Slater-Koster tables for density functional tight-binding (DFTB) are pre-computed molecular integrals constructed from atomic densities. In computing these atomic densities, a confinement potential is applied in order to mimic the atoms-in-molecule atomic densities. However, it is standard practice to apply the same confinement potential across all orbital types, this results in an inappropriate description of lone-pair interactions. In a recently study, our group introduced a confinement potential that is allowed to vary across all orbital types. Using this new orbital dependent confinement method (ODC) we are able to better capture lone-pair interactions, leading to a better geometrical description of oxy- and nitro-compounds without the need to introduce empirical corrections (Jenness et al., J. Chem. Theory Comp. 16, 6894 (2020)). An in-depth analysis of the molecular orbitals reveals that the origin of the improvement is due to the spacing between orbital eigenvalues being in better alignment with those from density functional theory (DFT). In this talk, we will explore these effects in addition to exploring the question of “If we are getting orbital energy differences better, then what about excitation energies?” Herein, we compare our results to standard TD-DFT, linear-response TD-DFT, and EOM-CCSD with our current DFTB model and previously used DFTB models in order to determine the influence of our orbital dependent scheme on this critical chemical phenomena.

Adventures in DFTB: Structural and optical properties

Fall 2022 abstracts, posters, and presentations will launch in mid-September.

ACS Fall 2022

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.