Halide promoted organotin-mediated carbohydrate benzylation: Mechanism and application

Article abstract of DOI:10.1016/j.tet.2013.02.024

In the present study, the mechanistic origin of the promoted organotin-mediated carbohydrate benzylation by halides was explored by the comparison of the activation ability of halides on benzylation of methyl β-d-galactoside. It was demonstrated that the

Full text of DOI:10.1016/j.tet.2013.02.024

Tetrahedron 69 (2013) 2693e2700
Contents lists available at SciVerse ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Halide promoted organotin-mediated carbohydrate benzylation:
mechanism and application
,
a,
Yixuan Zhou a, Jinyang Li a, Yingjie Zhan a, Zhichao Pei b , Hai Dong
*
*
a Huazhong University of Science & Technology, School of Chemistry & Chemical Engineering, Luoyu Road 1037, 430074 Wuhan, PR China
b State Key Laboratory of Crop Stress Biology in Arid Areas and College of Science, Northwest A&F University, Yangling, 712100 Shaanxi, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
In the present study, the mechanistic origin of the promoted organotin-mediated carbohydrate benzy-
lation by halides was explored by the comparison of the activation ability of halides on benzylation of
methyl b-D-galactoside. It was demonstrated that the improvement of benzylation in terms of speed,
Received 25 September 2012
Received in revised form 29 January 2013
Accepted 4 February 2013
yield and general convenience in the presence of halides was due to the formation of an activated oxide
species by the coordination of halide anions to the tetracoordinate tin atoms. The halide, being able to
form a stronger bond with tin, showed stronger activation ability. The catalytic amount of bromide
should be preferentially chosen as an activation additive in the reaction with consideration of economy,
convenience and moderate activation ability. The results were further applied to mono- and multi-
benzylation of additional carbohydrate structures. A range of prototype carbohydrate structures were
efficiently prepared with the guidance of the principle. In light of the previous studies and the present
experimental results, the general rules for organotin-mediated benzylation of carbohydrates have been
summarized.
Available online 13 February 2013
Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction
diols and polyols with organotin. Benzyl bromide was herein used
as an electrophile. The reaction was normally performed in the
Benzyl ethers are frequently used ether protecting groups in
carbohydrate chemistry, since they are particularly stable entities,
resistant to basic and acidic conditions, and also easily removed by
catalytic hydrogenation.1 Together with ester groups, they already
form the selective protection and deprotection of saccharide
building blocks prior to and after the glycosidic linkage step when
these building blocks are used as donors or acceptors for the syn-
thesis of oligosaccharides.2e7 Direct benzylation of carbohydrates
using benzyl halides requires extreme conditions, such as strong
base in polar solvents, which is difficult to lead to highly regiose-
lectivity for poly-hydroxyl compounds.8e10 Selective benzylation
usually requires special means, such as relying on the reductive
cleavage of benzylidene acetals,11 and depending on reagents ca-
pable of promoting regioselective alkylation of carbohydrates.
These reagents could be tin (IV), boron (IV), copper (II), mercury (II)
and nickel (II)-based complexes,12e15 as well as silver carbonate.16
However, the most widely used method for highly regioselective
benzylation involves the utilization of organotin.17,18 The regiose-
lectivity was provided by the introduction of dibutylstannylene
acetals or stannyl ethers, which was derived from the treatment of
presence of quaternary ammonium iodide,19 quaternary ammo-
nium bromide19,20 or cesium fluoride21,22 for the improvement in
terms of reaction rate, yield and general convenience. However, the
mechanistic origin of the improvement for the reaction by halides
has not been fully clarified. For example, the improvement of the
reaction by quaternary ammonium iodide cannot be simply at-
tributed to the generation of more reactive benzyl iodide, through
the replacement of bromine of alkyl bromide by iodine, since
quaternary bromide and cesium fluoride are also efficient active
additives.19,23 A more reasonable explanation may involve the
dibutylstannylene acetal intermediates, in which the coordination
of halide anions to tin enhances the nucleophilicity of one of
the bound oxygen atoms.19 However, this interpretation was
never demonstrated due to the unclear mechanism of the
dibutylstannylene-mediated protection. Furthermore, although
regioselective benzylation with the assistance of halides have been
reported in hundreds of protocols so far,17,18 a systematic study on
how the halides influence the reaction has never been performed.
Consequently, there always were some uncertainties on the utili-
zation of halides for any a specific dibutylstannylene-mediated
benzylation, including the choice of a halide (iodide, bromide,
chloride or fluoride), the amount of halide salt used (catalytic,
stoichiometric or excess amount), and the prediction of the ben-
zylation outcome (regioselectivity). These uncertainties have
* Corresponding authors. Tel.: þ86 27 87793243; fax: þ86 27 87793242; e-mail
addresses: peizc@nwsuaf.edu.cn (Z. Pei), hdong@mail.hust.edu.cn (H. Dong).
0040-4020/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tet.2013.02.024

2694
Y. Zhou et al. / Tetrahedron 69 (2013) 2693e2700
restricted the application of the method, especially on a regiose-
lective multiple carbohydrate benzylation, which is of particular
importance since two or more hydroxyl groups in polyols can be
benzylated in a single-step process. In the present study, the origin
of the promoted benzylation by halide anions was addressed,
where the effect of the quaternary ammonium iodide, bromide,
chloride, and fluoride on the dibutylstannylene-mediated benzy-
major 3-O-alkylated product and a minor 6-O-alkylated product were
obtained after subsequent alkylation in toluene in the presence of
tetrabutylammonium halide salts. It is not necessary to employ either
stoichiometric or excess amounts of ammoniumhalides.23 A tentative
mechanism has been proposed to explain the absence of 4-O-alky-
lated product and the formation of 6-O-alkylated product.29 Thus, we
chose the 3/6-O-regioselective dibutylstannylene-mediated benzy-
lation of methyl
b-
D-galactoside was investigated, and the process
lation of methyl b-D-galactoside 1 as a model to investigate the effects
was also applied on
benzylation.
a
regioselective multiple carbohydrate
of tetrabutylammonium halides on this simple and convenient re-
action (Scheme 1).
