Mechanistic studies and methods to prevent aglycon transfer of thioglycosides

Article abstract of DOI:10.1021/ja063247q

Thioglycosides have been employed extensively for the synthesis of complex oligosaccharides, carbohydrate libraries, and mimetics of O-glycosides. While very useful, aglycon transfer is a problematic side reaction with thioglycosides. In this paper, a series of mechanistic studies are described. The aglycon transfer process is shown to affect both armed and disarmed thioglycosides, cause anomerization of the carbon-sulfur bond of a thioglycoside, and destroy the product of a glycosylation reaction. The results indicate that the aglycon transfer process can be a major problem for a wide range of thioglycosides. This side reaction is especially important to consider when carrying out complex reactions such as solid-phase glycosylations, one-pot or orthogonal multicomponent glycosylations, and construction of carbohydrate libraries. To prevent transfer, a number of modified aglycons were examined. The 2,6-dimethylphenyl (DMP) aglycon was found to effectively block transfer in a variety of model studies and glycosylation reactions. The DMP group can be installed in one step from a commercially available thiol (2,6- dimethylthiophenol) and is useable as a glycosyl donor. On the basis of these features, the DMP group is proposed as a convenient and improved aglycon for thioglycosides.

Full text of DOI:10.1021/ja063247q

Published on Web 08/15/2006
Mechanistic Studies and Methods To Prevent Aglycon
Transfer of Thioglycosides
Zhitao Li and Jeffrey C. Gildersleeve*
Contribution from the Laboratory of Medicinal Chemistry, Center for Cancer Research,
NCI-Frederick, 376 Boyles Street, Building 376, Room 109, Frederick, Maryland 21702
Received May 9, 2006; E-mail: gildersleevej@ncifcrf.gov
Abstract: Thioglycosides have been employed extensively for the synthesis of complex oligosaccharides,
carbohydrate libraries, and mimetics of O-glycosides. While very useful, aglycon transfer is a problematic
side reaction with thioglycosides. In this paper, a series of mechanistic studies are described. The aglycon
transfer process is shown to affect both armed and disarmed thioglycosides, cause anomerization of the
carbon-sulfur bond of a thioglycoside, and destroy the product of a glycosylation reaction. The results
indicate that the aglycon transfer process can be a major problem for a wide range of thioglycosides. This
side reaction is especially important to consider when carrying out complex reactions such as solid-phase
glycosylations, one-pot or orthogonal multicomponent glycosylations, and construction of carbohydrate
libraries. To prevent transfer, a number of modified aglycons were examined. The 2,6-dimethylphenyl (DMP)
aglycon was found to effectively block transfer in a variety of model studies and glycosylation reactions.
The DMP group can be installed in one step from a commercially available thiol (2,6-dimethylthiophenol)
and is useable as a glycosyl donor. On the basis of these features, the DMP group is proposed as a
convenient and improved aglycon for thioglycosides.
Introduction
precursor. Not surprisingly, thioglycosides have been used
frequently for the synthesis of glycosylated natural products and
Oligosaccharides play important roles in many biological
processes such as inflammation, bacterial and viral infection,
and protein folding. In addition, carbohydrates are critical
components of many natural products and drugs such as
erythromycin, vancomycin, and doxorubicin. Unfortunately,
carbohydrates and glycosylated natural products can be par-
ticularly difficult to obtain from natural sources, especially in
homogeneous form. Therefore, chemical synthesis is an impor-
tant tool for gaining access to these compounds. Furthermore,
synthesis allows one to obtain unnatural derivatives which can
be useful for studying relationships between structure and
function and improving the activity of oligosaccharides and
glycosylated drugs. However, synthesis of carbohydrates re-
mains a challenging area of organic chemistry.
Alkyl and aryl thioglycosides are extremely versatile and
convenient derivatives for carbohydrate synthesis.1,2 The sulfide
group is easy to install and stable to a wide range of reaction
conditions. As a result, it can be incorporated at an early stage
and carried through many steps of a synthesis, a feature that
permits highly convergent approaches to the synthesis of
complex oligosaccharides (Scheme 1). Once a suitable building
block has been made, thioglycosides can be readily activated
as glycosyl donors. Alternatively, they can be converted into
other types of glycosyl donors such as glycosyl halides,3
sulfoxides,4 and sulfones.5 Therefore, one can test a variety of
glycosylation methods and conditions via a common synthetic
biologically interesting oligosaccharides,1,2 solid-phase synthe-
sis of oligosaccharides,6,7 the construction of carbohydrate
libraries,8, 9 and programmable and orthogonal one-pot glycosyla-
tions.10-12 Thioglycosides are also used often as mimetics
of O-glycosides for the development of glycosyltransferase
inhibitors, ligands for lectins, and carbohydrate-based vac-
cines.13,14
(3) For some examples and key references, see: (a) Nicolaou, K. C.; Dolle,
R. E.; Papahatjis, D. P.; Randall, J. L. J. Am. Chem. Soc. 1984, 106, 4189-
4192. (b) Kartha, K. P. R.; Field, R. A. Tetrahedron Lett. 1997, 38, 8233-
8236. (c) Sugiyama, S.; Diakur, J. M. Org. Lett. 2000, 2, 2713-2715.
(4) For some examples and key references, see: (a) Kahne, D.; Walker, S.;
Cheng, Y.; Engen, D. J. Am. Chem. Soc. 1989, 111, 6881-6882. (b)
Kakarla, R.; Dulina, R. G.; Hatzenbuhler, N. T.; Hui, Y. W.; Sofia, M. J.
J. Org. Chem. 1996, 61, 8347-8349. (c) Chen, M. Y.; Patkar, L. N.; Chen,
H. T.; Lin, C. C. Carbohydr. Res. 2003, 338, 1327-1332. (d) Chen, M.
Y.; Patkar, L. N.; Lin, C. C. J. Org. Chem. 2004, 69, 2884-2887. (e)
Agnihotri, G.; Misra, A. K. Tetrahedron Lett. 2005, 46, 8113-8116.
(5) For examples, see: (a) Chang, G. X.; Lowary, T. L. Org. Lett. 2000, 2,
1505-1508. (b) Brown, D. S.; Ley, S. V.; Vile, S.; Thompson, M.
Tetrahedron 1991, 47, 1329-1342.
