Regioselective green anomeric deacetylation catalyzed by lanthanide triflates

Article abstract of DOI:10.1016/j.tetlet.2008.01.106

Lanthanide triflates, especially Nd(OTf)₃, efficiently catalyze the regioselective transesterification of anomeric acetates. This method offers an efficient solution for the otherwise difficult removal of methyl uronates anomeric acetates as well as a green alternative to published protocols since the lanthanide catalysts are non-toxic and may be easily recycled and reused.

Full text of DOI:10.1016/j.tetlet.2008.01.106

Available online at www.sciencedirect.com
Tetrahedron Letters 49 (2008) 2163–2165
Regioselective green anomeric deacetylation catalyzed
by lanthanide triflates
Anh Tuan Tran, Sophie Deydier, David Bonnaff e´ , Christine Le Narvor *
LCOM-eG2M, Institut de Chimie Mol e´ culaire et des mat e´ riaux d’Orsay, CNRS UMR 8182, B aˆ t 420, Univ Paris-Sud, Orsay F-91405, France
Received 17 December 2007; revised 14 January 2008; accepted 18 January 2008
Available online 30 January 2008
Abstract
Lanthanide triflates, especially Nd(OTf) , efficiently catalyze the regioselective transesterification of anomeric acetates. This method
3
offers an efficient solution for the otherwise difficult removal of methyl uronates anomeric acetates as well as a green alternative to
published protocols since the lanthanide catalysts are non-toxic and may be easily recycled and reused.
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords: Anomeric deacetylation; Lanthanide triflates; Methyl uronates; Transesterification
Selective anomeric deacetylation is often a key step in
oligosaccharide synthesis commonly used for the prepara-
tion of anomeric trichloroacetimidate, one of the most used
glycosyl donor. A great number of reagents are reported
to perform this transformation. Most of them involve
selective transamidation using nitrogenous nucleophiles
low conversion and the more dramatic problems encoun-
tered with other nitrogen nucleophiles, may be due to side
reactions on the methoxycarbonyl group of 1. We reasoned
that such unwanted transamidation could be avoided by
using an acid transesterification. In this regard, HCl in
1
MeOH is known to allow selective hydrolysis of acetate
2
3
4
11
such as benzylamine, hydrazine, and piperidine. Alterna-
tive protocols involve transesterification catalyzed by
groups in the presence of benzoates,
however, we
doubted that such reagent could allow the regioselective
transesterification of anomeric acetates without competi-
tive formation of methyl glycoside.
5
organic tin oxides, solvolysis promoted by tin tetrachlo-
ride or conversion to an anomeric halogenide using HBr
followed by hydrolysis using silver salts. All these methods
6
7
Lanthanide triflate Lx(OTf) is used as mild Lewis acid
3
1
2
use stoichiometric amounts of reagents that are often toxic.
Moreover, when applied to compound 1 (an important
synthon involved in the synthesis of heparan sulfate frag-
in many organic transformations and selective deprotec-
tion of the bidentate methoxyacetyl group in the presence
of other monodentate ester groups has been achieved by
8
13
ments), none gave neither satisfactory nor reproducible
transesterification in MeOH using Yb(OTf) as a catalyst.
3
results. Indeed, conversion to the anomeric bromide with
We thus wondered whether Lx(OTf) salts could catalyze
the regioselective deacetylation of anomeric acetates espe-
cially on uronic derivatives.
3
9
10
TiBr4 followed by hydrolysis using mercuric salts gave
high yields on small scales but led to unreproducible yields
on larger ones. Transamidation with piperidine proved to
be the most reliable method giving the desired compound
In view of its importance in ongoing work in our labo-
1
4
ratory, methyl-1,2,3-tri-O-acetyl-3-O-benzyl-b-L-idopyra-
nuronate 1 was selected as a model to screen the efficiency
of different lanthanide triflates to catalyze the regioselective
transesterification of anomeric acetates with MeOH
(Scheme 1). Compound 1 was, thus, treated in MeOH with
5 mol % Yb(OTf) , Eu(OTf) , Sm(OTf) or Nd(OTf) . The
2
in reproducible but unsatisfactory 60% yield. Such a
*
Corresponding author. Tel.: +33 (0)169154719.
