Boosting Molecular Simulations with AI: Applications to Ice Chemistry, Thermal Modeling, and Materials Discovery

Molecular modeling, rooted in density functional theory (DFT) and molecular dynamics (MD), has been a cornerstone in physical chemistry, and materials science for nearly four decades. This approach offers a magnifying lens into the atomistic texture of complex systems, enabling the interpretation of experimental data, the elucidation of structure-function relationships of materials, and the exploration of extreme conditions. However, its efficacy is hindered by its inherently high computational costs, restricting the size and complexity of systems accessible to accurate electronic structure calculations to a mere few hundred atoms, whereas the use of empirical potentials grants access to billion-atom models but at the expense of diminished accuracy and transferability.

In this presentation, I will demonstrate how the integration of Artificial Intelligence (AI) is catalyzing a paradigm shift, conjugating precision, complexity, and computational efficiency. Examples will include the application of machine learning to study the structural and electronic properties of ice surfaces, thermal modeling of transistors, and the discovery of novel inorganic materials for renewable energy harvesting and conversion.