Predicting the Artificial Acceleration in Coarse-Grained Molecular Dynamics Simulation of Polymer Melts

A predictive model is proposed for the mobility acceleration factor, defined as ratio of the coarse-grained (CG) diffusion coefficient to the all-atom (AA) diffusion coefficient, in polymer melt systems. To this end, a number of polymers are selected for AA simulations at 450 K, followed by Iterative Boltzmann Inversion (IBI) coarse-graining and subsequent CG simulations to compute the AA and CG diffusion coefficients. After identifying the key parameters influencing the mobility acceleration factor, a functional form for predicting the acceleration is proposed based on these effective parameters. A fitting procedure is then carried out to determine the unknown constants associated with the model. As the next step, the developed model is applied to predict the mobility acceleration factor across various systems at different temperatures, which exposes the need to refine the model by incorporating temperature as an additional influential parameter. Inspired by the Arrhenius equation, the predictive model is revised to incorporate temperature, resulting in a formulation capable of predicting the mobility acceleration factor in polymer melts across a range of temperatures. Thus, the final model predicts the artificial mobility increase in coarse-grained systems using three parameters characterizing the local environment of the monomers and temperature. The model shows an average absolute deviation of 5.6 from the true values, demonstrating its reliability and predictive capability.