Balancing Iridium Speciation in Ir/NiO Single‐Atom Catalysts for Overall Water Splitting Authors: Yongfang Zhou, Yu Mao, Cuizhu Ye, Zhanghao Ren, Shanghai Wei, Hui Yang, Yufei Zhao, Anton Tadich, Aaron Marshall, Geoffrey IN Waterhouse Published: 2026 ABSTRACT ABSTRACTElectrocatalysts capable of driving overall water splitting at high current densities and different pH remain elusive. Herein, we demonstrate that iridium single atoms supported on NiO nanosheets (Irx‐NiO) can serve as promising bifunctional OER/HER catalysts in alkaline, neutral, and acidic media. An Ir8%‐NiO catalyst required low overpotentials of 177, 275, and 174 mV for the oxygen evolution reaction (OER), and 29, 54, and 24 mV for the hydrogen evolution reaction (HER), at 10 mA cm−2 in 1.0 m KOH, 0.1 m PBS, and 0.5 m H2SO4, respectively. When applied as the anode and cathode catalysts in an anion exchange membrane water electrolyser (AEMWE), Ir8%‐NiO achieved 1.0 A cm−2 at 2.02 V with good operational stability. Spectroscopic and theoretical investigations revealed that the coexistence of embedded and surface‐adsorbed Ir atoms in the Ir8%‐NiO catalyst underpinned its bifunctional OER/HER activity. Embedded Ir promoted Ni2+ → Ni3+ oxidation and NiOOH formation during OER, while surface‐adsorbed Ir atoms provided sites for *H adsorption, enabling efficient HER. The dual Ir sites optimize *OH and *H binding energies, resulting in exceptional bifunctional activity. These findings highlight the importance of controlling Ir speciation and metal–support interactions to achieve overall water splitting across a wide pH range.Keywords: iridium single atom catalyst, NiO support, overall water splitting, surface and embedded Ir atoms, water electrolyser Introduction Hydrogen (H2), due to its high gravimetric energy density and diverse production sources (including water, biomass feedstocks, and municipal waste), has long been regarded as a next‐generation energy carrier to replace fossil fuels [1, 2, 3, 4, 5]. Presently, hydrogen is mainly produced through steam methane reforming or coal gasification, both of which are energy‐intensive processes that generate substantial CO2 emissions [6]. To support the transition toward a green hydrogen economy, alternative green hydrogen production technologies must be found. Among these, electrocatalytic water splitting powered by renewable electricity represents the most attractive route for the large‐scale production of high‐purity hydrogen with net‐zero carbon emissions [7]. Electrocatalytic water splitting involves the anodic 4‐electron oxygen evolution reaction (OER) and the 2‐electron hydrogen evolution reaction (HER), with overall water splitting kinetics typically limited by the more complex anodic OER [8, 9, 10, 11, 12]. Traditionally, IrO2 or RuO2 have been used by researchers for the OER (due to their modest overpotentials), whilst Pt/C has been the preferred HER catalyst. However, the low abundance and high cost of Ir, Ru and Pt are obstacles to practical applications in water splitting [13, 14, 15, 16]. Furthermore, RuO2‐based OER catalysts are unstable under acidic conditions, which limits their range of application for water electrolysis. As a result, considerable effort is presently being devoted to the discovery of more cost‐effective alternatives to these conventional precious metal‐based OER and HER catalysts, specifically electrocatalysts that can work across a broad pH range (e.g, 0.5 m H2SO4 through to 1 m KOH) [17, 18]. Especially attractive in this context are robust bifunctional electrocatalysts capable of efficiently driving both the OER and HER reactions with low overpotentials at industrially relevant current densities (1 A cm−2) [19, 20, 21, 22]. To date, few electrocatalysts have been reported that can satisfy all these requirements. With the aim of developing lower‐cost electrocatalysts for the OER and HER, researchers have explored a diverse range of materials and synthetic strategies. For the OER, nickel oxide (NiO) and nickel‐iron layered double hydroxide (NiFe‐LDH) materials stand out for their low‐cost, exceptional activity and promising corrosion resistance during the OER in alkaline media [23, 24, 25]. These oxides reconstruct to form active NiOOH and Ni(Fe)OOH phases, respectively, under the highly polarizing conditions of the OER. While these oxyhydroxide phases exhibit excellent OER activity in alkaline media, their electronic and structural properties are not favorable for hydrogen adsorption, resulting in poor alkaline HER activity when applied as cathode catalysts [26]. Furthermore, these catalysts are unstable in acidic media. In recent years, metal single atom catalysts (SACs) have emerged as an exciting new class of catalysts for both the OER and HER. SACs comprise atomically dispersed metals on conductive supports, typically nitrogen‐doped carbon supports due to the relative ease of preparation of M‐N‐C SACs. SACs offer the theoretical advantage of ∼100% metal utilization efficiency in electrocatalysis, meaning high catalytic activities can often be achieved at very low metal loadings (which is particularly important for SACs based on precious metals like Ir, Ru, or Pt). As the field of SACs continues to expand, researchers are now beginning to explore the use of oxide supports in the design of SACs for the OER and HER. Taking iridium as an example, several recent studies [27, 28, 29, 30, 31, 32, 33] have shown that Ir single atoms on different oxide supports can deliver superior alkaline OER activity and durability compared to IrO2. Yin et al. [29] anchored Ir single atoms to NiOOH to form an Ir1@NiOOH electrocatalyst. The introduction of Ir was shown to shorten M─O bonds (M = Ir, Ni) and enhance the M─O covalency, changing the OER pathway from a single‐metal site mechanism (SMSM) in NiOOH to a dual‐metal‐site mechanism (DMSM) in Ir1@NiOOH. During alkaline OER, Ir1@NiOOH achieved low overpotentials of 142 and 308 mV at 10 and 1000 mA cm−2, respectively. Kumar et al. [27] prepared a series of Ir/NiO catalysts, exploring the effect of the local coordination of Ir single atoms on the OER performance. Using x‐ray absorption spectroscopy at the Ir L3‐edge, it was found that Ir atoms partially embedded within the NiO surface (Iremb‐NiO) showed stronger Ir‐Ni second coordination shell interactions compared to Ir single‐atoms adsorbed on the NiO surface. The stronger metal−support interactions in Iremb‐NiO enhanced the alkaline OER mass activity and catalyst durability by suppressing Ir‐dissolution. Electrochemical tests at various pH values established that high‐valence atomic Ir sites embedded on the NiOOH surface facilitated a lattice oxygen mechanism (LOM) for the OER. Somewhat surprisingly, the bifunctional OER/HER activity of iridium single atom catalysts on nickel oxide supports has received little attention to date, prompting a detailed investigation. We hypothesized that by immobilizing Ir single atoms on NiO nanosheets, then optimizing the Ir loading to achieve a mixture of embedded and adsorbed Ir sites, it should be possible to realize efficient and stable bifunctional OER and HER activity across a wide pH range. In this work, a series of Irx‐NiO catalysts (x = 1–12 wt.%) were synthesized using hydrothermally‐grown NiO nanosheets (on carbon cloth) as a support. High‐angle annular dark‐field scanning transmission electron microscopy (HAADF‐STEM) and Ir L3‐edge x‐ray absorption spectroscopy (XAS) confirmed the presence of atomically dispersed Ir species without any detectable IrO2 in the Irx‐NiO catalysts. With increasing Ir loading, the local coordination of Ir evolved from embedded atoms at low loadings (< 4 wt.%), to an even mix of embedded and surface‐adsorbed Ir species at intermediate loadings (8 wt.%), with the proportion of surface adsorbed Ir atoms increasing at even higher loadings (up to 12 wt.%). At the optimal Ir loading of 8 wt.