Computational mechanistic study of PMe3 and N-heterocyclic carbene catalyzed intramolecular Morita-Baylis-Hillman-like cycloalkylations: the origins of the different reactivity.

It has been experimentally reported that PMe(3) catalyzes the cycloalkylation of the pendant halogenated α,β-unsaturated ketones (e.g., 1) at the C(α) position, whereas NHC catalyzes the cycloalkylation of the pendant halogenated α,β-unsaturated esters (e.g., 2) at the C(β) position. A DFT study has been performed to understand the detailed reaction mechanisms and the causes for the different ring closure positions in the two cycloalkylations promoted by the similar catalysts. Both reactions undergo three stages that include the nucleophilic addition of the catalysts (PMe(3) or NHC) to the substrates (1 or 2) and subsequent ring closure via S(N)2 mechanism, the H-elimination by the bases (KOH or K(3)PO(4)), and the catalyst release giving the final products. The first stage in the NHC-catalyzed reaction involves a favorable H-transfer step leading to the umpolung Michael acceptor, which is more stable than the nucleophilic addition complex. The H-transfer turns off the possibility to close ring at C(α) position for the reaction. In contrast, the similar H-transfer in the PMe(3)-catalyzed reaction is thermodynamically unfavorable. According to the electronic structure of the addition complex of PMe(3) to substrate (1), we rationalize why the experimentally isolated intermediate in the PMe(3)-catalyzed reaction is the trans-cyclic ketophosphonium salt, rather than its cis isomer that favors electrostatic attraction. The differences and the similarities among the two reactions and the traditional MBH reaction are discussed.