Synthesis and characteristics of the catalyst materials
A bottom-up route for the synthesis of nickel telluride and metal nitride heterostructures (NiCoN-NiTe and NiFeN-NiTe) is illustrated schematically in Figure 1c. Commercial Ni foam serves both as a supporting substrate and metal resource for the NiTe, and the nanorod arrays were uniformly prepared via a hydrothermal method (Figure 2a). The following in-situ electrochemical deposition process in nickel cobalt or nickel iron nitrate solution realizes the epitaxial growth of NiCo and NiFe hydroxide layers, respectively, with a leaf-like morphology on the NiTe surface. Energy dispersive spectroscopy (EDS) characterization indicated increased distributions of Fe, Co, and O, which proves the element introduction (Figure S1 in the Supporting Information). Afterwards, the as-prepared NiTe-NiFe(OH)x and NiTe-NiCo(OH)x precursors were then heated at 400 ℃ in an Ar + NH3 mixed-atmosphere flow to convert them to bimetallic nitrides. The scanning electron microscope (SEM) images and EDS mapping in Figure 2b, c and Figure S2, S3 show that the morphology of the three-dimensional (3D) nanoarray is maintained throughout, and the nitride shells packed with nanosheets are tightly anchored on the NiTe surface with a homogeneous distribution.
Further nanostructured characterization of the as-prepared samples was observed by transmission electron microscopy (TEM). The NiTe crystal plane in Figure S4 verifies the presence of nickel telluride. Meanwhile, both NiTe-NiCoN and NiTe-NiFeN show similar nanostructures with well-defined heterogeneous interfaces. In the case of the NiTe-NiCoN sample, the NiCo2N (200) and NiTe (111) facets can be distinctly observed to be in intimate contact (Figure 2d) with the lattice spacing of. 0.21 nm and 0.31 nm, respectively. The lattice fringes of NiTe-NiFeN in the high-resolution TEM (HRTEM) image of Figure 2e exhibit interplanar spacing of 0.31 nm and 0.19 nm, corresponding to NiTe (101) and Ni3FeN (111) planes, respectively. which is consistent with analysis of the apparent diffraction rings in the selected area electron diffraction (SAED) patterns (Figure S5), The corresponding high-angle annular dark field ‒ scanning TEM (HAADF-STEM) elemental mappings and line scan results in Figure 2f, 2g directly verify the presence of Ni, Fe, N and Ni, Co, N, as well as the internal enriched distribution of Te. All these characterizations together demonstrate the formation of heterostructured NiTe-NiCoN and NiTe-NiFeN with a delicate 3D core-shell nanostructure.
X-ray diffraction (XRD) was performed to characterize the crystal patterns of all the as-prepared samples. In Figure 3a, it is shown that, except for the large diffraction peaks of Ni, the peaks at 31.4°, 43.2°,46.2°, 56.7°, and 58.6° are completely matched with NiTe (JCPUS 89-2018). After ammoniation, the newly emerged peaks at 41.5°, 48.3°, and 70.6° strongly confirm the final formation of crystalline dual-phase NiFe3N (JCPUS 50-1434), while the peak at 41.7° of the as-prepared NiCo2N proves the same evolution. X-ray photoelectron spectroscopy (XPS) was conducted to clarify the surface chemical compositions and coordinated electronic environment (Figure 3b-f and Figure S6). All the survey spectra identify the co‐existence of Ni and Te, originating from NiTe. Fe, O, and N can be detected on NiTe-NiFeN, and Co, O, and N on NiTe-NiCoN, respectively. The atomic ratio of Ni:Fe:Te:N is ~3:31:20:25:21, and that of Ni:Co:Te:N is ~3:31:20:25:21, consistent with the SEM-EDS results but slightly different from the inductively coupled plasma (ICP) results (Table S1), why also imply the surficial distribution of N, Co, and Fe. The high-resolution Ni 2p spectra are shown in Figure 3b, NiTe exhibits two main peaks for Ni 2p3/2 and Ni 2p1/2, which are located at 856.3 and 874.1 eV, and peaks for emerging Ni-N species were identified at 852.4 and 852.3 eV after nitriding. 21,41 In the Te 3d spectra of Figure 3c, two peaks at 572.7 and 582.8 eV correspond to Te2- 3d5/2 and Te2- 3d3/2, and a pair of oxidation peaks were also observed, respectively. 36,42 The Co 2p spectra of NiTe-NiCoN in Figure 3d were deconvoluted into Co 2p3/2 Co0, Co2+, and Co3+ peaks at 778.6, 780.6, and 786.4 creV, respectively. 43,44 In the case of NiTe-NiFeN, the Fe 2p3/2 and Fe 2p1/2 peaks (Figure 3e) were located at 711.6 and 724.1 eV. 11,22 The above results verify the existence of Co and Fe with high valence in nitrides. The N 1s peak (Figure 3f) at 398.7 eV is ascribed to the metallic nitrides species, while the peaks at 401.3 eV and 400.6 eV correspond to the N-H bonding. 45 Significantly, compared with NiTe, a negative binding energy shift of both NiTe-NiCoN and NiTe-NiFeN is shown in the Ni 2p and Te 3d spectra, which appears because the interfacial electron redistribution leads to electron accumulation on the nitrides, generally considered to be beneficial for activating the electrocatalytic properties (details in Table S2). 34,36,46
