β-directing effect of electron-withdrawing groups at O-3, O-4, and O-6 positions and α-directing effect by remote participation of 3-O-acyl and 6-O-acetyl groups of donors in mannopyranosylations

Article abstract of DOI:10.1021/ja907252u

Mannosylations of various acceptors with donors possessing an electron-withdrawing o-trifluoromethylbenzenesulfonyl, benzylsulfonyl, p-nitrobenzoyl, benzoyl, or acetyl group at O-3, O-4, or O-6 positions were found to be β-selective except when donors had

Full text of DOI:10.1021/ja907252u

Published on Web 11/12/2009
ꢀ-Directing Effect of Electron-Withdrawing Groups at O-3, O-4,
and O-6 Positions and r-Directing Effect by Remote
Participation of 3-O-Acyl and 6-O-Acetyl Groups of Donors in
Mannopyranosylations
Ju Yuel Baek, Bo-Young Lee, Myung Gi Jo, and Kwan Soo Kim*
Center for BioactiVe Molecular Hybrids and the Department of Chemistry, Yonsei UniVersity,
Seoul 120-749, Korea
Received August 27, 2009; E-mail: kwan@yonsei.ac.kr
Abstract: Mannosylations of various acceptors with donors possessing an electron-withdrawing o-
trifluoromethylbenzenesulfonyl, benzylsulfonyl, p-nitrobenzoyl, benzoyl, or acetyl group at O-3, O-4, or O-6
positions were found to be ꢀ-selective except when donors had 3-O-acyl and 6-O-acetyl groups, which
afforded R-mannosides as major products. The R-directing effect of 3-O-acyl and 6-O-acetyl groups was
attributed to their remote participation, and the isolation of a stable bicyclic trichlorooxazine ring resulting
from the intramolecular trapping of the anomeric oxocarbenium ion by 3-O-trichloroacetimidoyl group
provided evidence for this remote participation. The triflate anion, counteranion of the mannosyl
oxocarbenium ion, was essential for the ꢀ-selectivity, and covalent R-mannosyl triflates with an electron-
withdrawing group at O-3, O-4, or O-6 were detected by low-temperature NMR. The strongly electron-
withdrawing sulfonyl groups, which exhibited a higher ꢀ-directing effect in the mannosylation, made the
R-mannosyl triflates more stable than the weakly electron-withdrawing acyl groups. We therefore proposed
the mechanism for the ꢀ-mannosylation and the origin of the ꢀ-directing effect: the electron-withdrawing
groups would stabilize the R-mannosyl triflate intermediate, and the subsequent reaction of the R-triflate
(or its contact ion pair) with the acceptor would afford the ꢀ-mannoside. The ꢀ-selective mannosylation of
a sterically demanding acceptor was achieved by employing a donor possessing two strongly electron-
withdrawing benzylsulfonyl groups at O-4 and O-6 positions.
strategies for ꢀ-mannopyranosylation have been developed,4
Introduction
including a method using mannosyl halides with silver salt
promoters5 and the indirect intramolecular aglycon delivery
approach.6 Recently, Crich and co-workers have made a
significant breakthrough in ꢀ-mannoside synthesis by employing
4,6-O-benzylidene-protected mannopyranosyl sulfoxides or thio-
mannopyranosides as mannosyl donors.7 The 4,6-O-acetal effect
on the construction of ꢀ-mannosyl linkages was further dem-
onstrated with other mannosyl donors by us8,9 and other
The development of stereoselective glycosylation methodolo-
gies has attracted a great deal of attention over the past decade
due to the important roles of complex oligosaccharides in many
fundamental life-sustaining processes.1 Although there are many
methods for the stereoselective formation of glycosyl linkages,2
the stereoselective construction of the 1,2-cis-ꢀ-D-mannopyra-
nosyl linkage still poses a great challenge, while the formation
of the 1,2-trans-R-mannosyl bond is relatively straightforward,
utilizing neighboring group participation.3 Several innovative
(4) For reviews on the ꢀ-mannosylation, see: (a) Barresi, F.; Hindsgaul,
O. In Modern Methods in Carbohydrate Synthesis; Khan, S. H., O’Neil,
R. A., Eds.; Harwood Academic Publishers: Amsterdam, 1996; pp
251-276. (b) Gridley, J. J.; Osborn, H. M. I. J. Chem. Soc., Perkin
Trans. 1 2000, 1471–1491. (c) Demchenko, A. V. Curr. Org. Chem.
2003, 7, 35–79.
(1) (a) Varki, A. Glycobiology 1993, 3, 97–130. (b) Dwek, R. A. Chem.
ReV. 1996, 96, 683–720. (c) Bertozzi, C. R.; Kiessling, L. L. Science
2001, 291, 2357–2364.
(2) For reviews on the glycosylation, see: (a) Toshima, K.; Tatsuta, K.
Chem. ReV. 1993, 93, 1503–1531. (b) Boons, G.-J. Tetrahedron 1996,
52, 1095–1121. (c) Davis, B. G. J. Chem. Soc., Perkin Trans. 1 2000,
2137–2160. (d) Ernst, B., Hart, G. W., Sinay¨, P., Eds. Carbohydrates
in Chemistry and Biology; Wiley-VCH: Weinheim, 2000; Vol. 1, pp
5-237.
(5) (a) Paulsen, H.; Lockhoff, O. Chem. Ber. 1981, 114, 3102–3114. (b)
Garegg, P. J.; Ossowski, P. Acta Chem. Scand. 1983, 337, 249–258.
(6) (a) Barresi, F.; Hindsgaul, O. J. Am. Chem. Soc. 1991, 113, 9376–
9377. (b) Stork, G.; Kim, G. J. Am. Chem. Soc. 1992, 114, 1087–
1088. (c) Ito, Y.; Ogawa, T. Angew. Chem., Int. Ed. Engl. 1994, 33,
1765–1767. (d) Ziegler, T.; Lemanski, G. Angew. Chem., Int. Ed. 1998,
37, 3129–3132.
(3) For neighboring group participation in glycosylation, see: (a) Capon,
B. Chem. ReV. 1969, 69, 407–498. (b) Nukada, T.; Berces, A.; Zgierski,
M. Z.; Whitfield, D. M. J. Am. Chem. Soc. 1998, 120, 13291–13295.
(c) Green, L. G.; Ley, S. V. In Carbohydrates in Chemistry and
Biology; Ernst, B., Hart, G. W., Sinay, P., Eds.; Wiley-VCH:
Weinheim, 2000; Vol. 1, pp 427-448. (d) Demchenko, A. V. In
Handbook of Chemical Glycosylation: AdVances in StereoselectiVity
and Therapeutic ReleVance; Demchenko, A. V., Ed.; Wiley-VCH:
Weinheim, 2008; pp 1-27.
(7) (a) Crich, D.; Sun, S. J. Org. Chem. 1996, 61, 4506–4507. (b) Crich,
D.; Sun, S. J. Org. Chem. 1997, 62, 1198–1199. (c) Crich, D.; Sun,
S. J. Am. Chem. Soc. 1998, 120, 435–436. (d) Crich, D.; Sun, S.
Tetrahedron 1998, 54, 8321–8348.
(8) (a) Kim, K. S.; Kim, J. H.; Lee, Y. J.; Park, J. J. Am. Chem. Soc.
2001, 123, 8477–8481. (b) Baek, J. Y.; Choi, T.-J.; Jeon, H. B.; Kim,
K. S. Angew. Chem., Int. Ed. 2006, 45, 7436–7440.
9
10.1021/ja907252u CCC: $40.75  2009 American Chemical Society
J. AM. CHEM. SOC. 2009, 131, 17705–17713 17705

A R T I C L E S
Baek et al.
