Large‐Scale Synthesis of Spinel NixMn3‐xO4 Solid Solution Immobilized with Iridium Single Atoms for Efficient Alkaline Seawater Electrolysis

Abstract Seawater electrolysis not only affords a promising approach to produce clean hydrogen fuel but also alleviates the bottleneck of freshwater feeds. Here, a novel strategy for large‐scale preparing spinel NixMn3‐xO4 solid solution immobilized with iridium single‐atoms (Ir‐SAs) is developed by the sol–gel method. Benefitting from the surface‐exposed Ir‐SAs, Ir1/Ni1.6Mn1.4O4 reveals boosted oxygen evolution reaction (OER) performance, achieving overpotentials of 330 and 350 mV at current densities of 100 and 200 mA cm–2 in alkaline seawater. Moreover, only a cell voltage of 1.50 V is required to reach 500 mA cm–2 with assembled Ir1/Ni1.6Mn1.4O4‖Pt/C electrode pair under the industrial operating condition. The experimental characterizations and theoretical calculations highlight the effect of Ir‐SAs on improving the intrinsic OER activity and facilitating surface charge transfer kinetics, and evidence the energetically stabilized *OOH and the destabilized chloride ion adsorption in Ir1/Ni1.6Mn1.4O4. This work demonstrates an effective method to produce efficient alkaline seawater electrocatalyst massively.


Introduction
Employing renewable electricity combined with water electrolyzers to produce hydrogen presents an appealing and sustainable strategy to combat climate changes and secure energy. [1,2] DOI: 10.1002/advs.202200529 Principally, the overall efficiency of water electrolysis is hampered by oxygen evolution reaction (OER) in consideration of its sluggish, multistep protoncoupled electron transfer process. [3] Over the past few years, numerous works have been devoted to explore efficient OER catalysts, including transition metal oxides, [4][5][6] (oxy)hydroxides, [7][8][9] sulfides, [10][11][12] phosphides, [13][14][15] and nitrides. [16][17][18] Moreover, part of them even prevails over the benchmark IrO 2 /RuO 2 in the OER activity and stability, [19][20][21] which certainly invigorates the blossom of the water electrolyzer technique. However, the freshwater feeds may become a bottleneck for large-scale water electrolysis. Given the abundant natural resources as well as the improved ionic conductivity due to dissolved salts, seawater thereby becomes the optimal choice to alleviate this issue. The main challenge for seawater electrolysis is the competitive active chlorine species formation reactions (ACSFRs), including the chlorine evolution reaction in low pH and chlorine oxidation reactions in high pH to generate hypochlorite. [22] Thus, OER selectivity for seawater electrolysis is highly essential. According to the Pourbaix diagram for the oxygen evolution reaction and chloride chemistry, OER selectivity over ACSFRs can be achieved with the maximum potential difference of 490 mV in the high pH value. [23,24] Therefore, considering the kinetics and standard potentials, an alkaline environment is more favorable to avoid hypochlorite formation during seawater electrolysis. Besides, considering the industrial requirement for delivering large current density (>500 mA cm -2 ) under seawater electrolyzers, the design of high-performance seawater electrocatalysts is still challenging. [25] The primary principle to design seawater electrocatalysts is to improve the corrosion resistance of electrocatalysts/seawater interfaces, that is, to stabilize -OOH formation or destabilize chloride ion adsorption, thereby enhancing the OER selectivity. [26] Unfortunately, the recently reported strategies for seawater electrocatalysts design are still involved in doping modification, defects construction, and surface engineering to lower the d-band centers, scilicet mainly concerning the improvement of the OER activity. [27][28][29][30] The up-to-date seawater electrocatalysts also basically follow those in water electrolyzers, whereas few can meet the www.advancedsciencenews.com www.advancedscience.com industrially mandated overpotential of 300 mV at 500 mA cm -2 with a cell voltage of below 1.60 V. Besides, apart from the prerequisite OER efficiency and stability, industrial seawater electrolyzer also require the electrocatalysts that can be easily scaled up, which is hardly achieved by template-based synthesis or exfoliation process. [31,32] Therefore, developing innovative seawater electrocatalysts with efficient OER activity and high selectivity, and mass-productive characteristics is of great necessity.
