An easy and versatile approach for the regioselective De-O-benzylation of protected sugars based on the I2/Et3SiH combined system

Article abstract of DOI:10.1002/chem.201003332

The use of cheap and easy to handle reagents, such as I₂ and Et₃SiH, at low temperature allows the regioselective removal of benzyl protecting groups from highly O-benzylated carbohydrates. The observed regioselectivity is dependent

Full text of DOI:10.1002/chem.201003332

FULL PAPER
DOI: 10.1002/chem.201003332
An Easy and Versatile Approach for the Regioselective De-O-benzylation of
Protected Sugars Based on the I2/Et3SiH Combined System
Antonello Pastore,[a] Silvia Valerio,[b] Matteo Adinolfi,[a] and Alfonso Iadonisi*[a]
Abstract: The use of cheap and easy to
handle reagents, such as I2 and Et3SiH,
at low temperature allows the regiose-
lective removal of benzyl protecting
groups from highly O-benzylated car-
bohydrates. The observed regioselectiv-
ity is dependent on the nature of the
precursor, the least accessible carbinol
often being liberated. A mechanistic
investigation reveals that in situ gener-
ated HI is the promoter of the process,
whereas the regioselectivity appears to
be mainly controlled by steric effects.
However, the presence of an electron
withdrawing acyl protecting group can
switch the regioselectivity to favour de-
protection of the carbinol position far-
thest from the ester group. The proto-
col is experimentally simple and pro-
vides straightforward access in useful
yields to a wide range of partially pro-
tected mono- and disaccharide building
blocks that are valuable for the synthe-
sis of either biologically useful oligo-
saccharides or highly functionalised
chiral compounds. Partially protected
sugars thus obtained can also be cou-
pled in situ with a glycosyl donor, as il-
lustrated by the one-pot synthesis of a
Lewis X mimic from fully protected
precursors.
Keywords: benzyl ethers · carbohy-
drates · protecting groups · regio-
AHCTUNGTRENNGsUN electivity
Introduction
allows tuning of the reactivity of carbohydrate building
blocks in glycosylation chemistry; a large number of exam-
ples highlight the “arming” effect of benzyl groups on both
glycosyl donors and acceptors in comparison with the de-
pressed reactivity of sugars equipped with “disarming” acyl
groups.[3] Furthermore, benzyl groups are widely used in
stereocontrolled glycosylations in which protecting, but not
participating, groups are needed on the glycosyl donor.[4]
Rapid access to partially protected building blocks is one
of the most important goals[5] in modern organic synthesis.
In this regard, several procedures for the regioselective de-
O-benzylation of densely O-benzylated saccharide precur-
sors have been described in the literature. Among these,
only a few examples rely on controlled hydrogenolysis[6] and
most approaches require acidic conditions. Indeed, the re-
gioselective de-O-benzylation of saccharide compounds can
be effected with stoichiometric amounts of strong, sensitive
Lewis acids, such as SnCl4 or TiCl4,[7] excess alanes (diisobu-
tylaluminium hydride (DIBAL), triisobutylaluminum
(TIBAL)) at high temperatures[8,9] or under microwave acti-
vation,[10] BCl3 at low temperature,[11] or excess CrCl2 and
LiI at high temperature.[12] In these cases, regioselectivity is
critically dependent on the possibility of the saccharide sub-
strate to generate a chelated structure with the Lewis acid
present in each reagent system, whereas benzyl removal is
accomplished by a nucleophilic agent introduced either as
the counteranion of the chelated metal or as an independent
species. Chelation control can also serve in the regioselec-
tive de-O-benzylation of non-saccharide substrates under
acidic conditions.[13] Other regioselective de-O-benzylation
approaches for carbohydrates require the pre-existence of
suitable functionalities next to the site to be deprotected,
such as a free hydroxyl[14] or allyl group.[15]
The benzyl group is one of the most frequently used protect-
ing groups in organic synthesis for carbinol derivatisation.[1]
Several advantages justify the popularity of this protecting
group: a) the protection step can occur under either acidic
or basic conditions, b) a benzyl group can be installed at
multiple positions of molecules rich in functional groups,
such as carbohydrates, and c) the protecting group is stable
under a wide range of chemical conditions and can be re-
moved under mild conditions.[1] Hydrogenolysis is the most
commonly used method for de-O-benzylation and is espe-
cially suitable for the total deprotection of poly-O-benzylat-
ed substrates. Consequently, benzyl protecting groups are
frequently employed in organic synthesis for the manipula-
tion of polyhydroxylated targets and their removal is often
performed at advanced stages in synthetic schemes. Protect-
ed carbohydrates are frequently exploited as convenient
precursors for chiral compounds[2] through synthetic
schemes that routinely rely on O-benzylated intermediates.
In addition, the use of O-benzylated saccharide derivatives
[a] Dr. A. Pastore, Prof. M. Adinolfi, Dr. A. Iadonisi
Dipartimento di Chimica Organica e Biochimica
Univeresitꢀ degli Studi di Napoli “Federico II”
Via Cinthia 4, 80126 Naples (Italy)
Fax : (+39)81-674393
E-mail: iadonisi@unina.it
[b] Dr. S. Valerio
International Chemical Industry
81030 Cellole (CE) (Italy)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201003332.
Chem. Eur. J. 2011, 17, 5881 – 5889
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5881

Very recently, Crich and Vinogradova showed that highly
benzylated manno- and rhamno-derivatives can be depro-
tected with moderate selectivity at O-4 by treatment with a
stoichiometric oxidant, such as 2,3-dichloro-5,6-dicyanoben-
zoquinone (DDQ).[16] The primary positions can be de-O-
benzylated (and concomitantly triethylsilylated) with excel-
lent regioselectivity on prolonged exposure of benzylated
appeared to be a feasible task when working at low temper-
ature. Galactoside 1 displayed a high propensity to be de-
protected at O-4 (Scheme 1) and it was chosen as the model
compound for optimising several experimental parameters,
such as the order of reagent addition, their stoichiometry
and the temperature (Table 1).
mono-, di- and trisaccharides to the Et3SiH/
bined system, at high temperature under a CO atmo-
sphere.[17] Acetolysis is another approach that often leads to
ACHTUGNTREN[UNNG Co2(CO)8] com-
ACHTUNGTRENNUNG
selective de-O-benzylation (and concomitant acetylation) of
the primary positions, but in this case undesired cleavage of
glycosidic bonds might occur.[18] Trimethylsilyl iodide
(TMSI) has also been reported to produce selective de-O-
benzylation of primary carbinols (via the corresponding tri-
methylsilyl ethers).[19]
Scheme 1. Regioselective de-O-benzylation of model compound 1 (Bn=
benzyl).