(6) For some recent examples employing thioglycoside acceptors, see: (a) Yan,
L.; Taylor, C. M.; Goodnow, R.; Kahne, D. J. Am. Chem. Soc. 1994, 116,
6953-6954. (b) Rademann, J.; Geyer, A.; Schmidt, R. R. Angew. Chem.,
Int. Ed. Engl. 1998, 37, 1241-1245. (c) Ferguson, J.; Marzabadi, C.
Tetrahedron Lett. 2003, 44, 3573-3577.
(7) Seeberger, P. H.; Haase, W. C. Chem. ReV. 2000, 100, 4349-4393.
(8) For some recent examples, see: (a) Zhu, T.; Boons, G. J. Angew. Chem.,
Int. Ed. Engl. 1998, 37, 1898-1900. (b) Sofia, M. J. et al. J. Med. Chem.
1999, 42, 3193-3198. (c) Ye, X. S.; Wong, C. H. J. Org. Chem. 2000, 65,
2410-2431.
(9) Liang, R.; Yan, L.; Loebach, J.; Ge, M.; Uozumi, Y.; Sekanina, K.; Horan,
N.; Gildersleeve, J.; Thompson, C.; Smith, A.; Biswas, K.; Still, W. C.;
Kahne, D. Science 1996, 274, 1520-1522.
(10) Koeller, K. M.; Wong, C. H. Chem. ReV. 2000, 100, 4465-4493.
(11) Codee, J. D. C.; Litjens, R.; van den Bos, L. J.; Overkleeft, H. S.; van der
Marel, G. A. Chem. Soc. ReV. 2005, 34, 769-782.
(1) Toshima, K.; Tatsuta, K. Chem. ReV. 1993, 93, 1503-1531.
(2) Garegg, P. J. Thioglycosides as glycosyl donors in oligosaccharide synthesis.
In AdVances In Carbohydrate Chemistry And Biochemistry; Academic
Press: New York, 1997; Vol. 52, pp 179-205.
(12) Wang, Y. H.; Zhang, L. H.; Ye, X. S. Comb. Chem. High-Throughput
Screen. 2006, 9, 63-75.
9
11612
J. AM. CHEM. SOC. 2006, 128, 11612-11619
10.1021/ja063247q Not subject to U.S. copyright. Publ. 2006 Am. Chem. Soc.

Aglycon Transfer of Thioglycosides
A R T I C L E S
Scheme 1. Thioglycosides Are Highly Versatile Carbohydrate Derivatives
While thioglycosides possess many beneficial features, ag-
lycon transfer can be a problematic side reaction. When a
glycosyl acceptor possessing a thioglycoside aglycon is reacted
with a glycosyl donor, the sulfide group can be transferred from
the acceptor to the donor (Scheme 1). Aglycon transfer of
thioglycosides has been reported in a number of papers over
the last 15 years.15-28 It can occur with a variety of glycosyl
donors using a range of reaction conditions. Unfortunately, the
factors that determine whether aglycon transfer will occur for
a particular donor/acceptor pair are not well understood, and
even small changes to protecting groups or reaction conditions
can have a dramatic effect on the outcome of a reaction.
Therefore, aglycon transfer has been exceedingly difficult to
predict even for experienced carbohydrate chemists. Moreover,
the limited understanding of this process has also made it
difficult to develop convenient, reliable, and general strategies
to block aglycon transfer. Given the prevalence of thioglycosides
in carbohydrate chemistry, more detailed information on the
mechanism of transfer and factors that affect transfer are
necessary.
construction of carbohydrate libraries, and one-pot glycosylation
reactions involving thioglycosides. Finally, a simple and efficient
strategy for avoiding transfer was developed and utilized to
overcome a problematic glycosylation required for the synthesis
of a GalNAcR1-3Gal-linked disaccharide found on human
mucins.
Results and Discussion
Our group has been developing a carbohydrate microarray
as a new tool for basic and translational cancer research.29,30
Carbohydrate microarrays contain many different carbohydrate
structures immobilized on a solid support in a miniaturized
fashion. The unique format is designed for high-throughput
evaluation of carbohydrate-macromolecule interactions with
a minimal amount of sample. One of the key challenges for the
development of a carbohydrate array is obtaining a large set of
structurally defined, homogeneous glycans for the array. Since
the vast majority of carbohydrates are not readily accessible,
our group relies heavily on organic synthesis to acquire them.
Like many other carbohydrate chemists, we employed
thioglycosides as key intermediates for the syntheses of glycans.
During the construction of a series of structurally related human
disaccharide glycans, aglycon transfer became a serious problem.
The key step for the synthesis of each disaccharide was
formation of the glycosidic linkages (Scheme 2). Coupling of
donor 1 with thioglycoside 2 and donor 4 with thioglycoside 5
produced disaccharides 3 and 6 in high yield. However, the
related glycosylation between chloride 7 and thioglycoside 8
to form the glycosidic linkage of GalNAcR1-3Gal derivative
9 was completely unsuccessful (Scheme 2). Sulfide 10 was
found to be the major side product in the reaction resulting from
transfer of the thiophenyl aglycon from acceptor 8 to donor 7.
Equivalent results were obtained with imidate 4.
In this paper, a series of experiments are described which
provide new insights into the transfer process. On the basis of
our studies, aglycon transfer has the potential to be a problem
for nearly any thioglycoside. Consideration of aglycon transfer
is especially important for solid-phase oligosaccharide synthesis,
(13) For some recent reviews, see: (a) Driguez, H. Thiooligosaccharides in
glycobiology. In Glycoscience Synthesis Of Substrate Analogs And Mi-
metics; Springer Verlag: Berlin, 1997; Vol. 187, pp 85-116. (b) Witczak,
Z. J. Curr. Med. Chem. 1999, 6, 165-178. (c) Pachamuthu, K.; Schmidt,
R. R. Chem. ReV. 2006, 106, 160-187.
(14) For some recent examples, see: (a) Knapp, S.; Myers, D. S. J. Org. Chem.
2002, 67, 2995-2999. (b) Bundle, D. R.; Rich, J. R.; Jacques, S.; Yu, H.
N.; Nitz, M.; Ling, C. C. Angew. Chem., Int. Ed. Engl. 2005, 44, 7725-
7729. (c) Thayer, D. A.; Yu, H. N.; Galan, M. C.; Wong, C. H. Angew.
Chem., Int. Ed. Engl. 2005, 44, 4596-4599.
(15) Kihlberg, J.; Eichler, E.; Bundle, D. R. Carbohydr. Res. 1991, 211, 59-
75.
