Influence of alkali and alkaline earth metal salts on glucose conversion to 5-hydroxymethylfurfural in an aqueous system

Article abstract of DOI:10.1016/j.catcom.2012.10.011

The aqueous phase acid-catalyzed reaction of glucose to 5-hydroxymethylfurfural (HMF) was performed with the addition of alkali and alkaline earth metal salts. Previous studies investigated this reaction in a biphasic system using salts to enhance the partitioning of HMF and suppress undesired products thereby increasing the overall selectivity. The results of this study suggest the presence of alkaline earth salt/glucose interactions that affect glucose conversion and HMF yield. While sulfate salts tended to lead to higher activity than chloride salts, the further addition of HCl led to decreased conversion through a competitive anion effect.

Full text of DOI:10.1016/j.catcom.2012.10.011

Catalysis Communications 30 (2013) 1–4
Contents lists available at SciVerse ScienceDirect
Catalysis Communications
journal homepage: www.elsevier.com/locate/catcom
Short Communication
Influence of alkali and alkaline earth metal salts on glucose conversion to
5-hydroxymethylfurfural in an aqueous system
Elliot Combs a, Basak Cinlar a, Yomaira Pagan-Torres b, James A. Dumesic b, Brent H. Shanks a,
⁎
a
Department of Chemical and Biological Engineering, Iowa State University, 1140L BRL, Ames, IA 50011, USA
b
Department of Chemical and Biological Engineering, University of Wisconsin-Madison, 3014 Engineering Hall, 1415 Engineering Drive, Madison, WI, 53706, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 19 July 2012
Received in revised form 27 September 2012
Accepted 17 October 2012
Available online 22 October 2012
The aqueous phase acid-catalyzed reaction of glucose to 5-hydroxymethylfurfural (HMF) was performed
with the addition of alkali and alkaline earth metal salts. Previous studies investigated this reaction in a bi-
phasic system using salts to enhance the partitioning of HMF and suppress undesired products thereby in-
creasing the overall selectivity. The results of this study suggest the presence of alkaline earth salt/glucose
interactions that affect glucose conversion and HMF yield. While sulfate salts tended to lead to higher activity
than chloride salts, the further addition of HCl led to decreased conversion through a competitive anion
effect.
Keywords:
Glucose
5-Hydroxymethylfurfural
Dehydration
© 2012 Elsevier B.V. All rights reserved.
Salt complexation
1. Introduction
complexes with glucose, thereby activating the glucosyl ring and hold-
ing it in a particular orientation that enables selective dehydration. If the
The efficient production of versatile intermediate compounds from
biomass constitutes a challenge for establishing a sustainable chemical
industry. One such potential intermediate in the production of poly-
mers and fine chemicals is 5-hydroxymethylfurfural (HMF) [1]. The
ideal production of HMF is via dehydration of a hexose such as glucose
or its polysaccharide due to high abundance and low cost. Although
considerable research has been performed on HMF production from
fructose, its efficient production from glucose presents a greater chal-
lenge due to the higher stability of the glucopyranose ring [2].
Most efforts directed towards HMF production have utilized one
of the following two strategies: combining isomerization of glucose
to fructose with subsequent dehydration of fructose, or employing
metal complexes to activate the glucose ring selectively for the for-
mation of HMF. In the former approach, both a Bronsted acid and a
base or Lewis acid are needed in concert as isomerization is a base-
or Lewis acid-catalyzed reaction and dehydration is catalyzed primar-
ily by protons [3]. In the latter approach, glucose complexation with
metals (such as copper or chromium in the presence of ionic liquids)
has been shown to activate the glucose ring for selective HMF produc-
tion [2].
same effect could be achieved with complexation agents in water, the
disadvantages of ionic liquids could be avoided [5].
Several groups have recently explored the biphasic conversion of
glucose to HMF using a reactive aqueous phase and an immiscible or-
ganic phase which selectively extracts HMF as it is produced. This
strategy enhances the selectivity towards HMF by minimizing the for-
mation of side products due to the tendency of HMF to degrade in
acidic aqueous environments [1]. In one approach, a homogeneous
Lewis acid is used in concert with a strong Brønsted acid (HCl) to fa-
cilitate the isomerization of glucose into fructose and the subsequent
dehydration to HMF [6]. Other work has demonstrated the possibility
of producing HMF from glucose through the use of water-compatible
Lewis acids in a nearly neutral pH environment, alleviating processing
concerns in terms of the materials of construction and environmental
impact [7]. Both studies feature the use of concentrated salts to in-
crease the in situ partitioning of HMF to the organic phase. Previous
work has investigated the catalytic effects of a range of inorganic
salts on the conversion of glucose and other sugars to HMF [8]. How-
ever, the results do not provide sufficient consideration for the salts'
effects in conjunction with an HCl catalyst, as may be present in a bi-
phasic reaction system.
