Cation-controlled diffusion of chloride ions during electrochemical chlorine evolution in acidic media

Cation-controlled diffusion of chloride ions during electrochemical chlorine evolution in acidic media | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Cation-controlled diffusion of chloride ions during electrochemical chlorine evolution in acidic media Ryuhei Nakamura, Taejung Lim, Hideshi Ooka This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5786052/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 08 Dec, 2025 Read the published version in Nature Chemistry → Version 1 posted You are reading this latest preprint version Abstract Impurity ions pose a major challenge towards diversifying water usage for electrolysis. In particular, millimolar-level chloride impurities remaining in reverse osmosis filtrates significantly diminish the selectivity and longevity of water electrolyzers. Here, we show that alkali metal cations can regulate the diffusion coefficient of chloride ions, enabling suppression of chlorine evolution during water electrolysis under diffusion-limiting conditions. Evidence of the cation dependency is provided by positive intercepts in both Levich and modified Koutecký−Levich plots using a rotating ring disk electrode, indicating the presence of an additional, cation-dependent diffusion layer that suppresses chloride diffusion. Numerical simulations based on the double diffusion model quantify this effect, resulting in a linear correlation between the cation-dependent diffusion barrier and the structural entropy of cation hydration. These findings suggest that the cation-dependent structuring of water significantly influences mass transport, which is particularly important at practical current densities where impurity ions are diffusion-limited. Physical sciences/Chemistry/Electrochemistry/Electrocatalysis Physical sciences/Chemistry/Physical chemistry/Chemical physics Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Water electrolysis driven by renewable electricity is promising as a sustainable method to produce valuable chemicals, such as dihydrogen, hydrocarbons, and alcohols 1 , 2 . However, this process requires large amounts of water, and at least 1.23 million tons of pure water per day (Mtpd) is needed by 2030 for green H 2 production alone 2 , 3 . Although this huge demand for water can be met using seawater reverse osmosis (SWRO) technology, which is capable of producing more than 60 Mtpd of pure water 4 , existing water electrolyzers require reagent-grade water containing less than 85 nM NaCl (ASTM Type II) to prevent unwanted reactions and degradation of electrolyzer components 3 , 5 , 6 . To utilize SWRO filtrate for the existing water electrolyzer systems, multiple deionization processes are required to remove ion impurities 6 , 7 . However, these additional purification steps increase the system installation costs and operational complexity, and are associated with design limitations 3 , 5 . Therefore, developing water electrolyzers that are stable in the presence of millimolar-scale impurities is needed to reduce costs and engineering complexity of industrial-scale water electrolysis. The dominant impurities remaining in reverse osmosis filtrates are monovalent and hydrophobic ions due to dielectric, steric, and Donnan effects 8 . During the water splitting, chloride ions (Cl − ) represent a particularly problematic impurity because the chlorine evolution reaction (CER; standard reduction potential, E ° = +1.36 V) occurs as a side reaction, competing with the desired oxygen evolution reaction (OER; E ° = +1.23 V) 9 – 11 . At the charge-transfer limit where slow electron transfer determines the reaction rate, Cl − concentrations as low as 20 mM can halve the OER activity of IrO 2 [12,13], which is the most industrially relevant OER catalyst 2 , 3 , 5 , 10 , 12 – 14 . This selectivity issue may be due to the sharing of active sites or common intermediates by the CER and OER, leading to the intrinsic difficulty of suppressing the CER 9 – 14 . To date, attempts to control electrocatalytic selectivity by tuning electrolyte composition have been effective for a number of electrochemical reactions 15 – 22 , even without catalyst modification. However, although the diffusion of a few tens of millimolar reactants is limited at high current densities (> 0.1 A∙cm –2 ), the effects of electrolyte composition on diffusion have been far less explored compared to those on charge transfer 21 , 22 . Thus, limiting Cl − diffusion may hold a key to suppress the impurity-driven CER during water electrolysis 23 . Here, we demonstrate that alkali metal cations (M + = Li + , Na + , K + , and Cs + ) modulate the Cl − diffusion coefficient during the acidic CER. Using a rotating ring-disk electrode (RRDE) and electrolytes supplemented with MCl, the diffusion-limiting current densities of CER ( j lim ) on amorphous iridium oxide (IrO x ) were shown to follow the order CsCl > KCl > HCl > NaCl > LiCl, with j lim being suppressed by up to 33%. Contrary to conventional R(R)DE theory, j lim measured at different rotations per minute (rpm) values produced positive intercepts in both Levich and modified Koutecký−Levich plots. Such intercepts have only been reported for electrodes coated with semi-permeable polymers 24 , 25 , leading us to hypothesize the presence of cation-dependent and rpm-independent diffusion layer. Numerical simulations based on the double diffusion model quantified the cation-dependent diffusion barriers of this additional layer. The identification of a linear correlation between the diffusion barrier and the structural entropy of hydration (Δ struc S ) 26 suggests that the cation-dependent hydration structure influences Cl − diffusion and the selectivity between the CER and OER. Results and Discussion Cation dependence of diffusion-limiting CER. An RRDE composed of a glassy carbon (GC) disk and polycrystalline Pt (poly Pt) ring was used. An amorphous IrO x colloid was electrodeposited onto the GC disk (IrO x /GC) following a reported method 12 (see Methods section). Adjusting the rotations per minute (rpm) of the RRDE allowed analysing the mass transport during electrochemical measurements, while simultaneously enabling measurement of the CER and OER partial current densities (Supplementary Fig. S1 ). Cyclic voltammograms (CVs) using the RRDE were conducted at 3,600 rpm in an H 2 SO 4 solution (pH 0.90 ± 0.03) supplemented with either 50 mM HCl or 50 mM MCl (M = Li + , Na + , K + , and Cs + ; Fig. 1 ). For all electrolytes, the disk current densities ( j disk ) increased as the potential increased versus the reversible hydrogen electrode (V RHE ), as the CER and OER occurred (Fig. 1 a). A constant potential of 0.95 V RHE was applied to the Pt ring to selectively reduce Cl 2 generated from the disk 12 , 13 . Using this setup, more Cl 2 was generated from the IrO x /GC disk in electrolyte containing Cs + than Li + because a more negative ring current density (− j ring ) was detected in the presence of Cs + (Fig. 1 b). To directly compare the cation dependence of j ring , the partial current densities of CER ( j CER ) and OER ( j OER ) were calculated for each electrolyte (Fig. 1 c, see Methods ). When the iR -corrected potential exceeded 1.65 V RHE , the observed plateau of j CER ( j lim ) was almost potential-independent, demonstrating that the CER rate was limited by the mass transport of chloride ions (Cl − ). In addition, j lim measured in the presence of CsCl was around 1.5 times higher than that of LiCl for all examined rpm values (Supplementary Fig. S2 ), demonstrating that the influence of cations on Cl − transport was rpm-independent. j OER decreased as j lim increased for the different cations, consistent with the competing nature of two reactions 9 – 13 , 27 . Accordingly, the CER selectivity in the presence of Li + was at least 20% lower than in Cs + (Fig. 1 d). Tafel slopes of the CER and the OER in the absence of Cl − were consistent with the representative values for the CER and OER in acidic media 9 , 10 , 27 , 28 (Supplementary Figs. S3 − 5 ), demonstrating a minimal influence of electrolytic cation on the intrinsic activity of IrO x on both CER 10 and OER 27 , 28 . Overall, the RRDE measurements demonstrate that the CER activity at the mass-transport limit is cation-dependent and emphasize the importance of electrolyte composition in controlling catalytic selectivity between CER and OER. Levich and Koutecký−Levich analyses. To investigate the cation effects on Cl − transport during CER, j lim was plotted with respect to the electrode rpm (Supplementary Fig. S2 ) using the following Levich Eq. 2 5,29 : $$\:{j}_{\text{l}\text{i}\text{m}}=0.62\bullet\:n\bullet\:F\bullet\:{C}_{\text{b}\text{u}\text{l}\text{k}}\bullet\:{\nu\:}^{-1/6}\bullet\:{D}^{2/3}\bullet\:{\omega\:}^{1/2}$$ 1 where n represents the number of electrons transferred for CER ( n = 2), F is the Faraday constant (96,485 C∙mol − 1 ), C bulk is the bulk Cl − concentration (mol∙cm − 3 ), ν is the kinematic viscosity of the electrolyte (cm 2 ∙s − 1 ), D is the Cl − diffusion coefficient in bulk electrolyte (cm 2 ∙s − 1 ), and ω is the rotational speed of RRDE (rad∙s − 1 or π/30 rpm). In all electrolytes, j lim was linearly proportional to ω 1/2 , indicating that Cl − diffusion predominantly contributed to mass transport for the CER (Fig. 2 a). Based on Eq. ( 1 ), the cation-dependent values of j lim are expected to be due to changes in D because changes in ν resulted in shifts of less than 1 mA∙cm − 2 (Supplementary Fig. S6 ). However, conventional RRDE theory does not consider D to be dependent on the electrolyte composition, contrary to the values of D estimated from the Levich slopes (Fig. 2 b). Furthermore, all the cations show a positive intercept value exceeding > 6 mA∙cm − 2 , which is significant, as j lim must extrapolate towards zero at zero rpm according to Eq. ( 1 ). So far, positive intercepts have been interpreted as resulting from non-diffusional factors 25 . For example, when electron transfer rate is not sufficiently fast for Cl − diffusion, the intercept corresponds to the kinetic current density of charge transfer ( j K ), as represented in the Koutecký−Levich (KL) Eq. 2 5,29 : $$\:{j}_{\text{C}\text{E}\text{R}}^{-1}={j}_{\text{l}\text{i}\text{m}}^{-1}+{j}_{\text{K}}^{-1}$$ 2 where j K indicates the kinetic current density of charge transfer for the CER. However, the ratio of j K / j lim exceeded > 500 for all electrolytes and rpm values, demonstrating that charge transfer was sufficiently fast (Supplementary Fig. S7 ). Therefore, the positive intercepts made by Eq. ( 1 ) must indicate the cation effect on Cl − diffusion, suggesting the presence of another rpm-independent diffusion layer. Double diffusion model of semi-permeable layer. Gough and Leypoldt proposed a modified KL equation by including rpm-independent current densities ( j f ), which result from the reactant diffusion in a semi-permeable polymer film 24 , 25 . The following equation is applied for the diffusion-limiting condition: $$\:{j}_{\text{l}\text{i}\text{m}}^{-1}={j}_{\text{c}\text{o}\text{n}\text{v}}^{-1}+{j}_{\text{f}}^{-1}=\frac{1}{0.62\bullet\:n\bullet\:F\bullet\:{C}_{\text{b}\text{u}\text{l}\text{k}}\bullet\:{\nu\:}^{-1/6}\bullet\:{{D}_{\text{c}\text{o}\text{n}\text{v}}}^{2/3}}\bullet\:{\omega\:}^{-1/2}+\frac{{L}_{\text{f}}}{n\bullet\:F\bullet\:{C}_{\text{f}}\bullet\:{D}_{\text{f}}}$$ 3 The separation of j lim −1 into the reciprocal current densities of conventional ( j conv −1 ) and film-originated ( j f −1 ) diffusion layers newly defines the Cl − diffusion coefficient in the conventional diffusion layer ( D conv ), the Cl − diffusion coefficient ( D f ) and Cl − concentration in a film ( C f ), and the film thickness ( L f ). The value of j f −1 is rpm-independent and determined solely by the properties of the semi-permeable polymer. For example, gaseous reactants, such as H 2 and O 2 , exhibit a linear correlation between j f −1 and Nafion thickness 30 , 31 . Here, we revisited previous results for H 2 and O 2 diffusion through Nafion-coated electrodes and confirmed that D conv is nearly independent of L f (Supplementary Figs. S8 − 10 ). Although no polymer film was present on the IrO x /GC disk in this study, Eq. ( 3 ) can be used to rationalize the cation-dependent intercepts ( j f −1 ) by considering an additional, rpm-independent diffusion barrier, similar to that arising from polymer films (Fig. 2 c,d). Consistent with the L f -independent D conv values of polymer films 30 , 31 (Supplementary Figs. S8d − 10d ), the values of D conv calculated by Eq. ( 3 ) were less cation-dependent than D by Eq. ( 1 ) (Fig. 2 b), indicating that this cation-dependent layer suppressed Cl − diffusion similar to polymer films. Although the thickness of cation-dependent layer was not experimentally controllable, we hypothesized that an increase in M + concentration could intensify the effects of the cation-dependent barrier, if the layer is controlled by cations. Effects of cation concentration on Cl − diffusion. To observe how M + concentration influences Cl − diffusion, chronoamperometry (CA) was conducted in H 2 SO 4 electrolyte supplemented with 25 mM MCl (pH 0.90 ± 0.03), and M 2 SO 4 was added during the measurements to increase M + concentration. The diffusion-limiting potential of 2.1 V RHE (no iR -correction; Supplementary Fig. S11 ) was maintained during the measurements. Current densities from the last half of each CA (7.5 s) were normalized to analyse steady state behaviour (Supplementary Fig. S12 ). Both Levich and j lim − 1 versus ω − 1/2 plots exhibited good linearities in all measured M 2 SO 4 concentrations (Supplementary Fig. S13 ). j f −1 values measured in electrolyte containing 25 mM MCl were also cation-dependent (Fig. 3 a). Adding M 2 SO 4 gradually decreased j f −1 , with Cs 2 SO 4 yielding the most rapid decrease of j f −1 within M + examined in this study. In addition, the mixing of Li 2 SO 4 and Cs 2 SO 4 revealed that Cs + had a greater effect on j f −1 than Li + (Supplementary Fig. S14 ). This finding indicates that M + concentration affects j f −1 but cannot change the order of j f −1 made by different M + . The slight decrease in j f −1 upon adding M 2 SO 4 might be due to the increased total electrolyte concentration. Consistent with the effect of different M + on Cl − diffusion (Fig. 2 ), Eq. ( 3 ) described cation-independence of D conv for different M + concentrations (Fig. 3 b), whereas Eq. ( 1 ) presented cation-dependent D (Fig. 3 c). Taken together, the effects of cation species (Fig. 2 ) and concentration (Fig. 3 ) on Cl − diffusion reveal that the cation dependency in Cl − diffusion coefficients is due to the cation-dependent diffusion layers. Numerical simulation of the cation effects on Cl − diffusion. Rpm-independent diffusion barriers are a characteristic feature of polymer-coated electrodes 24 , 25 . This phenomenon is generally attributed to a discontinuous concentration gradient of the reactant, due to different reactant solubilities across the polymer-electrolyte interface 25 , 32 . However, our unmodified IrO x electrode has no physical boundaries, leading us to hypothesize that the reactant concentration gradient is continuous across electrolyte interfaces (see Supplementary Note S1 ). In our proposed double diffusion model for Cl − diffusion (Fig. 4 a), the cation-dependent layer is located inside the conventional diffusion layer, whose thickness follows Levich theory ( δ = 1.61∙ D conv 1/3 ∙ ν 1/6 ∙ ω − 1/2 ). The same notational parameters used in Eq. ( 3 ) can be applied to this model. Based on this model, j lim at sufficiently high j K can be expressed as: $$\:{j}_{\text{l}\text{i}\text{m}}=n\bullet\:F\bullet\:{C}_{\text{b}\text{u}\text{l}\text{k}}\bullet\:{\left(\frac{{L}_{\text{f}}}{{D}_{\text{f}}}+\frac{\delta\:}{{D}_{\text{c}\text{o}\text{n}\text{v}}}\right)}^{-1}=\frac{n\bullet\:F\bullet\:{C}_{\text{b}\text{u}\text{l}\text{k}}}{\alpha\:+{\beta\:}^{{\prime\:}}\bullet\:{\omega\:}^{-1/2}}$$ 4 Its inversion yields: $$\:{j}_{\text{l}\text{i}\text{m}}^{-1}={j}_{\text{c}\text{o}\text{n}\text{v}}^{-1}+{j}_{\text{f}}^{-1}=\frac{{\beta\:}^{{\prime\:}}}{n\bullet\:F\bullet\:{C}_{\text{b}\text{u}\text{l}\text{k}}}\bullet\:{\omega\:}^{-1/2}+\frac{\alpha\:}{n\bullet\:F\bullet\:{C}_{\text{b}\text{u}\text{l}\text{k}}}$$ 5 where α and β indicate L f / D f and δ / D conv , respectively. β ′ is the rpm-independent component of β , determined by β = β ′ ∙ω −1/2 . Since the only two free variables in this model are α and β , they can be determined unambiguously from the slopes and intercepts in Fig. 2 c. The cation effects from both species (Fig. 2 ) and concentrations (Fig. 3 ) were applied to the double diffusion model at finite ω (Fig. 4 b). According to the obtained slope of 1.8 in the comparison of j f −1 between 25 mM and 50 mM MCl (Supplementary Fig. S15 ), C bulk is also important to calculate α . Because of the continuous concentration gradient, the boundary Cl − concentration between two diffusion layers ( C f ) depends on ω and is equal to C bulk at infinite ω (Fig. 4 a). Even so, $$\:{j}_{\text{l}\text{i}\text{m}(\omega\:,k\to\:\infty\:)}^{-1}={{j}_{\text{f}}}^{-1}=\frac{{L}_{\text{f}}}{n\bullet\:F\bullet\:{D}_{\text{f}}\bullet\:{C}_{\text{b}\text{u}\text{l}\text{k}}}>0$$ 6 demonstrates that L f and D f reproduce the positive intercept of the experimental plots in an rpm-independent manner (Figs. 2 d and 3 a). Numerical simulations of the double diffusion model combined with Butler − Volmer kinetics (details in Supplementary Note S1 ) produced sigmoidal j CER plots (Fig. 4 c), which matched well with the experimentally determined j CER (Fig. 1 c). Thus, without considering the potential dependence of α , the cation-dependent layer is unrelated to the applied potential. The simulated Levich plots were curved, indicating that β has a greater influence on j lim as β approaches zero due to infinite ω (Fig. 4 d). Conversely, the influence of α diminished as β goes infinite, yielding a zero intercept at ω = 0. The simulated j lim − 1 versus ω − 1/2 plots were linear, confirming that the conventional diffusion layer is entirely rpm-dependent (Fig. 4 e). Therefore, Eq. ( 5 ) of double diffusion model is more suitable to rationalize the present results than Eq. ( 1 ), which could not explain the anomalous, cation-dependent intercepts. Parameters obtained from the double diffusion model are summarized in Table 1 . For the conventional diffusion layer, β was weakly cation-dependent following the same series of the mutual diffusion coefficients of bulk MCl solutions 33 ( D MCl ; Supplementary Fig. S6c ). The obtained D conv values are therefore reasonable, as they are only ~ 2 times lower than D MCl values. On the contrary, the experimental measurements of L f are challenging because the double diffusion model is only applicable at the high j lim with intensive Cl 2 and O 2 evolutions. In an attempt to estimate L f , we considered the electrical double layer (EDL; Supplementary Fig. S16 ), which exist on the innermost electrode-electrolyte interface during electrocatalysis 25 , 34 . The EDL is composed of the