Glycosylation mediated - BAIL in aqueous solution

Article abstract of DOI:10.1016/j.carres.2013.08.009

The use of Br?nsted acid ionic liquid (BAIL) as a catalyst for the activation of unreactive and unprotected glycosyl donors has been demonstrated for the first time in aqueous solution.

Full text of DOI:10.1016/j.carres.2013.08.009

Carbohydrate Research 381 (2013) 12–18
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Glycosylation mediated—BAIL in aqueous solution
Sébastien Delacroix a,b,c, Jean-Pierre Bonnet b,c, Matthieu Courty b,c, Denis Postel a,c
,
Albert Nguyen Van Nhien a,c,
⇑
a Laboratoire des Glucides CNRS FRE3517, 33 rue Saint-Leu, 80039 Amiens, France
b Laboratoire de Réactivité et de Chimie des Solides UMR7314, 33 rue Saint-Leu, 80039 Amiens, France
c Institut de Chimie de Picardie FR3085, 33 rue Saint-Leu, 80000 Amiens, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 8 May 2013
Received in revised form 25 July 2013
Accepted 14 August 2013
Available online 23 August 2013
The use of Brønsted acid ionic liquid (BAIL) as a catalyst for the activation of unreactive and unprotected
glycosyl donors has been demonstrated for the first time in aqueous solution.
Ó 2013 Elsevier Ltd. All rights reserved.
Keywords:
Brønsted acid ionic liquid
Tetrazole
Glycosylation
Aqueous media
1. Introduction
Brønsted acid ionic liquids (BAILs)2 are a class of ionic liquids
that can be used as solvent and/or as acid catalysts. Among all
Carbohydrates are abundant natural resources and renewable
feedstock that can be used particularly in chemistry in various
application fields. Linking carbohydrate units by glycosylation
has proven to be one of the most difficult synthetic processes to
control stereoselectivity. Among all the existing strategies for the
glycosylation of sugars, Fischer’s glycosylation is the earliest and
the simplest. Despite many innovative glycosylation strategies1
developed to improve the overall yield, chemoselectivity, regiose-
lectivity, and stereoselectivity (from neighboring group participa-
tion of 2-O-acyl protected glycosyl donors to intramolecular
strategies, use of asymmetric transition metal complexes and more
recently asymmetric Brønsted acid catalysis), most of them require
temporary protection of both acceptor and donor.
The development of a useful and environmentally friendly gly-
cosylation, which is one of the most important and fundamental
transformation reactions of carbohydrates, is now becoming more
and more important, and urgently needed both in laboratory and in
industry. With the approach to minimize the use of hazardous
materials or volatile solvents, Ionic Liquids (ILs) have attracted
much attention because of their advantageous properties as a reac-
tion media, which include negligible vapor pressure and high ther-
mal stability.
the good reasons to study ionic liquids as alternative Brønsted
acids3 in acid catalyzed reactions, one of them is the possibility
to adjust solubility properties by different cation/anion combina-
tions to allow a systematic optimization of the biphasic reaction.
For example, BAILs have been studied in Mannich reaction,4
Friedel–Crafts alkylation,5 Fischer indole synthesis,6 and esterifica-
tion.7–9
Glycosylation using unprotected and unactivated donors10 is of-
ten preferable as it can reduce the number of steps, allow for a dif-
ferent stereochemistry, and increase prospects for further process
modifications. However, their use involves a number of challenges,
including poor solubility in most conventional organic solvents.
Glycosylation reactions performed in Ionic Liquids (ILs) are now
emerging. Linhardt11,12 and Augé13 have recently reported a conve-
nient method to catalyze O-glycosylation with free carbohydrates.
Whereas Linhardt was studying the sole 1-ethyl-3-methylimidazo-
lium benzoate [emIm][ba] in the presence of Amberlite IR-120 (H+)
or p-toluenesulfonic acid (TsOH) to promote the glycosylation in
the presence of a large excess of alcohol acceptor, Augé screened
various commercially available ILs demonstrating that the best
combination was obtained for 1-butyl-3-methylimidazolium tri-
flate [BMIM][OTf] and the robust lewis acid scandium triflate
Sc(OTf)3 with 5 equiv of alcohol. In both cases,
major product.
a-anomer was the
The combination of ILs and Lewis acid has also been achieved to
perform the b-selectivity by using protected and activated glycosyl
⇑
Corresponding author. Tel.: +33 322827569; fax: +33 322827568.
E-mail address: albert.nguyen-van-nhien@u-picardie.fr (A.N. Van Nhien).
0008-6215/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.carres.2013.08.009

S. Delacroix et al. / Carbohydrate Research 381 (2013) 12–18
13
donors such as glycosyl trichloroacetamidate,14,15 glycosyl
phosphite,16 glycosyl fluoride,17 and thioglycoside.18 A wide range
of acid reagents such as montmorillonite K10,19 amberlite resin,20
N
HTFSI, H2O
rt, 5h
N
H3N
H2N
N
N
HN
HN
N
NTf2
H2SO4–silica,21 AuCl3,22 FeCl3, ZnCl2, BF3.OEt, TMSOTf, Cu(OTf)2
23
N
also provided the O-glycosyl derivatives on unreactive and unpro-
tected sugars. Moreover, many of the Lewis acids are moisture sen-
sitive and metal triflates or gold catalysts are highly expensive.
Therefore, the search for newer and efficient synthetic methodolo-
gies with greater efficiency and convenient procedures continues
to be a challenge.
5
5a
(>99%)
Scheme 3.
Besides the use of ILs as a new tool for green processes, the use
of water as a solvent has also received considerable attention in
synthetic organic chemistry for several reasons. Water will reduce
the use of harmful organic solvents and it is also unnecessary to
dry co-solvents, substrates and reagents leading to time and en-
ergy savings.
The chemical yield is essentially quantitative. Moreover, since nei-
ther reaction produces byproducts (except water), the ILs synthesis
is 100% atom efficient.