The results raised several important questions. First, how did
aglycon transfer occur? Second, why did it occur in one reaction
and not the others? Finally, how could transfer be avoided in
order to complete the synthesis of the mucin glycan?
(16) Knapp, S.; Nandan, S. R. J. Org. Chem. 1994, 59, 281-283.
(17) Leigh, D. A.; Smart, J. P.; Truscello, A. M. Carbohydr. Res. 1995, 276,
417-424.
(18) Belot, F.; Jacquinet, J. C. Carbohydr. Res. 1996, 290, 79-86.
(19) Du, Y. G.; Lin, J. H.; Linhardt, R. J. J. Carbohydr. Chem. 1997, 16, 1327-
1344.
(20) Yu, H.; Yu, B.; Wu, X. Y.; Hui, Y. Z.; Han, X. W. J. Chem. Soc., Perkin
Trans. 1 2000, 9, 1445-1453.
Mechanism of Transfer and Factors Affecting the Out-
come. Aglycon transfer is generally considered to proceed via
activation of the glycosyl donor31 followed by glycosylation of
the sulfur atom of the thioglycoside to form sulfonium ion 11
(Scheme 3). The sulfonium ion can break down in several ways.
(21) Zhu, T.; Boons, G. J. Carbohydr. Res. 2000, 329, 709-715.
(22) Sherman, A. A.; Yudina, O. N.; Mironov, Y. V.; Sukhova, E. V.; Shashkov,
A. S.; Menshov, V. M.; Nifantiev, N. E. Carbohydr. Res. 2001, 336, 13-
46.
(23) Cheshev, P. E.; Kononov, L. O.; Tsvetkov, Y. E.; Shashkov, A. S.;
Nifantiev, N. E. Russ. J. Bioorg. Chem. 2002, 28, 419-429.
(24) Geurtsen, R.; Boons, G. J. Tetrahedron Lett. 2002, 43, 9429-9431.
(25) Tanaka, H.; Adachi, M.; Takahashi, T. Tetrahedron Lett. 2004, 45, 1433-
1436.
(29) Manimala, J.; Li, Z.; Jain, A.; VedBrat, S.; Gildersleeve, J. C. ChemBio-
Chem 2005, 6, 2229-2241.
(26) Xue, J.; Khaja, S. D.; Locke, R. D.; Matta, K. L. Synlett 2004, 861-865.
(27) Codee, J. D. C.; Stubba, B.; Schiattarella, M.; Overkleeft, H. S.; van
Boeckel, C. A. A.; van Boom, J. H.; van der Marel, G. A. J. Am. Chem.
Soc. 2005, 127, 3767-3773.
(30) Manimala, J. C.; Roach, T. A.; Li, Z.; Gildersleeve, J. C. Angew. Chem.,
Int. Ed. Engl. 2006, 45, 3607-3610.
(31) For simplicity, the activated species has been represented as an oxocarbe-
nium ion. However, other activated intermediates such as glycosyl triflates
and dioxolenium ions may be present as well.
(28) Sun, J. S.; Han, X. W.; Yu, B. Org. Lett. 2005, 7, 1935-1938.
9
J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006 11613

A R T I C L E S
Li and Gildersleeve
Scheme 2. Synthesis of Mucin Glycans and Aglycon Transfer
Scheme 3. Mechanism of Aglycon Transfer
Cleavage of the bond between the sulfur atom and the anomeric
carbon of the glycosyl donor (arrow a) reforms the original
thioglycoside and activated donor (initial reactants). Alterna-
tively, cleavage of the bond between the sulfur atom and the
anomeric carbon of the acceptor (arrow b) results in transfer
of the thiophenyl aglycon group to the glycosyl donor along
with formation of the oxocarbenium ion (or other activated
intermediate) derived from the thioglycoside. In prin-
ciple, polymerization can occur if the activated intermediate
derived from the acceptor glycosylates another acceptor mol-
ecule.
On the basis of the mechanistic picture, a number of factors
could play a role in determining whether transfer occurs.32 Since
the sulfide and hydroxyl compete for the activated intermediate,
9
11614 J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006

Aglycon Transfer of Thioglycosides
A R T I C L E S
Scheme 4. Model Studies on Transfer with Armed vs Disarmed Thioglycosides
one key factor is the relative reactivity of the sulfur atom and
oxygen atom. The nucleophilicity of the hydroxyl can be
affected by its stereochemistry (axial vs equatorial), its position
on the ring, the size and electronic nature of nearby protecting
groups, and hydrogen bonding. The nucleophilicity of the sulfide
will be affected by the steric and electronic effects of the
aglycon, the stereochemistry of the bond between the anomeric
carbon and sulfur atom (axial vs equatorial), and the nearby
protecting groups. The reaction conditions can also affect the
competition between the sulfide and hydroxyl. For example,
kinetic selectivity favoring the better nucleophile may be eroded
at higher temperatures. Thus, factors that affect the activation
temperature such as the solvent, activating agent, and type of
glycosyl donor may affect transfer.
A second factor that can determine whether aglycon transfer
will occur in a particular reaction is the electronic nature of the
donor and acceptor. The electronic nature of a carbohydrate
derivative is frequently described as “armed” or “disarmed”.33,34
These terms refer to the ease or difficulty of activating a sugar
as a glycosyl donor. Disarmed sugars are harder to activate;
they generally contain functionality that would destabilize an
oxocarbenium ion such as electron-withdrawing groups (e.g.,
esters and azides) and/or higher degrees of oxygenation.35
Armed sugars are easier to activate; they typically contain
functionality that is less destabilizing such as ether protecting
groups on the hydroxyls (e.g., O-benzyl, O-allyl, and O-silyl
groups) and/or deoxygenated positions on the ring. All previ-
ously reported cases of aglycon transfer have involved armed
thioglycoside acceptors. In fact, Yu et al. directly compared
several armed and disarmed thioglycosides and found that only
the armed thioglycosides suffered from transfer.20 The authors
rationalized this observation in terms of relative stabilities of
oxocarbenium ions; the reaction is driven to produce the more
stable oxocarbenium ion. On the basis of this hypothesis, any
thioglycoside could undergo transfer as long as the oxocarbe-
nium ion derived from the thioglycoside is more stable than
the oxocarbenium ion derived from the glycosyl donor. It is
important to note that activation of a glycosyl donor can produce
a number of other activated intermediates such as glycosyl
triflates and dioxolenium ions. Therefore, a more general
phrasing of this theory is that the reaction is driven to the more
stable set of activated intermediates.