inner Helmholtz double layer and the outer diffuse layer, which were omitted in our double diffusion model (Fig. 4 a,b) because cations negligibly affect charge transfer (Supplementary Fig. S3 ). The EDL is highly ion-concentrated 34 – 39 . The cation concentrations of the diffuse layer need to be higher than the bulk to complement the anion-concentrated Helmholtz layer, collectively balancing the positively charged electrode surface. Given the rpm-independent nature of the EDL 25 , 40 , the cation-dependent layer can be hypothesized as an extension of the outer diffuse layer of the EDL. Based on this assumption, we calculated D f assuming L f to be in the range of 1 − 10 nm (below < 0.2% of δ ), which is larger than the EDL thickness (~ 0.9 nm) calculated by the Gouy − Chapman theory at a minimum of total electrolyte concentration of this study (0.125 M) 25 . The expected D f is at least two orders of magnitude lower than D conv , indicating that cation-dependent layers are a significant barrier to Cl − diffusion. Correlation between the cation-dependent diffusion barrier and the structure of water network. Although ions can be highly concentrated in EDL, ~ 56 M of H 2 O solvent is sufficient to fully hydrate each cation 35 . Solvated ions that interact with nearby water molecules, inducing disruption of the original hydrogen-bond network, are called “water-structure breaker” according to Hofmeister’s classification 41 . To identify which properties of cations represent the cation-dependent diffusion barrier, L f/ D f was compared with the structural entropy of hydration 26 , 42 (∆ struc S ; Fig. 5 ) based on previous studies that examined the relationships among thermodynamic entropies, viscosities, and diffusion coefficients in aqueous solutions 39 , 41 – 44 . The following equation, proposed by Marcus 26 , 42 , was used to determine ∆ struc S values for the monovalent ions: $$\:{\varDelta\:}_{\text{s}\text{t}\text{r}\text{u}\text{c}}S\:\left(\text{J}\bullet\:{\text{K}}^{-1}\bullet\:{\text{m}\text{o}\text{l}}^{-1}\right)=40-606\bullet\:B\:$$ 7 where B indicates the viscosity B-coefficient, which mainly depends on the ion-solvent interactions of the Jones − Dole empirical expression 45 at 25°C. The values of experimental B selected by Jenkins and Marcus were used 26 . Since the B values of an aqueous MCl solution are expected to be valid up to 4 M [46], we considered that Eq. ( 7 ) is applicable for cation-dependent layer as an extension of ion-concentrated EDL. A linear correlation between L f/ D f and ∆ struc S (Fig. 5 ) supports the hypothesis that the cation-dependent water structure represents the cation-dependent diffusion barrier. Specifically, the negative ∆ struc S values of Li + and Na + indicate that the solvation of these cations results in an organized water network, whereas the positive ∆ struc S values of K + and Cs + represent a disrupted water network. In the absence of M + , H + exhibits slightly negative ∆ struc S , consistent with the RRDE measurements performed in H 2 SO 4 electrolyte supplemented with HCl (Supplementary Figs. S2c and S17 ). This finding indicates that blank H 2 SO 4 electrolyte enables solvated M + to organize or disrupt H 3 O + -H 2 O network, decreasing or increasing the diffusion barrier against Cl − , respectively. Since ∆ struc S excludes electrostatic forces 26 , 42 , the structural effect of cations predominates in suppressing Cl − diffusion. Including electrostatic effects in thermodynamic parameters also exhibits a linear relationship between L f/ D f and ∆ struc S for M + species (Supplementary Fig. S18 ). Conclusion In this work, RRDE analyses were used to study the cation effects on suppressing Cl − diffusion in acidic electrolytes with tens of millimolar Cl − concentrations. The diffusion-limiting CER current densities ( j lim ) was cation-dependent and significantly influenced the CER selectivity. This suggests that the OER selectivity in the presence of Li + can be further increased in higher applied potentials because the OER is not limited by diffusion. By measuring different cation species and concentrations, the cation-dependent intercepts ( j f −1 ) in j lim − 1 versus ω − 1/2 plots indicated that the cation-dependent layer existed independently of electrode rpm. The proposed double diffusion model attributes the rpm-independent diffusion barrier against Cl − diffusion to the cation-dependent layer. This finding allows us to understand cation-dependent Cl − diffusion based on the diffusion theory of semi-permeable layers. Among the cations examined here, the strength of the diffusion barriers was ranked in the order of LiCl > NaCl > HCl > KCl > CsCl and exhibited a clear linearity with the structural entropy of hydration (∆ struc S ) of each cation. This result indicates that the rigid hydrogen bond network in the presence of Li + and Na + decreases the Cl − diffusion coefficients in the cation-dependent layer ( D f ) due to negative ∆ struc S . On the contrary, Cs + and K + have positive ∆ struc S , indicating that the hydrogen bond network is disrupted, exhibiting increased D f . Given that diffusion is independent of the specific reactant-catalyst combination, ion-dependent reactant diffusion is also applicable to other electrochemical reactions having issues with ionic impurities, and thus may promote the development of efficient low-grade water electrolysis. Methods Chemicals. Li 2 SO 4 (99.7%) and Na 2 IrCl 6 ∙6H 2 O (33.9% Ir) were purchased from Thermo Scientific. H 2 SO 4 (97%), HClO 4 (70%), anhydrous LiCl (> 99.0%), KCl (> 99.5%), CsCl (> 99.0%), Na 2 SO 4 (> 99.0%), and K 2 SO 4 (> 99.0%) were purchased from Fujifilm Wako Pure Chemicals. Cs 2 SO 4 (> 99.0%) and Na 3 IrCl 6 ∙ x H 2 O (35 − 40% Ir) were purchased from Sigma-Aldrich. HCl (35 − 37%), NaCl (> 99.5%), and NaOH (97%) were purchased from Junsei Chemical. 1 M HCl standard solution was purchased from Hayashi Pure Chemical. All chemicals were used as received unless otherwise noted. A Milli-Q® EQ-7008 system (Merck Millipore) was used to prepare DI water (18.2 MΩ∙cm). General electrochemical procedures. All experiments were carried out in a fume hood at room temperature (20 − 25°C). Before electrochemical measurements, all glassware was boiled in DI water for 1 h, washed with DI water several times, and then dried in a 70°C oven, sequentially. When not being used, all glassware was stored in a 6 M HCl bath to remove ionic residues. Electrochemical measurements were conducted using two-compartment borosilicate glass cells divided by the pretreated Nafion 115 membrane (0.005-inch thickness, Fuel Cell Earth), unless otherwise noted. The Nafion 115 membranes (2×2 cm 2 ) were boiled in 3 wt% of H 2 O 2 , DI water, 1 M of H 2 SO 4 , and DI water for one hour at a time, sequentially. Before use, the pretreated Nafion 115 was stored in DI water for up to 3 months. A dual-channel potentiostat (SP-150e, BioLogic) was used for all electrochemical experiments. The GC disk of HR2-RD1 GC-Pt RRDE with polyether ether ketone (PEEK) spacer and shroud (Hokuto Denko), while its rpm was controlled by a HR-500 digital rotator (Hokoto Denko). We noted that the PEEK spacer was indispensable to prevent the Cl 2 /O 2 bubbles lodged on the RRDE spacer that severely disturbed the laminar flow from disk to ring (Supplementary Fig. S19 ). The geometrical area of GC disk and poly Pt ring were 0.196 cm 2 (radius, r = 2.5 mm) and 0.265 cm 2 (inner r = 2.75 mm and outer r = 4 mm), respectively. To obtain mirror-finished GC and Pt surfaces before all measurements, electrodes were hand-polished on microcloth (EC Frontier) with MetaDi Supreme diamond paste suspension (Buehler) of different particle sizes from 6, 1, to 0.05 µm, sequentially. Before changing the polishing pad with different suspension, RRDE was washed with copious amounts of DI water. Displayed potentials in electrochemical data were 85% iR-corrected after measurements (denoted as E − iR ), except for chronoamperometry, IrO x electrodeposition, and collection efficiency measurements. Electrochemical impedance spectroscopy (EIS) was conducted at an open circuit potential (OCP) to measure the solution resistance in the high-frequency domain (100 − 100,000 Hz) at a zero-degree phase angle. The EIS was measured before and after conducting the set of measurements. Measured OCPs were around 0.4 V RHE and 1.1 V RHE before and after measurements, independent of electrode rpm, respectively. The blank 0.1 M H 2 SO 4 electrolytes were prepared by diluting 97% H 2 SO 4 with DI water. After adding MCl and M 2 SO 4 (M = Li + , Na + , K + , and Cs + ) into the blank 0.1 M H 2 SO 4 , the solution pH was adjusted to the range of 0.90 ± 0.03 by adding a few drops of 97% H 2 SO 4 . For the H 2 SO 4 electrolyte supplemented with HCl, 1 M HCl standard solution was diluted with DI water and then adjusted to pH 0.90 ± 0.03 by adding 97% H 2 SO 4 . A ROSS pH electrode (8220BNWP, Thermo Scientific) connected with a digital meter (Orion Star A211, Thermo Scientific) was used to measure the solution pH. For the RRDE experiments, 70 mL of electrolyte in the working electrode compartment was directly purged with Ar gas (Taiyo Nippon Sanso JFP, G3 grade, 99.999%) for more than 10 minutes and kept a gas blanket during measurements (0.2 L∙min − 1 ). The reference electrode was a KCl-saturated Ag/AgCl electrode (EC Frontier, RE-7A, V Ag/AgCl = + 0.197 V vs. standard hydrogen electrode) with a Vycor porous glass frit, equipped with another glass-fritted junction filled with measured electrolytes to minimize Ag + and Cl − leakage. Unless noted otherwise, all potential units in this manuscript were noted in the reversible hydrogen electrode (RHE) scale by the equation as V RHE = V Ag/AgCl + (0.059∙pH) V + 0.197 V. A Pt coil (r = 1 mm and length = 5 cm) annealed by hydrogen flame was used as a counter electrode. All standard errors of the mean were calculated based on three independent measurements unless noted otherwise. Preparation of IrO x /GC disk electrodes. The IrO x /GC disk electrode was prepared by the colloidal electrodeposition of IrO x on the GC disk of polished GC-Pt RRDE following the previous report 12 , 13 . First, a metastable IrO x colloid suspension was obtained from an alkaline hydrolysis of Ir(III). 28.4 mg of Na 3 IrCl 6 · x H 2 O was mixed with 30 mL of 0.1 M NaOH (2 mM of nominal Ir). The pale-yellow solution was moved into jacketed glassware and heated using a water circulator set to 85°C (solution temperature: 75°C) for 20 minutes at 300 rpm of stirring. After a few minutes, the color of solution turned into transparent grey, light violet, and blue, sequentially. The colloidal solution was moved into a 50 mL beaker placed in an ice bath. 70% HClO 4 was added to acidify the solution to pH ~ 1.5 under mild stirring, changing the solution color to dark purple. Subsequently, a few drops of 10.1 M NaOH solution were added to make the solution alkaline (pH ~ 12.5), which shifted the solution color to violet. The solution was stored for up to 3 months at 4°C. After mechanical polishing, RRDE was rinsed and sonicated in DI water for 1 minute. For electrodeposition, the RRDE was vertically immersed into the IrO x colloidal solution in a single-compartment 5 mL borosilicate glass beaker, where a ring-shaped Pt wire annealed by hydrogen flame (r = 5 mm) was placed as a counter electrode. A 4 mL aliquot of the alkaline solution was acidified by adding 70% HClO 4 down to pH ~ 1.55. Care must be taken not to make the IrO x solution too acidic, which promotes OER and CER rates and lowers the electrodeposition rate. On the contrary, too high pH led to an excessive deposition of rough IrO x on GC disk and spontaneous deposition on the Pt ring. During electrodeposition, RRDE was rotated at 600 rpm without gas purging. To check the potential where IrO x deposition starts, RRDE was scanned at 0.25 V s − 1 in the range of 0.16 V − 1.36 V Ag/AgCl (Supplementary Fig. S20a ). Depending on the pH of IrO x solution, chronoamperometry (CA) at 1.25 ± 0.01 V Ag/AgCl for ~ 300 ± 30 s was conducted at around 20 mV positive potential at the onset of mixed CER and OER 12 (~ 0.8 mA, Supplementary Fig. S20b ). After electrodeposition, a reflective bluish-violet film was formed on the GC disk. The electrodeposited RRDE was washed with a copious amount of DI water and then heated in a 75°C oven for 6 h (ramping rate = 0.45°C∙min − 1 ). Sudden temperature increase could exfoliate the IrO x film from the GC substrate. The dehydration of IrO x increases the stability of IrO x film on GC substrate, reducing the amount of electrochemically dissolved Ir 47 and the physical exfoliation by Cl 2 and O 2 bubble evolutions during RRDE measurements. After dehydration, the color of IrO x /GC film changed to a reflective black. Electrochemical measurements using a rotating ring-disk electrode. For all RRDE measurements, the dual-channel potentiostat was connected by the mode of working electrode to ground. The distance between the reference electrode and RRDE was kept constant at around 2 cm to minimize the bubble interruption and the change of solution resistance. Before RRDE experiments, both IrO x /GC disk and Pt ring electrodes were electrochemically cleaned in a blank 0.1 M H 2 SO 4 solution (pH 0.90 ± 0.03) at 3600 rpm. The IrO x deposited on the Pt ring was removed within potential cycling in the range of 1.3–2.2 V RHE at 0.5 V s − 1 until decreasing OER activity was stabilized (Supplementary Fig. S21a ). Additionally, 40 potential cycling was conducted to electropolish the Pt surface in the range of − 0.1–1.7 V RHE at 0.5 V∙s − 1 (Supplementary Fig. S21b ). The CV of electrochemically cleaned Pt ring exhibited the characteristic peaks of poly Pt without IrO x peak in the range of 0.05–1.4 V RHE at 0.05 V s − 1 (Supplementary Fig. S21c ). The IrO x /GC disk electrode was also electropolished by 20-potential cycling between 1.3–1.6 V RHE at 0.5 V∙s − 1 (into the OER region). The electrochemical cleaning did not significantly change the IrO x surface (Supplementary Fig. S21d ). The separation of CER and OER current densities using RRDE was conducted following the methods of Vos et al. 12 A sufficient rpm of RRDE ( ω ; rad∙s − 1 ) generates the laminar flow from the inner disk to the outer ring, allowing the Pt ring to reduce the generated Cl 2 from the IrO x /GC disk back to Cl − (Supplementary Fig. S1 ). The spontaneous hydrolysis of Cl 2 into ClO − can be suppressed by the sufficiently low pH 0.9. Since poly Pt cannot reduce O 2 at 0.95 V RHE , the ring detects Cl 2 during a catalyst operation. Disk current density ( j disk ) and ring current density ( j ring ) can be respectively converted into j CER and j OER via $$\:{j}_{\text{C}\text{E}\text{R}}=\left|\frac{{j}_{\text{r}\text{i}\text{n}\text{g}}}{{N}_{\text{c}\text{o}\text{l}}}\right|$$ $$\:{j}_{\text{O}\text{E}\text{R}}=\:{j}_{\text{d}\text{i}\text{s}\text{k}}-\:{j}_{\text{C}\text{E}\text{R}}=\:{j}_{\text{d}\text{i}\text{s}\text{k}}-\left|\frac{{j}_{\text{r}\text{i}\text{n}\text{g}}}{{N}_{\text{c}\text{o}\text{l}}}\right|$$ The collection efficiency ( N col ) of the blank GC-Pt RRDE was determined by measuring the [Fe(CN) 6 ] 3− /[Fe(CN) 6 ] 4− redox couple using chronoamperometric RRDE measurements in a solution of 10 mM K 3 Fe[CN] 6 supplemented with 0.1 M MCl in the range of 1,225 − 3,600 rpm (Supplementary Fig. S22 ). To measure the background current, 0.5 V Ag/AgCl were respectively applied on GC disk and Pt ring for 60 s. Subsequently, CA was recorded at the disk potential of − 0.3 V Ag/AgCl and the ring potential of 0.5 V Ag/AgCl at each electrode rpm for 60 s. After subtracting the background current, N col was measured for each cation and electrode rpm. All measured N col exhibited 0.5% of standard deviation from the theoretical N col (0.474) given by the manufacturer. For the data measured in HCl (absence of alkali metal cations), N col measured in KCl was used. Since the IrO x /GC disk underwent highly oxidative potential, j ring may include a dissolution current of IrO x . However, j ring above − 1 mA∙cm − 2 was measured in the H 2 SO 4 electrolyte supplemented with 1 mM of each [Ir III Cl 6 ] 3− and [Ir IV Cl 6 ] 2− (pH 0.9; Supplementary Fig. S23 ), of which is excessive compared to the amount of dissolved Ir species during the OER at 1.65 V RHE (no iR-correction) 48 . The value of j ring was only 5% of the lowest j ring measured in the electrolyte containing 50 mM LiCl (~ 20 mA∙cm − 2 ; Fig. 1 b). Thus, the dissolution current was not considered for calculating j CER , j OER , and CER selectivity. The electrocatalytic selectivity of CER can be calculated from $$\:\text{C}\text{E}\text{R}\:\text{s}\text{e}\text{l}\text{e}\text{c}\text{t}\text{i}\text{v}\text{i}\text{t}\text{y}=\:1-\left(\text{O}\text{E}\text{R}\:\text{s}\text{e}\text{l}\text{e}\text{c}\text{t}\text{i}\text{v}\text{i}\text{t}\text{y}\right)=\frac{0.5\bullet\:{j}_{\text{C}\text{E}\text{R}}}{0.5\bullet\:{j}_{\text{C}\text{E}\text{R}}+0.25{\bullet\:j}_{\text{O}\text{E}\text{R}}}$$ Before RRDE measurements for OER and CER, IrO x /GC and Pt ring were preconditioned by 20 scans in the range of 1.2–2.2 V at 0.1 V∙s − 1 (into the OER/CER region), while the Pt ring was kept at 0.95 V to detect Cl 2 generated by the IrO x /GC disk electrode in MCl-containing electrolytes. The current densities in individual CVs of both disk and ring did not significantly change after 15 scans. This step stabilized the IrO x /GC and its OER/CER response to avoid a gradual current change during the measurements. The IrO x /GC disk electrode was scanned in the range of 1.2–2.2 V at 0.01 V∙s − 1 , while the Pt ring was kept at 0.95 V RHE . Each disk and ring current were corrected with the background currents measured at 1.21 V RHE , which was below the E 0 of OER/CER and Ir(IV/V) redox in acidic media 13 . After CER and OER measurements, reduction in current densities in CV (0.3–1.5 V RHE , 0.01 V∙s − 1 ; Supplementary Fig. S24 ) indicated that the electrochemically active surface area of IrO x decreased by electrochemical Ir dissolution during CER and OER at the high current densities. The CAs were measured after electrochemical cleaning, identical to the CV measurements. The applied potentials were fixed at 2.1 V RHE (no iR-correction), where the diffusion limit was selected based on CVs. Each CA was done in 15 s and consecutively measured in different rpm and ion concentrations without solution exchange. Characterization of Ir oxides. Scanning electron microscopy (SEM; Quattro S, Thermo Fisher Scientific) was operated at an accelerating voltage of 30 kV to take the SEM images and energy dispersive X-ray (EDX) spectra of the IrO x surface. Before measurements, all samples were washed with copious DI water and dried in the fume hood overnight. All SEM images and EDX spectra were obtained at 30 keV. The SEM images of the electrodeposited IrO x /GC show a partially cracked surface covered by IrO x NPs with 0.2 − 1 µm diameters (Supplementary Fig. S25a,b ). The SEM images of the IrO x surface after electrochemical cleaning by 20 potential cycling in 0.3–1.6 V RHE also exhibited the surface cracks and IrO x particles, indicating the