Formation and stability of oxocarbenium ions from glycosides

Article abstract of DOI:10.1002/jms.880

Structural, protecting group and leaving group effects in the formation of oxocarbenium intermediates were studied in the gas phase. It is found that significant stabilization of oxocarbenium cations is achieved by protecting groups that interact with the

Full text of DOI:10.1002/jms.880

JOURNAL OF MASS SPECTROMETRY
J. Mass Spectrom. 2005; 40: 1055–1063
Published online 22 June 2005 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jms.880
Formation and stability of oxocarbenium ions from
glycosides
Chagit Denekamp∗ and Yana Sandlers
Department of Chemistry, Institute of Catalysis Science and Technology, Technion – Israel Institute of Technology, Haifa 32000, Israel
Received 10 February 2005; Accepted 11 April 2005
Structural, protecting group and leaving group effects in the formation of oxocarbenium intermediates
were studied in the gas phase. It is found that significant stabilization of oxocarbenium cations is achieved
by protecting groups that interact with the cationic center via neighboring group participation despite the
electron-withdrawing character of these moieties. On the other hand, ethereal protecting groups do not
facilitate the formation of oxocarbenium intermediates. The experimental findings are supported by DFT
calculations that show the following order of stabilization by the group adjacent to the cationic center:
RCO > SiR3 > R, where R is an alkyl group. This indicates that the SN1-like mechanism that is commonly
proposed for this reaction is not always valid. Moderate leaving group effect is also detected in a series of
thioaryl glucopyranosides. Copyright  2005 John Wiley & Sons, Ltd.
KEYWORDS: glycosylation; oxocarbenium; mass spectrometry; electrospray ionization
groups. Much effort has been invested in the design of suit-
able leaving groups at the anomeric position, which allow
selective activation of the donor, while the functions of the
counterpart sugar (acceptor) are all protected except for the
site that is to be linked.15–20 Several reports show that a large
enhancement in the rate of glycosylation is observed when
the hydroxy group at position 4 of the sugar ring of the
donor is axial (e.g. galactose).21–24 Other kinetic effects that
were observed are attributed to the nature of the protect-
ing groups.21–23,25,26 Thus, electron-withdrawing protecting
groups are thought to destabilize the oxocarbenium interme-
diate while electron-donating moieties stabilize the reaction
intermediate and thus enhance the reaction.21–23 An ultimate
goal in this field is to establish a reactivity scale that will
allow the design of the oligosaccharides and prepare them
in a one pot synthesis of a mixture of donors with different
reactivities.21,27,28
Mass spectrometry allows the isolation of reactive
intermediates in the gas phase for the study of their
relative stability. Electrospray ionization (ESI) enables the
formation of cationized sugar moieties from solution in order
to follow their decomposition in the gas phase. Collision
induced dissociation (CID) spectra of protonated or ionized
isomeric pyranosides have been measured by others using
chemical ionization (CI), electron impact (EI) or fast atom
bombardment (FAB) as ionization methods.29–34 The purpose
of these studies was to develop methodologies for isomeric
distinction using mass spectrometry or the determination of
linkage position.35 The experimental conditions chosen for
this investigation are less energetic than previous studies as
the ions are produced with electrospray ionization and the
CID carried out using low energy on [M C NH4]C ion. We
report here a systematic study that was conducted in order to
elucidate the structural, protecting group and leaving group
INTRODUCTION
The study of carbohydrate biology has seen enormous
growth in the past decade, and as a result there is an increas-
ing need for synthetic methods for their preparation.1–4 The
key reaction in the synthesis of oligosaccharides is the gly-
cosylation step, in which the various monosaccharide units
are bound according to a well designed sequence. The acid-
catalyzed synthetic strategies for glycosylation are based on
the idea that the reactions are stepwise SN1-like, mediated by
the formation of oxocabenium species.3 Several mechanisms
have been considered either dissociative or partially disso-
ciative through the participation of ion pairs.5–12 Regarding
the stereoselectivity of the glycosylation (˛ vs ˇ), a useful
synthetic strategy is based on anchimeric assistance of 2-O-
acyl groups of the so-called donors, supporting the notion
that oxocarbenium intermediates are indeed formed in these
reactions.13,14 It has also been proposed that the involve-
ment of solvent-separated ion pair between the anomeric
carbon and the leaving group result in better reactivity
in polar solvents that stabilize charge-separated species.2
For reactions performed under basic conditions, concerted
SN2-like mechanisms were also considered. Specifically for
glucopyranosides, a stepwise transacetalization mechanism
that involves ring opening and closure was proposed.3 There
are several factors that may affect the course of reaction: sol-
vent, leaving group at the anomeric position, the Lewis acid
(promoter) that is being used to activate the glycosyl donor,
and, as carbohydrates possess several hydroxy groups, the
stereochemistry of the donor and the nature of the protecting
ŁCorrespondence to: Chagit Denekamp, Department of Chemistry,
Technion – Israel Institute of Technology, Haifa 32000, Israel.
E-mail: chchagit@tx.technion.ac.il
Contract/grant sponsor: Israel Science Foundation.
Copyright  2005 John Wiley & Sons, Ltd.

1056
C. Denekamp and Y. Sandlers
effects on the formation of oxocarbenium intermediates
from protected glucosides and galactosides using ESI FT-
ICR (Fourier Transform Ion Cyclotron Resonance) mass
spectrometry.
7.0 mmol) was dissolved in 40 ml of freshly distilled DMF
°
and cooled to ꢀ10 C under nitrogen. NaH (1.7 g, 70 mmol)
was added, the mixture was stirred for 10 minutes at
room temperature followed by dropwise addition of benzyl
bromide (7.6 ml, 70 mmol). The reaction mixture was stirred
°
at room temperature for 5 h, quenched by water at 0 C
EXPERIMENTAL
and extracted with ethylacetate. The organic layer was
dried over MgSO4 and concentrated. Purification with flash
chromatography (ethylacetate/hexane, 5 : 1) yielded a white
solid that was recrystallized from an ethylacetate/hexane
mixture to afford pure (2) (3.1 g, 70%). 1H NMR (CDCl3)
υ 7.42 (d, J D 9 Hz, 3H), 7.23 (m, aromatic protons), 6.95
(d, J D 8.1 Hz, 3H), 4.91 (d, J D 11.4 Hz, 1H), 4.72 (m,2H),
4.69 (d, J D 3.6 Hz, 1H), 4.58 (d, J D 1.5 Hz, 2H), 4.55 (d,
J D 1.8 Hz, 1H), 4.36 (q, J D 13.8 Hz, 2H), 3.94 (d, J D 2.4 Hz,
1H), 3.84 (t, J D 9.3 Hz, 1H), 3.53 (m, 4H). 13C NMR (CDCl3)
υ 127–132 (aromatic carbons), 88.47, 84.62, 77.75, 77.65, 76.06,
74.84, 73.99, 73.96, 73.14, 69.18, 21.5. HRMS: C41O5NSH46
[M C NH4]C calcd 664.3091, found 664.3120.
