Structure- and Morphology-Controlled Synthesis of Hexagonal Ni2–xZnxP Nanocrystals and Their Composition-Dependent Electrocatalytic Activity for Hydrogen Evolution Reaction

Nickel phosphides are an emerging class of earth-abundant catalysts for hydrogen generation through water electrolysis. However, the hydrogen evolution reaction (HER) activity of Ni2P is lower than that of benchmark Pt group catalysts. To address this limitation, an integrated theoretical and experimental study was performed to enhance the HER activity and stability of hexagonal Ni2P through doping with synergistic transition metals. Among the nine dopants computationally studied, zinc emerged as an ideal candidate due to its ability to modulate the hydrogen binding free energy (ΔGH) closer to a thermoneutral value. Consequently, phase pure hexagonal Ni2–xZnxP nanocrystals (NCs) with a solid spherical morphology, variable compositions (x = 0–17.14%), and size in the range of 6.8 ± 1.1–9.1 ± 1.1 nm were colloidally synthesized to investigate the HER activity and stability in alkaline electrolytes. As predicted, the HER performance was observed to be composition-dependent with Zn compositions (x) of 0.03, 0.07, and 0.15 demonstrating superior activity with overpotentials (η–10) of 188.67, 170.01, and 135.35 mV, respectively at a current density of −10 mA/cm2, in comparison to Ni2P NCs (216.2 ± 4.4 mV). Conversely, Ni2–xZnxP NCs with x = 0.01, 0.38, 0.44, and 0.50 compositions showed a notable decrease in HER activity, with corresponding η–10 of 225.3 ± 3.2, 269.9 ± 4.3, 276.4 ± 3.7 and 263.9 ± 4.9 mV, respectively. The highest HER active catalyst was determined to be Ni1.85Zn0.15P NCs, featuring a Zn concentration of 5.24%, consistent with composition-dependent ΔGH calculations. The highest performing Ni1.85Zn0.15P NCs displayed a Heyrovsky HER mechanism, enhanced kinetics and electrochemically active surface area (ECSA), and superior corrosion tolerance with a negligible increase of η–10 after 10 h of continuous HER. This study provides critical insights into enhancing the performance of metal phosphides through doping-induced electronic structure variation, paving the way for the design of high-efficiency and durable nanostructures for heterogeneous catalytic studies.


■ INTRODUCTION
Hydrogen generation via electrocatalysis-enabled water splitting offers an exciting opportunity to decrease our reliance on fossil fuels and address critical issues with energy sustainability.However, large-scale hydrogen production is hindered by the availability of earth-abundant catalysts with comparable or higher activity to expensive noble metal catalysts.Nickel phosphides have emerged as a burgeoning noble metal free alternative for the hydrogen evolution reaction (HER) because of high activity and stability compared to other nonnoble metal catalysts. 1−3 However, they do not exhibit low overpotentials (η) for HER and oxygen (OER) evolution and high corrosion tolerance in acid and alkali compared to noble metal catalysts.−10 The crystal structure of Ni 2 P comprises two alternating stoichiometric planes, Ni 3 P and Ni 3 P 2 , along the [0001] direction. 11The Ni-rich Ni 3 P 2 surface, exposed by the Ni 2 P (001) facet, is the most active surface for reversible binding of hydrogen, providing highly active Ni sites (hydride-acceptor) and less active P sites (proton-acceptor) for the water splitting reaction. 12However, the catalytic activity of the Ni 2 P (001) surface is not optimal for water splitting, because the Ni sites bond strongly to hydrogen, impeding the facile removal of H 2 . 12Therefore, an increase in surface metal sites along with systematic modulation of the surface affinity is needed to further improve the HER activity, which can be achieved via admixing of synergistic metal atoms.To understand how different dopants and doping-induced structures impact the HER activity for designing efficient electrocatalysts, theoretical calculations are essential.−20 Recent literature indicates that nickel phosphides doped with V, 14,21,22 W, 23 Al, 24 Mo, 25 Fe, 26 Co, 26 and Mn 27 show improved HER performance in comparison to monometallic counterparts.For instance, V-doped Ni 2 P nanosheets (NS) showed a notable increase in HER activity with an overpotential (η) of 85 mV compared to Ni 2 P NS (143 mV) at a current density (j) of −10 mA/cm 2 (η −10 ). 21Similarly, Mndoped Ni 2 P NS showed enhanced HER performance (η −20 = 185 mV) in comparison to Ni 2 P NS (η −20 = 103 mV). 27A composition-dependent study exploring the HER activity of Ni 2−x Mo x P nanocrystals (NCs) reported a prominent increase in HER performance for homogeneously alloyed Ni 1.87 Mo 0.13 P NCs (η −10= 101 mV) in comparison to Ni 2 P NCs (244 mV). 13nterestingly, the larger heterogeneous Ni 2−x Mo x P NCs displaying the same dopant concentration (i.e., Ni 1.87 Mo 0.13 P) showed a higher HER activity (η −10 = 96 mV) in comparison to heterogeneous Ni 2 P NCs (η −10 = 156 mV). 28In both cases, the Ni 1.87 Mo 0.13 P composition displayed the lowest Tafel slopes, which were comparable to benchmark Pt/C catalyst, suggesting a pronounced increase in HER kinetics upon heteroatom doping.These reports illustrate the importance of optimal dopants and their critical concentration for modulating the surface affinity of nickel phosphides in the development of high-efficiency and earth-abundant HER electrocatalysts.Although the influence early transition metal dopants have been extensively studied, a few experimental studies on late transition metal-doped Ni 2 P for HER have been reported. 20,29Moreover, a systematic study on the influence of late transition metal dopants (i.e., Zn and Cu) on HER activity and stability of nickel phosphides, to our knowledge, has not been reported.This work aims to increase the arsenal of transition metal-doped Ni 2 P catalysts by investigating the composition-dependent effects on the HER activity and stability of discrete Zn-doped Ni 2 P NCs.