E-mail address: chlenarv@icmo.u-psud.fr (C. Le Narvor).
3
3
3
3
0
040-4039/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2008.01.106

2164
A. T. Tran et al. / Tetrahedron Letters 49 (2008) 2163–2165
OBn
O
OBn
O
OBn
OH
OH
OBn
O
O
Ln(OTf)3
MeOOC
OH
MeOOC
MeOOC
OAc
MeOOC
H
OAc
OAc
MeOH, 0º C
OAc
OAc
OAc
OAc
OAc
OAc
OBn
1
2
3
2
O
Scheme 1. Regioselective deacetylation of compound 1.
OMe
O
OAc
O
MeOOC
OAc
OH
H
OBn
reactions, followed by TLC led to the exclusive or highly
selective formation of a product identified as the hemi-
acetal 2. Isolated yields (Table 1) were determined after
purification by simple extraction followed by flash chroma-
tography, and showed that quasi-quantitative yields were
OH
OAc
4
OAc
5
Scheme 2. Proposed mechanism for the in situ formation of 5.
1
8
3
+
3+
3+
1,2,3,4-tetra-O-acetyl-a-D-galactopyranuronate 7
were
obtained with Nd and Sm triflates, while Eu and
3
+
treated in the conditions used for compound 1, without fur-
ther optimisation of the reaction conditions (Scheme 3).
Hemiacetals 8 and 9 were obtained in respective 88% and
5% isolated yields, showing the efficiency of the method
on other methyl uronates than 1. Such results are similar
Yb salts gave, respectively, 85 and 75% isolated yields.
To our surprise, the lower yields obtained with
Yb(OTf) and Eu(OTf)3 were not due to a diminished
regioselectivity, a lower conversion or an over-deacetyl-
ation, but to the rearrangement of the pyranosyl derivative
into the furanosyl isomer 5, whose structure was un-
ambiguously established by NMR (COSY and H– C
HSQC). Since methyl 1,2,5-tri-O-acetyl-3-O-benzyl-a,b-
L-idofuranonate has already been characterized, a small
3
7
to the one obtained when Bu SnO or Bu SnOMe are used
2
3
2
5
1
13
as catalyst. However, it is worth noting that, in this later
case, stoichiometric amounts of tin derivatives were used
and not recycled, whereas, in our case, 5 mol % of
1
5
9
1
2
1
13
recyclable and environmentally friendly Nd(OTf) was
3
amount of 5 was acetylated. The H and C NMR data
of the result product were identical to those already pub-
lished. As shown in Scheme 2, we propose that the isomeri-
sation of compound 2 into the furano derivative 5 proceeds
via a migration of an acetyl group between position 4 and
position 5 in the open form of these hemiacetals. It seems
sufficient.
To test whether the electronic effect of the carboxy-
methyl group in the uronate substrates has an impact on
the regioselectivity of the reaction, 1,2,3,4-tetra-O-acetyl-
b-D-xylopyranose 10 was used as the next substrate
3
+
3+
3+
(Scheme 4). As for compound 1, a set of Lx(OTf) salts
3
thus that Yb and Eu are better catalysts than Sm
3
+
were screened to test their efficiency in the anomeric
deacetylation reaction. No major difference between them
was detected since 80–85% isolated yields of 14 were
obtained irrespective of the catalyst used (Table 2). Thus,
as expected, the presence of a carboxymethyl group in
position 5 of a peracetylated sugar is not a prerequisite
or Nd in promoting such acetyl migration leading to a
greater amount of furanosyl compound 5. Therefore, it
may be proposed that the higher yields obtained with
Sm and Nd are due to their lower ability to catalyze
the unwanted isomerisation of pyranose 2 into furanose
3
+
3+
5
(
. A preparative reaction was then carried on 11 mmol
4.7 g) of compound 1 using Nd(OTf) and gave 2 in 90%
to the regioselectivity and efficiency of Lx(OTf) mediated
3
3
1
6
anomeric deacetylation.
yield. The treatment of the reaction is very simple and
the catalyst may be easily recycled from the aqueous phase
and reused, after evaporation of the water, at least 3 times
without loss of activity.