It was a general concept that the regioselectivity of the
dibutylstannylene-mediated protection was controlled by complex
stannylene structures.18 However, we have recently suggested that
the regioselectivity is more likely to be controlled by stereo-
electronic effects of the parent carbohydrate structure,24 which has
been supported by further experiments on organoboron- and
organosilicon-mediated protection.25,26 Thereby, the origin of the
regioselectivity for dibutylstannylene-mediated protection could
be depicted to proceed via a selective SneO bond cleavage in
a stannylene acetal ring, enhanced by a pentacoordinate tin atom.
For dibutylstannylene-mediated benzylation (Fig. 1), as suggested,
Bu
Bu
OH
OH
HO
HO
O
Bu2SnO
Sn
BnBr
O
O
OMe
1
OMe
Bu4NX
O
OH
OH
OH
OBn
HO
BnO
HO
HO
O
O
+
OMe
2
OMe
3
OH
major
OH
minor
Scheme 1. Dibutylstannylene-mediated benzylation of galactoside 1.
Br
X
X
Bu2Sn
BnO
Bu2Sn
Bu2Sn
BnO
BnBr
M+
O
M+Br
+
+
M+X
O
O
O
+
+
X
R1
R2
R1
R2
R1
R2
SnBu2
O
SnBu2
O O
O
M+X
M+
+
X
X
Br
R1
R2
R1
R2
SnBu2
SnBu2
SnBu2
BnBr
M+
O
OBn
M+X
M+Br
O
O
O
OBn
R1
R2
R1
R2
R1
R2
Fig. 1. A proposed mechanism for the activation of tin intermediates by halides. X¼F, Cl, Br or I.
the coordination of the halide anion to the tetracoordinate tin atom
enhances the selective SneO bond cleavage to give a reactive ox-
ygen anion coordinated to ions (Mþ), followed by a nucleophilic
attack of the oxygen species to an electrophile (BnBr). Conse-
quently, the regioselectivity was controlled by both the preferential
cleavage of SneO bond and the nucleophilicity of the oxygen spe-
cies. The formation and cleavage of SneX (halides) bond and SneO
bond were reversible, and finally equilibria were likely to form. The
formation of a stronger SneX bond would therefore lead to more
reactive oxygen anions, which markedly accelerated the reaction. If
this mechanism was valid, it could be proposed that the halide
anion, forming stronger bond with tin, shows a higher activation
ability for benzylation. This would be analogous to our previous
studies where anions activated a cascade inversion by supramo-
lecular effect.27 Then the activation ability should follow the order:
Fꢀ>Clꢀ>Brꢀ>Iꢀ. The bromide and iodide should not show higher
activation ability than chloride and fluoride in the reaction. In order
to demonstrate this, a systematic study on how the halides in-
fluence the dibutylstannylene-mediated benzylation was initiated
in the following experiments.
In order to compare the activation ability of halide anions, the
benzylation reactions were performed under the same conditions
with various halides. The conversion rates were recorded after the
same time intervals. The activation ability of various halides can be
reflected by the comparison of the conversion rates. Thus, the
methyl b-D-galactoside 1 (50 mg) was chosen to react with 1.1 equiv
of dibutyltin oxide in toluene to form the dibutylstannylene acetals.
The dibutylstannylene acetals were subsequently allowed to react
with 1.5 equiv of benzyl bromide in 4 mL toluene at 90 ꢁC for 10 h
with the separate addition of 0.5 equiv of tetrabutylammonium
iodide, bromide, chloride and fluoride. The conversion rates and the
3/6-O-benzylated product ratios were both determined by 1H NMR
spectra (Table S1a in SI). It can be seen in Fig. 2a. The recorded
conversion rates at the 10 h time point were too subtle for the
activation ability of the different halides to be distinguished. The
analogous product ratios of 3/6-OBn were achieved with the sep-
arate addition of chloride, bromide and iodide; whereas more 6-
OBn products formed with the addition of fluoride, in which the
product ratio of 3/6-OBn with fluoride was 56/21 in comparison to
64/7 with chloride, 66/8 with bromide and 65/9 with iodide.
When the reactions were tested in 2 mL toluene under the same
conditions, the analogous conversion rates were reached in 2 h (b in
Fig. 2), indicating the reactions were concentration dependent. The
conversion rates reached approximately 90% in 6 h. There were
slight differences in the conversion rates for all the halides. More 6-
OBn products formed with fluoride anion, thereby showing a poor
regioselectivity. Upon analysis of the results, we realized that the
reactions proceeded faster than expected. Thus the conversion
2. Results and discussion
The 3/6-O-regioselectivity has been reported for the
dibutylstannylene-mediated alkylation of galactosides with the acti-
vation of halide anions.28,29 The dibutylstannylene acetals were
generated by heating the unprotected galactosides with dibutyltin
oxide under reflux in toluene until a clear solution was formed. A

Y. Zhou et al. / Tetrahedron 69 (2013) 2693e2700
2695
(a)100
(b)
100
conversion
3-OBn
conversion
3-OBn
90
90
80
70
60
50
40
30
20
10
0
6-OBn
6-OBn
80
70
60
50
40
30
20
10
0
F
Cl
Br
I
F
Cl
Br
I
Anion
Anion
Fig. 2. The recorded conversion rates and 3/6-OBn ratios for benzylation of galactoside 1. (a) Reaction in 4 mL toluene for 10 h; (b) reaction in 2 mL toluene for 2 h.
rates must be recorded in a suitable short reaction time period,
when the conversion rates remain distinct. Therefore, in further
experiments, the reactions were performed under the same con-
ditions using less (1.1 equiv) benzyl bromide in order to avoid the
formation of di-OBn products. The conversion rates and the 3/6-O-
benzylated product ratios were determined by 1H NMR spectrum
with the reactions proceeding for 5 min, 2 h, 4 h and 6 h. The results
were outlined in Fig. 3 (also seen in Table S2 in SI). The recorded
conversion rate with fluoride at the 5 min time point was 58% in
comparison to 34% with chloride, 34% with bromide and 33% with
iodide, indicating a distinctly higher activation ability of fluoride
than other halides. At the 2 h reaction time point, the conversion
rates for all halides approached close to each other. The relative
poor regioselectivity for fluoride was also shown.
