Defect chemistry in electrocatalysis: mechanistic insights and design principles for sustainable energy conversion and environmental remediation

1 Introduction

The accelerating transition toward a sustainable, carbon-neutral society is increasingly defined by the convergence of two global imperatives: decarbonizing industrial chemical manufacturing and mitigating the widespread degradation of aquatic ecosystems. These challenges are not independent; they are interconnected manifestations of a global energy–environment nexus where resource consumption, pollutant generation, and greenhouse gas emissions are thermodynamically and kinetically coupled (Xia et al., 2022; Glavič et al., 2023; Schnyder et al., 2025). Ammonia synthesis occupies a uniquely critical position within this nexus. As the foundation of global agriculture and a highly promising zero-carbon energy carrier, its production currently relies almost entirely on the century-old Haber–Bosch process. Operating under extreme thermodynamic constraints (350 °C–550 °C, 150–250 bar) to activate the highly stable N≡N bond (941 kJ mol^-1), this process consumes approximately 1%–2% of global energy supply and accounts for nearly 2% of anthropogenic CO₂ emissions (Ghavam et al., 2021; Zhang et al., 2024g; Ribeiro and Santos, 2025). Concurrently, rapid industrialization has triggered a water pollution crisis characterized by the continuous discharge of recalcitrant organics, toxic heavy metals, and excess nitrogenous species (e.g., NO₃ ⁻) into aquatic environments. Recent advances further highlight the dual role of nitrate pollution as both an environmental liability and a potential feedstock for decentralized ammonia production. In particular, oxygen-vacancy-assisted Cu–Co catalysts have demonstrated enhanced nitrate-to-ammonia conversion even at ultralow nitrate concentrations, reinforcing the importance of defect-mediated interfacial regulation under environmentally relevant conditions (Fan et al., 2025; Xu, 2025).

Conventional remediation technologies, which often rely on chemical-intensive and energy-demanding advanced oxidation processes (AOPs), frequently fail to achieve complete mineralization and rarely allow for resource recovery (Gopalakrishnan et al., 2023; Satyam and Patra, 2025; Oladipo, 2026). Recent studies also demonstrate that defect and coordination-environment engineering can redirect pollutant-removal pathways by modulating interfacial electron transfer and reactive oxygen species formation. For example, metal/support interactions have been shown to promote high-valent cobalt–oxo species during PMS activation, while atomically dispersed Fe sites with tailored coordination environments can facilitate electron-transfer pathways for antibiotic degradation (Yang L. et al., 2022; Fan et al., 2024; Ma D. et al., 2026).

Consequently, the development of electrified technologies capable of simultaneously advancing clean energy conversion and environmental remediation has become a central scientific priority. Electrocatalysis offers a transformative platform to address these dual challenges by utilizing renewable electricity to drive chemical transformations under ambient conditions. However, the widespread industrial deployment of systems such as the nitrogen reduction reaction (NRR), nitrate reduction reaction (NO₃RR), and advanced electrochemical oxidation is fundamentally bottlenecked by the intrinsic limitations of pristine catalyst surfaces. Idealized, perfectly crystalline metal or metal-oxide lattices are universally constrained by energy scaling relations—a manifestation of the Sabatier principle dictating that the binding energies of sequential reaction intermediates (e.g., *N and *NH₂ in NRR, or *O and *OH in water oxidation) are linearly correlated (Medford et al., 2015; Gao Q. et al., 2022; Yang et al., 2023a; Yi et al., 2026). Consequently, optimizing a pristine surface to activate a stable reactant inevitably leads to the over-stabilization of subsequent intermediates, poisoning the active site and severely limiting reaction kinetics and Faradaic efficiency. To surpass conventional performance limits, electrocatalyst design must evolve beyond compositional tuning to directly disrupt these universal scaling relations.

Defect engineering has emerged as the most powerful paradigm for achieving this disruption. In contemporary electrocatalysis, structural imperfections—such as vacancies, heteroatom dopants, dislocations, and lattice strain—are no longer viewed as unavoidable crystallographic anomalies, but rather as precision electronic regulators (Mete et al., 2024; Zhang Y. et al., 2024; 2025d; Mudoi and Saikia, 2025). By deliberately introducing specific defects, local lattice symmetry is broken, triggering orbital reconstruction. This manifests as highly localized perturbations in the electronic structure: shifting the transition-metal d-band center relative to the Fermi level, inducing work-function reductions (often by 0.3–1.0 eV), and driving localized charge accumulation or depletion (Jiao et al., 2022; Yuan et al., 2026; Zhu Z. et al., 2026). Because these electronic modulations are spatially confined to the defect site, they selectively alter the adsorption free energies (ΔG_ads) of specific intermediates without identically affecting others. For example, oxygen-vacancy formation in transition metal oxides can lower the activation barrier for N₂ or NO₃ ⁻ adsorption by 0.2–0.8 eV while simultaneously suppressing the competing hydrogen evolution reaction (HER) by destabilizing *H adsorption (Xiao et al., 2021; Ling et al., 2023; Zhang M. et al., 2023).

Despite the explosion of literature surrounding defect-engineered electrocatalysts, current understanding remains highly fragmented. Most studies remain siloed, treating green ammonia synthesis and wastewater remediation as distinct disciplines. This review instead examines the common physical-chemical principles through which defects regulate catalytic interfaces, while recognizing that the relevant descriptors, intermediates, rate-determining steps, and reaction pathways remain reaction-specific. The activation of N₂ for ammonia synthesis, the electroreduction of NO₃ ⁻ for concurrent water treatment and nutrient recovery, and the activation of peroxymonosulfate (PMS) for pollutant degradation can all involve defect-mediated charge redistribution, coordination changes, electronic-state modulation, and altered interfacial reaction energetics, although these effects manifest differently across reaction classes (Miller et al., 2023; Sun J. et al., 2023; Pei et al., 2024). By framing both energy conversion and environmental remediation through a common but reaction-specific electronic-structure lens, this review establishes defect chemistry as a transferable framework rather than a universal mechanistic equivalence for catalytic design.

A significant portion of current literature relies on ex-situ characterization (e.g., ultra-high vacuum X-ray photoelectron spectroscopy) to confirm the presence of vacancies or dopants, implicitly assuming these features survive under harsh, operando electrochemical conditions (e.g., cathodic polarization at −0.8 V vs. RHE in aqueous media). This assumption is frequently incorrect. Defect sites are thermodynamically metastable and highly susceptible to healing, migration, or triggering broader surface reconstruction under applied bias (Zhou et al., 2021; Xiao et al., 2024; Wang X. et al., 2026). Therefore, this review continuously distinguishes between “as-synthesized” structural defects and the dynamic, true active sites that evolve during reaction turnover. Moving beyond descriptive catalyst catalogues, the present review systematically interrogates the causal pipeline of defect electrocatalysis: from the thermodynamics of defect formation to electronic structure modulation, intermediate adsorption, and ultimate macroscopic performance. By synthesizing recent breakthroughs in both energy-focused NRR/NO₃RR and environment-focused advanced oxidation and reduction processes, we extract universal structure–activity descriptors. Ultimately, identifying how atomic-scale defects dynamically govern interfacial electron transfer will provide the definitive mechanistic foundation required to design the next-generation of multifunctional, AI-guided electrocatalysts for a carbon-neutral and circular economy.

To ensure a balanced synthesis of recent literature, publications from 2019 to 2026 were identified using Web of Science, Scopus, and Google Scholar. Searches combined defect-related terms, including “oxygen vacancy,” “single-atom,” “d-band center,” and “operando reconstruction,” with application-specific terms such as “nitrogen reduction,” “nitrate electroreduction,” “electro-Fenton,” and “persulfate activation.” Experimental and theoretical studies were considered separately when assessing quantitative performance and mechanistic evidence. For NRR, particular attention was given to ¹⁵N₂ isotope labeling, Ar-saturated and open-circuit controls, catalyst/electrolyte blanks, and quantitative ammonia analysis, with the completeness of these controls used to assess evidentiary strength. For defect thermodynamics and reaction mechanisms, greater weight was given to studies incorporating electrochemical-relevant treatments, such as solvation and computational hydrogen electrode approaches, where applicable. Quantitative data and mechanistic assignments were extracted together with their reported experimental or computational conditions, and values obtained under substantially different conditions were not treated as directly comparable.

2 Fundamental defect chemistry and electronic structure regulation

The catalytic behavior of an electrocatalyst is fundamentally governed by the atomic and electronic structure of its interface. While idealized crystalline models provide a useful baseline, real materials operate far from thermodynamic perfection. In contemporary electrocatalysis, structural imperfections are not viewed as anomalous deviations, but as precision levers capable of disrupting long-range lattice periodicity to tailor local electronic environments. Because ideal, pristine crystals are universally bound by linear scaling relations that dictate rigid, non-optimal intermediate binding energies, defect chemistry has emerged as the most critical paradigm for uncoupling these intermediates and bypassing inherent catalytic limitations (Kolding et al., 2025; Zhang et al., 2025e). Establishing a rigorous understanding of how defects form, how they manipulate localized quantum states, and how they survive under applied bias is essential for the rational design of both energy-conversion and water-remediation platforms.

Importantly, defect engineering should not be regarded as a single-variable strategy in which increasing defect concentration automatically improves catalytic performance. A defect is catalytically meaningful only when its formation, stability, electronic consequences, and interaction with reaction intermediates are considered together. Accordingly, the framework adopted here distinguishes four interconnected levels: (i) defect formation and thermodynamic stability, (ii) local electronic and interfacial-state modulation, (iii) reaction-specific intermediate energetics and kinetics, and (iv) the resulting catalytic performance under operating conditions. This distinction is particularly important because the same defect type can produce different catalytic consequences depending on the host material, defect charge state, electrolyte, applied potential, and reaction environment.

2.1 The hierarchical classification of defects: How do structural discontinuities localize charge?

Electrocatalytically relevant defects alter the electronic environment across varying spatial dimensions, necessitating a hierarchical classification into point, line, planar, and interfacial defects. The type and dimensionality of a defect determine whether its electronic influence is highly localized or distributed across an extended interface.

