Communications to the Editor
J. Am. Chem. Soc., Vol. 122, No. 27, 2000 6503
ticity and thereby affords 7 or 6 starting from 3 or 4 respectively
(entries 1-2, Table 1).
Scheme 1a
It is noteworthy that the o-OBOC substituent permits the
cascade to be stopped at intermediate A by introducing the NaBH4
at low temperatures (0 °C) for short reaction times. Treatment of
the corresponding alcohol A in a separate pot with an excess of
the appropriate Grignard reagent reinitiates the cascade and
produces the anticipated ring alkylated adducts in good yields
[Table 1, entries 3, 14].
Next, the reactions of 3 and 4 with organolithium reagents were
examined. It was found that in the case of aldehyde 4, addition
of an organolithium at -78 °C followed by the subsequent
addition of a hydride slowly produced the corresponding alkylated
materials 7-8 in 24 h [Table 1, entries 4, 5]. However, in the
case of entry 6, addition of methyllithium (1.05 equiv, -78 °C)
to ketone 3 did not lead to the anticipated product 10, instead
styrene 9 formed quite rapidly even at low temperature.13
Phenols ortho-substituted with R-branched chains (cf. 10-13)
were obtained either by adding 2 equiv of a Grignard reagent
initially or by adding the first equivalent at low temperature,
observing the disappearance of the starting material by TLC and
then adding the second Grignard reagent [Table 1, entries 7-8].
Alternatively, R-branched phenols were prepared by adding a
lithium species to the aldehyde 4, warming to 0 °C, followed by
addition of the appropriate organomagnesium reagent [Table 1,
entries 9, 10]. However, in the case described by entry 9, 12% of
7 was also observed. Indeed, in cases where the second incoming
nucleophile experienced significant nonbonded interactions, re-
duction of the quinone methide D often predominated. In entry
11 for example, only a small amount (<25%) of the bis-isopropyl
adduct is observed, instead 14 is formed as the major product by
reduction of D with i-PrMgCl.
If the intermediate that emerges from the addition of an
organomagnesium reagent is to be subsequently reduced in 1,4-
fashion with NaBH4, then a two-pot procedure is required. The
cascade can be stopped by protonation of the magnesium alkoxide
at low temperature after a short reaction time. After separation
and drying, treatment of the intermediate corresponding to A in
a second pot with an excess of NaBH4 reinitiates the cascade
and produces the desired adducts [Table 1, entries 12, 13]. Both
8 and 15 were constructed in this manner; however, in the case
of 15, 20% of the product mixture was the regioisomer that had
resulted from 1,6-hydride addition. A simple solution to this
problem is to reverse the process. Thus, if 4 is first reduced with
NaBH4 and the cascade halted, then treatment of intermediate A
with an excess of the vinyl Grignard reagent reinitiates the cascade
and affords 15 cleanly [Table 1, entry 14].
a (a) 0 °C, 2.0 equiv of NaBH4 1 h. 88%. (b) -78 °C, MeLi (1.05
equiv) 30 min, AcCl (3.0 equiv) 75%. (c) -78 °C, t-BuMgBr (1.05 equiv),
EVE (10 equiv) 18 (56%). (d) -78 °C, MeMgBr (1.05 equiv), EVE, 18
(77%). (e) -78 °C, MeMgBr (4.0 equiv), EVE, 18 (17%) + 17 (77%).
4 could arise via intermediate C [M ) Na] or [M ) BRx], while
the styrene 9 in entry 6 could arise from several other pathways.13
To further illuminate this issue, ketone 3 was submitted to NaBH4.
Quenching after a short interval led to alcohol 16 exclusively.
Similarly treatment of the aldehyde 4 at -78 °C with MeLi (1.0
equiv) followed by a low-temperature addition of acetyl chloride
yielded 17. Treatment of either 16 or 17 at -78 °C with a
Grignard, followed by the addition of ethyl vinyl ether (EVE)
led to 18, (>20:1/endo:exo). Compound 18 was also constructed
in a single pot by treatment of 4 at -78 °C with MeMgBr (1.05
equiv) followed after a short interval by the addition of EVE (10
equiv). However, higher yields were obtained if EVE was used
as the solvent for the reaction. Similar cycloaddition reactions
could not be initiated from 4, 16, or 17 with organolithium
reagents even at elevated temperatures. Instead, all of these
reactions yielded complex mixtures of unidentified products. If,
however, MgBr2‚Et2O were added soon after addition of the
organolithium reagent, 18 formed smoothly. Thus, these experi-
ments seem to indicate that the metal or its corresponding salt
plays some role in the conversion of C f D.
In summary, salient features of this new procedure include the
use of o-OBOC substituted aryl ketones and aldehydes in
combination with various organomagnesium reagents to generate
o-quinone methides that undergo subsequent 1,4-conjugate ad-
dition as well as Diels Alder reactions to produce a wide range
of ortho ring alkylated phenols and chromans. Several generalities
have emerged. First, if a Grignard reagent is followed by NaBH4,
or if NaBH4 is used to initiate the cascade, then the process
requires protonation of intermediate A and a two-pot sequence.
Second, the use of Grignard reagents or of an organolithium
reagent followed by a Grignard reagent or MgBr2‚Et2O expedites
the formation of D and possibly the subsequent reaction. In some
cases, reduction occurs instead of 1,4-addition. However, this can
be avoided by raising reaction temperature, by employing RMgCl,
by using diethyl ether as the solvent, or by some combination of
these tactics.
Verification of the proposed intermediates A, B, C, and D
required further experimentation (Scheme 1). All attempts to
isolate D were unfruitful. However, the formation of products in
good yield when RLi preceded RMgX and in moderate yield when
RMgX preceded RLi seemed to indicate its presence. Curiously,
all attempts to use organolithium reagents as both R1M and R2M
failed. Upon further consideration the only indication that D had
formed, when organolithium species were added, were the
products described by entry 4 and entry 6. However, 7 in entry
Acknowledgment. Research Grants from Research Corporation
(R10296), the UC Cancer Committee on Research (19990641), and the
National Science Foundation (CHE-9971211) are greatly appreciated.
R.W.V. and J.N.C. acknowledge funds provided by the John H. Tokuyama
Memorial Graduate Fellowship and Robert H. DeWolfe Undergraduate
Research Award, respectively.
(13) Compound 9 was observed when adding MeLi to 3, but 9 was not
observed when adding MeMgBr to 3. The two pathways shown below may
explain the formation of 9. One involves a 1,5-sigmatropic shift while the
other one involves an intermolecular deprotonation of E, which could arise
by a divergent collapse of intermediate B (Scheme 1).
Supporting Information Available: A general procedure, charac-
terization, and 1H NMR spectra for compounds 3-4, 6-18 (PDF). This
JA994209S
Van De Water, Ryan W.
