The Origin of High Stereoselectivity in Di-tert-butylsilylene-Directed α-Galactosylation

Article abstract of DOI:10.1021/acs.joc.6b01685

The origin of the high α(1,2-cis)-stereoselectivity in the reaction of galactosyl and galactosaminyl donors with a di-tert-butylsilylene (DTBS) group with several nucleophiles has been elucidated by means of experimental and computational approaches. DTBS overcomes any other cyclic protecting groups examined to date and the β(1,2-trans)-directing effect due to the neighboring participation by CO groups at C2. Requirements for the α(1,2-cis)-stereoselectivity are as follows: (1) generation of an oxocarbenium ion; (2) a galacto-type glycosyl donor with a cyclic protecting group bridging O4 and O6 to form a six-membered ring; (3) through-space electron donation from O4 and O6 into the empty p-orbital of the anomeric carbon to stabilize the oxocarbenium intermediate; (4) steric hindrance due to bulky alkyl substituents on the cyclic protecting group to prevent nucleophilic attack from the β-face; and (5) a 4,6-O-silylene structure. Furthermore, it was found that the strong stereodirecting effect of the DTBS group was unique and specific among the various cyclic protecting groups examined.

Full text of DOI:10.1021/acs.joc.6b01685

Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
The Origin of High Stereoselectivity in Di-tert-
butylsilylene-directed alpha-Galactosylation
Akihiro Imamura, Naomi Matsuzawa, Shizuo Sakai, Taro Udagawa,
Shinya Nakashima, Hiromune Ando, Hideharu Ishida, and Makoto Kiso
J. Org. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.joc.6b01685 • Publication Date (Web): 20 Sep 2016
Downloaded from http://pubs.acs.org on September 23, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
The Journal of Organic Chemistry is published by the American Chemical Society.
1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
The Origin of High Stereoselectivity in Di-tert-butylsilylene-directed
α-Galactosylation
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
,
†
†
†
§
†
Akihiro Imamura,* Naomi Matsuzawa, Shizuo Sakai, Taro Udagawa, Shinya Nakashima,
†
,‡
†
†,‡
Hiromune Ando, Hideharu Ishida and Makoto Kiso
†
Department of Applied Bio-organic Chemistry, Faculty of Applied Biological Sciences, Gifu
University, 1-1 Yanagido, Gifu 501-1193, Japan; e-mail: aimamura@gifu-u.ac.jp
‡
Institute for Integrated Cell-Material Sciences (WPI-iCeMS), Kyoto University, Yoshida
Ushinomiya-cho, Sakyo-ku, Kyoto 606-8501, Japan
§
Department of Chemistry and Biomolecular Science, Faculty of Engineering, Gifu University,
-1 Yanagido, Gifu 501-1193, Japan
1
1
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 2 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Abstract
The origin of the high α(1,2-cis)-stereoselectivity in the reaction of galactosyl and
galactosaminyl donors with a di-tert-butylsilylene (DTBS) group with several nucleophiles has
been elucidated by means of experimental and computational approaches. DTBS overcomes any
other cyclic protecting groups examined to date and the β(1,2-trans)-directing effect due to the
neighboring participation by CO groups at C2. Requirements for the α(1,2-cis)-stereoselectivity
are as follows: (1) generation of an oxocarbenium ion; (2) a galacto-type glycosyl donor with a
cyclic protecting group bridging O4 and O6 to form a six-membered ring; (3) through-space
electron donation from O4 and O6 into the empty p-orbital of the anomeric carbon to stabilize
the oxocarbenium intermediate; (4) steric hindrance due to bulky alkyl substituents on the cyclic
protecting group to prevent nucleophilic attack from the β-face; and (5) a 4,6-O-silylene
structure. Furthermore, it was found that the strong stereo-directing effect of the DTBS group
was unique and specific among the various cyclic protecting groups examined.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Introduction
The pursuit of stereoselectivity in chemical glycosylation reactions is a key to success in the
synthesis of glycoconjugates. To date, many factors in glycosylation have been identified that
1
potentially affect the stereochemistry of glycoside product. Among these factors, the presence
of an acyl protecting group at the C2 position of the glycosyl donor is one of the most useful
2
factors for forming 1,2-trans-glycosides. Meanwhile, the synthesis of 1,2-cis-glycosides has
been plagued by lower stereoselectivity due to a lack of reliable factors for controlling the
glycosylation reaction. Hence, synthetic chemists have focused on developing methods for
2
ACS Paragon Plus Environment

Page 3 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
3
1
,2-cis-stereoselective glycosylation. In particular, the formation of the α(1,2-cis)-glycosidic
linkage between galactosamine and serine/threonine has garnered much attention because it is
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
the core linkage of natural O-glycans in glycoproteins. However, the construction of the
α-N-acetylgalactosamine (α-GalNAc) linkage is usually difficult because the presence of the
2
-acetamide group leads to the formation of the β(1,2-trans)-glycoside through neighboring
group participation or the undesirable 1,2-oxazoline derivative through intramolecular
cyclization. Thus, 2-azido-2-deoxy-galactosyl donors have been used in glycosylation reactions
5
to maximize the anomeric effect, resulting in the formation of α-glycosides. Although
2-azidogalactosaminyl (GalN) donors are useful for synthesizing glycans containing α-GalNAc,
the glycosylation products are often α/β mixtures with moderate stereoselectivity. In 2003, we
6
developed a di-tert-butylsilylene (DTBS)-directed α-stereoselective galactosylation and found
that both GalN and galactosyl donors bearing the DTBS group as a protecting group bridging
the C4 and C6 positions gave unusual high α-stereoselectivity (Figure 1). Notably, this method
was compatible with a wide variety of leaving groups, activator systems, reaction temperatures,
solvent systems, and protecting groups (even in the presence of a neighboring acyl group at C2).
High α-selectivity was almost completely independent of the nucleophile (the sugar acceptor),
which could be an alcohol, either linear or sterically hindered, a phenol, Ser, Thr and partially
7
protected monosaccharides. Furthermore, DTBS-directed α-galactosylation was used to
7
efficiently synthesize various α-GalNAc/Gal-containing glycans, such as α-GalNAc-Ser/Thr,
8
8
α-galactosyl ceramides, isoglobotrihexosylceramide, the α-tetrasaccharyl-Ser segment of
7
9
10
glycophorin A, p-nitrophenyl O-glycan cores, 4-methylumbelliferyl T-antigen, human ABO
11
12
histo-blood group antigens, and some gangliosides. Also, glycosyl donors that are
3
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 4 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
conformationally restricted by the DTBS group can be effective for a wide variety of
1
3
14
stereoselective glycosylations, such as β-arabinofuranosylation, β-galactofuranosylation,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
5
16
α-sialylation, and β-mannosylation. These reports demonstrate that the DTBS group can
serve as a factor for controlling glycosylation stereoselectivity. Several groups have reported
theoretical and experimental studies regarding the ring restriction of glycosyl donors
17,18
responsible for the reactivity of the donor and the stereochemical outcome of glycosylation.
DTBS
Si
O
Si
O
O
O
+ ROH
O
X
O
PGO
PGO
LG
O
–
LG-H
O
OR
X: N or O
X
LG: Leaving group
PG: Protecting group
R'
R'
α-glycoside
(predominantly formed)
Neighboring
acyl group
Figure 1. Outline of the DTBS-directed α-galactosylation
In previous papers, we proposed a reaction mechanism to explain the high α-stereoselectivity of
the DTBS-directed reaction, which is summarized as follows (Figure 2). The previous X-ray
crystallographic analysis of a DTBS-protected donor showed a six-membered ring composed of
a 4,6-O-DTBS ketal moiety and a C4-C5-C6 bond in a nearly half-chair conformation, which
resulted in the tert-butyl group being positioned close to the anomeric carbon, even in the
6
ground state. This implies that the steric effect of the DTBS group during the course of the
1
9
glycosylation leads to the α-stereoselectivity. Furthermore, through-space electron donation
may contribute to the high α-stereoselectivity. Because of through-space electrostatic
stabilization of the oxocarbenium ion by an axially oriented electronegative substituent at C4,
the group on the C2 oxygen or nitrogen cannot interact effectively with the oxocarbenium ion in
4
ACS Paragon Plus Environment

