Chemical Engineering Journal 254 (2014) 333–339
Contents lists available at ScienceDirect
Chemical Engineering Journal
journal homepage: www.elsevier.com/locate/cej
Catalytic conversion of glucose in dimethylsulfoxide/water binary mix
with chromium trichloride: Role of water on the product distribution
Songyan Jia a, Zhanwei Xu a, Z.Conrad Zhang a,b,⇑
a
b
Dalian National Laboratory for Clean Energy, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China
State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
Tuning water content in DMSO leads
to desirable product distribution from
glucose.
Pure DMSO solvent causes undesired
dehydration of glucose and low HMF
yield.
The available water in DMSO
effectively suppressed undesired
dehydration of glucose.
DMSO/H2O with vw 6 0.3 is capable
of stabilizing HMF.
a r t i c l e
i n f o
Article history:
Received 10 March 2014
Received in revised form 25 May 2014
Accepted 28 May 2014
Available online 4 June 2014
Keywords:
Glucose conversion
5-Hydroxymethylfurfural
DMSO/H2O mixed solvent
Water content
Product distribution
a b s t r a c t
The production of 5-Hydroxymethylfurfural (HMF) from hexoses is a stoichiometric dehydration process.
Water content in a solvent is expected to play an important role in HMF formation by affecting the
equilibrium and the reaction kinetics. In this work, the impact of water content on the catalytic conversion of glucose was investigated in detail in different dimethylsulfoxide (DMSO)/H2O mixtures (vw = 0–1)
with chromium trichloride hexahydrate (CrCl36H2O) as the catalyst at 110–130 °C. Water content in the
binary mix was found to dominantly affect the product distribution. Anhydrous DMSO system is favored
for HMF formation from glucose but caused a number of side reactions, especially the undesired dehydration of glucose into cellobiose. Adding an appropriate amount of water in DMSO (vw = 0.17–0.50) was
found to significantly suppress the undesired dehydration side reactions while preserving high HMF yield
over the CrCl36H2O catalyst, therefore remarkably improving the total selectivity of HMF and fructose
from glucose conversion. While CrCl36H2O was essential in isomerizing glucose into fructose, hydrochloric acid (HCl) from CrCl36H2O hydrolysis in DMSO/H2O mixed system catalyzed the dehydration of in situ
formed fructose to HMF. Although effective water removal was pronounced toward improving HMF yield
from hexose dehydration in previous work, the results of this work indicate that a controlled amount of
water in the non-aqueous system is favorable to drive the thermodynamic equilibrium for high HMF
yield and desired product selectivity, providing reference information on designing a one-pot process
for HMF synthesis from cellulosic materials.
Ó 2014 Elsevier B.V. All rights reserved.
⇑ Corresponding author at: Dalian National Laboratory for Clean Energy, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China.
Fax: +86 411 84379462.
E-mail address: [email protected] (Z. Zhang).
http://dx.doi.org/10.1016/j.cej.2014.05.121
1385-8947/Ó 2014 Elsevier B.V. All rights reserved.
334
S. Jia et al. / Chemical Engineering Journal 254 (2014) 333–339
1. Introduction
Biomass, the most abundant and available renewable carbon
source on the earth, has attracted a great deal of attention in view
of the diminishing fossil fuels [1]. Biomass consists primarily of
lignocellulose, which includes polysaccharides (cellulose and
hemicellulose) and lignin [2]. Compared with the complex and
recalcitrant lignin, the polysaccharides have relatively regular
structures with monosaccharide units and can be hydrolyzed into
various sugars [2,3]. Thus, the development of biomass biorefinery
is critically dependent on catalytic carbohydrate conversions with
high selectivity.
5-Hydroxymethylfurfural (HMF) has recently been regarded as
one important precursor for the production of biofuels and biobased compounds [4]. Dehydration of hexoses into HMF has
emerged as a promising path toward sustainable development
[4]. For the above purpose, fructose is a good starting material,
which can be smoothly converted into HMF in varieties of reaction
media, such as water/organic solvent biphasic system [5],
dimethylsulfoxide (DMSO) [6], dimethylacetamide (DMA)-lithium
chloride (LiCl) [7], and ionic liquid [8], with many different
catalysts, including mineral acid [9], organic acid [10], metal salt
[11], solid acid [12], and functionalized ionic liquid [13]. However,
fructose is of limited abundance in nature and typically at a higher
cost than glucose [14].
Glucose is the most abundant monosaccharide in nature and, as
the main monomer, can be produced from the hydrolysis of starch
and cellulose [14]. However, the obstacle of effectively converting
glucose into HMF has slowed the development of HMF based biorefinery. Zhang and co-workers first reported that chromium chloride (CrCl2 or CrCl3) catalyzed the efficient production of HMF
from glucose in ionic liquids by enabling the isomerization of glucose into fructose [15]. Since then, ionic liquids have been mostly
employed as the reaction media for the HMF production from glucose based feedstocks with chromium chloride [16–20]. Except
ionic liquids, other solvents, such as DMA and DMSO, have been
employed for glucose conversion as well, and it was found that
chromium chloride was also effective in those systems for the
isomerization of glucose into fructose [7,21,22]. The versatility of
chromium chloride catalyst in different solvents offers more
options toward the development of alternative reaction systems
for saccharide conversions.
In previous works, it was reported that DMSO could improve
the HMF yield from fructose by suppressing side reactions [5,23].
