Scheme 1
.
Mukaiyama-Michael Reactions of
Scheme 3. Reactions of Methyl
2-Diazo-3,7-dioxo-5(E)-octenoate
TBSO-Substituted Vinyldiazoacetate 1
expectation that an enolizable enedione6 would be formed
whose chemistry could lead us to cycloaddition products that
contained the diazoester functionality. As anticipated, the
direct Mukaiyama-Michael product (3) was unstable, un-
dergoing elimination under the reaction conditions to form
R,ꢀ-unsaturated ketone 4 exclusively (Scheme 2),7 but the
occurred (36% conversion to 6). Performing the reaction at
elevated temperature with silica gel or by adding acetic acid to
the system produced the same results. When 2.0 molar equiv
of 4 N aqueous HCl was used, 5 was produced in 21% isolated
yield. However, assuming that enolization of 4 is one of the
key steps for this transformation, we anticipated that the addition
of base would facilitate this reaction. With 5.0 equiv of
triethylamine, 5 was obtained in 45% isolated yield; however,
the yield of 5 was further improved when only a catalytic
amount of aqueous sodium hydroxide (0.10 mol/L NaOH, 10
mol %) was employed, and under these catalytic conditions, 5
was isolated in 83% yield.
Scheme 2. Mukaiyama-Michael Reaction with
4-Methoxy-3(E)-buten-2-one
A plausible mechanistic pathway for resorcinol formation
is presented in Scheme 4. Removal of the most acidic proton
from compound 4 produces the conjugated enolate anions
that are depicted by intermediates 7.
Isomerization of 7a in which the 5,6-positions have the
E-geometry to 7b in which the 5,6-positions have the
Z-geometry is critical to the second isomerization in which
the two ends are wrapped together (7c) through a conjugated
triene that is appropriately arranged for pericyclization.
Similar 6-π electrocyclizations involving enolate derivatives
have been reported,8 although none have involved a diazo
compound.
resulting enedione underwent an unexpected transformation
resulting in the formation of a substituted resorcinol.
Examination of this process showed a diverse chemistry that
we now report.
The diazonium ion intermediate 8 resulting from peri-
cyclization is suitably positioned to undergo loss of
dinitrogen in concert with methyl migration to form
intermediate 9 that is the tautomer of the observed
resorcinol product. The conversion of 8 to 9 is, to our
knowledge, unprecedented; also, instead of undergoing
methyl migration to the carbon bearing the diazonium ion
that would be a semipinacol rearrangement9 or, alterna-
tively, forming an epoxide10 with loss of dinitrogen, the
methyl group migrates in the reverse direction to that of
dinitrogen extrusion. The formation of the phenolate anion
that can continue proton removal from reactant 4 is
consistent with the need for only a substoichiometric
amount of base to complete the reaction; after the reaction
is initiated, the transformation is self-sustained.
Attempted chromatographic purification of 4 on silica gel
resulted in the loss of 4, but the highly substituted resorcinol
5 was isolated in low yield (13%) (Scheme 3).
Various conditions were employed to optimize formation of
this unexpected reaction product by first treating 1 and 2 with
zinc triflate and then, after concentrating the mixture but without
isolating 4, adding a suitable promoter to catalyze the formation
of the resorcinol product. Since compound 4 decomposes upon
contact with silica gel, silica gel was added to the reaction
system; however, only migration of the enone double bond
(6) Schuler, M.; Duvvuru, D.; Retailleau, P.; Betzer, J.-F.; Marinetti,
A. Org. Lett. 2009, 11, 4406.
(7) Compound 3 was produced when the zinc triflate catalyzed reaction
was performed in the presence of molecular sieves 4A. A mixture of cis
and trans isomer in about 2/1 ratio was observed by NMR.
Org. Lett., Vol. 12, No. 19, 2010
4305