Directing-protecting groups for carbohydrates. Design, conformational study, synthesis and application to regioselective functionalization

Article abstract of DOI:10.1016/j.tet.2005.04.060

A novel concept of regioselective transformation of secondary hydroxyl groups in carbohydrates is presented. First, the relative reactivity of the free hydroxyl groups of onoprotected d-glucose derivatives was assessed using acetylation as a model reactio

Full text of DOI:10.1016/j.tet.2005.04.060

Tetrahedron 61 (2005) 6839–6853
Directing-protecting groups for carbohydrates.
Design, conformational study, synthesis and application
to regioselective functionalization
a
b
Nicolas Moitessier,a,b, Pablo Englebienne and Yves Chapleur
*
a
´ ´
Department of Chemistry, McGill University, 801 Sherbrooke St. W., Montreal, Quebec, Canada H3A 2K6
´ ´ ´
Groupe SUCRES, Unite Mixte 7565 CNRS, Universite Henri Poincare-Nancy 1, B.P. 239, F-54506 Nancy-Vandoeuvre, France
b
Received 31 August 2004; revised 14 April 2005; accepted 28 April 2005
Available online 31 May 2005
Abstract—A novel concept of regioselective transformation of secondary hydroxyl groups in carbohydrates is presented. First, the relative
reactivity of the free hydroxyl groups of onoprotected D-glucose derivatives was assessed using acetylation as a model reaction. As a result,
acylation of these polyols gave a mixture of monosubstituted products in which the 3-O functionalized derivatives predominated. Novel
hydrogen bond acceptor protecting groups were next designed to modulate the 4-OH and 3-OH reactivity in the hope to mediate higher
regioselective transformations. A molecular modeling study later validated by spectroscopic analysis predicted additional intramolecular
hydrogen bonds between the hydroxyl groups and pyridyl-containing protecting groups. Taking advantage of this induced hydrogen bond
network, we achieved regioselective acetylation of the hydroxyl group at position 3 without protecting any secondary hydroxyl groups of the
carbohydrate moiety. This designed protecting/directing group increased the nucleophilicity and the steric hindrance of position 3. As a
result, optimization of the reaction conditions enabled the monoacetylation (not affected by steric hindrance) of 6-O-protected
glucopyranosides at position 3 and selective silylation (affected by steric hindrance) of position 2 in high isolated yields and
regioselectivities. This result certainly opens doors to the regioselective open glycosylation of carbohydrates.
q 2005 Elsevier Ltd. All rights reserved.
1. Introduction
Carbohydrates are biologically relevant molecules whose
potential in medicinal chemistry is still under-evaluated.
This partly stems from the lack of universal synthetic
approaches for the synthesis of oligosaccharides in a stereo-
and regiocontrolled manner. Regioselective functionaliza-
tion of the secondary hydroxyl groups has been achieved
using metals (tin,1 copper2), Lewis acids,3 or designed
bases.4 However, these efficient methods have been
restricted to specific transformations. More recently,
innovative approaches exploiting the hydrogen bond net-
work of the carbohydrate alcohols have appeared.5 In all
these approaches, the regioselective transformations rely on
the modulation of the relative reactivity of the many
hydroxyl groups of the carbohydrates. The relative reac-
tivity of the secondary hydroxyl groups depends strongly on
both their acidity and their nucleophilicity, which are
modulated by intramolecular H-bonds (Fig. 1).6 One of the
Figure 1. Intramolecular H-bond network as proposed by Yoshida and
co-workers.4,7
first reports on the evaluation of the relative reactivity of
secondary hydroxyl groups in monosaccharides appeared
recently.7 In this report, Yoshida and co-workers shed light
on intramolecular H-bonds, which control the relative
reactivity of the four hydroxyl groups (Fig. 1) and relate
the enhanced reactivity of the 3-hydroxyl group to the
hydrogen bond network.8 More recently, ab initio calcu-
lations questioned whether these hydrogen bonds exist.9
An efficient and convenient regioselective strategy would
allow regiocontrolled manipulation of the hydroxyl groups
without extensive recourse to protection/deprotection steps.
Such an approach implies the control of the relative
reactivity of the secondary hydroxyl groups of the
Keywords: Carbohydrate; Relative reactivity; Regioselectivity; Open
glycosylation; Hydrogen bond; Pyridyl ring; Acetylation.
*
Corresponding author. Tel.: C514 398 8543; fax: C514 398 3797;
e-mail: nicolas.moitessier@mcgill.ca
0040–4020/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2005.04.060

6840
N. Moitessier et al. / Tetrahedron 61 (2005) 6839–6853
carbohydrate unit. We reasoned that a protecting group at
position 6—the easiest to install—could modulate the
reactivity of the secondary hydroxyl groups. Thus, we first
focused on the effect of 6-hydroxyl protecting groups on the
relative reactivity of the free secondary hydroxyls.5a Herein,
we wish to provide a full report on the design, preparation,
solution conformation study and use of original pyridyl-
containing protecting groups, which served as a basis for
regiocontrolled transformation of monoprotected glycosyl
acceptors. Even though the regioselective glycosylation is
our primary concern, acetylation was found to be a more
convenient model reaction. Thus a variety of acceptors were
monoacetylated under kinetic conditions and the regio-
selectivity was evaluated.
hydroxyl groups was also evaluated by means of the
xylopyranoside derivative 1k. Other protecting groups were
investigated including TMS, PhSi(Me2), and CF3C(O).
However, they were found to be prone to migration or
cleavage under the acetylation conditions employed during
the course of this study.
Thus, compound 1a was subjected to regioselective
protection by standard methods including reaction with
the suitable chloride reagent either in DMF in presence
of imidazole at room temperature, in pyridine at 0 8C or
in collidine at K40 8C as reported by Yamamoto and
co-workers (Scheme 1).20 Benzylidene derivatives were
regioselectively prepared on treatment with the appropriate
aryl dimethylacetal reagents in presence of a catalytic
amount of PTSA.
2. Results and discussion
2.1. Preparation of 6-O-monoprotected
glucopyranosides
Monosaccharides are molecules with limited conformation-
al freedom. Thus the shape of the substrates should not be
affected much by the functionalization of the 6-OH. In
contrast, the influence of the 6-O protecting group on the
electrostatic potential of the molecule or on the cooperative
network of intramolecular hydrogen bonds should influence
the relative reactivity of the three secondary hydroxyl
groups.
Scheme 1. (a) TrCl, pyridine, 81% (1b); (b) TBDPSCl, imidazole, DMF,
82% (1c); (c) TBSCl, pyridine, 78% (1d); (d) AcCl, collidine, K40 8C,
55% (1e); (e) (CCl3CO)2O, collidine, K20 8C, 13% (1f); (f) PhCOCl,
collidine, K10 8C, 49% (1g); (g) ArCH(OMe)2, PTSA, 81% (1h), 90% (1i),
82% (1j).
2.2. Acetylation as a model reaction
A variety of protecting groups were installed at the position
6 of methyl-a-D-glucopyranoside 1a. The protecting groups
(aromatic or aliphatic, silyl ether or ester) depicted
in Figure 2 were chosen due to their common use in
carbohydrate chemistry. In addition, these protecting groups
can be quantitatively removed by hydrogenation or under
acidic conditions. Although the removal of the protecting
group is not greatly important in the present study, it will be
crucial when glycosylation will be studied. The influence of
the hydroxymethyl group in the reactivity of the secondary
Although we were concerned by the glycosylation reaction,
analysis of the steric and electronic contributions by direct
study of the six possible isomers (regio- and stereoisomers)
would be tedious. Thus, our efforts, directed at evaluating
the relative nucleophilicity of the secondary hydroxyl
group, focused on regioselectivity of acetylation as a
model study.21 Application to the glycosylation reaction
will follow and will be reported in due course. The choice of
this reaction was dictated by the ease of evaluation of the
1
regioselectivity using routine H NMR on crude mixtures.
After usual work-up, the ratios of regioisomers were
conveniently and reliably estimated by integration of methyl
signals around 2 ppm and of either H-2, H-3 or H-4 peak
that shifted downfield due to acetylation of the adjacent
hydroxyl groups.
Among the parameters that can affect the regioselectivity of
the reaction are the reaction conditions and the chemical
nature of the protecting group at position 6. The reaction
conditions (K2CO3, Ac2O, DMAP) optimized by Yoshida
and co-workers led to a kinetic control of the reaction.4,7
Dichloromethane and THF, two solvents widely used for the
glycosylation reaction that we wish to model, were chosen22
and DMAP (5 mol%) and acetic anhydride (0.7 equiv) were
added as freshly prepared dichloromethane solutions. Low
concentrations of carbohydrate (0.5 mmol LK1) precluded
any aggregation or intermolecular interactions. However,
although the reaction proceeded smoothly in dichloro-
methane, the reaction rate was too low in THF and higher
concentrations were needed (2 mmol LK1) with 10 mol%
DMAP. No concentration effect was noticed since the ratio
Figure 2. 1a,10 1b,11 1c,12 1d,13 1e,14 1f, 1g,15 1h,16 1i,17 1j,18 1k.19