%, strong electronic coupling occurred between the embedded Ir atoms and Ni atoms of NiO promoting Ni oxidation (benefitting OER). Meanwhile, surface‐adsorbed Ir atoms provided accessible sites for hydrogen adsorption, boosting the HER activity. This dual‐site configuration balances metal–support interactions and active site accessibility, enabling efficient bifunctional catalysis across a wide pH range. These findings underscore the importance of controlling the Ir speciation (embedded and surface‐adsorbed) to achieve bifunctional OER/HER activity. Experimental Section Preparation of NiO NiO nanosheets were grown on carbon cloth via a simple hydrothermal method. Prior to NiO hydrothermal growth, 3 × 5 cm2 carbon cloth pieces were pretreated in 0.5 m H2SO4 under a constant potential of 1.69 V vs. reversible hydrogen electrode for 2 h. After this electrochemical treatment, the carbon cloth was thoroughly rinsed with deionized water until neutral pH was reached. This pretreatment step was performed to enhance the hydrophilicity of the carbon cloth, which improves NiO nanosheet growth uniformity and adhesion. In a typical synthesis, 89 mg of NH4F and 360.4 mg of urea were dissolved in 18 mL of deionized water. After ultrasonication to obtain a homogeneous solution, 12 mL of a 0.1 mol/L Ni(NO3)2·6H2O solution was added under constant stirring. The resulting solution and a piece of cleaned carbon cloth (2.2 × 4.5 cm) were then transferred to a 50 mL Teflon‐lined stainless‐steel autoclave and heated at 120°C for 12 h. Following natural cooling to room temperature, the modified carbon cloth was washed repeatedly with water and then dried at 60°C. The sample was then calcined at 350°C for 2 h in air, yielding NiO nanosheets on carbon cloth (denoted herein simply as NiO). Preparation of Irx‐NiO To prepare the Ir8%‐NiO catalyst, an impregnation solution was first prepared by dissolving 8 mg of IrCl3·xH2O in 10 mL of ethanol under magnetic stirring. A piece of carbon cloth supporting NiO nanosheets (1 cm × 2 cm) was then immersed in the 0.8 mg/mL IrCl3·xH2O impregnation solution for 10 min, then taken out and dried at 60°C. After calcination at 350°C for 2 h in air, Ir8%‐NiO was obtained. The Ir loading of Ir8%‐NiO was confirmed to be 7.96 wt.% (Table S1) by inductively coupled plasma atomic emission spectrometry (ICP‐AES). Ir‐NiO catalysts with different Ir loadings of 1%, 4%, 10%, and 12% were prepared by the same general method using IrCl3·xH2O impregnation solutions with concentrations of 0.1, 0.4, 1, and 1.2 mg/mL, respectively. The obtained samples are denoted as Ir1%‐NiO, Ir4%‐NiO, Ir10%‐NiO, and Ir12%‐NiO, respectively. Characterization The morphologies of the obtained materials were examined by transmission electron microscopy (TEM, Tecnai G2 F20) and scanning electron microscopy (SEM, HITACHI SU‐8010). High‐angle annular dark field scanning transmission electron microscopy (HAADF‐STEM) imaging was performed on an aberration‐corrected FEI Titan Themis G2 300 microscope. All images were acquired at 300 kV with a beam size of ∼0.6 Å, a convergence angle of 30 mrad, and collection angles of 39.4‐200 mrad. The crystalline structures of samples were determined by x‐ray diffraction (XRD, Bruker D8 Advance diffractometer) using Cu Kα radiation at 40 kV and 40 mA. Inductively coupled plasma atomic emission spectrometry (ICP‐AES) measurements were performed on a Thermo Fisher Scientific iCAP‐Q. X‐ray photoelectron spectroscopy (XPS) analyses were performed on an AXIS SUPRA x‐ray photoelectron spectrometer (SHIMADZU‐KRATOS) using a Mg‐Kα x‐ray source. Synchrotron x‐ray absorption near‐edge spectroscopy (XANES) and extended x‐ray absorption fine‐structure (EXAFS) data at the Ir L3‐edge and Ni K‐edge were obtained on the MEX1 beamline at the Australian Synchrotron. Electrochemical Measurements Electrochemical data for all samples were collected on a CHI 760E electrochemical workstation (CH Instruments, Inc., Shanghai). A conventional three‐electrode system, consisting of a Hg/HgO or SCE reference electrode, a graphite rod counter electrode, and a catalyst‐modified carbon cloth working electrode, was used. The geometric active area of the working electrodes used for OER and HER measurements was 1.0 cm2, defined by the exposed area of the electrode in contact with the electrolyte. Electrolytes used were 1 m KOH, 0.1 m PBS or 0.5 m H2SO4. All potentials were reported on the reversible hydrogen electrode (RHE) scale. After twenty cyclic voltammetry (CV) scans, the polarization data were collected using linear sweep voltammetry (LSV) at a scan rate of 5 mV s−1. All polarization curves were corrected for ohmic‐drop compensation with the ohmic resistance obtained by electrochemical impedance spectroscopy (EIS). The EIS measurements were performed with a 5 mV amplitude over a frequency range from 100 000 to 0.1 Hz. The stability of the catalyst‐modified working electrodes was first measured by CV tests at 10 mV s−1 for 50 cycles, followed by chronopotentiometry stability tests. All experiments were carried out at room temperature. Electrode Fabrication and MEA Tests For the anion‐exchange membrane water electrolyser (AEMWE) single‐cell test, a membrane electrode assembly (MEA) was fabricated using an anion exchange membrane (Fumasep FAA‐3‐PK‐130, Fuel Cell Store), Ir8%‐NiO or RuO2 coated on carbon cloth as the anode, and Ir8%‐NiO or Pt/C‐20% coated on a gas diffusion layer as the cathode. MEA was prepared using the catalyst‐coated substrate method. The areal loading of the catalyst on the carbon cloth was 2.0 mg/cm2, unless otherwise specified. The electrochemical tests on the single cell assembled with the MEA were carried out using a SP‐300 potentiostat (BioLogic Science Instruments) in a two‐electrode cell with full serpentine flow‐fields on both sides (1 cm ×1 cm) at 85°C. PTFE gaskets were used to make the cell leak‐tight. 1.0 m KOH was used as the electrolyte for the AEMWE. The electrolyte was supplied to the assembled single cells at a flow rate of 5 mL min−1. LSV curves were measured at a scan rate of 5 mV s−1. Computational Methods Density functional theory calculations for the oxygen evolution reaction and hydrogen evolution reaction on model catalyst surfaces were performed in accordance with the methods described in our recently published work [34] and described in detail in the Supporting Information. Results and Discussion Structural Characterization Irx‐NiO electrocatalysts with iridium loadings (x) ranging from 1–12 wt.% were synthesized by a simple two‐step strategy (Figure S1). NiO nanosheets were first grown on carbon cloth. SEM and TEM characterization studies (Figures S2 and S3) showed the carbon cloth fibers to be uniformly coated by thin NiO nanosheets. After adsorption of Ir species from solution, followed by calcination at 350°C, Irx‐NiO electrocatalysts with similar morphologies to the NiO support were obtained. Microscopy characterization data for the Ir8%‐NiO catalyst is shown in Figure 1a–h. The SEM images (Figure 1a–c) showed Ir8%‐NiO possesses the same nanosheet morphology as NiO. TEM images of Ir8%‐NiO (Figure 1d–h) showed distinct lattice fringes at high magnifications, most of which could readily be indexed to the (111) crystal facets of NiO (Figure 1i). No lattice fringes due to IrO2 were seen, suggesting the iridium was atomically dispersed in the sample. Inductively coupled plasma atomic emission spectroscopy (ICP‐AES) confirmed that Ir loading in the Ir8%‐NiO sample was 7.96 wt.