Organic Carbonates: Efficient Extraction Solvents for the Synthesis of HMF in Aqueous Media with Cerium Phosphates as Catalysts

Article abstract of DOI:10.1002/cssc.201501181

We describe a process for the selective conversion of C₆-polyols into 5-hydroxymethylfurfural (5-HMF) in biphasic systems of organic carbonate/water (OC/W), with cerium(IV) phosphates as catalysts. Different reaction parameters such as the OC/W ratio, catalyst loading, reaction time, and temperature, were investigated for the dehydration of fructose. Under the best reaction conditions, a yield of 67.7 % with a selectivity of 93.2 % was achieved at 423 K after 6 h of reaction using [(Ce(PO₄)_1.5(H₂O)(H₃O)_0.5(H₂O)_0.5)] as the catalyst. A maximum yield of 70 % with the same selectivity was achieved after 12 h. At the end of the reaction, the catalyst was removed by centrifugation, the organic phase was separated from water and evaporated in vacuo (with solvent recovery), and solid 5-HMF was isolated (purity >99 %). The recovery and reuse of the catalyst and the relationship between the structure of the OC and the efficiency of the extraction are discussed. The OC/W system influences the lifetime of the catalysts positively compared to only water.

Full text of DOI:10.1002/cssc.201501181

DOI: 10.1002/cssc.201501181
Full Papers
Organic Carbonates: Efficient Extraction Solvents for the
Synthesis of HMF in Aqueous Media with Cerium
Phosphates as Catalysts
Angela Dibenedetto,*[a, b] Michele Aresta,[b, c] Luigi di Bitonto,[b] and Carlo Pastore[b]
We describe a process for the selective conversion of C6-poly-
ols into 5-hydroxymethylfurfural (5-HMF) in biphasic systems of
organic carbonate/water (OC/W), with cerium(IV) phosphates
as catalysts. Different reaction parameters such as the OC/W
ratio, catalyst loading, reaction time, and temperature, were in-
vestigated for the dehydration of fructose. Under the best re-
action conditions, a yield of 67.7% with a selectivity of 93.2%
was achieved at 423 K after 6 h of reaction using [(Ce(-
PO4)1.5(H2O)(H3O)0.5(H2O)0.5)] as the catalyst. A maximum yield of
70% with the same selectivity was achieved after 12 h. At the
end of the reaction, the catalyst was removed by centrifuga-
tion, the organic phase was separated from water and evapo-
rated in vacuo (with solvent recovery), and solid 5-HMF was
isolated (purity >99%). The recovery and reuse of the catalyst
and the relationship between the structure of the OC and the
efficiency of the extraction are discussed. The OC/W system in-
fluences the lifetime of the catalysts positively compared to
only water.
Introduction
The rapid development of the world economy and the grow-
ing use of fossil fuels (oil, coal, and natural gas) are boosting
the search for sustainable alternative sources for both the
energy and chemical industries.[1] New chemistry is growing
based on the development of innovative catalytic reactions
that are able to convert renewable second-generation feed-
stock such as lignocellulosic biomass, almost ubiquitously,[2]
into chemicals and fuels that can replace those obtained cur-
rently from fossil carbon.[3] Cellulose can be converted into
a series of furan-based compounds that can be used as plat-
form molecules for the synthesis of chemicals, materials, and
fuels.[4] One of these, 5-hydroxymethylfurfural (5-HMF)
(Scheme 1)[5] is a versatile intermediate obtained from glucose,
which is first isomerized into fructose using enzymes or basic
catalysts and eventually dehydrated to 5-HMF under acid con-
ditions. Furan derivates and nonfuran compounds (Scheme 1)
are used as monomers for the synthesis of polymeric materials,
macrocyclic compounds, and other intermediates.
Scheme 1. 5-HMF: a key intermediate for the production of chemicals.
The dehydration of fructose to 5-HMF has been performed
in various reaction media in the presence of catalysts such as
mineral acids,[6] ion-exchange resins,[7] H-form zeolites,[8] and
metal phosphates.[9,10] The use of water as a solvent has been
attempted with mixed success, the use of water, which reduces
costs with respect to organic solvents, increases the problems
of the rehydration of 5-HMF to levulinic and formic acids and
polycondensation to humins, which reduces the selectivity to-
wards the target product. Moreover, mineral acids show a low
selectivity towards 5-HMF in aqueous media (55–63%) because
of the formation of humins and other secondary products. Ex-
change resins limit the formation of byproducts through the
selective absorption of 5-HMF (selectivity=70–80%), but 5-
HMF must be extracted subsequently with organic solvents
(DMSO[11] and methyl isobutyl ketone (MIBK)[12] have been used
so far). The addition of metal chlorides (CrIII or AlIII)[13] and the
use of ionic liquids (ILs)[14] or DMSO[15] as cosolvents improve
[a] Prof. A. Dibenedetto
Department of Chemistry
University of Bari
Via Orabona n. 4, 70126 Bari (Italy)
E-mail: angela.dibenedetto@uniba.it
[b] Prof. A. Dibenedetto, Prof. M. Aresta, Dr. L. di Bitonto, Dr. C. Pastore+
CIRCC
Via Celso Ulpiani n. 27, 70126 Bari (Italy)
[c] Prof. M. Aresta
Department of Chemical and Biomolecular Engineering
NUS
4 Engineering Drive 4, Singapore 117585 (Singapore)
[+] Present address:
Water Research Institute (IRSA-CNR)
National Research Council
Via de Blasio 5, 70132 Bari (Italy)
ChemSusChem 2016, 9, 118 – 125
118
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
the selectivity towards 5-HMF (up to 85%), but the isolation of
the target product is not so simple. If ILs (the application of
which on a large scale is questioned for their potential toxicity)
are used, 5-HMF has been extracted using a low-boiling-point
solvent, such as dichloromethane (DCM),[16,17] ethyl acetate
(EA),[18] THF,[19] diethyl ether,[20] and acetone.[21] Moreover, the
hydrophilic character of 5-HMF makes its extraction from the
aqueous phase difficult.[22] DMSO[23] and MIBK[24] have found
application but have high boiling points (462 and 389 K, re-
spectively) and this makes the final recovery of 5-HMF quite
complex. However, although several improvements have been
achieved in this field in the last few years, more efficient sepa-
ration techniques must be developed to make the production
of 5-HMF economically viable for large-scale production.
