Determination of the side-products formed during the nitroxide-mediated bleach oxidation of glucose to glucaric acid

Article abstract of DOI:10.1016/S0008-6215(02)00072-1

The side products formed in the TEMPO-mediated oxidation of glucose to glucaric acid were determined by GC. Next to glucaric acid, gluconic acid, the intermediate in the oxidation, the degradation products, oxalic acid, tartronic acid, meso-(erythraric) a

Full text of DOI:10.1016/S0008-6215(02)00072-1

Carbohydrate Research 337 (2002) 1059–1063
www.elsevier.com/locate/carres
Note
Determination of the side-products formed during the
nitroxide-mediated bleach oxidation of glucose to glucaric acid
Mathias Ibert,a Francis Marsais,a,* Nabyl Merbouh,b Christian Bru¨cknerb,*
aUMR 1064-IRCOF-INSA de Rouen, BP08, F-76131 Mont Saint Aignan cedex, France
bDepartment of Chemistry, Uni6ersity of Connecticut, Storrs, CT 06269-3060, USA
Received 3 December 2001; accepted 22 February 2002
Abstract
The side products formed in the TEMPO-mediated oxidation of glucose to glucaric acid were determined by GC. Next to
glucaric acid, gluconic acid, the intermediate in the oxidation, the degradation products, oxalic acid, tartronic acid, meso-
(erythraric) and DL-threaric (tartaric) acid were detected. Chiral GC determined the DL-tartaric acid to be non-racemic mixtures
of L- and D-tartaric acids, with inverse D/L-ratios depending on the oxidation of D- or L-glucose. The origin of all degradation
products is rationalized. This study details a fast screening method to optimize the reaction conditions toward minimal
degradation. © 2002 Published by Elsevier Science Ltd.
Keywords: Glucaric acid; TEMPO-mediated oxidations; GC analysis of hydroxyacids
mercially available nitroxide catalysts.12,13,18 The system
TEMPO–NaBr–NaOCl is utilized in the high yield
1. Introduction
oxidation of unprotected mono- and polysaccha-
rides.10,16,19,20 Thus, glucaric acid, isolated as its
monopotassium or disodium salt, can be made under
carefully controlled pH and temperature conditions
from glucose in yields approaching 90%.10 Deviations
from the ideal protocol decreases the yield of glucaric
acid and increases the formation of degradation (over-
oxidation) products.
Glucaric acid has potentially a number of industrially
relevant applications in polymer, food, and medicinal
chemistry.1–3 Though naturally occurring,3 it is most
conveniently prepared by oxidation of
D-glucose using
a number of oxidation methodologies.4–11 The most
established method uses nitric acid as oxidant but pro-
duces, as its monopotassium salt, glucaric acid in only
ꢀ40% yield.5,6 The remaining 60% of the starting
material is degraded in due course of the reaction to a
complex mixture consisting, in part, of carbonate, tar-
taric, oxalic, and ‘5-ketogluconic acid’.5
What are the over-oxidation products? Some older
methods for the determination of the over-oxidation
products oxalic and tartaric acid have been de-
scribed.5,21–23 However, these methods are not
amenable to the convenient determination of a larger
N-Oxide mediated oxidations in aqueous base using
bleach (NaOCl) or other terminal oxidants have, in
recent years, proven to be promising for the selective
oxidation of alcohols in general,12–15 and the produc-
tion of aldaric acids from aldoses in particular.10,11,16,17
2,2,6,6-Tetramethyl-1-piperidinyloxy (TEMPO) and its
4-acetamido derivative are commonly used and com-
1
number of samples. H NMR methods for the determi-
nation of the side products are, due to overlapping
peaks or the absence of protons in the side products not
ideal either.24 13C NMR methods are time consuming
and allow for the quantitative determination of the side
products only under special conditions.25 Can one de-
vise a simple and reliable method to determine the
amount and identity of the side products? Do the side
products illuminate particular degradation pathways?
This contribution evaluates a fast method for the
determination of the short-chain carboxylic acids side
* Corresponding authors. Tel.: +33-2-35522475 (F.M.);
tel.: +1-860-4862743 (C.B.).
E-mail addresses: francis.marsais@insa-rouen.fr (F. Mar-
sais), c.bruckner@uconn.edu (C. Bru¨ckner).
0008-6215/02/$ - see front matter © 2002 Published by Elsevier Science Ltd.
