Which DFT functional for excited states of organic dyes?

I am benchmarking TD-DFT for a series of push–pull organic dyes. B3LYP underestimates charge-transfer excitations badly, as expected. CAM-B3LYP and ωB97X-D look better but disagree with each other by up to 0.3 eV. Without experimental spectra for half of the set, how would you choose? Is it worth running a few higher-level reference calculations?

Finish signup to comment

Comments

@pawel_michalski

43d

Yes, I am leaning towards a handful of reference calculations on the smallest dyes, then picking the functional that tracks them best.

@natalia_mazur

43d

Not my field, but in astro we would always anchor to a smaller high-accuracy reference set before trusting the cheap method on everything. Same logic?