Synergistically promoting proton-coupled electron transfer of oxygen reduction with dual atomic sites on high-curvature carbon onions for highly efficient Zn–air batteries

Abstract

The oxygen reduction reaction is significantly important for metal–air batteries, yet the sluggish reaction kinetics limited by slow proton-coupled electron transfer has hindered their further application. Here, dual Cr and Fe atoms are incorporated into a high-curvature carbon nano-onion (Onion-CrFe_DSA) to demonstrate significantly synergistic regulation of proton generation and transfer from the dual catalytic centers. The electrochemical measurements confirmed that the dual-site configuration in Onion-CrFe_DSA could enhance the ORR performance with a half-wave potential of 0.916 V and a kinetic current density of 26.45 mA cm⁻². Besides, a high turnover frequency (TOF) of 7.44 s⁻¹ together with a high mass activity of 51.42 A mg_Fe⁻¹ are also achieved for the as-obtained dual atomic catalysts. A series of operando techniques, combined with density functional theory calculations, revealed that Cr atoms mainly contribute to water dissociation and proton transfer, while Fe atoms predominantly catalyze oxygen reduction and intermediate conversion. Moreover, the Onion-CrFe_DSA-assembled aqueous and quasi-solid zinc–air batteries achieve impressively high maximum power densities of 314.7 mW cm⁻² and 163.3 mW cm⁻², representing a top-tier Fe-based ORR catalyst. This work proposes a feasible way to enhance the ORR performance by engineering adjacent catalytic centers to cooperatively mediate the proton-coupled electron transfer.

---

Introduction

Clean energy conversion and storage have attracted tremendous attention owing to their great potential for replacing fossil fuels.^1–4 Zinc–air batteries (ZABs) are typical materials that exhibit high energy conversion efficiency with net zero-carbon emissions.^5–8 However, the sluggish cathodic oxygen reduction reaction (ORR) kinetics still relies on the use of Pt-based catalysts, which suffer from the scarcity and high prices.^9–11 Among various newly discovered catalysts, carbon-supported transition metal single-atom catalysts (SACs) have been widely studied owing to their large specific surface area, high atom utilization, and tunable local configuration.^12,13 Thus, Fe-based SACs with nitrogen coordination configuration (Fe–N–C) have shown ideal ORR performance.^14–16 To further improve the intrinsic performance of Fe–N–C active centers, tremendous efforts have been made, such as introducing heteroatoms,^17–19 breaking asymmetric N₄ coordination,²⁰ and inducing local charge polarizations.^21,22 Beyond single-atom catalysts (SACs), dual-atom catalysts (DACs) have recently emerged as an advanced class of materials, offering the potential for synergistic effects that break the scaling relationships. For the ORR, which involves multiple proton-coupled electron transfer steps, designing DACs with spatially distinct sites to separately optimize the water dissociation (proton generation)²³ and oxygen reduction steps represents a highly promising strategy.^24–26

High-curvature morphological substrates have been demonstrated to induce local charge polarizations and enhance the accessibility of active sites.^27,28 In addition, strained molecules supported on high curvature supports can directly regulate the potential-determining step (PDS) of the ORR via lowering the energy barrier.²⁹ Given that the alkaline ORR involves multiple proton-coupled electron transfer (PCET) steps, the protons mainly originate from water dissociation.^30,31 Constructing adjacent centers for water dissociation is a feasible way to promote the PCET process. Cr, with multivalent states and strong oxophilicity, has shown superior capacity in water dissociation.³² For example, Cr can act as a Lewis acid site to promote OH⁻ adsorption and water activation.³³ Wan et al. demonstrated that the incorporation of Cr atoms can disturb the hydrogen bond network during NO₃ reduction, which facilitates the hydrogenation of intermediates.³⁴ Hence, constructing adjustment catalytic centers to balance water dissociation and proton transfer for the promoted PCET process is of great significance for the further development of the ORR.

In this work, a high-curvature carbon-supported dual single-atom (DSA) catalyst was designed and fabricated to investigate the synergistic enhancement of proton generation/transfer and oxygen reduction during the ORR. Cr atoms with favourable water adsorption capability were first embedded into a carbon nano-onion (marked as Onion-Cr_SA). Then, FePc was anchored on the Onion-Cr_SA substrate, yielding a dual-atom catalyst with Cr-embedded carbon nano-onion-supported FePc (Onion-CrFe_DSA). The designed Onion-CrFe_DSA exhibits outstanding ORR performance with a high half-wave potential (E_1/2) of 0.916 V and a kinetic current density (J_k) of 26.45 mA cm⁻² (at 0.88 V). Besides, the Onion-CrFe_DSA-assembled aqueous and flexible ZABs exhibit maximum power densities of 314.7 mW cm⁻² and 163.31 mW cm⁻², significantly higher than that of commercial 20% Pt/C. A series of operando characterization studies combining theoretical calculations demonstrated that the synergistic effects between adjacent Cr and Fe sites, as well as the high-curvature carbon substrate, can significantly alter the local electronic structure and facilitate proton generation and transfer. This study highlights the critical role of adjacent catalytic centers in promoting the PCET process for ORR kinetics, delivering valuable insight and an accessible approach for dual-site synergy on a curved carbon substrate for the rational design of high-performance electrocatalysts.

Results and discussion

Synthesis and characterization

The synthetic strategy of Onion-CrFe_DSA is illustrated in Fig. 1a. First, Cr single atoms were incorporated into the carbon nano-onion framework to form an Onion-Cr_SA substrate. Then, FePc molecules were anchored onto the Onion-Cr_SA substrate to fabricate a carbon nano-onion-supported CrFe dual single-atom catalyst. X-ray diffraction (XRD) was performed to investigate the crystal structure of the as-obtained samples. The XRD pattern of Onion-CrFe_DSA exhibits two broad peaks at around 26° and 44°, attributed to the (002) and (101) diffraction planes of the carbon framework. In particular, the XRD patterns of the as-prepared samples show similar features to those of the carbon nano-onion, indicating that the incorporation of Cr and Fe atoms has a negligible influence on the crystal phases (Fig. S1). Besides, no metallic Cr and Fe peaks could be observed in Onion-CrFe_DSA, implying the atomic dispersion of metal atoms rather than aggregation into metallic nanoparticles. The morphological information for the Onion-CrFe_DSA and references was further obtained using an electron microscope. As shown in Fig. 1b, the low-magnification transmission electron microscopy (TEM) images of Onion-CrFe show a morphology composed of a series of aggregated carbon nano-onions. In addition, the high-resolution TEM (HRTEM) images further confirm the high-curvature features of the onion-like morphology, with diameters near 5 nm (Fig. 1c, Fig. S2 and S3). It should be noted that the carbon nano-onion morphology can not only provide a high curvature surface to expose more active sites, but also plays key roles in inducing local charge polarization. Moreover, no obvious metal nanoparticles were detected, which is consistent with previous XRD results. Aberration-corrected high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) revealed that isolated metal single atoms were highly dispersed on the carbon nano-onion matrix, indicating the successful synthesis of Onion-CrFe_DSA. Besides, a series of dual-atom pairs could be observed in Onion-CrFe_DSA compared to Onion-Cr_SA and Onion-Fe_SA, implying the successful fabrication of dual-atom pairs (Fig. 1d, Fig. S4 and S5).^35,36 The energy dispersive X-ray spectroscopy (EDS) elemental mapping images showed the uniform dispersion of C, N, O, Cr, and Fe atoms (Fig. 1e). In comparison, the uniform dispersion of the specific elements could also be observed in the as-prepared counterparts (Fig. S6 and S7). The metal loading of the as-prepared samples was also confirmed by Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP-AES), as shown in Table S1.

