Calcium‐Mediated Fe─N Bond Reinforcement for Ultra‐Stable Oxygen Reduction Reaction

ABSTRACT

The practical deployment of atomically dispersed Fe–N–C catalysts for the oxygen reduction reaction (ORR) is severely hampered by the electrochemical leaching of Fe active sites. Inspired by the stabilizing role of Ca²⁺ in metalloenzyme active sites, a novel Fe–Ca dual‐atom sites catalyst (Ca/Fe–N–C) is constructed on amorphous porous carbon nanosheets. The introduced Ca atom acts as an “electronic modulator” and “structural stabilizer,” which effectively lowers the oxidation state of Fe and reinforces Fe–N coordination bond. This ingenious design results in an exceptional ORR catalyst with the half‐wave potential of 0.912 V in alkaline media and unprecedented durability, exhibiting negligible decay after 80000 cycles. When integrated into Zn‐air batteries (ZABs), the Ca/Fe–N–C‐based cathode delivers a peak power density of 215 mW cm⁻² and sustains operation for exceeding 1110 h, markedly superior to benchmark Pt/C. This work not only unveils the pivotal role of alkaline‐earth metals in stabilizing transition‐metal sites but also establishes a general paradigm for designing durable atomically dispersed catalysts for advanced energy conversion devices.

A novel Fe–Ca dual‐site catalyst anchored on amorphous porous carbon nanosheets delivers exceptional oxygen reduction activity and long‐term durability. Ca atom modulates the electronic structure of Fe centers, reinforces Fe–N coordination, and accelerates O₂ adsorption and activation. This work extends dual‐atom catalysts to alkaline earth metals and offers a paradigm beyond the activity‐stability trade‐off.

1. Introduction

Fuel Cells (FCs) and Zn‐Air batteries (ZABs) rely on ORR as a key step, yet its sluggish kinetics remain a major barrier to improving device efficiency [1, 2, 3]. Although Pt‐based materials exhibit outstanding ORR activity, the broad application is limited by cost and durability, driving the research of non‐precious metal alternatives [4, 5, 6]. Fe–N–C has drawn great attention, largely due to its remarkable intrinsic ORR performance. However, their operational stability, particularly in alkaline media, is critically challenged by the vulnerability of the Fe‐N₄ bond to OH⁻ attack. This can lead Fe–N bond cleavage, metal leaching, and accelerated carbon corrosion, significantly compromising catalytic performance [7, 8, 9]. Thus, improving durability of Fe–N–C is imperative for advancing FCs and ZABs applications.

To tackle the limited stability of Fe–N–C, developing dual‐atom active sites has gained increasing attention [10, 11, 12]. Introducing a secondary metal atom adjacent to the Fe‐N₄ center enables electronic reconfiguration and generates a confinement effect, thereby suppressing Fe–N bond cleavage [13, 14, 15]. At present, dual‐atom catalysts have mainly focused on transition metal combinations (Fe–Co, Fe–Ni, and Fe–Mn). However, the large differences in reduction potential among d‐block transition metals limit the application potential of transition metal dual atom pairs. Furthermore, the participation of 3d orbitals may induce competitive electron interactions with Fe active sites, thereby affecting electron transfer processes [16, 17, 18]. Notably, the potential of incorporating alkaline earth metals (Mg and Ca) in combination with transition metals has been largely overlooked and remains an underexplored area. Despite the reported existence of Mg–N–C and Ca–N–C, prior studies have focused on constructing isolated metal sites and have not been employed to optimize the durability of Fe–N–C [19, 20, 21]. Notably, Ca²⁺ is known to stabilize active sites in metalloenzymes, thereby maintaining structural integrity, offering a conceptual inspiration for catalyst design [22, 23]. Furthermore, the large ionic radius and low electronegativity of Ca can induce local lattice coordination distortion and modulate the spin state of the Fe center, which is expected to enhance catalytic activity [24].

