Fischer–Tropsch study over impregnated silica-supported cobalt–iron nanocatalyst

Article abstract of DOI:10.1007/s13738-016-0975-y

Fischer–Tropsch reaction was carried out over the impregnated 60?%Co/40?%Fe/90wt%SiO₂/1.5wt%K catalyst in a fixed-bed micro-reactor over a range of operating conditions. Reaction rate expression on the basis of the Langmuir–Hinshelwood–Hougen–Watson type models for the FT reactions was derived. Here seven kinetic expressions for CO consumption were proposed, and interaction between adsorption CO and dissociated adsorption hydrogen as the controlling step gave the most conceivable kinetic model. According to one factor at a time method, the product distributions in FT synthesis are found to be strongly influenced by temperature and pressure, and maximum hydrocarbon selectivity C₂–C₃ light olefins are obtained at 260?°C and 8?bar. The activation energy for CO consumption is calculated as 106.2?kJ?mol^?1.

Full text of DOI:10.1007/s13738-016-0975-y

J IRAN CHEM SOC
DOI 10.1007/s13738-016-0975-y
ORIGINAL PAPER
Fischer–Tropsch study over impregnated silica‑supported
cobalt–iron nanocatalyst
Mohammad Abdollahi1 · Hossein Atashi1 · Farshad Farhchi Tabrizi1,2
Mohsen Mansouri3
·
Received: 7 April 2016 / Accepted: 9 September 2016
© Iranian Chemical Society 2016
Abstract Fischer–Tropsch reaction was carried out over
the impregnated 60 %Co/40 %Fe/90wt%SiO2/1.5wt%K
catalyst in a fixed-bed micro-reactor over a range of oper-
ating conditions. Reaction rate expression on the basis of
the Langmuir–Hinshelwood–Hougen–Watson type mod-
els for the FT reactions was derived. Here seven kinetic
expressions for CO consumption were proposed, and inter-
action between adsorption CO and dissociated adsorption
hydrogen as the controlling step gave the most conceivable
kinetic model. According to one factor at a time method,
the product distributions in FT synthesis are found to
be strongly influenced by temperature and pressure, and
maximum hydrocarbon selectivity C2–C3 light olefins are
obtained at 260 °C and 8 bar. The activation energy for CO
Introduction
Fischer–Tropsch synthesis (FTS) is the most promising
development option for the environmentally sound pro-
duction of hydrocarbons and fuels from synthesis gas at
elevated pressures and temperatures due to the large verity
of products such as C2–C3 alkenes, gasoline and diesel that
can be obtained [1, 2]. Selective formation of such distribu-
tion of products by FTS requires a deep molecular study of
the CO hydrogenation reactions over a wide range of cata-
lysts [3]. Iron and cobalt are the most promising FTS cata-
lysts, and the combination of Co–Fe catalysts leads to Co–
Fe alloy formation which favors the formation of alkenes
with or without promoters [4–7]. Interest in the develop-
ment of more selective and active Co–Fe catalyst and more
effective FTS process technologies and reactors increased
dramatically in the past decades [2, 8]. The FTS process
with Co catalysts is not complex due to the irreversible and
expensive changes in the catalyst; however with Fe cata-
lyst, the physical and chemical nature of the FTS with iron
catalyst is significantly improved by the wax and carbon
deposition due to the formation of carbides and magnetite
[9]. The FTS-preferred catalysts for hydrocarbon produc-
tion is cobalt-based catalysts due to their low activity and
selectivity for the water-gas shift reaction and high activ-
ity and selectivity for long-chain paraffins [10]. The past
researches on the FTS catalysts in the patent literature indi-
cate that the Fe catalyst could influence quite dramatically
by the addition of small amounts of Co which enhance the
activity of this catalyst in comparison with the other metal
oxides and the Co–Fe combination is the most promising
catalyst for light olefin production from synthesis gas [3,
4]. According to the past researches, a few catalysts are
suitable to improve the C2–C3 fraction due to the thermo-
dynamic and kinetic limitations of the FTS reactions, and
consumption is calculated as 106.2 kJ mol−1
.
Keywords Fischer–Tropsch synthesis · Co–Fe catalyst ·
Kinetics modeling · Light olefins
* Hossein Atashi
h.ateshy@hamoon.usb.ac.ir; atashi.usb@gmail.com
* Mohsen Mansouri
mansouri2010@yahoo.com; m.mansouri@ilam.ac.ir
1
Department of Chemical Engineering, University of Sistan
and Baluchestan, Zahedan 98164-161, Iran
2
Department of Chemical Engineering, School of Chemical
and Petroleum Engineering, Shiraz University, Shiraz, Iran
3
Department of Chemical Engineering, University of Ilam,
Ilam 69315-516, Iran
1 3