The results prompted us to further investigate the effect on
benzylation with the addition of varied amounts of tetrabuty-
lammonium halides (Table S3 in SI). When 0.2 equiv of halide an-
ions were employed, the conversion rates for all halides in 2 h were
around 32e48%, indicating the decreased reactivity. When stoi-
chiometric amounts of halides were employed, the conversion
rates in 2 h were roughly analogous to those when 0.5 equiv of
halides were employed. This indicated that the reactivity could not
be noticeably improved by increasing the amounts of halide addi-
tives once the used amount of halide was more than 0.5 equiv. The
activation ability was also compared by the comparison of the
conversion rates at the 5 min time point. The rates for fluoride were
27%, 58% and 73%, respectively when 0.2, 0.5 and 1.0 equiv of
fluoride anions were employed separately. However, the difference
in rates for the other halides was smaller, being around 21e22%,
33e34% and 35e42% corresponding to the employment of 0.2, 0.5,
and 1.0 equiv of halides. The results indicated that the fluoride
showed higher activation ability as expected, which was likely to
originate from the formation of strong SneF bond (Fig. 4a). It was
supposed that the dibutyltin acetal intermediates involve benzy-
lation as monomers in toluene at high temperature (100 ꢁC).24
Pentacoordinate tin atoms were thereby formed instantly when
the added halides coordinated to tetracoordinate tin atom of the
stannylene acetal monomers. Pentacoordinate-tetracoordinate
equilibria of tin were formed then, since tetracoordinate tin atom
could also generate from pentacoordinate tin atom by the breakage
of the SneO or SneX (X stands for halide anions) bond (Figs. 1 and
4). The stronger SneX bond corresponded to the formation of more
SneX bonds and less SneO bonds, leading to more oxygen anions
and less X anions. Weaker SneX bond corresponded to the for-
mation of less SneX bonds and more SneO bonds, leading to less
oxygen anions and more X anions. Therefore, once the tetrabuty-
lammonium fluoride was added to the reaction mixture (Fig. 4a),
the fluoride coordinated to the tin atom to form stronger SneF
conversion(F)
conversion(Cl)
conversion(Br)
conversion(I)
3-OBn (F)
3-OBn (Cl)
3-OBn (Br)
3-OBn (I)
100
90
80
70
60
50
40
30
20
10
0
time
Fig. 3. The recorded conversion rates and 3/6-OBn ratios for benzylation of galactoside 1 with time.

2696
Y. Zhou et al. / Tetrahedron 69 (2013) 2693e2700
F
Sn
Bu
OH
O
(a)
O
Bu
OMe
O
OH
Bu
Bu
Bu
F
F
Sn
Sn
Bu
Bu
Bu
OH
O
OH
OH
O
O
Sn
O
O
O
BnBr
+
OMe
TBABr
+
TBAF
OMe
X
OMe
O
TBA
BnO
O
OH
OH
OH
(b)
OH
O
Bu
Sn
O
OMe
Bu
X = Cl. Br, I
O
OH
Bu
Bu
Bu
X
X
Sn
Sn
Bu
Bu
Bu
OH
OH
OH
O
O
O
Sn
O
O
O
BnBr
TBAX
+
+
TBABr
OMe
OMe
OMe
BnO
O
TBA
O
OH
OH
OH
Fig. 4. Proposed approaches for promoted dibutylstannylene-mediated benzylation of galactoside 1 by halides. (a) With Fꢀ; (b) with Clꢀ, Brꢀ or Iꢀ
.
bond instantly, leading to the breakage of the weaker SneO bond.
The benzylation then proceeded through nucleophilic attack of the
oxygen species on benzyl bromide in a very short time (shorter
than 5 min). With the formation of benzylation products, the bro-
mide anions were generated, displacing the fluoride anion. When
the most of the fluoride anions coordinated to the tin atoms, further
benzylation was activated by the generation of bromide anions. As
a result, the reaction would proceed slowly by then. That is the
reason why 0.2 and 0.5 equiv of fluoride anion only led to ap-
proximately 20% and 50% conversion rate, respectively in 5 min.
When chloride, bromide and iodide were employed (Fig. 4b), these
anions could also coordinate to tin to form SneCl, SneBr and SneI
bonds, respectively. However, the SneO bond was stronger in this
case, leading to less active oxygen species. With the formation of
benzylation products, the bromide anions were also formed. As
a result, there were subtle differences of activation ability among
the three halides. For iodide, there was a possibility that the lower
activation ability was compensated by the formation of more re-
active benzyl iodide. These results demonstrated our hypothesis
and also indicated that it was not a preferential choice to employ
tetrabutylammonium fluoride in the reaction unless a rapid re-
action was necessary and a stoichiometric amount of tetrabuty-
lammonium fluoride additive was employed. Consequently, for
organotin mediated benzylation, the optimum choice when using
halides was to employ 0.5 equiv of bromide anions as activation
additive, which should also be in more benzylation cases, including
multiple carbohydrate benzylation.
organotin reagent in order to compare the present results with the
multiple esterification. Thereby, the unprotected methyl galacto-
side 1, glucosides 4 and 5, and mannoside 6 were treated with
1.1e3.3 equiv of dibutyltin oxide to form stannylene intermediates.
