Effect of cobalt metal loading on Fischer–Tropsch synthesis activities over Co/γ-Al2O3 catalysts: CO conversion, C5+ productivity, and α value

Article abstract of DOI:10.1007/s11164-019-03839-8

This study investigates the effect of cobalt loading (5–20?wt% Co) on the physical properties of the alumina support and cobalt particle size of Co/γ-Al₂O₃ catalysts for Fischer–Tropsch synthesis (FTS) reaction (T = 210?°C (set), P = 20?bar, H₂/CO = 2). To characterize the catalysts and correlate these characteristics with their catalytic activities in FTS, N₂ adsorption, inductively coupled plasma, X-ray diffraction (XRD), transmission electron microscopy (TEM), and X-ray photoelectron spectroscopy (XPS) studies were conducted. N₂ adsorption and XRD analyses showed that as the cobalt loading was increased up to 15?wt%, the number of cobalt oxide particles increased by the direct interaction between cobalt oxide and the alumina support surface, but that when the cobalt loading was further increased to 20?wt%, the particle size of the cobalt oxide increased abruptly due to the additional cobalt loading on the previously loaded cobalt oxide. These physical properties of the supported cobalt catalysts were attributed to the pore structure of the alumina support. From the TEM micrographs, the size of cobalt particles was roughly estimated to increase from 20 to 50?nm at a cobalt loading up to 15?wt% Co/γ-Al₂O₃ to 70–100?nm at 20?wt% cobalt loading. For the 5–15?wt% Co/γ-Al₂O₃ catalysts, CO conversion and C₅₊ liquid oil productivity increased with increasing cobalt loading because they were strongly proportional to the number of cobalt particle active sites. However, the 20?wt% Co/γ-Al₂O₃ catalyst showed the highest value of α because the larger cobalt particles increased the opportunity for chain growth. The XPS data also supported the greater reducibility of the cobalt species in 20?wt% Co/γ-Al₂O₃ and hence its larger size compared to that at low cobalt loading.

Full text of DOI:10.1007/s11164-019-03839-8

Research on Chemical Intermediates
https://doi.org/10.1007/s11164-019-03839-8
Efect of cobalt metal loading on Fischer–Tropsch synthesis
activities over Co/γ‑Al2O3 catalysts: CO conversion, C5+
productivity, and α value
Jung‑Il Yang1 · Chang Hyun Ko2
Received: 10 July 2018 / Accepted: 27 April 2019
© Springer Nature B.V. 2019
Abstract
This study investigates the efect of cobalt loading (5–20 wt% Co) on the physi-
cal properties of the alumina support and cobalt particle size of Co/γ-Al2O3 cata-
lysts for Fischer–Tropsch synthesis (FTS) reaction (T=210 °C (set), P=20 bar,
H2/CO=2). To characterize the catalysts and correlate these characteristics with
their catalytic activities in FTS, N2 adsorption, inductively coupled plasma, X-ray
difraction (XRD), transmission electron microscopy (TEM), and X-ray photoelec-
tron spectroscopy (XPS) studies were conducted. N2 adsorption and XRD analyses
showed that as the cobalt loading was increased up to 15 wt%, the number of cobalt
oxide particles increased by the direct interaction between cobalt oxide and the
alumina support surface, but that when the cobalt loading was further increased to
20 wt%, the particle size of the cobalt oxide increased abruptly due to the additional
cobalt loading on the previously loaded cobalt oxide. These physical properties of
the supported cobalt catalysts were attributed to the pore structure of the alumina
support. From the TEM micrographs, the size of cobalt particles was roughly esti-
mated to increase from 20 to 50 nm at a cobalt loading up to 15 wt% Co/γ-Al2O3 to
70–100 nm at 20 wt% cobalt loading. For the 5–15 wt% Co/γ-Al2O3 catalysts, CO
conversion and C5+ liquid oil productivity increased with increasing cobalt loading
because they were strongly proportional to the number of cobalt particle active sites.
However, the 20 wt% Co/γ-Al2O3 catalyst showed the highest value of α because
the larger cobalt particles increased the opportunity for chain growth. The XPS data
also supported the greater reducibility of the cobalt species in 20 wt% Co/γ-Al2O3
and hence its larger size compared to that at low cobalt loading.
Keywords Fischer–Tropsch synthesis · Co/γ-Al2O3 · Cobalt loading · CO
conversion and C5+ liquid oil productivity · α value
*
*
Jung-Il Yang
yangji@kier.re.kr
Chang Hyun Ko
chko@jnu.ac.kr
Extended author information available on the last page of the article
Vol.:(0123456789)
1 3

