Bidirectional mapping between self-consistent field theory and molecular dynamics

Abstract

A bidirectional mapping scheme that bridges particle-based and field-based descriptions for polymers is presented. Initial application is made to immiscible homopolymer blends. The forward mapping (upscaling) approach is based on the use of molecular dynamics simulations to calculate interfacial density profiles for polymer molecular weights that can be readily relaxed using standard simulation methods. These profiles are used to determine the optimal, effective interaction parameter that appears in the one-parameter self-consistent field theory treatment employed in the present work. Reverse mapping from a field representation to a particle-based description is accomplished by the application of a density-biased Monte Carlo method that generates representative chain configurations in the blend using statistical weights derived from fields obtained from self-consistent field theory.