Assessing the Interplay between Functional-Driven and Density-Driven Errors in DFT Models of Water

• DOI: 10.1021/acs.jctc.2c00050
• Publication Date: 2022-05-04T13:59:48Z
• Authors: Etienne Palos; Eleftherios Lambros; Steven Swee; Jie Hu; Saswata Dasgupta; Francesco Paesani
• Keywords: 13. Climate action; 0103 physical sciences; Water; Molecular Dynamics Simulation; 01 natural sciences; 6. Clean water

We investigate the interplay between functional-driven and density-driven errors in different density functional approximations within theory (DFT) implications of these for simulations water with DFT-based data-driven potentials. Specifically, we quantify two widely used dispersion-corrected functionals derived generalized gradient approximation (GGA), namely BLYP-D3 revPBE-D3, modern meta-GGA functionals, strongly constrained appropriately normed (SCAN) B97M-rV. The effects on interaction energies are first assessed clusters BEGDB dataset. Further insights into nature gained from applying absolutely localized molecular orbital energy decomposition analysis (ALMO-EDA) to energies, which demonstrates that correlated interactions. discuss cases where density-corrected DFT (DC-DFT) models display higher accuracy than original reducing leads larger deviations reference due presence large errors. Finally, dynamics performed many-body potentials DC-DFT data determine effect minimizing has description liquid water. Besides rationalizing performance water, believe our findings unveil fundamental relations shortcomings some common requirements accurate descriptions interactions, will aid development a consistent, framework machine-learned condensed-phase systems.