The intermediates were subsequently treated with 1.5e4.0 equiv of
benzyl bromide in dry toluene at 100 ꢁC for 6e12 h in the presence
of 0.5 equiv of tetrabutylammonium bromide. The resulting mix-
tures were directly analyzed by 1H NMR in DMSO-d6 after removal
of the solvent. The product distribution by NMR ratio was displayed
in Table 1.
For methyl
b-D-galactoside 1, the mono-protection had been
fully studied in the previous and present studies. The 3-hydoxyl
group was regioselectively benzylated accompanied with a minor
benzylation of 6-hydroxyl group, whereas the benzylation of 4-
hydroxyl group was never found. Thus, the utilization of 2.2 equiv
of dibutyltin oxide and 3.0 equiv of benzyl bromide led to a major
expected product 7 (72%) where both the 3- and 6-hydroxyl groups
were benzylated and a minor 3-position protected product 2 (28%).
The regioselective mono-benzylation of the
a-D-glucoside 4 and b-
D-glucoside 5 had never been reported. Herein, the utilization of
1.1 equiv of dibutyltin oxide and 1.5 equiv of benzyl bromide led to
major products 8 (67%) and 10 (65%) where the 2-hydroxyl groups
were benzylated. The utilization of 2.2 equiv of dibutyltin oxide and
3.0 equiv of benzyl bromide led to major products 9 (67%) and 11
(56%) where both the 2- and 6-hydroxyl groups were benzylated.
However, when we used 3.3 equiv of dibutyltin oxide and 4.0 equiv
of benzyl bromide, expecting to obtain products with one free hy-
droxyl group, we failed and still obtained major products with two
free hydroxyl groups. Even much more than 4.0 equiv of benzyl
bromide were employed, the compounds with one free hydroxyl
group could not be found from the reaction mixtures. An un-
For the dibutylstannylene-mediated multiple carbohydrate es-
terification, products with one or two free hydroxyl groups were
produced by use of excess (2e3 equiv) organotin reagent in a one-
pot process in light of description.30,31 We wondered if the analo-
gous product patterns could also be achieved by an organotin-
mediated multiple carbohydrate benzylation, where the formed
stannylene intermediates would be treated with benzyl bromide in
the presence of 0.5 equiv of tetrabutylammonium bromide. Thus
expected product for the multiple benzylation of methyl b-D-glu-
coside 5 was compound 12 (23%) where both the 2- and 3-hydroxyl
groups were benzylated. Analogous results were obtained in the
multiple benzylation of a-D-mannoside 6. It was known that the
the results obtained from the benzylation of methyl
1 were applied to multiple benzylation and three other carbohy-
drate structures, methyl -glucoside 4, methyl -glucoside 5,
and methyl -mannoside 6 (Table 1). Although it has been re-
ported that benzylation of two hydroxyl groups in -glucoside
and -mannoside was made by use of 1 equiv equiv of dibutyltin
oxide in the presence of i-Pr2Net (DIPEA),20 we still use excess
b
-
D
-galactoside
utilization of 1.1 equiv of dibutyltin oxide led to a major product 13
where the 3-hydroxyl group was benzylated.32 At present we
employed 2e3 equiv of dibutyltin oxide and 3e4 equiv of benzyl
bromide, which led to a major product 14 (51e69%) where both the
3- and 6-hydroxyl groups were benzylated. An unexpected product
15 was also produced in the reactions. Upon analysis of the
resulting product compositions, a reactivity order of the hydroxyl
a
-D
b-D
a-D
a-D
a-D

Y. Zhou et al. / Tetrahedron 69 (2013) 2693e2700
2697
Table 1
One-pot organotin-mediated multiple carbohydrate benzylationa
O
Bu2SnO
Toluene
BnBr
O
HO
BnO
HO
OMe
OMe
Bu4NBr
Substrate
BnBr (equiv)
Bu2SnO (equiv)
NMR ratio (%)
Product 1
Product 2
Product 3
OH
OBn
O
OH
HO
HO
HO
BnO
HO
BnO
O
O
3.0
2.2
72
28
OMe
1
OMe
7
OMe
2
OH
OH
OH
OH
OH
OBn
O
1.5
3.0
4.0
1.1
2.2
3.3
67
33
33
33
67
67
O
O
HO
HO
HO
HO
HO
HO
4
8
9
HO
BnO
BnO
OMe
OMe
OMe
OH
O
OH
OBn
O
OH
1.5
3.0
4.0
1.1
2.2
3.3
65
22
21
35
56
56
O
O
22
23
HO
HO
HO
HO
HO
HO
HO
BnO
OMe
OMe
10
OMe
OMe
11
5
12
OH
OBn
OBn
OBn
3.0
4.0
4.0
2.2
2.2
3.3
38
28
14
51
69
67
11
3
19
HO
HO
HO
OBn
O
OH
O
OH
O
OBn
O
HO
HO
HO
BnO
HO
BnO
HO
HO
OMe
13
OMe
15
6
11
OBn
OMe
OMe
a
Reaction conditions: reactant (100 mg), Bu2SnO (1.1e3.3 equiv), TBAB (0.5 equiv), BnBr (1.1e4.0 equiv), Toluene (2 mL), 90 ꢁC, 12 h.
Br
Sn
groups for benzylation of b-D-galactoside was arranged as 3>6>2, 4
(a)
Bu
Bu
in comparison to the previously reported 3>6>2>4 for esterifica-
Br
tion.31 The reactivity order for benzylation of
b-D-glucoside was
TBA
BnO
O
arranged as 2>6>3>4 in comparison to the previously reported
6>3¼2>4 for esterification. The reactivity order for benzylation of
R1
R2
a-D-glucoside was arranged as 2>6>3, 4 in comparison to the
previously reported 6>2>3>4 for esterification. The reactivity or-
der for benzylation of -mannoside was arranged as 3>6>2>4 in
Br
Activated
a-D
Bu2Sn
comparison to the previously reported 6>3>2>4 for esterification.
