Recent Advances and Perspectives of Metal Clusters for Electrochemical Nitrogen Reduction

Abstract

The rapid intensification of anthropogenic nitrogen cycle perturbations and the energy‐intensive nature of Haber–Bosch process have motivated the pursuit of sustainable routes for ammonia synthesis. Electrocatalytic nitrogen reduction reaction (NRR) represents a compelling strategy to convert inert N₂ into NH₃, yet its practical realization remains constrained by formidable activation barriers, sluggish kinetics, and strong competition from hydrogen evolution reaction. Well‐defined metal clusters have emerged as a distinctive catalytic platform, offering tunable electronic structures and active motifs that bridge the gap between single‐atom catalysts and nanoparticles. Our review establishes a unified framework for understanding cluster‐enabled NRR, beginning with the physicochemical principles of N₂ activation, including orbital interactions, reaction energetics, and key activity descriptors, while considering electrolyte and proton‐coupled effects. We then analyze cluster characteristics such as size‐dependent electronic behavior, geometric configurations, and support interactions, highlighting their implications for catalytic selectivity. A systematic overview of electrocatalytic platforms is presented, including metal clusters, bimetallic systems, heterostructures, and nitrogenase‐inspired architectures, followed by critical perspectives on reproducibility, durability, and future development pathways. Looking ahead, the integration of operando characterization, theoretical modeling, and scalable synthesis is expected to accelerate the rational development of efficient and reliable cluster‐based NRR systems.

From N₂ activation fundamentals to advanced catalyst architectures, we establish a comprehensive framework for metal cluster‐based electrocatalytic nitrogen reduction in this review. It identifies the electronic and structural features governing activity and selectivity, evaluates current state‐of‐the‐art systems, and outlines future directions toward reliable ambient ammonia synthesis.

1. Introduction

Ammonia synthesis occupies a foundational role in sustaining modern society while simultaneously emerging as a strategic component in future carbon‐neutral energy systems. Its global production, exceeding 180 million tons per year, underpins agricultural productivity through nitrogen fertilization and is increasingly being evaluated as a hydrogen carrier due to its high energy density and established infrastructure [1]. Nevertheless, the industrial fixation of nitrogen remains dominated by the Haber–Bosch process, which is intrinsically energy‐intensive and responsible for significant CO₂ emissions due to its reliance on fossil‐derived hydrogen and elevated operating conditions [2]. These limitations have stimulated extensive efforts to identify alternative nitrogen fixation routes capable of operating under ambient conditions. From a molecular standpoint, the challenge originates from the exceptional thermodynamic stability and kinetic inertness of N₂, characterized by a strong triple bond and limited dipole moment, which collectively hinder its adsorption and activation on catalyst surfaces. Consequently, achieving efficient nitrogen conversion under mild conditions continues to represent a central objective in catalysis and energy research [3, 4, 5].

Electrocatalytic nitrogen reduction offers a compelling strategy by directly coupling renewable electricity with catalytic interfaces to drive ammonia formation through proton‐coupled electron transfer (PCET) sequences. This approach enables operation at ambient temperature and pressure while providing tunable reaction environments through applied potential and electrolyte engineering [6, 7]. Despite these conceptual advantages, the practical implementation of nitrogen reduction reaction (NRR) remains constrained by intrinsically slow activation kinetics and poor selectivity. In particular, the formation of the first hydrogenated intermediate (*NNH) constitutes a high‐energy barrier, which often limits overall reaction rates [8, 9, 10]. Concurrently, the hydrogen evolution reaction (HER) proceeds preferentially under similar electrochemical conditions, diverting electrons and protons away from nitrogen reduction [11]. Recent investigations have demonstrated that modifying catalyst electronic structures can alter adsorption energetics and intermediate stabilization; for example, atomically dispersed Fe–N–C systems and defect‐engineered Mo‐based catalysts have shown measurable NRR activity, yet their performance remains limited by scaling constraints and insufficient suppression of HER [12, 13, 14, 15]. These findings underscore the necessity for catalytic platforms that can simultaneously optimize N₂ adsorption, intermediate stabilization, and proton transfer kinetics. A critical limitation in current NRR catalysis arises from the intrinsic coupling between adsorption energies of key intermediates, which restricts independent optimization of elementary reaction steps. Catalysts that bind N₂ strongly often suffer from sluggish hydrogenation or product desorption, whereas weak‐binding surfaces fail to sufficiently activate the molecule. This trade‐off is further complicated by the difficulty in accurately quantifying ammonia at trace levels, where contamination from environmental or experimental sources can lead to misleading results [16, 17, 18, 19]. Advanced isotopic labeling experiments, such as those employing ¹⁵N₂, have revealed that many previously reported systems suffer from overestimated activity, highlighting the importance of rigorous validation protocols. Furthermore, electrolyte composition, local pH, and interfacial electric fields have been shown to influence reaction pathways and selectivity, adding additional layers of complexity to catalyst evaluation [20, 21, 22, 23]. Accordingly, progress in this field necessitates a combined focus on catalyst design, mechanistic elucidation, and methodological standardization.

Within this framework, metal clusters have emerged as a particularly promising class of electrocatalysts due to their ability to bridge the gap between isolated atomic sites and extended surfaces. Unlike single‐atom catalysts, which provide uniform but often electronically constrained active centers, and nanoparticles, which exhibit heterogeneous and less controllable surface sites, metal clusters offer discrete ensembles with tunable nuclearity and coordination environments [24]. This structural versatility enables precise modulation of electronic density and d‐band characteristics, thereby enhancing backdonation into the antibonding orbitals of adsorbed N₂ and facilitating bond weakening [25, 26]. Moreover, the presence of multiple adjacent metal atoms enables cooperative activation mechanisms, where different atoms within the cluster can stabilize distinct intermediates or participate in sequential protonation steps. Experimental evidence supports these advantages; for instance, subnanometer Ru clusters anchored on nitrogen‐doped carbon have demonstrated enhanced Faradaic efficiency due to optimized N₂ adsorption configurations, while Mo‐based clusters confined in porous matrices exhibit improved catalytic turnover attributed to synergistic electronic interactions among neighboring metal centers. Similarly, Fe‐containing clusters stabilized on conductive supports have been shown to promote associative NRR pathways through facilitated *NNH formation and reduced activation barriers, as corroborated by density functional theory (DFT) calculations and operando spectroscopic measurements [27, 28, 29, 30, 31].

Despite these advances, a unified understanding of structure–activity relationships in cluster‐based NRR systems remains underdeveloped. Many studies report improved performance without fully resolving the underlying mechanistic origins, and discrepancies in experimental conditions often hinder direct comparison. In particular, the interplay between cluster size, composition, and support interactions is not yet comprehensively understood, limiting the rational design of next‐generation catalysts. Furthermore, issues related to catalyst stability, dynamic structural evolution under reaction conditions, and reproducibility continue to challenge the field. These considerations highlight the necessity for a systematic and critically evaluated synthesis of current knowledge, with emphasis on mechanistic clarity and experimentally validated insights. In response to these challenges, the present review provides a focused and concept‐driven analysis of metal cluster‐based electrocatalysts for nitrogen reduction. We first delineate the fundamental principles governing N₂ activation, including its electronic structure, plausible reaction pathways, and the descriptors that dictate catalytic performance, while also addressing the influence of competing HER and electrolyte effects. Subsequently, we examine the structural and electronic characteristics of metal clusters, emphasizing their classification, size‐dependent electronic evolution, geometric configurations, and interactions with supporting materials, alongside comparisons with single‐atom catalysts and nanoparticles. Building on these insights, we discuss rational design strategies encompassing size control, compositional engineering, coordination environment optimization, and interfacial modulation. The review further surveys synthetic methodologies, including bottom‐up and top‐down approaches, in situ formation, and stabilization strategies, with consideration of scalability. A comprehensive assessment of reported electrocatalytic systems is then presented, covering noble metal clusters, transition‐metal clusters, bimetallic systems, heterostructures, and nitrogenase‐inspired architectures, followed by a critical comparison of their performance metrics. Finally, we address key challenges and ongoing controversies, including issues of false‐positive detection, lack of standardized evaluation protocols, catalyst durability, and intrinsically low current densities, thereby outlining future directions for the development of efficient and reliable cluster‐based NRR catalysts.

2. Fundamentals of N₂ Activation and Nitrogen Reduction Reaction Mechanisms

2.1. Electronic Structure of N₂

The intrinsic inertness of molecular nitrogen arises from its highly stable electronic configuration, characterized by a fully occupied bonding manifold and energetically inaccessible antibonding π* orbitals. As a consequence, the activation of N₂ in electrocatalytic systems fundamentally requires electron injection into these π* orbitals, thereby weakening the N≡N bond. This process is classically described by the Dewar–Chatt–Duncanson model, in which σ‐donation from N₂ to the metal is complemented by π‐backdonation from filled metal d orbitals into N₂ antibonding orbitals. However, on extended metal surfaces, this interaction is often insufficient to induce significant bond polarization under ambient conditions, thereby necessitating alternative electronic environments. Metal clusters introduce a distinct paradigm because their discrete electronic states, low coordination numbers, and intermetallic charge redistribution collectively enhance orbital overlap with N₂. In particular, atomically precise Au₄Pt₂ clusters supported on defective graphene were shown to activate N₂ via a confined interfacial electronic structure in which Pt atoms enrich the local electron density, facilitating π‐backdonation and inducing measurable N—N bond elongation [21]. Complementary theoretical screening of M₃ clusters anchored on N‐doped graphene further demonstrated that clusters, such as Ni₃, Ru₃, and Ir₃, possess optimal frontier orbital alignment with N₂, enabling simultaneous σ‐donation and π‐backdonation through a multicenter “push–pull” mechanism that is absent in single‐atom systems [22]. Moreover, compositional modulation within clusters can amplify this effect. For example, the influence of cluster composition on N₂ activation has been demonstrated in several theoretical studies of heterometallic clusters [23]. In these systems, electropositive promoter atoms modify the electronic distribution of the transition‐metal core, increasing electron density available for π‐backdonation into N₂ antibonding orbitals. Such charge redistribution weakens the N≡N bond, increases N—N bond elongation, and lowers the energetic requirements for subsequent hydrogenation steps. The significance of these studies lies in establishing that N₂ activation can be regulated not only through metal identity but also through deliberate electronic polarization within multiatom clusters. Nevertheless, most conclusions are derived from idealized computational models, and experimental verification of the proposed charge transfer mechanisms under realistic electrochemical conditions remains limited. These findings collectively establish that N₂ activation on metal clusters is not governed solely by adsorption strength, but rather by cooperative electronic coupling across multiple metal centers, which enables efficient population of antibonding orbitals and initiates bond weakening.

2.2. Reaction Pathways

Following adsorption, the reduction of N₂ proceeds through PCET steps that define three principal mechanistic pathways: distal, alternating, and enzymatic (Figure 1). The distal pathway involves sequential hydrogenation of one nitrogen atom to release NH₃ prior to activation of the second nitrogen, whereas the alternating pathway proceeds via interleaved hydrogenation of both nitrogen atoms. In contrast, the enzymatic pathway involves side‐on adsorption and simultaneous multisite hydrogenation, analogous to the mechanism of nitrogenase enzymes. The selection of these pathways is strongly dependent on adsorption geometry and electronic structure. Computational studies have shown that end‐on adsorption typically favors distal or alternating mechanisms due to asymmetric charge distribution, whereas side‐on adsorption lowers the barrier for enzymatic‐type hydrogenation [24]. Metal clusters are uniquely capable of stabilizing multiple adsorption geometries due to their heterogeneous active sites and flexible electronic structure. For instance, TM₃ clusters confined on porous BN were found to provide spatially adjacent active centers that stabilize *NNH and *NHNH intermediates through cooperative interactions, thereby lowering the overall reaction barriers compared to isolated sites [25]. Experimental evidence further supports this mechanistic versatility. In Au₄Pt₂ cluster systems, the confined cluster–support interface was shown to stabilize activated N₂ and its hydrogenated intermediates, suggesting that pathway selection can be governed by interfacial electronic confinement rather than by intrinsic metal identity alone [11]. Similarly, theoretical investigations on trimeric clusters have revealed that the presence of multiple metal atoms allows redistribution of electron density during PCET steps, enabling more favorable energetics for intermediate formation along alternating or enzymatic pathways [23]. Thus, metal clusters enable dynamic pathway modulation by integrating geometric flexibility with electronic cooperativity.

FIGURE 1.

Schematic illustration of different fundamental properties of metal clusters on carbon‐based supports for NRR.

2.3. Key Descriptors for Nitrogen Reduction Reaction Activity

The identification of reliable descriptors is essential for rationalizing and predicting NRR activity. Among these, the free energy of the first hydrogenation step (*N₂ → *NNH) is widely regarded as a critical descriptor because it represents the coupling between N₂ activation and proton transfer. However, this single descriptor is insufficient for cluster‐based systems, where multiple active sites and nonuniform electronic structures introduce additional complexity. Classical studies on transition‐metal surfaces established that adsorption–energy scaling relationships impose intrinsic limitations on catalytic performance, particularly due to correlated binding energies of *NNH and *NH₂ intermediates [26]. Metal clusters, however, can partially circumvent these constraints. In M₃–NG systems, Ni₃ and Ru₃ clusters were predicted to exhibit low limiting potentials because electron density is distributed across the cluster, enabling independent optimization of N₂ activation and intermediate stabilization [12]. Similarly, studies on trimeric and tetrameric clusters supported on defective carbon substrates demonstrated that cluster nuclearity and local coordination significantly influence adsorption energetics, thereby altering the conventional scaling relations [27]. Beyond adsorption energies, additional descriptors, such as charge transfer to N₂, d‐band or frontier orbital alignment, and N—N bond elongation, must be considered. For instance, increased Bader charge accumulation on adsorbed N₂ correlates strongly with reduced activation barriers, highlighting the importance of electron‐rich active sites [13]. Furthermore, the relative free‐energy difference between *NNH and *H adsorption provides a quantitative measure of NRR selectivity over HER, emphasizing that multiparameter descriptor frameworks are necessary for accurately capturing catalytic behavior in cluster systems.

While adsorption energies and limiting potentials remain valuable screening metrics, recent advances indicate that static thermodynamic descriptors alone are insufficient to capture the catalytic behavior of electrochemical NRR under realistic operating conditions. In particular, constant‐potential DFT, grand‐canonical simulations, and potential‐dependent ab initio molecular dynamics (AIMD) have demonstrated that the electronic structure of active sites evolves continuously with applied bias and interfacial charging. As a consequence, descriptor frameworks are shifting from static adsorption energies toward dynamic quantities that explicitly account for electrode potential, solvent fluctuations, and catalyst restructuring [32]. Emerging descriptors include potential‐dependent charge transfer to adsorbed N₂, interfacial electric‐field strength, hydrogen‐bond network stabilization of *NNH, free‐energy‐span descriptors under applied potential, and potential‐resolved adsorption energies ΔG(U). These quantities capture phenomena that are absent in conventional computational hydrogen electrode (CHE) treatments, including field‐induced stabilization of charged intermediates and dynamic solvent‐assisted proton transfer. Recent studies further reveal that scaling relationships derived from static surface calculations can change substantially when explicit potential and solvent effects are included. Grand‐canonical simulations incorporating both chemical and electrochemical elementary steps have generated revised NRR activity volcano relationships, demonstrating that catalyst rankings may differ from those predicted by conventional CHE analysis [22]. Likewise, explicit‐solvent calculations show that hydrogen‐bond rearrangement and electrolyte polarization can significantly lower the free‐energy barrier for the first hydrogenation step (*N₂ → *NNH), thereby altering the identification of the rate‐determining step. Rather than proceeding through idealized proton transfer events, the formation of key intermediates, such as *NNH, is often governed by transient solvent networks that dynamically shuttle protons from the electrolyte to the catalyst surface. Consequently, the activation barrier and even the identity of the rate‐determining step may differ substantially from predictions based on static CHE calculations. These findings have stimulated the development of dynamic descriptors that explicitly incorporate interfacial behavior. Examples include potential‐dependent charge transfer to adsorbed N₂, electric‐field‐induced stabilization energies, solvent reorganization free energies, hydrogen‐bond coordination numbers surrounding reactive intermediates, and proton transfer probability distributions obtained from molecular dynamics trajectories. Such descriptors provide a more realistic representation of catalytic performance because they simultaneously account for electrode potential, solvent fluctuations, and catalyst restructuring. For cluster‐based catalysts, where electronic states and coordination environments are particularly sensitive to interfacial charging, incorporation of these dynamic descriptors is expected to significantly improve the predictive power of theoretical screening and mechanistic analysis [33].

Beyond conventional thermodynamic descriptors, kinetic descriptors are increasingly recognized as essential for evaluating NRR performance. These include activation barriers for elementary protonation steps, degree‐of‐rate‐control coefficients, turnover frequency‐determining intermediates, and potential‐dependent reaction fluxes obtained from microkinetic simulations. Unlike adsorption energies, which describe thermodynamic feasibility, kinetic descriptors directly quantify the rates of competing reaction pathways and therefore provide a more rigorous basis for predicting catalytic selectivity. For cluster‐based catalysts, where multiple active sites and adsorption configurations may coexist, microkinetic analysis is particularly valuable because it can capture the dynamic interplay among parallel reaction channels, surface coverages, and competitive HER processes. Hence, future descriptor development for cluster‐based NRR catalysts should integrate potential‐dependent electronic structure, interfacial solvation dynamics, catalyst restructuring effects, and kinetic parameters to provide a more realistic description of catalytic activity and selectivity under operating conditions [34].

2.4. Competing Hydrogen Evolution Reaction

The HER represents the primary competing process in aqueous NRR due to its favorable kinetics and lower activation barriers. The competition arises because protons are more readily reduced than N₂ is activated, particularly on catalysts that strongly bind hydrogen. Consequently, the suppression of HER while maintaining efficient N₂ activation is a central challenge in electrocatalyst design. Mechanistic studies have shown that HER/NRR selectivity is governed by the relative adsorption free energies of *H and nitrogenated intermediates, as well as by their coadsorption dynamics under operating potentials [28]. In cluster‐based systems, this competition can be modulated through electronic and geometric effects. For example, Au hollow nanocages exhibited enhanced ammonia selectivity compared to bulk Au, attributed to confinement‐induced modulation of surface energetics that disfavors hydrogen adsorption [29]. Additionally, modifying the electronic structure of active sites can further suppress HER. In Ru‐based systems, the incorporation of oxophilic components was shown to weaken *H adsorption while preserving N₂ activation capability, thereby improving Faradaic efficiency for ammonia synthesis [30]. Although these studies span both single‐atom and cluster regimes, they converge on a common principle: Selective NRR requires a delicate balance in which *NNH formation is energetically favored over *H adsorption. Metal clusters, with their tunable electronic heterogeneity, provide a platform to achieve this balance by decoupling adsorption sites for hydrogen and nitrogen species.

While thermodynamic descriptors, such as ΔG(*NNH) and ΔG(*H), provide useful indicators of NRR–HER competition, catalytic selectivity is ultimately governed by reaction kinetics. In practice, the pathway exhibiting the lowest free‐energy barrier is not always the dominant route under electrochemical conditions because reaction rates depend on activation barriers, surface coverages, proton availability, and potential‐dependent elementary‐step kinetics. Consequently, increasing attention has been directed toward microkinetic modeling approaches that explicitly account for the rates of individual proton‐coupled electron transfer steps. Within a microkinetic framework, NRR selectivity is determined by the relative rates of nitrogen reduction and hydrogen evolution rather than solely by adsorption energetics. Rate constants for competing elementary reactions can be evaluated using transition‐state theory or potential‐dependent kinetic models, allowing calculation of surface coverages, turnover frequencies, and current distributions under realistic operating conditions. Such analyses frequently reveal that catalysts predicted to be favorable from thermodynamic considerations alone may exhibit poor Faradaic efficiency because HER remains kinetically dominant at practical electrode potentials. Recent studies have therefore emphasized the importance of combining thermodynamic descriptors with kinetic metrics, including activation barriers for *NNH formation, potential‐dependent rate constants, and reaction flux distributions. These kinetic descriptors provide a more realistic assessment of catalytic selectivity and offer valuable guidance for designing cluster catalysts capable of suppressing HER while maintaining efficient ammonia synthesis [35].

