Performances of promoted cobalt catalysts supported on mesoporous alumina for Fischer-Tropsch synthesis

Article abstract of DOI:10.1016/j.apcata.2017.08.032

Noble metals (Ru, Re, Pt, Ir) and transition metals (Mn, Y) with their atomic ratio of 0.01 to cobalt were used as promoters on the cobalt catalysts supported on mesoporous alumina. The cobalt loading was 15 wt% and all catalysts were tested for Fischer-Tropsch synthesis in a fixed-bed reactor. The process conditions for activity test were: 230 ° C and 400 psi, GHSV = 900 h^?1, and [formula omitted]=2.0. The selectivity test for the promoted and unpromoted catalysts was performed at CO conversion ≈50% by adjusting the syngas flowrate. Addition of promoters increased the cobalt dispersion on mesoporous alumina support. The highest increase in dispersion from 4.7% to 7.6% resulted from addition of Ru. Promoters positively impacted the cobalt dispersion in the decreasing order of: Ru < Re < Pt < Ir < Mn < Y. The reduction temperature for all promoted catalysts were shifted to lower temperature. The maximum decrease in temperature happened with platinum by ΔT ～ ? 150°C for the 1st reduction temperature and ΔT ～ ? 270°C for the 2nd reduction temperature. The reducibility of the CoO_x species on mesoporous alumina was extensively improved by noble metals addition and modestly enhanced by Mn and Y. Significant increase in CO conversion and C₅₊ of the promoted catalysts were observed. CO conversion was increased from ～60% to ～86, 84, 82, 80% as a result of Ru, Re, Ir, Pt addition. Addition of Ru and Re promoters favored the heavy molecular weight hydrocarbons’ (C₅₊) selectivity and improved it from 81.7% to 84.2 and 83.1%, respectively.

Full text of DOI:10.1016/j.apcata.2017.08.032

Accepted Manuscript
Title: Performances of promoted cobalt catalysts supported on
mesoporous alumina for Fischer-Tropsch synthesis
Authors: Vahid Vosoughi, Ajay K. Dalai, Nicolas Abatzoglou,
Yongfeng Hu
PII:
S0926-860X(17)30412-X
DOI:
Reference:
http://dx.doi.org/10.1016/j.apcata.2017.08.032
APCATA 16387
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
3-5-2017
7-8-2017
21-8-2017
Please cite this article as: Vahid Vosoughi, Ajay K.Dalai, Nicolas Abatzoglou,
Yongfeng Hu, Performances of promoted cobalt catalysts supported on
mesoporous alumina for Fischer-Tropsch synthesis, Applied Catalysis A,
Generalhttp://dx.doi.org/10.1016/j.apcata.2017.08.032
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Performances of promoted cobalt catalysts supported on mesoporous alumina for Fischer-
Tropsch synthesis
Vahid Vosoughi1, Ajay K. Dalai1,*, Nicolas Abatzoglou2, Yongfeng Hu3
1Catalysis and Chemical Reaction Eng. Lab., Department of Chemical and Biological Eng.,
University of Saskatchewan, SK, Canada, S7N 5A9
2Chemical and Biotechnological Eng. Dep., Université de Sherbrooke, QC, Canada, J1K 2R1
3SXRMB Beamline, Canadian Light Source, Saskatoon, SK, Canada, S7N 2V3
*Corresponding author: ajay.dalai@usask.ca
Graphical abstract
1

Highlights
 Nobel metals as promoters significantly improved the reducibility of Co/mAl2O3
catalysts.
 The order of promoters’ impact on the enhancement of reducibility for Co/mAl2O3
catalysts was Y < Mn < Re < Pt < Ru < Ir.
 Mn and Y slightly affected the Co/mAl2O3 catalysts’ properties and performance for
FTS.
 Ru and Re had the highest promotional impact on the activity and selectivity of
Co/mAl2O3 catalyst for FTS.
 Re-, Mn-, and Y-promoted Co catalysts showed the lowest deactivation in FTS.
Abstract
Noble metals (Ru, Re, Pt, Ir) and transition metals (Mn, Y) with their atomic ratio of 0.01 to cobalt
were used as promoters on the cobalt catalyst supported on mesoporous alumina. The cobalt
loading was 15 wt % and all catalysts were tested for Fischer-Tropsch synthesis in a fixed-bed
reactor. The process conditions for activity test were: 230 ˚C and 400 psi, GHSV = 900 h−1, and
퐻
2
= 2.0. The selectivity test for the promoted and unpromoted catalysts was performed at CO
퐶푂
conversion ≈ 50% by adjusting the syngas flowrate. Addition of promoters increased the cobalt
dispersion on mesoporous alumina support. The highest increase in dispersion from 4.7% to 7.6%
resulted from addition of Ru. Promoters positively impacted the cobalt dispersion in the decreasing
order of: Ru < Re < Pt < Ir < Mn < Y. The reduction temperature for all promoted catalysts were
shifted to lower temperature. The maximum decrease in temperature happened with platinum by
Δ푇~ − 150 °퐶 for 1st reduction temperature and Δ푇~ − 270 °퐶 for 2nd reduction temperature. The
reducibility of the CoOx species on mesoporous alumina was extensively improved by noble
metals addition and modestly enhanced by Mn and Y. Significant increase in CO conversion and
C5+ of the promoted catalysts were observed. CO conversion was increased from ~ 60 % to ~ 86,
84, 82, 80 % as a result of Ru, Re, Ir, Pt addition. Addition of Ru and Re promoters favored the
heavy molecular weight hydrocarbons’ (C5+) selectivity and improved it from 81.7 % to 84.2 and
83.1 %, respectively.
2

Keywords: Fischer-Tropsch synthesis, Mesoporous alumina, Cobalt catalyst, CO hydrogenation,
Promoter, Noble metal, Transition metal
1. Introduction
It is crucial nowadays to develop the alternative (non-crude oil-based) and environmentally sound
routes to produce transportation fuels and hydrocarbon−based specialty products. Fischer-Tropsch
synthesis (FTS) and associated technologies such as gas-to-liquid (GTL), coal-to-liquid (CTL),
biomass-to-liquid (BTL), currently seem to be promising options to synthesize the value added
liquid hydrocarbons (HCs). The XTL (X: natural gas, coal, biomass, petroleum coke, etc.)
technologies consist of three major steps: (1) syngas (mixture of H2 and CO) production and
purification from carbonaceous resources, (2) syngas polymerization to wide range of long chain
HCs through FT reaction, and (3) FT products work−up (refining) [1,2]. FTS potentially can
convert the syngas to light hydrocarbons (C1−C4), premium middle distillate cuts (naphtha, jet
fuel, diesel), and heavier solid waxes (C22+). Aside from the impact of feed composition, reactor
type, and operational conditions on productivity and product distribution of FTS, the catalyst
characteristics including the support material, active metal, additive, and preparation method are
crucial for the better catalysts’ performance [1,3]. Among group VIII transition metals, iron and
cobalt distinctively favor carbon monoxide hydrogenation and are commercially used in FT
catalysts. Iron catalyst has higher water-gas-shift (WGS) activity, therefore is more suitable for
CTL and BTL with lower H2:CO ratio (even 0.6−1.0) syngas. In contrast, cobalt performs better
with higher H2:CO ratio (e.g.1.8−2.3) syngas driven from natural gas [4,5]. In addition, it is less
active toward WGS reaction and CO2 production, but more selective to CH4 and paraffinic C5+.
Higher intrinsic activity and better stability make the Co as a commercially preferred catalyst for
FTS targeting long−chain alkanic hydrocarbons. However, understanding the catalysis, designing,
and developing of more active, selective, and stable cobalt−based catalyst for FTS have been a
continuing interest of research in feed-to-liquid (XTL) technology [6–8].
It is well-established that the nature of carrier (support) and its textural properties, active metal
precursor and its loading quantity, nature/amount of additives and catalyst preparation methods
would significantly influence the physico-chemical properties of a supported cobalt−based
catalyst. These properties include cobalt dispersion, crystallite size, degree of reduction, electronic
structure, number/structure of active sites, and cobalt-support interaction. Most of the mentioned
3