Having found an effective protocol for the preparation
of compound 2, we wondered whether other methyl uro-
nates could also be regioselectively deacetylated using
To test further the scope and limitations of our method,
we checked if the regioselectivity of the Lx(OTf) catalyzed
reaction was maintained with sugars containing an acetyl
group in a primary position. Peracetylated sugars 11 and
3
1
2 were, thus, used as next substrates in the deacetylation
reaction (Scheme 4). Once more and in each case, the selec-
tivity for the anomeric acetate was maintained whatever
the salt used and good to excellent yields were obtained
in both cases. It should be noted that in these cases too,
Nd(OTf)3 gave better isolated yields (Table 2). When
a-D-glucose pentaacetate 13 was used as substrate, lower
isolated yields between 62% and 68% were obtained. Such
3
+
Nd
catalyzed transesterification. So, methyl 1,2,3,4-
tetra-O-acetyl-b-D-glucopyranuronate
1
7
6
and methyl
Table 1
Yields obtained for the anomeric deacetylation of compound 1 depending
on the lanthanide catalyst used
a
b
Lanthanide
Yb(OTf)
Eu(OTf)
Reaction time (min)
Yields (%)
R4
R4
COOMe
O
Nd(OTf)3
COOMe
O
3
180
90
90
70
85
92
95
R3
AcO
R3
AcO
3
R1
MeOH, r.t., 4 h.
OH
Sm(OTf)
Nd(OTf)
a
3
AcO
OAc
R2
3
90
6
. R =OAc, R =H, R =OAc, R =H
8. R = OAc, R =H
1 2 3 4
3
4
Reactions were carried out in anhydr MeOH (2.4 mL) containing
7
. R =H, R =OAc, R =H, R =OAc
3 4
9. R = H, R =OAc
1
2
3
4
(
50 mg, 0.11 mmol) and 5 mol % Lx(OTf)
Isolated yields after purification.
3
at 0 °C.
b
Scheme 3. Efficient anomeric deacetylation of methyl uronates.

A. T. Tran et al. / Tetrahedron Letters 49 (2008) 2163–2165
2165
R4
4. Rowell, R. M.; Feather, M. S. Carbohydr. Res. 1967, 4, 486–491.
5. Nudelman, A.; Herzig, J.; Gottlieb, H. E.; Keinan, E.; Sterling, J.
Carbohydr. Res. 1987, 162, 145–152.
R4
R5
R5
O
Ln(OTf)3
O
R3
AcO
R3
AcO
R2
OH
AcO
MeOH, r.t.
OAc
6. Banaszk, A.; Bordas Cornet, X.; Zamojski, A. Carbohydr. Res. 1985,
R1
1
44, 344–345.
1
1
1
1
0. R1 =H, R2 =OAc, R3=OAc, R4 =H,R5=H
1
1
1
4. R3 =OAc, R4 = H, R5=H
5. R3 = H, R4 =OAc, R5=CH2OAc
6. R3= OAc, R4=H, R5=CH2OAc
7. Allen, P. Z. Methods Carbohydr. Chem. 1962, 1, 372–373.
8. Dilhas, A.; Bonnaff e´ , D. Carbohydr. Res. 2003, 338, 681–686.
1. R1 = H, R2= OAc, R3=H, R4=OAc, R5=CH2OAc
2. R1 = H, R2= OAc, R3=OAc, R4=H, R5=CH2OAc
3. R1 = OAc, R2= H, R3=OAc, R4=H, R5=CH2OAc
9
. Jacquinet, J. C.; Petitou, M.; Duchaussoy, P.; Lederman, I.; Choay,
J.; Torri, G.; Sina y¨ , P. Carbohydr. Res. 1984, 130, 221–241.
Scheme 4. A general method for anomeric deacetylation.
