Advantage, limitations, and performance insight of commonly used computation chemistry methods for solvation free energy estimation

The accurate calculation of free energy is a cornerstone of modern computational chemistry, essential for predicting the spontaneity of chemical reactions and the stability of molecular systems. Solvation free energy ΔG𝑠𝑜𝑙𝑣 is a fundamental thermodynamic quantity that describes the free energy change associated with transferring a solute from the gas phase into a solvent. It is particularly useful for estimating reaction rates and pathways of organic reactions, most of which occur in the liquid phase, and has important applications in organic synthesis, biochemistry, and drug discovery.1 The field has evolved significantly, moving from foundational statistical mechanics methods such as Free Energy Perturbation (FEP) and Thermodynamic Integration (TI) to more sophisticated multiscale and hybrid approaches.2,3 A central and persistent challenge is the delicate balance between achieving high accuracy in quantum mechanical descriptions of electronic structure and performing the extensive configurational sampling required to capture entropic effects.4 This report provides an overview of the theoretical frameworks, implementations, and predictive accuracy of commonly used methods for calculating solvation free energy ΔG𝑠𝑜𝑙𝑣 , assessing alchemical methods, quantum chemistry methods, hybrid QM/MM approaches, while also addressing practical considerations. Finally, we examine the transformative role of emerging methods such as machine learning, which are poised to accelerate these calculations, making rigorous, large-scale free energy simulations feasible and better balancing the cost-accuracy trade-off that has long constrained the field.5,6