Fig. 1 Synthesis and basic characterization of Onion-CrFe_DSA. (a) Synthetic strategy for Onion-CrFe_DSA (the orange, cyan, blue, gray, red and pink balls stand for the Fe, Cr, N, C, O and H atoms, respectively). (b) TEM and (c) HRTEM images of the as-prepared Onion-CrFe_DSA. (d) HAADF-STEM image of Onion-CrFe_DSA with pointed dual-atom pairs. (e) Elemental mapping of Onion-CrFe_DSA and the corresponding STEM image.

The surface chemical states and the surface compositions were further studied by X-ray photoelectron spectroscopy (XPS) and X-ray absorption near edge spectroscopy (XANES). As shown in Fig. S8a, the Fe 2p XPS spectrum of Onion-CrFe_DSA displays peaks at 709.86 and 723.23 eV, corresponding to Fe 2p_3/2 and 2p_1/2 of the Fe²⁺ species with a four-nitrogen coordinated configuration.³⁷ Besides, the Fe 2p XPS spectrum of Onion-CrFe_DSA is similar to that of Onion-Fe_SA, indicating a consistent local configuration of Fe atoms (Fig. S8b). The Cr 2p XPS spectra of Onion-CrFe_DSA split into two saturated peaks at 577.27 and 587.08 eV that could be assigned to the Cr 2p_3/2 and 2p_1/2 of Cr³⁺ valence states with adsorbed oxygen species on the Cr atoms. Besides, the binding energies of Cr 2p_3/2 and 2p_1/2 for Onion-CrFe_DSA are nearly 0.3 eV higher than those in Onion-Cr_SA (577.05 and 586.85 eV for Cr 2p_3/2 and 2p_1/2), indicating that more electrons have shifted from Cr species to adjacent atoms (Fig. S9). The N 1s XPS spectra of the series of samples imply the coexistence of the graphitic, pyrrolic, metallic, and pyridinic-N species (Fig. S10). The C K-edge XANES spectrum of Onion-CrFe_DSA exhibits two distinct dominant peaks originating from 1s to π* (nearly 285 eV) and 1s to σ* (nearly 292 eV) while a series of peaks (marked as B) is located between π* and σ* bonds mainly induced by structural defects or dangling bonds with other heteroatoms (e.g., C–O/N and C=O).³⁸ As shown in Fig. S11, the Onion-CrFe_DSA and Onion-Fe_SA substrates exhibit a higher intensity of peak B than the Onion-Cr_SA substrate, which may be because of an increase in the number of C–N bonds induced by FePc incorporation.^39,40

Considering the difficulties of clarifying the dual single atom catalysts, hard X-ray absorption fine structure (XAFS) is a powerful technology to clarify the local coordination environment of the Cr and Fe atoms. As shown in Fig. 2a, the XANES spectrum for the Cr K-edge of Onion-CrFe_DSA shows a different feature from that for Onion-Cr_SA, implying a modified local configuration induced by FePc incorporation. Besides, the pre-peak originating from the 1s to 4p shakedown transition confirms the D_4h symmetry of the Cr atoms with a square planar coordination. The higher intensity of pre-peak A for Onion-CrFe_DSA than Onion-Cr_SA implies broken symmetry and structural distortion.^21,41 The Onion-CrFe_DSA substrate also exhibits a higher white-line peak intensity, indicating more unoccupied states in the 3d orbitals (inset of Fig. 2a). The Fourier transform extended-XAFS (FT-EXAFS) curve displayed in R-space for Cr atoms shows two separate peaks at 1.49 and 2.45 Å (without phase correction), which can be assigned to the Cr–N/C and cambered second-shell Cr–C bonds of the Onion-Cr_SA sample. After incorporating Fe atoms, the second-shell coordination shifts to a larger bond length of 2.53 Å (without phase correction), along with a higher coordination number, which may be due to the extra Cr–Fe scattering (Fig. 2b). The EXAFS fitting results further confirmed the local geometric configuration of the Cr atoms (Fig. 2c, S12, Table S2). In the Fe K-edge XANES spectrum of Onion-CrFe_DSA, the pre-edge A (1s to 4p transition) is significantly higher than those of Onion-Fe_SA and FePc, which may be due to the rearranged local geometrical configuration (Fig. 2d).²¹ Besides, the lower white line peak of Onion-CrFe_DSA shows the charge polarization between Cr and Fe atoms (inset of Fig. 2d).⁴² The scattering peaks at 1.45 Å and 2.33 Å in the R-space curve of Onion-CrFe_DSA could be assigned to the Fe–N/C and second-shell Fe–C/Cr bonds (Fig. 2e). Besides, the second-shell bond length of Fe–C/Cr in Onion-CrFe_DSA (2.33 Å) is smaller than that of Onion-Fe_SA (2.45 Å), which is similar to the Cr K-edge analysis (Fig. 2f, S13, Table S3). Considering that the integration of Cr and Fe atoms may lead to the rearrangement of local electronic structures, the wavelet transformed EXAFS (WT-EXAFS) plots can offer direct observation of coordination environments. As illustrated in Fig. 2g and h, the intensity maximum at around 4 Å⁻¹ is attributed to Cr–N scattering, while the peak at approximately 6.5 Å⁻¹ is assigned to Cr–C/Fe scattering. The Cr–C/Fe scattering in Onion-CrFe_DSA shows a distinct feature and a slightly larger k value than that in Onion-Cr_SA, indicating a modulated local electronic structure induced by the incorporation of Fe atoms. A similar phenomenon can also be observed in Fig. 2i and j. The enhanced second-shell scattering of the Fe–C/Cr bond in Onion-CrFe_DSA demonstrates the strongest interaction between Cr and Fe atoms.^43,44

Fig. 2 Characterization of the fine electronic structure. (a) Cr K-edge and (d) Fe K-edge XANES spectra of the as-obtained Onion-CrFe_DSA, Onion-Cr_SA, Onion-Fe_SA, and necessary counterparts. FT-EXAFS curves for (b) the Cr K-edge and (e) the Fe K-edge displayed in R-space for the as-obtained Onion-CrFe_DSA, Onion-Cr_SA, Onion-Fe_SA, and necessary counterparts. EXAFS fitting curves for (c) the Cr K-edge and (f) the Fe K-edge of Onion-CrFe_DSA. (g) and (j) WT-EXAFS plots for the Cr K-edge and the Fe K-edge of Onion-Cr_SA, Onion-CrFe_DSA and Onion-Fe_SA.

