Integrated, cascading enzyme-/chemocatalytic cellulose conversion using catalysts based on mesoporous silica nanoparticles

Article abstract of DOI:10.1002/cssc.201402605

This article reports a novel approach to deconstructing cellulose into 5-hydroxymethylfurfural (HMF) with a high yield (46.1%) by integrating a sequential enzyme cascade technique in an aqueous system with solid acid catalysis in an organic-solvent system. We executed the rational design and synthesis of mesoporous silica nanoparticles (MSNs) with various pore sizes and surface functionalities, which proved to be useful for the immobilization of various enzymes (i.e., cellulase and isomerase) and nanoparticles (i.e., magnetic Fe₃O₄) and for functionalization of various acid groups (i.e., H₂PO₃, COOH, and SO₃H). We separately applied the synthesized biocatalysts (i.e., cellulase-Fe₃O₄@MSN and isomerase-Fe₃O₄@MSN) and chemical catalysts (i.e., HSO₃-MSN) in a sequential cellulose-to-glucose, glucose-to-fructose, and fructose-to-HMF conversion, respectively, across both aqueous- and organic-solvent systems after the optimization of reaction conditions (e.g., reaction temperature, water ratio, catalyst amount). The integrated enzymatic and chemocatalytic concept in this study could be an effective and economically friendly process for various catalytic applications.

Full text of DOI:10.1002/cssc.201402605

CHEMSUSCHEM
COMMUNICATIONS
DOI: 10.1002/cssc.201402605
Integrated, Cascading Enzyme-/Chemocatalytic Cellulose
Conversion using Catalysts based on Mesoporous Silica
Nanoparticles
Yi-Chun Lee, Saikat Dutta, and Kevin C.-W. Wu*[a]
This article reports a novel approach to deconstructing cellu-
lose into 5-hydroxymethylfurfural (HMF) with a high yield
(46.1%) by integrating a sequential enzyme cascade technique
in an aqueous system with solid acid catalysis in an organic-
solvent system. We executed the rational design and synthesis
of mesoporous silica nanoparticles (MSNs) with various pore
sizes and surface functionalities, which proved to be useful for
the immobilization of various enzymes (i.e., cellulase and iso-
merase) and nanoparticles (i.e., magnetic Fe3O4) and for func-
tionalization of various acid groups (i.e., H2PO3, COOH, and
SO3H). We separately applied the synthesized biocatalysts (i.e.,
cellulase-Fe3O4@MSN and isomerase-Fe3O4@MSN) and chemical
catalysts (i.e., HSO3-MSN) in a sequential cellulose-to-glucose,
glucose-to-fructose, and fructose-to-HMF conversion, respec-
tively, across both aqueous- and organic-solvent systems after
the optimization of reaction conditions (e.g., reaction tempera-
ture, water ratio, catalyst amount). The integrated enzymatic
and chemocatalytic concept in this study could be an effective
and economically friendly process for various catalytic applica-
tions.
plentiful renewable carbon. Enzyme-based biocatalysis is con-
sidered an alternative to the use of chemicals because
enzyme-based biocatalysis provides higher product selectivity
and can be triggered under milder reaction conditions.[3] The
consolidation of cascading enzymatic reactions in a single
vessel has numerous benefits, such as decreased unit opera-
tions, decreased reactor volume, increased volumetric and
space-time yields, and shortened cycle times. The major ad-
vantage of the cascading strategy is that the coupling of steps
forces even unfavorable equilibria towards the formation of
desired products .[4] However, the major problems of cascading
enzymatic reactions involve the stability and recyclability of
the enzymes, particularly because the different enzymes are
used under different reaction conditions. Therefore, researchers
have preferred a method involving the preparation of magnet-
ic and porous solid particles that can prevent enzymes from
denaturing and allow easy recyclability.[4]
We and other researchers have used mesoporous silica ma-
terials with large surface areas, adjustable pore sizes, and im-
proved thermal stability as fillers,[5] and diverse surface func-
tionalities to immobilize enzymes.[6] The encapsulated enzymes
maintain their efficacy and show increased stability and recy-
clability. In particular, for cellulosic biomass conversion, we
demonstrated that cellulase and glucose isomerase can be
separately immobilized into iron oxide-encapsulated mesopo-
rous silica nanoparticles (i.e., cellulase-Fe3O4@MSN and isomer-
ase-Fe3O4@MSN) for cellulose-to-glucose and glucose-to-fruc-
tose conversion sequences, respectively.[6d] The main advant-
age of such cascading enzymatic reactions is that the enzyme-
immobilized Fe3O4@MSN material can be separated easily after
each reaction, by using a magnet. In this manner, successful
cascading of enzymatic reactions with a maximum fructose
yield of 51% can be achieved under optimized reaction condi-
tions, including buffer composition, reaction temperature time,
and pH values. From a biofuel production perspective, scien-
tists have considered further conversion of fructose to 5-hy-
droxymethylfurfural (HMF), a platform chemical, to be scientifi-
cally valuable but technologically challenging.[7] Therefore, we
attempted to solve this problem by integrating the enzyme
cascade sequence in water systems with a chemical dehydra-
tion process for fructose-to-HMF conversion in an organic sol-
vent system containing a series of functionalized MSN-based
catalysts (Scheme 1).
