A new, efficient glycosylation method for oligosaccharide synthesis under neutral conditions: Preparation and use of new DISAL donors

Article abstract of DOI:10.1021/jo0057654

Efficient, stereoselective glycosylation methods are required for the synthesis of complex oligosaccharides as tools in glycobiology. All glycosylation methods, which have found wide acceptance, rely on Lewis acid activation of glycosyl donors prior to glycosylation. Here, we present a new and efficient method for glycosylation under neutral or mildly basic conditions. Glycosides of methyl 2-hydroxy-3,5-dinitrobenzoate (DISAL) and its para regioisomer, methyl 4-hydroxy-3,5-dinitrobenzoate, were prepared by nucleophilic aromatic substitution. In a first demonstration of their potential as glycosyl donors, stereospecific glycosylation of methanol was achieved. In the glycosylation of more hindered alcohols, the β-donor proved more reactive, and α-glucosides were predominantly formed. Glycosylation of protected monosaccharides, with free 6-0H or 3-OH, proceeded smoothly in 1-methyl-2-pyrrolidinone (NMP) at 40-60 °C in the absence of Lewis acids and bases in good to excellent yields. Glycosylation of 3-OH gave the α-linked disaccharide only.

Full text of DOI:10.1021/jo0057654

6268
J . Org. Chem. 2001, 66, 6268-6275
A New , Efficien t Glycosyla tion Meth od for Oligosa cch a r id e
Syn th esis u n d er Neu tr a l Con d ition s: P r ep a r a tion a n d Use of New
DISAL Don or s
Lars Petersen and Knud J . J ensen*
Department of Chemistry, Building 201, Kemitorvet, Technical University of Denmark,
DK-2800 Kgs. Lyngby, Denmark
kj@kemi.dtu.dk
Received December 11, 2000
Efficient, stereoselective glycosylation methods are required for the synthesis of complex oligosac-
charides as tools in glycobiology. All glycosylation methods, which have found wide acceptance,
rely on Lewis acid activation of glycosyl donors prior to glycosylation. Here, we present a new and
efficient method for glycosylation under neutral or mildly basic conditions. Glycosides of methyl
2-hydroxy-3,5-dinitrobenzoate (DISAL) and its para regioisomer, methyl 4-hydroxy-3,5-dinitroben-
zoate, were prepared by nucleophilic aromatic substitution. In a first demonstration of their potential
as glycosyl donors, stereospecific glycosylation of methanol was achieved. In the glycosylation of
more hindered alcohols, the â-donor proved more reactive, and R-glucosides were predominantly
formed. Glycosylation of protected monosaccharides, with free 6-OH or 3-OH, proceeded smoothly
in 1-methyl-2-pyrrolidinone (NMP) at 40-60 °C in the absence of Lewis acids and bases in good to
excellent yields. Glycosylation of 3-OH gave the R-linked disaccharide only.
Sch em e 1
In tr od u ction
In recent years, a composite picture of the biological
roles of oligosaccharides has emerged.1 Cell surface
oligosaccharides mediate cell-cell recognition through
multiple simultaneous interactions with receptors on the
cognate cell, but they also mediate the binding of bacte-
rial and viral pathogens to target cells prior to infection.
Cancer cells display an array of tumor-associated anti-
gens, many of which are carbohydrate based. Glycans of
glycopeptides and -proteins isolated from natural sources
or after gene expression2 are often heterogeneous. Easy
access to pure and diverse oligosaccharides would hand
the glycobiologists an important tool in the study of the
biological roles of glycans, especially in glycoproteins.
Furthermore, reagents that inhibit, modify, or mimic
such specific carbohydrate-protein interactions are of
great pharmaceutical importance and may lead to new
pharmaceuticals for immunology, inflammation, and
oncology.3
Numerous studies on the Lewis acid (electrophile)-
promoted glycosylation of alcohols have been reported,
and an ever increasing number of glycosylation proce-
dures have been published.4 In practically all cases,
activation by a Lewis acid promoter gives rise to an
oxocarbenium ion I, which in the case of a 2-O-acyl donor
gives a dioxocarbenium ion II (Scheme 1), both as ion
pairs. Nucleophilic attack by an alcohol on I can occur
on either the R- or â-face to give the R- and â-glycosides
III and IV, respectively. Attack on dioxocarbenium ion
II by an alcohol gives â-glycoside IV or ortho ester V,
which under acidic conditions can rearrange to the
â-glycoside. Though many features of these reactions thus
are well understood, the complex nature of the reactants
often makes it difficult to predict the stereochemical
outcome, and consequently, it may be difficult to predict
which of the current glycosylation procedures is prefer-
able for establishing a particular glycosidic linkage.
Glycosylation of phenols by glycosyl bromides,5 chlo-
rides,6 and fluorides7 under basic conditions, in which the
phenoxide is formed, can proceed with a high degree of
inversion at the anomeric center. However, this approach
has generally not been successful for aliphatic alcohols,
* To whom correspondence should be addressed. Phone: (+45) 45
25 21 41. Fax: (+45) 45 93 39 68.
(1) (a) Varki, A. Glycobiology 1993, 3, 97-130. (b) Molecular
Glycobiology; Fukuda, M., Hindsgaul, O., Eds.; IRL Press: Oxford,
1994. (c) Yarema, K. J ., Bertozzi, C. R. Curr. Opin. Chem. Biol. 1998,
49-61.
(2) J enkins, N.; Parekh, R. B.; J ames, D. C. Nature Biotechnol. 1996,
14, 975-981.
(3) Sears, P.; Wong, C.-H. Angew. Chem., Int. Ed. Engl. 1999, 38,
2300-2324.
(4) (a) Paulsen, H. In Modern Methods in Carbohydrate Synthesis;
Khan, S. H., O’Neill, R., Eds; Harwood Academic Publisher: Foster
City, 1996; Vol. 1, pp 1-19. (b) Schmidt, R. R. Angew. Chem. 1986,
98, 213-236. (c) Sinay, P. Pure Appl. Chem. 1991, 63, 519-528. (d)
Paulsen, H. Angew. Chem., Int. Ed. Engl. 1982, 21, 155-173. (e) Davis,
B. G. J . Chem. Soc., Perkin Trans. 1 2000, 2137-2160.
10.1021/jo0057654 CCC: $20.00 © 2001 American Chemical Society
Published on Web 08/18/2001

Glycosylation Method for Oligosaccharide Synthesis
J . Org. Chem., Vol. 66, No. 19, 2001 6269
Sch em e 2
Sch em e 3a
especially if they are of the complexity of carbohydrate
derivatives. On the other hand, glycosyl iodides have
proven their value for the glycosylation of aliphatic
alcohols in saccharides under mild, nonacidic conditions
by an in situ anomerization mechanism.8,4d However, the
main disadvantage of glycosyl iodides is their lack of
stability and often the need to prepare them in situ.8,9
We are interested in developing new and efficient
methods for glycosylation under nonacidic conditions10
as this promises to simplify the reaction conditions and
avoid at least some of the problems associated with Lewis
acid promoted glycosylations. We reasoned that glyco-
sides of phenols carrying sufficiently electron-withdraw-
ing substituents11 could possibly serve as glycosyl donors
under neutral or mildly basic conditions (Scheme 2).
Glycosides of 2,4-dinitrophenol, while labile to nucleo-
philes such as ammonia in methanol,12 appeared not to
be reactive enough in glycosylation reactions. To obtain
more reactive glycosyl donors, we prepared glycosides of
methyl 2-hydroxy-3,5-dinitrobenzoate13 (DISAL) and meth-
yl 4-hydroxy-3,5-dinitrobenzoate and applied them suc-
cessfully to glycosylations. The name DISAL was derived
from the former compound, a dinitrosalicylic acid deriva-
tive. For safety concerns, trinitrophenyl glycosides were
not tested.
a
Reagents and conditions: (a) fuming HNO3 in fuming H2SO4
(65% oleum) (1:2), 1 h at 90 °C (84%); (b) (COCl)2 (1.2 equiv), cat.
DMF in CH2Cl2, 1 h at 25 °C; (c) MeOH (97%, two steps); (d)
fuming HNO3 in concentrated H2SO4 (1:2), 17 h at 90 °C (75%);
(e) (COCl)2 (1.2 equiv), cat. DMF in THF, 15 min at 25 °C; (f)
MeOH in hexane (60%).