Herein, we describe the wide and unprecedented scope of
the I2/Et3SiH combined system in the regioselective de-O-
benzylation of sugars. This reagent permits the smooth gen-
eration of anhydrous HI and several useful applications
have already been established in our laboratory, for exam-
ple, the combination of I2 (stoichiometric) and Et3SiH (cata-
lytic) is an effective activating system for trihaloacetimidate
glycosyl donors[20] and it was successively exploited by Taka-
hashi and co-workers in their protocols aimed at the stereo-
selective synthesis of b-2,6-dideoxyglycosides.[21] An analo-
gous stoichiometric combination of these reagents in metha-
nol was found to be effective for the cleavage of benzyli-
dene acetals.[22] Additionally, stoichiometric amounts of both
I2 and Et3SiH allow glycosyl iodides to be generated from 1-
O-acylated precursors within a few minutes; this rapid pro-
cess has been incorporated into several procedures, which
have lead to a variety of useful saccharide intermediates,
such as 1,2-orthoesters,[23] ethylidenes,[23] glycals,[23] thio- and
selenoglycosides,[24] and 2-O-deprotected allyl glycosides,[25]
as well as glyco-conjugates with estradiol[26] or melanogenic
5,6-dioxyindole.[27] Additional synthetically useful applica-
tions of I2/Et3SiH have been reported by other groups for
the regioselective reductive ring opening of benzylidenes,[28]
the reductive Ferrier rearrangement of glycals[29] and the
Friedel–Crafts cyclisation of aryl-substituted propargyl alco-
hols.[30]
Table 1. The optimisation of the de-O-benzylation conditions.[a]
Iodine
[equiv]
Triethylsilane
[equiv]
T
[8C]
t
Isolated yield of 2
[%]
G
E
[min]
1
2
3
4
5
6
7
0.5
0.5
1.0
1.25
1.25[b]
1.25
1.25
0.5
1.0
1.0
1.25
1.25[b]
1.25
1.25
À20
60
60
50
60
30
15
15
29
À20 to À5
À20 to À15
À25 to 0
À20
35 (42)[c]
43
48
45 (67)[c]
À20 to À10
À20
60
50
[a] General conditions: Et3SiH was added at the indicated temperature
to a solution of 1 and I2 in anhydrous CH2Cl2. The reaction was quenched
at the final temperature by addition of pyridine. [b] I2 and Et3SiH were
premixed in CH2Cl2 and the resulting solution was added to 1 dissolved
in CH2Cl2. [c] % conversion.
These experiments show the effectiveness of the simplest
conceivable experimental procedure, the addition of Et3SiH
to a preformed solution of I2 and the substrate in CH2Cl2.
Other solvents, such as toluene, 1,2-dichloroethane, acetoni-
trile and dioxane, were found to be much less suitable and
slower reactions were observed. Investigation into the stoi-
chiometry of the reagents revealed that a slight excess of
both is required to obtain high conversion of the starting
material (compare Table 1, entries 4–7 with entries 1–3).
Use of less than one equivalent of either reagent provided
incomplete de-O-benzylation (Table 1, entries 1 and 2), sug-
gesting that only HI, and not Et3SiI, is responsible for the
deprotection (see below for other evidence of this). Temper-
ature was found to be critical for maximising the yield while
allowing the reaction to proceed at an elevated rate. In fact,
TLC analysis suggested that further deprotection of the
mono-de-O-benzylated products was competitive with the
initial de-O-benzylation step of the fully protected saccha-
ride reagent. Thus, temperature control was pivotal for
tuning the course of these potential processes. Under the
optimised conditions (Table 1, entry 6), d-galacto precursor
1 provided 4-OH derivative 2 in an appreciable isolated
yield (60%) after only 15 min at a temperature of between
À20 and À108C. Upon further warming, the yield of 2 de-
creased owing to formation of larger amounts of products
The reaction of Et3SiH and I2 provides HI and Et3SiI
under very simple experimental conditions.[31] Given that HI
and Me3SiI are known reagents for ether cleavage,[19,32,33] we
were prompted to explore the feasibility of applying the
convenient I2/Et3SiH system to the deprotection of carbohy-
drates, with a special focus on the difficult issue of the regio-
selective de-O-benzylation of sugars.
Results and Discussion
In preliminary experiments, we examined the behaviour of
tetra-O-benzylated methyl a-glycopyranosides in the pres-
ence of a slight excess of I2 and Et3SiH in CH2Cl2; the regio-
selective de-O-benzylation of these substrates immediately
5882
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 5881 – 5889

Regioselective De-O-benzylation
FULL PAPER
that had undergone multiple de-O-benzylations. In compari-
son with other known procedures for the regioselective de-
O-benzylation of sugars, this approach looked advantageous
for several reasons, which include the reaction rapidity, the
use of a slight excess of cheap and easy-to-handle reagents,
and the non-demanding experimental precautions (no inert
atmosphere was required for high yields to be achieved).
The procedure was then applied to gluco- and manno-pre-
cursors 3 and 4 (Table 2). A predominance of the 4-OH re-
gioisomer was again recorded but in both cases the corre-
scope, especially in view of the wider chemical versatility of
allyl aglycons. The stability of allyl groups installed on non-
anomeric carbinols also led to an improvement in the regio-
selectivity of the de-O-benzylation of manno-configured
sugars, for which partially O-allylated precursors can be ef-
fectively prepared by stannylidene-mediated protection.[34]
As shown in Table 3, entries 2 and 3, 4-O deprotection of 3-
O-allylated derivatives 11 and 13 occurred in satisfactory
yields and high regioselectivity; unlike the behaviour of the
corresponding 3-O-benzylated counterpart
4
(Table 2,
entry 2), no products deprotected at O-3 were detected
from these 3-O-allylated precursors.
