Hu Group at Georgia Tech MSE

Our group at Georgia Tech advances the discovery of high-performance, low-cost materials for next-generation energy technologies. Our research is guided by the belief that transformative materials discovery requires robust, transferable design principles grounded in an understanding of why materials behave as they do. We develop computational frameworks that connect atomistic modeling, high-throughput screening, and agentic AI. Density functional theory, molecular dynamics, and machine-learned interatomic potentials reveal atomic-scale mechanisms and structure-property relationships. High-throughput screening extends these insights across thousands of materials to identify general trends and assess their transferability across chemical and structural spaces. Our agentic AI frameworks incorporate chemical reasoning and domain knowledge to generate hypotheses, test and refine them through computational evaluation, and translate findings into new design principles. We apply these approaches to understand and design dynamic materials interfaces, focusing on (1) electrochemical interfaces for catalysis under realistic reaction conditions and (2) organic-inorganic interfaces in quantum-confined nanomaterials for optoelectronics.