N. Moitessier et al. / Tetrahedron 61 (2005) 6839–6853
Table 1. Regioselectivity of the acetylation of the secondary hydroxyl groups
6841
Entry
Compd
Protecting
group
Solvent
2 (%)a
3 (%)a
4 (%)a
Ratio
1
2
3
4
5
6
7
8
1b
Ph3C
THF
14
13
7
68
74
80
67
74
66
68
62
69
57
56
66
62
53
55
51
55
43
75
74
18
13
13
18
14
15
16
20
12
25
27
23
—
—
—
—
—
—
17
17
1: 4.8: 1.2
1: 5.5: 1.0
1: 11.8: 1.0
1: 4.4: 1.2
CH2Cl2
CHCl3
THF
CH2Cl2
THF
CH2Cl2
THF
CH2Cl2
THF
THF
CH2Cl2
THFc
CH2Cl2
THFc
CH2Cl2
THFc
CH2Cl2
THF
CH2Cl2
1c
1d
1e
t-Bu(Ph)2Si
t-Bu(Me)2Si
MeC(O)
15
12
19
16
17
19
18
17
11
38
47
45
49
45
57
8
1: 6.0: 1.123
1: 3.5: 0.8
1: 4.3: 1.0
1: 3.6: 1.2
1: 3.8: 0.6
1: 3.6: 2.0
1: 3.1: 1.6
1: 5.9: 2.0
1: 1.6: –
9
10
11
12
13
14
15
16
17
18
19
20
1f
1g
Cl3CC(O)b
PhC(O)
1h
1i
MeOPhCH
PhCH
c
c
c
1: 1.1: –
1: 1.3: –
1: 1.0: –
1: 1.3: –
1: 0.75: –
1: 11.6: 2.0
1: 8.5: 2.0
1j
NO2PhCH
1k
1k
—
—
9
a Based on reacted material, 40–65% conversion.
b Insoluble in dichloromethane.
c Reacted for 2 h.
remained the same at concentrations of carbohydrate
ranging from 0.5 to 2 mmol LK1. Thus, the series of
pyranosides were subjected to acetylation and the results are
summarized in Table 1 with an accuracy of about 3%. All
the triols were allowed to react for one hour and conversion
rates of 40–65% were observed. Benzylidene derivatives
were found to be less reactive, requiring longer reaction
times to give similar conversions. Although this was a
purely qualitative observation, it revealed the lowest
intrinsic reactivity of this class of compounds.
THF to dichloromethane led to an increase in the amount of
3-OAc isomers when triols were reacted but to a decrease
in the 3-OAc in the benzylidene series (1h, 1i, 1j, entries
13–18). The higher regioselectivity in dichloromethane
compared to THF may stem from the postulated intra-
molecular hydrogen bonds, which are believed to be
stronger in dichloromethane. These preliminary data
prompted us to further explore the influence of the
intramolecular hydrogen bonds.
2.3. Design and synthesis of second generation protecting
groups
The data summarized in Table 1 clearly indicates that the
steric effects were negligible since 1b and 1k (entries 2 and
20) led to similar regioselectivities. Thus, the following
study will primarily focus on the differences of reactivity
between each hydroxyl group. The data also reveals that the
electronic effects, although weak, affected the regioselec-
tivity. Indeed, the 3-OAc/2-OAc ratios in either dichloro-
methane or THF correlate well with the electron-donating/
withdrawing properties of the protecting groups (1h, 1i and
1j, entries 13–18). These benzylidene derivatives (1h–1j)
exhibited a different reactivity pattern with a higher
preference for position 2 relative to the 6-O monoprotected
compounds (1b–1g).
Although this data indicates clear electronic and solvent
effects in the regioselectivity, they are not useful from a
practical viewpoint. In addition, these monoprotected
pyranosides were reacted with 0.7 equiv of acetic anhydride
and a maximum conversion of 70% is to be expected.
Reaction with a stoichiometric amount of acetic anhydride
led to mixtures of mono-, di- and tri-acetylated products
along with unreacted material. The overacetylation can be
explained by the small difference in reactivity of the
hydroxyl groups and the low hindrance around the
monoacetylated products. Hence, the next stage of our
research program involved the development of original
protecting/directing groups that would mediate regioselec-
tive functionalization of the polyols. We postulated that
hydrogen bond acceptor moieties would perturb the
hydrogen bond network and modulate the relative reactivity
of the free OH groups. During the course of our work, a
similar approach was used with glucosamine derivatives
A difference between silyl ethers (1c, 1d, entries 4–7), ether
(1b, entries 1–3) and esters (1e, 1f, 1g, entries 8–12) is
discernible, ethers and silyl ethers leading to the highest
regioselectivities. In the four examples shown in entries 8, 9,
11 and 12, the amount of 4-OAc was slightly enhanced in
THF compared to dichloromethane. Similarly, going from

6842
N. Moitessier et al. / Tetrahedron 61 (2005) 6839–6853
and led to enhancement of reactivity of the 6-OH.24 The
authors used a picolinyl protecting group which experienced
an intramolecular hydrogen bond.
The design of such groups calls for complementary spatial
positions of the H-bond acceptor and donor. Molecular
modeling was therefore instrumental at this stage. Our
previous work have demonstrated the predictive power of
molecular mechanics methods which was herein exploited
in the design of novel protecting/directing groups.25 Thus, a
series of protecting groups were virtually screened (mol-
ecular mechanics studies) and the ones which were found to
interact with OH-4 were selected for synthesis. As a result,
2-pyridyl counterparts of the trityl (the most regioselective
so far, see Table 1) and benzoyl groups were prepared
(Fig. 3) and installed onto 1a leading to 1l and 1m,
respectively. Although one can expect a role of this pyridyl
rings in the acetylation reaction, the hydrogen bond (if
strong enough) should shut down the nucleophilicity of the
nitrogen. This concern will be addressed further in this
report. To improve the effect of these selected protecting
groups (1l and 1m), other polyaromatic moieties were
prepared (1n and 1p). Finally, to assess the computational
study predictions, the thiophene derivative 1o which was
predicted to behave as compound 1b, was also prepared.
Figure 4. Proposed solution conformations of 1b and 1l and observed
coupling constants.
1
1D and 2D H NMR spectroscopic analysis in deuterated
chloroform and DMSO confirmed this hypothesis (Figs. 4
and 5). The OH peaks were first unambiguously assigned
The appropriate chloride reagents were synthesized accord-
ing to literature procedures.26 For example, 1n was prepared
from 4-dimethylaminopyridine, which was regioselectively
metallated according to Fort’s procedure26c,d and reacted
with benzophenone. The tertiary alcohol was subsequently
chlorinated and reacted with the free carbohydrate 1a.
Figure 3. 1l–1p. Tol: p-tolyl, Th: 2-thiophenyl.
2.4. Solution conformation analysis
The molecular modeling study suggested that the most
energetically favored conformations of 1l would exhibit the
postulated H-bond (Fig. 4). After synthesis, the amorphous
solid was unfortunately not suitable for X-ray diffraction
analysis.
Figure 5. 1H NMR spectra of (a) 1b, 1o, 1l, 1p in DMSO-d6, (b) 1b, 1o in
CDCl3 and (b) 1l, 1p in CDCl3 (4 mg/mL). OH-4’s are missing on the
bottom spectra (7–8 ppm).

N. Moitessier et al. / Tetrahedron 61 (2005) 6839–6853
6843
using COSY experiments. As expected, the spectra of 1b
and 1l were similar in DMSO-d6 (Fig. 5a), a solvent known
to disrupt the hydrogen bonding. Going from 1b to 1l,
characteristic shifts of the OH-4 peak (Dd (OH-4)Z
4.73 ppm) and of the H-4 signal (Dd (H-4)Z0.55 ppm)
were observed in CDCl3 (Fig. 5b and c, Dd represents the
chemical shift difference between 1b and 1l spectra). More
in-depth NMR spectra analysis also brought some infor-
mation about both the solution conformation and the H bond
network (Figs. 4 and 5).27 For instance, H-3 shifted
downfield (Dd (H-3)Z0.15 ppm). This shift was attributed
to a new hydrogen bond between the activated oxygen O-4
and OH-3 thus leading to the proposed structures shown in
Figure 4. More interestingly, the characteristic change in the
H-6 peak pattern was observed. The values for J5,6a and J5,6b
(3.5 and 8.5 Hz) measured for 1b correspond to a gt
conformation of the C-5–C-6 bond.28 The values measured
for 1l (J5,6a,J5,6bZ3.5 and 3.5 Hz) confirmed the computa-
tionally predicted conformational change and a gg confor-
mation. These NMR experiments indicate that the observed
solution conformations match well with the predicted
models.
stronger for 1n which features the more basic para-
substituted protecting group. The OH-4 peak for 1l is well
resolved for any concentration of DMSO whereas it is
spread over the spectrum for 1n and even not observed in
neat CDCl3. This broadness was attributed to the pyridinium
character of this proton. Extrapolation of the curve predicts
a chemical shift of around 9 ppm for the 4-OH, a value that
is rarely observed for a secondary alcohol. Again the role of
the DMAP ring in the acetylation will have to be assessed.
Although we succeeded in inducing an hydrogen bond in 1l,
any attempt to remove the protecting group either by
standard methods (TFA, formic acid) or more recent
procedures such as acetyl chloride,29 cerium chloride/
sodium iodide30 failed. Only hydrogenolysis (H2, 30 bar,
10% Pd/C) slowly removed this trityl-like group. The low
yielding introduction of the DMAP-containing group in 1n
was also a limitation to its practical use. These observations
called for other protecting groups, which maintain hydrogen
bond acceptor properties while being easy to prepare,
introduce and remove. The deprotection issue could be
addressed by substituting the protonatable pyridyl ring for a
non-basic thiophene or furan ring, which would also act as
a hydrogen bond acceptor, though weaker. However, the
computational study predicted no hydrogen between these
heterocycles and the carbohydrate OH’s. To validate these
predictions, 1o was selected as a negative reference and
prepared following the same strategy as for 1l. As expected,
the similarity of the 1H NMR spectra of 1b and 1o indicated
that the desired H bond did not occur (Fig. 5b). The
preparation of the corresponding furan analog appeared to
be low yielding. Keeping the nitrogen was therefore
essential for the success of our approach. The problematic
stability was tackled by adding electron-donating groups on
the phenyl rings. Two methoxy groups were introduced
however providing a fairly unstable dimethoxytrityl-like
moiety. Introducing two methyl groups was a more
successful strategy affording an appropriately stable ether
bond (1p, Fig. 3). Gratifyingly, the hydrogen bond was
observed (Fig. 5c) and this protecting group was easily
removed by a solution of TFA in dichloromethane.
Careful examination of the 1H NMR spectra also suggested
that substituting the phenyl ring in 1g (RZPhCO) for a
pyridyl (1m, RZPyrCO) led to a weak intramolecular
H-bond. The small shift of the H-4 peak and the broadness
of the OH peaks were presumably the result of a
perturbation of the H-bond network by the pyridyl group
or the formation of aggregates.
In an attempt to further increase the hydrogen bond strength,
the pyridine nitrogen was electron-enriched by a dimethyl-
amino group (1n). The recorded spectra in DMSO and
CDCl3 were similar to those of 1l. For the sake of
quantitative comparison, NMR solvent titration was next
carried out (Fig. 6). As can be observed in Figure 6, the
hydrogen bond observed for 1l is strong and a substantial
amount of DMSO is required to disrupt it. It appears even
2.5. Design, synthesis and solution conformation analysis
of third generation protecting groups
An additional protecting group (1q, Fig. 7) was further
designed that would participate in two intramolecular
hydrogen bonds. In order to differentiate between hydrogen
bond and steric effect, 1r was also prepared.
Figure 6. Chemical shift dependence of OH protons as a function of solvent
composition for 1l (a) and 1n (b).
Figure 7. Compounds 1q and 1r.

