DFT and experimental exploration of the mechanism of InCl3-catalyzed type II cycloisomerization of 1,6-enynes: identifying InCl2(+) as the catalytic species and answering why nonconjugated dienes are generated

InCl(3) and other In(III) species have been widely applied as catalysts in many reactions. However, what are the real catalytic species of these reactions? Through DFT calculations and experimental investigation of the mechanism and regioselectivity of InCl(3)-catalyzed cycloisomerization reactions of 1,6-enynes (here all discussed 1,6-enynes are ene-internal-alkyne molecules), we propose that the catalytic species of this reaction is the in situ generated InCl(2)(+). Further electrospray ionization high-resolution mass spectroscopy (ESI-HRMS) supported the existence of InCl(2)(+) in acetonitrile solution. This finding of InCl(2)(+) as the catalytic species suggests that other reactions catalyzed by In(III) species could also have cationic In(III) species as the real catalysts. DFT calculations revealed that the catalytic cycle of the cycloisomerization of 1,6-enynes catalyzed by InCl(3) starts from InCl(2)(+) coordination to the alkyne of the substrate, generating a vinyl cation. Then nonclassical cyclopropanation of the vinyl cation to the alkene part of the substrate gives a homoallylic cation, which undergoes a novel homoallylic cation rearrangement involving a [1,3]-carbon shift to give the more stable homoallylic cation 15. Finally InCl(2)(+) cation coordination assisted nonconjugated [1,2]-hydride shifts deliver the final nonconjugated diene products. The preference of generating nonconjugated dienes instead of conjugated dienes in the cycloisomerization reaction is mainly due to two reasons: coordination of the InCl(2)(+) to the alkene part in [1,2]-H shift transition states disfavors the conjugated [1,2]-H shifts that generate cations adjacent to the positively charged alkene, and coordination of InCl(2)(+) to the nonconjugated diene product is stronger than coordination to the conjugated diene, making nonconjugated [1,2]-H shift transition states lower in energy than conjugated [1,2]-H shift transition states, on the basis of the Hammond postulate. DFT calculations predicted that the conjugated [1,2]-H shifts could become favored if the electron-donating methyl substituent in the alkyne moiety of the 1,6-enyne is replaced by a H atom. This prediction of producing a conjugated diene has been verified experimentally. Rationalization about why type II rather than type I products were obtained using InCl(3) as the catalyst in the cycloisomerization of 1,6-enynes has also been investigated computationally.