Crystal aluminum enolate

The few crystal structures obtained from aluminum enolates that are less important in synthesis than their boron counterparts reveal dimeric aggregates (Scheme 3.5). This maybe illustrated by the enolates 15 [43a], obtained fromiV,Af-dimethyl methyl glycinate through transmetallation of the lithium enolate, and 16 [43b] that was prepared by direct enolization of 2,4,6-trimethylacetophenone and trimethylaluminum. Both dimers feature an AI2O2 core unit and clearly demonstrate the O-bound character of aluminum enolates. 

The structure of catalyst 428 was proposed as a result of the several experiments shown in Sch. 60 and discussed below [89]. Firstly, it was observed that treatment of ALB catalyst 394 (Sch. 51) with methyllithium produced a solution from which the hexacoordinate aluminum species 434 (M = Li) could be crystallized in 43 % yield. The same compound could also be obtained from solutions prepared from 394 and nBuLi, and the sodium enolate of 425. Solid-state X-ray analysis of this compound revealed that it has the same structiu-e as the species 417 (Sch. 56) isolated by Feringa and coworkers during the preparation of ALB with excess BINOL (Sch. 55) [86]. The tris-BINOL(tris-lithium) alimunum complex 434 is not the active catalyst in the Michael addition of phosphonate 425 to cyclohexenone because the use of this material as catalyst gave the Michael adduct 426 in 28 % yield and 57 % ee which is dramatically lower than obtained by use of catalyst 428 (Sch. 59). In addition, the use of catalyst 434 (M = Li) gave the alkene product 429 in 13 % yield, a product that was not seen with catalyst 428. Additional evidence comes from the reaction between 425 and cyclopentenone with catalyst 434 (M = Li) which gives the adduct 427 in 78 % yield and 12 % ee. 

---