PII: S0008-6215(02)00072-1

1060
M. Ibert et al. / Carbohydrate Research 337 (2002) 1059–1063
bleach as terminal oxidant, and we propose a degrada-
tion pathway of glucose (and/or gluconic or glucaric
acid) under these reaction conditions.
products by gas chromatography. Using GC and
NMR, we will describe the side products of the nitrox-
ide-mediated oxidation of glucose to glucaric acid using
2. Results and discussion
Glucose oxidation.—The oxidation protocol chosen
for the oxidation of glucose is similar to that described
by us in earlier contributions.10,26 A bleach solution is,
under tight pH and temperature control, added to an
aqueous NaOH solution of glucose containing TEMPO
and the co-catalyst, bromide. The oxidation products
are isolated from the reaction mixture by precipitation
1
with EtOH. At pH 11.7, H NMR-spectroscopic yields
close to 90% of sodium glucarate can be obtained.26 At
pH 10.5, the percentage of side-products increases on
the expense of glucarate yields—they drop to as low as
25%. Analysis of the crude reaction mixtures by 1H
NMR, quantitative 13C NMR, as well as a number of
precipitation procedures allowed the determination of
the over-oxidation products to be tartrate, oxalate, and
carbonate.10,24 Fig. 1 shows a comparison of the 1H
NMR spectra of a reaction mixture prepared at pH
10.5 and 11.7, clearly displaying the increased number
of non-glucaric acid peaks. Overlapping peaks and the
inherent difficulty in discerning minor components in
mixtures using NMR led us to investigate an alternative
analytical method of the identification of the side
products.
1
Fig. 1. H NMR (300 MHz, D2O) spectrum of crude reaction
mixtures of nitroxide-mediated oxidations of glucose to glu-
caric acid using bleach as primary oxidant, performed at the
pH indicated (for experimental details see Section 3). ꢀ
indicates signals characteristic of gluconate; ꢁ indicates sig-
nals for glucarate; and * indicates the signals for the tartrates
(meso- and DL-form).24
Deri6atization of the oxidation product-mixture and
GC analysis.—The oximation–trimethylsilylation and
subsequent gas-chromatographic analysis of carbohy-
drates is a well established technique.27–30 Thus, the
precipitated, dried and powdered reaction mixture was
converted into its volatile trimethylsilyl ethers and es-
ters. Fig. 2 shows the GC trace of the silylated reaction
mixture resulting from the oxidation of D-glucose. Next
to glucaric acid and gluconic acid, the degradation
products, oxalic acid (major product), tartronic acid
(traces), meso- (erythraric) and DL-threaric (tartaric)
acid (minor and major products, 1:6 ratio) can be
detected. While not revealing unexpected degradation
products, the trace demonstrates the power of the GC
method for the convenient and rapid analysis of such
reaction mixtures. It also confirms the absence of
D-
xylo-5-hexulosonic ‘5-ketogluconic’ acid, a common
oxidation product in the nitric acid oxidation of
glucose.5,23
Interpretation of the oxidation products.—NMR in-
vestigations of the mechanism of the hypochlorite oxi-
dation have shown that the oxidation of glucose to
gluconic acid is fast and accompanied by the formation
of relatively few side products.11 In fact, this fast and
selective oxidation has been utilized for the preparation
of gluconate.31,32 The N-oxide-mediated oxidation of
Fig. 2. GC trace of a silylated crude reaction mixture. Peak
assignment, retention time, rel. amount (reference compound
L
-malic acid=100%): (1) oxalic acid, 6.73 min, 30.8%; (2),
tartronic acid, 13.33 min, 0.4%; (3), -malic acid, 15.98 min;
L
(4) meso-tartaric acid, 18.61 min, 3.7%; (5) DL-tartaric acid,
19.84 min, 23.9%; (6) gluconic acid, 28.29 min, 23.8%; (7)
glucaric acid, 28.43 min, 17.4%. The unassigned peaks at low
retention times can be attributed to reagent peaks.

M. Ibert et al. / Carbohydrate Research 337 (2002) 1059–1063
1061
Evidently, the observed high selectivity for the conver-
sion of glucose to glucaric acid is the result of a
fortuitous scaling of the rates of the various competing
oxidation reactions and the low concentration of oxi-
dants during the experiment.
