Effect of La2O3-dopping on the Al2O 3 supported cobalt catalyst for Fischer-Tropsch synthesis

Article abstract of DOI:10.1016/j.molcata.2010.06.025

Alumina was doped with lanthana, La₂O₃, by either impregnation or co-precipitation. The doped and undoped alumina was then used as supports for the preparation of cobalt catalysts by incipient wetness impregnation. The catalysts prep

Full text of DOI:10.1016/j.molcata.2010.06.025

Journal of Molecular Catalysis A: Chemical 330 (2010) 10–17
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Effect of La2O3-dopping on the Al2O3 supported cobalt catalyst for
Fischer-Tropsch synthesis
Zhe Caia, Jinlin Lia,∗, Kongyong Liewb, Juncheng Hua
a Key Laboratory of Catalysis and Materials Science of the State Ethnic Affairs Commission & Ministry of Education, South-Central University for Nationalities, Wuhan, China
b Faculty of Industrial Science and Technology, University Malaysia Pahang, 26500 Kuantan, Malaysia
a r t i c l e i n f o
a b s t r a c t
Article history:
Alumina was doped with lanthana, La2O3, by either impregnation or co-precipitation. The doped and
undoped alumina was then used as supports for the preparation of cobalt catalysts by incipient wetness
impregnation. The catalysts prepared were characterized by nitrogen adsorption–desorption, powder
X-ray diffraction, hydrogen temperature programmed reduction, hydrogen temperature programmed
desorption, O2-titration and diffuse reflectance infrared Fourier transform spectroscopy, and the catalytic
properties for Fischer-Tropsch synthesis (FTS) have been evaluated. The results showed that La2O3-
dopping impacted the catalytic performance significantly. The catalyst with the La2O3-doped support
prepared by co-precipitation showed better reducibility, activity and product selectivities. The higher
ratio of the octahedral to tetrahedral cobalt species of the co-precipitated catalyst relative to the catalyst
doped by impregnation might have caused the enhanced performance. The presence of the hydrocarbonyl
species, distinguished by its vibrational feature, is directly related to CO conversion and chain-growth.
© 2010 Elsevier B.V. All rights reserved.
Received 26 March 2010
Received in revised form 23 June 2010
Accepted 25 June 2010
Available online 3 July 2010
Keywords:
Fischer-Tropsch synthesis
La2O3-doped Al2O3
Cobalt
1. Introduction
be more easily reduced than the cobalt in the tetrahedral envi-
ronment and hence, display better catalytic properties. Rare earth
Fischer-Tropsch synthesis (FTS) attracts much attention as an
option for producing clean transportation fuels and chemicals by
conversion of natural gas or coal [1,2]. Cobalt-based catalysts are
some of the most important candidates for FTS to produce high
molecular weight paraffinic waxes that can be hydrocracked to
produce lubricants and diesel fuels [3]. To increase the number of
cobalt active sites, the cobalt is dispersed as clusters on high sur-
face area supports such as SiO2, TiO2 or Al2O3. Alumina is often
used due to its favorable mechanical properties [4]. Previous stud-
ies indicated that cobalt supported on alumina was not completely
reduced to the metallic state because of significant metal-support
interactions [3,5]. Chen and Zhang [6] found that when the metal
loading was lower than the dispersion capacity in the ␥-Al2O3 sup-
ported transition metal (Ni, Co or Cu) oxide catalysts, the supported
M2+ ions were preferentially incorporated into the tetrahedral
vacancies of ␥-Al2O3, and that the amount of M2+ ions incorporated
into the octahedral vacancies increased with increased M2+ loading.
Since octahedral M2+ ions are easier to be reduced to the metallic
state than the tetrahedral M2+ ions, researchers [7] tried to increase
the octahedral ratio of the M2+ (Ni2+ in the study) to increase the
degree of reduction. Similarly, many researchers [8–10] reported
that the octahedrally coordinated cobalt FTS catalyst could also
metal oxides are often used as promoters in cobalt [11–13] and iron
[14] based CO hydrogenation catalysts (e.g., FTS [15]). Vada et al.
[12] studied a La3+/Co/Al2O3 system and found that La3+ increased
the overall activity and chain-growth probability when a low La3+
loading was used. Ren et al. [16] studied NiO/Al2O3 catalysts with
La3+ promoter, andfoundthatthe dispersedLa3+ species on␥-Al2O3
increased the ratio of octahedral Ni2+ ions to the tetrahedral ones,
and thus increased the degree of reduction of the catalysts. Ledford
et al. [13] studied two methods of preparing lanthanum promoted
Co/Al2O3 catalysts. They observed promotional effects only when
La3+ was impregnated first followed by cobalt, and for higher La3+
loadings, a La–Co mixed oxide was suggested to be formed and, in
turn, enhanced the dispersion of cobalt. In the present study, a sup-
port was prepared by co-precipitation consisting of ␥-Al2O3 doped
with La2O3 and then it was subsequently loaded with cobalt. It was
expected that when the lanthana was used as a dopant, a greater
fraction of octahedral sites will be formed on the surface of ␥-
Al2O3 crystallites and provide sites for Co ions during impregnation.
