Utilizing imogolite nanotubes as a tunable catalytic material for the selective isomerization of glucose to fructose

Article abstract of DOI:10.1016/j.cattod.2018.07.059

The isomerization of glucose to fructose is an important step in the conversion of biomass to valuable fuels and chemicals. A key challenge for the isomerization reaction is achieving high selectivity towards fructose using recyclable and inexpensive catalysts. In this study, the isomerization of glucose to fructose is investigated through using imogolite nanotubes as a catalyst. Imogolite is a single-walled aluminosilicate nanotube characterized by surface areas of 200-400 m²/g and pore widths near 1 nm. Imogolite is a highly tunable structure and can be modified through substitution of Si with Ge or through functionalization of methyl groups to the inner surface. These modifications change the surface properties of the nanotubes and enable tuning of the catalytic performance. Imogolite nanotubes are successfully used as a heterogeneous catalyst for the isomerization of glucose to fructose. Of the compositions tested, aluminosilicate imogolite is the most active material for the conversion of glucose, achieving a glucose conversion of 30% and a fructose selectivity of 45%. Catalyst recycling experiments reveals that organic content accumulates on the nanotubes that results in a minor reduction in conversion while maintaining similar catalytic selectivity. The catalyst can be washed with aqueous ammonia, allowing the productivity of fructose to be recovered. Overall, imogolite nanotubes are an active and tunable catalytic platform with moderate selectivity for the isomerization of glucose to fructose.

Full text of DOI:10.1016/j.cattod.2018.07.059

Accepted Manuscript
Title: <!–<query id="Q2">Your article is registered as
¨
belonging to the Special Issue/Collection entitled Honoring
Umit S. Ozkan. If this is NOT correct and your article is a
¨
regular item or belongs to a different Special Issue please
contact I.Samikannu@elsevier.com immediately prior to
returning your corrections.</query>–>Utilizing imogolite
nanotubes as a tunable catalytic material for the selective
isomerization of glucose to fructose
Authors: Nathaniel Olson, Nitish Deshpande, Seval Gunduz,
Umit S. Ozkan, Nicholas A. Brunelli
PII:
S0920-5861(18)31339-7
DOI:
Reference:
https://doi.org/10.1016/j.cattod.2018.07.059
CATTOD 11604
To appear in:
Catalysis Today
Received date:
Revised date:
Accepted date:
4-5-2018
8-7-2018
31-7-2018
Please cite this article as: Olson N, Deshpande N, Gunduz S, Ozkan US,
Brunelli NA, Utilizing imogolite nanotubes as a tunable catalytic material for
the selective isomerization of glucose to fructose, Catalysis Today (2018),
https://doi.org/10.1016/j.cattod.2018.07.059
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Utilizing Imogolite Nanotubes as a Tunable Catalytic Material for the Selective
Isomerization of Glucose to Fructose
Nathaniel Olson, Nitish Deshpande, Seval Gunduz, Umit S. Ozkan, Nicholas A. Brunelli*
William G. Lowrie Department of Chemical and Biomolecular Engineering, 151 W. Woodruff
Ave, Columbus, OH 43210
*Corresponding Author: brunelli.2@osu.edu; Twitter: @OSUChemEProfBru
Graphical abstract
Highlights
 Imogolite is an active catalyst for the isomerization of glucose to fructose
 The structure of imogolite can be modified with germanium and methyl groups
 Imogolite nanotubes has maintains significant catalytic activity and selectivity
 Adsorbed organic species can be removed through a simple base wash
1

Abstract
The isomerization of glucose to fructose is an important step in the conversion of biomass
to valuable fuels and chemicals. A key challenge for the isomerization reaction is achieving high
selectivity towards fructose using recyclable and inexpensive catalysts. In this study, the
isomerization of glucose to fructose is investigated through using imogolite nanotubes as a
catalyst. Imogolite is a single-walled aluminosilicate nanotube characterized by surface areas of
200-400 m2/g and pore widths near 1 nm. Imogolite is a highly tunable structure and can be
modified through substitution of Si with Ge or through functionalization of methyl groups to the
inner surface. These modifications change the surface properties of the nanotubes and enable
tuning of the catalytic performance. Imogolite nanotubes are successfully used as a heterogeneous
catalyst for the isomerization of glucose to fructose. Of the compositions tested, aluminosilicate
imogolite is the most active material for the conversion of glucose, achieving a glucose conversion
of 30% and a fructose selectivity of 45%. Catalyst recycling experiments reveals that organic
content accumulates on the nanotubes that results in a minor reduction in conversion while
maintaining similar catalytic selectivity. The catalyst can be washed with aqueous ammonia,
allowing the productivity of fructose to be recovered. Overall, imogolite nanotubes are an active
and tunable catalytic platform with moderate selectivity for the isomerization of glucose to
fructose.
Keywords: Biomass utilization; glucose isomerization; imogolite; nanotube
Keywords
Biomass; Glucose isomerization; Imogolite; aluminosilicate material; Surface characterization
Abbreviations
 IMO = imogolite
 AlSi = aluminosilicate
1. Introduction
Creating sustainable methods to produce chemicals and fuels can be achieved through
identifying robust and stable catalytic materials for conversion of biomass. Indeed, the primary
2