Linear or cyclic organic carbonates (OCs) have a low toxicity
and a low environmental impact and are used widely as sol-
vents,[25] monomers in the polymer industry,[26] for the produc-
tion of other carbonates,[27] as alkylating or carboxylating
agents, and in the agrochemical and pharmaceutical indus-
tries.[28] Nevertheless, they have never been used as extracting
agents in the production of 5-HMF. In this work, we present
a new application of dimethyl carbonate (DMC) and its conge-
ners as extracting solvents in the dehydration of fructose for
the synthesis of 5-HMF in an aqueous medium with cerium(IV)
phosphates as catalysts.[10] We have investigated the influence
of various process parameters, such as the volume ratio of or-
ganic carbonate to water (OC/W), the catalyst loading, the re-
action time, and the temperature, to find the optimal reaction
conditions and maximize the production of 5-HMF, which was
isolated easily at the end of the reaction as a >99% pure
solid.
Scheme 2. Synthesis of 5-HMF starting from glucose isomerized into fruc-
tose.
boiling point and the reactive nature of 5-HMF at high temper-
atures, the recovery of the target product is very difficult.
Recently, we reported the catalytic activity of cerium(IV) phos-
phates in the dehydration of fructose in aqueous media[10] and
demonstrated that they are active in the production of 5-HMF
with a yield of 43–52% (T=423 K). We have also shown that
after 6 h of reaction the selectivity towards 5-HMF decreases
from 79 to 68% because of the production of humins and
other secondary products. As a result of their limited solubility
in water and low environmental impact,[25] OCs find industrial
application in several fields and, as we show here, can be used
for the extraction of 5-HMF from the aqueous phase during
the reaction. The effect of different volume ratios of DMC/
water in the dehydration of fructose using cerium(IV) phos-
phates as catalysts is presented in Figure 1.
With an increase of the amount of DMC (up to DMC/water=
3), an increase of selectivity is observed in the production of
HMF for all catalysts tested (from a maximum of 79% in only
Results and Discussion
Effect of the volume ratio of DMC/water in the dehydration
of fructose
As discussed in the Introduction, 5-HMF is synthesized mainly
through the acid-catalyzed dehydration of fructose (obtained
from the isomerization of glucose) that loses three molecules
of water (Scheme 2).
Disaccharides (sucrose[29] and cellobiose[30]), polysaccharides
(inulin[31]), and even cellulose itself[32] can be used as starting
materials, the latter after depolymerization. Extensive research
is currently underway to develop efficient, cheap, and sustain-
able commercial methods for the preparation of 5-HMF.
As mentioned above, the use of water as a reaction
medium, although economically viable, increases issues such
as the low selectivity towards 5-HMF, caused by either the re-
hydration of the latter to afford levulinic and formic acid or the
polycondensation of 5-HMF to give humins (Scheme 1). To find
a solution to such problems, alternative approaches have been
investigated, namely, the extraction of 5-HMF from water with
organic/water biphasic systems or the direct use of organic sol-
vents as reaction media. DMSO[23] is used widely as a reaction
solvent for the dehydration of fructose to give a selectivity
around 85%, but, as mentioned before, because of its high
Figure 1. Effect of the different DMC/water volume ratios in the dehydration
of fructose using CeP1 (top), CeP2 (middle), and CeP3 (bottom) as catalysts.
Reaction conditions: fructose=0.64 g, catalyst=0.1 g, total volume of sol-
vent=12.8 mL, T=423 K, t=3 h.
ChemSusChem 2016, 9, 118 – 125
www.chemsuschem.org
119
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
water to over 90%). The continuous extraction of 5-HMF from
the aqueous phase minimizes the time it spends in the protic
solvent and, thus, its degradation/polymerization. In addition,
because of the limited miscibility of DMC and water, the furan-
ic compound is extracted without any cross-contamination.
