Alumina-promoted cobalt and iron xerogels as catalyst for the Fischer-Tropsch synthesis

Article abstract of DOI:10.1039/c3cy00483j

Sol-gel chemistry offers versatile new ways to prepare catalysts with tunable compositions and in different forms of application. In particular the option to obtain catalytic coatings is of interest for microreactors or the deposition of catalysts on monolithic supports. The present study explored a sol-gel approach for Fe and Co Fischer-Tropsch catalysts based on the so-called epoxide addition method. As a support and structural promoter, respectively, alumina was added. In the case of cobalt, the role of different precursors was studied, whereas in the case of iron, varying Fe-Al₂O₃ ratios were investigated. In both cases, ambient drying resulted in xerogels with high specific surface areas. The performance of the xerogels as catalysts for the Fischer-Tropsch (FT) reaction was studied in a fixed bed reactor and compared to cobalt- and iron-based reference catalysts synthesized by established impregnation and precipitation methods, respectively. All catalysts were carefully characterized in the as-prepared state as well as after catalysis with respect to their reducibility, stability and the distribution of the FT active component. Our study proves the suitability of the applied sol-gel technique to prepare highly active alumina-promoted Fe as well as alumina-supported Co FT catalysts and provides insight into the structure-performance relationships of these systems. The Royal Society of Chemistry.

Full text of DOI:10.1039/c3cy00483j

Catalysis
Science & Technology
View Article Online
View Journal | View Issue
PAPER
Alumina-promoted cobalt and iron xerogels as catalyst
for the Fischer–Tropsch synthesis
Cite this: Catal. Sci. Technol., 2013, 3,
256
3
M. Minnermann, B. Neumann, V. Zielasek and M. Bäumer*
Sol–gel chemistry offers versatile new ways to prepare catalysts with tunable compositions and in different forms
of application. In particular the option to obtain catalytic coatings is of interest for microreactors or the deposition
of catalysts on monolithic supports. The present study explored a sol–gel approach for Fe and Co Fischer–Tropsch
catalysts based on the so-called epoxide addition method. As a support and structural promoter, respectively,
alumina was added. In the case of cobalt, the role of different precursors was studied, whereas in the case of
2 3
iron, varying Fe–Al O ratios were investigated. In both cases, ambient drying resulted in xerogels with high specific
surface areas. The performance of the xerogels as catalysts for the Fischer–Tropsch (FT) reaction was studied in a
fixed bed reactor and compared to cobalt- and iron-based reference catalysts synthesized by established impreg-
nation and precipitation methods, respectively. All catalysts were carefully characterized in the as-prepared state
as well as after catalysis with respect to their reducibility, stability and the distribution of the FT active compo-
nent. Our study proves the suitability of the applied sol–gel technique to prepare highly active alumina-promoted
Fe as well as alumina-supported Co FT catalysts and provides insight into the structure–performance relationships
of these systems.
Received 8th July 2013,
Accepted 13th September 2013
DOI: 10.1039/c3cy00483j
www.rsc.org/catalysis
1. Introduction
stabilize the catalysts (Fe) or to enhance the dispersion of
1,2,6,7
the active component (Co).
The addition of a support is
Over the last years, the Fischer–Tropsch (FT) synthesis has
gained increasing attention again as an industrial process for
the production of synthetic fuels and chemicals as an alterna-
tive to the production of fuels from crude oil. Using either
other fossil feedstocks or feedstocks from renewable sour-
ces, the FT reaction is an integral part of the so-called XTL
especially important for cobalt-based catalysts because of
cobalt's higher material cost compared to iron. Therefore,
5,8
state of the art Co catalysts are prepared by impregnating
high surface area oxides, such as alumina or silica, with load-
9
ings of typically 15–30 wt% Co. Cobalt oxide nanoparticles
are generated in the pores of the impregnated oxide support
during calcination and turned into metal particles during a
subsequent reduction step. In contrast, iron's low cost allows
its direct use as catalyst. However, the addition of a struc-
tural promoter (SiO , Al O ) helps to disperse the iron phase
(X to liquid) processes converting syngas derived from natu-
ral gas (GTL), coal (CTL) or biomass (BTL) to liquid fuels. In
this way, high quality diesel, free of sulfur and aromatics,
can be produced over iron- or cobalt-based catalysts in the
so called low temperature process (LTFT). Both metals differ
in terms of activity, selectivity, and long-term stability. Their
advantages and disadvantages have been compared in several
reports considering the applied syngas feedstock, process
2
2 3
and to prevent sintering and attrition. Accordingly, e.g., the
1,5,10
2
Ruhrchemie iron LTFT catalyst is promoted by 20 wt% SiO .
Novel synthesis concepts for FT catalysts are still scarce in
industry. However, in the more fundamentally oriented litera-
ture new synthetic methods, such as colloidally prepared
nanoparticle catalysts, eggshell or monolithic catalysts have
been reported (see ref. 2, 11–16 and refrerences therein).
Among these techniques, sol–gel chemistry offers a particu-
larly rich field of possibilities to tune compositions and to
process the catalysts. In view of newly emerging ways to per-
form the FT synthesis in micro reactors or using monolithic
catalysts, the option to derive catalytic coatings by sol–gel che-
mistry is particularly appealing. For example, micro reactors or
monoliths can be covered by a porous catalyst layer by dip coa-
1
–5
conditions, the price etc.
For industrial applications, Fe- and Co-based FT catalysts
are still predominantly synthesized by established precipita-
tion and impregnation methods that are based on empirical
approaches and go back to compositions of early FT catalysts.
Fe- as well as Co-based catalysts contain different additives.
Besides chemical and reduction promoters, non-reducible
oxides are typically used as structural promoters to mechanically
Institute of Applied and Physical Chemistry, University of Bremen, Germany.
E-mail: mbaeumer@uni-bremen.de
17
ting with the sol or by pouring the sol through the channels.
3256 | Catal. Sci. Technol., 2013, 3, 3256–3267
This journal is © The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Paper
It is important to note, however, that the preparation tech-
nique has a strong influence on the catalytic properties.
the performance with the distribution and the reducibility of
the active sites as well as with the thermal stability of the
systems. Note that the synthesis was carried out in a similar
way for both metals; yet because the alumina content was gene-
rally very different for both systems (high for Co; low for Fe),
we will speak of “alumina supported Co” and “alumina
promoted Fe” in the following. The results in terms of activity
and selectivity are compared to reference catalysts synthe-
sized by impregnation (Co) and precipitation (Fe). Applying
the epoxide route, we will demonstrate that our catalysts
show FT activities similar to that of the reference catalysts
but exhibit a superior thermal stability.
2
,18–20
The performance is generally determined by a set of inter-
relating and sometimes conflicting factors, such as disper-
sion, reducibility and stability against sintering, oxidation or
coking. Therefore, a thorough characterization of catalysts pre-
pared by any new technique is essential.
In the present study, we followed such an approach explo-
ring the potential of a novel sol–gel chemistry route – the so-
called epoxide addition method (EAM) – for the preparation
of alumina iron and alumina cobalt FT catalysts. In contrast
to classical sol–gel methods based on alkoxides which are
limited only to a few transition metal oxides, cheap and easy-
21
to-handle metal salts can be applied in the EAM. Using this
technique, epoxides serve as gelation initiators of simple
hydrated metal salts dissolved in alcohol or water in a room
2. Experimental section
2.1 Catalyst preparation
21
temperature synthesis.
