A Straightforward and Efficient Approach for Enhancing Energetics within the Framework of Density-Corrected Density Functional Theory

Add to your list(s) Download to your calendar using vCal

• Daniel Graf, University of Cambridge, Cambridge, United Kingdom
• Monday 12 June 2023, 14:30-15:00
• Yusuf Hamied Department of Chemistry, Todd-Hamied Room.

If you have a question about this talk, please contact Maria-Andreea Filip.

https://zoom.us/j/92447982065?pwd=RkhaYkM5VTZPZ3pYSHptUXlRSkppQT09

Density functional theory (DFT) is widely recognised for its cost-effectiveness and is extensively used in chemistry, physics, and material science. However, standard density functional approximations (DFAs) suffer from various inherent issues, one prominent among them being the over-delocalisation of charge. This charge over-delocalisation can lead to inaccurate densities and, consequently, significant errors in energetic properties. Previous research by Burke and colleagues¹ demonstrated that these so-called density-driven errors can be separated from functional errors and efficiently addressed by evaluating the density functional on the Hartree-Fock (HF) density instead of the self-consistent one, leading to the development of density-corrected Hartree-Fock density functional theory (DC(HF)-DFT).

In this forthcoming talk, we present a straightforward yet efficient procedure for conducting DC(HF)-DFT calculations. Our approach involves incorporating a non-interacting kinetic energy-based heuristic to determine whether the self-consistent DFA density should be replaced by the HF density. Additionally, we introduce a procedure that optimises the utilisation of computed quantities required for DC(HF)-DFT. This involves evaluating a related hybrid functional with minimal computational cost, enabling us not only to “correct” the density but also the functional itself. Furthermore, we extend these ideas to double-hybrids, enhancing the applicability of our procedures.

To substantiate the effectiveness and efficiency of our proposed methodologies, we showcase results obtained from a diverse range of chemical problems sourced from the GMTKN55 [2] database. The results demonstrate that our procedures significantly enhance the accuracy of DFA calculations while maintaining computational efficiency to a large extent.

S. Vuckovic, S. Song, J. Kozlowski, E. Sim, K. Burke J. Chem. Theory Comput. 15, 6636 (2019).

L. Goerigk, A. Hansen, C. Bauer, S. Ehrlich, A. Najibi, S. Grimme Phys. Chem. Chem. Phys. 19, 32184 (2017).

This talk is part of the Lennard-Jones Centre series.

This talk is included in these lists:

• Hanchen DaDaDash
• Lennard-Jones Centre
• Yusuf Hamied Department of Chemistry, Todd-Hamied Room

Note that ex-directory lists are not shown.