Isomerization of d-glucose into d-fructose with a heterogeneous catalyst in organic solvents

Article abstract of DOI:10.1016/j.catcom.2013.05.004

The isomerization of d-glucose into d-fructose was investigated in view to afford carbohydrate source for the production of 5-hydroxymethylfurfural (HMF) without the use of toxic metallic catalyst. The process uses a basic heterogeneous catalyst in organic solvent mixtures to lead up to 72% selectivity of d-fructose for 68% conversion of d-glucose. The catalyst has been recycled four times with little decrease of the selectivity. The resulting carbohydrates solution can be directly dehydrated to afford HMF.

Full text of DOI:10.1016/j.catcom.2013.05.004

Isomerization of D-glucose into D-fructose with a heterogeneous catalyst in
organic solvents
Solenne Despax, Boris Estrine, Norbert Hoffmann, Jean Le Bras, Sinisa
Marinkovic, Jacques Muzart
PII:
DOI:
Reference:
S1566-7367(13)00170-2
doi: 10.1016/j.catcom.2013.05.004
CATCOM 3498
To appear in:
Catalysis Communications
Received date:
Revised date:
Accepted date:
12 March 2013
2 May 2013
4 May 2013
Please cite this article as: Solenne Despax, Boris Estrine, Norbert Hoffmann, Jean Le
Bras, Sinisa Marinkovic, Jacques Muzart, Isomerization of D-glucose into D-fructose
with a heterogeneous catalyst in organic solvents, Catalysis Communications (2013), doi:
1
0.1016/j.catcom.2013.05.004
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

ACCEPTED MANUSCRIPT
Isomerization of D-glucose into D-fructose
with a heterogeneous catalyst in organic
solvents
a,b
a*1
b
b
Solenne DESPAX , Boris ESTRINE , Norbert HOFFMANN , Jean LE BRAS , Sinisa
a
b
MARINKOVIC , Jacques MUZART
a
Agro-industrie Recherches et Développements, Green Chemistry Department, POMACLE,
France
b
Institut de Chimie Moléculaire de Reims, UMR 7312, CNRS-Université de Reims
Champagne-Ardenne, UFR des Sciences Exactes et Naturelles, REIMS, France
*1
Corresponding author. Tel.:+33 326-054-284; fax: +33 326-054-289
E-mail address: b.estrine@a-r-d.fr
Abstract
The isomerization of D-glucose into D-fructose was investigated in view to afford
carbohydrates source for the production of 5-hydroxymethylfurfural (HMF) without the use of
toxic metallic catalyst. The process uses a basic heterogeneous catalyst in organic solvent
mixtures to lead up to 72% selectivity of D-fructose for 68% conversion of D-glucose. The
catalyst has been recycled four times with little decrease of the selectivity. The resulting
carbohydrates solution can be directly dehydrated to afford HMF.
Keywords
D-glucose; D-fructose; isomerization; heterogeneous catalyst; organic solvent; dehydration.
1

ACCEPTED MANUSCRIPT
1
. Introduction
Accelerated depletion of non-renewable sources of energy and rising oil prices have prompted
a surge in research to develop alternatives to fuels and chemicals from renewable materials. In
this way, 5-hydroxymethylfurfural (HMF) is a biobased platform chemical with high
potential. HMF serves as starting material for a wide variety of biobased chemical
intermediates (e.g., levulinic acid, furandicarboxylic acid, dimethyl furan, γ-valerolactone,
and dihydroxymethylfuran). The manufacture of 5-hydroxymethylfurfural (HMF) through the
dehydration reaction of sugars is one of the most important approaches to transform biomass
into useful chemicals [1, 2]. HMF is preferentially synthesized by dehydration of
monosaccharides. The dehydration is more effective from D-fructose than from D-glucose
[
3], but D-glucose is more available than D-fructose. Consequently, the isomerization of D-
glucose into D-fructose appears to be an important step for the efficient formation of HMF
cf. Figure 1).
(
Fig. 1. Two-step D-Glucose conversion into HMF.
Industrially, the isomerization of D-glucose into D-fructose is carried out with enzymes but
this suffers from various drawbacks such as the need of low concentrated solutions of D-
glucose with high purity and limited operating temperature range. Overall, this limits the
interest of an enzymatic process for producing D-fructose in the context of HMF [3, 4, 5].
Preparation of D-fructose can also be done by the isomerization of D-glucose with chemical
catalysts. Most isomerization reactions mediated by chemical catalysis are done in aqueous
solution. Isomerization of aldoses by the action of aqueous bases such as Ca(OH) , NaOH,
2
KOH is well known. This isomerization reaction is called the ‘Lobry de Bruyn-van
Eckeinstein transposition’ according to the name of the inventors [6, 7]. Despite the use of
inexpensive catalysts, these reactions suffer from low concentration of glucose and long
reaction time [8, 9]. Moreover, these processes have the disadvantage to random degrade
carbohydrates giving low yields and requiring complexes steps in order to remove
degradation products. On the other hand, chromium complexes are known to directly afford
2

