Sucrose tricarboxylate by sonocatalysed TEMPO-mediated oxidation

Article abstract of DOI:10.1016/S0008-6215(00)00046-X

Oxidation of sucrose by the NaOCl/TEMPO system provided sucrose tricarboxylate without the addition of sodium bromide as co-catalyst when high-frequency (500 kHz) ultrasound was applied, in contrast to very limited conversion without sonication. In the pr

Full text of DOI:10.1016/S0008-6215(00)00046-X

www.elsevier.nl/locate/carres
Carbohydrate Research 326 (2000) 176–184
Sucrose tricarboxylate by sonocatalysed
TEMPO-mediated oxidation
Sandrine Lemoine, Ce´cile Thomazeau, David Joannard, Ste´phane Trombotto,
Ge´rard Descotes, Alain Bouchu, Yves Queneau *
Unite´ Mixte de Sucrochimie CNRS Be´ghin-Say (UMR 143), c/o Eridania Be´ghin-Say, C.E.I.,
27 boule6ard du 11 No6embre 1918, BP 2132, F-69603 Villeurbanne, France
Received 18 May 1999; received in revised form 17 December 1999; accepted 27 January 2000
Abstract
Oxidation of sucrose by the NaOCl/TEMPO system provided sucrose tricarboxylate without the addition of
sodium bromide as co-catalyst when high-frequency (500 kHz) ultrasound was applied, in contrast to very limited
conversion without sonication. In the presence of sodium bromide, sonication also caused acceleration of the
oxidation. The rate increase due to sonication of the oxidant system prior to sucrose addition suggests that
ultrasound acts at the level of the formation of the nitrosonium ion, the active oxidising species in the catalytic cycle.
© 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Sucrose; Oxidation; Ultrasound; TEMPO; Uronic acids
tem [15–20]. Its efficiency was demonstrated
1. Introduction
by Davis and Flitsch [21] and van Bekkum
and co-workers [22–24], either for partially
protected or unprotected monosaccharides, or
with oligo- or polysaccharides, as well as more
recently by Schnatbaum and Scha¨fer using
electrochemical regeneration of the TEMPO
reagent [25]. In the ‘classical’ method, sodium
bromide is needed to catalyse the formation of
the C9H18N+ꢁO ion, which is the active oxi-
dising species. The exact role of the halide
ions has been related to steric factors or disso-
ciation constants of ClOH or BrOH
[16,23,26]. Rychnovsky and co-workers re-
cently reported some mechanistic studies on
the role of added salts in the case of m-CPBA-
promoted oxidations using TEMPO [27] as
well as on the influence of the redox potential
of the nitroxyl radicals [28]. New clues on the
mechanism could also be obtained from our
recent work on the sonocatalysis of the reac-
Polyhydroxy–polycarboxylates are interest-
ing targets for carbohydrates as organic raw
materials, due to their possible applications as
cation-sequestering agents [1–3]. These poly-
carboxylates can be obtained by various meth-
ods, which can either involve cleavage of
carbonꢀcarbon bonds by glycol oxidation
[4,5], or by selective oxidation of the primary
alcohol functionality keeping intact the carbo-
hydrate backbone. In this latter case, the oxi-
dation of carbohydrates to uronic acids can be
achieved by heterogeneous catalysis over plat-
inum [6–11], electrocatalysis [12,13], or bio-
conversion [14]. An increasingly popular
method is the use of the NaOCl/TEMPO sys-
* Corresponding author. Tel.: +33-472-442989; fax: +33-
472-442991.
E-mail address: queneau@univ-lyon1.fr (Y. Queneau)
0008-6215/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 0 8 - 6 2 1 5 ( 0 0 ) 0 0 0 4 6 - X

S. Lemoine et al. / Carbohydrate Research 326 (2000) 176–184
177
(2,2,6,6-tetramethyl-1-piperidinyloxy, 0.4 mol
equivalents and 0.0065 molar equivalents, re-
spectively, per primary hydroxyl group),
maintained at pH 10.5 by addition of sodium
hydroxide (pH-stat) and at ca. 5 °C, the typi-
cal conditions optimised by van Bekkum and
co-workers [23] (Scheme 1).
Scheme 1.
Preparation of an authentic sample of sodium
or triethylammonium sucrose tricarboxy-
lates.—For identification and quantification
purposes, authentic samples of 1 and 5 were
prepared from a mixture of carboxylates, ob-
tained by the Pt/O2 method [6], which was
subjected to an esterification–saponification
sequence as shown in Scheme 2. Treatment
with methyl iodide in DMF afforded a com-
plicated mixture of products, from which the
trimethyl ester of the triacid 2 could be iso-
lated, and further transformed (Ac2O–
pyridine) into its penta-O-acetyl derivative 4,
which was fully characterised. Comparison of
tion [29] using methyl a- -glucopyranoside as
D
model substrate, showing that the reaction can
proceed without sodium bromide, as well as
from the results reported in a patent claiming
that upon heating, sodium bromide was not
essential [30]. When high-frequency ultra-
sound (500 kHz) was applied to the system,
we observed an acceleration of the process.
