Molybdenum and fluorine co-doping induces lattice oxygen activation in Ni-Fe spinel oxides for enhanced oxygen evolution

INTRODUCTION

Green hydrogen, widely recognized as a clean and sustainable energy carrier, plays a pivotal role in the global transformation of the energy landscape^[1-3]. The primary environmentally friendly and energy-effective strategy for large-scale production of hydrogen is water electrolysis driven by renewable energy sources^[4-7]. However, the overall efficiency of this process is substantially constrained by the sluggish kinetics of the anodic oxygen evolution reaction (OER)^[8-10] since it involves a complex four-electron transfer process. The sluggish kinetics necessitate high overpotentials in electrolyzers, leading to significant energy loss, not conducive to achieving global target of carbon peak and carbon neutrality^[11-13]. This reaction predominantly follows the adsorbate evolution mechanism (AEM), in which oxygen-containing intermediates are sequentially adsorbed and desorbed at the same active site on the catalyst surface^[14,15]. This pathway establishes intrinsic linear scaling relationships among the adsorption energies of these intermediates, thereby imposing a fundamental thermodynamic overpotential limit on the OER^[16,17]. To overcome this limitation, the lattice oxygen oxidation mechanism (LOM) has emerged as a promising alternative^[18,19]. By enabling direct participation of lattice oxygen from the catalyst in molecular oxygen formation, LOM bypasses the scaling relations inherent in AEM and offers a viable route to significantly enhance intrinsic OER activity^[20-22]. Therefore, developing earth-abundant, cost-effective catalysts capable of activating lattice oxygen is crucial for scalable water splitting technologies.

In recent years, spinel oxides, particularly iron-based spinel (e.g., FeCo₂O₄, NiFe₂O₄, etc.), have emerged as highly promising electrocatalysts for the OER due to their low cost, earth abundance, and excellent electrical conductivity^[23-25]. However, their intrinsic stable electronic structures and strong metal-oxygen bonds typically constrain the OER to the conventional AEM pathway, which limits catalytic efficiency^[26,27]. This fundamental limitation highlights the critical need for precise electronic structure modulation to weaken metal-oxygen bonding and facilitate lattice oxygen activation, thereby promoting the transition from AEM to the more favorable LOM^[28,29]. Heteroatom doping is widely recognized as an effective approach to modulate the electronic structure and enhance catalytic activity^[30-33]. Among various dopants, high-valence molybdenum (Mo) and highly electronegative fluorine (F) have drawn significant attention due to their unique electronic characteristics^[34-37]. Specifically, high-valence Mo can partially substitute metal cations in AB₂O₄ spinel lattice, effectively altering the electronic distribution and creating electron-rich or electron-deficient domains of catalyst interface, thereby optimizing the intrinsic activity for water oxidation^[38,39]. Concurrently, the strong electron-withdrawing nature of F tends to distort the M-O bonding network, which facilitates lattice oxygen activation and promotes the LOM pathway^[40,41]. Furthermore, the large-scale application and commercialization of alkaline water electrolysis require catalysts with not only high intrinsic activity but also long-term stability under practical operating conditions. However, under high current density operation, catalytic performance is often limited by structural degradation. Therefore, the development of spinel-based catalysts featuring precisely engineered structures, enhanced catalytic efficiency, and exceptional durability is crucial for sustainable water splitting at industrially relevant current densities - yet significant challenges remain.

