NMR evidence for the participation of triflated ionic liquids in glycosylation reaction mechanisms

Article abstract of DOI:10.1016/j.carres.2006.02.021

A systematic low-temperature NMR study of a glycosylation reaction was performed in the presence of different ionic liquids and acidic catalysts. The influence of the triflate anion derived from [emim][OTf] on the stereochemistry of the glycosylation prod

Full text of DOI:10.1016/j.carres.2006.02.021

Carbohydrate Research 341 (2006) 903–908
Note
NMR evidence for the participation of triflated ionic liquids
in glycosylation reaction mechanisms
a
b
b
c
b,
*
Anna Rencurosi, Luigi Lay, Giovanni Russo, Enrico Caneva and Laura Poletti
a
CNR, Istituto di Scienze e Tecnologie Molecolari, via Golgi, 19, 20133 Milan, Italy
University of Milan, Dipartimento di Chimica Organica e Industriale, via Venezian, 21, 20133 Milan, Italy
University of Milan, Centro Interdipartimentale Grandi Apparecchiature (C.I.G.A.), via Golgi, 19, 20133 Milan, Italy
b
c
Received 22 December 2005; received in revised form 17 February 2006; accepted 21 February 2006
Available online 10 March 2006
Abstract—A systematic low-temperature NMR study of a glycosylation reaction was performed in the presence of different ionic
liquids and acidic catalysts. The influence of the triflate anion derived from [emim][OTf] on the stereochemistry of the glycosylation
products was evaluated.
Ó 2006 Elsevier Ltd. All rights reserved.
Keywords: Ionic liquids; Glycosylation; Low-temperature NMR
The growing awareness of the multiple biological func-
tions of carbohydrates over the past 30 years has stim-
1a,b, which contain a non-participating group on O-2,
by comparing their glycosylation properties in a classical
solvent (dichloromethane) and ILs composed of both a
coordinating and a non-coordinating anion (Scheme 1
1
ulated the development of numerous glycosylation
2
protocols. However, influencing the reaction course
6
towards a unique anomeric product still often requires
a demanding effort in regulating coupling conditions
and Table 1). Our results, supported by low-tempera-
1
ture H NMR experiments, assessed the influence of
3
and protecting group pattern, because these reactions
[emim][OTf] on the stereochemical outcome of the reac-
tion through the formation of a transient a-glycosyl tri-
flate, which shields the a-face of the oxonium ion and
biases the attack of the acceptor from the b-face.
proceed through many possible intermediates, the most
4
common one being an oxocarbenium ion.
We recently described the effectiveness and scope of a
glycosylation reaction using trichloroacetimidate donors
Very little is known about the influence of ILs in reac-
tion mechanisms and only few papers have appeared in
the literature describing investigations of reaction mech-
5
in ionic liquids (ILs). The reaction conditions were mild
and the ILs could be recycled while retaining the cata-
lytic properties of the Lewis acid. Interestingly, when
performed in [emim][OTf] with trichloroacetimidates
bearing non-participating groups on O-2, the reaction
afforded glycoside products without the need for an
acidic catalyst. Moreover, the b-glycoside was obtained
as the major product, even in the absence of the anchi-
meric assistance of the protecting group on O-2.
7
anisms in ILs, and in particular by NMR spectroscopy.
TMSOTf, r.t.
O
O
Ionic Liquid
X
O
R'O
R'O
1α,β
Y
2α,β
OH
We have subsequently performed a systematic study
on 2,3,4,6-tetra-O-benzyl glucose trichloroacetimidates
X or Y =
O
CCl3
NH
*
Corresponding author. Tel.: +39 02 50314063; fax: +39 02
0314061; e-mail: laura.poletti@unimi.it
Scheme 1. Glycosylation of isopropanol with a- and b-
trichloroacetimidates.
5
0
008-6215/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2006.02.021