The ILs derivatives 3a and 4a were obtained in a similar way
(Scheme 2). After a classical Z-protection of the amino group of
D- or L-phenylalanine, the carboxylic group was transformed into
However, the use of water as a reaction medium can also limit
the efficiency of the reaction especially if equilibrium is disadvan-
taged as in the case of steps involving dehydration. A representa-
tive example is the Fischer acid-catalyzed O-glycosylation. This
drawback could be easily circumvented. Kobayashi and co-work-
ers24 reported the use of dodecylbenzenesulfonic acid (DBSA) as
a surfactant-type to catalyze etherification. Some examples of
esterification catalyzed by BAILs have also been described.7–9
We therefore initiated an investigation to see if Brønsted acid
ionic liquid could be used to catalyze the simple glycosylation
using water as co-solvent.
amide by using Boc2O, NH4HCO3 in pyridine–acetonitrile. After
dehydration by POCl3, the nitrile derivatives were obtained. Addi-
tion of sodium azide in DMF afforded the cyclic tetrazoles 3 in 14%
and tetrazole 4 in 23% overall yield (over 5 steps). Those modest
yields are attributed to the last step of Cbz deprotection. In the case
of L-Pro (Scheme 1), a slight hydrogen pressure was exerted during
the hydrogenation step. These conditions would probably increase
this last step (performed under atmospheric pressure) as well as
the addition of a catalytic amount of acetic acid. Metathesis ex-
change with HTFSI led quantitatively to ILs 3a and 4a. Similarly,
the known alanine tetrazole 526 was exchanged with HTFSI to af-
ford IL 5a (Scheme 3).
2. Results and discussion
Thermal stability (Fig. 1).
Considering the wide variety of ionic liquids, we choose TFSIꢀ as
anion for its hydrophobic properties and its high stability and the
tetrazole ring as anion since they represent good bio-isosteres of
carboxylic acid. Moreover, they are planar and aromatic and gener-
ally have proven better reactivity and/or selectivity due to a great-
er charge-stabilization compared to their corresponding carboxylic
acid isoster.
2.1. Thermogravimetric analysis (TGA)
The stability of the synthesized BAILs was determined by TGA
measurements to prove their compatibility with the thermal glyco-
sylation reactions conditions (T = 80 °C). It appears that BAILs 3a
and 4a are stable until 150 °C whereas 2 and 5a start to decompose
around 178 °C and 190 °C respectively which correspond to the
degradation of the cation moiety, mostly into CO2 and N2. The
beginning of the degradation of the anion TFSIꢀ which is more sta-
ble than organic cations, starts around 320 °C for every IL.
As a preliminary study, we synthesized BAIL 2. Tetrazole-pro-
line 1 was readily prepared from natural L-proline (Scheme 1). Be-
cause of its high efficiency in a number of reactions,25 compound 1
is lately commercially available. The multistep synthesis will be
described for similar amino acids a little later. The metathesis ex-
change was then accomplished with an aqueous solution of HTFSI.
After elimination of water under vacuo, compound 2 was obtained.
The newly synthesized ILs were screened as solvent/catalysts
for the catalytic evaluation of direct glycosylation (Table 1). We
initiated our study with
D
-mannose which is known to give prefer-
entially
a-pyranoside and thus easy purification. In this reaction
vi
i, ii, iii
iv, v
N
N
L-proline
CN
N
N
N
Cbz
N
N
H
H2
HN
HN
N
N
NTf2
1
2
Scheme 1. Reagents and conditions: (i) NaOH 2 M, Cbz-Cl (1.5 equiv), 5 °C, 2 h (95%), (ii) Boc2O (1.5 equiv), NH4HCO3 (1.5 equiv), pyridine (0.7 equiv), CH3CN, rt, 68 h (90%),
(iii) cyanuric chloride, DMF, Argon, 1 h at 0 °C then rt for 3 h (>99%), (iv) NaN3 (1.1 equiv), NH4Cl (1.1 equiv), DMF, 90 °C, 8 h (>99%), (v) Pd/C (10%), H2, EtOH, 24 h, rt (90%), (vi)
(CF3SO2)2NH (1 equiv), H2O, rt, 5 h (>99%).
Ph
Ph
Ph
vi
D
-phenylalanine i, ii, iii
iv, v
N
N
H3N
NTf2
L
-phenylalanine
N
H2N
HN
Cbz
CN
N
HN
N
HN
N
3
3a
(D
)- 14% (over 5 steps)
(>99%)
4a
(>99%)
4
(L)- 23% (over 5 steps)
Scheme 2. Reagents and conditions: (i) NaOH 2 M, Cbz-Cl (1.1 equiv), 5 °C, 2 h, (ii) Boc2O (1.3 equiv), NH4HCO3 (1.2 equiv), pyridine (3 equiv), CH3CN, rt, 68 h, (iii) POCl3
(1.2 equiv) in DCM, pyridine, ꢀ10 °C, 3 h, (iv) NaN3 (1.1 equiv), NH4Cl (1.1 equiv), DMF, 90 °C, 8 h, (v) Pd/C (10%), H2, EtOH, 24 h, rt, (vi) (CF3SO2)2NH (1 equiv), H2O, rt, 5 h.

14
S. Delacroix et al. / Carbohydrate Research 381 (2013) 12–18
Table 2
Mass = f (T) (10 K/min, Ar)
Glycosylation of alcohols using BAIL 3a as catalysta
100
Alcohol
Yieldb
a
/bc
BAIL 3a & 4a
BAIL 2
n-Butan-1-ol
n-Heptan-1-ol
n-Octan-1-ol
n-Decan-1-ol
n-Dodecan-1-ol
Cyclohexanol
Allyl alcohol
61d
50
89/11
91/9
94/6
93/7
91/9
90/10
n.d.g
99/1
n.d.
69e
38
BAIL 5a
47
75
50
25
0
45f
Traces
25
Traces
Propargyl alcohol
Benzyl alcohol
a
Typical procedure:
D-mannose, octanol (5 equiv), H2O (2 equiv), and BAIL
(10 mol %) were stirred at 80 °C. After 72 h, water was eliminated by evaporation
and the reaction mixture was purified by flash chromatography.
b
Isolated yield.
Determined by NMR.
48 h reaction time.
24 h reaction times.
47% yield obtained after 24 h.
