Mn‐Induced Support Stabilization and Ir Electronic Activation Enable Acid‐Stable, Low‐Loading IrO₂ Water Oxidation

ABSTRACT

The practical application of low‐iridium catalysts for the acidic oxygen evolution reaction (OER) is primarily constrained by the intertwined issues of inadequate activity and stability. Incorporation of Mn into such low‐iridium catalysts is effective, yet the underlying mechanism remains unclear. This study addresses the mechanistic role of Mn doping in enhancing the activity and stability of low‐loading IrO₂/Co₃O₄ catalysts. By incorporating Mn³⁺ into the octahedral sites of Co₃O₄, Mn induces strong Mn─O covalency that reinforces the spinel lattice and stabilizes ultra‐low‐loading IrO₂ nanoparticles, delivering a 51 mV reduction in overpotential and a six‐fold enhancement in operational stability compared to its undoped counterpart at a current density of 10 mA cm⁻². In situ spectroscopic analyses and theoretical calculations decipher the dual role of Mn: it reinforces lattice integrity through strong covalent Mn─O bonds, suppressing ion leaching, while concurrently activating the Ir sites via interfacial Mn─O─Ir electron transfer, which optimizes intermediate adsorption and promotes the efficient oxide‐path mechanism (OPM). This work demonstrates that the targeted dual‐regulation of support chemistry establishes a general principle for designing high‐performance, low‐loading IrO₂ catalysts for acidic OER.

This study introduces a dual‐regulation strategy utilizing a Mn‐doped Co₃O₄ support to enhance the activity and stability of low‐loading IrO₂ catalysts for acidic water oxidation. Mn induces strong Mn─O covalency to reinforce the spinel lattice and activates Ir sites via interfacial electron transfer. This approach provides a general principle for designing high‐performance, acid‐stable electrocatalysts.

1. Introduction

The electrochemical oxygen evolution reaction (OER) under acidic conditions is a critical process for proton exchange membrane (PEM) water electrolysis, a promising technology for sustainable hydrogen production [1, 2, 3, 4, 5, 6]. However, the harsh acidic environment and sluggish OER kinetics impose a dual challenge: electrocatalysts must deliver both high intrinsic activity and long‐term structural stability to withstand corrosive conditions [7, 8, 9, 10, 11, 12]. Ir‐based materials currently set the benchmark catalysts for acidic OER, but suffer from their scarcity, high cost, and limited long‐term durability [7, 13, 14, 15]. Consequently, strategies that minimize Ir loading while enhancing performance via synergistic interactions with non‐noble supports have gained prominence [16, 17, 18, 19, 20, 21]. However, in the study of supported catalysts, research attention has heavily prioritized the chemical state of IrO₂, including its valence and coordination environment, allocating substantial characterization and computational resources to these aspects [22, 23, 24]. By comparison, the properties of the support and the underlying stabilization mechanisms have been far less explored. In fact, targeted investigation into support chemistry through rational design of the composition and properties of non‐noble metal oxide supports offers a promising route to achieve highly active and stable acidic OER electrocatalysts under low iridium loading. Nevertheless, systematic studies in this regard remain largely insufficient.

Manganese (Mn) has emerged as a uniquely advantageous element in acidic OER catalysis. Mn‐based oxides are not only among the most promising non‐precious catalysts themselves, but also serve as effective supports for IrO₂ and RuO₂, significantly enhancing their catalytic performance [25, 26, 27, 28, 29]. Moreover, even if a small amount of Mn is employed as a dopant in support, the OER activity and stability in acidic media can be improved significantly [30, 31, 32, 33]. For instance, Mn‐doped Co₃O₄ spinel structures have demonstrated improved corrosion resistance and electronic conductivity compared to pristine Co₃O₄, enabling better dispersion and stabilization of Ir species [34]. Besides, our previous study showed that Mn dopant in Co₃O₄ altered the electronic structure and reduced the leaching of RuO₂, leading to better activity and long‐term stability [24]. Despite these empirical advances, the atomic‐scale mechanisms underlying Mn's dual structural and electronic roles remain unresolved.

In this work, we present a case study in advanced support chemistry by introducing Mn into the octahedral sites of a Co₃O₄ support for IrO₂/Co₃O₄ catalysts under acidic OER conditions. The resulting IrO₂/Mn‐Co₃O₄ catalyst achieves a lower overpotential of 269 mV at a current density of 10 mA cm⁻², which is 51 mV lower than that of the undoped counterpart, and demonstrates more than six‐fold operational stability. Through a combination of spectroscopic and theoretical methods, we elucidate a mechanism of Mn‐Induced Dual‐Regulation of Support Chemistry: Mn not only reinforces the structural integrity of the support via strong covalent Mn─O bonds, suppressing the leaching of Co and Ir, but also enables interfacial electron transfer through Mn─O─Ir bridges, which raises the Ir d‐band center and enhances the adsorption of *OH intermediates. This dual modulation facilitates the oxide‐path mechanism and promotes the formation of O─O bonds. Our findings reveal how atomic‐level doping can strategically tailor support properties to achieve highly active and stable low‐iridium acidic OER electrocatalysts.

