Polarizable Embedding Strategies for the Simulation of Spectroscopic Properties in Aqueous Solution

Accurate modeling of solvent effects is essential for the reliable simulation of spectroscopic properties in condensed phase. In this presentation, I will discuss multiscale computational strategies combining quantum-mechanical descriptions of the solute with atomistic polarizable embedding models, with particular focus on QM/Fluctuating Charges (FQ) and QM/Fluctuating Charges and Dipoles (FQFμ )approaches.

The first part will address zwitterionic L-tryptophan in aqueous solution, where extensive classical molecular dynamics sampling is combined with TD-DFT calculations to simulate several spectroscopic properties like UV-Vis, ECD, NMR, IR, Raman, and ROA spectra. The results highlight the importance of solvent polarization in reproducing electronic, chiroptical, magnetic, and vibrational signatures.

The second part will focus on the application of the MCSCF/FQ framework to vibronic absorption spectroscopy in aqueous solution. In particular, results for benzene in water will be presented, showing how the combination of multireference electronic structure methods, polarizable embedding, and classical molecular dynamics sampling enables an accurate description of solvent effects on vibronic spectral profiles.

Finally, I will present ongoing work aimed at developing a computational protocol for the simulation of fluorescence spectra in aqueous solution, using phenol as a model system. This strategy relies on classical excited-state molecular dynamics, employing a force field reparametrized with the JOYCE program, to sample the relevant excited-state configurations prior to the calculation of emission properties.

Overall, these studies demonstrate the versatility of multiscale and polarizable computational approaches for the interpretation and prediction of spectroscopic observables in complex molecular environments.