Not determined.
c
d
e
f
g
Nonetheless when more water was added in our glycosylation
reaction (5 equiv instead of 2 equiv, entry 1), the conversion rate
was lower than 30%. A large excess of water probably shifts disad-
vantageously the equilibrium involved to the classical oxocarbeni-
um formation under thermodynamic control. Moreover, the
T (°C)
500
0
100
200
300
400
600
chirality brought by ILs 3a and 4a (
L-series vs
D-series) showed
no difference either on the
(69% vs 64%).
a/b ratio (94/6 vs 90/10) or on the yield
Figure 1. TGA curves of compounds 2, 3a, 4a, and 5a.
After screening the different ILs as catalyst, we focused our
attention toward the potential versatility of these BAILs and partic-
ularly 3a (Table 2) which offered the best yield in the model
reaction (Table 1) in the glycosylation of D-mannose with various
alcohols. Interestingly, with lipophilic alcohols such as heptan-1-
ol to dodecan-1-ol or cyclohexanol, the yields were obtained
Table 1
Optimization of the reaction conditionsa
HO
OH
HO
HO
OH
O
ILs (10 mol%)
O
HO
+ octanol
OH
HO
HO
H2O, 80°C, 24h
between 45% and 69% yield with a high
a-selectivity (>85%). In
OC8H17
contrast, with benzyl alcohol (more liphophilic alcohol), so as with
water-soluble allylic alcohol, no glycosylation products were
obtained. We supposed that these results could be attributed to
the fact that those alcohols were firstly dehydrated by the BAIL
and that, unlike propargyl alcohol, they could be stabilized by
resonance effects, thus preventing the O-glycosylation.
Entry
IL
Yieldb (%)
a
/bc
1
2
3
4
2
62
69d
64
46
91/9
94/6
90/10
80/20
3a
4a
5a
a
Typical procedure:
D-mannose, octanol (5 equiv), H2O (2 equiv), and IL
Finally, various sugars were tested to expand the scope and lim-
itation of this reaction and the results were summarized in Table 3.
(10 mol %) were stirred at 80 °C. After 24 h, water was eliminated by evaporation
and the reaction mixture purified by flash chromatography.
Starting from D-glucose and after 24 h at room temperature, an
b
Isolated yield.
Determined by NMR.
No reactivity when the reaction was performed at room temperature.
c
anomeric mixture of pyranoside A was obtained in a low 14% yield
(entry 1) accompanied by the formation of the kinetic furanoside B.
This moderate yield was ascribed to the weak reactivity of the
d
model, water was added to dissolve the unprotected sugar other-
starting D-glucose. To overcome the formation of the furanosides,
wise no reactivity was observed. Glycosylation of octanol with
D
-
the reaction was performed by using 30 mol % of BAIL 3a (entry
2). Surprisingly, the pyranoside A was only formed in a 1.5/1 ano-
mannose was carried out in the presence of the synthesized ILs.
After 24 h of reaction at 80 °C, water was eliminated under vac-
uum. Due to the small amount of water used in the following reac-
tions, water has not been recycled. Nonetheless, the reaction was
also controlled by mass spectrometry and the only detected com-
pounds were the alcohol, the sugar, the O-glycoside, the tetrazole
moiety and the presence of its anion TFSIꢀ detected by ESIꢀ (neg-
ative electrospray). No by-products were detected by MS and
TLC. As a matter of fact, these ILs gave encouraging results espe-
cially with BAILs 2 and 3a (62% and 69% yield respectively) with
meric ratio with a better 35% yield. In a similar fashion,
(entry 5) also led to a mixture of furanoside and pyranoside. Nev-
ertheless, when -mannose was treated with catalyst 3a, only the
D-galactose
D
more stable pyranoside was obtained in a good yield of 69% (entry
8). In view of these results, we performed the other reactions
involving the pentoses
D-fucose, D-xylose, D-arabinose, D-ribose
and -arabinose. Similarly, the glycoside derivatives were obtained
L
as a mixture of furanosides and pyranosides with yields ranging
from 52% to 79%. In all cases, the more stable pyranoside derivative
a
-selectivity higher than 90%. The catalyst 1 precursor of BAIL 2
was obtained except for
formed (entry 9).
D-ribose where the furanoside was mostly
was already known to be an efficient catalyst in many reactions.27
For example, Saito and Yamamoto28 have demonstrated that the
acidity of the protonic acids of catalyst 1 plays a critical role in
enhancing reactivity, catalyst efficiency, and enantioselectivity.
Adding that, increasing the water ratio leads to an enhancement
of the reactivity.
2.2. Determination of pKa
Many chemical transformations are sensitive to the acidity.
Therefore, knowledge of the pKa value is important in order to

S. Delacroix et al. / Carbohydrate Research 381 (2013) 12–18
15
Table 3
Table 4
pKa values of BAILs
Glycosylation of octanol using BAIL 3a as catalysta
O
O
BAILs
pKa1
pKa2
3a
IL
OC8H17
Aldose
+
2
2.25
2.37
2.46
2.58
1.70
1.83
9.69
8.34
8.29
8.78
H2O/octan-1-ol, 80°C
OC8H17
3a
4a
5a
HTFSI
7
HO
OH
n
n
B
A
/Abc
4.80
Entry
Carbohydrate
Yieldb
A:B
Aa
Ba
/Bbc
1
2
3
4
5
6
7
8
9
14d,e
35f
20g
69
75:25
100:0
60:40
100:0
80:20
80:20
70:30
75:25
15:85
1/1.1
1.5/1
2.2/1
1/1.5
/
D
D
D
D
D
D
D
-Glucose
-Glucose
-Galactose
-Mannose
-Fucose
1/2.5
/
H
15.6/1
2/1
N
N
N
NH
79
1/2
N
+
HTFSI
NTf2
N
N
N
52
1/1
1.3/1
1/3
-Xylose
H
68
2/1
-Arabinose
6
7 (>99%)
62
2.2/1
1.8/2
1/3.5
6.1/1
L
-arabinose
-Ribose
62
D
Scheme 4.
a
Typical procedure: Sugar, octanol (5 equiv), H2O (2 equiv), and BAIL (10 mol %)
were stirred at 80 °C. After 24 h, water was eliminated by evaporation and the
reaction mixture was purified by flash chromatography.
b
Table 5
Isolated yield.
