Modulating Coarse-Grained Dynamics by Perturbing Free Energy Landscapes

Area of Science:

• Computational chemistry
• Molecular dynamics simulations
• Statistical mechanics

Background:

• Simulating long-time dynamics of multiatomic molecules is computationally challenging.
• Traditional coarse-grained (CG) models often struggle to accurately capture energy barrier crossing dynamics.
• Preserving the structural integrity of all-atom (AA) systems during coarse-graining is crucial.

Purpose of the Study:

• To introduce a novel approach for describing long-time molecular dynamics.
• To accurately determine the self-diffusion coefficient in coarse-grained systems.
• To preserve the structural characteristics of all-atom systems in coarse-grained models.

Main Methods:

• Modulating the free energy landscape (FEL) to capture dominant features of energy-barrier crossing dynamics.
• Enhancing conservative force fields with high-frequency perturbations in CG systems.
• Utilizing theoretical arguments to demonstrate the preservation of lower-order distribution functions.
• Applying molecular dynamics simulations to various systems, including bulk and confined fluids.
• Incorporating machine learning (ML) optimized many-body potentials.

Main Results:

• The self-diffusion coefficient of CG systems can be accurately delineated by enhancing force fields with high-frequency perturbations.
• These perturbations do not alter lower-order distribution functions, preserving AA system structure.
• The approach is validated across systems with and without time scale separations, and in inhomogeneous nanochannel environments.
• Successful application to ML-optimized many-body potentials demonstrates broad utility.

Conclusions:

• The proposed method effectively describes long-time molecular dynamics by enhancing CG models.
• Accurate self-diffusion coefficients and preserved molecular structure are achieved.
• The approach is versatile, applicable to diverse systems and advanced potentials.