In all cases, the benzylation showed better regioselectivity. The
differences in the reactivity order and the regioselectivity between
benzylation and esterification were likely due to the reactivity of
the electrophiles and the intermolecular migration of stanny-
lenes.33 The products with one free hydroxyl group were difficult to
acquire, which was likely due to the deactivation of the bound
oxygen atom connecting the tetracoordinate tin, which was also
coordinated by bromide. The tentative explanation was outlined in
Fig. 5. The oxide anion was difficult to form through the cleavage of
SneO bond of a pentacoordinate tin coordinated by two bromides
since SneBr bond was weaker than SneO bond (Fig. 5a). For ex-
Br
BnO
O
BnO
OTBA
R2
+
TBABr
+
Bu2Sn
Br
R1
R2
R1
Deactivated
(b)
Bu
Sn
Bu
OBn
O
Br
O
Bu
O
Sn
O
O
Bu
BnBr
O
O
O
TBABr
Bu
Bu
BnO
Sn
Sn
OMe
OMe
Bu
Bu
Br
Deactivated
ample, for the benzylation of the a-D-glucoside (Fig. 5b), the oxides
Fig. 5. The tetracoordinate tin coordinated by a bromide leading to a deactivated oxide
species.
in 3- and 4-position were deactivated due to their connection with
the tetracoordinate tin atoms, which were coordinated by a bro-
mide. As a result, the benzylation of 3- or 4-positions could hardly
occur. However, as a benefit of the outcome, it means that there will
be no need to worry about per-benzylation of substrates even if
a large excess amount of benzyl bromide is employed.
Finally, the reaction conditions were optimized in order to ob-
tain good isolation yields (Table 2). The di-benzylation of galacto-
side 1 and mannoside 6 showed relatively low reactivity, requiring
more benzyl bromide and longer reaction time. Consequently, the
di-benzylated compounds 7 and 14 were isolated in 80% and 70%
yields, respectively. However, the benzylation of glucosides 4 and 5
showed relatively high reactivity. Decreasing the amount of benzyl
bromide and shortening the reaction time thereby led to high
isolation yields of mono benzylated glucosides 8 (83%) and 10
(81%). The high isolated yields of di-benzylated compounds 9 (85%)
and 11 (72%) were also obtained with more benzyl bromide. In
these results, one of the most valuable results was unexpectedly to
obtain 2-hydroxyl group benzylated glucosides 8 and 10 in high
yields. These two compounds are traditionally prepared by making
use of complex reaction sequences, being potential intermediates
to form important glycoside-type carbohydrate building block
acceptors.
In the light of previous conducted studies18,24 and the results of
the present experiments, the general rules for organotin-mediated
benzylation of carbohydrates can be summarized as follows:
2.1. For selective mono-benzylation of diols and polyols
The substrate should react with 1e1.1 equiv of dibutyltin oxide
to form dibutylstannylene acetal. Then the dibutylstannylene acetal

2698
Y. Zhou et al. / Tetrahedron 69 (2013) 2693e2700
In all cases, the trans-diols having higher reactivity,33 require
Table 2
The optimized one-pot multiple carbohydrate benzylation
less benzyl bromide (1.2 equiv), lower reaction temperature and
shorter reaction time. The cis-diols and 1,3-diols having lower re-
activity,33 require more benzyl bromide (1.5e2.5 equiv), higher
reaction temperature and longer reaction time. The amount of
solvent used should be as little as possible to increase the con-
centration of the substrate. The polar solvent DMF can be used
instead of the non-polar solvent toluene, which will increase the
reactivity since DMF molecules can also coordinate to the tin atoms
and polar solvent favors S2N reaction. In addition, any anion can be
employed in the reaction instead of bromide in principle. However,
with consideration of economy, convenience and moderate acti-
vation ability, the catalytic amount of tetrabutylammonium bro-
mide should be preferentially chosen as the activation additive. The
studies on benzylation by employing catalytic amount of dibutyltin
oxide could also be guided by the above general rules and are in
proceeding.
Substrate
OH
Product
Yield (%)
80[a]
OBn
O
HO
HO
HO
BnO
O
OMe
1
OMe
7
OH
OH
OH
OH
O
O
83[b]
HO
HO
HO
HO
4
8
HO
BnO
OMe
OMe
OH
O
OBn
O
85[c]
HO
HO
HO
HO
4
9
HO
BnO
OMe
OMe
OH
O
OH
O
81[b]
HO
HO
HO
HO
OMe
OMe
5
10
OH
OBn
3. Conclusions
OH
OBn
O
72[c]
O
By the study of the activation ability of halides on organotin-
mediated benzylation of methyl b-D-galactoside, the mechanism
HO
HO
HO
HO
OMe
OMe
11
5
OH
OBn
of the halide promoted benzylation was initially revealed, sug-
gesting an activated oxide species forming from the coordination of
the halide anion to the tetracoordinate tin atom. It was also dem-
onstrated that the halide, which could form a stronger bond with
tin showed stronger activation ability, thereby fluoride being the
strongest activation additive to the benzylation. However, a cata-
lytic amount of bromide was employed as the best activation ad-
ditive in the reaction owing to its regeneration and moderate
activation ability. The results were further applied to additional
carbohydrate structures and multiple carbohydrate benzylations. A
range of prototype carbohydrate structures were efficiently pre-
pared with the guidance of the principle. General rules for
organotin-mediated benzylation of carbohydrates have been
summarized in present studies.