Shimizu et al. reported that the in situ formed water from fructose
dehydration lowered the HMF yield in DMSO system over Amberlyst-15 [24], so that effective water removal was pronounced
toward the enhanced production of HMF [24]. Nevertheless, as
water formation is inevitable in a dehydration reaction, information on whether and how water content in non-aqueous media
affects HMF production from hexoses, especially glucose, is an
important subject of research and the existing literature remains
lacking in such information. Recently, a few studies focused on
the effect of water content in HMF chemistry by employing
DMSO/water (H2O) binary mix. Tsilomelekis et al. studied the origin of HMF stability in DMSO/deuterium oxide (D2O), and proposed
that HMF solvation by DMSO reduced its susceptibility to nucleophilic attack, therefore minimizing undesirable hydrolysis of
HMF and humin formation [25]. HMF solvation by DMSO remained
preferential at vD2O < 0.4 [25], indicating that an appropriate
amount of water in such a system does not lower the stability of
HMF. Kimura et al. studied the non-catalytic conversion of cellobiose in DMSO/water at high temperature (170 °C), in which higher
HMF yield (70%) was obtained after 23–26 h in the DMSO/water
mixtures with vw = 0.2–0.3 than that in pure DMSO (HMF yield of
45%) or water (HMF yield of 40%) [26]. In our previous study, it
was shown that isomerization of glucose into fructose catalyzed by
CrCl36H2O followed a general curve in water independent of
reaction variables, while adding a controlled amount of DMSO as
co-solvent significantly changed the reaction behavior [27]. For
example, when a DMSO/water mixture (v/v, 8/2) was used as the
solvent, the obtained HMF yield (11.0%) was comparable to that
(12.4%) in pure DMSO at 110 °C after 4 h [27]. The above studies
demonstrate that an appropriate amount of water may be tolerated and even more beneficial for HMF synthesis in DMSO.
It is noted that the DMSO/water mixture solvent has been well
studied. For instance, several characteristic parameters of the mixture, such as melting point, density and viscosity, deviate from the
ideal solution as a function of the molar fraction of water (vw)
[28,29]. As reported, water appears tetrahedrally hydrogen
bonded, linking each water molecule to four water molecule neighbors in tetrahedral geometry [30]. In DMSO/water binary mixtures
with higher water content (vw > 0.5), the local tetrahedral structure of water is still preserved [31,32], especially at vw P 0.7
[33], implying that the mixtures have water-like properties. However, in the DMSO-rich aqueous system (vw < 0.5), DMSO molecules retain the same molecular arrangement as in the pure state
[34]. Therefore, the DMSO/water binary mix may offer unique
properties for biomass conversion research.
In this work, we investigated the catalytic conversion of glucose
with chromium trichloride catalyst in varied DMSO/water binary
solvents. The objective was to explore how water content of the
mixed solvents may affect the catalytic conversion of glucose.
Achieving such an understanding may also help reconcile the large
difference between very high glucose conversion and relatively
low HMF yield in some previous literatures [21,22].
2. Experimental
2.1. Materials
D-Glucose (99%), D-fructose (99%), 1,6-anhydro-b-D-glucose
(AGP, 99%) and D-cellobiose (98%) were purchased from Alfa Aesar.
Chromium trichloride hexahydrate (CrCl36H2O, 96%) was purchased from Sigma–Aldrich. 5-Hydroxymethylfurfural (HMF, 98%)
and levulinic acid (LA, 99%) were purchased from Aladdin. Glycerol
(99%) and dimethylsulfoxide (DMSO, 99%) were purchased from
Sinopharm (China). Hydrochloric acid (HCl, 36–38 wt%) was provided by a local supplier. All the chemicals were used as received.
Deionized water (DI H2O) with a resistivity of 18.2 MX-cm was
produced by a Milli-Q Integral 5 system.
2.2. Typical reaction procedure
Glucose (50 mg) or fructose (50 mg), CrCl36H2O (3.7 mg,
5 mol% with respect to glucose or fructose) and total 1 mL of solvent were added into each reaction vial with a magnetic stir bar
for reaction. The reaction vial was sealed and it was used as an
autoclave. The reaction is conducted at autogenic pressure. It
should be specified that some solvents may be composed of both
DMSO and DI H2O, and the 1 mL was defined as the sum volume
of DMSO and H2O. Eight DMSO/H2O solutions with volume ratios
(lL/lL) of 1000/0, 950/50, 900/100, 800/200, 600/400, 400/600,
200/800 and 0/1000 were employed. Based on that the relative
densities of DMSO and H2O are 1.1 g/mL and 1.0 g/mL, and the
molecular weights of DMSO and H2O are 78.13 g/mol and
18.02 g/mol, respectively, the molar fraction of H2O (vw) in the
above DMSO/H2O solutions was calculated to be 0, 0.17, 0.30,
0.50, 0.72, 0.86, 0.94 and 1, respectively. The feedstocks have good
solubilities in either water or DMSO. For example, the solubility of
glucose in water and DMSO is 909 mg/mL and 540 mg/mL,
S. Jia et al. / Chemical Engineering Journal 254 (2014) 333–339
respectively, around room temperature [35,36]. As reported, fructose has higher solubility than other sugars as well as other sugar
alcohols in water [37], and it also has a good solubility in DMSO
[38]. Moreover, HMF is also soluble in water and many polar
organic solvents [39]. Therefore, it is believed that the used feedstocks and produced HMF completely dissolved in the above reaction systems. The reactions were operated on a heating and stirring
module (TS-18821, Thermo Scientific) with the stirring rate at
500 rpm for a specified time at the reaction temperature. After
the reaction was quenched in ice water bath, 1 mL of DI H2O and
the internal standard were added into the reaction mixture. Then,
a small amount of the sample mixture was taken out and further
diluted by DI H2O for 15 times and analyzed by high performance
liquid chromatography (HPLC).
