Synergistic effect of nano-Pt and Ni spine for HER in alkaline solution: hydrogen spillover from nano-Pt to Ni spine

The design of active, stable, and cost-effective electrocatalysts for the H2 evolution reaction (HER) in alkaline conditions is important for electrochemical systems such as the chloro-alkaline process and H2 production. Here we report catalysts comprising Pt on Ni single crystalline spines (Pt/Ni-SP) with high activity and stability for HER in alkaline solution with proposed mechanism. The Pt/Ni-SP catalysts are prepared by dispersing platinum nanoparticles (1.7–3.1 nm) on the single-crystalline spines (Ni-SP) of Ni urchin-like particles. The size and coverage of Pt nanoparticles on Ni-SP are increased with increases in the Pt loading amount. X-ray diffraction, high-resolution transmission electron microscopy, X-ray photoelectron spectroscopy, and X-ray absorption spectroscopy are performed to observe the structure of the Pt/Ni-SP catalyst. The catalysts achieve the mass activity of 1.11 A mg−1(Pt), comparing favorably to Pt/C catalysts with the mass activity of 0.33 A mg−1(Pt) at 0.05 V overpotential. The Tafel slope of the Pt/Ni-SP catalyst is approximately 30 mV dec−1, similar to that of Pt, while Pt/Ni-SP is very stable in alkaline solution, like Ni. The synergistic effect of Pt/Ni-SP is ascribed to H spillover from Pt to Ni.

(Ni-SPs) are used as support. Pt nanoparticles are uniformly dispersed over the Ni-SP with controllable coverage, size, and loading level. The Ni particles are highly conductive because of the high crystallinity of the Ni-SP. The preparation is simple, entailing Pt impregnation of Ni particles. The Pt nanoparticle-loaded Ni-SP catalyst exhibits much improved activity and stability in the HER in alkaline conditions compared to that of a commercial Pt/C catalyst (40% Pt on Vulcan XC72).

Results and Discussion
Deposition of Pt nanoparticles on urchin-like structures. Pt islands of various sizes and thicknesses are assembled on urchin-like nickel structures. Four catalysts are prepared and named as 0.75Pt/Ni-SP, 1Pt/Ni-SP, 2Pt/Ni-SP, and 5Pt/Ni-SP for 0.75, 1, 2, and 5 mol% Pt loading on the surface of the Ni metal, respectively. The base Ni particles have urchin-like structures ( Fig. 1(a1-d1)) and the Ni spines (SPs) of the particles show well-developed single-crystalline structures ( Fig. 1(a3-d3)). The Pt particle size of the Pt/Ni-SP catalysts is increased as the Pt loading amount increases ( Fig. 1(a2-d2)): ~1.8, ~2.0, ~2.3, and ~2.8 nm for 0.75Pt/Ni-SP, 1Pt/Ni-SP, 2Pt/ Ni-SP, and 5Pt/Ni-SP, respectively. The transmission electron microscopy (TEM) images show the good dispersion of Pt particles on the Ni-SPs. A particle size histogram was constructed for all prepared catalysts (Fig. S2). The compositions and distributions of the Pt particles were also observed by high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) images and corresponding energy-dispersive X-ray spectroscopy (EDS) mapping (Fig. S3), which showed Pt only on the surface of Ni. The EDS line profile in Fig. S4 also confirmed that the Pt particles were dispersed on the surface of Ni-SP. Table S1 shows the Pt loading amount as determined by an inductively coupled plasma optical emission spectroscopy (ICP-OES) analysis. The ICP-OES result confirmed that the designed Pt loading amount corresponded well to the actual loaded Pt amount.