We now report our results in the bleach-
TEMPO oxidation of sucrose for which little
is available in the literature [23,31], with a
focus on the effect of ultrasonic waves [32–
34].
1
the H NMR spectrum with that of sucrose
octaacetate showed the disappearance of the
signals corresponding to the protons at posi-
tions 6, 1% and 6%. As observed in the case of
the dicarboxy analogues for which NMR data
are available [7], H-5 and H-5% are shifted
downfield, as well as H-4%. Also based on the
study of NMR spectra (especially l and J
2. Results and discussion
An aqueous solution of sucrose was treated
with sodium hypochlorite (2.2 equivalents), in
the presence of sodium bromide and TEMPO
Scheme 2.

178
S. Lemoine et al. / Carbohydrate Research 326 (2000) 176–184
larger (8.6–13.8 mL) than the stoichiometric
amount (7.7 mL), indicating the formation of
derivatives containing more than three car-
boxylic acid functions. In these cases, new
peaks in the carboxyl region as well as new
anomeric carbons (l 99.7 and 93.2) were ob-
values for H-4% of the dimethyl ester of the
6,1%-diacid 3 that was isolated as side product,
in comparison with the triester 2), it was
possible to establish the conformation of the
fructosyl moiety, with an inversion of the twist
envelope placing the glucosyl moiety in a
pseudo-equatorial orientation for the triester 2
with a pseudo-equatorial H-4% (Dl= +0.43
ppm, DJ3%,4%= −1.9 Hz), whereas it is in a
pseudo-axial orientation for the diester 3.
Careful alkaline treatment (triethylamine in
methanol–water) of the protected triester 4,
followed by co-evaporation with water, led to
the tris-(triethylammonium) salt of the triacid
5 free from any other salt. Alternatively, treat-
ment of the triester 2 with stoichiometric
amounts of sodium hydroxide led to the
sodium tricarboxylate 1. The structure of the
triethylammonium salt 5 was deduced from
2D carbon–proton NMR spectroscopy mea-
surements, allowing the full assignment of all
hydrogen and carbon atoms. Carbon atoms at
C-5ꢀC-5%, and C-6ꢀC-6% provide broad signals,
resulting probably from an exchange between
the salt and protons of the solvent. Similar
observations were made in the case of the
trisodium salt 1 obtained by semi-preparative
size-exclusion chromatography from a mixture
prepared by the Pt/O2 method. This fact was
not observed in the case of the crude reaction
samples, in which large amounts of salts (e.g.,
NaCl, NaBr) are present, displacing largely
the equilibrium towards the carboxylate spe-
cies, unlike in pure water.
Effects of ultrasonic wa6es.—The rate of
sodium hydroxide addition is directly related
to the oxidation reaction rate, because of the
production of one H+ per catalytic cycle at
the TEMPO regeneration step by sympropor-
tionation of NꢀOH and N+ꢁO. The oxidation
of sucrose could thus be analysed through the
determination of the rate of addition of base,
the volume of added base, the analytical yield
(Dionex chromatography), and the 13C NMR
spectrum of the crude products. Effects of
sucrose concentration, amount of sodium
hypochlorite, of sodium bromide, of TEMPO,
and the activation conditions on the reaction
outcome are reported in Table 1.
13
served in the C NMR spectrum correspond-
ing to a pentacarboxylate, which could be
isolated in trace amount upon multiple precip-
itation from water–methanol mixtures. The
largest change was for C-2%, which indicates
some cleavage of the C-3%ꢀC-4% bond. Further-
more, the signals corresponding to C-3% and
C-4% in the tricarboxysucrose salts are absent.
This is consistent with a larger sensitivity of
the secondary alcohol functions towards oxi-
dation in a furanosyl ring compared with a
pyranosyl one [22,23]. If sodium hypochlorite
is added stepwise, the side oxidations are less
favoured, leading to improved yields [3].
In the preliminary report, we mentioned the
possibility of performing the reaction without
sodium bromide [29]. We first checked this in
the simpler case of the oxidation of methyl
a-D
-glucopyranoside to sodium (methyl a- -
D
glucopyranosid)uronate. Without sonication,
the reaction was very sluggish, with only 21%
conversion after 3 h. Ultrasound appeared to
be necessary to complete the reaction, with
increased yields based on the conversion of
the glucoside, but with a slower reaction rate
compared with the reaction in the presence of
sodium bromide (Table 2).