In this study, we introduced molybdenum and fluorine dopants into the nickel-iron-based spinel (Mo/F-NiFe₂O₄), aiming to enhance the oxygen evolution from a LOM pathway by cation-anion co-doping. Mo/F co-doping effectively elevates the valence band maximum of NiFe₂O₄, promoting activation of the LOM pathway during electrochemical OER. The Mo/F-NiFe₂O₄ catalyst shows excellent OER performance with low overpotentials of 247 and 311 mV to reach current densities of 50 and 100 mA cm^-2, respectively. It also exhibits remarkable stability in seawater electrolytes, maintaining stable operation for over 300 h at 100 mA cm^-2. Synchrotron radiation-based characterization and electron energy loss spectroscopy (EELS) confirmed the effective incorporation of Mo and F dopants, which promoted the formation of high-valence metal centers within the host lattice. X-ray photoelectron spectroscopy (XPS) and valence band measurements revealed that synergistic Mo/F co-doping shifted the maximum valence band (VBM) closer to the Fermi level. The upward VBM shift indicates a favorable electronic structure modification in the vicinity of the Fermi level, which is consistent with the observed changes in metal oxidation states from core-level spectroscopy and facilitates the activation of the LOM. Comprehensive electrochemical and physicochemical characterizations, supported by in situ electrochemical impedance spectroscopy (EIS) and ¹⁸O-labeled differential electrochemical mass spectrometry (DEMS), consistently demonstrated that Mo/F co-doping effectively modulates the AB₂O₄ lattice to enhance lattice oxygen reactivity, accelerate reaction kinetics, and improve charge transfer efficiency. This work thus presents a highly efficient, fundamentally advanced, and scalable catalyst for the alkaline OER.

EXPERIMENTAL

RESULTS AND DISCUSSION

Scanning electron microscopy (SEM) was applied to study the microstructural evolution of various NiFe₂O₄-based catalysts. As illustrated in Figure 1A, pristine NiFe₂O₄ displays a flower-like morphology consisting of two-dimensional nanosheets with smooth surfaces. Upon Mo doping [Figure 1B], the overall flower-like architecture is retained; however, the nanosheets exhibit increased surface wrinkling and are decorated with small particles, resulting in enhanced surface roughness. In the case of F-doped NiFe₂O₄ [Figure 1C], the nanosheet-assembled flower-like structures demonstrate a significantly rougher surface, with well-distributed nanoparticles evident across the framework. In contrast, the Mo/F-NiFe₂O₄ [Figure 1D] preserves the flower-like morphology while displaying smoother, thinner and cleaner nanosheets, indicating that synergistic Mo and F co-doping effectively regulates the anisotropic growth of crystal planes. Scanning transmission electron microscopy (STEM) and high-resolution transmission electron microscopy (HRTEM) were further employed to investigate the detailed microstructure at higher magnifications. As shown in Figure 1E and Supplementary Figure 1A-D, all NiFe₂O₄-based catalysts exhibit nanosheet-like morphologies, consistent with the SEM observations. In Figure 1F, Mo/F-NiFe₂O₄ displays lattice fringes with spacings of 0.19 and 0.25 nm, which are assigned to the (1 3 3) and (3 1 1) planes of NiFe₂O₄, respectively, confirming its high crystallinity. This finding is further corroborated by the selected-area electron diffraction (SAED) pattern in Figure 1G, where distinct diffraction spots can be indexed to the ($$\overline{2}$$ 2 $$\overline{2}$$), (1 3 $$\overline{1}$$), and (3 1 1) planes of NiFe₂O₄. In comparison, Supplementary Figure 2A-C displays the HRTEM images of the other three catalysts. The NiFe₂O₄ exhibits lattice spacings of 0.24 and 0.25 nm, corresponding to the ($$\overline{2}$$ 2 $$\overline{2}$$) and (1 3 $$\overline{1}$$) lattice planes, respectively; the Mo-NiFe₂O₄ sample shows lattice spacings of 0.21 and 0.24 nm, corresponding to the (4 0 0) and (2 2 2) planes; while the F-NiFe₂O₄ sample also reveals lattice spacings of 0.21 and 0.25 nm, indexed to the (4 0 0) and (2 2 2) planes of NiFe₂O₄. Supplementary Figure 3A-C presents the corresponding SAED patterns, all of which are in good agreement with the HRTEM observations, further confirming the spinel structure of NiFe₂O₄. Energy-loss spectroscopy (EELS) spectra [Figure 1H] show that both the Ni-L and Fe-L edges, arising from 2p → 3d electronic transitions, shift to higher energy losses upon Mo/F co-doping relative to the pristine spinel. This shift indicates an increase in the average oxidation states of Ni and Fe, as well as enhanced metal–oxygen covalency, driven by electron redistribution among Ni, Fe, Mo, and F species. Figure 1I presents the annular dark field (ADF)-STEM image and corresponding elemental mapping images of the Mo/F-NiFe₂O₄ catalyst, revealing a homogeneous distribution of Mo and F throughout the structure. Consistently, Supplementary Figure 4 further confirms the uniform incorporation of Mo and F dopants.

