C2-hydroxyglycosylation with glycal donors. Probing the mechanism of sulfonium-mediated oxygen transfer to glycal enol ethers

Article abstract of DOI:10.1021/ja025639c

The C2-hydroxyglycosylation reaction employing the reagent combination of a diaryl sulfoxide and triflic anhydride offers a novel method for glycal assembly whereby a hydroxyl functionality is stereoselectively installed at the C2-position of a glycal donor with concomitant glycosylation of a nucleophilic acceptor. Mechanistic investigations into this reaction revealed a novel process for sulfonium-mediated oxidation of glycal enol ethers in which the sulfoxide oxygen atom is stereoselectively transferred to the C2-position of the glycal. ¹⁸O-labeling studies revealed that the S-to-C2 oxygen-transfer process involves initial formation of a C1-O linkage followed by O-migration to C2, leading to the generation of an intermediate glycosyl 1,2-anhydropyranoside that serves as an in situ glycosylating agent. These findings are consistent with the initial formation of a C2-sulfonium-C1-oxosulfonium pyranosyl species upon activation of the glycal donor with Aryl₂SO·Tf₂O.

Full text of DOI:10.1021/ja025639c

Published on Web 06/04/2002
C2-Hydroxyglycosylation with Glycal Donors. Probing the
Mechanism of Sulfonium-Mediated Oxygen Transfer to Glycal
Enol Ethers
Eiji Honda and David Y. Gin*
Contribution from the Department of Chemistry, UniVersity of Illinois at Urbana-Champaign,
Urbana, Illinois 61801
Received January 18, 2002
Abstract: The C2-hydroxyglycosylation reaction employing the reagent combination of a diaryl sulfoxide
and triflic anhydride offers a novel method for glycal assembly whereby a hydroxyl functionality is
stereoselectively installed at the C2-position of a glycal donor with concomitant glycosylation of a nucleophilic
acceptor. Mechanistic investigations into this reaction revealed a novel process for sulfonium-mediated
oxidation of glycal enol ethers in which the sulfoxide oxygen atom is stereoselectively transferred to the
C2-position of the glycal. 18O-labeling studies revealed that the S-to-C2 oxygen-transfer process involves
initial formation of a C1-O linkage followed by O-migration to C2, leading to the generation of an intermediate
glycosyl 1,2-anhydropyranoside that serves as an in situ glycosylating agent. These findings are consistent
with the initial formation of a C2-sulfonium-C1-oxosulfonium pyranosyl species upon activation of the glycal
donor with Aryl2SO‚Tf2O.
Introduction
conjugates of varied structure and function. In particular,
oxidative glycosylation methods in which an oxygen substituent
The development of new methods for the preparation of
complex oligosaccharides and glycoconjugates is a challenging
endeavor in organic synthesis and glycobiology. Glycosylation
is a critical function that modulates important cellular processes1
and has prompted the development of numerous chemical
methods for the construction of glycosidic bonds.2 The use of
glycal substrates in oligosaccharide synthesis is an attractive
strategy for carbohydrate synthesis, as selective functionalization
at the C2-position of the glycal enol ether and anomeric bond
formation can be achieved in the glycosylation process. Elegant
methods are available by which oxygen,3 nitrogen,4 halide,5
sulfur, selenium,6 carbon,7 and hydrogen8 substituents can
be introduced at the C2-position to provide synthetic glyco-
is directly introduced at the C2-position of a glycal donor have
seen extensive utility in carbohydrate synthesis. This “glycal
assembly” approach, which typically involves direct glycal
epoxidation9 followed by oxirane ring opening of the resulting
1,2-anhydropyranoside10 by a nucleophilic glycosyl acceptor,
has seen recent remarkable advances in the context of the
preparation of complex natural products, immunostimulants, and
cell-surface oligosaccharides.11
(5) (a) Lemieux, R. U.; Levine, S. Can. J. Chem. 1965, 43, 2190-2198. (b)
Tatsuta, K.; Fujimoto, K.; Kinoshita, M.; Umezawa, S. Carbohydr. Res.
1977, 54, 85-104. (c) Thiem, J.; Karl, H.; Schwentner, J. Synthesis 1978,
696-698. (d) Burkart, M. D.; Zhang, Z.; Hung, S.-C.; Wong, C.-H. J. Am.
Chem. Soc. 1997, 119, 11743-11746.
(6) For C2-sulfur/selenium substituents, see: (a) Jaurand, G.; Beau, J.-M.;
Sinay¨, P. J. Chem. Soc., Chem. Commun. 1981, 572-573. (b) Ito, Y.;
Ogawa, T. Tetrahedron Lett. 1987, 28, 2723-2726. (c) Preuss, R.; Schmidt,
R. R. Synthesis 1988, 694-697. (d) Perez, M.; Beau, J.-M. Tetrahedron
Lett. 1989, 30, 75-78. (e) Grewal, G.; Kaila, N.; Franck, R. W. J. Org.
Chem. 1992, 57, 2084-2092. (f) Roush, W. R.; Sebesta, D. P.; James, R.
A. Tetrahedron 1997, 53, 8837-8852.
(7) (a) Linker, T.; Sommermann, T.; Kahlenberg, F. J. Am. Chem. Soc. 1997,
119, 9377-9384. (b) Beyer, J.; Madsen, R. J. Am. Chem. Soc. 1998, 120,
12137-12138.
(8) (a) Toshima, K.; Tatsuta, K.; Kinoshita, M. Bull. Chem. Soc. Jpn. 1988,
61, 2369-2381. (b) Bolitt, V.; Mioskowski, C.; Lee, S.-G.; Falck, J. R. J.
Org. Chem. 1990, 55, 5812-5813. (c) Kaila, N.; Blumenstein, M.;
Bielawska, H.; Franck, R. W. J. Org. Chem. 1992, 57, 4576-4578. (d)
Sabesan, S.; Neira, S. J. Org. Chem. 1991, 56, 5468-5472. (e) France, C.
J.; McFarlane, I. M.; Newton, C. G.; Pitchen, P.; Webster, M. Tetrahedron
Lett. 1993, 34, 1635-1638. (f) McDonald, F. E.; Reddy, K. S. Angew.
Chem., Int. Ed. 2001, 40, 3653-3655.
(9) (a) Halcomb, R. L.; Danishefsky, S. J. J. Am. Chem. Soc. 1989, 111, 6661-
6666. (b) Barili, P. L.; Berti, G.; D’Andrea, F.; Di Bussolo, V.; Granucci,
I. Tetrahedron 1992, 48, 6273-6284. (c) Bellucci, G.; Catelani, G.;
Chiappe, C.; D’Andrea, F. Tetrahedron Lett. 1994, 35, 8433-8436. (d)
Cavicchioli, M.; Mele, A.; Montanari, V.; Resnati, G. J. Chem. Soc., Chem.
Commun. 1995, 901-902. (e) Yang, D.; Wong, M.-K.; Yip, Y.-C. J. Org.
Chem. 1995, 60, 3887-3889. (f) Boehlow, T. R.; Buxton, P. C.; Grocock,
E. L.; Marples, B. A.; Waddington, V. L. Tetrahedron Lett. 1998, 39, 1839-
1842.
* To whom correspondence should be addressed. E-mail: gin@
scs.uiuc.edu.
(1) (a) Dwek, R. A. Chem. ReV. 1996, 96, 683-720. (b) Bertozzi, C. R.;
Kiessling, L. L. Science 2001, 291, 2357-2364. (c) Helenius, A.; Aebi,
M. Science 2001, 291, 2364-2369.
(2) (a) Toshima, K.; Tatsuta, K. Chem. ReV. 1993, 93, 1503-1531. (b) Schmidt,
R. R.; Kinzy, W. AdV. Carbohydr. Chem. Biochem. 1994, 50, 21-123. (c)
Davis, B. G. J. Chem. Soc., Perkin Trans. 1 1999, 3215-3237. (d) Davis,
B. G. J. Chem. Soc., Perkin Trans. 1 2000, 2137-2160.
(3) Danishefsky, S. J.; Bilodeau, M. T. Angew. Chem., Int. Ed. Engl. 1996,
35, 1380-1419 and references therein.