Herein, a comprehensive computational analysis of nine synergistic dopants (Cr, Zn, Mo, Mg, Fe, V, Cu, Ti, and W) was conducted to evaluate their influence on the hydrogen adsorption free energy (ΔG H ) on the hexagonal Ni 2 P (001) surface.Density functional theory (DFT) analysis indicates Zn as an optimal dopant that promotes H adsorption and desorption near the thermoneutral ΔG H on more energetically favorable binding sites.Accordingly, a series of phase pure Ni 2−x Zn x P NCs with hexagonal crystal structure, solid spherical morphology, and variable compositions were synthesized to investigate the compositional effects on HER activity in alkaline electrolytes.Binary Ni 2 P NCs displayed η −10 of 216.2 ± 4.4 mV, consistent with literature reports. 28In contrast, Ni 2−x Zn x P NCs showed significantly improved HER performance and a progressive decrease of η −10 from 188.7 ± 3.7, 170.0 ± 4.2, and 135.4 ± 3.4 mV for x = 0.03, 0.07, and 0.15 compositions, respectively.At concentrations above 5.24% Zn (x = 0.15), bimetallic NCs showed a notable decrease in HER activity compared to monometallic Ni 2 P NCs.The catalyst with the highest HER activity was determined to be Ni ■ EXPERIMENTAL SECTION Materials.Nickel(II) 2,4-pentanedionate (Ni(acac) 2 , 95%) and bis (2,4-pentanedionato), zinc(II) (Zn(acac) 2 , 96.0%), oleylamine (OLA, 70%), toluene, and ethanol were purchased from Fisher.Trin-octylphosphine (TOP, 97%) was purchased from Strem Chemicals.Carbon-coated 200 mesh copper grids were purchased from SPI Supplies.Titanium foil (thickness 0.25 mm; 99.7%) and platinum on graphitized carbon (Pt/C, 10% wt.) were purchased from Sigma-Aldrich.Graphite rods (6.15 mm × 102 mm, 99.9995%) were purchased from Alfa Aesar.Hg/HgO reference electrode filled with 1 P NCs was adopted from a literature procedure. 30Briefly, 1.0 mmol of Ni (acac) 2 , 10.0 mL of OLA, and 5.0 mL of TOP were combined in a 50 mL roundbottom flask under a nitrogen atmosphere and attached to a Schlenk line.This solution was degassed under vacuum at 120 °C for 20 min, purged with nitrogen, and the temperature was raised to 330 °C where it was kept for 2.5 h.Then, the reaction was cooled to room temperature and NCs were isolated with a mixture of toluene and ethanol and centrifuged for 10 min.This purification procedure was repeated a minimum of 3 times.The final NC product was dried under vacuum.
Synthesis of Ni 2−x Zn x P NCs.The synthesis of Ni 2−x Zn x P NCs was modified from the synthesis of Ni 2 P NCs.Ni(acac) 2 and Zn(acac) 2 precursors were combined with the desired nominal molar ratio (Table 1) and added to a round-bottom flask containing 10.0 mL of OLA and 5.0 mL of TOP.The flask was attached to a Schlenk line, and the mixture was degassed under vacuum at 120 °C for 20 min.Then, the flask was purged with nitrogen and the temperature was increased to 330 °C.The solution was kept under reflux at 330 °C for 2.5 h, and the purification procedure remained unchanged.
Physical Characterization.Powder X-ray diffraction (PXRD) patterns were recorded using an Empyrean multipurpose X-ray diffractometer equipped with a Cu Kα (λ = 1.5418Å) radiation source, under 45 kV and 40 mA operating conditions.The Scherrer formula was used to manually calculate the crystallite size based on the hexagonal Ni 2 P (111) reflection.Low-resolution transmission electron microscopy (TEM) images were acquired using a Jeol JEM-1400Plus transmission electron microscope equipped with a Gatan UltraScan 4000SP 4K x 4K CCD camera, operating at 120 kV.Highresolution TEM, scanning (STEM) images, and elemental maps were attained using a JEM-F200 cold FEG electron microscope operating at a voltage of 200 kV and equipped with an energy-dispersive X-ray analyzer.Sample preparation for TEM/high-resolution (HR)TEM involved drop-casting a 10 μL aliquot of a dilute Ni 2−x Zn x P NCtoluene solution onto carbon-coated copper grids and toluene was evaporated in ambient air prior to imaging.Elemental compositions were determined through inductively coupled plasma-optical emission spectroscopy (ICP-OES), using an Agilent Technologies 5110 ICP-OES equipped with a SPS4 autosampler.Samples were digested in acid (3:1 v:v HCl/HNO 3 ) and placed in a heated water bath (∼60 °C) for 24 h and then further diluted with Milli-Q-filtered water prior to analysis.X-Ray photoelectron spectra (XPS) were recorded using a PHI VersaProbe III Scanning XPS Microprobe.NCs were annealed in 5% H 2 :Ar at 450 °C for 2 h prior to XPS analysis.Regional scans were completed using a pass energy of 26.00 eV with 20 ms per step, and the number of sweeps was dependent upon the signal intensity of each element.FT-IR spectra were recorded using a Thermo Scientific Nicolet iS50 FT-IR.
Theoretical Calculations.To investigate the most effective metal-doped nickel phosphides for HER, a comprehensive study on the Ni 2 P (001) surface doped with a series of synergistic metal (Cr, Zn, Mo, Mg, Fe, V, Cu, Ti, and W) atoms was conducted using density functional theory (DFT) within the Vienna Ab initio Simulation Package (VASP) 31 and the Projector-Augmented-Wave (PAW) method.The electrocatalytic performance was assessed by calculating the ΔG H . Utilizing the Bell−Evans−Polanyi principle, we focused on ΔG binding interactions, allowing us to bypass the computationally demanding calculations of reaction barriers. 15The ΔG H at the hydrogen adsorption site was considered a critical descriptor of each catalyst's intrinsic activity. 12An ideal catalyst is designed with a ΔG H value close to 0 eV, aligning with the Sabatier principle, ensuring that hydrogen binds neither too weakly nor too strongly to the surface. 32Therefore, smaller |ΔG H | values indicate a higher HER activity. 33Because of the influence of ΔG H values from the computational level, 16−18 |ΔG H | <0.1 eV was adopted as the optimal criterion 32 to identify the most active heteroatom-doped Ni 2 P catalyst.