characteristics of a catalyst are mainly interconnected with the last two, i.e., the strength of cobalt-
support bonding and the quantity of cobalt active sites. The type of support material and promoters
can directly affect these two crucial features of the active cobalt metal and indirectly modify the
other physico-chemical properties and catalytic performance in FTS. Promoters for instance are
the compounds that can be doped on active metal at very small quantity to enhance its catalytical
performance [9,10]. These elements can function as either a structural, electronic, textural modifier
or play a role of stabilizer and poison-resisting agent for the catalyst. These modifiers can influence
the surface properties of the support, affect the crystallite size and morphology of the active metal
and catalyst lifetime [10,11].
The structural promotion refers to influences on the cobalt−support interaction which results in
higher cobalt dispersion, enhanced reducibility and stabilized Co active sites (ensemble effect)
[10,12]. These might lead to an increase in the conversion/yield or a decrease in the Co particles
sintering/agglomeration during the FT reaction [9,11]. It has been speculated that the electronic
promoter may affect the adjacent atoms environment (ligand effect) and the active site geometry
through intimate mixing with active metal [11,12]. These can induce changes on the fundamental
pathways (elementary steps) of FT reaction (CO and H2 adsorption, C1 monomer formation, chain
growth and termination) [13,14].The electronic/chemical alteration can directly affect the intrinsic
site activity and selectivity of Co catalyst in FTS [10–12,15].
Transition metals (e.g. Ni, Cu, Mo, V, Cr, Nb, Zr, Ti) including the noble metals (e.g. Ru, Re, Ir,
Pt, Pt, Pd, Rh, Os, Au, Ag), rare earth metals (e.g. Ce, La, Sc, Pr, Y), alkali and alkaline earth
metals (Li, Na, K, Rb, Cs, Ca, Mg) have been studied as promoters for the supported cobalt catalyst
in FTS [9,11,16–22]. Platinum, ruthenium, rhenium, iridium from noble metals and manganese,
lanthanum, zirconium among transition metals are the most patented promoters for FT Co catalyst
[23–26]. The high cost, low natural abundance, and impact of precious metals on the cobalt
catalytic performance are the key factors to be considered in screening the effective 2nd metal for
FT catalyst. For instance, Pt is about 3 times expensive than Ru but 5 times abundant. Re is the
most scarce noble metal (~ 7 times less than Pt) but the most inexpensive option (~ 10 time cheaper
than Pt) [27].
Some of the parameters, which can significantly influence the promotional effect of the elements
on Co-supported FT catalyst are: (1) the nature of support surface and its textural properties, (2)
4

cobalt and promoter loading, (3) preparation method of the promoted cobalt catalyst, and (4) FTS
process condition. All of these variables affect the number of active sites, their stability, and cobalt-
support interaction which make it difficult to compare the impact of a promoter solely on the
performance of cobalt catalyst in FT reaction [9,11]. The promotional mechanism of the second
metal added to Co catalyst for FTS have not been elucidated comprehensively. However, transition
metals and their oxides are generally added to ensure the selectivity for heavier hydrocarbons and
stability of the cobalt catalyst [9,28]. The noble metals are also well−known for their considerable
improvement in the reducibility and dispersion of cobalt catalysts supported on SiO2 [29,30], TiO2
[31,32], Al2O3 [33,34].
In our earlier work [35], the mesoporous alumina was synthesized with high surface area (>400
m2/g) and large pore diameter (> 10 nm), also was loaded with Co for FTS favoring C5+ HC. The
objective of this work was to continue development of the mesoporous alumina supported cobalt
catalyst for FTS via addition of the 2nd metal as promoter. Since the higher cobalt dispersion and
extent of reduction (> 80%) are beneficial for HCs yield and selectivity, their improvement via
employing promoter metals was explored on the synthesized mesoporous alumina support. In this
regard, the commercially used noble metals (Ru, Re, Pt, Ir) with 1% molar ratio of promoter to Co
were chosen in this study. The same concentration level of manganese (Mn) and yttrium (Y) from
transition metals were also examined. There is a limited research in the literature studying the
promotional effect of Y on Co based catalyst in FTS. The impact of adding the transition and noble
metals on physico-chemical properties of the cobalt catalyst using mesoporous alumina support
were investigated employing in-depth characterization. Further, all corresponding promoted cobalt
based catalysts were tested for their activity and selectivity performance for Fischer-Tropsch
synthesis in a fixed−bed reaction system.
2 Experimental methods
2.1 Materials
For the mesoporous alumina support synthesis, aluminum isopropoxide Al(O-i-Pr)3 as a precursor
of alumina, and pluronic F127 [(EO)106(PO)70(EO)106] as a structure-directing agent were
purchased from Sigma–Aldrich, Oakville, Canada. Also, anhydrous ethanol and isopropanol as
solvents used were procured from Fisher Scientific, Ottawa, Canada. The following active and
5

promoter metal precursors were purchased from Alfa Aesar, Tewksbury, United States: cobalt
nitrate [Co(NO3)2.6H2O, 99.9%], Mmanganese (II) nitrate hydrate [Mn(NO3)2.xH2O, 99.999%],
rhenium (VII) oxide [R2O7, 99.99%], ruthenium (III) nitrosyl nitrate [Ru(NO)(NO3)2, 99.9%],
platinum (IV) nitrate solution [PtN4O12, Pt 15% w/w], yttrium (III) nitrate hexahydrate
[Y(NO3)3.6H2O, 99.9%], Iridium (III) chloride hydrate [IrCl3.xH2O, 99.9%].
2.2. Support synthesis
Non-ionic templating sol-gel method was adopted to synthesize the mesoporous alumina (mAl2O3)
as a support material. In this method, aluminum alkoxide (e.g. isopropoxide) is hydrolyzed in the
presence of a long chain structure-directing agent. This template was triblock copolymer
surfactant, Pluronic F127, with a hydrophilic poly ethylene oxide (PPO) connected to two
hydrophilic poly ethylene oxide (PEO) chains [36]. Employing this templating agent, with
appropriate ratios of precursor and solvents secure the high specific surface area (>300 m2/g) and
pore volume (>1 cm3/g) with mesoporous structure. These polymers are relatively non−toxic,
biodegradable, inexpensive, and commercially available compounds. Moreover, employing
different ratio of H2O/Al in the synthesis provides the flexibility of tuning the pore diameter
[37,38]. The moderately large pore diameter of 10-15 nm was targeted during the synthesis
procedure to meet the recommended properties for cobalt catalyst in FTS [7,39]. The chosen molar
ratio of the precursors and solvents for this synthesis was 1.0 Al(o-i-Pr)3: 0.01 F127: 8.0 EtOH:
6.0 iPrOH: 4.0 H2O. Aluminum iso-propoxide and non-ionic F127 copolymer were separately
dissolved in the mixture of ethanol and isopropanol at 50 ºC. Appropriate amount of water was
introduced into the polymer solution before adding the aluminum iso-propoxide solution into it.
Following the 4 h stirring of the suspension at 50 ºC, it was aged at room temperature for 24 h.
The resulting gel was treated at two thermal stages of 80 ºC and 150 ºC for 24 h each. The thickened
gel was then filtered, dried and calcined at 600 ºC for 6 h. Further details of the synthesis procedure
for the mesoporous alumina support have been explained elsewhere [35,37-38].
2.3. Catalyst preparation
To prepare the promoted cobalt catalysts, incipient wetness impregnation (IWI) method was
employed to fill the pores of the support. The required amount of cobalt precursor to secure 15 wt
% of Co loading on the catalyst along with promoter precursor were dissolved in the anhydrous
6

