A novel glycosidation of glycosyl fluoride using a designed ionic liquid and its effect on the stereoselectivity

Article abstract of DOI:10.1016/j.tetlet.2004.07.128

The glycosidations of glucopyranosyl fluoride and alcohols using an ionic liquid containing a protic acid effectively proceeded under mild conditions to afford the corresponding glycosides in good to high yields. The stereoselectivity of the glycosidation

Full text of DOI:10.1016/j.tetlet.2004.07.128

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 7043–7047
A novel glycosidation of glycosyl fluoride using a designed
ionic liquid and its effect on the stereoselectivity
Kaname Sasaki, Shuichi Matsumura and Kazunobu Toshima*
Department of Applied Chemistry, Faculty of Science and Technology, Keio University, 3-14-1 Hiyoshi, Kohoku-ku,
Yokohama 223-8522, Japan
Received 29 June 2004; revised 26 July 2004; accepted 28 July 2004
Available online 13 August 2004
Abstract—The glycosidations of glucopyranosyl fluoride and alcohols using an ionic liquid containing a protic acid effectively pro-
ceeded under mild conditions to afford the corresponding glycosides in good to high yields. The stereoselectivity of the glycosidation
was significantly affected by the ionic liquid solvent. 1-n-Hexyl-3-methylimidazolium trifluoromethanesulfonate (C
taining
6
mim[OTf]), con-
trifluoromethanesulfonate anion, and 1-(3-cyanopropyl)-3-methylimidazolium trifluoromethanesulfonimidide
a
(
CNC
3
2
mim[NTf ]), possessing a cyano group at the side chain of the imidazolium cation, gave the b-stereoselectivity.
ꢀ
2004 Elsevier Ltd. All rights reserved.
The development of a practical and environmentally
benign glycosidation, which is one of the most impor-
tant and fundamental transformation reactions of
carbohydrates, is now becoming more and more impor-
tant, and urgently needed both in the laboratory and in
with an anion species. Therefore, if an oxonium interme-
diate could interact with an anion species contained in
ionic liquid, stereoselective glycosidation induced by
ionic liquid would be achieved. Although Poletti et al.
showed that the stereoselectivity of a glycosidation using
a glycosyl trichloroacetimidate was changed by the ionic
1
2
industry. The greening of chemistry in the field of gly-
cosidation reactions may include the use of an environ-
mentally benign catalyst and solvent, both of which
could be reused. Recently, ionic liquids have been
described as one of the most promising environmentally
6
liquid, a systematic study related to the effect of the
ionic liquid on the stereoselectivity of the glycosidation
reaction has never been reported. Herein, we report
the stereoselective glycosidation of glycosyl fluoride
and alcohols controlled by a designed ionic liquid and
its effect on the stereoselectivity (Fig. 1).
3
benign reaction media. In this context, we previously
reported the glycosidation of glucopyranosyl phosphite
4
using an ionic liquid containing a protic acid, and the
glycosidations of glycals and glycosyl trichloroacetimi-
dates in ionic liquids were independently reported by
7
In this study, we selected a glycosyl fluoride as the gly-
cosyl donor because glycosyl fluoride is effectively acti-
5
6
8
Yadav et al. and Poletti et al., respectively. In these
studies, the recyclability of the ionic liquids as greener
solvents for the glycosidation reactions was well demon-
strated. Another more promising feature of ionic liquid
is its designability. Through our previous study, we have
expected that stereoselectivity of glycosidation would be
strongly influenced by ionic liquid as a solvent because
many glycosidation reactions involve a cationic species,
namely, an oxonium intermediate, which could interact
vated by both Lewis and protic acids. For the
reaction solvent, we chose several ionic substances,
which are liquids at room temperature, 25ꢁC, having a
low viscosity, and have the ability to dissolve the glyco-
syl donor and several glycosyl acceptors as shown in
Figure 1. Thus, we first examined the glycosidations of
the glucopyranosyl fluoride 1b (a/b = 1/>20) and cyclo-
hexylmethanol (2) (2.0equiv to 1) using several ionic
liquids (0.1M for 1b) containing a protic acid (3mol%
to ionic liquid (IL)) possessing a common anion with
the ionic liquid, such as 1-n-hexyl-3-methylimidazolium
tetrafluoroborate (C mim[BF ]) with HBF , 1-n-hexyl-
6
4
4
Keywords: Carbohydrate; Glycosidation; Glycosyl fluoride; Ionic liq-
uid; Designability.