The lack of observed transfer with disarmed thioglycosides
could also be explained via a kinetic argument: cleavage of
the C-S bond for disarmed thioglycosides may be too slow
under the reaction conditions. On the basis of this rationale,
highly disarmed thioglycosides would not be expected to transfer
regardless of the armed/disarmed nature of the donor. If true,
one could avoid transfer by using disarmed thioglycosides.
However, it would be important to know how deactivated a
thioglycoside must be to avoid transfer.
(32) While this paper focuses on other factors, it should be noted that the relative
rates of decomposition of the activated intermediates and glycosylation may
also affect the outcome of a reaction.
(33) Mootoo, D. R.; Konradsson, P.; Udodong, U.; Fraser-Reid, B. J. Am. Chem.
Soc. 1988, 110, 5583-5584.
To distinguish between kinetic and thermodynamic explana-
tions and probe the scope of the aglycon transfer problem, a
series of model glycosylation reactions were carried out. Various
armed and disarmed donors were activated in the presence of
highly armed, highly disarmed, and intermediate thioglycosides
(34) Fraser-Reid, B.; Wu, Z. F.; Udodong, U. E.; Ottosson, H. J. Org. Chem.
1990, 55, 6068-6070.
(35) For a nice paper illustrating the effects of various protecting groups on the
relative reactivities of glycosyl sulfides, see: Zhang, Z.; Ollmann, I. R.;
Ye, X.-S.; Wischnat, R.; Baasov, T.; Wong, C.-H. J. Am. Chem. Soc. 1999,
121, 734-753.
9
J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006 11615

A R T I C L E S
Li and Gildersleeve
(Scheme 4). To simplify analysis of the products and eliminate
competition with the hydroxyl, fully protected thioglycosides
were utilized. First, the disarmed imidate 436 (three disarming
ester protecting groups and one disarming azide group) was
activated in the presence of armed thioglycoside 12.37 As
expected from both theories, transfer product 1038 was isolated
in 100% yield from the reaction mixture.39 Next, the disarmed
imidate 1340 was reacted with thioglycoside 14 of intermediate
reactivity (i.e., two disarming ester protecting groups and two
ethers). Again, the transfer product (15) was isolated in 100%
yield. Finally, disarmed imidate 4 was reacted with disarmed
thioglycoside 15.41,42 Since the donor and thioglycoside are both
highly disarmed, one might expect the oxocarbenium ions (or
other intermediates) to be of similar stability. On the basis of a
thermodynamic rationale, one would predict a mixture of transfer
and starting material. In contrast, one would predict no transfer
based on the kinetic argument since the thioglycoside is highly
disarmed. In actuality, the transfer product (10, 35%) was
isolated from the reaction along with some recovered starting
material (40%). Therefore, the C-S bond of highly disarmed
thioglycosides can be cleaved under standard reaction conditions.
While transfer is more likely to occur with armed thioglycosides,
it has the potential to occur with any thioglycoside.
One additional reaction was investigated to further probe the
mechanism and scope of the aglycon transfer process. The
mechanistic pathway described in Scheme 3 is a reversible
process. If the carbon-sulfur bond of the thioglycoside can
break and reform, the transfer process could provide a pathway
to equilibrate the anomeric center of the thioglycoside. More-
over, equilibration could potentially occur even when formation
of the transfer product is not energetically favored (e.g., the
thioglycoside is more disarmed than the glycosyl donor). To
investigate this possibility, the intermediate thioglycoside 14
was reacted with a more armed donor, 1643 (reaction D, Scheme
4). As expected, no transfer product was isolated from the
reaction. Interestingly, the thioglycoside (14) was recovered as
an alpha/beta mixture, indicating that anomerization had oc-
curred. Thus, even when transfer is not energetically favored,
the process can still be detrimental in a reaction by producing
a mixture of products.
On the basis of the mechanism, glycosylation on the sulfur
atom of the product of a glycosylation reaction should also be
a viable pathway in reactions. Therefore, excess equivalents of
donor could destroy the product of a glycosylation reaction or
cause anomerization of the carbon-sulfur bond. With excess
equivalents of donor, destruction could become a problem even
when glycosylation of the hydroxyl is much faster than
glycosylation of the sulfur atom. To examine this possibility,
acceptor 5 was reacted with either 1 or 2 equivalents of donor
4 (Scheme 2). The free hydroxyl of the acceptor is a primary
alcohol and should react rapidly with activated donor. However,
the donor is similar or slightly more disarmed than the acceptor.
Therefore, transfer should be energetically favorable. With 1
equivalent, the reaction proceeded smoothly to form disaccharide
6 in 76% yield. However, with 2 equivalents of donor, only
25% of disaccharide 6 was isolated. Moreover, aglycon transfer
product 10 was obtained in 60% yield. Thus, destruction of
product by excess equivalents of donor leads to a much lower
yield. While excess equivalents can drive reactions to comple-
tion, they may drive the wrong reaction to completion.
Preventing Aglycon Transfer. Taken together, the above
experiments indicate that aglycon transfer can be a major
problem with a wide range of thioglycosides. One obvious
strategy to avoid this problem is to remove the thioglycoside
functional group. With this approach, however, one loses all
the advantageous features of thioglycosides. Therefore, synthetic
strategies that minimize side reactions while retaining the
advantageous features of thioglycosides would be best.
One approach to avoid transfer involves modifying the
aglycon. Transfer occurs via nucleophilic attack of the sulfur
atom on the activated glycosyl donor. Therefore, strategies to
reduce the reactivity of the sulfur could minimize transfer. This
approach has been examined by other groups, but a general and
convenient alternative has not yet emerged.19,23,24,27 One of the
key challenges is reducing the reactivity of the sulfur atom
sufficiently to prevent transfer but not so drastically that the
thiol group becomes difficult to install and/or utilize at a later
stage in a synthesis. For example, thioglycosides with a para-
nitrophenyl aglycon have substantially reduced nucleophilicity
but cannot be activated as glycosyl donors even under very
strong activating conditions (e.g., NIS/TfOH).44 Therefore, one
must find a balance between reduced nucleophilicity and
adequate reactivity. Ideally, one would also like a starting thiol
that is commercially available, inexpensive, and nontoxic.
To identify a suitable aglycon group, a series of model
reactions were carried out. The first model reaction involved
activating disarmed donor 4 in the presence of disarmed
thioglycosides 15a-m (Table 1). The reaction of the parent
thiophenyl aglycon produces a 1:1 mixture of transfer product
and starting material. Therefore, the model system would be
sensitive to even modest changes in yield of transfer product.
