(62o) Elevating Density Functional Theory Towards Chemical Accuracy for Condensed Phase Simulations through Machine Learning and Many-Body Techniques

Density functional theory (DFT) has been extensively used to model the properties of condensed-phase systems.

Albeit maintaining a good balance between accuracy and efficiency, no density functional has so

far achieved the degree of accuracy necessary to correctly predict the properties of hydrated systems across

the entire phase diagram. Here, we present density-corrected SCAN (DC-SCAN) method for

water and other hydrated systems which, minimizing density-driven errors, elevate the accuracy of the SCAN functional to

that of coupled-cluster theory, the “gold standard” for chemical accuracy. Building upon the

accuracy and efficiency of DC-SCAN within a many-body formalism, we introduce a data-driven

many-body potential energy function, the MB-SCAN(DC)PEF, that can quantitatively reproduce

coupled-cluster reference energetics of water clusters. Importantly, the properties of liquid

water calculated from molecular dynamics simulations carried out with the MB-SCAN(DC) PEF

are found to be in excellent agreement with the experimental data, which thus demonstrates

that MB-SCAN(DC) is effectively the first DFT-based model that correctly describes water from

the gas to the liquid phase. Apart from neutral water, we present extensive calculations aimed at determining the accuracy

of the DC-SCAN functional for various aqueous systems. DC-SCAN shows remarkable consistency

in reproducing reference data obtained at the coupled-cluster level of theory, with minimal loss

of accuracy. The accuracy of DC-SCAN has been further extended by integrating the capabilities of neural network potentials, enabling the generation of reactive machine-learned potentials. These were used to investigate rare-events such as the autoionization of liquid water within the confinement of a carbon nanotube, with the help of enhanced sampling.