ACCEPTED MANUSCRIPT
HMF from D-glucose via an isomerization step, but their use causes toxicity problems [10-
1
3]. From our knowledge, the best results are disclosed in a French patent [14]. Using
potassium aluminate, high yields of D-fructose are obtained but the catalyst need to be
neutralized by acid prior to be removed by filtration. Other processes have been designed in
aqueous solutions with cation-exchanged zeolites or hydrotalcites as catalysts [15-18]. These
methods require high temperature (95 °C) and nitrogen pressure (8 bars) to avoid oxidation
reactions. Under these conditions, the D-fructose selectivity was high but only for a low D-
glucose conversion and initial concentration (9 wt %). Moreover, leaching leads to reduced
activity of the recycled catalyst. More recently, Essayem’s team [19] used basic hybrid
catalysts in a lower amount (5 wt % compared to the D-glucose) but at 100°C under 20 bar of
nitrogen to afford D-fructose with a good selectivity but with a moderate yield. The
recyclability of the catalyst was not studied but infrared analysis has shown modifications of
the catalyst.
As all the reactions described previously are realized in water, the subsequent dehydration of
the crude mixture into HMF might remain problematic to limit hydration of HMF into
levulinic and formic acids [20]. As our goal was to design a process that could be industrially
implemented providing D-fructose source that could be directly used without any purification
in a HMF process, the use of organic solvents instead of water was a possibility. The only
study that had used organic solvents for the isomerization of D-glucose into D-fructose was
from Jung and co-workers [21]. They realized the isomerization in DMF in the presence of
Mg-Al hydrotalcites catalysts (32 wt % compared to D-glucose amount) at 80 °C for 3 h. D-
fructose was obtained with good selectivities but D-glucose conversions were lower than 50%
and decreased in trials carried out with the recycled catalyst. The presence of other
monosaccharides such as D-mannose was not examined; the selectivity of D-fructose could
therefore be overestimated. The productivity of the process was moderate because D-glucose
initial concentrations were very low (3 wt %). We didn’t explored isomerization in ionic
liquids due to limited D-glucose solubility under 70°C compared to solvents studied in our
work and potential viscosity problems for recovering catalyst by filtration [22]. Herein, we
report a process that enables good D-fructose selectivities in widely used organic solvents in
view to directly dehydrate the resulting carbohydrates solution into HMF. This process uses
high initial concentration of D-glucose and DMSO was selected as a component of choice in
solvents mixtures due to the high solubility of D-glucose at low temperature. DMSO is also
advantageous on the basis of former studies that revealed its positive role in non catalyzed
3

