Efficient conversion of fructose to 5-hydroxymethylfurfural by functionalized γ-Al2O3 beads

Article abstract of DOI:10.1002/aoc.4821

Millimeter size γ-Al₂O₃ beads were prepared by alginate assisted sol–gel method and grafting organic groups with propyl sulfonic acid and alkyl groups as functionalized γ-Al₂O₃ bead catalysts for fructose dehydr

Full text of DOI:10.1002/aoc.4821

Received: 11 September 2018
DOI: 10.1002/aoc.4821
Revised: 6 January 2019
Accepted: 7 January 2019
F U L L P A P E R
Efficient conversion of fructose to 5‐hydroxymethylfurfural
by functionalized γ‐Al2O3 beads
Fang Lin1
| Kang Wang1
| Lan Gao2 | Xin Guo2
1 Tianjin Key Lab of Membrane Science
and Desalination Technology, Chemical
Engineering Research Center, School of
Chemical Engineering and Technology,
Tianjin University, Tianjin 300350, China
2 Research Institute of Petroleum
Processing, SINOPEC, Beijing 100013,
China
Millimeter size γ‐Al2O3 beads were prepared by alginate assisted sol–gel
method and grafting organic groups with propyl sulfonic acid and alkyl groups
as functionalized γ‐Al2O3 bead catalysts for fructose dehydration to 5‐
hydroxymethylfurfural (5‐HMF). Experiment results showed that the porous
structure of γ‐Al2O3 beads was favorable to the loading and dispersion of active
components, and had an obvious effect on the properties of the catalyst. The
lower calcination temperature of γ‐Al2O3 beads increased the specific surface
area, the hydrophobicity and the activity of catalysts. Competition between
the reaction of alkyl groups and ‐SH groups with surface hydroxyl during the
preparation process of the catalyst influenced greatly the acid site densities,
hydrophobic properties and activity of the catalyst. With an increase in the
alkyl group chain, the hydrophobicity of catalysts increased obviously and
the activity of the catalyst was enhanced. The most hydrophobic catalyst C16‐
SO3H‐γ‐Al2O3–650°C exhibited the highest yield of 5‐HMF (84%) under the fol-
lowing reaction conditions: reaction medium of dimethylsulfoxide/H2O (V/V,
4:1), catalyst amount of 30 mg, temperature of 110°C and reaction time of 4 hr.
Correspondence
Kang Wang, Tianjin Key Lab of
Membrane Science and Desalination
Technology, Chemical Engineering
Research Center, School of Chemical
Engineering and Technology, Tianjin
University, Tianjin, 300350, China.
Email: 18322693646@163.com
Funding information
Natural Science Foundation of Tianjin,
China, Grant/Award Number:
16JCYBJC19600; National Key R&D Pro-
gram of China, Grant/Award Number:
2017YFB0306502
KEYWORDS
5‐hydroxymethylfurfural, calcination temperature, fructose, functionalized γ‐Al2O3 bead, pore
structure
1 | INTRODUCTION
selectivity, and directly decide the prospect of industrial
production.
Among various biomass‐derived valuable chemicals
reported to date, because 5‐hydroxymethylfurfural (5‐
HMF) is easily converted into a variety of high‐value‐
added chemicals, it is regarded as a key platform
compound, such as, 2,5‐diformylfuran (2,5‐DFF),[1,2] 2,5‐
furandicarboxylic acid (2,5‐FDCA),[3,4] C9–C15 alkanes[5]
and others.[6] In addition, levulinic acid, the hydrolysate
of HMF, can be converted into methyl tetrahydrofuran
(MTHF) as a gasoline additive.[7] Therefore, the study of
the transformation of biomass into 5‐HMF has received
considerable attention. However, there were obstacles in
the preparation process of 5‐HMF, including the selection
of catalysts and solvents, which affect its yield and
As reported previously, fructose has been studied
widely as a starting material for the production of HMF
due to its high activity and efficiency towards HMF pro-
duction. In recent years, the dehydration of fructose into
5‐HMF has shown a rapid progression in developing cat-
alytic systems. In this process, many people study homo-
geneous catalytic systems, such as mineral acids,[8] metal
halides[9] or organic acids,[10] which have shown moder-
ate yields (40–60%) of 5‐HMF. At the same time, there
are several serious drawbacks, such as separation of the
products, recycling of catalysts, and the corrosion of
materials, which are not conducive to large‐scale indus-
trial production. Recently, it was reported that EDTA
Appl Organometal Chem. 2019;33:e4821.
wileyonlinelibrary.com/journal/aoc
© 2019 John Wiley & Sons, Ltd.
1 of 13
https://doi.org/10.1002/aoc.4821

