Electronic structure modulation of Ru–NiCu bimetallic phosphide for highly efficient and stable overall water splitting in alkaline and seawater conditions Authors: Junjie Du, Qihao Zhang, Heran Ren, Zhipeng Su, Tenglong Ma, Rong-Da Zhao, Ning Xie, Liang Liu, Fufa Wu Published: 2026 ABSTRACT To meet the urgent demand for low-cost and highly active electrocatalysts in water electrolysis for hydrogen production, a ruthenium-doped nickel–copper bimetallic phosphide electrocatalyst (Ru–NiCuP@NF) was successfully prepared. Electrochemical measurements reveal that in 1.0 M KOH electrolyte, the as-prepared Ru–NiCuP@NF catalyst requires only 108 mV for the hydrogen evolution reaction (HER) and 243 mV for the oxygen evolution reaction (OER) at a current density of 10 mA cm−2. This performance is attributed to the metal doping, which effectively modulates the electronic structure and provides robust water dissociation capability. In alkaline seawater, the overpotentials required to achieve 10 mA cm−2 are 120 mV for HER and 278 mV for OER. Furthermore, when assembled into a two-electrode overall water splitting system, a cell voltage of only 1.28 V is required to deliver a current density of 10 mA cm−2 in 1.0 M KOH. Introduction Amid the escalating global energy crises and environmental challenges, the advancement of clean, renewable, and efficient energy conversion technologies has emerged as a crucial strategy. Hydrogen energy, recognized as a sustainable energy source with elevated energy density and non-polluting combustion byproducts, is deemed an optimal alternative to fossil fuels. Water electrolysis for hydrogen production, utilizing benefits such as process simplicity, elevated product purity, and compatibility with renewable energy sources like wind and solar power,1–3 has emerged as a fundamental method for large-scale hydrogen production. Alkaline water electrolysis is preferable for industrial-scale applications owing to its inexpensive equipment costs, moderate reaction conditions, and enhanced safety. This process is constrained by the slow kinetics of the HER and OER, necessitating effective electrocatalysts to reduce reaction energy barriers and improve energy conversion efficiency. Nickel–copper bimetallic catalysts have emerged as focal points of research owing to their copious resources and adjustable intrinsic activity; yet, problems persist, including sluggish HER/OER kinetics in alkaline environments and inadequate stability at elevated current densities.4,5 Noble metal catalysts, such as platinum (Pt) and iridium (Ir), exhibit superior activity; nonetheless, their scarcity and elevated costs significantly restrict large-scale utilization.6–8 Consequently, the present research emphasis in electrocatalysis is on the development of low-cost, high-activity, and durable non-noble metal-based catalysts using doping engineering to accurately adjust electronic structures and interfacial properties. Transition metal phosphides (TMPs) have recently attracted considerable interest in electrocatalysis, regarded as optimal substitutes for noble metals (e.g., Pt, Ir, Ru). In comparison to transition metal sulfides,9,10 selenides,11,12 and carbides,13,14 transition metal phosphides have distinct advantages. The metallic properties confer outstanding electrical conductivity to materials, whereas the adjustable coordination environment provided by phosphorus atoms efficiently enhances the adsorption/desorption energy of intermediates. Moreover, TMPs provide significant compositional tunability and a variety of synthesis techniques.15–17 Nonetheless, the inherent catalytic efficacy of pure TMPs continues to fall short of that of noble metals, chiefly due to the inadequacy of their active sites' electronic structure in substantially reducing the energy barriers of critical reaction steps.18,19 Consequently, the incorporation of a minimal quantity of noble metals by doping has demonstrated efficacy and cost-effectiveness.20–22 Ruthenium (Ru), a platinum group metal, is a quintessential transition metal characterized by exceptional oxidation resistance, substantial chemical stability, and a diverse array of valence states. Despite the exceptional catalytic activity of pristine Ru metal, its elevated cost and finite availability hinder extensive usage. Thus, integrating a minimal quantity of Ru as a dopant into transition metal complexes (e.g., phosphides,23,24 sulfides,25,26 selenides27,28) has emerged as a favored approach to optimize cost and