(4) (a) Lemieux, R. U.; Ratcliffe, R. M. Can. J. Chem. 1979, 57, 1244-1251.
(b) Leblanc, Y.; Fitzsimmons, B. J.; Springer, J. P.; Rokach, J. J. Am. Chem.
Soc. 1989, 111, 2995-3000. (c) Griffith, D. A.; Danishefsky, S. J. J. Am.
Chem. Soc. 1990, 112, 5811-5819. (d) Czernecki, S.; Ayadi, E. Can. J.
Chem. 1995, 73, 343-350. (e) Rubinstenn, G.; Esnault, J.; Mallet, J.-M.;
Sinay¨, P. Tetrahedron: Asymmetry 1997, 8, 1327-1336. (f) Du Bois, J.;
Tomooka, C. S.; Hong, J.; Carreira, E. M. J. Am. Chem. Soc. 1997, 119,
3179-3180. (g) Das, J.; Schmidt, R. R. Eur. J. Org. Chem. 1998, 1609-
1613. (h) Di Bussolo, V.; Liu, J.; Huffman, L. G., Jr.; Gin, D. Y. Angew.
Chem., Int. Ed. 2000, 39, 204-207. (i) Nicolaou, K. C.; Baran, P. S.; Zhong,
Y.-L.; Vega, J. A. Angew. Chem., Int. Ed. 2000, 39, 2525-2529. (j) Kan,
C.; Long, C. M.; Paul, M.; Ring, C. M.; Tully, S. E.; Rojas, C. M. Org.
Lett. 2001, 3, 381-384.
9
10.1021/ja025639c CCC: $22.00 © 2002 American Chemical Society
J. AM. CHEM. SOC. 2002, 124, 7343-7352
7343

A R T I C L E S
Honda and Gin
Scheme 1
Chart 1
Existing methods for glycal epoxidation involve the use of
well-known alkene oxidants such as peroxyacids and dioxirane
reagents.9 Recently, we have developed a new method for C2-
hydroxyglycosylation (Scheme 1) with glycal donors (1),12
wherein a mixture of excess diaryl sulfoxide and Tf2O is
employed as the glycal oxidant. Following initial glycal activa-
tion with this reagent combination, subsequent sequential
addition of methyl alcohol and a tertiary amine (R′3N), followed
by the glycosyl acceptor (Nu-H) and a Lewis acid such as ZnCl2,
yields C2-hydroxypyranosides 2 in a one-pot procedure. The
process offers a new method for sulfonium-mediated oxygen
transfer to glycals and thus presents a novel means for
glycoconjugate construction. We report herein our investigations
into the identification and characterization of reactive glycosyl
intermediates in the process, thereby providing key insights into
the pathway of oxygen transfer.
Results and Discussion
The C2-hydroxyglycosylation reaction depicted in Scheme
1 involves a one-pot procedure, in which an oxygen-transfer
reagent is generated in situ from the addition of triflic anhydride
(1.5 equiv) to a solution of glucal 1 (1 equiv), diphenyl sulfoxide
(3 equiv), and 2,6-di-tert-butyl-4-methylpyridine (DTBMP, 3-4
equiv) in dichloromethane at -78 °C. After 1 h at -40 °C,
anhydrous methyl alcohol (1 equiv) and triethylamine (3 equiv)
are introduced. The reaction is allowed to proceed at 23 °C for
1 h, at which time the glycosyl acceptor (Nu-H, 2-3 equiv)
and a Lewis acid (ZnCl2, 1-2 equiv) are added to afford the
C2-hydroxy-â-D-glucopyranoside product 2 in good yields and
high diastereoselectivity. The method was found to be amenable
to the preparation of a variety of O- and N-glycoconjugates
(Chart 1).12,13
Scheme 2
Proposed Reaction Pathways. The transformation in Scheme
1 provides a useful method for oxygen transfer to glycal donors
with concomitant anomeric bond construction. The reaction
pathway appears to involve a novel process for oxygen transfer
to glycals, especially in light of the paucity of existing methods
glycosylation, although it is likely that the mechanism of the
transformation proceeds via initial electrophilic activation of
the glycal donor by the Ph2SO‚Tf2O complex.
Our initial observations revealed several key aspects of the
C2-hydroxyglucosylation. First, nearly stoichiometric quantities
of Ph2S are generated in these glycosylation reactions, which
is indicative of the sulfoxide reagent serving as the oxidant.
This hypothesis was tested by an 18O-labeling experiment
(Scheme 2) in which Ph2S18O (96% 18O incorporation) was
employed as the sulfoxide reagent in an oxidative glycosylation
with tri-O-benzyl-D-glucal (3) as the donor and sodium azide
as a non-oxygen nucleophilic acceptor. The C2-hydroxy glycosyl
azide 4 (79%) is formed with 93% 18O incorporation (as
measured by FAB+ mass spectrometry) along with 0.7 equiv
of Ph2S. This result clearly indicates that the oxidation process
proceeds through a stereoselective oxygen transfer from the
sulfoxide reagent to the C2-position of the glycal donor. Second,
monitoring of the oxidative glycosylation reaction with tri-O-
benzyl-D-glucal (3) by 1H NMR revealed that the 1,2-anhydro-
pyranoside 5 was formed as the principal carbohydrate species
in solution following the addition of MeOH, prior to the
for sulfoxide-mediated oxidation of electron-rich alkenes.14
A
number of reaction pathways can be envisioned for the oxidative
(10) For pioneering work in the synthesis and utility of 1,2-anhydropyranosides,
see: (a) Brigl, P. Z. Physiol. Chem. 1922, 245-256. (b) Hardegger, E.; de
Pascual, J. HelV. Chim. Acta. 1948, 31, 281-286. (c) Lemieux, R. U.;
Huber, G. J. Am. Chem. Soc. 1953, 75, 4118. (d) Sondheimer, S. J.;
Yamaguchi, H.; Schuerch, C. Carbohydr. Res. 1979, 74, 327-332. (e)
Klein, L. L.; McWhorter, W. W., Jr.; Ko, S. S.; Pfaff, K.-P.; Kishi, Y.;
Uemura, D.; Hirata, Y. J. Am. Chem. Soc. 1982, 104, 7362-7364. (f)
Trumbo, D. L.; Schuerch, C. Carbohydr. Res. 1985, 135, 195-202. (g)
Bellosta, V.; Czernecki, S. J. Chem. Soc., Chem. Commun. 1989, 199-
200.
(11) (a) Park, T. K.; Kim, I. J.; Hu, S.; Bilodeau, M. T.; Randolph, J. T.;
Kwon, O.; Danishefsky, S. J. J. Am. Chem. Soc. 1996, 118, 11488-11500.
(b) Seeberger, P. H.; Eckhardt, M.; Gutteridge, C. E.; Danishefsky, S. J.
J. Am. Chem. Soc. 1997, 119, 10064-10072. (c) Deshpande, P. P.;
Danishefsky, S. J. Nature 1997, 387, 164-166. (d) Kuduk, S. D.; Schwarz,
J. B.; Chen, X.-T.; Glunz, P. W.; Sames, D.; Ragupathi, G.; Livingston,
P. O.; Danishefsky, S. J. J. Am. Chem. Soc. 1998, 120, 12474-12485.
(12) Di Bussolo, V.; Kim, Y.-J.; Gin, D. Y. J. Am. Chem. Soc. 1998, 120,
13515-13516.
(13) Kim, Y.-J.; Gin, D. Y. Org. Lett. 2001, 3, 1801-1804.
(14) Nenajdenko, V. G.; Vertelezkij, P. V.; Gridnev, I. D.; Shevchenko, N. E.;
Balenkova, E. S. Tetrahedron 1997, 53, 8173-8180.
9
7344 J. AM. CHEM. SOC. VOL. 124, NO. 25, 2002

Probing the Mechanism of C2-Hydroxyglycosylation
A R T I C L E S
Scheme 3
Scheme 4
introduction of the nucleophilic acceptor.15 Thus, the formation
of the C2-hydroxyglycoside products illustrated by 4 is likely
the result of nucleophilic ring opening of 1,2-anhydropyranoside
intermediates such as 5. It is also significant that the formation
of the R-glucal epoxide 5 occurs only after the addition of the
nucleophile (i.e., MeOH) to the activated glycal intermediate.