Mixed cobalt oxide/alumina xerogels with 22 wt% Co loading
as well as iron oxide/alumina xerogels with different weight
ratios were synthesized by the epoxide addition method follow-
Like other sol–gel techniques, the epoxide route allows to
adjust the size and shape of the particles forming an inter-
connected 3D network, and thus the specific surface area and
2
1
ing recipes reported Gash et al. In addition, pure alumina
xerogels were prepared for comparison. The mixed cobalt
oxide/alumina and pure alumina xerogels are denoted as Co–N,
Co–Cl, Al–N, or Al–Cl, depending on the precursor used
(N: nitrate; Cl: chloride, cf. Table 1). The notation of the iron
oxide/alumina xerogels (synthesized only from metal nitrates)
Fe : Al O 1 : X indicates the ratio of Fe : Al O by weight (cf. Table 2).
22,23
the porosity of the final material.
Moreover, mixed metal
oxide materials with uniform dispersions and variable stoichio-
24–28
metries have already been synthesized by this technique.
The products can be obtained as powders and monoliths or
29,30
can directly be coated onto a support (wafer, foam).
In
summary, the tunable compositional and structural proper-
ties from nanometer to millimeter range make these mate-
rials attractive for applications in catalysis, either as such or
as support for an active metal.
2
3
2
3
3 2 2
Cobalt nitrate hexahydrate (Co(NO ) ·6H O, Jansen Chimica,
p.a.), cobalt chloride hexahydrate (CoCl ·6H O, Riedel-de-
2
2
Haen, 99%), iron nitrate nonahydrate (Fe(NO ) ·9H O, Sigma-
3
3
2
To the best of our knowledge, the usage of the epo-
xide addition method as a one-step synthesis of alumina-
supported Co FT catalysts has never been reported in
literature so far. Other sol–gel techniques mainly based on
alkoxide precursors were already tested and controversially
discussed in the literature as summarized in ref. 2. Uniform
distributions on the support were achieved even for Co load-
3 3 2
Aldrich, 98%), aluminum nitrate nonahydrate (Al(NO ) ·9H O,
Roth, ≥98%) and aluminum chloride hexahydrate (AlCl ·6H O,
3
2
Alfa Aesar, 99%) were used as received as precursors. All reac-
tions were performed under ambient conditions.
Depending on the desired catalyst and composition, the
metal salts were dissolved in absolute ethanol (1.7 g EtOH/1
mmol metal salt, Sigma-Aldrich) under stirring using a poly-
ethylene vial. After complete dissolution, propylene oxide
(Acros Organic, 99.5%) as a gelation initiator was added. The
molar ratio of propylene oxide to the sum of the metal ions
was 11. The solution was briefly stirred to ensure homogene-
ity before the stir bar was removed. Within several minutes, a
gel formed and was allowed to age for 24 h. Afterwards etha-
nol was exchanged by decanting the old ethanol and adding
fresh ethanol which, in the following, was changed 3 times
in 3 days to remove residues from the synthesis. Xerogels
were obtained after air-drying of the resulting gels at room
temperature for at least 5 days in a fume hood. Pure and
Co-containing alumina xerogels were calcined at 400 °C
while xerogels containing iron were treated at 350 °C. The
31
ings higher than those of impregnated catalysts, but –
because of a stronger chemical interaction between Co and the
alumina as compared to impregnated catalysts – the accessi-
bility of Co in the support matrix and the reduction were
32
reported to be difficult.
Contrary to Co FT systems, pure iron oxide catalysts pre-
pared by the EAM have already been tested as FT catalysts by
3
3
Bali and co-workers. Their study was motivated by the
reported high specific surface area of iron oxide aero- and
xerogels synthesized by the epoxide route as compared to
2
1
2 3
other sol–gel derived Fe O solids. The performance of their
iron oxide aerogel – subsequently impregnated with potassium
as a chemical promoter – as catalyst for the FT reaction ini-
tially was found to be similar to that of a reference catalyst
exhibiting a similarly high surface area. However, the specific
surface area of the aerogel catalyst collapsed during operation
−
1
heating rate was 1 °C min and all xerogels were heated for
4 h in a static oven in air. Catalyst calcination tempe-
ratures were chosen based on TGA/DTA measurements of
the dried xerogels (above these temperatures no mass loss
was recorded).
(
35 h on stream) diminishing the reaction rates per gram of
1
catalyst significantly.
In our study, we investigated mixed Fe/Al and Co/Al
xerogels prepared by the EAM as FT catalysts and correlated
A Co reference catalyst with 22 wt% Co loading was prepared
by incipient wetness impregnation (catalyst notation: Co–IWI).
This journal is © The Royal Society of Chemistry 2013
Catal. Sci. Technol., 2013, 3, 3256–3267 | 3257

View Article Online
Paper
Catalysis Science & Technology
Table 1 BET surface areas after different catalyst treatments (a–d) and porosity measurements of the calcined catalysts. The xerogels are either prepared from
metal chloride (Cl) or by nitrate (N) salts
2
−1
BET specific surface area [m g ] after different treatments
Pore
Max. pore size
distribution
[nm]
a
a
volume
3
−1
Notation
a) Drying
b) Calcination
c) FT 430 °C
d) FT 800 °C
[cm g
]
Co–N
Co–Cl
Al–N
Al–Cl
Co–IWI
369
255
618
373
—
364
225
763
363
206
190
209
—
—
99
191
152
—
—
81
1.1
1.1
0.66
0.65
0.55
9
14
4
4
8
a
Calculated using Barrett–Joyner–Halenda (BJH) method; for calcined catalyst only.
2
−1
Briefly, aluminum oxide (Alfa Aesar, 270 m g ) was pre-
with an integrated Gatan Image Filter Model 2001. EF-TEM
maps of Co, Fe, Al, O and Cl were acquired on the basis of the
three-window method at ionization edges of the elements.
For the preparation of the TEM grids, a small amount of each
sample was suspended in ethanol and ultrasonicated for at
least 5 minutes. Afterwards, a droplet (12 μL) was placed on
carbon-coated copper grids.
calcined in air at 600 °C for 4 h. An aqueous solution of
−1
cobalt nitrate (0.33 mol L ) was added to the pretreated
alumina support until visible wetness was observed. The
procedure was repeated ten times with drying at 130 °C fol-
lowing each impregnation step. Analogously to the xerogels,
the catalyst was then calcined at 400 °C for 4 h in a static
3
4,35
oven in air.
The specific surface areas of all catalysts were determined
by single point BET measurements after drying, calcination,
and FT tests. The powder samples were measured at a relative
pressure of 0.3, using a βeta Scientific Corp. Instrument
(model 4200). Before data collection, the samples were out-
gassed at 150 °C for 2 h under He flow. Each measurement was
conducted at the boiling point of liquid nitrogen (77 K).
In addition, adsorption–desorption isotherms were collected
for the calcined catalysts within a pressure range of P/P0
between 0.01 and 0.99 using a Quantachrome Nova 2000e
instrument. The pore size distributions were calculated from
the desorption branch of the isotherms based on the Barrett–
Joyner–Halenda (BJH) model. Before each measurement, the
samples were pretreated at 150 °C for 15 h under vacuum.
The reduction behavior of the calcined catalysts was ana-
lyzed by collecting temperature-programmed reduction (TPR)
profiles, employing a Quantachrome CHEMBET-3000 instru-
An iron reference catalyst was synthesized by co-precipitation.