2.1.1 Point defects

Zero-dimensional lattice imperfections—primarily vacancies and heteroatom dopants—are the most extensively engineered defect class. Vacancies function as highly localized charge perturbations. In ionic solids, their generation is accurately described by Kröger–Vink notation. For example, the formation of a doubly ionized oxygen vacancy in a transition metal oxide leaves behind localized electrons that reduce adjacent metal cations, generating highly reactive, lower-valence active centers (Krishnan et al., 2022; Zheng et al., 2024). These anionic vacancies are critical across disparate fields: in green ammonia synthesis, they can provide localized electronic states that facilitate N₂ or NO₃ ⁻ adsorption and activation, while in advanced wastewater remediation, oxygen-deficient sites can participate in electron transfer and oxidant activation. However, the chemical consequence of a vacancy is not universal: whether it acts predominantly as an electron donor, adsorption site, redox center, or precursor to surface reconstruction depends on the host lattice and its local coordination environment. Thus, a nominally identical oxygen vacancy should not automatically be assigned the same catalytic function across different materials or reactions. Conversely, cationic defects (e.g., metal vacancies) alter the oxidation states of surrounding anions, creating electrophilic centers that excel at stabilizing oxygenated intermediates during the oxygen evolution reaction (OER) or organic electro-oxidation.

2.1.2 Extended and interfacial defects

Line defects (dislocations) and planar defects (grain boundaries, twin boundaries) introduce medium-to long-range structural discontinuities. Dislocations generate substantial lattice strain (often >2%), which systematically alters the degree of orbital overlap between adjacent atoms. Under tensile strain, the distance between metal atoms increases, narrowing the d-band and shifting its center closer to the Fermi level, which characteristically strengthens adsorbate binding (Fang et al., 2024; Luo et al., 2025; 2026). At the highest level of complexity, interfacial defects (heterojunctions and phase boundaries) induce massive, directional charge redistribution due to work function mismatch between distinct phases.

Rather than treating these defect classes as a simple hierarchy of catalytic importance, their dimensionality should be understood as a determinant of the spatial range over which structural and electronic perturbations propagate. Point defects primarily perturb local coordination and electronic states, whereas extended and interfacial defects can additionally modify strain, charge-transfer pathways, interfacial electric fields, and the distribution of active sites. The hierarchical classification of these defects, alongside their characteristic synthesis approaches and thermodynamic stability, is summarized in Table 1.

TABLE 1. Classification, thermodynamic descriptors, and representative formation energies of electrocatalytic defects.

Representative literature (Ouyang et al., 2010; Medasani et al., 2015; Mayeshiba and Morgan, 2017; Ma et al., 2019; Wan et al., 2021; Baldassarri et al., 2023; Katnagallu et al., 2023; Li and Walsh, 2025; Wang M. et al., 2025; Wei Z. et al., 2025; Yang T. et al., 2025). Ranges are provided only where a common normalization is meaningful; values should not be interpreted as directly comparable across different defect dimensionalities.

The numerical ranges are representative literature values rather than universal constants. Formation energies depend strongly on host composition, surface orientation, local coordination, defect charge state, chemical potentials, computational functional, supercell size, and treatment of solvation. For extended defects, line or area-normalized energies are used and therefore cannot be directly ranked against point-defect formation energies. Under electrochemical conditions, the relevant stability may instead be governed by a potential- and environment-dependent formation free energy, ΔG_f (U,pH), which can differ substantially from a vacuum-phase DFT, formation energy. Accordingly, the values in this table are intended to illustrate the magnitude and diversity of reported defect thermodynamics rather than establish a universal stability ranking.

2.1.3 Critical perspective on defect classification

Thermodynamic quantities associated with defects vary substantially with crystal structure, surface orientation, local coordination environment, defect charge state, chemical potentials, and computational methodology. Moreover, the appropriate thermodynamic descriptor depends on defect dimensionality: point and substitutional defects are commonly described by formation energies per defect, whereas dislocations and planar defects require line- or area-normalized quantities, and heterointerfaces require system-specific interface or adhesion energies. Consequently, these quantities should not be directly ranked across different defect classes unless their thermodynamic definitions and normalization bases are equivalent. Table 1 therefore focuses on the classification of defect types, their appropriate thermodynamic descriptors, and their representative electronic consequences rather than presenting a single numerical ranking of defect stability.

A definitive conclusion emerging from this classification is that defect concentration alone is not a reliable predictor of catalytic performance. Instead, macroscopic activity is dictated by the precise interplay between operando defect stability, dynamic electronic structure modulation, and intermediate-binding energetics. Consequently, future catalyst design must pivot away from merely maximizing point defect density, prioritizing instead multi-defect synergy and interfacial polarization to deliberately bypass pristine-lattice scaling limitations. This distinction provides an important principle for the subsequent sections: defect density is a structural descriptor, whereas catalytic activity is an emergent property arising from defect identity, local electronic structure, interfacial environment, and reaction-specific energetics.

2.2 Defect formation thermodynamics: Can “as-synthesized” defects survive under applied bias?

While static density functional theory (DFT) frequently predicts enhanced catalytic activity for defect-rich surfaces, the fundamental limitation of such predictions lies in defect thermodynamics. The formation of a charged defect X in charge state q is governed by its formation energy, commonly expressed for a solid-state reference system as:where E def q and E p r i s t i n e are the total energies of the defective and pristine systems, respectively; n _*i* and μ _*i* denote the number and chemical potential of atoms exchanged with the corresponding reservoirs; q is the defect charge state; E _*F* is the Fermi-level position referenced to the valence-band maximum E _*VBM*; ΔV is the electrostatic potential-alignment correction (Van de Walle and Neugebauer, 2004; Menéndez-Proupin and Orellana, 2016; Broberg et al., 2023); and E _*corr* accounts for finite-size and other corrections associated with charged-defect calculations.

Δ E f q X = E def q − E p r i s t i n e − ∑ i n i μ i + q E F + E V B M + Δ V + E c o r r

A critical vulnerability in current electrocatalysis literature is the conflation of ex-situ thermodynamic stability with operando stability. The conventional expression above is a useful solid-state reference, but it does not by itself describe the free energy of a defect at a polarized solid–liquid interface. Under electrochemical conditions, the relevant quantity is more appropriately considered as a defect formation free energy that includes the chemical and electrochemical reservoirs, solvation, entropy, and potential-dependent contributions:where U is the electrode potential and ΔG _{*electrochemical*} represents the contribution associated with exchange of charged species and/or proton-coupled species with the electrochemical reservoirs. The exact potential and pH dependence depends on the defect chemistry and the chosen thermodynamic convention; therefore, an isolated vacuum DFT formation energy should not be interpreted directly as an operando defect stability descriptor. The chemical potential of relevant species (e.g., oxygen or hydrogen) is not necessarily constant; in an electrochemical cell, it is coupled to the surrounding chemical reservoirs and, for charged or proton-coupled processes, to the electrode potential and pH. Therefore, a high concentration of oxygen vacancies engineered under high-temperature argon annealing may become thermodynamically unstable when subjected to electrochemical polarization, potentially undergoing hydroxylation, water-mediated healing, migration, dissolution, or broader surface reconstruction (Wang X. et al., 2026).

Δ G f q U , p H = Δ E f q + Δ G s o l v + Δ G Z P E − T Δ S + Δ G c o r r + Δ G e l e c t r o c h e m i c a l U , p H

This distinction is particularly important for oxygen-deficient oxides. An oxygen vacancy observed before electrolysis may represent the precursor state rather than the catalytically dominant state during turnover. Under aqueous bias, water, hydroxide, adsorbed intermediates, dissolved ions, and the applied potential can collectively alter the relative stability of vacancy-containing and reconstructed configurations. Accordingly, defect stability should be evaluated under conditions that approximate the actual electrochemical environment whenever possible. Furthermore, defect equilibrium is intimately tied to configurational entropy. The equilibrium defect concentration (C _*d*) follows a Boltzmann distribution governed by the Gibbs free energy of formation:

C d ∝ exp − Δ G f k B T

While high temperatures during synthesis drive entropic stabilization (ΔS_f), room-temperature electrocatalysis operates in a regime where the enthalpic penalty (ΔH_f) dominates (Kormányos et al., 2023; Hu et al., 2026; Meng et al., 2026). Consequently, many highly defective electrocatalysts exist in a kinetically trapped, metastable state. The most robust electrocatalyst designs are those where the formation energy of the defect is dynamically stabilized by the specific adsorption of the target reactant (e.g., NO₃ ⁻ or N₂), allowing the defect to persist through the catalytic cycle rather than undergoing irreversible reconstruction. Accordingly, defect formation energy should be interpreted as one component of a broader stability landscape rather than as a standalone predictor of catalytic durability.

2.3 Electronic structure modulation: breaking linear scaling relations

The ultimate purpose of engineering a defect is not geometric, but quantum-mechanical. Point defects and strain fields disrupt the periodic potential of the lattice, triggering orbital reconstruction that dictates adsorption thermodynamics and charge-transfer kinetics. The d-band center (ϵ d) remains the most ubiquitous electronic descriptor. However, a rigid-band shift is often insufficient to describe complex, multi-electron processes. Defects induce local asymmetry that splits the transition metal d-orbitals (e.g., e_g and t_2g orbitals in octahedral fields). By explicitly tuning the occupancy of the e_g orbitals—which point directly toward adsorbates like *OH or *OOH—defects can decouple the binding energies of these species. For example, specific heteroatom doping adjacent to a vacancy can alter the local spin-polarization and spatial orbital overlap, successfully bypassing the universal ≈3.2 eV scaling limit between *OH and *OOH binding free energies (Fang et al., 2023; Zheng et al., 2025; Zhang Y. et al., 2026).