Page 5 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
the intermediate. This lack of interaction, together with the steric effect of the DTBS group,
results in the prominent α-stereoselectivity. However, this mechanism for the DTBS-directed
α-galactosylation is still only a hypothesis and needs experimental verification. Therefore, to
gain insight into the origin of the α-selectivity, we examined how the steric and electronic
effects of the substituents on both O4 and O6 of the glycosyl donor influenced stereoselectivity.
Here, we report a systematic study using experimental and computational approaches in
combination to ascertain the origin of stereoselectivity in DTBS-directed α-galactosylation.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CH3
Si
Si
Si
O
Si
O CH
CH3
Si
ROH
O
Half-chair
Conformation
O
O
O
O
3
O
O
O
β-face
HO-R
O
O
O
O
O
PGO
R'
PGO
LG
PGO
PGO
PGO
X
X
O
R'
X
LG
X
X
OR
O
O
α-face attack
O
R'
O
(predominantly)
R'
Oxocarbenium ion
R'
X: N or O
LG: Leaving group
PG: Protecting group
α-glycoside
Through-space
electron donation
from O4 and O6
Figure 2. Plausible mechanism for the α-selectivity in DTBS-directed
α-galactosylation
5
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 6 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Results and Discussion
In this study, we focused on the relationship between stereoselectivity and the magnitude of
steric and electronic effects of substituents on O4/O6 of the glycosyl donor. For this purpose,
we designed various GalN and glucosaminyl (GlcN) donors in which the only structural
variation is the protecting groups on O4 and O6 (Figure 3). Selected protecting groups were a
diisopropylsilylene group (less bulky than the DTBS group; 2), various acetal groups that
differed in bulkiness and electronic properties (3–9), cyclic groups that differed in ring size (10,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2), and acyclic groups (11, 13–15). Donors 1–16 were synthesized as outlined in Scheme 1.
Starting from the known compound phenyl
-deoxy-2-phthalimide-1-thio-β-D-galactopyranoside (17), 4,6-O-benzylidene-protected donor
2
0
2
6
was prepared according to a typical acetalization procedure followed by benzylation at C3.
Si
Si
Si
Ac
Ac
Donor 1
16
2
Me
Me
15*
(GlcN donor)
3
1
4*
4
O
O
O
BnO
SPh
NPhth
5
1
3*
core structure of GalN donors
NO2
OMe
12
6
Si
Si
O Si
Si
Cl
7
1
1*
*
acyclic type donor
10
8
9
Figure 3. Structures of glycosyl donors used in this study
6
ACS Paragon Plus Environment

Page 7 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Acidic hydrolysis then gave 4,6-diol product 18, which was the common intermediate for
synthesizing the other GalN donors. Compound 18 was transformed into GalN donors 1–15
using appropriate reaction conditions for each. For silylene-bridged donors, 18 was reacted with
dialkylsilyl ditriflate in the presence of base, giving 1 (92%) and 2 (86%). Treatment of 18 with
pivalaldehyde, dimethylacetal, or dimethoxymethane in the presence of camphorsulfonic acid
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
CSA) afforded alkylidene donors 3 (84%), 4 (92%), and 5 (84%), respectively. Substituted
benzylidene donors were derived from 18 by treatment with 4-nitrobenzaldehyde or
anisaldehyde dimethyl acetal under acidic conditions, affording 7 (94%) and 8 (89%),
respectively. Similarly, the reaction of 18 with 2-chloroacetaldehyde diethyl acetal in the
presence of camphorsulfonic acid afforded 9 in 74% yield. 1′,1′,3′,3′-Tetraisopropyldisiloxane
(
2
TIPDS)-protected donor was prepared by treating 18 with TIPDSCl and imidazole in DMF,
giving 10 in 88% yield. The reaction of 18 with tert-butyldimethylsilyl (TBS) triflate, imidazole
along with 2,6-di-tert-butyl-4-methylpyridine as an acid scavenger gave acyclic
4,6-di-O-TBS-protected donor 11 in 96% yield. 4,6-O-Xylylene-protected donor 12 was
prepared by the Williamson ether synthesis. Similarly, acyclic donors 13 and 14 were obtained
in yield of 85% and 68%, respectively. The reaction of 18 with acetic anhydride in pyridine
afforded 4,6-O-acetyl-protected donor 15 in quantitative yield. Meanwhile, glucosamine donor
2
1
1
6 bearing the DTBS group was prepared from the known glucosamine derivative 19 in two
steps. Hydrolysis of the benzylidene acetal under acidic conditions gave 20, which was reacted
with DTBS(OTf) in pyridine to afford 16 in good yield. The glucosyl acceptor (23) was
2
synthesized as follows and used in all the glycosylation experiments. The hydroxyl groups of
22
the known 2-(trimethylsilyl)ethyl 4,6-O-benzylidene-β-D-glucopyranoside 21 were acetylated
7
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 8 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
to give 22, and subsequent reductive opening of the benzylidene acetal furnished 6-OH glucosyl
acceptor 23 in 79% yield over the two steps.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
8
ACS Paragon Plus Environment

