Reinvestigation of the synthesis of per-benzylated glycosylethenes: A new result

Article abstract of DOI:10.1021/jo034795e

Addition of vinylmagnesium bromide on the perbenzylated glucono-1, 5-lactone gave, after reduction with Et₃SiH/BF₃· Et₂O, a mixture of the desired β-C-vinyl glucoside 1 and the unexpected β-C-but-3-enyl glucoside 3 resulting from double addition of vinylmagnesium bromide on the lactone. Similar results have been obtained with the perbenzylated galactono-1,5-lactone. This side reaction was then been explored to prepare β-C-but-3-enyl glycosides and other β-C-glycosyl derivatives by employing different Grignard reagents. An alternative approach to the per-benzylated glycosylethenes has been studied and compared.

Full text of DOI:10.1021/jo034795e

SCHEME 1
Rein vestiga tion of th e Syn th esis of
P er -ben zyla ted Glycosyleth en es: A New
Resu lt
J uan Xie,* Franc¸ois Durrat, and J ean-Marc Vale´ry
Laboratoire de Chimie des Glucides, CNRS UMR 7613,
Universite´ Pierre et Marie Curie, 4 Place J ussieu, 75005
Paris, France
TABLE 1.
xie@ccr.jussieu.fr
Received J une 10, 2003
Abstr a ct: Addition of vinylmagnesium bromide on the per-
benzylated glucono-1,5-lactone gave, after reduction with
Et3SiH/BF3‚Et2O, a mixture of the desired â-C-vinyl gluco-
side 1 and the unexpected â-C-but-3-enyl glucoside 3 result-
ing from double addition of vinylmagnesium bromide on the
lactone. Similar results have been obtained with the per-
benzylated galactono-1,5-lactone. This side reaction was then
been explored to prepare â-C-but-3-enyl glycosides and other
â-C-glycosyl derivatives by employing different Grignard
reagents. An alternative approach to the per-benzylated
glycosylethenes has been studied and compared.
products (yield)
CH2dCHMgBr
mono-adduct
bis-adduct
entry lactone
(equiv)
(1,5)
(3,6)
1
2
3
4
5
6
2
2
2
4
4
4
1.2
1.5
4
1.2
1.5
4
1 (49.3%)
1 (32.6%)
3 (13.8%)
3 (17.9%)
3 (89%)
5 (50%)
5 (25%)
6 (28.4%)
6 (47%)
Glycosylethenes (e.g. (2,3,4,6-tetra-O-benzyl-â-D-glu-
copyranosyl)ethene, 1) are versatile synthetic intermedi-
ates which have been used for the synthesis of C-glycosyl
amino acids1,2 and carbasugars.3 During the course of
studies on the synthesis of sugar amino acids, we needed
to prepare the â-C-vinyl compound 1 that could be readily
transformed into desired sugar amino acid derivatives.4
The first synthesis of this compound was reported by
Kraus and Molina, by treating the per-benzylated lactone
2 with 1.5 equiv of vinylmagnesium bromide at -78 °C
followed by reduction with Et3SiH and BF3‚Et2O, in 60%
yield with two steps (Scheme 1).5 However, this reaction
was found quite tricky in our hands. Our first attempt
using the described procedure led to two products: the
desired â-C-vinyl compound 1 together with the â-C-but-
3-enyl derivative 3 in 32.6% and 17.9% isolated yield
(Table 1, entry 2). The physical and spectral properties
of 3 were essentially identical with those reported, since
compound 3 has already been prepared by nucleophilic
addition of but-3-enylmagesium bromide on the lactone
2 followed by reduction.6,7 It is to be noticed that Dondoni
and colleagues have obtained the same yield of 1 (32.5%)
using similar reaction conditions.1 Nevertheless, the
formation of 3 has never been reported under these
conditions.
To improve the yield of the desired product 1, we tried
to modify the reaction conditions. The 1/3 ratio was
insensitive to the reaction concentration. Raising the
reaction temperature led to a complex mixture. When
conducted at lower temperature (-100 °C), the reaction
was extremely slow, and the majority of the starting
lactone was recovered. The best yield of 1 (49.3%) was
obtained with 1.2 equiv of nucleophile, along with 13.8%
of 3 (entry 1). A lower conversion of the starting lactone
was observed when using <1.2 equiv of vinylmagnesium
bromide. When the reaction was conducted with 4 equiv
of vinylmagnesium bromide, compound 3 was isolated
exclusively in 89% yield (entry 3).