Resu lts a n d Discu ssion
Syn th esis of P r ecu r sor s a n d Don or s. Glycosides of
2,4-dinitrophenol are most conveniently prepared by
nucleophilic aromatic substitution of 1-fluoro-2,4-dini-
trobenzene with partially protected carbohydrate deriva-
tives with a free 1-OH in DMF in the presence of DBU
or DABCO.14 Thus, Koeners et al. prepared 2,4-dinitro-
phenyl 2,3,4,6-tetra-O-benzyl-â-D-glucopyranoside (1) from
2,3,4,6-tetra-O-benzyl-D-glucopyranose (2). To access gly-
cosides of methyl 2-hydroxy-3,5-dinitrobenzoate and meth-
yl 4-hydroxy-3,5-dinitrobenzoate, we required the corre-
sponding fluoro precursors. Nitration of inexpensive
2-fluorobenzoic acid in a preformed mixture of 65%
fuming sulfuric acid and fuming nitric acid (2:1)15 gave
the expected 2-fluoro-3,5-dinitrobenzoic acid (3) in a good
yield of 84% (Scheme 3), which was an improvement over
a previous report.16 Reaction with oxalyl chloride in
dichloromethane containing a catalytic amount of DMF
converted the benzoic acid derivative 3 smoothly to the
corresponding acid chloride, which was reacted im-
mediately with excess methanol to give a high yield of
methyl ester 4 after crystallization (Scheme 3). 4-Fluo-
robenzoic acid was dinitrated under somewhat milder
conditions giving a slightly higher yield (75%) of 4-fluoro-
3,5-dinitrobenzoic acid (5) than previously reported.17
Benzoic acid derivative 5 was converted as before to the
corresponding methyl ester 6 in 60% yield after crystal-
lization. Higher lability of the aryl fluoride was observed
in 6, but it was successfully synthesized under mild
reaction conditions. Even though 4 and 6 both are very
good electrophiles, they store well as crystalline com-
pounds.
(5) (a) Glaser, E.; Wulwek, W. Biochem. Zeitschrift 1924, 145, 514-
534. (b) Abbas, S. A.; Matta, K. L Carbohydr. Res. 1983, 124, 115-
121.
(6) (a) Ballardie, F.; Capon, B.; Sutherland, J . D. G.; Cocker, D.;
Sinnott, M. J . Chem. Soc., Perkin Trans. 1 1973, 2418-1429. (b)
Courtin-Duchateau, M.-C.; Veyrie`res, A. Carbohydr. Res. 1978, 65, 23-
33. (c) Delmotte, F. M.; Privat, J .-P. D. J .; Monsigny, M. L. P.
Carbohydr. Res. 1975, 40, 353-364.
(7) (a) Voznyi, Y. V.; Kalicheva, I. S.; Galoyan, A. A. Bioorg. Khim.
1982, 8, 1388-1392. (b) Voznyi, Y. V.; Kalicheva, I. S.; Galoyan, A. A.
Bioorg. Khim. 1985, 11, 970-972.
(8) Hadd, M. J .; Gervay, J . Carbohydr. Res. 1999, 320, 61-69.
(9) Caputo, R.; Kunz, H.; Mastroianni, D.; Palumbo, G.; Pedatella,
S.; Solla, F. Eur. J . Org. Chem. 1999, 11, 3147-3150.
(10) Recent reports have described the use of lithium ions for the
activation of glycosyl trichloroacetimidates, fluorides, and phosphites,
as well as epoxides. However, the lithium salts still act as, albeit mild
Lewis acids, hence the reaction conditions are not neutral in the strict
sense used in this paper: (a) Lubineau, A.; Drouillat, B. J . Carbohydr.
Chem. 1997, 16, 1179-1186. (b) Schene, H.; Waldmann, H. Chem.
Commun. 1998, 2759-2760. (c) Schmid, U.; Waldmann, H. Chem. Eur.
J . 1998, 4, 494-501. (d) Schene, H.; Waldmann, H. Synthesis 1999,
1411-1422.
(11) The synthesis of glycosyl donors by nucleophilic aromatic
substitution is conceptually related to Mukaiyama’s and Schmidt’s
preparation of hetaryl glycosides from electron-poor heterocycles.
However, in contrast to the glycosylation method presented here,
Mukaiyama’s donors required activated, trimethylsilylated nucleoside
bases and Schmidt’s hetaryl glycosides required the presence of a Lewis
acid (TMSOTf or BF3‚OEt2) for efficient glycosylation of alcohols: (a)
Mukaiyama, T.; Hashimoto, Y.; Hayashi, Y.; Shoda, S. Chem. Lett.
1984, 557-560. (b) Huchel, U.; Schmidt, C.; Schmidt, R. R. Tetrahedron
Lett. 1995, 36, 9457-9460. (c) Huchel, U.; Schmidt, C.; Schmidt, R. R.
Eur. J . Org. Chem. 1998, 1353-1360.
(12) (a) Glaser, E.; Thaler, A. C. Arch. Pharm. 1926, 228-237. (b)
Hengstenberg, W.; Wallenfels, K. Carbohydr. Res. 1969, 11, 85-91.
(13) Aryl glycosides of salicylic acid are significantly more labile to
acid hydrolysis than the corresponding phenyl glycosides. This has been
ascribed to acid catalysis by the free benzoic acid moiety; see: (a)
Capon, B. Tetrahedron Lett 1963, 14, 9111-913. (b) Capon, B.; Smith,
M. C.; Anderson, E.; Dahm, R. H.; Sankey, G. H. J . Chem. Soc. B 1969,
1038-1047.
Next, with these fluorobenzene derivatives synthe-
sized, we turned to the formation of aryl glycosides as
potential glycosyl donors. It was observed, as expected,
that nucleophilic aromatic substitution was not feasible
on 3, which has a free carboxylic acid. However, forma-
tion of the aryl glycoside from the methyl esters 4 and 6
proceeded in high yields with short reaction times
(Scheme 4). Initial experiments showed that high con-
centration of base, ions (even chloride ions), or weak
nucleophiles were not well tolerated in the reaction
mixture. Therefore, a double-base system was developed
(14) Koeners, H. J .; de Kok, A. J .; Romers, C.; van Boom, J . H.
Recueil 1980, 99, 355-362.
(15) The same yield was obtained in a mixture of concentrated
sulfuric acid and fuming nitric acid, but longer reaction times were
required.

6270 J . Org. Chem., Vol. 66, No. 19, 2001
Petersen and J ensen
Sch em e 4
Sch em e 5
Ta ble 2. Resu lts fr om Solvolysis Exp er im en ts
entry
DISAL (R/â)
acceptor
product
yielda (%) (R/â)
1
2
3
4
5
6
7r,â (8.4:1)
7r,â (1:14.1)
10r
MeOHb
MeOHb
MeOHc
MeOHc
i-PrOHd
i-PrOHd
13r,â
13r,â
13â
98 (1:6.1)
98 (12.3:1)
98
10â
13r
98
7r,â (8.4:1)
7r,â (1:14.1)
14r,â
14r,â
97 (1.7:1)
98 (12.7:1)
Ta ble 1. F or m a tion of Ar yl Glycosid es by
RP-HPLC (215 nm). 30 °C for 6 h. c 25 °C for 2 h. 30 °C for
Ba se-Ca ta lyzed Nu cleop h ilic Ar om a tic Su bstitu tion
a
b
d
16 h + 40 °C for 24 h.
hemiacetal
DISAL
time (h) glycoside
yield (%)
entry
base
(R/âa)
(R/âb)
yield compared to ortho glycoside 7 was due to a higher
lability of the para aryl glycoside.
Finally, the ortho donor derived from 2,3,4,6-tetra-O-
benzyl-D-mannopyranose21 (11) was synthesized from 4
via the DMAP protocol and isolated in 75% yield as the
pure R-anomer (12R).22,23
1
2
3
4
5
6
7
8
DMAP
DMAP
DMAP
DMAP
DMAP
DMP
2 (3:1)
2r
0.5
0.5
0.5
0.5
0.5
2
7r,â
7r,â
9r,â
10r,â
12r
7r,â
7r,â
9r,â
89 (2.1:1)
89 (8.4:1)
92 (6.7:1)
65 (4.1:1)
75
89 (1:9)
78 (1:14.1)
87 (1:8)
8 (7:1)
2 (6:1)
11 (5:1)
2 (3:1)
2 (4:1)
8 (4:1)
DMP
DMP
2 + 1c
2
The selective synthesis of 7R or 7â was also carried
out. Tetrabenzylated glucose 2 was recrystallized twice
form ethyl acetate to yield the pure R-anomer,24 which
was reacted with 4 under the DMAP protocol. This
yielded donor 7R (R/â 8.4:1) in the same yield as before
(89%, entry 2); however, the presence of some â-glycoside
indicated that even in fast reactions some anomerization
of 1-OH occurred prior to formation of the aryl glycoside.
