Unraveling the Catalytic Effect of Hydrogen Adsorption on Pt Nanoparticle Shape-Change

Understanding atomic-level processes in surface science and heterogeneous catalysis is complicated by the wide range of time- and length-scales needed for realistic simulations for comparison to experiments. To accelerate molecular dynamics simulations, we rely on machine-learning methods to capture interatomic interactions with quantum-mechanical accuracy. Our method (FLARE) enables autonomous selection of training sets for reactive systems, based on an adaptive closed-loop algorithm that constructs accurate and uncertainty-aware Bayesian force fields on-the-fly from molecular dynamics simulations. From this accelerated sampling routine, a machine-learned force field (MLFF) is built that can access long time- and length-scale simulations of catalytic reaction dynamics. We use ML-accelerated MD simulations to study heterogeneous reactions and nanoparticle shape-changes that are catalyzed under exposure to reactive atmospheres. These observations allow for improved opportunities in material design, where material exposure to temperature or adsorbates can yield shape-changes or reconstruction, and consequently different catalytic behavior under operating conditions.<br/><br/>To illustrate these phenomena, we present the catalytic effect of hydrogen adsorption on small Pt nanoparticle shape-change [1]. The MLFF provides an accurate description of these systems under the effects of temperature and reactive adsorbates, which are corroborated by experimental observations, establishing confidence in the ability of the MLFF to provide atomic resolution regarding the underlying mechanisms driving these processes.<br/><br/>[1] Owen, C.J., Marcella, N., Xie, Y., Vandermause, J., Nuzzo, R.G., Frenkel, A.I., Kozinsky, B. “Unraveling the Catalytic Effect of Hydrogen Adsorption on Pt Nanoparticle Shape-Change,” arXiv preprint: arXiv:2306.00901 (2023).