J.-I. Yang, C. H. Ko
Introduction
Clean liquid fuels can be manufactured from synthesis gas, a mixture of H2 and
CO, which is produced from reforming, and/or shift reaction, or gasifcation of
methane, coal, and biomass [1]. Potential world oil shortage is strongly pushing
the development of alternative technologies for the manufacture of transporta-
tion fuels. Furthermore, as these fuels contain no aromatics, sulfur, or nitrogen-
containing compounds, their high purity will easily satisfy the upcoming stricter
environmental regulations in both Europe and the USA [2, 3].
Fischer–Tropsch synthesis (FTS) is used to convert synthesis gas into synthetic
liquid fuels and is, therefore, a crucial technology for gas-to-liquid (GTL) pro-
cess. Although many metals exhibit FTS activity, only cobalt (Co) and iron (Fe)
catalysts are in industrial use [4, 5]. These two catalysts share the following simi-
larities in FTS: they are very active, produce a wide range of long-chained hydro-
carbons, and aford a product distribution that follows the Anderson–Shultz–Flory
equation [5]. However, cobalt is active in its Co metal form and iron is active as
Fe carbides [5, 6]. Cobalt catalysts, although much more expensive than iron, are
used in smaller amounts and can be recycled for several years by regeneration
(oxidation/reduction) [5, 7]. Cobalt catalysts have been used for FTS because of
their high activity towards the production of long-chained hydrocarbons, wax,
and long catalyst life about 10-fold longer than Fe [7–9]. Currently, cobalt cata-
lysts are used in commercial GTL plants of Oryx GTL by Sasol/QP and Pearl
GTL by Shell [10]. In the FTS reaction, Sasol has developed a Co-based slurry
bubble column reactor (SBCR) with a capacity of 34,000 bpd and Shell has been
developing a Co-based fxed-bed reactor (FBR) with a capacity of 140,000 bpd
[10, 11]. Therefore, there are strong eforts for the optimization and tailoring of
cobalt catalysts for specifc applications [1].
Martinez et al. reported the infuence of cobalt loading, cobalt precursor,
and promoters on the physico-chemical and catalytic properties of Co/SBA-15
catalysts for the FTS reaction [12]. The Co/SBA-15 catalysts supported the pre-
ferred formation of C5+ hydrocarbon because of its high reducibility, while the
less reducible, low-cobalt-loading sample exhibited a product distribution shifted
toward the formation of lighter hydrocarbons (methane, C2–C4). Storsœter et al.
investigated the performance of unpromoted and Re-promoted cobalt catalysts
supported on γ-Al2O3, SiO2, and TiO2 in terms of mass transfer limitations and
the structural parameter, χ [13]. They reported that the χ values for these catalysts
were all in the range 7–52 × 1016 m−1 and that the reaction rates and selectivities
should not be limited by the difusion of CO to the active sites. Furthermore,
although the reactants were in the gas phase, the support pores were flled with
liquid products, and difusion in the liquid phase was several orders of magni-
tude slower than that in the gas phase. Barbier et al. observed the size sensitiv-
ity of cobalt particles in the FTS reaction, in which the intrinsic activity and the
chain growth probability, α, frstly increased and then stabilized with increasing
particle size [14]. They also reported the efcient control of cobalt reducibility
as one of the key issues in the design of highly active cobalt alumina-supported
1 3

