Download as pdf or txt
Download as pdf or txt
You are on page 1of 15

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 4 2 1 8 5 e4 2 1 9 9

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.elsevier.com/locate/he

Light olefins synthesis from CO2 hydrogenation


over mixed FeeCoeK supported on micro-
mesoporous carbon catalysts

Thongthai Witoon a,b,*, Thanapha Numpilai a,b,c, Khanin Nueangnoraj d,


Chin Kui Cheng e, Metta Chareonpanich a,b, Jumras Limtrakul f
a
Center of Excellence on Petrochemical and Materials Technology, Department of Chemical Engineering, Faculty of
Engineering, Kasetsart University, Bangkok, 10900, Thailand
b
Center for Advanced Studies in Nanotechnology for Chemical, Food and Agricultural Industries, KU Institute for
Advanced Studies, Kasetsart University, Bangkok, 10900, Thailand
c
Department of Environmental Science, Faculty of Science and Technology, Thammasat University, Pathumthani,
12120, Thailand
d
School of Bio-Chemical Engineering and Technology, Sirindhorn International Institute of Technology (SIIT),
Thammasat University, Thailand
e
Center for Catalysis and Separation (CeCaS), Department of Chemical Engineering, College of Engineering, Khalifa
University, P.O. Box 127788, Abu Dhabi, United Arab Emirates
f
Department of Materials Science and Engineering, School of Molecular Science and Engineering, Vidyasirimedhi
Institute of Science and Technology, Rayong, 21210, Thailand

highlights graphical abstract

 FeeCoeK loaded on microporous


and micro-mesoporous carbon
supports.
 Micropores confined particle sizes
of metal oxides and level of
aggregates.
 Mesopores allowed aggregates of
metal oxides providing an inti-
mate contact of particles.
 FeeCoeK supported on micropo-
rous carbon produced a high frac-
tion of CH4.
 FeeCoeK supported on micro-
mesoporous carbon provided a
high light olefins yield.

article info abstract

Article history: Recycling CO2 into light olefins is a promising approach to reduce CO2 emissions. To pro-
Received 8 August 2021 mote this technology, an efficient catalyst with high activity and selectivity towards light

* Corresponding author. Center of Excellence on Petrochemical and Materials Technology, Department of Chemical Engineering, Faculty
of Engineering, Kasetsart University, Bangkok, 10900, Thailand.
E-mail address: fengttwi@ku.ac.th (T. Witoon).
https://doi.org/10.1016/j.ijhydene.2021.10.265
0360-3199/© 2021 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.
42186 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 4 2 1 8 5 e4 2 1 9 9

Received in revised form olefins is imperative. In this work, a series of FeeCoeK supported on micro-mesoporous
7 October 2021 carbon (MMC) and microporous carbon (MC) with different metal loading contents (20
Accepted 31 October 2021 e80 wt%) were prepared for CO2 hydrogenation to light olefins. Impregnating mixed metal
Available online 21 November 2021 oxides on both MMC and MC reduced particle sizes and enhanced their dispersion and
reducibility, yielding a higher CO2 conversion compared to the unsupported FeeCoeK
Keywords: catalyst. The metal oxides were highly dispersed inside the micropores of MC support,
CO2 hydrogenation achieving the highest CO2 conversion. However, the high dispersion of metal oxides inside
Light olefins micropores led to the formation of isolated particles with a low interfacial contact with
Fe-based catalysts each other, resulting in a low light olefins selectivity. The MMC support provided a lower
Micro-mesoporous carbon degree of metal dispersion, creating more interfacial contact area, promoting the selec-
Microporous carbon tivity towards olefins. Overall, the 60 wt% FeeCoeK supported on MMC catalyst exhibited
the highest light olefins yield of 10.8% at 400  C and 20 bar and excellent stability.
© 2021 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

