See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/273579752

Light olefin synthesis from CO2 hydrogenation

over K-promoted Fe–Co bimetallic catalysts

Article in Catalysis Today · February 2015

Impact Factor: 3.89 · DOI: 10.1016/j.cattod.2015.01.011

CITATIONS READS

2 80

4 authors, including:

Naoto Koizumi Chunshan Song

Pennsylvania State University Pennsylvania State University
64 PUBLICATIONS 735 CITATIONS 480 PUBLICATIONS 12,366 CITATIONS

SEE PROFILE SEE PROFILE

Pattarapan Prasassarakich
Chulalongkorn University
94 PUBLICATIONS 1,121 CITATIONS

SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate, Available from: Chunshan Song
letting you access and read them immediately. Retrieved on: 18 April 2016
 Catalysis Today 251 (2015) 34–40

Contents lists available at ScienceDirect

Catalysis Today
journal homepage: www.elsevier.com/locate/cattod

Light olefin synthesis from CO2 hydrogenation over K-promoted

Fe–Co bimetallic catalysts
Ratchprapa Satthawong a,b , Naoto Koizumi a , Chunshan Song a,c,∗ ,
Pattarapan Prasassarakich b
a
Clean Fuels & Catalysis Program, EMS Energy Institute, and Department of Energy & Mineral Engineering, The Pennsylvania State University,
209 Academic Projects Building, University Park, PA 16802, USA
b
Department of Chemical Technology, Faculty of Science, Chulalongkorn University, Phayathai, Bangkok 10330, Thailand
c
PSU-DUT Joint Center for Energy Research, The Pennsylvania State University, University Park, PA 16802, USA

a r t i c l e i n f o a b s t r a c t

Article history: This work investigates the effects of potassium addition on light olefin production from CO2 hydro-
Received 5 August 2014 genation over the Al2 O3 -supported Fe–Co bimetallic catalysts. The addition of small amount of K on the
Received in revised form Fe–Co catalyst significantly enhanced C2 –C4 olefin formation. The C2 –C4 olefins predominated among
31 December 2014
C2 + hydrocarbons at K/Fe atomic ratio of 1. Temperature-programmed desorption experiments demon-
Accepted 4 January 2015
strated that K addition diminished the weakly adsorbed hydrogen but enhanced CO2 adsorption on the
Available online 14 February 2015
catalyst surface, which led to significantly enhanced production of light olefins. For C2 –C4 olefin synthe-
sis at 573 K and 1.0 MPa, the K-promoted Fe–Co/Al2 O3 catalyst also showed better performance than the
Keywords:
CO2 hydrogenation
K-promoted Fe–Mn/Al2 O3 catalyst which is known as a good catalyst for olefin production, under the
Light olefins present conditions employed. The pathway of olefin formation from CO2 hydrogenation was analyzed.
K-promoter Light olefins were formed via two-step reactions over K-promoted Fe–Co catalysts with the formation
Fe–Co bimetallic catalyst of CO or CO-like intermediate as the first step. Some of these olefins could be further hydrogenated to
Higher hydrocarbons paraffins, depending on the type and concentration of chemisorbed hydrogen on the catalyst surface.
Adsorption properties © 2015 Elsevier B.V. All rights reserved.

1. Introduction catalysts used for CO2 hydrogenation to hydrocarbons are based on

Carbon dioxide (CO2 ) utilization has recently become attrac-
tive due to the increasing concern on the effects of greenhouse
gas CO2 accumulation in the atmosphere, and its potential as a
source of carbon for synthesizing chemicals, materials and fuels.
The catalytic CO2 hydrogenation using hydrogen produced with
renewable energy source is of particular interest, since it can con-
vert CO2 to hydrocarbons and alcohols. Light olefins (C2 –C4 ) are
particularly important building blocks for chemicals and polymers,
thus the development of efficient catalysts for CO2 hydrogenation
to olefins is highly desirable, which could reduce both the current
CO2 emission levels and the dependence of chemical production on
the depleting fossil fuels [1–7].
CO2 hydrogenation to hydrocarbons could be viewed as an
alternative to the conventional Fischer–Tropsch synthesis (FTS),
where CO2 is used as a carbon source instead of CO. Until now, the
∗ Corresponding author at: Clean Fuels & Catalysis Program, EMS Energy Institute,
and Department of Energy & Mineral Engineering, Pennsylvania State University,
209 Academic Projects Building, University Park, PA 16802, USA.
E-mail address: csong@psu.edu (C. Song).
those that are also used for FTS such as Fe, Co, Ni and Ru catalysts
[8]. However, when supported Co (or Ni or Ru) catalysts are used
for CO2 hydrogenation, CH4 is the dominant product while little or
no C2 + hydrocarbons are produced; whereas higher hydrocarbons
could be synthesized over Fe catalyst [9–14]. Therefore, the atten-
tions have been paid on improving the catalytic performances of
Fe-based catalysts by adding some promoters such as Li, K and Mn
or selecting the proper support materials [15–21].
Potassium has been regarded as an effective promoter for CO and
CO2 hydrogenation to higher hydrocarbons over Fe-based catalysts
[14,19,20,22], and the physicochemical properties of K-promoted
Fe-based catalysts have been studied in relation to the hydrogena-
tion activity. These studies revealed that K addition enhanced CO2
adsorption, while H2 adsorption on Fe catalysts was suppressed,
resulting in the decrease in CH4 formation and increase in olefin
content in the products [16,23–28]. Many researchers also pro-
posed that K could facilitate the formation of iron carbide, which
is considered as the active phase for CO and CO2 hydrogenation to
higher hydrocarbons [23,27,29].
Recently, a significant promotion of C2 + hydrocarbons forma-
tion from CO2 hydrogenation over Fe–Co/Al2 O3 bimetallic catalysts
with a low Co content at Co/(Co + Fe) atomic ratio of 0.17 was

