Tuning the location of niobia/carbon composites in a biphasic reaction: Dehydration of d-glucose to 5-hydroxymethylfurfural

Article abstract of DOI:10.1007/s10562-013-1004-8

The conversion of glucose to 5-hydroxymethylfurfural (HMF) was studied in a biphasic system with niobia solid acid catalysts. We used carbon supported niobia since it shows improved hydrothermal stability. However, the carbon supported catalysts tend to be located in the organic phase due to their hydrophobic nature, while the reactant glucose is present in the aqueous phase. Therefore, we functionalized the carbon black support to increase the degree of hydrophilicity, allowing us to locate the catalyst either in the organic phase, in the aqueous phase or at the interface. The effect of the location of the niobia/carbon composites on the catalytic performance was investigated in the conversion of d-glucose to HMF in a biphasic system. It was found that the niobia catalyst located in the aqueous phase showed the highest d-glucose conversion (twofold greater than the others). We also compared the performance of catalysts obtained by deposition of niobia on functionalized carbon with a catalyst prepared by deposition precipitation-carbonization (DPC). The latter catalyst provides crystalline niobia particles while the other catalysts yield amorphous niobia. The crystalline niobia is more active for the acid catalyzed dehydration of isopropanol in the vapor phase. The DPC method provides a simple one pot synthesis to generate niobia particles embedded in a carbon support.

Full text of DOI:10.1007/s10562-013-1004-8

Catal Lett (2013) 143:509–516
DOI 10.1007/s10562-013-1004-8
Tuning the Location of Niobia/Carbon Composites in a Biphasic
Reaction: Dehydration of D-Glucose to 5-Hydroxymethylfurfural
Haifeng Xiong
Abhaya K. Datye
•
Tianfu Wang
•
Brent H. Shanks
•
Received: 14 February 2013 / Accepted: 31 March 2013 / Published online: 10 April 2013
Ó Springer Science+Business Media New York 2013
Abstract The conversion of glucose to 5-hydroxymeth-
ylfurfural (HMF) was studied in a biphasic system with
niobia solid acid catalysts. We used carbon supported
niobia since it shows improved hydrothermal stability.
However, the carbon supported catalysts tend to be located
in the organic phase due to their hydrophobic nature, while
the reactant glucose is present in the aqueous phase.
Therefore, we functionalized the carbon black support to
increase the degree of hydrophilicity, allowing us to locate
the catalyst either in the organic phase, in the aqueous
phase or at the interface. The effect of the location of the
niobia/carbon composites on the catalytic performance was
investigated in the conversion of D-glucose to HMF in a
biphasic system. It was found that the niobia catalyst
located in the aqueous phase showed the highest D-glucose
conversion (twofold greater than the others). We also
compared the performance of catalysts obtained by depo-
sition of niobia on functionalized carbon with a catalyst
prepared by deposition precipitation-carbonization (DPC).
The latter catalyst provides crystalline niobia particles
while the other catalysts yield amorphous niobia. The
crystalline niobia is more active for the acid catalyzed
dehydration of isopropanol in the vapor phase. The DPC
method provides a simple one pot synthesis to generate
niobia particles embedded in a carbon support.
Keywords Niobia/carbon acid catalyst Á Biphasic
reaction Á Hydrophilic carbon Á D-glucose conversion to
5-hydroxymethylfurfural
1 Introduction
5-hydroxymethylfurfural (HMF) has been suggested as an
important biomass-derived chemical intermediate for the
production of polymers, fine chemicals and transportation
fuels [1–5]. The production of HMF from dehydration of
D-glucose is a promising route because D-glucose is an
abundant monosaccharide. The conversion of D-glucose to
HMF carried out in a single phase (aqueous phase or ionic
liquids [6, 7]) leads to a low HMF selectivity because HMF
can undergo further reactions to form byproducts under the
reaction conditions. Use of a biphasic system of two
immiscible solvents (often water and a hydrophobic
organic liquid), can avoid the aforementioned issue
because HMF produced in the aqueous phase can migrate
into the organic phase to avoid the further reaction [8].
