Dehydrative glycosylation with the Hendrickson reagent

Article abstract of DOI:10.1021/jo2015856

The Hendrickson reagent is able to perform efficiently dehydrative glycosylation of 1-hydroxyglycosyl donors. The reaction occurs under mild conditions through an anomeric oxophosphonium intermediate detected by nuclear magnetic resonance. Further insight into the mechanism was gained by ¹⁸O labeling of anomeric OH.

Full text of DOI:10.1021/jo2015856

Note
pubs.acs.org/joc
Dehydrative Glycosylation with the Hendrickson Reagent
Matteo Mossotti and Luigi Panza*
Dipartimento di Scienze Chimiche, Alimentari, Farmaceutiche e Farmacologiche, Universita
28100 Novara, Italy
̀
del Piemonte Orientale, Via Bovio 6,
S
* Supporting Information
ABSTRACT: The Hendrickson reagent is able to perform efficiently
dehydrative glycosylation of 1-hydroxyglycosyl donors. The reaction occurs
under mild conditions through an anomeric oxophosphonium intermediate
detected by nuclear magnetic resonance. Further insight into the mechanism
was gained by 18O labeling of anomeric OH.
he importance of the glycosylation reaction is appreciated
today because of the biological relevance of oligosacchar-
formation of a reactive intermediate anomeric oxophospho-
nium salt.9
T
ides and glycoconjugates.1 On the other hand, despite the
progresses made in developing new procedures for the
synthesis of glycosides, such a reaction remains demanding
and no universal method for the synthesis of any glycosidic
linkage has emerged.
Careful bibliographic examination revealed only a paper by
Mukaiyama,10 who tested the POP reagent just for the
glycosylation of furanoses. In spite of the interest in the
reaction, the investigation was not extended, especially in the
case of more interesting pyranose hemiacetal donors,
prompting us to pursue our study.
The common way to perform glycosylations involves the
derivatization of the anomeric position of a sugar with a latent
leaving group, which is then activated in the presence of a
glycosyl acceptor to form the glycosidic bond.2 Less attention
has been paid to glycosidic bond formation through
dehydrative glycosylation, where the hydroxyl group of a
hemiacetal sugar can be converted into a leaving group in situ
and exploited directly for glycosylation.
Various conditions have been used for dehydrative
glycosylation, from the classical Fisher glycosylation protocol
to the elegant works of Gin et al., which exploited anomeric
oxosulfonium donors,3 and Shingu et al., who developed the
activation of 1-OH sugars by their in situ conversion to
bromides under dehydrative conditions using the Appel
reagent.4
Although the Hendrickson reagent is commonly used in the
presence of a base (see, e.g., ref 5), we initially considered the
possibility of activating the hemiacetal donor without addition
of a base as it is known that phosphine oxides are weak bases;11
therefore, the phosphine oxide liberated from POP reagent
upon activation of the hydroxyl group should buffer the formed
triflic acid. The POP reagent (1.1 equiv) was generated
according to a literature method8 from 3.3 equiv of Ph3PO
and 1.1 equiv of Tf2O at 0 °C in CH2Cl2; to the so formed
reagent was added 1 equiv of 2,3,4,6-tetra-O-benzylgalactopyr-
anose 1a. The reaction took place immediately, giving rise to an
almost quantitative yield of perbenzylated galacto trehaloses.
This observation suggests that the self-condensation between
the activated donor and the 1-OH sugar is faster than the
reaction of 1-OH with the POP reagent. Therefore, we
modified the order of addition of the donor and acceptor,
adding the acceptor (2-propanol, 2 equiv) to the POP reagent
before the donor. Under such conditions, the glycoside was
formed in fair yield (58%) (Scheme 1).
Among different dehydrating agents, the so-called Hen-
drickson reagent (POP), triphenylphosphonium anhydride
trifluoromethanesulfonate, was used for the synthesis of esters
and amides through oxophosphonium species.5−7
The result prompted us to employ more complex acceptors.
The use of cholesterol as an acceptor (2 equiv) gave an
intriguing result. Besides the expected glycoside 4a in 44% yield
and an α:β ratio of 63:37, we observed the formation of an
inseparable mixture of Δ3,5-cholestadiene12 and cholesteryl
ether13 (Scheme 1). Further attempts to glycosylate protected
sugar acceptors revealed that, despite the presence of
phosphine oxide, the reaction conditions were too acidic and
acid labile protecting groups, such as silyl ethers, did not
survive.14
Figure 1. Hendrickson POP reagent.
In an interesting work,8 it was also mechanistically studied in
comparison with the Mitsunobu reaction for ester formation.
