AM+ dependence of HER catalytic activity and capacitance in alkaline media
First, the HER catalytic activity of polycrystalline Pt electrodes in 0.1 M alkaline media (pH = 13, LiOH, NaOH, KOH) was investigated through linear sweep voltammetry (LSV) and cyclic voltammetry (CV). The LSV curves depicted in Fig. 1(a) reveal a clear dependence on cation identity for the HER performance of Pt electrode. Notably, Pt exhibited the highest HER catalytic activity in LiOH, whereas its activity was the lowest in KOH, consistent with earlier findings3,11,13,24,25.
The CV curves, depicted in Fig. 1(b) and Figs. S1(a-c), demonstrate a larger oxygen overpotential region and a smaller hydrogen overpotential region on the Pt surface. An adsorption peak of H (underpotential deposition peak, Hupd) is observed before the onset of hydrogen evolution1,25,26. The positive shift of Hupd with the variation of alkaline cations from Li+ to K+ suggests that the presence of K+ ions strengthen the HBE, consistent with a suppression of the HER activity. In contrast, the lower HBE associated with Li+ facilitates the HER catalytic process, in line with previous experimental investigations on polycrystalline Pt(100) and (110) surfaces27. This observation contrasts with the behaviour observed on Pt(111), where there was minimal change in Hupd, while a similar decrease in HER activity was observed when moving from acidic to alkaline environments19. Based on these, it is evident that the overall behaviour of our sample is primarily governed by the Pt(100) and (110) surfaces. However, it is important to note that the conclusion may exhibit bias, as alkaline media present a more complex chemical environment where the impact of AM+ on catalytic activity, alongside that of H+, must not be overlooked28. Unfortunately, due to the substantial oxygen overpotential region, the adsorption of AM+ on Pt electrode is obscured in the oxygen region, as shown in Figs. S1(a-c), posing challenges in clearly illustrating this phenomenon in a standard blank CV. Therefore, to thoroughly investigate the effects of AM+ on polycrystalline Pt catalysis, alternative electrochemical measurement techniques need to be devised to overcome these limitations and provide a more profound insight into the system.
According to the double layer theory, solvated cations accumulate within the OHP29, as illustrated in Figs. 2(a, b) and Figs. S2(a, b). Therefore, by determining the concentration of AM+ within the EDL, its influence on HER process can be revealed. To achieve this, we employed electrochemical impedance spectroscopy (EIS) to directly examine the differential capacitance curve of the polycrystalline Pt surface30,31. The variation in potential induces the adsorption of ions at the electrode interface, leading to changes in EDL capacitance. It is important to note that the capacitance (Cd) at the electrode interface is not a constant value, but rather varies with the change in potential. Its expression is as follows32:
$${\text{C}}_{\text{d}}\text{=}\frac{\text{∂}{\text{σ}}^{\text{M}}}{\text{∂φ}}$$
where σM represents the charge carried by the metal, i.e., the residual charge density on the electrode surface, and φ is the electric potential.
The differential capacitance curves of Pt in 0.1 M alkaline solutions are shown in Fig. 2(c). To compare the effects of different pH values and the identities of the anions and cations, we conducted measurements of the differential capacitance curves in H2SO4, H2SO4 + KCl, KCl, and KCl + KOH solutions (see Fig. S3 in the Supplementary Information). In H2SO4 solution, as shown in Fig. S3(a), H+ displays two adsorption peaks (0.10 ~ 0.25 V vs. RHE), consistent with the CV curves provided in Fig S4. By adding KCl to H2SO4 solution [Fig. S3(b)], a new adsorption peak emerges at a higher potential than the H+ adsorption peaks, which is attributed to the adsorption of K+ (0.30 ~ 0.45 V vs. RHE), while the adsorption peak in the far right (~ 0.60 V vs. RHE) belongs to OH− [Fig. S3(e)]. This distinction is also evident when comparing the results of the KCl solution and the KCl + KOH mixture, as shown in Figs. S3(c)-(d), which is consistent with the previous report33.
Figure 2(c) illustrates that the adsorption peaks of Li+, Na+, and K+ occur at 0.41, 0.43, and 0.44 V vs. RHE, respectively, following the sequence of Li+ < Na+ < K+. As depicted in Fig. 2(c), the differential capacitances of these ions, reflecting their non-covalent interactions, which are 122.36, 140.36, and 171.42 µF cm− 2 for LiOH, NaOH, and KOH solutions, respectively, support the relationship of Li+ < Na+ < K+,3,7 aligning with the trend of adsorption potential variation. Furthermore, Fig. 2(d) present a systematic comparison of the HER activity in alkaline media with AM+ adsorption characteristics at Pt electrode interface using overpotential at 10 mA cm− 2. It is evident that the catalytic activity of Pt electrode (Li+ > Na+ > K+) is inversely correlated with the adsorption capacity and potential of AM+ (Li+ < Na+ < K+). To gain a deeper insight into the influence of AM+ on HER catalytic performance, a Gouy-Chapman-Stern (GCS) double layer model was utilized for a comprehensive analysis, (refer to Section C in the Supplementary Information for details)34,35. The potential distribution of EDL in different alkaline media is depicted in Figs. 2(a,b) and Fig. S2(b). The centres of OHP positions (x2) for Li+, Na+, and K+ are calculated to be 5.68, 4.95, and 4.05 Å, respectively, in good agreement with previous report using DFT calculations.3 The EDL potential drops (Ψ0) are determined to be 99.8, 109.9, and 124.0 mV for Li+, Na+, and K+, respectively (for detailed information, refer to Table S1 in the Supplementary Information).