Page 9 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Ph
O
1
) PhCH(OMe)2
CSA, MeCN
HO OH
O
O
HO OH
O
80% aq AcOH
O
HO
SPh
NPhth
BnO
SPh
NPhth
BnO
SPh
NPhth
8
0 °C
7%
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2) BnBr, NaH
DMF
9
6
9% (2 steps)
1
7
6
18
DTBS(OTf)2
Py, 92%
DIPS(OTf)2
Me CCHO
CSA, MeCN
84%
3
MeCH(OMe)2
CSA, MeCN
92%
CH (OMe)
2
2
Et N, DMF
CSA, MeCN
rt to 60 °C
84%
3
8
6%
Si
Si
4
5
1
2
3
p-NO PhCHO
p-MPCH(OMe)2 ClCH CH(OEt)
2
CSA, MeCN
89%
TIPDSCl2
Im, DMF
TBSOTf
Im, DTBMP
DMF, 96%
2
2
CSA, DMF
100 °C, 94%
CSA, MeCN
rt to 70 °C, 74% 88%
NO2
OMe
Cl
O Si
Si
Si
Si
9
10
11
7
8
α,α'-dibromo-o-xylene
NaH, DMF
BnBr, NaH
DMF
85%
MeI, NaH
DMF
Ac O, Py
2
DMAP
quant.
38%
68%
Me
Ac
Me
Ac
1
4
15
12
13
OH
O
Si
O
Ph
O
80% aq AcOH
0 °C
DTBS(OTf)2
Py
O
O
O
BnO
HO
O
SPh
NPhth
SPh
SPh
NPhth
BnO
BnO
6
NPhth
8
6%
1
9
20
(2 steps)
16
(GlcN donor)
Ac O
2
PhBCl2
TESH
OH
O
Ph
O
DMAP
Ph
O
O
O
OH
O
O
HO
BnO
AcO
OSE
OSE
OSE
AcO
THF
OAc
CH2Cl2
4Å MS
OAc
9
8%
2
1
22
23
-78 °C
8
1%
SE: 2-(Trimethylsilyl)ethyl
Scheme 1. Preparation of glycosyl donors (1-16) and glucose acceptor 23
9
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 10 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Next, GalN/GlcN donors 1–16 and glucosyl acceptor 23 were used in glycosidation experiments
to examine the relationship between substituents and stereoselectivity in relation to the cyclic
protecting groups at C4 and C6 of the glycosyl donor. All glycosylation reactions were carried
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
3
2 2
out using the NIS-TfOH promoter system in CH Cl at 0 °C. Moreover, almost all the
glycosidations proceeded rapidly, requiring a reaction time less than 30 min. This high
18d
reactivity can be explained by Bols’ work which found that the dihedral angle based on the
exocyclic C5–O6 bond can strongly influence the reactivity of carbohydrates and gg conformer
is the most reactive among tg, gg, and gt conformers. Specifically, the C5–O6 bond in the
4,6-O-ring-fused galactosyl donors must receive the imposition of the relatively reactive gg
conformer. Figure 4 shows the results of glycosidation experiments. The α/β ratios were
calculated from each of the isomers isolated after column chromatography.
Donor 1 bearing the DTBS group was glycosidated with acceptor 23 and the corresponding
disaccharide 24 was obtained in 94% isolated yield with prominent α-selectivity. This result was
6
-12
consistent with our previous works. Notably, the α-stereo-directing effect of the DTBS group
overrides the effect of neighboring group participation, which has been reliably used for
β(1,2-trans)-glycoside formation. We have attributed high α-stereoselectivity to steric effect
arising from bulky tert-butyl substituents of the DTBS group. Thus, silylene-type donor 2 was
examined, in which both tert-butyl substituents of 1 were replaced with less bulky isopropyl
groups. The reaction gave glycosidated product 25 in 99% yield with 79:21 (α/β)
stereoselectivity, which was lower than the α-stereoselectivity obtained with the tert-butyl
groups. Thus, the bulk of the substituents in the silylene structure played a role in the observed
stereoselectivity. Next, to assess the steric effect of substituents on the cyclic protecting group,
1
0
ACS Paragon Plus Environment

Page 11 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
cyclic acetal-type donors 3–5 were examined, the steric bulk of which decreased in the order of
2
4
3 (steric parameter E
s
: EsCMe3 = –1.54) to 4 (EsMe = 0.00) and 5 (EsH = 1.24).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
S)-tert-Butylmethylidene donor 3 afforded the corresponding disaccharide 26 in 93% yield
RO
OR
NIS
TfOH
O
OH
O
GalN donors 1–15
or
BnO
BnO
AcO
PhthN
BnO
O
+
OSE
CH Cl
O
OAc
2
2
GlcN donor 16
OSE
4Å MS
0 ºC
AcO
2
3
OAc
[1.5 equiv]
[1.0 equiv]
Disaccharides 24–39
Si
O
Si
O
O
O
O
O
O
O
O
O
O
O
BnO
SPh
BnO
SPh
NPhth
BnO
SPh
NPhth
BnO
SPh
NPhth
NPhth
1
2
3
4
2
4: 94% (100 : 0)
25: 99% (79 : 21)
26: 93% (36 : 64)
27: 85% (29 : 71)
NO2
OMe
O
O
O
O
O
O
O
O
O
O
O
O
BnO
SPh
BnO
SPh
BnO
SPh
BnO
SPh
NPhth
NPhth
NPhth
NPhth
5
6
7
8
28: 74% (16 : 84)
29: 97% (13 : 87)
30: 65% (15 : 85)
31: 92% (32 : 68)
Cl
O Si
Si
Si
O
O
Si
O
O
O
O
O
O
O
O
O
O
BnO
SPh
BnO
SPh
NPhth
BnO
SPh
NPhth
BnO
SPh
NPhth
NPhth
9
10
33: 99% (21 : 79)
11
12
32: 93% (20 : 80)
34: 97% (28 : 72)
35: 73% (8 : 92)
O
O
O
O
O
O
AcO OAc
O
Si
O
O
O
BnO
SPh
NPhth
BnO
SPh
NPhth
BnO
SPh
NPhth
BnO
SPh
NPhth
1
3
14
37: 79% (9 : 91)
15
38: 97% (5 : 95)
16
3
6: 98% (9 : 91)
39: 97% (0 : 100)
Figure 4. Glycosidation of GalN/GlcN donors 1–16 with glucosyl acceptor 23.
Percentages are isolated yields of the corresponding disaccharides obtained after
1
1
column chromatography. Values in parentheses are α/β stereoselectivities calculated
ACS Paragon Plus Environment
based on the yield of isolated isomers.

