Dehydration of carbohydrates to 2-furaldehydes in ionic liquids by catalysis with ion exchange resins

Article abstract of DOI:10.1016/j.catcom.2012.07.002

The dehydration of several sugars, including pentoses, hexoses, di, tri, and polysaccharides, in ionic liquids with acidic ion-exchange resins as heterogeneous catalysts was investigated. Good 2-furaldehydes recovery yields, reaching 92% in some cases, were achieved when Dowex 50W ion-exchange resins and 1-n-butyl-3-methylimidazolium chloride ([C₄mim]Cl) were used. Our results show that this aproach could represent a promising route towards the cost-efficient production of 2-furaldehydes from carbohydrate-based feedstocks.

Full text of DOI:10.1016/j.catcom.2012.07.002

Catalysis Communications 27 (2012) 88–91
Contents lists available at SciVerse ScienceDirect
Catalysis Communications
journal homepage: www.elsevier.com/locate/catcom
Short Communication
Dehydration of carbohydrates to 2-furaldehydes in ionic liquids by catalysis with ion
exchange resins
a
a
a
Viviana Heguaburu a,b, , Jaime Franco , Luis Reina , Carlos Tabarez , Guillermo Moyna b,1, Patrick Moyna a
⁎
a
Departamento de Tecnología Química, Facultad de Química, Universidad de la República, Avenida General Flores 2124, Montevideo 11800, Uruguay
Departamento de Química del Litoral, Polo Agroalimentario y Agroindustrial de Paysandú, Universidad de la República, Ruta 3 km 363, Paysandú 60000, Uruguay
b
a r t i c l e i n f o
a b s t r a c t
Article history:
The dehydration of several sugars, including pentoses, hexoses, di, tri, and polysaccharides, in ionic liquids with
acidic ion-exchange resins as heterogeneous catalysts was investigated. Good 2‐furaldehydes recovery yields,
reaching 92% in some cases, were achieved when Dowex® 50W ion‐exchange resins and 1‐n‐butyl‐3‐
methylimidazolium chloride ([C4mim]Cl) were used. Our results show that this aproach could represent a prom-
ising route towards the cost-efficient production of 2-furaldehydes from carbohydrate-based feedstocks.
© 2012 Elsevier B.V. All rights reserved.
Received 12 April 2012
Received in revised form 27 June 2012
Accepted 3 July 2012
Available online 10 July 2012
Keywords:
Ion-exchange resins
Ionic liquids
Dehydration
2-Furaldehydes
Carbohydrates
1. Introduction
disposing of volatile organic compounds (VOCs). Given their ability to
dissolve cellulose and lignocellulosic materials, ILs have been proposed
Interest in the chemistry of biomass-derived compounds has led to
efforts aimed on the use of lignocellulosic materials as alternative
sources of chemicals and energy. The dissolution of sugars and their
dehydration into renewable feedstocks, such as 2-furaldehydes, are an
example. These compounds represent valuable building blocks that
could be obtained from biomass and used to synthesize a broad range
of value added compounds currently derived from non-renewable
sources [1,2].
5-(Hydroxymethyl)furfural (HMF) can be prepared by acid cata-
lyzed dehydration of hexoses (Fig. 1) [3]. Similarly, furfural can be
obtained by dehydration of pentoses. Both are produced industrially
from agricultural byproducts such as sugarcane bagasse and corn cobs
using methods that date back to the early 1900s. These processes have
relatively high costs, operate at less than 50% yield, use vast ammounts
of steam, and generate of large quantities of effluent waste [4,5]. These
acid dehydrations have been studied using a broad range of catalytic
systems and solvents. On the other hand, many catalysts, such as min-
eral and organic acids and salts, have been investigated using ionic
liquids (ILs) as reaction media [6–8]. ILs possess a number of advantages
over conventional solvents, including negligible vapor pressure,
non-flammability, and thermal stability [9]. These characteristics have
made them attractive to industries, as their use has the potential of alle-
viating regulatory constraints and manufacturing costs associated with
as media for the generation of biomass-based renewable energy
[10–12]. Consequently, they could play an important role in the develop-
ment of alternative and bio-renewable energy sources and feedstocks, as
well as making the traditional routes cleaner and more efficient. The use
of heterogeneous catalysts for these reactions has been limited to
ion-exchange resins [13–16] and zeolites [17–19], and only a few report
using ILs as media [20–25]. In this communication we present the dehy-
dration of sugars and their conversion into HMF and related compounds
using ILs as solvents and several heterogeneous catalysts. We initially
surveyed a number of catalysts for the dehydration of fructose, glucose,
arabinose, and xylose. The results obtained with ion-exchange resins
were particularly encouraging, and the study was thus extended to a
broader series of mono, di, tri, and polysaccharides as models to evaluate
the reaction on carbohydrate-based biomass. Additionally, chloride and
bromide salts were added in certain cases, as the anions in these salts
could enhance the solvation of the carbohydrates in the IL [26], as well
as to promote the aldose to ketose isomerization process [27].