http://dx.doi.org/10.1016/j.cattod.2015.01.011
0920-5861/© 2015 Elsevier B.V. All rights reserved.
 R. Satthawong et al. / Catalysis Today 251 (2015) 34–40 35

reported from our laboratory [22]. The temperature-programmed

desorption coupled with experimental investigation of the cata-
lysts responsible for CO2 and H2 activation and carbon carbon
bond formation on this Fe–Co bimetallic system revealed that mod-
erately adsorbed hydrogen species played an important role in
C2 + hydrocarbons formation. The addition of K-promoter to these
bimetallic catalysts led to more effective synthesis of olefin-rich
long chain hydrocarbons.
The K promoted Fe–Co catalyst would have high activity for light
olefin formation from CO2 hydrogenation. However, the studies on
the effects of the K-promoter have mainly dealt with monometal-
lic Fe catalysts [16,23,27,29], while the effect of K loading on C2 –C4
olefin formation, and the pathways for these olefins formation on
Fe–Co bimetallic catalysts have not been reported. Furthermore,
the combination of Fe and Co led to unique adsorption properties
for H2 and CO2 . It is of great interest to investigate how the K addi-
tion affects the H2 and CO2 adsorption properties and relations with
CO2 hydrogenation activities of these catalysts. Fig. 1. Effects of space time W/F on CO2 conversion over Fe–Co(X)/K(0.3)/Al2 O3
In this work, the K-promoted Fe, Co and Fe–Co bimetallic catalysts.

catalysts (Co/(Co + Fe) = 0.17 atom-Co atom−1 ) with different K/Fe

atomic ratios (K/Co ratio for the catalyst with Co alone) were (ca. 0.20–0.25 g) were outgassed at 363 K for 1 h and 473 K for 12 h
tested for CO2 hydrogenation and characterized by temperature- under flowing N2 before analysis.
programmed desorption (TPD) to elucidate the influence of
K-promoter on adsorption properties of H2 and CO2 on the Fe–Co 2.4. Temperature-programmed desorption (TPD) of H2 and CO2
bimetallic catalysts, and their impact on the olefin formation from
CO2 hydrogenation. The adsorption properties of H2 and CO2 on the pre-reduced cat-
alyst were evaluated by TPD using a Micromeritics AutoChem 2910
Characterization System. Before starting the TPD analysis, approx-
2. Experimental
imately 150 mg of catalyst was pre-reduced in situ under a H2 flow
at 673 K for 2 h. For H2 -TPD experiment, the reduced catalyst was
2.1. Catalyst preparation
cooled down to 200 K using an isopropanol–liquid N2 mixture to
prevent desorption of weakly adsorbed H2 . The reduced catalyst
K-promoted catalysts were prepared by a two-step incipient
was subsequently purged with 0.95 vol.% Ar/He at this temperature
wetness impregnation method using ␥-Al2 O3 (Sasol PURALOX TH
for 20 min, followed by the TPD analysis with heating to 1173 K at
100/150) as a support. K2 CO3 (Sigma-Aldrich, ≥ 99.0%) aque-
10 K min−1 . The temperature program started at around 280 K.
ous solution was firstly impregnated onto Al2 O3 support and the
For CO2 -TPD, after pre-reduction of the catalyst, the reactor was
sample was dried at 383 K overnight. Concentration of K in the
flushed with 0.95 vol.% Ar/He at 573 K for 30 min. The reduced cat-
impregnating solution was adjusted to obtain desired K/Fe atomic
alyst was subsequently dosed with 15 vol.% CO2 /He (30 mL min−1 )
ratios (0.0–1.0). The aqueous solution containing Fe(NO3 )3 ·9H2 O
at this temperature for 1 h, and cooled down to ambient tempera-
(Sigma-Aldrich, 99.99%) and/or Co(NO3 )2 ·6H2 O (Sigma-Aldrich,
ture. The catalyst was then flushed with 0.95 vol.% Ar/He for 20 min.
≥98%) was then co-impregnated onto the K/Al2 O3 sample. The
The TPD analysis was performed according to the procedures as
impregnated sample was subsequently dried at 383 K and calcined
that for the H2 -TPD analysis above. The effluent gas was analyzed
at 673 K in air [22]. The total metal (Fe + Co) loading was kept con-
using a quadrupole mass spectrometer (DM200M, Ametek Dycor
stant at 15 wt% (support weight basis) with the Co/(Co + Fe) atomic
Dymaxion) for both H2 - and CO2 -TPD analyses.
ratios of 0.0, 0.17 and 1.0. The catalysts prepared in this work
are denoted as Fe–Co(X)/K(Y)/Al2 O3 , where X and Y represent the
Co/(Co + Fe) atomic ratio and the K/Fe atomic ratio (or K/Co atomic 3. Results and discussions
ratio for catalyst with Co alone without Fe), respectively.
3.1. Pathway for olefin formation from CO2 hydrogenation

