DFT‐D3 for Solid‐State Materials: A Data‐Driven Perspective on Accuracy

Abstract

Dispersion interactions are essential for accurate modeling of materials, yet standard Density Functional Theory (DFT) lacks explicit treatment. Grimme's DFT‐D3 method adds empirical dispersion corrections and performs well for molecular systems, but its reliability in solids remains uncertain. Here, DFT‐D3, including zero‐damping and Becke–Johnson variants, is systematically assessed across more than 2,000 lithium‐, potassium‐, magnesium‐, zinc‐, copper‐, and manganese‐based solid‐state binary compounds using a homegrown automated workflow integrating VASP, pymatgen, and a custom D3 interface. It is found that D3 often introduces artificial secondary minima in potential energy surfaces, particularly in metallic and densely packed solids, leading to unphysical stabilization and significant energy errors. The successor scheme D3S is also examined, and although it reduces dispersion errors relative to D3, it still produces artificial stabilization in ionic and metallic systems, indicating that its molecular‐level improvements do not fully extend to solids. Comparisons with many‐body dispersion (MBD@rsSCS) calculations show that these artifacts arise from abrupt coordination‐number‐dependent variations in the coefficients. Since the errors lack consistent correlation with structural or chemical descriptors, they remain difficult to predict. Overall, DFT‐D3 is reliable for covalent frameworks and low‐metal‐content intercalation systems but requires caution in dense, metal‐rich materials to avoid misleading overbinding.