Also, it was anticipated that the co-precipitation procedure would
favor octahedral site formation relative to the case of impregnating
the lanthana, hence displacing Co to the preferred octahedral sites
providing a more facile reduction. In this contribution, a series of
lanthana doped alumina supports was prepared by impregnation
or co-precipitation. Cobalt was added to both the La2O3-doped and
undoped alumina by incipient wetness impregnation. The follow-
ing techniques were used to characterize the catalysts, including
∗
Corresponding author. Tel.: +86 027 67843016; fax: +86 027 67842752.
E-mail addresses: lij@mail.scuec.edu.cn, jinlinli@hotmail.com (J. Li).
1381-1169/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2010.06.025

Z. Cai et al. / Journal of Molecular Catalysis A: Chemical 330 (2010) 10–17
11
nitrogen adsorption–desorption, powder X-ray diffraction (XRD),
hydrogen temperature programmed reduction (H2-TPR), hydrogen
temperature programmed desorption (H2-TPD) and O2-titration
and diffuse reflectance infrared Fourier transform spectroscopy
(DRIFTS). FTS reaction tests were carried out using a fixed bed
reactor.
at a relative pressure of 0.99. Pore size distributions were eval-
uated from the desorption branches of the isotherms using the
Barrett–Joyner–Halenda (BJH) model.
2.2.3. Hydrogen temperature programmed reduction
The reduction profiles of the catalysts were measured by hydro-
gen temperature programmed reduction (H2-TPR) experiments
employing a Zeton Altamira AMI-200 unit. Calcined catalysts
(∼0.10 g) were placed in a U-shape quartz reactor, with a thermo-
couple for continuous temperature measurement. Samples were
first pretreated by flowing high purity argon at 423 K, and cooled
to 323 K, to remove water or other residual contaminants. Then,
10% H2/Ar (constant flow rate of 30 ml/min) was switched on and
the temperature was raised from 323 to 1073 K at a rate of 10 K/min,
and held for 30 min at 1073 K. Hydrogen consumption was moni-
tored using a thermal conductivity detector (TCD).
2. Experimental
2.1. Catalyst preparation
2.1.1. Preparation of supports
The La2O3-doped alumina support was prepared by co-
precipitation as follows: known amounts of aluminum nitrate and
lanthamum nitrate solutions were mixed in appropriate propor-
tions. Ammonium hydroxide was added to the mixed solution
to precipitate the hydroxide. The pH value of the solution was
kept constant at 9.0 0.1. The obtained support is denoted as LAP
(2.9 wt% La2O3). Alumina support without La2O3 prepared by co-
precipitation was labeled as AP.
The support with the same loading of La2O3 was also prepared
by incipient wetness impregnation with the desired amount of
aqueous lanthanum nitrate on AP. The excess water was evaporated
in a rotary evaporator at around 363 K. This support was named
LAI. A commercial La2O3-doped support (SASOL Germany GmbH),
labeled as LAS, was used as a reference. All the materials were dried
in an oven overnight at 393 K, crushed, and sieved to obtain 120
mesh particles, followed by calcination at 1023 K for 6 h.
2.2.4. Hydrogen temperature programmed desorption
The dispersion and crystallite size of cobalt were measured by
hydrogen temperature programmed desorption (H2-TPD) using a
U-shaped quartz reactor and the Zeton Altamira AMI-200 unit. The
catalysts (∼0.15 g) were reduced at 723 K for 12 h and cooled to
373 K in flowing hydrogen. Prior to increasing the temperature
from 373 to 723 K at a rate of 10 K/min, the samples were held
at 373 K for 1 h under an argon stream to drive away weakly bound
physisorbed species. Then the catalysts were held at 723 K under
flowing argon to desorb the remaining chemisorbed hydrogen;
meanwhile, the TCD detector began to record the signal until it
returned to the baseline. The amount of desorbed hydrogen was cal-
culated by comparing the integrated TPD spectrum with the mean
areas of calibrated hydrogen pulses. The formula for the calculation
has been shown in previous works [18].
2.1.2. Preparation of cobalt catalysts
The 12 wt% cobalt catalysts were prepared by impregnating the
different synthesized supports with aqueous solutions of cobalt
nitrate. The samples were dried at 393 K and calcined in air at 623 K
for 6 h, hereafter denoted as CAP, CLAI, CLAS and CLAP.
2.2.5. O2-titration
The degree of reduction was measured by O2-titration done in
the Zeton Altamira AMI-200 unit. The samples which had been
reduced in pure hydrogen at 723 K for 12 h were reoxidized at 723 K
by introducing pulses of high purity oxygen, until there was no
further consumption of O2, detected by the thermal conductivity
detector located downstream. The reference gas for TCD was high
purity helium. All flow rates were set to 30 ml/min. The extent of
catalyst reduction was calculated assuming stoichiometric reoxi-
dation of Co to Co3O4. The formula for the calculation is described
elsewhere [18].
2.2. Catalyst characterization
2.2.1. Powder X-ray diffraction
The X-ray diffraction (XRD) measurements for the samples were
carried out using a Brukers D8 powder X-ray diffractometer with
monochromatic CuK␣ radiation and a VANTEC-1 detector over a 2ꢀ
range of 20–80◦ with a step size of 0.0167◦, and from 35◦ to 40◦ with
a step size of 0.008◦ steps for the crystallite size of cobalt oxide. The
in situ XRD measurement was carried out as follows. The sample,
placed in the XRK 900 reactor chamber (Anton paar) mounted on a
goniometer, was first purged with pure nitrogen at 423 K for 1 h to
flush away water and impurities. Pure hydrogen gas was introduced
to the cell after cooling to room temperature. The temperature was
then increased to 723 K at a rate of 1 K/min, followed by several
scans taken in an interval of 1 h between scans. The sample temper-
ature and reduction gas admitted into the in situ cell were reliably
controlled and measured. The crystallite phase was detected by
comparing the patterns with those in the standard powder XRD
file compiled by the Joint Committee on Powder Diffraction Stan-
dards (JCPDS). The average Co3O4 crystallite size was calculated by
line broadening analysis using the Scherrer equation [17].