Homogeneous Lewis acid catalysts have been extensively
used in the conversion of D-glucose to HMF in biphasic
reaction systems [9–11]. However, these homogeneous
catalysts have several drawbacks including the production
of waste, difficulty in separation from products and the
recovery and reuse of the catalysts [12]. Thus, it is
important to develop a heterogeneous acid catalyst for this
reaction [13].
H. Xiong Á A. K. Datye (&)
Department of Chemical and Nuclear Engineering and Center
for Micro-engineered Materials, University of New Mexico,
Albuquerque, NM 87131, USA
Recently, several studies have explored the use of het-
erogeneous catalysts in the conversion of glucose to HMF
e-mail: datye@unm.edu
[
9, 12–14]. For example, in conjunction with a Brønsted
acid, tin-containing silica has been proved to catalyze the
conversion of glucose [15]. In ionic liquid–water mixtures,
T. Wang Á B. H. Shanks
Chemical and Biological Engineering, Iowa State University,
Ames, IA 50011, USA
ZrO was found to promote the formation of HMF with the
2
123

5
10
H. Xiong et al.
-
addition of Cl or HSO4 anions [6]. Thus, it was noted
-
functionalized carbon black. Subsequently, the temperature
was raised to 90 °C. After allowing sufficient time (17 h) for
the hydrolysisof the urea, the sample was filteredand washed
with deionized water, followed by drying at 100 °C for 10 h
that the addition of some salts or ions such as HCl, CrCl or
2
-
HSO4 can increase the yield of HMF. Without the addi-
tives, mesoporous TiO nanospheres produced HMF with
2
selectivity range from 4–30 % [14]. Typical solid acid
catalysts are metal oxide based and they are not stable
under hydrothermal conditions in the aqueous phase,
leading to loss of surface area and reactivity [16]. Our
recent findings indicated that carbon-loaded oxide com-
posites show improved hydrothermal stability [17].
and calcining at 250 °C for 4 h in a flow of N . The samples
2
were denoted as Nb/CB-1-DP and Nb/CB-2-DP (CB stands
for carbon black and DP stands for deposition precipitation).
Nb/CS-HT A niobia/carbon catalyst (nominal 10 wt%
niobium loading) was prepared by a method we have called
deposition precipitation-carbonization (DPC). The synthe-
sis involved the use of 4 g of D-glucose dissolved in
250 mL dionized water and mixed with ammonium nio-
bium oxalate (0.6724 g) and urea (0.2826 g). The mixture
was placed in an autoclave and held at 200 °C for 12 h.
The sample was pyrolyzed at 400 °C for 4 h in a flow of N2
and denoted as Nb/CS-HT (CS stands for carbon spheres
and HT stands for hydrothermal synthesis).
In general, pure carbon materials are hydrophobic [8, 18,
1
9], so, they partition into the organic phase in a biphasic
reaction system. The hydrophobic surface of carbon mate-
rials can be tailored by introducing functional groups or
oxides, so that the carbon materials could be located at the
interface with the aqueous phase or within the aqueous phase.
The different locations of carbon materials in a biphasic
system can influence the performance of these catalysts,
which is the focus of this study. We describe methods to tune
the location of the carbon materials in biphasic solvents.
Niobium oxide is a solid acid catalyst that can be used
extensively in important biomass reactions, such as dehy-
dration, aldol condensation, hydrolysis and ketonization.
Recently, pure bulk niobia was shown to be active for the
dehydration of glucose to 5-(hydroxymethyl)furfural in the
aqueous phase with a HMF yield of 20 % [12]. In this
study, we generated nanostructured niobia/carbon com-
posites whose properties were tailored for applications in
biphasic reactions. By controlling the degree of function-
alization, the niobia/carbon composites were located in
different locations in the water/oil biphasic solvent system
2.2 Catalyst Characterization
N adsorption/desorption isotherms were recorded using a
2
Quantachrome Autosorb-1 instrument. Prior to the experi-
ment, the sample was out gassed at 200 °C for 6 h. The
surface area was obtained using the BET method using
adsorption data over a relative pressure range from 0.05 to
0.30. The total pore volumes were calculated from the
amount of N vapor adsorbed at a relative pressure of 0.99.