2 of 13
LIN ET AL.
was an efficient organic acid catalyst that could improve
HMF yield up to 89 3%, and could be easily recycled
by cooling the reaction mixture to room temperature.
However, further experiments are necessary to investigate
its potential in practical applications.[11] In contrast, het-
erogeneous acid catalysts that avoid the aforementioned
disadvantage have been widely utilized for biomass con-
version into 5‐HMF, such as supported heteropoly
acids,[12,13] metal phosphates,[14] cation exchange
resins,[15,16] sulfated zirconia,[17] Metal Organic Frame-
work (MOFs),[18] acid‐functionalized silicas,[19,20] and C‐
based solid acids.[21–23] They have the advantage of being
easily separated and recyclable, making them more suit-
able for industrial production applications.
Furthermore, both the yield of 5‐HMF and selectivity
depend greatly on the reaction medium. Wang et al. sug-
gested that the strong interaction between solvents and
beta zeolite (Hβ) has a crucial influence on the catalytic
performance, and also suggested an effective approach
to improving the selectivity by regulating the solvents.[24]
It was shown that high yields and selectivity of 5‐HMF
can be obtained in ionic liquids[25] and some organic sol-
vents, such as dimethylsulfoxide (DMSO).[26,27] Indeed,
the use of ionic liquids can not only improve the selectiv-
ity of HMF because of its extraction with 5‐HMF, but also
reaction rates were increased, and levulinic acid and
humin side‐products were not formed.[28] Nonetheless,
from the view of industrial application cost, most ionic
liquids are expensive and unsuitable for mass production.
DMSO has an important positive effect on the reaction
mainly because of the following aspects. First, it facili-
tates the formation of the furanoid form of fructose,
which can be easily dehydrated to HMF. Second, DMSO
can act as both an electron acceptor and an electron
donator to improve the dehydration of the sugars. Third,
DMSO can prevent the formation of byproducts such as
levulinic acid and humins from HMF. Moreover, as a ver-
satile polar aprotic solvent, DMSO can dissolve both polar
and non‐polar compounds, which makes it an effective
solvent for sugars and the products.
that thiol groups promote the isomerization of fructose
to form furanose, and favor the dehydration reac-
tion.[29,30] In the two‐phase catalytic system, the product
5‐HMF was continuously extracted from the aqueous
phase into the organic phase. So the rehydration of 5‐
HMF can hardly proceed because of its separation from
water in a large extent. Lv et al. introduced MIBK into
the reaction system, which can extract HMF into the
MIBK phase, improve HMF selectivity in the reaction,
and suppress the formation and deposition of humins
on the catalyst surface.[31] It can be seen from the process
of solvent use, that blocking the rehydration of 5‐HMF
was an important means to enhance the yield and selec-
tivity, therefore the use of hydrophobic solid catalyst has
been studied. Wang et al. synthesized a superhydrophobic
acid catalyst (P‐SO3H‐154) for fructose dehydration to 5‐
HMF, and the yield of 5‐HMF as the sole product
increased to 99% in the three‐phase solvent system.[32]
They believed that stabilizing 5‐HMF and inhibiting rehy-
dration of 5‐HMF was due to the isolation of the acid sites
in the catalyst from the oxygen atom of water molecules
in the reaction system. Recently, Yang et al. ingeniously
designed a series of hydrophobic silica nanoparticle cata-
lysts, which also successfully restrained the rehydration
of 5‐HMF in the DMSO–water solvent system, thereby
improving the 5‐HMF yield and selectivity.[33]
These catalyst systems can largely improve the 5‐HMF
yield and selectivity, but there are still some problems
such as the recovery, and utilization operation of powder
catalysts is relatively complex. Yao et al. designed a mag-
netically recoverable carbonaceous material catalyst,
which could be recovered by an external magnetic field
after reaction.[34] Nevertheless, there was also a certain
catalyst loss and high cost of industrial production in
the recovery process. It was important to develop a more
efficient catalytic system for the industrial production of
5‐HMF with lower cost and better performance.
In recent years, aluminum oxide (Al2O3) was an
abundant, cheap and commercially available material.
Due to its various forms, good thermal stability, high spe-
cific surface and acid surface, and easy to be functional-
ized by various functional groups, it has been widely
used in the field of catalysis. For example, some
researchers control the acidity of the alumina by using
inorganic acid groups.[35,36] Furthermore, the alumina
has abundant superficial hydroxyl groups, which were
adequate for some functionalization of functional groups
and modifying the acidity.[37] Recently, De la Rosa and
co‐workers synthesized an organic group‐modified alu-
mina by the sol–gel method, and grafted thiol and
sulfonic groups.[38,39] After 24 hr of continuous reaction,
the loss of sulfur content in the catalyst was 15%, and
the fructose conversion yield and selectivity to 5‐HMF
However, aqueous systems are more economically
and environmentally friendly as solvents for 5‐HMF pro-
duction. Unfortunately, water as a reaction medium was
conducive to side‐effects, such as the conversion of
HMF unstable intermediates into polymers, humins and
5‐HMF rapid rehydration into levilininc acid, formic acid,
etc., resulting in low 5‐HMF selectivity and yield in aque-
ous media.[19] So, in order to improve the 5‐HMF yield
and further reduce the cost of application, biphasic
solvent systems appeared. For instance, Román‐Leshkov
et al. studied the effect of a two‐phase reaction solvent
containing primary alcohols, diketones and others on
the dehydration of fructose to HMF. They demonstrated

LIN ET AL.
3 of 13
did not undergo significant changes (less than 5%), which
showed an efficient stability of the catalyst.[39]
were analyzed by SEM. As indicated in Figure 2, the sur-
face and cross‐section morphology of γ‐Al2O3–650°C
beads were loose porous structure, and rich macropore
in micron size and mesoporous structure could be
observed. The macroporous number of SO3H‐γ‐Al2O3–
650°C beads was significantly reduced. After being loaded
with alkylation reagents, the skeleton structure of func-
tionalized γ‐Al2O3 beads became pyknotic. It was demon-
strated that both the surface and the interior of the beads
have been alkylated and sulfonated successfully. When
the calcination temperature of the support increased to
1000°C, the section morphology of C16‐SO3H‐γ‐Al2O3–
1000°C became flatter, which inferred a reduction in spe-
cific surface area.
Herein, in order to overcome the problem of the
recovery of powder catalyst and improve the performance
of the catalyst, millimeter size γ‐Al2O3 beads (1.8 mm)
prepared by the alginate assisted sol–gel method[40,41] were
used as supports and modified with propyl sulfonic acid
and alkyl groups as functionalized γ‐Al2O3 bead catalyst.
The effects of the preparation process, such as alkylation,
sulfonation and calcination temperature of γ‐Al2O3 beads
on the pore structure, acidity, hydrophobicity and activity
of catalyst were investigated by scanning electron micros-
copy (SEM), Brunauer–Emmett–Teller (BET), Fourier
transform‐infrared (FT‐IR) analysis, energy‐dispersive
X‐ray spectroscopy (EDS), acid–base titration[S1], water
contact angles (CAs) and the dehydration of fructose
into 5‐HMF. The optimization of reaction conditions
and the reusability of the catalyst were also presented.
2.1.2 | Pore distribution of catalysts
The pore structure of the catalyst, especially the specific
surface area, is closely related to its activity and selectiv-
ity. Figure 3 showed the effects of modification on the dis-
tribution of the pore size of γ‐Al2O3 beads. It can be seen
from Figure 3 that the pore distribution peak shifts to
small pore after the sulfonation of Al2O3–650°C beads,
which indicates that the acidification process caused a
certain damage to the macropore structure of the alumina
particles and promoted the formation of small pores.
However, when alkylation was carried out prior to
the sulfonation process, the pore distribution peak
of C3‐SO3H‐γ‐Al2O3 moved to large pores, and that of
C8‐SO3H‐γ‐Al2O3 was close to that of γ‐Al2O3 beads.
Furthermore, the pore distribution area of C16‐SO3H‐γ‐
Al2O3 was reduced obviously because of the blockage of
pores by long alkanes. It seems that the alkylation
reaction depresses the destruction of aluminum pore
structure by sulfonation reaction. The catalyst with
larger average pore size was beneficial to the diffusion
of the reactants.
2 | RESULTS AND DISCUSSION
2.1 | Characterization of the
functionalized γ‐Al2O3 beads
2.1.1 | Morphology of catalysts
As shown in Figure 1, the C16‐SO3H‐γ‐Al2O3–650°C were
white, smooth surface, uniform in size with about 1.8
mm, and can be directly used as fillers in industrial pro-
duction, which was in accordance with the appearance
of γ‐Al2O3 beads. Table 1 showed that the crush strength
of the catalyst decreased after alkylation and sulfonation.
However, the crush strength of all catalysts was over 36 N
per bead, which could satisfy the requirement of practical
application.
In order to understand the surface and cross‐section
morphology of functionalized γ‐Al2O3 beads, the samples
Table 1 showed the BET data of samples. It can be
seen clearly that the specific surface area of SO3H‐γ‐
Al2O3–650°C increased and that of C16‐SO3H‐γ‐Al2O3–
650°C decreased due to the blockage of pores by long
alkanes. When the catalyst was only sulfonated, both
the pore volume and the average pore diameter were
reduced, which indicated that the pore structure of the
catalyst was destroyed. However, the longer alkyl chain
grafted to the γ‐Al2O3 beads, the greater steric hindrance
for grafting of organic –SO3H groups, the pore volume
and average pore diameter of the catalyst gradually
increased. This was consistent with the results from
Figure 3 that the alkylation reaction protected from
the damage of aluminum pore structure under strong
acidity treatment.
FIGURE 1 Optical photo of C16‐SO3H‐γ‐Al2O3–650°C