2.5. Role of Electrolyte

The electrolyte environment plays a decisive role in governing NRR kinetics, selectivity, and catalyst stability. Beyond serving as an ionic conductor, the electrolyte determines proton availability, cation distribution, interfacial electric fields, and N₂ solubility, all of which directly influence reaction pathways. Cation effects are particularly significant. Experimental studies on Bi nanocrystals demonstrated that K⁺ ions accumulate at the electrode interface, stabilizing negatively charged intermediates such as *NNH through electrostatic interactions and thereby enhancing NRR activity [31]. This cation‐induced stabilization is especially relevant for cluster catalysts, where localized charge density can be strongly influenced by the surrounding ionic environment. Furthermore, electrolyte composition can modulate the balance between NRR and HER. Low proton‐activity environments, such as nonaqueous or ionic liquid electrolytes, suppress hydrogen evolution while increasing N₂ solubility, thereby improving ammonia selectivity. In lithium‐mediated systems, the formation of a solid–electrolyte interphase was shown to regulate species transport and stabilize reactive intermediates, effectively becoming an integral part of the catalytic interface [36]. Importantly, the electrolyte can also influence the structural stability of metal clusters. Due to their small size and high surface energy, clusters are susceptible to reconstruction or oxidation under electrochemical conditions. Operando studies indicate that electrolyte composition can alter cluster oxidation states and coordination environments, thereby affecting catalytic performance. Consequently, optimal NRR performance requires a synergistic design of both catalyst and electrolyte, where proton activity, cation identity, and interfacial structure are cooptimized to stabilize N₂‐derived intermediates while suppressing competing reactions.

Recent advances have substantially revised the conventional understanding of alkali metal cation effects in electrochemical NRR. Traditionally, cations were viewed primarily as electrostatic stabilizers of negatively charged intermediates such as *NNH. However, emerging theoretical and experimental evidence indicates that alkali cations actively participate in the activation of molecular nitrogen through localized electric‐field effects within the electrochemical double layer. Under operating potentials, partially dehydrated cations accumulate near the catalyst surface and generate strong local electric fields that polarize adsorbed N₂ molecules, facilitate electron transfer into antibonding orbitals, and reduce the activation barrier associated with the first proton‐coupled electron transfer step. Recent studies further demonstrate that the magnitude of this effect depends on cation identity, hydration structure, and interfacial organization. Larger alkali cations with lower hydration energies can approach the electrode surface more closely, producing stronger local field effects and more efficient stabilization of activated nitrogen species. Consequently, cation promotion should not be viewed solely as a thermodynamic stabilization phenomenon but rather as a dynamic electrostatic mechanism that continuously influences adsorption geometry, charge transfer, and reaction kinetics. This emerging perspective transforms the electrochemical double layer from a passive ionic environment into an active component of the catalytic interface that directly contributes to nitrogen activation [37].

An important development in recent NRR theory is the transition from implicit electrolyte approximations toward explicit descriptions of the electrochemical interface. While continuum solvation models remain useful for estimating bulk electrostatic effects, they neglect the atomistic organization of water molecules and therefore cannot accurately describe the explicit hydrogen‐bonding networks that develop near catalytically active sites. Potential‐dependent molecular dynamics and constant‐potential AIMD simulations have shown that solvent molecules are not passive spectators but active participants in proton‐coupled electron transfer processes. Explicit water networks dynamically reorganize around adsorbed N₂ and hydrogenated intermediates, creating transient hydrogen‐bond structures that influence reaction energetics and selectivity. Proton transfer frequently proceeds through dynamically evolving hydrogen‐bond networks rather than through isolated protonation events, with water molecules functioning as transient proton relays that continuously reorganize in response to electrode potential, local electric fields, and adsorbate configuration. As a consequence, solvent fluctuations may substantially alter reaction barriers, stabilize charged intermediates, and influence pathway selection [38]. In several reported systems, explicit inclusion of water molecules significantly lowered the calculated free‐energy barrier for *NNH formation compared with implicit‐solvent predictions, bringing theoretical results into closer agreement with experimental observations. These findings are particularly relevant for metal cluster catalysts because their highly localized charge distributions generate strong interactions with surrounding solvent molecules and are highly sensitive to interfacial charging. Cluster active sites often undergo reversible structural reconstruction under operating potentials, accompanied by potential‐dependent changes in oxidation state, coordination environment, solvent orientation, and ion accumulation. The resulting hydrogen‐bond networks become an integral component of the catalytic interface and can influence both nitrogen activation and competing hydrogen evolution processes. Consequently, the catalyst–electrolyte interface becomes highly dynamic, and realistic modeling of cluster‐based NRR systems should combine explicit solvation, electrode‐potential control, and dynamic structural evolution. Such an approach provides a direct bridge between theoretical predictions and operando spectroscopic observations, enabling more reliable identification of catalytically relevant active states [16].

Beyond cation‐specific effects, growing evidence indicates that the electrochemical double layer should be considered a dynamic and potential‐dependent reaction medium. Molecular simulations reveal that electrolyte ions, solvent molecules, and adsorbed intermediates continuously reorganize under applied bias, generating transient electric fields and hydrogen‐bond networks that influence reaction pathways in real time. Such behavior is particularly important for cluster‐based catalysts because their electronic structure is highly sensitive to local charge redistribution. As a result, catalyst performance cannot be fully understood using static adsorption models alone [18]. Instead, realistic descriptions increasingly require explicit consideration of electric‐field effects, electrolyte fluctuations, and potential‐dependent interfacial dynamics. This dynamic interpretation of the double layer provides a conceptual bridge between electrolyte engineering and catalyst design. By controlling cation identity, electrolyte composition, and interfacial structure, it becomes possible to manipulate local electric fields and selectively stabilize nitrogen activation pathways while suppressing competing hydrogen evolution. Therefore, future theoretical investigations of cluster‐based NRR systems should increasingly employ explicit solvation methodologies, including AIMD, constant‐potential simulations, and hybrid explicit/implicit approaches, to achieve realistic descriptions of reaction energetics and interfacial behavior [23]. These emerging concepts are expected to become increasingly important in the rational development of next‐generation cluster‐based NRR electrocatalysts.

3. Metal Clusters: Structural and Electronic Features

3.1. Definition and Classification

Metal clusters are catalytically distinct ensembles containing several to tens of metal atoms, positioned between isolated single atoms and nanoparticles. Their importance for electrocatalytic NRR arises from the coexistence of high atom utilization, multiple adjacent metal centers, and nonbulk electronic structures. Unlike SACs, which generally provide one metal–adsorbate interaction point, clusters can bind N₂ through terminal, bridging, or side‐on configurations, thereby enabling cooperative N–N polarization. Conversely, unlike nanoparticles, clusters retain discrete or semidiscrete electronic states, making their catalytic response highly sensitive to nuclearity, composition, and support coordination. Accordingly, clusters used for N₂ activation may be classified into atomically precise ligand‐derived clusters, support‐confined subnanometer clusters, and computationally designed single‐cluster catalysts. Wang et al. reported Fe–Cu clusters confined in graphitic carbon nitride cavities as “subnanometer reactors,” where spatially neighboring Fe and Cu atoms promoted cooperative N₂ activation and improved electrochemical ammonia formation [39]. In another experimental example, Cu₆(SMPP)₆ nanoclusters supported on graphene oxide were used for NRR, demonstrating that atomically precise copper clusters can provide defined metal nuclearity and stable catalytic motifs under alkaline electrolysis [40]. From the theoretical side, Fe₄/GaS single‐cluster catalysts were proposed as highly efficient N₂‐activation platforms because the Fe₄ unit simultaneously donates electrons into two antibonding π channels of N₂, giving an ultralow calculated overpotential [41]. These studies clarify that cluster classification should not be based only on size; it must also include nuclearity precision, confinement mode, and the number of electronically coupled metal atoms. Although these representative studies collectively demonstrate the promise of metal clusters for NRR, their significance extends beyond simple activity enhancement. The Fe–Cu cluster system highlights the importance of cooperative multimetal activation, while atomically precise Cu₆ clusters illustrate how defined nuclearity enables direct correlation between structure and catalytic behavior. Likewise, the Fe₄/GaS model provides mechanistic insight into how multicenter electron donation can facilitate N₂ activation. However, several limitations remain. Most experimental studies report catalytic performance under different electrolyte compositions, potentials, and ammonia quantification protocols, making direct comparison difficult. Furthermore, many theoretically predicted cluster structures have not yet been synthesized or verified under operando conditions, raising uncertainty regarding their real catalytic identity during NRR. Therefore, future work should focus on correlating atomically resolved structural characterization with standardized catalytic evaluation to establish reliable structure–activity relationships.

3.2. Electronic Structure Evolution

The electronic structure of metal clusters evolves from molecule‐like orbitals to quasimetallic bands as nuclearity increases. This evolution is central to NRR because N₂ activation requires σ‐donation from N₂ to metal states and π‐backdonation from metal d orbitals into N₂ antibonding orbitals. Small clusters can therefore display higher N₂ activation efficiency than larger particles when their frontier orbitals align with the π* orbitals of N₂. Recent studies on electrocatalytic N₂ reduction on metal nanoclusters emphasized that this size‐dependent electronic structure is a key reason clusters can outperform conventional nanoparticles, particularly when their discrete states promote charge transfer into adsorbed N₂ [24]. Complementing these insights, Li et al. employed advanced ab initio methodologies, integrating single‐reference RCCSD(T) with multireference DMRG‐SC‐NEVPT2 treatments, to elucidate the fundamental factors governing dinitrogen activation on gas‐phase metal clusters [42]. Their analysis identifies the N₂ adsorption energy (ΔE _ads) as a decisive descriptor controlling the initial activation step, directly linking adsorption energetics with the efficiency of electron transfer into N₂ antibonding orbitals. This finding provides a quantitative framework that reinforces the importance of frontier orbital alignment discussed above. Notably, ΔE _ads governs the extent of N≡N bond weakening prior to dissociation and enables consistent rationalization of reactivity trends across diverse cluster systems. Guided by this principle, several previously unexplored clusters, including FeV₂S₂ ^–, TaV₂C₂ ^–, and TaV₂C₃ ^–, were predicted and experimentally validated to exhibit high activation capability, highlighting how discrete electronic states and adsorption energetics jointly determine catalytic performance.

Electronic evolution is also strongly composition‐dependent. In Nb₂M trimetallic clusters, systematic DFT analysis showed that the third metal atom tunes the electron density of the Nb₂ core, thereby changing N₂ adsorption strength, N–N elongation, and hydrogenation energetics [43]. Similarly, transition‐metal clusters supported on black phosphorene were reported to activate N₂ through bidirectional donation–backdonation between metal d orbitals and N₂ molecular orbitals; this study directly linked enhanced N₂ capture to electronic coupling between the cluster and the support [44]. In graphdiyne‐supported TM₃ clusters, first‐principles screening further showed that the conductive, π‐conjugated support can regulate cluster charge density and stabilize key *NNH intermediates [45]. Extending the concept of electronic tunability beyond composition, Zhang et al. demonstrated that lattice strain and defect engineering can further modulate the electronic structure and catalytic behavior of transition‐metal systems [46]. Their DFT study on strained Ti₂CO₂ surfaces with oxygen vacancies revealed that biaxial tensile strain shifts the O p‐band center downward and optimizes the Ti d‐band position, thereby enhancing electron donation to adsorbed N₂. At an optimal strain of 2.5%, a Sabatier‐type maximum in activity was observed, governed by the adsorption energetics of *N₂ and NNH intermediates. Orbital‐level analysis confirmed that σ‐donation from N₂ (3σ) to Ti d _z ² states and π‐backdonation from d _xz / d _yz orbitals into N₂ 2π orbitals are central to activation, mirroring the fundamental mechanisms described for clusters. These results emphasize that not only size and composition but also strain and defect‐induced electronic restructuring can control adsorption strength and reaction energetics. Therefore, the electronic structure of clusters should be discussed as a dynamic function of size, composition, and support‐induced charge redistribution rather than as a simple d‐band shift.

The studies discussed above collectively establish that electronic structure engineering is the primary origin of enhanced NRR performance in metal clusters. Their major significance lies in demonstrating that cluster nuclearity and composition can modulate charge transfer to N₂ independently of conventional bulk‐metal electronic descriptors. This offers a potential route to partially overcome adsorption–energy scaling relationships that limit traditional catalysts. Nevertheless, most current interpretations rely heavily on DFT and static electronic descriptors. Experimental verification of charge transfer processes, orbital interactions, and potential‐dependent electronic restructuring remains limited. Consequently, the mechanistic link between calculated electronic properties and experimentally observed catalytic performance is still incomplete, emphasizing the need for operando spectroscopic validation.

3.3. Geometric Configurations

Cluster geometry determines whether N₂ binds end‐on, side‐on, or in a bridge configuration. This geometric factor is mechanistically important because end‐on adsorption generally promotes distal or alternating hydrogenation, whereas side‐on or bridged adsorption can activate both nitrogen atoms simultaneously. In triple‐atom catalysts on graphdiyne/graphene heterostructures, the three adjacent metal atoms created dense active centers with high theoretical metal loading, allowing multisite N₂ adsorption and improved NRR energetics [47]. This work shows that cluster geometry can be used to increase both active‐site density and cooperative activation. Recent studies further highlight that cluster nuclearity and coordination environment play a decisive role in defining these geometric motifs and their catalytic consequences. For example, Teja et al. demonstrated that Ni species supported on γ‐Al₂O₃(110) exhibit strongly nuclearity‐dependent structures and reactivity [48]. While isolated Ni atoms occupy bridge sites coordinated to surface oxygen, Ni₄ clusters adopt distorted tetrahedral geometries that create multiple adjacent metal sites. This geometric transformation enables stronger orbital overlap with N₂ π* orbitals, leading to enhanced activation and a significantly reduced limiting potential (−0.576 V) compared to the single‐atom case (−1.272 V). The results clearly show that cluster formation not only alters electronic structure but also introduces cooperative adsorption geometries that favor more efficient hydrogenation pathways. Geometric confinement further stabilizes catalytically useful configurations. In Fe–Cu clusters confined within graphitic carbon nitride cavities, the support cavity restricted metal migration and maintained short Fe–Cu distances, which strengthened intermetallic cooperation during N₂ adsorption and hydrogenation. In Ni₄ and Ni₁₃ clusters supported on different substrate phases, Yang et al. showed that the same metal cluster can display different NRR activity depending on whether the support phase stabilizes a favorable cluster geometry and charge state [49].

In addition, systematic size‐dependent studies reinforce that cluster geometry evolves with nuclearity and directly impacts catalytic performance. Jin et al. investigated Fe _n clusters (n = 1–10) supported on anatase TiO₂(101), revealing that specific cluster sizes, such as Fe₃, Fe₅, and Fe₈, adopt particularly stable geometries due to strong Fe–O–Ti interfacial bonding [50]. These geometries not only prevent aggregation but also create distinct local coordination environments that govern N₂ adsorption modes and activation strength. Larger clusters exhibit structural rearrangements and enhanced metal–support interactions, which in turn influence charge transfer, magnetic properties, and the energetics of key intermediates. Notably, Fe₅/TiO₂ and Fe₈/TiO₂ show favorable pathways with low barriers for the first hydrogenation step, emphasizing how optimized geometry and interface structure can lower reaction energetics. More recently, triatomic 3d transition‐metal clusters on graphdiyne were screened as NRR catalysts, and the study concluded that cluster planarity, metal–metal distance, and N₂ adsorption geometry collectively determine the limiting potential [51]. Thus, cluster geometry is not a static structural descriptor; it directly controls reaction pathway selection and the energy of the first hydrogenation step.

3.4. Support Interactions

The support is not only a stabilizing matrix; it is an electronic partner that can control cluster charge, geometry, and resistance to aggregation. In NRR, this is especially critical because weakly anchored clusters may reconstruct, dissolve, or agglomerate under applied potential. Reviews on single‐cluster catalysis have emphasized that support choice determines whether atomically precise clusters remain isolated and whether the cluster–support interface contributes directly to catalytic bond activation. For N₂ activation, supports with vacancies, heteroatoms, and π‐conjugated frameworks are especially valuable because they can withdraw or donate charge to the cluster. Experimental evidence further reinforces the decisive role of the support in modulating interfacial charge transfer and catalytic performance. For instance, Shehzad et al. synthesized atomically precise tetrametallic clusters, Cu₄Cl₄(PPh₃)₄ and Ag₄Cl₄(PPh₃)₄, and anchored them onto TiO₂ supports, demonstrating that the oxide interface significantly enhances NRR activity [28]. The Cu‐based system exhibited superior ammonia production rates and Faradaic efficiency, which were attributed to stronger electronic coupling with the TiO₂ surface. This interaction facilitates interfacial charge transfer, improves N₂ adsorption, and increases the availability of active sites. Importantly, comparison with unsupported clusters confirmed that catalytic performance is not solely an intrinsic property of the cluster but is strongly governed by the cluster–support ensemble. Black phosphorene‐supported transition‐metal clusters provide an instructive example: The support participates in stabilizing electron‐rich cluster states that strengthen donation–backdonation with N₂ [52]. Likewise, graphdiyne provides acetylenic pores and π conductivity, enabling stable anchoring of TM₃ clusters while preserving electron transport for PCET steps [53]. An additional consideration for framework‐confined cluster catalysts is the efficiency of electron transport across the support. Although metal–organic frameworks (MOFs) and covalent organic frameworks (COFs) provide highly ordered confinement environments capable of stabilizing small clusters and preventing agglomeration, the insulating nature of many framework structures can limit charge transfer between the electrode and catalytic sites. As a result, the effectiveness of cluster–support electronic coupling may depend not only on local interfacial charge redistribution but also on the ability of the host framework to mediate long‐range electron transport through tunneling or hopping processes. This factor becomes increasingly important when active clusters are embedded within the interior of porous architectures rather than being directly exposed to conductive substrates.

