1 Introduction

The high dependence on fossil fuels worldwide has resulted in serious environmental pollution and climate change, leading to energy crises. It is imperative for us to conform to the development of the times, phase out fossil fuels, and strive to reduce the usage of non-renewable energy by develo** clean energy [1, 2]. Hydrogen has apparent advantages over other energy sources because of its extensive origins, high combustion calorific value, and eco-friendly production possibilities. However, conventional large-scale hydrogen production technologies mainly include coal gasification and steam reforming, which result in massive emissions of environment-unfriendly CO2 gas [3]. Compared to other methods, water-splitting appears to be a perspective and environmentally friendly method for hydrogen production. Despite this advantage, the hydrogen production efficiency is further restricted by oxygen evolution reaction (OER)’s kinetics at the anodes during water splitting, which causes the high overall energy consumption during electrolysis. At the moment, noble metal Pt-based catalysts are outstanding catalysts for water-splitting because of their perfect adsorption properties in volcanic type activity trends and several representative electrocatalytic qualities including excellent electrochemical activity, chemical stability and corrosion resistance. However, the price of precious metals is expensive due to their extreme rarity and high demand. As a result, it is believed that monometallic catalysts based on noble metals are less likely to be used in industrial production [4,5,6].

Hence, develo** bi-functional catalysts with low platinum load recently becomes an effective strategy for hydrogen evolution via organic-water co-electrolysis, which can not only decrease energy consumption of hydrogen evolution by reducing the cell voltage, but also concurrently obtain value-added organic chemical products at the anode, making this reaction academically and industrially significant [7]. Formic acid has a higher market value compared to methanol, it is an important intermediate in the chemical industry with higher value (> 539 €/ton) and commonly used for synthesizing various fine chemicals. Methanol has received wide attention in comparison to other small organic compounds including urea, ethanol, glycerol, amine, formaldehyde [8, 9], hydrazine hydrate [10, 11], furfural [12, 13], and 5-hydroxymethylfurfural owing to its strong oxidation reactivity, high water solubility, and affordable pricing (about 350 €/ton) [22]. Although the commercially used platinum catalysts have good performance in hydrogen production, they have the drawbacks of low earth content. Further, as a bi-functional catalyst, the efficiency of such co-electrolysis for formate-H2 co-production is limited by CO poisoning during the methanol oxidation reaction (MOR), preventing high current densities from steady working overtime [24].

Reducing the loading of precious metal Pt and combining it with chalcogenide multiphase compounds may be an effective strategy to further decrease the voltage of water electrolysis and boost the efficiency of hydrogen production. By now, some previous works have found that various heterojunction nanomaterials formed by the combination of precious metal Pt and chalcogenide compounds have a significant effect on enhancing the activity of catalysts for HER, OER, CO2RR, and other catalytic reactions [25, 26]. The key reasons are the synergistic effects of the constructed interface which was generated from the contacting region due to influence each other to increase the catalyst activity significantly [27,28,29]. Nowadays, the design of low platinum loading catalysts typically includes some strategies. Among them, platinum nanoparticles are integrated into specific carrier to enhance the dispersibility and mass activity of platinum [1. This work provides an approach towards rational design of efficient bi-functional electrocatalysts through precise construction of interfaces.

Scheme 1
scheme 1

Diagram of the as-synthesized DMD Pt–Ni3S2 HNCs used as electrodes for MOR and HER

