Atomically Dispersed Platinum Modulated by Sulfide as an Efficient Electrocatalyst for Hydrogen Evolution Reaction

Abstract Catalytically active metals atomically dispersed on supports presents the ultimate atom utilization efficiency and cost‐effective pathway for electrocatalyst design. Optimizing the coordination nature of metal atoms represents the advanced strategy for enhancing the catalytic activity and the selectivity of single‐atom catalysts (SACs). Here, we designed a transition‐metal based sulfide‐Ni3S2 with abundant exposed Ni vacancies created by the interaction between chloride ions and the functional groups on the surface of Ni3S2 for the anchoring of atomically dispersed Pt (PtSA‐Ni3S2). The theoretical calculation reveals that unique Pt‐Ni3S2 support interaction increases the d orbital electron occupation at the Fermi level and leads to a shift‐down of the d ‐band center, which energetically enhances H2O adsorption and provides the optimum H binding sites. Introducing Pt into Ni position in Ni3S2 system can efficiently enhance electronic field distribution and construct a metallic‐state feature on the Pt sites by the orbital hybridization between S‐3p and Pt‐5d for improved reaction kinetics. Finally, the fabricated PtSA‐Ni3S2 SAC is supported by Ag nanowires network to construct a seamless conductive three‐dimensional (3D) nanostructure (PtSA‐Ni3S2@Ag NWs), and the developed catalyst shows an extremely great mass activity of 7.6 A mg−1 with 27‐time higher than the commercial Pt/C HER catalyst.


Introduction
The noble-metals based catalysts play significant roles in renewable energy utilization, storage, and conversion, [1,2] and have been widely used to drive the various catalytic reactions DOI: 10.1002/advs.202100347 including nitrogen reduction reaction, [3,4] hydrogen evolution reaction (HER), [5,6] oxygen evolution reaction, [7,8] and the oxygen reduction reaction. [9,10] However, the high cost caused by scarcity hinders the widespread application of these catalysts. Single-atom catalysts (SACs), with atomically distributed noble-metal on supports equipped with extremely high specific activity and significantly large atom utilization efficiency, have attracted considerable attention. [11] The electrochemical activity of single-atom sites highly depends on the chemical and physical properties of the SACs, and specifically, the microenvironments of the single atom, including geometric construction and electronic coordination are extremely critical in determinating the performances of the catalysts. [12,13] Therefore, exploring and understanding the structure-activity relationship of the singleatom in SACs is imperative from the aspects of materials synthesis and reaction mechanism.
For the SACs, the strong interaction between the metal atoms and support is required to avoid the formation of aggregated clusters and particles by considering the stability. [14,15] However, the strong metal-support interaction will result in a large charge loss of metal atoms, and gift a cation-like positive electrons states to the metal atom. [16,17] These high-valence metal atoms are more favorable for electro-oxidation reactions rather than electroreduction processes due to the less d orbital electrons of SACs involved in the reaction. [13] Recently, the metal atoms with lowcoordinated electrons transfer and the close-to-zero valence state resembling a metallic-state atom were designed by anchoring Pt on the specific site of N-C framework and exhibited a superior activity for electroreduction reactions. [18] Whereas, the application of such metallic-state atoms is embarrassed due to the unaccessible fabrication approach and unsatisfied durability. Further, the local electron distribution around metal sites and d-band position of SACs is sensitive to the electron interaction between the metal atom and support, [16,17] and especially, the locally enhanced electron density on the single-metal sites induced by appropriate metal-support interaction has been expected to efficiently accelerate the catalytic reaction kinetics toward HER. [19,20] Nevertheless the development of such unique electronic field distribution feature in SACs are extremely sluggish due to the lack of suitable support. Therefore, it is challenging but significant to fabricate the metallic-state single atoms in the catalyst with the optimized d orbital structure and locally enhanced electronic field distribution by increase the local electron density on the metal atoms for satisfactory HER by considering the stability. Apart from the electronic structure factors, the catalytic activity of SACs is also extremely limited by low exposure efficiency of metal atoms sites and inefficient electrons supply ability due to the introduction of insulating binder (e.g., Nafion and polytetrafluoroethylene) in electrode framework, [13,14,19,21,22] leading to the inefficient conversion of the adsorbed hydrogen (M-H ads ) to H 2 in the Heyrovsky or Tafel step. [23][24][25][26][27] Interesting, architectural nanostructure engineering, in the non-noble metal catalysts field, provides an appealing platform to realize a seamlessly conductive system by integrating active materials on current collectors. [28,29] Inspired by the above, the integration of electronic structure adjustment strategy and architectural nanostructure engineering may provide some new avenues and insights for the design of highly efficient SACs.