Nevertheless, the d-band center should be treated as a qualitative or semi-quantitative descriptor rather than a universal predictor. Its relevance depends on the electronic configuration of the active center, adsorbate orbital character, surface geometry, and degree of covalency. Other descriptors—including the work function, projected density of states, charge-transfer descriptors, adsorption free energies, local magnetic moments, and interfacial electric fields—may provide greater mechanistic relevance for specific defect–reaction combinations.

This decoupling is exactly what allows advanced defect-engineered catalysts to simultaneously accelerate water oxidation and suppress toxic byproduct formation in wastewater remediation. Additionally, defects profoundly alter the density of states (DOS) and work function (Φ). Vacancy formation frequently introduces mid-gap defect states near the Fermi level, transitioning wide-bandgap insulators (like pristine TiO₂) into highly conductive, heavily doped semiconductors. This localized density of states functions as an electron reservoir, facilitating ultra-fast interfacial charge transfer. However, increased electronic conductivity should not be equated automatically with enhanced catalytic activity. A defect-induced state may facilitate charge transfer while simultaneously over-stabilizing an intermediate, promoting an undesired reaction pathway, or becoming chemically unstable under reaction conditions. Thus, electronic descriptors must ultimately be connected to reaction-specific thermodynamic and kinetic quantities.

Simultaneously, the redistribution of electron density—quantifiable via Bader charge analysis—establishes strong local electric fields at the solid-liquid interface. These localized polarization fields orient polar solvent molecules and reactants, dramatically lowering the kinetic barrier for initial electron transfer (Ren et al., 2023). The mechanistic cascade from structural defect generation to observable macroscopic performance is conceptually illustrated in Figure 1.

2.3.1 Comparative and reaction-specific perspective

Although defect engineering provides a common strategy for modifying interfacial electronic structure, the resulting catalytic consequences are inherently reaction-specific. For NRR, defect-induced charge redistribution, work-function modulation, and changes in metal–nitrogen orbital interactions primarily influence N₂ adsorption and the energetics of hydrogenated N₂ intermediates, with proton/electron transfer and HER competition strongly affecting the overall pathway. In NO₃RR, the same electronic descriptors can regulate nitrate adsorption and sequential N–O bond cleavage, but the reaction additionally involves strongly solvated nitrate species and multiple proton-coupled electron-transfer steps. In electro-Fenton chemistry, by contrast, defect chemistry primarily regulates metal redox cycling and H₂O₂ activation rather than adsorption of a gaseous reactant, whereas PMS-based oxidation can proceed through surface electron transfer and either radical or non-radical pathways. Thus, d-band position, work function, charge redistribution, and adsorption free energy should be regarded as transferable electronic descriptors, rather than universal predictors of activity. Their mechanistic significance depends on the reactant state, solvation environment, dominant intermediates, and rate-determining step of each reaction. This distinction is central to the present review: defect chemistry provides a common language for describing interfacial electronic regulation, but it does not imply identical reaction mechanisms or a universal activity descriptor across energy-conversion and environmental-remediation processes.

2.3.2 Critical synthesis and emerging perspectives

Current consensus firmly establishes that defects act as localized electronic modulators capable of bypassing pristine-lattice scaling relations through orbital reconstruction and charge redistribution. However, a major knowledge gap persists regarding the dynamic evolution of these electronic states under operating conditions. Contradictory reports regarding the stability of oxygen vacancies in aqueous environments suggest that many reported “defect-mediated” mechanisms may actually be the result of transient surface reconstructions or operando metal leaching. Moving forward, the simple correlation between initial defect density and catalytic activity must be challenged. Future advancements will require coupling potential-dependent DFT models with operando spectroscopy to track exactly how a defect’s electronic signature fluctuates in response to applied bias and intermediate binding.

2.4 Defect-to-performance framework: From structural perturbation to reaction-specific catalysis

The preceding discussion establishes a four-stage defect–catalysis relationship:

Defect structure → Operando stability → Electronic/interfacial modulation → Reaction-specific energetics → Catalytic outcome.

At the first level, defect identity, dimensionality, concentration, and local coordination determine the accessible structural configurations. At the second level, applied potential, pH, solvent, adsorbates, and ion composition determine whether these configurations persist or reconstruct under operating conditions. At the third level, the surviving defect states modify local charge density, orbital occupation, work function, spin state, conductivity, and interfacial electric fields. At the fourth level, these changes influence reaction-specific descriptors, including adsorption free energies, proton/electron-transfer energetics, transition-state barriers, and intermediate stabilization. The resulting catalytic activity and selectivity therefore emerge from the coupling of structural, electronic, thermodynamic, and kinetic effects rather than from defect concentration alone. This framework also establishes an important boundary for cross-reaction comparison. NRR, NO₃RR, electro-Fenton chemistry, persulfate activation, and pollutant electroreduction may share defect-mediated electronic and interfacial principles, but their dominant descriptors and elementary reaction steps are not interchangeable. The subsequent sections therefore use the common defect framework while retaining reaction-specific mechanistic analysis.

3 Defect-controlled electrocatalytic mechanisms and operando interrogation

The catalytic influence of defects extends far beyond simply increasing the geometric density of active sites. While literature frequently correlates increased defect concentration with enhanced performance, the fundamental origin of this enhancement lies in the ability of defects to dynamically regulate charge-transfer kinetics, alter adsorption thermodynamics, and reshape multi-proton/multi-electron energy landscapes (Jia et al., 2019; Huang P. et al., 2022; Mao et al., 2022; Zhang et al., 2025d). Many of the most critical reactions in sustainable energy and environmental remediation—such as the nitrogen reduction reaction (NRR), nitrate reduction reaction (NO₃RR), and the generation of reactive oxygen species (ROS) from persulfate activation—are kinetically bottlenecked by intermediate stabilization and competing parasitic reactions. Defects function as atomic-scale electronic regulators, bypassing pristine-lattice limitations to exert disproportionate control over the overall reaction pathway.

3.1 Interfacial charge transfer dynamics: How do defects act as electron reservoirs?

Efficient electron transfer from the bulk catalyst to the solid-liquid interface is a prerequisite for high catalytic turnover. However, in semiconducting metal oxides and purely carbonaceous supports, intrinsic conductivity is often too poor to sustain industrial-level current densities. Defects fundamentally alter this limitation by generating localized mid-gap states that act as electron reservoirs (Nakayama, 2025; Shi et al., 2025). Anion vacancies provide a specific example of this phenomenon. The removal of a lattice oxygen atom frequently leaves behind excess electrons that localize on adjacent transition metal centers (e.g., reducing Ti⁴⁺ to Ti³⁺ in TiO₂). This localized electron accumulation generates regions of high charge density perfectly suited to activate electron-deficient molecules. In both NRR and advanced oxidation processes, these electron-rich centers act as direct donors. For example, the transfer of an electron from a localized vacancy site into the antibonding π* orbital of N₂ weakens the highly stable N≡N bond.

In exact parallel, identical oxygen vacancies act as direct electron donors for the homolytic cleavage of the peroxide bond in peroxymonosulfate (PMS), generating the highly oxidative sulfate radical (S O 4 · −) for wastewater remediation (Lim et al., 2019; Zhao et al., 2024; Cao S. et al., 2026).

Furthermore, interfacial defects (such as phase boundaries in heterojunctions) generate spontaneous internal electric fields. When two distinct crystallographic phases interact, their mismatched work functions drive interfacial charge redistribution until Fermi-level equilibration occurs. The resulting built-in electric field establishes a directional charge-transfer pathway, effectively suppressing electron-hole recombination and accelerating electron flux directly to the highly localized active sites. Therefore, enhanced conductivity and localized adsorption are not independent variables; they are coupled manifestations of the same defect-induced charge redistribution.

3.2 Adsorption thermodynamics: bypassing the sabatier constraints

Following rapid electron transfer, the optimization of intermediate binding energies is paramount. According to the Sabatier principle, an ideal catalyst must bind a reactant strongly enough to activate it, but weakly enough to allow product desorption. Pristine catalyst surfaces are fundamentally hindered by linear scaling relations, which enforce rigid correlations between the adsorption free energies (ΔG) of sequential intermediates. Defects provide a spatial and electronic asymmetry capable of breaking these constraints [10, 36]. In NRR, the primary kinetic bottleneck is the initial adsorption and activation of non-polar N₂ (*N₂ → *N₂H). Defect sites—particularly dual-metal vacancies or single-atom dopants adjacent to a vacancy—alter the local d-band configuration to allow simultaneous σ-donation (from N₂ to the empty metal orbital) and π-backdonation (from the metal d-orbital into the N₂ π* orbital). This synergistic ‘push-pull’ mechanism modifies the thermodynamics of N₂ activation, bringing the free energy of *N₂H formation closer to thermoneutrality (Jin et al., 2020; Wang M. et al., 2023; Zou et al., 2024; Long et al., 2026).

Similarly, in oxygen electrocatalysis and electro-Fenton systems, pristine catalysts are plagued by a universal energy gap of approximately 3.2 eV between *OH and *OOH binding energies. By introducing a heteroatom dopant that induces a localized spin-polarization effect, the binding of *OOH can be selectively stabilized via hydrogen bonding with the dopant, without equally strengthening the binding of the preceding *OH intermediate (Zhou et al., 2023; Fu et al., 2026; Li et al., 2026). The thermodynamic consequences of this defect-induced electronic tuning are conceptually illustrated in Figure 2, which compares the reaction free-energy landscapes of pristine and defect-engineered surfaces and highlights how localized electronic perturbations alter intermediate energetics and can alleviate thermodynamic limitations associated with conventional scaling relationships.

3.3 Reaction pathway engineering and selectivity control

Altering the free energy of a single intermediate rarely improves catalysis if it inadvertently accelerates a competing reaction. The true hallmark of effective defect engineering is reaction pathway engineering—the ability to selectively shift the rate-determining step (RDS) and suppress parasitic side reactions (Wang D. et al., 2026). The hydrogen evolution reaction (HER) is the universal parasite in aqueous cathodic electrocatalysis. In standard aqueous electrolytes, protons are vastly more mobile and abundant than dissolved N₂ or NO₃ ⁻. If a defect simply lowers the work function uniformly, the surface will be immediately poisoned by strongly bound hydrogen (H), funneling all Faradaic efficiency into H₂ gas.