2.2. Activity test
CO2 hydrogenation over the Fe–Co(X)/K(Y)/Al2 O3 catalysts was
conducted in a high-pressure fixed-bed reactor at 573 K and
1.1 MPa, and the detailed procedure was reported elsewhere [22].
Typically, 0.2 g of the catalyst was used and the feed gas flow rate
was 12 mL(STP) min−1 (W/F of 1.01 g s mL−1 ). To examine the path-
way for C2 –C4 olefin formation from CO2 hydrogenation, W/F was
varied from 0.01 to 17.5 g s mL−1 by changing the catalyst weight
and feed gas flow rate.
2.3. Measurement of the physical properties
The surface area (SA), pore volume (PV) and average
pore diameter (Dp ) were measured using TriStar II analyzer
(Micromeritics). The properties were determined from the N2
adsorption/desorption isotherms measured at 77 K. All the samples
To investigate the pathway for olefin formation from CO2 hydro-
genation, the catalysts were tested over a wide range of W/F at
573 K and 1.1 MPa. Conversions and product selectivities at about
15–16 h on stream were used for evaluating the activity and selec-
tivity of the catalysts. Stable CO2 conversion and selectivity could
be obtained after 3–4 h on stream for all catalysts. Effects of space
time W/F (the ratio of catalyst mass to the feed flow rate) on CO2
conversion of Fe–Co(X)/K(0.3)/Al2 O3 catalyst are shown in Fig. 1.
The CO2 conversion increased with increasing W/F and gradually
reached constant values at high W/F for all catalysts. It is noted that
the equilibrium CO2 conversion at this reaction condition is around
70%. K-promoted Fe/Al2 O3 catalyst provided lowest CO2 hydro-
genation activity, while Co/K(0.3)/Al2 O3 catalyst exhibited highest
activity and the equilibrium conversion was reached at around W/F
of 6 g s mL−1 . The results show that the addition of a small amount of
Co to Fe catalyst at the atomic ratio of Co/(Co + Fe) = 0.17 increased
the CO2 hydrogenation activity.
36 R. Satthawong et al. / Catalysis Today 251 (2015) 34–40
Fig. 2. GC-FID chromatograms of gas-phase hydrocarbons from CO2 hydrogena-
tion on Fe–Co(0.17)/K(Y)/Al2 O3 catalysts with different K/Fe atomic ratio at W/F of
1.01 g s mL−1 .
Fig. 2 illustrates GC-FID chromatograms of gas-phase hydro-
carbons with Fe–Co(0.17)/Al2 O3 catalysts with different K loading
levels at W/F of 1.01 g s mL−1 . Linear paraffins were dominant prod-
ucts with Fe–Co(0.17)/K(0.3)/Al2 O3 , but the production of light
olefins was observed along with paraffinic hydrocarbon prod-
ucts (middle chromatogram). Further increasing K loading to
Fe–Co(0.17)/K(1.0)/Al2 O3 dramatically enhanced olefin production
to the level where C2 –C4 olefins become dominant products in the
light hydrocarbons.
To study the pathways for these hydrocarbons, the CO and
hydrocarbon selectivity were plotted as a function of CO2 con-
version as shown in Fig. 3. For all catalysts, similar trends
were observed. The CO selectivity decreased with increasing CO2
Fig. 3. CO and hydrocarbon selectivity of Fe–Co(X)/K(0.3)/Al2 O3 catalysts as a func-
tion of CO2 conversion.
Fig. 4. Product selectivity as a function of CO2 conversion from CO2 hydrogenation
over Fe–Co(0.17)/K(0.3)/Al2 O3 catalysts.
conversion, while hydrocarbon selectivity increased sharply with
increasing conversion. These results suggest that CO or CO-like sur-
face intermediate was mainly formed as a primary product from
CO2 hydrogenation, possibly through reverse water–gas shift reac-
tion (RWGS) for CO formation, and the produced CO was further
hydrogenated to hydrocarbons (FTS) [30], even for the K-promoted
Co catalyst. The catalysts studied in this work were quite active at
573 K and 1.1 MPa; it was difficult to obtain very low conversion
data. Thus, we would not exclude the presence of a direct hydro-
genation route suggested by Riedel et al. [30], but the contribution
of such route would be small under the conditions used in the
present work. Though K-promoted Co catalyst exhibited highest
CO2 hydrogenation activity, CO selectivity decreased quickly with
increasing CO2 conversion and CH4 was the predominant hydro-
carbon product (not shown here) compared with those of Fe and
Fe–Co catalysts, due to the high hydrogenation ability of Co [22].
It is also worth noting that Fe–Co catalyst showed higher activity
than Fe catalyst (Fig. 1), but the changes of CO and hydrocarbon
selectivity with both catalysts were similar. It could be concluded
that the addition of small amount of Co increased both RWGS and
FTS activities at a similar degree, resulting in an enhancement of
CO2 conversion without changing the product selectivity.
The light olefin and paraffin selectivity together with CO selec-
tivity of Fe–Co(0.17)/K(0.3)/Al2 O3 catalyst also plotted as a function
of CO2 conversion and shown in Fig. 4. Both C2 –C4 olefins and
paraffins were observed at the CO2 conversion of 14%. The paraf-
fin selectivity increased drastically with the conversion, while the
olefin selectivity increased approaching the maximum value and
then decreased. Thus, light olefins and paraffins (C2 –C4 ) would form
simultaneously from hydrogenation of the produced CO, but some
of these olefins could be further hydrogenated to paraffins, depend-
ing on the concentration of chemisorbed hydrogen on the catalyst
surface. Hydrogenation of C2 –C4 olefins was also confirmed by a
sharp decrease in the olefin–paraffin ratio with increasing the con-
version (see Fig. 4). Therefore, suppressing olefin hydrogenation
by tailoring the surface coverage of CO2 and H2 would cause the
increase in the olefin content from CO2 hydrogenation.
3.2. Light olefin synthesis from CO2 hydrogenation
To modify the surface coverage of H2 and CO2 on the catalyst
surface, the amount of potassium promoter of Fe–Co catalysts was
then varied. The activity and selectivity of Fe–Co(0.17)/K(Y)/Al2 O3
catalysts (Y = 0.0–1.0) were investigated at hydrogenation
condition of 573 K and 1.1 MPa (1.01 g s mL−1 ). The product
selectivities were plotted as a function of the K/Fe atomic ratio,
 R. Satthawong et al. / Catalysis Today 251 (2015) 34–40 37