Iron nitrate (0.6 M) and aluminum nitrate with the desired
O ratio were used as precursors. The hot aqueous
2 3
precursor solution was continuously precipitated in
stirred round-bottom flask at 82 °C using hot ammonia
solution (2.7 M). Fluid pumps were used to adjust the flow
Fe : Al
a
rate of the precursor and the NH
3
solution in order to
achieve a constant precipitation pH of about 6. The precipi-
tate was cooled with ice and then washed with cold distilled
water (about 4 L) by vacuum filtration. After drying at 50 °C
48 h) and 120 °C (12 h), the catalyst was calcined analo-
gously to the iron-based xerogels at 350 °C for 4 h in a static
(
3
6,37
oven in air.
2
.2 Catalyst characterization
−
1
The calcined xerogels and the freshly-prepared reference cata-
lysts as well as the catalysts after FT reaction were character-
ized by transmission electron microscopy (TEM) using a FEI
Tecnai F20 S-TWIN microscope. The microscope was operated
at an accelerating voltage of 200 kV. TEM and energy-filtered
images (EF-TEM) were recorded with a slow-scan CCD camera
ment. Passing a gas mixture of 5 vol.% H in Ar (60 ml min )
2
through the reactor, hydrogen uptake was monitored with a
thermal conductivity detector (TCD), heating the sample from
−
1
room temperature up to 1000 °C with a rate of 10 °C min .
Before each measurement, the catalysts were purged in N
2
flow at 200 °C for 30 min to remove any traces of water.
Table 2 BET surface areas after different catalyst treatments (a–c) and porosity measurements of the calcined xerogels prepared from nitrate salts
2
−1
Fe :
Al
BET specific surface area [m g ] after different treatments
2
O
3
Pore
Max. pore size
distribution
[nm]
a
a
ratio
[wt%]
volume
3
−1
Notation
Fe
a) Drying
b) Calcination
c) FT 430 °C
[cm g
]
100
80.0
66.7
60.0
33.3
9.1
534
517
512
520
592
572
618
—
49
250
270
382
395
486
763
92
7
65
102
103
124
360
—
0.18
0.22
0.28
0.26
0.33
0.62
0.66
0.26
14
4
4
4
4
5
4
8
Fe : Al
Fe : Al
Fe : Al
Fe : Al
Fe : Al
O
2 3
O
2 3
O
2 3
O
2 3
O
2 3
1 : 0.25
1 : 0.5
1 : 1
1 : 2
1 : 10
Al O
2 3
0.0
80.0
Precip. 1 : 0.25
43
a
Calculated using Barrett–Joyner–Halenda (BJH) method; for calcined catalyst only.
3258 | Catal. Sci. Technol., 2013, 3, 3256–3267
This journal is © The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Paper
Thermogravimetric (TG) analysis was used to record the
mass loss during heat treatment of the dried xerogel catalysts
in order to determine the xerogel calcination temperatures.
Using a Netsch STA and a Linseis instrument, selected sam-
ples were heated from room temperature to 600–1000 °C with
the sample was transferred in a glass vial and the catalyst
was separated from the non-magnetic SiC and some residual
quartz wool by sliding a magnet at the outer surface of the
glass vial. The separated catalyst powder was then transferred
in another vial.
−
1
−1
a rate of 2 °C min in a synthetic air stream (50 mL min ).
3. Results and discussion
2
.3 Catalyst evaluation in the FT reaction
3
.1 Synthesis and structural characterization
The Fischer Tropsch (FT) performance of the Co and Fe-based
xerogels and reference catalysts were measured in a fixed-bed
reactor system. For each catalytic test, 0.124 g (180–250 μm)
of the cobalt catalyst was diluted with inert SiC (3 g, 180–250 μm)
and loaded in the reactor using quartz wool to fix the sam-
ple. In contrast, the amount of iron catalyst was calculated
so that 0.2 g (180–250 μm) of Fe was used irrespective of the
Fe : Al O ratio and mixed with SiC (2.7–3 g). An axially placed
After ambient drying of the pink (Co–Cl) and purple (Co–N)
wet gels and their calcination in air, black monoliths were
obtained which were crushed for the catalytic measurements.
In the case of iron, the brownish color of the pure Fe xerogels
became lighter with increasing aluminum content. All calcined
catalysts were characterized by TEM/EF-TEM, nitrogen
physisorption and TPR measurements in order to determine
the distribution, stability and reducibility of the FT active
metal in dependence of the applied precursor in the case of
Co and the varying metal-to-promoter ratio in the case of Fe.
As shown by the TEM images depicted in Fig. 1 and 2, both,
Co- and Fe-based xerogels consist of interconnected nano-
particles, forming a three-dimensional network after calcina-
tion. The structure visible on different length scales (here
nano- and micrometer) apparently depends on the anion of
the precursor. As observed in the case of Co, the xerogel
derived from the chloride salt (Co–Cl) exhibits a web-like
fibrous nanostructure with ligaments of different length up to
50 nm. In contrast, interconnected spheres of about 3–8 nm
in a sponge-like arrangement form when using nitrate precur-
sors in the case of both, iron- and cobalt-based xerogels. This
structural difference has already been observed for pure alumina
gels synthesized with different anions and was explained by
different growth rates and complexation characteristics, depen-
2
3
thermocouple in contact with the catalyst bed monitored and
controlled the temperature inside the reactor. Gas flow rates
were adjusted by mass flow controllers (Bronkhorst Mättig).
A N pressure line after the reactor ensured constant pressure
2
during the reaction in the reactor system.
In situ activation of the catalysts prior to the FT reaction
−
1
was conducted at 1 bar in pure H (60 mL min ). Cobalt cata-
2
−
1
lysts were treated at 430 °C for 16 h (heating rate 3 °C min
till 200 °C, 1 °C min till 430 °C) or at 800 °C for 3 h
heating rate 3 °C min , see following section for details)
−
1
−
1
(
while activation of iron catalysts was carried out at 350 °C for
−
1
−1
1
6 h (heating rate 2 °C min till 150 °C, 1 °C min
till 350 °C). After reduction, the temperature was allowed to
cool down to 150 °C and the system was flushed with Ar. FT
measurements were performed at 5 bar and 230 °C (Fe) or
2
50 °C (Co). (Note that slightly higher reaction temperatures
for the Co catalysts were applied for the catalytic measure-
ments as compared to standard conditions (220–230 °C) to
increase the conversion, i.e. to partly compensate for the
2
2
ding on the precursor anion and the solvent. Elemental
mappings included in Fig. 1 and 2 demonstrate the homoge-
neous distribution of the FT-active components Co and Fe
on the nanometer scale for all investigated calcined catalysts.
This is even observed for samples with a 10-fold higher
−1
lower total pressure.) The total gas flow was set to 30 mL min
H : CO : Ar = 6 : 3 : 1) with Ar used as internal standard.
(
2
The FT performance (activity and selectivity) was deter-
mined over 20 h on stream monitored with an on-line gas
chromatograph (Hewlett-Packard 5890 II). Analysis of the per-
manent gases H , CO, N , CH , CO and Ar as internal stan-
2 3
alumina content by weight (Fe : Al O 1 : 10, Fig. 2).
All alumina-containing xerogels exhibit high specific surface
areas (SSA) and narrow pore size distributions with pore diame-
ters in the mesoporous range after calcination (see Table 1 and 2).
