| Abstract
| - The reaction enthalpy (298 K), barrier (0 K), and activation energy and preexponential factor (600−800 K)have been examined computationally for the abstraction of hydrogen from benzene by the methyl radical, toassess their sensitivity to the applied level of theory. The computational methods considered include high-level composite procedures, including W1, G3-RAD, G3(MP2)-RAD, and CBS-QB3, as well as conventionalab initio and density functional theory (DFT) methods, with the latter two classes employing the 6-31G(d),6-31+G(d,p) and/or 6-311+G(3df,2p) basis sets, and including ZPVE/thermal corrections obtained from 6-31G(d) or 6-31+G(d,p) calculations. Virtually all the theoretical procedures except UMP2 are found to givegeometries that are suitable for subsequent calculation of the reaction enthalpy and barrier. For the reactionenthalpy, W1, G3-RAD, and URCCSD(T) give best agreement with experiment, while the large-basis-setDFT procedures slightly underestimate the endothermicity. The reaction barrier is slightly more sensitive tothe choice of basis set and/or correlation level, with URCCSD(T) and the low-cost BMK method providingvalues in close agreement with the benchmark G3-RAD value. Inspection of the theoretically calculated rateparameters reveals a minor dependence on the level of theory for the preexponential factor. There is moresensitivity for the activation energy, with a reasonable agreement with experiment being obtained for the G3methods and the hybrid functionals BMK, BB1K, and MPW1K, especially in combination with the 6-311+G(3df,2p) basis set. Overall, the high-level G3-RAD composite procedure, URCCSD(T), and the cost-effectiveDFT methods BMK, BB1K, and MPW1K give the best results among the methods assessed for calculatingthe thermochemistry and kinetics of hydrogen abstraction by the methyl radical from benzene.
|