ethanol for co−impregnation on the dried mesoporous alumina support. The atomic ratio of the
added promoter metals to cobalt metal was 0.01 which is equivalent to 0.27, 0.5, 0.52, 0.28, 0.15,
0.24 wt % of Ru, Re, Pt, Ir, Mn, and Y in the catalyst, respectively. All catalysts were vacuum
dried in 60 °C for 12 h and then calcined at 400 °C for 3 hours.
2.4. Catalyst Characterization
Nitrogen physisorption: The adsorption and desorption amount of nitrogen at cryogenic
temperature of 77 K was measured using Micromeritics ASAP 2020 instrument. To strip all
possible volatile content, about 0.2 g of the samples was vacuum heated first at 50 mTorr and 200
ºC for 4 hours. The specific surface area and pore dimensions were calculated using the Brunauer-
Emmett-Teller (BET) and Barrett-Joyner-Halenda (BJH) method [40,41].
H2 chemisorption: The hydrogen adsorption amount to estimate the number of cobalt atoms
available on the surface was measured using chemisorption (ASAP 2020 instrument). 0.1 g of the
calcined catalysts was first reduced by hydrogen at 400 ºC for 150 min then cooled down to 35 ºC
under 1.3×10−5 Pa vacuum. Then the hydrogen gas was introduced on sample to determine the H2
uptake and cobalt atoms accordingly, assuming H:Co stoichiometry of 1:1 for chemisorption [40–
42].
measured Co0atoms on the surface
D =
(1)
total Co0 atoms in the sample
Where D is the fraction of available cobalt atoms (Co0) in the surface available for chemisorption
out of the whole cobalt available in the sample [34,42].
Temperature programmed reduction (TPR): To study the impact of promoter on reducibility of
the mesoporous supported cobalt catalysts AutoChem Micromeritics 2950 HP instrument was
used, which is equipped with a thermal conductivity detector. Prior to the analysis, the catalyst
samples were degassed first at 100 ºC, then 35 ºC by introducing the helium gas into the stainless
steel U−shaped sample holder. Reduction profiles of the catalysts were acquired by increasing the
temperature with 10 ºC/min ramping steps to 800 ºC. This reduction behavior was obtained under
50 mL/min flow of 10 % H2/N2 (v/v) gas and recorded using the thermal conductivity of the outlet
7

gas via a detector [40,41]. Degree of reduction (DOR) for the catalysts was calculated by taking
the ratio of the consumed H2 for reduction from TPR analysis over the required stochiometric H2
to reduce all cobalt oxides (Co3O4) to metallic Co. Then, the dispersion of Co atoms was corrected
by associating it with the extent of reduced cobalt atoms [34,42]:
measured Co0atoms on the surface
Reduced Co0 atoms in the sample
Dcorr.% =
× 100
D %
=
(2)
DOR
In order to calculate the Co0 crystallite size based on the Co dispersion amount, the cobalt clusters
were assumed to be in spherical shape with a site density of 14.6 atoms/nm2 [34,42]:
96
0
Co
d
=
chem.
Dcorr.
%
96
=
× DOR
(3)
D %
X−ray diffraction (XRD): To carry out the powder XRD analysis for the catalysts, the Bruker
Advance D8, series II, diffractometer with monochromatic CuKα radiation was used. The
operating voltage and current were 40 kV and 40 mA, respectively. The broad angle XRD from
10 to 90º was achieved at 2º/min scan rate and 2 s step time. To estimate the mean diameter of the
cobalt oxide crystallites, Debye–Scherrer equation was applied at 2θ = 36.8º:
Kλ
180
π
Co O
XRD
3
4
d
=
×
(4)
β cos θ
Co O
XRD
3
Where, d
4 is the average thickness of crystallites (nm); K is assumed 0.89 as shape factor
constant for sphere crystallites; λ is X−ray wavelength = 0.154 (nm); β is broadening at half-max
intensity (full width at half maximum) in degree; and θB = 2θ/2 is Bragg angle (diffraction angle)
[41–43].
Transmission electron microscopy (TEM): To examine the morphology of the samples through
their micrographs, the JEOL 2011 transmission electron microscope was used. The standard
sample preparation was followed to carry out the analysis [40,41].
8

X−ray absorption near−edge structure (XANES) spectroscopy: The Co K−edge, XANES of the
promoted and unpromoted cobalt catalyst supported on mesoporous alumina were obtained at the
Soft X−ray Microanalysis Beamline (SXRMB) of the Canadian Light Source with medium energy
range of 1.7–10 keV. To acquire the scan for X-ray spectra the monochromator equipped with
double−crystal silicon (1 1 1) was utilized. The double−sided conductive carbon tape was used to
disperse the sample on it. The data for XANES spectrum was collected in two modes: total electron
yield by measuring the sample drain current; and partial fluorescence yield using Si(Li) drift
detector. The XANES spectra were corrected with background subtraction, and normalized to the
incident photon flux and to unity using polynomials..
2.5. Catalyst performance study for Fischer-Tropsch Synthesis
Fischer–Tropsch reaction was carried out in the fixed−bed micro reactor with 1 cm internal
diameter and 50 cm length. To ensure uniform mass flow and heat transfer across the reactor, five
different layers of silicon carbides (SiC) from coarse to fine toward the reaction zone were used.
Moreover, for each run, one gram of the powder catalyst was diluted with 7 g of SiC (90 mesh).
Prior to FT reaction, in-situ reduction occurred using ultra−high purity (UHP) hydrogen. To reduce
the catalyst, the reactor temperature was raised with 1 ºC/min step to 400 ºC and maintained for
18 h under 30 ml/min flow of H2 and atmospheric pressure. Following the reduction, the reactor
was cooled down to 180 ºC and pressurized to 400 psi, and the inlet gas was switched to syngas
(H2:60%, CO:30%, Ar: 10%). After syngas introduction, the reactor temperature was gradually
raised to 230 ºC with heating ramp of 0.5 ºC/min. Brook mass flow controller and PID temperature
controller were used to adjust the flow and temperature to desired values for FT reaction. The
products were first passed through a container wrapped with heating tape and maintained at about
100 ºC to collect heavy HCs and waxes in liquid form. The lighter products and water are
condensed in cold trap which was maintained at around 0 ºC. The non−condensable reactant and
product gas mixture (CO, Ar, CO2, C1−C4) was analyzed through gas chromatograph (Shimadzu
GC−2014) in 8 h intervals, otherwise sent to exhaust gas line [41,44,45]. Liquid hydrocarbons
were collected from the hot and cold traps and analyzed via an off−line GC.
After doing mass balance over the argon as a non−reactive component in the reaction, following
equations were used to estimate the activity and selectivity the catalyst in FTS:
9