performance. Recent research have shown that Ru-doped nickel-based phosphide catalysts (e.g., Ru–NiCoP, Ru–NiMoP) have exceptional efficacy in alkaline water electrolysis. Zhao et al. documented a ruthenium-doped nickel–cobalt phosphide nanowire array electrocatalyst that attained an overpotential of 108 mV for the HER and 264 mV for the OER at a current density of 100 mA cm−2. The enhancement is ascribed to the microstructural modification caused by Ru doping, which reveals additional active sites and facilitates charge transfer, thus improving the catalytic activity for water splitting.29 Sadhanala et al. synthesized a ruthenium-modified bimetallic phosphide (Ru22NiMoP2/NF). At a current density of 10 mA cm−2, this material exhibited HER overpotentials of 60 mV in 1.0 M KOH and 52 mV in alkaline seawater, with OER overpotentials of 250 mV and 240 mV under identical circumstances, with performance predominantly dependent on the synergistic interaction between Ru and NiMoP2.30 Motivated by these results, trace Ru doping can alter the electronic structure and interfacial characteristics of catalysts,31–38 hence improving their water electrolysis efficiency. In this work, a ruthenium-doped nickel–copper bimetallic phosphide electrocatalyst supported on nickel foam (Ru–NiCuP@NF) was successfully prepared via a hydrothermal method followed by low-temperature phosphorization at 300 °C. The low-temperature phosphorization preserves the porous structure of the substrate while introducing phosphorus sites and forming abundant defects and interfacial structures. For the HER, Ru doping compensates for the insufficient water dissociation capability of NiCuP, enabling rapid hydrogen evolution. For the OER, Ru doping breaks the energy barrier bottleneck of pure Ni-based materials, achieving low-overpotential oxygen evolution and significantly lowering the energy barriers for both the HER and OER. Moreover, Ru has a higher electronegativity than Ni and Cu; upon doping, electrons are transferred from Ni/Cu to Ru, forming electron-rich Ru sites and electron-deficient Ni/Cu sites. This creates dual active centers that synergistically accelerate water dissociation and hydrogen desorption. Electrochemical tests show that in 1.0 M KOH and alkaline seawater, the overpotentials for the HER at a current density of 10 mA cm−2 are 108 mV and 120 mV, respectively. For the OER, the overpotentials at 10 mA cm−2 are 243 mV and 278 mV, respectively. Furthermore, in a two-electrode overall water splitting cell, the cell voltages required to achieve 10 mA cm−2 in alkaline and alkaline seawater are only 1.28 V and 1.64 V, respectively, demonstrating excellent durability. Experimental Materials Nickel foam (NF), ruthenium(iii) chloride trihydrate (Cl3H6O3Ru, 98% purity) were purchased from Shanghai Macklin Biochemical Technology Co., Ltd. Nickel(ii) nitrate hexahydrate (Ni(NO3)2·6H2O), copper(ii) nitrate trihydrate (Cu(NO3)2·3H2O, 99% purity), urea (H2NCONH2), ammonium fluoride (NH4F), and sodium hypophosphite monohydrate (NaH2PO2·H2O) were purchased from Shanghai Aladdin Reagent Co, Ltd. The natural seawater was collected from Qingdao, China, and filtered through a 0.45 µm cellulose acetate membrane to remove suspended solids and plankton. No acid or alkaline pretreatment was applied to preserve the native ionic composition. The filtered seawater was then directly mixed with KOH to prepare the alkaline seawater electrolyte (1.0 M KOH + natural seawater). Preparation of ruthenium-doped nickel–copper precursor Nickel foam (4.0 cm × 3.0 cm) was immersed in hydrochloric acid (1 M) for 15 minutes, ultrasonicated, and then rinsed thoroughly with deionized water. In a Teflon-lined autoclave, 0.2 mmol of Cl3H6O3Ru, 2 mmol of Ni(NO3)2·6H2O, 2 mmol of Cu(NO3)2·3H2O, 15 mmol of NH4F, 15 mmol of urea, and 60 mL of deionized water were added and stirred for 30 minutes. The pre-treated nickel foam was immersed into the prepared precursor solution. The sealed autoclave was placed in an oven for a hydrothermal reaction at 120 °C for 10 h. After natural cooling, the prepared ruthenium-doped nickel–copper precursor was taken out and dried at 60 °C for 24 h. Additionally, nickel–copper precursor, ruthenium–nickel precursor, and ruthenium–copper precursor were prepared using the same method. The mass loading of the catalyst on the nickel foam was determined by the weight difference method. A blank nickel foam was weighed as m1 (0.4434 g). After the complete preparation process, the sample was dried at 60 °C under vacuum for 12 h and weighed as m2 (0.4720 g). The loading was calculated as (m2 − m1)/geometric area, giving a value of approximately 2.38 mg cm−2 for Ru–NiCuP@NF. The Rux–NiCuP@NF catalysts with different ruthenium doping contents (x = 0.1, 0.3, and 0.4 mmol) were prepared following the same procedure described above. Preparation of Ru–NiCuP@NF, NiCuP@NF, RuCuP@NF, and RuNiP@NF The prepared ruthenium-doped nickel–copper precursor was placed in a porcelain boat positioned upstream in a tube furnace. 