significant potential of metal imide-enhanced plasma catalysis for future decentralized and energy-efficient ammonia synthesis processes.

  To investigate the durability of the prepared catalysts, long-term stability tests lasting 500 h were carried out. As shown in Fig. 2a, all three catalysts maintained their high NH₃ yield throughout the tests, demonstrating excellent stability. XRD patterns and TEM images (Fig. S7 in Supporting information) showed negligible changes in the crystal and morphological structures of the catalysts, further confirming their durability under continuous plasma exposure conditions. In addition, the ammonia produced during the long-term tests were collected using a rotary evaporation method [23]. As shown in Fig. 2b, approximately 94.2% of the produced NH₃ was successfully collected as solid ammonium chloride (NH₄Cl), and its high purity was verified by XRD analysis. Over the full 500-h plasma-catalytic reaction, the Ca₂NH, Ca₂N, and Ca₃N₂ catalyst produced 175.1, 129.3, and 100.5 g of solid NH₄Cl products, respectively. These results demonstrate not only the substantial ammonia production capacity of the catalysts, but also their excellent long-term operational stability under plasma conditions. Electron paramagnetic resonance (EPR) spectroscopy and positron annihilation lifetime spectroscopy (PALS) were employed to probe the plasma-induced release of lattice nitrogen and hydrogen from the catalyst surfaces and to investigate the distribution of anion vacancies at the atomic scale. As shown in Fig. 2c, all three catalysts exhibited EPR signals at g = 2.003 after the 500-h longevity tests, corresponding to electrons captured in anion vacancies [12]. In contrast, the fresh catalysts showed no detectable EPR signals (Fig. S8 in Supporting information), confirming that these vacancies were generated during the plasma-catalytic reaction. Notably, the EPR signal intensity of Ca₂NH is significantly higher than that of Ca₂N and Ca₃N₂, which may be attributed to the simultaneous formation of both nitrogen and hydrogen vacancies in Ca₂NH during the reaction. Furthermore, after treating the used catalysts with N₂ plasma, the EPR signals of Ca₂N and Ca₃N₂ disappeared completely, whereas Ca₂NH retained a relatively strong vacancy signal (Fig. 2d). This observation provides direct evidence that both nitrogen and hydrogen vacancies are generated in Ca₂NH under reaction conditions. PALS measurements further elucidated the characteristic of these defects. As displayed in Fig. 2e, the positron decay in Ca₂NH proceeds significantly more slowly than in Ca₂N and Ca₃N₂, indicating that Ca₂NH possesses a higher proportion of long-lifetime positrons. According to the previous report, PALS spectra can be deconvoluted into three components of τ₁, τ₂, and τ₃, with the intensities of I₁, I₂, and I₃, respectively [24-26]. Here, τ₁ corresponds to free positron annihilation and annihilation within isolated anion vacancies; τ₂ arises from positrons trapped by larger size defects, specifically hydrogen-vacancy clusters; and τ₃ is attributed to orthopositronium annihilation within large voids. Significantly, Ca₂NH exhibits a considerably larger τ₂ value of 501.3 ps (Table S6 in Supporting information) along with a markedly higher I₂ intensity of 68.7% (Fig. 2f) than Ca₂N and Ca₃N₂. These results indicate that the vacancy clusters on the Ca₂NH surface primarily consist of hydrogen vacancies, consistent with their more extensive surface distribution and lower formation energies compared with nitrogen vacancies.

  Figure 2

  

  Figure 2.  (a) Stability of the catalysts over a 500-h reaction teat (input power = 40 W, gas flow rate = 50 mL/min, N₂:H₂ = 2:3, catalyst loading = 1 g, reaction temperature = 25 ℃, pressure = 0.1 MPa). (b) XRD patterns of NH₄Cl product obtained after 500 h of reaction with various catalysts. (c) EPR spectra of the catalysts after 500 h of reaction. (d) EPR spectra of the used catalysts after being treated with N₂ plasma at 40 W. (e) PALS results of the catalysts after 500 h of reaction. (f) Intensities of I₁, I₂, and I₃ in the catalysts obtained via PALS.