2.2.6. Diffuse reflectance infrared Fourier transform spectroscopy
(DRIFTS)
Adsorption properties of catalysts were measured by in situ
DRIFTS with a Nicolet NEXUS 6700 FTIR instrument. The spectral
resolution was 4 cm−1 and 32 scans were obtained for each spec-
trum. The DRIFTS unit contains a mercury cadmium telluride (MCT)
detector, and a diffuse reflectance attachment was used. High
purity carbon monoxide (>99.9%) and syngas (H2/CO = 2) served as
probe gases. Nitrogen (>99.999%) was used as the purge gas and
hydrogen (>99.999%) was used as the reducing gas. The gases were
cleaned prior to use by passing through gas purifiers. The catalysts
(30–40 mg) were placed in an infrared cell with ZnSe windows and
reduced in situ at 723 K for 10 h under atmospheric pressure in a
hydrogen stream at a flow rate of 20 ml/min. After the reduction
procedure, the system was cooled to the desired temperature and
the background spectra were recorded before the introduction of
probe gases. Each spectrum was referenced to a spectrum of the
catalyst collected at the desired temperature under H2 flow before
CO or syngas adsorption. Circulating water was used to cool the
body of the reaction chamber. A thermocouple located directly in
the sample was used for temperature control.
2.2.2. BET surface area and pore size distribution
The BET surface area and pore size distribution were measured
by nitrogen adsorption–desorption at the boiling temperature
of nitrogen (77 K) using
a Quantachrome Autosorb-1-C-MS.
Before the measurement, the samples were outgassed at 473 K
for >6 h. The surface area was obtained by applying the
Brunauer–Emmett–Teller (BET) model for adsorption in a rela-
tive pressure range of 0.05–0.30. The total pore volume of each
sample was calculated from the amount of N2 vapor adsorbed

12
Z. Cai et al. / Journal of Molecular Catalysis A: Chemical 330 (2010) 10–17
Table 1
Surface area, average pore diameter and pore volume.
Sample
Surface
Pore volume
(cm3/g)
Average pore
diameter (nm)
Crystallite
area (m2/g)
diameter (nm)a
AP
LAI
168.8
160.0
143.9
199.7
143.6
127.7
121.9
173.2
0.438
0.417
0.548
0.497
0.335
0.306
0.430
0.377
7.81
9.58
12.37
7.81
7.86
9.55
–
–
–
–
LAS
LAP
CAP
CLAI
CLAS
CLAP
14.02
14.14
14.17
14.49
12.38
7.81
a
Average Co3O4 particle diameter obtained from XRD line broadening analysis.
2.3. Catalytic evaluation
FTS was performed in a fixed bed reactor (id = 12 mm) at 503 K
and 1.0 MPa. The catalysts (ca. 0.5 g) were mixed with 5 g of car-
borundum to minimize the temperature gradient and reduced in
Fig. 1. X-ray diffraction patterns for calcined catalysts (᭹: Co3O4; ꢀ: ␥-Al2O3).
situ with H2 at atmosphere pressure. With H2 flowing (6 NL h−1 g−1
,
298 K, 0.1 MPa), the reactor temperature was raised from room
temperature to 373 K at 2 K/min, increased to 723 K at a rate of
1 K/min, and held for 10 h. The reactor was then cooled to 373 K.
At that temperature, syngas, with space velocity of 4 SL g−1 h−1
(273 K, 0.1 MPa), was switched on and the pressure was increased to
1.0 MPa. The reactor temperature was increased to 503 K in 16 h and
the reaction was performed at 503 K. The reaction products were
collected after more than 80 h of operation to achieve a good mass
balance at close to steady state, typically. The carbon monoxide
conversion and hydrocarbon selectivities were measured at steady
state conditions over the period of operation. The discharged gases
were analyzed online by an Agilent MicroGC 3000A gas chromato-
graph (GC). The analysis of the liquid product, which was collected
in the cold trap (269 K), was performed using an Agilent 6890N
GC equipped with a flame ionization detector (FID). The solid wax,
which was collected in the hot trap (373 K), was analyzed using an
Agilent 7890A GC.
persed on the support [19]. The size data of the Co3O4 crystallites
are shown in Table 1. Crystallite size of Co3O4 increased slightly on
dopping with La2O3, suggesting that La2O3 may have influenced
the Co3O4 crystal growth.
The in situ XRD patterns of catalysts reduced at 723 K under
H2 are shown in Figs. 2–5. For all the catalysts, the reduction with
H2 at 723 K for 5 h resulted in the reduction of most of the Co3O4
to metal Co. When the reduction temperature reached 573 K, the
Co3O4 phase of all the catalysts transformed to CoO, indicating that
Co3O4 reduction is fast regardless of the support composition and
structure. However, the diffraction peaks of Co metal for catalysts
doped with La2O3 can be found immediately as soon as the temper-
ature reached 723 K, but for CAP (without La2O3-dopping), these
peaks appeared only after having been reduced at 723 K for 1 h.