Herein, we propose a charge‐regulation approach to overcome the stability challenges associated with Fe–N–C. We hypothesize that introducing neighboring Ca atoms could mimic this biological function, thereby electronically and structurally reinforcing the vulnerable Fe‐N₄ sites. Through a solid‐state mechanochemistry‐assisted confined approach, we successfully fabricate Fe–Ca dual‐atom sites embedded within amorphous porous carbon nanosheets (denoted as Ca/Fe‐N‐C). Comprehensive experimental characterization combined with theoretical calculations reveals that the Ca atom serves as an electron donor, induce charge accumulation at the Fe‐N₄ site, and substantially strengthening the Fe─N bond. This synergistic effect not only lowers the energy barrier for O₂ activation but also dramatically elevates the dissolution energy of Fe, efficiently suppressing demetalation. Consequently, the Ca/Fe‐N‐C delivers remarkable ORR performance and exceptional durability of over 1110 h, which is further translated into superior performance in practical ZABs. This work transcends the conventional combination of transition metals and highlights the untapped potential of s‐block elements in constructing ultra‐stable atomic interfaces for sustainable energy conversion.

2. Results and Discussion

Recent studies have shown that Ca²⁺ can regulate and stabilize the active sites of metalloenzymes, effectively reducing the risk of inactivation [25]. Furthermore, Ca possesses intrinsic physicochemical properties, including the absence of partially filled d orbitals, low electronegativity, and large ionic radius, enabling modulation of Fe–N active sites via coordination distortion and charge polarization [26]. Inspired by these features, Ca site was introduced into Fe–N–C to enhance the activity and structural integrity. As illustrated in Figure 1a, Ca/Fe‐N‐C was synthesized via a solid‐state mechanochemical‐assisted confined pyrolysis method. Ball milling of ZnO, NaCl, metal acetylacetonates, and 2‐methylimidazole generated a NaCl‐confined precursor. Subsequently, N₂ atmosphere pyrolysis induced carbonization and Ca/Fe‐N_x site formation. After removal of NaCl and subsequent acid etching, a hierarchically porous N‐doped carbon nanosheet featuring Fe/Ca dual‐atom sites was successfully obtained. X‐ray diffraction (XRD) patterns of as‐obtained samples (Figure S1) display feature peaks near 25° and 44°, consistent with the typical diffraction of amorphous carbon, confirming that no metallic Fe and Ca phases [27, 28]. The high I_D/I_G value of Ca/Fe‐N‐C confirms its abundant defect structures (Figure S2). Furthermore, the Ca/Fe‐N‐C possesses a large specific surface area (1284.9 m² g⁻¹), with the pore structure predominantly composed of micro‐, meso‐, and macropores (Figure S3 and Table S1). This promotes active site accessibility and mass transfer, which is of great significance for promoting the ORR process [29, 30]. Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) analysis revealed that Ca/NC, Fe–N–C, and Ca/Fe‐N‐C all display sheet‐like morphology and highly porous structures with abundant edge sites, which could promote sufficient diffusion of the electrolyte ions (Figure S4, Figure 1b; Figures S5 and S6). Notably, high‐resolution TEM (HRTEM) image illustrates its amorphous structures of Ca/Fe‐N‐C, and selected area electron diffraction (SAED) corroborates this observation (Figure 1c). The energy‐dispersive X‐ray spectroscopy (EDX) mappingof Ca/Fe‐N‐C (Figure 1d) reveals a homogeneous distribution of C, N, Ca, and Fe across the amorphous carbon support. Aberration‐corrected high‐angle annular dark‐field scanning TEM (AC HAADF‐STEM) result suggests that Fe and Ca in Ca/Fe‐N‐C are atomically dispersed, and the paired bright spots (yellow circles) confirm the successful construction of Fe–Ca sites (Figure 1e). The 2D and 3D resolution maps within the red‐marked region provide clear confirmation of Fe–Ca dual sites formation (Figure 1f,g). The measured interatomic distance between Fe and Ca atoms is approximately 2.92 Å, indicating the presence of electronic interactions of Fe and Ca. Although a small fraction of isolated Fe or Ca single‐atom sites (green circles) can still be observed in Figure 1e, Fe–Ca dual sites constitute 83.2% of the total sites, confirming their dominant presence in Ca/Fe‐N‐C (Figure 1h). Inductively coupled plasma optical emission spectroscopy (ICP‐OES) analysis determined the Fe and Ca contents in Ca/Fe‐N‐C to be 0.72 and 0.59 wt.%, respectively (Table S2).

