Green Chemistry

CRITICAL REVIEW
Cite this: Green Chem., 2014, 16,
548
Received 10th July 2013,
Accepted 5th November 2013
DOI: 10.1039/c3gc41365a
www.rsc.org/greenchem
Catalytic dehydration of C
6
carbohydrates for the
production of hydroxymethylfurfural (HMF) as a
versatile platform chemical
Tianfu Wang,
a,b
Michael W. Nolte
a
and Brent H. Shanks*
a,b
Interest in utilizing biorenewable feedstocks to produce fuels and chemicals has risen greatly in the past
decade due to the economic, political and environmental concerns associated with diminishing pet-
roleum reserves. A fundamental challenge lying ahead in the development of ecient processes to utilize
biomass feedstock is that, unlike their petroleum counterparts, biomass contains an excess amount of
oxygen. Therefore, catalytic strategies such as dehydration and hydrogenolysis amongst others have been
extensively studied as platform technologies for deoxygenation. In this review, we primarily discuss the
catalytic dehydration of C
6
carbohydrates to 5-hydroxymethylfurfural, which has attracted much attention
due to the versatility of using furanic compounds as an important platform intermediate to synthesize
various chemicals. The emphasis is on the fundamental mechanistic chemistry so as to provide insights
for further catalyst/catalytic system design. After separately discussing fructose and glucose dehydration,
this review summarizes recent progress with bi-functional catalyst systems for tandem glucose/fructose
isomerization and subsequent fructose dehydration, thereby realizing highly selective HMF production
directly from the more abundant and cheaper C
6
sugar feedstock, glucose.
1. Introduction
1.1. Incorporation of biomass into conventional fuel and
chemical production: alternative to petroleum feedstock
Throughout the 20
th
century, the carbon supply for the pro-
duction of chemicals and fuels to support societys needs has
been deeply rooted in fossil carbon reserves. Through develop-
ing these feedstocks, numerous technological breakthroughs
have occurred, such as the development of porous zeolitic
materials for the refinery,
1
the synthesis of unprecedented
quantities of new synthetic materials with wide ranging applica-
bility (e.g., nylon, carbon fiber, etc.), and the exploitation of
petroleum-sourced chemicals for application in human health
and nutritional applications
2
to name a few. In conjunction
with these technological advancements, the global population
has experienced a growth to about 7 billion people while living
standards have also improved significantly, which in turn has
resulted in the accelerated depletion of fossil carbon reserves.
Due to the rapidly expanding worldwide population, particularly
in emerging countries, it is projected that over the next two
decades, the world energy consumption and petroleum demand
will increase by over 30%.
3,4
Therefore, it is imperative that
alternative carbon sources are developed to at least augment
petroleum-sourced chemicals. Taking all these considerations
into account, biomass exists as a unique and promising candi-
date, with potential for improved sustainability and biodegrad-
ability relative to its petroleum counterpart.
5
In the United
States alone, an estimated 1.3 billion tons of dry biomass is pro-
duced annually from either agriculture or forest waste.
6
Catalyst development faces a multitude of unique technical,
economic, and logistical challenges to utilize biorenewable
feedstocks. In contrast to the petroleum-based industries,
where unfunctionalized alkanes are the primary starting
material, a biorenewables industry will use a feedstock with
extensive and diverse functionality derived from carbohydrates
and lignin, which have excess oxygen relative to current chemi-
cal products. In order to use these compounds for chemicals,
the excess functionality will need to be selectively removed to
produce the same or similar molecules to those used in the
petrochemical industry. New chemistries and novel catalytic
systems, that may not be as thoroughly studied as those of the
established petrochemical industry, need to be developed by
the catalysis community to address the over-functionalization
challenge in biorenewables conversion.
7,8
Of particular interest
are the dehydration and hydrogenolysis reactions, which can
eectively reduce the oxygen content and can be applicable
to various biorenewables.
9,10
The complexity and diverse
a
Chemical and Biological Engineering Department, Iowa State University, Ames, IA
50011, USA. E-mail: [email protected]
b
NSF Engineering Research Center for Biorenewable Chemicals (CBiRC), Ames, IA
50011, USA
548 | Green Chem., 2014, 16, 548572 This journal is The Royal Society of Chemistry 2014
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
View Journal | View Issue
reactivity of biomass molecules necessitate the systematic
development of appropriate approaches for dierent com-
ponents in plant biomass.
1.2. Lignocellulosic biomass
First generation biorenewable feedstocks generally refer to vege-
table oil and starch/sucrose, from which biodiesel and ethanol,
respectively, can be produced. However, concerns arise when
these feedstocks are used because vegetable oil, sucrose and
starch are also food resources. Moreover, the quantity that can
be generated economically is rather limited relative to the total
production of fuels and chemicals. Hence, they can only be qua-
lified as a transitional solution to the overall energy and
materials requirements needed by a growing world population.
From the long-term perspective, second generation biorenew-
able feedstocks, which are typically derived from lignocellulose,
are required to meet the renewable carbon demands.
Lignocellulose is the major component found in plant and
algae biomass, and the estimated annual production of ligno-
cellulosic biomass is 2 10
11
metric tons.
11,12
Lignocellulose
consists of three major fragments: cellulose, hemicellulose
and lignin with its representative structure shown in Fig. 1.
Depending on plant type, the relative portion of cellulose,
hemicellulose and lignin found in the biomass is dierent.
For example, in dried hardwoods, cellulose accounts for
4045% by weight, hemicellulose is typically 2440%, and
lignin is 1825%.
13
In cornstover, the percentages are around
2748%, 1317%, 1431%, respectively.
13
Cellulose is the most abundant biopolymer, consisting of
D-glucose units connected via -1,4 linkages. Unlike the -1,4
linkages in starch, extensive hydrogen bonding exists in the
cellulose structure, making hydrolysis dicult.
14
Dilute acid
hydrolysis and enzymatic hydrolysis are commonly used to lib-
erate the monosaccharide glucose units.
15,16
The acid hydro-
lysis step can present hurdles for subsequent conversion steps
since the use of base is typically required to neutralize the
excess acids. Quite often, side products, such as dehydration
products of glucose, can be also found in the euent. The pres-
ence of side products can be harmful for the microorganisms
used in downstream fermentation steps.
17
Enzymatic hydrolysis,
on the other hand, can suer from low turnover rates, high cost
associated with the need for large quantities of enzyme, and
diculty in catalyst separation from solution.
18,19
Hemicellulose is composed of a mixture of several dierent
types of carbohydrates including mainly xylose and arabi-
nose.
20
These C
5
sugars are not as easily fermentable as their
C
6
counterpart glucose, however recent literature articles have
demonstrated that xylitol (hydrogenation product of xylose)
and L-arabinose can serve as functional sugar additives or
replacements for high-fructose corn syrup-derived foods and
beverages.
21
It has been shown that the addition of 23 wt% of
L-arabinose into sucrose can reduce glucose uptake by about
70% while maintaining a similar sweetness level to sucrose. As
a result, these C
5
sugar derivatives can serve as a health food
resource.
21
The utilization of biomass for high value food or
pharmaceutical applications like these can greatly improve the
comprehensive economics of lignocellulose conversion and
facilitate the likelihood of process commercialization. The
dehydration of C
5
sugars can yield furfural, which is currently
commercially produced as a platform chemical having appli-
cations ranging from solvents to resins and fuel additives.
22,23
Lignin is a highly cross-linked phenolic polymer
24
and the
robust aromatic COC ether bonds present throughout make
it very dicult to deconstruct into monomeric constitutes.
25
Therefore, ecient methods to depolymerize lignin are still a
major goal. On the other hand, its aromatic-based structure
gives lignin a high heating value. Lignin can potentially be
upgraded into transportation fuels by hydrogenation and/or
hydrodeoxygenation. A study by Lercher and coworkers found
a method to quantitatively depolymerize lignin to its mono-
meric constitutes using a combination of base (as a catalyst)
and boric acid (as a capping agent).
26
Later, the same group
also reported that using a dual-function catalyst system, com-
prised of Pd/C and H
3
PO
4
, can be used to hydrogenate the
monomer phenols in an aqueous media to form cycloalkanes,
an ideal fuel.
27,28
Dumesic et al. found in the processing steps
to convert cellulose and hemicellulose to chemicals and fuels, a
solvent derived from lignin is particularly interesting for use in
biphasic reaction systems in which the organic phase is the
lignin-based phenol and the aqueous layer is the reactive media
containing the carbohydrates and catalyst. The in situ produced
furanics (5-hydroxymethylfurfural, furfural) or lactones
(gamma-valerolactone, GVL) can be selectively extracted into the
organic phase, making the process sustainable with both reac-
tants and organic solvent being derived from lignocellulose.
29
Processes designed to convert the entire carbon supply into
useful chemicals and fuels are preferred, in order to realize
the economical utilization of lignocellulose. If one portion is
not converted, significant amounts of waste will be generated
that will necessitate appropriate handling.
2. HMF as a platform chemical
HMF (Fig. 2) is a heterocyclic furanic molecule substituted in
2,5-position with hydroxide and aldehyde functionalities.
Fig. 1 Lignocellulose biomass composition: cellulose, hemicellulose,
and lignin.
Green Chemistry Critical Review
This journal is The Royal Society of Chemistry 2014 Green Chem., 2014, 16, 548572 | 549
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
Because of its unique chemical structure, the eort in utilizing
HMF as a starting point for the synthesis of functional replace-
ment or direct replacement alternatives to the current com-
modity chemicals mainly results from three molecular
attributes. First, HMF is an ,-bifunctional molecule with
substituents in both the 2 and 5 positions, so it can be either
oxidized to a dicarboxylic acid or reduced to a diol. Both of
which can be used for the synthesis of polymers. Second, HMF
is a relatively unsaturated aromatic compound and can be
upgraded to fuel molecules via hydrogenation. Finally, the
heterocyclic structure of furans can be found in an array of bio-
logically active molecules with pharmaceutical applications.
The current review will focus on the first two applications and
the last topic can be found in a review published recently.
30
2.1. HMF derivatives as monomers for polymeric materials
2.1.1. HMF to furandicarboxylic acid (FDCA). It has been
shown that PEF ( polyethylene furanoate), made from FDCA
and ethylene glycol, has similar physical, mechanical and
chemical properties to the petroleum-based analog, PET
( polyethylene terephthalate).
3133
FDCA is produced through
the oxidation of HMF, which has been reported for a number
of homogeneous and heterogeneous systems. In a study by
Corma and co-workers, 99 mol% yield of FDCA was achieved
using Au/CeO
2
and Au/TiO
2
as catalysts, in aqueous media.
34
Gorbanev et al. systematically studied the eect of added bases
and oxygen pressure on the HMF oxidation reactions with Au/
TiO
2
under ambient temperature at 30 C.
35
They concluded
the bases are required for such conversion and reported a
71 mol% yield of FDCA at full conversion of HMF. In order to
improve the activity and selectivity as well as stability of the
catalysts, bimetallic catalysts have also been reported for HMF
oxidation. Pasini et al. reported the bimetallic goldcopper on
TiO
2
, in which a yield of 9099% of FDCA was obtained at
95 C for 4 h.
36
In a study published by Abu-Omar and co-
workers, a modification of a homogenous oxidation system,
consisting of bimetallic zinc and cobalt bromides and acetic
acid, gave greater than a 60 mol% yield with 1 atm oxygen.
37
Davis and co-workers used supported precious metal catalysts
(Pt/C, Pd/C and Au/TiO
2
) under high pH conditions with mole-
cular oxygen in an aqueous media leading to HMF being quan-
titatively converted FDCA in high yields. For the Pt/C system,
about an 80 mol% FDCA yield was obtained.
38
A plausible
reaction mechanism and pathway has been proposed by the
same research group (Fig. 3).
39
One of the disadvantages of
these systems is that they need the addition of excess amounts
of base. To address this problem, Abu-Omar and co-workers
recently reported that 79 mol% FDCA can be achieved with an
Fig. 2 Chemical structure of HMF.
Fig. 3 Overall reaction scheme and proposed mechanism for the oxidation of HMF in aqueous solution in the presence of excess base (OH