HO
BnO
HO
BnO
OH
OH
O
70[a]
O
HO
HO
6
14
OMe
OMe
Reaction conditions: [a] 2.1 equiv Bu2SnO, 0.5 equiv TBAB, 5 equiv BnBr, 100 ꢁC,
12 h; [b] 1.1 equiv Bu2SnO, 0.5 equiv TBAB, 1.2 equiv BnBr, 100 ꢁC, 8 h; [c] 2.1 equiv
Bu2SnO, 0.5 equiv TBAB, 4 equiv BnBr, 100 ꢁC, 10 h.
should react with 1.2e1.5 equiv of benzyl bromide in a suitable
volume of toluene at 90e100 ꢁC for 6e12 h with the addition of
0.5 equiv of tetrabutylammonium bromide. For diols, the regiose-
lectivity is controlled by steric and stereoelectronic effect of the
substrate itself, and thus it can be predicted according to the parent
diol structure.24 Therefore, the hydroxyl groups adjacent to axial
substituents are preferentially benzylated in the trans-diols; the
equatorial hydroxyl groups are preferentially benzylated in the cis-
diols; and the primary hydroxyl groups are preferentially benzy-
lated in the 1,3-diols. For polyols, it must be confirmed, which two
hydroxyl groups form the dibutylstannylene acetal with dibutyltin
oxide first. Then the regioselectivity can also be predicted by the
steric and stereoelectronic effect principle. Therefore, the equato-
rial hydroxyl group adjacent to an axial hydroxyl group is prefer-
entially benzylated in a polyol having an axial hydroxyl group, e.g.,
mono-benzylation of galactoside and mannoside; and the equato-
rial hydroxyl group adjacent to an axial substituent or anomeric
center is preferentially benzylated in a polyol without having an
axial hydroxyl group, e.g., mono-benzylation of glucoside.
4. Experimental section
4.1. General
All commercially available starting materials and solvents were
reagent grade and used without further purification. Chemical re-
actions were monitored with thin-layer chromatography using
precoated silica gel 60 (0.25 mm thickness) plates (Macher-
eyeNagel). Flash column chromatography was performed on silica
gel 60 (SDS 0.040e0.063 mm). 1H NMR spectra were recorded with
a Bruker Avance 400 instrument at 298 K in DMSO-d6, using the
residual signals from DMSO-d6 (1H:
d
¼2.50 ppm) as internal stan-
dard. 1H peak assignments were made by first order analysis of the
spectra, supported by standard 1He1H correlation spectroscopy
(COSY).
2.2. For selective multi-benzylation of polyols
4.2. General procedure for benzylation
Every two hydroxyl groups require 1e1.1 equiv of dibutyltin
oxide to form dibutylstannylene acetals and 2e3 equiv of benzyl
bromide to react with one dibutylstannylene acetal. Therefore, the
polyols with four hydroxyl groups should react with 2e2.2 equiv of
dibutyltin oxide to form dibutylstannylene acetals. The dibutyl-
stannylene acetals should react with 4e6 equiv of benzyl bromide
in a suitable volume of toluene at 90e100 ꢁC for 6e12 h with the
addition of 0.5 equiv of tetrabutylammonium bromide. The regio-
selectivities are dependent on the dibutylstannylene acetals and
the stereoelectronic effects of diols. Thus, the regioselectivities can
also be predicted by the steric and stereoelectronic effect principle
mentioned above (Scheme S1 in SI).
The starting material (0.515 mmol) and dibutyltin oxide
(1.1e3.3 equiv) was dissolved in 20 mL anhydrous toluene, and
refluxed for 1 h. After evaporation of the solvent till 2 mL, benzyl
bromide (1.1e4.5 equiv) was added dropwise, and then allowed to
react at 90e100 ꢁC for 6e12 h. The resulting mixture was directly
analyzed by 1H NMR in DMSO-d6.
4.3. Methyl 3,6-di-O-benzyl-b-D
-galactopyranoside (7)32
Methyl
b-D-galactopyranoside (200 mg, 1.03 mmol) and dibu-
tyltin oxide (566 mg, 2.266 mmol) was dissolved in 40 mL

Y. Zhou et al. / Tetrahedron 69 (2013) 2693e2700
2699
anhydrous toluene, and refluxed for 1 h. After evaporation of the
NMR (DMSO-d6, 400 MHz):
d
¼7.23e7.39 (m, 10H, 2ꢂPh), 5.21 (d,
solvent till 4 mL remained, benzyl bromide (612
m
l, 5.15 mmol) was
J¼5.6, 1H, 4-OH), 5.17 (d, J¼6, 1H, 3-OH), 4.53e4.49 (m, 4H,
2ꢂPhCH2), 4.27 (d, J¼8, 1H, 1-H), 3.74e3.77 (d, J¼12, 1H, 6a-H),
3.51e3.55 (dd, J1¼4, J2¼10, 1H, 6b-H), 3.43 (s, 3H, OMe), 3.30e3.41
(m, 2H, 4-H, 5-H), 3.09e3.15 (m, 1H, 3-H), 3.02 (t, J¼8, 1H, 2-H)
ppm.
added dropwise, and then allowed to react at 100 ꢁC for 12 h. The
resulting mixture was directly purified by flash column chroma-
tography (hexane/ethyl acetate, 2:1). To give 309 mg of product
(80%). 1H NMR (DMSO-d6, 400 MHz):
d
¼7.26e7.43 (m, 10H, 2ꢂPh),
5.11 (d, J¼5.2, 1H, 2-OH), 4.52e4.70 (m, 4H, 2ꢂPhCH2), 4.64 (d,
J¼5.6, 1 H, 4-OH), 4.06 (d, J¼7.6, 1H, 1-H), 3.89 (s, 1H, 4-H),
3.47e3.64 (m, 4H, 2-H, 5-H, 6a-H, 6b-H), 3.39 (s, 3H, OMe),
3.25e3.28 (dd, J1¼3.2, J2¼9.6, 1H, 3-H) ppm.