As the ion size of AM+ increases, their solvation effect weakens gradually, leading to a decrease in the charge screening effect. This phenomenon is more prominent for weakly solvated ions (such as K+), resulting in enhanced non-covalent electrostatic responses. The local accumulation rate of K+ is 72% and 36% higher than that of Li+ and Na+, respectively. Electrochemically, this is manifested by a higher adsorption potential, closer x2, and increased Cd36–38, which collectively contribute to a greater potential drop (Ψ0) across the entire double layer region. Consequently, this inhibits the HER activity of Pt to a great extent.
Correlation between EDL structure and HER activity through directional external electric field (DEEF)
The interface behaviour is influenced by a significant negative built-in E-field (ΔE = E-Epzc) at the interface, which arises from the positive shift of the potential of zero charge (Epzc) at large pH (Epzc=Eopzc + 0.059 pH; Epzc= 1.087 V vs. RHE for Pt at pH = 13).39 Therefore, when an external E-field is applied, the local pH will change, leading to alterations in the E-field at the interface and the rearrangement of the interfacial structured water molecules. This dynamic process can facilitate easier proton transfer, thereby enhancing HER process. On the other hand, AM+ ions are attracted by the negative E-field at the electrode interface, prompting their accumulation in the vicinity of the EDL region and resulting in alterations to the local structure of water molecules, H-bond network, cation concentration, and pH value. To comprehensively unveil the complex interplay between the properties of the EDL and the HER activity, we effectively controlled the local ion concentration at the solid-liquid interface using the DEEF method. For detailed information on the experimental measurements and modulation mechanisms, please refer to the Experimental Section and the Supplementary Information.
In Fig. 3(a) and Figs. S5(a, c), the obtained LSV curves clearly demonstrate a direct correlation between HER activity and the intensity of DEEF. It is indicated that the HER catalytic efficiency exhibits a gradual improvement as the DEEF strength increases. Additionally, chronoamperometry results, as depicted in Figs. S6(a-c), support this argument, in line with our initial expectations. The controlled manipulation of AM+ distribution in alkaline solutions through DEEF has systematic effects on HER catalytic performance. Furthermore, CV results displayed in Fig. 3(b) and Figs. S5(b, d) reveal a consistent decrease in current density with the escalation of DEEF strength, accompanied by a reduction in the intensity of Hupd (refer to Fig. S8 and Table S3). This trend suggests a progressive drift of positive charges (AM+, H+) away from Pt electrode under the influence of DEEF. Importantly, the peak positions in hydrogen and oxygen overpotential regions remain unchanged, indicating that DEEF does not impact the ion adsorption potentials. This observation implies that variations in HER efficiency are solely due to changes in the accumulation of ions within the EDL region.
Figure S9 shows that the response of HER performance to the DEEF is opposite under acidic conditions (0.05 M H2SO4) compared to alkaline ones. This discrepancy arises because in acidic conditions H+ serve as the primary reactants, which are drifted away by DEEF. This comparative analysis confirms the effectiveness of DEEF in regulating local ion concentrations and emphasise the significant impact of the local cationic environment on HER performance. However, in CV curves, the adsorption peak in the oxygen overpotential region overlaps with the signal of AM+ adsorption on Pt electrode, posing challenges for quantitative analysis. Therefore, in the following, we adopted a more sensitive differential capacitance method to investigate the adsorption behaviour of AM+ at the interface.
Combining with the observations from Fig. 2(c), it is evidenced that ion adsorption on Pt predominantly occurs within the potential range of 0 to 0.7 V vs. RHE, a non-Faradaic region where the HER does not occur. A single differential capacitance curve may not accurately capture the ion states in EDL at solid-liquid interface during HER process. Therefore, we refined our detection methodology, as outlined in the Experimental Section and illustrated in Fig. S10. The data acquired through this refined measurement setup are referred to as dynamic differential capacitance (DDC). Analysis of the results presented in Fig. 3(c) and Figs. S11(a, d) demonstrates a reduction in the adsorption quantity of AM+ in EDL with increasing DEEF strength in alkaline environments (pH = 13, LiOH, NaOH, KOH). Specifically, the capacitance values for Li+, Na+, and K+ decreased by 14% (from 137.68 to 117.98 µF cm− 2), 13% (from 146.27 to 127.34 µF cm− 2), and 25% (from 146.56 to 110.65 µF cm− 2). The surface densities for Li+, Na+, and K+ are decreased by 24.34%, 21.19% and 40.11%, respectively. These results indicate that DEEF effectively modulates ions dynamics in EDL. Importantly, it only influences ion capacitance while preserving the adsorption potential unaffected. This observation aligns with the CV analysis, affirming the reliability of our experimental measurements.
