Alumina-supported iron oxide nanoparticles as Fischer-Tropsch catalysts: Effect of particle size of iron oxide

Article abstract of DOI:10.1016/j.molcata.2010.03.025

The Fischer-Tropsch synthesis of unpromoted and nano-sized iron oxide supported on δ-Al₂O₃ was investigated using a fixed-bed reactor. The catalysts prepared from pre-synthesized iron oxide with varying particle size (2-12 nm) showed much higher catalytic activities than the one prepared by using conventional impregnation method. The best results for CO conversion were obtained when the catalyst had Fe particle size of 6.1 nm. With an increase in particle size, the reduction degree and C₅+ selectivity was increased, whereas CH₄ selectivity and the uptake of adsorbed CO were decreased. Turnover frequency (TOF) at 300 °C was increased from 0.02 to 0.16 s^-1 when d(Fe⁰) was increased from 2.4 to 6.2 nm, and then it remains almost constant up to a particle size of 11.5 nm. Particle sizes of prepared iron oxide were analyzed by XRD and TEM, and the reduction behaviors of Fe/Al₂O₃ catalysts were studied by H₂-TPR. The effective iron size, metal dispersion and reduction degree of Fe/Al₂O₃ catalysts were measured by CO chemisorption and O₂ titration.

Full text of DOI:10.1016/j.molcata.2010.03.025

Journal of Molecular Catalysis A: Chemical 323 (2010) 84–90
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Alumina-supported iron oxide nanoparticles as Fischer–Tropsch catalysts:
Effect of particle size of iron oxide
Jo-Yong Parka,b, Yun-Jo Leea,∗, Pawan K. Khannaa,c, Ki-Won Juna,∗∗, Jong Wook Baea, Young Ho Kimb
a Petroleum Displacement Technology Research Center, Korea Research Institute of Chemical Technology (KRICT), P.O. BOX 107, Yuseong, Daejeon 305-600, Republic of Korea
b Department of Fine Chemical Engineering and Applied Chemistry, BK21-E2M, Chungnam National University, Daejeon 305-764, Republic of Korea
c Nanoscience Laboratory, Centre for Materials for Electronics Technology (C-MET), Panchwati, off Pashan Road, Pune 411 008, India
a r t i c l e i n f o
a b s t r a c t
Article history:
The Fischer–Tropsch synthesis of unpromoted and nano-sized iron oxide supported on ␦-Al2O3 was
investigated using a fixed-bed reactor. The catalysts prepared from pre-synthesized iron oxide with
varying particle size (2–12 nm) showed much higher catalytic activities than the one prepared by using
conventional impregnation method. The best results for CO conversion were obtained when the catalyst
had Fe particle size of 6.1 nm. With an increase in particle size, the reduction degree and C5+ selectivity was
increased, whereas CH4 selectivity and the uptake of adsorbed CO were decreased. Turnover frequency
(TOF) at 300 ◦C was increased from 0.02 to 0.16 s−1 when d(Fe0) was increased from 2.4 to 6.2 nm, and
then it remains almost constant up to a particle size of 11.5 nm. Particle sizes of prepared iron oxide were
analyzed by XRD and TEM, and the reduction behaviors of Fe/Al2O3 catalysts were studied by H2-TPR.
The effective iron size, metal dispersion and reduction degree of Fe/Al2O3 catalysts were measured by
CO chemisorption and O2 titration.
Received 14 January 2010
Received in revised form 15 March 2010
Accepted 20 March 2010
Available online 27 March 2010
Keywords:
Fe/Al2O3 catalyst
Size-defined nanoparticle
Fischer–Tropsch synthesis
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
Usually Fe-based FT catalysts have been prepared by the conven-
tional methods, i.e. impregnation or coprecipitation [15,18,19]. In
In recent years, the catalytic activity of the materials has been
reported to be dependent on the particle size and is considered as
one of the main factors that affects the catalytic activity and selec-
tivity during Fischer–Tropsch Synthesis (FTS). Many researchers
[1–13] have studied the size effect of cobalt metal in FT catalyst.
While turnover frequency (TOF) depends on cobalt particle size for
smaller size (<8 nm), it is independent of the particle size for bigger
size (>8 nm) [1].
Iron-based FT catalysts have not much persuaded particularly
with respect to the particle size effect [14,15]. Mabaso et al.
reported the effect of crystal size on carbon supported iron catalysts
prepared via precipitation using reverse micelle systems where cat-
alysts with smaller metal than 7–9 nm have showed lower TOF
and higher selectivity of methane compared to the bigger-sized
catalysts [14]. Although recently a good account of reaction mech-
anism for FTS using iron catalysts have been documented by Davis
[4,5]. It has been suggested that pre-reduced metallic iron is mostly
changed to iron oxide surrounded by an iron carbide layer in bulk
composition under FTS reaction condition [16]. The iron carbides
are considered as a catalytic active phase [17].
such methods, it is difficult to control an iron crystal size in the cata-
lysts where the crystal has broad distribution in particle size. Such
unavoidable variations in the crystallite size of the catalyst may
reduce catalytic efficiency because the catalytic activity is maxi-
mized in optimum crystal size in FTS reaction [14]. Furthermore
a problem which often associates with these conventional meth-
ods is agglomeration of particles and sintering of active metal on
the surface during the calcinations [20]. It has also been reported
that in the supported iron catalyst it is very difficult to completely
reduce the small-sized iron species due to strong metal–support
interaction when the iron loading is lower than 10% [21,22].
