Inverse Design of Block Polymer Materials with Desired Nanoscale Structure and Macroscale Properties

Abstract

The rational design of novel polymers with tailored material properties has been a long-standing challenge in the field due to the large number of possible polymer design variables. To accelerate this design process, there is a critical need to develop novel tools to aid in the inverse design process and efficiently explore the high-dimensional polymer design space. Optimizing macroscale material properties for polymeric systems is difficult as properties are dictated by features on a multitude of length scales, ranging from the chosen monomer chemistries to the chain level design to larger-scale domain structures. In this work, we present an efficient high-throughput in-silico based framework to effectively design high-performance polymers with desired multi-scale nanostructure and macroscale properties, which we call RAPSIDY 2.0 - Rapid Analysis of Polymer Structure and Inverse Design strategY 2.0. This new version of RAPSIDY builds upon our previous work, RAPSIDY 1.0, which focused purely on identifying polymer designs that stabilized a desired nanoscale morphology. In RAPSIDY 2.0 we use a combination of molecular dynamics simulations and Bayesian optimization driven active learning to optimally query high-dimensional polymer design spaces and propose promising design candidates that simultaneously stabilize a selected nanoscale morphology and exhibit desired macroscale material properties. We utilize MD simulations with polymer chains preplaced into selected nanoscale morphologies and perform virtual experiments to determine the stability of the chosen polymer design within the target morphology and calculate the desired macroscale material properties (e.g., thermal conductivity). Our methodology directly addresses the unique challenge associated with copolymers, whose macroscale properties are a function of both their chain design and mesoscale morphology, which are coupled.