The Journal of Organic Chemistry
Page 12 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
along with 36:64 (α/β) stereoselectivity. Similarly, (S)-ethylidene donor 4 and methylidene
donor 5 were coupled with 23, giving 27 in 85% yield (α/β = 29:71) and 28 in 74% yield (α/β =
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
6:84), respectively. These results strongly suggested that formation of the α-product could
decrease as the steric bulk was decreased; in other words, the stereoselectivity was most likely
to relate to the bulkiness of substituents at C4 and C6. Importantly, comparison of 1 with 3
showed the significance of the silylene structure for obtaining high α-selectivity, because both
the compounds contained a tert-butyl substituent as a source of steric hindrance, but it was
connected to silicon in 1 and carbon in 3. This suggests that the spatial position of the tert-butyl
group in the oxocarbenium intermediate and in the ground state might give rise to the
6
stereoselectivity. Based on X-ray crystallographic data of the DTBS donor, the endo-tert-butyl
group in 1 is closer to the anomeric carbon compared to an endo-substituent of general cyclic
acetal protecting groups. When 1 is activated to form the corresponding oxocarbenium
intermediate, the tert-butyl group should move much closer to the anomeric carbon and prevent
β-face nucleophilic attack by the glycosyl acceptor, leading to the observed α-selectivity. In
contrast, the tert-butyl group in 3 is expected to be positioned far from the anomeric carbon in
the oxocarbenium intermediate and would not sufficiently block the β-face. Next, we
investigated the effect on stereoselectivity of the electronic environment at O4 and O6 of the
glycosyl donor, because we expected through-space electron donation to contribute to the
stereoselectivity. Glycosidation of benzylidene donor 6 with 23 proceeded smoothly and gave
the corresponding disaccharide 29 in excellent yield with 13:87 (α/β) stereoselectivity.
Aryl-substituted benzylidene-type donors, namely, 7 with an electron-withdrawing group
(
2
p-NO ) and 8 with an electron-donating group (p-OMe), were then examined for their
1
2
ACS Paragon Plus Environment

Page 13 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
stereoselectivity. Donor 8 had clearly higher α-selectivity compared with the parent benzylidene
donor 6, whereas 7 showed no difference in stereoselectivity compared with 6. These results
could be a consequence of the different through-space stabilization of positive charge at the
anomeric center affecting the position of substituents on the cyclic protecting group upon
formation of the oxocarbenium intermediate stage. To further test this hypothesis, we designed
donor 9 with electronegative chlorine atom on the ethylidene moiety; then, we compared the
results with those for 4 (85% yield, α/β = 29:71). Decreased α-selectivity was observed for 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
93% yield, α/β = 20:80), even though its chlorine atom (EsCH2Cl –0.24) is bulkier than the
hydrogen atom (EsCH3 0.00) on 4. Next, to examine the stereoselectivity arising from ring size
formed by cyclic protecting groups, donors 10 and 12 with larger ring sizes were investigated.
First, donor 10 with a TIPDS group, which is bulkier than the DTBS group, was subjected to
glycosidation with 23, producing disaccharide 33 in 99% yield with 21:79 (α/β) stereoselectivity.
This experiment eliminated the possibility that the α-selectivity was due to steric effects alone.
Second, the glycosidation of xylylene-protected donor 12, which contains a nine-membered ring,
gave the β-linked disaccharide predominantly (α/β = 8:92). We then compared acyclic-type
donors 11 and 13 with 10 and 12, respectively. 4,6-Di-O-TBS-protected donor 11 gave
stereoselectivity similar to that of 10, and 4,6-di-O-Bn-protected donor 13 gave stereoselectivity
nearly identical to that of 12. In additional experiments, 4,6-di-O-Me-protected donor 14 with
less steric bulk and donor 15 with acyl protecting groups on both O4 and O6 were examined. As
expected based on reported findings in carbohydrate chemistry, both the donors showed a strong
preference for the formation of the β-product. Finally, 4,6-O-DTBS-protected GlcN donor 16
was coupled with 23 under reaction conditions identical to those used with the GalN donors.
1
3
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 14 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The corresponding disaccharide 39 was obtained exclusively as the β-product with perfect
stereoselectivity. From these experiments using donors 10–16, we can conclude that just the
presence of a silicon atom and a cyclic structure was not sufficient for α-selectivity. The
requirements for marked α-selectivity were a silylene structure forming a six-membered ring
and distinct steric hindrance due to bulky substituents.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Encouraged by the results of these experimental studies, we next shifted our attention to a
(a)ꢀ
(b)ꢀ
(c)ꢀ
Si
O
O
H C
O
3
O
O
N
O
O
6ꢀ
4ꢀ
5ꢀ
2ꢀ 1ꢀ
4H3ꢀ
3
ꢀ
Figure 5. Optimized structure of the oxocarbenium intermediate liberated from a
DTBS-protected galactosamine donor obtained by DFT calculations
(
M062X/6-31G*). (a) Chemical structure produced in ChemDraw. (b) Ball-and-stick
representation of the whole structure (c) Ball-and-stick representation of the sugar
ring in a partial structure.
computational study to examine the geometry of the postulated oxocarbenium intermediates.
Density functional theory (DFT) calculations were performed to optimize the structures. Figure
5
3
shows the optimized structure of the oxocarbenium intermediate formed from phenyl
-O-acetyl-2-deoxy-4,6-O-di-tert-butylsilylene-2-phthalimide-1-thio-β-D-galactoside. The
DFT-optimized geometry revealed that the sugar ring with the DTBS group at the C4 and C6
1
4
ACS Paragon Plus Environment

