Reciprocal donor acceptor selectivity (RDAS): A new concept for matching donors with acceptors

Article abstract of DOI:10.1139/v02-137

Lemieux's extensive work on replacement reactions at the anomeric center helped to establish the fact that the O-2-protecting group of a donor exerts powerful control over stereoselectivity in glycoside coupling reactions. This manuscript shows that the O-2-protecting group of a donor also exerts powerful, indeed sometimes total, control over regioselectivity in glycosidation of diols. The latter acceptors also exhibit preferences over the donor, thereby providing evidence for the concept of reciprocal donor acceptor selectivity (RDAS). The latter concept is put to the test by simultaneously presenting an acceptor diol with equivalent amounts of two donors, in the hope of achieving double differential glycosidation leading to one-pot assembly of a trisaccharide. When the pair of donors did not conform to RDAS principles the reaction did not proceed beyond a dissacharide. However, when the pair was RDAS sanctioned, a single trisaccharide (out of four possibilities) was obtained.

Full text of DOI:10.1139/v02-137

1075
Reciprocal donor acceptor selectivity (RDAS):
A new concept for “ matching” donors with
acceptors
Bert Fraser-Reid, J. Cristobal Lopez, K.V. Radhakrishnan, Mateusz Mach,
Urs Schlueter, Ana Gomez, and Clara Uriel
Abstract: Lemieux’s extensive work on replacement reactions at the anomeric center helped to establish the fact that
the O-2-protecting group of a donor exerts powerful control over stereoselectivity in glycoside coupling reactions. This
manuscript shows that the O-2-protecting group of a donor also exerts powerful, indeed sometimes total, control over
regioselectivity in glycosidation of diols. The latter acceptors also exhibit preferences over the donor, thereby providing
evidence for the concept of reciprocal donor acceptor selectivity (RDAS). The latter concept is put to the test by simul-
taneously presenting an acceptor diol with equivalent amounts of two donors, in the hope of achieving double differen-
tial glycosidation leading to one-pot assembly of a trisaccharide. When the pair of donors did not conform to RDAS
principles the reaction did not proceed beyond a dissacharide. However, when the pair was RDAS sanctioned, a single
trisaccharide (out of four possibilities) was obtained.
Key words: regiocontrolled glycosidation, armed and disarmed donors, di- and trioxolenium ions, oxocarbenium ion.
Résumé : Les nombreux travaux de Lemieux sur les réactions de remplacement au niveau du centre anomérique on
permis d’établir le fait que le groupe protecteur en position O-2 d’un donneur exerce un contrôle important sur le
stéréosélectivité des réactions de couplage des glycosides. Les résultats présentés dans ce travail montrent que le
groupe protecteur en position O-2 exerce aussi un contrôle important, et quelquefois total, sur la régiosélectivité de la
glycosidation des diols. Ces accepteurs présentent aussi des préférences par rapport au donneur et ceci correspond à
des indications en faveur du concept d’une sélectivité réciproque donneur accepteur (SRDA). On a évalué ce dernier
concept en présentant un diol accepteur à des quantités équivalentes de deux donneurs, en espérant obtenir une
glycosidation doublement différentielle qui conduirait à un assemblage monotope d’un trisaccharide. Dans les cas où la
paire de donneurs n’est pas conforme au concept de SRDA, la réaction ne procède pas au-delà du disaccharide.
Toutefois, lorsque la paire était conforme au concept de SRDA, on n’obtient qu’un seul trisaccharide (parmi les quatre
possibilités).
Mots clés : glycosidation, donneurs armés et désarmés, ions di- et trioxolénium, ion oxocarbénium.
[Traduit par la Rédaction] Fraser-Reid et al. 1087
Introduction
pation in organic chemistry (3). Thus Isbell’s work showed
that an O-2 acyl protecting group would foster 1,2-trans
glycoside synthesis (1J3) a process in which a dioxolenium
ion (2) was subsequently (4) implicated (Scheme 1a). The
corollary of this insight was that a nonparticipating group
would enhance the chances of obtaining the corresponding
1,2-cis stereoisomer (5).
By the time Lemieux began his studies on “Replacement
reactions at the anomeric center” (1), Isbell’s elegant ratio-
nalization for the stereoselectivity observed in reactions of
“acetyl glycoside halides” had been recognized as a monu-
mental mechanistic breakthrough (2) (Scheme 1a), which
was to give rise to the concept of neighboring group partici-
However, enhance did not mean ensure, as was already
evident to Lemieux from repeated failures to conquer su-
crose using fully benzylated donors, which had led instead
to excellent syntheses of “ꢀ-sucrose” (6). In this context,
Lemieux’s historic breakthrough utilizing Briggl’s anhydride
as a donor (7) (Scheme 1b) was seemingly contra-intuitive
since epoxide cleavage should (and usually did) proceed
from the ꢀ-face. However, prior investigations had given evi-
dence of O-6 involvement (Scheme 1b) when “large”
nucleophilic acceptors were used (8). In addition to being
another type of “neighboring group participation”,
Lemieux’s synthesis demonstrated the importance of
“matching” donor with acceptor, although the concept had
not been explicitly stated as such (see below).
Received 13 February 2002. Published on the NRC Research
Press Web site at http://canjchem.nrc.ca on 28 August 2002.
Dedicated to the memory, teachings, and wisdom of Ray
Lemieux, our Doktorvater/vater.
B. Fraser-Reid,1 K.V. Radhakrishnan, M. Mach, and
U. Schlueter. Natural Products and Glycotechnology
Research Institute Inc. (NPG), 4118 Swarthmore Road,
Durham, NC 27706, U.S.A.
J.C. Lopez, A. Gomez, and C. Uriel. Instituto de Quimica
Organica General (CSIC), Juan de la Cierva 3, Madrid 28006,
Spain.
1Corresponding author (e-mail: dglucose@aol.com).
Can. J. Chem. 80: 1075–1087 (2002)
DOI: 10.1139/V02-137
© 2002 NRC Canada

1076
Can. J. Chem. Vol. 80, 2002
Scheme 1.
Following Isbell’s lead, the O-2 protecting group has been
justifiably established as an implement for anomeric
stereocontrol — as is apparent from various textbooks deal-
ing with carbohydrate synthesis (9). Nevertheless, recent re-
ports (10) from our laboratory suggest that an O-2 protecting
group can also provide a ready, low-tech implement for
achieving regio control in glycosidation of polyhydroxylic
acceptors.
These observations have profound importance for match-
ing donor with acceptor, a concept, which has long been
championed by Paulsen et al. (11). However, that matching
modality is not to be confused with the subsequent, and now
common usage of the term for “double stereodifferentiation”
as introduced by Masamune et al. (12) in connection with
the coupling of two enantiopure partners. In this regard,
Masamune’s usage is ably demonstrated by Spijker and
VanBoeckel’s example (13) showing that a given acceptor
gave better yields with a D donor than its L enantiomer. On
the other hand, Paulsen’s matching concept has also demon-
strated in VanBoeckel’s collaboration with Petitou (14) on
heparin analogs, in which prodigious preliminary experi-
mentation served to reveal the best “match” for each linkage
of the oligosaccharide array.
alkylation and (or) acylation reagent in question. For exam-
ple, the (presumably) “more reactive” equatorial C-6-OH
was rejected by the more reactive benzylating agent.
We were interested to see if the regioselectivity shown for
4 in Scheme 1c could be extended to glycosidation reac-
tions. For this experiment, Lemieux’s procedure (16) al-
lowed ready preparation of orthoesters such as 5 (17)
(Scheme 2). Acid-catalyzed rearrangement gave the dis-
armed 2-O-acyl donor (NPGAC) 6, which was transformed
by two routine reactions into the armed (18) counterpart 7.
Compounds 8–10 were obtained similarly. Thus, a variety of
n-pentenyl donors, armed and disarmed (18) to different ex-
tents, were available which could be made to react under
similar conditions on a level playing field, so that valid com-
parisons could be drawn.
When diol 4 was treated with 1 equiv of the manno
NPGAC 6a, the only compound obtained was the pseudo
disaccharide 11a (Scheme 3a and Table 1, entry i). The
NPOE 5a followed similar regioselectivity giving 11a exclu-
sively — indeed in the better yield of 73% (Table 1, entry ii).
By contrast, the armed counterpart 7a, gave a mixture of
regioisomers 11b and 12b in a 1:3 ratio (Table 1, entry iii).
In all of these studies (above and below), assignment of
the regiosomeric structures was readily accomplished by
acylating the material in question, and determining the cou-
plings of the newly downshifted proton (these are shown in
the Experimental section). For example in 11a, acetylation
gave rise to a double doublet at 5.84 ppm (J = 2.8 and
2.4 Hz) readily assignable to the inositol’s H-2. By contrast
the benzoate of 12 gave rise to a double doublet at 5.76 ppm
(J = 9.6 and 10 Hz) as required by the inositol’s H-6.
The trends in Scheme 3a for inositol 4 prompted us to ex-
amine the comparable mannoside diol 13. The results were
Results
Our interest in this question of regioselectivity emanated
from the observation reported in 1998 (15) that the myo-
inositol diol 4, underwent acylation at C-6-OH exclusively,
but alkylation primarily at C-2-OH, when a unimolar amount
of the pertinent reagent was used (Scheme 1c). These differ-
ences in regioselectivity could not be rationalized by imput-
ing “more or less” reactivity to either the OH or to the
© 2002 NRC Canada

