The reaction of sydnones with bromine in acetic anhydride revisited: a new route to 5-substituted-3-aryl-1,3,4-oxadiazol-2( 3H )-ones from N -aryl- N -bromocarbonylhydrazines

The reaction of 3-phenylsydnone with bromine in acetic anhydride to form 5-methyl-3-phenyl-1,3,4-oxadiazol-2( 3H )-one has been reexamined and improved. A new mechanism involving a bromocarbonylhydrazine species is proposed and its intermediacy is supported by the observation that it reacts with acetic anhydride to yield the corresponding 1,3,4-oxadiazol-2( 3H )-one. The process has been expanded to the use of acid chlorides and a novel synthesis of 5-substituted-3-aryl-1,3,4-oxadiazol-2( 3H )-ones has been developed.


Introduction
Sydnones (c.f. 1) are members of the class of compounds known as mesoionic and have been studied extensively. 1 In 1946, Kenner and Mackay prepared 4-bromo-3-phenylsydnone (2, R = H) from the reaction of 3-phenylsydnone (1, R = H) with bromine in acetic acid. 2 Later, Baker, Ollis and Poole modified the process to use acetic anhydride as solvent. 3However, when Stansfield 4 utilized the latter protocol he observed a vigorous evolution of gas at 30-40 o C and he isolated 5-methyl-3-phenyl-1,3,4-oxadiazol-2(3H)-one (3a) instead of the bromosydnone 2 (R = H) [Scheme 1, route a].Over 40 years later, Badami et al. extended the process to the synthesis of a variety of 5-methyl-3-aryl-1,3,4-oxadiazol-2(3H)-ones (3, R = various) 5 and they proposed a mechanism involving a 1,3-dipolar cycloaddition between 2 and the carbonyl group of the anhydride.They later concluded that HBr, which is formed in situ, is important for the process, but only as a catalyst for the 1,3-dipolar cycloaddition mechanism. 6These mechanistic suggestions are surprising, especially since Yeh et al. 7 showed in 1994 that treatment of 4-bromo-3-phenylsydnone (2, R = H) with HX (X = Cl, Br) cleaves the ring to form an isolable bromocarbonyl hydrazine derivative (4, R = H).Given Yeh's results, it seemed likely to us that, rather than an unprecedented 1,3-dipolar cycloaddition process, the formation of the 1,3,4-oxadiazol-2(3H)-ones 3 (from 1) instead involves the intermediacy of the corresponding bromocarbonyl hydrazine species 4 (via reaction of 2 with HBr formed in situ) and subsequent reaction of the latter with acetic anhydride (Scheme 1, route b).
The reaction of sydnones with Br 2 in Ac 2 O.