%. The scanning transmission electron microscopy (STEM) image (Figure 1j) and corresponding energy‐dispersive spectroscopy (EDS) mapping images confirmed a uniform distribution of Ir species throughout the NiO nanosheets without any obvious aggregation as would be expected for Ir metal or IrO2 nanoparticles. The aberration‐corrected high‐angle annular dark‐field scanning transmission electron microscopy (HAADF‐STEM) images of Ir8%‐NiO (Figure 1k–m) showed isolated bright spots with small sizes (∼0.3 nm) on the surface of the NiO nanosheets, verifying the presence of atomically‐dispersed iridium. Morphological characterization of the Ir8%‐NiO catalyst. (a–c) SEM images, (d–g) TEM images, (h,i) HRTEM images and the corresponding FFT patterns, (j) STEM and corresponding elemental mapping images, (k–m) Aberration‐corrected HAADF‐STEM images of Ir8%‐NiO. Figure 2a shows XRD patterns for NiO and Ir8%‐NiO, both of which are dominated by peaks from the carbon cloth and NiO (the XRD pattern for the carbon cloth used in this work is shown in Figure S4). After introduction of iridium, no new peaks appeared, indicating the absence of any crystalline Ir metal or iridium oxide phases. Figure S5 shows the XRD patterns for all the Irx‐NiO samples with different Ir contents (from 1 to 12 wt.% Ir). All XRD patterns were near identical, showing peaks due to NiO and the carbon cloth. Even at Ir loadings up to 12%, no peaks due to Ir metal or IrO2 were detected. Next, x‐ray photoelectron spectroscopy (XPS) measurements were carried out to investigate the near‐surface region chemical compositions of the Irx‐NiO catalysts. Figure 2b shows the XPS survey spectrum for NiO revealing the presence of Ni, O, and C, with the carbon signal mainly being due to the carbon cloth and/or adventitious hydrocarbons. The XPS survey spectrum of Ir8%‐NiO was clearly different from those of pristine NiO, showing additional peaks due to iridium. The O 1s spectra (Figure 2c) for NiO and Ir8%‐NiO were deconvoluted into two peaks. The peak at lowest binding energy was due to lattice oxygen (O2−) in NiO, whilst the broader peak at high binding energy likely contained contributions from surface hydroxyl groups (OHads) and defective oxygen atoms near defects such as cation vacancies [35]. Quantitative analysis of the O 1s XPS spectra shows that increasing Ir loading progressively decreases the ratio of lattice oxygen to surface‐adsorbed species (Figure S6 and Table S2), including hydroxyl groups and defect‐related oxygen: 1.07 (NiO), 0.82 (Ir1%‐NiO), 0.62 (Ir4%‐NiO), 0.51 (Ir8%‐NiO), 0.42 (Ir10%‐NiO), and 0.19 (Ir12%‐NiO). At first glance, it might appear that the Ir atoms are filling oxygen lattice sites, thus leading to attenuation of the lattice oxygen signal around 530 eV. However, the much larger size of an iridium atom compared to an oxygen atom would prevent such a substitution. Instead, it is more realistic that the atomically dispersed Ir atoms induce lattice distortions, leading to the creation of oxygen vacancies (with attenuation of the 530 eV peak). The Ni 2p XPS spectra of NiO and Irx‐NiO catalysts were deconvoluted into three sets of peaks, each with a 2:1 area ratio and a spin‐orbit splitting of ∼19.8 eV (Figure 2d and Figure S7). The peaks were due to the Ni 2p3/2 and Ni 2p1/2 signals of Ni2+ (Ni 2p3/2 = 854.3 eV), Ni3+ (Ni 2p3/2 = 856.2 eV), and Ni2+ “shake‐up” satellites (Ni 2p3/2 = 861.6 eV) [36], respectively. The presence of Ni3+ in the samples was likely due to partial oxidation of Ni2+ during air annealing at 350°C, along with electronic interactions with Ir that help stabilize higher oxidation states of nickel (see below). The Ni3+/Ni2+ ratio was higher for Ir8%‐NiO than NiO. Figure 2e shows high‐resolution Ir 4f spectra for IrO2 and Ir8%‐NiO. The Ir 4f spectrum of IrO2 showed two sets of peaks, each with a 4:3 area ratio and a spin orbit splitting of 2.9 eV. The intense set of peaks at 62.0 and 65.0 eV could readily be assigned to the Ir 4f7/2 and Ir 4f5/2 peaks, respectively, of Ir4+ in IrO2. The weaker set of peaks at 64.0 and 67.1 eV, are attributed to either Ir4+ shake‐up satellites or a more oxidized surface iridium species, likely an Ir6+ species [37, 38]. In the case of Ir8%‐NiO, the Ir 4f spectrum partially overlapped with the Ni 3p spectrum of NiO. For Ir8%‐NiO, the Ir 4f peaks were shifted to lower binding energies compared to IrO2, suggestive of the presence of Ir3 + rather than Ir4+. Taken together with the higher Ni3+/Ni2+ ratio seen in Ni 2p spectrum for Ir8%‐NiO (compared to NiO), the data points to a redistribution of electron density between Ni and Ir, whereby Ir single atoms accept electrons from Ni2+ (leading to more surface Ni3+). Figure 2f shows N2 adsorption–desorption isotherms for NiO and Ir8%‐NiO. The Brunauer‐Emmett‐Teller (BET) specific surface area of Ir8%‐NiO (73.09 m2 g−1) was smaller than that of NiO (116.51 m2 g−1), with the adsorption‐desorption hysteresis loop seen for NiO missing for Ir8%‐NiO. These changes are attributed to grain growth and partial pore collapse during the second calcination step used to prepare Ir8%‐NiO. Structural characterization of the Irx‐NiO catalysts. (a) XRD patterns of Ir8%‐NiO and NiO. (b) XPS survey spectra of Ir8%‐NiO, NiO and IrO2. High‐resolution XPS spectra for Ir8%‐NiO, NiO and IrO2: (c) O 1s spectra, (d) Ni 2p spectra, and (e) Ir 4f spectra. (f) Nitrogen adsorption−desorption isotherms collected at 77 K for NiO and Ir8%‐NiO. To further investigate the electronic structure and local coordination environment of Ir single atoms in the Irx‐NiO catalysts, synchrotron‐based x‐ray absorption near edge structure (XANES) and extended x‐ray absorption fine structure (EXAFS) measurements were conducted. The Ir L3‐edge XANES spectra (Figure 3a) show that the absorption edge positions of the Irx‐NiO samples were much closer to IrO2 than Ir metal, suggesting the presence of oxidised states of Ir (either Ir3+ or Ir4+) [39, 40]. However, the Irx‐NiO samples exhibit a more pronounced white‐line intensity compared to IrO2, particularly for the Ir8%‐NiO catalyst. Since the white line reflects the density of unoccupied 5d states [41, 42, 43], this enhancement indicates that Ir in the Irx‐NiO samples experiences a more perturbed electronic environment than in bulk IrO2. The evolution of white‐line intensity across the series provides further insight into the electronic structure of Ir as a function of loading. At low Ir contents (1–4 wt.%), Ir atoms are most likely incorporated into the NiO lattice by filling Ni2+ vacancies (producing embedded Ir species), as supported by the XPS results showing a lower surface oxidation state (Ir3+) compared with IrO2. Lattice incorporation modifies the local coordination symmetry of Ir and increases the number of unoccupied 5d states, thereby contributing to the enhanced white‐line intensity. When the Ir loading reaches 4 wt.%, all the Ni2+ vacancies are almost filled with Ir atoms, thus any additional Ir begins to deposit on the surface. At 8 wt.% loading, both lattice‐incorporated and surface‐adsorbed Ir species coexist in roughly equal proportions. Further increasing the Ir loading to 10–12 wt.% results only in additional surface‐adsorbed Ir, whilst the amount of embedded Ir atoms remains unchanged. Notably, the white‐line intensity remains higher than that of IrO2 even at highest Ir loadings used in this work (12 wt.%), indicating that the embedded Ir species still dominate the unoccupied 5d‐state density and electronic structure of the catalysts. Electronic structure and local coordination geometry of the Irx‐NiO catalysts. (a) Ir L3 edge XANES spectra, (b) FT‐EXAFS spectra in R‐space (c) Ni K edge XANES spectra, (d) FT‐EXAFS spectra in R‐space for the various Irx‐NiO catalysts. (e) WT‐EXAFS spectra for Ir8%‐NiO, IrO2, NiO, and Ir foil. To gain deeper insight into the local structure of the Irx‐NiO catalysts, Ir L3‐edge EXAFS spectra were analyzed in detail (Figure 3b). The R‐space plots showed a first shell Ir─O peak at ∼1.6 Å and second shell Ir─Ni peak at ∼2.7 Å. Both peaks intensified with increasing Ir loadings up to 4 wt.%, which is explained by Ir atoms being embedded into the NiO lattice (with both Ir─O and Ir─O─Ni interactions being important). At 4 wt.