2. Experimental
The following catalysts were assayed: 3 Å, 4 Å, 5 Å, and 13X molec-
ular sieves, Montmorillonite, Montmorillonite K10, zeolites ZSM-5 and
HZSM-5 (prepared according to Zhang et al. [28]), Amberlyst® 15,
Amberlite® IR120, Dowex® 50Wx4-50, Dowex® 50Wx8-100, and
Dowex® 50Wx8-200. Dehydrations were performed using sealed
tubes in concentrations of 10% (w/w) for both the carbohydrate and
the heterogeneous catalyst, using 3 mL of [C4mim]Cl and keeping the
⁎
Corresponding author. Tel.: +598 2924 8194; fax: +598 2924 1906.
E-mail address: vheguab@fq.edu.uy (V. Heguaburu).
Tel./fax: +598 4722 7950.
1
1566-7367/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.catcom.2012.07.002

V. Heguaburu et al. / Catalysis Communications 27 (2012) 88–91
89
glucose and fructose, respectively. Dowex® resins in several forms
showed better HMF yields from fructose, ranging from 44 to 60%.
Glucose dehydration reactions showed large differences. Dowex®
50W-1x4 resin resulted in a 3% HMF yield, while under the same con-
ditions the yield from fructose was 54%. Dowex® 50Wx8-100 and
200 showed moderate HMF yields from glucose (34 and 44 % HMF
yields, respectively), while yields from fructose were higher when
these resins were used (44 and 60%, respectively).
Fig. 1. Dissolution and dehydration of sugars to 2-furaldehydes.
Given the results using CrCl3 with no catalyst, the addition of salts
to the resin trials was evaluated. However, the results were erratic.
Addition of CrCl3 to Amberlyst® 15 resin lowered the HMF yield in
the case of glucose (21%), but slightly increased it when fructose
was used as substrate (39%). In the case Dowex® 50Wx8-200, HMF
recovery yields were lower for fructose (44%) and glucose (33%).
Addition of lithium salts to the mixtures gave yields of HMF similar
to the original reaction. HMF yields from fructose decreased to 46%
when LiCl was added, and was 60% when LiBr was added. A similar
profile was observed for glucose, which showed HMF yields of 39
and 50% with LiCl and LiBr, respectively.
temperature at 100 °C with continuous stirring for 6 hrs. In specific
experiments, 1% of different salt co-catalysts reported in the literature,
including CrCl3 LiBr, and LiCl, were used. The reaction mixtures were
then diluted with water (10 mL), and extracted with ethyl acetate
(4×10 mL). The combined organic extracts were dried with Na2SO4, fil-
tered, and the solvent evaporated under reduced pressure to obtain the
corresponding 2-furaldehyde as the only product for the cases where
the reaction took place.
The identity of the products was confirmed by 1H NMR analyses.
These were performed on a Bruker AVANCE 400 spectrometer operat-
ing at 1H frequency of 400.13 MHz using CDCl3 as solvent. Chemical
shifts (δ) are in ppm relative to the residual solvent signal
(7.28 ppm), and coupling constants (J) are reported in Hz. Furfural:
δ 9.70 (s, 1H), 7.72 (dd, J=0.8, J=1.7, 1H), 7.26 (dd, J=1.7, J=3.6,
1H), 6.63 (dd, J=1.7, J=3.6, 1H). Methylfurfural: δ 9.52 (s, 1H), 7.20
(d, J=3.6, 1H), 6.26 (qd, J=0.6, J=3.6, 1H), 2.44 (d, J=0.6, 3H).
5-(Hydroxymethyl)furfural: δ 9.48 (s, 1H), 7.20 (d, J=3.5, 1H), 6.48
(d, J=3.5, 1H), 4.64 (s, 2H), 4.03 (bs, 1H).
3.2. Effects of temperature, reaction time, and catalyst recycling on HMF
yields
Having established that Dowex® 50Wx8-200 gave the best results,
the yield profiles with varying temperature and time were studied to
establish the optimal operating conditions. These experiments were
carried using glucose as a standard and in the same conditions as the
initial trials (3 mL of [C4mim]Cl, 10% sugar and catalyst, 6 h). The yields
obtained with different temperatures are presented in Fig. 3.
