Charge transfer across the metal/oxide interface determines the rate of CO2 hydrogenation to methanol over Cu/ZnO catalysts

Charge transfer across the metal/oxide interface determines the rate of CO2 hydrogenation to methanol over Cu/ZnO catalysts | 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 Physical Sciences - Article Charge transfer across the metal/oxide interface determines the rate of CO 2 hydrogenation to methanol over Cu/ZnO catalysts Jakob Christensen, Jonas Boysen, Rasmus Svensson, Niels Nielsen, and 9 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6939552/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Composite nanomaterials are crucial in a wide range of applications with the industrial Cu/ZnO catalyst used to convert CO 2 into methanol being an important example. The reaction rate scales with the Cu surface area 1,2 implying that the rate-limiting part of the reaction occurs on the metal surface. However, the turnover frequency (TOF, rate per Cu surface atom) is one order of magnitude higher for Cu supported on ZnO compared to pure Cu samples 3–6 . This materials synergy, which is responsible for 90% of the catalytic activity, is still poorly understood 7–9 , and the copper in Cu/ZnO is sometimes described 10,11 as an element with properties radically different from the pristine metal. An understanding of this synergy thus has a direct impact on Power-to-X processes to produce methanol for storage of renewable energy and on heterogeneous catalysis in general. Here we establish the mechanism for the hugely important conversion of CO2 into methanol and how it is accelerated on the composite nanomaterial. We show that charge from donor states created by ad-/absorbed H in the ZnO transfers to the metal and distributes across the metal surface. The surface charging influences the bonding of the adsorbates on the metal and lowers the energy barrier for HCOOH dissociation, which we identify as the rate-limiting step in methanol formation. The lower barrier accounts for the order of magnitude increase in the methanol synthesis rate. The charge transfer phenomenon is general in nature and found to be essential for understanding catalytic phenomena and thus for the rational development of improved future catalysts. Physical sciences/Chemistry/Catalysis/Heterogeneous catalysis Physical sciences/Physics/Chemical physics Physical sciences/Chemistry/Catalysis/Catalytic mechanisms Physical sciences/Physics/Condensed-matter physics/Surfaces, interfaces and thin films Figures Figure 1 Figure 2 Figure 3 Figure 4 Full Text Methanol is synthesised from CO/CO 2 /H 2 mixtures at 50-100 bar and 220-280 ℃. 10 The process relies on Cu/ZnO/Al 2 O 3 catalysts with Cu nanoparticles (typically 5-10 nm) dispersed in a matrix of ZnO nanoparticles (the support) with low levels of Al 2 O 3 as a structural stabilizer. 10–13 Methanol is formed from CO 2 and H 2 , 14,15 while CO removes the inhibiting 16 co-produced water by water-gas shift 4 . The reaction rate scales linearly with the Cu surface area (Fig. E1a), but the TOF for Cu/ZnO is one order of magnitude higher than for unsupported Cu (Fig. E1b), and the underlying mechanisms of this synergy and of the reaction steps remain poorly understood. In situ infrared (IR) spectroscopy shows that formate (HCOO) is present on the metallic surface (Fig. E1c) and involved in the reaction (Fig. E1d). When conducting CO 2 hydrogenation with a 9:1 H 2 :D 2 mixture, where DCOO is enriched in the adsorbed formate 17 , the enrichment of methanol molecules with a characteristic C-D bond matches the enrichment of adsorbed DCOO (Fig. 1a), thus proving that methanol is formed from adsorbed formate. The activation of CO 2 to HCOO through an Eley-Rideal mechanism 18 is fast compared to methanol synthesis and can be regarded as quasi equilibrated (Fig. E1e), meaning that the rate-limiting step occurs later in the mechanism. To identify where formate is located on the catalyst surface, we pre-adsorbed DCOO (from DCOOD) on unsupported Cu and Pt and on 10 wt% Pt/ZnO. We conducted a temperature programmed hydrogenation (TPH) of the adsorbed DCOO into CDH 2 OH (Fig. 1b). As unsupported Pt only forms small amounts of methanol (Fig. 1b), the Pt/ZnO system gives the reactivity for metal-assisted hydrogenation of formate on ZnO. The results in Fig. 1b show that even without a synergy between Cu and ZnO, the emergence of methanol formation on Cu occurs at a 55 °C lower temperature than formate on ZnO. This corresponds to an activation energy difference of 0.1 eV that amounts to a 10-fold higher rate from HCOO-Cu compared to HCOO-ZnO at typical reaction temperatures, indicating that reaction on the metal surface is responsible for at least 90% of the turnovers. Further TPH experiments described in the supplementary information confirms that HCOO-ZnO represents the least active species in Cu/ZnO samples. Metal-assisted hydrogenation of formates on ZnO thus represents one type of synergy but this is only responsible for a minority of turnovers on Cu/ZnO and a second and more important synergy creates an accelerated reaction on the metallic surface. Recent computational work 19–21 suggests that formate is converted via formic acid. This is validated by Fig. 1c, which shows that the conversion of HCOO and HCOOH becomes possible at the same temperature when formic acid pulses are injected during TPH of pre-adsorbed formate. Their conversions must consequently share the same rate-limiting step, which must be the later step, namely HCOOH conversion. As the HCOOH conversion is independent of the surrounding gas, the reaction does not revert to HCOO and then hydrogenate to methanol from there (Fig. 1d). To identify the dominant HCOOH conversion pathway, we injected a pulse of DCOOD into a flow of H 2 at 200 ℃ passing over Cu/ZnO/Al 2 O 3 . Fig. 1e shows that this produces CDH 2 OH as the dominant methanol isotope. The major preservation of the C-D bond from the original DCOOD despite the possibilities for H/D scrambling indicates that the conversion of formic acid proceeds through the scission of the (HCO)-(OH) bond into an HCO moiety that is then hydrogenated to methanol. Fig. 1d shows that the methanol yield from a formic acid pulse depends on the formic acid isotope but not on the use of D 2 or H 2 . The independence of D 2 /H 2 suggests that only formic acid is involved in the rate-limiting step, ruling out the previously most studied paths 10,22,23 via H 2 COOH, H 2 COO or COOH, as illustrated in Fig. E2. Consistent with a rate-limiting HCOOH scission the first order H 2 dependence of the reaction also supports that it requires two protonations of CO 2 to pass the rate-limiting step (Fig. E1f). Dosing formaldehyde (CH 2 O) during CO 2 hydrogenation greatly accelerates the reaction, which shows that once protonated HCO and H 2 CO species have been formed, the remaining pathway to methanol is fast (Fig. E1g). The proposed mechanism also explains the autocatalytic behaviour in the presence of methanol 16 as esterification of HCOOH means that a weaker 24,25 (HCO)-(OCH 3 ) bond needs to be broken. We can thus identify the unimolecular (HCO)-(OH) scission into HCO and OH on Cu as the dominant methanol forming pathway, and the key to understanding the Cu-ZnO synergy is to understand this reaction step (Fig. 1f). To study the properties of Cu in the catalyst we conducted X-ray photoelectron spectroscopy (XPS) on a Cu/ZnO/Al 2 O 3 sample (11 nm Cu particles) and a pure Cu reference. Fig. 2a shows that after exposure to 1 bar H 2 at 175 ℃, the Cu surface is negatively charged in Cu/ZnO compared to pure Cu as evident from a downshift in electron binding energy. The measured shifts are substantial given the subtlety of Cu binding energy differences (Cu 0 /Cu + /Cu 2+ only differ by 0.5 eV). When the H 2 -treated Cu/ZnO/Al 2 O 3 sample is heated in vacuum to 250 ℃ the amount of ad-/absorbed hydrogen in the surface region of ZnO phase is halved (Fig. E3a), and Fig. 2a shows that this makes the Cu phase in Cu/ZnO/Al 2 O 3 more similar to pristine Cu metal. This reveals that stored hydrogen on/in the catalyst contributes to the altered state of Cu. Remarkably, this also demonstrates that fundamental properties of a catalyst, such as the charge of the surface atoms, changes dynamically with the conditions. The XPS shifts are also present in an ambient pressure of 14 mbar H 2 at 250 ℃ (Fig. E3b). To understand the reason for the altered nature of a metal in a supported catalyst we prepared and pre-reduced a 2 wt% Pt/ZnO sample, which due to platinum’s nobility can be transferred through air to ex situ analysis by transmission electron microscopy (TEM, Fig. 2b) and XPS (Fig. 2c). The TEM results in Fig. 2b show that the Pt nanoparticles are not alloyed with Zn or decorated by ZnO layers via so-called strong metal support interactions (SMSI). The XPS results in Fig. 2c again show that the metal surface is negatively charged compared to pure Pt. From the XPS-results in Fig. 2a we can also conclude that for metal/ZnO systems, the presence of adsorbed hydrogen plays a major role in this phenomenon. We analysed the hydrogen uptake in 1 bar H 2 at 175 ℃ for a series of Cu/ZnO/Al 2 O 3 samples, and Fig. 2d shows that the ability of ZnO to take up hydrogen increases by orders of magnitude when Cu and ZnO are in contact. TPD was used to distinguish between adsorption on Cu and ZnO (see Fig. E3c). IR spectroscopy during H 2 /D 2 exchange shows that uptake rate of atomic hydrogen is kinetically assisted by spillover 26 from the metal (Fig. E3d), but the increased uptake for Cu/ZnO samples in Fig. 2d is a thermodynamic effect, as evident from long-term tests to establish the equilibrium coverage (Fig. E3e). Density functional theory (DFT) calculations predict that heterolytic (H 2 →H - +H + ) adsorption is associated with a small adsorption energy (E ads (2H) = -0.05 eV) on a ZnO( ) surface (Fig. E3e). By contrast, the adsorption of hydrogen on ZnO( ) next to a supported Cu particle prefers homolytic adsorption (H 2 →2H + +2e - ) with a high energy of E ads (2H) = -1.84 eV (Fig. 2e). Owing to the strong bonding of H to ZnO in the presence of Cu, H is stable in vacuum at room temperature (Fig 2a) and at elevated temperatures in the presence of H 2 (Fig. 2d), but not at elevated temperatures in vacuum (Fig. E3a). The DFT calculations in Fig. 2e show that adsorption of H on a ZnO( ) surface containing a cluster of Pt or Cu is associated with H + /e - charge separation and an electron transfer to the metal. The electron transfer to the metal cluster is shown to increase with the amount of adsorbed hydrogen (Fig. 2e). Hydrogen is a shallow donor in ZnO 27 , and the energy diagram 27–29 shows that the strong bonding of H for Cu/ZnO is the result of the exothermic electron transfer to the metal (Fig. 2f). The DFT calculations show that the stabilisation of adsorbed H is most pronounced for adsorption at the interface. However, stabilisation is present also for H adsorbed further away from the metal particle and H absorption into the bulk of ZnO close to the metal particles (Fig. E3f). Analyses in the supporting information show that different metal cluster sizes or ZnO terminations and the existence of CuZn alloying do not change the conclusions regarding the charge transfer. The analyses also show that although a minor charge transfer between adjacent phases can occur without the presence of donors, it is the presence of donors, such as hydrogen in ZnO, that mainly determines the charge transfer. Oxygen vacancies can also contribute to charge transfer 29 , but have a lower driving force for this than hydrogen (Fig. 2f). These results illustrate that H + /e - charge separation occurs between connected solids as previously observed across thin oxide films 30 and at metal/liquid interfaces 31–33 . Continuum modelling of a hemispherical particle biased by 1 V relative to the support shows that the charge transferred to the metal is distributed on the metal surface and that such a particle becomes surrounded by an electrical field that is stronger at the interface but surrounds the entire particle (Fig. 2g). The uptake of H is driven by a transfer of charge that then distributes on the metal surface, which explains why H uptake correlates with Cu surface area (Fig. 2d).The prior state of the art understanding 34,35 has been that significant electronic metal-support interactions in catalysts were limited to small metal clusters, where most of the metal atoms have an interface to the support. We show that these effects are also relevant for the sizable nanoparticles present in most industrial catalysts. The explanation lies in our discovery that donor/acceptor states from dopants, such as the contained hydrogen in ZnO, represent a larger reservoir for charge transfer than previously realized. Adsorbates on the catalyst are affected by the charge transfer. TPD experiments show that formate, the major adsorbate during methanol synthesis, is generally stabilized by ca. 0.2 eV on Cu/ZnO samples compared to pure Cu (Fig. 3a). From DFT calculations where the system is modelled as a Cu rod on a ZnO( ) surface (Fig. 3b), we can conclude that this is because the charge transferred from donors in the adjacent ZnO phase provides additional electrons that participate in the HCOO-Cu bond. Fig. 3b shows that the stabilizing effect is strongest for adsorbates right at the interface where hydrogen bonds to OH on the support also participates in the stabilisation, but even on the wider surface far away from the interface the adsorbates are also stabilised by ca. 0.2 eV due to electrons from the support. The internal bonds in the adsorbed molecules are also affected. IR spectroscopy reveals that the C-O frequency in CO adsorbed on the 2 wt% Pt/ZnO sample is redshifted by ca. 30-40 cm -1 compared to pure Pt (Fig. 3b). DFT calculations for a Pt 54 cluster shows that this arises from the negative charge transferred from the support (mainly from H dopants) to the metal. (Fig. 3d). The DFT calculations also show that the modifications of the internal bonds are strongest at the interface but present across the metal surface (Fig. E4). In nature, enzymes are believed to direct chemical reactions by means of charged sites and electric fields, and based on the CO frequency shifts in Fig. 3c the field strengths on this inorganic sample resemble levels estimated 36–38 at the active sites in enzymes. Inorganic, industrial catalysts with fields extending between the constituents could, thus, have closer analogies to enzymes than previously realized. This reveals new possibilities for the development of inorganic catalysts with inspiration from enzymatic processes (and vice versa). To evaluate how the rate-limiting (HCO)-(OH) scission on Cu is affected by charge transfer, the Cu/ZnO system was evaluated by the DFT model with a Cu rod on a ZnO( ) surface with/without adsorbed H on the ZnO (Fig. 4a) or as a 48 atom Cu(211) slab with/without an artificially added electron (Fig. 4b). Fig. 4 shows the calculated potential energy landscape for the HCOOH dissociation and illustrates the importance of charge transfer for the reaction on the Cu surface. Fig. 4a shows that the barrier is lowered by 0.1 eV when H is adsorbed in ZnO, which corresponds to a one order of magnitude increase in the reaction rate at typical reaction temperatures of 220-280 ℃. Consequently, charge transfer from the support to the metal strongly modifies the reactions occurring on the metal phase, as the rate of a catalytic reaction on the metal surface is essentially determined by the concentration of electron donors on the adjoining support. Fig. 4b shows that adding one electron to a Cu(211) slab causes a considerable 0.4 eV lowering of the barrier for HCOOH dissociation, which verifies that the facile scission is a result of charge transfer. The strong effect also implies that there may be a possibility for even more active catalysts, through enhancements of the charging effects. Fig. 4 shows that the HCOOH mechanism provides a favourable route to formation of formyl (HCO) species, because HCOOH has a Cu-O bound HCO moiety, whereas CO hydrogenation produces formyl from Cu-C bound CO, which can explain the differences in reaction rates from CO and CO 2 . 