4 of 13
LIN ET AL.
TABLE 1 The crush strength and BET data of samples
Crush strength
(N per bead)
Specific surface
Pore volume
(cm3 g−1
Average pore
diameter (nm)
Sample
area (m2 g−1
)
)
γ‐Al2O3–650°C
56.8
39.4
37.0
36.5
38.4
36.9
68.6
65.5
65.3
46.5
211.5
280.6
259.0
259.1
176.3
170.2
182.8
167.7
137.9
121.5
0.5259
0.4503
0.4679
0.5088
0.5418
0.5048
0.4858
0.4583
0.4743
0.4143
9.947
6.419
SO3H‐γ‐Al2O3–650°C
C3‐SO3H‐γ‐Al2O3–650°C
C8‐SO3H‐γ‐Al2O3–650°C
C16‐SO3H‐γ‐Al2O3–650°C
C16‐SO3H‐γ‐Al2O3–650°C (the fifth cycle)
γ‐Al2O3–850°C
7.226
7.856
8.435
8.246
10.630
10.221
16.090
13.642
C16‐SO3H‐γ‐Al2O3–850°C
γ‐Al2O3–1000°C
C16‐SO3H‐γ‐Al2O3–1000°C
FIGURE 2 (a) Scanning electron microscopy (SEM) images of the surface of different functionalized γ‐Al2O3 beads. (b) SEM images of the
cross‐section of different functionalized γ‐Al2O3 beads
In addition, the calcination temperature obviously
affects the pore structure of supports. The small hole of
alumina supports collapse into the large hole under high‐
temperature calcinations, which lessens the specific sur-
face area of alumina supports and catalysts in Table 1.
groups were incorporated successfully, and –SH groups
were converted to –SO3H groups. As the grafted alkyl
chain length increases, the ‐CH2 vibrational bands
at approximately 2928 and 2846 cm−1 became strong,
and –SO3H groups vibrational bands at 1170 cm−1 and
1098 cm−1 decreased. Yang considered that the phenome-
non was caused by steric hindrance in the structure, the
longer alkyl chain grafted to the γ‐Al2O3 beads, the greater
steric hindrance for grafting of organic –SO3H groups.[33]
This experimental result was also in agreement with
that of pore distribution, i.e. the alkylation reaction
depresses the destruction of aluminum pore structure by
sulfonation reaction.
Moreover, the calcination temperature also influ-
enced the load of functional groups. With the increase
of calcination temperature, the stretching vibration of
CH2 became weaker and that of –SO3H enhanced. It
was well known that the specific surface area and the
active hydroxyl groups of alumina were reduced with an
increase in calcination temperature, which led to a
decrease in the number of cross‐linked trimethoxy and
2.1.3 | Surface functional groups of
catalysts
In order to investigate the surface functional groups of the
as‐prepared catalysts, FT‐IR analysis was carried out. As
shown in Figure 4, for γ‐Al2O3–650°C beads, the absorp-
tion bands at 3456 cm−1 (O‐H stretching vibration), 1650
cm−1 (O‐H bending vibration),[38] 1536 cm−1 and 1434
cm−1 (Al‐O group) were observed. After the addition of
functional groups, IR bands at about 2928 cm−1 and
2846 cm−1 ascribed to the stretching vibration of CH2 in
the hydrocarbon chain,[42] and 1170 cm−1 and 1098 cm−1
attributed to the characteristic adsorption of –SO3H
groups[43] were detected, indicating that surface functional

LIN ET AL.
5 of 13
FIGURE 2 Continued.
Moreover, as listed in Table 2, S and C were detected
on the outer and inner surface of the catalyst, showing
that the functional groups were loaded successfully on
the outer surface and inside of the pore, which was con-
sistent with the experimental results by SEM.
The acid site densities and hydrophobic properties of
catalysts were related to S and C content of catalysts,
respectively. The acid site densities were determined by
acid–base titration using NaOH[S1], as listed in Table 2.
Obviously, the acid site densities agreed well with the S
content of catalysts, and the values of acid site density/S
content on the catalyst cross‐section were in the range
of 29.4–30.5 mmol g−1 (g indicates the quality of the S ele-
ment), which remains approximately constant, indicating
that the acid sites mainly resulted from SO3H groups.
With the increase in alkyl chain length, the acid site den-
sity gradually decreased due to the reduction of S content.
In addition, the increase of calcination temperature for γ‐
Al2O3 beads can raise S content and enhance acid site
density. This result is ascribed to the decrease of C con-
tent. In order to ensure the consistency of the ratio of acid
density and substrate, the mass ratio of functionalized γ‐
Al2O3 beads used in the following fructose dehydration
reactions was adjusted according to their acidity densities.
To study the hydrophobic properties of functionalized
γ‐Al2O3 beads, samples were dispersed in the two‐phase
solvent system containing dichloromethane and water,
respectively. As shown in Figure 5, SO3H‐γ‐Al2O3–650°C
and C3‐SO3H‐γ‐Al2O3–650°C beads stayed in the oil–
water interface and were unable to enter the oil phase,
while C8‐SO3H‐γ‐Al2O3–650°C and C16‐SO3H‐γ‐Al2O3–
FIGURE 3 Distribution of pore size
CH2, then relatively more –SO3H were loaded in the sub-
sequent sulfonation process.
To verify the above results, the element analysis
results of catalysts by EDS were shown in Table 2. With
the increase in carbon chain length of the alkylation
reagent, the content of C increased, but that of S
decreased. After alkylation, steric hindrance of long‐
chain alkanes led to lessen the subsequent ‐SH binding
with supports. In addition, under higher calcination tem-
perature, C content decreased and S content increased, at
the same time the total content of C and S reduced, which
was mainly due to the competition between alkyl chain
and ‐SH groups and the reduction of hydroxyl groups
on alumina. These results were in accordance with those
of FT‐IR analysis.