ACCEPTED MANUSCRIPT
dehydration processes of carbohydrates [23, 24]. Finally, its use allows efficient and selective
HMF recovery by liquid-liquid extraction as described by Dumesic’s team [25].
2
. Experimental
2
.1.Materials
D-fructose, D-glucose and D-mannose were purchased from Sigma Aldrich and used without
purification. Solvents (dimethylsulfoxide, glycerol, propylene glycol, ethylene glycol) were
purchased from Acros Organics and used without further purification. Sodium aluminate
(
NaAlO ) was purchased from VWR. Aqueous concentrated D-glucose solutions were from
2
Chamtor.
2
.2.Methods
General procedure for the isomerization of anhydrous D-glucose into D-fructose
A round bottom glass flask equipped with a magnetic bar was charged with D-glucose (10 g)
and solvents. After stirring at 55° C till dissolution of D-glucose, the catalyst was added and
the resulting mixture was stirred at 55° C for 3 h.
General procedure for the isomerization of D-glucose syrup into D-fructose
A round bottom glass flask equipped with a magnetic bar was charged with a syrup of D-
glucose (10 g, 70% dry matter) and solvents. After stirring at 55°C till dissolution of D-
glucose, the catalyst was added and resulting mixture was stirred at 55°C for 3 h.
Recycling experiments
At the end of the reaction, the mixture was centrifuged (4000 rev/min at 20° C) for 5 min. The
solution was removed and the catalyst was washed with DMSO and ethyl acetate in order to
eliminate traces of carbohydrates. Then it was dried under 1 mbar at 40 °C for 1 hour.
General procedure for the dehydration into HMF
A round bottom glass flask equipped with a magnetic bar was charged with the supernatant of
the isomerisation reaction and was stirred at 55°C under 100 mbar for 15 min. in order to
remove water. Appropriate amounts of DMSO and HCl 37% (pH ~2) were added to the
4

ACCEPTED MANUSCRIPT
mixture. This latter was heated to the desired temperature (150°C) and stirred at this
temperature for 1 hour.
Analytical methods
Carbohydrates amounts were determined by ion-exchange chromatography technique and
quantified thanks to calibration curves obtained using commercial internal standards (D-
fructose, D-glucose and D-mannose). The analyses were performed with Dionex HPLC
composed of a PDA detector, a P580 pump, an AS100 automatic injector and a STH 585
column oven. Carbohydrates were analyzed with CARBOPAC MA1 column and precolumn
(4*250 mm and 4*50 mm; eluents: A: 480 mM of NaOH (50 %) in water and B: 600 mM of
NaOH (50 %) in water; flow: 0.7 mL/min; oven temperature: 30°C).
ICP-AES procedure
ICP-AES (Inductively coupled plasma atomic emission spectroscopy) analyses were
performed by SGS Multilab. An aliquot of 0.5 g of the sample was mineralized with nitric
acid (10 ml) under heat. Dilution of the sample in deionised water (25 ml) was done before
injection. The analyses were performed with Varian 820-MS. The concentrations of the
elements were determined thanks to the calibration curves.
Glucose conversion and Fructose yield and selectivity were calculated using the following
equations:
Glucose Conversion (%) =
((mol of glucose initial – mol of glucose final) / mol of glucose initial) X 100
Fructose Selectivity (%) =
(mol of fructose produced / (mol of glucose initial – mol of glucose final)) X 100
5

ACCEPTED MANUSCRIPT
3
. Results and discussion
Inspired by former work [14], preliminary experiments have been carried out using
commercially available sodium aluminate. In water (Table 1, entry 1), sodium aluminate is an
efficient catalyst as 70% selectivity in D-fructose for 59% of D-glucose conversion were
obtained with a minimal loss of carbohydrates and low concentration of D-mannose.
Complementary analysis allowed to identify organic acids such as formic, lactic, glycolic
acids that have been described as degradation products in former studies [26]. Due to good
water solubility, it was impossible to recover the catalyst by filtration, confirming literature
results [14]. This problem was solved when the reaction was carried out in organic solvent
mixtures (Table 1, Entries 2-10). Three hours of reaction appeared to be the best compromise
for high D-fructose yield and minimal degradation of carbohydrates (see ESI for D-fructose
production as a function of time). This residence time is interesting in the view of designing
an intensive process at industrial level. After carrying out experiments with DMSO and
various DMSO-water mixtures (Table 1, entries 2-4) which led to good results, we finally
found that solvent mixtures composed of ethylene glycol (EG) or propylene glycol (PG) with
DMSO allowed high yields of D-fructose (Table 1, entries 5, 6 and 9). Even if it is not easy to
compare reactions that reached different D-glucose conversions, it seems that water and
glycols have an important impact on catalyst activity probably because they both favor proton
exchange (Table 1, Entries 1 and 7) unlike DMSO (Table 1, Entry 2). Noteworthy, when
reaction is carried out in the presence of water, the selectivity of D-fructose remain high
(Table 1, Entry 8). Finally, best result was obtained starting from commercially available D-
glucose concentrated syrup (72% selectivity in D-fructose - Table 1, entry 10). This latter
result is among the best reported D-fructose selectivities using such a high concentration of
D-glucose [21]. Here, the association of water, PG and DMSO is responsible for such high D-
fructose yields. In addition, the DMSO facilitates the catalyst filtration.
6

