(448h) Homogeneous Catalysis of Ketene Production By Triethylphosphate
Stable intermediate species and transition-state structures included in the proposed mechanism were simulated using computational quantum chemistry software (Gaussian 09). Initial structure exploration and optimization were carried out using DFT methods at a B3LYP/6-31G(d,p) level of theory. Refined structures and their thermochemical properties were calculated with the composite method CBS-QB3. Hindered-rotation anharmonicity was considered for hydroxyl and methyl rotations in the structures using rotational energy profiles calculated at a B3LYP/6-31G++(d,p) level of theory. For electronically barrierless reactions, the transition state was calculated as the highest-free-energy structure in the reaction path, according to variational transition-state theory. Variational transition states were initially found using a B3LYP/6-31G++(d,p) level of theory and then refined using CBS-QB3. Rates of reaction were calculated using the Mesmer master equation code. Flow-reactor predictions using the resulting mechanisms and Chemkin Pro were compared against experimental flow-reactor yields as determined by molecular-beam mass-spectrometry. Potential pathways were evaluated by comparing reaction kinetics at the high-pressure limit against what was observed experimentally.
Eight possible pathways were considered. Only one pathway had rates fast enough to explain the experimental yields, the pathway involving metaphosphoric acid as a reactive intermediate. This mechanism begins with formation of phosphoric acid from triethylphosphate by successive eliminations of ethylene. Phosphoric acid then goes through a catalytic cycle in which it dehydrates to form metaphosphoric acid, associates with acetic acid to form acetylphosphate, and finally dissociates to yield ketene and a reformed phosphoric acid molecule. Predicted yields agreed with experimental yields within an order of magnitude without requiring any parameter fitting.
The authors gratefully acknowledge funding of this research by the Eastman Chemical Company and computational resources provided by the Extreme Science and Engineering Discovery Environment (XSEDE), which is supported by National Science Foundation grant number ACI-1548562.