904
A. Rencurosi et al. / Carbohydrate Research 341 (2006) 903–908
a
Table 1. Glycosylation of isopropanol with a- and b-2,3,4,6-tetra-O-
benzyl glucopyranose trichloroacetimidates 1 in different solvents
Table 2. Low-temperature NMR experiments
a
b
c
Exp. # Donor Ionic liquid (equiv)
Lewis acid (equiv)
b
Entry Donor Solvent
Product Yield
%)
a/b Ratio
1
1
a
b
1a
1b
1a
1b
1a
1b
1a
1b
[emim][OTf] (1.0)
[emim][OTf] (0.6)
TMSOTf (0.01)
TMSOTf (0.01)
TMSOTf (0.03)
TMSOTf (up to 0.26)
TMSOTf (0.2)
TMSOTf (0.1)
BF
BF
(
1
2
3
4
1a
CH
[emim][OTf]
[emim][OTf]
2
Cl
2
2a,b
85
67
79
65
16/84
16/84
15/85
25/75
2a
2b
3a
3b
4a
4b
[bmim][PF
[bmim][PF
—
6
] (1.8)
] (1.6)
6
c
CH
emim][OTf] 1:1
[bmim][PF
2
Cl
2
/
—
[
[emim][OTf] (1.6)
[emim][OTf] (1.4)
3
ÆEt
ÆEt
2
O (0.45)
O (0.45)
5
6
]
98
86
18/82
70/30
3
2
a
2 2
In a typical experiment, the donor was dissolved in CD Cl in an
6
7
8
9
1b
2 2
CH Cl
[emim][OTf]
[emim][OTf]
2a,b
NMR tube together with the IL (except entries 3a and 3b) and cooled
to À78 °C. The Lewis acid was subsequently added at the same
temperature and the NMR spectra were recorded at regular time
intervals.
The exact amount of the IL was ascertained through integration of
the signals of the H-2 of the IL and of the NH of the donor before
addition of the Lewis acid.
Quant. 45/55
85
72
c
20/80
54/46
CH
emim][OTf] 1:1
[bmim][PF
2 2
Cl /
[
b
c
1
0
6
]
98
76/24
a
All reactions were carried out with 20 equiv of isopropanol,
.01 equiv of TMSOTf in 0.5 mL of solvent at room temperature.
Determined by NMR spectroscopy.
0
The exact amount of Lewis acid was determined through the inte-
b
c
gration of Si(CH
3 3 3 2 2
) signal for TMSOTf and of (CH CH ) O signal
Reaction performed without Lewis acid catalysis.
for BF ÆEt O.
3
2
Therefore, we describe here a more detailed study on the
possible intermediates generated upon acidic activation
of glycosyl donors 1a,b in the solvents reported in Table
to the b-glucosyl triflate was detected during the exper-
iment. The peaks corresponding to 1a and to the a-
glucosyl triflate disappeared after 40 and 100 min,
respectively, affording degradation by-products. The
same experiment was performed with donor 1a (Table
2, Experiment 1a), which, on the contrary, gave a slug-
gish reaction. The H-1 peak intensity decreased slowly
and only a small amount of the a-triflate was generated.
In a subsequent experiment, an IL with a non-coordi-
nating anion was used, under similar conditions
(0.03 equiv of TMSOTf, Experiments 2a,b). In this way,
we hoped to detect any alternative intermediate that
could explain the inversion of the anomeric configuration
in the products observed in this IL (Table 1, entries 5 and
10). Under these conditions, the reaction of donor 1a, as
well as of donor 1b, was very slow and no formation of the
anomeric triflate was detected, not even after 150 min.
This is in line with a mechanism where the direct attack
of the acceptor on trichloroacetimidate affords the glyco-
side product with the opposite anomeric configuration.
1
, using low-temperature NMR spectroscopy.
The H NMR experiments were performed as out-
1
lined in Scheme 2 and Table 2. Donor 1a or 1b was
dissolved in CD Cl together with the IL (except
2
2
Experiments 3a and 3b) and the sample was cooled to
À78 °C. The exact amount of the IL was calculated by
integration of NH of the donor and H-2 of the IL.
The acidic catalyst was subsequently added at the same
temperature and the NMR spectra were recorded at
regular time intervals. Experiment 1b revealed that 1b
(
H-1, doublet at 5.74 ppm, J = 7.4 Hz; NH, singlet at
8
.83 ppm) was converted within 30 min to glucosyl imi-
date 1a (H-1, broad doublet at 6.40 ppm, J = 3.4 Hz;
NH, singlet at 8.71 ppm) and to the a-glucosyl triflate,
whose chemical shift (a doublet at d 6.16 ppm with
J = 2.9 Hz) was consistent with the data reported in
8
the literature (Fig. 1a and b). No peak corresponding
a)
OBn
OBn
OBn
O
O
BnO
BnO
O
CD2Cl2, -78 C
˚
BnO
BnO
+
CCl3
BnO
BnO
+
degradation
by-products
O
CCl3
BnO
IL x eq.
Lewis Acid.
O
OBn
BnO
NH
OTf
1
α
2
1β
NH
b)
OBn
O
OBn
O
CD2Cl2, -78 C
BnO
˚
BnO
BnO
+
degradation
by-products
BnO
BnO
IL x eq.
Lewis Acid.
O
CCl3
BnO
OTf
2
1α
NH
Scheme 2. Low-temperature NMR experiments.