Here, ΔE, ΔE ZPE , T, and ΔS indicate the electronic adsorption energy obtained through DFT, zero-point energy, temperature, and entropy contributions, respectively.A commonly utilized factor of 0.24 eV 22,36 was applied to encompass all contributions in entropic and zero-point energy (ΔE ZPE − TΔS).Consequently, ΔG H was simplified to ΔE + 0.24 eV.The Perdew−Burke−Ernzerhof (PBE) functional 37 was employed to obtain structures, while PBE with the Grimme D3(BJ) dispersion correction 38,39 was employed for ΔE calculations.A periodically repeated slab with six layers and a 15 Å vacuum (cell parameters a = b = 5.891 Å and c = 25.120Å) was employed, with atoms at the bottom layer fixed to their bulk positions, to model ΔG H on the Ni 2 P (001) surface with Ni 3 P 2 termination.The emphasis on the Ni 3 P 2 termination was due to it being the most active Ni 2 P surface termination. 40,41A plane wave cutoff energy of 520 eV and a Γ-centered 5 × 5 × 1 grid of k-points were utilized.In our HER calculations, one Ni atom was substituted with one of the nine doped metals to identify the most stable doping site, resulting in a 5.56% dopant concentration.We also investigated the dependence of ΔG H on Zn concentration.For the 11.11% Zn concentration, we maintained the same cell parameters used for the 5.56% Zn concentration but replaced two Ni atoms with Zn atoms to find the most stable structure.To generate a 2.78% Zn computation, a larger cell (a = 11.782Å, b = 5.891 Å, and c = 25.120Å) was created by substituting one Ni with a Zn atom, and the most stable structure was then determined.To validate a direct comparison of ΔG H bindings in two cells with different Zn concentrations, the first and second ΔG H bindings on the pristine Ni 2 P (001) surface with Ni 3 P 2 termination were also calculated.In the smaller cell, the first and second ΔG H bindings were −0.42 and 0.15 eV, respectively.Correspondingly, for the larger cell, bindings were −0.47 and 0.15 eV, respectively.The marginal energy difference observed ensures the validity of comparing the binding energies of two cells with different Zn compositions.
Fabrication of Working Electrodes.Working electrodes were fabricated on Ti foils using Ni 2−x Zn x P NCs and commercial Pt/C catalysts.The foils were cut into rectangles (0.5 cm × 0.4 cm) and then subjected to a multistep cleaning process that includes sonication in a mixture of acetone and ethanol (1:1, v:v) for 10 min, soaking in a mixture of 1 M HCl and 30% H 2 O 2 (1:1, v:v) for 20 min, and rinsing via sonication in Milli-Q water (18 Ω).To fabricate catalyst ink, 4 mg of Ni 2−x Zn x P or commercial Pt/C catalyst was mixed with 380 μL of isopropanol and 20 μL of nafion and sonicated in an ice water bath for 10 min.The catalyst ink (40 μL) was drop cast onto the Ti foil in 10 μL of aliquots.The catalyst-coated Ti foils were then annealed at 450 °C for 2 h under 5% H 2 :Ar to remove residual surface ligands.Then, foils were attached to a Ag-plated Cu wire using Ag paint, which ensures ohmic contact between the Cu wire and Ti foil.A two-part epoxy was used to cover the Cu wire and Ag paint, leaving only the catalyst area (0.20 cm 2 ) exposed for HER studies.The epoxy was left dry in ambient air for 24 h prior to electrocatalytic experiments.
Electrochemical Measurements.A CHI 760E electrochemical workstation was used to study the HER activity of Ni 2−x Zn x P NCs and Pt/C in nitrogen-saturated 1 M KOH at room temperature.A conventional three-electrode electrochemical cell was employed for all experiments where a graphite rod, Hg/HgO (1 M NaOH) electrode, and Ni 2−x Zn x P NCs or Pt/C-coated Ti foil were used as counter, reference, and working electrodes, respectively.Overpotentials were reported with respect to the scale of reversible hydrogen electrode (RHE) potential by using the conversion formula: E RHE = E Exp + E Hg/HgO 0 + 0.05916pH.The current density (j, mA/cm 2 ) was calculated using the geometrical surface area of the electrode (0.2 cm 2 ).Working electrodes were electrochemically cleaned and activated by performing cyclic voltammetry (CV) and sweeping the potential from 0.2 to 0.1 V at different scan rates.Polarization curves were recorded using linear sweep voltammetry (LSV) by scanning the potential from 0.2004 to −0.4996 V (vs RHE) at a scan rate of 5 mV/ s.−44 CVs were recorded within the non-Faradaic region at various scan rates (50, 100, 200, and 400 mV/s), which were utilized to determine the difference in double layer capacitive current.This difference in capacitive current [Δj = (j a − j c )/2] was plotted against the scan rate, and corresponding slopes were generated through linear fit analysis to determine the C DL .Finally, ECSA was calculated by dividing the C DL with specific capacitance (C s = 0.040 mF/cm 2 ). 42,43,45Chronopotentiometry analysis was performed at a current density of −10 mA/cm 2 for 10 h in N 2 -saturated 1 M KOH to investigate the stability of all catalysts.