Notably, although the solubility of DMC in water increases
with temperature (12.7% at 293 K and 14.3% at 353 K[33]) we
have ascertained (by using a glass apparatus) that two liquid
phases exist at 395 K. We have ascertained that the system is
still biphasic at the reaction temperature (423 K) by using an
autoclave with a sapphire window that allows us to observe
the reaction system. The miscibility of the OC with water plays
a key role in the transfer of 5-HMF: the lower the miscibility,
the more difficult the interphase transfer (see data for the
other OCs). An increase of the OC/W ratio to 3 with a constant
total volume decreases the conversion yield (Figure 1). Thus,
water plays an important role in the conversion of polyols and
its content should be controlled carefully to maximize the solu-
bilization of fructose, which increases the conversion rate and
minimizes the negative effects and loss in selectivity. DMC
alone is a poorer solvent than water for fructose and this does
not favor dehydration.
Figure 3. Effect of the different catalyst/fructose weight ratio in the dehydra-
tion of fructose using CeP3 as the catalyst. Reaction conditions: fructo-
se=0.64 g, DMC=9.6 mL, water=3.2 mL (volume ratio DMC/water=3),
T=423 K, t=3 h.
Figure 2). This further demonstrates that the permanence of 5-
HMF in water causes its decomposition.
Effect of the catalyst loading and the reaction time
The effect of the different catalyst/fructose weight ratios in the
dehydration process in the synthesis of 5-HMF using CeP3 as
the catalyst (which is the most active catalyst and has been
used in all following tests) is shown in Figure 3. An increase of
the amount of catalyst from 15.6 to 35% w/w (catalyst/fruc-
tose) led to an increase in the production of 5-HMF (from 26.6
to 52.3%) with the same selectivity (>90%). A higher dehydra-
tion rate with the increase of the catalyst loading may be at-
tributed to the availability, nature, and number of active
sites[10] that promote the production of 5-HMF, which is not re-
hydrated as it is extracted into DMC.
We have also performed an experiment in which we kept
the amount of fructose (0.64 g) and temperature (423 K) con-
stant, which led the same conversion of fructose (40Æ2%) in
the same volume of water (10 mL), and different amounts of
DMC (0, 1, 2.5, 5, 10 mL) were added. We kept the volume of
water constant because its decrease might affect the rate of
conversion of fructose. Such conditions cannot be compared
with those used in the experiments shown in Figure 1, in
which a constant total volume (W+OC) was used (12.8 mL)
and the reactions were run for 3 h. With this experiment, we
aimed to demonstrate the effect of DMC on the extraction of
5-HMF at almost the same conversion of fructose. We found
that the extraction of 5-HMF into DMC improves the selectivity
(Figure 2).
Notably, a further increase of the catalyst/substrate ratio
does not improve the conversion much. An increase of the cat-
alyst loading from 32 to 50% w/w catalyst/substrate increased
the yield of 5-HMF by only 1.9% from 52.3 to 54.2%.
The reaction appears to be an equilibrium, in which in the
biphasic system is shifted to the right because the product is
extracted into the organic phase. Notably, water alone can cat-
alyze the dehydration of fructose at 423 K but the conversion
rate and yield are much lower than those observed in the pres-
ence of the catalyst.
Notably, if DMC is added in an amount such that it may give
rise to a single phase (below its solubility in water, second
point in Figure 2), the benefit of its presence is not observed
as the same selectivity is found as in water (73%, first point in
We have proved that the yield of 5-HMF is only 9.3% (com-
pared to 52.3% with CeP3) and the selectivity towards 5-HMF
decreases to 45.4% (from >93%). Therefore, the role of the
catalyst is relevant as well as the role of the organic solvent.
The effect of the reaction time on the dehydration process
was also investigated. Reaction times were set at 3, 6, 9, and
12 h (Figure 4).
The reaction time has a remarkable effect on the conversion
of fructose into 5-HMF. The conversion of fructose increases
sharply up to 6 h reaction time, after which the 5-HMF yield is
67.7% with a selectivity of 93.2%. Interestingly, under such re-
action conditions, only small amounts of levulinic acid and glu-
cose were formed (~1%). This confirms that 5-HMF produced
during the reaction is extracted selectively from the aqueous
medium to avoid secondary reactions (isomerization, conden-
sation, and cross-polymerization[22]) that lead to the formation
Figure 2. Effect of the use of DMC as the extraction solvent at a constant
water volume and a constant conversion of fructose at 40Æ2%. An increase
of selectivity is observed.
ChemSusChem 2016, 9, 118 – 125
www.chemsuschem.org
120
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Figure 4. Effect of the reaction time in the dehydration of fructose using
CeP3 as the catalyst. Reaction conditions: fructose=0.64 g, catalyst=0.224 g
(weight ratio catalyst/fructose=35%), DMC=9.6 mL, water=3.2 mL
(volume ratio DMC/water=3), T=423 K.
Figure 6. 1H NMR spectrum of 5-HMF isolated from the dehydration of fruc-
tose (299 K, CDCl3).
of humins and byproducts. After 6 h there is no substantial in-
crease of the yield of 5-HMF, which is 70% after 12 h. A longer
reaction time does not improve the yield as the decomposition
of 5-HMF may increase with time.
