Atomic-Scale Imaging of Built-In Electric Field Reveals Fe–O_v Synergistic Enhancement for Oxygen Evolution Reaction

Abstract

Imaging difficulties of the built-in electric field within water-splitting catalysts have hampered the acquisition of visual evidence for establishing the direct relationship between electronic structure and catalytic behavior. Herein, modeling on a cobalt–nickel bimetallic oxide comodified by iron doping and oxygen vacancies (O_v), we establish an explicit and comprehensive picture of the enhanced built-in electric field arising from the synergy effect of Fe and O_v at the atomic-level by utilizing differential phase-contrast scanning transmission electron microscopy. Heterogeneous Fe ions are revealed to introduce additional electronic conduction pathways for enhancing the overall electric field intensity. While O_v acted as absorption stations for attracting electrons and leading to a localized accumulation of electron density, Fe and O_v synergistically induce a strong and uniform built-in electric field, thereby enhancing charge migration. In situ spectroscopies further reveal that Fe doping promotes OH* intermediate adsorption by enhancing electric fields, and oxygen vacancies induce localized electric fields to facilitate the lattice oxygen oxidation mechanism. NiCo_2–x Fe _x O_δ fully exploits the Fe–O_v synergistic effect to achieve long-term stability and high catalytic activity in the anodic oxygen evolution reaction (OER). We anticipate this work can elucidate the interplay between cation–anion coregulation and the built-in electric field, offering a rational design strategy for high-performance OER electrocatalysts.

Introduction

Oxygen evolution reaction (OER), serving as the anodic half-reaction in clean energy technologies, involves a complex four-electron transfer process that limits the overall efficiency of water splitting. − Transition-metal oxides (TMOs) are regarded as typical matrix catalytic materials due to their flexible elemental composition and crystal structure. − Recent studies have revealed the crucial role of built-in electric field modulation within TMOs in efficient OER catalysis. The efficient built-in electric field can drive interface charge rearrangement, induce ligand field distortion, facilitate electron migration to the active site, and thereby optimize the adsorption behavior of catalytic reaction intermediates, thereby enhancing the overall catalytic efficiency. − Mingkui Wang and his collaborators constructed a NiFe LDH/FeMoO heterostructure catalyst, using the oxygen vacancies in NiFe LDH and the built-in electric field at the material interface to optimize the electronic state, allowing lattice oxygen to directly participate in the oxygen evolution reaction. Dang et al. constructed NFC catalysts, whose unique asymmetric built-in electric field induces electron redistribution at the interface, optimizing the adsorption behavior of hydrogen and oxygen intermediates. However, precise control over the strength and distribution of the built-in electric field to enhance catalytic activity remains technically challenging.

Regulating the internal chemical environment of catalysts through cation doping or anion defects offers a viable approach to optimizing the electronic structure and built-in electric field properties. For instance, introducing heteroatomic cations can modulate electron distribution and band structure, − while anion defects such as oxygen vacancies (O_v) can form localized states that promote charge trapping and separation. − Both strategies enhance OER activity from distinct perspectives. However, in systems where heteroatoms and O_v coexist, the microscopic mechanism governing their synergistic regulation of the intrinsic structure and catalytic behavior of TMOs remains unclear, particularly lacking direct evidence from the built-in electric field and charge migration image. Therefore, visualizing the built-in electric field is crucial for elucidating the intrinsic mechanism by which cation doping and O_v synergistically enhance the catalytic performance of TMOs.

Differential phase contrast scanning transmission electron microscopy (DPC-STEM) has enabled direct mapping of charge density distributions and electric field polarities within materials by detecting phase shifts as electrons penetrate the sample. − This technique has been successfully applied to analyze charge transfer at heterojunction interfaces, reveal the polarization fields of ferroelectric domain walls, and visualize localized charge rearrangements around single site defects. Lu and collaborators employed DPC-STEM to map the polarization electric fields around antisite defects in monolayer molybdenum disulfide (MoS₂). Zhang et al. applied DPC-STEM to reveal reduced electric field energy at the Ni₃S₂/NM-I electrode interface, indicating no barrier hindering electron transport. Tan et al. used DPC-STEM to demonstrate an asymmetric electric field distribution on Cu surfaces arising from the arrangement of Ag and Sn atoms. These studies illustrate that DPC-STEM transcends conventional electron microscopy analysis, providing a distinctive tool for visualizing and understanding the built-in electric field in catalytic materials.

