The peripheral environment: key to advanced single-atom catalysis for CO₂ reduction

Abstract

Electrocatalytic reduction of CO₂ into value-added fuels and chemicals is a crucial pathway to achieving carbon neutrality and energy regeneration. Single-atom catalysts (SACs) are considered among the most promising catalysts for CO₂ electroreduction due to their high atom utilization and well-defined active site structures. However, the traditional design concept centered on the central metal atom and the first coordination shell is insufficient to fully explain the differences in activity, selectivity, and stability exhibited by catalysts under operating conditions. Here, we emphasize the role of peripheral factors, including higher coordination shells, support defects, pore confinement, and neighboring metal species. Through an “outside-in” mechanism, these factors regulate the electronic structure, adsorption behavior, and stability of single-atom sites, while their interplay with solvation, electrolytes, and interfacial electric fields under operating conditions dynamically reshapes the coordination environment and reaction energetics. We systematically summarize the experimental and theoretical progress of peripheral environment regulation for CO₂ electroreduction over SACs, including construction and identification of peripheral structures through precise synthesis and operando characterization, as well as theoretical understanding using electronic structure analysis, explicit solvent dynamics, and constant-potential simulations. The remaining challenges and opportunities in unified descriptor construction, dynamic mechanism analysis, and industrial applications, are also discussed. This perspective reframes single-atom catalysis by shifting the focus from the isolated metal center to the full catalytic active domain, providing a new design paradigm for efficient, stable, and selective CO₂ electroreduction.

The peripheral environment is the hidden key to advanced single-atom catalysis for CO₂ reduction.

1. Introduction

The electrocatalytic reduction of carbon dioxide (CO₂RR) into value-added fuels and chemicals is one of the central strategies for mitigating greenhouse gas emissions and achieving carbon cycling and carbon neutrality.^1,2 As the share of renewable energy increases, the efficient conversion of intermittent electricity into storable chemical energy has emerged as a critical challenge.³ Single-atom catalysts (SACs), featuring nearly 100% atom utilization, well-defined active-site structures, and tunable electronic configurations, are widely regarded as one of the most promising CO₂RR catalytic materials.^4–6 By precisely tuning the geometric coordination environment of the central metal atom and the Gibbs free energy matching between the active site and key reaction intermediates, SACs can effectively promote the selective conversion of CO₂ into CO, formate, methanol, and other value-added products.⁷

Precise synthetic methods and advanced characterization techniques provide the dual foundation for ensuring the well-defined structures and superior performance of SACs.^8–10 These advances enable the accurate identification of isolated metal atoms anchored on support materials, including their coordination geometry and electronic properties. This progress has driven the rapid development of a structural design strategy centered on the “central metal atom + first coordination shell”, a paradigm that has largely guided the rational design of SACs to date. However, this paradigm is showing its limits. A growing body of experimental and theoretical evidence indicates that focusing exclusively on the central metal atom and its first coordination shell cannot explain the substantial differences in catalytic activity, stability, and C₁/C₂₊ product selectivity observed under practical reaction conditions.¹¹

What causes these differences? The answer may lie in the “peripheral environment”, a broad category of factors located beyond the core metal–ligand coordination unit that can nevertheless influence individual single-atom sites through an “outside-in” mechanism.¹² To clarify this concept, we define the peripheral environment as the collection of structural and physicochemical factors that modulate the coordination structure, electronic properties, and reaction behavior of single-atom sites. As shown in Fig. 1, peripheral modulation refers to atoms and structural motifs within the catalyst that lie beyond the first coordination shell yet maintain electronic or structural coupling with the central metal. These include higher coordination shells,^13,14 support defects and edges,^15,16 pore confinement and framework effects,^17–19 as well as neighboring species or adjacent sites.^20,21 Beyond these relatively static structural features, the coordination environment of single-atom site can also be dynamically regulated by its surrounding solvent, electrolyte, pH, and interfacial electric field. Therein, interfacial water networks, hydrogen-bonding interactions, local ion distributions, and electric-field effects can alter the solvation structure, charge distribution, adsorption configurations, and coordination state of the active site under operating conditions.^22,23 To distinguish true peripheral effects from changes in the primary active site, we adopt the preservation of the core coordination motif as the primary practical criterion; that is, peripheral factors are modified while the central metal, the types of ligands in the first coordination shell, and the coordination number remain constant. When such controlled comparative experiments cannot be conducted, site-resolved structural, spectroscopic, or theoretical evidence may be used to determine the contribution of peripheral factors. Through these site-associated interactions, peripheral regulation can systematically modulate the electronic structure of the active center, intermediate adsorption, transition-state energetics, and proton-coupled electron transfer, thereby reshaping the free-energy landscape and kinetic pathways of CO₂RR through an “outside-in” mechanism.

Fig. 1. A schematic diagram of five key aspects affecting the peripheral effects of CO₂RR over SACs.

The central message of this perspective is simple but powerful: beyond the first coordination shell, the peripheral environment matters. Understanding single-atom catalysis requires looking beyond the isolated metal center to consider the full catalytic ensemble, including the central atom, its coordination shells, the support, and the dynamic reaction interface. This shift in focus not only represents a new frontier in single-atom catalysis but also directly addresses a critical bottleneck in the design of high-performance CO₂RR catalysts.

On this basis, we systematically summarize recent experimental advances in engineering peripheral structures, from higher coordination shells to pore confinement, and show how theoretical simulations have begun to unravel the “outside-in” electronic and dynamic mechanisms. By integrating these two perspectives, we establish a design framework that moves beyond the first coordination shell, offering enhanced explanatory power and predictive capability for future catalyst development. We argue that the traditional “central metal atom + first coordination shell” paradigm, while powerful, is incomplete. Accordingly, the peripheral environment should be considered alongside the first coordination shell as a key design variable.

2. Experimental evidence of peripheral environment effects

This section focuses on how the surrounding environment (i.e., peripheral structural modulation) regulates the catalytic performance of SACs in the CO₂RR, and discusses relevant experimental observations and rational design strategies. Three key questions guide our discussion: (i) how far can the influence of peripheral coordination extend? (ii) how do support defects reshape the electronic landscape of single-atom sites? (iii) can spatial confinement create catalytic environments unavailable to open surfaces? In general, peripheral structure modulation can be categorized into three types: coordination environment engineering, support and defect engineering, and spatial confinement and framework effects.