6 of 13
LIN ET AL.
FIGURE 4 Fourier transform‐infrared
(FT‐IR) spectra of functionalized γ‐Al2O3
beads
TABLE 2 The S and C content, and the surface properties of catalysts
S content (wt%)
Surface/cross‐section
C content (wt%)
Surface/cross‐section
Acid site
density (mmol g−1
Photo of
water CA
Sample
)
CA
SO3H‐γ‐Al2O3–650°C
2.25/1.82
28.2/25.2
0.55
17.1°
C3‐SO3H‐γ‐Al2O3–650°C
C8‐SO3H‐γ‐Al2O3–650°C
1.38/1.16
35.0/32.0
44.0/43.6
0.35
0.32
32.8°
54.9°
1.26/1.05
1.13/1.02
C16‐SO3H‐γ‐Al2O3–650°C
61.6/60.0
61.0/59.6
0.30
0.29
88.4°
83.5°
C16‐SO3H‐γ‐Al2O3–650°C
1.09/0.98
(the fifth cycle)
C16‐SO3H‐γ‐Al2O3–850°C
C16‐SO3H‐γ‐Al2O3–1000°C
CA, contact angle.
1.30/1.07
1.72/1.34
52.9/47.8
48.8/47.9
0.32
0.41
45.3°
38.6°
650°C beads completely sank in the bottom of dichloro-
methane. This phenomenon showed intuitively that the
increase in alkyl chain length greatly enhanced the
hydrophobicity of the catalyst. In order to further quanti-
tatively compare the hydrophobicity of functionalized γ‐
Al2O3 beads, the water droplet CAs were measured. As
shown in Table 2, the water CA for SO3H‐γ‐Al2O3–
650°C in pure water was about 17.1°, and this increased
gradually to 32.8o, 54.9o and 88.4o with the increase in
alkyl chain from C3 to C16, respectively. This could be
attributed to different hydrophobic properties of the alkyl
chain. Because the hydrophobic properties of the catalyst
helped in stabilizing 5‐HMF and inhibiting rehydration of
5‐HMF due to the isolation of the acid sites in the catalyst

LIN ET AL.
7 of 13
FIGURE 5 Macro picture for dispersion of (a) SO3H‐γ‐Al2O3–650 °C, (b) C3‐SO3H‐γ‐Al2O3–650 °C, (c) C8‐SO3H‐γ‐Al2O3–650°C, (d) C16‐
SO3H‐γ‐Al2O3–650°C, (e) C16‐SO3H‐γ‐Al2O3–850°C, (f) C16‐SO3H‐γAl2O3–1000°C and (g) C16‐SO3H‐γ‐Al2O3–650°C (the fifth cycle) in the
two‐phase system of water and dichloromethane
from the oxygen atom of water molecules in the reaction
system, the alkylation of the catalyst would bring benefit
to dehydration of fructose into 5‐HMF.
Al2O3–650°C,
C3‐SO3H‐γ‐Al2O3–650°C,
C8‐SO3H‐γ‐
Al2O3–650°C and C16‐SO3H‐γ‐Al2O3–650°C. In addition,
the γ‐Al2O3 beads were also used as blank experiments.
The effect of the hydrophobicity of functionalized γ‐
Al2O3 beads on the dehydration of fructose to 5‐HMF
was shown in Figure 6. The γ‐Al2O3 beads as catalyst
had a low 5‐HMF yield. But the yields of 5‐HMF obtained
with SO3H‐γ‐Al2O3–650°C, C3‐SO3H‐γ‐Al2O3–650°C, C8‐
SO3H‐γ‐Al2O3–650°C and C16‐SO3H‐γ‐Al2O3–650°C were
52%, 66%, 70% and 84%, respectively, which indicates that
the hydrophobic properties of the functionalized γ‐Al2O3
beads could increase the yield of 5‐HMF, the stronger
the hydrophobicity the higher the yield of 5‐HMF. This
reaction further confirmed that hydrophobic groups in
the γ‐Al2O3 beads can effectively isolate ‐SO3H groups
with H2O molecules, hindering the progression of 5‐
HMF rehydration, which was also consistent with Yang's
research.[33]
In addition, the hydrophobicity of the catalyst
decreased by using high‐temperature calcined γ‐Al2O3
beads as supports. As shown in Figure 5, the higher the
calcination temperature, the less γ‐Al2O3 beads are dis-
persed in the oil phase. It was also found from Table 2
that the water CAs of the catalyst with high‐temperature
calcinations decreased. The main reason was the reduc-
tion of grafting alkyl chain content when γ‐Al2O3 beads
were calcined at high temperature.
The relationship between surface functional group
and pore structure was further discussed. It can be seen
from BET data in Table 1 that the specific surface area
of the catalyst decreases with an increase in the length
of the carbon chain, which should not benefit improving
the hydrophobicity of the catalyst, but the hydrophobicity
of catalyst increased. Thus, the contribution of carbon
chain increase was a critical factor. Of course, the surface
area and pore volume of supports should not be too small
to ensure that the catalyst has enough catalytic surface.
Furthermore, for the catalysts with the same carbon
chain, C16‐SO3H‐γ‐Al2O3 system, the decrease of specific
surface area of catalysts by using a high calcination tem-
perature will reduce the hydrophobicity of the catalyst.
2.2.2 | Effect of the calcination tempera-
ture of γ‐Al2O3 beads on dehydration of
fructose into 5‐HMF
The three functionalized γ‐Al2O3 beads with the different
calcination temperatures were tested in catalytic dehydra-
tion of fructose to 5‐HMF. Figure 7 shows that the 5‐HMF
yields obtained from fructose with C16‐SO3H‐γ‐Al2O3–
650°C, C16‐SO3H‐γ‐Al2O3–850°C and C16‐SO3H‐γ‐Al2O3–
1000°C were 84%, 68% and 63%, respectively, which
shows that the pore structure of functionalized γ‐Al2O3
beads has a great effect on the yields of 5‐HMF. It was
known from the analysis of hydrophobic properties that
the decrease in the specific surface area and the active
hydroxyl groups of alumina would lead to lower hydro-
phobicity of functionalized γ‐Al2O3 beads. Although the
average pore diameter of γ‐Al2O3 beads increased and
the diffusion resistance decreased by using high calcina-
tion temperature, the specific surface area of γ‐Al2O3
2.2 | Catalytic performances
2.2.1 | Effect of alkyl chain length on
dehydration of fructose into 5‐HMF
The four functionalized γ‐Al2O3 beads (SO3H‐γ‐Al2O3–
650°C, C3‐SO3H‐γ‐Al2O3–650°C, C8‐SO3H‐γ‐Al2O3–650°C
and C16‐SO3H‐γ‐Al2O3–650°C) as solid acids were used
to evaluate the fructose dehydration to 5‐HMF. In order
to ensure the consistency of acid density and fructose
ratio in these parallel experiments, the catalyst amount
was 20, 25, 30 and 30 mg, respectively, for SO3H‐γ‐