The relative peak intensity of the reduced cobalt in the catalyst
doped with La2O3 was higher than the undoped catalyst, which
suggests that dopping with La2O3 enhanced the reduction of CoO
phase to Co. For all catalysts except CLAP, peaks of CoO are still
found even after reduction in H2 at 723 K for 5 h. It is concluded
that CLAI showed similar reduction properties to that of CLAS, and
CLAP exhibited the highest extent of reduction.
3. Results and discussion
3.1. BET surface area and porosity
The surface area, average pore diameter and pore volume data of
the supports and the cobalt catalysts are listed in Table 1. Lanthana
dopping has a marginal influence on the pore volume. However, the
porosity of the supports was significantly affected by the prepara-
tion method. The co-precipitation method (LAP support) resulted
in the highest BET surface area, while the commercial support (LAS)
gave the lowest surface area but the widest average pore size. As can
be seen in Table 1, La2O3-dopping by the method of impregnation
increased the average pore diameter from 7.81 nm (for undoped
alumina support AP) to 9.58 nm (LAI), indicating that the impreg-
nation of lanthana caused some blocking of the narrower pores
relative to undoped AP. The impregnation with cobalt lowered both
the support surface area and the pore volume, and did not change
the average pore size, which suggests that some cobalt species have
entered the pore of the supports.
3.2. X-ray diffraction results
XRD patterns of the catalysts are presented in Fig. 1. The diffrac-
tion peaks at 45.7◦ and 66.6◦ are attributed to the ␥-Al2O3 support
and the peaks at 31.4◦, 36.9◦, 45.0◦, 59.5◦ and 65.5◦ are assigned
to the Co3O4 phase, which exists on all the catalysts [15]. The
diffraction peak of La2O3 phase could not be detected from the
XRD patterns of catalysts, indicating that the La2O3 was highly dis-
Fig. 2. In situ X-ray diffraction patterns of CAP catalyst reduced in pure H2 (᭹: Co3O4;
ꢀ: ␥-Al2O3; +: CoO; #:Co0).

Z. Cai et al. / Journal of Molecular Catalysis A: Chemical 330 (2010) 10–17
13
Fig. 6. TPR spectra of calcined catalysts.
Fig. 3. In situ X-ray diffraction patterns of CLAI catalyst reduced in pure H2 (᭹:
Co3O4; ꢀ: ␥-Al2O3; +: CoO; #:Co0).
3.3. H2-temperature programmed reduction
The TPR spectra of catalysts are shown in Fig. 6. Three reduc-
tion peaks for Co/Al2O3 are observed at temperature ranges of
<573, 573–673 and >673 K, respectively. The peak at 573–673 K is
attributed to the reduction of Co3O4 to CoO [17]. The broad peak at
high temperature (>673 K) is ascribed to the subsequent reduction
of CoO to Co0 [20]. The appearance of the peaks at the different
reduction temperature suggests that all catalysts exhibited the dif-
ferent interactions between the metal and support [21,22]. Besides
two main reduction peaks, a small peak is observed below 573 K.
This peak could be attributed to the reductive decomposition of
residual Co(NO3)2 [23]. The reduction process was consistent with
the results of in situ XRD.
Table 2 presents the reduction degree calculated from the TPR
profile. All the Co3O4 can be reduced to Co0 in the whole temper-
ature program process for CLAP, In the temperature range above
673 K, assuming the H2 consumption to completely reduce CoO to
Co to be 2.04 mmol/gcatal, CoO was reduced to Co metal completely
for catalyst CLAP, while for the other catalysts, 70–86% of the CoO
could be reduced by the amount of H2 above this temperature. This
is due to the lower reducibility of these cobalt species with stronger
interaction with the supports. It is noticed that the reduction degree
of catalyst CLAP was above 100% which may be caused by the H2
consumption of the nitrate ions.
Fig. 4. In situ X-ray diffraction patterns of CLAS catalyst reduced in pure H2 (᭹:
Co3O4; ꢀ: ␥-Al2O3; +: CoO; #:Co0).
All reduction peaks of CAP appeared at higher reduction tem-
perature than the catalysts doped with La2O3, indicating that
La2O3-dopping increased catalyst reducibility. The methods of
La2O3-dopping markedly affected the reduction properties of
Co/Al2O3 catalysts. It should be noted that the second reduction
step (>673 K) of several catalysts comprised of broad peaks, which
may be due to different degrees of interaction between cobalt and
the support [24]. Also, the existence of cobalt aluminate cannot be
excluded, as it typically reduces at or above 1273 K. The reduction
Table 2
Reduction degree and H2 consumption for the TPR profile calculation.
Catalysts H2 consumption H2 consumption H2 consumption Reduction
for the peak
<673 K
for the peak
>673 K
for 373–1073 K
(mmol/gcat)
degree (%)a
(mmol/gcat)
(mmol/gcat)
CAP
0.62
0.70
0.72
0.82
1.43
1.68
1.75
2.03
2.05
2.38
2.47
2.85
76
88
91
CLAI
CLAS
CLAP
105
Fig. 5. In situ X-ray diffraction patterns of CLAP catalyst reduced in pure H2 (᭹:
Co3O4; ꢀ: ␥-Al2O3; +: CoO; #:Co0).
a
H2 consumption should be 2.72 mmol/gcat to reduce the Co3O4 to the metal
cobalt.