1 g of sodium hypophosphite was placed in another porcelain boat positioned downstream. The temperature was ramped to 300 °C at a rate of 2 °C min−1 and held for 2 h under an Ar atmosphere. After the tube furnace cooled down, the ruthenium-doped nickel–copper bimetallic phosphide electrocatalyst was obtained. NiCuP@NF, RuCuP@NF, and RuNiP@NF were prepared using the same method. Characterization techniques The phase structure of the catalysts was characterized by X-ray powder diffraction (XRD, Shimadzu-7000, Cu Kα). X-ray photoelectron spectroscopy (XPS, ESCALAB 250, with an Al Kα source) was used to analyze the catalysts before and after the stability test. The morphology of the catalysts was examined by scanning electron microscopy (SEM, Gemini 300-71-31) and transmission electron microscopy (TEM, JEM-2100 PLUS) to observe the microstructure and dispersion of the active components. Electrochemical measurements Electrochemical tests were performed using a CHI 760E workstation (Shanghai Chenhua Instrument Co., China) in 1.0 M KOH and alkaline seawater, using Ru–NiCuP@NF as the working electrode, a mercury/mercury oxide electrode as the reference electrode, and a graphite rod or platinum sheet (10 mm × 15 mm) as the counter electrode. All measured potentials were calibrated to the reversible hydrogen electrode (RHE) using the formula: ERHE = EHg/HgO + 0.059 × pH + 0.098 V. Calibration of the RHE potential was performed as follows, In a 1.0 M KOH electrolyte, with a platinum plate as the counter electrode and Hg/HgO as the reference electrode, high-purity H2 was bubbled for 30 min to saturate the electrolyte. CV was then conducted around 0 V to obtain the equilibrium potential for hydrogen evolution. Based on the measured equilibrium potential, the potential was corrected using the corresponding formula to ensure that all measured potentials were uniformly converted to the RHE scale. Linear sweep voltammetry (LSV) for HER was performed in the range of −1.35 to −0.9 V at 5 mV s−1. LSV for OER was performed in the range of −0.2 to 0.75 V at 2 mV s−1. The HER proceeds with faster kinetics, therefore, a scan rate of 5 mV s−1 can be employed to shorten the testing duration while avoiding surface reconstruction of the catalyst induced by prolonged polarization. In contrast, the OER exhibits sluggish kinetics and is prone to surface phase reconstruction during the oxidation process. Hence, a lower scan rate of 2 mV s−1 is adopted to minimize measurement errors arising from polarization effects, thereby yielding more authentic and stable polarization curves and ensuring accurate and reliable Tafel slope fitting. Electrochemical impedance spectroscopy (EIS) was conducted in the frequency range of 105 to 10−2 Hz with an amplitude of 5 mV. LSV curves were corrected using 90% iR compensation, with the compensation value obtained from EIS. Cyclic voltammetry (CV) was performed at scan rates of 10, 20, 30, 40, and 50 mV s−1 within the same potential range to determine the electrochemical double-layer capacitance (Cdl), which was used to evaluate the electrochemically active surface area (ECSA) of the catalysts for HER and OER. Long-term stability was assessed via chronoamperometry. Results and discussion A ruthenium-doped nickel–copper phosphide spherical electrocatalyst supported on nickel foam was synthesized via a hydrothermal method followed by a low-temperature phosphorization process. The preparation procedure is illustrated in Fig. 1a. After phosphorization, the color of the nickel foam changed from reddish-brown to black. During the hydrothermal process, the trace doping of Ru not only achieved uniform dispersion but also optimized the electronic structure of the NiCu precursor through electronic modulation, laying the foundation for subsequent enhancement of catalytic performance. Owing to the higher electronegativity of Ru compared to Ni and Cu, during the reductive deposition, Ru3+ undergoes strong electronic interaction