The highest reported HMF yields to date have been reported using
ionic liquids [2]. However, the processing of ionic liquids is significantly
challenged due to high costs associated with product purification and
the sensitivity of ionic liquids to water [4]. It is believed the high activity
observed using ionic liquids can be attributed to their ability to form
An in-depth effect of alkali and alkaline earth metal salts in satu-
rated aqueous solutions has been examined for the dehydration of
glucose and fructose [9,10]. While the addition of NaCl, CaCl2 and
MgCl2 did not improve the HMF yields from fructose significantly,
their saturated solutions did result in improved yields from glucose.
This was attributed to complexation of the salts with glucose, in
which the cations bound to the oxygens of glucose, creating a
⁎
Corresponding author. Tel.: +1 515 294 1895; fax: +1 515 294 1269.
E-mail address: bshanks@iastate.edu (B.H. Shanks).
1566-7367/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.catcom.2012.10.011

2
E. Combs et al. / Catalysis Communications 30 (2013) 1–4
positively charged molecule that could subsequently interact with
water molecules in the proper orientation to facilitate dehydration
[9]. The anions' role in the mechanism was also explored. 13C NMR
spectra of glucose in saturated salt solutions at room temperature re-
vealed that chloride ions (notably in MgCl2 and LiCl) donated electron
density to the C-1, C-3, and C-5 carbons of glucose [11]. Sulfate ions,
although more bulky than halides, were found in a separate study
to donate electron density indiscriminately to the carbons of glucose
[9]. The sulfate ions were found to enhance the HMF selectivity
through the suppression of condensation reactions forming insoluble
humins, which was noticed in the reaction with chloride ions. Sulfate
ions were also posited to participate in hydrogen bonding with water
and with hydroxyl protons of glucose, thereby contributing further to
the complexation matrix [9].
Although this alkali/alkaline earth metal salt complexation has
been demonstrated in saturated solutions at room temperature, com-
prehensive studies have not been performed at lower concentrations
and elevated temperatures in the presence of HCl. This communica-
tion reports the direct effects of salt interaction under typical condi-
tions that have been utilized in past reaction studies [1,12].
thick-walled glass vial reactor under ambient pressure. The glucose
time-conversion results for these experiments are shown in
Fig. 1-a. The final results after 60 min are also detailed in Table 1 (En-
tries 1–4 and 6). Rapid heat transfer in the vials was assumed and no
time delay in the kinetic profile was taken into account. All of the chlo-
ride salts improved glucose conversion relative to the base case of HCl
without salt addition. CaCl2 and MgCl2 addition led to the greatest im-
provement in conversion. As shown in Fig. 1-b, the increased rate of glu-
cose conversion corresponded to higher HMF yields. The alkali metal
salts (NaCl and KCl) increased the yield only slightly. Of the salts, the
MgCl2 produced the highest HMF yields. However, the conversion
leveled off after ca. 40 min while the HMF yield began to decrease, indi-
cating a successive HMF degradation as the reaction continued. This
confirms findings from literature on the interaction of chloride ions
with the HMF molecule leading to its degradation [9,10].
It has been shown in the literature that the presence of alkaline
earth metal salts accelerates glucose decomposition rates [9,10].
When saturated solutions of MgCl2 and CaCl2 were examined for
their ability to decompose glucose, Ca2+ was found to be more selec-
tive towards HMF, while Mg2+ showed higher selectivity towards
levulinic acid, one of the HMF degradation products [10]. It is reason-
able to presume in this study that the higher activity of MgCl2 led to
more HMF degradation, producing levulinic acid after the 40 minute
mark. The higher rates of HMF formation and degradation from
Mg2+ may be due to a stronger interaction relative to Ca2+, leading
to the further degradation of HMF and the formation of side products.
Alkaline earth metal cations have been shown to form bidentate com-
plexes involving the C-1 and C-3 hydroxyl groups of glucose [13]. It is
possible therefore that the formation of these complexes can assist in
the conversion of glucose to HMF and its subsequent degradation.