FIGURE 1.

(a) the fabrication process of Ca/Fe‐N‐C, (b) TEM image of Ca/Fe‐N‐C, (c) HRTEM image of Ca/Fe‐N‐C, (d) EDS mapping image of Ca/Fe‐N‐C, (e) AC HAADF‐STEM image of Ca/Fe‐N‐C, (f) 2D atomic resolution map of Fe–Ca sites from e, (g) 3D model of Fe–Ca sites from g, (h) density profiles of Fe–Ca sites.

The X‐ray photoelectron spectroscopy (XPS) spectra of as‐obtained samples display prominent characteristic peaks attributable to C, N, and O elements (Figure S7). Low contents of Fe and Ca result in correspondingly weak signals in the measurements. Table S3 lists the element contents of different samples determined from XPS full spectrum. The C 1s spectra reveal that Ca/Fe‐N‐C possesses a greater proportion of sp³‐hybridized carbon, yielding a reduced sp²/sp³ ratio of 1.99 than Fe–N–C and Ca–N–C. This suggests a more defective amorphous carbon structure in Ca/Fe‐N‐C, conducive to stabilizing the active sites (Figure S8a–c) [31, 32]. The N 1s spectra of Ca–N–C, Fe–N–C, and Ca/Fe‐N‐C can be resolved into five distinct species, corresponding to pyridinic N (398.3 eV), metal‐N_x (399.4 eV), pyrrolic N (400.2 eV) graphitic N (401 eV), and oxidized N (403.9 eV), respectively (Figure S8d,e) [33, 34]. Importantly, the presence of M‐N_x confirms the coordination of Fe/Ca with N atoms. The Fe 2p and Ca 2p spectra of Ca/Fe‐N‐C, Fe–N–C, and Ca–N–C shown no obvious characteristic peaks, likely resulting from the minimal Fe and Ca loading (Figure S9). In addition, X‐ray absorption spectroscopy (XAS) was further employed to elucidate the coordination environment of Fe–Ca diatomic site. The Fe K‐edge X‐ray absorption near‐edge structure (XANES) spectra of both Fe‐N‐C and Ca/Fe‐N‐C are located between those of FeO and Fe₂O₃, implying an average oxidation state of Fe close to +3 (Figure 2a) [35, 36]. Notably, the Fe‐K edge front in Ca/Fe‐N‐C shifts to lower energy than Fe–N–C, demonstrating a lower oxidation at Fe sites in the former (Figure 2b) [37]. The coordination structure of the central metal atom was further analyzed using Fe K‐edge Fourier transform extended X‐ray absorption fine structure spectroscopy (FT‐EXAFS). In Figure 2c, the single scattering path corresponding to the Fe–N paths is observed at approximately 1.5 Å in Fe‐N‐C and Ca/Fe‐N‐C, while no strong peaks associated with Fe–Fe coordination [38]. Meanwhile, the Fe‐N coordination distance in Ca/Fe‐N‐C is shorter than in Fe–N–C, indicating bond reinforcement contributes to enhanced stability [39]. The Ca K‐edge XANES curves for Ca/Fe‐N‐C show absorption edge positions close to those of the Ca‐containing reference samples, indicating that Ca presenting an oxidation state (Figure 2d) [40]. Due to the influence of CaO edge‐front peak, the average oxidation state of Ca cannot be directly determined from the absorption edge. The linear fit of between absorption intensity and average oxidation state, Ca in Ca–N–C displays an average oxidation state of +1.5, which is higher than observed in Ca–N–C (+1.35) (Figure 2e). The increased oxidation state of Ca indicates charge redistribution between Ca and Fe sites, where electron transfer from Ca to Fe decreases the oxidation state of Fe while simultaneously elevating that of Ca. Meanwhile, the Ca K‐edge EXAFS spectrum exhibits a feature at ∼1.64 Å, reflecting the presence of Ca–N scattering path (Figure 2f). Notably, no signals of Ca‐Ca or Ca‐O‐Ca bonding are observed, confirming the atomic dispersion of Ca species [41, 42]. The EXAFS fitting curves in k‐ and R‐space reveal that Fe and Ca in all three samples are four N atoms coordination environment (Figures S10–S12 and Table S3). The wavelet transform (WT) EXAFS analysis further confirms that Ca/Fe‐N‐C displays a dominant intensity maximum around 4.0 Å⁻¹, characteristic of Fe/Ca‐N coordination, without evidence of other significant scattering contributions (Figure 2f). All the results demonstrate that Ca and Fe atoms are interconnected through N‐bridging bonds. The Ca‐N₄ sites effectively modulates Fe‐N₄ electronic states, strengthens the Fe–N bonds, and consequently boosts ORR performance.