X-ray absorption spectroscopy (XAS) was performed to further investigate the electronic structural properties of the heterostructures. In Figure 3g, the X-ray absorption near edge spectroscopy (XANES) spectra of the Ni K-edge near-edge structure curves show that the Ni oxidation species coexist in NiTe, NiTe-NiCoN, and NiTe-NiFeN. The NiTe-NiCoN and NiTe-NiFeN exhibit lower energy levels than NiTe, which will cause a low-level shift of the Ni d-band center and subsequently reduce the bonding strength of intermediates (H*, *OH, *O, and *OOH) on the catalyst surface, to the benefit of the reaction kinetics.36,37 Work-function (WF) tests were carried out by scanning Kelvin probe microscopy (Figure 3h). The calculated values of both NiTe-NiFeN (5.11 eV) and NiTe-NiCoN (5.12 eV) are lower than that of NiTe (5.22 eV), suggesting sufficient electron pathways across the heterogeneous interface.46 Subsequently, the liquid contact angle and gas evolution angle of different nano-array structures were measured and are compared in Figure S7. Evidently, NiTe-NiCoN and NiTe-NiFeN with the array structure have better immersion and gas release behavior, thereby ensuring stable solid-liquid-gas interfaces during the electrocatalytic process. In order to explore the role of tge heterogeneous interfaces in depth, we also prepared NiCoN and NiFeN samples without NiTe cores for comparison (labeled as NiCoN-NF and NiFeN-NF, respectively). The SEM images in Figure S8 show the flat morphology without NiTe support, and the EDS analysis confirms the uniform distribution of metals and nitrogen. The XRD spectra in Figure S8c characterize the construction of Ni3FeN and NiCo2N with results that are similar to those above. The subsequent electrocatalytic comparisons will visually demonstrate the role of heterogeneous structures.
Electrocatalytic hydrogen and oxygen evolution performance
The advantages of NiTe-NiCoN and NiTe-NiFeN (heterointerface effect, 3D nanostructure, bimetal synergy) offer promising electrocatalysis performance. The alkaline HER activity of NiTe-NiCoN was first evaluated by linear sweep voltammetry measurements (LSV) in a typical three-electrode system, where all the as-prepared samples and a commercial candidate (Pt/C) were also included for comparison. All the curves were corrected by iR compensation (85%). As shown in Figure 4a, NiTe-NiCoN exhibits remarkable activity with a small overpotential of 62 and 240 mV to deliver current densities of 10 and 500 mA cm-2, respectively, which is obviously better than those of NiTe (Ƞ10 = 274 mV) and NiCoN-NF (Ƞ10 = 90 mV), and comparable to that of Pt/C (Ƞ10 = 37 mV). The reaction kinetics of the HER determines the catalytic activity, which can be visualized intuitively through the derived Tafel slope. In Figure 4b, the calculated Tafel slope value of NiTe-NiCoN is 48 mV dec−1, which is relatively lower than for NiCoN-NF (95 mV dec−1) and NiTe (198 mV dec−1), indicating that the HER process follows a Volmer–Heyrovsky mechanism with optimized catalytic kinetics. For coupling with NiTe-NiCoN catalyst for efficient overall water electrolysis, we simultaneously examined the OER activity of the as-prepared NiTe-NiFeN and its control group. In Figure 4d, NiTe-NiFeN shows exceptional OER activity with a Ƞ10 of 211 mV and Ƞ500 of 300 mV, which is much better than for NiFeN-NF (Ƞ10 = 259 mV), NiTe (Ƞ10 = 339 mV), and IrO2 (Ƞ10 = 251 mV). Meanwhile, the relatively lowest Tafel slope of NiTe-NiFeN (39 mV dec-1) compared with NiFeN-NF (51 mV dec-1) and NiTe (124 mV dec-1) directly verifies the rapid OER kinetics of the heterogeneous interface and bimetal synergy (Figure 4f). It is worth mentioning that both NiTe-NiCoN and NiTe-NiFeN are also competitive in recent representative research work (Table S3), and further HER performance of NiTe-NiFeN and OER performance of NiTe-NiCoN prove the electrocatalytic pertinence (Figure S9).
