Coarse-grained conformational surface hopping: Methodology and transferability

Abstract

Coarse-grained (CG) conformational surface hopping (SH) adapts the concept of multisurface dynamics, initially developed to describe electronic transitions in chemical reactions, to accurately describe classical molecular dynamics at a reduced level. The SH scheme couples distinct conformational basins (states), each described by its own force field (surface), resulting in a significant improvement of the approximation to the many-body potential of mean force [Phys. Rev. Lett. 121, 256002 (2018)]. The present study first describes CG SH in more detail, through both a toy model and a three-bead model of hexane. We further extend the methodology to non-bonded interactions and report its impact on liquid properties. Finally, we investigate the transferability of the surfaces to distinct systems and thermodynamic state points, through a simple tuning of the state probabilities. In particular, applications to variations in temperature and chemical composition show good agreement with reference atomistic calculations, introducing a promising "weak-transferability regime," where CG force fields can be shared across thermodynamic and chemical neighborhoods.

arXiv comment15 pages, 7 figures

References in corpus (11)

• A Performance and Cost Assessment of Machine Learning Interatomic Potentials
• Dissipative particle dynamics for interacting systems
• Beware of density dependent pair potentials
• Representability problems for coarse-grained water potentials
• DeePCG: constructing coarse-grained models via deep neural networks
• Temperature-transferable coarse-graining of ionic liquids with dual graph convolutional neural networks
• Understanding three-body contributions to coarse-grained force-fields
• Encoding and Selecting Coarse-Grain Mapping Operators with Hierarchical Graphs
• Consistent interpretation of molecular simulation kinetics using Markov state models biased with external information
• Accurate structure-based coarse-graining leads to consistent barrier-crossing dynamics
• Concurrent parametrization against static and kinetic information leads to more robust coarse-grained force fields

Cited by in corpus (2)

• Dynamical properties across different coarse-grained models for ionic liquids
• Broad chemical transferability in structure-based coarse-graining