An Expedient Synthesis of β-Phenyl Substituted Baylis-Hillman and Aza-Baylis-Hillman Adducts

Abstract

During the last two decades notable improvements in the Baylis-Hillman chemistry have been achieved in view of its reaction rate and synthetic applications of Baylis-Hillman adducts. However, synthesis of β-branched Baylis-Hillman adducts has still remained as a difficult task. Synthesis of these compounds has been carried out either via the vinylalumination of activated carbonyl compounds or SmI2mediated reaction of α-halo-α,β-unsaturated esters with carbonyls. However, these methods suffer from the use of expensive/moisture-sensitive reagents and α,β-acetylenic esters as starting materials which are not easily accessible. For the synthesis of poly-substituted benzenes and pyridines we required β-phenyl Baylis-Hillman adducts such as 3a. Thus, we examined the synthesis of β-phenyl Baylis-Hillman adduct by following the successive FriedelCrafts reaction of Baylis-Hillman adduct 1a to 2a, bromination at the benzylic position of 2a with NBS (N-bromosuccinimide), and the final substitution reaction with water as a nucleophile, as depicted in Scheme 1. The starting material 2a (E) was prepared according to the reported method by the Friedel-Crafts reaction of 1a and benzene in the presence of H2SO4 in moderate yield (68%). Trace amounts of the corresponding Z-isomer was removed during the column separation stage. Bromination of 2a with NBS in CCl4 in the presence of AIBN produced the corresponding allylic bromides (I) and (II) which turned out too unstable to be isolated. The bromide (II) was generated via the bromination after allylic rearrangement of the initially generated allylic radical (vide infra, Scheme 2). During the bromination reaction we observed the formation of trace amounts of 3a, which might be produced by the substitution reaction of the intermediate bromides with trace amounts of water in the reaction mixture. Thus, we decided to prepare 3a without isolation of the bromide intermediates. The actual experiment was carried out as follows: bromination of 2a (NBS, CCl4, AIBN, reflux, 1 h), filtration, concentration, and followed by the reaction in aqueous DMSO (80 C, 1 h). By following the procedure we obtained 3a-Z (61%) and 3a-E (21%). The stereochemistry of 3a could be assigned based on the chemical shift of vinyl proton by comparison with the reported data. The vinyl proton of 3a-Z appeared at δ = 6.93 ppm, while that of the Eform at δ = 7.97 ppm. As depicted in Scheme 2, both isomers 3a-Z and 3a-E can be formed by following different pathways due to allylic rearrangement in the bromination stage and the competition between SN2 and SN2' pathways in the substitution reaction. Encouraged by the results we carried out the synthesis of some analogous derivatives 3b-f and the results are summarized in Table 1. Irrespective of the electron-withdrawing groups (-COOEt, -COMe, -CN) we obtained desired products 3b-d in moderate yields (53-68%, entries 2-4). However, we could not isolate the minor components (3c-E