Ring-closing olefin metathesis on ruthenium carbene complexes: model DFT study of stereochemistry

Ring-closing metathesis (RCM) is the key step in a recently reported synthesis of salicylihalamide and related model compounds. Experimentally, the stereochemistry of the resulting cycloolefin (cis/trans) depends strongly on the substituents that are present in the diene substrate. To gain insight into the factors that govern the observed stereochemistry, density functional theory (DFT) calculations have been carried out for a simplified dichloro(2-propylidene)(imidazole-2-ylidene)ruthenium catalyst I, as well as for the real catalyst II with two mesityl substituents on the imidazole ring. Four model substrates are considered, which are closely related to the systems studied experimentally, and in each case, two pathways A and B are possible since the RCM reaction can be initiated by coordination of either of the two diene double bonds to the metal center. The first metathesis yields a carbene intermediate, which can then undergo a second metathesis by ring closure, metallacycle formation, and metallacycle cleavage to give the final cycloolefin complex. According to the DFT calculations, the stereochemistry is always determined in the second metathesis reaction, but the rate-determining step may be different for different catalysts, substrates, and pathways. The ancillary N-heterocyclic carbene ligand lies in the Ru-Cl-Cl plane in the simplified catalyst I, but is perpendicular to it in the real catalyst II, and this affects the relative energies of the relevant intermediates and transition states. Likewise, the introduction of methyl substituents in the diene substrates influences these relative energies appreciably. Good agreement with the experimentally observed stereochemistry is only found when using the real catalyst II and the largest model substrates in the DFT calculations.