Due to the enhanced HMF yields with MgCl2 addition, the effects
of varying its concentration were explored. Entries 5, 6, 7, and 8 of
Table 1 illustrate the effects of MgCl2 at concentrations of 0.4 M,
0.8 M, 1.6 M, and 3.2 M, respectively. The 5 wt.% glucose feed solu-
tion corresponded to a glucose concentration of approximately
0.3 M. As predicted, an increase in salt concentration led directly to
an increase in the overall conversion. However, the decrease in
2. Experimental
2.1. Materials
Hydrochloric acid (12 N, Fisher), glucose (99%, Fisher), sodium
chloride (99%, Fisher), potassium chloride (99%, Fisher), magnesium
chloride hexahydrate (99%, Fisher), calcium chloride dihydrate (99%,
Fisher), magnesium sulfate heptahydrate (98%, EMD), calcium sulfate
(98%, EMD), and calcium phosphate dibasic (30.0%–31.7% as Ca, Fisher)
were used as purchased.
2.2. Reaction experiments
To prepare the reaction mixture, hydrochloric acid was added to
nanopure water until a pH value of 1.5 was achieved using a glass
electrode (6.0233.100, Metrohm) connected to a Metrohm 798 MPT
Titrino automatic titrator. Glucose and salt were added to achieve
5 wt.% glucose and the desired concentration of salt. For the biphasic
reactions, the organic phase was prepared by mixing methyl-isobutyl
ketone (MIBK) with 2-butanol in the ratio of 7:3 (w/w) with an
aqueous/organic phase ratio of 1:2 (w/w) as described in the litera-
ture [1,12].
a)
The reaction experiments were performed at 160 °C in 10 ml
thick-walled glass vial batch reactors under stir bar agitation of
400 rpm. The vials were heated by submersion in a preheated bath
of Dow Corning 550 Fluid on a Fisher Scientific Isotemp hot plate
with an attached thermocouple. Samples were taken post-reaction
after cooling the vials rapidly in an ice bath.
0
10
20
30
40
50
60
70
2.3. UPLC analysis
Time (min)
b)
6
5
4
3
2
1
0
Samples were filtered through a 0.2 μm nylon filter (Cobert
Assoc.) prior to dilution. Analysis of the samples was performed
using a Waters ACQUITY UPLC H-Class system. HMF analysis was
performed using an ACQUITY UPLC BEH C18 column (Waters
186002350) at 35 °C and an ACQUITY UPLC photodiode array (PDA)
eλ detector at 254 nm. Glucose analysis was performed using an
Acquity UPLC BEH Amide column (Waters 186004801) at 65 °C and
a Waters 2414 Refractive Index (RI) detector at 50 °C.
3. Results and discussions
0
10
20
30
40
50
60
70
Time (min)
Initially, the effect of salts on glucose conversion was examined
by adding 0.8 M Group I–II chlorides (NaCl, KCl, CaCl2, and MgCl2)
into a 5 wt.% glucose solution of pH 1.5 at 160 °C in a sealed
Fig. 1. Glucose conversion (a) and HMF yield (b) in the presence of 0.8 M salt solutions,
5 wt.% glucose, T=160 °C; (HCl only (□), HCl with: NaCl (○), KCl (Δ), MgCl2 (+), CaCl2 (◊)).

E. Combs et al. / Catalysis Communications 30 (2013) 1–4
3
Table 1
although the selectivity dropped. This indicated a competitive effect
between the chloride and sulfate anions, leading to a tradeoff of glu-
cose conversion (with sulfate ions) and selective HMF production
(with the incorporation of both sulfate and chloride ions). It is possi-
ble that the chloride donation of electron density to glucose C-1, C-3,
and C-5 occurred more readily than the donation from the bulkier
sulfate group. As a result, the sulfate ions interacted predominantly
with the surrounding water molecules instead of the glucose itself,
leading to lower overall conversion while at the same time providing
conditions for HMF to be more selectively formed.
An organic phase was added to the system as discussed in the lit-
erature [1,12]. As shown in Fig. 2, both MgCl2 and MgSO4 experienced
gains in conversion and selectivity with MgSO4 performing better
than MgCl2, much like the aqueous system. This result was due to a
stabilizing effect of the organic phase on HMF, hindering its serial
conversion following glucose dehydration. The partition coefficient
was also calculated for all three systems, defined as the ratio of
HMF concentration in the organic phase to that in the aqueous
phase. This coefficient followed the increasing order of HCl only
bMgCl2 bMgSO4, with values of 1.41, 1.98, and 2.24 respectively, indi-
cating greater extracting potential of the organic phase through the
increased ionic strength of the aqueous phase. This demonstrated a
two-fold benefit of using metal salts in such a biphasic system, taking
into account both the enhancement in reactivity and the improved in
situ partitioning to the organic phase.