A. Rencurosi et al. / Carbohydrate Research 341 (2006) 903–908
905
1
Figure 1. Partial H NMR spectrum from Experiments 1 (Table 2). (a) Experiment 1a: The singlet at 9.06 ppm corresponds to the H-2 proton of
emim][OTf], whose chemical shift slightly varies in the presence of the Lewis acid. The singlet at 8.71 ppm corresponds to the NH proton of 1a; the
signal at 6.40 ppm (a singlet that is resolved to a doublet during the course of the experiment, J = 3.4 Hz) is the anomeric proton of 1a; the doublet at
.16 ppm (J = 2.9 Hz) corresponds to the anomeric proton of a-triflate 2. (b) Experiment 1b: The singlet at 8.83 ppm corresponds to the NH of 1b;
the doublet at 5.74 ppm (J = 7.4 Hz) corresponds to the anomeric proton of 1b.
[
6

906
A. Rencurosi et al. / Carbohydrate Research 341 (2006) 903–908
1
9
Nevertheless, when the amount of the Lewis acid was
raised to 0.26 equiv (Experiment 2b), compound 1b
quickly anomerized to 1a, and was then converted to
the a-glucosyl triflate, which was stable up to 6 h under
these conditions. This behaviour raised a critical issue be-
cause a relatively low amount of TMSOTf (0.2 equiv) was
sufficient to induce the formation of an anomeric triflate,
the a-triflate observed in Experiment 1 could derive from
both the [emim][OTf] and from the Lewis acid. Experi-
ments 3, performed without the addition of IL but using
the a-glucosyl triflate before degradation (Fig. 2b).
F
NMR spectroscopy (Fig. 3a and b) revealed a peak at
d À77.9 ppm that appeared in both the experiments, in
agreement with the formation of the anomeric a-triflate
observed in the H NMR spectra. The chemical shift of
this peak is consistent with the values corresponding to
an anomeric triflate reported in the literature by Lowary
1
8
b
and co-workers.
Taken together, our results support a direct participa-
tion of the anion of [emim][OTf] in the glycosylation
reaction by the formation of a transient a-glycosyl tri-
flate, which biases the attack of the glycosyl acceptor
from the b-face. On the other hand, in the presence of
0
.1 equiv of TMSOTf, supported this hypothesis. In fact,
under these conditions, both the donors formed the a-tri-
flate after 3 min. Donor 1b also displayed the concomi-
tant anomerization to its a-anomer.
[bmim][PF ] no transient species was detected by
6
To gain definitive evidence that the triflate anion of
emim][OTf] was able to provide the anomeric a-triflate,
in Experiments 4 boron trifluoride–diethyl ether com-
NMR experiments. Therefore, in the glycosylation reac-
tion this IL displays a behaviour similar to classical non-
coordinating solvents, where the direct attack of the
acceptor to the donor is favoured, providing a product
with the opposite configuration.
[
plex (BF ÆEt O) was used as the promoter. As this Lewis
3
2
acid is weaker than TMSOTf, more equivalents were
added to the mixture (0.45 equiv). After the addition
9
ILs can be therefore considered as ‘designer solvents’
1
19
of BF ÆEt O, H and F NMR spectra were recorded
for glycosylation reactions, as they can act during the
course of the reaction by modulating the stereoselectiv-
ity of the products. This method provides a new way to
simplify the construction of oligosaccharides without
the use of complex protecting group patterns. Further
3
2
at regular time intervals. Donor 1a was completely con-
verted within 30 min to the anomeric a-triflate, which
was stable up to 2 h (Fig. 2a). Donor 1b, in turn, epimer-
ized to the a-donor and was partially transformed into
1
Figure 2. Partial H NMR spectrum of Experiment 4 (Table 2). (a) Experiment 4a: The broad singlet at 6.40 ppm is the anomeric proton of 1a; the
peak at 6.16 ppm, which in this experiment appears as a broad singlet, corresponds to the anomeric proton of a-triflate 2. (b) Experiment 4b: The
doublet at 5.74 ppm, (J = 7.4 Hz) corresponds to the anomeric proton of 1b; the anomeric proton of a-triflate 2 appears as a doublet at 6.16 ppm
(
J = 2.9 Hz).

