Quantum Chemical Microsolvation via Free Energy–Based Solvation Site Identification

Abstract In quantum chemical descriptions of solutions, solvents are treated either implicitly, accounting only for bulk properties, or explicitly by including solvent molecules in the calculations. Although explicit treatments are more accurate, they are often computationally prohibitive. As a cost-effective alternative, microsolvation models have emerged that include only a limited number of solvent molecules to capture the most relevant solute–solvent interactions. However, determining their optimal number, positions, and orientations remains a challenge. Existing approaches typically either build up solvent clusters in a stepwise, order-dependent manner or select solvent molecules based on geometric criteria without directly quantifying interaction strength. Here, we introduce the Free Energy Based Identification of Solvation Sites (FEBISS) 2, a revised microsolvation protocol based on Grid Inhomogeneous Solvation Theory (GIST) applied to molecular dynamics trajectories. The method identifies thermodynamically favorable solvation sites from Free Energy contributions. It automatically places, orients, and ranks solvent molecules by interaction strength, using a statistically averaged, order-independent orientation procedure that is applicable to arbitrary rigid solvents rather than water alone. This approach provides a physically grounded criterion for selecting the number of solvent molecules in microsolvated clusters. We demonstrate the applicability of the approach across a range of systems, from small organic molecules to transition-metal catalysts, and for various solvents. We show that explicit microsolvation affects the electronic structure even in systems traditionally considered weakly solvent-sensitive, highlighting FEBISS 2 as a practical methodology to set up physically meaningful microsolvation clusters for quantum chemical studies.