6844
N. Moitessier et al. / Tetrahedron 61 (2005) 6839–6853
and 9b. Finally, catalytic hydrogenation and subsequent
chlorination of the tertiary alcohols 10a and 10b led to 11a
and 11b, respectively. Prior protonation of both pyridyl
rings was necessary to avoid hydrolysis of these reagents.
The trityl chloride-like compounds 11a and 11b were finally
reacted with 1a in pyridine yielding the intended target
compounds 1q and 1r.
To further validate the presence of the postulated hydrogen
bonds, the dependence of the solvent composition on the
chemical shift was studied. From comparison of Figure 8c
with Figure 6a, we hypothesized that the hydrogen bond
with OH-4 still exists in 1q and seems to be approximately
as strong as in 1l. The NMR titration curve for OH-3 reveals
the presence of the expected additional hydrogen bond,
although weaker than the hydrogen bond with OH-4.
2.6. Role of the pyridyl ring in the acetylation reaction
Prior to the use of these protecting groups, we confirmed the
existence of the hydrogen bond in dichloromethane used as
a solvent for the acetylation reaction. The similarity
between the 1H NMR spectra in deuterated dichloromethane
and chloroform confirmed the strong hydrogen bond
between OH-4 and the pyridyl ring of 1l (Fig. 9). Thus,
the data obtained through extensive NMR studies carried
out in deuterated chloroform can be transferred to the
reactions in dichloromethane.
Scheme 2. (a) TolMgBr, THF, rt then TolMgBr, K78 8C, 45%;
(b) 2-picoline, n-BuLi, THF, K78 8C then 6, HMPT, 83% (7a);
(c) toluene, t-BuLi, THF, K78 8C then 6, HMPT, 35% (7b, along with
30% 6); (d) H2, 10% Pd/C, EtOH/THF, 60% (8a); (e) LiAlH4, THF, 0 8C,
70% (8b); (f) NaHMDS, THF, K78 8C, then PhNTf2, K50 8C, 76% (9a),
83% (9b); (g) H2, 10% Pd/C, EtOH/THF, 72% (10a), 94% (10b); (h) HCl,
H2O, reflux then SOCl2, AcCl; (i) pyridine, 1a, 63% (1q, two steps along
with 23% of 10a), 50% (1r, two steps, along with 29% of 10b). Tol: p-tolyl.
In order to rule out any catalytic role of the pyridyl-
containing protecting groups, reactions were carried out in
absence of DMAP (Table 2, entries 3, 9 and 18). The lack of
reaction demonstrates that the hydrogen bond is strong
enough to prevent the pyridyl ring from reacting with acetic
anhydride or that the pyridyl rings are poor nucleophiles.
Indeed, the pyridyl ring is known to be a much weaker
catalyst than dimethylaminopyridine (DMAP) and is
unlikely competing with DMAP.31 Reactions in CD2Cl2
monitored by 1H NMR confirmed the lack in reactivity of 1l
with acetic anhydride as well as the poor catalytic properties
of 2-picoline used as a model for the pyridyl-containing
groups. Even with 2 equiv of 2-picoline, compound 1l in
CD2Cl2 in presence of acetic anhydride remained
unchanged after 2 h. As a comparison, addition of a
The synthesis of 1q and 1r is illustrated in Scheme 2. It
began with two regioselective Grignard addition to the
pyridine derivative 5 followed by a third addition of
picolinyl lithium or p-tolyl lithium to the second carbonyl
group. These successive additions led to the enol 7a and the
ketone 7b, which were subsequently reduced into 10a and
10b. For this purpose, catalytic hydrogenation of 7a and
hydride reduction of 7b afforded compounds 8a and 8b
along with over-reduced products resulting from concomi-
tant cleavage of the tertiary alcohol group. Triflation
followed by in situ elimination was achieved using
PhN(Tf)2 as a mild triflating reagent, yielding olefins 9a
Figure 8. (a) NMR spectra of 1q (bottom) and 1r (top) in CDCl3 (dilution: 4 mg/mL), (b) Proposed modeled structure and (c) OH proton chemical shift vs.
DMSO/CDCl3 composition for 1r.

N. Moitessier et al. / Tetrahedron 61 (2005) 6839–6853
6845
Figure 9. 1H NMR spectra of 1l in CD2Cl2 (dilution: 4 mg/mL).
catalytic amount of DMAP led to completion within an
hour.
compared to the data obtained with the trityl and benzoyl
derivatives 1b and 1g (Table 1, entries 1–3, 11 and 12).
From the proposed structural models, we can expect that
O-3 would be the most reactive. O-4 is both activated by
the pyridyl ring (OH-4/N bond) and deactivated by the
hydrogen bond with OH-3 (O-4/HO-3). O-3 does not
interact with any hydrogen while OH-3 interacts with O-4
2.7. Regioselectivity of the acetylation reaction
The prepared monoprotected substrates were subjected to
acetylation conditions (Table 2) and the ratios were
Table 2. Nucleophilicity of the secondary hydroxyl groups
Entry
Compd
Protecting group
Pyr(Ph)2C
Conditions
2 (%)a
3(%)a
4(%)a
Ratio
1
2
1l
THF
CH2Cl2
15
16
69
79
16
5
1: 4.6: 1.1
1: 5.0: 0.34
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
CH2Cl2, no DMAP
CH2Cl2, no DMAP, 2-picoline
CHCl3
THF
CH2Cl2
THF
CH2Cl2
CH2Cl2, no DMAP
THF
NR
NR
74b
66
80
77
76
NR
70
19b
12
6
12
15
7b
22
1: 3.9: 0.40
1: 5.4: 1.8
1: 14.3: 2.5
1: 4.6: 1.1
1: 5.0: 0.34
b
1m
1n
PyrC(O)
14
DMAP(Ph)2C
(11)c
(9)c
1o
1p
Th(Ph)2C
13
10
15
11
22b
13
11
17
15
12
6
1: 5.3: 1.3
1: 7.4: 1.4
1: 4.8: 0.8
1: 7.2: 0.5
1: 3.2: 0.35
1: 5.8: 0.84
1: 7.2: 0.66
CH2Cl2
THF
CH2Cl2
75
73
Pyr(Tol)2C
83
CHCl3
THF
CH2Cl2
71b
76
7b
11
6
b
1q
1r
Pyr-(CH2)2-Pyr(Tol)2C
Ph-(CH2)2-Pyr(Tol)2C
83
CH2Cl2, no DMAP
CHCl3
THF
CH2Cl2
b
CHCl3
NR
NR
78
81
93b
—
17
14
7b
5
5
1: 4.5: 0.30
1: 5.8: 0.38
1: 13.2: !0.1
!3b
1
a Based on recovered starting material, 40–65% conversion, measured on crude H NMR, NR: no reaction.
b Conversions of 10%; ratios are given but are not highly accurate.
c Amount of 3,4-di-O-Ac derivative (see text).

6846
N. Moitessier et al. / Tetrahedron 61 (2005) 6839–6853
Figure 10. Intramolecular catalysis.
(O-4/HO-3). In fact, the observed 3-OAc/4-OAc ratio
increased remarkably by substituting a phenyl of the trityl
group for a pyridyl ring (1b: 5.5:1, Table 1, entry 2, 1l: 14:1,
Table 2, entry 2). This significant loss of reactivity of the
4-hydroxyl group presumably arose from the observed
intramolecular hydrogen bond formed between the 4-OH
and the pyridyl ring. More surprisingly, even the DMAP-
based protecting group of 1n did not catalyze the reaction
(entry 10). In addition, when 1n was reacted in presence of a
catalytic amount of DMAP, the 3,4-di-O-acetylated com-
pound was isolated instead of the 4-OAc derivative. This
unexpected behavior was rationalized as shown in Figure
10. The strong hydrogen bond between the pyridyl ring and
4-OH prevents any catalytic action of the protecting group.
However, when 4-OH is converted into the 4-OAc, this
bond no longer exists and the protecting group acts as a
catalyst resulting in acetylation of 3-OH. This is an
additional evidence of the strength of the hydrogen bond
experienced by this designed protecting group. Similar
behaviors were not observed with the other pyridyl-
containing protecting groups which do not compete with
DMAP.
Fortunately, the reactivity was good enough in dichloro-
methane to allow for the estimation of the second pyridyl
ring effect. As can be seen in Table 2 (entries 16 and 17
vs. 13 and 14) the additional aromatic ring does not
significantly modify the regioselectivity neither in THF nor
in dichloromethane.
2.8. Regioselective acetylation, silylation and
pivaloylation on preparative scale
At this stage, we thought to check the practical usefulness of
the proposed strategy based on directing/protecting groups.
The previously proposed preparation of the protecting group
was judged to be lengthy and hazardous on a large scale.
The original synthesis of 10a was initially achieved in five
steps and only 12% overall yield. We sought alternative
reaction sequences to the synthetic pathway previously
envisaged which would provide the expected tertiary
alcohol in higher yields. Exploratory experiments were
thus conducted and other strategies were envisaged inclu-
ding the use of 2,6-dibromopyridine as a starting material.
Ultimately, the Sonogashira coupling32 was advantageously
employed and afforded the intermediate 10a in a much
higher overall yield (76% in four steps, Scheme 3). Readily
available dissymmetric pyridine derivative 1433 (prepared
by high yielding methylation of commercially available
bromopicolinic acid 13) was coupled with 2-ethynylpyri-
dine to afford dipyridine compound 15 in excellent yield.
Reduction under standard conditions followed by conden-
sation of 2 equiv of a Grignard reagent on the ester group
completed the synthesis of 10a.
The solvent effect previously observed for 1g (Table 1,
entries 11 and 12) became more pronounced for 1m (Table
2, entries 6 and 7). The two electron donating methyl groups
of 1p presumably increased the pyridyl nitrogen partial
charge resulting in a stronger hydrogen bond hence the
observed enhanced regioselectivities (entries 13 and 14 vs. 1
and 2). More surprising was the discernible loss of reactivity
of the carbohydrate hydroxyl groups in chloroform. One
hydrogen bond reduced dramatically the conversion rate
(around 10% after 1 h, entries 5, 15 and 22) while the
additional pyridyl ring in 1q completely inhibited the
reaction (entry 19). The formation of the hydrogen bonds
might result in a significant decrease of the oxygen partial
charges, hence a decrease of their nucleophilicity.
With a scalable synthesis of the protecting group in hand,
the preparative use of this protecting group could now be
evaluated. For this purpose, a slight excess of acetic
anhydride was reacted with 1q in presence of DMAP
(Scheme 4). For the comparison purpose, 1p was reacted