A. Rencurosi et al. / Carbohydrate Research 341 (2006) 903–908
907
19
Figure 3. Partial F NMR spectrum of Experiments 4a (a) and 4b (b) (Table 2). The singlet at À77.9 ppm corresponds to the anomeric a-triflate; the
singlet at À81.2 ppm corresponds to the triflate of [emim][OTf].
studies on the influence of differently coordinating ionic
liquids in glycosylation reactions are under investigation
and the results of these studies will be reported in due
course.
ble at these conditions and therefore the shimming of
2
the sample (lock on the H nuclei of CD Cl solvent)
2
19
2
1
and the acquisition of the first H or F spectrum
(labelled with t = 0) was performed. The donor was
activated by addition of the Lewis acid (equivalents
according to Table 2) at À78 °C. A second spectrum
was acquired after the necessary manipulation and lock
1. Experimental
1
19
stabilization time (t = 5 min). H and F experiments
35 min each) were subsequently acquired at regular
(
1
.1. General methods
time intervals until the no changes were evident in the
spectrum.
1
19
H NMR and F NMR experiments were performed
using a 500 MHz Bruker Avance spectrometer, equipped
1
3
13
19
1
with a 5 mm QNP probe ( C, P, F/ H coils) and
Acknowledgements
with an N evaporator cooling system unit (able to work
2
down to À150 °C), which was used for experiments ac-
This work was supported by MIUR (COFIN 2004, pro-
tocol number 2004039212). A.R. is grateful to Regione
Lombardia (Assessorato alla Sanit a` ) for financial
support.
quired at low temperature.
1
.2. Representative experiment
Glucosyl imidate 1 (1 equiv) was dissolved in CD Cl in
Supplementary data
2
2
a 5 mm NMR tube in the presence of the ionic liquid
equivalents according to Table 2, calculated by integra-
tion of the NH of the donor and the H-2 of the IL) and
the sample was cooled to À78 °C. The mixture was sta-
(
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.carres.
2006.02.021.

908
A. Rencurosi et al. / Carbohydrate Research 341 (2006) 903–908
4. Demchenko, A. Curr. Org. Chem. 2003, 7, 35–79, and
References
references cited therein.
5
6
7
. Poletti, L.; Rencurosi, A.; Lay, L.; Russo, G. Synlett 2003,
1
2
. (a) Dwek, R. A. Chem. Rev. 1996, 96, 683–720; (b) Sears,
P.; Wong, C.-H. Angew. Chem., Int. Ed. 1999, 38, 2300–
324; (c) Nicolau, K. C.; Mitchell, H. J. Angew. Chem., Int.
Ed. 2001, 40, 1576–1624.
. (a) Banoub, J.; Boullanger, P.; Lafont, D. Chem. Rev. 1992,
1
5, 2297–2300.
. Rencurosi, A.; Lay, L.; Russo, G.; Caneva, E.; Poletti, L.
J. Org. Chem. 2005, 70, 7765–7768.
. (a) Nara, S. J.; Harjani, J. R.; Salunkhe, M. M. J. Org.
Chem. 2001, 66, 8616–8620; (b) Kim, H. S.; Kim, Y. J.; Bae,
J. Y.; Kim, S. J.; Lah, M. S.; Chin, C. S. Organometallics
2
92, 1167–1195; (b) Toshima, K.; Tatsuta, K. Chem. Rev.
1993, 93, 1503–1531; (c) Schmidt, R. R.; Kinzy, W. Adv.
2
003, 22, 2498–2504.
. (a) Crich, D.; Sun, S. Tetrahedron 1998, 54, 8321–8348;
b) Callam, C. S.; Gadikota, R. R.; Krein, D. M.;
Lowary, T. L. J. Am. Chem. Soc. 2003, 125, 13112–
3119.
Carbohydr. Chem. Biochem. 1994, 50, 21–123; (d) Taylor, C.
M. Tetrahedron 1998, 54, 11317–11362; (e) Seeberger, P. H.;
Haase, W.-C. Chem. Rev. 2000, 100, 4349–4393.
8
9
(
3
. (a) Nukada, T.; Berces, A.; Zgierski, M. Z.; Whitfield, D.
M. J. Am. Chem. Soc. 1998, 120, 13291–13295; (b) Schmidt,
R. R.; Michael, J. J. Carbohydr. Chem. 1985, 4, 141–169; (c)
Schmidt, R. R.; Behrendt, M.; Toepfer, A. Synlett 1990, 11,
1
. Sheldon, R. Chem. Commun. 2001, 2399–2407, and refer-
ences cited therein.
694–696.