Finally, the formation of the C2-hydroxyglycoside products
proceed efficiently only in the presence of excess sulfoxide
reagent. These initial observations provide a useful mechanistic
framework from which plausible reaction pathways for the
oxygen-transfer process can be investigated.
Two possible reaction pathways for Ph2SO-mediated C2-
hydroxyglycosylation consistent with the above observations are
illustrated in Schemes 3 and 4. In the first (Scheme 3), the
reagent combination of Ph2SO‚Tf2O is considered to generate
diphenylsulfide bis(triflate) (6) in situ, which can engage in
electrophilic activation of the glycal donor at C2. The resulting
C2-functionalization of 3 generates a C2-sulfonium-C1-oxo-
carbenium (or glycosyl) triflate intermediate 7, which incor-
porates electrophilic sites at C1 and C2. In the presence of
excess sulfoxide reagent, O-substitution of excess Ph2SO at
the anomeric position leads to the C2-sulfonium-C1-oxo-
sulfonium pyranosyl intermediate 8. Addition of an initial
hydroxyl nucleophile such as MeOH then leads to addition of
the alcohol at the anomeric oxosulfonium group, generating a
transient anomeric σ-sulfurane species 9. Fragmentation of
9 with concomitant intramolecular displacement of the C2-
sulfonium substituent results in formation of the observed 1,2-
anhydropyranoside 5, accompanied by expulsion of Ph2S and
Ph2SOMe+‚TfO-, which is hydrolyzed to Ph2SO upon aqueous
workup. Finally, introduction and glycosylation of the nucleo-
philic acceptor (Nu-H) by 5 generates the C2-hydroxy-â-
glucopyranoside 10. It is worth noting that the stereochemical
course of the reaction involves a glucal f â-glucopyranoside
transformation by way of the R-1,2-anhydropyranoside 5.
According to Scheme 3, this requires a sterically demanding
â-approach of the sulfonium reagent 6 onto the glycal 3 (i.e.,
syn to the C3-alkoxy substituent) to provide 7/8. This is
plausible, given that the sulfonium electrophile would assume
a favorable â-axial approach onto the C2-position of the glucal16
in a chairlike transition structure, resulting in a net 1,2-trans-
diaxial functionalization of the endocyclic π-system in 3 to
form 8.
An alternate reaction pathway (Scheme 4) invokes a different
sulfonium reagent in the glycal activation process. In situ
generation of diphenylsulfide bis(triflate) (6) in the presence of
excess Ph2SO leads to the formation of the oxygen-bridged bis-
(sulfonium) salt 11 prior to glycal activation. If electrophilic
activation of glycal 3 were to occur by addition of the C2-π-
nucleophile onto a sulfonium center within 11, the reaction
pathway depicted in Scheme 3 would ensue. Alternatively,
addition of the C2-π-nucleophile of 3 to the bridging oxygen
functionality in 11 results in expulsion of Ph2S and formation
of the C2-oxosulfonium-C1-pyranosyl triflate intermediate 12.
Upon introduction of 1 equiv of MeOH as a sacrificial
nucleophile, its addition to the C2-oxosulfonium functionality
in 12 initiates the scission of the S-O bond, leading to
intramolecular expulsion of the anomeric triflate functionality
at C1 to generate the observed 1,2-anhydropyranoside interme-
diate (i.e., 12 f 13 f 5). In this proposed reaction pathway
(Scheme 4), the stereochemical course of the activation process
is distinct from that of Scheme 3. In this case, R-approach of
11, consistent with reagent approach under steric control, leads
to initial transfer of the oxygen atom to the R-face of the glycal
at C2.17 Nevertheless, this reaction pathway is also consistent
with the above observations in the C2-hydroxyglycosylation
reaction, which include (1) the transfer of an oxygen atom from
the sulfoxide reagent to the C2-position of the glycal donor;
(2) the requirement for excess sulfoxide reagent for the reaction
to proceed; (3) the generation of Ph2S as a principal byproduct;
(16) â-Approach of electrophilic reagents onto the C2-position of protected glucal
substrates are not uncommon. See, for example, refs 3, 5a,b, and 6a.
(17) For examples of R-approach of electrophilic reagents onto the C2-position
of protected glucal substrates, see ref 6c-f.
(15) This assignment was verified by independent synthesis. See ref 9c.
9
J. AM. CHEM. SOC. VOL. 124, NO. 25, 2002 7345

A R T I C L E S
Honda and Gin
Scheme 5
Scheme 6 a
a Reagents and conditions: (a) Ac2O, HBr, AcOH; Zn, CuSO4; PhCH2Cl,
KOH.
Scheme 7 a
and (4) the formation of the 1,2-anhydropyranoside 5 as a key
intermediate in the O-transfer process.
Detection of Reactive Glycosyl Intermediates. A key feature
which distinguishes between the reaction pathways outlined in
Schemes 3 and 4 is the stage at which the C2-O bond is formed
during the C2-hydroxyglycosylation reaction. In Scheme 3,
glycal oxidation proceeds by formation of the R-C1-O bond
prior to oxirane formation; in Scheme 4, the R-C2-O linkage
is established prior to oxirane formation. Thus, detection and
characterization of the putative reactive intermediates 8 and 9
(Scheme 3) or 12 and 13 (Scheme 4) would reveal the course
of oxygen delivery from sulfoxide to glycal.
To determine whether the initial glycosyl intermediate formed
in the oxidative glycosylation incorporates a C1-O or a C2-O
linkage, low-temperature NMR monitoring of the glycal activa-
tion process was performed, employing heavy-atom 18O-labeled
sulfoxide reagents. It is known that 13C NMR resonances
experience a small but significant upfield chemical shift
perturbation (i.e., ∼0.01-0.05 ppm) when directly bound to
18O relative to that of a 13C-16O linkage;18 consequently, this
technique should be amenable to establishing the nature of C-O
connectivity in the reactive glycosyl intermediates generated
in the C2-hydroxyglycosylation process. Activation of a glycal
donor 3 (Scheme 5) with the reagent combination of Ph2SO‚
Tf2O would, according to the proposed pathways in Schemes 3
and 4, generate two different intermediates, either 8 or 12. If
the C1-oxosulfonium intermediate 8 were formed as the initial
species with Ph2S18O as the sulfoxide reagent, an upfield 13C1
chemical shift perturbation would be observed relative to
reaction with unlabeled Ph2S16O.19 Conversely, if the C2-
oxosulfonium intermediate 12 were generated in the activation
process, a 13C2 upfield chemical shift perturbation in the range
of 0.01-0.05 ppm would be observed with Ph2S18O as the
b
a δC1 values measured at -40 °C. δC1 values measured at -20 °C.
sulfoxide reagent.20 To this end, both the 13C1- and 13C2-labeled
3,4,6-tri-O-benzyl-D-glucal substrates 3a and 3b (Scheme 6)
were prepared from 13C1- and 13C2-labeled glucose (14a/b),
respectively,21 and employed in this study to ensure adequate
detection and unambiguous assignment of both the 13C1 and
13C2 resonances of the putative reactive intermediates.
The first series of low-temperature 13C NMR investigations
employed the 13C1-labeled glucal 3a (Scheme 7, δC1 ) 143.889
in CD2Cl2) with a focus on the reaction pathway in Scheme 3.
Upon treatment of this donor with the reagent combination of
Ph2S16O (3 equiv) and Tf2O (1.5 equiv) at -78 °C, followed
(20) It is worth noting that in the proposed formation of either diphenylsulfide
bis(triflate) (6) or bis(sulfonium) species (11), 50% of the triflate anions
could conceivably incorporate 18O if rapid exchange of the triflate anions
were to occur at the sulfonium center(s). A mechanism involving oxygen
transfer from sulfoxide to triflate to glycal is unlikely, as this would erode
the amount of 18O incorporation on the C2-hydroxyglycoside product. The
possibility of incorporation of 18O into 50% of the triflate anions in the
activation process might also lead to a 13C1 upfield chemical shift
perturbation in the intermediates 7 (Scheme 3) or 12/13 (Scheme 4) if a
covalent C1-O-triflate bond were to exist in these species; however, if
this were the case, the relative intensity of the upfield-shifted 13C1-18O
resonance would be significantly diminished compared with the amount
of 18O incorporation in the sulfoxide reagent.