X-ray diffraction (XRD) analysis is performed to observe the crystallinities of the prepared catalysts, as shown in Fig. 2 200), and (220) planes, respectively, indicating that no structural change occurs in the Ni particles during the process of Pt loading. The Pt (111) plane reflection is clear in the patterns of 5Pt/Ni-SP and 2Pt/Ni-SP, but not in those of 1Pt/ Ni-SP and 0.75Pt/Ni-SP because of the low Pt loading amounts. Using Scherrer's formula, the average particle sizes can be calculated. For the Ni urchin-like substrates, the average crystal size remains at 25.7 nm for all the prepared catalysts, while the crystal size of Pt increases as the Pt loading is increased. The Pt average crystal sizes are recorded as 1.67 and 2.8 nm for 2Pt/Ni-SP and 5Pt/Ni-SP, respectively, slightly smaller than the sizes observed in high-resolution TEM (HR-TEM) analysis.
X-ray photoelectron spectroscopy (XPS) measurements were performed to elucidate the surface compositions and chemistries of the prepared catalysts. The spectra of Ni 2p and Pt 4 f are shown in Fig. S1. The results are summarized in Table 1. For all the pristine samples, the binding energies for Pt 4f 7/2 and Pt 4f 5/2 are observed at 71.4 and 74.6 eV, respectively, corresponding to the Pt 4f 7/2 and Pt 4f 5/2 levels with a binding energy difference of 3.2 eV and indicating the metallic state of Pt 18 . After the deconvolution of Ni 2p 3/2 , three peaks are observed at 852.6 eV, 854.2 eV, and 856.2 eV, corresponding to Ni (0) , NiO, and Ni(OH) 2 , respectively 24,25 . The content of Ni(OH) 2 is near 50% on the surfaces of all prepared samples; that of NiO also remains consistent at approximately 30%. However, the metallic content slightly decreases as the amount of Pt loading is increased, confirming that the Pt is attached to Ni metallic sites. The XPS analysis of the used catalyst after reaction of 50 ks at −1.5 V vs. Hg/HgO showed no NiO peak (Fig. S11), because the passivated Ni(OH) 2 and NiO on the surface of Ni were reduced to metallic Ni during HER in alkaline conditions, as described by MacDougall et al. 26 . The XPS analysis of the used catalyst also confirmed the reduction of Pt oxides to Pt metal; after deconvolution, no oxide peaks were observed for the loaded Pt catalyst. On the surface of the catalyst, the amount of Pt atoms increases as the loading of Pt is increased; Pt atoms occupy 15%, 30%, and 47% of the Ni surface for 1Pt/Ni-SP, 2Pt/Ni-SP, and 5Pt/Ni-SP, respectively. X-ray absorption near-edge structure (XANES) and extended X-ray absorption fine-structure (EXAFS) measurements are performed to obtain insight on the local structure of the Pt-loaded Ni-SP catalyst in order to determine the detailed mechanism of electrochemical HER on the catalyst. XANES provides information on the oxidation states based on edge positions and white line intensity. Figures 3(a) and S5(a) clearly exhibit no substantial change in the edge positions of the prepared Ni catalysts compared to that of standard Ni metal. The Ni K-edge position remains at 8333-8334 (eV), indicating a small amount of Ni + 2 present on the surface of Ni, as observed in the XPS analysis. For large amounts of NiO on the surface of Ni, the absorption K-edge would be shifted to a higher value reaching 8344-8347 (eV) 27 . The white line intensities, positions of the Ni K-edges, and shapes of the EXAFS spectra for all Pt/Ni-SP samples exhibit no noticeable changes, as shown in Fig. S5 (a,b), indicating that the electronic and atomic structures are similar for all samples.
