The ozonide anion: A theoretical study

Abstract

Complete active space self-consistent field (CASSCF) and CASSCF second-order perturbation theory (CASPT2) methods have been used for the geometry optimization and calculation of harmonic and fundamental frequencies of the ozonide ion (O−3) and the ozonide lithium complex (Li+O−3). For O−3 harmonic frequencies have also been obtained using the coupled-cluster method, CCSD(T). Infrared intensities are computed from dipole moment derivatives at the CASSCF level. The predicted equilibrium geometry for O−3 is ROO=1.361 Å and ∠OOO=115.4°, and the fundamental frequencies are ν1=989 cm−1, ν2=556 cm−1, ν3=870 cm−1 [experimental values are ROO=1.36±0.02 Å, ∠OOO=111.8±2.0°, ν1=975(50) cm−1, ν2=550(50) cm−1, ν3=880(50) cm−1]. Corresponding data for the lithium ozonide complex have also been obtained. The presented data contradict the previous interpretation of the IR and Raman spectrum obtained after codeposition of ozone and alkali atoms in N2, argon, or neon matrices. The presence of the lithium cation raises the asymmetric stretch frequency to about 940 cm−1, which is contradictory to assumptions made in the assignments of the matrix spectra. Calculations made in a dielectric medium strongly suggest that the effect of the matrix on the IR spectrum is small for O−3 itself. The dissociation and atomization energies of O−3 are found to be in agreement with experiment.