ACCEPTED MANUSCRIPT
a
Table 1. D-Glucose isomerization in various solvents by NaAlO catalyst.
2
Entry
Solvent mixture
wt/wt ratio)
D-Glucose
D-Fructose
D-Mannose
conversion (%) selectivity (%) selectivity (%)
(
1
2
3
4
5
6
7
8
Water
DMSO
59
47
62
63
46
60
68
65
60
68
70
33
59
44
51
50
40
50
47
72
3
1
2
1
2
3
3
3
3
3
DMSO/Water (2/8)
DMSO/Water (8/2)
DMSO/EG (7/3)
DMSO/EG (3/7)
EG
EG/Water (7/3)
DMSO/PG (3/7)
DMSO/PG/Water (2/5/3)
9
b
1
0
a
Reaction conditions: 40 wt % D-glucose, catalyst/D-glucose = 32 wt%, 55 °C, 3 h.
D-Glucose aqueous solution was used (70% dry matter)
b
%
D-glucose
% D-fructose
% D-mannose
40
35
30
25
20
15
10
5
0
Fig. 2. Screening of various catalysts for the isomerization of D-glucose in DMSO/PG/Water (2/5/3).
Reaction conditions: D-glucose syrup (40 wt%, 70% dry matter), catalyst/D-glucose = 70 mol%
(
molar equivalent of 32 wt% of NaAlO /D-glucose), 55 °C, 3 h.
2
In order to explain the singular behavior of sodium aluminate, complementary trials were
carried out on D-fructose and D-mannose in the condition described in Table 1, Entry 10 (see
7

ACCEPTED MANUSCRIPT
ESI for details). Besides the fact that concentrations of carbohydrates evolved as described in
former studies for D-fructose [9], they showed that D-glucose was converted more quickly
and selectively than D-fructose and D-mannose during the first hour of reaction. Moreover,
the difference with published results [9] lies in the fact that sodium aluminate also efficiently
converted D-mannose into D-fructose during the first 3 hours of reaction explaining the high
selectivity in D-fructose observed when D-glucose was used as substrate.
Under the same experimental conditions, the use of other soluble metallic acid catalysts,
soluble base catalysts or aluminum-based heterogeneous catalyst did not improved D-fructose
yield (Fig. 2). We rather observed significant carbohydrates degradation. It is also interesting
to point out the results of metallic acid catalysts that are usually used for direct conversion of
D-glucose into HMF and which are known to proceed through two steps isomerization-
dehydration reaction [27] and which under our experimental conditions are not efficient. In
the present study, we used a lower temperature and higher carbohydrate concentrations and
this can explain the poor results obtained with these metallic catalysts.
The results obtained with sodium aluminate led to higher yields than those described by Jung
and co-workers [21] and were obtained on more concentrated D-glucose organic solutions.
The catalyst recyclability was studied in the DMSO/PG solvent system. The presence of water
in D-glucose leads to a dramatic decrease of the selectivity (cf. Table 2, entries 1-3). As
sodium aluminate is water soluble, ICP-AES analyses (see ESI for analysis details) revealed
important leaching of aluminum (18 %) and sodium (30 %) cations in the filtrates. Use of
anhydrous D-glucose at the same concentration led to lower D-fructose selectivity with fresh
and recycled catalyst which is not surprising in view of the role of water displayed by results
of table 1. Due to agglomeration of the carbohydrates on the solid catalyst (Table 2, entries 4
and 5) the recovery of the catalyst became impossible after the second cycle. A lower D-
glucose concentration allowed catalyst recovery and afforded a good D-fructose selectivity
with minimum losses of the catalyst activity (cf. Table 2, entries 6-9). However, lowering the
initial carbohydrate concentration did not completely excluded carbohydrate agglomeration
phenomenon. The results of leaching of sodium especially (aluminium leaching remaining
low) indicated a probable modification of the catalyst composition and this is in favor with the
lower conversion obtained after the first cycle recycling (Table 2, entry 7). Further leaching of
sodium was then less problematic for activity of the catalyst in consecutive recycling (Table
2
, entries 8, 9). Although sodium continued to be transferred to the liquid phase at a low level
(4%), the process allows maintaining sufficient amount of aluminium in the recycled catalyst
for the next isomerization trials (see ESI for analysis details). Although the D-glucose
8