■ RESULTS AND DISCUSSION Theoretical Calculations.The lowest first and second ΔG H values on Cr, Zn, Mo, Mg, Fe, V, Cu, Ti, and W-doped Ni 2 P (001) surfaces with the Ni 3 P 2 termination are depicted in Figure 1, along with the corresponding configurations in the Supporting Information, Figure S1.The first ΔG H on the pristine Ni 2 P surface at the Ni 3 -hollow site is −0.42 eV, which aligns with the first ΔG H obtained in previous studies. 16,19,41owever, such a large hydrogen binding energy indicates that pristine Ni 2 P is not catalytically active for HER at low hydrogen coverage due to the difficulty in H 2 desorption.The second most stable hydrogen adsorption site is at the Ni−Ni bridge with a second ΔG H of 0.15 eV, consistent with a prior report. 16The binding of the second hydrogen also facilitates the first hydrogen binding at the Ni−Ni bridge site.This observation indicates that the binding of both hydrogens at the Ni−Ni bridge site is more energetically favorable than one hydrogen adsorption at the Ni 3 hollow site and another at the Ni−Ni bridge site.The Ni−Ni bridge site is more catalytically active than the Ni 3 -hollow site, and this behavior can be attributed to the enhanced HER activity of Ni 2 P reported in the literature. 13,46The HER activity of Ni 2 P can be enhanced by reducing ΔG H on the Ni 3 -hollow site and increasing ΔG H on the Ni−Ni bridge site.Therefore, the metal doping strategy, adopted in this study, could prove useful for improving the HER activity of hexagonal Ni 2 P. 2 As shown in Figure 1, doping can substantially modify the HER activity of the Ni 3 P 2 -terminated Ni 2 P (001) surface.The most promising dopants for HER are Cr, Zn, and Mo based on the overall values for the first and second ΔG H . Specifically, Cr doping modifies the first ΔG H from −0.42 (pristine Ni 2 P) to −0.24 eV and the second ΔG H from 0.15 (pristine Ni 2 P) to 0.13 eV.For Cr doping, both the first and second hydrogens occupy Cr−Ni bridge sites.Doping with Zn significantly decreases the first ΔG H to −0.10 eV, but it also results in the second ΔG H becoming more repulsive at 0.33 eV.For Zn doping, the first hydrogen occupies the Ni−Ni bridge site, and the second hydrogen is located on the Ni−P bridge site.Similarly, Mo doping changes the first and second ΔG H to −0.32 and −0.15 eV, respectively.For Mo doping, the first hydrogen still occupies the hollow site, but the second hydrogen locates at the Mo atop site.The other dopants (Mg, Fe, V, Cu, Ti, and W) investigated in this study did not substantially improve the ΔG H in comparison to pristine Ni 2 P.
Synthesis and Characterization of Ni 2 P and Ni 2−x Zn x P NCs.Computational analysis of the nine dopants suggests that Zn would be a promising dopant for hexagonal Ni 2 P. Consequently, Ni 2−x Zn x P NCs were synthesized to experimentally investigate the composition-dependent HER activity in 1 M KOH.The colloidal synthesis of phase pure Ni 2 P NCs involves the thermal decomposition of Ni(acac) 2 in an OLA and TOP system, where OLA acts as a surfactant and TOP acts as both a surfactant and phosphorus precursor.−49 Furthermore, the thermal decomposition using P/Ni molar ratios >2.24 results in Ni x P y nucleating species, which form amorphous particles at 240 °C and then crystallize to form smaller solid particles above 300 °C. 14,32We found that phase pure solid Ni 2 P NCs can be reproducibly synthesized under reaction conditions of 330 °C for 2.5 h with a P:Ni ratio of 11.2.The incorporation of Zn(acac) 2 did not impact the reaction as Ni 2−x Zn x P NCs retained the hexagonal Ni 2 P structure with no evidence of elemental Zn or Zn 3 P 2 impurity phases.
Zn-doped Ni 2 P compositions were investigated using ICP-OES, and the concentration of Zn varied from x = 0.01−0.50,corresponding to 0.41−17.14%experimental Zn concentrations (Table 1).The experimental atomic percent of Zn was repeatedly found to be lower than the nominal moles of Zn(acac) 2 used in the synthesis.For instance, Ni 2−x Zn x P NCs produced with a nominal 10% Zn(acac) 2 typically yields an experimental Zn concentration ranging from 2.60 to 5.24%, whereas NCs synthesized with a nominal 15% Zn(acac) 2 show experimental Zn concentration from 5.50 to 8.61%.In contrast, alloyed NCs with significantly higher Zn (13.23−15.32%)compositions can be targeted with a nominal Zn(acac) 2 concentration of 25%.
Structure, crystallinity, and phase purity of Ni 2−x Zn x P NCs were confirmed through PXRD, and diffraction patterns that show variable Zn compositions (0−17.14%)are shown in Figure 2. All samples displayed a hexagonal crystal structure consistent with the binary Ni 2 P reference pattern (JCPDS file No. 01-074-1385) with no evidence of impurity phases of nickel phosphides.Although no apparent shifts in the diffraction patterns were observed, slight broadening of Bragg reflections was noted and confirmed with increased full width at half maxima (fwhm) of Zn-doped samples.This observation is consistent with an earlier report, where no shifts in Bragg reflections were noted for Ni 5−x Zn x P 4 NCs because of the similar size of Ni and Zn. 50This combined with broader diffraction peaks obtained for Ni 2−x Zn x P NCs make it more challenging to distinguish minor peak shifts.Nevertheless, the effect of Zn doping was evident in the crystallite size estimated by applying the Scherrer formula to the (111) reflection.Overall, it appears that admixing of Zn tends to decrease the crystallite size, as the size decreased from 9.0 nm for binary Ni 2 P to 8.2 and 7.8 nm for x = 0.15 and x = 0.44 compositions, respectively.However, at x = 0.50, crystallite size displayed a deviation from this trend and increased to 8.4 nm.The average particle size calculated from TEM images also reflects an oscillating effect with Zn concentration, where an increase was observed at lower Zn (x <0.15) compositions and then a decrease at higher Zn (x >0.15) compositions.
