Catalytic regioselective oxidation of glycosides

Article abstract of DOI:10.1002/anie.201301662

Discrimination among equals: A catalytic method for the selective oxidation of unprotected glycosides, both monosaccharides and disaccharides, has been developed. The resulting ketosaccharides are isolated in moderate to excellent yields. This approach provides a basis for protecting-group-free synthetic transformations of carbohydrates. Copyright

Full text of DOI:10.1002/anie.201301662

Angewandte
Chemie
DOI: 10.1002/anie.201301662
Synthetic Methods
Catalytic Regioselective Oxidation of Glycosides**
Manuel Jꢀger, Marcel Hartmann, Johannes G. de Vries,* and Adriaan J. Minnaard*
Currently, functional group transformations on carbohydrates
are highly reliant on the use of protecting groups. These
protecting groups serve different functions, 1) to protect all
hydroxy groups except one to allow selective modification,
including glycosidic bond formation, of the remaining hy-
droxy group,[1] 2) to steer the reactivity at the anomeric center
by stabilizing or destabilizing the incipient oxonium ion
(arming, disarming)[2,3] and to allow stereoselective glycosidic
bond formation through anchimeric assistance (neighboring
group participation),[4] 3) to allow solubility of carbohydrates
in nonpolar organic solvents and purification by silica gel
chromatography. As a consequence, the preparation of
a desired carbohydrate, even a straightforward derivative of
by choice of the substrate, the use of stoichiometric amounts
of organotin reagents limits the use of this approach.
The enzymatic oxidation of several carbohydrates, includ-
ing glycosides, has been described by Kçpper and co-work-
ers.[10] By using pyranose oxidase, selective oxidation at C2
and C3 was achieved, depending on the substrate. For
reducing carbohydrates, the yields were generally high, but
for glycosides low yields were observed. The activity of this
enzyme is rather low. This and the substrates being restricted
to the b-anomer, have prohibited the application of this
method. Another enzymatic approach has been described by
Haltrich et al. in which a fungal pyranose dehydrogenase was
able to oxidize a series of carbohydrates on C1, C2, C3, C1,3’,
or C2,3’.[11,12] However, yields of isolated products were not
reported.
We describe here the first catalytic, regioselective oxida-
tion of unprotected pyranosyl glycosides, both mono- and
disaccharides, to the corresponding ketosaccharides. Given
the wide scope and high selectivity of the reaction, this
approach is a significant step towards protecting-group-free
carbohydrate synthesis.
Inspiration was obtained from recent work of Waymouth
and co-workers on the palladium-catalyzed oxidation of
glycerol to dihydroxy acetone.[13] Their cationic 2,9-dimethyl-
phenanthroline (neocuproine) palladium complex selectively
oxidizes the secondary hydroxy group with excellent selec-
tivity and yield. We wondered whether this approach would
also be able to discriminate between multiple secondary
hydroxy groups. This would then possibly provide a catalytic
method for the oxidation of unprotected carbohydrates to
their corresponding keto sugars.
a
commercially available monosaccharide, frequently
requires a multistep route comprising protection and depro-
tection steps. The selective modification, in particular oxida-
tion, of hydroxy groups in unprotected carbohydrates is
therefore highly desired. At least, it would remedy the
necessity to balance the protecting-group strategies for
protection and glycosidic bond formation.
Nevertheless, the selective oxidation of carbohydrates, of
which pyranosides are the most important representatives, is
a longstanding challenge in organic chemistry. The selective
oxidation of the primary hydroxy group in pyranosides,
chemically by using the 2,2,6,6-tetramethyl-1-piperidinyloxy
radical (TEMPO)[5] or enzymatically by uridine 5’-diphos-
phoglucose dehydrogenase,[6] has been well-described. In
contrast, the selective oxidation, or even conversion in
general, of the secondary hydroxy groups is extremely
difficult and barely known.[7] Tsuda et al. have described the
selective oxidation of several methyl glycosides with stoichio-
metric bistributyltin oxide and bromine.[8,9] Although good
yields were obtained and the regioselectivity could be steered
We commenced to study this hypothesis by treating
methyl a-d-glucopyranoside (2) with catalyst 1 (2.5 mol%)
in aqueous acetonitrile by using benzoquinone as the terminal
oxidant (Scheme 1). 1H NMR and IR spectroscopy indicated
[*] M. Jꢀger, M. Hartmann, Prof. Dr. J. G. de Vries,
Prof. Dr. A. J. Minnaard
Stratingh Institute for Chemistry, University of Groningen
Nijenborgh 7, 9747 AG Groningen (The Netherlands)
E-mail: A.J.Minnaard@rug.nl
Homepage: http://www.rug.nl/research/bio-organic-chemistry/
minnaard/
Scheme 1. Selective oxidation of methyl a-d-glucopyranoside.