2.1. Coordination environment engineering

What is the reach of peripheral coordination? Single-metal atoms are anchored on supports via chemical bonds, which is critical for maintaining isolated dispersion.^24,25 Conventional wisdom holds that only the first coordination shell matters. However, mounting evidence challenges this view. Our previous study found the important role of peripheral nitrogen species on a single-atom Ru₁ center. The inner-shell N atoms coordinated with Ru₁ for stabilization, while outer-shell N atoms, located farther away, promoted electron accumulation at the Ru₁ center, enhancing charge repulsion between Ru₁ and product and thus promoting product desorption.²⁶ A further study showed that increasing the distance between the N dopant and the single-atom metal resulted in different energetic changes during the reaction. Notably, even stochastic, rather than precisely positioned, N dopants could tailor the d-band center of single-atom Ir₁, optimizing intermediate adsorption strength and enhancing propane dehydrogenation activity and selectivity.²⁷ These findings suggest that peripheral coordination effects operate over a longer range than previously appreciated, and that precise positioning may not always be necessary.

Taking a step further, we asked: can the geometry of peripheral rings, not just the identity of peripheral atoms, dictate catalytic outcome? To this end, we prepared Fe–priN and Fe–proN by pyrolysis of Fe(iii)-phenanthroline (Fe-phen) complex and Fe(iii) tetraphenylporphyrin (FeTpp) precursors, respectively.²⁸ AC HAADF-STEM confirmed the atomic dispersion of Fe (Fig. 2, Fe–proN vs. Fe–priN), while N 1s XPS and Fe K-edge XANES supported Fe–N₄ sites embedded in six- and five-membered ring frameworks. This method could be universal to prepare SACs with six-membered-ring or five-membered-ring N coordinated structures. Although the pyridinic and pyrrolic N atoms directly coordinated to Fe belong to the first coordination shell, they are embedded in distinct peripheral ring frameworks. As further discussed in Section 3.1, these ring geometries can alter the electronic states of the coordinated N atoms and the Fe–N interaction, thereby transmitting their influence inward to the Fe center through an “outside-in” pathway. The two catalysts with similar Fe contents were evaluated under comparable electrochemical conditions, providing a suitable paired system for assessing the influence of peripheral ring geometry. Compared with the Fe–priN catalyst, which features a six-membered-ring N coordination motif, the Fe–proN catalyst with a five-membered-ring N coordination structure not only enhanced the stability of the active center but also exhibited significantly higher activity toward CO₂RR. Fe–proN achieved an FE_CO of 73.1% at −0.6 V vs. RHE, compared with 24.3% for Fe–priN, and exhibited a 6.2-fold higher CO partial current density at −0.7 V. Critically, both catalysts share the same Fe–N₄ first coordination shell; only the peripheral ring structure differs. This provides unambiguous evidence that peripheral geometry alone can dramatically alter catalytic performance through electronic structure modulation.

Fig. 2. This comparative diagram of single-atom catalysts links the peripheral environmental engineering with key mechanistic steps of the CO₂ reduction reaction in SAC-based catalytic systems. Representative systems are positioned based on five categories of peripheral regulatory factors (highly coordinated shells, support defects/edges, confinement/framework, adjacent metals/nearby sites, and dynamic interfaces) and their main mechanistic steps (CO₂ activation, *COOH formation, CO desorption, C–C coupling, and proton transfer/interfacial transport). The diagram also indicates the main target products and their corresponding references. Each system is positioned according to the dominant role emphasized in the cited study, which does not exclude additional influence on other reaction steps.

Furthermore, the second layer of heteroatom doping provides a controllable strategy for adjusting the peripheral environment. Cu–C₃N₄–S catalyst was prepared via urea polymerization, heteroatom substitution, Cu impregnation, and heat treatment.²¹ AC HAADF-STEM confirmed the single-atom dispersion of Cu, while XAS, EXAFS, XPS, and solid-state NMR results collectively supported the formation of the Cu–N₃ first coordination structure and the peripheral S–N bonds, indicating that S is mainly located in the second coordination layer at Cu sites (the specific control strategies are summarized in Table 1). This catalyst achieved a C₂H₄ faradaic efficiency of 60.2% at −0.9 V vs. RHE and maintained over 58% after 24 h of electrolysis, demonstrating that the second-layer S environment can be experimentally constructed and used to improve the selectivity and stability of C₂H₄.

Table 1. Comparative summary of representative peripheral-environment strategies for CO₂RR over SAC-based catalytic systems. The comparison includes the peripheral motif, preservation of the primary coordination shell, synthesis/control strategy, key characterization, mechanistic role, catalytic performance, electrochemical conditions, and corresponding references^a.