Herein, utilizing DPC-STEM and in situ spectroscopic techniques, we for the first time directly image the built-in electric field within a series of nickel–cobalt-based metal oxides, comprehensively elucidating the unique synergistic enhancement mechanism of Fe doping and O_v for OER. The introduction of cationic Fe is revealed to enhance the overall built-in electric field for reducing the activation energy of the key intermediate OH*, thereby strengthening the adsorbate evolving mechanism (AEM). Meanwhile, the localized accumulation of electron density induced by O_v modulates the electronic orbital arrangement of the active metal and the free energy of reaction, which further promotes the partial surface reconstruction to activate the lattice oxygen oxidation mechanism (LOM). Under the synergistic regulation of Fe and O_v, NiCo_2–x Fe _x O_δ develops a strong and uniform built-in electric field, compared to other oxides with single-factor regulation, to achieve a more efficient charge migration. The Fe–O_v synergistic effect comprehensively utilizes the high stability of AEM and the high activity of LOM to significantly enhance the OER performance. This work has showcased visual and direct evidence for elaborating the synergistic regulatory effect of cation doping and anion vacancies and linking the built-in electric field with the catalytic behavior of TMO-based catalysts.

Results and Discussion

Fe Substitution at Co Sites and Identification of O_v

NiCo₂O _x (NCO _x ) and Fe-doped NiCo₂O _x (NCFO _x ) were synthesized via a one-pot hydrothermal method, and O_v was introduced into the matrix oxide through controlled calcination under a regulated atmosphere to obtain NiCo₂O_δ (NCO_δ, δ < x) and Fe-doped NiCo₂O_δ (NCFO_δ) (the synthesis procedure is detailed in Figure S1). The powder X-ray diffraction (XRD) characteristic peaks of the as-prepared catalysts match the space group of Fm3̅m (Figure a). The shift of the characteristic peaks reveals that lattice expansion occurred in NCFO _x due to the doping of larger Fe³⁺ ions (0.645 Å), whereas the lattice contraction in NCO_δ could be attributed to O_v-induced strain. It is worth noting that the coexistence of Fe doping and O_v led to a net lattice contraction, which could be further corroborated by the blue shift of the characteristic peaks in Raman spectra (Figure S2). Atomic-resolution high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) images revealed that this series of oxides exhibits a typical antispinel structure with the crystal orientation of [011] (Figure b and Figures S3-S4). The similar ion occupation of Co and Fe could be directly visualized by the atomic STEM-energy dispersive X-ray (STEM-EDX) elemental mapping (Figure b and Figures S5-S8), suggesting that the Fe ions are incorporated into the oxide lattice by substituting Co. X-ray photoelectron spectroscopy (XPS) of the Fe 2p region confirmed the octahedral coordination environment of Fe (III) (709.85 eV-Fe 2p_3/2, 718.25 eV-Fe 2p_1/2), which is consistent with that of the substituted Co site (Figure S9). Density functional theory (DFT) calculations further demonstrated that Fe ions are more inclined to substitute the Co ions at the octahedral sites due to the lower formation energy (−1.69 eV) (Figure c).

1.

Fine structure characterization of catalyst materials. (a) XRD spectra of the four catalysts and zoomed-in comparisons of the (311) characteristic peaks (right panel). (b) Atomic-resolution HAADF-STEM image of NCFO_δ crystals obtained from the [011] direction and elemental mappings of Ni, Co, and Fe. (c) Formation energies of Fe-doped catalysts calculated by DFT and a schematic diagram. (d) Calculated O_v content differences in the catalysts from XPS data. (e) Atomic-resolution iDPC-STEM image of NCFO_δ crystals along the [011] direction. (f) Atomic contrast distribution map derived from (e). The annotated region indicates oxygen atoms with lower contrast. (g) The pixel intensity profile of the line in (e). (h) DFT-calculated formation energy for catalyst O_v formation and a schematic diagram. (i) Schematic diagram of catalytic-level atomic structural changes.