Efect of cobalt metal loading on Fischer–Tropsch synthesis…
FTS catalysts. Blanchard et al. reported the successful use of the cobalt catalyst
for the production of alcohols so long as both metallic cobalt and cobalt oxide
were in a close association within the cobalt catalyst [15]. Furthermore, consid-
ering CO difusional restrictions within the catalyst, an eggshell cobalt catalyst
was developed for increasing FTS rates and C5+ selectivity [16]. Further research
has examined the dependency of the FTS activity on the reducibility of cobalt
catalysts. Jacobs et al. investigated the support, loading, and promoter efects on
the reducibility of the catalysts and found that alumina support, increased cobalt
loading, and addition of Ru, Pt, and Re were favorable for improving the reduc-
ibility of the catalysts [17]. The promoting efect of MnO was attributed to the
lower degree of cobalt reduction because the manganese retards cobalt reduction
[18]. The chain growth probability increased, and the product distribution shifted
toward olefnic products at increased MnO loadings. Furthermore, the catalyst life
and contaminants tolerance of the cobalt catalysts used in FTS are important due
to their relatively high cost. Although the oxidation of nano-sized metallic cobalt
to cobalt oxide during FTS has been considered a major deactivation mechanism
[19–21], van de Loosdrecht reported that oxidation can be ruled out as a major
deactivation mechanism [22]. Oxidation has also been eliminated as a potential
deactivation mechanism of cobalt crystallites ≥ 6 nm supported on alumina. In
their recent research on the poisoning of cobalt catalyst used for FTS, Sparks
et al. reported that a H2S level of about 500 ppb was the threshold for a detectable
decrease in activity [23]. Recently, Farzanfar et al. reported inorganic complex
precursor route for preparation of high-temperature Fischer–Tropsch synthesis
Ni–Co nanocatalysts and they showed that the novel procedure was more advan-
tageous than impregnation and co-precipitation methods for the preparation of
efective and durable cobalt catalysts for Fischer–Tropsch synthesis [24]. Also,
Hajjar et al. investigated the efect of applying nanoporous graphene to Fis-
cher–Tropsch synthesis catalysts [25]. The novel catalysts, Co/graphene oxide
and Co/nanoporous graphene, yielded heavier hydrocarbons compared with the
Co/Al2O3 catalyst, possibly due to the high surface area and intrinsic properties
of the carbon nanostructures as efective hydrogen carriers.
This work investigates the catalytic activities of Co/γ-Al2O3 catalysts in FTS accord-
ing to the cobalt loading. In particular, N2 adsorption, inductively coupled plasma
(ICP), X-ray difraction (XRD), transmission electron microscopy (TEM), and X-ray
photoelectron spectroscopy (XPS) studies are conducted to characterize the cobalt cata-
lysts according to the cobalt loading and to correlate these catalyst characteristics with
their catalytic activities in FTS. Especially, the infuence of cobalt loading (5–20 wt%)
on the distribution of liquid oil products is discussed in terms of the chain growth prob-
ability, α. Finally, we investigate the efects of the number and size of cobalt active sites
in the cobalt catalysts on the FTS activities in terms of CO conversion, C5+ productiv-
ity, and α value.
1 3

J.-I. Yang, C. H. Ko
Experimental
Preparation of catalysts
Metallic cobalt supported on γ-Al2O3 (Co/γ-Al2O3) supports were prepared as cata-
lysts at four cobalt concentrations (5, 10, 15, and 20 wt%) by the wet impregnation
of cobalt nitrate (Co(NO3)2, Aldrich) on γ-Al2O3. After impregnation, these sam-
ples were calcined at 400 °C for 8 h to obtain cobalt oxide supported on γ-Al2O3
(Co3O4/γ-Al2O3). Prior to reaction, these samples were pre-reduced in situ with a
hydrogen fow of 300 ml/min for 16 h at 350 °C to obtain Co/γ-Al2O3.
Characterization of catalysts
XRD patterns for Co/γ-Al2O3 and Co3O4/γ-Al2O3 catalysts were obtained in the
2θ range from 20° to 80° at intervals of 0.02° using a Rigaku Difractometer (D/
MAX Ultima III) with Cu Kα radiation at 40 kV and 40 mA. TEM images of the
Co/γ-Al2O3 catalysts were obtained by Technai F30S-Twin (FEI) microscopy with
an operating voltage of 300 kV. The surface areas of catalysts were measured by
N2 adsorption at 77 K using ASAP2010 (micrometrics). The cobalt composition in
the catalyst was measured by ICP-atomic emission spectroscopy (ICP-AES) using
Optima 4300™ DV ICP-OES (PerkinElmer). The oxidation states of the cobalt in
the catalyst were investigated by XPS (MultiLab2000 by VG) using AlKα radiation
(1486.6 eV).
Catalytic experiments
The FTS reaction was carried out in a conventional, tubular FBR with a diameter of
25.4 mm and a length of 300 mm. A schematic diagram of the experimental appa-
ratus is shown in Fig. 1. Reaction heat was supplied by an electrical furnace with
a PID controller (Hanyoung Science). The reaction temperature was monitored by
a thermocouple (K type, nickel-chromium vs. nickel-aluminum, Omega Engineer-
ing, Inc.) installed at the lower end of the catalyst bed through the inlet of the reac-
tor. Mass fow controllers (Brooks, 5850E) were used for precise fow control of the
reactants (H2/CO=2 molar ratio) under pressurized reaction conditions by a back
pressure regulator (Tescom). The liquid products were completely recovered by a
cold trap working with a cold ethylene glycol (0 °C) circulating system, which was
fnally collected from the product reservoir. The total fow of gas products and unre-
acted reactants was measured by using a dry gas meter (Shinogawa, 1 L/rev) and the
products gas was sampled to analyze the gas composition by using an on-line sam-
pler located after the dry gas meter.
Reactant and product analysis
The product distribution in the liquid products was analyzed by a simulated distil-
lation (SIMDIS) technique using the ASTM D2887 method. A gas chromatograph
1 3

