Origin of Regioselectivity in the Dehydrogenation of Alkanes by Pincer–Iridium Complexes: A Combined Experimental and Computational Study

PCP-pincer (κ3-2,6-C6H3(CH2PR2)2) iridium complexes have been reported to catalyze the transfer dehydrogenation of n-alkanes with high regioselectivity for the terminal position. We find that the very closely related PCOP (κ3-2,6-C6H3(CH2PR2)­(OPR2)) and POCOP (κ3-2,6-C6H3(OPR2)2) complexes, in contrast, afford no such regioselectivity. The difference is a true kinetic phenomenon, i.e., it is not a result of isomerization subsequent to the formation of free α-olefin. In addition to direct observation of the distribution of n-alkane dehydrogenation products over time, the pronounced difference in regioselectivity is confirmed through intermolecular competition studies of the reverse reaction (olefin transfer hydrogenation) and of the dehydrogenation of cycloalkane vs n-alkane. Electronic structure (DFT) calculations indicate that the rate- and selectivity-determining step for dehydrogenation by the (PCP)Ir complexes is β-H transfer. C–H activation at the primary position is much more favorable than at secondary positions, but this is not responsible for the terminal regioselectivity; indeed, the formation of α-olefin via C2–H addition and transfer of the C1–H bond is calculated to be slightly more favorable than dehydrogenation proceeding via C1–H addition. For both PCP and POCOP complexes, the formation of the α-olefin iridium dihydride complex is more facile than the formation of internal-olefin complexes. The next step in the catalytic pathway, loss of olefin, is calculated to have an activation energy that is significantly greater than the metal–ligand (thermodynamic) bond energy. In the case of POCOP complexes, the loss of olefin, rather than β-H transfer, is the rate- and selectivity-determining step. The hydrocarbon moiety in the transition state for olefin loss has the character of a fully formed olefin; this favors the formation of internal olefin. The different regioselectivity of (POCOP)Ir vs (PCP)Ir catalysts is thus attributable to the different rate-determining steps of their respective catalytic cycles; this in turn can be explained in terms of different electronic effects of O versus CH2 linker exerted through the pincer aromatic ring.