  DownLoad: Full-Size Img PowerPoint

  Isotope tracking experiments were performed to determine whether lattice atoms on the catalyst surfaces participate in the plasma-catalytic NH₃ synthesis process. As presented in the ¹H nuclear magnetic resonance (¹H NMR) spectra (Fig. 3a), both ¹⁴NH₄⁺ and ¹⁵NH₄⁺ were detected when a ¹⁵N₂/H₂ mixture was used as the feed gas in the plasma reactor, providing direct evidence that NH₃ synthesis over the prepared catalysts proceeds via the MvK mechanism [27]. Furthermore, when Ca₂NH was used as the catalyst and a ¹⁴N₂/D₂ mixture was introduced, distinct peaks at 7.05, 7.14 and 7.23 ppm appeared in the ¹H NMR spectrum (Fig. 3b), corresponding to the characteristic signals of ND₄⁺ [28]. In addition, typical NH₄⁺ peaks at 7.03, 7.12, and 7.20 ppm were also observed, indicating that a portion of the hydrogen in the produced ammonia originated from the lattice hydrogen atoms within the Ca₂NH catalysts. Following the D₂ isotope tracking experiments, the XRD spectrum of the Ca₂NH catalyst (Fig. S9 in Supporting information) revealed a slight shift in the crystal planes relative to the fresh samples, consistent with the introduction of deuterium (D) into the lattice. The nearly unchanged spacing between the diffraction peaks confirmed that the crystal structures of Ca₂NH remained intact, with deuterium merely substituting for hydrogen in the lattice. These findings further confirm that the lattice hydrogen in Ca₂NH actively participated in the hydrogenation of N₂ molecules. During the plasma-catalytic NH₃ synthesis reaction, lattice hydrogen is released from the Ca₂NH catalyst and converted into H^• radicals, while H₂ plasma replenishes the resulting hydrogen vacancies. This reversible equilibrium between lattice hydrogen and H^• radicals not only enhances the reaction efficiency but also contributes to the remarkable catalytic stability of the system.

  Figure 3

  

  Figure 3.  (a) ¹⁵N₂ isotope tracking experiments for various catalysts. (b) D₂ isotope tracking experiments for various catalysts. Kinetic isotope effects (H/D) on NH₃ synthesis rates for (c) Ca₂NH, (d) Ca₂N and (e) Ca₃N₂. The reaction rate dependence on (f) P_N2 and (g) P_H2 under plasma-catalytic conditions for various catalysts.

  DownLoad: Full-Size Img PowerPoint

  To gain deeper insight into the catalytic enhancement of metal imides in NH₃ synthesis, kinetic experiments were conducted, beginning with the investigation of the hydrogen/deuterium (H/D) kinetic isotopic effect (KIE) to assess the influence of Ca₂NH on H₂ dissociation. As shown in Figs. 3c–e, substituting H₂ with D₂ led to a significant decrease in NH₃ production for both Ca₂N and Ca₃N₂ catalysts, with corresponding KIE values of 1.36 and 1.41, respectively. In contrast, the Ca₂NH catalyst exhibited a notably lower KIE value of only 1.08, suggesting its superior ability to dissociate H₂ molecules and generate H^• radicals, which represents an essential factor contributing to its outstanding performance in NH₃ synthesis [29,30]. Furthermore, the reaction orders with respect to N₂ partial pressure (P_N2) were determined to be 0.07, 0.18 and 0.22 for the Ca₂NH, Ca₂N, and Ca₃N₂ systems, respectively (Fig. 3f). The extremely low N₂ reaction order for the Ca₂NH system suggests that N₂ hydrogenation is highly favorable in this catalytic system [11]. Additionally, the relatively higher reaction orders for H₂ in the NTP-Ca₂NH system (Fig. 3g) further support the conclusion that the Ca₂NH-enhanced reaction is primarily driven by the reversible equilibrium between lattice hydrogen and H^• radicals, which is consistent with the results from the isotope tracking experiments and KIE analysis.