In Fig. 3(d) and Fig. S11(b, e), it is evident that the HER activity of Pt exhibits an inverse relationship with the AM+ adsorption quantity. This indicates that higher HER catalytic activity is associated with a lower adsorption quantity of AM+ in EDL, as characterized using the GCS model. Figure 3(e) and Figure S11(c, f) illustrate the potential distribution in EDL for various alkaline media. The position of OHP (x2) increases with rising DEEF intensity, shifting from 5.05 to 5.89 Å for Li+, 4.75 to 5.46 Å for Na+, and 4.74 to 6.28 Å for K+. As x2 increases, the potential drop Ψ0 decreases from 108.5 to 95.8 mV for Li+, from 113.0 to 102.3 mV for Na+, and from 113.2 to 89.8 mV for K+ (refer to Table S4). These experimental results indicate that closer proximity of x2 and enhanced adsorption in EDL region lead to higher Ψ0, resulting in more pronounced inhibition of HER catalytic activity on Pt.
Both DDC and CV measurements highlight that as DEEF strength increases, the adsorption of OH− in EDL decreases. This observation contradicts the expected outcome if OH− were solely influenced by E-field forces, as per the anticipated accumulation near Pt electrode (cathode) leading to increased adsorption. This phenomenon suggests that the electrostatic interaction between AM+ and OH− in EDL results in the formation of OHδ−-(H2O)x-AM+ clusters20,40,41. Previous studies have also emphasized the presence of hydrated complexes cantered on AM+ formed through electrostatic interactions. For instance, Jia and collaborators13 identified the existence of OHad-(H2O)x-AM+ complexes in EDL region, which are involved in regulating the transport properties of OH− into the bulk region. The reduction in OH− within EDL is attributed to the drift of OHδ−-(H2O)x-AM+ clusters under the influence of DEEF. This implies that the presence of OHδ−-(H2O)x-AM+ clusters in EDL exhibits an important effect in facilitating the Volmer process in water electrolysis.
Effect of EDL structure on catalytic kinetics
After investigating the impact of ionic adsorption characteristics on the HER catalytic activity of Pt in alkaline and acidic systems, we further explore its catalytic kinetic mechanism. In acidic solution, the HER catalytic steps are as follows:
Volmer: \({\text{H}}^{\text{+}}\text{ }\text{+ }{\text{e}}^{\text{-}}\text{ + * → }{\text{H}}^{\text{*}}\)
Heyrovsky: \({\text{H}}^{\text{+}}\text{ }\text{+ }{\text{e}}^{\text{-}}\text{ + }{\text{H}}^{\text{*}}\text{ → }{\text{H}}_{\text{2}}\text{ + }\text{*}\)
Tafel: \(\text{2}{\text{H}}^{\text{*}}\text{ → }{\text{H}}_{\text{2}}\text{ + 2*}\)
where, * represents the active site, and H* represents the adsorbed hydrogen atom. In alkaline systems, the Volmer and Heyrovsky steps are different and become the following process:
Volmer: \({\text{H}}_{\text{2}}\text{O + }{\text{e}}^{\text{-}}\text{ + * → }{\text{H}}^{\text{*}}\text{ + }{\text{OH}}^{\text{-}}\)
Heyrovsky: \({\text{H}}_{\text{2}}\text{O +}{\text{ e}}^{\text{-}}\text{ +}{\text{ H}}^{\text{*}}\text{ →}{\text{ H}}_{\text{2}}\text{ + * }\text{+ }{\text{OH}}^{\text{-}}\)
The fundamental difference between alkaline and acidic catalysis stems from the source of hydrogen species (H*), particularly the dissociation of water leading to the generation of OH−.42 During the progression of the catalytic reaction, an accumulation of OH− in EDL at the electrode interface may occur if they does not diffuse efficiently, which impede the rapid dissociation of water, consequently slowing down the kinetic process13,15,19,43.
The Tafel slopes of Pt in alkaline systems are shown in Fig. 4(a) and Figs. S12(a, b). In the absence of DEEF (0 V), highest rate of the kinetic process was observed in LiOH solution (153 mV dec− 1), followed by NaOH (155 mV dec− 1). This suggests that the rate-limiting step is the breaking of the O − H bond (Volmer step), which has a theoretical value of approximately 120 mV dec− 1.44,45 The variations in kinetics observed in different alkaline solutions can be attributed to the structure of EDL on their surfaces.
