Ab initio molecular orbital calculations with split-valence plus polarization basis sets and incorporating valenceelectron correlation have been performed to determine the equilibrium structure of ethyloxonium (I CH,CH,OH, 1 +) and examine its modes of unimolecular dissociation. An asymmetric structure (1) is predicted to be the most stable form of ethyloxonium, but a second conformational isomer of C, symmetry lies only 1.4 kJ mol-' higher in energy than 1. Four unimolecular decomposition pathways for 1 have been examined involving loss of H, , CH, , H,O or C,H, . The most stable fragmentation products, lying 65 kJ mol-' above 1, are associated with the H, elimination reaction. However, large barriers of 257 and 223 kJ mol-' have to be surmounted for H, and CH, loss, respectively. On the other hand, elimination of either C,H, or H,O from ethyloxonium can proceed without a barrier to the reverse associations and, with total endothermicities of 130 and 160 kJ mol-', respectively, these reactions are expected to dominate at lower energies. A second important equilibrium structure on the surface is a hydrogen-bridged complex, lying 53 kJ mol-' above 1. This complex is involved in the C,H, elimination reaction, acts as an intermediate in the proton-transfer reaction connecting [C,Hs]+ + H,O and C,H, + IH,0] + and plays an important role in the isotopic scrambling that has been observed experimentally in the elimination of either H,O or C2H4 from ethyloxonium. The proton affinity of ethanol was calculated as 799 kJ mol-', in close agreement with the experimental value of 794 kJ mol-' . et al. ' '

In this work, theoretical rate constants are estimated for mercury oxidation reactions by hydrogen chloride that may occur in the flue gases of coal combustion. Rate constants are calculated using transition state theory at the quadratic configuration interaction (QCI) level of theory with single and double excitations, and are compared to results obtained from density functional theory, both including high level pseudopotentials for mercury. Thermodynamic and kinetic data from the literature are used to assess the accuracy of the theoretical calculations when possible. Validation of the chosen methods and basis sets is based upon previous and current research on mercury reactions involving chlorine. The present research shows that the QCISD method with the 1992 Stevens et al. basis set leads to the most accurate kinetic and thermodynamic results for the oxidation of mercury via chlorine containing molecules. Also, a comparison of the heats of reaction data for a series of mercury oxidation reactions reveals that the density functional method, B3LYP, with the 1997 Stuttgart basis set provides reasonably accurate results for these large systems.

No abstract

No abstract