(18) See, inter alia: (a) Jameson, C. J. J. Chem. Phys. 1977, 66, 4983-4988.
(b) Risley, J. M.; Van Etten, R. L. J. Am. Chem. Soc. 1979, 101, 252-
253. (c) Darensbourg, D. J. J. Organomet. Chem. 1979, 174, C70-C76.
(d) Vederas, J. C. J. Am. Chem. Soc. 1980, 102, 374-376. (e) Vederas, J.
C. Can. J. Chem. 1982, 60, 1637-1642. (f) Risley, J. M.; Van Etten, R. L.
In NMRsBasic Principles and Progress; Gunther, H., Ed.; Springer-
Verlag: Berlin, 1990; Vol. 22, pp 81-168.
(19) Any chemical shift perturbation at 13C2 in 8 would likely be too small to
be observed at low temperature. 18O chemical shift perturbations on â-
and γ-13C signals can be detected, although magnitudes of ∆δ are typically
<0.01 ppm. See, for example: (a) Moore, R. N.; Diakur, J.; Nakashima,
T. T.; McLaren, S. L.; Vederas, J. C. J. Chem. Soc., Chem. Commun. 1981,
501-502. (b) Mega, T. L.; Van Etten, R. L. J. Am. Chem. Soc. 1993, 115,
12056-12059.
(21) Shull, B. K.; Wu, Z.; Koreeda, M. J. Carbohydr. Chem. 1996, 15, 955-
964.
9
7346 J. AM. CHEM. SOC. VOL. 124, NO. 25, 2002

Probing the Mechanism of C2-Hydroxyglycosylation
A R T I C L E S
Scheme 8 a
Figure 1.
by warming to -40 °C, a principal glycosyl intermediate,
assigned as 8a (Scheme 7, Figure 1i), bearing a 13C1 resonance
at δ104.108 was observed. Recooling of the reaction to -78
°C and subsequent addition of the initial sacrificial nucleophile
(MeOH, 1 equiv) and Et3N (3 equiv) led to the immediate
formation of a second glycosyl intermediate, assigned as 9a.
Upon warming of the reaction to -20 °C, steady conversion to
the 1,2-anhydropyranoside 5a (δC1 ) 77.559)15 was observed.
With the detection of two principal intermediates in the
oxidative glycosylation reaction prior to oxirane formation, the
analogous 13C1 NMR tracking experiment was performed
employing 60% 18O-labeled diphenyl sulfoxide to probe the
structure of the reactive species.22 Treatment of 3a with 60%
Ph2S18O and Tf2O (3.0 and 1.5 equiv, respectively) led to the
formation of an intermediate 8a with two distinct 13C1
resonances, one at δ 104.108 and a second of slightly larger
intensity at δ 104.064 (Scheme 7, Figure 1ii). This chemical
shift perturbation (∆δC1 ) 0.044 ppm upfield) upon incorpora-
tion of Ph2S18O as the oxidant is indicative of a 13C1-18O bond,
consistent with the initial intermediate of glycal activation being
the glycosyl oxosulfonium species 8a. Subsequent addition of
MeOH (1 equiv) and Et3N (3 equiv) at -78 °C and warming
to -20 °C afforded the intermediate 9a with two 13C1
resonances, one at δ 89.361 and the other at δ 89.346. It should
be noted that these two 16/18O isotopomer 13C1 resonances in
9a could not be resolved at temperatures below -20 °C due to
peak broadening; however, upon warming to -20 °C during
the conversion to the 1,2-anhydropyranoside 5a, the ∆δ value
of 0.015 ppm could be clearly discerned in 9a (Figure 1ii). Given
the marked upfield 13C1 chemical shift of this second inter-
mediate (9a) compared to that of the first (8a), it is reasonable
to propose that the introduction of MeOH generates the glycosyl
sulfurane intermediate 9a, which leads to the 1,2-anhydro-
pyranoside 5a. Finally, examination of the 13C1 resonances in
the 1,2-anhydropyranoside 5a revealed, not surprisingly, two
signals, one at δ 77.559 (13C1-16O) and the other at δ 77.522
(13C1-18O). These initial experiments therefore are consistent
with the formation of the intermediates 8 and 9 of Scheme 3,
b
a δC2 values measured at -40 °C. δC2 values measured at -20 °C.
both bearing a C1 linkage to the sulfoxide oxygen atom prior
to formation of the 1,2-anhydropyranoside 5.23
To evaluate the alternative pathway of Scheme 4 for oxygen
transfer in the C2-hydroxyglucosylation reaction, the analogous
complementary 13C2 NMR tracking experiments were performed
with 13C2-labeled tri-O-benzyl-D-glucal (3b, δC2 99.185).
Activation of 3b (Scheme 8, Figure 2) with either Ph2S16O or
Ph2S16/18O (60% 18O incorporation) and Tf2O in CD2Cl2 at low
temperature generates the first glycosyl intermediate 8b with
only a single discernible 13C2 resonance at δ 58.152, even when
60% Ph2S18O was employed (Scheme 8, Figure 2).24 Following
the addition of MeOH and Et3N to 8b, the second intermediate
9b is generated, incorporating a single 13C2 resonance (δ
62.415), even when 60% Ph2S16/18O was employed. Finally,
conversion of the second intermediate 9b to the 1,2-anhydro-
pyranoside 5b at -20 °C produced two 13C2 resonances when
(23) In all of the low-temperature NMR experiments reported herein, initial
activation of the glucal is accompanied by the formation of ∼20% of the
C2-sulfonium glucal (below) as a byproduct, presumably a result of
elimination of H2 from the putative reactive glycosyl species 7 (Scheme
3). This minor competitive pathway could not be avoided, as small
temperature fluctuations (i.e., elevations) are inevitably associated with the
transfer of the NMR reaction tube to the precooled NMR probe following
introduction of Tf2O. The formation of the C2-sulfonium glucal is also
consistent with the pathway depicted in Scheme 3, if it is indeed formed
in a manifold diverted from the productive pathway.
(22) The chemical shift perturbation values (∆δ) to be measured are small (0.01-
0.05 ppm), and minor chemical shift fluctuations of this magnitude may
occur between NMR experiments, especially at variable temperatures. Thus,
use of 60% 18O-labeled Ph2SO, as opposed to >90% 18O incorporation,
ensures the detection of isotope-induced ∆δ values in the form of two
distinct resonances of nearly equal intensity.
(24) Two-bond 18O-C1-13C2 chemical shift perturbations (typically <0.01 ppm;
see ref 19) were not detected, presumably due to some degree of peak
broadening at low temperature.
9
J. AM. CHEM. SOC. VOL. 124, NO. 25, 2002 7347

A R T I C L E S
Honda and Gin
Scheme 9 a
Figure 2.
a Warmed to -20 °C for a short time (∼6 min) to allow for acquisition
60% Ph2S18O was employed, one at δ 52.445 (13C2-16O) and
the other at δ 52.415 (13C2-18O), revealing an 18O-isotope-
induced ∆δC2 value of 0.030 ppm upfield. These results indicate
that both of the glycosyl intermediates 8b and 9b are devoid of
a C2-O linkage, which is not consistent with the expectations
for reaction via Scheme 4. Only when 9b is converted to the
1,2-anhydropyranoside 5b through the formation of a C2-O
linkage is the 18O-heavy-atom 13C2 chemical shift perturbation
detected.25
1
3
of H NMR data with resolved JHH values.
