GPU-accelerated Wang-Landau Simulation of Polymer Crystallization

While semicrystalline polymers are prevalent in today’s markets and homes, many facets of the crystallization process remain mysterious. For example, connectivity and stiffness in polymers complicate the nucleation process greatly, introducing nematically-ordered phases. There has been recent interest in understanding the relative stability of these phases, and their role in mediating the melt-to-crystal transition in polymers. To investigate this claim, we have investigated the possibility of using a GPU-accelerated Wang-Landau Monte Carlo algorithm with an efficient polymer move set to sample the full density of states (DOS) of a crystallizing polymer melt. In principle, the DOS contains full thermodynamic information, making it possible to construct a free energy surface based on crystallization-relevant order parameters. We will discuss the challenges with this approach and our recent progress.