The decrease of the reaction rate in the
presence of oxygen (in brackets in Table 2)
can be attributed to the trapping by O2 of
some of the species that might be involved in
the catalysis [35].
It was checked that no reaction occurred
under ultrasound either without NaOCl or
without TEMPO. This, along with the effect
of presonication of the oxidant system (vide
infra), might suggest that ultrasound promotes
the formation of other species able to partici-
pate in the oxidation cycle, increasing the rate
of formation of N+ꢁO cations, as could be the
reason for a more efficient reaction in the
presence of bromides compared with
chlorides.
Using an excess of sodium hypochlorite led
to lower yields. The volume of added base was
In the case of sucrose, as for methyl a-
D
-
glucopyranoside, the sonocatalysis permits a

S. Lemoine et al. / Carbohydrate Research 326 (2000) 176–184
179
Table 1
Reaction of sucrose at 592 °C using 0.4 equivalents of NaBr a
Sucrose concentration
(mM)
NaOCl
TEMPO
Conditions
Rate b
(mL of NaOH/min)
1 c (%)
(equivalents) a
(equivalents) a
d
40
40
40
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
2.2
2.2
2.2
2.2
2.2
2.2
2.2
2.2
2.2
2.2
2.2
2.2
2.2
3.2
3.2
3.2
0.0065
0.0065
0.0065
0.0065
0.0065
0.025
0.025
0.06
0.8
1.0
1.9
54
55
74
68
71
66
67
74
58
70
55
63
53
57
64
16
63
21 f
46 f
78
80
20 kHz
500 kHz
d
0.22
0.39
0.65
1.37
0.79
1.41
1.35
1.45
2.5
500 kHz
d
500 kHz
d
0.06
500 kHz
d
0.125
0.125
0.20
500 kHz
d
0.20
500 kHz
3.4
d
0.0065
0.0065
0.0065
0.0065
0.125
0.125
0.0065
0.0065
0.27
0.34
0.18
0.36
2.1
2.8
0.13
0.20
500 MHz
d,e
3.2
3.2
3.2
500 MHz e
d
500 MHz
d
2.2 Fraction g
2.2 Fraction g
500 MHz
a Per primary hydroxyl groups.
b Maximum slope of the base addition versus time.
c Analytical yield (ion exchange chromatography).
d Without sonication.
e Reaction stopped after the stoechiometric amount of base added.
f 175% base addition indicating overoxidation.
g NaOCl added in 10 portions.
Table 2
Oxidation of methyl a-
D
-glucopyranoside to sodium (methyl a-D-glucopyranosid)uronate: influence of the presence of sodium
bromide (reactions stopped after 3 h) a
Conditions
NaOH addition rate b
(mL/min)
Glucoside
Yield d (%)
Yield/conversion
conversion c (%)
no Br−
Br−
0.7
no Br−
Br−
97
no Br− Br− no Br−
Br−
76
Without sonication
21
19
74
90
(–)
0.1
0.3
(29)
69
84
(12)
59
82
(41)
86
98
20 kHz, 0.26 W/mL, 13 mm probe
500 kHz, 0.22 W/mL, 35 mm probe
0.9
1.9
98
96
74
63
77
64
(0.2)
(71)
(69)
(98)
a Same conditions as in Table 1 except for sodium bromide; in brackets are given the results for the reaction in the presence
of oxygen.
b Maximum slope of the base addition versus time.
c Measured by HPLC on NH2-grafted column.
d Analytical yields (anion exchange chromatography).
used. This is due to the relative rates of the
oxidation and to the decomposition of sodium
hypochlorite, potentially affected by ultra-
reaction in the absence of sodium bromide
(Fig. 1), although with a smaller conversion
rate when only 0.65 mol% of TEMPO was

180
S. Lemoine et al. / Carbohydrate Research 326 (2000) 176–184
yield of sodium sucrose tricarboxylate in less
than 2 h at 5 °C.
In a final set of experiments, we investigated
the effect of the presonication of the oxidant
mixture (30 min at 5 °C) in the absence of
sucrose. Compared with the classical reaction,
this procedure led to a much faster reaction:
using 0.65 mol% of TEMPO and in the pres-
ence of NaBr, reaction rates were 0.94 and
0.33 mL/min (for reactions with or without
continued sonication after sucrose addition,
respectively), compared with 0.39 and 0.22
mL/min for the corresponding reactions with-
out presonication (Table 1). The yield of the
reaction combining ultrasound before and
during the sucrose oxidation was increased to
86% compared with 71%, whereas the ‘preson-
icated’ reaction further conducted without
sonication, provided a 56% instead of 68%
yield. Using 6 mol% of TEMPO and in the
absence of NaBr, rates were also increased
(3.0 and 0.74 mL/min, respectively, compared
with 0.6 and 0.13 mL/min), showing an accel-
eration ratio larger than the regular classical
Fig. 1. Addition of base as a function of time for the
oxidation of sucrose (2.2 equivalents of NaOCl, 0.0065 equiv-
alents of TEMPO, no NaBr) with or without sonication (500
kHz).
