Peaks and pitfalls of electrocatalytic descriptor models at the example of CO2 reduction

Peaks and pitfalls of electrocatalytic descriptor models at the example of CO2 reduction | 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 Analysis Peaks and pitfalls of electrocatalytic descriptor models at the example of CO2 reduction Jihun Oh, Beomil Kim, Seungchang Han, Suneon Wang, Stefan Ringe This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5559232/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 13 Apr, 2026 Read the published version in Nature Catalysis → Version 1 posted You are reading this latest preprint version Abstract Electrocatalysis advances rely on the development of efficient catalysts. Systematic material design hinges on identifying activity and selectivity descriptors. While adsorption energy descriptors have helped predict new materials, they are typically based on pure metals, uncertain of their applicability to complex materials like alloys. Here, we systematically analyze the validity of descriptor models for the electrochemical reduction of CO 2 (CO 2 RR). For this, we prepare gold, silver, and palladium alloys of variable composition and confirm experimentally the continuous variation of the d-band center (i.e. the CO adsorption energy) and work function (i.e. the potential of zero charge). Our results indicate that while the d-band center is the decisive factor for CO production, it, along with the work function, fails to fully explain the production of HCOO − and H 2 . Designing a copper-like alloy based on the matching of these descriptor values showed no formation of C 2 products (as commonly expected for copper). This breakdown of the descriptor model is explained from first-principles calculations by the heterogeneity of the surface leading to different deactivation pathways for C 2 product formation. Our results highlight the problems in transferring conventional descriptor models to more complex, heterogeneous materials motivating future developments. Physical sciences/Chemistry/Electrochemistry/Electrocatalysis Physical sciences/Materials science/Materials for energy and catalysis/Electrocatalysis Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Electrochemical CO 2 reduction reaction (CO 2 RR) has attracted intense interest in the context of carbon capture and utilization (CCU) technology. Carbon monoxide (CO), hydrocarbons (e.g., CH 4 , C 2 H 4 , etc.), and alcohols (e.g., C 2 H 5 OH and CH 3 CH 2 CH 2 OH, etc.) can be formed from CO 2 RR along with H 2 from the competitive hydrogen evolution reaction (HER). 1-4 To reach competitiveness with established petrochemical processes, material design is key to enhancing the turnover and product selectivity of the reaction. Systematic material design requires the availability of performance descriptors, and first-principles calculations have been instrumental in deriving and rationalizing such. 5 Density functional theory (DFT)--based calculations have revealed that often a simple or two adsorption energy descriptors are enough to predict the full free energy diagram and thus the turnover frequency or selectivity of a catalyst. 6-9 The resulting activity volcanos can be used for high-throughput screening of materials, possibly facilitated by machine learning. 10,11 In the case of CO 2 RR, the binding affinity between carbon monoxide (CO) and a catalyst’s surface (i.e., the CO binding strength) has been suggested as a descriptor for metallic catalysts. 12 Weaker CO adsorption thereby facilitates the desorption of CO granting Au and Ag the CO-evolving CO 2 RR catalysts. Conversely, stronger adsorption such as to Ni 13 and Pt 14 , leads to poisoning/blocking of the active sites, preventing CO 2 RR and facilitating the competitive hydrogen evolution reaction (HER). This argument was used to explain why Cu with a moderate binding of CO is the only transition metal-based catalyst leading to significant amounts of higher reduced C 2+ products such as ethylene, ethanol, and 1-propanol. 15 For transition metal catalysts, the CO adsorption energy correlates well with the d-band center of the metal, 8,16-18 providing a physical rational for explaining observed trends. Recently, it has been, however, pointed out that in addition to the CO adsorption energy, the driving force of a catalyst to transfer charge to the adsorbate should be considered. 6 This force is related to the work function of the metal and correlates with the potential of zero charge (PZC) measured electrochemically. The charge-transfer descriptor emerges from the fact that during adsorption, charge transfers from the metal to the adsorbate leading to the formation of an electric dipole. This electric dipole interacts with the surrounding electric field from the electric double layer (EDL) that is formed at potentials deviating from the PZC. 6,19,20 In case of C 2+ production, this field-effect leads to a decrease in the kinetic barrier for CO dimer formation. 21 Due to the difference in the field dependence of different adsorbates, the PZC or any other charge-transfer related descriptor 22 must be considered as an independent descriptor and spans together with the CO adsorption energy a two-dimensional activity and selectivity volcano for CO 2 RR. 6,23-27 The current descriptor framework or CO 2 RR has been developed for single-element transition metal catalysts. It has remained unclear if they could be used for practical catalyst design of more complex materials, the ultimate purpose of any descriptor model. Complications arise in the measurements of the CO binding strength and PZC which are challenging and often limited in operando CO 2 RR environments. 8 Moreover, the performance and selectivity of CO 2 RR depend on various factors, such as the electrode material, the catalyst structure, the electrolyte composition, the applied potential, the local environment near the electrode, and electrolytic cell designs. 28 This makes it problematic to compare different published datasets and use them to evaluate the ability of descriptors. Here, we investigate the reliability of these descriptor models by studying the CO 2 RR activities of binary and ternary alloys of gold (Au), silver (Ag), and palladium (Pd) in a gas-fed flow cell. We measure the d-band center and work function which we consider as descriptors due to their close correlation with the reported CO binding strength and PZC descriptors, respectively. The Au, Ag, and Pd alloys with varying compositions have nearly identical morphology, crystallographic orientation and crystal structure, and a uniform composition without phase segregation, justifying our attempt to correlate CO 2 RR behavior with the measured descriptors. From this, we identify that the d-band center is the key descriptor for CO and HCOOH production with a contribution of work function for HCOOH, but fails to describe C 2+ product formation. We also synthesize AuAgPd alloys with the d-band center and work function matching those of Cu, but contrary to the expectations from Cu and the descriptor model, find no C 2+ products. From electric double layer-aware DFT calculations, we show this to be related to the heterogeneous bonding environments on the surface leading to various deactivation pathways for C 2+ product formation. Our results highlight the chances and challenges of applying descriptor models for designing electrocatalysts. Results Au-Ag-Pd alloys with tunable electronic structure We synthesized Au-Ag-Pd alloys with various compositions using a co-sputtering method that yields alloy films with uniform composition across large areas (Fig. 1 a). The metal alloys were deposited on a chemically inert, nano-fibrous PTFE membrane gas diffusion electrode (GDE) as a carbon-based GDE may produce hydrogen (H 2 ) at negative applied potential, 27 potentially confounding the CO 2 RR activities of catalysts (Fig. 1 b). The compositions of the as-deposited alloy films are measured using X-ray photoelectron spectroscopy (XPS) and scanning electron microscope-energy dispersive spectroscopy (SEM-EDS) (Supplementary Table 1). The alloys maintain metal ratio uniformly from near surface to bulk as can be seen by comparing two spectroscopy analyses. To investigate the morphology and spatial distribution of the chemical composition of the alloy films on PTFE membranes, electron microscopy was conducted. The SEM images indicate that the AuAg, AuPd, and AgPd binary alloy films were conformally deposited on the PTFE membranes, maintaining the fibrous nature of the GDE (the left panels of Fig. 1 c and Supplementary Fig. 1). In addition, all the alloy films exhibit similar morphology across the different compositions. The SEM-EDS analyses on each binary alloy composed of Au, Ag, and Pd indicate that all metals are uniformly distributed across the PTFE membrane without noticeable secondary phase precipitations. The cross-sectional transmission electron microscope-energy dispersive spectroscopy (TEM-EDS) images also confirm the formation of the Au 3 Ag 1 Pd 3 catalyst with uniform chemical composition throughout the film thickness (Fig. 1 d). The absence of the phase separation in our alloys is consistent with the complete miscibility in the phase diagrams of the binary alloys of Au, Ag, and Pd. 29 , 30 Through TEM-EDS and angle resolved X-ray photoelectron spectroscopy (ARXPS) analyses confirmed that the composition alteration in both the near-surface and bulk of the ternary alloys, during CO 2 RR, is less than 10 atomic percent (Supplementary Fig. 2, 3, 4). Furthermore, the X-ray diffraction (XRD) measurements indicate that all pure metals and alloys have the face-centered cubic (fcc) crystal structure with a dominant (111) out-of-plane texture with similar crystallite sizes (Supplementary Fig. 5, Supplementary Table 2). Note also that no second phases are seen in XRD. Therefore, our catalysts (co-)sputtered on PTFE membranes are an ideal platform to evaluate and compare the intrinsic CO 2 RR properties of the alloy catalysts because they are single-phase fcc crystals with similar morphology, crystallographic orientation, and uniform chemical composition over entire films. Figures 2 a and b show contour plots of the d-band centers and work function of pure metals (Au, Ag, and Pd) and their binary and ternary alloys, measured by ultraviolet photoemission spectroscopy (UPS) (see the measured values in Supplementary Table 3). The d-band centers of Au (−4.41 eV) and Ag (−5.41) are located significantly lower relative to the Fermi energy level compared to the one of Pd (−1.58 eV). When alloying Au with Pd or Ag with Pd, the resulting d-band center of alloys shifts to an intermediate position relative to that of pure Au and Pd or Ag and Pd, respectively (Fig. 2 a and Supplementary Fig. 6). For example, the d-band center moves toward to the Fermi energy level when Pd is introduced to Au or Ag alloys, whereas AuAg alloys maintain a rather similar d-band center position (Supplementary Table 3). Similarly, alloying alters the work function to a value that lies between those of the constituent metals. Figure 2 b shows that Au (5.17 eV) and Pd (5.12 eV) have similar work function values, higher than the value of Ag (4.46 eV). Hence, alloys of Au and Pd exhibit nearly the same work function of ~ 5.1 eV and the introduction of Ag into Au or Pd makes the work functions of AuAg and AgPd alloys lower than their bulk counterpart (Supplementary Table 3). These visual representations offer clear insights, enabling us to identify viable combinations of d-band center and work function achievable through alloying Au, Ag, and Pd. Therefore, a catalyst with the desired d-band center positions from −1.58 to −5.41 eV and work function from 4.46 to 5.17 eV can be fabricated from alloying two or three metals of Au, Ag, and Pd. Descriptor-dependent CO 2 RR properties of Au, Ag, and Pd alloys The co-sputtered alloys with varying Ag-Au-Pd ratio allow us to form a catalyst with independently controlled CO 2 RR descriptors. To evaluate the connection between the descriptors and CO 2 RR performance and selectivity, each alloy with a different d-band center and work function was subject to CO 2 electrolysis in a 1 M KHCO 3 electrolyte at three distinct potentials of −0.6, −0.7, and −0.8 V RHE in a custom-built flow cell. Electrochemical currents are normalized by the electrochemical surface area (ECSA) of each catalyst from the EDL capacitance (DLC) (Supplementary Table 4). CO, HCOO − , and H 2 , were the only major CO 2 RR products measured from pure and binary metals of Ag, Au, and Pd whereas Cu also produces hydrocarbons and alcohols (Supplementary Fig. 7–11). Figures 3 a-c show the contour plots of the partial current density for CO production normalized by ECSA ( j CO,ECSA ) as a function of the d-band center and work function of each alloy at different potentials from −0.6 to −0.8 V RHE . Shown in Supplementary Fig. 12 are the contour plots for j total,ECSA and j CO2RR,ECSA as a function of d-band center and work function. At −0.6 and −0.7 V RHE , the contour lines for CO production are curved, but nearly parallel to the work function, suggesting that the d-band center is the dominant descriptor (Fig. 3 a and 3 b). When the cathodic potential was increased to −0.8 V RHE , however, the contour lines become more bent, forming a ridge connecting Ag and Pd as an apex and base, respectively (Fig. 3 c). The scatter plots also reveals a strong correlation of the d-band center with j CO,ECSA , indicated by R-square value from 0.80 to 0.89 (Supplementary Fig. 13). Note that j CO,ECSA of the binary alloys lacks a linear correlation with the alloy composition (Supplementary Fig. 11). In contrast, the work function exhibits a much weaker correlation with j CO,ECSA than the d-band center does (Fig. 3 a-c). Scatter analysis indicates that the linear correlation of the logarithm of j CO,ECSA with the work function increases with applied potentials but remains limited to a maximum R-square value of 0.457 at −0.8 V RHE , (Supplementary Fig. 13). Therefore, the d-band center position (i.e., the CO binding strength) of the catalysts is the primary descriptor for CO 2 -to-CO activity at all potentials whereas the work function (i.e., PZC) of the catalysts can be designated as the secondary descriptor that weights in only at high potentials. Simply, in the material space that we looked at, a catalyst with weaker CO binding strength has a high chance of producing more CO from CO 2 RR. This stands in contrast to the volcano-like dependence on the CO adsorption energy proposed from theoretical works on bare transition metal surfaces. 