operating these reactions in tandem needs to be carried out at


Introduction high reaction temperature (300e350  C) for achieving a high
CO to be converted into hydrocarbons. Fe-based catalysts are
Human activities like combustion of fossil fuels have contin- more favored because they offer short-chain hydrocarbons as
uously increased CO2 levels in the atmosphere which irre- main products unlike the Co-based catalysts, yielding mainly
versibly affects the climate change. As a consequence, CH4 [22,23]. The activity and selectivity of light olefins pro-
mitigation of CO2 emissions is an urgent necessity. Various nouncedly enhanced by alkali promoters (i.e., Li, Na, K and Cs)
approaches, comprising CO2 capture, storage and utilization as well as secondary metals (i.e., Cu, Co, Mn and Zn)
have been employed to curtail the CO2 concentrations [1e3]. [17,24e28], which was attributed to the modifying adsorption-
Among them, utilizing CO2 as a carbon source for producing desorption properties of CO2 and H2 on the catalyst surface.
high-value chemicals with large market share and growth, The alkali promoters are responsible for donating electrons to
especially light olefins (i.e., ethylene, propylene and butylene) active Fe atoms, strengthening the interaction between
has received much attention [4e9]. Light olefins, generally adsorbed H and the surface of Fe, suppressing the hydroge-
produced by the petroleum-derived naphtha cracking process nation activity and thus bringing about more light olefins
and ethane/propane dehydrogenation [10e12], are important formation. Meanwhile, it enhances strong basic sites which
platform molecules in the production of polymers, synthetic contribute to the increase in the CO2 adsorption. The sec-
fibers, packaging, construction materials, and so on. With the ondary metals can manipulate the products distribution by
dwindling supply of fossil fuels resources and the rapidly neither accreting nor restraining the chain-growth reaction,
growing demands for light olefins, converting CO2 to light which contributes to boosting the light olefins productivity.
olefins becomes more and more attractive. Furthermore, confinement of Fe nanoparticles into porous
The conversion of CO2 to light olefins can take place matrices also potentially brings benefits for the catalytic per-
through two different paths: (1) CO2 hydrogenation to meth- formance [29,30]. Compared to the oxide materials, a moder-
anol (MeOH)/dimethyl ether (DME) over In2O3-based, ZnOx- ate metalsupport interaction between carbonaceous
ZrO2 and spinel oxide (i.e., ZnGa2O4 and ZnAl2O4) catalysts substrates and Fe nanoparticles is more beneficial to the
followed by its conversion into light olefins over acid sites of reduction of Fe species, boosting active phase generation and
zeolites [13e16], and (2) CO2 transformation to CO by a reverse thus resulting in a higher catalytic performance [31e33].
water gas shift (RWGS) reaction and subsequent CO conver- Carbon materials in the form of activated carbon (AC),
sion to light olefins via FischereTropsch (FT) synthesis carbon nanotubes (CNTs), carbon nanofibres (CNFs), gra-
[17e19]. Despite the satisfied selectivity towards light olefins phene oxide and metal organic frameworks (MOFs) are
in hydrocarbon products (>70%), the former route suffers from extensively used for active metals dispersion due to their
the poor performance in term of productivity (light olefins high accessible surface areas, multitudinous pore structure,
yield <10%) due to thermodynamic limitation of CO2-to-MeOH ultrahigh porosity and tunable functionalities [18,34e41]. In
reaction. In the latter approach, beyond 10% light olefins yield addition, a superior heat-transfer property of carbon mate-
can be achieved as a high overall CO2 conversion associated rials is useful for a highly exothermic reaction, alleviating a
with the thermodynamically favored RWGS reaction, making hot spot in the catalyst bed [35]. In comparison to those
it more attractive for industrial-scale applications [6,20,21]. carbon materials, AC which can be prepared from residual
However, this process is restricted by low selectivity towards biomass and waste materials is more attractive from both
light olefins (<50%) regardless of diverse mixture of hydro- economics and sustainability perspective. Regarding the
carbons products, requiring an efficient catalyst for highly large-scale manufacturing, the production of advanced car-
selective transformation of CO2 and H2 to desired light olefins. bon materials like graphene and CNTs is restricted by their
Fe- and Co-based catalysts can function efficiently for both high cost relevant to low mass production of high-quality
RWGS and FT synthesis reactions in the latter route. However,
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 4 2 1 8 5 e4 2 1 9 9 42187