Effect of salts in the aqueous phase on glucose conversion and molar HMF selectivity
and yield after 60 min at 160 °C.
Entry Catalyst Salt
Salt
Conversion
Selectivity
(mol%)
S.D.
Yield
(mol%)
S.D.
concentration (mol%)
(M)
S.D.
1
2
3
4
5
6
7
8
HCl
HCl
HCl
HCl
HCl
HCl
HCl
HCl
HCl
HCl
HCl
HCl
–
–
–
10.0 1.3
13.4 1.2
14.4
23.1 1.9
24.9 0.8
23.6
2.3 0.1
3.3 0.2
3.4
NaCl
KCl
0.8
0.8
0.8
0.4
0.8
1.6
3.2
CaCl2
MgCl2
MgCl2
MgCl2
MgCl2
18.0 4.4
11.1 2.3
15.3 2.4
21.1 1.4
31.4 2.4
24.0 1.7
27.7 0.8
11.9 1.2
22.0 5.1
25.0 6.7
28.8 4.7
20.4 1.7
22.5 1.6
26.3 2.4
40.2 3.8
14.8 1.5
3.8 0.1
2.7 0.2
4.3 0.4
4.3 0.2
7.0 0.03
6.3 0.2
11.2 1.1
1.7 0.04
0.99 0.067
2.6 0.06
3.9 0.19
7.1 0.89
11.1 0.2
7.7 1.0
11.9 0.1
9
MgSO4 0.8
MgSO4 1.6
10
11a
12a
13
14
15
16
17
18
CaSO4
0.8
CaHPO4 0.8
5.16 1.25 20.2 6.0
MgCl2
MgCl2
MgCl2
MgCl2
0.4
0.8
1.6
3.2
11.3 1.2
15.6 1.2
20.7 1.0
23.6 0.9
45.1 0.7
40.7 4.2
23.6 2.2
25.0 2.8
34.1 2.7
47.2 1.2
17.2 2.0
29.5 3.0
–
–
–
–
MgSO4 0.8
MgSO4 1.6
–
a
Salt not completely soluble, reaction solution was an opaque salt/glucose suspension.
selectivity with 1.6 M MgCl2 demonstrated the aforementioned effect
of MgCl2 in the degradation of HMF in acidic media.
Further experiments were performed to probe the role of the
anion in activating glucose in the selective conversion to HMF. The
salts MgSO4, CaSO4, and CaHPO4 were selected to screen different an-
ions with alkaline earth cations. Reactions were performed under the
same conditions as the chloride salts at concentrations of 0.8 M in the
presence of HCl. The solubility of CaSO4 and CaHPO4 in water was lim-
ited, even at elevated temperatures, and they formed viscous suspen-
sions when prepared in solution. The reaction results with these salts
are given in Entries 9–12 in Table 1. The low activities of CaSO4 and
CaHPO4 were likely due to internal transport limitations within the
solution due to the increased viscosity and density of insoluble parti-
cles. The activity of MgSO4, however, was found to be higher than that
of MgCl2, and a higher concentration of MgSO4 led to greater HMF se-
lectivity over MgCl2 (Entries 9 and 10 vs. 6 and 7). This indicated a
stronger effect of the sulfate ion on activating the glucose ring leading
to higher conversion. The literature has discussed the complexation
of magnesium ions with the glucose ring via ring activation [11]. Be-
cause the selectivity was found to be higher with MgSO4 than with
MgCl2 in the presence of HCl, this suggested that sulfate ions do not
complex with the HMF molecule to facilitate its degradation as is
the case with chloride ions.
4. Conclusions
The introduction of alkali and alkaline earth salts, especially mag-
nesium salts, into the acidic media was found to enhance both the
glucose conversion and the yield of HMF. This enhancement was like-
ly via glucose complexation with the salts. The activity of MgSO4 was
greater than MgCl2, and the HMF selectivity increased with higher
sulfate ion concentrations. Selectivity in this system increased further
with the addition of HCl, although the conversion decreased. Im-
proved yields in the biphasic system with these added salts could
be explained by a combination of increased activity and enhanced
in situ partitioning of HMF into the organic phase to stabilize HMF
and prevent side product formation.