As the Hendrickson reagent was quite rarely used, we thought
to try such a reagent in dehydrative glycosylation. It is worth
mentioning that Mukaiyama’s group described the use of
triphenylphosphine oxide to catalyze glycosylation reactions of
glycosyl bromides and iodides through the postulated
Received: July 29, 2011
Published: October 4, 2011
© 2011 American Chemical Society
9122
dx.doi.org/10.1021/jo2015856|J. Org. Chem. 2011, 76, 9122−9126

The Journal of Organic Chemistry
Note
Then we moved to 2,3,4,6-tetra-O-benzylglucopyranose 1b
as a donor. The activation was performed under the same
conditions previously described. Surprisingly, and in a manner
different from that of galactose, after 2 h the formation of a
relevant amount of trehalose was observed. This fact was quite
unexpected as it is known that tetrabenzylglucosyl donors are
less reactive than the corresponding galactosyl derivatives.15
Monitoring the reaction provided evidence that the formation
of trehalose was not immediate but its amount increased with
reaction time. Decreasing the temperature to −20 °C resulted
in an almost complete suppression of formation of the
undesired trehaloses, and addition of 2-propanol afforded the
desired glycoside 3b.
Scheme 1. Glycosylation under Acidic Conditions
The same acceptors previously glycosylated with 1a as a
donor were also used for reactions with 1b at −20 °C, giving
similar results.
Some experiments were performed with different solvents
and additives to improve the α:β ratio. The use of diethyl ether
and acetonitrile led to practical problems because of the low
solubility of the POP reagent; the reaction performed in DMF
gave a low yield, while the use of tetramethylurea instead of
DIPEA gave a similar yield and did not improve the
stereoselectivity. Finally, tetrabenzoylgalactose was used as a
donor, but no product formation was observed even the
reaction mixture was warmed, probably because it is too
disarmed.
From the examination of the anomeric ratio of the products
obtained in our set of experiments, and with the hypothesis that
alkoxyphosphonium ions are the reactive intermediates, some
information can be inferred. First, the different ratio between
the anomeric alkoxyphosphonium ions (see below) and the
related glycosides suggests that the α/β intermediate species
are in fast equilibrium; moreover, the amount of α anomer in
the glycoside mixture, which in some cases (Table 1, entry 4) is
the only product, indicates that the reaction does not occur
exclusively with direct displacement with inversion and that the
β intermediate appears to be more reactive than the α anomer
as can be deduced from the fact that less reactive nucleophiles
gave higher α:β ratios.16
We therefore moved toward the use of a stronger base and
ran the reaction under slightly modified conditions with respect
to those described previously (see Experimental Section),
adding a mixture of DIPEA and compound 1a to the
Hendrickson reagent and then leaving the mixture for 2 h at
0 °C to allow its activation by conversion into the
phosphonium salt. Under these conditions, formation of the
galacto trehalose was almost completely suppressed, and upon
addition of the acceptor, the expected glycosides were formed
in good yields.
The glycosylation reaction was then extended using other
acceptors, as shown in Table 1, in good yield.
It is worth noting that it was possible to satisfactorily
glycosylate a poorly reactive 4-OH (entry 4, product 6a);
moreover, the method is compatible with the presence of a
thioglycoside, therefore opening the possibility of orthogonal
glycosylations.
a
Table 1. Glycosylation of Tetrabenzyl Donors under Basic Conditions
a
b
c
See Experimental Section for reaction conditions. Isolated yield. Determined by 13C NMR.
9123
dx.doi.org/10.1021/jo2015856|J. Org. Chem. 2011, 76, 9122−9126

The Journal of Organic Chemistry
Note
To improve our understanding of the mechanism either
under acidic or under basic conditions, we performed the
reaction under both conditions using C1-OH 18O-labeled
tetrabenzylgalactose (prepared from tetrabenzylgalactose, POP
reagent, and 97% H218O; 77 ± 5% sugar labeling) as a donor
and 2-propanol as an acceptor.
Two possible patways can be envisaged for the glycosylation
reaction: the anomeric alkoxyphosphonium ion can be attacked
by the acceptor (Scheme 3, path a), or less probably, the
alkoxyphosphonium ion derived from the acceptor can be
substituted with the hemiacetal hydroxyl group (Scheme 3,
path b). One mechanism or the other could be effective
depending on the conditions, the order of addition of two
reagents, and the rate of the equilibria between the different
ions in solution.
45.0 ppm was also present, probably because of the excess of
POP reagent in fast equilibrium with Ph3PO liberated during
activation of the anomeric hydroxyl group. The 13C NMR
spectrum revealed a doublet for the anomeric carbon at 103.2
ppm with a JC,P of 9.17 Hz, while the minor isomer was not
detectable. The 1H NMR spectrum also exhibited the presence
of a doublet of doublets (J = 3.1 and 6.7 Hz) because of an
anomeric proton at 5.59 ppm. Unfortunately, again it was not
possible to detect the signal for the other anomer as it was most
likely hidden under the signal of CDHCl2 around 5.3 ppm. The
chemical shifts in the 31P NMR spectrum and the coupling
1
constant in the H NMR spectrum strongly suggest that the
major anomer is the α one.
Upon addition of the acceptor, the signal at 5.6 ppm in the
1H NMR spectrum quickly almost disappeared as well as the
signals at 66.1 and 64.0 ppm in the 31P NMR spectrum, where a
new signal at 59.6 ppm, which can be attributed to the
phosphonium salt of the acceptor, became visible.
Running the reaction under basic conditions led to the
recovery of 18O-labeled Ph3PO (25 ± 5% 18O incorpo-
ration); surprisingly, under acidic conditions, the 18O was found
neither on Ph3PO nor on the glycoside.