workers.10 A combination of torsional and electronic disarming
effects was ascribed to the stereoselectivity in the 4,6-O-
benzylidene directed ꢀ-mannosylation: the benzylidene acetal
disarms the donor not only by raising the activation barrier by
opposing the flattening that is required in the oxocarbenium ion11
but also by restricting the C5-C6 bond to the most electron-
withdrawing tg conformer.12 On the other hand, Schuerch and
co-workers reported earlier that a nonparticipating, strongly
electron-withdrawing group such as the alky or aryl sulfonyl
group at O-2 in the mannosyl donor led to the preferential
formation of the ꢀ-mannopyranoside.13 Recently, the scope and
limitation of Schuerch’s selective ꢀ-mannosylation were further
investigated, and explanations for the ꢀ-selectivity were provided
by Schmidt and Crich. Schmidt and co-workers reported that
mannosylations of one acceptor with three 2-O-benzylsulfonly
mannosyl donors were ꢀ-selective and proposed that the strong
dipole produced by the O-2 sulfonyl group would favor the
formation of the mannosyl oxocarbenium ion in a twist-boat
conformation over the half-chair conformation, and that the
twist-boat oxocarbenium ion intermediate would be preferen-
tially attacked by the acceptor from the ꢀ-side.14 Crich and
colleagues reported that with 2-O-sulfonyl-thiomannosides as
donors, mannosylations of less hindered alcohols were ꢀ-selec-
tive, but those of sterically demanding glucose 4-OH gave
R-mannosides as major products. They reasoned that the
electron-withdrawing effect of the O-2 sulfonyl group would
destabilize the oxocarbenium ion, thereby shifting the equilib-
rium toward a covalent R-mannosyl triflate, which would react
with an acceptor to generate the ꢀ-mannoside.15 However, the
effect on the mannosylation stereochemistry of nonparticipating,
strongly electron-withdrawing protective groups at the O-3, O-4,
or O-6 positions of the donors has not been addressed. However,
the ꢀ-directing effects of the O-4-benzylsulfonyl group in
D-olivose (2,6-dideoxy-D-arabino-hexose) and D-oliose (2,6-
dideoxy-D-lyxo-hexose) donors,16 of the C-5-carboxylate ester
group in D-mannuronate donors,17 and of the C-6-fluorine in
6-fluoro-D-rhamnose donors18 have been investigated. The origin
of the ꢀ-selectivity was ascribed to the reaction of the glycosyl
reaction of the R-glycosyl triflate intermediates with the
acceptor, in the case of the fluoro sugars.18 In this study, we
investigated the directing effect of electron-withdrawing protect-
ing groups at O-3, O-4, and O-6 positions of mannosyl donors
on the outcome of the stereochemistry in mannosylations. We
were also interested in the magnitude of the directing effect by
each of the groups at O-3, O-4, and O-6 of mannosyl donors,
because their distances from the anomeric center are different
and all of them are in remote position from the anomeric center
as compared to groups at O-2. Ley’s work on quantification of
the deactivating effect of electron-withdrawing protective groups
at each position of thiomannoside donors in mannosylations
indicates that the electron-withdrawing benzoyl group at O-2
is more deactivating than benzoyl groups at other positions, and
thus the influence of the positions is in the order of 2 > 6 > 4
> 3.20
When the electron-withdrawing, potentially participating
protective groups are present at O-3, O-4, and O-6 of mannosyl
donors, it is difficult to distinguish the electron-withdrawing
effect from the effect of remote participation of the protective
groups on the outcome of the stereochemistry in the manno-
sylation. In mannosylations with donors possessing electron-
withdrawing, participating groups at O-3 and O-6, remote
participation and electron-withdrawing would have opposite
effects on the outcome of the stereochemistry: the R-manno-
side would be favored by the remote participation, while the
ꢀ-mannoside might be formed preferentially by the electron-
withdrawing effect, if the electron-withdrawing effect is ꢀ-di-
recting. On the other hand, in mannosylations with donors
possessing electron-withdrawing, participating groups at O-4,
both effects would contribute to the generation of the ꢀ-manno-
side if the electron-withdrawing effect is ꢀ-directing. Although
numerous examples for mannosylations with electron-withdraw-
ing, potentially participating groups at O-3, O-4, and O-6 of
donors can be found in the literature, most of them do not
provide any information on remote participation, and only a
handful of reports discuss the remote participation of the
protective groups in mannosylations. There have been reports
both opposed to and in favor of the remote participation by
O-3 acyl groups of mannosyl donors. Thus, van Boeckel and
colleagues oppose the remote participation of the 3-O-acetyl
group in mannosylations, even though mannosylation with a
donor bearing the 3-O-acetyl group afforded more R-mannoside
than that with the corresponding donor bearing the 3-O-benzyl
group. Their reasoning was that mannosylations with donors
with the 3-O-acetyl group afforded almost the same ratio of
the R- and ꢀ-mannosides as mannosylations with donors with
the 3-O-trichloroacetyl group, which they believed to be a poor
participating group.21 Reports by Crich and co-workers sup-
ported the participation by the 3-O-benzoyl and 3-O-chloroactyl
groups of mannosyl donors, because mannosylation with them
resulted in predominantly R-anomers.22 Contradictory views are
also found in discussions of the remote participation by the 4-O-
acyl and 4-O-thiocarbamoyl groups of mannosyl donors. Thus,
van Boeckel and colleagues are opposed to the remote participa-
tion of 4-O-acetyl group based on the same argument as that
for the 3-O-acetyl group.21 Demchenko and co-workers, on the
3
oxocarbenium ion intermediate of the H4 conformation with
the acceptor, in the case of the mannuronates,19 and to the
(9) Kim, K. S.; Fulse, D. B.; Baek, J. Y.; Lee, B.-Y.; Jeon, H. B. J. Am.
Chem. Soc. 2008, 130, 8537–8547.
(10) (a) Weingart, R.; Schmidt, R. R. Tetrahedron Lett. 2000, 41, 8753–
8758. (b) Tanaka, S.-i.; Takashina, M.; Tokimoto, H.; Fujimoto, Y.;
Tanaka, K.; Fukase, K. Synlett 2005, 2325–2328. (c) Code´e, J. D. C.;
Hossain, L. H.; Seeberger, P. H. Org. Lett. 2005, 7, 3251–3254.
(11) (a) Fraser-Reid, B.; Wu, Z. C.; Andrews, C. W.; Skowronski, E. J. Am.
Chem. Soc. 1991, 113, 1434–1435. (b) Andrews, C. W.; Rodebaugh,
R.; Fraser-Reid, B. J. Org. Chem. 1996, 61, 5280–5289.
(12) Jesnen, H. H.; Nordstrom, L. U.; Bols, M. J. Am. Chem. Soc. 2004,
126, 9205–9213.
(13) (a) Srivastava, V. K.; Schuerch, C. Carbohydr. Res. 1980, 79, C13–
C16. (b) Srivastava, V. K.; Schuerch, C. J. Org. Chem. 1981, 46, 1121–
1126. (c) Awad, L. F.; El Ashry, E. S. H.; Schuerch, C. Bull. Chem.
Soc. Jpn. 1986, 59, 1587–1592.
(14) Abdel-Rahman, A. A.-H.; Jonke, S.; El Ashry, E. S. H.; Schmidt, R. R.
Angew. Chem., Int. Ed. 2002, 41, 2972–2974.
(15) Crich, D.; Hutton, T. K.; Banerjee, A.; Jayalath, P.; Picione, J.
Tetrahedron: Asymmetry 2005, 16, 105–119.
(16) Tanaka, H.; Yoshizawa, A.; Takahashi, T. Angew. Chem., Int. Ed.
2007, 46, 2505–2507.
(17) van den Bos, L. J.; Dinkelaar, J.; Overkleeft, H. S.; van der Marel,
G. A. J. Am. Chem. Soc. 2006, 128, 13066–13067.
(20) Douglas, N. L.; Ley, S. V.; Lu¨cking, U.; Warreiner, S. L. J. Chem.
Soc., Perkin Trans. 1 1998, 51–65.
(18) Crich, D.; Vinogradova, O. J. Am. Chem. Soc. 2007, 129, 11756–
11765.
(21) van Boeckel, C. A. A.; Beetz, T.; van Aelst, S. F. Tetrahedron 1984,
40, 4097–4107.
(19) Code´e, J. D. C.; van den Bos, L. J.; de Jong, A.-R.; Dinkelaar, J.;
Lodder, G.; Overkleeft, H. S.; van der Marel, G. A. J. Org. Chem.
2009, 74, 38–47.
(22) (a) Crich, D.; Cai, W.; Dai, Z. J. Org. Chem. 2000, 65, 1291–1297.
(b) Crich, D.; Yao, Q. J. Am. Chem. Soc. 2004, 126, 8232–8236.
9
17706 J. AM. CHEM. SOC. VOL. 131, NO. 48, 2009

ꢀ- and R-Directing Effects in Mannopyranosylations
A R T I C L E S
benzylsulfonyl groups were selected as nonparticipating, strongly
electron-withdrawing protective groups, and p-nitrobenzoyl,
benzoyl, and acetyl groups were chosen as potentially participat-
ing, weakly electron-withdrawing protective groups. For com-
parison purposes, tetra-O-benzyl-mannosyl trichloroacetimidate
17 and tetra-O-benzyl-mannose 18 were selected as standard
donors bearing no electron-withdrawing groups. Primary hy-
droxy sugars 19, 20, and 22 and a secondary hydroxy sugar 21
were chosen as acceptors as shown in Figure 1. Mannosylations
were carried out with mannosyl trichloroacetimidates 1-5 and
9-16 using TMSOTf as a promoter26 and with anomeric
hydroxy sugars 6-8 using phthalic anhydride/Tf2O as promot-
ers.9 Here, we report the results of mannosylations of various
acceptors with donors 1-18, which support the ꢀ-directing effect
of nonparticipating, electron-withdrawing groups and the R-di-
recting effect by the remote participation of some acyl groups
at O-3 and O-6 of donors. We also detected R-mannosyl triflate
intermediates by low-temperature NMR and isolated a bridging
bicyclic compound as evidence for remote participation.