Mass spectrometry
All ESI FT-ICR experiments were carried out using Bruker
BioAPEX III 47e FT-ICR spectrometer (Bruker Analytical
Systems, Inc., Billerica, MA) equipped with a 4.7 T super-
conducting magnet, an external source (Apollo ESI Source),
and an infinity analyzer cell. The samples were dissolved in
an ammonium acetate solution (10.0 mg ammonium acetate
in 100 ml CH3OH) and introduced into the ESI source at
a flow rate of 0.3 ml hꢀ1. Ions were detected using the
broadband detection mode covering a mass range from
50 to 3000 amu. Typically, 8 individual transients were
accumulated to improve the signal-to-noise ratio. For CID
experiments, precursor ions were isolated using swept fre-
quency ejection pulses of 250–500 µs duration to eject all
other ions. A pulsed valve introduced the argon collision
gas prior to ion activation. With the pulsed valve open for
30 ms, a peak pressure of ¾8 exp-7 mbar was obtained.
The precursor ions were excited using a variable amplitude
off-resonance excitation pulse.
p-methylphenyl 2, 3, 4, 6-tetra-O-methyl-1-thio-
ˇ-D-galactopyranoside (3)
p-Methylphenyl-1-thio-ˇ-D-galactopyranoside
(0.3 g,
1.057 mmol, see preparation of (1)) was dissolved in 40 ml of
°
freshly distilled DMF and cooled to ꢀ10 C under nitrogen
atmosphere. NaH (0.31 g, 12.5 mmol) was then added and
the mixture was allowed to warm up to room temperature
and was further stirred for 10 min. Methyl iodide (1.6 ml,
25 mmol) was added dropwise and the reaction mixture
Materials
p-methylphenyl 2, 3, 4, 6-tetra-O-acetyl-1-thio-
ˇ-D-galactopyranoside (1)
°
was further stirred for 5 h, cooled to 0 C, quenched with
Galactose pentaacetate (33.1 g, 0.084 mol) and p-thiocresol
water and extracted with ethylacetate. The organic layer was
dried over MgSO4 and concentrated. Flash chromatography
(ethylacetate/hexane, 3 : 1) yielded a white solid that was
recrystallized from an ethylacetate/hexane mixture to afford
pure (3) (0.29 g, 60%).1H NMR (CDCl3) υ 7.40 (d, J D 8.1 Hz,
2H), 7.04 (d, J D 7.8 Hz, 2H), 4.44 (d, J D 7.8 Hz, 1H), 3.72 (d,
J D 3 Hz, 1H), 3.6 (m, 4H), 3.58 (s, 3H) 3.56 (s, 3H), 3.52 (s,
3H), 3.34 (s, 3H), 3.14 (dd, J D 9, 3 Hz, 1H). 13C NMR (CDCl3)
υ 132.7, 129.8, 88.5, 86.1, 79.5, 77.75, 75.16, 70.85, 61.59, 61.47,
59.47, 58.49, 21.4. HRMS: C17O5NSH30 [M C NH4]C calcd
360.1839, found 360.1847.
(5.8 g, 0.13 mol) were dissolved in CH2Cl2, under nitrogen, at
°
0 C. Boron trifluoride diethyl etherate (14 ml, 0.11 mol) was
added and the reaction mixture was stirred for 18 h at room
temperature. After dilution with CH2Cl2, the organic layer
was washed with saturated NaHCO3 and brine. The organic
layer was dried over Na2SO4, the solvent was removed and
the residue was purified by column chromatography on
silica gel (hexane/ethylacetate 3 : 1) to afford (8) in 57% yield
(20.2 g). 1H NMR (CDCl3) υ 7.37 (dd, J D 8.1 Hz, 2H), 7.08 (dd,
J D 7.8 Hz, 2H), 5.37 (d, J D 5.7 Hz, 1H), 5.16 (t, J D 5.4 Hz,
1H), 4.98 (dd, J D 9.9, 3.3 Hz, 1H), 4.6 (d, J D 9.9 Hz, 1H),
4.05 (m, 2H), 3.86 (t, J D 6.9 Hz, 1H), 2.3 (s,3H), 2.09 (s,3H),
2.07 (s,3H), 2.07 (s,3H), 1.94 (s,3H). 13C NMR (CDCl3) υ 170.7,
170.5, 170.4, 169.8, 133.4, 133.2, 130.1, 129.9, 21.5, 21.2, 21.0,
20.93, 20.89. HRMS: C21O9NSH30 [M C NH4]C calcd 472.1636,
found 472.1653.
p-methylphenyl 2, 3, 4, 6-tetra-O-trimethyl silyl-1-thio-
ˇ-D-glucopyranoside (4)
To a solution of p-methylphenyl-1-thio-ˇ-D-glucopyranoside
(0.13 g, 0.045 mmol) (synthesized according to the described
above procedure for the preparation of p-methylphenyl-1-
°
thio-ˇ-D-galactopyranoside) in dry pyridine at 0 C, TMSCl
p-methylphenyl 2, 3, 4, 6-tetra-O-benzyl-1-thio-
ˇ-D-galactopyranoside (2)
Galactopyranoside (1) (10.6 g, 0.02409 mol) was dissolved in
(0.25 ml, 1.9 mmol) and catalytic amount of 4-DMAP were
added. The reaction was carried out at room temperature
while monitored by TLC (ethylacetate/hexane, 6 : 4) and
was quenched by addition of ethylacetate. The mixture was
washed with saturated NaHCO3 and brine, after which the
organic layer was dried over MgSO4 to allow (5) (0.17 g,
63%). 1H NMR (CDCl3) υ 7.48 (d, J D 4.8, 2H), 7.22 (d, J D 4.8,
2H), 4.75 (m,1H), 4.33 (m,1H), 4.00 (m,1H), 3.75 (m,1H), 3.60
(m, 2H), 3.35 (m, 1H), 0.37 (s, 3H), 0.30 (s, 3H), 0.28 (s,
3H), 0.193 (s, 3H). HRMS: C25O5NSSiH54 [M C NH4]C calcd
592.2794, found 592.2829.