6 , 31 The contour plots reveal that HCOO − production has a rather different dependence on the electronic structure parameters of the catalysts (Fig. 3 d-f). HCOO − production exhibits a volcano-like activity with a peak near Au 3 Ag 1 for all potentials. Firstly, j HCOO−,ECSA increases with the d-band center and the R-square values are increased as the potential increases (Supplementary Fig. 14). Secondly, the work function dependence stands out for weakly-binding metals with a d-band center lower than −3.5 eV. A clear volcano-like dependence on the work function is identified for these alloys (Supplementary Fig. 15a-c). This stands in contrast to recent work on bare transition metal surfaces suggesting a continuous switch of selectivity between CO and formate when tuning the work function. 6 Thirdly, HCOO − production and the d-band center have a linear relationship for strongly binding metals with a d-band center higher than −3.0 eV (Supplementary Fig. 15d-f). Note that the relationship between j HCOO − ,ECSA and the work function is not clearly visible when all catalysts are compared with the work function as a single parameter because the d-band center-dependence is lumped together in a simple scatter plot. These findings suggest that CO 2 -to-HCOO − activity depends on both the CO binding strength and PZC and the influence of the latter on HCOO − formation is limited to a catalyst with weak CO binding strength. We also investigated the H 2 evolution reaction (HER) behaviors of our catalysts during CO 2 RR (Supplementary Fig. 16–18). Contour plots show volcano-like HER activity with a peak near Au 3 Ag 1 (Supplementary Fig. 17). A close examination also shows that the contour lines are horizontal at low work function but change vertically as the work function increases (Supplementary Fig. 18). This behavior implies that in the low work function region, the proton-transfer rate depends significantly on the electric field across the EDL likely via water structuring 32 . Conversely, the d-band center becomes a prominent descriptor in the high work function region, likely due to the saturation of the water structure due to the more negatively charged electrode. 10 It must be noted that H 2 and HCOO − production display a similar dependence on the work functions of electrodes, implying that they share reaction intermediates. Indeed, it is suggested that HCOO is formed by a reaction of CO 2 with hydrogen adsorbed on a surface. 33 – 3 6 As a further test for the descriptor model, we designed an alloy with a d-band center and work function comparable to that of Cu. It is well-known that Cu is the only metal that produces C 2+ hydrocarbons and alcohols with significant FE. 12 , 37 Previous descriptor models have explained this behavior by the unique CO binding strength and work function of Cu. 6 , 38 , 39 Therefore, we anticipate that a catalyst with an imitated electronic structure of Cu would equally generate C 2+ products. In our alloy system, by incorporating Au into Ag 1 Pd 3 to form ternary catalysts, we can tune simultaneously the d-band center and work function. By this method, we fabricated Au 1.5 Ag 1 Pd 3 and Au 3 Ag 1 Pd 3 with a similar d-band center and work function as Cu (Fig. 4 a, b). To our surprise, CO 2 electrolysis on these Cu-like AuAgPd alloys revealed no generation of any C 2+ products at all potentials (Fig. 4 c and Supplementary Fig. 19). In contrast to the general opinion, this indicates a limitation of the descriptor model to explain Cu’s unique ability for C-C coupling. To investigate the reason for this shortcoming, we turned to a computational study of the alloy systems. For this, we considered the key elementary reaction steps that are currently thought to be limiting the formation of C 2+ products, 40,41 CO 2 adsorption, CO 2 reduction to *CO, second CO 2 adsorption, reduction of the second *CO 2 to *CO, and finally coupling of the two *CO adsorbates: \(\:\text{C}{\text{O}}_{2}\left(\text{g}\right){+}^{\text{*}}\rightleftharpoons\:{}_{\:}{}^{\text{*}}\text{C}{\text{O}}_{2}\) \(\:{}_{\:}{}^{\text{*}}\text{C}{\text{O}}_{2}+2{\text{H}}^{+}+2{\text{e}}^{-}\rightleftharpoons\:{}_{\:}{}^{\text{*}}\text{C}\text{O}+{\text{H}}_{2}\text{O}\left(\text{l}\right)\) \(\:{}_{\:}{}^{\text{*}}\text{C}\text{O}\rightleftharpoons\:\text{C}\text{O}\left(\text{g}\right)\) \(\:{}_{\:}{}^{\text{*}}\text{C}\text{O}+\text{C}{\text{O}}_{2}\left(\text{g}\right)\rightleftharpoons\:{}_{\:}{}^{\text{*}}\text{C}{\text{O}}_{\:}+{}_{\:}{}^{\text{*}}\text{C}{\text{O}}_{2}\) \(\:{}_{\:}{}^{\text{*}}\text{C}{\text{O}}_{\:}+{}_{\:}{}^{\text{*}}\text{C}{\text{O}}_{2}+2{\text{H}}^{+}+2{\text{e}}^{-}\rightleftharpoons\:2{}_{\:}{}^{\text{*}}\text{C}\text{O}+{\text{H}}_{2}\text{O}\left(\text{l}\right)\) \(\:2{}_{\:}{}^{\text{*}}\text{C}{\text{O}}_{\:}\rightleftharpoons\:{}_{\:}{}^{\text{*}}\text{O}\text{C}\text{C}\text{O}\) (1) Due to the strong dependence of CO 2 adsorption and CO coupling on the interfacial electric field, it is important to perform all DFT calculations under a realistic surface charge density/electric field condition. We utilized an implicit solvation model to add excess electrons to the system which are compensated by realistically distributed countercharges in the electrolyte. The constant charge scheme is the same as outlined in our previous work 6 , 24 , 25 , 42 and is described in more detail in the Supplementary Materials. The complexity of the alloys makes it practically impossible to digitally reconstruct the atomistic structure of the experimental system. Thus, instead, we decided to construct simplified model surfaces to mimic the limiting edges of the surface composition space, and thus the limiting factors for different compositions. In practice, this corresponds to single-atom alloys constructed from the three metals. The intermediates were positioned close to the dopant metal on the most stable adsorption site to explore the variations induced in the free energy diagram from the electronic structure perturbation. In line with the XRD analyses, we chose the face-centered cubic crystal structure and the (100) facet exposure that is known to be the active of the abundant sites for C-C coupling 43 (Supplementary Fig. 5, 20). Figure 5 shows the thermodynamic barriers for three selected intermediate steps for binary systems at −1.5 V SHE . Based on the data in Fig. 5 a, it can be noted that Cu demonstrates CO 2 adsorption behavior like that of Ag alloys, while Au and Pd alloys exhibit notably lower barriers. However, as shown in Fig. 5 b, CO is expected to be desorbed immediately after the reduction of *CO 2 in the systems without Pd. On the other hand, for the systems with Pd, the formation of *OCCO is very unfavored due to strong CO binding, as seen in Fig. 5 c. This shows that the strong ability of Cu to form C 2+ products comes from the fact that it avoids all three deactivation pathways, by providing facile CO 2 adsorption, avoiding *CO desorption, and enabling *OCCO formation. From the investigations of the dilute limits, we conclude that Pd is needed for providing sufficient *CO coverage, and Au/Ag for reducing the C-C coupling barrier. A homogeneous mixture of those metals would be beneficial but might be then limited by the adsorption of the second CO 2 molecule that has to approach closely the Pd site. These results highlight the complications of designing heterogeneous metal alloy surfaces using simple descriptor models in particular if two active sites are involved in the reaction process like for the formation of C 2+ products. Conclusions In summary, we employed the co-sputtering technique to fabricate pure metals, binary alloys, and ternary alloys using Au, Ag, and Pd metals. This technique enables continuous variation of the d-band center and work function of the alloys, measured through UPS, corresponding to altered CO adsorption energies and PZC. We then investigated the performance of the descriptors in predicting electrochemical performance, i.e. CO 2 reduction rate to CO, HCOO − , and H 2 . From this, the d-band center was found to be the primary descriptor for CO production, while both the d-band center and work function are descriptors for HCOO − production, showing a volcano-like 2-dimensional activity plot. Furthermore, by designing a ternary alloy mimicking the descriptor values of copper, we did not find any C 2 product formation suggesting limitations of the descriptor model. Through charge-dependent DFT calculations, we reveal that the heterogeneity of the adsorption energies across the alloy surface results in different deactivation pathways which prevent C-C coupling from happening. Our research validates the importance of these descriptors for predicting the performance of CO 2 RR but also highlights the challenges in particular when applying it to multi-site reaction pathways like C 2+ formation. Methods Catalyst preparation Every catalyst including Au, Ag, Pd, Cu, AuPd alloys, AgPd alloys, AuAg alloys, and AuAgPd alloys is deposited by sputtering on PTFE membrane (Aspired Laminated, Hydrophobic, Polypropylene Backer, 0.45 Micron, STERLITECH) The Au (99.99%, iTasco), Ag (99.99%, iTasco), Pd (99.99%, iTasco) and Cu (99.998%, iTasco) sputtering targets were powered by direct current (DC). Deposition time was 5 min for every catalyst and argon (Ar) gas was purged on the chamber of sputter with 5 mTorr during the deposition time. The composition of each alloy was controlled by the power of each gun. Material characterization SEM images were obtained with Magellan400 from FEI. TEM, STEM and TEM-EDS image were obtained with Talos F200X from FEI. XRD patterns were obtained with SmartLab from RIGAKU. XPS and UPS were measured with Axis-Supra from Kratos. Electrochemical CO 2 RR in a flow reactor All measurements for electrochemical CO 2 RR were conducted in a gas diffusion electrode-based flow cell, consisting of two electrolyte chambers and one gas chamber. An anion exchange membrane, Selemion membrane, separated the catholyte chamber from the anolyte chamber. An Ag/AgCl reference electrode (Saturated KCl, RE-1CP) was positioned near the cathode electrode, and a Nickel Iron Copper Molybdenum foil (Alfa Aesar) served as the anode. Both the catholyte and anolyte were prepared with 1 M KHCO 3 . Constant potential was applied to the cathode through an electrochemical workstation (VSP/VMP3B-5, BioLogic). Prior to each measurement, a 5 minute pre-reduction step at 40 mA cm − 2 was performed. For potentials of −0.6 V RHE , −0.7 V RHE , and −0.8 V RHE , the constant potential was applied to the cathode for 40 minutes. Uncompensated resistance was measured and automatically compensated by potential-stat. The catholyte was continuously circulated at a rate of 13.24 ml/min using a peristaltic pump (Major Science), and CO 2 gas was introduced at 20 sccm through a mass flow controller (MKP) during the reaction. The outlet gas flow rate was monitored by an electronic flow meter (Agilent). Gas products from CO 2 RR were analyzed using gas chromatography (3000 Micro GC, INFICON), while liquid products were assessed using high performance liquid chromatography (YL9100) and headspace GC (YL Instruments). The Faradaic efficiency (FE) for a specific product was determined by the formula: FE j (%) = \(\:\frac{{\text{n}}_{\text{j}}\text{ ∙ }{\text{z}}_{\text{j}}\text{ ∙ }\text{F}}{\text{Q}}\) (2) where n j is the moles of the specific product measured from GC and LC; z j is the number of electrons required to produce the product; F is the Faraday’s constant, and Q is the total charge applied during the measurement. The partial current density for a specific product was calculated by multiplying the FE of the product by the total current density. DFT calculation Using Vienna Ab-initio Simulation Package (VASP) version 6.4.1 software, 44 we performed the density functional theory (DFT) calculations with the projector-augmented wave (PAW) 45 pseudopotentials provided along with VASP. The standard version 46 of the pseudopotentials generated by VASP has been used for Au, Pd, Ag, C, and O. The calculations were carried out with the Bayesian error estimation functional with van der Waals correlation (BEEF-vdW) 47 to consider the precise description of the surficial systems. Additionally, implicit solvation with planar counter charge was included using VASPsol 48 , 49 to simulate the experimental conditions of solvent, pH, and electric potential. The PZC of each surficial system, defined with the Fermi level from the planar average potential along the c direction of the structure and the vacuum level, was evaluated using the following equation: $$\:{U}_{\text{P}\text{Z}\text{C}}=-\left({E}_{\text{F}}+{E}_{\text{S}\text{H}\text{I}\text{F}\text{T}}\right)+{\Delta\:}{U}_{\text{S}\text{H}\text{E}}^{\text{e}\text{x}\text{p}}=-\left({E}_{\text{F}}+{E}_{\text{S}\text{H}\text{I}\text{F}\text{T}}\right)-4.6$$ 3 where \(\:{E}_{\text{F}}\) is the Fermi level, \(\:{E}_{\text{S}\text{H}\text{I}\text{F}\text{T}}\) is a correction constant for the reference electrostatic potential, and \(\:{\Delta\:}{U}_{\text{S}\text{H}\text{E}}^{\text{e}\text{x}\text{p}}\) is a constant shift in between the experimental and the computed PZC. 49 Bulk structures were generated using ASE 50 and relaxed with a kinetic energy cutoff for the plane waves of 600 eV and the smallest allowed spacing in the k-grid of 0.1 Å − 1 with gamma-centered mesh. For the surface slabs, symmetric 4-layered 3x3 (100) surfaces were built from the bulk crystal using the CatKit Python package from SUNCAT ( https://github.com/SUNCAT-Center/CatKit ) and relaxed with a kinetic energy cutoff of 400 eV and a smallest allowed spacing in the k-grid of 0.1 Å − 1 with gamma-centered mesh. The minimum length of the lattice parameter c vertical to the a and b lattice plane was 25 Å. Each adsorbate was positioned symmetrically at both sides of the surface near the dopant for the adsorption energy calculation. All calculations were performed using Gaussian smearing with a width of 0.05 eV. Geometry relaxations were performed until the norm of each atom’s forces reached 0.05 eV/Å. For a few systems, an additional net charge was induced to obtain a reliable adsorption structure. After the relaxation, DFT energy was calculated by removing the charge within the constraint of the structure. To calculate the grand-canonical potential, charge-dependent DFT energy \(\:{{\Delta\:}}_{i}E\left(\sigma\:\left(U\right)\right)\) must be evaluated with the following second-order approximated equation: $$\:{{\Delta\:}}_{i}E\left(\sigma\:\left(U\right)\right)={{\Delta\:}}_{i}{E}_{0}+{{\Delta\:}}_{i}a\cdot\:\sigma\:+{{\Delta\:}}_{i}b\cdot\:{\sigma\:}^{2}$$ 4 where \(\:{{\Delta\:}}_{i}{E}_{0}\) is the DFT energy at the zero charge, \(\:{{\Delta\:}}_{i}a\) , \(\:{{\Delta\:}}_{i}b\) are the coefficients for the charge dependence (Supplementary Table 5). This expression is inserted into the grand-canonical free energy: \(\:{{\Delta\:}}_{i}{\Omega\:}\left(U,\text{p}\text{H}\right)={{\Delta\:}}_{i}E\left(\sigma\:\left(U\right)\right)+{{\Delta\:}}_{i}{E}_{\text{T},\text{S},\text{Z}\text{P}\text{E}}^{^\circ\:}+0.0592\cdot\:\text{p}\text{H}+\text{e}U\) \(\:\sigma\:\left(U\right)=C\cdot\:(U-{U}_{\text{P}\text{Z}\text{C}})\) (5) with the DFT energy \(\:{{\Delta\:}}_{i}E\left(\sigma\:\left(U\right)\right)\) , the surface charge density \(\:\sigma\:\) , the electrode potential vs. SHE \(\:U\) , thermal and zero-point energy (ZPE) correction term \(\:{{\Delta\:}}_{i}{E}_{\text{T},\text{S},\text{Z}\text{P}\text{E}}^{^\circ\:}\) , the capacitance of the system \(\:C\) and the potential of zero charge (PZC) \(\:{U}_{\text{P}\text{Z}\text{C}}\) . The surface charge density is thus determined from the PZC and the capacitance, for which we took the experimental value of 20 µF/cm 2 as an approximate metal-independent value. Also, to reduce the computational cost, we approximated the thermal and zero-point energy correction to be the same for all systems as Cu. We direct readers to the method section for further details. Additionally, + 0.33 eV and + 0.09 eV of gas phase correction for DFT energy has been added for CO 2 and H 2 . 