To form 7â selectively, 4 was added slowly to a stirred
solution of 2 (R/â 4:1) and DMP, in what amounts to a
dynamic kinetic resolution. After isolation, 78% of donor
7â was obtained with only a minor amount of the
R-anomer (R/â 1:14.1, entry 7). These results are sum-
marized in Table 1. It should be emphasized that DISAL
donors do not anomerize under the conditions used for
their synthesis, i.e., in nonpolar solvents.
a
Approximately anomeric distribution in starting material (13
C
NMR). Determined by RP-HPLC. c Aryl fluoride added over 2 h
and then stirred for an additional 1 h.
b
where only a catalytic amount of the soluble base was
used as a “shuttle” and a solid base (Li2CO3) acted as a
drain for liberated HF from the reaction mixture. The
solid base could then be used in excess without affecting
the reaction. An extensive screening of potential soluble
bases (see the Supporting Information) pointed to 4-(N,N-
dimethylamino)pyridine (DMAP, 0.3 equiv)18 and 1,4-
dimethylpiperazine (DMP, 0.5 equiv) as very effective.
The use of DMAP as base gave an R/â ratio similar to
the starting 1-OH derivative (Table 1). Interestingly,
formation of aryl â-glycoside 7â was favored using DMP
as soluble base, and somewhat longer reaction times were
required.19 These two protocols were also applied to
benzoyl-protected glucopyranose derivative 8 to give the
benzoyl-protected glycoside 9R,â.20 Also, the benzyl-
protected para donor 10 was synthesized from 2 and 6
in a DMAP-catalyzed reaction in 65% yield. The lower
Solvolysis Exp er im en ts. In a first test of their
potential as glycosyl donors, aryl glycoside 7R,â was
treated with simple alcohols under solvolytic conditions
(Scheme 5, Table 2). Thus, treatment of 7R (R/â 8.4:1)
with methanol gave the methyl â-D-glucoside 13â (R/â
1:6.1). Similarly, 7â (R/â 1:14.1) with methanol gave the
(16) Goldstein, H.; Giddey, A. Helv. Chim. Acta 1954, 37, 1121-
1125.
(17) Nielsen, A. T.; Chafin, A. P.; Christian, S. L. J . Org. Chem.
1984, 49, 4575-4580.
(20) Whereas the anomers of benzyl-protected 7R,â proved difficult
to separate, the analogous benzoyl protected anomers 9R,â were easily
separated by chromatography on silica gel.
(18) This also removed the slight excess of aryl fluoride (1.2 equiv),
presumably through formation of an ionic compound by substitution
of the fluoride by the heterocyclic nitrogen of DMAP. This highly
colored compound was not a reaction intermediate as addition of 2 to
a preformed mixture of 4 and DMAP yielded no aryl glycoside.
(19) With respect to the carbohydrate, this reaction is an anomeric
O-arylation in which the equatorial (â) anomeric oxide oxygen is likely
to be more reactive than the axial (R) (see ref 4b). The DMP-promoted
arylation of 1-OH derivative 2 thus seems to form the kinetic product
by anomeric O-arylation of the more reactive equatorial (â) oxide
oxygen.
(21) Koto, S.; Morishima, N.; Miyata, Y.; Zen, S. Bull. Chem. Soc.
J pn. 1976, 49, 2639-2640
1
(22) Heteronuclear nondecoupled NMR showed J C1,
indicating R-configuration.
) 178.1 Hz
H1
(23) In the synthesis of a related aryl glycoside, the DMP promoted
reaction between 11 and an aryl fluoride gave predominantly the
â-mannoside, similar to the reaction of 2 to give â-glucoside 7â:
Laursen, J . B.; Petersen, L.; J ensen, K. J . Org. Lett. 2001, 2, 687-
690.
(24) Oscarson, S.; Sehgelmeble, F. W. J . Am. Chem. Soc. 2000, 122,
8869-8872.

Glycosylation Method for Oligosaccharide Synthesis
J . Org. Chem., Vol. 66, No. 19, 2001 6271
Ta ble 3. Mod el Glycosyla tion of Cycloh exa n ol, Solven t Eva lu a tion a
DISAL donor
cyclohexanol
(equiv)
additives
(equiv)
yieldb of 16
(%) (R/â)
entry
(R/â)
solvent
DMA
NMP
NMP
1
2
3
7 (2.1:1)
7 (2.1:1)
7 (2.1:1)
8
8
5
Et3N (6)
Et3N (6)
2,6-lutidine (5),
4 Å sieves
41 (2.2:1)
71 (2.4:1)
71 (2.4:1)
4
5
6
7
8
7 (2.1:1)
7 (1:9)
5
5
5
5
5
5
5
NMP/CH3CN
NMP/CH3CN
CH3NO2c
NMP/ CH3NO2
THFc,d
80 (2.1:1)
80 (2.1:1)
81 (3.4:1)
84 (1.9:1)
78e (8.1:1)
73 (2.0:1)
86 (2.1:1)
7 (1.6:1)
7 (1.6:1)
7 (2.1:1)
7 (1.6:1)
10r
Et3N (6)
9
10
NMP/THF
NMP
a
b
d
Temperature 40 °C, time 17 h. RP-HPLC (215 nm). c 60 °C. 40 h. e Isolated yield.
methyl R-D-glucoside 13R (R/â 12.3:1).25 That the gly-
cosylation proceeded with inversion was further demon-
strated by treating pure 10R and 10â with methanol,
which gave 13â and 13R, respectively. From 10R 93% of
pure 13â was isolated. It is thus a stereospecific gly-
cosylation. Furthermore, sterically hindered alcohols also
reacted with 7R,â. The isopropyl glucoside 14R,â was thus
formed in high yields (Table 2, entries 5 and 6) and with
a high degree of inversion for 7â whereas 7R was less
reactive and gave an R/â ratio of 1.7:1. Similarly, reaction
of 7â with sterically hindered tert-butyl alcohol proceeded
smoothly to give 89% of 15R,â, whereas 7R did not react.
The lack of inversion for reaction of 7R with 2-propanol
can be explained by in situ anomerization of DISAL donor
7R prior to the slow glycosylation (Table 2, entry 5).
Op tim iza tion of Glycosyla tion Con d ition s. Next,
we turned our attention to the glycosylation of cyclohex-
anol as a model substrate. The anomers of cyclohexyl
glucoside 16R,â can readily be separated by RP-HPLC.
In initial studies, DMF caused total breakdown of gly-
cosyl donor 7. This has also been observed for a glycosyl
bromide under Koenigs-Knorr glycosylation conditions.26
Other solvents such as DMSO, acetone, 1,3-dimethyl-
3,4,5,6-tetrahydro-2(1H)-pyrimidone (DMPU), or N,N-
diethylformamide were also not compatible with 7. Ethyl
acetate, dioxane, dimethoxyethane, dichloromethane, 1,2-
dichloroethane, or toluene were tolerated but gave slug-
gish reactions. The best solvents proved to be N,N-
dimethylacetamide (DMA), 1-methyl-2-pyrrolidinone
(NMP), or mixtures of the latter with acetonitrile or
nitromethane. Nitromethane27 and acetonitrile in them-
selves were good solvents, but required somewhat higher
reaction temperatures.
base with 2,6-lutidine, addition of 4 Å molecular sieves,
and use of argon atmosphere did not improve the yield
(Table 3, entry 3). However, a solvent mixture of NMP
and acetonitrile gave a higher yield (Table 3, entry 4).
Interestingly, the R/â ratio in glycosyl donor 7 did not
affect the yield or R/â ratio of product cyclohexyl glycoside
16. In all cases, the R/â ratio was about 2:1 (Table 3,
entries 4 and 5). Similarly, in nitromethane a high
amount of 16R was obtained, though a higher tempera-
ture was needed to drive the reaction to completion
(Table 3, entry 6). A mixture of NMP and nitromethane
gave the cleanest reaction and highest yield, with an R/â
ratio of 2.1:1 (Table 3, entry 7). Use of THF as solvent
gave incomplete conversion using the standard protocol,
but using more forcing conditions, the product was
isolated in 78% yield with even higher R-selectivity (Table
3, entry 8) than in nitromethane. Changing the solvent
to NMP-THF again gave complete conversion and an
R/â ratio of 2.1:1 (Table 3, entry 9). The para donor 10
showed similar glycosyl donor properties as 7. For the
glycosylation of cyclohexanol, NMP again gave the best
results (Table 3, entry 10). For ease of comparison,
reaction times were kept constant; however, in neat NMP
glycosylations of cyclohexanol (5 equiv) with 7 or 10 were
complete within 2 h at 40 °C as observed by the
disappearance of the donor.
