(661k) Determining Surface-Specific Hubbard-U Corrections and Identifying Key Adsorbates on Reducible and Non-Reducible Transition Metal Oxide Catalysts

Transition metal oxides (TMOs) are commonly used as catalysts and catalyst supports in a variety of chemical transformations. TMOs are classified as reducible or non-reducible depending on the energy cost of breaking the metal-oxygen bond, and the reducibility is one of the key characteristics for catalysts (Puigdollers et al.,2017). Computational tools like density functional theory (DFT) are often used to study the TMO catalyzed reactions, and for strongly correlated systems like TMOs require DFT+U approach to reduce the systematic error caused by electron delocalization (Wang et al.,2006). However, commonly reported Hubbard U-corrections (HUC) reflect bulk properties but fail to reproduce surface properties like surface-adsorbate interactions (Hassan-Legault et al.,2019). Meanwhile, experimental surface characterization techniques like X-ray Photoelectron Spectroscopy (XPS) have difficulties identifying key surface adsorbates or reaction intermediates corresponding to the XPS shifts observed. A synergistic application of XPS and DFT+U can be used to determine the surface HUC, as well as unidentified adsorbed surface moieties on TMO surface (Thang et al.,2018). In this work, we include both reducible (CeO₂, Co₃O₄, and NiO) and non-reducible TMOs (Y₂O₃ and ZrO₂). Their experimental XPS shifts data are obtained from published literature and compared with the core-level binding energy shifts which are calculated with various HUC values and probable adsorbates using Vienne Ab-initio Simulation Package(VASP). From this study, both goals are achieved simultaneously: (i)The determination of surface HUC, which brings better understandings regarding the electron delocalization penalty patterns, and (ii)the assignment of adsorbate/surface species corresponding to experimental XPS shifts, which can help advanced characterization method like in-situ XPS to study reaction mechanism. For all TMOs investigated, the surface HUC value is lower than bulk HUC. This project will further extend to complicated systems such as doped/modified TMO catalysts, which will provide systematic strategies for choosing HUC value and guides for catalyst design and development.