shown, however, that the C1 position of this compound is
the most reactive site toward electrophiles such as formal-
dehyde.5 Thus, we chose to block reaction at the 1 and 8
positions by utilizing 1,8-dialkyl derivatives of 3 (i.e., R )
alkyl), forcing reaction at the C3 and C6 positions. Further-
more, symmetrical 1,8 substitution would avoid complica-
tions from the generation of regioisomers in the cyclized
products.
Scheme 3a
Unfortunately, initial studies of 1,8-dialkyl-2,7-dihydrox-
ynaphthalenes (3, R ) alkyl) indicated that these derivatives
were unstable to the acid and base conditions that are
typically employed for calixarene formation. We circum-
vented these issues of instability by protecting the naphthols
as methyl ethers.
Patterning our strategy after a convenient synthesis of
octamethylcalix[4]resorcinarene,6 we chose to integrate a
carbinol moiety at C3 (Scheme 2). We anticipated that
a Reagents and conditions: (a) allyl bromide, K2CO3, acetone,
reflux, 77%; (b) N,N-diethylaniline, reflux; (c) CH3I, K2CO3, THF,
reflux, 53% (two steps); (d) H2, Pd/C, EtOAc, 95%; (e) n-BuLi,
TMEDA, Et2O; DMF, 57%; (f) NaBH4, EtOH, 92%.
Scheme 2
Nevertheless, directed lithiation with n-BuLi, followed by
quenching with N,N-dimethylformamide, gave aldehyde 13
as a single regioisomer, which was reduced to carbinol 14
upon treatment with sodium borohydride.8
In the event, compound 14 was dehydrated with acid to
give calixarene-type products. A range of conditions were
screened, varying both the type and concentration of acid,
as well as reaction temperature. Mass spectrometry indicated
that each reaction produced uncyclized dimer, trimer, and
tetramer, as well as cyclic trimer and tetramer. The major
product under most conditions was cyclic trimer.
Under optimal conditions, treatment of carbinol 14 with
0.1 equiv of trifluoromethanesulfonic acid at 0 °C, followed
by warming to room temperature, gave 23% isolated yield
of a single compound, which according to mass spectrometry,
was a cyclic trimer. 1H and 13C NMR spectra did not support
a fully symmetrical structure, such as 7, as each naphthyl
ring had two aromatic singlets and two different methoxy
groups in the 1H NMR spectrum. Atropisomerism was ruled
out by variable temperature experiments. The spectra could
be explained by a structure such as 15, in which alkylation
of the presumed cationic intermediate occurred at C5, meta
to the methoxy carbon, rather than at C6 (Scheme 4).
treatment of a compound such as 5 with acid would generate
a carbocation that would react at the remaining electrophilic
carbon (C6) of a second unit of 5, eventually producing
calixarene oligomers such as 6 and 7.
Therefore, 1,8-diallyl-2,7-dihydroxynaphthalene (10) was
synthesized in two steps from 2,7-dihydroxynaphthalene (8)
through an intramolecular Claisen rearrangement7 (Scheme
3). Compound 10 was unstable even to silica and was taken
on crude to the methylation reaction giving compound 11.
The allyl groups were reduced in order to prevent rearrange-
ments under the acid-promoted calixarene formation condi-
tions. The desired carbinol moiety was incorporated at C3
by formylation, followed by reduction. Electrophilic formy-
lation of compound 12 (TiCl4, MeOCHCl2) gave low
conversion to a mixture of regioisomers (ca. 1:1) of aldehyde-
containing products. The lack of selectivity in the addition
of this electrophile to compound 12 would portend the
eventual outcome of the calixarene cyclization reaction.
Scheme 4
578
Org. Lett., Vol. 3, No. 4, 2001