, 1982 (1985). LCAO local de!sity calculations for ozone yield a ground state geometry in good agreement with experiment (R = 1.27 A vs. 1.278 A (exp.), 0 117.5" vs. 116.8" (exp.)). A second local minimum is found about 45 kcal/mol higher for a cyclic geometry (R = 1.44 A, 8 = 60"). For S, the calculations predict a bent ground state (R = 2.00 A, 0 = 1 16") with the cyclic geometry (R = 2.125 A, 0 = 58") about 15 kcal/mol higher.MARIO MORIN, ANIKO E. FOTI et DENNIS R. SALAHUB. Can. J. Chem. 63, 1982Chem. 63, (1985. Des rCsultats de calculs du type LCAO densit6 locale sont en bon accord avec I'expCrience pour la gComCtrie de I'ttat fondamental de I'ozone (R = 1.27A vs. 1.278 A (exp.); 0 = 117.5" vs. 116.8" (exp.)). Un deuxikme minimum local pour une gComCtrie cyclique (R = 1.44 A, 0 = 60' 1 est prCdit a une Cnergie plus 6levCe d'environ 45 kcal/mol. Pour S3 les calcujs prkdisent une gComCtrie ouverte (R = 2.00 A, 0 = 116") pour I'Ctat fondamental avec la forme cyclique (R = 2.125 A, 0 = 58") i peu prks 15 kcal/mol plus haute.
IntroductionThe ozone molecule has been the object of numerous experimental (1-6) and theoretical (7-13) studies. Because of its high biradical character (7, 9, 11, 12) it has become a classic example of a case where single determinantal Hartree-Fock theory is inadequate. In the usual wave function theory at least two determinants must be mixed to achieve a reasonable qualitative description. On the other hand, X a (14-16) calculations, which should be viewed as approximate applications of density functional theory (17, 18), have yielded the correct singlet ground state and accurate values for the ionization potentials and many low-lying excitation energies (8, 13). These X a calculations were performed with either the scattered-wave (16) or LCAO (Gaussian) (19,20) version of the method and only the equilibrium geometry was considered. A direct optimization of the geometry was carried out by Laidlaw and Trsic (10) using the discrete variational (21, 22) X a (Hartree-Fock-Slater) method. 'Their results were discouraging in that a closed geometry having a bond angle near 60" and involving six n electrons was predicted to lie lower in energy than the experimentally observed open form (8 = 1 17") which involves four n electrons.Ozone's third row congener thiozone is less well characterized. Although S1 has been observed in sulfur vapour and liquid and its ionization potential is known (23) its structure has not been experimentally determined. Only a few electronic structure calculations have been performed for S3. The Xa-SW method was applied (24) to the study of structural changes accompanying oxidation or reduction. Examination of the orbital energies and wave functions led to the suggestion that an open form with an angle of about 120" would be a reasonable choice for the ground state.