Effect of the reaction temperature
The effect of the reaction temperature on the yield of 5-HMF is
shown in Figure 5. The reaction temperature affects the dehy-
dration of fructose significantly. If the dehydration of fructose
was performed at 353 K, the reaction was very slow. Only 6.4%
yield of HMF was obtained after 6 h of reaction.
An increase the reaction temperature from 353 to 423 K led
to a rapid increase of the production of 5-HMF with a maxi-
mum yield of 67.7% at 423 K (6 h, selectivity=93.2% in both
cases). A further increase of the temperature causes a larger
amount of humins with the formation of dark solids that sepa-
rate at the bottom of the reactor.
Figure 7. 13C NMR spectrum of 5-HMF isolated from the dehydration of fruc-
tose (299 K, CDCl3).
The NMR data obtained show clearly that the sample of 5-
HMF produced during the dehydration process is quite pure
(purity >99%) as the presence of secondary products is ex-
cluded.
Effect the nature of the OC on 5-HMF yield and selectivity
We have used several OCs, DMC, diethyl carbonate (DEC), 1,2-
propene carbonate (1,2-PC), and diallyl carbonate (DAlC), to es-
tablish if the molecular structure of the OC plays a role in the
production of 5-HMF in an aqueous medium. The effect of the
organic phase in the dehydration of fructose using CeP3 as the
catalyst is shown in Figure 8.
Figure 5. Effect of the reaction temperature in the dehydration of fructose
using CeP3 as the catalyst. Reaction conditions: fructose=0.64 g, cata-
lyst=0.224 g (weight ratio catalyst/fructose=35%), water=9.6 mL,
DMC=3.2 mL (volume ratio DMC/water=3), t=6 h.
From the data, it is possible to observe a decrease of the 5-
HMF yield from 67.7 to 36.5% under the same experimental
conditions following the order DMC@DEC>1,2-PC>DAlC. Ap-
parently, there is a correlation of the structure and activity of
the OC. An increase of the length of the alkyl chain (from
methyl to ethyl and allyl) reduced the activity as well as the se-
lectivity towards 5-HMF (from 93.2 for DMC to 80.1% for
DAlC). 1,2-PC behaves like DEC. DMC shows the best extraction
properties as the organic solvent in the dehydration process
Isolation and characterization of the final product
As reported above, the isolation of 5-HMF was performed by
simple evaporation of the organic solvent under vacuum, and
an orange solid was obtained, which was analyzed.
1
The H and 13C NMR spectra of 5-HMF isolated from the de-
hydration of fructose using CeP3 as catalyst are shown in Fig-
ures 6 and 7, respectively.
ChemSusChem 2016, 9, 118 – 125
www.chemsuschem.org
121
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
activation of the catalyst is much lower than in water. Phos-
phoric acid is not formed extensively in solution; otherwise it
would have affected the OC significantly and converted it back
to alcohol and CO2. However, if the catalyst is calcined at 823 K
after the fifth cycle it is able to produce 5-HMF with the same
yield and selectivity as in the first run. The use of the OC as
the extracting solvent improves the already positive per-
formance of the catalyst observed in only water.[10] Indeed, the
catalyst shows a longer life and improved activity in the bipha-
sic system with respect to that in only water, which is an addi-
tional benefit of the biphasic system.
Figure 8. Effect of the different organic phase in the dehydration of fructose.
Reaction conditions: fructose=0.64 g, CeP3=0.224 g (weight ratio catalyst/
fructose=35%), OC=9.6 mL, water=3.2 mL (volume ratio organic phase/
water=3), T=423 K, t=6 h.
Conclusions
for the synthesis of HMF, which influences the conversion yield
(by shifting the equilibrium) and the selectivity (by avoiding
permanence in water). Compared to alternative solvents, such
as DMSO[23] or MIBK,[24] or other organic carbonates,[34] DMC
has low boiling point (363 K) that allows its distillation and re-
covery and reuse in several reaction cycles with minimal cost.
We removed DMC by distillation at 333 K under vacuum,
which avoids the deterioration of 5-HMF. Notably, DMC pres-
ents a much better partition coefficient than MIBK (2.31 for
DMC/W compared to 0.9 for MIBK/W[35]) and, thus, is more ef-
fective in the extraction of 5-HMF. The systems used in this
study show a high conversion and selectivity compared to
other systems described in the literature.
The role played by dimethyl carbonate (DMC) and other organ-
ic carbonates to enhance the selectivity towards 5-hydroxyme-
thylfurfural (5-HMF) in the dehydration of fructose in aqueous
media has been investigated. DMC is by far the best cosolvent
among those tested and afforded a high yield and selectivity.
The use of the biphasic water/DMC system allows the continu-
ous extraction of 5-HMF from the aqueous medium, which re-
duces the formation of humins and other byproducts. Under
the best conditions, a 5-HMF yield of >70% with a selectivity
of 93.3% was achieved in the dehydration of fructose using
CeP3 as the catalyst. With respect to an aqueous medium, the
catalyst is recovered easily and reusable in several reaction
cycles with only a slight loss of the catalytic activity after five
runs caused by the adsorption of the polyol on its surface.