N. Moitessier et al. / Tetrahedron 61 (2005) 6839–6853
6847
1p (entry 4). The observed ratios confirmed the role played
by the terminal pyridyl ring in 1q. Indeed, when the reaction
was carried out at K78 8C, the amount of isolated 3p (3p/
other isomers, 2.0:1) is much lower than when 1q was
acetylated (3q/other isomers, 5.7:1).
1b was previously found to provide the 3-O isomer as the
major compound (Table 1) with 0.7 equiv of acetic
anhydride. However, when 1b was reacted with 1 or more
equivalent of acetic anhydride, a complex mixture of the
3-O isomer along with unreacted material, diacetylated and
triacetylated products was obtained. This low selectivity
precludes the use of the trityl group to achieve regioselec-
tive acetylation.
Scheme 3. a) Ref 33, quant.; (b) 2-ethynylpyridine, PdCl2(PPh3)2, CuI,
Et3N/THF 1:1, 60 8C, 96%; (c) H2, 10% Pd/C, EtOH/THF, 90%;
(d) TolMgBr, THF, 0 8C, 88%. Tol: p-tolyl.
The successful application of this methodology to the
regioselective acetylation of the hydroxyl at position 3 led
us to consider the regioselective protection of the position 2.
Although the reactivity of the hydroxyl group on position 3
has been increased by the designed protecting group, the
steric hindrance at this position has also been increased. The
acetylation was found to be unaffected by the steric
hindrance of the protecting group (Table 1, entries 2 and
20). We reasoned that a bulky reagent would be more
sensitive to the steric effects. Thus the pivaloylation and
silylation reactions were selected. Table 3 entries 5 and 6
summarize the data for these two reactions. Optimal
conditions for pivaloylation have been developed and
shown in Table 3. Although the increase of the steric
effects has been observed, the reversal of regioselectivity
between positions 2 and 3 has not been reached. Moreover,
under the same conditions (Table 3, entries 1–4). In a hope
to introduce a single acetyl group by increasing the
hydrogen bond strength, the temperature was lowered.
Under the conditions used previously, no reaction occurred.
Increasing the amount of DMAP restored a reasonable rate.
Eventually, regioisomer 3q was isolated in 70% yield along
with the regioisomer 2q (13%) and overacetylated products
(17%) (entry 1). When the reaction was carried out at
K60 8C, the preference for O-3 was further enhanced
(entry 2). Upon cooling further, the reaction was even more
selective and led to the isolation of 3q in 85% yield. It is
notable that the second major compound is the 2,3-O-
diacetylated product presumably arising from overacetyl-
ation of 3q. In order to attribute this high regioselectivity to
the protecting group, the same reaction was performed with
Scheme 4. Regioselective protection of 1q.
Table 3. Preparative scale application
Entry
Compd
Protecting group
Conditions
2 (%)a
3 (%)a
4 (%)a
Diprotectedb
c
1
2
3
4
5
6
1q
Pyr-(CH2)2-Pyr(Tol)2C
Ac2O, DMAP, K30 8C
Ac2O, DMAP, K60 8C
Ac2O, DMAP, K78 8C
Ac2O, DMAP, K78 8C
PivCl, DMAP, Et3N, K78 8C
TBSOTf, 2,6-lutidine, K78 8C
13
10
8
14
31
67
70
81
85
67
32
23
—
—
—
—
—
—
17 (13)
9 (6)
7 (5)
19 (9)
0
c
c
c
1p
1q
Pyr(Tol)2C
Pyr-(CH2)2-Pyr(Tol)2C
c
c
7
a Measured on crude 1H NMR, accuracy G2%.
b Amount of 2,3-di-O-acetyl derivative in brackets.
c Not detected.

6848
N. Moitessier et al. / Tetrahedron 61 (2005) 6839–6853
the conversion rate is low and the use of different
nucleophilic bases did not improve neither the conversion
nor the ratio. We next investigated the silylation reaction.
Again a series of nucleophilic bases including DMAP,
2,6-lutidine and imidazole were tried with either TBSCl
or TBSOTf. A single combination shown in Table 3 did
provide complete conversion. Gratifyingly, the expected
reversal of selectivity was observed and further validated
the concept of protecting/directing group. This last result
demonstrated that the steric effects can balance the higher
reactivity of position 3. Thus a small reagent would react
with position 3 (most nucleophilic) and a bulky reagent
would react with position 2 (least hindered). This is a good
indication for the coming studies on regioselective glyco-
sylation (Scheme 4).
Chemical shifts are reported in ppm using the residual of
chloroform as internal standard (7.27 ppm for 1H and
77.0 ppm for 13C, respectively). Mass spectra were recorded
on a Trio 1000 Thermo Quest spectrometer in the electron
impact mode or a Platform Micromass in the electrospray
mode. Elemental analyses were obtained on a Perkin Elmer
240C microanalyser. Analytical thin-layer chromatography
was performed on Merck 60 F254 pre-coated silica gel
plates. Visualization was performed by UV or by develop-
ment using KMnO4, H2SO4/MeOH or Mo/Ce solutions.
Preparative chromatography was performed on silica gel 60
(230–40 mesh ASTM) at increased pressure.
4.1.1. Methyl6-O-(diphenyl-(2-pyridyl)methyl)a-D-gluco-
pyranoside (1l). To a solution of methyl a-D-glucopyrano-
side 1a (5.0 g, 25.7 mmol) in pyridine (100 mL) was added
diphenyl-(2-pyridyl)methyl chloride (5.6 g, 20 mmol, pre-
pared from benzophenone and 2-pyridyl lithium as
described in the litterature26). The mixture was stirred for
48 h then concentrated. The residue was extracted with
CH2Cl2, washed with water and brine, dried over Na2SO4
and concentrated in vacuo. Purification by chromatography
(CH2Cl2/MeOH, 1:0 then 19:1) afforded compound 1l,
which was recrystallized (CH2Cl2/hexanes) (4.8 g, 55%,
white powder); RfZ0.37 (CH2Cl2/MeOH, 9:1); [a]D K3.5
(c 1.2, CHCl3); mp 132 8C; IR (neat/NaCl) 3400,
3. Conclusion
For the last 20 years or so, regioselective manipulation of
carbohydrate hydroxyl groups has been addressed with
challenging strategies. In this context, we carried out
exploratory experiments directed toward the evaluation of
the effect of the 6-O protecting group on the relative
reactivity of the secondary alcohols. With a designed
protecting group installed at position 6, we have observed
the expected decrease in reactivity of the 4-OH and a
concomitant increase in reactivity of the adjacent 3-OH. The
designed H-bond acceptor protecting groups operate by
partly modifying the intramolecular H-bond network. Thus,
we have established a plausible strategy for modulation of
the relative reactivity of the 2-, 3- and 4-OH’s consistent
with the experimentally observed hydrogen bonds. The
preparative applicability has been demonstrated and a high
yielding synthesis of the protecting group was proposed.
Furthermore, we have shown that the designed protecting/
directing group can be used to regioselectively react the
positions 2 or 3, the selectivity being tuned by the size of the
reagent.
1
1587 cmK1; H NMR (250 MHz, CDCl3) d 8.57 (d, 1H,
JZ4.5 Hz), 7.63 (dd, 1H, JZ7.5, 8.0 Hz), 7.50–7.22 (m,
11H), 7.16 (dd, 1H, JZ4.5, 7.5 Hz), 4.78 (d, 1H, JZ
3.5 Hz), 4.11 (dd, 1H, JZ9.0, 9.5 Hz), 3.85 (dd, 1H, JZ9.5,
9.5 Hz), 3.75–3.65 (m, 2H), 3.58 (dd, 1H, JZ3.5, 9.5 Hz),
(3.34 (s, 3H), 3.13 (dd, 1H, JZ3.5, 9.5 Hz), 2.95 (bs, 1H),
2.20 (d, 1H, JZ9.5 Hz); 13C NMR (65 MHz, CDCl3) d
161.8, 147.8, 143.6, 142.2, 136.3, 129.2, 128.4, 127.6,
127.3, 127.0, 123.7, 121.6, 99.4, 86.5, 73.5, 71.9, 71.0,
70.4, 64.7, 54.7; LRMS (EIC, m/z, %): 438 (11) (MCHC),
437 (13) (MC), 260 (100) (Pyr(Ph)2COC), 244 (92)
(Pyr(Ph)2CC); HRMS (m/z): [MCH]C calcd for
C25H27NO6 437.18384, found 437.18463; Anal. calcd for
C25H27NO6: C, 68.63; H, 6.22; N, 3.20; found: C, 68.57; H,
6.24; N, 3.18.
Further design and synthesis of more synthetically acces-
sible protecting groups and their application to regioselec-
tive glycosylation and alkylation of glucose, mannose and
galactose is underway and will be reported in due course.
This concept might also be extended to acyclic polyols.
4.1.2. Methyl 6-O-picolinyl a-D-glucopyranoside (1m).
To a solution of methyl a-D-glucopyranoside 1a (5.0 g,
25.7 mmol) in collidine (40 mL) was added picolinyl
chloride (3.5 g, 25 mmol) at K20 8C. The resulting mixture
was stirred for 3 h at K20 8C and for 6 h at room
temperature. After evaporation, the residue was extracted
with EtOH/CH2Cl2 1:1, filtrated and concentrated in vacuo.
The residue was purified three times by chromatography
(MeOH/CH2Cl2, 9:1) to afford compound 1m (1.9 g, 26%,
yellowish gum); RfZ0.31 (CH2Cl2/MeOH, 9:1);
[a]DC90.4 (c 1.1, CHCl3); IR (neat/NaCl) 3418,
4. Experimental
4.1. General methods
Solvents were distilled and dried by standard methods; THF
and ether, from Na/benzophenone; and CH2Cl2 from P2O5.
All commercially available reagents were used without
1
further purification. 4 A molecular sieves were dried at
1731 cmK1; H NMR (250 MHz, CDCl3) d 8.76 (d, 1H,
˚
100 8C prior to use. Melting points are uncorrected and
recorded with a Bu¨chi capillary tube melting-point appara-
tus. Optical rotations were measured on a Perkin Elmer 141
polarimeter in a 1 dm cell at 20 8C. FTIR spectra were
recorded on Perkin Elmer Spectrum 1000 on NaCl windows
JZ4.5 Hz), 8.13 (d, 1H, JZ8.0 Hz), 7.86 (ddd, 1H, JZ1.5,
8.0, 8.0 Hz), 7.50 (ddd, 1H, JZ1.5, 4.5, 8.0 Hz), 4.88 (d,
1H, JZ3.5 Hz), 4.83 (dd, 1H, JZ4.5, 12.0 Hz), 4.60 (dd,
1H, JZ2.0, 12.0 Hz), 4,25 (m, 1H), 3.90 (m, 1H), 3.88 (m,
1H), 3.82 (dd, 1H, JZ9.5, 9.5 Hz), 3.60 (m, 2H), 3.45 (s,
3H), 1.98 (m, 2H); 13C NMR (65 MHz, CDCl3) d 164.9,
147.5 (2), 137.3, 127.1, 125.3, 99.6, 73.9, 71.9, 70.1, 69.6,
64.7, 55.2; LRMS (EIC, m/z, %): 300 (21) (MCHC), 124
1
or KBr pellets. H and 13C NMR spectra were recorded on
Bruker AC 250 or DRX 400 spectrometers (250 and
400 MHz, respectively), or Varian Mercury 300 (300 MHz).