materials, long time consumption and multiple processing supports were examined by N2 physisorption at 196  C
steps [42e44]. The production cost of graphene, graphene (Micromeritics 3 Flex) after outgassing at 200  C under vacuum
oxide and CNTs are in the ranges of $50e10,000/kg, while AC for 24 h.
has relatively low price ($3e8/kg) [44]. Besides, a high envi- Transmission electron microscope (TEM) images were
ronmental impact of a graphene-based device production attained on a JEOL JEM-2010 microscope operated at 200 kV.
which provided two-times higher global warming potential Powder X-ray diffraction (XRD) patterns of the calcined,
(GWP) than an activated carbon-based supercapacitor [45], reduced and spent FeeCoeK supported on carbon materials
makes AC more environmentally friendly and potentially were recorded with a Bruker D8 Advance Phaser diffrac-
applicable for widespread applications, including catalytic tometer using Cu-Ka radiation (l ¼ 1.5406  A). The 2q was
system. The Fe-based catalysts supported on AC have been scanned from 25 to 70 with a step of 0.05 . The Fe crystallite
previously investigated in hydrocarbons, including light size of the reduced catalysts were calculated using the
olefins synthesis from syngas conversion (CO þ H2), Scherrer Equation [46].
achieving high olefins selectivity of 39.4% and 30% over Thermogravimetric (TG) analysis experiment was con-
FeeMneK/AC [40] and FeeCueMn/AC [41] catalysts, respec- ducted on a Mettler-Toledo TGA/DSC 3þ analyzer by heating a
tively. However, applying AC as an effective support for Fe- sample from 30 to 400  C at a heating rate of 10  C min1 in a
based catalyst in light olefins synthesis from CO2 hydroge- flow of N2 (75 mL min1).
nation has not yet been implemented. Moreover, there are no H2 temperature-programmed reduction (H2-TPR) was per-
reports regarding the influence of coexisting pore sizes formed with a Micromeritics AutoChem II 2920 apparatus. The
ranging from micropores to mesopores in the structure of as-prepared catalyst (0.1 g) was mounted in a U-type quartz
AC. Herein, a series of FeeCoeK oxide supported on AC tube and heated from room temperature to 150  C with a
materials with different metal loading contents were pre- ramping rate of 10  C min1 under Ar flow (50 mL min1) for
pared. The effects of pore structures including bimodal 30 min. Then, the temperature was lowered to 50  C and gas
micro-mesoporous carbon (MMC) and unimodal micropo- flow was switched to 10% H2/Ar. The response to TCD for the
rous carbon (MC) on the physicochemical properties of H2 consumption was monitored continuously upon rising
mixed FeeCoeK oxides supported catalysts as well as their temperature from 50  C to 900  C with a heating rate of 10  C
catalytic performance in CO2 hydrogenation to light olefins min1.
were investigated. Temperature-programmed desorption of CO2 (CO2-TPD)
and H2 (H2-TPD) measurements were carried on the same
apparatus as that in TPR analysis. 0.1 g sample was firstly
Experimental reduced in situ with 10% H2/Ar (50 mL min1) at 400  C for 5 h,
and then cooled to 50  C under Ar flow (50 mL min1). After-
Catalyst preparation ward, 10% CO2/He or 10% H2/Ar flow was passed over the
catalyst bed for 1 h, followed by flushing with Ar for at least
The carbon materials with different pore characteristics, 30 min to eliminate the physisorbed species until the baseline
including micro-mesoporous carbon (MMC) and microporous was restored. The desorption curve was acquired by rising
carbon (MC) were purchased from Sigma-Aldrich, Singapore. temperature from 50  C to 600  C under Ar flow at a heating
A series of FeeCoeK oxide supported on carbon materials rate of 10  C min1.
were prepared by two-step wetness impregnation method.
Briefly, the carbon support was first impregnated with desired Catalytic activity test
amount of KNO3 solution in order to avoid a block of metal
active sites by K deposition. The mixture was stirred at 60  C The catalytic CO2 hydrogenation over prepared FeeCoeK
for 4 h and then dried at 100  C for 24 h. Subsequently, the oxide supported on carbon materials were executed in a fixed-
KNO3/carbon was further impregnated with mixed solution bed reactor. Prior to the reaction, 1 g of catalyst mixed with 3 g
containing Fe(NO3)3$9H2O and Co(NO3)2$6H2O. The molar ratio of SiO2 sand was loaded in the stainless-steel tube and
of Fe:Co:K was 1:0.2:0.5. The sample was eventually dried reduced in H2 flowing at 400  C for 5 h under. After cooling to
overnight at 100  C and calcined in a flow of N2 at 400 for 4 h. 300  C, the feed stream was switched to H2/CO2/N2 (3/1/1)
The resulting catalysts were labeled as xFe-Co-K/yMMC and mixed gas with total flow rate of 75 mL min1, and the reaction
xFe-Co-K/yMC, in which x and y represented weight percent temperature and pressure were raised to 320  C and 20 bar.
of FeeCoeK and carbon in the total weight of catalysts, The effluent gases were collected every hour for 4 h and then
respectively. The unsupported FeeCoeK catalyst was also the reaction temperature increased to 340  C with a heating of
employed for purpose of comparison. In Brief, Fe(NO3)3$9H2O, 2  C min1. The method was repeated for the reaction tem-
Co(NO3)2$6H2O and KNO3 were dissolved in 5 mL of deionized perature of 340  C, 350  C, 360  C, 380  C and 400  C. The outlet
water. The mixture was stirred at 60  C for 4 h, followed by gas compositions were analyzed by two gas chromatographs
dried at 100  C for 24 h, and eventually calcined in a flow of N2 (GC). CO2, CO, H2 and N2 were measured by GC-14B (Shimadzu)
at 400 for 4 h. with a thermal conductivity detector (TCD) and a Shin Carbon
ST column. While, GC-8A (Shimadzu) connected with a flame
Catalyst characterization ionization detector (FID) and Porapak Q column was employed
for separating C1eC4 hydrocarbons, and C5eC18 hydrocarbons
The physical properties, including surface area, pore volume were analyzed with an OV-1 column. The CO2 conversion,
and pore diameter of calcined samples as well as carbon selectivities to CO and hydrocarbon, hydrocarbon
42188 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 4 2 1 8 5 e4 2 1 9 9