Efect of cobalt metal loading on Fischer–Tropsch synthesis…
Fig. 1 Schematic diagram of the experimental apparatus for Fischer–Tropsch synthesis (FTS) reaction
(Agilent 6890 series) equipped with a SIMDIS 2884 column (Analytical Controls)
and an auto injector system was used for analysis. The compositions of both gas
products and reactants were analyzed by using a gas chromatograph (HP5890 GC)
equipped with a Carboseive II column. Furthermore, the details of reactant and
product analysis have been comprehensively explained in our previous study [26].
Results and discussion
Efect of cobalt loading on the physical properties of the supported catalysts
Table 1 lists the chemical composition and textural properties (BET surface area,
total pore volume, and average pore diameter), as obtained by N2 adsorption, of pure
γ-alumina and the four supported cobalt catalysts according to the cobalt loading.
The cobalt loading infuenced the surface area and pore volume of the catalysts. The
surface area and pore volume of the catalyst linearly decreased as the cobalt load-
ing was increased from 5 to 15 wt%, which Martinez et al. attributed to a partial
blockage of the support pores by cobalt oxides cluster and/or partial collapse of the
porous structure with the increase of cobalt loading. Therefore, the increased cobalt
loading may be related with the interaction between cobalt oxide and the alumina
support surface [12]. On the other hand, as the cobalt loading was increased from at
the 15–20 wt%, the surface area, pore volume, and average pore diameter remained
1 3

J.-I. Yang, C. H. Ko
Table 1 Chemical composition and textural properties obtained by N2 adsorption of pure γ-alumina and
supported cobalt catalyst
Catalyst
Co (wt%)
SBET (m2/g)
Total pore volume
(cm3/g)
Average pore
diameter (nm)
γ-Al2O3
–
5.0
144
122
0.22
0.20
60
65
5 wt%
Co/γ-Al2O3
10 wt%
9.4
119
100
100
0.18
0.16
0.16
61
64
64
Co/γ-Al2O3
15 wt%
12.5
16.6
Co/γ-Al2O3
20 wt%
Co/γ-Al2O3
unchanged, which indicated that the 15 wt% cobalt loading was sufcient for the
complete coverage of alumina surface with cobalt oxide and the complete flling of
pores in alumina support. Increasing the cobalt loading to 20 wt%, did not further
afect the interaction between the cobalt oxide and the surface of the alumina sup-
port, but it did increase the size of the previously loaded cobalt oxide particles in the
alumina support.
Efect of cobalt loading on cobalt particle size
As metallic cobalt is the active site for FTS reaction, the particle size of metallic
cobalt is an important factor for catalytic activity [27]. However, measuring the par-
ticle size of metallic cobalt in a catalyst is relatively difcult because cobalt is imme-
diately oxidized to cobalt oxide upon air exposure, even in ambient condition. Thus,
the cobalt particle size was estimated indirectly by measuring the size of the cobalt
oxide particles after the calcination step in the preparation of Co/γ-Al2O3. As shown
in Fig. 2, Co3O4 was the only phase detected in the XRD patterns for the cobalt
species after this calcination step, and the other peaks were assigned to γ-Al2O3.
The full width at half maximum (FWHM) peak for cobalt oxide refects the particle
size of the corresponding cobalt oxide. Storsœter et al. used the most intense Co3O4
peak, 2θ=36.9°, to calculate the Co3O4 particle size [13]. Especially, the XRD pat-
tern of 5 wt% Co/γ-Al2O3 presented only small broad cobalt oxide peaks in a pat-
tern that was very similar to that of the pure γ-Al2O3 support. As the cobalt loading
was increased to 15 wt%, the Co3O4 peak of 2θ=36.9° clearly appeared and its size
gradually increased, which was mainly attributed to the increased number of cobalt
oxide particles as the cobalt loading was increased to 15 wt% in the alumina pores
because of the interaction between cobalt oxide and the alumina support surface.
However, when the cobalt loading was further increased to 20 wt%, the peak of
cobalt oxide at 2θ=36.9° sharply increased, as shown in Fig. 2e. Therefore, it can
be deduced that the particle size of cobalt oxide increased abruptly at this 20 wt%
cobalt loading due to the efect of this additional cobalt loading on the previously
loaded cobalt oxide.
1 3

