Development of Multiscale Embedding Methods for Computational Vibrational Spectroscopy

This Ph.D. thesis presents advancements in multiscale embedding methods, with a focus on the fragment-based polarizable embedding quantum mechanics (PE-QM) approach for computational vibrational spectroscopy. The research emphasizes the development of methodologies for efficient and accurate modeling of vibrational properties, including harmonic infrared (IR) and Raman spectra, through the use of analytical Hessians.

A comprehensive workflow is introduced to facilitate the integration of PE-QM into vibrational spectroscopy simulations. The methodology is further evaluated through benchmarks against experimental data for solute–solvent systems. Building on these developments, the thesis explores extensions toward anharmonic corrections within an open-ended response theory framework. To support these efforts, the opensource software package PyFraME is introduced, offering a modular and automated platform for high-performance computations in multiscale embedding.

These advancements refine computational vibrational spectroscopy methods, address current limitations, and establish a foundation for the study of nonlinear vibrational properties and complex biomolecular systems.