Figure 1. Mannosyl donors and acceptors.
Results and Discussion
other hand, argued that ꢀ-selective mannosylations with donors
bearing a p-methoxybenzoyl and a thiocarbamoyl groups at O-4
could be attributed to the remote participation of the groups at
O-4.23 Reports on the potential participation by O-6 protective
groups of mannosyl donors are rare, but there have been reports
both in favor of and opposed to the participation by O-6
protective groups of glucosyl and galactosyl donors.24 Recently,
Crich and co-workers performed a trapping experiment of the
bridging dioxocarbenium ion intermediates resulting from
possible neighboring group participation by nonvicinal esters
of donors in glycosylations and concluded that neighboring
group participation from 3-O-equatorial, 4-O-axial and -equato-
rial, and 6-O-positions does not occur under typical glycosy-
lation conditions.25 This result suggests that no remote partici-
pations occur by acyl groups at O-3, O-4, and O-6 of donors in
mannosylations. These results and evidence on the potential
remote participation by the electron-withdrawing groups in
mannosyl donors are contradictory, somewhat confusing, and
not at all conclusive.
The earlier reports on the ꢀ-directing effect and the potential
remote participation by electron-withdrawing groups of donors
on the mannosylation stereochemistry suggest the following
questions: (1) Do electron-withdrawing protective groups at O-3,
O-4, and O-6 of donors also enhance the ꢀ-selectivity in the
mannosylation like those at O-2? (2) If so, what makes the
electron-withdrawing groups facilitate the formation of ꢀ-manno-
sides and what is the reactive intermediate? (3) Are acyl
protective groups at O-3, O-4, and O-6 of donors involved in
the remote participation in mannosylations? To address these
questions, a systematic study was required, which could also
lead to the development of a new method for the synthesis of
ꢀ-mannosides. Therefore, we designed mannosyl donors 1-16,
bearing a nonparticipating or potentially participating, electron-
withdrawing protective group at the O-3, O-4, or O-6 positions,
as shown in Figure 1. o-Trifluoromethylbenzenesulfonyl and
Synthesis of Mannosyl Donors and Mannosylation Condi-
tions. Introduction of electron-withdrawing groups (EWGs) to
donors was achieved by sulfonylation and acylation of the
hydroxy group at the O-3, O-4, or O-6 position of allyl tri-O-
benzyl-mannosides. The resulting fully protected allyl manno-
sides were converted into mannosyl trichloroacetimidates 1-5
and 9-16 by deallylation with PdCl2 in methanol, followed by
the reaction of the resultant anomeric hydroxy sugars with
trichloroacetonitrile in the presence of a base. Anomeric hydroxy
sugar donors 6-8 were obtained by deallylation of the corre-
sponding fully protected ally mannosides with PdCl2 in methanol
(see the Supporting Information).
For mannosylation with trichloroacetimidate donors, the
coupling was carried out by addition of TMSOTf (0.15 equiv)
to a solution of the acceptor (1.0 equiv) and of the donor (1.5
equiv) in CH2Cl2 at -78 °C, followed by stirring the solution
for 2 h at -78 °C. The reaction was quenched by addition of
triethylamine at -78 °C. The amount of TMSOTf (0.15 or 1.0
equiv) did not significantly affect the yield and the ꢀ/R ratio of
the mannosylation product, although the reaction was a little
faster with 1.0 equiv of TMSOTf.
Mannosylation with anomeric hydroxy sugar donors was
carried out in the following one-pot sequence: (i) reaction of
the donor (1.0 equiv), phthalic anhydride (1.1 equiv), and 1,8-
diazobicyclo[5,4,0]undec-7-ene (DBU, 1.2 equiv) in the pres-
ence of 4A molecular sieves for 15 min at room temperature in
CH2Cl2; (ii) addition of Tf2O (1.5 equiv) to this solution at -78
°C and stirring for further 15 min; and (iii) addition of the
acceptor (1.5 equiv) at -78 °C and allowing the reaction mixture
to warm over 1 h to 0 °C. The reaction was quenched with
saturated aqueous NaHCO3 at 0 °C.
Mannosylations with Donors Possessing an Electron-Withdraw-
ing Group at O-4. Mannosylations of benzyl-protected primary
hydroxy sugar 19 with mannosyl donors with strongly electron-
withdrawing sulfonyl groups at O-4, including 4-O-(o-trifluo-
(23) De Meo, C.; Kamat, M. N.; Demchenko, A. V. Eur. J. Org. Chem.
2005, 706–711.
(26) For reviews for glycosyl trichloroacetimidates, see: (a) Zhu, X.;
Schmidt, R. R. In Handbook of Chemical Glycosylation: AdVances in
StereoselectiVity and Therapeutic ReleVance; Demchenko, A. V., Ed.;
Wiley-VCH: Weinheim, 2008; pp 143-185. (b) Schmidt, R. R.; Zhu,
X. In Glycoscience: Chemistry and Chemical Biology, 2nd ed.; Fraser-
Reid, B. O., Tatsuta, K., Thiem, J., Eds.; Springer-Verlag: Berlin, 2008;
pp 452-524.
(24) (a) Eby, R.; Schuerch, C. Carbohydr. Res. 1974, 34, 79–90. (b) Lin,
C.-C.; Shimazaki, M.; Heck, M.-P.; Aoki, S.; Wang, R.; Kimura, T.;
Ritze`n, H.; Takayama, S.; Wu, S.-H.; Weitz-Schmidt, G.; Wong, C.-
H. J. Am. Chem. Soc. 1996, 118, 6826–6840. (c) Mukaiyama, T.;
Suenaga, M.; Chiba, H.; Jona, H. Chem. Lett. 2002, 56–57.
(25) Crich, D.; Hu, T.; Cai, F. J. Org. Chem. 2008, 73, 8942–8953.
9
J. AM. CHEM. SOC. VOL. 131, NO. 48, 2009 17707

A R T I C L E S
Baek et al.
Table 1. Mannosylations with Trichloroacetimidate Donors 1-5
Having EWGs at O-4
Table 2. Mannosylations with Anomeric Hydroxy Sugar Donors
6-8 Having EWGs at O-4
entry acceptor ROH donor
EWG
product (yield, %)a ratiob ꢀ/R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
19
19
19
19
19
19
20
20
20
20
20
20
21
21
21
21
21
1
2
3
4
5
17
1
2
3
4
5
17
2
3
4
5
17
SO2C6H4o-CF3
SO2Bn
Bzp-NO2
Bz
23 (81)
24 (80)
25 (91)
26 (73)
27 (85)
28 (91)
29 (81)
30 (86)
31 (95)
32 (79)
33 (87)
34 (96)
35 (80)
36 (82)
37 (85)
38 (81)
39 (87)
15.5/1
10.7/1
7.2/1
7.1/1
4.0/1
2.7/1
7.2/1
5.0/1
5.0/1
4.6/1
4.1/1
2.5/1
3.5/1
2.9/1
2.3/1
2.1/1a
1/2.7
entry
acceptor ROH
donor
EWG
product (yield, %)a
ratiob ꢀ/R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
19
19
19
20
20
20
20
21
21
21
21
22
22
22
6
7
8
6
7
8
18
6
7
8
18
6
7
8
SO2Bn
Bzp-NO2
Bz
24 (85)
25 (88)
26 (89)
30 (88)
31 (89)
32 (88)
34 (86)
35 (87)
36 (88)
37 (85)
39 (81)
40 (90)
41 (85)
42 (89)
9.2/1
4.7/1
6.1/1
4.9/1
3.0/1
3.6/1
1.6/1
3.1/1
1.5/1
1.3/1
R onlya
7.3/1
2.8/1
3.0/1
Ac
SO2Bn
Bzp-NO2
Bz
(Bn)c
SO2C6H4o-CF3
SO2Bn
Bzp-NO2
Bz
(Bn)c
SO2Bn
Bzp-NO2
Bz
Ac
(Bn)c
(Bn)c
SO2Bn
Bzp-NO2
Bz
SO2Bn
Bzp-NO2
Bz
Ac
(Bn)c
a Determined after isolation. b The ratio was determined by
LC-mass. c A standard donor for comparison.
a Determined after isolation. b The ratio was determined by LC-mass.
c A standard donor for comparison.
groups at O-4. The results shown in Table 2 were quite similar
to those with the trichloroacetimidate donors 2-4. Thus, the
benzylsulfonyl and acyl groups of donors 6-8 exhibited the
ꢀ-directing effect in mannosylations of primary hydroxy sugar
acceptors 19 and 20, and the effect of the benzylsulfonyl group
was stronger than those of acyl groups (entries 1-7 in Table
2). Although mannosylations of secondary hydroxy sugar
acceptor 21 with 6-8, on the other hand, exhibited poor
ꢀ-selectivity (entries 8-10), the ꢀ-directing effects of the
sulfonyl and acyl groups contrasted to the R only yield of
disaccharide 39 obtained from the mannosylation of 21 with
tetra-O-benzyl-mannose 18 (entry 11). Mannosylations of ben-
zoyl-protected primary hydroxy sugar acceptor 22 with donors
6-8 were also ꢀ-selective (entries 12-14).