To detect the signal for the other anomeric proton, we
repeated the experiment in CDCl3. After activation for 30 min,
the 31P NMR spectrum showed again two peaks at 66.2 and
64.1 ppm in a 26:74 ratio. The proton spectrum revealed a
situation more complex than the one that was expected,
showing, apart from a doublet of doublets at 5.47 ppm (J = 3.1
and 6.7 Hz), other signals around 5.3 ppm that may due to the
β anomer together with other unidentified signals that did not
allow confident identification of the signal due to the β anomer.
Comparing the rate and stereoselection of our reactions,
especially in the case of more reactive acceptors, with those
described with Appel-Lee reagent in ref 4 suggests that in our
case the reaction is much faster but much less stereoselective.
The rate difference, together with the observation of the
prevalence of the α anomer of the alkoxyphosphonium
intermediate, can help to explain the difference in stereo-
selectivity. In fact, in both cases in situ anomerization probably
takes place. However, in less reactive anomeric bromides, the
difference in the activation energy for the attack on the more
reactive β anomer with respect to the α one allows generation
of mainly the α glycoside, while in more reactive anomeric
alkoxyphosphonium, the activation energy difference is not
enough to minimize the reactivity of the α-alkoxyphosphonium
giving rise to an anomeric mixture of the glycosides.
We could therefore conclude that the more likely pathway
presumably followed by the reaction is path a (Scheme 2)
Scheme 2. Possible Mechanisms for the Glycosylation
Reaction
under both acidic and basic conditions, but under acidic
conditions, the 18O may be scrambled, ending with the loss of
labeling in water (see Scheme 3).
In conclusion, an approach to dehydrative glycosylation with
1-hydroxyglycopyranosyl donors by means of the POP
Hendrickson reagent has been established. The use of 18O-
labeled donors indicates that the activation occurs through the
formation of an alkoxyphosphonium intermediate, further
confirmed by NMR analysis of the dehydrative glycosylation
process. The operative conditions are quite simple, avoiding
exceedingly low temperatures and the unwanted formation of
donor self-condensation products.
The obtained results open the possibility of improving the
reaction by testing other phosphine oxides and modifying the
conditions; moreover, the procedure will be applied to
reactions with different donors and acceptors and to activate
the anomeric position to promote the substitution reaction with
nucleophiles other than oxygen.
Scheme 3. Possible Loss of Labeling under Acidic
Conditions
From a practical point of view, the basic conditions provided
far better results; therefore, the mechanism of the reaction
under acidic conditions was not further investigated.
On the other hand, to improve our understanding of the
mechanism under basic conditions, the detection of reactive
intermediates was attempted by NMR at low temperatures. In
an NMR tube, tetrabenzylgalactose 1a (50 mg) was added at
−10 °C to a suspension of 2 equiv of POP reagent17 in CD2Cl2
(giving a solution that was submitted to 1H, 13C, and 31P NMR
at the same temperature). The 31P spectrum exhibited a main
peak at 64.08 ppm compatible with the formation of a major
intermediate alkoxyphosphonium ion together with a minor
peak at 66.16 ppm in an 88:12 ratio.18 A broad peak around
EXPERIMENTAL SECTION
■
General Experimental Methods. Flash column chromatography
was performed on Merck silica gel 60 (0.040−0.063 mm), following
the procedure described in ref 19. Reactions were monitored by TLC
on Merk silica gel 60 F254 plates, and the compounds were detected by
9124
dx.doi.org/10.1021/jo2015856|J. Org. Chem. 2011, 76, 9122−9126

The Journal of Organic Chemistry
Note
examination under UV light and by charring with 10% sulfuric acid in
methanol. Solvents were removed under reduced pressure below 40
°C. CH2Cl2 was dried and stored over 4 Å molecular sieves. All
reactions (if not specifically including water as a reactant, solvent, or
cosolvent) were performed under an Ar atmosphere, in oven-dried or
3.83 (dd, 0.58 H, J = 7.6, 9.8 Hz, H-2β), 3.65−3.45 (m, 3.16 H, H-3β,
H-5β, 2 H-6), 2.51−2.25 (m, 2 H), 2.08−1.79 (m, 5 H), 1.68−0.82
(m, 38 H), 0.70 (s, 3 H); 13C NMR (CDCl3, 75 MHz) δ 141.2 (s, C-5
chol. of α anomer), 141.1 (s, C-5 chol. of β anomer), 139.5−138.4
(cluster of s, quaternary Ar), 128.7−127.5 (cluster of d, CH Ar), 122.0
(d, C-6 chol. of β anomer), 121.9 (d, C-6 chol. of α anomer), 102.8 (d,
C-1β), 95.8 (d, C-1α), 82.7 (d), 80.0 (d), 79.8 (d), 79.6 (d), 77.1 (d),
76.9 (d), 75.6 (t), 75.5 (d), 75.1 (t), 74.8 (t), 73.9 (d), 73.8 (t), 73.7
(t), 73.7 (d), 73.5 (t), 73.5 (t), 73.4 (t), 69.6 (d), 69.5 (t), 69.4 (t),
57.1 (d), 57.1 (d), 56.5 (d), 50.5 (d), 50.5 (d), 42.7 (t, 2 C), 42.7 (s, 2
C), 40.1 (t), 40.0 (t), 39.8 (t), 39.3 (t), 37.6 (t), 37.5 (t), 37.1 (q),
36.5 (q), 36.1 (d), 32.3 (t, 2 C), 32.2 (d, 2 C), 28.6 (t, 2 C), 28.3 (d, 2
C), 24.6 (t, 2 C), 24.1 (t, 2 C), 23.1 (q, 2 C), 22.9 (q, 2 C), 21.4 (t, 2
C), 19.7 (q, 2 C), 19.0 (q, 2 C), 12.1 (q, 2 C). Elemental anal. Calcd
for C61H80O6: C, 80.57%; H, 8.87%. Found: C, 80.43%; H, 8.65%.
