arXiv subjects

Damien Laage

Publications and source records attributed to Damien Laage.

4 recordsLinked to original sources

arXiv2025

Liquid water under vibrational strong coupling: an extended cavity Born-Oppenheimer molecular dynamics study

A computational study of liquid water when the system is coupled with a (model) Fabry-Perot cavity is reported. At this end, the Cavity Born-Oppenheimer Molecular Dynamics approach proposed recently (Li et al., Proc. Nat. Acad. Sci. USA, 2020, 117, 18324-18331) is employed and different properties of water under vibrational strong coupling (VSC) are investigated. Different cavity frequencies are considered, corresponding to different modes in the IR spectrum of liquid water: high frequency (corresponding to O-H stretching modes), medium frequency (corresponding to water molecule bending) and low frequencies (corresponding to librational modes). Simulations were done both using classical and quantum nuclear dynamics, this last via Ring Polymer Molecular Dynamics. Similar effects of the cavity are obtained in both cases. Namely, whereas the infrared spectrum is observed to be split for all cavity frequencies, no effects on structural properties are detected. In addition, transport and dynamical properties, including the diffusion coefficient, molecular reorientation and hydrogen bond (HB) jump exchange times, show no effect due to cavity coupling when an extended statistical analysis is performed.

physics.chem-ph↗

arXiv2024

ArcaNN: automated enhanced sampling generation of training sets for chemically reactive machine learning interatomic potentials

The emergence of artificial intelligence has profoundly impacted computational chemistry, particularly through machine-learned potentials (MLPs), which offer a balance of accuracy and efficiency in calculating atomic energies and forces to be used in molecular dynamics simulations. These MLPs have significantly advanced molecular dynamics simulations across various applications, including large-scale simulations of materials, interfaces, and chemical reactions. Despite these advances, the construction of training datasets - a critical component for the accuracy of MLPs - has not received proportional attention. This is particularly critical for chemical reactivity which depends on rare barrier-crossing events. Here we address this gap by introducing ArcaNN, a comprehensive framework designed for generating training datasets for reactive MLPs. ArcaNN employs a concurrent learning approach combined with advanced sampling techniques to ensure accurate representation of high-energy geometries. The framework integrates automated processes for iterative training, exploration, new configuration selection, and energy and force labeling, while ensuring reproducibility and documentation. We demonstrate ArcaNN's capabilities through a paradigm nucleophilic substitution reaction in solution, showcasing its effectiveness, the uniformly low error of the resulting MLP everywhere along the chemical reaction coordinate, and its potential for broad applications in reactive molecular dynamics. We also provide guidelines on how to assess the quality of a NNP for a reactive system.

physics.chem-ph↗

arXiv2023

Dielectric response of confined water films: Insights from classical DFT

We re-examine the problem of the dielectric response of highly polar liquids such as water in confinement between two walls using a simple two-variable density functional theory involving number and polarisation densities. In the longitudinal polarisation case where a perturbing field is applied perpendicularly to the walls, we show that the notion of local dielectric constant, although ill-defined at a microscopic level, makes sense when a coarse-graining over the typical size of a particle is introduced. The approach makes it possible to study the effective dielectric response of thin liquid films of various thicknesses in connection to the recent experiments of [Fumagalli et al. , Science, 2018, 360, 1339-1342], and to discuss the notion interfacial dielectric constant. We argue that the observed properties as function of slab dimension, in particular the very low dielectric constants of the order of 2-3 measured for thin slabs of 1 nm thickness do not highlight any special property of water but can be recovered for a generic polar solvent having similar particle size and the same high dielectric constant. Regarding the transverse polarisation case where the perturbing field is parallel to the walls, the associated effective dielectric constant as a function of the slab dimension reaches bulk-like values at much shorter widths than in the longitudinal case.

cond-mat.stat-mech↗

arXiv2022

Explicit models of motions to understand protein side-chain dynamics

Nuclear magnetic relaxation is widely used to probe protein dynamics. For decades, most analyses of relaxation in proteins have relied successfully on the model-free approach, forgoing mechanistic descriptions of motions. Model-free types of correlation functions cannot describe a large carbon-13 relaxation dataset in protein sidechains. Here, we use molecular dynamics simulations to design explicit models of motion and solve Fokker-Planck diffusion equations. These models of motion provide better agreement with relaxation data, mechanistic insight and a direct link to configuration entropy.

physics.chem-ph↗