English

• Hydrogen (H₂) from overall water splitting (OWS) based on electrolysis, since its zero-carbon emission, superior energy density, and extensive source reserve, is generally recognized as one of the promising and renewable energy carriers for replacing fossil fuels [1–3]. Since existing commercial OWS systems are largely dependent on limited freshwater, using ocean seawater, which accounts for 96.5% of the Earth’s total water stored, for hydrogen production presents a fundamental advantage [4,5]. Nevertheless, there are two significant issues with this seawater electrolysis that urgently require a solution: the oxygen evolution reaction (OER, 1.23 V versus reversible hydrogen electrode (vs. RHE)) possesses complicated four-electron transfer kinetics leading to high overpotentials and extra electricity expense [6]; especially at high voltages and currents requirements, about 0.5 mol/L Cl^- in seawater easily undergo competitive chlorine oxidation reaction (ClOR, 1.71 V vs. RHE), generating poisonous Cl₂ or ClO^- that cause corrosion of electrode materials [7,8]. Note that the substitution of thermodynamically favorable, easily oxidized, and value-added organic small molecule oxidation, including urea, alcohols, glucose, hydrazine, and 5-hydroxymethylfurfural, etc., for sluggish OER has been identified as is a promising alternative strategy to reduce energy consumption and enable energy-efficient hydrogen generation [9,10]. Among them, the sulfion oxidation reaction (SOR) demonstrates a low oxidation potential of −0.48 V vs. RHE, which also allows the extraction of valuable solid sulfur from wastewater containing S^2- or H₂S ions under ambient conditions, making it stand out [11]. Therefore, the coupling assistance system via the integration of anodic SOR and cathodic hydrogen evolution reaction (HER) would likely achieve both industrial-scale hydrogen production and transformation/degradation of sulfur-based pollutants, while the development of advanced, affordable, and efficient electrocatalysts is the key to guarantee the stable functioning of this technology.

  To date, with the exception of precious metal materials (e.g., Pt, Ru, and Ir), some non-precious transition metal electrocatalysts dependent on cost-effectiveness, catalytic activity, and long-life have been reported in abundance, such as Mo-based (e.g., MoS₂, Mo₂C/MoC, MoP, and MoO₃) and Ni-based (e.g., Co-Ni₃S₂–2, CoNi@NGs, and NiS₂) materials are non-ignorable categories available for HER and SOR, respectively [12–14]. In terms of Mo-based materials, it behaves with a d-band electronic structure similar to that of Pt to optimize the H^* adsorption. For example, Wu et al. found that the reducing conditions transformed nitrogen-doped Mo nanoclusters on Ti₃C₂T_X into six-coordinated monoatomic Mo structures, exposing more active sites and thus enhancing HER catalytic efficiency [15]. Hou et al. discovered that Mo-Co_0.85Se_VSe/NC nanosheets exhibit strong catalytic activity towards alkaline HER because of their optimized electronic structure, leading to the creation of enriched electronic Co site and rapid water dissociation [16]. Nonetheless, in the Ni₄Mo alloy catalyst, the in situ dissolution and polymerization of Mo formed new alkaline HER activity promotion of MoO₄^2- and Mo₂O₇^2- [17]. In another ingenious design of microstructured nickel-based electrocatalysts, the nanoparticle NiMo alloy core is encapsulated by an N-doped carbon layer shell (NiMo-5%@NC), in which the confinement effect of the carbon shell reasonably lowers the binding energy with oxygen-containing species [18]. Several nanocatalysts composed of Ni have been successfully constructed that are capable of sulfion oxidation at low potentials. For instance, as an efficient SOR catalyst, the Ni-Co-S/NF in situ converted into highly amorphous Ni, Co-based sulfides within the electrocatalytic reactions, presenting enriched defects and unsaturated sites [19]. To circumvent the long-confounding problem of solid sulfur passivation in SOR, Zhang and colleagues proposed a self-cleaning NiS₂ electrode exhibiting sulfophobic phenomena, i.e., its poor interaction with sulfur species [20]. Despite these successful attempts mentioned above, HER/SOR catalysts typically exhibit severe activity loss and unsatisfactory kinetics owing to the complex composition of natural seawater, the corrosive nature of alkali additives, the passivation of solid sulfur products, etc.

  Here, we present a novel chemical dealloying and following sulfurization approach of constructing two kinds of nanoporous alloy ligament, i.e., nanoporous NiMo alloy (np-NiMo) and Mo-doped NiS₂ compound (np-NiMo-S), which serve as electrocatalytic HER (cathode) and OER/SOR (anode) materials, respectively. As expected, using np-NiMo and np-NiMo-S as electrocatalysts, compared with the conventional overall seawater splitting system (1.547 V), the coupling alkaline seawater electrolysis system combining the HER with the SOR requires only 0.532 V total voltage to obtain 50 mA/cm², enabling energy-saving and high-efficiency hydrogen generation and sulfur recovery. Further research has demonstrated that this particular nanoporous ligament structure and the uniform development of nanosheets on its surface lead to more accessible electrochemical catalytic sites and that improved kinetic pathways along the interpenetrating nanopore channels facilitate electron transfer and mass transport. Theoretical studies further verified that the AlNi₃/Al₅Mo heterostructure in np-NiMo and the Mo doping of np-NiMo-S are the origin of the activity that leads to their excellent HER and SOR, respectively. These results offer valuable insights into the creation and application of high-activity electrocatalysts for economical, sustainable, and efficient green hydrogen production from seawater.