Determined by NMR.
13% yield with 5 equiv of water.
14% yield using BAIL 4a.
Glycosylation of octan-1-ol using BAIL 7 (10 mol %) as catalysta
c
d
O
O
e
OC8H17
BAIL 7
+
f
Aldose
Yield was obtained when BAIL was used in 30 mol %.
g
H2O, octan-1-ol
80°C, 24 h
OC8H17
18% yield with 4 equiv of water.
HO
OH
n
n
A
B
decide about their possible application as reaction media, espe-
cially for the Fischer’s glycosylation. The pKa values of the BAILs
2, 3a, 4a, and 5a (Table 4) have been determined by titration with
NaOH in water. As one might expect, all the ILs showed 2 pKa val-
ues. pKa1 were ranging from 2.25 to 2.58. The strongest acid in the
series is the IL 2 (pKa1 = 2.25). Meanwhile, the second deprotona-
tion step (pKa2) for all the BAILs were ranging from 8.29 to 9.69.
A direct correlation between the acidity of the synthesized ILs
2–-5a and the yield of octyl mannoside (Table 1) showed that
the acidity of the alanine derivative 5a (pKa1: 2.58) led to a lower
yield (46%) than for the more acidic IL 2, 3a, and 4a which led to
62–69% yield (pKa1: 2.25–2.46). Because no direct correlation can
be established between the experimental results obtained so far
and the different pKa values, we nonetheless believe that a stronger
pKa value could increase the hydrogen-bonding interaction be-
tween the cation of IL and the sugar starting material. In addition,
the chirality of the synthesized 2–5a seems to have a small impact
Entry
Aldose
Yieldb
A:Bc
A
a
/Abc
B
a
/Bbc
1
2
3
4
5
6
54
40
19
68
64
72
100:0
100:0
73:27
94:6
92/8
/
/
D
D
D
D
D
D
-Mannose
-Glucose
-Galactose
-Xylose
70/30
70/30
64/36
66/34
64/36
31/69
44/56
22/78
86/14
72:28
17:83
-Arabinose
-Ribose
a
Typical procedure: sugar, octanol (5 equiv), H2O (2 equiv), and BAIL (10 mol %)
were stirred at 80 °C. After 24 h, water was eliminated by evaporation and the
reaction mixture was purified by flash chromatography.
b
Isolated yield.
Determined by NMR.
c
Yield of the reaction involving D-xylose was slightly enhanced with
BAIL 7 (68%) in comparison with BAIL 3a (52%, Table 3, entry 6).
Keeping in mind that pKa1 value of 7 (1.83) seems to favourably
affect the formation of O-glycoside, we decided to undertake the
synthesis of two new analogs of 7 with a longest alkyl chain in or-
der to study the influence of their lipophily and their impact on the
O-glycosylation reactions. Alkyl chains ranging from 7 to 11 car-
bons were installed at the position 5 of BAIL 7. Syntheses of BAILs
10 and 13 are depicted in Scheme 5.
BAILs 10 and 13 were prepared in three steps starting from
commercially available alkylnitrile derivatives 8 and 11. Firstly,
they were converted into the corresponding tetrazoles 9 and 12
in 99% and 87% yields, respectively, by reaction with a mixture of
trimethylsilyl azide and dibutyltin oxide in toluene. Protonation
of 9 and 12 by HTFSI led quantitatively to BAILs 10 and 13. TGA
curves showed that ILs 10 and 13 are stable until 200 and 210 °C,
respectively.
both on the stereoselectivity (a/b) and the furanose/pyranose pro-
portion. We then decided to undertake the synthesis of a more
acidic derivative and unsubstituted tetrazolic BAIL 7 (Scheme 4).
Starting from commercially available tetrazole 6, a metathesis ex-
change between 6 with HTFSI led to the new BAIL 7 in a quantita-
tive yield with an acidity (pKa1 = 1.83, Table 4) being comparable
with HTFSI (pKa = 1.90, Table 4). The pKa2 value (4.80) corresponds
approximately to the pKa of the corresponding tetrazole without
considering any counter-anion effects of TFSIꢀ. TGA curve revealed
that ILs 7 is stable until 180° (degradation of the tetrazolium part
mainly into N2).
BAIL 7 was then studied in the direct O-glycosylation reaction.
Compared to the others BAILs, 7 led to the octyl mannoside deriv-
ative (Table 5, entry 1) in a similar yield. A slight modification has
been observed for D-glucose (entry 2) where the glycoside was ob-
tained in 40% yield in the lone pyranoside form whereas in Table 3,
the glycoside has been obtained in 14% yield as a mixture of pyra-
nose and furanose.
2.3. Determination of pKa
Similar results were expected with
-arabinose and -ribose, but the yield could not be improved com-
pared to the result obtained in Table 3 (entries 3, 5, 7, 8, and 9).
D
-galactose,
D
-fucose,
D
- and
The pKa values of the BAILs 7, 10, and 13 (Table 6) have been
determined by titration with NaOH in EtOH. As expected and ob-
served for BAILs 2, 3a, 4a, and 5a, all the ILs showed 2 pKa values.
L
D

16
S. Delacroix et al. / Carbohydrate Research 381 (2013) 12–18
TMSN3 (2 equiv.),
Bu2SnO (1 equiv.)
N
N
N
N N
HTFSI (1,10 equiv.)
CN
n
N
TFSI
N
H
N
Toluene, 100°C, 24 h
H2O, 50 or 74 °C,
2 or 3 h
n
n
H2
8
9
10
13
(n=6)
(n=6, >99 %)
(n=10, 87 %)
(n=6, >99 %)
(n=10, >99 %)
11
12
(n=10)
Scheme 5.