Representative peripheral-environment strategies for CO2RR over SACs
Representative system	Peripheral motif	First shell held constant?	Synthesis/control strategy	Key characterization	Main mechanistic role	Product/key performance	Electrolyte/cell/	References
Fe–proN vs. Fe–priN	Peripheral ring geometry (5 membered vs. 6 membered N ring)	Yes (Fe–N4)	Different Fe–N precursors; pyrolysis	AC HAADF-STEM; XAFS	Tunes Fe–N interaction; optimizes *COOH/*CO adsorption	CO; FE 73.1% at −0.6 V	Aqueous CO2RR	28
Cu–C3N4–S	Second-shell S around Cu–N3	Approximately yes (Cu–N3 preserved)	Urea polymerization and heat treatment	HAADF-STEM; XAFS; ssNMR	Stabilizes CO-related intermediates; lowers C–C coupling barrier	C2H4; FE 60.2% at −0.9 V; >58% after 24 h	Aqueous CO2RR	21
CuN4/pr-h-NC	Pyrrolic-N/pyridinic-N triadic-site environment	Yes (Cu–N4 retained, N configuration varied)	CuPc-based synthesis + H2/Ar control of pyrrolic-N	AC HAADF-STEM; XAFS; ATR-FTIR; DEMS	Raises CO coverage and promotes *COH formation/C–C coupling	C2H5OH; FE 79.6%, j = 5.0 mA cm−2 at −0.6 V; FE >70% over 45 h	0.1 M KHCO3: H-cell; flow cell: 1 M KOH	66
Cu/CeO2–Ov	Oxygen vacancy adjacent to Cu single atom	Not strictly	Low-temperature annealing under different atmospheres	STEM; XAFS; defect-sensitive spectroscopy	Activates adjacent site; promotes deeper reduction	CH4-selective CO2RR	Aqueous CO2RR	32
e-Fe-NGM	Edge-rich graphene nanomesh surrounding FeN4 sites	Yes (Fe–N4 retained; edge carbon environment modified)	MOF-derived self-exfoliation for edge-rich Fe–N4 sites	AC HAADF-STEM; XAFS	Enhances *COOH formation, CO desorption, and interfacial proton transfer	CO; FECO >98%; TOF 6648 h−1; >1500 min stability	0.5 M KHCO3: H-cell; flow-cell	39
NiTPP in CNTs	Confinement + higher-shell π-network tuning	Yes (primary Ni–N; secondary C preserved)	CNT-confined NiTPP; thermolysis	AC HAADF-STEM; XAFS	Promotes *COOH formation and *CO desorption	Enhanced CO2-to-CO activity	Aqueous CO2RR	43
ZnPc-pz/CuPc-pz COFs	π-Conjugated framework/interlayer charge transport	N/A	Phthalocyanine-pyrene condensation	PXRD; Hall effect; THz spectroscopy; DFT	Framework conjugation governs charge transport	Not a CO2RR study; band gap ≈1.2 eV; mobility up to ≈5 cm2 V−1 s−1	N/A	47
FeNCs/FeSAs-NPC	Adjacent Fe nanoclusters as electron reservoir	Approximately yes (e-N atomic motif retained)	Supramolecular co-assembly + pyrolysis	AC HAADF-STEM; XAFS; ATR-IR	Electron reservoir lowers the *COOH formation barrier	CO; FE >90% from −0.40 to −0.85 V; 97.7% at −0.60 V	0.5 M KHCO3: H-cell	60
Ni1Fe1–NC	Neighboring Ni–Fe sites with tuned spacing	No (cooperative dual-site ensemble)	Fe in ZIF-8; dual-solvent Ni confinement; pyrolysis	AC HAADF-STEM; XAFS	Lowers *COOH barrier; promotes *CO desorption	CO; TOF 976 h−1 at −1.0 V	Aqueous CO2RR	20
Adjacent Cu–N3 sites	High-density neighboring Cu–N3 sites	Yes (Cu–N3 retained, site spacing varied)	Silica-mediated HOF-templated synthesis	AC HAADF-STEM/EELS; XAFS; in situ SERS	Promotes C–C coupling between adjacent Cu–N3 sites	C2H5OH; FE 81.9%; j = 35.6 mA cm−2; TOF 3.46 h−1; >25 h	0.5 M CsHCO3: H-cell	67
ZrO2/CuN1O2	Adjacent ZrO2-assisted dynamic reconstruction	No (CuN4 → CuN1O2)	PCN-222(Cu) pyrolysis + HER-induced reconstruction	AC HAADF-STEM; in situ Raman; XAFS	Forms stable Cu–N/O sites and optimizes intermediate adsorption	CH4; FE 87.3% at −500 mA cm−2, 80.2% at −1000 mA cm−2; >80% over 25 h	1 M KOH: Flow cell/MEA	74
Sn–N4–C vs. polyatomic Sn	pH-electric-field-coupled dynamic interface	Benchmark comparison of two active-site types	Coupled simulation and experimental validation	First-principles calculations; ML-force-field MD; microkinetics	Site-dependent *OCHO stabilization under interfacial field	Distinct pH-dependent HCOOH activity trends	Benchmark pH-dependent CO2RR study	80
Cu–N4–C	Dynamic electrode–electrolyte interface/H-induced reconstruction	No	Operando electrochemical environment	Constant-potential hybrid-solvation dynamic simulation	Promotes dynamic Cu reconstruction and C–C coupling through transient Cu clusters	C2+ products; dynamic active ensemble beyond initial SAC structure	Aqueous CO2RR	75
K+-regulated Fe site	Interfacial ion accumulation and H-bond-network reorganization	Yes (active Fe site retained)	Electrolyte-cation engineering	Interfacial analysis + simulation	Exposes Fe site; facilitates CO2 access; suppresses HER	Improved acidic CO2RR selectivity	Acidic CO2RR	77
C3–Ni–Cl	Asymmetric carbon-vacancy environment	No	Carbon-vacancy engineering	Local coordination analysis; XAFS	Asymmetric electron distribution and abundant CO2 adsorption space	CO; near 100% selectivity in acidic media	Acidic electrolyte	38

^a“Held constant?” indicates whether the primary first coordination shell is unchanged in the comparison. This operational criterion helps distinguish peripheral effects from direct first-shell changes.

From another perspective, the addition of a second metal in the periphery can also affect the properties of the isolated metal center. For example, by first incorporating Fe into ZIF-8, then confining the Ni precursor within the framework using a dual-solvent method, and finally preparing Ni₁Fe₁–NC diatomic catalysts with different atomic spacings via high-temperature pyrolysis, a Ni₁Fe₁–NC diatomic catalyst was obtained (Fig. 2, Ni₁Fe₁–NC).²⁰ The N-bridged Ni–N–Fe sites with an appropriate Ni–Fe distance achieved a TOF of 976 h⁻¹ at −1.0 V vs. RHE, both substantially higher than those of the corresponding Ni–N–C and Fe–N–C catalysts.

Together, these studies demonstrate that coordination environments beyond the first shell are pivotal for optimizing CO₂RR performance. The emerging picture is that peripheral coordination, whether through distant heteroatoms, ring geometry, or second-metal neighbors, exerts measurable and often decisive effects on catalytic behavior, even when the first coordination shell remains unchanged.

2.2. Support and defect engineering

Beyond coordination atoms, can defects in the support, which are often considered undesirable, be turned into an advantage? For carbon materials prepared by pyrolysis, intrinsic defects such as edges, vacancies, holes, and topological defects are inevitable. These defects can break the integrity of the structure and charge state of the support. Rather than viewing defects as nuisances to be eliminated, recent studies have turned them into design levers.²⁹ The synergy between Cu single atoms and oxygen vacancies in UiO-66 promoted the formation of *HCCOH and subsequent ethanol production.³⁰ In another case, atomically dispersed Cu on defective black anatase TiO₂ was favorable for CO₂RR to CH₄, with oxygen vacancies enhancing hydrogen supply by promoting water dissociation.³¹ Moreover, oxygen vacancy concentration can be manipulated through varying reaction atmosphere such as air, Ar, and H₂. As shown in Fig. 2 (Cu/CeO₂–O_v), single-atom Cu on ceria under an H₂ atmosphere, which generated the highest concentration of oxygen vacancies, facilitated CO₂ activation and conversion to CH₄.³² These examples reveal a general strategy: oxygen vacancies can serve as electron donors or hydrogen-bonding sites to modulate the local environment of adjacent single atoms.

