1,2 shift of one ene group (in the diene) or the aryl group (in the allyl-aromatic analog), followed by bond formation between the lateral carbons of the non-migrating moiety:[1][2]
Discovery
This rearrangement was originally encountered in the
semibullvalene.[3] Once the mechanism was recognized as general by Howard Zimmerman
in 1967, it was clear that the structural requirement was two π groups attached to an sp3-hybridized carbon, and then a variety of further examples was obtained.
Notable examples
One example was the photolysis of Mariano's compound, 3,3‑dimethyl-1,1,5,5‑tetraphenyl-1,4‑pentadiene. In this
symmetric diene, the active π bonds are conjugated to arenes, which does not inhibit the reaction.[4][5][6]
Another was the asymmetric Pratt diene. Pratt's diene demonstrates that the reaction preferentially cyclopropanates aryl substituents, because the reaction pathway preserves the
The barrelene rearrangement is more complex than the Mariano and Pratt examples since there are two sp3-hybridized carbons. Each bridgehead carbon has three (ethylenic) π bonds, and any two can undergo the di‑π-methane rearrangement. Moreover, unlike the acyclic Mariano and Pratt dienes, the barrelene reaction requires a triplet excited state. Thus acetone is used in the barrelene reaction; acetone captures the light and then delivers triplet excitation to the barrelene reactant. In the final step of the rearrangement there is a spin flip, to provide paired electrons and a new σ bond.
As excited-state probe
The dependence of the di-π-methane rearrangement on the
cis-trans interconversion of diene double bonds (i.e. free rotation). In acyclic dienes, this free rotation leads to diradical reconnection, short-circuiting
the di-π-methane process. Singlet excited states do not rotate and may thus undergo the di-π-methane mechanism. For cyclic dienes, as in the barrelene example, the ring structure can prevent free-rotatory dissipation, and may in fact require bond rotation to complete the rearrangement.