electrochemical cleaning did not significantly change the surface morphology (Supplementary Fig. S25c,d ). After CER and OER measurements in the electrolyte supplemented with 50 mM of LiCl and CsCl, the IrO x surface became roughened by enlarged cracks (Supplementary Fig. S25e–h ), presumably due to intensive Cl 2 /O 2 gas evolution. However, crack formation did not increase the electrochemically active surface area, as shown in the decreased current densities in CV after measurements (Supplementary Fig. S24 ). The EDX spectra were measured before and after RRDE measurements (Supplementary Fig. S26 ), revealing that Cs + was not detected on IrO x /GC surface after RRDE measurements. Li + might not be detectable in EDX spectra. The results are consistent with the previous report that the incorporation of M + into the IrO x lattice was done by calcining the mixture solution of IrCl 3 and alkali metal hydroxides precursors at 400°C [49]. Declarations Author contributions T.L., H.O., and R.N. conceived the project and wrote the paper. T.L. conducted material synthesis, characterization, and electrochemical analyses. H.O. simulated the double diffusion model. All authors approved the final version of the manuscript. Acknowledgements This work was supported by GteX Program Japan (Grant Number: JPMJGX23H2) and MEXT Program: Data Creation and Utilization-Type Material Research and Development Project (Grant Number: JPMXP1122712807). T.L. received financial support from RIKEN Special Postdoctoral Researchers (SPDR) fellowship and RIKEN Incentive Research Projects (IRP). We appreciate Prof. A. Yamaguchi of Institute of Science Tokyo and Prof. K. Sakaushi of National Institute for Materials Science for experimental support on electrochemical measurements. We are grateful to Dr. A. Li and Dr. A. Koishi of RIKEN Center for Sustainable Resource Science and Dr. K. Yatsuzuka of University of Tokyo for the discussion on material synthesis, electrochemistry, and water network. We acknowledge Materials Characterization Support Team in RIKEN Center for Emergent Matter Science for technical support on SEM measurements. Data availability All relevant data generated and analysed during this study are included in this article and its Supplementary Information. Data for the main and supplementary figures are available in Figshare. References Graves, C., Ebbesen, S. D., Mogensen, M. & Lackner, K. S. Sustainable hydrocarbon fuels by recycling CO 2 and H 2 O with renewable or nuclear energy. Renew. Sustain. Energy Rev. 15, 1–23 (2011). IRENA. Green hydrogen cost reduction: scaling up electrolysers to meet the 1.5⁰C climate goal. International Renewable Energy Agency. (2020). Mayyas, A., Ruth, M., Pivovar, B., Bender, G. & Wipke, K. Manufacturing cost analysis for proton exchange membrane water electrolyzers. National Renewable Energy Laboratory. https://www.nrel.gov/docs/fy10osti/72740.pdf (2019). Eke, J., Yusuf, A., Giwa, A. & Sodiq, A. The global status of desalination: an assessment of current desalination technologies, plants and capacity. Desalination 495, 114633 (2020). Becker, H. et al. Impact of impurities on water electrolysis: a review. Sustain. Energy Fuels 7, 1565–1603 (2023). DuPont. FilmTec™ reverse osmosis membranes technical manual. DuPont. https://www.dupont.com/content/dam/dupont/amer/us/en/water-solutions/public/documents/en/RO-NF-FilmTec-Manual-45-D01504-en.pdf (2023). Lim, Y. J., Goh, K., Kurihara, M. & Wang, R. Seawater desalination by reverse osmosis: Current development and future challenges in membrane fabrication – a review. J. Membr. Sci. 629, 119292 (2021). Wang, L. et al. Salt and water transport in reverse osmosis membranes: beyond the solution-diffusion model. Environ. Sci. Technol. 55, 16665–16675 (2021). Trasatti, S. Electrocatalysis in the anodic evolution of oxygen and chlorine. Electrochim. Acta 29, 1503–1512 (1984). Karlsson, R. K. B. & Cornell, A. Selectivity between oxygen and chlorine evolution in the chlor-alkali and chlorate processes. Chem. Rev. 116, 2982–3028 (2016). Exner, K. S. Controlling stability and selectivity in the competing chlorine and oxygen evolution reaction over transition metal oxide electrodes. ChemElectroChem 6, 3401–3409 (2019). Vos, J. G., Wezendonk, T. A., Jeremiasse, A. W. & Koper, M. T. M. MnO x /IrO x as selective oxygen evolution electrocatalyst in acidic chloride solution. J. Am. Chem. Soc. 140, 10270–10281 (2018). Vos, J. G. et al. Selectivity trends between oxygen evolution and chlorine evolution on iridium-based double perovskites in acidic media. ACS Catal. 9, 8561–8574 (2019). Izumiya, K. et al. Anodically deposited manganese oxide and manganese–tungsten oxide electrodes for oxygen evolution from seawater. Electrochim. Acta 43, 3303–3312 (1998). Strmcnik, D. et al. The role of non-covalent interactions in electrocatalytic fuel-cell reactions on platinum. Nat. Chem. 1, 466–472 (2009). Colic, V., Pohl, M. D., Scieszka, D. & Bandarenka, A. S. Influence of the electrolyte composition on the activity and selectivity of electrocatalytic centers. Catal. Today 262, 24–35 (2016). Singh, M. R., Kwon, Y., Lum, Y., Ager, J. W., III & Bell, A. T. Hydrolysis of electrolyte cations enhances the electrochemical reduction of CO 2 over Ag and Cu. J. Am. Chem. Soc. 138, 13006–13012 (2016). Wintrich, D. et al. Enhancing the selectivity between oxygen and chlorine towards chlorine during the anodic chlorine evolution reaction on a dimensionally stable anode. ChemElectroChem 6, 3108–3112 (2019). Rao, R. R. et al. pH- and cation-dependent water oxidation on rutile RuO 2 (110). J. Phys. Chem. C 125, 8195–8207 (2021). Ovalle, V. J., Hsu, Y.-S., Agrawal, N., Janik, M. J. & Waegele, M. M. Correlating hydration free energy and specific adsorption of alkali metal cations during CO 2 electroreduction on Au. Nat. Catal. 5, 624–632 (2022). Gu, J. et al. Modulating electric field distribution by alkali cations for CO 2 electroreduction in strongly acidic medium. Nat. Catal. 5, 268–276 (2022). Li, X. Y. et al. Mechanism of cations suppressing proton diffusion kinetics for electrocatalysis. Angew. Chem. Int. Ed. 62, e202218669 (2023). Costentin, C. & Saveant, J. M. Catalysis at the nanoscale may change selectivity. Proc. Natl. Acad. Sci. U.S.A. 113, 11756–11758 (2016). Gough, D. A. & Leypoldt, J. K. Membrane-covered, rotated disk electrode. Anal. Chem. 51, 439–444 (1979). Bard, A. J., Faulkner, L. R. & White, H. S. Electrochemical methods: fundamentals and applications . 3rd ed, (John Wiley & Sons, 2022). Jenkins, H. D. B. & Marcus, Y. Viscosity B-coefficients of ions in solution. Chem. Rev. 95, 2695–2724 (1995). Shinagawa, T., Garcia-Esparza, A. T. & Takanabe, K. Insight on Tafel slopes from a microkinetic analysis of aqueous electrocatalysis for energy conversion. Sci. Rep. 5, 13801 (2015). Nong, H. N. et al. Key role of chemistry versus bias in electrocatalytic oxygen evolution. Nature 587, 408–413 (2020). Koutecký, J. & Levich, V. G. The use of a rotating disk electrode in the studies of electrochemical kinetics and electrolytic processes. Zh. Fiz. Khim. 32, 1565–1575 (1958). Watanabe, M., Igarashi, H. & Yosioka, K. An experimental prediction of the preparation condition of Nafion-coated catalyst layers for PEFCs. Electrochim. Acta 40, 239–334 (1995). Schmidt, T. J. et al. Characterization of high-surface-area electrocatalysts using a rotating disk electrode configuration. J. Electrochem. Soc. 145, 2354–2358 (1998). Knoche, K. L., Hettige, C., Zook-Gerdau, L. A. & Leddy, J. Film permeation by rotating disk voltammetry at electrodes modified with electrochemically inert layers and heterogeneous composites. J. Electrochem. Soc. 163, H3173–H3181 (2016). Rard, J. A. & Miller, D. G. Mutual diffusion coefficients of Nal-H 2 O and LiCl-H 2 O at 25°C from Rayleigh interferometry. J. Sol. Chem. 12, 413–425 (1983). Gebbie, M. A., Liu, B., Guo, W., Anderson, S. R. & Johnstone, S. G. Linking electric double Layer formation to electrocatalytic activity. ACS Catal. 13, 16222–16239 (2023). Garlyyev, B., Xue, S., Watzele, S., Scieszka, D. & Bandarenka, A. S. Influence of the nature of the alkali metal cations on the electrical double-layer capacitance of model Pt(111) and Au(111) electrodes. J. Phys. Chem. Lett. 9, 1927–1930 (2018). Bohra, D., Chaudhry, J. H., Burdyny, T., Pidko, E. A. & Smith, W. A. Modeling the electrical double layer to understand the reaction environment in a CO 2 electrocatalytic system. Energy Environ. Sci. 12, 3380–3389 (2019). Huang, J., Li, M., Eslamibidgoli, M. J., Eikerling, M. & Gross, A. Cation overcrowding effect on the oxygen evolution reaction. JACS Au 1, 1752–1765 (2021). Lee, S. S., Koishi, A., Bourg, I. C. & Fenter, P. Ion correlations drive charge overscreening and heterogeneous nucleation at solid–aqueous electrolyte interfaces. Proc. Natl. Acad. Sci. U.S.A. 118, e2105154118 (2021). Ojha, K., Doblhoff-Dier, K. & Koper, M. T. M. Double-layer structure of the Pt(111)-aqueous electrolyte interface. Proc. Natl. Acad. Sci. U.S.A. 119, e2116016119 (2022). Cameron, A. P., Davey, S. B., Callahan, C. L. & Donne, S. W. Communication—demonstrating the role of mass transport in double layer formation. J. Electrochem. Soc. 169 (2022). Assaf, K. I. & Nau, W. M. The chaotropic effect as an assembly motif in chemistry. Angew. Chem. Int. Ed. 57, 13968–13981 (2018). Marcus, Y. Viscosity B-coefficients, structural entropies and heat capacities, and the effects of ions on the structure of water. J. Solution Chem. 23, 831–848 (1994). Rosenfeld, Y. Relation between the transport coefficients and the internal entropy of simple systems. Phys. Rev. A 15, 2545–2549 (1977). Chopra, R., Truskett, T. M. & Errington, J. R. On the use of excess entropy scaling to describe the dynamic properties of water. J. Phys. Chem. B 114, 10558–10566 (2010). Jones, G. & Dole, M. The viscosity of aqueous solutions of strong electrolytes with special reference to barium chloride. J. Am. Chem. Soc. 51, 2950–2964 (1929). Breslau, B. R. & Miller, I. F. On the viscosity of concentrated aqueous electrolyte solution. J. Phys. Chem. 74, 1056–1061 (1970). Geiger, S. et al. Activity and stability of electrochemically and thermally treated iridium for the oxygen evolution reaction. J. Electrochem. Soc. 163, F3132–F3138 (2016). Geiger, S. et al. The stability number as a metric for electrocatalyst stability benchmarking. Nat. Catal. 1, 508–515 (2018). Gao, J. et al. Breaking long-range order in iridium oxide by alkali ion for efficient water oxidation. J. Am. Chem. Soc. 141, 3014–3023 (2019). Tables Table 1 Diffusion parameters in the numerical simulation based on the double diffusion model for 50 mM MCl at 3,600 rpm. Cation Species α = L f /D f (s∙cm − 1 ) β = δ/D conv (s∙cm − 1 ) δ (µm) D conv (×10 − 5 cm 2 ∙s − 1 ) D f for L f = 1−10 nm (×10 − 5 cm 2 ∙s − 1 ) Li + 45.9 117.3 6.99 0.596 0.00022 − 0.00218 Na + 33.5 96.0 7.73 0.805 0.00030 − 0.00299 H + 27.4 93.2 7.85 0.842 0.00037 − 0.00365 K + 15.1 100.3 7.57 0.755 0.00066 − 0.00663 Cs + 14.0 90.3 7.97 0.883 0.00072 − 0.00715 Additional Declarations There is NO Competing Interest. Supplementary Files 20250108NatChemSupplementaryInformation.docx Supplementary Information Cite Share Download PDF Status: Published Journal Publication published 08 Dec, 2025 Read the published version in Nature Chemistry → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5786052","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":402652628,"identity":"55e07e15-4896-48c0-88a4-729f99d84392","order_by":0,"name":"Ryuhei Nakamura","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABAElEQVRIiWNgGAWjYHACNgYGAyDFA+HJMTCzQUWJ1WJMpBYGhJbEBvyqGRjMZ+SYPfhQwJDH33P46WbePXbp29vZEh8XMPDl4dIicyPH3HCGAUOxxNk2s9s8z5Jz5xxmO2w8g4GtGJcWCYkcM2keA6B7zjMAtRxgzp3BzN4mzcPABnQhHi1/gFrmn2f/BtRSny7BzN7+m6AWYIglbjjbA7LlcIIEM9sxZrxaeJ6VSfYYSCRuPHOm7OacA8cNZzCzJQOdiscv7MnbJH78sUmcdyZ92403B6rlJfiPGX7mqTiGM8QYBBLAOtGFDY4l4NTCfwC7eA1uLaNgFIyCUTDSAABhYE3A5/fkJQAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0003-0743-8534","institution":"RIKEN /Japan","correspondingAuthor":true,"prefix":"","firstName":"Ryuhei","middleName":"","lastName":"Nakamura","suffix":""},{"id":402652629,"identity":"68c55989-3fc7-4bd9-9e98-a4bf84579c48","order_by":1,"name":"Taejung Lim","email":"","orcid":"","institution":"RIKEN /Japan","correspondingAuthor":false,"prefix":"","firstName":"Taejung","middleName":"","lastName":"Lim","suffix":""},{"id":402652630,"identity":"067f882b-3273-41a9-97cf-64b77ac8d4a0","order_by":2,"name":"Hideshi Ooka","email":"","orcid":"https://orcid.org/0000-0002-6921-6796","institution":"RIKEN","correspondingAuthor":false,"prefix":"","firstName":"Hideshi","middleName":"","lastName":"Ooka","suffix":""}],"badges":[],"createdAt":"2025-01-08 06:10:23","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5786052/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5786052/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41557-025-02014-4","type":"published","date":"2025-12-08T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":74021861,"identity":"a44b5fbe-88d8-421d-a2b8-2ad53f5ac9cf","added_by":"auto","created_at":"2025-01-17 05:10:08","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":2238498,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCation dependence of CER and OER current densities.\u003c/strong\u003e \u003cstrong\u003e(a)\u003c/strong\u003e Anodic scans of cyclic voltammograms on the IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e/GC disk (\u003cem\u003ej\u003c/em\u003e\u003csub\u003edisk\u003c/sub\u003e) at 3,600 rpm in an Ar-saturated 0.1 M H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e electrolyte supplemented with various alkali metal chlorides (MCl; 50 mM, pH 0.90±0.03). The applied potential (\u003cem\u003eE\u003c/em\u003e−\u003cem\u003eiR\u003c/em\u003e) on the disk was corrected for 85% of the solution resistance. Scan rate: 10 mV∙s\u003csup\u003e−1\u003c/sup\u003e \u003cstrong\u003e(b) \u003c/strong\u003eCurrent densities of the poly Pt ring (\u003cem\u003ej\u003c/em\u003e\u003csub\u003ering\u003c/sub\u003e) measured at 0.95 V\u003csub\u003eRHE\u003c/sub\u003e. \u003cstrong\u003e(c)\u003c/strong\u003e Plotted values for \u003cem\u003ej\u003c/em\u003e\u003csub\u003eCER\u003c/sub\u003e (solid lines) and \u003cem\u003ej\u003c/em\u003e\u003csub\u003eOER\u003c/sub\u003e (dashed lines) calculated from \u003cem\u003ej\u003c/em\u003e\u003csub\u003edisk\u003c/sub\u003e and \u003cem\u003ej\u003c/em\u003e\u003csub\u003ering\u003c/sub\u003e in Fig. \u003cstrong\u003e1a,b\u003c/strong\u003e (see \u003cstrong\u003eMethods\u003c/strong\u003e).\u003cstrong\u003e (d) \u003c/strong\u003eCER selectivity calculated from Fig. \u003cstrong\u003e1c\u003c/strong\u003e (see \u003cstrong\u003eMethods\u003c/strong\u003e). Three independent data sets were normalized for each line.\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-5786052/v1/71a73e0ac751dbbcb701e8be.png"},{"id":74021865,"identity":"10d0807a-e6ba-4473-b592-1fdfe0aa8248","added_by":"auto","created_at":"2025-01-17 05:10:08","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":2809368,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eLevich (\u003c/strong\u003e\u003cem\u003e\u003cstrong\u003ej\u003c/strong\u003e\u003c/em\u003e\u003csub\u003e\u003cstrong\u003elim\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e versus \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eω\u003c/strong\u003e\u003c/em\u003e\u003csup\u003e\u003cstrong\u003e1/2\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e) and \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003ej\u003c/strong\u003e\u003c/em\u003e\u003csub\u003e\u003cstrong\u003elim\u003c/strong\u003e\u003c/sub\u003e\u003csup\u003e\u003cem\u003e\u003cstrong\u003e−\u003c/strong\u003e\u003c/em\u003e\u003c/sup\u003e\u003csup\u003e\u003cstrong\u003e1\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e versus \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eω\u003c/strong\u003e\u003c/em\u003e\u003csup\u003e\u003cem\u003e\u003cstrong\u003e−\u003c/strong\u003e\u003c/em\u003e\u003c/sup\u003e\u003csup\u003e\u003cstrong\u003e1/2\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e plots to analyse the effect of cation species on Cl\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e−\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e diffusion. (a)\u003c/strong\u003e Levich plots of \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e versus the square-root of the rpm (\u003cem\u003eω\u003c/em\u003e\u003csup\u003e1/2\u003c/sup\u003e). The V\u003csub\u003eRHE\u003c/sub\u003e values in parentheses indicated the \u003cem\u003eE−iR\u003c/em\u003e at the displayed \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e. Dashed lines indicate linear regression. \u003cstrong\u003e(b)\u003c/strong\u003e The diffusion coefficients, \u003cem\u003eD\u003c/em\u003e (open grey diamond) and \u003cem\u003eD\u003c/em\u003e\u003csub\u003econv\u003c/sub\u003e (red circle), were calculated from Fig. \u003cstrong\u003e2a,c\u003c/strong\u003e, respectively, and are presented for the electrolyte supplemented with the indicated MCl or HCl. \u003cstrong\u003e(c)\u003c/strong\u003e \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e\u003csup\u003e−1\u003c/sup\u003e versus ω\u003csup\u003e−1/2\u003c/sup\u003e plots. The V\u003csub\u003eRHE\u003c/sub\u003e values in parentheses indicated the \u003cem\u003eE−iR\u003c/em\u003e at the displayed \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e. Dashed lines indicate linear regression. \u003cstrong\u003e(d)\u003c/strong\u003e The intercept of \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e\u003csup\u003e\u003cem\u003e−\u003c/em\u003e\u003c/sup\u003e\u003csup\u003e1\u003c/sup\u003e versus \u003cem\u003eω\u003c/em\u003e\u003csup\u003e\u003cem\u003e−\u003c/em\u003e\u003c/sup\u003e\u003csup\u003e1/2\u003c/sup\u003e plot (\u003cem\u003ej\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e\u003csup\u003e−1\u003c/sup\u003e). Three independent data sets were normalized for each data point, and the standard error of the mean was calculated.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-5786052/v1/668d1e04e5810def1c0790db.png"},{"id":74023227,"identity":"9c5d2985-121f-4c97-9433-3d480d6e0ac6","added_by":"auto","created_at":"2025-01-17 05:34:08","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":2102976,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEffect of cation concentration on intercept of \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003ej\u003c/strong\u003e\u003c/em\u003e\u003csub\u003e\u003cstrong\u003elim\u003c/strong\u003e\u003c/sub\u003e\u003csup\u003e\u003cem\u003e\u003cstrong\u003e−\u003c/strong\u003e\u003c/em\u003e\u003c/sup\u003e\u003csup\u003e\u003cstrong\u003e1\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e versus \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eω\u003c/strong\u003e\u003c/em\u003e\u003csup\u003e\u003cem\u003e\u003cstrong\u003e−\u003c/strong\u003e\u003c/em\u003e\u003c/sup\u003e\u003csup\u003e\u003cstrong\u003e1/2\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e plot and Cl\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e−\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e diffusion coefficients. (a) \u003c/strong\u003eInfluence of increasing M\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e concentration on \u003cem\u003ej\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e\u003csup\u003e−1\u003c/sup\u003e. Normalized current densities of chronoamperogram (CA) were obtained at 2.1 V\u003csub\u003eRHE\u003c/sub\u003e (no \u003cem\u003eiR\u003c/em\u003e-corrections) in 0.1 M H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e electrolyte initially supplemented with 25 mM MCl (pH 0.90±0.03). The last half of CA current densities (7.5 s) were normalized to analyse steady state. The measurements were performed by adding M\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e into the initial electrolyte incrementally to increase cation concentration from 25 mM to 125 mM (Supplementary Figs. \u003cstrong\u003eS11−13\u003c/strong\u003e). \u003cstrong\u003e(b)\u003c/strong\u003e \u003cem\u003eD\u003c/em\u003e\u003csub\u003econv\u003c/sub\u003e calculated from equation (3) as a function of increasing M\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e concentration.