Page 15 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
4
positions adopts the H
3
conformation in the oxocarbenium intermediate. We also calculated the
Mulliken and natural charges on atoms in selected glycosyl donors, considering that the
electronegativity of oxygen atoms at C4 and C6 might affect the through-space electron
donation. As shown in Figure 6, oxygen atoms next to a silicon atom (1, 2, 10, and 11) have a
higher electron density compared with oxygen atoms next to a carbon atom (3 and 6). This
result suggests the stronger through-space electron donation from oxygen atoms next to a
silicon atom means that substituents on O4 and/or O6 are closer to the anomeric center, leading
to increased formation of the α-product.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Taking together the experimental results and the computational results, we next propose an
Figure 6. Calculated Mulliken charges (top numbers) and natural charges (bottom
numbers) on O4 and O6 of glycosyl donors 1, 2, 3, 6, 10, and 11.
improved reaction mechanism for the DTBS-directed α-galactosylation (Figure 7). First, upon
1
5
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 16 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
activation of the phenylthioglycoside under the NIS-TfOH promoter system, the buildup of
positive charge on the sulfur atom requires electrons to be stabilized. When the sugar ring
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
adopts the C
1
conformation, there is inefficient overlap to provide electron density from O5 to
1
,4
the C1 antibonding orbital. In contrast, when the ring adopts the B conformation, the
anti-periplanar electron pair on O5 can participate as in the E1 elimination reaction. Considering
1,4
steric effects alone, there is not enough flexibility to provide access to the B conformation
4
from the C
1
conformation due to steric interactions between the pseudoaxial sulfonium ion and
4
1,4
the bulky DTBS group. However, a conformational flip from the C
explained by stereoelectronic effects. Subsequent departure of the benzenesulfenyl iodide
ultimately giving PhSSPh and I ) generates the oxocarbenium intermediate. Importantly,
although typical pyranosyl oxocarbenium intermediates can usually as adopt a variety of
1
to the B could be
(
2
4
4
3
1
25
3 3 4 5
conformations, such as E, E , H , H , S , and B2,5, here the conformational freedom of the
4
3
sugar ring is limited to the half-chair H conformer because other conformers are unattainable
because of the fused ring system formed by the DTBS group. Once the oxocarbenium
intermediate is generated, the C2 phthalate group will interact with the electron-poor anomeric
center through neighboring group participation, thus producing a dioxolenium ion. A
preliminary computational study suggested that no reaction pathway could exist without
neighboring group participation due to the phthalate group. Usually, 1,2-trans glycosides are
produced by nucleophilic attack by an alcohol acceptor from β-face onto this sort of
dioxolenium ion. In the DTBS-directed galactosylation, however, the dioxolenium ion would be
cleaved by through-space electron donation because both O4 and O6 are in an axial orientation.
In other words, the coplanar alignment of the O4-C4 and O6-C6 bonds with the empty p-orbital
1
6
ACS Paragon Plus Environment

Page 17 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
generated on C1 inevitably causes electrostatic attraction between the partially negatively
charged oxygen atom of the C4 alkoxy groups and the positively charged carbon atom of the
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
6
27
28
oxocarbenium ion. The Woerpel, Bols, and Crich groups have independently reported the
importance of the electrostatic effect of both the axially oriented C4 alkoxy and the exocyclic
C5 alkoxymethyl groups in stabilizing the oxocarbenium ion. Breaking the five-membered ring
4
in the dioxolenium ion would re-form the oxocarbenium ion in the H
3
conformer, which can
then undergo nucleophilic attack by the alcohol acceptor, predominantly from the exo side
α-face) because endo attack on the β-face is blocked by the substantial steric hindrance of the
(
tert-butyl group. Endo or exo nucleophilic attack can determine the stereochemical outcome in
the glycosylation. First, endo attack and exo attack generate the distinct exploded transition
states A and B, respectively. Next, transition states A and B can access a twist-boat-like
conformer (C) and a chair-like conformer (D), respectively, to maximize orbital overlap
between the approaching alcohol acceptor and the developing lone pair on oxygen, resulting in
3
29
glycosidated products where the resulting sp -hydridized orbitals are anti-coplanar. The
chair-like conformer (D) is more stable than C, which is in a kinetically disfavored
twist-boat-like conformation that must undergo a ring-flip to reach the more stable equatorial
β-product. Moreover, C clearly suffers from unfavorable steric clash between the tert-butyl
substituent and the incoming alcohol acceptor. Hence, the nucleophilic attack predominantly
occurs via the pathway leading to the α-product. Importantly, the steric hindrance of the DTBS
group can hamper the β-face approach of even the smallest alcohol, methanol, giving rise to
7
predominantly the α-linked methylglycoside. This study has demonstrated that the preferential
4
formation of the H
3
oxocarbenium intermediate due to the conformational constraints imposed
1
7
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 18 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
by the DTBS group and subsequent access to the kinetically stable chair-like transition state
provide high stereoselectivity. Indeed, the powerful steric effect of the DTBS group is the most
important factor in the observed α-selectivity. Whitfield and co-workers have reported that, to
control the stereochemistry of glycosylations, glycosyl donors that can form face-discriminated
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
0
cations and access only one ring conformation should be used. Furthermore, they proposed
that bridged donors are one possible approach. Our results support their account of
stereoselective glycosylation.
1
8
ACS Paragon Plus Environment

Page 19 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
H
O
I
Si
Si
I
Si
N
O
O
Ph
S
O
O
O
O
O
I
O
O
O
PGO
SPh
S
overlap
Ph
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
PhthN
PhthN
PhSSPh
I
2
4C1
1,4B
Through-space
electron donation
Si
Si
empty
p-orbital
Si
O
O
O
O
O
O
O
O
O
O
N
O
O
N
N
O
Neighboring group
participation
O
O
Breaking
five-membered ring
4
Oxocarbenium ion ( H
3
)
Dioxolenium ion
Nucleophile
ROH)
(
endo attack
(β-face)
Si
Si
O
O
O
O
O
O
O
N
O
O
N
4H
3
O
exo attack
(
α-face)
Si
Si
O
H
O
O
O
R
O
O
O
N
O
O
Exploded TS
N
O
R
O
O
H
A
B
Steric conflict
Si
Si
O
O
RO
O
overlap
PhthN
O
O
O
PhthN
RO
overlap
Chair
more stable)
C
Twist-boat
less stable)
D
(
(
Disfavored
Favored
Si
Si
Si
Si
O
O
O
O
O
RO
O
PGO
RO
O
O
ring-flip
O
PGO
PhthN
O
OR
O
PGO
PGO
O
PhthN
PhthN
PhthN
OR
α-product
major)
β-product
minor)
(
(
Figure 7. Proposed mechanism of DTBS-directed α-galactosylation.
1
9
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 20 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Si
Si
O
O
O
O
Ag-silicate
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
O
O
TrocO
TrocO
OR
TrocHN
TrocHN
Br
β-glycoside (Major product)
Si
Si
O
Si
SN2-like
O R
O
O
R
R
δ
O
O
O
O
O
O
H
O
O
H
δ
TrocO
TrocHN
H
O
R'
NH Brδ
HN
Br
R'
Br
+
O
δ
+
Ag
Ag
+
Ag
insoluble Ag-silicate
Figure 8. Proposed mechanism of β-selective glycosidation using a glycosyl bromide
donor with activation by insoluble silver silicate.
7
We have previously made two intriguing findings regarding DTBS-directed α-galactosylation.
One is the exceptional case where β-selectivity (6:77, α/β ratio) is obtained by using glycosyl
31
bromide donor with activation by insoluble silver silicate. This remarkable result can be
rationalized by an S
on insoluble silver silicate and the nucleophilic attack of the alcohol acceptor mainly occurs
along an equatorial trajectory via an S 2-like pathway where the oxocarbenium ion is formed
N
2-like reaction mechanism. As illustrated in Figure 8, the reaction proceeds
N
but remains linked to silicate anionic surface, in such a way that the nucleophile (HO-R) can
only attack by/through the β-face, predominantly forming the β-glycoside. This result suggests
that the generation of the "distinct" oxocarbenium ion is required for the stereo-directing effect
of the DTBS group.
Our second intriguing finding is the formation of 2-OH α-galactoside as a byproduct when
2
-O-benzoyl-protected galactosyl donors were used. As expected, the DTBS-protected
2
0
ACS Paragon Plus Environment

