(191a) Searching for a Molecular-Level Understanding of Polymer Crystal Nucleation | AIChE

(191a) Searching for a Molecular-Level Understanding of Polymer Crystal Nucleation

Authors 

Tree, D. - Presenter, Brigham Young University
Hundreds of millions of tons of semi-crystalline polymers are produced each year, ranging from commodity plastics to engineering polymers to specialized classes of polymers such as semiconductors. Despite their ubiquity and technological importance, the molecular mechanisms of crystallization remain hotly debated. We argue that the polymer physics of the process of nucleation is a critical unsolved fundamental problem in the field that holds the key to understanding both quiescent and flow-induced crystallization processes. Our approach to this problem has focused on computing phase diagrams and free energy landscapes (FELs) using advanced Monte Carlo methods including Wang-Landau (WL) and the Expanded Ensemble Density of States (EXEDOS) methods. Somewhat counterintuitively, these methods for computing equilibrium properties allow us to understand the fundamental forces driving the highly non-equilibrium process of crystallization.

Our work has led to two insights: one methodological and one with broader relevance for the physics of polymer crystallization. Methodologically, we find that WL and EXEDOS calculations of equilibrium phases and energy barriers are especially helpful for elucidating the connections between polymer chemistry and nucleation. Regarding the polymer physics of crystal nucleation, we find evidence that crystal nucleation is quite sensitive to chain flexibility, but rather insensitive to other molecular parameters. These latter findings are still preliminary, but if they generalize to a broader class of molecules, they will have significant implications for the long-standing debate over multistep crystallization processes and even perhaps for the ability to categorize different polymers into “universality classes” with regards to crystallization.

Funding Acknowledgement: We acknowledge support from the ACS PRF (59244-DN16), Brigham Young University, and computational support from BYU’s Office of Research Computing.