\u003cstrong\u003e (c)\u003c/strong\u003e \u003cem\u003eD\u003c/em\u003e calculated from equation (1) as a function of increasing M\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e concentration. Three independent data sets were normalized for each data point, and the standard error of the mean was calculated.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-5786052/v1/d467c69bc4a85409b9fc6672.png"},{"id":74021870,"identity":"994ceb1a-0d84-465f-8b73-4c5825b09eca","added_by":"auto","created_at":"2025-01-17 05:10:09","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":3561347,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSchematic model for double diffusion and simulated Levich (\u003c/strong\u003e\u003cem\u003e\u003cstrong\u003ej\u003c/strong\u003e\u003c/em\u003e\u003csub\u003e\u003cstrong\u003elim\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e versus \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eω\u003c/strong\u003e\u003c/em\u003e\u003csup\u003e\u003cstrong\u003e1/2\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e) and \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003ej\u003c/strong\u003e\u003c/em\u003e\u003csub\u003e\u003cstrong\u003elim\u003c/strong\u003e\u003c/sub\u003e\u003csup\u003e\u003cem\u003e\u003cstrong\u003e−\u003c/strong\u003e\u003c/em\u003e\u003c/sup\u003e\u003csup\u003e\u003cstrong\u003e1\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e versus \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eω\u003c/strong\u003e\u003c/em\u003e\u003csup\u003e\u003cem\u003e\u003cstrong\u003e−\u003c/strong\u003e\u003c/em\u003e\u003c/sup\u003e\u003csup\u003e\u003cstrong\u003e1/2\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e plots.\u003c/strong\u003e \u003cstrong\u003e(a)\u003c/strong\u003e In the proposed double diffusion model for cation-controlled Cl\u003csup\u003e−\u003c/sup\u003e diffusion, the conventional diffusion layer (thickness: \u003cem\u003eδ\u003c/em\u003e, rpm-dependent) is much thicker than the cation-dependent layer (thickness:\u003cem\u003e L\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e, rpm-independent). At infinite \u003cem\u003eω\u003c/em\u003e, the Cl\u003csup\u003e−\u003c/sup\u003e concentration at the boundary of the two diffusion layers (\u003cem\u003eC\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e) is equal to the bulk Cl\u003csup\u003e−\u003c/sup\u003e concentration (\u003cem\u003eC\u003c/em\u003e\u003csub\u003ebulk\u003c/sub\u003e), as \u003cem\u003eδ\u003c/em\u003e =0. \u003cstrong\u003e(b)\u003c/strong\u003e Schematic image showing the double diffusion model that reflects the effect of cation species (Fig. \u003cstrong\u003e2\u003c/strong\u003e) and concentration (Fig. \u003cstrong\u003e3\u003c/strong\u003e) on Cl\u003csup\u003e−\u003c/sup\u003e diffusion at finite \u003cem\u003eω \u003c/em\u003e(Li\u003csup\u003e+\u003c/sup\u003e and Cs\u003csup\u003e+\u003c/sup\u003e for example). \u003cstrong\u003e(c)\u003c/strong\u003e Experimental and simulated CV of CER (details in Supplementary \u003cstrong\u003eNote\u003c/strong\u003e \u003cstrong\u003eS1\u003c/strong\u003e). The experimental data is identical to Fig. \u003cstrong\u003e1c\u003c/strong\u003e and was used to fit the CER rate constant (\u003cem\u003ek\u003c/em\u003e). \u003cstrong\u003e(d)\u003c/strong\u003e Experimental and simulated Levich plots of CER. The experimental data was identical to Fig. \u003cstrong\u003e2a\u003c/strong\u003e. \u003cstrong\u003e(e)\u003c/strong\u003e Experimental and simulated \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e\u003csup\u003e\u003cem\u003e−\u003c/em\u003e\u003c/sup\u003e\u003csup\u003e1\u003c/sup\u003e versus \u003cem\u003eω\u003c/em\u003e\u003csup\u003e\u003cem\u003e−\u003c/em\u003e\u003c/sup\u003e\u003csup\u003e1/2\u003c/sup\u003e plots of CER. The experimental data was identical to Fig. \u003cstrong\u003e2c\u003c/strong\u003e.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-5786052/v1/ed699fd6746e3a6613b3060c.png"},{"id":74021879,"identity":"ea3f407d-9d79-44ad-a094-16bd82cb0fd1","added_by":"auto","created_at":"2025-01-17 05:10:09","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":558593,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDiffusion barrier of the cation-dependent layer (\u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eα, L\u003c/strong\u003e\u003c/em\u003e\u003csub\u003e\u003cstrong\u003ef\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e/\u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eD\u003c/strong\u003e\u003c/em\u003e\u003csub\u003e\u003cstrong\u003ef\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e) as a function of the structural entropy of hydration (Δ\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003estruc\u003c/strong\u003e\u003c/sub\u003e\u003cem\u003e\u003cstrong\u003eS\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e). \u003c/strong\u003eThe values of \u003cem\u003eL\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e/\u003cem\u003eD\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e\u003cem\u003e \u003c/em\u003ewere obtained from \u003cstrong\u003eTable 1\u003c/strong\u003e. ∆\u003csub\u003estruc\u003c/sub\u003e\u003cem\u003eS\u003c/em\u003e for each cation was calculated using equation (7) based on the selected B-coefficients at 25 °C (\u003cem\u003eB\u003c/em\u003e = 0.146, 0.085, 0.068, −0.009, and −0.047 within ±0.005 L∙mol\u003csup\u003e−1\u003c/sup\u003e for Li\u003csup\u003e+\u003c/sup\u003e, Na\u003csup\u003e+\u003c/sup\u003e, H\u003csup\u003e+\u003c/sup\u003e, K\u003csup\u003e+\u003c/sup\u003e, and Cs\u003csup\u003e+\u003c/sup\u003e, respectively)\u003csup\u003e26\u003c/sup\u003e. The dashed line indicates linear regression (\u003cem\u003eR\u003c/em\u003e\u003csup\u003e2\u003c/sup\u003e = 0.959). Three independent data sets were normalized for each \u003cem\u003eα\u003c/em\u003e value, and the standard error of the mean was calculated.\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-5786052/v1/d956e2c98ab04e0331960481.png"},{"id":97770895,"identity":"b71352f6-e860-426a-8dc0-11f687554cfb","added_by":"auto","created_at":"2025-12-09 08:06:24","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":12765847,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5786052/v1/2fea6cb6-08a7-433e-90af-a4ac21d99522.pdf"},{"id":74023229,"identity":"96174340-8f01-4284-b097-c7f84be593d4","added_by":"auto","created_at":"2025-01-17 05:34:09","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":18781224,"visible":true,"origin":"","legend":"Supplementary Information","description":"","filename":"20250108NatChemSupplementaryInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-5786052/v1/f24ddd67d83c0db6b1150f4d.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Cation-controlled diffusion of chloride ions during electrochemical chlorine evolution in acidic media","fulltext":[{"header":"Introduction","content":"\u003cp\u003eWater electrolysis driven by renewable electricity is promising as a sustainable method to produce valuable chemicals, such as dihydrogen, hydrocarbons, and alcohols\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. However, this process requires large amounts of water, and at least 1.23\u0026nbsp;million tons of pure water per day (Mtpd) is needed by 2030 for green H\u003csub\u003e2\u003c/sub\u003e production alone\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. Although this huge demand for water can be met using seawater reverse osmosis (SWRO) technology, which is capable of producing more than 60 Mtpd of pure water\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e, existing water electrolyzers require reagent-grade water containing less than 85 nM NaCl (ASTM Type II) to prevent unwanted reactions and degradation of electrolyzer components\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. To utilize SWRO filtrate for the existing water electrolyzer systems, multiple deionization processes are required to remove ion impurities\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. However, these additional purification steps increase the system installation costs and operational complexity, and are associated with design limitations\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Therefore, developing water electrolyzers that are stable in the presence of millimolar-scale impurities is needed to reduce costs and engineering complexity of industrial-scale water electrolysis.\u003c/p\u003e \u003cp\u003eThe dominant impurities remaining in reverse osmosis filtrates are monovalent and hydrophobic ions due to dielectric, steric, and Donnan effects\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. During the water splitting, chloride ions (Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e) represent a particularly problematic impurity because the chlorine evolution reaction (CER; standard reduction potential, \u003cem\u003eE\u003c/em\u003e\u0026deg; = +1.36 V) occurs as a side reaction, competing with the desired oxygen evolution reaction (OER; \u003cem\u003eE\u003c/em\u003e\u0026deg; = +1.23 V)\u003csup\u003e\u003cspan additionalcitationids=\"CR10\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. At the charge-transfer limit where slow electron transfer determines the reaction rate, Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e concentrations as low as 20 mM can halve the OER activity of IrO\u003csub\u003e2\u003c/sub\u003e [12,13], which is the most industrially relevant OER catalyst\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan additionalcitationids=\"CR13\" citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. This selectivity issue may be due to the sharing of active sites or common intermediates by the CER and OER, leading to the intrinsic difficulty of suppressing the CER\u003csup\u003e\u003cspan additionalcitationids=\"CR10 CR11 CR12 CR13\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. To date, attempts to control electrocatalytic selectivity by tuning electrolyte composition have been effective for a number of electrochemical reactions\u003csup\u003e\u003cspan additionalcitationids=\"CR16 CR17 CR18 CR19 CR20 CR21\" citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e, even without catalyst modification. However, although the diffusion of a few tens of millimolar reactants is limited at high current densities (\u0026gt;\u0026thinsp;0.1 A∙cm\u003csup\u003e\u0026ndash;2\u003c/sup\u003e), the effects of electrolyte composition on diffusion have been far less explored compared to those on charge transfer\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e,\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. Thus, limiting Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion may hold a key to suppress the impurity-driven CER during water electrolysis\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eHere, we demonstrate that alkali metal cations (M\u003csup\u003e+\u003c/sup\u003e = Li\u003csup\u003e+\u003c/sup\u003e, Na\u003csup\u003e+\u003c/sup\u003e, K\u003csup\u003e+\u003c/sup\u003e, and Cs\u003csup\u003e+\u003c/sup\u003e) modulate the Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion coefficient during the acidic CER. Using a rotating ring-disk electrode (RRDE) and electrolytes supplemented with MCl, the diffusion-limiting current densities of CER (\u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e) on amorphous iridium oxide (IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e) were shown to follow the order CsCl\u0026thinsp;\u0026gt;\u0026thinsp;KCl\u0026thinsp;\u0026gt;\u0026thinsp;HCl\u0026thinsp;\u0026gt;\u0026thinsp;NaCl\u0026thinsp;\u0026gt;\u0026thinsp;LiCl, with \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e being suppressed by up to 33%. Contrary to conventional R(R)DE theory, \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e measured at different rotations per minute (rpm) values produced positive intercepts in both Levich and modified Kouteck\u0026yacute;\u0026minus;Levich plots. Such intercepts have only been reported for electrodes coated with semi-permeable polymers\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e, leading us to hypothesize the presence of cation-dependent and rpm-independent diffusion layer. Numerical simulations based on the double diffusion model quantified the cation-dependent diffusion barriers of this additional layer. The identification of a linear correlation between the diffusion barrier and the structural entropy of hydration (Δ\u003csub\u003estruc\u003c/sub\u003e\u003cem\u003eS\u003c/em\u003e)\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e suggests that the cation-dependent hydration structure influences Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion and the selectivity between the CER and OER.\u003c/p\u003e"},{"header":"Results and Discussion","content":"\u003cp\u003e \u003cb\u003eCation dependence of diffusion-limiting CER.\u003c/b\u003e An RRDE composed of a glassy carbon (GC) disk and polycrystalline Pt (poly Pt) ring was used. An amorphous IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e colloid was electrodeposited onto the GC disk (IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e/GC) following a reported method\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e (see Methods section). Adjusting the rotations per minute (rpm) of the RRDE allowed analysing the mass transport during electrochemical measurements, while simultaneously enabling measurement of the CER and OER partial current densities (Supplementary Fig. \u003cb\u003eS1\u003c/b\u003e). Cyclic voltammograms (CVs) using the RRDE were conducted at 3,600 rpm in an H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e solution (pH 0.90\u0026thinsp;\u0026plusmn;\u0026thinsp;0.03) supplemented with either 50 mM HCl or 50 mM MCl (M\u0026thinsp;=\u0026thinsp;Li\u003csup\u003e+\u003c/sup\u003e, Na\u003csup\u003e+\u003c/sup\u003e, K\u003csup\u003e+\u003c/sup\u003e, and Cs\u003csup\u003e+\u003c/sup\u003e; Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). For all electrolytes, the disk current densities (\u003cem\u003ej\u003c/em\u003e\u003csub\u003edisk\u003c/sub\u003e) increased as the potential increased versus the reversible hydrogen electrode (V\u003csub\u003eRHE\u003c/sub\u003e), as the CER and OER occurred (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). A constant potential of 0.95 V\u003csub\u003eRHE\u003c/sub\u003e was applied to the Pt ring to selectively reduce Cl\u003csub\u003e2\u003c/sub\u003e generated from the disk\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. Using this setup, more Cl\u003csub\u003e2\u003c/sub\u003e was generated from the IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e/GC disk in electrolyte containing Cs\u003csup\u003e+\u003c/sup\u003e than Li\u003csup\u003e+\u003c/sup\u003e because a more negative ring current density (\u0026minus;\u0026thinsp;\u003cem\u003ej\u003c/em\u003e\u003csub\u003ering\u003c/sub\u003e) was detected in the presence of Cs\u003csup\u003e+\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb).\u003c/p\u003e \u003cp\u003eTo directly compare the cation dependence of \u003cem\u003ej\u003c/em\u003e\u003csub\u003ering\u003c/sub\u003e, the partial current densities of CER (\u003cem\u003ej\u003c/em\u003e\u003csub\u003eCER\u003c/sub\u003e) and OER (\u003cem\u003ej\u003c/em\u003e\u003csub\u003eOER\u003c/sub\u003e) were calculated for each electrolyte (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec, see \u003cb\u003eMethods\u003c/b\u003e). When the \u003cem\u003eiR\u003c/em\u003e-corrected potential exceeded 1.65 V\u003csub\u003eRHE\u003c/sub\u003e, the observed plateau of \u003cem\u003ej\u003c/em\u003e\u003csub\u003eCER\u003c/sub\u003e (\u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e) was almost potential-independent, demonstrating that the CER rate was limited by the mass transport of chloride ions (Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e). In addition, \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e measured in the presence of CsCl was around 1.5 times higher than that of LiCl for all examined rpm values (Supplementary Fig. \u003cb\u003eS2\u003c/b\u003e), demonstrating that the influence of cations on Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e transport was rpm-independent. \u003cem\u003ej\u003c/em\u003e\u003csub\u003eOER\u003c/sub\u003e decreased as \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e increased for the different cations, consistent with the competing nature of two reactions\u003csup\u003e\u003cspan additionalcitationids=\"CR10 CR11 CR12\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. Accordingly, the CER selectivity in the presence of Li\u003csup\u003e+\u003c/sup\u003e was at least 20% lower than in Cs\u003csup\u003e+\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed). Tafel slopes of the CER and the OER in the absence of Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e were consistent with the representative values for the CER and OER in acidic media\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e (Supplementary Figs. \u003cb\u003eS3\u0026thinsp;\u0026minus;\u0026thinsp;5\u003c/b\u003e), demonstrating a minimal influence of electrolytic cation on the intrinsic activity of IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e on both CER\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e and OER\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. Overall, the RRDE measurements demonstrate that the CER activity at the mass-transport limit is cation-dependent and emphasize the importance of electrolyte composition in controlling catalytic selectivity between CER and OER.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eLevich and Kouteck\u0026yacute;\u0026minus;Levich analyses.\u003c/b\u003e To investigate the cation effects on Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e transport during CER, \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e was plotted with respect to the electrode rpm (Supplementary Fig. \u003cb\u003eS2\u003c/b\u003e) using the following Levich Eq.