%, the first shell Ir─O and second shell Ir─Ni peaks exhibit comparable intensities. With increasing Ir loading up to 8 wt.%, the Ir─O peak intensified relative to the Ir─Ni peak, indicating a distinctive structural shift toward Ir─O coordination and reduced second shell Ir─Ni interactions. The data reflects the emergence of surface‐adsorbed Ir species alongside the originally formed embedded Ir species. At Ir loadings of 10 and 12 wt.%, the Ir─Ni peak was very weak with the spectra dominated by Ir─O signals of surface‐adsorbed Ir species (becoming more similar to that of the IrO2 reference sample). These spectral trends are further supported by EXAFS fitting results (Figures S8–S13 and Tables S3–S8), which reveal a clear evolution in the local coordination environment of Ir atoms. At an Ir loading of 4 wt.%, Ir atoms were predominantly embedded within the NiO lattice, evidenced by a first shell Ir─O coordination number of 6 and a second shell Ir─Ni coordination number of 7. As the Ir loading was increased from 4 to 10 wt.%, the Ir─O coordination number remained the same (i.e. 6), but Ir─Ni coordination progressively decreased due to the increasing abundance of surface‐adsorbed Ir atoms. The EXAFS fitting of the 12 wt.% Ir sample revealed the presence of a minor Ir─Cl scattering path (CN = 1.6, R = 2.34 ± 0.02 Å), which was absent at lower loadings. This clearly originated from the IrCl3 precursor used to prepare the catalysts. In summary, the Ir L3‐edge data provides good evidence for the formation of embedded Ir species at loadings up to 4 wt.%, together with the co‐existence of surface‐adsorbed Ir species at higher loadings (which lower the average Ir─Ni coordination in the samples). The Ni K‐edge XANES spectra (Figure 3c) revealed that the Irx‐NiO catalysts exhibited slightly higher absorption edge onset energies compared to NiO, indicating partial oxidation of Ni2+ to Ni3+ (consistent with the XPS findings). The white line intensities for the Irx‐NiO catalysts were also higher than that of NiO, indicating more unoccupied 4p states. For the Irx‐NiO catalysts, a slight decrease in the white line intensity was observed at very high Ir loadings, possibly due to structural saturation or weakened Ir─Ni interactions caused by excessive surface‐adsorbed Ir atoms. Complementary insights were provided by the Ni K‐edge EXAFS spectra (Figure 3d), which showed Ni─O and Ni─Ni peaks at ∼1.6 and ∼2.5 Å, respectively, in R‐space. Compared to pure NiO (Ni─O CN = 5.6; Ni─Ni CN = 12.3), the Irx‐NiO samples exhibit a slightly increased Ni─O coordination number (CN = 6) and a reduced Ni─Ni coordination number (CN = 10). These changes suggest that Ir incorporation induces local structural disorder and disrupts Ni─Ni connectivity, likely due to the larger atomic radius of Ir. Although embedded Ir atoms are expected to form Ni─Ir coordination at ∼2.7 Å, this position overlaps with the broad Ni─Ni peak. As a result, any Ni─Ir signals are likely masked by Ni─Ni scattering and cannot be resolved as a distinct feature. As discussed above, at low Ir loadings (1%–4%), Ir atoms are proposed to embed into the NiO lattice as isolated single atoms. Due to the larger size of Ir atoms compared to Ni atoms, this embedding of Ir atoms in NiO is expected to induce lattice disorder. The decrease in Ni‐Ni peaks intensity on going from NiO to Ir4%‐NiO is attributed to increased structural disorder in the Ni lattice with the embedding of iridium atoms. At higher Ir loadings up to 10 wt.%, the Ni‐Ni peak intensity does not change much, since the additional Ir atoms are simply surface adsorbed on the catalyst surface and hence should not disrupt the NiO lattice. Considering the preserved Ni─O coordination revealed by EXAFS fitting, together with the increased contribution of hydroxyl and defect‐related oxygen species in the O 1s XPS spectra, it is likely that the attenuation of second shell Ni─Ni FT‐EXAFS peak at high Ir loading results from local lattice disorder rather than degradation of the NiO framework. The wavelet transform (WT) EXAFS spectra (Figure 3e) provide simultaneous resolution in both k and R spaces, enabling discrimination of backscattering atoms with different atomic numbers. For the Ir8%‐NiO sample, a dominant intensity maximum is observed at approximately R ≈ 1.5 Å with a relatively low k value (∼6–8 Å−1), which is characteristic of Ir─O coordination. Notably, no obvious high‐k feature (k > 10 Å−1) is detected, indicating the absence of Ir─Ir metallic bonding. In contrast, the Ir foil exhibits a strong WT maximum at higher k (∼12–14 Å−1) and R ≈ 2.5 Å, corresponding to Ir─Ir scattering, confirming its metallic nature. Similarly, IrO2 shows contributions from both Ir─O (low k) and Ir─Ir (higher k) scattering paths, reflecting its crystalline oxide structure. The Ni K‐edge WT‐EXAFS shows slight changes in the R–k distribution compared with pure NiO, consistent with minor local distortions of the NiO lattice induced by Ir incorporation. These observations provide direct evidence for the coexistence of lattice‐embedded and surface‐adsorbed Ir in the Ir‐NiO catalysts (for Ir loadings > 4 wt.%), as well as subtle structural perturbations of the NiO framework. To quantify the distribution, the Ir8%‐NiO sample was subjected to an acid treatment (0.1 m HCl) to leach the surface‐adsorbed Ir single atoms. Subsequent ICP‐AES analysis revealed an Ir mass loss of 48.2%, suggesting that nearly half (51.8%) of the Ir atoms were embedded within the NiO lattice. This is consistent with the results obtained from EXAFS. Furthermore, post‐washing XAS analysis (Figure S14) showed that the oxidation state remained in the +4 region, and the Ir‐O coordination environment stayed intact. Notably, the Ir‐Ni scattering path in the EXAFS spectra became more pronounced, confirming that the remaining Ir species are structurally integrated into the NiO host. Remarkably, after the removal of surface‐adsorbed species from the Ir8%‐NiO sample, the coordination numbers (CN) for Ir‐O (CN ≈ 6) and Ir‐Ni (CN ≈ 7) were found to be nearly identical to those obtained from the fresh Ir4%‐NiO sample (for which all Ir atoms are embedded in the NiO lattice). This comparative analysis effectively addresses the averaging limitation of bulk EXAFS and confirms the co‐existence of almost equimolar amounts of lattice‐embedded and surface‐adsorbed Ir sites in the Ir8%‐NiO catalyst. Electrochemical Measurements The unique electronic structures of the Irx‐NiO catalysts, featuring different ratios of embedded and surface‐adsorbed iridium atoms, were expected to significantly impact their catalytic behaviour for the OER. The OER performance of all the Irx‐NiO catalysts and a commercial Ir/C catalyst (supported on carbon cloth) were evaluated in 1 m KOH, 0.1 m PBS and 0.5 m H2SO4 to compare their activity and durability as wide pH OER catalysts. For experiments conducted under neutral conditions, some minor perturbances were observed in the LSV curves, which can be attributed to transient bubble dynamics and local electrolyte disturbances. These do not significantly affect the determination of onset potentials or Tafel slopes. The OER onset potentials were determined from a region where the catalytic current clearly dominates over the Ni redox peak, ensuring accurate evaluation of OER performance. The iR‐corrected polarization curves obtained in 1 m KOH electrolyte (Figure 4a) demonstrate that the Ir8%‐NiO catalyst exhibits superior OER performance compared to all other Irx‐NiO counterparts as well as commercial Ir/C. The Ir8%‐NiO catalyst achieved OER current densities of 10, 50, and 100 mA cm−2 at overpotentials of only 177, 251, and 274 mV, respectively, which were significantly lower than values for commercial Ir/C (Figure 4d). Furthermore, Ir8%‐NiO exhibited superior activity compared to most recently reported alkaline OER catalysts (Table S9), demonstrating exceptional performance