J IRAN CHEM SOC
some examples such as iron- and/or cobalt-based catalysts
on partially reducible oxide supports such as TiO2 instead
of the conventional supports such as Al2O3 are reported in
the literature [3]. The general effects of promoters on the
iron catalyst behavior have been studied in many researches
in the past decade, and the most essential promoters in Fe
catalysts for the FTS process are the group IA metals such
as potassium which have obvious effect on both the activity
and the selectivity of Fe catalysts [3].
mechanism over cobalt–iron catalyst under various operat-
ing conditions and (5) assessment of operating variables
such as reaction temperatures and pressures on the cata-
lytic performance. We also characterized the precursors and
calcined catalysts (before and after the test) using various
techniques such as XRD, SEM, EDS and BET.
Experimental
The FTS kinetic has been studied significantly, and
many researchers have been attempting to postulate the rate
of FTS reaction using a Langmuir–Hinshelwood–Hougen–
Watson (LHHW) or a power law rate equation [11–14]. An
overview of the rate expressions developed for Fe catalysts
is given by Huff and Satterfield [15] and for Co catalysts
has been summarized by Yates and Satterfield [16]. For
cobalt catalysts, the first researches show that the CO con-
sumption is proportional to the rate of desorption of hydro-
carbon chains growing on the catalyst surface where this
growing chain concentration is empirically associated with
the second order of hydrogen partial pressure and first order
of CO partial pressure [17]. Huff and Satterfield [15] and
Outi et al. [18] declared that the rate of CO consumption
can be modeled by assuming the rate of the monomer for-
mation as the rate-controlling step for a reduced fused mag-
netite and a cobalt catalyst in the same way. This assump-
tion indicates that all elementary reactions except the
controlling step are close to equilibrium and the developed
rate equation describes the experimental results practically
well in certain ranges of experimental conditions [17].
Studies giving the activity/selectivity relationships,
product distributions and kinetic expressions for CO hydro-
genation on impregnated silica-supported cobalt–iron cata-
lysts are not to be found in the literature. Mirzaei et al. [3]
reports the effect of operating conditions such as reaction
temperatures and pressures and hydrogen to CO molar ratio
in the catalytic performance on the SiO2-supported Fe–Co
catalyst prepared by sol–gel route. However, it is particu-
larly important to know that no work on FTS reaction over
impregnated potassium-promoted Co–Fe/SiO2 catalyst has
been done and consequently its kinetic study has not been
reported up to now. Thus, in the present work, for the first
time, we attempted to investigate the kinetic and mecha-
nism of the CO hydrogenation on the SiO2-supported Fe–
Co catalyst promoted by potassium, which was prepared by
impregnation procedure.
Catalyst preparation
Using conventional method of impregnation,
a
60 %Co/40 %Fe/90wt%SiO2/1.5wt%K catalyst, which has
a high value of olefin/paraffin ratio and the chain growth
probability, was prepared as discussed in this section. The
catalysts were prepared by incipient impregnation of SiO2
with aqueous cobalt nitrate (Co(NO3)2·6H2O) (0.5 M)
(99 %, Merck) and iron nitrate (Fe(NO3)2·6H2O) (0.5 M)
(99 %, Merck) and potassium nitrate (KNO3) (0.5 M)
(99 %, Merck) solutions. The SiO2 support was first cal-
cined at 600 °C in flowing air for 6 h before impregnation.
For 60 %Co/40 %Fe/90wt%SiO2/1.5wt%K catalyst, the
solution of proper amount of cobalt, iron and potassium
nitrate was prepared and was directly dispersed through
a spray needle onto the support. The impregnated support
was then dried at 120 °C for 16 h. In order to obtain the
final catalyst, the precursor was then calcined at 550 °C for
6 h.
Characterization techniques
Powder X-ray diffraction (XRD) measurements were
performed using a Bruker AXS Company, D8 Advance
diffractometer (Germany). Scans were taken with a 2θ
step size of 0.02 and a counting time of 1.0 s using Cu
Kα radiation source generated at 40 kV and 30 mA.
Specimens for XRD were prepared by compaction into
a glass-backed aluminum sample holder. Data were col-
lected over a 2θ range from 4° to 70°, and phases were
identified by matching experimental patterns to entries
in the Diffracplus Version 6.0 indexing software iden-
tified phases. The average size of the nanocrystallites
was estimated by using Debye–Scherrer formula. Par-
ticle size (nm) = 0.9 λ/β cosθ, where λ, β and θ are
the X-ray wavelength, the full width at half maximum
(FWHM) of the diffraction peak and the Bragg diffrac-
tion angle, respectively. The BET surface area (BET)
was measured using a N2 adsorption–desorption iso-
therm at liquid nitrogen temperature (−196 °C), using a
NOVA 2000 instrument (Quantachrome, USA). Prior to
the adsorption–desorption measurements, all of the sam-
ples comprising precursors and calcined catalysts were
Therefore, the main objectives of the present work are:
(1) preparation of SiO2-supported Fe–Co/K nanocatalyst
using impregnation method, (2) testing of this prepared
catalysts at the operational conditions for CO hydrogena-
tion, (3) investigation of the structure and physicochemi-
cal properties of this catalyst, (4) establishing of kinetic
models for the FTS reaction systemically based on detailed
1 3

J IRAN CHEM SOC
Fig. 1 Schematic representa-
tion of the reactor used. 1 Gas
cylinders, 2 pressure regulators,
3 needle valves, 4 mass flow
controllers, 5 monometers, 6
non-return valves, 7 ball valves,
8 tubular furnace, 9 reactor,
10 catalyst bed, 11 trap, 12
condenser, 13 silica gel column,
and 14 gas chromatograph (GC)
degassed at 200 °C in a N2 flow for 3 h to remove the
moisture and other adsorbates. The morphology of cata-
lysts and their precursors was observed by means of an
S-360 Oxford Eng scanning electron microscopy (USA).
Also, the elemental composition was determined using
the EDS (the energy-dispersive X-ray spectroscopy)
(IE300X) analysis.
of catalyst weight shows that external diffusion can be
neglected above this GHSV. Therefore, the kinetic experi-
ments were investigated free from internal and external
mass-transfer limitations.
All steady-state kinetic experimental data were collected
in the differential fixed-bed micro-reactor with a maximum
conversion below 25 % to minimize temperature and con-
centration gradients using the following operating condi-
tions: Experiments were conducted with mixtures of CO,
H2 and nitrogen in a temperature range from 230 to 280 °C,
H2/CO feed ratios of 1–1.5, pressure range of 4–16 bar and
GHSV = 4500 h−1. The data of catalytic test are repre-
sented in Table 1. The stability of the catalyst was inves-
tigated by reiteration of central points of the designs in the
middle and the end of the experiments. When we choose to
consider the rate to be constant at all points within the reac-
tor we have a differential flow reactor. The CO conversion
and selectivity of olefin and paraffin products account from
Eqs. (1) and (2), respectively.
Catalyst testing
Figure 1 shows a schematic of the experimental setup. The
kind of fixed-bed micro-reactor that made for Fischer–
Tropsch synthesis is stainless steel with an inner diameter
of 20 mm. Three mass flow controllers (Brooks Model,
5850E) were utilized to adjust automatically the flow
rate of the inlet gases comprising CO, H2 and N2 (purity
of 99.99 %). Mixture of CO, H2 and N2 was subsequently
entered into the reactor, which was placed inside a tubu-
lar furnace (Atbin, Model ATU 150-15). Temperature of
the reaction was measured and controlled by a thermocou-
ple inserted into the catalytic bed and visually monitored
by a computer. To achieve the isothermal conditions in a
catalytic bed, the catalyst was diluted with an inert mate-
rial (quartz). The catalyst was in situ pre-reduced at atmos-
pheric pressure under H2–N2 flow (N2/H2 = 1, flow rate
of each gas = 30 ml min−1), at 400 °C for 16 h. In every
experiment 1.0 g catalyst was loaded, and to ensure attain-
ing the steady-state operating conditions the reactor oper-
ated about 12 h.
Moles of COin − Moles of COout
CO conversion (%) =
× 100
Moles of COin
(1)
Moles of carbon in product j
Selectivity of j product (%) =
× 100
Moles of converted CO
(2)
The experimental reaction rate was calculated as
follows:
Because the catalyst was extremely fine particles, intra-
particle diffusion could be neglected. Increasing of the
gas hourly space velocity (GHSV) to the value in which
the CO conversion was almost the same for a variety
Rate of CO conversion =
(Fractional conversion) × (Input flow rate of CO)
Weight of the catalyst
(3)
1 3

