Hydrides cyclic imines, stereoselectivity

Examination of the enantioselectivities in Table 7.5 indicates a striking difference in selectivity achieved in the reduction of cyclic (entries 1-8) vs. acyclic imines (entries 9-11). The former is very nearly 100% stereoselective. The simple reason for this is that the acyclic imines are mixtures of E and Z stereoisomers, which reduce to enantiomeric amines vide infra). The mechanism proposed for this reduction is shown in Scheme 7.11 [86]. The putative titanium(III) hydride catalyst is formed in situ by sequential treatment of the titanocene BINOL complex with butyllithium and phenylsilane. The latter reagent serves to stabilize the catalyst. Kinetic studies show that the reduction of cyclic imines is first order in hydrogen and first order in titanium but zero order in imine. This (and other evidence) is consistent with a fast 1,2-insertion followed by a slow hydrogenolysis (a-bond metathesis), as indicated [86]. Although P-hydride elimination of the titanium amide intermediate is possible, it appears to be slow relative to the hydrogenolysis. 

---