Full size image

2 Experimental Methods

2.1 Synthesis of Pt–Ni3S2 Catalyst

The one-pot solution-based syntheses were performed using a standard Schlenk vacuum line technique under argon atmosphere. In a standard synthesis, 2 mmol (0.5138 g) Ni(acac)2 and OLA (20 mL) were fully dissolved in a round-bottom Schlenk flask (100 mL) at room temperature. The flask was degassed under vacuum at 80 °C for 0.5 h to remove oxygen and other low-boiling-point organic solvent. Subsequently, the reaction was programmed to be 220 °C with a ramp of 5 °C min−1 after backfilling with Ar in oil bath. At the same time, 1 mmol (0.2184 g) DPDS and 3 mL OLA, 0.1 mmol (0.04 g) Pt(acac)2 and 1 mL OLA was separately mixed in a glass vial, then preheated to 80 °C on a hot plate to form a clear solution. When the flask reaches 220 °C, the diphenyl disulfide solution was injected into the metal solution by syringe. After injection, the temperature drops to 210–215 °C and the reaction was allowed to maintain at 215 °C for 10 min with continuous stirring. After 10 min dwelling for the growth of Ni3S2 NPs, the acetylacetone platinum solution was further injected into the Ni3S2 dispersion by syringe. With constant stirring, the reaction was allowed to continue at that temperature for 5 min. After stop** the reaction, the flask was taken out of the oil bath and allowed to naturally cool to room temperature. The product was dissolved in toluene and the solution was centrifuged at 12,000 rpm during 10 min for nanoparticles separation. Finally, the as-synthesized Pt–Ni3S2 nanocrystals were thoroughly purified by multiple precipitation and re-dispersion steps using toluene and isopropanol.

2.2 Electrocatalytic Experiments

Electrochemical measurements were achieved by CHI760E (CH Instruments, Inc. Shanghai, China) electrochemical analyzer at room temperature with standard three-electrode system. The sample was chosen as the working electrode and a saturated Ag/AgCl (sat. KCl) and the platinum foil were chosen as the reference electrode and the counter electrode, respectively. Two-electrode system using Pt–Ni3S2 sample as both the cathode and anode was utilized to estimate the performance of water splitting. All the potentials were converted to the reversible hydrogen electrode (RHE) according to the Nernst equation, the conversion equation is E(RHE) = E(Ag/AgCl) + 1.0205 V (Fig. S5). 20 cycles of CVs cans were conducted for every working electrode with electrocatalyst before the data collection. The linear sweep voltammetry (LSV) curves were recorded at a scan rate of 5 mV s−1 without iR-correction. Electrochemical impedance spectroscopy (EIS) was undertaken at 1.49 V (vs. RHE) with AC amplitude of 5 mV and frequency range of 0.01–100 kHz. Moreover, the cyclic voltammetry (CV) curves were collected in the potential range from 1.02 to 1.12 V (vs. RHE) with the scan rates of 20, 40, 60, 80, 100 and 120 mV s−1 to obtain the double-layer capacitance of electrocatalyst. Ion Chromatography (CIC-D120, SHENGHAN, China) was employed to identify and quantify the value-added chemical product (formate). The generated H2 in the cathode compartment were determined by gas chromatography (5977B MSD, Agilent Technologies). At least three chromatographic trace curves were collected for statistical analysis.

2.3 Catalyst Characterization

The morphological information of Pt–Ni3S2 catalyst was characterized by field emission transmission electron microscopy (FETEM JEM-F200) including SAED and EDS elemental map** functions. Crystallographic and purity information on Pt–Ni3S2 catalyst were obtained by X-ray powder diffraction (XRD, RIGAKU Smartlab), The elemental components of Pt–Ni3S2 was studied by X-ray photoelectron spectroscopy (XPS, ESCALAB 250** images in Fig. 1f explicitly manifest evenly distributions of all three elements within the Pt–Ni3S2. Figures 1g and S1 present the clear structures of single Ni3S2 nanocrystal plates with the size of ~ 15 nm. Furthermore, the extra HRTEM image demonstrates a substantial amount of clear lattice fringes corresponded to the (110) facets of Ni3S2 in Fig. 1h. The matching EDS elemental map**s confirm (Fig. 1i) the existence of Ni and S. Notably, various lattice planes of highly dispersed Ni3S2 nanocrystals (PDF # 00-044-1418) agree well with the distinct diffraction rings shown in Fig. 1j.