Deconvolution of the O 1s spectrum revealed that the O_v content of NCO_δ increased by 10.2% compared to NCO _x , while that of NCFO_δ increased by 11.2% relative to NCFO _x (Figure d, Figures S10-S11 and Table S1). , The results obtained through electron paramagnetic resonance (EPR) spectroscopy further confirmed the differences in the content of O_v in the catalysts (Figure S12). Corresponding shifts of the oxygen peak position in the electron energy loss spectrum (EELS) similarly validate the increasing trend (Figure S13). Moreover, integrated differential phase contrast STEM (iDPC-STEM) imaging and the corresponding atomic contrast distribution map have revealed localized contrast reduction at oxygen sites in NCFO_δ, which could be utilized to precisely determine the location of the oxygen vacancies (Figure e,f and Figures S14-S15). Concurrently, we observed low contrast at metal sites near O_v, which corresponds to doped Fe atoms (Figure g). Specifically, theoretical calculations and atomic-level EDS mapping further revealed that the O_v are more readily formed near the Fe-doped sites, suggesting a potential synergistic interaction between the Fe ions and O_v (Figure h and Figure S16). These results collectively indicate that the atmosphere control strategy can effectively construct stable high-O_v catalytic materials with heterogeneous element doping (Figure i). Critically, the anchoring interaction between O_v and Fe ions could produce a synergistic effect, which implies an optimized and regulated electronic structure of the active metal within the oxide matrix.

Homogeneous Built-In Electric Field Distribution Enabled by Fe–O_v Synergy

DPC-STEM was employed to map the electric field information within these catalysts (as schematically illustrated in Figure a; see Supporting Information for technical principles). The atomic-level electric field information on NCO _x along the [110] orientation is shown in Figure b, where the arrow direction denotes the direction of the electric field, and color gradients represent the magnitude of the electric field strength. The overall electric field distribution of NCO _x is relatively uniform, but the electric field strength is relatively weak (Figures S17-S18). The electric field within the material serves as the primary driving force for internal charge migration. An efficient charge migration is crucial for optimizing the adsorption strength of intermediates during the catalytic process. To further analyze charge migration driven by varying electric field strengths, we performed distribution of relaxation time (DRT) analysis on the electrochemical impedance spectra (EIS) of the catalytic materials. The results indicate that NCO _x exhibits a high charge transfer resistance during this relaxation time, while the resistances of NCFO _x , NCO_δ, and NCFO_δ gradually decrease (Figures S19-S20).

2.

Analysis of built-in electric fields in catalytic materials. (a) Schematic diagram of the working principle of the DPC-STEM technique. (b) Atomic-level built-in electric field image of NCO _x obtained via DPC-STEM. (c–e) Atomic-level built-in electric field images of catalytic materials obtained via DPC-STEM, corresponding to NCFO _x , NCO_δ, and NCFO_δ, respectively. (f) Comparison of the average electric field strength, the 95th percentile, and the median electric field strength of catalytic materials obtained from DPC-STEM. (g) Schematic diagram of the regulatory role of Fe and O_v in the built-in electric field.