1
1
0. Dilhas, A.; Bonnaff e´ , D. Tetrahedron Lett. 2004, 45, 3643–3645.
1. Byramova, N.; Ovchinnikov, M. V.; Backinowsky, L. V.; Kochetkov,
N. K. Carbohydr. Res 1983, 124, C8–C11.
Table 2
12. Kobayashi, S.; Sugiura, M.; Kitagawa, H.; Lam, W. W. L. Chem. Rev
2002, 102, 2227–2302.
13. Hanamoto, T.; Sugimoto, Y.; Yokoyama, Y.; Inanaga, J. J. Org.
Chem. 1996, 61, 4491–4492.
Yields obtained for anomeric deacetylation depending on the lanthanide
catalyst used
Starting
material
Products Yb(OTf)
3
Eu(OTf)
3
Sm(OTf)
3
Nd(OTf)
3
1
4. (a) Hamza, D.; Lucas, R.; Feizi, T.; Chai, W.; Bonnaff e´ , D.;
Lubineau, A. Chem. Bio. Chem. 2006, 7, 1856–1858; (b) Lubineau,
A.; Lortat-Jacob, H.; Gavard, O.; Sarrazin, S.; Bonnaff e´ , D. Chem.
1
1
1
1
0
1
2
3
14
15
16
16
85
68
61
67
81
78
62
62
85
82
79
67
81
82
81
68
Eur. J. 2004, 10, 4265–4282.
1
1
5. Compound 5 a/b: H NMR (360 MHz, CDCl
3
): d 5.68 (dd, 1H, J1,2
4
1
0
1
1
1
.5 Hz, J1,OH 6.0 Hz, H-1b), 5.60 (d, 0.5H, J4,5 6.5 Hz, H-5a), 5.39 (d,
H, J4,5 3.0 Hz, H-5b), 5.28 (d, 1H, J1,OH 10.0 Hz, H-1a), 5.21 (d,
.5H, J2,3 1.0 Hz, H-2a), 5.08 (dd, 1H, J2,3 7.0 Hz, H-2b), 4.79 (dd,
0H, J3,4 7.0 Hz, H-4b), 4.73 (d, 0.5H, Jgem 11.0 Hz, Ph–CH), 4.67 (d,
H, Jgem 12.0 Hz, Ph–CH), 4.66 (dd, 0.5H, J3,4 6.0 Hz, H-4a), 4.59 (d,
H, Jgem 12.0 Hz, Ph–CH), 4.55 (d, 0.5H, Jgem 11.0 Hz, Ph–CH), 4.51
Reactions were carried out in anhydr MeOH (6 mL) containing starting
materials (100 mg) and 5 mol % Lx(OTf)
purification.
3
at rt. Yields were obtained after
a lower reactivity of a acetates has been previously
observed in transesterification promoted by tin oxides.
3
(t, 1H, H-3b), 4.19 (br d, 0.5H, H-3a), 3.74 (s, 3H, COOCH ), 3.66 (s,
5
1.8H, COOCH
OCOCH ), 2.13 (s, 3H, OCOCH
90 MHz, CDCl ): d 1 1.2 (C-1a), 94.2 (C-1b), 80.7 (C-3a), 80.0
C-4a), 79.2 (C-3b), 79.0 (C-2a), 77.2 (C-2b), 75.6 (C-4b), 73.3 (CH ),
3.0 (CH ), 70,6 (C-5a, C-5b), 53.0 (COOCH ), 52.5 (COOCH ), 20.8
(OCOCH ), 20.7 (OCOCH ), 20.6 (OCOCH ). C18H22O9 (MW =
3
), 3.53 (d, 0.5H, OHa), 3.10 (d, 1H, OHb), 2.18 (s, 3H,
1
3
3
3 3
), 2.11 (s, 3H, OCOCH ). C NMR
We have, thus, developed an effective method to selec-
tively deacetylate the anomeric position of carbohydrates.