challenge for biomass conversion is achieving high selectivity with a stable heterogeneous
catalytic material. One important chemical reaction for biomass conversion is the isomerization of
glucose to fructose [1]. This reaction remains an important bottleneck in producing the important
intermediate 5-hydroxymethylfurfuraldehyde (HMF), which is considered an important chemical
to produce to enable large scale biomass processing [2]. The importance of this overall reaction
sequence merits investigation of alternative catalytic materials.
Several intriguing catalysts for this reaction have been investigated, including enzymes [3],
Lewis acids [4], inorganic Brønsted bases [5,6], and organic bases [7,8]. The most common
industrial catalyst for isomerization of glucose to fructose is the enzyme glucose isomerase.
Though enzymes are widely used for food applications, it is necessary to maintain strict control of
pH, temperature, and feed purity to avoid catalyst deactivation [3]. To enable large scale biomass
utilization, it is desirable to develop more robust catalytic materials. Seminal work discovered that
glucose isomerization could be catalyzed by the Lewis acidic zeolite Sn-Beta [4]. This work has
sparked interest in catalysts for the isomerization of glucose to fructose, but commercial utilization
would require improving catalyst lifetime since Sn-Beta undergoes deactivation via leaching of Sn
active sites [9]. Several inorganic base catalysts have been identified for the reaction including
Ca(OH)2 and NaOH [5,6]. The yield of fructose achieved with these catalysts is thought to be low
because of the instability of monosaccharides when exposed to high concentrations of base
catalysts. Intriguing work has also investigated organic base such as triethylamine [7,8].
Homogeneous triethylamine exhibited good catalytic activity and selectivity, but homogeneous
catalysts are more difficult to recycle than heterogeneous catalysts. This work induced researchers
to investigate heterogeneous catalysts such as poly(ethylene imine) (PEI) [7] and tertiary amines
immobilized on silica materials [8]. These catalysts tended to deactivate via leaching and
formation of an acidic byproduct that is currently unknown, though the catalysts could be
regenerated through washing with pure water. Combining the ideas of using zeolites and base
catalyst, researchers have investigated sodium-exchanged zeolites [10]. These catalysts are active
and selective, but deactivate via leaching of the basic sodium species, making constant
regeneration of the active site necessary. The concept of base catalyzed glucose isomerization
certainly has merit, provided that the deactivation could be limited.
3

An interesting alternative material that has received limited investigation is imogolite.
Imogolite is a naturally occurring aluminosilicate nanotube with an empirical formula of
(OH)3Al2O3SiOH that was first discovered in 1962 in volcanic-ash derived soils in Japan [11]. A
synthetic procedure was first described in 1977 based upon the conditions in which imogolite
formed [12]. This procedure has been optimized over the years, and it can be synthesized with
monodisperse and tunable dimensions and composition [13–15]. Imogolite nanotubes have a
diameter of 2 nm and range in length from 100 to 1000 nm. Currently, limited characterization has
been performed to determine the nature of the surface properties of imogolite as a catalytic
material, but it is known that the accessible surface area on the outer wall of the imogolite nanotube
is comprised of a gibbsite sheet [16,17] and provides the most likely source of catalytic activity.
Gibbsite, which is a mineral form of aluminum hydroxide, displays amphoteric behavior [18],
displaying both weakly Bronsted basic and Lewis acidic behavior that could be used to catalyze
this reaction, which are both active for glucose isomerization. Compared to conventional gibbsite
alumina [19], the nanostructured nature of imogolite offers a larger surface area that can have a
greater number of catalytic sites with the additional intriguing aspect that tuning nanotube structure
could further modify the catalytic behavior.
Several modifications have been performed on imogolite that could provide the ability to
control its properties on the nanoscale. Aluminogermanate imogolite has been synthesized with up
to 100% substitution of silicon with germanium. The resulting nanotubes have larger, tunable
diameters and can form double-walled nanotubes [14,20,21]. Interior surface modification with
organic groups has also been performed using acetyl chloride, methyltrimethoxysilane, and
trichloro-silane [15]. The organic groups adjust the composition, hydrophobicity, and dimensions
of the nanotube. Iron-doped imogolite has been synthesized as a potential catalyst for oxidation
reactions [22]. Gold nanoparticle and imogolite composites have been fabricated for use as a
catalyst or a coloring agent [23]. Additionally, silver nanoparticles were immobilized on imogolite
and showed antibacterial properties [24]. It has also been shown that the phosphate groups of DNA
interact with the aluminum hydroxide groups on imogolite, which was utilized to form
imogolite/DNA hybrid hydrogels [25]. These different compositions suggest that imogolite is a
tunable platform.
4

The high level of modification offered by imogolite can serve as a tunable catalytic platform.
Yet, limited work has been done on application of imogolite nanotubes, especially in the realm of
catalysis [26–30]. To evaluate the promise that imogolite holds as tunable heterogeneous catalytic
platform, it is important to elucidate structure-function relationships of the modifications that have
been performed. Identification of design elements of imogolite would offer the ability to tune its
properties and enable its use as an effective catalytic platform.
In this work, imogolite is investigated as a catalyst for the selective isomerization of glucose
to fructose. The work examines the effect of modifying the composition of the imogolite nanotubes
through including germanium and methyl substitution to alter the surface properties and catalytic
performance. The most selective material is subjected to recycle experiments to determine the
robustness of the catalytic material. Overall, this work provides insight on the catalytic activity in
imogolite and its potential as a tunable, heterogeneous catalytic platform.
2. Experimental Methods
2.1. Chemicals
The following chemicals are used and stored in a glovebox: tetraethylorthosilicate (TOES;
98%, Acros Organics), germanium(IV) ethoxide (97%, BeanTown Chemical), aluminum-tri-sec-
butoxide (95%, Alfa Aesar), methyltrimethoxysilane (97%, Alfa Aesar). The following chemicals
are also used: perchloric acid (70% by wt., BDH), hydrochloric acid (36.5-38%, BDH), ammonia
(28-30%, Sigma Aldrich), dextrose (USP grade, Fisher), and D-mannitol (99%, Amresco). De-
ionized (DI) water is obtained by a house supply that produces Milli-Q quality water.
2.2. Imogolite Synthesis
The synthesis of imogolite nanotubes has been achieved through methods described
previously [31]. Perchloric acid is diluted to 38 mM from a 70 wt% (11.6 M) stock solution. A
solution is made with 5.454 g of 70 wt% perchloric acid (HClO4) and 1000 mL of DI water.
Precursors to imogolite are mixed in a glove box filled with nitrogen. For aluminosilicate imogolite
(Al-Si IMO), aluminum-tri-sec-butoxide (4.680 g) is mixed with TEOS (1.979 g) in a glass vial
and shaken vigorously. This precursor mixture is then added to a 38 mM HClO4 solution (250 mL)
in a 2 L round-bottom flask under stirring at room temperature. The final molar ratio is
Si:Al:HClO4 of 1:2:1. The solutions are stirred at room temperature for 24 h. The solution is then
5