Fig. 1
figure 1

Composition characterizations and morphology of the Pt–Ni3S2 and Ni3S2 nanocrystals. a XRD results; b TEM images, c HAADF-STEM image and SAED pattern (inset), d–e HRTEM images and f EDS elemental map**s of Pt–Ni3S2; g–h HRTEM, i EDS map**, and j SAED pattern of Ni3S2 nanocrystals

Full size image

For more details on surface chemical states between Pt and Ni3S2 catalysts was obtained by resorting to the XPS. In Fig. 2a, Ni 2p XPS spectra of Pt–Ni3S2 presents two visible peaks at 873.4 and 855.64 eV that should be allocated to Ni 2p1/2 and Ni 2p3/2 orbits of Ni2+ [54]. After Pt nanocrystals is anchored to Ni3S2 nanocrystals, there was a positive shift in the binding energy of Ni 2p1/2 and Ni 2p3/2 of Ni2+ in Pt–Ni3S2, indicating that the platinum atom influences the electronic structure of the element Ni in the Pt–Ni3S2 to result in a higher Ni valence state. In addition, the lower content of Ni0 species of Pt–Ni3S2 compared with Ni3S2 may be associated with lightly absorption of energy from Pt nanocrystals placed on the Pt–Ni3S2 [54]. Furthermore, XPS spectrum in O 1s orbital (Fig. S3b) shows that the surface of Pt–Ni3S2 and Ni3S2 contains some OH− and water species. It is easy to adsorb a small amount of oxygen-containing species due to the small particle size of the catalyst, which is also very conducive to the activation of methanol [60,61,62]. The FT-EXAFS of Pt–Ni3S2 exhibits prominent peaks in the region 1.8–2.8 Å (Fig. 2f). The EXAFS of Pt–Ni3S2 displays the length of the Pt–Pt bond (~ 2.42 Å), which is approximately 6.2% shorter than that in Pt foil (~ 2.58 Å), demonstrating that the short-range ordered Pt–Pt bond is affected by Ni3S2 to make it form Pt–Pt(Ni) chemical bonding interactions at the Pt–Ni3S2 interface [34, 35, 63]. The Pt L3-edge EXAFS results also indicated the existence of Pt–S bond at ~ 1.92 Å, on which the powerful correlation between Pt and Ni3S2 is beneficial for anchoring Pt atoms to make Pt nanocrystals uniformly dispersed on the surface of Ni3S2 nanocrystals [64, 65]. Additionally, the WT-EXAFS analyses for Pt–Ni3S2, Ni3S2, Ni foil, Pt foil, and Ni(OH)2 are given in Figs. 2g–n and S4 for further investigating the coordination environment of Ni, S and Pt atoms. The WT-EXAFS spectra in Ni K-edge for Ni3S2 and Pt–Ni3S2 (Fig. 2g–l) show a maximum intensity at ~ 6.35 and ~ 5.70 Å−1, which further explain that the Ni–S bond length of Pt–Ni3S2 is shorter than the pristine Ni3S2 [25]. In addition, a WT intensity maximum (Fig. 2i–n) at about 9.8 Å−1 was attributed to the Pt–Pt scattering in Pt foil. In contrast, the intensity maximum of Pt–Ni3S2 is near 6.5 Å−1, assigned to the Pt–Pt(Ni) contribution [17, 27, S16c–f) of used Pt–Ni3S2 show good dispersion of elemental of Ni, S, C and Pt. In particular, elemental of Pt maintains excellent dispersibility after such drastic MOR process by chronoamperometry (I–t) at 2.12 V (vs. RHE) with an initial current density of ~ 1000 mA cm−2 for 72 h. The supplementary XPS and TEM results indicate that the hetero-structure of Pt–Ni3S2 is highly stable for electrocatalytic evolution at large current density, which further confirms its high potential for practical applications in environmental and energy fields.

3.3 Theoretical Study and Mechanistic Insight

To deeply understand the origin of the good performance of the Pt–Ni3S2 for MOR, DFT computations are further performed to gain insight into the interaction between Pt and Ni3S2 nanocrystals. The schematic models of Pt–Ni3S2, Ni3S2, and Pt foil catalyst are shown in Fig. 4a. These electronic changes were marked by the differential charge density of Pt–Ni3S2 interface, in which blue stands for electronic consumption state and yellow stands for electronic accumulation state (Fig. 4b). The charge density of Ni3S2 nanocrystals was weakened after introducing Pt nanocrystals, which laterally reflects the higher oxidation state of element Ni. Further, we calculated the charge distribution to quantitatively evaluate these charges. It can be clearly seen that the electrons at the interface are clustered on the Pt side.