The formation of 3 resulted from two additions of
vinylmagnesium bromide on the lactone 2 (Figure 1): the
first addition led to the deprotonated hemiketal A, which
is in equilibrium with the enone B; the second underwent
either a Michael fashion or a SN 2′ mechanism and
furnished the hydroxy enone or the hemiketal 7 after
hydrolysis.
To check this hypothesis, we then tried to trap the
postulated intermediates with other nucleophiles. Lac-
tone 2 was first treated with 1.2 equiv of vinylmagnesium
bromide over 2.5 h at -78 °C, followed by addition of 4
equiv of phenylmagnesium bromide (2.5 h at -78 °C).
Reduction of the obtained hemiketal with TMSOTf/Et3-
SiH afforded the desired compound 8 in 70% yield
(Scheme 2). However, all attempts to react the protonated
hemiketal A (compound 11, Scheme 3) with vinylmag-
nesium bromide failed (use of excess of nucleophile,
activation with nBuLi or Na, etc.). This might be due to
the difference of the lactol stereocenter of the deproto-
nated hemiketal A, formed probably by an R-attack of
the nucleophile on the lactone, with the protonated (and
also equilibrated) form 11 where the hydroxy group
prefers the R position. Although the formation of enone
* To whom correspondence should be addressed. Phone: 33-1-44-
27-58-93. Fax: 33-1-44-27-55-13.
(1) Dondoni, A.; Giovannini, P. P.; Marra, A. J . Chem. Soc., Perkin
Trans. 1 2001, 2380-2388.
(2) Westermann, B.; Walter, A.; Flo¨rke, U.; Altenbach, H.-J . Org.
Lett. 2001, 3, 1375-1378.
(3) Wang, W.; Zhang, Y.; Zhou, H.; Ble´riot, Y.; Sinay¨, P. Eur. J . Org.
Chem. 2001, 1053-1059.
(4) Xie, J . Eur. J . Org. Chem. 2002, 3411-3418.
(5) Kraus, G. A.; Molina, M. T. J . Org. Chem. 1988, 53, 752-753.
(6) Best, W. M.; Ferro, V.; Harle, J .; Stick, R. V.; Tilbrook, D. M. G.
Aust. J . Chem. 1997, 50, 463-472.
(7) Cipolla, L.; Nicotra, F.; Vismara, E.; Guerrini, M. Tetrahedron
1997, 53, 6163-6170.
10.1021/jo034795e CCC: $25.00 © 2003 American Chemical Society
Published on Web 09/06/2003
7896
J . Org. Chem. 2003, 68, 7896-7898

F IGURE 1. Proposed mechanism for the formation of the hemiketal 7.
SCHEME 2
prop-1-enylmagnesium bromide under the same condition
also furnished the product of double addition: compound
9 was isolated in 51% yield as a mixture of four diaster-
eoisomers (Scheme 2). Similar results were obtained with
the galactonolactone 4. As shown in Table 1, treatment
of lactone 4 with 1.5 equiv of vinylmagnesium bromide
gave, after reduction, 25% of 5 (product of single addition)
and 28.4% of 6 (product of double addition) (entry 5). Use
of 1.2 equiv of nucleophile led to the â-C-vinyl galactoside
5 in 50% yields (entry 4). The known compound 6, already
prepared by addition of but-3-enylmagesium bromide on
lactone 4,9,10 has been obtained in 47% yield by using 4
equiv of vinylmagnesium bromide (entry 6).
SCHEME 3
To avoid the formation of bis-adduct, we decided to
follow a different synthetic approach. In fact, Nakai and
colleagues have recently reported the synthesis of the
hemiketal 11 from the per-benzylated D-glucose 10 in
97% yield (Scheme 3).11 Their strategy implied the
addition of vinylmagnesium bromide on 10, followed by
oxidation of the so-formed allylic alcohol with MnO2.
Following this procedure, we have observed that the
oxidation step is very slow (more than one week’s
reaction) and a large excess of oxidant (three times in
weight) is needed. Reduction of the crude hemiketal 11
with Et3SiH/BF3‚Et2O (or TMSOTf) afforded compound
1 contaminated by an unidentified byproduct with similar
polarity. This has made the purification of 1 extremely
difficult. Compound 1 was isolated only in 35% total yield.
In the case of the galacto derivative, this reaction
sequence worked better: compound 5 was isolated in 65%
overall yield, without any difficulty of purification.
by double addition of vinylmagnesium bromide to esters
has already been reported,8 to the best of our knowledge,
this is the first report of formation of this kind of product
in the case of sugar lactone. We have therefore explored
the scope of this reaction. Treatment of 2 with 4 equiv of
In summary, we have demonstrated that nucleophilic
addition of vinylmagnesium bromide on the per-benzyl-
ated glycono-1,5-lactones led to the formation of double
conjugated addition product, along with the desired single
addition product. The reaction conditions have then been
(8) (a) Boccara, N.; Maitte, P. C. R. Hebd. Seances Acad. Sci. 1963,
257, 3427-3430. (b) Watanabe, S.; Suga, K.; Fujita, T.; Saito, N. Aust.
J . Chem. 1977, 30, 427-431. (c) Comins, D. L.; Herrick, J . J .
Tetrahedron Lett. 1984, 25, 1321-1324. (d) Larson, G. L.; Montes de
Lo´pez-Cepero, I.; Torres, L. E. Tetrahedron Lett. 1984, 25, 1672-1676.
(e) Larson, G. L.; Hernadez, D.; Montes de Lo´pez-Cepero, I.; Torres,
L. E. J . Org. Chem. 1985, 50, 5260-5267. (f) Larson, G. L.; Soderquist,
J . A.; Claudio M. R. Synth. Commun. 1990, 20, 1095-1104.
(9) Zheng, W.; DeMattei, J . A.; Wu, J .-P.; Duan, J . J .-W.; Cook, L.
R.; Oinuma, H.; Kishi, Y. J . Am. Chem. Soc. 1996, 118, 7946-7968.
(10) Wellner, E.; Gustafsson, T.; Ba¨cklund, J .; Holmdahl, R.; Kihl-
berg, J . ChemBioChem 2000, 1, 272-280.
(11) Tomooka, K.; Nakamura, Y.; Nakai, T. Synlett 1995, 321-322.
J . Org. Chem, Vol. 68, No. 20, 2003 7897

(c 1, CHCl3)]; 1H NMR (250 MHz, CDCl3) δ 1.43-1.57 (m, 1 H),
1.78-1.91 (m, 1 H), 1.92-2.27 (m, 2 H), 3.14 (td, 1 H, J ) 2.5,
9.3 Hz), 3.39-3.63 (m, 5 H), 3.89 (d, 1 H, J ) 2.5 Hz), 4.33 and
4.40 (2d, 2 H, J ) 11.8 Hz), 4.53-4.69 (m, 4 H), 4.83-4.94 (m,
4 H), 5.65-5.82 (m, 1 H), 7.14-7.31 (m, 20 H). 13C NMR (62.9
MHz, CDCl3) δ 29.9, 31.1, 69.2, 72.3, 73.6; 73.8, 74.5, 75.6, 77.2,
79.1, 79.3, 85.0, 114.7, 127.7, 127.8, 128.0, 128.2, 128.3, 128.5,
138.1, 138.5, 138.6, 138.9.
optimized to prepare the per-benzylated glycosylethenes
1 and 5. The second method explored, through an
alternative approach to the intermediate hemiketals 11
and 13, afforded a better result only for compound 5 (65%
total yield instead of 50%). The observed side reaction
with vinylmagnesium bromide has been successfully
explored to prepare the â-C-but-3-enyl glycosides and
other â-C-glycosides.
2-Meth yl-1-(2′,3′,4′,6′-tetr a -O-ben zyl-â-D-glu cop yr a n osyl)-
3-p en ten e (9): 51% of a mixture of four diastereoisomers.