A thermal treatment at 823 K can reactivate the catalyst com-
pletely. In addition, 5-HMF is isolated easily at the end of the
dehydration process as DMC is evaporated easily, recovered,
and reused. The use of the biphasic DMC/water system repre-
sents an innovative method for the synthesis of 5-HMF.
Recycling of CeP3
Finally, the catalyst recovery and reusability was tested using
CeP3, the most active of the catalyst tested. After each reac-
tion, the catalyst was separated from the reaction mixture by
centrifugation, washed with water, and dried under a nitrogen
flow at 333 K before use in the next run. The yields of 5-HMF
obtained in five consecutive runs using the same catalyst used
repeatedly are shown in Figure 9. We observed a moderate re-
duction of the yield of 5-HMF (from 67.7 to 59.8%) and the
same selectivity.
As we have shown elsewhere,[10] the deactivation of the cat-
alyst is mainly caused by the adsorption of the organic polyol
on the surface of the catalyst and, to a lower extent, to the
loss of phosphate. With the use of the biphasic system, the de-
Experimental Section
Materials
Cerium ammonium nitrate ꢀ99% (by titration); potassium phos-
phate tribasic monohydrate ꢀ 95%; cerium(IV) oxide fused, pieces,
3–6 mm, 99.9% trace metals basis; phosphoric acid 85wt% in H2O,
99.99% trace metals basis; fructose ꢀ99%; glucose ꢀ99.5%; levu-
linic acid ꢀ99%; dimethyl carbonate ꢀ99% (1=1.069 gmLÀ1);
and the other OCs were ACS-grade reagents purchased from
Sigma–Aldrich.
Catalyst preparation
The synthesis and characterization of catalysts CeP1, CeP2, and
CeP3 was reported previously.[10] Here we summarize the synthetic
aspects briefly for the reader’s convenience. All catalysts gave satis-
factory elemental analyses. We also report the BET area and the
volume of NH3 adsorbed by each catalyst as a measure of the
number of acid sites.
Synthesis of Ce3(PO4)4 (CeP1)
Figure 9. Recycling tests of CeP3. Reaction conditions: fructose=0.64 g, cat-
alyst=0.224 g (weight ratio catalyst/fructose=35%), DMC=9.6 mL, wa-
ter=3.2 mL (volume ratio DMC/water=3), T=423 K, t=6 h.
This catalyst was prepared as described in Ref. [10] from
Ce(NO3)6(NH4)2 and K3PO4 in stoichiometric amount (P/Ce=3 molar
ChemSusChem 2016, 9, 118 – 125
www.chemsuschem.org
122
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
ratio). The yellow precipitate isolated was washed with water and
dried at 403 K. BET À1surface area: 16.7 m2 gÀ1
; Acid sites:
2.14 VNH mLadsorbed gcatalyst
.
3
Synthesis of [(Ce(PO4)(HPO4)0.5(H2O)0.5)] (CeP2)
This catalyst was prepared as described in Ref. [10] from CeO2 and
H3PO4 (PO4/Ce=2.4 molar ratio). The pale yellow suspension was
heated in an iron steel autoclave at 473 K for 5 h, and the light
yellow product was collected by filtration, washed with distilled
water, and dried under
a
nitrogen flow. À1BET surface area:
16.2 m2 gÀ1; Acid sites: 2.09 VNH mLadsorbed gcatalyst
.
3
Synthesis of [(Ce(PO4)1.5(H2O)(H3O)0.5(H2O)0.5)] (CeP3)
This catalyst was prepared as described in Ref. [10]. Ce(NO3)6(NH4)2
was dissolved in HNO3, and then H3PO4 was added (reaction condi-
tions: [Ce4+]=0.1m, PO4/Ce=2 and final [HNO3]=5.8m). The solu-
tion, aged at 368 K for 18 h, yielded a yellow solid that was collect-
ed by filtration, washed with distilled water, and dried under a ni-
trogen flow. BET sÀu1rface area: 18.9 m2 gÀ1
;
Acid sites:
2.42 VNH mLadsorbed gcatalyst
.
3
Our results show that CeP3 is the most acid of the catalysts used,
has the largest surface area, and displays the best activity.
Catalytic tests
Typical procedure for the dehydration of fructose into 5-HMF
in H2O/DMC
Figure 10. 100 mL stainless-steel reactor used in the dehydration of fructose.
The catalytic dehydration of fructose for the synthesis of 5-HMF
was studied in a 100 mL stainless-steel reactor equipped with
a withdrawal valve and an electrical heating jacket (Figure 10).[36]
than that of DMC. The conversion yields and selectivity are report-
ed in Figure 8 and summarized in Table 1. Notably, because of the
different boiling points, DMC is the most suited organic solvent for
the easy recovery of both the solvent and 5-HMF, which was ob-
tained as a solid after DMC evaporation at 333 K.