DL-tartaric acid forms over the meso-form in a 6:1
ratio, indicating a pronounced regioselectivity for the
degradation reaction. Likely, this differentiation reflects
the effects of coordination of glucarate to the sodium
ions in solution. The oxygen atoms attached to C-1,
C-2, and C-3 of glucaric, as well as gluconic acid, have
been shown by NMR, UV–CD spectroscopy, and X-
ray crystallography studies to bind preferentially over
all other oxygen atoms to a number of metal ions.34–39
Coordination to, in this case, the sodium cations in the
alkaline reaction solution may serve as a steric and/or
electronic protection towards C-2ꢀC-3 cleavage, direct-
ing degradation toward C-4ꢀC-5 cleavage. Alterna-
tively, the cation may induce a conformationally
restricted, cyclic chelate structure. This may set up sets
of cis- and trans-diols exhibiting differing cleavage
rates. These explanations suggest a counter-ion influ-
ence on the outcome of the oxidation reaction, a hy-
pothesis as yet not tested. Oxalic acid and CO2 (or
CO23−), the ultimate degradation products are gener-
ated by a number of degradation reactions.40
Fig. 3. Proposed genealogy of the degradation products ob-
served in the nitroxide-mediated oxidation of -glucose to
-glucaric acid.
D
D
the primary alcohol of the gluconate to a carboxylate is
slow. In the absence of ideal reaction conditions, this
conversion is accompanied by extensive over-oxidation.
It is noteworthy, though, that no side products with
primary alcohol termini other than gluconic acid were
detected.
Remarkably, no direct evidence for any C-1ꢀC-2
cleavage or C-5ꢀC-6 cleavage are found, that is, no
arabinose or xylose (or their corresponding -onic, -aric,
or -uronic acids) were detected. While the preparation
Fig. 3 depicts the proposed genealogy of the degrada-
tion products with glucaric acid at its origin. We realize
that glucaric acid is not the only possible source of the
degradation products detected. Pure gluconic or glu-
caric acid, exposed to the oxidation conditions, exhibit
very similar degradation patterns as those shown in
Fig. 2. Thus, the degradation products detected by GC
(oxalic, tartronic, and meso- and DL-tartaric acid) can
be rationalized by the degradations steps indicated:
C-3ꢀC-4 cleavage of glucaric acid forms tartronic acid;
cleavage between C-2ꢀC-3 and C-4ꢀC-5 generates tar-
of arabinose from
D-glucose using hypochlorite is
known, this reaction requires acidic conditions and long
reaction times in the absence of a catalyst.31,32 How-
ever, there is indirect evidence that (fast) terminal car-
bon cleavages occur: Chiral GC analysis of the silylated
oxidation mixture generated from oxidation of
cose indicated the DL-tartaric acid to be a 16:3 mixture
of - to -tartaric acid (Fig. 4(B)). The -form of
D-glu-
L
D
L
tartaric acid is the expected degradation product result-
taric acids: meso-tartaric acid in the former and
L-
ing from C-4ꢀC-5 cleavage of glucaric acid. The origin
threaric ( -tartaric) acid in the latter case. Hypochlorite
L
of the
the
epimerization are too mild.41 The only way we can
perceive the formation of -tartrate is by consecutive
C-1ꢀC-2 and C-5ꢀC-6 cleavages of -glucaric (or any of
its precursors). The larger quantity of -tartaric acid
D-form is less obvious. It cannot originate from
and hypobromite in the absence of a catalyst are gener-
ally not known to be able to cleave diols, although
examples to the contrary are known.33 However, con-
trol experiments in which glucaric acid was exposed to
a several-fold molar excess bleach in the absence of the
nitroxide catalyst but in the presence and absence of
bromide at high pH indicated that, over 24 h, glucaric
acid was entirely oxidized to the point where no CH
L
-form because the reaction conditions for an
D
D
L
and the absence of any five-carbon acids in the oxida-
tion mixture imply a very efficient cleavage reaction.
We cannot offer an explanation for this circumstantial
1
signals were detectable in the H NMR spectrum. Like-
evidence. If this hypothesis for the origin of
acid is in general true, oxidation of -glucose under
identical reaction conditions ought to produce an in-
verse amount of - to -tartaric acid. As Fig. 4(C)
D-glucaric
wise, reaction of glucarate with a severalfold stoichio-
metric excess of independently prepared 4-
acetamido–TEMPO-based oxammonium salt18 under
basic conditions resulted in complete decomposition.
Thus, the oxammonium salt as well as the terminal
oxidant are capable of over-oxidizing glucaric acid.
L
L
D
demonstrates, this is indeed observed.
In summary, the degradation products formed during
the nitroxide-mediated oxidation of glucose to glucaric

1062
M. Ibert et al. / Carbohydrate Research 337 (2002) 1059–1063
acid under basic conditions using bleach as terminal
oxidant were determined by GC analysis. A rationale
for their formation is offered. We have thus provided
the necessary analytical tools for the facile optimization
of the reaction conditions for this or related reactions
in industrial or laboratory settings.
with EtOH, and the pH was adjusted to 8.5 with 1.0 M
HCl. The solution was concentrated to 50 mL, and the
oxidation products were precipitated by addition of
EtOH (150 mL). The precipitate was dissolved in water
(50 mL), and the precipitation operation was repeated.
