Dinuclear Intermediates - Direct Observation

In the approaches toward a direct asymmetric Mannich reaction by enolate formation with the metal of the catalyst also, the well-proved systems of the analogous aldol reactions were widely applied. Here, it is referred to some of these protocols wherein a metal enolate is involved, as least as assumed and plausible intermediate [183]. Shibasaki and coworkers used a dinuclear zinc complex derived from linked BINOL ligand 371 for catalyst in direct Mannich reactions of a-hydroxy ketones 370 with Af-diphenylphosphinoyl imines 369 to give ti-configured a-hydroxy-P-amino ketones 372 in high yield, diastereoselectivity, and enantioselectivity (Scheme 5.97) [184]. The authors postulate the metal to form a chelated zinc enolate by double deprotonation of the a-hydroxy ketone. This enolate approaches with its Si-face to the Si-face of the imine, as illustrated by the transition state model 373, in agreement with the observed stereochemical outcome. It is remarkable that opposite simple diastereoselectivity arises from the Mannich reaction (anti-selective) and the previously reported syn-selective aldol reaction [185], although the zinc enolates... 

An example of a kinetic study of the mercury(n)-catalysed aquation of a dinuclear chloro-complex of cobalt(in) is provided by that of the newly prepared and characterized dicobalt cations (19) with 2< <6. The kinetic pattern is that of a two-stage reaction, with the second step only slightly slower than the first. The reactivity of these complexes increases steadily as the polymethylene bridge lengthens the observed first-order rate constant for removal of the first chloride is 3.3 x 10 s when = 2 and 51 x 10 s when n = 6, at 25 °C and a mercury(n) concentration of 0.16 mol 1 . The intermediacy of chloride-bridged (Co—Cl— Hg) species is assumed, though there is no direct evidence for such intermediates in these reactions. 

---