Table 2. De-O-benzylation of tetra-O-benzylated gluco- and manno-pyr-
The result obtained for allyl lactoside 15 was rather sur-
prising (Table 3, entry 4), as it was quickly deprotected in
very good yield at O-3 (gluco residue) rather than at the ex-
pected O-4’ position (galacto residue). This outcome broke
the trend inferred from Tables 1 and 2 (4-OH products in all
cases); the previously observed higher reactivity of glucose
compared with galactose (compare activation temperatures
in Tables 1 and 2) was initially considered in order to ac-
count for the apparently anomalous result. However, this
hypothesis was immediately contradicted by results obtained
from other disaccharide precursors 17, 19 and 21 (Table 3,
entries 5–7); regardless of the nature of the saccharide com-
ponents, these compounds were selectively deprotected at
the secondary carbinol of the reducing terminus adjacent to
the glicoside bond. Very interestingly, 3-O-linked manno-
disaccharide 21 (Table 3, entry 7) was preferentially depro-
tected at the axial O-2 position, oriented cis to the encum-
bered 3-O-glycosylated position, rather than at O-4. Nota-
bly, the deprotection yields of disaccharides were typically
higher than for the monosaccharide precursors despite the
larger number of potential deprotection events. Also worthy
of note is the stability of the glycoside linkages on the disac-
charides under the acidic conditions employed, despite the
“arming” effect of the benzyl protecting groups, which is ex-
pected to facilitate acid-promoted scission of inter-saccha-
ride linkages.[35]
anoside precursors.[a]
Reagent
t
T
Products, combined yield
A
(conversion), regioisomeric ratio
À55 to
À30
1
2
70
75
À40 to
À10
[a] General conditions: Et3SiH (1.25 equiv) was added to a solution of 3
or 4 and I2 (1.25 equiv) in anhydrous CH2Cl2 at the starting temperature.
The reaction was quenched at the final temperature by addition of pyri-
dine.
sponding 3-OH regioisomer was isolated with the main
product, as an inseparable mixture. The regioselectivity
ratio was strongly dependent on the precursor and was
much higher for manno precursor 4 (7/8, 8:1) than for gluco
derivative 3 (5/6, 3.8:1). Interestingly, in these cases, de-O-
benzylation started at lower temperatures than with the ga-
ACHTUNGTRENNUNGlacto precursor, but gradual warming of the reaction
medium ensured completion in relatively short times.
Although the yields of these preliminary experiments
might not appear especially high, it should be noted that
standard routes to 4-OH compounds 2, 5 and 7 from the
corresponding methyl glycosides would require two addi-
tional synthetic steps (initial 4,6-O-benzylidenation followed
As with monosaccharide and disaccharide precursors, fur-
anoside derivatives displayed a peculiar regioselectivity
trend. Glucofuranoside 23 (Table 3, entry 8) was selectively
de-O-benzylated to give 5-OH regioisomer 24 in a compara-
ble yield (58%) to that obtained with TiCl4.[7b] This was the
only case found in which the Et3SiH/I2 system gave both a
yield and regioselectivity similar to previously established
procedures. Mannofuranoside 25 (Table 3, entry 9) was also
debenzylated at O-5, but in this case the putative furanoside
intermediate deprotected at O-5 evolved to give pyranoside
26,[36] which has a free OH at C-4, and minor amounts of an
unidentified product (ca. 10%) with neither a free hydroxyl
group nor the allyl aglycon.[37]
by regioselective reductive ring opening) with chroma
ACHTUNGTNERtNUNG og-
ACHTUNGTRENNUNGraphic purifications. Additionally, it is pertinent that the re-
cently reported attempt at regioselective deprotection of 4
with DDQ gave an inseparable mixture of 7 and 8 in a
lower yield (31%) and regioselectivity (7:8, less than
3:1),[16a] thus showing the potential of the approach present-
ed herein with respect to the current state of the art of re-
gioselective de-O-benzylation of carbohydrates.
Other saccharide precursors bearing multiple benzyl pro-
tecting groups were then screened and further interesting re-
sults emerged (Table 3). Similar to a-methyl glycosides
(Tables 1 and 2), a-allyl glycosides proved to be compatible
with this protocol (compare Table 3, entry 1 with Table 1,
entry 6), thus offering a useful extension to the reaction
At this stage, a general survey of the obtained results and
their comparison with previously reported approaches indi-
cated an unprecedented direction of regioselectivity. As
mentioned above, only the O-4 deprotection of monosac-
charides resembles the results described by Crich (stoichio-
metric DDQ), the applicability of which, however, is re-
Chem. Eur. J. 2011, 17, 5881 – 5889
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5883

A. Iadonisi et al.
Table 3. Regioselective de-O-benzylation of pyranoside or furanoside monosaccha-
ACHTUNGTRENNUNG
rides, and disaccharides.[a]
in high yield upon exposure to I2 (0.6 equiv) and
propandithiol (1.2 equiv; Scheme 2). Under these
conditions, the reaction rate and yield were both
similar to those observed with I2 and Et3SiH. This
outcome supports the hypothesis that in situ gener-
ated HI acts as the de-O-benzylation agent.[32] On
the other hand, the negligible contribution of Et3SiI
to the process is supported by the reported stability
at 08C of saccharide benzyl ethers upon exposure
to Me3SiI, which is supposedly more reactive in
ether cleavage reactions than Et3SiI, under the con-
ditions adopted for the synthesis of glycosyl iodides
of per-O-benzylated sugars.[39] Additionally, in sharp
contrast with the results found for the I2/Et3SiH
system, the regioselectivity of Me3SiI-mediated de-
O-benzylation leads to preferential modification of
the sterically more accessible primary positions.[19]
In fact, only in the reaction described in Table 3,
entry 3 did we isolate minor quantities (5%) of the
4-O-triethylsilylated ether of 14.
On the basis of this evidence and according to a
well-established general mechanism for ether scis-
sion,[32b] HI is expected to activate the benzyloxy
group by O-protonation to favour the subsequent
benzyl removal by nucleophilic attack of the iodide
anion and release of the alcohol. Consistently,
benzyl iodide was isolated as a by-product of the re-
actions.