with Ni0 and Cu0, resulting in a decrease in the electron cloud density around Ni and Cu atoms and creating a slight electron-deficient state. (a) Schematic illustration of the fabrication process of Ru–NiCuP@NF (b–d) SEM images of Ru–NiCuP@NF (e and f) TEM and HRTEM images of Ru–NiCuP (g) SAED pattern (h) HAADF-STEM image and corresponding element mapping. To determine the phase and chemical composition of the target catalyst, XRD analysis was performed. The XRD pattern of Ru–NiCuP@NF (Fig. 2a) shows several distinct diffraction peaks at 18.82°, 33.08°, 35.56°, 40.24°, 44.16°, 47°, 53.74°, and 74.04°, corresponding to the (100), (101), (200), (111), (201), (210), (300), and (212) planes of nickel phosphide (Ni2P PDF#1385), respectively. Other relatively strong characteristic peaks at 38.72°, 41.26°, 43.5°, 45.84°, and 59.92° correspond to the (210), (121), (300), (113), and (204) planes of copper phosphide (Cu3P PDF#1330), respectively. The XRD pattern of Ru–NiCuP@NF is highly similar to that of NiCuP@NF. No other ruthenium compounds were detected, indicating successful doping of Ru into the NiCuP lattice.39,40 (a) XRD patterns of sample (b–f) XPS spectra of Ru–NiCuP@NF before and after electrolysis. The microstructures of the Ru–NiCuP catalysts supported on nickel foam were characterized by SEM, and the results are shown in Fig. 3. After phosphorization, NiCuP exhibits a dense, agglomerated flower-sphere morphology on the surface of the nickel foam (Fig. S1a). Although phosphorization is completed, the active components are mainly stacked on the nickel foam surface as large particles, forming a relatively dense layered structure. This morphology leads to a limited specific surface area and insufficient exposure of active sites. Moreover, the close packing between the particles hinders electrolyte penetration, and the gas bubbles generated during the reaction tend to be trapped inside the agglomerates, making it difficult for them to detach. Consequently, the active sites become blocked, limiting charge transport and reaction kinetics. In contrast, the introduction of Ru significantly alters the growth mode of the phosphorized product, leading to the formation of a three-dimensional open “nanosphere” hierarchical structure (Fig. 1c). At the microscopic level, it evolves into a highly open three-dimensional nanosheet network (Fig. 1d). These nanosheets are nanoscale in thickness and grow intercrossedly, creating abundant edge and step sites (Fig. 1b). This hierarchical porous structure endows the material with an exceptionally large specific surface area and abundant surface active sites, thereby substantially increasing the ECSA of the catalyst and providing a greater number of active sites for HER and OER. In addition, the morphology of RuNiP tends to be dense and block-like (Fig. S1b), whereas Ru–NiCuP exhibits a structure rich in cracks and hierarchical pores. This porous structure not only increases the density of active sites but also greatly facilitates electrolyte infiltration and the diffusion and desorption of reaction gases, thereby optimizing mass transfer efficiency. Meanwhile, RuCuP displays a larger-scale petal-like block stacking morphology (Fig. S1c). These morphological differences endow Ru–NiCuP with a substantially larger active specific surface area and more efficient mass transfer pathways, thus exposing more accessible catalytic active sites and favoring the ECSA and providing more active sites for the HER and OER. (a) HER polarization curves of sample in 1.0 M KOH (b) corresponding Tafel plots (c) Cdl values (d) Nyquist plots (e) Bode phase plots of EIS measured for Ru–NiCuP@NF in 1.0 M KOH at various applied potentials (f) LSV curves of Ru–NiCuP@NF in 1.0 M KOH before and after 72 h of continuous testing at a current density of 24 mA cm−2. The morphology of the target catalyst was further observed by TEM. Ru–NiCuP exhibits a nanoscale granular morphology (Fig. 1e and S3a–c) with uniform particle dispersion and no obvious agglomeration, indicating good crystallinity of the catalyst. In the high-resolution transmission electron microscopy (HRTEM) images, alternating bright and dark lattice fringes can be clearly observed (Fig. 1f), with interplanar spacings of 0.307 nm and 0.297 nm, corresponding to the (012) plane of Cu3P and the (110) plane of Ni2P, respectively. Meanwhile, selected area electron diffraction (SAED) patterns reveal multiple diffraction rings (Fig. 1g), corresponding to the (101), (200), (201) planes