4,15 Finally, our results also show that the primary location of the turnovers on a catalyst will be a function of the reaction conditions. The sites directly at the interface between the constituent phases experience the strongest modifications, whereas sites on the wider Cu surface experience a smaller but substantial modification (Figs. 2 and 3). Our DFT results show that the stabilisation of adsorbed formate is so strong right at the interface that it is favourable to establish a high coverage (Fig. E5). The high coverage precludes the non-adjacent sites required for the products in the rate-limiting step (see Fig. 4). Under conditions favouring such coverages, the sites immediately at the interface will play a limited role and the more moderately affected sites further into the Cu particle will dominate turnovers. At industrial high pressures favouring high coverages the TOF for methanol synthesis is lower for smaller Cu particles 39 , even though smaller particles have relatively more metal-oxide interface sites at the periphery 40 , and this indication of a limited role of the sites right at the interface is explained by our results. The effects reported here are consistent with all the previous suggestions advanced to explain the role of structural features in Cu/ZnO catalysts, namely structure sensitivities, alloying effects and oxide decoration of the metal in SMSI effects. Step sites on Cu have been reported 10 to be more active, and Fig. 4b shows that charge transfer effects are also very important for step sites. Steps and edges also become focal points for electrical fields from charged crystals 41 . Encapsulation of metal particles by SMSI can occur 42 , but the industrial suppliers of Cu/ZnO catalysts report 2,12,43 that no significant SMSI effects occur under typical reaction conditions. However, the electron transfer effects reported here would be important regardless of the existence of partial encapsulation. In fact, the strong electronic interactions could be a major factor influencing the morphology of catalytic materials causing SMSI and restructuring. The Cu-ZnO synergy can be observed on single crystal model systems 3,5 , which consist of 3D islands of ZnO on the Cu surface 44 , and such a geometry would also be affected by the charge transfer dynamics observed here. As noted, the electron transfer effects occur regardless of the existence of CuZn alloying, and the electronic interactions can account for the synergy without invoking alloying. The extent of CuZn-alloying in Cu/ZnO catalysts depends on the presence of CO in the atmosphere 5,45,46 whereas the rate enhancement from ZnO is independent of CO 4 . There is thus no direct evidence of a positive effect of alloying, although not all influences of alloying can be clarified with the presently available information. In conclusion, we have elucidated the mechanism and active sites for the hydrogenation of CO 2 into methanol with Cu/ZnO catalysts and identified the electronic interactions between the Cu and ZnO phases that cause a synergistic acceleration of the reaction. The significance of the charge transfer effects identified here implies that an accurate representation of these is essential for understanding reactions on multicomponent catalysts and explains why Cu almost behaves as a different element in the Cu/ZnO catalyst. The ability to transfer charge between adjoining phases in nanomaterials and thereby regulate the concentration of donor/acceptor states may not only be essential for heterogeneous catalysts but also for many other applications ranging from ad-/absorbent materials to electronic components. Declarations Acknowledgements This work was supported by the Villum Foundation Center for Science of Sustainable Fuels and Chemicals, Villum Foundation V-SUSTAIN Grant 9455. We would also like to acknowledge the Pioneer Center of Accelerating P2X Materials Discovery (CAPeX), DNRF grant P3, for support. Financial support from the Swedish Research Council (2020-05191) is acknowledged. The Competence Centre for Catalysis (KCK) is hosted by Chalmers University of Technology and is financially supported by the Swedish Energy Agency and the member companies Johnson Matthey, Perstorp, Powercell, Preem, Scania CV, Umicore, and Volvo Group. The calculations have been performed at PDC and NSC via a NAISS grant (NAISS 2025/3-36). We acknowledge the MAX IV Laboratory for beamtime on the HIPPI beamline under proposal 20252200. Research conducted at MAX IV, a Swedish national user facility, is supported by Vetenskapsrådet (Swedish Research Council, VR) under contract 2018-07152, Vinnova (Swedish Governmental Agency for Innovation Systems) under contract 2018-04969 and Formas under contract 2019-02496. Contributions JAB, JMC, TEA, NDN, JLF, LB-E: Catalytic experiments/analysis. RS&HG: DFT calculations/analysis. ML&CDD: TEM experiments/analysis, JMC, AS, JAB, J-CG: XPS measurements/analysis. HG, ADJ&JMC: acquiring funding, research supervision and planning. JMC: Original idea, initiation of research experiments and data analysis. RS, HG, ADJ, JMC&JAB: First draft of manuscript. All: comments on manuscript. References Baltes, C., Vukojević, S. & Schüth, F. Correlations between synthesis, precursor, and catalyst structure and activity of a large set of CuO/ZnO/Al 2 O 3 catalysts for methanol synthesis. J. Catal. 258, 334–344 (2008). Barrow, N. et al. Doubling the life of Cu/ZnO methanol synthesis catalysts via use of Si as a structural promoter to inhibit sintering. Sci. Adv. 10, eadk2081 (2024). Nakamura, I., Fujitani, T., Uchijima, T. & Nakamura, J. A model catalyst for methanol synthesis: Zn‐deposited and Zn‐free Cu surfaces. J. Vac. Sci. Technol. A 14, 1464 (1998). Nielsen, N. D., Jensen, A. D. & Christensen, J. M. The roles of CO and CO 2 in high pressure methanol synthesis over Cu-based catalysts. J. Catal. 393, 324–334 (2021). Kattel, S., Ramírez, P. J., Chen, J. G., Rodriguez, J. A. & Liu, P. Active sites for CO 2 hydrogenation to methanol on Cu/ZnO catalysts. Science 355, 1296–1299 (2017). Graciani, J. et al. Highly active copper-ceria and copper-ceria-titania catalysts for methanol synthesis from CO 2 . Science 345, 546–550 (2014). Pacchioni, G. From CO 2 to Methanol on Cu/ZnO/Al 2 O 3 Industrial Catalyst. What Do We Know about the Active Phase and the Reaction Mechanism? ACS Catal. 14, 2730–2745 (2024). Beck, A., Newton, M. A., van de Water, L. G. A. & van Bokhoven, J. A. The Enigma of Methanol Synthesis by Cu/ZnO/Al 2 O 3 -Based Catalysts. Chem. Rev. 124, 4543–4678 (2024). Ye, J. et al. Hydrogenation of CO 2 for sustainable fuel and chemical production. Science 387, eadn9388 (2025). Behrens, M. et al. The active site of methanol synthesis over Cu/ZnO/Al 2 O 3 industrial catalysts. Science 336, 893–897 (2012). Schlögl, R. Chemical Batteries with CO 2 . Angew. Chem. Intl. Ed. 61, e202007397 (2022). Kuld, S. et al. Quantifying the promotion of Cu catalysts by ZnO for methanol synthesis. Science 352, 969–974 (2016). Kondrat, S. A. et al. Stable amorphous georgeite as a precursor to a high-activity catalyst. Nature 531, 83–87 (2016). Chinchen, G. C., Denny, P. J., Parker, D. G., Spencer, M. S. & Whan, D. A. Mechanism of methanol synthesis from CO 2 /CO/H 2 mixtures over copper/zinc oxide/alumina catalysts: use of 14 C-labelled reactants. Appl. Catal. 30, 333–338 (1987). Nielsen, N. D., Thrane, J., Jensen, A. D. & Christensen, J. M. Bifunctional Synergy in CO Hydrogenation to Methanol with Supported Cu. Catal. Lett. 150, 1427–1433 (2020). Thrane, J. et al. Methanol-Assisted Autocatalysis in Catalytic Methanol Synthesis. Angew. Chem. Intl. Ed. 59, 18189–18193 (2020). Kunkes, E. L., Studt, F., Abild-Pedersen, F., Schlögl, R. & Behrens, M. Hydrogenation of CO2 to methanol and CO on Cu/ZnO/Al2O3: Is there a common intermediate or not? J. Catal. 328, 43–48 (2015). Quan, J. et al. Vibration-driven reaction of CO 2 on Cu surfaces via Eley–Rideal-type mechanism. Nat. Chem. 11, 722–729 (2019). Higham, M. D. et al. Mechanism of CO 2 conversion to methanol over Cu(110) and Cu(100) surfaces. Dalton Transactions 49, 8478–8497 (2020). Grabow, L. C. & Mavrikakis, M. Mechanism of methanol synthesis on cu through CO 2 and CO hydrogenation. ACS Catal. 1, 365–384 (2011). Cao, A., Wang, Z., Li, H., Elnabawy, A. O. & Nørskov, J. K. New insights on CO and CO 2 hydrogenation for methanol synthesis: The key role of adsorbate-adsorbate interactions on Cu and the highly active MgO-Cu interface. J. Catal. 400, 325–331 (2021). Takeyasu, K. et al. Hydrogenation of Formate Species Using Atomic Hydrogen on a Cu(111) Model Catalyst. J. Am. Chem. Soc. 144, 12158–12166 (2022). Zhao, Y. F. et al. Insight into methanol synthesis from CO 2 hydrogenation on Cu(111): Complex reaction network and the effects of H 2 O. J. Catal. 281, 199–211 (2011). Blanksby, S. J. & Ellison, G. B. Bond Dissociation Energies of Organic Molecules. Acc. Chem. Res. 36, 255–263 (2003). Metcalfe, W. K., Simmie, J. M. & Curran, H. J. Ab Initio Chemical Kinetics of Methyl Formate Decomposition: The Simplest Model Biodiesel. J. Phys. Chem. A 114, 5478–5484 (2010). Karim, W. et al. Catalyst support effects on hydrogen spillover. Nature 541, 68–71 (2017). Van de Walle, C. G. & Neugebauer, J. Universal alignment of hydrogen levels in semiconductors, insulators and solutions. Nature 423, 626–628 (2003). Allen, M. W. & Durbin, S. M. Influence of oxygen vacancies on Schottky contacts to ZnO. Appl. Phys. Lett. 92, 122110 (2008). Frost, J. C. Junction effect interactions in methanol synthesis catalysts. Nature 334, 577–580 (1988). Honkala, K., Hellman, A. & Grönbeck, H. Water Dissociation on MgO/Ag(100): Support Induced Stabilization or Electron Pairing? J. Phys. Chem. C 114, 7070–7075 (2010). Westendorff, K. S., Hülsey, M. J., Wesley, T. S., Román-Leshkov, Y. & Surendranath, Y. Electrically driven proton transfer promotes Brønsted acid catalysis by orders of magnitude. Science 383, 757–763 (2024). Wesley, T. S., Román-Leshkov, Y. & Surendranath, Y. Spontaneous Electric Fields Play a Key Role in Thermochemical Catalysis at Metal-Liquid Interfaces. ACS Cent. Sci. 7, 1045–1055 (2021). Svensson, R. & Grönbeck, H. Spontaneous Charge Separation at the Metal-Water Interface. ChemPhysChem 25, e202400099 (2024). Xu, M. et al. Renaissance of Strong Metal–Support Interactions. J. Am. Chem. Soc. 146, 2290–2307 (2024). Lykhach, Y. et al. Counting electrons on supported nanoparticles. Nat. Mater. 15, 284–288 (2015). Welborn, V. V., Ruiz Pestana, L. & Head-Gordon, T. Computational optimization of electric fields for better catalysis design. Nat. Catal. 1, 649–655 (2018). Zheng, C., Ji, Z., Mathews, I. I. & Boxer, S. G. Enhanced active-site electric field accelerates enzyme catalysis. Nat. Chem. 15, 1715–1721 (2023). Fried, S. D., Bagchi, S. & Boxer, S. G. Extreme electric fields power catalysis in the active site of ketosteroid isomerase. Science 346, 1510–1514 (2014). Van Den Berg, R. et al. Structure sensitivity of Cu and CuZn catalysts relevant to industrial methanol synthesis. Nat. Commun. 7, 13057 (2016). Cargnello, M. et al. Control of metal nanocrystal size reveals metal-support interface role for ceria catalysts. Science 341, 771–773 (2013). LaCount, M. D., Lambeets, S. V., Perea, D. E., Prozorov, T. & Kathmann, S. M. Electric Fields at Interfaces. J. Phys. Chem. C 129, 15489–15506 (2025). Lunkenbein, T. et al. Formation of a ZnO Overlayer in Industrial Cu/ZnO/Al 2 O 3 Catalysts Induced by Strong Metal–Support Interactions. Angew. Chem. Intl. Ed. 54, 4544-4548 (2015). Hansen, P. L. et al. Atom-resolved imaging of dynamic shape changes in supported copper nanocrystals. Science 295, 2053–2055 (2002). Orozco, I. et al. Understanding Methanol Synthesis on Inverse ZnO/CuO x /Cu Catalysts: Stability of CH 3 O Species and Dynamic Nature of the Surface. J. Phys. Chem. C 125, 6673–6683 (2021). Amann, P. et al. The state of zinc in methanol synthesis over a Zn/ZnO/Cu(211) model catalyst. Science 376, 603–608 (2022). Beck, A. et al. Following the structure of copper-zinc-alumina across the pressure gap in carbon dioxide hydrogenation. Nat. Catal. 4, 488–497 (2021). Methods 2.1 Catalyst synthesis Cu/ZnO/Al 2 O 3 , Cu/ZnO and unsupported ZnO were made by the precipitation method of Baltes et al. 1 as described elsewhere 15 . We varied the Cu surface area by varying the composition, the addition rate of the precursors and by calcining the samples to different temperatures. A list of preparation conditions and sample properties are given in the supplementary information. We also tested that a commercial methanol synthesis catalyst from Topsoe to ensure that it was on the same rate vs Cu area line in Fig. E1a as our in house samples. By agreement with the supplier, we do not disclose the sample product code. Our unsupported Cu sample was a Raney TM -type, sponge Cu purchased from Strem Chemicals. Our handling of the sample and further characterizations of this material can be found elsewhere 4,15,51,52 . The sample contains a small concentration of Al 2 O 3 as a structural stabilizer. For spectroscopic tests we also used a commercial CuO sample (Sigma Aldrich, nanopowder). Pt samples were synthesized by impregnating Pt on ZnO or SiO 2 . Here we used either our precipitated ZnO or commercial ZnO (Sigma-Aldrich, 7.7 m 2 /g, ≥ 99.99% trace metal basis) and SiO 2 (SS61138, 250 m 2 /g, Saint Gobain), see supplementary information for details. Platinum(II)2,4-pentanedionate (Alfa Aesar, ≥ 99.977%) was dissolved in a small excess of ethanol in a beaker on a heating plate, then the support was added followed by gradual evaporation of ethanol. The standard calcination of the samples was heating (1 °C/min) to 330 °C in an air flow (1 NL/min) and holding this temperature for 3 hr. All samples were pelletized/crushed/sieved to obtain a particle size of 150-300 μm. As a pure Pt reference we used a commercial platinum black sample (Merck, ≥ 99.95%). 2.2 Experimental setup and gases Methanol synthesis and temperature programmed desorption (TPD) experiments were performed in a high-pressure reactor setup described elsewhere 15 . Gases used in these experiments are 5% H 2 /Ar (both 99.999%), 5000 ppmv CO 2 in N 2 (99.999%), 99.9999% H 2 , 99.9999% Ar, 99.995% CO 2 , 1% N 2 O/He (both 99.999%) and 99.97% CO from Air Liquide as well as D 2 (99.8%-D) from Linde gas. 2.3 Catalyst reduction and pre-treatments Catalyst samples were reduced in situ using a 100 NmL/min flow of 5% H 2 /Ar at atmospheric pressure, by heating (1 °C/min) to 175 °C and holding this temperature for 2 h. For Pt/ZnO, we also tested reduction with the same program using 10% CO/Ar. 2.4 Catalytic tests Methanol synthesis experiments were conducted in the high-pressure setup 15 after pre-reduction. Methanol synthesis activity is investigated at low conversion conditions far from equilibrium at mild (T = 135 °C, 1 atm, GHSV = 48000 NmL/g cat /h) and industry-type, high-pressure conditions (20 bar, GHSV = 24∙10 5 NmL/g cat /h), but still at low conversion with at most 2000 ppmv methanol produced. When measuring reaction orders in hydrogen, the total gas flow is fixed by replacing part of the hydrogen flow with argon. 2.5 Determination of stored hydrogen in ZnO TPD of adsorbed hydrogen was conducted by treating the pre-reduced sample (typically 250 mg) to 1 bar of H 2 at 175 °C for 1 h (or up to 96 h in the case of pure ZnO) followed by rapid quench cooling as described by Nielsen et al. 51 . The system was then thoroughly flushed with Ar and subjected to a TPD in 50 NmL/min Ar flow with a 2 °C/min heating rate. The H 2 desorption was followed with a mass spectrometer and the MS signal was converted into concentration by calibration against a certified H 2 /Ar gas mixture. The desorption below 100 °C is due to H on Cu metal 53–55 and the higher temperature desorption is due to H on ZnO 56,57 . Adsorption energies were calculated from the TPD peak temperature using the Redhead equation and an assumed pre-factor of 10 13 s -1 . 