Once we had identified the actual promoter of
the process, some consideration was given to ration-
alising the origin of the observed regioselectivity.
Unlike many of the previously reported methods
operating under Lewis acidic conditions, HI-in-
duced de-O-benzylation cannot be regioselectively
controlled by chelation effects. On the other hand,
we observe that in all examined cases the liberation
of poorly accessible carbinols was favoured and in
no case were O-benzylated primary positions signif-
icantly affected. Thus, a reasonable mechanistic
sketch could involve the initial HI-promoted O-pro-
tonation occurring reversibly at the benzyloxy
groups; the following de-O-benzylation step, pro-
moted by the nearby iodide anion, is somehow ki-
netically controlled by the relief of the steric strain
associated with the expulsion of the benzyl group.
This effect could account for both the preferential
deprotection (Tables 1 and 2) at the O-4 position,
Reagent
T [8C]
t
Product, isolated
yield (conversion)
À40
to
1
50 min
À25
À60
2
3
2.5 h
to 0
À65
2.5 h
to À5
4
5
À15
25 min
À60
to
1 h
À20
À60
to
À20
6
7
1.5 h
À65
to
2.5 h
+10
À60
to
À20
8
9
1 h
À60
to
75 min
À20
[a] General conditions: Et3SiH (1.25 equiv) was added to a solution of the benzylated
sugar and I2 (1.25 equiv) in anhydrous CH2Cl2 at the starting temperature. The reac-
tions were quenched by addition of pyridine (entries 1–3) or solid NaHCO3 (entries 4–
9; 2.5–5 equiv) at the final temperature.
stricted to manno- and rhamno-configured substrates to
date.[16a]
To better explain the origin of the regioselectivity ob-
tained, investigations were carried out to ascertain whether
HI was actually the agent responsible for the deprotection.
To this end, galacto precursor 1 was treated with a mixture
of I2 and propandithiol, in accordance with the procedure
described some years ago by Koreeda and co-workers for
the in situ generation of anhydrous HI.[38] As expected, the
Scheme 2. Regioselective de-O-benzylation of model compound 1 with I2
reaction once more afforded free 4-OH galacto derivative 2
(0.6 equiv) and propandithiol (1.2 equiv).
5884
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 5881 – 5889

Regioselective De-O-benzylation
FULL PAPER
Table 4. Regioselective de-O-benzylation of 6-deoxy glycopyranosides
the most encumbered for pyranoside monosaccharides, and
the highest regioselectivity associated with removal of the
axially oriented galactose 4-O-benzyl group (compare
Tables 1 and 2). Furthermore, the apparently anomalous be-
haviour of disaccharide derivatives (Table 3, entries 4–7) is
consistent with this hypothesis if the steric crowding around
a benzylated carbinol group flanking a glycosidation site is
acknowledged. Similar reasoning can also be applied to ra-
tionalise the liberation of the 5-OH in furan derivatives
(Table 3, entries 8 and 9), which, in the case of entry 8, for-
tuitously provides a result similar to a de-O-benzylation oc-
curring under chelation control. Some cases of acid-promot-
ed ether or acetal scission that appear in the literature seem
to be governed by the “steric-strain release effect” invoked
here.[40]
and monosaccharide b-pyranosides.[a]
Reagent
T
[8C]
t
Quenching
conditions
Products, yield
(conversion)
À60
1
2
2 h
A
B
to 0
27
27
À60 2 h
28 (a/b 1)
40% (50%)
to 0
À60
2 h
3
C
to 0
To further extend the substrate scope of this reaction,
other kinds of precursor, such as 2,3,4-tri-O-benzylated 6-
deoxy pyranosides and b-configured 2,3,4,6-tetra-O-benzy-
lated glycopyranosides, were exposed to the I2/Et3SiH
system. Initial experiments were frustrating as complex mix-
tures were observed by TLC analysis. However, upon pyri-
dine or lutidine quenching, the TLC profile of the reaction
mixtures became simpler and most products displayed
higher mobility than during the course of the reaction. This
apparently strange observation was ascribed to the in situ
generation of glycosyl iodide intermediates as a conse-
quence of the higher anomeric reactivity of 6-deoxyglyco-
sides and b-O-linked glycosides, which could result in a fast
HI-promoted expulsion of the aglycon alcohol. Glycosyl io-
dides of “armed” densely O-benzylated sugars are not very
stable, so that they are probably degraded under TLC elu-
tion conditions to give more polar hydrolysis products. Ex-
periments conducted with varied amounts of I2 and Et3SiH
showed that anomeric iodination was faster than de-O-ben-
zylation at any site, and thus part of the I2/Et3SiH reagent
was consumed for this process. Interestingly, quenching of
the reaction with bases as mild as pyridine or lutidine in-
duced reattachment of the initial aglycon alcohol with poor
selectivity to give mixtures of a- and b-glycosides starting
from anomerically pure substrates. These observations led
us to foresee that addition of sufficient I2 and Et3SiH could
induce concomitant de-O-benzylation together with the
faster anomeric iodination. Thus, upon quenching, the re-
À60
1 h
4
5
27
D
C
to 0
À65
2 h
to 0
À65
6[b]
31
2 h
D
to 0
À25
to
7[b]
1 h
D
À5
À60
to
À10
8
9
80 min
40 min
A
D
À30
to
À15
A
ACHTUNGTNERmNUNG er-
ACHTUNGTRENNUNG
[a] General conditions: Et3SiH (1.8 equiv) was added at the starting tem-
perature to a solution of the benzylated sugar and I2 (1.8–2.1 equiv) in
anhydrous CH2Cl2. The mixture was allowed to warm up to the final tem-
perature. The reaction was quenched at the final temperature according
to the following conditions A–D, and then the reaction vessel was al-
lowed to warm to RT. Conditions A: allyl alcohol (5 equiv) and NaHCO3
(5 equiv); B: allyl alcohol (5 equiv) and lutidine (5 equiv); C: saturated
aqueous sodium carbonate; D: a solution of lutidine (7 equiv) and acetic
acid (4 equiv) in CH2Cl2. [b] Products 33 and 37 were slightly contaminat-
ed by an unidentified regioisomer.
synthetic elaboration of the sugar entails subsequent remov-
al of the aglycon, as in, for example, the preparation of gly-
cosyl donors for oligosaccharide synthesis.
