Oxygen Reduction Reactions of Catalysts with Asymmetric Atomic Structures: Mechanisms, Applications, and Challenges

Abstract

Asymmetric-atomic-structure catalysts can modulate the interactions between active sites and intermediates through their unique electronic filling states and asymmetric charge distribution, breaking the linear relationship between adsorption energy and activity, thereby enhancing the catalytic performance of the oxygen reduction reaction (ORR). By introducing heteroelements, vacancies, or clusters into symmetric-atomic-structure catalysts (e.g., M-N₄), asymmetric configurations (such as M-N_x, M-N_x-S/B/O, etc.) can be formed. These modifications substantially alter their internal structure, trigger charge redistribution, and create asymmetric sites to reduce reaction energy barriers, effectively regulating the adsorption strength of oxygen intermediates and significantly improving ORR performance. This review systematically summarizes recent advancements in asymmetric-atomic-structure catalysts for ORR, elucidating the intrinsic “structure–performance–application” relationships to provide theoretical guidance for developing high-performance asymmetric atomic catalysts. First, the ORR mechanisms, including the two-electron and four-electron pathways, are introduced. Furthermore, strategies to modulate catalyst selectivity and activity through doping with metallic/nonmetallic elements or introducing defects are discussed. Finally, prospects for asymmetric-atomic-structure catalysts in next-generation energy storage and conversion technologies are outlined, offering novel insights to overcome current ORR performance bottlenecks.

1. Introduction

The depletion of fossil fuels and the associated energy demands and environmental issues have inevitably spurred interest in developing new energy sources and energy storage/conversion technologies, including hydrogen fuel cells, zinc–air batteries, and others [1,2,3,4,5,6,7,8,9]. In these devices, the oxygen reduction reaction (ORR) plays a pivotal role. The ORR, a core reaction in fuel cells and metal–air batteries, can proceed via two distinct pathways in flow cells depending on the number of electrons transferred during electrocatalysis: the two-electron pathway, generating hydrogen peroxide (H₂O₂) as an intermediate (O₂ + 2H⁺ + 2e⁻ → H₂O₂) [10,11], or the four-electron pathway, producing water (H₂O) as the final product (O₂ + 4H⁺ + 4e⁻ → 2H₂O) [12]. Despite being an exothermic reaction, practical applications of ORR still face significant challenges in energy utilization efficiency [13]. The high bond energy of O-O leads to a high activation energy barrier, resulting in sluggish reaction kinetics [14]; the actual operating potential is lower than the theoretical value, necessitating additional energy input and compromising energy efficiency [15]; the inherent complexity of the two-electron and four-electron transfer pathways severely limits energy conversion efficiency and device longevity [16]. Furthermore, catalyst instability in acidic or alkaline electrolytes, where they are prone to degradation or poisoning [17,18,19], poses additional critical barriers. These challenges collectively hinder the large-scale commercialization of ORR-based technologies. Platinum (Pt) is widely recognized as the most effective and commonly used ORR catalyst [20]. Although nanotechnology has paved the way for advancements in Pt catalysts—such as precise control of size and morphology to create Pt nanowires, exposure of high-index facets in Pt nanoparticles, and alloying strategies to form nanocrystals, nanosheets, and nanocages [21,22]—these approaches have mitigated catalytic activity as a limiting factor for ORR. However, the low atomic utilization efficiency of Pt (where only surface atoms participate in reactions, with less than 20% of internal atoms being utilized) poses significant challenges to the cost and performance of fuel cells [23,24,25]. These issues mean that the high cost and limited durability of Pt catalysts remain major obstacles to their widespread commercial application [26,27]. Consequently, the development of non-precious-metal catalyst materials with cost-effectiveness, abundant raw material availability, high ORR activity, excellent stability, and well-defined selectivity has become an imperative objective. In recent years, non-precious-metal symmetric atomic catalysts (e.g., M-N₄ sites with highly symmetric coordination structures) [28,29,30,31] have attracted significant attention due to their cost-effectiveness, high stability, universal synthesis of symmetric configurations, highly ordered active site distribution, and modifiable ligand environments. However, their performance remains constrained by low active site density, insufficient exposure efficiency, imbalanced adsorption strength of intermediates, and metal leaching in acidic media. This has facilitated the emergence of asymmetric atomic engineering strategies as an innovative paradigm for ORR catalyst design [32]. The core principle of asymmetric atomic engineering lies in constructing heterogeneous coordination microenvironments, facilitating precise modulation of the geometric and electronic structures of active centers. This approach breaks the performance limitations of conventional catalysts, fostering synergistic enhancement of activity, selectivity, and stability.

Since the early 21st century, conventional ORR catalysts have predominantly manifested highly symmetric coordination environments [33] (e.g., the planar square symmetry of Fe-N₄). The homogeneous electron distribution in symmetric structures facilitates uniform adsorption and activation of O₂, while modulation of intermediate adsorption energies helps prevent over-adsorption-induced active site poisoning. However, limitations such as restricted electronic structure tunability, corrosion susceptibility in acidic media, and low conductivity have impeded breakthroughs in catalyst stability, activity, and selectivity. To date, significant theoretical advancements have been achieved in understanding asymmetric-atomic-structure catalysts for the ORR. The core challenge of both 2e⁻ and 4e⁻ORR pathways lies in balancing the kinetics of intermediate adsorption and desorption [34]. In 2e⁻ORR, asymmetric coordination lowers the d-band center position of the active metal atom toward the Fermi level, weakening *OOH adsorption and enabling its direct desorption to produce H₂O₂. Conversely, for 4e⁻ORR, elevating the d-band center strengthens the adsorption of intermediates (*O, *OH), while asymmetric coordination promotes O₂ dissociation to achieve complete reduction to H₂O. The adsorption configuration of O₂ critically dictates pathway selectivity: end-on adsorption preserves the O-O bond (favoring 2e⁻ORR), whereas side-on adsorption cleaves it (driving 4e⁻ORR) [35]. Asymmetric catalysts optimize O₂ adsorption modes by tailoring the electronic structure of metal centers through heterogeneous coordination environments. Synergistic effects between elements further elevate ORR performance [36]. For instance, one element activates O₂ via adsorption, while another facilitates the desorption of oxygenated intermediates, ensuring efficient reaction kinetics. Moreover, it has been reported that the choice of carbon supports profoundly impacts catalytic activity [37]. The porous structure of carbon governs reactant/product transport efficiency: hierarchical porous hollow carbon spheres accelerate mass transfer due to their high specific surface area and open channels [38], while graphitic carbon enhances conductivity, reduces charge transfer resistance, and improves ORR kinetics. Additionally, 3D porous carbon supports provide abundant anchoring sites to disperse active centers, maximizing active site exposure [39]. These insights collectively establish a robust framework for designing high-performance asymmetric ORR catalysts.

Although asymmetric atom-structured catalysts have made progress in oxygen reduction reaction, they still have problems like poor structural stability, weak anti-poisoning ability, and high demand for structural control precision. Also, there is a lack of research on the mechanism of asymmetric atom-structured catalysts in two-electron and four-electron pathways (Figure 1) and summaries on activity enhancement. In this review, we outline the mechanism and application of ORR in the two-electron and four-electron pathways of asymmetric-atomic-structure catalysts, and provide feasible strategies for the rational design of asymmetric-atomic-structure catalysts. This paper introduces the oxygen reduction reaction mechanisms of the two-electron pathway and the four-electron pathway. In addition, this review synthesizes the reasons for the improvement of catalyst activity, including reasonably regulating the charge coordination environment, defect initiation, and equilibrium adsorption energy through element doping. Finally, by comprehensively looking forward to the development prospects of asymmetric-atomic-structure catalysts, some design principles of the catalysts are proposed to provide new design paradigms for the subsequent generation of asymmetric-atomic-structure oxygen reduction reaction catalysts.

2. Pathway and Mechanism of Oxygen Reduction Reaction

2.1. Two-Electron Pathway and Mechanism of Oxygen Reduction Reaction

The 2e⁻ORR pathway involves the catalytic reduction of molecular oxygen through a 2e⁻ transfer process, producing hydrogen peroxide (H₂O₂) or superoxide radicals (O₂⁻) as characteristic products. In this process, the adsorption mode of O₂ molecules on the catalyst surface (end-on or side-on) critically determines the reaction pathway. End-on adsorption, where O₂ binds to the catalyst surface (e.g., Pt) via a single oxygen atom, may instigate O-O bond cleavage, favoring the 4e⁻ pathway [40]. Conversely, side-on adsorption, characterized by parallel alignment of O₂ on surfaces (e.g., carbon-based materials), preserves the O-O bond and promotes the 2e⁻ pathway [41]. Mechanistically, adsorbed O₂ undergoes proton-coupled electron transfer (PCET) to form intermediates such as *OOH (in acidic media) or *O₂⁻ (in alkaline media), contingent upon pH. These intermediates subsequently accept a second electron and combine with protons, reducing to H₂O₂ (acidic) or HO₂⁻ (alkaline). Although less energy-efficient than the 4e⁻ route (direct H₂O formation), the 2e⁻ pathway is indispensable in applications like H₂O₂ synthesis, biomedical systems, and wastewater treatment [42,43,44].

The 2e⁻ pathway is fundamentally dictated by the partial reduction of O₂ molecules through the acceptance of two electrons, yielding H₂O₂ as the primary product. Its selectivity hinges critically on the catalyst’s adsorption strength toward the key intermediate (*OOH). Weak adsorption promotes H₂O₂ desorption, suppressing further progression to the 4e⁻ pathway, while strong adsorption drives complete O₂ reduction via the 4e⁻ route. Thus, refining the adsorption energy of *OOH through heteroatom doping—which modifies the electronic coordination configuration of active sites—is a pivotal strategy to enhance 2e⁻ selectivity (Figure 2a–c) [45]. Additionally, introducing heteroatoms into a catalyst’s stable lattice generates electron-rich or electron-deficient regions, creating vacancy defects [46]. In asymmetric atomic structures, such defects alter the local electron density and charge distribution, fine-tuning the catalyst’s O₂ adsorption energy and activation capacity. Defects further regulate the adsorption strength of critical intermediates (e.g., *OOH) on the catalyst surface (Figure 2d–g) [47]. An idealized adsorption strength stabilizes intermediates, facilitating subsequent electron transfer steps without premature desorption or over-adsorption that could divert the reaction toward the 4e⁻ pathway [48]. When a catalyst contains a single metal element as the active center, incorporating nonmetallic heteroatoms disrupts the symmetric coordination structure. The electronegativity mismatch between the metal and nonmetal induces electronic redistribution or a downshift of electron orbitals at the metal center, enhancing ORR activity [49]. Moreover, nonmetallic doping reduces the metal center’s adsorption strength for oxygen-containing intermediates, promoting their conversion to H₂O₂ [50].