Further electrochemical analysis were adopted to investigate the properties of all the as-prepared catalysts. The electrochemical surface area (ECSA) calculated from the double-layer capacitance (Cdl) was used to evaluate the density of active sites. In Figure S10, the derived values of NiTe-NiCoN and NiTe-NiFeN are 38.5 mF cm-2 and 37.4 mF cm-2, respectively, which are obviously higher than for NiCoN-NF (19.1 mF cm-2), NiFeN (29.3 mF cm-2), and NiTe (13.8 mF cm-2). The calculated ECSA values intuitively illustrate the enlarged active surface area achieved by the unique core-shell structure. Moreover, the normalized curves also exhibit the best activity of NiTe-NiCoN and NiTe-NiFeN, indicating that factors other than the active surface area are also criticial. Assuming that all the loaded metal ions functioned as positive active sites, the turnover frequency (TOF) values were then measured under the HER overpotential of 150 mV. The calculated TOF of NiTe-NiCoN is 0.314 s−1, which is obviously higher than those of NiTe (0.006 s−1), NiCoN (0.132 s−1), and NiTe-Ni3N (0.065−1). Under the overpotential of 300 mV, NiTe-NiFeN exhibits a higher TOF of 0.225 s−1 towards the OER than NiFeN-NF (0.082 s−1) and NiTe (0.005 s−1), suggesting improved intrinsic activity (Figure S11). Additionally, electrochemical impedance spectroscopy (EIS) measurements were conducted, and the Nyquist plots are shown in Figure S12. The corresponding fitting results in Table S4 demonstrate that in the HER, NiTe-NiCoN has a lower charge-transfer resistance (Rct) of 3.52 Ω than those of NiCoN-NF (6.47 Ω) and NiTe (53.68 Ω). Similarly, in the OER, NiTe-NiFeN also shows the lowest Rct of 4.64 Ω compared with the other samples (NiFeN-NF: 8.85 Ω, NiTe: 38.4 Ω), as expected, which can be ascribed to the faster electron transport during the HER and OER processes. Impressively, the integrated radar charts of Figure 4c and 4f, taken together, exhibit the advantages of both NiTe-NiCoN and NiTe-NiFeN, evidently demonstrating the enhanced comprehensive properties that benefit from the rational design of the bimetallic composition as well as the suitable heterostructures.
Moreover, an electrocatalytic operating durablity assessment was performed by chronopotentiometric testing under different current densities in 1 M KOH electrolyte. As displayed in Figure 4g and 4h, the best candidates, NiTe-NiCoN and NiTe-NiFeN. exhibited ultra-long stability at 10, 20. 50, 100, and 500 mA cm-2 without significant attenuation after 60 h of operation. In Figure S13, the polarization curves after the stability tests also showed well preserved activity, and the collected SEM images comfirmed the good structural stability. Afterwards, the XPS spectra shown in Figure S14 confirmed the composition integrity. Notablely, NiTe-NiCoN and NiTe-NiFeN exhibit obviously decreased surface N and Te distributions, while the emergence of an -OH peak in the O 1s spectrum at 532.1 eV, accompanying by the hydroxide evolution of the nickel-based components, identifies the surface reconstruction evolving the Ni-based species (NiCo-OH and NiFe-OOH).47,48 Overall, the above-mentioned analyses prove the successful construction of heterostructured NiTe-NiCoN and NiTe-NiFeN catalyst with promising performance in both the HER and the OER.