Scheme 10
These heavy-atom labeling investigations confirm the C1-
to-C2 pathway (consistent with Scheme 3) for oxygen transfer
from the sulfoxide reagent to the glycal donor and thus reveal
key details concerning the structures of the initial activated
glycal intermediates. Further structural characterization of the
intermediate d21-9. Low-temperature 1H NMR analysis (-20 °C)
3
of d21-9 reveals a JHH coupling pattern distinct from that of
1
d21-8. The relatively large values of 3JH3H4 (9.1 Hz; axial-axial)
activated glycosyl intermediates by H NMR analysis allows
3
and JH4H5 (9.1 Hz; axial-axial), the relatively small values
insight into the stereochemical and conformational properties
of 8 and 9. These studies employed tri-O-d7-benzyl-D-glucal
(d21-3, Scheme 9)26 as the glucal donor in the C2-hydroxygly-
cosylation in order to facilitate identification and assignment
of the pyranoside 1H resonances within the intermediates d21-8
and d21-9 observed at low temperature. Thus, 1H NMR analysis
of d21-8 at -20 °C27 reveals that it does not exist in a traditional
4C1 chair conformation. Rather, the relatively large 3JH4H5 (9.1
3
3
of JH1H2 (<1 Hz; equatorial-equatorial) and JH2H3 (4.1 Hz;
1
equatorial-axial), and the large value of JC1H1 (171 Hz;
C1-H1 equatorial) are indicative of d21-9 being an R-manno-
pyranosyl-like substrate in a 4C1 chair conformation. These data,
in conjunction with the previous 18O-labeling experiments, are
consistent with the intermediacy of glycosyl species such as 8
and 9 in the glycal epoxidation pathway outlined in Scheme 3.
C2-Hydroxymannosylation. The use of Ph2SO as the
sulfoxide reagent for the C2-hydroxyglucosylation leads to the
stereoselective formation of C2-hydroxy-â-glucopyranosides
when glucal donors are employed. We have recently reported
that a complementary manifold for C2-hydroxyglycosylation to
generate C2-hydroxy-R-mannopyranosides 15 (Scheme 10) from
glucal donors 1 can be accessed when dibenzothiophene 5-oxide
(DBTO) is employed as the sulfoxide reagent.30 This remarkable
reversal of diastereoselectivity with a seemingly subtle change
in sulfoxide reagent thus allows for the direct conversion of
glucal donors to mannopyranosides, a transformation that
heretofore has required at least a four-step synthetic sequence.11b
3
Hz) value and the relatively small 3JH1H2 (5.5 Hz),28 JH2H3 (3.2
Hz), and 3JH3H4 (2.3 Hz) values are consistent with a twist-boat
conformation (d21-8), in which the C2-sulfonium group adopts
a pseudoequatorial orientation.29 In the subsequent stage of
the oxidation reaction, addition of MeOH (1 equiv) and Et3N
(3 equiv) to d21-8 generates the putative glycosyl sulfurane
(25) The relative intensities of the 13C1 and 13C2 resonances which exhibit
18O isotope-inducted ∆δC values (Figures 1ii and 2ii) all reflect the ratio
of 18O incorporation in the sulfoxide reagent (60% 18O). Therefore, it
is unlikely that the observed isotope-induced ∆δC values are due to
13C-18OTf linkages arising from the CF3SO218O- counterion byproduct,
as this would generate a significantly diminished 13C-18O resonance
intensity compared to that observed.
(26) Koto, S.; Asami, K.; Hirooka, M.; Nagura, K.; Takizawa, M.; Yamamoto,
S.; Okamoto, N.; Sato, M.; Tajima, H.; Yoshida, T.; Nonaka, N.; Sato, T.;
Zen, S.; Yago, K.; Tomonaga, F. Bull. Chem. Soc. Jpn. 1999, 72, 765-
777.
1
(29) A JC1H1 value of 189 Hz was observed for 8a. For D-pyranosides in a
1
4C1 conformation, JC1H1 values ) 170 Hz are indicative of an equatorial
1
(27) Although the conversion of d21-3 to d21-8 proceeds rapidly at -78 °C,
the solution of d21-8 was warmed to -20 °C for a short interval (∼6 min)
to aquire 1H NMR data before recooling of the reaction to -78 °C for
the next step. Warming of d21-8 to -20 °C allowed for the acquisition of
C1-H1 linkage (R-glycoside), where the magnitude of JC1H1 increases
with the electron-withdrawing capacity of the anomeric substituent.
However, it is worth noting that the twist-boat conformation implied by
1
the 1H NMR data might preclude direct correlation of JC1H1 magnitudes
3
1H NMR data with adequately resolved JHH values.
with anomeric configuration, especially with cationic substituents on the
pyranose ring. See: Duus, J. Ø.; Gotfredsen, C. H.; Bock, K. Chem. ReV.
2000, 100, 4589-4614.
(28) The observed δH2 value of 6.12 ppm is also consistent with that of R-protons
in alkyldiphenylsulfonium salts. See, for example: Trost, B. M.; Bogdanow-
icz, M. J. J. Am. Chem. Soc. 1973, 95, 5298-5307.
(30) Kim, J.-Y.; Di Bussolo, V.; Gin, D. Y. Org. Lett. 2001, 3, 303-306.
9
7348 J. AM. CHEM. SOC. VOL. 124, NO. 25, 2002

Probing the Mechanism of C2-Hydroxyglycosylation
A R T I C L E S
Scheme 11
Scheme 12 a
Initial mechanistic investigations into this transformation
revealed that (1) excess DBTO is required for the transformation
to proceed; (2) dibenzothiophene is formed as a principal
byproduct of the reaction; (3) the sulfoxide oxygen atom is
transferred to the C2 position of the glycal donor, as evidenced
by 18O-labeling of the sulfoxide reagent (Scheme 11) in the C2-
hydroxyglycosylation of NaN3 to form 16; and (4) the â-1,2-
anhydropyranoside 17 is formed as the key intermediate in the
oxygen-transfer process prior to the glycosidic coupling event.15
It is reasonable that the oxidative glycosylation with glucals
employing DBTO proceeds via an oxygen-transfer pathway
similar to that employing Ph2SO. However, given the marked
difference in stereochemical outcome of the two reactions (i.e.,
Scheme 1 vs Scheme 10), we sought to determine whether the
reactions involving DBTO also involved a C1-to-C2 oxygen-
transfer pathway similar to that depicted in Scheme 3.
a δC1 values measured at -20 °C in separate experiments.
13C1-labeled tri-O-benzyl-D-glucal (3a) was treated with the
reagent combination of excess DBTO (3 equiv) and Tf2O (1.5
equiv) in CD2Cl2 at -78 °C (Scheme 12).31 Monitoring of the
reaction of 3a with DBTO‚Tf2O at -78 °C by 13C NMR
revealed the presence of a principal intermediate, assigned as
18a (Scheme 12). This initial intermediate is stable at -78 °C
but undergoes decomposition to the C2-vinylsulfonium salt 19
upon temperature elevation above -40 °C. With the increased
propensity for the DBTO-derived glycosyl intermediate 18a to
form the glycal sulfonium salt 19, the detection of an isotope-
induced ∆δC1 value when DBT16/18O was used was particularly
challenging in that resolution of ∆δC1 in 18a was not possible
at temperatures below -40 °C due to broadening of the 13C1
resonances. Consequently, the reaction mixture of glucal 3a and
DBT16/18O was activated with Tf2O at -78 °C and then warmed
to -20 °C. At this elevated temperature, the unwanted conver-
sion of the reactive glycosyl intermediate 18a to the C2-
sulfonium glycal salt 19 was rapid; however, the 13C1 resonance
of 18a could be monitored transiently at -20 °C as it diminished
during the decomposition process. Under these 13C NMR
detection conditions, one can observe the putative glycosyl
intermediate 18a, which incorporates a 13C1 resonance at δ
97.852 (Scheme 12) when unlabeled DBTO was employed in
Figure 3.
be detected at -20 °C, indicative of the presence of a C1-O
linkage within 18a (Scheme 12, Figure 3ii).
The next step in the C2-hydroxymannosylation reaction, the
introduction of a sacrificial nucleophile (MeOH) to initiate
epoxide ring closure, was investigated by treatment of a solution
of the glucal donor 3a, DBTO, and Tf2O in CD2Cl2 at -78 °C
with methyl alcohol (1 equiv) and Et3N (3 equiv) with 13C NMR
monitoring. Unlike the previous series of experiments with
Ph2SO, in which the putative glycosyl oxosulfurane 9 could be
detected as a second glycosyl intermediate prior to oxirane
formation, a similar intermediate could not be detected in the
corresponding oxidation employing DBT16O (Scheme 12).