  The steps involved in the synthesis are described in Fig. 1, Al₈₉Ni₁₀Mo₁ ribbons, which are a precursor material, were prepared by an arc melting and melt spinning technique. Then, the Al₈₉Ni₁₀Mo₁ undergoes dealloying and sulfurization operations in succession, the resulting np-NiMo and np-NiMo-S were utilized for cathodic HER and anodic OER/SOR, respectively.

  Figure 1

  

  Figure 1.  Schematic diagram of the synthetic process of np-NiMo and np-NiMo-S.

  DownLoad: Full-Size Img PowerPoint

  Representative scanning electron microscopy (SEM, Figs. 2a and f) and transmission electron microscopy (TEM, Figs. 2b and g) images of dealloyed np-NiMo and sulfured np-NiMo-S displayed continuous interconnected ligaments network and abundant lamellar nanopore channels. With the exception of minor sintering following the sulfidation treatment, the magnified TEM images (Figs. 2c and h) further demonstrated the well-dispersed nanosheets on the surface of each ligament and that their nanosheet structure was mostly well preserved. The apparent diffraction rings consisting of numerous discrete spots, as shown in the selected area electron diffraction (SAED) patterns (inset of Figs. 2c and h), which suggested the polycrystalline nature of these two nanoligaments for np-NiMo and np-NiMo-S [21]. And the enlarged SAED patterns further reveal the AlNi₃, Al₅Mo, and NiS₂ phases (Figs. S1 and S2 in Supporting information). In the high-resolution TEM (HRTEM) image (Fig. 2d) of np-NiMo, the well-defined interplanar spacings are spaced at 0.215 and 0.230 nm, with each consistent with the (111) and (111) lattice planes of AlNi₃ and Al₅Mo crystalline, and the presence of abundant heterogeneous interfaces in the AlNi₃/Al₅Mo heterostructure are also found. The corresponding energy-dispersive X-ray (EDX) elemental mappings exhibit the co-existing of Ni, Mo, Al, and O, along with their even distribution (Fig. 2e). Furthermore, as for np-NiMo-S’ HRTEM image (Fig. 2i), the predominant (220) plane of NiS₂ possesses a measured D-spacing of 0.207 nm, combining uniformly distributed Ni, Mo, Al, and S, which confirms the successful transformation from an alloy compound to corresponding metal sulfide. The lamellar nanopore channels and continuous ligament networks of two kinds of np-NiMo and np-NiMo-S naturally encourage the exposure of many active sites for the activation and adsorption of species like reactants and important intermediates [22].

  Figure 2

  

  Figure 2.  (a) SEM image, (b, c) TEM images (inset of c: SAED pattern), (d) HRTEM image, and (e) the EDX elemental mapping images for np-NiMo. (f) SEM image, (g, h) TEM images (inset of h: SAED pattern), (i) HRTEM image, and (j) the EDX elemental mapping images for np-NiMo-S.

  DownLoad: Full-Size Img PowerPoint

  The XRD characterization of those materials was adopted to investigate their crystal phases in Fig. 3a and Fig. S3 (Supporting information). The Al₈₉Ni₁₀Mo₁ alloy ribbons demonstrate a multiphase nanostructure, which consists of the Al phase (PDF #04–0787), the corresponding oxides Al₂O₃ (PDF #04–0877), and intermetallic compounds: AlNi₃ (ICSD No. 98–060–8811) as well as Al₅Mo (PDF #00–025–1132). After dealloying, the Al phase of np-NiMo disappears due to selective corrosion by KOH, while the unstable Al component remaining AlNi₃ and Al₅Mo is partially oxidized to Al₂O₃ in air, resulting in the formation of a porous heterostructure. The diffraction peaks of the np-NiMo-S catalyst were attributed to the NiS₂ phase (PDF #11–0099), verifying that the sulfurization achieves the conversion to Mo-doped NiS₂. XRD patterns of NF and its derived NiS_x are shown in Figs. S4 and S5 (Supporting information), individually. To further probe their elemental composition, chemical states, and bonding configuration, X-ray photoelectron spectroscopy (XPS) measurement was employed. Fig. 3b shows the high-resolution Ni 2p XPS spectra of np-NiMo, in which two minor peaks emerging at 852.6 and 869.7 eV belong to metallic Ni. The two other groups of peaks centered on 855.8/873.7 and 857.2/875.6 eV are attributed to Ni²⁺ and Ni³⁺ species, respectively, as well as two associated satellite peaks at 862.0 and 880.2 eV, which is attributed to the slight surface oxidation. After incorporating sulfur element into the np-NiMo alloy, the enhanced peak intensity of Ni²⁺ and the disappearance of Ni³⁺ peaks are observed, and Ni 2p peaks of the np-NiMo-S show a positive shift (~0.5 eV for Ni²⁺, 0.8 eV for Ni⁰), indicating an increase in the oxidation state of Ni [23,24]. In Fig. 3c of Mo 3d_5/2 region of np-NiMo, the three deconvoluted peaks with binding energies of 227.8, 230.1, and 232.4 eV, ascribed to Mo⁰, Mo⁴⁺, and Mo⁶⁺, respectively. Similar to the case of Ni 2p, the Mo 3d peaks are also shifted to higher binding energy (~0.7 eV for Mo⁴⁺, 0.4 eV for Mo⁶⁺), and, importantly, a new S 2s peak appeared at 226.8 eV, while no metallic Mo was found [25]. As seen in Fig. 3d, before and after sulfur introduction, the main Al³⁺ peaks are at 74.1 and 74.4 eV, respectively, and the remaining peaks (68.0 and 71.0 eV; 68.1 and 71.5 eV) are attributed to the presence of Ni 3p [26]. As for the detailed O 1s spectrum (Fig. 3e, np-NiMo and np-NiMo-S), which could be deconvoluted into lattice oxygen (O_L, 530.5 and 530.9 eV), oxygen vacancies (O_V, 531.6 and 531.9 eV), and adsorbed water molecules (532.7 and 533.4 eV) [27]. For the high-resolution S 2p spectra of np-NiMo-S (Fig. 3f), sulfur components are deconvoluted into five peaks: the binding energies at 161.6 eV (2p_3/2) and 162.7 eV (2p_1/2) for S^2- of Mo-S bondings similar to natural MoS₂, the peaks at 162.6 and 163.8 eV are assigned to the S₂^2- 2p_3/2 and S₂^2- 2p_1/2 of Ni-S bondings, respectively, whereas the other peak with a higher binding energy of 167.9 eV could be oxidized SO₄^2- species exposed to air [28,29]. On the other hand, compared to np-NiMo, it can be clearly observed that the valence states of np-NiMo-S with the introduction of the sulfur component, including Ni 2p, Mo 3d, Al 2p, and O 1s, are all positively shifted towards higher bonding energies to varying degrees, suggesting S-induced partial electron transfer from Ni/Mo/O atoms to S and enhanced electronic interactions between them [30]. According to those analysis results, it would be concluded that the preparation of np-NiMo (AlNi₃/Al₅Mo heterostructures) and np-NiMo-S (Mo doped NiS₂) was achieved successfully.