As a general trend, the SSA of the xerogels decreases as com-
pared to pure alumina xerogel when adding cobalt or iron, irres-
pective of the precursor (chloride or nitrate). Yet, in the case
of the Co-containing xerogel (Co–Cl and Co–N), the precursor
anion influences the resulting surface area. The sponge-like
arrangement of spherical nanoparticles formed by nitrates
leads to a higher overall surface area as compared to the
chloride-based xerogels.
In case of the iron oxide/alumina xerogels prepared from
nitrates, the data allow judging the influence of the iron to
aluminium ratio on the SSA. After drying, the SSA of the pure
and of all alumina containing Fe xerogels are in the same
range increasing slightly with higher alumina contents but
significantly decreasing during calcination depending on the
2
2
4
2
dard was performed with a molecular sieve column, using a
thermal conductivity detector (TCD). The molar gas flows
were determined by multiplying the component-to-Ar peak
ratio, the adjusted Ar molar flow and corresponding calibra-
tion factors. The molar flows were then used to calculate the
conversion of CO as well as CH and CO yields. A flame ioni-
4
2
zation detector (FID) equipped with a fused silicone capillary
column (FS-OV-1-CB-2) was used to analyze the organic com-
ponents up to C8 by referring their peak areas to that of
methane. Hydrocarbon selectivities were then calculated based
on the methane flow determined by TCD analysis.
After 20 h time on stream, the catalyst was cooled down in
−
1
Ar atmosphere (30 mL min ) and subsequently recovered
under ambient conditions. The catalyst bed (catalyst + SiC)
was then separated from the quartz wool by hand. Afterwards,
This journal is © The Royal Society of Chemistry 2013
Catal. Sci. Technol., 2013, 3, 3256–3267 | 3259

View Article Online
Paper
Catalysis Science & Technology
Fig. 1 EFTEM images at zero-loss as well as Co and Al mappings of alumina-supported cobalt catalysts after calcination: (I) Co–Cl xerogel prepared from metal chloride salts
and (II) Co–N xerogel synthesized based on metal nitrate salts.
alumina content. The extent of shrinkage increases as the
alumina content is decreased. The most drastic loss of specific
surface area occurred for the pure iron oxide xerogel due to
sintering as also corroborated by the TEM results showing
2 3 2 3
Fig. 2 EFTEM images at zero-loss as well as Fe and Al mappings of alumina-promoted iron catalysts after calcination: (I) Fe : Al O 1 : 10, (II) Fe : Al O 1 : 1 and (III) pure iron
oxide xerogel catalyst. All catalysts were synthesized from metal nitrate salts.
3260 | Catal. Sci. Technol., 2013, 3, 3256–3267
This journal is © The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Paper
fused iron oxide particles, each with a diameter of about 30 nm
(Fig. 2, III). Obviously, alumina promotion allows the prepara-
tion of stable Co-based xerogels (no significant sintering
during calcination; note that the EAM method was not suit-
able to obtain pure cobalt oxide gels so that a comparison
between pure Co xerogels and alumina stabilized Co xerogels
is not possible) and it leads to an enhanced sintering stabi-
lity for Fe-based xerogels (cf. SSA loss after calcination bet-
ween alumina promoted Fe and pure Fe xerogels), ensuring
high SSAs for both systems which is important for the
catalytic reaction.
Besides the surface areas, the pore characteristics of the
calcined catalysts were determined by nitrogen physisorption
measurements. All xerogels exhibited type IV isotherms. The
most frequent pore diameter (maximum of pore size distribu-
tion) for all alumina-containing Fe xerogels after calcination
was about 4 nm as a result of maintaining the sponge-like
structure. In contrast, sintering and the concomitant collapse
of SSA during calcination of pure iron oxide resulted in a
shift of the maximum of the pore size distribution to larger
diameters (14 nm) in conjunction with lower pore volumes
Fig. 4 TPR profile of the calcined pure Fe and Fe–Al xerogel catalysts and reference
catalyst prepared by precipitation. All profiles were recorded with a heating rate of
−
1
1
0 °C min .
the FT synthesis in order to achieve a highly active FT cata-
lyst. A comparison of the investigated Co-based catalysts
reveals that the preparation method as well as the metal pre-
cursor applied for the xerogel synthesis strongly affect the reduc-
tion behavior. With respect to the applied preparation method,
the reference catalyst prepared by impregnation shows the
standard reduction behavior of an alumina-supported Co
catalyst. As shown in Fig. 3, a two step reduction of Co O
(Table 2). The structural difference of the Co xerogels derived
from nitrates and chlorides leads to different maxima of the
pore size distribution of 9 nm (Co–N) and 14 nm (Co–Cl) being
slightly higher than those for the iron xerogels (Table 1).
The origin of the difference in pore diameter compared to pure
alumina or iron-alumina gels is not yet clear.
3
4
to CoO and then to metallic cobalt occurred at around 420 °C
as indicated by the two overlapping peaks. The second peak
3
.2 Reduction behavior
38
Aiming to compare the reduction behavior of Co and Fe
xerogels to that of the reference catalysts temperature-
programmed reduction (TPR) experiments using hydrogen
were performed. The results for the Co and Fe-based catalysts
are presented in Fig. 3 and 4, respectively.
between 500–700 °C can be attributed to the reduction of
cobalt oxide in close contact to the alumina support and cobalt
3
8–40
aluminates which are only reduced at high temperatures.
Contrary to the impregnation catalyst, the Co xerogels are
more difficult to reduce to the metal as evidenced by the domi-
nating high temperature peaks shown in Fig. 3. This observa-
tion can be attributed to the formation of cobalt aluminates
during calcination as a consequence of the homogenous Co
distribution within the alumina matrix. Interestingly, the redu-
cibility of the chloride-based xerogel is slightly better than that
of the Co–N catalyst as judged from the lower temperatures
of the reduction peaks.
For Co-based FT catalysts, cobalt oxide has to be reduced
to metallic Co in the reduction process performed prior to
In order to prepare highly active Fe-based FT catalysts, iron
oxide has to be transformed into FT active iron carbides. This
2 3
carburization process can be suppressed by strong Fe–Al O
4
1
interactions. Therefore, TPR measurements using H were
2
2 3
performed to study the degree of Fe–Al O interactions of the
xerogel catalysts with varying alumina content and of the refer-
ence catalyst for comparison.
The TPR profiles of the calcined Fe xerogels and the refer-
ence catalyst prepared by precipitation are presented in Fig. 4.
In all cases, the typical three-step reduction of hematite (Fe O )
2
3
is observed with a first reduction peak at about 400 °C which
can be assigned to magnetite (Fe ) formation. The second
broad peak ranging from 500 to 1000 °C represents the reduc-
3 4
O
Fig. 3 TPR profiles of the calcined alumina stabilized Co xerogels and the reference
−1
catalysts recorded with a heating rate of 10 °C min
.
tion of Fe O
3 4
to metallic iron or, above 570 °C, to the metal via
This journal is © The Royal Society of Chemistry 2013
Catal. Sci. Technol., 2013, 3, 3256–3267 | 3261

View Article Online
Paper
Catalysis Science & Technology
4
2,43
reduction to wüstite (FeO).
The data allow the conclusion
3.3.1 FT performance of Co-based catalysts. Since the TPR
measurements revealed the limited reducibility of the Co xero-
gels at lower temperatures, the FT performance of these cata-
lysts was tested after a standard reduction procedure at 430 °C
that the applied preparation method (EAM vs. classical precipi-
tation) has no effect on the reduction behavior if the same
metal-to promoter ratios are compared (Fe : Al O = 1 : 0.25).
