Accurately describing electronic properties of molecules in the condensed phase is challenging for theory since wavefunction methods are too expensive to model excited states for large molecules, let alone include explicit solvent . Density functional theory (DFT) is more affordable, but often over-polarizes the system and has spuriously low-energy charge-transfer states due to its delocalization error and lack of exact exchange. Long-range corrected (LC) or range separated hybrid (RSH) density functionals can be tuned to adjust the amount exact exchange in the system, which can greatly improve the delocalization error of DFT.I will present theoretical results for a test-set of 14 charge-transfer chromophores for which we have experimental solution phase absorption spectra and hyper-Rayleigh scattering measurements of the hyperpolarizabilty.  The amount of exact exchange that gives the best results for these experimental solution phase properties, as well as tuning to match Koopmans' theorem, is examined. Opposite size-dependent exact exchange trends are observed for excitation energies and Koopman's theorem: larger amounts of exact exchange are required for accurate excitation energies of large systems, while enforcing that that the ionization potential matches the HOMO energy for Koopmans' theorem requires less exact exchange for larger systems. Overall, these results suggest that tuning the amount of exact exchange based on Koopmans' theorem may be dominated by the delocalization error of DFT for large systems.  In addition, I will show that exact exchange tuning for Koopman's theorem when using a polarizable continuum model (PCM) yields very small amounts of exact exchange, resulting in a close to 'pure' density functional.

Optimum DFT exchange for accurate calculation of excitation energies and hyperpolarizabilities of charge-transfer chromophores

248th National Meeting (2014)

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