Influence of quantum corrections on the predicted isobaric heat capacity of polarizable water models.

This article focuses on evaluating quantum correction methods for predicting the isobaric heat capacity (Cp) of water using classical molecular dynamics (MD) simulations with flexible polarizable water models. Two quantum correction approaches are compared: the method of Horn et al., which uses fixed fundamental vibrational frequencies from experiments, and the method of Berens et al., which employs the full vibrational spectra derived from the MD simulations themselves, allowing temperature-dependent frequency shifts to be captured. Using the iAMOEBA and AMOEBA14 water models, the study finds that the Berens et al. method provides larger and temperature-dependent quantum corrections that improve agreement with experimental Cp data, particularly for iAMOEBA, which achieves an average relative error of 1.3%. The work also demonstrates that applying quantum corrections as additive terms to Cp computed via the fluctuation formula is more reliable than incorporating energy corrections before polynomial fitting, and it cautions against using experimental ideal gas contributions with polarizable models. These findings highlight the importance of model-based, temperature-sensitive quantum corrections for accurate thermodynamic property predictions in classical MD simulations of water.