(558d) Physical Descriptors for Predicting Charge Transfer at the Metal-Support Interface

The performance of heterogeneous catalysts featuring metal nanoclusters on oxide supports can be significantly impacted by charge transfer at the metal-support interface. Such charge transfer alters the oxidation state of the metal, thus changing the catalyst’s activity and selectivity. Charge transfer can also influence the metal’s adhesion to the surface, affecting sintering rates and cluster size distributions. The extent of charge transfer is controlled by the underlying electronic structure of the metal and the oxide, which in turn can be tuned by introducing dopants in the surface, as well as adsorbed ligands on the surface, that act to donate or withdraw electron density. Here, we employ density functional theory (DFT) together with statistical learning (SL) to identify physical descriptors that can be used to predict the nature of charge transfer at the metal-support interface in the presence of dopants and adsorbates. We use DFT to generate adsorption energy and charge distribution data for a range of metal-support pairs in the presence of both dopants and adsorbates derived from typical catalytic reaction environments. Using this data, we apply SL to scan a feature-space of candidate descriptors for physical properties of the metal, support, dopant, and adsorbate that most significantly influence the electronic structure of the metal-support interface (i.e., combinations of ionization energy, electron affinity, electronegativity, etc.). Such descriptors are essential for developing ways to tune the electronic structure of the metal and support to control charge transfer. This methodology provides predictive models for enhancing catalyst performance via charge transfer at the metal-support interface.