It is well known that the agglomeration and sintering of iron-
based catalyst may lead to low catalytic activity. This therefore,
warrants some smart strategy for use of iron-based catalysts. One
certain way to overcome these drawbacks, is to apply innovative
synthesis of nanoparticles of the catalysts, e.g. nano-iron oxide par-
ticles that can be easily dispersed in organic medium and can be
homogenously loaded on to the support materials.
In this paper, we present design and preparation of the
size-defined Fe/Al2O3 catalyst using pre-synthesized ␥-Fe2O3
nanoparticles. Catalysts based on such size-defined particles are
expected to prevent the agglomeration of small particles by
sintering of active metal. This may be possible because the pre-
synthesized size-defined nanoparticles of iron oxide are dispersible
and can be homogenously impregnated on the alumina support
∗
Corresponding author. Tel.: +82 42 860 7630; fax: +82 42 860 7388.
Corresponding author. Tel.: +82 42 860 7671; fax: +82 42 860 7388.
E-mail addresses: yjlee@krict.re.kr (Y.-J. Lee), kwjun@krict.re.kr (K.-W. Jun).
∗∗
1381-1169/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2010.03.025

J.-Y. Park et al. / Journal of Molecular Catalysis A: Chemical 323 (2010) 84–90
85
and thus making them more active in FTS. We thus expect that
the size-defined Fe/Al2O3 catalyst have better scope in activity
and selectivity in FTS. In present work, a relationship between
catalytic activity and crystal size of iron oxide is studied by using X-
ray diffraction (XRD), transmission electron microscopy (TEM), H2
temperature-programmed reduction (H2-TPR), CO chemisorption
and O2 titration.
2. Experimental
2.1. Preparation of size-defined iron oxide (ꢀ-Fe2O3)
Size-defined iron oxides were prepared by the solvothermal
method. The detailed procedure of this method has been published
elsewhere [23]. The iron-oleate precursor was synthesized from
iron nitrate, oleic acid and NH3 (30%) solution. The iron-oleate pre-
cursor was added in octadecene and then heated to 130–250 ◦C for
5 h. After cooling to room temperature, the resulting nanoparticles
of iron oxide were collected after repeated washing with acetone
and were then air-dried. Finally, the nanoparticles were dispersed
in hexane. The nano-sized iron oxide samples are designated as
x-Fe where x means reaction temperature in Celsius.
2.2. Catalyst preparation
Fig. 1. X-ray diffraction patterns of the size-defined iron oxide: (a) 130-Fe, (b) 180-
Fe, (c) 200-Fe, (d) 230-Fe, (e) 250-Fe.
Catalyst support was prepared by modification of our previ-
ous method [24]. ␥-Al2O3 with high surface area (∼350 m2/g)
was calcined at 1000 ◦C to prepare ␦-Al2O3 support (surface area
of 127 m2/g) having less reactive surface. Supported catalysts of
Fe/Al2O3 = 5/100 wt ratio (denoted to x-FeAl, x means the synthesis
temperature of the nanoparticles) were prepared by impregnation
of the pre-synthesized ␥-Fe2O3 nanoparticles (x-Fe) solution dis-
persed in hexane. The prepared catalysts were dried at 60 ◦C for
12 h and then calcined at 400 ◦C for 5 h in static air. For compari-
son, the reference catalyst denoted to i-FeAl was also prepared by
conventional impregnation method by homogenously mixing the
␦-Al2O3 with iron nitrate in ethanol.
450 ◦C for 12 h under flowing H2, and then the gas was changed
to He flow and cooled to 35 ◦C. The CO chemisorption was carried
out at 35 ◦C. A Fe: CO stoichiometry is assumed to 2:1. After CO
chemisorption measurement, the sample was re-oxidized at 450 ◦C
by 10% O2 in helium to determine the extent of reduction. Disper-
sion percentage (D%) of the iron metal was calculated according to
the equation as has been described in the literature [15,18].
1.117X
D% =
Wf
where X (␮ mol/g) is the total CO uptake, W is the weight percentage
of iron, and f is the fraction of iron reduced to the metal determined
from O2 titration. Average particle size (dp (nm)) was calculated
from D% assuming spherical metal crystallite of uniform diameter.