Complementary theoretical studies also demonstrate how support‐induced electronic effects regulate catalytic activity. Pei et al. systematically investigated TM₃ clusters anchored on graphitic carbon nitride (TM₃@g‐C₃N₄) and identified several highly active systems, such as Rh₃, V₃, and Ru₃, with low limiting potentials and efficient enzymatic pathways [54]. Their analysis revealed a clear correlation between catalytic activity and the position of the d‐band center, which is tuned by interfacial charge transfer between the cluster and the support. This charge redistribution governs adsorption strength, intermediate stabilization, and overall reaction energetics, further confirming that the support acts as an active electronic component rather than a passive scaffold. Beyond static stabilization, recent work on reversible metal cluster formation on nitrogen‐doped carbon showed that metal single atoms and clusters can interconvert under electrochemical bias, highlighting that cluster identity under operating conditions may differ from the as‐synthesized structure. Therefore, rigorous NRR studies must connect ex situ cluster structure with operando stability, because the real catalytic unit may be an electrochemically reconstructed cluster–support ensemble. A growing body of operando evidence indicates that this structural evolution is governed by potential‐induced aggregation/redispersion dynamics. Under cathodic polarization, changes in metal oxidation state and surface charge can significantly alter metal–support interaction strengths, thereby modifying the thermodynamic stability of isolated atoms and small clusters. When metal–support binding becomes weaker than metal–metal interactions, mobile atomic species may migrate across the support surface and aggregate into larger clusters through diffusion‐controlled coalescence. Conversely, under less reducing potentials or during potential cycling, oxidation and support‐mediated coordination can promote fragmentation of larger clusters, leading to partial or complete redispersion into smaller clusters or isolated metal sites. For cluster‐based NRR catalysts, these processes are particularly important because aggregation and redispersion directly affect active‐site nuclearity, electronic structure, and adsorption behavior. The resulting catalytic interface therefore exists as a dynamic population of interconverting species rather than a single static structure. Recent operando X‐ray absorption spectroscopy, environmental transmission electron microscopy, and electrochemical microscopy studies have revealed that such transformations can occur reversibly during electrocatalysis, with the equilibrium distribution strongly influenced by electrode potential, electrolyte composition, defect density, and support coordination environment. Importantly, potential‐induced restructuring may either enhance or diminish catalytic activity depending on whether the newly formed species provide more favorable nitrogen activation pathways. Consequently, catalyst durability should not be evaluated solely in terms of resistance to aggregation, but rather through understanding the dynamic equilibrium among isolated atoms, clusters, and larger aggregates under realistic reaction conditions. The significance of support engineering extends beyond preventing cluster agglomeration. The cited studies reveal that supports actively participate in electronic regulation, interfacial charge transfer, and stabilization of reaction intermediates, effectively functioning as part of the catalytic site. However, a common limitation of many reported systems is the lack of direct evidence, demonstrating that the proposed cluster–support interfaces remain unchanged during long‐term electrolysis. Potential‐induced restructuring, atom migration, and support oxidation may alter the nature of the active site and complicate mechanistic interpretation. Therefore, future studies should combine operando X‐ray absorption spectroscopy, environmental transmission electron microscopy, and isotope‐labeled NRR measurements to establish whether the proposed catalytic motifs persist under realistic reaction conditions.

3.5. Comparison with SACs and Nanoparticles

Compared with SACs, metal clusters provide adjacent metal centers that can cooperatively bind and polarize N₂ (Figure 2). This distinction is fundamental because NRR involves multielectron, multiproton chemistry, and the stabilization of *N₂, *NNH, *NHNH, and *NH₂ intermediates may require more than one metal atom. Recent studies on atomically dispersed catalysts provide an important contrast that highlights the unique advantages of clusters. For example, Kong et al. demonstrated that SACs on Ti₃C₂O₂ MXene can exhibit high catalytic performance with optimized adsorption energetics and suppressed side reactions when atomic dispersion is preserved [27]. However, aggregation into small clusters led to reduced activity and selectivity, illustrating that nuclearity strongly influences catalytic performance and reaction pathways. At the same time, cooperative multisite effects, absent in isolated SACs, have been shown to significantly enhance NRR performance. Fan et al. reported a heteronuclear FeMo dual‐site catalyst that achieved an NH₃ production rate of 1.07 mg h^–1 mg_metal ^–1 and a Faradaic efficiency of 21.7% at −0.25 V versus reversible hydrogen electrode (RHE), substantially outperforming corresponding single‐atom systems [10]. This improved performance originates from synergistic Fe–Mo interactions that enhance N₂ adsorption, promote π‐backdonation, and lower the energy barrier for the first hydrogenation step. Studies on structural evolution of single atoms and clusters further show that the boundary between SACs and clusters is dynamic; under reaction conditions, single atoms may aggregate into clusters, while clusters may redisperse, making operando identification essential [55]. Supporting this, Alemany‐Molina et al. showed that catalytic performance depends on the balance between single atoms and few‐atom clusters stabilized on nitrogen‐rich porous carbon, where Cu‐based systems exhibited superior activity, selectivity, and durability due to favorable electronic structure and accessible active sites [56].

FIGURE 2.

Summary comparison of various properties for cluster‐based NRR electrocatalysts.

Compared with nanoparticles, clusters offer higher atomic utilization and more uniform active motifs. Nanoparticles often expose many poorly defined facets that can favor HER, whereas clusters can be engineered to provide specific N₂ binding ensembles. The Nature Methods Primer on nitrogen activation also stressed that genuine N₂ reduction requires strict isotope and contamination control, which is especially important for cluster systems because high apparent activity can be misleading if ammonia contamination is not excluded [57]. Finally, recent studies on cocatalytic metal–support interactions in atomically dispersed catalysts provide a useful conceptual bridge: The active site may not be the metal alone but a coupled metal–support unit, a principle even more relevant for clusters because several metal atoms and nearby support atoms can jointly stabilize intermediates [58]. Therefore, clusters should not be treated as “larger SACs” or “smaller nanoparticles”; they represent a distinct catalytic class where nuclearity, intermetallic coupling, support participation, and reconstruction collectively define NRR performance in terms of activity, selectivity, and stability.

4. Design Principles for Metal Cluster‐based NRR Catalysts

4.1. Size Effects

The electrocatalytic reduction of N₂ on metal clusters is dictated by a subtle interplay between quantum‐confined electronic structure and cooperative multiatom reactivity, which fundamentally distinguishes these systems from extended crystalline surfaces. In contrast to periodic lattices, subnanometric clusters present discrete electronic states and nonuniform coordination environments, leading to size‐specific catalytic behavior that evolves in a nonmonotonic manner. This phenomenon has been rigorously demonstrated in atomically precise clusters, where modulation of nuclearity directly perturbs frontier orbital alignment and, consequently, the efficiency of π‐backdonation into the antibonding orbitals of N₂. Fundamental gas‐phase studies further reinforce the critical role of nuclearity and electronic structure in governing nitrogen‐related bond activation. For example, Chuhan et al. investigated ammonia activation on 3d transition‐metal clusters using laser ablation and reflectron time‐of‐flight mass spectrometry, revealing strong size‐ and metal‐dependent reactivity trends [7]. Early transition metals, such as Ti and V, exhibited efficient nitride formation, while Ni _n ⁺ clusters (n ≥ 3) demonstrated complete NH₃ dehydrogenation accompanied by H₂ formation, directly evidencing cooperative multiatom reactivity. These results highlight how increasing nuclearity enables multistep bond activation via electron redistribution and cooperative adsorption, providing a mechanistic basis that parallels N₂ activation processes in electrocatalysis.

Recent spectroscopic and computational investigations on cobalt and iron clusters anchored on nitrogen‐doped carbon substrates revealed that clusters containing three to five metal centers maximize charge polarization at the adsorption interface, thereby stabilizing the *NNH intermediate without excessively strengthening *H adsorption, a critical requirement for suppressing the competing HER [59]. Building on this foundation, it has become evident that multiatom ensembles enable reaction pathways inaccessible to isolated single‐atom catalysts. In particular, operando studies on Fe₂ and Fe₃ moieties embedded within carbon matrices demonstrated a marked reduction in the activation barrier for the first proton–electron transfer step, attributed to synergistic orbital overlap between adjacent Fe centers that facilitates end‐on to side‐on N₂ reorientation [60]. This cooperative mechanism is further corroborated by DFT analyses of Mn and Cr trimers supported on graphitic frameworks, where dual‐site adsorption geometries enable simultaneous weakening of the N≡N bond and stabilization of protonated intermediates, effectively circumventing conventional scaling limitations [61]. Consequently, triatomic clusters have emerged as a structurally minimal yet electronically sufficient motif for promoting efficient nitrogen activation. In parallel, support effects have been shown to critically modulate cluster reactivity by mediating charge transfer and stabilizing specific nuclearities. For instance, ruthenium clusters dispersed on reducible oxides, such as TiO₂ and CeO₂, exhibit dynamic electron exchange with oxygen vacancies, leading to an electron‐rich metal center that enhances N₂ adsorption while attenuating hydrogen binding [62]. Similarly, molybdenum clusters confined within porous carbon nitride frameworks demonstrate enhanced catalytic turnover due to the presence of multiple anchoring sites that prevent aggregation while simultaneously enabling diverse adsorption configurations for nitrogen and its hydrogenated intermediates [63]. These findings underscore the necessity of considering both cluster nuclearity and support‐induced electronic perturbations as codeterminants of catalytic performance.

Furthermore, the role of geometric heterogeneity within clusters has been elucidated through advanced microscopy and theoretical modeling. Studies on subnanometer nickel and palladium clusters reveal that low‐coordination edge and corner atoms dominate catalytic activity, as these sites possess elevated d‐band centers that facilitate stronger interaction with N₂ while maintaining moderate binding energies for reaction intermediates [64]. Importantly, this distribution of active sites evolves with cluster size, leading to an optimal regime in which the proportion of undercoordinated atoms is maximized without compromising structural integrity. Indeed, excessively small clusters, although highly active, tend to undergo sintering under electrochemical conditions, whereas slightly larger clusters achieve a balance between reactivity and durability [65]. Beyond support‐mediated effects, cluster nuclearity itself provides an independent design parameter for regulating catalytic behavior. Variations in cluster size directly influence orbital occupancy, charge delocalization, and the distribution of low‐coordination active sites, thereby determining the balance between nitrogen activation, intermediate stabilization, and catalyst durability.

Complementary investigations on Ir‐based clusters have shown that precise control over cluster size (n = 4–7) enables fine‐tuning of the d‐band center, thereby achieving an optimal balance between N₂ activation and intermediate desorption kinetics [66, 67]. While numerous studies report superior activity for specific cluster sizes, their broader significance lies in revealing that catalytic performance is governed by a balance between electronic confinement and multisite cooperativity rather than by minimizing cluster size alone. Importantly, several investigations identify optimal nuclearities in the few‐atom regime, suggesting that maximum activity emerges from a compromise between N₂ activation strength, intermediate stabilization, and structural stability. However, the majority of reported size effects are derived from model calculations or ex situ characterization. Because cluster nuclearity may evolve during electrocatalysis, the actual active species can differ substantially from the initially synthesized structure. This remains a major challenge for establishing universally applicable size–activity relationships. Hence, these advances establish that the catalytic efficiency of metal clusters in nitrogen electroreduction is governed not by minimal size alone, but by a convergence of electronic confinement, multisite cooperativity, and structural robustness (Figure 3). The identification of optimal nuclearities, typically within the few‐atom regime, reflects a delicate balance wherein sufficient electronic delocalization coexists with a high density of low‐coordination active sites. Importantly, the role of cluster size should be distinguished from composition‐ and interface‐driven effects discussed in subsequent sections. Future progress will therefore depend on precise control of cluster nuclearity while integrating complementary strategies involving composition tuning, coordination engineering, support design, and interfacial modulation to optimize the reaction landscape for sustainable ammonia synthesis.

FIGURE 3.

Schematic illustration of different design principles for cluster‐based NRR electrocatalysts.

4.2. Composition Tuning

The incorporation of compositional heterogeneity within subnanometric clusters provides a decisive handle for disentangling the energetics of N₂ activation from subsequent PCET steps. Unlike monometallic systems, heterometallic ensembles inherently generate asymmetric charge distributions, enabling spatial separation of functional roles across distinct atomic sites. Recent theoretical investigations further substantiate the critical role of heterometallic composition in promoting efficient N₂ activation. For example, Huang et al. systematically examined molybdenum‐based clusters incorporating alkali and alkaline‐earth promoters, demonstrating that pristine Mo₂ exhibits negligible N₂ activation, whereas heterometallic Mo₂M (M = Li, Na, K, Mg, Ca) clusters significantly enhance adsorption and activation through charge redistribution [3]. In particular, Mo₂Ca and Ca‐rich clusters show substantially improved activation characteristics owing to strong electron donation from electropositive dopants into N₂ antibonding orbitals. The resulting bond elongation, vibrational softening, and reduced bond order facilitate subsequent hydrogenation processes. Notably, cooperative adsorption involving both end‐on and side‐on interactions emerges as an effective mechanism for destabilizing the N≡N bond, highlighting the importance of multimetal synergy in catalytic design. More broadly, heterometallic clusters enable electronic complementarity among constituent elements, allowing distinct atomic sites to perform different catalytic functions. Electron‐rich centers can preferentially participate in π‐backdonation to activate N₂, whereas neighboring sites stabilize partially hydrogenated intermediates through moderated adsorption strengths. This functional differentiation provides a powerful strategy for improving catalytic selectivity and reducing kinetic barriers associated with the first protonation step.

Expanding on this paradigm, alloying‐induced charge redistribution has been systematically explored in late‐transition‐metal clusters. DFT studies reveal that electronegativity differences between constituent metals induce directional electron transfer within the cluster, thereby optimizing adsorption configurations of key intermediates and lowering free‐energy barriers for *NNH formation while simultaneously suppressing excessive hydrogen adsorption [68]. Importantly, this polarization effect is highly sensitive to atomic arrangement, indicating that catalytic behavior depends not only on composition but also on site‐specific ordering within the cluster. Furthermore, heterometallic clusters provide opportunities to mimic the functional specialization observed in biological catalytic systems. In such architectures, one metal center can preferentially adsorb and activate N₂, whereas adjacent centers facilitate sequential protonation and electron transfer steps. This cooperative division of labor promotes stabilization of reactive intermediates and improves overall reaction kinetics. In parallel, ligand and coordination environment engineering has emerged as an additional dimension of compositional control. Atomically precise copper‐based clusters functionalized with electron‐donating ligands exhibit modified electronic density at the metal core, selectively weakening hydrogen adsorption without compromising N₂ activation capability [69]. This observation highlights that compositional tuning extends beyond metallic constituents and can include nonmetallic components that regulate catalytic selectivity through electronic effects.

Recent computational studies further suggest that increasing compositional complexity may provide new opportunities for overcoming intrinsic scaling relationships. The presence of multiple electronically distinct centers creates the possibility of independently modulating adsorption energies of *N₂, *NNH, and *NH₂ intermediates, thereby decoupling elementary reaction steps that are typically correlated on monometallic surfaces [69]. Such behavior offers a promising framework for rational catalyst design aimed at maximizing both activity and selectivity. Equally important, compositional tuning can be employed to regulate competition between nitrogen reduction and hydrogen evolution. By adjusting electronic structure and d‐band characteristics, heterometallic clusters can selectively weaken hydrogen adsorption while preserving sufficient interaction with activated nitrogen species [70]. This capability is particularly important under aqueous electrochemical conditions, where HER frequently dominates catalytic behavior. Hence, these findings demonstrate that compositional engineering of metal clusters enables precise control over electronic structure and adsorption energetics through charge polarization, orbital hybridization, and cooperative interactions between constituent elements. The resulting functional differentiation of atomic sites provides a robust strategy for enhancing catalytic activity and selectivity. The principal significance of heterometallic cluster design is its ability to create functionally differentiated active sites that independently regulate N₂ adsorption, intermediate stabilization, and hydrogen evolution suppression. This strategy provides one of the most promising routes for breaking conventional scaling relationships that constrain monometallic catalysts. Nevertheless, increasing compositional complexity also introduces challenges, including synthetic reproducibility, compositional heterogeneity, and uncertainty regarding the exact atomic arrangement under reaction conditions. In many cases, theoretical models assume idealized atomic distributions that may not accurately represent experimentally synthesized catalysts. Therefore, future efforts should prioritize atomically precise synthesis and operando characterization to verify the proposed synergistic mechanisms.

4.3. Coordination Environment

The local coordination environment of metal clusters constitutes a decisive structural parameter that governs electron density distribution, adsorption configuration of dinitrogen, and the stabilization of key hydrogenated intermediates, thereby dictating the energetics of the potential‐determining *N₂ → *NNH transformation. In contrast to crystalline surfaces, cluster‐based systems can be immobilized within heteroatom‐functionalized matrices, which impose well‐defined M–N, M–C, or M–X (X = O, S, P) coordination geometries and enable deliberate modulation of orbital interactions at the active site. Recent advances in coordination engineering at the atomic scale further highlight how subtle variations in ligand environment directly influence catalytic behavior. For instance, Chen et al. demonstrated that tailoring the coordination structure of Ni single‐atom sites from symmetric Ni–N₄ to an asymmetric Ni–N₁–N₄ configuration significantly enhances N₂ activation [11]. The improved activity originates from symmetry breaking within the coordination environment, which promotes dz² orbital participation near the Fermi level and strengthens p–d hybridization with N₂, thereby lowering the energy barriers associated with key hydrogenation steps. These findings demonstrate that coordination asymmetry can be leveraged to optimize electronic structure and reaction energetics. Such structural precision allows direct tuning of π‐backdonation and σ‐donation processes that are central to N≡N bond activation. A representative illustration arises from dual‐metal sites embedded in nitrogen‐doped carbon frameworks. Advanced spectroscopic and theoretical studies have shown that Fe–Fe and Co–Co dimers coordinated by pyridinic nitrogen exhibit markedly enhanced N₂ adsorption and activation compared with their mononuclear counterparts owing to cooperative electron sharing between adjacent metal centers that stabilizes the *NNH intermediate through a bridging adsorption mode [46]. In a similar vein, Ni‐based dimeric motifs anchored on graphene derivatives redistribute electron density across the metal–metal axis, thereby lowering the free‐energy barrier associated with the first proton‐coupled electron transfer step [71].

Complementary insights from well‐defined gas‐phase cluster models further elucidate the role of coordination and heterometallic synergy in N₂ activation. Wang et al. investigated yttrium–cobalt carbide cluster anions (Y_1–3CoC_1,2 ^–), revealing strong size‐ and composition‐dependent reactivity toward N₂ [12]. The observed activation behavior originates from preferential N₂ binding at electron‐rich metal centers and efficient charge transfer into antibonding orbitals, while carbide ligands further modify the local coordination environment and facilitate multistep activation processes. These findings underscore how heterometallic coordination and ligand effects cooperatively regulate adsorption geometry and activation efficiency at the molecular level. These studies collectively establish that paired coordination environments enable dual‐site activation pathways inaccessible to isolated single‐atom catalysts. Progressing to higher nuclearities, triatomic clusters confined within carbon matrices introduce additional configurational flexibility that directly impacts reaction kinetics. For instance, Fe₃ and Co₃ clusters stabilized through nitrogen coordination can accommodate both end‐on and side‐on N₂ adsorption geometries, facilitating alternative hydrogenation pathways and reducing kinetic bottlenecks associated with conventional scaling relationships [72]. This structural adaptability enables dynamic reorganization of adsorption configurations throughout the catalytic cycle, thereby enhancing catalytic efficiency. The influence of coordination extends beyond nitrogen ligation to include interactions with other heteroatoms and supports. Ruthenium clusters confined within graphitic carbon nitride matrices exemplify this effect, where strong Ru—N bonding stabilizes subnanometric clusters against aggregation while simultaneously promoting charge transfer to adsorbed N₂ molecules [73]. Likewise, molybdenum clusters coordinated within nitrogen‐rich porous frameworks exhibit increased electron density at the metal center, strengthening π‐backdonation and facilitating subsequent hydrogenation steps [74].

Beyond nitrogen coordination, sulfur‐ and oxygen‐containing environments provide additional opportunities for electronic regulation. Sulfur‐coordinated Fe and Co clusters, inspired by biological Fe–S cofactors, stabilize activated nitrogen species through a combination of electron donation and structural flexibility, enabling reaction pathways reminiscent of enzymatic nitrogen fixation [75]. Concurrently, oxide‐supported clusters, such as Fe and Ru species, interfaced with TiO₂ or CeO₂ benefit from metal–oxygen interactions that regulate electron transfer through lattice coupling, thereby tuning adsorption energies of both N₂ and hydrogenated intermediates [50]. Defect engineering further amplifies the influence of coordination by generating anchoring sites that stabilize clusters while preserving optimal metal–metal distances required for cooperative activation. Vacancy‐rich graphene, graphdiyne, and C₂N substrates have been shown to immobilize heterometallic dimers and trimers with asymmetric coordination environments, inducing directional charge polarization and selectively favoring *NNH formation over competing *H adsorption [76]. In particular, heterodimeric systems anchored at defect sites exhibit pronounced electronic asymmetry, enabling site‐specific adsorption and reaction pathways that improve NRR selectivity. Equally important, the coordination environment plays a central role in maintaining catalyst stability under electrochemical conditions. Equally important, the coordination environment governs not only catalytic activity but also long‐term structural stability. Strong coordination interactions between metal clusters and heteroatom‐rich supports can substantially increase metal anchoring energies, suppress surface diffusion, and reduce the probability of cluster aggregation during electrochemical operation. This stabilization effect is particularly important for subnanometer clusters because their high surface energies make them thermodynamically prone to restructuring. Recent operando studies indicate that nitrogen‐coordinated and defect‐engineered supports can preserve cluster dispersion over extended electrolysis periods while simultaneously maintaining favorable electronic structures for N₂ activation. However, catalyst stability remains insufficiently understood because reversible aggregation, redispersion, and support‐induced restructuring may occur under applied potentials. Consequently, future studies should combine operando spectroscopy with long‐duration durability testing to identify the true active structures responsible for sustained NRR performance.