FIGURE 2.

(a) Fe K‐edge XANES spectra of Ca/Fe‐N‐C and comparison samples, (b) valence state fitting of Ca/Fe‐N‐C and comparison samples, (c) Fourier transform k3‐weighted Fe K‐edge EXAFS spectra of Ca/Fe‐N–C and comparison samples, (d) Ca K‐edge XANES spectra of Ca/Fe‐N‐C and comparison samples, (e) valence state fitting of Ca of Ca/Fe‐N‐C and comparison samples, (f) Fourier transform k3‐weighted Ca K‐edge EXAFS spectra of Ca/Fe‐N‐C and comparison samples, (g) WT of Ca/Fe‐N‐C and comparison samples.

The ORR activities of prepared catalysts were investigated in 0.1 m KOH solution. Prior to the electrochemical measurements, the reference electrode was carefully calibrated (Figure S13). Figure S14 presents the cyclic voltammetry (CV) curves of the catalysts in both N₂‐ and O₂‑saturated electrolytes, where all samples exhibit clear cathodic peaks under the O₂‑saturated condition. Notably, the cathodic peak potential of Ca/Fe‐N‐C is higher than the other catalysts, suggesting its remarkable ORR performance, as further confirmed through linear sweep voltammetry (LSV) analysis. In Figure 3a,b, the half‐wave potential (E_1/2) of Ca/Fe‐N‐C is 0.912 V, significantly outperforming NC (0.830 V), Ca/NC (0.850 V), Fe–N–C (0.875 V), and Pt/C (0.856 V). The ORR kinetic current densities (J_k) of the above catalysts at 0.85 V were further evaluated by LSV curves. Compared to other catalysts, the J_k of Ca/Fe‐N‐C could reach 51.4 mA cm⁻², which suggests the fast ORR kinetic process (Figure 3b). These results indicate that Ca species incorporation enhances catalytic performance. To verify the interaction between Ca and Fe species, we supplemented the electrochemical performance of a physically mixed Ca–N–C and Fe–N–C. The mixed catalyst exhibits a significantly lower E_1/2 (0.876 V) than Ca/Fe‐N‐C, indicating that the electron donor–acceptor interaction between Ca and Fe species is crucial in boosting ORR performance (Figure S15). In addition, the Tafel slope of Ca/Fe‐N‐C (52.2 mV dec⁻¹) is lower than reference catalysts, indicating its superior kinetic behavior (Figure S16). In Figure S17, electrochemical double‐layer capacitance (C_dl) test results reveal that the largest C_dl value for Ca/Fe‐N‐C is 18.82 mF cm⁻², indicating its maximum electrochemical surface area (ECSA) [43]. Moreover, electrochemical impedance spectroscopy (EIS) confirms that Ca/Fe‐N‐C has the smallest charge transfer resistance compare to the reference catalysts, further demonstrating higher ORR kinetics and excellent catalytic performance (Figure S18). Apart from ORR activity, selectivity also serves as a key parameter to evaluate ORR catalysts. The catalytic selectivity of the prepared catalysts and commercialized Pt/C was examined by rotating ring disk electrode (RRDE). Ca/Fe‐N‐C catalyst exhibits a lower H₂O₂ yield of less than 5%, indicating its excellent capability for O─O bond breaking (Figure 3c). The electron transfer numbers of Ca/Fe‐N‐C catalyst determined from RRDE tests and Koutecky–Levich (K–L) plots ranged from 3.92 to 3.98, further confirming their efficient four‐electron ORR pathway (Figure S19).

