(561d) Coarse-Grain Models with a Parameterizable Friction Coefficient: Recovering Structure, Dynamics, and Viscosity in Polymer Melts

All-atom (AA) molecular dynamics simulations of polymers capture behaviors exhibited on the order of picoseconds up to nanoseconds or beyond, limited primarily by computational expense. Coarse-grain (CG) representations of polymers aim to reduce computational effort and enable simulations to bridge to macroscopically relevant length and time scales, while preserving some features of the AA model. Here, we study a CG simulation model that aims to preserve both chemical specificity (typical of systematic CG methods) and dynamics (typical of phenomenological CG models) of a polymer melt. The model is parameterized in two steps. First, we generate the conservative part of the force-field using the iterative Boltzmann inversion (IBI) method to preserve the chemical specificity of the AA structure. Second, we recover the dynamics by introducing a Langevin thermostat, and thus a tunable friction coefficient, that we parameterize to correct for the sped-up dynamics of the IBI-generated force-field. We recently showed that we can recover AA dynamics by parameterizing the friction coefficient of the CG representation and compared the parameterization across various measures of translational and rotational motion [J. Chem. Phys. 154, 084114 (2021)]. Here, we test the parameterization of the dynamics compared to those targeted to recover a material property, the zero-shear viscosity. We show that the viscosity-based friction is consistent with the other measures and that viscosity may be more simply predicted by using the dynamics-based friction measures. We also discuss the chain length- and temperature-dependence of the friction parameterization method.