Beyond oxygen vacancy, nitrogen or carbon vacancy is a common type of defect in carbon materials. Under continuous plasma striking, pyrrolic N-defects around Ni sites could be transformed into a stable pyridinic N dominant Ni–N₂ structure.³³ These defective N sites provided abundant space for CO₂ adsorption and activation, leading to a lower energy barrier for CO₂ electroreduction to CO. Unsaturated single-atom sites, such as Ni–N₃-vacancy and Ni–N₂-vacancy, are favorable for the production of more unpaired 3d electrons, thus reducing the free energy for key intermediate formation and boosting CO₂RR performance.^34–36 For a single-atom Fe₁ catalyst, a peripheral N vacancy promoted electron delocalization to adjacent Fe₁, which was efficient for the activation of reactants.³⁷ Constructing an asymmetric carbon vacancy in SACs is also an effective way for improving CO₂RR performance. Lu et al. prepared C₃–Ni–Cl single atoms via carbon vacancy engineering, and asymmetric electron distribution in C₃–Ni–Cl promoted CO₂ electroreduction, even in strongly acidic media (pH = 1), realizing near 100% CO selectivity.³⁸ In some cases, single-metal sites at the edges of graphene-like carbon supports are considered more active than those in-plane. It was reported that edge-type FeN₄ sites on the edge-rich N-doped graphene nanomesh induced a strong local electrostatic field (Fig. 2, e-Fe-NGM), promoting the supply of interfacial H₂O and the protonation process of CO₂.³⁹

These findings not only highlight the critical role of vacancies in the activation of reactants but also elucidate their significance in the design and structural regulation of SACs. The emerging picture is that defects, traditionally viewed as imperfections, can be powerful design tools when harnessed deliberately, creating unsaturated coordination environments and modulated electronic structures that are often inaccessible in perfect lattices. Local support features, such as nearby vacancies, edges, and strain, are therefore considered peripheral factors only when their contribution to a specific single-atom site can be distinguished from changes in the primary active-site structure by controlled comparison or site-resolved evidence.

2.3. Spatial confinement and framework effects

Can confining single atoms within porous frameworks create catalytic environments that are fundamentally different from open surfaces? Metal–organic frameworks (MOFs) consist of polydentate organic ligands and metal-containing inorganic building blocks, forming porous and rigid three-dimensional interconnected frameworks.⁴⁰ Covalent organic frameworks (COFs) are emerging organic molecular frameworks linked by covalent bonds, with tunable micro/macro pores, high surface area, good chemical stability, and an adjustable internal chemical environment.⁴¹ Moreover, carbon nanotubes can provide confined space and stabilize single atoms. Therefore, pores and framework structures cannot be regarded as catalyst supports alone. When the constraints around identifiable single-atom sites, such as local pore geometry and framework ligands, directly regulate the electronic state or reaction behavior of active centers, materials such as MOFs, COFs, and carbon nanotubes can show great potential in regulating the microenvironment of single-atom active centers.⁴²

Wu et al. used carbon nanotubes confined nickel tetraphenylporphyrin as the precursor. Through temperature-dependent thermolysis, they preserved the primary Ni–N coordination (first shell) and secondary C coordination (second shell), while systematically eliminating π-electron networks in higher coordination shells (Fig. 2, NiTPP in CNTs). Experimental and theoretical results revealed that the elimination of long-range π-electrons delocalization promoted *COOH formation and *CO desorption. This work elucidates the key role of higher coordination shells in tuning carbon nanotube confined SACs.⁴³ Ni ions confined within the pores of ZIF-8 could occupy the Zn nodes and form Ni–N coordination, endowing excellent CO₂RR activity.⁴⁴ Despite significant progress, the insulating properties and long yet narrow diffusion channels of MOFs result in low single-atom utilization. Notably, downsizing MOF particles into quantum dots shortens the mass transfer route and enhances electron transfer, resulting in higher active-site utilization and improved CO₂RR performance.⁴⁵ Furthermore, Chen et al. found dual Fe sites on ordered channel walls of MOFs enabled efficient C–C coupling to C₂H₅OH and C₂H₄.⁴⁶

COFs featuring accessible active single-sites may serve as a bridge between homogeneous and heterogeneous electrocatalysis. As shown in Fig. 2 (ZnPc-pz/CuPc-pz COFs), consisting of Zn- and Cu-phthalocyanine based pyrazine-linked COFs had a negligible effect on the electrical properties, showing potential as semiconductors for electronics.⁴⁷ When single-atom Fe was anchored on COFs, the synergy between Fe single atoms and COF hosts reduced the reaction energy barrier.⁴⁸ Besides, Ni-phthalocyanine-based COF linked by pyrazine units was designed for CO₂RR. Owing to the good electron transfer ability and facilely accessible active sites, NiPc-COF exhibited high CO selectivity and long-term stability.⁴⁹ Using interlayer π–π stacking, engineering single-atom metal Cu, Ni, and Fe on the bipyridine-linked 2D conjugated COF was efficient for C–C coupling to yield ethanol, originating from the cooperation of two adjacent sites from neighboring layers.⁵⁰ In this context, neighboring metal atoms are considered peripheral regulators only when they remain structurally distinct and modulate an identifiable single-atom center; when two metal atoms directly cooperate in intermediate adsorption or C–C bond formation, the active structure is more appropriately described as a cooperative multi-site ensemble. Moreover, the porous and highly ordered Mn PhenPy COF showed superior performance for CO production.⁵¹ This represents the first successful application of a COF-based catalyst in a catholyte-free MEA configuration, opening a new avenue for advancing CO₂ electroreduction technologies.

To sum up, N atoms are commonly used to anchor single-metal atoms through direct coordination with the metal in the first shell. However, the peripheral environment, involving the distribution, type, and distance of N atoms around the single-atom center, plays a pivotal role in determining electronic structures and catalytic performance. This challenges the assumption that the first coordination shell alone dictates catalytic behavior. In addition, the peripheral defects, such as oxygen vacancies, N or C vacancies, and pore vacancies, also exert a clear influence on the metal center. Importantly, peripheral environment regulation may also introduce trade-offs. Defects, neighboring sites, ligands, pores, and reconstructed interfaces can enhance activity or selectivity, but may simultaneously promote metal leaching, aggregation, coordination loss, pore blocking, or irreversible structural changes. Thus, peripheral environments should be evaluated by balancing catalytic performance, structural stability, site accessibility, and mass transport. Rather than viewing these trade-offs as mere drawbacks, they define an optimal “design window” within which peripheral engineering maximizes catalytic benefit without compromising operational integrity.

To apply this criterion across representative systems, Table 1 compares whether the primary coordination shell is preserved during peripheral regulation, together with the corresponding synthesis/control strategy, structural characterization, mechanistic role, catalytic performance, and electrochemical conditions. This comparison distinguishes peripheral regulation with a retained primary coordination motif from cases involving active-site reconstruction or cooperative multi-site structures.

Overall, the peripheral effect of SACs can be understood in terms of three aspects: (i) long-range electronic induction. Long-range interaction between the metal center and the peripheral atom could change the electron distribution, d-band center, and oxidation state of single-atom sites, thereby optimizing their electronic features and catalytic activity. (ii) Synergistic behavior with single-metal atoms as an active center. The orbitals of the metal center that participate in hybridization with reactive species can shift with the presence of peripheral atoms, changing orientation or promoting desorption, thereby regulating the activity and selectivity. (iii) Confinement and conductivity enhancement. In the case of single-atom sites confined within MOFs, COFs, and carbon nanotubes, neighboring metal nodes or ligands induce both electronic and steric effects. Peripheral atoms can serve as an electron donor or acceptor to modulate electron density around the Fermi level. For MOFs and COFs in particular, the design of electron donor–acceptor conjugations offers a promising route for enhancing the conductivity.