All electron-withdrawing protecting groups, SO2C6H4o-CF3,
SO2Bn, p-NO2Bz, Bz, and Ac, at O-4 of donors promoted
ꢀ-selective mannosylation, and the magnitude and the trend of
their ꢀ-directing effect were not much affected by different
leaving groups at the anomeric center. It was also noteworthy
that the more strongly electron-withdrawing groups, SO2C6H4o-
CF3 and SO2Bn, exhibited a somewhat higher ꢀ-directing effect
than did the weakly electron-withdrawing acyl groups, p-NO2Bz,
Bz, and Ac.27 As compared to the ꢀ-directing effect of electron-
withdrawing groups at O-2,13-15 the ꢀ-directing effect of those
at O-4 was not smaller and even greater in certain cases. The
stereochemistries at the newly generated anomeric centers of
the product disaccharides were unequivocally determined on
the basis of characteristic one-bond C1′-H1′ coupling con-
stants.28
romethylbenzenesulfonyl)-tri-O-benzyl-mannosyl trichloroace-
timidate, 1, and 4-O-benzylsulfonyl-tri-O-benzyl-mannosyl
trichloroacetimidate, 2, were highly ꢀ-selective, yielding man-
nosyl disaccharides 23 (ꢀ/R ) 15.5:1) and 24 (ꢀ/R ) 10.7:1),
respectively, with a large excess of ꢀ-anomers in high yields
(entries 1 and 2 in Table 1). The acyl groups at O-4 of mannosyl
trichloroacetimidate donors 3-5 also made the mannosylation
of 19 highly ꢀ-selective, although less pronounced than the 4-O-
sulfonyl groups. Thus, mannosylations of 19 with 4-O-p-
nitrobenzoyl-mannosyl imidate, 3, 4-O-benzoyl-mannosyl im-
idate, 4, and 4-O-acetyl-mannosyl imidate, 5, provided mannosyl
disaccharides, 25 (ꢀ/R ) 7.2:1), 26 (ꢀ/R ) 7.1:1), and 27 (ꢀ/R
) 4.0:1), respectively, favoring ꢀ-anomers (entries 3-5). For
comparison, we carried out the mannosylation of 19 with
2,3,4,6-tetra-O-benzyl-mannosyl trichloroacetimidate, 17, pos-
sessing no electron-withdrawing protective groups, and obtained
a mixture of R- and ꢀ-mannosyl disaccharide, 28 (ꢀ/R ) 2.7:1)
(entry 6). The ꢀ-directing effect of the 4-O-sulfonyl group was
confirmed in the mannosylation of a different acceptor, benzoyl-
protected glucose acceptor, 20. Thus, mannosylations of 20 with
1 and 2 provided mannosyl disaccharides 29 (ꢀ/R ) 7.2:1) and
30 (ꢀ/R ) 5.0:1), respectively, with an excess of ꢀ-anomers
(entries 7 and 8). With 4-O-acyl-mannosyl imidate donors 3-5,
mannosylations of 20 were also ꢀ-selective, giving 31 (EWG
) p-NO2Bz, ꢀ/R ) 5.0:1), 32 (EWG ) Bz, ꢀ/R ) 4.6:1), and
33 (EWG ) Ac, ꢀ/R ) 4.1:1), respectively, in high yields
(entries 9-11). In the mannosylation of secondary hydroxy
acceptor 21 with mannosyl imidate donors 2-5, the ꢀ/R ratios
of the products decreased (entries 13-16) as compared to those
in the mannosylations of primary hydroxy acceptors. Neverthe-
less, when their ꢀ/R ratios were compared to the ꢀ/R ratio, 1/2.7,
of product 39 (no EWG) from the mannosylation of 21 with
fully benzyl-protected acceptor 17 (entry 17), the ꢀ-directing
effect by the sulfonyl group of 2 and acyl groups of 3, 4, and
5 appeared quite substantial.
Mannosylations with Donors Possessing an Electron-Withdraw-
ing Group at O-6. The mannosylation of benzyl-protected
primary hydroxy acceptor 19 with 6-O-benzylsulfonyl-tri-O-
(27) The field/inductive parameter F of SO2Ph is +0.58, and that of SO2Me
is +0.53, whereas those of Bz and Ac are +0.31 and +0.33,
respectively. On the other hand, the parameter F of Bn is known to
be-0.04. See: Hansch, C.; Leo, A.; Taft, R. W. Chem. ReV. 1991, 91,
165–195.
To examine the leaving group effect in the present manno-
sylation, we carried out mannosylations of acceptors 19-22 with
anomeric hydroxy sugar donors 6-8 with electron-withdrawing
(28) Bock, K.; Pedersen, C. J. Chem. Soc., Perkin Trans. 2 1974, 293–
297.
9
17708 J. AM. CHEM. SOC. VOL. 131, NO. 48, 2009

ꢀ- and R-Directing Effects in Mannopyranosylations
A R T I C L E S
Table 3. Mannosylations with Trichloroacetimidate Donors 9-12
Having EWGs at O-6
Table 4. Mannosylations with Trichloroacetimidate Donors 13-16
Possessing EWGs at O-3
entry
acceptor ROH
donor
EWG
product (yield, %)a
ratiob ꢀ/R
entry
acceptor ROH
donor
EWG
product (yield, %)a
ratiob ꢀ/R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
19
19
19
19
19
20
20
20
20
20
21
21
21
21
21
9
SO2Bn
Bzp-NO2
Bz
43 (76)
44 (83)
45 (83)
46 (84)
27 (91)
47 (85)
48 (85)
49 (85)
50 (85)
34 (96)
51 (70)
52 (85)
53 (85)
54 (82)
39 (87)
13.8/1
5.2/1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
19
19
19
19
19
20
20
20
20
20
21
21
21
21
21
13
14
15
16
17
13
14
15
16
17
13
14
15
16
17
SO2Bn
Bzp-NO2
Bz
55 (95)
56 (92)
57 (81)
58 (91)
27 (91)
59 (93)
60 (89)
61 (88)
62 (94)
34 (96)
63 (85)
64 (80)
65 (91)
66 (92)
39 (87)
15.9/1
1/1.7
1/9.8
1/25.9
2.7/1
10.2/1
1/2.3
1/29.6
1/39.0
2.5/1
11.8/1
1/3.8
1/19.8
1/40.4
1/2.7
10
11
12
17
9
10
11
12
17
9
10
11
12
17
6.8/1a
R onlya
2.7/1
Ac
Ac
(Bn)c
SO2Bn
Bzp-NO2
Bz
(Bn)c
SO2Bn
Bzp-NO2
Bz
9.1/1
3.5/1
4.0/1a
1/2.2
Ac
Ac
(Bn)c
SO2Bn
Bzp-NO2
Bz
2.5/1
5.6/1
(Bn)c
SO2Bn
Bzp-NO2
Bz
2.5/1
2.5/1a
R onlya
1/2.7
Ac
Ac
(Bn)c
(Bn)c
a Determined after isolation. b The ratio was determined by
a Determined after isolation. b The ratio was determined by
LC-mass. c A standard donor for comparison.
LC-mass. c A standard donor for comparison.
benzyl-mannsoyl trichloroacetimidate 9, a mannosyl donor
having a strongly electron-withdrawing group at O-6, was highly
ꢀ-selective, yielding mannosyl disaccharide 43 (entry 1 in Table
3). The p-nitrobenzoyl and benzoyl groups at O-6 of mannosyl
donors 10 and 11 also exhibited a ꢀ-directing effect in the
mannosylation of 19 (entries 2 and 3 in Table 3). Surprisingly,
however, the acetyl group at O-6 of mannosyl donor 12 in the
mannosylation of 19 exhibited a strong R-directing effect,
yielding R-mannosyl disaccharide 46 exclusively (entry 4),
whereas the mannosylation of 19 with the standard compound
17 with all benzyl-protecting groups gave an anomeric mixture
of disaccharide 27 (entry 5). Mannosylations of benzoyl-
protected primary hydroxy acceptor 20 with donors 9-12,
possessing electron-withdrawing groups at O-6, showed a trend
(entries 6-8) similar to mannosylations of 19 with 9-12.