FIGURE 3.

(a) LSV curves of NC, Ca–N–C, Ca/Fe‐N‐C and Pt/C, (b) J_k and E_1/2 of various catalyst at 0.85 V, (c) electron transfer numbers (n) and H₂O₂ yield of Ca/Fe‐N‐C and commercial Pt/C, (d) LSV curves of Ca/Fe‐N‐C and Fe–N–C before and after accelerated durability tests, (e) chronoamperometric tests of different catalysts at 0.65 V vs RHE, (f) dissolution energy of Fe sites on Fe–N–C and Ca/Fe‐N‐C, (g) comparison of stability and E_1/2 of Ca/Fe‐N‐C and other catalysts.

Stability is crucial for evaluating ORR catalysts. After 80 000 cycles, Ca/Fe‐N‐C exhibited a minor E_1/2 of 13 mV, whereas the Fe–N–C showed a 21 mV decline in E_1/2 after just 30 000 cycles (Figure 3d; Figure S20a,b). Chronoamperometry measurement further confirms this trend, with Ca/Fe‐N‐C maintained more than 99.2% current retention after 120 000 s of continuous operation (Figure 3e). To clarify the origin of enhanced stability, we evaluated the solubility energies of the corresponding catalysts. Compared to Fe–N–C (0.56), Ca/Fe‐N‐C exhibited a higher dissociation energy for Fe sites (0.94), indicating substantially improved resistance against dissolution (Figure 3f). This improvement is attributed to Ca incorporation, which modulates the local electronic structure and coordination environment of Fe centers, thereby suppressing Fe leaching. Furthermore, the Ca/Fe‐N‐C demonstrated excellent methanol tolerance. In Figure S21, the activity of Ca/Fe‐N‐C remained nearly unchanged after the introduction of methanol compared to Pt/C. In addition, the stability characterization after 80 000 cycles via SEM, TEM, and HRTEM confirmed that Ca/Fe‐N‐C retains its amorphous sheet‐like porous structure without observable structural degradation (Figure S22a–c). No Fe or Ca‐related crystalline compounds were detected in AC‐HAADF, EDS mapping, and XRD patterns, indicating that Fe and Ca remain in an atomically dispersed state after long‐term cycling (Figures S22d,e and S23). Furthermore, ICP‐MS analysis of the electrolyte after 20 000 cycles confirmed that the introduction of Ca atom effectively inhibits Fe dissolution (Table S4). Collectively, the above results highlight the beneficial effect of Ca incorporation in enhancing the longevity of the Fe–N–C. Through combining exceptional activity with remarkable stability, the designed Ca/Fe‐N‐C catalyst surpasses the majority of current single and dual‐atom catalysts (Figure 3g; Table S5).