2
Scanning electron microscopy (SEM) was performed on a
Hitachi S-5200, with a resolution of 0.5 nm at 30 kV and 1.7
nm at 1 kV. Scanning transmission electron microscopy
(STEM) was carried out in a JEOL 2010F microscope. The
powders were deposited on holey carbon support films after
being dispersed in ethanol. An electron probe, diameter of
0.2 nm, was scanned over the specimen, and electrons
scattered at high angles were collected to form the images.
The image contrast in the HAADF (high angle annular dark
field) mode is atomic number dependent and is dependent
also on the sample thickness in each pixel being imaged.
Thermogravimetric analyses (TGA) was performed with a
SDT Q600 TGA using nitrogen or air as the purge gas and a
heating rate of 10 °C/min. The flow rate of purge gas was
always 50 mL/min. Fourier transform infrared spectra
(FTIR) were recorded on a Thermo Nicolet 6700 FTIR
(
water and sec-butyl phenol). The effect of catalyst loca-
tion on the catalytic performance was investigated in the
conversion of D-glucose to 5-hydroxymethylfurfural
(
HMF) in this biphasic system.
2
Experimental
2
.1 Niobia/Carbon Catalyst Preparation
Nb/CB-DP Commercially-available carbon black (CABOT,
Vulcan XC 72R) was pretreated in 50 % HNO at 80 and
3
-
1
1
20 °C for 8 h, respectively. After filtration and washing by
spectrometer, at a spectral resolution of 4 cm , using 64
scans per spectrum.
water till pH = 7, the obtained material was dried at 120 °C
for 12 h and denoted as CB-1 and CB-2, respectively. Two
niobia/carbon catalysts (nominally 10 wt% niobium load-
ing) were prepared by homogeneous deposition precipitation
2.3 Catalytic Activity and Hydrothermal Stability
(
DP) using the pretreated carbon supports. Urea was used as
2.3.1 Isopropanol Dehydration
the precipitating agent and the detailed synthesis route was as
follows: ammonium niobium oxalate (0.41 g) and urea
Isopropanol dehydration was carried out in a fixed-bed flow
reactor as a general probe of the acidic properties of the cat-
alysts. A mass of ca. 20 mg of catalyst was used. The reactor
(
0.27 g; 2.5 mol urea per mole of niobium) were dissolved
in deionized water (250 mL) and added to 1 g of the
1
23

Tuning the Location of Niobia/Carbon Composites
511
was purged by flowing ultra high purity (UHP) Ar (ca.
and CB-2). As can be seen, the as-received carbon black
shows spherical particles of graphitic carbon (Fig. 1a).
After acid treatment at 80 °C we do not see much change
in the morphology (Fig. 1b). High-resolution TEM shows
that the carbon spheres are composed of curved graphitic
sheets within the spherical particles (Fig. 1c). However,
after acid treatment at 120 °C, the carbon black morphol-
ogy changed significantly. The TEM images indicate that
the primary particles are now hollow, which can explain
the increased pore volume of the functionalized carbon
(Fig. 1d). The observed hollow carbon black structure is a
result of the poor degree of graphitization of the interior
making it more susceptible to etching when the carbon is
treated with nitric acid. The harsher acid treatment creates
functional groups transforming the carbon surface from
being hydrophobic to becoming hydrophilic so that the
carbon can be easily dispersed in polar solvents, such as
ethanol.