Here, by combining theoretical simulation techniques, we design a transition-metal sulfide-Ni 3 S 2 , with abundant exposed Ni vacancies created by the interaction between chloride ions and the functional groups on the surface of Ni 3 S 2 , to anchor atomically dispersed Pt for efficiently electrocatalytic hydrogen evolution, Specifically, the Pt atoms undergo a redistribution in the local electronic structure after coordinating with Ni 3 S 2 support and lead to a locally enhanced electronic field distribution by increasing the electron density on the top of Pt sites. Meanwhile, metallic-state single Pt atoms are obtained by the orbital hybridization between S-3p and Pt-5d after introducing Pt into Ni position in the Ni 3 S 2 system for improved reaction kinetics. Moreover, Pt-Ni 3 S 2 support interaction increases the d orbital electron occupation at the Fermi level and leads to a shift-down of the d-band center, which energetically enhances H 2 O adsorption and provides the optimum H binding sites. By considering the superiority of architectural nanostructure, Ni 3 S 2 toiled Pt single atom is integrated on Ag nanowires network (Pt SA -Ni 3 S 2 @Ag NWs) to form a seamlessly conductive nanostructure for the more facile generation and release of H 2 . As such, the designed Pt SA -Ni 3 S 2 @Ag NWs demonstrate outstanding HER performance with 27-fold higher mass activity than the commercial Pt/C catalyst and long-term durability with 5000 cycles or 30 h.

Results and Discussion
First, theoretical investigations were carried out to elucidate the hidden mechanisms of Ni 3 S 2 tailored Pt single atom on the electronic structure and HER process of catalyst. According to the optimized crystal structures ( Figure S1, Supporting Information), the formation energy of Pt immobilized at surface Ni positions in Ni 3 S 2 is −0.33 eV, significantly lower than that at S positions (7.83 eV), bottom Ni site (−0.23 eV), and surface S-S bridge site (−0.26 eV) in Ni 3 S 2 , suggesting the preference of surface cation occupancy of incorporated Pt. From the charge density difference analysis, the replacement of Ni site by Pt atom causes the redistribution of the local electrons around the Pt site by metal-support coordination due to the different electronegativity among Pt (2.2), S (2.5), and S (2.5) (Figure 1a (Figure 1c), which is more favorable for boosting HER kinetics. [19] The electrons redistribution could affect the partial density of states (PDOS) of the SACs. Compared with the Ni 3 S 2 system, the total DOS ()of single-atom Pt immobilized Ni 3 S 2 show a higher electronic occupied state near the Fermi level ( Figure 1d; and Figure S4, Supporting Information), leading to the higher intrinsic electrical conductivity and larger carrier concentration. Specifically, the improved DOS of the Pt SA -Ni 3 S 2 near the Fermi level mainly derives from the contribution of the Pt 5d orbitals electrons ( Figure S5, Supporting Information), suggesting the strong orbital hybridization between S-3p and Pt-5d after introducing Pt into Ni position in the Ni 3 S 2 system. Besides, the Pt-5d band in Ni 3 S 2 shows a substantially wide range for overlapping with H-1s and H 2 O-2p orbitals (Figure 1e). Meanwhile, the surface Ni-3d band also exhibits optimal overlapping with H-1s and H 2 O-2p orbitals. Therefore, the anchored single-atom Pt site plays a protecting role for stabilizing the Ni valence state against corrosion, and also acts as a distributary role for reducing the deactivation of reaction sites by binding H and O species under over-binding of intermediates on the Nisites. [30] Further, the reactants-support interaction in Pt SA -Ni 3 S 2 system was explored by performing the d-band center model, and the Pt SA -Ni 3 S 2 shows the lowest d-band center (−1.40 eV, Figure 1f)) comparing with that of Ni 3 S 2 (−1.22 eV) and V Ni -Ni 3 S 2 (−1.26 eV), which could efficiently reduce the adsorption energy and meanwhile promote the desorption ability for H atoms on the catalyst surface. [31,32] Besides, the metal-support interactions could be quantified by performing Bader charge analysis of the Pt atom (Figure 1g), and the charge transfer (absolute value of Bader charge value) between Pt atom and Ni 3 S 2 support is only 0.017 e, much smaller than 0.044 e for Pt foil and ever lower than that of reported SACs recently (Table S2 (Supporting Information), 0.188 e for Pt SA in CoSe 2−x , [33] 0.23 e for Pt SA in C 1 N 1 , [13] and 0.47 e for Pt SA in Co(OH) 2 ), [20] which suggests a metallic electron state for Pt atom anchoring in Ni 3 S 2 support, leading to an extremely active catalytic property for HER. [13,18] From all of the above, the increased electrons occupation near Femi level, locally enhanced electrons density distribution on the top of Pt sites and shift-down of the d-band center deriving from the metallic-state Pt induced by the orbital hybridization between S-3p and Pt-5d after introducing Pt into Ni position in Ni 3 S 2 system are expected to optimize the reactant binding energy and promote the reaction kinetics for HER.