It thus appears that the determining factor for the
reaction rate was the solvent effect. If the solvent had a
sufficiently high polarity, such as NMP, the glycosyla-
tions occurred smoothly.
These experiments clearly demonstrated that aryl
glycosides 7 and 10 were indeed glycosyl donors under
neutral conditions. Nonnucleophilic bases (triethylamine,
Hu¨nig’s base, 2,6-lutidine, etc.) were also tolerated under
these reaction conditions but were not required to
promote the reaction. The fact that glycosylations also
occurred in the presence of bases indicated that the
glycosylations were not autocatalytical promoted by the
released, acidic methyl 2-hydroxy-3,5-dinitrobenzoate.
In a control experiment, 2,4-dinitrophenyl 2,3,4,6-tetra-
O-benzyl-â-D-glucopyranoside, 1, gave no reaction or even
breakdown when treated with cyclohexanol (5 equiv) and
2,6-lutidine in NMP for 4 days at 40 °C,28 thus confirming
our initial hypothesis that more electron withdrawing
groups were required on the leaving group.
In the glycosylation of cyclohexanol (Table 3), NMP
proved superior to DMA in the presence of triethylamine
(Table 3, entries 1 and 2), whereas substitution of the
(25) From 7 (R/â 1:9) in CD3OD-CH3CN (1:5) 75% yield of deuter-
ated methyl glycoside (R/â 5:1) was obtained. 1H and 13C NMR showed
no trace of a putative OCH3 aglycon clearly indicating that formation
of the methyl glycoside was not due to transfer of the ester methoxide
but due to an intermolecular attack by CD3OD.
(26) Lloyd, P. F.; Roberts, G. P. J . Chem. Soc. 1963, 2962-2971.
(27) An important issue was the potential deprotonation of ni-
tromethane by bases stronger than 2,6-lutidine (e.g., triethylamine)
to give a stabilized anion. This anion appeared to be able to cleave the
aryl glucoside as reactions under these conditions gave 2 and a new
aromatic compound, indicative of cleavage by nucleophilic aromatic
substitution at the aryl leaving group via a Meisenheimer complex (for
a review of Meisenheimer complexes see: Terrier, F. Chem. Rev. 1982,
82, 77-152). Thus, when nitromethane was used, a judicious choice
of base was required. These observations are in agreement with
literature reports describing the attack of morpholine on a peracety-
lated 2,4-dinitrophenyl glycoside to give N-(2,4-dinitrophenyl)morpho-
line (see: Lindberg, B. Acta Chem. Scand. 1950, 4, 49-51).
A general tendency in the glycosylation with DISAL
donors was that more hindered alcohols gave increasing
(28) The stability of the 2,4-dinitrophenyl â-D-glucoside 1 to ethanol
is implicit in the procedure by van Boom, as the final products were
recrystallized from ethanol (see ref 14).

6272 J . Org. Chem., Vol. 66, No. 19, 2001
Petersen and J ensen
Sch em e 6
F igu r e 1. In situ anomerization experiment of 7â in NMP
with methyl 2-hydroxy-3,5-dinitrobenzoate (1 equiv): 7R ([),
7â (9), hydrolysis (b).
charide 20 was formed in somewhat lower yield (Table
4, entries 5-7). It proved advantageous to raise the
reaction temperature to 60 °C as unreacted donor was
still observed after 17 h at 40 °C. Interestingly, only the
R-anomer of disaccharide 20 was obtained. The lower
yield obtained when using DIEA as base (3 equiv)
appeared to be due to increased formation of 2,3,4,6-tetra-
O-benzyl-2-hydroxy-D-glucal, which was the major byprod-
uct in all cases. The yield for the glycosylation of sec-
alcohol 19 was improved to 74% by addition of a second
1.5 equiv of donor 7 after 6h (Table 4, entry 8). Tetra-
benzyl glucose 2 which carries a free 1-OH was glycosy-
lated to yield trehalose derivatives (Table 4, entry 9). By
NMR, the products isolated were determined to be 21R,R
(18%) and 21R,â (28%), but no â,â product could be
observed.
The para glycosyl donor 10 also proved effective in the
glycosylation of monosaccharide 17 under the same
conditions as for 7 to give the disaccharide 18 in 82%
yield (Table 4, entry 10). Glycosylation of the secondary
hydroxyl in 19 proceeded in 46% yield of the pure
R-anomer 20R (Table 4, entry 11). Mannose donor 12R
also proved to be an efficient donor under these conditions
to produce the 1,6-linked disaccharide (22R,â) in good
yields (Table 4, entries 12 and 13).
F igu r e 2. In situ anomerization experiment of 7R in NMP
with methyl 2-hydroxy-3,5-dinitrobenzoate (1 equiv): 7R ([),
7â (9), hydrolysis (b).
degrees of R-selectivity. We speculated that the lack of
inversion in the glycosylation of alcohols other than
methanol and the observed R-selectivity was due to in
situ anomerization of the DISAL donor prior to slower
reaction with the alcohol. To test this hypothesis, donor
7 was dissolved in NMP or CH3NO2 together with methyl
2-hydroxy-3,5-dinitrobenzoate29 at 20 °C and the change
in composition was monitored by HPLC. In NMP, donor
7â (R/â 1:14.1) anomerized within 2 h to give an R/â ratio
of 3:1 (Figure 1). Similarly, donor 7â (R/â 8.4:1) anomer-
ized within 1-2 h to give an R/â ratio of 3:1 (Figure 2).
However, in CH3NO2 no anomerization was observed.
Furthermore, alkyl glycosides were stable under these
mild conditions and did not anomerize. It thus appears
likely that the R-selectivity in the glycosylation of more
hindered alcohols can be explained by an in situ ano-
merization of the DISAL donor prior to nucleophilic
attack by the alcohol.
We also studied benzoyl protected aryl glycoside 9R,â
as a potential glycosyl donor, but in model studies this
proved unsuccessful.30
The much increased reactivity of the DISAL glycosyl
donors compared to the corresponding 2,4-dinitrophenyl
glycosides can at least partially be attributed to the
additional electron withdrawing substituent on the leav-
ing group. Heterolysis of the glycosidic bond to form the
oxonium phenoxide complex should be favored by polar
Disa cch a r id e Syn th esis. We then turned to the
glycosylation of monosaccharide acceptors (Scheme 6). It
was rewarding to observe that reaction of diisopropy-
lidene protected galactose 17 carrying a primary hydroxy
with glycosyl donor 7 gave the resultant disaccharide 18
in yields from 74 to 79% as R/â mixtures. Reactions were
carried out as above in NMP either neat or in mixtures
with acetonitrile or nitromethane (Table 4, entries 1-3)
in the absence of added base and with an excess of
acceptor (2 equiv). The yields could be further improved
using an excess of donor (1.5 equiv) to 90% of disaccharide
18 after column chromatography (Table 4, entry 4).
Turning to the diisopropylidene-protected glucose deriva-
tive 19, which carries a free secondary hydroxy, disac-
(30) The R-configured tetrabenzoate 9R failed to react with neat
methanol and application of more forcing conditions with cyclohexanol
using the strong base 2-tert-butyl-1,1,3,3-tetramethylguanidine (TBT-
MG) in neat NMP resulted only in breakdown of the donor and
formation of 1,2,3,4,6-penta-O-benzoyl-â-D-glucopyranose. The aryl
â-glycoside 9â gave the ortho ester when treated with cyclohexanol
under basic conditions in THF. These results clearly show that benzoyl
protection gives a significantly less reactive donor, with the R-anomer
being practically unreactive and the â-anomer giving the ortho ester.
This is due to intermolecular benzoyl transfer and the neighboring
group participation from the 2-O-benzoyl moiety, respectively. Under
the nonacidic conditions employed here the ortho ester cannot rear-
range to the corresponding â-glycoside.
(29) Bartlett, P. D.; Trachtenberg, E. N. J . Am. Chem. Soc. 1958,
80, 5808-5810.

