Tehahedron Leum, Vo1.31. No.17, pp 24532456.1990 Priited in Great Britain oo4o4039190 $3.00 + .oo Pergamon Rcss plc STEREOSELECTIVE ALDOL REACTIONS OF y-THIOBUTYROLACTONE: THE BENZALDEHYDE ANOMALY. Cesare Gennari*. Ambrogio Oliva, Francesco Molinari, Umberto Piarulli Dipartimento di Chimica Organica e Industriale, Universita di Milano, Centro CNR Sost.Org.Nat., via Venezian 2 1,20133 Milano, Italy Abstract- Different protocols (lithium etwlate reactions, fluoride catalyzed and Lewis acid mediated silyl ketene acetal reactions) were studied to achieve stereoselectivity in the aldol reactions of y-thiobutyrolactone: in all cases benzaldehyde showed a striking peculiarity compared to aliphatic aldehydes. It is generally known and frequently cited that lithium Z enolates are more stereoselective (depending on the size of R1) than E enolates.’ zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 0 HO RCHO Lie R Y R b Rl/$/ syn Z R$ = Ryt; Recent reports on the stereoselectivity of the aldol reactions of lithium enolates derived from cyclic ketones with aldehydes2s3 have questioned previous data commonly accepted: it has been shown that the low anti-syn ratio obtained in the condensation of cyclohexanone lithium enolate with benzaldehyde (52:48 at -72°C)4b is due to equilibration, and that the kinetic ratio is at least 51. 2a3b Other aldehydes are highly anti selective (7-100:1).28 Here we report that the lithium enolate derived from y-thiobutyrolactone reacts with various aldehydes with high anti selectivity, except for benzaldehyde which gives a ca. 1:l ratio (Table I). The observed ratio (5644, entry 1) is close to the real kinetic ratio, which is estimated to be ca. 50:50. This was proved by the following experiments: (a) pure anti adduct 1 (R=Ph) gave only traces (l-2%) of syn adduct 2 (R=Ph) when treated with LDA under the same reaction conditions (-78’C, 3 min) (b) pure syn adduct 2 gave small amounts of anti 1 (10%) when treated with LDA under the same reaction conditions (c) equilibration (-20°C, 3h) favors the anti vs. the syn aldolate (entry 2, 66:34). If we believe that the results shown in Table I are nicely accomodated by the Zimmermann-Traxler chair transition statet I, it is difficult to understand why the kinetic ratio, which seems to follow roughly the steric demand of R, falls with benzaldehyde. The only reasonable explanation is that the competing boat transition state 1 II is stabilized when R=Ph. This behavior is not unprecedented: cyclohexanone derived enolborates are syn selective with aromatic aldehydes and anti selective with isobutyraldehyde.5 The results given above demonstrate once more that while Z enolates lead reliably to syn aldols, the stereochemical course of lithium E enolates remains unpredictable. Even minor changes in the enolate or in the 2453