TEM images of Ni 2−x Zn x P NCs with variable compositions are shown in Figure 3 and in the Supporting Information, Figures S2−S4.Binary Ni 2 P NCs possessed a solid spherical morphology, which was maintained in ternary NCs with Zn content up to 17.14%.While preserving the narrow size dispersity of Ni 2 P NCs, incorporation of Zn displayed an interesting effect on the size of ternary NCs.The average size of binary Ni 2 P NCs was 8.7 ± 1.5 nm.Upon doping with x = 0.01, the average size decreased to 7.4 ± 1.2 nm.However, as x increased to 0.03, a slight increase in size to 9.1 ± 1.1 nm was observed.This trend of slightly larger particles continued with Zn compositions up to 0.15 where the average size was measured to be 8.9 ± 1.4 nm.However, beyond x = 0.15, the average size decreased to 8.5, 6.8, and 7.1 nm at x = 0.23, 0.38, and 0.50 compositions, respectively.Lattice spacing for binary Ni 2 P NCs was found to be 2.2 Å, corresponding to the (111)  reflection, consistent with the literature, 47 whereas a slightly contracted (111) lattice spacing of 1.9 Å was obtained for the highest Zn composition (x = 0.50).Bright field TEM images and STEM-EDS elemental maps of Ni 1.77 Zn 0.23 P NCs are shown in Figure 3J-M and in the Supporting Information, Figure S5.Elemental maps confirm the presence of Ni, Zn, and P in all samples and homogeneous admixing of all elements with no evidence of segregation.This has been confirmed with Ni 2−x Zn x P NCs with variable compositions up to x = 0.50 (Supporting Information, Figures S6 and S7), which support no evidence of segregation at the highest Zn concentration.Surface characteristics were investigated using XPS, and spectra obtained for Ni 2 P, Ni 1.85 Zn 0.15 P, and Ni 1.56 Zn 0.44 P are shown in Figure 4 and in the Supporting Information, Figure S8.In the Ni 2p region for Ni 2 P, a partial positive charge was observed for nickel with the 2p 3/2 peak at 852.96 eV and the 2p 1/2 peak at 870.28 eV.Peaks at 856.43 and 856.30eV represent the presence of NiO and Ni satellites, respectively. 28he P 2p region for Ni 2 P exhibits two doublets at 129.52 and 133.37 eV, which correspond to a metal phosphide 51 and oxidized phosphorus, respectively. 28Ni 1.85 Zn 0.15 P NCs displayed similar characteristics with a partial positive charge on Ni represented by the Ni 2p 3/2 peak at 853.21 eV and Ni 2p 1/2 peak at 870.43 eV.NiO and Ni satellite moieties were also present.The metal phosphide and oxidize P doublets in the P 2p region of Ni 1.85 Zn 0.15 P NCs were demonstrated with peaks at 129.86 and 134.70 eV, respectively.Within the Zn 2p region of the Zn-doped sample, the peak at 1022.66 eV indicates a slight partial positive charge on Zn surface species. 52The higher Zn-doped composition, Ni 1.56 Zn 0.44 P, exhibited similar surface characteristics.Interestingly, as the Zn composition increased from x = 0, 0.15, to 0.44, the partial positive charge on Ni increases as the Ni 2p 3/2 peak shifted from 852.96 to 853.21 then to 853.28 eV, respectively.P also underwent a similar change as the partial negative charge decreased by 0.44 eV, and the P δ− peak shifts from 129.52 to 129.86 then to 129.96 eV, for x = 0, 0.18, and 0.40 compositions, respectively.Although rather dramatic shifts were observed in Ni and P surface moieties as a function of Zn concentration, the partial positive charge on Zn showed a slight shift to higher binding energies (1022.66−1022.70eV) as the x increased from 0.15 to 0.44.Since Zn is less electronegative than Ni, a shift of Ni δ+ to lower binding energies is expected.Instead, the peaks for both Ni δ+ and P δ− shift to higher binding energies.Furthermore, the oxidized P and Ni moieties become more pronounced as the Zn content increases from x = 0 to 0.44, suggesting that the presence of Zn causes Ni 2 P to be more susceptible to surface oxidation, leading to the higher binding energies observed for Ni δ+ and P δ− peaks.
Electrocatalytic Activity of Ni 2−x Zn x P NCs for HER.The HER activity of Ni 2−x Zn x P NCs was investigated using linear sweep voltammetry (LSV) in a three-electrode electrochemical cell in N 2 -saturated 1 M KOH solution.The catalytic activities of bare Ti foil and commercial Pt/C (wt.10%) standard were also investigated for comparison.A range of solvents such as isopropanol, ethanol, isopropanol + ethanol (1:1), isopropanol + water (1:1), and ethanol + water (1:1) were used to fabricate catalyst inks both in the presence and absence of nafion and isopropanol produced consistent HER data for all samples examined.In the presence of nafion, catalysts showed the highest HER activity as it prevents loss from the working electrode when scanning at higher negative potentials.Bulky organic ligands such as OLA and TOP impede the electrical conductivity and obstruct active surface sites resulting in a lower HER activity.Hence, the working electrodes were annealed at 450 °C for 2 h under a 5% H 2 /Ar atmosphere to remove organic surface functionalities while maintaining the hexagonal Ni 2 P structure and composition (Supporting Information, Figure S10).Annealing also improves the ohmic contact between the Ti foil and Ni 2−x Zn x P NCs.The FT-IR spectra of annealed samples confirm the removal of OLA/TOP ligands (Supporting Information, Figure S10), which was evident from the loss of C−H stretches at 2938 and 2859 cm −1 , C�C stretch at 1650 cm −1 , C−N stretch at 1468 cm −1 , and the C−P stretch at 1068 cm −1 .