Herein, inspired by the impressive selectivity and stability of Mn-based oxides in the acidic electrolyte, [33,34] we consider that the Mn-based oxides may also represent good selectivity in the alkaline electrolyte due to the similar chlorine ions adsorbing process by surface polarization. Generally, the intrinsic activity of OER electrocatalysts can be tuned by elemental doping or introducing vacancies. [35,36] Herein, benefitting from the structural similarity of cubic spinel NiMn 2 O 4 and Ni 2 MnO 4 , we consider that tuning the OER activity of NiMn-based oxide can be possibly achieved by preparing Ni x Mn 3-x O 4 solid solution. Furthermore, massively producing Ni x Mn 3-x O 4 through the sol-gel process is experimentally feasible. [37] Therefore, we employed cubic spinel Ni x Mn 3-x O 4 solid solution in the seawater electrolyzer, which manifested excellent chlorine oxidation resistance. Meanwhile, we introduced Ir single atoms into Ni x Mn 3-x O 4 to further enhance intrinsic OER performance and increase the number of active sites to alleviate the effect of possible insoluble precipitates. The Ir 1 /Ni 1.6 Mn 1.4 O 4 reveals low overpotentials of 330 and 350 mV to achieve the current densities of 100 and 200 mA cm -2 . Moreover, only a cell voltage of 1.50 V is required to reach 500 mA cm -2 with assembled Ir 1 /Ni 1.6 Mn 1.4 O 4 ‖Pt/C electrode pair under the industrial operating condition, demonstrative of the feasibility for Ir 1 /Ni 1.6 Mn 1.4 O 4 employed in alkaline seawater electrolyzer. In addition, the surface structural regulation and the plausible mechanism of Ir single atoms on the enhanced OER performance are also discussed. The surface elemental charge state changes involved in Ir 1 /Ni 1.6 Mn 1.4 O 4 were first measured by X-ray photoelectron spectroscopy ( Figure S9, Supporting Information). Generally, the position of binding energy for a certain atom is affected by its coordination environment or valence state. [38] Herein, the Ir 4f peak in Ir 1 /Ni 1.6 Mn 1.4 O 4 shifts toward lower binding energy relative to IrO 2 , indicative of its lower valence state than that in IrO 2 . [39] We further employed the X-ray absorption near-edge structure (XANES) spectra and density functional theory (DFT) calculations to illustrate the coordination information of Ir atoms in the   [40] Meanwhile, only one intensity maximum at 6 Å -1 can be observed from the wavelet transformed contour plots for Ir 1 /Ni 1.6 Mn 1.4 O 4 (Figure 2c), which is assigned to Ir-O coordination without Ir-Ir signal. The FT-EXAFS results referenced to our theoretical most stable Ir 1 /Ni 1.6 Mn 1.4 O 4 geometry are further fitted in k and R spaces to investigate the coordination configuration. As revealed in Figure 2d,e, the FT-EXAFS fitting results comply well with the measured results (Table S2, Supporting Information), and the coordination number for Ir-O is 1.9, which evidences the Ir-O 2 center in Ir 1 /Ni 1.6 Mn 1.4 O 4 , also consistent with our DFT calculated result.

Electrocatalytic OER Performance of Ir 1 /Ni 1.6 Mn 1.4 O 4
The electrocatalytic performances of Ir 1 /Ni 1.6 Mn 1.4 O 4 for OER are investigated in a standard three-electrode configuration in alkaline seawater. The salinity and OER polarization curves for  Figure S16, Supporting Information), where the peak position assigned to Ir-O vibration is nearly unchanged, illustrative of its coordination structural robustness. [41] Besides, the stability of the Ir single atom in Ir 1 /Ni 1.6 Mn 1.4 O 4 is also theoretically calculated, where its binding energy (−7.36 eV, Figure S10, Supporting Information) is much smaller than the corresponding cohesive energy (6.94 eV atom −1 ), [42] indicative of the thermodynamical stability of Ir single atoms. Notably, benefitting from the sol-gel process for preparing Ni 1.6 Mn 1.4 O 4 , Ir 1 /Ni 1.6 Mn 1.4 O 4 can be easily mass-produced in the hectogram scale (inset of Figure 3f), rendering the industrial application possible. Meanwhile, the gas evolution during the OER is measured by gas chromatography to subsequently calculate the Faradaic efficiency (FE) ( Figure S17, Supporting Information). The ca. 100% efficiency for Ir 1 /Ni 1.6 Mn 1.4 O 4 indicates the polarization current solely consumed in the OER process.