) and
either Pt or Au.
39
Reproduced from ref. 39.
Critical Review Green Chemistry
550 | Green Chem., 2014, 16, 548572 This journal is The Royal Society of Chemistry 2014
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
Fe
III
-porphyrin complex supported on porous carbon materials
without the need to add a base.
40
2.1.2. Diols from HMF. In addition to work reported for
the oxidation of HMF to produce diacids, interest in converting
HMF to diols has also been examined. Alamillo and co-
workers reported in their study, dihydroxymethyltetrahydro-
furan (DHMTHF) can be obtained in as high as 90% yield by
using a -alumina supported ruthenium catalyst to catalyze the
hydrogenation starting with HMF.
41
DHMTHF can be used as
a monomer for polymer synthesis, or other applications. Using
copper chromite as a catalyst for HMF hydrogenation can be
advantageous when the hydrogenation is preferred to proceed
without saturating the furan ring.
42
2.2. Fuel molecules from HMF
2.2.1. Furanic compounds. In the past decade, etherifica-
tion of HMF to 5-alkoxymethylfurfural ethers and hydrogeno-
lysis of HMF (to form dimethylfuran, DMF) have been proposed
for producing liquid transportation fuels (Fig. 4). In a report by
Bell and co-workers, in which an alcohol is used as both reac-
tion media and reactant, HMF can undergo acid catalyzed ether-
ification to form furfural ethers.
43
When Pt/C is used as the
catalyst in the reaction mixture, reductive etherification also
occurs, resulting in furfural di-ethers. These ether molecules
are good candidates as liquid fuels with high heating values.
Kraus et al. reported that in a binary system of sulfonic-functio-
nalized ionic liquids and hexane, fructose could be directly con-
verted to 5-alkoxymethylfurfural ethers with yields of greater
than 50 mol%. This reactive media can be recycled and the
authors showed that the 5-alkoxymethylfurfural ethers yields
didnt drop significantly over five consecutive runs.
44
Tsapatsis
and co-workers used glucose as a feedstock directly, with a com-
bined catalyst system involving Sn-beta zeolite, a Lewis acid
catalyst, and Amberlyst 131, a Brnsted acid. Glucose was
hypothesized to undergo isomerization to fructose followed by
dehydration/etherification to produce 5-alkoxymethylfurfural
ethers with yields of 31 mol%.
45
Amongst all the hydrogenation catalysts, copper has been
shown to preferentially perform CO bond hydrogenolysis over
CC bonds and thus can minimize aromatic hydrogenation.
9
Because of this, numerous copper-based catalysts have
been reported to selectively hydrogenate HMF to DMF as an
attractive fuel additive.
46
In 2007, Romn-Leshkov et al.
reported HMF can undergo selective hydrogenolysis over
copperruthenium catalysts in butanol solvent, yielding over
70 mol% DMF.
47
Another study by Shanks and co-workers
focused on the elucidation of the active speciation of copper in
copper-chromite catalysts and their activity on CO hydrogeno-
lysis with furanic model compounds.
9
Chidambaram et al.
reported in their study glucose can be directly converted to
DMF in a two-step, one-pot manner in an ionic liquidaceto-
nitrile mixed solvent.
48
While almost quantitative yield of
HMF from glucose was observed, less success was achieved in
the second-step hydrogenolysis to DMF (with 15 mol% yield
on Pd/C catalyst).
2.2.2. Liquid alkanes. Dumesic and co-workers devised
synthesis routes for C
9
, C
12
and C
15
alkanes using HMF and
acetone as starting materials through a two-step condensation/
hydrogenation and dehydration pathway.
49
If a 1 : 1 HMF
to acetone ratio was used, the aldol condensation gave a
C
9
species, which can undergo a second-step hydrogenation/
dehydration to form C
9
alkanes. Similarly, C
12
and C
15
alkanes
can be produced using self-condensation of HMF resulting
from use of a 2 : 1 HMF to acetone ratio. Although the econ-
omics of this pathway will rely on the high yield of HMF pro-
duced from lignocellulose feedstock and inexpensive
hydrogen, this process clearly provides an alternative to
produce liquid alkanes directly from cellulose-derived glucose.
The base catalyzed aldol condensation method to upgrade
HMF to alkanes later has been extended to the biphasic
system
50
and solid bases.
51
In the first case, Dumesic and co-
workers reported a biphasic system where the aldol conden-
sates were continuously extracted from the basic aqueous
phase to prevent degradation. In the later case, Shen et al.
examined the aldol condensation reaction between furfural
and acetone over several solid base catalysts: MgOZrO
2
, NaY
and nitrogen-substituted NaY at 120 C. Their results showed
MgOZrO
2
to be the most selective catalyst with almost 100%
selectivity for moderate conversions (50%). Recently reported
work by Corma et al.
52,53
and Yang et al.
54
have used similar
aldol-type strategies to obtain diesel range liquid alkanes from
HMF or furfural.
In general, HMF is a versatile platform chemical, which can
be converted to a broad array of fuel and chemical products.
While the promise of using HMF is quite significant, the econ-
omical production of HMF at large scale is the major obstacle
in unleashing the great potential of this platform chemical.
Understanding the chemistry of carbohydrate dehydration to
HMF is key providing insights into the proper catalyst design
principles, which can then be coupled with reactor configur-
ation and process optimization to improve the eciency of the
transformation.
3. Hexose dehydration mechanism
HMF is a triple dehydration product of hexoses (Fig. 5) with
both glucose and fructose explored as feedstocks. The
Fig. 4 Conversion of HMF to DMF and 5-alkoxymethylfurfural ethers as
biofuels.
Green Chemistry Critical Review
This journal is The Royal Society of Chemistry 2014 Green Chem., 2014, 16, 548572 | 551
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
dehydration reaction has been catalyzed by both homogeneous
and heterogeneous catalysts in aqueous, organic, and, most
recently, ionic liquid media. A number of excellent review
articles have been dedicated to the production of HMF with
several recent ones including, HMF and furfural synthesis
work summarized by Hu et al.,
55
mechanistic aspects of C
6
sugar conversion to HMF by Qian,
56
heterogeneous catalyst
systems for both HMF and furfural production by Karinen and
co-workers,
57
and HMF production using ionic liquids as
solvents by Sthlberg et al.
58
and Zakrzewska et al.
12
Most
recently, a published review article by Van Putten et al.
30
has
updated the current pilot and pre-commercial HMF production
status and their applications. In the present review, we first
briefly review the current hypothesis about hexose dehydration
and then discuss fructose and glucose dehydration, before
reviewing recent progress made in integrating glucosefructose
isomerization and subsequent dehydration to produce HMF.
No definitive conclusion has been made on the reaction
mechanism for hexose dehydration in the literature. The
hexose dehydration reaction to form HMF is generally
accompanied by numerous other side reactions, which lead to
the generation of humins and organic acids.
59,60
In addition,
other molecules produced during hexose dehydration, particu-
larly organic acids such as levulinic (LA) and formic acid, can
self-catalyze the dehydration of hexose.
59
Other factors thought
to be important are solvent type, temperature, and catalysts.
61,62
Several mechanisms have been proposed for hexose dehydration
and, in general, can be divided into pathways involving
cyclic
56,6164
or acyclic intermediates
6567
(Fig. 6). Most of these
studies were carried out with fructose in aqueous media.
In one study supporting the cyclic mechanism, Amarase-
kara et al. reported that the anomeric composition of D-fruc-
tose altered when heated to 150 C with an increase in the
-furanose form relative to the -pyranose tautomer.
64
Using
1
H and
13
C NMR spectra, they identified the key intermediate
(4R,5R)-4-hydroxy-5-hydroxymethyl-4,5-dihydrofuran-2-carb-
aldehyde when using dimethylsulfoxide (DMSO) as the
solvent, indicating the reaction proceed via a cyclic intermedi-
ate. In another combined experimental and computational
study by Antal et al., fructose conversion was examined with
mineral acids as the catalyst.
66
Reaction kinetics was measured
by monitoring the temporal composition of key products such
as HMF, furfural, and levulinic acid. The authors found the
cyclic pathway was more likely to account for HMF formation
by observing a rapid kinetic profile when using 2,5-anhydro-
D-mannose (an intermediate in the hypothesized cyclic
pathway) as the reactant. The authors also performed the reac-
tion in deuterium oxide (D
2
O) and the lack of CD bonds in
the final HMF molecules contradicted the acyclic mechanism
where ketoenol tautomerism will otherwise incorporate a D
atom into the formed HMF. In 2011, Assary et al. examined the
energetics of assumed intermediates of fructose dehydration
using high-level quantum chemical methods.
68
They found
the acidity of the solution was essential to lower the activation
energy barrier by proceeding via protonated intermediates.
The calculated activation barrier at neutral conditions at 298 K
was found to be as large as 74 kcal mol
1
, making the trans-
formation under this condition very unfavorable. However, the
cyclic forms of intermediates formed by combined protonation
and hydrogen transfer can lower the activation barriers, facili-
tating fructose to HMF transformation.
Alternative mechanisms for HMF formation also involve an
acyclic pathway, in which a 1,2-enediol has been suggested as
an intermediate. Moreau et al. studied fructose conversion
using H-mordenites as a solid catalyst in aqueous media.
65
Based on the detected formation of other hexoses like
mannose and glucose, they suggested that fructose first iso-
merized to glucose and mannose via a 1,2-enediol. This
hypothesis can also explain the observation of furfural and
hydroxyacetylfuran, which have been postulated through their
proposed reaction scheme. Unfortunately, no spectroscopic or
labeling experiments were performed to confirm this hypo-
thesis. Work by Qian et al. employed ab initio calculations to
study the possible pathways of both C
6
and C
5
sugars and they
Fig. 6 Possible hexose dehydration pathways (decomposition via fruc-
tose isomerization (); acyclic route (); cyclic route ()).
10
Reprinted
from Applied Catalysis A: General, 450, Basak Cinlar, Tianfu Wang Brent
H. Shanks, Kinetics of monosaccharide conversion in the presence of
homogeneous Bronsted acids, 237242, Copyright (2013), with per-
mission from Elsevier.
Fig. 5 Triple dehydration of hexose to synthesize HMF.
Critical Review Green Chemistry
552 | Green Chem., 2014, 16, 548572 This journal is The Royal Society of Chemistry 2014
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
concluded the rate-limiting step of the transformation from
hexose to HMF is the protonation of the hydroxyl group and
suggested the reaction proceeded via the acyclic pathway.
69
Extensive research and mechanistic studies have also been
conducted on C
5
sugar dehydration, which can be analogous
to work done on C
6
sugar and provide further information
about the transformation.
7075
While the more abundant glucose is a preferred feedstock
over fructose for HMF production, the rate of glucose conver-
sion and selectivity to HMF are both lower than with fructose.
The explanation for this observation is attributed to the fact
that the cyclic furanose tautomer and the pyranose form of
fructose co-exist in water. This leads to superior reactivity and
HMF selectivity for fructose relative to glucose, which exists in
aqueous solution with the pyranose form as the prevalent tauto-
mer. The observed slower kinetics of glucose conversion relative
to fructose also supports the cyclic mechanism for fructose
dehydration to HMF, which hypothesizes a pathway involving
the furanose form as an intermediate. To develop an ecient
glucose to HMF conversion process, an eective method is
needed that easily integrates glucose isomerization to fructose
and subsequent fructose dehydration to HMF. However, several
considerations should be taken into account for the choice of
glucosefructose isomerization catalysts so as to be compatible
with the Brnsted acid catalyst needed for fructose dehydration.
Traditional glucose isomerization catalysts are bases or
enzymes,
76,77
but these are not generally suitable for integration
due to the base will neutralizing the Brnsted acid, and because
enzyme activity and stability can be adversely aected in
strongly acidic conditions. New catalysts that can function
under the prerequisite condition need to be developed and in
the past several years, systems such as CrCl
x
/ionic liquid, Lewis
acidic zeolites and MCl
x
in water have achieved significant
progress with relatively high HMF yields (>60 mol% directly
from glucose). The common point of these three systems is the
use of a Lewis acid for glucose isomerization to fructose in combi-
nation with a Brnsted acid for fructose dehydration to HMF.
In 2007, Zhao et al. reported the use of sugar-solubilizing
1-ethyl-3-methylimidazolium chloride, [EMIM]Cl, as a solvent,
where fructose can be converted to HMF in yields of 80 mol%
with 18 mol% H
2
SO
4
(relative to fructose) as a catalyst at
80 C.
78
Glucose conversion was also studied and CrCl
x
Lewis
acidic salts were identified as the most eective catalyst to
obtain high yields of HMF (68 mol% to 70 mol%). Using NMR
spectroscopy, the authors proposed the role of CrCl
x
in cata-
lyzing glucose conversion (Fig. 7). They ascribed the activity of
CrCl
x
to the interaction between glucose and CrCl
x
which
leads to: (1) rapid mutarotation of glucose generating an equi-
librium mixture of - and -anomers, and (2) isomerization of
glucose to fructose via a formal hydride transfer from C-2 posi-
tion to C-1 position through a [EMIM]

mediated enediol
intermediate. Later in 2010, Pidko and co-workers proposed a
dierent mechanism for glucose isomerization with Cr
2+
cata-
lyst in [EMIM]Cl, based on data obtained from kinetic experi-
ments, in situ X-ray absorption spectroscopy (XAS) and density
functional theory (DFT).
79
They concluded that a single Cr
center was responsible for the equilibrium between the open
and closed form of glucose. In contrast to [CrCl
3
]

proposed by
Zhao et al., Pidko et al. suggested a mechanism for glucose
fructose isomerization involving the formation of a transient
binuclear complex between two Cr
2+
centers, analogous to the
active site of hexose isomerase enzymes.
79
A similar enzymatic
mechanism was proposed by Meilleur et al. in their crystallo-
graphic study of D-xylose isomerase.
80
Another breakthrough was made by Davis et al. in 2010.
81
They found a Sn-beta solid Lewis acid zeolite can isomerize
