Introduction

Hydrogen has emerged as a promising clean energy carrier due to its high energy density and zero-emission utilization1. Alkaline water electrolyzers (AWEs), which directly convert electricity from renewable energy sources into high-purity hydrogen, are a leading technology for green hydrogen production2. However, the hydrogen evolution reaction (HER) at the cathode of AWEs exhibits slower kinetics compared to acidic proton exchange membrane water electrolyzers, limiting energy efficiency even with noble Pt-based catalysts3,4. Research has shown that HER kinetics are influenced not only by the choice of catalyst but also significantly by alkali cations (AM+) present in the electric double layer (EDL)5,6,7,8. Intriguingly, synergistic effects between AM+ and surface structures have been identified by recent studies9,10,11, offering a new pathway for optimizing the electrochemical microenvironment around the active sites to enhance HER kinetics.

Cation effects in electrocatalytic reactions have attracted considerable interest due to their critical role in modulating reaction kinetics and mechanism. Several hypotheses have been proposed to explain how cations enhance HER kinetics, including (i) facilitating reactant adsorption or stabilizing intermediates via electrostatic interaction12,13,14,15,16,17,18,19,20,21, (ii) modulating hydrogen-bonding networks22,23,24, (iii) altering local pH13,25,26,27, etc. While these studies provide important insights, they do not yet offer a unified atomic-level understanding of cation–surface synergy under realistic interfacial conditions. As a result, the atomic-level mechanism underlying this cation-surface structure synergy remains unclear, impeding the rational design of catalyst/electrolyte interfaces for improved alkaline HER performance. In addition to the HER, solvated cations have also been demonstrated to play a decisive role in determining activities of other electrocatalytic reactions28,29,30.

The surface structure of the catalyst is also a critical factor in determining the kinetics of the HER in alkaline media. For low-index Pt surfaces, HER activity follows the trend Pt(111) < Pt(100) < Pt(110), which is attributed to differences in the population of low-energy H* binding states (active H* states toward HER) across these facets7,8,31,32,33,34. Beyond the intrinsic Pt–H* chemical interactions, facet-dependent surface charge states also influence the orientation of interfacial water molecules via electrostatic interactions31,35,36. This, in turn, may further modulate HER kinetics by altering solvent reorientation behavior as part of the concerted process in the HER kinetics.

Recent studies highlight the synergistic interplay between cationic effects and surface structure in governing HER kinetics. The Koper group examined how AM+ influence HER activity on Pt single-crystal surfaces10. They found that cations strongly interact with Pt step sites, significantly enhancing HER activity, whereas their impact on Pt(111) and Pt(100) terraces is negligible. This suggests that step-edge and cation interaction is critical for modulating the HER kinetics on stepped Pt surfaces. Similarly, work by the Mukerjee and Jia groups revealed that adsorbed hydroxyl-water-cation adducts (OH*-(H2O)x-AM+) in the double-layer region facilitate OH* removal into the bulk, forming OH−-(H2O)x-AM+ complexes37. This process, explained by hard-soft acid-base theory, accelerates the HER. However, due to the limited space and time resolution of experimental characterization techniques, the atomic-level understandings into the synergy between step sites and AM+ on HER kinetics are still elusive.

Density functional theory (DFT) calculations have provided crucial atomic-level insights into how AM+ and surface structure independently influence HER kinetics6,22,24,34,38,39,40,41. Recently, ab initio molecular dynamics (AIMD) simulations—which explicitly include solvent and ions under constant-potential conditions—have become increasingly important. These simulations more realistically model the electrode/electrolyte interface, capturing key atomic and electronic details within the EDL. For example, Chen et al. used AIMD to demonstrate how cations disrupt hydrogen bonding networks in the EDL, significantly impacting HER kinetics24. However, current computational models primarily focus on flat surfaces, failing to fully explain the experimentally observed synergy between AM+ and step sites in enhancing HER activity.