Nafion-coated Ti foil electrodes showed a negligible HER activity in 1 M KOH and displayed the highest overpotentials (η −1 = 440.7 mV).In contrast, binary Ni 2 P showed significantly higher HER activity, producing η −1 and η −10 at 110.1 ± 2.1 and 216.2 ± 4.4 mV, respectively, which are lower than those reported for phase pure Ni NCs 53 and consistent with literature reports of Ni 2 P NCs. 28PS spectra of Ni 2−x Zn x P NCs suggest a shift in Ni δ+ and the P δ− peaks to higher binding energies with increasing Zn composition.However, the HER activity was not observed to increase linearly with Zn concentration and displayed an oscillating effect on the Zn composition.Specifically, when the Zn content in Ni 2−x Zn x P NCs is minimal (x = 0.01 equiv to 0.41%) or significantly higher, x = 0.38 (13.23%), 0.44 (15.32%), and 0.50 (17.14%), the Ni 2−x Zn x P NCs showed an η −10 of 225.3 ± 3.2, 269.9 ± 4.3, 276.4 ± 3.7, and 263.9 ± 4.9 mV, respectively, which are higher than binary Ni 2 P (216.2 ±  4.4 mV) suggesting a notable decrease in HER activity (Figure 5).However, Ni 2−x Zn x P NCs with x = 0.03 (1.26%), 0.07 (2.60%), 0.10 (3.69%), 0.15 (5.24%), and 0.23 (8.17%) compositions showed an η −10 of 188.7 ± 3.7, 170.0 ± 4.2, 210.4 ± 2.8, 135.4 ± 3.4, and 198.3 ± 2.1 mV, respectively, suggesting a higher catalytic activity than binary Ni 2 P. Interestingly, Ni 1.85 Zn 0.15 P NCs showed the lowest η −10 of 135.4 ± 3.4 mV among all samples investigated.To understand the catalytic behavior of Ni 2−x Zn x P NCs with increasing current density, the η values were probed at j = −20, −50, and −100 mA/cm 2 .All samples displayed a similar compositiondependent trend observed at j = −10 mA/cm 2 (Figure 5 and Table 2).The HER activities of Ni 2−x Zn x P NCs were compared with the commercial Pt/C catalyst.Pt/C requires an η −10 of 64.3 ± 2.1 mV, which is lower than the highest performing Ni 1.85 Zn 0.15 P NCs.However, at higher current densities, the difference between the η (Pt/C vs Ni 1.85 Zn 0.15 P NCs) becomes minimal, and when j ≥−45.1 mA/cm 2 , the Ni 1.85 Zn 0.15 P NCs outperformed the HER activity of Pt/C.Specifically, Pt/C showed η −50 and η −100 of 259.9 ± 5.8 and 415.3 ± 6.1 mV, while Ni 1.85 Zn 0.15 P NCs showed η −50 and η −100 of 250.4 ± 5.1 and 350.1 ± 6.2 mV, respectively.The enhanced HER activity demonstrated by Ni 1.85 Zn 0.15 P NCs at higher j values is crucial for the development of efficient HER catalysts for industrial applications.
The HER activity is also known to vary based on the mass loading of the catalyst. 54Therefore, mass-normalized current density was calculated to eliminate any influence from mass loading.The highest performing Ni 1.85 Zn 0.15 P NCs required η of 112.4,251.3, and 348.8 mV to reach j = −10, −100, and −200 mA/mg, respectively, which are lower than other Zndoped Ni 2 P catalysts (Figure 5D).The ECSAs of Ni 2−x Zn x P NCs were calculated to investigate the surface specific activity of all samples.ECSA was determined by calculating double layer capacitance (C DL ), derived from CVs scanned in the non-Faradaic regions at different scan rates.The computed ECSAs of Ni 2−x Zn x P NCs varied from 0.036 to 0.402 cm 2 without any correlation with the Zn composition (Supporting Information, Figure S11).
To further understand the effect of Zn doping on the reaction kinetics and mechanisms, Tafel slopes were computed from the LSV plots.Commercial Pt/C and Ni 2−x Zn x P NCs produced Tafel slopes of 68.7 ± 1.1 to 157.2 ± 0.5 mV/dec, suggesting Volmer−Heyrovsky mixed HER mechanism (Figure 6 and Table 2). 42,43,45  Chronopotentiometry was employed to investigate the effects of Zn doping on the stability and durability of Ni 2−x Zn x P NCs.The stability of binary Ni 2 P and the highest active Ni 1.85 Zn 0.15 P NCs were investigated at j = −10 mA/cm 2 for 10 h and compared with the Pt/C catalyst (10% wt.).Ni 2 P and Ni 1.85 Zn 0.15 P NCs exhibit similar electrochemical stability in 1 M KOH and showed a modest increase in η −10 to 274.1 and 172.9 mV (corresponding to a 26.78 and 27.69% increase), respectively.In contrast, Pt/C displayed a notable increase in η −10 to 87.8 mV (36.55% increase) after 10 h of HER.Physical characterization of Ni 2−x Zn x P NCs after catalysis suggests that the morphology and crystal structure are retained, although a few minor peaks corresponding to tetragonal Ni 12 P 5 were observed for certain samples (Supporting Information, Figures S13 and S14).