for the OER in an alkaline electrolyte. The Ir8%‐NiO catalyst exhibited a large double‐layer capacitance (Cdl) of 82.2 mF cm−2 in alkaline solution (Figures S15 and S16), which was significantly higher than Ir/C (30.3 mF cm−2) and NiO (2.4 mF cm−2). This indicates that the Ir atoms on Ir8%‐NiO were accessible and created abundant active sites for the OER. The OER reaction kinetics of Ir8%‐NiO were analyzed using the Tafel plot method, with the Ir/C and NiO as reference catalysts (Figure 4g). As expected, the Tafel slope of Ir8%‐NiO (72.7 mV dec−1) was much lower than those of Ir/C (98.7 mV dec−1) and NiO (124.4 mV dec−1), indicating faster OER kinetics on Ir8%‐NiO. Furthermore, the stability of Ir8%‐NiO during the OER in 1 m KOH was evaluated using chronopotentiometric testing over a 120 h period (Figure 4j). At a high current density of 100 mA cm−2, the potential only increased by 12 mV over 120 h in the alkaline electrolyte, revealing Ir8%‐NiO possessed promising stability (which we attribute to the strong metal–support interaction which prevents the over‐oxidation and dissolution of Ir). Additionally, the polarization curve recorded after the chronopotentiometric test closely resembled the initial curve (inset in Figure 4j), further confirming the promising stability and durability of Ir8%‐NiO. Inspired by these findings, we next evaluated the OER activity of the Irx‐NiO catalysts in a neutral environment (0.1 m PBS). As shown in Figure 4b, Ir8%‐NiO again outperformed all the other catalysts in terms of OER activity. Specifically, Ir8%‐NiO achieved current densities of 10 and 50 mA cm−2 at overpotentials of 275 and 377 mV (Figure 4e), respectively. In contrast, commercial Ir/C required overpotentials of 540 and 584 mV to achieve the same current densities. This confirmed the exceptional catalytic performance of Ir8%‐NiO for the OER in neutral media. The low Tafel slope of 92.3 mV dec−1 for Ir8%‐NiO (Figure 4h) demonstrated superior OER kinetics compared to Ir/C (119.4 mV dec−1) and NiO (302.8 mV dec−1). Figure 4k shows the chronopotentiometric curve of Ir8%‐NiO during OER in 0.1 m PBS, revealing promising stability at 100 mA cm−2 over the duration of the 120 h stability test (voltage increase of only 44 mV). Additionally, the polarization curves before and after the 120 h‐test revealed negligible performance degradation (Figure 4k, inset). Compared to recently reported OER electrocatalysts in neutral electrolytes (Table S10), Ir8%‐NiO demonstrates superior electrocatalytic activity and durability. Under acidic conditions, the Ir8%‐NiO also demonstrated superior OER activity and kinetics to its Irx‐NiO counterparts and Ir/C (Figure 4c,f,i), with overpotentials of 174, 231 and 259 mV needed to reach current densities of 10, 50 and 100 mA cm−2, respectively, with a low Tafel slope of 70.5 mV dec−1. Compared to recently reported oxide‐based OER electrocatalysts in acidic media, Ir8%‐NiO offers remarkable electrocatalytic activity (Table S11). After the 120 h chronopotentiometric test, the potential increased by only 12 mV (Figure 4l), indicating almost no degradation in catalytic activity. Furthermore, the polarization curve after the stability test was identical to the initial curve (inset in Figure 4l), confirming the good stability of the Ir8%‐NiO catalyst for the OER in a 0.5 m H2SO4 electrolyte. Notably, when the current density was increased to 100 mA cm−2 in acidic media (Figure S17), the catalyst showed an obvious rise in potential, indicating significantly reduced stability compared with the test at 10 mA cm−2. This result suggests that the oxidation of NiO to NiOOH, along with Ni3+ leaching, becomes more significant under harsher oxidative regimes, reflecting the intrinsic stability limitation of Ni‐based oxides in strongly acidic environments. The electrochemical impedance spectroscopy (EIS) Nyquist plots in Figure S18 show that Ir8%‐NiO exhibited smaller semicircle diameters in the low‐frequency region in alkaline, neutral, and acidic solutions compared to all the other Irx‐NiO catalysts and Ir/C, indicating faster charge transfer on the Ir8%‐NiO electrocatalyst. Figure S19 shows multistep chronopotentiometric curves for Ir8%‐NiO in alkaline, neutral and acidic solutions with current densities ranging from 10 to 500 mA cm−2. The potential change occurred immediately at the beginning of each current density step and remained unchanged thereafter, indicating excellent mass transport and stability of the Ir8%‐NiO catalyst. SEM and TEM images of the Ir8%‐NiO catalyst obtained after the stability tests (Figures S20–S22) reveal that the overall morphology remained largely unchanged, indicating good structural integrity during prolonged operation. XPS characterization after the OER stability tests in alkaline solution (Figure S23) showed a clear increase in the Ni3+ content, consistent with the formation of NiOOH or more specifically Ni(Ir)OOH, with no aggregation of Ir being observed. In contrast, under acidic OER conditions, the Ni3+/Ni2+ ratio decreased (Figure S25), suggesting that NiOOH formation was either suppressed or the resulting phase was unstable due to Ni3+ leaching. This interpretation is further supported by post‐OER XRD analysis (Figure S26), which showed no distinct diffraction peaks corresponding to NiOOH. The absence of these peaks indicates that any NiOOH formed under acidic conditions was likely of low crystallinity or amorphous in nature. This conclusion is corroborated by the TEM images (Figures S20–S22), where the NiOOH domains appear poorly ordered. Although pristine NiO is inherently unstable in acidic environments, the incorporation of Ir was found to modulate the electronic structure, thereby stabilizing the NiO lattice to some degree. ICP‐AES measurements after the OER stability tests (Table S12) further confirmed suppressed Ni dissolution in the Ir─NiO system, although some leaching still occurred. Collectively, these results demonstrate that Ir incorporation enhances the chemical and structural stability of NiO in acidic media to some extent, with the catalyst maintaining impressive long‐term durability under alkaline conditions. Next, Ir‐normalized mass activities were calculated for the different catalysts (Figure S27), with Ir8%‐NiO offering the best mass activity. OER performance of the Irx‐NiO catalysts. (a–c) OER polarization curves. (d–f) Overpotentials of the catalysts at 10, 50, and 100 mA cm−2. (g–i) Tafel slope of Ir8%‐NiO, Ir/C, and NiO. (j–l) Chronopotentiometric curves of Ir8%‐NiO. The insets show a comparison of the polarization curves before and after the chronopotentiometric tests. Measurements were performed in 1 m KOH (a,d,g,j), 0.1 m PBS (b,e,h,k), and 0.5 m H2SO4 (c,f,i,l), respectively. Recently, we reported that the immobilization of Ru single atoms on Co3O4 nanorods created an efficient bifunctional electrocatalyst for overall water splitting [34]. This prompted us to examine the HER activity of the Irx‐NiO catalysts alongside NiO, Pt/C and Ir/C. The as‐synthesized NiO showed low activity for HER in alkaline, neutral and acidic media (Figure S28). The iR‐corrected polarization curves from the HER tests in 1 m KOH (Figure 5a) reveal that Ir8%‐NiO showed excellent HER activity in terms of overpotential compared to the other Irx‐NiO samples. The slight deviations observed at high current densities for Ir4%‐NiO and Ir12%‐NiO are likely due to transient hydrogen bubble dynamics, which are common in HER measurements where there is vigorous gas evolution. As expected, the commercial Pt/C catalyst exhibited very good HER catalytic activity, with overpotentials of 17, 50 and 83 mV at current densities of 10, 50 and 100 mA cm−2, respectively, in alkaline media (Figure 5d). The HER activity of Ir8%‐NiO was comparable, with overpotentials of only 29, 45, and 58 mV required to achieve the same current densities. The overpotential at a current density of 10 mA cm−2 in alkaline conditions for Ir8%‐NiO was lower than most recently reported HER electrocatalysts (Table S13), demonstrating its remarkable bifunctional OER/HER activity. Additionally, compared with the other Irx‐NiO catalysts, the charge transfer resistance (Rct) of Ir8%‐NiO was the smallest (Figure S29a), indicating Ir8%‐NiO allowed ultrafast charge transfer during HER. The HER activity of the Ir10%‐NiO catalyst was similar to Ir8%‐NiO, but increasing the Ir loading to 12 wt.