Figure 1. Morphology and structure characterizations of NiFe₂O₄-based catalysts. SEM images of (A) NiFe₂O₄, (B) Mo-NiFe₂O₄, (C)F-NiFe₂O₄ and (D) Mo/F-NiFe₂O₄; (E) TEM image, (F) iFFT-filtered HRTEM image, and (G) SAED pattern of Mo/F-NiFe₂O₄; (H) Ni L-edge and Fe L-edge EELS spectra of NiFe₂O₄, Mo-NiFe₂O₄, F-NiFe₂O₄ and Mo/F-NiFe₂O₄; (I) ADF-STEM and corresponding EELS maps of Mo/F-NiFe₂O₄.

The phase structure and chemical composition of the NiFe₂O₄-based catalysts was further investigated at bulk aspects. The X-ray diffraction (XRD) pattern shown in Supplementary Figure 5A displays diffraction peaks at 30.3°, 35.6°, and 63.1°, assigned to the (220), (311), and (440) crystal planes of NiFe₂O₄ spinel (PDF No. 10-0325), respectively. In addition, weak diffraction peaks observed at 44.2°, 51.2°, and 76.4° correspond to the (111), (200), and (220) planes of metallic Ni (PDF No. 04-0850), which are attributed to the underlying NF substrate. Meanwhile, as shown in Figure 2A, Supplementary Figure 5, and Supplementary Table 1, the refined lattice parameter increased from 8.32163 Å (NiFe₂O₄) to 8.34101 Å for the Mo/F-NiFe₂O₄. This lattice expansion provides direct structural evidence for the successful incorporation of Mo and F into the spinel framework. To further verify the phase composition and surface chemical states, XPS measurements were performed. As shown in Supplementary Figure 6, the survey XPS spectrum confirms the presence of Ni, Fe, O, Mo, and F in the heteroatom-doped catalysts. As shown in Supplementary Table 2, relative to the undoped NiFe₂O₄ (23.29%), F-doping significantly increases the surface oxygen content (63.12%), while Mo-doping alone results in a considerably smaller change (29.45%). Importantly, the Mo/F co-doped catalyst attains an optimal, intermediate oxygen content (33.54%). This state of balance is of crucial importance as it adequately activates the lattice oxygen mechanism to boost activity while preventing excessive lattice oxygen that might undermine structural stability during operation^[42]. High-resolution O 1s spectra of different catalysts are presented in Figure 2B, revealing three distinct peaks at 529.8, 531.2, and 532.3 eV, assigned to lattice oxygen (M-O), oxygen vacancies (O_V), and adsorbed water species, respectively^[43]. The specific atomic content of different oxygen species can be seen in Supplementary Table 3. Notably, the binding energy of the M-O peak gradually decreases with Mo and F doping. Specifically, the Mo/F-NiFe₂O₄ exhibits a 0.50 eV negative shift compared to pristine NiFe₂O₄, whereas the single-doped Mo-NiFe₂O₄ and F-NiFe₂O₄ samples display smaller shifts of 0.30 and 0.10 eV, respectively. As shown in Supplementary Figure 7, the Ni 2p spectrum exhibits characteristic peaks at 852.8, 855.2, 856.4, 861.1 and 863 eV, which are attributed to Ni⁰, Ni²⁺, Ni³⁺, and the satellite peak, respectively^[44-46]. Likewise, Supplementary Figure 8 presents the Fe 2p spectrum with distinct peaks at 710.8, 713.2, and 717.3 eV, assigned to Fe²⁺, Fe³⁺, and their corresponding satellite feature^[47-49], respectively. To confirm the successful incorporation of Mo and F dopants, Mo 3d and F 1s