2
3
For both methods a similar H2 consumption and reduction
peaks in the same temperature range are then observed.
as well as after high temperature H treatment at 800 °C. The
2
results are shown in Fig. 5 and Table 3. (In order to differentiate
between the pretreatments in the following text, the applied
reduction temperatures are added to the catalyst notations.)
After low temperature reduction, the xerogel catalysts showed
a distinctly inferior FT performance as compared to the refer-
ence impregnation catalyst, irrespective of the applied pre-
cursor. As shown in Fig. 5, the xerogel activities, expressed by
the CO conversion during 20 h on stream were significantly
lower than the activity of the reference catalyst (Co–IWI 430 °C).
Whereas the xerogels lead to a conversion of 5% and 11%,
respectively (Co–N 430 °C and Co–Cl 430 °C), the reference
catalyst showed a conversion of about 38% CO. This result is
not surprising. Due to the inferior reducibility of the xerogels,
the number of active Co sites generated during reduction
at 430 °C is significantly lower as compared to the IWI refer-
ence catalyst.
(Similar TPR profiles have also been observed for pure iron
oxide prepared by both techniques (not shown)). This is not
too surprising in view of the similar chemistry taking place.
During co-precipitation and xerogel synthesis, aluminum and
iron species form simultaneously and the resulting oxy-hydroxides
1
0,21
are calcined afterwards in both cases.
Yet, the profiles for the other xerogels with higher alumina
contents reveal that the amount of alumina added influences
the reducibility of iron oxide. With increasing content, the
reduction peaks shift to higher temperatures and peak broad-
ening is observed in agreement with reports for Fe–Al O FT
2
3
catalysts prepared by synthesis routes other than sol–gel tech-
4
4,45
niques.
Moreover, a decrease in the overall extent of
con-
interactions with increas-
reduction is found as judged by the decreasing total H
sumption due to stronger Fe–Al O
2 3
2
ing alumina content.
The TEM results of the spent catalysts are shown in Fig. 6.
Interestingly, the Co–Cl catalyst showed slight sintering
(Co particles of 21 ± 7 nm formed during the catalytic treatment),
whereas in the case of the IWI-catalyst and the Co–N catalyst
the Co particles basically kept their size. The lower stability
of the Co–Cl-catalyst might be a result of chloride residuals
(detected by EFTEM analysis (not shown)) known to promote
the metal atom mobility on the support. In view of this sin-
tering tendency, it is surprising that the Co–Cl catalyst exhibited a
two-fold higher FT activity than the Co–N catalyst.
This fact, however, may be explained by the more pronounced
tendency of the Co–N xerogel to form aluminates which
cannot be distinguished from FT active metallic Co in the
EFTEM mappings.
A special case is the Fe : Al 1 : 10 catalyst with 10-fold
higher alumina weight content differing from all other xerogel
catalysts investigated in this study. For this catalyst, a small
hump between 600–750 °C was recorded but a pronounced
high temperature peak above 500 °C was missing. Also, this
catalyst was observed to be non-magnetic after reduction by
48
2
H . These observations indicate that a high alumina content
inhibits the reduction to metallic iron by stabilization of the
41
FeO phase and – perhaps more importantly – due to the for-
mation of hardly reducible iron aluminates during the xerogel
18,46
synthesis.
For the latter species, Józwiak et al. detected a
three-step reduction pathway of FeAlO in analogy to hematite
3
2 3
(Fe O ) reduction. However, the reduction of the iron aluminate
47
to metallic iron could not be accomplished below 1000 °C.
Summarizing the results, sintering during calcination is
prevented for the xerogels ensuring a high specific surface
area which is important for the catalytic reaction. Yet, as a
consequence of the fine dispersion of Co and Fe, respectively,
in the alumina matrix and the tendency to aluminate forma-
tion, the reducibility decreases due to the intimate contact
between aluminum and the FT active metal. In the following
section, we will see how this affects the use of the xerogels as
FT catalysts.
3
.3 Evaluation of Co and Fe xerogel catalysts in the FT
reaction
The FT performance of the xerogels and the reference cata-
lysts was investigated in a fixed bed reactor at 230 °C (Fe)
and 250 °C (Co), respectively. After the catalytic experiments,
the catalysts were once again characterized by BET and TEM/
EF-TEM measurements. In the following, these results are
discussed separately for the Co and Fe systems.
Fig. 5 CO conversion of differently activated Co-based catalysts during 20 h TOS
after H
2
activation. FT reaction was performed in a fixed bed reactor at 250 °C,
−1
5 bar, 30 mL min (H
2
: CO = 2 : 1) using 0.126 g catalyst (22 wt% Co).
3262 | Catal. Sci. Technol., 2013, 3, 3256–3267
This journal is © The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Paper
−1
Table 3 FT performance of differently activated Co catalysts after 20 h TOS. FT reaction was performed in a fixed bed reactor at 250 °C, 5 bar, 30 mL min (H
2
:CO = 2:1)
using 0.126 g catalyst (22 wt% Co)
Avg.
CO
Selectivity (wt%)
Olefin/paraffin
Conv.
Productivity
(%)
a
b
Notation
(%)
CO
2
CH
4
2
C –C
4
C
5+
C
4
C
5
C
6
Co–N 430 °C
Co–Cl 430 °C
Co–IWI 430 °C
Co–N 800 °C
Co–Cl 800 °C
Co–IWI 800 °C
5
11
38
39
<3
25
<1
<1
1
1
—
14
14
14
21
—
10
11
12
13
19
—
8
74
73
72
59
—
81
1.2
0.3
1.2
0.1
—
0.2
0.60
0.09
0.70
0.04
—
0.04
0.30
0.04
0.30
0.02
—
0.02
4
8
27
23
—
20
1
a
b
Avg. CO-conv.: average CO conversion (±3%). Productivity: (conversion × C5+)/100.
The selectivity patterns of the cobalt catalysts are listed in
Table 3. All Co catalysts exhibited similar methane and C5+
selectivities. Only the olefin-to-paraffin ratio was significantly
lower for the xerogel synthesized from chloride precursors.
The reason for this increased paraffin formation is probably
the larger size of the metallic Co particles formed in this
case, resulting in a higher hydrogenation activity.
Since the FT performance of the Co xerogels is impaired
by the limited reduction of the Co at low temperatures, it can
39% when applying the high temperature treatment. In other
words, the activity of Co–N 800 °C was as high as for refer-
ence catalysts after standard reduction (Co–IWI 430 °C).
The TEM images and Co maps in Fig. 6 and Fig. 7 confirm
sintering as the origin for the diminished FT activity. For the
IWI catalyst, the Co particle size increased from about 11 nm
to 16 nm, resulting in 32% less Co surface area. This reduc-
tion in active surface area directly correlates with a decrease
of CO conversion from about 38 to 25%, corresponding to a
reduction of activity by 34%. Contrary, for catalyst Co–Cl 800 °C
the observed sintering of Co particles (particle size: 28 ± 7 nm)
cannot be the only explanation for its activity loss. (Note that
this catalyst was totally FT inactive after high temperature
reduction.) In addition, a drastic loss of total SSA of this cata-
lyst was observed (Table 1) possibly involving Co particles
being “buried” and no longer accessible for the FT reaction.
