Mechanism of CO2 Conversion to Methanol on a Highly Representative Model Cu/ZnO Interface

CO 2 -Hydrogenation: Computational modelling of methanol synthesis by the Cu/ZnO catalyst highlights the key role of interfacial sites with formate and methoxy intermediates and spectators. The least energy-demanding path for MeOH formation at the most active interfacial site, starting from activated, chemisorbed, CO 2 , is identified. • The interfacial Cu/ZnO site is determined to be the active site for MeOH formation. • The least energy-demanding mechanism for MeOH formation proceeds via HCOO. • CO 2 activation takes place at the interfacial Cu/ZnO site. • Water evolution occurs at Cu sites, avoiding poisoning of the interfacial site. • MeO* hydrogenation to MeOH is the most energy-demanding process. The mechanism of CO 2 hydrogenation to methanol is modelled using plane-wave DFT applied to a representative model Cu 8 -ZnO catalyst system (CZ), obtained via unbiased Monte Carlo exploration of Cu cluster growth over a reconstructed polar ZnO surface. Enhanced CO 2 adsorption and activation is found at the active Cu/ZnO interfacial site – resembling a V O vacancy – compared to sites on other Cu-based systems. Three competing methanol formation mechanisms (the formate, carboxyl and CO hydrogenation pathways) are investigated; the least energy-demanding pathway followed the formate mechanism: CO 2 * → HCOO* → H 2 COO* → H 2 COOH* → H 2 CO* → H 3 CO* → H 3 COH. We report the coexistence of several formate adsorbates, some of which being highly stable spectators that were observed spectroscopically. Only one higher energy interfacial Cu/ZnO formate species is a true intermediate relevant for catalysis, undergoing subsequent hydrogenation to methanol. The methoxy intermediate is also highly stable, in agreement with its spectroscopic observation. The most energy-demanding elementary process is hydrogenation of methoxy to methanol ( E a = 1.20 eV). Furthermore, the calculations indicate the possible role of CO and H 2 CO* in scavenging surface O* by forming CO 2 * or H 2 COO*, thus preventing the poisoning of active sites. Finally, water is expected to form from O* on a pure Cu site only, but not the Cu/ZnO interfacial site relevant for MeOH production. The calculations presented provide valuable new insights that allow a more complete rationalisation of experimental observations. They suggest the key steps to enhance catalysis involves destabilizing the long-lived H 3 CO* favouriting its hydrogenation and fast desorption or stabilizing competing intermediates such as H 2 COH*.