Thus, based on the optimized electronic structure of Pt SA -Ni 3 S 2 as the above analysis, the potential HER activity is predicted. Generally, the HER pathway consists of three main steps comprising the initial water adsorption H 2 O*, the generated intermediate H* as well as final molecular H 2 ( Figure 1h). The free energy (∆G H* ) of hydrogen atom adsorption is a reasonable descriptor to estimate HER activity for various catalysts. According to the density functional theory (DFT) calculation, the ∆G H* of Pt SA -Ni 3 S 2 (−0.06 eV) are closer to 0 eV than that of Ni 3 S 2 (0.49 eV), Pt foil (−0.10 eV), and recently reported SACs (Table S2, Supporting Information), benefitting from the promoted electrons occupation near Femi level, down-shift d-band center as well as local-enhance electrons density on Pt site due to the orbital hybridization between S-3p and Pt-5d after introducing Pt into Ni position in Ni 3 S 2 system as above discussion. In the alkaline condition, the water molecular splitting kinetics from the Volmer step (step 2 in Figure 1h) determine the overall HER rate. Based on the DFT simulation, Pt SA -Ni 3 S 2 shows more favorable H 2 O adsorption with −0.67 eV comparing with Ni 3 S 2 , Pt foil, and recently reported SACs (Table S2, Supporting Information), and the improved H 2 O adsorption on Pt SA -Ni 3 S 2 site derives from the more d electrons as the real active centers of catalyst contributed by metallic-state Pt 5d orbital ( Figure 1g; and Figure S5, Supporting Information). [13] Thus, transition-metal based sulfide-Ni 3 S 2 immobilized Pt single atom provides an efficient approach to modulate the electronic structure of SACs and is predicted to show the most outstanding HER catalytic activity.
Experimentally, we designed Ni 3 S 2 immobilized Pt singleatom catalyst with a 3D cable-shape structure on cloth fiberloaded Ag NWs support (Pt SA -Ni 3 S 2 @Ag NWs) by a four-step route illustrated in Figure 2a, in which Pt atoms immobilized at Ni positions in Ni 3 S 2 is realized by Ni vacancies creation and subsequent Pt single atom anchoring. In detail, metal Ni was electrodeposited on Ag NWs by chronoamperometry method (Figure S6, Supporting Information). Then, the Ag NW supported Ni was converted into Ni 3 S 2 (Ni 3 S 2 @Ag NWs) by the selective sulfuration process in Thioacetamide (TAA) aqueous solution. Finally, the formation of Ni vacancies and single-atom Pt Fabrication and structural characterizations of Pt SA -Ni 3 S 2 @Ag NWs. a) Schematic illustration of the in-situ preparation of Ni 3 S 2 phase via sulfuration reaction of metallic Ni on Ag nanowires substrate followed by the electrochemical anchoring of single-atom Pt. b) XRD patterns of bare Ag NWs, Ni@Ag NWs, and Pt SA -Ni 3 S 2 @Ag NWs. c) SEM images of Pt SA -Ni 3 S 2 @Ag NWs. d) HRTEM image of Pt SA -Ni 3 S 2 and the corresponding FFT pattern of Pt SA -Ni 3 S 2 @Ag NWs in insert of c). e,f) Atomic-resolution HAADF-STEM images of Pt SA -Ni 3 S 2 . g) The simulated structure corresponding to the STEM image of Pt SA -Ni 3 S 2 in e). h) The illustration of the DFT-optimized structure of Pt SA -Ni 3 S 2 @Ag NWs in f). i-l) EDS element mapping of Pt SA -Ni 3 S 2 @Ag NWs.