2. Results & Discussions

2.1. Engineering Support Chemistry: Atomic Incorporation of Mn³⁺ and IrO₂ Stabilization

To explore the influence of Mn doping on the spinel support, IrO₂/Mn‐Co₃O₄ supported on carbon cloth (CC) was synthesized using ZIF‐67 as the precursor. The synthesis process involved sequential immersion in KMnO₄ and IrCl₃ aqueous solutions to incorporate Mn and Ir elements, followed by calcination at 300 °C in a muffle furnace (Figure 1a). Inductively coupled plasma optical emission spectrometry analysis (ICP‐OES) reveals that the contents of Ir and Mn were 14.5% and 2.9% (Table S1), respectively. For comparison, the control sample of IrO₂/Co₃O₄ with similar Ir content (15.2%) was prepared under analogous conditions without KMnO₄.

FIGURE 1.

Synthesis and characterization of the catalysts. (a) Schematic of the fabrication process for IrO₂/Mn‐Co₃O₄. (b) XRD patterns of the samples. (c) SEM images of IrO₂/Mn‐Co₃O₄. (d) AC‐HAADF‐STEM and EDS mapping of IrO₂/Mn‐Co₃O₄. (e) HAADF‐STEM of IrO₂/Mn‐Co₃O₄ image taken along the [111] zone axis, accompanied by the corresponding structural model diagram and 3D surface electron density map. (f) HAADF‐STEM image of IrO₂/Mn‐Co₃O₄, where the Ir atom is highlighted within the dotted red circle in the yellow and green square areas, followed by a 3D surface electron density map of the selected yellow and green square areas. Representative images are shown from at least three independent observations (n = 3).

Besides for the peaks for CC, the X‐ray diffraction (XRD, Figure 1b) patterns of all samples primarily exhibit diffraction peaks consistent with the spinel Co₃O₄ phase (PDF# 42–1467), while the patterns for IrO₂ is absent due to the small size of the Ir species. Scanning electron microscopy (SEM) shows IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄ generally inherit the nanosheet morphology of ZIF‐67, indicating that the synthetic steps for introducing Mn or Ir do not affect the morphology (Figure 1c and Figures S1–S3).

The fine structure of IrO₂/Mn‐Co₃O₄ is first investigated by using aberration‐corrected high‐angle annular dark‐field scanning transmission electron microscopy (AC‐HAADF‐STEM) and corresponding high‐resolution energy‐dispersive spectroscopy (EDS) mapping (Figure 1d), which indicates that Ir, Co, and Mn disperse homogeneously at the scale of 10 nm. Upon closer observation, the HAADF‐STEM image taken along the [111] zone axis clearly reveals that the Mn‐Co₃O₄ possesses a highly ordered crystalline lattice (Figure 1e). The corresponding structural model diagram demonstrates the orderly arrangement of Co, Mn, and O atoms, represented by pink, blue, and green dots, respectively. Mn atoms clearly substitute partial Co sites within the lattice, confirming successful incorporation of Mn into the Co₃O₄ spinel structure. 3D surface electron density maps, derived from the HAADF‐STEM images, visualize atomic distribution, where the bright peaks indicate positions of atoms with heights proportional to atomic number or density. The obtained lattice image matches closely with the theoretical simulation of Mn‐doped spinel (Mn‐Co₃O₄), providing strong evidence for uniform Mn substitution into Co lattice sites.

Further magnified HAADF‐STEM images distinctly illustrate the structure of Ir species in IrO₂/Mn‐Co₃O₄ (Figure 1f). Localized bright spots highlighted by red circles, accompanied by corresponding 3D electron density maps, indicate enhanced Z‐contrast regions consistent with ultra‐small IrO₂ clusters (<2 nm) and several Ir single atoms anchored on the Mn‐Co₃O₄ support. These bright spots manifest as sharp peaks distributed uniformly, confirming that Ir exists predominantly as a single atom or extremely small cluster rather than aggregated nanoparticles, which is also supported by the HAADF‐STEM images taken in a different area (Figure S4). Moreover, HRTEM and corresponding EDS mapping of IrO₂/Co₃O₄ (Figure S5) indicate that the IrO₂ in the control group also exists as small clusters. The similar fine structure of IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄ lays the foundation for understanding the effect of Mn.

Raman spectroscopy further elucidates the atomic‐scale structural perturbations induced by Mn and IrO₂. In the Raman spectra (Figure 2a), the peaks observed at 188, 468, 519, 603, and 670 cm⁻¹ are assigned to the F_2g(1), E_g, F_2g(2), F_2g(3), and A_1g vibration modes of pristine Co₃O₄ [35, 36, 37]. Upon Mn³⁺ doping, Jahn‐Teller distortion enhances the Co─O bond in the octahedral sites, blue‐shifting the A_1g mode, whereas the F_2g(1) peak, dominated by tetrahedral sites vibrations, remains unchanged [38, 39]. This selective shift indicates that Mn³⁺ selectively substitutes for Co³⁺ (octahedral sites) in the octahedral sites rather than for Co²⁺ (tetrahedral sites) in the tetrahedral ones. Conversely, IrO₂ decoration introduces surface tensile stress, elongating Co─O bonds and reducing vibrational frequencies, thereby shifting the A_1g peak toward a red shift. The coexistence of Mn doping and IrO₂ decoration creates a competitive modulation of the local bonding environment. When these opposing effects reach equilibrium, the net shift in Raman peak positions becomes negligible, as observed experimentally.

