Comparison between hybrid functionals free of adjustable parameters and symmetry-adapted cluster–configuration interaction for electronically excited states of organic compounds: TD-PBE0-1/3 is better than expected

Abstract

Developing and benchmarking methodological approaches able to accurately predict the electronically excited states are major challenges for theoretical chemists. In this work, we analyze the performances of the hybrid density functionals based on Perdew, Burke, and Ernzerhof (PBE) with a single non-fitted parameter for Hartree–Fock (HF) exchange and without invoking to any fitting process for predicting the valence and Rydberg excitation energies of organic compounds in the framework of time-dependent density functional theory (TD-DFT). As a wave function theory-based approach in the context of excite states calculations, symmetry-adapted cluster–configuration interaction (SAC–CI) method including single- and double-linked excitation operators has been considered. Our test calculations show that the SAC–CI gives the best performance for Rydberg excited states, while for valence excited states the PBE hybrid functional with parameter 1/3 for HF exchange (PBE0-1/3) outperforms other methods. Overall, of the several combinations of functionals and parameters, the TD-PBE0-1/3 is found to offer the best performance with respect to others and its validity compared to SAC–CI has also been verified by computing low-lying excited states of a few molecules. From the viewpoint of compromising between accuracy and computational cost, these findings reveal that how nonempirical hybrid functionals are economical to model electronic excitation energies of various systems and still much better achievements can be obtained in this respect.