Electrocatalytic performance of the as-prepared samples

The electrocatalytic ORR performance of the series of as-prepared samples was then evaluated using rotating disk electrode (RDE) measurements in 0.1 M KOH electrolyte. The cyclic voltammetry (CV) curve of Onion-CrFe_DSA exhibited a prominent oxygen reduction peak in an O₂-saturated electrolyte, while no obvious reduction peak was detected in an N₂-saturated electrolyte, indicating its superior oxygen reduction capacity (Fig. S14). The linear sweep voltammetry (LSV) curves at 1600 rpm were also obtained to further compare the ORR performance (Fig. 3a). Onion-CrFe_DSA exhibits an onset potential (E_onset) of 0.986 V (vs. RHE) along with a half-wave potential (E_1/2) of 0.916 V (vs. RHE), which is better than those of Onion-Cr_SA (0.839 and 0.682), Onion-Fe_SA (0.949 and 0.865), and 20% Pt/C (0.999 and 0.884). Besides, the smallest Tafel slope of 65.42 mV dec⁻¹ for Onion-CrFe_DSA demonstrates its fast reaction kinetics (Fig. 3b). Considering the potential spatial distance induced by FePc loading, the higher loading of FePc was also measured to confirm the favourable loading of Fe and the spatial distance of Cr and Fe atoms (Fig. S15). In addition, the amplified gram-scale synthesis of Onion-CrFe_DSA also shows a similar LSV curve at 1600 rpm, indicating the reproducibility of the as-prepared sample (Fig. S16). The ORR kinetics were also investigated by measuring the LSV curves at various rotating speeds ranging from 225 to 2025 rpm (Fig. 3c and S17). The limited current density (J_L) increased with increasing rotation speed, which originates from the shortened diffusion distance and accelerated mass transport at the electrode surface. Besides, the Koutecky–Levich (K–L) plots obtained from the LSV curves show good linearity with fitted values near four, indicating a four-electron reaction pathway (Fig. S18).⁴⁵ Rotating ring-disk electrode (RRDE) measurement was then conducted to evaluate the electron transfer number and H₂O₂ yield (Fig. S19). As shown in Fig. 3d, Onion-CrFe_DSA displays an electron transfer number of 3.99 and an exceptionally low H₂O₂ yield of 0.19% at 0.50 V, which is much superior to those of the contrast samples.

Fig. 3 Evaluation of the ORR performance. (a) LSV curves of Onion-CrFe_DSA, Onion-Fe_SA, Onion-Cr_SA, Onion and 20% Pt/C in 0.1 M KOH electrolyte at 1600 rpm. (b) The corresponding Tafel slopes of Onion-CrFe_DSA, Onion-Fe_SA, Onion-Cr_SA, Onion and 20% Pt/C obtained from the non-Faraday zones. (c) LSV curves of Onion-CrFe_DSA at various rotating rates. (d) H₂O₂ yield and the electron transfer number obtained during the ORR process of Onion-CrFe_DSA, Onion-Fe_SA, Onion-Cr_SA, Onion and 20% Pt/C. (e) Comparison of ORR performance between Onion-CrFe_DSA and the as-prepared counterparts. (f) Chronoamperometric responses of Onion-CrFe_DSA and 20% Pt/C with methanol injected at around 200 s. (g) LSV curves obtained before and after 10 000 CV cycles and (h) durability test performed for 12 h at 0.55 V.

Furthermore, the kinetic current density (J_k) of Onion-CrFe_DSA was calculated to be 26.45 mA cm⁻² at 0.88 V, which is 7 and 3.94 times higher than those of Onion-Fe_SA and 20% Pt/C, respectively (Fig. S20a). As key parameters reflecting the intrinsic catalytic performance of the active sites, the turnover frequencies (TOFs) and mass activities of the series samples were also calculated.⁴⁶ As shown in Fig. S20b, the TOF value of Onion-CrFe_DSA is 7.44 s⁻¹ (at 0.8 V vs. RHE), which is 8.18 and 4.96 times higher than those of Onion-Fe_SA and 20% Pt/C, respectively. Similar results were also observed in the calculation of mass activities, where Onion-CrFe_DSA exhibits the highest mass activity of 51.42 A mg_Fe⁻¹. As shown in Fig. 3e, the ORR performance of Onion-CrFe_DSA surpassed those of the as-prepared counterparts, indicating the superior regulatory effect of dual-atom catalysts supported on high-curvature substrates. The double layer capacitance (C_dl) value calculated by CV measurements of Onion-CrFe_DSA was found to be 4.63 mF cm⁻², which is larger than those of Onion-Fe_SA, Onion-Cr_SA, and Onion (4.35 mF cm⁻²) (Fig. S21). Subsequently, the SCN⁻ ion poisoning tests confirmed the key role of FePc in the ORR (Fig. S22). The current density of Onion-CrFe_DSA shows negligible attenuation to 97.03% after methanol injection at 200s, indicating the favorable resistance to methanol (Fig. 3f). From the stability test, only a slight E_1/2 degradation of 10 mV occurred after 10 000 consecutive CV cycles, while the relative current remaining at 95.13% after a 12 h durability test at 0.55 V could be observed, indicating superior stability (Fig. 3g and h). Besides, the leached Fe and Cr atoms were calculated to be 3.6% and 0.3% after the stability test, indicating the high resistance to degradation of the dual atoms.