2.3. Activity test
Typically, 0.5 g of the Fe/Al2O3 catalyst was reduced in flow-
ing 5% H2/Ar (200 cm3/min) at 450 ◦C for 12 h. FTS activities of
the reduced catalysts were tested using a fixed-bed reactor. The
activity tests were conducted under the following reaction condi-
tions; T = 280–300 ◦C; Pg = 10 kgf/cm2; SV (L/kgcat/h) = 3600; feed
compositions H2/CO/Ar = 63.2/31.3/5.5; mol%.
96
dp
=
D%
3. Results and discussion
2.4. Catalyst characterization
3.1. Analysis of size-defined iron oxide
The catalysts processed from pre-synthesized ␥-Fe2O3 nanopar-
ticles or those obtained via impregnation method were studied by
XRD measurement using a Rigaku D/MAX-2200V diffractometer
with Cu/K-alpha radiation (ꢁ = 0.154056 nm). TEM was performed
with TECNAI G20 instrument for the study of the particle mor-
phology and for obtaining exact particle size of iron oxide. TPR
was recorded using a Micromeritics ASAP 2920 equipped with a
thermal conductive detector (TCD). The samples were first pre-
treated in a helium flow up to 200 ◦C and kept for 2 h to remove
the adsorbed water, and then cooling to 100 ◦C. The reducing gas
containing 5% H2/Ar mixture was passed over the samples at a
flow rate of 30 cm3/min. The samples were heated from 100 to
1000 ◦C with heating rate of 10 ◦C/min and kept at that temper-
ature for 30 min. The number of active metal on the surface in
the Fe-based catalyst was measured by CO chemisorption at 35 ◦C
in conventional volumetric adsorption system using micrometrics
ASAP 2020C instrument. The sample (0.1 g) was reduced in situ at
The crystal phase of the prepared iron oxide nanoparticles of
varying size is studied using XRD pattern. Fig. 1(a)–(e) shows the
XRD spectra of iron oxide prepared from iron-oleate precursor in
octadecene at different temperatures, i.e. 130, 180, 200, 230 and
250 ◦C, respectively. The XRD patterns of so-obtained iron oxide
reveals a face-centered cubic crystal structure which relates well
with ␥-Fe2O3 and all relevant diffraction peaks are well matched
with the reported value for this product [25]. In the present study,
the particle size can be varied by varying the reaction temperature.
The distinguishable difference in the XRD patterns clearly indicates
the temperature dependency of the nanoparticles. It is observed
that on increasing reaction temperature, the particle size increases
and the peaks become more prominent. Thus the XRD peaks for
the product that was obtained at 130 ◦C is very shallow and poorly
resolved, and the XRD pattern of the product synthesized at 250 ◦C
shows well-defined Bragg’s reflections, indicative of comparatively

86
J.-Y. Park et al. / Journal of Molecular Catalysis A: Chemical 323 (2010) 84–90
Fig. 2. TEM images of the size-defined iron oxide: (a) 130-Fe, (b) 180-Fe, (c) 200-Fe, (d) 230-Fe, (e) 250-Fe.
bigger particles. The crystallite size was calculated using Scherer’s
equation [26] taking diffraction peak at 2ꢂ = 36 and is estimated to
be between 2.0 and 12.4 nm (Table 1).
Fig. 2 shows TEM micrographs of prepared iron oxide nanopar-
ticles. TEM images clearly indicate that most of the nanoparticles
are spherical and uniform irrespective of their preparation tem-
perature. Average particle size of iron oxide increases from about
2.0 to 12.0 nm with increasing the reaction temperature and this
finding is similar to the results calculated from the XRD line broad-
ening. The bigger particles are more uniform in their dimensions
than the smaller particles simply because of no agglomeration due
to presence of optimum surfactant amount around the nanoparti-
cles. Also the preparation at higher temperature reveals that the
particles are more crystalline and the lattice fringes can be seen
from the TEM pictures in Fig. 2(e). Table 1 shows the comparison
between the particles calculated from the XRD patterns and those
obtained form TEM analysis.
Table 1
Particle size of the size-defined iron oxide (␥-Fe2O3).
Sample No.
Iron oxide particle size (nm)
XRD
3.2. Analysis of Fe/Al2O3 catalysts derived from pre-synthesized
ꢀ-Fe2O3
TEM
130-Fe
180-Fe
200-Fe
230-Fe
250-Fe
<2.0
4.6
6.2
9.3
12.4
2.0
4.5
6.1
9.0
12.0
Fig. 3 shows the XRD patterns of the supported and calcined
Fe/Al2O3 catalysts. The broad peaks at 2ꢂ values of 20.0, 31.1, 36.6,
39.3, 44.9, 59.9 and 67.3◦ are associated with ␦-Al2O3 (marked with
blank rectangles) [27]. Similarly, reflections at 2ꢂ value of 24.1, 32.9,

J.-Y. Park et al. / Journal of Molecular Catalysis A: Chemical 323 (2010) 84–90
87
Fig. 5. H2-TPR profiles of the size-defined Fe/Al2O3 catalysts: (a) i-FeAl, (b) 130-FeAl,
(c) 180-FeAl, (d) 200-FeAl, (e) 230-FeAl, (f) 250-FeAl.
interaction. Iron oxide nanoparticle on the supported 200-FeAl cat-
alyst shows reasonably homogenous distribution on the ␦-Al2O3
support and has spherical morphology.