sound [36], preventing the completion of the
reaction even when sodium hypochlorite was
added stepwise. Since excess of sodium
hypochlorite led to overoxidation, the alterna-
tive was to increase the rate of the reaction by
using larger amounts of TEMPO, but keeping
it well below the stoichiometric amount (Table
3). The optimised conditions in the absence of
sodium bromide are the use of 0.06 equiva-
lents of TEMPO, providing a 79% analytical
sonochemical one. These observations are
+
’
consistent with a faster NꢀO “N ꢁO oxida-
tion step mediated by ultrasound.
Table 3
Reaction of sucrose at 592 °C without NaBr
Sucrose concentration (mM)
NaOCl
TEMPO
Conditions
Rate b
(mL of NaOH/min)
1 c (%)
(equivalents) a
(equivalents) a
d
40
40
40
40
40
40
40
40
40
40
40
40
40
40
80
80
2.2
2.2
2.2
4.4
2.2
2.2
2.2
2.2
2.2
2.2
3.2
3.2
0.0065
0.0065
0.0065
0.0065
0.06
20 kHz
500 kHz
0.16
0.17
0.26
0.13
0.6
0.25
0.86
2.2
23
18
9
500 kHz
d
16
79
16
71
41
63
29
48
26
48
50
72
0.06
500 kHz
d
0.125
0.125
0.20
0.20
0.06
0.06
0.06
0.06
0.06
500 kHz
d
500 kHz
4.8
d
0.35
1.18
0.09
0.18
0.5
500 kHz
d
2.2 Fraction e
2.2 Fraction e
2.2
500 kHz
d
2.2
0.06
500 MHz
3.4
a Per primary hydroxyl groups.
b Maximum slope of the base addition versus time.
c Analytical yield (ion exchange chromatography).
d Without sonication.
e NaOCl added in 10 portions.

S. Lemoine et al. / Carbohydrate Research 326 (2000) 176–184
181
3. Conclusions
40% 1 M NaOAc in 100 mM NaOH–30%
water; flow rate 1 mL/min). Accuracy was
2–6% over the 15–80% yield range, as esti-
mated from different assays. Results given in
the Tables are mean values over two or three
measurements.
The sonocatalysis of the TEMPO-mediated
oxidation is confirmed in the case of sucrose.
Ultrasound could act at the level of the forma-
tion of the nitrosonium (N+ꢁO) salt through
an easier formation of oxidant species, poten-
tially via homolytic cleavage of chlorine or
bromine or their hypohalous acids. Sucrose
tricarboxylate can thus be obtained in good
yields without the usually necessary sodium
bromide. These observations provide new ele-
ments on the role of the bromide ions in the
catalytic process even under conventional
conditions.
Ultrasound equipment.—For the experi-
ments at 20 kHz, the ultrasound generator
was a 300 W (electric power) Vibra-Cell ap-
paratus, coupled with titanium horns having a
13, 25 or 35 mm diameter (0.1 to 0.6 W/mL
acoustic power). Acoustic power was evalu-
ated by calorimetric measurement. For the
500 MHz experiments, the transductor is a
lead titanate–zirconate ceramic pasted to-
gether with a stainless steel plate (35 mm
diameter, 0.08 to 0.22 W/mL acoustic power).
Hydrogen peroxide titration.—The hydro-
gen peroxide formation rate was evaluated by
the iodometric method (I−3 ions are quantified
at 352 nm). The starting soln (pH 2, aq
H2SO4) was degassed by bubbling argon for
15 min before the ultrasonic irradiation at
15 °C. Samples (250 mL) were taken at inter-
vals of 10 min for 1 h. Potassium iodide (0.1
M) and ammonium molybdate (NH4)6-
Mo6O24·4H2O (0.1 mM) were kept refrigerated
and in the dark before use. Absorbance at 352
nm was measured after a 10 min delay and
compared with a calibration curve.
General procedure for the oxidation reac-
tions.—The same systems were used for both
the classical and the sonochemical reaction. A
connection to a second flask allowed the per-
manent presence of the electrode of the pH-
stat without any risk of deterioration due to
ultrasound. Temperature was maintained by
fluid circulation and measured using a ther-
mocouple. The reaction volume was 200 mL.