4.4. Electronic Structure Modulation

Precise manipulation of the electronic structure in metal clusters constitutes a central strategy for promoting dinitrogen activation while simultaneously attenuating competitive hydrogen evolution. In contrast to bulk catalysts, subnanometric clusters exhibit highly tunable d‐band centers, discrete frontier orbitals, and pronounced charge polarization, all of which directly govern the efficiency of electron transfer into the antibonding π* orbitals of N₂. Consequently, the alignment between cluster‐derived electronic states and the molecular orbitals of adsorbed nitrogen emerges as a key descriptor for catalytic performance. Recent advances in data‐driven modeling further reinforce the centrality of electronic descriptors in governing N₂ activation. For example, Wang et al. developed a graph neural network framework trained on experimentally characterized clusters, demonstrating that electronic features, such as valence configuration, natural charge, and subgraph charge distribution, serve as decisive predictors of reactivity [26]. The model revealed that clusters possessing more delocalized charge distributions within the metal core generally exhibit enhanced activation capability because they facilitate efficient electron transfer into N₂ antibonding orbitals. In contrast, strongly polarized metal–ligand interfaces tend to reduce activation efficiency, highlighting the importance of balanced electronic environments. These findings provide a quantitative validation of electronic structure principles and establish a framework for accelerated catalyst discovery. Recent theoretical and experimental studies have further demonstrated that optimal d‐band positioning facilitates efficient π‐backdonation without excessively stabilizing reaction intermediates. Strong electronic coupling between clusters and supports can induce significant redistribution of electron density, thereby stabilizing key hydrogenated intermediates while simultaneously weakening hydrogen adsorption. Such behavior is particularly advantageous for achieving high selectivity under ambient electrochemical conditions. These observations highlight the dual role of supports as both structural stabilizers and electronic modulators.

Beyond intrinsic electronic structure, extrinsic charge regulation through dopants or heteroatom incorporation provides an additional mechanism for tuning catalytic activity. Electron‐donating promoters can increase charge density at catalytically active metal centers, thereby strengthening backdonation into N₂ antibonding orbitals and facilitating bond activation [77]. Similarly, heteroatom incorporation within support matrices can generate localized charge polarization at the cluster–support interface, selectively promoting nitrogen adsorption while suppressing competing hydrogen adsorption [78]. This interplay between cluster and support illustrates the importance of interfacial electronic regulation in determining catalytic behavior. External stimuli provide yet another dimension for electronic structure control. Zhang et al. demonstrated that applied electric fields can dynamically regulate N₂ activation on cluster systems [13]. Variations in field strength alter adsorption energetics, intermediate stability, and reaction barriers by modulating electron donation into antibonding orbitals. These observations indicate that electrostatic tuning offers a powerful strategy for manipulating catalytic pathways under realistic operating conditions. Moreover, cluster nuclearity and composition strongly influence electronic configuration and reaction mechanisms. Comparative studies of small transition‐metal clusters reveal that differences in d‐electron count, orbital occupancy, and spin state can produce distinct adsorption geometries and protonation pathways [79]. Certain electronic configurations favor associative mechanisms, whereas others facilitate partial dissociative activation, illustrating the strong dependence of reaction pathways on electronic structure.

Low‐dimensional supports further contribute to electronic regulation by providing extended charge‐transport networks and stabilizing transient intermediates through delocalized electronic interactions [80]. Such systems facilitate rapid electron injection into adsorbed N₂ while promoting efficient desorption of downstream products. More broadly, the electronic environment established at the cluster–support interface plays a decisive role in determining catalytic turnover and selectivity. Importantly, electronic heterogeneity within clusters offers a pathway for overcoming conventional scaling relationships that limit catalytic efficiency on extended surfaces. Nonuniform charge distributions enable independent modulation of adsorption energies for critical reaction intermediates, thereby decoupling elementary reaction steps that are typically correlated on monometallic surfaces [54]. This departure from linear scaling behavior is consistently supported by charge density analyses that correlate enhanced N₂ activation with localized electron accumulation at specific catalytic sites [81]. Hence, these insights establish that electronic structure modulation, achieved through d‐band engineering, charge redistribution, interfacial coupling, dopant regulation, and external‐field control, provides a robust framework for selectively activating nitrogen while suppressing competing hydrogen evolution. By integrating these principles with emerging data‐driven approaches and dynamic electronic tuning strategies, it becomes possible to tailor adsorption energetics and reaction pathways with unprecedented precision.

4.5. Support Engineering

The physicochemical characteristics of the support play a determinative role in governing the stability, dispersion, and electronic structure of metal clusters, thereby directly influencing nitrogen activation pathways under electrochemical conditions. In contrast to inert substrates, advanced supports actively participate in charge redistribution, modulate adsorption energetics, and define the accessibility of N₂ molecules to catalytically relevant sites. Consequently, the cluster–support interface has emerged as a critical locus for tuning catalytic performance in nitrogen reduction systems. Defect‐rich semiconductor supports provide a particularly effective platform for stabilizing highly dispersed metal clusters through strong electronic coupling and interfacial charge redistribution [41]. Such supports create abundant anchoring sites that suppress aggregation while simultaneously generating localized electronic environments capable of promoting N₂ adsorption, facilitating charge transfer, and regulating reaction energetics. These observations highlight the fundamental importance of defect‐mediated cluster–support interactions in controlling catalytic behavior. Carbon‐based architectures, particularly graphene derivatives, have been extensively explored due to their high electrical conductivity and structural versatility. Defect‐engineered and heteroatom‐doped graphene frameworks provide anchoring sites that stabilize subnanometric clusters through strong metal–support interactions while simultaneously facilitating rapid electron transfer to adsorbed N₂ molecules. Nitrogen‐coordinated carbon matrices induce electron‐rich metal centers that enhance π‐backdonation into N₂ antibonding orbitals, thereby lowering activation barriers associated with the initial hydrogenation step [82]. Furthermore, boron‐ and sulfur‐doped carbon supports introduce localized electronic asymmetry that promotes selective adsorption of nitrogen relative to hydrogen species.

Beyond graphene, emerging carbon allotropes, such as graphdiyne and related sp–sp² hybridized networks, offer unique opportunities for cluster immobilization. Their intrinsic porosity and extended π‐conjugation enable both physical confinement and efficient charge transport, thereby stabilizing small metal ensembles while maintaining continuous electron supply during catalysis [83]. Similarly, nitrogen‐rich frameworks, such as C₂N, provide well‐defined cavities that prevent cluster agglomeration and create confined reaction environments capable of influencing adsorption configurations and reaction pathways [84]. Oxide supports introduce an additional dimension of electronic regulation through metal–oxygen interactions and defect chemistry. Reducible oxides, including TiO₂ and CeO₂, effectively mediate charge transfer via oxygen vacancies, dynamically adjusting the oxidation state of supported clusters under reaction conditions. Such interfacial effects can enhance N₂ adsorption while simultaneously weakening hydrogen binding, thereby improving catalytic selectivity [85]. Moreover, strong metal–support interactions contribute significantly to the stabilization of subnanometric clusters against sintering and structural degradation. Two‐dimensional transition‐metal carbides and nitrides (MXenes) have recently emerged as conductive supports with highly tunable surface chemistry. Their metallic conductivity, combined with controllable surface terminations such as –O, –OH, and –F groups, enables precise regulation of interfacial charge transfer and adsorption energetics [86]. This tunability allows optimization of the balance between reactant activation and product desorption while maintaining efficient charge transport throughout the catalytic cycle.

Porous crystalline supports, such as MOFs and COFs, provide additional opportunities for stabilizing atomically precise clusters and preventing aggregation. However, their application in electrocatalytic NRR is often constrained by limited intrinsic electrical conductivity. Consequently, catalytic performance in framework‐confined systems depends not only on active‐site chemistry but also on the efficiency of long‐range charge transport through the host matrix. Recent studies suggest that electron transport within confined framework architectures generally proceeds through a combination of tunneling and hopping mechanisms. In tunneling‐dominated systems, electrons migrate between spatially separated clusters or redox‐active nodes through short‐range quantum‐mechanical coupling [87]. In contrast, hopping transport involves sequential charge transfer among adjacent metal centers, conjugated linkers, or framework defects. The relative contributions of these pathways depend strongly on cluster spacing, linker conjugation, framework topology, and cluster–framework electronic coupling. To address conductivity limitations, increasing attention has been directed toward conductive MOFs, π‐conjugated COFs, donor–acceptor framework architectures, and hybrid materials incorporating graphene, carbon nanotubes, or conductive carbon networks. These strategies improve charge mobility while preserving the confinement effects necessary for stabilizing subnanometric clusters. Consequently, charge‐transport efficiency should be considered alongside cluster nuclearity, electronic structure, and adsorption energetics when designing framework‐confined NRR catalysts [88, 89, 90, 91].

In addition to planar supports, porous crystalline materials, such as zeolites, provide confinement effects that are highly beneficial for cluster stabilization. Their well‐defined microporous environments restrict cluster migration and aggregation while generating local electrostatic fields capable of influencing N₂ adsorption and activation [92]. Other low‐dimensional supports, including boron nitride and black phosphorus, offer complementary electronic properties arising from intrinsic polarity and anisotropic electronic structures. These materials can induce directional charge transfer at the interface and thereby regulate adsorption geometries of nitrogen‐containing intermediates [93]. Importantly, support selection also governs catalyst durability under electrochemical conditions. Operando spectroscopic investigations demonstrate that strong anchoring interactions, particularly those associated with heteroatom coordination and defect sites, effectively inhibit cluster migration and coalescence during prolonged operation [56]. Such stabilization is essential for preserving active‐site integrity and maintaining consistent catalytic performance. Hence, support engineering integrates structural confinement, electronic modulation, and interfacial synergy to regulate cluster stability, charge transfer, adsorption energetics, and reaction selectivity. By tailoring support composition, defect chemistry, conductivity, and confinement characteristics, it becomes possible to simultaneously enhance nitrogen activation, suppress competing reactions, and improve catalyst durability. Representative catalytic performances arising from these support‐engineering strategies are summarized separately in Section 6 to maintain a clear distinction between mechanistic design principles and catalyst benchmarking.

4.6. Interfacial Effects

The catalytic interface formed between the metal cluster, its support, and the electrolyte constitutes the operative reaction zone for electrochemical nitrogen reduction, where electronic coupling, local electric fields, and adsorption dynamics collectively determine activity and selectivity. Unlike isolated active sites, these interfacial regions exhibit emergent properties arising from strong electronic communication across phase boundaries, thereby enabling reaction pathways inaccessible in homogeneous or bulk systems. Recent studies demonstrate that deliberate interfacial engineering can significantly enhance catalytic performance through synergistic coupling between cluster sites and supporting matrices [60]. Interfacial charge redistribution modifies local electronic structure, regulates adsorption energetics, and influences competition between nitrogen reduction and parasitic side reactions. These observations establish the catalyst–support interface as a critical design element for controlling reaction pathways. At the cluster–support boundary, orbital hybridization and charge transfer generate localized electronic environments that facilitate N₂ activation. Atomically precise cluster systems interfaced with conductive supports create confined electronic states capable of enhancing π‐backdonation into N₂ antibonding orbitals while stabilizing partially hydrogenated intermediates through interfacial polarization effects [94]. Similarly, subnanometric transition‐metal clusters embedded within heteroatom‐rich matrices function as spatially confined catalytic domains in which cooperative interactions between metal centers and surrounding frameworks reduce kinetic barriers associated with the formation of key reaction intermediates [95]. These findings highlight the critical role of geometric confinement and electronic communication in amplifying interfacial reactivity.

Strong coupling between clusters and electronically active supports further enhances catalytic performance by enabling efficient charge transfer across the interface. In such systems, the support acts not merely as a structural scaffold but as an active participant in the catalytic cycle, continuously regulating electron density at the active site [96]. Importantly, the magnitude of this coupling can be tuned through surface chemistry, defect engineering, and support composition, providing a versatile strategy for optimizing catalytic behavior. In addition to structural and electronic factors, electrochemical operation introduces dynamic interfacial phenomena that exert a profound influence on reaction kinetics. Under applied potentials, electrolyte cations accumulate near the electrode surface, generating local electric fields that stabilize charged intermediates and lower activation barriers for proton‐coupled electron transfer processes [97]. Simultaneously, solvent reorganization and ion‐specific interactions continuously modify adsorption energetics, emphasizing the importance of the electrochemical double layer in catalyst design. Equally significant is the dynamic nature of cluster‐based interfaces under operating conditions. Operando spectroscopic investigations reveal that subnanometric clusters can undergo reversible structural rearrangements during catalysis, generating transient active configurations that differ from their initial structures [98]. Such restructuring can alter coordination environments, electronic properties, and adsorption geometries, thereby influencing catalytic pathways in real time. Consequently, active sites identified under ex situ conditions may not accurately represent the dominant catalytic species present during operation.

Potential‐induced aggregation and redispersion processes provide a mechanistic explanation for many of these structural transformations. Because electrochemical polarization continuously modifies local charge density and adsorption energetics, the relative stability of different cluster nuclearities may evolve throughout catalyst operation. Understanding such dynamic restructuring behavior is therefore essential for establishing realistic structure–activity relationships and for designing catalysts capable of maintaining favorable active‐site distributions during long‐term electrolysis [6]. Recent computational investigations further demonstrate that explicit hydrogen‐bonding networks constitute an active component of the electrochemical interface. Rather than behaving as a homogeneous dielectric medium, the solvent forms structured hydration environments around both catalysts and adsorbed intermediates. These networks continuously reorganize during catalysis and can facilitate proton transport through Grotthuss‐type relay mechanisms, thereby influencing proton‐coupled electron transfer kinetics [34, 89]. Such effects are particularly important for cluster‐based catalysts because local electric fields and charge polarization generated by small metal ensembles strongly affect solvent orientation and hydrogen‐bond connectivity.

Potential‐dependent interfacial dynamics have also emerged as a critical factor governing catalyst performance. Recent studies indicate that applied potential can dynamically modify charge transfer between clusters and supports, producing transient electronic configurations that are absent under equilibrium conditions [17]. These observations suggest that the active state of cluster‐based NRR catalysts should be regarded as a dynamic electrochemical ensemble rather than a static geometric structure. Electric fields generated within the electrochemical double layer provide an additional mechanism for regulating catalytic behavior. In cluster‐based systems, where charge density is highly localized and responsive to external perturbations, electric‐field‐induced polarization can substantially alter adsorption configurations, charge transfer processes, and proton‐transport pathways [12]. Consequently, catalytic activity is governed not solely by catalyst composition but by the collective properties of the cluster–support–electrolyte ensemble. Furthermore, the reliability of reported NRR activity depends critically on rigorous experimental validation. Isotope‐labeling experiments using ¹⁵N₂, combined with stringent contamination‐control protocols, remain essential for confirming genuine ammonia production and distinguishing authentic catalytic activity from false‐positive signals [99]. These standards are indispensable for establishing reliable structure–activity relationships and ensuring meaningful comparison among catalyst systems. Recent advances in machine learning‐assisted catalyst discovery further emphasize the importance of interfacial descriptors. Computational screening studies identify parameters such as interfacial charge density, electronegativity mismatch, and orbital overlap as key predictors of catalytic performance [100]. These descriptors provide a quantitative framework for rational optimization of cluster–support interfaces. Hence, interfacial engineering integrates electronic coupling, geometric confinement, electrolyte interactions, solvent dynamics, and electric‐field effects to define the functional landscape of cluster‐based nitrogen reduction catalysts. By tailoring the cluster–support–electrolyte interface, it becomes possible to simultaneously enhance N₂ activation, stabilize key intermediates, and suppress competing reactions.

5. Synthetic Strategies

5.1. Bottom‐up Synthesis

Bottom‐up synthetic strategies offer an unparalleled platform for constructing metal clusters with precisely defined nuclearity, geometric configuration, and electronic structure, parameters that directly dictate the energetics of N₂ adsorption and activation. In contrast to top‐down methodologies, these approaches enable deliberate encoding of catalytic functionality at the atomic level, thereby establishing a direct structure–activity relationship for electrocatalytic nitrogen reduction. Recent advances in molecular cluster catalysis further reinforce the power of bottom‐up design in achieving well‐defined catalytic performance. For instance, Maman et al. investigated an atomically precise nickel–thiolate cluster, [Ni₆(PET)₁₂], operating as a homogeneous electrocatalyst, demonstrating selective ammonia formation with a Faradaic efficiency of 25% and a production rate of 16.2 μg h^–1 cm^–2 at −2.3 V versus Fc⁺/Fc [9]. The well‐defined tiara‐like coordination environment enables controlled electron delivery and efficient N₂ activation via both distal and alternating pathways (Figure 4a), with *N₂H identified as the rate‐limiting intermediate. These results highlight how precise control over nuclearity and coordination geometry directly translates into measurable catalytic activity and selectivity. Atomically precise clusters represent the most rigorously controlled systems within this paradigm. For instance, Li et al. demonstrated that Au₄Pt₂ clusters immobilized on defect‐engineered graphene exhibit markedly enhanced N₂ activation, originating from interfacial charge redistribution that enriches Pt‐centered electron density and promotes efficient π‐backdonation into antibonding orbitals of adsorbed N₂ [21]. This study highlights how discrete nuclearity governs adsorption energetics and reaction pathways. In a related context, thiolate‐stabilized gold clusters reported by Shichibu et al. revealed that ligand‐directed synthesis allows fine modulation of electronic states, thereby tuning catalytic selectivity at the molecular level [101].

FIGURE 4.

(a) Schematic representation of the pathway for N₂ fixation using Ni₆(PET)₁₂ catalyst [9], Reproduced with permission [9]. Copyright 2023, John Wiley & Sons. (b) The energy level cleavage for *NH2NH2* and bismuth ions, the transitional form of nitrogen in the nitrogen reduction [8], Reproduced with permission [8]. Copyright 2023, John Wiley & Sons. (c) Mass transfer/nitrate ions adsorption and promotion process over Cu _x –N₄ catalyst [47], Reproduced with permission [47]. Copyright 2025, American Chemical Society. (d) Synthesis process of NU‐902(M) [65]. Reproduced with permission [65]. Copyright 2026, SciOpen.