Fig. 3. XRD patterns of the size-defined Fe/Al2O3 catalysts: (a) ␦-Al2O3 support, (b)
130-FeAl, (c) 180-FeAl, (d) 200-FeAl, (e) 230-FeAl, (f) 250-FeAl.
TPR studies were performed using H2 on the supported cata-
lysts. Fig. 5 shows the TPR results for the Fe/Al2O3 catalysts with
different sizes of iron oxide. The figure also shows TPR of the cata-
lyst prepared by using impregnation method (i-FeAl). It is reported
that when hematite iron oxide is brought under the reducing envi-
ronment of hydrogen gas, it changes to magnetite form and then
to iron monoxide. The further reduction of iron oxide also makes
it possible to be converted to metallic iron as reported in the lit-
erature [4,5] and as published by us recently [12]. In the present
case, the TPR patterns of Fe/Al2O3 catalysts exhibit two peaks at
about 400 and 560 ◦C, and a broad reduction feature between 600
and 800 ◦C, indicating that the reduction process could occur in two
stages similar to that has been described by others [4,5,12,19,29].
Thus based on the known fact the reduction of such loaded cata-
lyst go through a two-stage reduction profile i.e. in the first-stage
␣-Fe2O3 is reduced to Fe3O4 at 400 ◦C which in the second-stage
undergoes further reduction process to possibly form a mixture of
FeO and Fe or the Fe alone at around 560 ◦C. Since the second-stage
reduction occurs at higher temperature, it may be considered for
conversion of the iron oxide into metallic iron. Such an observation
has also been reported for promoted iron catalyst [4,5,12]. In case
of silica supported iron catalyst, it has been reported that iron sil-
icate [29] may form which hinders the complete reduction of the
iron species. Thus, in the present case third-peak can be associated
with iron oxide layer which contacts on the alumina surface [30].
As the particle size of iron oxide on the Fe/Al2O3 catalyst increases,
the intensity and area of first reduction peak also tend to increase
35.6, 40.8, 49.3, 54.0, 62.3 and 64.0◦ are due to ␣-Fe2O3 (marked
as black rectangles) and match well with the reported data on
hematite iron oxide [28]. It can be seen that some peaks for alumina
and iron oxide are overlapped, and thus the accurate peak width of
iron oxides is difficult to ascertain. After heating Fe/Al2O3 catalysts
in air at 400 ◦C for 5 h, the color of the calcined Fe/Al2O3 cata-
lysts was changed from black to red showing phase change from
original maghemite (␥-Fe2O3) to final hematite (␣-Fe2O3) which is
consistent with the reported observation [15].
Similar to pre-synthesized iron oxide, an increase in peak inten-
sity was observed with increasing particle size of the catalyst even
after calcinations. Careful examination of the XRD pattern shows
more intense and well-defined peaks for 250-FeAl catalyst as com-
pared to that of 130-FeAl. Thus from XRD patterns, it is clear that
the initial difference of particle size of iron oxide continues even
when these have been loaded in support and calcined to tempera-
ture as high as 400 ◦C. In other words, the particle sizes of the iron
oxide between in Fe/Al2O3 catalyst and in as-prepared ␥-Fe2O3 are
similar, which means that there is no significant agglomeration of
iron oxide nanoparticles in Fe/Al2O3 after calcination.
Fig. 4 shows TEM images of ␦-Al2O3 support and supported 200-
FeAl catalyst. The alumina support has fiber-like morphology with
particle size of 20–40 nm in length and 5–10 nm in width as shown
in Fig. 4(a). The ␦-Al2O3 was prepared by the calcinations of ␥-Al2O3
at high temperature of 1000 ◦C to reduce the number of sur-
face hydroxyl group and consequently to minimize metal–support
Fig. 4. TEM images of (a) ␦-Al2O3 support and (b) supported 200-FeAl catalyst.

88
J.-Y. Park et al. / Journal of Molecular Catalysis A: Chemical 323 (2010) 84–90
Table 2
Iron particle size as determined by CO chemisorption and O2 titration of size-defined Fe/Al2O3 catalysts after reduction at 450 ◦C.
Sample
CO uptake (␮ mol/g)
Uncorrected
Uncorrected Fe
diameter (nm)b
O2 uptake
(␮ mol/g)
Reduction
degree (%)c
Corrected dispersion
(%)d
Corrected Fe diameter
(nm)e
dispersion (%)a
i-FeAl
35.8
31.9
24.5
20.7
16.3
13.2
8.4
7.5
5.7
4.9
3.8
3.1
11.4
12.8
16.7
19.8
25.1
31.0
136
142
171
202
215
238
21.3
22.2
26.7
31.6
33.6
37.2
39.5
33.7
21.5
15.4
11.4
8.3
2.4
2.8
4.5
6.2
8.4
130-FeAl
180-FeAl
200-FeAl
230-FeAl
250-FeAl
11.5
a
Dispersion (D) = surface Fe0 atom/total Fe atom × 100, assumed stoichiometric adsorpton ratio of CO/Fe = 1/2.