2.6 Cu surface area determination The Cu surface area was determined by N 2 O-Reactive frontal chromatography (N 2 O-RFC) 58 at 60 °C using a 15 NmL/min flow of 1 mol% N 2 O/He. The Cu surface area is derived from the N 2 O uptake, a Cu:O stoichiometry of 2:1 53 , and an average Cu surface atom density of 1.47∙10 19 Cu atoms/m 2 . The Cu surface area was determined for Cu-based catalysts reduced in 5% H 2 /Ar at 175 °C. The Cu surface area in catalysts that are exposed to industrial-like reaction conditions (T = 260 °C, P = 20 bar) was determined after normal reduction followed by a heating to 260 °C in Ar (1 °C/min) to capture the sintering of the samples at higher temperatures. 2.7 Infrared spectroscopy In situ Infrared (IR) measurements were performed with undiluted samples in a domed reactor cell and Praying Mantis DRIFTS unit from Harrick Scientific Products installed in a Nicolet iS50 FTIR spectrometer equipped with a liquid N 2 cooled MCT detector. Details of the setup are described elsewhere 15 . Each IR spectra is an average of 76 scans with a resolution of 4 cm -1 . For the IR measurements on adsorbed CO, the sample cup was filled with the Pt sample (ca 30 mg) and the background was measured in Air without flow through the system at 20 °C. This was done because even background traces of CO remaining in the gas lines were sufficient to cause high CO coverages on the Pt. The sample was then reduced in 100 NmL/min of 5% H 2 /Ar at 1 atm with a ramping rate of 5 °C /min to 175 °C. The temperature was decreased to 20 °C where CO is adsorbed from a flow of 0.64 % CO in Ar. The shown spectra are at 20 °C after the gaseous CO had been purged from the system by Ar flow. For studies of adsorbates under reaction conditions the samples, either Cu/ZnO/Al 2 O 3 or pure ZnO were pre-reduced and cooled to 135 °C where a background was collected. We then performed either H 2 /D 2 switch experiments or exposed the sample to syngas followed by exchange of H 2 with D 2 . 2.8 Scanning transmission electron microscopy STEM (scanning transmission electron microscopy) imaging was made of Pt/ZnO pre-reduced elsewhere and transferred through air to the microscope. A Titan Analytical 80-300ST TEM (probe corrected) equipped with an Oxford windowless EDX detector (X-Max 100TLE) was used for these measurements, and the microscope was operated with an acceleration voltage of 300 kV and an electron dose rate of 150 pA. The measurements were conducted at room temperature and ultra-high vacuum (10 −10 bar). Prior to the measurements, a Pt/ZnO sample was placed on a lacey carbon supported Au grid, which was loaded into the microscope in a CuZn or Be holder. Both STEM Energy-dispersive X-ray spectroscopy (EDX) and STEM high angular annular dark field (HAADF) were performed, and the image resolution was 1024x1024 and 512x512, respectively. 2.9 BET and surface area determinations For Cu/ZnO/Al 2 O 3 samples, the oxide surface area was determined as the difference between the N 2 O-RFC measurement of the Cu area and the BET measurement of the total surface area. BET analysis was performed on a pre-reduced and N 2 O passivated sample using a Quantacrhome NOVAtouch Gas sorption analyzer. In the NOVAtouch analyzer, the sample was vacuum degassed at 150 °C (10 °C/min) for 16 h before being subjected to a 6-points BET measurement. 2.10 Ambient pressure X-ray photoelectron spectroscopy The experiments have been performed at the Solid-Gas Endstation 59 of the HIPPIE beamline 60 , MAX IV Laboratory, Lund, Sweden. The measurements were of Cu/ZnO/Al 2 O 3 and pure Cu samples at an X-ray energy of 1580 eV. The samples were pre-reduced (1 atm 5% H 2 /Ar for 2 h at 175 °C, then 2 h in 250 °C and finally 2 h in 100% H 2 at 250 °C) in the catalytic reactor, passivated by exposure to 1% N 2 O at 60 °C and transported to the synchrotron in this state. As pure Cu reference, we used either the Raney type Cu used for the catalytic tests or commercial CuO (Sigma Aldrich, nanopowder) and these both gave identical results. At the synchrotron, the samples were dropcast from an ethanol slurry onto sample holders covered by gold foil. The samples were treated in 1 bar H 2 at 175 °C and transferred through vacuum to the measurement cell to produce Fig. 2a. Alternatively, the samples were transferred directly to the measurement cell and then first measured in 14 mbar He at room temperature and then heated to 250 °C in 14 mbar H 2 and measured at this condition. For pure Cu samples, we used the 84 eV 4f 7 / 2 binding energy of the underlying Au foil and the 284.8 eV 1s binding energy of adventitious carbon as energy references and these both gave identical results (see supplementary information). For the Cu/ZnO/Al 2 O 3 samples, we used the 1021.73±0.37 eV 2p 3 / 2 of Zn in ZnO as energy reference. This value was determined as the average of the reference spectra in the NIST database 61 and this is the value measured for Zn in high level XPS studies of Cu/ZnO materials 5,45,62 . Adventitious carbon gave a similar energy calibration and this was used to validate the approach (see supplementary information). This consistency in the energy references ensures the reliability of our reported values. A relaxation model 63 was used to confirm that the observed shifts are intrinsic effects (see supplementary information). All data treatment was made without background subtraction, while for figures a Shirley background has been subtracted for visual clarity. 2.11 Ex situ X-ray photoelectron spectroscopy The XPS measurements of Pt/ZnO, Pt/SiO 2 and Pt black were performed in UHV with an ESCALAB Xi+ instrument from ThermoFisher Scientific, using a monochromatic Al Kα (1486.6 eV) X-ray source. The powdered samples were reduced ex situ and then transferred through air and placed onto an adhesive copper tape (3M Electrical Tape). A low energy electron flood gun was used for the neutralization of surface charge buildup. The binding energies were calibrated by using Au foil. The diameter of the analysis area was approximately 650 μm. In situ sputtering was also used to remove oxygen adsorbed during sample transfer but this had no detectable impact on measured binding energies. 2.12 Mechanistic investigations To dose formic acid into the IR cell, we installed a Swagelok ¼-inch T fitting on the gas inlet line immediately before the cell. One leg of the T fitting has a GC rubber membrane that allows injection of liquid into the gas stream entering the cell using a GC syringe. We then performed experiments where a sample (unsupported Cu, Cu/ZnO/Al 2 O 3 or ZnO) was loaded into the sample cup and pre-reduced. Formic acid was injected (2 µL = 53 µmol) into a 50 NmL/min Ar flow at atmospheric pressure to pre-adsorb HCOO at 40 °C, the sample was then flushed in Ar flow and a TPD was conducted. We either used HCOOH (99.5%, Fisher Scientific) or DCCOD (≥ 95% D, Fisher Scientific). The IR spectrometer monitoring the IR cell was used to confirm that adsorbed formate had been produced. To evaluate the formate adsorption energy, the CO 2 concentration during a TPD (50 NmL/min Ar flow, 2 °C/min) was monitored by mass spectrometry, and the MS signal was calibrated against a certified CO 2 /Ar gas mixture from Air Liquide. To compare the isotopic distribution of surface formates and formed methanol, we used the TPD experiments to calculate the relative stability and coverage of HCOO and DCOO (see supplementary information for details). We then conducted methanol synthesis with a D 2 /H 2 /CO 2 gas mixture and measured masses 31-36 by mass spectrometry. The formate coverage was determined from integration of the CO 2 signal during the TPD. Because the appearance energies, ionization potentials and fragmentation patterns are very similar for all the H/D-methanol isotopes 64–66 it is considered justifiable to treat the fragmentation of mixed isotopes as identical and randomized. This allowed us to deconvolute the mass spectrometry measurements into a distribution of methanol isotopes (see supplementary information for details). We performed TPH measurements of adsorbed DCOO on various samples by first pre-adsorbing DCOO by the method described above and then heating (2 °C/min) in a 50 NmL/min H 2 flow while monitoring the product formation by mass spectrometry. To test if HCOOH conversion is the rate-limiting step in the conversion of HCOO, we first made a TPH of pre-adsorbed HCOO. We then made an identical experiment where we injected another 53 µmol HCOOH pulse at 121 °C before HCOO can be converted and an additional pulse at 53 µmol HCOOH pulse at 205 °C after HCOO can be converted. During TPH of pre-adsorbed HCOO, the majority of the formate decomposes to CO 2 and the second pulse replenishes adsorbed HCOO which is the reason for the higher general methanol formation in the multi-pulse experiment in Fig. 1c. To study formic acid conversion, either formic acid-2H or formic acid-2D were injected into a 100 NmL/min flow of H 2 or D 2 and the gas entered a cell kept at 200 °C filled with pre-reduced Cu/ZnO/Al 2 O 3 or unsupported Cu. We followed the resulting methanol formation by mass spectrometry and GC measurements and used the abovementioned method to deconvolute the methanol isotopes. Here the MS measurements were calibrated against simultaneous GC measurements. As described in the supplementary information, Cu/ZnO/Al 2 O 3 and unsupported Cu both give identical information in the first pulse, but in subsequent pulses Cu/ZnO/Al 2 O 3 shows isotopic scrambling because formed methanol re-adsorbs on ZnO as methoxide that undergoes H/D exchange with the gas phase. Unsupported gave no such signs of readsorption and for that reason the kinetic isotope effect investigations in Fig. 1e were made with unsupported Cu. 2.13 Computational method Density functional theory calculations were performed using the Vienna Ab Initio Simulation package VASP 67–70 . The projector-augmented wave method was applied to describe the interactions between the core- and valence electrons 71 . The considered valence electrons are 1s 1 (H), 2s 2 2p 2 (C), 2s 2 2p 4 (O), 3d 10 4s 1 (Cu), 3d 10 4p 2 (Zn), and 6s 1 5d 9 (Pt). The exchange and correlation functional was described using the generalized gradient approximation proposed by Perdew, Burke, and Ernzerhof (PBE) 72 . To describe the localised electrons in ZnO, a rotationally invariant Hubbard-U correction of 7.5 eV was applied to the 3d electrons in Zn 73 . The plane waves were truncated at 500 eV in the expansion of the Kohn-Sham orbitals. The electronic structure was considered converged when the change in Kohn-Sham eigenvalues and electronic energy, between two succeeding iterations are below 1 × 10 −6 eV. The structural optimization was performed using the conjugate-gradient method, and the structures are considered converged when the maximum force acting on any nuclei is below 0.03 eV/Å. The hexagonal lattice constants of bulk ZnO were calculated to be a = b = 3.19 Å and c = 5.12 Å, slightly smaller than the experimental values of a = b = 3.24 Å and c = 5.18 Å. The ZnO( ) surface was represented by a p (6x5) cell, constituted of four ZnO layers, of which the bottom two layers were kept fixed to their bulk positions. The periodic ZnO surfaces were separated by a vacuum layer of ~22 Å. The large vacuum layer was included to allow for the inclusion of the supported nanoparticles (NPs). The lattice constants of bulk Cu and bulk Pt were calculated to be 3.64 Å and 3.97 Å, respectively, in fair agreement with the experimental values of 3.61 Å and 3.92 Å, respectively. The k-point sampling was performed using the Γ-point. The energy of gas phase H 2 and CO 2 were calculated using a (30, 31, 32) Å vacuum box. Vibrational modes were calculated using the finite differences method, assuming the harmonic approximation. The analysis of the electron density distribution was investigated with Bader charge analysis 74 and charge density difference analysis. The effects of the adsorbed hydrogen-induced charge transfer are investigated further for the adsorption of HCOO and the dissociation of HCOOH to HCO and OH. For the dissociation of HCOOH, both the reaction energy and the energy barriers were calculated. The barriers were calculated using the climbing image nudged elastic band (CI-NEB) method, 75 and the transition states were verified by vibrational analyses. 2.14 Continuum modelling Continuum modelling was conducted using Comsol Multiphysics 6.2. We started from the geometries developed by Binninger et al. 76 but changed the system to one where support and particle are touching. We used a system of perfect conductors and a 1 V bias between particle and support. These calculations are only used for qualitative results, so we do not show numbers in the legend of Fig. 2g. Additional references Nakano, H., Nakamura, I., Fujitani, T. & Nakamura, J. Structure-dependent kinetics for synthesis and decomposition of formate species over Cu(111) and Cu(110) model catalysts. J. Phys. Chem. B 105, 1355–1365 (2001). Ruehl, G., Harman, S. E., Gluth, O. M., Lavoy, D. H. & Campbell, C. T. Energetics of Adsorbed Formate and Formic Acid on Cu(111) by Calorimetry. ACS Catal. 12, 10950–10960 (2022). Yasumura, H., Mizutani, A., Nagatsuka, N., Watanabe, K. & Koitaya, T. Surface Orientation Dependent Dissociation of Formic Acid and Adsorption States of Formate on Copper. J. Phys. Chem. C 129, 1216–1227 (2025). Hansen, S. I. et al. An Adsorption Isotherm That Includes the Interactions between Adsorbates. J. Phys. Chem. C 129, 5287–5300 (2025). Nielsen, N. D., Jensen, A. D. & Christensen, J. M. Quantification of Formate and Oxygen Coverages on Cu Under Industrial Methanol Synthesis Conditions. Catal. Lett. 150, 2447–2456 (2020). Nielsen, N. D., Smitshuysen, T. E. L., Damsgaard, C. D., Jensen, A. D. & Christensen, J. M. Characterization of oxide-supported Cu by infrared measurements on adsorbed CO. Surf. Sci. 703, 121725 (2021). Chatterjee, R. et al. Mapping Support Interactions in Copper Catalysts. Top. Catal. 62, (2019). Muhler, M., Nielsen, L. P., Törnqvist, E., Clausen, B. S. & Topsøe, H. Temperature-programmed desorption of H 2 as a tool to determine metal surface areas of Cu catalysts. Catal. Lett. 14, 241–249 (1992). Fichtl, M. B. & Hinrichsen, O. On the temperature programmed desorption of hydrogen from polycrystalline copper. Catal. Lett. 144, 2114–2120 (2014). Doh, W. H., Roy, P. C. & Kim, C. M. Interaction of Hydrogen with ZnO: Surface Adsorption versus Bulk Diffusion. Langmuir 26, 16278–16281 (2010). Barański, A. & Gałuszka, J. Temperature-programmed desorption studies of the hydrogen-zinc oxide system. J. Catal. 44, 259–270 (1976). Chinchen, G. C., Hay, C. M., Vandervell, H. D. & Waugh, K. C. The measurement of copper surface areas by reactive frontal chromatography. J. Catal. 103, 79–86 (1987). Jones, R. et al. Multimodal Ambient Pressure Sample Environment for the HIPPIE Solid-Gas Endstation at MAX IV Laboratory. Photon Science , in press doi:10.1021/PHOTONSCI.5C00021. Zhu, S. et al. HIPPIE: a new platform for ambient-pressure X-ray photoelectron spectroscopy at the MAX IV Laboratory. J. Synchrotron Rad. 28, 624–636 (2021). P.J. Linstrom & W.G. Mallard. NIST Chemistry WebBook, NIST Standard Reference Database Number 69. (National Institute of Standards and Technology, Gaithersburg MD, 2025). Kuld, S., Conradsen, C., Moses, P. G., Chorkendorff, I. & Sehested, J. Quantification of Zinc Atoms in a Surface Alloy on Copper in an Industrial-Type Methanol Synthesis Catalyst. Angew. Chem. Intl. Ed. 53, 5941–5945 (2014). Bahl, M. K., Tsai, S. C. & Chung, Y. W. Auger and photoemission investigations of the platinum-SrTiO 3 (100) interface: Relaxation and chemical-shift effects. Phys. Rev. B 21, 1344 (1980). Nishimura, T., Niwa, Y., Tsuchiya, T. & Nozoye, H. Ionic dissociation of methanol studied by photoelectron–photoion coincidence spectroscopy. J. Chem. Phys. 72, 2222–2225 (1980). Omura, I., Kaneko, T., Yamada, Y. & Tanaka, K. Mass Spectrometric Studies of Photoionization. V. Methanol and Methanol-d1. J. Phys. Soc. Jpn. 27, 981–984 (1969). Beynon, J. H., Fontaine, A. E. & Lester, G. R. Mass spectrometry: the mass spectrum of methanol. Part I. Thermochemical information. Intl. J. Mass Spec. 1, 1–24 (1968). Kresse, G. & Hafner, J. Ab initio molecular dynamics for liquid metals. Phys. Rev. B 47, 558–561 (1993). Kresse, G. & Hafner, J. Ab initio molecular dynamics for open-shell transition metals. Phys. Rev. B 48, 13115–13118 (1993). Kresse, G. & Hafner, J. Ab initio molecular-dynamics simulation of the liquid-metal–amorphous-semiconductor transition in germanium. Phys. Rev. B 49, 14251–14269 (1994). Kresse, G. & Furthmüller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 54, 11169–11186 (1996). Blöchl, P. E. Projector augmented-wave method. Phys. Rev. B 50, 17953–17979 (1994). Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 77, 3865–3868 (1996). Dudarev, S. & Botton, G. Electron-energy-loss spectra and the structural stability of nickel oxide: An LSDA+U study. Phys. Rev. B 57, 1505–1509 (1998). Henkelman, G., Arnaldsson, A. & Jónsson, H. A fast and robust algorithm for Bader decomposition of charge density. Comput. Mater. Sci. 36, 354–360 (2006). Henkelman, G. et al. A climbing image nudged elastic band method for finding saddle points and minimum energy paths. J. Chem. Phys. 113, 9901–9904 (2000). Binninger, T., Schmidt, T. J. & Kramer, D. Capacitive electronic metal-support interactions: Outer surface charging of supported catalyst particles. Phys. Rev. B 96, 165405 (2017). Additional Declarations There is NO Competing Interest. Supplementary Files SIrevised.docx Supplementary information SupplementaryFigures.docx Cite Share Download PDF Status: Under Review 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-6939552","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Physical Sciences - Article","associatedPublications":[],"authors":[{"id":587585985,"identity":"dc5a6838-cab6-4e88-bb0e-7321c0002442","order_by":0,"name":"Jakob Christensen","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAn0lEQVRIiWNgGAWjYBACxgYGxocNYJqBQbKBSC3Mhg0JDIw9RGsBAjZJ0rQwTzv8rHLmDwbZ/RIJjDdnEOWw2WlmNzckMBj3SCQwW24gTksO280HCQyJQC1skg+I1VJIuhbGDTAtRDoszVhyRpqEcc+Zh82WRHnfcHbyw489Njay7e3JB2/2EKWlAUxJMEATABFAnjhlo2AUjIJRMKIBAGZhMntfgxAfAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-2495-8905","institution":"Technical University of Denmark","correspondingAuthor":true,"prefix":"","firstName":"Jakob","middleName":"","lastName":"Christensen","suffix":""},{"id":587585986,"identity":"f09e8f13-c4f0-4654-b71d-827011ef400e","order_by":1,"name":"Jonas Boysen","email":"","orcid":"","institution":"Technical University of Denmark","correspondingAuthor":false,"prefix":"","firstName":"Jonas","middleName":"","lastName":"Boysen","suffix":""},{"id":587585987,"identity":"433bb87f-3aad-4e72-8ac2-7b1c1619ad6f","order_by":2,"name":"Rasmus Svensson","email":"","orcid":"","institution":"Chalmers University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Rasmus","middleName":"","lastName":"Svensson","suffix":""},{"id":587585988,"identity":"fc8a1502-6078-461e-9a7b-f51612463ab3","order_by":3,"name":"Niels Nielsen","email":"","orcid":"","institution":"Technical University of Denmark","correspondingAuthor":false,"prefix":"","firstName":"Niels","middleName":"","lastName":"Nielsen","suffix":""},{"id":587585989,"identity":"8d0cd513-8845-49b5-aeb6-c820495dcd08","order_by":4,"name":"Mads Lützen","email":"","orcid":"","institution":"Technical University of Denmark","correspondingAuthor":false,"prefix":"","firstName":"Mads","middleName":"","lastName":"Lützen","suffix":""},{"id":587585990,"identity":"20a0ef1d-ad36-400e-a4f4-8424ea3922cf","order_by":5,"name":"Lin Bigom-Eriksen","email":"","orcid":"","institution":"Technical University of Denmark","correspondingAuthor":false,"prefix":"","firstName":"Lin","middleName":"","lastName":"Bigom-Eriksen","suffix":""},{"id":587585991,"identity":"137287bb-02ff-4b1e-b321-ee1a038d1584","order_by":6,"name":"Jacob Fischer","email":"","orcid":"https://orcid.org/0009-0001-4230-0111","institution":"Technical University of Denmark","correspondingAuthor":false,"prefix":"","firstName":"Jacob","middleName":"","lastName":"Fischer","suffix":""},{"id":587585992,"identity":"86a1e322-d8b3-4d28-ab93-b2538fda0bb6","order_by":7,"name":"Jean-Claude Grivel","email":"","orcid":"","institution":"Technical University of Denmark","correspondingAuthor":false,"prefix":"","firstName":"Jean-Claude","middleName":"","lastName":"Grivel","suffix":""},{"id":587585993,"identity":"667e0a8f-6db3-4c4d-9116-fc63958e40fa","order_by":8,"name":"Christian Damsgaard","email":"","orcid":"","institution":"Technical University of Denmark","correspondingAuthor":false,"prefix":"","firstName":"Christian","middleName":"","lastName":"Damsgaard","suffix":""},{"id":587585994,"identity":"48d08f74-ee03-4022-a397-8d0dabfa6fdb","order_by":9,"name":"Henrik Grönbeck","email":"","orcid":"https://orcid.org/0000-0002-8709-2889","institution":"Chalmers University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Henrik","middleName":"","lastName":"Grönbeck","suffix":""},{"id":587585995,"identity":"7179a4b9-3255-4934-9a57-5cf70fad6dd1","order_by":10,"name":"Anker Jensen","email":"","orcid":"https://orcid.org/0000-0002-7341-4859","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Anker","middleName":"","lastName":"Jensen","suffix":""},{"id":587585996,"identity":"ad9c40de-fc38-4284-b06e-d04aabb301a7","order_by":11,"name":"Andrey Shavorskiy","email":"","orcid":"","institution":"MAX IV Laboratory, Lund University","correspondingAuthor":false,"prefix":"","firstName":"Andrey","middleName":"","lastName":"Shavorskiy","suffix":""},{"id":587585997,"identity":"e2ab3447-e31e-4a9a-b9b7-7178f733e0dc","order_by":12,"name":"Thomas Andersen","email":"","orcid":"","institution":"Technical University of Denmark","correspondingAuthor":false,"prefix":"","firstName":"Thomas","middleName":"","lastName":"Andersen","suffix":""}],"badges":[],"createdAt":"2025-06-20 14:05:21","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6939552/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6939552/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":102491296,"identity":"697e256a-30e9-4ce1-8461-bad1b2ac3e85","added_by":"auto","created_at":"2026-02-12 08:43:04","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1341193,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe mechanism of methanol synthesis from CO\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e.\u003c/strong\u003e \u003cstrong\u003ea\u003c/strong\u003e Distribution of the methanol isotopes and the adsorbed formate isotopes during methanol synthesis with Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3 \u003c/sub\u003efrom a 72/8/20:H\u003csub\u003e2\u003c/sub\u003e/D\u003csub\u003e2\u003c/sub\u003e/CO\u003csub\u003e2\u003c/sub\u003e feed. \u003cstrong\u003eb \u003c/strong\u003eThe relative CDH\u003csub\u003e2\u003c/sub\u003eOH production in TPH of pre-adsorbed DCOO, signals scaled to the same metal/ZnO area ratio as in the Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e sample. \u003cstrong\u003ec \u003c/strong\u003eTPH of pre-adsorbed HCOO on Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e compared to an experiment with HCOOH pulse injections during the TPH. \u003cstrong\u003ed\u003c/strong\u003e Methanol yields when 379 µmol/g formic acid is pulsed into H\u003csub\u003e2\u003c/sub\u003e/D\u003csub\u003e2\u003c/sub\u003e passing over unsupported Cu at 200 ℃. \u003cstrong\u003ee\u003c/strong\u003e The methanol isotope distribution when a pulse of DCOOD is injected into a flow of H\u003csub\u003e2\u003c/sub\u003e passing over Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e at 200 ℃. \u003cstrong\u003ef\u003c/strong\u003e The mechanism most consistent with the measurements. Colours: O: red, C: black, Cu: tan, H: white.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6939552/v1/812b8cee2c9835a6e4ed61b5.png"},{"id":102491413,"identity":"20335595-d45f-490e-8ee4-826216651fa1","added_by":"auto","created_at":"2026-02-12 08:43:13","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":2592746,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCharge transfer and synergistic interactions in Metal/ZnO systems.\u003c/strong\u003e \u003cstrong\u003ea\u003c/strong\u003e Cu 2p\u003csub\u003e3/2\u003c/sub\u003e XPS measurements of Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e and pure Cu after in situ exposure to 1 bar H\u003csub\u003e2\u003c/sub\u003e at 175 ℃ and after the H\u003csub\u003e2\u003c/sub\u003e-treated sample has been heated to 250 ℃ in vacuum. \u003cstrong\u003eb\u003c/strong\u003e TEM image of a 5 nm Pt particle in Pt/ZnO and elemental distributions around an 8 nm particle. \u003cstrong\u003ec\u003c/strong\u003e Pt 4f XPS measurements comparing Pt/ZnO and unsupported platinum. \u003cstrong\u003ed\u003c/strong\u003e H\u003csub\u003e2\u003c/sub\u003e uptake (1 bar H\u003csub\u003e2\u003c/sub\u003e, 175 ℃) on ZnO as a function of Cu surface area in Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e samples. \u003cstrong\u003ee\u003c/strong\u003e Charge transferred (relative to H-free ZnO) to a Pt\u003csub\u003e54\u003c/sub\u003e or Cu\u003csub\u003e54\u003c/sub\u003e cluster on ZnO() as a function of the number of adsorbed H atoms on the oxide and charge density difference analysis showing electron enrichment (yellow) on the metal and depletion (light grey/blue) around the adsorbed H. \u003cstrong\u003ef\u003c/strong\u003e Electron energy diagram\u003csup\u003e27–29\u003c/sup\u003e for Cu/ZnO \u003cstrong\u003eg\u003c/strong\u003e Electrical field around a hemisphere biased by 1 V relative to the support (spherical particle superimposed for illustration).\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6939552/v1/53405c4f0a76e72e348ad1b1.png"},{"id":102491460,"identity":"2d19340d-2734-4ac6-a7ff-9325ec5f0c86","added_by":"auto","created_at":"2026-02-12 08:43:20","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1499630,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe effect of charge transfer on surface atoms and adsorbates. a \u003c/strong\u003eTPD of formate adsorbed on Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e and pure Cu or ZnO. Signals from Cu and ZnO are scaled to the same area as in the Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e sample. \u003cstrong\u003eb \u003c/strong\u003eDFT calculations of how much formate on Cu is stabilized by charge transfer from a row of adsorbed H on the ZnO. \u003cstrong\u003ec\u003c/strong\u003e IR spectra of adsorbed CO on Pt samples and deconvolution from DFT calculations. \u003cstrong\u003ed \u003c/strong\u003eCalculated C-O stretching frequency as a function of the charge on the Pt\u003csub\u003e54\u003c/sub\u003e particle with a full Pt/ZnO() system (orange) or e\u003csup\u003e-\u003c/sup\u003e added to a free Pt cluster (blue). Pt: grey, other colours as Fig. 1.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6939552/v1/9df533313f8d039439d35e88.png"},{"id":102491458,"identity":"0a981214-252d-47f6-9ee3-5aa90541ef3f","added_by":"auto","created_at":"2026-02-12 08:43:19","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":2232307,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCharge transfer across the metal/oxide interface determines the rate of CO\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e hydrogenation to methanol. a \u003c/strong\u003ePotential energy diagram for the conversion of CO\u003csub\u003e2\u003c/sub\u003e via (HCO)-(OH) scission on the edge of the Cu rod/ZnO() system without H on the ZnO (blue) and with H on the ZnO (green) and images of initial, transition and final states. Colours as Fig. 1. \u003cstrong\u003eb\u003c/strong\u003e Energy diagram for (HCO)-(OH) scission on a 4-layer p(1x4) Cu(211) surface with/without an electron added to the unit cell and images of initial and final states. Colours as Fig. 1.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6939552/v1/97f598c19255c5c7e2155055.png"},{"id":103049262,"identity":"ec573f53-e696-4855-9ab8-ba306fa7eab2","added_by":"auto","created_at":"2026-02-20 07:39:16","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":10948933,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6939552/v1/e0fe74d4-b63e-43c2-88b2-2f4bc53ff3e1.pdf"},{"id":102491329,"identity":"ef932e99-5d85-4586-8b59-26279afa22e6","added_by":"auto","created_at":"2026-02-12 08:43:08","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":24819791,"visible":true,"origin":"","legend":"Supplementary information","description":"","filename":"SIrevised.docx","url":"https://assets-eu.researchsquare.com/files/rs-6939552/v1/abbc7aa5620b8e8e4c60e71d.docx"},{"id":102491461,"identity":"57c2bf7a-3953-433d-9714-286169e12445","added_by":"auto","created_at":"2026-02-12 08:43:20","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":13921548,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigures.docx","url":"https://assets-eu.researchsquare.com/files/rs-6939552/v1/9a90aae76ab41e08ac452d19.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"\u003cp\u003eCharge transfer across the metal/oxide interface determines the rate of CO\u003csub\u003e2\u003c/sub\u003e hydrogenation to methanol over Cu/ZnO catalysts\u003c/p\u003e","fulltext":[{"header":"Full Text","content":"\u003cp\u003eMethanol is synthesised from CO/CO\u003csub\u003e2\u003c/sub\u003e/H\u003csub\u003e2\u0026nbsp;\u003c/sub\u003emixtures\u003csub\u003e\u0026nbsp;\u003c/sub\u003eat 50-100 bar and 220-280 ℃.\u003csup\u003e10\u003c/sup\u003e The process relies on Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e catalysts with Cu nanoparticles (typically 5-10 nm) dispersed in a matrix of ZnO nanoparticles (the support) with low levels of Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e as a structural stabilizer.