-1
2
5 mL min ) for 30 min and then the temperature was
increased to 180 °C. The reaction was performed at 180 °C
and atmospheric pressure with Ar (ca. 25 mL min ) as car-
-1
-1
rier gas and isopropanol (0.002 mL min ) was pumped to
the reactor by a HPLC pump (Eldex). Reactants and products
were analyzed with an on-line GC (Varian 3800) equipped
with a capillary column (Porapak-T) and a FID detector.
2
.3.2 Conversion of Glucose to 5-Hydroxymethylfurfural
HMF)
(
A biphasic reaction system (water and sec-butyl phenol) was
used to study the conversion of D-glucose to 5-hydroxym-
ethylfurfural [20]. The mass ratio of sec-butyl phenol (SBP)
solvent to water was 2:1. Here, SBP is an extracting solvent
for HMF. The advantage of using SBP is that higher HMF
yields can be obtained, as compared to other organic sol-
vents such as tetrahydrofuran (THF), methyl isobutyl ketone
Representative SEM/STEM images of niobia/carbon
catalysts prepared by deposition precipitation are shown in
Fig. 2. The SEM image of Nb/CB-1-DP shows niobia
particles with an average size of 20 nm dispersed on the
carbon black surface (Fig. 1a). The STEM image shows
that the niobia clusters of ca. 20 nm are actually composed
of smaller niobia particles with an average size of 3 nm
(Fig. 2b). The STEM image of Nb/CB-2-DP shows niobia
clusters around 20 nm in diameter, hence there is no sig-
nificant difference in niobia particle morphology (Fig. 2c)
compared to Nb/CB-1-DP. The HR-TEM image of Nb/CB-
2-DP shows that the niobia particles are amorphous
(Fig. 2d, inset).
(
MIBK), and 2-butanol [20, 21]. The reaction was carried
out in an Alltech reactor. 0.1 g catalyst was used and the
reaction temperature is 170 °C and reaction time is 120 min.
After reaction, liquid effluents were collected and compo-
sitions were quantified in a Waters 1525 HPLC system
equipped with a 2998 PDA UV detector and a 2414
refractive index detector maintained at 60 °C. Aqueous
phase samples were analyzed using a PL Hi-Plex H-form
carbohydrate column at 80 °C, using 5 mM H SO as the
2
4
-
1
mobile phase at a flow rate of 0.6 mL min . Organic phase
samples were analyzed using a Zorbax SB-C18 reverse
phase column (Agilent) at 35 °C, with an methanol:water
-
1
(
8:2 v/v) mixed solvent at a flow rate of 0.7 mL min
.
Representative SEM/STEM images of Nb/CS-HT pre-
pared by the DPC method are shown in Fig. 3. The method
has been described in more detail elsewhere [17]. As
shown in Fig. 3a, the Nb/CS-HT composite shows a
spherical morphology with a diameter of 20–50 nm. These
spheres can be easily crushed yielding a powder that con-
tains niobia particles with a size of ca. 8 nm (Fig. 3b). The
high-resolution TEM image of Nb/CS-HT shows lattice
fringes (Fig. 3b, inset) indicating that the niobia particles
are crystalline, which is different from the amorphous
niobia seen on both Nb/CB-1-DP and Nb/CB-2-DP. The
synthesis of this catalyst involved a higher temperature
(200 °C) under hydrothermal conditions (22 bar) which
causes the niobia to become crystalline.
3
Results and Discussion
3
.1 N Physisorption
2
Table 1 displays the N physisorption results for the niobia/
2
carbon composites. As can be seen, all the niobia/carbon
2
catalysts showed a surface area [100 m /g. Compared to
the as-received carbon black, the Nb/CB-2-DP and Nb/CB-
1
-DP samples show higher pore volumes as a result of the
reaction with nitric acid at elevated temperatures. The
increase in pore volume is due to the formation of macro
pores due to the etching of the carbon, but this also leads to
a loss of surface area from Nb/CB-1-DP to Nb/CB-2-DP
likely due to a loss of the micropores.