Unfortunately, only a complex mixture of intermediates was
generated at this stage, as evidenced by multiple 13C1 NMR
the activation (Figure 3i). When 61% 18O-labeled DBT16/18
O
was employed, an isotope-induced ∆δC1 of 0.030 ppm could
(31) It is worth noting that the previously reported optimized reaction conditions
for the conversion of glucals to C2-hydroxy-R-mannopyranosides employed
5 equiv of DBTO (Scheme 10), as these reaction conditions provide
R-manno:â-gluco diastereoselectivities of at least 5:1. However, since the
reactions in the current NMR studies were performed in 5-mm sample tubes
at low temperature where mixing of the reagents is less efficient, only 3
equiv of DBTO was employed to ensure complete dissolution of the
sulfoxide reagent at low temperatures. The use of only 3 equiv of DBTO
in the NMR reactions led to erosion of the diastereoselectivity of oxygen
transfer to the glycal (3:1, 17:5).
9
J. AM. CHEM. SOC. VOL. 124, NO. 25, 2002 7349

A R T I C L E S
Honda and Gin
Scheme 13 a
Figure 4.
the R-face,17 followed by the addition of excess DBTO onto
C1 from the â-face to provide 18. Addition of the sacrificial
nucleophile (MeOH) then serves to close the epoxide function-
ality with concomitant expulsion of dibenzothiophene. The C2-
hydroxymannosylation process therefore appears to be analogous
to that of C2-hydroxyglucosylation involving Ph2SO, with the
obvious key difference being the reversal of stereoselectivity
in the oxygen transfer. While the remarkable stereochemical
contrast of the two reactions (Scheme 1 versus Scheme 10) is
likely the result of opposing facial selectivities in the initial
sulfonium-mediated glycal activation processes, it is worth
noting that the origin of the stereochemical reversal in the C2-
hydroxyglycosylation reaction employing Ph2SO versus DBTO
is as yet unclear and is currently under investigation.
a δC2 values measured at -20 °C in separate experiments.
signals at low temperature upon addition of MeOH and Et3N
to 18a. However, upon warming of the reaction mixture to -20
°C, the 13C1 signals indicated formation of the â-1,2-anhydro-
pyranoside 17a (δC1 78.863, Scheme 12, Figure 3i) as the
principal carbohydrate intermediate. When the same experiment
was conducted with 61% 18O-labeled DBT16/18O, the expected
∆δC1 (0.037 ppm upfield) was detected at -20 °C for the â-1,2-
anhydropyranoside 17a (Figure 3ii). These observations establish
that a glycosyl intermediate such as 18a, incorporating a C1-O
linkage, is formed prior to the â-1,2-anhydropyranoside.
The complementary experiments employing the 13C2-labeled
glucal 3b were also performed to further characterize the initial
activated glycosyl intermediate in this transformation (Scheme
13). Treatment of glucal 3b with the reagent combination of
61% 18O-labeled DBT16/18O and Tf2O as above led to the
formation of the glycosyl intermediate with only a single 13C2
resonance at δ 66.584 at -20 °C, indicating the absence of a
13C2-16/18O linkage (18b, Scheme 13, Figure 4). Following the
addition of MeOH and Et3N, two 13C2 resonances were observed
(δ 54.423 and δ 54.379; ∆δC2 ) 0.044 ppm), consistent with
the 13C2-16/18O linkage present in the â-1,2-anhydropyranoside
17b.
These results for the DBTO-mediated C2-hydroxymannosyl-
ation are consistent with a reaction pathway involving initial
transfer of the sulfoxide oxygen to the C1-position of the glucal
donor, followed by oxygen migration from C1 to C2 via epoxide
closure. A hypothesis consistent with the findings in Schemes
12 and 13 invokes the intermediacy of a â-glycosyl oxosulfo-
nium intermediate 18a/b as the activated glycosyl intermediate,
incorporating an R-C2-sulfonium moiety.32 Such an intermediate
could arise from the reagent combination of DBTO and Tf2O,
assuming a sterically favored electrophilic activation of C2 from
Conclusion
The C2-hydroxyglycosylation reaction employing the reagent
combination of a diaryl sulfoxide and triflic anhydride provides
a novel process for oxygen transfer to the enol ether functionality
of glycal donors. Initial 18O-labeling studies have established
that the C2-hydroxyglucosylation process employing Ph2SO‚
Tf2O proceeds by the net transfer of the sulfoxide oxygen atom
to the C2-position of the glucal, a process that also involves
the in situ generation of a glucosyl-1,2-anhydropyranoside
intermediate. Tracking of the C2-hydroxyglucosylation reaction
by low-temperature NMR spectroscopy employing 13C-labeled
glycal donors and 18O-labeled diphenyl sulfoxide clearly showed
that formation of the intermediate R-1,2-anhydropyranoside 5
proceeds by initial formation of a glycosyl species incorporating
a C1-O linkage followed by oxygen transfer to C2 by way of
oxirane ring closure. Further spectroscopic characterization of
the activated glycosyl species in this transformation is consistent
with the intermediacy of a C2-sulfonium-C1-oxosulfonium
glycosyl species 8 and a C2-sulfonium-C1-σ-sulfurane species
9 as precursors to the glucosyl-1,2-anhydropyranoside 5. Analo-
gous investigations into the C2-hydroxymannosylation reaction
(32) Detailed NMR structural characterization of the putative anomeric oxo-
sulfonium intermediate 18 was, unfortunately, not possible given the lability
of the intermediate at temperatures above -40 °C. At lower temperatures,
1
proton-proton coupling constants could not be resolved; only a JC1H1
coupling constant of 184 Hz could be extracted from the NMR data at
1
-78 °C. Although this magnitude of JC1H1 may be indicative of an
equatorial C1-H1 linkage in 18a, the absence of additional information
makes it difficult to secure the structural and conformational properties of
18 beyond its C1-O connectivity.
9
7350 J. AM. CHEM. SOC. VOL. 124, NO. 25, 2002

Probing the Mechanism of C2-Hydroxyglycosylation
A R T I C L E S
hexane) to afford O-methyl [methyl 4-O-acetyl-3-O-(tert-butyldimeth-
ylsilyl)-â-D-glucuronyl-(1f6)]-2,3,4-tri-O-benzyl-R-D-glucopyrano-
side (63 mg, 78%): colorless oil, Rf ) 0.64 (50% ethyl acetate in
employing DBTO‚Tf2O revealed a similar C1-to-C2 oxygen-
transfer process to afford a mannosyl-1,2-anhydropyranoside
intermediate 17. These findings offer critical mechanistic
insights into a new method for enol ether oxidation in the context
of C2-hydroxyglycosylation.
1
hexane); H NMR (500 MHz, CDCl3) δ 7.37-7.27 (m, 15H, ArH),
4.99 (d, 1H, J ) 10.9 Hz, benzylic-H), 4.94 (dd, 1H, J ) 9.2, 9.7 Hz),
4.90 (d, 1H, J ) 10.9 Hz, benzylic-H), 4.81 (d, 1H, J ) 10.5 Hz,
benzylic-H), 4.79 (d, 1H, J ) 12.2 Hz, benzylic-H), 4.66 (d, 1H, J )
12.4 Hz, benzylic-H), 4.59 (d, 1H, J ) 10.7 Hz, benzylic-H), 4.58 (d,
1H, J ) 3.4 Hz, H-1), 4.19 (d, 1H, J ) 7.9 Hz, H-1′), 4.11 (dd, 1H, J
) 2.4, 11.4 Hz), 3.99 (t, 1H, J ) 9.2 Hz), 3.82-3.78 (m, 1H), 3.78 (d,
1H, J ) 10.1 Hz), 3.70 (s, 3H, CO2Me), 3.67-3.63 (m, 2H), 3.51 (dd,
1H, J ) 3.4, 9.6 Hz), 3.48-3.44 (m, 2H), 3.37 (s, 3H, OMe), 2.38 (d,
1H, J ) 2.6 Hz, OH), 2.05 (s, 3H, OCOMe), 0.84 (s, 9H, tBu), 0.09
(s, 3H, SiMe), 0.053 (s, 3H, SiMe); 13C NMR (126 MHz, CDCl3) δ
169.69, 167.95, 138.63, 138.21, 138.03, 128.50, 128.46, 128.42, 128.16,
128.01, 127.98, 127.81, 127.80, 127.67, 103.26, 98.04, 81.95, 79.69,
77.99, 75.78, 74.99, 74.47, 73.90, 73.37, 72.96, 71.98, 69.71, 68.94,
55.29, 52.66, 25.61, 20.85, 18.04, -4.35, -4.90; FTIR (neat film) 3486,
2930, 2362, 1750, 1454, 1372, 1297, 1232, 1030, 839 780, 737, 698
cm-1; HRMS (FAB) m/z calcd for C43H58O13SiNa (M + Na) 833.3547,
found 833.3544.