Efect of cobalt metal loading on Fischer–Tropsch synthesis…
4500
4000
3500
3000
2500
(e)
(d)
2000
#
1500
#
#
#
#
^
^
(c)
(b)
#
1000
500
0
(a)
20
30
40
50
60
70
80
2θ [ degree ]
Fig. 2 X-ray difraction (XRD) patterns for a γ-alumina, b 5 wt% Co/γ-Al2O3, c 10 wt% Co/γ-Al2O3, d
15 wt% Co/γ-Al2O3, and e 20 wt% Co/γ-Al2O3 after calcination at 400 °C for 8 h. Phases denoted are (#)
cobalt oxide (Co3O4), and (^) γ-alumina
3600
3000
2400
1800
1200
600
+
*
^
(d)
(c)
!
^
^
+
*
(b)
(a)
0
20
30
40
50
60
70
80
2θ [ degree ]
Fig. 3 X-ray difraction (XRD) patterns for a 5 wt% Co/γ-Al2O3, b 10 wt% Co/γ-Al2O3, c 15 wt% Co/γ-
Al2O3, and d 20 wt% Co/γ-Al2O3 after FTS reaction. Phases denoted are (*) metallic cobalt, (+) cobalt
carbide (CoC8), (!) aluminum cobalt carbide (AlCo3C), and (^) γ-alumina
1 3

J.-I. Yang, C. H. Ko
After the FTS reaction, the Co/γ-Al2O3 catalysts were retained for further XRD
measurement. Figure 3 displays the XRD patterns of the four Co/γ-Al2O3 cata-
lysts after FTS reaction. Small and broad peaks corresponding to metallic cobalt
appeared. The peaks’ intensity and broadness (FWHM) suggested that 20 wt% Co/γ-
Al2O3 had the largest metallic cobalt particles.
These results obtained from the XRD patterns agreed well with the previous BET
results. The BET analysis suggested that the cobalt particle size should be sharply
increased at 20 wt% cobalt loading due to the saturation of the alumina pore struc-
ture at the 15 wt% cobalt loading, i.e., the optimum cobalt loading was 15 wt% for
particle size control. Finally, the trend in the cobalt particle size was more clearly
explained by using the XRD analysis.
TEM images of the Co/γ-Al2O3 catalysts after FTS reaction confrmed the
Co particle size, which was deduced from the XRD results. As shown in Fig. 4,
the cobalt particle size, actually the metallic cobalt particles covered by cobalt
Fig. 4 Transmission electron microscopic (TEM) images of supported cobalt catalysts: a 5 wt% Co/γ-
Al2O3, b 10 wt% Co/γ-Al2O3, c 15 wt% Co/γ-Al2O3, and d 20 wt% Co/γ-Al2O3 after FTS reaction
1 3

Efect of cobalt metal loading on Fischer–Tropsch synthesis…
carbide, increased with increasing cobalt loading. The cobalt particles were
roughly estimated to be sized 20–50 nm. Especially, the cobalt particle size
increased drastically to 70–100 nm at 20 wt% cobalt loading. The cobalt particles
began to agglomerate during the catalyst preparation due to the high cobalt pre-
cursor concentration.
Fig. 5 Catalytic activities of 5 wt% Co/γ-Al2O3 (a) and CO conversion of 5 wt%, 10 wt%, 15 wt%,
and 20 wt% Co/γ-Al2O3 catalysts (b) in a fxed bed reactor (FBR) for the FTS reaction (H2/CO=2,
Ftotal =100 ml/min, Wcat =4.5 g, P=2.0 MPa)
1 3