  Figure 3

  

  Figure 3.  (a) XRD patterns of np-NiMo, and np-NiMo-S. The high-resolution XPS spectra of (b) Ni 2p, (c) Mo 3d, (d) Al 2p, (e) O 1s, and (f) S 2p for np-NiMo and np-NiMo-S.

  DownLoad: Full-Size Img PowerPoint

  Using a three-electrode configuration, the as-prepared NF, NiS_x, np-NiMo, np-NiMo-S, PtC, and RuO₂ samples were examined in an alkaline seawater solution (1 mol/L KOH + seawater) to evaluate their HER and OER electrocatalytic properties. For HER assessment of np-NiMo, as shown in linear sweep voltammetry (LSV) polarization curves (Fig. 4a), the cleaned NF, as a conductive substrate, delivers the largest overpotential, demonstrating its inferior HER activity and ensuring minimal interference. Nevertheless, the np-NiMo displays an exceptionally low overpotential of just 134 and 178 mV, offering the geometric current densities of 50 and 100 mA/cm², respectively, which are even much superior to that of NiS_x (230 and 260 mV), NF (247 and 281 mV), and np-NiMo-S (258 and 297 mV), closing to PtC (41 and 75 mV) and excellent HER electrocatalysts reported in recent years (Table S1 in Supporting information). The polarization curves without iR compensation also support these results (Fig. S6 in Supporting information). Accordingly, in the polarization curves yielded Tafel plots (Fig. 4b), the small slope value of 96.1 mV/dec for np-NiMo among all electrolytic samples (135.3 mV/dec for NiS_x, 154.3 mV/dec for NF, 161.4 mV/dec for np-NiMo-S, and 62.8 mV/dec for PtC) was discovered, suggesting the favorable HER kinetics, Volmer-Heyrovsky mechanism, and a Heyrovsky rate-determining step [31,32]. The charge transfer resistance (R_ct) from Nyquist plots showed similar trends as those of LSV data (Fig. 4c), that is, the np-NiMo performs lower R_ct values (2.36 Ω) over the NiS_x (6.88 Ω), NF (9.48 Ω), and np-NiMo-S (11.15 Ω), except for PtC (1.41 Ω), emphasizing faster charge transfer rate and more active sites of np-NiMo among them for H⁺ reduction [33,34]. Besides, under 50 mA/cm², its catalytic stability measurement through Chronopotentiometry (CP) was also conducted, and the applied potential (−0.231 V → −0.289 V vs. RHE) almost remains nearly unchangeable after 11 h (Fig. 4g). Two possible explanations cause of the observed improvement in HER capability of np-NiMo throughout the electrocatalytic process: AlNi₃/Al₅Mo heterostructure, continuous ligament networks with abundant nanopore channels, these factors are conducive to promoting water dissociation, separating the generated H₂, and avoiding gas blockages [35,36].

  Figure 4

  

  Figure 4.  HER performance of PtC, np-NiMo, NiS_x, NF, and np-NiMo-S: (a) LSV curves, (b) Tafel plots, (c) Nyquist plots, and (g) CP stability curves. OER performance of RuO₂, np-NiMo-S, NiS_x, np-NiMo, and NF: (d) LSV curves, (e) Tafel plots, (f) Nyquist plots, and (h) CP stability curves.