FIGURE 2.

(a) Raman spectroscopy measurements for IrO₂/Mn‐Co₃O₄, IrO₂/Co₃O₄, Mn‐Co₃O_4, and Co₃O₄. (b) Co K‐edge XANES of IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄ with other reference samples. (c) Ir L₃‐edge XANES of IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄ with other reference samples. (d) Fourier transformed magnitudes of Co K‐edge EXAFS spectra for IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄ with other reference samples. (e) Fourier transformed magnitudes of Ir L₃‐edge EXAFS spectra for IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄ with other reference samples. The inset zooms in on the white line feature. IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄ for XPS spectra: (f) Co 2p (g) Ir 4f (h) Mn 2p. Representative spectra are shown from three independent samples (n = 3).

The electronic structure of Ir and Co in IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄ was analyzed through X‐ray absorption fine structure spectra (XAFS). The rising edge of the Co K‐edge X‐ray absorption near‐edge structure (XANES) spectra exhibits a slight negative energy shift of IrO₂/Mn‐Co₃O₄ compared to IrO₂/Co₃O₄ (Figure 2b), indicative of a lower average oxidation state of Co atoms after Mn doping. This phenomenon can be primarily attributed to the partial substitution of octahedral Co³⁺ by Mn³⁺ ion in the octahedral sites of spinel Co₃O₄, consistent with previous reports [33, 34]. The Ir L₃‐edge absorption maxima of both IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄ shifted toward higher energy values, closely aligning with the characteristic position of IrO₂ (Figure 2c). This energy shift confirms the predominant +4 oxidation state of Ir species in both catalysts [40]. Notably, the slightly reduced intensity of the L₃‐edge white line of IrO₂/Mn‐Co₃O₄ relative to IrO₂/Co₃O₄ suggests a stronger electron transfer via interfacial Ir─O─Mn compared to Ir─O─Co [34].

To further elucidate the atomic coordination environments of Ir and Co, Fourier‐transformed k³‐weighted extended X‐ray absorption fine structure (FT‐EXAFS) analyses were conducted. The Co K‐edge FT‐EXAFS spectra (Figure 2d) of both IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄ reveals that Co─O and Co─Co/Mn scattering paths at 1.9 Å and 2.8 Å, respectively. Quantitative EXAFS fitting reveals that the Co─O coordination number (CN) in IrO₂/Co₃O₄ is 5.3, with an average bond length of 1.93 Å. Upon Mn incorporation, the Co─O CN decreases to 4.9 (Table S2), while the bond length remains constant at 1.93 Å. Additionally, the Co─Co/Mn CN decreases from 8.4 to 7.3, implying that Mn doping does not significantly alter the bond length but decreases the coordination number, exposing more active sites and thereby enhancing the catalytic activity. At the Ir L_3‐edge, FT‐EXAFS analysis revealed dominant Ir─O coordination at 1.5 Å for both IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄, closely resembling the characteristic IrO₂ configuration (Figure 2e). Moreover, the Ir‐Ir coordination at ∼2.7 Å confirms that Ir mainly exists as IrO₂ clusters, consistent with the previous HADDF‐STEM observation. The conclusion drawn by FT‐EXAFS can also be supported by the wavelet transform EXAFS (WT‐EXAFS) analysis in Figure S6.

X‐ray photoelectron spectroscopy (XPS) was employed to probe the chemical states and electronic interactions of all surface elements. The survey spectra of both samples indicate the purity in terms of elements (Figure S7). The Co 2p spectrum (Figure 2f) exhibits two pairs of peaks: the dominant peaks at 780.3 and 795.4 eV are assigned to Co^{3 +}, while the minor peaks at 781.9 and 797.2 eV correspond to Co^{2 +} [41, 42]. After Mn incorporation, all Co 2p peaks shift toward lower binding energies, accompanied by a marked reduction in the Co^{3 +}/Co^{2 +} ratio to 39.1% from 47.4%. This confirms a lower average Co oxidation state, in agreement with the XANES results. Furthermore, Mn 2p spectra (Figure 2h) further reveal the co‐existence of Mn^{3 +} and Mn^{4 +} species whose binding energy locates at 641.6 and 643.2 eV, respectively [32]. The higher electronegativity of Mn induces a strong electron‐withdrawing effect, redistributing electron density toward Mn─O bonds and weakening adjacent Co─O interactions, which is further confirmed by the negative shift of lattice O species in the O 2p spectrum (Figure S8). Finally, Ir 4f spectra (Figure 2g) of IrO₂/Mn‐Co₃O₄ show characteristic peaks at 60.4 and 63.4 eV, corresponding to Ir⁴⁺ (4f 7/2 and 4f 5/2) [43]. Compared to IrO₂/Co₃O₄, the peaks shift toward lower binding energies, indicating an increased electron density around the Ir atoms induced by Mn doping. The electron‐rich IrO₂ may facilitate the adsorption and desorption of reactive intermediates during the catalytic cycle, ultimately enhancing the OER activity.