N. Moitessier et al. / Tetrahedron 61 (2005) 6839–6853
6849
(100) (PyrCOOHCHC), 106 (90) (PyrCOC); HRMS (m/z):
[MCH]C calcd for C13H18NO7 300.1083, found 300.1093.
diluted with chloroform and washed with satd NaHCO3,
water and brine, dried over Na2SO4 and concentrated in
vacuo. To the obtained residue was added a solution of
methyl a-D-glucopyranoside 1a (7.0 g, 36.3 mmol) in
pyridine (150 mL). The mixture was stirred for 48 h then
concentrated and purified twice by chromatography
(CH2Cl2/MeOH, 1:0 then 19:1) to afford compound 1p,
which was recrystallized (CH2Cl2/hexanes) into highly pure
compound (2.85 g, 50%, yellow powder); RfZ0.38
(CH2Cl21/MeOH, 9:1); [a]DC0.6 (c 1.1, CHCl3); mp
112 8C; H NMR (250 MHz, CDCl3) d 8.54 (d, 1H, JZ
5.0 Hz), 7.63 (ddd, JZ1.5, 7.5, 7.5 Hz), 7.40–7.05 (m,
10H), 4.76 (d, 1H, JZ3.5 Hz), 4.05 (dd, 1H, JZ9.0,
9.0 Hz), 3.85 (dd, 1H, JZ9.0, 9.0 Hz), 3.70 (m, 2H), 3.60
(dd, 1H, JZ3.5, 9.5 Hz), 3.34 (s, 3H), 3.22 (dd, 1H, JZ3.5,
9.5 Hz), 2.97 (bs, 1H), 2.38 (s, 3H), 2.32 (s, 3H), 2.22 (d,
1H, JZ9.5 Hz); 13C NMR (65 MHz, CDCl3) d 161.9, 147.8,
141.1, 139.2, 137.2, 136.8, 136.4, 129.4, 128.5, 128.4,
128.3, 124.4, 121.8, 99.6, 86.5, 73.7, 72.2, 71.3, 70.5, 64.9,
55.0, 20.9, 20.8; LRMS (EIC, m/z, %): 466 (10) (MCHC),
465 (28) (MC), 288 (100) (Pyr(Tol)2COC), 273 (75)
(Pyr(Tol)2CCHC), 272 (70) (Pyr(Tol)2CC); Anal. calcd
for C27H31NO6: C, 69.66; H, 6.71; N, 3.01; found: C, 69.48;
H, 6.77; N, 3.00.
4.1.3. Methyl 6-O-diphenyl-(2-(4-dimethylaminopyri-
dyl)methyl a-D-glucopyranoside (1n). A suspension of
diphenyl-(2-(4-dimethylaminopyridyl)methanol (4.4 g,
14.5 mmol, prepared from 4-dimethylaminopyridine and
benzophenone as described by Fort)26 in water (50 mL) and
conc. HCl (5 mL) was refluxed for 1 h. The resulting
mixture was concentrated and the residue precipitated in
ether. A solution of this white powder in AcCl (12 mL) and
SOCl2 (18 mL) was stirred for 72 h then concentrated at rt
and co-evaporated twice with toluene. To this crude mixture
was added a solution of pyranoside 1a (10 g, 51.4 mmol) in
pyridine (200 mL). The resulting mixture was stirred for a
further 48 h, concentrated, extracted with CH2Cl2, filtrated,
washed with water and brine, dried over Na2SO4 and
concentrated in vacuo. The residue was purified by
chromatography (CH2Cl2/MeOH, from 19:1 to 9:1) to
afford compound 1n (1.2 g, 17%, white powder); RfZ0.21
(CH2Cl2/MeOH, 9:1); [a]D1C16.1 (c 1.0, CHCl3); IR (neat/
NaCl) 3403, 1586 cmK1; H NMR (400 MHz, CDCl3) d
8.17 (d, 1H, JZ4.0 Hz), 7.52–7.44 (m, 4H), 7.36–7.24 (m,
4H), 6.47 (s, 1H), 6.41 (m, 1H), 4.75 (br s, 1H), 4.06 (dd,
1H, JZ7.5, 9.0 Hz), 3.86 (dd, 1H, JZ8.5, 9.0 Hz), 3.73–
3.63 (m, 3H), 3.32 (s, 3H), 3.09 (d, 1H, JZ9.5 Hz), 2.91 (s,
7H), 2.30 (bs, 1H); 13C NMR (100 MHz, CDCl3) d 160.6,
154.3, 147.3, 144.9, 142.4, 130.1, 129.9, 128.3, 128.1,
127.8, 127.7, 127.6, 127.2, 121.6, 107.6, 105.5, 99.6, 86.9,
74.1, 72.5, 71.1, 70.8, 65.3, 55.1, 39.0; LRMS (EIC, m/z,
%): 481 (29) (MCHC), 287 (100) (DMAP(Ph)2COC);
Anal. calcd for C27H32N2O6: C, 67.48; H, 6.71; N, 5.83;
found: C, 67.55; H, 6.70; N, 5.80.
4.2. Preparation of compounds 1q and 1r
4.2.1. 6-(Hydroxy-di-p-tolyl-methyl)-pyridine-2-carb-
oxylic acid methyl ester (6). To a solution of dimethyl-
2,6-pyridinedicarboxylate 5 (5.0 g, 25.6 mmol) in THF
(200 mL) was added p-tolylmagnesium bromide (1 M
solution in ether, 25 mL, 25 mmol) at rt. After stirring for
30 min, another portion of p-tolylmagnesium bromide
(20 mL, 20 mmol) was added at K78 8C. The mixture
was stirred at K78 8C for 1 h then allowed to warm up to rt
over 2 h and quenched with satd NH4Cl. The mixture was
diluted with CHCl3 then washed with satd NaHCO3, water
and brine, dried over Na2SO4 and concentrated. The residue
was purified by chromatography (H/EA, 4:1 then 7:3) to
afford the alcohol 6 (4.04 g, 45%, brownish oil); RfZ0.45
4.1.4. Methyl 6-O-diphenyl-(thiophenyl)methyl a-D-glu-
copyranoside (1o). Using the procedure as for compound
1m, a solution of methyl a-D-glucopyranoside 1a (4.56 g,
23.5 mmol) in pyridine (100 mL) and diphenylthiophenyl-
methyl chloride (2.55 g, 8.97 mmol, prepared from 4,40-
dimethyl benzophenone and bromopyridine as described in
the litterature1) afforded, after chromatography (CH2Cl2/
MeOH, 1:0 then 19:1), compound 1o, which was pre-
cipitated (CH2Cl2/hexanes) (2.05 g, 52%, white powder);
RfZ0.49 (CH2Cl2/MeOH, 9:1); [a]DC64.2 (c 1.0, CHCl3);
1
(H/EA, 7:3); IR (neat/NaCl) 3427, 1728, 1587 cmK1; H
NMR (250 MHz, CDCl3) d 8.07 (d, 1H, JZ7.3 Hz), 7.78
(dd, 1H, JZ7.3, 7.3 Hz), 7.33 (d, 1H, JZ7.3 Hz), 7.17 (d,
4H, JZ8.0 Hz), 7.12 (d, 4H, JZ8.0 Hz), 6.29 (s, 1H), 3.98
(s, 3H), 2.35 (s, 6H); 13C NMR (65 MHz, CDCl3) d 164.8,
163.5, 145.6, 142.7, 137.0, 136.7, 128.3, 127.7, 125.8,
123.3, 80.3, 52.2, 20.6; LRMS (EIC, m/z, %): 348 (4) (MC
HC), 346 (11) (MC), 119 (100), 91 (63) (TolC); HRMS
(m/z): [MCH]C calcd for C22H22NO3 348.1599, found
348.1592.
1
mp 150 8C; H NMR (250 MHz, CDCl3) d 7.52 (m, 3H),
7.39–7.22 (m, 9H), 7.00 (m, 1H), 4.79 (d, 1H, JZ3.5 Hz),
3.75–3.62 (m, 2H), 3.62–3.45 (m, 4H), 3.45 (s, 3H), 2.61 (s,
1H), 2.51 (d, 1H, JZ1.0 Hz), 2.08 (d, 1H, JZ9.5 Hz); 13C
NMR (65 MHz, CDCl3) d 144.0, 143.8, 128.3, 127.9, 127.5,
126.4, 126.1, 99.0, 84.8, 74.6, 72.1, 71.6, 70.0, 64.0, 55.2;
LRMS (EIC, m/z, %): 443 (3) (MCHC), 442 (8) (MC),
265 (10) (Th(Ph)2COC), 249 (100) (Th(Ph)2CC); Anal.
calcd for C24H26O6S: C, 65.14; H, 5.92; found: C, 64.89; H,
5.85.
4.2.2. 1-[6-(Hydroxy-di-p-tolyl-methyl)-pyridin-2-yl]-2-
pyridin-2-yl-ethanol (7a). To a solution of 2-picoline
(4.7 mL, 47.3 mmol) in THF (100 mL) at K78 8C was
added BuLi (1.6 M solution in hexanes, 27.3 mL,
44.5 mmol) dropwise. After stirring for 30 min, this solution
was cannulated to a solution of alcohol 6 (4.7 g, 13.5 mmol)
in THF (200 mL) at K78 8C. The resulting mixture was
stirred for a further 1.5 h at K40 8C then quenched with satd
NH4Cl. The mixture was diluted with CHCl3 then washed
with satd NaHCO3, water and brine, dried over Na2SO4 and
concentrated in vacuo. The residue was filtrated on a silica
gel pad then precipitated in hexanes to afford the enol 7a
4.1.5. Methyl 6-O-ditolyl-(2-pyridyl)methyl a-D-gluco-
pyranoside (1p). To a suspension of NaH (730 mg,
18.2 mmol) in THF (20 mL) was added a solution of
ditolyl-(2-pyridyl)methanol (3.5 g, 12.1 mmol, prepared
from 4,40-dimethyl benzophenone and bromopyridine as
described in the literature26) in THF (100 mL) at 0 8C. After
stirring for 1 h, SOCl2 (1.15 mL, 15.7 mmol) was added and
the mixture was stirred for a further 5 h. The mixture was

