Organic Letters
Letter
a
lactone from the β-position. However, the initial attack from
the β-position in both reports led to skeletal rearrangements of
the butenolide framework in a reaction cascade. With our
Table 1. Optimization Studies
7
ongoing interest and endeavors in butenolide chemistry, we
envisioned a regioselective β-attack of deconjugated buteno-
lides in an enol ester-type reactivity (unexplored).
8
Among various reports on α-addition, only a few were based
on Lewis acid catalyzed α-addition with retention of a double
bond without isomerization toward more stable α,β-unsatu-
rated butenolide. Toward this goal, Boukouvalas et al. have
b
yield
9
entry
1
catalyst
solvent
temp (°C) time (min)
3a
4
successfully achieved the α-addition of 2-furanolates regiose-
lectively using an Sn-enolate-based chelation-controlled strat-
egy. Recently, Hartwig and co-workers exquisitely demon-
strated Ir-catalyzed regio- and enantioselective α-allylation of
BF ·OEt2
CH Cl2
0
0
5
83
37
<5
28
67
78
14
69
86
3
2
c,d
2
3
4
5
6
7
8
9
1
1
1
BF ·OEt2
CH Cl2
15
5
33
74
45
−
3
2
e
Bi(OTf)3
Cu(OTf)2
Sc(OTf)3
AgOTf
CH Cl2
0
2
CH Cl2
0
15
5
2
10
trimethylsilyloxyfuran. In fact, in a few synthetic method-
CH Cl2
0
2
ologies, α-addition with silyloxyfurans was reported as a minor
CH Cl2
0
10
15
15
5
8
2
11
product but not explored to a large extent. Recently, Zhou
and co-workers reported an enantioselective α-addition/
transesterification of deconjugated butenolides with o-quinone
f
La(OTf)3
Bi(OTf)3
Bi(OTf)3
BH*a
BH*a
BH*b
CH Cl2
0
6
2
THF
0
12
CH CN
0
3
1
2
methides. To the best of our knowledge, there is no report of
Lewis acid catalyzed nucleophilic addition from the α,β,γ-
positions of butenolides on a p-QMs as a single substrate.
In recent years, p-quinone methides have been explored
extensively due to their unique ability as powerful Michael
0
1
2
toluene
toluene
toluene
25
60
25
24 h
24 h
24 h
NR
NR
NR
a
Unless otherwise stated, the reaction was performed with p-QMs 1a
0.17 mmol, 50 mg), α-angelica lactone 2a (0.17 mmol, 15 μL), and
Lewis acid/BH* (20 mol %) in 2 mL of solvent at the specified
(
13
acceptors with a variety of nucleophiles. In 2004, Eklund and
1
4
b
c
co-workers elegantly showed the oxidative metabolism of
plant lignan hydroxymatairesinol to its corresponding butyr-
olactone lignins, isohydroxymatairesinol and epi-isohydroxyma-
tairesinol, via a p-QMs intermediate. Recently, our group also
temperature. Isolated yields. 4 Å molecular sieves (50 mg).
d
e
Formation of 3a′ was observed in 11% yield. No reaction when
performed in the presence of 4 Å molecular sieves, and the yield of 3a
f
with 10 mol % of Bi(OTf) was 55%. Yields brsm 1a. NR = no
reaction. BH* = Appropriate chiral phosphoric acid (for the structure
3
reported a Tf NH-catalyzed 1,6-conjugate addition reaction of
2
15
p-QMs with vinyl azide. In continuation herein, we report a
highly efficient and regioselective 1,6-conjugate addition of
deconjugated butenolides and silyloxyfurans to p-QMs
catalyzed by Lewis acid, leading to a diversely substituted
butenolide-derived diarylmethane scaffold.
considered attempting the asymmetric version of same β-
addition reaction. To this end, we tested chiral phosphoric acids
containing bulky groups on the BINOL backbone such as (S)-
TRIP catalyst, but unfortunately, the reaction did not work
(entries 10 and 11).
This could probably be attributed to the inefficiency of the
catalyst to activate p-QMs toward nucleophilic attack. A similar
set of disappointing results was obtained on switching to chiral
metal phosphates with a view to activate p-QMs through the
interaction of its lone pair with the Lewis acidic metal (entry
12).
To investigate our hypothesis, we started our exploration of
an β-addition reaction using α-angelica lactone 2a and p-QMs
1
a as a model substrate. Table 1 summarizes the effect of
several parameters on this reaction. Initially, when 20 mol % of
BF ·OEt was used to catalyze the reaction between 1a and 2a
3
2
in CH Cl at 0 °C, it resulted in the formation of undesired
2
2
hydrolyzed product 4 exclusively in 5 min only via β-attack
entry 1). The formation of product 4 gave us the idea about
(
the reaction proceeding via an enol ester-type reactivity. The
intermediate oxonium ion thus formed was quenched due to
the presence of traces of moisture, thus accounting for the
formation of hydrolyzed product 4. To rationalize our concept
and to minimize the side product, we attempted the reaction
using activated molecular sieves (entry 2). Though this has
resulted in the formation of desired product 3a in 33% yield
along with product 3a′ having an isomerized exo-double bond
in 11% yield, the formation of hydrolyzed product 4 in 37%
yield could not be suppressed. Use of other Lewis acids seemed
to be the best alternative for improving yields and selectivity of
After having attempted the chiral catalysts in triggering the β-
addition, we considered inducing chirality employing a chiral
auxiliary. Accordingly, we prepared (−)-menthol-incorporated
1
7
p-QMs and subjected it to the standardized reaction
conditions, but to our dismay, we ended up obtaining the
undesired auxiliary-cleaved product 5 in 91% yield. The
formation of 5 may be attributed to steric crowding toward
the incoming nucleophile, thus facilitating intramolecular 1,6-
conjugate addition by ester carbonyl followed by elimination of
the menthol moiety (Scheme 2). On the other hand, BF ·OEt
16
3
2
worked well in the case of chiral p-QMs to deliver the required
1
product 3a. Interestingly, 20 mol % of Bi(OTf) afforded 74%
product 6 in 74% yield with dr ∼ 1:1 ( H NMR analysis).
3
yield of product 3a with the formation of a trace amount of 4
The effect of temperature on 1,6-conjugate addition reactions
was also studied. To our surprise, the reaction did not work at
lower temperature (−78 °C). Interestingly, though the reaction
failed to proceed at 0 °C for more than 2 days, as soon as it was
brought to room temperature the formation of product 6 was
observed within 5 min. Thus, we observed that the influence of
chiral auxiliary on the reactive site is minimal, which can be
(entry 3). Screening of other Lewis acids such as Cu(OTf)2,
Sc(OTf) , AgOTf, and La(OTf) was ineffective in terms of
3
3
product selectivity and yields (entries 4−7). With the
promising result of Bi(OTf) , we further screened its efficacy
3
in other solvents such as THF and CH CN but ended with
3
unsatisfactory results (entries 8 and 9). As this reaction led to
the formation of product 3a in a nearly racemic form, we
attributed not only to the presence of spacer (−CO −), thus
2
B
Org. Lett. XXXX, XXX, XXX−XXX