J IRAN CHEM SOC
Table 1 Summary of experimental conditions and results for the kinetic tests at PTot = 4–16 bar, H2/CO = 1–1.5 and GHSV = 4500 h−1
Number of data
T (°C)
PH (bar)
PCO (bar)
FCO (mmol min−1
)
XCO (%)
−rCO × 10 (mmol
2
gr−1 cat. min−1
)
1
230
240
250
270
230
240
230
241
260
270
250
260
273
280
277
280
283
240
250
258
240
250
260
240
250
260
270
230
240
250
270
1.534
1.557
1.585
1.37
1.547
1.555
1.563
1.485
3.087
3.082
4.979
4.632
5.997
5.945
1.023
0.992
0.976
0.975
1.928
2.869
3.721
1.216
1.195
1.099
2.388
2.366
2.205
3.601
3.506
3.248
3.274
4.858
4.779
4.707
4.303
2.774
2.734
2.695
2.466
5.535
5.418
8.927
8.127
10.147
9.874
1.764
1.678
1.612
1.590
3.161
4.679
6.036
2.138
2.061
1.867
4.198
4.080
3.731
6.330
6.046
5.496
5.438
8.710
8.401
8.117
7.147
3.161
2.749
7.436
16.92
3.470
3.699
2.338
3.493
6.295
10.149
19.882
22.388
23.569
23.895
24.565
24.254
24.00
5.078
6.686
14.092
6.763
7.573
13.872
6.2
0.877
0.752
2.00
4.18
1.92
2.00
2.09
2.84
6.39
10.02
3.51
3.76
3.80
3.80
7.77
11.81
14.48
1.09
1.38
2.63
2.84
3.09
5.18
3.92
5.26
8.48
9.98
4.43
5.59
6.51
11.39
2
3
4
5
3.088
3.075
4.928
4.802
6.028
5.862
0.983
0.944
0.84
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
0.845
1.634
2.83
3.213
1.967
1.915
1.663
3.819
3.757
3.34
5.713
5.457
4.887
4.829
7.836
7.553
7.378
6.383
8.702
15.423
18.351
5.081
6.659
8.025
15.949
Development of kinetic equations
The carbide or carbene mechanism was proposed by
Fischer and Tropsch in 1926 [23]. In this mechanism,
adsorbed CO is separated to C and O, and the carbide is
then hydrogenated to the monomer (CHx). The methylene
monomer polymerizes to alkyl species surface that termi-
nate to products. It is widely supported despite the fact
that it does not account for the formation of oxygenates.
The hydoxycarbene or enol mechanism was proposed by
Storch et al. in the 1950s [24]. Dissociative adsorption of
H2 and molecular adsorption of CO followed by the hydro-
genation of adsorbed CO with adsorbed hydrogen, form an
oxygenated intermediate. This intermediate species, react
with another adsorbed hydrogen, form water and adsorbed
carbon. This reaction pathway is known as carbide mecha-
nism. While this mechanism illustrates the formation of
The experiments were performed at pressures between 4 and
16 bar, temperatures between 230 and 280 °C and H2/CO
feed ratio between 1/1 and 1.5/1 to study the kinetic of FTS
for hydrocarbons formation over the impregnated catalyst
of 60 %Co/40 %Fe/90wt%SiO2/1.5wt%K based on cata-
lyst weight in order to drive a appropriate expression for CO
consumption. Table 1 shows the operating conditions and
some results. These results are used in rate equation expres-
sion for FTS reaction. The studied FTS kinetic by researcher
has been presented and many mechanistic schemes have
been proposed, but all the proposed mechanisms in the lit-
erature [19–25], the surface carbide and enolic mechanisms
have been noted more than other mechanisms.
1 3

