A large number of naturally occurring compounds of the polysubstituted 2-hydroxyphenoxybenzoic acid lactone (depsidone) type were isolated by Asahina ( 2 ) from lichens. Noyce and Weldon and others ( 3 ) succeeded in the synthesis of the parent lactone, depsidone (Ia) and several simple homologs. Accordingly we became interested in the possibility of synthesizing a nitrogen analog, which we have termed depsazidonc (Ib), and which does not occur naturally either as the parent o r a s derivatives. 0 U R l a . x = o I l a . R = H I b , X = N H I l b , R a C H 3We therefore wish to r e p o r t the synthesis of (Ib), and a novel rearrangement which it may be caused to undergo. Since Ia was obtained by Noyce and Weldon (3) from 2-(2-hydroxyphenoxy)benzoic acid by lactonization, we followed a s i m i l a r route to Ib.The intermediate, N-2 -h ydroxyphenylanthranil ic acid (4) (IIa), was prepared from o-bromobenzoic acid and o-aminophenol by an Ullmann type condensation c a rried out in dimethylformamide with cuprous chloride catalyst and potassium carbonate as the base. Since only 38% yield was obtained in this procedure, we attempted to obtain a higher yield of IIa by blocking the phenolic hydroxyl group. Thus N-2-methoxyphenylanthranilic acid (I%) was prepared in 79.1% yield from obromobenzoic acid and o -anisidine. However, we w e r e unable to demethylate IIb (5a). Furthermore, o-benzyloxyaniline (5b) which on condensation with o-bromobenzoic acid and reductive debenzylation was expected to give IIa, failed to condense with o-bromobenzoic acid.N-(2-Hydroxyphenyl)anthranilic acid (IIa) was then lactonized to Ib by standing in ethereal thionyl chloride-pyridine for three days, followed by neutral-ization and work-up.A pure product was obtained only after passage through a column of silica gel, and recrystallization from benzene. When stored in a d r y a m b e r bottle under nitrogen the compound gradually decomposed as indicated by the lowering of the melting point.When this lactone was treated with sodium hydride in xylene, with o r without an equivalent of DMF, a new substance was obtained the infrared spectrum of which was different from Ib. Whereas Ib showed strong absorption a t 5 . 8 microns, the product now absorbed at 6 . 1 microns, which indicated the p r e sence of a t e r t i a r y amide. The melting point changed from 160' to 275". The new material was resistant to ordinary hydrolysis, but treatment with sodium hydroxide in propylene glycol gave recovered unreacted material and some IIa. The n m r spectrum suggested that the carbonyl group was not in the s a m e plane as the benzene rings. In b, a peak in the n m r spectrum at 482 c p s was assigned to the hydrogen ortko to the carbonyl group. In 111, the ortho proton absorption was at higher field, under the other aromatic protons. A molecular weight determination by m a s s spectrography gave a value of 422, twice that of Ib, (211). Solvents w e r e tightly bound and required drying at 200" at 5 microns for removal. F r o m these data i t was concluded t...