Fig. 4
figure 4

Density functional theory (DFT) computations. a Theoretical models of Pt foil, Ni3S2, and Pt–Ni3S2; b Charge density difference plot at the Pt–Ni3S2 interface; c the adsorption energy diagrams of CH3OH and d the desorption energy diagrams of HCOOH on the surfaces of Pt–Ni3S2, Ni3S2, and Pt foil; e Schematic illustration of MOR mechanisms for Pt–Ni3S2 electrocatalyst

Full size image

These results indicate that the electronic interaction caused an electron aggregation effect on the Pt species and enhanced the oxidation state of Ni species, resulting in a significantly lower potential [69]. As shown in Fig. 1b, the onset potential of Pt–Ni3S2 is 1.31 V (vs. RHE) and its overpotential is 1.35 V (vs. RHE) at 10 mA cm−2, indicating that the electro-oxidation of adsorbed methanol is probably driven by the active Niδ+ sites with higher valence (δ > 2, e.g., probably Ni–OOH), which is consistent with the reported literature about nickel-based electrocatalyst [15, 70]. As the potential gradually increases, methanol is decomposed to formic acid on the surface of the catalyst. The adsorption energy of CH3OH* and the desorption energy of HCOOH* is different due to the different electronic valence states of surface Ni on Pt–Ni3S2, Ni3S2, and Pt. Hence, their onset potentials and overpotentials are different accordingly. Further, the adsorption energy of CH3OH* and the desorption energy of HCOOH* on the surfaces of Pt–Ni3S2, Ni3S2, and Pt foil was calculated to analyze the adsorption/desorption performance of the products via DFT (Figs. 4c, d and S17–S19). The Pt–Ni3S2 shows stronger adsorption energy (− 0.80 eV) for CH3OH* than Pt (− 0.47 eV) and Ni3S2 (− 0.19 eV), which indicates that Pt–Ni3S2 has a strong methanol adsorption capacity. It is beneficial for the initial adsorption and activation of CH3OH [71]. Figure 4d shows that the energy span for *HCOOH → * + HCOOH is − 0.31 eV on Pt–Ni3S2 surface, which is lower than those on Pt foil (0.22 eV) and Ni3S2 (0.07 eV), indicating that this step is thermodynamically easier to occur on Pt–Ni3S2 surface where Pt acts as the catalyst promoter. Therefore, the strong chemical interaction between Pt and Ni3S2 can modulate the electronic structure of Pt–Ni3S2, which further facilitates the formation of high-valent Ni and enhances the adsorption energies of intermediates for MOR, resulting in MOR activity [72]. Based on the aforementioned analysis, an active center mechanism for Pt–Ni3S2 to catalyze MOR is discussed (Fig. 4e). The methanol molecule, initially adsorbed on the catalyst surface via forming Ni3+–O bond, will undergo successive de-protonation and C–H bond cleavage steps with the assistance of Pt site. Finally, the intermediate at the Ni site is formed and easily converted into HCOO− due to the high concentrations of OH− in the electrolyte [15, 23, 69].