Introducing nonmetallic elements such as N and O into the catalyst breaks the symmetry of the catalyst structure. This alters orbital hybridization, lowers the d-band center of the central atom, and modifies the electronic structure. Consequently, it weakens the adsorption strength of intermediates along the two-electron pathway, shifting the adsorption strength towards favoring desorption. Furthermore, catalysts featuring bimetallic asymmetric coordination promote the two-electron pathway due to their unique side-on adsorption configuration. Additionally, defect-reconstructed catalysts, such as those with dissolved cation vacancies, alter the oxidation state of the central atom. This optimizes the d-band, lowers the adsorption energy of intermediates, and steers the reaction towards the two-electron pathway.

2.2. Four-Electron Pathway and Mechanism of Oxygen Reduction Reaction

The 4e⁻ORR pathway is pivotal for attaining high-efficiency energy conversion, with water (H₂O, acidic media) or hydroxide ions (OH⁻, alkaline media) as final products, offering high energy efficiency and environmental compatibility. A critical step in the 4e⁻ pathway is the cleavage of the O-O bond in molecular oxygen, which typically serves as the rate-determining step (RDS).

The 4e⁻ORR pathway revolves around the complete cleavage of the O-O bond in molecular oxygen (O₂), achieving full reduction to H₂O [51]. In this process, O₂ molecules adsorb onto the catalyst surface via chemisorption, where the O-O bond is stretched as electrons transfer from the catalyst to O₂’s antibonding *π orbitals, leading to direct dissociation into two O atoms or stepwise reduction via PCET to form *OOH. These oxygen atoms sequentially accept electrons and protons, ultimately generating H₂O (in acidic media) or OH⁻ (in alkaline media). The complete reduction of O₂ through 4e⁻ transfer maximizes energy release, offering superior energy efficiency. The selectivity of the 4e⁻ORR pathway is governed by the adsorption strength of the *OOH intermediate, which dictates reaction pathway bifurcation. The electronic structure and charge distribution of the catalyst modulate the active site’s interaction with intermediates, making heteroatom doping a key strategy to tailor the local coordination environment of the central active site and augment 4e⁻ selectivity (Figure 3a–e) [52]. Additionally, defect engineering introduces vacancies near active sites, disrupting the periodic atomic arrangement and altering the central atom’s electronic distribution to favor the 4e⁻ pathway. Finally, embedding metal clusters into the catalyst structure leverages synergistic effects between clusters and single atoms to adjust the d-band center position of the metal center (Figure 3f,g). This balances the adsorption capacity for oxygen-containing intermediates, facilitating multi-step electron transfer in ORR [53]. Clusters also enhance the structural stability of asymmetric atomic configurations, preventing aggregation or leaching of single atoms during catalysis and prolonging catalyst durability [54].

Introducing nonmetallic elements such as B and S into the catalyst to form an asymmetric structure lowers the oxidation state of the central active atom and alters the electron distribution in its d-orbitals. This simultaneously enhances O₂ adsorption while accelerating the cleavage of the O-O bond. The co-coordination of multiple nonmetallic elements modifies the coordination structure of the central metal atom, thereby influencing the hybridization mode of its atomic orbitals. The synergistic effect between the nonmetallic elements and the metal elements promotes the conversion of *O to *OH, thereby biasing the oxygen reduction reaction (ORR) towards the four-electron pathway.

2.3. Research on Reaction Pathways

2.3.1. Rotating Ring–Disk Electrode Technique

The Rotating Ring–Disk Electrode (RRDE) technique is an advanced research method combining hydrodynamic regulation and electrochemical detection, widely used to investigate electrocatalytic mechanisms (e.g., ORR) [55]. The operational mechanism of the RRDE relies on forced convection generated by electrode rotation and selective detection of intermediates at the ring electrode, enabling high-sensitivity real-time dynamic analysis of ORR pathways. In RRDE measurements, the ORR occurs at the disk electrode, while intermediate products are selectively oxidized at the ring electrode. By analyzing the ratio of ring current to disk current, the H₂O₂ yield and electron transfer number can be calculated. For instance, Cui et al. [56] documented that the prepared BIM-Co₂Zn₈-500 catalyst achieved a H₂O₂ selectivity of 92.11% and an electron transfer number of 2.15 in the 0.05–0.65 V vs. RHE range, highlighting its exceptional selectivity for the 2e⁻ORR pathway (Figure 4a). Wang et al. [57] demonstrated that the C@PVI-(NCTPP)Fe-800 catalyst exhibited a H₂O₂ yield below 2% and an electron transfer number greater than 3.95 within the 0.2–0.8 V vs. RHE range (Figure 4b), confirming its predominance in the efficient 4e⁻ORR pathway. The Co-N₅ catalyst constructed by Zhang et al. [58] exhibited a high H₂O₂ selectivity of up to 92.5% in 1 M KOH within the potential window of 0.3–0.7 V vs. RHE using the RRDE technique, with the electron transfer number (n) stably maintained between 2.0 and 2.3. This confirms that the dominant reaction pathway is two-electron oxygen reduction. The Fe_SA-Fe₃C/NC catalyst synthesized by Huang et al. [59] consistently showed an H₂O₂ yield below 5.2% across a wide potential window of 0.2–0.8 V vs. RHE. The electron transfer number calculated via RRDE was n ≈ 4.0, further confirming that the ORR process proceeds efficiently via the four-electron pathway.

2.3.2. In Situ Raman Spectroscopy

In situ Raman spectroscopy is a real-time monitoring method that integrates Raman spectroscopy with in situ reaction conditions, enabling dynamic surveillance of molecular structures, chemical bond changes, and reaction intermediate formation under actual reaction environments, providing direct evidence for elucidating ORR mechanisms [62]. Raman spectroscopy is predicated upon the inelastic scattering effect, detecting frequency shifts caused by interactions between incident light and molecular vibrational modes to reveal molecular vibrations, rotations, and lattice vibrations. In situ Raman spectroscopy couples Raman spectrometers with reaction systems (e.g., electrochemical cells) via specially designed reaction cells or probes, allowing real-time monitoring of structural changes during reactions. The “in situ” feature ensures real-time observation without disrupting the sample’s native state or environment. For example, Pan et al. [60] demonstrated that for the Fe/NSC-vd catalyst in 0.1 M KOH, in situ Raman spectra revealed rapid generation and transformation of *OOH intermediates as the potential decreased, while the protonation process of O₂⁻ was slower (Figure 4c). This suggests that Fe-N coordination structures likely serve as the primary active centers for O₂ adsorption and activation under alkaline conditions. In contrast, in 0.5 M H₂SO₄, in situ Raman spectra showed slower conversion of O₂ to O₂⁻, while *OOH intermediate formation and transformation remained pronounced, indicating that Fe-S coordination structures may dominate as active centers for pivotal ORR steps in acidic media (Figure 4d). These findings highlight dynamic structural changes and intermediate adsorption behaviors during the 4e⁻ORR pathway of the Fe/NSC-vd catalyst under divergent electrolytes, confirming distinct reaction mechanisms in alkaline and acidic conditions.

2.3.3. In Situ ATR-SEIRAS

In situ attenuated total reflection surface-enhanced infrared absorption spectroscopy (ATR-SEIRAS) is a highly sensitive surface analysis technique combining attenuated total reflection (ATR) and surface-enhanced infrared absorption spectroscopy (SEIRAS) [63]. It enables real-time monitoring of molecular vibrational information of surface-adsorbed species in practical reaction environments (e.g., electrochemical interfaces, catalytic reactions). The surface enhancement effect grants this technique ultra-high sensitivity for detecting surface species, allowing identification of trace amounts of low-concentration substances and even sub-monolayer-level surface-selective detection, surpassing the limitations of conventional infrared spectroscopy. The in situ monitoring capability permits real-time observation in the actual ORR environment without complex sample pretreatment or labeling, capturing transient intermediates and kinetic information during reactions. With surface specificity, high sensitivity, and environmental compatibility, in situ ATR-SEIRAS has become one of the most effective techniques for elucidating molecular behavior at solid/liquid/gas interfaces. Exemplifying this, Gao et al. [61] analyzed reaction intermediates in the CoMn-NSC catalyst in a 0.1M KOH electrolyte using in situ ATR-SEIRAS: the O-O stretching vibration at 1240 cm⁻¹ indicated the presence of superoxide intermediates (O₂⁻), the HOH bending vibration at 1638 cm⁻¹ reflected interactions between water molecules and intermediates, and the broad peak at 3000–3550 cm⁻¹ corresponded to O-H stretching vibrations, confirming hydroxyl adsorption (Figure 4e,f). By tracking potential-dependent spectral changes, the generation of oxygen-containing intermediates and transformation of superoxide intermediates were monitored, validating the CoMn-NSC catalyst’s efficacy in promoting the 4e⁻ORR pathway.