X-ray absorption spectroscopy (XAS) studies were mainly performed to observe the local structures and oxidation states of the Ni and Pt atoms in the proposed catalyst. In Fig. 3(a), and Fig. 3(b), Ni K-edge XANES and corresponding EXAFS spectra of the 5Pt/Ni-SP catalyst are compared with those of Ni metal, respectively. No noticeable differences appear between these two sets of spectra. Therefore, the EXAFS spectra of the Ni and Pt atoms in the 5Pt/Ni-SP catalyst are fitted with theoretical standards of the Ni K-edge, as shown in Fig. 3(c), and the Pt L 3 -edge, as shown in Fig. 3(d). In the Ni K-edge EXAFS spectra, the first prominent peak appears at 2.11 Å. The fitting results of the Ni K-edge EXAFS show that this peak represents the Ni-Ni coordination shell with the bond length of 2.483 Å 28 . The other low-intensity peaks, labeled as Ni-Ni2, Ni-Ni3, and Ni-Ni4, are present in the Ni K-edge EXAFS spectra at higher R values and represent higher Ni-Ni coordination shells in the Ni metal. The Ni K-edge EXAFS spectrum shows good agreement with that of Ni metal and no representative peak for Ni-Pt bonds is observed. The Pt L 3 -edge fit with Pt standard metal also shows good agreement. The Pt L 3 -edge EXAFS fitting for 5Pt/Ni-SP is shown in Fig. 3d, with results presented in Table 2. The coordination number of Pt-Pt in the first shell is 8.34, less than the theoretical value of 12. This low coordination number for Pt is expected because HER Performance. The HER activities of the prepared catalysts were investigated using linear-sweep voltammetry (LSV) at the scan rate of 50 mV·s −1 from 0.15 V vs. a reversible hydrogen electrode (RHE) to −0.57 V vs. RHE at room temperature in 1 M NaOH solution. The same total weights of catalysts were loaded on glassy carbon, with the net weight of Pt varying depending on the amount of Pt loading on the Ni substrate. The Pt loading    amount on the commercial Pt/C catalyst (40% Pt on Vulcan XC 72) was equal to the Pt loading on the 5Pt/Ni-SP catalyst. The specific activity of the 5Pt/Ni-SP catalyst was much higher than that of the Pt/C catalyst; meanwhile, that of the 2Pt/Ni-SP was comparable to that of the Pt/C catalyst. The mass activity of 2Pt/Ni-SP was approximately 3.15 times higher than that of the Pt/C catalyst, as 2Pt/Ni-SP contains 2.5 times less Pt than the Pt/C catalyst. The specific activity of the prepared Pt/Ni-SP catalysts is increased as the Pt loading amount is increased, as shown in Fig. 4(a). The initial activities of 5Pt/Ni-SP, 2Pt/Ni-SP, 1Pt/Ni-SP, and 0.75Pt/Ni-SP are −317 mA cm −2 , −265 mA cm −2 , −228 mA cm −2 , and −164 mA cm −2 , respectively, while that of Pt/C is −245 mA cm −2 .
The stability of a catalyst is among the most important features from a practical perspective. The stabilities of the prepared samples are investigated at the potential of −1.5 V vs Hg/HgO for 100 ks (Fig. 4(b)). Commercial Pt/C begins to deactivate; at 10 ks, 34% of the initial activity is lost, as reported in the literature 10 , while 57% of the initial activity is lost overall before 100 ks of the reaction. Huajie et al. observed a 39.4% loss in activity for the first 4 ks of a 20% commercial Pt/C catalyst 31 .
For the Pt/Ni-SP catalysts, all specimens show initial increases in activity, a common phenomenon for the Ni-based HER catalysts because surface NiO present on the surface is reduced, as has been explained previously 26 . The reduction of surface NiO to Ni metal provides more active surface area for electrocatalysis and more vacant sites for H 2 diffusion. XPS analysis of the used catalyst also confirmed the absence of NiO on the surface of the catalysts (Fig. S11). The Pt/Ni-SP catalysts with Pt loading >1 mol% show some deactivation after a few kiloseconds. To investigate this deactivation, TEM analysis of the used catalyst was performed; agglomerations of Pt particles on the Ni base are found for 2Pt/Ni-SP (Fig. S9) and 5Pt/Ni-SP (Fig. S10), while in the low-loading catalysts  of 1Pt/Ni-SP (Fig. S8) and 0.75Pt/Ni-SP (Fig. S7), no agglomeration was observed. In 5Pt/Ni-SP and 2Pt/Ni-SP, after this initial degradation, no further deactivation occurs before 100 ks. Therefore, the initial deactivation of the 2Pt/Ni-SP and 5Pt/Ni-SP catalysts is ascribed to Pt sintering.