Benzyl 2,3,4,6-tetra-O-benzyl-D-galactopyranosyl-(1→6)-
2,3,4-tri-O-benzyl-β-D-glucopyranoside (5a).22 was obtained by
method A as a colorless syrup in 98% yield (60:40 α:β ratio): Rf = 0.5
[8:2 (v/v) cyclohexane/ethyl acetate]; 1H NMR (CDCl3, 300 MHz) δ
7.5−7.2 (m, 40 H), 5.12 (d, 0.6 H, J = 3.4 Hz, H-1′α), 5.05−4.38 [m,
17.4 H, 8 CH2Ph, H-1′β, H-1(α1′→6), H-1(β1′→6)], 4.22 [br d, 0.4
H, J = 10.6 Hz, H-6(β1′→6)], 4.12−4.00 (m, 1.4 H), 4.0−3.8 (m, 3.2
H), 3.76−3.32 (m, 7 H); 13C NMR (CDCl3, 75 MHz) δ 139.5−137.9
(cluster of s, Ar), 128.8−127.8 (cluster of d, CH Ar), 104.7 (d, C-1′β),
103.0 [d, C-1(β1′→6)], 102.9 [d, C-1(α1′→6)], 98.4 (d, C-1′α), 85.2
(d), 85.1 (d), 82.9 (d), 82.7 (d), 82.6 (d), 79.9 (d), 78.9 (d), 78.8 (d),
78.4 (d), 76.2 (t), 76.0 (t), 75.8 (d), 75.6 (t), 75.5 (t), 75.4 (t), 75.2
(t), 74.0 (d), 74.0 (t), 73.8 (t), 73.5 (t), 73.0 (d), 71.6 (t), 71.4 (t),
69.8 (d), 69.5 (t), 69.1 (t), 68.8 (t), 66.6 (t). Elemental anal. Calcd for
C68H70O11: C, 76.81%; H, 6.64%. Found: C, 76.73%; H, 6.55%.
Phenyl tetra-O-benzyl-α-D-galactopyranosyl-(1→4)-2,3-di-
O-benzoyl-6-O-tert-butyldiphenylsilyl-1-thio-β-D-glucopyrano-
1
microwave oven-dried glassware. Unless noted otherwise, H NMR
spectra were recorded in CDCl3 at 300 MHz, 13C NMR spectra at 75
1
MHz with chloroform (7.27 ppm for H, 77.20 ppm for 13C) as an
internal reference, and 31P NMR spectra in CDCl3 at 121.5 MHz with
H3PO4 as an external reference. Chemical shifts (δ) are given in parts
per million; multiplicities are indicated as s (singlet), d (doublet), t
(triplet), q (quartet), m (multiplet), and br (broad). Coupling
constants (J) are reported in hertz.
Glycosylation Reactions. Method A. Typical Procedure from
a Galactosyl Donor. To a solution of triphenylphosphine oxide (175
mg, 6.75 equiv) in dry CH2Cl2 (3 mL) was added dropwise the
trifluoromethanesulfonic anhydride (0.046 mL, 3 equiv) at 0 °C. After
the mixture had been stirred for 30 min at the same temperature, a
solution of 2,3,4,6-tetra-O-benzyl-D-galactopyranose (100 mg, 2 equiv)
and diisopropylethylamine (0.054 mL, 3.3 equiv) in dry CH2Cl2 (2
mL) was added dropwise. Stirring was continued for 30 min at 0 °C,
and then a solution of acceptor (1 equiv) and diisopropylethylamine
(0.054 mL, 3.3 equiv) in dry CH2Cl2 (2 mL) was added dropwise, at
the same temperature. After the reaction had reached completion
(∼15 min), the mixture was diluted with EtOAc and washed with a
saturated NaHCO3 solution, water, and brine; the organic phase was
dried over Na2SO4 and concentrated in vacuo. The residue was
purified by flash chromatography (cyclohexane/ethyl acetate) to afford
the desired compound (3a, 4a, 5a, or 6a). All products were
characterized as anomeric mixtures.
