Theoretical studies on the electronic spectrum of selenium dioxide. Comparison with ozone and sulfur dioxide Friedrich Grein * Department of Chemistry and Centre for Laser, Atomic and Molecular Sciences, University of New Brunswick, P.O. Box 4400, Fredericton, N.B., Canada E3B5A3 article info Article history: Received 12 February 2009 Accepted 30 March 2009 Available online 5 April 2009 Keywords: Selenium dioxide MRCI CCSD(T) DFT Excited state geometries Vertical excitation energies Adiabatic excitation energies Potential curves 1-B 2 state Conical intersection 1-A 2 with 1-B 1 abstract MRCI vertical excitations energies and oscillator strengths of SeO 2 were calculated for singlet and triplet valence states up to about 8 eV, and for low-lying 5s-Rydberg states starting at 8.5 eV. CCSD(T) and DFT (B3PW91) geometry optimizations were performed for 20 excited states. There is good agreement with the experimental geometry of the ground state and of 1 1 B 2 . The calculated vibrational frequencies for the ground state are close to the observed frequencies. Relaxed DFT potential curves are given for the lowest singlet states in C 2v symmetry, and for the lowest 1 A 00 state, which bridges the conical intersection between 1 1 A 2 and 1 1 B 1 . Results are reported for adiabatic electron affinity, ionization potential, and dis- sociation energy. The SeOO structure lies about 3.4 eV above the OSeO ground state. Comparison is made with vertical excitation energies of first singlet and triplet states of O 3 and SO 2 . Ó 2009 Elsevier B.V. All rights reserved. 1. Introduction Experimental [1] and theoretical [2] studies have been reported for the X 1 A 1 ground state of SeO 2 . Recently, Crowther and Brown [3,4] performed laser excitation spectroscopy on the 313 nm band system, believed to be C 1 B 2  X 1 A 1 , with vibrational and rotational studies in the 292–327 nm range. They obtained the equilibrium geometry of C 1 B 2 , its adiabatic excitation energy and the vibra- tional frequencies m 1 and m 2 . Earlier work on this system by King and McLean [5,6] suggested a double minimum in the potential en- ergy curve of the q 0 3 vibration, indicating a distortion of the C state to C s symmetry. However, this could not be confirmed by Crowther and Brown. The antisymmetric stretch frequency m 3 is suggested to lie slightly higher than the symmetric stretch m 1 . No theoretical studies could be found in the literature on excited states of SeO 2 . Many experimental and theoretical studies have been reported on the electronic spectrum of the related molecules O 3 [7] and SO 2 [8]. Interestingly, work has also been done on TeO 2 , for which the analysis of the electronic spectrum between 345 and 406 nm gave evidence of a 1 B 2 state [9], which was confirmed by theoret- ical studies of Lee et al. [10]. In this paper, high-level multireference configuration interac- tion (MRCI) calculations will be performed at the ground state geometry, to give vertical excitation energies and oscillator strengths for valence and low-lying Rydberg states of SeO 2 , of sin- glet and triplet spin multiplicity. Vertical excitation energies calcu- lated by time-dependent density functional theory (TD-DFT), coupled cluster with single and double substitutions including non- iterative triple excitations (CCSD(T)) and DFT methods will be com- pared with MRCI results. Geometries for the ground state and C 2v excited states, including the observed 1 1 B 2 state, will be optimized by CCSD(T) and DFT. Geometries and vibrational frequencies are to be compared with experimental values for the ground state and for 1 1 B 2 . Geometry optimizations in C s symmetry will be performed for the lowest 1 A 00 state, corresponding to the lower C s state of the con- ical intersection between 1 1 A 2 and 1 1 B 1 . In addition, adiabatic elec- tron affinities and ionization potentials, as well as dissociation energies will be reported. While the main purpose of this paper re- lates to present and future spectroscopic work on SeO 2 , the perfor- mance of CCSD(T) and DFT methods for geometries and energies of excited states will also be tested. Such studies have previously been done on ozone [11] and NO 2 [12]. As described in these papers, sin- gle-determinant methods such as CCSD(T) and DFT can be applied to many, but not all excited states of symmetric molecules. They do not work for electronic configurations where an orbital is re- placed by another one of the same irrep. 0301-0104/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.chemphys.2009.03.021 * Tel.: +1 506 453 4776; fax: +1 506 453 4981. E-mail address: fritz@unb.ca Chemical Physics 360 (2009) 1–6 Contents lists available at ScienceDirect Chemical Physics journal homepage: www.elsevier.com/locate/chemphys