2.3.4. Density Functional Theory

Density functional theory (DFT) calculations, a quantum mechanical computational method, have become a core tool in materials simulation by efficiently solving electron density [64]. The Hohenberg–Kohn theorem, one of the cornerstones of DFT, states that the ground-state energy of a system is a unique functional of electron density, meaning all ground-state properties are derivable from a given electron density distribution. The Kohn–Sham equations, another theoretical foundation of DFT, incorporate the exchange–correlation energy functional, which describes the exchange and correlation interactions between electrons. Accurate calculation of this energy is critical to DFT precision. Although approximative treatments of the exchange–correlation functional introduce errors, DFT demonstrates significant value in predicting material properties and guiding experiments. In ORR studies, DFT is applied to analyze the geometric and electronic structures of catalytic sites [65]. For example, Zhao et al. [66] revealed that high-shell heteroatoms in the FeCoN₆ catalyst regulate charge distribution through DFT combined with machine learning, significantly reducing overpotential. Jia et al. [67] synthesized the O-PtCo-FeNC catalyst and used DFT to demonstrate that Fe sites lower the d-band center of Pt, weakening intermediate adsorption energy and enhancing ORR activity and stability. DFT calculations also determine reaction mechanisms and pathways by analyzing intermediate adsorption and free energy landscapes. Liu et al. [68] constructed four structural models for the Fe_SA/N,S-PHLC catalyst via DFT, highlighting the correlation between the FeN₃SOH/Pyridine N model and oxygen intermediates. By calculating Gibbs free energy diagrams for different models at U = 1.23 V, the FeN₃SOH/Pyridine N model exhibited relatively low overpotential during *OOH transformation, favoring O-OH bond cleavage and promoting the 4e⁻ORR pathway.

DFT has become an indispensable tool in ORR research, enabling comprehensive analysis from atomic-level mechanisms to macroscopic performance prediction. It deepens understanding of ORR mechanisms, accelerates the development of low-cost, high-performance catalysts, and drives advancements in technologies such as fuel cells and metal–air batteries.

2.3.5. X-Ray Absorption Near-Edge Structure

X-ray absorption near-edge structure (XANES) is a cutting-edge spectroscopic technique based on synchrotron radiation sources. It analyzes the local electronic structure and geometric environment of materials by detecting changes in the probability of core electron excitation to unoccupied orbitals for specific elements. Its working principle involves measuring the X-ray absorption coefficient of a specific element in the region near its absorption edge (typically from 10–20 eV below the edge to 30–50 eV above the edge). When the energy of the incident X-ray photon reaches a level sufficient to excite a core electron of the specific element into an unoccupied state (the conduction band or molecular orbital), the absorption coefficient exhibits a sharp rise (the absorption edge) [69]. As revealed by Co K-edge XANES spectroscopy analysis of the ZIF-ZC91@Co(OH)₂-V_Co catalyst synthesized by Zhang et al. [70], the Co K-edge absorption edge shifts towards higher energy. This indicates an increase in the Co oxidation state from +2 to nearly +3. This valence state elevation arises because the cation vacancies (V_Co) induce electron loss from surrounding Co atoms, triggering an electronic restructuring. This optimization enhances the catalyst’s adsorption capability for the *OOH intermediate, consequently improving H₂O₂ selectivity. Analysis of both the Cu K-edge and L-edge XANES for the Cu-S₁N₃ asymmetric structure catalyst synthesized by Jiang et al. [71] revealed that the Cu L-edge position lies between those of CuPc and CuS. Combined with the K-edge absorption edge position, this indicates an average Cu oxidation state of +1.97. During the oxygen reduction reaction (ORR) process, the absorption edge shifted towards lower energy and the white line intensity decreased, confirming a reduction in the Cu oxidation state from +2 to +1. This demonstrates that the low-valent Cu⁺ species serves as the ORR active center. The results confirm the existence of C–S and N–Cu bonds, providing evidence for the direct coordination of S to Cu.

2.3.6. Fourier Transform Infrared Spectroscopy

Fourier Transform Infrared Spectroscopy (FTIR) is a powerful technique used to obtain the infrared absorption or emission spectra of solid, liquid, or gas samples. It investigates the molecular structure, chemical bonds, and functional groups of substances based on their selective absorption of infrared radiation. The working principle of FTIR involves using an interferometer to generate an interferogram that contains information across all infrared frequencies. This interferogram passes through the sample. A detector then receives the interferogram signal modified by the sample’s absorption. Subsequently, a computer performs a Fourier transform on this complex interferogram signal, converting it into the conventional infrared absorption spectrum. In situ Fourier Transform Infrared Spectroscopy (In situ FTIR) is an advanced technique that enables real-time monitoring of dynamic structural changes in materials under authentic reaction conditions (such as specific temperature, pressure, atmosphere, or electrochemical environments). This is achieved by integrating a specialized reaction cell (e.g., high-temperature/high-pressure cell, electrochemical cell, catalytic reaction cell) into the FTIR optical path, allowing the infrared beam to pass directly through the sample while it is in its operational state. By rapidly acquiring time-resolved interferograms and performing real-time Fourier transforms, instantaneous spectra revealing molecular bond vibrations/rotations during the reaction are obtained. This allows for the capture of real-time information on intermediates, surface-adsorbed species, and reaction pathways [72]. For the ZnCo-MTF catalyst prepared by Liu et al. [73], Fourier Transform Infrared Spectroscopy (FTIR) revealed that the precursor material ZnCo-ZIF exhibited characteristic peaks of 2-methylimidazole (e.g., C=N stretching vibration, C-H bending vibration, etc.). During the reaction process, the characteristic peaks of 2-methylimidazole gradually diminished, while those of 1,2,3-triazole progressively intensified (e.g., N-H/N-N vibrations, triazole ring skeletal vibrations). The final product, ZnCo-MTF, displayed only the strong characteristic peaks of 1,2,3-triazole, with the 2-methylimidazole peaks completely absent. This demonstrates that the FTIR spectrum of the final product matches the characteristic peaks of the pure 1,2,3-triazole ligand, indicating the complete replacement of 2-methylimidazole by 1,2,3-triazole within the ZnCo-MTF framework, resulting in a new metal–organic framework (MOF) structure. This transformation provides the structural foundation for the subsequently enhanced catalytic performance (high H₂O₂ production rate/selectivity). During the synthesis of the Fe-N₄-SM catalyst constructed by Shen et al. [74], the precursor containing the defect-prone ligand 5-ATZ exhibited a Zn-N vibration peak at 423.0 cm⁻¹. The displacement of this Zn-N vibration peak, coupled with the preservation of the main framework characteristics, confirmed the successful synthesis of the dual-ligand ZTIF precursor. Furthermore, this analysis revealed that the thermal instability of 5-ATZ underpins the subsequent selective C-N bond cleavage and serves as the structural basis for atomic strain modulation.

3. Control of Oxygen Reduction Reaction Pathways

3.1. Control of Two-Electron Pathways

It is a reasonable choice to introduce one or more nonmetallic elements instead of the nonmetallic elements in the original structure to form an asymmetric atomic structure in order to control the tendency of the catalyst to follow the two-electron pathway in the reaction, so as to improve the catalytic efficacy of 2e⁻ORR [75].

The incorporation of heteroelements can create catalysts with asymmetric configurations, which induces uneven charge distribution at the central active sites and modifies electron orbital states. As elucidated by Ivan Parkin et al. [76], a surface engineering strategy was employed to synthesize atomically dispersed Co on N-doped carbon black, forming an asymmetric coordination configuration (Co-C/N/O). The asymmetric structure facilitates hybridization between Co 3d orbitals and C/N/O p-orbitals, which lowers the d-band center of Co, modulates the electronic structure, and weakens the adsorption strength of intermediates (e.g., *OOH) (Figure 5a,b), thereby preventing over-strong adsorption that would drive the 4e⁻ pathway. Oxygen atoms act as electron-withdrawing sites from Co, while N and C serve as electron-donating components, establishing a charge gradient that stabilizes *OOH adsorption. DFT calculations reveal a positive correlation between the vertical displacement of Co atoms from the graphene plane and the *OOH adsorption energy, highlighting the potential for geometric configuration optimization to fine-tune catalytic activity. Similarly, Zhang et al. [58] synthesized a Co-N₅ catalyst by precisely controlling the coordination number of Co atoms through the thermal stability differences of precursors (Figure 5c). The introduction of an additional N atom in the Co-N₅ configuration induces uneven charge distribution and electronic rearrangement, breaking the linear scaling relations (LSRs) intrinsic to conventional symmetric Co-N₄ catalysts and thereby resolving the activity–selectivity trade-off. DFT calculations demonstrate that the altered d-orbital occupancy of Co optimizes the adsorption energy of *OOH intermediates, preventing both O-O bond cleavage (4e⁻ pathway) and insufficient O₂ activation (blocked 2e⁻ pathway). In situ Raman spectroscopy reveals that Co-N₅ sites stabilize *OOH intermediates and slow their consumption rate. KSCN poisoning experiments further confirm that atomically dispersed Co serves as the catalytically active center, with changes in d-orbital states being the key factor for enhanced performance.

Additionally, the incorporation of heteroatoms modulates the adsorption capability of the central active atom toward oxygen-containing intermediates, tailoring the adsorption strength to favor *OOH desorption. Gao et al. [77] synthesized an asymmetrically coordinated Zn-N₃O catalyst by adsorbing Zn ions onto polypyrrole (PPy) to achieve an elevated Zn loading of 11.34 wt%, forming Zn-N/O bonds (Figure 5d). The introduction of O-atom coordination withdraws electrons from Zn, increasing its oxidation state and redistributing its electron cloud density. This induces a contraction of Zn’s d-orbital electron density, which strengthens the interaction between Zn’s d-orbitals and the antibonding orbitals of *OOH, thereby diminishing the O-O bond dissociation energy barrier and optimizing the adsorption energy of Zn toward *OOH intermediates to favor the 2e⁻ORR pathway. DFT calculations reveal that the density of states (DOS) near the Fermi level in the Zn-N₃O structure significantly increases, indicating enhanced hybridization between Zn’s d-orbitals and O’s p-orbitals, which improves intermediate adsorption. Zn K-edge EXAFS spectroscopy further confirms structural modulation by O-coordination, showing altered coordination numbers and bond lengths between Zn and N/O atoms. Additionally, Han et al. [78] developed a high-coordination Co-N₅ moiety by precisely regulating the coordination environment of Co atoms, leveraging adjacent electron-withdrawing epoxy groups to construct a Co-N₅-O-C catalyst. By adjusting the number of N atoms in the first coordination shell of Co single-atom catalysts (SACs) or partially substituting them with O atoms, Co-N_xO_y SACs were formed. This modification alters the charge distribution and d-band center of the metal atom, achieving an optimized electronic structure that shifts the reaction pathway toward the 2e⁻ORR. The high-coordination N atoms modulate Co’s d-band center, fine-tuning the adsorption energy of the *OOH intermediate. DFT calculations demonstrate that the Co-N₅ moiety exhibits a reduced overpotential for the 2e⁻ORR compared to the Co-N₄ configuration, thereby enhancing pathway selectivity toward H₂O₂ production.