  A comprehensive in situ characterization system was employed to further elucidate the origin of the outstanding performance of Ca₂NH. In situ diffuse-reflectance infrared Fourier transform spectroscopy (DRIFTS) was first conducted to identify surface intermediates during plasma-catalytic ammonia synthesis. As displayed in Fig. 4a, stretching vibrations corresponding to N≡N (1974 cm⁻¹) and N=N (1426 cm⁻¹) were simultaneously observed, indicating the formation of surface adsorbed NNH species during the reaction. To further confirm this result, isotope substitution experiments were performed using a ¹⁵N₂/H₂ mixture. Notably, the N≡N and N=N stretching bands red-shifted to 1902 and 1381 cm⁻¹, respectively. These shifts are consistent with the expected isotope effect (1974 cm⁻¹ × (14/15)^1/2 = 1907 cm⁻¹, 1439 cm⁻¹ × (14/15)^1/2 = 1377 cm⁻¹), thereby confirming the formation of surface adsorbed NNH intermediate [31]. The presence of these NNH species is attributed to reaction intermediates generated via the MvK mechanism on the catalyst surfaces. Moreover, 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) was employed as a spin-trapping agent for in situ radical trapping experiments. For the Ca₂NH catalyst, a characteristic EPR signal with an intensity ratio of 1:1:2:1:2:1:2:1:1 was observed (Fig. 4b), which can be unambiguously assigned to gaseous H^• radicals [32-34]. In contrast, no detectable H^• signal was observed when Ca₂N or Ca₃N₂ was used, indicating that the H^• concentration in these systems was below the detection limit. These results demonstrate that hydrogen atoms from Ca₂NH can be released into the gas phase under plasma exposure, thereby generating H^• radicals while simultaneously creating hydrogen vacancies on the catalyst surface. Optical emission spectroscopy (OES) was further employed to validate these findings. As shown in Fig. 4c, the presence of the H_α Balmer emission line confirms the dissociation of H₂ into gas-phase H^• radical. The significantly stronger OES signal observed for the Ca₂NH system is consistent with the EPR results, further supporting the enhanced H^• generation over Ca₂NH. Therefore, we propose that the Ca₂NH catalyst not only synthesizes ammonia through the MvK mechanism but also that H^• radicals released from the lattice induce a gas-phase radical pathway for N₂ reduction. To further substantiate the existence of this radical-dominated reaction pathway, radical trapping experiments targeting NNH^• radicals were performed. As exhibited in Fig. 4d, EPR signals corresponding to gaseous NNH^• radicals were exclusively detected in the Ca₂NH system. The identification of this key nitrogen-containing intermediate provides compelling evidence that H^• radicals derived from Ca₂NH participate in N₂ reduction through a gas-phase radical pathway [35]. Furthermore, a self-designed in situ molecular-beam mass spectrometry (MBMS) system (Fig. 4e) was employed to directly monitor gas-phase N_xH_y radicals during plasma-catalytic ammonia synthesis [36]. As shown in Fig. 4f, markedly enhanced signals for NNH^• (m/z 29), NHNH^• (m/z 30), NHNH₂^• (m/z 31), NH₂NH₂^• (m/z 32), NH₂NH₃^• (m/z 33) and NH₂^• (m/z 16) were detected with the Ca₂NH catalyst. These findings strongly demonstrate that H^• radicals derived from Ca₂NH readily hydrogenate N₂ molecules to form NNH^• intermediates, which are further transformed into a series of N_xH_y radicals and finally reduced to NH₃ through a continuous gas-phase reaction pathway.

  Figure 4

  

  Figure 4.  (a) In situ DRIFTS measurements conducted with various catalysts at input powers of 40, 60, 80, 100 and 120 W (gas flow rate = 50 mL/min, N₂:H₂ = 2:3, catalyst loading = 1 g, reaction temperature = 25 ℃, pressure = 0.1 MPa). (b) DMPO-involved EPR spectra for plasma-induced H^• species. (c) OES spectra for the plasma-catalytic system with various catalysts. (d) DMPO-involved EPR spectra for plasma-induced NNH^• species. (e) Schematic diagram of the in situ MBMS system. (f) MS signals of gas phase N_xH_y radicals detected by MBMS during plasma-catalytic ammonia synthesis with various catalysts.