Firstly, water molecules, acting as reactants, must pass through the EDL to participate in the HER on the electrode surface. The presence of EDL diminishes the reactivity of the reactant towards the electrode, which is quantified as the total potential drop Ψ0 across the entire EDL [refer to Fig. S2(b)]. When the oxophilic group AM+ is present in EDL, this total potential drop is modulated, being 99.8, 109.9, and 124.0 mV for for Li+, Na+, and K+, respectively. This indicates that compared to LiOH, water molecules encounter a more substantial kinetic barrier in KOH when crossing the EDL, resulting in more facile kinetic processes in Li+ as compared to K+, with the order being Li+ > Na+ > K+.
Secondly, as an oxophilic group, AM+ captures the OH− generated during water hydrolysis through electrostatic interactions and stabilizes them by forming an OHδ−-(H2O)x-AM+ clusters. The capability to capture OH− is contingent on the adsorption capacity of AM+ in EDL. The stabilization of OH− by AM+ in EDL adversely affects the kinetics of HER catalysis11, as depicted schematically in Figs. 4(b, c). The E-field of weakly solvated cations (K+) is inadequately shielded, leading to a stronger electrostatic attraction and stabilization of OH−, whereas strongly solvated (Li+) possess a weaker ability to capture OH−. This contrast is evident in the differential capacitance curves [see Fig. 2(c)], where the OH− capacitance in LiOH, NaOH and KOH are 119.99, 135.02, and 166.99 µF cm− 2, respectively. The above two factors collectively result in the minimal inhibition of the hydrolysis step in LiOH solution and the most pronounced inhibition in KOH solution.
The influence of AM+ in EDL region on the kinetics of HER process can be elucidated by systematically adjusting DEEF, as illustrated in Fig. 4(a) and Fig. S12(a, b). As the intensity of DEEF increases, both the Tafel slopes and the capacitances of AM+ decrease, leading to a shift in the rate-limiting step of the HER kinetics. At the DEEF strength of 25 V, the Tafel slopes decrease to 105 mV dec− 1 (LiOH), 95 mV dec− 1 (NaOH), and 93 mV dec− 1 (KOH). This suggests that the rate-limiting step could potentially be either the proton recombination (Volmer-Tafel) step or the electrochemical desorption (Volmer-Heyrovsky) step, with this change in reaction rate solely attributed to the variation of AM+ in EDL. As AM+ moves further away from the electrode surface, Ψ0 gradually decreases to 95.8, 102.3, and 89.8 mV for Li+, Na+, and K+, respectively. Simultaneously, the adsorption amounts of OH− also decrease to 122.35, 126.78, and 112.02 µF cm− 2 (as shown in Table S6). The reduction in Ψ0 facilitates the access of water molecules to the electrode interface, while the decrease in OH− promotes the progression of the Volmer and Heyrovsky reactions. These combined effects influence the dynamics of Pt catalysis, leading to a shift in the rate-limiting step.
Furthermore, we systematically investigated the influence of interfacial OH− on the Tafel slope. Simulation results indicate that as the interface OH− concentration decreases, the surface hydrogen coverage (θH) increases at the same overpotential. The calculated Tafel slope, as a function of overpotential, reflects this change through the increase in θH. At θH ≈ 0.6, a transition in the Tafel slope occurs, consistent with previous reports44,46–48. Meanwhile, with decreasing OH− concentration, this transition occurs at lower overpotentials. More detailed information on the Tafel analysis can be found in Figs. S13, Table S7, and Figs. S14 of the Supplementary Information.
The application of E-field serves to drive away AM+ species and reduce their surface aggregation, consequently shifting the rate-determining step of water decomposition. Instead of water splitting being the limiting factor, the desorption of surface-bound hydrogen atoms (H*) emerges as a prominent step. Water decomposition requires the involvement of electrons, and E-field enhances electron accumulation at the electrode surface, thereby promoting interfacial water splitting. Additionally, E-field facilitates the efficient desorption of H* species and reinforces the Tafel process. It is important to emphasize that our investigation on HER kinetics, as determined by the Tafel slopes, reveals that the boundaries between the three rate-determining steps, i.e. the breaking of O-H bonds, subsequent activation of H*, and the desorption of H2, are not clearly delineated in alkaline conditions. These processes involve cooperative interactions, where the interplay between them influences the overall reaction kinetics. This highlights the complex nature of the HER mechanism in alkaline environments and underlines the importance of considering multiple factors when optimising electrocatalytic processes.
The effect of AM+ on Pt catalytic kinetics was further investigated from an energetic perspective. Temperature dependent LSV activity tests were performed within the range of 30°C to 50°C as shown in Fig. 4(d) and Fig. S15(a,b). The experimental data obtained was fitted using the Butler-Volmer equation to derive the exchange current density (j0), which was then used as a function of temperature to calculate the apparent activation energy (EA,app). According to the transition state theory of electrode dynamics, the transition from reactants to products involves crossing a potential energy barrier. A higher potential energy barrier signifies a more challenging electrode reaction, leading to a slower kinetic process. By examining the relationship between temperature, j0, and EA,app, we gain insights into the energetic landscape of the electrocatalytic process on Pt surfaces in the presence of AM+. This analysis provides a fundamental understanding of the mechanisms governing the kinetics of the reaction and sheds light on the interplay between temperature, activation energy, and reaction rate in electrochemical systems.