glucose to fructose with high activity in water under both
neutral and acidic conditions. The performance of the catalyst
under acidic conditions opened up new opportunities for the
facile integration of the isomerization step with the Brnsted
acid-catalyzed fructose dehydration. Indeed, the same group
later reported that using Sn-beta and HCl in a biphasic system,
glucose can be directly converted to HMF with over 70 mol%
selectivities.
82
The mechanism of the isomerization has been
studied using isotopic labeling experiments (D and
13
C) and
NMR spectroscopy.
76,77
A hydride transfer mechanism has
been suggested by the experimental results (Fig. 8). Kinetic
experiments performed on two forms of glucose (one
Fig. 7 Proposed metal chlorides interaction with glucose to produce HMF.
78
From H. Zhao, J. E. Holladay, H. Brown and Z. C. Zhang, Science,
2007, 316, 15971600. Reprinted with permission from AAAS.
Green Chemistry Critical Review
This journal is The Royal Society of Chemistry 2014 Green Chem., 2014, 16, 548572 | 553
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
protonated and one deuterated at the C-2 position) to compare
initial reaction rate displayed a kinetic isotope eect of 1.98,
suggesting the proton at the C-2 position was involved in the
rate-limiting step. Mass spectroscopy proved the prevalent C-D
incorporation in the product fructose, supporting the hydride
transfer mechanism. In addition,
1
H and
13
C NMR data were
in agreement with the proposed mechanism.
In addition to the above mentioned work using hetero-
geneous Sn-beta as the isomerization catalyst, Shanks and co-
workers,
83,84
Abu-Omar and co-workers,
85,86
and Vlachos and
co-workers
87,88
have been focusing on using homogeneous
Lewis acidic metal chloride salts (e.g., AlCl
3
, CrCl
3
, YbCl
3
) in
combination with a Brnsted acid (HCl) to perform the iso-
merization/dehydration reaction. These reactions were carried
out in biphasic systems and high HMF yields (60 mol%) were
obtained directly from glucose, with fructose being observed
as an intermediate. Although more kinetic and mechanistic
work is required to understand the reaction network, the
authors proposed a similar hydride mechanism for the
glucosefructose isomerization.
Although the dehydration of hexose to produce HMF is the
reaction of interest, other side reactions can occur under the
reaction conditions. For example, HMF can undergo a sub-
sequent rehydration pathway to form levulinic acid in the
presence of water.
89
It is also well-known that soluble and inso-
luble humins can be formed from either direct decomposition
of hexose or cross-reaction between the formed HMF and
hexose.
90
Since water is a side-product that is concurrently pro-
duced during the hexose dehydration process, it is also necess-
ary to consider this limitation for reactions being carried out
under non-aqueous solvent such as organic solvents or ionic
liquids.
In accordance with the considerations stated above, strat-
egies that have been employed to improve HMF yield have
been focused largely on how to minimize the sequential reac-
tions of HMF to form undesirable byproducts. In general,
under the relatively harsh reaction condition for hexose dehy-
dration, further reaction of HMF can readily take place. Thus a
multitude of physical and chemical approaches have been con-
sidered to improve HMF stability: (1) adding a cosolvent (gen-
erally an organic phase) into the aqueous media, which has
several advantages relative to the dehydration reaction network
including changing the sugar tautomerism equilibrium to
favor the enhanced occurrence of the more reactive tautomer
due to the presence of a polar non-protonic solvent and sup-
pressing side reactions that lead to the formation of humin
and levulinic acid due to dilution from the cosolvent; (2)
chemically converting the HMF produced during the dehy-
dration to a less reactive derivative under reaction conditions
with one example being performing the reaction in an alcohol
to form HMF-derived ether compounds; and (3) removing the
HMF from the reactive aqueous media by contacting with an
extracting organic phase, or introducing adsorbents that can
selectively enrich HMF into the reaction system.
4. Fructose dehydration
The highest reported yields of HMF from hexoses use fructose
as the starting reactant. The fructose dehydration reaction has
been carried out using homogeneous and heterogeneous cata-
lysts in monophasic systems and biphasic systems, as well as
in ionic liquid solutions. Representative literature articles will
be reviewed looking at each of the three solvent systems indivi-
dually. Tables 13 summarize the results of fructose dehy-
dration in these systems.
Because of its higher reactivity compared to glucose, fruc-
tose was first used as the model C6 sugar for the dehydration
reaction. The early work in this field primarily focused on the
monophasic system, being composed of either pure water or a
waterorganic mixed solvent. Since fructose dehydration is
widely-accepted as a Brnsted acid catalyzed reaction, eorts
in exploring dierent types of acids, e.g., mineral, organic,
resins and metal oxides, have been extensively reported as
detailed in the discussion below.
4.1 Fructose dehydration in monophasic systems
In 1977, Kuster et al. developed analytical procedures for the
quantification of fructose and its dehydration products.
91
The
procedures were used in a series of articles to find the eect of
catalyst concentration, water concentration, and pH
9294
on
dehydration rate and product yields. Using a reaction tempera-
ture of 95 C, ambient pressure, and higher acid concentration
(2 M HCl), gave the highest yields of HMF in the study, 30%,
at a conversion of 75%.
93
In the next set of experiments, poly-
ethyleneglycol-600 (PG-600) was added as a diluent. The
highest PG-600 concentration used, 70% PG-600, gave the
maximum yield of HMF at 70% and 90% conversion.
94
In
the final study, pH was controlled using the addition of formic
acid.
92
It should be noted that in this final set of experiments,
Fig. 8 Proposed mechanism of glucose isomerization to fructose (M =
various metal Lewis acid centers). Adapted from ref. 76.
Critical Review Green Chemistry
554 | Green Chem., 2014, 16, 548572 This journal is The Royal Society of Chemistry 2014
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
the reaction conditions used were 175 C and 50 atm. The
maximum yields of HMF were obtained at pH of 2.73.0 where
yield was 55% at 70% conversion after 6090 minutes.
When the pH was not controlled, it was found that during
reaction the pH decreased due to the formation of acids. For
example, if no acid catalyst was initially added, the pH was
measured to have decreased from 7.0 to 3.2.
A few years later, Nakamura and Morikawa used a strongly
acidic ion-exchange resin, Diaion PK-216, in DMSO.
95
At 80 C,
a yield of 30% was obtained after 40 min. The yield increased
to 90% for a reaction time of 500 min. Nakamura and Mori-
kawa also used an Amberlite IR-118 resin at 60 C for the con-
tinuous dehydration of fructose and were able to achieve yields
of 8095% for reaction times up to 900 hours. Fructose was
also dehydrated in pure DMSO without a catalyst by Musau
and Munavu.
96
In a 1 : 8.5 fructoseDMSO solution at 150 C,
an HMF yield of up to 92% was obtained.
Van Dam et al. examined polyethylene glycol as a cosolvent
with DMSO to increase the yield of HMF from fructose.
97
Using 0.5 M fructose and 1 M p-toluenesulfonic acid at 88 C
and a 300 minute reaction time, the HMF yield increased from
20% with no cosolvent (65% conversion) to 30% with
50 vol% PEG-200 and up to 45% with 50 vol% PEG-4000.
Interestingly, HMF yield decreased to <5% with the addition of
50 vol% ethylene glycol. The conversion of fructose with an
added cosolvent was >90% for each of the three types of ethyl-
ene glycol. Van Dam et al. also experimented with changes in
initial fructose concentration, acidity, and the addition of
metal salts, but the eects on HMF yield were less pronounced
when compared with a cosolvent addition.
Vinke and van Bekkum used activated carbon to adsorb
HMF once it was produced to limit the rehydration to LA.
98
Two dierent systems were used in the work; first, a one-pot
reactor with the catalyst and adsorbent, and second, a
batch reactor with continuous adsorption using separate
catalyst and adsorbent beds. In the one-pot reactor, 9 g of
fructose, 20 g of activated carbon, and 40 g of acidic
ion-exchange catalyst (or an equivalent amount of HCl in
50 mL of water for homogeneously catalyzed reactions) were
mixed in 150 mL of water. For continuous adsorption, a fruc-
tose solution was circulated between the catalyst column and
then the adsorbent column. In both cases, reactions were
carried out at 90 C, and in the continuous system, the adsorp-
tion column was operated at 25 C. In both systems, the
maximum yield achieved was 50% with an HMF selectivity of
6065%.
Carlini et al. used a solid acid niobium catalyst (niobic acid
or niobium phosphate) in an aqueous environment at
100 C.
99,100
HMF selectivity was high (>90%) at low conver-
sions (<30%). When the reaction time was increased, fructose
conversion increased while the HMF selectivity dramatically
decreased. In a separate study, a range of dierent metal-
vanadyl phosphate catalysts were used at a slightly lower temp-
erature, 80 C.
101
Table 1 Fructose dehydration in monophasic system
Fructose
loading Solvent Catalyst Catalyst loading T/C Time
Conv
(%)
Selec
(%)
Yield
(%) Ref.
0.25 M Water HCl 2 M 95 24 min 75 41 31 91
0.5 M Water HCl 1 M 95 30 min 92 76 70 91
0.5 M DMSO Diaion PK-216 10 meq. 80 40 min 30 95
1.7 M DMSO 150 120 min 92 96
0.5 M Water pTSA 1 M 88 300 min 65 20 97
0.5 M 1 : 1 (v/v) waterPEG-200 pTSA 1 M 88 300 min >90 30 97
0.5 M 1 : 1 (v/v) waterPEG-4000 pTSA 1 M 88 300 min >90 45 97
0.05 M Water H
2
SO
4
2 mM 250 35 s 93 50 62
6 wt% Water H
3
PO
4
-treated niobic acid 3.5 wt% 100 30 min 31.2 93.3 29.1 99
6 wt% Water Niobium phosphate 4.3 wt% 100 30 min 25.1 91.8 23.0 99
6 wt% Water VOP 3.6 wt% 80 30 min 46.6 86.3 40.2 100
6 wt% Water FeVOP 3.5 wt% 80 30 min 41.5 87.1 36.2 100
8.3 wt% Water a-TiO
2
8.3 wt% 200 5 min 100 20 20 103
8.3 wt% Water m/c-ZrO
2
8.3 wt% 200 5 min 85 15 13 103
6 wt% Water a-TiP 3.3 wt% 100 30 min 29.1 98.3 28.6 104
6 wt% Water g-TiP 3.3 wt% 100 30 min 36.7 96.1 35.3 104
6 wt% Water C-ZrP
2
O
7
3.3 wt% 100 30 min 44.4 99.8 44.3 104
2 wt% 3 : 7 (w/w) wateracetone Dowex 50wx8 2 wt% 150 15 min 95.1 77.2 73.4 106
0.5 M i-Propanol HCl 5 mol% 120 120 min 83 109
20 wt% 1 : 9 waterTEAB Amberlyst-15 10 wt% 100 15 min 94 98 92 110
0.5 M DMSO [NMP]CH
3
SO
3
7.5 mol% 90 120 min 83 87.2 72.3 112
0.56 M DMSO SILnP 6.3 wt% 130 30 min 99.9 63 63 113
8.3 wt% Sulfolane LiCl 8.3 wt% 90 120 min 67 114
0.56 M NMP FeCl
3
Et
4
NBr 0.056 M FeCl
3
,
0.1 M Et
4
NBr
90 120 min 100 86 86 115
0.28 M DMSO TiO
2
4.2 wt% 140 5 min 53.2 105
10 wt% DMA H
2
SO
4
NaBr 6 mol% H
2
SO
4
,
10 wt% NaBr
100 120 min 93 111
0.3 M 9 : 1 (v/v) 1,4-dioxaneDMSO Amberlyst-15 110 3 min 98 94 92 116
2 wt% 1 : 9 waterDHMTHF Amberlyst-70 0.07 wt% 130 25 min 85 70 60 118
3 wt% DMSO Nafion(15)/MCF 0.1 mM H
+
90 120 min 94 95 89.3 119
Green Chemistry Critical Review
This journal is The Royal Society of Chemistry 2014 Green Chem., 2014, 16, 548572 | 555
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
In most instances, HMF selectivity was 8095% and fruc-
tose conversion was 4565%. In contrast, though, the
decrease in HMF selectivity at longer reaction times was not as
dramatic as with the niobium catalysts. Carniti et al. also used
niobic acid and niobium phosphate as catalysts, but in a flow
reactor.
102
With a temperature between 90110 C, there was
an increasing trend between selectivity and conversion, with
the maximum reported being a selectivity of 35% at 75% con-
version. However, catalyst deactivation was present in both
systems. After 24 hours on stream, fructose conversion was
60% for niobium phosphate and 40% for niobic acid. The
conversion decreased to <20% for both catalysts after
200 hours on stream.
Watanabe et al. examined TiO
2
and ZrO
2
metal oxides in
hot compressed water for glucose and fructose dehydration.
103
From fructose, HMF yield was 20% with TiO
2
and 15% with
ZrO
2
, with conversions being 90100% for each case. The reac-
tion conditions were as follows: 200 C temperature, 25 bar Ar,
and a 5 minute reaction time. Benvenuti et al. used dierent
heterogeneous Ti and Zr phosphate catalysts.
104
At 100 C, an
aqueous solution of 6 wt% fructose was converted with a cubic
zirconium pyrophosphate (C-ZrP
2
O
7
) catalyst giving an HMF
selectivity of 99.8% (44.4% conversion) after 30 minutes. For
longer reaction times the selectivity dropped to 86% after
1 hour (52.2% conversion) and to 82.1% after 2 hours (52.8%
conversion). The Ti phosphate catalysts gave selectivities of
9698% at conversions of 2535% after 30 minutes. The
authors concluded the both Lewis acid and Brnsted acid sites
were active in the reaction, and as the Lewis acid site strength
was increased the HMF yield in improved. Dutta et al. used
microwave heating and a TiO
2
catalyst to obtain up to a 54%
yield in DMSO at 140 C with 300 W of microwave power for
5 minutes.
105
Lower yields were aorded in water, waterMIBK,
and acetonitrile systems, and with decreased microwave power.
Qi et al. used an ion-exchange resin in an acetonewater
mixture with microwave heating.
106
The microwave irradiation
heated the 5 g reaction mixture to 150 C within 30 s. After ten
minutes at reaction temperature in an 85 : 15 (w/w) acetone
water mixture, fructose conversion was 94% with 73.4% HMF
selectivity. Decreasing the amount of acetone required increas-
ing reaction times to reach comparable fructose conversions.
Bicker et al. also used acetonewater mixtures but at sub- and
supercritical conditions (180300 C, 50300 bar).
107
Using
sulfuric acid (10 mM) as a catalyst at 180 C and 200 bar, the
highest yield for HMF was obtained in 90 : 10 acetonewater
where the selectivity was 75% at 90% conversion.
Le et al. experimented with acetic acid and formic acid as
catalysts in high temperature liquid water.
108
The temperature
was varied between 180 C and 220 C and without a catalyst
HMF yield was 50% at nearly full conversion. When either
acetic acid or formic acid were added at 10.8 mg mL
1
, the
yield increased to 60% at 100% conversion. Their work
suggested acetic acid may promote the dehydration of fructose
to HMF, while having little influence on the rehydration of
HMF to levulinic acid, whereas formic acid appeared to cata-
lyze both steps. T
a
b
l
e
2
F
r
u
c
t
o
s
e
d
e
h
y
d
r
a
t
i
o
n
i
n
b
i
p
h
a
s
i
c
s
y
s
t
e
m
F
r
u
c
t
o
s
e
l
o
a
d
i
n
g
A
q
u
e
o
u
s
p
h
a
s
e
O
r
g
a
n
i
c
p
h
a
s
e
A
q
:
O
r
g
p
h
a
s
e
r
a
t
i
o
C
a
t
a
l
y
s
t
C
a
t
a
l
y
s
t
l
o
a
d
i
n
g
T
/