Results and Discussion
First, it was important to reproduce the reported synthesis of 3a [Scheme 1, route a], to provide an authentic sample of the product and to act as a benchmark for the process.Using the reported conditions (bromine in acetic anhydride added to 1 (R = H) at 0 o C, warming to 60 o C, then addition to water and standing overnight) afforded 3a in low yield and purity and two recrystallizations were required to afford pure product.The structure of the isolated product was confirmed as 3a from its 1 H-NMR and 13 C-NMR spectra, however, extra peaks, most noticeably a singlet at δ 4.27 in the 1 H-NMR spectrum, were present in the crude material and, after analysis by GC-MS, it was concluded that these were due to the presence of 5-bromomethyl-3-phenyl-1,3,4-oxadiazol-2(3H)-one (3b).Indeed, separation by column chromatography allowed for complete analysis of the by-product ( 1 H-NMR, 13 C-NMR, GC-MS) and its identity as 3b was confirmed by its independent synthesis from the reaction of the hydrazine salt 4 (R = H, X = Cl) with bromoacetyl chloride (vide infra, Scheme 2, R 1 = BrCH 2 ; Table 1, Entry 2).Modifications to the initial procedure demonstrated that the amount of byproduct could be reduced drastically and that 3a could be obtained with <2% impurity before recrystallization (<0.5% after recrystallization; GC/MS analysis).It is probable that 3b arises from α-bromination of the acetic anhydride to form both mono-and di-bromoacetic anhydrides, which then react competitively with the bromocarbonylhydrazine intermediate 4 (R = H).
With a quantity of the desired oxadiazolone 3a in hand, we turned now to the mechanism of the transformation shown in Scheme 1, path a.Since we conjectured that the key intermediate, N-phenyl-Nbromocarbonylhydrazine salt 4 (R = H), was formed from 4-bromo-3-phenylsydnone (2, R = H) [also formed in situ], we prepared the latter from 3-phenylsydnone (1, R = H) by bromination with Br 2 / NaHCO 3 24 and converted it into the salt 4 (R = H, X = Cl) using the method reported by Yeh et al. 7 Treatment of the salt 4 with acetic anhydride gave the expected oxadiazolone 3a in reasonable yield (Scheme 2), a result which suggests strongly that the mechanism of the overall 3-phenylsydnone (1, R = H) to 3a transformation involves a bromocarbonylhydrazine salt intermediate rather than the 1,3-dipolar cycloaddition mechanism proposed by Badami et al.Further support for this avenue is provided by Badami's observation that treatment of the 4bromosydnone 2 (R = H) in acetic anhydride with HBr also yields the oxadiazolone product 3a.These results give a high degree of certainty to our mechanistic proposal delineated in Scheme 1 (path b).
While we had been able to prepare oxadiazolinone 3a from the bromocarbonyl hydrazine salt 4 (R = H, X = Cl) as a test of our mechanistic proposal, as a practical synthetic avenue to oxadiazolinones, the method suffered from the same major deficiency inherent in the original sydnone 1 to oxadiazolinone 3a conversion (Scheme 1, path a), viz. the use of the anhydride as both reactant and solvent.Accordingly, we explored the transformation of 4 (R = H, X = Cl) into 3a using a variety of solvents and reduced amounts of acetic anhydride.From these studies, it was determined that the use of 1,2-dimethoxyethane (DME) as solvent gave the best results and the optimal protocol was with 2 eq. of acetic anhydride at 65 °C for 2 h.Extension to acid chlorides, including acetyl chloride, (Table 1, Entries 1-12) yielded the corresponding oxadiazolinones 3a-l in good yields and the overall method provides a novel approach to the latter.The present findings exhibit further the utility of 3-arylsydnones as precursors to useful heterocycles and, accordingly, given the rather efficient avenues to sydnones, 25 including direct, one-pot avenues from N-substituted glycines, 26 the process may find considerable utility.

Conclusions
The conversion of 3-phenylsydnone (1, R = H) into 5-methyl-3-phenyl-1,3,4-oxadiazol-2(3H)-one (3a) has been reinvestigated, the major impurity (3b) identified and the protocol improved to minimize the amount of the latter.A new mechanism via a bromocarbonylhydrazine salt intermediate 4 (R = H) is proposed for the procedure and is supported by the observation that 4 reacts with acetic anhydride to yield 3a.The procedure has been extended to the use of acid chlorides and a novel synthesis of 5-substituted-3-aryl-1,3,4-oxadiazol-2(3H)-ones 3 from N-aryl-N-bromocarbonylhydrazine salts 4 (X = Cl) has been developed.Most of the products (3a-j) are derived from 3-phenylsydnone (1, R = H) and, accordingly, the substituent at the 3-position in the products is consistently a phenyl group.However, as a proof of concept, two other sydnones (1, R = 4-Br and 4-Cl) were converted into the corresponding bromocarbonylhydrazines 4 (R = 4-Br and 4-Cl, X = Cl), and the latter reacted with acetic anhydride in DME to form the corresponding oxadiazol-2(3H)-ones 3l and 3m, respectively.It is planned to extend this protocol to the use of other sydnone-derived bromocarbonylhydrazine salts and acid chlorides in order to better delineate the scope and limitations.