The results of our investigation are provided in Fig. 4(e). Taking into account the solvation effect, the EA,app follows the order of K+ (31.11 kJ/mol) > Na+ (22.02 kJ/mol) > Li+ (15.44 kJ/mol). This observation indicates a direct correlation between the kinetics of HER and the structure of EDL, where the proximity of AM+ ions to the electrode interface influences the rate of the kinetic process, consistent with prior research findings11,49,50. Further to this, we also explored the impact of DEEF on the variation of EA,app in a KOH solution. The results are presented in Fig. 4(f) and Figs. S16(a-d). With the increase in DEEF, the decrease in K+ adsorption in OHP results in a decrease in EA,app from 31.11 to 14.52 kJ mol− 1. This observation further demonstrates the key role of AM+ adsorption in EDL on the hydrolysis dissociation step of HER catalysis. The EA,app values obtained through Tafel extrapolation reaffirm this conclusion, as depicted in Fig. S17.
Density functional theory (DFT) analysis of interfacial dynamics
There is an ongoing debate regarding whether AM+ directly adsorbs onto or simply accumulates in OHP. In order to better understand the impact of the distribution characteristic of AM+ on the HER activity, we employed DFT to calculate the hydration structure of Pt/H2O in various solutions. In Fig. 5(a-c), our calculations clearly demonstrate that cations do not directly adsorb onto the Pt(111) surface. Instead, they are separated from the surface by a layer of water molecules. Specifically, Li+ and Na+ are separated from the Pt(111) surface by a layer of water molecules, maintaining a coordination number (CN) of 4 and oscillating between the first and second layers of water molecules, while K+ showed a CN of 5 and penetrated through the first water layer without specific adsorption onto the Pt surface. The simulated CN values (4 for Li+, 4 for Na+, and 5 for K+) and the cation positions align with previous findings as well3. Furthermore, the distances between Li+, Na+, and K+ and the Pt(111) surface are 5.22, 4.43, and 3.02 Å, respectively, consistent with the results reported in reference3. The results also agreed with the distances calculated using the GCS model (5.68 Å, 4.95 Å, and 4.05 Å for Li+, Na+, and K+), albeit slightly larger due to assumptions made (see Section C in the Supplementary Information), which further validated the calculations. Overall, the DFT simulations provided valuable insights into the hydration structure and behaviour of cations near the Pt(111) surface, shedding light on the distribution state of AM+ and its impact on HER activity.
Previous studies have demonstrated that (H2O)x-AM+ clusters located in EDL significantly affect the catalytic activity of Pt. To further investigate the impact of these clusters on the kinetic processes, we focused on the interactions between water molecules, AM+, and Pt(111) in the system, as H2O is the primary reactant in the alkaline HER. In the alkaline Volmer step, which involves H2O dissociation and is known to be the rate-determining step45, the interaction between Pt and H atoms in the interfacial H2O is crucial. Remarkably, our DFT calculations reveal that AM+ significantly influences the restructuring of the interfacial H2O molecules. AM+ weakens the interaction between H2O and Pt in the EDL, with the degree of weakening following the sequence Li+ < Na+ < K+. These findings are consistent with experimental observations that AM+ aggregation in OHP weakens the response of H2O to Pt interface, thereby directly impeding the kinetics of H2O dissociation. These insights provide valuable information about how AM+ clusters influence the kinetic processes involved in the alkaline HER on Pt surfaces.
In addition to influencing the distribution and structuring of interfacial H2O, AM+ also plays a significant role in modulating the orientations of O-H bonding within H2O molecules, which can serve as an indicator of the catalytic activity in HER process. To investigate this aspect, we analyzed the distribution of O-H bonds in the presence of various AM+ species. In the systems containing Li+, Na+, and K+, the two-dimensional distributions of the angle (θ) between the bisector of O-H bonds in H2O and the surface normal at different heights (h) above the electrode surface are shown in Figs. 5(d-f). The histograms depict the distribution profiles of the bonding angles exhibiting two main peaks, corresponding to the first water layer situated around ~ 3.4 Å and the second water layer around ~ 6 Å. More importantly, it is evident that the presence of cations significantly alters the distribution patterns of interfacial H2O molecules, particularly in the first H2O layer. Compared to the distribution of interfacial H2O in a pure H2O environment (~ 3.38 Å, ~ 92.98º), as illustrated in Fig. S18, the presence of Li+, Na+, and K+ make the H2O molecules being concentrated at (~ 3.24 Å, ~ 118.24º), (~ 3.38 Å, ~ 117.34º), and (~ 3.47 Å, ~ 95.29º), respectively. The order of distribution angles aligns with Li+ > Na+ > K+. In a pure H2O environment, θ is close to 90º, suggesting that they are predominantly horizontally aligned with the Pt surface. However, with the presence of small AM+ species (such as Li+), θ gradually increases, signifying a bending of the H atoms in H2O towards the Pt surface, which is favorable for promoting the HER process. This observation highlights how different AM+ species influence the orientations of O-H bonds in H2O, which in turn can impact the overall HER activity on Pt surfaces.