6850
N. Moitessier et al. / Tetrahedron 61 (2005) 6839–6853
1
(4.6 g, 83%, yellow powder); RfZ0.40 (H/EA, 1:1); H
NMR (250 MHz, CDCl3, 4:1 mixture of isomers) d 8.52 (d,
0.2H, JZ4.5 Hz), 8.35 (d, 0.8H, JZ4.5 Hz), 8.01 (d, 0.2H,
JZ4.5 Hz), 7.92 (d, 0.8H, JZ4.5 Hz), 7.82–7.55 (m, 2H),
7.33–7.01 (m, 12.2H), 6.83 (s, 0.8H), 6.52 (s, 0.8H), 5.83 (s,
0.2H), 2.35 (s, 6H); 13C NMR (65 MHz, CDCl3) d 162.2,
161.1, 158.1, 151.5, 149.6, 144.4, 143.5, 142.9, 137.4,
137.3, 137.0, 136.9, 136.5, 128.6, 128.5, 128.1, 127.9,
126.3, 124.1, 123.0, 122.5, 121.6, 120.8, 119.2, 118.6, 96.2,
80.4, 21.0; LRMS (EIC, m/z, %): 409 (19) (MCHC), 408
(48) (MC), 91 (TolC); HRMS (m/z): [MCH]C calcd for
C27H25N2O2 409.1916, found 409.1915.
NaOH then water. The resulting mixture was extracted with
CH2Cl2. The aqueous layers were mixed then washed with
water and brine, dried over Na2SO4 and concentrated in
vacuo. The residue was purified by chromatography (H/EA,
1:1) to afford diol 8b (560 mg, 70%, colorless oil); RfZ0.32
1
(H/EA, 4:1); IR (neat/NaCl) 3411, 1591, 1575 cmK1; H
NMR (250 MHz, CDCl3) d 7.60 (dd, 1H, JZ7.5, 7.5 Hz),
7.30–7.07 (m, 15H), 7.02 (d, 1H, JZ7.5 Hz), 5.71 (bs, 1H),
5.04 (dd, 1H, JZ5.0, 7.5 Hz), 3.20 (dd, 1H, JZ5.0,
14.0 Hz), 3.05 (dd, 1H, JZ7.5, 14.0 Hz), 3.30–2.80 (bs,
1H), 2.37 (s, 6H); 13C NMR (65 MHz, CDCl3) d 162.6,
159.9, 143.3, 143.2, 137.0, 136.9, 129.6, 128.5, 128.3,
128.0, 126.5, 121.6, 119.1, 80.6, 74.4, 44.6, 21.0; LRMS
(EIC, m/z, %): 410 (15) (MCHC), 409 (56) (MC), 300
(74), 119 (80), 91 (100) (TolC); HRMS (m/z): [MCH]C
calcd for C28H28NO2 410.2120, found 410.2137.
4.2.3. 1-[6-(Hydroxy-di-p-tolyl-methyl)-pyridin-2-yl]-2-
phenyl-ethanone (7b). To a solution of toluene (12.8 mL,
120 mmol) in THF (125 mL) at K78 8C was added t-BuLi
(1.7 M solution in pentane, 25 mL, 42.3 mmol) dropwise.
After stirring for 1 h, this solution was cannulated onto a
solution of alcohol 6 (4.2 g, 12.1 mmol) in THF (125 mL) at
K78 8C. The resulting mixture was stirred for a further 1.5 h
at K50 8C then was quenched with satd NH4Cl, diluted with
CHCl3, washed with satd NaHCO3, brine, dried over
Na2SO4 and concentrated in vacuo. The residue was purified
by chromatography (H/EA, 4:1) to afford ketone 7b (1.70 g,
35%, colorless oil) along with starting material 6 (1.25 g,
30%); RfZ0.57 (H/EA, 4:1); IR (neat/NaCl) 3452, 1697,
1583 cmK1; 1H NMR (250 MHz, CDCl3) d 7.99 (d, 1H, JZ
7.5 Hz), 7.78 (d, 1H, JZ7.5, 7.5 Hz), 7.35 (d, 1H, JZ
7.5 Hz), 7.32–7.10 (m, 13H), 5.68 (s, 1H), 4.49 (s, 2H), 2.36
(s, 6H); 13C NMR (65 MHz, CDCl3) d 163.2, 150.8, 142.7,
137.4, 137.0, 134.2, 130.4, 129.6, 128.6, 128.4, 128.0,
127.9, 127.7, 126.6, 126.2, 120.9, 80.7, 44.5, 20.9; LRMS
(EIC, m/z, %): 408 (12) (MCHC), 407 (32) (MC), 389
(50), 119 (90), 91 (100) (TolC); HRMS (m/z): [MKOH]C
calcd for C28H24NO 390.1858, found 390.1866.
4.2.6. [6-(2-Pyridin-2-yl-vinyl)-pyridin-2-yl]-di-p-tolyl-
methanol (9a). To a solution of diol 8a (1.85 g,
4.50 mmol) in THF (130 mL) at K78 8C was added
NaHMDS (2 M solution in THF, 5.6 mL, 11.2 mmol)
followed by a solution of PhNTf2 (1.93 g, 5.41 mmol) in
THF (20 mL). The resulting mixture was stirred for 2 h then
quenched with satd NH4Cl, diluted with CHCl3, washed
with water and brine, dried over Na2SO4 then concentrated
in vacuo. The residue was purified by chromatography
(H/EA, 9:1 then 3:2) to afford olefin 9a (1.33 g, 76%,
colorless oil); RfZ0.33 (H/EA, 1:1); IR (neat/NaCl) 3370,
1
1586, 1567 cmK1; H NMR (250 MHz, CDCl3) d 8.64 (d,
1H, JZ4.5 Hz), 7.72 (m, 3H), 7.64 (dd, 1H, JZ7.5, 7.5 Hz),
7.44 (d, 1H, JZ8.0 Hz), 7.44 (d, 1H, JZ7.5 Hz), 7.21 (m,
5H), 7.11 (d, 4H, JZ9.0 Hz), 7.03 (d, 1H, JZ8.0 Hz), 6.65
(bs, 1H), 2.36 (s, 6H); 13C NMR (65 MHz, CDCl3) d 162.9,
154.7, 152.6, 149.6, 143.4, 137.0, 136.8, 136.6, 132.2,
130.8, 128.5, 128.0, 123.2, 122.7, 121.9, 121.6, 80.4, 20.9;
LRMS (EIC, m/z, %): 393 (18) (MCHC), 392 (88) (MC),
301 (78), 182 (63), 181 (94), 119 (100), 91 (98) (TolC).
4.2.4. 1-[6-(Hydroxy-di-p-tolyl-methyl)-pyridin-2-yl]-2-
pyridin-2-yl-ethanol (8a). A solution of enol 7a (3.5 g,
8.55 mmol) in EtOH/THF (1:2, 250 mL) was stirred under
hydrogen (1 atm) in presence of 10% Pd/C (3.3 g) for 24 h.
The catalyst was filtered off and the solution concentrated in
vacuo. The residue was purified by chromatography (H/EA,
2:1 then 1:1 then 1:4) to afford diol 8a (2.11 g, 60%,
yellowish powder); RfZ0.38 (H/EA, 1:1); IR (neat/NaCl)
4.2.7. (6-Styryl-pyridin-2-yl)-di-p-tolyl-methanol (9b).
Using the same procedure as for olefin 9a, a solution of
diol 8b (540 mg, 1.32 mmol) in THF (40 mL) at K78 8C,
NaHMDS (2 M solution in THF, 1.65 mL, 3.30 mmol) and
a solution of PhNTf2 (565 mg, 1.58 mmol) in THF (10 mL)
afforded, after chromatography (H/EA, 1:0 then 9:1), olefin
9b (430 mg, 83%, colorless oil); RfZ0.60 (H/EA, 4:1); IR
1
3383, 1593, 1572 cmK1; H NMR (250 MHz, CDCl3) d
1
8.49 (d, 1H, JZ4.5 Hz), 7.62 (dd, 1H, 7.5, 7.5 Hz), 7.56 (dd,
1H, JZ8.0, 9.5 Hz), 7.52 (d, 1H, JZ7.3 Hz), 7.21–7.08 (m,
9H), 7.06 (d, 1H, JZ8.0 Hz), 6.97 (d, 1H, JZ7.3 Hz), 6.25
(s, 1H), 6.15 (bs, 1H), 5.26 (dd, 1H, JZ3.0, 8.5 Hz), 3.42
(dd, 1H, JZ3.0, 14.5 Hz), 3.22 (dd, 1H, JZ8.5, 14.5 Hz),
2.36 (s, 3H), 2.34 (s, 3H); 13C NMR (65 MHz, CDCl3) d
161.8, 160.9, 159.1, 148.1, 143.4, 143.3, 136.8, 136.5,
128.3, 127.8, 123.8, 121.4, 120.9, 118.7, 80.2, 73.3, 43.2,
20.8; LRMS (EIC, m/z, %): 411 (19) (MCHC), 410 (63)
(MC), 300 (52), 299 (47), 119 (93), 93 (100), 91 (78)
(TolC); HRMS (m/z): [MCH]C calcd for C27H27N2O2
411.2072, found 411.2083.
(neat/NaCl) 3368, 1584, 1566 cmK1; H NMR (250 MHz,
CDCl3) d 7.73–7.56 (m, 3H), 7.45–7.26 (m, 4H), 7.24 (d,
2H, JZ7.5 Hz), 7.14 (d, 2H, JZ7.5 Hz), 7.01 (d, 1H, JZ
8.0 Hz), 6.71 (bs, 1H), 2.37 (s, 6H); 13C NMR (65 MHz,
CDCl3) d 162.8, 153.3, 143.5, 136.9, 136.8, 136.4, 133.3,
128.7, 128.5, 128.4, 128.1, 127.1, 121.3, 120.5, 80.4, 21.0;
LRMS (EIC, m/z, %): 392 (14) (MCHC), 391 (54) (MC),
300 (61), 180 (82), 119 (100), 91 (81) (TolC); HRMS (m/z):
[MCH]C calcd for C28H26NO 392.2014, found 392.2031.
4.2.8. [6-(2-Pyridin-2-yl-ethyl)-pyridin-2-yl]-di-p-tolyl-
methanol (10a). A solution of olefin 9a (1.31 g,
3.34 mmol) in EtOH/THF (1:1, 100 mL) in presence of
10% Pd/C (700 mg) was stirred under H2 (1 atm) for 1.5 h.
The suspension was filtered and the filtrate concentrated in
vacuo. The residue was purified by chromatography (H/EA,
4:1 then 1:2) to afford alcohol 10a (941 mg, 72%, colorless
oil); RfZ0.45 (H/EA, 1:1); IR (neat/NaCl) 3350, 1590,
4.2.5. 1-[6-(Hydroxy-di-p-tolyl-methyl)-pyridin-2-yl]-2-
phenyl-ethanol (8b). To a suspension of LiAlH4 (225 mg,
5.9 mmol) in THF (75 mL) at 0 8C was added a solution of
ketone 7b (801 mg, 1.97 mmol) in THF (25 mL) dropwise.
After stirring for 1 h, water was added followed by 3 N