a 64-ml mixture of MeOH/ethylacetate (1 : 1). After cooling
°
to 0 C, a solution of sodium methoxide in methanol (4.8 ml,
0.024 mol) was added and the reaction was allowed to
warm to room temperature. The reaction was quenched
by addition of Amberlite HR-120 (HC) in methanol, then
filtrated and concentrated. The residue was purified by
flash chromatography on silica gel (CH2Cl2/MeOH 5 : 1) to
afford p-methylphenyl-1-thio-ˇ-D-galactopyranoside (5.5 g,
80%). p-Methylphenyl 1-thio-ˇ-D-galactopyranoside (2.0 g,
Copyright  2005 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2005; 40: 1055–1063

Stability of oxocarbenium ions from glycosides 1057
was purified by column chromatography on silica gel (hex-
ane/ethylacetate 1 : 1) to allow (8) as an anomeric mixture.
1H NMR (CDCl3) υ 7.45 (m, 2H, aromatic protons), 7.40 (d,
J D 7.83 Hz, 4H), 7.27 (d, J D 7.83, 4H), 5.88(d, J D 4.7 Hz,
1H), 5.4 (t, J D 10 Hz, 1H), 5.18 (t, J D 9.39 Hz, 1H), 5.07 (t,
J D 6.27, 1H), 5.02 (m, 1H), 4.99 (dd, 5.73 Hz, 1H), 4.93 (t,
J D 9.39, 1H), 4.67 (d, J D 9.39 Hz, 1H), 4.53 (dd, J D 4.7,
1H), 4.17 (d, J D 3.77, 2H), 3.99 (d, J D 3.77, 2H), 3.69 (dd,
J D 4.52 Hz, 1H), 2.09 (s, 3H), 2.06 (s, 3H), 2.04 (s, 3H),
2.01 (s, 3H), 1.9 (s, 3H), 1.97 (s, 3H), 1.95 (s, 3H). CI-MS
([M-H]ꢀ) 439.0.
1, 2, 3, 4, 6-O-benzoyl-ˇ-D-glucopyranoside (5)
Benzoyl chloride (3.9 ml, 33.5 mmol) was added to a solution
of D-glucose (1.1 g, 6.1 mmol) in dry pyridine, at 0 C
°
followed by catalytic amount of 4-DMAP. The reaction was
stirred overnight at room temperature, then quenched by
addition of ethylacetate and washed with aqueous H2SO4,
saturated NaHCO3 and brine. The organic layer was dried
over MgSO4 and concentrated. The product was crystallized
from ethylacetate/hexane to allow (5) (2.8 g, 68%). 1H NMR
(CDCl3) υ 7.26–8.1 (m, aromatic protons), 6.82 (d, J D 3.6 Hz,
1H), 6.25 (t, J D 9.6 Hz, 1H), 5.80 (t, J D 9.9 Hz, 1H), 5.63
(dd, J D 14.1, 3.6 Hz, 1H), 4.56 (m, 2H), 4.42 (m, 1H). 13C
NMR (CDCl3) υ 166.7, 166.4, 165.7, 165.4, 164.9, 128.7–130.5
(aromatic carbons), 90.3, 70.76, 70.75, 70.71, 69.1, 62.7. HRMS:
C41H36O11N [M C NH4]C calcd 718.2283, found 718.2324.
p-methylphenyl 2, 3, 4, 6-tetra-O-acetyl-1-thio-
ˇ-D-glucopyranoside (9)
(9) was prepared from glucose pentaacetate in the same
manner as described for the preparation of (1). 1H NMR
(CDCl3) υ 7.37 (dd, J D 4.31 Hz, 2H), 7.1 (dd, J D 4.66 Hz,
2H), 5.17 (dd, J D 9.14 Hz, 1H), 5.0 (t, J D 9.78 Hz, 1H), 4.91
(t, J D 9.25, 1H), 4.61 (d, J D 9.9 Hz, 1H), 4.17 (d, 2H), 3.68
(m, 1H), 2.3 (s, 3H), 2.07 (s, 3H), 2.06 (s, 3H), 1.99 (s, 3H), 1.96
(s, 3H). CI-MS ([M-H]ꢀ) 453.1.
p-nitrophenyl 2, 3, 4, 6-tetra-O-acetyl-1-thio-
ˇ-D-glucopyranoside (6)
To the solution of 2, 3, 4, 6-Tetra-O-acetyl-1-bromo-˛-
glucopyranoside (5 g, 12.17 mmol) and TBAHS (6.15 g,
18.3 mmol) in 50 ml of ethyl acetate, Na2CO31 M (50 ml) and
p-nitrothiophenol (3.2 g, 24.35 mmol) was added and the
reaction mixture was stirred overnight at room temperature.
After dilution with ethylacetate, the organic layer was
washed with saturated NaHCO3 and brine. The organic
layer was dried over Na2SO4, the solvent was removed
and the residue was purified by crystallization from
hexane/ethylacetate to afford (6) in 83% yield. 1H NMR
(CDCl3) υ 8.14 (d J D 8.64 Hz, 2H), 7.57 (d, J D 8.67 Hz, 2H),
5.26 (dd, J D 9.19 Hz, 1H), 5.08 (t, J D 4.83 Hz, 1H), 5.03 (t,
J D 4.59 Hz, 1H), 4.84 (d, J D 9.98 Hz, 1H), 4.21 (d, J D 5 Hz,
2H), 3.8 (m, 1H), 2.09 (s, 3H), 2.06 (s, 3H), 2.03 (s, 3H), 1.99 (s,
3H). CI-MS (MHC) 485.0.