immobilization was performed by an electrochemical process with 400 cyclic voltammetry between 0 and −0.4 V in 1 m KOH solution with 50 × 10 −6 m PtCl 6 2− . The X-ray diffraction (XRD) patterns (Figure 2b) confirm the phase evolution of metallic Ni to Ni 3 S 2 , and the diffraction peaks of Pt SA -Ni 3 S 2 @Ag NWs mainly derive from Ni 3 S 2 and metallic Ag. The absence of Pt characteristic peaks implies the singleatom distribution of Pt instead of nanoparticles or clusters in Ni 3 S 2 . Scanning electron microscopy (SEM, Figure 2c; and Figure S7, Supporting Information) suggests the densely packed Pt SA -Ni 3 S 2 nanoparticles uniformly grow on Ag NWs to form a seamless contact nanoarchitecture via sulfuration and subsequent electrochemical process, which shows a different appearance with bare Ag NWs ( Figure S8, Supporting Information) and Ni coated Ag NWs ( Figure S9, Supporting Information). Further, the Pt SA -Ni 3 S 2 @Ag NWs were characterized by transmission electron microscopy (TEM), and a distinct cable-shape configuration is presented ( Figure S10a, Supporting Information). Be-sides, interconnected pores were observed in the Pt SA -Ni 3 S 2 shell layer ( Figure S10b, Supporting Information), which is favorable for hydrogen generation and release. The high-resolution TEM (HRTEM, Figure 2d) image of Pt SA -Ni 3 S 2 @Ag NWs demonstrates well-resolved lattice fringes with 0.20 and 0.24 nm, corresponding to the exposed (202) and (003) crystallographic plane of Ni 3 S 2 , respectively. The derived fast Fourier transform (FFT) image (the inset in Figure 2d) also shows the related crystallographic plane of (101), (003), (202), and (015) from the Ni 3 S 2 phase, suggesting the surface coupling of Pt-support does not change the crystal structure of Ni 3 S 2 .
The high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) image presents uniformly dispersed bright spots on the Ni 3 S 2 substrate (Figure 2e  atoms are fixed in Ni 3 S 2 . Subsequently, a STEM simulation based on the DFT-optimized structure was performed to verify the microstructure of single-atom Pt immobilized Ni 3 S 2 (Figure 2g,h). The simulated result suggests that the single Pt atoms were anchored in the sites of Ni vacancies by binding with the nearest S atoms in Ni 3 S 2 support. The corresponding energy-dispersive spectroscopy (EDS) mapping analysis corroborates well the uniform dispersion of Pt atoms throughout the Ni 3 S 2 (Figure 2i-l; and Figure S11, Supporting Information). While increasing the electrodeposition cyclic voltammetry (CV) cycle to 500 will result in the formation of Pt nanoclusters during the Pt immobilization process (Pt NC -Ni 3 S 2 @Ag NWs, Figures S12-S14, Supporting Information), suggesting the continuous deposition of Pt atoms during the repeated polarization scanning, and the optimal HER catalytic performance could be obtained.
The corresponding mechanism was unraveled by theoretical investigations and experimental characterization. As shown in Figure 3a, the formation of Ni vacancies and Pt single atoms anchoring results from the facilitation of electrochemical reaction in alkaline media altering the chemical composition of Ni 3 S 2 @Ag NWs surface. Based on the theoretical calculation, the Ni-terminated Ni 3 S 2 provides the optimal site for water absorption ( Figure 3b; and Figure S15, Supporting Information) and hydrogen evolution ( Figure S16, Supporting Information). By capturing electrons, the Ni-terminated basal planes are capable to reduce water into H 2 , and the generated −OH was grafted to the Ni terminal by the Ni-O band, causing the formation of Ni-OH intermediate confirmed by X-ray photoelectron spectroscopy (XPS) (Figure 3b). [34,35] Further, the following reactions on Ni-OH intermediate include two possible routes. One is the restoration of Ni-OH intermediate to the original Ni-Ni 3 S 2 terminal by the deabsorption of the −OH due to the weak binding between Ni and −OH comparing with the strong Ni-S bond. [36] Another route involves the local corrosion of low-valence Ni in Ni-OH intermediate supported by Ni 3 S 2 induced by the chloride ions originating from PtCl 6 2− in alkaline media to form high-valence Ni 2+ -OH due to the strong depassivation ability and penetrating performance of Cl − ions. [37][38][39] Subsequently, by combining OH − ions, the Ni 2+ -OH intermediate was transformed to the metal hydroxide Ni(OH) 2 in the alkaline electrolyte to form a chemistry-stable OH-Ni-OH leaving group. [40] Consequently, the Ni vacancies on the surface of the Ni 3 S 2 (V Ni -Ni 3 S 2 ) were created for subsequent immobilization of Pt atoms. X-ray near-edge absorption spectra (XANES) for Ni K edge of V Ni -Ni 3 S 2 are depicted in Figure S17a (Supporting Information) and show that the white line intensity of V Ni -Ni 3 S 2 is lower than that of NiO, but higher than that of the pure Ni 3 S 2 . This proves that Ni atoms in V Ni -Ni 3 S 2 are positively charged, [7,8] implying Ni vacancies formation. By Bader charge analysis deriving from DFT calculation ( Figure S17b, Supporting Information), the average charge value of Ni atoms in V Ni -Ni 3 S 2 (+0.28 e) is high than that of Ni atoms in Ni 3 S 2 (+0.10 e), further implying the increased degree of nickel oxidization state and Ni vacancies formation. As shown in the HAADF-STEM (Figure 3c; and Figure S17c, Supporting Information) image, after the same electrochemical process with Pt SA -Ni 3 S 2 @Ag NWs but replacing PtCl 6 2− by NaCl in KOH solution to avoid the deposition of Pt species, the obtained Ni 3 S 2 shows a large number of Ni vacancies (V Ni -Ni 3 S 2 ). By comparing the STEM results of Pt SA -Ni 3 S 2 @Ag NWs and V Ni -Ni 3 S 2 @Ag NWs, the Pt atoms are finely anchored in the sites of Ni vacancies. Combining the analysis in theoretical investigations (Figure 1e), the anchored Pt atoms at Ni positions under cathodic potential will protect adjacent Ni against corrosion. Consequently, a stable single-atom Pt-Ni 3 S 2 support system could be obtained. Besides, the release of reduced H 2 will serve as the dynamic template to generate interconnected pores in the Ni 3 S 2 layer, [25] leading to more accessible basal planes for Ni vacancies defects creation and Pt atoms immobilization.