14
Z. Cai et al. / Journal of Molecular Catalysis A: Chemical 330 (2010) 10–17
Table 3
H2-temperature programmed desorption and O2 pulse reoxidation results for catalysts.
Sample
H2 desorbed (␮mol/g)
duncorrected (%)a
Duncorrected (nm)b
O2 uptake (␮mol/g)
dcorrected (%)c
Reducibility (%)
Dcorrected (nm)d
CAP
183.2
189.0
193.5
197.5
17.99
18.56
19.00
19.39
5.7
5.6
5.4
5.3
579.3
734.3
736.8
813.6
42.37
34.49
35.19
32.51
42.46
53.82
54.00
59.64
2.4
3.0
2.9
3.2
CLAI
CLAS
CLAP
a
The uncorrected catalyst dispersion.
The uncorrected cobalt metal cluster diameter.
The corrected catalyst dispersion.
b
c
d
The corrected cobalt metal cluster diameter.
peak above 673 K shifted to a lower temperature for catalyst CLAP
compared with the catalysts CLAI and CLAS, and the intensity of
this broad peak increased. All the above results indicated that CLAP
exhibits the best reduction properties among the catalysts studied.
more active for FTS than those with cobalt cations in tetrahedral
sites [9,10].
3.5. Fischer-Tropsch synthesis catalytic tests
Results of the FTS tests are displayed in Table 4. CO conversion
and selectivities to heavy hydrocarbons increased with dopping of
La2O3.The activity of the La2O3-doped catalysts increased in the
order of CLAI < CLAS < CLAP which is consistent with the order of
reducibility. Selectivity for methane showed an opposite trend to
the activity. CLAP catalyst displayed the highest FT activity and the
lowest methane selectivity among the catalysts studied.
3.4. H2-temperature programmed desorption and O2-oxidation
Table 3 presents the H2-temperature programmed desorption
and O2-oxidation data. The La2O3-dopping increased the degree of
reduction of the catalysts significantly, and the method of La2O3-
dopping affected the metal cobalt cluster size, dispersion and
reducibility. Catalyst reducibility increased in the order of CLAI,
CLAS to CLAP. Furthermore, the cobalt cluster size was increased
by the dopping of La2O3 and was slightly affected by the dopping
method. Khassin et al. [9] reported that the Co2+ cations which
entered the support structure and occupied the octahedral posi-
tions could be reduced at lower temperature. For Mg2+ and Zn2+
promoted catalysts, promoter cations occupied the tetrahedral
sites of the supports and caused the Co2+ cations to occupy mainly
octahedral sites, thus increasing the extent of reduction of the
catalyst. Catalyst performance parameters of the promoted cata-
lysts were significantly better than the unpromoted catalysts. In
the present work, it could be considered that in CLAP, the La2O3-
dopping process carried out at the same time with the formation
of ␥-Al2O3 crystallite phase and the dispersed lanthanum species
on ␥-Al2O3 might inhibit the incorporation of cobalt species from
entering the tetrahedral sites of ␥-Al2O3. When the La2O3-dopping
process was carried out after the formation of the support crystal-
lite phase in the CLAI catalyst, the impregnated La2O3 could not
inhibit the tetrahedral sites to the same degree as in the CLAP
catalyst, and thus the CLAI catalyst exhibited a lower degree of
reduction relative to the CLAP catalyst [6,9,10]. Similar finding was
reported by Ren et al. [16] in their studies of La2O3 promoted
NiO/Al2O3 in which they found from the UV–VIS DRS, XRD and
TPR results that the dispersed La3+ species inhibited the incor-
poration of Ni2+ ions into tetrahedral vacancies on the surface of
support and increased the ratio of Ni2+ ions incorporated into the
octahedral vacancies of ␥-Al2O3. Khassin et al. [8] studied Co–Al
co-precipitated catalysts for FTS, they observed that Co2+ occupied
mostly the octahedral sites in Mg2+ and Zn2+ promoted samples,
and had higher reducibility and better catalytic performance. The
catalysts with cobalt cations in octahedral sites were found to be
Previous studies on cobalt-based catalysts for FTS found that
high reducibility provided more cobalt active sites and hence,
increased the CO conversion rate by increasing the active site den-
sity [25–30]. In the present work, lanthanum oxide facilitated the
reduction of cobalt oxide dispersed on the supports, and the impact
was more significant in the case of the CLAP catalyst leading to
an increase in CO conversion. Methane selectivity was 13.19% for
the case of the unpromoted catalyst, CAP; however, with La2O3-
dopping, methane selectivity decreased significantly to 9.14% for
CLAS and 8.05% for CLAP. Vada et al. [12] reported that in the stud-
ies of the La3+ promotion in the Co/Al2O3 catalysts, the promoted
La3+ decreased the methane selectivity and increase the long chain-
growth ability. High methane selectivity has been reported in the
case of catalysts possessing low reducibility and high dispersion
[31], and is typically attributed to the presence of unreduced cobalt
oxides which catalyzed the WGS reaction, resulting in a higher
H2/CO ratio on the surface Co0 sites, which in turn tends toward
the hydrogenation of adsorbed species to lower carbon number
product. It should be mentioned that CLAP has the highest reduc-
tion degree, and CO2 selectivity was 0.49%, much lower than those
of CLAI (1.50%) and CLAS (1.64%); this lower WGS activity is thus
consistent with lower CH4 selectivity.