XPS spectra were acquired. As shown in Supplementary Figure 9, the Mo 3d spectra of both Mo-NiFe₂O₄ and Mo/F-NiFe₂O₄ exhibit well-resolved doublet peaks at 232.1 and 235.2 eV, which are assigned to Mo 3d_5/2 and Mo 3d_3/2, respectively, confirming the presence of Mo⁶⁺ species^[50,51]. Similarly, the high-resolution F 1s spectra [Supplementary Figure 10] display a sharp peak at 684.0 eV, which is characteristic of M-F, thereby verifying the successful substitution of F into the crystal structure^[52]. To investigate the impact of doping on the electronic band structure, XPS valence band spectra were analyzed [Figure 2C]. The energy difference between the VBM and the Fermi level is 1.25 eV for pristine NiFe₂O₄. In contrast, the VBM of Mo/F-NiFe₂O₄ shifts upward to 0.97 eV, whereas Mo-NiFe₂O₄ and F-NiFe₂O₄ exhibit smaller shifts of 1.18 and 1.21 eV, respectively. This upward shift in the VBM of the Mo/F co-doped sample suggests a reduced difficulty in excitation of valence electrons to the Fermi level, promoting more efficient charge transfer and accelerating reaction kinetics. As a result, Mo/F-NiFe₂O₄ demonstrates reduced charge-transfer resistance and improved intrinsic OER activity. X-ray absorption spectroscopy (XAS) based on synchrotron radiation was employed to study the local coordination environment and valence state variations of the catalysts. As shown in Figure 2D, the Mo/F-NiFe₂O₄ exhibits a higher Ni K-edge energy compared to both NiFe₂O₄ and Ni foil, indicating that the dual synergistic modulation by Mo and F promotes the formation of higher-valence Ni species in NiFe₂O₄. These high-valence Ni species are conducive to enhancing the intrinsic OER activity and facilitating the activation of the LOM^[53,54]. The k³-weighted Fourier-transformed extended X-ray absorption fine structure (FT-EXAFS) spectra of Mo/F-NiFe₂O₄ and pristine NiFe₂O₄ [Figure 2E] reveal not only the characteristic Ni-O and Ni-Ni coordination shells but also additional Ni-Mo and Ni-F scattering paths, further confirming the successful incorporation of Mo and F into the lattice structure^[54]. This structural modification suggests that under dual modulation of Mo and F, the Ni-O bond energy is reduced, facilitating bond cleavage. These observations are supported by wavelet transform (WT) analyses in Figure 2F and Supplementary Figure 11. Similarly, as shown in the Fe K-edge of Figure 2G, Mo/F-NiFe₂O₄ exhibits a higher oxidation state at the Fe K-edge compared to both NiFe₂O₄ and Fe foil, further indicating that the synergistic incorporation of Mo and F promotes the formation of higher-valence Fe species in NiFe₂O₄, which is conducive to activating the LOM^[55]. The FT-EXAFS spectra [Figure 2H] of Mo/F-NiFe₂O₄ and pristine NiFe₂O₄ reveal additional Fe-M and Fe-F coordination shells beyond the typical Fe-O and Fe-Fe bonds, providing direct evidence for the successful doping of Mo and F into the structure. Furthermore, the dual modulation by Mo and F weakens the Fe-O bond strength, thereby facilitating its cleavage during catalytic processes^[55]. These observations are consistent with the WT analyses presented in Figure 2I and Supplementary Figure 12.