Contrary to the IWI catalyst and the Co–Cl xerogel, a com-
parison of the TEM images of the Co–N 800 °C and Co–N
430 °C xerogel reveals no apparent change in Co particle size.
4
9
2
be expected that H treatment at high temperatures should
balance this disadvantage. Yet the question arises whether
increased sintering occurs under such harsh conditions thus
counteracting the better reduction.
In Fig. 5 and Table 3, summarizing the catalytic results,
two opposite trends can be observed. The FT activity of the
IWI reference catalyst and the Co–Cl 800 °C xerogel catalyst
indeed decreases as compared to the activity after the stan-
dard reduction procedure. Yet, in contrast, the CO conversion
of the Co–N catalyst increases drastically from 5 (430 °C) to
Fig. 6 EFTEM images at zero-loss and Co mappings of the Co–Cl (A), Co–N (B) xerogel and reference Co-IWI (C) catalysts after reduction treatment at 430 °C and
FT reaction.
This journal is © The Royal Society of Chemistry 2013
Catal. Sci. Technol., 2013, 3, 3256–3267 | 3263

View Article Online
Paper
Catalysis Science & Technology
This result is confirmed by BET measurements of the cata-
lysts after FT reaction, showing constant overall SSA (Table 1).
Apparently the noticeable increase in the FT activity of cata-
lyst Co–N 800 °C can be attributed to the excellent stabiliza-
tion of small Co particles within the sponge-like alumina
matrix even during high temperature treatment necessary to
achieve a higher degree of reduction (cf. TPR results). The
selectivity pattern is in agreement with this conclusion. A
decrease of C5+ and increased methane formation was
observed for the Co–N 800 °C xerogel catalyst as compared to
the IWI catalyst reduced at 430 °C. This can be explained
by the formation of small metallic Co particles below 6 nm
2–3 times higher than the FT activity of the pure iron xerogel
catalyst showing a CO conversion of just 6% (Fig. 8).
The FT results of the two borderline cases with maximum
or no Al O content prove the necessity of structural promo-
2
3
tion with alumina within certain limits in order to provide a
sufficient number of active iron surface sites for the catalytic
reaction. The reason is revealed by the reduction behavior
and the structural analysis showing, on the one hand, that
the unpromoted pure iron oxide xerogel catalyst strongly sin-
ters, forming iron particles up to 150 nm after catalysis
(TEM: Fig. 9) and low SSA (BET: Table 2). On the other hand,
2
TEM images of the Fe : Al O3 1 : 10 xerogel indicate that, after
49–51
during the high temperature H treatment.
FT reaction, the catalyst exhibits a similar structure as in the
as-prepared state with homogeneously distributed iron and
aluminum (compare with Fig. 2 and 9) and just a few bigger
iron nanoparticles as observed for the other FT active cata-
lysts as well. However, their density is significantly lower for
2
In conclusion, the EAM method enables the preparation
of alumina-supported Co FT catalysts being competitive with
the reference catalyst prepared by the established impregna-
tion technique when applying the EAM method with nitrate
precursors and using high temperature reduction. Note that
the stability of the small Co particles during the high
temperature reduction also implies a high thermal stability
and possibly long term stability against sintering.
2 3
the Fe : Al O 1 : 10 xerogel (compare Fig. 9, inset) suggesting
that the inactivity of the alumina-rich xerogel is due to the
formation of stable iron aluminates (as discussed in section 3.3)
which can be hardly converted to FT active iron carbide species.
In spite of the similar FT activities for all other investi-
3
.3.2 FT performance of Fe-based catalysts. The FT per-
formances of alumina-promoted iron xerogels with varying
Fe–Al ratios and of the reference catalyst synthesized by
2 3
gated xerogel catalysts with Fe–Al O ratios varying in the
2
O
3
range of 1 : 2 to 4 : 1, the structural analysis after reaction
points to differences in the catalysts morphology in depen-
dence of the alumina content. With decreasing alumina con-
tent enhanced sintering is observed, leading to bigger iron
particles and consequently lower specific iron surface area as
exemplarily shown in Fig. 9. Whereas the presented zero-loss
co-precipitation are summarized in Table 4. A number of
interesting conclusions can be drawn: comparing, e.g., the
2 3
Fe : Al O 1 : 0.25 catalyst with the reference catalyst prepared
with the same Fe–Al O ratio, a similar activity is observed,
2
3
indicating that the preparation method is of minor impor-
tance (see also discussion in section 3.2). Yet, the FT perfor-
mance depends in a stepwise fashion on the alumina content.
images and iron mappings for the Fe : Al
reveal an average iron particle size of about 20 nm after cata-
lysis, ~37 nm particles are detectable for the Fe : Al 4 : 1
xerogel. In accordance, the SSA of the catalysts after FT reaction
2 3
O 1 : 1 xerogel
While the Fe: Al
O
2 3
1 : 10 sample was FT inactive, all other
2 3
O
alumina promoted xerogels showed similar activities being
Fig. 7 EFTEM images at zero-loss and Co mappings of the Co–Cl (D), Co–N (E) xerogel and reference Co–IWI (F) catalysts after reduction treatment at 800 °C and FT
reaction.
3264 | Catal. Sci. Technol., 2013, 3, 3256–3267
This journal is © The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Paper
Table 4 FT performance of iron-based xerogel catalysts with varying Fe : Al
2
O
3
ratios after 20 h TOS. FT reaction was performed in a fixed bed reactor at 230 °C,
−1
5
bar, 30 mL min (H
2
: CO = 2 : 1) using 200 mg Fe
Selectivity (wt%)
Olefin/paraffin
Productivity
Avg. CO
Conv. (%)
a
b
Notation
CO
2
CH
4
2
C –C
4
C
5+
C
4
C
5
C
6
(%)
Fe
6
16
15
15
14
—
14
8
7
6
8
8
—
7
18
8
8
9
9
—
9
31
16
15
17
18
—
16
43
69
71
66
65
—
68
1.7
1.8
2.1
2.4
2.6
—
2.0
1.2
1.2
1.5
1.7
1.9
—
1.4
0.8
0.7
0.9
1.0
1.2
—
0.8
3
11
11
10
9
—
10
Fe : Al 1 : 0.25
Fe : Al 1 : 0.5
Fe : Al 1 : 1
Fe : Al 1 : 2
Fe : Al 1 : 10
Precip. 1 : 0.25
a
b
Avg. CO-conv.: average CO conversion (±3%). Productivity: (conversion × C5+)/100.
is the reduction/carburization behavior which increases as the
alumina content is decreased, as discussed in section 3.3.
Therefore, the observed similar activities of the catalysts
with Fe: Al O ratios in the range of 1 : 2 to 4 : 1 may be the
2
3
result of the two counteracting trends basically compensa-
ting each other.
The selectivities of the catalysts with respect to hydrocar-
2
bon and CO formation and the olefin-to-paraffin ratio are
presented in Table 4. The pure iron oxide catalyst exhibits an
inferior selectivity, i.e., a higher methane and lower C5+ for-
mation, compared to all other xerogel and reference catalysts,
most likely as a result of lower CO conversion. As for the
activities, also the selectivity pattern for the reference catalyst
and the xerogel catalysts are similar. With varying alumina
content, the formation of CO , CH and C was not signifi-
2
4
5+
Fig. 8 CO conversion of iron based xerogel catalysts with different Fe : Al
during 20 h TOS after H activation. FT reaction was performed in a fixed bed
reactor at 230 °C, 5 bar, 30 mL min (H
2
O
3
ratios
cantly affected (Table 4).