2.3. Analyses
HPLC analysis was performed on an Agilent 1260 series with a
refractive index detector and a PL Hi-Plex H column
(300 7.7 mm, 8 lm). The mobile phase was 0.005 M H2SO4 aqueous solution at a flow rate of 0.6 mL/min and the volume for each
injection was 10 lL. The column and detector temperatures were
65 °C and 50 °C, respectively. Authentic chemical compounds were
used to identify the retention times. Glycerol was added as the
internal standard for the quantitative calculations. The pH values
were measured by Core Module 3 (CM3) System (Freeslate).
3. Results and discussion
The production of HMF from hexoses is a stoichiometric
dehydration process in a solvent. The effect of water content on
the catalytic conversion of glucose was studied in DMSO/H2O mixtures with CrCl36H2O. The initial investigations were carried out in
various DMSO/H2O ratios at 110 °C.
As illustrated in Fig. 1a, after the experiments were run for 2 h,
the glucose conversion for all the systems reached a similar level of
23%. However, the product distribution showed different
patterns. In anhydrous DMSO, the HMF yield (9.6%) was the highest among all the systems with only a little fructose (3%) detected.
As reported, CrCl36H2O is a good catalyst for isomerizing glucose
into fructose [15,27,40,41]. Anhydrous DMSO is a strong dehydration medium, which facilitates HMF formation from the in situ
formed fructose. As a general trend, with increasing vw from 0 to
1, HMF yield was decreased along with increased fructose yield,
indicating that water addition suppressed the dehydration of
335
fructose into HMF. It is noted that, however, in the system with
vw = 0.17, the HMF yield (6.4%) remained close to that in anhydrous DMSO, while fructose yield was much increased (11.2%).
The total yield of HMF and fructose was much higher than that
in anhydrous DMSO. With vw > 0.50, HMF production was essentially quenched, and fructose was the primary product as a result
of the isomerization of glucose over CrCl36H2O, which is consistant with previous work [27]. Cellobiose, 1,6-anhydro-b-Dglucopyranose/glucofuranose (AGP/AGF) and levulinic acid (LA)
were also detected. Cellobiose, AGP and AGF are the products by
undesired intermolecular or intramolecular dehydration of glucose. LA is a downstream product due to the hydrolytic cracking
of HMF, with formic acid (FA) as a concomitant product [5,23].
However, the LA yields in the above systems were generally lower
than 2% and no FA was detected by HPLC, so they were not discussed herein. The total yield of cellobiose, AGF and AGP is hereafter called by-product yield. Notably, quantities of by-products
(9.1%) were produced in anhydrous DMSO, where cellobiose was
the main component with a yield of 7.5%. With increasing vw,
the yield of by-product dropped substantially to a very low level,
which can be attributed to the increased water concentration that
suppressed the dehydration of fructose into HMF as well as the
undesired dehydration of glucose into by-products.
Given in Fig. 1b, when the reaction time was extended to 3 h,
the glucose conversion in anhydrous DMSO was much improved
by 12%, while the other aqueous systems gave an average of
5% increase, indicating that water addition did retard the glucose
conversion. The product distribution was comparable to that
obtained at 2 h. The formation of by-products was significantly
enhanced in anhydrous DMSO, especially cellobiose, of which the
yield increased from 7.5% to 12.4%. However, the by-product yields
just slightly increased in other water containing systems (Fig. 1b),
which further demonstrates that water addition could drive the
thermodynamic equilibrium to circumvent undesired dehydration
of glucose. The HMF production remained at a very low level with
vw > 0.50, possibly because the DMSO/H2O mixed solvents displayed water-like properties [31–33], and the potential of DMSO
for the dehydration of in situ formed fructose was much inhibited
under these conditions. Remarkably, with vw = 0.17, the HMF yield
(13.5%) was higher than that (11.3%) in anhydrous DMSO system,
evidently demonstrating that an appropriate amount of water
could be tolerated and beneficial for a non-aqueous system in producing HMF from glucose. With vw = 0.30 and 0.50, the HMF yields
were 8.7% and 7.5%, respectively, both close to that in anhydrous
DMSO. The above results imply that higher water content up to
Fig. 1. Catalytic conversion of glucose in different DMSO/H2O mixtures at 110 °C. (a) 2 h; (b) 3 h. Conditions: 50 mg of glucose with CrCl36H2O (5 mol% with respect to
glucose) was added into 1 mL of DMSO/H2O solvent and heated at 110 °C for different times.
336
S. Jia et al. / Chemical Engineering Journal 254 (2014) 333–339
vw = 0.50 is also not significantly inhibitive for HMF production,
because DMSO molecules still retain similar feature as its pure
state at vw < 0.5 according to the well studied properties of the
DMSO/H2O binary mixture [34].