Page 21 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
galactosyl donors gave predominantly α-glycosides in >70% yield. However, non-negligible
amounts (~15%) of the corresponding 2-OH α-galactosides were also observed as by-products
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
Figure 9). This phenomenon could be explained by a route involving the formation of an
orthoester derivative. Usually, the benzoyl group is selected instead of the acetyl group to
prevent the formation of the orthoester and it works well in most cases. However, coexistence
of the benzoyl and DTBS group may make access of the alcohol acceptor to anomeric center at
the oxocarbenium ion stage limited due to the steric clash, allowing for the nucleophilic attack
at relatively less hindered carbonyl carbon on the benzoyl group at C2 to form the orthoester
derivative. In addition, nucleophilic attack by the alcohol acceptor at the carbonyl carbon C2,
giving the orthoester, could occur from the exo side because of the higher energy barrier to the
3
0
endo attack. Some of the orthoester formed could then be cleaved to the 2-OH oxocarbenium
ion to release the benzoylated acceptor and form the 2-OH oxocarbenium ion, which is then
glycosidated with the acceptor to furnish the 2-OH α-glycoside. We have not detected the
benzoylated acceptor in the reaction mixture so far but the formation of orthoester derivatives
has been actually observed. When 2-O-benzyl-protected DTBS-protected galactosyl donors
were used, the corresponding α-glycosides were obtained in excellent yields without the
3
2
formation of the 2-OH byproduct.
2
1
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 22 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Si
Si
Si
O
O
O
O
O
O
ROH
O
O
O
+
BzO
L
BzO
BzO
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
BzO
OR
HO
OBz
OR
α-glycoside
71–86%)
2-OH α-glycoside
(ca. 10%)
(
activation
Si
Si
Si
O
O
O
O
O
O
O
H
R
O
O
O
O
O
O
O
O
O
minor
Ph
Through-space
electron donation
major
Si
Si
O
O
O
+ ROH
O
O
BzO
O
O
Ph
O
O
O
O
R
O
O
O
α-face attack
predominantly)
Ph
Orthoester derivative
(
H
O
Si
BzO
Si
O
O
O
O
BzO
RO
O
O
H
O
BzO
R
O
BzO
O
Ph
BzO
OR
α-product
Major product
BzOR
tBu
Si
O
Si
O
O
O
O
+ ROH
BzO
O
HO
OR
HO
Minor product
Figure 9. Proposed mechanism for the formation of 2-OH derivatives in DTBS-directed
α-glycosidation using 2-O-benzoyl-protected galactosyl donors.
2
2
ACS Paragon Plus Environment

Page 23 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Conclusions
By experimental and computational studies, we have elucidated the origin of the α-selectivity in
DTBS-directed α-galactosylation. Glycosidation experiments using various GalN donors, which
differed in terms of only the protecting group at the C4 and C6 positions, demonstrated that the
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
presence of the DTBS group was essential for strong α-selectivity. The following main
33
requirements were found for α-selectivity. (1) Generation of the oxocarbenium ion:
well-defined S
A
N
1-like mechanism that produces the oxocarbenium ion results in the
electron-poor anomeric center necessary for modulating the stereoselectivity. (2) Galacto-type
donors with cyclic protecting group bridging O4 and O6 to form six-membered ring: These
4
donors preferentially adopt the H
3
conformation upon formation of the oxocarbenium ion,
allowing for nucleophilic attack by the glycosyl acceptor preferentially from exo side (α-face)
because of both stereoelectronic and steric effects. (3) Through-space electron donation: The
axially oriented C4 alkoxy and the exocyclic C5 alkoxy groups exert a powerful electrostatic
effect on the empty p-orbital that is formed after liberation of the leaving group, thus stabilizing
the oxocarbenium ion. Moreover, the effect is strengthened by the C4-O4 and C6-O6 bonds
being coplanar with the empty p-orbital. This effect is maximized when the donor has a cyclic
protecting group enabling the formation of a six-membered ring. (4) Steric hindrance due to
bulky alkyl substituents appended to cyclic protecting group: Steric bulk hinders nucleophilic
attack on the β-face and also affects the stability in the transition state. (5) 4,6-O-Silylene
structure: X-ray crystallographic analysis of the DTBS donor showed that one of the bulky alkyl
groups in the silylene structure is positioned very close to the anomeric carbon, even in the
ground state. The substituent can be positioned much closer to the anomeric carbon in the
2
3
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 24 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
oxocarbenium intermediate, resulting in severe steric clash during the glycosylation (this is
supported by the DFT calculations).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The DTBS group fulfills all these requirements well. These characteristics of the DTBS group
may explain the high stereoselectivity of DTBS-directed α-galactosylation. Furthermore, of all
the cyclic protecting groups examined, only the DTBS group provided strong α-selectivity.
Thus, we suggest that the unique and specific effect of the DTBS group in α-selective
galactosylation be called the “DTBS effect”. Encouraged by the discovery of the DTBS effect,
we are working on the development of stereoselective glycosylations using other ring-restricted
glycosyl donors, the results of which will be reported in due course.
Experimental Section
General procedure for chemical synthesis: All reactions were carried out under a positive
pressure of argon unless otherwise noted. All chemicals were purchased from commercial
suppliers and used as received without further purification. Molecular sieves were dried at
3
00 °C for 2 h in a muffle furnace prior to use. Solvents for reaction media were dried over
molecular sieves and used without purification. Thin-layer chromatography (TLC) was
performed on TLC plates (silica gel 60F254 on glass plate). Compounds were detected on TLC
2 4
plates by either exposure to UV light (2536 Å) or development in 10% H SO in ethanol
followed by heating. Silica gel (80 mesh and 300 mesh) was used for flash column
chromatography. The quantity of silica gel was usually 100 to 200-fold the weight of the crude
sample to be purified. Solvent systems for chromatography are specified in volume ratios.
2
4
ACS Paragon Plus Environment