J.-I. Yang, C. H. Ko
Performance of Co/γ‑Al2O3 catalysts
The catalytic activity of Co/γ-Al2O3 at the four cobalt loadings, was investigated
in an FBR under typical FTS conditions of 2.0 MPa, 210 °C (set), H2/CO=2, and
Wcat/F=0.75 gcat h L−1. Figure 5 shows the performance of the 5 wt% Co/γ-Al2O3
catalyst during 80 h reaction. As shown in Fig. 5, the reaction temperature initially
increased sharply to 236 °C due to the exothermic reaction heat generated in the FTS
reaction, resulting in both high CO conversion and high CH4 selectivity. After 30 h
reaction, the following steady-state reaction activities were obtained: 36.4% CO con-
version, 17.0% CH4 selectivity, and 1.52% CO2 selectivity at 217 °C. Furthermore,
the C5+ selectivity was 36.5% and the C5+ liquid oil productivity reached 30.3 mL
−1
kg h−1. The chain growth probability, α, was calculated to be 0.839 based on the
cat
distribution of liquid oil products. The contact time (Wcat/F) of 0.75 gcat h L−1 for
the 5 wt% Co/γ-Al2O3 catalyst in the FTS reaction was confrmed as being accept-
able for C5+ liquid oil production at 217 °C. It has previously been reported that the
contact time is a critical point for controlling the internal mass transfer limitation
induced by the produced hydrocarbons and that the liquid hydrocarbons produced
during the short contact time block the pores of the catalyst, leading to catalyst deac-
tivation [2, 28].
Table 2 summarizes the catalytic activities of Co/γ-Al2O3 in the FTS reaction
at the four cobalt loadings. CO conversion increased as the cobalt loading was
increased up to 15 wt%, but then decreased at the higher cobalt loading of 20 wt%.
This pattern of increasing and decreasing CO conversion was mainly ascribed to
Table 2 Efect of cobalt loading
Catalyst
5 wt%
10 wt%
15 wt%
20 wt%
on the catalytic activities of
Co/γ-Al2O3 for FTS reaction in
a fxed bed reactor (FBR)
Co/γ-Al2O3 Co/γ-Al2O3 Co/γ-Al2O3 Co/γ-Al2O3
Temp. (oC)
217
220
230
220
CO conv. (%) 36.4
Product sel. (%)
50.1
74.2
48.6
CO2
CH4
C2–C4
C5+
1.52
17.0
45.0
36.5
11.0
32.3
29.1
27.6
9.80
29.3
33.9
26.9
9.92
28.3
26.2
35.6
Liquid oil productivity
−1
(ml kg−1
h
)
cat
Gasoline
15.4
10.2
4.70
30.3
26.1
13.9
6.71
49.0
16.6
8.69
24.8
14.9
8.08
(C5–C12)
Diesel
(C13–C18)
Wax
(C19+
)
Total
46.7
74.3
47.8
Chain growth 0.839
0.834
0.831
0.855
probabil-
ity, α
1 3

Efect of cobalt metal loading on Fischer–Tropsch synthesis…
the number of cobalt active sites provided by these catalysts. As described above
in the catalysts characterization, the cobalt particle size in the Co/γ-Al2O3 catalysts
was unchanged at cobalt loading up to 15 wt%, as also supported by the XRD and
TEM analyses (Figs. 2, 3, 4). Thus, the increase in the number of active sites in the
catalysts may be directly related with the increase of cobalt loading. However, 20
wt% Co/γ-Al2O3 actually ofered fewer active sites than 15 wt% Co/γ-Al2O3 because
larger cobalt particles were formed in the preparation of 20 wt% Co/γ-Al2O3.
The reaction temperature distribution was almost identical to the dependency of
the CO conversion on the cobalt loading because FTS is a highly exothermic reac-
tion and the generated reaction heat is proportional to the conversion.
Not only CO conversion, but also C5+ liquid oil productivity was proportional
to the number of cobalt particles because those activities were closely related to
the number of active sites. However, 20 wt% Co/γ-Al2O3 showed the highest chain
growth probability, α, as it was strongly related to the size of the cobalt active sites
because larger cobalt particles aforded more chances for chain growth [14]. Fur-
thermore, as the high reaction temperature was unfavorable for chain growth, the
chain growth probability, α, and the C5+ liquid oil selectivity were inversely propor-
tional to the reaction temperature for the similarly sized cobalt particles in the 5, 10,
and 15 wt% Co/γ-Al2O3 catalysts [14, 26].
The XPS data for the Co/γ-Al2O3 catalysts after the FTS reaction are presented in
Fig. 6. It was reported that the position of Co 2p3/2 peaks is related to the chemical
status of the cobalt species [1]. Cobalt metallic species were identifed using the Co
2p3/2 peak at 777.9 eV and the Co3O4 spinel phase was characterized by the Co 2p3/2
peak at 780.2 eV and a low intensity of the shake-up satellite peak at ca. 787 eV [1, 12,
Co2p3/2
Co2p1/2
780.85 eV
780.15 eV
Satellite
(d)
(c)
(b)
(a)
770
780
790
800
810
Binding Energy [ eV ]
Fig. 6 Co 2p XPS spectra of the supported cobalt catalysts: a 5 wt% Co/γ-Al2O3, b 10 wt% Co/γ-Al2O3,
c 15 wt% Co/γ-Al2O3, and d 20 wt% Co/γ-Al2O3 after FTS reaction
1 3