diluted by a factor of 3.8 with DI water (700 mL) and stirred at 95°C for 4 days. After this period,
the solution is allowed to cool to room temperature and concentrated using one of two methods.
The first method involves concentrating the reaction mixture by a factor of 10 to approximately
100 mL using a rotary evaporator. Materials concentrated using a rotary evaporator are labeled
with “R.” All materials modified with germanium and methyl functionalization are concentrated
using the rotovap method. For the second method of concentration, a 30 wt% ammonia solution is
added dropwise until gelation of the nanotube solution occurs and a pH of 10 is reached. The gels
are then centrifuged at 9000 rpm for 10 min, and the supernatant is discarded. The nanotubes are
dispersed by adding a few drops of hydrochloric acid to the gel, which results in some fuming to
be observed. Materials concentrated using this acid-base addition method are labeled with “AB”.
The gels obtained by both methods are added to a 15,000 kDa membrane submerged in DI water.
The water is exchanged daily for 5 days. The purified gels are dried at 80°C, and the resulting
solids are ground vigorously to obtain imogolite nanotubes as a fine powder.
For germanium imogolite (Ge-x-IMO), aluminum-tri-sec-butoxide is mixed with
germanium(IV) ethoxide and tetraethylorthosilicate in a molar ratio of Si:Ge:Al:HClO4 of (1-
x):x:2:1. For Ge-0.3-IMO, this corresponds to combining 4.680 g aluminum-tri-sec-butoxide,
0.721 g germanium(IV) ethoxide, and 1.385 g TEOS. For Ge-0.8-IMO, the mixture consists of
4.680 g aluminum-tri-sec-butoxide, 1.922 g germanium(IV) ethoxide, and 0.396 g TEOS. For
methyl imogolite (Me-x-IMO), aluminum-tri-sec-butoxide is mixed with methyltrimethoxysilane
(Me-Si) and tetraethylorthosilicate in a molar ratio of Si:Me-Si:Al:HClO4 of (1-x):x:2:1. For Me-
0.5-IMO, this corresponds to 4.680 g aluminum-tri-sec-butoxide, 0.647 g methyltrimethoxysilane,
and 0.990 g TEOS. For Me-1.0-IMO, the mixture consists of 4.680 g aluminum-tri-sec-butoxide
and 1.294 g methyltrimethoxysilane.
2.3. Material Characterization
The materials are characterized using standard techniques, including nitrogen
physisorption, X-ray diffraction (XRD), and thermogravimetric analysis with differential scanning
calorimetry (TGA-DSC). The textural properties are analyzed using a Micromeritics 3Flex surface
characterization analyzer. Before analysis, the materials are degassed at 250°C using the SmartVac
Prep. The surface area is calculated using the BET method and the pore size is calculated using
the HK method. Powder X-ray diffraction data are collected on a Bruker D8 Advance powder
6

diffractometer (40 kV, 40 mA, sealed Cu X-ray tube) equipped with a Lynxeye XE-T position
sensitive detector in Bragg-Brentano geometry. TGA-DSC are performed using STA 449 F5
Jupiter® (NETZSCH Instruments). The analysis is performed under flowing air (20 mL/min) and
nitrogen (20 mL/min) at a ramp rate of 10°C/min from 30°C to 900°C followed by a 5 min hold at
900°C.
Methanol pulse experiments are conducted using a homebuilt system coupled with an MKS
Cirrus bench-top residual gas analyzer. The samples (10 mg of each catalyst) are packed in a ¼ in
OD quartz tube reactor placed in a tubular furnace. Helium is used as carrier gas at a flow rate of
30 sccm, passing the stream through a methanol bubbler that is maintained at room temperature to
produce a vapor concentration of 16% CH3OH in He. This stream is passed into a 500 µL sample
loop attached to a 6-port valve that can be switched with a manual actuator. After filling the sample
loop with methanol saturated helium, the valve is actuated to send the methanol containing pulse
to the sample. The exit stream is analyzed using an MKS Cirrus bench-top residual gas analyzer.
Pulses are sent until three equal-area peaks are observed in MS signal m/z=31, which indicates
that the surface is saturated with methanol.
2.4. Catalytic Testing
As illustrated in Scheme 1, catalytic testing is performed using a method that is similar to
previous work [8]. A bulk solution containing 4.0 g glucose and 36.0 g DI water is prepared. One
gram of the bulk solution is combined with 100 mg of catalyst in a 5 mL pressure tube. The tubes
are submerged in a silicone oil bath at 100°C at 420 rpm. After 24 hours, the tubes are transferred
to an ice bath for 15 min. The solution is diluted with 2 g of 0.3 M d-mannitol solution (internal
standard for HPLC measurements) and 30 g of DI water. The solution is centrifuged at 9000 rpm
for 15 min and the supernatant is filtered using a 0.22 μm nylon (Ø = 13 mm) syringe filter. The
filtered solution is analyzed using High Performance Liquid Chromatography (HPLC) from
Waters (Acquity) equipped with a Waters Sugar Pak-1 column (including pre-column filter) and
refractive index (RI) detector. DI water is used as the mobile phase at a flow rate of 0.20 mL/min
and a column temperature of 70°C. This analysis is used to quantify the concentrations of glucose,
fructose, and mannitol. The selectivity of the catalyst is calculated as the amount of fructose
produced per amount of glucose converted. The reported values are accurate to within 5%.
2.5. Catalyst Reuse Testing
7