Fraser-Reid et al.
1077
Scheme 2.
even more striking in that both the NPOE 5b and the armed
NPGALK 7b, reacted to give respectively, 14 only, and 15
mainly.
That these remarkable regioselectivities were not confined
to manno-NPOEs was established by showing that the gluco-
and galacto-NPOEs 8b and 8c, were also regioselective for
the C-6-OH of 4 (Scheme 3c) giving disaccharides 16a and
16b in 76% and 73% yields, respectively.
The results in Scheme 3a and 3b (Table 1, entries i–v)
challenged our assumptions about which hydroxyl group of
compounds 4 and 13 would (should!) be the more reactive.
Thus, if the accepted wisdom correlating reactivity and ori-
entation were applied, the equatorial-OH of each should
have been glycosylated more readily by the armed donor 7,
than by the disarmed counterpart 5.
We therefore decided to examine the merits of a steric
hindrance-based rationalization by employing altroside 17
(Scheme 4), the two hydroxyl groups of which present an
unequivocal choice of steric environments. We found, as in-
dicated in Scheme 4, that the armed donor 7b showed little
discrimination between the two sites of 17, although the
unencumbered C-2-OH should have been a clear favorite. In-
stead, the C-2-mannosylation product 19 was minor, whereas
the C-3-regioisomer 18a was major in a 1:2 ratio) (Table 1,
entry vi).
The anomeric glucosides 20 and 23 were logical choices
since the relative reactivities of the hydroxyl groups have
been frequently tested with respect to acylating and
alkylating reagents (19). According to the results of some
earlier glycosidation studies (20, 21), there would seem to
be very little difference in reactivity between the C-2- and
C-3-OH groups. To leverage these slight differences, the re-
actions were begun at –78°C and allowed to warm to room
temperature gradually. Under these non-forcing conditions,
the donors tended to decompose, and so substantial amounts
of the acceptor diols were recovered. Nevertheless, the reac-
tivities were clearly discernible as shown in Scheme 5. Thus,
the ꢁ-diol 20, reacted with the manno-NPOE to give 21a and
22a in a 7.4:1 ratio. Similarly, the gluco-NPGAC donor 9,
gave overwhelming preferences for the C-2-OH producing
21b and 22b in a 7.6:1 ratio. On the other hand, with the
“armed donors” NPG 7b and 10, the regioselectivities were
unremarkable, sending out mixed messages about their pref-
erences. Thus, C-3-OH was favored in the former giving
mainly 22c (22) with armed 7b, but C-2-OH was favored
(slightly) with the armed donor, the latter 10 giving mainly
21d (22) (Table 1, entries xii and xiii).
Comparison of the “same” reaction in entries x and xi of
Table 1 is instructive. The yield at the higher temperature is
nearly twice as high as at the lower temperature, but with a
50% sacrifice in regioselectively.
For the ꢀ-anomer 23, the regioselectivities were less pro-
nounced; however, NPOE 5b and NPGAC 9 were consistent
in their preference for C-2-OH (Table 1, entries xiv and xv),
On the other hand, the very hindered C-3-OH was se-
lected exclusively by NPOEs 5b and 5c, to give 18b and
18c, respectively, reaching quantitative proportions (92%)
with the NPOE 5c. (Table 1, entries vii and viii).
© 2002 NRC Canada

1078
Can. J. Chem. Vol. 80, 2002
Scheme 3.
whereas NPGALK 7b and 10, again sent mixed messages in
their preferences for C-2-OH and (or) C-3-OH (Table 1, en-
tries xvi and xvii).
selectivities. The formation of 18 and 19 therefore reveal the
interplay of reciprocal donor acceptor selectivity (RDAS).
Although the concept of RDAS emanated from
regioselective preferences observed in a series of diols (see
Table 1), the concept also has profound implications for
glycosidic coupling, even where the acceptor has only one
hydroxyl group. Unfortunately, gathering pertinent literature
examples is problematic because (i) reactions that fail, or
proceed in unflattering yields, are rarely reported; (ii) forc-
ing conditions may have been used to eke out the coveted
success, instead of going-with-the-flow; (iii) yields may
have been enhanced by changing reagents and (or) condi-
tions, thereby vitiating the ability to draw comparisons.
Fortunately, reports from van Boom’s laboratory provide a
vivid example of RDAS (23, 24) in which failure and suc-
cess related to a specific target, similar reactants and condi-
tions having been maintained. Thus, in attempting to prepare
the trisaccharide 27 (Scheme 6), the 2-O-benzoyl thio-
glycoside donor 26, reacted with disaccharide 11, to provide
only a 10% yield of the desired target material (23). How-
ever, when the same donor (26) was exposed to disaccharide
12, the target was produced in the excellent yield of 84%
(24). The results clearly indicated that the disarmed donor
Discussion
The concept of reciprocal donor acceptor selectivity
(RDAS)
The data above provide a new angle to the donor–acceptor
matching principle adumbrated by Paulsen and co-workers
(5, 11). The regioselectivities observed in Schemes 3–5
clearly cannot be accommodated by traditional explanations
based on reactivity and (or) steric hindrance.
This conclusion is dramatically exemplified in the results
of altroside 17 (Scheme 4 and Table 1, entries vi–viii) where
the exclusive preference of the C-3-OH for NPOEs 5b and
5c, a remarkable 92% in the case of the latter, must be re-
garded as surprising. But even more surprising is the fact
that the unencumbered C-2-OH of 17 underwent no detect-
able coupling with NPOE 5b or 5c.
Thus, the results in Scheme 4 implied that the donor as
well as the acceptor were expressing strong independent
© 2002 NRC Canada

Fraser-Reid et al.
1079
Table 1. Regioselective glycosidation of acceptor diols.
Scheme 4.
26, was “matched” with the C-6-OH of 12, but not the C-2-
OH of 11.
The concept of RDAS suggests that one-pot, target-
oriented, double glycosidations should be possible when ap-
propriate partners are employed. A case in our laboratory
involving non-reciprocal donors is seen in Scheme 7a. Thus,
treatment of the diol acceptor 4, with 2 equiv of donor 7a
stopped at disaccharide stage 12b. Difficulty in proceeding
to trisaccharide 28 was evident from the fact, that subse-
quent treatment of disaccharide 12b with 3 equiv of the
armed donor (7a) afforded only a 14% yield of 28a (25).
Note that our poor result (25) coincides with van Boom’s
10% yield in Scheme 6a confirming that these are couplings
not RDAS sanctioned.
© 2002 NRC Canada

1080
Can. J. Chem. Vol. 80, 2002
Scheme 5.
Scheme 6.
© 2002 NRC Canada

Fraser-Reid et al.
1081
Scheme 7.
By contrast a 1:1:1 mixture of the diol 4 + NPOE 5b +
NPGALK 7a in the presence of 2.5 equiv of the promoter
gave the trisaccharide 28b in unoptimized yield of 39%
(Scheme 7b). Similarly diol 13 + donors 5b and 7b gave
trisaccharide 29 in 54% yield. Clearly, in both cases, each of
the donors had gone to its RDAS-sanctioned site.
distributed over two (dioxolenium) and three (trioxolenium)
oxygens, 30 and 31, respectively. The charge in this com-
plex may be described as “diffuse” in contrast to that on the
armed oxocarbenium ion 32 where the charge is more “com-
pact”. Experiments and theoretical studies are underway to
determine whether the diffuse–compact dichotomy is valid,
and the results will be described in due course.
Analysis of factors which have an impact on the diols’
selectivities is more difficult. But as shown for Scheme 4 in
the Discussion section, RDAS is not based on steric hin-
drance of the acceptor’s OH.
Rationalization
A rationalization of the basis for RDAS obviously re-
quires analysis of both the donor and acceptor. With respect
to donors, an examination of Table 1 shows that the NPOE–
NPGAC group displays similar preferences qualitatively, and
is generally more discriminating than NPGALK. Are these
choices related to the donor’s reactivity? This question
prompted us to determine the relative reactivities of donors
5b, 6b, and 7b. We have recently shown that oxidative hy-
drolysis provides a ready method for comparing relative re-
activities of any two n-pentenyl glycosyl donors (26). Using
this method, the ranking of the donors was found to be 5b >
7b > 6b. These observations show that the regioselectivities
are not dependent on the relative reactivities of the donors,
since the most and least reactive donors, NPOE 5b and
NPGAC 6b, respectively, showed regioselectivity for the
same hydroxyl (C-6-OH) of the diol 4. Thus, RDAS is not
based on donor reactivity.
The most thorough investigation of glycosidation of
polyhydroxylic acceptors has been undertaken by Vasella in
connection with his scholarly development of the chemistry
of glycosylidene carbenes (33) as glycosyl donors (27)
(Scheme 9). These donors are unique in that they proceed to
the critical oxocarbenium ion intermediate (34) by bond for-
mation to a proton, whereas “classical donors” (35) achieve
the same objective by bond cleavage (sooner or later!), usu-
ally electrophile induced, of the anomeric leaving group. For
33, the more acidic OH group should therefore be the pre-
ferred glycosidation site. In turn, hydrogen bonding within
the diol 28 can be a determining influence on the acidity of
the OH concerned.
The results obtained by us thus far, do not provide any in-
sight into the implications of OH-acidity and (or) hydrogen
bonding. The role of such factors await future study.
As shown in Scheme 8, NPOE and NPGAC are able to
produce the same intermediate complex where the charge is
© 2002 NRC Canada