All tests, unless otherwise specified, were performed using a con-
stant total volume of solvents (OC+W) following the standard pro-
cedure described in the literature. Fructose (0.64 g) was dissolved
in a biphasic system of DMC/water (12.8 mL) in a glass reactor, in
which the catalyst under study (0.1 g, weight ratio of catalyst/fruc-
tose=15.6%) and a magnetic stirrer were placed. The glass reactor
was then transferred into the steel reactor, which was closed and
heated for 3 h at 423 K under atmospheric pressure of air. At the
end of the reaction, the catalyst was recovered by centrifugation,
washed with water (3”2 mL), and dried under a nitrogen flow at
333 K. The two liquid phases were separated, and the organic
phase was washed with water (3”2 mL) to remove impurities and
evaporated under vacuum (333 K) to leave 5-HMF as an orange
Table 1. Reaction conditions, conversion yield, and selectivity for various
OCs.[a]
OC ([mL])
Water
[mL]
Conversion of fructose
[%]
Selectivity to 5-HMF
[%]
DMC (9.6)
DEC (9.6)
1,2-PC (9.6)
DAlC (9.6)
3.2
3.2
3.2
3.2
72.6
49.4
45.9
43.5
93.2
87.8
85.3
81.0
1
solid that was analyzed by H and 13C NMR spectroscopy, IR spec-
[a] Conditions: 0.64 g of fructose, 0.224 g of CeP3 (weight ratio catalyst/
fructose=0.35), 9.6 mL of OC, 3.2 mL of water (volume ratio organic
phase/water=3), T=423 K, t=6 h.
troscopy, and GC–MS (purity>99%).
1
5-HMF: H NMR (600 MHz, CDCl3): d=2.99 (s, 1H, ÀOH), 4.73 (s, 2H,
ÀCH2OH), 6.53 (d, 1H, ÀCH=), 7.21 (d, 1H, ÀCH=), 9.59 ppm (s, 1H,
ÀCHO); 13C NMR (600 MHz, CDCl3): d=57.99 (ÀCH2OH), 110.42 (À
CH=), 152.75 (=CÀCHO), 161.05 (=CÀCH2OH), 178.11 ppm (ÀCHO);
FTIR: n˜ =3200–3000 (vs, OÀH stretch), 2930 (m, CÀH stretch), 2820
and 2750 (w, CHO stretch), 1658 (s, C=O stretch), 1520 (s, C=C
stretch), 1429 (s, CH2 bend), 1072 cmÀ1 (s, CH bend); MS (m/z): 50,
69, 81, 97, 109.
Test on the extraction capacity of DMC working at constant
conversion of fructose
Fructose (0.64 g) was dissolved in water (10 mL) in the presence of
CeP3 (0.22 g). The reaction was performed at 423 K for the time
necessary to reach 40Æ2% conversion. The yield of 5-HMF was
73% as expected.
Use of other OCs
DEC, DAlC, and 1,2-PC were used with water under the same reac-
tion conditions as reported in above. Their efficiency was lower
The reaction was repeated four more times at the same tempera-
ture with the same amount of fructose, catalyst, and water in each
ChemSusChem 2016, 9, 118 – 125
www.chemsuschem.org
123
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[8] a) L. Hu, Z. Wu, J. Xu, Y. Sun, L. Lin, S. Liu, Chem. Eng. J. 2014, 244, 137–
144; b) H. Abou-Yousef, E. B. Hassan, J. Ind. Eng. Chem. 2014, 20, 1952–
1957; c) V. V. Ordomsky, J. van der Schaaf, J. C. Schouten, T. A. Nijhuis, J.
Catal. 2012, 287, 68–75.
[9] a) H. Xu, Z. Miao, H. Zhao, J. Yang, J. Zhao, H. Song, N. Liang, L. Chou,
Fuel 2015, 145, 234–240; b) I. JimØnez-Morales, A. Teckchandani-Ortiz,
J. Santamaría-Gonzalez, P. Maireless-Torres, A. JimØnez-López, Appl.
test with the addition of DMC (1.0, 2.5, 5, and 10 mL). The 5-HMF
yield was 73, 83, 87, 93%, respectively. Different reaction times
were necessary to reach the same conversion of fructose of ꢁ40Æ
2%.
Analytical methods
Catal. B 2014, 144, 22–28; c) V. Ordomsky, J. Van der Schaaf, J. C.
Schouten, T. A. Nijhuis, ChemSusChem 2013, 6, 1697–1707; d) Y. Zhang,
J. Wang, J. Ren, X. Liu, X. Li, Y. Xia, G. Lu, Y. Wang, Catal. Sci. Technol.
2012, 2, 2485–2491; J. Zhuang, L. Lin, C. Pang, Y. Liu, Adv. Mater. Res.
2011, 236–238, 134–137.
5-HMF was analyzed by using a Jasco 5 HPLC with a UV detector
at 284 nm and an ion-exclusion Phenomenex Rezex RHM-Monosac-
charide H+ (8% cross-linked sulfonated styrene-divinylbenzene)
column 300”7.8 mm at 343 K. Sulfuric acid (0.005n) was used as
the mobile phase, and the flow rate was 0.6 mLminÀ1. The concen-
tration of fructose, glucose, levulinic acid, and other byproducts
was determined by using a refractive index (RI) detector. GC–MS
was performed by using a Shimadzu 17A gas chromatograph (ca-
pillary column: 30 m; MDN-5s, Ø 0.25 mm, 0.25 mm film) coupled
to a Shimadzu QP5050A mass spectrometer. Diffuse reflectance in-
frared Fourier transform spectra were recorded by using a Shimad-
zu Prestige 21 instrument equipped with the Shimadzu DRS-8000
[10] A. Dibenedetto, M. Aresta, C. Pastore, L. di Bitonto, A. Angelini, E. Quar-
anta, RSC Adv. 2015, 5, 26941–26948.