In the study, using constant-potential AIMD (CP-AIMD) simulations, we explore how Na+ modulates the HER kinetics in alkaline media on flat Pt(111) versus stepped Pt(311) surfaces. Our simulations reveal a stark contrast in interfacial water orientation at identical electrode potentials: while water adsorbs weakly on Pt(111), undercoordinated step sites on Pt(311) drive strong chemical interactions and tip effects, enabling specific water adsorption and water-mediated Na+ stabilization, forming a Pt–H2O*–Na+(H2O)x adduct on the stepped surface. This difference critically alters the EDL, generating distinct surface charge states on Pt(311) and a stronger interfacial electric field. These effects lower the water dissociation barrier, a key elementary step in alkaline HER, via partial O–H bond dissociation in interfacial water molecules induced by the strong interface electric field. In contrast, solvated Na+ stays farther away from Pt(111), resulting in a weaker electric field and water activation. Our study shed new light on the synergistic effects between AM+ and step sites, elucidating determinant factors for accelerating HER kinetics, which could facilitate the rational design of electrochemical microenvironment in the EDL for the enhanced HER activity in alkaline media.

Computational methods

All calculations were performed using the Vienna ab initio simulation package (VASP)42,43 with the Perdew–Burke–Ernzerhof functional44 and projector augmented wave pseudopotentials45. We used a 400 eV plane-wave cutoff and convergence criteria of 10–5 eV for electronic energy and 0.05 eV Å−1 for atomic forces. Transition states of electrocatalytic reactions were located using the electrochemical nudged elastic band (eNEB) method46, a constant-potential extension of the climbing-image nudged elastic band (CI-NEB) method47,48.

Pt(111) and Pt(311) surfaces were modeled as 4 × 4 four-layer slabs with the bottom two layers fixed for eNEB calculations; for CP-AIMD simulations, they were modeled as 6 × 6 three-layer slabs. Herein, the Pt(311) surface was employed to model a stepped surface at a tractable computational cost, despite being known to undergo a (1 × 2) missing-row reconstruction under electrochemical conditions49,50. A 25 Å vacuum layer prevented periodic interactions, and a 2 × 2 × 1 Γ-centered k-point mesh sampled the Brillouin zone. Constant-potential calculations followed our previous work51, with the electrolyte modeled using VASPsol’s implicit polarizable continuum model (PCM) approach (setting water dielectric constant ε = 80 and no cavitation)52,53. The Poisson–Boltzmann equation with 3 Å Debye length described counter-ion distributions. Analysis of the bound charge density (RHOB file, Fig. S2) indicates charge leakage into solvent cavities and around the Na⁺ ion. This leakage can induce partial Na⁺ desolvation and may weaken ion-specific interfacial effects. We note this as a known limitation of the linear PCM model used in VASPsol.

CP-AIMD simulations were performed at –0.7 V vs. standard hydrogen electrode (SHE) with the value of the absolute SHE potential set to 4.1 V. The value was calibrated to reproduce the experimental Gibbs free energy of the H⁺/H2 reaction (see Fig. S1 and Supplementary Note 1 for details). The model of explicit water solvent for CP-AIMD employed 64 H2O molecules at a density of 1 g cm−3 and 1 Na+. The bottom layer was fixed while the top two layers were allowed to move around. The configuration of water solvation was first equilibrated by using classical molecular dynamics simulations employing LAMMPS with TIP/3P force field for 1000 ps to obtain a structurally sound hydrogen-bond network (input script is shown in Supplementary Note 3)54,55,56. The relaxed water slab was positioned ~3.7 Å above the Pt surface, and a Na⁺ ion was introduced into a water network cavity ~4.5 Å from the metal. This combined interface then underwent 10 ps of AIMD equilibration (1 fs timestep) at 300 K and zero charge, using a Nosé–Hoover thermostat to minimize residual forces and produce the starting configuration for CP-AIMD simulations. The constant-potential conditions were achieved by adjusting the number of electrons in the system for every 5 steps. The CP-AIMD simulations for the Pt(111) and Pt(311) systems were performed for total durations of 20 and 30 ps, respectively, with a timestep of 1 fs at T = 300 K to ensure proper convergence to equilibrium.