Table 6
pKa values of BAILs
Table 8
Glycosylation of alcohols using BAIL 13 as catalysta
BAILs
pKa1
pKa2
Alcohol
Yieldb
a
/bc
7
10
13
0.20
0.34
0.53
5.45
6.14
5.11
n-Butan-1-ol
19
67
66
58
42
36
31
Traces
90/10
93/7
92/8
90/10
96/4
n-Heptan-1-ol
n-Decan-1-ol
n-Dodecan-1-ol
Cyclohexanol
Allyl alcohol
88/22
95/5
Propargyl alcohol
Benzyl acohol
Table 7
n.d.d
Screening of tetrazole-derived BAILsa
a
HO
OH
Typical procedure: sugar, octanol (5 equiv), H2O (2 equiv), and BAIL (10 mol %)
HO
OH
were stirred at 80 °C. After 24 h, water was eliminated by evaporation and the
O
ILs (10 mol%)
H2O, 80°C, 24h
O
HO
HO
reaction mixture purified by flash chromatography.
+ octanol
HO
b
HO
Isolated yield.
Determined by NMR.
Not determined.
OH
c
OC8H17
d
Entry
BAIL
Yeildb (%)
a
/bc
1
2
3
7
10
13
54
64
62
92/8
92/8
91/9
Table 9
Glycosylation of octan-1-ol using BAIL 10 (10 mol %) as catalysta
a
O
Typical procedure:
D-mannose, octanol (5 equiv), H2O (2 equiv), and IL
O
OC8H17
10
BAIL
(10 mol %) were stirred at 80 °C. After 24 h, water was eliminated by evaporation
+
Aldose
and the reaction mixture was purified by flash chromatography.
H2O, octan-1-ol
80°C, 24 h
OC8H17
b
HO
OH
Isolated yield.
Determined by NMR.
n
n
c
A
B
Yieldb
A:Bc
A
a
/Abc
B
a
/Bbc
pKa1 were ranging from 0.20 to 0.53. The strongest acid in the ser-
ies is IL 7 (pKa1 = 0.20). Meanwhile, the second deprotonation step
(pKa2) for all the BAILs were ranging from 5.11-6.14.
Entry
Aldose
1
2
3
4
5
6
64
53
22
72
76
72
100:0
100:0
80:20
94:6
92/8
/
/
D
D
D
D
D
D
-Mannose
-Glucose
-Galactose
-Xylose
72/28
74/26
66/34
67/33
67/33
Compounds 10 and 13 were studied in the reaction model
21/79
50/50
25/75
84/16
involving
D-mannose and octan-1-ol (Table 7). Following our
hypothesis, 10 which is structurally similar to octan-1-ol affords
the O-glycoside derivative with 10% yield higher than tetrazole 7.
BAIL 13 with the longest alkyl chain slightly decreases the yield
by 62% suggesting that the reaction can be improved if the alcohols
and the BAILs are structurally close in terms of the alkyl chain
length. Moreover the results are opposed to pKa1 values suggesting
that strong pKa1 is necessary to optimize the reaction but cannot
rule the O-glycosylation by itself. Interactions between the nature
of the alcohol and the BAIL must be considered. In Table 1, with
close pKa1, BAIL 5a which is probably less lipophilic clearly shows
a yield drop of about 10% compared to the more lipophilic 2 and
3a/4a. In this series, at similar pKa, the more lipohilic BAIL (3a/
4a > 2 > 5a) gave better results.
76:24
15:85
-Arabinose
-Ribose
a
Typical procedure: sugar, octanol (5 equiv), H2O (2 equiv), and BAIL (10 mol %)
were stirred at 80 °C. After 24 h, water was eliminated by evaporation and the
reaction mixture was purified by flash chromatography.
b
Isolated yield.
Determined by NMR.
c
(Table 9). Interestingly, compared with BAIL 3a (Table 3), BAIL 10
allowed to increase yields in every glycosylation reaction. No dif-
ferences were observed concerning the
the pyranose/furanose form with this catalyst. In addition, BAIL
10 permitted to improve yields in every case compared with the
results obtained with BAIL 7 (Table 5).
a/b anomer ratios and
Next, BAIL 13 with the longest alkyl chain was screened in the
reaction between
our delight, the fatty alcohols heptan-, decan-, and dodecan-1-ol
gave better results (58–67% yield) with a classical -selectivity ru-
D-mannose and different alcohols (Table 8). To
3. Conclusions
a
led under thermodynamic control. Interestingly, the yield dropped
down to 19% for the reaction with butan-1-ol, highlighting the
importance of interactions between the alkyl chain of BAIL 13
and alcohols.
A series of aminotetrazoles and alkyltetrazoles have been
synthesized and used in catalytic quantity as BAIL in water for the
direct O-glycosylation of unreactive and unprotected sugars (pen-
toses and hexoses). The chirality broughtabout by 3a and 4a affected
Surprisingly, allylic alcohol, miscible in water, which showed no
reactivity with BAIL 3a (Table 2) allowed the formation of 36% of
the corresponding glycosyl derivative with BAIL 13 (See Table 8).
Finally, the O-glycosylation of octan-1-ol with different sugars
was re-examined but in this case hexyltetrazole 10 was used
the reactivity less in terms of yield and selectivity (a/b or pyranose/
furanose). The second series of alkyltetrazoles showed that the reac-
tivity can be enhanced when the alkyl chain length of the IL and the
alcohol alkyl chain are similar. Attempts to recycle the catalyst were
unsuccessful. In fact, after flash chromatography, tetrazoles were

S. Delacroix et al. / Carbohydrate Research 381 (2013) 12–18
17
+
the only recovered compounds. Modification of the anion could
[M+] = 140 m/z, [Mꢀ] = 279.6 m/z; HRMS calcd for C5H10N5
:
probably modifythe reactivity ofthe alkyltetrazole. Others reactions
such as esterification catalyzed by BAIL would also be tested. This
work is in progress and will be reported in due course.
140.0936, found: 140.0949. Contains 0.90% of water.
4.2.2. [(S)-2-Phenyl-1-(1H-tetrazol-5-yl)ethanaminium][NTf2]
(3a)
Following the general method A, (S)-2-phenyl-1-(1H-tetrazol-
5-yl)ethanamine (1.01 g, 5.33 mmol, 1.00 equiv) and HTFSI
4. Experimental part
3
(1.87 g, 5.86 mmol, 1.10 equiv) were dissolved in 25 mL of water
4.1. Materials and methods
to afford compound 3a (2.48 g, >99%) as a slightly yellow solid.