distribution, olefins to paraffins ratio and light olefins yield 1.1 nm and 34 nm, respectively (Fig. 1b). Unlike the isotherm of
were calculated on a molar carbon basis according to the MMC, the isotherm of MC was a type I (Fig. 1c) which was
equations (1)e(7): indicative of micropores. The MC displayed a narrow pore size
distribution in the range of 1e3 nm (Fig. 1d). The unsupported
nproducts  100
CO2 conversion ð%Þ ¼ (1) FeeCoeK exhibited type II isotherm, representative of non-
nproducts þ moles CO2;out
porous and macroporous materials (Fig. 1a). The pore size
distribution of the unsupported FeeCoeK was 20e100 nm
X
n
nproducts ¼ moles COout þ i  moles of hydrocarboni ; i created by aggregates of metal oxides particles that give rise to
i¼1 pores. The isotherm of FeeCoeK oxides loaded onto MMC
¼ atomic number of carbon displayed between that of MMC and unsupported FeeCoeK
(2) (Fig. 1a), and the volume of gas adsorbed significantly
decreased at P/P0 < 0.1 and 0.5 < P/P0 < 0.99, indicating that
moles COout  100 particles of metal oxides were placed on a portion of micro-
Selectivity of CO ð%Þ ¼ (3) pores and mesopores (Fig. 1b). Loading mixed FeeCoeK oxides
nproducts
onto MC (40FeeCoeK/60MC) led to a significant reduction of
volume of gas adsorbed at P/P0 < 0.1, indicating the deposition

P
n
i  moles of hydrocarboni  100
i¼1
Selectivity of hydrocarbon ð%Þ ¼ (4)
nproducts

of metal oxides inside the micropores. Further increasing


metal content to 60 wt% (60FeeCoeK/40MC) increased the
volume of gas adsorbed at P/P0 of 0.99, suggesting the gener-
i moles of hydrocarboni 100
Hydrocarbon distributioni ð%Þ¼ ation of macropores. The results demonstrated that micro-
Pn
i moles of hydrocarboni pores of MC support were filled at 60 wt% metal loading, and
i¼1
part of the remaining metal oxides was aggregated them-
(5)
selves outside micropores. The BET surface area, pore volume
and average pore diameter of MMC, MC and all catalysts are
Olefins to paraffins ratio ¼ given in Table 1.
2 x moles of C2 H4 þ 3 x moles of C3 H6 þ 4 x moles of C4 H8 Transmission electron microscope (TEM) and scanning
2 x moles of C2 H6 þ 3 x moles of C3 H8 þ 4 x moles of C4 H10 electron microscope (SEM) were performed to capture the
(6) structure of FeeCoeK oxide supported on carbons materials;