3. Theoretical understanding of peripheral environment effects

An in-depth understanding of the peripheral effect in SACs is crucial for precise catalyst design and uncovering reaction mechanisms. To understand why these effects occur, we turn to theoretical simulations. It is well accepted that the catalytic activity and selectivity of SACs fundamentally depend on the electronic structure of the metal center. Beyond the intrinsic active site, the higher-dimensional peripheral microenvironment can dynamically couple with the catalytic center, endowing the system with new catalytic properties. Theoretical calculations enable a quantitative understanding of how electronic structures govern the adsorption of key intermediates and the associated reaction energy barriers. This section systematically summarizes recent first-principles studies, progressing from the electronic origin of peripheral effects (Section 3.1), to their manifestation in reaction selectivity (Section 3.2), and finally to their dynamic evolution under operating conditions (Section 3.3).

3.1. Electronic structure as the primary relay

According to the classical d-band model, the position of the d-band center is strongly correlated with adsorption strength. With a nearly constant bandwidth, an upshift of the d-band center typically lowers the orthogonalization energy while enhancing the hybridization energy, thereby strengthening adsorbate-surface interactions.^52,53 Therefore, the first coordination shell can directly regulate the d-band center through strong d–p orbital hybridization with the metal center, inducing charge redistribution. But can peripheral species achieve similar modulation without direct bonding? Theoretical calculations answer yes.

Beyond the first coordination shell, peripheral species can further modulate the electronic state of the metal center via long-range interactions, thereby influencing catalytic performance.⁵⁴ Second-shell S or P doping is a representative example, but its effect should not be viewed as a simple and universal electron-donating or electron-withdrawing process. Instead, the modulation direction depends on the intrinsic electronic structure of the central metal, the distribution of metal-related states, and the key reaction step to be optimized. In Fe–N–C systems, where CO₂ activation and *COOH formation are critical for CO production, S or P doping can polarize the surrounding N/C framework, enhance electron donation from Fe sites to CO₂ or *COOH, stabilize *COOH, and lower its formation barrier. S-doped FeN₄ systems further show a proton-feeding effect by promoting water dissociation, while P-doped FeN₄P enhances electronic localization at Fe sites, facilitating CO₂ adsorption and activation.^55,56

In contrast, Ni single-atom sites require a balance between *COOH formation and *CO desorption; thus, second-shell S doping can downshift the Ni d-band center, weaken Ni–CO binding, and promote CO release while still assisting *COOH formation.⁵⁷ Nevertheless, the d-band center should be used only as a qualitative descriptor in such systems, rather than as a universal explanation for all SACs. For non-transition metal-based SACs, the conventional transition-metal d-band model is less applicable. In P-SnN₄-CNT, second-shell P regulates Sn-related s/p states and promotes the reaction-induced conversion of Sn⁴⁺ to Sn²⁺, which is considered the active state for CO formation.⁵⁸ Therefore, the different electronic responses induced by S/P doping in Fe, Ni, and Sn systems are not contradictory, but reflect the metal-dependent nature of peripheral modulation. These interpretations are further supported by XAS/EXAFS, in situ spectroscopy, kinetic analysis, and quasi-in situ characterization, which provide evidence for second-shell structures, intermediate evolution, proton-feeding behavior, and valence-state changes.

In addition, M–N–C systems supported on different carbon frameworks exhibit distinct catalytic performances toward CO₂RR.⁵⁹ This suggests that peripheral carbon species surrounding the metal sites play a crucial role in determining CO₂RR activity. This effect is illustrated by the Fe–proN and Fe–priN systems reported by Zhao et al.²⁸ Although the pyridinic and pyrrolic N atoms directly coordinate with Fe, their electronic states are influenced by the five- and six-membered ring frameworks in which they are embedded. Theoretical calculations reveal that pyrrolic N within five-membered rings enhances d–p coupling with the metal center, stabilizes Fe in a higher oxidation state, and optimizes the adsorption of key intermediates such as *COOH and *CO, thereby enhancing catalytic activity for CO₂ conversion to CO. This electronic relay from the peripheral ring framework through the coordinated N atoms to the Fe center represents an “outside-in” regulation pathway (Fig. 3a) and explains the experimental performance difference between Fe–proN and Fe–priN discussed in Section 2.1. Conversely, suppressing π-electron delocalization by cleaving peripheral β-carbons in NiTPP enriches electron density at the Ni site and stabilizes the *COOH intermediate, thereby significantly improving catalytic performance.⁴³

Fig. 3. Schematic illustration of the “outside-in” regulation of single-atom centers through (a) peripheral coordination, (b) neighboring sites, (c) surrounding environment, and (d) interfacial microenvironment modulation. These regulation modes tune the electronic structure, catalytic synergy, dynamic evolution, and local environment of single-atom centers, thereby enhancing the activity, selectivity, and stability of CO₂ electroreduction.

Furthermore, peripheral modulation is not limited to the intrinsic catalyst framework. Functional groups or neighboring clusters can regulate the electronic structure of the metal center through electron-donating and electron-withdrawing effects. For example, Fe clusters near Fe atoms can act as electron pools, increasing the electron density of Fe atoms at the reaction center, providing more favorable conditions for adsorbed intermediates, thereby improving reaction activity (Fig. 2, FeNCs/FeSAs-NPC, and Table 1 summarizes the control strategies).⁶⁰ Moreover, Ni nanoclusters can alter the electronic structure of Ni atomic sites. Ni nanoclusters disrupt the overall charge symmetry of individual Ni atoms. This causes more charge to accumulate on the side closest to the Ni nanoclusters, leading to enhanced adsorption of intermediates. The reduction in charge at Ni atomic sites indicates that the Ni nanoclusters provide additional electrons for the entire CO₂RR process.⁶¹ In Co-porphyrin systems, the introduction of nitro ligands induces d-orbital splitting, facilitating electron transfer from the Co 3d_z² orbital to the CO₂ π* orbital, thereby strengthening CO₂ binding.⁶²

Collectively, these findings demonstrate that peripheral effects extend beyond conventional heteroatom doping. Adjacent clusters, ligands, and functional groups can finely tune the electronic structure of single-atom sites through electronic induction, orbital coupling, and delocalization effects, thereby markedly influencing catalytic activity and selectivity. These studies indicate that electronic-structure modulation is a major pathway through which peripheral factors influence catalytic behavior.