In addition, the thioglycoside substrates for this model reaction
required only a single synthetic step which allowed us to readily
evaluate an assortment of aglycons. Initially, we focused on
systematically varying the substituents on the phenyl ring of
thiophenol. Results are summarized in Table 1. In agreement
with the mechanistic analysis, electron-donating groups in-
creased the amount of transfer while electron-withdrawing
groups decreased the yield of transfer product. In fact, many of
Product Destruction. Thioglycoside aglycons are frequently
used for both solid-phase synthesis of oligosaccharides and
construction of libraries of carbohydrates. When synthesizing
compounds on a solid support or synthesizing large collections
of compounds, product purification can be extremely difficult
or completely impractical. Moreover, monitoring reaction
progress can be equally challenging. Therefore, obtaining a high
yield is critical. To achieve this, chemists frequently use a large
excess of reagents and starting materials in order to drive
reactions to completion. In addition, glycosylation reactions are
repeated multiple times. Increasing the equivalents of glycosyl
donor is also a common strategy for improving yields of
standard solution-phase glycosylations.
(36) Grundler, G.; Schmidt, R. R. Liebigs Ann. Chem. 1984, 1826-1847.
(37) Garegg, P. J.; Hultberg, H.; Lindberg, C. Carbohydr. Res. 1980, 83, 157-
162.
(38) Andreotti, A. H.; Kahne, D. J. Am. Chem. Soc. 1993, 115, 3352-3353.
(39) In the absence of the glycosyl donor (imidate 4), the thioglycoside is stable
to the reaction conditions.
(40) Schmidt, R. R.; Stumpp, M. Liebigs Ann. Chem. 1983, 1249-1256.
(41) Ferrier, R. J.; Furneaux, R. H. Carbohydr. Res. 1976, 52, 63-68.
(42) Khiar, N.; Martin-Lomas, M. J. Org. Chem. 1995, 60, 7017-7021.
(43) Schmidt, R. R.; Stumpp, M. Liebigs Ann. Chem. 1984, 680-691.
(44) Cao, S. D.; Gan, Z. H.; Roy, R. Carbohydr. Res. 1999, 318, 75-81.
9
11616 J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006

Aglycon Transfer of Thioglycosides
A R T I C L E S
Table 1. Effects of Modified Aglycons on Transfer in Model System 1 and on the Activation Temperature of the Thioglycoside
entry
aglycon R group
recovered starting material, %
transfer product, %
thioglycoside activation temp., °
Ca
1
2
3
4
5
6
7
8
15a ) Ph
40
50
90
100
50
100
100
50
100
30
100
100
100
80
35b
40
10
-45
15b ) 2-FPh
15c ) 2-ClPh
15d ) 2-(CF3)Ph
15e ) 3-FPh
none
>15
20b
15f ) 3-ClPh
none
none
44
none
70
none
none
none
16
-45
-50
15g ) 3-BrPh
15h ) 3-MePh
9
15i ) 3-(CF3)Ph
15j ) 4-(OMe)Ph
15k ) 4-(CF3)Ph
15l ) 3,5-Cl2Ph
0
10
11
12
13
14
15
16
17
18
19
20
0
>15
0
15m ) 3,5-(CF3)2Ph
15n ) 1-naphthyl
15o ) 2-naphthyl
15p ) adamantanyl (R)
15q ) adamantanyl (â)
15r ) 2,6-Me2Ph (â DMP)
15s ) 2,6-Me2Ph (R DMP)
15t ) 2,6-Cl2Ph (R)
60
40
100
100
100
100
100
none
none
none
none
none
-40
-50
-15
-50
-15
a In a separate experiment, compounds 15a-t were cooled to -60 °C and NIS/TMSOTf was added. The reaction was slowly warmed while monitoring
by TLC to determine the activation temperature. b A number of minor components were also observed by TLC.
Table 2. Preventing Aglycon Transfer in a More Challenging
Model System
the modified aglycons with electron-withdrawing groups such
as 3-chlorophenyl, 3-trifluoromethyl, and 3,5-dichlorophenyl
completely blocked transfer in this first model system. Modified
aglycons performing well in model system 1 were also tested
for activation using NIS/TMSOTf. Thioglycosides were treated
with NIS and TMSOTf at -60 °C. The reaction mixture was
monitored while slowly warming, and the temperature at which
activation occurred was measured (Table 1).
Next, the best candidates were examined in a second, more
challenging aglycon transfer system. Donor 4 was activated in
the presence of perbenzylated thioglycosides 12a, 12f, 12i, and
12l. In this model system, the thioglycosides are highly armed
and the donor is highly disarmed. Therefore, this represents an
entry
aglycon R group
recovered starting material, % transfer product, %
extreme mismatch and a highly challenging system for blocking
transfer. For reference, reaction of the parent thiophenyl
glycoside results in 100% transfer. The results are summarized
in Table 2. None of the aglycons with electron-withdrawing
groups were sufficiently deactivated to block transfer in this
very difficult reaction. Moreover, many of the thioglycosides
with modified aglycons required significantly higher tempera-
tures to activate using NIH/TMSOTf. For these reasons, phenyl
rings with electron-withdrawing substituents did not appear to
be a viable, general solution for the aglycon transfer problem.
Glycosylation of the sulfur atom produces a sterically
congested sulfonium ion (structure 11, Scheme 3). Therefore,
it was anticipated that bulky aglycons might also block transfer,
even in the absence of electron-withdrawing groups. In addition,
one might be able to overcome difficulties associated with
activation by using small activating agents. In support of this
approach, the Boons group recently reported the use of a
dicyclohexylmethyl thio aglycon to block transfer.24 We incor-
porated this aglycon into the perbenzylated galactose model
system and evaluated its ability to prevent transfer (12u, Table
1
2
3
4
5
6
7
8
12a ) Ph
none
100
77
67
75
50
none
none
15
12f ) 3-ClPh
10 (R)
20 (R)
10 (R)
30 (1:1 R:â)
81 (â)
80
12i ) 3-(CF3)Ph
12l ) 3,5-Cl2Ph
12q ) adamantyl (â)
12r ) 2,6-Me2Ph (â DMP)
12s ) 2,6-Me2Ph (R DMP)
12u ) CH(cHexyl)2
50
2). While performing better than most of the modified phenyl
aglycons, transfer still occurred. In addition, the corresponding
thiol starting material was not commercially available. To
identify a more optimal and convenient aglycon, a variety of
commercially available thiols were considered. The sterically
hindered adamantyl, 1-naphthyl, 2-naphthyl, and 2,6-dimeth-
ylphenyl thiols were chosen for further evaluation.45 Each was
incorporated into galactose, evaluated in both model systems,
and tested for activation (Tables 1 and 2). Happily, the 2,6-
(45) 2,6-Dimethylthiophenol was purchased from Oakwood Products, Inc. (West
Columbia, SC) and is available from Aldrich (St. Louis, MO).