  DownLoad: Full-Size Img PowerPoint

  In the meantime, the OER catalytic activity of the as-acquired np-NiMo-S sample was also studied. From the LSV curves (Fig. 4d), commercial RuO₂ catalyst is applied as a comparison and possessed the lowest overpotential of 264 and 306 mV to deliver 50 as well as 100 mA/cm², respectively. The np-NiMo-S required only 314 and 368 mV, which were much smaller than those of NiS_x (379 and 427 mV), np-NiMo (455 and 528 mV), and NF (497 and 542 mV). And the np-NiMo-S is also equally good when the polarization curves are not iR-compensated or LSV measurements are performed from the positive direction scanning (Figs. S7 and S8 in Supporting information). Compared to NiS_x (126.2 mV/dec; 3.10 Ω), np-NiMo (137.0 mV/dec; 6.46 Ω), and NF (153.5 mV/dec; 6.63 Ω), the Tafel slope and R_ct value of np-NiMo-S (94.8 mV/dec; 2.55 Ω) is the smallest besides RuO₂ (65.1 mV/dec; 1.94 Ω), implying that its oxygen production rate and charge transfer kinetics are faster (Fig. 4f) [37,38]. During the whole CP measurement in Fig. 4h, the potential-time curves from 1.604 V to 1.580 V vs. RHE of np-NiMo-S were relatively stable for 11 h at 50 mA/cm², and its crystal structure remained basically unchanged (Fig. S9 in Supporting information). The double-layer capacitance (C_dl) of np-NiMo and np-NiMo-S is significantly higher than that of NF and its sulfides of NiS_x, suggesting that they have the highest number of active reactive sites exposed per unit area and the key role of Mo addition (Figs. S10 and S11 in Supporting information) [39,40]. Notably, NiMo alloy and its sulfides have opposite electrocatalytic behavioral properties towards HER and OER, and their turnover frequency (TOF) is higher than the other three control samples in their respective applications, confirming the superior intrinsic catalytic activity of these two materials (Figs. S12 and S13 in Supporting information) [41]. Hence, in addition to the structural advantages of nanoporous ligaments that both have, Mo-doped NiS₂ of np-NiMo-S is mainly responsible for the OER against the responsibility of dealloyed np-NiMo with AlNi₃/Al₅Mo heterostructure for HER.

  With the aim of evaluating the electrocatalytic SOR, the previous alkaline seawater solution was replaced with an alkaline seawater electrolyte containing Na₂S (1 mol/L KOH + 1 mol/L Na₂S + seawater). Notably, the presence of the reducing Ni-S bond in NiS₂ can effectively promote the transfer of electrons from sulfur species to the catalyst [42]. In this sense, during sulfurization, the np-NiMo precursor developed a new np-NiMo-S with Mo-doped NiS₂ structure, and it is anticipated that np-NiMo-S exhibits strong SOR activity. As expected, Fig. 5a shows that with the addition of Na₂S, the np-NiMo-S is able to drive the SOR at a current density of 50 and 100 mA/cm² with a potential of 0.364 and 0.486 V vs. RHE, which are below that of the comparison samples, for example, RuO₂ (0.392 and 0.518 V), NiS_x (0.588 and 0.647 V), NF (0.616 and 0.692 V), and np-NiMo (0.647 and ~0.720 V), and this result is consistent with the polarization curves without iR compensation (Fig. S14 in Supporting information). At the same time, the SOR catalytic activity of the np-NiMo-S electrode is comparable to or even better than that of other published work (Table S2 in Supporting information). In contrast, without Na₂S, a slow OER requires overcoming 4.2 and 3.3 times higher potential (1.544 and 1.598 V) under the same operating current densities in Fig. 4d. Meantime, the Tafel slopes over SOR and OER of np-NiMo-S catalyst, as indicated in Fig. S15 (Supporting information) and Fig. 4e, were 32.2 and 94.8 mV/dec, respectively, suggesting that the OER was less efficient in mass transfer and that the SOR had faster reaction kinetics [43,44]. And, the interfacial charge transfer process was analyzed by Nyquist plots (Fig. S16 in Supporting information), the np-NiMo-S had the lowest R_ct of all the catalysts studied, indicating a fast electron transfer rate at the catalyst/electrolyte interface [45,46]. The high TOF value (8.55 s^-1) of np-NiMo-S highlights its ability to promote SOR at higher efficiency (Fig. S17 in Supporting information) [47]. We determined the stability of np-NiMo-S via CP at 50 mA/cm² for 11 h (Fig. S18 in Supporting information), in which the required operating potential decreases gradually with increasing electrolysis time, indicating its robust SOR activity. This enhanced SOR performance might be explained by the stronger oxide species generated on the catalyst surface and the catalyst’s weak interaction with sulfur products, which could keep the surface clean during SOR cycling and confirm the self-cleaning effect that prevented electrode passivation [20].

  Figure 5

  

  Figure 5.  (a) SOR LSV curves for np-NiMo-S, RuO₂, NiS_x, NF, and np-NiMo. (b) The voltage gap (ΔV) between HER (1 mol/L KOH + seawater) of np-NiMo and OER (1 mol/L KOH + seawater) or SOR (1 mol/L KOH + 1 mol/L Na₂S + seawater) of np-NiMo-S. (c) Comparison of LSV curves between HER||SOR with HER||OER seawater splitting systems assembled by np-NiMo cathode and np-NiMo-S anode. (d) Digital photo of the HER||SOR seawater splitting system. (e) CP stability test of HER||SOR seawater splitting systems. (f) Faradaic efficiency measurements for HER. (g) The UV–vis spectra of electrolytes of SOR anode at different reaction times (inset: the corresponding optical photo). (h) XRD pattern of the collected sulfur product (inset: the corresponding optical photo).