The resulting solid was washed with 4:1 EtOH–water,
and dried at 50 °C under diminished pressure. An
NMR of the mother liquors proved complete precipita-
tion of the hydroxyacids.
3. Experimental
GC analysis of the oxidation products.—The pow-
dered reaction mixture (1–5 mg) was dissolved in 1 mL
of an anhyd pyridine stock solution (100 mL) contain-
ing the internal reference compounds glutaric acid (50
Materials.—All chemicals, including anhyd
-glucose and TEMPO were analytical-grade commer-
cial products (Aldrich) and were used without further
purification.
D- and
L
mg) and
L-malic acid (50 mg) (it was confirmed that
Glucose oxidation procedure.—D-Glucose (5 g, 27.8
neither compound was an over-oxidation product of
glucose). Methoxylamine hydrochloride (30–40 mg)
and 500 mL of [bis(trimethylsilyl)trifluoroacetamide]
(BSTFA, containing 1% Me3SiCl) are added. The mix-
ture is sonicated for 5 min, and the resulting solution is
heated to 60–70 °C for 30 min. About 1 mL of the
resulting solution is injected into the gas chro-
matograph for analysis. All peak assignments were
confirmed with genuine materials.
mmol), TEMPO (45 mg, 0.01 equiv), and NaBr (0.40 g,
0.15 equiv), were dissolved in water (60 mL) in a 200
mL beaker equipped with a thermometer and a pH
electrode connected to a Logilap™ system. The reac-
tion vessel was cooled in a salt-ice mixture to 0 °C. In a
separate vessel, a 1.8 M NaClO solution (51 mL, equal
to 2.2 equiv ClO−/mmol primary OH plus 1.1 equiv
ClO−/mmol of hemiacetal OH) was adjusted to the
desired pH with 4 M HCl and the solution was cooled
to 0 °C. Controlled by the Logilap™ system, the chilled
oxidant was added to the glucose solution at a rate not
allowing the reaction temperature to exceed 5 °C. The
pH was held constant at the desired pH by automated
titration with an 1.5 M NaOH solution. When the
oxidation was completed (all oxidant added and no
further pH shift detectable), the reaction was quenched
Instrumentation.—NMR spectra were recorded on
Bruker AC300F or DRX400 spectrometers, operating
1
at H NMR frequencies of 300 and 400 MHz, respec-
tively. The non-chiral GC spectra were recorded on a
Varian 3400 using a J&W Scientific DB1 column (30 m,
0.32 mm diameter, 0.25 mm film). Injector temperature
300 °C, column temperature ramped from 80 to 250 °C
at a rate of 5 °C/min and from 250 to 320 °C at a rate
of 10 °C/min. Carrier gas rate 78 mL/min. FID detector
at 300 °C. Under these conditions, all compounds were
eluted between 4 and 34 min. The chiral GC was
performed using a Hewlett–Packard 5890 instrument
equipped with a Supelco Beta-Dex 120 column (15 m,
0.25 mm diameter, 0.25 mm film). Injector temperature
250 °C. The spectrum was recorded isothermally at
100 °C, carrier gas rate 50 cm/s and a split ratio of 1:20.
FID detector at 250 °C.
References
1. Pamuk V.; Yilmaz M.; Alicilar A. J. Chem. Technol.
Biotechnol. 2001, 76, 186–190.
2. Chen L.; Kiely D. E. J. Org. Chem. 1996, 61, 5847–5851.
3. Walaszek Z. Cancer Lett. 1990, 54, 1–8.
4. Mehltretter, C. L. US Patent 2,472,168, 1949.
5. Mehltretter C. L.; Rist C. E. J. Agr. Food. Chem. 1953, 1,
779–783.
6. Mehltretter C. L. In
D-Glucaric Acid; Whistler R. L.;
Fig. 4. Truncated chiral GC traces of (A) DL-tartaric acid
Wolfrom M. L., Eds.; Academic Press: New York, 1962;
Vol. 2, pp 46–48.
7. Ro¨per H. In Selecti6e Oxidation of
(
D
-tartaric acid, 52.4 min,
standard -malic acid, 23.07 min. (B) Oxidation product
mixture resulting from the oxidation of -glucose (15% -,
85% -tartaric acid). (C) Oxidation product mixture resulting
from the oxidation of -glucose (90% -, 10% -tartaric acid).
