Excited-State Absorption of Uracil in the Gas Phase: Mapping the Main Decay Paths by Different Electronic Structure Methods

We present a computational study of the one-photon and excited-state absorption (ESA) from the two lowest energy excited states of uracil in the gas phase: an n pi* dark state (1n) and the lowest energy bright pi pi* state (1 pi). The predictions of six different linear response electronic structure methods, namely, TD-CAM-B3LYP, EOM-CCSD, EOM-CC3, ADC(2), ADC(2)-x, and ADC(3) are critically compared. In general, the spectral shapes predicted by TD-CAM-B3LYP, EOM-CCSD, EOM-CC3, and ADC(3) are fairly similar, though the quality of TD-CAM-B3LYP slightly deteriorates in the high-energy region. By computing the spectra at some key structures on different potential energy surfaces (PES), that is, the Franck-Condon point, the 1n minimum, and structures representative of different regions of the 1 pi PES, we obtain important insights into the shift of the ESA spectra, following the motion of the wavepacket on the excited-state PES. Though 1 pi has larger ESA than 1n, some spectral regions are dominated by these latter signals. Aside from its methodological interest, we thus obtain interesting indications to interpret transient absorption spectra to disentangle the photoactivated dynamics of nucleobases.

Automated summary

This study examines the computational predictions of one-photon and excited-state absorption spectra of uracil, comparing six different linear response electronic structure methods. The spectral shapes predicted by some methods are fairly similar, although there may be slight deterioration in quality in the high-energy region. Insight into the shift of ESA spectra on different potential energy surfaces is obtained, revealing that despite 1 pi having larger ESA, some regions are dominated by signals from the 1n state.