Results and discussion

CP-AIMD simulations

To obtain a realistic description of the atomic structures of the EDL under HER conditions, we conducted CP-AIMD simulations at a constant potential of −0.7 V vs. SHE (approximately 0 V vs. RHE at pH = 12) for Pt(111) and Pt(311) surfaces interfacing with explicit water solvent and Na+ cations. Representative models of the interfacial EDL structures on the Pt(111) and Pt(311) electrodes are presented in Fig. 1a, c, respectively. The front and top views of the Pt(111) and Pt(311) models are shown in Fig. S3.

Fig. 1: Atomic structures and evolution of EDLs at Pt(111) and Pt(311) electrode–electrolyte interfaces.
Fig. 1: Atomic structures and evolution of EDLs at Pt(111) and Pt(311) electrode–electrolyte interfaces.The alternative text for this image may have been generated using AI.
Full size image

a, c Representative atomic structural models of the Pt(111)/electrolyte and Pt(311)/electrolyte interfaces. b, d Evolution of the surface charge density (σ = (q + 1)/2S, where q is the doped charge, 1 represents the electron transferred from sodium atom to the surface, and S is the surface area) and the electric potential (USHE) of the EDLs on Pt(111) and Pt(311) surfaces during CP-AIMD simulations at USHE = − 0.7 V at 300 K.

As shown in Fig. 1b, d, the surface charge density (σ) and electric potential (USHE) on both Pt(111) and (311) surfaces stabilize over time, confirming that the systems reached equilibrium and that the obtained EDL structures reliably represent the electrode/electrolyte interface under the studied electrochemical conditions. We note that the calculated σ does not accurately represent the true surface charge density, as the surface model used was asymmetric. The convergence of the system energy throughout the simulation is exhibited in Fig. S4.

The equilibrium σ’s of both Pt(111) and (311) interfaces are negative, indicating that their potentials of zero charge (PZCs) are higher than the target electrode potential (USHE = − 0.7 V), resulting in electron doping into the models. However, the σ of Pt(111) (~−37.5 μC cm−2) is significantly more negative than that of Pt(311) (~−25 μC  cm−2), suggesting distinct PZCs and double layer capacitances for the two facets57,58. As shown in Fig. S5, the implicit solvation model predicts an intrinsic PZC for Pt(311) that is 0.42 eV lower than that of the flat Pt surface.

Facet-dependent interfacial water orientation and Na+ location

The detailed atomic structures of the stern layer on Pt(111) and Pt(311) are presented in Fig. 2a, b, respectively, with the interfacial solvent layers highlighted in green. Notably, distinct water layer arrangements are observed between the two surfaces. On Pt(111), a compact layer of water molecules forms, characterized by vertically oriented O–H bonds pointing downward (with hydrogens facing the surface). This configuration arises from the negatively charged surface, which induces downward alignment of water dipoles to optimize electrostatic interactions. On Pt(311), in addition to the H-down water molecules, chemically adsorbed H2O* species are observed at the step edges due to strong interactions with undercoordinated Pt atoms. This leads to a more disordered arrangement of water molecules above the surface.

Fig. 2: Interfacial water structure and orientational analysis at electrified Pt(111) and Pt(311) interfaces.
Fig. 2: Interfacial water structure and orientational analysis at electrified Pt(111) and Pt(311) interfaces.The alternative text for this image may have been generated using AI.
Full size image

a, b The stern layer (0–3.5 Å above the surface) of the electrified aqueous Pt(111) and Pt(311) interfaces. The green shade highlights the interfacial water structures. c Illustration of the definition of angles α and β. α is the angle between the surface normal and the bisector of the two OH bonds, which is used to discriminate H-up (α < 90°) and H-down (α > 90°) configurations. β is the angle between the surface normal and each one of the OH bonds, which describes how much the OH bonds tilt away from the surface normal. d, e Distributions of α and β for the orientations of interfacial water molecules in the aqueous Pt(111) and Pt(311) interfaces.