Catalyst reuse testing is also performed to evaluate the ability to recycle the catalytic
material. The reaction began with 300 mg of catalyst and 3.0 g of 10 wt% glucose solution.
Following the reaction, the solution is diluted and centrifuged at 9000 rpm for 15 minutes, as
described previously. The supernatant is completely removed to separate the catalyst. DI water (25
mL) is added, and the centrifugation step is repeated. This step is performed two additional times
for a total of three washes. The catalyst is dried overnight at 80°C. The mass of the catalyst is
obtained through weighing the recovered material and then added to an amount of 10 wt% glucose
solution that corresponds to 100 mg catalyst per gram of 10 wt% glucose solution.
Acid and base treatment are applied in separate tests to regenerate the catalyst after a single
reaction. Post reaction, the catalyst is recovered through centrifugation, as described above. For
acid treatment, 100 mL of a 1.0 M HCl solution is prepared. For base treatment, 100 mL of a 1.0
M NH4OH solution is prepared. A solution of either acid or base (25 mL) is added to 50 mg of
dried catalyst obtained after a single reaction. The solution is mixed and then centrifuged at 9000
rpm for 15 minutes. The supernatant is completely removed to separate the catalyst. An additional
25 mL of acid or base solution is added for a total of four washes. DI water (25 mL) is added and
the centrifugation step is repeated. This step is performed two additional times for a total of three
DI water washes. The catalyst is dried overnight at 80°C prior to TGA/DSC measurements.
3. Results and Discussion
3.1. Catalyst Synthesis and Characterization
Imogolite nanotubes are successfully synthesized utilizing the procedure described
previously [31]. For all variations of imogolite, the solution turns cloudy upon combination of
precursors and the HClO4 solution. For the Al-Si and Ge-IMO, the solution then turns clear
approximately one hour after heating to 95°C. The Me-IMO remains cloudy throughout the
synthesis, potentially because of the increased hydrophobicity of the material [15]. Concentration
via the rotovap method yields a cloudy solution whereas the acid-base method yields a gel-like
solution. After drying, both result in a sheet-like material that must be ground vigorously to yield
imogolite as a fine powder.
The different imogolite nanotubes are characterized using standard techniques such as
powder XRD and nitrogen physisorption. The XRD pattern of Al-Si IMO is compared to Ge- and
8

Me-IMO in Figure 1. The important characteristic peaks of imogolite occur near 4, 9, 13, 28, and
40° 2θ. The characteristic peaks of imogolite remain in the Ge-IMO samples. New peaks begin to
form near 10° for 30% Ge and increase in intensity at 80% Ge, which are consistent with previous
attempts to synthesize Ge-IMO [14,32]. The shift of the first peak to smaller 2θ indicates an
increase in the inner diameter of the nanotube that is expected since Ge is a larger heteroatom,
requiring a larger radius of curvature. The Me-IMO patterns in Figure 1 contain the characteristic
peaks of imogolite. With increasing methyl content, a sharp feature forms near 20° that indicates
the formation of an impurity, corresponding to gibbsite [33].
The effects of composition on surface area and pore properties of imogolite are analyzed
with nitrogen physisorption (Figure S1). Table 1 provides a summary of imogolite materials
analyzed with nitrogen physisorption along with sites per surface area analyzed with methanol
pulse experiments. A single synthesis mixture of AlSi imogolite is split into two to analyze the
effect of the concentration method on the surface properties of imogolite. For the rotovap method,
the sample has a surface area of 273 m2/g with a pore width of 0.88 nm. For the acid-base method,
the surface area is 288 m2/g with a pore width of 0.88 nm. These results demonstrate that the
method of concentration does not have a significant impact on the surface properties of imogolite.
Germanium substitution results in changes to the surface area and pore width (Figure S1;
Table 1). The surface area of Ge-IMO is lower than that of AlSi IMO. Introduction of Ge into IMO
also results in an increased pore width according to nitrogen physisorption, which is corroborated
by the shift to the left in the first peak of the XRD patterns. At 30% Ge, a pore width of 0.91 nm
is achieved, which is greater than the pure IMO. The pore width further increases to 0.95 nm for
80% Ge. The increase in pore width can be explained by the increase in bond length from Si-O to
Ge-O, as has been previously reported [34], and is consistent with XRD results.
Methyl substitution leads to an increase in pore size from 0.94 nm at 50% methyl to 1.10
nm at 100% methyl (Figure S2; Table 1). The increased pore size could be attributed to the
presence of methyl groups on the inner surface of the nanotubes. The surface areas for these
materials are generally higher than AlSi IMO, 456 and 366 m2/g at 50% and 100% methyl,
respectively. This increase may result from the increased hydrophobicity of the material. Since
AlSi IMO has a high density of internal silanols, water remains strongly adsorbed to higher
temperatures whereas the Me-IMO has hydrophobic methyl groups that enables water to be
9

removed at low temperatures. TGA-DSC measurements indicate that a lower mass loss for
temperature less than 200°C for the methyl substituted materials (Figure S3 and S4). Though all
materials undergo heating under vacuum to remove any adsorbed compounds, it is possible water
remains more strongly adsorbed to pure imogolite than the modified materials [15]. Indeed, it has
been shown previously that the degas temperature has a strong effect on measured surface area.
While heating can help remove water, it can also cause irreversible structure collapse [17].
Therefore, a consistent temperature of 250°C is used for all sample degassing.
3.2. Glucose to Fructose Isomerization Results
Catalytic performance is evaluated through testing kinetic performance of AlSi imogolite
concentrated using the rotovap method. Figure 2 depicts the observed conversion and selectivity
at different time points over 48 hours for IMO-R. Conversion steadily increases with time and
selectivity decreases with time. Although the isomerization of glucose to fructose is an
equilibrium-limited reaction, an equilibrium composition of glucose and fructose is not achieved
using an imogolite catalyst. This test also demonstrates the activity of imogolite for the
isomerization of glucose. After 48 hours, the imogolite catalyst converts 35% of the glucose with
40% selectivity for fructose. Conversion continues to increase at 48 hours, providing evidence the
sites are not poisoned and the catalyst does not undergo deactivation, unlike previous organic
amine based catalysts [8].
The effect of concentration method for AlSi imogolite on catalytic activity is studied. At
100°C, no difference is observed in the conversion and selectivity of imogolite concentrated using
the rotovap and acid-base methods, as shown in Figure 3. The reaction is also run under inert
nitrogen atmosphere. Previous work has found that catalytic activity results are sensitive to the
presence of air [8] since the presence of CO2 can alter the pH of the reaction mixture. Relative to
air, running the reaction under inert conditions decreases the observed conversion from 30% to
18% and selectivity increases from 45% to 57% in 24 h. Although the selectivity is higher, this
selectivity is at a lower conversion. Comparing the results under nitrogen and under air, it is found
that all of the results lie along a general trend that is decreasing selectivity at higher conversions.
The imogolite catalyst is also tested at the temperature of 120°C to increase the glucose
conversion achieved in 24 hours. This temperature is also convenient since it is the temperature
commonly used for fructose dehydration to HMF, a reaction that is commonly coupled with
10