To achieve high ammonia selectivity in both NRR and NO₃RR, defects must exhibit differential stabilization. For instance, introducing a localized tensile strain around an oxygen vacancy can narrow the transition metal d-band just enough to ensure the binding of NO₃ is highly exothermic, while making the adsorption of *H endothermic. This thermodynamic blockade essentially shuts down the HER pathway, ensuring that the complex eight-electron reduction of NO₃ ⁻ to NH₃ proceeds unhindered (Wu et al., 2023; Zhang M. et al., 2023; Dildar et al., 2026). Furthermore, the stabilization of specific partially reduced nitrogenous intermediates (e.g., *NO₂, *NO, *NHO) by undercoordinated defect sites directly dictates the final product distribution. If *NO is bound too weakly at a defect, it will desorb as toxic NO gas; if it is bound strongly within the protective coordination sphere of a metal vacancy, it continues down the hydrogenation cascade toward NH₃. This pathway redirection confirms that defects do not merely lower overall energy barriers; they actively steer the reaction mechanism through complex multi-electron networks.

3.4 Operando interrogation: the dynamic reality of defects

A fundamental and persistent flaw in much of the current electrocatalysis literature is the assumption that defects synthesized ex-situ (e.g., in a vacuum furnace) are identical to the active sites functioning under applied electrochemical bias. This assumption is frequently incorrect. Defect sites are intrinsically metastable. Under realistic operating conditions—involving high overpotentials, aggressive solvent interactions, and concentrated ion fluxes—catalysts continuously undergo dynamic reconstruction (Mosquera-Lois et al., 2024; Ding and Fu, 2026). If a defect heals via irreversible oxygen incorporation during anodic polarization, or migrates and agglomerates during cathodic polarization, the catalyst characterized before the reaction no longer exists. Consequently, static characterization cannot establish reliable structure–activity relationships. Establishing the true identity of the active site requires advanced operando interrogation. Importantly, however, no individual operando technique can establish the complete causal chain from defect existence to catalytic participation. A rigorous assignment of defect-mediated activity therefore requires the convergence of structural, electronic, chemical, kinetic, and theoretical evidence obtained under reaction-relevant conditions.

A useful evidence hierarchy can be established by separating six experimentally distinct questions: (i) does the defect exist, (ii) where is it located, (iii) at what concentration or population does it occur, (iv) what is its oxidation state and local electronic configuration, (v) does its structure dynamically reconstruct under applied potential, and (vi) does the defect itself participate in the rate-determining catalytic pathway? These questions should not be treated as interchangeable. For example, observation of a vacancy before electrolysis establishes its initial presence but does not demonstrate that the vacancy survives under bias; likewise, a potential-dependent change in metal oxidation state does not by itself prove that the corresponding site is catalytically active.

Accordingly, operando characterization should be interpreted as a complementary evidence matrix rather than as a collection of independent “proofs.” Operando X-ray absorption spectroscopy (XAS) is particularly valuable for tracking average or element-specific changes in oxidation state, coordination number, and local structure, but its spatial resolution is generally insufficient to assign a signal uniquely to an individual vacancy or defect unless combined with appropriate structural models. Operando Raman and FTIR spectroscopy can provide important information on surface species and transient intermediates, but the observation of a vibrational band alone does not establish that the species is bound specifically to a defect site. Spatially resolved techniques, such as scanning probe or electron-based methods where experimentally feasible, can provide complementary information on defect distribution, whereas X-ray photoelectron spectroscopy and related spectroscopies can constrain chemical states but should be interpreted cautiously when measurements are performed outside the actual electrochemical environment.

Importantly, the evidentiary strength of these measurements depends on the specific mechanistic question being tested. Defect existence or chemical-state changes should therefore be distinguished from spatial assignment, dynamic reconstruction, and catalytic participation. No single technique can generally establish all of these attributes simultaneously; in particular, spectroscopic detection of a defect or intermediate does not by itself demonstrate that the species constitutes the rate-determining active site. Robust causal assignment instead requires convergence among complementary operando measurements, controlled defect-density/activity relationships, isotopic experiments where appropriate, and potential-dependent theoretical modeling.

Table 2 compares conventional ex-situ techniques against state-of-the-art dynamic methodologies. For instance, operando X-ray absorption spectroscopy (XAS) can track the dynamic shifting of metal oxidation states in real-time, while operando Raman and FTIR spectroscopies provide valuable vibrational signatures of surface-bound species and transient reaction intermediates, including potentially elusive species such as *N₂H or *NHO; however, assigning these intermediates specifically to defect sites requires corroboration from spatially resolved characterization, defect-density/activity relationships, isotopic labeling, or theory (Fabbri et al., 2017; Huang Y.-C. et al., 2022; Surya et al., 2025; Yoon et al., 2025; Lee et al., 2026). The distinction between defect identification and active-site identification is particularly important. A defect can be experimentally confirmed without being the dominant catalytic site, while an apparent correlation between defect concentration and activity can arise from concurrent changes in surface area, conductivity, morphology, oxidation state, or electrolyte adsorption. Consequently, causal attribution should ideally satisfy multiple independent criteria: (1) the defect is detected under operating conditions; (2) its population changes systematically with controlled defect density; (3) the corresponding catalytic activity or selectivity changes proportionally within an appropriate range; (4) isotopic or spectroscopic evidence connects the defect-associated interface to the relevant reaction intermediate; and (5) potential-dependent theoretical calculations reproduce the experimentally observed trends. This convergent-evidence approach provides a substantially stronger basis for active-site assignment than relying on a single spectroscopic signature or post-reaction characterization.

TABLE 2. Advanced characterization techniques: Ex-situ vs. operando interrogation of dynamic defects.

Note on Ex-situ Limitations: Readers must exercise extreme caution when evaluating literature that correlates electrocatalytic performance solely with ex-situ XPS or standard TEM data. Defects are highly metastable (Wang et al., 2020; Zhang Y. et al., 2023; Tehranchi et al., 2024). A vacancy characterized under ultra-high vacuum at room temperature is thermodynamically and kinetically distinct from a vacancy subjected to an aqueous electrolyte and a −0.8 V overpotential. Without operando validation, it is impossible to determine whether a claimed defect acts as the true catalytic center or simply triggers an irreversible surface reconstruction.

These experimental insights are only meaningful when rigorously coupled with theoretical models. However, standard static Density Functional Theory (DFT) in a vacuum is insufficient to model an aqueous, polarized interface. The frontier of mechanistic validation now requires ab initio molecular dynamics (AIMD) and constant-potential microkinetic modeling that explicitly includes solvent interactions. In particular, theoretical descriptions should account, where relevant, for explicit or implicit solvation, applied potential, electrolyte composition, interfacial electric fields, adsorbate coverage, and potential-dependent defect populations. These factors can substantially modify defect formation energetics, charge distribution, intermediate adsorption, and reaction barriers relative to predictions obtained from an isolated vacuum surface.

Only by capturing the dynamic evolution of defects as they continuously form, migrate, and participate in intermediate binding can the field move beyond correlation toward true, predictive causation (Liu et al., 2022; Zhao Z. et al., 2023; Jia et al., 2026). A particularly important emerging concept is that the “active site” should not necessarily be regarded as a static crystallographic defect. Under operating conditions, the catalytically relevant configuration may instead be a dynamically populated defect–adsorbate–solvent ensemble whose structure and electronic state depend on potential and reaction environment. Thus, an oxygen vacancy observed after synthesis may transform into a hydroxylated, reconstructed, or adsorbate-stabilized configuration during electrolysis while retaining a mechanistic relationship to the initially engineered defect. This distinction reconciles apparently contradictory reports in which nominally identical defect-rich catalysts exhibit different activities under different electrolytes or applied potentials.

The resulting evidence hierarchy can therefore be summarized as follows: defect existence establishes structural plausibility; defect localization establishes spatial assignment; spectroscopic and electronic measurements establish chemical-state changes; potential-dependent measurements establish dynamic reconstruction; and only convergent kinetic, isotopic, spectroscopic, and theoretical evidence can support causal active-site attribution. This framework is particularly important for NRR and NO₃RR, where trace nitrogen contamination can mimic catalytic ammonia production, and for advanced oxidation processes, where radical or non-radical pathways may coexist and evolve with surface reconstruction.

Overall, operando interrogation should therefore move from the question “Is the engineered defect present?” to the more demanding questions “What defect configuration exists under reaction conditions?”, “How does its population evolve with potential?”, and “Does this dynamically populated configuration control the observed reaction pathway and rate?” Addressing these questions will be essential for transforming defect engineering from a predominantly structure–performance correlation into a predictive structure–mechanism framework. The resulting evidence hierarchy for identifying dynamic defect active sites is schematically summarized in Figure 3, which connects defect existence and localization with electronic-state evolution, intermediate identification, dynamic reconstruction, and ultimately causal catalytic participation.

3.4.1 Critical perspective for table 2

While ex-situ methods provide baseline structural confirmation, operando techniques definitively reveal that active defect populations evolve dynamically under applied potential. The true working catalyst frequently bears little resemblance to the as-synthesized material. Consequently, establishing robust, causal structure–activity relationships requires abandoning isolated ex-situ claims in favor of multi-modal operando interrogation (e.g., combining XAS for electronic evolution with Raman for intermediate tracking) validated against explicit-solvent theoretical models.

