Systematic structure variation of catalysts has revealed that chemo-and regioselectivity of several phase transfer reactions can be influenced strongly by catalyst structure.Whereas typical quaternary ammonium and phosphonium ions show little selectivity control, there are two groups of catalysts with opposite steering effects:(1) Small and sterically unencumbered ammonium salts as well as some benzo-crown ethers, (2) sterically hindered and large catalyst as well as highly delocalized cations. Rather spectacular effects are found, for example, in the following reactions ((a) to (c); Scheme (a) NMe4CI and dibenzo-18-crown-6 catalyzed 97-99 % selectivity for A, AsPh4Br gives a 52 : 40 : 8 mixture of A, B, (b) "Cetrimid (hexadecyltrimethylammonium bromide) leads to a D/(E + F) ratio of 92 : 1 whekas AsPh4C1 switches reaction channels to a 1 : 91 ratio [3]. (c) O/C ratios for the first reaction step G/(H + I) vary between 0.7-1.1 (catalysts: PPhqBr, benzo-15crown-5, NMe4Br) and 21 (catalysts: Kryptofix [2.2.2], (Me2N)3P=N=P(NMe&BF4) [4]. 1).c PI.The present study addresses the question whether catalyst structure influences can also be found in the ZIE diastereoselectivity of Darzens reactions and mechanistically related cyclopropane syntheses from Michael acceptors and a-halocarbanions. In both of these reactions stereoselection occurs before ring closure. We have investigated phase transfer catalytic conversions (i) of benzaldehyde with ethyl chloroacetate to give (Z/E)-2-phenyl-3-ethoxycarbonyloxirane (l), (ii) of benzaldehyde with 2-chloro-2-phenylacetonitrile to yield the isomeric 2-cyano-2,3-diphenyloxiranes (Z), and (iii) of methacrylonitrile with 2-chloropropionitrile to (Z/E)-1,2-dicyano-l,2-dimethylcyclopropane (3) under standardized conditions. Scheme 2Reaction (i) has been the subject of intensive previous studies under non-PTC conditions. Influences of solvent, base and additional substituents were investigated [5,6]. The Makosza group introduced PTC conditions (solid K2CO3, TEBA as catalyst) and observed Z/E ratio of approximately 1 [7]. Further work by a Japanese group showed that this PTC conversion is sensitive towards many factors: concentration of catalyst, molar ratio of reactants, solvent, and temperature [8]. Under certain conditions, the Z/E ratio varied between 0.44 (dibenzo-18-crown-6) and 0.1 (tetrabutylphosphonium bromide).Literature is also available on some previous PTC experiments concerning reaction (ii). Whereas Z/E ratios of almost l was reported for the noncatalyzed process, TEBA and dibenzo-18-crownd gave a value of 8 under certain conditions [9]. Similarly, reaction (iii) has been performed before under PTC by the Polish group of Makosza and Jon'czyk. TEBA gives a Z/E ratio of 0.8 under the conditions used [lo].Our own results are given in Table 1. Because of the mentioned previous experiences, all experiments were conducted under carefully controlled circumstances. The recorded conversions and Z/E ratios are averages of 3-7 runs. Due to the heterogeneous nature of the systems, the...