ACCEPTED MANUSCRIPT
concentration was lower than with aqueous D-glucose syrup, this later was still high
compared to those of the literature and the productivity of the isomerization remained
significant.
The mixture obtained from the isomerization process was then dehydrated into HMF. Yields
up to 82% were obtained (see ESI for details). The aluminium amount leached in the
carbohydrates mixture seems to have a beneficial impact on the dehydration process and its
role is till under investigation. However, the behavior of sodium aluminate in organic solvents
isomerization process urges us to report our results. Our present work is also focusing on the
recycling of the solvent mixture and the scale up of our method. Preliminary results have
already shown a high rate of solvent recycling and these results will be published later.
a
Table 2. Catalyst recycling for the isomerization reaction.
Entry
D-Glucose
Solvents
Cycle
D-Glucose
D-Fructose
D-Mannose
conc. (wt %)
conversion (%) selectivity (%) selectivity (%)
b
1
40
40
40
DMSO/PG/Water
1
2
3
68
61
59
72
16
17
3
3
3
(
2/5/3)
DMSO/PG/Water
2/5/3)
DMSO/PG/Water
2/5/3)
b
2
(
b
3
(
4
5
6
7
8
40
40
29
29
29
29
DMSO/PG (3/7)
DMSO/PG (3/7)
DMSO/PG (3/7)
DMSO/PG (3/7)
DMSO/PG (3/7)
DMSO/PG (3/7)
1
2
1
2
3
4
60
70
66
49
49
49
47
15
49
48
47
47
3
1
4
3
2
2
9
a
Reaction conditions: 29 or 40 wt% D-glucose in DMSO/PG, NaAlO /D-glucose = 32 wt%, 55°C, 3h.
2
b
D-Glucose aqueous solution was used (70% dry matter)
4
. Conclusions
The sodium aluminate-catalyzed isomerization of D-glucose into D-fructose can be carried
out with good selectivity and high yields using organic solvents and D-glucose at 70% dry
matter. The recyclability of the catalyst was however improved with anhydrous D-glucose.
9

