Multiscale simulation frameworks for light-matter interaction in AI-designed materials
Abstract
We describe multiscale simulation protocols that connect light–material interaction and materials-by-design strategies. The frameworks include classical, semiclassical and quantum-mechanical descriptions of photons, electrons and atoms, and link ab initio approaches to large-scale dynamics and optimization algorithms. The schemes help provide insight into the fundamental physics and design of new materials-photonic crystals, plasmonic nanostructures, and organic photovoltaics-that offer promise for resolving fundamental energy and information processing problems. They represent proof-of-principle demonstration of the ability of multiscale methods to tackle the interplay of atomistic, continuum, and field-level phenomena as well as the potential of artificial intelligence in materials design. One of the long-term goals is also to include the schemes in a general data-driven-discovery framework for the computational design of functional materials in the context of different fields related to industrial and mission of government.