Taking the electric field distribution of NCO _x as the reference, NCFO _x demonstrates an overall enhancement in electric field intensity, as indicated by the blue arrows in Figure c (Figures S21-S22). In contrast, NCO_δ displays a locally intensified electric field distribution, yielding a nonuniform distribution state (circled by red dashed lines in Figure d, Figures S23-S24). Notably, the DPC image of NCFO_δ reveals the strongest and most uniform electric field intensity, without any signs of excessive localization or attenuation (Figure e and Figures S25-S27). These observations indicate that the Fe doping introduced additional electronic conduction pathways, thereby enhancing the overall electric field intensity, while O_v acted as absorption stations for electrons, strongly attracting electrons and leading to a localized accumulation of electron density. More importantly, the coexistence of Fe and O_v in NCFO_δ produces a synergistic effect. The uniformizing effect of Fe extends the local polarization induced by O_v into an overall enhanced built-in electric field, which promotes stronger charge migration within the catalyst. To provide quantitative support for this synergistic electric field enhancement effect, we performed a pixel-by-pixel statistical analysis of the DPC-STEM images and calculated metrics such as electric field intensity. The results show that NCFO_δ exhibits the highest average electric field intensity, 95th percentile electric field intensity, and median electric field intensity, quantitatively confirming a significant enhancement in electric field intensity (Figure f, Figures S28–S30 and Table S2). Analysis of the statistical plot of electric field intensity values and a comparison of the maximum and minimum potential differences reveal that NCFO_δ exhibits both the highest intensity and the widest distribution, further validating the uniformity of its electric field distribution (Figure S31). These statistical data transform the qualitative descriptions of “stronger” and “more uniform” into quantifiable values, further establishing the central role of the Fe–O_v synergy in modulating the intrinsic electric field. Figure g illustrates the schematic representation of Fe and O_v for electric field regulation. The Tauc plots derived from the ultraviolet–visible (UV–vis) diffuse reflectance spectra of the catalytic materials show that NCFO_δ has the narrowest bandgap, reflecting its strongest charge-transfer capability (Figure S32). , To bridge the gap between the static built-in electric field visualized by DPC-STEM and the dynamic electrocatalytic kinetics, we performed in situ EIS, in situ DRT, and in situ Bode analysis. The results show that among the four catalysts, the electrochemical impedance of NCFO_δ is the lowest, the voltage response speed is the fastest, and the resistance of charge bulk transport and interface electron transfer during the reaction process is the smallest (Figures S33-S35). − Furthermore, in situ Bode peak shift reveals that NCFO_δ demonstrates superior initial charge transfer activity and lower voltage dependence (Figures S36-S37). This phenomenon is completely consistent with the expression of the built-in electric field strength in DPC-STEM. A stronger and more uniform static electric field reduces the activation barrier of charge transfer. These results demonstrate that our modification of charge migration through regulating the catalyst’s built-in electric field directly influences the OER process.

Intermediate Adsorption Regulation and Reaction Pathway Identification

The optimization of charge migration dynamics by the built-in electric field ultimately requires precise regulation of intermediate adsorption during the OER. In situ Fourier transform infrared spectroscopy (FT-IR) directly tracked the dynamic adsorption behavior of key reaction intermediates, further investigating the synergistic effects of Fe doping and O_v on the evolution of reaction intermediates. A vibrational band at 1232 cm^–1 corresponding to OOH* adsorbate was detected in all catalysts during the OER, indicating that the adsorbate evolution mechanism (AEM) serves as the primary reaction pathway (Figure a–d). − In NCO_δ, the metal–oxygen (M–O*) characteristic oxygen-containing intermediate representing the LOM mechanism was observed. The relative vibration intensity of adsorbed OOH_ads (M–OOH_ads) and OH_ads (M–OH_ads) revealed that Fe doping in NCFO _x enhanced the adsorption of OH* intermediates, thereby strengthening the AEM mechanism (Figure e and Figure S38). In NCO_δ, the high-O_v strengthened M–O* interaction. Notably, both effects have been concurrently amplified in NCFO_δ (Figure d–f). Furthermore, the discrepancy in reaction mechanisms occurred on the surface of these catalysts was further elucidated by in situ differential electrochemical mass spectrometry (DEMS) with ¹⁸O isotopic labeling (Figure g,h, Figure S39 and Table S3). The ${\mathrm{O}}_2^{34}$ / ${\mathrm{O}}_2^{32}$ ratio remained below 10% for all catalysts, confirming the dominance of the AEM pathway. The ${\mathrm{O}}_2^{34}$ / ${\mathrm{O}}_2^{32}$ ratio decreased to 0.69% for NCFO _x compared to NCO _x , which is consistent with enhanced OH* adsorption observed by in situ FT-IR. The partial contribution from the lattice oxygen mechanism (LOM) could be proved by the pronounced increased ratio value of NCO_δ, which is consistent with the enhanced M–O* interaction observed in in situ FT-IR spectroscopy. Notably, NCFO_δ also exhibits a slightly lower ${\mathrm{O}}_2^{34}$ / ${\mathrm{O}}_2^{32}$ ratio (6.50%) compared to NCO_δ (11.26%), reflecting its appropriate LOM contribution (Figure i). This differentiated regulation of intermediate adsorption energy and reaction pathways also aligns with DPC-STEM results, stemming from distinct charge migration characteristics within the catalytic material. Fe doping induces enhanced electric fields that strengthen charge migration, suppressing excessive activation of lattice oxygen while maintaining AEM dominance by promoting OH* adsorption. Conversely, O_v-induced local electric fields enhance the covalent nature of M–O* bonds to activate the LOM pathway. In NCFO_δ, we calculated the synergy factor of the cooperative effect between Fe and O_v (Table S4). The synergistic regulatory effects of these two interactions help achieve the optimal integration of AEM stability and LOM activity.