In situ FTIR spectroscopy was used to monitor the formation and transformation of intermediate species during the ORR process, providing real‐time insight into the reaction mechanism (Figure 4a,b). Both catalysts display broad absorption peaks at ∼3350 cm⁻¹, assigned to adsorbed ^*OH intermediate, indicating the involvement of hydroxyl‐related intermediates during ORR [44, 45]. The features at ∼1435 and 1297 cm⁻¹ are characteristic of ^*O₂ ⁻ and ^*OOH related species. Notably, the Ca/Fe‐N‐C exhibits significantly more intense and distinctly potential‐dependent signals for the ^*O₂ ⁻ and ^*OOH intermediates compared with Fe–N–C, suggesting that Ca incorporation effectively promotes O₂ activation and subsequent ^*OOH formation. Additionally, we also conducted in situ Raman spectroscopy to investigate the oxygen intermediates in reaction process (Figure 4c,d). A characteristic band at ∼1142 cm⁻¹, assigned to the superoxide (^*O₂ ⁻) intermediate, is observed for both catalysts. However, the Ca/Fe‐N‐C catalyst exhibits the markedly stronger ^*O₂ ⁻ signal with a clearer potential dependence, suggesting a more efficient initial activation of O₂ [46]. In comparison with Fe–N–C, the band at ∼1578 cm⁻¹ corresponding to the ^*OOH intermediate is significantly enhanced on Ca/Fe‐N‐C, further confirming the facilitated formation and stabilization of ^*OOH species upon Ca incorporation.

FIGURE 4.

In situ ATR‐FTIR spectra of (a) Fe–N–C and (b) Ca/Fe‐N‐C, in situ Raman spectra of (c) Fe–N–C and (d) Ca/Fe‐N‐C (e) free energy diagrams, (f) charge density difference, (g) DOS of Fe sites, (h,i) COHP analysis and the corresponding ICOHP values of the Fe─N bond, (j) demetalation process of Fe sites on Fe–N–C and Ca/Fe‐N‐C.

DFT calculations were performed to further elucidate reaction mechanism and the role of Ca in enhancing both activity and stability. Optimization models of Fe‐N₄ (Fe–N–C) and Ca/Fe‐N₄ (Ca/Fe‐N‐C) were constructed for comparative analysis, and the adsorption behaviors of oxygen‐containing intermediates on the above models were investigated (Figure S24). Compared to Fe–N–C (−0.74 eV), the Ca/Fe‐N‐C model displays a higher O₂ adsorption energy (−1.2 eV), implying that Fe–Ca dual site can more stably adsorb and activate O₂ (Figure S25). To further experimentally verify this theoretical prediction, O₂ adsorption/desorption isotherms were measured. Compare to Fe–N–C, Ca/Fe‐N‐C possesses an enhanced O₂ accommodation capacity, demonstrating enhanced interaction between the catalyst surface and O₂ (Figure S26). At U = 0 V, all elementary steps on the two models proceed via spontaneous exothermic reactions (Figure 4e; Figures S27 and S28). At U = 1.23 V, the potential limiting steps (PLS) for Fe–N–C is the ^*O → ^*OH step, with a relatively high barrier of 0.74 eV. In contrast, after introducing Ca atom, the PLS shifts to the O₂ → ^*OOH step, and the reaction energy barrier is significantly reduced (0.44 eV). These results indicate that the Ca atom not only weakens the adsorption strength of intermediates but also alters the RDS, thereby accelerating the overall ORR kinetics. To better verify the theoretical results under more realistic conditions, ab initio molecular dynamics (AIMD) simulations and constant‐potential free energy calculations were additionally carried out (Figure S29). The AIMD results show that the Grand Canonical Energy (GCE) and temperature rapidly converge and fluctuate within a narrow range, while the Fe–Ca interatomic distance remains nearly constant without bond dissociation or structural reconstruction, confirming the dynamic stability of the dual‐atom configuration (Figures S30 and S31). More importantly, the free energy profiles obtained under AIMD and constant‐potential conditions exhibit a consistently downhill trend, with the potential‐determining step showing a minimal free energy change (∼0.03 V), corresponding to an onset potential of ∼0.82 V and an overpotential of ∼0.41 V, slightly lower than those derived from the conventional static DFT calculations (Figure S32). These results indicate that the solvation effects and applied potential do not alter the overall reaction trend.

