Data-Driven Theoretical\nDesign of Anion Cluster-Based\nSodium Antiperovskite Superionic Conductors

Sodium antiperovskite materials (APs) are a promising\nclass of\nsolid-state electrolytes owing to their high structural tolerance\nand good formability. However, few APs have been synthesized experimentally,\nindicating the necessity of exploring potential chemical spaces with\nhigher ionic conductivities. Herein, through a combined particle swarm\noptimization algorithm, high-throughput first-principles calculations,\nab initio molecular dynamics, and long time-scale machine-learning\nmolecular dynamics simulations, strategies based on site-exchanging\nand anion clusters are shown to simultaneously enhance the thermal\nstability and sodium diffusivity in the designed APs. Among these\nAPs, the highest theoretical ionic conductivity of 39.05 mS/cm is\nachieved with Na<sub>3</sub>BrSO<sub>4</sub> at room temperature due\nto the strong coupling of cluster rotation and sodium migration. We\nhighlight that not only the rotation dynamics but also its coupling\nwith Na diffusion contribute to the high ionic conductivity, as confirmed\nby the proposed local difference frequency center to evaluate the\ncoupling degree. Our work designs promising site-exchanging APs and\noffers insights into the coupling between anion rotation and cation\nmigration, which can effectively guide the design of superionic conductors\nwith cluster rotation dynamics.