To compare the electrocatalytic activity before and after the chronopotentiometry study, polarization data were recorded and are shown in Figure 7B,D.The LSV data suggest that, after the durability test, Ni 2 P showed a significant decrease in the HER activity characterized by an increase of η −10 from 216.2 to 280.9 mV, η −50 from 342.4 to 441.9 mV, and η −100 from 433.4 to 564.6 mV, corresponding to an overall ∼30% increase from initial η values.In contrast, commercial Pt/C showed an increase of η −10 from 64.3 to 102.1 mV (58.79% increase), an increase of η −50 from 259.9 to 310.0 mV (19.27% increase), and an increase of η −100 from 415.3 to 455.5 mV (9.68% increase).The highest performing Ni 1,85 Zn 0.15 P NCs demonstrated excellent stability in alkaline electrolytes with a negligible change in η after 10 h of continuous HER.For Ni 1.85 Zn 0.15 P NCs, the η −10 changed from 135.4 to 142.0 (4.87% increase), the η −50 changed from 250.4 to 260.5 mV (4.03% increase), and the η −100 changed from 350.1 to 353.6 mV (0.99% increase).LSV plots recorded after the stability test were used to calculate Tafel slopes.An obvious increase in the Tafel slope from 111.3 ± 1.2 to 222.7 ± 2.0 mV/dec (100.08% increase) was observed for binary Ni 2 P, whereas the slope for Pt/C increased from 68.7 ± 1.1 to 100.5 ± 0.8 mV/dec, corresponding to a 46.28% increase.These increases in slope values indicate that, after the stability test, Ni 2 P and Pt/C favor the Volmer HER mechanism, attributing to slower HER kinetics and/or decreased electrocatalytic efficiency.In contrast, Ni 1.85 Zn 0.15 P NCs demonstrated a decrease in the Tafel slope from 72.1 ± 1.0 to 49.5 ± 0.9 (31.35% decrease), suggesting more dominant Heyrovsky type reaction with enhanced HER kinetics.This suggests that, with appropriate Zn doping, both the HER activity and stability of hexagonal Ni 2 P can be greatly improved.

■ CONCLUSIONS
In this work, theoretical and experimental methods were integrated to identify the most promising HER dopant and dopant concentration for hexagonal Ni 2 P and the HER activity and stability of Ni 2−x Zn x P NCs were systematically probed as a function of Zn composition.DFT studies suggest that the Zn composition has a direct effect on the ΔG H of Ni 2−x Zn x P NCs where 5.56% Zn is predicted to lower the ΔG H more favorably than 2.78 and 11.11% Zn.Accordingly, hexagonal Ni 2−x Zn x P NCs with variable Zn compositions (x = 0−0.50)were synthesized via a colloidal route and homogeneous alloy formation with no evidence of Zn segregation was confirmed through STEM-EDS.XPS spectra indicate modulation of surface polarization due to changes in binding energies of Ni and P moieties upon Zn incorporation into Ni 2 P NCs.At low and high Zn compositions, a decrease in HER activity was noted as x = 0.01, 0.38, 0.44, and 0.50 compositions exhibit η −10 of 225.3 ± 3.2, 269.9 ± 4.3, 276.4 ± 3.7, and 263.9 ± 4.9 mV, respectively, which are higher than binary Ni 2 P NCs (216.9 mV).The lowest η −10 of 135.4 ± 3.4 mV was achieved for Ni 1.85 Zn 0.15 P NCs (5.24% Zn) consistent with DFT studies, which predicted the optimum HER activity at ∼5.56% Zn composition.Moreover, Ni 1.85 Zn 0.15 P NCs outperformed the HER activity of Pt/C (10% wt.) standard at j ≥−45.1 mA/cm 2 with an η −50 of 250.4 ± 5.1 mV compared to an η −50 of 259.9 ± 5.8 mV achieved for Pt/C.A comparison of polarization curves recorded at various j values before and after 10 h of HER indicates that Ni 1.85 Zn 0.15 P NCs display the highest corrosion tolerance and retain its high activity for a longer duration compared to benchmark Pt/C.More specifically, Ni 2 P NCs displayed an increase in η −10 from 216.2 to 280.9 mV, whereas Ni 1.85 Zn 0.15 P NCs and Pt/C showed increases from 135.4 to 142.0 mV and 64.3 to 102.1 mV, respectively.In summary, the dopant identity and concentration are critical parameters when designing an efficient electrocatalyst.When these considerations are optimized, the catalytic activity and stability can be improved to such an extent that it can rival expensive noble metal catalysts.This study continues the inquisition into the optimization of earth-abundant catalysts for water electrolysis reaction and demonstrates how doping can be employed to modulate the efficiency and stability of catalytic nanostructures.

Figure 1 .
Figure 1.Hydrogen adsorption free energies (ΔG H ) for the Cr, Zn, Mo, Mg, Fe, V, Cu, Ti, and W-doped Ni 3 P 2 -terminated Ni 2 P (001) surface calculated at the level of PBE+D3(BJ) with a plane wave cutoff energy of 520 eV and a Γ-centered 5 × 5 × 1 grid of k-points.The circles and squares indicate the lowest ΔG H for the first and second H adsorption, respectively.The blue and red dashed lines are the lowest ΔG H values for the first and second H adsorptions on pristine Ni 2 P, respectively.The yellow band highlights the ±0.1 eV region around the optimal ΔG H = 0 value.

Figure 5 .
Figure 5. [A] Geometrical area-normalized polarization plots of Ni 2−x Zn x P NCs and commercial Pt/C in N 2 -saturated 1 M KOH at a 5 mV/s scan rate.[B] Variation of overpotential at different current densities (−10, −20, −50, and −100 mA/cm 2 ) as a function of Zn content for Ni 2−x Zn x P NCs.[C] A bar plot diagram showing the comparison of catalytic activity of Ni 2 P, Ni 1.85 Zn 0.15 P, and Pt/C at different current densities.[D] Massnormalized LSV plots of Ni 2−x Zn x P NCs in 1 M KOH at a 5 mV/s scan rate.

Figure 6 .
Figure 6.[A] Tafel plots derived from polarization curves of Ni 2−x Zn x P and commercial Pt/C in N 2 -saturated 1 M KOH at a scan rate of 5 mV/s and [B] a bar plot diagram demonstrating the variation of Tafel slopes as a function of Zn composition and Pt/C.

Figure 7 .
Figure 7. [A] Chronopotentiometry study of Ni 2 P, Ni 1.85 Zn 0.15 P, and commercial Pt/C electrocatalysts at j = −10 mA/cm 2 for 10 h in N 2 -saturated 1 M KOH.[B] Polarization curves, [C] corresponding Tafel slopes, and [D] a bar plot diagram showing the variation of HER overpotentials before (solid line) and after (dotted line) the stability test.