3.2. Steering reaction paths: from C₁ to C₂₊ selectivity

Electronic-structure modulation can alter reaction pathways by changing the relative stability of key intermediates and transition states. Theoretical studies have revealed how specific peripheral environments selectively stabilize different reaction intermediates. For example, a peripheral environment that slightly weakens the binding of *CO may facilitate its desorption, thereby promoting the formation of CO as the final product.⁶³ In contrast, a peripheral environment that optimizes *CO adsorption at specific surface coverages can promote subsequent *CO dimerization.⁶⁴ SACs, with high atomic efficiency, tunable electronic structures, and well-defined active sites, exhibit distinct advantages in CO₂ activation and intermediate regulation, enabling efficient production of C₁ products such as CO. However, the isolated nature of single-atom sites limits cooperative interactions required for C–C coupling, thereby restricting the efficiency and selectivity toward C₂₊ products.⁶⁵ Neighboring metal sites and support structures can partly overcome this limitation by introducing cooperative interactions (Fig. 3b).

Recent studies indicate that the peripheral environment can alter the reaction pathways for C₂₊ products by modulating the energy states of intermediates and their spatial transformation processes. In the Cu–C₃N₄–S system (Fig. 2, Cu–C₃N₄–S), S atoms in the second coordination shell exert electronic control over the central Cu site. This peripheral doping enhances the Cu site's ability to activate CO₂ and optimizes the stability of CO-related intermediates, thereby lowering the C–C coupling energy barrier to 0.67 eV—compared to 0.74 eV and 0.78 eV for the corresponding P- and B-doped systems.²¹ Experimental analysis reveals a configuration where the Cu center retains a Cu–N₃ first coordination shell structure, while the S atom is positioned in the periphery, bonded to a coordinating nitrogen. This second-shell structure facilitates C₂H₄ formation by promoting CO₂ activation and stabilizing CO-derived intermediates. Furthermore, in Cu–N–C systems rich in pyrrolic nitrogen (Fig. 2, CuN₄/pr-h-NC and Table 1), pyrrolic-N lowers the energy barrier for *COOH formation and increases CO coverage. Subsequently, CO migrates to an adjacent pyridinic-N site and is protonated to form COH; this COH species then undergoes *CO–*COH coupling with another CO at the single-atom Cu site to form *COCOH, thereby promoting ethanol production.⁶⁶ These findings highlight that the peripheral environment can modulate metal center activity not only through electron transfer but also thereby influencing C₂₊ product selectivity.

Spatial proximity between active sites provides a complementary route for promoting C–C coupling. Xia et al.⁶⁷ showed that increasing the Cu–N₃ site density shortens the distance between neighboring Cu sites and increases the probability of *CO–*CO interaction (Fig. 2, adjacent Cu–N₃ sites). DFT calculations revealed that C–C coupling is unfavorable at isolated Cu–N₃ sites (ΔG = 1.704 eV) but becomes thermodynamically favorable at adjacent double Cu–N₃ sites (ΔG = −0.976 eV). Structural relaxation further lowers the coupling energy from 0.844 to 0.647 eV, emphasizing the importance of both site proximity and structural flexibility. Experimentally, while the higher copper loading increased the number of available sites and may have partially contributed to the increase in current density, the ethanol turnover frequency (TOF) also increased from 0.63 h⁻¹ to 3.46 h⁻¹. Combined with the more favorable C–C coupling energy at adjacent Cu–N₃ sites, this result supports an intrinsic proximity-dependent synergistic effect, rather than simply an increase in site density. Moreover, different metals exhibit distinct binding strengths toward key intermediates. Dual-atom systems can function as tandem catalytic sites, modulating reaction pathways or tuning *CO coverage, thereby promoting subsequent *CO dimerization.⁶⁸ These theoretical insights directly support the experimental observation (Section 2.1) that Ni₁Fe₁ dual-atom catalysts with an appropriate Ni–Fe distance enhance CO₂RR performance through synergistic effects. These examples indicate that neighboring metal atoms can function as peripheral components by regulating or cooperating with atomically dispersed metal centers. When several sites directly participate in the reaction, they are more appropriately described as cooperative active ensembles rather than isolated single-atom sites.

Furthermore, the broader support structure can also influence catalytic behavior even when the local coordination environment is similar. Even with similar local active sites, variations in peripheral structures can markedly alter catalytic behavior. Cui et al.⁶⁹ demonstrated that for carbon-based SACs with similar Fe–N₄ sites, the global support structure (e.g., N-doped carbon vs. phthalocyanine) induces distinct potentials of zero charge (PZC). This difference modulates interfacial electric fields and solvent orientation, thereby altering the energetics and activation barriers of competing pathways (e.g., *CO protonation to *CHO vs. *CO dimerization). In more precisely engineered environments, single-atom sites embedded in two-dimensional organic frameworks or multicoordinated scaffolds can exhibit reaction pathways distinct from conventional catalytic systems. Goddard et al.⁷⁰ constructed a class of two-dimensional metal–organic framework (2D-MOF) models based on metalloporphyrin units and revealed through grand-canonical quantum mechanical (GC-QM) calculations under constant potential conditions that organic ligands (e.g., porphyrins) can introduce reaction pathways distinct from those on metal surfaces, enabling the catalytic centers in MOFs to exhibit single-atom-like catalytic behavior, thereby facilitating the formation of C₂ products.

Collectively, these studies demonstrate that the “outside-in” peripheral modulation strategy not only regulates the electronic structure of SAC active sites but also significantly influences CO₂RR pathways and product selectivity. Changes in the peripheral environment modulate the electronic structure of the support and the spatial distribution of single atoms, and further influence metal–ligand bond lengths, electron donation, and site cooperativity, thereby enabling the synergistic optimization of catalytic activity and product selectivity. Thus, electronic restructuring directly translates into observable changes in reaction outcomes.

3.3. Periphery-interface dynamics: the electrochemical microenvironment

The discussion so far has established that peripheral factors alter electronic structure and reaction pathways. However, both the electronic structure and electrocatalytic pathways are not static. Instead, they evolve under operating conditions. The peripheral effects should not be simply understood as static structural design, but rather as emergent regulatory outcomes arising from the strong dynamic coupling between the catalyst and its surrounding environment during the reaction process.⁷¹ This realization marks a fundamental shift. We no longer simply design static structures, but explore the dynamic coupling of electrocatalytic systems, in which the peripheral microenvironment modulation plays an active and constantly evolving role.

Under operating conditions, dynamic structural evolution induced by the interfacial microenvironment is both crucial and universal in determining the overall performance of electrocatalysts. For example, Zhang et al.⁷² proposed an H activation mechanism. This study demonstrated that *H on metal surfaces is not merely a competing species, but rather an invisible dominator that drives the structural evolution of catalyst surfaces. *H cooperates with reaction intermediates to induce a dissolution–redeposition cycle on the metal surface, revealing that hydrogen gradually weakens Cu–Cu bonds and evolves from a surface state to a subsurface state under high potentials, thereby serving as the fundamental driving force for surface reconstruction.