C
T
i
m
e
C
o
n
v
(
%
)
S
e
l
e
c
(
%
)
Y
i
e
l
d
(
%
)
R
e
f
.
1
.
2
M
W
a
t
e
r
M
I
B
K
1
:
9
(
v
/
v
)
L
e
w
a
t
i
t
S
C
1
0
2
7
.
5
m
e
q
u
i
v
.
H
+
8
8
4
h
7
2
4
7
3
4
1
2
1
0
.
1
6

1
.
1
M
W
a
t
e
r
M
I
B
K
1
:
5
(
v
/
v
)
H
-
Z
S
M
5
1
.
5

2
.
5
w
t
%
1
6
5
6
0
m
i
n
9
0
5
9
5
3
1
2
4
0
.
1
6

1
.
1
M
W
a
t
e
r
M
I
B
K
1
:
5
(
v
/
v
)
H
Y
-
f
a
u
j
a
s
i
t
e
,
S
i
/
A
l
=
1
0
1
.
5

2
.
5
w
t
%
1
6
5
6
0
m
i
n
7
4
5
6
4
1
1
2
4
0
.
1
6

1
.
1
M
W
a
t
e
r
M
I
B
K
1
:
5
(
v
/
v
)
H
-
m
o
r
d
e
n
i
t
e
,
S
i
/
A
l
=
1
1
1
.
5

2
.
5
w
t
%
1
6
5
6
0
m
i
n
7
6
9
1
6
9
1
2
4
3
0
w
t
%
7
:
3
(
w
/
w
)
(
8
:
2
w
a
t
e
r

D
M
S
O
)
/
P
V
P
7
:
3
(
w
/
w
)
M
I
B
K

2
-
b
u
t
a
n
o
l
1
:
2
(
w
/
w
)
H
C
l
0
.
2
5
M
1
8
0
2
.
5

3
m
i
n
8
2
8
3
6
8
1
2
5
5
0
w
t
%
7
:
3
(
w
/
w
)
(
8
:
2
w
a
t
e
r

D
M
S
O
)
/
P
V
P
7
:
3
(
w
/
w
)
M
I
B
K

2
-
b
u
t
a
n
o
l
1
:
2
(
w
/
w
)
H
C
l
0
.
2
5
M
1
8
0
2
.
5

3
m
i
n
9
2
7
7
7
1
1
2
5
1
0
w
t
%
1
:
1
(
w
/
w
)
w
a
t
e
r

D
M
S
O
7
:
3
(
w
/
w
)
M
I
B
K

2
-
b
u
t
a
n
o
l
1
:
2
(
w
/
w
)
H
C
l
1
.
0
p
H
1
7
0
4
m
i
n
9
5
8
9
8
5
1
2
7
1
0
w
t
%
4
:
6
(
w
/
w
)
w
a
t
e
r

N
M
P
M
I
B
K
1
:
1
(
w
/
w
)
R
e
s
i
n
1
0
0
w
t
%
9
0
1
8
h
9
8
8
5
8
3
1
2
8
2
0
w
t
%
[
B
M
I
M
]
C
l
T
H
F
1
:
2
0
(
w
/
w
)
W
C
l
4
1
0
m
o
l
%
5
0
4
h

6
0
1
2
9
3
0
w
t
%
W
a
t
e
r
7
:
3
(
w
/
w
)
M
I
B
K

2
-
b
u
t
a
n
o
l
1
:
2
(
w
/
w
)
T
a
a
-
S
B
A
-
1
5
3
w
t
%
1
8
0
3
0
m
i
n
6
6
7
4
5
8
1
3
0
3
0
w
t
%
W
a
t
e
r
7
:
3
(
w
/
w
)
M
I
B
K

2
-
b
u
t
a
n
o
l
1
:
2
(
w
/
w
)
T
E
S
A
S
-
S
B
A
-
1
5
9
w
t
%
1
3
0
1
4
1
m
i
n
8
4
7
1
6
0
1
3
1
6
w
t
%
W
a
t
e
r
2
-
B
u
t
a
n
o
l
2
:
3
(
v
/
v
)
N
b
2
O
5

x
H
2
O
0
.
5
w
t
%
1
6
0
5
0
m
i
n
9
0
9
9
8
9
1
3
2
3
0
w
t
%
W
a
t
e
r
M
I
B
K
1
:
4
(
v
/
v
)
B
(
O
H
)
3
1
0
w
t
%
1
5
0
4
5
m
i
n
5
2
5
4
2
8
1
3
3
3
0
w
t
%
W
a
t
e
r
M
I
B
K
1
:
4
(
v
/
v
)
B
(
O
H
)
3
+
N
a
C
l
1
0
w
t
%
+
5
w
t
%
1
5
0
4
5
m
i
n
7
0
6
4
4
5
1
3
3
0
.
3
7
M
W
a
t
e
r
M
I
B
K
1
:
3
(
v
/
v
)
H
-
m
o
r
d
e
n
i
t
e
1
.
3
w
t
%
1
6
5
1
5
0
m
i
n
5
0
5
0
2
5
1
3
4
3
0
w
t
%
W
a
t
e
r
M
I
B
K
1
:
3
(
v
/
v
)
A
g
3
P
W
1
2
O
4
0
1
w
t
%
1
2
0
6
0
m
i
n
8
3
9
4
7
8
1
3
5
3
0
w
t
%
W
a
t
e
r
M
I
B
K
1
:
3
(
v
/
v
)
C
s
2
.
5
H
0
.
5
P
W
1
2
O
4
0
6
.
4
w
t
%
1
1
5
6
0
m
i
n
7
8
9
5
7
4
1
3
6
2
0
w
t
%
W
a
t
e
r
A
c
e
t
o
n
i
t
r
i
l
e
1
:
8
(
w
/
w
)
[
M
B
C
I
m
]
S
O
3
C
l
1
m
o
l
%
8
0
4
h

8
8
.
7
1
3
7
0
.
5
6
M
W
a
t
e
r
D
C
M
1
:
1
(
v
/
v
)
H
C
l
3
2
%
1
0
0
1
.
6
7
m
i
n