Experimental Section
General Procedures. All reactions were performed in flame-dried
modified Schlenk (Kjeldahl shape) flasks fitted with a glass stopper or
rubber septa under a positive pressure of argon, unless otherwise noted.
Low-temperature NMR experiments were performed in 5-mm NMR
tubes (dried under a stream of N2 gas) topped with rubber septa under
a positive pressure of argon. Air- and moisture-sensitive liquids and
solutions were transferred via syringe or stainless steel cannula. Organic
solutions were concentrated by rotary evaporation below 30 °C at ca.
25 Torr. Flash column chromatography was performed employing 230-
400-mesh silica gel. Thin-layer chromatography (analytical and pre-
parative) was performed using glass plates precoated to a depth of 0.25
mm with 230-400-mesh silica gel impregnated with a fluorescent
indicator (254 nm).
3,4,6-Tri-O-benzyl-â-D-glucopyranosyl azide (4).34 To the solution
of 3,4,6-tri-O-benzyl-D-glucal (3) (50 mg, 0.12 mmol, 1 equiv), diphenyl
sulfoxide (73 mg, 0.36 mmol, 3 equiv), and 2,4,6-tri-tert-butylpyridine
(104 mg, 0.42 mmol, 3.5 equiv) in dichloromethane (3.3 mL) was added
trifluoromethanesulfonic anhydride (30 µL, 0.18 mmol, 1.5 equiv) at
-78 °C. The resulting solution was stirred at -78 °C for 30 min, and
then at -40 °C for 1 h. Methyl alcohol (4.9 µL, 0.12 mmol, 1 equiv)
and triethylamine (50 µL, 0.36 mmol, 3 equiv) were added sequentially
at -40 °C. The solution was stirred at this temperature for 0.5 h, at 0
°C for 0.5 h, and finally at 23 °C for 1 h. The solvent was removed
under reduced pressure at 0 °C, and the residue was then dissolved in
dry CH3CN (5 mL). NaN3 (23 mg, 0.35 mmol, 3 equiv) and LiClO4
(63 mg, 0.60 mmol, 5 equiv) were added sequentially to the reaction
mixture at 0 °C, and the resulting suspension was stirred at 0 °C for
30 min and then at 23 °C for 12 h. The reaction mixture was partitioned
between dichloromethane (20 mL) and H2O (20 mL), and the aqueous
layer was further extracted with dichloromethane (3 × 10 mL). The
combined organic layers were dried (Na2SO4) and concentrated, and
the residue was purified by flash column chromatography (17% ethyl
acetate in hexane) to afford 3,4,6-tri-O-benzyl-â-D-glucopyranosyl azide
(4) (44 mg, 77%) as a white solid (from ethyl acetate-hexane): mp
Materials. Dichloromethane, diisopropylethylamine, triethylamine,
and acetonitrile were distilled from calcium hydride at 760 Torr. CD2-
Cl2 was stored over CaH2 and vacuum transferred immediately prior
to use. Methyl alcohol was distilled from Mg/I2. 18O-labeled diphenyl
sulfoxide (Ph2S18O) and 18O-labeled dibenzothiophene oxide (DBT18O)
were prepared following literature procedures.33 Trifluoromethane-
sulfonic anhydride (Tf2O) was triply distilled from phosphorous
pentoxide.
Instrumentation. Infrared (IR) spectra were obtained using a Perkin-
Elmer Spectrum BX spectrophotometer referenced to polystyrene
standard. Data are presented as frequency of absorption (cm-1). Proton
and carbon-13 nuclear magnetic resonance (1H NMR or 13C NMR)
spectra were recorded on Varian 500 and Varian Inova 500 NMR
spectrometers; chemical shifts are expressed in parts per million (δ
scale) downfield from tetramethylsilane and are referenced to residual
protium in the NMR solvent (CHCl3, δH 7.26; CHDCl2, δH 5.32; CDCl3,
δC 77.0; CD2Cl2, δC 53.8). Data are presented as follows: chemical
shift, multiplicity (s ) singlet, d ) doublet, t ) triplet, m ) multiplet
and/or multiple resonances), integration, coupling constant in hertz and
assignment. Melting points were recorded with a Fisher melting point
apparatus and are uncorrected.
1
63-64 °C; Rf ) 0.60 (50% ethyl acetate in hexane); H NMR (500
Typical Procedure for C2-Hydroxyglycosylation: O-Methyl
[Methyl 4-O-acetyl-3-O-(tert-butyldimethylsilyl)-â-D-glucuronyl-
(1f6)]-2,3,4-tri-O-benzyl-r-D-glucopyranoside. Trifluoromethane-
sulfonic anhydride (25 µL, 0.15 mmol, 1.5 equiv) was added to a
solution of methyl 4-O-acetyl-3-O-(tert-butyldimethylsilyl)-glucuronate-
D-glycal (33 mg, 0.1 mmol, 1 equiv), diphenyl sulfoxide (61 mg, 0.3
mmol, 3 equiv), and 2,4,6-tri-tert-butylpyridine (86 mg, 0.35 mmol,
3.5 equiv) in dichloromethane (3 mL) at -78 °C. The resulting solution
was stirred at -78 °C for 30 min and then at -40 °C for 1 h. Methyl
alcohol (4.2 µL, 0.1 mmol, 1 equiv) and triethylamine (42 µL, 0.3 mmol,
3 equiv) were added sequentially at -40 °C. The solution was stirred
at this temperature for 30 min and then at 0 °C for 1 h and at 23 °C for
1 h. A solution of methyl 2,3,4-tri-O-benzyl-R-D-glucopyranoside (139
mg, 0.3 mmol, 3 equiv) in dichloromethane (3 mL) was then added,
and the solution was cooled to -78 °C. Zinc chloride (1.0 M in ether,
200 µL, 0.2 mmol, 2 equiv) was added, and the mixture was stirred at
-78 °C for 30 min, at -40 °C for 1 h, at 0 °C for 1 h, and finally at
23 °C for 2 h. The reaction mixture was partitioned between
dichloromethane (20 mL) and H2O (20 mL), and the aqueous layer
was further extracted with dichloromethane (2 × 10 mL). The combined
organic layers were dried (Na2SO4) and concentrated, and the residue
was purified by flash column chromatography (20% ethyl acetate in
MHz, CDCl3) δ 7.37-7.28 (m, 13H, ArH), 7.18-7.17 (m, 2H, ArH),
4.89 (d, 1H, J ) 11.4 Hz, benzylic-H), 4.84 (d, 1H, J ) 11.2 Hz,
benzylic-H), 4.82 (d, 1H, J ) 10.7 Hz, benzylic-H), 4.63 (d, 1H, J )
12.2 Hz, benzylic-H), 4.56 (d, 1H, J ) 10.7 Hz, benzylic-H), 4.55 (d,
1H, J ) 12.2 Hz, benzylic-H), 4.53 (d, 1H, J ) 8.6 Hz, H-1), 3.75
(dd, 1H, J ) 2.6, 11.4 Hz, H-6), 3.73 (dd, 1H, J ) 3.8, 11.2 Hz, H-6),
3.67 (dd, 1H, J ) 9.0, 9.6 Hz, H-4), 3.55 (t, 1H, J ) 9.0 Hz, H-3),
3.54 (ddd, 1H, J ) 2.4, 3.9, 9.9 Hz, H-5), 3.45 (dt, 1H, 2.6, 8.8 Hz,
H-2), 2.24 (d, 1H, J ) 2.6 Hz, OH); MS (FAB) m/z (relative intensity,
%) 498.2 (M + Na, 27), 472.3 (16), 177.1 (100); HRMS (FAB) m/z
calcd for C27H29N3O5Na (M + Na) 498.2007, found 498.2005.