CO conversion percentage: COconv.
%
Mole COin − Mole COout
Mole COin
( )
5
=
× 100
Selectivity of CO2: SCO
=
%
2
Mole CO2,out
× 100
(6)
Mole COin − Mole COout
Mole Cn × 100
Mole COin − Mole COout − Mole CO2,out
(
)
Selectivity of light HCs 푛 = 1, 2, 3, 4 : SC % =
(7)
(8)
n
(
)
Selectivity of long − chain HCs n ≥ 5 : SC % = 1 − (SC + SC + SC + SC )
5+
1
2
3
4
3. Result and Discussion
3.1. Characterization
As shown in Table 1, co−impregnation of cobalt along with different promoters had rather similar
impact on the textural properties of the mesoporous alumina. Deposition of cobalt oxides during
the calcination partially fills the porosity of the support. The 15 wt % Co0 loading in the calcined
catalysts is equivalent to the presence of ~ 20 wt. % Co3O4. If the contribution of cobalt oxides in
the surface area is neglected, it is expected that after the catalysts are prepared and calcined, the
support would lose around 20% of its specific surface area (SSA). The unpromoted and all
promoted catalysts revealed the SSA in 310−330 m2/g limit which shows 18.5−23% decrease in
the SSA of the support (405 m2/g), conforming the expected decline. Pore volume of all catalysts
dropped from 1.55 to 1.11−1.28 cm3/g. However, not considerable changes were observed for pore
diameter of corresponding catalyst considering the experimental error of ±0.2 nm. The mesoporous
alumina support exhibited the pore diameter of 11.8 nm and the corresponding unpromoted and
promoted catalysts showed the average pore diameter between 11.3−12.0 nm.
To study the impact of added promoters on dispersion of the cobalt catalysts supported on
mesoporous alumina, H2 chemisorption was preferred over the CO chemisorption [46] due to the
difficulty in determining the number of chemisorbed CO atoms on a Co atom (stoichiometry of
CO:Co ~ 1−2) [47,48]. H2 chemisorption results shown in Table 2 indicate that the promotion with
10

noble metals (Ru, Re, Pt, Ir) led to higher cobalt dispersion (D%) than transition metals (Mn and
Y). The highest increase in apparent dispersion of Co in the Co/mAl2O3 catalyst was obtained by
the addition of Ru and Re promoters, changing the Co dispersion from 4.8% to 7.6% and 7.1%,
respectively. The order of improvement in dispersion of Co because of using different promoters
is as follows: Ru > Re > Pt > Ir > Mn > Y; the transition metals only slightly improved the Co
apparent dispersion (from 4.8 to ~ 5.3%). The interaction of promoters with Co, support and their
location is complicated. Using aberrated−corrected scanning transmission electron microscopy,
Shannon et al. found that the isolated atoms of the Re, Pt, and Ir were present on the Co surface,
whereas the Ru appeared in patches but all connected with Co particles [49]. However, it is
believed that the presence of promoter during the oxidation treatment (calcination) facilitates the
nucleation of cobalt oxide crystallites and crystallization sites. Therefore, the probability of the
smaller cobalt crystallite formation increases which results in the enhancement of the cobalt
dispersion and active site density [50].
Temperature programmed reduction (TPR) profiles in Figure 1 present the reduction behavior of
unprompted and promoted Co/mAl2O3 catalysts. It is evident that the noble metals considerably
shifted the reduction temperature to lower degrees, while the transition metals impact on
reducibility were not significant. It is well-established for the supported cobalt that at least two
prevailing peaks appear on its TPR profile. The first peak represents reduction of Co (II, III) to Co
(II) crystallites, Co3O4  CoO, which reportedly occurs in range of 200−300 ºC. The second peak
signifies reduction of Co (II) to zero valance cobalt, CoO  Co0, which might take places within
400−500 ºC [51,52]. In this work, the first and second reduction peaks of unpromoted Co/mAl2O3
appeared at 260−330 ºC and relatively broad range of 350−620 ºC, respectively. The calcination
temperature and cobalt-alumina interaction could underlie the location of the reduction peaks [51].
The broad-ranging reduction stage of CoO  Co0 can be attributed to the reduction of two different
cobalt species which starts at lower temperatures and extends to the higher temperatures due to:
(1) the reduction of strongly bonded Co3O4 to the support, and (2) the reduction of strongly bonded
CoO to the support or those cobalt ions which could be embedded in the alumina lattice and formed
hardly reducible species [9,51].
11

All promoters used in this work increased the cobalt reducibility for the mesoporous alumina
supported cobalt catalyst, and shifted either both or one of the occurring reduction ranges to lower
temperature. As illustrated in Figure 1, the first maxima of reduction was shifted from 300 ºC to
186, 296, 146, 188, 220, 229 ºC and the second maxima from 592 ºC to 325, 352, 333, 345, 567,
574 ºC, as a result of adding Ru, Re, Pt, Ir, Mn, Y promoters, respectively. The first stage of
reduction (Co3O4 → CoO) seems to be facilitated more with noble metals (i.e. ΔT =112−154 ºC)
than with Mn and Y transition metals (i.e. ΔT = 71−80 ºC). However, Re was almost ineffective
on shifting the maxima for the first stage of the reduction which concurred with findings of other
research groups [11,53]. It is believed that the rhenium oxide itself reduces at higher temperature
(~ 350 ºC) that may be the reason why the first reduction stage of the cobalt was not facilitated
much by rhenium [9,53]. Regarding the second peak of reduction, all noble metals exhibited
considerable impact on shifting the maxima of CoO → Co0 stage to lower temperatures (i.e.
ΔT =140−270 ºC) forming distinct narrow peak as compared to unprompted Co/mAl2O3 catalyst,
whereas the Mn and Y lowered the second peak reduction temperature only by 25 and 18 ºC,
respectively. The corresponding surface area for the main peaks was also estimated using
deconvolution of the reduction profiles and presented in Table A3 in Appendix A. Supplementary
data.
Despite the fact that the conditions for TPR analysis were not the same as the condition for in-situ
reduction of the catalyst before FT reaction in terms of purity of hydrogen (10% vs UHP),
reduction temperature (35 to 700 ºC vs 400 ºC), and duration (2−3 h vs 18 h), the measured H2
uptake and estimated DOR could be used as basis for comparison. To estimate the stochiometric
amount of H2, it is assumed that all Co3O4 were converted to Co0 following reactions (9) and (10)
[54]:
Co3O4 + H2 → 3CoO + H2O
3CoO + 3H2 → 3Co0 + 3H2O
(9)
(10)
12

As it is presented in Table 2, addition of the noble and transition metals to Co improved the
reducibility of the corresponding mesoporous alumina supported cobalt catalysts. Structural
contribution of the added promoters on the cobalt dispersion and its reducibility was in the same
order of enhancement (Y < Mn < Re < Pt < Ru < Ir). It was observed that the 2nd metals (noble)
added to the Co/mAl2O3 catalyst in 1% molar ratio level, considerably facilitated the reduction of
the catalyst. Degree of reduction (DOR) for unpromoted catalyst was enhanced from ~ 61% to ~
82, 84, 85, and 87% by addition of Re, Pt, Ru, and Ir promoters, respectively. The same order of
reducibility for the fresh calcined Co/alumina catalysts was reported in other study [17]. Mn and
Y also improved the extent of reduction for the corresponding promoted catalysts by absolute
percentage increase of 10 and 13, respectively, as compared to the unpromoted counterpart catalyst
(Co/mAl2O3). As it is evidenced, at equal ratio of the promoter to cobalt atoms, the noble metals
are much more effective promoters than transition metals by lessening the catalyst reduction
temperature, thereby, (i) facilitating the reduction of hardly−reducible small cobalt crystallites
interacting strongly with the support, and (ii) increasing the number of Co0 active sites on the
surface [9,55].
The ease of reductions could be associated with rapid dissociation of hydrogen molecules on the
noble metal surface and following diffusion of hydrogen atoms (spillover) to neighboring cobalt
oxide species [50,56]. Different electronic properties of transition metals (as substrates) exhibit
different affinity toward adsorbed atoms or molecules (adsorbates). Iglesia et al. [57] disclosed
formation of Ru−Co oxides species and bimetallic alloying during the calcination and catalyst
activation, respectively, using x-ray adsorption spectroscopy (XAS) technique. Formation of
RuOx, which are easily-reducible and can be mobile to bond with CoOx, assists the reduction of
CoOx ensembles during H2 activation of the catalyst (structural promotion) as well [58]. The
chemisorption model, the d−band theory, developed by Hammer and Nørskov [59] also
successfully explains the different reactivity of transition metals toward gaseous adsorbates (e.g.
H2). As one moves from the bottom right to top left through the d−block metals of periodic table,
the contribution of unpaired d-electrons to surface bonding of adsorbate increases. This can be
related to degree of filling (df) in the d−band states of metals where it becomes more depopulated
as one shifts to the earlier transition metals in the left which results in formation of stronger
chemisorption bonds (e.g. dPf t = 0.9 and dfY = 0.2) [59]. Appropriate chemisorption strength
13