3.6. Diffuse reflectance infrared Fourier transform spectroscopy
3.6.1. DRIFTS spectra of CO adsorbed on reduced catalysts at
room temperature
The spectra of CO adsorbed on the surface of undoped and
La2O3-doped catalysts are shown in Fig. 7. The DRIFTS spectra
were recorded after reduction of the catalyst in hydrogen, introduc-
Table 4
FTS reaction CO conversion and product distribution.
Catalysts
CO conversion (%)
CO2 selectivity (%)
Hydrocarbon selectivity (%)
CH4
C2
C3
C4
C5+
CAP
28.04
32.76
33.06
34.92
1.88
1.50
1.64
0.49
13.19
8.57
9.14
8.05
1.44
0.83
1.15
0.91
2.96
2.31
2.32
2.39
5.26
4.01
3.99
3.81
75.28
82.79
81.75
84.36
CLAI
CLAS
CLAP
The catalysts were reduced in H2 at 723 K for 10 h before FTS.
Reaction conditions: H2/CO = 2, space velocity of syngas: 4 SL g−1 h−1 (273 K, 0.1 MPa), temperature: 503 K; pressure: 1.0 MPa.

Z. Cai et al. / Journal of Molecular Catalysis A: Chemical 330 (2010) 10–17
15
Fig. 9. DRIFTS spectra of syngas reaction on reduced CLAI.
Fig. 7. DRIFTS spectra of CO adsorbed on reduced catalysts at 303 K.
to be strongest among all the catalysts. The methane selectivity was
the highest in this La2O3-doped catalyst, with intermediate activ-
ity. We propose that this band was due to multi-carbonyl species
on Co␦+ with weak electron-donor properties [37].
tion of CO at room temperature (∼1 h), and purging with flowing
N2.
Several bands located at 2054, 2033, 1976, 1935, 1891, and
1801 cm−1 in addition to two broad bands appearing at 2171 and
2120 cm−1 could be clearly resolved for all the catalysts. The bands
at 2171 and 2120 cm−1 have been assigned to residual weakly
adsorbed carbon monoxide [32]. The band positioned at 1801 cm−1
could be assigned to CO adsorbed on metal cobalt sites [33]. The
band at 1891 cm−1 arises either from CO adsorbed in a bridged posi-
tion on Co (0001) or from monocarbonyls adsorbed on defect sites,
edges, or corners [34]. The bands located at 1935 and 1976 cm−1
are attributed to CO adsorbed on Co0 in the bridge conformation. It
should be noted that the relative intensities of bands positioned at
approximately 1891, 1935 and 2033 cm−1 correspond to the same
trend for CO conversion and the opposing trend towards methane
selectivity (see Section 3.5), leading to the speculation that CO
adsorbed on the active cobalt phase sites resulted in the high prob-
ability for chain growth and low CH4 selectivity. Previous work [35]
reported that similar results were obtained for the in situ DRIFTS
on Co/SiO2 catalysts.
3.6.2. DRIFTS spectra of syngas reaction on catalyst surfaces
The DRIFTS spectra corresponding to in situ syngas reaction over
reduced catalysts are shown in Figs. 8–11, recorded after catalyst
reduction and introduction of syngas (H2/CO = 2). The temperature
was raised from 303 to 673 K at a rate of 5 K min−1, and syngas flow
rate was 6 ml/min.
The spectra could be divided into four regions. There are four
bands in the region of 2800–3100 cm−1, corresponding to the ␯(CH)
modes of CH4 (3015 cm−1), CH3 (2968 cm−1), and CH2 (2927 and
2864 cm−1) [38–40]. The band at ∼2350 cm−1 is typical of CO2,
while that at ∼2143 cm−1 is due to residual gas phase CO. Those
in the region 1900–2100 cm−1 are linear- and bridge-bonded CO
adsorbed on cobalt sites. From the spectra of syngas adsorbed on the
catalysts and temperature programmed desorption experiments,
the following features are noted:
For the band located at around 2060 cm−1, Rygh et al. [36]
assigned it to overlapping contributions from Co(CO)n, CoHCO,
Co␦+-CO (0 <ı <2), and CO linearly adsorbed on Co (0001). This band
was observed in our previous studies, where we assigned it to CO
adsorbed on metallic cobalt sites [35]. In the present work, the
intensity of the band at 2054 cm−1 for the CLAS catalyst was found
(1) Hydrocarbon species, located in the region 2800–3100 cm−1
,
appeared on catalyst CAP at 473 K; which is 20 K higher than
for the other catalysts. The intensity of these bands on CLAP and
CLAS catalysts increases rapidly with increasing temperature.
In contrast, the hydrocarbon band intensities for catalysts CAP
Fig. 8. DRIFTS spectra of syngas reaction on reduced CAP.
Fig. 10. DRIFTS spectra of syngas reaction on reduced CLAS.

16
Z. Cai et al. / Journal of Molecular Catalysis A: Chemical 330 (2010) 10–17
also observed a band at around 2015–2029 cm−1 on Co/SiO2
catalysts in previous studies, and ascribed it to hydrocarbonyl
species [35]. From the current results, the new band at around
2060 cm−1 appeared at 413 K, followed by the appearance of
hydrocarbon bands (at the region of 2800–3100 cm−1) at 453 K.