1082
Can. J. Chem. Vol. 80, 2002
Scheme 8.
Iodosuccinimide (0.1219 mmol) was added to the solution,
and after stirring for 3 min, TDBMSOTf (0.027 mmol) was
added. The reaction was quenched after 20 min with 10%
aqueous sodium thiosulfate and saturated aqueous sodium
bicarbonate, extracted with CH2Cl2, and worked up in the
usual way. After work up the mixture was separated by flash
chromatography to give the disaccharide products.
(b) To a solution of the diol (0.208 mmol) and the
glycosyl donor (0.16 mmol) in anhyd CH2Cl2 (3 mL) were
added molecular sieves 5 Å (100 mg) and NIS (73 mg,
0.32 mmol) at –30°C. After stirring for 5 min, BF3·OEt2
(6 L, 0.048 mmol) was added. The reaction was quenched
after 15 min and processed as in part (a).
(c) To a solution of diol (0.208 mmol) and glycosyl donor
(0.16 mmol) in anhyd CH2Cl2 (3 mL) were added molecular
sieves 5 Å (97 mg) and NIS (73 mg, 0.32 mmol) at –30°C.
After stirring for 5 min, BF3·OEt2 (6 L, 0.048 mmol) was
added. The mixture was allowed to warm to 0°C. After
40 min the reaction was quenched and processed as in part
(a).
(d) The diol (0.177 mmol, 50 mg) and the glycosyl donor
(0.177 mmol) were dissolved in a small quantity of a di-
chloromethane–toluene mixture, azeotroped to dryness, and
then dried under vacuum for ~30 min. The residue was taken
up in dry dichloromethane (2 mL for ꢁ-acceptor, 10 mL for
ꢀ-acceptor) and molecular sieves (4 Å) were added. The
reaction mixture was kept under an argon atmosphere
at –78°C (acetone – dry ice) and 68 mg (0.266 mmol,
1.5 equiv) of N-iodosuccinimide were added. After 3 min
stirring, 10 L (0.05 mmol, 0.3 equiv) of TBDMSOTf were
added. After 15 min at –78°C the reaction mixture was al-
lowed to warm up to room temperature. The reaction was
quenched after 45 min with 10% aqueous sodium thio-
sulphate and saturated aqueous sodium bicarbonate solution,
Experimental
General procedures
All NMR spectra were recorded on a GE 300 or a Varian
400 MHz NMR spectrometer and chemical shifts are re-
ported relative to internal TMS. Chemical ionization (CI)
mass spectroscopy was done on a Hewlett–Packard 5988A
GC–MS using 1% ammonia in methane as the reagent gas,
with a source temperature of 100°C, at 1 torr (1 torr =
133.322 Pa). High-resolution mass spectra (HR-MS) and fast
atom bombardment (FAB) analyses were recorded with a
JEOL JMS-SX 102A mass spectrometer operating at 10 K
resolution, using a dithiothreitol–dithioerythritol or m-
nitrobenzyl alcohol as the matrix with xenon as the fast
atom. Elemental analyses were conducted at Atlantic micro-
lab, Norcross, GA. All reactions were conducted under an
argon atmosphere. Thin layer chromatography (TLC), was
done on plates (Riedel-de Haen, coated with silicagel 60F
254) and were detected by UV or by spraying or dipping in
ammonium molybdate solution and subsequent heating.
The regioisomers were separated by flash chromatogra-
phy, evaporated to dryness, and weighed to obtain the ratios
reported in Schemes 3–5 and Table 1. Flash column chroma-
tography was performed on silica gel (spectrum SIL 58, 230
ꢂ400 mesh, grade 60) using mixtures of hexane and ethyl
acetate as eluants. Dichloromethane and toluene were dis-
tilled from CaH2. N-Iodosuccinimide was purchased from
Aldrich and recrystallized from hot CH2Cl2.
Glycosidation conditions
(a) The diol (~0.092 mmol) and glycosyl donor
(~0.119 mmol) were dissolved in a small quantity of tolu-
ene, azeotroped to dryness, and then dissolved in CH2Cl2
(2 mL) at 0°C under an argon atmosphere. N-
© 2002 NRC Canada