Fe particle size calculated from CO chemisorptions using d(Fe) = 96/D.
Calculated from O2 uptake.
b
c
d
e
Corrected dispersion (D) = surface Fe0 atom/total reduced Fe0 atom × 100 = surface Fe0 atom/(total Fe atom × reduced fraction) × 100.
Corrected Fe diameter = uncorrected Fe diameter × reduced Fe fraction.
Table 3
and shift to lower temperature. Also, the tail of third-peak moves to
lower temperature with increasing particle size (Fig. 5(a)–(f)). The
increased intensity of first and second peak suggests easy and more
favored reduction for the bigger particle of Fe/Al2O3 catalyst within
a size range of 2–12.0 nm (TEM). The formation of metallic iron may
therefore be easier from the bigger particles. The presence of tail
from the third-peak shows that the reduction of the small particles
is difficult because of the fact that these particles with high sur-
face area have reasonably strong interaction with the surfaces of
alumina support thus hindering the reduction of the smaller par-
ticles that are strongly bound to alumina leading to formation of
iron-aluminate at higher temperature. It is therefore appropriate
to suggest that the different particle size lead to different reduction
degree due to differences in their interaction with alumina. The TPR
studies thus can be summarized as per the following equation:
CO conversion and product distribution at 280 ◦C.a
Sample No. CO conversion
(%)
Conversion of CO
to CO2 (%)
Hydrocarbon selectivity (%)
CH4
C2-C4 C5+
i-FeAl
5.8
16.7
25.3
32.3
21.7
18.8
1.2
2.6
4.4
5.3
3.2
3.0
25.0 38.5
24.3 37.2
19.2 35.6
17.2 33.6
14.3 31.1
13.2 30.2
36.5
38.5
45.2
49.2
54.6
56.6
130-FeAl
180-FeAl
200-FeAl
230-FeAl
250-FeAl
a
Reaction conditions: T = 280 ◦C; Pg = 10 kgf/cm2; SV (L/kgcat/h) = 3600; feed com-
positions (H2/CO/Ar = 63.2/31.3/5.5; mol%).
By comparing the O2 titration of iron oxide on alumina support
at different particle size, it can be seen that the reduction degree
increases with increasing iron particle size. The low reduction
degree of the i-FeAl catalyst can be attributed to the small iron oxide
particle that can easily form iron–aluminate during calcinations or
reduction process. Further the reduction of the iron–aluminate is
extremely difficult and impractical at the adopted experimental
temperature and therefore it is ineffective in the FTS. In addition,
the size-defined Fe/Al2O3 catalyst prepared with small particle may
also have strong metal–support interaction which suppresses the
reducibility. Thus the degree of reduction is different for small par-
ticles and it has maximum value for bigger particles due to the fact
that bigger particles have smaller contact area with the support,
consequently resulting in more reducibility of iron oxide phase to
iron metal. These results well coincide with TPR results of Fig. 5.
Fe2O3 → Fe3O4 → FeO → Fe
(1)
Chemisorption studies were performed using carbon monox-
ide gas which is useful for determination of the particle size for
the reduced species. Table 2 shows the results of CO chemisorp-
tion and O2 titration of the size-defined Fe/Al2O3 catalysts. High CO
consumption is observed for the catalyst prepared by impregnation
method, i.e. i-FeAl as well as for sample 130-FeAl compared to the
catalysts that have bigger particles. The higher CO chemisorption
therefore can mainly be considered because i-FeAl and 130-FeAl
catalysts due to their small particle size, have higher dispersion and
higher surface area of the metallic iron. The calculations of parti-
cle size from O2 titration and CO chemisorptions show that the
size of initial iron oxide estimated from XRD line broadening and
TEM do not differ much from the final Fe/Al2O3 catalyst (Table 2).
Particle size from this calculation is found to be between 2.4 nm
(i-FeAl) and 11.5 nm (250-FeAl). Considering that the oxide shrinks
to about 80% of the size of the original oxide when the nanopar-
ticles of iron oxides are reduced to its metallic particles, the size
of supported iron oxide calculated by the chemisorptions gives a
little higher value especially in the case of smaller particles. Parti-
cle size of unsupported 130-Fe is 2.0 nm (TEM), but after loading
the size of iron metal of 130-FeAl increases to 2.8 nm (chemisorp-
tions). Whereas the size of 250-Fe with bigger particle just changes
marginally from 12.0 nm (TEM) to 11.5 nm (chemisorptions). It
means that smaller particles are more prone to be aggregated and
sintered than bigger particles when it is loaded and calcined.