4 Defect-engineered electrocatalysts for green ammonia synthesis

The decarbonization of ammonia production represents one of the most significant challenges in sustainable chemical manufacturing. Although the Haber–Bosch process remains the industrial benchmark, its extreme thermodynamic requirements and reliance on fossil-derived hydrogen are fundamentally incompatible with carbon neutrality. Electrocatalytic ammonia synthesis offers a zero-carbon alternative, utilizing renewable electricity to drive the nitrogen reduction reaction (NRR) or the nitrate reduction reaction (NO₃RR) under ambient conditions (Fang et al., 2024; Zhang B. et al., 2024). However, both pathways require navigating complex, multi-proton/multi-electron transfer networks while strictly controlling adsorption thermodynamics to suppress the ubiquitous hydrogen evolution reaction (HER). In this context, defect engineering has emerged not merely as a method to increase active site density, but as the primary lever for manipulating the localized electronic structures required to break conventional adsorption scaling relations.

4.1 NRR kinetics: overcoming the 941 kJ/mol N≡N activation barrier

The fundamental bottleneck of ambient NRR is the initial adsorption and activation of the highly inert, non-polar N≡N triple bond. On pristine metal or metal-oxide surfaces, N₂ adsorption is thermodynamically unfavorable due to a lack of highly localized, high-energy electronic states capable of hybridizing with the adsorbate. Defect engineering circumvents this limitation by creating highly asymmetric coordination environments that facilitate a “push-pull” electronic interaction. Oxygen vacancies in transition-metal oxides (e.g., NiMnO₃, CeO₂, and Fe₃O₄) generate electron-rich metal centers with upward-shifted d-band positions (Li et al., 2022; Yang X. et al., 2023). These localized defect states spontaneously donate electron density into the antibonding π* orbitals of adsorbed N₂, significantly weakening the triple bond. When integrated into rationally designed heterojunctions, such as Cu-[OX]-Ce interfaces or TiO_2-x supported on covalent organic frameworks, the interfacial electric field restricts proton accumulation. This thermodynamically favors *N₂ binding while raising the activation barrier for *H adsorption (Jing et al., 2022; Mishra et al., 2024).

This defect activation extends beyond oxides; sulfur vacancies in metal sulfides (e.g., ZnS, MoS₂, and NiCo₂S₄) induce profound p-band and d-band reconstructions, narrowing the bandgap to accelerate interfacial electron transfer directly into the N₂ adsorbate (Fei et al., 2022; Mushtaq et al., 2024; Chen et al., 2026). To maximize atomic utilization, recent advances have coupled single-metal atoms with specific defect sites. Anchoring single Ru atoms near oxygen vacancies on SnO_2-x, or embedding them within 5- and 7-membered topological carbon rings, physically distorts the coordination sphere (Han et al., 2023; Zheng et al., 2023). This precise defect-induced strain can stretch the N≡N bond from its native length to 1.146 Å, drastically lowering the free energy required for the first proton-coupled electron transfer step (*N₂ → *N₂H) (Han et al., 2023).

Despite these elegant mechanistic rationalizations, ambient NRR remains plagued by low ammonia yields and HER dominance. While defect engineering can lower the N₂ activation barrier, the thermodynamic stability of these specific defect sites—such as isolated S-vacancies or topological carbon defects—under continuous cathodic polarization remains highly questionable. Consequently, state-of-the-art NRR performance (summarized in Table 3) must increasingly be validated through rigorous isotopic labeling and operando stability tracking to definitively confirm that reported activity originates from true catalytic turnover rather than defect degradation or environmental contamination.

TABLE 3. Defect-engineered catalysts for N₂ reduction: performance metrics.

NRR Data Integrity and Evidence Hierarchy: The reduction of N₂ to NH₃ is susceptible to false-positive signals from nitrogen-containing contaminants, catalyst/electrolyte impurities, and feed-gas contamination. Accordingly, Table 3 audits 15N₂ isotope labeling, Ar-saturated blanks, open-circuit/dark controls, catalyst/electrolyte blanks, feed-gas purification, and contaminant analysis. Studies with more complete validation protocols provide stronger evidentiary support, whereas studies lacking isotope confirmation or essential blank controls should be interpreted cautiously. Where sufficient production-rate data were available, j N H 3 was calculated using j N H 3 = 3 F m ˙ N H 3 / M N H 3; NR = not reported.

Data-comparison note: NH₃ productivity, FE, and j N H 3 are reported under study-specific conditions and should not be directly ranked when electrode area, catalyst loading, electrolyte, potential, or test duration differ. DFT-derived adsorption energies and reaction barriers are reported as mechanistic interpretations rather than independent proof of the experimentally active site.

4.1.1 Critical perspective for table 3

A comprehensive analysis of state-of-the-art nitrogenous reduction systems reveals that catalytic performance is no longer limited by the mere generation of active sites, but by the precise spatial and electronic management of competitive proton reduction. As demonstrated across these diverse architectures—ranging from p-block single atoms (Sb) to complex dynamic interfaces—effective defect engineering systematically breaks pristine scaling relations. By either kinetically trapping protons, creating electrostatic repulsive fields, or altering d/p-band overlap to exclusively stabilize *N₂H over *H, multi-defect and dynamically generated vacancy systems now represent the undisputed frontier for achieving commercially relevant Faradaic efficiencies in ambient N₂ activation.

4.2 NO₃RR: navigating complex multi-electron pathways and selectivity control

Given the severe kinetic constraints of NRR, the electroreduction of nitrate (NO₃RR) has rapidly emerged as a more technologically viable route for green ammonia synthesis. Because nitrate (NO₃ ⁻) is already chemically activated, the primary electrocatalytic challenge shifts from bond activation to intermediate management. The complete reduction of NO₃ ⁻ to NH₃ involves eight electrons and nine protons. Therefore, catalytic selectivity is dictated by the ability of a defect site to continuously stabilize partially reduced intermediates (*NO₂, *NO, *NHO) without excessively binding them to the point of surface poisoning.

Oxygen vacancies in oxides, such as Cu/Fe₂O₃, Fe₂TiO₅, and amorphous RuO₂, excel at initiating this cascade by providing electrophilic centers that strongly bind the NO₃ ⁻ oxyanion, facilitating rapid deoxygenation via localized charge transfer (Wang Y. et al., 2022; Du et al., 2023; Yang et al., 2024). Furthermore, designing heterostructures like Cu/Cu₂O with abundant interfacial oxygen vacancies ensures a continuous supply of atomic hydrogen from adjacent domains to rapidly hydrogenate the critical *NOH intermediate (Shi et al., 2024). However, the assumption that higher defect density inherently yields better performance is a dangerous oversimplification. Recent mechanistic studies, such as those on CoO_x/Fe₃O₄, reveal a critical threshold where excessive oxygen vacancy concentrations trap the *NO intermediate too strongly. This entirely arrests the hydrogenation process, leading to high selectivity for the toxic NO₂ ⁻ byproduct rather than NH₃ (Zhou et al., 2025), underscoring the necessity of precision defect optimization over mere defect maximization.

To bypass the linear scaling relations that typically link the binding energies of *NO and *NHO, advanced non-oxygen defect systems have been engineered. Nitrogen vacancies in Cu₃N/GDY matrices (Zhang et al., 2024f), phosphorus vacancies in CoP-Ni₂P (Gao et al., 2023), and sulfur vacancies coupled with Cu-doping in SnS₂ (Li et al., 2023b) or Ni₃Co₆S₈ (Tao et al., 2023) create highly asymmetric charge distributions. Similarly, in MoS₂, carefully engineered sulfur vacancies explicitly couple H⁺ generation with adjacent N species, effectively turning the parasitic HER mechanism into a localized proton-supply engine for NH₃ synthesis (Wang J. et al., 2023). The pinnacle of intermediate tuning is achieved through single-atom and defect synergies. Introducing single Fe atoms into TiO₂ directly triggers dense interfacial oxygen vacancies that suppress proton reduction (Wang Z. et al., 2023), while precisely engineered Cu(I)-N₃C₁ single-atom sites break standard structural symmetry (Xue et al., 2022).

Unlike conventional symmetric Cu-N₄ sites, the asymmetric N₃C₁ coordination localizes charge to optimally balance the adsorption energies of both *NO₃ and *H, thermodynamically steering the eight-electron reduction exclusively toward NH₃. A comparable symmetry-breaking strategy relies on topological line defects (e.g., 5-8-5 membered rings) in Fe-N-C catalysts, where geometric distortion induces an asymmetric charge distribution on the Fe-N₄ center, fundamentally altering its interaction with NO₃ ⁻ (Zhu et al., 2024). Furthermore, filling specific oxygen vacancies with sulfur species (e.g., S-Co₃O₄) can dramatically reduce the free energy change of *H formation, accelerating the reduction of multiple nitrogen oxyanions (Niu et al., 2023). The correlation between these precise defect architectures and NO₃RR benchmarking metrics is detailed in Table 4.

TABLE 4. Benchmarks in defect-driven NO₃RR.

NO₃RR Performance and Data-Comparison Note: The electroreduction of NO₃ ⁻to NH₃ is an 8-electron transfer process. Where sufficient production-rate data were available, ammonia partial current density (j _NH3) was calculated using the conversion factor 1 μg h^-1 NH₃ ≈ 0.0126 mA; NR, not reported. NH₃ productivity, Faradaic Efficiency (FE), and partial current densities are reported under study-specific conditions (varying feed concentrations, catalyst loadings, and cell configurations) and should not be directly ranked. Furthermore, while the listed studies report high selectivity, claims of commercial relevance remain premature without achieving industrial-level current densities (>100 mA cm^-2) and demonstrating ultra-long-term stability under practical operational conditions.

4.2.1 Critical perspective for table 4

A comprehensive analysis of state-of-the-art NO₃RR systems reveals that superior Faradaic efficiency is intrinsically linked to mastery over intermediate management. A defining breakthrough in this domain is the deliberate breaking of active-site symmetry. Whether achieved through heteroatom doping (e.g., P-N-C) or through non-doping topological line defects (e.g., 5-8-5 carbon rings), asymmetric sites generate a localized “push-pull” charge polarization that makes initial nitrate adsorption highly favorable without irreversibly trapping final products. Furthermore, spatial isolation in single-atom catalysts inherently eliminates competitive N-N coupling, definitively choking off N₂ gas evolution. Crucially, defect engineering in NO₃RR has evolved to hijack the HER mechanism rather than merely outcompete it; sulfur-doped and vacancy-rich heterostructures deliberately accelerate the Volmer step to generate *H, but structurally poison the Heyrovsky step. This “H-trapping” paradigm confines protons to the catalyst surface, forcing rapid hydrogenation of adjacent nitrogenous species. Finally, while atomic oxygen vacancies serve as powerful thermodynamic sinks to spontaneously cleave N-O bonds, the emergence of “multiscale” architectures emphasizes that incorporating nanoscale porosity is equally vital to overcome the severe mass-transport limitations of bulky nitrate ions.