Fraser-Reid et al.
1083
Scheme 9.
Anal. calcd. for C79H82O13Si: C 74.86, H 6.52; found: C 74.66,
H 6.59.
1,3,4,5-Tetra-O-benzyl-2-O-(2,3,4-tri-O-benzyl-6-O-[tert-
butyldiphenylsilyl]-ꢁ-D-mannopyranosyl)-D-myo-inositol
(12b) and 1,3,4,5-tetra-O-benzyl-6-O-(2,3,4-tri-O-benzyl-6-
O-[tert-butyldiphenylsilyl]-ꢁ-D-mannopyranosyl)-D-myo-
inositol (11b)
The diol 4 and glycosyl donor 7a were reacted under the
general glycosidation conditions. The crude residue obtained
after work up on purification by flash column chromatogra-
phy (EtOAc–hexanes, 1:4) afforded 12b (49 mg, 49%) and
11b (15 mg, 16%). Acetylation of 11b gave a colorless
extracted with dichloromethane, dried with sodium sulfate,
and purified by column chromatography using silica gel
(hexanes – ethyl acetate, 4:1 J 1:1).
1
paste: H NMR (400 MHz, CDCl3) ꢃ: 7.71–6.88 (m, Ar),
Identification of regioisomers
1
5.84 (dd, J = 2.8, 2.4 Hz, 1H), 5.49 (d, J = 1.2 Hz, 1H)
4.91ꢂ4.32 (m, 14H), 4.18 (dd, J = 10, 10 Hz, 1H), 4.06 (dd,
J = 10 Hz, 9.6 Hz, 1H), 3.92–3.76 (m, 4H), 3.59–3.57 (m,
2H), 3.44 (dd, J = 10 Hz, 2.8 Hz, 1H), 3.37 (dd, J = 9.6,
2.4 Hz, 1H), 3.28 (dd, J = 9.6, 9.6 Hz, 1H), 2.12 (s, 3H),
0.99 (s, 9H). FAB-MS m/z: 1251.56 [M – 1]+. Data for 12:
1H NMR (400 MHz, CDCl3) ꢃ: 7.75–6.92 (m, 45H, Ar),
5.35 (d, J = 1.2 Hz, 1H, H-1), 4.95ꢂ4.41 (m, 14H, Bn), 4.37
(t, J = 2.4 Hz, 1H), 4.26 (dd, J = 10, 9.6 Hz, 1H), 3.98–3.85
(m, 3H), 3.79–3.71 (m, 3H), 3.57 (d, J = 11.2 Hz, 1H),
3.32–3.25 (m, 2H), 3.19–3.16 (dd, J = 9.6, 2 Hz, 1H), 2.39
(br s, 1H, OH), 1.02 (s, 9H, t-Bu). 13C NMR ꢃ: 138.99,
138.67, 138.57, 137.73, 137.43, 136.096, 135.64, 133.91,
133.61, 129.42, 129.36, 128.62, 128.47, 128.347, 128.29,
128.22, 128.18, 128.13, 128.06, 128.02, 127.96, 127.74, 127.56,
127.47, 127.36, 127.33, 127.17, 127.11, 127.06, 98.00 (C-1),
83.22, 80.85, 80.47, 79.38, 79.15, 75.78, 75.61, 75.21,
74.43, 72.90, 72.71, 72.16, 72.10, 70.08, 62.70, 26.80,
19.33. FAB-MS m/z: 1209.6 [M – 1]+. Benzoylation under
standard conditions afforded a colorless paste: 1H NMR
(400 MHz, CDCl3) ꢃ: 8.17–6.95 (m, Ar), 5.76 (dd, J = 9.6,
10 Hz, 1H), 5.40 (br s, 1H), 4.99ꢂ4.41 (m, 15H), 4.27
(dd, J = 10, 10 Hz, 1H), 4.02–3.86 (m, 4H), 4.27 (dd, J =
11.2, 2.8 Hz, 1H), 3.60–3.51 (m, 2H), 3.42 (dd, J = 10,
2 Hz, 1H), 3.30 (dd, J = 10, 2.3 Hz, 1H), 1.03 (s, 9H).
FAB-MS m/z: 1312.45 [M – 1]+, 1354.40 [M + K]+, 1447.42
[M + Cs]+.
The H NMR spectra of the chromatographically pure re-
action product, before and after acylation, were examined
for new signals in the 6.0–5.4 ppm region. The coupling
contant(s) of the new proton gave unequivocal information
about its proton’s environment. In the H NMR data listed
below, the “new” signal upon acylation is shown in italics.
1
1,3,4,5-Tetra-O-benzyl-6-O-(2-O-benzoyl-3,4-di-O-benzyl-
6-O-[tert-butyldiphenylsilyl]-ꢁ-D-mannopyranosyl)-D-myo-
inositol (11a)
(a) The diol 4 and the NPGAC donor 6a were reacted
under conditions (a). Standard work up and flash column
chromatography (EtOAc–hexanes, 1:4) afforded 11a (65 mg,
58%) as a colorless paste.
(b) The diol 4 and glycosyl donor 5a were reacted under
conditions (a) for 20 min. Standard work up, and flash col-
umn chromatography of the crude residue (EtOAc–hexanes,
1:4) afforded 11a (80 mg, 73%) as a colorless paste.
1H NMR (400 MHz, CDCl3) ꢃ: 8.12–6.84 (m, 45H, Ar-H),
5.78 (dd, J = 1.6 Hz, 2 Hz, 1H), 5.60 (d, J = 1.6 Hz, 1H,
H-1), 4.90ꢂ4.55 (m, 12H, BnCH2), 4.22 (dd, J = 9.6, 10 Hz,
1H), 4.17ꢂ4.12 (m, 2H), 4.05 (dd, J = 3.2, 10 Hz, 1H),
3.98–3.91 (m, 2H), 3.58–3.48 (m, 2H), 3.37 (dd, J = 2.8,
3.2 Hz, 1H), 3.34 (dd, J = 3.2, 3.2 Hz, 1H), 3.27 (dd, J =
9.6, 9.2 Hz, 1H), 2.36 (br s, 1H, OH), 1.06 (s, 9H).
13C NMR (100 MHz, CDCl3) ꢃ: 165.74 (C=O Ph), 139.09,
138.53, 138.25, 137.86, 137.16, 135.93, 135.64, 133.78,
133.31, 132.94, 130.00, 129.36, 128.57, 128.46, 128.39,
128.27, 128.22, 128.03, 127.96, 127.88, 127.55, 127.49,
127.09, 98.26 (C-1), 81.34, 81.06, 80.61, 79.50, 78.41,
75.81, 75.51, 74.93, 74.92, 74.08, 72.63, 72.10, 71.47,
69.31, 66.44, 62.39, 26.84, 19.31. Acetylation of 11a gave a
Methyl-3,6-di-O-allyl-4-O-(2-O-benzoyl-3,4,6-tri-O-benzyl-
ꢁ-D-mannopyranosyl)-ꢁ-D-mannopyranoside (14)
The diol 13 and glycosyl donor 5b were reacted under
conditions (a). The crude residue on flash column chroma-
tography (EtOAc–hexanes, 1:3) afforded 14 (69%) as a col-
orless paste. 1H NMR (400 MHz, CDCl3) ꢃ: 8.07–7.19
(m, 20H, Ar-H), 6.04–5.85 (m, 2H), 5.67 (dd, J = 2, 2.8 Hz,
1H), 5.42 (d, J = 1.6 Hz, 1H, H-1), 5.33–5.09 (m, 4H),
4.89ꢂ4.73 (m, 4H), 4.59ꢂ4.52 (m, 3H), 4.14–3.88 (m, 10H),
3.78–3.63 (m, 5H), 3.39 (s, 3H), 2.42 (s, 1H, OH).
13C NMR (100 MHz, CDCl3) ꢃ: 165.48, (OCOPh), 138.45,
138.04, 134.83, 134.02, 133.01, 129.92, 128.31, 128.27,
127.91, 127.56, 127.50, 118.33, 116.83, 99.98 (C-1), 99.51
(C-1>), 79.72, 78.15, 75.21, 74.15, 73.50, 73.13, 72.60,
72.44, 71.53, 70.46, 70.19, 69.33, 69.08, 67.54, 54.99. FAB-
MS m/z: 809.3 [M – 1]+, 811.3 [M + 1]+. Acetylation of 14
afforded the acetate as a colorless paste: 1H NMR
(400 MHz, CDCl3) ꢃ: 8.00–7.13 (m, 20H, Ar-H), 5.92–5.81
(m, 2H), 5.65 (dd, J = 2.4, 2.8 Hz, 1H), 5.38 (d, J = 2 Hz,
1
colorless paste. H NMR (400 MHz, CDCl3) ꢃ: 8.12–6.83
(m, 45H, Ar-H), 5.84 (dd, J = 2.8, 2.4 Hz, 1H), 5.80 (dd, J =
1.6, 2 Hz, 1H), 5.60 (d, J = 1.2 Hz, 1H, H-1) 4.01ꢂ4.47 (m,
12H, BnCH2,s), 4.14 (dd, J = 9.6, 10 Hz, 1H), 4.07–3.96
(m, 3H), 3.83 (dd, J = 9.6, 9.6 Hz, 1H), 3.61–3.52 (m, 2H),
3.46 (dd, J = 3.2, 3.2 Hz, 1H), 3.44 (dd, J = 2.8, 3.2 Hz,
1H), 3.31 (dd, J = 9.6, 9.6 Hz, 1H), 2.10 (s, 3H), 1.05 (s,
9H). 13C NMR (100 MHz, CDCl3) ꢃ: 170.39 (OCOCH3),
165.97 (OCOPh), 139.08, 138.49, 138.17, 137.65, 137.52,
136.82, 135.94, 135.66, 133.65, 133.36, 132.95, 130.01,
129.41, 129.29, 128.86, 128.48, 128.42, 128.22, 128.12,
128.02, 127.83, 127.60, 127.57, 127.43, 127.22, 127.12,
97.99 (C-1), 81.68, 81.08, 78.72, 78.32, 78.12, 75.99, 75.76,
75.03, 74.66, 74.06, 72.17, 71.86, 71.42, 69.05, 65.93,
62.55, 26.83, 21.03, 19.27. FAB-MS m/z: 1265.56 [M – 1]+.
© 2002 NRC Canada