  DownLoad: Full-Size Img PowerPoint

  Density functional theory (DFT) calculations were performed to further elucidate the reaction pathways and identify the rate-determining steps (RDS), with insights obtained from the in situ characterization results. These calculations primarily serve to compare the RDS among different catalysts, thereby fundamentally accounting for the differences in reaction rates and the resulting catalytic activity trends. In situ DRIFTS, OES and MBMS measurements collectively confirm that ammonia synthesis proceeds predominantly via a surface MvK mechanism on Ca₂N and Ca₃N₂ surfaces. As illustrated in Fig. 5a, the RDS in this MvK pathway corresponds to the hydrogenation of terminal N atoms in N₂ molecules adsorbed at nitrogen vacancies. The calculated thermodynamic energy barriers for the RDS are 2.630 eV for Ca₂N and 3.028 eV for Ca₃N₂. In contrast, for the Ca₂NH catalyst, MBMS and DFT results reveal that surface lattice hydrogen atoms can be released into the gas phase under plasma discharge conditions, with an energy barrier of 2.476 eV (Fig. 5b). This process generates H^• radicals while simultaneously creating hydrogen vacancies that serve as active sites for H₂ adsorption and dissociation. Upon H₂ adsorption, one hydrogen atom occupies the surface vacancy, while the other remains weakly bound to it. Subsequent plasma-induced cleavage of the H–H bond produces a lattice hydrogen atom and a gaseous H^• radical. This reversible H-release/H-refill cycle enables a continuous and energy-efficient supply of reactive H^• species for N₂ reduction. The overall gas-phase reaction pathway (Fig. 5c) indicates that, in the presence of abundant H^• radicals, N₂ molecules are readily hydrogenated to form NNH^• radicals with an ultralow energy barrier of 0.123 eV. These NNH^• radicals are further converted to a series of N_xH_y intermediates and ultimately reduced to NH₃ through a spontaneous gas-phase reaction pathway. Meanwhile, the released H^• radicals also act as reactive hydrogen species that promote the surface MvK mechanism on the Ca₂NH catalyst. As shown in Fig. 5d, the H^•-involved MvK pathway on Ca₂NH exhibits a significantly lower thermodynamic energy barrier compared to the conventional MvK pathways on Ca₂N and Ca₃N₂ surfaces. Within this dual-mechanism strategy, the release of H^• radicals from the Ca₂NH surface is identified as the RDS of ammonia synthesis. The corresponding energy barrier (2.476 eV) is lower than those associated with the RDS in the conventional MvK pathways on Ca₂N (2.630 eV) and Ca₃N₂ (3.028 eV). These results demonstrate that the Ca₂NH catalyst participate in N₂ reduction through two parallel and complementary pathways, specifically an H^•-enhanced MvK pathway and an H^• radical-dominated gas-phase pathway. As depicted in Fig. 5e, a fraction of N₂ molecules is captured and activated on the Ca₂NH surface via the H^•-enhanced MvK mechanism, while unabsorbed N₂ molecules in the plasma phase are hydrogenated by H^• radicals released from the catalyst. The integration of these pathways enables simultaneous heterogeneous and homogeneous activation of N₂ in the plasma environment, thereby overcoming the limitations of single-site or single-pathway mechanisms and achieving comprehensive N₂ activation throughout the entire reaction space.

  Figure 5

  

  Figure 5.  (a) Free-energy diagram for the MvK pathway on the Ca₂N and Ca₃N₂ catalysts. (b) Free-energy diagram for the plasma-catalytic H-release/H-refill cycle on the Ca₂NH catalyst. (c) Free-energy diagram illustrating the H^• radical-dominated N₂ reduction pathway via plasma-induced gas reactions. (d) Free-energy diagram illustrating the H^• radical-enhanced MvK pathway on the Ca₂NH catalyst. (e) Schematic diagram of the two plasma-catalytic reaction pathways using the Ca₂NH catalyst.