Solvated Na+ also adopts distinct configurations in the Stern layer over the two Pt surfaces. Figure S6 shows the statistical analysis of the Na+ height on the Pt(111) and Pt(311) surfaces. The Na+ resides above the closely bounded water layer on the Pt(111) surface, staying away from the surface by about 4.69 ± 0.57 Å. On the Pt(311) surface, Na+ locates closer to the surface at about 2.34 ± 0.36 Å due to the formation of a Pt–H2O*–Na+(H2O)x adduct, with Na+ tightly coordinated to an adsorbed water molecule. The probability distributions (Fig. S6) further indicate that Na+ resides significantly closer to the electrode on Pt(311) with a narrower distribution, whereas on Pt(111) the distribution is broader, suggesting a more diffuse and fluctuating interfacial position due to weaker bound to the surface. The time evolution of the Na+–Pt(111) distance is shown in Fig. S7. The Na+ ion initially resides 2.84 Å above the surface but rapidly moves beyond 4.0 Å as the H-down water layer forms. This spontaneous displacement indicates that the final position of the Na+ ion is not an artifact of the initial configuration but results from a specific underlying mechanism. Similarly, Fig. S7 also shows that the Pt–H2O*–Na+(H2O)x adduct is persistently observed during the CP-AIMD simulation.

In addition to the chemical stabilization by the adsorbed water, we propose that the tip effect at step sites also acts in concert to stabilize Na+ closer to the surface on Pt(311). As shown in Fig. S8, the adsorption of a solvated Na+ ion (with four water molecules) is consistently stronger on the Pt(311) surface than on Pt(111) across all examined potentials. This is further demonstrated by the equilibrium Na+-surface distance, which is ~4.3 Å on Pt(111) but only ~3.8 Å on Pt(311) in this model calculation. We propose that the tip effect contributes to this stronger adsorption, as the enhanced electric field at the Pt(311) step edges exerts a stronger electrostatic force on the Na+ ion, which simultaneously stabilizes the Na+ ion and promotes greater electron transfer to the surface, as evidenced in Fig. S8.

To conduct a more quantitative analysis of the EDL structures at the aqueous Pt(111) and Pt(311) interfaces, we define two angles, α and β, which describe the orientations of water molecules relative to the catalyst’s surface, as illustrated in Fig. 2c. Specifically, α distinguishes between H-up and H-down configurations, while β indicates more specific orientations of the two OH bonds in a water molecule. Together, α and β enable precise discrimination between different water configurations.

Figure 2d, e present the distributions of the α and β angles for interfacial water molecules on the Pt(111) and Pt(311) surfaces. On the Pt(111) surface, the α angle displays a single peak at 140°, while the β angle exhibits two distinct peaks at approximately 90° and 160°. These distributions indicate a consistent, uniform orientation of water molecules at the aqueous Pt(111) interface, where one OH bond points nearly vertically toward the surface and the other OH bond tilts away from the surface (see schematic in Fig. 2d). On the Pt(311) surface, the α angle distribution shows two prominent peaks at approximately 80° and 140°, corresponding to two distinct water configurations: a lying-down configuration with OH bonds parallel to the surface plane and a configuration with one OH bond pointing downward, similar to that on Pt(111), as shown in Fig. 2e. Consistently, the β angle distribution exhibits peaks near 90° and 160°, with the 90° peak being more intense due to the higher prevalence of the parallel OH orientation.

The α/β distributions for the solvent layer transitioning to the bulk solvent layer (3.5–7.0 Å above the surface) is also analyzed as shown in Fig. S9. The results indicate that for both the Pt(111) and Pt(311) interfaces, the transition layer generally shows broader α/β distributions, which indicates more random molecular orientations than that in the stern layer.

Previous studies of water structures at the Pt(111)/aqueous interface at PZC typically identify a mixture of flat-adsorbed and H-down configurations57,59,60,61. In contrast, our calculations for the negatively charged Pt(111) surface do not show these flat-adsorbed configurations. The stronger electrostatic interaction with the hydrogen atoms of water favors H-down orientations. Our observed H-down configurations, with α ≈ 140° and β ≈ 90°/160°, are consistent with previous reports.

