Exploring the space of meta-generalized gradient approximations for materials
Oral-In-person · Withdrawn
Abstract
Meta-generalized gradient (meta-GGA) approximations are often called “semilocal", because the exchange-correlation energy is found as a single integral over the three-dimensional space of an energy density at position r defined by the density and orbitals in an infinitesimal neighborhood of r. The functional form of recent meta-GGA’s, however, allows more flexibility than any previous semilocal approximation, with a high impact on applications [1], potentially competing with the more expensive hybrid density functionals.
Our hypothesis is that it is possible to design a computationally efficient meta-GGA that (1) preserves most or all the accuracy of r2SCAN [2] where r2SCAN is accurate, and (2) significantly further improves upon r2SCAN for the self-interaction-contaminated properties. This talk will give an overview of current efforts to address novel construction schemes of meta-GGAs with applications in materials.
Our hypothesis is that it is possible to design a computationally efficient meta-GGA that (1) preserves most or all the accuracy of r2SCAN [2] where r2SCAN is accurate, and (2) significantly further improves upon r2SCAN for the self-interaction-contaminated properties. This talk will give an overview of current efforts to address novel construction schemes of meta-GGAs with applications in materials.
–
Publication: [1] A. Giri, C. Shahi, A. Ruzsinszky, Isostructural α-γ phase transition in cerium from the perspective of meta-generalized gradient approximations, Phys. Rev. B 112, 125115 (2025)
[2] J.W. Furness, A.D. Kaplan, J. Ning, J.P Perdew and J. Sun, Accurate and Numerically Efficient r2SCAN Meta-Generalized Gradient Approximation, J. Phys. Chem. Lett. 11, 8208 (2020)
Presenters
-
Adrienn Ruzsinszky
- Tulane University