Furthermore, the electronic structure of IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄ is investigated theoretically by analyzing the charge density differences between IrO₂ and Co₃O₄ in the two systems. It is evident that Mn doping enhances the electron transfer between IrO₂ and the support (Figure S9). To provide a more specific and quantitative comparison, we calculated the Bader charges (Table S3). The results indicate that the electron count associated with Ir increases significantly after Mn doping, which aligns with the findings from XPS and XAS analyses. With the incorporation of Mn, the electron transfer from the support to the IrO₂ is further enhanced via interfacial Ir‐O‐Mn. This leads to a reduction in the oxidation state of Ir, increasing its oxygenophilicity and consequently enhancing the adsorption of oxygen‐containing intermediates.

Based on HADDF‐STEM, Raman, XAFS, XPS, and DFT calculations, it can be concluded that the incorporation of Mn not only reduces the average state of Co in Co₃O₄ supports by substituting the octahedral Co³⁺, but also endows IrO₂ clusters with a slightly higher valence state and electron‐deficient structure.

2.2. Dual‐Enhanced OER Performance of IrO₂/Mn‐Co₃O₄ Through Modified Support Chemistry

The OER performance of IrO₂/Mn‐Co₃O₄, IrO₂/Co₃O₄, bare spinel oxides, and commercial IrO₂ (Com. IrO₂) was tested by using a typical three‐electrode system in 0.5 m H₂SO₄ electrolyte. Before testing, the reference electrode has been calibrated experimentally according to literature (Figure S10) [44]. Of note is that key synthesis parameters influencing OER performance, such as the KMnO₄ concentration during Mn doping, were systematically optimized prior to comparative evaluation of the IrO₂/Mn‐Co₃O₄ catalysts (Figure S11). As shown in the linear sweep voltammetry (LSV) curves in Figure 3a, the IrO₂/Mn‐Co₃O₄ catalyst exhibits the best catalytic activity, achieving a low overpotential of only 269 mV at a current density of 10 mA/cm². This performance significantly surpasses those of IrO₂/Co₃O₄ (320 mV), Mn‐Co₃O₄ (459 mV), Co₃O₄ (521 mV), and Com. IrO₂ (357 mV). Meanwhile, IrO₂/Mn‐Co₃O₄ presents the lowest Tafel slope of 70.1 mV/dec among all catalysts, which is smaller than IrO₂/Co₃O₄ (75.2 mV/dec), Mn‐Co₃O₄ (109.1 mV/dec), Co₃O₄ (113.5 mV/dec), and Com. IrO₂ (83.3 mV/dec), suggesting that the simultaneous incorporation of Mn dopants and IrO₂ nanoparticles significantly accelerates OER kinetics by enhancing charge‐transfer efficiency and optimizing active sites. (Figure 3b).

FIGURE 3.

Electrocatalytic OER performance of the as‐made electrocatalysts in 0.5 m H₂SO₄. (a) LSV polarization curves (85% iR‐corrected) of the prepared samples. (b) Tafel plots for the corresponding samples. (c) Plots showing the extraction of the C _dl. (d) Performance comparison of IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄. (e) The chronopotentiometry curves of IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄ at a current density of 10 mA/cm². (f) Schematic of the PEMWE cell. (g) Polarization curves of the PEMWE device with IrO₂/Mn‐Co₃O₄ (Ir loading: ∼0.5 mg_Ir cm⁻²) as anode catalysts at 60°C in ultrapure water. (h) Chronopotentiometry curves of the PEMWE device with IrO₂/Mn‐Co₃O₄ as anode catalysts operated at 1000 mA cm⁻² at 60°C in ultrapure water. Representative curves are shown from three independent experiments (n = 3).

To further evaluate the charge transfer resistance (R _ct), electrochemical impedance spectroscopy (EIS) is carried out under the same potential. Compared with IrO₂/Co₃O₄ (4.94 Ω), Mn‐Co₃O₄ (66.88 Ω), Co₃O₄ (67.03 Ω), and commercial IrO₂ (8.07 Ω), IrO₂/Mn‐Co₃O₄ exhibits the lowest charge transfer resistance, which is only 2.12 Ω, highlighting the essential roles of Mn doping and IrO₂ decoration in promoting electrical conductivity and facilitating electron transport. (Figures S12 and S13). To better understand the intrinsic activity enhancement, electrochemical double‐layer capacitance (C _dl), correlated to the electrochemically active surface area (ECSA), was measured through cyclic voltammetry (CV) test (Figure S14). As depicted in Figure 3c, IrO₂/Mn‐Co₃O₄ exhibits the highest C _dl of 15.80 mF/cm², substantially larger than those of IrO₂/Co₃O₄ (13.20 mF cm⁻²), Mn/Co₃O₄ (3.94 mF/cm²), Co₃O₄ (2.95 mF/cm²), and commercial IrO₂ (7.06 mF/cm²). After ECSA normalization, IrO₂/Mn‐Co₃O₄ retains superior intrinsic catalytic activity (Figure S15). Moreover, the mass activity normalized by Ir loading further verifies the superior catalytic efficiency of IrO₂/Mn‐Co₃O₄ (106.1 A g_Ir ⁻¹ at 1.53 V), higher than IrO₂/Co₃O₄ (92.1 A g_Ir ⁻¹).