As oxyphilic groups, cations interact both covalently and non-covalently with the O in H2O molecules within EDL, leading to a reorganization of the interfacial H2O molecule structure. The extent of this reorganization is closely related to the position of AM+. For K+, it is situated in the first water layer, forcing the surrounding H2O molecules to reorient to form an asymmetric hydration structure of (H2O)x-K+. In this asymmetric (H2O)x-K+ clusters, fewer H atoms will be closer to the Pt surface to trigger the HER reaction. Essentially, this asymmetrical hydration structure restricts the interaction between the Pt interface and the hydrogen atoms. With the K+ center and the interfacial H2O layer occupying a similar plane, the resulting angle distribution is similar to that of pure H2O. In contrast, Li+ and Na+ are located between the first and second H2O layers, where their covalent and non-covalent interactions with oxygen induce a more favourable orientation distribution of H2O molecules in the first layer. The non-covalent interactions can be exemplified by the removal of solvated H2O molecules, as shown in Fig. S19, which demonstrates comparable outcomes to those obtained without the removal of solvated H2O molecules.
At the same time, we also noted that this restructuring of interfacial H2O diminishes its direct chemical adsorption with Pt (Pt-O), as illustrated in Fig. S20. Additionally, the presence of AM+ influences the distance between the layers of H2O molecules and Pt surface, following the order of Li+ < Na+ < K+, as seen in Table S8. This finding aligns with our experimental data, where weakly solvated aggregation of AM+ in OHP weakens the response of H2O to Pt interface, thereby directly impeding the kinetics of water dissociation. In short, a larger θ and a shorter distance imply that water molecules adopt a more favourable orientation for electrolysis, leading to a reduction in resistance, a shorter charge transfer distance, and an acceleration of the Volmer process, following the order of Li+ > Na+ > K+.
Comparison with previous models
Although there have been many preliminary studies, the mechanism between non-covalent interaction between AM+ and the reaction products is still not clear. Previous studies have preliminarily elucidated the observed HER trend by studying the changes in hydrolysis energy caused by cation interactions. Cations promote the interaction with hydrolysis products, including the identity of AM+, interfacial potentials, pH value, local alkalinity, ion concentration, and electrode surface properties, thereby impacting the energy barrier of kinetic steps2,4,36,40,51,52. The effects of AM+ can be divided into the following types: 1) promoting water decomposition, e.g. in Ni(OH)2-Pt system52; 2) altering the HBE; 3) forming OHad-(H2O)x-AM+ clusters within EDL to facilitate the diffuse of OHad into bulk13; 4) changing reaction dynamics from Heyrovsky (for Li) to Volmer (for K).
Previous study5 has proposed that the presence of OH− promotes Pt-H binding, potentially leading to a decrease in HER performance. However, this conclusion primarily considers the competitive adsorption of OH* and H* and predominant factors in thermodynamic considerations. It is crucial to note that changes in Hupd can be influenced by both pH-induced Nernstian and non-Nernstian variations, highlighting the necessity of incorporating kinetic aspects. Our findings align with the perspective that factors beyond the HBE significantly impact the overall water hydrolysis reaction. As a result, our discussion will broaden its scope to encompass a wider range of factors that influence HER performance, moving beyond the sole focus on HBE. By considering a more comprehensive array of influences, including kinetic factors, we aim to provide a more holistic understanding of the intricate interplay of variables that dictate the efficiency of HER process.
In alkaline environments, hydrated AM+ in EDL serve to stabilize the transition state of H2O dissociation and the resulting hydroxide products11. However, while this stabilization of intermediates is beneficial for certain aspects, it may be detrimental to the efficiency of HER catalysis. The presence of AM+ that stabilize OH− at the catalyst interface has the potential to hinder HER effectiveness by either obstructing the active sites or forming aggregates in OHP2. As the AM+ amount increases, its capacity to remove OH− diminishes. For active metal catalysts such as Pt, the slower OH− removal at the interface leads to an increase in the thermodynamic barrier, which inhibits the HER rate. It is demonstrated that metal catalysts with a strong affinity for OH− adsorption tend to exhibit a lower barrier for water hydrolysis, emphasizing the significance of effectively removing hydroxide species from the electrode interface from a kinetic standpoint. Liu et al.13 proposed that AM+ aids in the transport of water to the surface and the elimination of surface hydrolysis by-products, rather than directly facilitating hydrolysis. The stability of hydrolysis products in the presence of AM+ is linked to the local build-in E-field, which exerts a more pronounced influence on negatively charged OH− compared to neutral H*. However, the precise microscopic mechanism driving this process remains unclear and warrants further investigation to achieve a comprehensive understanding.
