(28b) Investigating Primary Nucleation in Polymer Melts Using GPU-Accelerated Wang-Landau Simulations

The importance of crystallinity in a polymer material comes from its drastic impact on its mechanical, optical, and other material properties.
Despite its importance, the process of crystallization is imperfectly understood, which is in no small part due to the highly non-equilibrium nature of the crystallization process.
The primary nucleation process has recently drawn renewed interest following experimental observations that cast doubt on the accuracy of classical nucleation theory.
Several authors have sought alternative explanations, including the proposal of an intermediate mesomorphic phase aiding in the crystallization transition.
To test this hypothesis, we have constructed a GPU-accelerated Wang-Landau Monte Carlo (WLMC) algorithm with an advanced polymer move set to directly sample the density of states of a melt of a coarse-grained polymer model undergoing primary nucleation.
Preliminary results from our method show an increase in sampling efficiency of nearly two orders of magnitude compared to an equivalent serial algorithm, allowing us to efficiently sample a large, dense melt as it undergoes nucleation.
From our WLMC simulations, we are able to directly calculate a free energy, revealing the location of the crystallization transition as a function of a variety of molecular parameters.
Interestingly, for a freely-jointed chain model with no bending or torsion potentials, the free energy predicted from our model is in good agreement with classical nucleation theory, revealing that chain connectivity alone does not give rise to an intermediate phase.

Funding Acknowledgement: We acknowledge financial support from the American Chemical Society
Petroleum Research Fund (PRF\# 59244-DN16) and BYU Board of Trustees as well as computational
resources from the BYU Office of Research Computing and Fulton Supercomputing Lab.