SGMSC: A New Software for Searching the Global-Minimum Structure of Surface-Supported Clusters by the Improved Basin...

Surface-supported clusters (SCCs) exhibit unique electronic, catalytic, and structural properties that make them indispensable in fields such as heterogeneous catalysis, nanoelectronics, and energy storage. The structure corresponding to the global energy minimum (GM) of these clusters is critical to investigating their intrinsic properties and practical applications, as it represents the most thermodynamically stable configuration. However, searching for the most stable structure of surface-supported clusters is a significant computational difficulty due to the high-dimensional potential energy surface (PES) and the complex interplay between cluster-cluster and cluster-support interactions. In this work, we present SGMSC (Global-minimum Search of Surface-supported Cluster), a software proposed to find the potential global minimum configurations of surface-supported clusters. SGMSC combines a refined basin hopping (BH) method alongside embedded atom potentials (EAM), leveraging the stochastic sampling capability of BH and the reliability of EAM in describing interatomic interactions. To validate the accuracy and efficiency of SGMSC package, we performed GM structural searches for a series of surface-supported clusters, including small Cu7@M (M = Zr, W, Ti, Ta, Pt, Pd, Pb, Ni, Mo, Mg, Fe, Cu, Co, Au, Al, Ag), medium Pt10@Au(100), and large Au20@Pt(110). The possible GM structures predicted by EAM remain the lowest-energy isomers after DFT relaxation, which indicates that the EAM-predicted structural conclusions are qualitatively reliable. Additionally, the GM structures of Pt10@Au(100) and Au20@Pt(110) differ significantly from their gas-phase counterparts, highlighting the crucial role of cluster-support interactions in determining the stable configurations of surface-supported clusters. SGMSC provides a prescreening tool that balances computational cost and structural rationality, suitable for large-scale rapid screening of supported cluster structures.