Glycosylation Method for Oligosaccharide Synthesis
J . Org. Chem., Vol. 66, No. 19, 2001 6273
Ta ble 4. Glycosyla tion of Mon osa cch a r id e Accep tor s
DISAL donor
acceptor
(equiv)
yielda (%)
(R/â)
entry
(R/â)
solvent
disaccharide
1
2
3
4
5
6
7
8
9
10
11
12
13
7 (3.7:1)
7 (3.7:1)
7 (3.7:1)
7 (3.7:1)
7 (3.7:1)
7 (4.6:1)
7 (4.6:1)
7 (4.6:1)
7 (3.7:1)
10 (4.1:1)
10 (4.1:1)
12r
17 (2.0)
17 (2.0)
17 (2.0)
NMP/CH3CNb
NMP/CH3NO2b
NMPb
18r,â
18r,â
18r,â
18r,â
20r
74 (2.1:1)
78 (2.1:1)
79 (2.3:1)
90 (2.4:1)
36
42
33
17 (0.67)
19 (0.67)
19 (0.67)
19 (0.67)
19 (0.33)
2 (0.67)
17 (0.67)
19 (0.67)
17 (0.67)
17 (0.67)
NMPb
NMPb
NMPc
20r
NMPc,d
NMPc
20r
20r
74
NMPb
21
18 (R,R), 28 (R,â)
82 (2.3:1)
46
78 (1.4:1)
84 (1.6:1)
NMPb
18r,â
NMPb
20r
NMPb
22r,â
22r,â
12r
NMPc
a
b
d
Isolated yields. 40 °C for 17 h. c 60 °C for 17 h. Et3N was added (3 equiv).
from Fluka (Sigma-Aldrich Denmark). 1,4-Dimethylpiperazine
was fractionally distilled and stored over 4 Å molecular sieves.
Melting points were uncorrected. All solvents were distilled
and/or stored over 3 or 4 Å molecular sieves as appropriate.
1H NMR spectra were recorded on either a Varian Mercury
300 operating at 300.06 MHz equipped with a 4-nuclei probe
or a Varian Unity Inova 500 operating at 499.87 MHz equipped
solvents such as NMP, and this corresponds well with
our findings. However, we speculated that an additional
factor for the increased reactivity could be displacement
of the phenoxide by nucleophilic attack of the aglycon
carbonyl oxygen to form an oxonium ion. The reactive
species thus formed could then react to form O-glycosides.
However, as the glycosyl donor 10 with a para carboxy
methyl moiety behaved much as the ortho analogue (7),
this potential pathway seemed insignificant. Participa-
tion of the nucleophilic solvent by interception of reactive
intermediates or attack by the solvent on the donor to
form other reactive species is another potential pathway.
with a z- (single axis) PFG inverse detection C-H-P probe. 13
C
NMR were recorded on a Varian Mercury 300 operating at
75.46 MHz with relaxation delays of up to 30 s to detect all
signals arising from quaternary carbons. Chemical shift (δ)
values are in ppm, coupling constants (J ) are in Hz. All
assignments were supported by 2D homonuclear chemical-shift
correlation spectroscopy (gCOSY) and heteronuclear single
quantum correlated spectroscopy (gHSQC) experiments. Thin-
layer chromatography was performed on Merck silica gel 60
F254 plates and spots were visualized by UV light at 254 nm
and/or spraying with 10% aqueous H2SO4 followed by heating.
Vacuum liquid chromatography was carried out on Merck
silica gel 60H. HPLC analyses were carried out on a Waters
system (600 control unit, 996 photodiode array (PDA) detector,
717 Plus autosampler, Millenium32 control software) on a
Waters Nova-Pak C18 column (3.9 × 5.0 mm cartridge; 4 µm
particle size) using a linear gradient of 0.1% aqueous TFA (A)
and 0.1% TFA in CH3CN (B): 0 min: 0% B, 2 min: 0% B, 5
min: 50% B, 12 min: 95% B, 13 min: 95% B, 13.5 min: 0%
B, 20 min: 0% B. Monitoring was from 200 to 400 nm,
integrations were performed at 215 and 265 nm, and indi-
vidual peaks were analyzed by their UV spectra. The purity
of compounds was determined from integrations at 215 nm.
Prep. HPLC was performed on a Waters system with a Delta
600 pump and a 996 PDA detector using a stack of three 40 ×
100 mm Prep Nova-Pak HR C18 6 µm 60 Å units eluting with
gradients of H2O-CH3CN. MS analyses were performed on a
Micromass LCT mass spectrometer.
Con clu sion s
A new type of glycosyl donors was synthesized in high
yields and with good control of the R/â ratio. Benzyl-
protected (i.e., “armed”) aryl glycosides 7, 10, and 12 were
efficient glycosyl donors for O-glycosylation under neutral
or mild basic conditions. Glycosylation of methanol under
solvolytic conditions was stereospecific with inversion of
anomeric configuration, whereas with sterically more
demanding alcohols, R-selectivity was observed, most
likely by in situ anomerization of the DISAL donor.
DISAL donors were shown to undergo in situ anomer-
ization under the reaction conditions. Glycosylations of
monosaccharides and cyclohexanol proceeded in high
yields and with R-selectivity in various solvents with
NMP as the solvent of choice. The exclusive formation of
R-linked disaccharides in glycosylation of sec-alcohol 19
amounts to a dynamic kinetic resolution. The ‘normal′
solvent effects used to direct the R/â ratio in Lewis acid
promoted glycosylations, were not observed here. These
donors are the first members in a new class of glycosyl
donors, which are stable upon storage, yet do not require
activation by a Lewis acid for efficient glycosylations.
Compared to glycosyl iodides, the DISAL glycosyl donors
have distinct advantages as they are stable on storage
for a prolonged period, yet appear to be more reactive.
Furthermore, the phenolic leaving group lends itself to
further “fine-tuning” of reactivity of the glycosyl donor.
Finally, aryl fluoride precursors were easily synthesized
in large quantities from inexpensive starting materials.
2-F lu or o-3,5-d in itr oben zoic Acid (3). To 65% fuming
sulfuric acid (160 mL) in a 500 mL conical flask cooled on an
ice bath was slowly added fuming nitric acid (80 mL) over 15
min. The ice-cold and now clear mixture was stirred effectively
while 2-fluorobenzoic acid (14.01 g, 0.10 mol) was suspended
herein in small portions over 5 min. The temperature was
slowly raised to 90 °C by means of an oil bath and the reaction
mixture again became clear. The mixture was kept at 90 °C
for 1 h, during which time the reaction mixture showed slight
foaming and evolution of brown gas. When the reaction
subsided, a drop was withdrawn and distributed between ethyl
acetate and water. TLC of the organic phase showed only one
compound (Rf 0.10-0.15; EtOAc-AcOH, 100:1). The reaction
mixture was poured onto crushed ice (1 L), diluted to ap-
proximately 1.5 L with water, and extracted with ethyl acetate
(4 × 100 mL). The combined organic phases were dried
(MgSO4) and concentrated to yield slightly off-white crystals
(19.22 g, 84%). Mp: 193-196 °C (lit.16 mp 200 °C). Purity:
>99%. Anal. Calcd for C7H3FN2O6: C, 36.54; H, 1.31; N, 12.17.
Exp er im en ta l Section
Gen er a l P r oced u r es. 2,3,4,6-Tetra-O-benzyl-D-glucose was
purchased from Pfanstiehl (Cheshire, U.K.) and CMS Chemi-
cals Ltd (Oxford, U.K.); the anomeric distribution was deter-
mined from the intensity of the anomeric carbons in 13C NMR.
2-Fluorobenzoic acid and 4-fluorobenzoic acid were obtained
Found: C, 36.64; H, 0.88; N, 12.18. 1H NMR (500 MHz,
4
acetone-d6): δ ) 12.3 (br s, 1H, COOH), 9.14 (dd, 1H, J H,F
)

6274 J . Org. Chem., Vol. 66, No. 19, 2001
Petersen and J ensen
4
4
4
5.8 Hz, J H,H ) 3.1 Hz), 9.05 (dd, 1H, J H,F ) 5.5 Hz, J H,H
)
)
11, Li2CO3 (2 equiv), and aryl fluoride 4 or 6 (1.2 equiv) were
suspended in CH2Cl2 (1-2 mL/mmol; dried over 4 Å molecular
sieves). Reactions proceeded in the presence of either DMAP
or DMP. (1) For reactions with DMAP catalysis, DMAP (0.3
equiv) was dissolved in CH2Cl2 (0.5 mL) and added to the
suspension in five portions over 20 min. After addition of the
first portion, the color changed from slightly discolored to
strongly yellow/brown. After stirring for an additional 10 min,
products were isolated by VLC chromatography (30 × 80 mm;
CH2Cl2-Et2O 1:0 f 19:1). Concentration of appropriate frac-
tions gave the yields in Table 1. (2) For reactions with 1,4-
dimethylpiperazine catalysis, DMP (distilled and stored over
4 Å molecular sieves; 0.5 equiv) was added and the mixture
stirred for 2-3 h. Remaining aryl fluoride was then quenched
by addition of DMAP18 (0.3 equiv). After the mixture was
stirred for an additional 15 min, isolation as above gave yields
listed in Table 1.