Further, the electronic states of Pt atoms in Pt SA -Ni 3 S 2 were verified by XPS, X-ray absorption near-edge spectroscopy (XANES), and extended X-ray absorption fine spectroscopy (EX-AFS) measurements. The binding energy in high-resolution Pt 4f XPS spectra (Figure 4a) shows a slightly positive shift for Pt SA -Ni 3 S 2 compared to Pt foil. Moreover, the Ni 2p and S 2p XPS peaks show an inapparent evolution in the binding energy after Pt anchoring ( Figure S18, Supporting Information), suggesting a weak electrons transfer between Pt atoms and Ni 3 S 2 support. The corresponding electronic states of Pt atoms in Pt SA -Ni 3 S 2 were further explored by XANES measurements (Figure 4b), in which Pt SA -Ni 3 S 2 show slightly enhanced intensity than Pt foil but extremely low than PtO 2 , proving less electron loss of Pt atoms in Ni 3 S 2 as theoretical calculation in Figure 1g. [13] The Fourier transforms spectrum of the k3-weighted EXAFS oscillations of Pt SA -Ni 3 S 2 exhibits a prominent peak in the region 1.8-2.4 Å (Figure 4c), corresponding to the Pt-Cl and Pt-S first coordination shell. [41,42] The absence of Pt-Pt contribution peaks at about 2.7 Å confirms atomically dispersed Pt in Ni 3 S 2 . The fitting results of the Fourier transform curves give a coordination number of 3 for Pt-S contribution (Figure 4c; and Table S1, Supporting Information), confirming that the single Pt atoms were anchored in the sites of Ni vacancies by binding with the nearest three S atoms in Ni 3 S 2 support (Figure S1c, Supporting Information). To further verify the coordination conditions of Pt atoms, the wavelet transforms (WT, Figure 4d-f) analysis was performed, which can provide powerful resolution in the radial distance in the R and K space. A WT intensity maximum at about 5 and 12 Å −1 are attributed to the Pt-O and Pt-Pt contributions, respectively. [43] In contrast, the intensity maximum of Pt SA -Ni 3 S 2 is near 8 Å −1 , assigned to the Pt-S contribution, and the Pt-Pt coordination with an intensity maximum at ≈12 Å −1 is not detected, confirming the presence of isolated Pt atoms in Ni 3 S 2 .
Based on the theoretical calculation and structural characterizations, for the first time, we reveal that Ni 3 S 2 tuned singleatom Pt tends to present a promising activity for HER by the integration of electronic structure adjustment strategy and architectural nanostructure engineering. Experimentally, the electrocatalytic performance of seamlessly conductive Pt SA -Ni 3 S 2 @Ag NWs with 3D architecture for HER was verified in alkaline electrolyte, and the original Ni 3 S 2 @Ag NWs, Ag NWs, and Ni foam supported commercial Pt/C (Pt/C@Ni foam) were also measured for comparison. Besides, Pt SA -Ni 3 S 2 @Ni foam was fabricated under the same conditions as Pt SA -Ni 3 S 2 @Ag NWs but replacing Ag NWs with Ni Foam. As demonstrated in Figure 5a, the Pt SA -Ni 3 S 2 @Ag NWs shows the highest HER performance among all catalysts, and only needs 33 and 137 mV to achieve  Figure 5b). c) The mass activity of Pt atom in Pt SA -Ni 3 S 2 @Ag NWs, Pt SA -Ni 3 S 2 @Ni foam, and state-of-theart Pt/C on Ni foam. The digital photos of fabricated catalysts d) Pt SA -Ni 3 S 2 @Ag NWs, e) Pt SA -Ni 3 S 2 @Ni foam, and f) Pt/C@Ni foam at the initial state and an overpotential of 150 mV in 1 m KOH electrolyte for HER. g) TOF plots of the Pt SA -Ni 3 S 2 @Ag NWs, Pt SA -Ni 3 S 2 @Ni foam, and Pt/C@Ni foam after normalizing to Pt atom sites.