However, the mannosylation of 20 with 12, possessing the 6-O-
acetyl group, was R-selective, affording mannoside 50 (ꢀ/R )
1:2.2) with a slight excess of the R-anomer (entry 9), while the
mannosylation of 20 with the standard compound 17 was
ꢀ-selective, giving 34 (ꢀ/R ) 2.5:1) (entry 10). The mannosy-
lations of the secondary hydroxy acceptor 21 with 9-11 also
exhibited the ꢀ-directing effect, giving disaccharides 51, 52,
and 53, respectively, favoring ꢀ-anomers (entries 11-13).
However, the acetyl group at O-6 of donor 12 exhibited a strong
R-directing effect in the mannosylation of 21, so that the
mannosylation product was exclusively R-disaccharide 54 (entry
14).
The electron-withdrawing benzylsulfonyl, p-nitrobenzoyl, and
benzoyl groups at O-6 of mannosyl donors exerted the ꢀ-direct-
ing effect, and this was more pronounced in the strongly
electron-withdrawing benzylsulfonyl group than in the weakly
electron-withdrawing p-nitrobenzoyl and benzoyl groups. It was
also noteworthy that the ꢀ-directing effect of electron-withdraw-
ing groups at O-6 was not smaller or even greater than the
ꢀ-directing effect of those at O-2.13-15 A surprising result was
that the acetyl group at O-6 of mannosyl donors exhibited an
R-directing effect, to give exclusively or predominantly R-dis-
accharides. The remote participation of the 6-O-acetyl group
of donors would be a possible explanation for its R-directing
effect.
Mannosylations with Donors Possessing an Electron-Withdraw-
ing Group at O-3. The mannosylation of the benzyl-protected
primary hydroxy acceptor 19 with mannosyl trichloroacetimidate
donor 13, possessing the benzylsulfonyl group at O-3, yielded
mannosyl disaccharide 55 (ꢀ/R ) 15.9:1) with a large excess
of the ꢀ-anomer (entry 1 in Table 4). However, the mannosy-
lation of 19 with donor 14, having p-nitrobenzoyl group at O-3,
was not ꢀ-selective, giving disaccharide 56 (ꢀ/R ) 1:1.7) with
a slight excess of R-anomer (entry 2). The benzoyl group at
O-3 of donor 15 exhibited a stronger R-directing effect than
the p-nitrobenzoyl group in the mannosylation of 19, and product
disaccharide 57 (ꢀ/R ) 1:9.8) contained predominantly the
R-anomer (entry 3). In the mannosylation of 19, the strongest
R-directing effect was obtained when there was an acetyl group
at O-3. Thus, the mannosylation of 19 with 16, with an acetyl
group at O-3, afforded almost exclusively R-disaccharide 58
(ꢀ/R ) 1:25.9) (entry 4). It is noteworthy that the mannosylation
of 19 with the standard donor 17, with a benzyl group at O-3,
gave mannoside 27 (ꢀ/R ) 2.7:1) with a little excess of the
ꢀ-anomer (entry 5). The high ꢀ-directing effect of the benzyl-
sulfonyl group was also observed in the mannosylation of
benzoyl-protected primary hydroxy acceptor 20 with donor 13,
with the benzylsulfonyl group at O-3, providing disaccharide
59 (ꢀ/R ) 10.2:1) with a large excess of the ꢀ-anomer (entry 6
in Table 4). In contrast, all three acyl groups showed an
R-directing effect in the mannosylation of 20. Thus, mannosy-
lations of 20 with donors 14-16, possessing an acyl group at
O-3, provided disaccharides favoring R-anomers, 60, 61, and
62, respectively (entries 7-9). Mannosylations of the secondary
hydroxy acceptor 21 with donors 13-16 also indicated that the
benzylsulfonyl group at O-3 was ꢀ-directing, giving 63 (ꢀ/R )
11.8:1) (entry 11), whereas the p-nitrobenzoyl group was weakly
R-directing, yielding 64 (ꢀ/R ) 1:3.8) (entry 12), and the
benzoyl and acetyl groups at O-3 were strongly R-directing,
giving 65 (ꢀ/R ) 1:19.8) and 66 (ꢀ/R ) 1:40.4), respectively
(entries 13 and 14).
We observed that nonparticipating, strongly electron-with-
drawing benzylsulfonyl groups at O-3 of mannosyl donors were
highly ꢀ-directing, while all acyl groups at O-3 were R-directing
9
J. AM. CHEM. SOC. VOL. 131, NO. 48, 2009 17709

A R T I C L E S
Baek et al.
Scheme 1. Formation of 68 by Intramolecular Trapping of the
Anomeric Oxocarbenium Ion Intermediate Resulting from the
Activation of 67
in mannosylations. Among three acyl groups, the acetyl group
exhibited the strongest R-directing effect, whereas the p-
nitrobenzoyl group at O-3 showed the weakest R-directing
effect. Mannosylations with donors having the O-3 acetyl group
probably proceeded almost exclusively through the six-
membered ring 2-methyl-1,3-dioxanium ion intermediate, in
which the sugar ring is in the 1C4 conformation, resulting from
the remote participation of the acetyl group. Mannosylations
with donors having the 3-O-benzoyl group might also have
proceeded through the 1,3-dioxanium ion pathway by the remote
participation of the benzoyl group, whereas the 3-O-p-nitroben-
zoyl group might have participated only partially during the
mannosylation.
Figure 2. Dioxocarbenium ions A and B are possible intermediates for
the formation of R-mannosides, but C might not be involved in mannosy-
lation pathways.
remote participation by 3-O-acyl and 6-O-acetyl groups, but
no participation by 4-O-acyl groups, could be explained by
comparing the relative stabilities of dioxocarbenium ion inter-
mediates resulting from possible remote participation of the acyl
groups. The participation of 3-O-acyl groups would generate
relatively stable bicyclic six-membered ring dioxocarbenium ion
1
intermediate A, in which the sugar ring is in a chair, C4,
conformation, and mannosylations with 3-O-acyl-mannosyl
donors would occur exclusively or predominantly through
intermediate A, as shown in Figure 2. The 6-O-acyl group
participation would generate the less stable seven-membered
dioxocarbenium ion B, having the sugar ring in a chair
conformation, and mannosylation with only the 6-O-acetyl-
mannosyl donor might occur through the bridging dioxocarbe-
nium ion B pathway. Nevertheless, the possibility cannot be
excluded that a small portion of the R-mannosyl products were
produced by the partial remote participation of the 6-O-benzoyl
and the 6-O-p-nitrobenzoyl groups. Participation of the 4-O-
acyl group would give the least stable seven-membered ring
dioxocarbenium ion C, with the sugar ring in a boat conforma-
tion, so that mannosylations with 4-O-acyl-mannosyl donors
would not occur through the pathway involving the interme-
diate C.