½
a 2D0
ꢁ
+17° (c = 0.13, CH3OH); IR (ATR):
1603, 1515, 1495, 1369, 1325, 1181, 1125, 1053, 795, 743, 701;
T°g: ꢀ7.3 °C; T°c: ꢀ; T°m: 90 °C;
95 °C = 2740 cP; 1H NMR (CD3OD,
m
(cmꢀ1) 3136, 1653, 1628,
Optical rotations were recorded in CH3OH or CH2Cl2 solution on
a
digital micropolarimeter Perkin Elmer 343™. 1H NMR
g
(300.13 MHz) and 13C NMR (75.47 MHz) spectra were recorded
in CD3OD, D2O, or CDCl3 (internal Me4Si) respectively, on a Bru-
kerTM Avance-300 spectrometer. TLC was performed on Silica
F254 and detection by UV light at 254 nm or by charring with p-
anisaldehyde–H2SO4–AcOH–EtOH reagent. FTIR spectra were ob-
tained on a AVATARTM 320 neat using ATR and are reported in
cmꢀ1. Mass spectral data were acquired on a WATERS Micromass™
Q-TOFF spectrometer. Viscosities were recorded on a BrookfieldTM
viscosimeter DVII + Pro. TGA experiments were carried out on a
NetzschÒ STA 449 C JupiterÒ coupled with a quadrupole mass
spectrometer QMS 403 C AëlosÒ. Samples were placed in a
PtꢀRh + Al2O3 crucible and then heated up to 1200 °C under air
or under an atmosphere of Ar, at 10 K minꢀ1. To determine Tg, Tc,
and Tm, DSC experiments were carried out on a NetzschÒ DSC
204F1 heat flux differential calorimeter at a heating rate of
10 K minꢀ1, from ꢀ100 to 100 °C (cycle done two times), under a
constant argon flow of 200 mL minꢀ1. Calibration was performed
with indium. Data processing was performed with Proteus soft-
wareÒ. When Tg were detected, they were measured at the inter-
section of the two tangents of the second heating curve at the
start of the endothermic phenomenon. Column chromatography
was effected on Silica Gel 60 (230 mesh) or using Combiflash com-
panionÒ/TS with cyclohexane, ethyl acetate and methanol, distilled
before use.
300 MHz): d (ppm) 7.03–7.37 (m, 5H, Ph), 5.10 (s, NH2), 5.02 (t,
1H, J3CH,CH2 = 7.6 Hz, CH), 3.40 (d, 2H, J3CH2,CH = 7.6 Hz, CH2); 13C
NMR (CD3OD, 75 MHz): d (ppm) 159.4 (NC@N), 136.3, 130.5,
130.2, 129, 127.6 (Ph), 121.3 (q, J1C,F = 320 Hz, 2 ꢂ CF3), 49.3 (CH),
40.0 (CH2); MS (ESI): [M+] = 189.7 m/z, [Mꢀ] = 279.8 m/z; HRMS
calcd For C9H12N5+: 190.1093, found: 190.1102. Contains 1.30% of
water.
4.2.3. [(R)-2-Phenyl-1-(1H-tetrazol-5-yl)ethanaminium][NTf2]
(4a)
Following the general method A, (R)-2-phenyl-1-(1H-tetrazol-
5-yl)ethanamine
4 (0.41 g, 2.15 mmol, 1.00 equiv) and HTFSI
(1.87 g, 2.43 mmol, 1.13 equiv) were dissolved in 15 mL of water
to afford compound 4a (2.48 g, >99%) as a slightly yellow solid.
½
a 2D0
ꢁ
ꢀ15° (c = 0.13, CH3OH); IR (ATR):
1629, 1603, 1515, 1495, 1370, 1325, 1181, 1125, 1053, 796, 743,
95 °C = 2740 cP; 1H NMR (CD3-
m
(cmꢀ1) 3125, 1652,
701; T°g: 3.4 °C; T°c: ꢀ; T°m: 85 °C;
g
OD, 300 MHz): d (ppm) 7.09–7.43 (m, 5H, Ph), 5.13 (s, NH2), 5.02 (t,
1H, J3CH,CH2 = 7.6 Hz, CH), 3.40 (d, 2H, J3CH2,CH = 7.6 Hz, CH2); 13C
NMR (CD3OD, 75 MHz): d (ppm) 159.3 (NC@N), 135.2, 130.4,
130.2, 129.0 (Ph), 121.2 (q, J1C,F = 320 Hz, 2 ꢂ CF3), 49.3 (CH), 40
(CH2); MS (ESI): [M+] = 189.7 m/z, [Mꢀ] = 279.8 m/z; HRMS calcd.
For C9H12N5+: 190.1093, found: 190.1101. Contains 0.58% of water.
4.2.4. [(S)-1-Methyl-1-(1H-tetrazol-5-yl)ethanaminium][NTf2]
(5a)
4.2. General procedure
Following the general method A, (S)-1-(1H-tetrazol-5-yl)ethan-
4.2.1. Method A: Synthesis of Brønsted acidic ionic liquids
(BAILs) by protonation of tetrazole derivatives with HTFSI
1.10 Equiv of HTFSI (90% wt. in water) was added to an aqueous
amine
5 (0.17 g, 1.50 mmol, 1.00 equiv) and HTFSI (0.49 g,
1.58 mmol, 1.05 equiv) were dissolved in 7 mL of water to afford
5a (0.59 g,>99%) as a slightly yellow solid. ½a D20
ꢀ4° (c = 0.15, CH3-
ꢁ
OH); IR (ATR):
m
(cmꢀ1) 3561, 3163, 1652, 1520, 1343, 1184, 1128,
solution of tetrazole-substituted a-amino acid, 1H-tetrazole-1-ium
1052, 792, 743, 654; T°g: ꢀ44.3 °C; T°c: ꢀ; T°m: ꢀ;
g
25 °C = n.d.; 1H
or 1H-5-alkyled tetrazole-1-ium (1.00 equiv). After stirring at least
30 min at rt, water was eliminated by freeze drying to afford pure
ionic liquid.