1084
Can. J. Chem. Vol. 80, 2002
1H, H-1), 5.21 (d, J = 1.2 Hz, 1H) 5.19 (dd, J = 2, 3.2 Hz,
1H), 5.17 (d, J = 1.6 Hz, 1H), 5.08–5.05 (m, 2H), 4.82ꢂ4.64
(m, 4H), 4.52ꢂ4.47 (m, 3H), 4.06–3.62 (m, 14H), 3.32 (s,
3H), 2.02 (s, 3H).
128.11, 127.83, 127.68, 127.48, 127.25, 126.36, 125.99,
102.25 (PhCH), 101.79 (C-1), 95.00 (C-1>), 79.65, 77.42,
75.08, 74.78, 74.67, 72.76, 72.50, 71.22, 70.33, 69.96,
1
69.34, 69.15, 67.32, 58.76, 55.31. Data for 19: H NMR
(200 MHz, CDCl3) ꢃ: 7.57–7.18 (m, 25H), 5.63 (s, 1H), 4.98
(d, J = 2.1 Hz, 1H), 4.92ꢂ4.50 (m, 9H), 4.33 (dd, J = 10.1,
4.8 Hz, 1H), 4.12 (dd, J = 10.1, 4.9 Hz, 1H), 3.98–3.69 (m,
10H), 3.29 (s, 3H). 13C NMR (75 MHz, CDCl3) ꢃ: 138.41,
138.31, 138.11, 137.42, 129.06, 128.37, 128.24, 128.07,
127.79, 127.48, 126.27, 102.25 (PhCH), 100.35 (C-1), 99.51
(C-1>), 79.41, 76.76, 76.68, 75.50, 75.12, 73.39, 72.93,
72.68, 72.46, 69.57, 69.25, 67.68, 58.07, 55.50. API-ES m/z:
828 [M + Na]+.
Methyl-3,6-di-O-allyl-2-O-(2,3,4,6-tetra-O-benzyl-ꢁ-D-
mannopyranosyl)-ꢁ-D-mannopyranoside (15)
The diol 13 and glycosyl donor 7b were reacted under
conditions (a) for 3 min. Standard work up followed by flash
column chromatography (EtOAc–hexanes, 1:3) afforded 15
1
(23%) as a colorless paste. H NMR (400 MHz, CDCl3) ꢃ:
7.35–7.15 (m, 20H, Ar-H), 5.95–5.81 (m, 2H), 5.30–5.11
(m, 4H), 5.09 (br s, 1H, H-1), 4.88ꢂ4.46 (m, 12H), 4.10–
3.35 (m, 21H), 3.24 (s, 3H). FAB-MS m/z: 795 [M – 1]+.
1
Benzoylation afforded the corresponding benzoate: H NMR
Methyl-4,6-O-benzylidene-3-O-(2,3,4,6-tetra-O-benzoyl-ꢁ-
D-mannopyranosyl)-ꢁ-D-altropyranoside (18c)
(300 MHz, CDCl3) ꢃ: 8.13–8.04 (d, 2H, Ar-H), 7.64–7.14
(m, 23H, Ar-H), 5.85–5.06 (m, 2H), 5.36 (dd, J = 9.6,
9.3 Hz, 1H), 5.19 (br s, 1H), 5.15ꢂ4.99 (m, 4H), 4.91ꢂ4.80
(m, 2H), 4.72ꢂ4.46 (m, 7H), 4.07–3.85 (m, 11H), 3.74 (br s,
2H), 3.55 (br s, 2H), 3.29 (s, 3H). 13C NMR (100 MHz,
CDCl3) ꢃ: 172.25 (OCOPH), 138.60, 138.59, 138.58,
134.72, 134.64, 134.00, 133.30, 130.41, 129.92, 129.48,
128.69, 128.61, 128.55, 128.21, 128.11, 127.93, 127.70,
127.63, 117.45, 117.03, 100.01 (C-1), 100.20 (C-1), 79.64,
75.14, 75.08, 74.73, 74.46, 73.46, 72.69, 72.62, 72.37,
72.08, 71.59, 70.45, 70.15, 69.76, 55.11.
A solution of diol 17 and glycosyl donor 5c were reacted
under conditions (b) for 15 min. After work up the crude
residue was purified by flash chromatography (hexane–
EtOAc, 4:1) to give the O-3 mannosylated product 18c in
1
92% yield. H NMR (300 MHz, CDCl3) ꢃ: 8.10–7.18 (m,
25H), 6.12 (t, J = 10.1 Hz, 1H), 5.93 (dd, J = 10.1, 3.3 Hz,
1H), 5.73 (dd, J = 3.2, 1.8 Hz, H1), 5.65 (s, 1H), 5.25 (d, J =
1.6 Hz, 1H), 4.79 (dt, J = 10.1, 2.6 Hz, 1H), 4.70 (s, 1H),
4.55ꢂ4.28 (m, 4H), 4.11 (dd, J = 9.5, 3.2 Hz, 1H), 3.98 (dd,
J = 12.5, 3.2 Hz, 1H), 3.84 (t, J = 10.2 Hz, 1H), 3.51 (s,
3H). 13C NMR (75 MHz, CDCl3) ꢃ: 166.13 (OC(O)Ph),
165.52 (OC(O)Ph), 165.35 (×2) (OC(O)Ph), 137.65, 133.36,
133.32, 133.03, 132.87, 130.01, 129.76, 129.70, 129.35,
129.26, 129.11, 128.52, 128.35, 128.25, 128.21, 126.14,
102.28 (PhCH), 101.79 (C-1), 95.63 (C-1>), 74.82, 73.09,
70.63, 70.32, 69.33, 68.58, 68.39, 66.46, 62.01, 58.78,
55.51. API-ES m/z: 883 [M + Na]+.
Methyl-4,6-O-benzylidene-3-O-(2-O-benzoyl-3,4,6-tri-O-
benzyl-ꢁ-D-mannopyranosyl)-ꢁ-D-altropyranoside (18b)
The altroside diol 17 and the NPOE glycosyl donor 5b,
were allowed to react under conditions (b) for 15 min. After
work up the crude material was purified by flash chromatog-
raphy (hexane– EtOAc, 3:2) to give the 3-O-mannosylated
1
derivative of 18b in 52% yield. H NMR (300 MHz, CDCl3)
ꢃ: 8.09–7.06, (m, 25H), 5.60 (m, 1H), 5.55 (s, 1H), 5.14 (d,
J = 1.8 Hz, 1H), 4.87 (d, J = 10.9 Hz, 1H), 4.81 (d, J =
11.2 Hz, 1H), 4.63 (d, J = 12.4 Hz, 1H), 4.61 (br s, 1H),
4.58 (d, J = 12.1 Hz, 1H), 4.49 (d, J = 10.9 Hz, 1H), 4.33
(m, 10H), 3.77 (m. 1H), 3.62 (dd, J = 11.0, 3.7 Hz, 1H),
3.43 (dd, J = 11.0, 1.6 Hz, 1H), 3.40 (s, 3H). 13C NMR
(75 MHz, CDCl3) ꢃ: 165.88 (OC(O)Ph), 138.81, 138.42,
137.98, 137.42, 133.10, 129.96, 129.88, 128.94, 128.37,
128.26, 128.21, 128.13, 127.85, 127.59, 127.46, 127.36, 127.32,
126.34, 102.33 (PhCH), 101.83 (C-1), 95.57 (C-1>), 77.88,
74.92, 74.72, 74.31, 72.92, 71.93, 71.63, 71.08, 69.42, 69.29,
68.61, 67.93, 58.67, 55.57. API-ES m/z: 837 [M + NH4]+.
Methyl-4,6-O-benzylidene-2 and 3-O-(3,4,6-tri-O-benzyl-
2-O-benzoyl-ꢁ-D-mannopyranosyl)-ꢁ-D-glucopyranoside
(21a and 22a)
The diol 20 and donor 5b reacted under conditions (d) to
give 21a and 22a in a 7.4:1 ratio.
For 21a
1
Rf = 0.31 (hexanes–EtOAc, 2:1). H NMR (300 MHz,
CDCl3) ꢃ: 8.03 (d, 2H), 7.16–7.56 (m, 23H), 5.72 (dd, 1H),
5.60 (s, 1H), 5.47 (d, 1H), 4.86–6.94 (m, 3H), 4.74 (d, 1H),
4.73 (d, 1H), 4.53 (d, 1H), 4.52 (d, 1H), 4.49 (d, 1H), 4.36
(dd, 1H), 4.30 (dd, 1H), 3.99ꢂ4.09 (m, 3H), 3.71–3.90 (m,
4H), 3.38 (s, 3H), 2.03 (s, 3H). LR-FAB-MS m/z: 817 [M – H]+.
Methyl-4,6-O-benzylidene-3-O and 2-O (2,3,4,6-tetra-O-
benzyl-ꢁ-D-mannopyranosyl)-ꢁ-D-altropyranosides (18a
and 19)
For 22a
Rf = 0.48 (hexanes–EtOAc, 2:1). 1H NMR (300 MHz,
CDCl3) ꢃ: 8.07 (d, 2H), 7.15–7.60 (m, 23H), 5.64 (dd, 1H),
5.56 (s, 1H), 5.10 (d, 1H), 4.83ꢂ4.87 (m, 2H), 4.80 (d, 1H),
4.71 (d, 1H), 4.60 (d, 1H), 4.54 (d, 1H), 4.52 (d, 1H), 4.29
(dd, 1H), 4.15ꢂ4.24 (m, 2H), 4.03ꢂ4.11 (m, 2H), 3.86 (dd,
1H), 3.69–3.82 (m, 5H), 3.47 (dd, 1H), 3.43 (s, 3H), 3.20 (d,
1H). LR-FAB-MS m/z: : 941 [M + Na]+, 817 [M – H]+.
To a solution of diol 17 and glycosyl donor 7b were re-
acted under conditions (a). After stirring for 5 min,
BF3·OEt2 (6 mL, 0.048 mmol) was added. The mixture was
allowed to warm to 0°C. After 90 min the reaction was
quenched and after work up the crude residue was purified
by flash column chromatography (hexane–EtOAc, 7:3) to
give the O-3 and O-2 mannosylated products 18a and 19 in
2:1 ratio in 37% combined yield. Data for O-3 product 18a:
1H NMR (300 MHz, CDCl3) ꢃ: 7.56–7.01 (m, 25H), 5.54 (s,
1H), 5.00 (d, J = 1.8 Hz, 1H), 4.92ꢂ4.44 (m, 9H), 4.30–3.43
(m, 12H), 3.30 (s, 3H). 13C NMR (75 MHz, CDCl3) ꢃ:
139.05, 138.88, 138.48, 137.62, 132.11, 128.80, 128.26,
Methyl-4,6-O-benzylidene-2 and 2-O-(3,4,6-tri-O-benzyl-
2-O-benzoyl-ꢀ-D-mannopyranosyl)-ꢁ-D-glucopyranoside
(21b and 22b)
(a) The diol 20 and donor 9 reacted under conditions
(d) to give 21b and 22b in a 7.6:1 ratio in 25% overall yield.
© 2002 NRC Canada