  DownLoad: Full-Size Img PowerPoint

  In summary, this work demonstrates that the metal imide Ca₂NH acts as a highly effective dual-functional catalyst for plasma-catalytic ammonia synthesis under mild conditions. By integrating nitride-like N₂ activation capability with hydride-like H^• release behavior, Ca₂NH enables the simultaneous progression of two complementary pathways: A surface MvK mechanism and an H^• radical-dominated gas-phase reduction route. This coupled mechanism significantly lowers the activation barrier for N₂ hydrogenation and maximizes the utilization of plasma-generated reactive species, resulting in an exceptional NH₃ yield of 371.8 mg h⁻¹ g⁻¹ and an energy efficiency of 9.30 g/kWh. In situ characterization, isotope labeling, and kinetic analyses collectively reveal that lattice hydrogen in Ca₂NH dynamically participates in the reaction through a reversible H-release/H-refill cycle, which continuously supplies H^• radicals for both surface and gas-phase N₂ activation while sustaining excellent catalyst stability over 500 h of continuous operation. These findings establish metal imides as a promising catalyst platform for plasma catalysis and highlight a general strategy for coupling heterogeneous surface reactions with homogeneous radical pathways to achieve efficient N₂ fixation. This work provides new insights into the design of advanced plasma-catalytic systems and paves the way for the development of sustainable, energy-efficient ammonia synthesis technologies compatible with renewable energy sources.

  CRediT authorship contribution statement

  Shijian Luo: Writing – original draft, Visualization, Software, Methodology, Investigation, Formal analysis, Conceptualization. Hao Chen: Software, Data curation. Yuran Yang: Software, Formal analysis. Yang Song: Validation, Data curation. Yongduo Liu: Validation, Software. Daojun Long: Validation, Resources, Funding acquisition. Siguo Chen: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition, Formal analysis, Conceptualization. Zidong Wei: Supervision, Resources, Project administration.

  Declaration of competing interest

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

  Acknowledgments

  This study was financially supported by the National Key R&D Program of China (No. 2024YFA1509201), the National Natural Science Foundation of China (Nos. 22579021, 22178034, 22508028), the Fundamental Research Funds for the Central Universities (No. 2025CDJ-IAISYB-048), the Postdoctoral Fellowship Program and China Postdoctoral Science Foundation (No. BX20250316), and the Chongqing Postdoctoral Science Special Foundation (No. 2024CQBSHTB3120).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2026.113084.

  
  
• 1. [1]

     H.R. Li, X. Kang, R. Gao, et al., Chin. Chem. Lett. 36 (2025) 109958. doi: 10.1016/j.cclet.2024.109958

  2. [2]

     X. Fu, J.B. Pedersen, Y. Zhou, et al., Science 379 (2023) 707–712. doi: 10.1126/science.adf4403

  3. [3]

     X. Peng, X. Hu, R. Mao, et al., Chin. Chem. Lett. 37 (2026) 111820. doi: 10.1016/j.cclet.2025.111820

  4. [4]

     C. Zhang, S. Shi, B. Fang, et al., Chin. Chem. Lett. 34 (2023) 107237. doi: 10.1016/j.cclet.2022.02.042

  5. [5]

     Y. Wang, M. Batmunkh, H. Mao, et al., Chin. Chem. Lett. 33 (2022) 394–398. doi: 10.1016/j.cclet.2021.05.025

  6. [6]

     K.H.R. Rouwenhorst, Y. Engelmann, K. Van ’T Veer, et al., Green Chem. 22 (2020) 6258–6287. doi: 10.1039/d0gc02058c

  7. [7]

     P. Mehta, P. Barboun, F.A. Herrera, et al., Nat. Catal. 1 (2018) 269–275. doi: 10.1038/s41929-018-0045-1

  8. [8]

     X. Zhang, Y. Liu, M. Zhang, et al., Chem 6 (2020) 3312–3328. doi: 10.1016/j.chempr.2020.09.016

  9. [9]

     Z. Zhang, C. Kondratowicz, J. Smith, et al., ACS Energy Lett. 10 (2025) 3349–3358. doi: 10.1021/acsenergylett.5c01034

  10. [10]

      Y. Wang, W. Yang, S. Xu, et al., J. Am. Chem. Soc. 144 (2022) 12020–12031. doi: 10.1021/jacs.2c01950

  11. [11]

