Abstract
Criegee intermediates (i.e., carbonyl oxides with two radical sites) are known to be important atmospheric reagents; however, our knowledge of their reaction kinetics is still limited. Although experimental methods have been developed to directly measure the reaction rate constants of stabilized Criegee intermediates, the experimental results cover limited temperature ranges and do not completely agree well with one another. Here we investigate the unimolecular reaction of acetone oxide [(CH(3))(2)COO] and its bimolecular reaction with H(2)O to obtain rate constants with quantitative accuracy comparable to experimental accuracy. We do this by using CCSDT(Q)/CBS//CCSD(T)-F12a/DZ-F12 benchmark results to select and validate exchange-correlation functionals, which are then used for direct dynamics calculations by variational transition state theory with small-curvature tunneling and torsional and high-frequency anharmonicity. We find that tunneling is very significant in the unimolecular reaction of (CH(3))(2)COO and its bimolecular reaction with H(2)O. We show that the atmospheric lifetimes of (CH(3))(2)COO depend on temperature and that the unimolecular reaction of (CH(3))(2)COO is the dominant decay mode above 240 K, while the (CH(3))(2)COO + SO(2) reaction can compete with the corresponding unimolecular reaction below 240 K when the SO(2) concentration is 9 × 10(10) molecules per cubic centimeter. We also find that experimental results may not be sufficiently accurate for the unimolecular reaction of (CH(3))(2)COO above 310 K. Not only does the present investigation provide insights into the decay of (CH(3))(2)COO in the atmosphere, but it also provides an illustration of how to use theoretical methods to predict quantitative rate constants of medium-sized Criegee intermediates.