NMR (CD3OD, 300 MHz): d (ppm) 5.10 (s, NH2), 4.90 (q, 1H, CH, J3-
CH,CH3 = 6.9 Hz), 1.75 (d, 3H, CH3, J3CH3,CH = 7.0 Hz); 13C NMR (D2O,
75 MHz): d (ppm) 156.2 (NC@N), 118.9 (q, J1C,F = 320 Hz, 2 ꢂ CF3),
42.2 (CH), 16.8 (CH3); MS (ESI): [M+] = 114.1 m/z, [Mꢀ] = 280 m/z;
HRMS calcd. for C3H8N5+: 114.0780, found 114.0786. Contains
1.05% of water.
4.2.2. [(S)-2-(1H-Tetrazol-5-yl)pyrrolidinium][NTf2] (2)
4
3
5
2
1
N
4.2.5. [1H-Tetrazol-1-ium][NTf2] (7)
N
N
H2
Tf2N
Following the general method A, 1H-tetrazole 6 (CAS number:
288-94-8) (0.50 g, 7.00 mmol, 1.00 equiv) and HTFSI (2.40 g,
7.70 mmol, 1.10 equiv) were dissolved in 10 mL of water to afford
HN
N
Following the general method A, (S)-5-(pyrrolidin-2-yl)-1H-tetra-
zole 129 (1.00 g, 7.19 mmol, 1.00 equiv) and HTFSI (2.53 g,
7.91 mmol, 1.10 equiv) were dissolved in 25 mL of water to afford
compound 2 (3.00 g, >99%) as a slightly orange transparent liquid.
BAIL 7 (2.45 g, >99%) as a yellow transparent liquid. IR (ATR):
m
(cmꢀ1) 3114, 2893, 1539, 1340, 1317, 1186, 1119, 1052, 792, 743,
653, 640; T°g: ꢀ49.6 °C; T°c: ꢀ; T°m: ꢀ;
g
25 °C = 1500 cP; 1H NMR
½
a 2D0
ꢁ
+17° (c = 0.13, CH3OH); IR (ATR):
m
(cmꢀ1) 3178, 2362, 2337,
(CD3OD, 300 MHz): d (ppm) 9.48 (s, 1H, H-5), 7.90 (s, NH); 13C
NMR (CD3OD, 75 MHz): d (ppm) 143.4 (C-5), 120.8 (q, J1C,F = 320.1 -
Hz, 2 ꢂ CF3); MS (ESI): [M+] = 71.1 m/z, [Mꢀ] = 279.8 m/z; HRMS
calcd for CH3N4+: 71.0358, found 71.0343. Contains 1.75% of water.
1653, 1344, 1181, 1128, 1053, 791, 742, 653; T°g: ꢀ38.3 °C; T°c:
ꢀ; T°m: ꢀ;
g
25 °C = 9099 cP; 1H NMR (D2O, 300 MHz): d (ppm)
9.56 s, 1H, NH), 9.15 (s, 1H, NH), 5.96 (s, 1H, NH), 5.11 (t, 1H, J3-
2,CH2 = 7.6 Hz, H-2), 3.42-3.67 (m, 2H, H-5a, H-5b), 2.53–2.75 (m,
1 H, H-3a), 2.12–2.43 (m, 3 H, H-3b, H-4a, H-4b); 13C NMR (D2O,
75 MHz): d (ppm) 158.2 (NC@N), 120.9 (q, J1C,F = 320 Hz, 2 ꢂ CF3),
54.8 (C-2), 47.1 (C-5), 30.5 (C-3), 24.2 (C-4); MS (ESI):
4.2.6. [5-Heptyl-1H-tetrazol-1-ium][NTf2] (10)
Following the general method A, 5-heptyl-1H-tetrazole 930
(0.50 g, 2.97 mmol, 1.00 equiv) and HTFSI (1.02 g, 3.03 mmol,

18
S. Delacroix et al. / Carbohydrate Research 381 (2013) 12–18
4. Zhao, G.; Jiang, T.; Gao, H.; Han, B.; Huang, J.; Sun, D. Green Chem. 2004, 6, 75–
77.
5. Wasserscheid, P.; Sesing, M.; Korth, W. Green Chem. 2002, 4, 134–138.
6. Xu, D.-Q.; Yang, W.-L.; Luo, S.-P.; Wang, B.-T.; Wu, J.; Xu, Z.-Y. Eur. J. Org. Chem.
2007, 2007, 1007–1012.
7. Cole, A. C.; Jensen, J. L.; Ntai, I.; Tran, K. L. T.; Weaver, K. J.; Forbes, D. C.; Davis, J.
H. J. Am. Chem. Soc. 2002, 124, 5962–5963.
8. Zhu, H.-P.; Yang, F.; Tang, J.; He, M.-Y. Green Chem. 2003, 5, 38–39.
9. Fraga-Dubreuil, J.; Bourahla, K.; Rahmouni, M.; Bazureau, J. P.; Hamelin, J. Catal.
Commun. 2002, 3, 185–190.
10. Hanessian, S.; Lou, B. Chem. Rev. 2000, 100, 4443–4464.
11. Park, T.-J.; Weïwer, M.; Yuan, X.; Baytas, S. N.; Munoz, E. M.; Murugesan, S.;
Linhardt, R. J. Carbohydr. Res. 2007, 342, 614–620.
12. Munoz, F. J.; Andre, S.; Gabius, H.-J.; Sinisterra, J. V.; Hernaiz, M. J.; Linhardt, R.
J. Green Chem. 2009, 11, 373–379.
1.10 equiv) were dissolved in 10 mL of water to afford BAIL 10
(1.33 g, >99%) as a slightly yellow syrup. IR (ATR):
(cmꢀ1) 3521,
3124, 2934, 2861, 1608, 1347, 1198, 1136, 1059, 794, 743, 656;
m
T°g: ꢀ62.7 °C; T°c: ꢀ; T°m: ꢀ;
g
25 °C = 689 cP; 1H NMR (CD3OD,
300 MHz): d (ppm) 5.30 (s, NH), 3.03 (t, 2H, J3CH2,CH2 = 7.8 Hz,
@CCH2), 1.70–1.89 (m, 2 H, CH2), 1.20–1.47 (m, 6 H, 3 x CH2),
0.88 (t, 3H, J3CH3,CH2 = 6.8 Hz, CH3); 13C NMR (CD3OD, 75 MHz): d
(ppm) 157.7 (C5), 121.2 (q, J1C,F = 320.5 Hz, 2 ꢂ CF3), 32.8, 30.0,
29.8, 28.0, 23.9, 23.6 (6 ꢂ CH2), 14.4 (CH3); MS (ESI):
[M+] = 169.0 m/z, [Mꢀ] = 279.7 m/z; HRMS calcd. for C8H17N4
169.14477, found 169.14487. Contains 1.18% of water.