p-methoxyphenyl 2, 3, 4, 6-tetra-O-acetyl-1-thio-
ˇ-D-glucopyranoside (10)
ˇ Glucose pentaacetate (5 g, 12.8 mmol) and p-metoxyben-
zenethiol (2.4 ml, 19.32 mmol) were dissolved in CH2Cl2,
°
under argon, at 0 C. TMS triflate was added and the reaction
mixture was stirred overnight at room temperature. After
dilution with CH2Cl2, the organic layer was washed with
saturated NaHCO3 and brine. The organic layer was dried
over Na2SO4, the solvent was removed and the residue
was purified by column chromatography on silica gel
(hexane/ethylacetate) to afford (10) in 37% yield (20.2 g). 1H
NMR (CDCl3) υ 7.42 (d, J D 8.85 Hz, 2H), 6.83 (d, J D 8.63 Hz,
2H), 5.2 (dd, J D 9.42 Hz, 1H), 4.97 (t, J D 9.87 Hz, 1H),
4.87 (t, J D 9.43, 3.3 Hz, 1H), 4.53 (d, J D 9.95 Hz, 1H),
4.17 (d, J D 3.58 Hz, 2H), 3.79 (s, 1H), 3.65 (m, 1H), 2.08
(s,3H), 2.06 (s,3H), 1.99 (s,3H), 1.96 (s,3H). CI-MS ([M-H]ꢀ)
469.0.
4-acetamidophenyl 2, 3, 4, 6-tetra-O-acetyl-1-thio-
ˇ-D-glucopyranoside (7)
A solution of (6) (1.95 g, 5.18 mmol) in ethyl alcohol (60 ml)
and SnCl2*H2O (6 g, 27 mmol) was refluxed for 3 h, then
it was cooled and NaHCO3 was added to adjust pH 7.
After dilution with ethylacetate, the organic layer was
washed with saturated NaHCO3 and brine. The organic
layer was dried over MgSO4, the solvent was removed
and the residue was purified by flash chromatography to
allow pure 4-aminophenyl 2, 3, 4, 6-Tetra-O-acetyl-1-thio-ˇ-
D-glucopyranoside, that was acetylated by acetic anhydride
in pyridine. Flash chromatography yielded pure (7). 1H NMR
(CDCl3) υ 7.3–7.44 (m, 4H), 5.18 (t, J D 9.15 Hz, 1H), 4.99 (t,
J D 9.66 Hz, 1H), 4.89 (t, J D 9.31 Hz, 1H), 4.17 (m, 2H), 3.66
(m, 1H), 2.17 (s, 3H), 2.07 (s, 3H), 2.02 (s, 3H), 1.99 (s, 3H),
1.96 (s, 3H), 1.64 (s, 1H). CI-MS (MHC) 496.1.
p-methylphenylsulfenyl 2, 3, 4, 6-tetra-O-benzyl-
ˇ-D-galactopyranoside (13)
Into a 25-ml round-bottomed flask was weighted 2.5 g
of Merck silica gel 60 that has been equilibrated with
°
atmosphere at 110 C for 72 h. The flask was stoppered
and the content was allowed to cool to room temperature
after which 0.5 ml of water was added and the absorbent
was tumbled on rotary evaporator at atmospheric pressure
until uniformly free-flowing. A solution of (2) (0.646 g,
1.0 mmol) in 5 ml CH2Cl2 was added with stirring followed
by addition of 462 mg of OXONE. The mixture was
stirred at room temperature for 18 h. The silica was
then removed by filtration and washed with ethylacetate.
Then the organic fraction was washed with 30 ml of
saturated, aqueous solution of FeSO4, dried over Na2SO4 and
concentrated. Further purification was achieved by column
chromatography on silica gel to allow a mixture of the
diastereomeric sulphoxides (13). 1H NMR (CDCl3) υ 7.32(
d, j D 8 Hz, 2H), 7.23 (m, 20H, aromatic protons), 7.23(m,
2H), 4.97 (d, J D 3.9 Hz, 1H), 4.92 (d, J D 10 Hz, 1H), 4.7
Tiophenyl 2, 3, 4, 6-tetra-O-acetyl-1-thio-
ˇ-D-glucopyranoside (8)
Glucose pentaacetate (5 g, 12.9 mmol) and thiophenol
(1.4 ml, 13.5 mmol) were dissolved in CH2Cl2, under argon,
°
at 0 C. TMS triflate (0.2 ml) was added and the reaction
mixture was stirred for 18 h at room temperature. After
dilution with CH2Cl2, the organic layer was washed with
saturated NaHCO3 and brine. The organic layer was dried
over Na2SO4, the solvent was removed and the residue
Copyright  2005 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2005; 40: 1055–1063

1058
C. Denekamp and Y. Sandlers
[M+NH4]+
472.2
O
STol
OAc
AcO
AcO
331.1
BnO
OAc
1
169.0
221.1
[M+NH4]+
664.4
O
STol
OBn
BnO
539.3
181.1
647.4
OBn
2
91.1
[M+NH4]+
360.2
O
STol
OMe
MeO
MeO
311.2
343.2
111.0
OMe
3
100 150 200 250 300 350 400 450 500 550 600 650
m/z
Figure 1. CID spectra that were recorded for ESI-generated [M C NH4]C ions of galactose based sugars (1), (2) and (3) that contain
an S-Tol group at the anomeric position and acetyl (Ac), benzyl (Bn) and methyl (Me) protecting groups, respectively.
(d, 3.6 Hz, 2H), 4.57 (d, J D 11.7 Hz, 1H), 4.39 (t, J D 9.6 Hz,
1H), 4.11 (s,2H), 3.8 (m, 2H)3.59(m, 2H), 3.38 (m, 2H), 2.26
(s, 3H).
13C NMR (CDCl3) υ 140.8, 137.8, 137.5, 137.4, 137.4,
129–124.9 (aromatic carbons), 93.59, 8.79, 78.75, 75.47,
73.92, 73.4, 72.99, 72.76, 72.18, 68.72. HRMS: C41 H42O6Na
[M C Na]C calcd 685.2594, found 685.2594.
acetylated (1), benzylated (2) or methylated (3). The thiotolyl
group is frequently used as a leaving group in glycosylation21
since it is readily activated with commonly used promoters
such as NIS-TfOH or DMTST, and it was therefore chosen as
a leaving group for this study.