After several experiments the use of 1.8 equivalents of
both I2 and Et3SiH emerged as the best option for perform-
ing the regioselective de-O-benzylation of sugars that are es-
pecially reactive at the anomeric positions. Application of
these conditions to l-rhamno precursor 27 (Table 4, en-
tries 1–4) resulted in regioselective deprotection at O-3,
whereas O-4 deprotection had previously been recorded in
the presence of the bulkier benzyloxy group at C-6 (Table 2,
entry 2) or under the DDQ methodology.[16a] Different
quenching conditions (Table 4, conditions A–D) were also
developed in order to tune the anomeric composition of the
product, or even to change the nature of the anomeric sub-
Chem. Eur. J. 2011, 17, 5881 – 5889
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5885

A. Iadonisi et al.
Table 5. Regioselective de-O-benzylation of 4-O-acylated galactopyrano-
stituent. These data show that reattachment of the aglycon
alcohol can proceed with variable stereoselectivity according
to the nature of the base (Table 4, entries 1 and 2). On the
other hand, hydrolysis of the glycosyl iodide was also carried
out by quenching with saturated aqueous sodium bicarbon-
ate (Table 4, entry 3) even though partial reattachment of
the original alcohol aglycon was also recorded. Even more
interestingly, the aglycon of the starting compound could be
exchanged for a synthetically versatile acetate group upon
quenching with a premixed solution of acetic acid and luti-
dine in CH2Cl2 (Table 4, entry 4). The quenching with pyri-
dine, applied in our initial experiments on precursors unaf-
fected at the anomeric position by I2/Et3SiH (Tables 1–3),
was soon abandoned after consistently isolating glycosyl pyr-
idinium species that reduced the overall yield of the process.
Application of these procedures to l-fucose confirmed the
greater amenability of 6-deoxy sugars to O-3 deprotection
with concurrent modification at the anomeric position
(Table 4, entries 5–6). Interestingly, per-O-benzylated b-glu-
coside 34 and b-galactoside 38 also revealed anomeric re-
sides.[a]
Reagent
t
T
Product,
A
[8C]
yield [%]
1
2
90
45
À20 to À5
À20 to À5
[a] General conditions: Et3SiH (1.25 equiv) was added to a solution of 41
or 43 and I2 (1.25 equiv) in anhydrous CH2Cl2 at the starting temperature.
The reaction was quenched at the final temperature by addition of pyri-
dine. Bz=benzoyl, TES=triethylsilyl.
also gave the 6-O-debenzylation product in 12% yield, as
well as minor amounts (less than 10%) of the 3-O-depro-
tected derivative. This was the only case in which liberation
of the primary carbinol was observed to an appreciable
extent. Starting from compound 43 (Table 5, entry 2), the
liberation of the 2-OH was partially accompanied by 6-O-
triethylsilylation (17%). As mentioned above, triethylsilyla-
tion of benzylated substrates is an uncommon event upon
treatment with I2/Et3SiH. The directing effect of acyl groups
on regioselectivity further expands the synthetic potential of
this method; selective 2-O-deprotection, which has never
been observed when starting from per-O-benzylated mono-
saccharides (Tables 1, 2 and 3), are now possible starting
from 4-O-acylated precursors.
ACHTUNGTRENNUNGactivity towards I2 and Et3SiH, as shown by the aglycon ex-
change in Table 4, entry 7 and the partial anomerisation in
Table 4, entry 8. The regioselectivity of the de-O-benzylation
was consistent with that of the corresponding a-pyranoside
precursors (compare with Tables 1 and 2), and small
amounts of 3,4-diol 37 (slightly contaminated by another 1-
O-acetylated derivative) were also isolated from the gluco
precursor. The 1-O-acetylated galacto derivative 39 (Table 4,
entry 9), which is prone to anomeric iodination with I2/
Et3SiH,[23] also reacted with a regioselectivity mirroring the
behaviour of the corresponding a-methyl glycoside
1
(Table 1). At this stage it is interesting to recall that no evi-
dence of HI-promoted de-O-allylation was recorded for b-
O-allylated disaccharides 15, 17, and 19 (Table 3, entries 4–
6). This evidence highlights that de-O-benzylation of disac-
charide substrates is much faster than their anomeric iodina-
tion, in stark contrast with the behaviour of monosaccharide
b-glycosides.
The usefulness of this methodology as a tool to streamline
the synthesis of biologically useful oligosaccharides was
demonstrated by use of acceptors 2 and 16, rapidly prepared
in high yields as shown above. Coupling of acceptor 2 with
galactosyl trifluoroacetimidate[41] donor 46 under catalytic
[42]
activation by Bi
G
quickly afforded protected gala-
A further extension of the reaction scope was pursued by
examining the effect of acyl protecting groups on substrates
containing multiple benzyl groups. The electron withdrawing
properties of acyl groups were expected to reduce the basic
character of adjacent benzylated oxygen atoms, interfering
with the preliminary protonation of the benzyloxy group
necessary to trigger the iodide-induced benzyl removal. This
effect could switch the normal regioselectivity (controlled
by steric factors) to favour deprotection of carbinol sites fur-
ther from the acyloxy groups. To test this effect, compound
2 was acetylated and benzoylated under standard conditions
to yield derivatives 41 and 43 (Table 5). Both derivatives
were then exposed to the I2/Et3SiH system at low tempera-
ture. Interestingly, the regioselectivity of the process was sig-
nificantly influenced by the presence of the ester functional-
ity, the benzyl group further away (at O-2) was preferential-
ly removed in synthetically useful yields (Table 5, entries 1
and 2). In these cases minor amounts of pure by-products
were isolated. In particular, compound 41 (Table 5, entry 1)
biose 47 (Scheme 3) in a very high yield and with a selectivi-
ty. Galabiose is contained in a range of biologically relevant
oligosaccharide sequences, and this disaccharide fragment is
frequently investigated for its potential role in tuning the
adhesion of pathogenic bacteria.[43] In another application,
the de-O-benzylation procedure was incorporated into a
one-pot sequence yielding Lewis X mimic 49, in which the
glucosamine residue is replaced by glucose (Scheme 3).[44]
For this purpose, fully protected lactose 15 was deprotected
at O-3 as shown in Table 3, entry 4, then fucosyl imidate
48[45] was added directly to the reaction vessel with addition-
al iodine.[20,46,47] The procedure afforded 49 in an acceptable
31% overall yield and, to the best of our knowledge, this is
the first example in which a benzyl ether cleavage has been
applied to a one-pot sequence for oligosaccharide synthesis.