To a solution of the carbohydrate substrate (8
mmol based on primary hydroxyl group, i.e.,
4. Experimental
General.—Methyl a- -glucopyranoside was
D
purchased from Sigma, whereas all other
chemicals were purchased from Aldrich. Su-
crose was obtained from Be´ghin-Say S.A. Re-
actions were monitored by thin-layer
chromatography (TLC) using aluminium Sil-
ica Gel plates (60 F254). Flash-chromatogra-
phy separations were performed using E.
Merck Silica Gel 60H (40–63 m) under 1 bar
pressure. Chromatography solvents were pur-
chased from Carlo–Erba S.A. Nuclear mag-
netic resonance spectra were recorded on
Bruker spectrometers at frequencies from 200
1
to 500 MHz for H and from 50 to 125 MHz
13
for C. FAB mass spectra were recorded with
Zab2-Seq VG Micromass (for compound 5)
and ZabSpec TOF Micromass (for compound
1) instruments. Semi-preparative size-exclu-
sion chromatography was performed on
Biogel P2 (Bio–Rad) and elution with water
with RI detection. Analytical ion-exchange
chromatography was performed on
a
DIONEX DX-500 chromatograph using a
Carbopac PA1 column (4×250 mm) with
electrochemical detection (PED) in pulse am-
perometric mode (Au and Ag/AgCl elec-
trodes, with a +0.1, +0.6, −0.8 V pulsation
cycle) and elution with NaOH and NaOAc
mixtures (0–3 min: 40% 200 mM NaOH–10%
1 M NaOAc in 100 mM NaOH–50% water;
3–10 min: linear gradient, 30% 200 mM
NaOH–40% 1 M NaOAc in 100 mM NaOH–
30% water; 10–25 min: 30% 200 mM NaOH–
1.55 g of methyl a- -glucopyranoside or 0.91
D
g of sucrose) was added NaBr (0.32 g, 0.4
equiv per primary hydroxyl group) and
TEMPO (8 mg for 0.65 mol% per primary
hydroxyl group). This mixture was cooled at
5 °C. An approx 12% NaOCl soln (typically
8.5 mL, 2.2 equiv based on primary hydroxyl
groups) adjusted at pH 10 by addition of 4 M
HCl and also cooled at 5 °C was added to the

182
S. Lemoine et al. / Carbohydrate Research 326 (2000) 176–184
reaction mixture, and the pH was adjusted to
10.5 by addition of 0.5 M NaOH. The reac-
tion temperature was maintained at 592 °C
and the pH was kept constant by addition of
0.5 M NaOH. The reaction was stopped (most
often after that a stoichiometric amount of
base was added) by quenching excess oxidant
with EtOH (10 mL) and the mixture was
neutralised by addition of 4 M HCl. The
resulting solution was then concentrated,
freeze-dried and identified by ion exchange
chromatography and NMR spectroscopy in
comparison with authentic standards prepared
as described in the following section. In the
case of overoxidation of sucrose, a pentacar-
boxylate was detected in the reaction mixture
[13C NMR (D2O, 50 MHz): l 71.6, 72.7, 73.3,
73.6, 78.7 (C-2,3,4,5,5%), 93.2 (C-1), 99.7 (C-
2%), 173.1, 173.8, 174.3, 175.4, 177.6 (5
COONa)].
1 H, H-3%), 4.29 (d, 1 H, H-5%), 3.77, 3.81, 3.84
(3s, 9 H, 3 OMe), 3.75 (t, 1 H, J2,3 9.6, J3,4 9.2
Hz, H-3), 3.51 (dd, 1 H, H-2), 3.47 (dd, 1 H,
H-4) 3. 13C NMR (CD3OD, 75 MHz): l
172.5, 171.6, 169.9 (C-6,1%,6%), 104.6 (C-2%),
96.3 (C-1), 82.0, 80.7, 77.3, 74.5, 74.0, 73.7,
73.2 (C-2,3,4,5,3%,4%,5%), 53.9, 53.1, 53.0 (OMe).
Partial separation led to the isolation of a
small amount of the dimethyl ester 3 (Rf 0.33).
1H NMR (CD3OD, 300 MHz): l 5.51 (d, 1 H,
J1,2 3.7 Hz, H-1), 4.41 (d, 1 H, J4,5 9.2 Hz,
H-5), 4.29 (d, 1 H, J3%,4% 8.2 Hz, H-3%), 4.02 (t,
1 H, J4%,5% 8.2 Hz, H-4%), 3.93 (m, 1 H, H-5%),
3.81, 3.84 (2s, 6 H, 2 OMe), 3.76 (m, 3 H,
H-3,6%ab), 3.60 (t, 1 H, J3,4 9.2 Hz, H-4), 3.55
(dd, 1 H, J2,3 9.2 Hz, H-2); 13C NMR
(CD3OD, 75 MHz): l 172.0, 171.0 (C-6,1%),
103.1 (C-2%), 95.7 (C-1), 84.8, 81.1, 75.2, 74.4,
74.0, 73.0, 72.9 (C-2,3,4,5,3%,4%,5%), 63.4 (C-6%),
54.0, 53.3 (OMe).