Since researchers discovered the
value of biomass as feedstock for
the production of fuels, building-
block chemicals, and advanced
materials, a considerable amount
of research has been conducted
on the transformation of lignocel-
lulosic biomass into commodity
chemicals and liquid fuels, using
numerous techniques, to reduce
the dependence of the economy
on petrochemicals.[1] The physico-
chemical recalcitrance of cellulose
COVER
limits its rapid and cost-effective
degradation.[2]
From the perspective of plant cell wall degradation, cellulo-
lytic and hemicellulolytic enzymes that can deconstruct cellu-
lose into fermentable sugars can facilitate the utilization of
[a] Y.-C. Lee, Dr. S. Dutta, Prof. Dr. K. C.-W. Wu
Department of Chemical Engineering
National Taiwan University
Herein, we set out to achieve an effective cellulose-to-glu-
cose-to-fructose-to-HMF conversion sequence by integrating
a water (i.e., enzyme) system with an organic solvent (i.e.,
chemical) system. For water systems, we prepared enzyme-im-
No. 1, Sec. 4, Roosevelt Road, Taipei 10617 (Taiwan)
E-mail: kevinwu@ntu.edu.tw
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cssc.201402605.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 3241 – 3246 3241

CHEMSUSCHEM
COMMUNICATIONS
www.chemsuschem.org
50 mg of Fe3O4@MSN, the maximum amount of im-
mobilized enzyme was 7.3 mg (36.5 units) for cellu-
lase and 0.65 mg (19.1 units) for isomerase.
To prepare sulfonic acid-functionalized MSNs, we
first synthesized MSNs with a small particle size (ap-
proximately 276 nm) and pore size (4.7 nm), and
grafted the thiol-containing organosilane onto the
MSNs. The thiol groups were then converted to sul-
fonic-acid groups through oxidation in the presence
of H2O2. We determined the acid strength (0.8–2.0)
and amount (1.29 mmol(H+)gÀ1) of the final
HSO3@MSN material by using based an indicator
Scheme 1. An integrated enzyme cascade-chemocatalytic conversion of cellulose oligo-
mers into HMF in aqueous (enzyme) and organic (chemical) media with enzyme and acid
functionalized mesoporous silica nanoparticles, respectively.
mobilized Fe3O4@MSN materials to use as biocatalysts, and for
chemical systems, we synthesized sulfonic-acid-functionalized
MSN materials for use as catalysts. Only fructose-to-HMF con-
versions were run in organic solvent, because enzymatic con-
versions cannot be performed in such media. Therefore, in ad-
dition to the preparation of various MSN-based catalysts, opti-
mization of reaction conditions such as the volume ratio of or-
ganic to aqueous phase and reaction temperature were critical
for maximizing the final yield of HMF.
(phenol red) and titration, respectively.
The enzyme-chemocatalytic cascading of the cellulose-to-
HMF conversion involves three steps: (1) cellulose-to-glucose
conversion (catalyst: cellulase-Fe3O4@MSN), (2) glucose-to-fruc-
tose conversion (catalyst: isomerase-Fe3O4@MSN), and (3) fruc-
tose-to-HMF conversion (catalyst: HSO3-MSN). We previously
optimized the reaction conditions for the sequential reaction
of the first two steps, and achieved a maximum yield of fruc-
tose (51%) under the following conditions: 508C, 24 h, and
a phosphate buffer with pH 4.8 for cellulose-to-glucose con-
version; and 708C, 24 h, and a phosphate buffer with pH 7.5
for glucose-to-fructose conversion.[7d] The most difficult task is
to continue the sequence from the first two steps in water-
based systems to the third step in an organic solvent (i.e.,
DMSO)-based system. To solve this problem, we optimized
three parameters: (1) reaction temperature and time, (2) the
volume ratio of DMSO to water, and (3) the amount of HSO3-
MSN catalyst.
We modified a synthesis process of Fe3O4 nanoparticles and
MSNs presented in a previous report,[7d] and present details of
the synthesis and functionalization processes using enzymes
and acids in the Experimental Section. We also characterized
the enzymes (cellulose, isomerase)-immobilized, Fe3O4-loaded
MSN and sulfonic-acid-functionalized MSN (HSO3-MSN) by
using scanning electron microscopy (SEM), N2 sorption iso-
therms, solid-state NMR, and UV-Vis spectroscopy. Table 1 sum-
marizes the properties of the prepared MSN catalysts, includ-
ing particle size, surface area and pore size, acidity, and acid
amount.