The electrocatalytic HER performance of Pt–Ni3S2 and Ni3S2 catalysts was studied using a standard three-electrode system at room temperature. Meanwhile, Commercial 20% Pt/C was used as a comparative sample to compare the HER activity of Pt–Ni3S2 and Ni3S2. The LSV curves (Fig. 5a) show that Pt–Ni3S2 exhibited the higher HER activity with a lower η10 of 61 mV (vs. RHE) than that of pure Ni3S2 (270 mV) without iR-compensation. The data with iR-compensation (Fig. S22) also confirms the performance tendency of these three kinds of electrocatalysts. Interestingly, Pt–Ni3S2 gradually surpassed the 20% Pt/C at a high current density region. That is, Pt–Ni3S2 requires an ultralow working potential of 440 mV (587 mV for 20% Pt/C) to achieve 300 mA cm−2. As histogram (Fig. 5b) also visually shows its advantages over other catalysts. Such an excellent HER performance of Pt–Ni3S2 also exceeds that of mostly reported catalysts (Table S5). Mass activity is also an important parameter for evaluating the electrocatalytic activity. As given in Fig. S20, Pt–Ni3S2/CC exhibits much higher mass activity than 20% Pt/C/CC at higher overpotential. The HER kinetics and mechanism are uncovered by Tafel slopes in Fig. 5c, and the Pt–Ni3S2 with 102 mV dec−1 is lower than 20% Pt/C (111 mV dec−1) and Ni3S2 (236 mV dec−1). This result is further validated by the lowest Rct of 2.4 Ω in the EIS spectra in Fig. S23. Further, ECSA normalized polarization curve (Fig. S24) shows the Cdl value of Pt–Ni3S2/CC is 18.75 mF cm−2, which is strongly superior to those of Ni3S2/CC (11.64 mF cm−2) and CC (1.86 mF cm−2), reflecting that the Pt–Ni3S2 has a higher amount of catalytic active sites. In addition, the Pt–Ni3S2 shows nearly 100% Faradaic efficiency by comparing theoretical value and measured value (Figs. 5d and S25). The cycling stability of Pt–Ni3S2 catalyst has been further verified by LSV scanning, in which the polarization curves measured by chronoamperometry after 48 h almost overlaps with the initial polarization curve, signifying the strong HER stability (Fig. 5e). The used Pt–Ni3S2 after HER stability tests are further characterized by XPS (Fig. S26) and HRTEM (Fig. S27), in which the results are similar to the those after MOR stability tests, indicating that the used Pt–Ni3S2 can operate stably for 24 h at a high current density of 100 mA cm−2 with remained excellent dispersibility of Pt atoms. TOF is another important figure of merit used to reveal the intrinsic electrocatalytic activity. As shown in Fig. S21, the Pt–Ni3S2/CC show much higher TOF values over the whole potential ranges than the other catalysts for HER. The TOF values of Pt–Ni3S2/CC were 1.44 and 2.88 s−1 at the overpotential of 50 and 100 mV in 1.0 mol L−1 KOH, respectively, which are all superior to other catalysts. Moreover, DFT calculations are applied to provide the mechanistic understandings for the high activity of Pt–Ni3S2 toward HER. Normally, the key steps affecting the reaction rate for alkaline HER include water adsorption and hydrogen desorption. Hence, the Gibbs free energy for hydrogen adsorption (ΔGH*) on electrocatalysts are calculated in this study. Figures 5f–g and S28–S30 show the binding models of H2O molecule and H atom at the active sites of Pt on Pt–Ni3S2. According to DFT simulation (Fig. 5h), it should be a rate determining step from the dissociation of *H2O to the formation of *H on Pt–Ni3S2, since a larger energy gap exists between *H2O (− 0.71 eV) and *H (− 0.19 eV). Hence, the HER performance on 20% Pt/C is better than that on Pt–Ni3S2 at lower overpotential, which is indicated by the LSV results in Fig. 5a. However, the current density of HER on Pt–Ni3S2 (Fig. 5a) is obviously larger than that on 20% Pt/C at higher overpotential, probably because the energy barrier between *H2O and *H is overcome on such condition [62, 65, 73]. Furthermore, the ΔGH* is calculated to determine the activity of catalysts during the adsorption of hydrogen atoms. According to the DFT calculation results, the ∆GH* of Pt–Ni3S2 (− 0.19 eV) is closer to 0 eV compared with Ni3S2 (0.46 eV) and Pt foil (− 0.30 eV), demonstrating a more favorable H* desorption (Fig. 5h), probably owing to the fact the electron-enriched Pt atoms at Pt–Ni3S2 are not easy to be oxidized and can efficiently adsorb hydrogen species to obtain more moderate H binding energy. This could dramatically facilitate the conversion of intermediates and desorption of H2, thereby improving the performance of the catalyst for long-term stable alkaline electrolysis of aquatic hydrogen [6, 74, 75].