2
In the literature, alumina promotion is reported to enhance
chain termination and hydrogenation in favor of light paraffin
formation due to decreased surface basicity. The latter was
observed to suppress the CO adsorption and, at the same time,
−1
2
: CO = 2 : 1) using 200 mg Fe.
decreases with decreasing alumina content (Table 2). The
other factor being important in addition to the sintering behavior
enhances H adsorption which results in stronger hydrogenation
2
2 3 2 3 2 3
Fig. 9 EFTEM images at zero-loss and Fe mappings of pure Fe (A), Fe : Al O 1 : 0.25 (B), Fe : Al O 1 : 1 (C) and Fe : Al O 1 : 10 (D) xerogel catalysts after FT reaction.
This journal is © The Royal Society of Chemistry 2013
Catal. Sci. Technol., 2013, 3, 3256–3267 | 3265

View Article Online
Paper
ability of the catalysts.
observed for the xerogel catalysts studied here. Rather, the
olefin-to-paraffin ratio shown in Table 4 increases with
higher alumina content, indicating a decreased hydroge-
nation activity. For further understanding, the surface acid-
ity of the xerogel catalysts as a function of the alumina
content needs to be investigated because the surface proper-
ties were shown to be not only influenced by the alumina
Catalysis Science & Technology
10,18,41,44
53,54
However, this effect was not
expensive noble metal promoters
(Pt, Pd, Ru), usually
added to impregnation catalysts in order to facilitate reduc-
tion, could be reduced or completely replaced by a high tem-
perature catalyst treatment when using a nitrate precursor.
Having proven that the systems exhibit similar activities
as compared to catalysts prepared by standard preparation
techniques, the sol–gel approach opens a wealth of further
tuning options and possibilities for novel applications. With
respect to the surface chemical properties of the catalyst, its
acidity or the degree of intermixing with further promoters
through addition at varying stages of gel formation can be
1
9
content but also by the preparation method or the type of
5
2
alumina support.
Summarizing, the performance of the xerogel catalysts is
as good as for the co-precipitated catalyst. Alumina promo-
tion significantly enhances the FT activity due to stabilization
of iron nanoparticles against sintering within a certain con-
centration range. A broad range of alumina contents can be
chosen to ensure a good compromise between reducibility/
carburization and stability.
5
5
tailored. In terms of the catalysts morphology, not only
powders, but also monoliths and thin films with variable
porosity from meso- to macrometer range can be synthesized
2
3,29,30,56
by the EAM.
This allows the coating of microreactors
or the development of novel monolithic FT catalysts concepts
bringing about advantages with respect to heat and mass
transfer, pressure drop, and the possibility of tuning the
product selectivity (see e.g. ref. 2, 57, 58).
4. Conclusion
In this report, the suitability of the epoxide addition method
as an alternative to well-established impregnation and precip-
itation preparation techniques for the synthesis of alumina-
supported Co and alumina-promoted Fe FT catalysts is
demonstrated. The EAM as a special sol–gel technique enables
the synthesis of high surface area FT catalysts with variable
stochiometries/metal loadings, using cheap metal salts as pre-
cursors. The synthesis parameters, such as the applied precur-
sor and the metal-to-alumina ratio, determine the properties
of the resulting xerogel catalyst in terms of distribution, sta-
bility and reducibility of the FT active metal and thus, the
FT performance. The Co- and Fe-based xerogels showed a FT
performance that was competitive to that of reference cata-
lysts synthesized by established impregnation and precipita-
tion techniques but partly exhibited higher thermal stabilities.
In the case of Fe, the catalytic results reveal that the con-
tent of the structural promoter is important for the FT perfor-
mance. Intermediate Fe : Al O ratios in the range of 1 : 2 to
Acknowledgements
The authors gratefully acknowledge financial support of the
EWE AG and the DFG. We thank Mireille Bipoun for some of
the syntheses and characterizations, Haibin Zheng from the
group of Prof. Helena Hagelin-Weaver (University of Florida)
for some of the TPR measurements of the iron oxide xerogels
and Karsten Thiel (Fraunhofer Institute for Manufacturing
Technology and Advanced Materials (IFAM), Bremen) for
assistance in TEM.
References
1 M. E. Dry, in Fischer-Tropsch Technology, Elsevier Science,
Amsterdam, 2004, vol. 152, pp. 1–700.
2 A. Y. Khodakov, W. Chu and P. Fongarland, Chem. Rev.,
2007, 107, 1692–1744.
3 B. H. Davis, Ind. Eng. Chem. Res., 2007, 46, 8938–8945.
4 P. J. van Berge and R. C. Everson, in Studies in Surface
Science and Catalysis, Elsevier, 1997, vol. 107, pp. 207–212.
5 E. van Steen and M. Claeys, Chem. Eng. Technol., 2008, 31,
655–666.
2
3
4
: 1 by weight lead to superior FT performances of the xerogels
compared to promotion with no or with too high alumina
amounts which suffered from sintering or formation of irre-
ducible aluminates, respectively.
In the case of Co, it was shown that the FT performance
is affected by the precursor salt used for the synthesis
and by the reduction temperature employed for activating
the catalysts prior to FT synthesis. The use of chloride-
containing precursors for Co FT catalyst synthesis lead to
competitive FT performances neither after standard reduc-
tion nor after high temperature treatment as a result of an
inferior stability and low Co dispersion due to sintering.
Contrary, the FT activity of the Co xerogel catalyst derived
from nitrates was competitive with the performance of the
reference catalyst after high temperature activation due to
excellent stabilization of small metallic Co particles within
the alumina matrix. Based on this high temperature stability
6 H. Schulz, Appl. Catal., A, 1999, 186, 3–12.
7 E. de Smit and B. M. Weckhuysen, Chem. Soc. Rev., 2008, 37,
2758–2781.
8 M. E. Dry, Appl. Catal., A, 2004, 276, 1–3.
9 A. Y. Khodakov, Catal. Today, 2009, 144, 251–257.
10 D. B. Bukur, X. Lang, D. Mukesh, W. H. Zimmerman,
M. P. Rosynek and C. P. Li, Ind. Eng. Chem. Res., 1990, 29,
1588–1599.
11 A. Gual, C. Godard, S. Castillón, D. Curulla-Ferré and
C. Claver, Catal. Today, 2012, 183, 154–171.
12 S. A. Gardezi, J. T. Wolan and B. Joseph, Appl. Catal., A,
2012, 447–448, 151–163.
13 Y. Q. Zhuang, M. Claeys and E. van Steen, Appl. Catal., A,
2006, 301, 138–142.
(as compared to the standard IWI catalyst), the amount of
3266 | Catal. Sci. Technol., 2013, 3, 3256–3267
This journal is © The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Paper
1
4 V. R. Calderone, N. R. Shiju, D. Curulla-Ferré, S. Chambrey,
A. Khodakov, A. Rose, J. Thiessen, A. Jess and G. Rothenberg,
Angew. Chem., 2013, 125, 4493–4497.
35 E. Iglesia, S. L. Soled and R. A. Fiato, J. Catal., 1992, 137,
212–224.
36 T. Riedel, G. Schaub, K. W. Jun and K. W. Lee, Ind. Eng.
Chem. Res., 2001, 40, 1355–1363.