Fig. 6. Effect of K/Fe atomic ratio (Y) on C2 –C4 olefin yield and O/P ratio from CO2
hydrogenation over Fe–Co(0.17)/K(Y)/Al2 O3 catalysts [22].

Fig. 5. Effect of K/Fe atomic ratio (Y) on the product selectivity from CO2 hydro-
genation over Fe–Co(0.17)/K(Y)/Al2 O3 catalysts [22].
as shown in Fig. 5. Without K-promoter, no olefin was observed
in the FID chromatogram of the gas-phase products as shown in
Fig. 2. Increasing K content resulted in a significant increase in
light olefins (C2 –C4 ) selectivity, at the expense of the light paraffin
selectivity. It is worth noting that CH4 selectivity drastically
decreased with increasing K content, while C5 + hydrocarbons
notably increased. Since the CO2 conversion was hardly changed
with the increasing K loading (from 0.3 to 1.0, Table 1), these
results suggest that the catalyst adsorption properties of H2 and
CO2 changed significantly with K content, as discussed in the
Sections 3.4 and 3.5.
The light olefin yield and the olefin to paraffin atomic ratio
(O/P) for the Fe–Co/K catalysts are illustrated in Fig. 6. Light olefin
yield was significantly increased with increasing K-promoter up to
0.5 atom atom−1 . Further increasing K content did not increase the
olefin yield, while the O/P ratio increased sharply approaching 5
at K/Fe = 1 atom atom−1 [22]. These results clearly show that the
K addition could suppress the olefin hydrogenation significantly,
leading to a predominance of olefins in the light hydrocarbon prod-
ucts. CO2 conversion and C2 –C4 olefin yield on the K-promoted
Fe–Co bimetallic catalyst are also compared with those on the cat-
alysts reported previously (Table 1).
A K-promoted Fe–Mn/Al2 O3 catalyst was also prepared accord-
ing to the literature [19], and tested at the present hydrogenation
condition of 573 K and 1.1 MPa. This table demonstrates that at the
same reaction condition, Fe–Co bimetallic catalyst with K/Fe of 0.5
and 1.0 exhibited higher CO2 conversion and light olefin yield than
the promising Fe-based catalysts such as Y-zeolite supported Fe–Ce
[31] and K-promoted Fe–Mn/Al2 O3 catalysts.
3.3. Physical properties of calcined catalysts
The physical properties of calcined K-promoted Fe–Co(0.17)/
Al2 O3 catalysts are provided in Table 1. The Brunauer–
Emmett–Teller (BET) surface area and pore volume of the catalysts
decreased with increasing K content, while average pore diameter
Dp barely changed with increasing K loading (from zero to 0.5).
The physical properties of the K-promoted catalysts suggest that
some portions of alumina micropores were covered by potassium
oxides, leading to the decrease in surface area and pore volume
[16,26].
3.4. H2 adsorption properties of K-promoted Fe–Co catalysts
3.4.1. H2 -TPD profiles of Fe–Co(0.17)/K(Y)/Al2 O3 catalysts
For the study on hydrogen adsorption states on the Al2 O3 -
supported K-promoted Fe–Co bimetallic catalysts by H2 -TPD (the
Al2 O3 support and K(1.0)/Al2 O3 analyzed as references), the TPD
profiles are illustrated in Fig. 7. The H2 desorption from the reduced
Fe–Co(0.17)/Al2 O3 (Y = 0) catalyst mainly occurred at 360, 565
(weakly adsorbed H2 ) and 770 K (moderately adsorbed H2 ) with