Fraser-Reid et al.
1085
For 21b
For 22d (ꢁ -anomer)
1
1H NMR (300 MHz, CDCl3) ꢃ: 8.01–7.98 (m, 2H), 7.57–
7.52 (m, 1H), 7.44–7.10 (m, 22H), 5.45 (t, J = 9.6 Hz, 1H,
H-3), 5.41 (s, 1H, PhCH), 5.29 (dd, J = 8.1, 9.6 Hz, 1H,
H-2>), 5.01 (d, J = 3.6 Hz, 1H, H-1), 4.82ꢂ4.51 (m, 7H),
4.25 (dd, J = 5.1, 10.2 Hz, 1H, H-6a), 3.92–3.81 (m, 2H),
3.74–3.56 (m, 6H), 3.45 (dd, J = 9.0, 9.6 Hz, 1H), 3.38 (s,
3H, OCH3), 1.41 (s, 3H, OC(O)CH3). FAB+-MS m/z (%):
859 ([M – H]+, 100), 883 ([M + Na]+, 21).
Selected H NMR (300 MHz, CDCl3) peaks ꢃ: 5.58 (d,
J = 3.0 Hz, 1H), 5.46 (s, 1H, PhCH), 2.02 (s, 3H,
OC(O)CH3). Lit. (28) ꢃ value: 5.58 (d, J = 3.5 Hz, 1H, H-1>),
5.46 (s, 1H), 2.03 (s, 3H).
For 22d (ꢀ-anomer)
1
Selected H NMR (300 MHz, CDCl3) peaks ꢃ: 5.50 (s,
1H, PhCH), 1.94 (s, 3H, OC(O)CH3). Lit. (28) ꢃ value: 5.51
(s, 1H), 1.96 (s, 3H).
For 22b
1H NMR (300 MHz, CDCl3) ꢃ: 7.98–7.95 (m, 2H), 7.59–
7.54 (m, 1H), 7.45–7.08 (m, 22H), 5.44 (s, 1H, PhCH), 5.22
(dd, J = 7.8, 8.1 Hz, 1H, H-2>), 4.84ꢂ4.47 (m, 9H),
4.26ꢂ4.20 (m, 2H), 3.81–3.67 (m, 5H), 3.53–3.52 (m, 2H),
3.32–3.29 (m, 1H), 3.29 (s, 3H, OCH3), 1.80 (s, 3H,
OC(O)CH3). FAB+-MS m/z (%): 859 ([M – H]+, 100).
(b) When the reaction was carried out at 0°C instead of
–78°C, the ratio of 21b to 22b was 3:1 and the overall yield
was 42%.
Methyl-4,6-O-benzylidene-2 and 3-O-(3,4,6-tri-O-benzyl-
2-O-benzoyl-ꢁ-D-mannopyranosyl)-ꢀ-D-glucopyranoside
(24a and 25a)
The diol 23 and donor 5b were reacted under conditions
(d) to give 24a and 25a in a 3.5:1 ratio and 36% overall
yield.
For 24a
1H NMR (300 MHz, CDCl3) ꢃ: 8.07 (d, 2H), 7.17–7.56
(m, 23), 5.65 (dd, 1H), 5.52 (s, 1H), 5.32 (dd, 1H), 5.30 (d,
1H), 4.84 (d, 1H), 4.81 (d, 1H), 4.72 (d, 1H), 5.59 (d, 1H),
4.54 (d, 1H), 4.53 (d, 1H), 4.37 (d, 1H), 6.33 (dd, 1H), 4.24
(dd, 1H), 4.00ꢂ4.12 (m, 2H), 3.74–3.87 (m, 5H), 3.56 (s, 3H),
3.44 (ddd, 1H), 2.01 (s, 3H). LR-FAB-MS m/z: 817 [M – H]+.
Methyl-4,6-O-benzylidene-2 and 3-O-(2–3,4,6-tetra-O-
benzyl-ꢁ-D-mannopyranosyl)-ꢁ-D-glucopyranoside (21c and
22c)
The diol 20 and donor 7b reacted under conditions (d) to
give 21c and 22c in a 1:3 ratio and 30% yield.
For 25a
For 21c
Rf = 0.32 (hexanes–EtOAc, 2:1). 1H NMR (300 MHz,
CDCl3) ꢃ: 8.06 (d, 2H), 7.17–7.58 (m, 23H), 5.75 (dd, 1H),
5.58 (s, 1H), 5.38 (d, 1H), 4.85 (d, 1H), 4.78 (d, 1H), 4.56
(d, 1H), 4.54 (d, 1H), 4.53 (d, 1H), 4.37 (dd, 1H), 4.28 (d,
1H), 4.05ꢂ4.16 (m, 2H), 3.84–3.93 (m, 3H), 3.79 (dd, 1H),
3.66 (dd, 1H), 3.57 (s, 3H), 3.38–3.51 (m, 2H), 2.85 (d, 1H).
LR-FAB-MS m/z: 817 [M – H]+.
Rf = 0.23 (hexanes–EtOAc, 4:1). 1H NMR (300 MHz,
CDCl3) ꢃ: 7.05–7.50 (m, 25H), 5.53 (s, 1H), 5.41 (d, 1H),
4.92 (d, 1H), 4.88 (d, 1H), 4.78 (dd, 1H), 4.68 (d, 1H),
4.48ꢂ4.58 (m, 4H), 4.38 (d, 1H), 4.35 (d, 1H), 4.29 (dd,
1H), 3.92ꢂ4.01 (m, 3H), 3.88 (ddd, 1H), 3.72–3.82 (m, 4H),
3.62 (dd, 1H), 3.33–3.40 (m, 1H), 3.38 (s, 3H), 2.02 (s, 3H).
LR-FAB-MS m/z: 803 [M – H]+.
Methyl-4,6-O-benzylidene-2 and 3-O-(3,4,6-tri-O-benzyl-
2-O-benzoyl-ꢀ-D-mannopyranosyl)-ꢀ-D-glucopyranoside
(24b and 25b)
The diol 23 and donor 9 reacted under conditions (d) to
give an 1.8:1 mixture (11% overall yield) of 29b and 25b.
For 22c
Rf = 0.28 (hexanes–EtOAc, 4:1). 1H NMR (300 MHz,
CDCl3) ꢃ: 7.10–7.65 (m, 25H), 5.67 (dd, 1H), 5.46 (s, 1H),
4.78ꢂ4.89 (m, 5H), 6.66 (d, 1H), 6.65 (d, 1H), 4.61 (d, 1H),
4.54 (d, 1H), 6.69 (d, 1H), 4.26 (dd, 1H), 3.67–3.96 (m, 8H),
3.50 (dd, 1H), 3.33 (s, 3H), 3.28–3.38 (m, 2H), 1.98 (s, 3H).
LR-FAB-MS m/z: 827 [M + Na]+, 803 [M – H]+.
For 24b
1H NMR (300 MHz, CDCl3) ꢃ: 8.02–7.99 (m, 2H), 7.59–
7.09 (m, 23H) 5.38 (s, 1H, PhCH), 5.24 (dd, J = 8.1, 9.7 Hz,
1H, H-2>), 5.17 (dd, J = 7.8, 9.3 Hz, 1H, H-3), 4.94ꢂ4.55
(m, 7H), 4.29 (dd, J = 5.1, 10.5 Hz, 1H, H-6a), 3.81–3.42
(m, 10H), 3.48 (s, 3H, OCH3), 1.77 (s, 3H, OC(O)CH3).
Methyl-4,6-O-benzylidene-2 and 3-O-(2–3,4,6-tetra-O-
benzyl-ꢁ/ꢀ-D-glucopyranosyl-ꢁ-D-glucopyranoside (21d
and 22d)
The diol 20 and donor 10 were reacted under conditions
(d) to give 21d and 22d in an 1.2:1 ratio.
For 25b
1H NMR (300 MHz, CDCl3) ꢃ: 7.99–7.97 (m, 2H), 7.73–
7.69 (m, 1H), 7.46–7.09 (m, 22H), 5.40 (s, 1H, PhCH), 5.20
(dd, J = 8.1, 8.7 Hz, 1H, H-2>), 4.97 (dd, J = 7.8, 9.0 Hz,
1H, H-2), 4.82ꢂ4.20 (m, 9H), 3.95 (dd, J = 8.7, 9.0 Hz, 1H),
3.83–3.57 (m, 5H), 3.50 (br d, J = 3.6 Hz, 2H), 3.37 (s, 3H,
OCH3), 3.32–3.28 (m, 1H), 2.17 (s, 3H, OC(O)CH3).
For 21d (ꢁ-anomer)
Selected 1H NMR (300 MHz, CDCl3) peaks ꢃ: 5.61 (t, J =
9.6 Hz, 1H), 5.46 (s, 1H, PhCH), 2.00 (s, 3H, OC(O)CH3).
Lit. (28) ꢃ value: 5.62 (t, J = 9.7 Hz, 1H, H-3), 5.47 (s, 1H),
2.01 (s, 3H).
For 21d (ꢀ-anomer)
Methyl-4,6-O-benzylidene-2 and 3-O-(2–3,4,6-tetra-O-
benzyl-ꢁ-D-mannopyranosyl)-ꢀ-D-glucopyranoside (24c
and 25c)
The diol 23 and donor 7b reacted under conditions (d) to
give 24c and 25c in an 1:3 ratio (35% overall yield).
Selected 1H NMR (300 MHz, CDCl3) peaks ꢃ: 5.65 (t, J =
9.6 Hz, 1H), 5.47 (s, 1H, PhCH), 1.76 (s, 3H, OC(O)CH3).
Lit. (28) ꢃ value: 5.65 (t, J = 9.8 Hz, 1H, H-3), 5.48 (s, 1H),
1.78 (s, 3H).
© 2002 NRC Canada