\u003csup\u003e10\u0026ndash;13\u003c/sup\u003e\u003csup\u003e\u0026nbsp;\u003c/sup\u003eMethanol is formed from CO\u003csub\u003e2\u003c/sub\u003e and H\u003csub\u003e2\u003c/sub\u003e,\u003csup\u003e14,15\u003c/sup\u003e while CO removes the inhibiting\u003csup\u003e16\u003c/sup\u003e co-produced water by water-gas shift\u003csup\u003e4\u003c/sup\u003e. The reaction rate scales linearly with the Cu surface area (Fig. E1a), but the TOF for Cu/ZnO is one order of magnitude higher than for unsupported Cu (Fig. E1b), and the underlying mechanisms of this synergy and of the reaction steps remain poorly understood. In situ infrared (IR) spectroscopy shows that formate (HCOO) is present on the metallic surface (Fig. E1c) and involved in the reaction (Fig. E1d). When conducting CO\u003csub\u003e2\u003c/sub\u003e hydrogenation with a 9:1 H\u003csub\u003e2\u003c/sub\u003e:D\u003csub\u003e2\u003c/sub\u003e mixture, where DCOO is enriched in the adsorbed formate\u003csup\u003e17\u003c/sup\u003e, the enrichment of methanol molecules with a characteristic C-D bond matches the enrichment of adsorbed DCOO (Fig. 1a), thus proving that methanol is formed from adsorbed formate. The activation of CO\u003csub\u003e2\u003c/sub\u003e to HCOO through an Eley-Rideal mechanism\u003csup\u003e18\u003c/sup\u003e is fast compared to methanol synthesis and can be regarded as quasi equilibrated (Fig. E1e), meaning that the rate-limiting step occurs later in the mechanism. To identify where formate is located on the catalyst surface, we pre-adsorbed DCOO (from DCOOD) on unsupported Cu and Pt and on 10 wt% Pt/ZnO. We conducted a temperature programmed hydrogenation (TPH) of the adsorbed DCOO into CDH\u003csub\u003e2\u003c/sub\u003eOH (Fig. 1b). As unsupported Pt only forms small amounts of methanol (Fig. 1b), the Pt/ZnO system gives the reactivity for metal-assisted hydrogenation of formate on ZnO. The results in Fig. 1b show that even without a synergy between Cu and ZnO, the emergence of methanol formation on Cu occurs at a 55 \u0026deg;C lower temperature than formate on ZnO. This corresponds to an activation energy difference of 0.1 eV that amounts to a 10-fold higher rate from HCOO-Cu compared to HCOO-ZnO at typical reaction temperatures, indicating that reaction on the metal surface is responsible for at least 90% of the turnovers. Further TPH experiments described in the supplementary information confirms that HCOO-ZnO represents the least active species in Cu/ZnO samples. Metal-assisted hydrogenation of formates on ZnO thus represents one type of synergy but this is only responsible for a minority of turnovers on Cu/ZnO and a second and more important synergy creates an accelerated reaction on the metallic surface. Recent computational work\u003csup\u003e19\u0026ndash;21\u003c/sup\u003e suggests that formate is converted via formic acid. This is validated by Fig. 1c, which shows that the conversion of HCOO and HCOOH becomes possible at the same temperature when formic acid pulses are injected during TPH of pre-adsorbed formate. Their conversions must consequently share the same rate-limiting step, which must be the later step, namely HCOOH conversion. As the HCOOH conversion is independent of the surrounding gas, the reaction does not revert to HCOO and then hydrogenate to methanol from there (Fig. 1d). To identify the dominant HCOOH conversion pathway, we injected a pulse of DCOOD into a flow of H\u003csub\u003e2\u003c/sub\u003e at 200 ℃ passing over Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e. Fig. 1e shows that this produces CDH\u003csub\u003e2\u003c/sub\u003eOH as the dominant methanol isotope. The major preservation of the C-D bond from the original DCOOD despite the possibilities for H/D scrambling indicates that the conversion of formic acid proceeds through the scission of the (HCO)-(OH) bond into an HCO moiety that is then hydrogenated to methanol. Fig. 1d shows that the methanol yield from a formic acid pulse depends on the formic acid isotope but not on the use of D\u003csub\u003e2\u003c/sub\u003e or H\u003csub\u003e2\u003c/sub\u003e. The independence of D\u003csub\u003e2\u003c/sub\u003e/H\u003csub\u003e2\u003c/sub\u003e suggests that only formic acid is involved in the rate-limiting step, ruling out the previously most studied paths\u003csup\u003e10,22,23\u003c/sup\u003e via H\u003csub\u003e2\u003c/sub\u003eCOOH,\u003csup\u003e\u0026nbsp;\u003c/sup\u003eH\u003csub\u003e2\u003c/sub\u003eCOO or COOH, as illustrated in Fig. E2. Consistent with a rate-limiting HCOOH scission the first order H\u003csub\u003e2\u003c/sub\u003e dependence of the reaction also supports that it requires two protonations of CO\u003csub\u003e2\u003c/sub\u003e to pass the rate-limiting step (Fig. E1f). Dosing formaldehyde (CH\u003csub\u003e2\u003c/sub\u003eO) during CO\u003csub\u003e2\u003c/sub\u003e hydrogenation greatly accelerates the reaction, which shows that once protonated HCO and H\u003csub\u003e2\u003c/sub\u003eCO species have been formed, the remaining pathway to methanol is fast (Fig. E1g). The proposed mechanism also explains the autocatalytic behaviour in the presence of methanol\u003csup\u003e16\u003c/sup\u003e as esterification of HCOOH means that a weaker\u003csup\u003e24,25\u003c/sup\u003e (HCO)-(OCH\u003csub\u003e3\u003c/sub\u003e) bond needs to be broken. We can thus identify the unimolecular (HCO)-(OH) scission into HCO and OH on Cu as the dominant methanol forming pathway, and the key to understanding the Cu-ZnO synergy is to understand this reaction step (Fig. 1f).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo study the properties of Cu in the catalyst we conducted X-ray photoelectron spectroscopy (XPS) on a Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e sample (11 nm Cu particles) and a pure Cu reference. Fig. 2a shows that after exposure to 1 bar H\u003csub\u003e2\u003c/sub\u003e at 175 ℃, the Cu surface is negatively charged in Cu/ZnO compared to pure Cu as evident from a downshift in electron binding energy. The measured shifts are substantial given the subtlety of Cu binding energy differences (Cu\u003csup\u003e0\u003c/sup\u003e/Cu\u003csup\u003e+\u003c/sup\u003e/Cu\u003csup\u003e2+\u003c/sup\u003e only differ by 0.5 eV). When the H\u003csub\u003e2\u003c/sub\u003e-treated Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e sample is heated in vacuum to 250 ℃ the amount of ad-/absorbed hydrogen in the surface region of ZnO phase is halved (Fig. E3a), and Fig. 2a shows that this makes the Cu phase in Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u0026nbsp;\u003c/sub\u003emore similar\u003csub\u003e\u0026nbsp;\u003c/sub\u003eto pristine Cu metal. This reveals that stored hydrogen on/in the catalyst contributes to the altered state of Cu. Remarkably, this also demonstrates that fundamental properties of a catalyst, such as the charge of the surface atoms, changes dynamically with the conditions. The XPS shifts are also present in an ambient pressure of 14 mbar H\u003csub\u003e2\u003c/sub\u003e at 250 ℃ (Fig. E3b). To understand the reason for the altered nature of a metal in a supported catalyst we prepared and pre-reduced a 2 wt% Pt/ZnO sample, which due to platinum\u0026rsquo;s nobility can be transferred through air to ex situ analysis by transmission electron microscopy (TEM, Fig. 2b) and XPS (Fig. 2c). The TEM results in Fig. 2b show that the Pt nanoparticles are not alloyed with Zn or decorated by ZnO layers via so-called strong metal support interactions (SMSI). The XPS results in Fig. 2c again show that the metal surface is negatively charged compared to pure Pt. From the XPS-results in Fig. 2a we can also conclude that for metal/ZnO systems, the presence of adsorbed hydrogen plays a major role in this phenomenon. We analysed the hydrogen uptake in 1 bar H\u003csub\u003e2\u003c/sub\u003e at 175 ℃ for a series of Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e samples, and Fig. 2d shows that the ability of ZnO to take up hydrogen increases by orders of magnitude when Cu and ZnO are in contact. TPD was used to distinguish between adsorption on Cu and ZnO (see Fig. E3c). IR spectroscopy during H\u003csub\u003e2\u003c/sub\u003e/D\u003csub\u003e2\u003c/sub\u003e exchange shows that uptake rate of atomic hydrogen is kinetically assisted by spillover\u003csup\u003e26\u003c/sup\u003e from the metal (Fig. E3d), but the increased uptake for Cu/ZnO samples in Fig. 2d is a thermodynamic effect, as evident from long-term tests to establish the equilibrium coverage (Fig. E3e). Density functional theory (DFT) calculations predict that heterolytic (H\u003csub\u003e2\u003c/sub\u003e\u0026rarr;H\u003csup\u003e-\u003c/sup\u003e+H\u003csup\u003e+\u003c/sup\u003e) adsorption is associated with a small adsorption energy (E\u003csub\u003eads\u003c/sub\u003e(2H) = -0.05 eV)\u0026nbsp;on a ZnO(\u0026nbsp;) surface (Fig. E3e). By contrast, the adsorption of hydrogen on\u0026nbsp;ZnO(\u0026nbsp;) next to a supported Cu particle prefers homolytic adsorption (H\u003csub\u003e2\u003c/sub\u003e\u0026rarr;2H\u003csup\u003e+\u003c/sup\u003e+2e\u003csup\u003e-\u003c/sup\u003e) with a high energy of E\u003csub\u003eads\u003c/sub\u003e(2H) = -1.84 eV (Fig. 2e). Owing to the strong bonding of H to ZnO in the presence of Cu, H is stable in vacuum at room temperature (Fig 2a) and at elevated temperatures in the presence of H\u003csub\u003e2\u003c/sub\u003e (Fig. 2d), but not at elevated temperatures in vacuum (Fig. E3a). The DFT calculations in Fig. 2e show that adsorption of H on a ZnO( ) surface containing a cluster of Pt or Cu is associated with H\u003csup\u003e+\u003c/sup\u003e/e\u003csup\u003e-\u003c/sup\u003e charge separation and an electron transfer to the metal. The electron transfer to the metal cluster is shown to increase with the amount of adsorbed hydrogen (Fig. 2e). Hydrogen is a shallow donor in ZnO\u003csup\u003e27\u003c/sup\u003e, and the energy diagram\u003csup\u003e27\u0026ndash;29\u003c/sup\u003e shows\u003csup\u003e\u0026nbsp;\u003c/sup\u003ethat the strong bonding of H for Cu/ZnO is the result of the exothermic electron transfer to the metal (Fig. 2f). The DFT calculations show that the stabilisation of adsorbed H is most pronounced for adsorption at the interface. However, stabilisation is present also for H adsorbed further away from the metal particle and H absorption into the bulk of ZnO close to the metal particles (Fig. E3f). Analyses in the supporting information show that different metal cluster sizes or ZnO terminations and the existence of CuZn alloying do not change the conclusions regarding the charge transfer. The analyses also show that although a minor charge transfer between adjacent phases can occur without the presence of donors, it is the presence of donors, such as hydrogen in ZnO, that mainly determines the charge transfer. Oxygen vacancies can also contribute to charge transfer\u003csup\u003e29\u003c/sup\u003e, but have a lower driving force for this than hydrogen (Fig. 2f).\u0026nbsp;These results illustrate that H\u003csup\u003e+\u003c/sup\u003e/e\u003csup\u003e-\u0026nbsp;\u003c/sup\u003echarge separation occurs between connected solids as previously observed across thin oxide films\u003csup\u003e30\u003c/sup\u003e and at metal/liquid interfaces\u003csup\u003e31\u0026ndash;33\u003c/sup\u003e. Continuum modelling of a hemispherical particle biased by 1 V relative to the support\u003csup\u003e\u0026nbsp;\u003c/sup\u003eshows that the charge transferred to the metal is distributed on the metal surface and that such a particle becomes surrounded by an electrical field that is stronger at the interface but surrounds the entire particle (Fig. 2g). The uptake of H is driven by a transfer of charge that then distributes on the metal surface, which explains why H uptake correlates with Cu surface area (Fig. 2d).The prior state of the art understanding\u003csup\u003e34,35\u003c/sup\u003e has been that significant electronic metal-support interactions in catalysts were limited to small metal clusters, where most of the metal atoms have an interface to the support. We show that these effects are also relevant for the sizable nanoparticles present in most industrial catalysts. The explanation lies in our discovery that donor/acceptor states from dopants, such as the contained hydrogen in ZnO, represent a larger reservoir for charge transfer than previously realized.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAdsorbates on the catalyst are affected by the charge transfer. TPD experiments show that formate, the major adsorbate during methanol synthesis, is generally stabilized by ca. 0.2 eV on Cu/ZnO samples compared to pure Cu (Fig. 3a). From DFT calculations where the system is modelled as a Cu rod on a ZnO(\u0026nbsp;) surface (Fig. 3b), we can conclude that this is because the charge transferred from donors in the adjacent ZnO phase provides additional electrons that participate in the HCOO-Cu bond. Fig. 3b shows that the stabilizing effect is strongest for adsorbates right at the interface where hydrogen bonds to OH on the support also participates in the stabilisation, but even on the wider surface far away from the interface the adsorbates are also stabilised by ca. 0.2 eV due to electrons from the support. The internal bonds in the adsorbed molecules are also affected. IR spectroscopy reveals that the C-O frequency in CO adsorbed on the 2 wt% Pt/ZnO sample is redshifted by ca. 30-40 cm\u003csup\u003e-1\u003c/sup\u003e compared to pure Pt (Fig. 3b). DFT calculations for a Pt\u003csub\u003e54\u003c/sub\u003e cluster shows that this arises from the negative charge transferred from the support (mainly from H dopants) to the metal. (Fig. 3d). The DFT calculations also show that the modifications of the internal bonds are strongest at the interface but present across the metal surface (Fig. E4). In nature, enzymes are believed to direct chemical reactions by means of charged sites and electric fields, and based on the CO frequency shifts in Fig. 3c the field strengths on this inorganic sample resemble levels estimated\u003csup\u003e36\u0026ndash;38\u003c/sup\u003e at the active sites in enzymes. Inorganic, industrial catalysts with fields extending between the constituents could, thus, have closer analogies to enzymes than previously realized. This reveals new possibilities for the development of inorganic catalysts with inspiration from enzymatic processes (and vice versa).\u003c/p\u003e\n\u003cp\u003eTo evaluate how the rate-limiting (HCO)-(OH) scission on Cu is affected by charge transfer, the Cu/ZnO system was evaluated by the DFT model with a Cu rod on a ZnO(\u0026nbsp;) surface with/without adsorbed H on the ZnO (Fig. 4a) or as a 48 atom Cu(211) slab with/without an artificially added electron (Fig. 4b). Fig. 4 shows the calculated potential energy landscape for the HCOOH dissociation and illustrates the importance of charge transfer for the\u0026nbsp;reaction on the Cu surface. Fig. 4a shows that the barrier is lowered by 0.1 eV when H is adsorbed in ZnO, which corresponds to a one order of magnitude increase in the reaction rate at typical reaction temperatures of 220-280 ℃.\u0026nbsp;Consequently, charge transfer from the support to the metal strongly modifies the reactions occurring on the metal phase, as the rate of a catalytic reaction on the metal surface is essentially determined by the concentration of electron donors on the adjoining support. Fig. 4b shows that adding one electron to a Cu(211) slab causes a considerable 0.4 eV lowering of the barrier for HCOOH dissociation, which verifies that the facile scission is a result of charge transfer. The strong effect also implies that there may be a possibility for even more active catalysts, through enhancements of the charging effects. Fig. 4 shows that the HCOOH mechanism provides a favourable route to formation of formyl (HCO) species, because HCOOH has a Cu-O bound HCO moiety, whereas CO hydrogenation produces formyl from Cu-C bound CO, which can explain the differences in reaction rates from CO and CO\u003csub\u003e2\u003c/sub\u003e.\u003csup\u003e4,15\u003c/sup\u003e Finally, our results also show that the primary location of the turnovers on a catalyst will be a function of the reaction conditions. The sites directly at the interface between the constituent phases experience the strongest modifications, whereas sites on the wider Cu surface experience a smaller but substantial modification (Figs. 2 and 3). Our DFT results show that the stabilisation of adsorbed formate is so strong right at the interface that it is favourable to establish a high coverage (Fig. E5). The high coverage precludes the non-adjacent sites required for the products in the rate-limiting step (see Fig. 4). Under conditions favouring such coverages, the sites immediately at the interface will play a limited role and the more moderately affected sites further into the Cu particle will dominate turnovers. At industrial high pressures favouring high coverages the TOF for methanol synthesis is lower for smaller Cu particles\u003csup\u003e39\u003c/sup\u003e, even though smaller particles have relatively\u0026nbsp;more metal-oxide interface sites at the periphery\u003csup\u003e40\u003c/sup\u003e, and this indication of a limited role of the sites right at the interface is explained by our results. The effects reported here are consistent with all the previous suggestions advanced to explain the role of structural features in Cu/ZnO catalysts, namely structure sensitivities, alloying effects and oxide decoration of the metal in SMSI effects. Step sites on Cu have been reported\u003csup\u003e10\u003c/sup\u003e to be more active, and Fig. 4b shows that charge transfer effects are also very important for step sites. Steps and edges also become focal points for electrical fields from charged crystals\u003csup\u003e41\u003c/sup\u003e. Encapsulation of metal particles by SMSI can occur\u003csup\u003e42\u003c/sup\u003e, but the industrial suppliers of Cu/ZnO catalysts report\u003csup\u003e2,12,43\u003c/sup\u003e that no significant SMSI effects occur under typical reaction conditions. However, the electron transfer effects reported here would be important regardless of the existence of partial encapsulation. In fact, the strong electronic interactions could be a major factor influencing the morphology of catalytic materials causing SMSI and restructuring. The Cu-ZnO synergy can be observed on single crystal model systems\u003csup\u003e3,5\u003c/sup\u003e, which consist of 3D islands of ZnO on the Cu surface\u003csup\u003e44\u003c/sup\u003e, and such a geometry would also be affected by the charge transfer dynamics observed here. As noted, the electron transfer effects occur regardless of the existence of CuZn alloying, and the electronic interactions can account for the synergy without invoking alloying. The extent of CuZn-alloying in Cu/ZnO catalysts depends on the presence of CO in the atmosphere\u003csup\u003e5,45,46\u003c/sup\u003e whereas the rate enhancement from ZnO is independent of CO\u003csup\u003e4\u003c/sup\u003e. There is thus no direct evidence of a positive effect of alloying, although not all influences of alloying can be clarified with the presently available information.