J.-I. Yang, C. H. Ko
29]. For the 5, 10, and 15 wt% Co/γ-Al2O3 catalysts, the binding energy of the Co 2p3/2
level was 780.85 eV, which decreased to 780.15 eV at the 20 wt% loading. This shift
of the Co 2p3/2 peak toward a lower binding energy at the 20 wt% loading defnitely
implied that the cobalt species in 20 wt% Co/γ-Al2O3 was more favorable for reducibil-
ity and larger than the others.
Conclusions
The following conclusions were derived from the results of the present work:
1. As the cobalt loading on the catalyst was increased from 5 to 15 wt%, the surface
area and pore volume of the catalysts linearly decreased because of partial block-
age of the alumina support pores by cobalt oxide cluster and/or partial collapse
of the pore structure with increasing cobalt loading. However, at 20 wt% cobalt
loading, the surface area, pore volume, and average pore diameter of the cata-
lyst remained unchanged from the corresponding values of 15 wt% Co/γ-Al2O3.
Thus, the optimum cobalt loading was 15 wt% because the complete coverage of
alumina surface with cobalt oxide and the complete flling pores in alumina was
achieved at this loading.
2. XRD analysis revealed that the number of cobalt oxide particles increased as the
cobalt loading was increased up to 15 wt% due to the interaction between cobalt
oxide and the alumina support surface, whereas the particle size of the cobalt
oxide increased as the cobalt loading was further increased to 20 wt% because of
the additional cobalt loading on the previously loaded cobalt oxide. Furthermore,
TEM analysis revealed cobalt oxide particle sizes of 20–50 nm and 70–100 nm
at cobalt loadings of up to 15 wt% and 20 wt%, respectively.
3. As the cobalt loading in the Co/γ-Al2O3 catalysts was increased from 5 to 15
wt%, both CO conversion and C5+ liquid oil productivity increased due to their
high dependency on the number of cobalt particle active sites. However, the CO
conversion and C5+ liquid oil productivity inversely dropped to 48.6% and 47.8 ml
−1
kg h−1, respectively, at the high cobalt loading of 20 wt%, due to the increased
cat
cobalt particle size. Especially, the 20 wt% Co/γ-Al2O3 catalyst showed the high-
est chain growth probability, α, because the larger cobalt active particles aforded
more opportunities for chain growth. Furthermore, the XPS data also supported
the more favorable reducibility of the cobalt species in 20 wt% Co/γ-Al2O3, and
hence its larger size compared to the other loadings.
Acknowledgements This work was supported by the Research and Development Program of the Korea
Institute of Energy Research (KIER) (No. B9-2442-05).
1 3