  DownLoad: Full-Size Img PowerPoint

  Encouragement by the excellent HER and SOR electrocatalytic characteristics of np-NiMo and np-NiMo-S, respectively, using SOR instead of OER is expected to be a workable approach for producing valuable hydrogen with not much energy consumption (Fig. 5b). To this end, we assembled two kinds of two-electrode HER||SOR and HER||OER seawater electrolysis system in Fig. 5d, including H-cell with cation-exchange membrane, np-NiMo cathode for HER (1 mol/L KOH + seawater), and np-NiMo-S anode for SOR (1 mol/L KOH + seawater + 1 mol/L Na₂S) or OER (1 mol/L KOH + seawater). As shown in Fig. 5c, when the thermodynamically superior SOR was substituted instead of the conventional OER coupled to HER, the necessary cell voltage could be drastically dropped from 1.547, 1.709, and 1.830 V to 0.532, 0.633, and 0.771 V, with drive current densities of 50, 100, and 200 mA/cm², respectively, which means energy savings of up to about 60% [48]. And SOR-based coupled electrolysis systems are equally advantageous when the LSV profiles are not iR-compensated (Fig. S19 in Supporting information). Then, the stability of the HER||SOR coupling electrolysis system was measured. During the operation of 34 h at a current density of 50 mA/cm², the CP curve was relatively stable (from ~0.639 V to ~0.702 V) after replacing the anode electrolyte with a fresh one every 11 h (Fig. 5e), whereas the higher voltage of the HER||OER system rose from 1.721 V to 1.783 V within the 24 h period (Fig. S20 in Supporting information). The observed increase in applied potential during the stability tests may be attributed to the complex composition of natural seawater, including Cl^-, Na⁺, Mg²⁺, Ca²⁺, bacteria/microorganisms, and small particulate impurities that can interfere with electrochemical processes [49]. In particular, the presence of Cl^- corrodes most metal-based electrocatalysts, leading to activity damage and insufficient stability of the electrocatalysts under high current operating conditions [50]. Moreover, the np-NiMo and np-NiMo-S in HER||SOR coupled system was characterized by XRD after 34 h operation and it was found that the corresponding characteristic diffraction peaks were still present compared to before electrolysis (Figs. S21 and S22 in Supporting information). For cathodic HER, the Faraday efficiency of the produced hydrogen was determined, which is close to 100% (Fig. 5f and Fig. S23 in Supporting information). As for another anodic SOR, the electrolyte gradually changes from colorless to light yellow until dark yellow over time. This color change belongs to the gradually increasing concentration of short-chain polysulfide intermediates (S₄^2- to S₆^2-), and three clear UV–vis absorption peaks at 315, 370, and 425 nm confirmed their existence (Fig. 5g). The appearance of shifted absorption peaks near 315 nm may be attributed to the partially overlapping wavelength regions of the generated S₄^2- and S₆^2- polysulfide ions, as well as to the different concentrations of S₄^2- and S₆^2- polysulfide anions at different reaction times [51,52]. The reacted anode electrolyte was acidified, centrifuged, and dried sequentially, and the resulting precipitate was characterized by XRD and confirmed to be a sulfur powder (Fig. 5h). According to the aforementioned findings, a two-electrode HER||SOR coupling device composed of np-NiMo and its derivative np-NiMo-S sulfide not only produces hydrogen at the cathode with energy efficiency but also recovers sulfur ions at the anode in an ecologically benign manner.

  In order to elucidate the nature of AlNi₃/Al₅Mo heterostructures in np-NiMo with high HER activity and Mo dopant in np-NiMo-S in enhancing the SOR activity, density functional theory (DFT) simulations were carried out for calculating the reaction barrier [53]. For comparison, the np-NiMo, consisting of AlNi₃ (111)/Al₅Mo (111), and two other control samples, Al₅Mo (111) facet, and AlNi₃ (111) facet, were chosen as computational models. The calculated density of states (DOS, Fig. 6a, Figs. S24 and S25 in Supporting information) indicates a strong interaction between the 4d orbitals of Mo atoms and the 3d orbitals of Ni atoms in np-NiMo on the neighborhood of the Fermi energy level, and the transfer of more charge from Al₅Mo to AlNi₃ after the formation of the heterostructure, leading to good electrical conductivity and strong electronic interactions (Fig. S26 in Supporting information) [54]. In alkaline HER (Fig. 6b), H₂O is sequentially adsorbed on the active site, dissociated into reactive *H, and combined to form H₂. It can be found that the H₂O adsorption process of three crystal structures is thermodynamically spontaneous, including np-NiMo (−1.07 eV), Al₅Mo (−2.07 eV), and AlNi₃ (−1.07 eV), which is favorable for subsequent water splitting and hydrogen-coupling. As for them, the H₂O dissociation is the potential-determining step (PDS), while the np-NiMo only needs to overcome the lower energy barrier of 1.53 eV compared to Al₅Mo (2.40 eV) and AlNi₃ (2.63 eV). And np-NiMo further exhibits a moderate strength of ^*H adsorption close to 0 eV, implying that it has a better hydrogen evolution ability [55,56].

