Density functionals in a 1-electron reduced density matrix theory framework - toward a solution to the static correlation error
ORAL
Abstract
Kohn Sham Density Functional Theory (DFT) has become a primary method for investigating chemical systems due to its excellent cost to accuracy ratio. However, it is known to struggle in the treatment of strongly correlated systems due to its single reference nature. Previously, we've combined DFT with 1-Electron Reduced Density Matrix Functional Theory (1-RDMFT) to address this limitation. By combining these two methodologies we are able to capture strong correlation effects with a 1-RDM based functional while relying on existing DFT functionals to capture the remaining dynamical correlation. While this framework yields large improvements as compared to standard DFT in the treatment of multi-reference problems, it relies on existing DFT functionals, many of which were developed under the assumption that the 1-RDM remains idempotent. Our approach breaks this assumption. Here, we benchmark a range of available density functionals for their ability to capture strong correlation and work towards developing a task-specific density functional for use within our 1-RDMFT framework.
–
Publication: The Journal of Physical Chemistry Letters 13 (6), 1382-1388<br>Journal of Chemical Theory and Computation 18 (11), 6600-6607<br>Physical Review Letters 131 (24), 243003<br>Phys. Rev. A 110, L040802<br>Manuscript in preparation
Presenters
-
Daniel Gibney
University of Minnesota
Authors
-
Daniel Gibney
University of Minnesota
-
Jan Niklas Boyn
University of Minnesota