Table 1
Physical properties and CO2 hydrogenation activities and selectivities of Fe–Co(0.17)/K(Y)/Al2 O3 and the catalysts from previous works.

Catalyst SA (m2 g−1 ) PV (cm3 g−1 ) Dp (nm) CO2 conv. (%) Selectivity (C-mol%) C2 = –C4 = yield (%)
+e
CH4 C2 CO

Al2 O3 a 139 1.03 24 – – – – –

K/Al2 O3 109 0.78 23 – – – – –
Fe–Co/Al2 O3 a , b 138 0.72 18 25 44 43 13 0
Fe–Co/K(0.3)/Al2 O3 a , b 127 0.66 17 34 25 64 12 3
Fe–Co/K(0.5)/Al2 O3 b 114 0.58 17 34 19 68 13 7
Fe–Co/K(1.0)/Al2 O3 b 82 0.59 22 31 13 69 18 7
Fe17 Mn12 K8/Al2 O3 b , c NA NA NA 19 7 43 50 4
Fe/HYd NA NA NA 10 44 17 39 0
Fe/KYd NA NA NA 18 8 60 32 4
Fe–Ce/HYd NA NA NA 12 49 20 31 1
Fe–Ce/KYd NA NA NA 20 6 60 34 5
a
From Ref. [14].
b
From Ref. [22].
c
Same composition and preparation procedure as Ref. [19].
d
2.02 g s mL−1 (17 wt% metal loading) from Ref. [31].
e
Including small amount of alcohols.
38 R. Satthawong et al. / Catalysis Today 251 (2015) 34–40
Fig. 7. Effect of K/Fe atomic ratio (Y) on the H2 -TPD profiles of fresh
Fe–Co(0.17)/K(Y)/Al2 O3 catalysts.
a small desorption peak at 980 K. The weakly and moderately
adsorbed H2 on the Fe-containing catalysts were originated from
atomic hydrogen adsorbed on the metal surface, and H2 adsorbed
on the Fe0 or alloy surface in intimate contact with unreduced
metal oxide, respectively [32–34]. The H2 desorption at higher
temperature (980 K) might be originated from H2 that had spilt-
over from metal onto the support surface rather than the surface
hydroxyl groups on the ␥-Al2 O3 surface, since no H2 desorption
peak was observed from the H2 -TPD profile of ␥-Al2 O3 support.
The addition of a small amount of K (Y = 0.3) to Fe–Co catalysts
suppressed the H2 desorption especially the weakly adsorbed H2 ,
without changing the TPD profile. However, further increasing K
content (Y = 1.0) affected the TPD profile, where the peak intensity
Fig. 9. Schematic illustration of the proposed model for the adsorption states of hydrogen on the Fe–Co(0.17)/Al2 O3 catalysts (A) without K, (B) with K/Fe = 0.3 and (C)
K/Fe = 1.0 atomic ratio.
Fig. 8. Deconvolution of the H2 -TPD profiles of the reduced
Fe–Co(0.17)/K(0.3)/Al2 O3 catalyst. (◦◦◦◦◦◦◦◦) measured TPD profile, (---------)
deconvoluted peaks, ( ) summation of the deconvoluted peaks.
at 640 K increased. This peak would be ascribed to the desorption
of hydrogen on K oxides or decomposition of KAlH4 species [19],
corresponding to the weak desorption peak observed at 650 K in
the TPD profile of K(1.0)/Al2 O3 (Fig. 7).
3.4.2. Deconvolution analysis of the H2 -TPD profiles
To investigate the effects of K on the H2 adsorption proper-
ties quantitatively, the overlapped peaks in the H2 -TPD profiles of
Fe–Co(0.17)/K(Y)/Al2 O3 catalysts were deconvoluted using Gauss-
ian functions. The quantities of different types of adsorbed
hydrogen were determined by integrating these Gaussian peaks,
and presented in Table 2. Fig. 8 illustrates the deconvolution result
of the reduced Fe–Co(0.17)/K(0.3)/Al2 O3 catalyst as a represen-
tative example. The curve created from the summation of the
 R. Satthawong et al. / Catalysis Today 251 (2015) 34–40 39