favoring the hydrogen spillover and facilitating the cobalt oxide reduction, follows the Sabatier
principle and demonstrates volcano type trend among transition metals. The later elements onward
the right columns of transition metals (e.g. Ag, Au) show poor dissociative chemisorption and the
earlier elements (e.g. Zr, Sc) on the left columns of transition metals form strong bond that
hydrogen atoms cannot desorb easily after adsorption, whereas the noble metals behave in between
[59,60].
Apparent dispersion increased by ~ 2−3 absolute % (from 4.7% to 6.6−7.6%) owing to addition
of the noble metals to cobalt in 1% molar ratio of promoter (Table 2), whereas Mn and Y had
insignificant impact on increasing the fraction of cobalt atoms available on the surface (from 4.7
to 5.2 and 5.4, respectively). This might be associated with strong interaction of H2 with Mn and
Y which become difficult for dissociated H atoms to desorb, spill over, and reduce the neighboring
CoOx species to metallic Co. During the reduction, the MnOx and YOx species remain in their
oxidic form while they adsorb the reducing hydrogen strongly. As mentioned above, the promoter
loading level and catalyst preparation method can affect the MnOx and YOx species location in the
catalyst to be either on the CoOx surface or alumina support. As reviewed by Morales et al [11],
Mn could aid the dispersion of cobalt if it is bound with the support. Considering the fact that all
cobalt species after reduction may not be in form of Co0 and still remain in oxidised form, degree
of reduction needs to be taken into account. After calculating the DOR, the apparent dispersion of
the catalysts was corrected (Dcorr.%) as shown in Table 2. According to the equation (2), the bigger
the degree of reduction (DOR → 1), the smaller difference would be between the apparent cobalt
dispersion and the corrected dispersion value (D% → Dcorr.%). The higher dispersion of cobalt
could be translated to more available cobalt surface area and active sites for the CO hydrogenation
0
in FTS. The average surface Co0 crystallite diameter (dCchoem.) was also corrected based on the
percentage of the Co atoms which are present in the reduced form. As shown in equation (3), the
higher metal dispersion results in smaller crystallite of the cobalt metal. The Ru exhibited the
highest impact on improving the dispersion of Co/mAl2O3 which resulted in the smallest Co0 size
of 10.7 nm. The catalyst with the lowest corrected dispersion (DY% = 7.2) led to the largest average
Co0 crystalline diameter of 13.4 nm. Given that the average surface diameter of the Co0 crystallites
for unpromoted catalyst was estimated of 12.5±0.5 nm, addition of one atom promoter for every
14

0
100 atoms of the cobalt did not noticeably change the dcChoem. for the promoted catalysts. This could
be related to strongly interacting Co particles with the alumina support.
HRTEM images (Fig. 2) also confirm the analogy among unpromoted and promoted cobalt
catalysts. These micrographs do not reveal information about the doped promoters position
whether they interacted with cobalt surface or segregated on the support. However, the images
from calcined catalysts exhibited well distribution of cobalt crystallites on the mesoporous alumina
support. Mn-promoted catalyst exhibited somewhat more coalescence of Co particles with rather
bigger sizes. The EDAX mapping for unpromoted and two promoted catalysts were appended in
Fig. A2 of Appendix A. (Supplementary data) identifying the present Co and Al elements.
The wide-angle XRD patterns of the unpromoted and promoted cobalt catalysts supported on the
mesoporous alumina as well as the spectra for the powder CoO and Co3O4 are demonstrated in
Figure 3. The corresponding XRD spectra for the synthesized alumina was discussed elsewhere
[35]. The reflections from (311), (222), (400), and (440) planes in γ-Al2O3 cubic structure appear
at 2θ: 37º, 39º, 46º, and 66º respectively [61,62]. CoO and Co3O4 are the main possible cobalt
oxides would present in a supported and calcined cobalt catalysts. Diffraction peaks at 42.7º and
61.9º, which represent the CoO, were not distinguished for all the unpromoted and promoted cobalt
catalysts supported on the mAl2O3, whereas the Co(II, III) was the prevalent cobalt oxide phase in
the series as evidenced by reflections from different planes of Co3O4 at 2θ of 19.0º, 31.5º, 36.8º,
38º, 45º, 59.6º, 65.5º [63].
The XRD patterns were also used to estimate the Co3O4 crystallite size in the catalyst employing
the equation (4). FWHM method for estimation of angular breadth (β) was applied at 2θ = 36.8º,
the most intense peak of Co3O4 phase representing (311) face of crystallites, to obtain the average
width of cobalt oxide crystallites in the catalysts. As the results shown in Table 2, the crystallite
Co O
XRD
3
size (d
4) of promoted cobalt catalysts from XRD analysis also verifies that the noble metal
addition did not notably influence the cobalt oxide size on mesoporous alumina (−0.4 to +1.0 nm
change from 10.8 nm). However, doping Mn and Y resulted slightly more change in the cobalt
oxide crystallite size as compared to noble metals, increasing it from 10.8 nm to 13.2 and 13.5,
respectively. The increase in the crystallite size might be attributed to weakening of the
15

Co−support interaction which in turn causes the migration of smaller atoms toward larger cobalt
crystallites. In the case of manganese, deposition of MnO2 on CoOx and formation of Co3−xMnxO4
(0 < x < 1.4) spinels have been also evidenced [11]. The XRD results of Co3O4 width were
0
multiplied by coefficient of 0.75 to estimate the Co0 size (dXCoRD) presented in Table 2. The Co0 size
trend estimated from XRD corroborate with the findings of the H2 chemisorption analysis.
Nonetheless, the crystallite size of Co0 approximated from chemisorption analysis, for the
corresponding catalysts, are considerably larger. The difference in the average Co0 size determined
using two methods could be ascribed to: (i) enlargement of cobalt ensembles during the reduction
in H2 chemisorption analysis and (ii) inaccessibility of noticeable amount of the present metal for
H2 chemisorption.
The Co K-edge normalized XANES spectra of unpromoted and promoted cobalt catalysts
supported on the mesoporous alumina are shown in Figure 4. The reference spectra for cobalt
oxide II (CoO) and cobalt oxide II & III (Co3O4) standards are also shown in Figure 4. The intense
peak or edge jump occurs at ~ 7730 eV owing to the dipole allowed 1s to 4p electronic transition.
The edge jump for CoO is 4eV lower than that of Co3O4. This is due to the higher oxidation state
of cobalt in Co3O4 as it requires higher energy to pull out additional electron. The XANES spectra
for all calcined cobalt catalysts were fitted using linear curve fitting with Co2+ (CoO) and Co+2,+3
(Co3O4) references and it resulted in the presence of > 99 % Co3O4. The fitting data for the XANES
spectra, obtained from Athena software were provided in Table A1 in Appendix A. This
corroborates with XRD results showing the prevalence of the Co (II, III) phase in all catalysts.
3.2. Fischer-Tropsch synthesis
The promoted cobalt catalysts supported on the mesoporous alumina were tested in commercial
condition of 230 °C and 400 Psi with H2/CO ratio of 2. To compare the CO conversion activity
(%) and productivity (gHC.g−1cat.h−1) of all promoted cobalt catalysts, the FT reaction was
performed at the same gas hourly space velocity (GHSV) of 900 h−1. The operating conditions
used in this research were based on the optimization study published elsewhere [35]. Moreover,
the selectivity of all catalysts for main products (CO2, CH4, C2−C4, C5+) was evaluated at similar
CO conversion of ~ 50 % by adjusting the syngas inflow. As shown in Table 3, the catalysts
16