3. Results and discussion
At 50 °C, dissolution was incomplete and only trace amounts of HMF
could be detected. At 70 and 80 °C clean solutions were obtained, indi-
cating that polymerization was controlled. However, yields of only
8 and 35%, respectively, were obtained at these temperatures. At 90 °C
the yield increased to 50%, comparable to the 44% yield obtained at
100 °C. Raising the temperature to 110 °C resulted in a HMF yield of
43%. An intense browning of the reaction mixtures was also observed
at this temperature, suggesting that polymerization had taken place.
Time profiles were established at a temperature of 100 °C on the
basis of the above mentioned results (Fig. 4). For glucose, a HMF yield
of 11% was reached in the first hour, increasing to 21% in 2 hours, 53%
at 3 hours and to 70% at 4 hours. A slight decrease to 67% was observed
after 5 hours and at 6 hours the yield was down to 44%. At the same
time, the darkening of the reaction mixture was intense, suggesting
once again that polymerization reactions were occurring. Therefore,
reaction times longer than 4 hours should be avoided.
3.1. Production of HMF from fructose and glucose with different catalysts
The results obtained for the dehydrations of fructose and glucose
into HMF with the different catalysts are summarized in Fig. 2. No de-
hydration products were observed with no added catalyst. Addition
of CrCl3 to the neat IL reaction mixture yielded HMF in 47% in the
case of fructose and 26% in the case of glucose. The results using mo-
lecular sieves or zeolites were not encouraging. Except when a molec-
ular sieve of 5 Å pore size was used with fructose, which yielded HMF
in 4%, these slightly acidic catalysts resulted in no dehydration
products. Using ZSM-5 and its acidic form gave higher yields with
fructose (14 and 20% yields respectively), but poor yields with glu-
cose (3% in both cases). Montmorillonite clay in its natural and acidic
form (K10) showed low yields of HMF (3% or less). Acidic ion-
exchange resins showed better results. Although using Amberlite®
IR120 resulted in no production of HMF with either monosaccharide,
Amberlyst® 15 catalysis afforded HMF in 61 and 34% yields from
The recyclability of the resin was then evaluated. Based on the
data obtained as a function of temperature and time, 3 hour reaction
cycles at 100 °C were used for these studies. In the first trial, the
Fig. 2. Dehydration of fructose and glucose in IL solution with different heterogeneous catalysts.

90
V. Heguaburu et al. / Catalysis Communications 27 (2012) 88–91
3.3. Production of furfural from xylose and arabinose with different catalysts
Two pentoses, xylose and arabinose, were tested with the same
set of heterogeneous catalysts and reaction conditions used for the
hexoses. In this case, a positive reaction resulted in the production
of furfural. Fig. 5 summarizes the results obtained.
As it was observed previously with the hexoses, there was no
reaction in the neat IL. Addition of 1% CrCl3 generated furfural in 33
and 30% yields for xylose and arabinose, respectively. Treatment with
molecular sieves and zeolites ZSM-5 and HZSM-5 did not produce furfu-
ral. The same held for natural Montmorillonite and K10 clay. In the case
of resins, Amberlite® IR120 showed no activity, while Amberlyst® 15
produced furfural in yields of 17% with xylose and 6% with arabinose.
The addition of CrCl3 to the Amberlyst® 15 reaction mixtures raised
the yields to 19 and 29%, respectively. Dowex® resins showed good re-
sults. Using Dowex® 50Wx8-100, a 92% yield of furfural from arabinose
was achieved. With Dowex® 50Wx8-200, the yield attained was 75%.
Addition of CrCl3 to the latter mixture lowered the yields (50%), as did
the use of lithium salts (44% for LiCl and 23% for LiBr). Xylose showed
lower yields than arabinose, producing furfural in 59% yield when
Dowex® 50Wx8-100 was used, and only 28% for Dowex® 50Wx8-
200. Again, addition of CrCl3 to the mixture containing the 200-mesh
resin lowered the yield to 8%, while LiCl and LiBr raised the yields to
56 and 68% respectively.
Fig. 3. HMF yields as a function of temperature using Dowex® 50Wx8-200 as catalyst
and glucose as substrate.
Despite the distinct behavior of the four tested carbohydrates,
acidic ion-exchange resins such as Dowex® 50Wx8 and Amberlyst®
15 show the most promising results overall. In some cases the yields
can be modulated by the addition of salts to the mixture.