Method B. Typical Procedure from a Glucosyl Donor. To a
solution of triphenylphosphine oxide (175 mg, 6.75 equiv) in dry
CH2Cl2 (3 mL) was added dropwise the trifluoromethanesulfonic
anhydride (0.046 mL, 3 equiv) at 0 °C. After the mixture had been
stirred for 30 min at the same temperature, the reaction mixture was
cooled to −20 °C and a solution of 2,3,4,6-tetra-O-benzyl-D-
glucopyranose (100 mg, 2 equiv) and diisopropylethylamine (0.054
mL, 3.3 equiv) in dry CH2Cl2 (2 mL) was added dropwise. Stirring
was continued for 30 min at −20 °C, and then a solution of acceptor
(1 equiv) and diisopropylethylamine (0.054 mL, 3.3 equiv) in dry
CH2Cl2 (2 mL) was added dropwise, at the same temperature. After
the reaction had reached completion (∼15 min), the mixture was
diluted with EtOAc and washed with a saturated NaHCO3 solution,
water, and brine; the organic phase was dried over Na2SO4 and
concentrated in vacuo. The residue was purified by flash
chromatography (cyclohexane/ethyl acetate) to afford the desired
compound (3b, 4b, 5b, or 6b). All products were characterized as
anomeric mixtures.
side (6a). was obtained by method A as a colorless syrup in 75% yield
20
(α only): Rf = 0.59 [8:2 (v/v) cyclohexane/ethyl acetate]; [α]D
=
+43.5° (c = 1.00, CHCl3); 1H NMR (CDCl3, 300 MHz) δ 7.97−7.90
(m, 4 H), 7.77 (q, 4 H, J = 6.1, 18.7 Hz), 7.55−7.05 (m, 37 H), 5.83
(t, 1 H, J = 9.5, H-3), 5.48 (t, 1 H, J = 9.5, H-2), 5.15 (d, 1 H, J = 3.1
Hz, H-1′α), 4.99 (d, 1 H, J = 9.8 Hz, H-1), 4.83 (d, 1 H, J = 11.0 Hz,
CHPh), 4.57−4.45 (m, 3 H, CHPh, CH2Ph), 4.35−4.16 (m, 6 H, H-4,
H-4′, H-6a, CHPh, CH2Ph), 3.95 (d, 1 H, J = 10.4, H-6b), 3.93 (d, 1
H, J = 12.5 Hz, CHPh), 3.84−3.63 (m, 4 H, H-5, H-2′, H-3′, H-5′),
3.33 (d, 2 H, J = 6.1 Hz, H-6′a,b), 1.07 (s, 9 H, 3 CH3); 13C NMR
(CDCl3, 75 MHz) δ 164.6 (s, CO), 164.2 (s, CO), 137.8 (s),
137.5 (s), 137.2 (s), 136.8 (s), 134.8 (d), 134.6 (d), 132.6 (s), 132.2
(s), 131.9 (d), 131.8 (s), 131.6 (d), 130.4 (d), 128.2 (s), 128.6−126.2
(cluster of d, CH Ar), 97.9 (d, C-1′), 85.4 (C-1), 79.2 (d), 77.5 (d),
74.7 (d), 74.5 (d), 73.9 (d), 73.8 (d), 73.3 (t), 72.1 (t, 2 C), 71.4 (t),
69.8 (d), 69.3 (d), 67.9 (t), 62.1 (t), 25.8 (q, 3 CH3), 18.1 [s,
C(CH3)3]. Elemental anal. Calcd for C76H76O12SSi: C, 73.52%; H,
6.17%. Found: C, 73.43%; H, 6.08%.
Isopropyl 2,3,4,6-tetra-O-benzylgalactopyranoside (3a).20
was obtained by method A as a colorless syrup in 98% yield (44:56 α:β
ratio): Rf = 0.55 [8:2 (v/v) cyclohexane/ethyl acetate]; 1H NMR
(CDCl3, 300 MHz) δ 7.45−7.20 (m, 20 H), 4.99 (d, 0.44 H, J = 3.5
Hz, H-1α), 4.98−4.39 (m, 8.56 H, 4 CH2Ph and H-1β), 4.09−3.86 [m,
3.32 H, H-2α, H-3α, H-4α, H-5α, H-4β, CH(CH3)2], 3.82 (dd, 1 H, J
= 7.6, 9.8 Hz, H-2β), 3.63−3.47 (m, 3.12 H, H-3β, H-5β, 2 H-6),
1.31−1.19 (m, 6 H, 2 CH3); 13C NMR (CDCl3, 75 MHz) δ 139.5−
138.4 (cluster of s, Ar), 128.7−127.8 (cluster of d, CH Ar), 102.8 (d,
C-1β), 95.8 (d, C-1α), 82.7 (d), 79.9 (d), 79.6 (d), 76.8 (d), 75.5 (d),
75.4 (t), 75.0 (t), 74.7 (t), 73.9 (d), 73.8 (t), 73.7 (t), 73.7 (d), 73.5
(t), 73.5 (t), 73.4 (t), 72.4 (d), 69.5 (d, 2 C), 69.4 (t, 2 C), 23.9 (q),
23.5 (q), 22.4 (q), 21.6 (q). Elemental anal. Calcd for C37H42O6: C,
76.26%; H, 7.26%. Found: C, 76.31%; H, 7.41%.