Gen er a l Meth od for Solvolysis, Mod el Glycosyla tion s,
a n d HP LC An a lysis (Ta bles 2 a n d 3). The initial evaluation
of solvents for the glycosylation was performed by reacting
glycosyl donors 7R,â or 10R,â (typically 0.02 mmol) with
cyclohexanol and either triethylamine, 2,6-lutidine, or in the
absence of base in the solvent or neat alcohol as indicated
(400-600 µL) on an Eppendorf Thermomixer 5436 (combined
heater/shaker). After the mixture was shaken for the indicated
period, a sample (10-40 µL) was diluted in acetonitrile (0.70-
1.00 mL) and analyzed by analytical RP-HPLC. Reported
yields are based on integrated areas.
Gen er a l Meth od for Glycosyla tion s a n d Isola tion
(Ta bles 4). The glycosyl donor (7R,â, 9R,â, 10R,â, or 12R,â)
and acceptor (cyclohexanol, 17, or 19) were placed in a 1.5 mL
Eppendorf tube, the indicated solvent (1 mL) added, and the
mixture shaken on an Eppendorf Thermomixer 5436 for the
time and temperature stated. After the indicated reaction time,
reaction mixtures with volatile solvents were concentrated and
products isolated by VLC chromatography, whereas NMP-
containing reactions were extracted prior to chromatography.
Reaction mixtures were diluted with ethyl acetate (30 mL) and
washed with brine (2 × 20 mL), aqueous NaOH (0.5 M; 2 ×
20 mL), and brine (2 × 20 mL). The organic phases were dried
(Na2SO4) and concentrated, the residues were dissolved in a
small volume of CH2Cl2, and products were isolated by VLC
chromatography (20 × 80 mm, gradient of toluene, 20 mL,
ethyl acetate-toluene 1:60, 20 mL, 1:20, 40 mL, and 1:10, 100
mL). Reactions in NMP were injected directly into the pre-
parative HPLC. Appropriate fractions were concentrated in
vacuo at 30 °C, traces of toluene were removed by coevapora-
tion twice with CH2Cl2, and the product was dried under high
vacuum to give the isolated yields in Table 4.
3.1 Hz). 13C NMR (75 MHz, acetone-d6): δ ) 161.5 (d, J C,F
3
3.1 Hz, COOH), 158.1 (d, 1J C,F ) 283.3 Hz), 143.2, 139.3, 131.9
3
2
(d, J C,F ) 3.6 Hz), 125.5, 123.4 (d, J C,F ) 12.5 Hz).
Meth yl 2-F lu or o-3,5-d in itr oben zoa te (4). Benzoic acid
derivative 3 (6.90 g, 30 mmol) was suspended in CH2Cl2 (50
mL), and DMF (0.050 mL) and oxalyl chloride (2.83 mL, 33
mmol) were added. After being stirred at room temperature
for 1 h, the reaction had become a clear solution and a small
sample was withdrawn and diluted with dry methanol; TLC
showed complete conversion to the methyl ester (Rf 0.84;
EtOAc-AcOH, 100:1). Dry methanol (6 mL) was added, and
the reaction mixture was stirred for 30 min. Elution through
a VLC column (30 × 30 mm) with CH2Cl2 (100 mL) gave after
evaporation of the relevant fractions an oil that was redis-
solved in pentane (50 mL) and CH2Cl2 (50 mL). The flask was
placed in a freezer, and when the crystallization had set in,
the solvent was partially evaporated, hexane (50 mL) was
added, and the procedure was repeated twice. Filtration of the
suspension gave a solid that after washing (hexane, 2 × 30
mL) and drying yielded crystals (7.10 g, 97%). Mp: 73-74 °C.
Purity: >99%. Anal. Calcd for C8H5FN2O6: C, 39.36; H, 2.06;
N, 11.47. Found: C, 39.53; H, 1.80; N, 11.47. 1H NMR (300
MHz, CDCl3): δ ) 9.05-9.00 (m, 2H), 4.06 (s, 3H); 13C NMR
(75 MHz, CDCl3): δ ) 161.3 (d, 3J C,F ) 4.2 Hz, COOMe), 158.0
1
3
(d, J C,F ) 287.4 Hz), 143.0, 139.1, 132.0 (d, J C,F ) 6.3 Hz),
125.3, 123.3 (d, J C,F ) 11.9 Hz), 54.0 (Me).
2
4-F lu or o-3,5-d in itr oben zoic Acid (5). Prepared from
4-fluorobenzoic acid by a literature procedure17 (76% yield).
Mp: 230-232 °C (lit.17 mp 235-237 °C). Purity: 98%. 1H NMR
(300 MHz, acetone-d6): δ ) 11.0 (br. s., 1H, COOH), 8.97 (d,
2H, 4J H,F ) 6.6 Hz). 13C NMR (75 MHz, acetone-d6): δ ) 162.9
(d, J C,F ) 1.3 Hz, COOH), 152.1 (d, J C,F ) 283.9 Hz), 139.4,
131.7 (d, J C,F ) 1.1 Hz), 127.5.
Meth yl 4-F lu or o-3,5-d in itr oben zoa te (6). Benzoic acid
derivative 5 (1.15 g, 5 mmol) was dissolved in THF (5 mL),
followed by addition of DMF (5 µL) and oxalyl chloride (0.52
mL, 6 mmol). After being stirred for 15 min, the mixture was
added to a stirred solution of methanol (1.5 mL) in hexane
(50 mL) cooled on an ice bath. After 5 min, the solid was
collected by filtration, washed (hexane, 2 × 10 mL), and dried
under vacuum (0.73 g; 60%). Mp: 96-97 °C (lit.31 mp 93.5 °C).
Purity: 98%. 1H NMR (300 MHz, CDCl3): δ ) 8.93 (d, 2H,
4J H,F ) 6.1 Hz), 4.05 (s, 3H); 13C NMR (75 MHz, CDCl3): δ )
5
1
3
5
1
162.4 (d, J C,F ) 1.1 Hz, COOMe), 151.9 (d, J C,F ) 288.5 Hz),
139.2, 131.7 (d, J C,F ) 1.1 Hz), 127.3 (d, J C,F ) 5.7 Hz), 53.9
(Me).
3
2
2,3,4,6-Tetr a -O-ben zoyl-D-glu cop yr a n ose (8). 1,2,3,4,6-
Penta-O-benzoyl-D-glucopyranoside32 (2.80 g, 4.0 mmol; R/â )
1:6) was dissolved in DMF (25 mL), and upon cooling to 3 °C,
hydrazinium acetate (1.11 g, 12.0 mmol; 3 equiv) was added.
After being stirred for 17 h, the mixture was diluted with ethyl
acetate (60 mL) and washed with brine (2 × 30 mL), aqueous
HCl (1M, 2 × 30 mL), and brine (2 × 30 mL). The organic
phase was dried (MgSO4), concentrated, and dried under high
vacuum to give a solid (2.39 g, 100%) containing the title
compound (R/â ) 4.1:1) in 97% purity. The purity was
increased by dissolving the solid in diethyl ether (25 mL) and
filtering through 5 mm of silica gel using additional diethyl
ether (30 mL). Heptane (80 mL) was added to the filtrate, and
the mixture was left in a beaker overnight allowing the diethyl
ether to evaporate. The solid was collected by filtration, washed
with heptane (2 × 15 mL), and air dried. This yielded a pure
product (1.57 g; 82%) as an anomeric mixture (R/â 4:1). 1H
NMR data for the R-anomer was identical to literature
values.33
P h ysica l Da ta for Glycosyl Don or s 7r,â, 9r,â, 10r,â,
a n d 12r. 2,4-Dinitro-6-(methoxy carbonyl)phenyl 2,3,4,6,-
1
tetra-O-benzyl-R-D-glucopyranoside (7R). H NMR (500 MHz,
4
4
CDCl3): δ ) 8.66 (d, 1H, J H,H ) 2.9 Hz), 8.54 (d, 1H, J H,H
)
2.9 Hz), 7.40-6.96 (m, 20H), 5.42 (d, 1H, 3J 1,2 ) 3.4 Hz, H-1R),
2
2
4.91 (dd, 2H, J gem ) 16.2 and 10.7 Hz, Bn), 4.86 (d, 1H, J gem
2
) 11.1 Hz, Bn), 4.64 (d, 1H, J gem ) 11.5 Hz, Bn), 4.55 (dd,
2H, 2J gem ) 11.1 and 3.4 Hz, Bn), 4.47 (d, 1H, 2J gem ) 12.0 Hz,
Bn), 4.34 (d, 1H, J gem ) 11.5 Hz, Bn), 4.11 (t, 1H, 3J ) 9.4
2
3
3
3
Hz, H-3), 3.99 (ddd, 1H, J 4,5 ) 9.4, J 5,6 ) 3.4, J 5,6′ ) 2.1 Hz,
2
H-5), 3.90 (s, 3H, OMe), 3.74 (dd, 1H, J gem ) 11.1 Hz, and
3J 5,6 ) 3.4 Hz, H-6), 3.73 (t, 1H, J ) 9.4 Hz, H-4), 3.64 (dd,
3
1H, 2J gem ) 11.1 Hz, 3J 5,6′ ) 2.1 Hz, H-6′), 3.62 (dd, 1H, 3J 2,3
)
9.4, 3J 1,2 ) 3.4 Hz, H-2); 13C NMR (75 MHz, CDCl3): δ ) 163.9
(carbonyl), 154.4, 144.9, 141.9, 138.6, 138.3, 137.9, 137.1, 129.8,
128.8-127.6 (m; benzyl), 123.5, 104.4 (C-1R), 81.6, 81.1, 76.7,
76.1, 75.2, 74.5, 74.1, 73.7, 68.2, 53.6 (OMe); ESI (positive
mode) of an anomeric mixture: m/z calcd for C42H42N2O13 (M
+ NH4)+ 782.3; found 782.3, 800.1 (M + H2O + NH4)+. The
title compound was stored as a foam at -10 °C.
