Accelerating Discovery of Ternary Chiral Materials via Large-Scale Random Crystal Structure Prediction

Abstract: Chiral inorganic crystals with topological characteristics, prized for their exotic properties and fundamental interest, remain scarce in existing database. This work establishes a viable route for their large-scale discovery by integrating universal machine learning interatomic potentials (uMLIPs) for high-throughput structure optimization with the broad exploration capability of Random Structure Search (RSS). We implemented this combined uMLIP-RSS workflow to perform massive variable-composition crystal structure prediction across ternary systems, specifically targeting chiral space groups. High-throughput uMLIP-based optimization and stability screening of over 20 million randomly generated chiral structures identified numerous potentially stable phases out of existing database. Subsequent validation by first-principles confirmed over 120 new chiral inorganic crystals with promising functional applications, including topological characteristics, nonlinear optics, and superconductivity. Notably, this set includes materials exhibiting remarkable quantum phenomena, such as the nonlinear Hall effect driven by berry curvature dipole, quantum metric and symmetry-enforced six-fold topological points, long Fermi arcs and large magnetoresistance. This work substantially expands the pool of chiral functional materials and demonstrates a scalable, efficient strategy for predictive discovery in complex materials.