2 x moles of C2 H4 þ 3 x moles of C3 H6 þ 4 x moles of C4 H8


Yield of light olefins ð%Þ ¼ x 100 (7)
nproducts þ moles CO2;out

the results are shown in Figs. 2 and 3, respectively. TEM image


of the unsupported FeeCoeK showed an aggregation of par-
ticles with their size larger than 10 nm (Fig. 2a). However, its
Results and discussion morphology (Fig. 3a) was inhomogeneous as the K-rich phase
confirmed by EDX analysis (Fig. S1) was present. Impregnating
Catalyst characterization 40 wt% FeeCoeK onto MMC (Fig. 2b) enhanced the distribution
of metal oxides and reduced their particle sizes. Nevertheless,
Fig. 1 shows N2 sorption isotherm and pore size distribution of an aggregation of particles on the external surface of carbon
micro-mesoporous carbon (MMC), microporous carbon (MC), material was also observed (Fig. 3b). In comparison to the
unsupported FeeCoeK, and FeeCoeK oxide supported on MMC support, the distribution of metal oxides (40 wt%) over
MMC and MC with different metal loading contents. As shown the microporous carbon (Fig. 2c) was even better and the
in Fig. 1a, the isotherm of MMC was a composite type I and IV particle sizes of metal oxides were smaller attributed to the
with H3 hysteresis loop, a characteristic of micro-mesoporous regulation of particle sizes by the micropores (1e3 nm) of MC.
materials consisting of slit-shaped pores. The pore size dis- In addition, there was a tiny portion of particles existed on the
tribution of MMC ranged 1e2 nm and 2e100 nm centered at external surface of microporous carbon (Fig. 3c). Increasing
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 4 2 1 8 5 e4 2 1 9 9 42189

Fig. 1 e N2 sorption isotherm (a and b) and pore size distribution (c and d) of micro-mesoporous carbon (MMC), microporous
carbon (MC), calcined unsupported FeeCoeK oxide catalyst, and calcined FeeCoeK oxide supported on carbons materials
(MMC and MC) with different metal loading contents.

the metal loading content to 60 wt% (Fig. 2d) led to the for- 0493), Fe2O3 (PDF 01-071-5088), and Co3O4 (PDF 00-042-1467).
mation of larger particle sizes aggregated outside the pores. The diffraction lines representative of Fe2O3, Co3O4, and KNO3
SEM images clearly revealed that number of particles gathered phases were no longer detectable, and Fe3O4 (PDF 01-076-0955)
on the outside of carbon materials steadily increased with phase appeared instead for all FeeCoeK supported on carbon
increasing metal loading contents (Fig. 3def). materials. In addition, CoO (PDF 01-089-2803) phase was also
Fig. 4 shows the XRD patterns of unsupported FeeCoeK visible for the catalysts with low metal loading contents
sample and FeeCoeK supported on carbon materials (MMC (20FeeCoeK/80MMC and 40FeeCoeK/60MC). The results
and MC) with different metal loading contents. The XRD suggested that the carbon materials facilitated the reduction
pattern of the unsupported FeeCoeK displayed the presence of metal oxides during the calcination stage. Nevertheless, the
of mixed phases of metal oxides, including rhombohedral diffraction line of Fe3O4 at 35.5 of all FeeCoeK supported on
KNO3 (PDF 01-078-7937), orthorhombic KNO3 (PDF 00-001- carbon materials became broader as the carbon contents
42190 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 4 2 1 8 5 e4 2 1 9 9

decreased, especially for 80FeeCoeK/20MMC (e). Note that the


Table 1 e BET surface area, pore volume, average pore
broad diffraction line overlapped with a diffraction line of
diameter and metals composition determined by SEM-
EDX of MMC, MC and FeeCoeK supported on MMC and Fe2O3 (35.7 ). The results suggested that Fe2O3 might coexist
MC with different metal loadings. with Fe3O4 for the samples with lower carbon contents.
Fig. 5 displays the TG analysis of MMC, unsupported
Catalysts BET surface Pore Average
area volume pore FeeCoeK, and FeeCoeK supported on carbon materials (MMC
and MC) with different metal loading contents. The TG profile
(m2g1) (cm3g1) diameter
(nm) of MMC showed one step of weight loss at 30e100  C due to the
desorption of adsorbed moisture. The weight of MMC
Micro-mesoporous 782 0.84 4.2
remained unchanged at temperatures higher than 100  C,
carbon (MMC)
Microporous carbon (MC) 1757 0.95 2.2
indicating its high thermal stability. Note that the TG profile of
FeeCoeK 5.5 0.03 23.6 MC was similar to that of MMC. The unsupported FeeCoeK
20FeeCoeK/80MMC 496 0.50 4.2 exhibited two steps of weight loss at 30e100  C and 200e400  C
40FeeCoeK/60MMC 275 0.31 4.5 corresponded to the evaporation of water molecules (2.2%)
60FeeCoeK/40MMC 109 0.18 6.5 and the decomposition of iron and cobalt nitrates precursors
80FeeCoeK/20MMC 36 0.11 11.6
(15.42%), respectively. All FeeCoeK supported on carbon ma-
40FeeCoeK/60MC 438 0.27 2.3
terials also presented two steps of weight loss. The weight loss
60FeeCoeK/40MC 45 0.07 6.4
in the second step was 7.96%, 12.88%, 16.8%, 19.32%, 11.79%,

Fig. 2 e TEM images of unsupported FeeCoeK oxide catalyst (a) and FeeCoeK oxide supported on carbons materials (MMC
and MC) with different metal loading contents: 40FeeCoeK/60MMC (b), 40FeeCoeK/60MC (c) and 60FeeCoeK/40MC (d).
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 4 2 1 8 5 e4 2 1 9 9 42191

Fig. 3 e SEM images of unsupported FeeCoeK oxide catalyst (a) and FeeCoeK oxide supported on carbons materials (MMC
and MC) with different metal loading contents: 40FeeCoeK/60MMC (b), 40FeeCoeK/60MC (c), 60FeeCoeK/40MMC (d),
60FeeCoeK/40MC (e) and 80FeeCoeK/20MMC (f).

and 16.05% for 20FeeCoeK/80MMC, 40FeeCoeK/60MMC, results were in good agreement with the XRD analysis (Fig. 4).
60FeeCoeK/40MMC, 80FeeCoeK/20MMC, 40FeeCoeK/60MC, SEM-EDS was used to determine the metals composition of
and 60FeeCoeK/40MC, respectively. Based on the decompo- different catalysts; the results are listed in Table 2. The molar
sition of iron and cobalt nitrates precursors for the 100 wt% ratio of Fe:Co:K of all catalysts was closed to the starting
FeeCoeK catalyst, the weight loss in the second step was metals composition (Fe:Co:K ¼ 1:0.2:0.5), while the carbon
theoretically proportional to 3.08%, 6.17%, 9.25%, and 12.34% content of all catalysts was slightly lower than their starting
for 20 wt%, 40 wt%, 60 wt%, and 80 wt% metal loading, values. In addition, the amount of Fe, Co and K for
respectively. The larger weight loss in the second step of all 80FeeCoeK/20MMC was greater than those of unsupported
FeeCoeK supported on carbon materials was attributed to FeeCoeK (100 wt% FeeCoeK), confirming the removal of ox-
losses of oxygen atoms of Fe2O3 and Co3O4 phases and carbon ygen atoms of Fe2O3 and Co3O4 phases.
atoms of carbon materials caused by the interaction of those The reduction behavior of metal oxides in different cata-
metal oxides with carbon to produce CO and CO2 (Fig. 5c). The lysts was investigated by H2-TPR (Fig. 6). The unsupported
42192 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 4 2 1 8 5 e4 2 1 9 9