the current density of 10 and 150 mA cm −2 , respectively, suggesting significantly superior activity than the Pt SA -Ni 3 S 2 @Ni foam, Ni 3 S 2 @Ag NWs, Ag NWs, and even the commercial Pt/C on Ni foam at an overpotential of >80 mV ( Figure S19, Supporting Information), in which the decrease of the current density of Pt/C at the high potential should be attributed to the application of Nafion as the binder and poor nanostructure morphology of Pt/C electrode, hindering the rapid generation and release of H 2 . Besides, the Pt SA -Ni 3 S 2 @Ag NWs exhibit a similar Tafel slope (34.70 mV dec −1 ) with Pt SA -Ni 3 S 2 @Ni foam and Pt/C@Ni foam towards HER ( Figure S20, Supporting Information), suggesting Pt-like kinetics deriving from the introduction of single-atom Pt. Moreover, the single-atom Pt anchored Ni 3 S 2 @Ag NWs also possess higher activity than the Ni 3 S 2 @Ag NWs supported Pt particles ( Figure S12, Supporting Information) due to the more efficient synergy of the single Pt atom and the Ni 3 S 2 host. Apart from the outstanding metal-support interaction, the Ag NWs are introduced to provide facile electron transport pathways through the entire catalyst, and consequently, the charge transfer resistance (R ct ) between Pt SA -Ni 3 S 2 @Ag NWs and reactants (H 2 O*, M-H abs and H*) reach a significantly low value of 0.95 Ω (Figure S21, Supporting Information), leading to a faster faradaic reaction. [26,27] Moreover, Pt SA -Ni 3 S 2 @Ag NWs exhibits a 5.1-time increase in double-layer capacitance (C dl ) over Ni 3 S 2 @Ag NWs ( Figure S22, Supporting Information), suggesting more exposed sites for Pt atoms immobilization and HER due to the formation of a large number of voids in Ni 3 S 2 during Pt anchoring process ( Figure S10, Supporting Information). Considering the ac-tual applications, Pt SA -Ni 3 S 2 @Ag NWs also shows excellent stability with negligible decay in HER activity for 5000 cycles or 30 h (Figure 5b). The corresponding structural measurements including SEM, STEM images, EDS mapping, XRD, and XPS data ( Figures S24 and S25, Supporting Information) are performed to confirm the single-atom dispersion and intact structure of Pt SA -Ni 3 S 2 @Ag NWs after the stability test, suggesting the high stability of Ni 3 S 2 @Ag NWs immobilized single-atom Pt by Ni vacancies. Furthermore, the mass activity of Pt SA -Ni 3 S 2 @Ag NWs normalized to the Pt loading mass (1.47 wt%, measured by inductively coupled plasma-mass spectrometry) at an overpotential of 150 mV is 7.6 A mg −1 (Figure 5c), which is 27-fold high than that of the commercial Pt/C catalyst (20 wt% Pt/C, 0.28 A mg −1 ). As a result, a large hydrogen generation rate was observed at room temperature for Pt SA -Ni 3 S 2 @Ag NWs (Figure 5d) comparing with Pt SA -Ni 3 S 2 @Ni foam ( Figure 5e) and Pt/C@Ni foam (Figure 5f). Especially, an aggregative H 2 cloud was presented on the top of the Pt SA -Ni 3 S 2 @Ag NWs electrode, indicating the high efficiency of M-H ads generation and facile conversion to H 2 . Besides, comparing with Ni foam supported commercial Pt/C electrode (Figure 5f) fabricated by using Nafion as the binder, no large bubble absorbs on the surface of Pt SA -Ni 3 S 2 @Ag NWs electrode, which could be attributed to the shift-down of the d-band center (Figure 1f), more optimal ∆G H* with −0.06 eV than that of Pt foil (−0.10 eV, Figure 1h), as well as the formation of abundant voids (Figure 2c; and Figures S10 and S23, Supporting Information) of Ni 3 S 2 immobilized single-atom Pt coating on Ag NWs network for facile desorption and release of H 2 . To quantify the output www.advancedsciencenews.com www.advancedscience.com efficacy of H 2 , the turnover frequency (TOF, Note S1, Supporting Information) per Pt atom was performed. As shown in Figure 5g, Pt SA -Ni 3 S 2 @Ag NWs catalyst shows a high TOF value of 2.91 H 2 s −1 , which is 1.6 and 19.4 times higher than that of Pt SA -Ni 3 S 2 @Ni foam and Pt/C@Ni foam, respectively, demonstrating the extremely high activity of Pt SA -Ni 3 S 2 @Ag NWs for HER. For the Faradaic efficiency, we roughly collect the H 2 by a drainage method to estimate the experimental gas amount. Faradaic efficiency can be determined by the ratio of H 2 (experimental) to H 2 (theoretical) in a percentage form. To determine the Faradaic efficiency of the Pt SA -Ni 3 S 2 @Ag NWs hybrid catalyst for HER, the electrolysis was performed in 1 m KOH with an applied constant current density of 100 mA cm −2 for 100 s. The Faradaic efficiency of about 90% was obtained for HER, indicating that the generated charges were almost consumed for H 2 generation. As summarized in Table S2 (Supporting Information), our Pt SA -Ni 3 S 2 @Ag NWs also show superior HER performances than the recently reported SACs, confirming that the improved HER activity of the fabricated Pt SA -Ni 3 S 2 @Ag NWs mainly derives from the more favorable reconstruction of the electronic-state structure after single-atom Pt anchoring on Ni 3 S 2 support.