Overall seawater electrolysis performance
In order to overcome the constraints of scarce freshwater resources and expand the environmentally-friendly process of water electrolysis, we then investigated the HER and OER performance of our samples in a simulated seawater electrolysis system (1 M KOH + 0.5 M NaCl) and in a real seawater electrolysis system (1 M KOH+ natural seawater, only after standing still for 12 h) (Figure S15). Under the chloride-rich conditions, the corrosion polarization curves (Figure S16) demonstrated that the nitride coating could significantly reduce the corrosion current, which improves the Cl-corrosion resistance to better adapt to the seawater system. LSV measurements were then performed and are shown in Figure 5a, where NiTe-NiCoN and NiTe-NiFeN still exhibit HER and OER activity comparable to thosepure alkaline solution and have better activity than the other samples (Figure 5b and Figure S17), and demonstrate excellent catalytic performance when working together. The decline of about 50 mV in the HER and OER in alkaline seawater is due to the hindrance of insoluble impurities (e.g., Ca(OH)2, Mg(OH)2) (Figure S18).8,49 Encouragingly, a non-membraneous two-electrode electrolyzer was built to achieve overall seawater-alkaline electrolysis, which employted the desirable anodic candidates NiTe-NiFeN and NiTe-NiCoN, respectively, to promote the concept of sustainable hydrogen production driven by solar/wind power (Figure 5c). In Figure 5d, the polarization curves of the NiFeN-NiTe || NiCoN-NiTe couple exhibit the same excellent overall activity in the pure alkaline electrolytes with 1.57 V and 1.75 V required to produce 100 and 500 mA cm-2, respectively, while it also shows similar performance in alkaline simulated seawater electrolyte. It is noteworthy that this couple required a cell voltage of 1.65 V to deliver 100 mA cm-2, and 1.84 V at 500 mA cm-2 in 1 M KOH + seawater electrolyte with IR compensation. Compared with recently reported overall electrolysis performances (Figure 5d), the as-constructed couple could be an ideal candidate for efficient hydrogen production.8,13,50-56 The electrocatalytic performance of NiFeN-NiTe || NiCoN-NiTe in further high temperature testing (60℃) shows significantly improved catalytic activity, demonstrating the positive response of reactivity activity to temperature changes, while also suggesting a new direction for heat utilization (Figure S19). A self-installed solar-powered system was assembled with an output voltage of 1.9 V (Figure S20a and Video S1) and actuated by a wind-powered system at about 1.7 V (Figure S20b and Video S2). All of these systems realized electrolysis of the seawater system, demonstrating that these coupled systems have the potential to utilize renewable energy and provide intermittent storage (solar/wind energy to hydrogen energy).
In Figure 5e, the fabricated electrolysis system (with or without the seawater medium) drives 10 mA cm-2, 100 mA cm-2, and an industrial current density of 500 mA cm-2 to investigate the long-term chronopotentiometric response and judge the stability of the catalyst for industrial applications. It can be easily seen that both systems have ultra-long stability of more than 200 h and that the attenuation is within 5%. The calculated Faradaic efficiency (FE) of hydrogen evolution in the simulated seawater full-cell, as measured by both gas chromatography (GC) and drainage gas collection, was approximately 98 % (Figure 5f), and the volume ratio of the two-pole product was 2:1(Figure S21a). Subsequently, it was detected by ultraviolet-visible spectroscopy (UV-VIS) that no ClO- was produced after the anode reaction (Figure S21b), which confirms the effective inhibition of chlorine oxidation. All these results taken together revealed the excellent energy efficiency and high selectivity in the electrolysis process, confirming the feasibility of the system-integration concept to achieve industrial-scale and environmentally friendly hydrogen production.
