Computational Vibrational Spectroscopy

TL;DR

This paper reviews computational methods used in vibrational spectroscopy to analyze molecular dynamics, including physics-based models, machine learning force fields, and conformational sampling techniques.

Contribution

It provides a comprehensive summary of recent computer-based approaches for linking molecular structure to vibrational spectra, highlighting advances in modeling and data analysis.

Findings

Physics-based empirical energy functions are widely used.

Machine-learned force fields improve spectral predictions.

Map-based approaches facilitate conformational sampling.

Abstract

Vibrational spectroscopy is a powerful technique to characterize the near-equilibrium dynamics of molecules in the gas- and the condensed-phase. This contribution summarizes efforts from computer-based methods to gain insight into the relationship between structure and spectroscopic response. Methods for this purpose include physics-based empirical energy functions, machine-learned force fields, and methods that separate sampling conformational space and determining the data for spectral analysis such as map-based approaches.