Conformation and Cope Rearrangement of sym-Oxepin Oxides

The synthesis of a homologous series of transoid, bridged sym-oxepin oxides (loa-c) is described. The lower homologues, IOa,b, do not undergo facile Cope rearrangement to the sym-oxepin oxides 15a,b. The generation of transoid 1Oc led rapidly to the production of the Cope rearrangement product 15c. The differing reactivity in the series loa-c is attributed to the inability of lOa,b to interconvert with their cisoid isomers, 14a,b. The production of 15c is thought to occur via ring inversion of transoid 1Oc to cisoid 14c, followed by rapid Cope rearrangement (14c + 15c). Under forcing conditions IOa,b undergo epoxide opening and a subsequent rearrangement. I n an elegant scheme Neuss and co-workers in 1968 postulated' the intermediacy of an oxepin oxide (1, Scheme I) during the fungal biogenesis of the aranotins (e.g., acetylaranotin, 2). Thus, it was suggested, the stereochemistry of the dihydrooxepin moiety of the aranotins is established by intramolecular displacement a t carbon with Walden inversion in enzyme-bound epoxide 1. The first syntheses of a n oxepin oxide were communicated by Klein and Grimme,2a and by in 1974-197s. Subsequently, we detailed3 our conversion of benzene oxide oxepin to sym-oxepin oxide (3, R = H, Scheme [ I ) , and studied the conformation and Cope rearrangement of 3 (R = H ) by ' H N M R ~ p e c t r o s c o p y . ~ Other studies revealed the Cope rearrangement of a methylated derivatives and helped define the scope6 of our synthetic approach to oxepin oxides. Further, we reported the synthesis of the bridgehead diene loa7 (Scheme 1V) and characterized a derivative of 10a by X-ray crystal The possible intermediacy of an enzyme-bound oxepin oxide in biogenesis (Scheme I ) raises an intriguing question of stereochemistry. A priori one must consider two stereochemical outcomes for the Cope rearrangement of a chiral oxepin oxide (4, Scheme 111). In principle, 4 could interconvert, via Cope rearrangement, with its diastereomer 5 (5 # 4) or with its rotamer 6 (6 = 4). The interconversion 4 F= 5 would proceed via a transition state resembling cisoid conformation 3a (Scheme 11); the degenerate rearrangement 4 F= 6 would proceed via a transoid transition state9 (cf. 3b, Scheme 11). Thus, the stereochemical integrity of an intermediate, chiral oxepin oxide would depend on the rate and the geometrical requirements of the Cope rearrangement. Herein we report that oxepin oxides locked in transoid conformations d o not Scheme I