Figure 1 presents CID spectra of [M C NH4]C ions
that were measured for (1)–(3). The [M C NH4]C ion
(m/z 472) of the acetylated thiotolyl galactose (1) gives
rise to an abundant oxocarbenium ion of m/z 331 (relative
abundance (RA) D 52%). This ion further dissociates to
afford consecutive products at lower masses. However, the
CID of the [M C NH4]C ion of methoxy derivatized (3) gives
rise to very low abundance of the oxocarbenium ion of
m/z 219 (RA D 0.05%), while the analogous oxocarbenium
ion of m/z 523 is totally absent from the CID spectrum of
benzyloxy protected (2). Instead, fragmentation reactions
that occur involve the losses of ammonia and benzyl alcohol
or methanol with retention of the thiotolyl group in (2) and
(3), respectively.
p-methylphenylsulfenyl 2, 3, 4, 6-tetra-O-acetyl-
ˇ-D-galactopyranoside (14)
A diastereomeric mixture of (14) was prepared in the same
manner as (13) from p-methylphenyl 2, 3, 4, 6-tetra-O-
acetyl-1-thio-ˇ-D-galopyranoside. Further purification was
achieved by column chromatography on silica gel to allow a
mixture of the diastereomeric sulfoxides. 1H NMR (CDCl3) υ
7.51 (m, 2H), 7.28 (d, J D 7.8, 2H), 5.42 (m, 1H), 5.31(m, 1H),
5.01 (m,1H), 4.27 (d, J D 9.9 Hz, 0.5H), 4.19 (m, 0.5H), 4.01 (d,
J D 7.2 Hz, 0.5H), 3.98 (m, 1H), 3.95 (d, J D 6.3 Hz, 0.5H), 3.8
(m, 1H), 2.39 (s,3H), 1.98 (s, 3H), 1.97 (s, 3H), 1.94 (s.3H), 1.91
(s,3H). ESI-MS (MHC) 471.2.
We further examined the fragmentation behavior of the
[M C NH4]C ion of per-silylated glucose-based (4) (Fig. 2).
The RA of the oxocarbenium ion of m/z 451 in the CID
spectrum of (4) is 11%, much higher than the abundance of
oxocarbenium ions in the CID spectra of ethereal galactose-
based (2) and (3) but significantly lower than what is seen in
the CID spectrum of (1), in which the RA of the oxocarbenium
ion of m/z 331 is higher than 60% under the same energetic
conditions. Apparently, silyl ethers enhance the formation of
the oxocarbenium cation and stabilize it more than ethers but
far less than acetyl groups. On the other hand, prominent loss
of ammonia and benzoic acid is observed in the CID spectrum
of per-benzyloxy (5) (Fig. 2). The resulting ion of m/z 579 (an
oxocarbenium ion, RA D 45%) is relatively stable and an
Theoretical methods
Calculations were carried out using Gaussian 98 package of
programs.36 The clusters and molecules under study were
optimized at B3LYP/6-31G* hybride density functional level
of theory and the optimized structures were analyzed using
analytical frequencies calculations. Energies are presented
with zero point energy correction.
RESULTS
Protecting group effect
Galactose-based (1)–(3) (Fig. 1) possess a thiotolyl leaving
group at the anomeric position while all other positions are
Copyright  2005 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2005; 40: 1055–1063

Stability of oxocarbenium ions from glycosides 1059
[M+NH4]+
592.4
O
STol
361.2
TMSO
TMSO
OTMS
OTMS
4
451.3
575.4
105.0
[M+NH4]+
718.4
O
OBz
OBz
BzO
BzO
OBz
5
231.1
579.2
100 150 200 250 300 350 400 450 500 550 600 650 700 750
m/z
Figure 2. CID spectra that were recorded for ESI-generated [M C NH4]C ions of glucose based (4) and (5) that contain trimethylsilyl
(TMS) and benzoyl (Bz) protecting groups, respectively.
[M+NH4]+
472.1
O
S
AcO
AcO
OAc
OAc
9
169.0
211.0
109.0
275.0
271.0
331.1
50
100
150
200
250
300
m/z
350
400
450
500
Figure 3. CID spectrum measured for the [M C NH4]C ion of glucose-based 9.
MS3 measurement shows that it does not readily dissociate
(not shown).
example, Fig. 3 presents the CID spectrum of ammonium
cationized glucose-based (9) that contains a thiotolyl group
at the anomeric carbon. Upon collisions, this [M C NH4]C ion
forms an [M C NH4 –NH3 –3HOAc]C ion of m/z 275 with a
relative abundance of 34%.
CID spectra of ammonium cationized (6–10) show
similar oxocarbenium to [M C NH4]C ion ratios indicating
no apparent effect of the nature of the leaving group on
the cleavage at the anomeric position (Scheme 1). However,
the small intensities of [M C NH4 –NH3 –180]C ions that were
measured depend on the substituent at the thioaryl group, as
summarized in Table 1. It is worth adding that an analogous
ion is absent from the CID spectrum of the isomeric galactose
per-acetate (1).
Leaving group effect
CID spectra were also measured for a series of glucopyra-
nosides that possess a variety of thioaryl leaving groups.
Ammonium cationized (6–10) undergo fragmentation in
two different ways, upon collisions. The major fragmenta-
tion pathway involves the loss of ammonia and substituted
thiophenol to form an oxocarbenium ion of m/z 331. This
ion further dissociates to afford product ions at m/z 271,
211, 169 and 109 by consecutive losses of acetic acid or
ketene. Another, less favored route of reaction is the loss
of ammonia and three acetic acid moieties in one step,
giving rise to an ion that retains the thioaryl group. For
Copyright  2005 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2005; 40: 1055–1063

1060
C. Denekamp and Y. Sandlers
The relative reactivity of stereoisomeric acetylated ˛- and
ˇ- glycosides towards the formation of oxocarbenium ions
at m/z 331 was examined by us previously.37 Comparison
of CID spectra of their [M C NH4]C ions shows that trans
diacetoxy groups at C1-C2 are more reactive, towards the
loss of acetic acid, as the group at C2 enhances the reaction
by neighboring group participation. It was also found that the
presence of an axial substituent at C4 (i.e. galactose) results in
favored formation of oxocarbenium ions. Moreover, ˇ-xylose
per-acetate and ˛-fucose per-acetate that lack an acetyl group
at C6 exhibit reactivities that are comparable with ˇ-glucose
per-acetate with trans diacetoxy groups at C1-C2.
Protecting groups
According to the presented experimental results, the pro-
tecting groups at the different positions show the strongest
influence on the stability of the oxocarbenium intermediate.