Markovic et al.53 suggest that water decomposition is a pivotal aspect of alkaline HER, with oxophilic groups potentially expediting this process14,54. It is argued that stable surface OHad can enhance alkaline HER by promoting the hydrolysis of the Volmer process. The stable adsorption of OH* on the surface is believed to boost HER, which is supported by DFT calculations3 of OH adsorption energy in the presence of Li+, Na+, and K+ ions. The stabilizing effect of cations on OH− at the Pt interface may involve direct non-covalent interactions or electrostatic effects induced by the cation. The interaction between OHad and AM+ is essentially electrostatic. In a Li+ environment, the strong E-field of Li+ results in significant electron density redistribution, leading to stronger adsorption of the polar OH−. Although AM+ is predominantly present in the aqueous layer and does not adsorb onto Pt, larger AM+ ions reside closer to Pt surface. The AM+ concentration on the surface, influenced by the coverage and polarity of OHad, significantly differs from that in the bulk solution. This is corroborated by our experimental results from capacitance measurements, indicating that the local alkaline concentration is 5–8 times higher than that of the bulk solution. It was agreed that the HER activity is strongly correlated with the AM+ coverage, which plays a role in influencing the OHad coverage.
Koper et al.23 argued that the HER mechanism involves the concentration of near-surface pH and AM+ rather than direct involvement of OH− species. This perspective is supported by other studies37,38, which highlight that high concentrations of large AM+ can increase the surface charge density at the electrode interface, creating a more robust interfacial E-field that stabilizes the reaction intermediate OH−. While the predominant role of the local cation concentration effect has been emphasized in many works, some researchers45 have linked cation-induced hydrolysis energetics to the observed HER rate trends. Studies suggest that cation interactions with the hydrolysis transition state can influence the kinetics of Pt catalysis, potentially lowering the energy barrier of the Volmer step. Furthermore, DFT calculations also support that the Heyrovsky step is rate-limiting for Pt23,45,55–57. Our research results indicate that the HER activity of Pt in alkaline solutions is mainly controlled by the first electron transfer step (Volmer process), with a minor contribution from the second electron transfer step (Heyrovsky process). We also support the significant effect of AM+ concentration changes on HER. Modulation of the AM+ concentration at the interface by an E-field allows it to dissociate from the complex formed with OH−, favouring HER properties.
It has been suggested that hydrated AM+(H2O)x ions interact with OHad53, forming H*-OH−-AM+ clusters that hinder the reaction site. These intricate interfacial hydrates, along with the water structure, can significantly influence the HER process. Wang et al.58 investigated how interfacial electronic states enhance water activation in HER. There are complex covalent and non-covalent interactions in EDL. In particular, H3O+ is observed in EDL even in alkaline solutions, as evidenced in our measurements, [refer to Fig. S3(d-e)]. Non-covalent interactions primarily contribute to the formation of interfacial water structures and are crucial in the HER mechanism. The interactions include hydrogen bonding, hydrated AM+, OHad-M+(H2O)x clusters, and OH−-H2O-AM+ hydrates. In a related study41, it is proposed that in the low potential region, AM+ and OH− do not adsorb, but water can form Pt-H2Oad-[(H2O)x-AM]+ clusters which can exchange with H* to produce H2.
Kinetic modelling indicates that Li+ plays a critical role in shaping the hydrogen bonding network and reducing the barrier for the HER. Conversely, large AM+ ions can disrupt this structure and impede the hydrogen reaction process. While kinetic models emphasize the significance of the Volmer reaction on Pt, some experimental and DFT studies propose that the Heyrovsky step may be pivotal for the HER mechanism. Further research is necessary to screen these divergent perspectives and to achieve a comprehensive understanding of the HER process on Pt surfaces.
The dynamics of potential within EDL and how external factors influence them are crucial for HER, which are impacted by local ion concentrations, potential gradients, and reactant chemical potentials59. These relationships can be better understood through the classical GCS model. Forces operating within the electrolyte layer include classical forces that compress the layer and entropy-induced forces that expand it. Altering the ionic concentration and potential energy gradient at the interface can significantly influence the disruption of hydrogen bonding in water. It is important to highlight that the Helmholtz model tends to overestimate interfacial capacitance and overlooks the role of entropy in these processes.
Our findings demonstrate that hydrogen adsorption on Pt surface exhibits two distinct adsorption potentials, with the first one decreasing and the second potential increasing as pH value increase, see Fig. S3. This variation highlights the intricate nature of H adsorption on Pt surface. Furthermore, the adsorption characteristics of AM+ are closely correlated to interfacial capacitance. The rise in interfacial capacitance aligns with the trend of AM+ adsorption potential. Moreover, the manipulation of E-field alters the intensity of Hupd only while maintaining a constant peak position, which is similar to the behaviour of interfacial capacitance. These observations suggest that ion transfer is more influenced by kinetic rather than thermodynamic (energetic) factors. Our DFT simulations reveal that hydrated AM+ migrates closer to the surface as its size increases, with distances of 5.22, 4.43, and 3.02 Å for Li+, Na+, and K+, respectively. At these distances, AM+ is unable to adsorb onto Pt surface, which is different from Koper's perspective1. As a result, the distances between the interfacial H2O and Pt increase to 4.22, 4.26, and 4.54 Å for Li+, Na+, and K+, respectively, from 4.14 Å in pure water. Our results align with those of Xu et al.,60 who also do not support AM+ adsorption on Pt. They proposed that the non-Nernstian pH shift is a consequence of pH-dependent water structure. There is no exchange between Had and OHad, but it does exist for Had and H2Oad.