\u0026nbsp;2\u003csup\u003e5,29\u003c/sup\u003e:\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:{j}_{\\text{l}\\text{i}\\text{m}}=0.62\\bullet\\:n\\bullet\\:F\\bullet\\:{C}_{\\text{b}\\text{u}\\text{l}\\text{k}}\\bullet\\:{\\nu\\:}^{-1/6}\\bullet\\:{D}^{2/3}\\bullet\\:{\\omega\\:}^{1/2}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cem\u003en\u003c/em\u003e represents the number of electrons transferred for CER (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2), \u003cem\u003eF\u003c/em\u003e is the Faraday constant (96,485 C∙mol\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e), \u003cem\u003eC\u003c/em\u003e\u003csub\u003ebulk\u003c/sub\u003e is the bulk Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e concentration (mol∙cm\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e), \u003cem\u003eν\u003c/em\u003e is the kinematic viscosity of the electrolyte (cm\u003csup\u003e2\u003c/sup\u003e∙s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e), \u003cem\u003eD\u003c/em\u003e is the Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion coefficient in bulk electrolyte (cm\u003csup\u003e2\u003c/sup\u003e∙s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e), and \u003cem\u003eω\u003c/em\u003e is the rotational speed of RRDE (rad∙s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e or π/30 rpm). In all electrolytes, \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e was linearly proportional to \u003cem\u003eω\u003c/em\u003e\u003csup\u003e1/2\u003c/sup\u003e, indicating that Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion predominantly contributed to mass transport for the CER (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). Based on Eq.\u0026nbsp;(\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), the cation-dependent values of \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e are expected to be due to changes in \u003cem\u003eD\u003c/em\u003e because changes in \u003cem\u003eν\u003c/em\u003e resulted in shifts of less than 1 mA∙cm\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e (Supplementary Fig. \u003cb\u003eS6\u003c/b\u003e). However, conventional RRDE theory does not consider \u003cem\u003eD\u003c/em\u003e to be dependent on the electrolyte composition, contrary to the values of \u003cem\u003eD\u003c/em\u003e estimated from the Levich slopes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb). Furthermore, all the cations show a positive intercept value exceeding\u0026thinsp;\u0026gt;\u0026thinsp;6 mA∙cm\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e, which is significant, as \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e must extrapolate towards zero at zero rpm according to Eq.\u0026nbsp;(\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eSo far, positive intercepts have been interpreted as resulting from non-diffusional factors\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. For example, when electron transfer rate is not sufficiently fast for Cl\u003csup\u003e\u003cem\u003e\u0026minus;\u003c/em\u003e\u003c/sup\u003e diffusion, the intercept corresponds to the kinetic current density of charge transfer (\u003cem\u003ej\u003c/em\u003e\u003csub\u003eK\u003c/sub\u003e), as represented in the Kouteck\u0026yacute;\u0026minus;Levich (KL) Eq.\u0026nbsp;2\u003csup\u003e5,29\u003c/sup\u003e:\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$$\\:{j}_{\\text{C}\\text{E}\\text{R}}^{-1}={j}_{\\text{l}\\text{i}\\text{m}}^{-1}+{j}_{\\text{K}}^{-1}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e2\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cem\u003ej\u003c/em\u003e\u003csub\u003eK\u003c/sub\u003e indicates the kinetic current density of charge transfer for the CER. However, the ratio of \u003cem\u003ej\u003c/em\u003e\u003csub\u003eK\u003c/sub\u003e/\u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e exceeded\u0026thinsp;\u0026gt;\u0026thinsp;500 for all electrolytes and rpm values, demonstrating that charge transfer was sufficiently fast (Supplementary Fig. \u003cb\u003eS7\u003c/b\u003e). Therefore, the positive intercepts made by Eq.\u0026nbsp;(\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) must indicate the cation effect on Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion, suggesting the presence of another rpm-independent diffusion layer.\u003c/p\u003e \u003cp\u003e \u003cb\u003eDouble diffusion model of semi-permeable layer.\u003c/b\u003e Gough and Leypoldt proposed a modified KL equation by including rpm-independent current densities (\u003cem\u003ej\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e), which result from the reactant diffusion in a semi-permeable polymer film\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. The following equation is applied for the diffusion-limiting condition:\u003cdiv id=\"Equ3\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ3\" name=\"EquationSource\"\u003e\n$$\\:{j}_{\\text{l}\\text{i}\\text{m}}^{-1}={j}_{\\text{c}\\text{o}\\text{n}\\text{v}}^{-1}+{j}_{\\text{f}}^{-1}=\\frac{1}{0.62\\bullet\\:n\\bullet\\:F\\bullet\\:{C}_{\\text{b}\\text{u}\\text{l}\\text{k}}\\bullet\\:{\\nu\\:}^{-1/6}\\bullet\\:{{D}_{\\text{c}\\text{o}\\text{n}\\text{v}}}^{2/3}}\\bullet\\:{\\omega\\:}^{-1/2}+\\frac{{L}_{\\text{f}}}{n\\bullet\\:F\\bullet\\:{C}_{\\text{f}}\\bullet\\:{D}_{\\text{f}}}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e3\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eThe separation of \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e\u003csup\u003e\u0026minus;1\u003c/sup\u003e into the reciprocal current densities of conventional (\u003cem\u003ej\u003c/em\u003e\u003csub\u003econv\u003c/sub\u003e\u003csup\u003e\u0026minus;1\u003c/sup\u003e) and film-originated (\u003cem\u003ej\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e\u003csup\u003e\u0026minus;1\u003c/sup\u003e) diffusion layers newly defines the Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion coefficient in the conventional diffusion layer (\u003cem\u003eD\u003c/em\u003e\u003csub\u003econv\u003c/sub\u003e), the Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion coefficient (\u003cem\u003eD\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e) and Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e concentration in a film (\u003cem\u003eC\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e), and the film thickness (\u003cem\u003eL\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e). The value of \u003cem\u003ej\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e\u003csup\u003e\u0026minus;1\u003c/sup\u003e is rpm-independent and determined solely by the properties of the semi-permeable polymer. For example, gaseous reactants, such as H\u003csub\u003e2\u003c/sub\u003e and O\u003csub\u003e2\u003c/sub\u003e, exhibit a linear correlation between \u003cem\u003ej\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e\u003csup\u003e\u0026minus;1\u003c/sup\u003e and Nafion thickness\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. Here, we revisited previous results for H\u003csub\u003e2\u003c/sub\u003e and O\u003csub\u003e2\u003c/sub\u003e diffusion through Nafion-coated electrodes and confirmed that \u003cem\u003eD\u003c/em\u003e\u003csub\u003econv\u003c/sub\u003e is nearly independent of \u003cem\u003eL\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e (Supplementary Figs. \u003cb\u003eS8\u0026thinsp;\u0026minus;\u0026thinsp;10\u003c/b\u003e).\u003c/p\u003e \u003cp\u003eAlthough no polymer film was present on the IrO\u003csub\u003ex\u003c/sub\u003e/GC disk in this study, Eq.\u0026nbsp;(\u003cspan refid=\"Equ3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) can be used to rationalize the cation-dependent intercepts (\u003cem\u003ej\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e\u003csup\u003e\u0026minus;1\u003c/sup\u003e) by considering an additional, rpm-independent diffusion barrier, similar to that arising from polymer films (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec,d). Consistent with the \u003cem\u003eL\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e-independent \u003cem\u003eD\u003c/em\u003e\u003csub\u003econv\u003c/sub\u003e values of polymer films\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e (Supplementary Figs. \u003cb\u003eS8d\u0026thinsp;\u0026minus;\u0026thinsp;10d\u003c/b\u003e), the values of \u003cem\u003eD\u003c/em\u003e\u003csub\u003econv\u003c/sub\u003e calculated by Eq.\u0026nbsp;(\u003cspan refid=\"Equ3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) were less cation-dependent than \u003cem\u003eD\u003c/em\u003e by Eq.\u0026nbsp;(\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb), indicating that this cation-dependent layer suppressed Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion similar to polymer films. Although the thickness of cation-dependent layer was not experimentally controllable, we hypothesized that an increase in M\u003csup\u003e+\u003c/sup\u003e concentration could intensify the effects of the cation-dependent barrier, if the layer is controlled by cations.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eEffects of cation concentration on Cl\u003c/b\u003e \u003csup\u003e \u003cb\u003e\u0026minus;\u003c/b\u003e \u003c/sup\u003e \u003cb\u003ediffusion.\u003c/b\u003e To observe how M\u003csup\u003e+\u003c/sup\u003e concentration influences Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion, chronoamperometry (CA) was conducted in H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e electrolyte supplemented with 25 mM MCl (pH 0.90\u0026thinsp;\u0026plusmn;\u0026thinsp;0.03), and M\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e was added during the measurements to increase M\u003csup\u003e+\u003c/sup\u003e concentration. The diffusion-limiting potential of 2.1 V\u003csub\u003eRHE\u003c/sub\u003e (no \u003cem\u003eiR\u003c/em\u003e-correction; Supplementary Fig. \u003cb\u003eS11\u003c/b\u003e) was maintained during the measurements. Current densities from the last half of each CA (7.5 s) were normalized to analyse steady state behaviour (Supplementary Fig. \u003cb\u003eS12\u003c/b\u003e). Both Levich and \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e\u003csup\u003e\u003cem\u003e\u0026minus;\u003c/em\u003e1\u003c/sup\u003e versus \u003cem\u003eω\u003c/em\u003e\u003csup\u003e\u003cem\u003e\u0026minus;\u003c/em\u003e1/2\u003c/sup\u003e plots exhibited good linearities in all measured M\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e concentrations (Supplementary Fig. \u003cb\u003eS13\u003c/b\u003e). \u003cem\u003ej\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e\u003csup\u003e\u0026minus;1\u003c/sup\u003e values measured in electrolyte containing 25 mM MCl were also cation-dependent (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). Adding M\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e gradually decreased \u003cem\u003ej\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e\u003csup\u003e\u0026minus;1\u003c/sup\u003e, with Cs\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e yielding the most rapid decrease of \u003cem\u003ej\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e\u003csup\u003e\u0026minus;1\u003c/sup\u003e within M\u003csup\u003e+\u003c/sup\u003e examined in this study. In addition, the mixing of Li\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e and Cs\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e revealed that Cs\u003csup\u003e+\u003c/sup\u003e had a greater effect on \u003cem\u003ej\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e\u003csup\u003e\u0026minus;1\u003c/sup\u003e than Li\u003csup\u003e+\u003c/sup\u003e (Supplementary Fig. \u003cb\u003eS14\u003c/b\u003e). This finding indicates that M\u003csup\u003e+\u003c/sup\u003e concentration affects \u003cem\u003ej\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e\u003csup\u003e\u0026minus;1\u003c/sup\u003e but cannot change the order of \u003cem\u003ej\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e\u003csup\u003e\u0026minus;1\u003c/sup\u003e made by different M\u003csup\u003e+\u003c/sup\u003e. The slight decrease in \u003cem\u003ej\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e\u003csup\u003e\u0026minus;1\u003c/sup\u003e upon adding M\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e might be due to the increased total electrolyte concentration. Consistent with the effect of different M\u003csup\u003e+\u003c/sup\u003e on Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), Eq.\u0026nbsp;(\u003cspan refid=\"Equ3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) described cation-independence of \u003cem\u003eD\u003c/em\u003e\u003csub\u003econv\u003c/sub\u003e for different M\u003csup\u003e+\u003c/sup\u003e concentrations (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb), whereas Eq.\u0026nbsp;(\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) presented cation-dependent \u003cem\u003eD\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). Taken together, the effects of cation species (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) and concentration (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) on Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion reveal that the cation dependency in Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion coefficients is due to the cation-dependent diffusion layers.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eNumerical simulation of the cation effects on Cl\u003c/b\u003e \u003csup\u003e \u003cb\u003e\u0026minus;\u003c/b\u003e \u003c/sup\u003e \u003cb\u003ediffusion.\u003c/b\u003e Rpm-independent diffusion barriers are a characteristic feature of polymer-coated electrodes\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. This phenomenon is generally attributed to a discontinuous concentration gradient of the reactant, due to different reactant solubilities across the polymer-electrolyte interface\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e,\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. However, our unmodified IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e electrode has no physical boundaries, leading us to hypothesize that the reactant concentration gradient is continuous across electrolyte interfaces (see Supplementary \u003cb\u003eNote S1\u003c/b\u003e). In our proposed double diffusion model for Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea), the cation-dependent layer is located inside the conventional diffusion layer, whose thickness follows Levich theory (\u003cem\u003eδ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.61∙\u003cem\u003eD\u003c/em\u003e\u003csub\u003econv\u003c/sub\u003e\u003csup\u003e1/3\u003c/sup\u003e∙\u003cem\u003eν\u003c/em\u003e\u003csup\u003e1/6\u003c/sup\u003e∙\u003cem\u003eω\u003c/em\u003e\u003csup\u003e\u003cem\u003e\u0026minus;\u003c/em\u003e1/2\u003c/sup\u003e). The same notational parameters used in Eq.\u0026nbsp;(\u003cspan refid=\"Equ3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) can be applied to this model. Based on this model, \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e at sufficiently high \u003cem\u003ej\u003c/em\u003e\u003csub\u003eK\u003c/sub\u003e can be expressed as:\u003cdiv id=\"Equ4\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ4\" name=\"EquationSource\"\u003e\n$$\\:{j}_{\\text{l}\\text{i}\\text{m}}=n\\bullet\\:F\\bullet\\:{C}_{\\text{b}\\text{u}\\text{l}\\text{k}}\\bullet\\:{\\left(\\frac{{L}_{\\text{f}}}{{D}_{\\text{f}}}+\\frac{\\delta\\:}{{D}_{\\text{c}\\text{o}\\text{n}\\text{v}}}\\right)}^{-1}=\\frac{n\\bullet\\:F\\bullet\\:{C}_{\\text{b}\\text{u}\\text{l}\\text{k}}}{\\alpha\\:+{\\beta\\:}^{{\\prime\\:}}\\bullet\\:{\\omega\\:}^{-1/2}}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e4\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eIts inversion yields:\u003cdiv id=\"Equ5\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ5\" name=\"EquationSource\"\u003e\n$$\\:{j}_{\\text{l}\\text{i}\\text{m}}^{-1}={j}_{\\text{c}\\text{o}\\text{n}\\text{v}}^{-1}+{j}_{\\text{f}}^{-1}=\\frac{{\\beta\\:}^{{\\prime\\:}}}{n\\bullet\\:F\\bullet\\:{C}_{\\text{b}\\text{u}\\text{l}\\text{k}}}\\bullet\\:{\\omega\\:}^{-1/2}+\\frac{\\alpha\\:}{n\\bullet\\:F\\bullet\\:{C}_{\\text{b}\\text{u}\\text{l}\\text{k}}}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e5\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cem\u003eα\u003c/em\u003e and \u003cem\u003eβ\u003c/em\u003e indicate \u003cem\u003eL\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e/\u003cem\u003eD\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e and \u003cem\u003eδ\u003c/em\u003e/\u003cem\u003eD\u003c/em\u003e\u003csub\u003econv\u003c/sub\u003e, respectively. \u003cem\u003eβ\u003c/em\u003e\u0026prime; is the rpm-independent component of \u003cem\u003eβ\u003c/em\u003e, determined by \u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;\u003cem\u003eβ\u003c/em\u003e\u0026prime;\u003cem\u003e∙ω\u003c/em\u003e\u003csup\u003e\u0026minus;1/2\u003c/sup\u003e. Since the only two free variables in this model are \u003cem\u003eα\u003c/em\u003e and \u003cem\u003eβ\u003c/em\u003e, they can be determined unambiguously from the slopes and intercepts in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec. The cation effects from both species (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) and concentrations (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) were applied to the double diffusion model at finite \u003cem\u003eω\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb). According to the obtained slope of 1.8 in the comparison of \u003cem\u003ej\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e\u003csup\u003e\u0026minus;1\u003c/sup\u003e between 25 mM and 50 mM MCl (Supplementary Fig. \u003cb\u003eS15\u003c/b\u003e), \u003cem\u003eC\u003c/em\u003e\u003csub\u003ebulk\u003c/sub\u003e is also important to calculate \u003cem\u003eα\u003c/em\u003e. Because of the continuous concentration gradient, the boundary Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e concentration between two diffusion layers (\u003cem\u003eC\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e) depends on \u003cem\u003eω\u003c/em\u003e and is equal to \u003cem\u003eC\u003c/em\u003e\u003csub\u003ebulk\u003c/sub\u003e at infinite \u003cem\u003eω\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea). Even so,\u003cdiv id=\"Equ6\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ6\" name=\"EquationSource\"\u003e\n$$\\:{j}_{\\text{l}\\text{i}\\text{m}(\\omega\\:,k\\to\\:\\infty\\:)}^{-1}={{j}_{\\text{f}}}^{-1}=\\frac{{L}_{\\text{f}}}{n\\bullet\\:F\\bullet\\:{D}_{\\text{f}}\\bullet\\:{C}_{\\text{b}\\text{u}\\text{l}\\text{k}}}\u0026gt;0$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e6\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003edemonstrates that \u003cem\u003eL\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e and \u003cem\u003eD\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e reproduce the positive intercept of the experimental plots in an rpm-independent manner (Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eNumerical simulations of the double diffusion model combined with Butler\u0026thinsp;\u0026minus;\u0026thinsp;Volmer kinetics (details in Supplementary \u003cb\u003eNote S1\u003c/b\u003e) produced sigmoidal \u003cem\u003ej\u003c/em\u003e\u003csub\u003eCER\u003c/sub\u003e plots (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec), which matched well with the experimentally determined \u003cem\u003ej\u003c/em\u003e\u003csub\u003eCER\u003c/sub\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec). Thus, without considering the potential dependence of \u003cem\u003eα\u003c/em\u003e, the cation-dependent layer is unrelated to the applied potential. The simulated Levich plots were curved, indicating that \u003cem\u003eβ\u003c/em\u003e has a greater influence on \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e as \u003cem\u003eβ\u003c/em\u003e approaches zero due to infinite \u003cem\u003eω\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed). Conversely, the influence of \u003cem\u003eα\u003c/em\u003e diminished as \u003cem\u003eβ\u003c/em\u003e goes infinite, yielding a zero intercept at \u003cem\u003eω\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0. The simulated \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e\u003csup\u003e\u003cem\u003e\u0026minus;\u003c/em\u003e1\u003c/sup\u003e versus \u003cem\u003eω\u003c/em\u003e\u003csup\u003e\u003cem\u003e\u0026minus;\u003c/em\u003e1/2\u003c/sup\u003e plots were linear, confirming that the conventional diffusion layer is entirely rpm-dependent (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee). Therefore, Eq.