51 General examples of free energy diagrams via the approach above are presented (Supplementary Fig. 21). Declarations Competing Interests The authors declare no competing interests. Author contributions B.K., S.H., S.R., and J.O. wrote the manuscript. B.K. performed the experiments and analyzed the data. S.H. and S.R. contributed DFT calculation. S.W. edited and supported the manuscripts. J.O. design the experiments and supervised the research, data analysis. All authors contributed to this project and have given approval to the final version of the manuscript. Acknowledgements This research was supported by a grant from the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (grant no. 2021R1A2C3007280, RS-2024-00435493). S.R. additionally acknowledges financial support from the NRF, funded by the Ministry of Science and ICT (grant no. 2021R1C1C1008776). We also acknowledge the support of the National Supercomputing Center of Korea Institute of Science and Technology Information (KISTI) with super-computing resources, including technical support (no. KSC-2023-CRE-0468). Data availability Source data to generate figures and tables are available from the corresponding authors. References Xu, Q. et al. Descriptor for Hydrogen Evolution Catalysts Based on the Bulk Band Structure Effect. ACS Catalysis 10, 5042–5048 (2020). https://doi.org:10.1021/acscatal.9b05539 Kim, B. et al. Trace-Level Cobalt Dopants Enhance CO 2 Electroreduction and Ethylene Formation on Copper. ACS Energy Letters 8, 3356–3364 (2023). https://doi.org:10.1021/acsenergylett.3c00418 Hong, S. et al. Tuning the C1/C2 Selectivity of Electrochemical CO 2 Reduction on Cu–CeO2 Nanorods by Oxidation State Control. Advanced Materials 35, 2208996 (2023). https://doi.org:https://doi.org/10.1002/adma.202208996 Kim, B. et al. Over a 15.9% Solar-to-CO Conversion from Dilute CO 2 Streams Catalyzed by Gold Nanoclusters Exhibiting a High CO2 Binding Affinity. ACS Energy Letters 5, 749–757 (2020). https://doi.org:10.1021/acsenergylett.9b02511 Peng, H. et al. The role of atomic carbon in directing electrochemical CO 2 reduction to multicarbon products. Energy & Environmental Science 14, 473–482 (2021). https://doi.org:10.1039/D0EE02826F Ringe, S. The importance of a charge transfer descriptor for screening potential CO 2 reduction electrocatalysts. Nature Communications 14, 2598 (2023). https://doi.org:10.1038/s41467-023-37929-4 Bagger, A., Ju, W., Varela, A. S., Strasser, P. & Rossmeisl, J. Electrochemical CO 2 Reduction: A Classification Problem. ChemPhysChem 18, 3266–3273 (2017). https://doi.org:https://doi.org/10.1002/cphc.201700736 Gao, W. et al. CO Binding Energy is an Incomplete Descriptor of Cu-Based Catalysts for the Electrochemical CO 2 Reduction Reaction. Angewandte Chemie International Edition 62, e202313798 (2023). https://doi.org:https://doi.org/10.1002/anie.202313798 Seh, Z. W. et al. Combining theory and experiment in electrocatalysis: Insights into materials design. Science 355, eaad4998 (2017). https://doi.org:doi:10.1126/science.aad4998 Cao, L. Recent advances in the application of machine-learning algorithms to predict adsorption energies. Trends in Chemistry 4, 347–360 (2022). https://doi.org : https://doi.org/10.1016/j.trechm.2022.01.012 Lan, J. et al. AdsorbML: a leap in efficiency for adsorption energy calculations using generalizable machine learning potentials. npj Computational Materials 9, 172 (2023). https://doi.org:10.1038/s41524-023-01121-5 Kuhl, K. P. et al. Electrocatalytic Conversion of Carbon Dioxide to Methane and Methanol on Transition Metal Surfaces. Journal of the American Chemical Society 136, 14107–14113 (2014). https://doi.org:10.1021/ja505791r Hori, Y., Wakebe, H., Tsukamoto, T. & Koga, O. Electrocatalytic process of CO selectivity in electrochemical reduction of CO 2 at metal electrodes in aqueous media. Electrochimica Acta 39, 1833–1839 (1994). https://doi.org:https://doi.org/10.1016/0013-4686(94)85172-7 Chung, D. Y. et al. Inhibition of CO poisoning on Pt catalyst coupled with the reduction of toxic hexavalent chromium in a dual-functional fuel cell. Scientific Reports 4, 7450 (2014). https://doi.org:10.1038/srep07450 Hori, Y., Takahashi, R., Yoshinami, Y. & Murata, A. Electrochemical Reduction of CO at a Copper Electrode. The Journal of Physical Chemistry B 101, 7075–7081 (1997). https://doi.org:10.1021/jp970284i Nørskov, J. K., Abild-Pedersen, F., Studt, F. & Bligaard, T. Density functional theory in surface chemistry and catalysis. Proceedings of the National Academy of Sciences 108, 937–943 (2011). https://doi.org:10.1073/pnas.1006652108 Hammer, B., Morikawa, Y. & Nørskov, J. K. CO Chemisorption at Metal Surfaces and Overlayers. Physical Review Letters 76, 2141–2144 (1996). https://doi.org:10.1103/PhysRevLett.76.2141 Dickens, C. F., Montoya, J. H., Kulkarni, A. R., Bajdich, M. & Nørskov, J. K. An electronic structure descriptor for oxygen reactivity at metal and metal-oxide surfaces. Surface Science 681, 122–129 (2019). https://doi.org:https://doi.org/10.1016/j.susc.2018.11.019 Trasatti, S. Work function, electronegativity, and electrochemical behaviour of metals: II. Potentials of zero charge and “electrochemical” work functions. Journal of Electroanalytical Chemistry and Interfacial Electrochemistry 33, 351–378 (1971). https://doi.org:https://doi.org/10.1016/S0022-0728(71)80123-7 Jackson, C., Smith, G., Russell, A. E., Levecque, P. & Kramer, D. Electronic metal-support interactions in vacuum vs. electrolyte. Nature Communications 11, 1470 (2020). https://doi.org:10.1038/s41467-020-15307-8 Montoya, J. H., Shi, C., Chan, K. & Nørskov, J. K. Theoretical Insights into a CO Dimerization Mechanism in CO 2 Electroreduction. The Journal of Physical Chemistry Letters 6, 2032–2037 (2015). https://doi.org:10.1021/acs.jpclett.5b00722 Karmodak, N., Vijay, S., Kastlunger, G. & Chan, K. Computational Screening of Single and Di-Atom Catalysts for Electrochemical CO 2 Reduction. ACS Catalysis 12, 4818–4824 (2022). https://doi.org:10.1021/acscatal.1c05750 Sandberg, R. B., Montoya, J. H., Chan, K. & Nørskov, J. K. CO-CO coupling on Cu facets: Coverage, strain and field effects. Surface Science 654, 56–62 (2016). https://doi.org:https://doi.org/10.1016/j.susc.2016.08.006 Ringe, S. et al. Understanding cation effects in electrochemical CO 2 reduction. Energy & Environmental Science 12, 3001–3014 (2019). https://doi.org:10.1039/C9EE01341E Ringe, S. et al. Double layer charging driven carbon dioxide adsorption limits the rate of electrochemical carbon dioxide reduction on Gold. Nature Communications 11, 33 (2020). https://doi.org:10.1038/s41467-019-13777-z Chen, L. D., Urushihara, M., Chan, K. & Nørskov, J. K. Electric Field Effects in Electrochemical CO 2 Reduction. ACS Catalysis 6, 7133–7139 (2016). https://doi.org:10.1021/acscatal.6b02299 Dong, W. J. et al. Electric-field-driven electrochemical CO 2 reduction of sharpened Sn/Cu catalysts. Applied Surface Science 565, 150460 (2021). https://doi.org: https://doi.org/10.1016/j.apsusc.2021.150460 Tan, Y. C. et al. Pitfalls and Protocols: Evaluating Catalysts for CO 2 Reduction in Electrolyzers Based on Gas Diffusion Electrodes. ACS Energy Letters 7, 2012–2023 (2022). https://doi.org:10.1021/acsenergylett.2c00763 Ghosh, G., Kantner, C. & Olson, G. B. Thermodynamic modeling of the Pd-X (X = Ag, Co, Fe, Ni) systems. Journal of Phase Equilibria 20, 295–308 (1999). https://doi.org:10.1361/105497199770335811 Okamoto, H. & Massalski, T. B. The Au – Pd (Gold-Palladium) system. Bulletin of Alloy Phase Diagrams 6, 229–235 (1985). https://doi.org:10.1007/BF02880404 Hansen, H. A., Varley, J. B., Peterson, A. A. & Nørskov, J. K. Understanding Trends in the Electrocatalytic Activity of Metals and Enzymes for CO 2 Reduction to CO. The Journal of Physical Chemistry Letters 4, 388–392 (2013). https://doi.org:10.1021/jz3021155 Ringe, S. Cation effects on electrocatalytic reduction processes at the example of the hydrogen evolution reaction. Current Opinion in Electrochemistry 39, 101268 (2023). https://doi.org: https://doi.org/10.1016/j.coelec.2023.101268 Lee, M.-Y., Ringe, S., Kim, H., Kang, S. & Kwon, Y. Electric Field Mediated Selectivity Switching of Electrochemical CO 2 Reduction from Formate to CO on Carbon Supported Sn. ACS Energy Letters 5, 2987–2994 (2020). https://doi.org:10.1021/acsenergylett.0c01387 Kortlever, R., Shen, J., Schouten, K. J. P., Calle-Vallejo, F. & Koper, M. T. M. Catalysts and Reaction Pathways for the Electrochemical Reduction of Carbon Dioxide. The Journal of Physical Chemistry Letters 6, 4073–4082 (2015). https://doi.org:10.1021/acs.jpclett.5b01559 Sun, Z., Ma, T., Tao, H., Fan, Q. & Han, B. Fundamentals and Challenges of Electrochemical CO 2 Reduction Using Two-Dimensional Materials. Chem 3, 560–587 (2017). https://doi.org:https://doi.org/10.1016/j.chempr.2017.09.009 Zou, J., Liang, G., Lee, C.-Y. & Wallace, G. G. Progress and perspectives for electrochemical CO 2 reduction to formate. Materials Today Energy 38, 101433 (2023). https://doi.org:https://doi.org/10.1016/j.mtener.2023.101433 Kuhl, K. P., Cave, E. R., Abram, D. N. & Jaramillo, T. F. New insights into the electrochemical reduction of carbon dioxide on metallic copper surfaces. Energy & Environmental Science 5, 7050–7059 (2012). https://doi.org:10.1039/C2EE21234J Ma, M. et al. Electrochemical reduction of CO 2 on compositionally variant Au-Pt bimetallic thin films. Nano Energy 42, 51–57 (2017). https://doi.org: https://doi.org/10.1016/j.nanoen.2017.09.043 Liu, K. et al. Electronic Effects Determine the Selectivity of Planar Au–Cu Bimetallic Thin Films for Electrochemical CO 2 Reduction. ACS Applied Materials & Interfaces 11, 16546–16555 (2019). https://doi.org:10.1021/acsami.9b01553 Kastlunger, G. et al. Using pH Dependence to Understand Mechanisms in Electrochemical CO Reduction. ACS Catalysis 12, 4344–4357 (2022). https://doi.org:10.1021/acscatal.1c05520 Liu, X. et al. pH effects on the electrochemical reduction of CO 2 towards C2 products on stepped copper. Nature Communications 10, 32 (2019). https://doi.org:10.1038/s41467-018-07970-9 Kim, D. H. et al. Selective electrochemical reduction of nitric oxide to hydroxylamine by atomically dispersed iron catalyst. Nature Communications 12, 1856 (2021). https://doi.org:10.1038/s41467-021-22147-7 Hori, Y., Takahashi, I., Koga, O. & Hoshi, N. Electrochemical reduction of carbon dioxide at various series of copper single crystal electrodes. Journal of Molecular Catalysis A: Chemical 199, 39–47 (2003). https://doi.org:https://doi.org/10.1016/S1381-1169(03)00016-5 Kresse, G. & Furthmüller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Physical Review B 54, 11169–11186 (1996). https://doi.org:10.1103/PhysRevB.54.11169 Blöchl, P. E. Projector augmented-wave method. Physical Review B 50, 17953–17979 (1994). https://doi.org:10.1103/PhysRevB.50.17953 Kresse, G. & Joubert, D. From ultrasoft pseudopotentials to the projector augmented-wave method. Physical Review B 59, 1758–1775 (1999). https://doi.org:10.1103/PhysRevB.59.1758 Wellendorff, J. et al. Density functionals for surface science: Exchange-correlation model development with Bayesian error estimation. Physical Review B 85, 235149 (2012). https://doi.org:10.1103/PhysRevB.85.235149 Mathew, K., Sundararaman, R., Letchworth-Weaver, K., Arias, T. A. & Hennig, R. G. Implicit solvation model for density-functional study of nanocrystal surfaces and reaction pathways. The Journal of Chemical Physics 140 (2014). https://doi.org:10.1063/1.4865107 Mathew, K., Kolluru, V. S. C., Mula, S., Steinmann, S. N. & Hennig, R. G. Implicit self-consistent electrolyte model in plane-wave density-functional theory. The Journal of Chemical Physics 151 (2019). https://doi.org:10.1063/1.5132354 Hjorth Larsen, A. et al. The atomic simulation environment—a Python library for working with atoms. Journal of Physics: Condensed Matter 29, 273002 (2017). https://doi.org:10.1088/1361-648X/aa680e Studt, F., Abild-Pedersen, F., Varley, J. B. & Nørskov, J. K. CO and CO 2 Hydrogenation to Methanol Calculated Using the BEEF-vdW Functional. Catalysis Letters 143, 71–73 (2013). https://doi.org:10.1007/s10562-012-0947-5 Additional Declarations There is NO Competing Interest. Supplementary Files 241127SupportingInformationNatureCatalysis.docx Peaks and pitfalls of electrocatalytic descriptor models at the example of CO2 reduction Cite Share Download PDF Status: Published Journal Publication published 13 Apr, 2026 Read the published version in Nature Catalysis → 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-5559232","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Analysis","associatedPublications":[],"authors":[{"id":400704047,"identity":"7c405e64-4759-49f6-a5e1-536da1522a39","order_by":0,"name":"Jihun Oh","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA3klEQVRIiWNgGAWjYBACxmYeMJ3AD6QPQMTYiNQi2QPSkkCEFgYGqBaDM2CKCC3M7bzHpAsYbPKMzxx+ePDnDwZ5/ga2tA/4HcaXJj2DIa3Y7GybwWGeBAbDGQfYDs8g4BczaR6Gw4nbzjMYHAY6jHEDA3szXofBtWzuZ/9w8EcCgz3xWjbw9hgcADoscQMD22FCWoyteQzSEmecOVNwmCdNInnGYbZkvFoM+88Y3uapsEns70nf/PGHjY1tf3ubMX4tDSDSAM6XAIY7Xg0MDPIE5EfBKBgFo2AUMDAAAMe7QL0Vz2wcAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0001-6465-6736","institution":"Korea Advanced Institute of Science and Technology","correspondingAuthor":true,"prefix":"","firstName":"Jihun","middleName":"","lastName":"Oh","suffix":""},{"id":400704048,"identity":"0067eab4-f787-4b98-8d78-7af769f9a3dd","order_by":1,"name":"Beomil Kim","email":"","orcid":"","institution":"Korea Advanced Institute of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Beomil","middleName":"","lastName":"Kim","suffix":""},{"id":400704049,"identity":"fee8931e-c720-4ff8-9541-bedda1661a97","order_by":2,"name":"Seungchang Han","email":"","orcid":"https://orcid.org/0009-0009-1533-9350","institution":"Korea University","correspondingAuthor":false,"prefix":"","firstName":"Seungchang","middleName":"","lastName":"Han","suffix":""},{"id":400704050,"identity":"903a6933-8a67-4980-80c5-8395b847ad81","order_by":3,"name":"Suneon Wang","email":"","orcid":"https://orcid.org/0000-0002-5573-8710","institution":"Korea Advanced Institute of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Suneon","middleName":"","lastName":"Wang","suffix":""},{"id":400704051,"identity":"81b9ede6-3bf8-4939-9f0b-2b2d6d2645d7","order_by":4,"name":"Stefan Ringe","email":"","orcid":"https://orcid.org/0000-0002-7804-1406","institution":"Korea University","correspondingAuthor":false,"prefix":"","firstName":"Stefan","middleName":"","lastName":"Ringe","suffix":""}],"badges":[],"createdAt":"2024-12-01 15:40:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5559232/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5559232/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41929-026-01526-7","type":"published","date":"2026-04-13T04:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":73730647,"identity":"a359ea91-29a2-469c-b478-ca7b253713fa","added_by":"auto","created_at":"2025-01-14 05:31:45","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1453626,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSynthesis method and the morphology of the catalysts.