N. Moitessier et al. / Tetrahedron 61 (2005) 6839–6853
6851
1574 cmK1; 1H NMR (250 MHz, CDCl3) d 8.42 (d, 1H, JZ
3.5 Hz), 7.38 (2dd, 2H, JZ8.0, 8.0 Hz), 7.08–6.90 (m,
11H), 6.81 (d, 1H, JZ7.5 Hz), 6.40 (bs, 1H), 3.15 (m, 4H),
2.35 (s, 6H); 13C NMR (65 MHz, CDCl3) d 162.2, 160.5,
158.6, 148.9, 143.4, 136.4, 136.3, 135.9, 128.2, 127.7,
122.8, 121.1, 120.8, 119.9, 80.0, 37.1, 37.0, 20.7; LRMS
(EIC, m/z, %): 395 (18) (MCHC), 394 (72) (MC), 303
(88), 156 (88), 119 (90), 91 (100) (TolC); HRMS (m/z):
[MCH]C calcd for C27H27N2O 395.2123, found 395.2131.
procedure as for compound 1p, alcohol 10b (401 mg,
1.02 mmol) in water (10 mL) and conc. HCl (1 mL) led to a
white powder that was reacted with AcCl (4 mL) and SOCl2
(6 mL) then with pyranoside 1a (1.1 g, 5.7 mmol) in
pyridine (20 mL) to afford, after chromatography
(CH2Cl2/MeOH, 49:1 then 19:1), compound 1r (265 mg,
50%) along with recovered alcohol 10b (115 mg, 29%);
RfZ0.48 (CH2Cl2/MeOH, 9:1); [a]DC0.5 (c 0.8, CHCl3);
mp 72 8C; IR (neat/NaCl) 3401, 1586, 1574 cmK1; 1H NMR
(250 MHz, CDCl3) d 7.47 (dd, 1H, JZ7.5, 8.0 Hz), 7.31
(d, 1H, JZ8.0 Hz), 7.29–7.08 (m, 13H), 6.91 (d, 1H, JZ
7.5 Hz), 6.75 (s, 1H), 4.71 (d, 1H, JZ3.5 Hz), 3.88–3.80 (m,
2H), 3.76 (m, 1H), 3.64 (ddd, 1H, JZ3.5, 9.0, 9.0 Hz), 3.41
(ddd, 1H, JZ3.5, 9.0 Hz), 3.34 (s, 3H), 3.20 (ddd, 1H, JZ
5.0, 9.0 Hz), 3.13 (dd, 2H, JZ6.5, 8.0 Hz), 2.98 (dd, 2H,
JZ6.5, 8.0 Hz), 2.82 (s, 1H), 2.35 (s, 6H), 2.28 (d, 1H, JZ
9.0 Hz); 13C NMR (65 MHz, CDCl3) d 161.5, 160.6, 141.1,
140.3, 140.1, 137.2, 137.1, 136.6, 129.1, 128.9, 128.6,
128.5, 128.2, 125.8, 122.2, 121.3, 99.5, 86.8, 74.0, 72.9,
72.3, 69.8, 65.9, 55.3, 39.3, 36.5, 21.0; LRMS (EIC, m/z,
%): 570 (14) (MCHC), 569 (27) (MC), 393 (25), 392 (88),
377 (100) (PGCHC), 376 (60) (PGC), 376 (100)
(PGKHC), 284 (72), 119 (83), 91 (63) (TolC); HRMS
(m/z): [MCH]C calcd for C35H40NO6 570.2855, found
570.2903.
4.2.9. (6-Phenethyl-pyridin-2-yl)-di-p-tolyl-methanol
(10b). Using the same procedure as for alcohol 10a, a
solution of olefin 9b (480 mg, 1.09 mmol) in EtOH (25 mL)
afforded, after chromatography (H/EA, 1:0 then 9:1) alcohol
10b (403 mg, 94%, colorless oil); RfZ0.68 (H/EA, 4:1); IR
1
(neat/NaCl) 3378, 1591, 1574 cmK1; H NMR (250 MHz,
CDCl3) d 7.50 (dd, 1H, JZ7.5, 8.0 Hz), 7.28–7.16 (m, 5H),
7.18 (d, 2H, JZ7.5 Hz), 7.12 (d, 2H, JZ7.5 Hz), 6.99
(d, 1H, JZ7.5 Hz), 6.93 (d, 1H, JZ8.0 Hz), 6.61 (s, 1H),
3.18–3.03 (m, 4H), 2.35(s, 6H);13C NMR (65 MHz, CDCl3) d
162.4, 158.9, 143.6, 141.2, 136.6, 136.5, 128.4, 128.2, 128.0,
125.8, 121.4, 120.2, 80.2, 39.4, 35.2, 20.9; LRMS (EIC, m/z,
%): 394 (11) (MCHC), 393 (49) (MC), 302 (45), 156 (48),
119 (76), 91 (100) (TolC); HRMS (m/z): [MCH]C calcd for
C28H28NO 394.2171, found 394.2445.
4.2.10. Methyl 6-O-[6-(2-Pyridin-2-yl-ethyl)-pyridin-2-
yl]-di-p-tolyl-methyl-a-D-glucopyranoside (1q). A sus-
pension of alcohol 10a (910 mg, 2.3 mmol) in water
(18 mL) and conc. HCl (2 mL) was refluxed for 1 h. The
resulting mixture was concentrated and the residue
precipitated in ether. A solution of this white powder in
AcCl (4 mL) and SOCl2 (6 mL) both freshly distilled was
stirred for 48 h then concentrated at rt and co-evaporated
twice with toluene. To the crude alkyl chloride 11a was
added a solution of pyranoside 1a (1.8 g, 9.2 mmol) in
pyridine (40 mL) and the mixture was stirred for 48 h,
concentrated, extracted with CH2Cl2, filtrated, washed with
water and brine, dried over Na2SO4 and concentrated in
vacuo. The residue was purified by chromatography
(CH2Cl2/MeOH, from 49:1 to 9:1) to afford protected
compound 1q (820 mg, 63%, white powder) along with
recovered alcohol 10a (205 mg, 23%); RfZ0.39 (CH2Cl2/
MeOH, 9:1); [a]DC56.0 (c 0.3, CHCl3); mp 92 8C; IR
4.3. Optimized synthesis of compound 10a
4.3.1. 6-Pyridin-2-ylethynyl-pyridine-2-carboxylic acid
methyl ester (15). To a solution of bromopicolinic acid
methyl ester 14 (1.88 g, 8.7 mmol.) and 2-ethynylpyridine
in a mixture of triethylamine and THF (80 mL, 1:1) were
added copper iodide (33 mg, 0.17 mmol) and PdCl2(PPh3)2
(244 mg, 0.35 mmol). The resulting mixture was heated to
60 8C and stirred for 2 h. The solid suspension was filtered,
concentrated in vacuo and the residue was purified by flash
chromatography (H/A/CH2Cl2, 4:1:1 then 1:4:1) to afford
compound 15 (1.99 g, 96%, light yellow solid); RfZ0.40
(EA); IR (neat/NaCl) 1718 cmK1 1H NMR (400 MHz,
;
CDCl3) d 8.61 (d, 1H, JZ4.5 Hz), 8.08 (dd, 1H, JZ1.5,
7.0 Hz), 7.84 (dd, 1H, JZ8.0, 8.0 Hz), 7.80–7.58 (m, 3H),
7.26 (m, 1H), 3.97 (s, 3H); 13C NMR (65 MHz, CDCl3) d
164.2, 149.5, 147.6, 142.0, 141.5, 136.9, 135.6, 130.1,
127.1, 123.9, 123.0, 88.1, 86.4, 52.3; LRMS (EIC, m/z, %):
261.5 (100) (MCNaC), 239.5 (66) (MCHC); Anal. calcd
for C14H10N2O2: C, 70.58; H, 4.23; N, 11.76; found: C,
70.47; H, 4.22. N, 11.54.
1
(neat/NaCl) 3400, 1589, 1573 cmK1; H NMR (250 MHz,
CDCl3) d 8.59 (dd, 1H, JZ1.5, 5.5 Hz), 7.55 (ddd, 1H, JZ
1.5, 8.0, 8.0 Hz), 7.53 (dd, 1H, JZ8.0, 8.0 Hz), 7.38 (d, 2H,
JZ8.0 Hz), 7.28 (d, 2H, JZ8.0 Hz), 7.26 (dd, 1H, JZ5.5,
8.0 Hz), 7.17–7.09 (m, 6H), 7.05 (d, 1H, JZ7.5 Hz), 4.71
(d, 1H, JZ3.5 Hz), 4.70 (bs, 1H), 3.94 (m, 1H), 3.85 (m,
2H), 3.65 (m, 1H), 3.35 (s, 3H), 3.33–3.05 (m, 6H), 2.37 (s,
3H), 2.33 (s, 3H), 2.29 (d, 1H, JZ9.5 Hz); 13C NMR
(65 MHz, CDCl3) d 161.7, 160.6, 160.3, 149.2, 141.1,
139.3, 137.3, 136.8, 136.5, 129.5, 128.4, 128.3, 123.1,
122.2, 121.3, 121.2, 99.5, 86.9, 73.7, 73.5, 72.6, 69.7, 66.7,
55.2, 38.9, 38.0, 20.9; LRMS (EIC, m/z, %): 571 (2) (MC
HC), 570 (5) (MC), 394 (21), 393 (76), 378 (69) (PGCHC),
377 (56) (PGC), 376 (100) (PGKHC), 284 (75), 119 (50),
91 (36) (TolC); HRMS (m/z): [MCH]C calcd for
C34H39N2O6 571.2808, found 571.2796.
4.3.2. 6-(2-Pyridin-2-yl-ethyl)-pyridine-2-carboxylic
acid methyl ester (16). A suspension of ester 15 (1.94 g,
8.15 mmol) and 10% Pd/C (1.9 g) was stirred for 24 h under
hydrogen, filtered and concentrated in vacuo. The residue
was next purified by chromatography (CH2/Cl2/MeOH, 99:1
then 94:6) to afford compound 16 (1.78 g, 90%, light yellow
solid); RfZ0.48 (CH2Cl2/MeOH); IR (neat/NaCl) 1740,
1723 cmK1; 1H NMR (400 MHz, CDCl3) d 8.55 (d, 1H, JZ
4.5 Hz), 7.96 (d, 1H, JZ8.0 Hz), 7.70 (dd, 1H, JZ8.0,
8.0 Hz), 7.56 (ddd, 1H, JZ2.0, 8.0, 8.0 Hz), 7.30 (d, 1H,
JZ8.0 Hz), 7.12 (m, 2H), 4.01 (s, 3H), 3.38 (m, 2H), 3.26
(m, 2H); 13C NMR (65 MHz, CDCl3) d 165.8, 161.6, 160.5,
149.1, 147.4, 137.0, 136.1, 126.3, 122.9, 122.6, 121.1, 52.7,
37.8, 37.7; LRMS (EIC, m/z, %): 265.5 (32) (MCNaC),
4.2.11. Methyl 6-O-[6-phenethyl-pyridin-2-yl]-di-p-tolyl-
methyl-a-D-glucopyranoside (1r). Following the same