3
-methylimidazolium
trifluoromethanesulfonimidide
(C mim[NTf ]) with HNTf , 1-n-hexyl-3-methylimid-
*
Corresponding author. Tel./fax: +81 45 566 1576; e-mail: toshima@
applc.keio.ac.jp
6
2
2
azolium perchlorate (C mim[ClO ]) with HClO and
6
4
4
0
040-4039/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.07.128

7044
K. Sasaki et al. / Tetrahedron Letters 45 (2004) 7043–7047
Figure 1. Glycosidations of glycosyl fluoride and alcohols using an acid–ionic liquid.
1
-n-hexyl-3-methylimidazolium trifluoromethanesulfo-
ent of the configuration at the anomeric center of the
glycosyl donor 1, and this reaction proceeded via the
SN1 type pathway and involved a cationic species, that
is, an oxonium intermediate. Furthermore, since no epi-
merization of the generated glycosidic bond occurred
during the glycosidation reaction, the stereoselectivity
of the glycosidation was due to kinetic control. There-
fore, the a-stereoselectivity observed using C mim[NTf ]
nate (C mim[OTf]) with HOTf at 25ꢁC. These results
are summarized in Table 1. Although the glycosidation
of 1 using C mim[BF ] with HBF needed a long reac-
6
6
4
4
tion time and gave a moderate yield of the correspond-
ing glycoside 8 (Table 1, entry 1), other glycosidations
proceeded very smoothly to afford 8 in good to high
yields (Table 1, entries 2–4). These results clearly indi-
cated, for the first time, that the glycosyl fluoride was
effectively activated by an acid–ionic liquid, namely,
an ionic liquid containing a protic acid, and coupled
with an alcohol. It was noteworthy that the stereoselec-
tivity of the glycosidation was highly dependent on the
catalyst–solvent system. Among the examined acid–
ionic liquids, C mim[NTf ] containing HNTf provided
6
2
containing HNTf must result from the anomeric effect,
2
while the b-stereoselectivity shown using C mim[ClO ]
6
4
with HClO or C mim[OTf] with HOTf must arise from
the a-oriented coordination of the perchlorate or tri-
fluoromethanesulfonate anion with the oxonium
4
6
9
intermediate.
6
2
2
the a-stereoselectivity with good yield (Table 1, entry 2),
At this stage, we had a question as to which anion is a
real factor to induce the unusual b-stereoselectivity with-
out neighboring group participation, the anion from the
ionic liquid or from the protic acid. To answer this ques-
tion, we next examined the glycosidations of 1 and 2
using C mim[NTf ] with HOTf and C mim[OTf] with
HNTf , both of which contained a protic acid possessing
2
a different anion from the ionic liquid (Table 1, entries 6
and 7). Comparing these results with those using
while C mim[OTf] with HOTf afforded the highest b-
6
stereoselectivity with high yield (Table 1, entry 4). It
was also confirmed that although the reactivity of 1b
was higher than that of the corresponding a-anomer
1
a (a/b = >20/1), the stereoselectivity of the glycosida-
tion using 1b was quite similar to that using 1a (Table
, entry 4 vs 5). These results clearly showed that the
6
2
6
1
stereoselectivity of the glycosidation is highly independ-
Table 1. Glycosidations of 1 and 2 in ionic liquids containing protic acids
a
b
Entry
Donor
Ionic liquid
Protic acid
HBF
HNTf
Time/h
Yield/%
a/b Ratio
1
2
3
4
5
6
7
1b
1b
1b
1b
1a
1b
1b
C
C
C
C
C
C
C
6
6
6
6
6
6
6
mim[BF
mim[NTf
mim[ClO
4
]
4
24
1
41
86
83
89
85
82
88
47/53
68/32
30/70
24/76
26/74
52/48
25/75
2
]
]
2
4
HClO
HOTf
HOTf
HOTf
4
1
mim[OTf]
mim[OTf]
mim[NTf
mim[OTf]
1
4
2
]
1
HNTf
2
1
a
Isolated yields after purification by column chromatography.
b
ꢂ
a/b Ratios were determined by HPLC analysis (column, CrestPak C18S , 4.6 · 150mm; eluent, 10% H O in MeCN; flow rate, 1.0mL/min, 40ꢁC;
detection, UV 250nm).