[11] K. Shimizu, R. Uozumi, A. Satsuma, Catal. Commun. 2009, 10, 1849–
1853.
[12] X. Tong, Y. Ma, Y. Li, Appl. Catal. A 2010, 385, 1–13.
[13] a) S. De, B. Saha, R. Luque, Bioresour. Technol. 2015, 178, 108–118; b) H.
Xianglin, D. Tiansheng, Z. Yulei, Faming Zhuanli Shenqing, CN
101941957A, 2011.
[14] a) V. Degirmenci, E. J. M. Hensen, Environ. Prog. Sustainable Energy 2014,
33, 657–662; b) L. Hu, Y. Sun, L. Lin, Ind. Eng. Chem. Res. 2012, 51,
1099–1104.
1
basic apparatus modified for our purposes.[37] The H and 13C NMR
[15] T. Tuercke, S. Panic, S. Loebbecke, Chem. Eng. Technol. 2009, 32, 1815–
1822.
[16] J. N. Chheda, Y. Roman-Leshkov, J. A. Dumesic, Green Chem. 2007, 9,
342–350.
[17] J. N. Chheda, J. A. Dumesic, Catal. Today 2007, 123, 59–70.
[18] a) S. Q. Hu, Z. F. Zhang, J. L. Song, Y. X. Zhou, B. X. Han, Green Chem.
2009, 11, 1746–1749; b) S. Q. Hu, Z. F. Zhang, Y. X. Zhou, J. L. Song, H. L.
Fan, B. X. Han, Green Chem. 2009, 11, 873–877.
[19] J. Y. G. Chan, Y. G. Zhang, ChemSusChem 2009, 2, 731–734.
[20] H. P. Yan, Y. Yang, D. M. Tong, X. Xiang and C. W. Hu, Catal. Commun.
2009, 10, 1558–1563.
spectra were recorded by using a BRUKER 600 MHz spectrometer.
The surface characterization of the catalysts was performed by
using a Pulse ChemiSorb 2750 Micromeritics instrument. The analy-
sis of the acid sites was performed using NH3 as a probe gas using
100 mg of catalyst. The samples were pretreated under a N2
(30 mLminÀ1) flow at 373 K. The pulsed chemisorption was per-
formed with NH3 using He as the carrier (30 mLminÀ1). Tempera-
ture-programmed desorption was performed under He flow at
30 mLminÀ1
.
[21] Y. T. Zhang, H. B. Du, X. H. Qian, E. Y. X. Chen, Energy Fuels 2010, 24,
2410–2417.
[22] a) A. Cukalovic, C. V. Stevens, Green Chem. 2010, 12, 1201–1206;
b) B. F. M. Kuster, Starch/Staerke 1990, 42, 314–321.
Acknowledgements
[23] a) S. Despax, C. Maurer, B. Estrine, J. Le Bras, N. Hoffmann, S. Marinkovic,
J. Muzart, Catal. Commun. 2014, 51, 5–9; b) Z. Hu, B. Liu, Z. Zhang, L.
Chen, Ind. Crops Prod. 2013, 50, 264–269; c) H. Li, Q. Zhang, X. Liu, F.
Chang, D. Hu, Y. Zhang, W. Xue, S. Yang, RSC Adv. 2013, 3, 3648–3654.
[24] a) Y. Romµn-Leshkov, C. J. Barrett, Z. Y. Liu, J. A. Dumesic, Nature 2007,
447, 982–985; b) S. Lima, P. Neves, M. M. Antunes, M. Pillinger, N. Igna-
tye, A. A. Valente, Appl. Catal. A 2009, 363, 93–99.
We have received funding from the ENERBIOCHEM Project PON
01966. IC2R is acknowledged for scientific support and loan of
catalysts. VALBIOR Project 33 (Apulia Region) is thanked for the
use of equipment (HPLC, parallel synthesizer, centrifuge evapora-
tor).
[25] a) D. Ballivet, A. Dibenedetto in Carbon Dioxide as Chemical Feedstock,
(Ed.: M. Aresta), Wiley-VCH, Weinheim, 2010; b) V. Chankeshwara Sunay,
Synlett 2008, 4, 624–625.
[26] a) H. William, H. Singh, A. Singh, J. Jimenez, Q. Xu, US 20150025217A1,
2015; b) J. H. Park, J. Y. Jeon, J. J. Lee, Y. Jang, J. K. Varghese, B. Y. Lee,
Macromolecules 2013, 46, 3301–3308.
[27] a) Q. Wang, C. Li, M. Guo, S. Luo, C. Hu, Inorg. Chem. Front. 2015, 2, 47–
54; b) A. Dibenedetto, A. Angelini, L. di Bitonto, E. De Giglio, S. Cometa,
M. Aresta, ChemSusChem 2014, 7, 1155–1161.
Keywords: biphasic catalysis · carbohydrates · cerium · solvent
effects · structure–activity relationships
[1] C. McGlade, J. Speirs, S. Sorrel, Energy 2013, 55, 571–584.