\u003c/strong\u003e \u003cstrong\u003ea,\u003c/strong\u003e Schematic of sputtering method. \u003cstrong\u003eb,\u003c/strong\u003e SEM image of the PTFE substrate. \u003cstrong\u003ec,\u003c/strong\u003e SEM-EDS of Au\u003csub\u003e1\u003c/sub\u003ePd\u003csub\u003e1\u003c/sub\u003e, Ag\u003csub\u003e1\u003c/sub\u003ePd\u003csub\u003e1\u003c/sub\u003e, and Au\u003csub\u003e1\u003c/sub\u003eAg\u003csub\u003e1\u003c/sub\u003e alloys and \u003cstrong\u003ed,\u003c/strong\u003e TEM-EDS for the Au\u003csub\u003e3\u003c/sub\u003eAg\u003csub\u003e1\u003c/sub\u003ePd\u003csub\u003e3\u003c/sub\u003e alloy before CO\u003csub\u003e2\u003c/sub\u003eRR.\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-5559232/v1/e02e556fcf26847fc1f4bfff.png"},{"id":73729364,"identity":"618d0e47-1824-4256-bd90-704a2b8122b8","added_by":"auto","created_at":"2025-01-14 05:07:45","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":104488,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eContour plots for the d-band center and work function of the catalysts.\u003c/strong\u003e \u003cstrong\u003ea-b,\u003c/strong\u003e The contour plot for (\u003cstrong\u003ea\u003c/strong\u003e) the d-band center and (\u003cstrong\u003eb\u003c/strong\u003e) work function of the catalysts relative to the compositions of Au, Ag, and Pd.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-5559232/v1/db4f2ffc887db393e52f5b0d.png"},{"id":73729367,"identity":"138e7934-8907-4e7d-8fe5-78dd08672ea6","added_by":"auto","created_at":"2025-01-14 05:07:45","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":221657,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCO\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003eRR properties as functions of the d-band center and work function. a-c,\u003c/strong\u003e Contour plots for the logarithm of \u003cem\u003ej\u003c/em\u003e\u003csub\u003eCO,ECSA\u003c/sub\u003e as a function of the d-band center and work function for every alloy at (a) -0.6 V\u003csub\u003eRHE\u003c/sub\u003e, (b) -0.7 V\u003csub\u003eRHE\u003c/sub\u003e, and (c) -0.8 V\u003csub\u003eRHE\u003c/sub\u003e. \u003cstrong\u003ed-f, \u003c/strong\u003eContour plots for the logarithm of \u003cem\u003ej\u003c/em\u003e\u003csub\u003eHCOO\u003c/sub\u003e\u003csup\u003e-\u003c/sup\u003e\u003csub\u003e,ECSA\u003c/sub\u003e as a function of the d-band center and work function for every alloy at (d) -0.6 V\u003csub\u003eRHE\u003c/sub\u003e, (e) -0.7 V\u003csub\u003eRHE\u003c/sub\u003e, and (f) -0.8 V\u003csub\u003eRHE\u003c/sub\u003e.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-5559232/v1/d5983b6bd8b5b521ba733151.png"},{"id":73730648,"identity":"9026662f-d953-47be-8557-8d1c1daf21ab","added_by":"auto","created_at":"2025-01-14 05:31:45","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":80035,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eComparison of AuAgPd alloys and Cu metal. a-b, \u003c/strong\u003e(\u003cstrong\u003ea\u003c/strong\u003e) The d-band and (\u003cstrong\u003eb\u003c/strong\u003e) work function of Ag\u003csub\u003e1\u003c/sub\u003ePd\u003csub\u003e3\u003c/sub\u003e, AuAgPd alloys, and Cu metal. \u003cstrong\u003ec,\u003c/strong\u003e CO\u003csub\u003e2\u003c/sub\u003eRR properties (Faradaic efficiency and total current density) of AuAgPd alloys and Cu metal measured at -0.8 V\u003csub\u003eRHE\u003c/sub\u003e.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-5559232/v1/8d4c82aced4ca58557669086.png"},{"id":73730650,"identity":"685a3501-9871-465a-acfb-09053978c463","added_by":"auto","created_at":"2025-01-14 05:31:46","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":223798,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThermodynamic barrier of major C\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e2+\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e production deactivation intermediate step at pH = 7, \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eU\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e = -1.5 V\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003eSHE\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e. a-c, \u003c/strong\u003e(\u003cstrong\u003ea\u003c/strong\u003e) CO\u003csub\u003e2\u003c/sub\u003e adsorption (\u003cstrong\u003eb\u003c/strong\u003e) CO desorption (\u003cstrong\u003ec\u003c/strong\u003e) OCCO formation. \u003cstrong\u003ed,\u003c/strong\u003e Summary of a-c.\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-5559232/v1/a95adeb4396e4602ffc72864.png"},{"id":106853923,"identity":"5c58774c-4bd7-45ee-a32e-1a0a872a2390","added_by":"auto","created_at":"2026-04-14 07:05:40","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2715320,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5559232/v1/94c6aad4-af37-4828-b2b4-bc7e303c6095.pdf"},{"id":73730655,"identity":"c151502f-2e41-4735-bc91-bcff50e88f87","added_by":"auto","created_at":"2025-01-14 05:32:12","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":10718407,"visible":true,"origin":"","legend":"Peaks and pitfalls of electrocatalytic descriptor models at the example of CO2 reduction","description":"","filename":"241127SupportingInformationNatureCatalysis.docx","url":"https://assets-eu.researchsquare.com/files/rs-5559232/v1/cc06a1bcbd1b99f0f6296b71.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Peaks and pitfalls of electrocatalytic descriptor models at the example of CO2 reduction","fulltext":[{"header":"Introduction","content":"\u003cp\u003eElectrochemical CO\u003csub\u003e2\u003c/sub\u003e reduction reaction (CO\u003csub\u003e2\u003c/sub\u003eRR) has attracted intense interest in the context of carbon capture and utilization (CCU) technology. Carbon monoxide (CO), hydrocarbons (e.g., CH\u003csub\u003e4\u003c/sub\u003e, C\u003csub\u003e2\u003c/sub\u003eH\u003csub\u003e4\u003c/sub\u003e, etc.), and alcohols (e.g., C\u003csub\u003e2\u003c/sub\u003eH\u003csub\u003e5\u003c/sub\u003eOH and CH\u003csub\u003e3\u003c/sub\u003eCH\u003csub\u003e2\u003c/sub\u003eCH\u003csub\u003e2\u003c/sub\u003eOH, etc.) can be formed from CO\u003csub\u003e2\u003c/sub\u003eRR along with H\u003csub\u003e2\u003c/sub\u003e from the competitive hydrogen evolution reaction (HER).\u003csup\u003e1-4\u003c/sup\u003e To reach competitiveness with established petrochemical processes, material design is key to enhancing the turnover and product selectivity of the reaction. Systematic material design requires the availability of performance descriptors, and \u003cem\u003efirst-principles\u003c/em\u003e calculations have been instrumental in deriving and rationalizing such.\u003csup\u003e5\u003c/sup\u003e Density functional theory (DFT)--based calculations have revealed that often a simple or two adsorption energy descriptors are enough to predict the full free energy diagram and thus the turnover frequency or selectivity of a catalyst.\u003csup\u003e6-9\u003c/sup\u003e The resulting activity volcanos can be used for high-throughput screening of materials, possibly facilitated by machine learning.\u003csup\u003e10,11\u003c/sup\u003e In the case of CO\u003csub\u003e2\u003c/sub\u003eRR, the binding affinity between carbon monoxide (CO) and a catalyst\u0026rsquo;s surface (i.e., the CO binding strength) has been suggested as a descriptor for metallic catalysts.\u003csup\u003e12\u003c/sup\u003e Weaker CO adsorption thereby facilitates the desorption of CO granting Au and Ag the CO-evolving CO\u003csub\u003e2\u003c/sub\u003eRR catalysts. Conversely, stronger adsorption such as to Ni\u003csup\u003e13\u003c/sup\u003e and Pt\u003csup\u003e14\u003c/sup\u003e, leads to poisoning/blocking of the active sites, preventing CO\u003csub\u003e2\u003c/sub\u003eRR and facilitating the competitive hydrogen evolution reaction (HER). This argument was used to explain why Cu with a moderate binding of CO is the only transition metal-based catalyst leading to significant amounts of higher reduced C\u003csub\u003e2+\u003c/sub\u003e products such as ethylene, ethanol, and 1-propanol.\u003csup\u003e15\u003c/sup\u003e For transition metal catalysts, the CO adsorption energy correlates well with the d-band center of the metal,\u003csup\u003e8,16-18\u003c/sup\u003e\u0026nbsp; providing a physical rational for explaining observed trends.\u003c/p\u003e\n\u003cp\u003eRecently, it has been, however, pointed out that in addition to the CO adsorption energy, the driving force of a catalyst to transfer charge to the adsorbate should be considered.\u003csup\u003e6\u003c/sup\u003e This force is related to the work function of the metal and correlates with the potential of zero charge (PZC) measured electrochemically. The charge-transfer descriptor emerges from the fact that during adsorption, charge transfers from the metal to the adsorbate leading to the formation of an electric dipole. This electric dipole interacts with the surrounding electric field from the electric double layer (EDL) that is formed at potentials deviating from the PZC.\u003csup\u003e6,19,20\u003c/sup\u003e In case of C\u003csub\u003e2+\u003c/sub\u003e production, this field-effect leads to a decrease in the kinetic barrier for CO dimer formation.\u003csup\u003e21\u003c/sup\u003e\u003csup\u003e\u0026nbsp;\u003c/sup\u003eDue to the difference in the field dependence of different adsorbates, the PZC or any other charge-transfer related descriptor\u003csup\u003e22\u003c/sup\u003e must be considered as an independent descriptor and spans together with the CO adsorption energy a two-dimensional activity and selectivity volcano for CO\u003csub\u003e2\u003c/sub\u003eRR.\u003csup\u003e6,23-27\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe current descriptor framework or CO\u003csub\u003e2\u003c/sub\u003eRR has been developed for single-element transition metal catalysts. It has remained unclear if they could be used for practical catalyst design of more complex materials, the ultimate purpose of any descriptor model. Complications arise in the measurements of the CO binding strength and PZC which are challenging and often limited\u003cem\u003e\u0026nbsp;in operando\u0026nbsp;\u003c/em\u003eCO\u003csub\u003e2\u003c/sub\u003eRR environments.\u003csup\u003e8\u003c/sup\u003e Moreover, the performance and selectivity of CO\u003csub\u003e2\u003c/sub\u003eRR depend on various factors, such as the electrode material, the catalyst structure, the electrolyte composition, the applied potential, the local environment near the electrode, and electrolytic cell designs.\u003csup\u003e28\u003c/sup\u003e This makes it problematic to compare different published datasets and use them to evaluate the ability of descriptors.\u003c/p\u003e\n\u003cp\u003eHere, we investigate the reliability of these descriptor models by studying the CO\u003csub\u003e2\u003c/sub\u003eRR activities of binary and ternary alloys of gold (Au), silver (Ag), and palladium (Pd) in a gas-fed flow cell. We measure the d-band center and work function which we consider as descriptors due to their close correlation with the reported CO binding strength and PZC descriptors, respectively. The Au, Ag, and Pd alloys with varying compositions have nearly identical morphology, crystallographic orientation and crystal structure, and a uniform composition without phase segregation, justifying our attempt to correlate CO\u003csub\u003e2\u003c/sub\u003eRR behavior with the measured descriptors. From this, we identify that the d-band center is the key descriptor for CO and HCOOH production with a contribution of work function for HCOOH, but fails to describe C\u003csub\u003e2+\u003c/sub\u003e product formation. We also synthesize AuAgPd alloys with the d-band center and work function matching those of Cu, but contrary to the expectations from Cu and the descriptor model, find no C\u003csub\u003e2+\u003c/sub\u003e products. From electric double layer-aware\u0026nbsp;DFT calculations, we show this to be related to the heterogeneous bonding environments on the surface leading to various deactivation pathways for C\u003csub\u003e2+\u003c/sub\u003e product formation. Our results highlight the chances and challenges of applying descriptor models for designing electrocatalysts.