2

K. Sasaki et al. / Tetrahedron Letters 45 (2004) 7043–7047
7045
C mim[NTf ] with HNTf and C mim[OTf] with HOTf,
6
and 2 in a 7:3 (mol/mol) mixture of C mim[OTf] and
6
2
2
6
it was made clear that the b-stereoselectivity was ob-
C mim[NTf ] at 0ꢁC for 24h gave the glycoside 8 in
6
2
served only when C mim[OTf] was used as the solvent.
high yield (91%) with good b-stereoselectivity (a/
b = 18/82).
6
These results demonstrated that the stereoselectivity of
the glycosidation was highly dependent on the ionic liq-
uid, not the protic acid, and the b-stereoselectivity was
induced by the a-coordination of the trifluoromethane-
Knowing these preferable results, we next carried out
the glycosidations of 1 and several alcohols 3–7 includ-
ing the sugar derivatives to examine the scope and lim-
itations of the present glycosidation. These results are
outlined in Table 3. It was found that other primary
alcohols 3 and 4 were effectively coupled with 1, as well
as 2, to afford the corresponding glycosides in high
yields with good b-stereoselectivities (Table 3, entries
1–3). When the reactions were performed using the sec-
ondary alcohols 5 and 6, satisfactory chemical yields
and stereoselectivities were obtained at the higher tem-
sulfonate anion from the ionic liquid, C mim[OTf], with
6
the oxonium intermediate.
With these new results in hand, we next attempted the
glycosidation reaction at a lower temperature to further
improve the b-stereoselectivity, because the a-triflate
intermediate must be held more tightly at lower temper-
atures. However, C mim[OTf] itself is unfortunately a
6
solid salt at 0ꢁC. Although 1-ethyl-3-methylimidazolium
trifluoromethanesulfonate (C mim[OTf]), which pos-
sesses a shorter alkyl chain, has a lower melting point,
it is much more hydroscopic than C mim[OTf]. Indeed,
perature of 25ꢁC in C mim[OTf] as the sole solvent
2
6
(Table 3, entries 4 and 5). Unfortunately, the low reac-
tive secondary alcohol 7, however, gave a poor result
under similar reaction conditions (Table 3).
6
the chemical yield of the glycosidation of 1 and 2 using
C mim[OTf] was lower than that employing C -
2
6
mim[OTf]. Therefore, the use of a mixture of C6-
mim[OTf] and C mim[NTf ] was attempted. From the
results summarized in Table 2, it was found that
although a longer reaction time (24h) was required at
Our attention was next turned to the modification of the
side chain of the imidazolium cation moiety of the ionic
liquid, C mim[NTf ]. Therefore, we newly designed
6
2
6
2
and synthesized 1-(2-ethoxyethyl)-3-methylimidazolium
trifluoromethanesulfonimidide (EtOC mim[NTf ]) pos-
0
ꢁC to obtain the high yield of the corresponding glyco-
side, the b-stereoselectivity increased as expected (Table
, entries 7–12). Furthermore, the b-stereoselectivity
2
2
sessing an ether function and 1-(3-cyanopropyl)-
3-methylimidazolium trifluoromethanesulfonimidide
(CNC mim[NTf ]) containing a cyano group. The
2
1
0
gradually increased as the ratio of C mim[OTf] in the
6
3
2
mixed ionic liquid increased. The same tendency was
also observed at 25ꢁC (Table 2, entries 1–6). These
results also showed that the trifluoromethanesulfonate
anion competed with the trifluoromethanesulfonimide
anion for the coordination with the oxonium cation gen-
erated from the glycosyl donor, and the b-stereoselectiv-
ity of the present glycosidation was induced by the
trifluoromethanesulfonate anion from the ionic liquid,
C mim[OTf]. Since the 8:2 (mol/mol) mixture of C -