Mechanism investigation using operando techniques

The integration of operando techniques offers a methodology for dynamically tracking active sites and intermediate evolution under working conditions. Here, the local coordination environments of Cr and Fe atoms in the reaction process were firstly tracked by operando XAFS. As shown in Fig. 4a, a slight change in the white line peak at 0.9 V can be observed compared to those at other potentials in the Cr K-edge XANES spectra, which may be attributed to the water dissociation on Cr sites that requires more charge transfer from the 3d orbitals.⁴⁷ In addition, the relative valence states of Cr atoms during the reaction were evaluated by referencing Cr foil and Cr₂O₃ as standard samples. As shown in Fig. 4b, the calculated valence states of Cr gradually decrease with decreasing potential, indicating the reduction of reactants during the ORR process.⁴⁸ When the applied potential is withdrawn, the valence state of Cr returns to 4.31, which is close to that at 1.0 V, implying a reversible local configuration of Cr atoms. Additionally, the higher valence states than two of Cr atoms are mainly due to their high metallicity, which allows more oxygen species to be adsorbed during the reaction process. The FT-EXAFS curves displayed in the R-space further confirm the evolution of the local configuration of Cr atoms (Fig. 4c, S23, Table S2). When a potential of 1.0 V is applied, the coordination number (CN) of Cr–N/C/O is found to be 5.43, which can be attributed to adsorbed water or oxygen molecules. In addition, the CNs of Cr–N/C/O are 4.45 and 5.34 at 0.9 V and 0.8 V, respectively, which could be attributed to the fast reaction kinetics during oxygen reduction. Interestingly, the CN decreases to 4.58 without an applied bias, which may be due to the reconstructed structure of the Cr sites.

Fig. 4 Operando characterization of Onion-CrFe_DSA during the ORR. (a) Cr K-edge and (d) Fe K-edge XANES spectra under different operation potentials. (b) Potential-dependent evolution of the calculated valence states of Cr in comparison with Cr foil and Cr₂O₃. (e) Potential-dependent evolution of the calculated valence states of Fe performed using Fe foil and Fe₂O₃ as standards. FT-EXAFS curves of (c) the Cr K-edge and (f) the Fe K-edge displayed in R-space at different operation potentials. (g) Operando ATR-SEIRAS spectra of Onion-CrFe_DSA at different working potentials during the ORR process. (h) Operando Raman spectra of Onion-CrFe_DSA at different working potentials during the ORR process.

Considering that the Fe atoms are widely recognized as high-performance ORR active centers, we further recorded the operando XAFS spectra of Fe. As shown in Fig. 4d, the XANES spectra of the Fe K-edge under different applied potentials were also recorded to evaluate the evolution of the active centers. The white line peaks show distinct changes compared to those of Cr during the ORR process, which may be induced by the strong interaction between reactants/intermediates and Fe atoms.^49,50 Besides, the calculated valence states of Fe at potentials from 1.0 V to 0.8 V are higher than that of the fresh sample, indicating the adsorption of oxygen or intermediates on the Fe atoms. When the potential is withdrawn, the valence states of Fe become close to that of the fresh sample, which could be attributed to the refreshed Fe sites (Fig. 4e). Furthermore, the EXAFS fitting results demonstrate the dynamic structural evolution of Fe (Fig. 4f, S24, Table S3). The CN of Fe–N/O for Onion-CrFe_DSA is slightly greater than four, which may originate from the extra coordination between Fe atoms and adsorbates. The CNs of Fe–N/O gradually decrease from 6.00 to 5.80 as the potential decreases from 1.0 V to 0.8 V, which is highly consistent with fast reaction kinetics. Additionally, the CN of Fe–N/O decreases to 3.91 after the bias is removed, implying reconstructed active sites without adsorbates. Considering the second-shell coordination, the dynamic changes in CNs indicate structural evolution of the FePc molecules. Hence, we can conclude from the operando XAFS measurements that there is a dynamic synergistic catalytic process between Cr and Fe atoms. However, the XAFS results can only provide information about the coordination environment of the central atoms, while the key intermediates related to the reaction pathways and their evolution remain unclear.

Considering the intermediate transformation, operando Raman and attenuated total reflection surface enhanced infrared absorption spectroscopy (ATR-SEIRAS) were conducted to precisely detect the key intermediates during the ORR process.⁵¹ As shown in Fig. 4g, the infrared bands at 1015 cm⁻¹ and 1126 cm⁻¹ originate from *OOH and O₂⁻, respectively, indicating the transformation of oxygen-associated intermediates.⁵² We find that the wavenumbers of these intermediates for Onion-CrFe_DSA are higher than those for Onion-Fe_SA (1008 and 1120 cm⁻¹), which may be due to the larger vibration frequency that facilitates the cleavage of the O–O bond to form *O (Fig. S25a). In the case of the Onion-Cr_SA sample, no obvious *OOH and O₂⁻ signals were observed during the ORR process (Fig. S25b), further demonstrating its sluggish ORR performance. Besides, the wavenumber of the H–O–H bond corresponding to water molecules is 1640 cm⁻¹ for Onion-CrFe_DSA, higher than those of Onion-Cr_SA (1635 cm⁻¹) and Onion-Fe_SA (1634 cm⁻¹), implying enhanced adsorption and water dissociation on Onion-CrFe_DSA. Considering that protons mainly originate from water molecules during the ORR process, both water dissociation and proton transfer are important to the PCET process. The infrared band in the range of 3200–3600 cm⁻¹ can be attributed to the interfacial water configuration and hydrogen bond networks.⁵³ The infrared bands of interfacial water for Onion-Cr_SA and Onion-Fe_SA gradually shift to lower wavenumbers during the ORR process, indicating the formation of more tetrahedrally coordinated water molecules with a denser hydrogen bond network. As for Onion-CrFe_DSA, there is a negligible shift of the infrared band, indicating the stable interfacial water structure during the reaction process. Considering the different catalytic performance of these as-prepared samples, we speculate that the dense hydrogen bond in Onion-Cr_SA and Onion-Fe_SA may hinder proton generation, leading to an insufficient proton supply. In addition to intermediate transfer, we also conducted operando Raman spectroscopy to further verify the reaction coordinate. As shown in Fig. 4h, peaks at 1150 and 1524 cm⁻¹ assigned to O₂⁻ and OOH*, which are regarded as key intermediates for the ORR, could be observed under applied potentials.^53,54 Compared to Onion-Fe_SA, the weaker intensity of O₂⁻ in Onion-CrFe_DSA suggests an accelerated protonation process (Fig. S26a).⁵⁴ Furthermore, only O₂⁻ was observed in Onion-Cr_SA, whereas OOH* was absent. This phenomenon is attributed to the strong oxygen absorption affinity and limited oxygen activation capacity of Cr atoms (Fig. S26b). Besides, the raised Cr–OH bond at 679 cm⁻¹ also demonstrated the strong affinity towards H₂O. Hence, we conclude from these operando characteristics that the introduction of adjacent Cr atoms effectively promotes proton generation and transfer during the ORR process.⁵⁵ We further prepared Onion-CoFe_DSA, Onion-MnFe_DSA and Onion-Fe_SA@WC to confirm the vital role of water dissociation in the ORR (Fig. S27); Onion-Fe_SA@WC shows the closest ORR performance to Onion-CrFe_DSA, which adequately confirms that the promoted water dissociation by WC can also enhance the PCET process.

