Density functional theory for charge transfer: The nature of the N-bands of porphyrins and chlorophylls revealed through CAM-B3LYP, CASPT2, and SAC-CI calculations

While density functional theory (DFT) has been proven to be extremely useful for the prediction of thermodynamic and spectroscopic properties of molecules, to date most functionals used in common implementations of DFT display a systematic failure to predict the properties of charge-transfer processes. While this is explicitly manifest in Rydberg transitions of atoms and molecules and in molecular charge-transfer spectroscopy, it also becomes critical for systems containing extended conjugation such as polyenes and other conducting polymers, porphyrins, chlorophylls, etc. A new density functional, a Coulomb-attenuated hybrid exchange-correlation functional (CAM-B3LYP), has recently been developed specifically to overcome these limitations, and it has been shown to properly predict molecular charge-transfer spectra. Here, we demonstrate that it predicts qualitatively reasonable spectra for porphyrin, some oligoporphyrins, and chlorophyll. However, alternate density functionals developed to overcome the same limitations such as currentdensity functional theory are shown, in their present implementation, to remain inadequate. The CAM-B3LYP results are shown to be in excellent agreement with complete-active-space plus second-order M011er - Plesset perturbation theory and symmetry-adapted cluster configuration interaction calculations: These depict the N and higher bands of porphyrins and chlorophylls as being charge-transfer bands associated with localization of molecular orbitals on individual pyrrole rings. The validity of the basic Gouterman model for the spectra of porphyrins and chlorophylls is confirmed, rejecting modern suggestions that non-Gouterman transitions lie close in energy to the Q-bands of chlorophylls. As porphyrins and chlorophylls provide useful paradigms for problems involving extended conjugation, the results obtained suggest that many significant areas of nanotechnology and biotechnology may now be realistically treated by cost-effective density-functional-based computational methods