Acknowledgments
This work was financially supported by the National Science Foun-
dation under Award No. EEC-0813570.
In order to gain a better perspective of these anion effects, reac-
tions were performed under the same conditions using only water
(without adding HCl) as the solvent. The salts MgCl2 and MgSO4
were both tested, the former varying in concentrations of 0.4 M,
0.8 M, 1.6 M, and 3.2 M, and the latter at 0.8 M and 1.6 M. The results
from these reactions are shown in Entries 13–18 of Table 1. Interest-
ingly, the lack of HCl did not decrease the overall glucose conversion.
This contradicts results from literature reporting the dependence of
glucose dehydration to HMF on the number of available protons
[14]. This implied that the acid-catalyzed dehydration of glucose can
occur without the addition of a strong acid to supply excess protons.
Instead, the addition of salts contributed ionic strength to and alters
the activity coefficient of the solution, lowering the effective pH and
facilitating the dehydration. In these reactions without a strong acid
catalyst, the values for conversion with MgCl2 did not change signifi-
cantly (except at 3.2 M), whereas an increase in selectivity was ob-
served at higher concentrations of MgCl2. The reaction with MgSO4,
on the other hand, experienced a great increase in glucose conversion
versus the reaction with HCl and the same concentration of MgSO4,
60
Conversion
50
40
30
20
10
0
Selectivity
Yield
mono
biphasic
mono
biphasic
mono
biphasic
HCl
MgCl2
MgSO4
Fig. 2. Effect of biphasic extraction on glucose conversion and HMF selectivity and yield
in the presence of HCl only, HCl with 0.8 M MgCl2, and HCl with 0.8 M MgSO4 after
60 min at 160 °C.

4
E. Combs et al. / Catalysis Communications 30 (2013) 1–4
[8] C.B. Rasrendra, I.G.B.N. Makertihartha, S. Adisasmito, H.J. Heeres, Topics in Catal-
References
ysis 53 (2010) 1241–1247.
[9] S.K. Tyrlik, D. Szerszen, M. Olejnik, W. Danikiewicz, Journal of Molecular Catalysis
A: Chemical 106 (1996) 223–233.
[10] S.K. Tyrlik, D. Szerszen, B. Kurzak, K. Bal, Starch/Staerke 47 (1995) 171–174.
[11] S.K. Tyrlik, D. Szerszen, S. Szymanski, New Journal of Chemistry 19 (1995)
1019–1021.
[12] J.N. Chheda, Y. Roman-Leshkov, J.A. Dumesic, Green Chemistry 9 (2007) 342–350.
[13] C.Y. Yang, X.S. Lu, W.G. Lin, W.G. Yang, W.G. Yao, Chemical Research in Chinese
Universities 22 (2006) 524–532.
[14] M. Watanabe, Y. Aizawa, T. Iida, T.M. Aida, C. Levy, K. Sue, H. Inomata, Carbohy-
drate Research 340 (2005) 1925–1930.
[1] Y. Roman-Leshkov, J.N. Chheda, J.A. Dumesic, Science 312 (2006) 1933–1937.
[2] H. Zhao, J.E. Holladay, H. Brown, Z.C. Zhang, Science 316 (2007) 1597–1600.
[3] M. Moliner, Y. Roman-Leshkov, M.E. Davis, Proceedings of the National Academy
of Science 107 (14) (2007) 6164–6168.
[4] C. Moreau, A. Finiels, L. Vanoye, Journal of Molecular Catalysis A: Chemical 253
(2006) 165–169.
[5] C. Lansalot-Matras, C. Moreau, Catalysis Communications 4 (2003) 517–520.
[6] Y.J. Pagan-Torres, T. Wang, J.M. Gallo, B.H. Shanks, J.A. Dumesic, Production of
5-hydroxymethylfurfural from Glucose Using a Combination of Lewis and Brønsted
Acid Catalysts in Water in a Biphasic Reactor with an Alkylphenol Solvent, ACS Catalysis
2 (2012) 930–934.
[7] T. Wang, Y.J. Pagan-Torres, E.J. Combs, J.A. Dumesic, B.H. Shanks, Water-compatible
Lewis acid-catalyzed conversion of carbohydrates to 5-hydroxymethylfurfural in a
biphasic solvent system, Topics in Catalysis 55 (2012) 657–662.