Theoretical calculations

To gain a deeper insight into the origin of the enhanced ORR performance in the high-curvature carbon framework and elucidate the synergistic catalytic roles of Cr–Fe dual sites, density functional theory (DFT) calculations were also conducted.^56–59 The high curvature carbon substrate was considered in constructing a theoretical model to clarify the intrinsic mechanism. A series of Onion-Cr_SA models with various Cr coordination environments was constructed to identify the most thermodynamically stable configuration (Fig. 5a, Fig. S28). Considering Cr is susceptible to being oxidized in aqueous solutions due to its high electrochemical activity, we introduced hydroxyl groups at exposed Cr sites to accurately simulate the reaction conditions. Since water dissociation is the initial step of the ORR in alkaline media and serves as the proton source for the subsequent PCET process, we first investigated this process on both Onion-Cr_SA and FePc surfaces. As shown in Fig. S29, water molecules exhibit stronger binding to the Cr site in Onion-Cr_SA, with a more negative adsorption energy (−0.41 eV) than the Fe site in FePc (−0.11 eV), indicating a clear site preference. Furthermore, the energy barrier for water dissociation at the Cr site is substantially lower (0.62 eV) compared to the Fe site (2.24 eV), identifying Cr as the primary active center for water adsorption and activation (Fig. 5b). With sufficient proton sources supplied via water dissociation, the subsequent PCET steps can be effectively triggered. To further clarify the overall reaction mechanism, Gibbs free energy profiles were then calculated for the complete ORR pathway (Fig. 5c and d and Fig. S30). The four sequential PCET steps reveal that when the Cr site (Onion-Cr_SA) acts as the active center, the potential-determining step (PDS) is the desorption of *OH with an energy barrier of 1.07 eV. In contrast, when the Fe site is involved, the PDS shifts to the conversion of *O₂ to *OOH, associated with a reduced energy barrier of 0.46 eV. These results suggest that, within the dual metal atoms, the Cr sites in Onion-Cr_SA excel in water dissociation and proton generation, while the Fe sites in FePc are better suited for oxygen reduction and transformation. Together, the Cr–Fe dual active centers could synergistically leverage their individual advantages to collectively promote the ORR activity.

Fig. 5 DFT calculations of the reaction mechanisms. (a) Top and side views of the FePc and Onion-Cr_SA catalysts, in which the orange, cyan, blue, gray, red, and pink balls stand for the Fe, Cr, N, C, O and H atoms, respectively. (b) Computed water dissociation pathway on the FePc and Onion-Cr_SA surfaces. (c) Gibbs free energy diagrams for the ORR on the FePc and Onion-Cr_SA surfaces. (d) The reaction pathway for the ORR with the key intermediates on the FePc catalyst. (e) Illustration of the d-band center shift for the metal site of FePc and Onion-Cr_SA (the orange, cyan, blue, gray, red and pink balls stand for the Fe, Cr, N, C, O and H atoms, respectively). (f) Illustration of the synergistically catalytic mechanism in this work.

To further demonstrate the catalytic roles of active sites, the electronic structures of the active centers were analyzed. As shown in Fig. 5e, the d-band center of the Cr site lies at −0.59 eV, which is close to the Fermi level, indicating strong interactions with the adsorbates. This specific electronic configuration facilitates efficient adsorption and activation of water molecules, thus promoting water dissociation. However, the strong binding affinity may impede desorption of oxygen-containing intermediates, making *OH desorption the PDS and limiting the contribution of the Cr site to the oxygen reduction process. As such, the Cr center primarily serves as the catalytic site dedicated to water dissociation. In contrast, the Fe site in FePc exhibits a deeper d-band center at −1.18 eV, leading to a weaker adsorption strength. While this limits the water activation capability, it enables balanced binding and desorption of oxygen species. As a result, the PDS shifts to the conversion of *O₂ to *OOH, and Fe acts as the preferred site for oxygen reduction. In addition, the reaction pathway for the entire system of Onion-CrFe_DSA was also simulated to further clarify the specific reaction process. The corresponding structural models of the onion framework are illustrated in Fig. S31, where FePc is anchored on the highly curved carbon nano-onion substrates adjacent to the Cr single-atom sites. The computed reaction profiles indicate that water dissociation at the Cr site of Onion-CrFe_DSA exhibits a relatively low activation barrier (0.33 eV), suggesting that Cr sites remain highly effective for proton generation. Meanwhile, for the ORR pathway, when the Fe site (Onion-CrFe_DSA) acts as the active center, a moderate energy barrier of 0.57 eV can be found for the PDS of *OH desorption. These energetically accessible barriers for both water dissociation and oxygen reduction further substantiate the distinct yet complementary roles of Cr and Fe centers in the Onion-CrFe_DSA catalyst. As reflected by the experimental and theoretical evidence, a synergistic mechanism can be proposed (Fig. 5f): (1) the high curvature carbon nano-onion offers ideal anchoring sites for the successful incorporation of Cr and Fe atoms during the synergistic process; (2) the Cr atoms mainly catalyze the water dissociation and generate more protons near the Fe atoms; (3) the generated protons migrate to the adjacent Fe sites and promote the oxygen transformation and reduction; and (4) the spatial separation and function cooperation of dual-atom Cr and Fe sites synergistically enhance the overall ORR process.

Catalytic performance in zinc–air batteries

Motivated by the superior ORR performance in alkaline electrolytes, a ZAB was assembled to evaluate its practical application (Fig. 6a).^6,14 The open-circuit voltage (OCV) of the Onion-CrFe_DSA-based ZAB reaches a high value of 1.55 V, exceeding that of Pt/C (1.45 V), and a light-emitting diode (LED) can be easily lit by two tandem batteries (Fig. 6b and S32). The Onion-CrFe_DSA-assembled ZAB delivered a maximum power density of 314.7 mW cm⁻², which is significantly higher than that of the commercial Pt/C-assembled one (144.6 mW cm⁻²) (Fig. 6c). Besides, the specific capacity of the Onion-CrFe_DSA-assembled ZAB was 820.2 mAh g_Zn⁻¹ (energy density of 1048.7 Wh kg_Zn⁻¹) at a current density of 10 mA cm⁻² (Fig. 6d). In addition, the galvanostatic discharge curves of the Onion-CrFe_DSA-assembled ZAB at current densities ranging from 5 to 50 mA cm⁻² showed a higher voltage than those of the Pt/C-assembled ZAB, and the discharge voltage recovered to 1.33 V when the current density was returned to 5 mA cm⁻², suggesting superior discharge capacity and reversibility (Fig. 6e). Furthermore, no significant attenuation was observed during the 32 h galvanostatic discharging test at a current density of 10 mA cm⁻², indicating favorable stability (Fig. 6f). Furthermore, a rechargeable ZAB was assembled using Onion-CrFe_DSA + RuO₂ as the cathode and Pt/C + RuO₂ as the anode counterpart. As shown in Fig. 6g, the Onion-CrFe_DSA- and RuO₂-assembled ZAB exhibits superior charge–discharge performance over 400 h (∼1200 cycles). Notably, the initial charge–discharge voltage gap was 0.74 V, increasing only slightly to 0.85 V after 400 h of testing, representing a significantly smaller voltage drift compared to Pt/C + RuO₂ (which increased from 0.69 V to 0.94 V within 60 h). The XPS and XAFS spectroscopy measurements also demonstrated the stability of the as-prepared Onion-CrFe_DSA (Fig. S33).