Building on these insights, systematic investigations of small clusters have elucidated the critical role of nuclearity in intermediate stabilization. Zheng et al. identified that triatomic configurations (Ni₃, Ru₃, Ir₃) anchored on N‐doped graphene exhibit optimal charge delocalization, which facilitates *NNH formation, a key kinetic bottleneck in NRR [22]. Consistently, Wang et al. demonstrated that Fe₃ clusters surpass isolated Fe1 sites by enabling cooperative multicenter electron donation, effectively lowering the activation barrier for the initial hydrogenation step [102]. These findings collectively underscore that multiatom ensembles provide electronic environments inaccessible to single‐atom catalysts. Furthermore, colloidal routes have enabled the synthesis of size‐controlled Ru clusters, where reduced coordination numbers and tunable d‐band centers correlate with improved nitrogen reduction performance [103]. Complementarily, molecular precursor strategies have yielded Mo‐based clusters that emulate the electronic characteristics of the FeMo‐cofactor, thereby promoting N₂ activation through bioinspired electronic structures [104]. Such approaches bridge homogeneous and heterogeneous catalysis by translating enzymatic motifs into solid‐state systems. Spatial confinement has emerged as an additional lever for tuning catalytic behavior. Wang et al. reported Fe–Cu clusters embedded within graphitic carbon nitride matrices, where geometric restriction enforces short metal–metal distances and enhances cooperative activation of N₂ while stabilizing key intermediates [39]. Similarly, clusters derived from MOF templates exhibit uniform dispersion and electronically modulated environments that suppress competing hydrogen evolution and improve NRR selectivity [105]. These observations demonstrate that confinement not only stabilizes active sites but also reshapes their electronic landscape.

Ligand chemistry further expands the design space of bottom‐up systems. Controlled ligand removal or electrochemical activation can expose catalytically active surfaces while preserving cluster integrity, generating dynamic interfaces under operating conditions [29]. In parallel, ligand‐induced electronic perturbations have been shown to modulate N₂ binding strength and alter reaction pathways, providing an additional dimension for catalyst optimization [106]. Theoretical analyses corroborate these experimental findings by revealing that small clusters can circumvent conventional scaling relationships that constrain bulk catalysts. DFT studies indicate that independent tuning of *NNH and *NH₂ adsorption energies is achievable in cluster‐based systems, thereby enabling lower overpotentials [107]. Charge density analyses further demonstrate that optimal electron distribution within clusters minimizes activation barriers for N₂ reduction steps [108]. Importantly, bottom‐up synthesis enables precise incorporation of multiple metal centers, facilitating synergistic interactions. Bimetallic clusters, such as Fe–Mo and Ru–Fe, exhibit complementary roles, where one component activates N₂ while the other stabilizes hydrogenated intermediates [26]. Extension to trimetallic architectures introduces additional degrees of freedom in electronic structure engineering, offering enhanced catalytic flexibility [109]. Finally, the integration of bottom‐up synthesis with descriptor‐based design has advanced the rational development of NRR catalysts. Parameters, such as d‐band center position, Bader charge, and adsorption free energies, provide predictive insight into catalytic performance, while emerging machine learning frameworks leverage structurally precise clusters to accelerate discovery and optimization [110, 111].

5.2. Top‐Down Approaches

Top‐down synthetic methodologies provide a complementary route for constructing metal cluster electrocatalysts for nitrogen reduction by inducing fragmentation, surface reconstruction, or defect generation within preexisting solids. In contrast to atomically programmed bottom‐up systems, these strategies inherently generate clusters in conjunction with lattice vacancies, local strain fields, and electronically coupled interfaces. Such features are particularly advantageous for electrocatalytic N₂ activation, where cooperative interactions between metal centers and surrounding defective environments can stabilize key intermediates (*N₂, *NNH, *NHNH) while kinetically suppressing the competing HER. Recent advances further demonstrate that top‐down reconstruction can extend beyond transition‐metal systems to include p‐block elements with unconventional electronic configurations. For example, Kang et al. reported that electrochemical reconstruction of BiOBr leads to the in situ formation of atomically dispersed Bi clusters with mixed‐valence Bi⁺/Bi²⁺ states, enabling efficient NRR catalysis through modulation of Bi 6p orbitals [8]. This system achieves an ammonia production rate of 113 μg h^–1 mg^–1 _cat and a Faradaic efficiency of 30%, highlighting competitive catalytic performance. The enhanced activity is attributed to strengthened coupling between Bi 6p orbitals and N₂ (Figure 4b), which facilitates electron transfer into antibonding orbitals while simultaneously suppressing HER. This work expands the scope of top‐down strategies by demonstrating that electronic structure engineering via p‐orbital tuning can provide an alternative pathway to efficient nitrogen activation. A defining characteristic of top‐down‐derived clusters is the emergence of electronically integrated cluster–support architectures. For example, Chen et al. demonstrated through first‐principles calculations that Mon clusters generated on Fe₅C₂ surfaces enable side‐on adsorption of N₂, accompanied by directional charge transfer from the Fe–Mo interface into antibonding orbitals of the adsorbate [112]. Notably, the Mo₅–Fe₅C₂ configuration minimized the limiting potential, whereas Mo₄–Fe₅C₂ exhibited superior suppression of parasitic proton reduction. This dual functionality highlights how the parent lattice actively modulates cluster electronic structure rather than acting as an inert scaffold. In a related system, in‐plane defect clusters engineered within MoS₂ basal planes were shown to transform an otherwise inert surface into a multicenter adsorption domain, facilitating associative hydrogenation pathways through side‐on N₂ binding [113]. These findings collectively underscore that top‐down restructuring can generate active motifs inaccessible in pristine materials.

Defect engineering represents a central design principle within this framework. Comprehensive analyses have established that vacancy ensembles, including oxygen, nitrogen, and sulfur vacancies, induce localized charge redistribution and create adsorption geometries favorable for N₂ activation [114, 115]. In particular, defect‐enriched iron monochalcogenides (FeS, FeSe) have been shown to expose coordinatively unsaturated Fe centers, which enhance N₂ binding affinity and lower the energy barrier for initial PCET steps [116]. Similarly, engineered defect–interface architectures in transition‐metal sulfides enable simultaneous modulation of conductivity and adsorption selectivity, thereby improving Faradaic efficiency toward ammonia formation [117]. These observations demonstrate that defect topology, rather than merely defect density, governs catalytic performance. Beyond static defect creation, electrochemical environments can dynamically reconstruct catalyst surfaces, giving rise to transient cluster species under operating conditions. Timoshenko et al. provided compelling evidence that atomically dispersed metal sites on nitrogen‐doped carbon can reversibly aggregate into subnanometer clusters during electrolysis and subsequently redisperse upon potential reversal [53]. This dynamic restructuring implies that the catalytically relevant active phase may differ substantially from the as‐synthesized material, emphasizing the necessity of operando characterization. Such behavior is particularly pertinent to NRR systems, where subtle changes in nuclearity can alter adsorption configurations and reaction energetics. In addition, top‐down derivation of clusters from larger nanoparticles offers a practical pathway to generate catalytically active ensembles with controlled support interactions. Yang et al. reported that supported Ni clusters exhibit NRR activity strongly dependent on both cluster size and substrate phase, indicating that electronic coupling with the support governs intermediate stabilization [49]. Complementary theoretical analyses further reveal that finite‐size electronic states and low‐coordination surface atoms characteristic of such clusters promote *NNH formation while disfavoring hydrogen adsorption, thereby improving selectivity toward ammonia [24]. These findings reinforce the notion that cluster–support synergy is an intrinsic feature of top‐down systems.

Despite these advantages, precise control over cluster nuclearity and geometric uniformity remains a fundamental challenge in top‐down synthesis. Consequently, rigorous validation protocols are indispensable. Reliable identification of active sites requires the integration of operando X‐ray absorption spectroscopy, in situ vibrational techniques, isotope‐labeled ¹⁵N₂ experiments, and high‐resolution postreaction microscopy. Without such corroboration, apparent ammonia production may arise from extrinsic sources, including nitrogen‐containing contaminants or competing reduction pathways. Therefore, the most robust top‐down studies are those that unify defect engineering, interfacial electronic analysis, dynamic reconstruction, and stringent product verification within a single mechanistic framework. In this context, top‐down approaches not only complement bottom‐up precision but also provide access to complex catalytic environments that closely resemble realistic operating conditions.

5.3. In Situ Formation

In situ formation strategies redefine the notion of catalyst identity in electrocatalytic nitrogen reduction by recognizing that the catalytically competent metal cluster is frequently generated under applied potential rather than preserved from the as‐synthesized state. Within cluster‐based NRR systems, electrochemical polarization, electrolyte composition, interfacial pH gradients, and adsorbate coverage collectively drive dynamic processes such as atom migration, cluster nucleation, lattice reduction, and defect reconfiguration. Consequently, the active phase should be regarded as a reaction‐induced construct, whose structure and electronic properties evolve in direct response to the operating environment. Recent experimental studies further substantiate the critical role of dynamic reconstruction in governing catalytic performance. For example, Shen et al. demonstrated that Cu single atoms anchored on nitrogen‐doped carbon nanocages undergo operando transformation into Cu _x clusters coordinated with nitrogen (Cu _x –N₄) under electrochemical conditions (Figure 4c) [47]. This in situ evolution generates highly active sites with optimized electronic structure, enabling an exceptional ammonia production rate of 99.4 mol h^–1 g_Cu ^–1 (2.30 mol h^–1 g_cat ^–1) and a Faradaic efficiency of 99.3%. The superior performance is attributed to reduced energy barriers for adsorption and protonation steps, as well as suppression of hydrogen evolution, highlighting how reconstruction can directly enhance catalytic efficiency. A compelling illustration of such dynamic restructuring was provided by Timoshenko et al., who demonstrated that Cu–N–C frameworks undergo reversible transformation between atomically dispersed Cu centers and subnanometric clusters or nanoparticles under potential modulation [53]. Although originally investigated in the context of CO₂ electroreduction, this work establishes a generalizable principle for nitrogen activation: Nitrogen‐coordinated carbon supports can reversibly stabilize and release metal ensembles depending on the electrochemical potential.

Complementary studies on Cu‐, Co‐, and Zn‐based M–N–C systems further confirm that electrochemical conditions can induce aggregation of isolated sites into few‐atom clusters, followed by redispersion upon potential reversal [118]. Such reversible evolution suggests that catalysts frequently categorized as single‐atom systems may, under NRR‐relevant conditions, operate through transient cluster motifs that provide multicenter adsorption geometries essential for N₂ activation. This concept extends naturally to catalysts derived from reducible compounds, including oxides, nitrides, carbides, and chalcogenides. Under cathodic bias, these materials often undergo partial reduction and lattice rearrangement, generating defect‐rich surfaces populated by low‐coordination metal clusters. Systematic analyses of electrochemical reconstruction in nitrogen‐cycle catalysis reveal that oxide‐derived materials can evolve into electronically distinct phases with altered adsorption energetics and intermediate stabilization capabilities [114]. For instance, conductivity‐driven restructuring of CuO _x under reductive potentials leads to the formation of cluster‐like Cu nanograins with mixed oxidation states, which significantly modify the binding strength of nitrogen‐containing intermediates [119]. Although such systems are frequently studied for nitrate reduction, the mechanistic implications, namely, cluster generation via reduction, dissolution–redeposition, and defect‐mediated nucleation, are directly transferable to N₂ reduction chemistry.

Importantly, the dynamic nature of in situ cluster formation necessitates a reassessment of how active sites are identified in NRR electrocatalysis. As emphasized in recent analyses, cluster size, oxidation state, and metal–support interaction are not static descriptors but evolve continuously during operation [24]. This variability challenges conventional postreaction characterization, which may fail to capture transient yet catalytically dominant species. Accordingly, recent perspectives highlight that single atoms, few‐atom clusters, and nanoparticles can interconvert under working conditions, and that accurate assignment of the active phase requires time‐resolved and potential‐dependent interrogation [57]. In this regard, operando techniques, such as X‐ray absorption spectroscopy, provide insight into coordination environment and valence evolution, while small‐angle X‐ray scattering enables direct tracking of cluster growth and dissolution processes [120]. The integration of these methods with electrochemical measurements is essential to establish causality between structural evolution and catalytic performance. Equally critical is the rigorous validation of ammonia production, particularly in systems undergoing continuous reconstruction. Best‐practice protocols emphasize the necessity of isotope‐labeled ¹⁵N₂ experiments, stringent contamination control, and quantitative correlation between spectroscopic signatures and catalytic outputs [121]. Without such measures, apparent activity enhancements may arise from extrinsic nitrogen sources or competing reduction pathways rather than genuine N₂ activation at reconstructed clusters. Overall, in situ formation represents a powerful yet experimentally demanding strategy for generating active metal clusters in NRR electrocatalysis. Its principal advantage lies in enabling access to highly reactive, defect‐rich cluster–support interfaces that are difficult to predefine synthetically. However, this advantage is accompanied by intrinsic complexity, as the catalyst structure is continuously evolving. Therefore, the most reliable implementations of this approach integrate reconstructable precursors, conductive defect‐engineered supports, controlled electrochemical activation protocols, and comprehensive operando characterization. Within such a framework, in situ formation transitions from an uncontrolled transformation into a deliberate strategy for constructing catalytically optimal subnanometric ensembles for nitrogen activation.

5.4. Stabilization Strategies

The stabilization of subnanometric metal clusters under electrochemical conditions remains a decisive factor governing their viability for nitrogen reduction. Owing to their intrinsically high surface energy and undercoordinated atomic configurations, such clusters are thermodynamically prone to coalescence, dissolution, or continuous restructuring during polarization. Consequently, effective stabilization must not only inhibit structural degradation but also preserve the electronic characteristics required for efficient N₂ adsorption and activation, while ensuring that active sites remain accessible under realistic reaction conditions. A primary strategy involves the establishment of strong metal–support interactions that simultaneously anchor clusters and regulate their electronic states. In nitrogen‐doped carbon matrices, coordination between transition metals and pyridinic or graphitic nitrogen sites provides robust stabilization against migration. Yan et al. demonstrated that Fe clusters confined within N‐coordinated carbon frameworks maintain structural integrity during electrolysis while exhibiting enhanced N₂ activation, attributed to efficient charge transfer mediated by Fe—N bonding environments [122]. Analogously, Zhao et al. reported that Co and Ni clusters stabilized through nitrogen coordination retain high dispersion and resist sintering under prolonged operation, with the coordination environment directly influencing adsorption energetics of nitrogen intermediates [123]. These findings indicate that heteroatom anchoring serves as both a geometric constraint and an electronic regulator, a combination that is particularly beneficial for NRR catalysis.

In parallel, oxide‐supported systems provide an alternative stabilization pathway through strong metal–oxygen interactions and defect‐mediated anchoring. For instance, Ru clusters supported on CeO₂ have been shown to localize preferentially at oxygen vacancy sites, where reduced Ce³⁺ centers facilitate electron donation to adsorbed N₂, thereby strengthening activation while preventing cluster migration [103]. Similarly, TiO₂‐supported metal clusters exhibit stabilized oxidation states under cathodic bias due to strong interfacial bonding, which suppresses aggregation and modulates catalytic behavior [124]. Notably, the redox‐active nature of oxide supports introduces an adaptive electronic environment, enabling dynamic adjustment of cluster properties during NRR operation. Spatial confinement further enhances cluster stability by physically restricting atomic mobility. Porous frameworks, such as graphitic carbon nitride, MOFs, and derived carbon matrices, offer well‐defined cavities that prevent coalescence while maintaining reactant accessibility. Wang et al. demonstrated that Fe–Cu clusters confined within g‐C₃N₄ exhibit sustained dispersion and improved catalytic performance, arising from enforced proximity between metal centers that promotes cooperative N₂ activation [39]. Likewise, clusters embedded in MOF‐derived carbon architectures show uniform distribution and resistance to sintering, resulting in prolonged catalytic lifetimes under electrochemical conditions [105]. These observations highlight that confinement not only stabilizes cluster size but also preserves the geometric arrangements necessary for multisite reaction pathways. Ligand‐mediated stabilization provides an additional dimension of control, particularly for atomically precise clusters. While ligands protect clusters during synthesis by preventing uncontrolled aggregation, their controlled removal or transformation under electrochemical conditions can expose catalytically active surfaces. Häkkinen et al. established that ligand–metal interactions significantly influence the electronic structure of clusters, thereby affecting adsorption strength and reaction selectivity [125]. In electrocatalytic environments, partial ligand desorption or restructuring can generate active configurations that balance stability with reactivity, suggesting that ligand dynamics should be considered an integral component of catalyst design.

Encapsulation strategies offer a complementary approach by introducing physical barriers that inhibit cluster growth without obstructing mass transport. Carbon‐coated clusters, for example, retain their nanoscale dimensions during extended electrolysis due to the protective effect of graphitic shells, which limit sintering while permitting diffusion of N₂ and protons [126]. Similarly, yolk–shell architectures provide internal void space that accommodates structural fluctuations while preventing aggregation, thereby maintaining catalytic activity over extended timescales. Such architectures illustrate how mechanical confinement can be combined with chemical accessibility to achieve long‐term stability. Beyond structural considerations, the electrochemical environment itself plays a nonnegligible role in stabilizing clusters. Electrolyte cations can accumulate at the electrode interface and influence local charge distribution, thereby modulating both cluster stability and intermediate binding. Hao et al. demonstrated that alkali metal cations can stabilize negatively charged intermediates and suppress dissolution processes, indirectly enhancing catalyst durability during nitrogen reduction [31]. Furthermore, solvent organization and ionic strength can alter interfacial interactions, indicating that stabilization must be evaluated within the full electrochemical context rather than as an intrinsic material property. Finally, it is increasingly evident that stability in cluster‐based systems is inherently dynamic. Operando investigations reveal that clusters can undergo reversible aggregation and redispersion depending on applied potential and reaction conditions. Timoshenko et al. showed that such dynamic equilibria can sustain a population of catalytically active clusters during operation, suggesting that stability should be interpreted as the persistence of an active ensemble rather than the immobility of a fixed structure [53]. This perspective shifts the focus from static preservation to controlled structural adaptability under working conditions. Overall, the stabilization of metal clusters for electrocatalytic nitrogen reduction requires an integrated strategy that combines strong interfacial anchoring, spatial confinement, electronic modulation, and environmental control. The most effective systems are those in which these elements operate synergistically, ensuring that cluster nuclearity and electronic structure are maintained while allowing sufficient flexibility for catalytic turnover. Such a holistic approach is essential for translating the intrinsic activity of subnanometric clusters into sustained and reliable ammonia production.

5.5. Scalability

The translation of metal cluster‐based electrocatalysts for nitrogen reduction from conceptual laboratory systems to technologically relevant platforms requires the reconciliation of atomic‐level precision with synthetic scalability, economic feasibility, and operational durability. Although bottom‐up approaches enable deterministic control over cluster nuclearity and electronic structure, their reliance on dilute conditions, elaborate ligand chemistry, and multistep purification protocols imposes intrinsic limitations on large‐scale implementation. Accordingly, scalable strategies must retain the defining catalytic attributes of clusters, high atom efficiency, cooperative multisite reactivity, and tunable electronic configurations, while ensuring reproducibility under practical manufacturing conditions. Recent advances in coordination‐framework‐based materials further demonstrate how structurally defined architectures can bridge atomic precision with scalability. For example, Li et al. developed NU‐902(M) MOF electrocatalysts (Figure 4d) with tunable transition‐metal centers, where NU‐902(Cu) achieved an ammonia production rate of 13.76 mg h^–1 mg_cat ^–1 and a Faradaic efficiency of 71.65% at −0.9 V versus RHE [65]. The high performance is attributed to the robust Zr–oxo backbone, well‐defined Cu–N₄ coordination sites, and a downshifted d‐band that suppresses HER while promoting efficient electron transfer and intermediate stabilization. Such systems highlight how scalable frameworks can integrate cluster‐like active sites with controlled electronic environments. Colloidal methodologies have progressed toward scale‐up through refined control over nucleation and growth kinetics. Vajda et al. established that cluster size selection can be preserved at increased production volumes by regulating precursor flux and reaction dynamics, enabling gram‐scale synthesis without substantial broadening of size distributions [106]. Nevertheless, extension to multimetallic systems introduces additional complexity, as compositional homogeneity must be maintained to avoid phase segregation that compromises catalytic uniformity. These challenges underscore the need for kinetic control strategies that remain effective under concentrated conditions.