Isopropyl 2,3,4,6-tetra-O-benzylglucopyranoside (3b).23
was obtained by method B as a white solid in 97% yield (25:75 α:β
ratio): Rf = 0.59 [8:2 (v/v) cyclohexane/ethyl acetate]; 1H NMR
(CDCl3, 300 MHz) δ 7.45−7.13 (m, 20 H), 4.90 (d, 0.25 H, J = 3.6
Hz, H-1α), 5.05−4.43 (m, 8.75 H, 4 CH2Ph and H-1β), 4.09−3.97 [m,
1 H, H-3α, CH(CH3)2β], 396−3.82 [m, 0.5 H, H-5α, CH(CH3)2α],
3.80−3.41 (m, 5.5 H, H-2α, H-2β, H-3β, H-4, H-5β, 2 H-6), 1.33 (d,
2.25 H, J = 6.1 Hz, CH3β), 1.30−1.22 (m, 3 H, CH3β and CH3α), 1.19
(d, 0.75 H, J = 6.1 Hz, CH3α); 13C NMR (CDCl3, 75 MHz) δ 139.2−
138.5 (cluster of s, Ar), 129.0−127.0 (cluster of d, CH Ar), 102.7 (d,
C-1β), 95.2 (d, C-1α), 85.3 (d), 82.7 (d), 82.6 (d), 80.4 (d), 78.5 (d),
78.4 (d), 76.1 (t, 2 C), 75.6 (t), 75.4 (t), 75.3 (t), 73.9 (d), 73.6 (t),
72.8 (d), 70.5 (d), 69.7 (t), 69.5 (d), 69.0 (t), 24.2 (q), 23.6 (q), 22.7
(q), 21.6 (q). Elemental anal. Calcd for C37H42O6: C, 76.26%; H,
7.26%. Found: C, 76.48%; H, 7.36%.
Cholesteryl 2,3,4,6-tetra-O-benzylgalactopyranoside
(4a).21 was obtained by method A as a light yellow syrup in 95%
yield (42:58 α:β ratio): Rf = 0.75 [8:2 (v/v) cyclohexane/ethyl
acetate]; H NMR (CDCl3, 300 MHz) δ 7.45−7.20 (m, 20 H), 5.34
Cholesteryl 2,3,4,6-tetra-O-benzylglucopyranoside (4b).24
was obtained by method B as a colorless syrup in 95% yield (27:73 α:β
ratio): Rf = 0.73 [8:2 (v/v) cyclohexane/ethyl acetate]; 1H NMR
(CDCl3, 300 MHz) δ 7.45−7.15 (m, 20 H), 5.37 (br d, 0.73 H, J = 4.9
Hz, H-6 chol. in β anomer), 5.31 (br d, 0.27 H, J = 4.9 Hz, H-6 chol. in
1
(br d, 0.58 H, J = 4.6 Hz, H-6 chol. in β anomer), 5.28 (br d, 0.42 H, J
= 4.3 Hz, H-6 chol. in α anomer), 5.01 (d, 0.42 H, J = 3.4 Hz, H-1α),
5.00−4.39 (m, 8.58 H, 4 CH2Ph and H-1β), 4.12−3.87 (m, 2.68 H, H-
2α, H-3α, H-4α, H-5α, H-3 chol.), 3.89 (d, 0.58 H, J = 2.4 Hz, H-4β),
9125
dx.doi.org/10.1021/jo2015856|J. Org. Chem. 2011, 76, 9122−9126

The Journal of Organic Chemistry
Note
α anomer), 5.06−4.43 (m, 9 H, 4 CH2Ph, H-1α and H-1β), 4.03 (t,
0.27 H, J = 9.4 Hz, H-3α), 3.91 (br d, 0.27 H, H-5α), 3.81−3.42 (m,
6.46 H, H-2α, H-2β, H-3β, H-4α, H-4β, H-5β, 2 H-6, H-3 chol.),
2.51−2.25 (m, 2 H), 2.15−1.79 (m, 5 H), 1.68−0.82 (m, 38 H), 0.70
(s, 3 H); 13C NMR (CDCl3, 75 MHz) δ 141.3 (s, C-5 chol. of α
anomer), 141.1 (s, C-5 chol. of β anomer), 139.5−138.4 (cluster of s,
quaternary Ar), 129.5−127.8 (cluster of d, CH Ar), 122.4 (d, C-6 chol.
of β anomer), 122.2 (d, C-6 chol. of α anomer), 102.7 (d, C-1β), 95.1
(d, C-1α), 85.3 (d), 82.8 (d), 82.6 (d), 80.4 (d), 80.1 (d), 78.5 (d),
78.4 (d), 77.0 (d), 76.2 (t), 75.6 (t), 75.5 (t), 75.3 (d), 73.9 (t), 73.6
(t), 70.5 (d), 69.7 (t), 69.0 (t), 57.2 (d), 56.6 (d), 50.7 (d), 50.6 (d),
42.8 (t), 42.8 (s), 40.1 (t), 40.4 (t), 40.3 (t), 40.0 (t), 39.6 (t), 37.8 (t),
37.2 (q), 36.7 (q), 36.3 (d), 32.4 (t, 2 C), 32.3 (d, 2 C), 28.7 (t, 2 C),
28.5 (d, 2 C), 24.7 (t, 2 C), 24.3 (t, 2 C), 23.3 (q, 2 C), 23.1 (q, 2 C),
21.6 (t, 2 C), 19.9 (q, 2 C), 19.2 (q, 2 C), 12.3 (q, 2 C). Elemental
anal. Calcd for C61H80O6: C, 80.57%; H, 8.87%. Found: C, 80.29%; H,
8.94%.