To explore whether DMSO with higher water content was also
effective for HMF production from glucose, we next investigated
the effect of reaction temperature. As illustrated in Fig. 2a, when
the reactions were run at 120 °C for 1 h, the glucose conversion
for all the systems were around 26%. In anhydrous DMSO, the
HMF yield was 12%. With increasing vw, the HMF yield decreased
from 9.7% (vw = 0.17) to 1.2% (vw = 1), while the fructose yield
increased accordingly. It was again found that much higher yield
(9%) of by-product was observed in anhydrous DMSO. Since
isomerization of glucose as well as the rate limiting step in the
dehydration of fructose are endothermic [42–44], it is not surprising that similar glucose conversions and HMF yields were obtained
in varied DMSO/H2O systems at 120 °C within short time (1 h) and
at 110 °C for 2 h (Figs. 1a and 2a). Dehydration of glucose into HMF
is considered to undergo a two-step tandem reaction, including
isomerization of glucose into fructose and dehydration of fructose
into HMF [15,17]. HMF and fructose are the desired products. As
seen in Fig. 2a, due to the large formation of by-products, the selectivity of desired product (56.2%) was low in anhydrous DMSO,
while the corresponding selectivities in other aqueous systems
stood at the level of 80%, mainly because water did suppress
the undesired dehydration from glucose.
As shown in Fig. 2b and c, when the reaction time was extended
to 2 h and 3 h, glucose conversions were significantly improved in
anhydrous DMSO than in other water containing systems. HMF
was more effectively produced in DMSO/H2O systems with
vw = 0.17–0.50. For 2 h reaction at 120 °C, the HMF yield (19.4%)
in vw = 0.17 system was higher than that (15.3%) in anhydrous
DMSO. The vw = 0.30 and 0.50 systems gave HMF yields of 14%
and 12.7%, respectively. Remarkably, for 3 h reaction, the HMF
yields in vw = 0.17, 0.30 and 0.50 systems were 21.5%, 20.1% and
17.6%, respectively, all of which were higher than that (16.6%) in
anhydrous DMSO. However, the selectivity of desired product
decreased with prolonging reaction time. After 3 h at 120 °C, the
selectivity of desired product dropped to only 33.8% in anhydrous
DMSO. Due to much increased formation of by-products, the selectivity of desired product decreased to 60.4% in vw = 0.17 system,
while the selectivity of desired product still remained at the level
of 70% with vw = 0.30–1 under these conditions. The difference
between glucose conversions and total product yields as shown
in Fig. 2 represents unaccounted by-products. However, as illustrated in Figs. S1 and S2, no obvious other products could be
detected by HPLC analysis except the known compounds. The difference could be ascribed to the formation of polymeric byproducts or humins [45]. Mass balance is an important indicator
for a reaction, and further study is essential for reducing mass loss
by circumventing undesired side reactions and humin formation.
When the reaction temperature was elevated to 130 °C, HMF
production was further improved. As illustrated in Fig. 2d, after
3 h, the HMF yields in DMSO/H2O with vw = 0.17, 0.30 and 0.50
reached up to 35%, 33.5% and 31.6%, respectively, which were
much higher than that (24.5%) in anhydrous DMSO. Although a
Fig. 2. Effect of temperature on the catalytic conversion of glucose in different DMSO/H2O mixtures. (a) 120 °C, 1 h; (b) 120 °C, 2 h; (c) 120 °C, 3 h; (d) 130 °C, 3 h. Conditions:
50 mg of glucose with CrCl36H2O (5 mol% with respect to glucose) was added into 1 mL of DMSO/H2O solvent and heated at 120 °C or 130 °C for different times.
S. Jia et al. / Chemical Engineering Journal 254 (2014) 333–339
337
little more by-products (22.5%) were produced in anhydrous DMSO
after 3 h at 130 °C, the cellobiose yield (9.2%) was lower than that
(13.4%) after 3 h at 120 °C, possibly because more cellobiose
underwent deep polymerization into cellotriose, cellotetrose and
even oligomers. The above results further demonstrate that an
appropriate aqueous DMSO system is more beneficial for the
HMF production under specified conditions.
The blank tests for glucose conversion were also studied in different DMSO/H2O systems. As seen in Table S1, when the reactions
were run at 120 °C for 2 h, negligible glucose conversions were
observed. However, the glucose conversion reached about 10.8%
with cellobiose (9.1%) as the primary product in anhydrous DMSO
at 130 °C for 2 h. The results indicate that pure DMSO system is
capable of causing the undesired direct dehydration of glucose
even in the absence of a catalyst. However, the available water in
DMSO could effectively suppress the side reactions, which explains
why lower yields of by-products were formed in aqueous systems.
As reported in the literature, fructose is generally considered as
the intermediate for the conversion of glucose into HMF [15,17,21].
We also studied how water content affected the conversion of fructose into HMF in varied DMSO/H2O systems with CrCl36H2O.
Fig. S3 presents the results for blank tests of fructose conversion
in different DMSO/H2O mixtures. It was found that fructose underwent effective dehydration into HMF in anhydrous DMSO even
without a catalyst, which is consistent with the existing literature
[23,46]. At 120 °C for 1 h, fructose conversion was 86% with HMF
yield of 56.4%, and at 130 °C for 1 h, fructose was almost
completely converted with HMF yield of 67.2%. However, the dehydration of fructose into HMF stopped when only a little water (i.e.
vw = 0.17) was added, indicating that the available water in the
system was detrimental to the conversion of fructose in the
absence of a catalyst.
When CrCl36H2O (5 mol% with respect to fructose) was added,
the dehydration of fructose was recovered at different levels in
different DMSO/H2O systems. As illustrated in Fig. 3, moderate to
high yield of HMF could be achieved in vw = 0.17–0.50 systems.