The change in intensity of hydrocarbon species is consistent
with that of the new band (Fig. 12). It should be pointed out that
the new band located at 2020–2060 cm−1 is not only directly
related to the bands of hydrocarbons (2800–3100 cm−1), but
also related to CO conversion and chain-growth. The new band
at around 2020–2060 cm−1 which is ascribed to the “hydrocar-
bonyl” species, may be a vibrational feature of some reaction
intermediate. With increasing temperature, the intensity of the
new band became stronger and shifted to lower wavenumbers,
indicating that more hydrocarbonyl species were formed and
that the C–O bond became gradually weaker at higher temper-
atures.
Fig. 11. DRIFTS spectra of syngas reaction on reduced CLAP.
4. Conclusions
The effect of preparation method of La2O3-doped Co/Al2O3 FTS
catalysts has been investigated by various characterization tech-
niques and catalytic testing using a fixed bed reactor. The following
conclusions can be drawn.
and CLAI were still very weak even at temperature of 553 K. This
suggests that CAP and CLAI catalysts have lower initial activity
and lower chain-growth probability relative to CLAP and CLAS.
(2) For either CLAP or CLAS, the band intensity of CO2 (at 2364 and
2338 cm−1) was relatively unchanged in the temperature range
of 453–513 K and became stronger at temperatures >533 K,
while the bands assigned to hydrocarbon product increased
with increasing temperature. These indicate that the products
were mainly hydrocarbons and little CO2 was produced during
this heating process. This observation suggests that the tem-
perature range of 453–513 K is appropriate for FT synthesis.
(3) All bands corresponding to linear- and bridge-bonded CO
became weaker with increasing temperature, and a new band
appeared at around 2060 cm−1 at 413 K. As the temperature
was increased further, the band shifted to lower wavenum-
bers to below 2030 cm−1 at 553 K. The assignment of the
band in the range of 2015–2060 cm−1 is still ambiguous since
its frequency is close to 2000 cm−1, the dividing line that
distinguishes between linear- and bridge-bonded CO absorp-
tion. Some researchers [41–43] attributed this new band at
2050–2060 cm−1 to a structure of the type HCoCO designated as
“hydrocarbonyl”. The formation of hydrocarbonyl was assumed
to be a prerequisite step in the overall reaction between coad-
sorbed H2 and CO. Chen et al. [44] ascribed the new band
at 2046 cm−1 to “carbonhydride”, which was supposed to be
formed by coadsorption of CO and H2 on Co0 active sites. We
(1) The dopping of La2O3 strongly influenced the catalytic perfor-
mance. The Co catalyst with the La2O3-doped support prepared
by co-precipitation was more easily reducible, resulting in
higher FTS activity and lower methane selectivity.
(2) DRIFTS studies show that there is more CO adsorbed on
cobalt species for CLAP than the other catalysts; thus, the
catalyst displayed higher reducibility, activity, and lower
methane selectivity. The intensity of bands corresponding to
CO adsorbed on the catalysts with different supports follows
the order of CLAP > CLAS > CLAI > CAP, which is consistent with
the order of CO conversion for FT synthesis and the product
selectivities.
(3) The intensity of the band attributed to hydrocarbonyl species
showed a direct relationship with both CO conversion and
chain-growth for the CLAP catalyst, while CAP and CLAI dis-
played lower initial activity and chain-growth probability
relative to CLAP and CLAS.
Acknowledgement
Financial support from National Natural Science Foundation
of China (Grant Nos: 20590360 and 20773166) are gratefully
acknowledged.
References
[1] H. Schulz, Appl. Catal. 186 (1999) 3–12.
[2] R.B. Anderson, Fischer-Tropsch Synthesis, Academic Press, NY, 1984.
[3] G. Jacobs, T.K. Das, Y.Q. Zhang, J.L. Li, G. Racoillet, B.H. Davis, Appl. Catal. A 233
(2002) 263–281.
[4] E. Iglesia, S.L. Soled, R.A. Fiato, G.H. Via, J. Catal. 143 (1993) 345–368.
[5] W.J. Wang, Y.W. Chen, Appl. Catal. 77 (1991) 223–233.
[6] Y. Chen, L.F. Zhang, Catal. Lett. 12 (1992) 51–62.
[7] S.B. Ren, J.H. Qiu, C.Y. Wang, B.L. Xu, Y.N. Fan, Y. Chen, Chin. J. Catal. 28 (2007)
651–656.
[8] A.A. Khassin, T.M. Yurieva, G.N. Kustiva, L.M. Plyasova, I.S. Itenberg, M.P.
Demeshkina, G.K. Chermashentseva, V.F. Anufrienko, V.I. Zaikovskii, T.V. Larina,
I.Y. Molina, V.N. Parmon, J. Mol. Catal. A 168 (2001) 209–224.
[9] A.A. Khassin, T.M. Yurieva, G.N. Kustiva, I.S. Itenberg, M.P. Demeshkina, T.A.
Krieger, L.M. Plyasova, G.K. Chermashentseva, V.N. Parmon, J. Mol. Catal. A 168
(2001) 193–207.
[10] A.A. Khassin, T.M. Yurieva, V.V. Kaichev, V.I. Bukhtiyarov, A.A. Budneva, E.A.
Paukshtis, V.N. Parmon, J. Mol. Catal. A 175 (2001) 189–204.