13C NMR Detection and Characterization of 8, 9, and 5. To a
solution of Ph2S16/18O (15 mg, 0.072 mmol, 3 equiv), 2,4,6-tri-tert-
butylpyridine (3 mg, 0.013 mmol, 0.5 equiv), and 3,4,6-tri-O-benzyl-
D-glucal (3a or 3b) (10 mg, 0.024 mmol, 1 equiv) in CD2Cl2 (0.66
mL) in a 5-mm NMR tube was added trifluoromethanesulfonic
anhydride (6 µL, 0.036 mmol, 3 equiv) at -78 °C. The reaction mixture
was briefly agitated (3 × 5 s; Fisher Vortex Genie 2) and placed in the
NMR probe at -78 °C and then warmed to -40 °C for the 13C NMR
characterization of 8a/b. The NMR sample was ejected from the NMR
probe and immediately cooled to -78 °C. To the reaction mixture was
(33) Sato, Y.; Kunieda, N.; Kinoshita, M. Chem. Lett. 1972, 1023-1025.
(34) Gordon, D. M.; Danishefsky, S. J. Carbohydr. Res. 1990, 206, 361-366.
9
J. AM. CHEM. SOC. VOL. 124, NO. 25, 2002 7351

A R T I C L E S
Honda and Gin
added methyl alcohol (1 µL, 0.012 mmol, 1 equiv) at -78 °C, and the
resulting mixture in the NMR tube was briefly agitated (3 × 5 s).
Triethylamine (10 µL, 0.072 mmol, 3 equiv) was then added at -78
°C, and the resulting mixture in the NMR tube was again briefly agitated
(3 × 5 s). The NMR reaction mixture was placed in the NMR probe
which had been recooled to -78 °C, and the reaction mixture was then
warmed to -20 °C for the 13C NMR characterization of 9a/b and 5a/b.
1H NMR Characterization of d21-8 and d21-9. To a solution
of Ph2SO (15 mg, 0.072 mmol, 3 equiv), 2,4,6-tri-tert-butylpyridine
(3 mg, 0.013 mmol, 0.5 equiv), and 3,4,6-tri-O-d7-benzyl-D-glucal
(d21-3) (10.5 mg, 0.024 mmol, 1 equiv) in CD2Cl2 (0.66 mL) in a
5-mm NMR tube was added trifluoromethanesulfonic anhydride (6 µL,
0.036 mmol, 3 equiv) at -78 °C. The reaction mixture in the NMR
tube was briefly agitated (3 × 5 s; Fisher Vortex Genie 2) and placed
NMR probe at -78 °C and then warmed to -20 °C for the 13C NMR
characterization of 18a/b.
13C NMR Characterization of 17a/b. To a solution of DBT16/18
O
(14 mg, 0.072 mmol, 3 equiv), 2,4,6-tri-tert-butylpyridine (3 mg, 0.013
mmol, 0.5 equiv), and 3,4,6-tri-O-benzyl-D-glucal (3a, 3b) (10 mg,
0.024 mmol, 1 equiv) in CD2Cl2 (0.66 mL) in a 5-mm NMR tube was
added trifluoromethanesulfonic anhydride (6 µL, 0.036 mmol, 3 equiv)
at -78 °C. The reaction mixture in the NMR tube was briefly agitated
(3 × 5 s; Fisher Vortex Genie 2) and placed in the NMR probe at -78
°C. After confirmation of the generation of 18a/b at -78 °C by 13C
NMR, the NMR sample was ejected from the NMR probe and
immediately cooled at -78 °C. To the reaction mixture was added
methyl alcohol (1 µL, 0.012 mmol, 1 equiv) at -78 °C, and the resulting
mixture was briefly agitated (3 × 5 s). Triethylamine (10 µL, 0.072
mmol, 3 equiv) was then added at -78 °C, and the resulting mixture
in the NMR tube was again briefly agitated (3 × 5 s). The NMR
reaction mixture was placed in the NMR probe at -78 °C, and the
reaction mixture was then warmed to -20 °C for the 13C NMR
characterization of 17a/b.
1
in the NMR probe at -40 °C and then warmed to -20 °C for the H
NMR characterization of d21-8. The NMR sample was ejected from
the NMR probe and immediately cooled to -78 °C. To the reaction
mixture was added methyl alcohol (1 µL, 0.012 mmol, 1 equiv) at
-78 °C, and the resulting mixture was briefly agitated (3 × 5 s). Tri-
ethylamine (10 µL, 0.072 mmol, 3 equiv) was then added at -78 °C,
and the resulting mixture in the NMR tube was again briefly agitated
(3 × 5 s). The NMR reaction mixture was placed in the NMR probe
which had been recooled to -40 °C, and the reaction mixture was then
Acknowledgment. This research was supported by the
National Institutes of Health (GM-58833) and Johnson &
Johnson. D.Y.G. is a Cottrell Scholar of Research Corporation.
NMR spectra were obtained in the Varian Oxford Instrument
Center for Excellence in NMR Laboratory, funded in part by
the W.M. Keck Foundation, the National Institutes of Health
(PHS 1 S10 RR10444), and the National Science Foundation
(NSF CHE 96-10502). The invaluable assistance of Dr. Vera
Mainz of the VOICE NMR Spectroscopy Lab at the University
of Illinois is gratefully recognized. We thank Professor Anthony
Serianni and Omicron Biochemicals, Inc., for generous gifts of
13C1- and 13C2-labeled D-glucose (14a/b).
1
warmed to -20 °C for the H NMR characterization of d21-9. d21-8:
1H NMR (500 MHz, CD2Cl2, -20 °C) δ 6.96 (d, 1H, J ) 5.5 Hz,
H-1), 6.12 (dd, 1H, J ) 3.2, 5.9 Hz, H-2), 3.97 (dd, 1H, J ) 2.3, 9.1
Hz, H-4), 3.81 (br d, 1H, J ) 8.8 Hz, H-5), 3.68 (br s, 1H, H-3), 3.22
(dd, 1H, J ) 3.4, 11.3 Hz, H-6), 2.46 (br d, 1H, J ) 10.7 Hz, H-6).
d21-9: 1H NMR (500 MHz, CD2Cl2, -20 °C) δ 5.55-5.54 (br m, 1H,
H-2), 4.96 (br s, 1H, H-1), 4.61 (dd, 1H, J ) 4.1, 8.8 Hz, H-3), 4.33
(t, 1H, J ) 9.1 Hz, H-4), 4.08 (br dt, 1H, J ) 9.7, 2.3 Hz, H-5), 3.81
(dd, 1H, J ) 3.3, 10.5 Hz, H-6), 3.61 (br dd, 1H, J ) 2.0, 10.8 Hz,
H-6).
13C NMR Detection and Characterization of 18a/b. To a solution
of DBT16/18O (14 mg, 0.072 mmol, 3 equiv), 2,4,6-tri-tert-butylpyridine
(3 mg, 0.013 mmol, 0.5 equiv), and 3,4,6-tri-O-benzyl-D-glucal (3a,
3b) (10 mg, 0.024 mmol, 1 equiv) in CD2Cl2 (0.66 mL) in a 5-mm
NMR tube was added trifluoromethanesulfonic anhydride (6 µL, 0.036
mmol, 3 equiv) at -78 °C. The reaction mixture in the NMR tube was
briefly agitated (3 × 5 s; Fisher Vortex Genie 2) and placed in the
Supporting Information Available: Analytical spectral data
for the C(2)-hydroxyglycoconjugates and preparative procedures
for the 13C- and d-labeled glucal donors (PDF). This material
is available free of charge via the Internet at http://pubs.acs.org.
JA025639C
9
7352 J. AM. CHEM. SOC. VOL. 124, NO. 25, 2002