However, HMF production remained at a low level when vw was
higher than 0.50, which is consistent with the low HMF yields from
glucose at vw > 0.50 as shown in Fig. 2a. CrCl36H2O could be
hydrolyzed into Cr species and HCl in an aqueous solution. In previous works, Cr species was proposed as the catalyst for the isomerization of glucose [27,40,41], while HCl worked for the
conversion of fructose [40]. Therefore, the different levels of HMF
yield in Fig. 3 could be attributed to the pH value in available
water. We then checked the above hypothesis by employing two
DMSO/H2O solutions with vw = 0.30 and 0.50. The vw = 0.30 system contained 3.7 mg of CrCl36H2O and 100 lL of water, while
the vw = 0.50 system contained 3.7 mg of CrCl36H2O and 200 lL
of water. On the assumption that HCl was formed by CrCl36H2O
hydrolysis in available water, the pH values of the two CrCl36H2O
solutions with concentrations of 3.7 mg/100 lL and 3.7 mg/200 lL
were 2.08 and 2.34, respectively, at room temperature. However, the conversion of fructose was conducted at 120 °C for 1 h,
high temperature would accelerate the hydrolysis of CrCl36H2O,
leading to a low pH value [27]. After the above CrCl36H2O solutions were heated at 120 °C for 1 h, the corresponding pH values
dropped to 1.09 and 1.42, respectively. Then two solutions with
pH of 1.09 and 1.42 were prepared by diluting concentrated HCl
solution, and used for fructose conversion. As shown in Fig. 4, both
fructose conversions and HMF yields in DMSO/H2O with vw = 0.30
(900 lL DMSO/100 lL H2O, pH(H2O) = 1.09) and vw = 0.50 (800 lL
DMSO/200 lL H2O, pH(H2O) = 1.42) were comparable to those
with CrCl36H2O as catalyst, indicating that the conversion of fructose in aqueous DMSO system is correlated with the pH value in
available water.
As shown in Fig. 3, glucose was also detected in the reaction.
According to previous literature, the isomerization between
glucose and fructose is reversible [42]. Most recent work also indicates that CrCl36H2O catalyzed the conversion of aldoses (i.e. glucose, xylose) and established a reversible path between aldose and
ketose (i.e. fructose, xylulose) [40,47]. Therefore, the formation of
glucose from fructose is expected. As shown in Fig. S4, when the
reaction time was extended or the temperature was elevated, the
HMF yields in DMSO/H2O systems with vw = 0.30 or 0.50 were further improved, either superior or very close to that in anhydrous
DMSO, which is consistent with the results that HMF production
from glucose was superior or close to that in anhydrous DMSO
(Fig. 2b–d). The results indicate that the negative effect of water
on the dehydration of fructose into HMF could be overcome by
adding a catalyst under specified conditions.
We finally tested the stability of HMF in different DMSO/H2O
mixtures. As seen in Table 1, HMF was essentially stable in
DMSO/H2O with vw 6 0.50 at 120 °C in the absence of a catalyst,
of which the recovery was higher than 95%. However, when the
vw was further increased, HMF recovery slightly decreased. Tsilomelekis et al. recently studied the origin of HMF stability in
DMSO/D2O mixtures [25]. It was found that DMSO could bind to
both the hydroxyl and carbonyl groups of HMF more strongly than
water. When HMF was solvated by DMSO, the LUMO energy of
HMF was higher, which indicates that solvation by DMSO reduces
Fig. 3. Catalytic conversion of fructose in different DMSO/H2O mixtures at 120 °C
for 1 h. Conditions: 50 mg of fructose and CrCl36H2O (5 mol% with respect to
fructose) were added into 1 mL of DMSO/H2O solvent and heated at 120 °C for 1 h.
Fig. 4. Comparison on the dehydration of fructose into HMF in DMSO/H2O with
vw = 0.30 and 0.50 in the presence of CrCl36H2O and HCl at 120 °C for 1 h.
338
S. Jia et al. / Chemical Engineering Journal 254 (2014) 333–339
Table 1
Stability of HMF in different DMSO/H2O systems.
vw
0
0.17
0.30
0.50
0.72
1
HMF recovery (%) without catalyst
HMF recoverya (%) with catalyst
96.3
97.3
96.1
96.0
96.9
95.3
96.8
89.2
92.9
89.1
92.2
86.6
Reaction conditions: 17 mg of HMF and 1 mL of solvent were added into each
reaction vial and heated at 120 °C for 2 h.
a
3.7 mg of CrCl36H2O (10 mol% with respect to HMF) was added.
its susceptibility to nucleophilic attack, therefore minimizing
undesirable hydrolysis of HMF and humin formation [25]. In
HMF/DMSO/D2O ternary mixtures, DMSO preferentially solvated
the carbonyl group of HMF if vD2O was less than 0.4 [25], which
may explain why HMF was more stable at vw = 0–0.50 without a
catalyst in this work. Since HMF was produced from glucose or
fructose under catalytic reaction conditions, the stability of HMF
was also examined in the presence of CrCl36H2O. HMF was still
stable at vw 6 0.30, while its recovery showed remarkable
decrease than those without catalyst at vw P 0.50, possibly
because the solvation of DMSO was disrupted to some extent,
and the catalyst, especially the formed HCl, caused the degradation
of HMF. In previous study, it is also reported that in DMSO/H2O
binary mixtures, the local tetrahedral structure of water is still preserved in the presence of DMSO molecules at higher water content
(vw > 0.5) [31–33]. However, in the DMSO-rich system (vw < 0.5),
DMSO molecules retain the same molecular arrangement as in
the pure state [34], which offers a perspective basis for the stability
of HMF. In brief, DMSO is a solvent capable of stabilizing HMF and
the stabilization persists in the presence of an appropriate amount
of water.