Figure 1(a) indicates that temperature notably influences
the yield of HMF. As the reaction temperature increased, the
HMF yield also increased both with and without the HSO3-MSN
catalyst. The fructose-to-HMF conversion is a dehydration reac-
tion, therefore a higher reaction temperature is preferred, and
numerous groups have used the same organic solvent (i.e.,
DMSO) at high temperatures to produce HMF.[8] Apparently,
the effect of the catalyst disappeares at a high temperature
(1208C), possibly due to the electrophilic nature of DMSO at
higher temperature. This facilitates formation of the a-furanose
anomeric form of d-fructose at the expense of the b-pyranose
form via the formation of a dihydrofuran-carbaldehyde inter-
mediate, as reported based on direct spectroscopic evi-
dence.[8d]
We first synthesized magnetic Fe3O4 nanoparticles with parti-
cle sizes of approximately 20 nm and characterized these
nanoparticles by transmission electron microscopy (TEM),
powder X-ray diffraction (XRD), and superconducting quantum
interference device (SQUID) measurements (Supporting Infor-
mation, Figures S1–S3). We then added the synthesized Fe3O4
nanoparticles to the precursor used for MSN synthesis (Sup-
porting Information, Figure S4). Subsequently, we used the
synthesized Fe3O4-loaded MSN (Fe3O4@MSN) samples as hosts
for the immobilization of enzymes. We maintained the particle
sizes of the Fe3O4@MSN between 500 and 600 nm and the
pore size of the Fe3O4@MSN at approximately 20 nm to effi-
ciently immobilize cellulase (hydrodynamic diameter ca. 8 nm)
and glucose isomerase (hydrodynamic diameter ca. 3 nm). We
determined the amount of the immobilized enzyme by using
UV-Vis spectroscopy, measuring absorbance at 280 nm. For
An exception occurred at a temperature of 1208C in the
presence of the catalyst, which implies that such a high tem-
perature produces intractable side products that decrease the
HMF yield. At room temperature, an HMF yield of approxi-
mately 37% was achieved in the
presence of the HSO3-MSN cata-
lyst, in contrast to the 0% HMF
Table 1. Characterization of enzyme- and acid-functionalized MSN-based catalysts.
yield without catalyst. This result
Catalyst
Particle
Specific surface Pore
Amount of
Acidity[f] Amount of acid[g]
clearly demonstrates the efficacy
size[a] [nm] area[b] [m2 gÀ1
]
size[c] [nm] enzyme[b] [mg] [pKa]
[mmol(H+)gÀ1
]
of our HSO3-MSN catalyst. How-
ever, based on the results of our
study, we determined that 608C
was the most economical and
cellulase-Fe3O4@MSN
isomerase-Fe3O4@MSN 645.1
HSO3-MSN 276.2
587.8
272.6
283.1
285.8
20.2
20.4
4.7
7.3
0.65
–
–
–
–
–
0.8–2.0 1.290
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 3241 – 3246 3242

CHEMSUSCHEM
COMMUNICATIONS
www.chemsuschem.org
alized MSNs (i.e., COOH-MSN,
H2PO3-MSN, and HSO3-MSN) for
the production of HMF under
the same reaction conditions.
Figure 1(d) illustrates that HSO3-
MSN was far more effective than
the other two MSNs for both
fructose conversion and HMF
yield. Acid density and strength
of these three acid-functional-
ized MSNs were measured by IR
and 29Si NMR spectroscopic char-
acterization (data not shown),
and the results showed that
strong Brønsted acid sites in
HSO3-MSN with a high amount
of 1.29 mmol(H+)gÀ1 were more
effective for the fructose-to-HMF
conversion as compared to the
as-synthesized MSN sample. Sev-
eral previous studies have pro-
vided similar results by using
mesoporous materials and spe-
cially designed ionic liquids.[1b,9]
Figure 1. (a) Reaction temperature in fructose conversion to HMF in integrated method using HSO3-MSN in DMSO
media. (b) Effect of water/DMSO (v/v) on the HMF yield (%) with and without the HSO3-MSN catalyst. (c) Influence
of the loading amount of catalyst on the yield of HMF. (d) Effect of the different acid functionalities (COOH, H2PO3,
HSO3) of MSN catalysts on the fructose conversion and HMF yield.
efficient temperature for producing HMF (Supporting Informa-
tion, Table S1).
To demonstrate the successful integration of the enzymatic
(water) reaction with the chemical (DMSO) reaction, we at-
tempted to perform a glucose-to-fructose and fructose-to-HMF
cascade conversion. Conditions for the glucose-to-fructose re-
action were as follows: phosphate buffer, pH 7.5, reaction tem-
perature=708C, reaction time=24 h, catalyst=isomerase-
Fe3O4@MSN (15 mg). After the reaction, we separated the iso-
merase-Fe3O4@MSN catalyst by using a magnet, and then
added the DMSO and HSO3-MSN catalyst. Conditions for the
fructose-to-HMF reaction were as follows: phosphate buffer/
DMSO mixed solution (water content=10 vol%), reaction
time=15 h, amount of catalyst=150 mg. Entries 1 to 3 in
Table 2 indicate that the reaction temperature has a substantial
influence on the final yield of HMF. At room temperature, the
HMF yield was nearly zero and the amount of fructose was the
Because we intended to integrate a water system with an
organic solvent system, the amount of water in the final organ-
ic/aqueous biphasic system was critical for the final HMF yield.
Figure 1(b) demonstrates that at a reaction temperature of
608C, the presence of H2O completely inhibits the production
of HMF, compared to a yield of 55% when no water was pres-
ent, in a pure DMSO system. In the presence of the HSO3-MSN
catalyst, we were able to generate HMF and maintain a yield
of over 60% when the water content was less than 33 vol%.
However, the HMF yield decreased to less than 20% when the
water content exceeded 50vol%, indicating that an increase in
water content causes a decrease of the HMF yield. This is be-
cause the fructose-to-HMF conversion is a dehydration reac-
tion, and the presence of water disturbs the equilibrium re-
quired for the formation of HMF. We obtained similar results
for a different organic solvent (i.e., THF) system. Therefore, we
determined that the optimal water content was 10 vol% [i.e.,
an aqueous/DMSO (v/v) ratio of 1:10; see Supporting Informa-
tion, Table S2].
Table 2. Production of HMF from glucose through a glucose-to-fructose-
to-HMF sequential reaction.[a]
Entry
T
[8C]
Conversion
[%]
Fructose
[mg]
Yield of
HMF [%]
The third factor affecting HMF yield is the amount of the
catalyst. Figure 1(c) indicates that for the same reaction condi-
tions (15 mg of fructose, 10% water content, and a tempera-
ture of 608C), the yield of HMF increased as the amount of
HSO3-MSN catalyst increased (from 2 to 150 mg). The HMF
yield peaked at 150 mg and retained the same value even
when more than 150 mg of catalyst was present. Therefore, we
determined that 150 mg was the optimal amount of catalyst
(Supporting Information, Table S3).
1
2
3
4[b]
5[c]
6[d]
27
60
90
60
60
60
57.8
66.3
70.5
63.8
56.3
55.1
0.045
0.005
0
0.009
0
0
46.1
45.6
40.1
4.1
0
3.7
[a] Reaction conditions for cellulose-to-glucose: phosphate buffer, pH 4.8,
T=508C, t=24 h, catalyst=cellulase-Fe3O4@MSN (15 mg). Reaction con-
ditions for glucose-to-fructose: phosphate buffer, pH 7.5, T=708C, t=
24 h, catalyst=isomerase-Fe3O4@MSN (15 mg). Reaction conditions for
fructose-to-HMF: phosphate buffer/DMSO mixed solution (water con-
tent=10 vol.%), t=15 h, catalyst=HSO3-MSN (0.15 g). [b] Catalyst=
HSO3-MSN (0.1 g). [c] No isomerase-Fe3O4@MSN. [d] No phosphate buffer.
Comparing the effects of the various acidic groups of the
MSN catalysts on the fructose-to-HMF conversion provided val-
uable information. We synthesized three organic-acid-function-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 3241 – 3246 3243

CHEMSUSCHEM
COMMUNICATIONS
www.chemsuschem.org
highest, indicating that the cascade reaction stopped after
fructose conversion. This result is consistent with the afore-
mentioned optimization of the fructose-to-HMF reaction. Al-
though an increased temperature (i.e., 908C) increases glucose
conversion, the final HMF yield did not increase as it did when
the temperature was 608C. The results of entries 2 and 4 sug-
gest that the amount of catalyst affects the conversion of fruc-
tose and the final yield of HMF. As negative controls, when the
isomerase-Fe3O4@MSN catalyst was not added in the phos-
phate buffer (entry 6) or was absent from the glucose-to-fruc-
tose reaction (entry 5), the subsequent fructose-to-HMF reac-
tion did not succeed, even with the addition of the HSO3-MSN
catalyst. These results indicate that the source of the second
step of the fructose-to-HMF conversion (i.e., fructose) is the
product of the first step of the glucose-to-fructose conversion.
Finally, we used cellulose as starting material and demon-
strated the cellulose-to-glucose-to-fructose-to-HMF cascade re-
action. We previously optimized the reaction conditions for the
sequential cellulose-to-glucose and glucose-to-fructose conver-
sions by using cellulase-Fe3O4@MSN and isomerase-
Fe3O4@MSN as the respective catalysts (Supporting Informa-
tion, Table S4). To cascade two enzymes, we used the same
medium (i.e., phosphate buffer) for cellulase and isomerase.[7d]
Therefore, after the cellulose-to-glucose conversion, we simply
collected the cellulase-Fe3O4@MSN catalyst by using a magnet,
increased the reaction temperature to 708C and the pH value
to 7.4, and added the second catalyst isomerase-Fe3O4@MSN.
After the glucose-to-fructose conversion, we again collected
the isomerase-Fe3O4@MSN catalyst by using a magnet, de-
creased the reaction temperature to 608C, and added the third
catalyst (HSO3-MSN) and the organic solvent (DMSO) (maintain-
ing the 1:10 ratio of buffer to DMSO). After reacting for 15 h,
the final yield of HMF reached 45.6%, nearly the same as that
using glucose as starting reactant (i.e., 46.1%). In this reaction,
the strongly acidic property of HSO3-MSN was useful, unlike
weak-acid surface sites of porous materials that act as suitable
candidates for biopolymer chain-breaking.[10,11] The very pur-
pose of utilizing the ionic liquid [BMIM]Cl for the pre-treatment
of cellulose is to disrupt the interactions between hydrogen-
bonded sheets in cellulose and solvation of microfibrils consist-
ing of a large number of glucan chains, which is possible via
disruption mechanism by [BMIM]Cl.[12(a)] In this case, nucleophil-
ic imidazole can attack and disrupt the hydrogen bonds in cel-
lulose by converting to a mixture of modified cellulose and
amorphous cellulose among which later is more prone to en-
zymatic hydrolysis.[12(b)]
Figure 2. Recyclability test of the HSO3-MSN for the HMF production in an
integrated, cascading method of cellulose conversion.
(e.g., ionic liquid, aqueous media, mechanochemical)[14] offer
selective production of HMF. Compared to these techniques,
we integrated enzymatic and chemocatalytic processes, and
the whole sequence can be triggered by selecting the desired
catalysts, controlling the pH value of the phosphate, and
switching the reaction medium from aqueous (enzymatic) to
organic (chemocatalytic) at the appropriate steps. These
unique features ultimately provide a higher yield of HMF than
do the aforementioned strategies.
In summary, we report for the first time an integrated
enzyme cascade with a chemocatalytic step for cellulose-to-
HMF conversion. This study demonstrates how cellulose de-
construction can be achieved using an enzyme cascade reac-
tion, using Fe3O4-loaded, enzyme-immobilized mesoporous
silica nanoparticles (MSN) materials that can be easily separat-
ed from the reaction medium at each step by applying an ex-
ternal magnetic force. We also demonstrate the effectiveness
of the cellulase-immobilized Fe3O4@MSN for deconstructing
glycosidic bonds of cellulose at pH 4.8, and of isomerase-im-
mobilized Fe3O4@MSN for the isomerization of glucose to fruc-
tose at pH 7.5, during the cascade reaction sequence. We then
integrated enzymatic reactions with the chemocatalytic reac-
tion of the fructose-to-HMF conversion by optimizing several
critical factors, such as reaction temperature, ratio of water to
organic solvent, and the amount of the acidic catalyst HSO3-
MSN, which caused the entire process to be a one-vessel cas-
cade reaction, offering cellulose deconstruction under milder
enzymatic conditions, and generating a high HMF yield of
45%. The results obtained in this study indicate that the con-
cept of integrating enzymatic and chemocatalytic biomass
processing can be an effective and economically friendly pro-
cess for various catalytic applications.
Experimental Section
We then performed the recycle test for the cellulose-to-HMF
cascade conversion. Figure 2 indicates that the yield of HMF
was similar for each recycle, indicating that the three catalysts
used in this study can be successfully recycled while maintain-
ing their catalytic ability. A decrease of HMF yield of ca. 7%
from the first to fifth run catalyzed by HSO3-MSN was found,
however, this loss is possibly due to surface deactivation of
MSNs and inaccessible HSO3 groups with each cycles.
Chemicals: Poly(oxyethylene) oleyl ether (Brij-97, C18H35EO10), am-
monia hydroxide (37%), hydrochloride acid (37%), iron(II) chloride
tetrahydrate, 3-aminopropyltrimethoxysilane (APTMS, 97%), di-
methyl phthalate (DOP), tetraethoxysilane (TEOS), ethanol (99.8%),
cellulase (Trichoderma reesei ATCC 26921), 1-butyl-3-methyl-imida-
zolium chloride (BMIM), cellulose (powder, ca. 20 mm), d-(+)-glu-
cose (>99.5%), d-(À)-fructose (>99%), sodium phosphate tribasic,
magnesium sulphate, and sodium chloride were purchased from
Sigma–Aldrich (Taipei, Taiwan). Citric acid, sodium hydroxide
(NaOH), and acetonitrile were purchased from J. T. Baker. Iron(III)
chloride hexahydrate (FeCl3·6H2O) was purchased from Alfa Aesar.
Researchers have developed numerous chemocatalytic sys-
tems for deconstructing cellulose and performing further con-
versions;[13] however, few cellulose deconstruction strategies
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 3241 – 3246 3244

CHEMSUSCHEM
COMMUNICATIONS
www.chemsuschem.org
Methyl alcohol was purchased from Mallinckrodt Chemical. Glucose
isomerase (purified from Streptomyces rubiginosus) was purchased
from Hampton Research. Organosilanes, such as 3-mercaptopropyl-
trimethoxysilane (MPTMS), 2-carbomethoxy- ethyltrimethoxysilane
(CTES), and diethylphosphato-ethyltriethoxysilane (DPTS), were
purchased from Gelest.
Pretreatment of cellulose with ionic liquids: Cellulose (50 mg)
was introduced into an ionic liquid (i.e., [BMIM]Cl, 0.95 mL) by stir-
ring at 1208C for 1 h. Methanol (3 mL) was added to the mixture
to quench the reaction and the resulting oligomer was separated
from the ionic liquid through centrifugation followed by washing
several times, by using methanol and water, and drying in lyophil-
izer.
Characterization of MSN-based catalysts: The morphology of
MSN-based catalysts was analyzed by using scanning electron mi-
croscopy (Nova Nano SEM) and transmission electron microscopy
(JEOL JEM 2100F). The porous properties were analyzed using ni-
trogen adsorption/desorption isotherms on a Micromeritics ASAP
2000 instrument. The specific surface area and pore size were cal-
culated using the Brunauer–Emmet–Teller and Barrett–Joyner–Ha-
lenda methods, respectively.
Synthesis of magnetite (Fe3O4) nanoparticles: FeCl3 (hexahydrate,
1.349 g) and FeCl2 (tetrahydrate, 0.781 g) were dissolved into de-
ionized water (600 mL) by stirring, followed by the addition of am-
monia hydroxide (1.5m) to the iron-containing aqueous solution
until the pH value of the medium increased to 9. The iron oxide
(i.e., Fe3O4) product was collected using external magnetic force
and washed several times by using deionized water and ethanol.
The resulting sample was re-dispersed in 600 mL of deionized
water for further use.
Sequential conversion of cellulose involving enzymatic and
chemical catalysis: To test the catalytic abilities of free enzymes
and immobilized enzymes, ILs-pretreated cellulose (15 mg) was
added to a citric buffer (1 mL) containing free cellulase or cellulase-
immobilized Fe3O4@MSNs at 508C for 1 d to cause cellulose-to-glu-
cose conversion. For the glucose-to-fructose conversion, glucose
(15 mg) was added to a phosphate buffer (1 mL) containing free
isomerase or isomerase-immobilized Fe3O4-loaded MSNs at 708C
for 1 d. For the fructose-to-HMF conversion, fructose (15 mg) was
added to a DMSO (5 mL) containing various acid-functionalized
MSNs (15 mg) at 608C for 1 d. For the cellulose-to-HMF conversion,
a sequential cellulose-to-glucose, glucose-to-fructose, and fructose-
to-HMF reaction was catalysed using three separate MSN-based
catalysts (i.e., cellulase-immobilized Fe3O4@MSNs for the first step,
isomerase-immobilized Fe3O4@MSNs for the second step, and
SO3H-functionalized MSNs for the third step). Typically, cellulase-im-
mobilized Fe3O4@MSNs (15 mg) were added to a phosphate buffer
(pH 4.8, 1 mL) containing ILs-pretreated cellulose (15 mg) at 508C.
After reacting for 1 d, the cellulase-immobilized Fe3O4@MSNs were
collected using a magnet, and the supernatant was transferred to
another vial that contained isomerase-immobilized Fe3O4@MSNs
(15 mg). Sodium hydroxide was then added to adjust the pH value
of the system to 7.5, and the reaction was conducted for another
day at 708C. After reaction, isomerase-immobilized Fe3O4@MSNs
were collected using an external magnet, and the supernatant was
transferred to another vial that contained DMSO and HSO3-func-
tionalized MSNs (15 mg). The reaction was run for 1 d at 608C to
obtain HMF as a final product.
Synthesis of Fe3O4-loaded MSN: Fe3O4-loaded MSN was synthe-
sized using a co-condensation method as follows. Brij-97 (6.92 mL)
was added to 180 mL (removed from 600 mL) of magnetite-con-
taining aqueous solution with stirring at room temperature. After
a complete dissolution of Brij-97, APTMS (0.3 mL) and DOP (0.8 mL)
were added to the Brij-97-manetite-contained medium with stir-
ring. After stirring for 30 min, TEOS (6.7 mL) was introduced to the
magnetite/Brig-97/APTMS/DOP solution, and the mixture was
stirred at RT for 1 d, followed by refluxing at 1008C for another
24 h. Finally, the precipitate was collected through filtration,
washed several times by using methanol to remove the surfactant,
and dried in lyophilizer. The resulting sample is denoted as
Fe3O4@MSNs.
