extended to studying the synthetic potential of these highly
functionalized bicyclic adducts. From a synthetic perspective,
these are interesting derivatives that possess a seven-
membered carbocyclic ring that is functionalized at every
position. Herein, we wish to report our studies on function-
alization of some of these derivatives through reactions
promoted by the use of silver ion.
Scheme 3. Silver-Promoted Arylation
Although there have been few studies on chemical
transformations of the halogenated cycloadducts, Tobey and
Law were able to show that the tetrachloro adducts could
be hydrolyzed to the 1,3-diketone in fair yield by treatment
with strong acid.7 In an attempt to find more mild hydrolytic
conditions, silver-promoted8 ionization of the allylic halides
was examined (Scheme 2). Treatment of the tetrabromo
complex mixture of products.10 One conceivable complica-
tion was that the initial product of the arylation contains a
more reactive allylic bromide than the starting compound.
This halide would be susceptible to further ionization and
subsequent reaction of the cation. We were also concerned
that the reaction became highly acidic, which could promote
alternative reactions. In an attempt to provide an acid
scavenger, an equivalent of silver oxide was added to the
reaction mixture. Under these conditions, a monoarylated
product was isolated in very high yield, and its structure was
determined to be the tertiary carbinol 7. The compound was
formed with a very high preference for the para isomer with
only minor amounts of the ortho product detected in the crude
NMR. A possible pathway to arrive at the adduct involves
initial formation of the allylic cation A upon treatment with
silver (Scheme 4).
Scheme 2. Synthesis of Ketones and Ketals
adduct of furan 4a with an aqueous solution of silver nitrate
lead to rapid hydrolysis of the geminal dibromide group to
produce the ketone 5a in very good yield. This type of
hydrolysis worked equally well on the adduct of 2,5-
dimethylfuran 4b and cylopentadiene 4d to give the highly
versatile dibromo enone moiety.
Scheme 4. Possible Mechanism for Arylation
It was also possible to employ a diol in the reaction to
directly produce the protected form of the dibromo enone
(Scheme 2). To avoid competitive hydrolysis, anhydrous
silver tetrafluoroborate in dichloromethane was required.
Under these conditions, the corresponding dioxolanes 6a-c
were produced in high yields. The use of Lewis acid to
promote conversion of a carbon-bromine bond to a carbon-
oxygen bond prompted us to examine if a similar process
could be used to effect the formation of carbon-carbon
bonds.
Initial attempts to promote arylation of the furan-derived
bromide 4a were unsuccessful with a variety of Lewis acid
catalysts. Treatment of an anisole solution of 4a with highly
reactive catalysts such as aluminum chloride or titanium
tetrachloride led to extensive decomposition of the materials,
while milder Lewis acids such as boron trifluoride etherate
failed to promote any reaction. The facility of the silver-
promoted hydrolysis of the bromides suggested that silver
salts9 may also be used to promote the arylation reaction
(Scheme 3).
Attack of anisole on the symmetrical cation should initially
give B11-12 and an equivalent of tetrafluoroboric acid that is
neutralized by silver oxide. Bromide B is poised for a second
ionization to give the highly stabilized cation C. Water,
generated from silver oxide, can intercept this nonsymmetri-
cal allylic cation at position a to give exo-alcohol D or at b
to give E after collapse of the initially formed bromohydrin.
Although E would appear to be the most stable adduct, the
addition of water was completely selective for formation of
the benzylic alcohol. This preference may reflect a greater
localization of the cationic charge at this position. With an
efficient set of reaction conditions determined, a study of
the arylation reaction was undertaken with several different
aromatic nucleophiles (Table 1).
Attempts to effect arylation with silver tetrafluroborate and
a large excess of anisole were sluggish and produced a
(7) Law, D. C. F.; Tobey, S. W. J. Am. Chem. Soc. 1968, 90, 2376.
(8) Warner, P. M.; Lu, S.-L.; Myers, E.; DeHaven, P. W.; Jacobsen, R.
A. J. Am. Chem. Soc. 1977, 99, 5102.
(9) Olah, G. A.; Kuhn, S. J.; Flood, S. H.; Hardie, B. A. Chem. Ind.
1957, 50.
(10) GC-MS and crude NMR data indicated that the products of the
reaction contained multiple aryl rings and a cleaved oxo-bridge.
(11) Compound B has not been observed in any of the reactions.
(12) AM1 calculations indicate that the exo and endo adducts from anisole
only differ by 0.88 kcal/mol in favor of the exo arene.
1998
Org. Lett., Vol. 4, No. 12, 2002