Indeed, reductive opening of benzylidenes is the most fre-
quently pursued strategy for liberating a carbinol position in
one-pot protocols of carbohydrate derivatisation and oligo-
saccharide synthesis.[5,48]
5886
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 5881 – 5889

Regioselective De-O-benzylation
FULL PAPER
er, the intermediate glycosyl io-
dides generated are reactive
enough to be converted into a
variety of synthetically useful
products upon suitable quench-
ing.
As demonstrated by the syn-
thesis of Lewis X mimic 49, the
de-O-benzylation protocol can
also be incorporated into one-
pot procedures for oligosacchar-
ide synthesis with an unprece-
dented use of the benzyl group
as a transient protecting group.
In addition to its wide applic-
ability, the disclosed protocol
exhibits a broad range of practi-
cal advantages over many of
the current methodologies for
regioselective
de-O-benzyla-
tion; reactions proceed rapidly
at low temperatures and are
performed without adopting
inert atmospheres and the re-
agents are cheap and easy to
handle and can be used in slight
stoichiometric excesses.
Scheme 3. Application of regioselectively deprotected products in the synthesis of biologically relevant oligo-
saccharides (All=allyl; AW MS=acid washed molecular sieves).
Conclusion
In this paper, we have demonstrated that the I2/Et3SiH com-
bined system is a very useful reagent for the regioselective
de-O-benzylation of highly O-benzylated substrates. The HI
generated in situ appears to be the actual promoter of the
process. Synthetically useful yields are frequently achieved
to give partially protected building blocks that are otherwise
accessible only through longer synthetic sequences. The ob-
tained results show that sterically hindered benzyl groups
are preferentially removed and the relief of steric strain ap-
pears to be the main factor influencing the regioselectivity.
This effect could explain other known selective processes of
ether and acetal scission that occur under acidic conditions
and might be exploited in the design of new, selective, acid-
promoted processes. Consistent with this mechanistic view,
each group of substrates examined displayed a well defined
regioselectivity trend; tetra-O-benzylated hexopyranoside
monosaccharides are preferentially deprotected at O-4,
Experimental Section
General procedure for de-O-benzylation with I2 and Et3SiH: Anhydrous
CH2Cl2 (16–20 mLmmolÀ1 I2) was added to the poly-O-benzylated deriv-
ative and I2 at room temperature (see Tables for the relative amounts of
reagents). After complete dissolution, the mixture was cooled to the
starting temperature of the reaction (see Tables) and then Et3SiH (see
Tables for the relative amount) was added by syringe. The reaction mix-
ture was allowed to gradually warm or maintained at the same tempera-
ture (see Tables) until TLC analysis displayed the optimal extent of con-
version. The reaction was quenched as indicated in the Tables. For condi-
tions A (Table 4), allyl alcohol (5 equiv) and solid NaHCO3 (5 equiv)
were added sequentially. For conditions B, allyl alcohol (5 equiv) and lu-
tidine (5 equiv) were added sequentially. For conditions C, saturated
aqueous sodium carbonate (5 mLmmolÀ1 I2) was added. For conditions
D, a solution of lutidine (7 equiv) and acetic acid (4 equiv) in CH2Cl2
(3 mLmmolÀ1 I2) was added. The mixture was allowed to warm to RT
and then diluted with CH2Cl2. The organic phase was washed with water
containing sodium thiosulfate (to remove residual amounts of iodine).
The aqueous phase was re-extracted with CH2Cl2, and the collected or-
ganic phases were dried and concentrated in vacuo. Silica-gel flash
column chromatography (eluent: hexane/ethyl acetate mixtures) yielded
the de-O-benzylated products reported in the tables.
hexoACHTUNGTRENNUNGfuranosides at O-5, 6-deoxy pyranosides at O-3, per-O-
benzylated disaccharides at a secondary carbinol site of the
reducing terminus adjacent to the glycoside bond. Interest-
ingly, the regioselectivity of the process can also be switched
by the insertion of acyl protecting groups on the substrate
that result in de-O-benzylation at the site farthest from the
acACHTUNGTRENNUNG
Acknowledgements
This work was supported by a grant from the Italian Ministry of Univer-
sity (PRIN 2008 project) and a grant from Naples University (Pro-
gramma FARO 2009). NMR and MS facilities of the Centro Interdiparti-
mentale di Metodologie Chimico-Fisiche dellꢂUniversitꢀ di Napoli
(CIMCF) are acknowledged.
stoichiometry) to substrates that are especially reactive at
the anomeric position (6-deoxy glycopyranosides and mono-
saccharide b-pyranosides) causes rapid anomeric iodination
at low temperature alongside the de-O-benzylation. Howev-
Chem. Eur. J. 2011, 17, 5881 – 5889
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5887

A. Iadonisi et al.
[18] a) R. Allerton, Jr., H. G. Fletcher, J. Am. Chem. Soc. 1954, 76,
1757–1760; b) B. Ganem, V. R. Small, Jr., J. Org. Chem. 1974, 39,
3728–3730; c) R. Eby, S. J. Sondheimer, C. Schuerch, Carbohydr.
Res. 1979, 73, 273–276; d) J.-I. Sakai, T. Takeda, Y. Ogihara, Carbo-
hydr. Res. 1981, 95, 125–131; e) P. Angibeaud, J.-P. Utille, Synthesis
1991, 737–738; f) G. Yang, X. Ding, F. Kong, Tetrahedron Lett. 1997,
38, 6725–6728; g) A. Brar, Y. D. Vankar, Tetrahedron Lett. 2006, 47,
5207–5210.
[19] A. Klemer, M. Bieber, H. Wilbers, Liebigs Ann. Chem. 1983, 1416–
1421.
[20] M. Adinolfi, G. Barone, A. Iadonisi, M. Schiattarella, Synlett 2002,
269–270.
