MB-pol(2023): Sub-chemical Accuracy for Water Simulations from the Gas to the Liquid Phase

XY Zhu and M Riera and EF Bull-Vulpe and F Paesani, JOURNAL OF CHEMICAL THEORY AND COMPUTATION, 19, 3551-3566 (2023).

DOI: 10.1021/acs.jctc.3c00326

Weuse the MB-pol theoretical/computational framework to introducea new family of data-driven many-body potential energy functions (PEFs)for water, named MB-pol(2023). By employing larger 2-body and 3-bodytraining sets, including an explicit machine-learned representationof 4-body energies, and adopting more sophisticated machine-learnedrepresentations of 2-body and 3-body energies, we demonstrate thatthe MB-pol(2023) PEFs achieve sub-chemical accuracy in modeling theenergetics of the hexamer isomers, outperforming both the originalMB-pol and q-AQUA PEFs, which currently provide the most accuratedescription of water clusters in the gas phase. Importantly, the MB-pol(2023)PEFs provide remarkable agreement with the experimental results forvarious properties of liquid water, improving upon the original MB-polPEF and effectively closing the gap with experimental measurements.

Return to Publications page