9
J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006 11617

A R T I C L E S
Li and Gildersleeve
Scheme 5. DMP Aglycon Blocks Transfera
a Reaction conditions: (a) AgOTf, di-tert-butylmethylpyridine, CH2Cl2, -60 to 0 ˚C (78% yield); (b) TMSOTf, diethylether, -20 to 0 ˚C (82% yield).
dimethylphenyl (DMP) aglycon completely blocked transfer in
both model systems (12r, 12s, 15r, and 15s). In addition, the
DMP thioglycoside could be activated under relatively mild
conditions. The 2,6-dichlorophenyl aglycon was also evaluated
but found to be suboptimal.46 On the basis of these model
studies, the DMP group satisfied the initial screening criteria
and appeared to be an excellent candidate as an alternative
aglycon for thioglycosides.
Next, the DMP aglycon was tested in several problematic
glycosylation reactions. First, the DMP aglycon was incorporate
into galactose derivative 18 and coupled with donor 1747 in an
effort to synthesize the corresponding alpha 1-3 linked disac-
charide (reaction A, Scheme 5). As mentioned previously, no
product was obtained in the glycosylation when a phenyl group
was used as the aglycon. With the DMP aglycon, however, a
78% yield of the desired disaccharide (19) was obtained. It is
important to note the DMP modified disaccharide could be
successfully transformed into the final target glycan via the same
reaction sequence and conditions previously used to convert the
thiophenyl-containing disaccharides 3 and 6 into the related
targets (see Supporting Information and ref 15). In a second
test case, the DMP aglycon was incorporated into galactose
derivative 20 and coupled with varying equivalents of donor 4
to determine if the modified aglycon could block product
destruction (reaction B, Scheme 5). Even with two equivalents
of donor, the disaccharide (21) could be obtained in 82% yield,
indicating that the modified aglycon effectively prevents transfer.
sulfoxides, and imidates present on the glycosyl donor and/or
acceptor.48-51 Glycosylation of the sulfur atom of a thioglycoside
can also occur, and this process leads to transfer of the aglycon
from the thioglycoside to the glycosyl donor. In this paper, it is
shown that (a) the transfer process can occur for a wide range
of thioglycosides, (b) even if the transfer product is not observed,
the process can lead to anomerization of the C-S bond of the
thioglycoside, and (c) excess equivalents of glycosyl donor can
lead to destruction of the product of a glycosylation reaction.
Taken together, the results show that the aglycon transfer process
can affect a wide range of thioglycosides. This side reaction is
especially important to consider when carrying out complex
reactions such as solid-phase glycosylations, one-pot or or-
thogonal multicomponent glycosylations, and construction of
carbohydrate libraries. Finally, the sterically hindered 2,6-
dimethylphenyl (DMP) aglycon was found to block glycosy-
lation of the sulfur atom and prevent aglycon transfer in a range
of model reactions and glycosylation reactions. The DMP group
can be easily installed using a commercially available thiol (2,6-
dimethylthiophenol) and utilized as a glycosyl donor at a later
stage in a synthesis. On the basis of the results, the DMP group
is a convenient and improved aglycon for thioglycosides.
Experimental Section
General Methods. Unless otherwise stated, reagents were obtained
from commercial suppliers and used without purification. NMR spectra
were recorded on a Unity Inova 400 Fourier transform NMR spec-
trometer. All proton NMR data was obtained at 400 MHz, and all carbon
NMR data was obtained at 100 MHz. Proton chemical shifts are
reported in parts per million (ppm) downfield from tetramethylsilane
(TMS) unless otherwise noted. Carbon chemical shifts are reported in
parts per million (ppm) downfield from TMS using CDCl3 as an internal
reference unless otherwise noted. Coupling constants (J) are reported
in hertz (Hz). Multiplicities are abbreviated as follows: singlet (s),
doublet (d), triplet (t), quartet (q), quintuplet (p), multiplet (m), and
broadened (br). High-resolution mass spectra were obtained on a VG
ZAB (University of California, Riverside, Mass Spectrometry Facility).
2,6-Dimethylthiophenol was purchased from Oakwood Products, Inc.
(West Columbia, SC)
Conclusions
Glycosylation reactions are one of the most challenging and
unpredictable chemical transformations known. Due to both
steric and electronic effects, hydroxyls found on carbohydrates
are especially poor nucleophiles. As a result, most glycosylation
methods are designed to produce extremely reactive intermedi-
ates such as glycosyl triflates and oxocarbenium ions. While
necessary for glycosylations, the high reactivity can lead to a
variety of side reactions. A better understanding of these
processes and methods to prevent them are critical for advancing
the field.
General Procedure for Aglycon Transfer Model Reactions.
Thioglycosides (0.1 mmol, 1 equiv) and glycosyl imidates (0.1 mmol,
One side reaction that has emerged as a general problem is
glycosylation of other nucleophilic groups such as amides,
(48) Schmidt, R. R.; Toepfer, A. Tetrahedron Lett. 1991, 32, 3353-3356.
(49) Gildersleeve, J.; Pascal, R. A.; Kahne, D. J. Am. Chem. Soc. 1998, 120,
5961-5969.
(50) Liao, L.; Auzanneau, F. I. Org. Lett. 2003, 5, 2607-2610.
(51) Liao, L.; Auzanneau, F. I. J. Org. Chem. 2005, 70, 6265-6273.
(46) The 2,6-dichlorophenyl aglycon was harder to install, harder to activate,
and generally produced lower yields.
(47) Lemieux, R. U.; Ratcliffe, R. M. Can. J. Chem. 1979, 57, 1244-1251.