[2] a) R. Parajuli, T. Dalgaard, U. Jørgensen, A. P. S. Adamsen, M. T. Knudsen,
M. Birkved, M. Gylling, J. K. Schjørring, Renewable Sustainable Energy
Rev. 2015, 43, 244–263; b) R. Schlçgl, ChemSusChem 2010, 3, 209–222.
[3] a) J.-L. Dubois in Biorefinery: from biomass to chemicals and fuels (Eds.:
M. Aresta, A. Dibenedetto, F. Dumeignil), De Gruyter, 2012, pp. 19–48;
b) J. Liu, H. Mooney, V. Hull, S. J. Davis, J. Gaskell, T. Hertel, J. Lubchenco,
K. C. Seto, P. Gleick, C. Kremen, S. Li, Science 2015, 347, 963–972; c) S. K.
Maity, Renewable Sustainable Energy Rev. 2015, 43, 1427–1445.
[4] N. Lucas, N. R. Kanna, A. S. Nagpure, G. Kokate, S. Chilukuri, J. Chem. Sci.
2014, 126, 403–413.
[5] R. J. van Putten, J. C. van der Waal, E. D. De Jong, C. B. Rasrendra, H. J.
Heeres, Chem. Rev. 2013, 113, 1499–1597.
[6] a) M. Watanabe, K. Yashiro, T. Nishi, PCT Int. Appl., WO 2013146085A1,
2013; b) S. T. Hansen, J. M. Woodley, A. Riisager, Carbohydr. Res. 2009,
344, 2568–2572.
´
[28] a) M. Litwinowicz, J. Kijenski, Sustainable Chem. Process 2015, 3, 1–7;
b) M. Carafa, E. Quaranta, Mini-Rev. Org. Chem. 2009, 6, 168–183.
[29] a) A. Jain, S. C. Jonnalagadda, K. V. Ramanujachary, A. Mugweru, Catal.
Commun. 2015, 58, 179–182; b) X. Tong, Y. Wang, G. Nie, Y. Yan, Environ.
Prog. Sustainable Energy 2015, 34, 207–210.
[30] a) Y. L. Song, Y. S. Qu, C. P. Huang, L. H. Ge, Y. X. Li, B. H. Chen, Adv.
Mater. Res. 2014, 1004–1005, 885–890; b) H. Kimura, K. Yoshida, Y.
Uosaki, M. Nakahara, J. Phys. Chem. A 2013, 117, 10987–10996.
[31] a) K. D. O. Vigier, A. Benguerba, J. Barrault, F. JØrôme, Green Chem. 2012,
14, 285–289; b) X. Qi, T. M. Aida, R. L. Smith, Jr., Green Chem. 2010, 12,
1855–1860.
[32] a) X. Cao, X. Peng, S. Sun, L. Zhong, W. Chen, S. Wang, R.-C. Sun, Carbo-
hydr. Res. 2015, 118, 44–51; b) L. Yang, X. Yan, S. Xu, H. Chen, H. Xia, S.
Zuo, RSC Adv. 2015, 5, 19900–19906; c) R. Weingarten, A. Rodriguez-
Beuerman, F. Cao, J. S. Luterbacher, D. M. Alonso, J. A. Dumesic, G. W.
Huber, ChemCatChem 2014, 6, 2229–2234; d) M. Yabushita, H. Kobaya-
shi, A. Fukuoka, Appl. Catal. B 2014, 145, 1–9.
[7] a) H. Liu, H. Wang, Y. Li, W. Yang, C. Song, H. Li, W. Zhu, Wei Jiang, RSC
Adv. 2015, 5, 9290–9297; b) J. PØrez-Maqueda, I. Arenas-Ligioiz, Ó.
López, J. Fernµndez-BolaÇos, Chem. Eng. Sci. 2014, 109, 244–250; c) L.
Wang, J. Zhang, L. Zhu, X. Meng, F.-S. Xiao, J. Energy Chem. 2013, 22,
241–244.
ChemSusChem 2016, 9, 118 – 125
www.chemsuschem.org
124
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[33] R. Stephenson, J. Stuart, J. Chem. Eng. Data 1986, 31, 56–70.
[34] M. Benoit, Y. Brissonnet, E. GuØlou, K. De Oliveira Vigier, J. Barrault, F.
JØrôme, ChemSusChem 2010, 3, 1304–1309.
[37] M. Aresta, A. Dibenedetto, C. Pastore, A. Angelini, B. Aresta, I. Pµpai, J.
Catal. 2010, 269, 44–52.
[35] Y. Roman-Leshkov, J. N. Chheda, J. A. Dumesic, Science 2006, 312, 1933–
1937.
[36] A. Dibenedetto, A. Angelini, A. Colucci, C. Pastore, L. di Bitonto, B. M.
Aresta and R. Comparelli, Inter. J. Renewable Energy Biofuels 2015, Article
ID 204112; http://www.ibimapublishing.com/journals/IJREB/AIP/204112/
a204112.html.
Received: September 1, 2015
Revised: October 28, 2015
Published online on December 16, 2015
ChemSusChem 2016, 9, 118 – 125
www.chemsuschem.org
125
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim