Computational Design of Complex Materials Using Information Theory: From Physics- to Data-driven Multi-scale Molecular Models

摘要

The development of novel materials with desirable properties, such as nanocomposites, polymers, colloids and biomolecular systems, relies heavily on the knowledge of their structure-property relationships. The prediction of such relationships is the subject of computational materials design. Molecular dynamics (MD) simulations at the atomistic level can provide quantitative information about structural and dynamical properties of molecular systems. The recent enormous advances in computational power allow us to perform intense atomistic-level simulations. However, the broad range of length and time scales appearing in such complex (e.g., macromolecular) materials still presents significant computational challenges, especially in tackling engineering and design tasks.