Gen er a l Meth od for P r ep a r a tion of Glycosyl Don or s.
In a dried flask, protected carbohydrate derivatives 2, 8, or
2,4-Din itr o-6-(m eth oxyca r bon yl)p h en yl 2,3,4,6-Tetr a -
O-ben zyl-â-D-glu cop yr a n osid e (7â). 1H NMR (300 MHz,
CDCl3): δ ) 8.81 (d, 1H, 4J ) 2.7 Hz), 8.71 (d, 1H, 4J ) 2.7
(31) Litvinenko, L. M.; Titskii, G. D.; Shumeiko, A. E. Zh. Org. Khim.
1977, 13, 767-770.
(32) Ness, R. K.; Hewitt, G.; Hudson, C. S. J . Am. Chem. Soc. 1950,
72, 2200-2205.
3
Hz), 7.45-7.13 (m, 20H), 5.18 (d, 1H, J 1,2 ) 7.0 Hz, H-1â),
2
2
5.03 (d, 1H, J gem )10.9 Hz), 4.96 (d, 1H, J gem ) 11.1 Hz),
(33) Salinas, A. E.; Sproviero, J . F.; Deulofeu, V. Carbohydr. Res.
1987, 170, 71-99.
2
4.84-4.77 (m, 3H), 4.56 (d, 1H, J gem ) 10.9 Hz), 4.50 (d, 1H,

Glycosylation Method for Oligosaccharide Synthesis
J . Org. Chem., Vol. 66, No. 19, 2001 6275
2
3
3
2J gem ) 11.8 Hz), 4.39 (d, 1H, J gem ) 11.9 Hz), 3.85 (s, 3H;
CDCl3): δ ) 8.80 (d, 1H, J ) 2.9 Hz, Ar-H), 8.68 (d, 1H, J )
2.9 Hz, Ar-H), 7.44-7.15 (m, 20H, benzyl), 5.63 (d, 1H, 3J 1,2
2.0 Hz, H-1), 4.86-4.36 (m, 8H, benzyl methylene), 4.32 (dd,
3
3
OMe), 3.76 (dd, 1H, J 2,3 ) 8.5, J 1,2 ) 7.0 Hz, H-2), 3.71 (dd,
)
1H, 3J ) 8.8 Hz, H-3), 3.63 (dd, 1H, 3J ) 9.1 Hz, H-4), 3.56
2
3
3
3
3
(br. d, 2H, J gem ) 4.5 Hz, 2 × H-6), 3.36 (dt, 1H, J 4,5 ) 9.2,
3J 4,5 ) 3.3 Hz, H-5); 13C NMR (75 MHz, CDCl3): δ ) 163.5
(carbonyl), 151.6, 146.7, 145.8, 143.2, 138.6, 138.2, 138.1, 138.1,
130.6, 129.3, 128.7-127.9 (m; benzyl), 123.0, 104.8 (C-1â), 84.5,
82.5, 77.5, 76.2, 75.9, 75.5, 75.4, 73.8, 69.0, 53.7 (OMe). The
title compound was stored as a foam at -10 °C.
1H, J 2,3 ) 3.0 Hz, J 2,1 ) 2.0 Hz, H-2), 4.13 (dd, 1H, J 3,4 )
3 3 3
9.2 Hz, J 4,5 ) 9.4 Hz, H-4), 4.00 (dd, 1H, J 2,3 ) 3.0 Hz, J 3,4
3
) 9.2 Hz, H-3), 3.90 (s, 3H, OMe), 3.72 (dd, 1H, J gem ) 10.8
3
3
3
Hz, J 5,6 ) 4.0 Hz, H-6), 3.63 (ddd, 1H, J 4,5 ) 9.4 Hz, J 5,6
)
3
3
4.0 Hz, J 5,6′ ) 1.7 Hz, H-5), 3.53 (dd, 1H, J gem ) 10.8 Hz,
3J 5,6′ ) 1.7 Hz, H-6′). Heteronuclear coupling constant: J C1,H1
2
2,4-Din itr o-6-(m eth oxyca r bon yl)p h en yl 2,3,4,6-Tetr a -
) 178.1 (R). 13C NMR (75 MHz, CDCl3): δ ) 163.1 (carbonyl),
153.7, 145.5, 142.1, 138.5, 138.5, 138.3, 129.8, 128.6-127.7 (m;
benzyl), 123.5, 104.9 (C-1), 79.2, 75.6, 75.2, 75.1, 73.9, 73.6,
73.3, 72.8, 68.8, 53.7. ESI (positive mode) m/z calcd for
O-ben zoyl-r-D-glu cop yr a n osid e (9r). Mp: 91-92 °C. 1H
4
NMR (300 MHz, CDCl3): δ ) 8.59 (d, 1H, J ) 2.8 Hz), 8.57
4
(d, 1H, J ) 2.8 Hz), 8.00-7.85 (m, 8H), 7.60-7.30 (m, 12H),
6.35 (t, 1H, 3J ) 9.8 Hz, H-3), 6.32 (d, 1H, 3J 1,2 ) 3.7 Hz, H-1R),
C
42H42N2O13 (M + NH4)+ 782.2687; found 782.0, 800.1 (M +
5.73 (t, 1H, 3J ) 9.8 Hz, H-4), 5.55 (dd, 1H, J 2,3 ) 9.8 Hz,
H2O + NH4)+. The title compound was stored as a foam at
-10 °C.
3
3J 1,2 ) 3.7 Hz, H-2), 4.58-4.39 (m, 3H; H-5, 6, and 6′), 4.00 (s,
3H; OMe). 13C NMR (75 MHz, CDCl3): δ ) 165.9, 165.8, 165.6,
165.4, 163.5, 151.4, 144.0, 141.7, 137.0, 134.0, 133.9, 133.6,
133.5, 130.3-128.6 (m), 127.4, 123.5, 99.0 (C-1R), 71.7, 71.3,
69.5, 68.6, 62.5, 54.1 (OMe). ESI (positive mode) of an anomeric
mixture: m/z calcd for C42H34N2O17 (M + NH4)+ 838.2; found
837.8. The title compound was stored as an off-white solid at
5 °C.
P h ysica l Da ta for Glycosid es 13r,â, 14r,â, 15r,â, 16r,â,
18r,â, 20r, a n d 22r,â: Meth yl 2,3,4,6-Tetr a -O-ben zyl-r,â-
D-glu cop yr a n osid e (13 r,â). Observed data were identical
to literature values.34
Isop r op yl 2,3,4,6-tetr a -O-ben zyl-r,â-D-glu cop yr a n osid e
(14r,â). Observed data were identical to literature values.35
ter t-Bu t yl 2,3,4,6-Tet r a -O-b en zyl-r,â-D-glu cop yr a n o-
sid e (15 r,â). Observed data were identical to literature
values.36
2,4-Din itr o-6-(m eth oxyca r bon yl)p h en yl 2,3,4,6-Tetr a -
O-ben zoyl-â-D-glu cop yr a n osid e (9â). Mp: 97.5-100 °C. 1H
4
NMR (300 MHz, CDCl3): δ ) 8.70 (d, 1H, J ) 2.9 Hz), 8.63
Cycloh exyl 2,3,4,6-Tetr a -O-ben zyl-r,â-D-glu cop yr a n o-
sid e (16r,â). Observed data were identical to literature
values.37
(d, 1H, 4J ) 2.9 Hz), 8.10-8.03 (m, 2H), 7.95-7.82 (m, 6H),
3
7.60-7.30 (m, 12H), 5.99 (t, 1H, J ) 9.6 Hz, H-3), 5.86 (dd,
3
3
3
1H, J 2,3 ) 9.7 Hz, J 1,2 ) 7.7 Hz, H-2), 5.67 (t, 1H, J ) 9.7
Hz, H-4), 5.66 (d, 1H, 3J 1,2 ) 7.7 Hz, H-1â), 4.50-4.38 (m, 2H;
H-6 and 6′), 4.15-4.07 (m, 1H; H-5), 3.80 (s, 3H; OMe). 13C
NMR (75 MHz, CDCl3): δ ) 165.9, 165.8, 165.2, 165.2, 162.9
(carbonyl), 150.8, 146.6, 143.7, 133.8, 133.6, 133.5, 130.9,
130.3-128.5 (m), 123.1, 102.4 (C-1â), 73.3, 72.6, 72.0, 69.5,
62.7, 53.5 (OMe). The title compound was stored as an off-
white solid at 5 °C.