[21] a) H. Tanaka, A. Yoshizawa, T. Takahashi, Angew. Chem. 2007, 119,
2557–2559; Angew. Chem. Int. Ed. 2007, 46, 2505–2507; b) H.
Tanaka, A. Yoshizawa, S. Chijiwa, J. Ueda, M. Takagi, K. Shin-Ya,
T. Takahashi, Chem. Asian J. 2009, 4, 1114–1125; c) H. Tanaka, S.
Yamaguchi, A. Yoshizawa, M. Takagi, K. Shin-ya, T. Takahashi,
Chem. Asian J. 2010, 5, 1407–1424.
[22] M. Adinolfi, L. De Napoli, G. Di Fabio, A. Iadonisi, D. Montesarch-
io, G. Piccialli, Tetrahedron 2002, 58, 6697–6704.
[23] M. Adinolfi, A. Iadonisi, A. Ravidꢀ, M. Schiattarella, Tetrahedron
Lett. 2003, 44, 7863–7866.
[24] S. Valerio, A. Iadonisi, M. Adinolfi, A. Ravidꢀ, J. Org. Chem. 2007,
72, 6097–6106.
[25] A. Pastore, M. Adinolfi, A. Iadonisi, Eur. J. Org. Chem. 2008, 6206–
6212.
[26] M. Adinolfi, A. Iadonisi, A. Pezzella, A. Ravidꢀ, Synlett 2005,
1848–1852.
[27] A. Pezzella, A. Iadonisi, S. Valerio, L. Panzella, A. Napolitano, M.
Adinolfi, M. dꢂIschia, J. Am. Chem. Soc. 2009, 131, 15270–15275.
[28] R. Panchadhayee, A. K. Misra, Synlett 2010, 1193–1196.
[29] J. S. Yadav, S. B. V. Reddy, K. Premalatha, T. Swamy, Tetrahedron
Lett. 2005, 46, 2687–2690.
[1] a) T. W. Greene, P. G. M. Wuts, Greeneꢀs Protective Groups in Or-
ganic Synthesis, 4th ed., Wiley, New York, 2006, pp. 102–120.
[2] K. C. Nicolaou, H. J. Mitchell, Angew. Chem. 2001, 113, 1624–1672;
Angew. Chem. Int. Ed. 2001, 40, 1576–1624.
[3] B. Fraser-Reid, Z. Wu, U. E. Udodong, H. B. Ottosson, J. Org.
Chem. 1990, 55, 6068–6070; for a recent review of oligosaccharide
synthesis, see: X. Zhu, R. R. Schmidt, Angew. Chem. 2009, 121,
1932–1967; Angew. Chem. Int. Ed. 2009, 48, 1900–1934.
[4] a) A. V. Demchenko, Synlett 2003, 1225–1240; b) L. K. Mydock,
A. V. Demchenko, Org. Biomol. Chem. 2010, 8, 497–510.
[5] a) C.-C. Wang, J.-C. Lee, S.-Y. Luo, S. S. Kulkarni, Y.-W. Huang, C.-
C. Lee, K.-L. Chang, S.-C. Hung, Nature 2007, 446, 896–899; b) A.
Francais, D. Urban, J.-M. Beau, Angew. Chem. 2007, 119, 8816–
8819; Angew. Chem. Int. Ed. 2007, 46, 8662–8665; c) C.-C. Wang,
S. S. Kulkarni, J.-C. Lee, S.-Y. Luo, S.-C. Hung, Nat. Protoc. 2008, 3,
97–113.
[6] a) D. Beaupere, I. Boutbaiba, A. Wadouachi, C. Frechou, G. De-
mailly, R. Uzan, New J. Chem. 1992, 16, 405–411; b) T. Bieg, W.
Szeja, Carbohydr. Res. 1990, 205, C10–C11; c) D. Beaupere, I. Bout-
baiba, G. Demailly, R. Zan, Carbohydr. Res. 1988, 180, 152–155;
d) M. del Carmen Cruzado, M. Martin-Lomas, Tetrahedron Lett.
1986, 27, 2497–2500.
[7] a) O. Martin, K. G. Kurz, S. P. Rao, J. Org. Chem. 1987, 52, 2922–
2925; b) H. Hori, Y. Nishida, H. Ohrui, H. Meguro, J. Org. Chem.
1989, 54, 1346–1353; c) A. Bouali, P. Boullanger, D. Lafont, B.
Fenet, Carbohydr. Res. 1992, 228, 81–93.
[8] a) C. Jia, Y. Zhang, L.-H. Zhang, P. Sinay, M. Sollogoub, Carbohydr.
Res. 2006, 341, 2135–2144; b) T. Lecourt, A. Herault, A. J. Pearce,
M. Sollogoub, P. Sinay, Chem. Eur. J. 2004, 10, 2960–2971; c) B. C.
Roizel, E. Cabianca, P. Rollin, P. Sinay, Tetrahedron 2002, 58, 9579–
9583.
[9] For applications to benzylated cyclodextrins, see: a) A. J. Pearce, P.
Sinay, Angew. Chem. 2000, 112, 3756–3758; Angew. Chem. Int. Ed.
2000, 39, 3610–3612; b) O. Bistri, P. Sinay, M. Sollogoub, Tetrahe-
dron Lett. 2005, 46, 7757–7760; c) O. Bistri, P. Sinay, J. Jimenez-Bar-
bero, M. Sollogoub, Chem. Eur. J. 2007, 13, 9757–9774; d) S. Guieu,
M. Sollogoub, J. Org. Chem. 2008, 73, 2819–2828; e) O. Lꢃpez, M.
Bols, Tetrahedron 2008, 64, 7587–7593; f) S. Guieu, M. Sollogoub,
Angew. Chem. 2008, 120, 7168–7171; Angew. Chem. Int. Ed. 2008,
47, 7060–7063; g) G. K. Rawal, S. Rani, C. C. Ling, Tetrahedron
Lett. 2009, 50, 4633–4636; h) G. Rawal, S. Rani, S. Ward, C.-C. Ling,
Org. Biomol. Chem. 2010, 8, 171–180; i) S. Guieu, E. Zaborova, Y.