Bimetallic or multimetallic-doped ORR catalysts typically orchestrate the adsorption strength of intermediates (*OOH) to achieve a balance where oxygen-containing intermediates neither undergo over-adsorption (leading to further reduction via the 4e⁻ pathway) nor under-adsorption (causing sluggish kinetics), thereby avoiding O-O bond cleavage [79,80]. In such catalysts, the complementary electronic properties of the two metals synergize: one metal adsorbs and activates O₂ to form *OOH, while the other weakens *OOH adsorption, enabling its desorption to generate H₂O₂. Additionally, differences in bimetallic coordination alter the electronic distribution of active sites, enhancing selectivity for the 2e⁻ pathway. The asymmetric coordination environment further induces hybrid orbital energy matching with O₂’s *π orbitals, facilitating single-electron transfer to produce superoxide radicals (O₂⁻), which subsequently undergo selective reduction to H₂O₂. This dual-metal synergy not only modulates intermediate adsorption but also tailors electronic interactions to stabilize the 2e⁻ pathway. Liu et al. [73] fabricated a hollow nanocubic-structured catalyst (ZnCo-MTF) using ZnCo-ZIF as a precursor and 1,2,3-triazole as a ligand. As shown in Figure 6a,b, the synergistic interaction between Zn and Co induces a unique side-on adsorption configuration for O₂: one oxygen atom weakly binds to Zn, while the other strongly interacts with Co. This binding mode effectively polarizes the O₂ molecule, lowering the O-O bond dissociation barrier and promoting desorption of the *OOH intermediate, thereby selectively driving the 2e⁻ pathway to generate H₂O₂. The coordination between 1,2,3-triazole ligands and Co forms *d-pπ delocalized orbitals**, increasing the occupation of Co-N antibonding orbitals and weakening Co-N bonds to activate the Co center. Simultaneously, Zn incorporation optimizes the electron density of Co via charge transfer, enhancing O₂ adsorption capability. DFT calculations reveal that the DOS of ZnCo-MTF spans the Fermi level, signifying superior conductivity compared to monometallic systems, which accelerates electron transfer kinetics.

Strategies to disrupt structural symmetry in catalysts extend beyond metal/nonmetal doping and include defect engineering to reconstruct the original framework. For instance, Zhang et al. [70] synthesized a ZIF-ZC91@Co(OH)₂-V_Co catalyst via electrochemical self-optimization reconstruction of ZIF-ZC91, where partial Zn²⁺ leaching generates cobalt cation vacancies. These vacancies, formed through selective Zn dissolution, modulate the catalyst’s electronic structure and surface morphology. The presence of cation vacancies optimizes the adsorption strength of *OOH intermediates (Figure 6c), suppresses O-O bond cleavage, and promotes the 2e⁻ORR pathway. Mechanistically, the vacancies alter the oxidation state of Co and refine its d-band characteristics, thereby reducing the adsorption energy of *OOH. DFT calculations confirm that the energy barrier for the 2e⁻ pathway is lower than that of the 4e⁻ route, effectively inhibiting O-O bond dissociation.

Notably, incorporating heteroelements into catalyst structures not only induces disparate charge distribution at the central atom but also creates defects near the active sites. These defects generate localized electron-rich regions or charge-asymmetric environments, stabilizing the *OOH intermediate (precursor to H₂O₂) and preventing its further reduction to *O or *OH, thereby blocking the 4e⁻ORR pathway. As demonstrated by Zhang et al. [81], a coaxial Co single-atom catalyst (Co_SA-N-C/CNTs) was synthesized via an SCVD strategy. Concentrated nitric acid oxidation was employed to introduce surface defects (e.g., carbon vacancies) and oxygen-containing functional groups on CNTs, forming defective CNTs (d-CNTs). These defects provide abundant anchoring sites, enabling the stable immobilization of Co atoms as single-atom sites on the carbon substrate to form high-density Co-N_x active centers. The defects alter the electronic state of the carbon substrate, modulating the electron density of Co-N_x sites and thereby tuning the adsorption strength of intermediates (*OOH) (Figure 6d,e). DFT calculations reveal that the ΔG*OOH of Co-N_x sites approaches the apex of the volcano plot, indicating that defect-induced electronic optimization achieves near-optimal adsorption energy for enhanced 2e⁻ORR selectivity. Additionally, the defect-engineered Co_SA-N-C/CNTs exhibit a microporous structure, where the pores and surface roughness (resulting from defects) accelerate the diffusion of reactants (O₂) and products (H₂O₂), while exposing more active sites to enhance catalytic performance.

3.2. Control of Four-Electron Paths

3.2.1. Incorporation Forms of Nonmetallic Elements

To modulate the adsorption energy of central active sites, nonmetallic elements are incorporated into inherently symmetric and stable catalysts, constituting asymmetric atomic structures [82]. Due to the higher electronegativity of nonmetallic elements compared to metallic counterparts, their incorporation alters the electron density distribution of the central active element. The nonmetallic dopant withdraws electrons from neighboring atoms (e.g., carbon or the central metal atom), creating electron-deficient regions. This polarization effect strengthens O₂ adsorption and activation, lowers the free energy barrier for oxygen adsorption reactions, and consequently augments ORR activity. For instance, Ding et al. [83] synthesized a single-atom Co catalyst with N/B asymmetric coordination (Co-N₃B) via a pyrolysis method and anchored it on boron–nitrogen Co-doped graphene (BNG) to form Co-N₃B active sites (Figure 7a). The incorporation of B atoms, owing to their low electronegativity, facilitates electron transfer to the Co center, reducing its oxidation state and increasing its electron density. This electronic modulation enhances Co’s ability to adsorb and activate oxygen intermediates (e.g., *O, *OOH). Projected density of states (PDOS) analysis from DFT reveals stronger d-p orbital hybridization between Co’s d-orbitals and O’s p-orbitals near the Fermi level in Co-N₃B, indicating enhanced orbital overlap that promotes the 4e⁻ORR pathway. Attenuated total reflection surface-enhanced infrared absorption spectroscopy (ATR-SEIRAS) experiments demonstrate that the Co-BNG catalyst exhibits stronger adsorption capacity for oxygen intermediates (e.g., *O₂⁻ and *OOH) and maintains excellent stability during prolonged reaction cycles. Similarly, Zhang et al. [84] fabricated a Co-N₃-C/CNT catalyst via a bidentate pre-coordination-assisted pyrolysis strategy. DFT calculations elucidate how the asymmetric low-coordination structure of Co-N₃ disrupts the symmetric charge distribution of Co-N₄. The reduced coordination (Co-N₃) induces asymmetric electron density distribution at the Co atom (Figure 7b–d), enhancing oxygen activation capabilities and optimizing the adsorption/desorption equilibrium of ORR intermediates. The Co-N₃ configuration exhibits lower energy barriers during ORR, particularly in the critical *O-to-*OH conversion step, demonstrating superior intrinsic catalytic activity compared to symmetric Co-N₄ counterparts. Furthermore, Yin et al. [85] successfully constructed a CoN₃S-PC catalyst by anchoring single Co atoms on porous carbon (PC) and subsequently doping S atoms into the first coordination shell of the Co active sites, forming an asymmetric CoN₃S coordination structure. The incorporation of S atoms induces electronic rearrangement and geometric distortion at the Co active center, disrupting its charge symmetry and enhancing interactions between Co 3d orbitals and oxygen molecules/intermediates (Figure 7e,f). This strengthens O₂ adsorption at the Co sites and accelerates O-O bond cleavage. DFT calculations reveal that S doping modifies the electronic distribution of Co’s d-orbitals, promoting O₂ adsorption/activation while suppressing H₂O₂ generation. Concurrently, the CoN₃S sites exhibit a reduced thermodynamic energy barrier and kinetic superiority for the 4e⁻ORR pathway, ensuring efficient and selective oxygen reduction.

The coordination of nonmetallic elements directly with the central active site increases the electron density peripherally to the active atom, thereby influencing the adsorption energy of intermediates. Jiang et al. [71] designed an asymmetrically coordinated Cu-S₁N₃ single-atom active site via an atomic interface engineering strategy, constructing a single-copper-atom catalyst (S-Cu-ISA/SNC) based on MOF-derived hierarchically porous carbon. The introduction of S atoms disrupts the conventional symmetric CuN₄ coordination, forming an asymmetric Cu-S₁N₃ active site that modifies the electronic structure around the central Cu atom. The relatively lower electronegativity of S atoms reduces the loss of valence electrons from Cu atoms, altering the local electron density distribution (Figure 8a) and thereby optimizing the adsorption free energy of ORR intermediates (e.g., *OOH, *O, *OH) (Figure 8b) while diminishing the reaction energy barrier. In situ soft X-ray absorption near-edge structure (XANES) tests and in situ FTIR spectroscopy elucidated the electronic and atomic structural evolution of Cu sites during ORR. DFT calculations, the Koutecky–Levich (K-L) equation, and RRDE tests collectively validated the catalyst’s performance, confirming the presence of low-valent Cu active centers, an electron transfer number (n ≈ 4), and a low H₂O₂ yield (<4%). Yang et al. [86] synthesized an FeN₃S/Cmeso catalyst by embedding single Fe atoms into a mesoporous carbon matrix and introducing S atoms to directly coordinate with the Fe center, forming an FeN3S coordination structure. The direct coordination of S atoms to the Fe center disrupts the symmetry of the FeN₄ configuration, increasing the electron density around Fe (Figure 8c). The incorporation of S lowers the d-band center of Fe, weakening the orbital overlap between active sites and oxygenated intermediates (e.g., *OH), thereby optimizing their adsorption energy (Figure 8d). DFT calculations demonstrate that the S-induced coordination enhances the electron cloud density of Fe atoms, which reduces the orbital overlap between active sites and oxygen-containing intermediates. This attenuated adsorption strength facilitates intermediate desorption and lowers the energy barrier during the oxygen reduction process.

