Organic Letters
Letter
Scheme 2. Synthesis of the Model Substrate 2-
Iodophenyl)acrylamidoglycopyranose 2a
revealed that stirring the reaction for only 30 min under
otherwise the same conditions led to a mixture of the desired
products β-3a and α-3a in an 85:15 ratio and the same yield for
the isolated β-3a (62%, entry 2). The screening reaction was
optimized with respect to the amount of the base. The use of
1
.5 equiv of n-Bu NOAc was found to be sufficient, giving rise
4
to β-3a in a good 62% yield (compare entries 2 and 4, 5).
Pleasingly, the yield of β-3a improved to 69% by omitting
TBAB as an additive (entry 6). In the next set of experiments,
pure (2-iodophenyl)-Z-acrylamide-β-glycopyranose (Zβ)-2a
was used as starting material. Under the optimal conditions,
N-glycosyl quinolin-2-one β-3a was formed in 76% yield as a
single β-isomer without any anomerization (entry 7). This
result suggested that α-3a was formed from (Zα)-2a. To clarify
this hypothesis, the coupling of pure (Zα)-2a was further
investigated under the optimized conditions. Accordingly, only
α-3a was isolated as a single anomer, and the coupling constant
verify the hypothesis that the (Eβ)-2a isomer does not react or
isomerize into the more reactive (Zβ)-2a, we carried out the
reaction with the pure (Eβ)-2a isomer. Under the optimized
conditions, no reaction occurred and the formation of
compound 3a was never detected (entry 9). Of note,
performing the reaction using the (2-bromophenyl)acrylamido-
glycopyranose instead of the iodinated derivative 2a furnished
the desired product 3a in 46% yield; however, the same
reaction with the (2-chlorophenyl)acrylamidoglycopyranose
failed.
a
Isomer ratios were measured by integration of well-resolved signals in
the 300 MHz 1H NMR (CDCl3).
9
glycopyranose employing Xu’s conditions (Scheme 2). In
contrast to Xu’s results, the expected coupling product (Zβ)-2a
was isolated together with two other byproducts in 75% yield
1
2
(
Scheme 2). After a tedious separation, 1D- and 2D-NMR
analysis revealed that these byproducts corresponded to the
(
Next, we continued our study by exploring the feasibility of
2
the intramolecular key step C(sp )−N bond formation of the 1-
amidosugar Zβ-2a contaminated with a small amount of its Eβ-
and Zα-isomers (Zβ/Zα/Eβ = 1:0.2:0.06). To our surprise,
heating the mixture in the presence of Pd(OAc) (5 mol %),
2
tetrabutylammonium acetate (3 equiv) as the base, and
tetrabutylammonium bromide (3 equiv) in dioxane at 130 °C
for 4 h led to N-β-glycosyl quinolin-2-one β-3a (J1,2 = 9.9 Hz)
in 63% yield (entry 1, Table 1) together with its anomer α-3a
derived from this cyclization of (Zα)-2a. Further optimization
With this encouraging result in hand, we next turned our
attention to examining the scope and limitations of the
coupling reaction of various β- or α-substituted (2-iodophenyl)-
Z-acrylamidosugars. As shown in Scheme 3, the amidosugars 2
were prepared from the coupling of substituted 3-(2-
iodophenyl)acrylic acids 1 as a mixture of Z/E-isomers (Z/E
Table 1. Survey of Reaction Conditions for the
Intramolecular N-Arylation of Tetraacetyl β-Amidoglucose
2
a
=
93:7) with 1-aminosugar derivatives. In all cases studied, the
a
desired coupling amidosugars Zβ-2 were obtained as the major
used crude for the next step. Gratifyingly, various substituted
(
2-iodophenyl)-Z-acrylamido-β-glycopyranoses reacted effi-
ciently to give the N-glycosylated quinolin-2-ones 3a−n as a
mixture of β and α anomers in ratios ranging from 1:0.1 to
1
:0.27 and yields up to 76%. Interestingly, this cross-coupling
tolerated the presence of C−halogen bonds (e.g., −Br, −Cl,
−F) which offers a platform for further metal-catalyzed cross-
coupling reactions (compounds 3b−e, 3h, 3k, and 3m,n). In
addition, the protocol was compatible with different amidosu-
gars such as galactoside and mannoside giving the correspond-
ing products 3g−i in 76%, 55%, and 68% yields, respectively.
This coupling is not limited to monoamidosaccharides, but also
works with peracetylated β-D-disaccharides derived from D-
cellobiose octaacetate. The N-glycosylated quinolin-2-ones 3j,k
were obtained in 64% and 60% yields, respectively.
Interestingly, this study was extended successfully to
unprotected amidosugars. As shown in Scheme 3, derivatives
ratio of 2a
Zβ/Zα/Eβ
temp
ratio
yield ofd
b
c
entry
(°C)
t (h) additive β-3a/α-3a
β-3a(%)
1
2
3
4
5
6
7
8
9
1:0.2:0.06
1:0.2:0.06
1:0.2:0.06
1:0.2:0.06
1:0.2:0.06
1:0.2:0.06
1:0:0
130
130
130
130
130
100
100
100
100
4
TBAB
TBAB
TBAB
TBAB
TBAB
−
85:15
85:15
−
85:15
85:15
85:15
100:0
0:100
−
63
62
0.5
0.25
0.5
0.5
1
e
−
f
63
g
62
69
1
−
76
h
0:1:0
1
−
72
0:0:1
1
−
00
a
2
a (1 equiv), Pd(OAc) (5 mol %), base (3 equiv), additive (3
2
b
2
l−n reacted efficiently under our optimized conditions leading
equiv), anhydrous dioxane (0.1 M). TBAB = nBu NBr (tetrabutyl-
4
c
1
to the product 3l−n bearing unprotected sugar moieties. Of
note in all cases, the β and α anomers were separated by SiO
amonium bromide). Ratio was determined by H NMR in the crude
reaction mixture based on the chemical shift (ppm) of the proton
2
d
flash chromatography or preparative HPLC and the NMR of
signal H2 for β-3a (δ = 5.90) and H3 for α-3a (δ = 6.15). Yield of
e
f
isolated β-3a. Only 25% of conversion of 2a. 2 equiv of base were
g
h
used. 1.5 equiv of base were used. Yield of isolated α-3a.
B
Org. Lett. XXXX, XXX, XXX−XXX