Machine learning models have proven effective in predicting diverse material properties, including mechanical strength, thermal conductivity, optical properties, and electronic structure properties. By analyzing extensive datasets, these models unveil intricate relationships between material compositions, structures, and properties, facilitating efficient screening and identification of desired materials. In our research, we focus on developing machine learning models for high-throughput virtual screening in materials discovery, aiming to replace time-consuming density functional theory (DFT) calculations. These models serve as valuable tools for accelerating the exploration of new materials with tailored characteristics.

The integration of machine learning (ML) and density functional theory (DFT) calculations revolutionizes materials discovery by enabling accelerated screening, predicting properties, optimizing materials, and enhancing our understanding of structure-property relationships. By harnessing large datasets, this approach empowers the prediction of material properties, facilitates the design of novel materials, and guides experimental efforts with greater efficiency. Ultimately, this synergy between ML and DFT drives the development of advanced materials with tailored properties, unlocking their potential for a wide range of applications.

Computational simulations at the atomic scale, inlcuding quantum chemistry (QC), density functional theory (DFT) calculations, and molecular dynamics (MD), play a crucial role in understanding experimental results. These simulations allow the observation of phenomena like bond breaking and surface reconstructing. Moreover, they provide valuable insights into catalytic activity by comparing different candidates and help unravel the underlying factors through electronic structure calculations. By bridging theory and experiment, computational simulations enhance our understanding of complex systems and facilitate the design of novel materials and processes.