Previous work on the electrified Pt(533)/aqueous interface with Na+ cations has also shown the presence of flat-adsorbed H2O at step sites and H-down H2O at terrace sites, both at the PZC and under a mild negative charged density of –26.4 µC/cm2, which aligns with our findings. Specifically, a Pt–H2O–Na⁺(H2O)x adduct was identified at the stepped sites at this surface charge density. Furthermore, a similar Pt–OH–Na+(H2O)ₓ species at step sites was previously proposed to contribute to the cation-dependent shift of the H/OH exchange peak on stepped Pt electrodes62.

Surface charge state of interfacial Pt atoms

The distinct orientations and specific adsorption of interfacial water molecules at the electrified aqueous Pt(111) and Pt(311) interfaces lead to different charge states of surface Pt atoms. Figure 3a, b show the column-wise averaged Bader charges for Pt atoms on the Pt(111) and Pt(311) surfaces, respectively. On the Pt(111) terrace, Pt atoms at the aqueous interface exhibit a negative charge of approximately −0.1 |e| at USHE = − 0.7 V, indicating electron accumulation at potentials below the PZC. In contrast to the bottom Pt/implicit-solvent interface, the explicit-solvent interface at the top of the slab exhibits much stronger electron enrichment. We attribute this difference to the electron transfer60 and electron push-back59 effects of explicit interfacial water molecules. These effects lead to distinct PZCs for the implicit and explicit Pt/aqueous interfaces, resulting in different surface charge states of Pt.

Fig. 3: Column-averaged Bader charges of surface Pt atoms on Pt(111) and Pt(311) surfaces.
Fig. 3: Column-averaged Bader charges of surface Pt atoms on Pt(111) and Pt(311) surfaces.The alternative text for this image may have been generated using AI.
Full size image

Column-wise average Bader charge of surface Pt atoms on (a) Pt(111) and (b) Pt(311) surfaces.

In contrast, the Pt(311)/explicit water interface exhibits significant heterogeneity in surface charge distribution due to variations in surface sites, as shown in Fig. 3b. On the stepped Pt sites, undercoordinated Pt atoms display a wide range of Bader charges, varying from approximately +0.1 |e| to below −0.2 |e|. The positively charged Pt atoms result from water adsorption, whereas the negatively charged Pt atoms remain adsorbate-free (see Fig. S10, where positive values indicate electron loss and negative values indicate electron gain). On the terrace sites, all Pt atoms are bare and uniformly negatively charged. Figure S10 further illustrates the atom-wise Bader charge distribution on the Pt(311) surface, confirming that the positively charged Pt atoms arise from water molecule adsorption. However, the positive charge at the Pt(311) step is unlikely to result from oxidation by adsorbed H2O, as evidence indicates that H2O acts as an electron donor to Pt sites60. Instead, this behavior is more plausibly attributed to the Smoluchowski effect63—a well-established electronic spillover phenomenon in which electrons redistribute from high-curvature regions (such as step atoms) toward flatter terrace regions, resulting in an apparent electron deficiency at the step.

Partial dissociation of interfacial H2O

The interfacial electric field in the stern layer strongly polarizes interfacial water molecules, leading to elongation or partial dissociation of O–H bonds in H2O. Due to their distinct surface charge distributions, the electrified Pt(111) and Pt(311)/aqueous interfaces are expected to exhibit different interfacial electric fields, resulting in varying degrees of interfacial water polarization. In Fig. 4a, we compare the O–H bond lengths from the CP-AIMD trajectory in two selected regions (highlighted in green and yellow), corresponding to electrolyte volumes without and with Na+, respectively. The statistical distributions of O–H bond lengths in the two regions are shown in Fig. 4b, c. The plots display the spatial variation of O–H bond lengths along z direction, with color gradients representing the probability of each bond length occurring. The equilibrium average O–H bond lengths are approximately 1.0 Å. As shown in Fig. 4b, O–H vibrations cause the bond length to fluctuate within a range of about ±0.08 Å around the equilibrium value in Region 1 of the Pt(111) surface without Na+. Notably, the elongation of the O–H bond is more pronounced than its shortening, particularly at the Pt(111)/aqueous interface. This asymmetry arises from the polarization effect of the negative electric field due to the negatively charged surface, which preferentially promotes O–H bond elongation. In region 2 of Pt(111) with Na+, the O–H vibration increases only slightly, likely due to the presence of Na+, which enhances the polarization of interfacial water.