of Ni2P and the (204) plane of Cu3P, respectively. These results are in good agreement with the standard lattice spacings of bimetallic phosphides, further confirming the successful doping of Ru atoms into the Ni2P and Cu3P lattices. XPS was employed to determine the chemical states, electron cloud distribution, and surface species composition of the elements on the catalyst surface. Fig. 2b–f shows the XPS spectra of the Ru–NiCuP@NF catalyst before and after the i–t stability test. Comparison reveals slight shifts in the characteristic XPS peaks of each element after the test. First, the survey XPS spectrum is presented. From the full spectrum (Fig. S2), characteristic peaks of Ni 2p, Cu 2p, P 2p, Ru 3d, and C 1s are clearly observed on the catalyst surface. The C 1s peak originates from surface contamination during measurement and residues from sample preparation, and no characteristic peaks of other impurity elements are detected. Fig. 2b displays the Cu 2p XPS spectrum and the corresponding peak-fitting results. The characteristic peaks are located at 952.3 eV, 953.6 eV, 932.7 eV, and 933.7 eV, corresponding to Cu 2p1/2 and Cu 2p3/2, respectively. In the Ni 2p XPS spectrum (Fig. 2c), characteristic peaks appear at 874.0 eV, 873.8 eV, 856.1 eV, and 855.9 eV, assigned to Ni 2p1/2 and Ni 2p3/2. In the P 2p XPS spectrum (Fig. 2d), characteristic peaks are observed at 130.2 eV, 130.5 eV, 129.4 eV, and 129.6 eV, corresponding to P 2p1/2 and P 2p3/2, respectively. In the Ru 3d XPS spectrum (Fig. 2e), characteristic peaks are located at 284.5 eV, 284.7 eV, 279.9 eV, and 281.3 eV, corresponding to Ru 3d3/2 and Ru 3d5/2, respectively. Similarly, in the Ru 3p XPS spectrum (Fig. 2f), the characteristic peaks at 483.8 eV, 460.5 eV, 484.8 eV, and 460 eV correspond to Ru 3p1/2 and Ru 3p3/2. In summary, the chemical valence states and electronic structures of the elements on the surface of the Ru–NiCuP@NF catalyst show no significant changes, with no obvious degradation in peak shape or peak intensity. This result is consistent with the continuous and stable long-term operation of the catalyst in the i–t stability test. Furthermore, morphological analysis of the sample after the i–t stability test reveals no significant changes in its morphology (Fig. S3d–f), fully confirming that the Ru–NiCuP@NF catalyst possesses excellent long-term electrocatalytic stability. Hydrogen evolution reaction measurements The HER performance was evaluated in 1.0 M KOH solution using LSV. The as-prepared Ru–NiCuP@NF exhibited superior performance. In addition, NiCuP@NF, RuCuP@NF, and RuNiP@NF were tested as control samples, and the results showed that Ru–NiCuP@NF possessed better performance, indicating that trace Ru doping can further enhance catalytic activity. At a current density of 10 mA cm−2, the overpotential of Ru–NiCuP@NF was only 108 mV, while those of NiCuP@NF, RuCuP@NF, and RuNiP@NF were 145 mV, 130 mV, and 167 mV, respectively (Fig. 3a). The Tafel slope is an important kinetic parameter for evaluating the HER. The fitted Tafel slope of Ru–NiCuP@NF was 125.79 mV dec−1, which is significantly lower than those of the control samples: NiCuP@NF (136.29 mV dec−1), RuCuP@NF (133.04 mV dec−1), and RuNiP@NF (133.51 mV dec−1) (Fig. 3b). This comparison indicates that Ru doping endows the target catalyst with superior HER activity. Furthermore, CV curves were measured at different scan rates in the non-faradaic potential region, and the Cdl was obtained by fitting the CV curves in the potential range of 0.04–0.14 V (Fig. S4). Obviously, the Cdl of Ru–NiCuP@NF (5.3 mF cm−2) is larger than those of NiCuP@NF (3.48 mF cm−2), RuNiP@NF (3.74 mF cm−2), and RuCuP@NF (4.53 mF cm−2) (Fig. 3c), indicating that Ru–NiCuP@NF possesses a higher ECSA. Meanwhile, the EIS results (Fig. 3d) show that the target catalyst has lower charge transfer resistance compared to the three control samples, demonstrating that Ru doping reduces the charge transfer resistance at the electrode interface and accelerates the electron transfer rate. This further confirms that the introduction of Ru indeed enhances the HER kinetics, which is consistent with the previous LSV results (Fig. 3a). To gain deeper insight into the HER kinetics of the Ru–NiCuP@NF electrocatalyst in alkaline medium, Multi-potential EIS measurements were conducted (Fig. 3e). The three-dimensional Bode phase plots at various applied potentials (vs. Hg/HgO) in 1.0 M KOH are presented. Within the tested potential range (−1.027 