Detection of r-Mannopyranosyl Triflate Intermediates and
Their Relationship with the ꢀ-Directing Effect. Activation of
the trichloroacetimidate donors by TMSOTf, and of the ano-
meric hydroxy sugar donors by phthalic anhydride/Tf2O, would
lead to mannosyl oxocarbenium ions with triflate counteranions,
which might be in equilibrium with covalent R-mannopyranosyl
triflates. Since the first detection of R-mannopyranosyl triflates
by Crich and colleagues,32 they have been quite well character-
ized by low-temperature NMR experiments.9,15,18,22a,33 In some
stereoselective ꢀ-mannosylations, the SN2-like reaction between
the R-mannopyranosyl triflate intermediate and the acceptor
alcohol is attributed to the generation of the ꢀ-manno-
side.9,15,18,30,34 In the present study, we performed a low-
temperature NMR study to detect mannosyl triflates intermedi-
ates and determine their stabilities, to better understand the
ꢀ-directing effect of the electron-withdrawing groups. When
trichloroacetimidate donors 2, 9, and 13, possessing the ben-
zylsulfonyl group, and trichloroacetimidate donors 4, 11, and
15, possessing the benzoyl group, were activated with 1.0 equiv
Further Evidence for the Participation of 3-O-Acyl and
6-O-Acetyl Groups in Mannosylations. To obtain more support
for the remote participation of 3-O-acyl and 6-O-acetyl groups
in mannosyl donors, we carried out experiments to trap anomeric
oxocarbenium ion intermediates by the intramolecular nucleo-
philic attack of the tert-butoxycarbonyl (Boc) group or the
trichloroacetimidoyl group at the O-3, O-4, or O-6 positions of
mannosyl donors. The trapping of nearby electrophilic centers
bytheBocgroup,leadingtotheformationofcycliccarbonates,25,29
and by the trichloroacetimidoyl group, leading to the formation
of cyclic oxazolines or oxazines,30 is quite well established. In
the present trapping experiment, although there was no sign of
the generation of cyclic products in the reaction of phenyl 1-thio-
3-Boc-mannoside with 1-benzenesulfinyl piperidine (BSP) and
Tf2O,31 we were able to obtain stable bicyclic product 68, having
a six-membered trichlorooxazine ring, in 85% yield by activation
of phenyl 1-thio-3-O-trichloroacetimidoyl- mannoside 67 with
1
BSP and Tf2O, as shown in Scheme 1. H NMR spectral data
clearly indicated that the 3-trichloroacetimidate 67 was in the
4C1 conformation, while the sugar ring of the bicyclic product
1
68 was in the C4 conformation. Two vicinal diaxial coupling
constants, JH3-H4 ) 9.7 Hz and JH4-H5 ) 9.7 Hz, were observed
in 67, whereas the two coupling constants, JH3-H4 and JH4-H5
,
of product 68 were both found to be 2.8 Hz. In contrast, no
bicyclic products were isolated at all when phenyl 1-thio-
mannosides possessing either the trichloroacetimidoyl group or
the Boc group at one of the 4- and 6-positions were activated
with BSP and Tf2O. Nevertheless, a small amount of a bicyclic
product with a seven-membered trichlorooxazepine ring was
detected in the product mixture by mass spectrometry, resulting
from the activation of phenyl 1-thio-6-trichloroacetimidoyl-
mannoside with BSP and Tf2O. The result of our trapping
experiment was consistent with our mannosylation results. The
(29) Bartlett, P. A.; Meadows, J. D.; Brown, E. G.; Morimoto, A.; Jernstedt,
K. K. J. Org. Chem. 1982, 47, 4013–4018.
(32) Crich, D.; Sun, S. J. Am. Chem. Soc. 1997, 119, 11217–11223.
(33) (a) Nokami, T.; Shibuya, A.; Tsuyama, H.; Suga, S.; Bowers, A. A.;
Crich, D.; Yoshida, J. J. Am. Chem. Soc. 2007, 129, 10922–10928.
(b) Crich, D.; Cai, W. J. Org. Chem. 1999, 64, 4926–4930.
(34) Crich, D.; Chandrasekera, N. S. Angew. Chem., Int. Ed. 2004, 43,
5386–5389.
(30) (a) Bongini, A.; Cardillo, G.; Orena, M.; Sandri, S.; Tomasini, C. J.
Org. Chem. 1986, 51, 4905–4910. (b) Sammes, P. G.; Thetford, D.
J. Chem. Soc., Perkin Trans. 1 1988, 111–123.
(31) For activation of thioglycosides with BSP/Tf2O, see: Crich, D.; Smith,
M. J. Am. Chem. Soc. 2001, 123, 9015–9020.
9
17710 J. AM. CHEM. SOC. VOL. 131, NO. 48, 2009

ꢀ- and R-Directing Effects in Mannopyranosylations
A R T I C L E S
Table 5. R-Mannopyranosyl Triflates in Low-Temperature NMR Experiments
donor
triflate
δ
H1, ppm (J1,2, Hz)
δC13, ppm
relative intensity of H1a
decomp. temp, °C
2
69
70
71
72
73
74
6.09
(d, 1.9)
6.16
(s)
6.11
(s)
6.21
(d, 1.6)
6.04
(s)
6.21
(d, 1.5)
105.2
0.60
-30 to -20
-40 to -30
-20 to -10
-40 to -30
-30 to -20
-50 to -40
(4-O-SO2Bn)
4
(4-O-Bz)
9
(6-O-SO2Bn)
11
(6-O-Bz)
13
(3-O-SO2Bn)
15
(3-O-Bz)
17
106.3
105.4
105.2
105.2
105.6
0.16
0.47
0.05
0.30
0.07
not detected
(tetra-O-Bn)
a The ratio of the H1 signal integration of the triflate and that of the corresponding donor.
Scheme 2. Mechanism for the Stereoselective ꢀ-Mannosylation
with Donors Possessing Electron-Withdrawing Group (EWG)
Table 6. Mannosylations of 20 with 75 and 76 Employing AgClO4
or AgOTf as Promoter
donor
R1
R2
promoter
product (yield %)a
ratiob ꢀ/R
75
75
76
76
SO2Bn
SO2Bn
Bn
Bn
Bn
SO2Bn
SO2Bn
AgClO4
AgOTf
AgClO4
AgOTf
30 (53)
30 (55)
59 (51)
59 (55)
1/2.1
2.7/1
1/2.1
3.9/1
of TMSOTf in CD2Cl2 at -60 °C, NMR spectra showed
generation of corresponding R-mannosyl triflates 69-74, al-
though their signal intensities and thermal stabilities differed,
Bn
a Determined after isolation. b The ratio was determined by
LC-mass.
1
as shown in Table 5. Their H and 13C chemical shift values
and the zero or small H1-H2 coupling constants indicated that
mannosyl triflates 69-74 all have R-configurations at their
anomeric centers. Stabilities of the triflates were determined by
measuring their relative H1 signal intensities at -60 °C and
their decomposition temperatures. Each of the relative H1 signal
intensities given in Table 5 is the ratio of the H1 signal
integration of the triflate and that of the corresponding trichlo-
roacetimidate donor at -60 °C. The decomposition temperature
was determined by observing the disappearance of the H1 signal
of the triflates while the temperature was raised from -60 to
-10 °C. In contrast, the NMR spectra showed no sign of the
generation of the triflate intermediate when tetra-O-benzyl-
mannosyl donor 17 was activated with TMSOTf. Triflates 69,
71, and 73, possessing the benzylsulfonyl group, were found
to be consistently more stable than triflates 70, 72, and 74,
possessing the benzoyl group. The results clearly indicated that
the strongly electron-withdrawing benzylsulfonyl group, which
exhibited the stronger ꢀ-directing effect in the mannosylation,
also made the R-mannosyl triflate intermediate more stable than
the weakly electron-withdrawing benzoyl group.
We therefore propose the following mechanism of the present
ꢀ-mannosylation and the origin of the ꢀ-directing effect of the
electron-withdrawing groups. Activation of donor D with
TMSOTf would generate mannosyl oxocarbenium ion E, which
is in equilibrium with R-mannosyl triflate F as shown in Scheme
2. The electron-withdrawing group on the sugar ring would
stabilize F, but might destabilize E so that the equilibrium would
shift toward the R-mannosyl triflate F. Next, F, or its contact
ion pair, would react with acceptor alcohol ROH in a SN2-like
fashion to give ꢀ-mannoside G.
In this mechanism, one of the crucial factors for the
ꢀ-selectivity is the triflate counteranion of the mannosyl
oxocarbenium ion E, which could produce the very reactive
but reasonably stable covalent R-mannosyl triflate F. We
therefore performed mannosylations of 20 with mannosyl
bromides 75 and 76, by employing either AgClO4 or AgOTf as
the promoter, to examine the counteranion effect of the
mannosyl oxocarbenium ion on the stereochemical outcome in
the present mannosylation. With AgClO4 as the promoter, the
mannosylation of 20 with the donor 75 bearing the O-4-
benzylsulfonyl group, in CH2Cl2 at -50 °C was R-selective and
gave disaccharide 30 (ꢀ/R ) 1:2.1), while the same mannosy-
lation with AgOTf as the promoter was ꢀ-selective and gave
30 (ꢀ/R ) 2.7:1) as shown in Table 6. Similarly, in the
mannosylation of 20 with 76, bearing the 3-O-benzylsulfonyl
group, employing AgClO4 as the promoter, product mannoside
59 had more R-anomer (ꢀ/R ) 1:2.1), whereas the ꢀ-mannoside
was found to be the major anomer in product 59 (ꢀ/R ) 3.9:1),
produced from the reaction of 20 with 76 employing AgOTf as
the promoter. The results clearly indicate that the counteranion
of the mannosyl oxocarbenium ion profoundly affects the
outcome of the stereochemistry of mannosylations. The activa-
tion of mannosyl bromide with AgOTf would generate the
mannosyl oxocarbenium ion, having the triflate counteranion,
which could form the covalent R-mannosyl triflate at equilib-
9
J. AM. CHEM. SOC. VOL. 131, NO. 48, 2009 17711

A R T I C L E S
Baek et al.