The co-coordination of multiple nonmetallic elements modifies the coordination number and coordination structure of the central metal atom, thereby influencing the hybridization mode of metal atomic orbitals [87]. Different coordination structures exhibit distinct geometric configurations and symmetries, which constrain or guide the direction and extent of hybridization in metal atomic orbitals. Following multi-nonmetallic element coordination, charge transfer occurs between the nonmetallic elements and the central metal atom. Nonmetallic elements tend to attract electrons, causing a partial transfer of electron density from the metal atom to the nonmetallic ligands and a reduction in the electron cloud density of the metal atom. This charge redistribution modifies the electronic configuration of the metal atom’s d-orbitals, affecting the formation and characteristics of its hybridized orbitals. Additionally, highly electronegative nonmetallic elements can lower the d-band center of metal sites through electronic modulation, geometric restructuring, and synergistic catalytic effects. This weakens the adsorption strength toward oxygen-containing intermediates (such as *OOH, *O, and *OH), bringing the adsorption energy closer to the peak of the “volcano curve”. Such optimization of reaction kinetics significantly promotes the 4e⁻ORR pathway while suppressing H₂O₂ generation routes, such as reducing the probability of *OOH protonation to form H₂O₂. As reported by Liu et al. [88], an asymmetric nitrogen–oxygen-coordinated tin single-atom catalyst (Sn-N/O-C) with Sn atoms as active centers was developed via a simple one-step pyrolysis strategy, forming asymmetric SnN₂O sites through the co-coordination of N and O atoms. The N and O co-coordination disrupted the symmetry of the Sn sites, optimized the p-orbital electron distribution of Sn, and induced strong hybridization between the Sn 5p orbitals and the O₂ 2p orbitals. This altered the theoretical d-band center of Sn, promoting electron transfer and accelerating O₂ activation and reduction (Figure 8e,f). DFT calculations revealed that the Sn atoms with broken symmetric coordination facilitated charge transfer and O₂ activation. RRDE measurements demonstrated a hydrogen peroxide (H₂O₂) yield of <5% in alkaline media and <3% in acidic media, with an electron transfer number close to 4, indicating the dominance of the 4e⁻ pathway. Similarly, Liu et al. [89] introduced O atoms into the first coordination shell of Fe and incorporated adjacent defects, inducing a transition of Fe from a low-spin state to a medium-spin state and forming a Fe-N₃O coordination structure, thereby synthesizing a single-atom iron catalyst (Fe-N/O-C). The coordination of N and O atoms in Fe-N/O-C redistributes the charge, increases the electron density at the Fe sites, and significantly weakens the adsorption of oxygen-containing species such as O₂, mitigating over-adsorption of O₂. Theoretical calculations reveal that the d-band center of Fe-N/O-C is lower than that of Fe-N-C, leading to weakened adsorption of intermediates. The formation of adsorbed *OOH from O₂ becomes the new rate-determining step (replacing the original *OH desorption), with a reduced energy barrier. In situ Raman spectroscopy further confirms this shift in the PDS. RRDE tests demonstrate an electron transfer number close to 4 and an extremely low H₂O₂ yield (<5%), indicating that a highly efficient 4e⁻ pathway dominates the reaction.

Figure 8. (a) The molecular orbital of *O adsorbed on Cu-S₁N₃ is in S-Cu-ISA/SNC. (b) ORR overpotential (ηORR) as a function of the free adsorption energy (∆G*O) of *O on different Cu central groups. (a,b) Reproduced with permission [71]. Copyright 2020, Springer Nature. (c) Charge density difference plot of FeN₃S₁. (d) Absorption energy of *OH intermediates at different active sites. (c,d) Reproduced with permission [86]. Copyright 2024, Elsevier Ltd. (e) The variation of Gibbs free energy of the ORR pathway at different Sn sites under a potential bias voltage of U = 0 V. (f) The charge density difference plot and relative charge state of Sn. (e,f) Reproduced with permission [88]. Copyright 2024, Wiley-VCH Verlag.

Figure 8. (a) The molecular orbital of *O adsorbed on Cu-S₁N₃ is in S-Cu-ISA/SNC. (b) ORR overpotential (ηORR) as a function of the free adsorption energy (∆G*O) of *O on different Cu central groups. (a,b) Reproduced with permission [71]. Copyright 2020, Springer Nature. (c) Charge density difference plot of FeN₃S₁. (d) Absorption energy of *OH intermediates at different active sites. (c,d) Reproduced with permission [86]. Copyright 2024, Elsevier Ltd. (e) The variation of Gibbs free energy of the ORR pathway at different Sn sites under a potential bias voltage of U = 0 V. (f) The charge density difference plot and relative charge state of Sn. (e,f) Reproduced with permission [88]. Copyright 2024, Wiley-VCH Verlag.

3.2.2. Co-Incorporation Forms of Nonmetallic and Metallic Elements

The introduction of metal elements beyond the central active site disrupts the local charge symmetry of the original catalyst structure and redistributes the electron cloud density of the active centers [90]. The incorporation of additional metal elements also narrows the energy band of the central active metal sites, enhancing electron mobility. This enables faster electron transfer during catalytic reactions, thereby accelerating the ORR rate. The synergistic interplay of multiple metal elements modulates the adsorption strength of active sites toward oxygen-containing intermediates (e.g., *O, *OOH), ensuring effective adsorption of these intermediates on the catalyst’s surface while preventing overly strong adsorption that impedes desorption. This equilibrium promotes the 4e⁻ transfer pathway. Additionally, multimetal systems provide diverse active sites that act synergistically and collectively participate in the ORR process [80]. As reported by Shang et al. [91], a three-step pyrolysis strategy was employed to anchor Se atoms and other metal atoms (e.g., Fe) onto a C₂N substrate, forming SeN₂-MN₂ active sites, thereby synthesizing an asymmetric Se-based diatomic SeM-C₂N catalyst (Figure 9a). Among these, SeFe-C₂N demonstrated optimal ORR performance. The heteronuclear Se atom induces short-range modulation to polarize the charge distribution of the Fe atom, reducing its oxidation state, enhancing the co-adsorption capacity for intermediates (e.g., *O and *OH), and forming heteronuclear diatomic sites, with diatomic sites accounting for 82% of the active centers. The introduction of Se increases the electron density at the Fe atom, creating a localized polarized electric field that lowers the ΔG value of the rate-determining step. This moderates the adsorption strength of intermediates, accelerates O₂ activation and electron transfer, and achieves more efficient conversion of O₂ to H₂O. DFT calculations integrated with operando XAS techniques confirmed that Se sites adsorb *OOH, while the Fe-Se dual sites co-adsorb *O, synergistically promoting the *O→*OH conversion. The K-L equation, fitted from LSV curves at varying rotation speeds, yielded an electron transfer number of n ≈ 4. RRDE measurements directly determined a H₂O₂ yield of <4%, and the calculated electron transfer number (based on ring and disk currents) ranged from n = 3.97 to 4.05, conclusively confirming the dominance of the 4e⁻ pathway. As reported by Chang et al. [92], a Fe-Co dual-site N-doped porous carbon catalyst (Fe,Co/N-C) was synthesized using a host–guest strategy. By adjusting the precursor ratio of FeCl₃ to Co-ZIF (optimal at Fe/Co = 1:1), the Fe,Co/N-C catalyst incorporated Fe-Co dual sites that provided more favorable active sites for O-O bond activation. Compared to single-metal sites, the dual sites significantly reduced the energy barrier for O-O bond cleavage, thereby enhancing catalytic activity. DFT calculations confirmed that upon O₂ adsorption on the Fe-Co dual sites, the O-O bond was stretched, and the cleavage energy barrier decreased to 0.25 eV (vs. 0.65 eV for a single Fe site), facilitating the four-electron pathway. K-L equation analysis and RRDE tests indicated an electron transfer number of n ≈ 3.96 and a H₂O₂ yield below 1.17%.

Incorporating low-electronegativity nonmetallic elements into catalysts possessing dual-metal sites can further modulate the distribution of surface electron cloud density and alter the spin state of metal sites. The interaction between low-electronegativity nonmetallic elements and metal sites enhances the electronic properties of the metal centers. For example, in Fe-based catalysts, the introduction of Si induces a transition of Fe ions from a low-spin state to a medium-spin state. Fe sites in the medium-spin state exhibit higher reactivity, improved oxygen adsorption capacity, and stronger interactions with oxygen-containing intermediates (e.g., *O and *OOH), thereby accelerating the ORR rate [94]. Additionally, low-electronegativity nonmetallic elements can form isolation layers during high-temperature treatments, preventing the aggregation of metal sites into nanoparticles and maintaining the dispersion and high activity of active sites. Finally, these elements tend to donate electrons to the dual-metal sites, causing an upward shift in the d-band center, which fortifies the adsorption of ORR intermediates (*OOH, *OH). As reported by Gao et al. [61], an adsorption–pyrolysis process was utilized to anchor adjacent S,N-coordinated Co atoms and N-coordinated Mn atoms (forming CoN₂S-MnN₃) onto a nitrogen-doped carbon substrate, constructing CoN₂S-MnN₃-2OH active sites and synthesizing the CoMn-NSC catalyst. In this structure, S atoms preferentially coordinate with Co to form CoN₂S, while Mn retains its N-coordinated MnN₃ configuration. The strong interactions between Co and Mn atoms, coupled with multi-electron filling effects induced by asymmetric charge distribution (Figure 9b), enhance orbital hybridization between Co and O₂, lower the O-O bond cleavage energy barrier, optimize adsorption energies of reaction intermediates, and improve electron transfer efficiency through dual-atom synergy. Here, Co acts as the primary active site for O₂ adsorption, while Mn serves as an auxiliary site that modulates the local electronic environment. This adjustment shifts the d-band center of Co closer to the Fermi level, strengthening its adsorption and activation capabilities for O₂, thereby boosting overall catalytic activity (Figure 9c). DFT calculations revealed that the d-band center of Co in the asymmetric structure enhances hybridization with the *π orbitals of O₂, improves antibonding electron filling efficiency, and optimizes adsorption strength. K-L equation analysis of polarization curves and RRDE tests confirmed an electron transfer number of n ≈ 4 and a H₂O₂ yield below 5%. In situ Raman spectroscopy and in situ ATR-SEIRAS further validated the efficient adsorption, activation, and reduction of O₂ on the CoMn-NSC catalyst during the reaction. Su et al. [95] employed a heteroatomic pair cooperative modulation strategy to alter the coordination environment of Fe sites by introducing Co and B heteroatomic pairs, thereby synthesizing the Fe-B-Co/NC catalyst. The introduction of Co atoms achieved electronic synergy between Fe and Co, modulating the d-band center position of Fe and balancing the competitive relationships between RDS during the oxygen reduction reaction. Concurrently, asymmetric coordination transformed Fe from an intermediate-spin state (t⁵_2ge¹_g) to a high-spin state (t⁴_2ge²_g). The high-spin Fe sites enhanced orbital hybridization with O₂, promoting O-O bond cleavage to directly generate *O intermediates (rather than accumulating *OOH), thereby enabling rapid four-electron reaction kinetics and improving oxygen reduction efficiency. K-L equation and RRDE tests confirmed that the Fe-B-Co/NC catalyst exhibited an average electron transfer number of 3.87 within the 0.5 V–0.65 V potential range, indicating a 4e⁻ pathway during ORR with a H₂O₂ yield below 6.9%. In situ EIS and SR-FTIR techniques confirmed that Fe sites accelerate O-O bond cleavage in *OOH to rapidly form *O intermediates. In 0.1 M KOH, the Fe-B-Co/NC catalyst demonstrated a half-wave potential (E_1/2) of 0.891 V and diffusion-limited current density (Jd) of 5.9 mA cm⁻², surpassing commercial Pt/C (0.851 V, 5.19 mA cm⁻²). During 50 h stability testing, 95% of the initial current density was retained.