      S. Luo, Y. Liu, Y. Song, et al., Chem. Commun. 61 (2025) 16066–16069. doi: 10.1039/d5cc04331j

  12. [12]

      S. Luo, Y. Liu, Y. Song, et al., Chem. Commun. 60 (2024) 3295–3298. doi: 10.1039/d4cc00042k

  13. [13]

      Y. Wang, J. Yang, Y. Sun, et al., Chem 10 (2024) 2590–2606. doi: 10.1016/j.chempr.2024.06.022

  14. [14]

      Y. Yang, L. Guo, S. Luo, et al., ACS Catal. 15 (2025) 14955–14965. doi: 10.1021/acscatal.5c01699

  15. [15]

      S. Luo, Y. Liu, L. Guo, et al., J. Mater. Chem. A 13 (2025) 30546–30553. doi: 10.1039/d5ta04447b

  16. [16]

      W. Gao, J. Guo, P. Wang, et al., Nat. Energy 3 (2018) 1067–1075. doi: 10.1038/s41560-018-0268-z

  17. [17]

      P. Chen, Z. Xiong, J. Luo, J. Lin, K.L. Tan, Nature 420 (2002) 302–304. doi: 10.1038/nature01210

  18. [18]

      H. Yu, X. Li, J. Zheng, ACS Catal. 14 (2024) 3139–3157. doi: 10.1021/acscatal.3c05696

  19. [19]

      M. Kitano, K. Yamagata, H. Hosono, Res. Chem. Intermed. 47 (2021) 235–248. doi: 10.1007/s11164-020-04332-3

  20. [20]

      H. Abe, Y. Niwa, M. Kitano, et al., Bull. Chem. Soc. Jpn. 90 (2017) 963–965. doi: 10.1246/bcsj.20170143

  21. [21]

      C. Smith, A.K. Hill, L. Torrente-Murciano, Energy Environ. Sci. 13 (2020) 331–344. doi: 10.1039/c9ee02873k

  22. [22]

      L. Hollevoet, F. Jardali, Y. Gorbanev, et al., Angew. Chem. Int. Ed. 59 (2020) 23825–23829. doi: 10.1002/anie.202011676

  23. [23]

      W. Liu, M. Xia, C. Zhao, et al., Nat. Commun. 15 (2024) 3524. doi: 10.1038/s41467-024-47765-9

  24. [24]

      L. Xu, Q. Wang, Q. Gu, et al., Angew. Chem. Int. Ed. 64 (2025) e202508091. doi: 10.1002/anie.202508091

  25. [25]

      H.J. Niu, N. Ran, W. Zhou, et al., J. Am. Chem. Soc. 147 (2025) 2607–2615. doi: 10.1021/jacs.4c14675

  26. [26]

      X.Y. Ran, T.J. Xiong, Y.P. Yang, et al., Chin. Chem. Lett. 37 (2026) 111864. doi: 10.1016/j.cclet.2025.111864

  27. [27]

      C. Hu, J. Zhao, R. Chang, et al., J. Am. Chem. Soc. 147 (2025) 37428–37436. doi: 10.1021/jacs.5c11505

  28. [28]

      J. Li, W. Yu, H. Yuan, et al., Nat. Commun. 15 (2024) 9499. doi: 10.24294/jipd9499

  29. [29]

      Y. Bai, Y. Zhang, J. Hu, et al., ACS Catal. 15 (2025) 1455–1466. doi: 10.1021/acscatal.4c06966

  30. [30]

      J. Lin, D. Liao, H. Zhang, H. Wan, K. Tsai, Chin. J. Catal. 31 (2010) 153–155. doi: 10.1016/S1872-2067(09)60040-1

  31. [31]

      Z. Zhang, K. Miyashita, T. Wu, et al., Nat. Chem. 17 (2025) 679–687. doi: 10.1038/s41557-025-01737-8

  32. [32]

      J. Li, G. Zhan, J. Yang, et al., J. Am. Chem. Soc. 142 (2020) 7036–7046. doi: 10.1021/jacs.0c00418

  33. [33]