The O2 titration results suggest different reduction degree for
iron oxide on alumina. As the particle size of iron oxide increases,
oxygen uptake also increases from initial 136 to 238 ␮mol/gcat. The
degree of reduction thus increases from 21.3% for i-FeAl to 37.2% for
250-FeAl with increasing particle size. The reduction degree of the
i-FeAl was lower than that of all Fe/Al2O3 catalysts having nano-
sized iron oxide (Table 2). Overall, the findings suggest that the
oxygen uptake is inversely proportional to the CO uptake.
3.3. FTS performance
3.3.1. Conversion and selectivity
The size-defined Fe/Al2O3 catalysts were tested in a fixed-
bed reactor at 280–300 ◦C at 10 bar under H2/CO = 2 and
GHSV = 3600 ml/(gcat h). The catalytic activities with different par-
ticle size are shown in Tables 3 and 4. The trends of catalytic
Table 4
CO conversion and product distribution at 300 ◦C.a
Sample No. CO conversion
(%)
Conversion of CO
to CO2 (%)
Hydrocarbon selectivity (%)
CH4
C2-C4 C5+
i-FeAl
10.5
24.7
36.2
47.8
32.6
28.1
5.1
7.5
11.8
13.1
9.4
27.8 42.1
26.1 39.0
24.2 38.6
22.9 37.3
20.2 34.7
18.1 33.1
30.1
34.9
37.2
39.8
45.1
48.8
130-FeAl
180-FeAl
200-FeAl
230-FeAl
250-FeAl
8.3
a
Reaction conditions: T = 300 ◦C; Pg = 10 kgf/cm2; SV (L/kgcat/h) = 3600; feed com-
positions (H2/CO/Ar = 63.2/31.3/5.5; mol%).

J.-Y. Park et al. / Journal of Molecular Catalysis A: Chemical 323 (2010) 84–90
89
Fig. 7. The influence of iron particle size on size-defined Fe/Al2O3 catalyst. Reaction
conditions: T = 280 and 300 ◦C; Pg = 10 kgf/cm2; SV (L/kgcat/h) = 3600; feed compo-
sitions (H2/CO/Ar = 63.2/31.3/5.5; mol%).
Fig. 6. Conversion of CO to CO2 by WGS reaction on size-defined Fe/Al2O3 catalyst:
reaction temperature = 280 and 300 ◦C, Pg = 10 kgf/cm2; SV (L/kgcat/h) = 3600; feed
compositions (H2/CO/Ar = 63.2/31.3/5.5; mol%).
performance of the Fe/Al2O3 catalysts at 280 ◦C are similar to results
obtained at 300 ◦C. The CO conversion of Fe/Al2O3 catalyst is clearly
seen to increase with increasing particle size of iron oxide in the
range of between 2.0 and 6.1 nm. On the contrary, the CO con-
version decreased with increasing the size of iron oxide from 6.1
to 12.0 nm. It can be seen from Tables 3 and 4 that by increasing
the particle size from 2.0 to 6.1 nm, the CO conversion of catalyst
increases from 16.7 to 32.3% at 280 ◦C and from 24.7 to 47.8% at
300 ◦C, respectively, and it appears that best catalytic activity can
be attained with particle diameter of 6.1 nm. Both, the CO conver-
sion and CO2 selectivity follow the same trends due to water gas
shift (WGS) reaction as shown in Fig. 6. Formation rate of CO2 by
WGS reaction is linearly correlated with CO conversion because
FT reaction of every CO molecule produces a water molecule. It is
well known that magnetite (Fe3O4) sites in iron-based FT catalyst
are the most active phase for WGS reaction [31]. However, its rate
is promoted at high temperature of 300 ◦C due to reaction kinetic
control.
ity in i-FeAl catalyst because of much smaller size of iron. Since
the difference in catalytic activity between size-defined catalyst
and i-FeAl catalyst is large, it can be easily concluded that the
particle size of iron oxide has great influence on the interaction
between iron oxide and Al2O3, and on the reduction/carburization
and the activity of catalysts. The catalytic performances of size-
defined Fe/Al2O3 catalyst are in agreement with the results of CO
chemisorption. To further verify such difference in the activity and
their size dependency, specific activity is plotted versus the iron
particle size (from chemisorption). Fig. 7 shows that when the iron
particle size increases from 2.0 (130-FeAl) to 6.1 nm (200-FeAl), the
activity increases rapidly due to improved reduction degree while
the activity from 6.1 nm (200-FeAl) to 12.0 nm (250-FeAl) decrease
due to the reduced surface area of Fe metal.
In Co-based FT catalysts, it has been reported that very small
crystallites of Co (<6 nm) have low activity per unit metal site and
high methane selectivity due to structure sensitivity of FT reaction
[1,2]. Therefore in iron-based FT catalysts, a similar behavior might
be expected.