3.

Analysis of Fe and O_v regulation on catalytic mechanisms and reaction mechanisms. (a–d) In situ FT-IR spectra obtained at different applied potentials: (a) NCO _x, (b) NCFO _x, (c) NCO_δ, and (d) NCFO_δ. (e,f) Trend diagrams of characteristic FTIR peak intensities for OOH_ads and M-O_ads. (g,h) DEMS signals of ${\mathrm{O}}_2^{32}$ , ${\mathrm{O}}_2^{34}$ , and ${\mathrm{O}}_2^{36}$ of catalysts from the reaction products cycled in the H₂ ¹⁸O aqueous KOH electrolyte: (g) NCO _x and (h) NCFO_δ. (i) Signal ratio and error bar of ${\mathrm{O}}_2^{32}$ to ${\mathrm{O}}_2^{34}$ during the catalyst OER process obtained from DEMS data. (j) Free energy diagram of the catalytic material.

Gibbs free energy changes during the OER were computed to quantitatively assess the reaction energy barriers (Figure j). , In the original NCO _x , the transition from OH* to O* represents the rate-determining step and possesses the highest thermodynamic energy barrier. Fe doping significantly reduces the activation energy for OH* formation, aligning with the enhanced adsorption of OH* observed through in situ FT-IR spectroscopy. In contrast, NCO_δ experiences a fundamental reorganization of the electronic structure, shifting the rate-determining step to the initial water dissociation (H₂O → OH). This shift confirms that O_v stabilizes the O* intermediate by strengthening the M–O* bonding, while simultaneously increasing the energy barrier of the initial adsorption step, which is correlated with the elevated LOM contribution detected by in situ DEMS. Notably, the NCFO_δ catalyst demonstrates an equilibrium distribution of free energy across all OER intermediates. The energy barrier of the rate-determining step (O* → OOH*) decreases significantly, resulting in a theoretical overpotential of 0.53 eV. These computational results have revealed that Fe doping promotes OH* adsorption during the early AEM steps, whereas O_v modulates M–O* bond covalency to facilitate the subsequent steps. The synergistic effect between these two key factors leads to an exceptional catalytic performance of NCFO_δ compared to that of individually modified catalysts.

Elucidating the Electronic Structure Modulation by Fe and O_v

The synergistic optimization of the electronic structure of active metals by Fe doping and high-O_v is the intrinsic mechanism that influences the built-in electric field strength within catalytic materials and enhances catalytic performance. The L₃-edge of Co and Ni in EELS was observed to shift toward the lower-energy direction due to Fe doping and high-O_v content, indicating a reduction in the valence states of Co and Ni (Figure a and Figure S40). Specifically, the dropped Fe ions with low electronegativity act as electron donors to inject electrons into the Co/Ni sites via the Fe–O–Co/Ni pathway, thereby reducing the valence states of Co and Ni. Meanwhile, O_v, a kind of defect carrying an effective positive charge, can transfer its bound electrons toward neighboring metal sites to maintain local charge neutrality, which directly increases the number of occupied electrons at Co/Ni sites. Especially in NCFO_δ, the locally low-coordination environment created by O_v enhances the electron-supplying tendency of Fe sites, while the Fe–O_v configuration provides a more direct pathway for electron transfer. This synergistic interaction between Fe and O_v therefore enables efficient charge modulation, producing a more pronounced L₃-edge shift of NCFO_δ compared to the other oxides. XPS results further proved the trend of decreasing valence states with the Ni²⁺ and Co²⁺ ratio progressively increasing (Figure b, Figures S41-S42 and Tables S5-S6). Moreover, the Bader charge of Ni and Co revealed the highest charge values of NCFO_δ (Figure c), which confirms that the electron-donating ability of dropped Fe ions and the electron-compensation effect of O_v synergistically optimize the electronic structure.

4.

