Determining quantum effects on water’s heat capacity

Though widespread in use within the physical sciences, classical molecular dynamics simulations — which predict the motions of molecules in a system — are not able to capture the nuclear quantum effects inherent to the heat capacity of water. As a result, scientists often apply these effects as corrections to their data in post-processing. But depending on the method used, the corrections can vary by up to a factor of four, accounting for up to one-third of the heat capacity’s total value.

These changes to heat capacity calculations are impactful, as researchers often use the results to, in turn, calculate additional parameters relating to macroscale properties in scientific and industrial applications. Savoia et al. worked to determine how best to accurately account for these corrections.

The researchers compared two common correction paradigms: the Horn method, which approximates quantum vibrations at a predefined temperature based on experimental data, and the Berens method, which calculates the vibrational states of the specific model being used. They derived water’s heat capacity for two different polarizable models of water at 12 different temperatures, with each correction method applied.

The two methods result in slight variations in their corrections, with the Berens method being the most accurate, but at a larger computational cost. Though the Horn method can be handy, its frequency selection is arbitrary, overlooks the temperature dependency on the frequencies, and may be incorrect for the specific water model being studied.

The authors hope their findings can guide researchers looking for better heat capacity predictions in their own work.

“This article serves as a sort of instruction manual to compute the heat capacity in future studies,” said author Elton Oyarzua.

Source: “Influence of quantum corrections on the predicted isobaric heat capacity of polarizable water models,” by Edoardo Savoia, Elton Oyarzua, B. D. Todd, and Richard J. Sadus, Journal of Chemical Physics (2025). The article can be accessed at https://doi.org/10.1063/5.0256589 .