Moreover, in the case of (2) and (3) that contain methyl and
benzyl groups, an oxocarbenium cation is not generated in
the gas phase at all. This could indicate that glycosylation
reactions with donors that contain ethereal protecting groups
proceed via a mechanism that does not involve the formation
of an oxocarbenium intermediate.
Alternatively, it can be argued that the ammonium
group in [M C NH4]C ions of (2) and (3) is not attached
at the anomeric position, therefore oxocarbenium cations
are absent from the CID spectra of (2) and (3) even if their
formation is favorable. Nevertheless, the basicity of acetyloxy
groups is higher than that of alkoxy groups and therefore
the behavior of (1) indicates that even if initial cationization
is not at the anomeric position, the ammonium cation (or
proton) can travel within the [M C NH4]C attachment ion. In
order to reassure this assumption a CID spectrum was also
measured for ammonium cationized unprotected methyl
glucopyranoside (11), in which the ammonium ion is likely
to interact with the methoxy group at the anomeric position
and the heterocyclic oxygen atom.
Scheme 1. Competitive routes of fragmentation: formation of
oxocarbenium ions versus combined loss of three acetic acid
moieties with retention of the thioaryl group.
Table 1. Relative abundances of [M C NH4 –3HOAc–NH3]C
ions in the CID spectra of (6)–(10) and relative intensities of
these ions corrected according to the total ion intensity
Substituent
RA
(%)
(X)
[M C NH4 –3HOAc–NH3]C (%)
(6)
(7)
(8)
(9)
(10)
NO2
NHAc
H
Me
Ome
2.4
8.9
9.7
10
5.6
30
32
34
54
13
DISCUSSION
A stepwise glycosylation process involves the formation of
an oxocarbenium intermediate, and the formation of this
intermediate may be the rate-determining step (Scheme 2). If
so, the stability of this intermediate as well as the rate of its
formation may determine the activity of the so-called donor
counterpart in this reaction.
The rate of formation of isomeric oxocarbenium ions
depends on stereochemical effects, dominated by the con-
figuration of a specific isomer, the nature of the protecting
groups and the leaving group at the anomeric position. The
systems that were chosen for this study were designed in
order to gain insight into each of these effects.
Nevertheless, the CID of the [M C NH4]C ion of (11)
(not shown) does not reveal the formation of an oxo-
carbenium ion of m/z 163. Instead, consecutive losses of
Scheme 2. Oxocarbenium intermediates formed in glycosylation reactions of galactose based glycosides. Glucose based systems
differ by the configuration at C4.
Copyright  2005 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2005; 40: 1055–1063

Stability of oxocarbenium ions from glycosides 1061
water molecules occur. This indicates that despite the
higher topical basicity at the anomeric position, pro-
ton migration and water loss from other sites takes
place.
An oxocarbenium ion is present in the CID spectra
of the [M C NH4]C ions of per-silylated glucose (4) (11%)
and per-benzyloxy (5) (45%) as seen in Fig. 2. Although
the RA of the analogous ion in the CID spectrum of per-
acetylated (1) is higher (¾60%, Fig. 1), MS3 experiments
reveal that the per-benzylated oxonium ion is more stable
than the per-acetylated analog upon collisional activation. It
is therefore concluded that in the case of ammonium cation-
ized glycosides, elimination of the group at the anomeric
position occurs when the resulting oxocarbenium cation
is stabilized by the substituents, with the following order
of stability: Bz > Ac > ꢀCH3ꢁ3Si > R, where R is an
alkyl group. When the formation of an oxocarbenium is
disfavored competing reactions, like alcohol elimination,
occur.
Scheme 3. Fraction of oxocarbenium ions in the sum intensity
of fragment ions present in the CID spectra of MHC ions of (13)
and 14, as a measure for stabilization by acetyl versus benzyl
groups.
Mobility of the cation (proton ammonium etc.) between
different basic sites can be avoided if the anomeric position
is significantly more basic. Thioaryls are used in solution
because they are selectively activated by specific reagents.
Since in the gas phase the promoter is an acid, it is desirable
to have a basic group at the anomeric position. For example,
sulphoxides are better candidates for the formation of MHC
ions than sulphides because of their higher proton affinity.
This is supported by the calculated topical proton affinity at
the two basic sites of the model system (12). DFT calculations
(B3LYP/6-31G* with ZPE correction) show that protonation
of the sulphoxide oxygen atom is favored by 29.0 kcal molꢀ1
when compared with protonation of the sulphide sulfur
atom. Indeed, we find that (13), that contains a sulphoxide
group at the anomeric position, gives rise to MHC ions upon
addition of ammonium acetate. This MHC ion is isolated in
the gas phase and dissociates to afford the corresponding
oxocarbenium ion of m/z 523 upon collisions, despite the
benzyl protecting groups. Nevertheless, a comparison of the
CID spectra that were measured for MHC ions of (13) and
(14) (Scheme 3) shows that the amount of oxocarbenium
ion (relative to the sum of fragment ions) in the CID of
the MHC ion of (14), that contains acetyloxy protecting
groups, is much higher than the amount of oxocarbenium
ion in the CID of the MHC ion of (13) (90% versus 26%,
respectively).
group interaction with acetyl groups at C2 and C4 of
glycosides. We could show that the strongest effect is
caused by the group at C2. Other experimental evidence
indicate that an acetyl group at C6 does not interact with the
cation. Here, in order to support the order of stabilization
by protecting groups at C2, as deduced from the relative
reactivity towards the formation of an oxocarbenium ion
upon CID, relative hydride affinities (HA) of model cations
(15)–(17) that contain an acetyl, methyl or trimethylsilyl
group at C2 were calculated. It is assumed that a cation that
shows a higher hydride affinity value is less stable. DFT
calculations, at the B3LYP/6–31GŁ level of theory (with ZPE
correction), indicate that acetoxy substituted (15) has the
smallest HA value while the relative HAs of model cations
(16) and (17) are 13.2 and 16.5 kcal molꢀ1 higher, respectively.
It is therefore shown that despite the electron-withdrawing
character of acyl groups they stabilize the oxocarbenium ion
efficiently through ꢂ overlap and the favored formation of a
5-membered ring.
Examination of the HOMO orbital of cation (15) (Fig. 4)
also leads to the conclusion that there is an overlap between
the nonbonding p-orbital of the acetyl oxygen atom and a
sp2 orbital of the anomeric carbon. On the other hand, no
analogous ꢂ interaction is identified in the HOMO orbitals
of model cations (16) and (17).