  Figure 6

  

  Figure 6.  (a) DOS diagrams of np-NiMo and np-NiMo-S. (b) Gibbs free energy diagram of np-NiMo for HER. (c) The calculated charger density differences of adsorbed S^2- on NiS₂ and np-NiMo-S surface. (d) Gibbs free energy diagram of np-NiMo-S for SOR.

  DownLoad: Full-Size Img PowerPoint

  On the other hand, the electrocatalytic properties of SOR have also been investigated. Compared to NiS₂, Mo-doped np-NiMo-S has enhanced electron states near the Fermi energy level for improving the electronic conductivity (Fig. 6a and Fig. S27 in Supporting information). The charge density difference further demonstrates that the np-NiMo-S substrate has more electron transfer (0.15 e^-) to the S atoms compared to the NiS₂ (200) facet (0.12 e^-), suggesting that the Mo sites have a significant chemisorption of S^2- and are more prone to SOR triggering (Fig. 6c) [48,57]. Fig. 6d displayed the Gibbs free energy variation of the SOR pathway (S^2- → ^*S → ^*S₂ → ^*S₃ → ^*S₄ → ^*S₈ → S₈). PDS is the oxidation of short-chain to long-chain sulfur (^*S₄ to ^*S₈), which leads to the enormous and quick accumulation of ^*S₄ as the main intermediate in an anodic electrolyte. This outcome is consistent with what is seen when SOR is performed on np-NiMo-S (Fig. 5g). Among them, the np-NiMo-S shows the lowest energy barrier (1.64 eV) for the rate-limiting step relative to NiS₂ (2.18 eV), indicating the highest activity for propelling SOR. In addition, it is spontaneous in the desorption step of ^*S₈, implying an inherent property of inhibiting sulfur passivation to mitigate the coverage of the reactive site by S₈ products [13,20].

  In summary, we have fabricated two interconnected alloy ligament networks with abundant lamellar nanopores, i.e., np-NiMo and np-NiMo-S, for high-efficiency hydrogen preparation as well as environmentally friendly desulfurization and sulfur recovery, respectively. As compared to the conventional HER||OER system, experimental results demonstrated that the SOR-assisted alkaline seawater splitting system could lower energy investment by about 60%. This superior catalytic activity is partly attributed to the advantages of their nanoporous ligament structure, which facilitates the exposure of more active sites, improved mass transfer/diffusion rates, and avoidance of gaseous hydrogen/solid sulfur blockage. On the other hand, theoretical calculations further show that the AlNi₃/Al₅Mo heterostructure in np-NiMo has strong electronic interactions, which are favorable for the adsorption of ^*H₂O and its dissociation, while Mo doping of NiS₂ in np-NiMo-S promotes the adsorption of S^2- and conversion to long-chain polysulfide intermediates (^*S₄ to ^*S₈). The energy-efficient hydrogen production from alkaline seawater and sulfur recovery from sulfur-rich wastewater reported in this work meet the human strategy of protecting the environment and building a green economy in response to climate change and is expected to move towards a safer and more sustainable pathway.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influencethe work reported in this paper.

  CRediT authorship contribution statement

  Miaosen Yang: Data curation. Junyang Ding: Data curation. Zhiwei Wang: Software. Jingwen Zhang: Investigation. Zimo Peng: Validation. Xijun Liu: Supervision, Conceptualization.

  Acknowledgments

  This work was financially supported by the Guangxi Natural Science Fund for Distinguished Young Scholars (No. 2024GXNSFFA010008), the Natural Science Foundation of Jilin Province of China (No. 20240101098JC), and the National Natural Science Foundation of China (No. 22469002).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.110861.

  
  
• 1. [1]

     P. Halder, M. Babaie, F. Salek, et al., Renew. Sust. Energ. Rev. 199 (2024) 114543. doi: 10.1016/j.rser.2024.114543

  2. [2]

     A.A. Feidenhans’l, Y.N. Regmi, C. Wei, et al., Chem. Rev. 124 (2024) 5617–5667. doi: 10.1021/acs.chemrev.3c00712

  3. [3]

     F. Cheng, X. Peng, L. Hu, et al., Nat. Commun. 13 (2022) 6486. doi: 10.1038/s41467-022-34278-6

  4. [4]

     L. Zhang, Z. Wang, J. Qiu, Adv. Mater. 34 (2022) 2109321. doi: 10.1002/adma.202109321

  5. [5]

     L. Chen, C. Yu, J. Dong, et al., Chem. Soc. Rev. 53 (2024) 7455–7488. doi: 10.1039/d3cs00822c

  6. [6]

     C. Lei, S. Lyu, J. Si, et al., ChemCatChem 11 (2019) 5855–5874. doi: 10.1002/cctc.201901707

  7. [7]

     J. Liu, S. Duan, H. Shi, et al., Angew. Chem. Int. Ed. 61 (2022) e202210753. doi: 10.1002/anie.202210753

  8. [8]

     J. Ding, Z. Peng, Z. Wang, et al., J. Mater. Chem. A 12 (2024) 28023–28031. doi: 10.1039/d4ta05277c

  9. [9]

     H. Yu, W. Wang, Q. Mao, et al., J. Mater. Chem. A 11 (2023) 2218–2224. doi: 10.1039/d2ta08730h

  10. [10]

      C. Yin, F. Yang, S. Wang, et al., Chin. J. Catal. 51 (2023) 225–236. doi: 10.1016/S1872-2067(23)64490-0

  11. [11]