Fig. 4. HMF yields as a function of reaction time using Dowex® 50Wx8-200 as catalyst
3.4. Dehydration of other carbohydrates using Dowex® 50Wx8-200
ion-exchange resin
and glucose as substrate.
Trials with Dowex® 50Wx8-200 were extended to several types of
carbohydrates and these results are presented in Fig. 6. These exper-
iments were carried out with an adapted set of parameters, using
3 mL of [C4mim]Cl, the appropriate carbohydrate and Dowex®
50Wx8-200 in 10% concentration, and heating for 3 hours at 100 °C.
In these conditions, ribose showed the highest yield of furfural
among pentoses, reaching 90%. Arabinose lyxose, and xylose yielded fur-
fural in 49%, 75%, and 28%, respectively. Hexoses displayed different pro-
files of reactivity, with galactose showing the lowest HMF yield (17%).
Sorbose and mannose showed moderate yields (48 and 41%, respective-
ly). As shown above, fructose and glucose generated HMF in 60 and 53%
yields, respectively. Rhamnose, a 6-deoxyhexose, was dehydrated in
these conditions to methylfurfural in 63% yield. Dehydration of the
disaccharides lactose, trehalose, maltose, and sucrose produced HMF in
catalyst was recovered, washed with acetone, air dried, and employed
in the second reaction cycle. A drop of 90% in the yield of HMF with
respect to the first catalytic cycle was observed in this case. In a sec-
ond trial, the resin was regenerated between cycles as indicated by
the manufacturer (i.e., washed with 10% H2SO4 and brought to pH 6
with distilled water), washed with acetone, and air dried. Despite
the resin acquired a dark brown color, the drop in the yield of HMF
in the second cycle was only 16% in this case. However, regenerating
the resing for a second time and using it in a third cycle leads to a
drop in HMF yield of 92%. These results indicate that the acid sites
of the resin require regeneration between catalytic cycles, and also
suggest that gradual poisoning with decomposition products takes
place.
Fig. 5. Dehydration of xylose and arabinose in IL solution with different heterogeneous catalysts.

V. Heguaburu et al. / Catalysis Communications 27 (2012) 88–91
91
Fig. 6. Dehydration of carbohydrates with Dowex® 50Wx8-200 ion exchange resin.
38, 51, 49, and 71% yields, respectively. It can be pointed out that su-
Acknowledgements
crose, the only disaccharide tested containing a fructose unit, showed
a yield similar to that of the monomer. The trisaccharide raffinose pro-
duced HMF in 47% yield. In the case of polysaccharides, inulin (a fructan)
reached a 67% yield, while microcrystalline cellulose resulted in low
yields (13%).
The authors wish to acknowledge the financial support of the Fondo
Sectorial de Energía (FSE-ANII) through project ANII-PR-FSE-2009-
1-06, and the assistance of the Programa de Desarrollo de las Ciencias
Basicas (PEDECIBA) and Facultad de Química for contributions to the
research project.
4. Conclusions
References
[1] F. Peng, J.L. Ren, F. Xu, R.C. Sun, In: Sustainable production of fuels, chemicals, and
fibers from forest biomass, ACS Symposium Series, 1067, 2011, p. 219.
[2] A.S. Mamman, J.M. Lee, Y.C. Kim, I.T. Hwang, N.J. Park, Y.K. Hwang, J.S. Chang, J.S.
Hwang, Biofuels, Bioproducts and Biorefining 2 (2008) 438.
[3] F.H. Newth, Advances in Carbohydrate Chemistry 6 (1951) 83.
[4] E. Erdmann, Berichte der deutschen chemischen Gesellschaft 43 (1910) 2391.
[5] B.F.M. Kuster, Starch 42 (1990) 314.
[6] M.E. Zakrzewska, E. Bogel-Lukasik, R. Bogel-Lukasik, Energy and Fuels 24 (2010)
737.
[7] M.E. Zakrzewska, E. Bogel-Lukasik, R. Bogel-Lukasik, Chemical Reviews 111
(2011) 397.
[8] T. Stahlberg, W. Fu, J.M. Woodley, A. Riisager, ChemSusChem 4 (2011) 451.
[9] J.P. Hallett, T. Welton, Chemical Reviews 111 (2011) 3508.
[10] A. Pinkert, K.N. Marsh, S. Pang, M.P. Staiger, Chemical Reviews 109 (2009) 6712.
[11] R.P. Swatloski, S.K. Spear, J.D. Holbrey, R.D. Rogers, Journal of the American Chemical
Society 124 (2002) 4974.