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec2\" class=\"Section2\"\u003e \u003ch2\u003eAu-Ag-Pd alloys with tunable electronic structure\u003c/h2\u003e \u003cp\u003eWe synthesized Au-Ag-Pd alloys with various compositions using a co-sputtering method that yields alloy films with uniform composition across large areas (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). The metal alloys were deposited on a chemically inert, nano-fibrous PTFE membrane gas diffusion electrode (GDE) as a carbon-based GDE may produce hydrogen (H\u003csub\u003e2\u003c/sub\u003e) at negative applied potential,\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e potentially confounding the CO\u003csub\u003e2\u003c/sub\u003eRR activities of catalysts (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). The compositions of the as-deposited alloy films are measured using X-ray photoelectron spectroscopy (XPS) and scanning electron microscope-energy dispersive spectroscopy (SEM-EDS) (Supplementary Table\u0026nbsp;1). The alloys maintain metal ratio uniformly from near surface to bulk as can be seen by comparing two spectroscopy analyses.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo investigate the morphology and spatial distribution of the chemical composition of the alloy films on PTFE membranes, electron microscopy was conducted. The SEM images indicate that the AuAg, AuPd, and AgPd binary alloy films were conformally deposited on the PTFE membranes, maintaining the fibrous nature of the GDE (the left panels of Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec and Supplementary Fig.\u0026nbsp;1). In addition, all the alloy films exhibit similar morphology across the different compositions. The SEM-EDS analyses on each binary alloy composed of Au, Ag, and Pd indicate that all metals are uniformly distributed across the PTFE membrane without noticeable secondary phase precipitations. The cross-sectional transmission electron microscope-energy dispersive spectroscopy (TEM-EDS) images also confirm the formation of the Au\u003csub\u003e3\u003c/sub\u003eAg\u003csub\u003e1\u003c/sub\u003ePd\u003csub\u003e3\u003c/sub\u003e catalyst with uniform chemical composition throughout the film thickness (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed). The absence of the phase separation in our alloys is consistent with the complete miscibility in the phase diagrams of the binary alloys of Au, Ag, and Pd.\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e Through TEM-EDS and angle resolved X-ray photoelectron spectroscopy (ARXPS) analyses confirmed that the composition alteration in both the near-surface and bulk of the ternary alloys, during CO\u003csub\u003e2\u003c/sub\u003eRR, is less than 10 atomic percent (Supplementary Fig.\u0026nbsp;2, 3, 4). Furthermore, the X-ray diffraction (XRD) measurements indicate that all pure metals and alloys have the face-centered cubic (fcc) crystal structure with a dominant (111) out-of-plane texture with similar crystallite sizes (Supplementary Fig.\u0026nbsp;5, Supplementary Table\u0026nbsp;2). Note also that no second phases are seen in XRD. Therefore, our catalysts (co-)sputtered on PTFE membranes are an ideal platform to evaluate and compare the intrinsic CO\u003csub\u003e2\u003c/sub\u003eRR properties of the alloy catalysts because they are single-phase fcc crystals with similar morphology, crystallographic orientation, and uniform chemical composition over entire films.\u003c/p\u003e \u003cp\u003eFigures \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea and b show contour plots of the d-band centers and work function of pure metals (Au, Ag, and Pd) and their binary and ternary alloys, measured by ultraviolet photoemission spectroscopy (UPS) (see the measured values in Supplementary Table\u0026nbsp;3). The d-band centers of Au (\u0026minus;4.41 eV) and Ag (\u0026minus;5.41) are located significantly lower relative to the Fermi energy level compared to the one of Pd (\u0026minus;1.58 eV). When alloying Au with Pd or Ag with Pd, the resulting d-band center of alloys shifts to an intermediate position relative to that of pure Au and Pd or Ag and Pd, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea and Supplementary Fig.\u0026nbsp;6). For example, the d-band center moves toward to the Fermi energy level when Pd is introduced to Au or Ag alloys, whereas AuAg alloys maintain a rather similar d-band center position (Supplementary Table\u0026nbsp;3).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eSimilarly, alloying alters the work function to a value that lies between those of the constituent metals. Figure\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb shows that Au (5.17 eV) and Pd (5.12 eV) have similar work function values, higher than the value of Ag (4.46 eV). Hence, alloys of Au and Pd exhibit nearly the same work function of ~\u0026thinsp;5.1 eV and the introduction of Ag into Au or Pd makes the work functions of AuAg and AgPd alloys lower than their bulk counterpart (Supplementary Table\u0026nbsp;3). These visual representations offer clear insights, enabling us to identify viable combinations of d-band center and work function achievable through alloying Au, Ag, and Pd. Therefore, a catalyst with the desired d-band center positions from \u0026minus;1.58 to \u0026minus;5.41 eV and work function from 4.46 to 5.17 eV can be fabricated from alloying two or three metals of Au, Ag, and Pd.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eDescriptor-dependent CO\u003csub\u003e2\u003c/sub\u003eRR properties of Au, Ag, and Pd alloys\u003c/h2\u003e \u003cp\u003eThe co-sputtered alloys with varying Ag-Au-Pd ratio allow us to form a catalyst with independently controlled CO\u003csub\u003e2\u003c/sub\u003eRR descriptors. To evaluate the connection between the descriptors and CO\u003csub\u003e2\u003c/sub\u003eRR performance and selectivity, each alloy with a different d-band center and work function was subject to CO\u003csub\u003e2\u003c/sub\u003e electrolysis in a 1 M KHCO\u003csub\u003e3\u003c/sub\u003e electrolyte at three distinct potentials of \u0026minus;0.6, \u0026minus;0.7, and \u0026minus;0.8 V\u003csub\u003eRHE\u003c/sub\u003e in a custom-built flow cell. Electrochemical currents are normalized by the electrochemical surface area (ECSA) of each catalyst from the EDL capacitance (DLC) (Supplementary Table\u0026nbsp;4). CO, HCOO\u003csup\u003e\u0026minus;\u003c/sup\u003e, and H\u003csub\u003e2\u003c/sub\u003e, were the only major CO\u003csub\u003e2\u003c/sub\u003eRR products measured from pure and binary metals of Ag, Au, and Pd whereas Cu also produces hydrocarbons and alcohols (Supplementary Fig.\u0026nbsp;7\u0026ndash;11).\u003c/p\u003e \u003cp\u003eFigures \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea-c show the contour plots of the partial current density for CO production normalized by ECSA (\u003cem\u003ej\u003c/em\u003e\u003csub\u003eCO,ECSA\u003c/sub\u003e) as a function of the d-band center and work function of each alloy at different potentials from \u0026minus;0.6 to \u0026minus;0.8 V\u003csub\u003eRHE\u003c/sub\u003e. Shown in Supplementary Fig.\u0026nbsp;12 are the contour plots for \u003cem\u003ej\u003c/em\u003e\u003csub\u003etotal,ECSA\u003c/sub\u003e and \u003cem\u003ej\u003c/em\u003e\u003csub\u003eCO2RR,ECSA\u003c/sub\u003e as a function of d-band center and work function. At \u0026minus;0.6 and \u0026minus;0.7 V\u003csub\u003eRHE\u003c/sub\u003e, the contour lines for CO production are curved, but nearly parallel to the work function, suggesting that the d-band center is the dominant descriptor (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb). When the cathodic potential was increased to \u0026minus;0.8 V\u003csub\u003eRHE\u003c/sub\u003e, however, the contour lines become more bent, forming a ridge connecting Ag and Pd as an apex and base, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). The scatter plots also reveals a strong correlation of the d-band center with \u003cem\u003ej\u003c/em\u003e\u003csub\u003eCO,ECSA\u003c/sub\u003e, indicated by R-square value from 0.80 to 0.89 (Supplementary Fig.\u0026nbsp;13). Note that \u003cem\u003ej\u003c/em\u003e\u003csub\u003eCO,ECSA\u003c/sub\u003e of the binary alloys lacks a linear correlation with the alloy composition (Supplementary Fig.\u0026nbsp;11). In contrast, the work function exhibits a much weaker correlation with \u003cem\u003ej\u003c/em\u003e\u003csub\u003eCO,ECSA\u003c/sub\u003e than the d-band center does (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea-c). Scatter analysis indicates that the linear correlation of the logarithm of \u003cem\u003ej\u003c/em\u003e\u003csub\u003eCO,ECSA\u003c/sub\u003e with the work function increases with applied potentials but remains limited to a maximum R-square value of 0.457 at \u0026minus;0.8 V\u003csub\u003eRHE\u003c/sub\u003e, (Supplementary Fig.\u0026nbsp;13). Therefore, the d-band center position (i.e., the CO binding strength) of the catalysts is the primary descriptor for CO\u003csub\u003e2\u003c/sub\u003e-to-CO activity at all potentials whereas the work function (i.e., PZC) of the catalysts can be designated as the secondary descriptor that weights in only at high potentials. Simply, in the material space that we looked at, a catalyst with weaker CO binding strength has a high chance of producing more CO from CO\u003csub\u003e2\u003c/sub\u003eRR. This stands in contrast to the volcano-like dependence on the CO adsorption energy proposed from theoretical works on bare transition metal surfaces.\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe contour plots reveal that HCOO\u003csup\u003e\u0026minus;\u003c/sup\u003e production has a rather different dependence on the electronic structure parameters of the catalysts (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed-f). HCOO\u003csup\u003e\u0026minus;\u003c/sup\u003e production exhibits a volcano-like activity with a peak near Au\u003csub\u003e3\u003c/sub\u003eAg\u003csub\u003e1\u003c/sub\u003e for all potentials. Firstly, \u003cem\u003ej\u003c/em\u003e\u003csub\u003eHCOO\u0026minus;,ECSA\u003c/sub\u003e increases with the d-band center and the R-square values are increased as the potential increases (Supplementary Fig.\u0026nbsp;14). Secondly, the work function dependence stands out for weakly-binding metals with a d-band center lower than \u0026minus;3.5 eV. A clear volcano-like dependence on the work function is identified for these alloys (Supplementary Fig.\u0026nbsp;15a-c). This stands in contrast to recent work on bare transition metal surfaces suggesting a continuous switch of selectivity between CO and formate when tuning the work function.\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e Thirdly, HCOO\u003csup\u003e\u0026minus;\u003c/sup\u003e production and the d-band center have a linear relationship for strongly binding metals with a d-band center higher than \u0026minus;3.0 eV (Supplementary Fig.\u0026nbsp;15d-f). Note that the relationship between \u003cem\u003ej\u003c/em\u003e\u003csub\u003eHCOO\u003c/sub\u003e\u003csup\u003e\u0026minus;\u003c/sup\u003e\u003csub\u003e,ECSA\u003c/sub\u003e and the work function is not clearly visible when all catalysts are compared with the work function as a single parameter because the d-band center-dependence is lumped together in a simple scatter plot. These findings suggest that CO\u003csub\u003e2\u003c/sub\u003e-to-HCOO\u003csup\u003e\u0026minus;\u003c/sup\u003e activity depends on both the CO binding strength and PZC and the influence of the latter on HCOO\u003csup\u003e\u0026minus;\u003c/sup\u003e formation is limited to a catalyst with weak CO binding strength.\u003c/p\u003e \u003cp\u003eWe also investigated the H\u003csub\u003e2\u003c/sub\u003e evolution reaction (HER) behaviors of our catalysts during CO\u003csub\u003e2\u003c/sub\u003eRR (Supplementary Fig.\u0026nbsp;16\u0026ndash;18). Contour plots show volcano-like HER activity with a peak near Au\u003csub\u003e3\u003c/sub\u003eAg\u003csub\u003e1\u003c/sub\u003e (Supplementary Fig.\u0026nbsp;17). A close examination also shows that the contour lines are horizontal at low work function but change vertically as the work function increases (Supplementary Fig.\u0026nbsp;18). This behavior implies that in the low work function region, the proton-transfer rate depends significantly on the electric field across the EDL likely via water structuring\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. Conversely, the d-band center becomes a prominent descriptor in the high work function region, likely due to the saturation of the water structure due to the more negatively charged electrode.\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e It must be noted that H\u003csub\u003e2\u003c/sub\u003e and HCOO\u003csup\u003e\u0026minus;\u003c/sup\u003e production display a similar dependence on the work functions of electrodes, implying that they share reaction intermediates. Indeed, it is suggested that HCOO is formed by a reaction of CO\u003csub\u003e2\u003c/sub\u003e with hydrogen adsorbed on a surface.\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e6\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eAs a further test for the descriptor model, we designed an alloy with a d-band center and work function comparable to that of Cu. It is well-known that Cu is the only metal that produces C\u003csub\u003e2+\u003c/sub\u003e hydrocarbons and alcohols with significant FE.\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e Previous descriptor models have explained this behavior by the unique CO binding strength and work function of Cu.\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e,\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e Therefore, we anticipate that a catalyst with an imitated electronic structure of Cu would equally generate C\u003csub\u003e2+\u003c/sub\u003e products. In our alloy system, by incorporating Au into Ag\u003csub\u003e1\u003c/sub\u003ePd\u003csub\u003e3\u003c/sub\u003e to form ternary catalysts, we can tune simultaneously the d-band center and work function. By this method, we fabricated Au\u003csub\u003e1.5\u003c/sub\u003eAg\u003csub\u003e1\u003c/sub\u003ePd\u003csub\u003e3\u003c/sub\u003e and Au\u003csub\u003e3\u003c/sub\u003eAg\u003csub\u003e1\u003c/sub\u003ePd\u003csub\u003e3\u003c/sub\u003e with a similar d-band center and work function as Cu (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea, b). To our surprise, CO\u003csub\u003e2\u003c/sub\u003e electrolysis on these Cu-like AuAgPd alloys revealed no generation of any C\u003csub\u003e2+\u003c/sub\u003e products at all potentials (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec and Supplementary Fig.\u0026nbsp;19). In contrast to the general opinion, this indicates a limitation of the descriptor model to explain Cu\u0026rsquo;s unique ability for C-C coupling.