37 D. B. Bukur, X. S. Lang, J. A. Rossin, W. H. Zimmerman,
M. P. Rosynek, E. B. Yeh and C. P. Li, Ind. Eng. Chem. Res.,
1989, 28, 1130–1140.
38 Y. T. Tsai, X. H. Mo, A. Campos, J. G. Goodwin and
J. J. Spivey, Appl. Catal., A, 2011, 396, 91–100.
1
1
1
5 R. Guettel, J. Knochen, U. Kunz, M. Kassing and T. Turek,
Ind. Eng. Chem. Res., 2008, 47, 6589–6597.
6 R. Y. Xie, D. B. Li, B. Hou, J. A. Wang, L. T. Jia and
Y. H. Sun, Catal. Commun., 2011, 12, 380–383.
7 V. S. Nagineni, S. H. Zhao, A. Potluri, Y. Liang,
U. Siriwardane, N. V. Seetala, J. Fang, J. Palmer and D. Kuila,
Ind. Eng. Chem. Res., 2005, 44, 5602–5607.
39 B. Jongsomjit, J. Panpranot and J. G. Goodwin, J. Catal.,
2001, 204, 98–109.
1
1
8 R. Snel, Ind. Eng. Chem. Res., 1989, 28, 654–659.
9 S. H. Kang, J. W. Bae, P. S. S. Prasad, S. J. Park, K. J. Woo
and K. W. Jun, Catal. Lett., 2009, 130, 630–636.
40 G. Jacobs, T. K. Das, Y. Zhang, J. Li, G. Racoillet and
B. H. Davis, Appl. Catal., A, 2002, 233, 263–281.
41 H. J. Wan, B. S. Wu, C. H. Zhang, H. W. Xiang, Y. W. Li,
B. F. Xu and F. Yi, Catal. Commun., 2007, 8, 1538–1545.
42 W. K. Jozwiak, E. Kaczmarek, T. P. Maniecki, W. Ignaczak
and W. Maniukiewicz, Appl. Catal., A, 2007, 326, 17–27.
43 J. Zieliński, I. Zglinicka, L. Znak and Z. Kaszkur, Appl. Catal.,
A, 2010, 381, 191–196.
2
2
0 C. L. Bianchi and V. Ragaini, Catal. Lett., 2004, 95, 61–65.
1 A. E. Gash, T. M. Tillotson, J. H. Satcher, J. F. Poco,
L. W. Hrubesh and R. L. Simpson, Chem. Mater., 2001, 13,
9
99–1007.
2
2 T. F. Baumann, A. E. Gash, S. C. Chinn, A. M. Sawvel,
R. S. Maxwell and J. H. Satcher, Chem. Mater., 2005, 17,
3
95–401.
44 H. J. Wan, B. S. Wu, X. An, T. Z. Li, Z. C. Tao, H. W. Xiang
and Y. W. Li, J. Nat. Gas Chem., 2007, 16, 130–138.
45 H.-J. Wan, B.-S. Wu, C.-H. Zhang, B.-T. Teng, Z.-C. Tao,
Y. Yang, Y.-L. Zhu, H.-W. Xiang and Y.-W. Li, Fuel, 2006, 85,
1371–1377.
2
2
2
2
2
2
2
3 Y. Kido, K. Nakanishi, A. Miyasaka and K. Kanamori, Chem.
Mater., 2012, 24, 2071–2077.
4 Y. Z. Guo, W. Meyer-Zaika, M. Muhler, S. Vukojevic and
M. Epple, Eur. J. Inorg. Chem., 2006, 4774–4781.
5 B. J. Clapsaddle, A. E. Gash, J. H. Satcher Jr and
R. L. Simpson, J. Non-Cryst. Solids, 2003, 331, 190–201.
6 P. Brown and L. J. Hope-Weeks, J. Sol-Gel Sci. Technol.,
46 J.-Y. Park, Y.-J. Lee, P. K. Khanna, K.-W. Jun, J. W. Bae and
Y. H. Kim, J. Mol. Catal. A: Chem., 2010, 323, 84–90.
47 W. K. Jozwiak, T. P. Maniecki, P. Mierczynski, K. Bawolak
and W. Maniukiewicz, Pol. J. Chem., 2009, 83, 2153–2162.
48 C. H. Bartholomew, Appl. Catal., A, 2001, 212, 17–60.
49 G. Prieto, A. Martínez, P. Concepción and R. Moreno-Tost,
J. Catal., 2009, 266, 129–144.
50 A. Y. Khodakov, A. Griboval-Constant, R. Bechara and
V. L. Zholobenko, J. Catal., 2002, 206, 230–241.
51 G. L. Bezemer, J. H. Bitter, H. Kuipers, H. Oosterbeek,
J. E. Holewijn, X. D. Xu, F. Kapteijn, A. J. van Dillen and
K. P. de Jong, J. Am. Chem. Soc., 2006, 128, 3956–3964.
52 S. H. Kang, J. W. Bae, P. S. S. Prasad and K. W. Jun, Catal.
Lett., 2008, 125, 264–270.
53 F. Diehl and A. Y. Khodakov, Oil Gas Sci. Technol., 2009, 64,
11–24.
54 W. Ma, G. Jacobs, R. A. Keogh, D. B. Bukur and B. H. Davis,
Appl. Catal., A, 2012, 437–438, 1–9.
55 N. Hüsing and U. Schubert, Angew. Chem., 1998, 110, 22–47.
56 Y. Kido, K. Nakanishi and K. Kanamori, RSC Adv., 2013, 3,
3661–3666.
57 C. G. Visconti, E. Tronconi, L. Lietti, G. Groppi, P. Forzatti,
C. Cristiani, R. Zennaro and S. Rossini, Appl. Catal., A,
2009, 370, 93–101.
2
009, 51, 238–243.
7 Y. Bang, J. G. Seo, M. H. Youn and I. K. Song, Int. J.
Hydrogen Energy, 2012, 37, 1436–1443.
8 S. K. Gill, P. Brown, M. T. Ogundiya and L. J. Hope-Weeks,
J. Sol-Gel Sci. Technol., 2010, 53, 635–640.
9 M. M. Koebel, D. Y. Nadargi, G. Jimenez-Cadena and
Y. E. Romanyuk, ACS Appl. Mater. Interfaces, 2012, 4,
2
464–2473.
0 C. D. Park, D. Magana and A. E. Stiegman, Chem. Mater.,
007, 19, 677–683.
1 K. Okabe, X. Li, M. Wei and H. Arakawa, Catal. Today,
004, 89, 431–438.
2 C. L. Bianchi, F. Martini and P. Moggi, Catal. Lett., 2001, 76,
5–69.
3
3
3
3
2
2
6
3 Sol–gel chemistry offers versatile new ways to prepare
catalysts with tunable compositions and in different forms
of application. In particular the option to obtain catalytic
coatings is of interest for microreactors or the deposition of
catalysts on monolithic supports. S. Bali, F. E. Huggins,
G. P. Huffman, R. D. Ernst, R. J. Pugmire and E. M. Eyring,
Energy Fuels, 2009, 23, 14–18.
3
4 M. Kraum and M. Baerns, Appl. Catal., A, 1999, 186,
58 W. Liu, J. L. Hu and Y. Wang, Catal. Today, 2009, 140,
142–148.
1
89–200.
This journal is © The Royal Society of Chemistry 2013
Catal. Sci. Technol., 2013, 3, 3256–3267 | 3267