6852
N. Moitessier et al. / Tetrahedron 61 (2005) 6839–6853
243.5 (100) (MCHC); Anal. calcd for C14H14N2O2: C,
69.41; H, 5.82; N, 11.56; found: C, 69.40; H, 5.85. N, 11.50.
2,6-lutidine (1.45 mL of a 0.5 M solution in CH2Cl2,
2 equiv) then TBSOTf (345 mg, 1.308 mmol, portionwise
as a 100 mg/mL CH2Cl2 solution, 3.5 equiv) at K78 8C.
After stirring for 3 h, 3 mL of methanol were added at
K78 8C, and after warming up to room temperature the
solution was filtrated on a silica gel pad and concentrated in
vacuo. Chromatography (DCM/MeOH, 99:1, then 95:5)
afforded monosilylated compound 18 along with the 3-O-
TBS isomer as a minor product and recovered starting
material 1q (26 mg); RfZ0.29 (CH2Cl2/MeOH, 19:1); IR
(neat/NaCl) 3500–3000 (broad), 3054, 2930, 1451,
4.4. Acetylation reaction
Acetylation—general procedure. To a solution of 1b–1r
(0.25 mmol) in CH2Cl2 (50 mL) or THF (25 mL) was added
K2CO3 (350 mg) then DMAP (0.25 mL of a 0.05 M solution
in CH2Cl2) and Ac2O (0.35 mL of a 0.5 M solution in
CH2Cl2). After stirring for 1 h at rt, the solution was filtrated
on a silica gel pad and concentrated in vacuo. Integration on
1H NMR spectra of the mixture permitted accurate measure
of the regioselectivity. COSY experimental were done to
assign the peaks.
1
1265 cmK1; H NMR (300 MHz, CDCl3) d 8.49 (dd, 1H,
JZ8.0, 2.5 Hz), 7.48 (dd, 1H, JZ8.0, 8.0 Hz), 7.47 (dd, 1H,
JZ8.0, 8.0 Hz), 7.30 (m, 4H), 7.23 (d, 1H, JZ8.0 Hz), 7.09
(m, 7H), 6.97 (d, 1H, JZ1.0 Hz), 4.56 (d, 1H, JZ5.4 Hz),
3.98–3.78 (m, 3H), 3.73–3.60 (m, 2H), 3.42 (d, 1H, JZ
5.0 Hz), 3.36 (m, 1H), 3.33 (s, 3H), 3.27 (d, 1H, JZ5.0 Hz),
3.19 (m, 5H), 2.46 (t, 1H, JZ8.0 Hz), 2.34 (s, 3H), 0.93 (s,
9H), 0.17 (s, 3H), 0.14 (s, 3H).; 13C NMR (75 MHz, CDCl3)
d 161.6, 160.7, 160.2, 149.7, 148.9, 140.5, 139.9, 137.1,
136.9, 136.6, 136.4, 129.5, 129.2, 128.7, 128.5, 128.4,
123.2, 122.1, 121.1, 100.4, 86.9, 74.0, 73.4, 73.1, 69.8, 66.4,
55.4, 38.9, 38.0, 26.0, 21.1, 18.4, K4.2, K4.5; HRMS
(m/z): [MCH]C calcd for C40H53N2O6Si 685.3673, found
685.3682.
Table 4. Representative chemical shifts of monoacetylated compounds
2, H-1/H-2
3, H-1/H-3
4, H-1/H-4
1b
1c
1d
1e
1f
1g
1h
1i
1j
1k
1l
1m
1n
1o
1p
1q
1r
4.92/4.73
4.89/4.70
4.88/4.69
4.91/4.72
4.94/4.70
5.04/–
4.96/4.82
4.96/4.82
4.97/4.79
4.91/4.50
5.02/4.70
4.92/–
4.90/4.87
4.93/4.74
4.90/4.85
4.88/4.73
4.86/4.81
4.78/5.05
4.74/5.10
4.75/5.10
4.80/5.08
4.78/5.05
4.82/5.12
4.77/5.33
4.80/5.32
4.81/5.32
4.77/5.05
4.79/5.26
4.78/5.13
4.65/5.23
4.80/5.08
4.75/5.24
4.76/5.28
4.71/5.26
4.86/4.86
4.82/4.84
4.79/4.82
4.83/4.88
4.80/4.84
4.99/4.94
—
—
—
4.88/4.62
4.85/5.00
— /5.01
Acknowledgements
4.87/4.88
4.90/4.97
4.85/4.94
4.88/4.95
We thank Dr. Philippe Gros (UMR 7565, Nancy, France) for
helpful discussions on pyridine chemistry and Mrs.
Sandrine Adach, Mrs. Eliane Eppiger and Mr. Claude
Mathieu for technical assistance. We also thank Professor
Stephen Hanessian and Mr. Eric Therrien for helpful
discussions.
4.4.1. Methyl 3-O-acetyl-6-O-[6-(2-Pyridin-2-yl-ethyl)-
pyridin-2-yl]-di-p-tolyl-methyl-a-D-glucopyranoside
(17). To a solution of pyranoside 1q (143 mg, 0.25 mmol) in
CH2Cl2 (25 mL) was added K2CO3 (350 mg) then DMAP
(2.0 mL of a 0.05 M solution in CH2Cl2, 0.4 equiv) and
Ac2O (0.60 mL of a 0.5 M solution in CH2Cl2, 1.2 equiv) at
K78 8C. After stirring for 5 h, the solution was filtrated on a
silica gel pad and concentrated in vacuo. Chromatography
(H/EA, 2:3 then 1:4) afforded monoacetylated compound 17
(130 mg, 85%) along with other isomers; RfZ0.46 (EA);
[a]DC32.9 (c 2.0, CHCl3); IR (neat/NaCl) 3378, 3215,
1739 cmK1; 1H NMR (250 MHz, CDCl3) d 8.50 (d, 1H, JZ
4.5 Hz), 7.54 (ddd, 1H, JZ1.5, 7.5, 7.5 Hz), 7.48 (dd, 1H,
JZ7.5, 7.5 Hz), 7.30 (m, 4H), 7.22 (d, 1H, JZ7.5 Hz), 7.14
(m, 7H), 7.00 (d, 1H, JZ7.5 Hz), 6.50–6.10 (bs, 1H), 5.26
(dd, 1H, JZ9.5, 9.5 Hz), 4.72 (d, 1H, JZ3.5 Hz), 3.99 (dd,
1, JZ9.5, 9.5 Hz), 3.83 (m, 1H), 3.64 (m, 1H), 3.52 (dd, JZ
3.5, 9.5 Hz), 3.37 (s, 3H), 3.30–3.10 (m, 5H), 2.36 (s, 3H),
2.35 (s, 3H), 2.12 (s, 3H); 13C NMR (65 MHz, CDCl3) d
171.5, 161.6, 161.0, 160.3, 148.8, 140.4, 140.1, 137.1,
136.6, 136.4, 129.0, 128.9, 128.5, 128.4, 123.3, 122.0,
121.2, 121.0, 99.6, 86.8, 75.4, 71.3, 70.5, 70.4, 65.7, 55.2,
38.6, 37.8, 21.0; HRMS (m/z): [MCH]C calcd for
C36H41N2O7 613.2914, found 613.2894.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.tet.2005.04.
060
References and notes
1. (a) David, S.; Hanessian, S. Tetrahedron 1985, 41, 643.
(b) Grindley, T. B. Adv. Carbohydr. Chem. Biochem. 1997, 53,
16.
2. Osborn, H. M. I.; Brome, V. A.; Harwood, L. M.; Suthers,
W. G. Carbohydr. Res. 2001, 332, 157.
3. Evtushenko, E. V. Carbohydr. Res. 1999, 316, 187.
4. (a) Kurahashi, T.; Mizutani, T.; Yoshida, J.-i. Tetrahedron
2002, 58, 8669. (b) Hu, G.; Vasella, A. Helv. Chim. Acta 2002,
86, 4369.
4.4.2. Methyl 2-O-tert-butyl dimethyl silyl-6-O-[6-(2-
pyridin-2-yl-ethyl)-pyridin-2-yl]-di-p-tolyl-methyl-a-D-
glucopyranoside (18). To a solution of pyranoside 1q
(209 mg, 0.367 mmol) in CH2Cl2 (35 mL) was added
5. (a) Moitessier, N.; Chapleur, Y. Tetrahedron Lett. 2003, 44,
1731. (b) Griswold, K. S.; Miller, S. J. Tetrahedron Lett. 2003,
59, 8869–8875. (c) Kattning, E.; Albert, M. Org. Lett. 2004, 6,
945–948. (d) Kurahashi, T.; Mizutani, T. Yoshida, J.-i.

N. Moitessier et al. / Tetrahedron 61 (2005) 6839–6853
6853
Tetrahedron 2002, 58, 8669. (e) Hu, G.; Vasella, A. Helv.
Chim. Acta 2002, 86, 4369 and references therein.
previously reported: Hanessian, S.; Kagotani, M. Carbohydr.
Res. 1990, 202, 67.
6. Haines, A. H. Adv. Carbohydr. Chem. Biochem. 1976, 33, 11.
7. Kurahashi, T.; Mizutani, T.; Yoshida, J.-i. J. Chem. Soc.,
Perkin Trans. 1 1999, 465. Although previous studies were
reported (Houdier, S.; Perez, S. J. Carbohydr. Chem. 1995, 14,
1117. Brewster, M. E.; Huang, M.-J.; Pop, E.; Pitha, J.; Dewar,
M. J. S.; Kaminski, J. J.; Bodor, N. Carbohydr. Res. 1993, 242,
53.) This report deals with pure kinetic data.
22. Toshima, K.; Tatsuka, K. Chem. Rev. 1993, 93, 1503.
23. Similar results were reported on the same substrate under
different conditions (Ac2O, pyridine in DMF ratio: 7/48/8, 1:
6.8: 1.1), Hanessian, S.; Kagotani, M. Carbohydr. Res. 1990,
202, 67.
24. Crich, D.; Dubkin, V. J. Am. Chem. Soc. 2001, 123,
6819–6825.
8. See also: Christofides, J. C.; Davies, D. B. J. Chem. Soc.,
Perkin Trans. 2 1987, 97.
25. Moitessier, N.; Henry, C.; Len, C.; Chapleur, Y. J. Org. Chem.
2002, 67, 7275–7282.
9. Klein, R. A. J. Am. Chem. Soc. 2002, 124, 13931.
10. Commercially available compound.
11. Helferich, B. Justus Liebigs Ann. Chem. 1924, 440, 1.
26. (a) Faller, J. W.; Ma, Y. J. Am. Chem. Soc. 1991, 113, 1579.
(b) Tzerpos, N. I.; Zarkadis, A. K.; Kreher, R. P.; Repas, L.;
Lehnig J. Chem. Soc., Perkin Trans. 1 1995, 755–761.
(c) Cuperly, D.; Gros, P.; Fort, Y. J. Org. Chem. 2002, 67,
238–241. (d) Gros, P.; Fort, Y. Eur. J. Org. Chem. 2002, 3375.
27. Gillet, B.; Nicole, D.; Delpuech, J.-J.; Gross, B. Org. Magn.
Res. 1981, 17, 28–36.
´
12. Hanessian, S.; Lavallee, J.-F. Can. J. Chem. 1975, 53, 2975.
13. Franke, F.; Guthrie, R. D. Aust. J. Chem. 1977, 30, 639.
14. Ishihara, K.; Kurihara, H.; Yamamoto, H. J. Org. Chem. 1993,
58, 3791.
15. Evans, M. E.; Long, L., Jr; Parrish, F. W. J. Org. Chem. 1968,
33, 1074.
28. Rockwell, G. D.; Grindley, T. B. J. Am. Chem. Soc. 1998, 120,
10953–10963.
16. Hanessian, S.; Plessas, N. R. J. Org. Chem. 1969, 34, 1035.
17. Johansson, R.; Samuelsson, B. J. Chem. Soc., Perkin Trans. 1
1984, 2371.
29. Bergmeier, S. C.; Arason, K. M. Tetrahedron Lett. 2000, 41,
5799.
30. Yadav, J. S.; Reddy, B. V. S. Synlett. 2000, 1275.
¨
¨
31. Hofle, G.; Steglich, W.; Vorbruggen, H. Angew. Chem. 1978,
90, 602–615.
18. Collins, P. M.; Oparaeche, N. N.; Munasinghe, V.; Ranjit, N.
J. Chem. Soc., Perkin Trans. 1 1975, 1700.
´
19. Moitessier, N.; Chretien, F.; Chapleur, Y.; Humeau, C.
Tetrahedron Lett. 1995, 36, 8023.
32. (a) Sonogashira, K.; Tohda, Y.; Hagihara, N. Tetrahedron Lett.
1975, 24, 4467. (b) Sonogashira, K. In Trost, B. M., Fleming,
I., Eds.; Comprehensive Organic Synthesis; Pergamon:
Oxford, 1991; Vol. 3, p 521.
20. Ishihara, K.; Kurihara, H.; Yamamoto, H. J. Org. Chem. 1993,
58, 3791.
21. A study on regioselectivity of the acetylation reaction was
33. Gilman, H.; Spatz, S. M. J. Org. Chem. 1951, 1485.