9
11618 J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006

Aglycon Transfer of Thioglycosides
A R T I C L E S
1 equiv) were dissolved in DCM (1 mL) and stirred with molecular
sieves at room temperature for 0.5 h. The reaction mixture was cooled
to -78 °C (dry ice/acetone bath) for 10 min and then warmed to -60
°C. TMSOTf (2 drops) was added via syringe at -60 °C, and the
reaction mixture was slowly warmed to 0 °C over 1.5 h. The
temperature was monitored externally. When the temperature reached
0 °C, the reaction mixture was quenched by addition of triethylamine
(0.1 mL), loaded directly onto a silica column, and purified by
chromatography.
2,6-Dimethylphenyl 2,3,4,6-Tetra-O-acetyl-1-thio-D-galactopyra-
noside (15r, 15s). BF3EtO2 (1 mL, 1.3 equiv) was added to a mixture
of galactose pentaacetate (2 g, 1 equiv) and 2,6-dimethylthiophenol45
(700 mg, 1 equiv) in CH2Cl2 and toluene (1:1, 8 mL). The reaction
mixture was stirred at 50 °C for 4 h and diluted with CH2Cl2 (10 mL).
The CH2Cl2 solution was washed with saturated NaHCO3 twice
followed by brine. Solvent was evaporated, and the product was purified
by chromatography (1:3 ethyl acetate/hexanes).
â product 15r (white powder, 1.4 g, 56% yield): Rf ) 0.3 (1:2 ethyl
acetate/hexanes); 1H NMR (CDCl3) δ 7.19-7.11 (m, 3 H), 5.38 (dd, J
) 3.2, 0.8 Hz, 1 H), 5.33 (t, J ) 10.0 Hz, 1 H), 5.01 (dd, J ) 10.0, 3.2
Hz, 1 H), 4.41 (d, J ) 10.0 Hz, 1 H), 4.10 (dd, J ) 11.2, 7.2 Hz, 1 H),
4.04 (dd, J ) 11.2, 7.2 Hz, 1 H), 3.74 (m, 1 H), 2.54 (s, 6 H), 2.18 (s,
3 H), 2.14 (s, 3 H), 1.99 (s, 3 H), 1.97 (s, 3 H); 13C NMR (CDCl3) δ
170.2, 170.0, 169.5, 144.0, 131.2, 129.3, 128.2, 89.2, 74.1, 71.9, 67.7,
67.2, 61.5, 22.3, 20.7, 20.6, 20.51, 20.49. HRMS [M + Na]+: calcd
for C22H28O9NaS 491.1352; found 491.1369.
1H NMR (CDCl3) δ 8.11-8.09 (m, 2 H), 7.60-7.56 (m, 1 H), 7.47-
7.43(m, 2 H), 7.13-7.04 (m, 3 H), 5.40 (dd, J ) 10.0, 7.2 Hz, 1 H),
4.43 (d, J ) 10.0 Hz, 1 H), 4.30 (dd, J ) 7.2, 5.6 Hz, 1 H), 4.23 (dd,
J ) 5.2, 2.0 Hz, 1 H), 3.76 (m, 1 H), 3.68-3.64 (m, 2 H), 3.18 (s, 3
H), 2.46 (s, 6 H), 1.65 (s, 3 H), 1.34 (s, 6 H), 1.33 (s, 3 H);13C NMR
(CDCl3) δ 165.4, 144.0, 133.1, 132.0, 129.9, 129.7, 128.9, 128.2, 128.0,
110.5, 100.1, 88.3, 75.5, 73.8, 72.4, 60.3, 48.6, 27.7, 26.2, 24.3, 24.2,
22.5.
The product (200 mg, 0.39 mmol) was dissolved in acetic acid
solution (80%, 3 mL) and heated to 70 °C for 3 h. The solvent was
evaporated, and the residue was dried by azeotropic distillation using
toluene. The residue was suspended in acetonitrile (10 mL) and treated
with benzaldehyde dimethyl acetal (119 mg) and camphorsulfonic acid
(catalytic amount). The mixture was stirred at room temperature for 4
h. The reaction was quenched by triethylamine (1 mL), and the product
was purified by chromatography (1:1 ethyl acetate/hexanes, with 0.1%
triethylamine) to give the desired product (18) as a white powder (130
mg, 68% yield for two steps): Rf ) 0.6 (1:1 ethyl acetate/hexanes);
1H NMR (CDCl3) δ 8.11-8.09 (m, 1 H), 7.60-7.53 (m, 3 H), 7.47-
7.39 (m, 5 H), 7.14-7.06 (m, 3 H), 5.53 (s, 1 H), 5.45 (t, J ) 10.0 Hz,
1 H), 4.59 (d, J ) 10.0 Hz, 1 H), 4.26-4.22 (m, 2 H), 3.99 (dd, J )
12.4, 1.6 Hz, 1 H), 3.85 (m, 1 H), 3.38 (m, 1 H), 2.50 (s, 6 H); 13C
NMR (CDCl3) δ 166.1, 144.3, 137.4, 133.1, 131.2, 129.9, 129.7, 129.3,
129.0, 128.31, 128.28, 128.1, 126.4, 101.4, 87.8, 75.6, 72.9, 91.9, 69.7,
69.2, 22.5. HRMS [M + Na]+: calcd for C28H28O6NaS 515.1504; found
515.1512.