Enzyme immobilization: For immobilization of cellulase,
Fe3O4@MSN (50 mg) was suspended in a citric buffer (10 mm, 2 mL,
pH 4.8). To this suspension, 1 mL of cellulase solution was added,
and the whole mixture was stirred at 48C for 1 d. For immobiliza-
tion of isomerase, the same amount of Fe3O4@MSN was suspended
in a phosphate buffer (20 mm sodium phosphate/0.15m sodium
chloride/5 mm magnesium sulfate, pH 7.5). Isomerase solution
(0.5 mL) was then added to the phosphate buffer and the whole
mixture was stirred at 48C for 1 d. Finally, the enzyme-immobilized,
Fe3O4-loaded MSNs (denoted as cellulase-Fe3O4@MSN and isomer-
ase-Fe3O4@MSN) were collected using a magnet. The enzyme re-
maining in the supernatant was considered as a non-adsorbed
enzyme, and the amount was measured using a UV-Vis spectrome-
ter at a wavelength of 280 nm. The final catalysts were washed
several times by using either a citric (for cellulase) or a phosphate
(for isomerase) buffer, and re-dispersed into a citric or a phosphate
buffer for further use.
Analysis of products: The products after reaction were analyzed
using a high-performance liquid chromatography system (ASI500
system) equipped with a Shodex NH2P50 4E column. Before the
products were subjected to HPLC, impurities were removed using
a syringe filter. Possible products such as cellobiose, glucose, fruc-
tose, and HMF were identified prior to product analysis, and cali-
bration curves were measured. The definition of conversion and
yield is described in the Supporting Information.
Functionalization of MSN with varying acid functional groups:
1 g of MSN was placed into a two-neck round-bottom flask, and
then degassed in vacuum at 1108C for 30 min. Toluene (30 mL)
was then added to the flask in nitrogen atmosphere. For function-
alization of the sulfonic group (SO3H), MPTMS (2.23 mL) was added
and the whole system was refluxed at 1108C. After reacting for 1 d,
the thiol-functionalized MSN sample was collected through centri-
fugation. The thiol group (SH) on the MSN was further converted
into a sulfonic group (SO3H) by reacting SH-MSN with H2O2 at RT
for 1 d. The final SO3H-functionalized MSN was collected and dried
in vacuum. Similarly, for functionalization of the carboxylic group
(COOH), CTES (2.66 mL) was first grafted onto MSN in a toluene
system at 1108C for 1 d. The sample (0.6 g) was then furthered
treated with H2SO4 (48 wt%, 90 mL) at 958C for 1 d. For functionali-
zation of the phosphoric group (H2PO3), DPTS (3.88 mL) was graft-
ed onto MSN in a toluene system at 1108C for 1 d. The sample
(0.6 g) was then furthered treated with HCl (37 wt%, 10 mL) at
908C for 1 d.
Acknowledgements
This research was supported by the Ministry of Science and Tech-
nology (MOST) of Taiwan (101-2628-E-002-015-MY3, 101-2623-E-
002-005-ET, 101-2923-E-002-012-MY3, and 103-2218-E-002-101) of
National Taiwan University (101R7842 and 102R7740), and the
Center of Strategic Materials Alliance for Research and Technolo-
gy (SMART Center) of National Taiwan University (102R104100).
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 3241 – 3246 3245

CHEMSUSCHEM
COMMUNICATIONS
www.chemsuschem.org
2007, 316, 1597–1600; c) J. Carlos Serrano-Ruiz, J. A. Dumesic, Energy
Environ. Sci. 2011, 4, 83–99; d) A. S. Amarasekara, L. D. Williams, C. C.
Ebede, Carbohydr. Res. 2008, 343, 3021–3024; e) Y. Y. Lee, K. C.-W. Wu,
Phys. Chem. Chem. Phys. 2012, 14, 13914–13917.
Keywords: biomass · enzyme catalysis · immobilization ·
mesoporous materials · renewable resources
[9] a) J. B. Binder, R. T. Raines, J. Am. Chem. Soc. 2009, 131, 1979–1985;
b) Z. Huang, W. Pan, H. Zhou, F. Qin, H. Xu, W. Shen, ChemSusChem
2013, 6, 1063–1069; c) W.-H. Peng, Y.-Y. Lee, C. Wu, K. C.-W. Wu, J.
Mater. Chem. 2012, 22, 23181–23185.
[1] a) J. Carlos Serrano-Ruiz, R. Luque, A. Sepulveda-Escribano, Chem. Soc.
Rev. 2011, 40, 5266–5281; b) M. E. Zakrzewska, E. Bogel-Lukasik, R.
Bogel-Lukasik, Chem. Rev. 2011, 111, 397–417; c) P. Gallezot, Chem. Soc.
Rev. 2012, 41, 1538–1558.
[10] a) A. Dutta, D. Gupta, A. K. Patra, B. Saha, A. Bhaumik, ChemSusChem
2014, 7, 925–933; b) M. G. Mazzotta, D. Gupta, B. Saha, A. K. Patra, A.
Bhaumik, M. M. Abu-Omar, ChemSusChem 2014, DOI: 10.1002/
cssc.201402007; c) Z. Fei, R. Scopelliti, G. Leurenczy, S. Katsyuba, N. Yan,
P. J. Dyson, ChemSusChem 2014, DOI: 10.1002/cssc.201301368.
[11] P.-W. Chung, A. Charmot, O. A. Olatunji-Ojo, K. A. Durkin, A. Katz, ACS
Catal. 2014, 4, 302–310.
