Benchmarking the performance of time-dependent density functional methods.

Area of Science:

• Quantum Chemistry
• Computational Chemistry
• Theoretical Chemistry

Background:

• Accurate calculation of vertical excitation energies is crucial for understanding molecular electronic properties.
• Density Functional Theory (DFT) is a widely used method, but its performance varies significantly with functional choice.
• A comprehensive assessment of modern DFT functionals against experimental data is needed.

Purpose of the Study:

• To evaluate the performance of 24 density functionals, including 14 meta-generalized gradient approximation (mGGA) functionals, for calculating vertical excitation energies.
• To compare these computational results against a robust experimental benchmark set.
• To identify the most accurate and reliable functionals for predicting excited states.

Main Methods:

• Assessment of 24 density functionals, with a focus on 14 mGGA functionals.
• Calculation of vertical excitation energies for 101 excited states across 14 small- to medium-sized molecules.
• Comparison against an experimental benchmark set encompassing singlet, triplet, valence, and Rydberg excited states.

Main Results:

• The global-hybrid (GH) PBE0 functional demonstrated strong overall performance with a mean absolute error (MAE) of 0.28 eV.
• The GH-mGGA M06-2X functional achieved a lower MAE of 0.26 eV but exhibited convergence issues.
• The Slater exchange and Volk-Wilk-Nusair (SVWN) local density approximation functional surpassed all tested non-GH GGAs, and M06-L showed the best performance among pure density functionals (MAE 0.41 eV).

Conclusions:

• Global-hybrid functionals, particularly PBE0, offer superior accuracy for vertical excitation energies.
• While M06-2X shows high accuracy, its convergence problems warrant caution.
• The study provides valuable guidance for selecting appropriate DFT functionals in computational chemistry for excited-state calculations.