Prof. Dr. J. G. de Vries
DSM Innovative Synthesis
BV P.O. box 18, 6160 MD Geleen (The Netherlands)
E-mail: Hans-JG.-Vries-de@dsm.com
the formation of a single oxidation product within 3 h. After
isolation (see below) and thorough 2D-NMR studies, this
product turned out to be 3.
The use of DMSO as the solvent considerably accelerated
the reaction and this encouraged the further optimization of
the catalyst system. The use of dichlorobenzoquinone
(DCBQ) instead of benzoquinone (BQ) led to a faster
reaction, but required a minimum catalyst loading of
[**] This research has been performed within the framework of the
CatchBio program. The authors gratefully acknowledge the support
of the Smart Mix Program of the Netherlands Ministry of Economic
Affairs, Agriculture and Innovation and the Netherlands Ministry of
Education, Culture and Science.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201301662.
Angew. Chem. Int. Ed. 2013, 52, 7809 –7812
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7809

.
Angewandte
Communications
Table 1: The selective catalytic oxidation of monosaccharides.
1.1 mol% to reach full conversion. With benzoquinone,
0.5 mol% already sufficed. The use of oxygen as the oxidant
resulted in a slower reaction and even 2.5 mol% of catalyst
was not enough to reach full conversion after 48 h.
Entry Substrate
Product
Method[a] Yield [%][b]
1
A
A
96
85
Initially, isolation of the product was problematic because
of its polarity and supposed acid and base sensitivity. When
the reaction was carried out in aqueous acetonitrile, it turned
out that 3 could be isolated pure and in 96% yield by
successive washings with ether and toluene, followed by
evaporation of the residual water. In case the reactions were
carried out in DMSO, percolation of the entire mixture over
a charcoal column with water as the eluent removed the
majority of this solvent, and subsequent chromatography on
silica gel by using a mixture of CH2Cl2/acetone/methanol/
water provided the pure product.[14]
With the catalyst system and product isolation estab-
lished, the scope of the reaction was studied. By using the
same conditions, both methyl b-d-glucopyranoside (Table 1,
entry 2) and methyl N-acetylamino-a-d-glycoside (entry 3)
were selectively oxidized in 85% and 89% yield after
isolation, respectively. Glucose and sorbitol were unselec-
tively oxidized.
Methyl-2-a-d-desoxyglucopyranoside (Table 1, entry 4)
was selectively oxidized using dioxane/DMSO (4:1) as
a solvent and the product could be isolated in 60% yield.
The anomeric phenyl-protected glycoside 10 (Table 1,
entry 5) was also cleanly oxidized to 11, which was isolated
in 73% yield.
Oxidation of thiophenyl glucopyranoside 12 gave 13 in
a clean reaction, but the product was isolated in only 47%
yield. Here 6.5 mol% of catalyst had to be used to drive the
reaction to completion.
Subsequently, the influence of the primary hydroxy group
on the reaction was studied, since we suspected its involve-
ment in the stereoselectivity. The tert-butyldimethylsilyl
substituent in 14 was completely removed during the reac-
tion[15] and the expected product was isolated in good yield.
The less-sensitive tert-butyldiphenylsilyl substituent in 15
withstood the reaction conditions, and in DMSO (which was
used because 15 did not dissolve in acetonitrile) oxidation was
complete in 15 min affording 16 in 66% yield. Furthermore,
benzoyl-substituted 17 was cleanly oxidized as well, though
product 18 was isolated in a decreased 45% yield. Apparently
neither the presence nor the nature, electron-donating or
-withdrawing, of the substituent has an influence on the
selectivity of the reaction.