      Y. Xi, C. Zhang, W. Tu, et al., Angew. Chem. Int. Ed. 64 (2025) e202505046. doi: 10.1002/anie.202505046

  34. [34]

      X. Feng, J. Liu, L. Chen, et al., J. Am. Chem. Soc. 145 (2023) 10259–10267. doi: 10.1021/jacs.3c01319

  35. [35]

      C. Zhang, Z.H. Wang, H. Wang, et al., Natl. Sci. Rev. 11 (2024) nwae251. doi: 10.1093/nsr/nwae251

  36. [36]

      H. Zhao, G. Song, Z. Chen, et al., ACS Energy Lett. 7 (2022) 53–58. doi: 10.1021/acsenergylett.1c02207

• 

  Figure 1  XRD patterns of (a) Ca₃N₂, (b) Ca₂N, and (c) Ca₂NH. (d) Ca K–edge XANES spectra of the catalysts. (e) NH₃ yield of the catalysts and catalyst-free conditions at varying input powers (gas flow rate = 50 mL/min, N₂:H₂ = 2:3, catalyst loading = 1 g, reaction temperature = 25 ℃, pressure = 0.1 MPa, sampling time = 1.0 h). (f) Energy efficiency of the catalysts and catalyst-free conditions at varying specific energy inputs (SEI) (gas flow rate = 50 mL/min, N₂:H₂ = 2:3, catalyst loading = 1 g, reaction temperature = 25 ℃, pressure = 0.1 MPa, sampling time = 1.0 h). (g) The modified Arrhenius plots for the catalytic systems; (h) Reported energy efficiency and NH₃ concentration in the plasma-catalytic systems over different catalysts. (i) Comparison of energy efficiency and NH₃ concentration in state-of-the-art sustainable ammonia synthesis technologies.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) Stability of the catalysts over a 500-h reaction teat (input power = 40 W, gas flow rate = 50 mL/min, N₂:H₂ = 2:3, catalyst loading = 1 g, reaction temperature = 25 ℃, pressure = 0.1 MPa). (b) XRD patterns of NH₄Cl product obtained after 500 h of reaction with various catalysts. (c) EPR spectra of the catalysts after 500 h of reaction. (d) EPR spectra of the used catalysts after being treated with N₂ plasma at 40 W. (e) PALS results of the catalysts after 500 h of reaction. (f) Intensities of I₁, I₂, and I₃ in the catalysts obtained via PALS.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) ¹⁵N₂ isotope tracking experiments for various catalysts. (b) D₂ isotope tracking experiments for various catalysts. Kinetic isotope effects (H/D) on NH₃ synthesis rates for (c) Ca₂NH, (d) Ca₂N and (e) Ca₃N₂. The reaction rate dependence on (f) P_N2 and (g) P_H2 under plasma-catalytic conditions for various catalysts.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) In situ DRIFTS measurements conducted with various catalysts at input powers of 40, 60, 80, 100 and 120 W (gas flow rate = 50 mL/min, N₂:H₂ = 2:3, catalyst loading = 1 g, reaction temperature = 25 ℃, pressure = 0.1 MPa). (b) DMPO-involved EPR spectra for plasma-induced H^• species. (c) OES spectra for the plasma-catalytic system with various catalysts. (d) DMPO-involved EPR spectra for plasma-induced NNH^• species. (e) Schematic diagram of the in situ MBMS system. (f) MS signals of gas phase N_xH_y radicals detected by MBMS during plasma-catalytic ammonia synthesis with various catalysts.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  (a) Free-energy diagram for the MvK pathway on the Ca₂N and Ca₃N₂ catalysts. (b) Free-energy diagram for the plasma-catalytic H-release/H-refill cycle on the Ca₂NH catalyst. (c) Free-energy diagram illustrating the H^• radical-dominated N₂ reduction pathway via plasma-induced gas reactions. (d) Free-energy diagram illustrating the H^• radical-enhanced MvK pathway on the Ca₂NH catalyst. (e) Schematic diagram of the two plasma-catalytic reaction pathways using the Ca₂NH catalyst.

  下载: 全尺寸图片 幻灯片
•