1086
Can. J. Chem. Vol. 80, 2002
For 25c
0.037 mmol, 40%). Note: To confirm the regioselectivity,
the pseudo-disaccharide 11a was coupled with 7a which af-
forded the same pseudo-trisaccharide 28. 1H NMR
(400 MHz, CDCl3) ꢃ: 8.10–6.94 (m, 65H, Ar-H), 5.83–5.81
(dd, J = 2.4, 2 Hz, 1H), 5.626–5.621 (d, J = 2 Hz, 1H,
H-1), 5.35 (s, 1H, H-1), 4.97ꢂ4.36 (m, 20H), 4.29–3.99 (m,
7H), 3.93–3.90 (dd, J = 2.8, 9.6 Hz, 1H), 3.85–3.59 (m, 4H),
3.48–3.26 (m, 5H), 1.03 (s, 9H). FAB-MS m/z: 1745.76
[M –1]+. Anal. calcd. for C111H114O17Si: C 76.26, H 6.57;
found: C 76.00, H 6.71.
Rf = 0.31 (hexanes–EtOAc, 2:1). 1H NMR (300 MHz,
CDCl3) ꢃ: 7.08–7.68 (m, 25H), 5.51 (s, 1H), 5.44 (d, 1H),
4.96 (dd, 1H), 4.86 (d, 1H), 4.63 (d, 1H), 4.53 (s, 2H),
4.48ꢂ4.57 (m, 3H), 4.44 (d, 1H), 4.38 (d, 1H), 4.36 (dd,
1H), 4.09 (dd, 1H), 3.64–3.93 (m, 8H), 3.49 (s, 3H), 3.44
(dd, 1H), 2.04 (s, 3H). LR-FAB-MS m/z: 803 [M – H]+.
For 24c
Rf = 0.31 (hexanes–EtOAc, 2:1). 1H NMR (300 MHz,
CDCl3) ꢃ: 7.13–7.43 (m, 25H), 5.45 (s, 1H), 5.34 (d, 1H),
5.24 (dd, 1H), 4.88 (d, 1H), 4.73 (s, 2H), 4.66 (d, 1H), 4.60
(d, 1H), 4.59 (d, 1H), 4.57 (d, 1H), 4.51 (d, 1H), 4.34 (dd,
1H), 4.33 (d, 1H), 3.99 (dd, 1H), 3.68–3.81 (m, 7H), 3.54
(dd, 1H), 3.46 (dd, 1H), 3.41 (s, 3H), 2.01 (s, 3H). LR-FAB-
MS m/z: 827 [M + Na]+, 803 [M – H]+.
Methyl-3,6-di-O-allyl-1-O-methyl-2-O-(2,3,4,6-tetra-O-
benzyl-ꢁ-D-mannopyranosyl)-4-O-(2-O-benzoyl-3,4,6-tri-O-
benzyl-ꢁ-D-mannopyranosyl)-ꢁ-D-manno-pyranoside (29)
The diol 13 (27.4 mg, 0.1 mmol) and glycosyl donors 5b
(68.42 mg, 0.11 mmol) and 7b (66.88 mg, 0.11 mmol) were
taken up together in a small quantity of toluene and
azeotroped to remove traces of water. The mixture was dis-
solved in CH2Cl2 (3 mL) at 0°C under an argon atmosphere.
N-Iodosuccinimide (58.5 mg, 0.26 mmol) was added to the
solution. After stirring for 3 min, TBDMSOTf (13.78 L,
0.06 mmol) was added. The reaction mixture was quenched
after 1 h with 10% aqueous sodium thiosulphate and satu-
rated sodium bicarbonate, and extracted with CH2Cl2. The
organic layer was removed under reduced pressure. The
crude residue on flash column chromatography (EtOAc–hex-
anes, 1:4) afforded the trisaccharide 29 (45 mg, 0.034 mmol,
54%) as the major product, along with the disaccharide 14
(20 mg, 0.025 mmol, 39%). Unreacted acceptor 13 (10 mg,
0.03 mmol) was recovered. Note: Regioselectivity was con-
firmed by coupling disaccharide 14 with 7b affording the
same trisaccharide 29 in 36% isolated yield. 1H NMR
(400 MHz, CDCl3) ꢃ: 8.09–8.06 (d, 2H, Ar-H), 7.58–7.06
(m, 38H, Ar-H), 5.94–5.84 (m, 2H), 5.70 (br s, 1H), 5.39 (br s,
1H), 5.26ꢂ4.99 (m, 5H), 4.95ꢂ4.46 (m, 15H), 4.12–3.61
(m, 21H), 3.24 (s, 3H). 13C NMR (100 MHz, CDCl3) ꢃ:
165.31 (OCOPh), 138.59, 138.53, 138.45, 138.39, 138.27,
138.07, 135.01, 134.01, 134.32, 133.04, 130.03, 129.96,
129.92, 128.39, 128.30, 128.28, 128.24, 128.19, 128.04,
127.98, 127.96, 127.83, 127.81, 127.69, 127.59, 127.48,
117.68, 116.41, 99.81 (C-1), 99.68 (C-1), 97.80 (C-1),
79.82, 79.43, 78.32, 75.26, 74.91, 74.76, 74.16, 74.00,
73.49, 73.36, 73.24, 72.56, 72.32, 72.27, 72.15, 71.55,
70.84, 70.70, 69.74, 69.48, 69.03, 54.81. Anal. calcd. for
C80H86O17: C 72.82, H 6.57; found: C 72.90, H 6.66.
Methyl-4,6-O-benzylidene-2 and 3-O-(2–3,4,6-tetra-O-
benzyl-ꢁ/ꢀ-D-glucopyranosyl)-ꢀ-D-glucopyranoside (24d
and 25d)
The diol 23 and donor 10 were reacted under conditions
(d) to give 24d and 25d in an 1.2:1 ratio (51% overall yield).
For 24d (ꢁ-anomer)
Selected H NMR (300 MHz, CDCl3) peaks ꢃ: 5.54 (d,
J = 3.6 Hz, 1H), 5.46 (s, 1H, PhCH), 2.00 (s, 3H,
OC(O)CH3). Lit. (28) ꢃ value: 5.55 (d, J = 3.5 Hz, 1H, H-1>),
5.47 (s, 1H), 2.01 (s, 3H).
1
For 24d (ꢀ-anomer)
Selected H NMR (300 MHz, CDCl3) peaks ꢃ: 5.48 (s,
1H, PhCH), 1.93 (s, 3H, OC(O)CH3). Lit. (28) ꢃ value: 5.49
(s, 1H), 1.94 (s, 3H).
1
For 25d (ꢁ-anomer)
Selected H NMR (300 MHz, CDCl3) peaks ꢃ: 5.57 (d,
J = 3.6 Hz, 1H), 3.50 (s, 3H, OCH3), 2.02 (s, 3H,
OC(O)CH3). Lit. (28) ꢃ value: 5.58 (d, J = 3.4 Hz, 1H, H-1>),
3.51 (s, 3H), 2.03 (s, 3H).
1
For 25d (ꢀ-anomer)
Selected H NMR (300 MHz, CDCl3) peaks ꢃ: 3.49 (s,
3H, OCH3), 3.21 (m, 1H), 1.92 (s, 3H, OC(O)CH3). Lit. (28)
ꢃ value: 3.49 (s, 3H), 3.22 (ddd, J = 2.3, 4.9, 9.2 Hz, 1H,
H-5>), 2.03 (s, 3H).
1
1,3,4,5-Tetra-O-benzyl-2-O-(2,3,4-tri-O-benzyl-6-O-[tert-
butyldiphenylsilyl]-ꢁ-D-mannopyranosyl-6-O-(2-O-benzoyl-
2,4,6-tri-O-benzyl-ꢁ-D-mannopyranosyl)-D-myo-inositol (28)
The diol 4 (50 mg, 0.092 mmol) and glycosyl donors 5b
(90 mg. 0.12 mmol) and 7a (75 mg, 0.12 mmol) were taken
up together in a small quantity of toluene and azeotroped to
remove traces of water. The mixture was dissolved in
CH2Cl2 (3 mL) at 0°C under an argon atmosphere. N-
Iodosuccinimide (54 mg, 0.24 mmol) was added to the solu-
tion. After stirring for 3 min, TBDMSOTf (12 L,
0.054 mmol) was added. The reaction was quenched after
1 h with 10% aqueous sodium thiosulphate and saturated so-
dium bicarbonate, and extracted with CH2Cl2. The organic
layer was removed under reduced pressure. The crude resi-
due on flash column chromatography (EtOAc–hexanes, 1:4)
afforded the pseudo-trisaccharide 28 (62 mg, 0.036 mmol,
39%) as the only trisaccharide along with 11a (40 mg,
Acknowledgements
BFR is grateful to the Human Frontier Science Program
Organization and Synthon Chiragenics of Monmouth Junc-
tion, NJ, U.S.A. for partial support of this work. C.U. thanks
the Comunidad Autónoma de Madrid for a postdoctoral fel-
lowship. Funding from the Dirección General de
Investigación Científica y Técnica (grant: PPQ2000-1330).
References
1. R.U. Lemieux. Adv. Carbohydr. Chem. 9, 1 (1954).
2. H.S. Isbell. Annu. Rev. Biochem. 9, 65 (1940); H.L. Frush and
H.S. Isbell. J. Res. Natl. Bur. Stand. (U.S.), 27, 413 (1941).
3. B. Capon and S.P. McManus. Neighboring group participation.
Plenum Press, New York. 1976.
© 2002 NRC Canada