Table 2
Amount of adsorbed H2 species over fresh Fe–Co(0.17)/K(Y)/Al2 O3 catalysts.

Catalyst Amount of adsorbed H2 /␮mol g-cat−1

Type I Type II Type II’ Type III Type IV Type (I + II)

Al2 O3 a 0 0 0 0 0 0
Fe–Co(0.17)/Al2 O3 b 27.2 52.4 0.0 90.3 38.5 79.6
Fe–Co(0.17)/K(0.3)/Al2 O3 b 19.4 40.9 0.0 55.9 46.1 60.3
Fe–Co(0.17)/K(1.0)/Al2 O3 b 16.9 41.8 8.6 52.8 37.0 58.7
K(1.0)/Al2 O3 b 0 0 8.7 0 0 0
a
Reduction at 673 K for 2 h.
b
Reduction at 673 K for 2 h and H2 adsorption at 573 K (5 K min−1 ).

Gaussian peaks fitted very well with the experimental data. These
Gaussian peaks are then denoted as type I–IV from low to high
temperature, representing the different adsorption strengths of
hydrogen on the catalyst surface. The addition of a small amount of
K (K/Fe = 0.3) clearly decreased the amount of type (I + II) H2 (weakly
adsorbed H2 , desorbed below 600 K) from 80 to 60 ␮mol g-cat−1 .
Further increase in the K/Fe atomic ratio from 0.3 to 1.0 caused
lower level of decrease in type I and type III hydrogen species.
The amount of type III H2 (moderately adsorbed H2 , desorbed at
700–800 K) changed in a manner similar to the type (I + II) H2 . This
is probably due to that some parts of the interface between reduced
and unreduced metal oxides (origin of type III H2 ) were also covered
by K (Fig. 9A and B).
Since type (I + II) H2 originates from hydrogen adsorbed on the
metal surface, the deconvolution results indicate the coverage of
the metal surface with K oxide species, as suggested for Fe catalysts
[23], resulting in a decrease of weakly adsorbed H2 (Fig. 9A and B).
A small contribution of hydrogen due to K oxide was observed in
the TPD profile of K(1.0)/Al2 O3 , and denoted as type II’ H2 . This
peak could be ascribed to the desorption of hydrogen on K oxides
or the decomposition of KAlH4 species [19]. The amount of type II’
hydrogen became significant at higher K loading (K/Fe = 1.0) which
is similar to that observed on K/Al2 O3 (without metal Fe or Co).
3.5. CO2 adsorption properties of Fe–Co(0.17)/K(Y)/Al2 O3
catalysts
The effect of K addition on the CO2 adsorption properties of
reduced Fe–Co(0.17)/K(Y)/Al2 O3 catalysts was studied by CO2 -TPD.
Fig. 10 shows the CO2 -TPD profiles. The ␥-Al2 O3 support was used
as a reference, where CO2 desorption mainly occurred at 350 K
(formation of bicarbonate species on a partially hydrated Al2 O3
surface [35]). For the reduced Fe–Co(0.17)/Al2 O3 catalysts (Y = 0),
CO2 desorption was observed at 350 and 530 K, respectively, which
mainly originated from bicarbonate and carbonate species formed
on the ␥-Al2 O3 surface. The amount of desorbed CO2 at 350 K
dramatically increased upon the addition of a small amount of K
(Y = 0.3). The CO2 -TPD profile significantly changed with further
increase in K content (Y = 1.0). The weakly adsorbed CO2 (des-
orbed at 350 K) was suppressed, while two desorption peaks at 630
and 880 K were clearly observed. The desorption peaks at 630 and
880 K might be originated from a bidentate carbonate species and
adsorbed CO2 on potassium ferrite (K2 Fe2 O4 ), respectively [36–38].
CO2 adsorption properties of catalysts with the K content. Accord-
ing to the H2 -TPD results (Section 3.4, Fig. 7), increasing K content of
catalyst resulted in the decrease in the amount of weakly adsorbed
H2 (type I + II), which was important for further hydrogenation of
produced olefins and CH4 formation [16,23,26–29]. From CO2 -TPD
results, the CO2 adsorption drastically increased with K content
(Section 3.5, Fig. 10), providing more chances of C C bond forma-
tion. The change in H2 and CO2 adsorption properties of theses
catalysts led to an increase in light olefin and C5 + hydrocarbons
selectivity. However, CO2 conversions with Fe–Co/K(0.3)/Al2 O3 ,
Fe–Co/K(0.5)/Al2 O3 and Fe–Co/K(1.0)/Al2 O3 catalysts are nearly the
same as summarized in Table 1. According to the H2 -TPD with the
deconvolution analysis of Fe–Co/K(Y)/Al2 O3 catalysts (Table 2), the
addition of small amount of K (Y = 0.3) significantly decreased the
amount of adsorbed H2 on the catalyst surface compared with that
on the unpromoted Fe–Co catalyst, which corresponds to higher
C2 + selectivity, higher olefin yield, and higher CO2 conversion. How-
ever, further increasing the K content (Y from 0.3 to 0.5 and 1.0) only
caused a small decrease in the amount of the adsorbed H2 species
(types 1, II, and III), corresponding to a nearly identical or a some-
what lower CO2 conversion but higher C2 + selectivity and higher
olefin yield among the K-promoted catalysts. Furthermore, in con-
trast to the H2 adsorption, CO2 adsorption significantly increased
with the K content.
CO selectivity slightly increased with K/Fe atomic ratio because
K could promote the RWGS activity of the catalysts by providing
the additional active sites to generate the formate species (inter-
mediate species of RWGS reaction), which was verified by DRIFT
analysis (not shown here) [24].