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo investigate the reason for this shortcoming, we turned to a computational study of the alloy systems. For this, we considered the key elementary reaction steps that are currently thought to be limiting the formation of C\u003csub\u003e2+\u003c/sub\u003e products,\u003csup\u003e40,41\u003c/sup\u003e CO\u003csub\u003e2\u003c/sub\u003e adsorption, CO\u003csub\u003e2\u003c/sub\u003e reduction to *CO, second CO\u003csub\u003e2\u003c/sub\u003e adsorption, reduction of the second *CO\u003csub\u003e2\u003c/sub\u003e to *CO, and finally coupling of the two *CO adsorbates:\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:\\text{C}{\\text{O}}_{2}\\left(\\text{g}\\right){+}^{\\text{*}}\\rightleftharpoons\\:{}_{\\:}{}^{\\text{*}}\\text{C}{\\text{O}}_{2}\\)\u003c/span\u003e \u003c/span\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:{}_{\\:}{}^{\\text{*}}\\text{C}{\\text{O}}_{2}+2{\\text{H}}^{+}+2{\\text{e}}^{-}\\rightleftharpoons\\:{}_{\\:}{}^{\\text{*}}\\text{C}\\text{O}+{\\text{H}}_{2}\\text{O}\\left(\\text{l}\\right)\\)\u003c/span\u003e \u003c/span\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:{}_{\\:}{}^{\\text{*}}\\text{C}\\text{O}\\rightleftharpoons\\:\\text{C}\\text{O}\\left(\\text{g}\\right)\\)\u003c/span\u003e \u003c/span\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:{}_{\\:}{}^{\\text{*}}\\text{C}\\text{O}+\\text{C}{\\text{O}}_{2}\\left(\\text{g}\\right)\\rightleftharpoons\\:{}_{\\:}{}^{\\text{*}}\\text{C}{\\text{O}}_{\\:}+{}_{\\:}{}^{\\text{*}}\\text{C}{\\text{O}}_{2}\\)\u003c/span\u003e \u003c/span\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:{}_{\\:}{}^{\\text{*}}\\text{C}{\\text{O}}_{\\:}+{}_{\\:}{}^{\\text{*}}\\text{C}{\\text{O}}_{2}+2{\\text{H}}^{+}+2{\\text{e}}^{-}\\rightleftharpoons\\:2{}_{\\:}{}^{\\text{*}}\\text{C}\\text{O}+{\\text{H}}_{2}\\text{O}\\left(\\text{l}\\right)\\)\u003c/span\u003e \u003c/span\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:2{}_{\\:}{}^{\\text{*}}\\text{C}{\\text{O}}_{\\:}\\rightleftharpoons\\:{}_{\\:}{}^{\\text{*}}\\text{O}\\text{C}\\text{C}\\text{O}\\)\u003c/span\u003e \u003c/span\u003e \u003cb\u003e(1)\u003c/b\u003e \u003c/p\u003e \u003cp\u003eDue to the strong dependence of CO\u003csub\u003e2\u003c/sub\u003e adsorption and CO coupling on the interfacial electric field, it is important to perform all DFT calculations under a realistic surface charge density/electric field condition. We utilized an implicit solvation model to add excess electrons to the system which are compensated by realistically distributed countercharges in the electrolyte. The constant charge scheme is the same as outlined in our previous work\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e,\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e and is described in more detail in the Supplementary Materials. The complexity of the alloys makes it practically impossible to digitally reconstruct the atomistic structure of the experimental system. Thus, instead, we decided to construct simplified model surfaces to mimic the limiting edges of the surface composition space, and thus the limiting factors for different compositions. In practice, this corresponds to single-atom alloys constructed from the three metals. The intermediates were positioned close to the dopant metal on the most stable adsorption site to explore the variations induced in the free energy diagram from the electronic structure perturbation. In line with the XRD analyses, we chose the face-centered cubic crystal structure and the (100) facet exposure that is known to be the active of the abundant sites for C-C coupling\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e (Supplementary Fig.\u0026nbsp;5, 20).\u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e shows the thermodynamic barriers for three selected intermediate steps for binary systems at \u0026minus;1.5 V\u003csub\u003eSHE\u003c/sub\u003e. Based on the data in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea, it can be noted that Cu demonstrates CO\u003csub\u003e2\u003c/sub\u003e adsorption behavior like that of Ag alloys, while Au and Pd alloys exhibit notably lower barriers. However, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb, CO is expected to be desorbed immediately after the reduction of *CO\u003csub\u003e2\u003c/sub\u003e in the systems without Pd. On the other hand, for the systems with Pd, the formation of *OCCO is very unfavored due to strong CO binding, as seen in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec. This shows that the strong ability of Cu to form C\u003csub\u003e2+\u003c/sub\u003e products comes from the fact that it avoids all three deactivation pathways, by providing facile CO\u003csub\u003e2\u003c/sub\u003e adsorption, avoiding *CO desorption, and enabling *OCCO formation. From the investigations of the dilute limits, we conclude that Pd is needed for providing sufficient *CO coverage, and Au/Ag for reducing the C-C coupling barrier. A homogeneous mixture of those metals would be beneficial but might be then limited by the adsorption of the second CO\u003csub\u003e2\u003c/sub\u003e molecule that has to approach closely the Pd site. These results highlight the complications of designing heterogeneous metal alloy surfaces using simple descriptor models in particular if two active sites are involved in the reaction process like for the formation of C\u003csub\u003e2+\u003c/sub\u003e products.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Conclusions","content":"\u003cp\u003eIn summary, we employed the co-sputtering technique to fabricate pure metals, binary alloys, and ternary alloys using Au, Ag, and Pd metals. This technique enables continuous variation of the d-band center and work function of the alloys, measured through UPS, corresponding to altered CO adsorption energies and PZC. We then investigated the performance of the descriptors in predicting electrochemical performance, i.e. CO\u003csub\u003e2\u003c/sub\u003e reduction rate to CO, HCOO\u003csup\u003e\u0026minus;\u003c/sup\u003e, and H\u003csub\u003e2\u003c/sub\u003e. From this, the d-band center was found to be the primary descriptor for CO production, while both the d-band center and work function are descriptors for HCOO\u003csup\u003e\u0026minus;\u003c/sup\u003e production, showing a volcano-like 2-dimensional activity plot. Furthermore, by designing a ternary alloy mimicking the descriptor values of copper, we did not find any C\u003csub\u003e2\u003c/sub\u003e product formation suggesting limitations of the descriptor model. Through charge-dependent DFT calculations, we reveal that the heterogeneity of the adsorption energies across the alloy surface results in different deactivation pathways which prevent C-C coupling from happening. Our research validates the importance of these descriptors for predicting the performance of CO\u003csub\u003e2\u003c/sub\u003eRR but also highlights the challenges in particular when applying it to multi-site reaction pathways like C\u003csub\u003e2+\u003c/sub\u003e formation.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eCatalyst preparation\u003c/h2\u003e \u003cp\u003eEvery catalyst including Au, Ag, Pd, Cu, AuPd alloys, AgPd alloys, AuAg alloys, and AuAgPd alloys is deposited by sputtering on PTFE membrane (Aspired Laminated, Hydrophobic, Polypropylene Backer, 0.45 Micron, STERLITECH) The Au (99.99%, iTasco), Ag (99.99%, iTasco), Pd (99.99%, iTasco) and Cu (99.998%, iTasco) sputtering targets were powered by direct current (DC). Deposition time was 5 min for every catalyst and argon (Ar) gas was purged on the chamber of sputter with 5 mTorr during the deposition time. The composition of each alloy was controlled by the power of each gun.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eMaterial characterization\u003c/h3\u003e\n\u003cp\u003eSEM images were obtained with Magellan400 from FEI. TEM, STEM and TEM-EDS image were obtained with Talos F200X from FEI. XRD patterns were obtained with SmartLab from RIGAKU. XPS and UPS were measured with Axis-Supra from Kratos.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eElectrochemical CO\u003csub\u003e2\u003c/sub\u003eRR in a flow reactor\u003c/h2\u003e \u003cp\u003eAll measurements for electrochemical CO\u003csub\u003e2\u003c/sub\u003eRR were conducted in a gas diffusion electrode-based flow cell, consisting of two electrolyte chambers and one gas chamber. An anion exchange membrane, Selemion membrane, separated the catholyte chamber from the anolyte chamber. An Ag/AgCl reference electrode (Saturated KCl, RE-1CP) was positioned near the cathode electrode, and a Nickel Iron Copper Molybdenum foil (Alfa Aesar) served as the anode. Both the catholyte and anolyte were prepared with 1 M KHCO\u003csub\u003e3\u003c/sub\u003e. Constant potential was applied to the cathode through an electrochemical workstation (VSP/VMP3B-5, BioLogic). Prior to each measurement, a 5 minute pre-reduction step at 40 mA cm\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e was performed.\u003c/p\u003e \u003cp\u003eFor potentials of \u0026minus;0.6 V\u003csub\u003eRHE\u003c/sub\u003e, \u0026minus;0.7 V\u003csub\u003eRHE\u003c/sub\u003e, and \u0026minus;0.8 V\u003csub\u003eRHE\u003c/sub\u003e, the constant potential was applied to the cathode for 40 minutes. Uncompensated resistance was measured and automatically compensated by potential-stat. The catholyte was continuously circulated at a rate of 13.24 ml/min using a peristaltic pump (Major Science), and CO\u003csub\u003e2\u003c/sub\u003e gas was introduced at 20 sccm through a mass flow controller (MKP) during the reaction. The outlet gas flow rate was monitored by an electronic flow meter (Agilent). Gas products from CO\u003csub\u003e2\u003c/sub\u003eRR were analyzed using gas chromatography (3000 Micro GC, INFICON), while liquid products were assessed using high performance liquid chromatography (YL9100) and headspace GC (YL Instruments).\u003c/p\u003e \u003cp\u003eThe Faradaic efficiency (FE) for a specific product was determined by the formula:\u003c/p\u003e \u003cp\u003eFE\u003csub\u003ej\u003c/sub\u003e (%) = \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\frac{{\\text{n}}_{\\text{j}}\\text{ ∙ }{\\text{z}}_{\\text{j}}\\text{ ∙ }\\text{F}}{\\text{Q}}\\)\u003c/span\u003e\u003c/span\u003e \u003cb\u003e(2)\u003c/b\u003e\u003c/p\u003e \u003cp\u003ewhere n\u003csub\u003ej\u003c/sub\u003e is the moles of the specific product measured from GC and LC; z\u003csub\u003ej\u003c/sub\u003e is the number of electrons required to produce the product; \u003cem\u003eF\u003c/em\u003e is the Faraday\u0026rsquo;s constant, and Q is the total charge applied during the measurement. The partial current density for a specific product was calculated by multiplying the FE of the product by the total current density.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eDFT calculation\u003c/h3\u003e\n\u003cp\u003eUsing Vienna Ab-initio Simulation Package (VASP) version 6.4.1 software,\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e we performed the density functional theory (DFT) calculations with the projector-augmented wave (PAW)\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e pseudopotentials provided along with VASP. The standard version\u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e of the pseudopotentials generated by VASP has been used for Au, Pd, Ag, C, and O. The calculations were carried out with the Bayesian error estimation functional with van der Waals correlation (BEEF-vdW)\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e to consider the precise description of the surficial systems. Additionally, implicit solvation with planar counter charge was included using VASPsol\u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e,\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e to simulate the experimental conditions of solvent, pH, and electric potential. The PZC of each surficial system, defined with the Fermi level from the planar average potential along the c direction of the structure and the vacuum level, was evaluated using the following equation:\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:{U}_{\\text{P}\\text{Z}\\text{C}}=-\\left({E}_{\\text{F}}+{E}_{\\text{S}\\text{H}\\text{I}\\text{F}\\text{T}}\\right)+{\\Delta\\:}{U}_{\\text{S}\\text{H}\\text{E}}^{\\text{e}\\text{x}\\text{p}}=-\\left({E}_{\\text{F}}+{E}_{\\text{S}\\text{H}\\text{I}\\text{F}\\text{T}}\\right)-4.6$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e3\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{E}_{\\text{F}}\\)\u003c/span\u003e\u003c/span\u003e is the Fermi level, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{E}_{\\text{S}\\text{H}\\text{I}\\text{F}\\text{T}}\\)\u003c/span\u003e\u003c/span\u003e is a correction constant for the reference electrostatic potential, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\Delta\\:}{U}_{\\text{S}\\text{H}\\text{E}}^{\\text{e}\\text{x}\\text{p}}\\)\u003c/span\u003e\u003c/span\u003e is a constant shift in between the experimental and the computed PZC.\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e Bulk structures were generated using ASE\u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e and relaxed with a kinetic energy cutoff for the plane waves of 600 eV and the smallest allowed spacing in the k-grid of 0.1 \u0026Aring;\u003csup\u003e\u0026minus;\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e with gamma-centered mesh. For the surface slabs, symmetric 4-layered 3x3 (100) surfaces were built from the bulk crystal using the CatKit Python package from SUNCAT (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/SUNCAT-Center/CatKit\u003c/span\u003e\u003cspan address=\"https://github.com/SUNCAT-Center/CatKit\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) and relaxed with a kinetic energy cutoff of 400 eV and a smallest allowed spacing in the k-grid of 0.1 \u0026Aring;\u003csup\u003e\u0026minus;\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e with gamma-centered mesh. The minimum length of the lattice parameter \u003cem\u003ec\u003c/em\u003e vertical to the \u003cem\u003ea\u003c/em\u003e and \u003cem\u003eb\u003c/em\u003e lattice plane was 25 \u0026Aring;. Each adsorbate was positioned symmetrically at both sides of the surface near the dopant for the adsorption energy calculation. All calculations were performed using Gaussian smearing with a width of 0.05 eV. Geometry relaxations were performed until the norm of each atom\u0026rsquo;s forces reached 0.05 eV/\u0026Aring;. For a few systems, an additional net charge was induced to obtain a reliable adsorption structure. After the relaxation, DFT energy was calculated by removing the charge within the constraint of the structure.