Stability and Dynamics\nof Zeolite-Confined Gold Nanoclusters

Nanoengineered metal@zeolite materials have recently\nemerged as\na promising class of catalysts for several industrially relevant reactions.\nThese materials, which consist of small transition metal nanoclusters\nconfined within three-dimensional zeolite pores, are interesting because\nthey show higher stability and better sintering resistance under reaction\nconditions. While several such hybrid catalysts have been reported\nexperimentally, key questions such as the impact of the zeolite frameworks\non the properties of the metal clusters are not well understood. To\naddress such knowledge gaps, in this study, we report a robust and\ntransferable machine learning-based potential (MLP) that is capable\nof describing the structure, stability, and dynamics of zeolite-confined\ngold nanoclusters. Specifically, we show that the resulting MLP maintains <i>ab initio</i> accuracy across a range of temperatures (300–1000\nK) and can be used to investigate time scales (>10 ns), length\nscales\n(ca. 10,000 atoms), and phenomena (e.g., ensemble-averaged stability\nand diffusivity) that are typically inaccessible using density functional\ntheory (DFT). Taken together, this study represents an important step\nin enabling the rational theory-guided design of metal@zeolite catalysts.