3.2.3. Other Incorporation Forms (Functional Groups, Defect Sites, Metallic Clusters, etc.)

Similar to 2e⁻ORR, strategies to augment 4e⁻ORR selectivity extend beyond doping with solely metallic or nonmetallic elements. Functional groups (e.g., OH⁻) can also be incorporated to coordinate with central active atoms, or combined with elements (e.g., S) for multi-coordination, optimizing the electronic structure of the central atom [96]. Additionally, carbon frameworks with pronounced edge effects (e.g., HsGDY) reconfigure the edge morphology of the catalyst structure, inducing unique electronic states at the edges. This effectively modulates the material’s band structure, enhancing pathway selectivity. Constructing asymmetric anchoring sites at the edges of the carbon substrate strengthens the bonding between metal atoms and the carbon matrix, reducing metal aggregation and leaching during reactions. As documented by Hou et al. [93], iridium single-atom catalysts (Ir-N-HsGDY) were synthesized by anchoring Ir atoms at the sp-N and pyridinic N sites of hydrogen-substituted graphdiyne (HsGDY) and coordinating them with hydroxyl groups (OH⁻), forming IrN₂(OH)₃ active sites. The introduction of OH⁻ ligands broadened the Ir d-band and displaced its center downward, enhancing orbital hybridization between Ir and O/N, optimizing the electronic structure (Figure 9d,e), weakening the adsorption strength of intermediates (e.g., *OH), and reducing reaction energy barriers. The macroporous structure of HsGDY provided space for OH⁻ incorporation, while its hierarchical design effectively isolated Ir precursors to prevent aggregation and ensure single-atom dispersion. DFT calculations confirmed that the IrN₂(OH)₃ active site exhibited the lowest ORR overpotential (0.54 eV), significantly outperforming unmodified IrN₂ (2.24 eV). When applied to PEMFCs, the Ir-N-HsGDY catalyst achieved a half-wave potential (E_1/2) of 0.82 V vs. RHE in acidic media (0.1 M HClO₄), comparable to commercial Pt/C with superior methanol tolerance. In alkaline media (0.1 M KOH), it demonstrated a half-wave potential (E_1/2) of 0.89 V vs. RHE within the potential window above 0.78 V, significantly outperforming commercial Pt/C, and exhibited only an 18 mV negative shift in E_1/2 after 5000 CV cycles, indicating high durability. Lv et al. [97] synthesized the Mn-N-HsGDY catalyst by asymmetrically anchoring Mn atoms as single atoms at edge sites of the HsGDY carbon substrate, incorporating two nitrogen species (sp-N and pyridinic N) and OH⁻ ligands to form asymmetric MnN₂(OH)₃ active sites, thereby disrupting the electronic state of conventional symmetric M-N₄ configurations. The strong coordination between sp-N (high electronegativity) and Mn, combined with the electron-withdrawing effect of OH⁻ ligands, synergistically drove the Mn center to a high oxidation state (+3 to +4), altered the electronic structure of Mn atoms, and reduced the desorption energy barrier of the intermediate *OH. DFT calculations elucidated that the asymmetric coordination of the MnN₂(OH)₃ site weakened *OH adsorption, lowering the RDS energy barrier to 0.502 eV, significantly outperforming symmetric sites (e.g., MnN₄). RRDE-derived electron transfer numbers and K-L equation analysis of LSV curve slopes at different rotation rates both yielded an average electron transfer number (n) close to 4, confirming the dominance of the 4e⁻ORR pathway.

Additionally, etching defects (e.g., vacancies) adjacent to the central atom that disrupt the original symmetric electronic distribution can also enhance 4e⁻ORR performance [98]. Defects alter the electron cloud distribution around the central active atom, modifying electron delocalization, thereby influencing the catalyst’s electronic structure. This creates an unsaturated coordination environment for atoms at defect sites, providing additional active sites for ORR [99]. Zhang et al. [100] prepared an atomically dispersed Fe-N-C catalyst modified with proximal carbon vacancies (Fe-N₄-V_c) through pyrolysis combined with urea etching to selectively cleave C-N bonds near Fe-N₄ sites, forming carbon vacancies (Vc). In the Fe-N₄-V_c structure, the introduced adjacent carbon vacancies induced redistribution of electron density in Fe-N₄, lowered the oxidation state of Fe (coexistence of Fe²⁺ and Fe³⁺ with an average oxidation state of +2.6), and generated asymmetric electronic states. This caused an upward shift of the Fe d-band center, reduced electron occupancy in antibonding orbitals, optimized adsorption behaviors of O₂ and intermediates, lowered reaction free energy barriers, and thereby enhanced ORR activity. DFT calculations demonstrated that Vc significantly reduced the adsorption energy of O₂ (from −2.27 eV to −1.66 eV) and optimized the adsorption free energy of intermediates (e.g., *OH). Among ORR free energy diagrams of distinct structures, Fe-N₄-V_c exhibited the smallest RDS energy barrier and the highest activity. Shen et al. [74] synthesized the Fe-N₄-SM catalyst by selectively cleaving C-N bonds around Fe-N₄ sites to construct an asymmetric atomic strain environment, thereby modulating the electronic structure of the Fe center. The selective cleavage of C-N bonds adjusted the asymmetric atomic strain environment at Fe sites, perturbed the symmetric electronic structure of Fe-N₄, modulated the d-band broadening and d-orbital electron distribution of the Fe center, and consequently altered the adsorption strength between the catalyst and ORR intermediates, enabling precise regulation of catalytic activity. RRDE tests and K-L equation analysis revealed an average electron transfer number of ~3.98 and H₂O₂ yield below 1%, confirming the dominance of the four-electron pathway. In situ Raman spectroscopy detected vibration peaks of *O₂ and *OOH intermediates, supported by theoretical models, further validating the four-electron mechanism. DFT calculations demonstrated that the RDS of Fe-N4-SM shifted from *OH desorption (energy barrier 0.83 eV) to the first electron transfer step (energy barrier 0.25 eV).

When functional groups (e.g., OH⁻) and defect vacancies are co-incorporated into a catalyst structure, they may exhibit a synergistic effect to jointly enhance the ORR performance. For example, Zhan et al. [101] employed a microenvironment engineering strategy to anchor Fe single atoms on a defective N-doped carbon (DNC) substrate by regulating micromorphology and interfacial microenvironments, synthesizing the Fe/DNC catalyst. Carbon vacancies reshaped the electronic properties of FeN₄(OH) sites via inductive effects, optimizing hybridization between Fe 3d and O₂ 2p orbitals, balancing energy barriers for *OOH formation and *OH reduction, and regulating charge distribution at Fe sites through electron induction to promote O₂ adsorption and intermediate desorption (e.g., *OH), thereby accelerating reaction kinetics. The introduction of axial hydroxyl (OH⁻) ligands disrupted the symmetry of FeN₄, enhancing adsorption capability for oxygen intermediates and reducing kinetic barriers. RRDE analysis revealed extremely low H₂O₂ yields (0.35% in alkaline, 6% in acidic, and <5% in neutral conditions) and an electron transfer number close to 4 (≈4.06), confirming the ascendancy of the 4e⁻ pathway and validating high ORR efficiency/selectivity. DFT simulations identified FeN₄(OH) as the true active center, with the defective model (FeN₄(OH)-A) exhibiting the lowest RDS energy barrier (0.43 eV), outperforming symmetric FeN₄(OH) (0.64 eV).

Finally, co-incorporating metal active atoms and metal clusters into the catalyst structure can also generate asymmetric atomic configurations [102]. The presence of metal clusters modifies the charge distribution and local coordination environment surrounding metal active sites, facilitating O₂ molecule access and adsorption at active sites while stabilizing the adsorption of subsequent reaction intermediates, thereby promoting 4e⁻ORR. Additionally, clusters and asymmetric atomic structures collectively create multi-active sites, enabling synergistic catalytic effects. Huang et al. [59] successfully synthesized the Fe_SA-Fe₃C/NC catalyst by encapsulating Fe atoms and Fe₃C clusters on nitrogen-doped carbon. The incorporation of Fe₃C clusters disrupted the symmetric electronic structure of FeN₄, depressed the d-band center of Fe atomic sites, thereby altering the adsorption strength of oxygen intermediates on Fe-N₄-C sites, reduced the adsorption energy of *OOH intermediates (0.38 eV), accelerated O-O bond cleavage, and promoted the 4e⁻ORR pathway. Additionally, the high specific surface area (443 m²/g) and mesoporous structure (4 nm pore size) of the nitrogen-doped carbon enhanced active site exposure and mass transfer efficiency. The synergistic interaction between atomically dispersed Fe-N₄ sites and Fe₃C clusters provided an efficient catalytic platform for ORR. DFT calculations demonstrated that the introduction of Fe₃C clusters increased the electron cloud density of Fe d-orbitals (Figure 9f), augmenting O₂ adsorption and activation while weakening *OH adsorption strength, favoring *OH desorption. This optimized the adsorption/desorption behavior of intermediates, reduced reaction energy barriers, and improved ORR kinetics and activity. In situ infrared spectroscopy confirmed that *OOH intermediate formation was the rate-determining step, and the FeSA-Fe₃C/NC catalyst effectively desorbed *OOH intermediates, thereby lowering reaction barriers and facilitating ORR progression.