4.3 Breaking the OER bottleneck: enabling anodic coupling

The industrial translation of cathodic ammonia synthesis, encompassing both NRR and NO₃RR, is severely constrained by the energetic demands of the conventional anodic half-reaction, the oxygen evolution reaction (OER). Although the thermodynamic equilibrium potential for OER is 1.23 V vs. RHE under standard conditions, practical OER operation requires an additional kinetic overpotential, making the anodic reaction a major contributor to the overall cell voltage. Importantly, the actual energy requirement of a coupled electrolyzer should therefore be evaluated using the measured full-cell voltage at a defined current density rather than by assigning the 1.23 V thermodynamic potential itself as an OER overpotential. An emerging paradigm in electrocatalysis is replacing OER with thermodynamically and kinetically more favorable organic oxidation reactions, coupling ammonia production with simultaneous wastewater treatment or value-added chemical synthesis (Li J. et al., 2023). Such anodic replacement can reduce the electrical energy demand while simultaneously converting an otherwise sacrificial anodic process into a useful chemical transformation.

Defect-engineered catalysts are inherently suited for these dual-function tandem systems (Wu Z. et al., 2024). For example, replacing OER with glycerol oxidation utilizing carbon-supported CoP catalysts enables a coupled NO₃RR–glycerol oxidation system that operates at 1.8 V cell voltage, simultaneously producing NH₃ and value-added formate (Li J. et al., 2023). The reported system achieved an NH₃ production rate of 15.2 mg h^-1 cm^-2 together with a formate production rate of 111.4 mg h^-1 cm^-2 at 1.8 V, illustrating the potential benefit of replacing an energetically demanding anodic reaction with a value-generating oxidation pathway (Li J. et al., 2023). Similarly, coupling NO₃RR with anodic formaldehyde oxidation over Cu₂O was reported to require a reported cell voltage of −0.19 V at 10 mA cm^-2, while achieving a high Faradaic efficiency of 99.77% for NO₃ ⁻ conversion to NH₃ and simultaneously producing formic acid from formaldehyde oxidation (Xiao et al., 2023). Because this reported value is a measured cell voltage for the coupled two-electrode configuration rather than an equilibrium thermodynamic potential, it should not be interpreted as a universal thermodynamic requirement for NO₃RR or as a directly comparable energy metric across different electrolyzer configurations.

Beyond simple voltage reduction, coupled architectures can also create process-level synergies by removing reaction intermediates that would otherwise accumulate and inhibit electrode performance. During anodic urea electrooxidation, for example, nitrite accumulation can inhibit the anodic process and reduce nitrogen selectivity. A recent cathode-assisted configuration coupled a MoNiOOH anode with a Cu/g-C₃N₄ cathode, where anodically generated NO₂ ⁻ was selectively reduced to N₂ at the cathode. This cathode–anode synergy suppressed nitrite accumulation and accelerated urea electrooxidation, demonstrating how spatially separated reactions can be coupled to alleviate intermediate inhibition (Yu et al., 2026). These examples also highlight an important distinction for defect-engineered systems: lowering the full-cell voltage does not, by itself, establish that defects are responsible for the improved energy efficiency. A causal defect–performance relationship requires the anodic and cathodic Faradaic efficiencies, current density, applied cell voltage, product selectivity, stability, and, where relevant, carbon and nitrogen balances to be evaluated under comparable operating conditions. Accordingly, comparisons of OER replacement should prioritize experimentally measured full-cell energy consumption rather than isolated half-cell potentials.

The future of defect-driven electrocatalysis therefore lies not only in optimizing cathodic active sites in isolation but also in designing integrated electrochemical architectures in which defect-regulated cathodic and anodic processes are thermodynamically and kinetically matched. Such systems can simultaneously address ammonia synthesis, pollutant removal, and value-added chemical production, but their practical significance must ultimately be judged using full-cell voltage, current density, Faradaic efficiency, product yield, stability, and energy consumption rather than single-electrode activity metrics.

5 Defect-driven advanced wastewater remediation

The increasing complexity of wastewater matrices has exposed the fundamental limitations of conventional physicochemical treatment technologies. Modern effluents are laden with recalcitrant organics—including per- and polyfluoroalkyl substances (PFAS), advanced pharmaceuticals, and endocrine disruptors—that resist traditional biodegradation. Consequently, advanced electrochemical remediation has pivoted from a strategy of simple pollutant transfer to one of complete, energy-efficient mineralization. However, the universal bottleneck in generating reactive oxygen species (ROS) or activating complex oxidants is identical to the challenge faced in green ammonia synthesis: pristine catalyst surfaces are bound by linear scaling relations that limit reaction kinetics and squander applied energy on parasitic side reactions (such as the four-electron oxygen evolution reaction, OER).Within this context, defect engineering has emerged as the definitive lever for bridging environmental remediation and fundamental electrocatalysis (Chen et al., 2023; Sun B. C. et al., 2023; Zhang M. et al., 2024; Murtaza et al., 2026). Structural imperfections—vacancies, heteroatom dopants, and synergistic single-atom sites—function not merely as high-energy binding sites, but as atomic-scale electronic regulators. By explicitly modulating local charge distributions and shifting d-band and p-band centers, defects dictate whether an oxidant is cleaved homolytically or heterolytically, steering the degradation mechanism between radical-mediated oxidation, highly selective non-radical pathways, and targeted pollutant electroreduction.

5.1 Electrochemical oxidation and electro-fenton: How do defects dictate ROS generation and Fe redox cycling?

Anodic electrochemical oxidation relies on the generation of highly reactive, surface-bound hydroxyl radicals (OH^•) from water. On pristine metal oxides, the thermodynamic barrier for water dissociation (H₂O→OH + H⁺ + e⁻) is often prohibitively high. Introducing oxygen vacancies into transition metal oxides (e.g., substoichiometric Ti₄O₇, CeO₂, or La-doped SnOx) creates localized, electron-deficient metal centers (Yang C. et al., 2022; Yang M. et al., 2025; Zhuo et al., 2023). These electrophilic sites drastically lower the adsorption free energy (ΔG) for water molecules while increasing the bulk conductivity of the anode. When further engineered into heterojunctions—such as N-doped graphene encapsulated on Ti₄O₇ or Ru-coupled TiO₂/SnO₂ interfaces—the resulting localized electric fields aggressively drive the complete mineralization of extremely recalcitrant molecules like GenX and glyphosate with vastly reduced energy consumption (He et al., 2024; Liang et al., 2025).

Similarly, specific topological defects in carbonaceous anodes (e.g., B,N-codoped carbon nanowalls lacking pyridinic-N) can directly hybridize with organic pollutants, allowing for direct anodic electro-oxidation without intermediate ROS generation (Pierpaoli et al., 2024).The importance of defects becomes even more pronounced in cathodic electro-Fenton (EF) systems, where catalytic performance hinges on two sequential bottlenecks: the selective two-electron oxygen reduction reaction (2e⁻ ORR) to generate H₂O₂, and the continuous regeneration of Fe²⁺ from Fe³⁺ to trigger Fenton radical generation. Pristine carbon cathodes typically suffer from poor 2e⁻ ORR selectivity due to the over-stabilization of the *OOH intermediate, driving the parasitic four-electron reduction to water. Defect engineering elegantly breaks this scaling relation, paralleling the thermodynamic principles utilized in ammonia synthesis.

The introduction of highly asymmetric defects—such as N-S codoping in carbon substrates, atomically dispersed Co sites constrained within defective carbon nanotubes, or asymmetric Fe-O-Ce oxygen vacancies—localizes electron density precisely where it is needed (Yang et al., 2023b; Cao C. et al., 2026; Cui et al., 2026; Guo et al., 2026). This electronic asymmetry ensures that *OOH is bound optimally for protonation to H₂O₂ while preventing the subsequent O–O bond cleavage that leads to water formation. Concurrently, incorporating mixed-valence defect sites in transition metal catalysts (e.g., Cu-doped LaCoO₃ perovskites or SO₄ ^2--modified Ce_xZr_1-xO₂) establishes rapid, localized electron-transfer channels that drastically accelerate the sluggish Fe³⁺/Fe²⁺ redox cycle, sustaining high radical yields without the need for continuous iron addition (Xie et al., 2022; Hu et al., 2024; Liu X. et al., 2025).

5.2 Persulfate activation: Can defects steer the selectivity between radical and non-radical pathways?

Beyond the electro-Fenton process, the electrochemical activation of peroxymonosulfate (PMS) and peroxydisulfate (PDS) provides a potent strategy for eliminating persistent pharmaceuticals. Traditionally, PMS/PDS activation has been viewed strictly through the lens of free-radical generation, where defect sites (e.g., oxygen vacancies in Co₃O₄ or sulfur vacancies in MoS₂) donate localized electrons to homolytically cleave the peroxide bond, yielding highly oxidative sulfate radicals (SO₄ ^•-) (Wang G. et al., 2022; Geng et al., 2024; Longfei et al., 2024; Li et al., 2025). However, a critical synthesis of recent literature reveals a profound limitation to the radical-centric approach: free radicals react indiscriminately. In real-world wastewater containing high concentrations of background organic matter, chloride ions, and carbonates, radicals are rapidly scavenged, severely suppressing target pollutant degradation.