Fig. 4: Spatially resolved O–H bond length distributions in Pt(111) and Pt(311)/aqueous interfaces with and without Na⁺.
Fig. 4: Spatially resolved O–H bond length distributions in Pt(111) and Pt(311)/aqueous interfaces with and without Na⁺.The alternative text for this image may have been generated using AI.
Full size image

a, d Top views of the Pt(111) and Pt(311)/aqueous models. Two electrolyte regions are color-coded: green represents region 1 (without Na⁺), and yellow represents region 2 (with Na⁺). b, c, e, f Statistical distributions of O–H bond lengths along the z direction in regions 1 and 2 in the Pt(111) and Pt(311)/aqueous models.

In the Pt(311) model, the Na+ ion stabilizes closer to the surface as a Pt–H2O–Na⁺(H2O)x complex, generating a stronger electric field that further polarizes interfacial O–H bonds. These bonds stretch beyond 1.12 Å—longer than those on Pt(311) without Na+ (Fig. 4e) or on Pt(111) without/with Na+ (Fig. 4b, c).

The enhanced interfacial water activation on Pt(311) can be attributed to the synergistic effects of the surface facet and Na+. The undercoordinated Pt sites facilitate the specific adsorption of water molecules and Na+, promoting the formation of a Pt–H2O–Na+(H2O)ₓ adduct near the Pt/aqueous interface. In contrast to the flat Pt(111) surface, where Na⁺ resides further from the interface, the Na+ on Pt(311) generates a stronger local electric field, leading to more pronounced polarization of interfacial water. These findings highlight the critical role of surface morphology and ion interactions in modulating interfacial water activation.

Synergistic cation-facet effects on the HER kinetics

To elucidate the effects of surface structure and interfacial cations on HER kinetics, we performed eNEB calculations for the alkaline Volmer and Heyrovsky steps on both Pt(111) and Pt(311) surfaces, examining systems with and without interfacial Na⁺. The atomic configurations were simplified based on CP-AIMD simulations. The key difference between the solvation models is that an explicit water region can alter the interfacial Debye length, thereby changing how charges are screened. This results in different surface charge conditions between explicit and implicit models. Quantifying this, we computed surface charge densities (σ) from the eNEB models, finding σ = −7.6 μC cm−2 for Pt(111) and σ = −11.9 μC cm−2 for Pt(311) with Na⁺. These values differ from our AIMD results, confirming that the solvent environment dictates interfacial charge. Crucially, however, both models consistently predict enhanced O–H activation of water at the Na+–Pt(311) interface. This agreement validates the transfer of insights from the explicit solvation model to the implicit model used in the eNEB calculations.

Figure 5a compares the minimum energy paths (MEPs) of the alkaline Volmer step at USHE = − 0.7 V on Pt(111) and Pt(311), both with and without Na+. The corresponding initial, transition, and final states are illustrated in Fig. S11. For the Volmer step on Pt(111) and Pt(311) surfaces without Na+, our initial configuration positioned four water molecules above the catalyst surface. During the reaction, one water molecule dissociates, releasing a proton that adsorbs onto the surface while the remaining OH− ion becomes solvated by the three other water molecules in the final state. For the Na⁺-involving models, a Na+ ion was placed approximately 4.4 Å above the Pt(111) surface. On the Pt(311) surface, the Na⁺ ion was positioned at 3.8 Å in the form of a Pt–H2O–Na+(H2O)3 adduct, mimicking the environment from the CP-AIMD simulations. In the eNEB calculation, the final state does not represent the true product of the Volmer step because the generated OH− remains near the interface. To model OH− diffusion into the bulk, we displaced the OH− ion—along with three solvating water molecules—to a position ~10 Å above the surface. The energy of this “bulk OH−” state is included in the MEP plots for Pt(311) (Fig. 5a), both with and without Na⁺. This state was not calculated for Pt(111) with Na⁺, as displacing the OH⁻ disrupts the Na⁺–H2O coordination, resulting in an unreliable energy.