ACCEPTED MANUSCRIPT
These results and the use of relatively high D-glucose concentrations are of potential
industrial interest. Moreover, the resulting crude carbohydrates mixtures can be efficiently
dehydrated into HMF.
Acknowledgments
The authors gratefully acknowledge financial support by the Champagne-Ardenne Region and
FEDER for this project (Furasynth program).
References
[
[
1] A. A. Rosatella, S. P. Simeonov, R. F. M. Frade and C. A. M. Afonso, Green Chem. 13
2011) 754-793
(
2] R.-J. van Putten, J. C. van der Waal, E. de Jong, C. B. Resrendra, H. J. Heeres, J. G. de
Vries, Chem. Rev. 113 (2013) 1499-1597
[
[
[
[
3] B. F. M. Kuster, Starch/Stärke 42 (1990) 314-321
4] R. L. Antrim, W. Colilla, B. J. Schnyder, Appl. Biochem. Bioeng. 2 (1979) 97-155
5] H. E. Vandam, A. P. G. Kieboom, H. Vanbekkum, Starch/Stärke 38 (1986) 95-101
6] C. A. Lobry de Bruyn, W. Alberda van Ekenstein, Rec. Trav. Chim. Pays-Bas 14 (1895)
2
03-216
7] C. A. Lobry de Bruyn, W. Alberda van Ekenstein, Rec. Trav. Chim. Pays-Bas 15 (1896)
2-96
8] S. J. Angyal, Topics Current Chem. 215 (2001) 1-14
[
9
[
[
9] H. S. El Khadem, S. Ennifar, H. S. Isbell, Carbohydr. Res. 169 (1987) 13-21
[
[
[
[
[
10] J. B. Binder, R. T. Raines, J. Am. Chem. Soc. 131 (2009) 1979-1985
11] H. Zhao, J. E. Holladay, H. Brown, Z. C. Zhang, Science 316 (2007) 1597-1600
12] G. Hong, Y. Zhang, J. Y. Ying, Angew. Chem. Int. Ed. 47 (2008) 9345-9348
13] C. Li, Z. Zhang, Z. K. Zhao, Tetrahedron Lett. 50 (2009) 5403-5405
14] F. Cornille, Preparation of ketose sugars by isomerisation of aldose sugars in the
presence of potassium aluminate, useful as low calorie-type sweeteners, Fr. Pat. (2003)
2
862 973
[
15] S. Lima, A. S. Dias, Z. Lin, P. Brandão, P. Ferreira, M. Pillinger, J. Rocha, V. Calvino-
Casilda, A. A. Valente, Appl. Catal. A-Gen. 339 (2008) 21-27
1
0

ACCEPTED MANUSCRIPT
[
16] M. Moliner, Y. Román-Leshkov, M. E. Davis, Proc. Natl. Acad. Sci. U. S. A. 107 (2010)
6
164-6168
[
[
17] C. Moreau, R. Durand, A. Roux, D. Tichit, Appl. Catal. A. 193 (2000) 257-264
18] J. Lecomte, A. Finiels, C. Moreau, Starch/Stärke 54 (2002) 75-79; C. Moreau, J.
Lecomte, A. Roux, Catal. Commun. 7 (2006) 941-944
[
19] R. O. L. Souza, D. P. Fabiano, C. Feche, F. Rataboul, D. Cardoso, N. Essayem, Catal.
Today 195 (2012) 114-119
[
[
[
20] M. J. Antal, W. S. L. Mok, Carbohydr. Res. 199 (1990) 91-109
21] S. Yu, E. Kim, S. Park, I. K. Song, J. C. Jung, Catal. Commun. 29 (2012) 63-67
22] L. Weina, H. Yucui, W. Weize, R. Shuhang, J. Yan, Z. Baogang, Ind. Eng. Chem. Res.,
5
0 (2011) 6952-6956
[
[
23] H. H. Szmant, D. D. Chundury, J. Chem. Technol. Biotechnol. 31 (1981) 135-145; R. M.
Musau, R. M. Munavu, Biomass 13 (1987) 67-74
24] B. Schneider, F. W. Lichtenthaler, G. Steinle, H. Schiweck, Liebigs Ann. Chem. (1985)
2
443-2453
[
[
[
25] J. N. Chheda, Y. Román-Leshkov, J. A. Dumesic, Green Chem. 9 (2007) 342-350
26] B. Y. Yang, R. Montgomery, Carbohydr. Res. 280 (1996) 27-45
27] H. Suqin, Z. Zhaofu, S. Jinliang, Z. Yinxi, H. Buxing, Green Chem. 11 (2009) 1746-
1
749
1
1

ACCEPTED MANUSCRIPT
Graphical abstract
1
2

ACCEPTED MANUSCRIPT
Highlights

The process designed for the isomerization of D-glucose into D-fructose used a
heterogeneous catalyst, in organic solvents.
NaAlO showed better activity than others tested catalysts and is recyclable without

2
decrease of D-fructose selectivity.
The resulting carbohydrates mixture could be directly dehydrated into HMF.

1
3