Fig. 6 Aqueous ZAB performance. (a) Schematic illustration of the Zn–air battery. (b) OCV measurement of the assembled batteries. (c) Polarization and power density curves of the Onion-CrFe_DSA- and 20% Pt/C-assembled ZABs. (d) Specific capacity calculated based on the consumed mass of Zn foil at a current density of 10 mA cm⁻² for Onion-CrFe_DSA and 20% Pt/C. (e) Galvanostatic discharging current densities of the ZABs ranging from 5 to 50 mA cm⁻². (f) Durability test of the ZAB for Onion-CrFe_DSA at a current density of 10 mA cm⁻². (g) Galvanostatic charge–discharge cycling curves at a current density of 10 mA cm⁻².

We further assembled a flexible ZAB with a polymer gel electrolyte to investigate its energy conversion performance (Fig. 7a).^60,61 Carbon cloth loaded with catalysts and Zn foil served as the air cathode and anode, respectively, while PAM/MMT was used as the electrolyte. The flexible state ZAB assembled with Onion-CrFe_DSA delivered a high open circuit potential of 1.41 V, which is higher than that of commercial Pt/C (1.31 V) (Fig. S34). The LED could be illuminated by the Onion-CrFe_DSA-based flexible ZAB at different bending angles (Fig. S35). Surprisingly, the maximum power density reached 163.31 mW cm⁻² for Onion-CrFe_DSA, which is significantly higher than the value for commercial Pt/C (116.15 mW cm⁻²) (Fig. 7c). Besides, the rate performance measured at various discharge current densities ranging from 1 mA cm⁻² to 5 mA cm⁻² demonstrated that Onion-CrFe_DSA exhibits better discharge performance (Fig. S36). Additionally, the Onion-CrFe_DSA-assembled flexible Zn–air battery delivered a specific capacity of 817.11 mAh g⁻¹, outperforming the Pt/C (20 wt%)-based counterpart (705.47 mAh g⁻¹) (Fig. S37). Furthermore, the Onion-CrFe_DSA-based flexible ZAB maintained stable charge/discharge cycling for 20 h at a current density of 2 mA cm⁻², indicating its excellent stability (Fig. 7e). Hence, the above conclusions fully confirm the significant application potential of Onion-CrFe_DSA for high-efficiency energy conversion and storage.

Fig. 7 Flexible ZAB performance of Onion-CrFe_DSA. (a) A quasi-solid state ZAB structure model diagram. (b) Discharging polarization curves and the corresponding power densities. (c) Galvanostatic charge–discharge cycling curves at a current density of 2 mA cm⁻².

Conclusions

In summary, a dual Cr and Fe electrocatalyst was fabricated to study the synergistic acceleration of the PCET process, driven by spatially adjacent catalytic centers for proton generation/transfer and oxygen reduction. The designed Onion-CrFe_DSA sample exhibits superior ORR performance with an E_1/2 of 0.916 V and a J_k of 26.45 mA cm⁻² along with a high TOF value (7.44 s⁻¹) and mass activity (51.42 A mg_Fe⁻¹). Besides, an aqueous ZAB assembled with Onion-CrFe_DSA reaches a high power density of 314.7 mW cm⁻² and a specific capacity of 820.2 mAh g_Zn⁻¹, along with ideal charge–discharge cycling performance. Operando XAFS measurements revealed that the Fe and Cr atoms undergo dynamic evolution along with the transformation of intermediates under applied potentials. Further operando Raman and ATR-SEIRAS spectroscopy combined with theoretical calculations verified that Cr atoms synergistically promote proton generation and transfer, while Fe atoms govern oxygen reduction and intermediate conversion, thereby accelerating the reaction kinetics.

In future, the strategy of decoupling proton supply and oxygen reduction tasks onto spatially adjacent but functionally specialized sites could be extended to other PCET-involved processes, such as the oxygen evolution and the carbon dioxide reduction reactions. Future research could focus on exploring more diverse metal pair combinations (e.g., pairing early and late transition metals) to further tailor the PCET process. Moreover, investigating the dynamic evolution of dual sites under realistic device operating conditions using advanced operando techniques will be crucial for bridging the gap between fundamental understanding and practical application. Finally, scaling-up the synthesis of these tailored dual-atom architectures while maintaining precise atomic control remains a key challenge and opportunity for the large-scale deployment of next-generation energy conversion technologies.

Author contributions

L. Song and L. Shan supervised the project. Y. Lin, Q. Wang and B. Geng carried out most of the experiments and data analysis. Y. Bao, and A. Ke helped to perform the electrochemical tests. C. Wang helped to assemble the Zn–air batteries and analyzed the data. H. Liu helped to perform the operando ATR-SEIRAS measurement and co-analyzed the data. X. Liu helped to carry out the operando Raman measurement. R. Zheng and L. Yang performed the theoretical calculations and gave helpful discussions. All the authors discussed the results and assisted during the manuscript preparation.

Conflicts of interest

The authors declare no competing financial interests.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information is available. The supplementary information containing materials and necessary data. See DOI: https://doi.org/10.1039/d5nr03918e.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (12225508, 12205301, and U23A20121), the National Key Research and Development Program of China (2022YFA1403203), the University Synergy Innovation Program of Anhui Province (GXXT-2023-036), the Anhui Provincial Department of Education University Natural Science Research Project (2023AH050113 and 2024AH050070), and the Natural Science Foundation of Anhui Province (2308085MB47). We thank the staff members of the Infrared Spectroscopy and Microspectroscopy Beamline (https://cstr.cn/31131.02.HLS.IRSM), XMCD (https://cstr.cn/31131.02.HLS.XMCD.b) at the Hefei Light Source (https://cstr.cn/31131.02.HLS), the Hard X-ray Absorption Fine Structure (XAFS) Spectroscopy Beamline (https://cstr.cn/31124.02.SSRF.BL14W1) and the experiment assist system (https://cstr.cn/31124.02.SSRF.LAB) at the Shanghai Synchrotron Radiation Facility (https://cstr.cn/31124.02.SSRF) for providing technical support and assistance in data collection and analysis. We also thank the Hefei Advanced Computing Center and the High-Performance Computing Platform of Anhui University for computational support.

