Multiscale Modeling for Catalytic Materials
Embedded Files
Our research advocates multi-scale modeling as a unifying framework to understand and design catalytic materials for valuable chemical transformations. Catalysis is inherently a hierarchical problem: atomic-scale electronic structures govern bond activation, mesoscopic surface motifs determine selectivity, and macroscopic reaction environments control efficiency and stability. By bridging quantum-level simulations, statistical and kinetic modeling, and data-driven machine learning approaches, we aim to connect fundamental physicochemical principles with experimentally observable catalytic performance. This integrated strategy enables rational exploration of complex catalytic systems, accelerates the discovery of functional materials, and provides predictive insights for sustainable chemical processes, including energy conversion, small-molecule activation, and green synthesis.
Machine Learning for Chemical Application
Our research focuses on advancing machine learning methodologies for chemical applications, with particular emphasis on molecular design toward novel and target-oriented properties. By integrating molecular representations, physics-informed descriptors, and data-driven learning architectures, we aim to uncover structure–property relationships that are difficult to access through conventional trial-and-error approaches. Machine learning serves not only as a predictive tool but also as a generative engine, enabling inverse design of molecules with tailored optical, electronic, catalytic, and functional characteristics. Through close coupling with quantum chemical calculations and experimental validation, our work seeks to accelerate molecular discovery, expand accessible chemical space, and provide interpretable insights for next-generation functional materials and molecular systems.
Page updated
Google Sites
Report abuse