Further purification by preparative thin-layer chromatography
(3:2 CH2Cl2/cyclohexane) allowed the separation into two major
isomers 9a and 9b. 9a : Rf 0.32 (3:2 CH2Cl2/cyclohexane), mp
78-80 °C; 1H NMR (250 MHz, CDCl3) δ 0.91 (d, 3 H, J ) 6.8
Hz), 1.35-1.61 (m, 2 H), 1.51 (dd, 3 H, J ) 1.8, 6.8 Hz), 2.80-
2.90 (m, 1 H), 3.13-3.28 (m, 3 H), 3.58-3.66 (m, 4 H), 4.45-
5.06 (m, 8 H), 4.97-5.06 (m, 1 H), 5.34-5.41 (m, 1 H), 7.21-
7.29 (m, 20 H). 13C NMR (62.9 MHz, CDCl3) δ 13.0, 21.9, 27.5,
39.3, 69.0, 73.6, 75.0, 75.3, 75.7, 78.7, 78.9, 82.8, 87.5, 123.7,
127.8, 128.0, 128.1, 128.6, 136.1, 138.4, 138.8, 159.4. 9b: Rf 0.27
(3:2 CH2Cl2/cyclohexane), mp 58-60 °C; 1H NMR (250 MHz,
CDCl3) δ 0.94 (d, 3 H, J ) 6.5 Hz), 1.33-1.70 (m, 2 H), 1.56 (dd,
3 H, J ) 1.0, 6.3 Hz), 2.72-2.88 (m, 1 H), 3.20-3.40 (m, 3 H),
3.58-3.68 (m, 4 H), 4.53-4.62 (m, 4 H), 4.77-4.85 (m, 4 H),
5.13-5.38 (m, 2 H), 7.21-7.31 (m, 20 H). 13C NMR (62.9 MHz,
CDCl3) δ 13.1, 20.1, 27.6, 39.3, 69.2, 73.6, 75.1, 75.4, 75.7, 77.3,
78.8, 79.2, 83.0, 87.6, 122.1, 127.8, 127.9, 128.1, 128.6, 137.4,
138.4, 138.8, 159.2. Anal. Calcd for C40H46O5: C, 77.18; H, 7.64.
Found: C, 77.36; H, 7.55.
Exp er im en ta l Section
Gen er a l Com m en ts. 1H and 13C NMR spectra were recorded
in CDCl3 solution. All assignments were confirmed by 1H/1H,
1H/13C correlations, and Dept 135. Column chromatography was
performed on E. Merck Silica Gel 60 (230-400 mesh). Analytical
thin-layer chromatography was performed on E. Merck alumi-
num percolated plates of Silica Gel 60F-254 with detection by
UV and by spraying with 6 N H2SO4 and heating about 2 min
at 300 °C. Dichloromethane was distilled over CaH2. THF was
distilled over sodium and benzophenone prior to use.
Gen er a l P r oced u r e for th e P r ep a r a tion of â-C-Alk en yl
Glycosid es (1, 3, 5, 6, a n d 9) by Ad d ition of Or ga n om a g-
n esiu m Rea gen t on th e La cton e. To a cold (-78 °C) solution
of lactone 2 or 4 (0.25 mmol) in THF (2.5 mL) under an argon
atmosphere was added dropwise the corresponding commercially
available organomagnesium reagent. After 2.5 h of stirring at
-78 °C, the reaction mixture was quenched with saturated NH4-
Cl and extracted with EtOAc. The combined organic layers were
washed with brine, dried (MgSO4), and concentrated to give the
corresponding hemiketal. To a cooled (-40 °C) solution of the
hemiketal in anhydrous acetonitrile (or dichloromethane when
TMSOTf was used) (1 mL) were added triethylsilane (3 equiv)
and boron trifluoride-diethyl ether (1 equiv) or TMSOTf (0.2
equiv). The solution was stirred 10 to 30 min at -40 °C
(monitoring by TLC), then quenched with saturated K2CO3 and
extracted with CH2Cl2. The combined organic layers were
washed with brine, dried (MgSO4), and concentrated. Purifica-
tion by preparative thin-layer chromatography (1:4 Et2O/cyclo-
hexane) afforded the corresponding â-C-glycoside.