3.3 Fourier Transform Infrared Spectroscopy (FTIR)
Figure 4 shows the FTIR spectra of the three niobia/carbon
catalysts. For Nb/CB-1-DP and Nb/CB-2-DP, two bands at
3
.2 Electron Microscopy
-
,693 and 1,510 cm were observed in the FTIR spectra
1
1
-
1
Figure 1 shows the TEM images of functionalized carbon
black with different extents of nitric acid treatment (CB-1
as well as an intense band at 1,800–2,600 cm (Fig. 4a,
b). These are ascribed to the formation of functional
123

5
12
H. Xiong et al.
Table 1 BET surface area, pore volume of the niobia/carbon catalysts
2
BET surface area (m /g)
3
Total pore volume (cm /g)
a
Sample
Nb
–
O
2 5
loading (wt%)
Carbon black (CB)
b
Nb/CB-1-DP
169
0.11
0.25
0.76
0.48
175.2
149.2
102.9
6.9
6.5
8.9
Nb/CB-2-DP
c
Nb/CS-HT
a
2 5
Nb O loading was determined by burning off the carbon in a TGA in air
b
c
CB-1 and CB-2 are carbon black functionalized by nitric acid at 80 °C and 120 °C, respectively
CS-HT: carbon prepared by hydrothermal synthesis [17]
Fig. 1 TEM images of carbon
materials: a as-received carbon
black; b carbon black
functionalized in nitric acid at
8
0 °C; c HR-TEM image of
carbon black functionalized in
nitric acid at 80 °C; d carbon
black functionalized in nitric
acid at 120 °C
oxygenate groups on the carbon surface. The bands at
both Nb/CB-1-DP and Nb/CB-2-DP. The FTIR spectrum
of Nb/CS-HT shows three bands at 1,705, 1,590 and
-
,693 and 1,510 cm are ascribed to the C=O and C=C
1
1
-
1,230 cm and no band was found at 1,800–2,800 cm
1
-1
bands, respectively [19, 22] and the bands at
.
-
,800–2,600 cm are ascribed to the C–O or C=O species.
1
-1
The bands at 1,705 and 1,590 cm are ascribed to the
C=O and C=C bands, respectively [23]. The band at
1
Further, although the Nb/CB-2-DP shows a similar FTIR
spectrum to Nb/CB-1-DP, the former shows more intense
peaks than the latter, this indicates that more functional
groups are present on the Nb/CB-2-DP catalyst surface. It
is interesting to note that the FTIR spectrum of Nb/CS-HT
exhibits completely different bands, in comparison with
-
1
1,230 cm is ascribed to the C–C band [24]. Thus, the
FTIR results revealed that a larger amount of functional
groups (C=O, C–O and C=C species) were contained on
the Nb/CB-1-DP and Nb/CB-2-DP surface and different
groups (C=C, C=O and C–C species) were present on the
1
23

Tuning the Location of Niobia/Carbon Composites
513
Fig. 2 SEM/STEM images of
Nb/CB-DP prepared by
deposition precipitation: a and b
Nb/CB-1-DP; c Nb/CB-2-DP; d
HR-TEM image of Nb/CB-2-
DP showing that the niobia
particles are amorphous (red
box)
Fig. 3 Electron microscopy of
Nb/CS-HT prepared by the DPC
method at 200 °C for 12 h: a
SEM image showing a wide
area view; b STEM image
showing clearly the niobia
particles and HR-TEM image
showing the lattice fringes of
niobia (insert)
Nb/CS-HT surface. This difference can be related to the
method of preparation, acid treatment in the case of the CB
samples compared with hydrothermal transformation of a
sugar followed by pyrolysis in the case of the CS-HT
sample.
3.4 TGA Results
The thermal stability of the niobia/carbon catalysts were
studied in a flow of N and shown in Fig. 5. For Nb/CB-1-
DP, \10 % weight loss was observed when the sample was
2
123

5
14
H. Xiong et al.
propylene formation rate range of 1.7–2.4 lmol -
-
1
-1
min
g
Nb. After treatment in liquid water at 200 °C for
2 h, the activity of all catalysts decreased by about 50 %.
a
60
40
20
0
1
The higher activity of the Nb/CS-HT catalyst is related to
the increased crystallinity of the niobia in this sample, as
detected via HR-TEM.
b
-1
-1
1693 cm
1510 cm
3.5.2 Conversion of D-Glucose to 5-Hydroxymethylfurfural
c
(HMF) in the Biphasic System
-
1
The catalytic behavior and the effect of different locations
of the niobia/carbon catalysts were investigated in the
conversion of D-glucose to 5-hydroxymethylfurfurl in a
biphasic system (water and sec-butyl phenol). The biphasic
system allows the SBP solvent to continuously extract the
HMF produced from the aqueous phase into the organic
phase, thereby minimizing HMF degradation. We first
investigated the locations of the niobia/carbon catalysts in
the biphasic system. As can be seen in Fig. 6, the Nb/CB-1-
DP was present exclusively in the organic phase (Fig. 6a),
while the Nb/CB-2-DP was entirely present in the aqueous
phase (Fig. 6c). Because both these catalyst supports are
functionalized by nitric acid but at different temperatures
and the catalysts are prepared by the same method, the
different locations are due to the different degree of
hydrophobicity/hydrophilicity on the carbon surface. It is
interesting to note that the Nb/CS-HT catalyst was located
at the interface between the aqueous and organic phase
suggesting it has an intermediate degree of hydrophobicity
(Fig. 6b).
1
705 cm
-1
1
590 cm
-
1
1
230 cm
2800
2400
2000
1600
1200
-
1
Wavenumbers (cm )
Fig. 4 FTIR spectra of the niobia/carbon catalysts: a Nb/CB-1-DP; b
Nb/CB-2-DP; c Nb/CS-DP
heated to 900 °C (Fig. 5a). The DTG curve shows that four
weight loss regions are observed and centered at 60, 250,
4
50 and 700 °C, respectively. The Nb/CB-2-DP catalyst
shows a similar weight loss behavior, while a weight loss
of 25 % was found when the temperature was increased to
9
00 °C (Fig. 5b). According to our previous study [19], the
peak at 60 °C was ascribed to the release of adsorbed water
in the sample and the peak between 300 and 600 °C
assigned to the removal of some functional groups on the
carbon sample. Thus, the Nb/CB-2-DP exhibited signifi-
cant greater amounts of surface functional groups, as
compared to Nb/CB-1-DP. This agrees well with the FTIR
result. For the Nb/CS-HT prepared by DPC, two weight
The catalytic reactivity of the niobia/carbon catalysts are
shown in Table 3. For the niobia catalysts supported on
carbon black, the Nb/CB-1-DP shows a glucose conversion
of 34 % while the Nb/CB-2-DP shows the highest glucose
conversion of 78 %. The higher glucose conversion for Nb/
CB-2-DP can be related to the niobia/carbon catalyst and
D-glucose being present in the same phase (aqueous phase)
thus ensuring intimate contact of reactants and catalyst.
The Nb/CS-HT shows the lowest conversion and HMF
selectivity since it is located at the interface, leading to a
little contact between catalyst and glucose. The HMF yield
for the niobia/carbon composites are in the range of
10–20 %, which compares favorably with the 20 % HMF
selectivity reported recently on pure niobia catalysts [12].
The niobia/carbon composite developed in this work are
more hydrothermally stable than pure niobia [26]. More-
over, although both the above reactions are acid catalyzed,
the dehydration reactivity of niobia/carbon catalysts in the
liquid phase does not scale with the dehydration of iso-
propanol in the vapor phase. This is because we are con-
ducting a biphasic reaction and the location of the catalyst,
in the aqueous phase, organic phase or the interface, played
a bigger role than the intrinsic catalytic activity.
loss peaks are found when it was heated in N (Fig. 5c).