performance data was acquired for 100−140 h of time-on-stream (TOS) and reaching the stable
condition.
Promotional effect of noble metals (NMs) on the CO conversion (%) and mass−based activity of
mesoporous alumina supported cobalt catalyst in FTS is substantial. Without promotion, the
Co/mAl2O3 catalyst converted ~ 60% of inlet carbon monoxide to CO2 and hydrocarbons, whereas
addition of NMs resulted in 20−26 absolute % increase in the converted CO and reached to 80.2,
81.6, 84.1, and 86.3 % for Ir, Pt, Re, and Ru, respectively. It is generally accepted that promoting
with NMs (i) improves the extent of reduction, and (ii) enhances the active metal dispersion [4].
The first two factors lead to higher active site density and the latter may contribute to boost the
intrinsic activity of the active sites [4]. Higher accessible cobalt atoms for the adsorbed reactants,
and rapid reactions of adsorbates on the surface sites consequently convert more of CO+H2
reactants to products. It is noteworthy that not all ensemble of cobalt atoms could be an active site
for the FT reaction, nonetheless, active site density is proportional to the number of surface atoms
[64].
Considerable improvements in CoOx reducibility and cobalt dispersion were evidenced in this
study for the promoted catalysts with NMs. Even though, iridium resulted in higher reducibility
among the NMs (~ 87%), ruthenium with highest reduced cobalt dispersion (9%) exhibited the
highest CO conversion (86.3 %) and hydrocarbon yield, 0.258 gHC.g−1cat.h−1. Obviously, the order
of activity in the catalysts follows the fraction of the available surface cobalt (Co0 dispersion).
Nevertheless, in the case of unpromoted and Ir−promoted Co/mAl2O3 catalysts with similar cobalt
dispersion (~ 7.6 %), it is the greater extent of reduction in CoIr/mAl2O3 (~ 87% vs. ~ 61 %) which
might have contributed to its higher CO conversion activity (~ 80% vs. ~ 60%). The catalytic
surface has a dynamic nature where active sites undergo reconstruction during the CO+H2
adsorption, reaction, and intermediates’ desorption [4]. The promoters interact with support
(structural) and cobalt (electronical) which may benefit maintaining the initial reducibility, active
sites density, or improving their resistance to deactivation and sintering during the surface atomic
rearrangement [4]. It is hard to distinguish the promoters precise impact on the electronic and
structural properties and consequently the catalyst performance. However, more in-depth
spectroscopic and in−situ characterization techniques are required to understand dynamic change
on the composition and morphology of the surface [11,65].
17

Addition of manganese and yttrium, in 1 % molar ratio concentration with respect to cobalt, on
mesoporous alumina supported cobalt catalyst raised the reducibility by ~ 10 and 14%,
respectively, while the cobalt dispersion was dropped to some extend (from 7.7 to 7.2%). This
trade−off between reducibility increase and dispersion decrease resulted in rather unchanged CO
conversion in FTS (from 60.2 to 61.1 %). The marginal improvement in performance of
Mn−promoted cobalt catalyst in FTS at lower loading of promoter (Mn/Co = 0.01) has already
been reported [66]. Regarding yttrium promotional effect in other research work, 1 wt% Y was
added to 20 wt% cobalt catalyst supported on the θ−Al2O3. TPR, H2 chemisorption, and XRD
characterization showed insignificant changes in the physico-chemical properties, along with
negligible variation in FTS activity and selectivity of the catalyst [18]. It is noteworthy that the
extent to which the promoter can modify the cobalt catalytic features and performance depends on
promoter loading, Co loading. loading sequence, support nature, catalyst preparation method,
calcination and reduction treatments [11].
The addition of promoter metal can contribute to cobalt performance via increasing the number of
available cobalt sites (textural/structural), enhancing the specific rate of FTS per surface site
(electronic/chemical), or altering the reactants concentration at vicinity of active sites and affecting
the product selectivity (synergistic) [11,13]. The estimated cobalt site-time-yield (STY) for the
promoted catalysts were in the range of 2.1 ×10−2 – 4.0 ×10−2 s−1 as compared to 1.7×10−2 s−1 for
unpromoted cobalt catalyst. These values for turn over frequency (TOF) are in good agreement
with reported TOF data for alumina supported cobalt catalysts in other works [4,64]. It has been
rationalized that the small increase in the magnitude of the specific rate of FTS per surface site
could be assigned to maintaining the active cobalt sites free from deactivating carbon and oxygen
atoms during the FTS reaction [11,57]. The slight changes in the site−based activity of promoted
cobalt catalyst in this work would be still consistent with structural−insensitivity assumption of
cobalt crystallite size above 6−10 nm in FT reaction [13,28,64,67].
The selectivity of promoted catalysts for FTs products is presented in Table 3. The Ru-promoted
cobalt catalyst exhibited the highest C5+ selectivity of 84.2% as compared to 81.7 % in unpromoted
Co/mAl2O3 counterpart at the similar CO conversion level of ~ 50%. Ruthenium itself is very
active catalyst for long chain hydrocarbons (C5+) as compared to other NMs (Re, Pt, Ir) which are
not active for CO dissociation [34]. Rhenium also exhibited beneficial impact on C5+ selectivity
18

by increasing it to 83.1 %, whereas, Pt and Ir adversely acted. Platinum−promoted cobalt catalyst
slightly decreased the C5+ selectivity to 80.7, while the iridium lowered it to 78.5%. The
promotional effect of manganese on long chain hydrocarbon was marginally positive, while, the
yttrium dropped it to 80.1%. As shown in Table 3, the CoRu/mAl2O3 catalyst with highest
selectivity toward long−chain hydrocarbons showed the lowest selectivity (12.5%) to methane
product, conversely the CoIr/mAl2O3 was the most selective catalyst for CH4 (17%) while being
the least selective to C5+ cuts. This signifies that the methanation and chain growth compete on the
common C1 monomer thus, less selectivity to heavier hydrocarbons indicates higher chain
termination probability on the corresponding catalysts [55].
Utilization of microkinetics and in-situ X-ray adsorption techniques such as EXAFS would be
crucial to understand underlying reasons for the activity and selectivity of differently promoted
cobalt catalyst during the FT reaction. As a result, associative/dissociative H2 and CO adsorption
(C1 monomer formation), oxygen desorption (H2O and CO2 formation), monomers coupling (chain
growth), and hydrogen addition / abstraction to / from CxHy intermediates on differently
coordinated active sites could be understood [13,14]. The higher activity and productivity of Ru
can be assigned to its ability to decelerate the cobalt active site deactivation by inhibiting carbon
deposition, slowing down formation of non-active cobalt aluminate, and facilitating oxygen
removal from the surface [9,58]. Tsubaki et al. speculated that the higher activity of the Ru-added
Co catalyst is attributed to the bridge-adsorbed CO which is more active (easily dissociate) than
linear−adsorbed CO [30]. It can be observed that the enhancement of CO conversion in the
promoted catalysts (Figure A1 in Appendix A.) can be mainly correlated with the increase in Co
dispersion. The increase in cobalt site atoms on the surface not only contributes to enhancement
of the catalyst activity, it also shifts the selectivity toward the higher C5+ HCs. As a result, CH4
selectivity decreases due to increased probability in α-olefin readsorption [58].
Source of CO2 product could be the result of water-gas-shift (WGS) (CO + H2O → CO2 + H2)
reaction and Boudouard reaction (C + CO → CO2). Platinum- and iridium-promoted cobalt
catalysts increased the CO2 selectivity from 2.4 to 3.8 and 3.3 %, respectively (at ~ 50% CO
conversion), whereas the Ru-added catalyst decreased it to 1.5%. The other promoters did not have
significant impact on CO2 selectivity.
19