Fig. 5: Reaction energetics and bond-specific vibrational analysis of the alkaline Volmer step on Pt(111) and Pt(311) surfaces.
Fig. 5: Reaction energetics and bond-specific vibrational analysis of the alkaline Volmer step on Pt(111) and Pt(311) surfaces.The alternative text for this image may have been generated using AI.
Full size image

a Minimum energy paths (MEPs) at USHE = –0.7 V for the alkaline Volmer reaction on Pt(111) and Pt(311) surfaces. b Potential-dependent activation energies (Ea) of the alkaline Volmer reaction on Pt(111) and Pt(311) surfaces. c Initial state of the Volmer reaction on Pt(111) and Pt(311) surfaces with Na⁺. The two O–H bonds of the dissociating water are designated by A1 (parallel to the surface) and A2 (perpendicular to the surface). The same designation is applied to surfaces without Na+. d Comparison of the vibrational frequencies of the A1 and A2 bonds in the initial structures of the Volmer reaction on Pt(111) and Pt(311) surfaces.

On the Pt(111) surface, the activation energy (Ea) of the Volmer step is ~0.65 eV without Na+ but decreases to 0.61 eV with Na+. In contrast, on the Pt(311) surface, Ea drops more substantially, from 0.59 eV (without Na+) to 0.45 eV (with Na+). The ~0.14 eV reduction in Ea on Pt(311) is significantly larger than that on Pt(111), highlighting the stronger promotional effect of Na+ on the stepped surface. Consistently, the reaction energy (ΔE) also decreases most markedly on Pt(311). The “bulk OH−” state further decreases the reaction energies on Pt(311) surfaces, both with and without Na⁺, likely due to the relief of repulsion between the negatively charged surface and the adsorbed OH−. The Pt(311)-Na⁺ surface exhibits a greater energy drop for this state. This is probably because Na⁺ induces electron enrichment on nearby Pt edge sites, leading to stronger repulsion with the interfacial OH−, as demonstrated in Fig. 3b.

Figure S12 shows that the observed reductions in Ea and ΔE with Na+ cannot be attributed to stronger H binding (see Supplementary Note 2 for computational details of the hydrogen adsorption energy). In fact, Na+ weakens H* adsorption on both Pt(111) and Pt(311)—an effect that would typically increase the activation and reaction energies for the Volmer step. Instead, this enhancement likely stems from Na+-induced local electrostatic interactions, as further elaborated below.

Figure 5b depicts the potential dependence of the Volmer step Ea for Pt(111) and Pt(311), both with and without Na+. The Ea decreases monotonically with increasingly negative potentials for all systems, and the slopes of the Ea–potential relations are similar, indicating that the potential response of the transition state is not substantially altered by the presence of Na⁺. However, the absolute barrier values exhibit clear differences. Across the studied potential range, Na⁺ exhibits the strongest cation effect on the Pt(311) surface, where it most significantly accelerates the reaction. The stronger cationic effect of Na⁺ on Pt(311) is attributed to its closer proximity to the active site, facilitated by the formation of a Pt–H2O–Na+(H2O)x adduct structure.

To better understand how Na+ promotes the Volmer reaction on stepped Pt surfaces, we computed the vibrational frequencies of the O–H bonds in the dissociating water molecule at the initial reaction state. As shown in Fig. 5c, we designated the two O–H bonds of the reacting water molecule as A1 and A2. Figure 5d compares the calculated vibrational frequencies of the A1 and A2 O–H bonds in the initial structures of the Volmer step, both with and without Na+. While the A1 bond frequencies remain relatively constant across all configurations, the A2 bond (oriented perpendicular to the surface) shows significant variations. On Pt(111), the A2 frequency decreases from 3279 cm−1 (without Na+) to 3171 cm−1 (with Na+). The effect is more pronounced on Pt(311), where the frequency shifts from 3280 to 2957 cm−1 with Na+ introduction. This frequency reduction corresponds to a weaker A2 bond strength, particularly evident on Na+-modified Pt(311), which corroborates with the partial dissociation of the interfacial water molecules observed in the CP-AIMD simulations. The perpendicular O–H bond (A2) of the dissociating water molecule shows the greatest bond weakening. This decreased bond strength directly lowers the activation energy barrier for water dissociation, thereby accelerating the Volmer reaction. The enhanced bond weakening on Pt(311) with Na+ explains the observed promotion of water activation in the Volmer step.