Fraser-Reid et al.
1087
4. For a convenient summary see: A.F. Bochkov and G.E. Zaikov.
Chemistry of the O-glycosidic bond. Pergamon Press, Oxford,
U.K. 1979. Chap. 2.
15. G. Anilkumar, Z.J. Jia, R. Kraehmer, and B. Fraser-Reid. J.
Chem. Soc. Perkin Trans. 1, 3591 (1999).
16. R. Lemieux, T. Takeda, and B.Y. Chung. ACS Symp. Ser. 39
(1976).
17. C. Roberts, R. Madsen, and B. Fraser-Reid. J. Am. Chem. Soc.
117, 1546 (1995); J.G. Allen and B. Fraser-Reid. J. Am.
Chem. Soc. 121, 468 (1999); C. Roberts, C.L. May, and
B. Fraser-Reid. Carbohydr. Lett. 1, 89 (1994).
18. D.R. Mootoo, P. Konradsson, U. Udodong, and B. Fraser-Reid.
J. Am. Chem. Soc. 110, 5583 (1988).
19. A.H. Haines. Adv. Carbohydr. Chem. Biochem. 33, 11 (1976);
J. Stanek, Jr. Top. Curr. Chem. 154, 209 (1990); A.C. Richardson
and J.M. Williams. Tetrahedron, 23, 1641 (1967); R.C. Chalk,
D.H. Ball, and L. Long, Jr. J. Org. Chem. 31, 1509 (1966);
A.K. Mitra, D.H. Ball, and L. Long, Jr. J. Org. Chem. 27, 160
(1962); H.M. Flowers. Carbohydr. Res. 100, 418 (1982);
H.M. Flowers. Carbohydr. Res. 39, 245 (1975).
20. (a) P.R. Muddasani, B. Bernet, and A. Vasella. Helv. Chim.
Acta. 77, 334 (1994); (b) P.R. Muddasani, E. Bozo, B. Bernet,
and A. Vasella. Helv. Chim Acta, 77, 257 (1994).
21. (a) K. Takeo. Carbohydr. Res. 77, 131 (1979); (b) S. Hanessian
and J. Banoub. Carbohydr. Res. 53, C13 (1977); (c) B. Coxon
and H.G. Fletcher. J. Org. Chem. 26, 2892 (1961).
22. These compounds were identified by comparing their NMR
data with those published in the following: P.R. Muddasani,
E. Bozo, B. Bernet, and A. Vasella. Helv. Chem. Acta. 77, 257
(1994).
5. H. Paulsen. Angew. Chem. Int. Ed. Engl. 21, 152 (1982).
6. A. Pictet and H. Vogel. Helv. Chim. Acta, 11, 436 (1928), and
subsequent papers; J.C. Irvine, J.W.H. Oldham, and A.F. Skin-
ner. J. Am. Chem. Soc. 51, 1279 (1929); H.H. Schlubach and
B. Middelhoff. Ann. 550, 1344 (1942); W.W. Binkley and
M.L. Wolfrom. J. Am. Chem. Soc. 68, 2171 (1946); A.
Klemer and B. Dietzel. Tetrahedron Lett. 275 (1970); R.K.
Ness and H.G. Fletcher, Jr. Carbohydr. Res. 17, 465 (1971).
7. R.U. Lemieux and G. Huber. J. Am. Chem. Soc. 78, 4117
(1956).
8. R.U. Lemieux. Can. J. Chem. 31, 949 (1953); R.U. Lemieux
and G. Huber. Can. J. Chem. 33, 128 (1955).
9. See for example: P. Collins and R. Ferrier. Monosaccharides:
Their chemistry and their roles in natural products. John Wiley
and Sons, Chichester, U.K. 1995. Chap. 3; T.K. Lindhorst. Es-
sentials of carbohydrate chemistry and biochemistry. Wiley–
VCH, Weinheim. 2000; H. Paulsen. In Modern methods in car-
bohydrate synthesis. Edited by S.H. Khan and R.A. O’Neill.
Barwood Academic Publishers, London, U.K. 1996; W.W. Pigman
and R.M. Goepp, Jr. Chemistry of the carbohydrates. Aca-
demic Press, New York. 1948. p. 159; G.-J. Boons. Tetrahe-
dron, 52, 1995 (1996); A.F. Bochkov and G.E. Zaikov.
Chemistry of the O-glycosidic bond. Pergamon Press, Oxford.
1979. Chap. 2 and Chap. 3.
10. G. Anilkumar, L.G. Nair, and B. Fraser-Reid. Org. Lett. 2,
2587 (2000); B. Fraser-Reid, J.C. Lopez, K.V. Radhakrishnan,
M. Mach, U. Schlueter, A.M. Gomez, and C. Uriel. J. Am.
Chem. Soc. 124, 3198 (2002).
11. H. Paulsen, M. Paal, D. Hadamczyk, and K.-M. Steiger.
Carbohydr. Res. 133, 71 (1984), C1-C5; H. Paulsen and
O. Lockhoff. Chem. Ber. 114, 3079 (1981); H. Paulsen, D.
Hadamczyk, W. Kutschker, and A. Bunsch. Liebigs. Ann.
Chem. 129 (1985).
12. S. Masamune, W. Choy, J.S. Petersen, and L.R. Sita. Angew.
Chem. Int. Ed. Engl. 23, 1 (1985).
13. N.M. Spijker and C.A.A. van Boeckel. Angew. Chem. 103,
179 (1990); Angew Chem. Int. Ed. Engl. 30, 180 (1991).
14. C.A.A. van Boeckel and M. Petitou. Angew Chem. 105, 1741
(1993); Angew Chem. Int. Ed. Engl. 32, 1671 (1993).
23. G.H. Veeneman, S.H. Van Leeuwen, H. Zurmond, and J.H. van
Boom. J. Carbohydr. Chem. 9, 783 (1990).
24. C.J.J. Elie, R. Verduyn, C.E. Dreff, D.M. Brounts, G.A. van
der Marel, and J.H. van Boom. Tetrahedron, 46, 8243 (1990).
25. G. Anilkumar, M. Gilbert, and B. Fraser-Reid. Tetrahedron,
56, 1993 (2000).
26. B.G. Wilson and B. Fraser-Reid. J. Org. Chem. 60, 317 (1995).
27. For excellent summaries of his work, see (a) A. Vasella. Bio-
organic chemistry carbohydrates. Edited by S.M. Hecht. Ox-
ford University Press, New York. 1999. Chap. 2; (b) A. Vasella.
A new approach to the synthesis of glycosides. Pure Appl.
Chem. 65, 731 (1994).
28. B. Bernet and A. Vasella. Helv. Chim. Acta. 83, 995 (2000).
© 2002 NRC Canada