L-tartaric acid, 54 min), internal
L
D-Glucose: Chiral
D
D
Intermediates for Industrial Utilization; Lichtenthaler F.
W., Ed.; VCH: Weinheim, 1991.
L
L
D
L

M. Ibert et al. / Carbohydrate Research 337 (2002) 1059–1063
1063
25. Derome A. E. Modern NMR Techniques for Chemistry
Research; Pergamon Press: Oxford, 1991.
26. Merbouh N.; Thaburet J. F.; Ibert M.; Marsais F.;
Bobbitt J. M. Carbohydr. Res. 2001, 336, 75–78.
27. Sweeley C. C.; Bentley R.; Makita M.; Wells W. W. J.
Am. Chem. Soc. 1963, 85, 2497–2507.
28. Heartherball D. A. J. Sci. Food Agr. 1974, 25, 1095–
1107.
29. Robards K.; Whitelaw M. J. Chromatogr. 1986, 373,
81–110.
30. Bartolozzi F.; Bertazzi G.; Bassi D.; Cristoferi G. J.
Chromatogr. A. 1997, 758, 99–107.
31. Whistler R. L.; Schweiger R. J. Am. Chem. Soc. 1959, 81,
5190–5192.
32. Whistler R. L.; Yagi K. J. Org. Chem. 1961, 26, 1050–
1052.
33. Hammad N.; Sutherland A. G. J. Chem. Res., Synop.
1996, 3, 158–159.
34. Sawyer D. T.; Brannan J. R. Anal. Chem. 1966, 38,
192–198.
35. Cook W. J.; Bugg C. E. Acta. Crystallogr., Sect. B. 1973,
29, 215–222.
36. Taga T.; Kuroda Y.; Osaki K. Bull. Chem. Soc. Jpn.
1977, 50, 3079–3083.
37. Lis T. Acta. Crystallogr., Sect. C 1984, 40, 376–378.
38. Venema F. R.; Peters J. A.; van Bekkum H. Recl. Tra6.
Chim. Pays-Bas 1993, 112, 445–450.
39. Daniele P. G.; De Stefano C.; Giuffre´ O.; Prenesti E.;
Sammartano S. J. Chem. Soc., Dalton Trans. 2002, 435–
440.
40. Floor M.; Schenk K. M.; Kieboom A. P. G.; van
Bekkum H. Starch/Sta¨rke 1989, 41, 303–309.
41. Holleman A. F. Org. Synth., Coll. I 1941, 491–497.
8. Kiely, D.; Carter, A.; Shrout, D. US Patent 5,599,977,
1995.
9. Besson M.; Fleche G. Recl. Tra6. Chim. Pays-Bas 1996,
115, 217–221.
10. Thaburet J. F.; Merbouh N.; Ibert M.; Marsais F.;
Queguiner G. Carbohydr. Res. 2001, 330, 21–29.
11. Merbouh, N.; Bobbitt, J. M.; Bru¨ckner, C. J. Carbohydr.
Chem. 2002, 21, 1–14.
12. Bobbitt J. M.; Flores C. L. Heterocycles 1988, 27, 509–
533.
13. de Nooy A. E. J.; Besemer A. C.; van Bekkum H.
Synthesis 1996, 1153–1175.
14. Zhao M.; Li J.; Mano E.; Song Z.; Tsahen D. M.;
Grabowski E. J. J.; Reider P. J. Org. Chem. 1999, 64,
2564–2566.
15. Merbouh N.; Bobbitt J. M.; Bru¨ckner C. Tetrahedron.
Lett. 2001, 8793–8796.
16. Davis N. J.; Flitsch S. L. Tetrahedron Lett. 1993, 34,
1181–1184.
17. de Nooy A. E. J.; Besemer A. C.; van Bekkum H.
Tetrahedron 1995, 51, 8023–8032.
18. Bobbitt J. M. J. Org. Chem. 1998, 63, 9367–9375.
19. de Nooy A. E. J.; Besemer A. C.; van Bekkum H. Recl.
Tra6. Chim. Pays-Bas 1994, 113, 165–166.
20. de Nooy A. E. J.; Besemer A. C.; van Bekkum H.
Carbohydr. Res. 1995, 269, 89–98.
21. Barch W. E. J. Am. Chem. Soc. 1933, 55, 3653–3658.
22. Dale K. J.; Rice W. F. J. Am. Chem. Soc. 1933, 55,
4984–4985.
23. Lesquibe F. Industr. Alim. Agr. 1967, 84, 1493–1494.
24. van Duin M.; Peters J. A.; Kieboom A. P. G.; van
Bekkum H. Magn. Reson. Chem. 1986, 24, 832–833.