It is essential to assess stability and long‐term durability in the practical application in water splitting [45, 46]. The overpotential of IrO₂/Mn‐Co₃O₄ only reduces 30 mV at 10 mA·cm⁻² after 1000 CV cycles (Figure S16). Notably, the long‐term durability is greatly improved from 15 h IrO₂/Co₃O₄ to about 100 h of IrO₂/Mn‐Co₃O₄, suggesting that a small amount of Mn in support can greatly affect the stability (Figure 3e). This is critically important for OER in acidic media. Quantitative ICP‐OES analysis of the electrolyte after the 100 h durability test reveals elemental retention rates of 55.8% for Co, 85.3% for Mn, and 82.7% for Ir (Table S4), further confirming the structural robustness of the Mn‐doped composite. To systematically evaluate the synergistic optimization mechanism of Mn doping on catalytic performance, Figure 3d comprehensively compares the electrochemical metrics of IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄. IrO₂/Mn‐Co₃O₄ demonstrates superior performance across all critical parameters. This comparative analysis conclusively demonstrates the pronounced “dual‐functional” effect of Mn doping, which not only enhances the intrinsic activity of the catalyst but also effectively mitigates the corrosion of active species by acidic media, thereby reinforcing its long‐term stability under harsh acidic conditions.

Building upon the exceptional catalytic performance and stability of IrO₂/Mn‐Co₃O₄ in three‐electrode systems, we further evaluated its practical viability by constructing a proton exchange membrane water electrolyzer (PEMWE). The PEMWE employed ultrapure water feed and Pt/C and IrO₂/Mn‐Co₃O₄ as cathode/anode catalysts (Figure 3f), respectively. As shown in Figure 3g, the current‐voltage polarization curves reveal that the IrO₂/Mn‐Co₃O₄‐based PEMWE achieves a cell voltage of 1.75 V at 1 A cm⁻² under 60°C operating conditions. Notably, these PEMWE measurements were conducted as a proof‐of‐concept device evaluation, without device‐level optimization of Ir loading. Despite this, the system demonstrates remarkable operational stability, maintaining continuous operation for over 400 h at 1000 mA cm⁻² (Figure 3h) with a minimal voltage degradation rate of approximately 0.07 mV h⁻¹. Comparative analysis with previously reported PEMWE catalysts (Table S5) confirms that the IrO₂/Mn‐Co₃O₄ composite exhibits superior overall performance metrics, combining both high efficiency and exceptional durability in practical electrolyzer applications.

2.3. Unraveling the Mn‐Mediated Stabilization Mechanism Toward OER

In order to further elucidate the mechanism by which Mn doping enhances the stability of the catalyst, we employed ICP‐OES to monitor the time‐evolution dissolution behavior of metal ions in solution during the acidic OER reaction for IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄ (Figure 4a,b). This suggests that Mn not only avoids causing lattice damage to the material but also functions as a structural stabilizer. In contrast, although Ir and Co in both samples dissolved fast in the first 10 h, after prolonged reaction times, the dissolved Co and Ir ions from IrO₂/Mn‐Co₃O₄ were significantly lower than those from the undoped IrO₂/Co₃O₄, which can also be supported by SEM and TEM of post‐reaction samples (Figures S17–S19). These results indicate that Mn doping effectively suppresses the dissolution of Co₃O₄ and IrO₂, thereby enhancing the structural and chemical stability of the catalyst in an acidic electrolytic environment.

FIGURE 4.

Dissolved ions concentrations measured for IrO₂/Mn‐Co₃O₄ (0.750 mg cm⁻²) and IrO₂/Co₃O₄ (0.725 mg cm⁻²) in 0.5 m H₂SO₄ electrolyte by ICP‐OES. (a) Co ions (b) Ir (left‐y‐axis) and Mn (right‐y‐axis). (c, d) in situ Raman spectroscopy measurements for IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄ recorded during acidic OER from 1.1 to 1.7 V versus RHE in 0.5 M H₂SO₄. (e) DOS plots of Co, Ir, O, and Mn states in IrO₂/ Mn‐Co₃O₄ and DOS plots of Co, Ir, and O states in IrO₂/Co₃O₄. Wave functions of the band‐edge states at the Γ point (isosurface level = 3e⁻⁹) (f) IrO₂/Co₃O₄ and (g) IrO₂/Mn‐Co₃O₄. 2D ELF contour plots for (h) IrO₂/ Co₃O₄ and (i) IrO₂/Mn‐Co₃O₄. The experimental data in (a‐d) represent results from three independent measurements (n = 3).

