Laplacian-level meta-GGA for the exchange-correlation energies of metals*

Common meta-generalized gradient approximations (meta-GGAs) for the exchange-correlation energy of many-electron systems use the non-interacting kinetic energy density (KED) to identify and describe local chemical properties. KED-level meta-GGAs are exceedingly accurate for diverse insulating systems, but often worsen GGA-level descriptions of metallic systems. This talk will motivate the development of a non-empirical Laplacian-level meta-GGA using semiclassical turning surface arguments [1] and observed trends. I will discuss how exact constraints can be built into a Laplacian-level meta-GGA and the limitations inherent to the level of approximation. The functional form, numeric efficiency, and accuracy of such a functional will also be presented, particularly for the structural, magnetic, and energetic properties of alkalis and transition metals. Comparisons to recent KED-level meta-GGAs [2] and an outlook on future meta-GGAs will be featured.

[1] A. D. Kaplan, S. J. Clark, K. Burke, and J. P. Perdew, arXiv:2007.01925 (2020).
[2] J. W. Furness, A. D. Kaplan, J. Ning, J. P. Perdew, and J. Sun, J. Phys. Chem. Lett. 11, 8208 (2020).