Conclusion
Based on the theoretical simulation, we design a transition-metal sulfide-Ni 3 S 2 , with abundant exposed Ni vacancies created by the interaction between chloride ions and the functional groups on the surface of Ni 3 S 2 , to anchor atomically dispersed Pt for efficiently electrocatalytic hydrogen evolution. Specifically, Ni 3 S 2 tailored Pt single atom shows a weak metal-support interaction, increasing the d orbital electron occupation near Femi level and shift-down of the d-band center for optimizing the reactants absorption during HER. The unique coordination environment of Pt sites is constructed by the orbital hybridization between S-3p and Pt-5d after introducing Pt into Ni position in Ni 3 S 2 system, inducing a locally enhanced electronic field distribution by increasing the electron density and metallic-state feature on the Pt sites, promoting the electrochemical HER kinetics. By integrating single-atom modification strategy and architectural nanostructure engineering, the fabricated 3D Pt SA -Ni 3 S 2 @Ag NWs exhibits outstanding HER catalytic activity in terms of high turnover frequencies (2.91 H 2 s −1 at 95 mV) and extremely great mass activity of 7.6 A mg −1 , exceeding the commercial Pt/C catalyst and the advanced SACs reported recently. This work presents a new avenue for developing high-activity single-atom electrocatalyst by tailoring the electronic structure and geometrical construction of SACs via transition-metal based sulfides.

Experimental Section
Synthesis of Ni@Ag NWs: First, an oil bath process was carried out to synthesize Ag nanowires (Ag NWs). In detail, an ethylene glycol solution containing FeCl 3 (7.19 × 10 −3 m), AgNO 3 (0.051 m), and polyvinylpyrrolidone (0.012 m) was kept at 110°C for 12 h by oil bath pan. Then, the generated yellow precipitate was collected and washed with acetone and alcohol to obtain Ag NWs. Subsequently, the cleaned Ag NWs were dispersed and stored in an alcohol solution. A conductive network was fabricated by spray coating Ag NWs on a flexible cloth fabric. Finally, the metallic Ni layer is grown on Ag NWs network by a facile electrodeposition process. Specifi-cally, the solution consisting of 0.10 m NiCl 2 , 0.09 m H 3 BO 3, and a solvent containing ethanol and deionized water with 2:5 in volume ratio. Then the electrodeposition of metallic Ni was performed by chronoamperometry with 1.2 V for 400 s in a standard three-electrode system on a workstation (CHI660E), in which graphite sheet acted as a counter electrode, saturated calomel electrode (SCE) acted as a reference electrode. Finally, the Ag NWs supported metallic Ni was cleaned and dried for stand-by.
Synthesis of Ni 3 S 2 @Ag NWs: Ni 3 S 2 @Ag NWs were prepared by performing a hydrothermal treatment on obtained Ni@Ag NWs. In brief, the Ni@Ag NWs coated cloth pieces were immersed in a 50 mL Teflonreaction vessel containing 30 mL of 3 mmol TAA aqueous solution and then kept at 120°C for 90 min. The obtained samples were cleaned with water and alcohol, and finally drying at room temperature, marking as Ni 3 S 2 @Ag NWs material.
Synthesis of V Ni -Ni 3 S 2 @Ag NWs: V Ni -Ni 3 S 2 @Ag NWs were fabricated by an electrochemical process in the three-electrode system, in which Ni 3 S 2 @Ag NWs coated cloth was performed as the working electrode, graphite sheet acted as a counter electrode, SCE acted as a reference electrode. The corresponding electrochemical process was carried out by multicycle cathode polarization in 1 m KOH solution containing 0.5 m NaCl with a scan rate of 50 mV s −1 between 0.00 and −0.50 V versus reversible hydrogen electrode (RHE) for 400 cycles.