References

1. J. Guo, Y. Zheng, Z. Hu, C. Zheng, J. Mao, K. Du, M. Jaroniec, S.-Z. Qiao and T. Ling, Nat. Energy, 2023, 8, 264–272.
2. Q.-P. Zhao, W.-X. Shi, J. Zhang, Z.-Y. Tian, Z.-M. Zhang, P. Zhang, Y. Wang, S.-Z. Qiao and T.-B. Lu, Nat. Synth., 2024, 3, 497–506.
3. K. Liu, J. Fu, Y. Lin, T. Luo, G. Ni, H. Li, Z. Lin and M. Liu, Nat. Commun., 2022, 13, 2075.
4. Y. Xin, Z. Liu, C. Wang, L. Wang, Z. Zhou, L. Yang, H. Hao, L. Jiang, D. Zhang and J. Jiang, J. Materiomics, 2025, 11, 100965.
5. F. Xiao, Q. Wang, G.-L. Xu, X. Qin, I. Hwang, C.-J. Sun, M. Liu, W. Hua, H.-W. Wu and S. Zhu, Nat. Catal., 2022, 5, 503–512.
6. B. Tang, Y. Zhou, Q. Ji, Z. Zhuang, L. Zhang, C. Wang, H. Hu, H. Wang, B. Mei, F. Song, S. Yang, B. M. Weckhuysen, H. Tan, D. Wang and W. Yan, Nat. Synth., 2024, 3, 878–890.
7. J. Liu, W. Chen, S. Yuan, T. Liu and Q. Wang, Energy Environ. Sci., 2024, 17, 249–259.
8. S. Lee, J. Choi, M. Kim, J. Park, M. Park and J. Cho, Chem. Sci., 2022, 13, 6159–6180.
9. Q. Zhou, X. He, M. Min, M. Song, P. Wang, W. Huang and C. Xiong, Composites, Part B, 2025, 307, 112877.
10. J. Ma, J. Li, Y. Tian, J. Liu and R. Wang, Rare Met., 2025, 44, 5412–5427.
11. Q. Zhou, M. Song, Y. Tian, M. Min, S. Cui, X. He and C. Xiong, J. Colloid Interface Sci., 2025, 677, 59–67.
12. W. Guo, Z. Wang, X. Wang and Y. Wu, Adv. Mater., 2021, 33, 2004287.
13. Y. Shang, X. Xu, B. Gao, S. Wang and X. Duan, Chem. Soc. Rev., 2021, 50, 5281–5322.
14. M. Liu, Y. Liu, X. Zhang, L. Li, X. Xue, M. Humayun, H. Yang, L. Sun, M. Bououdina, J. Zeng, D. Wang, R. Snyders, D. Wang, X. Wang and C. Wang, Angew. Chem., Int. Ed., 2025, e202504923.
15. J. Liu, Y. Liu, B. Nan, D. Wang, C. Allen, Z. Gong, G. He, K. Fu, G. Ye and H. Fei, Angew. Chem., Int. Ed., 2025, 64, e202425196.
16. S. Ji, Y. Mou, H. Liu, X. Lu, Y. Zhang, C. Guo, K. Sun, D. Liu, J. H. Horton, C. Wang, Y. Wang and Z. Li, Adv. Mater., 2024, 36, 2410121.
17. M. Du, B. Chu, Q. Wang, C. Li, Y. Lu, Z. Zhang, X. Xiao, C. Q. Xu, M. Gu, J. Li, H. Pang and Q. Xu, Adv. Mater., 2024, 36, 2412978.
18. R. Sui, B. Liu, C. Chen, X. Tan, C. He, D. Xin, B. Chen, Z. Xu, J. Li, W. Chen, Z. Zhuang, Z. Wang and C. Chen, J. Am. Chem. Soc., 2024, 146, 26442–26453.
19. J. Liu, Q. Ye, Y. Wu, J. Chen, Y. Ge, L. Zhang, R. Abazari and J. Qian, Sci. China: Chem., 2025, 68, 1859–1868.
20. S. Wei, R. Yang, Z. Wang, J. Zhang and X. H. Bu, Adv. Mater., 2024, 36, 2404692.
21. X. Li, X. Wu, Y. Zhao, Y. Lin, J. Zhao, C. Wu, H. Liu, L. Shan, L. Yang, L. Song and J. Jiang, Adv. Mater., 2023, 35, 2302467.
22. C. B. Musgrave III, J. Su, P. Xiong, Y. Song, L. Huang, Y. Liu, G. Li, Q. Zhang, Y. Xin, M. M.-J. Li, R. T. K. Kwok, J. W. Y. Lam, B. Z. Tang, W. A. Goddard and R. Ye, J. Am. Chem. Soc., 2025, 147, 3786–3795.
23. S. Sanati, R. Abazari and A. M. Kirillov, Inorg. Chem., 2024, 63, 15813–15820.
24. X. Li, H. Lei, L. Xie, N. Wang, W. Zhang and R. Cao, Acc. Chem. Res., 2022, 55, 878–892.
25. J. Sun, M. Shen, A.-J. Chang, C. Liang, C. Xiong, C. Hou, J. Li, P. Wang, J. Li and J. Huang, Chem. Eng. J., 2025, 503, 158551.
26. Y. Wu, J. Chen, J. Liu, L. Zhang, R. Abazari, T.-T. Li and J. Qian, Chem. Eng. J., 2025, 503, 158343.
27. X. Zhao, Y. Wang, J. Wen, L. Cao, F. Wang, K. Lei, R. Fang, L. Zhou and Y. Li, J. Am. Chem. Soc., 2025, 147, 19166–19177.
28. G. Chen, R. Lu, C. Li, J. Yu, X. Li, L. Ni, Q. Zhang, G. Zhu, S. Liu, J. Zhang, U. I. Kramm, Y. Zhao, G. Wu, J. Xie and X. Feng, Adv. Mater., 2023, 35, 2300907.
29. R. Cao, R. Thapa, H. Kim, X. Xu, G. M. Kim, Q. Li, N. Park, M. Liu and J. Cho, Nat. Commun., 2013, 4, 2076.
30. Y. Wu, Y. Zhang, L. Lan, T. Hu, S. Tang, D. Lützenkirchen-Hecht, K. Yuan and Y. Chen, Angew. Chem., Int. Ed., 2025, 64, e202502019.
31. D. G. Nocera, J. Am. Chem. Soc., 2022, 144, 1069–1081.