■
ASSOCIATED CONTENT * sı Supporting Information The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsaem.4c00539.A diagram demonstrating the lowest first and second ΔG H values on pristine Ni 2 P (001) with the Ni 3 P 2 termination along with corresponding 9 transition metaldoped surfaces; size histograms of Ni 2−x Zn x P NCs with variable compositions; LRTEM images of Ni 2−x Zn x P NCs with x = 0.03, 0.07, and 0.10 compositions; STEM images and elemental maps of Ni 2−x Zn x P NCs with x = 0.23, x = 0.15, and 0.50 compositions; XPS spectra of Ni 2 P NCs; FT-IR spectra of Ni 2 P NCs after annealing at 450 °C for 2 h; PXRD patterns of Ni 2−x Zn x P NCs after annealing at 450 °C for 2 h; CV curves and corresponding ECSA plots of Ni 2 P and Ni 1.85 Zn 0.15 P NCs; the lowest first and second ΔG H on Zn-doped Ni 2 P (001) with the Ni 3 P 2 termination at Zn concentrations of 2.78, 5.56, and 11.11%; and LRTEM and PXRD pattern of Ni 1.90 Zn 0.10 P NCs recorded after HER (PDF) ■ AUTHOR INFORMATION Corresponding Author Indika U. Arachchige − Department of Chemistry, Virginia Commonwealth University, Richmond, Virginia 23284-2006, United States; orcid.org/0000-0001-6025-5011;Email: iuarachchige@vcu.edu 1.85 Zn 0.15 P NCs, consistent with composition-dependent ΔG H calculations.The highest performing Ni 1.85 Zn 0.15 P NCs outperformed the HER activity of benchmark Pt/C (10% wt.) catalyst at j >−45 mA/cm 2 .Moreover, the stability of Ni 1.85 Zn 0.15 P NCs was superior to that of Ni 2 P NCs and Pt/C catalyst, where Ni 1.85 Zn 0.15 P NCs showed a 4.87% increase in η −10 after 10 h of HER in comparison to 29.92 and 58.79% increases noted for Ni 2 P and Pt/C electrocatalysts, respectively.

Table 1 .
Nominal and Experimental Compositions as well as Crystallite and Average Particle Sizes of Hexagonal Ni 2−x Zn x P NCs Produced via Colloidal Synthesis Elemental compositions of Ni, Zn, and P were obtained by ICP-OES.Each composition was determined by averaging three individual measurements per sample.b Crystallite size was calculated by applying the Scherer equation to the (111) reflection of the PXRD pattern. a

Table 2 .
Comparison of Overpotentials, Tafel Slopes, and ECSAs of Ni 2−x Zn x P Nanocatalysts Overpotentials and Tafel slopes were calculated by employing the LSV curves of three individually prepared electrodes, and the average values are presented.ECSA values were calculated by measuring the double layer capacitance (C DL ) of Ni 2−x Zn x P catalysts. a Binary Ni 2 P displayed a Tafel slope of 111.3 ± 1.2 mV/dec, whereas Ni 2−x Zn x P NCs with x = 0.01, 0.23, 0.38, 0.44, and 0.50 showed Tafel slopes of 128.6 ± 0.4, 133.4 ± 0.9, 135.7 ± 0.7, 157.2 ± 0.5, and 145.9 ± 0.3 mV/dec, respectively.These values suggest that the Volmer water adsorption (H 2 O + e − → H* + OH − ) step is the ratedetermining step.When Zn content was increased from x = 0.03 to 0.15, a decrease in the Tafel slope was noted.Ni 1.97 Zn 0.03 P, Ni 1.90 Zn 0.10 P, and Ni 1.85 Zn 0.15 P NCs produced Tafel slopes of 99.7 ± 1.1, 95.4 ± 1.1, and 72.1 ± 1.0 mV/dec, respectively, suggesting a Heyrovsky rate-determining step (H 2 O + H* + e − → H 2 + OH − ).The Tafel slope of the highest active Ni 1.85 Zn 0.15 P NCs (72.1 ± 1.0 mV/dec) was closer to that of Pt/C (68.7 ± 1.1 mV/dec), suggesting similar HER kinetics for both catalysts.DFT studies indicate that introducing Zn into Ni 2 P brings the ΔG H closer to a thermoneutral value, |ΔG H | ∼0 eV, that can improve the HER activity.Therefore, the dependence of ΔG H on Zn composition was also investigated to understand the high activity of the best-performing Ni 1.85 Zn 0.15 P NCs and composition-dependent trends in ΔG H . Specifically, at a Zn concentration of 2.78%, the first ΔG H remained unchanged at −0.42 eV, with hydrogen binding at the Ni 3 -hollow site, similar to pristine Ni 2 P.Meanwhile, the second ΔG H increased from 0.15 to −0.13 eV, with hydrogen binding at the Ni−Ni bridge.The overall ΔG H binding, represented as |ΔG H1 | + |ΔG H2 |, only decreased slightly from 0.57 eV in pristine Ni 2 P to 0.55 eV in Ni 2 P with 2.78% Zn.At a Zn concentration of 5.56%, the first and second ΔG H , with hydrogen binding at the Ni−Ni bridge and Ni−P bridge sites, resulted in a net ΔG H binding of 0.43 eV.With 11.11% Zn, the first and second ΔG H bindings decreased to 0.23 and 0.50 eV, respectively, producing a net ΔG H binding of 0.73 eV.These adsorptions occur at the Ni− Zn bridge and Zn−P bridge sites, respectively.The computational models for the hydrogen binding sites on Zn-doped Ni 2 P at variable Zn compositions are shown in Supporting Information, Figure S12.A smaller net ΔG H binding suggests higher HER performance, indicating that Ni 2 P with 5.56% Zn should exhibit the highest electrocatalytic activity.These findings are consistent with experimental observations, where a 5.24% Zn composition, corresponding to Ni 1.85 Zn 0.15 P NCs, demonstrated the highest HER activity.