Tafel slopes and exchange current densities of all the Pt/Ni-SP catalysts are calculated from Fig. 4(c) and tabulated in Table 3. From the values, increased loading of Pt is clearly correlated with increased exchange current values. The Tafel slope is a potential-dependent property 32 , so all the Tafel slopes were calculated in the range of 10 to 30 mV of applied overpotential. The Tafel slope value of the Pt/Ni-SP catalysts with the Pt amounts of (0.75, 1, 2, and 5) mol% is ~30 mV dec −1 , while that of the 0.5Pt/Ni-SP with 0.5 mol% Pt loading is ~115 mV dec −1 . The Tafel slopes of Ni-SP and Ni catalysts have been reported as ~120 mV dec −1 previously 23,33 ; therefore, 0.5Pt/Ni-SP behaves like a metallic Ni catalyst. The Tafel slope of ~30 mV dec −1 with a Ni catalyst is impossible at the low overpotential of 30 mV, as described by Lasia 34 . The chemisorption energy of Pt that results in minimum Gibbs's free energy, H 2 adsorption on Pt, is essentially thermoneutral and the dissociation of water is much easier on Pt surfaces than on Ni 35 . The adsorption of H 2 on Ni is endothermic reaction 36 and more energy is required to overcome the H 2 adsorption barrier than that on Pt 37 ; therefore, only Pt can show the low Tafel slope of ~30 mV dec −1 in the low overpotential range of 10 to 30 mV. Hence, it is plausible that the Pt/Ni-SP catalysts with Pt amounts >0.5 mol% behave like Pt in mechanistic aspects. From the Tafel slopes, the Tafel recombination reaction is the rate-determining step for HER in both the Pt/C and Pt/Ni-SP catalysts 38 with Pt amounts >0.5 mol%, while a Volmer reaction is the rate-determining step for Ni and 0.5Pt/Ni-SP. To confirm the stable activity of the catalysts, the kinetic parameters (Tafel slope and exchange current densities) are also observed after 50 ks of HER at −1.5 V vs. Hg/HgO (Fig. 4(d)) and tabulated in Table 3; these kinetic parameters corroborate the stability test results. The exchange current densities of the low-Pt-loaded catalysts (0.75Pt/Ni-SP and 1Pt/Ni-SP) are increased because the catalyst is activated by the reduction of NiO and Ni(OH) 2 from the surface of Ni-SP; meanwhile, the exchange current densities of 2Pt/Ni-SP and 5Pt/Ni-SP are decreased after 50 ks of reaction by Pt agglomeration-induced deactivation.
Cyclic voltammetry (CV) analysis is performed to observe the H 2 adsorption and desorption phenomena ( Fig. 5(a)). For CV analysis, the current is normalized to the total weight of Pt loaded, allowing clear observation. In the range of H 2 adsorption and desorption, urchin-like Ni supports show no current peaks and the double-layer current is also very low. All the Pt/Ni-SP catalysts exhibit increases in electrochemical surface area (ECSA) as the amount of Pt loading approaches 1 Pt mol%; the ECSA decreases with further increases in Pt loading. The ECSA of 5Pt/Ni-SP with the average Pt particle size of 2.8 nm is much larger than that of Pt/C with the average Pt particle size of 2.8 nm and equal Pt loading. The 5Pt/Ni-SP catalyst shows two different peak positions for the under-and overpotential desorption steps, but for Pt/Ni-SP catalysts with lower Pt loadings, both peaks are merged and broadened, indicating much larger ECSA while the Pt particle size is decreased as well. However, distinctive H 2 desorption peaks on Pt/C with the Pt particle size of 2.2 nm have been reported in the literature 39 . Therefore, the high ESCA and disappearance of these distinctive H 2 desorption peaks cannot be attributed to the  Table 3. Catalytic performance of prepared catalysts in 1-M NaOH solution for HER. small Pt particles alone. It is plausible that H spillover may continuously generate vacancy active sites at the Pt surface; the total ECSA would increase with increasing H spillover 40 . The H spillover mechanism is supported by the diffusional desorption peak of H prior to the reversible desorption peaks of H 2 41 . H desorption begins at much lower voltages on the Pt/Ni-SP catalysts than on the Pt/C catalyst, at 0.27 and 0.37 V vs. RHE, respectively, as shown in Fig. 5(a). Mavrikakis et al. have explained the relation between the binding energy of adsorbed H and the diffusion activation energy; the diffusional barrier is approximately12% of the binding energy of the adsorbed state 42 . At 298 K, the activation energy of surface H diffusion on Pt ranges from 10 kJ mol −1 to 25 kJ mol −1 , yielding the diffusion coefficient of 10 −5 cm 2 s −1 and causing instantaneous H migration at room temperature; by these steps, H spillover can be rationalized 43 . Furthermore, if the rate-determining step is the Tafel recombination reaction, H spillover can boost H 2 production 44 . Electrochemical impedance spectroscopy is also performed and discussed in supplementary information (Fig. S13 and Table S2) that corroborates with the kinetic parameters obtained from Tafel slope.