Although the formation of small amounts of regioiso-
meric products cannot be excluded, the variation in yield
(Table 1) is not due to a lack of selectivity of the reaction,
because the NMR spectra of the crude mixtures show
oxidation solely at C3 or at least with a selectivity of > 90%
(see the Supporting Information). However, purification of
these very polar compounds by flash chromatography dimin-
ished the yields.
2
3
4
A
B
89
60
5[c]
B
C
A
D
D
73
47
96
66
45
6
7[d]
8
9
[a] Method A: pyranoside (4 mmol), 1 (2.5 mol%), DCBQ (3 equiv),
MeCN/H2O 10:1, 0.3m. Method B: pyranoside (0.8 mmol),
1 (2.5 mol%), DCBQ (3 equiv), dioxane/DMSO 4:1, 0.3m. Method C:
pyranoside (0.8 mmol), 1 (6.5 mol%) added in portions, DCBQ
(3 equiv), dioxane/DMSO 4:1, 0.3m Method D: pyranoside (0.8 mmol),
1 (2.5 mol%), DCBQ (3 equiv), DMSO, 0.3m. [b] Yields of isolated
products. [c] 10 (0.4 mmol). [d] The TBS group was cleaved under the
reaction conditions. TBS=tert-butyldimethylsilyl, TBDPS=tert-butyldi-
phenylsilyl.
saccharides opens a multitude of opportunities to modify
and access naturally occurring saccharides without the con-
struction of glycosidic bonds. Also in these cases the reaction
turned out to be highly regioselective, providing 20 and 22 in
good yields after isolation. The identity of the products was
established by COSY-, HSQC-, and NOESY-NMR spectros-
copy.
The observed preference of the catalyst to selectively
oxidize the C3 OH group, also in connection with its substrate
scope, is intriguing and currently cannot be fully explained.
The stereochemistry at the anomeric center C1 (Table 1,
entries 1, 2, and 6) and the nature of the substituent (entries 5
and 6) is apparently not relevant for the reaction outcome.
Also the C2 OH group is not important, since it can be
To take our strategy a step further, we studied the
selective oxidation of methyl maltoside 19 and methyl
cellobioside 21, as representatives of disaccharides containing
an a- and b-glycosidic linkage (Scheme 2). Being able to
selectively oxidize one hydroxy group in more-complex
7810
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 7809 –7812

Angewandte
Chemie
nucleophilicity of single hydroxy groups in
glycosides remains difficult.
As a first demonstration of the synthetic
versatility of the current strategy, protecting-
group-free[19] syntheses of methyl a-d-allose
and methyl 3-amino-3-deoxy-a-d-allose (3-
epi-kanosamine) were established. Allose is
a rare monosaccharide, and currently pre-
pared in four to five steps.[20,21] Reduction of 3
with NaBH4 in methanol[22] leads directly to
methyl allose in 95% yield (Scheme 4).
Scheme 2. Selective oxidation of methyl b-d-maltoside and methyl b-d-cellobioside.
omitted (Table 1, entry 4) or replaced by an acetamido
substituent (entry 3). Methyl a-mannopyranoside and
methyl a-galactopyranoside, with axial C2 OH and axial C4
OH functional groups, respectively, are however not selec-
tively oxidized. The primary C6 OH group stays consequently
untouched, in line with the selective oxidation of glycerol,
thereby making this method complementary to the afore-
mentioned TEMPO-catalyzed oxidation.[5] Substitution does
not influence the selectivity of the oxidation.
A possible rationale for the regioselectivity could be
a kinetically controlled coordination of the catalyst to the C3
OH group, followed by deprotonation and subsequent
hydride abstraction. Whether simultaneous coordination to
the C4 OH group assists in this process is not clear; it would
explain the selective oxidation of the disaccharides 19 and 21
on the left residue, but steric hindrance could cause the same
effect. The hypothesis derives some support from a study by
Bols and co-workers.[16] Comparing pKa values in a series of
methyl glucosamines 23–26 (Scheme 3), the C3 NH2 turned
Scheme 4. Synthesis of methyl a-d-allose and methyl 3-epi-kanosa-
mine
Alternatively, 3 is converted into its corresponding O-
methyl oxime 28 and subsequently reduced with H2/Adamꢀs
catalyst[23] to afford methyl 3-epi-kanosamine 29 in 58%
overall yield as a single isomer, after peracylation to facilitate
isolation.