\u003c/p\u003e \u003cp\u003eTo calculate the grand-canonical potential, charge-dependent DFT energy \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{{\\Delta\\:}}_{i}E\\left(\\sigma\\:\\left(U\\right)\\right)\\)\u003c/span\u003e\u003c/span\u003e must be evaluated with the following second-order approximated equation:\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$$\\:{{\\Delta\\:}}_{i}E\\left(\\sigma\\:\\left(U\\right)\\right)={{\\Delta\\:}}_{i}{E}_{0}+{{\\Delta\\:}}_{i}a\\cdot\\:\\sigma\\:+{{\\Delta\\:}}_{i}b\\cdot\\:{\\sigma\\:}^{2}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e4\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{{\\Delta\\:}}_{i}{E}_{0}\\)\u003c/span\u003e\u003c/span\u003e is the DFT energy at the zero charge, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{{\\Delta\\:}}_{i}a\\)\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{{\\Delta\\:}}_{i}b\\)\u003c/span\u003e\u003c/span\u003e are the coefficients for the charge dependence (Supplementary Table\u0026nbsp;5). This expression is inserted into the grand-canonical free energy:\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:{{\\Delta\\:}}_{i}{\\Omega\\:}\\left(U,\\text{p}\\text{H}\\right)={{\\Delta\\:}}_{i}E\\left(\\sigma\\:\\left(U\\right)\\right)+{{\\Delta\\:}}_{i}{E}_{\\text{T},\\text{S},\\text{Z}\\text{P}\\text{E}}^{^\\circ\\:}+0.0592\\cdot\\:\\text{p}\\text{H}+\\text{e}U\\)\u003c/span\u003e \u003c/span\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:\\sigma\\:\\left(U\\right)=C\\cdot\\:(U-{U}_{\\text{P}\\text{Z}\\text{C}})\\)\u003c/span\u003e \u003c/span\u003e \u003cb\u003e(5)\u003c/b\u003e \u003c/p\u003e \u003cp\u003ewith the DFT energy \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{{\\Delta\\:}}_{i}E\\left(\\sigma\\:\\left(U\\right)\\right)\\)\u003c/span\u003e\u003c/span\u003e, the surface charge density \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\sigma\\:\\)\u003c/span\u003e\u003c/span\u003e, the electrode potential vs. SHE \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:U\\)\u003c/span\u003e\u003c/span\u003e, thermal and zero-point energy (ZPE) correction term \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{{\\Delta\\:}}_{i}{E}_{\\text{T},\\text{S},\\text{Z}\\text{P}\\text{E}}^{^\\circ\\:}\\)\u003c/span\u003e\u003c/span\u003e, the capacitance of the system \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:C\\)\u003c/span\u003e\u003c/span\u003e and the potential of zero charge (PZC) \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{U}_{\\text{P}\\text{Z}\\text{C}}\\)\u003c/span\u003e\u003c/span\u003e. The surface charge density is thus determined from the PZC and the capacitance, for which we took the experimental value of 20 \u0026micro;F/cm\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e as an approximate metal-independent value. Also, to reduce the computational cost, we approximated the thermal and zero-point energy correction to be the same for all systems as Cu. We direct readers to the method section for further details. Additionally, +\u0026thinsp;0.33 eV and +\u0026thinsp;0.09 eV of gas phase correction for DFT energy has been added for CO\u003csub\u003e2\u003c/sub\u003e and H\u003csub\u003e2\u003c/sub\u003e.\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e General examples of free energy diagrams via the approach above are presented (Supplementary Fig.\u0026nbsp;21).\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eCompeting Interests\u003c/h2\u003e \u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eAuthor contributions\u003c/h2\u003e \u003cp\u003eB.K., S.H., S.R., and J.O. wrote the manuscript. B.K. performed the experiments and analyzed the data. S.H. and S.R. contributed DFT calculation. S.W. edited and supported the manuscripts. J.O. design the experiments and supervised the research, data analysis. All authors contributed to this project and have given approval to the final version of the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgements\u003c/h2\u003e \u003cp\u003eThis research was supported by a grant from the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (grant no. 2021R1A2C3007280, RS-2024-00435493). S.R. additionally acknowledges financial support from the NRF, funded by the Ministry of Science and ICT (grant no. 2021R1C1C1008776). We also acknowledge the support of the National Supercomputing Center of Korea Institute of Science and Technology Information (KISTI) with super-computing resources, including technical support (no. KSC-2023-CRE-0468).\u003c/p\u003e\n\u003ch3\u003eData availability\u003c/h3\u003e\n\u003cp\u003eSource data to generate figures and tables are available from the corresponding authors.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eXu, Q. \u003cem\u003eet al.\u003c/em\u003e Descriptor for Hydrogen Evolution Catalysts Based on the Bulk Band Structure Effect. ACS Catalysis 10, 5042\u0026ndash;5048 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1021/acscatal.9b05539\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1021/acscatal.9b05539\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim, B. \u003cem\u003eet al.\u003c/em\u003e Trace-Level Cobalt Dopants Enhance CO\u003csub\u003e2\u003c/sub\u003e Electroreduction and Ethylene Formation on Copper. ACS Energy Letters 8, 3356\u0026ndash;3364 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1021/acsenergylett.3c00418\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1021/acsenergylett.3c00418\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHong, S. \u003cem\u003eet al.\u003c/em\u003e Tuning the C1/C2 Selectivity of Electrochemical CO\u003csub\u003e2\u003c/sub\u003e Reduction on Cu\u0026ndash;CeO2 Nanorods by Oxidation State Control. Advanced Materials 35, 2208996 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:https://doi.org/10.1002/adma.202208996\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1002/adma.202208996\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim, B. \u003cem\u003eet al.\u003c/em\u003e Over a 15.9% Solar-to-CO Conversion from Dilute CO\u003csub\u003e2\u003c/sub\u003e Streams Catalyzed by Gold Nanoclusters Exhibiting a High CO2 Binding Affinity. ACS Energy Letters 5, 749\u0026ndash;757 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1021/acsenergylett.9b02511\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1021/acsenergylett.9b02511\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeng, H. \u003cem\u003eet al.\u003c/em\u003e The role of atomic carbon in directing electrochemical CO\u003csub\u003e2\u003c/sub\u003e reduction to multicarbon products. Energy \u0026amp; Environmental Science 14, 473\u0026ndash;482 (2021). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1039/D0EE02826F\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1039/D0EE02826F\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRinge, S. The importance of a charge transfer descriptor for screening potential CO\u003csub\u003e2\u003c/sub\u003e reduction electrocatalysts. Nature Communications 14, 2598 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/s41467-023-37929-4\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/s41467-023-37929-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBagger, A., Ju, W., Varela, A. S., Strasser, P. \u0026amp; Rossmeisl, J. Electrochemical CO\u003csub\u003e2\u003c/sub\u003e Reduction: A Classification Problem. \u003cem\u003eChemPhysChem\u003c/em\u003e 18, 3266\u0026ndash;3273 (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:https://doi.org/10.1002/cphc.201700736\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1002/cphc.201700736\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGao, W. \u003cem\u003eet al.\u003c/em\u003e CO Binding Energy is an Incomplete Descriptor of Cu-Based Catalysts for the Electrochemical CO\u003csub\u003e2\u003c/sub\u003e Reduction Reaction. \u003cem\u003eAngewandte Chemie International Edition\u003c/em\u003e 62, e202313798 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:https://doi.org/10.1002/anie.202313798\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1002/anie.202313798\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSeh, Z. W. \u003cem\u003eet al.\u003c/em\u003e Combining theory and experiment in electrocatalysis: Insights into materials design. Science 355, eaad4998 (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:doi:10.1126/science.aad4998\u003c/span\u003e\u003cspan address=\"https://doi.org:doi:10.1126/science.aad4998\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCao, L. Recent advances in the application of machine-learning algorithms to predict adsorption energies. Trends in Chemistry 4, 347\u0026ndash;360 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org\u003c/span\u003e\u003cspan address=\"https://doi.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.trechm.2022.01.012\u003c/span\u003e\u003cspan address=\"10.1016/j.trechm.2022.01.012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLan, J. \u003cem\u003eet al.\u003c/em\u003e AdsorbML: a leap in efficiency for adsorption energy calculations using generalizable machine learning potentials. npj Computational Materials 9, 172 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/s41524-023-01121-5\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/s41524-023-01121-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKuhl, K. P. \u003cem\u003eet al.\u003c/em\u003e Electrocatalytic Conversion of Carbon Dioxide to Methane and Methanol on Transition Metal Surfaces. Journal of the American Chemical Society 136, 14107\u0026ndash;14113 (2014). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1021/ja505791r\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1021/ja505791r\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHori, Y., Wakebe, H., Tsukamoto, T. \u0026amp; Koga, O. Electrocatalytic process of CO selectivity in electrochemical reduction of CO\u003csub\u003e2\u003c/sub\u003e at metal electrodes in aqueous media. Electrochimica Acta 39, 1833\u0026ndash;1839 (1994). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:https://doi.org/10.1016/0013-4686(94)85172-7\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/0013-4686(94)85172-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChung, D. Y. \u003cem\u003eet al.\u003c/em\u003e Inhibition of CO poisoning on Pt catalyst coupled with the reduction of toxic hexavalent chromium in a dual-functional fuel cell. Scientific Reports 4, 7450 (2014). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/srep07450\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/srep07450\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHori, Y., Takahashi, R., Yoshinami, Y. \u0026amp; Murata, A. Electrochemical Reduction of CO at a Copper Electrode. The Journal of Physical Chemistry B 101, 7075\u0026ndash;7081 (1997). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1021/jp970284i\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1021/jp970284i\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eN\u0026oslash;rskov, J. K., Abild-Pedersen, F., Studt, F. \u0026amp; Bligaard, T. Density functional theory in surface chemistry and catalysis. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e 108, 937\u0026ndash;943 (2011). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1073/pnas.1006652108\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1073/pnas.1006652108\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHammer, B., Morikawa, Y. \u0026amp; N\u0026oslash;rskov, J. K. CO Chemisorption at Metal Surfaces and Overlayers. Physical Review Letters 76, 2141\u0026ndash;2144 (1996). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1103/PhysRevLett.76.2141\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1103/PhysRevLett.76.2141\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDickens, C. F., Montoya, J. H., Kulkarni, A. R., Bajdich, M. \u0026amp; N\u0026oslash;rskov, J. K. An electronic structure descriptor for oxygen reactivity at metal and metal-oxide surfaces. Surface Science 681, 122\u0026ndash;129 (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:https://doi.org/10.1016/j.susc.2018.11.019\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/j.susc.2018.11.019\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTrasatti, S. Work function, electronegativity, and electrochemical behaviour of metals: II. Potentials of zero charge and \u0026ldquo;electrochemical\u0026rdquo; work functions. Journal of Electroanalytical Chemistry and Interfacial Electrochemistry 33, 351\u0026ndash;378 (1971). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:https://doi.org/10.1016/S0022-0728(71)80123-7\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/S0022-0728(71)80123-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJackson, C., Smith, G., Russell, A. E., Levecque, P. \u0026amp; Kramer, D. Electronic metal-support interactions in vacuum vs. electrolyte. Nature Communications 11, 1470 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/s41467-020-15307-8\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/s41467-020-15307-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMontoya, J. H., Shi, C., Chan, K. \u0026amp; N\u0026oslash;rskov, J. K. Theoretical Insights into a CO Dimerization Mechanism in CO\u003csub\u003e2\u003c/sub\u003e Electroreduction. The Journal of Physical Chemistry Letters 6, 2032\u0026ndash;2037 (2015). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1021/acs.jpclett.5b00722\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1021/acs.jpclett.5b00722\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKarmodak, N., Vijay, S., Kastlunger, G. \u0026amp; Chan, K. Computational Screening of Single and Di-Atom Catalysts for Electrochemical CO\u003csub\u003e2\u003c/sub\u003e Reduction. ACS Catalysis 12, 4818\u0026ndash;4824 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1021/acscatal.1c05750\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1021/acscatal.1c05750\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSandberg, R. B., Montoya, J. H., Chan, K. \u0026amp; N\u0026oslash;rskov, J. K. CO-CO coupling on Cu facets: Coverage, strain and field effects. Surface Science 654, 56\u0026ndash;62 (2016). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:https://doi.org/10.1016/j.susc.2016.08.006\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/j.susc.2016.08.006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRinge, S. \u003cem\u003eet al.