8
1
1
3
8
2
3
w
t
%
W
a
t
e
r
n
-
B
u
t
a
n
o
l
3
:
1
(
v
/
v
)
T
i
O
2

2
0
0
3
m
i
n

1
8
1
3
9
Critical Review Green Chemistry
556 | Green Chem., 2014, 16, 548572 This journal is The Royal Society of Chemistry 2014
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
Lai and Zhang used alcohols as a solvent for the dehy-
dration reaction system.
109
A significant benefit in using
alcohol is the further reaction with HMF to form HMF ethers,
a more stable product that does not undergo polymerization
or rehydration. C
1
to C
4
alcohols were used with i-propanol
giving the highest yield of HMF, 87%. Reactions were carried
at 120 C for 2 h with 5 mol% HCl as the catalyst. Recycling
experiments were also performed by reusing evaporated and
recondensed solvent. HMF yields were similar across 5 separ-
ate runs, provided the 2 mol% of HCl was added before each
run.
Simeonov et al. used readily crystallizable salts as a solvent
to ease separation of HMF after production.
110
Tetraethyl-
ammonium bromide (TEAB) with an Amberlyst-15 catalyst pro-
duced yields of >92% with nearly 100% HMF selectivity.
Reactions were carried out at 100110 C for 1590 min.
Binder and Raines used various catalysts and additives in a
10 wt% fructose in DMA system.
111
The highest yields (>89%)
came with 6 mol% H
2
SO
4
and 10 wt% of alkali metal bromide
or iodide salts. The reaction temperature was 100 C and the
reaction time was 26 hours.
Tong and Li used a small amount of ionic liquid (7.5 mol%)
in DMSO as a catalyst.
112
Using the ionic liquid N-methyl-2-
pyrrolidonium methyl sulfonate [NMP]CH
3
SO
4
at 90 C a
72.3% yield of HMF at 87.2% selectivity was achieved after
2 hours of reaction. Sidhpuria et al. immobilized an ionic
liquid, 1-(tri-ethoxysilyl-propyl)-3-methyl-imidazolium hydro-
gen sulfate, onto silica nanoparticles of 300 or 600 nm size.
113
The highest HMF yield attained was 63% at full conversion.
The reaction was performed in pure DMSO at 130 C for
30 minutes.
Caes and Raines used sulfolane as a solvent with various
metal chloride salts as catalysts.
114
LiCl provided the highest
yield at 67% with the separate selectivity or conversion not
reported. Reactions were performed at 90 C for 2 hours with
equal weights of fructose and catalyst. Tong et al. used an iron
salt catalyst, FeCl
3
, with a co-catalyst.
115
At 90 C after 2 hours
in N-methylpyrrolidone, the highest yields achieved were
8286% at full conversion. The most active co-catalysts were
tetraethylammonium bromide (Et
4
NBr), NH
4
Br, and Et
4
NCl.
Using a flow reactor, Antal et al. added 2 mM H
2
SO
4
as a
catalyst to 0.05 M fructose in water and attained about 50%
HMF yield at 93% conversion under optimal conditions.
62
Reactions were carried out at 250 C and 34.5 MPa. Another
system for the continuous production of HMF from fructose
used an Amberlyst-15 resin catalyst with DMSO as a solvent.
116
Under optimal condition, a 92% HMF yield at 98% conversion
was attained. At 110 C, the catalyst system was relatively stable
up to 96 hours. In the system used, the yield was maximized
by decreasing the Amberlyst particle size to 0.2 mm, increasing
the flow rate to minimize external mass transfer limitations,
and increasing temperature to increase reaction rate.
Table 3 Fructose dehydration in ionic liquids
Fructose loading Solvent Catalyst
Catalyst
loading T (C) Time
Conv
(%)
Selec
(%)
Yield
(%) Ref.
0.5 M 5 : 3 (v/v) [BMIM]BF
4
DMSO 80 15 h 27 140
0.5 M 5 : 3 (v/v) [BMIM]BF
4
DMSO 80 32 h 36 140
0.5 M 5 : 3 (v/v) [BMIM]BF
4
DMSO Amberlyst-15 88 g L
1
80 32 h 75 140
0.5 M 5 : 3 (v/v) [BMIM]BF
4
DMSO Amberlyst-15 175 g L
1
80 32 h 87 140
0.5 M 5 : 3 (v/v) [BMIM]PF
6
DMSO Amberlyst-15 88 g L
1
80 24 h 70 140
0.5 M 5 : 3 (v/v) [BMIM]PF
6
DMSO Amberlyst-15 175 g L
1
80 24 h 80 140
10 wt% [EMIM]Cl 80 3 h 5 20 1 78
10 wt% [EMIM]Cl 100 3 h 58 69 40 78
10 wt% [EMIM]Cl 120 3 h 100 72 72 78
10 wt% [BMIM]Cl NHC-Cr 9 mol% 100 6 h 96 141
6.2 wt% ChClcitric acid 80 1 h 93.2 83.5 77.8 146
5 wt% [BMIM]Cl Amberlyst-15 5 wt% 120 1 min 99.3 82.8 82.2 147
5 wt% [BMIM]Cl + 6 wt% DMSO Amberlyst-15 5 wt% 25 6 h 90 78 70 144
5 wt% [BMIM]Cl GeCl
4
10 mol% 100 5 min 100 92.1 92.1 142
10 wt% [EMIM]Cl PTA 4 wt% 80 1 h 87 92 80 149
10 wt% [EMIM]Cl PTA/MIL-101 4 wt% 80 1 h 84 75 63 149
0.67 M [EMIM]HSO
4
MIBK 100 30 min 100 88 88 145
20 wt% IL + 35 wt% methanol 100 5 min 70 44
20 wt% IL + 45 wt% ethanol 100 40 min 64 44
80 wt% ChCl 120 90 min 70 150
250 wt% ChCl 120 90 min 50 150
40 wt% ChCl pTSA 10 mol% 100 30 min 67 151
40 wt% ChCl CrCl
3
10 mol% 100 30 min 60 151
40 wt% ChCl FeCl
3
10 mol% 100 30 min 59 151
10 wt% [BMIM]Cl Sulfonated lignin 2.5 wt% 100 10 min 98 96 94.3 143
10 wt% 1 : 1 (w/w) BHCglycerol 110 2 h 57 152
20 wt% 4 : 1 (w/w) BHCwater + MIBK 110 2 h 57 152
1 wt% [BMIM]Cl Sulfonated
carbonaceous solid
5 wt% 80 10 min 83 144
Green Chemistry Critical Review
This journal is The Royal Society of Chemistry 2014 Green Chem., 2014, 16, 548572 | 557
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
Another reaction system used a sulfonic acid-functionalized
mesoporous silica catalyst with a single phase mixture of 4 : 1
(w/w) THFwater.
117
While the silica has better ex situ thermal
stability than an Amberlyst-15 resin (170 C reaction tempera-
ture vs. 120 C max operating temperature for the resin), it was
less stable under reaction conditions. During reaction, the
catalyst was gradually deactivated as fructose conversion
decreased from 90% at 10 hours on stream to 40% after
60 hours. The selectivity to HMF ranged from 6070%. Tucker
et al. also experimented with a carbohydrate derived com-
pound, DHMTHF, as a cosolvent with water to give up to
70% HMF selectivity at 85% conversion.
118
The reaction
was carried out at 130 C, in 1 : 9 waterDHMTHF with Amber-
lyst-70 as the catalyst. Decreasing the amount of DHMTHF led
to decreases in the initial reaction rate and HMF yield.
Huang et al. impregnated Nafion onto mesocellular silica
foams (MCF) to use as an acid catalyst.
119
In DMSO at 90 C
for 2 hours, up to an 89.3% yield was achieved at 95.0% selecti-
vity. The yield was decreased if a dierent organic solvent was
used.
In summary, it was observed that if water was used as the
only solvent, HMF yields were typically low regardless of the
type of acid catalyst used. Unfortunately, the solubility of C6
sugars is generally limited in most of organic solvents except
for high boiling point non-protonic ones such as DMSO. Thus,
an organic solvent was typically mixed with water in order to
solubilize reactants and to a certain degree minimize HMF
degradation, resulting in higher HMF yields. Very high HMF
yields were obtained from high boiling point organic solvents
such as DMSO, which would ultimately cost a significant
amount of energy to separate HMF via distillation.
Despite numerous studies on the manipulation of solvent
composition in order to improve HMF yields, very few investi-
gations were devoted to the creation of favorable catalyst
systems. In this regard, a recent report by Dumesic et al.,
demonstrated that by creating a local microenvironment where
an acid catalyst and a PVP polymer were co-localized, the resul-
tant fructose tautomer favored the production of HMF.
120
This
approach suggests a promising direction for HMF production
in monophasic systems with emphasis on the design and
synthesis of a favorable local catalytic environment.
4.2 Fructose dehydration in biphasic systems
Another approach that has been explored involved the reaction
engineering strategy of contacting the reactive aqueous phase
containing both the sugars and the catalyst with an extracting
organic phase. With this concept, the produced HMF was
removed from the reactive aqueous phase, thereby minimizing
side reactions and over reactions.
Rigal et al. used a system consisting of water and methyliso-
butylketone (MIBK) in a 1 : 9 (v/v) waterorganic ratio with an
ion-exchange resin as an acid catalyst.
121
A range of dierent
gel-type and macroreticular strong acid resins were used with
varied H
+
capacity, surface area, porosity, pore volume, and
operating temperature. For reaction at 88 C and 4 h, HMF
yield was 1035% at selectivities of 3060%. After 15 h of
reaction, the highest yield of HMF was obtained, 56% at 66%
selectivity. In another study, Rigal et al. investigated the eect
of reaction time, temperature, waterMIBK ratio, catalyst con-
centration, fructose concentration, stir rate, and resin particle
size distribution.
122,123
In short, longer reaction times, higher
temperatures, or higher fructose concentrations had a ben-
eficial eect on HMF yield, but also led to higher LA yields.
Increasing the amount of MIBK used relative to the amount of
water had a positive eect on HMF selectivity, but a lower yield
was achieved. Increasing the amount of catalyst also increased
HMF yield, but had less of a negative eect on HMF selectivity.
Increasing the stir rate had a positive eect on HMF yield and
a limited eect on LA formation. Using smaller particles had a
positive eect on LA yield.
Rivalier et al. evaluated a range of zeolites for their use in a
continuous solidliquidliquid extraction reaction.
124
Zeolite
and fructose were added to a 1 : 5 (v/v) mixture of waterMIBK.
The reactors were pressurized with N
2
to 10 bar and heated to
165 C. After a 60 minute reaction time with no catalyst, HMF
selectivity was 38% at 32% conversion. A zeolite catalyst,
H-ZSM5, gave 59% HMF selectivity at 90% conversion. HY-fau-
jasite and H-mordenite were each tested at four dierent Si/Al
ratios. The highest yield from HY-faujasite was with a Si/Al =
10 or 15, where HMF yield was 40% at 75% conversion. The
highest yield from H-mordenite was with a Si/Al = 11 where
selectivity was 91% at 76% conversion. In a subsequent article,
H-mordenite was tested at six dierent Si/Al ratios.
65
However,
the new H-mordenite catalysts tested all gave lower yields of
HMF when compared to the H-mordenite with a Si/Al = 11.
Romn-Leshkov et al. used phase modifiers to promote the
partitioning of HMF to the organic phase after formation.
125
Specifically, polar aprotic solvents (DMSO or 1-methyl-2-pyrro-
lidinone (NMP)) and a hydrophilic polymer ( poly(1-vinyl-2-pyr-
rolidinone (PVP)) were added to the aqueous phase, while
2-butanol was added to MIBK in the organic phase. Using
three modifiers, DMSO, PVP, and 2-butanol, with an initial
fructose loading of 30 wt% in the aqueous phase, a reaction
temperature of 180 C, and a reaction time of 2.53 minutes,
an 83% HMF selectivity was achieved at 82% conversion. Fig. 9
shows the biphasic reaction design. The Dumesic group satu-
rated the aqueous phase with NaCl to help promote the parti-
tioning of HMF to the organic phase and thereby improved
selectivity and yield.
126
The authors also tested dierent
organic solvents for the extracting phase; primary and second-
ary alcohols, ketones, and cyclic ethers. Generally, C
4
solvents
gave the highest yields of HMF as compared with the C
3
, C
5
,
and C
6
compounds.
Chheda et al. was able to obtain high selectivities from fruc-
tose, 8790% at nearly full conversion
127
using an aqueous
phase of 1 : 1 (w/w) waterDMSO and an organic phase of 7 : 3
(w/w) MIBK2-butanol. HCl was used as the catalyst and reac-
tions were performed at 170 C for 48 minutes with a 10 wt%
initial fructose loading. In a separate publication, up to 85%
selectivity at 98% conversion was attained from fructose.
128
For this case, the aqueous phase was 4 : 6 (w/w) waterNMP,
MIBK the organic phase, 90 C reaction temperature, resin
Critical Review Green Chemistry
558 | Green Chem., 2014, 16, 548572 This journal is The Royal Society of Chemistry 2014
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
catalyst, and an 18 hour reaction time. DMSOwater mixtures
were also used and gave similar yields of HMF.
Chan et al. evaluated 12 dierent metal chloride catalysts
using a biphasic system.
129
In the system, the ionic liquid,
[BMIM]Cl, was used as the polar phase, and THF was used as
the organic phase. At low temperature (50 C), WCl
4
and WCl
6
were the most promising catalysts giving HMF yields of 60%
after 4 hours although the separate fructose conversion or
HMF selectivity were not given.
Bifunctional solid catalysts, consisting of thioether groups
and sulfonic acid groups integrated into ordered mesoporous
silica, achieved a 74% HMF selectivity at 66% fructose conver-
sion.
130
Reaction occurred at 180 C for 30 min with a 30 wt%
fructose loading and a 7 : 3 (w/w) MIBK2-butanol organic
phase. By comparison, when the sulfur groups were grafted
onto non-porous silica, HMF selectivity was 61% and conver-
sion 62%, while reaction time was four times longer, 120 min.
Further research with acid-functionalized mesoporous silica
showed similar yields of HMF.
131
At 130 C, a 30 wt% fructose
loading in water with a 7 : 3 MIBK2-butanol organic layer after
141 min gave a HMF yield of 60% at 84% conversion.
A solid acid, Nb
2
O
5
, was used by Yang et al. to attain up to
an 89% yield of HMF at 92% conversion.
132
The highest yield
was produced in a biphasic reactor with 20 mL of water and
30 mL of 2-butanol at 160 C for 50 min. Yields decreased
when the temperature was lowered or if the catalyst or fructose
concentrations were decreased. Another Lewis acid, boric acid,
B(OH)
3
, was used to get a yield of 28% at 52% conver-
sion.
133
With added NaCl, the yield increased to 45% at a
conversion of 70%. The temperature was 150 C for
45 minutes in 1 : 4 waterMIBK starting with 30 wt% fructose,
100 g L
1
B(OH)
3
, and 50 g L
1
NaCl.
Ordomsky et al. used zeolites in a waterMIBK system for
fructose dehydration.
134
The addition of MIBK led to an
increase in HMF selectivity when compared with a purely
aqueous system, 8090% to 3040%, respectively. Selectivity
was improved upon silylation and deactivation of the external
surface acid sites.
Fan et al. used a heteropolyacid, Ag
3
PW
12
O
40
, to obtain
yields of 7078% at 75100% conversion, with the highest
yield being 77.7% at 82.8% conversion.
135
Reactions were
carried out in a waterMIBK biphasic system at 120 C for
60120 min. The highest yields occurred at shorter reaction
times and higher amounts of MIBK. Ag
3
PW
12
O
40
also gave the
highest yield among the dierent catalysts tested in the study,
which included AgNO
3
, H
3
PW
12
O
40
, Cs
3
PW
12
O
40
, and HCl.
The eect of initial feedstock concentration and temperature
were also probed. Similar results as other literature studies
were obtained. Namely, lower fructose loadings and higher
temperatures led to higher HMF yields. Zhao et al. used
another heteropolyacid, Cs
2.5
H
0.5
PW
12
O
40
, and obtained up to
a 74% yield at 78% conversion.
136
The reactions were per-
formed at 115 C for 60 minutes in 1 : 3 waterMIBK with
30 wt% fructose initial loading. Zhao et al. also studied the
eect of reaction time, ratio of aqueous and organic solvents,
and catalyst loading. Generally, higher yields came at longer
reaction times, greater amounts of MIBK, and higher catalyst
loadings. Okano et al. used a biphasic system of water and
acetonitrile and the introduction of an ionic liquid catalyst,
1-methyl-3-(butyl-4-chlorosulfonyl)imidazolium chlorosulfate,
[MBCIm]SO
3
Cl.
137
The ionic liquid and HMF partitioned to
the organic, however, the ionic liquid was removed to a
pseudo-ternary phase by treatment with alumina leaving HMF
in the organic phase. Up to 88.7% yield of HMF was achieved
at 80 C after 4 h in this system.
Brasholz et al. used a continuous HMF production system
and obtained up to an 81% yield.
138
The aqueous phase con-
tained 0.56 M fructose and 32% HCl. The organic phase was
dichloromethane. Reactions were carried out at 100 C with a
residence time of 1.67 minutes. In a dierent flow reactor
setup, McNe et al. was able to get 18% yield of HMF.
139
The
reaction was performed at 200 C with a 3 minute residence
time, TiO
2
catalyst, 3 : 1 aqueousn-butanol ratio, and a 23 wt%
fructose loading.
Several insights can be gleaned from these studies. First,
while various acid catalysts (e.g., mineral acids, zeolites, resins,
and metal oxides) can function as eective dehydration
catalysts, the ultimate HMF yields not only depend on how
rapidly the catalysts can dehydrate fructose, but on also on
how eciently HMF can be extracted into the protective
organic phase. With this in mind, high salt levels were added
to the aqueous phase to favor HMF partitioning into organic
phase. Additionally, mass transfer of HMF through the con-
tacting area of the organicaqueous interface should be con-
sidered in order to extract produced HMF before it was
degraded.
Given these considerations, further development of an
appropriate catalyst system need to focus on the design of cata-
lysts that are more stable in the harsh reactive aqueous phase
conditions where high concentration of salts can be present in
the acidic hot liquid water. Relative to mass transfer concerns
at the organicaqueous interface, more aggressive mechanical
methods could be considered to improve mass transfer
eciently. Alternatively, process conditions leading to a water-
Fig. 9 Representation of biphasic reactor design for HMF pro-
duction.
125
From Y. Romn-Leshkov, J. N. Chheda and J. A. Dumesic,
Science, 2006, 312, 19331937. Reprinted with permission from AAAS.
Green Chemistry Critical Review
This journal is The Royal Society of Chemistry 2014 Green Chem., 2014, 16, 548572 | 559
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
in-oil or oil-in-water emulsion configuration in the biphasic
system could be utilized to increase the contact area.
4.3 Fructose dehydration in ionic liquid
Due to its superior ability to solubilize sugars as well as poten-
tially reinforce sugar molecule configurations that favor de-
hydration to HMF, ionic liquids (ILs) have been attracting
increased attention in recent years.
Lansalot-Matras and Moreau explored a hydrophilic
[BMIM]BF
4
or hydrophobic [BMIM]PF
6
ionic liquid with
DMSO as the cosolvent and Amberlyst resin as the acid cata-
lyst.
140
Under reaction at 80 C, with fructose in pure DMSO
only trace of amounts of HMF was formed. With the ionic
liquid present at 5 : 3 [BMIM]BF
4
DMSO, the yield of HMF was
27% after 15 hours and 36% after 32 hours. Upon addition of
Amberlyst-15, the yield of HMF increased to 75% after
32 hours, and when the catalyst amount was doubled, the
yield increased further to 87%. For the hydrophobic ionic
liquid, [BMIM]BF
4
, HMF yield was 70% after 24 hours and
when the Amberlyst amount was doubled the subsequent
HMF yield obtained was 80%.
Zhao et al. used [BMIM]Cl to achieve 72% yield at 100%
conversion at 120 C and 3 hours.
78
At 100 C, the yield was
40% at 58% conversion, and at 80 C, the yield was 1% at
5% conversion. Yong et al. were able to achieve up to a 96%
yield of HMF from fructose using N-heterocyclic carbene
(NHC)-chromium (Cr
II
and Cr
III
) complexes in [BMIM]Cl.
141
Reaction conditions were: 100 C, 6 hours, 10 wt% fructose
loading, and 9 mol% catalyst loading. High yields of HMF
were achieved with a GeCl
4
catalyst in [BMIM]Cl.
142
At 100 C
after 5 min, HMF yield was 92.1% at full conversion. GeCl
4
gave the highest yields as compared to 6 other metal chlorides
that were tested. As a result, GeCl
4
was subsequently examined
in the dehydration of other carbohydrates.
Xie et al. used sulfonated lignin as an acid catalyst in
[BMIM]Cl.
143
At 100 C, the maximum HMF yield was achieved
after 10 minutes with a value of 94% at 98% conversion.
The ionic liquid (83% yield at 100% conversion) aorded
higher yields of HMF when compared to the polar aprotic sol-
vents DMSO (80% yield at 100% conversion), NMP (55%
yield at 95% conversion), and dimethylacetamide (DMA,
35% yield at 75% conversion) at 100 C for 40 minutes.
Qi et al. sulfonated a cellulose-derived carbonaceous solid and
attained an 83% yield when the reaction was performed at
80 C for 10 minutes in [BMIM]Cl with a 1 wt% fructose
loading and 5 wt% catalyst loading.
144
Lima et al. obtained up
to 88% yield at full conversion in [BMIM]HSO
4
.
145
The reaction
was performed at 100 C for 30 minutes with MIBK as an
extracting solvent.
Hu et al. experimented with a wide range of ionic
liquids.
146
The highest yields were obtained in choline chlo-
ride with citric acid as a catalyst, particularly when ethyl
acetate was used to extract HMF. At 80 C and 1 hour, the yield
in choline chloride was 75% at 91% conversion. With inter-
mittent extraction, the yield increased to 86% at 94%
conversion, and with continuous extraction, the yield
increased further to 92% at 97% conversion.
Using an Amberlyst-15 resin, Qi et al. achieved an 83.3%
HMF yield and 98.6% fructose conversion in [BMIM]Cl at
80 C after 10 min.
147
When the temperature was increased to
120 C, the reaction time decreased to 1 min with a similar
yield and conversion. At 80 C, the catalyst and ionic liquid
solvent were able to be recycled for seven successive runs
without any appreciable loss in activity. In a dierent study,
Qi et al. added a co-solvent (acetone, DMSO, methanol,
ethanol, or ethyl acetate) to the [BMIM]Cl and catalytic resin
system.
148
At 25 C, after 6 hours the HMF yield was 78% at
90% conversion.
Zhang et al. encapsulated a heteropolyacid, phosphotungs-
tic acid (PTA), in a chromium-based metalorganic framework
(MIL-101).
149
Under reaction at 80 C for 60 minutes with
1-ethyl-3-methylimidazolium chloride as a solvent, up to a
63% yield was attained at 84% conversion. When PTA was
used without the metalorganic framework, the yield was 80%
at 87% conversion.
Kraus and Guney prepared 5-methoxy or 5-ethoxymethyl-
furfural from fructose in sulfonic acid-functionalized ionic
liquids.
44
The highest yield obtained with methanol as the
reagent was 70% (46% for HMF and 24% for MMF), and with
ethanol was 64% (21% for HMF and 43% for EMF). The reac-
tion was performed in a biphasic reactor with the ionic liquid
and alcohol as the reaction phase and hexane as the extracting
phase. The temperature used was 100 C with the highest
product yields being obtained at shorter reaction times.
Highly concentrated solutions of fructose in choline chlo-
ride were able to give a HMF yield of 70% using initial fructose
to choline chloride weight ratios of up to 4 : 1.
150
At a 2.5 : 1
weight ratio the optimal HMF yield was 50%. Reactions were
performed at 120 C under 40 bar CO
2
for 90 minutes. In a
2 : 3 (w/w) solution of fructosecholine chloride and 10 mol%
of p-toluenesulfonic acid, a 67% yield of HMF was obtained
after reaction at 100 C for 30 minutes.
151
Under the same
reaction conditions, a 60% yield was achieved with CrCl
3
as
the catalyst and was 59% yield with FeCl
3
.
Vigier et al. used betaine hydrochloride (BHC).
152
In a 1 : 1
(w/w) BHCglycerol system at 110 C and 10 wt% fructose
loading, a 57% yield was obtained. The same yield was
attained in 4 : 1 (w/w) BHCwater with MIBK as an extracting
solvent at 100 C after 120 minutes and starting with 20 wt%
fructose.
It has been hypothesized that the cationic moiety in ionic
liquids can interact with fructose molecules and orient them
in an optimized configuration for the dehydration reaction to
occur more readily. As such, the hydrophilic/hydrophobic pro-
perties and spatial positioning rendered by alkyl groups
attached to the cationic core could play an important role in
controlling the eectiveness of the ILs for fructose de-
hydration. In contrast, the choice of anion in the ILs seemed
to be less relevant for dehydration activity.
As discussed above, both Lewis acidic metal salts and
homogeneous or heterogeneous Brnsted acids have been
Critical Review Green Chemistry
560 | Green Chem., 2014, 16, 548572 This journal is The Royal Society of Chemistry 2014
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
employed. It was shown that higher yields were generally
obtained with stronger Lewis acids, e.g., WCl
6
and GeCl
4
.
However, metal salts such as these are also very hydrolysable
Lewis acids, meaning that they will readily generate protons in
the presence of water, which is released during dehydration.
Given this fact, more fundamental understanding is required
to decipher the relative contribution of the Lewis acidity of
metals versus Brnsted acidity generated from the hydrolysis.
5. Glucose dehydration
Compared to fructose, glucose is the preferred feedstock for
HMF production due to its greater availability and lower cost.
Therefore, a more recent shift in focus has occurred from fruc-
tose to glucose as the starting reactant. In this section, repre-
sentative glucose dehydration literature articles will be
reviewed in monophasic, biphasic and ionic liquid solvent
systems. Results for the glucose dehydration reaction from
these articles are summarized in Tables 46.
5.1. Glucose dehydration in monophasic systems
Early research on converting glucose to HMF was primarily
focused on monophasic systems with either water or an
organic as the reaction media. Glucose decomposition in pure
water is non-selective and HMF yields are typically very low.
Using a sugar-solubilizing polar organic aprotic solvent such
as DMSO can improve HMF yields significantly, possibly due
to fewer side-reactions that occur under non-aqueous con-
ditions. However, sugar solubility in most organic solvents is
low and, as such, a miscible waterorganic binary solvent has
been used to perform sugar dehydration reactions.
In 1962, Mednick tested a variety of acids, ammonium salts
and combinations of an acid and a base ( pyridine) for glucose
conversion in pure water (160190 C) or 1 : 1 waterdioxane
(200250 C).
153
In pure water, HMF was produced by various
catalysts with yields ranging from 3 mol% to 23 mol%, with
the highest yield of 23 mol% being obtained in the presence
of both H
3
PO
4
and (NH
4
)
2
HPO
4
. Increased yields of HMF were
observed when using combinations of acid and base catalysts
in the waterdioxane system at temperatures above 200 C. For
example, a 46 mol% HMF yield was achieved with tempera-
tures of 200228 C, catalyzed by H
3
PO
4
/pyridine.
Seri et al. reported lanthanide(III) cations were very active
catalysts to dehydrate glucose in water at 140 C to produce
HMF without generating much levulinic acid.
154,155
However,
low yields of HMF were obtained (<10 mol%) due to the for-
mation of significant amounts of insoluble humins.
Qi et al. used microwave heating to perform glucose de-
hydration with metal oxides, TiO
2
and ZrO
2
.
156
Microwave
heating was found to promote the formation of HMF. When
TiO
2
was used as the catalyst, a 19 mol% HMF yield was realized
in water at 200 C after 5 min of microwave heating at 90%
glucose conversion. 10 mol% was observed with ZrO
2
under the
same reaction conditions. In 2011, Dutta et al. synthesized
mesoporous TiO
2
in order to increase surface area and tested
the TiO
2
nanoparticles for carbohydrate dehydration.
105
Glucose
was readily converted by microwave heating with this catalyst, at
120 C for 5 min and a 25 mol% HMF yield was achieved.
Hara and co-workers have explored two solid acid systems
as heterogeneous catalysts for glucose dehydration in aqueous
media.
157,158
In one study, protonated titanate nanotubes with
high surface area (400 m
2
g
1
) were used in a system of
1 wt% glucose in aqueous media with an equal weight of cata-
lyst, and a reaction temperature of 120 C with reaction time of
3 h. A HMF yield of 14 mol% was achieved with titanate nano-
tubes, whereas only 2 mol% was observed with conventional
TiO
2
.
158
Another report from the same group examined niobic
acid, Nb
2
O
5
nH
2
O, as a heterogeneous water-tolerant Lewis
acid.
157
Using Raman and FT-IR spectra, they found that even
NbO
4
tetrahedra, on the surface of Nb
2
O
5
nH
2
O, formed NbO
4