8 of 13
LIN ET AL.
FIGURE 6 Effect of the catalyst hydrophobicity on the dehydration of fructose to 5‐HMF. Reaction conditions: fructose 90 mg, SO3H‐γ‐
Al2O3–650°C 20 mg, C3‐SO3H‐γ‐Al2O3–650°C 25 mg, C8‐SO3H‐γ‐Al2O3–650°C 30 mg, C16‐SO3H‐γ‐Al2O3–650°C 30 mg, solvents (DMSO/
H2O, V/V = 4:1), volume 1 mL, 110ºC, 4 hr
FIGURE 7 Effects of the pore structure
of C16‐SO3H‐γ‐Al2O3 on the dehydration
of fructose to 5‐HMF. Reaction conditions:
fructose 90 mg, C16‐SO3H‐γ‐Al2O3 beads
30 mg, solvents (DMSO/H2O, V/V = 4:1),
volume 1 mL, 110ºC, 4 hr
beads and hydrophobicity of functionalized γ‐Al2O3 beads
was the key factor affecting the yield of HMF.
medium, reaction time, temperature and catalyst
amount) were investigated. Above all, the influence of
reaction medium has been studied. According to previous
studies,[44,45] DMSO has been proven to be a good solvent
for the production of 5‐HMF and, due to economic and
environmental considerations, the biphasic solvents of
DMSO and H2O with different volume ratios were consid-
ered as reaction medium in this study. At 110°C and 4 hr
reaction time, the effect of reaction medium was shown
2.2.3 | Effect of reaction conditions on
dehydration of fructose into 5‐HMF
In order to be more conducive to industrial production
applications, the optimal reaction conditions (i.e. reaction

LIN ET AL.
9 of 13
FIGURE 8 Effect of reaction medium
on C16‐SO3H‐γ‐Al2O3–650°C catalyzed
synthesis of 5‐HMF from fructose.
Reaction conditions: fructose 90 mg, C16‐
SO3H‐γ‐Al2O3–650°C 30 mg, volume 1
mL, 110°C, 4 hr
in Figure 8. It was found that both fructose conversion
and the yield of 5‐HMF increased significantly with the
increase of DMSO ratio in the solvent system, when
DMSO/H2O (V/V, 4:1) or pure DMSO, the yield of 5‐
HMF reaches a maximum of 84% with fructose conver-
sion of 100%. It was indicated that reaction medium has
an important influence on fructose conversion, especially
5‐HMF yield. It was worth reminding that sulfonic acid
groups play a key role in this process. According to previ-
ous literature, it was reported that the most stable form of
fructose was the β‐D‐fructofuranoses in DMSO solution.
When Bronsted acid was present in the reaction system,
H+ prefers to interact with DMSO other than fructose,
forming [DMSOH]+ as the catalytically active species.
Catalytic fructose tautomerization to β‐D‐furan sugar for
further dehydration produced 5‐HMF.[46] In addition,
when the reaction system proceeded for more than 20
min, the color of the solution began to turn brown, grad-
ually deepening with time, indicating the production of
humins. However, when the reaction was completed, no
insoluble substance was found in the catalytic system.
Upon further detection by high‐performance liquid chro-
matography (HPLC), 5‐HMF was found to be the unique
product with full fructose conversion, which manifested
that soluble humins as byproducts were formed in this
system. There was also no levulinic acid as byproduct,
as was detected in previous studies.[47] The result may
be explained by the fact that DMSO can promote the
dehydration of fructose to a high 5‐HMF yield by main-
taining the stability of 5‐HMF formation. Thus, the vol-
ume ratio of DMSO/H2O (V/V, 4:1) was chosen as
reaction medium.
Catalyst amount was an important parameter for
obtaining the optimal yield of products. Figure 9 shows
FIGURE 9 Effect of catalyst amount on
C16‐SO3H‐γ‐Al2O3–650°C catalyzed
synthesis of 5‐HMF from fructose.
Reaction conditions: fructose 90 mg,
solvents (DMSO/H2O, V/V = 4:1), volume
1 mL, 110°C, 4 hr

10 of 13
LIN ET AL.
that the results were obtained in DMSO/H2O (V/V, 4:1)
at 110°C and 4 hr. In the absence of catalyst, 14% 5‐
HMF yield and 35% fructose conversion were obtained,
respectively. With an increase in the catalyst amount,
the 5‐HMF yield and selectivity and fructose conversion
increased significantly, which was due to the increase in
the availability of catalyst active sites. When the catalyst
amount increased to 30 mg, the 5‐HMF yield reached a
maximum of 84% with full fructose conversion. Neverthe-
less, further increasing the catalyst amount decreased the
5‐HMF yield to 64% (40 mg) and 67% (50 mg), which sug-
gests that superfluous catalyst will lead to the formation
of polymerization and other byproducts. With these
results, we considered the catalyst amount of 30 mg as
optimal for fructose dehydration to 5‐HMF.
summarized in Figure 10. It was found that the yield of
5‐HMF at different temperatures showed a similar trend
with the extension of reaction time, which was that the
5‐HMF yield increased rapidly and then decreased with
time. Because the optimal reaction time was over, there
would be soluble polymer and humin production in the
reaction. Moreover, different reaction temperatures and
reaction times correspond to different maximum 5‐HMF
yields. The yield of 5‐HMF reached a maximum of 84%
in all catalytic systems at 110°C for a reaction time of 4
hr. However, the 5‐HMF yield decreased gradually when
the temperature was further elevated to the higher range.
It was probable that the increase in reaction temperature
favors fructose dehydration, but it also increases the gen-
eration of soluble polymers and humin, thereby covering
the effective active sites of the catalyst resulting in a
decrease in the area of contact with the substrate.
Amarasekara et al. also reported similar results for the
decomposition of 5‐HMF to unknown byproducts at high
temperatures.[48]
Compared with Yang's research,[33] the optimal
amount of catalyst was reduced in this catalytic system.
The ratio of fructose to catalyst was 50 mg:28 mg in Yang's
study, while our ratio of fructose to catalyst was 90 mg:30
mg. This was mainly related to the properties of alumina
supports. On the one hand, the acid of the alumina beads
was stronger than that of nano‐silica, and the former had
stronger catalytic capacity. A 5‐HMF yield of about 10%
and fructose conversion of about 20% were obtained in
Yang's research with bare silica as catalyst. However, there
was a 5‐HMF yield of 30% and fructose conversion of 84%
using the pure alumina beads as catalyst in this paper. On
the other hand, the alumina catalyst had a high specific
surface area and abundant porous structure, which was
conducive to the loading and dispersion of active compo-
nents and was beneficial to the catalytic reaction.
In summary, we obtained the highest 5‐HMF yield of
84% with C16‐SO3H‐γ‐Al2O3–650°C under the following
reaction conditions: reaction medium of DMSO/H2O (V/
V, 4:1), catalyst amount of 30 mg, temperature of 110°C,
and reaction time of 4 hr.
2.3 | The reusability and stability of the
catalyst
The reusability and stability of the catalyst are important
parameters to be considered when using heterogeneous
catalysts in dehydration of fructose to 5‐HMF in indus-
trial applications. In Figure 11, the C16‐SO3H‐γ‐Al2O3–
650°C was recovered by filtering, treating with water
and dried in an oven overnight at 65°C after each run,
and the regeneration experiments were obtained under
the optimized reaction conditions. Figure 11 showed
that the reused catalyst, C16‐SO3H‐γ‐Al2O3–650°C, has
a slight decrease in catalytic activity, but still 79% of
5‐HMF yield was maintained after five runs. This
decrease was attributed to the formation of residuals
such as humins or some adsorbed intermediates, which
were deposited onto the catalyst surface, thereby clog-
ging the active sites for reaction. Moreover, the color
of the catalyst turns brown after five runs (Figure 6),
further indicating that the humin has been chemisorbed
to the catalyst surface, which will also decrease slightly
the surface area and pore volume of the catalyst
(Table 1). In addition, compared with the fresh C16‐
SO3H‐γ‐Al2O3–650°C, the reduction of acid site density,
decrease of C and S elemental content of the catalyst
(Table 2) as well as weakness of the FT‐IR bands at
Reaction time and temperature played a vital role in
the fructose dehydration to 5‐HMF, and the results were
FIGURE 10 Effects of reaction temperature and time on C16‐
SO3H‐γ‐Al2O3–650°C catalyzed synthesis of 5‐HMF from fructose.
Reaction conditions: fructose 90 mg, C16‐SO3H‐γ‐Al2O3–650°C 30
mg, solvents (DMSO/H2O, V/V = 4:1), volume 1 mL

LIN ET AL.
11 of 13
hand, the acid site densities and hydrophobic properties
of catalysts were affected by this competition. On the
other hand, the effect of competition between S and C
on the pore structure of catalysts also can be found by
pore distribution of catalysts. The macroporous structure
of the catalyst was damaged by the sulfonation process
and can be protected by the alkylation process. Moreover,
a less optimal amount of catalyst, a ratio of fructose to cat-
alyst of 90 mg:30 mg, and a lower optimal reaction tem-
perature of 110°C were presented in this paper. The
highest yield of 5‐HMF (84%) was obtained with the most
hydrophobic catalyst C16‐SO3H‐γ‐Al2O3–650°C in reaction
medium DMSO/H2O (V/V, 4:1). Finally, the synthesized
C16‐SO3H‐γ‐Al2O3–650°C had certain reusability and
stability for the dehydration of fructose to 5‐HMF, the 5‐
HMF yield of C16‐SO3H‐γ‐Al2O3–650°C after the fifth
cycle was still 79%. There is industry potential for func-
tional alumina as a catalyst for dehydration of fructose to
5‐HMF.
FIGURE 11 Recycling experiments of C16‐SO3H‐γ‐Al2O3–650°C
for the dehydration of fructose to 5‐HMF. Reaction conditions:
fructose 90 mg, C16‐SO3H‐γ‐Al2O3–650°C 30 mg, solvents (DMSO/
H2O, V/V = 4:1), volume 1 mL, 110ºC, 4 hr
1170 cm−1 and 1098 cm−1 (Figure 4) after five repeated
catalytic cycles revealed a slight loss in the active acid
sites, which can lead to a decrease of catalytic perfor-
mance. Although the water CA also dropped from
88.4° to 83.5°, most of the C16‐SO3H‐γ‐Al2O3–650°C
beads used five times can still sink in the oil phase
(Figure 5), suggesting that the hydrophobicity of C16‐
SO3H‐γ‐Al2O3–650°C was well maintained. It was worth
noting that the crush strength of C16‐SO3H‐γ‐Al2O3–
650°C beads after the fifth cycle was still 36.9 N per
bead, which could meet the requirement of industrial
applications. Therefore, the C16‐SO3H‐γ‐Al2O3–650°C
beads have excellent recyclability, and have great poten-
tial to be applied in industry.
4 | EXPERIMENTAL
4.1 | Materials and agents
In this work, the pseudo‐boehmite (PB) powder was pur-
chased from Tianjin Chemical Research and Design Insti-
tute. The ammonium alginate (ALG) was provided by
Bright Moon Seaweed Group Company. Fructose, (3‐
mercaptopropyl) trimethoxysilane [(CH3O)3Si(CH2)3SH,
95%], 5‐hydroxymethylfurfural, trimethoxy (hexadecyl)
silane [(CH3O)3Si(CH2)15CH3, 90%], trimethoxy (octyl)
silane [(CH3O)3Si(CH2)7CH3, 96%] and trimethoxy (pro-
pyl)silane [(CH3O)3Si(CH2)2CH3, 97%] were ordered by
McLean Reagent. The other agents (analytical grade)
were ordered by Guangfu Fine Chemical Industry
Research Institute (Tianjin, China).
3 | CONCLUSIONS
In conclusion, we successfully synthesized functional-
ized γ‐Al2O3 beads with size of about 1.8 mm and crush
strength over 36 N per bead. The porous structure of γ‐
Al2O3 beads facilitated the dispersion of functional
groups. The loading of alkyl and sulfonic groups on
the surface and internal pores of catalyst can be found
by analyzing catalyst performance. Therefore, the pore
structure of the supports would significantly affect the
performance of the catalyst. The increase in the calcina-
tion temperature would reduce the specific surface area
of supports, which cut down the hydrophobicity and
activity of catalysts. Furthermore, it was shown from
FT‐IR and EDS analyses that competition existed
between the loading of ‐CH2 and ‐SH groups due to
the space hindrance of the long carbon chain. On one
4.2 | Preparation of alumina beads
Alumina beads were prepared by previous work.[41] Then,
20 mL ALG solution (1.5 wt%) and 4 g PB powder were
added into 20 mL deionized water under stirring at high
speed for 15 min to form the homogeneous PB–ALG sus-
pension, which was evenly poured into a 0.5 M Ca (NO3)2
solution. Due to the ionized gelation of ALG with metal
Ca2+
, the semi‐rigid alumina beads were formed
instantly. These beads were treated with filtration, wash-
ing and acetic acid solution (0.5 M), and were finally dried
at 65°C overnight and calcined at different temperature
(650°C, 800°C and 1000°C) in a muffle furnace for 4 hr,
and samples of γ‐Al2O3‐T were produced (T referred to
calcination temperature).

12 of 13
LIN ET AL.
4.3 | Preparation of the functionalized
4.5 | Catalytic activity test
alumina beads
In this process, the reactant fructose (90 mg), reaction
medium (water and DMSO with different volume ratios)
and catalyst (0–50 mg) were successively added to a 15‐
mL glass reactor with a screw top of Teflon, and stirred
and heated at a certain stated temperature. Timing was
started when the oil bath temperature reached the set tem-
perature. After the reaction was completed, the reactor
was cooled down to room temperature in flowing cold
water, and then the sample was diluted to 50 mL with
ultrapure water. Then, 1 mL of the solution was taken
out with a syringe, filtered through a 0.22‐μm syringe fil-
ter, and the resulting sample was subjected to HPLC anal-
ysis using a refractive index detector. Using 0.004 M H2SO4
solution as the mobile phase, samples were separated
using an Aminex HPX‐87H 300 mm × 7.8 mm column,
and the column temperature was designed at 65°C. Identi-
fication of the compounds was calibrated with a reference
sample. Fructose conversion, 5‐HMF yield and selectivity
were carbon base evaluated as follows:
The synthetic method of functionalized alumina beads
was on the basis of previous research.[33] A mixture of
14 mL ultrapure water, 100 mL absolute ethanol and 8.1
mL ammonium hydroxide was stirred at 40°C for 10 min.
Then, the 2.24 g γ‐Al2O3–650°C was added into the
mixture. After 6 min, the prepared organosilane [i.e.
(CH3O)3Si(CH2)15CH3, (CH3O)3Si(CH2)7CH3, (CH3O)3Si(-
CH2)2CH3] and (CH3O)3Si(CH2)3SH were added to the
mixture at intervals of 5 min. The molar ratio of
[(CH3O)3Si(CH2)3SH], organosilane and alumina beads
was 1:1:4. After 24 hr, the alumina beads were filtered
and washed by ultrapure water, and filtered again until
the solution became neutral. Then, H2O2 (30 wt%) solu-
tion was added to oxidize the thiol (‐SH) groups to sulfonic
acid groups (–SO3X, X = NH4+) at 25°C for 8 hr, followed
by acidification with H2SO4 solution (0.8 M) at 25°C for 12
hr. The target alumina beads were obtained by filtering
and washing till the solution became neutral, and were
then dried at 80°C for 24 hr. Four functionalized alumina
beads with sulfonic groups and different alkyls (i.e. C0,
C3, C8, C16) were prepared, marked as SO3H‐γ‐Al2O3–
650°C, C3‐SO3H‐γ‐Al2O3–650°C, C8‐SO3H‐γ‐Al2O3–650°C
and C16‐SO3H‐γ‐Al2O3–650°C, respectively. In addition,
the C16‐SO3H‐γ‐Al2O3–850°C and C16‐SO3H‐γ‐Al2O3–
1000°C are also synthesized.
Fructose conversion (mol%) = (1 − fructose concen-
tration in product/fructose concentration in the sample);
5‐HMF yield (mol%) = moles of carbon in 5‐HMF/
moles of carbon loaded in fructose × 100%;
5‐HMF selectivity (mol%) = 5‐HMF yield/fructose
conversion × 100%.
ACKNOWLEDGEMENTS
The authors thank the National Key R&D Program of
China (No. 2017YFB0306502) and the Natural Science
Foundation of Tianjin, China (No. 16JCYBJC19600) for
financial support.
4.4 | Catalyst characterizations
Crush strength was measured by the particle strength
tester. The morphology of beads and C and S contents
of the alumina beads were recorded by Hitachi S‐4800
scanning electron microscope at 20 kV. Grinding the
functional γ‐Al2O3 beads into powders and pressing into
sheets, the water CAs of the sheet samples were analyzed
on an Optical Contact Angle Measuring Device
(Dataphysics, OCA15EC, Germany). Specific surface
areas and pore diameters were obtained from low‐
temperature N2 adsorption–desorption instrument
(Quantachrome Autosorb instrument from America).
The samples were outgassed for 8 hr under vacuum at
250°C prior to adsorption. The nitrogen adsorption and
desorption curves were obtained at −196°C liquid nitro-
gen. The surface area was determined by the multipoint
BET method, and the pore volume and pore size of the
samples were calculated by the BJH method. FT‐IR
spectroscopy analysis was achieved with a spectropho-
tometer (BIO‐Rad, Japan) with the wavenumber range
ORCID
Fang Lin https://orcid.org/0000-0002-7929-0527
Kang Wang https://orcid.org/0000-0002-1451-2360
REFERENCES
[1] X. Liu, J. Xiao, H. Ding, W. Zhong, Q. Xu, S. Su, D. Yin, Chem.
Eng. J. 2016, 283, 1315.
[2] I. Sádaba, Y. Y. Gorbanev, S. Kegnæs, S. S. R. Putluru, R. W.
Berg, A. Riisager, ChemCatChem 2013, 5, 284.
[3] N. K. Gupta, S. Nishimura, A. Takagaki, K. Ebitani, Green
Chem. 2011, 13, 824.
[4] L. Gao, K. Deng, J. Zheng, B. Liu, Z. Zhang, Chem. Eng. J. 2015,
270, 444.
[5] G. W. Huber, J. N. Chheda, C. J. Barrett, J. A. Dumesic, Science
2005, 308, 1446.
of 500–4000 cm−1 at a spectral resolution of 4 cm−1
.