\u003c/em\u003e Understanding cation effects in electrochemical CO\u003csub\u003e2\u003c/sub\u003e reduction. Energy \u0026amp; Environmental Science 12, 3001\u0026ndash;3014 (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1039/C9EE01341E\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1039/C9EE01341E\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRinge, S. \u003cem\u003eet al.\u003c/em\u003e Double layer charging driven carbon dioxide adsorption limits the rate of electrochemical carbon dioxide reduction on Gold. Nature Communications 11, 33 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/s41467-019-13777-z\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/s41467-019-13777-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, L. D., Urushihara, M., Chan, K. \u0026amp; N\u0026oslash;rskov, J. K. Electric Field Effects in Electrochemical CO\u003csub\u003e2\u003c/sub\u003e Reduction. ACS Catalysis 6, 7133\u0026ndash;7139 (2016). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1021/acscatal.6b02299\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1021/acscatal.6b02299\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDong, W. J. \u003cem\u003eet al.\u003c/em\u003e Electric-field-driven electrochemical CO\u003csub\u003e2\u003c/sub\u003e reduction of sharpened Sn/Cu catalysts. Applied Surface Science 565, 150460 (2021). https://doi.org:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.apsusc.2021.150460\u003c/span\u003e\u003cspan address=\"10.1016/j.apsusc.2021.150460\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTan, Y. C. \u003cem\u003eet al.\u003c/em\u003e Pitfalls and Protocols: Evaluating Catalysts for CO\u003csub\u003e2\u003c/sub\u003e Reduction in Electrolyzers Based on Gas Diffusion Electrodes. ACS Energy Letters 7, 2012\u0026ndash;2023 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1021/acsenergylett.2c00763\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1021/acsenergylett.2c00763\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGhosh, G., Kantner, C. \u0026amp; Olson, G. B. Thermodynamic modeling of the Pd-X (X\u0026thinsp;=\u0026thinsp;Ag, Co, Fe, Ni) systems. Journal of Phase Equilibria 20, 295\u0026ndash;308 (1999). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1361/105497199770335811\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1361/105497199770335811\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOkamoto, H. \u0026amp; Massalski, T. B. The Au\u0026thinsp;\u0026ndash;\u0026thinsp;Pd (Gold-Palladium) system. Bulletin of Alloy Phase Diagrams 6, 229\u0026ndash;235 (1985). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1007/BF02880404\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1007/BF02880404\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHansen, H. A., Varley, J. B., Peterson, A. A. \u0026amp; N\u0026oslash;rskov, J. K. Understanding Trends in the Electrocatalytic Activity of Metals and Enzymes for CO\u003csub\u003e2\u003c/sub\u003e Reduction to CO. The Journal of Physical Chemistry Letters 4, 388\u0026ndash;392 (2013). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1021/jz3021155\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1021/jz3021155\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRinge, S. Cation effects on electrocatalytic reduction processes at the example of the hydrogen evolution reaction. Current Opinion in Electrochemistry 39, 101268 (2023). https://doi.org:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.coelec.2023.101268\u003c/span\u003e\u003cspan address=\"10.1016/j.coelec.2023.101268\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee, M.-Y., Ringe, S., Kim, H., Kang, S. \u0026amp; Kwon, Y. Electric Field Mediated Selectivity Switching of Electrochemical CO\u003csub\u003e2\u003c/sub\u003e Reduction from Formate to CO on Carbon Supported Sn. ACS Energy Letters 5, 2987\u0026ndash;2994 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1021/acsenergylett.0c01387\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1021/acsenergylett.0c01387\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKortlever, R., Shen, J., Schouten, K. J. P., Calle-Vallejo, F. \u0026amp; Koper, M. T. M. Catalysts and Reaction Pathways for the Electrochemical Reduction of Carbon Dioxide. The Journal of Physical Chemistry Letters 6, 4073\u0026ndash;4082 (2015). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1021/acs.jpclett.5b01559\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1021/acs.jpclett.5b01559\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun, Z., Ma, T., Tao, H., Fan, Q. \u0026amp; Han, B. Fundamentals and Challenges of Electrochemical CO\u003csub\u003e2\u003c/sub\u003e Reduction Using Two-Dimensional Materials. Chem 3, 560\u0026ndash;587 (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:https://doi.org/10.1016/j.chempr.2017.09.009\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/j.chempr.2017.09.009\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZou, J., Liang, G., Lee, C.-Y. \u0026amp; Wallace, G. G. Progress and perspectives for electrochemical CO\u003csub\u003e2\u003c/sub\u003e reduction to formate. Materials Today Energy 38, 101433 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:https://doi.org/10.1016/j.mtener.2023.101433\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/j.mtener.2023.101433\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKuhl, K. P., Cave, E. R., Abram, D. N. \u0026amp; Jaramillo, T. F. New insights into the electrochemical reduction of carbon dioxide on metallic copper surfaces. Energy \u0026amp; Environmental Science 5, 7050\u0026ndash;7059 (2012). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1039/C2EE21234J\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1039/C2EE21234J\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMa, M. \u003cem\u003eet al.\u003c/em\u003e Electrochemical reduction of CO\u003csub\u003e2\u003c/sub\u003e on compositionally variant Au-Pt bimetallic thin films. Nano Energy 42, 51\u0026ndash;57 (2017). https://doi.org:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.nanoen.2017.09.043\u003c/span\u003e\u003cspan address=\"10.1016/j.nanoen.2017.09.043\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, K. \u003cem\u003eet al.\u003c/em\u003e Electronic Effects Determine the Selectivity of Planar Au\u0026ndash;Cu Bimetallic Thin Films for Electrochemical CO\u003csub\u003e2\u003c/sub\u003e Reduction. ACS Applied Materials \u0026amp; Interfaces 11, 16546\u0026ndash;16555 (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1021/acsami.9b01553\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1021/acsami.9b01553\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKastlunger, G. \u003cem\u003eet al.\u003c/em\u003e Using pH Dependence to Understand Mechanisms in Electrochemical CO Reduction. ACS Catalysis 12, 4344\u0026ndash;4357 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1021/acscatal.1c05520\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1021/acscatal.1c05520\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, X. \u003cem\u003eet al.\u003c/em\u003e pH effects on the electrochemical reduction of CO\u003csub\u003e2\u003c/sub\u003e towards C2 products on stepped copper. Nature Communications 10, 32 (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/s41467-018-07970-9\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/s41467-018-07970-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim, D. H. \u003cem\u003eet al.\u003c/em\u003e Selective electrochemical reduction of nitric oxide to hydroxylamine by atomically dispersed iron catalyst. Nature Communications 12, 1856 (2021). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/s41467-021-22147-7\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/s41467-021-22147-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHori, Y., Takahashi, I., Koga, O. \u0026amp; Hoshi, N. Electrochemical reduction of carbon dioxide at various series of copper single crystal electrodes. Journal of Molecular Catalysis A: Chemical 199, 39\u0026ndash;47 (2003). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:https://doi.org/10.1016/S1381-1169(03)00016-5\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/S1381-1169(03)00016-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKresse, G. \u0026amp; Furthm\u0026uuml;ller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Physical Review B 54, 11169\u0026ndash;11186 (1996). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1103/PhysRevB.54.11169\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1103/PhysRevB.54.11169\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBl\u0026ouml;chl, P. E. Projector augmented-wave method. Physical Review B 50, 17953\u0026ndash;17979 (1994). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1103/PhysRevB.50.17953\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1103/PhysRevB.50.17953\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKresse, G. \u0026amp; Joubert, D. From ultrasoft pseudopotentials to the projector augmented-wave method. Physical Review B 59, 1758\u0026ndash;1775 (1999). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1103/PhysRevB.59.1758\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1103/PhysRevB.59.1758\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWellendorff, J. \u003cem\u003eet al.\u003c/em\u003e Density functionals for surface science: Exchange-correlation model development with Bayesian error estimation. Physical Review B 85, 235149 (2012). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1103/PhysRevB.85.235149\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1103/PhysRevB.85.235149\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMathew, K., Sundararaman, R., Letchworth-Weaver, K., Arias, T. A. \u0026amp; Hennig, R. G. Implicit solvation model for density-functional study of nanocrystal surfaces and reaction pathways. The Journal of Chemical Physics 140 (2014). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1063/1.4865107\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1063/1.4865107\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMathew, K., Kolluru, V. S. C., Mula, S., Steinmann, S. N. \u0026amp; Hennig, R. G. Implicit self-consistent electrolyte model in plane-wave density-functional theory. The Journal of Chemical Physics 151 (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1063/1.5132354\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1063/1.5132354\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHjorth Larsen, A. \u003cem\u003eet al.\u003c/em\u003e The atomic simulation environment\u0026mdash;a Python library for working with atoms. Journal of Physics: Condensed Matter 29, 273002 (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1088/1361-648X/aa680e\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1088/1361-648X/aa680e\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStudt, F., Abild-Pedersen, F., Varley, J. B. \u0026amp; N\u0026oslash;rskov, J. K. CO and CO\u003csub\u003e2\u003c/sub\u003e Hydrogenation to Methanol Calculated Using the BEEF-vdW Functional. Catalysis Letters 143, 71\u0026ndash;73 (2013). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1007/s10562-012-0947-5\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1007/s10562-012-0947-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\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-5559232/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5559232/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eElectrocatalysis advances rely on the development of efficient catalysts. Systematic material design hinges on identifying activity and selectivity descriptors. While adsorption energy descriptors have helped predict new materials, they are typically based on pure metals, uncertain of their applicability to complex materials like alloys. Here, we systematically analyze the validity of descriptor models for the electrochemical reduction of CO\u003csub\u003e2\u003c/sub\u003e (CO\u003csub\u003e2\u003c/sub\u003eRR). For this, we prepare gold, silver, and palladium alloys of variable composition and confirm experimentally the continuous variation of the d-band center (i.e. the CO adsorption energy) and work function (i.e. the potential of zero charge). Our results indicate that while the d-band center is the decisive factor for CO production, it, along with the work function, fails to fully explain the production of HCOO\u003csup\u003e−\u003c/sup\u003e and H\u003csub\u003e2\u003c/sub\u003e. Designing a copper-like alloy based on the matching of these descriptor values showed no formation of C\u003csub\u003e2\u003c/sub\u003e products (as commonly expected for copper). This breakdown of the descriptor model is explained from \u003cem\u003efirst-principles\u003c/em\u003e calculations by the heterogeneity of the surface leading to different deactivation pathways for C\u003csub\u003e2\u003c/sub\u003e product formation. Our results highlight the problems in transferring conventional descriptor models to more complex, heterogeneous materials motivating future developments.\u003c/p\u003e","manuscriptTitle":"Peaks and pitfalls of electrocatalytic descriptor models at the example of CO2 reduction","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-01-14 05:07:40","doi":"10.21203/rs.3.rs-5559232/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-catalysis","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"natcatal","sideBox":"Learn more about [Nature Catalysis](http://www.nature.com/natcatal/)","snPcode":"","submissionUrl":"","title":"Nature Catalysis","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Research","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"5aca82d7-ecef-494b-9c92-f75f4d3eea23","owner":[],"postedDate":"January 14th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":42711301,"name":"Physical sciences/Chemistry/Electrochemistry/Electrocatalysis"},{"id":42711302,"name":"Physical sciences/Materials science/Materials for energy and catalysis/Electrocatalysis"}],"tags":[],"updatedAt":"2026-04-14T07:05:29+00:00","versionOfRecord":{"articleIdentity":"rs-5559232","link":"https://doi.org/10.1038/s41929-026-01526-7","journal":{"identity":"nature-catalysis","isVorOnly":false,"title":"Nature Catalysis"},"publishedOn":"2026-04-13 04:00:00","publishedOnDateReadable":"April 13th, 2026"},"versionCreatedAt":"2025-01-14 05:07:40","video":"","vorDoi":"10.1038/s41929-026-01526-7","vorDoiUrl":"https://doi.org/10.1038/s41929-026-01526-7","workflowStages":[]},"version":"v1","identity":"rs-5559232","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5559232","identity":"rs-5559232","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}