Figure 2. Physical properties of Mo/F-NiFe₂O₄. (A) XRD Rietveld refinement patterns; (B) High-resolution O 1s XPS spectrum; (C) XPS valence band spectrum; (D) XANES spectra of Ni K-edge; (E) Ni R-space EXAFS spectra; (F) WT analysis of Mo/F-NiFe₂O₄; (G) XANES spectra of Fe K-edge; (H) Fe R-space EXAFS spectra; (I) WT analysis of Mo/F-NiFe₂O₄.

The electrocatalytic OER performance of the synthesized NiFe₂O₄-based catalysts and commercial RuO₂ was evaluated in 1.0 M KOH solution. As shown from the polarization curves in Figure 3A and Supplementary Figure 13, Mo/F-NiFe₂O₄ exhibits superior OER activity compared with all other catalysts synthesized under the same conditions and RuO₂. Specifically, Mo/F-NiFe₂O₄ delivers current densities of 50 and 100 mA cm^-2 at overpotentials of 245 and 311 mV, which are significantly lower than those of Mo-NiFe₂O₄, F-NiFe₂O₄, NiFe₂O₄, and commercial RuO₂ [Figure 3B]. The Tafel slope was further employed to assess the electron transfer kinetics during the OER, in which Mo/F-NiFe₂O₄ exhibits the lowest Tafel slope (49.68 mV dec^-1) among the tested catalysts, compared to Mo-NiFe₂O₄ (53.78 mV dec^-1), F-NiFe₂O₄ (67.24 mV dec^-1), and NiFe₂O₄ (71.15 mV dec^-1), indicating more efficient OER kinetics under alkaline conditions [Figure 3C]. The properties at the catalytic interface were studied via evaluation of the electrochemical active surface area (ECSA) and normalized kinetic current (j_K^norm) since the ECSA is linearly proportional to the double-layer capacitance (C_dl) at non-Faradic potential range. Cyclic voltammetry (CV) curves with a non-Faradic potential range of 0.5 V-0.6 V at varying scan rates [Supplementary Figure 14]. As illustrated in Figure 3D, Mo/F-NiFe₂O₄ exhibits a C_dl value of 2.11 mF cm^-2, which is higher than that of NiFe₂O₄ (1.46 mF cm^-2) and NF (1.14 mF cm^-2), but much lower than those of Mo-NiFe₂O₄ (18.20 mF cm^-2) and F-NiFe₂O₄ (5.45 mF cm^-2). This indicates that Mo/F co-doping does not confer a substantial advantage in terms of ECSA. The higher C_dl values observed for Mo-NiFe₂O₄ and F-NiFe₂O₄ can be attributed to their comparatively rougher nanoparticle morphologies, as revealed by the SEM images in Figure 1. However, when the kinetic current density (j) is normalized by the C_dl value at a fixed potential of 1.48 V, Mo/F-NiFe₂O₄ displays the highest j_K^norm value (24.31 mA mF^-1), outperforming Mo-NiFe₂O₄ (1.75 mA mF^-1), F-NiFe₂O₄ (5.12 mA mF^-1), NiFe₂O₄ (15.54 mA mF^-1) and NF (0.12 mA mF^-1). This result clearly demonstrates that the superior intrinsic OER activity of Mo/F-NiFe₂O₄ stems from its optimally engineered electronic structure with appropriate ECSA. To elucidate the kinetic differences, EIS was employed to evaluate the charge transfer kinetics during the OER [Figure 3E]. As shown in Supplementary Figure 15, the equivalent circuit model was used to deconvolute the two distinct electrochemical processes occurring at the electrode/electrolyte interface. Among all synthesized catalysts, Mo/F-NiFe₂O₄ exhibits the lowest charge transfer resistance (R_ct) of 3.1 Ω, which is lower than those of Mo-NiFe₂O₄ (3.6 Ω), F-NiFe₂O₄ (3.7 Ω), and pristine NiFe₂O₄ (3.8 Ω), indicating superior charge transfer efficiency and enhanced OER kinetics. In addition to activity, stability represents another critical parameter for assessing OER performance. The long-term durability of Mo/F-NiFe₂O₄ was evaluated through a chronopotentiometric test (V-t curve) at a high current density of 100 mA cm^-2 [Figure 3F]. The catalyst exhibited stable OER performance over 130 h of continuous operation, underscoring its exceptional durability. Notably, the OER performance of Mo/F-NiFe₂O₄ outperforms that of most recently reported alkaline OER catalysts in the literature [Figure 3G and Supplementary Table 4].