In contrast, electrochemical deposition offers a scalable and inherently integrative route, wherein cluster formation occurs directly on conductive substrates under applied potential. Wang et al. demonstrated that controlled electrodeposition can yield highly dispersed metal clusters with strong substrate coupling, thereby enabling continuous fabrication of catalytically active electrodes without intermediate handling steps [127]. This approach not only minimizes material loss but also facilitates direct incorporation into electrochemical devices, representing a practical advantage for NRR systems where interfacial properties are critical. Support‐assisted strategies further contribute to scalability by leveraging bulk‐preparable host materials that stabilize cluster formation at high loadings. Porous carbon architectures derived from MOFs, nitrogen‐doped graphene, and graphitic carbon nitride provide structurally defined environments for anchoring clusters. Zhao et al. showed that MOF‐derived carbons enable uniform dispersion of metal clusters even at elevated concentrations, maintaining catalytic activity while allowing gram‐scale synthesis [105]. Similarly, large‐scale preparation of nitrogen‐functionalized carbon supports has enabled reproducible immobilization of transition‐metal clusters with consistent electronic properties across batches [122]. Complementary developments in coordination polymer systems further emphasize the role of multinuclear cluster motifs in scalable electrocatalysis. Wang et al. reported cobalt‐based coordination polymers (NJUZ‐2 and NJUZ‐3) featuring uniformly distributed multinuclear Co clusters, achieving Faradaic efficiencies up to 98.5% at −0.8 V versus RHE and exceptionally high ammonia production rates of 3370.6 mmol h^–1 g_cat ^–1 under flow‐cell conditions [62]. The high performance is attributed to cooperative interactions among Co centers, which facilitate adsorption and hydrogenation of intermediates while maintaining structural stability under harsh conditions. These results illustrate how ordered frameworks can combine scalability, durability, and high catalytic efficiency.

Top‐down methodologies inherently align with industrial scalability, as they utilize bulk materials and established processing techniques such as thermal treatment, chemical etching, and mechanical activation. For instance, controlled reduction and defect engineering in metal sulfides and oxides can generate cluster‐like active sites within extended solids, providing a cost‐effective route to high‐density catalytic ensembles [128]. However, such approaches require precise control over defect distribution and local structure to avoid loss of catalytic specificity, highlighting a key trade‐off between scalability and structural precision. Continuous‐flow synthesis has recently emerged as a promising platform for bridging laboratory control with industrial throughput. By regulating parameters such as precursor concentration, temperature, and residence time within flow reactors, reproducible formation of metal clusters can be achieved at larger scales. Advances in flow chemistry have demonstrated the feasibility of producing nanoclusters with consistent size and composition, while offering compatibility with automated processing and downstream integration [129]. This methodology provides a pathway toward continuous manufacturing of NRR catalysts with controlled properties. From a materials perspective, scalability is closely linked to elemental availability and cost considerations. While noble metals, such as Ru, Pt, and Au, exhibit favorable electronic structures for N₂ activation, their scarcity imposes economic constraints. Consequently, increasing emphasis has been placed on earth‐abundant metals, including Fe, Co, Ni, and Mo, which can achieve competitive catalytic performance when engineered at the cluster level to optimize electronic descriptors and adsorption energetics [130]. The rational substitution of noble metals with abundant alternatives is therefore essential for viable large‐scale deployment.

Equally critical is the retention of catalytic performance under prolonged electrochemical operation. Operando investigations reveal that cluster‐based catalysts may undergo structural evolution, including aggregation or redispersion, which can either enhance or degrade activity depending on system‐specific dynamics. Timoshenko et al. demonstrated that such transformations can be reversible under controlled potentials, suggesting that scalability must be coupled with strategies that preserve an active cluster population during extended electrolysis [53]. This requirement links synthetic scalability directly with stabilization and dynamic structural control. Finally, reproducibility and methodological standardization are indispensable for translating scalable synthesis into reliable performance metrics. Suryanto et al. emphasized that rigorous NRR evaluation necessitates strict contamination control, isotope‐labeled ¹⁵N₂ validation, and consistent reporting protocols to ensure that observed ammonia production originates from genuine nitrogen reduction [57]. This is particularly important for cluster catalysts, where subtle variations in nuclearity or electronic structure can lead to significant discrepancies in activity. Overall, scalable synthesis of metal cluster‐based NRR catalysts demands an integrated framework that combines controlled growth, robust support engineering, economically viable materials selection, and rigorous performance validation. The most promising strategies are those that preserve the essential physicochemical characteristics of clusters while enabling reproducible, large‐scale production, thereby bridging the gap between fundamental catalyst design and practical ammonia synthesis technologies.

6. Metal Cluster‐based Electrocatalysts for N₂ Activation

6.1. Noble Metal Clusters

Noble metal clusters represent a uniquely informative class of electrocatalysts for nitrogen reduction, positioned at the intersection of high electronic conductivity, chemical robustness, and tunable surface reactivity. While extended noble metal surfaces typically favor hydrogen evolution due to energetically accessible proton adsorption, the transition to subnanometric clusters fundamentally alters this behavior. In such confined regimes, reduced coordination numbers, discrete electronic states, and enhanced charge polarization reshape adsorption energetics, enabling N₂ activation pathways that are inaccessible on bulk analogues. As emphasized in recent analyses, nanocluster catalysis cannot be extrapolated directly from nanoparticle behavior; instead, catalytic performance emerges from a complex interplay between nuclearity, electronic quantization, and adsorbate‐induced charge redistribution [24]. Recent studies on atomically precise ligand‐protected clusters further highlight how dynamic surface restructuring under electrochemical conditions can activate otherwise inert noble metal systems. For example, Chen et al. demonstrated that the N‐heterocyclic carbene‐stabilized Au₁₃ cluster undergoes selective ligand dissociation under reductive potentials, exposing low‐coordination Au sites that become catalytically active, enabling stable syngas production with a CO/H₂ ratio of 0.8–1.1 across a wide potential window (−1.0 to −0.26 V) [18]. This electrochemically induced reorganization enhances surface accessibility and modifies adsorption energetics, illustrating how ligand dynamics can be leveraged to generate active sites in situ. Although investigated in CO₂ reduction, the underlying principle, controlled exposure of coordinatively unsaturated sites, directly applies to N₂ activation on noble metal clusters. Gold‐based systems provide a prototypical example of this transition. When the coordination environment is deliberately reduced, Au evolves from a relatively inert metal toward a catalytically competent platform for nitrogen activation. Zhang et al. demonstrated that low‐coordinated Au centers significantly enhance N₂ reduction under ambient conditions by strengthening adsorption through unsaturated surface sites while avoiding excessive hydrogen binding [131]. This behavior is further corroborated by size‐dependent studies in which monodisperse Au nanoparticles exhibited systematic variations in ammonia production as a function of particle diameter, confirming that the catalytic response is governed by surface coordination and quantum‐size effects rather than intrinsic metallic identity alone [132]. These findings collectively establish that the catalytic role of Au is activated only within a narrow structural window defined by low coordination and finite‐size electronic structure.

The incorporation of secondary metals into Au clusters introduces an additional level of electronic control. Shehzad et al. reported graphene‐supported Au₂Ag₂ and Au₄Cu₂ nanoclusters, where the latter displayed markedly enhanced nitrogen reduction performance, achieving a Faradaic efficiency approaching 50% under optimized conditions [133]. The improved activity was attributed to synergistic charge redistribution arising from heterometal incorporation, which modulates the adsorption strength of nitrogen intermediates while suppressing competing proton reduction. Such observations demonstrate that alloying enables precise tuning of electronic asymmetry within small clusters, thereby optimizing the balance between activation and selectivity. Complementary advances in atomically precise cluster synthesis further reinforce the importance of heterometallic design. Yao et al. constructed graphene‐supported Au₄Pt₂ clusters with well‐defined composition and coordination, demonstrating enhanced N₂ activation at the cluster–support interface, while substitution of Pt with Pd further improved ammonia yield and Faradaic efficiency relative to the Pt analogue [134]. These results highlight how precise compositional control enables systematic tuning of adsorption energetics and catalytic performance. Related insights emerge from Ag–Au alloy systems, where composition‐dependent electronic modulation governs catalytic behavior. In alloy nanotube architectures, the ratio of Ag to Au was found to directly influence nitrogen adsorption and hydrogen evolution competition, indicating that even in larger structures, alloy composition serves as a key descriptor for reactivity [127]. Although these systems extend beyond the strict cluster regime, they provide mechanistic guidance for designing bimetallic clusters in which electronic structure can be continuously tuned through compositional variation.

Palladium‐based clusters offer a complementary perspective due to their stronger interaction with nitrogen‐containing intermediates. However, this enhanced binding can also promote hydrogen evolution if contiguous metallic domains are formed. To address this limitation, Du et al. employed zeolite confinement to isolate Pd clusters within ZSM‐5 frameworks, thereby restricting aggregation and generating spatially discrete active ensembles [135]. This approach effectively balances the intrinsic reactivity of Pd with geometric constraints that suppress undesired proton reduction, highlighting the importance of confinement in regulating noble metal cluster behavior. Ruthenium remains a central model system for understanding the interplay between nitrogen activation and hydrogen competition. While Ru exhibits strong affinity for N₂‐derived intermediates, its catalytic selectivity is highly sensitive to surface hydrogen coverage. Grand‐canonical DFT studies have demonstrated that potential‐dependent adsorbate populations must be explicitly considered, as increasing hydrogen coverage can block active sites and suppress N₂ activation despite favorable intrinsic binding energies [30]. Complementary experimental–theoretical investigations have further shown that selectivity is governed by dynamic surface coverage under applied bias, rather than by zero‐potential descriptors alone [136]. These findings are particularly relevant for Ru clusters, where nuclearity directly influences both adsorption energetics and hydrogen coverage, rendering cluster size a critical determinant of catalytic selectivity. Broader insights into cluster catalysis can also be drawn from studies beyond nitrogen reduction. For instance, Yang et al. demonstrated that atomically precise Ag‐based clusters with controlled heterometal doping exhibit strong dependence of catalytic activity and selectivity on electronic structure (Figure 5a,b), where Ni‐doped clusters showed higher intrinsic activity due to improved charge transport and increased electrochemically active surface area [6]. Similarly, Zhao et al. showed that reactivity in heteronuclear Rh‐based clusters is governed by descriptors such as charge distribution, orbital alignment, and polarizability, with optimal activity observed in clusters containing 3–7 atoms (Figure 5c) and rate constants varying over several orders of magnitude [61]. These findings reinforce the general principle that catalytic performance in small clusters is dictated by a multidimensional interplay of electronic and geometric factors, which is directly transferable to N₂ activation chemistry.

FIGURE 5.

(a) Catalytic efficiency of silver nanoclusters using different amounts of KOH, and (b) different potentials in 0.2 M KOH [6], Reproduced with permission [6]. Copyright 2026, Elsevier B.V. (c) Comparison between experimental measurement and cluster size [61], Reproduced with permission [61]. Copyright 2024, American Chemical Society. (d) FE of NH₃ and H₂ at different potentials [29], Reproduced with permission [29]. Copyright 2025, American Chemical Society. (e) Limiting potential measurements using graphene‐supported Fe₃, Co₃, and Ni₃ clusters [92], Reproduced with permission [92]. Copyright 2023, Elsevier B.V. (f) Kinetic obstacles for breaching the N≡N bond using Cr₄ clusters/Ti₂CO₂ MXene [72]. Reproduced with permission [72]. Copyright 2025, John Wiley & Sons.

Given the typically low ammonia production rates associated with noble metal systems, rigorous validation protocols are indispensable. It is now well established that trace contamination from nitrogen‐containing species can lead to false‐positive results, necessitating stringent experimental controls. Suryanto et al. outlined comprehensive methodological standards, including isotope‐labeled ¹⁵N₂ experiments, contamination‐free setups, and quantitative analysis of reaction products [57]. Subsequent reviews have reinforced the necessity of combining multiple verification techniques, such as parallel blank tests and postreaction electrolyte analysis, to ensure that detected ammonia originates exclusively from N₂ reduction [41, 43, 44]. These requirements are particularly critical for noble metal clusters, where subtle variations in activity must be distinguished from background contributions. From a design perspective, the optimization of noble metal clusters for NRR can be rationalized through several interdependent parameters. First, cluster nuclearity must be minimized to maximize low‐coordination active sites while preserving sufficient electronic conductivity. Second, alloying strategies should be employed to introduce asymmetric charge distribution, thereby enhancing nitrogen activation while suppressing hydrogen adsorption. Third, confinement within conductive and defect‐rich supports is essential to prevent sintering and maintain isolated active ensembles. Finally, catalytic evaluation must account for potential‐dependent competition between hydrogen evolution and nitrogen reduction, as noble metals are particularly sensitive to changes in surface coverage under operating conditions. Overall, noble metal clusters are best understood not as practical ammonia synthesis catalysts in their current form, but as highly informative model systems for elucidating the fundamental principles governing electrocatalytic nitrogen activation. Their well‐defined electronic structures, tunable compositions, and sensitivity to coordination environment make them ideal platforms for disentangling the roles of finite‐size effects, alloy‐induced polarization, and interfacial charge transfer. Insights derived from these systems are therefore expected to guide the rational design of more abundant and scalable catalysts with comparable or superior performance.

6.2. Transition Metal Clusters

Transition‐metal clusters constitute the foundational platform for electrocatalytic nitrogen activation, owing to their partially filled d orbitals that enable synergistic σ‐donation from N₂ and π‐backdonation into antibonding orbitals. In contrast to isolated single‐atom centers, clustered metal ensembles introduce adjacent active sites capable of stabilizing multiple adsorption geometries, including end‐on, bridging, and side‐on configurations. This geometric and electronic flexibility is particularly critical for facilitating the first PCET step (*N₂ → *NNH), which is widely recognized as the kinetic bottleneck in NRR. Iron‐based clusters occupy a central position in this context due to their close analogy to the Fe‐rich active sites in nitrogenase enzymes. Computational investigations of Fe₃ clusters embedded within C₂N frameworks revealed that the intrinsic cavity structure stabilizes the trimer while enabling concerted electron donation to adsorbed N₂, thereby lowering the activation barrier for initial hydrogenation [137]. Expanding this concept, heterometallic trimers, such as FeCo₂ and NiCo₂, supported on vacancy‐engineered graphene were shown to exhibit reduced *NNH formation barriers through intermetallic charge redistribution, demonstrating that compositional tuning within earth‐abundant systems can effectively regulate reaction energetics [138]. Similarly, Fe₄ clusters anchored on GaS substrates were predicted to activate N₂ via simultaneous electron transfer into multiple antibonding orbitals, resulting in significantly reduced limiting potentials and enhanced catalytic efficiency [41].

Complementary experimental advances further validate the effectiveness of multinuclear transition‐metal ensembles. For instance, Das et al. reported a trinuclear nickel complex with a confined thiolate‐ligand environment that achieves highly selective N₂ reduction, delivering a Faradaic yield exceeding 80% (Figure 5d), a turnover number of 8.63 × 10⁴, and a turnover frequency of 11.98 s^–1 [29]. The enhanced performance arises from the spatially confined cavity, which governs N₂ binding geometry and directs a six‐electron reduction pathway toward NH₃ rather than hydrazine, highlighting the critical role of geometric confinement and ligand architecture in cluster‐based catalysis. These studies collectively underscore the importance of multicenter electron donation in Fe‐based cluster catalysis. Molybdenum‐containing clusters provide complementary functionality by stabilizing nitrogenated intermediates while maintaining strong orbital coupling with N₂. For instance, Mo₁₂ clusters supported on C₂N were shown to offer multiple active sites and parallel reaction pathways, thereby reducing the energetic penalty associated with sequential hydrogenation steps [139]. In a nitrogenase‐inspired design, Mon clusters interfaced with Fe₅C₂ substrates demonstrated that interfacial charge transfer plays a decisive role in N₂ activation, with cluster nuclearity governing the balance between nitrogen reduction and hydrogen evolution [112]. In addition, Mo‐based defect clusters within MoS₂ matrices were found to modify the basal‐plane electronic structure, creating localized adsorption pockets that favor side‐on N₂ binding and cooperative activation [140]. Cobalt and nickel clusters offer a distinct approach by enabling modulation of d‐band position to avoid excessive binding of late‐stage intermediates. Experimental and theoretical studies of Ni clusters supported on TiO₂ polymorphs revealed that catalytic activity is strongly influenced by support‐induced charge transfer, with identical cluster sizes exhibiting different behaviors depending on the substrate phase [49]. Comparative analyses of homonuclear Fe₃, Co₃, and Ni₃ clusters on graphene further demonstrated that metal identity dictates the balance between N₂ adsorption strength, *NNH formation kinetics, and hydrogen evolution suppression [141]. More recently, systematic screening of triatomic 3d transition‐metal clusters on graphdiyne identified Ti₃/GDY as a particularly promising system, where the conjugated support stabilizes the cluster while facilitating efficient electron transport to adsorbed N₂ [51].

Additional theoretical studies further highlight the importance of cluster nuclearity and support interaction. Liu et al. demonstrated that graphene‐supported Fe₃, Co₃, and Ni₃ clusters exhibit favorable limiting potentials of −0.60, −0.64, and −0.45 V (Figure 5e), respectively, indicating competitive NRR activity [92]. Similarly, Das et al. showed that larger icosahedral clusters, such as Fe₁₃, Co₁₃, and Ni₁₃, supported on graphyne enable strong N₂ adsorption and activation through significant charge transfer into antibonding orbitals, with reaction energetics and selectivity governed by cluster composition and metal–support coupling [81]. These findings reinforce that both cluster size and support interaction critically determine catalytic performance. The role of the support extends beyond structural stabilization to active electronic participation. Comprehensive screening of transition‐metal clusters on biphenylene sheets revealed that the substrate modulates both cluster stability and reaction energetics, with W₂, Ru₂, and Mo₄ clusters exhibiting favorable onset potentials for NRR [44]. Similarly, trimeric and tetrameric clusters supported on defective phosphorene–carbide substrates were shown to partially circumvent conventional scaling relationships, enabling independent tuning of intermediate adsorption energies [27]. Graphdiyne‐supported multimetallic clusters further demonstrated that heterometallic combinations can simultaneously optimize N₂ adsorption, *NNH stabilization, and hydrogen binding, providing a multidimensional design space for catalyst optimization [45].

Ruthenium‐based clusters bridge the gap between noble and base‐metal behavior, offering strong N₂ activation alongside tunable hydrogen adsorption. Ru₃ clusters confined within defective g‐C₃N₄ frameworks were predicted to outperform extended Ru surfaces by altering the energetics of the rate‐determining hydrogenation step [142]. In a related system, Ru₃ clusters anchored on Mo₂CO₂ MXene exhibited enhanced N₂ activation due to strong interfacial coupling, which facilitates charge transfer and stabilizes associative reaction pathways [143]. Further insights into multicenter catalytic mechanisms are provided by studies on tetranuclear clusters. Yu et al. showed that Cr₄ clusters supported on Ti₂CO₂ MXene exhibit low overpotential (0.29 V) and reduced kinetic barriers (≈0.40 eV) (Figure 5f), arising from cooperative orbital interactions among multiple metal centers that enable efficient reactant activation and intermediate stabilization [72]. These results highlight the potential of higher nuclearity clusters in enabling complex multielectron transfer reactions. At the molecular level, bioinspired Mo–Fe–S clusters provide direct evidence for cooperative nitrogen activation within multimetallic environments. Recent experimental work has shown that synthetic clusters mimicking the nitrogenase cofactor can mediate N₂ reduction at Fe centers, with Mo and S atoms contributing to electronic stabilization and proton transfer pathways [144]. This molecular insight reinforces the importance of cooperative metal–metal and metal–ligand interactions in cluster‐based catalysis. Overall, the performance of transition‐metal clusters in NRR cannot be attributed solely to elemental identity. Instead, catalytic behavior arises from the interplay of cluster nuclearity, metal–metal spacing, heterometallic charge polarization, support‐induced electronic effects, and adsorption geometry. Iron and molybdenum clusters typically provide strong N₂ activation but require careful control of intermediate binding energies; cobalt and nickel clusters offer improved selectivity but depend on support‐mediated electronic tuning; ruthenium clusters deliver high intrinsic activity but must avoid hydrogen‐dominated regimes. Consequently, the most effective transition‐metal cluster catalysts are those in which cluster composition and support are codesigned as a unified electronic system and validated under rigorous isotope‐labeled conditions to ensure genuine nitrogen reduction activity.