Electronic structure characterization of catalysts. (a) Peak positions of Co and Ni L₃-edge collected from the EELS spectra of catalysts. (b) Changes in Ni³⁺ and Co²⁺ content in the catalyst as calculated by XPS. (c) Bader charges of catalysts. (d,e) PDOS diagrams of the Co element of NCFO _x and NCFO_δ. (f,g) Schematic representations of electronic orbital distribution of NCFO _x and NCFO_δ.

Projected density of states (PDOS) calculations at octahedral sites , reveal that Fe doping and O_v significantly alter the electronic structure of Co. In contrast, the PDOS at the Ni site has a negligible effect (Figure S43), suggesting that the significant reduction in the Ni valence state detected by EELS is attributed to changes in electron transfer behavior caused by Fe doping and O_v, rather than the direct occupation of the Ni d orbitals. Compared with NCO _x , the up-spin component of Co ${\mathrm{d}}_{z^2}$ orbital in NCFO _x transitions from a partially occupied to fully occupied state. This increases the electron density of the up-spin ${\mathrm{d}}_{z^2}$ orbital, enabling the formation of a stronger σ bond with the OER initial intermediate OH*, thereby effectively enhancing the adsorption capacity for OH* (Figure S38). This finding is consistent with infrared spectroscopy results, which demonstrate that Fe doping effectively enhances OH* adsorption during the catalytic process. By comparing the PDOS of NCFO _x and NCFO_δ, it can be observed that the down-spin portion of the Co ${\mathrm{d}}_{z^2}$ orbital shifts downward in energy and becomes fully occupied by electrons after O_v introduction (indicated by the blue arrow in Figure d). Meanwhile, the down-spin portion of the $d_{xz}$ and $d_{yz}$ orbitals shifts upward in energy, raising their levels (indicated by the pink arrow in Figure e). This indicates a shift in the energy levels of the two sets of orbitals, prompting a rearrangement of electrons within them (green arrow in Figure g) and significantly influencing the adsorption behavior of key intermediates in the OER process. Figure f and g clearly contrasts the differences in electronic configurations between the NCFO _x and NCFO_δ systems during OH* and O* adsorption. In the OH* adsorption state, the partial occupation of the σ* antibonding orbital in the NCFO_δ system moderately weakens the initially strong adsorption force, creating a kinetically more favorable starting environment for the conversion step (left panels of Figure f,g). For the O* adsorption state, the introduction of O_v simultaneously weakens the adsorption of both OH* and O*, constrained by the linear scaling relationship of the intermediates. , However, since O* is already bound to the active site via a double bond, its antibonding orbital was partially occupied even before O_v introduction. Consequently, O_v exhibits a more pronounced weakening effect on OH* adsorption (Figure j), thereby promoting the conversion of OH* to O*. This process benefits from the rearrangement of Co’s electronic structure induced by O_v introduction. The fully occupied, energetically lowered ${\mathrm{d}}_{z^2}$ orbital acts as a strong σ donor, injecting electrons into electron-deficient O during dehydrogenation to stabilize the intermediate and lower the energy barrier. Concurrently, the $d_{yz}$ orbitalelevated in energy and better matched with O $p_x$ orbitalforms an efficient π feedback pathway with the fully vacant π* acceptor, further stabilizing O* and preventing its excessive adsorption (right panel of Figure g). Comparisons between NCO _x and NCO_δ similarly validate this perspective (Figures S44-S45).

Based on the above computational analysis, Fe doping and O_v introduction exhibit a distinct synergistic electronic structure regulation mechanism. The core role of Fe doping lies in optimizing the electronic occupation of the spin-up channel at the Co site, transforming the ${\mathrm{d}}_{z^2}$ orbital from a partially vacant state to a fully occupied state. This significantly enhances the material’s adsorption capacity for the initial intermediate OH*, laying the foundation for subsequent reactions. The core function of O_v introduction lies in reconstructing the orbital energy levels and occupation of the spin-down pathway. It shifts the ${\mathrm{d}}_{z^2}$ orbital energy level downward and fully occupies it, while simultaneously raising the d_yz orbital energy level and clearing its associated antibonding π orbital. The essence of this synergistic effect lies in Fe and O_v independently yet complementarily governing the electronic engineering of different spin directions and bonding pathways. Fe enhances the initial electron supply to σ orbitals (spin-up), ensuring effective capture of reactants, while O_v simultaneously optimizes the depth of σ orbital electron supply (spin-down) and the feedback capacity of π orbitals. Together, they drive the efficient conversion of OH* to O* and fine-tune the adsorption strength of each intermediate to overcome scaling limitations. Therefore, Fe and O_v do not act independently but achieve comprehensive multilevel optimization of the electronic structure at the Co active site through precise adjustment of orbital and spin channels.