Similarly, in single-atom catalysts, catalytic active sites can also undergo dynamic reconstruction under the influence of the surrounding environment (Fig. 3c). Hydrogen radicals in the solvent environment can attack the Cu–N bonds in Cu–N–C catalyst systems, leading to the leaching of Cu²⁺ species and thereby weakening the structural stability of the catalyst.⁷³ In composite systems, the partial cleavage of weakened Cu–N bonds generates unsaturated Cu sites. These newly formed unsaturated sites can further undergo dynamic reconstruction with neighboring metal oxide clusters, resulting in the formation of new Cu–N/O hybrid structures (Fig. 2, ZrO₂/CuN₁O₂). Consequently, the d-band center of Cu shifts closer to the Fermi level, leading to stronger adsorption toward key reaction intermediates.⁷⁴ Under negative potentials, copper atoms bound to the catalyst surface can collide with dissolved copper species in the aqueous solution to form transient Cu cluster structures, which serve as the true catalytic active sites for CO₂ reduction to ethanol. After potential reversal, these clusters can redisperse into isolated Cu atoms under the dominance of hydroxyl radicals (Fig. 2 Cu–N₄–C).⁷⁵ These transformations demonstrate the strong coupling between the central metal atoms and their surrounding environment. In particular, the transient Cu clusters should be understood as dynamically reconstructed active ensembles derived from an initially atomically dispersed catalyst, rather than as static single-atom sites. The key question then is whether we can predict and control this dynamic evolution of the peripheral environment, or whether we are merely observing it after the fact.

However, when this dynamic evolution leads to the irreversible loss or transformation of original active sites, it can have adverse effects. Under prolonged reducing conditions, the breaking of M–N and N–C bonds can disrupt the stability of the M–N–C structure, resulting in the loss of coordinated nitrogen species, metal leaching, and the transformation of isolated metal atoms into particles. In the Cu–N–C system,^73,74 adsorbed *H can weaken or break Cu–N bonds, reducing the number of stable active sites. Meanwhile, CO-containing intermediates may promote the detachment and aggregation of Cu species to form clusters, impairing the long-term stability of single-atom catalysts. Therefore, effective peripheral regulation requires both reversible structural evolution and a balance between single-atom stability and reactant accessibility.

Beyond active site reconstruction, the interfacial microenvironment itself undergoes dynamic reconstruction during the CO₂RR process, and such reconstruction can in turn influence the catalytic activity and selectivity of active sites from the outer environment inward (Fig. 3d). First, interfacial water molecules, as a major component of the reaction environment, can directly affect the CO₂RR performance of SACs. For example, Zhao et al.⁶³ found that water molecules can co-adsorb with intermediates such as *CO₂ and *COOH at active sites, which directly alters the electronic structure of the central atoms and exhibits metal-dependent regulation of catalytic activity that is unique to Ru. Moreover, water molecules can both coordinate with active sites and form extensive hydrogen-bonding networks around them, thereby indirectly influencing catalytic activity. In directly S-coordinated Ni–N catalysts (NiN₃S₁), the dynamic rearrangement of surrounding water molecules forms a localized covalent environment. Under different applied potentials, this localized environment cooperatively regulates the activation of CO₂ at Ni sites together with the active center, thereby enhancing electron transfer between Ni and CO₂.⁷⁶ This reveals a hierarchical dynamic coupling within the peripheral environment. Interfacial water therefore acts not only as a solvent but can also participate directly in intermediate stabilization, hydrogen-bonding, and proton-transfer processes near the active site.

Electrolyte ions and local pH can further regulate the interfacial water structure. Within the interfacial microenvironment, electrolyte ions can reconstruct the interfacial solvation environment through electrostatic interactions, thereby indirectly altering the intrinsic selectivity of SACs. As shown in Fig. 2 (K⁺-regulated Fe site), in the absence of K⁺ ions, interfacial water molecules form a dense and ordered hydrogen-bonding network that blocks the Fe active centers, hinders the approach of CO₂, and favors proton transfer for the hydrogen evolution reaction (HER). In contrast, the introduction of K⁺ ions disrupts the near-surface hydrogen-bonding network and attracts water molecules to form hydration shells around the ions, thereby exposing the Fe active sites and facilitating the approach and adsorption of CO₂, which consequently promotes CO₂RR.⁷⁷ In fact, the specific ion identity (Li⁺, K⁺, Na⁺, etc.) also exerts a dynamic regulatory influence on the microenvironment at the electrocatalytic interface, thereby affecting reaction processes such as HER and CO₂RR.^78,79 Beyond ion effects, recent simulation studies on pH-electric field coupling further indicate that external environmental modulation also influences site activity. Wang et al. discovered that the binding strength of OCHO species responds differently to the interfacial electric field on Sn–N₄–C SACs versus multi-atom Sn structures, resulting in markedly different pH-dependent activity trends (Fig. 2, Sn–N₄–C vs. polyatomic Sn).⁸⁰ These results demonstrate that external operational variables can remotely regulate specific sites and their surrounding microenvironments by altering factors such as local pH and the interfacial electric field.

These findings also illustrate the importance of matching computational models to the dynamic nature of the electrochemical interface. However, the predictive scope of computational studies is inevitably influenced by the balance between model realism and computational cost. Simplified active-site structures and idealized supports enable systematic screening, but may overlook site heterogeneity, defect distributions, and reaction-induced reconstruction. Similarly, implicit or limited solvation treatments capture general dielectric stabilization but cannot fully represent specific hydrogen-bond networks, ion accumulation, and proton-transfer processes. Fixed-electron or fixed-charge calculations may also deviate from electrochemical conditions, where the electrode charge changes with the applied potential.

Recent studies have addressed these issues by using experimentally constrained structural models and configurational sampling, hybrid or explicit-solvent simulations, and constant-potential or grand-canonical DFT. Explicit-solvent AIMD and enhanced-sampling methods further allow dynamic solvent reorganization, structural fluctuations, and reaction free-energy barriers to be evaluated, while microkinetic modelling connects elementary energetics with coverage-dependent activity and selectivity. Machine-learning potentials are increasingly used to extend the accessible time and length scales. A hierarchical combination of these approaches is therefore needed, because increasing realism in one aspect does not by itself remove uncertainties associated with active-site structure, interfacial models, and sampling.