Page 25 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
1
13
Evaporation and concentration were carried out in vacuo. H-NMR and C-NMR spectra were
1
recorded on a 400/500/600 MHz spectrometer. Chemical shifts in H-NMR spectra are
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
expressed in ppm (δ) relative to the Me Si signal adjusted to δ 0.00 ppm. Data are presented as
follows: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, dd = double
doublet, td = triple doublet, m = multiplet and/or multiple resonances), coupling constant in
1
1
hertz (Hz), integration, and position of the corresponding proton. H- H COSY method was
used to confirm the NMR peak assignments. High-resolution mass spectra (ESI-TOF MS) were
obtained. Optical rotations were measured with a high-sensitivity polarimeter.
Computational details: All DFT calculations were carried out in the GAUSSIAN09 software
34
35
package. We chose the M062X exchange-correlation functional and the 6-31G* electronic
basis set for calculations. Initial structures for the calculations were built with the aid of
36
GaussView 5 visualization software. To investigate the desired reactions, we
constructed a six-membered ring as chair-like form in the initial structure. The
geometries of complexes were fully optimized at the M062X/6-31G* level. We also performed
normal mode analysis to characterize the optimized structures and confirmed that all optimized
structures have no imaginary frequencies.
2
5
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 26 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Typical procedure for glycosidation experiments: Molecular sieves, 4Å (100 mg per 1 mL of
solvent) were added to a mixture of the glycosyl donor (1.5 equiv with respect to the acceptor)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2 2
and acceptor 23 in CH Cl at room temperature. After stirring for 30 min, the mixture was
cooled to 0 °C and N-iodosuccinimide (2.0 equiv with respect to the donor) and TfOH (0.2
equiv with respect to the donor) were added. The mixture was stirred for 30 min to 1 h while
monitoring the reaction by TLC. Once all the acceptor was consumed, the reaction was
quenched by the addition of saturated aq. NaHCO
3
. The reaction mixture was diluted with
CHCl and filtered through Celite. The filtrate was then washed sequentially with saturated aq.
3
3 2 2 3 2 4
NaHCO , saturated aq. Na S O , and brine. The organic layer was collected, dried over Na SO ,
and concentrated. The resulting residue was purified by silica gel column chromatography to
give the glycosylated products.
Phenyl
3
-O-benzyl-2-deoxy-4,6-O-di-tert-butylsilylene-2-phthalimide-1-thio-β-D-galactopyranosid
e (1). To a solution of 18 (200 mg, 0.407 mmol) in pyridine (2.0 mL) was added
di-tert-butylsilyl bis(trifluoromethanesulfonate) (223 µL, 0.611 mmol) at 0 ºC. After stirring for
1
h at rt, the completion of the reaction was confirmed by TLC (1:3 EtOAc–n-hexane). The
reaction was quenched by the addition of dry methanol at 0 ºC. Solvents were removed by
co-evaporation with toluene, and then the residue was diluted with CHCl , washed with 2M HCl,
3
2 3 2 4
H O, satd aq NaHCO , and brine. The organic layer was subsequently dried over Na SO ,
filtered, and concentrated. The resulting residue was purified by silica gel column
chromatography (1:5 EtOAc–n-hexane) to give 1 (236 mg, 92%) as a colorless viscous
2
6
ACS Paragon Plus Environment