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn conclusion, we have elucidated the mechanism and active sites for the hydrogenation of CO\u003csub\u003e2\u003c/sub\u003e into methanol with Cu/ZnO catalysts and identified the electronic interactions between the Cu and ZnO phases that cause a synergistic acceleration of the reaction. The significance of the charge transfer effects identified here implies that an accurate representation of these is essential for understanding reactions on multicomponent catalysts and explains why Cu almost behaves as a different element in the Cu/ZnO catalyst. The ability to transfer charge between adjoining phases in nanomaterials and thereby regulate the concentration of donor/acceptor states may not only be essential for heterogeneous catalysts but also for many other applications ranging from ad-/absorbent materials to electronic components.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the Villum Foundation Center for Science of Sustainable Fuels and Chemicals, Villum Foundation V-SUSTAIN Grant 9455. We would also like to acknowledge the Pioneer Center of Accelerating P2X Materials Discovery (CAPeX), DNRF grant P3, for support. Financial support from the Swedish Research Council (2020-05191) is acknowledged. The Competence Centre for Catalysis (KCK) is hosted by Chalmers University of Technology and is financially supported by the Swedish Energy Agency and the member companies Johnson Matthey, Perstorp, Powercell, Preem, Scania CV, Umicore, and Volvo Group. The calculations have been performed at PDC and NSC via a NAISS grant (NAISS 2025/3-36). We acknowledge the MAX IV Laboratory for beamtime on the HIPPI beamline under proposal 20252200.\u0026nbsp;Research conducted at MAX IV, a Swedish national user facility, is supported by Vetenskapsrådet (Swedish Research Council, VR) under contract 2018-07152, Vinnova (Swedish Governmental Agency for Innovation Systems) under contract 2018-04969 and Formas under contract 2019-02496.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eContributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eJAB, JMC, TEA, NDN, JLF, LB-E: Catalytic experiments/analysis. RS\u0026amp;HG: DFT calculations/analysis. ML\u0026amp;CDD: TEM experiments/analysis, JMC, AS, JAB, J-CG: XPS measurements/analysis. HG, ADJ\u0026amp;JMC: acquiring funding, research supervision and planning. JMC: Original idea, initiation of research experiments and data analysis. RS, HG, ADJ, JMC\u0026amp;JAB: First draft of manuscript. All: comments on manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBaltes, C., Vukojević, S. \u0026amp; Sch\u0026uuml;th, F. Correlations between synthesis, precursor, and catalyst structure and activity of a large set of CuO/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e catalysts for methanol synthesis. \u003cem\u003eJ. Catal.\u003c/em\u003e 258, 334\u0026ndash;344 (2008).\u003c/li\u003e\n\u003cli\u003eBarrow, N. \u003cem\u003eet al.\u003c/em\u003e Doubling the life of Cu/ZnO methanol synthesis catalysts via use of Si as a structural promoter to inhibit sintering. \u003cem\u003eSci. Adv.\u003c/em\u003e 10, eadk2081 (2024).\u003c/li\u003e\n\u003cli\u003eNakamura, I., Fujitani, T., Uchijima, T. \u0026amp; Nakamura, J. A model catalyst for methanol synthesis: Zn‐deposited and Zn‐free Cu surfaces. \u003cem\u003eJ. Vac. Sci. Technol. A\u003c/em\u003e 14, 1464 (1998).\u003c/li\u003e\n\u003cli\u003eNielsen, N. D., Jensen, A. D. \u0026amp; Christensen, J. M. The roles of CO and CO\u003csub\u003e2\u003c/sub\u003e in high pressure methanol synthesis over Cu-based catalysts. \u003cem\u003eJ. Catal.\u003c/em\u003e 393, 324\u0026ndash;334 (2021).\u003c/li\u003e\n\u003cli\u003eKattel, S., Ram\u0026iacute;rez, P. J., Chen, J. G., Rodriguez, J. A. \u0026amp; Liu, P. Active sites for CO\u003csub\u003e2\u003c/sub\u003e hydrogenation to methanol on Cu/ZnO catalysts. \u003cem\u003eScience \u003c/em\u003e355, 1296\u0026ndash;1299 (2017).\u003c/li\u003e\n\u003cli\u003eGraciani, J. \u003cem\u003eet al.\u003c/em\u003e Highly active copper-ceria and copper-ceria-titania catalysts for methanol synthesis from CO\u003csub\u003e2\u003c/sub\u003e. \u003cem\u003eScience \u003c/em\u003e345, 546\u0026ndash;550 (2014).\u003c/li\u003e\n\u003cli\u003ePacchioni, G. From CO\u003csub\u003e2\u003c/sub\u003e to Methanol on Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e Industrial Catalyst. What Do We Know about the Active Phase and the Reaction Mechanism? \u003cem\u003eACS Catal.\u003c/em\u003e 14, 2730\u0026ndash;2745 (2024).\u003c/li\u003e\n\u003cli\u003eBeck, A., Newton, M. A., van de Water, L. G. A. \u0026amp; van Bokhoven, J. A. The Enigma of Methanol Synthesis by Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e-Based Catalysts. \u003cem\u003eChem. Rev.\u003c/em\u003e 124, 4543\u0026ndash;4678 (2024).\u003c/li\u003e\n\u003cli\u003eYe, J. \u003cem\u003eet al.\u003c/em\u003e Hydrogenation of CO\u003csub\u003e2\u003c/sub\u003e for sustainable fuel and chemical production. \u003cem\u003eScience \u003c/em\u003e387, eadn9388 (2025).\u003c/li\u003e\n\u003cli\u003eBehrens, M. \u003cem\u003eet al.\u003c/em\u003e The active site of methanol synthesis over Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e industrial catalysts. \u003cem\u003eScience \u003c/em\u003e336, 893\u0026ndash;897 (2012).\u003c/li\u003e\n\u003cli\u003eSchl\u0026ouml;gl, R. Chemical Batteries with CO\u003csub\u003e2\u003c/sub\u003e. \u003cem\u003eAngew. Chem. Intl. Ed. \u003c/em\u003e61, e202007397 (2022).\u003c/li\u003e\n\u003cli\u003eKuld, S. \u003cem\u003eet al.\u003c/em\u003e Quantifying the promotion of Cu catalysts by ZnO for methanol synthesis. \u003cem\u003eScience \u003c/em\u003e352, 969\u0026ndash;974 (2016).\u003c/li\u003e\n\u003cli\u003eKondrat, S. A. \u003cem\u003eet al.\u003c/em\u003e Stable amorphous georgeite as a precursor to a high-activity catalyst. \u003cem\u003eNature \u003c/em\u003e531, 83\u0026ndash;87 (2016).\u003c/li\u003e\n\u003cli\u003eChinchen, G. C., Denny, P. J., Parker, D. G., Spencer, M. S. \u0026amp; Whan, D. A. Mechanism of methanol synthesis from CO\u003csub\u003e2\u003c/sub\u003e/CO/H\u003csub\u003e2\u003c/sub\u003e mixtures over copper/zinc oxide/alumina catalysts: use of \u003csup\u003e14\u003c/sup\u003eC-labelled reactants. \u003cem\u003eAppl. Catal.\u003c/em\u003e 30, 333\u0026ndash;338 (1987).\u003c/li\u003e\n\u003cli\u003eNielsen, N. D., Thrane, J., Jensen, A. D. \u0026amp; Christensen, J. M. Bifunctional Synergy in CO Hydrogenation to Methanol with Supported Cu. \u003cem\u003eCatal. Lett.\u003c/em\u003e 150, 1427\u0026ndash;1433 (2020).\u003c/li\u003e\n\u003cli\u003eThrane, J. \u003cem\u003eet al.\u003c/em\u003e Methanol-Assisted Autocatalysis in Catalytic Methanol Synthesis. \u003cem\u003eAngew. Chem. Intl. Ed.\u003c/em\u003e 59, 18189\u0026ndash;18193 (2020).\u003c/li\u003e\n\u003cli\u003eKunkes, E. L., Studt, F., Abild-Pedersen, F., Schl\u0026ouml;gl, R. \u0026amp; Behrens, M. Hydrogenation of CO2 to methanol and CO on Cu/ZnO/Al2O3: Is there a common intermediate or not? \u003cem\u003eJ. Catal.\u003c/em\u003e 328, 43\u0026ndash;48 (2015).\u003c/li\u003e\n\u003cli\u003eQuan, J. \u003cem\u003eet al.\u003c/em\u003e Vibration-driven reaction of CO\u003csub\u003e2\u003c/sub\u003e on Cu surfaces via Eley\u0026ndash;Rideal-type mechanism. \u003cem\u003eNat. Chem. \u003c/em\u003e11, 722\u0026ndash;729 (2019).\u003c/li\u003e\n\u003cli\u003eHigham, M. D. \u003cem\u003eet al.\u003c/em\u003e Mechanism of CO\u003csub\u003e2\u003c/sub\u003e conversion to methanol over Cu(110) and Cu(100) surfaces. \u003cem\u003eDalton Transactions\u003c/em\u003e 49, 8478\u0026ndash;8497 (2020).\u003c/li\u003e\n\u003cli\u003eGrabow, L. C. \u0026amp; Mavrikakis, M. Mechanism of methanol synthesis on cu through CO\u003csub\u003e2\u003c/sub\u003e and CO hydrogenation. \u003cem\u003eACS Catal.\u003c/em\u003e 1, 365\u0026ndash;384 (2011).\u003c/li\u003e\n\u003cli\u003eCao, A., Wang, Z., Li, H., Elnabawy, A. O. \u0026amp; N\u0026oslash;rskov, J. K. New insights on CO and CO\u003csub\u003e2\u003c/sub\u003e hydrogenation for methanol synthesis: The key role of adsorbate-adsorbate interactions on Cu and the highly active MgO-Cu interface. \u003cem\u003eJ. Catal.\u003c/em\u003e 400, 325\u0026ndash;331 (2021).\u003c/li\u003e\n\u003cli\u003eTakeyasu, K. \u003cem\u003eet al.\u003c/em\u003e Hydrogenation of Formate Species Using Atomic Hydrogen on a Cu(111) Model Catalyst. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e 144, 12158\u0026ndash;12166 (2022).\u003c/li\u003e\n\u003cli\u003eZhao, Y. F. \u003cem\u003eet al.\u003c/em\u003e Insight into methanol synthesis from CO\u003csub\u003e2\u003c/sub\u003e hydrogenation on Cu(111): Complex reaction network and the effects of H\u003csub\u003e2\u003c/sub\u003eO. \u003cem\u003eJ. Catal.\u003c/em\u003e 281, 199\u0026ndash;211 (2011).\u003c/li\u003e\n\u003cli\u003eBlanksby, S. J. \u0026amp; Ellison, G. B. Bond Dissociation Energies of Organic Molecules. \u003cem\u003eAcc. Chem. Res.\u003c/em\u003e 36, 255\u0026ndash;263 (2003).\u003c/li\u003e\n\u003cli\u003eMetcalfe, W. K., Simmie, J. M. \u0026amp; Curran, H. J. Ab Initio Chemical Kinetics of Methyl Formate Decomposition: The Simplest Model Biodiesel. \u003cem\u003eJ. Phys. Chem. A\u003c/em\u003e 114, 5478\u0026ndash;5484 (2010).\u003c/li\u003e\n\u003cli\u003eKarim, W. \u003cem\u003eet al.\u003c/em\u003e Catalyst support effects on hydrogen spillover. \u003cem\u003eNature\u003c/em\u003e 541, 68\u0026ndash;71 (2017).\u003c/li\u003e\n\u003cli\u003eVan de Walle, C. G. \u0026amp; Neugebauer, J. Universal alignment of hydrogen levels in semiconductors, insulators and solutions. \u003cem\u003eNature \u003c/em\u003e423, 626\u0026ndash;628 (2003).\u003c/li\u003e\n\u003cli\u003eAllen, M. W. \u0026amp; Durbin, S. M. Influence of oxygen vacancies on Schottky contacts to ZnO. \u003cem\u003eAppl. Phys. Lett.\u003c/em\u003e 92, 122110 (2008).\u003c/li\u003e\n\u003cli\u003eFrost, J. C. Junction effect interactions in methanol synthesis catalysts. \u003cem\u003eNature \u003c/em\u003e334, 577\u0026ndash;580 (1988).\u003c/li\u003e\n\u003cli\u003eHonkala, K., Hellman, A. \u0026amp; Gr\u0026ouml;nbeck, H. Water Dissociation on MgO/Ag(100): Support Induced Stabilization or Electron Pairing? \u003cem\u003eJ. Phys. Chem. C\u003c/em\u003e 114, 7070\u0026ndash;7075 (2010).\u003c/li\u003e\n\u003cli\u003eWestendorff, K. S., H\u0026uuml;lsey, M. J., Wesley, T. S., Rom\u0026aacute;n-Leshkov, Y. \u0026amp; Surendranath, Y. Electrically driven proton transfer promotes Br\u0026oslash;nsted acid catalysis by orders of magnitude. \u003cem\u003eScience \u003c/em\u003e383, 757\u0026ndash;763 (2024).\u003c/li\u003e\n\u003cli\u003eWesley, T. S., Rom\u0026aacute;n-Leshkov, Y. \u0026amp; Surendranath, Y. Spontaneous Electric Fields Play a Key Role in Thermochemical Catalysis at Metal-Liquid Interfaces. \u003cem\u003eACS Cent. Sci.\u003c/em\u003e 7, 1045\u0026ndash;1055 (2021).\u003c/li\u003e\n\u003cli\u003eSvensson, R. \u0026amp; Gr\u0026ouml;nbeck, H. Spontaneous Charge Separation at the Metal-Water Interface. \u003cem\u003eChemPhysChem\u003c/em\u003e 25, e202400099 (2024).\u003c/li\u003e\n\u003cli\u003eXu, M. \u003cem\u003eet al.\u003c/em\u003e Renaissance of Strong Metal\u0026ndash;Support Interactions. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e 146, 2290\u0026ndash;2307 (2024).\u003c/li\u003e\n\u003cli\u003eLykhach, Y. \u003cem\u003eet al.\u003c/em\u003e Counting electrons on supported nanoparticles. \u003cem\u003eNat. Mater. \u003c/em\u003e15, 284\u0026ndash;288 (2015).\u003c/li\u003e\n\u003cli\u003eWelborn, V. V., Ruiz Pestana, L. \u0026amp; Head-Gordon, T. Computational optimization of electric fields for better catalysis design. \u003cem\u003eNat. Catal. \u003c/em\u003e1, 649\u0026ndash;655 (2018).\u003c/li\u003e\n\u003cli\u003eZheng, C., Ji, Z., Mathews, I. I. \u0026amp; Boxer, S. G. Enhanced active-site electric field accelerates enzyme catalysis. \u003cem\u003eNat. Chem. \u003c/em\u003e15, 1715\u0026ndash;1721 (2023).\u003c/li\u003e\n\u003cli\u003eFried, S. D., Bagchi, S. \u0026amp; Boxer, S. G. Extreme electric fields power catalysis in the active site of ketosteroid isomerase. \u003cem\u003eScience \u003c/em\u003e346, 1510\u0026ndash;1514 (2014).\u003c/li\u003e\n\u003cli\u003eVan Den Berg, R. \u003cem\u003eet al.\u003c/em\u003e Structure sensitivity of Cu and CuZn catalysts relevant to industrial methanol synthesis. \u003cem\u003eNat. Commun. \u003c/em\u003e7, 13057 (2016).\u003c/li\u003e\n\u003cli\u003eCargnello, M. \u003cem\u003eet al.\u003c/em\u003e Control of metal nanocrystal size reveals metal-support interface role for ceria catalysts. \u003cem\u003eScience \u003c/em\u003e341, 771\u0026ndash;773 (2013).\u003c/li\u003e\n\u003cli\u003eLaCount, M. D., Lambeets, S. V., Perea, D. E., Prozorov, T. \u0026amp; Kathmann, S. M. Electric Fields at Interfaces. \u003cem\u003eJ. Phys. Chem. C\u003c/em\u003e 129, 15489\u0026ndash;15506 (2025).\u003c/li\u003e\n\u003cli\u003eLunkenbein, T. \u003cem\u003eet al.\u003c/em\u003e Formation of a ZnO Overlayer in Industrial Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e Catalysts Induced by Strong Metal\u0026ndash;Support Interactions. \u003cem\u003eAngew. Chem. Intl. Ed.\u003c/em\u003e 54, 4544-4548 (2015).\u003c/li\u003e\n\u003cli\u003eHansen, P. L. \u003cem\u003eet al.\u003c/em\u003e Atom-resolved imaging of dynamic shape changes in supported copper nanocrystals. \u003cem\u003eScience \u003c/em\u003e295, 2053\u0026ndash;2055 (2002).\u003c/li\u003e\n\u003cli\u003eOrozco, I. \u003cem\u003eet al.\u003c/em\u003e Understanding Methanol Synthesis on Inverse ZnO/CuO\u003csub\u003ex\u003c/sub\u003e/Cu Catalysts: Stability of CH\u003csub\u003e3\u003c/sub\u003eO Species and Dynamic Nature of the Surface. \u003cem\u003eJ. Phys. Chem. C\u003c/em\u003e 125, 6673\u0026ndash;6683 (2021).\u003c/li\u003e\n\u003cli\u003eAmann, P. \u003cem\u003eet al.\u003c/em\u003e The state of zinc in methanol synthesis over a Zn/ZnO/Cu(211) model catalyst. \u003cem\u003eScience \u003c/em\u003e376, 603\u0026ndash;608 (2022).\u003c/li\u003e\n\u003cli\u003eBeck, A. \u003cem\u003eet al.\u003c/em\u003e Following the structure of copper-zinc-alumina across the pressure gap in carbon dioxide hydrogenation. \u003cem\u003eNat. Catal. \u003c/em\u003e4, 488\u0026ndash;497 (2021).\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cem\u003e2.1 Catalyst synthesis\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eCu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e, Cu/ZnO and unsupported ZnO were made by the precipitation method of Baltes et al.