32. L. Zeng, Z. Zhao, Q. Huang, C. Zhou, W. Chen, K. Wang, M. Li, F. Lin, H. Luo and Y. Gu, J. Am. Chem. Soc., 2023, 145, 21432–21441.
33. X. Cao, H. Qin, J. Zhang, X. Chen and L. Jiao, J. Am. Chem. Soc., 2024, 146, 32049–32058.
34. Y. Wan, Y. Tang, Y. Zuo, K. Sun, Z. Zhuang, Y. Zheng, W. Yan, J. Zhang and R. Lv, Energy Environ. Sci., 2025, 18, 7460–7469.
35. X. Zhao, R. Fang, F. Wang, X. Kong and Y. Li, Nat. Commun., 2022, 13, 7873.
36. S. Tian, B. Wang, W. Gong, Z. He, Q. Xu, W. Chen, Q. Zhang, Y. Zhu, J. Yang, Q. Fu, J. Chen, Y. Bu, L. Gu, X. Sun, H. Zhao, D. Wang and Y. Li, Nat. Commun., 2021, 12, 3181.
37. Y. Lu, M. Zhu, S. Chen, J. Yao, T. Li, X. Wang and C. Tang, J. Am. Chem. Soc., 2024, 146, 23338–23347.
38. K. Feng, J. Zhong, B. Zhao, H. Zhang, L. Xu, X. Sun and S. T. Lee, Angew. Chem., Int. Ed., 2016, 55, 11950–11954.
39. Z. Mo, J. Di, P. Yan, C. Lv, X. Zhu, D. Liu, Y. Song, C. Liu, Q. Yu, H. Li, Y. Lei, H. Xu and Q. Yan, Small, 2020, 16, 2003914.
40. Z. Sun, C. Li, Z. Wei, F. Zhang, Z. Deng, K. Zhou, Y. Wang, J. Guo, J. Yang, Z. Xiang, P. Ma, H. Zhai, S. Li and W. Chen, Adv. Mater., 2024, 36, 2404665.
41. B. L. Geoghegan, Y. Liu, S. Peredkov, S. Dechert, F. Meyer, S. DeBeer and G. E. Cutsail III, J. Am. Chem. Soc., 2022, 144, 2520–2534.
42. G. S. Henderson, F. M. De Groot and B. J. Moulton, Rev. Mineral. Geochem., 2014, 78, 75–138.
43. Y. Wei, H. Xia, H. Lan, D. Xue, B. Zhao, Y. Yu, Y. Hu and J. N. Zhang, Adv. Energy Mater., 2024, 14, 2303011.
44. Y. Dai, B. Liu, Z. Zhang, P. Guo, C. Liu, Y. Zhang, L. Zhao and Z. Wang, Adv. Mater., 2023, 35, 2210757.
45. R. Zhou, Y. Zheng, M. Jaroniec and S.-Z. Qiao, ACS Catal., 2016, 6, 4720–4728.
46. Q. Luo, K. Wang, Q. Zhang, W. Ding, R. Wang, L. Li, S. Peng, D. Ji and X. Qin, Angew. Chem., Int. Ed., 2025, 64, e202413369.
47. W. Zhou, H. Su, W. Cheng, Y. Li, J. Jiang, M. Liu, F. Yu, W. Wang, S. Wei and Q. Liu, Nat. Commun., 2022, 13, 6414.
48. S. Mondal, D. Bagchi, M. Riyaz, S. Sarkar, A. K. Singh, C. Vinod and S. C. Peter, J. Am. Chem. Soc., 2022, 144, 11859–11869.
49. J. Feijóo, Y. Yang, M. V. F. Guzman, A. Vargas, C. Chen, C. J. Pollock and P. Yang, J. Am. Chem. Soc., 2023, 145, 20208–20213.
50. Z. Sun, H. Zhang, L. Cao, X. Liu, D. Wu, X. Shen, X. Zhang, Z. Chen, S. Ru, X. Zhu, Z. Xia, Q. Luo, F. Xu and T. Yao, Angew. Chem., Int. Ed., 2023, 62, e202217719.
51. H. Liu, Z. Qi and L. Song, J. Phys. Chem. C, 2021, 125, 24289–24300.
52. Y. Lin, L. Yu, L. Tang, F. Song, R. Schlögl and S. Heumann, ACS Catal., 2022, 12, 5345–5355.
53. Y. Zhao, H. C. Chen, X. Ma, J. Li, Q. Yuan, P. Zhang, M. Wang, J. Li, M. Li, S. Wang, H. Guo, R. Hu, K.-H. Tu, W. Zhu, X. Li, X. Yang and Y. Pan, Adv. Mater., 2024, 36, 2308243.
54. Y. L. Zhang, B. Liu, Y. K. Dai, L. X. Shen, P. Guo, Y. F. Xia, Z. Zhang, F. Kong, L. Zhao and Z. B. Wang, Angew. Chem., Int. Ed., 2024, 63, e202400577.
55. X. Cao, H. Qin, J. Zhang, X. Chen and L. Jiao, J. Am. Chem. Soc., 2024, 146, 32049–32058.
56. D. Liu, X. Li, S. Chen, H. Yan, C. Wang, C. Wu, Y. A. Haleem, S. Duan, J. Lu, B. Ge, P. M. Ajayan, Y. Luo, J. Jiang and L. Song, Nat. Energy, 2019, 4, 512–518.
57. X. Wu, J. Chen, M. Wang, X. Li, L. Yang, G. Li, L. Shan, X. Li, Y. Lin and J. Jiang, Chem. Commun., 2022, 58, 2914–2917.
58. Y. Zhu, Y. Jiang, H. Li, D. Zhang, L. Tao, X. Z. Fu, M. Liu and S. Wang, Angew. Chem., Int. Ed., 2024, 63, e202319370.
59. J. Su, C. B. Musgrave III, Y. Song, L. Huang, Y. Liu, G. Li, Y. Xin, P. Xiong, M. M.-J. Li and H. Wu, Nat. Catal., 2023, 6, 818–828.
60. S. Wang, M. Zhang, X. Mu, S. Liu, D. Wang and Z. Dai, Energy Environ. Sci., 2024, 17, 4847–4870.
61. J. Du, N. Zhang, W. Zhang, X. Shi, Y. Gong, R. Wang, H. Wang, J. Jin, L. Zhao and B. He, Adv. Funct. Mater., 2025, 35, 2415351.

---

Footnote

† These authors contributed equally to this work.

---