Deactivation pattern of the promoted catalysts was also explored to evaluate the effects of different
promoters used at the same atomic ratio on the stability of the Co/mAl2O3 catalysts in FT. The
time-on-stream data for CO conversion (%) and selectivities of the promoted catalysts were
provided in Table A2 in Appendix A. Comparison of different promoters in terms of stability
revealed that the Pt- and Ir-promoted catalysts lost the highest average % of the CO conversion by
~ 0.13 and ~ 0.1 % , respectively, in the last 50 h of the time-on-stream. Re, Ru, Mn, and Y lost
ℎ
approximately by 0.06, 0.08, 0.06, and 0.05 % , respectively. This evidenced that the transition
ℎ
metals, Re, and Ru promoted catalysts exhibit more stability compared to Ir and Pt promoted cobalt
catalysts supported on the mesoporous alumina. It is speculated that with reduction facilitating
promoters, more number of small cobalt oxides, strongly interacting with support, are reduced
during the catalyst activation. However, these small reduced cobalt crystallites are highly prone to
reoxidation in presence of water during the FT reaction [68]. The findings of this work confirm
the different deactivation hindrance of the promoter elements on the chosen level of concentrations
and operation conditions. Re from noble metals and the used transition metals exhibit better
stability as compared to the counterpart promoters in this study.
In summary, during the promoter screening for the cobalt catalyst all the advantages and
disadvantages need be addressed separately for individual elements. For instance, platinum is the
more naturally available, highly reducing, moderately active, selective and stable, but very
expensive option among the noble metals for FTS Co catalyst, whereas rhenium is naturally scarce
element, highly C5+ selective, productive, and stable, with quite reasonable price among the noble
metals. Ruthenium is moderately high price and very scarce element worldwide, however,
according to the results of this work, ruthenium is the most active and C5+ selective, moderately
stable, highly reducing and dispersing promoter for the cobalt catalyst.
4. Conclusions
Addition of Ru, Re, Pt, Ir noble metals (NMs) in mesoporous alumina supported cobalt catalysts
NM
1 atom
(
=
100 atom) significantly increased the cobalt dispersion, and degree of reduction. The impact
Co
of NMs on the cobalt crystallite size was not substantial. This was confirmed by calculation of Co0
size from H2 chemisorption and XRD analyses. Addition of Mn and Y transition metals in the
same atomic ratio (1 %) resulted in mild increase in the Co apparent dispersion and extent of
20

reduction. Prevalent Co-oxide species in all oxidized catalysts was Co3O4 which is confirmed by
XRD and XANES analyses. After testing all catalysts in the fixed-bed Fischer-Tropsch reactor,
the NM-promoted Co/mAl2O3 catalysts unveiled a noticeable raise (absolute percentage of ~
20−27 %) in CO conversion and hydrocarbon productivity. The selectivity of desired C5+ products
increased as result of Ru and Re addition to cobalt catalyst from 81.7 % to 84.2 and 83.1 %,
respectively. In addition, the selectivity of undesired CH4 product dropped from 15.4% for
unpromoted cobalt catalyst to 12.5 and 13.8 %, respectively, for CoRu/mAl2O3 and CoRe/mAl2O3
catalysts in FTS. The Pt and Mn exhibited slightly positive impact on the C5+ and CH4 selectivity,
whereas Ir and Y adversely affected the product selectivities. Ru was the most effective option
among the promoters added to mAl2O3 supported Co catalyst for enhancing the activity and C5+
selectivity, while, the employed transition metals exhibited to be more efficient promoters to slow
down the catalyst deactivation. Moreover, noble metals could affect the stability of cobalt catalysts
in considerably different manner. In this study, Re-promoted catalyst disclosed rather low loss in
CO conversion (~ 0.06%/h), comparable to Mn- and Y-promoted Co catalysts in FTS.
. Acknowledgment
Authors are thankful to Natural Sciences and Engineering Research Council, BioFuelNet Canada,
and University of Saskatchewan for their financial support, Canadian Light Source for providing
facility to perform the XANES experiments, Mohsen Shakouri, PhD student in Chemical
Engineering Department at University of Saskatchewan, due to his great assistance in compiling
the XANES data. Dr. Sandeep Badoga is highly appreciated due to his valuable advice and input
during the research work and manuscript writeup.
. Appendix A. Supplementary data
Supplementary data associated with this article can be found in the online version of this paper.
21

References
[1] P.M. Maitlis, A. de Klerk, in:, Greener Fischer-Tropsch Process Fuels Feed, Wiley-VCH
Verlag & Co. KgaA, 2013, pp. 1–15.
[2] A. Steynberg, M.E. Dry, in:, Fischer-Tropsch Technol, Elsevier Science & Technology
Books, 2004, pp. 1–63.
[3] A. de Klerk, in:, Fischer-Tropsch Refin, Wiley-VCH Verlag & Co. KgaA, 2011, pp. 1–19.
[4] C.H. Bartholomew, R.J. Farrauto, Hydrogen Production and Synthesis Gas Reactions,
Wiley Online Library, 2006.
[5] A.P. Vogel, B. Van Dyk, A.M. Saib, Catal Today 259 (2015) 323–330.
[6] B.H. Davis, in:, Ind Eng Chem Res, 2007, pp. 8938–8945.
[7] R. Luque, A.R. de la Osa, J.M. Campelo, A.A. Romero, J.L. Valverde, P. Sanchez, Energy
Environ Sci 5 (2012) 5186.
[8] E.F. Sousa-Aguiar, F.B. Noronha, J.A. Faro, Catal Sci Technol 1 (2011) 698–713.
[9] A.Y. Khodakov, W. Chu, P. Fongarland, Chem Rev 107 (2007) 1692–1744.
[10] B. Cornils, W.A. Herrmann, C. Wong, H. Zanthoff, R. Eds, Catalysis from A to Z, 2007.
[11] B.Y.F. Morales, B.M. Weckhuysen, in:, J. J. Spivey and K. M. Dooley (Ed.), Catalysis,
RSC Publishing, 2006, pp. 1–40.
[12] B.E. Koel, J. Kim, in:, Handb Heterog Catal, 2nd Ed., Wiley-VCH Verlag GmbH & Co.
KGaA., 2008, pp. 1593–1624.
[13] E. Iglesia, Appl Catal A Gen 161 (1997) 59–78.
[14] C.J. Weststrate, P. van Helden, J.W. Niemantsverdriet, Catal Today 275 (2016) 100–110.
[15] I. Chorkendorff, J.W. Niemantsverdriet, Concepts of Modern Catalysis and Kinetics, 2003.
[16] E. Rytter, N.E. Tsakoumis, A. Holmen, Catal Today 261 (2016) 3–16.
[17] T. Jermwongratanachai, G. Jacobs, W.D. Shafer, V.R.R. Pendyala, W. Ma, M.K.
Gnanamani, S. Hopps, G.A. Thomas, B. Kitiyanan, S. Khalid, B.H. Davis, Catal Today 228
(2014) 15–21.
[18] K. Shimura, T. Miyazawa, T. Hanaoka, S. Hirata, Appl Catal A Gen 494 (2015) 1–11.
[19] O.L. Eliseev, M. V. Tsapkina, O.S. Dement’eva, P.E. Davydov, a. V. Kazakov, a. L.
Lapidus, Kinet Catal 54 (2013) 207–212.
[20] H. Ming, B.G. Baker, M. Jasieniak, Appl Catal A Gen 381 (2010) 216–225.
[21] J.H. Den Otter, S.R. Nijveld, K.P. De Jong, ACS Catal 6 (2016) 1616–1623.
22

