Journal of the American Chemical Society, Vol.130, No.34, 11317-11327, 2008
Combined experimental and computational studies on carbon-carbon reductive elimination from bis(hydrocarbyl) complexes of (PCP)Ir
The reductive elimination of carbon-carbon bonds is one of the most fundamentally and synthetically important reaction steps in organometallic chemistry, yet relatively little is understood about the factors that govern the kinetics of this reaction. C-C elimination from complexes with the common d(6) six-coordinate configuration generally proceeds via prior ligand loss, which greatly complicates any attempt to directly measure the rates of the specific elimination step. We report the synthesis of a series of five-coordinate d(6) iridium complexes, ((PCP)-P-tBu)Ir(R)(R'), where R and R' are Me, Ph, and (phenyl-substituted) vinyl and alkynyl groups. For several of these complexes (R/R' = Ph/Vi, Me/Me, Me/Vi, Me/CCPh, and Vi/CCPh, where Vi = trans-CH=CHPh) we have measured the absolute rate of C-C elimination. For R/R' = Ph/Ph, Ph/Me, and Ph/CCPh, we obtain upper limits to the elimination rate; and for R/R' = CCPh/CCPh, a lower limit. In general, the rates decrease (activation barriers increase) according to the following order: acetylide < vinyl similar to Me < Ph. Density functional theory (DFT) calculations offer significant insight into the factors behind this order, in particular the slow rates for elimination of the vinyl and, especially, phenyl complexes. The transition states are calculated to involve rotation of the aryl or vinyl group around the Ir-C bond, prior to C-C elimination, such that the group to which it couples can add to the face of the aryl or vinyl group. This rotation is severely hindered by the presence of the phosphino-t-butyl groups that lie above and below the plane of the aryl/vinyl group in the ground state. Accordingly, calculations predict dramatically different relative rates of elimination from the much less sterically hindered complexes ((PCP)-P-H)Ir(R)(R'). For example, the barrier to elimination from ((PCP)-P-H)Ir(Me)(2) is 20 kcal/mol, which is 2 kcal/mol greater than from the ((PCP)-P-tBu)Ir analogue. In contrast, the activation enthalpies calculated for vinyl-vinyl and phenyl-phenyl elimination from ((PCP)-P-H)Ir are remarkably low, only 2 and 9 kcal/mol, respectively; these values are 16 and 22 kcal/mol less than those of the corresponding ((PCP)-P-tBu)Ir complexes. Moreover, since these eliminations are very nearly thermoneutral, the barriers are calculated to be equally low for the reverse reactions [C-C oxidative addition to ((PCP)-P-H)Ir]. The absence of differences in intraligand C-C bond lengths in the transition states relative to the ground states, combined with a comparison of calculated "face-on" and "planar" transition states for C-C coupling, suggests that the critical importance of the aryl/vinyl rotation is based on geometric or steric factors rather than electronic ones. Thus there is no evidence for participation of the pi or pi* orbitals of the aryl or vinyl groups in the formation of the C-C bond, although a small pi effect cannot be rigorously excluded. Likewise, the results do not support the hypothesis that the degree of directionality of the carbon-based orbital used for bonding to iridium (sp(3) > sp(2) > sp) plays an important role in this system in determining the barrier to reductive elimination.