2,6-Dimethylphenyl (3,4,6-Tri-O-acetyl-2-azido-2-deoxyl-R-D-ga-
lactopyranosyl)-(1f3)-4,6-O-benzylidene-2-O-benzoyl-1-thio-â-D-
galactopyranoside (19). A mixture of donor 1747 (70 mg, 1.6 mmol,
2 equiv), acceptor 18 (40 mg, 0.8 mmol, 1 equiv), and di-tert-
butylmethylpyridine (60 mg, 1.6 mmol, 2 equiv) was dried under
vacuum for 1 h and dissolved in CH2Cl2 (1.5 mL). Molecular sieves
(100 mg) were added, and the mixture was stirred and cooled to -78
°C. AgOTf (60 mg, 1.6 mmol, 2 equiv) was added to the reaction
mixture, and the mixture was allowed to warm to 0 °C over 2.5 h. The
reaction was quenched with NaHCO3 (1 mL), and the mixture was
filtered to remove the molecular sieves. The filtrate was extracted with
CH2Cl2 three times. The organic layers were combined and washed
with brine. The solvent was removed, and the residue was purified by
column chromatography to give desired product as a white solid (50
mg, 77% yield): Rf ) 0.5, (1:1 ethyl acetate/hexanes); 1H NMR
(CDCl3) δ 8.10-8.07 (m, 2 H), 7.59-7.54 (m, 3 H), 7.46-7.42 (m, 2
H), 7.39-7.33 (m, 3 H), 7.10-7.01 (m, 3 H), 5.69 (t, J ) 10.0 Hz, 1
H), 5.53 (s, 1 H), 5.16-5.12 (m, 2 H), 5.06 (dd, J ) 2.8, 0.8 Hz, 1 H),
4.61 (d, J ) 10.4 Hz,1 H), 4.37 (d, J ) 3.2 Hz, 1 H), 4.23 (dd, J )
12.4 Hz, 1.2 Hz, 1 H), 4.02-3.97 (m, 3 H), 3.80 (dd, J ) 11.2, 8.0
Hz, 1 H), 3.68 (dd, J ) 11.2, 8.0 Hz, 1 H), 3.50 (dd, J ) 12.4, 3.6 Hz,
1 H), 3.34 (d, J ) 0.8 Hz, 1 H) 2.45 (s, 6 H), 2.01 (s, 3 H), 1.95 (s, 3
H), 1.88 (s, 3 H); 13C NMR (CDCl3) δ 170.0, 169.8, 169.2, 164.9,
144.3, 137.4, 133.3, 131.3, 129.8, 129.5, 129.0, 128.9, 128.5, 128.12,
128.06, 126.2, 101.0, 95.6, 88.2, 77.1, 72.1, 69.6, 69.5, 69.3, 67.5, 67.0,
66.9, 60.7, 59.9, 22.5, 20.6, 20.50, 20.49.
R product 15s (white powder, 390 mg, 15% yield): Rf ) 0.4 (1:2
ethyl acetate/hexanes); 1H NMR (CDCl3) δ 7.20-7.08 (m, 3 H), 5.55
(d, J ) 4.8 Hz, 1 H), 5.53 (m, 1 H), 5.35 (m, 2 H), 4.77 (m, 1 H), 4.08
(m, 2 H), 2.49 (s, 6 H), 2.136 (s, 3 H), 2.135 (s, 3 H), 2.019 (s, 3 H),
2.017 (s, 3 H); 13C NMR (CDCl3) δ 170.2, 170.1, 169.8, 143.2, 130.7,
129.0, 128.4, 87.2, 68.0, 67.93, 67.87, 61.6, 22.0, 20.7, 20.6. HRMS
[M + Na]+: calcd for C22H28O9NaS 491.1352; found 491.1351.
2,6-Dimethylphenyl 4,6-O-Benzylidene-2-O-benzoyl-1-thio-â-D-
galactopyranoside (18). To a solution of compound 15r (1.4 g, 3.0
mmol) in methanol (10 mL) was added NaOMe (catalytic amount).
The reaction mixture was stirred at room temperature for 2 h and
quenched by adding ion-exchange resin. The resin was filtered, and
the solvent was evaporated to give 2,6-dimethylphenyl 1-thio-â-
galactopyranoside (900 mg, 100% yield): 1H NMR (DMSO-d6) δ
7.11-7.04 (m, 3 H), 5.11 (d, J ) 5.6 Hz, 1 H), 4.82 (d, J ) 5.2 Hz,
1 H), 4.45-4.40 (m, 2 H), 4.07 (d, J ) 9.6 Hz, 1 H), 3.61 (m, 1 H),
3.46-3.42 (m, 1 H), 3.36-3.30 (m, 1 H), 3.25-3.20 (m, 2 H), 3.13 (t,
J ) 6.4 Hz, 1 H), 2.45 (s, 6 H); 13C NMR (DMSO-d6) δ 144.2, 132.7,
128.9, 128.2, 91.6, 79.1, 75.1, 70.7, 68.4, 60.6, 22.7.
2,6-Dimethylphenyl thiogalactoside (300 mg, 1.0 mmol) and cam-
phorsulfonic acid (catalytic amount) were dissolved in acetone dimethyl
acetal (25 mL) and stirred at room temperature overnight. The reaction
was quenched by addition of triethylamine (1 mL), and the solvent
was evaporated. The product was purified by chromatography (1:1 ethyl
acetate/hexanes, 0.1% triethylamine) to yield a white solid (240 mg,
58% yield): Rf ) 0.4 (1:1 ethyl acetate/hexanes); 1H NMR (CDCl3) δ
7.11-7.05 (m, 3 H), 4.13 (d, J ) 10.4 Hz, 1 H), 4.10 (dd, J ) 5.6, 2.0
Hz, 1 H), 3.97 (dd, J ) 6.8, 5.2 Hz, 1 H), 3.66-3.64 (m, 1 H), 3.60-
3.55 (m, 3 H), 3.13 (s, 3 H), 2.88 (d, J ) 2.4 Hz, 1 H), 2.54 (s, 6 H),
1.49 (s, 3 H), 1.29 (s, 6 H), 1.27 (s, 3 H); 13C NMR (CDCl3) δ 143.9,
131.2, 129.1, 128.2, 110.0, 100.0, 90.9, 79.0, 75.8, 73.6, 72.8, 60.3,
48.5, 28.1, 26.2, 24.3, 24.2, 22.7.
Acknowledgment. We thank the UC Riverside Mass Spec-
troscopy Facility for high-resolution mass spectral analysis. We
thank Dr. Joseph Barchi, Jr. and Dr. Anthony Stapon for critical
review of this manuscript. This research was supported by the
Intramural Research Program of the NIH, NCI.
Supporting Information Available: Experimental details for
the syntheses of compounds 8, 12f, 12i, 12l, 12q, 12r, 12s, 14,
15b-q, 15t, 20, and 21; 1H and 13C NMR spectra for
compounds 12r, 12s, 14, 15r, 15s, 18, 19, 20, and 21; complete
author list for ref 8b. This material is available free of charge
via the Internet at http://pubs.acs.org.
The product (240 mg, 0.58 mmol) was dissolved in CH2Cl2 (5 mL),
followed by addition of benzoyl chloride (140 mg, 1 mmol), triethy-
lamine (2 mmol), and DMAP (catalytic amount). The solution was
stirred at room temperature for 5 h and diluted by CH2Cl2. The solution
was washed with 1 N HCl, saturated sodium bicarbonate, and brine.
The solvent was removed, and the product was purified by chroma-
tography (1:3 ethyl acetate/hexanes with 0.1% triethylamine) to afford
a white solid (200 mg, 67% yield): Rf ) 0.7 (1:1 ethyl acetate/hexanes);
JA063247Q
9
J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006 11619