Page 27 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
1
compound. [α]
D
+112.0 (c 1.0, CHCl
3
); H NMR (500 MHz, CDCl
3
): δ = 7.87–7.07 (m, 14 H,
NPhth, 2 Ph), 5.54 (d, J1,2 = 10.5 Hz, 1 H, 1-H), 4.87 (t, J2,3 = 10.7 Hz, 1 H, 2-H), 4.64 (d, Jgem
=
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
3
2.3 Hz, 1 H, PhCH
2
2
), 4.60 (d, J3,4 = 2.7 Hz, 1 H, 4-H), 4.44 (d, 1 H, PhCH ), 4.27 (dd, 1 H,
t
13
-H), 4.25 (m, 2 H, 6a-H, 6b-H), 3.50 (s, 1 H, 5-H), 1.20, 1.07 (2 s, 18 H, 2 Bu); C NMR (100
MHz, CDCl
3
): δ = 168.4, 167.6, 137.9, 134.1, 133.9, 133.6, 132.0, 131.8, 128.8, 128.3, 127.6,
1
27.4, 123.6, 123.1, 84.9, 75.6, 75.2, 70.0, 69.4, 67.5, 50.7, 27.7, 27.6, 23.5, 20.8. HRMS
+
6
(ESI-TOF): m/z calcd for C35H41NO SSi: 654.2316 [M+Na] ; found: 654.2318.
Phenyl
3-O-benzyl-2-deoxy-4,6-O-diisopropylsilylene-2-phthalimide-1-thio-β-D-galactopyranoside
2). To a solution of 18 (187 mg, 0.380 mmol) in DMF (3.8 mL) were added diisopropylsilyl
bis(trifluoromethanesulfonate) (170 µL, 0.580 mmol) and triethylamine (370 µL, 2.66 mmol) at
ºC. After stirring for 10 min at rt, the consumption of the starting material was confirmed by
(
0
TLC (2:3 EtOAc–n-hexane). The reaction was quenched by the addition of dry methanol at 0 ºC.
Solvents were removed by co-evaporation with toluene, and then the residue was diluted with
3 2 2 4
CHCl , washed with H O and brine. The organic layer was subsequently dried over Na SO ,
filtered, and concentrated. The resulting residue was purified by silica gel column
chromatography (1:5 EtOAc–n-hexane) to give 2 (198 mg, 86%) as a colorless viscous
1
compound. [α]
D
+108.5 (c 1.0, CHCl
3
); H NMR (400 MHz, CDCl
3
): δ = 7.90–6.99 (m, 14 H,
NPhth, 2 Ph), 5.53 (d, J1,2 = 10.5 Hz, 1 H, 1-H), 4.80 (t, J2,3 = 10.6 Hz, 1 H, 2-H), 4.63 (d, Jgem
2.4 Hz, 1 H, PhCH ), 4.51 (d, J3,4 = 2.7 Hz, 1 H, 4-H), 4.43 (d, 1 H, PhCH ), 4.27–4.17 (m, 3
H, 3-H, 6a-H, 6b-H), 3.53 (s, 1 H, 5-H), 1.26–1.03 (m, 14 H, 2 Pr); C NMR (100 MHz,
CDCl ): δ = 168.4, 167.4, 137.8, 134.0, 133.8, 132.9, 132.4, 131.7, 128.7, 128.2, 127.6, 127.6,
=
1
2
2
i
13
3
2
7
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 28 of 69
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
27.5, 123.6, 123.1, 84.4, 75.8, 75.3, 70.4, 69.2, 66.9, 50.7, 29.7, 17.1, 17.0, 16.9, 13.9, 12.7.
+
HRMS (ESI-TOF): m/z calcd for C33
Phenyl
-O-benzyl-4,6-O-tert-butylmethylidene-2-deoxy-2-phthalimide-1-thio-β-D-galactopyranosi
de (3). To a solution of 18 (50 mg, 0.105 mmol) in MeCN (349 µL) were added pivalaldehyde
46 µL, 0.418 mmol) and (±)-camphor-10-sulfonic acid (CSA) (5 mg, 21.0 µmol) at rt. After
stirring for 7.5 h at rt as the reaction was monitored by TLC (1:2 EtOAc–n-hexane, 1:15
MeOH–CHCl ), the reaction was quenched by the addition of triethylamine at 0 ºC and the
H
37NO
6
SSi: 626.2003 [M+Na] ; found: 626.2002.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
(
3
reaction mixture was then evaporated. The resulting residue was purified by silica gel column
chromatography (1:5 EtOAc–n-hexane) to give 3 (49 mg, 84%) as a white foamy compound.
1
[
α]
D
+2.3 (c 1.0, CHCl
3
); H NMR (500 MHz, CDCl
3
): δ = 7.85–7.07 (m, 14 H, NPhth, 2 Ph),
5
.53 (d, J1,2 = 10.4 Hz, 1 H, 1-H), 4.70 (t, J2,3 = 10.7 Hz, 1 H, 2-H), 4.54 (d, Jgem = 12.4 Hz, 1 H,
PhCH ), 4.38 (d, 1 H, PhCH ), 4.35 (dd, J3,4 = 3.5 Hz, 1 H, 3-H), 4.27 (dd, Jgem = 12.2 Hz, J5,6a
1.4 Hz, 1 H, 6a-H), 4.14 (s, 1 H, BuCH<), 4.05 (d, 1 H, 4-H), 3.81 (dd, J5,6b = 1.7 Hz, 1 H,
2
2
=
t
t
13
6
b-H), 3.49 (d, 1 H, 5-H), 0.99 (s, 9 H, Bu); C NMR (125 MHz, CDCl
3
): δ = 168.8, 167.0,
138.0, 134.0, 133.8, 132.8, 131.8, 131.8, 131.6, 128.7, 128.1, 127.7, 127.5, 127.3, 123.6, 123.1,
1
07.1, 83.0, 74.9, 72.1, 70.6, 69.2, 50.6, 35.2, 29.7, 24.7. HRMS (ESI-TOF): m/z calcd for
+
C
32
H
33NO
6
S: 582.1921 [M+Na] ; found: 582.1922.
Phenyl 3-O-benzyl-2-deoxy-4,6-O-ethylidene-2-phthalimide-1-thio-β-D-galactopyranoside
4). To a solution of 18 (281 mg, 0.572 mmol) in MeCN (1.9 mL) were added dimethylacetal
(182 µL, 1.72 mmol) and CSA (13 mg, 57.0 µmol) at rt. After stirring for 2.5 h at rt as the
reaction was monitored by TLC (1:30 MeOH–CHCl ), the reaction was quenched by the
(
3
2
8
ACS Paragon Plus Environment

Page 29 of 69
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
addition of triethylamine at 0 ºC and the reaction mixture was then evaporated. The resulting
residue was purified by silica gel column chromatography (1:2→2:3 EtOAc–n-hexane) to give
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
4
(273 mg, 92%) as a white foamy compound. [α]
D
–105.0 (c 1.0, CHCl
3
); H NMR (500 MHz,
CDCl
3
): δ = 7.71–7.00 (m, 14 H, NPhth, 2 Ph), 5.54 (d, J1,2 = 10.4 Hz, 1 H, 1-H), 4.75–4.68 (m,
2
3
H, 2-H, CH
3 2 2
CH<), 4.59 (d, Jgem = 12.6 Hz, 1 H, PhCH ), 4.37 (d, 1 H, PhCH ), 4.32 (dd, J3,4 =
.4 Hz, J2,3 = 10.7 Hz, 1 H, 3-H), 4.23 (dd, J5,6a = 1.5 Hz, Jgem = 12.2 Hz, 1 H, 6a-H), 4.07 (d, 1
H, 4-H), 3.83 (dd, J5,6b = 1.2 Hz, 1 H, 6b-H), 3.48 (d, 1 H, 5-H), 1.46 (d, JCH3CH<,CH3CH< = 5.0 Hz,
13
3
3 3
H, CH CH<); C NMR (125 MHz, CDCl ): δ = 168.5, 167.1, 137.6, 137.6, 134.0, 133.8,
132.8, 132.2, 131.8, 131.7, 128.7, 128.2, 128.0, 127.7, 127.7, 123.6, 123.1, 99.1, 83.5, 74.7,
7
6
2.1, 71.0, 70.1, 69.0, 50.7, 29.7, 21.1. HRMS (ESI-TOF): m/z calcd for C29H27NO S: 540.1451
+
[
M+Na] ; found: 540.1453.
Phenyl
-O-benzyl-2-deoxy-4,6-O-methylidene-2-phthalimide-1-thio-β-D-galactopyranoside
To a solution of 18 (200 mg, 0.407 mmol) in MeCN (814 µL) were added dimethoxymethane
2.53 mL, 28.5 mmol) and CSA (19 mg, 81.4 µmol) at rt. After stirring for 14 h at rt as the
reaction was monitored by TLC (1:30 MeOH–CHCl ), the reaction was warmed to 40 ºC and
3
(5).
(
3
the stirring was continued for 10 h at the same temperature. Then the reaction was warmed to 60
ºC and the stirring was continued for 10 h at the same temperature. The reaction was quenched
by the addition of triethylamine at 0 ºC and the reaction mixture was then evaporated. The
resulting residue was purified by silica gel column chromatography (1:1 EtOAc–n-hexane) to
1
give 5 (173 mg, 84%) as a white foamy compound. [α]
D
+0.2 (c 4.0, CHCl
3
); H NMR (500
3
MHz, CDCl ): δ = 7.88–7.00 (m, 14 H, NPhth, 2 Ph), 5.56 (d, J1,2 = 10.5 Hz, 1 H, 1-H), 5.28 (d,
2
9
ACS Paragon Plus Environment