The in situ Raman spectroscopy data (Figure 4c,d) reveal distinct structural evolution trends for the IrO₂/Co₃O₄ and IrO₂/Mn‐Co₃O₄ catalysts under increasing applied potentials. For the undoped IrO₂/Co₃O₄, the Co─O vibrational peak near 690 cm⁻¹ exhibits a redshift (toward lower wavenumbers) with increasing potential. This redshift correlates with a weakening of Co─O bonds, likely driven by the oxidation of Co³⁺ to Co⁴⁺ under acidic OER conditions. Furthermore, the in situ Raman spectra revealed a progressive intensity enhancement of the Co─O vibrational mode at ∼590 cm⁻¹ under increasing potentials, which is attributed to the formation of high‐valent Co⁴⁺ species during OER [47, 48, 49]. This observation aligns with the structural instability of undoped Co₃O₄, as the partial oxidation of Co³⁺ to Co⁴⁺ disrupts the native Co₃O₄ spinel framework (e.g., Co³⁺ [Co²⁺Co³⁺] O₄ → Co⁴⁺O₂‐like configurations). The intensified ∼590 cm⁻¹ mode likely corresponds to symmetry‐breaking vibrations induced by Co⁴⁺‐O bond contraction and local lattice distortion, consistent with prior reports on metastable Co⁴⁺‐rich oxide phases in acidic OER [48, 50, 51]. The formation of higher‐valent Co⁴⁺ species elongates Co─O bonds due to reduced electron density and increased lattice oxygen participation in OER (e.g., *O → O₂), accelerating structural degradation. The enlarged graph (660–720 cm⁻¹ region) further confirms this trend, where the peak position progressively shifts leftward (∼4 cm⁻¹ at OCP to 1.7 V), accompanied by intensity attenuation, indicative of bond rupture and oxygen vacancy generation.

In contrast, the Mn‐doped IrO₂/Mn‐Co₃O₄ demonstrates a blueshift (toward higher wavenumbers) for the Co─O peak under identical conditions. The blueshift (∼4 cm^{− 1} at OCP to 1.7 V) signifies strengthening of Co─O bonds, attributed to Mn³⁺ doping‐induced electronic modulation [33]. Mn species likely stabilize the Co³⁺/Co⁴⁺ redox couple through charge compensation, suppressing oxygen vacancy formation. Additionally, the stronger covalent Mn─O bonds (versus Co─O) mitigate lattice oxygen loss by stabilizing the Co₃O₄ framework [32]. The retained peak intensity in the 660–720 cm⁻¹ region further supports enhanced structural integrity. The electrochemical stability improvement (100 h for Mn‐doped catalyst) aligns with these observations. The Mn‐induced Co─O bond strengthening and suppressed lattice oxygen participation collectively inhibit acid corrosion and phase transitions. The stability enhancement is further rationalized by the synergistic IrO₂/Mn‐Co₃O₄ interactions, which optimize charge transfer kinetics and reduce overpotential‐driven structural stress.

Multiple theoretical calculations have been employed to deeply understand the doping effect of Mn on OER stability. Notably, DFT models were constructed using the as‐prepared lattice/cluster motif as an initial framework, while the mechanistic conclusions are constrained by post‐OER characterization and operando spectroscopic evidence of the stabilized active state. As can be seen from the DOS (Density of States) plot (Figure 4e), the electronic states of Co are generally higher than those of Ir across various energy ranges due to the significantly larger number of Co atoms compared to Ir atoms in the model. However, near the Fermi level, the electronic states of Co exhibit a noticeable decrease, even falling below those of Ir. This indicates that the reaction activity of Ir is significantly higher than that of Co, establishing Ir as the primary active site for the catalytic reaction, while Co functions as a cocatalyst and support. Mn primarily introduces new unoccupied states in the DOS above the Fermi level. The d‐band center values further corroborate this; the d‐band center of Mn lies relatively close to the Fermi level compared to the others, directly demonstrating that Mn introduces a portion of new density of states within the unoccupied orbitals. This increase in conduction band electronic states enhances the adsorption of oxygen‐containing intermediates by the support, thereby improving the stability of the system during the acidic OER process.

Furthermore, the CBM (Conduction Band Minimum) of IrO₂/Co₃O₄ and IrO₂/Mn‐Co₃O₄ was visualized (Figure 4f,g) to elucidate the effect of Mn doping on the electronic states of the catalyst. Clearly, compared to pure Co₃O₄, a higher concentration of unoccupied orbitals is localized around Mn atoms, consistent with the findings from the DOS plots and d‐band center analysis. This demonstrates that Mn possesses significantly stronger oxygen affinity within this system, which facilitates lattice oxygen regeneration and thereby maintains the structural stability. The calculated oxygen vacancy formation energies (E_v) are presented in Figure S20. The E_v for the lattice oxygen adjacent to the Mn site increases significantly to 2.317 eV, compared to 1.110 eV for pristine Co₃O₄. This provides direct thermodynamic evidence that Mn doping effectively suppresses the leaching of lattice oxygen by increasing the energy barrier for vacancy formation. Concurrently, the enhanced oxygen affinity of Mn strengthens the Mn‐O‐Ir interaction, leading to a tighter coupling between the IrO₂ and the support. Consequently, Ir receives a greater number of electrons from the support.