J IRAN CHEM SOC
Table 2 Elementary reaction set for Fischer–Tropsch synthesis
distribution pathways for the determination of kinetic mod-
els. For each model is summarized in Table 2 an elementary
reactions set on sites. In order to derive each kinetic model,
initially one of the elementary reactions was assumed as a
rate-determination step and all other steps were considered
at equilibrium. Therefore, all of the obtained models were
against experimental data fitted separately.
The development of the kinetic equations will be illus-
trated for model FT-III-3 (the third step elementary reaction
from the FT-III model). These model codes ascribe to the
set of elementary reactions, and the elementary reaction is
not at equilibrium (that is the rate-determining step, so in
this case reaction 3). The reaction rate of the rate-determin-
ing step is:
Model
FT-I
No
Elementary reaction
1
2
3
4
5
6
7
8
1
2
3
4
5
1
2
3
4
5
6
7
1
2
3
4
5
1
2
3
4
5
6
7
8
1
2
3
4
CO + s ↔ COs
COs + s ↔ Cs + Os
H2 + 2s ↔ 2Hs
Cs + Hs ↔ HCs + s
HCs + Hs ↔ H2Cs + s
Os + Hs → HOs + s
HOs + Hs → H2Os + s
H2Os ↔ H2O + s
FT-II
CO + s ↔ COs
COs + s ↔ Cs + Os
Cs + H2 ↔ H2Cs
Os + H2 → H2Os
−rFT-III-3 = k3 θCOθH
(4)
H2Os ↔ H2O + s
θCO is the surface fraction filled with the associative
adsorbed carbon monoxide and θH is the surface fraction
filled with the dissociative adsorbed hydrogen. Here the
assumption is that only surface of CO occupies a signifi-
cant fraction of the total number of sites. The fraction of
vacant sites, θS, can be calculated from the following bal-
ance equation:
FT-III
CO + s ↔ COs
H2 + 2s ↔ 2Hs
COs + Hs ↔ HCOs + s
HCOs + Hs ↔ Cs + H2Os
Cs + Hs ↔ CHs + s
CHs + Hs ↔ CH2s +s
H2Os ↔ H2O + s
FT-IV
FT-V
CO + s ↔ Cos
θS + θCO = 1
(5)
COs + s ↔ Cs + Os
Cs + H2 ↔ H2Cs
The surface fractions of CO and H can be calculated
from the site balance, the preceding reaction steps which
are at quasi-equilibrium:
Os + H2 → H2O + s
H2 + 2s ↔ 2Hs
k1
CO + s ←→ COs
CO + s ↔ Cos
H2 + 2s ↔ 2Hs
Cos + Hs ↔ HCOs + s
HCOs + Hs ↔ Cs + H2O + s
Cs + Hs ↔ CHs + s
CHs + Hs ↔ CH2s + s
Os + Hs → HOs + s
HOs + Hs → H2O + 2 s
CO + s ↔ COs
k1 PCOθS − k−1 θCO = 0
(6)
θCO = K1PCOθS
(7)
k1
K1 =
k−1
FT-VI
where K1 is the equilibrium constant of CO adsorption step.
COs + H2 → CHs + OHs
CHs + H2 + s ↔ CH2s + Hs
OHs + Hs → H2Os + s
k2
H2 + 2s ←→ 2Hs
k2 PH θS2 − k−2 θH2 = 0
(8)
2
oxygenates and was supported by Kummer et al. [25] who
used C-alcohols or alkenes as a co-feed and remarked that
these alcohols participated in the chain growth, neverthe-
less, the details of the chemistry of this mechanism are
unclear.
For the derivation of the model equations tested these two
mechanistic cases were considered. Competitive adsorption
on similar sites has been taken into account in both cases.
Six mechanisms were offered on the basis of various mon-
omer formation (elementary reactions) and carbon chain
θH = K20.5 PH0.5θS
(9)
2
k2
K2
=
k−2
where K2 is the equilibrium constant of dissociated hydro-
gen adsorption step.
Substituting Eq. (7) in Eq. (5), the concentration of free
active site can be expressed as:
1 3

J IRAN CHEM SOC
Table 3 Reaction rate expression proposed for FTS
Model
FT-I-4
Kinetic equation
kPH0.25
(1+aPC0.O5 +bPH0.25
0.5
P
CO
−RCO
−RCO
−RCO
=
=
=
2
)
kPH0.275
0.5
P
CO
FT-I-5
−0.25
0.5
CO
+bPH0.25
)
2
(1+aP PH
2
0.5
P
CO
kPH2
0.5
FT-II-3
FT-III-3
FT-IV-3
(1+aP +bPCO
)
CO
kPCOPH0.25
−RCO
−RCO
=
=
2
(1+aPCO
)
0.5
P
CO
Fig. 2 XRD patterns for both precursor and calcined catalysts con-
taining 60 %Co/40 %Fe/90 wt% SiO2/1.5 wt% K
kPH2
(1+aPCO+bPC0.O5 +cPH0.25
)
kPH2 PCO
(1+aPCO+bPH0.25
FT-V-4
−RCO
=
=
2
)
catalyst during the reaction and to detect the phases formed,
the catalyst after the test was characterized by XRD and its
phases were found to be C (hexagonal), Fe3Si (cubic), FeO
(cubic), CoO (cubic), Fe2C (monoclinic), Fe (cubic). Dur-
ing chemical reaction, some of the oxidic phases in the cal-
cined catalyst before the test transformed into metallic and
carbide phases; in the tested catalyst, there are oxidic and
carbide phases, both of which are active phases in the FTS.
Carbide phase is active in CO hydrogenation. Formation
of this phase with oxidic phases which are active to olefins
cause high performance of supported SiO2 catalyst [26].
The scanning electron microscopy (SEM) observations
have shown differences in morphology of precursor and
calcined catalysts before and after the reaction at 8 bar and
260 °C. The electron micrograph obtained from catalyst
precursor characterizes agglomerations of particles and
illustrates that this material comprise of particles with dif-
ferent size with none of defined geometrical shapes; this
material has a high density and homogeneous dispersion of
particles (Fig. 3a). The morphological feature was different
to the precursor sample after the calcinations at 550 °C and
showed that the agglomerate size was reduced compared
with the precursor (Fig. 3b). Therefore, the calcined cata-
lyst has disproportion agglomerate before the test. How-
ever, by agglomeration the size of the tested grains grew
larger (Fig. 3c), which may be due to sintering after the
reactions. In previous study by Galarrage [27], he observed
that temperature could cause agglomeration of these small
grains, which correlates with catalyst deactivation under
high temperature.
0.5
kP PH
FT-VI-2
CO
2
−RCO
0.5
CO
(1+aP
)
1
θS =
(10)
1 + K1PCO
By substitution of the surface fraction of CO and H in
Eq. (4), the final rate expression is obtained as follows:
0.5
k3 K1K20.5PCOP0.5
k P
P
CO
H2
H2
−rFT-III-3
=
=
(11)
2
2
(1 + K1PCO
)
(1 + aPCO)
Table 3 summarizes the final form of the different rate
expressions for the 7 possible kinetic models.
Results and discussion
Catalyst screening results
Characterization of 60 %Co/40 %Fe/90 wt% SiO2/1.5 wt%
K catalyst was carried out using XRD, and the patterns
of this catalyst in different stages of precursor, fresh cal-
cined catalyst (before the test) and used calcined catalyst
(after the test) at H2/CO = 1.25:1 are shown in Fig. 2. The
catalyst precursor was found to be amorphous. The pres-
ence of amorphous phases in the XRD pattern of the pre-
cursor makes the other phases undetectable. However, the
calcined catalyst before the test showed different phases,
and the actual phases identified in this catalyst were Co3O4
(cubic), Co2SiO4 (cubic), Fe2SiO4 (cubic), Fe3O4 (cubic),
CoFe2O4 (cubic). XRD patterns of calcined catalyst before
the test showed broadening of the peaks indicative of the
ultra-fine nature of the crystallite. The average value of the
crystal size in the potassium-promoted Co–Fe/SiO2 nano-
catalyst was determined to be about 21 nm by using Scher-
rer equation. In order to identify the changes in the calcined
The BET surface area measurement was used in order to
measure the specific surface area of the catalysts. The spe-
cific surface area of the catalyst tested at 8 bar and 260 °C
that is compared with the precursor and calcined catalysts
before the test is presented in Table 4. It is found that the
calcined catalysts before the test have higher specific sur-
face area (105.8 m2 g−1) than their precursor (97.2 m2 g−1);
this is the same SEM results which showed the agglomer-
ate size of calcined catalyst is less than its precursor and so
1 3

J IRAN CHEM SOC
Fig. 3 SEM images of 60 %Co/40 %Fe/90wt%SiO2/1.5wt%K catalyst in a precursor, b catalyst before the test and c catalyst after the test (H2/
CO = 1.25 at 8 bar and 260 °C)
Table 4 BET surface area (m2
calcined catalysts (before and after reactor test at 8 bar and 260 °C)
g
−1) results for both precursor and
samples. The elemental compositions of the precursor,
fresh and used calcined catalysts obtained from the EDS
data are given in Table 5. The EDS spectra of the precur-
sor and calcined samples are also illustrated in Fig. 4. The
EDS spectrum of the precursor showed the presence of N,
O, Co, Fe and Si (Table 5; Fig. 4a), which indicates that the
precursor comprised phases with hydroxyl and nitrate com-
pounds. The EDS spectrum of the fresh calcined catalyst
refers to the presence of O, Co, Fe and Si (Table 1; Fig. 4b),
which confirms that these materials are comprised of oxidic
phases of cobalt and iron. This is supported by the X-ray
diffraction data, which showed the existence of different
oxide phases of cobalt iron. The EDS spectrum of the cal-
cined catalyst after the test showed the presence of C, O,
Co, Fe and Si (Table 5; Fig. 4c). This obtained result for the
used calcined catalyst is also in good agreement with the
XRD data and confirms the existence of different metallic,
oxidic and carbidic phases in the used calcined catalyst.
Precursor
97.2
Calcined catalyst
(before reaction)
Calcined catalyst
(after reaction)
105.8
92.4
Table 5 EDS data of 60 %Co/40 %Fe/90wt%SiO2/1.5wt%K catalyst
in precursor and calcined states (before and after the test)
Catalyst
Elements
Weight %
Atomic %
Precursor
N
8.21
29.59
4.31
11.71
37.12
2.65
O
Co
Fe
Si
K
2.92
1.72
54.16
0.81
46.28
0.52
Calcined catalyst
(before the test)
O
31.62
4.89
42.07
3.11
Co
Fe
Si
K
Experimental results
3.12
2.25
59.25
1.12
51.84
0.73
One of the main factors with a marked effect on the cata-
lytic performance of the catalysts for FTS is the operational
conditions. Conventional FT catalyst screening consists of
comparing the catalytic performance at the same experi-
mental conditions (temperature, pressure, space velocity
and amount of catalyst or active phase). The measured out-
put values of catalyst screening are typically overall activ-
ity (syngas conversion), selectivities (methane, light and
heavy hydrocarbons, carbon dioxide), stability and cata-
lyst behavior during the start-up activation steps [29]. The
choice of the operating conditions and reactor represents
the major challenge of FT catalyst screening. The effect
of a range of operation variables such as reaction tempera-
ture and reaction pressure on the catalytic performance of
SiO2-supported Fe–Co/K catalyst was investigated in order
to identify and optimize the operational conditions for
FTS using OFAT method. The typical reaction results are
Calcined catalyst
(after the test)
C
8.13
15.12
33.18
3.92
O
26.63
4.49
Co
Fe
Si
K
3.16
2.31
56.67
0.92
44.59
0.88
leads to an increase in the BET specific surface area of the
calcined catalyst. The high specific surface area of calcined
catalyst before the test allows a high degree of metal dis-
persion [28].
The EDS characterization was also carried out at dif-
ferent stats of catalyst (precursor, fresh and used calcined)
in order to determine the mean element proportions of the
1 3

J IRAN CHEM SOC
Fig. 4 EDS spectrums of
potassium-promoted Fe–Co/
SiO2 catalyst in a precursor, b
calcined catalyst before the test
and c calcined catalyst after
the test
compared with respect to product selectivity and the forma-
tion of methane.
Effect of temperature on product selectivity
The catalyst was reduced in atmospheric pressure
in a flow stream of H2 with flow rate of 30 ml min−1 at
300 °C for 12 h, and then it was exposed to the stream
of H2 + CO. Eighty-three kinetic tests were carried out
over 60 %Co/40 %Fe/90wt%SiO2/1.5wt%K catalyst. The
results at each particular condition are the calculated aver-
age value containing at least two measurements, and the
average value was considered as characteristic data.
Effects of reaction temperature on the catalytic perfor-
mance of the 60 %Co/40 %Fe/90wt%SiO2/1.5wt%K pre-
pared using initial impregnation procedure was studied
from the schemes of product selectivity versus temperature
and showed results at P = 8 bar, H2/CO = 1.25 (Fig. 5) and
P = 12 bar, H2/CO = 1.25 (Fig. 6).
According to the obtained results (Figs. 5, 6) from OFAT
method, the optimum reaction temperature was 260 °C, at
1 3