An analytical sample of sodium (methyl
Preparation of the peracetylated triester
4.—Triester 2 (0.2 g, 0.47 mmol) was treated
in 10 mL of a 1:1 pyridine–Ac2O mixture for
12 h at rt. After evaporation of the solvent
(co-evaporation with toluene), silica gel chro-
matography of the residue (1:1 EtOAc–hex-
ane) provided the peracetylated triester 4, as a
white solid (80 mg, 27%); [h]2D1 +38 (c 1,
a- -glucopyranosid)uronate was prepared by
D
precipitation of a concd aq soln with EtOH.
The solid was washed with a 7:3 EtOH–water
mixture and the structure was confirmed by
13C NMR (50.32 MHz, D2O): l (ppm) 55.4
(OMe), 71.5 (C-2), 72.5 (C-4,5), 73.4 (C-3),
99.8 (C-1), 177.2 (C-6) [17]. The amount of
sodium ions (chloride and bromide) was
quantified by analytical ion chromatography
using an AS11 (4×250 mm) column on the
Dionex DX-500 system (30% 5 mM NaOH–
70% water, flow rate 2 mL/min).
1
CH2Cl2). H NMR (CDCl3, 300 MHz): l 5.91
(d, 1 H, J1,2 3.7 Hz, H-1), 5.62 (t, 1 H,
J3%,4%=J4%,5% 6.6 Hz, H-4%), 5.52 (dd, 1 H, J2,3
10.3, J3,4 9.6 Hz, H-3), 5.49 (d, 1 H, H-3%),
5.12 (dd, 1 H, J4,5 10.6 Hz, H-4), 4.88 (dd,
H-2), 4.61 (d, 1 H, H-5), 4.52 (d, 1 H, H-5%),
3.78, 3.74, 3.72 (3 s, 9 H, 3 CO2Me), 2.18,
2.09, 2.05, 2.02, 2.01 (5 s, 15 H, 5 Ac); 13C
NMR (CDCl3, 50 MHz): l 170.2, 169.8,
169.7, 169.6, 169.5 (5 Ac), 168.2, 167.9, 166.1
(C-6,1%,6%), 101.1 (C-2%), 92.1 (C-1), 78.7, 78.3,
75.4, 69.7, 69.2, 69.0, 68.4 (C-2,3,4,5,3%,4%,5%),
53.2, 52.8, 52.6 (3 OMe), 20.6, 20.5, 20.4, 20.0
(5 Ac). Anal. Calcd for C25O19H32: C, 47.18;
H, 5.07. Found: C, 47.08; H, 5.09.
Preparation of a calibration sample of tri-
ethylammonium sucrose tricarboxylate.—A 10
g crude sample of sodium sucrose tricarboxyl-
ate, prepared by the Pt/O2 method [6] (possi-
bly containing small amounts of correspond-
ing mono- or dicarboxylates), was mixed with
anhyd DMF (20 mL) and methyl iodide (20
mL). The mixture was stirred at room temper-
ature (rt) in the dark for 15 days. The solvents
were evaporated and the residue was purified
by flash-chromatography (30:10:9:1 CH2Cl2–
MeOH–acetone–water). Among fractions
containing the desired product (Rf 0.4), the
purest was evaporated providing the trimethyl
Deacetylation and saponification of 4 was
achieved by careful treatment with a 8:1:1
MeOH–NEt3 –water mixture at rt. Slow de-
protection of the three methyl ester functions
was followed over a 21 day period by TLC
(7:32 propan-2-ol–water). The mixture was
then co-evaporated several times with water in
order to eliminate all the triethylammonium
1
ester 2 as an oil (1.7 g, 18%). H NMR
(CD3OD, 300 MHz): l 5.43 (d, 1 H, J1,2 3.7
Hz, H-1), 4.47 (d, 1 H, J4,5 9.9 Hz, H-5), 4.45
(dd, 1 H, J3%,4% 6.3, J4%,5% 7.0 Hz, H-4%), 4.37 (d,

S. Lemoine et al. / Carbohydrate Research 326 (2000) 176–184
183
acetate formed, thus producing the pure tri-
ethylammonium sucrose tricarboxylate 5 in
quantitative yield. 2D C–H NMR spec-
troscopy at 300 and 500 MHz allowed full
la Recherche et de la Technologie for a grant
(contrat CIFRE), and C.T., the CNRS for a
grant. This work was partly supported by the
Programme Aliment Demain, convention
R94/10 and the COST D10 network on Inno-
vative Methods and Techniques for Chemical
transformations.
