Development and assessment of a new hybrid density functional model for thermochemical kinetics

A new hybrid Hartree-Fock-density functional model called the Becke88-Becke95 1-parameter model for kinetics (BB1K) was optimized against a database of three forward barrier heights, three reverse barrier heights, and three energies of reaction for the reactions in the BH6 representative barrier height database. We then assessed the newly developed BB1K method against a saddle point geometries database, a database of 42 barrier heights, the AE6 representative atomization energy database, a molecular geometries data set, and a set of 13 zero point energies. The results show that BB1K can give excellent saddle point geometries and barrier heights, and its performance for calculating atomization energies is 40% better than MPW1K. Using a mean mean unsigned error criterion that equally weights the errors in barrier heights and in bond energies, the new BB1K method outperforms all other DFT and hybrid DFT methods by a large margin, and we therefore conclude that it is the best density functional-type method for thermochemical kinetics.