To elucidate the origin of the reduced reaction energy barrier and the shift of RDS, the interaction between Ca and Fe atoms was examined through differential charge analysis. As shown in Figure 4f, Ca incorporation induces the redistribution of charge density around Fe‐N₄. Specifically, the Fe‐N₄ part in Ca/Fe‐N‐C exhibits significant charge accumulation compared to Fe–N–C, while electron depletion occurs near the Ca atom, confirming the electron‐donating nature of Ca atom. Bader charge analysis shows that the charge of Fe atom decreases from the original 1.24 to 1.09 after the introduction of Ca atoms, further confirming the electron supply from Ca atom to Fe atom, consistent with the XAS result (Table S6). Projected density of states (PDOS) calculations indicate that Ca incorporation shifts the d‐band center of Fe downward to −1.28 eV, which weakens over‐adsorption of oxygen‐containing intermediates typically observed on conventional Fe‐N₄ sites, while maintaining sufficient adsorption strength for efficient O₂ activation. To clarify the specific contributions of different d orbitals, the electron occupancy of individual orbitals was further analyzed. Compared to Fe–N–C, the introduction of Ca atom enhances the electron filling degree of the d_z ² orbital of Fe, which directly enhances the DOS of this orbital at the Fermi level, making it easier for it to exchange electrons with reactants (Figure S33). Thus, the d_z ² orbital controls the adsorption of O₂ and ^*OOH species, the increased filling degree can effectively balance the adsorption‐activation relationship of O₂ [47, 48].

To further understand the enhanced stability, the influence of Ca introduction on the Fe–N bond strength was analyzed via Crystal Orbital Hamilton Population (COHP). ICOHP analysis indicates that Ca/Fe‐N‐C exhibits a higher ICOHP value of 2.76, meaning that Ca incorporation enriches the electron density near the Fe‐N₄ site, thus reduces the oxidation state of Fe and strengthens the Fe–N bond (Figure 4h,i). Furthermore, the energy barrier for Fe–N bond dissociation was evaluated (Figure 4j). The formation energy between Fe and O₂ was computed, revealing a more negative value for Ca/Fe‐N‐C (−6.29 eV) than for Fe–N–C (−5.72 eV). Moreover, the Fe–N bond in Ca/Fe‐N‐C requires 1.18 eV to dissociate, substantially exceeding the 0.72 eV required for the Fe–N–C. In general, ICOHP and bond dissociation energy results clearly demonstrate that the introduction of Ca atom substantially enhances the structural stability of Fe–N–C.

Given its outstanding ORR activity, Ca/Fe‐N‐C was utilized as the air cathode in assembled aqueous ZAB to evaluate its practical application (Figure 5a). As shown in Figure 5b, the Ca/Fe‐N‐C‐based ZAB delivers an open circuit potential is 1.56 V, high than the 1.53 V for Pt/C. Discharge polarization curves (Figure 5c) reveal that the Ca/Fe‐N‐C‐based ZAB achieves a peak power density of 215 mW cm⁻², outperforming the Pt/C (127 mW cm⁻²) and recently reported single/dual atoms‐based ZABs. Moreover, the Ca/Fe‐N‐C‐based ZAB reaches a specific capacity of 783 mAh g⁻¹, surpassing both Pt/C and recently reported single‐atom‐based ZABs (Figure 5d). Notably, the specific capacity of ZAB assembled by Ca/Fe‐N‐C is 95.5% of the theoretical specific capacity, indicating that the ZAB achieves almost complete utilization of Zn. Benefiting from its outstanding discharge performance and operational stability, the Ca/Fe‐N‐C‐based ZAB can easily drive a light‐emitting diode panel, thereby providing compelling evidence for the practical potential of the Ca/Fe‐N‐C in real‐world energy storage devices (Figure 5e). Furthermore, continuous discharge tests across a range of different current densities were conducted to assess the rate performance of ZABs (Figure 5f). The device maintains a stable discharge voltage over the entire current density range, and it quickly returns to its initial value upon reverting from 50 to 1 mA cm⁻², highlighting the outstanding rate performance of the Ca/Fe‐N‐C‐based ZAB. More importantly, the cycling stability of Ca/Fe‐N‐C‐based ZAB was evaluated via a continuous charge–discharge test (Figure 5g). As expected, the Ca/Fe‐N‐C‐based ZAB exhibits narrower charge‐discharge voltage gaps and negligible voltage drop after 1110 h of continuous operation. On the contrary, the Pt/C‐based ZAB showed significant performance degradation after only 287 h. The comprehensive comparison reveals that the power density and durability of Ca/Fe‐N‐C‐based ZAB outperform most currently reported single/dual‐atom catalysts (Figure 5h; Table S7).

