Abstract
Macromolecular materials are fundamental in biological systems and common in synthetic materials due to their unique functions from the emergent collective properties, which are governed by underlying structural order. One major category of macromolecules, polymers can have intrinsic order in the sequential arrangement of their monomers, the overall structure of the chain, and their self-assembly with other polymer chains. Controlling these features of structural order enables the tailoring of polymer properties for specific applications. However, achieving precise sequence and structural control, as well as obtaining a comprehensive understanding of intra-chain self-assembly processes is significantly challenging for polymers, owing to system complexity and the difficulty of experimental characterization during synthesis and processing. To address these challenges, this dissertation employs computational approaches, including molecular dynamics simulations and machine learning tools, to investigate the emergence and control of order in copolymer sequence formation and hierarchical polymer self-assembly.
For copolymer sequence formation, we first examine the templating effect of pre-formed seed oligomers in step-growth copolymerizations. The results demonstrate that seed oligomers with blocky sequences can bias the resulting copolymers toward longer block lengths and accelerate autocatalysis, which is triggered by an emergent phase separation. We further explore sequence evolution under wet–dry cycling and in radical polymerization systems.
In addition, we develop a conditional generative machine learning model that predicts sequence distributions from synthesis conditions. This model successfully captures key statistical features obtained in molecular dynamics simulations. For self-assembly, we focus on the hierarchical organization of liquid crystalline block copolymers (LCBCPs). Specifically, we investigate the kinetics of the end-to-end coupling of LCBCP cylindrical micelles, which is driven by depletion interactions and follows a step-growth mechanism.
Overall, these findings advance our understanding of factors governing the sequence formation and intra-chain self-assembly processes in macromolecular systems towards more complex and realistic conditions. By combining computational simulations with theory and introducing data-driven generative tools, this dissertation provides new strategies for the rational design and generation of polymer sequences and hierarchically ordered polymeric structures. Ultimately, such strategies will facilitate an improved control over material properties and expand potential applications for macromolecules.