In conclusion, the possibility to perform protecting-group-
free synthetic transformations on carbohydrates has been
brought a step closer by developing a Pd-catalyzed regiose-
lective oxidation of pyranosyl glycosides. The applied Pd/
neocuproine catalyst distinguishes between the various sec-
ondary hydroxy groups and selective oxidizes the one at C3. A
catalyst loading of 0.5 mol% (1 mol% Pd) is sufficient for full
conversion on gram scale within hours at room temperature.
The products are isolated in high yields, and the substrate
scope is considerable, including both mono- and disacchar-
ides. The selective oxidation of more-complex tri- and
oligosaccharides is currently studied, and the approach
could assist in the preparation of building blocks for
automated carbohydrate synthesis.[24] Although the origin of
the regioselectivity is under study as well, kinetic product
formation seems to be the most likely explanation. Applica-
tion of the methodology is illustrated by the efficient synthesis
of methyl allose and methyl 3-epi-kanosamine in high yield
from methyl glucose.
Scheme 3. pKa values in a series of methyl glucosamines 23–26.
out to have the highest pKa value, which is an indication for
a higher basicity/nucleophilicity of the corresponding C3 OH
group.
A different study by Thiem and Matwiejuk, investigating
the influence of partial protection on the pKa value of the
residual free hydroxy groups of methyl glycosides, shows that
the isolated C4 OH group should be more basic/nucleophilic
than the C3 OH group.[17] Nevertheless it also shows that an
increasing number of free vicinal hydroxy groups increases
the acidity of the glycoside, probably because of intramolec-
ular hydrogen bonding. If the C3 and C6 hydroxy groups are
not protected, however, the basicity/nucleophilicity of the C3
OH group is even higher compared to the glycoside with only
one free hydroxy group. Furthermore it has been shown by Li
and Kalikanda that partial protection can lead to inversion of
the reactivity of glycosides.[18] Therefore an estimation of the
Experimental Section
Methyl-a-d-ribo-hexapyranoside-3-ulose (3): Methyl-a-glucopyrano-
side (777 mg, 4.0 mmol, 1.0 equiv) and 2,6-dichloro-1,4-benzoquinone
Angew. Chem. Int. Ed. 2013, 52, 7809 –7812
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7811

.
Angewandte
Communications
(2.12 g, 12.0 mmol, 3.0 equiv) were suspended in acetonitrile/water
(13 mL, 10:1, 0.3m in substrate). The catalyst [(2,9-dimethyl-1,10-
phenanthroline)-Pd(m-OAc)]2(OTf)2 (105 mg, 2.5 mol%) was added,
and the mixture was stirred at room temperature for 3 h. Toluene
(50 mL) was added, and the mixture was extracted twice with water
(7 mL). The combined water layers were washed once with ethyl
ether (35 mL), filtered, and concentrated in vacuo to give the pure
methyl-a-d-ribo-hexapyranosid-3-ulose (751 mg, 3.9 mmol) in 96%
yield as a dark brown solid. 1H NMR (400 MHz, 298 K, [D6]DMSO):
d = 4.95 (d, J = 4.2 Hz, 1H), 4.29 (dd, J = 4.2, 1.5 Hz, 1H), 4.07 (dd,
J = 9.8, 1.4 Hz, 1H), 3.69 (dd, J = 11.9, 1.9 Hz, 1H), 3.59 (dd, J = 11.9,
4.9 Hz, 1H), 3.46 (ddd, J = 9.7, 4.9, 1.8 Hz, 1H), 3.26 (s, 3H).
13C NMR (50 MHz, [D6]DMSO): d = 206.1, 102.2, 75.4, 74.6, 71.9,
60.7, 54.4. HRMS (ESI): m/z: calcd for C7H12O6Na ([M + Na+]):
[7] D. Lee, M. Taylor, Synthesis 2012, 3421 – 3431.