Conversely, in IrO₂/Mn‐Co₃O₄, Mn (purple) substitutes for some Co positions, significantly enhancing electron localization around Mn. The ELF value between Mn and O is higher, and the electron cloud is more evenly distributed between Mn and O, unlike the Co─O bond, where electrons are primarily localized on the O atom. The Mn─O bond exhibits a stronger covalent character. Since covalent bonds are inherently more stable than ionic bonds, the enhanced covalency of Mn─O bonds contributes to greater overall lattice stability, reducing the likelihood of metal dissolution and structural collapse, leading to higher stability of IrO₂/Mn‐Co₃O₄.

The effect of Mn doping on the OER stability can be summarized as below: after doping with Mn, which exhibits a stronger oxygen affinity, the structural stability of the spinel support can be enhanced through the formation of Mn‐O bonds with greater covalent character. Additionally, the interaction between IrO₂ and the support is strengthened via the formation of Mn‐O‐Ir bonds at the interface. This effect further mitigates lattice expansion in the in situ OER catalyst, slows down the dissolution of Ir and Co species, and ultimately enhances the overall OER stability.

2.4. Electronic Modulation and Promotion of the Oxide‐Path Reaction Through Support Modulation

The incorporation of Mn in Co₃O₄ not only drastically enhances the catalyst's long‐term stability, but also reduces the required overpotential. Electrocatalytic OER on metal oxides proceeds via three main pathways: the conventional adsorbate evolution mechanism (AEM) involving *OH, *O, and *OOH intermediates; the lattice‐oxygen mediated mechanism (LOM), which bypasses OOH by engaging lattice oxygen but risks structural collapse in acid; and the oxide‐path mechanism (OPM), where adjacent M─O species directly couple to O─O without extracting lattice oxygen, thus preserving integrity (Figure 5a–c). Unlike LOM, which accelerates M─O bond cleavage and catalyst degradation in acidic media, OPM offers fast kinetics without compromising structure. To unambiguously assign the active pathway in IrO₂/Mn‐Co₃O₄, we performed pH‐dependent kinetics, in situ ATR‐FTIR, and operando DEMS with isotope labeling [27, 52, 53, 54, 55, 56].

FIGURE 5.

(a–c) Schematic illustration of simplified OER mechanisms: AEM, LOM, and OPM. (d‐e) LSV curves for IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄ in H₂SO₄ at different pH values. (f) In situ ATR‐FTIR spectra results of IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄. (g) DEMS signals of O₂ products for IrO₂/Mn‐Co₃O₄ in the electrolyte using H₂ ¹⁸O as the solvent during three times of LSV. (h, i). Calculated models of IrO₂/Mn‐Co₃O₄ catalyst and Mn‐Co₃O₄ structure. (j) The determined ΔG of RDS via the AEM, LOM, and OPM pathway in different models. (k) Gibbs free energy illustration by IrO₂/Mn‐Co₃O₄ and IrO₂/Co₃O₄ catalysts during the OER process by OPM pathways. The experimental data in (d–g) are representative of three independent experiments (n = 3).

First, pH‐dependent OER kinetics (Figure 5d,e and Figure S21) were systematically evaluated across a pH range of 0 to 1.5. The IrO₂/Mn‐Co₃O₄ catalyst exhibited a pronounced pH‐dependent characteristic, reflected by a larger ρ_RHE value (16.45) compared to IrO₂/Co₃O₄ (9.11), excluding the AEM mechanism for both catalysts [57]. To further validate the formation of oxygen‐related intermediates, operando synchrotron Fourier‐transform infrared (FTIR) spectroscopy was conducted under acidic OER conditions. As depicted in Figure 5f, an absorption peak in IrO₂/Mn‐Co₃O₄ at 1102 cm⁻¹ emerged distinctly at potentials above 1.6 V (Figure 5f), attributed to the superoxo‐like species (*O₂ ²⁻) [32, 58, 59]. Similarly, the undoped IrO₂/Co₃O₄ also exhibited a characteristic peak at the same wavenumber (Figure S22). The gradual increase in peak intensity with increasing potential indicates the progressive accumulation of reactive intermediates, which also support the LOM or OPM mechanism.

Operando differential electrochemical mass spectrometry (DEMS) with in situ isotope labeling was employed to elucidate the oxygen evolution mechanism on IrO₂/Mn‐Co₃O₄, using 0.5 m H₂SO₄ in H₂ ¹⁸O as the electrolyte. For IrO₂/Mn‐Co₃O₄, the detection of ³²O₂, ³⁴O₂, and ³⁶O₂ signatures is characteristic of the OPM (Figure 5g) [27]. Notably, the ³⁶O₂ signal intensity exceeds that of ³²O₂ and ³⁴O₂ by two and three orders of magnitude, respectively. This overwhelming dominance of ³⁶O₂ unequivocally indicates that the oxygen product primarily originates from two ¹⁸O atoms in the isotope‐labeled electrolyte, thereby categorically excluding the LOM.