Considering the electronegativity of a methyl versus
trimethylsilyl groups, it is reasonable that model cation (17)
is less stable than (16). Indeed, the sum of Mulliken charges
Theoretical support for the effect of protecting
groups
In previous work, we calculated the stabilization of the
developing cation at the anomeric position by neighboring
Copyright  2005 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2005; 40: 1055–1063

1062
C. Denekamp and Y. Sandlers
Figure 4. Calculated HOMO orbitals (B3LYP/6-31GŁ) of model oxocarbenium cations (15), (16) and (17). A clear overlap between the
nonbonding p-orbital of the acetyl oxygen atom and a sp2 orbital of the anomeric carbon of (15) is observed. The electron density
around the methyl group in (17) is higher than the electron density on the trimethylsilyl group in (16).
on the trimethylsilyl group in (16) is 0.78 in comparison with
0.19 on the methyl group in (17). Also the HOMO orbitals
that are shown in Fig. 4 illustrate higher electron density
around the methyl group in (17) in comparison with the
electron density around the trimethylsilyl group in (16).
Our findings indicate that per-benzoylated protected
glucose (5) (Fig. 2) affords a particularly stable oxocarbenium
ion of m/z 579, more stable than the acetate analog. It is
known that benzoyl thio-donors are less reactive towards
glycosylation reactions than their acetate thio-analogs21. It
might be explained by the formation of a more stable
oxocarbenium ion in the presence of benzoyl protecting
groups, an intermediate that is too stable to react, possibly
an ion pair that collapses back to the starting material. In this
case, the rate-determining step is no longer the formation of
a cationic intermediate but the successive reaction with the
acceptor.
a mild leaving group effect when comparing a series of
thioaryl leaving groups.
Acknowledgement
This work was supported by the Israel Science Foundation.
REFERENCES
1. Dwek RA. Toward understanding the function of sugars. Chem.
Rev. 1996; 96: 683.
2. Jensen KJ. O-Glycosylations under neutral or basic conditions. J.
Chem. Soc., Perkin Trans. 1. 2002; 20: 2219.
3. Bennet AJ, Kitos TE. Mechanisms of glycopyranosyl and 5-
thioglycopyranosyl transfer reactions in solution. J. Chem. Soc.,
Perkin Trans. 2. 2002; 7: 1207.
4. Demchenko AV. Stereoselective chemical 1,2-cis O-glycosyla-
tion: from ‘sugarray’ to modern techniques of the 21st century.
Synlett. 2003; 9: 1225.
5. Zhu J, Bennet AJ. Hydrolysis of (2-deoxy-˛-D-glucopyranosyl)
pyridinium salts: the 2-deoxyglucosyl oxocarbenium is not
solvent-equilibrated in water. J. Am. Chem. Soc. 1998; 120: 3887.
6. Indurugalla D, Bennet AJ. A kinetic isotope effect study on the
hydrolysis reactions of methyl xylopyranosides and methyl
5-thioxylopyranosides: oxygen versus sulfur stabilization of
carbenium ions. J. Am. Chem. Soc. 2001; 123: 10 889.
7. Zhu J, Bennet AJ. Solvolyses of ˛-2-deoxy and ˇ-D-
glucopyranosyl 40-bromoisoquinolinium tetrafluoroborates. J.
Org. Chem. 2000; 65: 4423.
We also find that the effect of the leaving group on the rate
of formation of an oxocarbenium ion is small. This does not
mean that there is no leaving group effect in glycosylation. It
only suggests that if glycosylation proceeds via the formation
of an oxocarbenium intermediate it should be less sensitive
to the leaving group.
8. Stubbs JM, Marx D. Glycosidic bond formation in aqueous
solution: on the oxocarbenium intermediate. J. Am. Chem. Soc.
2003; 125: 10 960.
CONCLUSIONS
9. Zhang Y, Bommuswamy J, Sinnott ML. Kinetic isotope effect
study of transition states for the hydrolyses of ˛- and ˇ-
glucopyranosyl fluorides. J. Am. Chem. Soc. 1994; 116: 7557.
10. Huang X, Surry C, Hiebert T, Bennet AJ. Hydrolysis of (2-deoxy
ˇ-D-glucopyranosyl) pyridinium salts. J. Am. Chem. Soc. 1995;
117: 10 614.
11. Sinnott ML, Jencks WP. Solvolysis of D-glucopyranosyl
derivatives in mixtures of ethanol and 2,2,2-trifluoroethanol.
J. Am. Chem. Soc. 1980; 102: 2026.
12. Banait NS, Jencks WP. Reactions of anionic nucleophiles with
˛-D-glucopyranosyl fluoride in aqueous solution through a
concerted, ANDN (SN2) mechanism. J. Am. Chem. Soc. 1991;
113: 7951.
13. Nukada T, Berces A, Zgierski MZ, Whitfield DM. Exploring
the mechanism of neighboring group assisted glycosylation
reactions. J. Am. Chem. Soc. 1998; 120: 13 291.
Our experiments show that protecting groups play a
significant role in the generation and stabilization of
oxocarbenium intermediates. It is clear that oxocarbeniums
are not generated from [M C NH4]C ions in the presence
of alkyl protecting groups instead of carboxy groups at
proper positions. Adequate stabilization is achieved in the
present of trimethylsilyl groups. It is concluded that the
order of stabilization by protecting groups at C2 and/or
C4 are Bz > Ac > ꢀCH3ꢁ3Si > R, where R is an alkyl
group. These results point out that efficient glycosylation
reactions that occur with benzylated or methylated donors
might take place without the formation of an oxocarbenium
intermediate. Other mechanisms that can be considered are
SN2- like or ring opening (mutarotation). Moreover, the
formation of some stabilized oxocarbenium ions (e.g. 5) is
not necessarily a rate-determining step in this reaction. Thus,
protecting groups should be chosen with care in order to
allow the formation of a reactive intermediate. We find only
14. Nukada T, Berces A, Whitfield DM. Can the stereochemical
outcome of glycosylation reactions be controlled by the
conformational preferences of the glycosyl donor? Carbohydr.
Res. 2002; 337: 765.
15. Cura P, Aloui M, Kartha KPR, Field RA. Iodine and its
interhalogen compound: versatile reagents in carbohydrate
Copyright  2005 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2005; 40: 1055–1063

Stability of oxocarbenium ions from glycosides 1063
chemistry XII. Tuning promoter reactivity for thioglycoside
activation. Synlett. 2000; 9: 1279.
one-pot glycosylation using seven independent building blocks.