The Mass activities of the prepared catalysts at the overpotential of 0.05 V vs. RHE are also observed ( Fig. 5(b)) and show good agreement with the CV results for the ECSA. Overpotential of 0.05 V is selected for the mass activity measurement because the onset potential of Ni-SP was observed at 0.078 mV, so the activity contribution of Ni-SP at 0.05 V is negligible. The Pt-based initial mass activities of the prepared catalysts are 0.33 A mg −1 , 0.6 A mg −1 , 0.95 A mg −1 , 1.04 A mg −1 , and 0.55 A mg −1 for Pt/C, 0.75Pt/Ni-SP, 1Pt/Ni-SP, 2Pt/Ni-SP, and 5Pt/ Ni-SP, respectively. After the 50 ks of stability testing, these values have changed to 0.15 A mg −1 , 0.91 A mg −1 , 1.11 A mg −1 , 0.95 A mg −1 , and 0.47 A mg −1 , showing the same trend as the ECSA. In addition, the Ni-SP support enhances the stability of the loaded Pt. From the above data we can conclude that, for the best stability and activity with the least amount of Pt, 1Pt/Ni-SP is the best catalyst among those designed and tested here.

Conclusion
Here we have designed a catalyst by the addition of Pt on crystalline Ni spines in ethylene glycol. The sizes of the Pt particles on the Ni-SP supports ranged from 1.8 nm to 2.8 nm, depending on the amount of Pt loading. The mass activity was optimized with the 1 mol% Pt, 1Pt/Ni-SP catalyst have much higher mass activity than that of the commercial Pt/C catalyst. TEM and EXAFS analysis clearly showed that the isolated Pt particles were well dispersed on the crystalline Ni-SP supports. The rate-determining step for HER over the Pt/Ni-SP catalysts was changed from the Volmer reaction to the Tafel recombination reaction as the Pt loading on the Ni-SP was increased. Furthermore, a large ECSA was observed with the 1Pt/Ni-SP catalyst, attributed to H spillover from Pt to Ni. Ni metal has higher stability but lower activity than Pt. The lower activity of Ni arises from the slow rate of water dissociation (the Volmer step); the split H atoms on Pt and Ni recombine on Ni to produce H 2 . Therefore, it can be concluded that the small Pt particles on Ni-SP enhanced the catalyst mass activity, which was further enhanced by H spillover from the Pt surface to the crystalline Ni surface. No Pt agglomeration was observed with the 1Pt/Ni-SP catalyst, even after 50 ks of HER. By combining theory with practical results, the mass activity of Pt was increased from 0.33 A mg −1 in the commercial Pt catalyst to 1.11 A mg −1 for 1Pt/Ni-SP.