Figure 3. OER catalytic performance of Mo/F-NiFe₂O₄. (A) Polarization curves; (B) Histogram of overpotentials at current densities of 50.0 and 100.0 mA cm^-2; (C) Tafel plots; (D) Dot-line plots of C_dl (blue), and j_K^norm (red) at potential of 1.48 V; (E) Nyquist plots; (F) Time-dependent potential curves at a constant current density of 100 mA cm^-2 for over 130 h; (G) A comprehensive comparison of with recently reported other catalysts.

To verify whether Mo/F-NiFe₂O₄ activates the LOM during the OER, DEMS measurements were performed in real-time on both Mo/F-NiFe₂O₄ and pristine NiFe₂O₄ [Figure 4A and B]. Using 99.99% H₂¹⁸O as the solvent in a 1 M KOH electrolyte, a distinct signal was detected at m/z = 34 (¹⁶O¹⁸O) for Mo/F-NiFe₂O₄ during OER, whereas the signal at m/z = 36 (¹⁸O₂) was negligible. In contrast, pristine NiFe₂O₄ showed no detectable signal at m/z = 34 but exhibited a prominent peak at m/z = 36 (¹⁸O₂), indicating exclusive involvement of labeled oxygen from the electrolyte. These results clearly demonstrate that the synergistic modulation of Mo and F promotes the activation of the LOM in NiFe₂O₄ spinel, thereby significantly enhancing its intrinsic catalytic activity. To observe the structural evolution after the stability test, Raman spectroscopy was conducted before and after OER stability test. As clearly shown in Figure 4C, the pre-reaction sample displayed distinct characteristic peaks at approximately 621 and 1,154 cm^-1, which corresponded to NiFe₂O₄. In sharp contrast, the post-reaction catalyst presented a double peak around 551 and 678 cm^-1. Previous research findings indicate that these peaks correspond to the characteristic peaks of NiFeOOH, which strongly suggests a certain degree of surface reconstruction at the interface during the reaction^[26]. To further confirm the occurrence of surface reconstruction, XPS characterization was carried out on Mo/F-NiFe₂O₄ before and after the OER stability test. As depicted in Figure 4D and E, the relative content of oxygen vacancy species in the post-reconstruction sample increased from 50.07% to 61.21%, whereas the content of lattice oxygen species decreased from 26.01% to 6.77%. This firmly confirms the occurrence of surface reconstruction and further validates the LOM pathway. Furthermore, to gain deeper insight into the kinetic behavior of the catalytic process, in situ EIS was conducted [Figure 4F-I]. Analysis of the relaxation time and impedance magnitude reveals that Mo/F-NiFe₂O₄ exhibits the lowest R_ct and the fastest reaction kinetics among all catalysts, outperforming Mo-NFe₂O₄, F-NiFe₂O₄ and pristine NiFe₂O₄. These findings confirm that the superior intrinsic activity and accelerated reaction kinetics of Mo/F-iFe₂O₄ are directly attributed to the activation of the LOM.

Figure 4. Verification of the LOM pathway and interface changes of Mo/F-NiFe₂O₄. DEMS of (A) Mo/F-NiFe₂O₄ and (B) NiFe₂O₄; (C) Raman spectra ; (D) High-resolution XPS O 1s spectra; (E) Relative content of the different oxygen species of the Mo/F-NiFe₂O₄ before and after stability test; In-situ impedance testing diagram with corresponding relaxation time distribution diagram of (F) NiFe₂O₄, (G) Mo-NiFe₂O₄, (H) F-NiFe₂O₄ and (I) Mo/F-NiFe₂O₄.

