Reply to Comment on ‘Failure of time-dependent density functional methods for excitations in spatially separated systems’ by Andreas Dreuw and Martin Head-Gordon

We confirm our view that zero or small overlap between individual subsystems of a larger system due to spatial separation causes a failure of present-day time-dependent density-functional theory methods for certain excitation energies. This view, very generally, covers typical charge-transfer excitations between a donor and an acceptor as first analysed by Dreuw and Head-Gordon, excitations in symmetric supersystems with spatially well-separated components without donor–acceptor character, as well as cases lying inbetween.