6.3. Bimetallic Clusters

Bimetallic clusters provide a chemically versatile platform for electrocatalytic nitrogen reduction by enabling spatial and electronic differentiation of catalytic functions within a single subnanometric ensemble. In contrast to monometallic clusters, where adsorption and hydrogenation processes occur on electronically equivalent sites, heterometallic architectures allow the deliberate partitioning of roles: One metal center can preferentially activate N₂ via π‐backdonation, while the second modulates proton transfer kinetics, stabilizes hydrogenated intermediates, or attenuates competitive hydrogen evolution. Such cooperative behavior is particularly advantageous for overcoming the intrinsic challenge of balancing *N₂ activation with controlled hydrogenation along the associative pathway. Experimental evidence for this cooperative mechanism has been demonstrated in Pd–Cu amorphous nanoclusters supported on graphene, where Shi et al. reported enhanced ammonia formation under ambient conditions [145]. The improved activity was attributed to synergistic electronic interactions between Pd and Cu, which generate a heterogeneous charge distribution within the cluster, facilitating N₂ adsorption while moderating proton reduction kinetics. In a more structurally defined system, graphene‐supported Au₄Cu₂ and Au₂Ag₂ clusters exhibited distinct catalytic responses, with Au₄Cu₂ delivering superior ammonia yield and Faradaic efficiency [133]. This behavior highlights that both cluster nuclearity and heterometal incorporation dictate the electronic landscape available for N₂ activation. Similarly, atomically dispersed Au₄Pt₂ clusters on defective graphene were shown to create localized electron‐rich regions through dopant‐induced charge redistribution, providing a model system for understanding how heterometallic coupling enhances adsorption and activation of inert molecules [21].

Earth‐abundant bimetallic clusters further expand the design space by combining catalytic efficiency with improved scalability. Confined Fe–Cu clusters embedded within graphitic carbon nitride frameworks act as subnanometric reaction environments in which the short Fe–Cu separation enables efficient electron communication between adjacent sites [39]. This structural arrangement enhances N₂ adsorption and promotes cooperative hydrogenation, leading to significantly improved ammonia production relative to monometallic analogues. Complementary studies on FeCu alloy systems supported on carbon nitride indicate that Cu can function as an electronic modulator, increasing electron density at Fe centers and thereby strengthening N₂ binding while maintaining favorable hydrogenation kinetics [146]. These results reinforce the concept that Fe‐centered activation can be optimized through adjacent heterometallic charge redistribution. The Fe–Mo pairing represents a particularly instructive case, reflecting the cooperative functionality observed in biological nitrogenase. Nanoporous Fe₃N–MoN systems synthesized via molten‐salt routes demonstrated notable NRR performance, which was attributed to interfacial electronic coupling between Fe and Mo nitride domains that facilitates sequential electron and proton transfer steps [147]. Similarly, polyoxometalate‐derived Mo–Fe oxide systems containing mixed‐valence phases were proposed to accelerate N₂ activation through multimetal charge delocalization, enabling more efficient intermediate stabilization [148]. Although these materials are not strictly atomically precise clusters, they exemplify the same fundamental principle: Adjacent metal centers with complementary electronic properties can distribute catalytic tasks across the reaction coordinate.

Direct experimental validation of interfacial electronic asymmetry in well‐defined cluster systems has been reported by Ren et al., who constructed sub‐2 nm Pd–MoO _x clusters supported on carbon, achieving an ammonia production rate of 46.5 μg h^–1 mg^–1 (Figure 6a,b) and a Faradaic efficiency of 13.0% [4]. The enhanced performance originates from a polarized Mo–N–N–Pd interfacial configuration, where Pd induces strong d–π* coupling with N₂, weakening the N≡N bond while simultaneously lowering the free‐energy barriers for subsequent protonation steps. This work provides clear mechanistic evidence that heterometallic polarization can simultaneously optimize N₂ activation and hydrogenation kinetics. Theoretical studies provide further mechanistic clarity by illustrating how bimetallic clusters can overcome limitations inherent to single‐site catalysis. Wang et al. proposed that dual‐metal donor–acceptor motifs generate asymmetric charge distributions that enhance N₂ activation while suppressing hydrogen adsorption, offering a generalizable strategy for catalyst design [149]. In a related investigation, Pd–W heteronuclear dimers supported on C₂N were identified as stable and active NRR catalysts, where the interplay between Pd and W modifies both hydrogen adsorption energetics and the potential‐determining hydrogenation step [150]. More complex three‐atom clusters, such as Nb₂M systems, further demonstrate that heterometallic combinations can simultaneously optimize N₂ activation, intermediate stabilization, and product desorption through tailored d–σ and d–π* interactions [43]. A defining advantage of bimetallic clusters lies in their ability to partially circumvent scaling relationships that constrain catalytic optimization. Asymmetrical Fe–Co active sites have been shown to decouple adsorption energies of key intermediates by assigning distinct reaction steps to electronically differentiated metal centers [151]. Similarly, FeCo₂ and NiCo₂ trimers supported on vacancy‐engineered graphene exhibit reduced barriers for *NNH formation due to intermetallic charge redistribution, while maintaining manageable energetics for subsequent hydrogenation steps [138]. Broader screening of heteronuclear transition‐metal pairs on C₂N substrates confirms that bimetallic combinations can systematically tune stability and catalytic performance, although certain pairings may suffer from structural instability under operating conditions [152].

FIGURE 6.

(a) HAADF–STEM image and (b) NH₃ production rates [4], Reproduced with permission [4]. Copyright 2025, John Wiley & Sons. (c) Free energy for the generation of different intermediates during NRR in distal pathway [16], Reproduced with permission [16]. Copyright 2020, American Chemical Society. (d) Interpretable framework identifying key descriptor across homonuclear system [78]. Reproduced with permission [78]. Copyright 2025, American Chemical Society.

Beyond direct N₂ reduction, insights from adjacent nitrogen‐electroreduction reactions further validate the utility of bimetallic clusters. For example, Fe–Co nanoclusters have been shown to function as dual‐site catalysts for nitrate and nitrite reduction to ammonia, where tandem adsorption and hydrogenation processes occur across neighboring metal centers [153]. High‐throughput computational screening of bimetallic alloys likewise demonstrates that electronic descriptors derived from d‐state interactions can predict limiting potentials across diverse compositions, providing a rational framework for catalyst discovery [147]. Additional advances in atomically precise bimetallic clusters further highlight the role of electronic and ligand‐induced modulation. Dai et al. demonstrated that Cu₄Pt₂ clusters with electron‐withdrawing ligands achieve exceptional catalytic performance, reaching a Faradaic efficiency of 91.84% and an ammonia production rate of 13.65 mg_cat ^–1 h^–1 at −0.5 V [73]. The enhanced activity arises from a synergistic mechanism in which Cu sites activate reactants while Pt centers facilitate proton transfer, with ligand‐induced charge redistribution further strengthening this cooperative behavior. Although studied in nitrate reduction, the mechanistic insights into dual‐site functionality and electronic tuning are directly transferable to NRR systems. These studies collectively emphasize that bimetallic clusters should be regarded not as simple extensions of alloy nanoparticles, but as electronically heterogeneous ensembles in which local coordination environments and charge polarization govern catalytic pathways. To sum up, the effectiveness of bimetallic clusters in NRR arises from the integration of complementary functionalities within a single nanoscale architecture. Their catalytic behavior is dictated by the interplay between metal identity, cluster nuclearity, intermetallic distance, support‐induced electronic effects, and competition with hydrogen evolution. Systems based on Fe–Cu and Fe–Mo typically enhance N₂ activation through cooperative charge transfer, while noble–base metal combinations offer additional flexibility in tuning adsorption energetics. Ultimately, the most promising bimetallic cluster catalysts are those in which site differentiation, electronic asymmetry, and structural stability are cooptimized and validated under rigorous isotope‐labeling protocols to ensure reliable nitrogen‐to‐ammonia conversion.

6.4. Heterostructures

Heterostructured metal cluster electrocatalysts represent an advanced level of architectural control in nitrogen reduction systems, wherein catalytic performance emerges from the collective behavior of cluster, support, and interfacial defects rather than from the metal ensemble alone. In such configurations, interfacial charge redistribution, polarity gradients, and local electric fields cooperatively regulate adsorption geometry, intermediate stabilization, and proton accessibility. This integrated functionality is particularly critical for NRR, where activation of the inert N≡N bond, stabilization of *NNH intermediates, and suppression of hydrogen evolution must be simultaneously achieved within a narrow electrochemical potential window. Two‐dimensional heterostructures provide a versatile platform for constructing such integrated catalytic environments. Chen et al. demonstrated through systematic screening that triatomic clusters embedded within defective W₂N₃ matrices exhibit composition‐dependent NRR activity, with VNiCu, Fe₃, and Fe₂Co clusters identified as particularly effective due to strong metal–support orbital coupling and vacancy‐assisted charge transfer [154]. In this system, nitrogen vacancies serve both as anchoring sites and as electronic modulators, enabling efficient electron donation to adsorbed N₂. Similarly, transition‐metal clusters anchored on biphenylene sheets were shown to benefit from the unique electronic structure of the support, which stabilizes subnanometric ensembles while tuning adsorption energetics; W₂, Ru₂, and Mo₄ clusters were highlighted as promising candidates based on their favorable limiting potentials [44]. Recent advances further demonstrate how multimetallic cluster engineering combined with defect‐modified supports can dramatically enhance catalytic performance. For example, Chen et al. identified VNiCu@NV–W₂N₃ as an optimal heterostructured catalyst, exhibiting an exceptionally low limiting potential of −0.05 V, with strong spin‐dependent orbital coupling facilitating efficient N₂ activation and hydrogenation [155]. This study highlights the combined role of vacancy engineering, spin polarization, and cluster composition in defining interfacial reactivity. Carbon‐based heterostructures with extended π‐conjugation further enhance cluster performance by combining structural confinement with efficient charge transport. Graphdiyne‐supported triatomic clusters, for instance, exploit acetylenic linkages to immobilize metal atoms within well‐defined pores, enabling short metal–metal distances and facilitating cooperative N₂ activation [45]. In parallel, C₂N‐based systems provide nitrogen‐rich cavities that stabilize M₃ clusters while modulating their electronic density, thereby improving both adsorption strength and selectivity against competing hydrogen evolution [156]. These findings emphasize that the geometric and electronic characteristics of the host framework are inseparable from the catalytic properties of the embedded cluster.

Heterostructures based on oxides and carbides introduce an additional dimension through redox‐active interfaces. In TiO₂‐supported Ni cluster systems, the catalytic response was shown to depend strongly on the crystallographic phase of the support, with interfacial charge transfer and bonding interactions dictating NRR energetics [49]. A particularly instructive example is provided by Mon clusters interfaced with Fe₅C₂, where the carbide support and metal cluster form a cooperative catalytic unit that promotes side‐on N₂ adsorption and direct electron transfer; variations in Mo nuclearity were found to modulate both activity and selectivity with respect to hydrogen evolution [112]. Such systems demonstrate that heterostructures can distribute catalytic functions across multiple components rather than localizing them on a single active site. Nitride and carbon nitride supports are especially effective in constructing electronically coupled heterointerfaces due to their intrinsic nitrogen content. Fe–Cu alloy clusters anchored on carbon nitride frameworks were shown to generate electron‐enriched Fe sites through interfacial charge transfer, enhancing nitrogen activation and catalytic efficiency [146]. Moreover, defective g‐C₃N₄ supports have been predicted to stabilize small Ru₃ and Fe₃ clusters within vacancy sites, where confinement and electronic perturbation modify reaction energetics relative to extended metal surfaces [142]. Complementary insights from model catalyst systems further highlight the importance of coordination environment and electronic coupling at heterointerfaces. Sahoo et al. demonstrated that Fe–C₂N systems exhibit a lower onset potential (≈−0.7 V) compared to Au–C₂N (≈−1.6 V) (Figure 6c), reflecting stronger d–π* interactions and enhanced N₂ activation, although accompanied by stronger intermediate binding and HER competition [16]. These results reinforce the need to balance activation strength with intermediate desorption in heterostructured systems.

Layered heterostructures further expand the design space by enabling control over interlayer electronic communication. In Cr‐doped expanded graphite systems, modulation of interlayer spacing was shown to influence charge transfer to the active site, thereby affecting N₂ adsorption and reduction kinetics [80]. This observation suggests that catalytic activity can be tuned not only through direct metal–support interactions but also through long‐range electronic coupling within multilayered architectures. Recent proposals involving twisted two‐dimensional heterostructures introduce the concept of moiré‐induced charge modulation, offering a route to spatially periodic electronic fields that could regulate cluster activity at the nanoscale [57]. Defect engineering remains a critical component of heterostructure design. Vacancy sites, whether oxygen, nitrogen, or sulfur, create localized regions of enhanced electron density that facilitate N₂ adsorption and lower the activation barrier for initial hydrogenation steps [64]. In MoS₂‐based systems, the introduction of in‐plane defect clusters has been shown to convert inert basal planes into active multisite regions capable of stabilizing side‐on N₂ adsorption configurations [98]. These results highlight that defects are not merely structural imperfections but integral elements of the catalytic interface. Machine learning and data‐driven approaches are increasingly contributing to the rational design of such complex systems. Mou et al. developed an interpretable framework identifying key descriptors such as atomic charge distribution, spin density, and HOMO energy as governing factors for cluster reactivity, enabling accurate prediction of catalytic trends across homonuclear and heteronuclear systems (Figure 6d) [78]. These tools provide a powerful means to decode structure–reactivity relationships in heterostructured catalysts.

Further advancements in cluster–support systems highlight the role of cooperative multicenter interactions. Yu et al. showed that tetranuclear clusters on Ti₂CO₂ MXene supports achieve low overpotentials (≈0.35 V, reduced to 0.24 V upon Fe incorporation), driven by strong cluster–support charge transfer and orbital hybridization effects [55]. Similarly, alternative cluster‐based heterostructures, such as Pd₃@SnS₂, demonstrate how cooperative adsorption sites can enable unconventional reaction pathways, achieving low limiting potentials (≈−0.08 V) and reduced energy barriers for key bond‐forming steps [58]. An additional advantage of heterostructures lies in their ability to modulate selectivity by controlling hydrogen adsorption. Because hydrogen evolution competes strongly with nitrogen reduction, effective heterostructures must weaken *H binding while preserving favorable energetics for *N₂ and *NNH intermediates. This balance is evident in systems such as Ni cluster–TiO₂ interfaces and Mon–Fe₅C₂ heterostructures, where support‐induced electronic effects shift the relative energetics of NRR and HER pathways [52, 71]. Consequently, heterostructures provide multiple adjustable parameters, including cluster size, vacancy concentration, support phase, and interfacial polarity, that can be optimized to achieve selective nitrogen reduction. Overall, heterostructured metal cluster catalysts should be regarded as integrated interfacial systems in which catalytic performance arises from the coordinated interaction of cluster composition, support electronic structure, and defect chemistry. Their effectiveness in NRR is governed by the ability to simultaneously control charge transfer, adsorption geometry, and reaction selectivity within a unified framework. Future progress will depend on the precise engineering of these interfaces, combined with operando characterization and isotope‐labeled validation, to ensure that observed activity originates from genuine nitrogen reduction at well‐defined heterointerfaces.

6.5. Nitrogenase‐Inspired Systems

Nitrogenase‐inspired cluster catalysts represent a conceptually rigorous approach to N₂ activation, in which design is guided not by compositional similarity alone but by the functional architecture of the enzymatic cofactor. The biological FeMo cofactor operates through a highly integrated [MoFe₇S₉C] framework, where multimetallic cooperativity, sulfur‐bridged electron delocalization, and the presence of an interstitial light atom collectively define its reactivity. Spatzal et al. established that the central atom is a μ₆‐carbide, a discovery that redefined mechanistic interpretations by demonstrating that carbon plays a decisive role in organizing electron density across the Fe framework during catalysis [104]. This insight provides a critical design principle for heterogeneous systems: Interstitial atoms can act as electronic regulators rather than passive structural elements. Subsequent advances in synthetic cluster chemistry have reinforced the central role of Fe sites within sulfur‐rich environments. A landmark study by Arashiba et al. demonstrated that [Mo₃S₄Fe] cubane‐type clusters can catalytically reduce N₂ through Fe‐centered activation, even in the absence of nitrogen‐based ligands [144]. This result provides direct experimental evidence that sulfur‐coordinated Fe centers are capable of binding and transforming N₂, thereby validating the relevance of Fe–S motifs in catalyst design. More recently, the synthesis of FeMoco‐inspired clusters containing a trigonal‐prismatic [Fe₆C] core has enabled systematic exploration of how interstitial carbon redistributes charge within multimetallic frameworks, offering a structural analogue for probing carbide‐mediated electronic effects in heterogeneous catalysts [19]. Diversity among nitrogenase cofactors further refines these design concepts. Structural elucidation of the Fe‐only nitrogenase revealed an [8Fe:9S:C] cluster capable of sustaining N₂ reduction without Mo or V, underscoring that the essential requirement is not a specific metal identity but a properly organized Fe–S–C network [20]. In contrast, investigations of the FeV cofactor have shown that substitution of Mo with V alters the electronic distribution and redox behavior of the cluster, thereby influencing substrate reduction pathways [66]. Complementary analyses of FeFe cofactors and associated P‐clusters indicate that efficient catalysis depends on a broader electron transfer architecture, in which auxiliary clusters regulate charge accumulation and delivery to the active site [93]. These findings highlight that nitrogenase function arises from a network of interacting components rather than an isolated catalytic center.

Mechanistic studies of the enzymatic cycle provide further guidance for catalyst design. Computational analyses incorporating thermal fluctuations suggest that N₂ activation involves dynamic restructuring of the cofactor, transient exposure of Fe sites, and coupled release of H₂ during reduction [56]. In parallel, detailed investigations of proton transfer pathways demonstrate that proton delivery is tightly linked to the redox state of the cluster and cannot be decoupled from electron transfer processes [17]. These insights imply that effective electrocatalytic analogues must incorporate not only active metal sites but also local environments capable of directing proton flux with spatial and temporal precision. Molecular model systems serve as an important bridge between enzymatic chemistry and solid‐state catalysts. Diiron complexes have been shown to stabilize reduced nitrogen intermediates through cooperative binding, illustrating how adjacent Fe centers can distribute electron density during successive hydrogenation steps [24]. Earlier work on molecular Fe catalysts for N₂‐to‐NH₃ conversion further demonstrated that appropriate ligand environments and controlled proton/electron supply can enable complete reduction cycles [39]. These studies collectively suggest that heterogeneous cluster catalysts should replicate key features of molecular systems, including proximity of active centers and flexibility of coordination environments to accommodate reaction intermediates. From a materials perspective, sulfur‐rich inorganic frameworks offer a practical route to implementing nitrogenase‐inspired motifs. Iron‐based chalcogels, for example, have been shown to facilitate nitrogen fixation by providing extended Fe–S networks that mimic the electronic environment of biological cofactors [18]. Although primarily explored in photocatalytic contexts, these materials demonstrate that sulfur coordination can modulate electron density around Fe centers and promote N₂ activation. Studies on Fe‐based electrocatalysts similarly emphasize that low‐valent Fe sites, combined with Fe–S or Fe–N coordination environments, enhance adsorption and selectivity by stabilizing key intermediates [17].

