Computational strategies for polymer dielectrics design

Abstract

The present contribution provides a perspective on the degree to which modern computational methods can be harnessed to guide the design of polymeric dielectrics. A variety of methods, including quantum mechanical ab initio methods, classical force-field based molecular dynamics simulations, and data-driven paradigms, such as quantitative structure-property relationship and machine learning schemes, are discussed. Strategies to explore, search and screen chemical and configurational spaces extensively are also proposed. Some examples of computation-guided synthesis and understanding of real polymer dielectrics are also provided, highlighting the anticipated increasing role of such computational methods in the future design of polymer dielectrics.