2
The peak at 60 °C was ascribed to the release of adsorbed
water in the sample. The peak between 300 and 900 °C
assigned to the removal of some functional groups and the
decomposition of residual levulinic acid in the catalyst
based on NMR [25].
3
.5 Catalytic Performance and Hydrothermal Stability
3
.5.1 Isopropanol Dehydration in Gas-Phase Reactor
The gas-phase reaction of 2-propanol dehydration was used
to probe the nature of the surface acid sites in the niobia/
carbon solid acid catalysts. Further, a hydrothermal treat-
ment process in an autoclave at 200 °C for 12 h was used
to investigate the hydrothermal stability of the catalyst. The
results of 2-propanol dehydration at 180 °C for niobia/
carbon catalysts before and after hydrothermal treatment
are shown in Table 2. As can be seen, before the treatment
in liquid water, all niobia/carbon catalysts showed a
1
23

Tuning the Location of Niobia/Carbon Composites
515
2
Fig. 5 TGA and DTG data for niobia/carbon catalysts in N : a Nb/CB-1-DP; b Nb/CB-2-DP; c Nb/CS-HT
Table 2 Reactivity for the dehydration of 2-propanol before and
after hydrothermal treatment in liquid water at 200 °C for 12 h
of hydrophilicity, hence they are located in different
locations in the biphasic water/sec-butyl phenol system.
The niobia/carbon catalyst (Nb/CB-1-DP) prepared by DP
using a carbon black pretreated in nitric acid at 80 °C was
present in the organic phase, while the niobia/carbon cat-
alyst (Nb/CB-2-DP) prepared by DP using the carbon black
pretreated in nitric acid at 120 °C was mainly located in the
aqueous phase. The Nb/CS–HT was found to locate at
the interface between the organic and aqueous phases. The
different locations for the niobia/carbon catalysts were
found to significantly affect the catalytic behavior in the
conversion of D-glucose to 5-hydroxymethylfurfural
resulting from their different extent of contact with the
glucose. The Nb/CB-2-DP shows the highest D-glucose
conversion (twofold than the others dispersed in interface
or organic phase), which was ascribed to both niobia/car-
bon catalyst and D-glucose distributed in the same phase
(aqueous phase) that ensuring complete contact of reactant
and catalyst. The reactivity did not scale with the conver-
sion of isopropanol in the vapor phase, a reaction that
would probe the nature of the acid sites. This is because the
-
1
-1
g Nb)
Catalyst
Propylene formation rate (lmol min
Before treatment
in liquid water
After treatment
in liquid water
Nb/CS-HT
2.4
1.7
1.9
1.1
Nb/CB-1-DP
Nb/CB-2-DP
0.74
0.81
4
Conclusions
Nano-structured niobia/carbon catalysts were prepared by
two different preparation methods. The niobia/carbon black
catalysts prepared by deposition–precipitation (DP) show
niobia clusters of 20 nm, which are comprised of amor-
phous niobia particles of 3 nm. A niobia/carbon catalyst
(
Nb/CS–HT) prepared by deposition precipitation-carbo-
nization (DPC) shows crystalline niobia particles with a
size of 8 nm. The different catalysts have different degrees
123

