A comprehensive view of M–H addition across the RCCH bond: frustration culminating in ultimate union

Abstract

The reaction of MHCl(CO)L2 (L = PiPr3; M = Ru or Os) with more than a dozen terminal alkynes RCCH has been studied at variable temperatures and for a variety of R groups representing a wide range of steric and electronic effects. This sometimes reveals (for the slower osmium examples) formation of an η2-alkyne adduct, then the vinylidene OsHCl(CCHR)(CO)L2 and finally the η1-vinyl complex OsCl(CHCHR)(CO)L2. The rate of formation of the vinyl complex decreases with R according to the series primary > tertiary > secondary and electron-withdrawing > electron-donating. Deuterium labeling of OsHCl(CO)L2 at either Os or the alkyne sp carbons shows that isotope exchange between these two sites can be competitive with vinylidene and vinyl product formation, and thus can confuse some attempts to trace the fate of the hydride. When this complication is absent, conventional syn addition of Os–D to HCCR is established, to give Os(E-CHCDR). The rate of conversion to the vinyl product is not suppressed by added free PiPr3 . Taken together, these results are consistent with a mechanism of vinyl complex formation involving neither the adduct with H trans to RCCH, nor the vinylidene, but rather with direct alkyne attack cis to the hydride, which is also consistent with the considerable steric influence on the rate of vinyl formation. DFT (B3PW91) calculations show that the vinyl complex is the thermodynamically most stable product and thus is always the final observed product. The calculations also show that the “direct” addition of the alkyne occurs ia approach of the alkyne cis to M–H inside the H–M–Cl quadrant. This direct route is in fact calculated to be a multistep process with an alkyne intermediate that is not in a deep well and thus cannot be observed experimentally. Calculations also agree with the fact that the vinylidene and the vinyl complexes are obtained through two independent routes.