\u0026nbsp;(\u003cspan refid=\"Equ5\" class=\"InternalRef\"\u003e5\u003c/span\u003e) of double diffusion model is more suitable to rationalize the present results than Eq.\u0026nbsp;(\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), which could not explain the anomalous, cation-dependent intercepts. Parameters obtained from the double diffusion model are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003eFor the conventional diffusion layer, \u003cem\u003eβ\u003c/em\u003e was weakly cation-dependent following the same series of the mutual diffusion coefficients of bulk MCl solutions\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e (\u003cem\u003eD\u003c/em\u003e\u003csub\u003eMCl\u003c/sub\u003e; Supplementary Fig. \u003cb\u003eS6c\u003c/b\u003e). The obtained \u003cem\u003eD\u003c/em\u003e\u003csub\u003econv\u003c/sub\u003e values are therefore reasonable, as they are only\u0026thinsp;~\u0026thinsp;2 times lower than \u003cem\u003eD\u003c/em\u003e\u003csub\u003eMCl\u003c/sub\u003e values. On the contrary, the experimental measurements of \u003cem\u003eL\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e are challenging because the double diffusion model is only applicable at the high \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e with intensive Cl\u003csub\u003e2\u003c/sub\u003e and O\u003csub\u003e2\u003c/sub\u003e evolutions. In an attempt to estimate \u003cem\u003eL\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e, we considered the electrical double layer (EDL; Supplementary Fig. \u003cb\u003eS16\u003c/b\u003e), which exist on the innermost electrode-electrolyte interface during electrocatalysis\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e,\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. The EDL is composed of the inner Helmholtz double layer and the outer diffuse layer, which were omitted in our double diffusion model (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea,b) because cations negligibly affect charge transfer (Supplementary Fig. \u003cb\u003eS3\u003c/b\u003e). The EDL is highly ion-concentrated\u003csup\u003e\u003cspan additionalcitationids=\"CR35 CR36 CR37 CR38\" citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. The cation concentrations of the diffuse layer need to be higher than the bulk to complement the anion-concentrated Helmholtz layer, collectively balancing the positively charged electrode surface. Given the rpm-independent nature of the EDL\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e,\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e, the cation-dependent layer can be hypothesized as an extension of the outer diffuse layer of the EDL. Based on this assumption, we calculated \u003cem\u003eD\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e assuming \u003cem\u003eL\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e to be in the range of 1\u0026thinsp;\u0026minus;\u0026thinsp;10 nm (below \u0026lt;\u0026thinsp;0.2% of \u003cem\u003eδ\u003c/em\u003e), which is larger than the EDL thickness (~\u0026thinsp;0.9 nm) calculated by the Gouy\u0026thinsp;\u0026minus;\u0026thinsp;Chapman theory at a minimum of total electrolyte concentration of this study (0.125 M)\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. The expected \u003cem\u003eD\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e is at least two orders of magnitude lower than \u003cem\u003eD\u003c/em\u003e\u003csub\u003econv\u003c/sub\u003e, indicating that cation-dependent layers are a significant barrier to Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion.\u003c/p\u003e \u003cp\u003e \u003cb\u003eCorrelation between the cation-dependent diffusion barrier and the structure of water network.\u003c/b\u003e Although ions can be highly concentrated in EDL, ~\u0026thinsp;56 M of H\u003csub\u003e2\u003c/sub\u003eO solvent is sufficient to fully hydrate each cation\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. Solvated ions that interact with nearby water molecules, inducing disruption of the original hydrogen-bond network, are called \u0026ldquo;water-structure breaker\u0026rdquo; according to Hofmeister\u0026rsquo;s classification\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. To identify which properties of cations represent the cation-dependent diffusion barrier, \u003cem\u003eL\u003c/em\u003e\u003csub\u003ef/\u003c/sub\u003e\u003cem\u003eD\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e was compared with the structural entropy of hydration\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e (∆\u003csub\u003estruc\u003c/sub\u003e\u003cem\u003eS\u003c/em\u003e; Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e) based on previous studies that examined the relationships among thermodynamic entropies, viscosities, and diffusion coefficients in aqueous solutions\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e,\u003cspan additionalcitationids=\"CR42 CR43\" citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. The following equation, proposed by Marcus\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e, was used to determine ∆\u003csub\u003estruc\u003c/sub\u003e\u003cem\u003eS\u003c/em\u003e values for the monovalent ions:\u003cdiv id=\"Equ7\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ7\" name=\"EquationSource\"\u003e\n$$\\:{\\varDelta\\:}_{\\text{s}\\text{t}\\text{r}\\text{u}\\text{c}}S\\:\\left(\\text{J}\\bullet\\:{\\text{K}}^{-1}\\bullet\\:{\\text{m}\\text{o}\\text{l}}^{-1}\\right)=40-606\\bullet\\:B\\:$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e7\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cem\u003eB\u003c/em\u003e indicates the viscosity B-coefficient, which mainly depends on the ion-solvent interactions of the Jones\u0026thinsp;\u0026minus;\u0026thinsp;Dole empirical expression\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e at 25\u0026deg;C. The values of experimental \u003cem\u003eB\u003c/em\u003e selected by Jenkins and Marcus were used\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. Since the \u003cem\u003eB\u003c/em\u003e values of an aqueous MCl solution are expected to be valid up to 4 M [46], we considered that Eq.\u0026nbsp;(\u003cspan refid=\"Equ7\" class=\"InternalRef\"\u003e7\u003c/span\u003e) is applicable for cation-dependent layer as an extension of ion-concentrated EDL.\u003c/p\u003e \u003cp\u003eA linear correlation between \u003cem\u003eL\u003c/em\u003e\u003csub\u003ef/\u003c/sub\u003e\u003cem\u003eD\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e and ∆\u003csub\u003estruc\u003c/sub\u003e\u003cem\u003eS\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e) supports the hypothesis that the cation-dependent water structure represents the cation-dependent diffusion barrier. Specifically, the negative ∆\u003csub\u003estruc\u003c/sub\u003e\u003cem\u003eS\u003c/em\u003e values of Li\u003csup\u003e+\u003c/sup\u003e and Na\u003csup\u003e+\u003c/sup\u003e indicate that the solvation of these cations results in an organized water network, whereas the positive ∆\u003csub\u003estruc\u003c/sub\u003e\u003cem\u003eS\u003c/em\u003e values of K\u003csup\u003e+\u003c/sup\u003e and Cs\u003csup\u003e+\u003c/sup\u003e represent a disrupted water network. In the absence of M\u003csup\u003e+\u003c/sup\u003e, H\u003csup\u003e+\u003c/sup\u003e exhibits slightly negative ∆\u003csub\u003estruc\u003c/sub\u003e\u003cem\u003eS\u003c/em\u003e, consistent with the RRDE measurements performed in H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e electrolyte supplemented with HCl (Supplementary Figs. \u003cb\u003eS2c\u003c/b\u003e and \u003cb\u003eS17\u003c/b\u003e). This finding indicates that blank H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e electrolyte enables solvated M\u003csup\u003e+\u003c/sup\u003e to organize or disrupt H\u003csub\u003e3\u003c/sub\u003eO\u003csup\u003e+\u003c/sup\u003e-H\u003csub\u003e2\u003c/sub\u003eO network, decreasing or increasing the diffusion barrier against Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e, respectively. Since ∆\u003csub\u003estruc\u003c/sub\u003e\u003cem\u003eS\u003c/em\u003e excludes electrostatic forces\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e, the structural effect of cations predominates in suppressing Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion. Including electrostatic effects in thermodynamic parameters also exhibits a linear relationship between \u003cem\u003eL\u003c/em\u003e\u003csub\u003ef/\u003c/sub\u003e\u003cem\u003eD\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e and ∆\u003csub\u003estruc\u003c/sub\u003e\u003cem\u003eS\u003c/em\u003e for M\u003csup\u003e+\u003c/sup\u003e species (Supplementary Fig. \u003cb\u003eS18\u003c/b\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIn this work, RRDE analyses were used to study the cation effects on suppressing Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion in acidic electrolytes with tens of millimolar Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e concentrations. The diffusion-limiting CER current densities (\u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e) was cation-dependent and significantly influenced the CER selectivity. This suggests that the OER selectivity in the presence of Li\u003csup\u003e+\u003c/sup\u003e can be further increased in higher applied potentials because the OER is not limited by diffusion. By measuring different cation species and concentrations, the cation-dependent intercepts (\u003cem\u003ej\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e\u003csup\u003e\u0026minus;1\u003c/sup\u003e) in \u003cem\u003ej\u003c/em\u003e\u003csub\u003elim\u003c/sub\u003e\u003csup\u003e\u003cem\u003e\u0026minus;\u003c/em\u003e1\u003c/sup\u003e versus \u003cem\u003eω\u003c/em\u003e\u003csup\u003e\u003cem\u003e\u0026minus;\u003c/em\u003e1/2\u003c/sup\u003e plots indicated that the cation-dependent layer existed independently of electrode rpm. The proposed double diffusion model attributes the rpm-independent diffusion barrier against Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion to the cation-dependent layer. This finding allows us to understand cation-dependent Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion based on the diffusion theory of semi-permeable layers. Among the cations examined here, the strength of the diffusion barriers was ranked in the order of LiCl\u0026thinsp;\u0026gt;\u0026thinsp;NaCl\u0026thinsp;\u0026gt;\u0026thinsp;HCl\u0026thinsp;\u0026gt;\u0026thinsp;KCl\u0026thinsp;\u0026gt;\u0026thinsp;CsCl and exhibited a clear linearity with the structural entropy of hydration (∆\u003csub\u003estruc\u003c/sub\u003e\u003cem\u003eS\u003c/em\u003e) of each cation. This result indicates that the rigid hydrogen bond network in the presence of Li\u003csup\u003e+\u003c/sup\u003e and Na\u003csup\u003e+\u003c/sup\u003e decreases the Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e diffusion coefficients in the cation-dependent layer (\u003cem\u003eD\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e) due to negative ∆\u003csub\u003estruc\u003c/sub\u003e\u003cem\u003eS\u003c/em\u003e. On the contrary, Cs\u003csup\u003e+\u003c/sup\u003e and K\u003csup\u003e+\u003c/sup\u003e have positive ∆\u003csub\u003estruc\u003c/sub\u003e\u003cem\u003eS\u003c/em\u003e, indicating that the hydrogen bond network is disrupted, exhibiting increased \u003cem\u003eD\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e. Given that diffusion is independent of the specific reactant-catalyst combination, ion-dependent reactant diffusion is also applicable to other electrochemical reactions having issues with ionic impurities, and thus may promote the development of efficient low-grade water electrolysis.\u003c/p\u003e "},{"header":"Methods","content":"\u003cp\u003e \u003cb\u003eChemicals.\u003c/b\u003e Li\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e (99.7%) and Na\u003csub\u003e2\u003c/sub\u003eIrCl\u003csub\u003e6\u003c/sub\u003e∙6H\u003csub\u003e2\u003c/sub\u003eO (33.9% Ir) were purchased from Thermo Scientific. H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e (97%), HClO\u003csub\u003e4\u003c/sub\u003e (70%), anhydrous LiCl (\u0026gt;\u0026thinsp;99.0%), KCl (\u0026gt;\u0026thinsp;99.5%), CsCl (\u0026gt;\u0026thinsp;99.0%), Na\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e (\u0026gt;\u0026thinsp;99.0%), and K\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e (\u0026gt;\u0026thinsp;99.0%) were purchased from Fujifilm Wako Pure Chemicals. Cs\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e (\u0026gt;\u0026thinsp;99.0%) and Na\u003csub\u003e3\u003c/sub\u003eIrCl\u003csub\u003e6\u003c/sub\u003e∙\u003cem\u003ex\u003c/em\u003eH\u003csub\u003e2\u003c/sub\u003eO (35\u0026thinsp;\u0026minus;\u0026thinsp;40% Ir) were purchased from Sigma-Aldrich. HCl (35\u0026thinsp;\u0026minus;\u0026thinsp;37%), NaCl (\u0026gt;\u0026thinsp;99.5%), and NaOH (97%) were purchased from Junsei Chemical. 1 M HCl standard solution was purchased from Hayashi Pure Chemical. All chemicals were used as received unless otherwise noted. A Milli-Q\u0026reg; EQ-7008 system (Merck Millipore) was used to prepare DI water (18.2 MΩ∙cm).\u003c/p\u003e \u003cp\u003e \u003cb\u003eGeneral electrochemical procedures.\u003c/b\u003e All experiments were carried out in a fume hood at room temperature (20\u0026thinsp;\u0026minus;\u0026thinsp;25\u0026deg;C). Before electrochemical measurements, all glassware was boiled in DI water for 1 h, washed with DI water several times, and then dried in a 70\u0026deg;C oven, sequentially. When not being used, all glassware was stored in a 6 M HCl bath to remove ionic residues. Electrochemical measurements were conducted using two-compartment borosilicate glass cells divided by the pretreated Nafion 115 membrane (0.005-inch thickness, Fuel Cell Earth), unless otherwise noted. The Nafion 115 membranes (2\u0026times;2 cm\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e) were boiled in 3 wt% of H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e, DI water, 1 M of H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e, and DI water for one hour at a time, sequentially. Before use, the pretreated Nafion 115 was stored in DI water for up to 3 months. A dual-channel potentiostat (SP-150e, BioLogic) was used for all electrochemical experiments. The GC disk of HR2-RD1 GC-Pt RRDE with polyether ether ketone (PEEK) spacer and shroud (Hokuto Denko), while its rpm was controlled by a HR-500 digital rotator (Hokoto Denko). We noted that the PEEK spacer was indispensable to prevent the Cl\u003csub\u003e2\u003c/sub\u003e/O\u003csub\u003e2\u003c/sub\u003e bubbles lodged on the RRDE spacer that severely disturbed the laminar flow from disk to ring (Supplementary Fig. \u003cb\u003eS19\u003c/b\u003e). The geometrical area of GC disk and poly Pt ring were 0.196 cm\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e (radius, \u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2.5 mm) and 0.265 cm\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e (inner \u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2.75 mm and outer \u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;4 mm), respectively. To obtain mirror-finished GC and Pt surfaces before all measurements, electrodes were hand-polished on microcloth (EC Frontier) with MetaDi Supreme diamond paste suspension (Buehler) of different particle sizes from 6, 1, to 0.05 \u0026micro;m, sequentially. Before changing the polishing pad with different suspension, RRDE was washed with copious amounts of DI water. Displayed potentials in electrochemical data were 85% iR-corrected after measurements (denoted as \u003cem\u003eE\u0026thinsp;\u0026minus;\u0026thinsp;iR\u003c/em\u003e), except for chronoamperometry, IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e electrodeposition, and collection efficiency measurements. Electrochemical impedance spectroscopy (EIS) was conducted at an open circuit potential (OCP) to measure the solution resistance in the high-frequency domain (100\u0026thinsp;\u0026minus;\u0026thinsp;100,000 Hz) at a zero-degree phase angle. The EIS was measured before and after conducting the set of measurements. Measured OCPs were around 0.4 V\u003csub\u003eRHE\u003c/sub\u003e and 1.1 V\u003csub\u003eRHE\u003c/sub\u003e before and after measurements, independent of electrode rpm, respectively.\u003c/p\u003e \u003cp\u003eThe blank 0.1 M H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e electrolytes were prepared by diluting 97% H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e with DI water. After adding MCl and M\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e (M\u0026thinsp;=\u0026thinsp;Li\u003csup\u003e+\u003c/sup\u003e, Na\u003csup\u003e+\u003c/sup\u003e, K\u003csup\u003e+\u003c/sup\u003e, and Cs\u003csup\u003e+\u003c/sup\u003e) into the blank 0.1 M H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e, the solution pH was adjusted to the range of 0.90\u0026thinsp;\u0026plusmn;\u0026thinsp;0.03 by adding a few drops of 97% H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e. For the H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e electrolyte supplemented with HCl, 1 M HCl standard solution was diluted with DI water and then adjusted to pH 0.90\u0026thinsp;\u0026plusmn;\u0026thinsp;0.03 by adding 97% H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e. A ROSS pH electrode (8220BNWP, Thermo Scientific) connected with a digital meter (Orion Star A211, Thermo Scientific) was used to measure the solution pH. For the RRDE experiments, 70 mL of electrolyte in the working electrode compartment was directly purged with Ar gas (Taiyo Nippon Sanso JFP, G3 grade, 99.999%) for more than 10 minutes and kept a gas blanket during measurements (0.2 L∙min\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e). The reference electrode was a KCl-saturated Ag/AgCl electrode (EC Frontier, RE-7A, V\u003csub\u003eAg/AgCl\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;+\u0026thinsp;0.197 V vs. standard hydrogen electrode) with a Vycor porous glass frit, equipped with another glass-fritted junction filled with measured electrolytes to minimize Ag\u003csup\u003e+\u003c/sup\u003e and Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e leakage. Unless noted otherwise, all potential units in this manuscript were noted in the reversible hydrogen electrode (RHE) scale by the equation as V\u003csub\u003eRHE\u003c/sub\u003e = V\u003csub\u003eAg/AgCl\u003c/sub\u003e + (0.059∙pH) V\u0026thinsp;+\u0026thinsp;0.197 V. A Pt coil (r\u0026thinsp;=\u0026thinsp;1 mm and length\u0026thinsp;=\u0026thinsp;5 cm) annealed by hydrogen flame was used as a counter electrode. All standard errors of the mean were calculated based on three independent measurements unless noted otherwise.