glucose isomerization [35]. At the temperature of 120°C, the imogolite catalyst achieved a
conversion of 74%, which is higher than the conversion achieved in 48 hours at 100°C. In addition,
imogolite is able to produce a selectivity of 22% to fructose. In comparison to catalytic results at
100°C, the selectivity appears lower at 120°C, but the reaction mixture is also found to contain the
fructose dehydration product of HMF with an approximate 5-8% yield of HMF. This yield of HMF
indicates more fructose is being produced, but that the fructose is being dehydration to HMF. It is
likely that fructose dehydration is not catalyzed by imogolite, but rather occurs in the absence of a
catalyst, as concurrent work in our lab will demonstrate [36].
The two modifications to imogolite investigated for effect on catalytic activity are
replacing silicon with germanium and replacing the internal silanols with methyl groups. Adding
germanium to the framework affects both the curvature of the material and the electronic properties
of the material. Catalytic testing of the material with the 30% germanium material (Ge-0.3-IMO)
reveals a decrease in the conversion. The selectivity for fructose is lower than other materials at
similar conversion. For Ge-0.8-IMO, the conversion is further decreased, and the selectivity is
increased to 58%, compared to 45% for AlSi IMO. The decrease in activity may be explained by
the formation of double-walled nanotubes with Ge-0.8-IMO. The second outer wall prevents
access to catalytic sites on the first inner wall, reducing the number of catalytic sites available in
the reaction. The decreased conversion and increased selectivity for Ge-0.8-IMO fits well with the
general trend established in Figure 3.
Functionalization of imogolite with methyl groups also has an effect on the catalytic
activity. At low temperatures, the conversion of glucose is greatly decreased to 12% for Me-0.5-
IMO with a high selectivity of 64%. This data again falls along the linear trend discussed
previously, where selectivity increases as conversion decreases. While all other catalytic materials
fall along the same trend line, Me-1.0-IMO deviates from the trend established in Figure 4. Me-
1.0-IMO has similar conversion to AlSi imogolite, but at much lower selectivity for fructose. The
origin of the difference in conversion is investigated through using methanol titration experiments.
The methanol titration experiments allow quantification of the total active sites [37], including
Bronsted acid, Bronsted base, and Lewis acid sites. The titration experiments reveal that the methyl
substituted imogolite nanotubes have fewer catalytic sites than the pure IMO materials, as shown
in Table 1. Fewer sites would result in lower catalytic activity for Me-0.5-IMO. For the highest
11

methyl-substituted material (Me-1.0-IMO), the methanol titration experiments indicate the lowest
number of catalytic sites, but this material is found to have similar conversion as pure AlSi IMO.
These results suggest that the methyl substitution is altering catalytic sites in the material to
produce less selective sites. The observed change could be associated with the change in curvature
of the material.
In general, the modification of imogolite has an effect on the structure and catalytic activity
of the material that leads to decreased conversion of glucose with higher selectivities for fructose.
It is hypothesized the aluminum hydroxide groups on the outer wall of the nanotube play a key
role in the catalytic activity of imogolite. With 100% methyl substitution, the material still
demonstrates catalytic activity. This provides evidence that silanol groups are not necessary for
imogolite to convert glucose into fructose. Further investigation of imogolite and its active sites
may lend to its use as an effective heterogeneous catalyst for other important reactions. Here, it is
demonstrated that imogolite nanotubes are an effective catalyst for the isomerization of glucose to
fructose. Formation of HMF is a promising indication of its use as a catalyst for the dehydration
of fructose to HMF. Typically, the heterogeneous nature of the catalyst would also be tested
through performing a hot filtration test. However, attempts to separate the imogolite catalyst via
filtration proved to be difficult. Even during catalyst synthesis, it is found that filtration is difficult.
Therefore, the heterogeneous nature of the catalyst will be demonstrated through catalyst reuse
testing.
3.3. Catalyst Reuse Testing
Catalyst reuse experiments are used to test the robustness of aluminosilicate imogolite as a
catalyst. Since this testing requires collecting the material after a fixed amount of reaction time, it
is useful to scale up the initial amount of catalyst to ensure recovery of sufficient material for
subsequent tests. The recycle tests are performed at 100°C for 24 hours, achieving a conversion of
25% with a selectivity of 56%, similar to reactions performed at smaller scale. After the reaction,
the catalyst is separated from the reaction mixture through centrifugation. The material is washed
with distilled water and centrifuged three times before drying in the oven.
After drying in the oven, the catalyst is tested in a second reaction. The recovered material
is combined with the fructose solution, maintaining a ratio of catalyst to reaction mixture of 100
mg to 1 g of 10 wt% fructose. For the second 24 hour reaction cycle, the catalyst achieves a
12