Bleriot, G. Poli, A. Jutand, D. Madec, G. Prestat, M. Sollogoub,
Angew. Chem. 2010, 122, 2364–2368; Angew. Chem. Int. Ed. 2010,
49, 2314–2318.
[10] E. Zaborova, Y. Bleriot, M. Sollogoub, Tetrahedron Lett. 2010, 51,
1254–1256.
[11] J. Xie, M. Menand, J.-M. Valery, Carbohydr. Res. 2005, 340, 481–
487.
[12] J. R. Falck, D. K. Barma, S. K. Venkataraman, R. Baati, C. Mios-
kowski, Tetrahedron Lett. 2002, 43, 963–966.
[13] a) M. Kawana, Chem. Lett. 1981, 1541–1544; b) C. P. Casas, H. Ya-
mamoto, T. Yamato, J. Chem. Res. 2005, 694–696; c) G. G. Haralds-
son, J. E. Baldwin, Tetrahedron 1997, 53, 215–224.
[14] a) J. Madsen, M. Bols, Angew. Chem. 1998, 110, 3348–3350; Angew.
Chem. Int. Ed. 1998, 37, 3177–3178; b) J. Madsen, C. Viuf, M. Bols,
Chem. Eur. J. 2000, 6, 1140–1146.
[15] a) L. Cipolla, L. Lay, F. Nicotra, J. Org. Chem. 1997, 62, 6678–6691;
b) L. Lay, L. Cipolla, B. La Ferla, F. Peri, F. Nicotra, Eur. J. Org.
Chem. 1999, 3437–3440; c) L. Cipolla, B. La Ferla, L. Lay, F. Peri, F.
Nicotra, Tetrahedron: Asymmetry 2000, 11, 295–303.
[16] a) D. Crich, O. Vinogradova, J. Org. Chem. 2007, 72, 3581–3584;
b) for previous examples of the DDQ-mediated regioselective de-O-
benzylation of sugars, see: E. Cabianca, A. Tatibouet, P. Rollin, Pol.
J. Chem. 2005, 79, 317–322.
[17] Z.-J. Yin, B. Wang, Y.-B. Li, X.-B. Meng, Z-J. Li, Org. Lett. 2010, 12,
536–539.
[30] B. V. Subba Reddy, B. Brahma Reddy, K. V. Raghavendra Rao, J. S.
Yadav, Tetrahedron Lett. 2010, 51, 5697–5700.
[31] D. R. Deans, C. Eaborn, J. Chem. Soc. 1954, 3169–3173.
[32] For the use of HI in ether cleavage, see: a) C. A. Smith, J. B. Grutz-
ner, J. Org. Chem. 1976, 41, 367–370; b) V. M. Bhatt, S. U. Kulkarni,
Synthesis 1983, 249–282.
[33] For the use of TMSI in ether cleavage, see: M. E. Jung, M. A.
Lyster, J. Org. Chem. 1977, 42, 3761–3764.
[34] G. B. Yang, F.-Z. Kong, S. H. Zhou, Carbohydr. Res. 1991, 211, 179–
182.
[35] Per-O-benzylated saccharose, containing a more acid-labile glyco-
side linkage, decomposed upon exposure to iodine and Et3SiH at
low temperature.
[36] M. Adinolfi, D. Giacomini, A. Iadonisi, A. Quintavalla, S. Valerio,
Eur. J. Org. Chem. 2008, 2895–2899.
[37] In a reasonable hypothesis, the 5-OH furanoside intermediate could
be involved in an intramolecular trans-acetalation with expulsion of
allyl alcohol and generation of a 1,5-anhydro bicyclo structure.
[38] S. M. Chervin, P. Abada, M. Koreeda, Org. Lett. 2000, 2, 369–372.
[39] M. J. Hadd, J. Gervay, Carbohydr. Res. 1999, 320, 61–69.
[40] a) E. H. Vickery, L. F. Pahler, E. J. Eisenbraun, J. Org. Chem. 1979,
44, 4444–4446; b) J. L. Hendel, J.-W. Wang, T. A. Jackson, K. Hard-
meier, R. De Los Santos, F.-I. Auzanneau, J. Org. Chem. 2009, 74,
8321–8331.
[41] B. Yu, J. Sun, Chem. Commun. 2010, 46, 4668–4679.
[42] M. Adinolfi, A. Iadonisi, A. Ravidꢀ, S. Valerio, Tetrahedron Lett.
2006, 47, 2595–2599.
[43] a) P. I. Kitov, H. Shimizu, S. W. Homans, D. R. Bundle, J. Am.
Chem. Soc. 2003, 125, 3284–3294; b) J. A. F. Joosten, V. Loimaranta,
C. C. M. Appeldoorn, S. Haataja, F. A. E. Maate, R. M. J. Liskamp,
J. Finne, R. J. Pieters, J. Med. Chem. 2004, 47, 6499–6508; c) A. Sal-
minen, V. Loimaranta, J. A. F. Joosten, A. S. Khan, J. Hacker, R. J.
Pieters, J. Finne, J. Antimicrob. Chemother. 2007, 60, 495–501.
[44] For other examples of synthetic efforts towards analogous Lewis X
mimics, see: a) T. Tsukida, M. Yoshida, K. Kurokawa, Y. Nakai, T.
Achiba, T. Kiyoi, H. Kondo, J. Org. Chem. 1997, 62, 6876–6881;
5888
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 5881 – 5889

Regioselective De-O-benzylation
FULL PAPER
b) A. Asnani, F.-I. Auzanneau, Carbohydr. Res. 2008, 343, 1653–
1664.
[47] M. Adinolfi, A. Iadonisi, A. Ravidꢀ, M. Schiattarella, Synlett 2004,
275–278.
[45] M. Adinolfi, A. Iadonisi, A. Ravidꢀ, M. Schiattarella, J. Org. Chem.
2005, 70, 5316–5319.
[48] Y. Vohra, M. Vasan, A. Venot, G.-J. Boons, Org. Lett. 2008, 10,
3247–3250.
[46] K. P. Ravindranathan Kharta, T. S. Karkkainen, S. J. Marsh, R. A.
Field, Synlett 2001, 260–261.
Received: November 19, 2010
Published online: April 19, 2011
Chem. Eur. J. 2011, 17, 5881 – 5889
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5889