Table 7. Mannosylations of 77 with Donors Bearing One
Benzylsulfonyl and Two Benzylsulfonyl Groups
unusual C1′-H1′ coupling constant for the ꢀ-anomer of 82
1
might be attributed to a non-4C1 conformation, maybe the C4
conformation, of the sugar ring at the nonreducing end was
excluded by the measurement of exact values of all proton
vicinal coupling constants of the sugar ring.36 Finally, the NOE
experiments, in which the strong NOE interaction between H1′
and H2′ of the major product was observed, confirmed our
tentative assignment and thus enable us to unequivocally assign
the major product (JC1′-H1′ ) 171 Hz) as ꢀ-disaccharide 82ꢀ
and the minor product (JC1′-H1′ ) 174 Hz) as R-disaccharide
82r.
entry
donor
R1
R2
R3
product (yield, %)a
ratiob ꢀ/R
Conclusion
1
2
3
4
2
9
13
81
SO2Bn Bn
Bn
Bn
Bn
SO2Bn Bn
Bn SO2Bn
78 (70)
79 (86)
80 (78)
82 (84)
R only
3.8/1
1/3.7
6/1
Through a systematic study on mannosylations with donors
possessing an electron-withdrawing group at O-3, O-4, or O-6
positions, we clearly observed the ꢀ-directing effect of these
electron-withdrawing groups, except for 3-O-acyl and 6-O-acetyl
groups, which exhibited the R-directing effect in mannosylations.
The R-directing effect was ascribed to the remote participation
of 3-O-acyl and 6-O-acetyl groups, whereas the ꢀ-directing
effect was explained by the reaction between the acceptor and
the R-mannosyl triflate intermediate, which is stabilized by
electron-withdrawing groups. Evidence for the remote participa-
tion was provided, and the stability of the R-mannosyl triflates
was determined by a low-temperature NMR study. With a donor
possessing two benzylsulfonyl groups at O-4 and O-6 positions,
the ꢀ-selective mannosylation of a sterically demanding acceptor
was achieved.
SO2Bn SO2Bn Bn
a Determined after isolation. The ratio was determined by H NMR.
b
1
rium; then the reaction of the R-triflate with the acceptor would
provide the ꢀ-mannoside. In contrast, the activation of the
mannosyl bromide with AgClO4 would form the mannosyl
oxocarbenium ion, having the noncoordinating (or weakly
coordinating) perchlorate counteranion, so that the reaction of
the oxocarbenium ion with the acceptor would be the major
pathway to provide the R-mannoside as the major product.35
Mannosylation of a Sterically Demanding Secondary Hy-
droxy Sugar with the Donor Bearing Two Benzylsulfonyl Groups
at O-4 and O-6 Positions. Mannosylations of the secondary
hydroxy acceptor 21 were found to be less ꢀ-selective than those
of the primary hydroxy acceptors 19, 20, and 22. We examined
the ꢀ-selectivity of the mannosylation of another secondary
hydroxy sugar acceptor, 77, the mannosylation of which with
the donor bearing the strongly electron-withdrawing o-trifluo-
romethylbenzenesulfonyl group at the O-2 position was previ-
ously reported to be not ꢀ-selective, possibly because of steric
hindrance.15 In the present study, mannosylations of 77 with 2,
possessing the 4-O-benzylsulfonyl group, and with 13, possess-
ing 3-O-benzylsulfonyl groups, were not ꢀ-selective, either,
yielding disaccharides 78 (R only) and 80 (ꢀ/R ) 1:3.7),
respectively (entries 1 and 3 in Table 7), while the reaction of
77 with 9, bearing the 6-O-benzylsulfonyl group, was ꢀ-selec-
tive, giving 79 (ꢀ/R ) 3.8:1) (entry 2). To improve the
ꢀ-selectivity of the mannosylation of the sterically demanding
77, we prepared mannosyl donor 81, possessing two benzyl-
sulfonyl groups, both at O-4 and O-6 positions. The mannosy-
lation of 77 with 81, with two benzylsulfonyl groups, exhibited
a substantially improved ꢀ-selectivity, giving disaccharide 82
(ꢀ/R ) 6.0:1) in 84% yield (entry 4 in Table 7). Comparison
of their one-bond C1′-H1′ coupling constants and C1′ chemical
shifts led us to assign tentatively the major product as the
ꢀ-anomer 82ꢀ and the minor product as the R-anomer 82r.
However, we conducted more NMR studies for the unambiguous
assignment because C1′-H1′ coupling constant values, 171 Hz
at 99.5 ppm for the major product and 174 Hz at 98.9 ppm for
the minor product, both fall into the range of those for the
R-anomer of hexopyranosides.28 At first, the possibility that the
Experimental Section
General Procedure for Mannosylations with Mannosyl Trichlo-
roacetimidate Donors. A solution of a mannosyl trichloroacetimi-
date donor (0.15 mmol, 1.5 equiv) and a mannosyl acceptor (0.10
mmol, 1.0 equiv) in CH2Cl2 (5 mL) was stirred for 5 min at room
temperature and cooled to -78 °C. After the addition of TMSOTf
(0.015 mmol, 0.15 equiv), the reaction mixture was stirred at -78
°C for 2 h, quenched with triethylamine, and concentrated in vacuo.
The residue was purified by silica gel flash column chromatography.
General Procedure for Mannosylations with Anomeric Hy-
droxy Mannose Donors. A solution of an anomeric hydroxy
mannose donor (0.1 mmol, 1.0 equiv), phthalic anhydride (1.1
equiv), and 1,8-diazobicyclo[5,4,0]undec-7-ene (DBU, 1.2 equiv)
in CH2Cl2 (4 mL) in the presence of 4A molecular sieves was stirred
for 15 min at room temperature and cooled to -78 °C. Next, Tf2O
(1.5 equiv) was added, and the resulting solution was stirred for
further 15 min at -78 °C. After addition of a mannosyl acceptor
(1.5 equiv) in CH2Cl2 (2 mL) to the above solution via cannula,
the reaction mixture was stirred at -78 °C for 15 min, allowed to
warm over 1 h to 0 °C, quenched with saturated aqueous NaHCO3,
and extracted with CH2Cl2. The combined organic phase was
washed with brine, dried over MgSO4, and concentrated in vacuo.
The residue was purified by silica gel flash column chromatography.
Formation of Bicyclic Trichlorooxazine 68 by Intramolecular
Trapping of the Anomeric Oxocarbenium Ion Intermediate
Resulting from the Activation of 67. A solution of 67 (118 mg,
0.172 mmol) and BSP (43.1 mg, 0.206 mmol) in CH2Cl2 (5 mL)
in the presence of 4A molecular sieves was stirred for 15 min at
-60 °C. After addition of Tf2O (43.3 µL, 0.258 mmol), the reaction
mixture was stirred at -60 °C for 1 h, allowed to warm over 1 h
to 0 °C, quenched with saturated aqueous NaHCO3, and extracted
(35) For discussions on preferred conformations and stereoselective reac-
tions of pyranosyl oxocarbenium ions, see: (a) Yang, M. T.; Woerpel,
K. A. J. Org. Chem. 2009, 74, 545–553. (b) Lucero, C. G.; Woerpel,
K. A. J. Org. Chem. 2006, 71, 2641–2647. (c) Chamberland, S.; Ziller,
J. W.; Woerpel, K. A. J. Am. Chem. Soc. 2005, 127, 5322–5323. (d)
Ayala, L.; Lucero, C. G.; Romero, J. A. C.; Tabacco, S. A.; Woerpel,
K. A. J. Am. Chem. Soc. 2003, 125, 15521–15528.
(36) Because of the extensive overlap of sugar ring protons of 82ꢀ, the
assignment of proton signals and the measurement of proton vicinal
coupling constants in the sugar ring at the nonreducing end of 82ꢀ
were only possible after the selective COSY and TOCSY: JH1′-H2′
1.8 Hz, JH2′-H3′ ) 3.1 Hz, JH3′-H4′ ) 9.7 Hz, and JH4′-H5′ ) 9.7 Hz.
)
9
17712 J. AM. CHEM. SOC. VOL. 131, NO. 48, 2009

ꢀ- and R-Directing Effects in Mannopyranosylations
A R T I C L E S
with CH2Cl2. The combined organic phase was washed with brine,
dried over MgSO4, and concentrated in vacuo. The residue was
purified by silica gel flash column chromatography (hexane/EtOAc,
6:1) to afford compound 68 (84.6 mg, 85%): colorless oil, Rf )
0.35 (hexane/EtOAc, 5:1, v/v); 1H NMR (400 MHz, CDCl3) δ 3.38
(dd, J ) 9.6, 8.3 Hz, 1H, H-6), 3.45 (dd, J ) 9.7, 4.8 Hz, 1H,
H-6′), 3.96 (t, J ) 2.8 Hz, 1H, H-4), 3.98-4.04 (m, 1H, H-5),
4.08 (t, J ) 2.4 Hz, 1H, H-2), 4.44 and 4.50 (ABq, J ) 12.0 Hz,
2H), 4.55 and 4.62 (ABq, J ) 12.0 Hz, 2H), 4.59 and 4.61 (ABq,
J ) 12.0 Hz, 2H), 4.85 (dd, J ) 2.8, 2.4 Hz, 1H, H-3), 5.35 (t, J
) 2.4 Hz, 1H, H-1), 7.21-7.35 (m, 15H); 13C NMR (100 MHz,
CDCl3) δ 65.7, 70.7, 70.8, 71.7, 72.0, 73.47, 73.52, 75.6, 76.5,
127.8, 127.85, 127.88, 128.1, 128.2, 128.3, 128.5, 128.65, 128.69,
137.0, 137.2, 138.0, 155.8. HRMS calcd for C29H28Cl3NO5Na [M
+ Na]+, 598.0931; found, 598.0934.