H
2
O immediately upon contact with water. The NbO
4
H
2
O
adducts could still be coordinatively unsaturated and function
as eective Lewis acids. A HMF yield of 12 mol% was obtained
using this catalyst with reaction temperature of 120 C and reac-
tion time of 3 h. When Nb
2
O
5
nH
2
O was used together with
H
3
PO
4
, an improved HMF yield of 48 mol% was found.
Work in aprotic polar organic solvents primarily focused on
DMSO and DMF. Yan et al. prepared super acids SO
4
2
/ZrO
2
(CZS) and SO
4
2
/ZrO
2
Al
2
O
3
(CSZA) and tested their catalytic
Table 4 Glucose dehydration in monophasic system
Glucose loading Solvent Catalyst Catalyst loading T (C) Time Yield (%) Ref.
1.4 M Water H
3
PO
4
0.14 M 176189 20 min 12.4 153
1.4 M Water H
3
PO
4
/(NH
4
)
2
HPO
4
0.06 M/0.076 M 174180 20 min 23 153
2.8 M 1 : 1 (v/v) water/dioxane H
3
PO
4
/pyridine 0.18 M/0.6 M 220228 20 min 46 153
2 wt% Water TiO
2
50 wt% 200(MI)
a
5 min 19 156
2 wt% Water ZrO
2
50 wt% 200(MI)
a
5 min 10 156
5 wt% Water meso-TiO
2
100 wt% 120 5 min 25 115
1 wt% Water titanate nanotubes 100 wt% 120 180 min 14 158
1 wt% Water Nb
2
O
5
10 wt% 120 180 min 12 157
1 wt% Water Nb
2
O
5
(pretreated with H
3
PO
4
) 10 wt% 120 180 min 48 157
7.6 wt% DMSO CZS 20 wt% 130 240 min 19.0 159
7.6 wt% DMSO CSZA 20 wt% 130 240 min 48.0 159
0.18 M DMF Amberlyst-15/HT 100 wt% 100 180 min 42 160
4.5 wt% 9 : 1 (w/w) ethanol:water AlCl
3
40 wt% 160 15 min 57
b
161
a
MI, microwave irradiation.
b
Yields were based on combined EMF and HMF.
Green Chemistry Critical Review
This journal is The Royal Society of Chemistry 2014 Green Chem., 2014, 16, 548572 | 561
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
activities for glucose conversion in DMSO.
159
Using 130 C and
a reaction time of 4 h, glucose was completely converted with
both the CZS and CSZA catalysts, giving yields between 19 mol%
and 48 mol%, compared to 4 mol% in the absence of a cata-
lyst. Takagaki et al. performed glucose dehydration in DMF
catalyzed by a combination of solid acid and base catalysts.
160
A HMF yield of 42 mol% was obtained from glucose at a
conversion of 73% after 3 h of reaction time at 100 C.
Apart from these conventional aqueous and aprotic polar
solvents, the use of alcohols (mainly methanol and ethanol)
and carboxylic acids (mainly acetic acid) were also studied.
While lower HMF yields from glucose were typically obtained
with these solvents relative to DMSODMF, it could still be
advantageous to employ these solvents, taking into account
the easier isolation/purification of HMF from the reaction
mixture, and the fact that alcohols and carboxylic acids can react
with HMF under reaction conditions to form ethers and esters,
respectively, thereby stabilizing the dehydration product. Yang
et al. conducted glucose conversion in a 9 : 1 (w/w) ethanolwater
mixed solvent system, using AlCl
3
as the catalyst. 1 mmol
glucose, 0.4 mmol catalyst, and 4 g of mixed solvent were reacted
at 160 C for 15 min.
161
A 24 mol% HMF yield was obtained
together with its ether, 5-ethoxymethylfurfural at 33 mol%,
resulting in a 57 mol% overall furan yield. Similar protection
approaches have been employed with other solvents.
107,162,163
In summary, similar approaches were applied to glucose
dehydration as were explored with fructose. Again, separation
limitations due to the use of high boiling point solvents like
DMSO need to be considered for implementing these strat-
egies. The chemical protection approach of using alcohols or
acetic acid has been more extensively examined with glucose
than with fructose.
Earlier work focused on using Brnsted acids as catalysts,
whereas the utilization of bulk or nanostructured metal oxides
has received attention. Due to its lower activity compared to
fructose, glucose dehydration typically required harsher reac-
tion conditions (higher temperature, etc.), so the stability of
these metal oxide catalyst must be examined under these
typical glucose reaction conditions. Recent work by Sievers
and co-worker have shown that metal oxides can either
undergo structural transformations to that diminishes the
surface area or the metal atoms can leach under the reaction
conditions of acidic hot liquid water.
164,165
Unlike with fruc-
tose, little work with glucose has focused on creating a favor-
able local environment for dehydration in a heterogeneous
system, so eorts in this direction might have promise.
5.2 Glucose dehydration in biphasic system
Before the report by Dumesic and co-workers
166
on the fruc-
tose dehydration using MIBKwater biphasic configuration
that revitalized the interest in developing novel biphasic
systems for HMF production, the majority of literature on this
topic has focused on using fructose as the reactant.
121,167169
Table 6 Glucose dehydration in ionic liquids
Glucose loading Solvent Catalyst Catalyst loading T (C) Time Conv (%) Selec (%) Yield (%) Ref.
10 wt% DMALiBr (10%) CrBr
3
6 mol% 100 6 h 80 111
10 wt% DMA[EMIM]Cl (10%) CrCl
2
6 mol% 100 6 h 67 111
5 wt% BMImCl H
2
SO
4
13 mol% 120 4 h 98 12 12 174
10 wt% [C
4
mim]Cl CrCl
3
36 wt% 100
a
1 min 91 175
10 wt% [C
4
mim]Cl CrCl
3
36 wt% 100 60 min 17 175
23 wt% [EMim]BF
4
SnCl
4
10 mol% 100 3 h 61 177
5 wt% [BMIM]Cl CrCl
3
10 mol% 140
a
0.5 min 96 74 71 176
5 wt% [BMIM]Cl GeCl
4
10 mol% 140 30 min 48 176
5 wt% [BMIM]Cl Cr-HAP 60 wt% 140
a
150 s 78 51 40 181
20 wt% [EMIM]Cl Al(O
i
Pr)
3
10 mol% 140 6 h 49 179
20 wt% [EMIM]Cl AlEt
3
10 mol% 140 6 h 51 179
23 wt% [EMIM]Cl Cr-NP 10 mol% 140 6 h 49 180
a
Microwave irradiation.
Table 5 Glucose dehydration in biphasic system
Glucose
loading
Aqueous
phase
Organic
phase
A : O phase
ratio Catalyst
Catalyst
loading T (C)
Time
(min)
Conv
(%)
Selec
(%)
Yield
(%) Ref.
6 wt% Water 2-Butanol 2 : 3 (v/v) Nb
2
O
5
8 wt% 170 110 70 70 49 132
6 wt% Water 2-Butanol 2 : 3 (v/v) Ta
2
O
5
8 wt% 160 140 70 83 58 170
30 wt% WaterNaCl MIBK 1 : 4 (v/v) B(OH)
3
100 g/L 150 300 41 34 14 133
30 wt% Water MIBK 4 : 9 (v/v) Ag
3
PW
12
O
40
13.3 wt% 130 240 90 64 58 135
5 wt%
a
Water Cr[(DS)H
2
PW
12
O
40
]
3
0.015 M 170 240 100 70 70 172
5 wt% WaterNaCl SBP 1 : 2 (w/w) Nb/CB-2-DP 133 wt% 170 120 78 26 20 171
5 wt% WaterNaCl SBP 1 : 2 (w/w) Nb/CB-1-DP 133 wt% 170 120 34 52 18 171
6.5 wt% Water MIBK 1 : 3 (v/v) ZrPO 20 wt% 165 360 60 40 24 173
6.5 wt% Water MIBK 1 : 3 (v/v) ZrPO/Si 20 wt% 165 400 40 52 21 173
a
Cellulose was used instead of glucose.
Critical Review Green Chemistry
562 | Green Chem., 2014, 16, 548572 This journal is The Royal Society of Chemistry 2014
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
In this section, updated literature concerning the application
of biphasic systems for glucose conversion will be discussed.
In an eort to develop non-corrosive and non-toxic catalysts
(compared to conventional homogeneous mineral acids, e.g.,
H
2
SO
4
and HCl) for glucose conversion to HMF, Hansen et al.
used a combination of various salts and boric acid, H
3
BO
3
, to
catalyze glucose dehydration in water with MIBK as the extract-
ing organic solvent.
133
In the study, boric acid functioned as a
Lewis acid, which can chelate with diols, resulting in a transi-
ent release of a proton to the solution that could act as a
Brnsted acid to catalyze dehydration. After all the glucose
molecules were consumed, the solution pH returned to the
value determined by the boric acidwater equilibrium in solu-
tion. A 14 mol% HMF yield was reported at a glucose conver-
sion of 41% with a reaction temperature of 150 C and time of
300 min. Other extracting solvents (2-butanol and THF) were
also studied and similar conversion and selectivity results were
found.
Yang et al. examined the use of H
3
PO
4
treated niobic acid
(Nb
2
O
5
nH
2
O, NA-p) as catalysts for glucose conversion.
132
1.2 g glucose was dissolved in 20 mL water with 0.1 g NA-p,
and the solution was contacted with 30 mL of 2-butanol.
A HMF yield of 49 mol% was obtained with 70% glucose con-
version after 110 min reaction time at 170 C. However, HMF
yields dropped after 110 min with higher glucose conversion
due to the formation of undesired products. Catalyst recyclabi-
lity studies were performed by comparing sugar conversions
and HMF selectivities under identical reaction conditions. The
results showed that the catalyst could be reused at least
7 times without significant loss of catalytic performance. The
same group also reported tantalum compounds (Ta
2
O
5
nH
2
O,
TA-p) showed strong acidity and could catalyze carbohydrates
conversion to HMF as well.
170
A higher HMF selectivity com-
pared with Nb
2
O
5
nH
2
O was observed as a reaction time of
140 min and reaction temperature of 160 C led to 70% of the
glucose being converted to give a HMF yield of 58 mol%.
Recently, Xiong et al. prepared niobic acid/carbon hybrid
materials.
171
By tuning the hydrophobicity and hydrophilicity
of the catalyst, the partitioning of the catalyst could be altered
between the hydrophilic aqueous phase, hydrophobic organic
phase, or the interface. With the catalyst residing in the
aqueous phase, a HMF yield of 20 mol% with glucose conver-
sion of 78% at 170 C for 2 h was obtained.
Fan et al. reported glucose conversion to HMF in a water
MIBK biphasic system using solid heteropolyacid salt
(Ag
3
PW
12
O
40
) as the catalyst.
135
They proposed that the pre-
sence of both Lewis and Brnsted acidity could lead to
improved catalytic activity and HMF selectivity. The optimized
HMF yield was 76 mol% for a glucose conversion of 90% at
130 C and 4 h with a water to MIBK ratio of 1 : 2.25.
Work from the same group involved using a Brnsted acid
Lewis acid-surfactant combined heteropolyacid catalyst for
HMF production directly from cellulose.
172
Cr[(DS)-
H
2
PW
12
O
40
]
3
(DS is dodecyl sulfate) was synthesized as the
composite catalyst. The combined Lewis and Brnsted acidity
could catalyze cellulose hydrolysis to glucose with subsequent
dehydration to HMF and the hydrophobic core formed by
dodecyl group was speculated to provide a microenvironment
that protected the produced HMF. Using this strategy, cellulose
was quantitatively converted after 4 h at 170 C, yielding HMF
>70 mol%.
As discussed above, eective operation of these biphasic
systems requires high interfacial surface area for eective mass
transfer. In this regard, Ordomsky et al. prepared solid acid
zirconium phosphate (ZrPO) on an Al foam to address this
issue.
173
Silylation was also conducted to reduce the amount
of Lewis acidity on the surface of the ZrPO, while increasing
the density of Brnsted acid sites with the resultant material
denoted ZrPO/Si. Fewer Lewis sites on the ZrPO/Si led to a
decrease in activity but an increase in HMF selectivity, which
was due to suppression of humin formation by Lewis acid
sites.
As outlined in this section, glucose conversion in a biphasic
system faces the same challenges with mass transfer between
the organicaqueous interfaces as with fructose and harsher
reaction conditions are required to activate glucose. Strong in-
organic mineral acids and metal oxides have been applied
together with various salt additives in the biphasic reaction
system. To achieve satisfactory HMF yields, higher temperature
and longer reaction time were typically needed relative to the
fructose reaction.
As such, it would be advantageous to integrate the glucose
fructose isomerization with subsequent dehydration for
glucose conversion, which will be discussed in detail in
section 6.
5.3 Glucose dehydration in ionic liquids
Since the breakthrough work by Zhang et al.
78
in utilizing
ionic liquids (IL) as solvents and co-catalysts for HMF
production from carbohydrates (fructose and glucose), an
expanding interest has been reflected in the rapidly growing
number of reports in the literature in using ILs. Several excel-
lent reviews have summarized this topic.
30,58
In this section,
we will summarize the work on dierent compositions of ILs,
catalysts and reaction conditions and how these factors
aected glucose conversion activity and the ultimate HMF
yields.
Binder et al. reported a simple reaction system for the con-
version of glucose, fructose and full lignocellulosic biomass
using a solvent system that resembled the properties of ILs.
111
In their strategy, the polar aprotic organic solvent DMA was
modified with LiCl, which was used to successfully solubilize
carbohydrates. By optimizing the composition of CrX
n
(X = Cl,
Br; n = 2, 3) and [BMIM]Cl, the highest glucose molar yield of
80% was achieved with a catalyst system composed of DMA
LiBr (10%), 6 mol% CrBr
3
catalyst, and reaction conditions of
6 h at 100 C.
Later, a Brnsted acid (H
2
SO
4
) was added to the ionic liquid
[BMIm]Cl by Sievers et al. to test the eectiveness in converting
glucose, fructose and mannose to HMF.
174
Interestingly, a low
HMF yield of only 12 mol% was obtained from glucose, which
was significantly lower than the results by Zhang et al.
78
This
Green Chemistry Critical Review
This journal is The Royal Society of Chemistry 2014 Green Chem., 2014, 16, 548572 | 563
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
result suggested that for ecient glucose transformation to
HMF under this condition a Lewis acid (e.g., CrCl
2
) is
advantageous.
Microwave irradiation (MI) was found to play a key role in
controlling HMF selectivities when using ILs. Li et al. observed
that a HMF yield of 91 mol% could be achieved in [C
4
mim]Cl
with a CrCl
3
catalyst using MI at 100 C for 1 min.
175
A conven-
tional oil-bath heating method was used for comparison. For
similar conversion values (60 min of reaction in the conven-
tional system) only a 17 mol% yield was obtained. Qi et al.
reported similar observations on the enhancement eect of
using MI as the heating method. In their study, a 71 mol%
HMF yield was achieved in a reaction system with [BMIM]Cl
and a CrCl
3
catalyst.
176
Instead of using the eective CrCl
x
(x = 2, 3) catalyst employed
by in the above work, Hu et al. extended the catalysts to include
a commonly used Lewis acid, SnCl
4
.
177
They ran the reaction at
100 C for 3 h in [BMIM]BF
4
, obtaining a HMF yield of 61 mol
%. Zhang et al. later examined the use of GeCl
4
to replace the
toxic chromium and tin as a Lewis acid for glucose dehy-
dration.
142
The reaction conditions were 5 wt% glucose dissolved
in [BMIM]Cl, followed by heating at 120 C for 30 min. A moder-
ate HMF yield of 48 mol% was detected. To that end, Khokhlova
et al. also reported B
2
O
3
as an environmentally benign metal-
free promoter for carbohydrate conversion in ILs.
178
A study by Liu et al. demonstrated that ubiquitous alumi-
num alkyl or alkoxy compounds were eective Lewis acids for
glucose to HMF transformation in ILs.
179
At 120 C for 6 h,
HMF yields from glucose catalyzed by Al(O
i
Pr)
3
and AlEt
3
were
49 mol% and 51 mol%, respectively. Chen and co-workers also
reported Cr
0
nanoparticles could function as eective catalysts
for the glucose to HMF conversion.
180
A moderate yield of
49 mol% in [BMIM]Cl was obtained from 23 wt% glucose solu-
tion using reaction conditions of 120 C for 6 h.
To address the challenge of catalyst recycling and separation
from the reaction system, Zhang et al. prepared hydroxyapatite-
supported chromium chloride (Cr-HAP) and examined its
activity.
181
With MI heating at 150 C for 2.5 min, they reported
a HMF yield of 40 mol% at a glucose conversion of 78%.
These studies demonstrated that high HMF yields required
an optimized combination of ILs having appropriate hydro-
philic/hydrophobic properties with eective Brnsted or Lewis
acids. As was the case with fructose, stronger Lewis acids
generally led to higher HMF yields. In addition to the cost
associated with using ILs, the stability of ILs in contact with
water produced during dehydration need to be examined
under reaction conditions. An interesting direction is the
immobilization of ILs onto a solid catalyst support as shown
by Zhang et al.
181
However, one concern that could limit this
approach is the possible leaching of Lewis acidic salts that are
electrostatically bound on the catalyst surface. As an alterna-
tive, the complexation of a Lewis acid with a pre-organized
multi-valent ligand on the catalyst surface might be able to
overcome this issue.
6. Combined glucose to fructose
isomerization and fructose
dehydration to HMF strategy
After reviewing the current approaches used to convert glucose
to HMF, it is apparent that an aqueousorganic biphasic
system is generally attractive for improving HMF yields. MCl
x
/
ionic liquids were also used to convert both glucose and fruc-
tose to HMF in high yields. However, the commercial use of
expensive ILs as a processing solvent is challenging due to the
stability and purity requirements for the ILs.
182
Progress has
been made in combining a glucose isomerization catalyst with
a fructose dehydration catalyst in either a one-pot or two-
step configuration. Based on the knowledge that Brnsted
acid catalysts give the best HMF selectivities for the fructose
dehydration step, researchers have pursued the identification
of eective glucose isomerization catalysts and successes have
been reported in using bases, enzymes, or Lewis acids for this
purpose. As such, isomerization of glucose can therefore be
integrated with fructose dehydration to produce HMF (Fig. 10).
6.1. Combined base and Brnsted acid systems
In general, glucosefructose isomerization is known through
the Lobry-deBruynvan Ekenstein transformation, which is
catalyzed by a base. Takagaki et al. hypothesized a tandem
approach, which consisted of a solid base catalyst for isomeri-
zation of glucose to fructose and a solid acid catalyst for fruc-
tose dehydration.
183
After initial catalyst screening for the
activity of the two individual steps, MgAl hydrotalcite and
Amberlyst-15 were chosen as the solid base and acid, respecti-
vely. To create site isolation of acids and bases, an organic
aprotic solvent was the preferred solvent since it can minimize
the possibility of mutual interaction of an acidbase pair. DMF
was selected as the media for this one-pot transformation.
The combination of the base and acid catalysts was demon-
strated to be eective, achieving glucose conversion of 73%
with selectivity towards HMF of 58%.
Fig. 10 Synthesis of HMF from glucose using combined isomerization/dehydration catalysts.
Critical Review Green Chemistry
564 | Green Chem., 2014, 16, 548572 This journal is The Royal Society of Chemistry 2014
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
Despax et al. later examined the eectiveness of sodium
aluminate (NaAlO
2
) and HCl as a solid base and acid pair for
glucose conversion and achieved up to 82 mol% HMF yield.
184
DMSO was used as the solvent, which may cause concerns
about the ultimate separation and purification of HMF.