Electrocatalytic Performance for Overall Water Splitting
The Ir 1 /Ni 1.6 Mn 1.4 O 4 assembled with commercial Pt/C electrode is constructed as a two-electrode system for overall water splitting in alkaline seawater. As revealed in Figure 4a, the Ir 1 /Ni 1.6 Mn 1.4 O 4 ‖Pt/C catalytic couple affords current densities of 100 and 200 mA cm -2 at 1.62 and 1.69 V ( of 390 and 460 mV, respectively). In contrast, the commercial IrO 2 ‖Pt/C electrode pairs achieve the same current densities at 1.78 V and 1.98 V, respectively. The electrochemical stability for overall water splitting is investigated by long-term J-t measurement (Figure 4b).
No obvious decay is observed after continuous operation for 60 h, illustrative of the robust structural stability. In addition, the post-OER electrolyte is monitored by potassium iodide starch paper, where no color change appears, suggesting the excellent OER selectivity and anti-corrosion property in alkaline seawater circumstances. The gas evolutions in the two-electrode cell were also measured by gas chromatography (Figure 4c). Especially, the oxygen release rate approaches its theoretical value, which illustrates the high electron utilization efficiency with FE of O 2 above 99.0% ( Figure S18, Supporting Information). In addition, the industrial application requires a large current density (e.g., 500 and 1000 mA cm -2 ) in concentrated alkaline circumstances (typically 6 m KOH solution at 60˚C). [49] The Ir 1 /Ni 1.6 Mn 1.4 O 4 ‖Pt/C catalytic couple affords current densities of 500 and1000 mA cm -2 at only1.51 and 1.56 V in 6 m KOH electrolyte (Figure 4d), respectively, superior to the IrO 2 ‖Pt/C couple (The HER performance of Pt-C electrode is also supplemented in Figure S19, Supporting Information). A similar polarization curve is achieved for the Ir 1 /Ni 1.6 Mn 1.4 O 4 ‖Pt/C catalytic couple in alkaline seawater ( 500 = 1.50 V, 1000 = 1.56 V), indicating the feasibility of employing alkaline seawater for electrocatalytic water splitting. Given the operational stability as importing metric, the Ir 1 /Ni 1.6 Mn 1.4 O 4 ‖Pt/C catalytic couple can maintain the excellent electrocatalytic activity at 1.50 V with a large current density of 500 mA cm -2 over 50 h without apparent degradation in 6 m KOH + seawater at 60˚C ( Figure S20, Supporting Information). In addition, compared to other benchmarking electrocatalysts with large current densities (Figure 4e), the operation voltage for the Ir 1 /Ni 1.6 Mn 1.4 O 4 ‖Pt/C catalytic couple is still dominant, which renders it a promising industrial candidate for overall water splitting. Besides, the ideal power supply in the coastal areas could be abundant solar energy. [22] Therefore, the PV-electrolysis system comprising a commercial Si PV module connected to the two-electrode cell is constructed. As revealed in Figure 4f, the seawater electrolyzer driven by a commercial Si solar cell achieves an impressively high current of 1.04 A under a photovoltage of 2.85 V without generating hypochlorite, indicating the feasibility for electrocatalytic water splitting with the PV-electrolysis system.

Origin of the Enhanced OER Mechanism
To explore the origin of the efficient OER activity of the Ir 1 /Ni 1.6 Mn 1.4 O 4 catalyst, we further investigate its pseudocapacitance and charge transfer characteristics. Given the high dependence of electrochemical activity on the number of active sites, the electrochemical surface areas (ECSA) estimated by the electrochemical double-layer capacitances (C dl ) are calculated (Figure  5a), where the C dl value of Ir 1 /Ni 1.6 Mn 1.4 O 4 (76.17 mF cm -2 ), much higher than that of Ni 1.6 Mn 1.4 O 4 (29.98 mF cm -2 ), suggesting that the enhanced OER activity is contributed mainly by the increased number of active sites. To further determine whether the enhanced OER activity is solely contingent on the increased active sites, the ECSA-normalized OER polarization curves are calculated ( Figure S21, Supporting Information). The Ir 1 /Ni 1.6 Mn 1.4 O 4 still possesses lower overpotentials, illustrative of the Ir single atoms not only creating new active sites but also enhancing its intrinsic activity. To access the activation energy (E a ) of the surface oxygen evolution reaction, which can