\u003csup\u003e1\u003c/sup\u003e as described elsewhere\u003csup\u003e15\u003c/sup\u003e. We varied the Cu surface area by varying the composition, the addition rate of the precursors and by calcining the samples to different temperatures. A list of preparation conditions and sample properties are given in the supplementary information. We also tested that a commercial methanol synthesis catalyst from Topsoe to ensure that it was on the same rate vs Cu area line in Fig. E1a as our in house samples. By agreement with the supplier, we do not disclose the sample product code. Our unsupported Cu sample was a Raney\u003csup\u003eTM\u003c/sup\u003e-type, sponge Cu purchased from Strem Chemicals. Our handling of the sample and further characterizations of this material can be found elsewhere\u003csup\u003e4,15,51,52\u003c/sup\u003e. The sample contains a small concentration of Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e as a structural stabilizer. For spectroscopic tests we also used a commercial CuO sample (Sigma Aldrich, nanopowder). Pt samples were synthesized by impregnating Pt on ZnO or SiO\u003csub\u003e2\u003c/sub\u003e. Here we used either our precipitated ZnO or commercial ZnO (Sigma-Aldrich, 7.7 m\u003csup\u003e2\u003c/sup\u003e/g, \u0026ge; 99.99% trace metal basis) and SiO\u003csub\u003e2\u003c/sub\u003e (SS61138, 250 m\u003csup\u003e2\u003c/sup\u003e/g, Saint Gobain), see supplementary information for details. Platinum(II)2,4-pentanedionate (Alfa Aesar, \u0026ge; 99.977%) was dissolved in a small excess of ethanol in a beaker on a heating plate, then the support was added followed by gradual evaporation of ethanol. The standard calcination of the samples was heating (1 \u0026deg;C/min) to 330 \u0026deg;C in an air flow (1 NL/min) and holding this temperature for 3 hr. All samples were pelletized/crushed/sieved to obtain a particle size of 150-300 \u0026mu;m. As a pure Pt reference we used a commercial platinum black sample (Merck, \u0026ge; 99.95%).\u003c/p\u003e\n\u003cp\u003e\u003cbr /\u003e\u003cem\u003e2.2 Experimental setup and gases\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eMethanol synthesis and temperature programmed desorption (TPD) experiments were performed in a high-pressure reactor setup described elsewhere\u003csup\u003e15\u003c/sup\u003e. Gases used in these experiments are 5% H\u003csub\u003e2\u003c/sub\u003e/Ar (both 99.999%), 5000 ppmv CO\u003csub\u003e2\u003c/sub\u003e in N\u003csub\u003e2\u003c/sub\u003e (99.999%), 99.9999% H\u003csub\u003e2\u003c/sub\u003e, 99.9999% Ar, 99.995% CO\u003csub\u003e2\u003c/sub\u003e, 1% N\u003csub\u003e2\u003c/sub\u003eO/He (both 99.999%) and 99.97% CO from Air Liquide as well as D\u003csub\u003e2\u003c/sub\u003e (99.8%-D) from Linde gas.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e2.3 Catalyst reduction and pre-treatments\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eCatalyst samples were reduced in situ using a 100 NmL/min flow of 5% H\u003csub\u003e2\u003c/sub\u003e/Ar at atmospheric pressure, by heating (1 \u0026deg;C/min) to 175 \u0026deg;C and holding this temperature for 2 h. For Pt/ZnO, we also tested reduction with the same program using 10% CO/Ar.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e2.4 Catalytic tests\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eMethanol synthesis experiments were conducted in the high-pressure setup\u003csup\u003e15\u003c/sup\u003e after pre-reduction. Methanol synthesis activity is investigated at low conversion conditions far from equilibrium at mild (T = 135 \u0026deg;C, 1 atm, GHSV = 48000 NmL/g\u003csub\u003ecat\u003c/sub\u003e/h) and industry-type, high-pressure conditions (20 bar, GHSV = 24∙10\u003csup\u003e5\u003c/sup\u003e NmL/g\u003csub\u003ecat\u003c/sub\u003e/h), but still at low conversion with at most 2000 ppmv methanol produced. When measuring reaction orders in hydrogen, the total gas flow is fixed by replacing part of the hydrogen flow with argon.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e2.5 Determination of stored hydrogen in ZnO\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eTPD of adsorbed hydrogen was conducted by treating the pre-reduced sample (typically 250 mg) to 1 bar of H\u003csub\u003e2\u003c/sub\u003e at 175 \u0026deg;C for 1 h (or up to 96 h in the case of pure ZnO) followed by rapid quench cooling as described by Nielsen et al.\u003csup\u003e51\u003c/sup\u003e. The system was then thoroughly flushed with Ar and subjected to a TPD in 50 NmL/min Ar flow with a 2 \u0026deg;C/min heating rate. The H\u003csub\u003e2\u003c/sub\u003e desorption was followed with a mass spectrometer and the MS signal was converted into concentration by calibration against a certified H\u003csub\u003e2\u003c/sub\u003e/Ar gas mixture. The desorption below 100 \u0026deg;C is due to H on Cu metal\u003csup\u003e53\u0026ndash;55\u003c/sup\u003e and the higher temperature desorption is due to H on ZnO\u003csup\u003e56,57\u003c/sup\u003e. Adsorption energies were calculated from the TPD peak temperature using the Redhead equation and an assumed pre-factor of 10\u003csup\u003e13\u003c/sup\u003e s\u003csup\u003e-1\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e2.6 Cu surface area determination \u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe Cu surface area was determined by N\u003csub\u003e2\u003c/sub\u003eO-Reactive frontal chromatography (N\u003csub\u003e2\u003c/sub\u003eO-RFC)\u003csup\u003e58\u003c/sup\u003e at 60 \u0026deg;C using a 15 NmL/min flow of 1 mol% N\u003csub\u003e2\u003c/sub\u003eO/He. The Cu surface area is derived from the N\u003csub\u003e2\u003c/sub\u003eO uptake, a Cu:O stoichiometry of 2:1\u003csup\u003e53\u003c/sup\u003e, and an average Cu surface atom density of 1.47∙10\u003csup\u003e19\u003c/sup\u003e Cu atoms/m\u003csup\u003e2\u003c/sup\u003e. The Cu surface area was determined for Cu-based catalysts reduced in 5% H\u003csub\u003e2\u003c/sub\u003e/Ar at 175 \u0026deg;C. The Cu surface area in catalysts that are exposed to industrial-like reaction conditions (T = 260 \u0026deg;C, P = 20 bar) was determined after normal reduction followed by a heating to 260 \u0026deg;C in Ar (1 \u0026deg;C/min) to capture the sintering of the samples at higher temperatures.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e2.7 Infrared spectroscopy\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eIn situ Infrared (IR) measurements were performed with undiluted samples in a domed reactor cell and Praying Mantis DRIFTS unit from Harrick Scientific Products installed in a Nicolet iS50 FTIR spectrometer equipped with a liquid N\u003csub\u003e2\u003c/sub\u003e cooled MCT detector. Details of the setup are described elsewhere\u003csup\u003e15\u003c/sup\u003e. Each IR spectra is an average of 76 scans with a resolution of 4 cm\u003csup\u003e-1\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eFor the IR measurements on adsorbed CO, the sample cup was filled with the Pt sample (ca 30 mg) and the background was measured in Air without flow through the system at 20 \u0026deg;C. This was done because even background traces of CO remaining in the gas lines were sufficient to cause high CO coverages on the Pt. The sample was then reduced in 100 NmL/min of 5% H\u003csub\u003e2\u003c/sub\u003e/Ar at 1 atm with a ramping rate of 5 \u0026deg;C /min to 175 \u0026deg;C. The temperature was decreased to 20 \u0026deg;C where CO is adsorbed from a flow of 0.64 % CO in Ar. The shown spectra are at 20 \u0026deg;C after the gaseous CO had been purged from the system by Ar flow.\u003c/p\u003e\n\u003cp\u003eFor studies of adsorbates under reaction conditions the samples, either Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e or pure ZnO were pre-reduced and cooled to 135 \u0026deg;C where a background was collected. We then performed either H\u003csub\u003e2\u003c/sub\u003e/D\u003csub\u003e2\u003c/sub\u003e switch experiments or exposed the sample to syngas followed by exchange of H\u003csub\u003e2\u003c/sub\u003e with D\u003csub\u003e2\u003c/sub\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e2.8 Scanning transmission electron microscopy\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eSTEM (scanning transmission electron microscopy) imaging was made of Pt/ZnO pre-reduced elsewhere and transferred through air to the microscope. A Titan Analytical 80-300ST TEM (probe corrected) equipped with an Oxford windowless EDX detector (X-Max 100TLE) was used for these measurements, and the microscope was operated with an acceleration voltage of 300 kV and an electron dose rate of 150 pA. The measurements were conducted at room temperature and ultra-high vacuum (10\u003csup\u003e\u0026minus;10\u003c/sup\u003e bar). Prior to the measurements, a Pt/ZnO sample was placed on a lacey carbon supported Au grid, which was loaded into the microscope in a CuZn or Be holder. Both STEM Energy-dispersive X-ray spectroscopy (EDX) and STEM high angular annular dark field (HAADF) were performed, and the image resolution was 1024x1024 and 512x512, respectively.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e2.9 BET and surface area determinations\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eFor Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e samples, the oxide surface area was determined as the difference between the N\u003csub\u003e2\u003c/sub\u003eO-RFC measurement of the Cu area and the BET measurement of the total surface area. BET analysis was performed on a pre-reduced and N\u003csub\u003e2\u003c/sub\u003eO passivated sample using a Quantacrhome NOVAtouch Gas sorption analyzer. In the NOVAtouch analyzer, the sample was vacuum degassed at 150 \u0026deg;C (10 \u0026deg;C/min) for 16 h before being subjected to a 6-points BET measurement.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e2.10 Ambient pressure X-ray photoelectron spectroscopy\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe experiments have been performed at the Solid-Gas Endstation\u003csup\u003e59\u003c/sup\u003e of the HIPPIE beamline\u003csup\u003e60\u003c/sup\u003e, MAX IV Laboratory, Lund, Sweden. The measurements were of Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e and pure Cu samples at an X-ray energy of 1580 eV. The samples were pre-reduced (1 atm 5% H\u003csub\u003e2\u003c/sub\u003e/Ar for 2 h at 175 \u0026deg;C, then 2 h in 250 \u0026deg;C and finally 2 h in 100% H\u003csub\u003e2\u003c/sub\u003e at 250 \u0026deg;C) in the catalytic reactor, passivated by exposure to 1% N\u003csub\u003e2\u003c/sub\u003eO at 60 \u0026deg;C and transported to the synchrotron in this state. As pure Cu reference, we used either the Raney type Cu used for the catalytic tests or commercial CuO (Sigma Aldrich, nanopowder) and these both gave identical results. At the synchrotron, the samples were dropcast from an ethanol slurry onto sample holders covered by gold foil. The samples were treated in 1 bar H\u003csub\u003e2\u003c/sub\u003e at 175 \u0026deg;C and transferred through vacuum to the measurement cell to produce Fig. 2a. Alternatively, the samples were transferred directly to the measurement cell and then first measured in 14 mbar He at room temperature and then heated to 250 \u0026deg;C in 14 mbar H\u003csub\u003e2\u003c/sub\u003e and measured at this condition. For pure Cu samples, we used the 84 eV 4f\u003csub\u003e7\u003c/sub\u003e\u003csub\u003e/\u003c/sub\u003e\u003csub\u003e2\u003c/sub\u003e binding energy of the underlying Au foil and the 284.8 eV 1s binding energy of adventitious carbon as energy references and these both gave identical results (see supplementary information). For the Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e samples, we used the 1021.73\u0026plusmn;0.37 eV 2p\u003csub\u003e3\u003c/sub\u003e\u003csub\u003e/\u003c/sub\u003e\u003csub\u003e2\u003c/sub\u003e of Zn in ZnO as energy reference. This value was determined as the average of the reference spectra in the NIST database\u003csup\u003e61\u003c/sup\u003e and this is the value measured for Zn in high level XPS studies of Cu/ZnO materials\u003csup\u003e5,45,62\u003c/sup\u003e. Adventitious carbon gave a similar energy calibration and this was used to validate the approach (see supplementary information). This consistency in the energy references ensures the reliability of our reported values. A relaxation model\u003csup\u003e63\u003c/sup\u003e was used to confirm that the observed shifts are intrinsic effects (see supplementary information). All data treatment was made without background subtraction, while for figures a Shirley background has been subtracted for visual clarity.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e2.11 Ex situ X-ray photoelectron spectroscopy\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe XPS measurements of Pt/ZnO, Pt/SiO\u003csub\u003e2\u003c/sub\u003e and Pt black were performed in UHV with an ESCALAB Xi+ instrument from ThermoFisher Scientific, using a monochromatic Al K\u0026alpha; (1486.6 eV) X-ray source. The powdered samples were reduced ex situ and then transferred through air and placed onto an adhesive copper tape (3M Electrical Tape). A low energy electron flood gun was used for the neutralization of surface charge buildup. The binding energies were calibrated by using Au foil. The diameter of the analysis area was approximately 650 \u0026mu;m. In situ sputtering was also used to remove oxygen adsorbed during sample transfer but this had no detectable impact on measured binding energies.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e2.12 Mechanistic investigations\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eTo dose formic acid into the IR cell, we installed a Swagelok \u0026frac14;-inch T fitting on the gas inlet line immediately before the cell. One leg of the T fitting has a GC rubber membrane that allows injection of liquid into the gas stream entering the cell using a GC syringe. We then performed experiments where a sample (unsupported Cu, Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e or ZnO) was loaded into the sample cup and pre-reduced. Formic acid was injected (2 \u0026micro;L = 53 \u0026micro;mol) into a 50 NmL/min Ar flow at atmospheric pressure to pre-adsorb HCOO at 40 \u0026deg;C, the sample was then flushed in Ar flow and a TPD was conducted. We either used HCOOH (99.5%, Fisher Scientific) or DCCOD (\u0026ge; 95% D, Fisher Scientific). The IR spectrometer monitoring the IR cell was used to confirm that adsorbed formate had been produced. To evaluate the formate adsorption energy, the CO\u003csub\u003e2\u003c/sub\u003e concentration during a TPD (50 NmL/min Ar flow, 2 \u0026deg;C/min) was monitored by mass spectrometry, and the MS signal was calibrated against a certified CO\u003csub\u003e2\u003c/sub\u003e/Ar gas mixture from Air Liquide. To compare the isotopic distribution of surface formates and formed methanol, we used the TPD experiments to calculate the relative stability and coverage of HCOO and DCOO (see supplementary information for details). We then conducted methanol synthesis with a D\u003csub\u003e2\u003c/sub\u003e/H\u003csub\u003e2\u003c/sub\u003e/CO\u003csub\u003e2\u003c/sub\u003e gas mixture and measured masses 31-36 by mass spectrometry. The formate coverage was determined from integration of the CO\u003csub\u003e2\u003c/sub\u003e signal during the TPD. Because the appearance energies, ionization potentials and fragmentation patterns are very similar for all the H/D-methanol isotopes\u003csup\u003e64\u0026ndash;66\u003c/sup\u003e it is considered justifiable to treat the fragmentation of mixed isotopes as identical and randomized. This allowed us to deconvolute the mass spectrometry measurements into a distribution of methanol isotopes (see supplementary information for details). We performed TPH measurements of adsorbed DCOO on various samples by first pre-adsorbing DCOO by the method described above and then heating (2 \u0026deg;C/min) in a 50 NmL/min H\u003csub\u003e2\u003c/sub\u003e flow while monitoring the product formation by mass spectrometry. To test if HCOOH conversion is the rate-limiting step in the conversion of HCOO, we first made a TPH of pre-adsorbed HCOO. We then made an identical experiment where we injected another 53 \u0026micro;mol HCOOH pulse at 121 \u0026deg;C before HCOO can be converted and an additional pulse at 53 \u0026micro;mol HCOOH pulse at 205 \u0026deg;C after HCOO can be converted. During TPH of pre-adsorbed HCOO, the majority of the formate decomposes to CO\u003csub\u003e2\u003c/sub\u003e and the second pulse replenishes adsorbed HCOO which is the reason for the higher general methanol formation in the multi-pulse experiment in Fig. 1c. To study formic acid conversion, either formic acid-2H or formic acid-2D were injected into a 100 NmL/min flow of H\u003csub\u003e2\u003c/sub\u003e or D\u003csub\u003e2\u003c/sub\u003e and the gas entered a cell kept at 200 \u0026deg;C filled with pre-reduced Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e or unsupported Cu. We followed the resulting methanol formation by mass spectrometry and GC measurements and used the abovementioned method to deconvolute the methanol isotopes. Here the MS measurements were calibrated against simultaneous GC measurements. As described in the supplementary information, Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e and unsupported Cu both give identical information in the first pulse, but in subsequent pulses Cu/ZnO/Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e shows isotopic scrambling because formed methanol re-adsorbs on ZnO as methoxide that undergoes H/D exchange with the gas phase. Unsupported gave no such signs of readsorption and for that reason the kinetic isotope effect investigations in Fig. 1e were made with unsupported Cu.