to −1.234 V), the low-frequency phase angle peak in the Bode plots decreases with increasing potential, indicating the onset of the HER and a more efficient charge transfer process or adsorption/desorption process.41 Furthermore, as the overpotential increases, the width of the peak broadens, reflecting the complexity of the reaction pathway and the reconstruction of the surface state. Under strongly polarizing conditions, no obvious uncontrolled shrinkage of the low-frequency capacitive arc is observed, demonstrating that Ru doping effectively modulates the electronic structure of NiCuP and maintains excellent interfacial charge transfer capability, thereby significantly enhancing the intrinsic HER activity. Catalyst durability is also an important indicator for evaluating its performance. Chronoamperometry test results show that the catalyst maintains stable performance during continuous operation at a current density of 24 mA cm−2 for 72 h. After a long-term i–t test, the measured LSV curve shows an obvious shift (Fig. 3f), which may be related to the detachment of active components, a common phenomenon observed in the field. This indicates that the prepared Ru–NiCuP@NF possesses good durability. Meanwhile, after 72 h of testing, the XPS spectra reveal an obvious shift in the binding energies of Ni 2p, Cu 2p, Ru 3d, and P 2p compared to their initial states (Fig. 2b–f). This is precisely due to the active phase evolution and electronic structure reconstruction on the catalyst surface during long-term electrolysis, leading to this phenomenon. Overall, Ru–NiCuP@NF as a working electrode in 1.0 M KOH solution not only exhibits excellent performance but also demonstrates good durability. When alkaline seawater was used as the electrolyte, LSV testing showed that Ru–NiCuP@NF achieved an HER overpotential of 120 mV at 10 mA cm−2, significantly better than NiCuP@NF (154 mV), RuCuP@NF (176 mV), and RuNiP@NF (165 mV) (Fig. 4a). Performance decreased slightly compared to 1.0 M KOH, likely due to competitive adsorption of Cl− ions from seawater onto catalyst active sites,42,43 occupying active centers and hindering H* and OH* intermediate adsorption. Moreover, although the highly alkaline 1.0 M KOH medium suppresses Mg(OH)2 and Ca(OH)2 precipitation, local pH fluctuations at the electrode interface during continuous alkaline seawater electrolysis may still induce trace hydroxide deposition, partially blocking active sites and increasing interfacial charge transfer resistance. Nevertheless, benefiting from the Ru-induced electronic structure modulation and hierarchical porous architecture, Ru–NiCuP@NF maintains excellent HER activity in alkaline seawater. The Cdl obtained from CV at different scan rates was 31.9 mF cm−2, much higher than that of control samples (Fig. 4c). The Tafel slope was 109.5 mV dec−1, significantly lower than those of NiCuP@NF (145.7 mV dec−1), RuCuP@NF (148.1 mV dec−1), and RuNiP@NF (146.3 mV dec−1) (Fig. 4b), indicating not only better performance but also a lower HER energy barrier for Ru–NiCuP@NF in alkaline seawater. EIS measurements indicate that the target catalyst has a similar charge transfer resistance to the other control groups (Fig. 4d). Long-term i–t tests showed a stable current density for the catalyst in alkaline seawater over 105 h of continuous electrolysis (Fig. 4f). (a) HER polarization curves of sample in alkaline seawater (b) corresponding Tafel plots (c) Cdl values (d) Nyquist plots (e) Bode phase plots of EIS measured for Ru–NiCuP@NF alkaline seawater at various applied potentials (f) LSV curves of Ru–NiCuP@NF in alkaline seawater before and after 105 h of continuous testing at a current density of 40 mA cm−2. Correspondingly, to evaluate the HER catalytic stability and kinetic characteristics of the Ru–NiCuP@NF electrocatalyst in alkaline seawater, its multi-potential EIS response was comparatively examined in a mixed system of 1.0 M KOH + natural seawater. The results show that the target catalyst maintains favorable frequency response characteristics in this medium (Fig. 4e). Compared with the pure 1.0 M KOH environment (Fig. 3e), although high-concentration ions (e.g., Cl−, Ca2+, Mg2+) are introduced from seawater, the main phase peak continuously shifts to higher frequencies as the applied potential is moved negatively from −1.026 to −1.203 V. This indicates that even in the complex seawater environment, the charge transfer rate at the electrode surface can still be effectively enhanced with