Oxazolone-thiazolidine structure

Prior to the complete structural elucidation of penicillin some effort was directed towards the preparation of the erroneous oxazolone-thiazolidine structure (48) using (52) and penicillamine (53). Surprisingly trace amounts of bioactive penicillins corresponding to the fused P-lactam structure (49) were obtained 40). The first cyclisation of a penicilloate derivative was described by Sheehan 41) using the acid chloride of the phthalimido derivative (54) to give (55). 

The synthesis represented a triumph in that this important therapeutic agent had been synthesized, yet because of the obscurity of the reaction mechanism, it could not be used as conclusive proof of the structure. Presumably a rearrangement from the intended oxazolone-thiazolidine had occurred. 

In October 1943 I suggested to Chain that penicillin had a p-lactam structure, because 1 could find no basic group in it by potentiometric titrations and because in some of its properties it seemed to resemble an N-acylthiazolidine more closely than a thiazolidine itself. Chain was immediately convinced and the structure was also accepted by Wilson Baker. However, Sir Robert Robinson had already proposed a thiazolidine oxazolone structure (which we thought would be less stable than penicillin) and would hear of no other. He was not alone in his view. One eminent organic chemist in Oxford threatened to give up chemistry if the p-lactam structure was correct. 

Reports from the United States for November and December 1943, which reached London only in February 1944, indicated that the American work on the chemistry of penicillin G had closely parallelled that in Britain on penicillin F and that the Merck group favored the thiazolidine oxazolone structure. But in 1944 the p-lactam structure received strong support from R. B. Woodward at Harvard, who later gave a... 

---