Benzyl 2,3,4,6-tetra-O-benzyl-D-glucopyranosyl-(1→6)-
2,3,4-tri-O-benzyl-β-D-glucopyranoside (5b).25 was obtained by
method B as a white solid in 96% yield (37:63 α:β ratio): Rf = 0.5 [8:2
(v/v) cyclohexane/ethyl acetate]; 1H NMR (CDCl3, 300 MHz) δ
7.5−7.2 (m, 40 H), 5.03 (d, 0.37 H, J = 3.7 Hz, H-1′α), 5.08−4.41 [m,
16.63 H, 8 CH2Ph, H-1′β, H-1(α1′→6), H-1(β1′→6)], 4.24 [br d, 0.63
H, J = 10.5 Hz, H-6(β1′→6)], 4.28 (t, 0.37 H, J = 9.2 Hz, H-3′α), 3.93
(m, 0.37 H, H-5′α), 3.86−3.35 (m, 10.63 H); 13C NMR (CDCl3, 75
MHz) δ 139.2−137.8 (cluster of s, Ar), 128.9−127.7 (cluster of d, CH
Ar), 104.5 (d, C-1′β), 103.1 [d, C-1(β1′→6)], 102.7 [d, C-1(α1′→6)],
97.6 (d, C-1′α), 85.3 (d), 85.2 (d), 82.9 (d), 82.8 (d), 82.6 (d), 82.3
(d), 80.5 (d), 78.8 (d), 78.3 (d), 76.2 (t), 76.1 (t), 75.6 (d), 75.4 (t),
75.3 (t), 75.2 (t), 74.0 (t), 73.9 (t), 73.0 (t), 71.6 (t), 71.4 (t), 70.6
(d), 69.4 (t), 69.1 (t), 68.9 (t). Elemental anal. Calcd for C68H70O11:
C, 76.81%; H, 6.64%. Found: C, 76.89%; H, 6.77%.
REFERENCES
(1) Dwek, R. A. Chem. Rev. 1996, 96, 683−720.
(2) Zhu, X.; Schmidt, R. R. Angew. Chem., Int. Ed. 2009, 48, 2−37.
(3) Garcia, B. A.; Gin, D. Y. J. Am. Chem. Soc. 2000, 122, 4269−4279,
and references cited therein.
(4) Shingu, Y.; Miyachi, A.; Miura, Y.; Kobayashi, K.; Nishida, Y.
Carbohydr. Res. 2004, 340, 2236−2244, and references cited therein.
(5) Hendrickson, J. B.; Hussoin, M. S. J. Org. Chem. 1989, 54, 1144−
1149.
(6) Hendrickson, J. B.; Hussoin, M. S. J. Org. Chem. 1987, 52, 4137−
4139.
(7) For recent papers on the Hendrickson reagent, see: (a) Xu, P.;
Liu, G.-S.; Xi, J.; Wang, S.; Yao, Z.-J. Tetrahedron 2011, 67, 5455−
5460. (b) Moussa, Z.; Ahmed, S. A.; ElDouhaibi, A. S.; Al-Raqa, S. Y.
Tetrahedron Lett. 2010, 51, 1826−1831.
(8) Elson, K. E.; Jenkins, I. D.; Loughlin, W. A. Org. Biomol. Chem.
2003, 1, 2958−2965.
(9) Mukaiyama, T.; Kobashi, Y. Chem. Lett. 2004, 33, 10−11.
(10) Mukaiyama, T.; Suda, S. Chem. Lett. 1990, 1143−1146.
(11) Quin, L. D. A Guide to Organophosphorous Chemistry; John
Wiley and Sons: New York, 2000; Chapter 4, p 103.
(12) Xiong, Q.; Wilson, W. K.; Pang, J. Lipids 2007, 42, 87−96.
(13) Kowalski, J.; Łotowski, Z.; Morzycki, J. W.; Płoszyn
Sobkowiak, A.; Wilczewska, A. Z. Steroids 2008, 73, 543−548.
(14) As the results of the reaction conducted under acidic conditions
were unsatisfactory, detailed experimental conditions have not been
reported.
(15) Zhang, Z.; Ollmann, I. R.; Ye, X.-S.; Wischnat, R.; Baasov, T.;
Wong, C.-H. J. Am. Chem. Soc. 1999, 121, 734−753.
(16) For a related example, see ref 8. For a general discussion of the
in situ anomerization, see, for example: Kulkami, S. S.; Gervay-Hague,
J. In Handbook of Chemical Glycosylation; Demchenko, A. V., Ed.;
Wiley-VCH: Berlin, 2008; Chapter 2, pp 71 and ff.
(17) The chemical shift of the POP reagent was 75.5 ppm as
previously described. See, for example: Crich, D.; Dyker, H.
Tetrahedron Lett. 1989, 30, 475. Petersson, M. J.; Jenkins, I. D.;
Loughlin, W. A. J. Org. Chem. 2008, 73, 4691−4693.