results of the glycosidations of 1 and 2 using a protic
acid, HNTf , in these novel ionic liquids are described
2
in Figure 2. When EtOC mim[NTf ] was used as the
2
2
solvent, unfortunately, no stereoselectivity was
observed. However, in the case of CNC mim[NTf ], b-
3
2
stereoselectivity was induced. The chemical yield and
stereoselectivity were very similar to those using
C mim[OTf] under similar conditions. Compared the
6
result when using C mim[NTf ], the cyano group in
2
6
6
6
mim[OTf] and C mim[NTf ] became a solid (Table 2,
2
CNC mim[NTf ] significantly influenced the stereoselec-
3 2
6
entries 11 and 12), the highest b-stereoselectivity was
obtained when the 7:3 (mol/mol) mixture was employed
at 0ꢁC (Table 2, entry 10). Thus, the glycosidation of 1
tivity of the glycosidation. These results strongly sug-
1
1
gested that the cyano group in the side chain of the
imidazolium cation coordinated with the oxonium
Table 2. Glycosidations of 1 and 2 in mixed ionic liquids
Entry
Ratio of mixed ionic liquid (C
6
mim[NTf ]/C
2
6
mim[OTf])
25ꢁC, 1h (Entries 1–6)
0ꢁC, 24h (Entries 7–12)
b
c
b
c
Yield/%
a/b Ratio
Yield/%
a/b Ratio
1
2
3
4
5
6
, 7
10/0
7/3
82
90
89
89
93
89
52/48
39/61
30/70
27/73
25/75
24/76
81
90
89
91
––
––
52/48
24/76
21/79
18/82
––
, 8
, 9
5/5
3/7
, 10
, 11
, 12
a
a
2/8
0/10
––
a
Isolated yields after purification by column chromatography.
b
ꢂ
a/b Ratios were determined by HPLC analysis (column, CrestPak C18S , 4.6 · 150mm; eluent, 10% H
detection, UV 250nm).
2
O in MeCN; flow rate, 1.0mL/min, 40ꢁC;
c
The used ionic liquid is solid state at 0ꢁC.

7046
K. Sasaki et al. / Tetrahedron Letters 45 (2004) 7043–7047
Table 3. Glycosidations of 1 and several alcohols 2–7
a
b
c
Entry
Alcohol
Conditions
Product
Yield/%
a/b Ratio
1
2
3
4
5
6
2
3
4
5
6
7
A
A
A
B
B
B
8
9
91
95
99
91
77
54
18/82
20/80
13/87
29/71
28/72
42/58
10
11
12
13
a
b
c
Conditions A: C
6
mim[OTf]/C mim[NTf
6
2
] = 7/3, 0ꢁC, 24h. Conditions B: C
6
mim[OTf], 25ꢁC, 1h.
Isolated yields after purification by column chromatography.
a/b Ratios were determined by HPLC analysis (column, CrestPak C18S , 4.6 · 150mm; eluent, 10% H
O in MeCN for entry 3; flow rate, 1.0mL/min, 40ꢁC; detection, UV 250nm).
ꢂ
2
O in MeCN for entries 1, 2, and 4–6, 12.5%
H
2
2 2 3 2
Figure 2. Glycosidations of 1b and 2 using EtOC mim[NTf ] and CNC mim[NTf ].
intermediate more effectively than trifluoromethanesulf-
onimide anion. These results also indicated that the
interaction between the oxonium intermediate and tri-
fluoromethanesulfonimide anion was very weak and
did not affect the stereoselectivity of the glycosidation.
Boons, B. G. Tetrahedron 1996, 52, 1095; (e) Davis, B. G.
J. Chem. Soc., Perkin. Trans. 1 2000, 2137.
. Green Chemistry: Theory and Practice; Anastas, P. T.,
Warner, J. C., Eds.; Oxford University Press: Great
Britain, 1998.
2
3
. (a) Welton, T. Chem. Rev. 1999, 99, 2071; (b) Wasser-
scheid, P.; Keim, W. Angew. Chem., Int. Ed. 2000, 39,
In conclusion, we presented novel glycosidations of glu-
copyranosyl fluoride and alcohols using an ionic liquid
containing a protic acid. Furthermore, although the
stereoselectivity is still not very high, the effect of the
ionic liquids on the stereoselectivity of the glycosidation
was clearly demonstrated. These results should be
instructive for further research that employs ionic
liquids in stereoselective and environmentally benign
glycosidation reactions. Further studies along this line
are currently in progress.