be extracted from the slope of the Arrhenius plot, we measured the variation of current density along with the temperature (Figure 5b). [55] The Ir 1 /Ni 1.6 Mn 1.4 O 4 reveals a much smaller E a value (24.5 kJ mol -1 ) compared to that of Ni 1.6 Mn 1.4 O 4 (53.5 kJ mol -1 ), which demonstrates the importance of Ir single atoms in accelerating the surface OER kinetics. Besides, the charge transport behavior of Ir 1 /Ni 1.6 Mn 1.4 O 4 is investigated by electrochemical impedance spectra (EIS) measurements. As depicted in Figure 5c, the decreased interfacial charge transfer resistance (R ct ) in Ir 1 /Ni 1.6 Mn 1.4 O 4 illustrates its increased charge transfer rate. Moreover, the charge-carrier density deduced from the Mott-Schottky plot is calculated to evaluate the charge-carrier concentration (Figure 5d (Figure 5e), demonstrating the high-spin state of the Ir atom in Ir 1 /Ni 1.6 Mn 1.4 O 4 and indicating its improved conductivity. [56] Besides, as revealed in the optimized coordination model (Figure 5f), charge redistribution appears on the surface of Ir 1 /Ni 1. indicating that the easier absorption of OHto generate *OH in the OER process on Ir 1 /Ni 1.6 Mn 1.4 O 4 . [57] Additionally, we calculated the Gibbs free energy evolution of the crucial intermediates to illustrate the potential-determining step in the water oxidation process (Figures S22 and S23, and Tables S2-S5, Supporting Information). Notably, except for the reduced largest Gibbs free energy difference in Ir 1 /Ni 1.6 Mn 1.4 O 4 (ΔG = 0.404 eV) compared to that of Ni 1.6 Mn 1.4 O 4 (ΔG = 1.449 eV), the rate-determining step (RDS) has changed from the *O formation process to the *OOH formation process (Figure 5h,i), which is considered to be the fundamental reason for the reduced overpotential and improved OER kinetics in Ir 1 /Ni 1.6 Mn 1.4 O 4 . Meanwhile, the aggressive chloride ions in seawater corrode electrocatalysts generally through metal chloride-hydroxide formation mechanisms, which involve three steps, that is, chloride ion adsorption, dissolution by further coordination, and conversion from chloride to hydroxide. Herein, we first calculate the adsorption energies for chlorine ions on Ni 1.6 Mn 1.4 O 4 and Ir 1 /Ni 1.6 Mn 1.4 O 4 , which is determined to be 3.159 and 4.355 eV ( Figure S24 and Table S6

Conclusions
In summary, we developed a simple and effective approach to mass-productively synthesize Ir-SAs immobilized Ni 1. By evaluating its overall water-splitting performance, a cell voltage of only 1.50 V is required to reach 500 mA cm -2 with assembled Ir 1 /Ni 1.6 Mn 1.4 O 4 ‖Pt/C electrode pair under the industrial operating condition, superior to up-to-date alkaline seawater electrolyzer. The sol-gel strategy in this manuscript presents a promising way to massively prepare highly efficient electrolytes in alkaline seawater, which may inspire more excellent works on developing highly efficient alkaline seawater electrolyzers.

Synthesis of Ni x Mn 3-x O 4 Solid Solution:
The NiMn 2 O 4 powders were synthesized by a sol-gel method. Typically, 5 mmol of Ni(NO 3 ) 2 ·6H 2 O, 10 mmol of Mn(NO 3 ) 2 ·4H 2 O, and 50 mmol of citric acid were dissolved into 20 mL of distilled water. Subsequently, the NH 3 ·H 2 O was added dropwise into the above solution that was placed in an ice-water bath until the pH reached 8. Then, the above mixture was thermally treated at 60°C for 12 h to get a dry gel. Finally, the dry gel was transferred to the muffle furnace and heated at 800°C for 4 h with a heating rate of 5°C·min -1 . [58] The Ni x Mn 3-x O 4 solid solution samples were prepared by altering the ratio of the precursors.
Synthesis of Ir 1 /Ni 1.6 Mn 1.4 O 4 Nanocrystals: Typically, 0.5 g of Ni 1.6 Mn 1.4 O 4 was added into 10 mL ethanol solution with ultrasonic treatment for 5 min to obtain a well-dispersed mixture. Then, 114.6 μL IrCl 3 solution (50 mg mL -1 ) was added into the above dispersion dropwise under magnetic stirring and heated at 50°C to completely volatilize the ethanol. The obtained black powder was then transferred into a ceramic boat and thermal-treated at 400°C for 2 h under an Ar atmosphere. After cooling down to room temperature naturally, the powders were collected and denoted as Ir 1 /Ni 1.6 Mn 1.4 O 4 .