4. Application of Oxygen Reduction Reaction

4.1. Device for Producing H₂O₂

Currently, the traditional anthraquinone process remains the dominant technology for H₂O₂ production due to its maturity and high capacity. This method hinges upon the hydrogenation–oxidation cycle of 2-ethylanthraquinone over palladium catalysts (hydrogenation to form hydroanthraquinone, followed by oxidation to regenerate anthraquinone and release H₂O₂), integrated with extraction and purification for continuous production. It offers advantages such as mature technology, high productivity (annual production exceeding 100,000 tons per single unit), and high product purity. However, it faces environmental and economic challenges, including dependence on fossil-fuel-derived hydrogen, high energy consumption, solvent leakage, catalyst costs, and byproduct accumulation. With increasing environmental regulations and advancements in renewable energy technologies, emerging methods like electrochemical synthesis are gradually overcoming technical barriers. The electrochemical production of H₂O₂ via the two-electron oxygen reduction reaction (2e⁻ORR) offers a green, low-energy alternative to the anthraquinone process [103,104]. This approach directly utilizes water and oxygen from air as feedstocks, eliminating reliance on fossil-fuel hydrogen or organic solvents, thereby reducing carbon emissions and pollution risks at the source. Reactions proceed under ambient conditions, circumventing the high energy consumption and safety hazards of high-temperature/pressure processes, and generate no anthraquinone degradation byproducts, minimizing wastewater treatment demands. Additionally, catalyst design can selectively suppress the 4e⁻ORR pathway, markedly potentiating H₂O₂ selectivity.

The design of devices for the electrochemical production of H₂O₂ is critical to augmenting synthesis efficiency and practicality, with current key technological routes including gas diffusion electrode reactors and flow cell systems [105].

A gas diffusion electrode (GDE) reactor is a device that achieves efficient electrochemical reactions through a gas–liquid–solid three-phase interface [106]. Its core structure typically comprises four components: the gas diffusion layer (GDL), microporous layer (MPL), catalyst layer (CL), and current collector. Gas diffusion layer (Figure 10a): Constructed from porous hydrophobic materials (e.g., carbon fiber paper or polytetrafluoroethylene (PTFE) membranes), it ensures efficient gas transport and prevents electrolyte penetration through its pore structure. Microporous layer: A mixture of carbon powder and PTFE forms a hydrophobic and conductive intermediate layer, further refining pore structures, enhancing adhesion between the catalyst layer and GDL, and improving interfacial stability. Catalyst layer: Catalysts are loaded onto carbon supports via spraying, electrodeposition, or electrospinning. Current collector: A conductive metal mesh (e.g., nickel mesh) or carbon cloth provides electrical conduction and mechanical support. The GDE enables high-efficiency reactions by optimizing the three-phase interface (gas–electrolyte–catalyst). In H₂O₂ electrocatalytic synthesis, O₂ is directly conveyed to the catalyst surface, where H₂O₂ is generated via the 2e⁻ORR pathway, significantly reducing mass transfer resistance. Upon reaching the catalyst layer, O₂ forms a gas–liquid–solid three-phase interface with the electrolyte. At this interface, oxygen, ions in the electrolyte, and active sites on the electrode surface interact, driving complex electrochemical reactions. Under the influence of the electric field, ions in the electrolyte migrate to the catalyst layer surface and participate in the reaction with O₂ to produce H₂O₂. Zhang et al. [58] constructed the cathode of a three-phase flow cell device using a gas diffusion layer (GDL) made of carbon paper and a microporous layer (MPL) loaded with Co-N₅/NC catalyst. With 1.0 M KOH as the electrolyte in O₂-saturated alkaline conditions, O₂ was reduced to H₂O₂ via 2e⁻ORR, achieving a production rate of 16.1 mol g_cat⁻¹h⁻¹ at 150 mA cm⁻². Operating at 0.5 V vs. RHE for 20 h, the accumulated H₂O₂ concentration reached 427.7 mM (14,540 ppm) with a Faradaic efficiency (FE) >90%. Liu et al. [76] coated a Co-C/N/O catalyst onto a gas diffusion electrode and assembled a two-compartment flow cell separated by an NR-212 cation-exchange membrane to separate the anode and cathode chambers. In 1 M phosphate-buffered saline (PBS), galvanostatic electrolysis at 20 mA cm⁻² for 5 h yielded an H₂O₂ production rate of 4.72 mol g_cat⁻¹h⁻¹. After 5000 cycles of accelerated durability testing, the activity declined by only 20–39 mV, retaining stable H₂O₂ selectivity.

A flow cell system is an electrochemical device utilizing gas diffusion electrodes (GDEs) and electrocatalytic materials to enable electrocatalytic O₂ reduction at the gas–liquid–solid three-phase interface for H₂O₂ generation [108]. The core structure of the flow cell comprises the GDE (cathode), anode, flowing electrolyte system, membrane materials, and external circulation and storage systems (Figure 10b). The GDE, serving as the cathode, directly supplies gaseous O₂ to the reaction interface, forming a three-phase interface (gas–liquid–solid) and significantly enhancing oxygen mass transfer efficiency. The flowing electrolyte system circulates the electrolyte via pumps to mitigate concentration polarization and rapidly remove generated H₂O₂, preventing its further reduction to water or decomposition. Commonly used membrane materials include proton-exchange membranes (PEMs) or anion-exchange membranes (AEMs). External circulation and storage systems enable continuous electrolyte flow, regulating reactant concentration and flow rate. To enhance H₂O₂ production efficiency, designing superior catalysts remains a critical challenge, particularly in acidic environments where ORR tends to favor complete reduction to H₂O, necessitating precise structural tuning of catalysts to maintain H₂O₂ selectivity. Zhang et al. [63] achieved an H₂O₂ production rate exceeding 2000 mmol g_cat⁻¹h⁻¹ with a Faradaic efficiency (FE) >90% in an acidic electrolyte (0.1 M HClO₄) using a flow cell assembled with Co_SA-N-C/CNTs, maintaining stable performance for 100 h without decay. In alkaline environments, however, challenges such as insufficient catalyst activity or poor membrane stability necessitate improvements in catalyst performance and membrane durability to enhance H₂O₂ yields. Similarly, Han et al. [55] demonstrated that the Co-N₅-O-C catalyst delivered an H₂O₂ production rate of 11.3 mol g⁻¹h⁻¹ at 200 mA cm⁻² in 1M KOH, retaining ~80% FE after 24 h of continuous operation at 100 mA cm⁻².

Future research should focus on multiscale collaborative design, spanning from atomic-level modulation of catalyst electronic structures to optimization of reactor architectures, integrated with computational materials science, to achieve low-cost, high-concentration, and fully green process chains for electrochemical H₂O₂ synthesis, ultimately revolutionizing traditional chemical production paradigms.

4.2. Zinc–Air Battery

Aqueous zinc–air batteries are a novel category of chemical power source that utilize metallic zinc as the anode and oxygen from air as the cathode active material, combining characteristics of both secondary batteries and fuel cells [109]. These batteries typically consist of an air electrode, an alkaline or neutral electrolyte, a separator, and a zinc anode. During discharge, zinc at the anode is oxidized in the alkaline electrolyte to form zincate ions while releasing electrons. At the cathode, oxygen from the air is adsorbed onto the air diffusion electrode and reduced in the electrolyte to generate hydroxide ions (OH⁻), which react with zincate ions to produce zinc oxide (ZnO) and H₂O. Electrons flow from the anode to the cathode through an external circuit, generating electrical current. During charging, the anode reaction involves the oxidation of ZnO to Zn²⁺ and OH⁻ by losing electrons, while the cathode reaction reduces water to produce oxygen and OH⁻ (Figure 10c). However, aqueous zinc–air batteries suffer from sluggish ORR kinetics, necessitating highly efficient catalysts to reduce overpotentials and enhance reaction rates [110]. When catalysts with metal–nonmetal asymmetric atomic structures are applied, they exhibit superior performance compared to commercial Pt/C catalysts, offering improved stability and catalytic activity for practical applications. Catalysts can be applied in two forms: the first is their direct use as the cathode for catalytic reactions. Zhang et al. [84] employed Co-N₃-C/CNT as the air cathode, where the zinc–air battery demonstrated a maximum power density of 159.3 mW cm⁻² and a high specific capacity of 892.6 mAh g⁻¹, with no performance decay after 8000 cycles. Yang et al. [86] applied Fe-N₃S/C_meso as the cathode material in a zinc–air battery, attaining an open-circuit voltage of 1.45 V, a peak power density of 178 mW cm⁻², and stability exceeding 350 h, with 96.17% capacity retention over 100 h. Ren et al. [111] assembled a zinc–air battery using an asymmetric N,P-coordinated Co single-atom catalyst (d-CoN₃P) as the cathode, delivering a specific capacity of 842 mAh g⁻¹ and a power density of 175.7 mW cm⁻², exceeding Pt/C (662 mAh g⁻¹ and 123.8 mW cm⁻²), and retained stable performance after 1163 charge–discharge cycles. The second approach involves loading catalysts onto the cathode for catalytic reactions, where the loaded catalysts appear to outperform those directly used as cathodes. For example, Ding et al. [83] deposited a Co₁-BNG catalyst onto the anode to form a composite electrode, achieving a zinc–air battery with a maximum power density of 253 mW cm⁻² and durability exceeding 110 h. Jiang et al. [71] applied S-Cu-ISA/SNC as the air cathode catalyst in a zinc–air battery, demonstrating a maximum power density of 225 mW cm⁻² and negligible voltage decay during 50 h cycling tests, indicating exceptional discharge stability. Lin et al. [88] employed the Sn-N/O-C catalyst in a ZAB, achieving a maximum power density of 254 mW cm⁻², a specific capacity of 781 mAh g⁻¹, and stable performance over 350 h of discharge testing (Figure 10d). Metal–metal asymmetric-atomic-structure catalysts further enhance the performance of aqueous zinc–air batteries, with bimetallic synergistic catalysis exhibiting stronger performance than metal–nonmetal catalysts, yielding higher peak power densities. For instance, Shang et al. [91] applied the SeFe–C₂N catalyst as the cathode in a ZAB, achieving a high power density of 287.2 mW cm⁻² and a specific capacity of 764.8 mAh g⁻¹, with virtually imperceptible voltage decay over 380 h of charge–discharge cycling. Wang et al. [107] integrated a Co/Cu-N-C catalyst into a ZAB, attaining a peak power density of 256.1 mW cm⁻² and cycling stability exceeding 500 h.