      W. Wang, Q. Mao, S. Jiang, et al., Appl. Catal. B: Environ. 340 (2024) 123194. doi: 10.1016/j.apcatb.2023.123194

  12. [12]

      Y. Li, S. Zuo, Q. Li, et al., Appl. Catal. B: Environ. 318 (2022) 121832. doi: 10.1016/j.apcatb.2022.121832

  13. [13]

      M. Yang, T. Wei, C. Zeng, et al., Chem. Eng. J. 498 (2024) 155799. doi: 10.1016/j.cej.2024.155799

  14. [14]

      Z. Xiao, C. Lu, J. Wang, et al., Adv. Funct. Mater. 33 (2023) 2212183. doi: 10.1002/adfm.202212183

  15. [15]

      P.F. Wu, Y.Q. Yang, H.Y. Xi, et al., Small 20 (2024) 2306716. doi: 10.1002/smll.202306716

  16. [16]

      Q. Dai, L. Wang, K. Wang, et al., Adv. Funct. Mater. 32 (2022) 2109556. doi: 10.1002/adfm.202109556

  17. [17]

      W. Du, Y. Shi, W. Zhou, et al., Angew. Chem. Int. Ed. 60 (2021) 7051–7055. doi: 10.1002/anie.202015723

  18. [18]

      F. Yang, S. Lu, Y. Feng, et al., ACS Catal 14 (2024) 2324–2332. doi: 10.1021/acscatal.3c06058

  19. [19]

      T. Gao, X. Tang, X. Li, et al., Inorg. Chem. Front. 10 (2023) 1447–1456. doi: 10.1039/d2qi02629e

  20. [20]

      S. Zhang, Q. Zhou, Z. Shen, et al., Adv. Funct. Mater. 31 (2021) 2101922. doi: 10.1002/adfm.202101922

  21. [21]

      H.O. Ayoola, S.D. House, C.S. Bonifacio, et al., Acta Mater. 182 (2020) 257–266. doi: 10.1016/j.actamat.2019.10.027

  22. [22]

      Y. Zhang, Z. Li, K. Chen, et al., Adv. Energy Mater. 14 (2024) 2402309. doi: 10.1002/aenm.202402309

  23. [23]

      L. Li, H. Qiu, Y. Zhu, et al., Appl. Catal. B: Environ. 331 (2023) 122710. doi: 10.1016/j.apcatb.2023.122710

  24. [24]

      J. Cao, K. Wang, J. Chen, et al., Nano-Micro Lett. 11 (2019) 67. doi: 10.1007/s40820-019-0299-4

  25. [25]

      X. Huang, H. Xu, D. Cao, et al., Nano Energy 78 (2020) 105253. doi: 10.1016/j.nanoen.2020.105253

  26. [26]

      A.P. Sinha, T.S. Thomas, D. Mandal, Energy Storage Mater 63 (2023) 102988. doi: 10.1016/j.ensm.2023.102988

  27. [27]

      T. Tian, L. Huang, L. Ai, et al., J. Mater. Chem. A 5 (2017) 20985–20992. doi: 10.1039/C7TA06671F

  28. [28]

      L. Zeng, Z. Liu, K. Sun, et al., J. Mater. Chem. A 7 (2019) 25628–25640. doi: 10.1039/c9ta08030a

  29. [29]

      A. Ambrosi, M. Pumera, ACS Catal. 6 (2016) 3985–3993. doi: 10.1021/acscatal.6b00910

  30. [30]

      G. Qian, Y. Wang, L. Li, et al., Adv. Funct. Mater. 34 (2024) 2404055. doi: 10.1002/adfm.202404055

  31. [31]

      J.S. Tian, Y.C. Hu, W.F. Lu, et al., Carbon Energy 5 (2023) e322. doi: 10.1002/cey2.322

  32. [32]

      L.Y. Ozer, C. Garlisi, H. Oladipo, et al., J. Photochem. Photobiol. C Photochem. Rev. 33 (2017) 132–164. doi: 10.1016/j.jphotochemrev.2017.06.003

  33. [33]

      H. Cai, Z. Wang, G. Meng, et al., Inorg. Chem. 63 (2024) 20935–20939. doi: 10.1021/acs.inorgchem.4c03881

  34. [34]

      P. Liu, Y. Liu, K. Wang, et al., Nano Res 17 (2024) 7957–7966. doi: 10.1007/s12274-024-6799-7

  35. [35]

      X. Liu, M. Chen, J. Ma, et al., China Powder Sci. Technol. 30 (2024) 35–45.

  36. [36]

      C. Yuan, H. Lv, Y. Zhang, et al., Carbon 206 (2023) 237–245. doi: 10.1016/j.carbon.2023.02.022

  37. [37]

      D. Aegerter, E. Fabbri, N.S. Yüzbasi, et al., ACS Catal. 13 (2023) 15899–15909. doi: 10.1021/acscatal.3c04138

  38. [38]

      H.S. Lee, H. Shin, S. Park, et al., Joule 7 (2023) 1902–1919. doi: 10.1016/j.joule.2023.06.018

  39. [39]

      K. Chen, D. Ma, Y. Zhang, et al., Adv. Mater. 36 (2024) 2402160. doi: 10.1002/adma.202402160

  40. [40]

      F. Cheng, Z. Li, L. Wang, et al., Mater. Horiz. 8 (2021) 556–564. doi: 10.1039/d0mh01757d

  41. [41]