Org. Lett. 2002; 24: 4213.
16. Douglas NL, Ley SV, Lu¨cking U, Warriner SL. Tuning glycoside
reactivity: new tool for efficient olygosaccharide synthesis. J.
Chem. Soc., Perkin Trans. 1 1998; 1: 51.
17. Geurtsen R, Holmes DS, Boons G-J. Chemoselective glycosyla-
tions. 2. Differences in size of anomeric leaving groups can be
exploited in chemoselective glycosylations. J. Org. Chem. 1997;
62: 8145.
18. Roy R, Andersson FO, Letellier M. Active and latent thioglycosyl
donors in oligosaccharide synthesis. Application to the synthesis
of ˛-sialosides. Tet. Lett. 1992; 33: 6053.
28. Raghavan S, Kahne D. A one step synthesis of the ciclamycin
trisaccharide. J. Am. Chem. Soc. 1993; 115: 1580.
29. Domon B, Mu¨ller DR, Richter WJ. Identification of interglyco-
sidic linkages and sugar constituents in disaccharide subunits
of larger glycosides by tandem mass spectrometry. Org. Mass
Spectrom. 1989; 24: 357.
30. Vouros P, Mu¨ller DR, Richter WJ. Low-energy collision-induced
dissociation of B1-type sugar ions formed from peracetylated
methyl pentosides and methyl 6-deoxyhexosides. J. Mass
Spectrom. 1999; 34: 346.
19. Lahmann M, Oscarson S. Investigation of the reactivity
difference between thioglycoside donors with variant aglycon
parts. Can. J. Chem. 2002; 80: 889.
31. Kulkarni PS, Kulkarni SY, Pansare VS. Electron-impact and
chemical-ionization mass spectra of some tetraacetylated
anomeric glycosides. Org. Mass Spectrom. 1986; 21: 23.
32. Biemann K, De Jongh DC, Schnoes HK. Application of mass
spectrometry to structure problems. XIII. 1Acetates of pentoses
and hexoses. J. Am. Chem. Soc. 1963; 85: 1763.
33. van der Kerk SM, Blok-Tip L, van der Kerk-van Hoof A,
Heerma W, Haverkamp J. Differences in fragmentation behavior
between ˛- and ˇ-linked derivatized xylobiosides: explanation
in terms of sigma conjugation. Int. J. Mass Spectrom. 1994; 134:
41.
20. Bols M, Liang X, Jensen HH. Equatorial contra axial polar
substituents. The relation of
a
chemical reaction to
stereochemical substituent constants. J. Org. Chem. 2002; 67:
8970.
21. Zhang Z, Ollmann IR, Ye XS, Wischnat R, Baasov T, Wong C-H.
Programmable one-pot oligosaccharide synthesis. J. Am. Chem.
Soc. 1999; 121: 734.
22. Miljkovic M, Yeagley D, Deslongchamps P, Dory Y. Experimen-
tal and theoretical evidence of through-space electrostatic stabi-
lization of the incipient oxocarbenium ion by an axially oriented
electronegative substituent during glycopyranoside acetolysis.
J. Org. Chem. 1997; 62: 7597.
34. Guevremont R, Pathak VP, Wright JLC. Unimolecular and
collision-induced dissociation reactions of peracetylated methyl
glucopyranoside. Org. Mass Spectrom. 1990; 25: 101.
35. Cuyckens F, Shahat AA, Pieters L, Claeys M. Direct stereochem-
ical assignment of hexose and pentose residues in flavonoid
O-glycosides by fast atom bombardment and electrospray ion-
ization mass spectrometry. J. Mass Spectrom. 2002; 37: 1272.
36. Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA,
Cheeseman JR, Zakrzewski VG, Montgomery JA Jr., Strat-
mann RE, Burant JC, Dapprich S, Millam JM, Daniels AD,
Kudin KN, Strain MC, Farkas O, Tomasi J, Barone V, Cossi M,
Cammi R, Mennucci B, Pomelli C, Adamo C, Clifford S, Ochter-
ski J, Petersson GA, Ayala PY, Cui Q, Morokuma K, Malick DK,
Rabuck AD, Raghavachari K, Foresman JB, Cioslowski J,
Ortiz JV, Baboul AG, Stefanov BB, Liu G, Liashenko A,
Piskorz P, Komaromi I, Gomperts R, Martin RL, Fox DJ, Keith T,
Al-Laham MA, Peng CY, Nanayakkara A, Challacombe M,
Gill PMW, Johnson B, Chen W, Wong MW, Andres JL, Gonza-
lez C, Head-Gordon M, Replogle ES, Pople JA. Gaussian: Pitts-
burgh, 1998.
23. Crich D, dela Mora M, Vinod AU. Influence of the 4,6-
O-benzylidene, 4,6-O-phenylboronate, and 4,6-O-polystyryl-
boronate
protecting groups on the stereochemical
outcome of thioglycoside-based glycosylations mediated
by 1-benzenesulfinyl piperidine/triflic anhydride and N-
iodosuccinimide/trimethylsilyl triflate. J. Org. Chem. 2003; 68:
8142.
24. Romero JAC, Tabaco SA, Woerpel KA. Stereochemical reversal
of nucleophilic substitution reactions depending upon
substituent: reactions of heteroatom-substituted six-membered-
ring oxocarbenium ions through pseudoaxial conformers. J. Am.
Chem. Soc. 2000; 122: 168.
25. Fraser-Reid B, Wu Z, Andrews CW, Skowronski E. Torsional
effects in glycoside reactivity: saccharide couplings mediated
by acetal protecting groups. J. Am. Chem. Soc. 1991; 113: 1434.
26. Namchuk MN, McCarter JD, Becalski A, Andrews T, With-
ers SG. The role of sugar substituents in glycoside hydrolysis. J.
Am. Chem. Soc. 2000; 122: 1270.
37. Denekamp C, Sandlers Y. Anomeric distinction and oxonium
ion formation in acetylated glycosides. J. Mass Spectrom. In
Press.
27. Tanaka H, Adachi M, Tsukamoto H, Ikeda T, Yanada H,
Takahashi T. Synthesis of di-branched heptasaccharide by
Copyright  2005 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2005; 40: 1055–1063