Therefore, the electrocatalytic activity of SACs originates from the combined effects of the central metal, its coordination shell and peripheral microenvironment. Active sites and supporting environments cannot be considered in isolation. Thus, when understanding and designing single-atom catalysts, their inherent structural features should not be separated from the external effects and the dynamic evolution of the interfacial microenvironment. Instead, these systems should be viewed as a whole, where the central metal, its coordination shell, the support, and the dynamic interface of water molecules and ions collectively constitute a complete catalytic system. This integrated perspective transforms our understanding from passive observation, where we simply note that the peripheral environment exists, to active design, where we engineer its dynamic evolution. The next frontier is not just characterizing the peripheral environment, but predicting and programming its behavior.

4. Challenges and future perspectives

This Perspective has emphasized that the performance of SACs is determined not only by the central metal and its first coordination shell, but also by peripheral factors, including higher coordination shells, support defects, pore confinement, and neighboring metal species, as well as by the surrounding peripheral microenvironments. Experimental studies demonstrate that precise synthesis and in situ/operando characterization techniques can effectively identify and construct diverse types of peripheral environments and reveal their critical roles in regulating catalytic activity, product selectivity, and structural stability. Theoretical investigations, through electronic structure analysis, reaction energetics calculations, explicit-solvent ab initio molecular dynamics, and constant-potential simulations, further elucidate how the peripheral environment modulates the electronic structure of the central metal, the stability of reaction intermediates, and reaction pathways through an “outside-in” mechanism.

On this basis, we propose a framework of “peripheral structure – electronic regulation – reaction pathway – product selectivity”, elevating peripheral effects from scattered empirical observations to a catalytic design principle with both explanatory and predictive power. This framework provides a new theoretical foundation for the development of CO₂ electroreduction catalysts with high activity, excellent stability, and superior selectivity. Peripheral environment engineering is expected to become a central direction for the rational design of SACs, but several key challenges remain to be addressed (Fig. 4):

Fig. 4. This schematic diagram illustrates the challenges that urgently need to be addressed in the development of single-atom catalysts with peripheral effects for CO₂RR.

(i) Unified descriptors for peripheral effects. A key challenge is to establish quantifiable descriptors that link peripheral structures with electronic properties and catalytic performance. Compared with the first coordination shell, which can be described by bond lengths and coordination numbers, the peripheral environment contains more complex factors, including ring geometry, defect density, pore size, and interfacial water structure.⁸¹ Thus, effective descriptors should be predictive, experimentally accessible, and able to reflect synergistic effects among multiple peripheral factors. Traditional DFT calculations are important for understanding peripheral effects by revealing adsorption configurations, electronic structures, and reaction energetics. However, their predictive power is limited by the vast structural space, high computational cost, and simplified models, including idealized supports, static structures, limited solvation, fixed charge states, and simplified interfaces. Machine learning methods, such as active learning, surrogate models, and graph neural networks, can complement DFT by rapidly exploring diverse metal centers, dopants, coordination environments, defects, and neighboring sites.⁸² These methods may help identify interpretable descriptors linking peripheral structures with adsorption energies, reaction barriers, electronic properties, and product selectivity, but still require high-quality datasets, uncertainty assessment, and experimental validation.

(ii) Deep integration of experiment and theory. While precise synthesis, in situ/operando characterization, and advanced computational approaches have each made significant progress, they are often applied in isolation. The central question of how to design the peripheral environment cannot be answered by any single tool alone. A key future direction is the seamless integration of these tools, using theory to guide synthesis, in situ characterization to validate theoretical predictions, and machine learning to extract patterns from complex experimental data. This integrated approach is essential for systematically analyzing realistic active structures and dynamic interfaces under operating conditions. In this context, automated and robotic synthesis platforms⁸³ can systematically control parameters such as precursor ratios, auxiliary molecules, dopant sources, pyrolysis conditions, reaction atmospheres, and post-processing. These parameters are closely related to the formation of the second-layer environment, defects, and pore structures. By combining standardized structural characterization, electrochemical testing, and theoretical analysis, such platforms can decouple complex, intertwined peripheral effects and help establish clearer “synthesis-structure-performance” relationships.

(iii) Understanding dynamic processes. The static view of SACs is increasingly giving way to a dynamic picture in which active sites undergo continuous transformation, single-atom-to-cluster interconversion, reconstruction of interfacial water networks, and potential-driven structural evolution. Understanding dynamic interfaces requires answering three sequential questions. First, which components of the peripheral environment evolve? The answer includes active sites, water networks, and ion distributions. Second, how fast do these peripheral components evolve? Timescales range from picoseconds to seconds, spanning orders of magnitude that challenge both experiment and theory. Third, can we control the evolution of the peripheral environment through external variables such as applied potential, electrolyte composition, or peripheral design? Elucidating the fundamental roles of these dynamic processes in stabilizing key intermediates and facilitating C–C coupling remains a major scientific challenge. Meeting this challenge will require the development of operando techniques with high temporal and spatial resolution, as well as theoretical methods capable of simulating catalytic cycles on operationally relevant timescales.

(iv) Bridging the gap to practical applications. Most fundamental studies are conducted under idealized conditions, such as low current densities, short operation times, and simplified electrolyte compositions. Validating peripheral environment design strategies under practical operating conditions (high current densities, prolonged operation, and technically relevant electrolyte compositions) is essential for translating laboratory discoveries into industrial applications. To establish the practical relevance of peripheral environment engineering, device-level metrics, including technically relevant current densities, energy efficiency, CO₂ utilization, product crossover, membrane compatibility, and long-term stability, should be evaluated in addition to intrinsic activity and selectivity. Peripheral environment engineering may influence these metrics by regulating catalyst-layer wettability, local CO₂ availability, water and ion transport, product desorption, and the stability of isolated metal sites under high-rate operation. However, practical performance is jointly determined by the catalyst, porous electrode, electrolyte, membrane, and reactor configuration. Future studies should therefore integrate peripheral catalyst design with gas-diffusion electrodes, flow cells, and membrane-electrode assemblies, and validate their performance at technically relevant current densities, over prolonged operation, and with complete carbon-balance assessments. Addressing these issues will determine whether peripheral environment engineering remains an academic exercise or becomes an industrial reality.

Future progress will depend on closer integration of controlled synthesis, operando characterization, theoretical modelling, and data-driven methods. These approaches should enable peripheral environments to be identified and regulated with greater structural precision under realistic operating conditions. Extending such strategies to technically relevant current densities, long-term operation, and other electrochemical reactions will be important for assessing their broader applicability.

Author contributions

G. F.: writing original draft, writing – review & editing; Y. C.: writing original draft, writing – review & editing; J. Z.: visualization; F. W.: visualization; W. I.: supervision; J. L.: conceptualization, funding acquisition, supervision, project administration, writing – review & editing; X. F.: supervision; S. L.: conceptualization, funding acquisition, supervision, project administration, writing – review & editing.

Conflicts of interest

There are no conflicts to declare.

Data availability

No primary research results, software or code have been included and no new data were generated or analyzed as part of this review.