J IRAN CHEM SOC
Fig. 7 Effect of pressure on product selectivity at T = 260 °C and
H2/CO = 1.25
Fig. 5 Effect of temperature on product selectivity at P = 8 bar and
H2/CO = 1.25
performance of the cobalt iron oxide catalyst containing
60 %Co/40 %Fe/90wt%SiO2/1.5wt%K for production of
light olefins at GHSV = 4500 h−1. The product selectivi-
ties at different temperatures of 260 and 280 °C and H2/
CO = 1.25 are plotted as a function of pressure (Figs. 7, 8).
All these selectivity parameters reached steady-state values
following a short initial period. Methane and ethane selec-
tivities remained unchanged or even slowly decreased, but
the C+5 selectivity was increased with increasing total reac-
tion pressure. The results indicate that at the total pressure
of 12 bar, the catalyst showed a total selectivity of 54 %
with respect to C2–C3 light olefins. However, the light
olefins selectivities were changed and the results indicate
that at the total pressure of 8 bar and 260 °C, the catalyst
showed the highest total selectivity of 57 % with respect to
C2–C3 light olefins.
Fig. 6 Effect of temperature on product selectivity at P = 12 bar and
H2/CO = 1.25
which temperature of the total selectivity of C2–C3 light
olefin products was higher than those at the other reac-
tion temperatures under the same operating conditions.
The optimum temperature and pressure for different prod-
ucts are summarized in Table 6. The methane selectivity
increased with increasing in temperature at constant pres-
sure and H2/CO = 1.25 feed ratio (Figs. 5, 6). The maxi-
mum selectivity with respect to ethylene and propylene
occurs in 260 and 280 °C, respectively, as shown in Figs. 5,
6 and Table 6.
Kinetic study: kinetic parameter estimation
The kinetic data presented in Table 1 for CO conversion
were used for testing the seven models listed in Table 3.
The discrimination of the kinetic models and the estima-
tion of the kinetic parameters were performed by fitting
the experimental data of the components partial pressure to
the kinetics equations (Table 3). The unequal treatment of
the values of the kinetic parameters through best-fit model
(FT-III-3) was better determined using the Levenberg–Mar-
quardt algorithm by a multivariable nonlinear regression
method. To minimize the sum of the square of residuals
corresponding to difference between the experimental data
The effect of pressure on product selectivity
Number of tests were carried out to examine the
influence of the reaction pressure on the catalyst
Table 6 Optimum condition
for product distribution in H2/
C5+
Product
CH4
C2H4
C2H6
C3H6
C3H8
C4H8
C4H10
CO = 1.25
T (°C)
280
4
260
8
260
4
280
12
260
8
250
8
270
12
260
16
P (bar)
Selectivity (%)
16.6
32.55
14.95
26.17
22.77
15.0
10.12
14.06
1 3

J IRAN CHEM SOC
ꢁ
ꢃ
ꢂ
ꢂ
ꢂ
ꢂ
Nexp
ꢀ
exp
ri − rical
1
Nexp
cal
MARR =
× 100
(15)
riexp
i=1
exp
CO,i
r
and r
show the experimental and calculated CO
CO,i
conversion rate obtained from each kinetic model in ith
data point, and Nexp explain the number of experimental
data points with pure error variance ρ. To select the most
suitable kinetic expression, different statistical indices can
also be used to determine the quality of regression models
via the Polymath software 6.0. To find the most appropriate
model, the following conditions should be considered [30]:
achieved constants must be positive; coefficients of the
equation must obey from rules of Arrhenius and Van’t Hoff;
optimal model is the one which gives the lowest MARR.
According to the statistical results obtained by inserting
the data and models, the best model can be selected. Table 7
shows the statistical indicators for the FT kinetic models.
Based on the statistical criteria and by comparing the values
of R2 and RMSD, it was recognized that FT-III-3 is the most
appropriate model. Also, it can be seen that FT-III-3 model
has the lowest MARR among the other models that has been
used in the present study. This value is reasonable and shows
that the predicted values are 16.43 % different from the
observed values. Therefore, there are best fitted by a LHHW
Fig. 8 Effect of pressure on product selectivity at T = 280 °C and
H2/CO = 1.25
was the objective function and those calculated for the
kinetic models. The R2 value (reflects the amount of vari-
ance), root mean square deviation (RMSD) and mean abso-
lute relative residual (MARR) have been reported as meas-
ure of the goodness of fit:
Nexp
ꢀ
1
exp
r
CO,i
ρ =
(12)
Nexp
i=1
approach rate form −rCO = k PCOP
0.5/(1 + a PCO)2 where
H2

2
Nexp
�
�
�
2
the controlling step of the reaction rate is formation of mon-
omer CHO intermediate. Model FT-III-3 shows that domi-
nant mechanism on catalyst surface is based on dissociation
of hydrogen with associative carbon monoxide and form-
ing methyl monomers (enolic mechanism), which is same
as the previous study done by the Wojciechowski [31]. In
previous research on bimetallic cobalt catalysts (titania-
supported Co–Mn catalyst), formation of the monomer CH2
was done by reaction of adsorbed CO and hydrogen in two
steps that was assumed as dominant mechanism. Also based
on research carried out by Keyser et al. [32], the enolic
mechanism are much better than carbide mechanism for
bimetallic Co–Mn oxide catalysts. As it is shown in Table 7,
rCexOp,i − r
cal
CO,i





 i=1

R2 = 1 −
(13)
Nexp

�
�
�
2
rCexOp,i − ρ
i=1
and (RMSD) is described as:



2
Nexp
�
�
�
2
1
exp
cal

RMSD =
r
− r
CO,i
CO,i
(14)
Nexp
i=1
The MARR between experimental and calculated con-
sumption rate of CO is defined as:
Table 7 Estimated values of the statistical indicators for the FT kinetic models
Model
Kinetic parameter
Statistical indicator
E (kJ mol−1) R2 RMSD (10−5
k0 (x) (mmol g−1
a0 (x) (barx)
b0 (x) (barx)
c0 (x) (barx)
)
MARR (%)
min−1 barx)
FT-I-4
FT-I-5
6.002E + 02 (−1)
0.4958 (−1/2)
−0.0791 (−1/2)
2.060 (−1/2)
0.299 (−1)
92.41
91.94
0.929 1.67
18.28
18.54
18.37
16.43
18.89
17.87
22.82
1.089E + 04 (−5/4) 3.263 (−1/4)
0.921 1.76
0.924 1.73
0.942 1.49
0.925 1.73
0.933 1.62
0.849 2.66
FT-II-3 1.352E + 03 (−3/2) 1.103 (−1/2)
FT-III-3 9.758E + 03 (−3/2) 15.241 (−1)
91.79
106.02
FT-IV-3
FT-V-4
7.78E + 07 (−3/2) −2.2E + 05 (−1) −1.3E + 04 (−1/2) 2.E + 04 (−1/2) 91.26
9.12E + 06 (−2)
2.08 (−1)
2.88 (−1/2)
91.89
FT-VI-2 2.49E + 10 (−3/2) 5.33E + 06 (−1/2)
107.32
1 3

J IRAN CHEM SOC
3. A.A. Mirzaei, A. Beigbabaei, M. Galavy, A. Youssefi, A silica
supported Fe–Co bimetallic catalyst prepared by the sol/gel tech-
nique: operating conditions, catalytic properties and characteri-
zation. Fuel Process. Technol. 91, 335–347 (2010)
4. C. Cabet, A.C. Roger, A. Kiennemann, S. Läkamp, G. Pourroy,
Synthesis of new Fe–Co based metal/oxide composite materi-
als: application to the Fischer–Tropsch synthesis. J. Catal. 173,
64–73 (1998)
5. F. Tihay, A.C. Roger, A. Kiennemann, G. Pourroy, Fe–Co based
metal/spinel to produce light olefins from syngas. Catal. Today
58, 263–269 (2000)
6. V.A. de la Peña O’Shea, M.C. Alvarez-Galvan, J.M. Campos-
Martin, J.L.G. Fierro, Fischer–Tropsch synthesis on mono-and
bimetallic Co and Fe catalysts in fixed-bed and slurry reactors.
Appl. Catal. A 326, 65–73 (2007)
7. V.A. de la Peña O’Shea, M.C. Alvarez-Galvan, J.M. Campos-
Martin, N.N. Menendez, J.S.D. Tornero, J.L.G. Fierro, Surface
and structural features of Co–Fe oxide nanoparticles deposited on
a silica substrate. Eur. J. Inorg. Chem. 2006, 5057–5068 (2006)
8. A.N. Pour, S.M.K. Shahri, H.R. Bozorgzadeh, Y. Zamani, A.
Tavasoli, M.A. Marvast, Effect of Mg, La and Ca promoters on
the structure and catalytic behavior of iron-based catalysts in Fis-
cher–Tropsch synthesis. Appl. Catal. A 348, 201–208 (2008)
9. R. Zennaro, M. Tagliabue, C.H. Bartholomew, Kinetics of Fis-
cher–Tropsch synthesis on titania-supported cobalt. Catal. Today
58, 309–319 (2000)
the activation energy for the best-fitted model (FT-III-3)
was found to be 106.2 kJ mol−1 which is close to activa-
tion energies reported previously: 100 and 103 kJ mol−1
reported by Yang et al. [33] and Storch et al. [24], respec-
tively. Nevertheless, it was substantially lower than the
value of 142 kJ mol−1 reported by Reuel and Bartholomew
[34]. Reuel and Bartholomew’s value achieved at lower sig-
nificantly reactant partial pressure based on two data points,
while in this study it was based on thirty-one data points. It
suggested that the activation energy obtained for hydrocar-
bon formation for the diffusion interference is not signifi-
cant in the tests [35]. The presence of external mass-transfer
limitations could be discerned via measuring the apparent
activation energy with intraparticle diffusion limitations. A
control regime of external mass-transfer could lead to the
obvious energy activation of just a few kJ mol−1 [29].
Conclusion
An active silica-supported cobalt–iron nanocatalyst was pre-
pared by impregnation method and was tested for hydro-
genation of carbon monoxide to light olefins. The kinetic
experimental study was performed in a differential fixed-bed
micro-reactor by altering reaction pressure (4–16 bar), H2/
CO feed molar ratio (1–1.5) and space velocity (4500 h−1) at
the temperature range of 230–280 °C. Considering LHHW
adsorption theory in catalytic reactions, CO consumption rate
equations were defined by 6 mechanisms; consequently, 7
kinetic models were proposed. Essential kinetic data obtained
in the initial rate region illustrate that the enolic mechanism
with interaction between adsorption CO and separated adsorp-
tion hydrogen as the rate-controlling step provides the most
plausible kinetic model. The kinetic parameters estimated for
this kinetic model showed reasonable confidence intervals.
The OFAT method was applied to investigation of temperature
and pressure influence on the conversion of syngas into a wide
spectrum of hydrocarbons in FTS. The product distributions in
FT synthesis are strongly influenced by temperature and pres-
sure, and 260 °C and 8 bar are the optimum temperature and
pressure for obtaining C2–C3 light olefins hydrocarbons with
high selectivity. 106.2 kJ mol−1 was the activation energy for
the best-fitted model; this value of activation energy approves
that intraparticle mass transport is not significant.
10. A.A. Mirzaei, B. Shirzadi, H. Atashi, M. Mansouri, Modeling
and operating conditions optimization of Fischer–Tropsch syn-
thesis in a fixed-bed reactor. J. Ind. Eng. Chem. 18, 1515–1521
(2012)
11. M. Mansouri, H. Atashi, N. Setareshenas, Detailed Kinetic Study
of the FTS Over the Co-precipitated Co–Ce/SiO2 (Lambert Aca-
demic Publishing, Saarbrüchen, 2013)
12. H. Atashi, M. Mansouri, S.H. Hosseini, M. Khorram, A.A. Mir-
zaei, M. Karimi, G. Mansouri, Intrinsic kinetics of the Fischer–
Tropsch synthesis over an impregnated cobalt-potassium cata-
lyst. Korean J. Chem. Eng. 29, 304–309 (2012)
13. M. Mansouri, H. Atashi, F.F. Tabrizi, G. Mansouri, N. Setaresh-
enas, Fischer–Tropsch synthesis on cobalt–manganese nanocata-
lyst: studies on rate equations and operation conditions. Int. J.
Ind. Chem. 5, 1–9 (2014)
14. G.P. van der Laan, A.A.C.M. Beenackers, Catalysis reviews—
science and engineering. Catal. Rev. Sci. Eng. 41, 255–318
(1999)
15. G.A. Huff Jr., C.N. Satterfield, Intrinsic kinetics of the Fischer–
Tropsch synthesis on a reduced fused-magnetite catalyst. Ind.
Eng. Chem. Process Des. Dev. 23, 696–705 (1984)
16. I.C. Yates, C.N. Satterfield, Intrinsic kinetics of the Fischer–Tropsch
synthesis on a cobalt catalyst. Energy Fuels 5, 168–173 (1991)
17. E. van Steen, H. Schulz, Polymerisation kinetics of the Fischer–
Tropsch CO hydrogenation using iron and cobalt based catalysts.
Appl. Catal. A 186, 309–320 (1999)
18. A. Outi, I. Rautavuoma, H.S. van der Baan, Kinetics and mecha-
nism of the Fischer–Tropsch hydrocarbon synthesis on a cobalt
on alumina catalyst. Appl. Catal. 1, 247–272 (1981)
19. M. Mansouri, H. Atashi, F.F. Tabrizi, A.A. Mirzaei, G. Mansouri,
Kinetics studies of nano-structured cobalt–manganese oxide
catalysts in Fischer–Tropsch synthesis. J. Ind. Eng. Chem. 19,
1177–1183 (2013)
References
20. C.G. Visconti, E. Tronconi, L. Lietti, R. Zennaro, P. Forzatti,
Development of a complete kinetic model for the Fischer–
Tropsch synthesis over Co/Al2O3 catalysts. Chem. Eng. Sci. 62,
5338–5343 (2007)
21. T. Botao, C. Jie, W. Haijun, L. Jiqing, Z. Shaocheng, L. Ya, L.
Ying, G. Xiaohui, A corrected comprehensive kinetic model of
Fischer–Tropsch synthesis. Chin. J. Catal. 28, 687–695 (2008)
1. H. Atashi, F. Siami, A.A. Mirzaei, M. Sarkari, Kinetic study of
Fischer–Tropsch process on titania-supported cobalt–manganese
catalyst. J. Ind. Eng. Chem. 16, 952–961 (2010)
2. M. Mansouri, H. Atashi, A.A. Mirzaei, R. Jangi, Kinetics of the
Fischer–Tropsch synthesis on silica-supported cobalt–cerium
catalyst. Int. J. Ind. Chem. 4, 1–10 (2013)
1 3