To further validate the advantages of OPM and probe the origin of enhanced activity via Mn doping in support, we calculated the Gibbs free energies for IrO₂/Co₃O₄ (Figure 5h) and IrO₂/Mn‐Co₃O₄ (Figure 5i) under OPM, AEM, and LOM pathways. Figure 5j presents the Gibbs free energy changes for the rate‐determining steps of IrO₂/Mn‐Co₃O₄ across the three mechanisms. For OPM, the rate‐determining step is *OH+*OH → *O+*OH with an energy barrier of 1.51 eV, lower than those of AEM (1.65 eV, [*O → *OOH]) and LOM (1.79 eV, [*O+H → V_o+O₂]), confirming the superiority of OPM over LOM and AEM (Figures S23 and S24). Figure 5k further compares the Gibbs free energy changes of IrO₂/Co₃O₄ and IrO₂/Mn‐Co₃O₄. The Mn‐modified Ir sites exhibit stronger binding of oxygenated intermediates, ensuring that after OH deprotonation, the adsorbate remains coordinated as Ir─O. At optimal Ir─Ir separations, these adjacent Ir‐O species directly couple to form O‐O bonds via the oxide‐path mechanism. As a result, the energy barrier of RDS reduces from 2.75 to 1.51 eV, leading to the enhanced OER activity of the Mn‐doped sample. The doping effect of Mn on OER activity can be further explained by DOS. Compared with the d‐band center of Ir supported on pure Co₃O₄, the d‐band center of Ir on the Mn‐modified surface is significantly higher (Table S6). To further identify the primary active sites, the adsorption energies of *OH (ΔG_*OH) on Ir and Co sites were calculated (Figure S25). The Ir sites exhibit a significantly stronger adsorption affinity (ΔG_*OH = ‐0.989 eV) compared to the Co sites (ΔG_*OH = −0.112 eV). This significant energy difference confirms that the OER process preferentially initiates at the Ir centers, while the Mn‐modified support primarily serves to tune the electronic structure of the active site.

The effect of Mn doping on the OER activity can be summarized as below: Mn doping primarily enhances the OER activity of the catalyst through electronic modulation. The electron transfer effect within the Mn‐O‐Ir structure enriches the electron density of IrO₂, shifting the Ir d‐band center upward. This stronger *OH binding ensures that, upon deprotonation, the adsorbate remains as Ir─O, and at optimal Ir─Ir distances these adjacent Ir─O species directly couple (oxide‐path mechanism), thereby boosting overall OER activity (Figure 6). It should be emphasized that this exceptional durability is not dictated by a single isolated factor. Rather, the suppression of Co dissolution provides a robust macroscopic structural foundation, while the strengthened Mn‐O‐Ir interfacial interaction affords direct atomic‐level electronic protection against Ir over‐oxidation. These two factors act synergistically to deliver the 6‐fold improvement in the operational lifespan.

FIGURE 6.

The summarized dual roles of Mn dopants in Co₃O₄ for enhancing the activity and stability of low‐loading IrO₂/Co₃O₄ catalysts.

3. Summary

In summary, this work demonstrates that Mn doping provides an effective dual‐regulation strategy for tuning the structural and electronic chemistry of Co₃O₄ support, thereby overcoming the intertwined challenges of activity and stability in low‐Ir acidic OER catalysts. Compared to the undoped counterpart, IrO₂/Mn‐Co₃O₄ exhibits 6 times higher stability and a significantly reduced overpotential of 51 mV. Structural stabilization arises from the substitution of Mn³⁺ into octahedral Co sites, forming highly covalent Mn─O bonds that anchor the spinel lattice and suppress both Co/Ir dissolution and lattice collapse, as evidenced by in situ Raman, XAFS, XPS, and corroborated by DOS, CBM, and ELF analyses. Simultaneous electronic activation occurs via Mn─O─Ir electron donation, which shifts the Ir d‐band center upward, optimizing *OH adsorption so that, after deprotonation, Ir─O species remain intact and, at ideal Ir─Ir separations, undergo direct O─O coupling through the oxide‐path mechanism. These atomic‐scale insights reveal that Mn‐induced support stabilization and Ir electronic activation act synergistically to drive efficient and durable acidic OER, providing a clear design blueprint for the development of high‐performance, acid‐stable, and low‐iridium catalysts for sustainable hydrogen production. Beyond improving the long‐term electrocatalytic durability, the significantly suppressed Ir and Co dissolution via Mn doping also substantially lowers the potential ecotoxicological burden of the catalyst on aquatic environments [60].

Crucially, the Mn‐doped Co₃O₄ support functions as a universal and robust platform capable of stabilizing various active phases—from Ir/Ru clusters to dual‐atom pairs—by synergistically modulating interfacial electronic states and reinforcing the spinel lattice against acidic corrosion [15, 20, 33, 61, 62]. This strategy provides a generalizable blueprint for the rational design of high‐performance, acid‐stable electrocatalysts for industrial‐scale PEM water electrolysis.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: advs75667‐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.