3.6. Effects of H2 and CO2 adsorption properties on CO2

hydrogenation activity

As discussed in Section 3.2, the K addition resulted in a sig-

nificant increase in light olefin (C2 –C4 ) selectivity, at the expense
of light paraffin selectivity. Moreover, CH4 selectivity drasti-
cally decreased with K content, while C5 + hydrocarbon notably Fig. 10. Effect of K/Fe atomic ratio (Y) on the CO2 -TPD profiles of fresh
increased. These results could be explained by change of H2 and Fe–Co(0.17)/K(Y)/Al2 O3 catalysts.
40 R. Satthawong et al. / Catalysis Today 251 (2015) 34–40
4. Conclusion
The Fe–Co(0.17)/K(Y)/Al2 O3 catalysts with various potassium
content (Y = 0.0–1.0) were employed for the CO2 hydrogenation.
Fe–Co was an effective catalyst for synthesizing light olefins when
it was highly alkalized (Y = 1.0). Deconvoluted H2 -TPD profiles of
the K-promoted catalysts reveal that the weakly adsorbed hydro-
gen on the metal surface is diminished by adding K, indicating
the partial coverage of the metal surface by K oxide species,
which suppressed hydrogenation of the produced olefins, resulting
in the high olefin content in the hydrocarbon product. Increas-
ing K/Fe ratio decreased the amount of weakly adsorbed H2 and
increased CO2 adsorption resulting in a decrease in the hydrogena-
tion of produced olefins. Moreover, at the same reaction condition,
Fe–Co(0.17)/K(1.0)/Al2 O3 catalyst also showed higher CO2 conver-
sion and higher yield of light olefins than the K-promoted Fe–Ce and
Fe–Mn/Al2 O3 which are known as good catalysts for hydrocarbon
production from CO2 hydrogenation.
The pathway for CO2 hydrogenation to olefins over K-promoted
Fe–Co catalysts was analyzed. Light olefins were formed via two-
step reactions over K-promoted Fe–Co catalysts with the formation
of CO or CO-like intermediate as the first step in CO2 hydrogenation.
Some of these olefins could be further hydrogenated to paraffins,
depending on the type and concentration of chemisorbed hydrogen
on the catalyst surface.
Acknowledgments
This research was supported in part by the Pennsylvania State
University through Penn State Institutes of Energy and the Envi-
ronment. The authors also wish to thank the Thailand Research
Fund and Graduate School of Chulalongkorn University through the
Royal Golden Jubilee Ph.D. Program Scholarship to RS at PSU.
References
[1] G. Centi, S. Perathoner, Catal. Today 148 (2009) 191–205.
[2] W. Wang, S. Wang, X. Ma, J. Gong, Chem. Soc. Rev. 40 (2011) 3703–3727.
[3] C. Song, CO2 Conversion and Utilization: An Overview, in: C. Song, A.F. Gaffney,
K. Fujimoto (Eds.), CO2 conversion and utilization, ACS Symposium Series, vol.
809, American Chemical Society, Washington, DC, 2002, pp. 2–30 (Chapter 1).
[4] C. Song, Catal. Today 115 (2006) 2–32.
[5] Z. Jiang, T. Xiao, V.L. Kuznetsov, P.P. Edwards, Philos. Trans. R. Soc. A 368 (2010)
3343–3364.