Fe–O_v Synergistic Regulation of Catalytic Activity and Stability

OER performance was evaluated in 1 M KOH by linear sweep voltammetry (LSV). As shown in Figure a, NCFO _x and NCO_δ exhibited lower overpotential than NCO _x at 10 mA/cm², indicating that both Fe doping and increased O_v content have the effect of enhancing the catalytic performance. Notably, NCFO_δ achieved an overpotential of 276 mV. We have again calculated the synergy factor for the synergistic effect of the interaction between Fe and O_v (Table S7), which verified the synergistic enhancement from Fe–O_v comodification. Tafel analysis further confirmed that NCFO_δ exhibited faster kinetics with a slope of 83.89 mV/dec, compared with NCO _x , NCFO _x, and NCO_δ (Figure b). This remarkable enhancement in catalytic performance benefits from highly efficient charge migration enabled by the potent built-in electric field in NCFO_δ (Figure f), indicating that the Fe–O_v synergistically regulated electric field is the key factor for improving intrinsic activity. Moreover, the distinctive role of high-O_v in exposing active sites was corroborated by the significantly enhanced double-layer capacitance (C _dl) values of NCO_δ and NCFO_δ (Figure c and Figure S46), which correspond to larger electrochemical surface areas (ECSA) for these catalysts. To eliminate the influence of ECSA, we normalized the current density. The results showed that NCFO_δ still exhibited the highest catalytic activity, confirming its superior intrinsic activity (Table S8). We further synthesized control samples with different doping concentrations. The electrochemical test results similarly confirmed that NCFO_δ exhibits optimal activity and charge transfer capability (Figure S47 and the discussion sections).

5.

Analysis of OER performance and reconstruction behavior in catalysts. (a) LSV polarization curve measured at a scan rate of 5 mV/s (the internal illustration depicts the overpotential of the catalysts). (b) Tafel slope plot. (c) C _dl of the catalysts. (d,e) In situ Raman spectra of NCO_δ and NCFO_δ obtained at different applied potentials. (f,g) HRTEM images of catalysts NCO_δ and NCFO_δ after 50 h of OER. (h) The thickness of the reconstructed layers of the catalytic material obtained through HRTEM after different reaction times. (i) Polarization curve of the AEM membrane electrode assembly (MEA) composed of NCO _x and NCFO_δ materials at 80 °C. (j) Stability test of the NCFO_δ-based AEMWE system at a current density of 1.0 A/cm² in 1 M KOH solution.

In situ Raman spectroscopy was employed to probe the structural evolution of the catalysts under operating conditions. Three characteristic peaks at 479 cm^–1 (E_g–Ni–O₆), 521 cm^–1 (F_2g–Co–O₆), and 687 cm^–1 (A_1g–Co–O₄) were identified and assigned to the spinel framework (Figure d,e and Figures S48-S49). , NCO _x exhibited minimal spectral changes with the increasing potential. In contrast, NCO_δ underwent reversible structural reconstruction above 1.4 V, evidenced by the appearance of new bands at 452 cm^–1 and 565 cm^–1, which were assigned to NiOOH and (Fe,Co)­OOH species, respectively (Figure d). A similar reconstruction behavior of NCFO_δ under a 1.6 V applied bias voltage confirmed the promoting effect of O_v on the phase transformation (Figure e). Notably, NCFO _x maintained the spectral features across the measured potential range, indicating that Fe doping does not have the function of triggering the reconstruction. Similarly, HRTEM images show that for NCO_δ and NCFO_δ, the thickness of the surface reconstruction layer increases with the extension of the reaction time (Figure f,g and Figures S50-S53). However, the reconstruction layer of NCFO_δ is always thinner than that of NCO_δ at all time points (Figure h). This observation directly confirms the in situ Raman evidence. These results are consistent with the Fe doping-promoted AEM mechanism and O_v-promoted M–O* interaction observed in theoretical calculations, FT-IR, and DEMS. The enhanced electric field effect from Fe doping accelerates charge migration, effectively boosting the stability of the matrix oxide, while high-O_v induces surface restructuring under the localized enhanced electric field. In the NCFO_δ system, the synergistic interaction between Fe doping and high-O_v yields a stronger, more uniform electric field and enhanced charge migration. This combination effect simultaneously maintains structural stability and boosts catalytic activitythe sum of these combined effects exceeds the sum of their individual contributions.