\u003c/p\u003e\n\u003cp\u003e\u003cbr /\u003e\u003cem\u003e2.13 Computational method\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eDensity functional theory calculations were performed using the Vienna Ab Initio Simulation package VASP\u003csup\u003e67\u0026ndash;70\u003c/sup\u003e. The projector-augmented wave method was applied to describe the interactions between the core- and valence electrons\u003csup\u003e71\u003c/sup\u003e. The considered valence electrons are 1s\u003csup\u003e1\u003c/sup\u003e (H), 2s\u003csup\u003e2\u003c/sup\u003e2p\u003csup\u003e2\u003c/sup\u003e (C), 2s\u003csup\u003e2\u003c/sup\u003e2p\u003csup\u003e4\u003c/sup\u003e (O), 3d\u003csup\u003e10\u003c/sup\u003e4s\u003csup\u003e1\u003c/sup\u003e (Cu), 3d\u003csup\u003e10\u003c/sup\u003e4p\u003csup\u003e2\u003c/sup\u003e (Zn), and 6s\u003csup\u003e1\u003c/sup\u003e5d\u003csup\u003e9\u003c/sup\u003e (Pt). The exchange and correlation functional was described using the generalized gradient approximation proposed by Perdew, Burke, and Ernzerhof (PBE)\u003csup\u003e72\u003c/sup\u003e. To describe the localised electrons in ZnO, a rotationally invariant Hubbard-U correction of 7.5 eV was applied to the 3d electrons in Zn\u003csup\u003e73\u003c/sup\u003e. The plane waves were truncated at 500 eV in the expansion of the Kohn-Sham orbitals. The electronic structure was considered converged when the change in Kohn-Sham eigenvalues and electronic energy, between two succeeding iterations are below 1 \u0026times; 10\u003csup\u003e\u0026minus;6\u003c/sup\u003e eV. The structural optimization was performed using the conjugate-gradient method, and the structures are considered converged when the maximum force acting on any nuclei is below 0.03 eV/\u0026Aring;.\u003c/p\u003e\n\u003cp\u003eThe hexagonal lattice constants of bulk ZnO were calculated to be a = b = 3.19 \u0026Aring; and c = 5.12 \u0026Aring;, slightly smaller than the experimental values of a = b = 3.24 \u0026Aring; and c = 5.18 \u0026Aring;. The ZnO( ) surface was represented by a \u003cem\u003ep\u003c/em\u003e(6x5) cell, constituted of four ZnO layers, of which the bottom two layers were kept fixed to their bulk positions. The periodic ZnO surfaces were separated by a vacuum layer of ~22 \u0026Aring;. The large vacuum layer was included to allow for the inclusion of the supported nanoparticles (NPs). The lattice constants of bulk Cu and bulk Pt were calculated to be 3.64 \u0026Aring; and 3.97 \u0026Aring;, respectively, in fair agreement with the experimental values of 3.61 \u0026Aring; and 3.92 \u0026Aring;, respectively. The k-point sampling was performed using the \u0026Gamma;-point. The energy of gas phase H\u003csub\u003e2\u003c/sub\u003e and CO\u003csub\u003e2\u003c/sub\u003e were calculated using a (30, 31, 32) \u0026Aring; vacuum box. Vibrational modes were calculated using the finite differences method, assuming the harmonic approximation. The analysis of the electron density distribution was investigated with Bader charge analysis\u003csup\u003e74\u003c/sup\u003e and charge density difference analysis. The effects of the adsorbed hydrogen-induced charge transfer are investigated further for the adsorption of HCOO and the dissociation of HCOOH to HCO and OH. For the dissociation of HCOOH, both the reaction energy and the energy barriers were calculated. The barriers were calculated using the climbing image nudged elastic band (CI-NEB) method,\u003csup\u003e75\u003c/sup\u003e and the transition states were verified by vibrational analyses.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e2.14 Continuum modelling\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eContinuum modelling was conducted using Comsol Multiphysics 6.2. We started from the geometries developed by Binninger et al.\u003csup\u003e76\u003c/sup\u003e but changed the system to one where support and particle are touching. We used a system of perfect conductors and a 1 V bias between particle and support. These calculations are only used for qualitative results, so we do not show numbers in the legend of Fig. 2g.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAdditional references\u003c/strong\u003e\u003c/p\u003e\n\u003col start=\"47\"\u003e\n\u003cli\u003eNakano, H., Nakamura, I., Fujitani, T. \u0026amp; Nakamura, J. Structure-dependent kinetics for synthesis and decomposition of formate species over Cu(111) and Cu(110) model catalysts. \u003cem\u003eJ. Phys. Chem. B\u003c/em\u003e 105, 1355\u0026ndash;1365 (2001).\u003c/li\u003e\n\u003cli\u003eRuehl, G., Harman, S. E., Gluth, O. M., Lavoy, D. H. \u0026amp; Campbell, C. T. Energetics of Adsorbed Formate and Formic Acid on Cu(111) by Calorimetry. \u003cem\u003eACS Catal.\u003c/em\u003e 12, 10950\u0026ndash;10960 (2022).\u003c/li\u003e\n\u003cli\u003eYasumura, H., Mizutani, A., Nagatsuka, N., Watanabe, K. \u0026amp; Koitaya, T. Surface Orientation Dependent Dissociation of Formic Acid and Adsorption States of Formate on Copper. \u003cem\u003eJ. Phys. Chem. C\u003c/em\u003e 129, 1216\u0026ndash;1227 (2025).\u003c/li\u003e\n\u003cli\u003eHansen, S. I. et al. An Adsorption Isotherm That Includes the Interactions between Adsorbates. \u003cem\u003eJ. Phys. Chem. C\u003c/em\u003e 129, 5287\u0026ndash;5300 (2025).\u003c/li\u003e\n\u003cli\u003eNielsen, N. D., Jensen, A. D. \u0026amp; Christensen, J. M. Quantification of Formate and Oxygen Coverages on Cu Under Industrial Methanol Synthesis Conditions. \u003cem\u003eCatal. Lett.\u003c/em\u003e 150, 2447\u0026ndash;2456 (2020).\u003c/li\u003e\n\u003cli\u003eNielsen, N. D., Smitshuysen, T. E. L., Damsgaard, C. D., Jensen, A. D. \u0026amp; Christensen, J. M. Characterization of oxide-supported Cu by infrared measurements on adsorbed CO. \u003cem\u003eSurf. Sci.\u003c/em\u003e 703, 121725 (2021).\u003c/li\u003e\n\u003cli\u003eChatterjee, R. et al. Mapping Support Interactions in Copper Catalysts. \u003cem\u003eTop. Catal.\u003c/em\u003e 62, (2019).\u003c/li\u003e\n\u003cli\u003eMuhler, M., Nielsen, L. P., T\u0026ouml;rnqvist, E., Clausen, B. S. \u0026amp; Tops\u0026oslash;e, H. Temperature-programmed desorption of H\u003csub\u003e2\u003c/sub\u003e as a tool to determine metal surface areas of Cu catalysts. \u003cem\u003eCatal. Lett.\u003c/em\u003e 14, 241\u0026ndash;249 (1992).\u003c/li\u003e\n\u003cli\u003eFichtl, M. B. \u0026amp; Hinrichsen, O. On the temperature programmed desorption of hydrogen from polycrystalline copper. \u003cem\u003eCatal. Lett.\u003c/em\u003e 144, 2114\u0026ndash;2120 (2014).\u003c/li\u003e\n\u003cli\u003eDoh, W. H., Roy, P. C. \u0026amp; Kim, C. M. Interaction of Hydrogen with ZnO: Surface Adsorption versus Bulk Diffusion. \u003cem\u003eLangmuir\u003c/em\u003e 26, 16278\u0026ndash;16281 (2010).\u003c/li\u003e\n\u003cli\u003eBarański, A. \u0026amp; Gałuszka, J. Temperature-programmed desorption studies of the hydrogen-zinc oxide system. \u003cem\u003eJ. Catal.\u003c/em\u003e 44, 259\u0026ndash;270 (1976).\u003c/li\u003e\n\u003cli\u003eChinchen, G. C., Hay, C. M., Vandervell, H. D. \u0026amp; Waugh, K. C. The measurement of copper surface areas by reactive frontal chromatography. \u003cem\u003eJ. Catal.\u003c/em\u003e 103, 79\u0026ndash;86 (1987).\u003c/li\u003e\n\u003cli\u003eJones, R. et al. Multimodal Ambient Pressure Sample Environment for the HIPPIE Solid-Gas Endstation at MAX IV Laboratory. \u003cem\u003ePhoton Science\u003c/em\u003e, in press doi:10.1021/PHOTONSCI.5C00021.\u003c/li\u003e\n\u003cli\u003eZhu, S. et al. HIPPIE: a new platform for ambient-pressure X-ray photoelectron spectroscopy at the MAX IV Laboratory. \u003cem\u003eJ. Synchrotron Rad.\u003c/em\u003e 28, 624\u0026ndash;636 (2021).\u003c/li\u003e\n\u003cli\u003eP.J. Linstrom \u0026amp; W.G. Mallard. NIST Chemistry WebBook, NIST Standard Reference Database Number 69. (National Institute of Standards and Technology, Gaithersburg MD, 2025).\u003c/li\u003e\n\u003cli\u003eKuld, S., Conradsen, C., Moses, P. G., Chorkendorff, I. \u0026amp; Sehested, J. Quantification of Zinc Atoms in a Surface Alloy on Copper in an Industrial-Type Methanol Synthesis Catalyst. \u003cem\u003eAngew. Chem. Intl. Ed.\u003c/em\u003e 53, 5941\u0026ndash;5945 (2014).\u003c/li\u003e\n\u003cli\u003eBahl, M. K., Tsai, S. C. \u0026amp; Chung, Y. W. Auger and photoemission investigations of the platinum-SrTiO\u003csub\u003e3\u003c/sub\u003e(100) interface: Relaxation and chemical-shift effects. \u003cem\u003ePhys. Rev. B\u003c/em\u003e 21, 1344 (1980).\u003c/li\u003e\n\u003cli\u003eNishimura, T., Niwa, Y., Tsuchiya, T. \u0026amp; Nozoye, H. Ionic dissociation of methanol studied by photoelectron\u0026ndash;photoion coincidence spectroscopy. \u003cem\u003eJ. Chem. Phys.\u003c/em\u003e 72, 2222\u0026ndash;2225 (1980).\u003c/li\u003e\n\u003cli\u003eOmura, I., Kaneko, T., Yamada, Y. \u0026amp; Tanaka, K. Mass Spectrometric Studies of Photoionization. V. Methanol and Methanol-d1. \u003cem\u003eJ. Phys. Soc. Jpn.\u003c/em\u003e 27, 981\u0026ndash;984 (1969).\u003c/li\u003e\n\u003cli\u003eBeynon, J. H., Fontaine, A. E. \u0026amp; Lester, G. R. Mass spectrometry: the mass spectrum of methanol. Part I. Thermochemical information. \u003cem\u003eIntl. J. Mass Spec.\u003c/em\u003e 1, 1\u0026ndash;24 (1968).\u003c/li\u003e\n\u003cli\u003eKresse, G. \u0026amp; Hafner, J. Ab initio molecular dynamics for liquid metals. \u003cem\u003ePhys. Rev. B\u003c/em\u003e 47, 558\u0026ndash;561 (1993).\u003c/li\u003e\n\u003cli\u003eKresse, G. \u0026amp; Hafner, J. Ab initio molecular dynamics for open-shell transition metals. \u003cem\u003ePhys. Rev. B\u003c/em\u003e 48, 13115\u0026ndash;13118 (1993).\u003c/li\u003e\n\u003cli\u003eKresse, G. \u0026amp; Hafner, J. Ab initio molecular-dynamics simulation of the liquid-metal\u0026ndash;amorphous-semiconductor transition in germanium. \u003cem\u003ePhys. Rev. B\u003c/em\u003e 49, 14251\u0026ndash;14269 (1994).\u003c/li\u003e\n\u003cli\u003eKresse, G. \u0026amp; Furthm\u0026uuml;ller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. \u003cem\u003ePhys. Rev. B\u003c/em\u003e 54, 11169\u0026ndash;11186 (1996).\u003c/li\u003e\n\u003cli\u003eBl\u0026ouml;chl, P. E. Projector augmented-wave method. \u003cem\u003ePhys. Rev. B\u003c/em\u003e 50, 17953\u0026ndash;17979 (1994).\u003c/li\u003e\n\u003cli\u003ePerdew, J. P., Burke, K. \u0026amp; Ernzerhof, M. Generalized Gradient Approximation Made Simple. \u003cem\u003ePhys. Rev. Lett.\u003c/em\u003e 77, 3865\u0026ndash;3868 (1996).\u003c/li\u003e\n\u003cli\u003eDudarev, S. \u0026amp; Botton, G. Electron-energy-loss spectra and the structural stability of nickel oxide: An LSDA+U study. \u003cem\u003ePhys. Rev. B\u003c/em\u003e 57, 1505\u0026ndash;1509 (1998).\u003c/li\u003e\n\u003cli\u003eHenkelman, G., Arnaldsson, A. \u0026amp; J\u0026oacute;nsson, H. A fast and robust algorithm for Bader decomposition of charge density. \u003cem\u003eComput. Mater. Sci.\u003c/em\u003e 36, 354\u0026ndash;360 (2006).\u003c/li\u003e\n\u003cli\u003eHenkelman, G. et al. A climbing image nudged elastic band method for finding saddle points and minimum energy paths. \u003cem\u003eJ. Chem. Phys.\u003c/em\u003e 113, 9901\u0026ndash;9904 (2000).\u003c/li\u003e\n\u003cli\u003eBinninger, T., Schmidt, T. J. \u0026amp; Kramer, D. Capacitive electronic metal-support interactions: Outer surface charging of supported catalyst particles. \u003cem\u003ePhys. Rev. B\u003c/em\u003e 96, 165405 (2017).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"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-6939552/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6939552/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Composite nanomaterials are crucial in a wide range of applications with the industrial Cu/ZnO catalyst used to convert CO\u003csub\u003e2\u003c/sub\u003e into methanol being an important example. The reaction rate scales with the Cu surface area\u003csup\u003e1,2\u003c/sup\u003e implying that the rate-limiting part of the reaction occurs on the metal surface. However, the turnover frequency (TOF, rate per Cu surface atom) is one order of magnitude higher for Cu supported on ZnO compared to pure Cu samples\u003csup\u003e3–6\u003c/sup\u003e. This materials synergy, which is responsible for 90% of the catalytic activity, is still poorly understood\u003csup\u003e7–9\u003c/sup\u003e, and the copper in Cu/ZnO is sometimes described\u003csup\u003e10,11\u003c/sup\u003e as an element with properties radically different from the pristine metal. An understanding of this synergy thus has a direct impact on Power-to-X processes to produce methanol for storage of renewable energy and on heterogeneous catalysis in general. Here we establish the mechanism for the hugely important conversion of CO2 into methanol and how it is accelerated on the composite nanomaterial. We show that charge from donor states created by ad-/absorbed H in the ZnO transfers to the metal and distributes across the metal surface. The surface charging influences the bonding of the adsorbates on the metal and lowers the energy barrier for HCOOH dissociation, which we identify as the rate-limiting step in methanol formation. The lower barrier accounts for the order of magnitude increase in the methanol synthesis rate. The charge transfer phenomenon is general in nature and found to be essential for understanding catalytic phenomena and thus for the rational development of improved future catalysts.","manuscriptTitle":"Charge transfer across the metal/oxide interface determines the rate of CO2 hydrogenation to methanol over Cu/ZnO catalysts","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-12 08:42:59","doi":"10.21203/rs.3.rs-6939552/v1","editorialEvents":[],"status":"published","journal":{"display":false,"email":"[email protected]","identity":"nature","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"nature","sideBox":"Learn more about [Nature](http://www.nature.com/nature/)","snPcode":"","submissionUrl":"","title":"Nature","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"5645b887-4297-4e4c-9be0-c236c7ab0f6e","owner":[],"postedDate":"February 12th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":62791516,"name":"Physical sciences/Chemistry/Catalysis/Heterogeneous catalysis"},{"id":62791517,"name":"Physical sciences/Physics/Chemical physics"},{"id":62791518,"name":"Physical sciences/Chemistry/Catalysis/Catalytic mechanisms"},{"id":62791519,"name":"Physical sciences/Physics/Condensed-matter physics/Surfaces, interfaces and thin films"}],"tags":[],"updatedAt":"2026-03-18T14:41:06+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-12 08:42:59","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6939552","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6939552","identity":"rs-6939552","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}