% caused the activity to sharply decrease (possibly due to some agglomeration of Ir). With increasing Ir loading above 8 wt.%, EXAFS fitting reveals a progressive decrease in the Ir‐Ni coordination number despite a nearly constant Ir─O coordination, indicating that excessive Ir primarily accumulates as surface species with weakened interfacial coupling to the NiO lattice. This accumulation (excess) of surface Ir atoms lowers the intrinsic activity of HER‐active centers in the Ir12%‐NiO sample. The Tafel plots (Figure 5g) showed that Ir8%‐NiO exhibited a smaller HER Tafel slope (28.6 mV dec−1) compared to Pt/C (38.1 mV dec−1), Ir/C (81.6 mV dec−1), and NiO (152.3 mV dec−1) in alkaline media, demonstrating Ir8%‐NiO offered faster reaction kinetics in catalysing the HER. The Tafel slope values determined for Ir8%‐NiO also suggested that the HER process followed a Volmer‐Heyrovsky mechanism, where the rate‐determining step is the Heyrovsky step [44, 45]. The chronopotentiometric curve in Figure 5j showed no significant change in the potential needed to achieve a current density of 100 mA cm−2 over 120 h. During the 120 h‐test, the potential change was almost negligible (becoming more negative by only 8 mV). The HER polarization curves recorded before and after the chronopotentiometric test (inset in Figure 5j) were similar, indicating that the Ir8%‐NiO maintained good catalytic activity for HER during prolonged use. In neutral media (0.1 m PBS), the iR‐corrected polarization curves (Figure 5b) demonstrated the superior HER activity of Ir8%‐NiO compared to commercial Pt/C (and the other tested catalysts except Ir10%‐NiO for which the performance was similar to Ir8%‐NiO). This enhanced performance likely results from the combined contributions of the surface‐adsorbed Ir atoms and the NiO support. The overpotentials at current densities of 10 and 50 mA cm−2 were 54 and 259 mV (Figure 5e), respectively, which were slightly better than those of Pt/C. Furthermore, the HER performance of Ir8%‐NiO significantly outperforms recently reported neutral HER electrocatalysts (Table S14). A Tafel slope of 135.3 mV dec−1 was observed for Ir8%‐NiO (Figure 5h), which is smaller than that of Pt/C. After the 120 h stability test, the potential became more negative by only 22 mV. Additionally, there was no significant difference between the HER polarization curves before and after 120 h of testing (inset in Figure 5k). In acidic conditions, Ir8%‐NiO required very low overpotentials of only 24, 51, and 64 mV to achieve current densities of 10, 50, and 100 mA cm−2, respectively (Figure 5c,f). Additionally, it exhibited a smaller Tafel slope of 48.3 mV dec−1 (Figure 5i) compared to their counterparts. The overpotential of Ir8%‐NiO at a current density of 10 mA cm−2 is better than most recently reported acidic HER electrocatalysts (Table S15). Due to the intrinsic instability of NiO in acidic media, the HER stability test in 0.5 m H2SO4 was performed under relatively mild conditions (20 mA cm−2 for 50 h) to mitigate rapid structural degradation. Although Ir incorporation enhanced the stability compared to pristine NiO, the long‐term durability of the Ir‐NiO catalysts in acid was inferior to that achieved in alkaline or neutral electrolytes. Nevertheless, Ir8%‐NiO demonstrated impressive initial HER activity, sustaining continuous operation for 50 h in 0.5 m H2SO4 with only negligible performance loss (Figure 5l). As shown by our x‐ray absorption measurements, the incorporation of atomically dispersed Ir introduces strong interfacial interactions within the NiO matrix. XPS analysis indicates charge redistribution across the Ir‐O‐Ni framework, which enhances metal‐oxygen bonding and partially stabilizes the oxide lattice in acidic environments. Furthermore, embedded Ir species act as local coordination centers that mitigate rapid structural degradation, although some instability remains under high current densities. The mass activity for the HER was also calculated to assess the utilization efficiency of iridium. As with the OER tests, the Ir8%‐NiO catalyst offered the highest mass activity for HER, as detailed in Figure S30. HER performance of the Irx‐NiO catalysts. (a–c) HER polarization curves. (d–f) Overpotentials of the catalysts at 10, 50, and 100 mA cm−2. (g–i) Tafel slope of Ir8%‐NiO, Pt/C, Ir/C, and NiO. (j–l) Chronopotentiometric curves of Ir8%‐NiO. The insets show a comparison of the polarization curves before and after the chronopotentiometric tests. The measurements were performed in 1 m KOH (a,d,g,j), 0.1 m PBS (b,e,h,k), and 0.5 m H2SO4 (c,f,i,l), respectively. The correlation between the Ir speciation (embedded only or embedded + varying amounts of surface‐adsorbed Ir) and catalytic activity was evident across the Irx‐NiO series of catalysts. While increasing the Ir loading from 1% to 4% enhances Ir─Ni coordination and strengthens metal–support interactions, the OER activity does not peak at this stage. This suggests that excessively strong second shell Ir─Ni interactions may overly stabilize the electronic structure of Ir atoms, limiting their flexibility to participate in dynamic electron transfer and lattice oxygen activation processes. Such rigidity can hinder the modulation of Ni oxidation states and reduce overall catalytic efficiency. At 8 wt.% Ir loading, a balanced dual‐site configuration emerges, characterized by moderate Ir─Ni coordination (CN = 6) and the coexistence of embedded and surface‐adsorbed Ir species. Embedded Ir atoms maintain sufficient interaction with the NiO matrix to facilitate Ni2+ → Ni3+ oxidation during the OER, while surface‐adsorbed Ir atoms provide additional active sites and enhance electron transport pathways. This balanced structure offers both structural stability and electronic flexibility, resulting in the high OER activity. In addition to its superior OER performance, the 8 wt.% Ir sample also exhibits significantly enhanced HER activity. This improvement can be attributed to the dual‐site configuration: embedded Ir atoms contribute to electronic modulation of the NiO substrate, resulting in a highly conductive framework, while surface‐adsorbed Ir atoms provide highly exposed active sites for hydrogen adsorption (*H). The increased accessibility of these sites lowers the energy barrier for proton reduction and accelerates hydrogen evolution kinetics. These findings highlight that the speciation of Ir atoms and their degree of interaction with the NiO support play a critical role in realising bifunctional OER/HER catalytic activity. To explore the feasibility of utilizing Ir8%‐NiO as a wide pH OER/HER catalyst, it was applied as both the anode and cathode catalyst in water electrolysers with 1 m KOH, 0.1 m PBS and 0.5 m H2SO4 as the electrolytes. The performance of the electrolysers was assessed using LSV measurements. For comparison, an electrolyser was constructed with commercial RuO2 and Pt/C as the anode and cathode catalysts, respectively. The electrolyser containing the Ir8%‐NiO electrodes delivered overall water‐splitting current densities of 100 mA cm−2 at low operating voltages of 1.50 V in 1 M KOH and 1.51 V in 0.5 m H2SO4, respectively (Figure 6a,c). In neutral conditions, a voltage of only 1.50 V was needed to achieve a current density of 20 mA cm−2 (Figure 6b). These values were much lower than those of the commercial Pt/C(−)||RuO2(+) electrolyser