The calculated reaction kinetics for the Heyrovsky step, in which a proton released from a reacting water molecule combines with an adsorbed hydrogen atom (H*) to form an H–H bond, are shown in Fig. S13. The energy changes associated with “bulk OH⁻” formation and the entropic contribution from gas-phase H2 (ΔTSH₂ = 0.41 eV at standard conditions) are also included. On the Pt(111) surface, Na⁺ increases the Eₐ from 1.35 eV to 1.48 eV, whereas on the Pt(311) surface, Na⁺ slightly reduces Eₐ from 1.29 eV to 1.25 eV at USHE = − 0.7 V. These results suggest that Na⁺ inhibits the Heyrovsky reaction on Pt(111) but marginally promotes it on Pt(311). However, the overall cation effect on the Heyrovsky step is much weaker than that on the Volmer step. Consequently, the O–H stretching frequency of the reacting water molecule remains high even in the presence of Na⁺ (Fig. S14).

In the context of alkaline HER kinetics, we suggest that Na+ incorporation at stepped Pt surfaces markedly accelerates the Volmer step but has little impact on the Heyrovsky step. The increased H* coverage from the accelerated Volmer step weakens H* binding strength, thereby lowering the Heyrovsky step’s activation energy, which collectively boosts the overall HER rate.

Finally, we combine the Volmer and Heyrovsky steps to obtain the complete HER free energy landscape at \({U}_{{\mbox{SHE}}}=-0.7{\mbox{{0.17em}}V}\) (Fig. S15). The calculated profiles show that HER is endothermic by ~0.4 eV on Pt(311) but becomes exothermic by ~–0.4 eV on Na⁺-Pt(311) when the energy for bulk OH⁻ formation and the entropy of gas-phase H₂ are included. This difference in reaction free energy stems from the distinct initial state of interfacial water. Our structural (Fig. 4) and vibrational (Fig. 5d) analyses reveal greater O–H bond elongation at the Na⁺–Pt(311) interface, implying a higher local water dissociation constant (\({K}_{w}\)). This further demonstrates the cationic effect on the stepped Pt(311) surface. It is important to note that while the simulation assumes initial conditions of pH = 12 and \({U}_{{\mbox{SHE}}}=-0.7{\mbox{{0.17em}}V}\) (\({U}_{{\mbox{RHE}}}\approx 0{\mbox{{0.17em}}V}\)), pH is not an explicit controlled variable. The effective pH varies along the eNEB reaction coordinate. Consequently, the calculated free energy profile deviates from the thermodynamic expectation of \(\Delta G=0\) at \({U}_{{\mbox{RHE}}}=0{\mbox{{0.17em}}V}\) due to the absence of explicit pH control in the simulation.

Conclusion

In this study, we employed constant-potential DFT calculations and AIMD simulations to investigate the influence of Na⁺ on Pt(111) and Pt(311) surfaces and its role in the kinetics of the alkaline HER. Our results demonstrate that the stepped Pt(311) surface adsorbs H2O and Na+ more strongly than the terraced Pt(111) surface, leading to distinct interfacial solvent orientations and Na+ distributions within the EDL. Notably, on the Pt(311) surface, a Pt–H2O–Na+(H2O)x adduct forms at the step sites, positioning Na+ in close proximity to the surface. This near-surface Na⁺ enhances the local electric field, promoting partial O–H bond dissociation in adjacent water molecules. As a result, Na+ significantly reduces the activation energy of the Volmer step on Pt(311), thereby accelerating the HER kinetics. These computational insights unravel how AM+ and step sites modulate the alkaline HER, providing a theoretical framework for understanding synergistic facet-cation effects in electrochemical reactions and guiding the design of high-performance electrocatalysts.