A practical and robust process for the Darzens condensation of substituted benzaldehydes and subsequent a,b-epoxy rearrangement is reported. The process developed is both amenable to large scale and parallel synthesis. While electron-poor benzaldehydes gave mixtures of aryl ketones and 2-substituted aryl ketones in mediocre to low yields, electron-rich benzaldehydes were found to react in high yields with complete regioselectivity to form 2-substituted aryl ketones.In the course of one of our development program, we became interested in the well-documented Darzens condensation. 2 Since the seminal work of Erlenmeyer and Darzens, the methodology has been extensively used for aromatic aldehydes as well as for aromatic and aliphatic ketones. 3 During the reaction, a carbonyl is reacted with an a-halo ester in the presence of a base to give an epoxide. Of particular interest for us was the opening of the latter and decarboxylation upon thermolysis, to afford the corresponding one-carbon homologue of the starting aldehyde or ketone. 3,4 Development of a practical and robust metal-free one-carbon homologation process for access to ketone 1 from substituted and readily available benzaldehydes 2 (Scheme 1) was our initial objective. The stereochemistry of the Darzens condensation was of no relevance in this process.
Scheme 1 Retrosynthetic analysisA series of potent candidates containing a sulfone moiety was first brought to our attention (R 2 = CH 3 SO 2 in Scheme 1). Under classical conditions, we found that the desired product 5a did not form from the benzaldehyde 3a. Interestingly, the 4-methylsulfanyl-substituted benzaldehyde 3b reacted smoothly under identical conditions and afforded the desired ketone 5b in 73% isolated yield at the first attempt (Scheme 2). Careful investigation revealed that the Darzens condensation was not proceeding at all in the case of the sulfone 3a. A thorough literature search revealed that only sparse results suggesting an electronic bias in the Darzens reaction of aromatic aldehydes were reported. 5 Our project was a perfect opportunity to probe the substrate scope of the reaction while streamlining our process. Particular attention was paid to the practicality and robustness of the overall process to illustrate the importance of the methodology in process development.In the case of aldehyde 3b, the reaction was found to proceed well with the inexpensive (±)-2-chloropropionic acid methyl ester 4 and sodium methoxide at 0°C in a toluene-R 3