FIGURE 5.

(a) Schematic of the homemade ZABs, (b) open circuit potential plots, (c) discharge polarization curves and power density, (d) discharge plots at 5 mA cm⁻², (e) photograph of Ca/Fe‐N‐C based ZAB‐driven light‐emitting diode panel, (f) rate performance, (g) galvanostatic discharge‐charge cycling curves at 10 mA cm⁻², (h) power density and stability comparison of the Ca/Fe‐N‐C and other ZABs reported.

3. Conclusions

In summary, the Fe–Ca dual‐atom catalyst has been successfully engineered, demonstrating a breakthrough in concurrently achieving ultrahigh activity and exceptional durability for the ORR. The introduction of Ca is driven by its unique physicochemical properties, which can fundamentally regulate the electronic configuration of the Fe‐N₄ centers. It serves as an efficient electron donor, populating the critical d_z ² orbital of Fe to optimize O₂ adsorption, while simultaneously reinforcing the Fe–N covalent bond through charge redistribution, thereby rendering the active site highly resistant to electrochemical dissolution during the ORR process. The optimized Ca/Fe‐N‐C displayed an E_1/2 of 0.912 V and maintains exceptional stability throughout 80 000 cycles. Its practical utility is further confirmed in ZABs, which achieve a peak power density of 215 mW cm⁻² and a remarkable operating lifespan of 1110 h. Beyond presenting a superior electrocatalyst, this work introduces a novel paradigm of utilizing s‐block elements as electronic and structural modifiers for single‐atom catalysts. It offers a new opportunity to engineer next‐generation durable electrocatalysts for sustainable energy conversion and storage, moving them closer to commercial reality.

4. Experimental Section

Synthesis of Ca/Fe‐N‐C and Reference Electrocatalysts: In the typical synthesis, 4 g NaCl was first ball‐milled at 350 rpm for 30 min. Subsequently, 0.821 g 2‐methyl imidazole, 0.406 g ZnO, 0.12 g Ca(acac)₂, 0.04 g Fe(acac)_3, along with 1 mL ethanol were added, and ball‐milling continued for 1 h at the same speed. Under N₂, the obtained precursor mixture was ramped to 350°C and kept at this temperature for 1 h and then further raised to 950°C and held for 3 h. To remove the NaCl template and compounds associated with Fe and Ca, the pyrolyzed sample was treated with H₂O and 3 m HCl, respectively. After thorough washing and drying, the treated material was heated to 950°C for 1 h at 3°C/min, yielding the catalyst was denoted Ca/Fe‐N‐C. For Ca–N–C and Fe–N–C, the synthesis followed the same preparation method as for Ca/Fe‐N‐C, incorporating only the corresponding single metal precursors during preparation.

Conflicts of Interest

The authors declare no competing interests.

Supporting information

Supporting File: adma73684‐sup‐0001‐SuppMat.docx.

Data Availability Statement

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