increasing overpotential, maintaining excellent electrochemical reversibility. Within the same overpotential range, both the relative height of the phase angle peak and the frequency range are highly similar to those in the pure KOH system. These results confirm that the Ru doping strategy not only significantly enhances the intrinsic activity of NiCuP but also endows the catalyst with excellent anti-interference capability, allowing it to preserve its inherent HER mechanism and active site structure during seawater electrolysis, without the drastic deterioration of interfacial impedance caused by impurity adsorption. Oxygen evolution reaction measurements The OER performance of different samples in 1.0 M KOH was evaluated by LSV. The test results show that Ru–NiCuP@NF exhibits better OER performance than the other control samples. At a current density of 10 mA cm−2, its overpotential is 243 mV, while those of NiCuP@NF, RuCuP@NF, and RuNiP@NF are 247 mV, 266 mV, and 256 mV, respectively (Fig. 5a). This result indicates that Ru doping further enhances the OER activity of the nickel–copper based phosphide catalyst. In this alkaline medium, the Tafel slope of Ru–NiCuP@NF reaches 39.8 mV dec−1 (Fig. 5b), which is slightly higher than that of NiCuP@NF (35.1 mV dec−1), but much lower than those of RuNiP@NF (49.5 mV dec−1) and RuCuP@NF (74.8 mV dec−1), clearly indicating a lower OER energy barrier. In addition, by fitting the CV curves in the potential range of 1.35–1.45 V, the Cdl values were obtained and compared. It was found that the Cdl of Ru–NiCuP@NF (127.8 mF cm−2) is higher than those of RuNiP@NF (82.6 mF cm−2) and RuCuP@NF (111.2 mF cm−2) (Fig. 5c), indicating that it can expose more active sites. The EIS spectra show a small charge transfer resistance for the catalyst (Fig. 5d). As presented in the Bode plots of Fig. 5e, where impedance spectra were measured from 0.4 to 0.665 V, the low-frequency phase angle peak continuously decreases with increasing potential, indicating the onset of the OER. Meanwhile, continuous testing was conducted at a voltage of 0.8 V for 120 h, the catalyst exhibited exceptionally stable performance over the long-term operation. After long-term testing, the LSV curve showed a certain shift (Fig. 5f). Moreover, after durability testing, the XPS spectra revealed shifts in the binding energies of Ni 2p, Cu 2p, Ru 3d, and P 2p (Fig. 2b–f). (a) OER polarization curves of sample in 1.0 M KOH (b) corresponding Tafel plots (c) Cdl values (d) Nyquist plots (e) Bode phase plots of EIS measured for Ru–NiCuP@NF in 1.0 M KOH at various applied potentials (f) LSV curves of Ru–NiCuP@NF in 1.0 M KOH before and after 120 h of continuous testing at a constant voltage of 0.8 V. The same testing method was applied to evaluate OER performance in alkaline seawater. Ru–NiCuP@NF achieved an OER overpotential of 278 mV at 10 mA cm−2 (Fig. 6a), superior to NiCuP@NF (303 mV), RuCuP@NF (293 mV), and RuNiP@NF (311 mV). The Tafel slope for Ru–NiCuP@NF was 95.9 mV dec−1 (Fig. 6b), indicating a lower OER energy barrier. The Cdl of Ru–NiCuP@NF (207.2 mF cm−2) was higher than those of NiCuP@NF (37.3 mF cm−2), RuNiP@NF (24.3 mF cm−2), and RuCuP@NF (128.4 mF cm−2) (Fig. 6c), demonstrating that Ru doping significantly enhances active site exposure efficiency. EIS (Fig. 6d) revealed that Ru–NiCuP@NF had smaller charge transfer resistance (Rct), indicating faster OER kinetics. As shown in Fig. 6e, the Bode plots measured in the 1.0 M KOH + seawater system reveal that as the applied potential gradually increases from 0.466 to 0.746 V (vs. Hg/HgO), the impedance behavior of the catalyst exhibits significant changes. With increasing potential, the phase angle peak continuously decreases, and a single, clean peak shape is maintained throughout the entire tested potential range, without any additional shoulder peak in the mid-to-high frequency region. This strongly demonstrates that the catalyst possesses excellent selectivity for the OER in the alkaline seawater environment. Subsequently, to evaluate the operational stability of the target catalyst for the OER in different media, it was found that it could operate stably for 120 h in alkaline seawater at a voltage of 0.8 V, and the polarization curves before and after the test also exhibited a shift (Fig. 6f). (a) OER polarization curves of sample in alkaline seawater (b) corresponding Tafel plots (c) Cdl values (d) Nyquist plots (e) Bode phase plots of EIS measured for Ru–NiCuP@NF in alkaline seawater