Electrocatalytic mechanism analysis
Inspired by this remarkable electrocatalytic performance, density functional theory (DFT) calculations were performed to theoretically reveal the relationship between the synergistic effects of the heterostructure and the intrinsic electrocatalytic activity of both NiTe-NiCoN and NiTe-NiFeN. Generally, the alkaline HER consists of the initial catalyst acting on the H2O, the intermediate H* adsorption, and the final H2 desorption. The water oxidation pathway undergoes four coordinated electron transfer steps and generates a series of key oxygenated species (OH*, O*, and OOH*, where the active sites are labeled as *).57-59 Critically, the appropriate atomic sites and interface states determine the reduced thermodynamic barriers. The optimized kinetic models and the adsorbed configurations of the samples to be studied (NiTe, Ni3FeN, NiTe-NiFeN, NiCo2N, and NiTe-NiCoN) on the heterojunction surfaces are shown in Figure S22. For the calculated H2O adsorption and the intermediate adsorption energy of H*, NiTe-NiCoN displays a higher ΔEH2O of -0.61 eV than the NiTe value of -0.11 eV, indicating the strengthened H2O adsorption for improved water dissociation (Figure S23).60 In Figure 6a, the middle value of the Gibbs free energy, ΔGH* ≈ 0.41 eV, is better than for NiTe and NiCo2N, illustrating the favorable H* adsorption kinetics during the HER. In the case of the OER, the Ni site of NiTe undergoes the third electrochemical step (*O → *OOH) with the largest barrier of 2.32 eV, which is the rate-determining step (RDS) that results in slow kinetics. In contrast, the RDS of Ni3FeN (O*→OOH*) has a barrier of 1.82 eV on the major Ni sites. In the case of the interfacial NiTe-NiFeN, a significantly lower RDS barrier of 1.71 eV illustrates the optimized OER activity, which is mainly derived from the stronger adsorption of OH* and the favorable stabilization of intermediates (Figure 6b). To further explore the laws of adsorption energy change, the density of state (DOS) with the orbitals for different elements was employed for systematic analysis. In Figure 6c, the heterojunction d-band center is higher than those of NiTe, NiCoN, and NiFeN at the Fermi level, implying that the heterogeneous interface leads to a more rapid electron exchange at the catalyst–intermediate interface, as well as enhanced electron transfer ability, the same as in the impedance analysis. In the simulated electron density distribution (Figure 6d, 6e), obvious electron transfers from NiCo2N to NiTe and from Ni3FeN to NiTe were examined that were consistent with the above XPS phenomenon, further indicating the enriched electron and hole distributions on metal nitrides. The charge distribution between Ni-N-Co and Ni-N-Fe also show a tendency for electrons to move away from the Ni site and thus gain greater freedom, which favors the optimization of the adsorption energy modulation as well as surface –OOH reconstruction (Figure S24).
Previous reports have demonstrated the influence of electric field changes that were guided by nanostructures on the catalytic activity of the cathode (e.g., the CO2 reduction reaction (CO2RR) and the HER). As a semi-reaction that limits the efficiency of water electrolysis, the OER performance, structural characteristics, and electric field distribution should be similarly related, and worth systematic exploration. In Figure S25, the surface electric field distributions (measured by atomic force microscopy (AFM) and Kelvin probe force microscopy) of the as-prepared NiTe-NiFeN with different calcining temperatures (400 and 475 °C, respectively) together demonstrated that the heterostructured electrocatalyst with uneven surface structure enlarges the charge accumulation. We constructed a simulated electrochemical model of heterostructured nickel telluride and nickel-based nitride based on the TEM and AFM morphology to reveal their influence on anodic catalytic behavior by the finite-element numerical method (details in the Supporting Information). Obviously in Figure S26a, the core-shell structure with nanosheet wrapping results in a more uniform and stronger current response than the smooth rod surface, which means that the active site is fully mobilized. The calculated diffusion flux under the two models indicates that the increase in current density promotes the more rapid concentration of hydroxide radicals on the electrode surface (Figure S26b). The rich and high-curvature convex structure will generate a local electrostatic field environment, thereby changing the ion migration and distribution states in the electrocatalytic process, which will have a certain impact towards improved OER activity as well as selective anode reactions (OER and CER). Therefore, we comparatively studied the influence of different electric fields on the ion distribution in the Helmholtz layer. It can be seen in Figure 6f and Figure S27 that under the condition of high field strength generated by high curvature, the "hot spot" effect near the electrode area is particularly significant and has more charge accumulation, which results in enhanced local field strength and changes in the anion concentration. Furthermore, the linear relationship between the position of the Helmholtz layer and the electric field intensity, as well as with the ion concentration (Figure 6g) directly proves the stronger difference in anion concentration near the electrode surface in the high curvature structure. The OH- concentration is twice as high as that of Cl- (0.2 mol cm-3 compared to 0.1 mol cm-3 at 0.3 V), but the simulated Helmholtz layer under different potentials results in the same phenomena for both (Figure S28). OH- radicals have a more rapid transfer process, which makes the “hot spot” near the electrode area appear in the hydroxide-rich state and occupy more active sites to ensure rapid oxygen evolution reaction kinetics, as shown in the inset of Figure 6g. At the same time, this state can also hinder the adsorption of chloride ions, thereby inhibiting the CER as well as the chlorine corrosion,61. The competition-inhibition effect produced by the nanostructure provides a guarantee of an efficient and stable seawater electrolysis process (Figure 6h, Figure S29). To sum up, competitive theoretical simulation analyses demonstrate the synergetic effect of the heterostructure and the nanostructure advantage, which together determine the increased electrocatalytic performance.