When a Ne:SO2 mixture is subjected to Penning ionization and/or photoionization by neon atoms in their first excited states, between 16.6 and 16.85 eV, and the products are rapidly frozen at approximately 5 K, the infrared spectrum of the resulting deposit includes absorptions assigned with the aid of isotopic substitution studies to SO, SO2+, SO2−, (SO2)2−, and, tentatively, SO−. The fundamental and first overtone absorptions of SO lie 0.9 and 1.8 cm−1, respectively, below the gas-phase band centers. Ab initio calculations at the Hartree–Fock level show an instability in the v3 vibration of SO2+ which is avoided by higher-level calculations. The ν3 and ν1 fundamentals of SO2− isolated in solid neon are identified at 1086.2 and 990.8 cm−1, respectively. In agreement with an earlier proposal, the 1042 cm−1 absorption originally assigned to ν3 of SO2− trapped in solid argon is reassigned to MSO2, with M an alkali metal. Near the photodetachment threshold for SO2− isolated in a neon matrix, electron capture by SO2 nearest-neighbor pairs results in growth of infrared absorptions of (SO2)2−, which has been shown by gas-phase studies to have a significantly higher photodetachment threshold than does SO2−. The isotopic substitution studies require that the two sulfur atoms in (SO2)2− be nonequivalent, favoring the linking of the two SO2 units by a S ⋯ O bond.

The superoxide isomer of sulfur dioxide (Fig. 1) was first proposed by Myerson, Taylor, and Hanst in 1957 [J. Chem. Phys. 26, 1309 (1957)] as a possible intermediate in the combustion of CS2, COS, and H2S as well as a possible source of some troublesome ultraviolet absorptions in the spectra associated with those combustion processes. Subsequent experimental work on SO2 has also referred to the hypothesized asymmetric structure. Single reference post Hartree–Fock methods, including configuration interaction with single and double excitations (CISD), coupled cluster singles and doubles (CCSD), and coupled cluster singles and doubles with perturbative triples [CCSD(T)], as well as multireference configuration interaction (MRCISD) performed with CI natural orbitals (CINOs) have been employed in the interest of characterizing the relative energies of the open, ring and superoxide isomers of SO2. The largest basis used was a triple-ζ plus double polarization function set with f-type functions appended to each atom [TZ2P(f)]. The ring and superoxide isomers are predicted to lie approximately 111 and 104 kcal mol−1, respectively, above the open isomer ground state. Based upon these energy separations, it is predicted that neither the ring nor superoxide isomers are responsible for the troublesome UV absorption continuum, as postulated by Myerson et al. Moreover, neither the ring nor the superoxide structure is the source of the spectroscopic features very recently observed below 100 kcal mol−1 by Dai’s group.

Inspired by the recent experimental study of the radical anions HCCN− and HCNC− and by earlier examinations of HCCN, the equilibrium geometry of the HCNC molecule has been investigated using both self-consistent field (SCF) and configuration interaction methods including single and double excitations (CISD). The largest basis set used was a triple-ζ plus double polarization with diffuse functions and higher angular momentum functions appended to each atom [TZ2P(f,d)+diff]. Using this basis, the H–C–N equilibrium angle is predicted to be 128.5° at the CISD level of theory. Additionally, the zero point vibrational energy (ZPVE) corrected energy separation of the bent and linear conformations was predicted to be 10.1 kcal mol−1 at the CISD level of theory with the largest basis set employed. The barrier to linearity is 7.7 kcal mol−1 at the CCSD level of theory and 6.9 kcal mol−1 at the CCSD(T) level of theory, employing the CISD optimized geometries with a basis that was comprised of triple-ζ plus double polarization with higher angular momentum functions appended to each atom [TZ2P(f,d)]. These results were compared to those obtained in previous ab initio investigations of HCCN, which has been dubbed a quasilinear molecule by the most recent experimental investigators. HCNC is predicted to lie 22.2 kcal mol−1 above HCCN at the CISD level of theory, with a the TZ2P(f,d) basis. The differences between the two isomers are discussed and HCNC is predicted to be a definitively bent molecule, rather than quasilinear.