      X. Liu, J. He, S. Zhao, et al., Nat. Commun. 9 (2018) 4365. doi: 10.1038/s41467-018-06815-9

  42. [42]

      Y. Li, Y. Duan, K. Zhang, et al., Chem. Eng. J. 433 (2022) 134472. doi: 10.1016/j.cej.2021.134472

  43. [43]

      Y. Ji, Z. Yu, L. Yan, et al., China Powder Sci. Technol. 29 (2023) 100–107. doi: 10.3847/1538-3881/acad84

  44. [44]

      D. Liu, T. Liu, L. Zhang, et al., J. Mater. Chem. A 5 (2017) 3208–3213. doi: 10.1039/C6TA11127K

  45. [45]

      H. Wu, G. Fei, X. Gao, et al., China Powder Sci. Technol. 29 (2023) 70–80.

  46. [46]

      T. Wang, Q. Zhang, K. Lian, et al., J. Colloid Interf. Sci. 655 (2024) 176–186. doi: 10.1016/j.jcis.2023.10.157

  47. [47]

      S. Cho, Y. Oh, H.T. Nguyen, et al., Int. J. Hydrogen Energy 90 (2024) 747–756. doi: 10.1016/j.ijhydene.2024.10.039

  48. [48]

      Y. Pei, D. Li, C. Qiu, et al., Angew. Chem. Int. Ed. 63 (2024) e202411977. doi: 10.1002/anie.202411977

  49. [49]

      H. Xie, Z. Zhao, T. Liu, et al., Nature 612 (2022) 673–678. doi: 10.1038/s41586-022-05379-5

  50. [50]

      X. Chen, Y. Duan, H. Dou, et al., Chem. Eng. J. 499 (2024) 155925. doi: 10.1016/j.cej.2024.155925

  51. [51]

      J. Wang, M. Zhou, R. Fu, et al., Adv. Funct. Mater. 34 (2024) 2315326. doi: 10.1002/adfm.202315326

  52. [52]

      J. Häcker, D.H. Nguyen, T. Rommel, et al., ACS Energy Lett. 7 (2022) 1–9. doi: 10.1021/acsenergylett.1c02152

  53. [53]

      P.E. Blöchl, Phys. Rev. B 50 (1994) 17953–17979. doi: 10.1103/PhysRevB.50.17953

  54. [54]

      J. Zhang, X. Shang, H. Ren, et al., Adv. Mater. 31 (2019) 1905107. doi: 10.1002/adma.201905107

  55. [55]

      C. Cheng, W. Ao, H. Ren, et al., Appl. Catal. B: Environ. 331 (2023) 122681. doi: 10.1016/j.apcatb.2023.122681

  56. [56]

      S. Ma, B. Yu, B.Y. Xia, et al., Sci. China Mater. 67 (2024) 752–761. doi: 10.1007/s40843-023-2762-9

  57. [57]

      X. Li, W. Yu, Y. Wang, et al., Chin. Chem. Lett. 35 (2024) 109166. doi: 10.1016/j.cclet.2023.109166

• 

  Figure 1  Schematic diagram of the synthetic process of np-NiMo and np-NiMo-S.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) SEM image, (b, c) TEM images (inset of c: SAED pattern), (d) HRTEM image, and (e) the EDX elemental mapping images for np-NiMo. (f) SEM image, (g, h) TEM images (inset of h: SAED pattern), (i) HRTEM image, and (j) the EDX elemental mapping images for np-NiMo-S.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) XRD patterns of np-NiMo, and np-NiMo-S. The high-resolution XPS spectra of (b) Ni 2p, (c) Mo 3d, (d) Al 2p, (e) O 1s, and (f) S 2p for np-NiMo and np-NiMo-S.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  HER performance of PtC, np-NiMo, NiS_x, NF, and np-NiMo-S: (a) LSV curves, (b) Tafel plots, (c) Nyquist plots, and (g) CP stability curves. OER performance of RuO₂, np-NiMo-S, NiS_x, np-NiMo, and NF: (d) LSV curves, (e) Tafel plots, (f) Nyquist plots, and (h) CP stability curves.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  (a) SOR LSV curves for np-NiMo-S, RuO₂, NiS_x, NF, and np-NiMo. (b) The voltage gap (ΔV) between HER (1 mol/L KOH + seawater) of np-NiMo and OER (1 mol/L KOH + seawater) or SOR (1 mol/L KOH + 1 mol/L Na₂S + seawater) of np-NiMo-S. (c) Comparison of LSV curves between HER||SOR with HER||OER seawater splitting systems assembled by np-NiMo cathode and np-NiMo-S anode. (d) Digital photo of the HER||SOR seawater splitting system. (e) CP stability test of HER||SOR seawater splitting systems. (f) Faradaic efficiency measurements for HER. (g) The UV–vis spectra of electrolytes of SOR anode at different reaction times (inset: the corresponding optical photo). (h) XRD pattern of the collected sulfur product (inset: the corresponding optical photo).

  下载: 全尺寸图片 幻灯片
  

  Figure 6  (a) DOS diagrams of np-NiMo and np-NiMo-S. (b) Gibbs free energy diagram of np-NiMo for HER. (c) The calculated charger density differences of adsorbed S^2- on NiS₂ and np-NiMo-S surface. (d) Gibbs free energy diagram of np-NiMo-S for SOR.

  下载: 全尺寸图片 幻灯片
•