Disa cch a r id es 18r,â, 20r, a n d 22r,â. Observed data were
identical to literature values.38
Octa -O-ben zyl-r,r-tr eh a lose 21r,r. Selected data. 1H
NMR (300 MHz, CDCl3): δ ) 5.28 (d, 1H, 3J 1,2 ) 3.7 Hz, H-1R).
13C NMR (75 MHz, CDCl3): δ ) 94.6 ppm. HSQC showed
strong coupling between these signals. ESI (positive mode):
m/z calcd for C68H71O11 (M + H)+ 1063.5, found 1063.4; m/z
calcd for C68H72O12 (M + NH4)+ 1080.5, found 1080.5.
Octa -O-ben zyl-r,â-tr eh a lose 21r,â. Selected data. 1H
NMR (300 MHz, CDCl3): δ ) 5.17 (d, 1H, 3J 1,2 ) 3.4 Hz, H-1R),
4.59 (d, 1H, 3J 1,2 ) 7.7 Hz, H-1′â). 13C NMR (75 MHz, CDCl3):
δ ) 104.4 (C-1′â), 99.7 (C-1R). ESI (positive mode): m/z calcd
for C68H71O11 (M + H)+ 1063.5, found 1063.4; m/z calcd for
2,6-Din itr o-4-(m eth oxyca r bon yl)p h en yl 2,3,4,6-Tetr a -
1
O-ben zyl-r-D-glu cop yr a n osid e (10r). H NMR (500 MHz,
CDCl3): δ ) 8.49 (s, 2H), 7.40-7.00 (m, 20H; benzyl), 5.28 (d,
3
2
1H, J 1,2 ) 3.0 Hz, H-1R), 4.90 (2 × d, 2H, J gem ) 11.1 Hz),
2
2
4.83 (d, 1H, J gem ) 11.1 Hz), 4.65 (d, 1H, J gem ) 11.9 Hz),
2
2
4.55 (2×d, 2H, J gem )11.1 Hz), 4.43 (2×d, 2H, J gem ) 11.9
C
68H72O12 (M + NH4)+ 1080.5, found 1080.5.
Hz), 4.09 (t, 1H, 3J ) 9.4 Hz, H-3), 4.01 (s, 3H; OMe), 3.93
3
3
3
(ddd, 1H, J 4,5 ) 9.4 Hz, J 5,6 ) 3.0 Hz, J 5,6′ ) 1.7 Hz, H-5),
3.75 (dd, 1H, 2J gem )11.1 Hz, 3J 5,6′ ) 1.7 Hz, H-6′), 3.73 (t, 1H,
Ack n ow led gm en t. The authors thank the Alfred
3J ) 9.4 Hz, H-4), 3.63 (dd, 1H, J gem ) 11.1 Hz, J 5,6 ) 3.0
2
3
Benzon Foundation for a Bøje Benzon Fellowship (K.J.J.),
the Danish Technical Science Research Council (K.J .J .),
Danish Natural Science Research Council (K.J .J .),
Lundbeck Foundation (K.J .J .), and the Technical Uni-
versity of Denmark (L.P.).
Hz, H-6), 3.60 (dd, 1H, J 2,3 ) 9.4 Hz, J 1,2 ) 3.0 Hz, H-2). 13C
NMR (75 MHz, CDCl3): δ ) 163.1 (carbonyl), 147.3, 145.1,
138.7, 138.3, 138.0, 137.3, 129.5, 129.0-127.0 (m; benzyl),
125.5, 104.8 (C-1R), 81.5, 80.2, 76.6, 76.0, 75.2, 74.5, 74.4, 73.7,
68.1, 53.4 (OMe). ESI (positive mode) of an anomeric mix-
ture: m/z calcd for C42H42N2O13 (M + NH4)+ 782.3, found 782.0,
787.0 (M + Na)+. The title compound was stored as a foam at
-10 °C.
3
3
Su p p or tin g In for m a tion Ava ila ble: Detailed discussion
of the choice of bases and reaction kinetics for the formation
of glycosyl donors. This material is available free of charge
via the Internet at http://pubs.acs.org.
2,6-Din itr o-4-(m eth oxyca r bon yl)p h en yl 2,3,4,6-Tetr a -
O-ben zyl-â-D-glu cop yr a n osid e (10â). 1H NMR (500 MHz,
CDCl3): δ ) 8.60 (s, 2H), 7.45-7.10 (m, 20H; benzyl), 5.18 (d,
3
2
1H, J 1,2 ) 7.7 Hz, H-1â), 5.00 (d, 1H, J gem ) 10.7 Hz), 4.93
(d, 1H, 2J gem ) 11.1 Hz), 4.79 (dd, 2H, 2J gem ) 10.7 Hz, 2J gem
J O0057654
)
2.1 Hz), 4.77 (d, 1H, 2J gem ) 10.7 Hz), 4.54 (d, 1H, 2J gem ) 11.1
2
2
Hz), 4.47 (d, 1H, J gem ) 11.9 Hz), 4.41 (d, 1H, J gem ) 11.9
(34) (a) Wang, H.; Sun, L.; Glazebnik, S.; Zhao, K. Tetrahedron Lett.
1995, 36 (17), 2953-2956. (b) Fowler, P.; Bernet, B.; Vasella, A. Helv.
Chim. Acta 1996, 79, 269-287. (c) Dhawan, S. N.; Chick, T. L.; Goux,
W. J . Carbohydr. Res. 1988, 172, 297-307.
(35) Briner, K.; Vasella, A. Helv. Chim. Acta 1989, 72, 1371-1382.
(36) Yamanoi, T.; Nakamura, K.; Goto, M.; Furusawa, Y.; Takano,
M.; Fujioka, A.; Yanagihara, K.; Satoh, Y.; Hosokawa, H.; Inazu, T.
Bull. Chem. Soc. J pn. 1993, 66, 2617-2622.
(37) Kim, W.-S.; Hosono, S.; Sasai, H.; Shibasaki, M. Heterocycles
1996, 42 (2), 795-809.
(38) Physical data for 18R,â and 20R: Houdier, S.; Vottero, P. J . A.
Carbohydr. Res. 1992, 232, 349-352. 22R,â: Yamanoi, T.; Nakamura,
K.; Takeyama, H.; Yanagihara, K.; Inazu, T. Bull. Chem. Soc. J pn.
1994, 67, 1359-1366.
Hz), 4.02 (s, 3H; OMe), 3.78 (dd, 1H, 3J 2,3 ) 9.0 Hz, 3J 1,2 ) 7.7
3
3
Hz, H-2), 3.70 (t, 1H, J ) 9.0 Hz, H-3), 3.62 (dd, 1H, J 4,5
)
9.8 Hz, 3J 3,4 ) 9.0 Hz, H-4), 3.57 (br. d, 2H, 3J ) 3.4 Hz, 2×H-
3
3
6), 3.40 (dt, 1H, J 4,5 ) 9.8 Hz, J 5,6 ) 3.4 Hz, H-5). 13C NMR
(75 MHz, CDCl3): δ ) 163.1 (carbonyl), 146.4, 144.6, 138.6,
138.2, 138.1, 138.0, 129.1, 128.7-127.7 (m; benzyl), 127.2,
104.6 (C-1â), 84.3, 82.3, 77.3, 76.1, 75.9, 75.5, 75.3, 73.7, 68.9,
53.5 (OMe). The title compound was stored as a foam at -10
°C.
2,4-Din itr o-6-(m eth oxy a r bon yl)p h en yl 2,3,4,6-Tetr a -
O-ben zyl-r-D-m a n n op yr a n osid e (12r). 1H NMR (300 MHz,