and superior to almost all recently reported bifunctional OER/HER electrocatalysts (Table S16). Inspired by the impressive overall water splitting performance of Ir8%‐NiO, a membrane electrode assembly (MEA) was developed for use in an AEMWE single cell. This MEA (Figure 6d) was fabricated using the catalyst‐coated substrate (CCS) method, incorporating an anion exchange membrane (Fumasep FAA‐3‐PK‐130, Fuel Cell Store). It was tested in 1.0 m KOH at 85°C. The AEMWE single cell containing Ir8%‐NiO as both the anode and cathode catalysts achieved a current density of 1.0 A cm−2 at an operating voltage of 2.02 V (Figure 6e), far surpassing the performance of the device containing the commercial RuO2 and Pt/C catalysts. Notably, the relatively higher overpotential compared to the three‐electrode system is likely due to additional interfacial and transport resistances inherent to the membrane electrode assembly. To evaluate the stability of the AEMWE with Ir8%‐NiO, the single cell was subjected to chronopotentiometric measurements at 1.0 A cm−2 for 120 h (Figure 6f), during which the potential increased by only 6%. These results indicate that Ir8%‐NiO is a very promising overall water splitting catalyst possessing both high efficiency and durability for practical AEMWE applications. The structural integrity of Ir8%‐NiO was investigated after the MEA stability test (Figure S31), showing the structural integrity of the catalyst was maintained. Figures S32 and S33 show XAS and EXAFS spectra for Ir8%‐NiO before and after the MEA durability test. After OER, the atomically dispersed Ir species undergo local structural reorganization without compromising catalytic performance. The Ir L3‐edge XANES spectrum shows a positive shift in the white line position and a decrease in intensity, indicating an increase in the Ir oxidation state and possible changes in the local coordination environment. EXAFS analysis reveals reduced intensities and slight shifts in both Ir─O and Ir─Ni peaks, which are attributed to dynamic bonding and surface reconstruction under OER conditions rather than catalyst degradation. EXAFS fitting (Figure S34 and Table S17) further shows that the Ir–O coordination number decreased from ∼6 to 5, while the Ir─Ni coordination number dropped from ∼6 to 4. These changes suggest that Ir atoms undergo partial detachment from the NiO lattice and adopt a more open coordination environment during reaction (consistent with Ni(Ir)OOH formation). The lower Ir─O coordination may result from surface oxygen deficiency or dynamic interaction with electrolyte species, exposing Ir centers to the reaction interface. Similarly, the decreased Ir─Ni coordination implies weakened lattice anchoring, possibly due to surface migration of embedded Ir or electrochemical restructuring. Notably, these changes are observed after 120 h of continuous OER operation, indicating that the evolved Ir species were relatively stable and in a catalytically active configuration rather than products of transient activation or structural degradation. The catalyst maintained excellent OER activity and electrochemical durability throughout the test period, confirming the robustness of the restructured Ir sites. Concurrently, the NiO support undergoes a transformation into a NiOOH‐like phase. Ni K‐edge XANES shows a slight positive shift in the white line and a modest increase in intensity, indicating partial oxidation of Ni2+ to Ni3+. The corresponding EXAFS spectra showed peaks at ∼1.6 and ∼2.5 Å, corresponding to Ni─O and Ni─Ni coordination, respectively. These changes reflect the formation of a partially ordered NiOOH phase, characterized by layered Ni─O octahedral motifs and the development of medium‐range Ni─Ni correlations. However, both peaks remain less intense than those of crystalline NiO, implying that the resulting NiOOH structure retains significant disorder, likely due to Ir incorporation and electrochemical reconstruction during OER. Overall water splitting performance. (a–c) Polarization curves of Ir8%‐NiO || Ir8%‐NiO and RuO2 || Pt/C for water electrolysis in alkaline, neutral, and acidic solutions, respectively. (d) Schematic illustration of the structure of MEA for anion‐exchange membrane water electrolysis single cells. (e) Polarization curves for water electrolysis using Ir8%‐NiO as both the anode and cathode catalysts, compared with the commercial RuO2 and Pt/C. (f) Durability of the electrolyser based on Ir8%‐NiO at a constant current density of 1 A cm−2 for 120 h at 85°C. Under HER conditions following 120 h of MEA operation, the Ni K‐edge spectra show minimal changes, indicating that the NiO phase is largely preserved or partially converted to Ni(OH)2. The Ir L3‐edge spectra shift slightly to lower energy, accompanied by a decrease in white‐line intensity, consistent with partial reduction of Ir species. EXAFS fitting (Figure S35 and Table S18) reveals that the Ir─O coordination number remains at 5, while the Ir–Ni coordination number is 4, with a slight increase in Ir─Ni bond length from 3.01 ± 0.01 Å (pre‐reaction) to 3.03 ± 0.01 Å. Additionally, the emergence of Ir─Ir coordination signals suggests the formation of small Ir clusters or short‐range Ir─Ir interactions, likely induced by prolonged exposure to highly reducing conditions. However, while Ir atoms undergo partial reduction and local rearrangement under HER conditions, they largely retain an atomic dispersion and strong interaction with the NiO support. The modest elongation of Ir─Ni bond lengths and reduced peak intensity imply a slight weakening of Ir─Ni interactions, possibly due to lattice relaxation or surface restructuring. Together, these findings demonstrate that the Ir‐NiO system exhibits good structural stability under reductive conditions, with Ir species remaining catalytically accessible. Further, the local rearrangement of Ir atoms under OER/HER conditions do not compromise catalytic performance, indicating that these dynamic changes lead to a stable, active configuration rather than degradation. This structural adaptability under both oxidative and reductive environments contributes to the long‐term operational overall water splitting stability of the Ir‐NiO catalyst. Table S19 benchmarks the durability of the Ir8%‐NiO catalyst with other recently reported state‐of‐the‐art Ir‐based catalysts in alkaline media. DFT Calculations To gain deeper insights into the OER mechanism and the exceptional structural stability of Ir8%‐NiO, density functional theory (DFT) calculations were performed. Based on prior electrochemical observations, the NiO substrate is oxidized to NiOOH under OER conditions. Accordingly, four structural models were constructed for theoretical evaluation: pristine NiOOH, NiOOH with a single surface Ir atom (NiOOH_surface_Ir), NiOOH with a single embedded Ir atom (NiOOH_embedded_Ir), and NiOOH containing both embedded and surface Ir atoms (NiOOH_both_Ir), as illustrated in Figure 7a. Figure 7b–e show the calculated OER free energy profiles for these configurations, while Figures S36–S39 present the corresponding data under various pH conditions. Among the four models, NiOOH_both_Ir exhibits the lowest theoretical overpotential (0.52 eV) for OER, followed by NiOOH_surface_Ir (0.77 eV), NiOOH_embedded_Ir (0.95 eV), and pristine NiOOH (1.04 eV). For NiOOH, NiOOH_embedded_Ir, and NiOOH_both_Ir, the potential‐limiting step is the formation of OOH* from O* (O* + OH− → OOH* + e−), whereas for NiOOH_surface_Ir, the limiting step shifts to the conversion of OH* to O* (OH* + OH− → O* + H2O + e−). This suggests that the O* → OOH* step is energetically more favorable in the presence of both embedded and surface Ir, leading to a reduced overpotential (consistent with our experimental observations, where the Ir8%‐NiO catalyst with roughly equal amounts of embedded and surface‐adsorbed Ir species offered the best