A two-electrode configuration was employed to evaluate the potential of the electrocatalyst for overall water splitting, in which the synthesized catalyst served as anode and commercial Pt/C as cathode. In this configuration, water oxidation occurs at the anode, where OH^- ions are converted into O₂, while hydrogen evolution takes place at the cathode, producing H₂. Electrochemical measurements were performed in both 1.0 M KOH and alkaline seawater electrolytes. As shown in Figure 5A, under 1.0 M KOH conditions, the polarization curves of the optimized Mo/F-NiFe₂O₄ and commercial RuO₂ were performed for comparison. Specifically, the overall water-splitting voltages for Mo/F-NiFe₂O₄ were 1.54 and 1.59 V at current densities of 50 and 100 mA cm^-2, respectively, lower than those of commercial RuO₂ (1.60 and 1.66 V). Meanwhile, as illustrated in Figure 5B, in alkaline seawater, Mo/F-NiFe₂O₄ achieved a current density of 100 mA cm^-2 at only 1.65 V, highlighting its superior intrinsic catalytic activity. Considering working stability is another key criterion for assessing catalytic performance, the stability test at both alkaline and alkaline seawater electrolytes was performed. As shown in Figure 5C, the chronoamperometric test in 1.0 M KOH revealed that Mo/F Mo/F-NiFe₂O₄ exhibited negligible voltage fluctuations over 300 h of continuous operation. Similarly, Figure 5D demonstrates that under alkaline seawater conditions, Mo/F-NiFe₂O₄ maintained exceptional stability for more than 300 h at a high current density of 100 mA cm^-2, demonstrating excellent adaptability to complex environments and practical potential for industrial production. Additionally, to further evaluate the catalyst’s resistance to chlorine-related side reactions and OER-related Faraday efficiency (FE), iodination colorimetric reaction was performed. As shown in Supplementary Figure 16, Mo/F-NiFe₂O₄ exhibited negligible color change after exposure to the post-reaction electrolyte solution. In contrast, RuO₂ turned pale yellow after the stability test. This indicates that Mo/F-NiFe₂O₄ possesses excellent performance in suppressing competing chlorine (ClO^-) side reactions. The oxygen FE of Mo/F-NiFe₂O₄ was calculated using the displacement gas collection method [Supplementary Equation 2]. The pre-reaction FE [Supplementary Figure 17] was 96.6%, while the post-reaction efficiency was 93.1%, showing no significant decrease and demonstrating the outstanding OER stability of Mo/F-NiFe₂O₄.

Figure 5. Overall water splitting performance of Mo/F-NiFe₂O₄. (A) Polarization curves in 1.0 M KOH; (B) Polarization curves in 1.0 M KOH + Seawater; (C) Illustration of chronopotentiometric stability test in 1.0 M KOH. (D) Illustration of chronopotentiometric stability test in 1.0 M KOH + Seawater.

CONCLUSIONS

In summary, Mo/F-NiFe₂O₄ was successfully constructed for the OER in alkaline media. Mo/F co-doping modulated the electronic structure, increased Ni/Fe oxidation states, and shifted the VBM upward, inducing electron redistribution and weakening metal-oxygen bonds, thereby facilitating M-O bond formation and cleavage during catalysis. Furthermore, Mo/F-NiFe₂O₄ catalyst effectively activated LOM, leading to accelerated reaction kinetics and enhanced intrinsic activity. Owing to these synergistic effects, the optimized Mo/F-NiFe₂O₄ catalyst delivered a low overpotential of 247 mV at 50 mA cm^-2, and outstanding long-term stability over 130 h at 100 mA cm^-2, surpassing both pristine NiFe₂O₄ and single-doped counterparts. Practical water-splitting in both alkaline and alkaline seawater demonstrated improved activity and stability. This work shows case insights into how dual anion-cation co-dopants in regulating lattice oxygen participation in the OER, offering valuable guidance for practical water electrolysis applications.