\u003c/p\u003e \u003cp\u003e \u003cb\u003ePreparation of IrO\u003c/b\u003e \u003csub\u003e \u003cb\u003ex\u003c/b\u003e \u003c/sub\u003e \u003cb\u003e/GC disk electrodes.\u003c/b\u003e The IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e/GC disk electrode was prepared by the colloidal electrodeposition of IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e on the GC disk of polished GC-Pt RRDE following the previous report\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. First, a metastable IrO\u003csub\u003ex\u003c/sub\u003e colloid suspension was obtained from an alkaline hydrolysis of Ir(III). 28.4 mg of Na\u003csub\u003e3\u003c/sub\u003eIrCl\u003csub\u003e6\u003c/sub\u003e\u0026middot;\u003cem\u003ex\u003c/em\u003eH\u003csub\u003e2\u003c/sub\u003eO was mixed with 30 mL of 0.1 M NaOH (2 mM of nominal Ir). The pale-yellow solution was moved into jacketed glassware and heated using a water circulator set to 85\u0026deg;C (solution temperature: 75\u0026deg;C) for 20 minutes at 300 rpm of stirring. After a few minutes, the color of solution turned into transparent grey, light violet, and blue, sequentially. The colloidal solution was moved into a 50 mL beaker placed in an ice bath. 70% HClO\u003csub\u003e4\u003c/sub\u003e was added to acidify the solution to pH\u0026thinsp;~\u0026thinsp;1.5 under mild stirring, changing the solution color to dark purple. Subsequently, a few drops of 10.1 M NaOH solution were added to make the solution alkaline (pH\u0026thinsp;~\u0026thinsp;12.5), which shifted the solution color to violet. The solution was stored for up to 3 months at 4\u0026deg;C.\u003c/p\u003e \u003cp\u003eAfter mechanical polishing, RRDE was rinsed and sonicated in DI water for 1 minute. For electrodeposition, the RRDE was vertically immersed into the IrO\u003csub\u003e\u003cspan type=\"ItalicUnderline\" class=\"ItalicUnderline\" name=\"Emphasis\"\u003ex\u003c/span\u003e\u003c/sub\u003e colloidal solution in a single-compartment 5 mL borosilicate glass beaker, where a ring-shaped Pt wire annealed by hydrogen flame (r\u0026thinsp;=\u0026thinsp;5 mm) was placed as a counter electrode. A 4 mL aliquot of the alkaline solution was acidified by adding 70% HClO\u003csub\u003e4\u003c/sub\u003e down to pH\u0026thinsp;~\u0026thinsp;1.55. Care must be taken not to make the IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e solution too acidic, which promotes OER and CER rates and lowers the electrodeposition rate. On the contrary, too high pH led to an excessive deposition of rough IrO\u003csub\u003ex\u003c/sub\u003e on GC disk and spontaneous deposition on the Pt ring. During electrodeposition, RRDE was rotated at 600 rpm without gas purging. To check the potential where IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e deposition starts, RRDE was scanned at 0.25 V s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e in the range of 0.16 V\u0026thinsp;\u0026minus;\u0026thinsp;1.36 V\u003csub\u003eAg/AgCl\u003c/sub\u003e (Supplementary Fig. \u003cb\u003eS20a\u003c/b\u003e). Depending on the pH of IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e solution, chronoamperometry (CA) at 1.25\u0026thinsp;\u0026plusmn;\u0026thinsp;0.01 V\u003csub\u003eAg/AgCl\u003c/sub\u003e for ~\u0026thinsp;300\u0026thinsp;\u0026plusmn;\u0026thinsp;30 s was conducted at around 20 mV positive potential at the onset of mixed CER and OER\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e (~\u0026thinsp;0.8 mA, Supplementary Fig. \u003cb\u003eS20b\u003c/b\u003e). After electrodeposition, a reflective bluish-violet film was formed on the GC disk. The electrodeposited RRDE was washed with a copious amount of DI water and then heated in a 75\u0026deg;C oven for 6 h (ramping rate\u0026thinsp;=\u0026thinsp;0.45\u0026deg;C∙min\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e). Sudden temperature increase could exfoliate the IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e film from the GC substrate. The dehydration of IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e increases the stability of IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e film on GC substrate, reducing the amount of electrochemically dissolved Ir\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e and the physical exfoliation by Cl\u003csub\u003e2\u003c/sub\u003e and O\u003csub\u003e2\u003c/sub\u003e bubble evolutions during RRDE measurements. After dehydration, the color of IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e/GC film changed to a reflective black.\u003c/p\u003e \u003cp\u003e \u003cb\u003eElectrochemical measurements using a rotating ring-disk electrode.\u003c/b\u003e For all RRDE measurements, the dual-channel potentiostat was connected by the mode of working electrode to ground. The distance between the reference electrode and RRDE was kept constant at around 2 cm to minimize the bubble interruption and the change of solution resistance. Before RRDE experiments, both IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e/GC disk and Pt ring electrodes were electrochemically cleaned in a blank 0.1 M H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e solution (pH 0.90\u0026thinsp;\u0026plusmn;\u0026thinsp;0.03) at 3600 rpm. The IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e deposited on the Pt ring was removed within potential cycling in the range of 1.3\u0026ndash;2.2 V\u003csub\u003eRHE\u003c/sub\u003e at 0.5 V s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e until decreasing OER activity was stabilized (Supplementary Fig. \u003cb\u003eS21a\u003c/b\u003e). Additionally, 40 potential cycling was conducted to electropolish the Pt surface in the range of \u0026minus;\u0026thinsp;0.1\u0026ndash;1.7 V\u003csub\u003eRHE\u003c/sub\u003e at 0.5 V∙s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (Supplementary Fig. \u003cb\u003eS21b\u003c/b\u003e). The CV of electrochemically cleaned Pt ring exhibited the characteristic peaks of poly Pt without IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e peak in the range of 0.05\u0026ndash;1.4 V\u003csub\u003eRHE\u003c/sub\u003e at 0.05 V s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (Supplementary Fig. \u003cb\u003eS21c\u003c/b\u003e). The IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e/GC disk electrode was also electropolished by 20-potential cycling between 1.3\u0026ndash;1.6 V\u003csub\u003eRHE\u003c/sub\u003e at 0.5 V∙s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (into the OER region). The electrochemical cleaning did not significantly change the IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e surface (Supplementary Fig. \u003cb\u003eS21d\u003c/b\u003e).\u003c/p\u003e \u003cp\u003eThe separation of CER and OER current densities using RRDE was conducted following the methods of Vos et al.\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e A sufficient rpm of RRDE (\u003cem\u003eω\u003c/em\u003e; rad∙s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) generates the laminar flow from the inner disk to the outer ring, allowing the Pt ring to reduce the generated Cl\u003csub\u003e2\u003c/sub\u003e from the IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e/GC disk back to Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e (Supplementary Fig. \u003cb\u003eS1\u003c/b\u003e). The spontaneous hydrolysis of Cl\u003csub\u003e2\u003c/sub\u003e into ClO\u003csup\u003e\u0026minus;\u003c/sup\u003e can be suppressed by the sufficiently low pH 0.9. Since poly Pt cannot reduce O\u003csub\u003e2\u003c/sub\u003e at 0.95 V\u003csub\u003eRHE\u003c/sub\u003e, the ring detects Cl\u003csub\u003e2\u003c/sub\u003e during a catalyst operation. Disk current density (\u003cem\u003ej\u003c/em\u003e\u003csub\u003edisk\u003c/sub\u003e) and ring current density (\u003cem\u003ej\u003c/em\u003e\u003csub\u003ering\u003c/sub\u003e) can be respectively converted into \u003cem\u003ej\u003c/em\u003e\u003csub\u003eCER\u003c/sub\u003e and \u003cem\u003ej\u003c/em\u003e\u003csub\u003eOER\u003c/sub\u003e via\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:{j}_{\\text{C}\\text{E}\\text{R}}=\\left|\\frac{{j}_{\\text{r}\\text{i}\\text{n}\\text{g}}}{{N}_{\\text{c}\\text{o}\\text{l}}}\\right|$$\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Equb\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$\\:{j}_{\\text{O}\\text{E}\\text{R}}=\\:{j}_{\\text{d}\\text{i}\\text{s}\\text{k}}-\\:{j}_{\\text{C}\\text{E}\\text{R}}=\\:{j}_{\\text{d}\\text{i}\\text{s}\\text{k}}-\\left|\\frac{{j}_{\\text{r}\\text{i}\\text{n}\\text{g}}}{{N}_{\\text{c}\\text{o}\\text{l}}}\\right|$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eThe collection efficiency (\u003cem\u003eN\u003c/em\u003e\u003csub\u003ecol\u003c/sub\u003e) of the blank GC-Pt RRDE was determined by measuring the [Fe(CN)\u003csub\u003e6\u003c/sub\u003e]\u003csup\u003e3\u0026minus;\u003c/sup\u003e/[Fe(CN)\u003csub\u003e6\u003c/sub\u003e]\u003csup\u003e4\u0026minus;\u003c/sup\u003e redox couple using chronoamperometric RRDE measurements in a solution of 10 mM K\u003csub\u003e3\u003c/sub\u003eFe[CN]\u003csub\u003e6\u003c/sub\u003e supplemented with 0.1 M MCl in the range of 1,225\u0026thinsp;\u0026minus;\u0026thinsp;3,600 rpm (Supplementary Fig. \u003cb\u003eS22\u003c/b\u003e). To measure the background current, 0.5 V\u003csub\u003eAg/AgCl\u003c/sub\u003e were respectively applied on GC disk and Pt ring for 60 s. Subsequently, CA was recorded at the disk potential of \u0026minus;\u0026thinsp;0.3 V\u003csub\u003eAg/AgCl\u003c/sub\u003e and the ring potential of 0.5 V\u003csub\u003eAg/AgCl\u003c/sub\u003e at each electrode rpm for 60 s. After subtracting the background current, \u003cem\u003eN\u003c/em\u003e\u003csub\u003ecol\u003c/sub\u003e was measured for each cation and electrode rpm. All measured \u003cem\u003eN\u003c/em\u003e\u003csub\u003ecol\u003c/sub\u003e exhibited 0.5% of standard deviation from the theoretical \u003cem\u003eN\u003c/em\u003e\u003csub\u003ecol\u003c/sub\u003e (0.474) given by the manufacturer. For the data measured in HCl (absence of alkali metal cations), \u003cem\u003eN\u003c/em\u003e\u003csub\u003ecol\u003c/sub\u003e measured in KCl was used. Since the IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e/GC disk underwent highly oxidative potential, \u003cem\u003ej\u003c/em\u003e\u003csub\u003ering\u003c/sub\u003e may include a dissolution current of IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e. However, \u003cem\u003ej\u003c/em\u003e\u003csub\u003ering\u003c/sub\u003e above \u0026minus;\u0026thinsp;1 mA∙cm\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e was measured in the H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e electrolyte supplemented with 1 mM of each [Ir\u003csup\u003eIII\u003c/sup\u003eCl\u003csub\u003e6\u003c/sub\u003e]\u003csup\u003e3\u0026minus;\u003c/sup\u003e and [Ir\u003csup\u003eIV\u003c/sup\u003eCl\u003csub\u003e6\u003c/sub\u003e]\u003csup\u003e2\u0026minus;\u003c/sup\u003e (pH 0.9; Supplementary Fig. \u003cb\u003eS23\u003c/b\u003e), of which is excessive compared to the amount of dissolved Ir species during the OER at 1.65 V\u003csub\u003eRHE\u003c/sub\u003e (no iR-correction)\u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e. The value of \u003cem\u003ej\u003c/em\u003e\u003csub\u003ering\u003c/sub\u003e was only 5% of the lowest \u003cem\u003ej\u003c/em\u003e\u003csub\u003ering\u003c/sub\u003e measured in the electrolyte containing 50 mM LiCl (~\u0026thinsp;20 mA∙cm\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e; Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). Thus, the dissolution current was not considered for calculating \u003cem\u003ej\u003c/em\u003e\u003csub\u003eCER\u003c/sub\u003e, \u003cem\u003ej\u003c/em\u003e\u003csub\u003eOER\u003c/sub\u003e, and CER selectivity. The electrocatalytic selectivity of CER can be calculated from\u003cdiv id=\"Equc\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equc\" name=\"EquationSource\"\u003e\n$$\\:\\text{C}\\text{E}\\text{R}\\:\\text{s}\\text{e}\\text{l}\\text{e}\\text{c}\\text{t}\\text{i}\\text{v}\\text{i}\\text{t}\\text{y}=\\:1-\\left(\\text{O}\\text{E}\\text{R}\\:\\text{s}\\text{e}\\text{l}\\text{e}\\text{c}\\text{t}\\text{i}\\text{v}\\text{i}\\text{t}\\text{y}\\right)=\\frac{0.5\\bullet\\:{j}_{\\text{C}\\text{E}\\text{R}}}{0.5\\bullet\\:{j}_{\\text{C}\\text{E}\\text{R}}+0.25{\\bullet\\:j}_{\\text{O}\\text{E}\\text{R}}}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eBefore RRDE measurements for OER and CER, IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e/GC and Pt ring were preconditioned by 20 scans in the range of 1.2\u0026ndash;2.2 V at 0.1 V∙s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (into the OER/CER region), while the Pt ring was kept at 0.95 V to detect Cl\u003csub\u003e2\u003c/sub\u003e generated by the IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e/GC disk electrode in MCl-containing electrolytes. The current densities in individual CVs of both disk and ring did not significantly change after 15 scans. This step stabilized the IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e/GC and its OER/CER response to avoid a gradual current change during the measurements. The IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e/GC disk electrode was scanned in the range of 1.2\u0026ndash;2.2 V at 0.01 V∙s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, while the Pt ring was kept at 0.95 V\u003csub\u003eRHE\u003c/sub\u003e. Each disk and ring current were corrected with the background currents measured at 1.21 V\u003csub\u003eRHE\u003c/sub\u003e, which was below the \u003cem\u003eE\u003c/em\u003e\u003csup\u003e0\u003c/sup\u003e of OER/CER and Ir(IV/V) redox in acidic media\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. After CER and OER measurements, reduction in current densities in CV (0.3\u0026ndash;1.5 V\u003csub\u003eRHE\u003c/sub\u003e, 0.01 V∙s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e; Supplementary Fig. \u003cb\u003eS24\u003c/b\u003e) indicated that the electrochemically active surface area of IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e decreased by electrochemical Ir dissolution during CER and OER at the high current densities. The CAs were measured after electrochemical cleaning, identical to the CV measurements. The applied potentials were fixed at 2.1 V\u003csub\u003eRHE\u003c/sub\u003e (no iR-correction), where the diffusion limit was selected based on CVs. Each CA was done in 15 s and consecutively measured in different rpm and ion concentrations without solution exchange.\u003c/p\u003e \u003cp\u003e \u003cb\u003eCharacterization of Ir oxides.\u003c/b\u003e Scanning electron microscopy (SEM; Quattro S, Thermo Fisher Scientific) was operated at an accelerating voltage of 30 kV to take the SEM images and energy dispersive X-ray (EDX) spectra of the IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e surface. Before measurements, all samples were washed with copious DI water and dried in the fume hood overnight. All SEM images and EDX spectra were obtained at 30 keV. The SEM images of the electrodeposited IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e/GC show a partially cracked surface covered by IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e NPs with 0.2\u0026thinsp;\u0026minus;\u0026thinsp;1 \u0026micro;m diameters (Supplementary Fig. \u003cb\u003eS25a,b\u003c/b\u003e). The SEM images of the IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e surface after electrochemical cleaning by 20 potential cycling in 0.3\u0026ndash;1.6 V\u003csub\u003eRHE\u003c/sub\u003e also exhibited the surface cracks and IrO\u003csub\u003ex\u003c/sub\u003e particles, indicating the electrochemical cleaning did not significantly change the surface morphology (Supplementary Fig. \u003cb\u003eS25c,d\u003c/b\u003e). After CER and OER measurements in the electrolyte supplemented with 50 mM of LiCl and CsCl, the IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e surface became roughened by enlarged cracks (Supplementary Fig. \u003cb\u003eS25e\u0026ndash;h\u003c/b\u003e), presumably due to intensive Cl\u003csub\u003e2\u003c/sub\u003e/O\u003csub\u003e2\u003c/sub\u003e gas evolution. However, crack formation did not increase the electrochemically active surface area, as shown in the decreased current densities in CV after measurements (Supplementary Fig. \u003cb\u003eS24\u003c/b\u003e). The EDX spectra were measured before and after RRDE measurements (Supplementary Fig. \u003cb\u003eS26\u003c/b\u003e), revealing that Cs\u003csup\u003e+\u003c/sup\u003e was not detected on IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e/GC surface after RRDE measurements. Li\u003csup\u003e+\u003c/sup\u003e might not be detectable in EDX spectra. The results are consistent with the previous report that the incorporation of M\u003csup\u003e+\u003c/sup\u003e into the IrO\u003csub\u003e\u003cem\u003ex\u003c/em\u003e\u003c/sub\u003e lattice was done by calcining the mixture solution of IrCl\u003csub\u003e3\u003c/sub\u003e and alkali metal hydroxides precursors at 400\u0026deg;C [49].\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor contributions\u003c/h2\u003e \u003cp\u003eT.L., H.O., and R.N. conceived the project and wrote the paper. T.L. conducted material synthesis, characterization, and electrochemical analyses. H.O. simulated the double diffusion model. All authors approved the final version of the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgements\u003c/h2\u003e \u003cp\u003eThis work was supported by GteX Program Japan (Grant Number: JPMJGX23H2) and MEXT Program: Data Creation and Utilization-Type Material Research and Development Project (Grant Number: JPMXP1122712807). T.L. received financial support from RIKEN Special Postdoctoral Researchers (SPDR) fellowship and RIKEN Incentive Research Projects (IRP). We appreciate Prof. A. Yamaguchi of Institute of Science Tokyo and Prof. K. Sakaushi of National Institute for Materials Science for experimental support on electrochemical measurements. We are grateful to Dr. A. Li and Dr. A. Koishi of RIKEN Center for Sustainable Resource Science and Dr. K. Yatsuzuka of University of Tokyo for the discussion on material synthesis, electrochemistry, and water network. We acknowledge Materials Characterization Support Team in RIKEN Center for Emergent Matter Science for technical support on SEM measurements.