Synthesis of Pt SA -Ni 3 S 2 @Ag NWs: Pt SA -Ni 3 S 2 @Ag NWs were fabricated by an electrochemical process in the three-electrode system, in which Ni 3 S 2 @Ag NWs coated cloth was performed as the working electrode, graphite sheet acted as a counter electrode, SCE acted as a reference electrode. The corresponding electrochemical process was carried out by multicycle cathode polarization in 1 m KOH solution containing 50 × 10 −6 m H 2 PtCl 6 with a scan rate of 50 mV s −1 between 0.00 and −0.50 V versus RHE for 400 cycles.
Characterizations: The morphology of the fabricated catalysts in this study was observed by SEM (GeminiSEM 300). HRTEM images, HAADF-STEM images, and STEM-EDX mapping images were obtained by a TEM coupled with an energy spectrum analyzer (JEOL JEM2100). The Pt contents in Pt SA -Ni 3 S 2 @Ag NWs were measured by inductively coupled plasma optical emission spectrometry. The XPS spectra of elements in Pt SA -Ni 3 S 2 @Ag NWs were tested by a surface analysis system (ES-CALAB250Xi). The phase and crystal information of the fabricated catalysts were obtained by Cu K radiation in an XRD (Shimadzu). The EX-AFS measurement of the Pt SA -Ni 3 S 2 @Ag NWs at the Pt L 3 -edge was performed at 1W1B station at the Beijing Synchrotron Radiation Facility. The cloth-supported Pt SA -Ni 3 S 2 @Ag NWs were directly subjected to measurements. Data analysis and fitting were performed with Athena and Artemis in the Demeter package.
Electrochemical Measurements: All electrochemical tests were finished by an electrochemical workstation (CHI 660E) with a three-electrode configuration, in which fabricated catalysts in this work were employed as the working electrode, graphite sheet acted as a counter electrode, SCE acted as a reference electrode. All the presented potential in this work was calibrated versus RHE according to the experimental method. [13] HER performance tests were performed in 1 m KOH aqueous solution. Linear sweep voltammograms (LSV) with 95% iR-corrections were tested with a potential range from 0.05 to −0.5 V under a scan rate of 5 mV s −1 . Electrochemical impedance spectroscopy (EIS) was obtained by a frequency range from 100 k to 0.1 Hz with 200 mV of initial voltage. For the preparation of Pt/C@Ni foam, 5 mg 20 wt% Pt/C was dispersed in 0.9 mL alcohol containing 0.1 mL 5 wt% Nafion solution to form a homogeneous ink. Then, the obtained ink was coated on the Ni foam and dried in air to form a porous Pt/C@Ni foam electrode.
DFT Theoretical Calculations: All the DFT simulations were performed by the Vienna Ab-initio Simulation Package (VASP) [44][45][46] with the projector-augmented-wave method. [47,48] And the method DFT-D3 developed by Grimme et al. [49,50] was also employed to account for the dispersion correction and van der Waals' interaction into the conventional Kohn-Sham DFT potential energy calculation. The exchange-correlation energy was described by the Perdew-Burke-Ernzerhof (PBE) implementation of generalized gradient approximation (GGA-PBE). [51] Spin-polarized calculations have been performed with plane wave expansion truncated at 500 eV. 2 × 4 × 1 k-mesh grids were set for the self-consistent field calculation and structural relaxation. Electronic relaxation utilized the conjugategradient (CG) [52] with the energy convergence being 10 −5 eV. Geometry optimization was conducted by using the quasi-Newton algorithm [53,54] with residual forces on atoms low than 0.01 eV Å −1 . In all calculations, spin polarization was considered. A vacuum region thickness of 20 Å was built. The atoms at the top two layers were free for structural relaxation.
The H and H 2 O absorbing on Ni 3 S 2 , Pt SA -Ni 3 S 2, and Pt slab were investigated by comparing the formation energy of different sites. The equation for calculating adsorption enthalpy ΔE H* as the following Where the E slab+H is the total enthalpy of H adsorbing on the catalysts, the enthalpy of the catalysts is E slab , the H 2 enthalpy is E H2 . As similar, the equation for calculating the H 2 O adsorption enthalpy ∆E H2O as the following where ΔE H* represent the H adsorption energy and ∆E H2O* represent the H 2 O adsorption energy, and ΔE ZPE represents the difference related to the zero-point energy between the gas phase and the adsorbed state.

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