In the end, we use Fig. 5 to summarize the conversion of glucose
in DMSO/H2O mixed solvents. Anhydrous DMSO with CrCl36H2O
significantly accelerated glucose conversion, leading to effective
production of HMF. In the meantime, however, glucose is apt to
undergo undesired dehydration to produce cellobiose, AGP and
AGF as well at high temperature, reducing the selectivity of desired
product. In the literature, glucose was converted in DMSO at 120 or
140 °C with about 100% conversion and 50% yield of HMF [21,22],
but the large difference between glucose conversion and HMF yield
was not explained. The results in this work imply that by-product
such as cellobiose, AGP and AGF may account for bulk of the missing products. However, a number of other products remained
unknown, possibly humins. Compared with anhydrous DMSO system, the key role of added water in DMSO was to suppress the
undesired dehydration of glucose. Water addition also limited
the dehydration of in situ formed fructose into HMF in DMSO
without a catalyst. CrCl36H2O catalyzed the isomerization of glucose. HCl from the hydrolysis of CrCl36H2O was sufficient for the
dehydration of fructose into HMF. In aqueous solution, HMF also
is known to undergo hydrolysis to form LA and FA under acid catalyzed conditions [5,23]. However, only a very low amount of LA
was detected in this work, which may be explained by the following reasons: (1) LA was effectively synthesized from carbohydrates
at temperature up to 160–200 °C in strongly acidic media [48–50],
while the employed conditions in this work were much milder; (2)
at vw < 0.5, the DMSO/H2O systems preserved the DMSO-like properties, which may suppress the side reactions as mentioned above
[5,25,51]; (3) although the reaction media had water-like properties at vw > 0.5, HMF formation was much inhibited, and the low
HMF concentration in the media would not effectively drive the
equilibrium into LA.
4. Conclusions
Water content in DMSO/H2O binary mixture effectively affected
the product distribution in the catalytic conversion of glucose.
Anhydrous DMSO system was favored for the effective production
of HMF from glucose but also gave rise to many side reactions, such
as the undesired intermolecular and intramolecular dehydrations
of glucose. The key role of added water in DMSO was to suppress
undesired formation of cellobiose, AGP and APF so that to improve
the selectivity of desired product. Employing a catalyst in an
appropriate DMSO/H2O mixed solvent may overcome the limitation of water addition on the dehydration of fructose into HMF.
With vw = 0.17–0.50, HMF yield reached up to 35% after 3 h at
130 °C, which was much higher than that obtained in anhydrous
DMSO. Therefore, tuning the water content in DMSO is an effective
approach to achieve high HMF yield. CrCl36H2O catalyzed the
isomerization of glucose into fructose. The conversion of fructose
into HMF was achieved by HCl from the hydrolysis of CrCl36H2O,
and by a comparable acid concentration with the same pH value
in available water of the DMSO/H2O mixture. HMF was stable in
DMSO/H2O with vw 6 0.30 with or without CrCl36H2O in this
work, implying that a certain amount of water was tolerated in
DMSO in terms of HMF stability. In short, although effective water
removal is pronounced toward the enhanced production of HMF
from hexose dehydration in previous work, a system with an
appropriate amount of water is superior to an anhydrous system
in producing HMF by effectively circumventing side reactions.
Fig. 5. Schematic diagram on the conversion of glucose in DMSO/H2O binary mix system.
S. Jia et al. / Chemical Engineering Journal 254 (2014) 333–339
The results reported herein may provide reference information for
the one-pot synthesis of HMF from cellulosic materials, because
water is necessary for the hydrolysis of the polymer feedstocks.
Acknowledgements
This work was supported by the China Postdoctoral Science
Foundation (2013M530952, 2013M540236), National Natural
Science Foundation of China (21306186), and the Chinese Government ‘‘Thousand Talent’’ program funding.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.cej.2014.05.121.
References
[1] P. Gallezot, Conversion of biomass to selected chemical products, Chem. Soc.
Rev. 41 (2012) 1538–1558.
[2] J.A. Melero, J. Iglesias, A. Garcia, Biomass as renewable feedstock in standard
refinery units. Feasibility, opportunities and challenges, Energy Environ. Sci. 5
(2012) 7393–7420.
[3] P. Mäki-Arvela, B. Holmbom, T. Salmi, D.Y. Murzin, Recent progress in
synthesis of fine and specialty chemicals from wood and other biomass by
heterogeneous catalytic processes, Catal. Rev. 49 (2007) 197–340.
[4] R.v. Putten, J.C.v.d. Waal, E.d. Jong, C.B. Rasrendra, H.J. Heeres, J.G.d. Vries,
Hydroxymethylfurfural a versatile platform chemical made from renewable
resources, Chem. Rev. 113 (2013) 1499–1597.
[5] Y. Román-Leshkov, J.N. Chheda, J.A. Dumesic, Phase modifiers promote
efficient production of hydroxymethylfurfural from fructose, Science 312
(2006) 1933–1937.
