Regensburg 2022 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

CPP: Fachverband Chemische Physik und Polymerphysik

CPP 12: Poster 1

CPP 12.27: Poster

Monday, September 5, 2022, 18:00–20:00, P1

Correcting Coarse-Grained Dynamics of Molecular Liquids and Their Mixtures Using an Efficient Iterative Memory Reconstruction Method — •Madhusmita Tripathy, Viktor Klippenstein, and Nico FA van der Vegt — Eduard-Zintl-Institut für Anorganische und Physikalische Chemie, Technische Universität Darmstadt, 64287 Darmstadt, Germany

Generalized Langevin equation (GLE) based coarse-grained (CG) models are considered to be the most reliable models for dynamically consistent coarse-graining [1]. However, their implementation in molecular simulation is not straight-forward owing to their inherent complexity [2]. With an aim to employ computationally tractable GLE based CG models for dynamic coarse graining of complex molecular systems, we coarse-grain two molecular liquids and their mixtures at various compositions following a novel iterative optimization scheme. Using the memory kernel from an isotropic GLE model as a starting point, we use an efficient iterative memory reconstruction method, which can closely reproduce the underlying fine-grained (FG) dynamics, assessed in terms of the velocity auto-correlation function, within a few iterations. We use this iterative method to correct the artificial dynamic speed-up in CG molecular dynamics (MD) simulations of pure molecular liquids and the relative dynamic speed-up in their mixtures. Furthermore, we investigate the transferability of the resulting memory kernels to molecular mixtures with varying composition.

[1] Klippenstein et al. J. Phys. Chem. B 125 (19), 4931-4954 [2] Glatzel and Schilling, Europhys Lett. 136 36001 (2021)