Neither Cyclic HgO3 nor Biradicaloid OHgOO Participate in the Gas-phase Oxidation of Hg(0) by Ozone

Redox chemistry largely controls when and where mercury enters ecosystems from the atmosphere. Kinetics experiments in static chambers near room temperature indicated that ozone could oxidize gaseous elemental mercury, Hg(0), at a rate constant, that while low (3-80 × 10-20 molecule-1 cm3 s-1 at 293-298 K), is high enough to be important to its fate. However, there are no known gas-phase products that are kinetically feasible for this reaction, so it is now suspected that the chamber walls catalyzed the consumption of Hg(0) observed in experiments. The present work examines two products hypothesized by Calvert and Lindberg in 2005 to rationalize a gas phase Hg(0) + O3 reaction: OHgOO and a cyclic HgO3 compound. We do not find cyclic HgO3 as a minimum in potential energy; instead, it falls apart to an Hg-O3 van der Waals complex bound by about 1 kcal mol-1. The OHgOO molecule is biradicaloid, and was treated with multireference (MR) methods, including MRPT2 and MRCISD + Q. We find OHgOO lies >21 kcal mol-1 above Hg(0) + O3. We used the thermodynamic data to reanalyze the kinetic experiments. We find that the rate of loss of Hg(0) in experiment due to gas phase reactions would have corresponded to an apparent second-order rate constant ≈10-33 molecule-1 cm3 s-1 at 293-298 K, a full 13 orders of magnitude lower than the lowest reported experimental rate constant. We conclude that the experimental results arise from chemistry on the walls of reactors, and that the gas phase oxidation of Hg(0) by ozone is irrelevant in the atmosphere.