:
+
13. Auge, J.; Sizun, G. Green Chem. 2009, 11, 1179–1183.
14. Rencurosi, A.; Lay, L.; Russo, G.; Caneva, E.; Poletti, L. J. Org. Chem. 2005, 70,
7765–7768.
15. Galan, M. C.; Brunet, C.; Fuensanta, M. Tetrahedron Lett. 2009, 50, 442–445.
16. Sasaki, K.; Nagai, H.; Matsumura, S.; Toshima, K. Tetrahedron Lett. 2003, 44,
5605–5608.
17. Sasaki, K.; Matsumura, S.; Toshima, K. Tetrahedron Lett. 2004, 45, 7043–7047.
18. Galan, M. C.; Jouvin, K.; Alvarez-Dorta, D. Carbohydr. Res. 2010, 345, 45–49.
19. Roy, D. K.; Bordoloi, M. J. Carbohydr. Chem. 2008, 27, 300–307.
20. Bornaghi, L. F.; Poulsen, S.-A. Tetrahedron Lett. 2005, 46, 3485–3488.
21. Roy, B.; Mukhopadhyay, B. Tetrahedron Lett. 2007, 48, 3783–3787.
22. Mamidyala, S. K.; Finn, M. G. J. Org. Chem. 2009, 74, 8417–8420.
23. Euzen, R.; Guégan, J.-P.; Ferrières, V.; Plusquellec, D. J. Org. Chem. 2007, 72,
5743–5747.
24. Manabe, K.; Iimura, S.; Sun, X.-M.; Kobayashi, S. J. Am. Chem. Soc. 2002, 124,
11971–11978.
25. Cobb, A. J. A.; Shaw, D. M.; Longbottom, D. A.; Gold, J. B.; Ley, S. V. Org. Biomol.
Chem. 2005, 3, 84–96.
26. Ibrahem, I.; Zou, W.; Engqvist, M.; Xu, Y.; Córdova, A. Chem. Eur. J. 2005, 11,
7024–7029.
27. (a) Hartikka, A.; Arvidsson, P. I. Eur. J. Org. Chem. 2005, 2005, 4287–4295; (b)
Sundén, H.; Ibrahem, I.; Eriksson, L.; Córdova, A. Angew. Chem., Int. Ed. 2005, 44,
4877–4880; (c) Chowdari, N. S.; Ahmad, M.; Albertshofer, K.; Tanaka, F.;
Barbas, C. F. Org. Lett. 2006, 8, 2839–2842; (d) Hansen, H. M.; Longbottom, D.
A.; Ley, S. V. Chem. Commun. 2006, 4838–4840; (e) Jiao, P.; Kawasaki, M.;
Yamamoto, H. Angew. Chem., Int. Ed. 2009, 48, 3333–3336; (f) Nielsen, M.;
Jacobsen, C. B.; Paixa~o, M. R. W.; Holub, N.; Jørgensen, K. A. J. Am. Chem. Soc.
2009, 131, 10581–10586; (g) Odedra, A.; Seeberger, P. H. Angew. Chem., Int. Ed.
2009, 48, 2699–2702.
28. Torii, H.; Nakadai, M.; Ishihara, K.; Saito, S.; Yamamoto, H. Angew. Chem., Int. Ed.
2004, 43, 1983–1986.
29. Demko, Z. P.; Sharpless, K. B. J. Org. Chem. 2001, 66, 7945–7950.
30. Fürmeier, S.; Metzger, J. O. Eur. J. Org. Chem. 2003, 2003, 885–893.
4.2.7. [5-undecyl-1H-tetrazol-1-ium][NTf2] (13)
Following the general method A, 5-undecyl-1H-tetrazole 1230
(0.20 g, 0.89 mmol, 1.00 equiv) was heated at 80 °C, then HTFSI
(0.28 g, 0.89 mmol, 1.00 equiv) was added and the reaction mix-
ture was stirred for 30 min to afford BAIL 13 (0.45 g, >99%) as a
slightly yellow syrup. IR (ATR):
1607, 1468, 1347, 1200, 1135, 1060, 793, 743, 654; T°g: ꢀ; T°c:
25 °C = 729 cP; 1H NMR (CD3OD,
m
(cmꢀ1) 3522, 3111, 2928, 2857,
ꢀ30.7 °C; T°m: 11.4 °C;
g
300 MHz): d (ppm) 5.44 (s, NH), 3.01 (t, 2 H, J3CH2,CH2 = 7.8 Hz,
@CCH2), 1.70–1.90 (m, 2H, CH2), 1.17–1.47 (m, 16 H, 8 ꢂ CH2),
0.88 (t, 3H, J3CH3,CH2 = 6.8 Hz, CH3); 13C NMR (CD3OD, 75 MHz): d
(ppm) 157.7 (C5), 121.3 (q, J1C,F = 320.0 Hz, 2 ꢂ CF3), 33.1, 30.7,
30.6, 30.5, 30.2, 30.0, 28.1, 23.9, 23.8 (10 ꢂ CH2), 14.5 (CH3); MS
(ESI): [M+] = 225.1 m/z, [Mꢀ] = 279.7 m/z; HRMS calcd. for
C
12H25N4+: 225.2079, found 225.2088; Contains 0.83% of water.
Acknowledgements
A.N.V.N. thanks the Conseil Régional de Picardie for financial
support and M. Armand for fruitful discussion.
References
1. Cox, D. J.; Smith, M. D.; Fairbanks, A. J. Org. Lett. 2010, 12, 1452–1455.
2. Johnson, K. E.; Pagni, R. M.; Bartmess, J. Monatsh. Chem. 2007, 138, 1077–1101.
3. Akiyama, T. Chem. Rev. 2007, 107, 5744–5758.