Perturbative ensemble density functional theory applied to charge transfer excitations

Perturbative ensemble density functional theory (pEDFT) has recently been suggested as an easy-to-implement and low-cost route to robust prediction of molecular excitation energies. Because charge transfer excitation energies are known to be challenging to standard time-dependent density functional theory, it is of interest to examine how pEDFT performs in this important case. Here, we examine this issue analytically and numerically, with an emphasis of examining appropriate limiting cases. Specifically, we observe the emergence of a self-interaction-like term that adversely affects an accurate prediction.