[6] X. Tong, Y. Li, Efficient and selective dehydration of fructose to 5hydroxymethylfurfural catalyzed by Brønsted-acidic ionic liquids,
ChemSusChem 3 (2010) 350–355.
[7] J.B. Binder, R.T. Raines, Simple chemical transformation of lignocellulosic
biomass into furans for fuels and chemicals, J. Am. Chem. Soc. 131 (2009)
1979–1985.
[8] A.H. Jadhav, A. Chinnappan, R.H. Patil, S.V. Kostjuk, H. Kim, Green chemical
conversion of fructose into 5-hydroxymethylfurfural (HMF) using
unsymmetrical dicationic ionic liquids under mild reaction condition, Chem.
Eng. J. 243 (2014) 92–98.
[9] C. Sievers, I. Musin, T. Marzialetti, M.B.V. Olarte, P.K. Agrawal, C.W. Jones, Acidcatalyzed conversion of sugars and furfurals in an ionic-liquid phase,
ChemSusChem 2 (2009) 665–671.
[10] N. Jiang, R. Huang, W. Qi, R. Su, Z. He, Effect of formic acid on conversion of
fructose to 5-hydroxymethylfurfural in aqueous/butanol media, Bioenerg. Res.
5 (2012) 380–386.
[11] Z. Zhang, B. Liu, Z. Zhao, Conversion of fructose into 5-HMF catalyzed by GeCl4
in DMSO and [Bmim]Cl system at room temperature, Carbohydr. Polym. 88
(2012) 891–895.
[12] V.V. Ordomsky, J.C. Schouten, J. Schaaf, T.A. Nijhuis, Foam supported
sulfonated polystyrene as a new acidic material for catalytic reactions,
Chem. Eng. J. 207–208 (2012) 218–225.
[13] Q. Bao, K. Qiao, D. Tomida, C. Yokoyama, Preparation of 5hydroymethylfurfural by dehydration of fructose in the presence of acidic
ionic liquid, Catal. Commun. 9 (2008) 1383–1388.
[14] A. Corma, S. Iborra, A. Velty, Chemical routes for the transformation of biomass
into chemicals, Chem. Rev. 107 (2007) 2411–2502.
[15] H. Zhao, J.E. Holladay, H. Brown, Z.C. Zhang, Metal chlorides in ionic liquid
solvents convert sugars to 5-hydroxymethylfurfural, Science 316 (2007)
1597–1600.
[16] C. Li, Z. Zhang, Z.K. Zhao, Direct conversion of glucose and cellulose to 5hydroxymethylfurfural in ionic liquid under microwave irradiation,
Tetrahedron Lett. 50 (2009) 5403–5405.
[17] Y. Zhang, E.A. Pidko, E.J.M. Hensen, Molecular aspects of glucose dehydration
by chromium chlorides in ionic liquids, Chem. Eur. J. 17 (2011) 5281–5288.
[18] Y. Yi, M. Ha, J. Lee, C. Chung, New role of chromium fluoride: its catalytic action
on the synthesis of hydroxymethylfurfural in ionic liquid using raw plant
biomass and characterization of biomass hydrolysis, Chem. Eng. J. 180 (2012)
370–375.
[19] Y. Su, H.M. Brown, X. Huang, X. Zhou, J.E. Amonette, Z.C. Zhang, Single-step
conversion of cellulose to 5-hydroxymethylfurfural (HMF), a versatile platform
chemical, Appl. Catal. A 361 (2009) 117–122.
[20] J. Zhang, Y. Cao, H. Li, X. Ma, Kinetic studies on chromium-catalyzed
conversion of glucose into 5-hydroxymethylfurfural in alkylimidazolium
chloride ionic liquid, Chem. Eng. J. 237 (2014) 55–61.
339
[21] Z. Yuan, C. Xu, S. Cheng, M. Leitch, Catalytic conversion of glucose to 5hydroxymethyl furfural using inexpensive co-catalysts and solvents,
Carbohydr. Res. 346 (2011) 2019–2023.
[22] C.B. Rasrendra, J.N.M. Soetedjo, I.G.B.N. Makertihartha, S. Adisasmito, H.J.
Heeres, The catalytic conversion of D-glucose to 5-hydroxymethylfurfural in
DMSO using metal salts, Top. Catal. 55 (2012) 543–549.
[23] B.F.M. Kuster, 5-Hydroxymethylfurfural (HMF). A review focussing on its
manufacture, Starch/Stärke 42 (1990) 314–321.
[24] K. Shimizu, R. Uozumi, A. Satsuma, Enhanced production of
hydroxymethylfurfural from fructose with solid acid catalysts by simple
water removal methods, Catal. Commun. 10 (2009) 1849–1853.
[25] G. Tsilomelekis, T.R. Josephson, V. Nikolakis, S. Caratzoulas, Origin of 5hydroxymethylfurfural stability in water/dimethyl sulfoxide mixtures,
ChemSusChem 7 (2014) 117–126.
[26] H. Kimura, K. Yoshida, Y. Uosaki, M. Nakahara, Effect of water content on
conversion of D-cellobiose into 5-hydroxymethyl-2-furaldehyde in a dimethyl
sulfoxide-water mixture, J. Phys. Chem. A 117 (2013) 10987–10996.
[27] S. Jia, K. Liu, Z. Xu, P. Yan, W. Xu, X. Liu, Z.C. Zhang, Reaction media dominated
product selectivity in the isomerization of glucose by chromium trichloride:
from aqueous to non-aqueous systems, Catal. Today (2014), http://dx.doi.org/
10.1016/j.cattod.2014.02.038.