CH4 and C2-C4 selectivity of Fe/Al2O3 catalyst decreases with
increasing particle size from 2.0 to 12.0 nm in other words, C5+
selectivity increases with increasing particle size. The catalyst with
high CH4 selectivity indicates that dissociated hydrogen is present
on the catalyst surface. Because catalyst with smaller particle size
has higher tendency to get hydrogen adsorption rather than dis-
sociative adsorption of CO, they lead to increase population of
hydrogen species at the surface. The large number of hydrogen
species thus present on the surface is hydrogenated to CH4 or light
hydrocarbons [1,2,32]. The increasing C5+ selectivity with increas-
ing particle size of iron oxide could be due to easy activation and
fast movement resulting in easy reaction with the adsorbed CO
molecules. Under similar catalytic condition, the impregnated Fe
catalyst (i-FeAl) show five times lower activity than best Fe cat-
alyst, i.e. 200-FeAl. Such an observation hints towards a definite
size-dependent FTS where the particle size of 6.1 nm is most suited
for effective metal support interaction leading to efficient produc-
tion of hydrocarbon. Particle smaller than this dimension shows
low activity from the fact that their interaction with the support
is so strong that they exist as oxide on the surface of support and
part of the particles only exist as iron carbide in main and metallic
state in part, and thus eventually losing their active sites. Further-
more smaller metal particles with high surface energy are apt to
be oxidized by water produced during the FTS reaction [33]. The
loss of active sites therefore leads to suppression of catalytic activ-
3.3.2. Turnover frequency (TOF)
Since intrinsic activity may be related to the particle size, espe-
cially in FTS reaction, the correlation between TOF and iron particle
size can be a useful way to understand the catalyst efficiency. The
TOF is calculated using the dispersion obtained CO chemisorption
and O2 titration. The TOF data gathered at 280 and 300 ◦C based on
the particle size are plotted as shown in Fig. 8. The TOF increases
rapidly from 0.06 to 0.187 s−1 at 300 ◦C with the increase of the iron
particle sizes in the range of between 2.0 and 6.1 nm and thereafter,
it almost remains constant for the particles ranging from 6.1 to
12.0 nm. It is noteworthy that the TOF of the most active, 200-FeAl
catalyst is 8 times higher than that of i-FeAl (0.023 s−1 at 300 ◦C)
catalyst. This result shows that the size-defined Fe/Al2O3 catalyst is
more efficient and better suited for FTS than conventional catalyst.
The initial increase in turnover frequency with increasing particle
size up to about 6.1 nm could be due to the presence of optimum
active metal content on the surface but after this size domain, the
effect is not so-prominent and the TOF is more or less unaffected by
particle size. These results well coincide with those of cobalt-based
catalysts, where TOF showed a sharp decline below Co particle sizes
of 6 nm [1,2]. It is probably due to the high activation energy of CO
on small metal surface [34].

90
J.-Y. Park et al. / Journal of Molecular Catalysis A: Chemical 323 (2010) 84–90
“Project of next-generation novel technology development” of ITEP.
Dr. P.K. Khanna is thankful to Korea Federation of Science & Tech-
nology (KOSEF) for giving Brain-Pool fellowship.
References
[1] G.L. Bezemer, J.H. Bitter, H.P.C.E. Kuipers, H. Oosterbeek, J.E. Holewijn, X. Xu, F.
Kapteijn, A.J. van Dillen, K.P. de Jong, J. Am. Chem. Soc. 128 (2006) 3956.
[2] J.P. den Breejen, P.B. Radstake, G.L. Bezemer, J.H. Bitter, V. Frøseth, A. Holmen,
K.P. de Jong, J. Am. Chem. Soc. 131 (2009) 7197.
[3] J. Gaube, H.-F. Klein, J. Mol. Catal. A: Chem. 283 (2008) 60.
[4] B.H. Davis, Catal. Today 141 (2009) 25.
[5] M. Luo, R. O’Brien, B.H. Davis, Catal. Lett. 98 (2004) 17.
[6] E. Iglesia, S.L. Soled, R.A. Fiato, J. Catal. 137 (1992) 212.
[7] R. Oukaci, A.H. Singleton, J.G. Goodwin, Appl. Catal. A 186 (1999) 129.
[8] R.C. Reuel, C.H. Bartholomew, J. Catal. 85 (1984) 78.
[9] A. Martinez, C. Lopez, F. Marquez, I. Diaz, J. Catal. 220 (2003) 486.
[10] A.M. Saib, M. Claeys, E. van Steen, Catal. Today 71 (2002) 395.
[11] Y.-J. Lee, J.-Y. Park, K.-W. Jun, J.W. Bae, P.S. Sai Prasad, Catal. Lett. 130 (2009)
198.
[12] S.-H. Kang, J.W. Bae, P.S. Sai Prasad, K.-W. Jun, Catal. Lett. 125 (2008) 264.