1
assignment of all patterns: H NMR (D2O,
300 MHz): l 5.52 (d, 1 H, J1,2 3.5 Hz, H-1),
4.17 (t, 1 H, J3%,4% 7.8, J4%,5% 7.0 Hz, H-4%), 4.09
(d, 1 H, H-5%), 4.08 (d, 1 H, J4,5 9.5 Hz, H-5),
4.07 (d, 1 H, H-3%), 3.73 (t, 1 H, J3,4 9.5, J2,3
9.8 Hz, H-3), 3.44 (dd, 1 H, H-2), 3.34 (t, 1 H,
H-4), 3.20 (q, 8 H, J=7 Hz, 3 Et), 1.30 (t, 9
References
13
H, 3 Et); C NMR (D2O, 125 MHz): l 176.9,
[1] H. Ro¨per, in F.W. Lichtenthaler (Ed.), Carbohydrates as
Organic Raw Materials I, VCH, Weinheim, 1993, pp.
267–288.
[2] H. van Bekkum, in F.W. Lichtenthaler (Ed.), Carbohy-
drates as Organic Raw Materials I, VCH, Weinheim,
1993, pp. 289–310.
[3] A.C. Besemer, H. van Bekkum, Starch, 46 (1994) 95–
101.
[4] M. Floor, L.P.M. Hofsteede, W.P.T. Groenland, L.A.T.
Verhaar, A.P.G. Kieboom, H. Van Bekkum, Recl. Tra6.
Chim. Pays-Bas, 108 (1989) 384–392.
[5] D. de Wit, F. van rantwijk, L. Maat, A.P.G. Kieboom,
Recl. Tra6. Chim. Pays-Bas, 109 (1990) 518–522.
[6] W. Fritsche-Lang, E.I. Leupold, M. Schlingmann, Ger.
Offen. DE 3535720 (1987); Chem. Abstr., 107 (1987)
59,408.
[7] L.A. Edye, G.V. Meehan, G.N. Richards, J. Carbohydr.
Chem., 10 (1991) 11–23.
[8] L.A. Edye, G.V. Meehan, G.N. Richards, J. Carbohydr.
Chem., 13 (1994) 273–283.
[9] M. Kunz, H. Puke, C. Recker, L. Scheiwe, J. Kowalczyk,
Ger. Offen. DE 4307388 (1994); Chem. Abstr., 122 (1995)
56,411.
175.9 (C-6,6%), 174.7 (C-1%), 103.5 (C-2%), 93.9
(C-1), 80.1 (C-5%), 79.7 (C-3%), 75.7 (C-4%), 73.7
(C-5), 73.3 (C-3), 72.2 (C-4), 71.5 (C-2), 46.6
(Et), 8.3 (Et). High-resolution FABMS (thio-
glycerol matrix): calcd for C24H47O14N2: m/z
587.3027; found 587.3010 [(M−HNEt3+
2H)]+. Treatment of the triester 2 (30 mg) in
MeOH (4.3 mL) with NaOH (0.1 N, 1 equiv)
at rt led to the corresponding sodium salt 1
which could be further purified by size-exclu-
sion chromatography using Biogel P6 and elu-
1
tion with water. H NMR (D2O, 300 MHz): l
5.44 (d, 1 H, J1,2 3.7 Hz, H-1), 4.24 (t, 1 H,
J3%,4%=J4%,5% 7.7 Hz, H-4%), 4.17 (d, 1 H, J4,5 9.8
Hz, H-5), 4.17,4.15 (2 d, 2 H, H-3%, H-5%), 3.84
(t, 1 H, J2,3 =J3,4 9.6 Hz, H-3), 3.55 (dd, 1 H,
H-2), 3.41 (t, 1 H, H-4), 13C NMR (D2O, 75
MHz): l 177.3, 176.9 (C-6,6%), 174.9 (C-1%),
104.1 (C-2%), 94.6 (C-1), 80.6 (C-5%), 79.9 (C-
3%), 76.2 (C-4%), 73.5 (C-5), 73.3 (C-3), 72.4
(C-4), 71.7 (C-2). High-resolution FABMS for
1 (glycerol matrix): calcd for C12H14O14Na3:
m/z 451.0077; found 451.0097 [M+H]+.
[10] M. Kunz, C. Recker, Carbohydr. Eur., 13 (1995) 11–15.
[11] Austrian patent AT 401647 (1996); Chem. Abstr., 126
(1997) 93,664.
[12] P. Parpot, K.B. Kokoh, B. Beden, C. Lamy, Electrochim.
Acta, 38 (1993) 1679–1683.