LIN ET AL.
13 of 13
[6] S. Yao, X. Wang, Y. Jiang, F. Wu, X. Chen, X. Mu, ACS Sustain-
able Chem. Eng. 2014, 2, 173.
[33] Z. Yang, W. Qi, R. Huang, J. Fang, R. Su, Z. He, Chem. Eng. J.
2016, 296, 209.
[7] J. Zhang, E. Weitz, ACS Catal. 2012, 2, 1211.
[34] Y. Yao, Z. Gu, Y. Wang, H. J. Wang, W. Li, Russ. J. Gen. Chem.
2016, 86, 1698.
[8] Y. Yang, C. W. Hu, M. M. Abu‐Omar, Mol. Catal. A Chem.
2013, 376, 98.
[35] J. S. Valente, S. Falcón, E. Lima, M. A. Vera, P. Bosch, E. López
Salinas, Micropor. Mesopor. Mater. 2006, 94, 277.
[9] N. Jiang, R. Huang, W. Qi, R. X. Su, Z. M. He, Bio Energy Res.
2012, 5, 380.
[36] J. S. Valente, E. López Salinas, X. Bokhimi, J. Flores, A. M.
Maubert, E. Lima, Phys. Chem. C. 2009, 113, 16 476.
[10] H. Zhao, J. E. Holladay, H. Brown, Z. C. Zhang, Science 2007,
316, 1597.
[37] V. Smuleac, D. A. Butterfield, S. K. Sikdar, R. S. Varma, D.
Bhattacharyya, Membr. Sci. 2005, 251, 169.
[11] Z. Cao, M. Li, Y. Chen, T. Shen, C. L. Tang, C. J. Zhu, H. J.
Ying, Appl. Catal. A 2019, 569, 93.
[38] C. S. Maldonado, J. R. De la Rosa, C. J. Lucio‐Ortiz, J. S.
Valente, M. J. Castaldi, Fuel 2017, 198, 134.
[12] A. H. Jadhav, H. Kim, I. T. Hwang, Bioresour. Technol. 2013,
132, 342.
[39] F. J. Morales‐Leal, J. R. Rosa, C. J. Lucio‐Ortiz, D. A. De Haro‐
Del Rio, C. S. Maldonado, S. Wi, L. B. Casabianca, C. D. Garcia,
Appl. Catal. B 2019, 244, 250.
[13] Y. Qu, C. Huang, J. Zhang, B. Chen, Bioresour. Technol. 2012,
106, 170.
[40] K. Wang, G. Y. Meng, W. J. Yang, H. B. Yu, Y. M. Sun, X. Y. Li,
[14] P. Khemthong, P. Daorattanachai, N. Laosiripojana, K.
Tianjin Univ. (Sci. Technol.) 2014, 47, 1052. (in Chinese)
Faungnawakij, Cat. Com. 2012, 29, 96.
[41] A. H. Dong, K. Wang, S. Z. Zhu, G. B. Yang, X. T. Wang, Fuel
Process. Technol. 2017, 158, 218.
[15] X. H. Qi, M. Watanabe, T. M. Aida, R. L. Smith Jr., Green
Chem. 2008, 10, 799.
[42] D. Yang, B. Paul, W. Xu, Y. Yuan, E. Liu, X. Ke, Water Res.
2010, 44, 741.
[16] X. H. Qi, M. Watanabe, T. M. Aida, R. L. Smith Jr., Ind. Eng.
Chem. Res. 2008, 47, 9234.
[43] R. J. Kalbasi, A. R. Massah, B. Daneshvarnejad, Appl. Clay Sci.
2012, 55, 1.
[17] X. H. Qi, M. Watanabe, T. M. Aida, R. L. Smith Jr., Cat. Com.
2009, 10, 1771.
[44] J. Zhang, A. Das, R. S. Assary, L. A. Curtiss, E. Weitz, Appl.
Catal. B 2016, 181, 874.
[18] J. Chen, K. Li, L. Chen, R. Liu, X. Huang, D. Ye, Green Chem.
2014, 16, 2490.
[45] J. Wang, W. Xu, J. Ren, X. Liu, G. Lu, Y. Wang, Green Chem.
2011, 13, 2678.
[19] C. Tian, C. Bao, A. Binder, Z. Zhu, B. Hu, Y. Guo, B. Zhao, S.
Dai, Chem. Commun. 2013, 49, 8668.
[46] L. K. Ren, L. F. Zhu, T. Qi, J. Q. Tang, H. Q. Yang, C. W. Hu,
[20] X. Guo, Q. Cao, Y. Jiang, J. Guan, X. Wang, X. Mu, Carbohydr.
Res. 2012, 351, 35.
ACS Catal. 2017, 7, 2199.
[47] Y. Shen, J. K. Sun, Y. X. Yi, B. Wang, F. Xu, R. C. Sun, J. Mol.
Catal. A Chem. 2014, 394, 114.
[21] X. Qi, H. Guo, L. Li, R. L. Smith, ChemSusChem 2012, 5, 2215.
[22] H. Wang, Q. Kong, Y. Wang, T. Deng, C. Chen, X. Hou, Y. Zhu,
[48] A. S. Amarasekara, L. D. Williams, C. C. Ebede, Carbohydr.
Res. 2008, 343, 3021.
ChemCatChem 2014, 6, 728.
[23] L. Hu, X. Tang, Z. Wu, L. Lin, J. X. Xu, N. Xu, B. L. Dai, Chem.
Eng. J. 2015, 263, 299.
[24] Y. Q. Wang, G. Q. Ding, X. H. Yang, H. Y. Zheng, Y. L. Zhu, Y.
SUPPORTING INFORMATION
W. Li, Appl. Catal. B 2018, 235, 150.
Additional supporting information may be found online
in the Supporting Information section at the end of the
article.
[25] Y. Li, H. Liu, C. Song, X. Gu, H. Li, W. Zhu, S. Yin, C. Han,
Bioresour. Technol. 2013, 133, 347.
[26] D. A. Kotadia, S. S. Soni, Cat. Sci. Technol. 2013, 3, 469.
[27] X. Tong, Y. Li, ChemSusChem 2010, 3, 350.
[28] C. Lansalot‐Matras, C. Moreau, Cat. Com. 2003, 4, 517.
[29] Y. Román‐Leshkov, J. N. Chheda, J. A. Dumesic, Science 2006,
312, 1933.
How to cite this article: Lin F, Wang K, Gao L,
Guo X. Efficient conversion of fructose to 5‐
hydroxymethylfurfural by functionalized γ‐Al2O3
beads. Appl Organometal Chem. 2019;33:e4821.
https://doi.org/10.1002/aoc.4821
[30] Y. Román‐Leshkov, J. A. Dumesic, Top. Catal. 2009, 52, 297.
[31] G. Q. Lv, L. L. Deng, B. Q. Lu, J. L. Li, X. L. Hou, Y. X. Yang,
J. Clean. Prod. 2017, 142, 2244.
[32] L. Wang, H. Wang, F. Liu, A. Zheng, J. Zhang, Q. Sun, J. P. Lewis,
L. Zhu, X. Meng, F. S. Xiao, ChemSusChem 2014, 7, 402.