[13] G. Prieto, A. Martínez, P. Concepción, R. Moreno-Tost, J. Catal. 266 (2009) 129.
[14] E.I. Mabaso, E. van Steen, M. Claeys, DGMK Tagungsbericht 4 (2006) 93.
[15] V.K. Jones, L.R. Neubauer, C.H. Bartholomew, J. Phys. Chem. 90 (1986) 4832.
[16] M.E. Dry, in: J. Anderson, M. Boudart (Eds.), Catalysis Science & Technology, vol.
1, Springer-Verlag, Berlin, 1981, p. 160.
Fig. 8. The influence of iron particle size on the TOF. Reaction conditions:
T = 280 and 300 ◦C; Pg = 10 kgf/cm2; SV (L/kgcat/h) = 3600; feed compositions
(H2/CO/Ar = 63.2/31.3/5.5; mol%).
4. Conclusions
[17] D.B. Bukur, L. Nowicki, R.K. Manne, X.S. Lang, J. Catal. 155 (1995) 366.
[18] R.M.M. Abbaslou, A. Tavasoli, A.K. Dalai, Appl. Catal. A 355 (2009) 33.
[19] J.F. Bengoa, A.M. Alvarez, M.V. Cagnoli, N.G. Gallegos, S.G. Marchetti, Appl. Catal.
A 325 (2007) 68.
[20] Z. Tao, Y. Yang, M. Ding, T. Li, H. Xiang, Y. Li, Catal. Lett. 117 (2007) 130.
[21] G.B. Raupp, W.N. Delgass, J. Catal. 58 (1979) 337.
[22] M.V. Cagnoli, S.G. Marchett, N.G. Gallegos, A.M. Alvarez, A.A. Yeramián, R.C.
Mercader, J. Catal. 123 (1990) 21.
[23] Y.-J. Lee, K.-W. Jun, J.-Y. Park, H.S. Potdar, R.C. Chikate, J. Ind. Eng. Chem. 14
(2008) 38.
[24] S.M. Kim, Y.-J. Lee, K.-W. Jun, J.-Y. Park, H.S. Potdar, Mater. Chem. Phys. 104
(2007) 56.
[25] T. Taniguchi, K. Nakagawa, T. Watanabe, N. Matsushita, M. Yoshimura, J. Phys.
Chem. C 113 (2009) 839.
[26] Z. Jing, S. Wu, Mater. Lett. 58 (2006) 3637.
[27] Z. Jing, D. Han, S. Wu, Mater. Lett. 59 (2005) 804.
[28] Y. Jin, A.K. Datye, J. Catal. 196 (2000) 8.
[29] C.-H. Zhang, H.-J. Wan, Y. Yang, H.-W. Xiang, Y.-W. Li, Catal. Commun. 7 (2006)
733.
[30] H.-J. Wan, B.-S. Wu, C.-H. Zhang, H.-W. Xiang, Y.-W. Li, B.-F. Xu, F. Yi, Catal.
Commun. 8 (2007) 1538.
[31] H.B. Zhang, G.L. Schrader, J. Catal. 95 (1985) 325.
[32] N.G. Gallegos, A.M. Alvarez, M.V. Cagnoli, J.F. Bengoa, S.G. Marchetti, R.C. Mer-
cader, A.A. Yeramian, J. Catal. 161 (1996) 132.
[33] E. van Steen, M. Claeys, M.E. Dry, J. van de Loosdrecht, E.L. Viljoen, J.L. Visagie,
J. Phys. Chem. B 109 (2005) 3575.
[34] R.A. Van Santen, Acc. Chem. Res. 42 (2009) 57.
The Fe/Al2O3 catalyst prepared from pre-synthesized and size-
defined ␥-Fe2O3 nanoparticles were investigated for the FTS. It is
observed that small iron oxide particles lead to strong interactions
between the support and the iron precursor which shift degree of
reduction to higher temperature. Bigger iron oxide particles can be
reduced easily due to weak interaction with the support but the
bigger particles suppress the dispersion of iron oxide. Thus, there
seems to be an optimum particle size required for FTS. With increas-
ing particle size, CO adsorption of Fe/Al2O3 catalyst decreases and
iron dispersion decreases at the surface of support, whereas the
reduction degree increases with the increase of particle size. The
catalytic activity of sample 200-5Fe/Al (particle size 6.1 nm) is
found to be the best suited for FTS in the present studies. The CO
conversion in the FTS was dependent on particle size, and hydro-
carbon selectivities were strongly affected by the particle size for
catalysts in the size range of 2.0–12.0 nm. At 300 ◦C, TOF increases
from 0.06 to 0.187 s−1 when the particle size of the reduced catalyst
is increased from 2.0 to 6.1 nm. It can therefore be concluded that
particle size of iron oxide seems to be important for the catalytic
activity and selectivity of Fe/Al2O3 catalysts.
Acknowledgements
The authors would like to acknowledge the financial support
from the Korea Ministry of Knowledge Economy (MKE) through