[22] J. Cheng, P. Hu, P. Ellis, S. French, G. Kelly, C.M. Lok, J Catal 257 (2008) 221–228.
[23] S.W.T. Price, D.J. Martin, A.D. Parsons, W.A. Sławiński, A. Vamvakeros, S.J. Keylock,
A.M. Beale, J.F.W. Mosselmans, Sci Adv 3 (2017).
[24] J.L. Casci, C.M. Lok, M.D. Shannon, Catal Today 145 (2009) 38–44.
[25] R. Oukaci, A.H. Singleton, J.G. Goodwin, Appl Catal A Gen 186 (1999) 129–144.
[26] E. Rytter, A. Holmen, Catalysts 5 (2015) 478–499.
[27] F. Diehl and A. Y. Khodakov, Oil Gas Sci Technol 64 (2009) 11–24.
[28] J. van de Loosdrecht, F.G. Botes, I.M. Ciobica, A. Ferreira, P. Gibson, D.J. Moodley, A.M.
Saib, J.L. Visagie, C.J. Weststrate, J.W. Niemantsverdriet, in:, Compr Inorg Chem II, 2013,
pp. 525–557.
[29] J. Hong, P.A. Chernavskii, A.Y. Khodakov, W. Chu, Catal Today 140 (2009) 135–141.
[30] N. Tsubaki, S. Sun, K. Fujimoto, J Catal 199 (2001) 236–246.
[31] T.O. Eschemann, J. Oenema, K.P. De Jong, Catal Today 261 (2016) 60–66.
[32] G. Jacobs, T.K. Das, Y. Zhang, J. Li, G. Racoillet, B.H. Davis, Appl Catal A Gen 233 (2002)
263–281.
[33] K.M. Cook, H.D. Perez, C.H. Bartholomew, W.C. Hecker, Appl Catal A Gen 482 (2014)
275–286.
[34] W. Ma, G. Jacobs, R.A. Keogh, D.B. Bukur, B.H. Davis, Appl Catal A Gen 437 (2012) 1–
9.
[35] V. Vosoughi, S. Badoga, A.K. Dalai, N. Abatzoglou, Fuel Process Technol 162 (2017) 55–
65.
[36] C. Márquez-Alvarez, N. Žilková, J. Pérez-Pariente, J. Čejka, Catal Rev 50 (2008) 222–286.
[37] Z. Shan, J.C. Jansen, W. Zhou, T. Maschmeyer, Appl Catal A Gen 254 (2003) 339–343.
[38] C. Lesaint, W.R. Glomm, yvind Borg, S. Eri, E. Rytter, G. ye, Appl Catal A Gen 351 (2008)
131–135.
[39] E. van Steen, M. Claeys, Chem Eng Technol 31 (2008) 655–666.
[40] S. Badoga, K.C. Mouli, K.K. Soni, A.K. Dalai, J. Adjaye, Appl Catal B Environ 125 (2012)
67–84.
[41] V. Vosoughi, S. Badoga, A.K. Dalai, N. Abatzoglou, Ind Eng Chem Res 55 (2016) 6049–
6059.
[42] S. Rane, Ø. Borg, J. Yang, E. Rytter, A. Holmen, Appl Catal A Gen 388 (2010) 160–167.
23

[43] D. Schanke, J Catal 156 (1995) 85–95.
[44] R.M.M. Abbaslou, A. Tavassoli, J. Soltan, A.K. Dalai, Appl Catal A Gen 367 (2009) 47–
52.
[45] M. Trépanier, A.K. Dalai, N. Abatzoglou, Appl Catal A Gen 374 (2010) 79–86.
[46] G. Słowik, A. Gawryszuk-Rżysko, M. Greluk, A. Machocki, Catal Letters 146 (2016)
2173–2184.
[47] T.E.W. Jr., Catal Rev 8 (1974) 117–134.
[48] F. Pinna, Catal Today 41 (1998) 129–137.
[49] M.D. Shannon, C.M. Lok, J.L. Casci, J Catal 249 (2007) 41–51.
[50] F. Diehl, A.Y. Khodakov, Oil Gas Sci Technol 64 (2009) 11–24.
[51] W. Chu, P.A. Chernavskii, L. Gengembre, G.A. Pankina, P. Fongarland, A.Y. Khodakov, J
Catal 252 (2007) 215–230.
[52] A.M. Saib, A. Borgna, J. van de Loosdrecht, P.J. van Berge, J.W. Geus, J.W.
Niemantsverdriet, J Catal 239 (2006) 326–339.
[53] T.K. Das, G. Jacobs, P.M. Patterson, W.A. Conner, J. Li, B.H. Davis, Fuel 82 (2003) 805–
815.
[54] H.-Y. Lin, Y.-W. Chen, Mater Chem Phys 85 (2004) 171–175.
[55] E. Rytter, N.E. Tsakoumis, A. Holmen, Catal Today 261 (2016) 3–16.
[56] V. V Rozanov, O. V Krylov, Russ Chem Rev 66 (2007) 107–119.
[57] E. Iglesia, S.L. Soled, R.A. Fiato, G.H. Via, J Catal 143 (1993) 345–368.
[58] E. Iglesia, Appl Catal A Gen 161 (1997) 59–78.
[59] B. Hammer, J.K. Norskov, Adv Catal 45 (2000) 71–129.
[60] J.K. Nørskov, F. Abild-Pedersen, F. Studt, T. Bligaard, Proc Natl Acad Sci U S A 108
(2011) 937–43.
[61] H. Romar, A.H. Lillebø, P. Tynjälä, T. Hu, A. Holmen, E.A. Blekkan, U. Lassi, Top Catal
58 (2015) 887–895.
[62] R.-S. Zhou, R.L. Snyder, Acta Crystallogr Sect B 47 (1991) 617–630.
[63] D. Merino, I. Pérez-Miqueo, O. Sanz, M. Montes, Top Catal (2015) 1–12.
[64] F.H. Ribeiro, A.E.S. Von Wittenau, C.H. Bartholomew, G. Somorjai, A, Catal Rev 39
(1997) 49–76.
[65] G. Jacobs, W. Ma, P. Gao, B. Todic, T. Bhatelia, D.B. Bukur, B.H. Davis, Catal Today 214
24

(2013) 100–139.
[66] G.R. Johnson, S. Werner, A.T. Bell, ACS Catal 5 (2015) 5888–5903.
[67] G.L. Bezemer, J.H. Bitter, H.P.C.E. Kuipers, H. Oosterbeek, J.E. Holewijn, X. Xu, F.
Kapteijn, A.J. Van Diilen, K.P. De Jong, J Am Chem Soc 128 (2006) 3956–3964.
[68] N.E. Tsakoumis, M. Rønning, Ø. Borg, E. Rytter, A. Holmen, Catal Today 154 (2010) 162–
182.
25

Figure 1. Temperature−programmed profile for mesoporous alumina supported Co catalysts with different
noble and transient metal promoters
26

Figure 2. HRTEM images of (1) Co/mAl2O3, (2) CoRu/mAl2O3, (3) CoRe/mAl2O3,
(4)
CoPt/mAl2O3, (5) CoMn/mAl2O3.
27

CoY/mAl2O3
CoMn/mAl2O3
CoIr/mAl2O3
CoPt/mAl2O3
CoRe/mAl2O3
CoRu/mAl2O3
15Co/mAl2O3
CoO
Co3O4
10
20
30
40
50
2θ
60
70
80
90
Figure 3. XRD patterns for unprompted and promoted cobalt catalyst supported on the mesoporous alumina
28

CoO_Ref
Co3O4_Ref
Co/mAl2O3
1
0.5
0
CoRu/mAl2O3
CoRe/mAl2O3
CoPt/mAl2O3
CoIr/mAl2O3
CoMn/mAl2O3
CoY/mAl2O3
7690
7710
7730
7750
7770
7790
Photon Energy (eV)
Figure 4. XANES spectra for Co0 (dotted line), CoO (small−dashed), Co3O4 (large−dashed), unprompted
and promoted cobalt catalysts (solid lines) supported on the mesoporous alumina
29