Mechanism of Asymmetric Hydroformylation Reaction

As we have seen in Chapter 5, the mechanistic details of the hydroformylation reaction with rhodium triphenylphosphine complexes are well established. These mechanistic considerations may be modified and extrapolated to the chiral hydroformylation system. One important point to bear in mind is that bi-dentate rather than monodentate ligands are involved in the chiral hydroformylation system. 

A hypothetical catalytic cycle for asymmetric hydroformylation reaction is shown in Fig. 9.13. The precatalyst Rh(acac)(P-P) reacts with H2 and CO to give the square planar catalytic intermediate 9.47. Alkene addition to 9.47 can lead to the formation of 9.48, 9.49, and 9.50. The steric requirements of the chelating ligand would have to be such that the formation of 9.50 is avoided. This is because alkene insertion into the Rh-H bond in this case would lead to the formation of the linear rather than the branched alkyl. Both 9.48 and 9.49, which differ in the coordination positions of the phosphorus atoms, can give 9.51, which has the desired branched alkyl ligand. 

Two alternative postulates for enantioselection may be proposed. In the first all four diastereomers are formed in varying amounts, and the relative amounts are determined by their respective thermodynamic stabilities. Assuming approximately equal amounts of each diastereomer to be present, if one of them has a transition state that is about 2.5 kcal lower in energy than the transition states of the others, more than 90% enantioselection for 9.51 would result. The proposed mechanism for enantioselection is thus similar to that of asymmetric hydrogenation (see Section 9.3.1). In the second postulate only one such diastereomer is produced that is, the thermodynamic stability of one of the diastereomers is higher than that of the others. The stable diastereomer owing to its steric and electronic characteristics is converted to 9.51 in an enantioselec-tive manner. 

Either of the mechanisms postulated above is far from being well established. With 9.46 as a ligand, in the absence of alkene but in the presence of H2 and CO, a species such as 9.52 has been identified by NMR. This complex is sufficiently stable. No fluxional behavior is observed up to a temperature of 60°C. 

Obviously the preferred coordination positions are the phosphite in the axial position and the phosphine in the equatorial position. This, however, is not the case with 9.45. Here NMR studies indicate the formation of a relatively stable isomer, where both the phosphorus atoms are in the equatorial positions as in 9.48. The evidence gathered so far seems to indicate that for both 9.45 and 9.46 only one stable diastereomer is formed. As proposed in the second postulate, this diastereomer probably determines the stereochemical course of the subsequent catalytic steps. 

---