[12] D.A. Fort, R.C. Remsing, R.P. Swatloski, P. Moyna, G. Moyna, R.D. Rogers, Green
Chemistry 9 (2007) 63.
[13] X. Qi, M. Watanabe, T.M. Alda, R.L. Smith, Industrial and Engineering Chemistry
Research 47 (2008) 9234.
The acid degradation of sugars to 2-furaldehydes in ILs with several
heterogeneous catalysts was surveyed. Zeolites displayed poor or negli-
gible dehydration yields, probably due to their relatively low acid
strength. On the other hand, ion-exchange resins such as Dowex®
50W showed the best yields. The highest furfural yield was found to
be 90% in the case of ribose, and for pentoses the order of reactivity
was ribose>arabinose>lyxose>xylose. These results could be explained
by the presumed open chain mechanism postulated for the dehydration
of pentoses [29], in which ribose and arabinose share the same ketose
form (ribulose), both having the highest furfural yields. In addition, xylose
and lyxose dehydrations may proceed via their ketose form (xylulose),
and therefore presenting lower dehydration yields. If the dehydrations
proceed from structures with a furanose ring, the slight yield differences
of ribose and arabinose on the one hand, and xylose and lyxose on the
other could be better explained [30]. These results are not yet conclusive,
and NMR experiments are currently underway to decipher the dehydra-
tion mechanism. In the case of hexoses, no correlation between steric fac-
tors and observed reactivities could be found for the furanose forms, being
these data consistent with an open-chain form mechanism. The addition
of salts can have an influence on the yields, and these results also imply a
further evaluation of the alternative mechanisms suggested so far [29,30].
In the case of lithium salts, the hexoses evaluated showed little differ-
ences, but there were sharp decreases for arabinose and sharp increases
for xylose. CrCl3 seems to somehow disrupt the action of the resin catalyst
as the yields are lower in all cases.
[14] X. Qi, M. Watanabe, T.M. Aida, R.L. Smith, Green Chemistry 10 (2008) 799.
[15] Y. Nakamura, S. Morikawa, Bulletin of the Chemical Society of Japan 53 (1980)
3705.
[16] Y. Román-Leshkov, J.N. Chheda, J.A. Dumesic, Science 2006 (1933) 312.
[17] M. Moliner, Y. Román-Leshkov, M.E. Davis, Proceedings of the National Academy
of Sciences 107 (2010) 6164.
[18] T.R. Carlson, J. Jae, Y.C. Lin, G.A. Tompsett, G.W. Huber, Journal of Catalysis 270
(2010) 110.
[19] C. Moreau, R. Durand, S. Razigade, J. Duhamet, P. Faugeras, P. Rivalier, P. Ros,
G. Avignon, Applied Catalysis A: General 145 (1996) 211.
[20] M. Tan, L. Zhao, Y. Zhang, Biomass and Bioenergy 35 (2011) 1367.
[21] P. Wang, H. Yu, S. Zhan, S. Wang, Bioresource Technology 102 (2011) 4179.
[22] X. Qi, M. Watanabe, T.M. Aida, R.L. Smith, Green Chemistry 11 (2009) 1327.
[23] C. Lansalot-Matras, C. Moreau, Catalysis Communications 4 (2003) 517.
[24] X. Qi, M. Watanabe, T.M. Aida, R.L. Smith, ChemSusChem 2 (2009) 944.
[25] X. Qi, M. Watanabe, T.M. Aida, J.R.L. Smith, Green Chemistry 2010 (1855) 12.
[26] R.C. Remsing, R.P. Swatloski, R.D. Rogers, G. Moyna, Chemical Communications
(2006) 1271.
Although furfural and HMF are produced industrially by using high
temperatures and strong acid catalysts, this work shows that ion-
exchange resins in ILs allows for their production from sugars at
moderate temperatures. This process offers an interesting route to 2-
furaldehydes, and could thus be considered as a cost-efficient method
for the conversion of carbohydrate-based biomass into materials of
industrial interest.
[27] H. Zhao, J.E. Holladay, H. Brown, Z.C. Zhang, Science 316 (2007) 1597.
[28] W. Zhang, K. Wang, Y. Yu, H. He, Chemical Engineering Journal 163 (2010) 62.
[29] J.B. Binder, J.J. Blank, A.V. Cefali, R.T. Raines, ChemSusChem 3 (2010) 1268.
[30] E.R. Garrett, B.H. Dvorchik, Journal of Pharmaceutical Sciences 58 (1969) 813.