at various applied potentials (f) LSV curves of Ru–NiCuP@NF in alkaline seawater before and after 120 h of continuous testing at a constant voltage of 0.8 V (g) Stability test of Ru–NiCuP@NF in 1.0 M KOH at a constant voltage of 2 V for 120 h as well as LSV curves before and after the test. (h) Stability test of Ru–NiCuP@NF in alkaline seawater at a constant voltage of 2 V for 120 h as well as LSV curves before and after the test. Optimization of ruthenium doping content in Ru–NiCuP To further investigate the effect of ruthenium doping content on the electrocatalytic performance of Ru–NiCuP@NF catalysts, a series of Rux–NiCuP@NF catalysts with varying ruthenium doping amounts (x = 0.1, 0.3, and 0.4 mmol) were prepared. Electrochemical measurements revealed that Ru0.2–NiCuP@NF exhibited the optimal catalytic activity in 1.0 M KOH electrolyte. Specifically, it delivered a HER overpotential of merely 108 mV at a current density of 10 mA cm−2, outperforming Ru0.1–NiCuP@NF (133 mV), Ru0.3–NiCuP@NF (131 mV), and Ru0.4–NiCuP@NF (149 mV) (Fig. S8a and c). For the OER, an overpotential of 243 mV was achieved, which was also superior to those of Ru0.1–NiCuP@NF (250 mV), Ru0.3–NiCuP@NF (256 mV), and Ru0.4–NiCuP@NF (285 mV) (Fig. S8b and c). Similarly, in alkaline seawater electrolyte, Ru0.2–NiCuP@NF demonstrated remarkable catalytic performance at the identical current density, requiring only an HER overpotential of 120 mV, substantially lower than those of Ru0.1–NiCuP@NF (159 mV), Ru0.3–NiCuP@NF (141 mV), and Ru0.4–NiCuP@NF (149 mV) (Fig. S8d and f). Regarding the OER performance, although Ru0.3–NiCuP@NF exhibited a slightly lower overpotential of 238 mV, Ru0.2–NiCuP@NF displayed superior catalytic potential at high current densities (Fig. S8e and f). Based on the comprehensive evaluation of both HER and OER activities, the Ru0.2–NiCuP@NF catalyst with a ruthenium doping amount of 0.2 mmol was identified as the optimal candidate. Overall water splitting performance To evaluate the practical application potential of the Ru–NiCuP@NF electrocatalyst, a symmetric two-electrode overall water splitting cell was assembled using Ru–NiCuP@NF as both the cathode and the anode, and tested in 1.0 M KOH and alkaline seawater, respectively. For the 1.0 M KOH electrolyte, the cell voltage required to achieve a current density of 10 mA cm−2 is only 1.28 V (without ohmic correction). Meanwhile, Ru–NiCuP@NF was subjected to a continuous operation test at a constant voltage of 2 V for 120 h, and it was found that the cell voltage exhibited very low fluctuation, demonstrating excellent durability. After the long-term durability test, the polarization curves showed a certain shift (Fig. 6g). In alkaline seawater, due to the presence of high-concentration ions (e.g., Cl−, Ca2+, Mg2+) that may partially block active sites or increase ohmic resistance, the cell voltage required to achieve the same current density is slightly higher, at 1.64 V. Similarly, after 120 h of continuous testing, the cell voltage showed no significant fluctuation, and the polarization curves exhibited only a slight shift after the long-term test (Fig. 6h), further indicating that the as-prepared Ru–NiCuP@NF possesses excellent durability. Conclusion Based on the above analyses of phase, morphology, surface chemical state, and electrocatalytic performance, the mechanism by which Ru doping and low-temperature phosphorization. Ru doping can form a strong electronic synergy with Ni and Cu, causing electron deficiency on Ni and Cu atoms and lowering the reaction energy barrier. Benefiting from the above advantages, Ru–NiCuP@NF not only exhibits superior electrolytic activity but also demonstrates good durability. In alkaline and alkaline seawater media, the overpotentials for the HER at a current density of 10 mA cm−2 are 108 mV and 120 mV, respectively, and those for the OER are 243 mV and 278 mV, respectively. In an overall water splitting cell, cell voltages of only 1.28 V and 1.64 V are required to achieve a current density of 10 mA cm−2 in alkaline and alkaline seawater environments, respectively. This study provides a new design pathway for developing low-cost phosphide catalysts suitable for alkaline water electrolysis systems. This structure not only effectively prevents the detachment of active materials during long-term testing but also significantly enhances the long-term operational lifetime of the catalyst. Conflicts of interest There are no conflicts to declare. Supplementary Material