(2,3,4,6-Tet r a -O-b en zyl-â-D-glu cop yr a n osyl)et h en e (1):
mp 72-73 °C, [R]D +33 (c 1, CH2Cl2) [lit.1 mp 68-69 °C, [R]D
+28 (c 0.2, CHCl3)].
1-(2′,3′,4′,6′-T e t r a -O -b e n zy l-â-D -g lu c o p y r a n o s y l)-3-
bu ten e (3): mp 83-84 °C, [R]D +6.2 (c 1, CH2Cl2) [lit.7 mp 83-
84 °C, [R]D +3 (c 1.2, CHCl3)].
(2,3,4,6-Te t r a -O-b e n zyl-â-D-ga la ct op yr a n osyl)e t h e n e
(5): oil, [R]D +19.2 (c 1.3, CHCl3); 1H NMR (250 MHz, CDCl3) δ
3.52-3.64 (m, 3 H), 3.70-3.74 (m, 2 H), 3.96 (d, 1 H, J ) 2.8
Hz), 4.41 (dd, 2 H, J ) 11.8, 15.5 Hz), 4.60-4.96 (m, 6 H), 5.25
(dd, 1 H, J ) 1.5, 10.5 Hz), 5.40 (dd, 1 H, J ) 1.5, 17.3 Hz),
5.89-5.95 (m, 1 H), 7.21-7.36 (m, 20 H); 13C NMR (62.9 MHz,
CDCl3) δ 67.0, 72.6, 73.6, 74.0, 74.6, 75.4, 76.9, 79.0, 80.8, 84.4,
118.4, 127.6, 127.7, 127.8, 128.3, 128.4, 128.5, 135.5, 138.0, 138.4,
138.6, 138.9. Anal. Calcd for C36H38O5: C, 78.51; H, 6.97.
Found: C, 78.42; H, 7.06.
1-(2,3,4,6-Tetr a -O-ben zyl-â-D-glu cop yr a n osyl)-2-p h en yl-
eth a n e (8). To a cold (-78 °C) solution of lactone 2 (135 mg,
0.25 mmol) in THF (2.5 mL) under an argon atmosphere was
added dropwise vinylmagnesium bromide (1 M solution in THF,
300 µL, 0.3 mmol). After 2.5 h of stirring at -78 °C, phenyl-
magnesium bromide (1 M solution in THF, 1 mL, 1 mmol) was
added and the reaction was continued over 2.5 h at -78 °C. The
reaction mixture was then quenched with saturated NH4Cl and
treated as described in the general procedure. The corresponding
hemiketal was reduced with 3 equiv of Et3SiH and 0.6 equiv of
TMSOTf as described in the general procedure to afford, after
purification by preparative thin-layer chromatography (1:4 Et2O/
cyclohexane), 110 mg of a white solid: 70%, mp 90-91 °C, [R]D
-8.3 (c 0.7, CHCl3); 1H NMR (250 MHz, CDCl3) δ 1.83-1.93 (m,
1 H), 2.16-2.26 (m, 1 H), 2.72-2.85 (m, 1 H), 2.91-3.00 (m, 1
H), 3.28-3.50 (m, 3 H), 3.69-3.88 (m, 4 H), 4.64-4.78 (m, 4 H),
4.90-4.97 (m, 4 H), 7.22-7.45 (m, 25 H). 13C NMR (62.9 MHz,
CDCl3) δ 31.8, 33.5, 69.2, 73.6, 75.1, 75.4, 75.7, 78.4, 78.8, 79.0,
82.6, 87.4, 125.9, 127.7, 127.9, 128.0, 128.1, 128.5, 128.6, 128.7,
138.2, 138.3, 138.7, 142.2. Anal. Calcd for C42H44O5: C, 80.23;
H, 7.05. Found: C, 80.08; H, 7.19.
Ack n ow led gm en t. The authors thank Drs Se´bas-
tien Comesse, Luc Dechoux, and Louis Hamon for
fruitful discussions.
1-(2′,3′,4′,6′-Te t r a -O-b e n zyl-â-D -ga la ct op yr a n osyl)-3-
bu ten e (6): mp 54-56 °C, [R]D +2.2 (c 1, CHCl3) [lit.10 [R]D -2.0
J O034795E
7898 J . Org. Chem., Vol. 68, No. 20, 2003