The true pinnacle of defect-engineered persulfate activation lies in challenging this paradigm and harnessing non-radical pathways. Precise electronic manipulation at defect sites can fundamentally alter the PMS activation mechanism. For instance, regulating the second-shell coordination environment of single-atom Fe in a boron-doped carbon matrix drives directional electron transfer from the PMS molecule directly to the electrode (Wang Z. et al., 2025). This suppresses radical formation entirely, instead selectively generating singlet oxygen (¹O₂)—a highly targeted oxidant that attacks electron-rich organic pollutants while ignoring background scavengers (Zhao et al., 2024; Zhang S. et al., 2025; Zhang et al., 2026 C.). Similarly, oxygen vacancies in Fe₃O₄@N-C composites can bypass the radical cascade, utilizing a surface-confined charge-transfer complex that oxidizes nitroso-containing organics selectively across a wide pH range (Zhang Q. et al., 2024). This highlights a crucial design rule: the ultimate goal of defect engineering in advanced oxidation is not merely maximizing ROS density, but deliberately tuning electronic orbital overlap to secure mechanistic selectivity.

5.3 Electroreduction of priority pollutants: How do defect kinetics parallel cathodic ammonia synthesis?

While oxidative systems dominate wastewater literature, electroreductive technologies—which include the detoxification of heavy metals and the dechlorination of toxic organics—share identical mechanistic principles with cathodic ammonia synthesis (NRR and NO₃RR). The reduction of highly toxic, mobile Cr(VI) to benign Cr(III), or the recovery of radioactive U(VI), requires the catalyst to securely adsorb the oxoanion, rapidly donate electrons, and suppress the ubiquitous hydrogen evolution reaction (HER). Defect sites function as ideal localized electron donors. The introduction of engineered vacancies directly alters the density of states near the Fermi level, creating a thermodynamic sink for heavy metal oxyanions. For example, oxygen-vacancy-enriched CoOx or TiO₂ cathodes drastically reduce the adsorption energy of the uranyl ion (UO₂ ²⁺). The defect acts as an electron superhighway, accelerating charge transfer from the bulk semiconductor directly into the d-orbitals of the uranium species, enabling highly selective, low-potential uranium extraction from complex seawater matrices (Wang et al., 2024; Zhang Q. et al., 2025).

Furthermore, coupling defect engineering with dynamic operational modes—such as applying pulsed potentials to Fe²⁺-doped NiFe layered double hydroxides—allows for continuous, in situ active site regeneration, preventing the catalyst poisoning that typically plagues continuous cathodic reduction (Wang Z. et al., 2022). When advanced multi-defect carbon cathodes (e.g., dual-site N-doped, carbon-deficient structures coupled with uncoordinated Mo sites) are utilized, it becomes possible to seamlessly orchestrate the simultaneous reduction of Cr(VI) and oxidation of organic pollutants within a single, highly synergistic cell (Wu Y. et al., 2024).

5.4 Structure–activity relationships: a unified environmental–energy framework

Although wastewater remediation encompasses a vast array of target molecules and reaction conditions, a unified structure–activity framework unequivocally emerges when viewed through the lens of defect chemistry. The literature evaluated in this section and summarized in Table 5 demonstrates that macroscopic environmental performance is intrinsically linked to atomic-scale quantum mechanics. This relationship follows a definitive mechanistic cascade: specific defect architecture induces orbital reconstruction and d-band shifting, which subsequently tunes intermediate adsorption, steers the reaction pathway between radical and non-radical mechanisms, and ultimately dictates pollutant mineralization or detoxification. Crucially, this is the exact same cascade that governs N₂ and NO₃ ⁻ reduction to ammonia. The oxygen vacancy that lowers the barrier for *OH generation in anodic oxidation is governed by the same electronic principles as the vacancy that stabilizes the *N₂H intermediate in NRR. By recognizing that energy conversion and environmental remediation are not separate disciplines, but rather two sides of the same electrocatalytic coin, researchers can leverage defect informatics from one field to rapidly accelerate breakthroughs in the other.

TABLE 5. Defect-engineered electrocatalysts for advanced wastewater remediation.

Data-Comparability and Kinetic Context Note: The performance metrics provided illustrate the state-of-the-art capabilities of defect-engineered electrocatalysts. However, it must be rigorously noted that degradation kinetics are intrinsically dependent on the target organic molecule’s structure, initial loading concentration (C0), reactor geometry, and applied current density/bias. Consequently, removing 99% of a 10 mg/L pollutant cannot be mathematically equated to removing 99% of a 200 mg/L pollutant. Therefore, these values and kinetic rate constants (kobs) are intended to demonstrate mechanistic trends, shifts in ROS, pathways, and catalyst viability rather than to serve as a direct, absolute ranking of catalytic superiority across disparate studies.

5.4.1 Critical perspective for table 5

A profound paradigm shift is evident across state-of-the-art environmental electrocatalysis: the deliberate transition from indiscriminate radical generation to highly targeted, non-radical pathways. While traditional anodic oxidation and electro-Fenton systems relied on oxygen vacancies to maximize OH^• and SO₄ ^•- evolution, these radicals are aggressively scavenged by background organic matter in real wastewater. By leveraging precise defect engineering—such as Ce³⁺/Ce⁴⁺ redox cycling or trace Ru doping—catalysts can now structurally stretch peroxide bonds or instantly recombine intermediate radicals into singlet oxygen (¹O₂). This non-radical species specifically attacks electron-deficient pollutants (e.g., the nitroso groups in p-nitrophenol) while ignoring background scavengers. Furthermore, defect engineering successfully bypasses the universal kinetic bottleneck of electro-Fenton systems: the sluggish Fe(III)/Fe(II) reduction. The introduction of built-in defect electrons via sulfur vacancies or dual Co/Cu redox couples provides an inexhaustible, localized electron conduit that maintains perpetual catalytic turnover without requiring chemical addition.

6 Unified design principles and emerging paradigms

The rapid expansion of defect-engineered electrocatalysis has generated a vast repository of data spanning diverse materials, electrolytes, and target reactions. Despite the apparent complexity of the field, a profound convergence emerges: macroscopic catalytic activity is governed less by the bulk identity of the catalyst and more by how highly localized defects regulate the quantum-mechanical environment at the solid-liquid interface. Whether the target is ambient nitrogen reduction, nitrate conversion, or the generation of reactive oxygen species (ROS) for wastewater remediation, the underlying physical chemistry is identical. This mechanistic universality signals the end of the empirical “cook-and-look” era of catalyst development. The historical approach of indiscriminately introducing defects and blindly correlating them with performance metrics is fundamentally flawed, as defect concentration alone does not dictate catalytic behavior. Instead, the future of electrocatalysis relies on establishing quantitative, causal frameworks: leveraging multi-defect synergy to bypass universal scaling limits, employing artificial intelligence to map multidimensional design spaces, and utilizing operando techniques to track the dynamic evolution of defects under applied bias.

6.1 Universal descriptors: predicting catalytic turnover from quantum mechanics

A central challenge in contemporary electrocatalysis is establishing universal descriptors that predict performance across disparate chemical systems. Traditional geometric metrics, such as surface area or pore volume, fail entirely in defect-engineered systems, where atomic-scale perturbations dictate macroscopic output. Instead, predictive catalyst design must rely on a strict hierarchical cascade: specific defect architecture dictates the electronic descriptor, which tunes intermediate binding energies (ΔG), ultimately governing reaction kinetics. The d-band center (ϵ d) remains the preeminent electronic descriptor for transition-metal catalysts. Defect-induced spatial distortions alter the overlap between metal d-orbitals and adsorbate frontier orbitals, effectively shifting ϵ d relative to the Fermi level. However, a rigid band-shift model is often insufficient. Advanced descriptors increasingly incorporate the p-band center for non-metallic catalysts, the work function (Φ), and localized electron localization functions (ELF). The true value of these universal descriptors lies in their ability to map thermodynamic energy landscapes. By correlating an easily computable electronic descriptor with the adsorption free energies of key intermediates (e.g., Δ G * N 2 H for NRR, or Δ G * O H for ROS generation), researchers can establish “volcano plots” (Figure 4) that mathematically predict the theoretical maximum activity of a novel defect architecture before it is ever synthesized in the laboratory.

6.2 Multi-defect synergy: bypassing the sabatier limit

While isolated point defects (e.g., single oxygen vacancies) provide significant enhancements over pristine lattices, they remain fundamentally constrained by the Sabatier principle. An isolated defect cannot independently optimize the binding energies of multiple sequential intermediates without encountering linear scaling relations. The most sophisticated, high-performance systems currently reported rely on multi-defect synergy—strategically combining vacancies, dopants, and phase boundaries to decouple these intermediate binding energies and distribute catalytic functions. The coupling of a heteroatom dopant with an oxygen vacancy is a prime example of this synergistic decoupling. Introducing large, highly polarized rare-earth cations like Ce³⁺ into CoNi layered double hydroxides, or La³⁺ into SnO_x lattices, does not merely add a secondary active site. Instead, the dopant strongly polarizes the adjacent oxygen vacancy, preventing it from healing under anodic bias while creating an extreme local charge asymmetry that accelerates OH^• generation by over 100% compared to pure vacancy systems (Yang C. et al., 2022; Wei X. et al., 2025).

When applied to the electro-oxidation of antibiotics, these vacancy-dopant synergies achieve near-complete mineralization at fractions of the energy cost of state-of-the-art boron-doped diamond electrodes. This synergy extends effectively into metal-free systems and complex tandem configurations. In nitrogen self-doped biochar, the simultaneous presence of topological nitrogen vacancies and graphitic nitrogen establishes dual active sites: the graphitic N facilitates rapid H₂O₂ generation, while the highly electron-trapping N-vacancies instantly activate the peroxide to yield a precise mixture of both radical (OH^•, O₂ ^•-) and non-radical (¹O₂) oxidants (Zhang T. et al., 2022; Chen et al., 2023). At the device level, combining distinct defective electrodes—such as pairing an oxygen-vacancy-rich TiO₂ anode with a fluorine-doped porous carbon cathode—creates a closed-loop anodic oxidation-assisted electro-Fenton (AO-EF) system. The cooperative electron transfer across this macroscopic tandem setup can multiply degradation kinetic rate constants by over 40-fold relative to isolated half-cell reactions (Zhao et al., 2026). These complex cooperative mechanisms and their corresponding computational descriptors are categorized in Table 6. The prevailing consensus is clear: next-generation catalytic activity requires the deliberate orchestration of multiple defect types to break molecular scaling constraints.