\u003c/p\u003e\n\u003ch3\u003eData availability\u003c/h3\u003e\n\u003cp\u003eAll relevant data generated and analysed during this study are included in this article and its Supplementary Information. Data for the main and supplementary figures are available in Figshare.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eGraves, C., Ebbesen, S. D., Mogensen, M. \u0026amp; Lackner, K. S. Sustainable hydrocarbon fuels by recycling CO\u003csub\u003e2\u003c/sub\u003e and H\u003csub\u003e2\u003c/sub\u003eO with renewable or nuclear energy. Renew. Sustain. Energy Rev. 15, 1\u0026ndash;23 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIRENA. Green hydrogen cost reduction: scaling up electrolysers to meet the 1.5⁰C climate goal. International Renewable Energy Agency. (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMayyas, A., Ruth, M., Pivovar, B., Bender, G. \u0026amp; Wipke, K. Manufacturing cost analysis for proton exchange membrane water electrolyzers. National Renewable Energy Laboratory. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.nrel.gov/docs/fy10osti/72740.pdf\u003c/span\u003e\u003cspan address=\"https://www.nrel.gov/docs/fy10osti/72740.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEke, J., Yusuf, A., Giwa, A. \u0026amp; Sodiq, A. The global status of desalination: an assessment of current desalination technologies, plants and capacity. Desalination 495, 114633 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBecker, H. \u003cem\u003eet al.\u003c/em\u003e Impact of impurities on water electrolysis: a review. Sustain. Energy Fuels 7, 1565\u0026ndash;1603 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDuPont. FilmTec\u0026trade; reverse osmosis membranes technical manual. DuPont. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.dupont.com/content/dam/dupont/amer/us/en/water-solutions/public/documents/en/RO-NF-FilmTec-Manual-45-D01504-en.pdf\u003c/span\u003e\u003cspan address=\"https://www.dupont.com/content/dam/dupont/amer/us/en/water-solutions/public/documents/en/RO-NF-FilmTec-Manual-45-D01504-en.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLim, Y. J., Goh, K., Kurihara, M. \u0026amp; Wang, R. Seawater desalination by reverse osmosis: Current development and future challenges in membrane fabrication \u0026ndash; a review. J. Membr. Sci. 629, 119292 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, L. \u003cem\u003eet al.\u003c/em\u003e Salt and water transport in reverse osmosis membranes: beyond the solution-diffusion model. Environ. Sci. Technol. 55, 16665\u0026ndash;16675 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTrasatti, S. Electrocatalysis in the anodic evolution of oxygen and chlorine. Electrochim. Acta 29, 1503\u0026ndash;1512 (1984).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKarlsson, R. K. B. \u0026amp; Cornell, A. Selectivity between oxygen and chlorine evolution in the chlor-alkali and chlorate processes. Chem. Rev. 116, 2982\u0026ndash;3028 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eExner, K. S. Controlling stability and selectivity in the competing chlorine and oxygen evolution reaction over transition metal oxide electrodes. ChemElectroChem 6, 3401\u0026ndash;3409 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVos, J. G., Wezendonk, T. A., Jeremiasse, A. W. \u0026amp; Koper, M. T. M. MnO\u003csub\u003ex\u003c/sub\u003e/IrO\u003csub\u003ex\u003c/sub\u003e as selective oxygen evolution electrocatalyst in acidic chloride solution. J. Am. Chem. Soc. 140, 10270\u0026ndash;10281 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVos, J. G. \u003cem\u003eet al.\u003c/em\u003e Selectivity trends between oxygen evolution and chlorine evolution on iridium-based double perovskites in acidic media. ACS Catal. 9, 8561\u0026ndash;8574 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIzumiya, K. \u003cem\u003eet al.\u003c/em\u003e Anodically deposited manganese oxide and manganese\u0026ndash;tungsten oxide electrodes for oxygen evolution from seawater. Electrochim. Acta 43, 3303\u0026ndash;3312 (1998).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStrmcnik, D. \u003cem\u003eet al.\u003c/em\u003e The role of non-covalent interactions in electrocatalytic fuel-cell reactions on platinum. Nat. Chem. 1, 466\u0026ndash;472 (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eColic, V., Pohl, M. D., Scieszka, D. \u0026amp; Bandarenka, A. S. Influence of the electrolyte composition on the activity and selectivity of electrocatalytic centers. Catal. Today 262, 24\u0026ndash;35 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSingh, M. R., Kwon, Y., Lum, Y., Ager, J. W., III \u0026amp; Bell, A. T. Hydrolysis of electrolyte cations enhances the electrochemical reduction of CO\u003csub\u003e2\u003c/sub\u003e over Ag and Cu. J. Am. Chem. Soc. 138, 13006\u0026ndash;13012 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWintrich, D. \u003cem\u003eet al.\u003c/em\u003e Enhancing the selectivity between oxygen and chlorine towards chlorine during the anodic chlorine evolution reaction on a dimensionally stable anode. ChemElectroChem 6, 3108\u0026ndash;3112 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRao, R. R. \u003cem\u003eet al.\u003c/em\u003e pH- and cation-dependent water oxidation on rutile RuO\u003csub\u003e2\u003c/sub\u003e(110). J. Phys. Chem. C 125, 8195\u0026ndash;8207 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOvalle, V. J., Hsu, Y.-S., Agrawal, N., Janik, M. J. \u0026amp; Waegele, M. M. Correlating hydration free energy and specific adsorption of alkali metal cations during CO\u003csub\u003e2\u003c/sub\u003e electroreduction on Au. Nat. Catal. 5, 624\u0026ndash;632 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGu, J. \u003cem\u003eet al.\u003c/em\u003e Modulating electric field distribution by alkali cations for CO\u003csub\u003e2\u003c/sub\u003e electroreduction in strongly acidic medium. Nat. Catal. 5, 268\u0026ndash;276 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, X. Y. \u003cem\u003eet al.\u003c/em\u003e Mechanism of cations suppressing proton diffusion kinetics for electrocatalysis. Angew. Chem. Int. Ed. 62, e202218669 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCostentin, C. \u0026amp; Saveant, J. M. Catalysis at the nanoscale may change selectivity. \u003cem\u003eProc. Natl. Acad. Sci. U.S.A.\u003c/em\u003e 113, 11756\u0026ndash;11758 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGough, D. A. \u0026amp; Leypoldt, J. K. Membrane-covered, rotated disk electrode. Anal. Chem. 51, 439\u0026ndash;444 (1979).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBard, A. J., Faulkner, L. R. \u0026amp; White, H. S. \u003cem\u003eElectrochemical methods: fundamentals and applications\u003c/em\u003e. 3rd ed, (John Wiley \u0026amp; Sons, 2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJenkins, H. D. B. \u0026amp; Marcus, Y. Viscosity B-coefficients of ions in solution. Chem. Rev. 95, 2695\u0026ndash;2724 (1995).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShinagawa, T., Garcia-Esparza, A. T. \u0026amp; Takanabe, K. Insight on Tafel slopes from a microkinetic analysis of aqueous electrocatalysis for energy conversion. Sci. Rep. 5, 13801 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNong, H. N. \u003cem\u003eet al.\u003c/em\u003e Key role of chemistry versus bias in electrocatalytic oxygen evolution. Nature 587, 408\u0026ndash;413 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKouteck\u0026yacute;, J. \u0026amp; Levich, V. G. The use of a rotating disk electrode in the studies of electrochemical kinetics and electrolytic processes. Zh. Fiz. Khim. 32, 1565\u0026ndash;1575 (1958).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWatanabe, M., Igarashi, H. \u0026amp; Yosioka, K. An experimental prediction of the preparation condition of Nafion-coated catalyst layers for PEFCs. Electrochim. Acta 40, 239\u0026ndash;334 (1995).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchmidt, T. J. \u003cem\u003eet al.\u003c/em\u003e Characterization of high-surface-area electrocatalysts using a rotating disk electrode configuration. J. Electrochem. Soc. 145, 2354\u0026ndash;2358 (1998).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKnoche, K. L., Hettige, C., Zook-Gerdau, L. A. \u0026amp; Leddy, J. Film permeation by rotating disk voltammetry at electrodes modified with electrochemically inert layers and heterogeneous composites. J. Electrochem. Soc. 163, H3173\u0026ndash;H3181 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRard, J. A. \u0026amp; Miller, D. G. Mutual diffusion coefficients of Nal-H\u003csub\u003e2\u003c/sub\u003eO and LiCl-H\u003csub\u003e2\u003c/sub\u003eO at 25\u0026deg;C from Rayleigh interferometry. J. Sol. Chem. 12, 413\u0026ndash;425 (1983).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGebbie, M. A., Liu, B., Guo, W., Anderson, S. R. \u0026amp; Johnstone, S. G. Linking electric double Layer formation to electrocatalytic activity. ACS Catal. 13, 16222\u0026ndash;16239 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGarlyyev, B., Xue, S., Watzele, S., Scieszka, D. \u0026amp; Bandarenka, A. S. Influence of the nature of the alkali metal cations on the electrical double-layer capacitance of model Pt(111) and Au(111) electrodes. J. Phys. Chem. Lett. 9, 1927\u0026ndash;1930 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBohra, D., Chaudhry, J. H., Burdyny, T., Pidko, E. A. \u0026amp; Smith, W. A. Modeling the electrical double layer to understand the reaction environment in a CO\u003csub\u003e2\u003c/sub\u003e electrocatalytic system. Energy Environ. Sci. 12, 3380\u0026ndash;3389 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang, J., Li, M., Eslamibidgoli, M. J., Eikerling, M. \u0026amp; Gross, A. Cation overcrowding effect on the oxygen evolution reaction. JACS Au 1, 1752\u0026ndash;1765 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee, S. S., Koishi, A., Bourg, I. C. \u0026amp; Fenter, P. Ion correlations drive charge overscreening and heterogeneous nucleation at solid\u0026ndash;aqueous electrolyte interfaces. \u003cem\u003eProc. Natl. Acad. Sci. U.S.A.\u003c/em\u003e 118, e2105154118 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOjha, K., Doblhoff-Dier, K. \u0026amp; Koper, M. T. M. Double-layer structure of the Pt(111)-aqueous electrolyte interface. \u003cem\u003eProc. Natl. Acad. Sci. U.S.A.\u003c/em\u003e 119, e2116016119 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCameron, A. P., Davey, S. B., Callahan, C. L. \u0026amp; Donne, S. W. Communication\u0026mdash;demonstrating the role of mass transport in double layer formation. J. Electrochem. Soc. 169 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAssaf, K. I. \u0026amp; Nau, W. M. The chaotropic effect as an assembly motif in chemistry. Angew. Chem. Int. Ed. 57, 13968\u0026ndash;13981 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarcus, Y. Viscosity B-coefficients, structural entropies and heat capacities, and the effects of ions on the structure of water. J. Solution Chem. 23, 831\u0026ndash;848 (1994).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRosenfeld, Y. Relation between the transport coefficients and the internal entropy of simple systems. Phys. Rev. A 15, 2545\u0026ndash;2549 (1977).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChopra, R., Truskett, T. M. \u0026amp; Errington, J. R. On the use of excess entropy scaling to describe the dynamic properties of water. J. Phys. Chem. B 114, 10558\u0026ndash;10566 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJones, G. \u0026amp; Dole, M. The viscosity of aqueous solutions of strong electrolytes with special reference to barium chloride. J. Am. Chem. Soc. 51, 2950\u0026ndash;2964 (1929).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBreslau, B. R. \u0026amp; Miller, I. F. On the viscosity of concentrated aqueous electrolyte solution. J. Phys. Chem. 74, 1056\u0026ndash;1061 (1970).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGeiger, S. \u003cem\u003eet al.\u003c/em\u003e Activity and stability of electrochemically and thermally treated iridium for the oxygen evolution reaction. J. Electrochem. Soc. 163, F3132\u0026ndash;F3138 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGeiger, S. \u003cem\u003eet al.\u003c/em\u003e The stability number as a metric for electrocatalyst stability benchmarking. Nat. Catal. 1, 508\u0026ndash;515 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGao, J. \u003cem\u003eet al.\u003c/em\u003e Breaking long-range order in iridium oxide by alkali ion for efficient water oxidation. J. Am. Chem. Soc. 141, 3014\u0026ndash;3023 (2019).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDiffusion parameters in the numerical simulation based on the double diffusion model for 50 mM MCl at 3,600 rpm.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCation Species\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003eα\u0026thinsp;=\u0026thinsp;L\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e\u003cem\u003e/D\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e (s∙cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003eβ\u0026thinsp;=\u0026thinsp;δ/D\u003c/em\u003e\u003csub\u003econv\u003c/sub\u003e (s∙cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003eδ\u003c/em\u003e (\u0026micro;m)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cem\u003eD\u003c/em\u003e\u003csub\u003econv\u003c/sub\u003e \u003c/p\u003e \u003cp\u003e(\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;5\u003c/sup\u003e cm\u003csup\u003e2\u003c/sup\u003e∙s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cem\u003eD\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e for \u003cem\u003eL\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e = 1\u0026minus;10 nm\u003c/p\u003e \u003cp\u003e(\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;5\u003c/sup\u003e cm\u003csup\u003e2\u003c/sup\u003e∙s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLi\u003csup\u003e+\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e45.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e117.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e6.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.596\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e0.00022\u0026thinsp;\u0026minus;\u0026thinsp;0.00218\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNa\u003csup\u003e+\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e33.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e96.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e7.73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.805\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e0.00030\u0026thinsp;\u0026minus;\u0026thinsp;0.00299\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eH\u003csup\u003e+\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e27.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e93.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e7.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.842\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e0.00037\u0026thinsp;\u0026minus;\u0026thinsp;0.00365\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eK\u003csup\u003e+\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e15.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e100.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e7.57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.755\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e0.00066\u0026thinsp;\u0026minus;\u0026thinsp;0.00663\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCs\u003csup\u003e+\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e14.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e90.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e7.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.883\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c6\"\u003e \u003cp\u003e0.00072\u0026thinsp;\u0026minus;\u0026thinsp;0.00715\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5786052/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5786052/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eImpurity ions pose a major challenge towards diversifying water usage for electrolysis. In particular, millimolar-level chloride impurities remaining in reverse osmosis filtrates significantly diminish the selectivity and longevity of water electrolyzers. Here, we show that alkali metal cations can regulate the diffusion coefficient of chloride ions, enabling suppression of chlorine evolution during water electrolysis under diffusion-limiting conditions. Evidence of the cation dependency is provided by positive intercepts in both Levich and modified Kouteck\u0026yacute;\u0026minus;Levich plots using a rotating ring disk electrode, indicating the presence of an additional, cation-dependent diffusion layer that suppresses chloride diffusion. Numerical simulations based on the double diffusion model quantify this effect, resulting in a linear correlation between the cation-dependent diffusion barrier and the structural entropy of cation hydration. These findings suggest that the cation-dependent structuring of water significantly influences mass transport, which is particularly important at practical current densities where impurity ions are diffusion-limited.\u003c/p\u003e","manuscriptTitle":"Cation-controlled diffusion of chloride ions during electrochemical chlorine evolution in acidic media","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-01-17 05:10:04","doi":"10.21203/rs.3.rs-5786052/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-chemistry","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"nchem","sideBox":"Learn more about [Nature Chemistry](http://www.nature.com/nchem/)","snPcode":"","submissionUrl":"","title":"Nature Chemistry","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Research","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"66b62ffe-efce-4bb0-96c5-732572aed5dc","owner":[],"postedDate":"January 17th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":42924357,"name":"Physical sciences/Chemistry/Electrochemistry/Electrocatalysis"},{"id":42924358,"name":"Physical sciences/Chemistry/Physical chemistry/Chemical physics"}],"tags":[],"updatedAt":"2025-12-09T08:06:14+00:00","versionOfRecord":{"articleIdentity":"rs-5786052","link":"https://doi.org/10.1038/s41557-025-02014-4","journal":{"identity":"nature-chemistry","isVorOnly":false,"title":"Nature Chemistry"},"publishedOn":"2025-12-08 05:00:00","publishedOnDateReadable":"December 8th, 2025"},"versionCreatedAt":"2025-01-17 05:10:04","video":"","vorDoi":"10.1038/s41557-025-02014-4","vorDoiUrl":"https://doi.org/10.1038/s41557-025-02014-4","workflowStages":[]},"version":"v1","identity":"rs-5786052","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5786052","identity":"rs-5786052","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}