This method, using lanthanide triflate catalyzed transesteri-
fication, is especially valuable when working with uronic
acids methyl esters for whom more classical methods are
(
(
3
2
7
2
3
3
3
3
3
382.1): calcd C 56.54, H 5.80; found C 56.22, H 6.04. ESI-MS calcd
+
3
+
for C18
6. Crystalline methyl-1,2,4-tri-O-acetyl-3-O-benzyl-b-L-idopyranuronate
1) (4.7 g, 11 mmol) was dissolved in MeOH (225 mL) at 0 °C, then
mol % Nd(OTf) (0.332 g, 0.56 mmol) were added to the mixture.
After stirring for 1 h at 0 °C, the reaction was stopped. Water
(300 mL) was added and the compound was extracted with CH Cl
3 Â 200 mL). The organic phase was evaporated and the residue was
purified by column chromatography using a toluene: EtOAc gradient
9
H22NaO [M+Na] m/z = 405.1; found 405.1.
inefficient. Different Lx(OTf) salts were tested and Nd
3
1
gave better yields in most of the cases and thus seems to
be the best catalyst for this reaction. Most importantly, this
approach is a green alternative to previously described pro-
cedures, thanks to the low toxicity of lanthanide salts, the
possibility of recycling the catalyst and the easy and highly
tolerant experimental procedure.
(
5
3
2
2
(
1
(
8:2–7:3) to give 1 (3.75 g, 1.7 mmol, 90%). Compound 2 a/b:
NMR (300 MHz, CDCl ): d 7.4–7.3 (m, 8H, Ph), 5.33 (br d, 1H, J1,OH
6.0 Hz, H-1a), 5.22 (m, 1H, H-2a), 5.2 (br s, 0.6H, H-1b), 5.16 (m,
H
3
Acknowledgements
0
4
4
J
.6H, H-2b), 5.02 (d, 1H, J4,5 2.0 Hz, H-5a), 4.90 (m, 0.6H, H-4b),
.83 (m, 1H, H-4a), 4.82 (d, 1H, Ph–CHa), 4.78 (d, 1H, Ph–CHa),
We thank the Univ-Paris Sud, the CNRS for funding,
the Minist e` re de l’Education Nationale et de la Recherche
for a Ph.D. Grant (A.T.T.) and ENDOTIS-Pharma for a
Ph.D Grant (S.D.) and funding.
.76 (s, 1.2H, Ph–CH
1,OH 9.0 Hz, OH-a), 4.2 (t, 0.6H, J2,3 = J3,4 3 Hz, H-3b), 3.98 (m, 1H,
), 3.79 (s, 1.8H, COOCH ), 2.2–2 (4s,
9.6H, Ac). C NMR (90 MHz, CDCl ): d 170.3–169.0 (C=O), 128.9–
2
b), 4.72 (d, 0.6H, J4,5 2 Hz, H-5b), 4.28 (d, 1H,
H-3a), 3.81 (s, 3H, COOCH
3
3
1
3
3
1
(
6
28.0 (Ph), 93.2 (C-1a), 92.1 (C-1b), 73.7 (CH
C-3b, C-5b), 72.0 (C-3a), 68.3 (C-4b), 67.2 (C-2a, C-2b), 66.8 (C-4a),
5.85 (C-5a), 52.8 (COOCH ), 21.0 (OCOCH ), 20.8 (OCOCH ).
7. Nakajima, R.; Ono, M.; Aiso, S.; Akita, H. Chem. Pharm. Bull. 2005,
3, 684–687.
8. (a) Kramer, S.; Nolting, B.; Ott, A-J.; Vogel, C. J. Carbohydr. Chem.
000, 19, 891–921; (b) Vogel, C.; Boye, H.; Kristen, H. J. Prakt.
Chem. 1990, 332, 28–36.
2 2
), 73.3 (CH ), 72.9
References and notes
3
3
3
1
1
1
. Schmidt, R. R.; Kinzy, W. Adv. Carbohydr. Chem. Biochem. 1994, 50,
5
21–123.
2
3
. Helferich, B.; Portz, W. Chem. Ber. 1953, 86, 604.
. Excoffier, G.; Gagnaire, D.; Utille, J. P. Carbohydr. Res. 1975, 39,
2
368–373.