[28] D.H. Rasmussen, A.P. MacKenzie, Phase diagram for the system waterdimethylsulphoxide, Nature 220 (1968) 1315–1317.
[29] J.M.G. Cowie, P.M. Toporowski, Association in the binary liquid system
dimethylsulphoxide-water, Can. J. Chem. 39 (1961) 2240–2243.
[30] I. Brovchenko, A. Oleinikova, Interfacial and Confined Water, Elsevier, 2008.
[31] E. Antoniou, Structure-property relationships in associating polymer solutions.
(Dissertation) State University of New York at Buffalo, 2008.
[32] S.E. McLain, A.K. Soper, A. Luzar, Investigations on the structure of dimethyl
sulfoxide and acetone in aqueous solution, J. Chem. Phys. 127 (2007) 174515.
[33] N. Zhang, W. Li, C. Chen, J. Zuo, Molecular dynamics simulation of aggregation
in dimethyl sulfoxide-water binary mixture, Comput. Theor. Chem. 1017
(2013) 126–135.
[34] Y. Koga, Y. Kasahara, K. Yoshino, K. Nishikawa, Mixing schemes for aqueous
dimethyl sulfoxide: support by X-ray diffraction data, J. Solution Chem. 30
(2001) 885–893.
[35] http://en.wikipedia.org/wiki/Glucose (checked on May 23th, 2014).
[36] http://www.gaylordchemical.com/uploads/images/pdfs/literature/
102B_english.pdf (checked on May 23th, 2014).
[37] http://en.wikipedia.org/wiki/
Fructose#Fructose_solubility_and_crystallization (checked on May 23th,
2014).
[38] K. Tahvildari, S. Taghvaei, M. Nozari, The study of hydroxymethylfurfural as a
basic reagent for liquid alkanes fuel manufacture from agricultural wastes, Int.
J. Chem. Environ. Eng. 2 (2011) 62–68.
[39] http://ntp.niehs.nih.gov/ntp/htdocs/chem_background/exsumpdf/
hydroxymethyl.pdf (checked on May 23th, 2014).
[40] V. Choudhary, S.H. Mushrif, C. Ho, A. Anderko, V. Nikolakis, N.S. Marinkovic, A.I.
Frenkel, S.I. Sandler, D.G. Vlachos, Insights into the interplay of Lewis and
Brønsted acid catalysts in glucose and fructose conversion to 5(hydroxymethyl)furfural and levulinic acid in aqueous media, J. Am. Chem.
Soc. 135 (2013) 3997–4006.
[41] V. Choudhary, A.B. Pinar, R.F. Lobo, D.G. Vlachos, S.I. Sandler, Comparison of
homogeneous and heterogeneous catalysts for glucose-to-fructose
isomerization in aqueous media, ChemSusChem 6 (2013) 2369–2376.
[42] Y.B. Tewari, Thermodynamics of industrially-important, enzyme-catalyzed
reactions, Appl. Biochem. Biotechnol. 23 (1990) 187–203.
[43] C. E. Wyman, Aqueous Pretreatment of Plant Biomass for Biological and
Chemical Conversion to Fuels and Chemicals (This edition first published
2013), John Wiley & Sons Ltd.
[44] R.S. Assary, P.C. Redfern, J.R. Hammond, J. Greeley, L.A. Curtiss, Computational
studies of the thermochemistry for conversion of glucose to levulinic acid, J.
Phys. Chem. B 114 (2010) 9002–9009.
[45] A.A. Rosatella, S.P. Simeonov, R.F.M. Frade, C.A.M. Afonso, 5Hydroxymethylfurfural (HMF) as a building block platform: Biological
properties, synthesis and synthetic applications, Green Chem. 13 (2011)
754–793.
[46] A.S. Amarasekara, L.D. Williams, C.C. Ebede, Mechanism of the dehydration of
D-fructose to 5-hydroxymethylfurfural in dimethyl sulfoxide at 150 °C: an
NMR study, Carbohydr. Res. 343 (2008) 3021–3024.
[47] V. Choudhary, S.I. Sandler, D.G. Vlachos, Conversion of xylose to furfural using
Lewis and Brønsted acid catalysts in aqueous media, ACS Catal. 2 (2012) 2022–
2028.
[48] B. Girisuta, K. Dussan, D. Haverty, J.J. Leahy, M.H.B. Hayes, A kinetic study of
acid catalysed hydrolysis of sugar cane bagasse to levulinic acid, Chem. Eng. J.
217 (2013) 61–70.
[49] L. Peng, L. Lin, J. Zhang, J. Zhuang, B. Zhang, Y. Gong, Catalytic conversion of
cellulose to levulinic acid by metal chlorides, Molecules 15 (2010) 5258–5272.
[50] R. Weingarten, Y.T. Kim, G.A. Tompsett, A. Fernández, K.S. Han, E.W. Hagaman,
W.C. Conner Jr., J.A. Dumesic, G.W. Huber, Conversion of glucose into levulinic
acid with solid metal(IV) phosphate catalysts, J. Catal. 304 (2013) 123–134.
[51] J.N. Chheda, J.A. Dumesic, An overview of dehydration, aldol-condensation and
hydrogenation processes for production of liquid alkanes from biomassderived carbohydrates, Catal. Today 123 (2007) 59–70.