The absence of direct binding of AM+ to Pt surface is also consistent with recent study3, suggesting that AM+ is separated from Pt surface by water molecules in the first hydrated layer. It indirectly interacts with OHad, thereby modifying the energy barrier of hydrolysis3,23. AM+ indirectly interacts with OHad, thus changing the energy barrier of the potential distribution within EDL. Above, we have determined that that the Ψ0 of the corresponding EDL is at 99.8, 109.9, and 120 mV in Li+, Na+, and K+ solutions, respectively. This potential barrier reduces when the external E-field displaces the AM+, leading to a shift in system dynamics (see Fig. S12). This shift is also evident in the alteration of the Tafel slope, decreasing from 155 to 95 mV dec− 1. These results unveil the complex dynamics and interconnected properties at the interface that impact the HER.
Upon the application of an external E-field (with a voltage of 5 V), a significant amount of OHδ−-(H2O)x-AM+ clusters within the dispersion layer depart from the surface, reducing the build-up of OH− in EDL. This dispersion of OHδ−-(H2O)x-AM+ clusters eliminates the aggregation effect, leading to a convergence of the HER behaviours across the three distinct alkaline solutions, as depicted in Fig. S7(b) and Table S2. However, with a further increase in DEEF strength, the HER performance continues to improve, displaying variability even among the alkaline solutions (Li+ < Na+ < K+), as illustrated in Fig. S7(c). This observation may be intricately linked to the solvation effect of AM+. Our experimental findings indicate that the formation of OHδ−-(H2O)x-AM+ clusters by AM+ in EDL stabilizes OH− through electrostatic forces. These complexes impede the removal of OH− from the electrode interface and the migration of water molecules toward the interface. The extent of AM+ aggregation in EDL is closely associated with this hindrance, with larger AM+ ions being more prone to aggregation, thereby compromising the HER performance of Pt.
DFT calculations reveal a close relationship between the structure of (H2O)x-AM+ and AM+. The distances between Li+, Na+, K+ and solvating water molecules are found to be 2.01, 2.30, and 2.87 Å, respectively. Compared to (H2O)x-Na+ and (H2O)x-K+, (H2O)x-Li+ exhibits a more compact structure. The solvating water layer acts as a charge shield for AM+, and the denser the structure of (H2O)x-AM+, the more effective this shielding effect becomes. Consequently, (H2O)x-AM+ exhibits varying responses to the E-field, with the trend following (H2O)x-Li+ < (H2O)x-Na+ < (H2O)x-K+. Under low DEEF, (H2O)x-AM+ tends to distribute evenly within EDL, resulting in similar impediments from OHδ−-(H2O)x-AM+ clusters to HER. This uniform distribution accounts for the convergence of HER characteristics across the three distinct alkaline solutions. However, under high DEEF, more (H2O)x-K+ ions are dispersed within EDL by E-field, weakening their stabilizing effect on OH−. Consequently, the inhibitory impact of OHδ−-(H2O)x-AM+ clusters on HER is reversed (Li+ < Na+ < K+). This process simultaneously enhances the surface Volmer reaction and Tafel or Heyrovsky reactions (by reducing the concentration of the reaction product OH−), thereby boosting the HER performance. These results are consistent with the Tafel slope measurements discussed earlier.
While our research model is subject to limitations imposed by computational constraints, we emphasise the critical role of the non-covalent aggregation of OH− and surface solvated cations in the interfacial HER. Our comprehensive approach elucidates that cations do not directly engage in covalent bonding with Pt surface or OHad. Instead, solvated cations (H2O)x-AM+ aggregate on the surface, forming electrostatic interactions with interfacial OHδ−. Within the HER potential range, weakly hydrated AM+ tends to approach the surface, impeding proton transfer kinetics and hydroxide dissociation in alkaline environments. Among the AM+ ions electrostatically bound to OHδ−, Li+ ions, characterized by smaller atomic radii, induce a more pronounced local E-field, leading to a denser structure for (H2O)x-Li+ compared to the relatively relaxed structures of (H2O)x-K+. Under the influence of DEEF, the relaxed (H2O)x-K+ is dispersed to a greater extent. Our findings validate that the adsorption of surface OH− diminishes as the strength of DEEF increases. This study sheds light on the substantial influence of AM+ on the HER kinetics, offering valuable insights for improving the efficiency of renewable energy conversion in electrolysis cells.