[13] P. Parpot, K.B. Kokoh, B. Beden, E.M. Belgsir, J.M.
Leger, C. Lamy, Stud. Surf. Sci. Catal., 78 (1993) 439–
445.
[14] T. Ishiguro, M. Oka, T. Yamaguchi, I. Nogami, Eur. Pat
Appl. EP 599646 (1994); Chem. Abstr., 121 (1994)
132,399.
[15] P.L. Anelli, C. Biffi, F. Montanari, S. Quici, J. Org.
Chem., 52 (1987) 2559–2562.
Acknowledgements
[16] A.E.J. de Nooy, A.C. Besemer, H. van Bekkum, Synthe-
sis, (1996) 1153–1174.
[17] J.A. Cella, J.A. Kelley, E.F. Kenehan, J. Org. Chem., 40
(1975) 1860–1862.
[18] M.F. Semmelhack, C.S. Chou, D.A. Cortes, J. Am.
Chem. Soc., 105 (1983) 4492–4494.
[19] M.F. Semmelhack, C.R. Schmid, D.A. Cortes, Tetrahe-
dron Lett., 27 (1986) 1119–1122.
[20] S.D. Rychnovsky, T.L. McLernon, H. Rajapakse, J.
Org. Chem., 61 (1996) 1194–1195.
[21] N.J. Davis, S.L. Flitsch, Tetrahedron Lett., 34 (1993)
1181–1184.
[22] A.E.J. de Nooy, A.C. Besemer, H. van Bekkum, Recl.
Tra6. Chim. Pays-Bas, 113 (1994) 165–166.
[23] A.E.J. de Nooy, A.C. Besemer, H. van Bekkum, Carbo-
hydr. Res., 269 (1995) 89–98.
The authors thank Professor C. Pe´trier and
Dr J.-L. Luche (Universite´ de Savoie, Le
Bourget-du-Lac) for fruitful discussions, Dr
J.-P. Maˆıtre (Be´ghin-Say R&D, Villeurbanne)
for kindly providing some tricarboxysucrose
trisodium salt, and J.-M. Taconnat (Be´ghin-
Say R&D, Toulouse) for his help in the Biogel
separation of the same compound from a
crude NaOCl/TEMPO-mediated reaction mix-
ture. Financial support from the CNRS and
Be´ghin-Say is greatly acknowledged; S.B.
thanks Be´ghin-Say and l’Agence Nationale de

184
S. Lemoine et al. / Carbohydrate Research 326 (2000) 176–184
[24] D.L. Verraest, J.A. Peters, H. van Bekkum, Zuckerindus-
trie, 120 (1995) 799–803.
[25] K. Schnatbaum, H.J. Scha¨fer, Synthesis, (1999) 864–872.
[26] A.C. Besemer, H. van Bekkum, Starch, 46 (1994) 101–
106.
beet Sugarcane, Proc. S.P.R.I. Workshop, 1994, pp. 264–
293; Chem. Abstr., 124 (1996) 11,277.
[32] S. Brochette, G. Descotes, A. Bouchu, Y. Queneau,
N. Monnier, C. Pe´trier, J. Mol. Catal., 123 (1997) 123–
130.
[27] S.D. Rychnovsky, R. Vaidyanathan, J. Org. Chem., 64
(1999) 310–312.
[33] S. Brochette, G. Descotes, A. Bouchu, Y. Queneau, J.
Lemoine, J. Carbohydr. Chem., 17 (1998) 879–891.
[34] G. Descotes, Y. Queneau, in W. Praznik, A. Huber
(Eds.), Carbohydrates as Organic Raw Materials IV,
WUV, Wien, 1998, pp. 39–63.
[35] C. Pe´trier, M.-F. Lamy, A. Francony, A. Benahcene, B.
David, V. Renaudin, N. Gondrexon, J. Phys. Chem., 98
(1994) 10,514–10,520.
[28] S.D. Rychnovsky, R. Vaidyanathan, T. Beauchamp, R.
Lin, P.J. Farmer, J. Org. Chem., 64 (1999) 6745–6749.
[29] S. Lemoine, D. Joannard, G. Descotes, A. Bouchu, Y.
Queneau, J. Mol. Catal., 150 (1999) 31–36.
[30] G. Fle`che, EP 0 798 310 A1 (1996); Chem. Abstr., 127
(1997) 307619p.
[31] H. van Bekkum, D.L. Verraest, A.E.J. de Nooy, C.
Besemer, J.A. Peters, in: M.A. Clarke (Ed.), Prod. Sugar-
[36] T.J. Mason, A. Newman, J.P. Lorimer, J. Lindley, K.
Hutt, Ultrason. Sonochem., 3 (1996) 53–55.
.