4.8. Methyl 3, 6-di-O-benzyl-
a-D
-mannopyranoside (14)20
Methyl -mannopyranoside (200 mg, 1.03 mmol) and dibu-
a-D
tyltin oxide (566 mg, 2.266 mmol) was dissolved in 40 mL anhy-
drous toluene, and refluxed for 1 h. After evaporation of the solvent
4.4. Methyl 2-O-benzyl-
a-D
-glucopyranoside (8)34
till 4 mL remained, benzyl bromide (612 ml, 5.15 mmol) was added
dropwise, and then allowed to react at 100 ꢁC for 12 h. The resulting
mixture was directly purified by flash column chromatography
(hexane/ethyl acetate, 2:1). To give 270 mg of product (70%). 1H
Methyl -glucopyranoside (200 mg,1.03 mmol) and dibutyltin
oxide (282 mg, 1.133 mmol) was dissolved in 40 mL anhydrous
toluene, and refluxed for 1 h. After evaporation of the solvent till
a-D
4 mL remained, benzyl bromide (122 ml, 1.236 mmol) was added
NMR (DMSO-d6, 400 MHz):
d
¼7.26e7.43 (m, 10H, 2ꢂPh), 5.08 (d,
dropwise, and then allowed to react at 100 ꢁC for 8 h. The resulting
mixture was directly purified by flash column chromatography
(hexane/ethyl acetate, 1:1). To give 243 mg of product (83%). 1H
J¼6.4, 1H, 4-OH), 4.89 (d, J¼4.8, 1H, 2-OH), 4.68e4.55 (m, 4H,
2ꢂPhCH2), 5.56 (d, J¼1.6, 1H, 1-OH), 3.87 (s, 1H, 2-OH), 3.74 (d,
J¼9.6, 1H, 6a-H), 3.50e3.72 (m, 3H, 4-H, 5-H, 6b-H), 3.40 (dd, J1¼3.2,
J2¼9.2, 1H, 3-OH), 3.25 (s, 3H, OMe) ppm.
NMR (DMSO-d6, 400 MHz):
d
¼7.24e7.41 (m, 5H, Ph), 5.10 (d, J¼6,
1H, 4-OH), 4.97 (d, J¼7.2, 1H, 3-OH), 4.78 (s, 2H, PhCH2), 4.51e4.55
(m, 2H, 1-H, 6-OH), 3.62e3.66 (m, 1H, 6a-H), 3.43e3.50 (m, 2H, 3-H,
6b-H), 3.30e3.35 (m, 2H, 2-H, 5-H), 3.29 (s, 3H, OMe), 3.23e3.27 (m,
1H, 4-H) ppm.
Acknowledgements
This study was supported by the Fundamental Research Funds
for the Central Universities (HUST: 2012QN146), the National Na-
ture Science Foundation of China (Nos. 21272083), the Chutian
Project-Sponsored by Hubei Province and the Project-Sponsored by
the SRF for ROCS, SEM. The authors are also grateful to the staffs in
the Analytical and Test Center of HUST for support with the NMR
instruments.
4.5. Methyl 2, 6-di-O-benzyl-
a-D
-glucopyranoside (9)20
Methyl -glucopyranoside (200 mg,1.03 mmol) and dibutyltin
a-D
oxide (566 mg, 2.266 mmol) was dissolved in 40 mL anhydrous
toluene, and refluxed for 1 h. After evaporation of the solvent till
4 mL remained, benzyl bromide (490 ml, 4.412 mmol) was added
dropwise, and then allowed to react at 100 ꢁC for 10 h. The resulting
mixture was directly purified by flash column chromatography
(hexane/ethyl acetate, 2:1). To give 327 mg of product (85%). 1H
Supplementary data
Table S1, Table S2, Table S3, Scheme S1, 1H NMR data of com-
pounds 2, 3, 12, 13, 15 and 1H NMR-spectra of compound 7, 8, 9, 10,
11, and 14. Supplementary data related to this article can be found
at http://dx.doi.org/10.1016/j.tet.2013.02.024.
NMR (DMSO-d6, 400 MHz):
d
¼7.26e7.39 (m, 10H, 2ꢂPh), 5.11 (d,
J¼6, 1H, 4-OH), 5.07 (d, J¼5.2, 1H, 3-OH), 4.71 (d, J¼3.6, 1H, 1-H),
4.59e4.68 (m, 2H, PhCH2), 4.50 (s, 2H, PhCH2), 3.67e3.69 (d, J¼9.2,
1H, 6a-H), 3.50e3.56 (m, 3H, 3-H, 5-H, 6b-H), 3.26 (s, 3H, OMe),
3.18e3.21 (dd, J1¼3.6, J2¼9.6, 1H, 2-H), 3.11e3.13 (m, 1H, 4-H) ppm.
References and notes
4.6. Methyl 2-O-benzyl-
b-D
-glucopyranoside (10)35
1. Wuts, P. G. M.; Greene, T. W. Greene’s Protective Groups in Organic Synthesis;
John Wiley & Sons: Hoboken, NJ, 2007.
ꢀ
2. Gomez, A. M. In Glycoscience; Frascr-Reid, B., Ed.; Springer: Berlin, Germany,
Methyl -glucopyranoside (200 mg, 1.03 mmol) and dibutyltin
oxide (282 mg, 1.133 mmol) was dissolved in 40 mL anhydrous
toluene, and refluxed for 1 h. After evaporation of the solvent till
b-D
2008; pp 103e177.
3. Hsu, Y.; Lu, X. A.; Zulueta, M. M. L.; Tsai, C. M.; Lin, K. I.; Hung, S. C.; Wong, C. H.
J. Am. Chem. Soc. 2012, 134, 4549.
4. Hsu, C. H.; Hung, S. C.; Wu, C. Y.; Wong, C. H. Angew. Chem., Int. Ed. 2011, 50,
11872.