While it has been shown that the combination of base-cata-
lyzed isomerization of glucose to fructose and subsequent
Brnsted acid-catalyzed dehydration reaction can produce
HMF in relatively high yields, the approach has several limit-
ations. First, the byproduct water produced in the one-pot
glucose dehydration to HMF could lead to mutual destruction
of the base and acid moieties even if a neat organic solvent
was chosen initially (use of an adsorbent material might be
able to address this issue). Additionally, the requirement of a
non-aqueous environment for the base and acid moieties to
co-exist will limit the choice of solvent to one such as DMSO,
which will create added costs for the subsequent purification.
6.2. Combined enzymatic and Brnsted acid catalysts
systems
Enzymes are a potentially attractive catalysts for performing
glucose isomerization to fructose because of their high selectivity
under mild reaction conditions.
185
A multi-step novel reactor
design and separation scheme was shown that allowed for very
high yields of HMF directly from glucose (up to 90 mol%).
However, concerns still remain in the application of enzymes due
to the need to perform multiple and separate processing steps.
Huang et al. used a combination of glucose isomerase and
sodium tetraborate to transform glucose to fructose.
186
A fruc-
tose yield of up to 88 mol% (equilibrium without borate is
50% maximum yield) was realized by exploiting the superior
chelating eect of sodium tetraborate with ketoses (fructose)
relative to aldoses (glucose). After the separation of fructose
from the glucose isomerase, a second, HCl-catalyzed step for
fructose dehydration was conducted in a biphasic configur-
ation with 1-butanol as the organic extracting solvent. The
overall glucose conversion from glucose was 88% with the
final HMF yield reaching 63 mol%.
Grande et al. reported a chemo-enzymatic process that
could be performed in seawater as an abundantly available
and environmentally friendly solvent.
185
In their experiments,
they used immobilized D-glucosexylose isomerase (EC 5.3.1.5)
for the isomerization of glucose to fructose until the equili-
brium yield of fructose was reached (ca. 50 : 50 glucosefruc-
tose). Then in a separate second-step, fructose (without
glucose present) was dehydrated by oxalic acid in a biphasic
system with 2-methyltetrahydrofuran (2-MTHF) as the extract-
ing solvent to give 57 mol% yields in seawater. Although the
individual isomerization/dehydration steps could be per-
formed in seawater, an integrated process, which directly uses
the fructoseglucose mixture produced in the isomerization
step, would be required for practical application.
The ionic liquid N,N-dibutylethanolammonium octanoate
(DBAO) was also used by Riisager and co-workers as the media
for enzymatic glucose isomerization.
187
In their study, Sweet-
zyme was chosen as a model glucose isomerase and dierent
ILs were screened as a solvent for glucose to fructose isomeri-
zation. Kinetic analysis of glucose isomerization catalyzed by
Sweetzyme in DBAO was performed and it was revealed that a
longer reaction time (4 h) was required to reach the glucose
fructose equilibrium compared to an experiment done in an
aqueous media (0.5 h).
An interesting report by Li et al. outlined a novel simul-
taneous glucose isomerization and reactive-extraction strategy
to separate the in situ produced fructose (Fig. 11).
188
In the
first glucose isomerization step (12a), enzymes were used to
catalyze the transformation to an equilibrium mixture of
glucose and fructose. To shift the equilibrium to favor ketose
formation, an eective method could be removal of the ketose.
The authors devised a biphasic system in which a hydrophobic
aryl boronic acid (ABA) containing organic phase could continu-
ously and selectively extract the ketose into the organic phase,
whereas the aldose-form of the sugar was held in the aqueous
phase where the isomerization occurred. In a second step (12b),
contact with an acidic aqueous layer separated the conjugated
ketose-ABA complex. This integrated process resulted in a pure
ketose stream from aldose sugars, creating the possibility for
integration with a subsequent dehydration reaction step. Recent
work from the same research group described the use of this
strategy to perform a xylose to xylulose to furfural tandem reac-
tion. Under a relatively mild reaction temperature of 110 C,
they reported over an 80 mol% yield of furfural from xylose.
189
Simeonov et al. also studied the chemo-enzymatic strategy
for ecient glucose to HMF conversion.
110,190
In their reports,
Sweetzyme was applied as the isomerization catalyst in a wet
tetraethylammonium bromide (TEAB) media (Fig. 12). Con-
ducting isomerization and dehydration in the same media and
recycling the unreacted glucose from the second step allowed
the highest reported yield of HMF, 91 mol% from glucose.
In summary, studies integrating enzymatic isomerization of
glucose with chemo-catalytic dehydration have shown the
eectiveness of this approach in obtaining the highest
reported HMF yields. From a thermodynamic point of view,
the maximum yields of ketose from aldose are 50% and as a
result, the ketose must be removed from the equilibrium in
order to shift the reaction towards formation of the ketose.
Dierent approaches have been used, either relying on a com-
plexing agent like boric acid or by iteratively shifting the equili-
brium through conversion of the more reactive ketose to HMF
followed by repeated isomerization.
However, limitations also exist for this approach. While
chemo-catalytic dehydration is a relatively rapid reaction, the enzy-
matic isomerization requires a significantly longer time and as
such, the throughput dierence can represent a challenge in inte-
grating the two steps. In this regard, the development of some
hexose isomerase biomimics which could work in concert with a
subsequent dehydration catalyst could be an interesting area.
6.3 Combined Lewis acid and Brnsted acid systems
In contrast to the two-step requirement for the combined
chemo-enzymatic approach, combined Lewis and Brnsted
acid catalyst systems could provide a more easily integrated
Green Chemistry Critical Review
This journal is The Royal Society of Chemistry 2014 Green Chem., 2014, 16, 548572 | 565
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
one-step process in an aqueous media.
191
In general, a Lewis
acid is required for glucose isomerization to fructose, followed
by a Brnsted acid-catalyzed dehydration of fructose to HMF.
The Lewis acids reported in literature have been either hetero-
geneous zeolitic Lewis acids or homogeneous metal salts.
6.3.1. Solid Lewis acid and Brnsted acid systems. In 2010,
a tin-containing beta zeolite (Sn-beta) was reported to be an
active catalyst for the isomerization of glucose to fructose in
water with activities and selectivities rivaling those of
enzymes.
192
Subsequently, several mechanistic studies have
been performed to elucidate how the isomerization reaction
proceeds in the presence of Sn-beta.
76,77
It was shown that the
hydrophobic Sn-beta provides a microenvironment that
appears to be analogous to that in a metalloenzyme, glucose
isomerase. NMR, in conjunction with labeling experiments,
showed that glucose could enter the micropores of the zeolite
and undergo glucosefructose isomerization via a 1,2-hydride
transfer mechanism catalyzed by the tin Lewis acid center.
Unlike its enzyme counterparts, Sn-beta can function under
acidic conditions and elevated temperatures, which made it
Fig. 11 Process of (a) reactive extraction and (b) sugar recovery via stripping.
188
Reproduced from ref. 188.
Fig. 12 Integrated process for HMF production and isolation from glucose.
190
Reproduced from ref. 190 with permission from Wiley.
Critical Review Green Chemistry
566 | Green Chem., 2014, 16, 548572 This journal is The Royal Society of Chemistry 2014
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
possible to integrate the subsequent acid-catalyzed fructose to
HMF conversion.
Nikolla et al. exploited the combination of Sn-beta with HCl
for glucose conversion in a biphasic system (Fig. 13).
82
A HMF
selectivity of 55 mol% was obtained at a glucose conversion of
75% with the combination of Sn-beta and HCl at 160 C for
90 min in a water1-butanol biphasic system. When only Sn-
beta was used, a 18 mol% HMF selectivity at 75% glucose con-
version was observed and with only HCl a 40 mol% HMF
selectivity at 26% glucose conversion. The improved overall
HMF yield from Sn-beta/HCl was attributed to the synergistic
catalytic activity for the individual isomerization/dehydration
steps. Further optimization of the catalytic system gave a HMF
selectivity of 72 mol% at 79% glucose conversion with reaction
conditions of: waterTHF biphasic system, catalysis by a com-
bination of Sn-beta and HCl, reaction temperature of 180 C,
and reaction time of 70 min.
Lew et al. reported application of the combination of Sn-beta
and an acidic resin, Amberlyst, for the synthesis of 5-EMF directly
from glucose in ethanol.
45
A 31 mol% yield of EMF was obtained
with the combined Sn-beta/Amberlyst catalysts at 90 C.
While quite promising, Sn-beta is not a commercially avail-
able zeolite, is challenging to prepare and introduces environ-
mental concerns associated with the toxic element, tin. To
overcome these concerns, Saravanamurugan et al. used
dierent types of commercially available zeolites and described
a two-step pathway to make fructose from glucose in methanol
and water.
193
In the first step, glucose was isomerized to fruc-
tose and then the fructose was reacted with methanol to form
methyl fructoside. In the second step, methyl fructoside was
hydrolyzed by the addition of water, resulting in fructose pro-
duction (Fig. 14). Y-zeolites were found to be the most eective
catalysts. After optimization, high fructose yields (>50 mol%)
were achieved using this two-step process with H-USY(6).
However, the limited solubility of glucose in alcohol could be a
concern for this approach since only a 3 wt% glucose in metha-
nol was used, whereas a 45 wt% glucose aqueous solution was
used in the abovementioned work by the Davis group.
192
In addition to using Sn-beta for C
6
sugars conversion, a
considerable amount of work has also been reported on the
dehydration of C
5
sugars to furfural in high yields.
194
Unlike the limited reaction conditions required for
enzymes to perform isomerization of glucose, the inorganic
metalloenzyme mimics could carry out the reaction in harsher
conditions such as acidic conditions at higher temperature.
For this reason, it was possible to integrate isomerization and
dehydration in a one-pot reactor configuration. Biphasic
reactor system approach discussed earlier has been widely
adopted in this type of work as well to obtain high HMF yields.
6.3.2. MCl
x
-type Lewis acid and Brnsted acid systems.
Another area that has attracted growing interest for glucose to
HMF conversion involves the use of a combination of various
MCl
x
-type Lewis acidic metal salts (M represents the metal)
and Brnsted acids, typically HCl. It has been shown that the
solution pH plays a key role in controlling the Lewis acidity of
the MCl
x
catalyst.
195
In that regard, traditional moisture-sensi-
tive and water-compatible metal salts both have been tested as
catalysts for the glucose dehydration reaction under appropri-
ate pH conditions.
In 2012, Dumesic and co-workers reported a catalytic
system in which a combination of Lewis acidic metal salts
(e.g., AlCl
3
, SnCl
4
, GaCl
3
, YbCl
3
, LaCl
3
) and HCl were used for
glucose conversion in a biphasic system.
83
Depending on the
nature of the MCl
3
the dierent metal salts had dierent
Fig. 13 One-pot HMF production from glucose using Sn-beta zeolites.
82
Reprinted with permission from E. Nikolla, Y. Romn-Leshkov,
M. Moliner and M. E. Davis, ACS Catal., 2011, 1, 408410. Copyright 2011 American Chemical Society.
Fig. 14 Reaction pathway for fructose formation from glucose in
alcohol and aqueous media.
193
Reprinted with permission from
S. Saravanamurugan, M. Paniagua, J. A. Melero and A. Riisager, J. Am.
Chem. Soc., 2013, 130402104357006. Copyright 2013 American Chemi-
cal Society.
Green Chemistry Critical Review
This journal is The Royal Society of Chemistry 2014 Green Chem., 2014, 16, 548572 | 567
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
activity in catalyzing glucose conversion. In order to compare
selectivity towards HMF for various catalysts, a similar glucose
conversion (90%) was targeted. At this conversion level, the
HMF yields ranged from 40 mol% to 60 mol% based on
glucose. AlCl
3
gave the highest selectivity of 67 mol%. The
authors attributed the dierent activities of the MCl
3
catalysts
to their intrinsic Lewis character, such as the ionic radius of
the cation, hardness/softness of the Lewis acid, metal specia-
tion, etc. Another study by Wang et al. revealed that using
water-compatible Lewis acids (e.g., YbCl
3
, DyCl
3
, LaCl
3
), could
promote glucose conversion under higher solution pH values
(5).
84
Other carbohydrates such as cellobiose, starch, or even
cellulose could be converted to HMF in moderate yields.
Yang et al. achieved 61 mol% HMF yield from glucose
using a similar combined Lewis acid, AlCl
3
, and Brnsted acid,
HCl, after 10 min of microwave heating at 160 C in a water
THF biphasic system.
85
The authors attributed the activity of
this reaction system to a synergistic eect between AlCl
3
and
HCl. Another report from the same group extended this cata-
lyst system to perform xylose dehydration with a similar
aldoseketose isomerization followed by dehydration to give a
furfural yield of 75 mol%.
75
As CrCl
3
has been widely used in IL-based glucose de-
hydration systems, Vlachos and co-workers have successfully
demonstrated their activity for C
5
sugar dehydration in a
waterorganic biphasic system.
74,87
In a recent publication, the
authors have used experimental kinetic data and compu-
tational methods to decipher the interplay between Lewis and
Brnsted acidity in controlling the activities and selectivities in
C
6
sugar (glucose and fructose) dehydration.
88
Deng et al. reported the use of concentrated ZnCl
2
as a
water-compatible Lewis acid for the conversion of various
carbohydrates to HMF.
196
The high concentration of ZnCl
2
in
water was proposed to facilitate coordination with the
hydroxyl/carbonyl groups in glucose thereby catalyzing glucose
isomerization to fructose. The protons generated by the ZnCl
2
-
acidified solution could then catalyze fructose dehydration. At
120 C, after full conversion of the carbohydrate, 16 mol% and
53 mol% HMF yields were obtained from glucose and fructose,
respectively. A combination of ZnSO
4
and NaHSO
4
was
explored by Shi et al. to convert cellulose directly to HMF
giving a 53 mol% yield.
197
The reaction was run in a water
THF biphasic system with a reaction temperature of 170 C
and time of 60 min.
A multitude of Lewis acidic metal salts have been examined
for the ecacy in glucose dehydration. These salts can gene-
rally be classified as water-compatible and water-sensitive
Lewis acids, which behave dierently in response to solution
pH changes. However, under suitable pH values improved
HMF yields were obtained when using stronger Lewis acids. It
would be advantageous to immobilize the Lewis acid metal
salts to make heterogeneous systems, however, the high level
of added salts needed to facilitate HMF partitioning in bipha-
sic system would impose challenges. As a result, the alternative
monophasic reactor configuration discussed in section
4.1 may be worth considering.
7. Conclusion and outlook
Interest in HMF production has experienced a rapid expansion
over the past decade. Not only have researchers utilized
more abundantly available feedstocks, but also novel catalysts,
processing solvents, reaction engineering and HMF deri-
vatization have been explored leading to a revitalized
interest in HMF. Recent significant breakthroughs in the field
include, but are not limited to, the use of ILs as novel solvents
and co-catalysts and the combination of CrCl
x
/IL as an
eective catalyst system to convert C
6
sugars. Also, a novel reac-
tion engineering approach involving the reemergence and
modification of waterorganic biphasic systems have shown
dramatically improved HMF yields. Another important devel-
opment is the used of either heterogeneous or homogeneous
Lewis acid catalysts, which can activate C
6
sugars in aqueous
media.
Amongst all the catalyst/reaction systems reviewed, a com-
bined Lewis acid-catalyzed isomerization and subsequent
Brnsted acid-catalyzed dehydration that converts glucose to
HMF in either an aqueousorganic biphasic or ionic liquid
one-pot reactor configuration seems particularly promising.
While a large number of combined catalytic systems have been
reported to have ecacy in converting glucose to HMF with
good selectivities, very few studies have been performed to
examine the kinetics of the respective isomerization and
dehydration steps. Better understanding of the intrinsic
kinetics will allow for a systematic determination of the
optimal ratio of the two catalytic functions rather than the
empirical approach that has been primarily used to date.
This information will also provide a solid foundation for the
future development of processes and reactors for HMF pro-
duction. Additionally, many of the reported Lewis acidic metal
salts readily hydrolyze upon contact with water, which can
either be present as the solvent or being produced during the
reaction. Therefore, studies that can distinguish Brnsted and
Lewis acidity and their eect on the catalytic process are
important. Finally, an area that has received relatively little
attention is the evaluation of catalyst stability under real
process conditions. There is a need to identify catalyst systems
with little catalyst leaching and extended hydrothermal
stability.
Also, both life cycle and techno-economic analysis of the
HMF production strategies are required as well the environ-
mental impact of the upstream and downstream processing to
generate HMF from carbohydrate feedstocks. The rapid rate of
recent innovation in HMF production suggests that a viable
process will be realized.
Acknowledgements
The authors thank NSF award EEC-0813570 for providing
financial support.
Critical Review Green Chemistry
568 | Green Chem., 2014, 16, 548572 This journal is The Royal Society of Chemistry 2014
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
Notes and references
1 M. E. Davis, Nature, 2002, 417, 813821.
2 G. W. Stone, S. G. Ellis, D. A. Cox, J. Hermiller,
