Computational strategies for polymer dielectrics design

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.