Efect of cobalt metal loading on Fischer–Tropsch synthesis…
References
1. W. Chu, P.A. Chernavskii, L. Gengembre, G.A. Pankina, P. Fongarland, A.Y. Khodakov, J. Catal.
252, 215 (2007)
2. J.H. Yang, H.K. Kim, D.H. Chun, H.T. Lee, J.C. Hong, H. Jung, J.I. Yang, Fuel Process. Technol.
91, 285 (2010)
3. M.H. Rafq, H.A. Jakobsen, R. Schmid, J.E. Hustad, Fuel Process. Technol. 92, 893 (2011)
4. M.E. Dry, Catal. Today 71, 227 (2002)
5. I. Wender, Y. Zhang, L. Hou, J. Tierney, CFFS Annual Meeting, Roanoke, West Virginia, Aug. 1–4,
2004
6. H. Jung, J.I. Yang, J.H. Yang, H.T. Lee, D.H. Chun, H.J. Kim, Fuel Process. Technol. 91, 1839
(2010)
7. C.G. Visconti, E. Tronconi, L. Lietti, P. Forzatti, S. Rossini, R. Zennaro, Topics Catal. 54, 786
(2011)
8. R.L. Espinoza, Gas to Liquids (GTL). www.primethreelabs.com/assets/gtl.ppt
9. H. Schulz, Appl. Catal. A: Gen. 186, 3 (1999)
10. A. Forbes, Gas to Liquids 2007, London, Oct. 30–31, 2007
11. R. Guettel, U. Kunz, T. Turek, Chem. Eng. Technol. 31, 746 (2008)
12. A. Martinez, C. Lopez, F. Marquez, I. Diaz, J. Catal. 220, 486 (2003)
13. S. Storsœter, B. Totdal, J.C. Walmsley, B.S. Tanem, A. Holmen, J. Catal. 236, 139 (2005)
14. A. Barbier, A. Tuel, I. Arcon, A. Kodre, G.A. Martin, J. Catal. 200, 106 (2001)
15. M. Blanchard, H. Derule, P. Canesson, Catal. Lett. 2, 319 (1989)
16. E. Iglesia, S.L. Soled, J.E. Baumgartner, S.C. Reyes, J. Catal. 153, 108 (1995)
17. G. Jacobs, T.K. Das, Y. Zhang, J. Li, G. Racoillet, B.H. Davis, Appl. Catal. A: Gen. 233, 263 (2002)
18. G.L. Bezemer, P.B. Radstake, U. Falke, H. Oosterbeek, H.P.C.E. Kuipers, A.J. van Dillen, K.P. de
Jong, J. Catal. 237, 152 (2006)
19. P.J. van Berge, J. van de Loosdrecht, S. barradas, A.M. van der Kraan, Catal. Today 58, 321 (2000)
20. G. Jacobs, T.K. Das, P.M. Patterson, J. Li, L. Sanchez, B.H. Davis, Appl. Catal. A: Gen. 247, 335
(2003)
21. G. Jacobs, P.M. Patterson, T.K. Das, M. Luo, B.H. Davis, Appl. Catal. A: Gen. 270, 65 (2004)
22. J. van de Loosdrecht, B. Balzhinimaev, J.-A. Dalmon, J.W. Niemantsverdriet, S.V. Tsybulya, A.M.
Saib, P.J. van Berge, J.L. Visagie, Catal. Today 123, 293 (2007)
23. D.E. Sparks, G. Jacobs, M.K. Gnanamani, V.R.R. Pendyala, W. Ma, J. Kang, W.D. Shafer, R.A.
Keogh, U.M. Graham, P. Gao, B.H. Davis, Catal. Today 215, 67 (2013)
24. J. Farzanfar, A.R. Revani, Res. Chem. Intermed. 41, 8975 (2015)
25. Z. Hajjar, M.D. Rad, S. Soltanli, Res. Chem. Intermed. 43, 1341 (2017)
26. J.I. Yang, J.H. Yang, H.J. Kim, H. Jung, D.H. Chun, H.T. Lee, Fuel 89, 237 (2010)
27. E. Iglesia, Appl. Catal. A: Gen. 161, 59 (1997)
28. H.J. Kim, J.H. Ryu, H. Joo, J. Yoon, H. Jung, J.I. Yang, Res. Chem. Intermed. 34, 811 (2008)
29. F. Morales, O.L.J. Gijzeman, F.M.F. de Groot, B.M. Weckhuysen, Stud. Surf. Sci. Catal. 147, 271
(2004)
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published
maps and institutional afliations.
Afliations
Jung‑Il Yang1 · Chang Hyun Ko2
1
Clean Fuel Laboratory, Korea Institute of Energy Research, 152 Gajeong-ro, Yuseong-gu,
Daejeon 305-343, Korea
2
School of Applied Chemical Engineering, Chonnam National University, 77 Yongbong-ro,
Buk-gu, Gwangju 500-757, Korea
1 3