Fig. 5
figure 5

Electrocatalytic performance of all examined catalysts for HER without iR-compensation. a LSV curves; b bar diagram representing the overpotentials at different current densities; c Tafel curves; d the calculated theoretical values and the measured H2 amount; e the HER stability test after 48 h by the LSV curves; f the adsorbed H2O on Pt–Ni3S2 model structures; g the adsorbed H on Pt–Ni3S2 model structures; h the H2O and H adsorbing free energy diagrams on catalyst surfaces

Full size image

3.4 Overall Methanol Splitting Performance

Considering the excellent activity and stability of Pt–Ni3S2 for HER and MOR, we believe that Pt–Ni3S2 can be an excellent bi-functional catalyst in the decomposition of methanol- water. Hence, a two-electrode cell using Pt–Ni3S2 electrocatalysts as both electrodes were established in 1.0 mol L−1 KOH with the presence of 1.0 mol L−1 methanol (Fig. 6a). As indicated in Fig. 6b, the cell voltage for Pt–Ni3S2 in the methanol–water electrolyzer is merely 1.71 V to drive a current density of 100 mA cm−2, which is 176 mV lower than that in the overall water electrolyzer, demonstrating more efficient hydrogen production with the assistance of methanol selective upgrading. At the same time, the electrolytic performance is also compared with different dual-electrodes systems including Ni3S2||Ni3S2, and 20% Pt/C||20% Pt/C (Fig. 6c). The cell voltages reaching 50 mA cm−2 for the latter two systems are 1.78 V and 1.82 V, respectively, which are all higher than the Pt–Ni3S2 || Pt–Ni3S2 system (1.60 V). As the current density increases, the voltage difference between Pt–Ni3S2 and other comparison samples will be further increased. Furthermore, the 36 h’ chronoamperometry (I–t) measurement has been implemented by cyclically refilling the fresh electrolyte (1.0 mol L−1 KOH + 1.0 mol L−1 methanol). The I–t curve in Fig. 6e indicates that the Pt–Ni3S2 shows the advantages of excellent long-term stability and cyclic reusability for methanol–water co-electrolysis, demonstrating its good prospect of practical applications.

Fig. 6
figure 6

Schematic illustration of two-electrode system. a Illustration of the assembled electrocatalytic system of methanol–water electrolyzer; b Comparison of the overall water splitting and methanol–water co-electrolysis by using Pt–Ni3S2; c Comparing the co-electrolytic performances with other electrocatalysts; d the photo of the two-electrode configuration during operation; e the stability test of Pt–Ni3S2

Full size image

4 Conclusion

In summary, the dual-monodispersed Pt–Ni3S2 heterojunction nanocrystals (“DMD Pt–Ni3S2 HNCs”) with good dispersion rich interface defects are constructed as highly active electrocatalysts by anchoring platinum on Ni3S2 nanocrystals through the injection method. The “DMD Pt–Ni3S2 HNCs” with fully exposed active sites exhibit excellent bi-functional activity and stability towards HER and MOR, which require only 1.45 V (vs. RHE) to achieve 100 mA cm−2 for MOR and a low η10 of 61 mV for HER, respectively. Coupled with XAFS and DFT calculations, it shows that the electronic interactions at the interface of dual-monodispersed heterojunctions result in an asymmetrical charge distribution at Pt–Ni3S2 interface. On the one hand, the positive charge-enriched Ni3S2 area could promote and stabilize high-valent Ni sites to effectively optimize and facilitate the oxidation process of reaction intermediates, resulting in high electrocatalytic activity and selectivity for MOR. On the other hand, the negative charge-enriched Pt side is responsible for optimizing the H* conversion and H2 desorption to accelerate water dissociation, improving performance for HER. Further, an alkaline electrolysis cell of Pt–Ni3S2||Pt–Ni3S2 exhibits outstanding activity, which requires a low cell voltage of 1.60 V to drive a current density of 50 mA cm−2. The construction of heterostructured interfaces to modulate surface charge distribution provides a new pathway for superior bi-functional electrocatalysts to achieve the concurrent production of value-added formate and hydrogen.