3
772; (c) Sheldon, R. Chem. Commun. 2001, 2399; (d) Ionic
Liquid in Synthesis; Wasserscheid, P., Welton, T., Eds.;
Wiley-VCH: Weinheim, 2002.
4. Sasaki, K.; Nagai, H.; Matsumura, S.; Toshima, K.
Tetrahedron Lett. 2003, 44, 5605.
. Yadav, J. S.; Reddy, B. V. S.; Reddy, J. S. S. J. Chem.
Soc., Perkin Trans. 1 2002, 2390.
. Poletti, L.; Rencurosi, A.; Lay, L.; Russo, G. Synlett 2003,
2
. For reviews on glycosyl fluoride, see: (a) Shimizu, M.;
Togo, H.; Yokoyama, M. Synthesis 1998, 799; (b)
Toshima, K. Carbohydr. Res. 2000, 327, 15.
5
6
7
297.
8
. For glycosidations of glycosyl fluoride using protic acid,
see: (a) Toshima, K.; Kasumi, K.; Matsumura, S. Synlett
Acknowledgements
1
998, 643; (b) Toshima, K.; Kasumi, K.; Matsumura, S.
Synlett 1999, 813–815; (c) Mukaiyama, T.; Jona, H.;
Takeuchi, K. Chem. Lett. 2000, 696; (d) Jona, H.;
Takeuchi, K.; Mukaiyama, T. Chem. Lett. 2000, 1278;
This research was supported in part by Grant-in-Aid for
the 21st Century COE program ÔKEIO Life Conjugate
ChemistryÕ from the Ministry of Education, Culture,
Sports, Science, and Technology, Japan.
(
e) Jona, H.; Mandai, H.; Mukaiyama, T. Chem. Lett.
001, 426.
. Crich, D.; Sun, S. J. Org. Chem. 1996, 61, 4506.
0. The synthesis of newly designed ionic liquids will be
2
9
1
1
References and notes
reported elsewhere. CNC mim[NTf ]:
3
H
NMR
2
(
300MHz, neat): d 8.46 (1H, s, N–CH@N), 7.39 (1H, s,
1
. (a) Schmidt, R. R. Angew. Chem., Int. Ed. Engl. 1986, 25,
12; (b) Sina, P. Pure. Appl. Chem. 1991, 63, 519; (c)
Toshima, K.; Tatsuta, K. Chem. Rev. 1993, 93, 1503; (d)
CH@C), 7.33 (1H, s, CH@C), 4.20 (2H, t, J1,2 = 5.7Hz,
N–CH ), 3.81 (3H, s, N–CH ), 2.43 (2H, t, J = 5.7Hz,
2
2
3
2,3
1
3
CH –CN), 2.17 (2H, tt, J = J2,3 = 5.7Hz, CH2);
2 1,2
C

K. Sasaki et al. / Tetrahedron Letters 45 (2004) 7043–7047
7047
NMR (75MHz, neat): d 136.48, 124.22, 122.47, 120.21 (q,
C,F = 318.7Hz), 119.33, 48.38, 36.05, 25.78, 13.81.
(2H, q, J4,5 = 6.9Hz, CH
2
–CH
3
), 0.21 (3H, q, J4,5 =
1
3
J
6.9Hz, CH –CH ); C NMR (75MHz, neat): d 136.44,
2 3
1
EtOC mim[NTf ]: H NMR (300MHz, neat): d 7.68
123.26, 122.90, 119.90 (q, JC,F = 318.5Hz), 67.54, 66.18,
49.70, 35.66, 14.07.
2
2
(
1H, s, N–CH@N), 6.60 (1H, s, CH@C), 6.51 (1H, s,
CH@C), 3.43 (2H, t, J1,2 = 4.2Hz, N–CH ), 3.02 (3H, s,
N–CH ), 2.87 (2H, t, J1,2 = 4.2Hz, CH –CH –O), 2.60
2
11. Hashimoto, S.; Hayashi, M.; Noyori, R. Tetrahedron Lett.
1984, 25, 1379.
3
2
2