conversion of 21% with a selectivity of 49%. Compared to the first reaction cycle, the conversion
and the selectivity remain similar to the performance of the fresh catalyst. A third reaction cycle
results in conversion of 17% with a selectivity of 49%. These results indicate that the catalytic
activity is decreasing, but the catalyst remains selective for the isomerization of glucose to
fructose.
Following the three reactions, the catalyst is characterized to provide insight on the
observed decrease in conversion. It is confirmed with XRD that the material retains the imogolite
structure (Figure S5). Nitrogen physisorption (Figure S6) reveals a significant decrease in surface
area to 95 m2/g from 288 m2/g. Interestingly, the isotherm reveals loss of micropore volume
associated with the internal pores of the imogolite nanotubes where the silanols are present. The
measured pore width also decreased to 0.71 nm from 0.88 nm. It is also observed that the material
changed color from white to brown after the reactions. It is hypothesized organic species adsorb
to the inner surface and are not removed during the washing and centrifugation steps. The adsorbed
species would both decrease the surface area and reduce the amount of actual catalytic material
that would be added during the recycle experiments, reducing the observed conversion. Since the
decrease in surface area is much larger than the decrease in conversion, the results are consistent
with adsorption of organic species inside the nanotubes on sites that have limited or no catalytic
activity. This further corroborates the previous results that the catalytic activity is associated with
the external surface of imogolite.
TGA/DSC is performed to compare the quantity and composition of adsorbed species on
the imogolite catalyst. Figures 4 display the TGA data for the as-synthesized IMO and the recycled
IMO after three reactions. The mass loss of the recycled IMO is much greater than the as-
synthesized material, indicative of more adsorbed species on the material. DSC reveals less water
is adsorbed on the recycled imogolite and indicates additional organic species adsorbed onto the
recycled material. The results of TGA/DSC corroborate the hypothesis generated from the physical
appearance of the material and decrease in surface area. The results of the reuse testing could be
promising for flow reactors if the organic species catalyst could be removed from the surface of
the catalyst.
TGA results can typically be used to identify the calcination temperature required to
regenerate the catalyst. However, the imogolite nanotubes are known to collapse irreversibly at
13

temperatures greater than 300°C [38]. As an alternative, it is examined how different treatments
can be used to desorb organic species. It is found that acid treatment using 1 M HCl, caused the
material to suspend in solution, making it difficult to separate using filtration or centrifugation.
The solubility of imogolite in acidic solution is consistent with the synthesis conditions since
imogolite must be concentrated or gelled out of a clear acidic solution. Thus, acid treatment is
determined to not be a viable route for imogolite regeneration. Importantly, treatment with a 1 M
solution of ammonium hydroxide allowed the material to be separated using centrifugation. TGA
of the base-treated material compared to the untreated material is shown in Figure 4. The base-
treated imogolite has more adsorbed water, but lower amounts of adsorbed organic species. This
is a promising result that shows a potential route to catalyst regeneration through treatment with
dilute NH4OH. The base wash procedure is repeated on the catalyst subjected to three prior
catalytic cycles. The recovered catalytic material is utilized in a fourth catalytic cycle, achieving a
fructose yield of 13%, consistent with the first reaction cycle. This demonstrates that the fructose
productivity can be recovered through washing the catalyst with base.
4. Summary
The catalytic activity of imogolite nanotubes is investigated for the selective conversion of
glucose to fructose. The selectivity for fructose decreases as the conversion of glucose increases.
Introduction of methyl groups or germanium changes the surface properties and catalytic activity
of the material, resulting in a lower catalytic activity and/or selectivity than the aluminosilicate
IMO. Testing of imogolite nanotubes in reuse experiments show organic species accumulate in
imogolite. The catalyst still maintains selectivity for glucose isomerization to fructose even after
catalyst reuse. Post-reaction characterization reveals a decrease in surface area and an increase in
adsorbed organic species. Treatment with dilute base appears to desorb these species and could be
used to regenerate imogolite for further reaction. Overall, these results demonstrate that imogolite
could be used as a recyclable heterogeneous catalyst for glucose isomerization.
Acknowledgement
Funding sources: This work was supported by the American Chemical Society Petroleum
Research Fund (ACS-PRF 55946-DNI5), the National Science Foundation (NSF CBET
14

1605037; 1653587 and NSF CBET REU 1645126), and the Ohio State University Institute of
Materials Research (OSU IMR FG0138) for their financial support.
Conflict of Interest Statement
The authors have no conflicts of interest to declare.
Supplementary Data
Supplementary data includes nitrogen physisorption and TGA data for the materials used
in this manuscript.
15

References
[1] L. Shuai, J.S. Luterbacher, Organic Solvent Effects in Biomass Conversion Reactions,
ChemSusChem. 9 (2015) 133–155. doi:10.1002/cssc.201501148.
[2] Y. Román-Leshkov, J.N. Chheda, J.A. Dumesic, Phase modifiers promote efficient
production of hydroxymethylfurfural from fructose, Science. 312 (2006) 1933–1937.
doi:10.1126/science.1126337.
[3] S.H. Bhosale, M.B. Rao, V. V Deshpande, Molecular and industrial aspects of glucose
isomerase, Microbiol. Rev. 60 (1996) 280–300.
[4] M. Moliner, Y. Román-Leshkov, M.E. Davis, Tin-containing zeolites are highly active
catalysts for the isomerization of glucose in water, Proc. Natl. Acad. Sci. 107 (2010)
6164–6168. doi:10.1073/pnas.1002358107.
[5] B.Y. Yang, R. Montgomery, Alkaline degradation of glucose: effect of initial
concentration of reactants, Carbohydr. Res. 280 (1996) 27–45. doi:10.1016/0008-
6215(95)00294-4.
[6] M. Watanabe, Y. Aizawa, T. Iida, T.M. Aida, C. Levy, K. Sue, H. Inomata, Glucose
reactions with acid and base catalysts in hot compressed water at 473 K, Carbohydr. Res.
340 (2005) 1925–1930. doi:10.1016/j.carres.2005.06.017.
[7] C. Liu, J.M. Carraher, J.L. Swedberg, C.R. Herndon, C.N. Fleitman, J.-P. Tessonnier,
Selective Base-Catalyzed Isomerization of Glucose to Fructose, ACS Catal. 4 (2014)
4295–4298. doi:10.1021/cs501197w.
[8] N. Deshpande, L. Pattanaik, M.R. Whitaker, C.-T. Yang, L. Lin, N.A. Brunelli,
Selectively converting glucose to fructose using immobilized tertiary amines, J. Catal. 353
(2017) 205–210. doi:10.1016/j.jcat.2017.07.021.
[9] N. Rajabbeigi, A.I. Torres, C.M. Lew, B. Elyassi, L. Ren, Z. Wang, H. Je Cho, W. Fan, P.
Daoutidis, M. Tsapatsis, On the kinetics of the isomerization of glucose to fructose using
Sn-Beta, Chem. Eng. Sci. 116 (2014) 235–242. doi:10.1016/j.ces.2014.04.031.
16

[10] C. Moreau, R. Durand, A. Roux, D. Tichit, Isomerization of glucose into fructose in the
presence of cation-exchanged zeolites and hydrotalcites, Appl. Catal., A. 193 (2000) 257.
doi:10.1016/S0926-860X(99)00435-4.
[11] N. Yoshinaga, S. Aomine, Allophane in some ando soils, Soil Sci. Plant Nutr. 8 (1962) 6–
13. doi:10.1080/00380768.1962.10430983.
[12] V.C. Farmer, A.R. Fraser, J.M. Tait, Synthesis of imogolite: a tubular aluminium silicate
polymer, J. Chem. Soc. Chem. Commun. (1977) 462. doi:10.1039/c39770000462.
[13] S.M. Barrett, P.M. Budd, C. Price, The synthesis and characterization of imogolite, Eur.
Polym. J. 27 (1991) 609–612. doi:10.1016/0014-3057(91)90144-D.
[14] S. Mukherjee, V. Bartlow, S. Nair, Phenomenology of the growth of single-walled
aluminosilicate and aluminogermanate nanotubes of precise dimensions, Chem. Mater. 17
(2005) 4900–4909. doi:10.1021/cm0505852.
[15] D.-Y. Kang, J. Zang, C.W. Jones, S. Nair, Single-Walled Aluminosilicate Nanotubes with
Organic-Modified Interiors, J. Phys. Chem. C. 115 (2011) 7676–7685.
doi:10.1021/jp2010919.
[16] P.D.G. Cradwick, V.C. Farmer, J.D. Russell, C.R. Masson, K. Wada, N. Yoshinaga,
Imogolite, a Hydrated Aluminium Silicate of Tubular Structure, Nat. Phys. Sci. 240
(1972) 187–189. doi:10.1038/physci240187a0.
[17] W.C. Ackerman, D.M. Smith, J.C. Huling, Y.W. Kim, J.K. Bailey, C.J. Brinker, Gas
Vapor Adsorption in Imogolite - a Microporous Tubular Aluminosilicate, Langmuir. 9
(1993) 1051–1057.
[18] X. Yang, Z. Sun, D. Wang, W. Forsling, Surface acid-base properties and
hydration/dehydration mechanisms of aluminum (hydr)oxides, J. Colloid Interface Sci.
308 (2007) 395–404. doi:10.1016/j.jcis.2006.12.023.
[19] A.C. Vieira Coelho, H. De Souza Santos, P.K. Kiyohara, K.N.P. Marcos, P. De Souza
Santos, Surface area, crystal morphology and characterization of transition alumina
powders from a new gibbsite precursor, Mater. Res. 10 (2007) 183–189.
doi:10.1590/S1516-14392007000200015.
17

[20] S. Wada, K.O.J. Wada, Effects of Substiution of Germanium for Silicon in Imogolite,
Clays Clay Miner. 30 (1982) 123–128.
[21] A. Thill, P. Maillet, B. Guiose, O. Spalla, L. Belloni, P. Chaurand, M. Auffan, L. Olivi,
J.C. Rose, Physico-chemical control over the single- or double-wall structure of
aluminogermanate imogolite-like nanotubes., J. Am. Chem. Soc. 134 (2012) 3780–6.
doi:10.1021/ja209756j.
[22] M. Ookawa, M.W.A.J. Anabe, M. Suzuki, Synthesis and Characterization of Fe
Containing Imogolite, Energy. 284 (2006) 280–284.
doi:10.11362/jcssjclayscience1960.12.Supplement2_280.
[23] Y. Kuroda, K. Fukumoto, K. Kuroda, Uniform and high dispersion of gold nanoparticles
on imogolite nanotubes and assembly into morphologically controlled materials, Appl.
Clay Sci. 55 (2012) 10–17. doi:10.1016/j.clay.2011.07.004.
[24] G. Ipek Yucelen, R.E. Connell, J.R. Terbush, D.J. Westenberg, F. Dogan, Synthesis and
immobilization of silver nanoparticles on aluminosilicate nanotubes and their antibacterial
properties, Appl. Nanosci. 6 (2015) 607–614. doi:10.1007/s13204-015-0467-x.
[25] N. Jiravanichanun, K. Yamamoto, K. Kato, J. Kim, S. Horiuchi, W.O. Yah, H. Otsuka, A.
Takahara, Preparation and characterization of imogolite/DNA Hybrid Hydrogels,
Biomacromolecules. 13 (2012) 276–281. doi:10.1021/bm201616m.
[26] S. Imamura, T. Kokubu, T. Yamashita, Y. Okamoto, K. Kajiwara, H. Kanai, Shape-
selective copper-loaded Imogolite catalyst, J. Catal. 160 (1996) 137–139.
doi:10.1006/jcat.1996.0132.
[27] S. Imamura, Y. Hayashi, K. Kajiwara, H. Hoshino, C. Kaito, Imogolite: a possible new
type of shape-selective catalyst, Ind. Eng. Chem. Res. 32 (1993) 600–603.
doi:10.1021/ie00016a005.
[28] E. Bahadori, V. Vaiano, S. Esposito, M. Armandi, D. Sannino, B. Bonelli, Photo-activated
degradation of tartrazine by H2O2as catalyzed by both bare and Fe-doped methyl-
imogolite nanotubes, Catal. Today. 304 (2018) 199–207.
doi:10.1016/j.cattod.2017.08.003.
18