J IRAN CHEM SOC
22. J. Yang, Y. Liu, J. Chang, Y. Wang, L. Bai, L. Xu, H. Xiang, Y. Li,
B. Zhong, Detailed kinetics of Fischer–Tropsch synthesis on an
industrial Fe–Mn catalyst. Ind. Eng. Chem. Res. 42, 5066–5090
(2003)
29. A.Y. Khodakov, W. Chu, P. Fongarland, Advances in the devel-
opment of novel cobalt Fischer–Tropsch catalysts for synthesis
of long-chain hydrocarbons and clean fuels. Chem. Rev. 107,
1692–1744 (2007)
23. F. Fischer, H. Tropsch, The synthesis of petroleum at atmos-
pheric pressures from gasification products of coal. Brennst.
Chem. 7, 97–104 (1926)
30. G. Bercic, J. Levec, Intrinsic and global reaction of methanol
dehydration over γ-Al2O3 pellets. Ind. Eng. Chem. Res. 31,
1035–1040 (1992)
24. H.H. Storch, N. Golumbic, R.B. Anderson, The Fischer–Tropsch
Synthesis and Related Synthesis (Wiley, New York, 1951)
25. J.T. Kummer, H.H. Podgurski, W.B. Spencer, P.H. Emmett,
Mechanism studies of the Fischer–Tropsch synthesis: the addi-
tion of radioactive alcohol. J. Am. Chem. Soc. 73, 564–569
(1951)
26. H.B. Zhang, G.L. Schrader, Characterization of a fused iron cat-
alyst for Fischer–Tropsch synthesis by in situ laser Raman spec-
troscopy. J. Catal. 5, 325–332 (1985)
27. C.E. Galarrage, Heterogeneous catalyst for the synthesis of mid-
dle distillate hydrocarbons. M.Sc. Thesis. University of Western
Qntario, London, 1998
28. R. Bechara, D. Balloy, J.-Y. Dauphin, J. Grimblot, Influence
of the Characteristics of γ-aluminas on the dispersion and the
reducibility of supported cobalt catalysts. Chem. Mater. 11,
1703–1711 (1999)
31. B.W. Wojciechowski, The kinetics of the Fischer–Tropsch syn-
thesis. Catal. Rev. Sci. Eng. 30, 629–702 (1988)
32. M.J. Keyser, R.C. Everson, R.L. Espinoza, Fischer–Tropsch
kinetic studies with cobalt–manganese oxide catalysts. Ind. Eng.
Chem. Res. 39, 48–54 (2000)
33. C.H. Yang, F.E. Massoth, A.G. Oblad, Kinetics of CO + H2 reac-
tion over Co–Cu–Al2O3 catalyst. Adv. Chem. Ser. 178, 35–46
(1979)
34. R.C. Reuel, C.H. Bartholomew, Effects of support and dispersion
on the CO hydrogenation activity/selectivity properties of cobalt.
J. Catal. 85, 78–88 (1984)
35. A. Sari, Y. Zamani, S.A. Taheri, Intrinsic kinetics of Fischer–
Tropsch reactions over an industrial Co–Ru/γ-Al2O3 catalyst
in slurry phase reactor. Fuel Process. Technol. 90, 1305–1313
(2009)
1 3