[8] Y. Tsuda, M. Hanajima, N. Matsuhira, Y. Okuno, K. Kanemitsu,
Chem. Pharm. Bull. 1989, 37, 2344 – 2350.
[9] H.-M. Lui, Y. Sato, Y. Tsuda, Chem. Pharm. Bull. 1993, 41, 491 –
501.
[10] S. Freimund, A. Huwig, F. Giffhorn, S. Kçpper, Chem. Eur. J.
1998, 4, 2442 – 2455.
ˇ
[11] J. Volc, P. Sedmera, P. Halada, G. Daniel, V. Prikrylovꢂ, D.
Haltrich, J. Mol. Catal. B 2002, 17, 91 – 100.
ˇ
[12] P. Sedmera, P. Halada, E. Kubꢂtovꢂ, D. Haltrich, V. Prikrylovꢂ, J.
Volc, J. Mol. Catal. B 2006, 41, 32 – 42.
[13] R. M. Painter, D. M. Pearson, R. M. Waymouth, Angew. Chem.
2010, 122, 9646 – 9649; Angew. Chem. Int. Ed. 2010, 49, 9456 –
9459.
=
215.053, found: 215.052; IR nmax = 3436 (OH), 2947 (C–H), 1736 (C
[14] C. Besset, S. Chambert, B. Fenet, Y. Queneau, Tetrahedron Lett.
2009, 50, 7043 – 7047.
O), 1031 (C–O) cmÀ1
.
[15] B. H. Lipshutz, D. Pollart, J. Monforte, H. Kotsuki, Tetrahedron
Lett. 1985, 26, 705 – 708.
[16] C. M. Pedersen, J. Olsen, A. B. Brka, M. Bols, Chem. Eur. J.
2011, 17, 7080 – 7086.
Received: February 26, 2013
Revised: May 2, 2013
Published online: June 18, 2013
[17] M. Matwiejuk, J. Thiem, Eur. J. Org. Chem. 2012, 2180 – 2187.
[18] J. Kalikanda, Z. Li, Tetrahedron Lett. 2010, 51, 1550 – 1553.
[19] I. S. Young, P. S. Baran, Nat. Chem. 2009, 1, 193 – 205.
[20] F. R. D. Colombo, A. Scala, I. M. Taino, P. A. Taino, J. Carbo-
hydr. Chem. 1994, 13, 611 – 617.
[21] D. C. Baker, D. Horton, C. G. Tindall, Carbohydr. Res. 1972, 24,
192 – 197.
[22] W. Li, K. Koike, Y. Asada, T. Yoshikawa, T. Nikaido, Carbohydr.
Res. 2003, 338, 729 – 731.
[23] Y. Tsuda, Y. Okuno, M. Iwaki, K. Kanemitsu, Chem. Pharm.
Bull. 1989, 37, 2673 – 2678.
[24] M. Weishaupt, S. Eller, P. H. Seeberger, Methods Enzymol. 2010,
478, 463 – 484.
Keywords: carbohydrates · glycosides · oxidation · palladium ·
regioselectivity
.
[1] J. D. C. Codꢁe, A. Ali, H. S. Overkleeft, G. A. van der Marel, C.
R. Chim. 2011, 14, 178 – 193.
[2] C. M. Pedersen, L. U. Nordstrøm, M. Bols, J. Am. Chem. Soc.
2007, 129, 9222 – 9235.
[3] D. R. Mootoo, P. Konradsson, U. Udodong, B. Fraser-Reid, J.
Am. Chem. Soc. 1988, 110, 5583 – 5584.
[4] J. Guo, X.-S. Ye, Molecules 2010, 15, 7235 – 7265.
[5] N. J. Davis, S. L. Flitsch, Tetrahedron Lett. 1993, 34, 1181 – 1184.
[6] E. J. Toone, E. S. Simon, G. M. Whitesides, J. Org. Chem. 1991,
56, 5603 – 5606.
7812
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 7809 –7812