activity). (a) Optimized structures of O*, OH*, and OOH* adsorbed on (i) pristine NiOOH, (ii) NiOOH with surface Ir, (iii) NiOOH with embedded Ir, and (iv) NiOOH with both surface and embedded Ir. (b–e) OER energy profiles for the corresponding four configurations. (f) Charge density differences upon OH* adsorption on NiOOH with surface and embedded Ir (yellow: electron accumulation; blue: depletion; iso‐surface = 0.01 e/Å3). (g) H* adsorption structures on NiO with embedded Ir and with one to four surface Ir atoms. (h) HER energy profiles on Pt, and NiO with embedded and increasing numbers of surface Ir atoms. (i) d‐band center values of NiO with embedded Ir and one to four surface Ir atoms. The pristine NiOOH system displayed intrinsically poor OER activity. At low Ir doping levels, Ir atoms tend to occupy subsurface (embedded) positions, with OH* binding too strongly (−0.4 eV), hindering catalytic performance. With increasing Ir content, surface Ir sites begin to emerge, offering weaker OH* binding (∼0.98 eV) and substantially improved OER activity. Notably, on the NiOOH_both_Ir model, O* and OH* preferentially reside on the surface Ir atoms, while OOH* binds favorably to the embedded Ir site, despite the thermodynamically stronger adsorption of O* and OH* on the embedded site. This site selectivity arises from two factors for O* and OH*: (i) the prohibitively high migration barriers from the surface to the embedded Ir site, calculated as 0.97 eV for OH* and 1.22 eV for O* using the CI‐NEB method; and (ii) the over‐stabilization of these intermediates on the embedded site, which by the Sabatier principle would raise the limiting‐step free energy were the entire cycle confined to that site. In contrast, the weaker adsorption of OOH* on the surface Ir gives a much lower migration barrier (0.46 eV), allowing OOH* to relocate to the embedded Ir site, where it is favorably stabilized and lowers the free energy of the potential‐limiting O* → OOH* step. Each site therefore handles the intermediate for which its binding strength is closest to optimal: the surface Ir binds O* and OH* at moderate strength, while the embedded Ir stabilizes OOH*. This division of labor accounts for the superior OER activity of the NiOOH_both_Ir model relative to structures containing only surface Ir. To further elucidate the distinct electronic roles of the surface and embedded Ir sites, the charge density difference upon OH* adsorption was analyzed. As shown in Figure 7f, OH* gains 0.35 electrons upon adsorption on NiOOH_surface_Ir, whereas it gains 0.49 electrons on NiOOH_embedded_Ir. The lower charge accumulation on surface Ir indicates a weaker interaction with OH*, which facilitates subsequent reaction steps. This provides a mechanistic explanation for the enhanced OER performance associated with surface Ir sites (with co‐existing embedded Ir sites acting to lower the overpotential of the overall OER process). To elucidate the HER mechanism, a series of NiO‐based models were constructed featuring one embedded Ir atom and varying numbers of surface Ir atoms (from one to four). As shown in Figure 7g, the optimized structures illustrate H* adsorption on these configurations. The corresponding HER free energy profiles (Figure 7h) reveal a clear activity trend: Ir_sur_1 ≈ Ir_sur_2 ≈ Pt > Ir_sur_3 ≈ Ir_sur_4 > Ir_embedded. The embedded single‐atom Ir site exhibits overly strong H* adsorption (∼−0.95 eV), making it less favorable for HER catalysis. In contrast, surface Ir sites provide more moderate H* binding energies, thereby facilitating improved HER activity. Interestingly, as the number of surface Ir atoms increases, H* adsorption weakens progressively. Among all models, Ir_sur_1 and Ir_sur_2 show the most favorable energetics, achieving HER performance comparable to that of Pt. To gain further insight into the electronic origin of this trend, the projected density of states (DOS) was calculated for embedded Ir and surface Ir atoms (Figure S40), and the corresponding d‐band centers were extracted (Figure 7i). The results show that the d‐band center of surface Ir shifts upward with increasing Ir coverage, consistent with the observed enhancement in H* adsorption strength based on d‐band theory. Notably, the embedded Ir atom exhibits the highest d‐band center, aligning with its strongest H* binding energy. Together, these findings emphasize the pivotal role of surface Ir atoms in tuning the electronic structure and modulating adsorption behaviour. While embedded Ir alone is unable to promote efficient HER, the introduction and optimization of surface Ir sites enables fine control over adsorption energetics. The strong correlation between theoretical predictions and experimental results underscores the potential of rationally engineered Ir site distributions for designing high‐performance HER catalysts. Conclusion In this study, we synthesized a series of Irx‐NiO catalysts and identified Ir8%‐NiO as the optimal composition for achieving outstanding bifunctional OER/HER catalytic performance and stability across a wide pH range. Structural and spectroscopic analyses revealed a clear evolution in Ir configuration with increasing loading, transitioning from embedded atoms (at loadings up to 4 wt.%) to a coexistence of embedded and surface‐adsorbed Ir species at 8 wt.%. This dual‐site configuration enables a balance between strong metal–support interactions and accessible active sites, which is critical for efficient water splitting. At this optimal loading, embedded Ir atoms facilitate Ni2+ → Ni3+ oxidation and enhance OER kinetics. Simultaneously, surface‐adsorbed Ir atoms provide highly exposed sites for hydrogen adsorption, accelerating HER activity. Ir8%‐NiO exhibits low overpotentials for OER (177, 275, and 174 mV) and HER (29, 54, and 24 mV) in alkaline, neutral, and acidic media, respectively, positioning it as a promising overall water splitting catalyst. Membrane electrode assemblies incorporating Ir8%‐NiO as both anode and cathode deliver efficient and durable overall water splitting in alkaline conditions. Complementary DFT calculations reveal that embedded Ir sites promote site‐specific adsorption of OER intermediates, while surface‐adsorbed Ir atoms mitigate the overly strong H* binding of embedded Ir, tuning HER energetics toward an optimal range. The observed shift in the d‐band center with increasing surface Ir content further explains the enhanced HER performance. These findings underscore the importance of the Ir speciation and interaction strength with the NiO support in achieving bifunctional OER/HER catalytic efficiency. These results guide the development of high‐performance single‐atom catalysts for water splitting over a wide range of pH values. Author Contributions Hui Yang: formal analysis, data curation, Writing – review and editing, methodology. Yongfang Zhou: conceptualization, investigation, writing – original draft, methodology, validation, writing – review and editing, formal analysis, data curation. Yu Mao: conceptualization, investigation, writing – original draft, writing – review and editing, methodology, software, formal analysis, data curation. Zhanghao Ren: formal analysis, writing – review and editing. Yufei Zhao: methodology, writing – review and editing, formal analysis, data curation. Geoffrey I.N. Waterhouse: conceptualization, investigation, funding acquisition, resources, supervision, project administration, formal analysis, writing – review and editing, writing – original draft, methodology. Cuizhu Ye: formal analysis, writing – review and editing. Anton Tadich: methodology, data curation. Aaron Marshall: resources, writing – review and editing, methodology, funding acquisition. Shanghai Wei: formal analysis, data curation . Conflicts of Interest The authors declare no conflicts of interest. Supporting information Supporting File: smll74803‐sup‐0001‐SuppMat.pdf.