[12] a) A. S. Gross, A. T. Bell, J. W. Chu, J. Phys. Chem. B 2011, 115, 13433;
b) K. Okushita, E. Shikayama, J. Kikuchi, Biomacromolecules 2012, 13,
1323–1330.
[2] M. E. Himmel, S.-Y. Ding, D. K. Johnson, W. S. Adney, M. R. Nimlos, J. W.
Brady, T. D. Foust, Science 2007, 315, 804–807.
[3] a) U. T. Bornscheuer, G. W. Huisman, R. J. Kazlauskas, S. Lutz, J. C. Moore,
K. Robins, Nature 2012, 485, 185–194; b) J. A. Rollin, T. K. Tam, Y. H. P.
Zhang, Green Chem. 2013, 15, 1708–1719.
[4] a) Y. Chi, S. T. Scroggins, J. M. J. Frechet, J. Am. Chem. Soc. 2008, 130,
6322–6323; b) C. You, H. Chen, S. Myung, N. Sathitsuksanoh, H. Ma, X.-
Z. Zhang, J. Li, Y. H. P. Zhang, Proc. Natl. Acad. Sci. USA 2013, 110, 7182–
7187.
[5] a) A. Katz, M. E. Davis, Nature 2000, 403, 286–289; b) S. Shylesh, A. Wa-
gener, A. Seifert, S. Ernst, W. R. Thiel, Angew. Chem. Int. Ed. 2010, 49,
184–187; Angew. Chem. 2010, 122, 188–191; c) N. Suzuki, M. B. Zakaria,
Y.-D. Chiang, K. C. W. Wu, Y. Yamauchi, Phys. Chem. Chem. Phys. 2012,
14, 7427–7432.
[13] a) A. Wang, T. Zhang, Acc. Chem. Res. 2013, 46, 1377–1386; b) N. Ji, M.
Zheng, A. Wang, T. Zhang, J. G. Chen, ChemSusChem 2012, 5, 939–944;
c) Z. Tai, J. Zhang, A. Wang, J. Pang, M. Zheng, T. Zhang, ChemSusChem
2013, 6, 652–658.
[14] a) B. Kim, J. Jeong, D. Lee, S. Kim, H.-J. Yoon, Y.-S. Lee, J. K. Cho, Green
Chem. 2011, 13, 1503–1506; b) M. Tan, L. Zhao, Y. Zhang, Biomass Bio-
energy 2011, 35, 1367–1370; c) S. Zhao, M. Cheng, J. Li, J. Tian, X.
Wang, Chem. Commun. 2011, 47, 2176–2178; d) B. R. Caes, M. J. Palte,
R. T. Raines, Chem. Sci. 2013, 4, 196–199; e) J. Fan, M. De Bruyn, V. L. Bu-
darin, M. J. Gronnow, P. S. Shuttleworth, S. Breeden, D. J. Macquarrie,
J. H. Clark, J. Am. Chem. Soc. 2013, 135, 11728–11731; f) C. Tian, C. Bao,
A. Binder, Z. Zhu, B. Hu, Y. Guo, B. Zhao, S. Dai, Chem. Commun. 2013,
49, 8668–8670.
[6] a) H. Yamada, C. Urata, Y. Aoyama, S. Osada, Y. Yamauchi, K. Kuroda,
Chem. Mater. 2012, 24, 1462–1471; b) H. Yamada, C. Urata, H. Ujiie, Y.
Yamauchi, K. Kuroda, Nanoscale 2013, 5, 6145–6153; c) M. Nandi, J.
Mondal, K. Sarkar, Y. Yamauchi, A. Bhaumik, Chem. Commun. 2011, 47,
6677–6679; d) K. C.-W. Wu, Y. Yamauchi, J. Mater. Chem. 2012, 22,
1251–1256; e) Y. Yamauchi, J. Ceram. Soc. Jpn. 2013, 121, 831–840.
[7] a) A. Takimoto, T. Shiomi, K. Ino, T. Tsunoda, A. Kawai, F. Mizukami, K. Sa-
kaguchi, Microporous Mesoporous Mater. 2008, 116, 601–606; b) S. B.
Hartono, S. Z. Qiao, J. Liu, K. Jack, B. P. Ladewig, Z. Hao, G. Q. M. Lu, J.
Phys. Chem. C 2010, 114, 8353–8362; c) R. H.-Y. Chang, J. Jang, K. C.-W.
Wu, Green Chem. 2011, 13, 2844–2850; d) Y.-C. Lee, C.-T. Chen, Y.-T.
Chiu, K. C.-W. Wu, ChemCatChem 2013, 5, 2153–2157.
Received: June 29, 2014
Revised: August 8, 2014
[8] a) Y. Roman-Leshkov, J. N. Chheda, J. A. Dumesic, Science 2006, 312,
1933–1937; b) H. Zhao, J. E. Holladay, H. Brown, Z. C. Zhang, Science
Published online on September 26, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 3241 – 3246 3246