5
16
H. Xiong et al.
Fig. 6 Photographs of the
locations of niobia/carbon
catalysts dispersed in sec-butyl
phenol/water (v/v = 2):
a organic phase; b interface;
c aqueous phase
Table 3 Performance of niobia/carbon catalysts in the conversion of
D-glucose to 5-hydroxymethylfurfural (HMF)
6. Qi X, Watanabe M, Aida TM, Smith RL (2012) Bioresource
Technol 109:224
7
. Zhao H, Holladay JE, Brown H, Zhang ZC (2007) Science
16:1597
. Crossley S, Faria J, Shen M, Resasco DE (2010) Science 327:68
. Wang J, Ren J, Liu X, Xi J, Xia Q, Zu Y, Lu G, Wang Y (2012)
Green Chem 14:2506
Catalyst
Conversion Selectivity Yield Location
(%)
3
(%)
(%)
8
9
Nb/CS-HT
33
34
52
26
11
18
20
Interface
Nb/CB-1-DP 34
Nb/CB-2-DP 78
Organic phase
Aqueous phase
10. Wang T, Pag a´ n-Torres Y, Combs E, Dumesic JA, Shanks B
2012) Top Catal 55:657
1. Huang R, Qi W, Su R, He Z (2010) Chem Commun 46:1115
(
1
Reaction conditions: 5wt % D-glucose in water saturated with NaCl,
reaction temperature of 170 °C, organic to aqueous mass ratio of two,
reaction time of 2 h
12. Nakajima K, Baba Y, Noma R, Kitano M, N. Kondo J, Hayashi S,
Hara M (2011) J Am Chem Soc 133:4224
13. Fan C, Guan H, Zhang H, Wang J, Wang S, Wang X (2011)
Biomass Bioenergy 35:2659
1
1
1
1
1
4. De S, Dutta S, Patra AK, Bhaumik A, Saha B (2011) J Mater
Chem 21:17505
5. Nikolla E, Rom a´ n-Leshkov Y, Moliner M, Davis ME (2011)
ACS Catal 1:408
6. Pham HN, Anderson AE, Johnson RL, Schmidt-Rohr K, Datye
AK (2012) Angew Chem Int Ed 51:13163
7. Xiong HF, Pham HN, Datye AK (2013) J Catal. doi:10.1016/
j.jcat.2013.03.007
location of the catalyst played a bigger role than intrinsic
catalytic activity. The results suggest that the acid sites on
the niobia catalysts are not ideally suited for the glucose
conversion reaction, and the ratio of Lewis to Bronsted
acid need to be further tuned to achieve optimal reactivity.
8. Rodr ´ı guez-reinoso F (1998) Carbon 36:159
Acknowledgments This work is supported by the NSF Engineering
Research Center for Biorenewable Chemicals (CBiRC) supported by
award number EEC 0813570. The XRD facilities used in this work
were supported by NSF MRI grant CBET 0960256. We thank
19. Xiong HF, Moyo M, Motchelaho MAM, Jewell LL, Coville NJ
(2010) Appl Catal A Gen 388:168
20. Pag a´ n-Torres YJ, Wang T, Gallo JMR, Shanks BH, Dumesic JA
(2012) ACS Catal 2:930
21. Rom a´ n-Leshkov Y, Dumesic JA (2009) Top Catal 52:297
Dr. Hien Pham for N
2
physisorption measurements.
2
2
2
2
2
2. Moyo M, Motchelaho MAM, Xiong HF, Jewell LL, Coville NJ
2012) Appl Catal A Gen 413–414:223
3. Titirici MM, Thomas A, Yu S-H, M u¨ ller J-O, Antonietti M
2007) Chem Mater 19:4205
4. Shin Y, Wang L-Q, Bae I-T, Arey BW, Exarhos GJ (2008) J Phys
Chem C 112:14236
5. Baccile N, Laurent G, Babonneau F, Fayon F, Titirici M–M,
Antonietti M (2009) J Phys Chem C 113:9644
6. Pag a´ n-Torres YJ, Gallo JMR, Wang D, Pham HN, Libera JA,
Marshall CL, Elam JW, Datye AK, Dumesic JA (2011) ACS
Catal 1:1234
(
References
(
1
2
. Davis SE, Zope BN, Davis RJ (2012) Green Chem 14:143
. Buntara T, Noel S, Phua PH, Meli a´ n-Cabrera I, de Vries JG,
Heeres HJ (2011) Angew Chem Int Ed 50: 7083
3
4
. Chidambaram M, Bell AT (2010) Green Chem 12:1253
. Huber GW, Chheda JN, Barrett CJ, Dumesic JA (2005) Science
3
08:1446
5
. Roman-Leshkov Y, Barrett CJ, Liu ZY, Dumesic JA (2007)
Nature 447:982
1
23