To identify the phase composition of the restructured layer and its impact on the electronic structure of the catalytic material, we systematically characterized the postreaction samples. Raman and EDS analyses confirmed that the restructured layer consists of hydroxyl oxides generated by OER (Figures S54-S55). XPS analysis showed that the O_v content in the NCFO_δ bulk phase remained essentially stable after the reaction, while the O_v content in the NCO_δ phase continued to decrease (Figures S56-S58 and Table S9). Combined with EELS data, this further indicated that the valence states of Ni and Co in the NCFO_δ phase underwent the least change (Figures S59-S62). Additionally, postreaction bulk DPC-STEM imaging revealed that the electric field distribution was consistent with that before the reaction (Figure S63), while DEMS confirmed that a moderate LOM contribution was retained after restructuring (Figure S64). Further DFT calculations were performed to model the hydroxyl oxides formed in the restructured phase. The results indicated that Fe could still lower the OH* adsorption energy barrier, and PDOS analysis also revealed an energy level shift in the Co d orbitals that favored OH* adsorption (Figures S65-S66). These findings demonstrate that the Fe–O_v synergistic effect remains stable after restructuring, achieving a balance between structural stability and catalytic activity.

In anion exchange membrane water electrolysis (AEMWE) operating in 1 M KOH at 80 °C, the NCFO_δ-based membrane electrode assembly (MEA) achieved current densities of 0.957 A/cm² and 1.90 A/cm² at battery voltages of 1.70 and 1.87 V, respectively. In contrast, NCO _x -based MEA could only attain current densities of 0.572 A/cm² and 1.167 A/cm² at these voltages (Figure i and Figures S67-S68). More importantly, chronopotentiometry testing of the AEM was carried out at a current density of 1 A/cm² (Figure j). During the 1000 h chronopotentiometry test, the NCFO_δ electrode demonstrated exceptional stability, with no performance degradation observed. Inductively coupled plasma optical emission spectroscopy (ICP-OES) analysis revealed that during the OER process, both Ni and Co in NCO _x continued to dissolve, whereas the dissolution rate of metal ions in NCFO_δ increased slowly over time, with the dissolution concentration gradually stabilizing, further confirming the structural stability (Figure S69 and Table S10). This further validates the dual enhancement of activity and stability achieved through Fe–O_v synergistic effects, fully highlighting its application potential in industrial water electrolysis.

Conclusion

In summary, we directly imaged the atomic-level electric field in catalytic materials and elucidated the synergistic interaction between Fe doping and high-O_v. Fe doping enhanced the charge transport by altering electronic orbital occupation, and high-O_v modulated the orbital structure due to the charge concentration effect, which collectively imparted a significantly strengthened built-in electric field to NCFO_δ. The enhanced built-in electric field significantly improved the adsorption of the key intermediate OH_ads and suppressed material restructuring behavior. Thus, NCFO_δ can balance the long-term stability characteristic of the AEM with the high activity characteristic of the LOM, thereby improving catalytic performance and accelerating the OER. Our work has revealed the influence of cation–anion regulation on the atomic-level electric field within catalytic materials and clarified the impact of charge migration on catalytic behavior. These findings establish an advanced electron microscopy technology foundation for investigating the catalytic mechanisms governed by electronic-structure modulation.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacs.6c10106.

• Detailed experimental details, experimental methods (sample synthesis, characterization, testing, and DFT calculations), additional STEM, Raman, XPS, electrochemical, DEMS, and DFT calculation data (PDF)

#.

B.S. and Z.Z. contributed equally to this work.

The authors declare no competing financial interest.