(18) The 31P chemical shifts of alkoxyphosphonium salts that we
measured are different from those reported in ref 10 for furanoses but
are in agreement with the values usually reported in the literature. See,
for example: Camp, D.; Jenkins, I. D. J. Org. Chem. 1989, 54, 3045.
(19) Still, W. C.; Kahn, M.; Mitra, A. J. Org. Chem. 1978, 43, 2923−
2925.
■
́
ska, J.;
Phenyl tetra-O-benzyl-α-D-glucopyranosyl-(1→4)-2,3-di-O-
benzoyl-6-O-tert-butyldiphenylsilyl-1-thio-β-D-glucopyrano-
side (6b). was obtained by method B as a colorless syrup in 72% yield
20
(α only): Rf = 0.6 [8:2 (v/v) cyclohexane/ethyl acetate]; [α]D
=
+43.9° (c = 1.00, CHCl3); 1H NMR (CDCl3, 300 MHz) δ 7.99−7.93
(m, 4 H), 7.77 (q, 4 H, J = 6.1, 18.7 Hz), 7.55−7.05 (m, 37 H), 5.84
(t, 1 H, J = 9.1 Hz, H-3), 5.49 (t, 1 H, J = 9.5 Hz, H-2), 5.09 (d, 1 H, J
= 3.1 Hz, H-1α), 5.00 (d, 1 H, J = 9.8 Hz, H-1), 4.77 (d, 1 H, J = 11.0
Hz, CHPh), 4.65 (s, 2 H, CH2Ph), 4.45 (d, 1 H, J = 11.9 Hz, CHPh),
4.4 (d, 1 H, J = 11.0 Hz, CHPh), 4.36−4.21 (m, 4 H, 2 CHPh, H-4, H-
6a), 4.04 (d, 1 H, J = 11.64 Hz, H-6b), 3.96−3.82 (m, 3 H, H-3′,
CH2Ph), 3.77 (br d, 1 H, H-5′), 3.67 (br d, 1 H, H-5), 3.58−3.45 (m, 2
H, H-4′, H-6′a), 3.41−3.25 (m, 2 H, H-2′, H-6′b), 1.08 (s, 9 H, 3
CH3); 13C NMR (CDCl3, 75 MHz) δ 164.6 (s, CO), 164.3 (s, C
O), 137.8 (s), 137.4 (s), 137.1 (s), 136.9 (s), 134.9 (d), 134.6 (d),
132.7 (s), 132.6 (s), 132.1 (d), 131.9 (s), 131.7 (d), 130.5 (d), 128.8
(d), 128.5 (d), 128.4 (d), 128.3 (s), 127.9−126.2 (cluster of d, CH
Ar), 97.9 (d, C-1′), 85.5 (C-1), 80.4 (d), 79.2 (d), 78.2 (d), 76.3 (d),
74.7 (d), 74.4 (d), 74.3 (t), 73.8 (t), 72.3 (t), 71.4 (t), 70.6 (d), 69.9
(t), 67.5 (t), 61.9 (t), 25.9 (q, 3 CH3), 18.3 [s, C(CH3)3]. Elemental
anal. Calcd for C76H76O12SSi: C, 73.52%; H, 6.17%. Found: C,
73.73%; H, 6.25%.
(20) Mereyala, H. B.; Reddy, G. V. Tetrahedron 1991, 47, 6435−
6448.
(21) (a) Shingu, Y.; Nishida, Y.; Dohi, H.; Kobayashi, K. Org. Biomol.
Chem. 2003, 1, 2518−2521. (b) Nishizawa, M.; Garcia, D. M.; Shin,
T.; Yamada, H. Chem. Pharm. Bull. 1993, 41, 784−785.
(22) Hashimoto, S.; Honda, T.; Ikegami, S. Chem. Pharm. Bull. 1990,
38, 2323−2325.
(23) Bucher, C.; Gilmour, R. Angew. Chem., Int. Ed. 2010, 49, 8724−
8728.
(24) Vankaylapati, H.; Singh, G.; Tranoy, I. Tetrahedron: Asymmetry
2001, 12, 1373−1381.
(25) Hirooka, M.; Koto, S. Bull. Chem. Soc. Jpn. 1998, 71, 2893−
2902.
ASSOCIATED CONTENT
■
S
* Supporting Information
1H and 13C NMR spectra of the products and H, 13C, and 31
P
1
NOTE ADDED AFTER ASAP PUBLICATION
■
NMR spectra of the intermediate. This material is available free
of charge via the Internet at http://pubs.acs.org.
During the preparation of the manuscript the authors were not
aware of the works of Prof. Chapleur’s group on anomeric
aminophosphonium salts, see: Chretien, F.; Chapleur, Y.;
Castro, B.; Gross, B. J. Chem. Soc., Perkin Trans. 1 1980, 381−
384, and references cited therein. This paper published ASAP
on October 11, 2011 and then reposted with this note October
14, 2011.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: panza@pharm.unipmn.it.
DEDICATION
■
Dedicated to the memory of Prof. David Y. Gin.
9126
dx.doi.org/10.1021/jo2015856|J. Org. Chem. 2011, 76, 9122−9126