at -78 °C for 2 h, quenched with triethylamine, and concentrated
in vacuo. The residue was purified by silica gel flash column
chromatography (hexane/EtOAc/CH2Cl2, 3:1:1) to afford compound
82r (20 mg, 12%): colorless oil, Rf ) 0.38 (hexane/EtOAc/CH2Cl2,
1
3:1:1, v/v/v); [R]20 ) 20.7 (c 1.1, CHCl3); H NMR (400 MHz,
D
CDCl3) δ 3.36 (s, 3H), 3.38-3.44 (m, 2H), 3.51 (dd, J ) 11.2, 3.6
Hz, 1H), 3.55 (dd, J ) 11.2, 3.6 Hz, 1H), 3.65-3.71 (m, 2H),
3.78-3.91 (m, 2H), 3.93 (d, J ) 9.2 Hz, 1H), 4.01 (d, J ) 14.0
Hz, 1H), 4.13 (d, J ) 14.4 Hz, 1H), 4.22 (d, J ) 14.4 Hz, 1H),
4.24 (d, J ) 10.4 Hz, 1H), 4.33-4.39 (m, 2H), 4.46-4.58 (m,
6H), 4.65 (d, J ) 12.0 Hz, 1H), 4.86 (d, J ) 11.6 Hz, 1H), 5.01
(d, J ) 11.6 Hz, 1H), 5.02-5.08 (m, 2H), 5.41 (d, J ) 1.2 Hz, 1H,
H-1′), 7.12-7.41 (m, 35H); 13C NMR (100 MHz, CDCl3) δ 55.4,
57.6, 57.7, 68.4, 69.0, 69.5, 72.4, 72.8, 73.3, 73.4, 73.7, 74.9, 76.7,
77.12, 77.13, 80.4, 81.1, 97.9, 98.9 (C-1′, JC1′-H1′ ) 174 Hz), 127.4,
127.5, 127.6, 127.9, 128.0, 128.1, 128.3, 128.4, 128.52, 128.53,
128.6, 128.90, 128.92, 129.1, 129.3, 130.7, 131.0, 131.1, 137.6,
138.0, 138.3, 139.0. HRMS calcd for C62H66O15S2Na [M + Na]+,
1137.3741; found, 1137.3740.
Procedure for the Detection and Thermal Decomposition of
r-Mannosyl Triflates 69-74 in CD2Cl2 by Low-Temperature
NMR Analysis. A 5 mm NMR tube containing a solution of a
mannosyl trichloroacetimidate 2, 4, 9, 11, 13, 15, or 17 (0.05-0.10
mmol) in CD2Cl2 (500 µL) was placed in the NMR probe at room
Further elution provided compound 82ꢀ (118 mg, 70%): colorless
1
temperature, cooled to -60 °C, and then the reference H NMR
oil, Rf ) 0.35 (hexane/EtOAc/CH2Cl2, 3:1:1, v/v/v); [R]20 ) 8.9
D
spectrum was obtained. The NMR tube was removed from the NMR
probe, and TMSOTf (1.2 equiv) was added to the tube at -78 °C
in an acetone/dry ice bath. After being briefly agitated, the NMR
tube was placed in the precooled NMR probe at -60 °C, and the
1H NMR spectrum was recorded again. The conversion of the
mannosyl trichloroacetimidate into the R-mannosyl triflate was
almost instantaneous, and the 1H NMR spectrum showed the
anomeric proton signal of the R-mannosyl triflate. Also, then the
1H-13C HSQC spectrum was recorded to assign 13C chemical shifts
and to confirm the anomeric proton signal. The NMR probe
1
(c 5.7, CHCl3); H NMR (400 MHz, CDCl3) δ 3.42 (s, 3H), 3.56
(dd, J ) 9.6, 3.6 Hz, 1H), 3.63 (d, J ) 5.6 Hz, 1H), 3.68 (d, J )
5.6 Hz, 1H), 3.72-3.76 (m, 3H), 3.84-3.88 (m, 2H, H-3′, H-5′),
3.90 (dd, J ) 9.6, 2.8 Hz, 1H), 4.10 (dd, J ) 11.4, 4.4 Hz, 1H),
4.11 (d, J ) 11.6 Hz, 1H), 4.17 (d, J ) 9.2 Hz, 1H), 4.18 (d, J )
14.4 Hz, 1H), 4.20-4.28 (m, 2H), 4.29 (d, J ) 14.4 Hz, 1H), 4.34
(d, J ) 14.0 Hz, 1H), 4.41 and 4.44 (ABq, J ) 11.2 Hz, 2H), 4.48
(d, J ) 12.2 Hz, 1H), 4.53-4.65 (m, 4H), 4.68 (d, J ) 12.0 Hz,
1H), 4.95 (t, J ) 9.7 Hz, 1H, H-4′), 5.16 (d, J ) 12.0 Hz, 1H),
5.32 (d, J ) 1.8 Hz, 1H, H-1′), 7.02-7.38 (m, 35H); 13C NMR
(100 MHz, CDCl3) δ 55.5, 68.8, 69.2, 69.8, 70.0, 71.6, 72.5, 73.3,
73.5, 75.0, 75.1, 76.1, 76.7, 77.4, 80.1, 81.5, 97.8, 99.5 (C-1′, JC1′-
H1′ ) 171 Hz), 126.6, 127.4, 127.66, 127.73, 127.80, 127.84, 128.0,
128.2, 128.3, 128.4, 128.56, 128.60, 128.64, 128.7, 128.89, 128.93,
129.0, 130.7, 130.9, 137.3, 137.8, 137.9, 138.2, 138.8. HRMS calcd
for C62H66O15S2Na [M + Na]+, 1137.3741; found, 1137.3743.
1
temperature was increased by 10 °C increments, and H NMR
spectra were acquired at each temperature until the thermal
decomposition of the R-mannosyl triflate was completed.
General Procedure for Mannosylations of 20 with Mannosyl
Bromides 75 and 76 Employing AgClO4 or AgOTf as the
Promoter. A solution of a mannosyl bromide (70 mg, 0.105 mmol)
and a promoter (AgClO4 or AgOTf, 0.157 mmol) in CH2Cl2 (3
mL) in the presence of 4A molecular sieves was stirred for 10 min
at -50 °C. After addition of acceptor 20 (63.7 mg, 0.126 mmol)
to the above solution, the reaction mixture was stirred at -50 °C
for 2 h, quenched with saturated aqueous NaHCO3, and extracted
with CH2Cl2. The combined organic phase was washed with brine,
dried over MgSO4, and concentrated in vacuo. The residue was
purified by silica gel flash column chromatography.
Acknowledgment. This Article is dedicated with respect and
affection to the late Professor Chi Sun Hahn, an inspiring teacher
and mentor, for his contributions to the field of organic chemistry
in Korea. This work was supported by a grant from the Korea
Science and Engineering Foundation through the Center for
Bioactive Molecular Hybrids (CBMH). J.Y.B., B.-Y.L., and M.G.J.
thank the fellowship of the BK 21 program from the Ministry of
Education and Human Resources Development. We thank Dr. Hye-
Seo Park for her help with NMR studies.
Methyl (2,3-Di-O-benzyl-4,6-di-O-benzylsulfonyl-r-D-mannopy-
ranosyl)-(1f4)-2,3,6-tri-O-benzyl-r-D-mannopyranoside (82r) and
Methyl (2,3-Di-O-benzyl-O-4,6-di-O-benzylsulfonyl-ꢀ-D-mannopy-
ranosyl)-(1f4)-2,3,6-tri-O-benzyl-r-D-mannopyranoside (82ꢀ). A
solution of 2,3-di-O-benzyl-4,6-di-O-benzylsulfonyl-R-D-mannopy-
ranosyl trichloroacetimidate (81, 183 mg, 0.226 mmol) and acceptor
77 (70 mg, 0.151 mmol) in CH2Cl2 (5 mL) was stirred for 5 min
at room temperature and cooled to -78 °C. After the addition of
TMSOTf (4.1 µL, 0.0226 mmol), the reaction mixture was stirred
Supporting Information Available: Experimental procedure,
characterization data, and NMR spectra. This material is
available free of charge via the Internet at http://pubs.acs.org.
JA907252U
9
J. AM. CHEM. SOC. VOL. 131, NO. 48, 2009 17713