The biosynthetic pathway of nitrogenase cofactors provides additional inspiration for catalyst fabrication. The enzyme NifB assembles the FeMo precursor through fusion of smaller [4Fe–4S] clusters followed by insertion of carbide and sulfide species, indicating that active sites are constructed through controlled cluster aggregation rather than isolated atom placement [99]. Recent studies on cofactor trafficking and maturation further reveal that metal exchange, ligand incorporation, and scaffold‐assisted assembly are essential for achieving the final active structure [77]. Translating these principles to electrocatalysis suggests that synthetic strategies based on cluster fusion, confinement, and postsynthetic modification may yield more effective catalysts than conventional impregnation methods. Finally, rigorous validation remains essential for nitrogenase‐inspired systems. The presence of nitrogen‐containing ligands, supports, or precursors introduces a high risk of false‐positive ammonia detection. As emphasized by Suryanto et al., reliable demonstration of N₂ reduction requires isotope‐labeled ¹⁵N₂ experiments, comprehensive contamination control, and quantitative analysis of reaction products [57]. For biomimetic catalysts, additional verification through operando characterization is necessary to confirm that the active cluster structure is maintained under electrochemical conditions. Overall, nitrogenase‐inspired metal cluster catalysts should be designed according to functional principles rather than superficial compositional similarity. Key elements include sulfur‐bridged multimetallic frameworks, interstitial light atoms that regulate electron distribution, flexible bonding environments that enable dynamic site exposure, and controlled proton–electron delivery. Integration of these features within conductive and stable supports offers a pathway toward genuinely biomimetic electrocatalysts capable of efficient and selective nitrogen reduction.

6.6. Performance Benchmarking of Metal Cluster Electrocatalysts

Quantitative benchmarking of metal cluster‐based electrocatalysts for nitrogen reduction remains inherently complex, as reported ammonia production metrics are strongly influenced by experimental variables, including electrolyte composition, catalyst loading, applied potential, membrane configuration, and analytical methodology. Consequently, rigorous comparison must integrate multiple descriptors, namely NH₃ yield rate, Faradaic efficiency (FE), overpotential, hydrogen evolution suppression, operational stability, and isotope‐labeled validation, rather than relying on isolated performance values. Within this framework, the cited studies reveal clear trends that correlate catalytic behavior with cluster composition, nuclearity, and support interaction. Among noble metal clusters, graphene‐supported Au₄Cu₂ nanoclusters represent one of the highest‐performing systems reported under ambient conditions, achieving an NH₃ yield of 4.14 μg h^–1 cm^–2 and FE approaching 49.6% at −0.8 V versus RHE [133]. This performance exceeds that of Au₂Ag₂ clusters within the same study, demonstrating that both heterometal doping and cluster nuclearity critically influence catalytic output. The improved activity is attributed to Cu‐induced charge redistribution, which enhances N₂ adsorption while suppressing excessive hydrogen binding. In contrast, Pd clusters confined within ZSM‐5 zeolites exhibit a different performance profile: Although peak activity is lower, the system demonstrates superior structural stability and resistance to aggregation, highlighting that confinement can prioritize durability and sustained activity over maximum yield [135]. Thus, noble metal clusters illustrate a trade‐off between activity and stability depending on structural design. Transition‐metal and bimetallic clusters generally outperform monometallic noble systems when evaluated on a combined activity–selectivity basis. Confined Fe–Cu clusters embedded in g‐C₃N₄ cavities show a marked enhancement in performance, with ammonia yield nearly doubled and FE increased by ≈54% relative to single‐metal analogues, reaching ≈34% FE [39]. This improvement arises from cooperative interaction between Fe and Cu centers, which enables efficient N₂ adsorption and facilitates the *N₂ → *NNH step while maintaining moderate hydrogen binding. Supporting this trend, FeCu alloy/carbon nitride systems demonstrate that Cu acts as an electronic promoter, increasing electron density at Fe active sites and thereby strengthening N₂ activation [146]. These studies establish heterometallic synergy as a key determinant of improved catalytic performance. As shown in Table 1, the reported different electrocatalysts shows varying production rate, Faradaic efficiency, and stability. While some catalyst obtain high production rate, others provide superior Faradaic efficiency or long‐term stability.

TABLE 1.

Performance benchmarking of metal cluster electrocatalysts.

Catalyst name	Yield	Faradaic efficiency	Stability	References
Pd–MoO x /C	46.5 μg h−1 mg−1	13.0%	24 h	[4]
Ni20–FAN	46.73 μg h−1 mgcat −1	35.74%	24 h	[11]
Trinuclear nickel complex	TOF 11.98 s−1	80%	n/a	[29]
Fe–Cu/g‐C3N4	9.86 μg h−1 mg−1	34%	10 h	[39]
Re/In2Se3	26.63 μg h−1 cm−2	30.8%	96 h	[41]
Ru/Mn1.04–RuO2	52.6 μg mg−1 h−1	44.29%	72 h	[60]
Cu4Pt2	13.65 mgcat −1 h−1	91.84%	n/a	[73]
Graphene/Au4Cu2	4.14 μg h−1 cm−2	49.6%	12 h	[133]

Copper‐based clusters provide an additional benchmark due to their balance between cost and tunability. Atomically precise Cu clusters supported on graphene oxide exhibit moderate NH₃ production with improved selectivity compared to bulk Cu, indicating that controlled nuclearity and support coupling can mitigate HER competition [34]. More advanced ligand‐protected Cu₁₄ clusters further improve performance by stabilizing defined electronic structures, demonstrating that molecular‐level control can translate into enhanced catalytic efficiency and durability [20]. While these systems do not yet match the highest‐performing heterometallic clusters, they highlight the potential of earth‐abundant materials when precisely engineered. Theoretical studies provide a normalized basis for comparing intrinsic catalytic properties. Raju's comprehensive analysis shows that cluster nuclearity, electronic configuration, and adsorption mode collectively determine activity trends, with smaller clusters often exhibiting stronger N₂ activation but increased susceptibility to HER [24]. Complementarily, the dual‐metal donor–acceptor model proposed by Wang et al. explains why heterometallic clusters consistently outperform monometallic systems: Asymmetric charge distribution enhances N₂ activation while weakening hydrogen adsorption, thereby improving selectivity [149]. These descriptor‐based insights align closely with experimental observations across Fe–Cu, Au–Cu, and Pd–W systems. It is also essential to contextualize performance relative to other nitrogen reduction pathways. Systems, such as FeCo bimetallic nanoclusters and Ni‐based layered double hydroxides, demonstrate exceptionally high FE in nitrate or nitrite reduction, often exceeding values reported for direct N₂ reduction [24, 74]. However, these reactions involve different thermodynamic and kinetic constraints and should not be directly compared with N₂‐to‐NH₃ catalysts. Instead, they serve as mechanistic analogues for multielectron nitrogen hydrogenation processes and highlight the ability of cluster catalysts to mediate complex reaction pathways.

A comparative analysis of the cited studies reveals several consistent performance trends. First, heterometallic clusters (e.g., Fe–Cu, Au–Cu) outperform monometallic clusters due to enhanced charge polarization and cooperative adsorption. Second, confinement within porous or defect‐rich supports improves both stability and selectivity by suppressing aggregation and HER. Third, atomically precise or ligand‐defined clusters provide better control over electronic structure but may face challenges in scalability. Finally, systems that integrate cluster design with conductive, nitrogen‐affinitive supports exhibit the most balanced performance across activity, selectivity, and durability. Despite these advances, the field lacks standardized benchmarking protocols, which limits direct comparison across studies. Future performance evaluation should include: (i) NH₃ yield normalized to geometric area, catalyst mass, and electrochemically active surface area; (ii) FE reported over a defined potential range; (iii) long‐term stability under continuous operation; (iv) ¹⁵N₂ isotope verification to confirm nitrogen source; and (v) postreaction structural characterization to ensure cluster integrity. Without these criteria, high reported activity cannot be unequivocally attributed to genuine N₂ reduction. Overall, performance comparison across metal cluster NRR catalysts demonstrates that the most effective systems are those that combine heterometallic synergy, spatial confinement, and controlled electronic structure. Such integrated design strategies enable simultaneous optimization of N₂ activation, intermediate stabilization, and HER suppression, providing a clear pathway toward more efficient and reliable electrocatalytic nitrogen fixation.

7. Challenges and Controversies

7.1. Reliability of Ammonia Detection: False Positives and Contamination Pathways

The reliability of ammonia detection remains the most critical challenge in electrocatalytic nitrogen reduction, particularly for metal cluster‐based systems where intrinsic reaction rates are extremely low. Under such conditions, even trace contamination can dominate the measured NH₃ signal, leading to systematic overestimation of catalytic performance. Importantly, contamination is not limited to external sources but is often intrinsically linked to catalyst composition and experimental configuration. Comprehensive studies have shown that ammonia may originate from ambient air, laboratory consumables, membranes, and nitrogen‐containing precursors, while independent investigations demonstrated that comparable ammonia yields can be obtained without N₂ feed, indicating that system‐derived nitrogen sources frequently overshadow genuine N₂ activation. This issue is further amplified in cluster catalysts supported on nitrogen‐rich materials, where lattice or adsorbed nitrogen species can be released under electrochemical bias, mimicking catalytic nitrogen reduction. Even isotope‐labeling experiments require careful interpretation, as impurities in commercial ¹⁵N₂ gas can introduce false‐positive signals unless purification protocols are implemented. In parallel, analytical techniques, such as the indophenol blue method, are prone to interference from transition‐metal ions and organic species, necessitating cross‐validation using independent methods. Thus, false positives arise from a coupled interplay of contamination pathways, catalyst‐derived nitrogen, and analytical limitations, making rigorous validation indispensable for credible assessment of cluster‐mediated N₂ activation. Given the exceptionally low ammonia production rates typically observed for cluster‐based electrocatalysts, a standardized validation framework is essential for establishing the credibility of reported NRR activity. At a minimum, future studies should satisfy five fundamental criteria. First, NH₃ production must be quantified under both N₂ and Ar atmospheres using identical experimental conditions to exclude contributions from catalyst decomposition or environmental contamination. Second, isotope‐labeling experiments employing purified ¹⁵N₂ should be performed, with independent verification of isotopically labeled ammonia by ¹H NMR, mass spectrometry, or equivalent techniques. Third, rigorous blank control experiments, including open‐circuit conditions, catalyst‐free electrodes, and electrolyte‐only measurements, should demonstrate negligible NH₃ formation. Fourth, contamination assessment should include analysis of feed gases, membranes, electrolytes, laboratory consumables, and catalyst precursors to identify potential nitrogen‐containing impurities. Fifth, catalytic performance should be evaluated through reproducibility testing, including multiple independent experiments, Faradaic efficiency determination, NH₃ mass balance analysis, and extended stability measurements. For cluster‐based catalysts specifically, additional characterization before and after electrolysis should verify that nitrogen‐containing supports, ligands, or framework components are not serving as unintended nitrogen sources. Adoption of these minimum validation requirements would substantially improve the reliability, reproducibility, and comparability of reported NRR performances across different catalyst systems.

7.2. Lack of Standardization and Benchmarking in Nitrogen Reduction Reaction Evaluation

The absence of universally accepted benchmarking standards remains one of the most significant obstacles in electrocatalytic nitrogen reduction research. Reported H₃ yields and Faradaic efficiencies often vary dramatically due to differences in electrolyte composition, pH, catalyst loading, electrode architecture, gas purification procedures, and ammonia quantification methods. Consequently, comparisons between catalyst systems are frequently unreliable, and apparent performance improvements may reflect experimental artifacts rather than intrinsic catalytic activity. This challenge is particularly acute for metal cluster catalysts, where subtle changes in cluster nuclearity, coordination environment, or support interactions can produce large variations in electronic structure and catalytic behavior. Beyond methodological inconsistencies, the field has faced persistent concerns regarding false‐positive ammonia detection arising from airborne NH₃, nitrogen‐containing impurities in electrolytes and membranes, catalyst decomposition, laboratory contamination, and even impurities present in commercial ¹⁵N₂ gas supplies. Several landmark studies have demonstrated that many previously reported NRR activities cannot withstand rigorous isotope verification, highlighting the urgent need for standardized validation procedures. Importantly, Faradaic efficiency alone is an insufficient indicator of catalyst quality because high Faradaic efficiency values can be obtained at extremely low current densities that are far from practical operating conditions. Meaningful benchmarking should therefore simultaneously report ammonia yield rate, current density, NH₃ production rate normalized to catalyst loading and electrode area, stability, isotope‐confirmed product origin, and postreaction structural integrity. The benchmarking workflow shown in Table 2 is adapted from recently established community guidelines and best‐practice protocols for electrochemical NRR, which emphasize contamination elimination, rigorous ¹⁵N₂ isotope verification, multiple blank controls, reproducibility testing, and comprehensive postreaction characterization to ensure reliable identification of genuine N₂‐derived ammonia production [157, 158, 159, 160, 161].

TABLE 2.

Recommended benchmarking and validation protocol for electrocatalytic NRR.

Step	Validation requirement	Purpose
Contamination screening	Quantify NH3/NO x impurities in electrolytes, membranes, feed gases, cell components, and catalyst precursors before testing	Eliminates background nitrogen sources
Blank control I	Perform electrolysis under Ar atmosphere using identical conditions	Confirms that NH3 is not generated from catalyst decomposition or contamination
Blank control II	Conduct N2‐saturated experiments at open‐circuit potential without applied bias	Verifies that measured NH3 is electrochemically generated
Gas purification	Purify N2 and 15N2 streams using acid traps, alkaline scrubbers, or purification units	Removes residual NH3 and NO x contaminants
Isotope verification	Perform 15N2‐labeling experiments and verify 15NH4 + formation by nuclear magnetic resonance, gas chromatography–mass spectrometry (MS), liquid chromatography–MS, or isotope‐selective MS	Confirms ammonia originates from supplied N2
Quantification cross‐validation	Validate NH3 concentration using at least two independent analytical methods	Reduces analytical bias and false positives
Reproducibility assessment	Conduct ≥3 independent experiments and report standard deviations	Ensures statistical reliability
Activity benchmarking	Report NH3 yield rate, FE, partial current density, and normalization basis	Enables meaningful comparison across studies
Stability testing	Evaluate long‐term electrolysis performance and repeated cycling	Assesses catalyst durability
Postreaction characterization	Analyze catalyst using transmission electron microscopy (TEM), high‐angle annular dark‐field scanning transmission electron microscopy (HAADF–STEM), X‐ray photoelectron spectroscopy, X‐ray absorption spectroscopy (XAS), inductively coupled plasma–MS before and after reaction	Confirms structural integrity and active‐state evolution
Operando verification	Employ operando XAS, Raman, Fourier transform infrared spectroscopy, differential electrochemical mass spectrometry, or related methods	Identifies true catalytic species under working conditions
Interlaboratory validation	Independent replication by external laboratories	Establishes community‐wide reproducibility

7.3. Structural Instability and Dynamic Evolution of Metal Clusters

The structural stability of metal clusters under electrochemical conditions represents a fundamental limitation that distinguishes cluster catalysts from conventional nanomaterials. Due to their high surface energy and low coordination numbers, clusters are inherently prone to aggregation, dissolution, and dynamic restructuring during operation. Operando studies have revealed that the active catalytic species often evolves under applied potential, rather than remaining identical to the as‐synthesized structure. For example, reversible interconversion between isolated atoms and small clusters has been observed, indicating that the catalytically relevant ensemble is dynamically formed during electrochemical polarization. This behavior complicates mechanistic interpretation, as ex situ characterization may not accurately represent the active state. Additionally, support materials can undergo electrochemical degradation or phase transformation, weakening metal–support interactions and accelerating cluster migration. While confinement and encapsulation strategies can improve structural stability, they may simultaneously restrict access to active sites, creating a trade‐off between durability and catalytic activity. Consequently, stability in cluster‐based NRR systems should be viewed as a dynamic equilibrium governed by competing processes of aggregation and redispersion, requiring operando characterization to identify the true active configuration under working conditions.

7.4. Kinetic Limitations and Low Current Density

Low current density remains a major barrier to the practical implementation of electrocatalytic nitrogen reduction, reflecting intrinsic kinetic and transport limitations. The activation of the N≡N bond is associated with a substantial energy barrier, and the first hydrogenation step (*N₂ → *NNH) is widely recognized as the potential‐determining step in most catalytic systems. Even on optimized cluster catalysts, this step requires significant energy input, limiting overall reaction rates. At the same time, hydrogen evolution proceeds with lower activation barriers and competes directly for electrons and protons, reducing NRR selectivity. Although metal clusters can enhance N₂ adsorption and provide multiple active sites for cooperative activation, these advantages do not necessarily translate into high current densities. Mass transport limitations further exacerbate the problem, as the low solubility of N₂ in aqueous electrolytes restricts its availability at the electrode surface, leading to diffusion‐controlled reaction rates. As a result, even highly selective catalysts often operate far below industrially relevant current densities. Addressing this limitation requires an integrated approach combining catalyst design with reactor engineering, including gas‐diffusion electrodes, optimized electrolytes, and improved mass transport strategies. Ultimately, overcoming low current density will require simultaneous control of reaction kinetics, competing pathways, and transport phenomena.

7.5. Hydrogen Evolution Reaction Competition and Selectivity Constraints

A central controversy in electrocatalytic nitrogen reduction is the competition between NRR and HER, which fundamentally limits selectivity under aqueous conditions. Because proton reduction is kinetically more favorable, many catalysts that exhibit strong N₂ adsorption also bind hydrogen, leading to preferential HER activity at negative potentials. For metal cluster systems, this competition is particularly complex, as cluster nuclearity and electronic structure simultaneously influence both *N₂ activation and *H adsorption. Small clusters often exhibit enhanced electron density that promotes N₂ activation, but this same property can increase hydrogen binding, reducing selectivity. Conversely, strategies that weaken *H adsorption may also compromise N₂ activation, creating a delicate balance between activity and selectivity. Experimental and theoretical studies indicate that heterometallic clusters and defect‐engineered supports can partially decouple these competing interactions by introducing asymmetric charge distribution and site‐specific adsorption behavior. However, achieving complete suppression of HER remains challenging. Therefore, improving NRR selectivity requires precise control over electronic structure, adsorption energetics, and interfacial environment, highlighting the need for integrated catalyst and system‐level design.

7.6. Critical Assessment of Current Cluster‐Based Nitrogen Reduction Reaction Research

Despite the substantial progress achieved in cluster‐based NRR catalysis, several fundamental limitations continue to impede practical implementation. First, many reported ammonia production rates remain close to the detection limits of conventional analytical methods, increasing susceptibility to contamination‐related artifacts. Second, catalyst comparisons are often complicated by variations in electrolyte composition, gas purification procedures, isotope‐labeling protocols, and reporting metrics. Third, many mechanistic interpretations rely heavily on theoretical predictions without direct operando validation of the proposed active structures. Finally, catalyst durability remains insufficiently understood because cluster restructuring, aggregation, and redispersion may occur under electrochemical conditions. Consequently, future advances should focus not only on improving catalytic performance but also on establishing rigorous experimental standards, operando characterization methodologies, and reproducible benchmarking protocols that enable meaningful comparison across different catalyst systems.

Funding

This work was supported by the Prince Sattam bin Abdulaziz University project number (PSAU/2026/R/1447) and Implementation Plan for Scientific and Technological Innovation in Key Areas of the Inner Mongolia Autonomous Region (NMGLLY2025‐01).

Conflicts of Interest

The authors declare no conflicts of interest.

Aljohani Khalid, Hao Terigelema, Ahamd Irshad, Sebakhy Khaled O., Aljohani Bassam S., Ali Ijaz, Li Gao, Recent Advances and Perspectives of Metal Clusters for Electrochemical Nitrogen Reduction, ChemPlusChem 2026, 91, e70214. 10.1002/cplu.70214

Data Availability Statement

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