[11] J. Barrault, A. Guilleminot, Appl. Catal. 21 (1989) 307–312.
[12] S. Vada, B. Chen, J.G. Goodwin, J. Catal. 153 (1995) 224–231.
Fig. 12. DRIFTS spectra of syngas adsorbed on reduced catalysts at 503 K.

Z. Cai et al. / Journal of Molecular Catalysis A: Chemical 330 (2010) 10–17
17
[13] J.S. Ledford, M. Houalls, A. Proctor, D.M. Hercules, J. Phys. Chem. 93 (1989)
[28] J.L. Li, N.J. Coville, Appl. Catal. A 181 (1999) 201–208.
6770–6777.
[29] Y. Wang, M. Noguchi, Y. Takahashi, Y. Ohtsuka, Catal. Today 68 (2001) 3–9.
[14] A.N. Pour, S.M.K. Shahri, H.R. Bozorgzadeh, Y. Zamani, A. Tavasoli, M.A. Marvast,
Appl. Catal. A 348 (2008) 201–208.
[30] A.Y. Khodakov, R. Bechara, A. Griboval-Constant, Stud. Surf. Sci. Catal. 142
(2002) 1133–1140.
[15] G.J. Haddad, B. Chen, J.G. Goodwin, J. Catal. 161 (1996) 274–281.
[16] S.B. Ren, J.H. Qiu, C.Y. Wang, B.L. Xu, Y.N. Fan, Y. Chen, Chin. J. Inorg. Chem. 23
(2007) 1021–1028.
[17] B.D. Cullity, Elements of X-ray Diffraction, third ed., Additon-Wesley, Reading.
MA, 1967.
[31] R.C. Reuel, C.H. Bartholomew, J. Catal. 85 (1984) 63–77.
[32] S.L. Sun, K. Fujimoto, N. Tsubaki, M. Yamada, Catal. Commun. 4 (2003) 361–364.
[33] P. Concepcion, C. Lopez, A. Martinez, V.F. Puntes, J. Catal. 228 (2004) 321–332.
[34] G.Z. Bian, A. Oonuki, Y. Kobayashi, N. Koizumi, M. Yamada, Appl. Catal. A 219
(2001) 13–24.
[18] Y.H. Zhang, H.F. Xiong, K.Y. Liew, J.L. Li, J. Mol. Catal. 237 (2005)
172–181.
[19] B. Jongsomjit, J. Panpranot, J.G. Goodwin, J. Catal. 215 (2003) 66–77.
[20] A. Bao, K.Y. Liew, J.L. Li, J. Mol. Catal. A 304 (2009) 47–51.
[21] G. Jacobs, Y. Ji, B.H. Davis, D. Cronauer, A.J. Kropf, C.L. Marshall, Appl. Catal. A:
Gen. 333 (2007) 177–191.
[22] G. Jacobs, M.C. Ribeiro, W. Ma, Y. Ji, S. Khalid, P.T.A. Sumodjo, B.H. Davis, Appl.
Catal. A: Gen. 361 (2009) 137–151.
[23] A. Lapidus, A. Krylova, V. Kazanskii, V. Borovkov, A. Zaitsev, J. Rathousky, A.
Zukal, M. Jancalkova, Appl. Catal. 73 (1991) 65–81.
[24] Ø. Borg, E.A. Blekkan, S. Eri, D. Akporiaye, B. Vigerust, E. Rytter, A. Holmen, Top.
Catal. 45 (2007) 39–43.
[25] A. Martinez, C. Lopez, F. Marquez, I. Diaz, J. Catal. 220 (2003) 486–499.
[26] J.L. Zhang, J.G. Chen, J. Ren, Y.H. Sun, Appl. Catal. A 243 (2003) 121–133.
[27] A.R. Belamble, R. Oukaci, J.G. Goodwin, J. Catal. 166 (1997) 8–15.
[35] D.C. Song, J.L. Li, Q. Cai, J. Phys. Chem. C 111 (2007) 18970–18979.
[36] L.E.S. Rygh, O.H. Ellestad, P. Klaeboe, Phys. Chem. Chem. Phys.
1835–1846.
[37] L. Ji, J. Lin, H.C. Zeng, J. Phys. Chem. B 104 (2000) 1783–1790.
[38] M.J. Heal, E.C. Leisegang, R.G. Torrington, J. Catal. 51 (1978) 314–325.
[39] F.M.T. Mendes, C.A.C. Perez, F.B. Noronha, C.D.D. Souza, D.V. Cesar, H.J. Freund,
M. Schmal, J. Phys. Chem. B 110 (2006) 9155–9163.
[40] V. Curtis, C.P. Nicolaides, N.J. Coville, D. Hildebrandt, D. Glasser, Catal. Today 49
(1999) 33–40.
[41] G. Kadinov, C. Bonev, S. Todorova, A. Palazov, J. Chem. Soc. Faraday Trans. 94
(1998) 3027–3031.
[42] E.L. Rodrigues, J.M.C. Bueno, Appl. Catal. 232 (2002) 147–158.
[43] S.Z. Todorova, G.B. Kadinov, Res. Chem. Intermediat. 28 (2002) 291–301.
[44] Y.Q. Chen, M.C. Gong, X.H. Zhu, H. Ming, S.H. Tang, Chem. J. Chin. Univ. 4 (2002)
715–717.
2 (2000)