Pt nanoparticle deposition on Urchin-like Ni structures. For the synthesis, 58.7 mg (1 mmol) of urchin-like Ni nanoparticles and 0.0075, 0.01, 0.02, and 0.05 mmol of potassium hexachloroplatinate were dissolved in 20 mL of ethylene glycol and held for 3 h in a rotary reactor at room temperature to allow the complete dissolution and impregnation of Pt precursor on the Ni nanoparticles. After 5 min of bath sonication, the homogeneous solution was heated in a water bath at 90 °C for 1 h. After the reaction, the ethylene glycol was removed by centrifugation at 8000 rpm and washed with deionized water several times to remove residual ethylene glycol; the product was then dried in a vacuum oven at 25 °C for 5 h. For the deposition of Pt nanoparticles on the urchin-like Ni nanoparticles, the NPs were prepared in one batch. For ease of description, the prepared Pt-loaded catalysts are named as 0.75Pt/Ni-SP, 1Pt/Ni-SP, 2Pt/Ni-SP, and 5Pt/Ni-SP for 0.75, 1, 2, and 5 mol% Pt loading, respectively.
Characterization. Transmission electron microscopy (TEM) and high-resolution transmission electron microscopy (HR-TEM) were performed using a FEI Tecnai TEM G 2 (accelerating voltage of 200 kV). Powder X-ray diffraction (XRD) was performed using a Bruker D8 Avance diffractometer using Cu Kα radiation (λ = 1.5406 Å). XPS analyses were recorded on a PHI 5000 VersaProbe (Ulvac-PHI) with the background pressure of 2 × 10 −7 Pa and the spot size of 100 μm × 100 μm at an angle of 45°, wide scan pass energy of 117.4 eV and narrow scan pass energy of 46.95 eV. The concentrations of the catalysts were analyzed using Thermoscientific iCAP7000. X-ray absorption (XAS) data was collected at the 1D KIST-PAL beamline in the Pohang light source (PLS-II). The storage ring was operated at 2.5 GeV with an injection current of 360 mA. The intensity of the incident X-ray absorption I o was monitored with a N 2 -filled ionization chamber. The spectral energies were calibrated by using the first inflection points of Ni (8333 eV) and Pt (11564 eV) metal foil spectra as references. XANES data preprocessing and EXAFS data fitting with theory were performed by using the IFEFFIT-based program Demeter based on FEFF code 45,46 . Electrochemical measurements. All the electrochemical measurements were performed in a three-electrode cell at room temperature (25 °C), with Hg/HgO as the reference electrode and a Pt foil as the counter electrode, respectively. The working electrode was prepared with a glassy carbon rotating disk of 5 mm in diameter as the substrate. Typically, a mixture containing 4.0 mg of the electrocatalyst in a solvent of 0.9 mL Water and 0.1 mL 5 wt% Nafion solution (5 wt%, DuPont, USA) was ultrasonicated for 5 minutes to obtain a well-dispersed ink. The catalyst ink was then quantitatively transferred onto the surface of the glassy carbon electrode and dried at room temperature to obtain a thin film of the catalyst. The total catalyst loading was 0.204 mg cm −2 . The Pt/C commercial catalyst was deposited at 0.031 mg cm −2 , equivalent to the amount of Pt in the 5Pt/ Ni-SP catalyst. The electrochemical tests were performed in Ar-saturated 1 M NaOH aqueous solution at room temperature. CV scanning was performed at the potential scan rate of 50 mV s −1 in the range of 0.005 V to 0.93 V vs. RHE to determine the electrochemical surface areas. Linear sweeps to obtain the Tafel slopes were performed at the potential scan rate of 1 mV s −1 in the range of −60 mV to 0 mV vs RHE while the hydrogen evolution reaction polarization curve was obtained at the scan rate of 50 mVs −1 in the range of 0.2 V to −0.57 V vs Hg/HgO. Chronoamperometric analysis was also performed at −1.5 V (vs. Hg/HgO) for 50 ks and 100 ks. All the electrochemical studies were performed on an (Ivium Technologies, Iviumstat), coupled with a rotating-disk electrode (PINE-AFMARCE-5 mm dia).