C. OShaughnessy, J. T. Mann, M. Turco, R. Caputo,
P. Bergin, J. Greenberg, J. J. Popma, M. E. Russell and
T.-I. Investigators, New Eng. J. Med., 2004, 350, 221231.
3 J. A. Geboers, S. Van de Vyver, R. Ooms, B. O. de Beeck,
P. A. Jacobs and B. F. Sels, Catal. Sci. Technol., 2011, 1,
714726.
4 H. Kobayashi, H. Ohta and A. Fukuoka, Catal. Sci.
Technol., 2012, 2, 869883.
5 R. A. Gross and B. Kalra, Science, 2002, 297, 803807.
6 J. Q. Bond, D. M. Alonso, D. Wang, R. M. West and
J. A. Dumesic, Science, 2010, 327, 11101114.
7 B. H. Shanks, Ind. Eng. Chem. Res., 2010, 49.
8 B. H. Shanks, Abs. Pap. Am. Chem. Soc., 2011, 242.
9 K. L. Deutsch and B. H. Shanks, J. Catal., 2012, 285.
10 B. Cinlar, T. Wang and B. H. Shanks, Appl. Catal., A, 2013,
450, 237242.
11 S. Lima, M. M. Antunes, M. Pillinger and A. A. Valente,
ChemCatChem, 2011, 3, 16861706.
12 M. E. Zakrzewska, E. Bogel-ukasik and R. Bogel-ukasik,
Chem. Rev., 2010, 111, 397417.
13 H. Kobayashi and A. Fukuoka, Green Chem., 2013, 15,
17401763.
14 N. Mosier, C. Wyman, B. Dale, R. Elander, Y. Y. Lee,
M. Holtzapple and M. Ladisch, Bioresour. Technol., 2005,
96, 673686.
15 R. Rinaldi and F. Schueth, ChemSusChem, 2009, 2, 1096
1107.
16 R. H.-Y. Chang, J. Jang and K. C. W. Wu, Green Chem.,
2011, 13, 28442850.
17 J. Zaldivar, A. Martinez and L. O. Ingram, Biotechnol.
Bioeng., 1999, 65, 2433.
18 A. Takimoto, T. Shiomi, K. Ino, T. Tsunoda, A. Kawai,
F. Mizukami and K. Sakaguchi, Microporous Mesoporous
Mater., 2008, 116, 601606.
19 C. E. Wyman, B. E. Dale, R. T. Elander, M. Holtzapple,
M. R. Ladisch and Y. Y. Lee, Bioresour. Technol., 2005, 96,
19591966.
20 D. M. Alonso, S. G. Wettstein, M. A. Mellmer, E. I. Gurbuz
and J. A. Dumesic, Energy Environ. Sci., 2013, 6, 7680.
21 I. Krog-Mikkelsen, O. Hels, I. Tetens, J. J. Holst, J. R. Andersen
and K. Bukhave, Am. J. Clin. Nutr., 2011, 94, 472478.
22 A. S. Mamman, J.-M. Lee, Y.-C. Kim, I. T. Hwang,
N.-J. Park, Y. K. Hwang, J.-S. Chang and J.-S. Hwang, Bio-
fuels, Bioprod. Biorefin., 2008, 2, 438454.
23 J.-P. Lange, E. van der Heide, J. van Buijtenen and
R. Price, ChemSusChem, 2012, 5, 150166.
24 J. Zakzeski, P. C. A. Bruijnincx, A. L. Jongerius and
B. M. Weckhuysen, Chem. Rev., 2010, 110, 35523599.
25 A. Corma, S. Iborra and A. Velty, Chem. Rev., 2007, 107,
24112502.
26 V. M. Roberts, V. Stein, T. Reiner, A. Lemonidou, X. Li and
J. A. Lercher, Chem.Eur. J., 2011, 17, 59395948.
27 C. Zhao, J. He, A. A. Lemonidou, X. Li and J. A. Lercher,
J. Catal., 2011, 280, 816.
28 C. Zhao, Y. Kou, A. A. Lemonidou, X. Li and J. A. Lercher,
Chem. Commun., 2010, 46, 412414.
29 P. Azadi, R. Carrasquillo-Flores, Y. J. Pagan-Torres,
E. I. Guerbuez, R. Farnood and J. A. Dumesic, Green
Chem., 2012, 14, 15731576.
30 R.-J. van Putten, J. C. van der Waal, E. de Jong,
C. B. Rasrendra, H. J. Heeres and J. G. de Vries, Chem.
Rev., 2013, 130211074411009.
31 A. J. J. E. Eerhart, A. P. C. Faaij and M. K. Patel, Energy
Environ. Sci., 2012, 5, 6407.
32 G.-J. M. Gruter, L. Sipos and M. A. Dam, Comb. Chem.
High Throughput Screening, 2012, 15, 180188.
33 A. Gandini and M. N. Belgacem, Prog. Polym. Sci., 1997,
22, 12031379.
34 O. Casanova, S. Iborra and A. Corma, ChemSusChem,
2009, 2, 11381144.
35 Y. Y. Gorbanev, S. K. Klitgaard, J. M. Woodley,
C. H. Christensen and A. Riisager, ChemSusChem, 2009, 2,
672675.
36 T. Pasini, M. Piccinini, M. Blosi, R. Bonelli, S. Albonetti,
N. Dimitratos, J. A. Lopez-Sanchez, M. Sankar, Q. He,
C. J. Kiely, G. J. Hutchings and F. Cavani, Green Chem.,
2011, 13, 20912099.
37 B. Saha, S. Dutta and M. M. Abu-Omar, Catal. Sci.
Technol., 2012, 2, 7981.
38 S. E. Davis, L. R. Houk, E. C. Tamargo, A. K. Datye and
R. J. Davis, Catal. Today, 2011, 160, 5560.
39 S. E. Davis, B. N. Zope and R. J. Davis, Green Chem., 2012,
14, 143147.
40 B. Saha, D. Gupta, M. M. Abu-Omar, A. Modak and
A. Bhaumik, J. Catal., 2013, 299, 316320.
41 R. Alamillo, M. Tucker, M. Chia, Y. Pagan-Torres and
J. Dumesic, Green Chem., 2012, 14.
42 J. D. Garber, R. E. Jones and U. Torleif, US, 3083236, 1963.
43 M. Balakrishnan, E. R. Sacia and A. T. Bell, Green Chem.,
2012, 14, 1626.
44 G. A. Kraus and T. Guney, Green Chem., 2012, 14, 1593.
45 C. M. Lew, N. Rajabbeigi and M. Tsapatsis, Ind. Eng.
Chem. Res., 2012, 51, 53645366.
46 X. Tong, Y. Ma and Y. Li, Appl. Catal., A, 2010, 385,
113.
47 Y. Romn-Leshkov, C. J. Barrett, Z. Y. Liu and
J. A. Dumesic, Nature, 2007, 447, 982985.
48 M. Chidambaram and A. T. Bell, Green Chem., 2010, 12,
1253.
49 G. W. Huber, J. N. Chheda, C. J. Barrett and J. A. Dumesic,
Science, 2005, 308.
50 R. M. West, Z. Y. Liu, M. Peter, C. A. Gaertner and
J. A. Dumesic, J. Mol. Catal. A: Chem., 2008, 296.
51 W. Shen, G. A. Tompsett, K. D. Hammond, R. Xing,
F. Dogan, C. P. Grey, W. C. Conner Jr., S. M. Auerbach and
G. W. Huber, Appl. Catal., A, 2011, 392, 5768.
52 A. Corma, l. T. O. de and M. Renz, ChemSusChem, 2011, 4,
15741577.
Green Chemistry Critical Review
This journal is The Royal Society of Chemistry 2014 Green Chem., 2014, 16, 548572 | 569
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
53 A. Corma, l. T. O. de and M. Renz, Energy Environ. Sci.,
2012, 5, 63286344.
54 J. Yang, N. Li, G. Li, W. Wang, A. Wang, X. Wang, Y. Cong
and T. Zhang, ChemSusChem, 2013, 6, 11491152.
55 L. Hu, G. Zhao, W. Hao, X. Tang, Y. Sun, L. Lin and S. Liu,
RSC Adv., 2012, 2, 11184.
56 X. Qian, Top. Catal., 2012, 55, 218226.
57 R. Karinen, K. Vilonen and M. Niemel, ChemSusChem,
2011, 4, 10021016.
58 T. Sthlberg, W. Fu, J. M. Woodley and A. Riisager, Chem-
SusChem, 2011, 4, 451458.
59 A. Ranoux, K. Djanashvili, I. W. C. E. Arends and
U. Hanefeld, ACS Catal., 2013, 760763.
60 B. F. M. Kuster, Starch-Starke, 1990, 42, 314321.
61 R. S. Assary, P. C. Redfern, J. R. Hammond, J. Greeley and
L. A. Curtiss, J. Phys. Chem. B, 2010, 114, 90029009.
62 M. J. Antal Jr., W. S. L. Mok and G. N. Richards, Carbo-
hydr. Res., 1990, 199, 91109.
63 F. H. Newth, Adv. Carbohydr. Chem. Biochem., 1951, 6, 83
106.
64 A. S. Amarasekara, L. D. Williams and C. C. Ebede, Carbo-
hydr. Res., 2008, 343, 30213024.
65 C. Moreau, R. Durand, S. Razigade, J. Duhamet,
P. Faugeras, P. Rivalier, P. Ros and G. Avignon, Appl.
Catal., A, 1996, 145, 211224.
66 M. J. Antal, T. Leesomboon, W. S. Mok and
G. N. Richards, Carbohydr. Res., 1991, 217, 7185.
67 X. H. Qian, M. R. Nimlos, D. K. Johnson and
M. E. Himmel, Appl. Biochem. Biotechnol., 2005, 121, 989
997.
68 R. S. Assary, P. C. Redfern, J. Greeley and L. A. Curtiss,
J. Phys. Chem. B, 2011, 115, 43414349.
69 X. Qian, M. R. Nimlos, M. Davis, D. K. Johnson and
M. E. Himmel, Carbohydr. Res., 2005, 340, 23192327.
70 G. G. Ferenczy, G. Naray-Szabo and P. Varnai,
Int. J. Quantum Chem., 1999, 75, 215222.
71 T. D. Fenn, D. Ringe and G. A. Petsko, Biochemistry, 2004,
43, 64646474.
72 M. R. Nimlos, X. Qian, M. Davis, M. E. Himmel and
D. K. Johnson, J. Phys. Chem. A, 2006, 110, 1182411838.
73 G. Marcotullio and W. De Jong, Green Chem., 2010, 12,
17391746.
74 V. Choudhary, A. B. Pinar, S. I. Sandler, D. G. Vlachos and
R. F. Lobo, ACS Catal., 2011, 1, 17241728.
75 Y. Yang, C.-W. Hu and M. M. Abu-Omar, ChemSusChem,
2012, 5, 405410.
76 Y. Roman-Leshkov, M. Moliner, J. A. Labinger and
M. E. Davis, Angew. Chem., Int. Ed., 2010, 49.
77 R. Bermejo-Deval, R. S. Assary, E. Nikolla, M. Moliner,
Y. Roman-Leshkov, S.-J. Hwang, A. Palsdottir,
D. Silverman, R. F. Lobo, L. A. Curtiss and M. E. Davis,
Proc. Natl. Acad. Sci. U. S. A., 2012, 109.
78 H. Zhao, J. E. Holladay, H. Brown and Z. C. Zhang,
Science, 2007, 316, 15971600.
79 E. A. Pidko, V. Degirmenci, R. A. van Santen and
E. J. M. Hensen, Angew. Chem., Int. Ed., 2010, 49, 25302534.
80 F. Meilleur, E. H. Snell, M. J. van der Woerd, R. A. Judge
and D. A. A. Myles, Eur. Biophys. J. Biophys. Lett., 2006, 35,
601609.
81 M. Moliner, Y. Roman-Leshkov and M. E. Davis, Proc.
Natl. Acad. Sci. U. S. A., 2010, 107, 61646168.
82 E. Nikolla, Y. Romn-Leshkov, M. Moliner and
M. E. Davis, ACS Catal., 2011, 1, 408410.
83 Y. J. Pagan-Torres, T. Wang, J. M. R. Gallo, B. H. Shanks
and J. A. Dumesic, ACS Catal., 2012, 2.
84 T. Wang, Y. J. Pagan-Torres, E. J. Combs, J. A. Dumesic
and B. H. Shanks, Top. Catal., 2012, 55.
85 Y. Yang, C.-w. Hu and M. M. Abu-Omar, Green Chem.,
2012, 14, 509513.
86 S. Dutta, S. De, M. I. Alam, M. M. Abu-Omar and B. Saha,
J. Catal., 2012, 288, 815.
87 V. Choudhary, S. I. Sandler and D. G. Vlachos, Acs Catal.,
2012, 2, 20222028.
88 V. Choudhary, S. H. Mushrif, C. Ho, A. Anderko,
V. Nikolakis, N. S. Marinkovic, A. I. Frenkel, S. I. Sandler
and D. G. Vlachos, J. Am. Chem. Soc., 2013, 135, 3997
4006.
89 B. Girisuta, L. P. B. M. Janssen and H. J. Heeres, Green
Chem., 2006, 8, 701709.
90 S. K. R. Patil and C. R. F. Lund, Energy Fuels, 2011, 25,
47454755.
91 B. F. M. Kuster and L. M. Tebbens, Carbohydr. Res., 1977,
54, 158164.
92 B. F. M. Kuster and H. M. G. Temmink, Carbohydr. Res.,
1977, 54, 185191.
93 B. F. M. Kuster and H. S. van der Baan, Carbohydr. Res.,
1977, 54, 165176.
94 B. F. M. Kuster, Carbohydr. Res., 1977, 54, 177183.
95 Y. Nakamura and S. Morikawa, Bull. Chem. Soc. Jpn., 1980,
53, 37053706.
96 R. M. Musau and R. M. Munavu, Biomass, 1987, 13, 6774.
97 H. E. van Dam, A. P. G. Kieboom and H. van Bekkum,
Starch/Starke, 1986, 38, 95101.
98 P. Vinke and H. van Bekkum, Starch/Starke, 1992, 44, 90
96.
99 T. Armaroli, G. Busca, C. Carlini, M. Giuttari,
G. A. M. Raspolli and G. Sbrana, J. Mol. Catal. A: Chem.,
2000, 151, 233243.
100 C. Carlini, M. Giuttari, G. A. M. Raspolli, G. Sbrana,
T. Armaroli and G. Busca, Appl. Catal., A, 1999, 183, 295
302.
101 C. Carlini, P. Patrono, G. A. M. Raspolli and G. Sbrana,
Appl. Catal., A, 2004, 275, 111118.
102 P. Carniti, A. Gervasini, S. Biella and A. Auroux, Catal.
Today, 2006, 118, 373378.
103 M. Watanabe, Y. Aizawa, T. Iida, R. Nishimura and
H. Inomata, Appl. Catal., A, 2005, 295, 150156.
104 F. Benvenuti, C. Carlini, P. Patrono, G. A. M. Raspolli,
G. Sbrana, M. A. Massucci and P. Galli, Appl. Catal., A,
2000, 193, 147153.
105 S. Dutta, S. De, A. K. Patra, M. Sasidharan, A. Bhaumik
and B. Saha, Appl. Catal., A, 2011, 409410, 133139.
Critical Review Green Chemistry
570 | Green Chem., 2014, 16, 548572 This journal is The Royal Society of Chemistry 2014
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
106 X. Qi, M. Watanabe, T. M. Aida and R. L. Smith Jr., Green
Chem., 2008, 10, 799.
107 M. Bicker, J. Hirth and H. Vogel, Green Chem., 2003, 5,
280284.
108 Y. Li, X. Lu, L. Yuan and X. Liu, Biomass Bioenergy, 2009,
33, 11821187.
109 L. Lai and Y. Zhang, ChemSusChem, 2011, 4, 17451748.
110 S. P. Simeonov, J. A. S. Coelho and C. A. M. Afonso, Chem-
SusChem, 2012, 5, 13881391.
111 J. B. Binder and R. T. Raines, J. Am. Chem. Soc., 2009, 131,
19791985.
112 X. Tong and Y. Li, ChemSusChem, 2010, 3, 350355.
113 K. B. Sidhpuria, A. L. Daniel-da-Silva, T. Trindade and
J. A. P. Coutinho, Green Chem., 2011, 13, 340349.
114 B. R. Caes and R. T. Raines, ChemSusChem, 2011, 4, 353
356.
115 X. Tong, M. Li, N. Yan, Y. Ma, P. J. Dyson and Y. Li, Catal.
Today, 2011, 175, 524527.
116 C. Aellig and I. Hermans, ChemSusChem, 2012, 5, 1737
1742.
117 M. H. Tucker, A. J. Crisci, B. N. Wigington, N. Phadke,
R. Alamillo, J. Zhang, S. L. Scott and J. A. Dumesic, ACS
Catal., 2012, 2, 18651876.
118 M. H. Tucker, R. Alamillo, A. J. Crisci, G. M. Gonzalez,
S. L. Scott and J. A. Dumesic, ACS Sustainable Chem. Eng.,
2013, 130326083134008.
119 Z. Huang, W. Pan, H. Zhou, F. Qin, H. Xu and W. Shen,
ChemSusChem, 2013, 6, 10631069.
120 R. Alamillo, A. J. Crisci, J. M. R. Gallo, S. L. Scott and
J. A. Dumesic, Angew. Chem., Int. Ed., 2013, 52, 10349
10351.
121 L. Rigal, A. Gaset and J.-P. Gorrichon, Ind. Eng. Chem.
Prod. Res. Dev., 1981, 20, 719721.
122 L. Rigal, J.-P. Gorrichon, A. Gaset and J.-C. Heughebaert,
Biomass, 1985, 7, 2745.
123 L. Rigal and A. Gaset, Biomass, 1985, 8, 267276.
124 P. Rivalier, J. Duhamet, C. Moreau and R. Durand, Catal.
Today, 1995, 24, 165171.
125 Y. Romn-Leshkov, J. N. Chheda and J. A. Dumesic,
Science, 2006, 312, 19331937.
126 Y. Romn-Leshkov and J. A. Dumesic, Top. Catal., 2009,
52, 297303.
127 J. N. Chheda, Y. Romn-Leshkov and J. A. Dumesic, Green
Chem., 2007, 9, 342350.
128 J. N. Chheda and J. A. Dumesic, Catal. Today, 2007, 123,
5970.
129 J. Y. G. Chan and Y. Zhang, ChemSusChem, 2009, 2, 731
734.
130 A. J. Crisci, M. H. Tucker, J. A. Dumesic and S. L. Scott,
Top. Catal., 2010, 53, 11851192.
131 A. J. Crisci, M. H. Tucker, M.-Y. Lee, S. G. Jang,
J. A. Dumesic and S. L. Scott, ACS Catal., 2011, 1, 719728.
132 F. Yang, Q. Liu, X. Bai and Y. Du, Bioresour. Technol., 2011,
102, 34243429.
133 T. S. Hansen, J. Mielby and A. Riisager, Green Chem.,
2011, 13, 109114.
134 V. V. Ordomsky, d. S. J. van, J. C. Schouten and
T. A. Nijhuis, J. Catal., 2012, 287, 6875.
135 C. Fan, H. Guan, H. Zhang, J. Wang, S. Wang and
X. Wang, Biomass Bioenergy, 2011, 35, 26592665.
136 Q. Zhao, L. Wang, S. Zhao, X. Wang and S. Wang, Fuel,
2011, 90, 22892293.
137 T. Okano, K. Qiao, Q. Bao, D. Tomida, H. Hagiwara and
C. Yokoyama, Appl. Catal., A, 2013, 451, 15.
138 M. Brasholz, K. K. von, C. H. Hornung, S. Saubern and
J. Tsanaktsidis, Green Chem., 2011, 13, 11141117.
139 C. V. McNe, D. T. Nowlan, L. C. McNe, B. Yan and
R. L. Fedie, Appl. Catal., A, 2010, 384, 6569.
140 C. Lansalot-Matras and C. Moreau, Catal. Commun., 2003,
4, 517520.
141 G. Yong, Y. Zhang and J. Y. Ying, Angew. Chem., Int. Ed.,
2008, 47, 93459348.
142 Z. Zhang, Q. Wang, H. Xie, W. Liu and Z. K. Zhao, Chem-
SusChem, 2010, 4, 131138.
143 H. Xie, Z. K. Zhao and Q. Wang, ChemSusChem, 2012, 5,
901905.
144 X. Qi, H. Guo, L. Li and R. L. Smith, ChemSusChem, 2012,
5, 22152220.
145 S. Lima, P. Neves, M. M. Antunes, M. Pillinger, N. Ignatyev
and A. A. Valente, Appl. Catal., A, 2009, 363, 9399.
146 S. Hu, Z. Zhang, Y. Zhou, B. Han, H. Fan, W. Li, J. Song
and Y. Xie, Green Chem., 2008, 10, 12801283.
147 X. Qi, M. Watanabe, T. M. Aida and R. L. Smith Jr., Green
Chem., 2009, 11, 1327.
148 X. Qi, M. Watanabe, T. M. Aida and R. L. Smith Jr., Chem-
SusChem, 2009, 2, 944946.
149 Y. Zhang, V. Degirmenci, C. Li and E. J. M. Hensen, Chem-
SusChem, 2011, 4, 5964.
150 F. Liu, J. Barrault, O. V. K. De and F. Jerome, Chem-
SusChem, 2012, 5, 12231226.
151 F. Ilgen, D. Ott, D. Kralisch, C. Reil, A. Palmberger and
B. Koenig, Green Chem., 2009, 11, 19481954.
152 K. D. O. Vigier, A. Benguerba, J. Barrault and F. Jerome,
Green Chem., 2012, 14, 285289.
153 M. L. Mednick, J. Org. Chem., 1962, 27, 398403.
154 K. Seri, Y. Inoue and H. Ishida, Bull. Chem. Soc. Jpn., 2001,
74, 11451150.
155 H. Ishida and K. Seri, J. Mol. Catal. A: Chem., 1996, 112,
L163L165.
156 X. Qi, M. Watanabe, T. M. Aida and R. L. Smith Jr., Catal.
Commun., 2008, 9, 22442249.
157 K. Nakajima, Y. Baba, R. Noma, M. Kitano, J. N. Kondo,
S. Hayashi and M. Hara, J. Am. Chem. Soc., 2011, 133,
42244227.
158 M. Kitano, K. Nakajima, J. N. Kondo, S. Hayashi and
M. Hara, J. Am. Chem. Soc., 2010, 132, 66226623.
159 H. Yan, Y. Yang, D. Tong, X. Xiang and C. Hu, Catal.
Commun., 2009, 10, 15581563.
160 A. Takagaki, M. Ohara, S. Nishimura and K. Ebitani,
Chem. Commun., 2009, 6276.
161 Y. Yang, C. Hu and M. M. Abu-Omar, Bioresour. Technol.,
2012, 116, 190194.
Green Chemistry Critical Review
This journal is The Royal Society of Chemistry 2014 Green Chem., 2014, 16, 548572 | 571
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online
162 D. D. C. Munoz, M. A. Dam and G. J. M. Gruter, WO,
2011043661A1, 2011.
163 M. Bicker, D. Kaiser, L. Ott and H. Vogel, J. Supercrit.
Fluids, 2005, 36, 118126.
164 A. L. Jongerius, J. R. Copeland, G. S. Foo, J. P. Hofmann,
P. C. A. Bruijnincx, C. Sievers and B. M. Weckhuysen, ACS
Catal., 2013, 3, 464473.
165 R. M. Ravenelle, J. R. Copeland, A. H. Van Pelt,
J. C. Crittenden and C. Sievers, Top. Catal., 2012, 55, 162
174.
166 Y. Roman-Leshkov, J. N. Chheda and J. A. Dumesic,
Science, 2006, 312, 19331937.
167 A. C. Cope, US 2917520, 1959.
168 B. F. M. Kuster and H. J. C. van der Steen, Starch/Starke,
1977, 29, 99103.
169 D. Mercadier, L. Rigal, A. Gaset and J. P. Gorrichon,
J. Chem. Technol. Biotechnol., 1981, 31, 489496.
170 F. Yang, Q. Liu, M. Yue, X. Bai and Y. Du, Chem. Commun.,
2011, 47, 44694471.
171 H. Xiong, T. Wang, B. H. Shanks and A. K. Datye, Catal.
Lett., 2013, 143, 509516.
172 S. Zhao, M. Cheng, J. Li, J. Tian and X. Wang, Chem.
Commun., 2011, 47, 21762178.
173 V. V. Ordomsky, d. S. J. van, J. C. Schouten and
T. A. Nijhuis, ChemSusChem, 2013, 6, 16971707.
174 C. Sievers, I. Musin, T. Marzialetti, O. M. B. Valenzuela,
P. K. Agrawal and C. W. Jones, ChemSusChem, 2009, 2,
665671.
175 C. Li, Z. Zhang and Z. K. Zhao, Tetrahedron Lett., 2009, 50,
54035405.
176 X. Qi, M. Watanabe, T. M. Aida and J. R. L. Smith, Chem-
SusChem, 2010, 3, 10711077.
177 S. Hu, Z. Zhang, J. Song, Y. Zhou and B. Han, Green
Chem., 2009, 11, 1746.
178 E. A. Khokhlova, V. V. Kachala and V. P. Ananikov, Chem-
SusChem, 2012, 5, 783789.
179 D. Liu and E. Y. X. Chen, Appl. Catal., A, 2012,
435.
180 J. He, Y. Zhang and E. Y. X. Chen, ChemSusChem, 2013, 6,
6164.
181 Z. Zhang and Z. Zhao, Bioresour. Technol., 2011, 102,
39703972.
182 H. Zhao, H. M. Brown, J. E. Holladay and Z. C. Zhang,
Top. Catal., 2012, 55, 3337.
183 A. Takagaki, M. Ohara, S. Nishimura and K. Ebitani,
Chem. Commun., 2009, 62766278.
184 S. Despax, B. Estrine, N. Homann, B. J. Le, S. Marinkovic
and J. Muzart, Catal. Commun., 2013, 39, 3538.
185 P. M. Grande, C. Bergs and d. M. P. Dominguez, Chem-
SusChem, 2012, 5, 12031206.
186 R. Huang, W. Qi, R. Su and Z. He, Chem. Commun., 2010,
46, 11151117.
187 T. Stahlberg, J. M. Woodley and A. Riisager, Catal. Sci.
Technol., 2012, 2, 291295.
188 B. Li, P. Relue and S. Varanasi, Green Chem., 2012, 14,
24362444.
189 B. Li, S. Varanasi and P. A. Relue, Green Chem., 2013, 15,
21492157.
190 S. P. Simeonov, J. A. S. Coelho and C. A. M. Afonso, Chem-
SusChem, 2013, 6, 9971000.
191 Y. Roman-Leshkov and M. E. Davis, Acs Catal., 2011, 1,
15661580.
192 M. Moliner, Y. Roman-Leshkov and M. E. Davis, Proc.
Natl. Acad. Sci. U. S. A., 2010, 107, 61646168.
193 S. Saravanamurugan, M. Paniagua, J. A. Melero and
A. Riisager, J. Am. Chem. Soc., 2013, 130402104357006.
194 M. S. Holm, Y. J. Pagan-Torres, S. Saravanamurugan,
A. Riisager, J. A. Dumesic and E. Taarning, Green Chem.,
2012, 14.
195 S. Kobayashi, M. Ueno, S. Saito, Y. Mizuki, H. Ishitani and
Y. Yamashita, Proc. Natl. Acad. Sci. U. S. A., 2004, 101,
54765481.
196 T. Deng, X. Cui, Y. Qi, Y. Wang, X. Hou and Y. Zhu, Chem.
Commun., 2012, 48, 54945496.
197 N. Shi, Q. Liu, Q. Zhang, T. Wang and L. Ma, Green Chem.,
2013, 15, 19671974.
Critical Review Green Chemistry
572 | Green Chem., 2014, 16, 548572 This journal is The Royal Society of Chemistry 2014
P
u
b
l
i
s
h
e
d

o
n

0
5

N
o
v
e
m
b
e
r

2
0
1
3
.

D
o
w
n
l
o
a
d
e
d

b
y

U
n
i
v
e
r
s
i
t
y

o
f

U
t
a
h

o
n

0
7
/
0
5
/
2
0
1
4

1
6
:
1
7
:
4
1
.

View Article Online