TABLE 6. Multi-defect synergy and activity descriptors in electrocatalysis.

6.2.1 Critical perspective for table 6

The era of isolated point-defect engineering has reached its thermodynamic limit. As demonstrated across these synergistic architectures, single defects are fundamentally constrained by rigid intermediate scaling relations. Multi-defect configurations—such as coupling a symmetry-breaking vacancy with an isolated single-atom site, or anchoring vacancies within the built-in electric field of a phase boundary—represent the only viable strategy to independently decouple adsorption, activation, and desorption kinetics. Consequently, the frontier of electrocatalysis has shifted from merely increasing defect density to engineering complex, highly asymmetric electronic microenvironments.

Rather than assuming a universal descriptor across all reactions, defect engineering should be interpreted as a common strategy for modifying interfacial electronic structure, while the relevant descriptors and mechanistic consequences remain reaction-specific (Figure 5).

6.3 AI-guided and data-driven design: the digital catalyst revolution

The combinatorial complexity of multi-defect engineering—factoring in defect type, spatial distribution, doping elements, and support interactions—creates a design space too vast for experimental trial-and-error. Artificial intelligence (AI) and high-throughput computational screening are fundamentally disrupting this bottleneck (Gao S. et al., 2022; Rehman et al., 2022; Lun et al., 2025; Li and Wang, 2026). Digital catalyst discovery relies on density functional theory (DFT) to generate massive datasets of defect formation energies (ΔE _*f*), d-band shifts, and intermediate binding free energies. Machine-learning algorithms, such as graph neural networks and random forests, ingest these datasets to identify hidden, nonlinear structure–property relationships. Crucially, AI models can rapidly predict the thermodynamic stability of a proposed defect under specific electrochemical potentials, screening out thousands of configurations that would otherwise spontaneously heal or degrade operando. The field is rapidly moving toward a closed-loop framework: AI algorithms predict the optimal multi-defect configuration for a specific reaction (e.g., breaking the OH/OOH scaling relation for targeted singlet oxygen generation), autonomous robotic laboratories synthesize the material, and operando performance data is fed back into the algorithm to continuously refine the predictive model.

6.4 Dynamic defect electrocatalysis: the operando reality

Perhaps the most profound paradigm shift in modern electrocatalysis is the critical realization that defects are dynamically evolving entities (Zhang Y. et al., 2023; Wang X. et al., 2026). The historical assumption that a catalyst’s ex situ structure strictly dictates its working performance is fundamentally flawed. Under the extreme polarization, local pH gradients, and highly reactive intermediate coverage of a working electrochemical cell, crystalline lattices undergo continuous reconstruction. “As-synthesized” defects frequently act merely as precatalysts. During the first few minutes of applied bias, oxygen vacancies may migrate to the surface and amalgamate, metal dopants may selectively leach into the electrolyte, and the surface may transition from a crystalline oxide to an amorphous, defect-rich hydroxide layer. Consequently, the true active sites driving the reaction often exist only as transient, bias-induced states.

If defects evolve continuously, static ex situ characterization (e.g., vacuum XPS or standard TEM) provides an inaccurate picture of the catalytic mechanism. Future defect engineering must prioritize dynamic stability—designing lattices that intentionally reconstruct under bias to expose specific, highly active transient defect states, and subsequently revert to a stable resting state when the potential is removed. Understanding and controlling this operando defect mobility represents the ultimate frontier in designing multifunctional catalysts for both continuous energy conversion and aggressive wastewater matrices.

7 Translational outlook and future perspectives

Defect engineering has evolved from a niche materials-modification strategy into the dominant paradigm of modern electrocatalysis. By deliberately introducing structural discontinuities, researchers have successfully localized charge density, shifted d-band centers, and systematically bypassed the universal linear scaling relations that historically crippled the efficiency of pristine catalyst lattices. Nevertheless, a severe translational gap persists between academic demonstrations in idealized laboratory cells and the rigorous demands of industrial implementation. The field is currently undergoing a critical transition from a discovery-driven era into an engineering-driven era. The central scientific question is no longer whether defects can improve catalytic performance, but rather how highly active, metastable defect configurations can be precisely manufactured, dynamically sustained, and scaled into commercially viable architectures. Achieving this transition requires abandoning empirical trial-and-error optimization in favor of rigorous defect metrology, predictive artificial intelligence, and systems-level integration.

7.1 Defect metrology and atomic precision: moving beyond statistical ensembles

A fundamental limitation of current electrocatalysis research is the reliance on bulk synthesis techniques, such as thermal annealing in reducing atmospheres or plasma etching, which yield statistically random defect distributions. Consequently, the observed macroscopic catalytic activity is merely an ensemble average of diverse, uncharacterized active sites. This lack of atomic precision obfuscates true structure–activity relationships and renders computational modeling highly idealized.

The future of catalyst synthesis must transition from bulk defect generation to atomic-scale defect programming. Emerging methodologies—such as area-selective atomic layer deposition (ALD), molecular precursor self-assembly, and defect-seeded epitaxial growth—will be required to engineer specific coordination environments, such as precisely placing a single Ru atom adjacent to an oxygen vacancy without disrupting the surrounding lattice. Paralleling this synthetic precision, advanced defect metrology must become standard practice. Leveraging aberration-corrected high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) coupled with operando scanning tunneling microscopy (STM) will allow researchers to visualize the spatial distribution of defects and definitively confirm that the theoretical active site matches the physical reality.

7.2 Dynamic defect kinetics: overcoming lattice healing and operando degradation

While vacancies and dopants elevate catalytic activity, they simultaneously inject profound thermodynamic instability into the lattice. By definition, a defect is a high-energy structural state. Under the aggressive conditions of industrial electrocatalysis—which involve extreme polarization, radical generation, and complex wastewater matrices—defects exhibit an overwhelming thermodynamic drive to minimize their free energy via reconstruction, agglomeration, or lattice healing. The widespread academic reliance on ex situ stability testing before and after chronopotentiometry completely masks the dynamic reality of the interface. For example, oxygen vacancies frequently “heal” via the irreversible incorporation of oxygen species during anodic polarization, or trigger catastrophic lattice dissolution during intense cathodic bias. Moving forward, the objective cannot simply be to lock defects into rigid, thermodynamically unnatural states. Instead, the paradigm must shift toward engineering dynamic defect kinetics. Future high-performance catalysts will likely be designed as adaptive interfaces, where the applied overpotential continuously drives the reversible exsolution and dissolution of active dopants, or where the adsorption of a reactant such as NO₃ ⁻ spontaneously induces a transient vacancy that subsequently heals upon product desorption.

7.3 Industrial translation: bridging the gap from laboratory cells to techno-economic reality

The fundamental thermodynamics and kinetics of an electrocatalytic reaction change drastically as current densities scale from the 10-mA cm^-2 typical of laboratory-scale H-cells to the >500 mA cm^-2 required for industrial viability. At commercial current densities, mass transport limitations dominate, local pH gradients at the solid-liquid interface become extreme, and the chemical potentials of reactive intermediates shift dramatically. Consequently, defect-engineered catalysts must be seamlessly integrated into zero-gap membrane electrode assemblies (MEAs) and continuous-flow architectures. The design of these systems requires an intimate understanding of how gas-diffusion layers, ionomer networks, and defect-rich catalyst nanoparticles interact. A highly active defective catalyst is useless if its localized charge is insulated by poor ionomer integration or if its nanostructure collapses under high hydraulic flow rates.

It is important to stress that the industrial implementation is ultimately gated by techno-economic analysis (TEA) and life-cycle assessment (LCA). The synthesis of an atomically precise, multi-defect catalyst must not consume more energy or rely on more critically scarce elements, such as Ru or Ir, than the environmental or energy benefits it provides. Future research must rigorously evaluate catalyst performance through the dual lenses of the levelized cost of ammonia (LCOA) and complete mineralization efficiency per kilowatt-hour.

7.4 Closing vision: a unified defect paradigm for a circular economy

Historically, green energy conversion and environmental remediation have been treated as disparate disciplines, published in different journals and analyzed using distinct theoretical frameworks. This review definitively challenges that compartmentalization. The activation of the N≡N bond, the electroreduction of NO₃ ⁻, the two-electron reduction of O₂ to H₂O₂, and the selective generation of singlet oxygen (¹O₂) from persulfate are universally governed by the exact same principles of defect chemistry: localized charge redistribution, d-band shifting, and the breaking of linear scaling relations.

This mechanistic convergence provides a unified scientific language capable of accelerating breakthroughs in both fields simultaneously. More importantly, it unlocks the ultimate vision of a circular economy: multifunctional defect electrocatalysts operating within closed-loop systems.

In such integrated reactors, toxic nitrate runoff is no longer viewed merely as a pollutant requiring destruction, but as a heavily subsidized, thermodynamically favorable feedstock for decentralized green ammonia synthesis. Similarly, organic contaminants are not just mineralized; they are electro-oxidized at the anode to replace the sluggish oxygen evolution reaction, driving down the overall cell voltage for cathodic fuel production. Ultimately, defects are the atomic-scale regulators that will enable this electrified, closed-loop future. By marrying rigorous operando quantum mechanics with AI-guided discovery and industrial reactor engineering, defect electrocatalysis is poised to transcend materials science, becoming the foundational technology for a sustainable, carbon-neutral global infrastructure.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Valentín Briega-Martos, Cornell University, United States

Reviewed by: Xiaomin Xu, Curtin University, Australia

Raman Kumar, Guru Nanak Dev Engineering College, India

Cui Lai, Hunan University, China