Preparation of Ir 1 /Ni 1.6 Mn 1.4 O 4 and Pt-C Electrodes: Typically, 80 mg of catalyst and 10 mg of acetylene black were placed in a mortar. Subsequently, 50 μL of 5 wt.% Nafion and 25 μL of N-methylpyrrolidone were added dropwise. Then, a uniform ink was obtained after grinding for 30 min. The above suspension was spread on the Ti net and dried at 60°C. The catalyst loading was ≈4mg cm -2 . Characterization: The crystalline structures were analyzed by powder X-ray diffraction (D8, Bruker AXS) with Cu K radiation ( = 1.5418 Å). The morphology and microstructure were characterized using SEM (Hitachi, SU8010), high-resolution TEM (JEM-1011, JEOL), and spherical aberration-corrected TEM (JEM-ARM200F). The XPS (EscaLab 250Xi, Thermo scientific) technique with 30.0 eV pass energy and an Al K line excitation source was employed to identify the elemental compositions and bonding information. The gas evolutions were analyzed by gas chromatography (3420A, Beifen-Ruili Co. Ltd., China).
Electrochemical Measurements: The measurements were conducted on a CHI760E electrochemical workstation with a standard three-electrode system. The as-prepared samples were employed as working electrodes with an average catalyst loading of ≈4 mg cm -2 , and graphite rod and Hg/HgO electrode (1.0 m KOH) were used as counter and reference electrodes, respectively. All the measurements were carried out in O 2 -saturated electrolyte. The measured potentials were calibrated to the reversible hydrogen electrode (RHE) by the equation: E RHE = E Hg/HgO + 0.098 + 0.059 × pH. The EIS spectra were recorded in 0.5 m KOH+seawater at open-circuit potential with the frequency range from 1 MHz to 0.1 Hz. Mott−Schottky plots were recorded from 0 to 1.6 V versus Hg/HgO reference electrode with the frequency of 1 kHz.
Methods for Faradaic Efficiency and TOF Calculation: Details about the calculations of Faradic efficiency and turnover frequency (TOF) are shown below: where is the Faradic efficiency, m is the actual molar number of H 2 or O 2 , n is the number of reactive electrons, F is Faraday's constant (96 485.3 C mol -1 ), I is the current, and t is time.
where j represents the measured current density, S represents the surface area of the electrode (typically 1 cm 2 ), N represents the number of electrons required per mole of gas (H 2 or O 2 ), F is the Faraday's constant F (96 485.3 C mol -1 ), and n is the moles of metal atoms on the electrode. Among others, n is accumulated as all the additive Ir atoms. XAFS Measurements and Analysis: The X-ray absorption fine structure spectra (Ir L3-edge) were collected at Taiwan Synchrotron Radiation Facility (BSRF). The storage rings of BSRF were operated at 2.5 GeV with a maximum current of 250 mA. Using Si(111) double-crystal monochromator, the data collection was carried out in transmission mode using an ionization chamber. All spectra were collected in ambient conditions. The acquired EXAFS data were processed according to the standard procedures using the ATHENA module implemented in the IFEFFIT software packages. The k 3 -weighted EXAFS spectra were obtained by subtracting the post-edge background from the overall absorption and then normalizing with respect to the edge-jump step. [59,60] Subsequently, k 3 -weighted (k) data of Ir L-edge were Fourier transformed to real (R) space using hanging windows (dk = 1.0 Å -1 ) to separate the EXAFS contributions from different coordination shells. The least-squares curve parameter fitting was performed using the ARTEMIS module of IFEFFIT software packages to obtain the quantitative structural parameters around central atoms. [61] Theoretical Calculation: Spin-polarized DFT calculations were performed with the projected augmented wave method, as implemented in the Vienna Ab-initio Simulation Package. [62,63] The exchange-correction function was treated by the generalized gradient approximation (GGA) of Perdew−Burke−Ernzerhof functional, and the wave functions were expanded on a plane wave basis with an energy cutoff of 500 eV. The effective U-J values of 3.9, 6.2, and 0 eV were introduced to account for the strong on-site Coulomb repulsion of Mn, Ni, and Ir (no U correction) atoms, respectively. [64,65] The gamma-centered scheme for K-points grid sampling was applied for all the calculations. For all the calculations, the convergence criteria for the electronic and ionic relaxations are 10 −5 eV and 0.02 eV Å −1 , respectively.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.