Flexible solid-state zinc–air batteries (ZABs) are novel energy storage devices characterized by high theoretical energy density, augmented safety, and flexible bendability. Their working principle aligns with aqueous ZABs, with the core innovation being the replacement of traditional liquid electrolytes with solid-state electrolytes and the achievement of flexibility through material and structural design [112]. The key components of a flexible solid-state ZAB include an air cathode, zinc anode, solid-state electrolyte layer, and flexible encapsulation layer. The air cathode facilitates ORR. The zinc anode combines a metallic zinc layer with a flexible substrate to enable bending. The solid-state electrolyte layer conducts ions, prevents short circuits by isolating electrodes, and provides mechanical flexibility and strength. The flexible encapsulation layer is typically made of polymeric materials to block moisture ingress and maintain mechanical flexibility [113]. Cathode catalyst deactivation significantly degrades ORR kinetics and is a primary cause of capacity decay in flexible solid-state ZABs. For instance, Wang et al. [107] synthesized a Co/Cu-N-C catalyst that demonstrated a high specific capacity of 727.1 mAh g⁻¹ and stable power density of 113.2 mW cm⁻² under bending conditions in flexible ZABs (Figure 10e), outperforming Pt/C-based batteries (21.7 mW cm⁻²). Lv et al. [97] applied the Mn-N-HsGDY catalyst in a 1D solid-state ZAB, achieving an open-circuit voltage of 1.363 V and maintaining stable voltage output during bending to power a timer. In a 2D solid-state ZAB, it delivered a peak power density of 156.6 mW cm⁻², surpassing most reported flexible ZABs, and preserved stable voltage differences throughout 1400 min of charge–discharge testing.

4.3. Hydrogen–Oxygen Fuel Cell

Proton-exchange membrane fuel cells (PEMFCs) constitute highly efficient and clean energy conversion devices that directly transform chemical energy into electricity through hydrogen–oxygen electrochemical reactions [114]. Among these, the ORR at the cathode is one of the most critical processes. A PEMFC consists of gas diffusion layers (GDLs), membrane electrode assemblies (MEAs), and bipolar plates. The MEA, which includes catalyst layers (CLs) and a proton-exchange membrane (PEM), serves as the core of the electrochemical conversion process. The performance and efficiency of devices relying on this process depend heavily on the properties of ORR electrocatalysts and PEMs [115]. Thus, developing higher-activity ORR catalysts and cost-effective MEAs with high durability is essential for advancing these energy conversion technologies [116]. Fundamentally, PEMFCs operate as the reverse process of water electrolysis: H₂ and O₂ react in the presence of catalysts to produce H₂O while releasing electrical energy. In the operation of PEMFCs, H₂ enters the anode, where it splits into H⁺ and e⁻ under the action of catalysts. The H⁺ ions migrate through the PEM to the cathode, while the e⁻ are forced through an external circuit, generating an electric current to power electronic devices. At the cathode, O₂ combines with the migrating H⁺ and incoming e⁻ to drive the ORR, producing H₂O. To optimize PEMFC performance, high-performance ORR catalysts are essential. Catalyst particles must be integrated on both sides of the PEM to minimize the transport distances of H⁺ and e⁻, thereby enhancing reaction kinetics. The chemical stability of the catalysts and their compatibility with the membrane are critical to preventing degradation of both components and maintaining optimal performance. Due to the constrained availability and high cost of noble metal-based catalysts, asymmetric-atomic-structure catalysts have emerged as promising alternatives to Pt-based catalysts. For example, Liu et al. [89] reported a Fe-N/O-C cathode catalyst that achieved a peak power density of 1179 mW cm⁻² in PEMFCs (Figure 10f). Yin et al. [85] developed a CoN₃S-PC cathode catalyst that delivered a peak power density surpassing 1.32 W cm⁻² under O₂ at 80 °C. Operating at a constant voltage of 0.6 V, the CoN₃S-PC cathode maintained 53.3% of its current density after 37 h (ultimately stabilizing at approximately 300 mA cm⁻²). Chang et al. [92] synthesized a (Fe,Co)/N-C catalyst that achieved a peak power density of 505 mW cm⁻² in PEMFCs (353 K with 0.2 MPa backpressure), reaching 76% of the power density of commercial Pt/C. Additionally, in H₂/air fuel cells, it demonstrated negligible voltage decay when operated continuously for 100 h at constant current densities (600 and 1000 mA cm⁻²). These advancements highlight the potential of metal–nonmetal asymmetric coordination catalysts to outperform Pt/C in PEMFC applications, driving progress toward cost-effective and durable fuel cell technologies.

5. Conclusions and Prospects

In conclusion, asymmetric atomic structures have demonstrated exceptional performance in the ORR, showing practical applications in high-efficiency H₂O₂ production, metal–air batteries, and hydrogen–oxygen fuel cells, and have materialized as one of the most promising alternatives to traditional noble metal catalysts. Based on recent advancements in theoretical and experimental research, key design principles for developing next-generation high-performance asymmetric-atomic-structure catalysts have been summarized, providing critical guidance for optimizing catalytic activity and selectivity.

(i)

Design Principles for Metal Element Doping: In ORR catalysts, transition metal elements are the predominant choice for constructing asymmetric atomic structures owing to their unique electronic configurations and chemical properties. The partially filled d-orbitals of transition metals can hybridize with the antibonding *π orbitals of O₂, modulating the activation energy for O-O bond cleavage. The position of the d-band center critically governs the adsorption strength of oxygen intermediates (e.g., *OOH, *O, *OH). Furthermore, transition metals exhibit multiple oxidation states (e.g., Fe²⁺/Fe³⁺, Co²⁺/Co³⁺), empowering reversible electron transfer during ORR. This redox flexibility allows their oxidation states to dynamically adapt across reaction steps, thereby facilitating catalytic progression. Additionally, the atomic size and electronegativity of transition metals promote the formation of asymmetric coordination structures with other atoms or ligands. Such configurations induce asymmetric charge distribution and electronic effects, optimizing the adsorption and reactivity toward ORR intermediates. Finally, transition metals possess high chemical and thermal stability, enabling sustained catalytic activity under ORR operating conditions (e.g., acidic/alkaline environments, high potentials) while resisting structural degradation or corrosion.

(ii)

Design Principles for Nonmetal Element Doping: In the fabrication of asymmetric-atomic-structure catalysts, the incorporation of high-electronegativity nonmetal elements induces charge polarization through their strong electron-withdrawing capability, creating localized electron density differences. This effectively modulates the electronic structure of the catalyst, enhancing its adsorption capacity for reactants and catalytic activity. Additionally, synergistic co-doping of multiple high-electronegativity nonmetal elements can be considered to achieve more significant catalytic performance improvements. Interactions between different dopants may collectively modulate the electronic and geometric structures of the catalyst, generating synergistic effects. For low-electronegativity nonmetal elements, their weak electron-withdrawing or electron-donating capabilities can modulate the charge distribution, surface reactive sites, and intermediate adsorption behavior of the catalyst structure. Selecting elements with atomic radii matching the substrate can reduce the risk of lattice distortion and preserves structural stability.

(iii)

Design Principles for Other Dopants: In asymmetric-atomic-structure ORR catalysts, defect engineering amplifies catalytic activity and selectivity by introducing vacancies, edge sites, and other asymmetric configurations. These defects modulate electronic distribution, expose active sites, and optimize reaction pathways. Defects instigate localized states near the Fermi level, reducing the energy gap between O₂’s d-orbitals and the catalyst’s electronic states (e.g., carbon vacancies shift the Fermi level downward by 0.2–0.5 eV), thereby enhancing charge transfer efficiency. Finally, defects in porous structures expose additional active sites (e.g., defects in mesoporous carbon walls enhance mass transport and active site density).

(iv)

In summary, for the 2e⁻ pathway in the ORR, elements that weakly activate the O-O bond should be selected to reduce O-O bond cleavage capability, stabilize the *OOH intermediate, and retain the O-O bond, thereby enhancing H₂O₂ selectivity. The asymmetric configurations composed of N and O atoms in structures such as Co-N₅, Zn-N₃O and other structures promote the two-electron pathway of the oxygen reduction reaction. For the 4e⁻ pathway, the O-O bond cleavage capability must be strengthened, and active sites should integrate dual functions of O₂ adsorption and O-O bond scission to promote rapid adsorption/desorption equilibrium of oxygen-containing intermediates (*O, *OH). The presence of B and Fe elements in structures such as Co-N₃B and Fe-B-Co/NC promotes the four-electron pathway of the oxygen reduction reaction. Additionally, precise regulation of defect type, location, and concentration can optimize the electronic structure and reaction kinetics of active sites, enabling directional selection of ORR pathways (2e⁻/4e⁻). Finally, the choice of carbon matrix must comprehensively consider the synergistic effects of multi-element doping, defect engineering, hierarchical structures, and coordination modulation to achieve electronic structure optimization, rapid mass transport, and efficient exposure of active sites.