[29] B. Bonelli, I. Bottero, N. Ballarini, S. Passeri, F. Cavani, E. Garrone, IR spectroscopic and
catalytic characterization of the acidity of imogolite-based systems, J. Catal. 264 (2009)
15–30. doi:10.1016/j.jcat.2009.03.003.
[30] B. Bonelli, C. Zanzottera, M. Armandi, S. Esposito, E. Garrone, IR spectroscopic study of
the acidic properties of alumino-silicate single-walled nanotubes of the imogolite type,
Catal. Today. 218–219 (2013) 3–9. doi:10.1016/j.cattod.2013.02.004.
[31] D.-Y. Kang, N.A. Brunelli, G.I. Yucelen, A. Venkatasubramanian, J. Zang, J. Leisen, P.J.
Hesketh, C.W. Jones, S. Nair, Direct synthesis of single-walled aminoaluminosilicate
nanotubes with enhanced molecular adsorption selectivity, Nat. Commun. 5 (2014) 1–9.
doi:10.1038/ncomms4342.
[32] C. Levard, A. Masion, J. Rose, E. Doelsch, D. Borschneck, L. Olivi, P. Chaurand, C.
Dominici, F. Ziarelli, A. Thill, P. Maillet, J.Y. Bottero, Synthesis of Ge-imogolite:
influence of the hydrolysis ratio on the structure of the nanotubes, Phys. Chem. Chem.
Phys. 13 (2011) 14516–14522. doi:10.1039/c1cp20346k.
[33] B. Bonelli, Surface Chemical Modifications of Imogolite, 1st ed., Elsevier Ltd., 2016.
doi:10.1016/B978-0-08-100293-3.00012-1.
[34] S. Konduri, S. Mukherjee, S. Nair, Controlling nanotube dimensions: Correlation between
composition, diameter, and internal energy of single-walled mixed oxide nanotubes, ACS
Nano. 1 (2007) 393–402. doi:10.1021/nn700104e.
[35] S.G. Wettstein, D. Martin Alonso, E.I. Gurbuz, J.A. Dumesic, A roadmap for conversion
of lignocellulosic biomass to chemicals and fuels, Curr. Opin. Chem. Eng. 1 (2012) 218–
224. doi:10.1016/j.coche.2012.04.002.
[36] M.R. Whitaker, A. Parulkar, R. Joshi, N.A. Brunelli, Examining Acid Formation During
the Selective Dehydration of Fructose to 5-Hydroxymethylfurfural in DMSO and Water,
(In Prep.). (n.d.).
[37] M. Badlani, I.E. Wachs, Methanol: A “smart” chemical probe molecule, Catal. Letters. 75
(2001) 137–149. doi:10.1023/A:1016715520904.
[38] D. Kang, J. Zang, E.R. Wright, A.L. Mccanna, C.W. Jones, S. Nair, Dehydration,
19

Dehydroxylation, and Rehydroxylation of Single-Walled Aluminosilicate Nanotubes,
ACS Nano. 4 (2010) 4897–4907.
[39] J.M. de Bruijin, A.P.G. Kieboom, H. van Bekkum, Alkaline degradation of
monosaccharides V: Kinetics of the alkaline isomerization and degradation of
monosaccharides, Recl. Trav. Chim. Pays-Bas. 106 (1987) 35–43.
[40] J.M. de Bruijin, A.P.G. Kieboom, H. van Bekkum, Alkaline Degradation of
Monosaccharides. Part VII. A Mechanistic Picture, Starch-Stärke. 39 (1987) 23–28.
20

Scheme 1. Isomerization of glucose to fructose. Cross-section of an aluminosilicate imogolite
nanotube with average inner and outer diameters labeled [15].
Me-1.0-IMO
Me-0.5-IMO
Ge-0.80-IMO
Ge-0.3-IMO
AlSi IMO
10
20
30
40
50
2Q
Figure 1. Powder X-ray diffraction pattern of AlSi imogolite (black) compared to Ge-IMO of 30%
(blue) and 80% (red) and Me-IMO of 50% (green) and 100% (purple).
Table 1. Nitrogen physisorption and methanol pulse experiment results for imogolite. The surface area
BET surface areas and HK pore widths of the AlSi IMO are compared to the Me and Ge modified
materials. Sites per surface area and per gram determined by methanol pulse experiements.
Sites per
Surface Area
(mol/m2)
2.5
Sites per
Gram
(mmol/g)
BET Surface
Area (m2/g)
HK Pore
Width (nm)
Material
IMO-R
273
0.88
0.66
21

IMO-AB
288
456
366
229
253
0.88
0.94
1.10
0.91
0.95
-
0.98
0.28
-
-
0.40
0.10
-
Me-0.5-IMO
Me-1.0-IMO
Ge-0.3-IMO
Ge-0.8-IMO
-
-
50
40
30
20
10
0
90
80
70
60
50
40
30
20
10
Conversion
Selectivity
0
5
10 15 20 25 30 35 40 45 50
Time (h)
Figure 2. Kinetic evaluation of IMO-R over 48 hours. Kinetic evaluation of IMO-R over 48 hours.
Data points are collected at 2, 6, 16, 20, 24, and 48 hours. The test is performed using 100 mg of
IMO at 100 °C with 1 g of 10 wt% glucose solution. The glucose conversion is in blue (left axis)
and fructose selectivity is in red (right axis).
22

80
60
40
20
0
AlSi IMO
IMO-AB
IMO-R
IMO N2
Ge-0.3-IMO
Ge-0.8-IMO
Me-0.5-IMO
Me-1.0-IMO
KOH
0
10
20
Conversion (%)
30
40
Figure 3. A comparison of the conversion and selectivity for different imogolite catalysts
compared to KOH [39,40]. The reaction conditions are 100 mg of imogolite catalyst per 1 gram of
a 10 wt% glucose solution in air at a reaction temperature of 100°C. IMO N2 corresponds to using
the IMO catalyst under nitrogen. For the IMO catalysts, a dashed line is plotted to show trend of
conversion with increasing selectivity.
23

100
90
80
70
60
50
40
IMO
IMO (after 3 reactions)
IMO Base Wash
100 200 300 400 500 600 700 800 900
Temperature (°C)
Figure 4. TGA mass percent curve of as-synthesized IMO (black), the recycled IMO after three
reactions (green), and the IMO catalyst after washing with a 1 M NH4OH solution (red).
24