The chemical instability of isoindole (l_) is well documented2 and impeded its isolation and 0 G \ :NH 2 a 0 'N 2 detailed characterization for nearly eighty years, until 1972, when it was reported by Bonnett and Brown3, and later by Bornstein, Remy and Shields. 4 This ortho-quinonoidal nitrogen heterocycle is thermally unstable, reacts with atmospheric oxygen and undergoes DielsLAlder reactions in the pyrrole ring with dienophiles such as N-phenylmaleimide to yield bicyclic aclducts (2).
0
The great reactivity of isoindole contrasts sharply with the predictions of theoretical cslculatlons, performed at various levels of sophistication, all of which indicate the molecule to be aromatic.5 In view of the historical connection between aromaticity and diminished chemical reactivity, the potential coexistence in isoindole of both aromaticity and chemical instability seemed to us to present an apparent inconsistency which invited resolution. Accordingly, we undertook a spectral investigation of isoindole whose aims were to provide evidence for or against its aromaticity, and to characterize its electronic ground state, in order thereby to clarify the source of its reactivity. On the basis of our results, we propose herein a structural representation capable of accounting for both properties. 6
Isoindole can be produced pure, in nearly quantitative yields, by flash vacuum pyrolysis.4.7