[6] M. Mikkelsen, M. Jorgensen, F.C. Krebs, Energy Environ. Sci. 3 (2010) 43–81.
[7] M. Aresta, A. Dibenedetto, A. Angelini, J. CO2 Util. 3–4 (2013) 65–73.
[8] S. Krishnamoorthy, A. Li, E. Iglesia, Catal. Lett. 80 (2002) 77–86.
[9] G.D. Weatherbee, C.H. Bartholomew, J. Catal. 87 (1984) 352–362.
[10] M.R. Prairie, A. Renken, J.G. Highfield, K. Ravindranathan Thampi, M. Grätzel, J.
Catal. 129 (1991) 130–144.
[11] P.W. Yesgar, M. Sheintuch, J. Catal. 127 (1991) 576–594.
[12] G. Du, S. Lim, Y. Yang, C. Wang, L. Pfefferle, G.L. Haller, J. Catal. 249 (2007)
370–379.
[13] T. Riedel, M. Claeys, H. Schulz, G. Schaub, S.-S. Nam, K.-W. Jun, M.-J. Choi, G.
Kishan, K.-W. Lee, Appl. Catal. A 186 (1999) 201–213.
[14] R. Satthawong, N. Koizumi, C. Song, P. Prasassarakich, Top. Catal. 57 (2014)
588–594.
[15] S.-S. Nam, H. Kim, G. Kishan, M.-J. Choi, K.-W. Lee, Appl. Catal. A 179 (1999)
155–163.
[16] P.H. Choi, K.-W. Jun, S.-J. Lee, M.-J. Choi, K.-W. Lee, Catal. Lett. 40 (1996)
115–118.
[17] L. Xu, S. Bao, D.J. Houpt, S.H. Lambert, B.H. Davis, Catal. Today 36 (1997)
347–355.
[18] S.-C. Lee, J.-S. Kim, W.C. Shin, M.-J. Choi, S.-J. Choung, J. Mol. Catal. A: Chem. 301
(2009) 98–105.
[19] R.W. Dorner, D.R. Hardy, F.W. Williams, H.D. Willauer, Appl. Catal. A 373 (2010)
112–121.
[20] R.W. Dorner, D.R. Hardy, F.W. Williams, H.D. Willauer, Catal. Commun. 15
(2011) 88–92.
[21] T. Herranz, S. Rojas, F.J. Pérez-Alonso, M. Ojeda, P. Terreros, J.L.G. Fierro, Appl.
Catal. A 311 (2006) 66–75.
[22] R. Satthawong, N. Koizumi, C. Song, P. Prasassarakich, J. CO2 Util. 3–4 (2013)
102–106.
[23] M. Cubeiro, H. Morales, M. Goldwasser, M. Pérez-Zurita, F. González-Jiménez,
React. Kinet. Catal. Lett. 69 (2000) 259–264.
[24] C.-S. Chen, W.-H. Cheng, S.-S. Lin, Appl. Catal. A 238 (2003) 55–67.
[25] W.-P. Ma, Y.-J. Ding, L.-W. Lin, Ind. Eng. Chem. Res. 43 (2004) 2391–2398.
[26] H. Wan, B. Wu, C. Zhang, H. Xiang, Y. Li, J. Mol. Catal. A: Chem. 283 (2008)
33–42.
[27] G. Zhao, C. Zhang, S. Qin, H. Xiang, Y. Li, J. Mol. Catal. A: Chem. 286 (2008)
137–142.
[28] M. Trépanier, A. Tavasoli, A.K. Dalai, N. Abatzoglou, Appl. Catal. A 353 (2009)
193–202.
[29] Y. Yang, H.-W. Xiang, Y.-Y. Xu, L. Bai, Y.-W. Li, Appl. Catal. A 266 (2004) 181–194.
[30] T. Riedel, G. Schaub, K.-W. Jun, K.-W. Lee, Ind. Eng. Chem. Res. 40 (2001)
1355–1363.
[31] S.-S. Nam, G. Kishan, M.-W. Lee, M.-J. Choi, K.-W. Lee, J. Chem. Res. Synop. (1999)
344–345.
[32] F. Bozso, G. Ertl, M. Grunze, M. Weiss, Appl. Surf. Sci. 1 (1977) 103–119.
[33] J.M. Zowtiak, C.H. Bartholomew, J. Catal. 83 (1983) 107–120.
[34] C. Bartholomew, Catal. Lett. 7 (1990) 27–51.
[35] Y. Pan, C.-J. Liu, Q. Ge, Langmuir 24 (2008) 12410–12419.
[36] T. Montanari, L. Castoldi, L. Lietti, G. Busca, Appl. Catal. A 400 (2011) 61–69.
[37] M. Kantschewa, E.V. Albano, G. Etrtl, H. Knözinger, Appl. Catal. 8 (1983) 71–84.
[38] T. Hirano, Appl. Catal. 26 (1986) 65–79.