Molecular dynamics simulations on polymeric materials

Abstract

In the comparison of chain-growth polymerization in solution versus on surface, we found that surface-initiated polymer chains exhibited a markedly broader distribution of chain lengths and slower chain growth kinetics as compared to the solution phase process due to two key factors: (1) the formation of chain “loops” with both termini attached to the substrate via recombination and (2) the “starvation” effect, in which the live chain ends of short polymers are sterically shielded from monomers by the presence of longer neighboring chains. We further extended our reactive force fields to construct a dynamic self-assembly of “living” polymeric chains system, and demonstrated that the system could survive and maintain a steady state with a continuous supply of matter and energy, otherwise, it underwent spontaneous dissociation driven by thermodynamics.

Polymer materials have been revolutionizing the world since its invention. As of today we still fall short of completely comprehending them because of the huge built-in complexity of molecular chemical architecture. Computer simulations have become useful tools to enrich our understating of these materials. The current coarse-grained polymer models are intrinsically non-reactive, and relying on extrinsic Monte Carlo operations of bond creation/deletion to simulate chemical reactions, which poses great challenges to many aspects of studies. In this thesis, we developed reactive coarse-grained models for step-growth and chain-growth polymerization using a modified Lennard-Jones based pairwise force field. The kinetics of in silico step-growth synthesized linear polymers was examined to follow a step-wise reaction manner, which led to Flory-Schulz molecular weight distribution. The dispersity was further found to have a significant effect on the diffusion of polymer melts, and mechanical response of polymer glass. Additionally, the length of solvent in polymer melts was found to affect the diffusion considerably, which could explain the deviation from the theoretical Rouse model widely reported in literatures.