Assessing the Performances of CASPT2 and NEVPT2 for Vertical Excitation Energies br

Methods able to simultaneously account for both staticand dynamic electron correlations have often been employed, not only tomodel photochemical events but also to provide reference values forvertical transition energies, hence allowing benchmarking of lower-ordermodels. In this category, both the complete-active-space second-orderperturbation theory (CASPT2) and theN-electron valence state second-order perturbation theory (NEVPT2) are certainly popular, the latterpresenting the advantage of not requiring the application of the empiricalionization-potential-electron-affinity (IPEA) and level shifts. However,the actual accuracy of these multiconfigurational approaches is not settled yet. In this context, to assess the performances of theseapproaches, the present work relies on highly accurate (+/- 0.03 eV) aug-cc-pVTZ vertical transition energies for 284 excited states ofdiverse character (174 singlet, 110 triplet, 206 valence, 78 Rydberg, 78 n ->pi*, 119 pi ->pi*, and 9 double excitations) determinedin 35 small- to medium-sized organic molecules containing from three to six non-hydrogen atoms. The CASPT2 calculations areperformed with and without IPEA shift and compared to the partially contracted (PC) and strongly contracted (SC) variants ofNEVPT2. Wefind that both CASPT2 with IPEA shift and PC-NEVPT2 provide fairly reliable vertical transition energy estimates,with slight overestimations and mean absolute errors of 0.11 and 0.13 eV, respectively. These values are found to be rather uniformfor the various subgroups of transitions. The present work completes our previous benchmarks focused on single-reference wavefunction methods (J. Chem. Theory Comput.2018,14, 4360;J. Chem. Theory Comput.2020,16, 1711), hence allowing for a faircomparison between various families of electronic structure methods. In particular, we show that ADC(2), CCSD, and CASPT2deliver similar accuracies for excited states with a dominant single-excitation character.