4 mL remained, benzyl bromide (122 ml, 1.236 mmol) was added
dropwise, and then allowed to react at 100 ꢁC for 8 h. The resulting
mixture was directly purified by flash column chromatography
(hexane/ethyl acetate, 1:1). To give 237 mg of product (81%). 1H
5. Bucher, C.; Gilmour, R. Angew. Chem., Int. Ed. 2010, 49, 8724.
6. Wang, P.; Zhu, J. L.; Yuan, Y.; Danishefsky, S. J. J. Am. Chem. Soc. 2009, 131, 16669.
€
7. Dong, H.; Rahm, M.; Brinck, T.; Ramstrom, O. J. Am. Chem. Soc. 2008, 130, 15270.
8. Hamann, C. H.; Koch, R.; Pleus, S. J. Carbohydr. Chem. 2002, 21, 53.
9. Fugedi, P.; Nanasi, P. J. Carbohydr., Nucleosides and Nucleotides 1981, 8, 547.
10. Iwashige, T.; Saeki, H. Chem. Pharm. Bull. 1967, 15, 1803.
11. Garegg, P. J. Pure Appl. Chem. 1984, 56, 845.
12. Eby, R.; Webster, K. T.; Schuerch, C. Carbohydr. Res. 1984, 129, 111.
13. Osborn, H. M. I.; Brome, V. A.; Harwood, L. M.; Suthers, W. G. Carbohydr. Res.
2001, 332, 157.
NMR (DMSO-d6, 400 MHz):
d¼7.26e7.40 (m, 5H, Ph), 5.16 (d, J¼5.2,
1H, 3-OH), 5.02 (d, J¼4.8, 1H, 4-OH), 4.67e4.76 (m, 2H, PhCH2), 4.57
(t, J¼2.4, 1H, 6-OH), 4.23 (d, J¼8, 1H, 1-H), 3.67e3.68 (dd, J1¼4,
J2¼12, 1H, 6a-H), 3.45e3.48 (m, 1H, 6b-H), 3.43 (s, 3H, OMe),
3.28e3.31 (m, 1H, 3-H), 3.08e3.15 (m, 2H, 4-H, 5-H), 3.00 (t, J¼4,
1H, 2-H) ppm.
14. Gangadharmath, U. B.; Demchenko, A. V. Synlett 2004, 2191.
15. Chan, L.; Taylor, M. S. Org. Lett. 2011, 12, 3090.
16. Satish, M.; Vaibhav, A. D.; Prasad, V. B.; Ravindranathan, K. K. P. Carbohydr. Res.
2010, 345, 559.
4.7. Methyl 2, 6-di-O-benzyl-
b-D
-glucopyranoside (11)36
17. David, S.; Hanessian, S. Tetrahedron 1985, 41, 643.
18. Grindley, T. B. Adv. Carbohydr. Chem. Biochem. 1998, 53, 17.
19. David, S.; Thieffry, A.; Veyrieres, A. J. Chem. Soc., Perkin Trans. 1 1981, 1796.
20. Simas, A. B. C.; da Silva, A. A. T.; dos Santos Filho, T. J.; Barroso, P. T. W.
Tetrahedron Lett. 2009, 50, 2744.
Methyl -glucopyranoside (200 mg, 1.03 mmol) and dibutyltin
oxide (566 mg, 2.266 mmol) was dissolved in 40 mL anhydrous
toluene, and refluxed for 1 h. After evaporation of the solvent till
b-D
ꢀ
21. Danishefski, S. J.; Hungate, R. J. Am. Chem. Soc. 1986, 108, 2486.
22. Nagashima, N.; Ohno, M. Chem. Pharm. Bull. 1991, 39, 1972.
23. Simas, A. B. C.; Pais, K. C.; da Silva, A. A. T. J. Org. Chem. 2003, 68, 5426.
4 mL remained, benzyl bromide (490 ml, 4.412 mmol) was added
dropwise, and then allowed to react at 100 ꢁC for 10 h. The resulting
mixture was directly purified by flash column chromatography
(hexane/ethyl acetate, 2:1). To give 277 mg of product (72%). 1H
€
24. Dong, H.; Zhou, Y. X.; Pan, X. L.; Cui, F. C.; Liu, W.; Liu, J. Y.; Ramstrom, O. J. Org.
Chem. 2012, 77, 1457.

2700
Y. Zhou et al. / Tetrahedron 69 (2013) 2693e2700
€
€
25. Lee, D.; Taylor, M. S. J. Am. Chem. Soc. 2011, 133, 3724.
31. Dong, H.; Pei, Z. C.; Bystrom, S.; Ramstrom, O. J. Org. Chem. 2007, 72, 1499.
32. Yang, G. B.; Kong, F. Z.; Zhou, S. H. Carbohydr. Res. 1991, 211, 179.
33. David, S.; Malleron, A. Carbohydr. Res. 2000, 329, 215.
34. Matwiejuk, M. Chem. Commun. 2011, 47, 8379.
€
26. Zhou, Y. X.; Ramstrom, O.; Dong, H. Chem. Commun. 2012, 48, 5370.
€
27. Dong, H.; Pei, Z. C.; Ramstrom, O. Chem. Commun. 2008, 11, 1359.
28. Ballell, L.; Joosten, J. A. F.; el Maate, F. A.; Liskamp, R. M. J.; Pieters, R. J. Tetra-
hedron Lett. 2004, 45, 6685.
35. Haque, M. E.; Kikuchi, T.; Kimihiro, Y.; Tsuda, Y. Chem. Pharm. Bull. 1985, 33,
29. Kaji, E.; Shibayama, K.; In, K. Tetrahedron Lett. 2003, 44, 4881.
30. Zhang, Z. Y.; Wong, C. H. Tetrahedron 2002, 58, 6513.
2243.
36. Qin, H. P.; Grindley, T. B. J. Carbohydr. Chem. 1994, 13, 475.