FeeCoeK (a) showed H2 consumption at three temperature


ranges (a, b and g) centered at 350  C, 540  C and 830  C
assigned to a reduction of Fe2O3 to Fe3O4 and Co3O4 to CoO [26],
a transformation of Fe3O4 to FeO to metallic Fe, and CoO to
metallic Co, and a reduction of isolated FexOy particles
without the interaction with Co particles, respectively.
Impregnating FeeCoeK oxides onto the carbon supports (c-g)
remarkably altered the H2-TPR profiles as the broad peak at g
region disappeared, indicating an intimate contact of oxidic Fe
and Co. Nevertheless, a shoulder peak at g region of 20 wt%
metal oxides loaded on carbon material (b) was visible, sug-
gesting a lower contact of oxidic Fe and Co particles due to a
very high dispersion of each metal oxides particle. The
amount of H2 consumption for each temperature range was
reported in Table 3. The FeeCoeK oxide supported on carbon
materials exhibited a significant lower H2 consumption at a
region compared to the unsupported FeeCoeK because the
initial phase of Fe and Co species of all FeeCoeK oxide sup-
ported on carbon materials was mainly reduced to Fe3O4 and
CoO, respectively.
The phases of catalysts subsequent to the reduction stage
were examined with XRD (Fig. 7). All catalysts showed the
Fig. 4 e XRD patterns of calcined unsupported FeeCoeK presence of Fe3O4 (PDF 01-076-0955) and metallic Fe (PDF 00-
oxide catalyst (a) and calcined FeeCoeK oxide supported 006-0696) phases. However, Fe2O3 (PDF 01-071-5088) still
on carbons materials (MMC and MC) with different metal remained in the unsupported FeeCoeK corresponded to the
loading contents: 20FeeCoeK/80MMC (b), 40FeeCoeK/ H2-TPR analysis (Fig. 6). The metallic Fe crystallite size was
60MMC (c), 60FeeCoeK/40MMC (d), 80FeeCoeK/20MMC (e), calculated using the line width of Fe (110) reflection (44.7 )
40FeeCoeK/60MC (f) and 60FeeCoeK/40MC (g). with the Scherrer Equation (Table 3). The unsupported

Fig. 5 e TG profile (a) and its derivative (b) of micro-mesoporous carbon (MMC), unsupported FeeCoeK oxide catalyst and
FeeCoeK oxide supported on carbon materials (MMC and MC) with different metal loading contents and schematic diagram
representative of metal oxides reduction assisted by carbon materials (c).
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 4 2 1 8 5 e4 2 1 9 9 42193

Table 2 e Metals composition determined by SEM-EDS of Table 3 e H2 consumption of unsupported FeeCoeK and
calcined unsupported FeeCoeK and all FeeCoeK FeeCoeK oxide supported on carbon materials
supported on carbon materials. determined from the area under H2-TPR curve calibrated
Catalysts Element (wt%) Molar ratio by an injection of known H2 volume.
of Fe:Co:K Catalysts H2 consumption Fe
Fe Co K C O
(mmol g1) crystallite
FeeCoeK 36.3 8.7 12.6 e 42.4 1:0.23:0.50 size (nm)
А b g Total
20FeeCoeK/80MMC 9.5 1.9 2.8 71.4 14.4 1:0.19:0.43
40FeeCoeK/60MMC 17.7 2.8 4.4 56.5 18.6 1:0.19:0.44 FeeCoeK 369.5 277.5 198.1 845.1 41.4
60FeeCoeK/40MMC 20.8 4.9 7.0 36.7 30.6 1:0.22:0.48 20FeeCoeK/80MMC 44.2 127.1 31.8 213.1 22.4
80FeeCoeK/20MMC 37.1 8.8 12.8 17.8 23.5 1:0.22:0.49 40FeeCoeK/60MMC 43.9 203.7 247.6 31.9
40FeeCoeK/60MC 15.7 2.5 4.0 54.7 23.1 1:0.19:0.47 60FeeCoeK/40MMC 43.0 329.8 372.8 34.9
60FeeCoeK/40MC 24.9 5.4 8.6 30.6 30.5 1:0.20:0.49 80FeeCoeK/20MMC 72.8 479.3 552.1 40.4
40FeeCoeK/60MC 62.2 214.9 277.1 30.6
60FeeCoeK/40MC 49.5 317.7 367.2 40.2

Fig. 7 e XRD patterns of reduced FeeCoeK oxide catalyst (a)


and reduced FeeCoeK oxide supported on carbons
materials (MMC and MC) with different metal loading
contents: 20FeeCoeK/80MMC (b), 40FeeCoeK/60MMC (c),
Fig. 6 e H2-TPR profile of unsupported FeeCoeK oxide
60FeeCoeK/40MMC (d), 80FeeCoeK/20MMC (e),
catalyst (a) and FeeCoeK oxide supported on carbon
40FeeCoeK/60MC (f) and 60FeeCoeK/40MC (g).
materials (MMC and MC) with different metal loading
contents: 20FeeCoeK/80MMC (b), 40FeeCoeK/60MMC (c),
60FeeCoeK/40MMC (d), 80FeeCoeK/20MMC (e), desorption peak at 474  C was visible attributed to the
40FeeCoeK/60MC (f) and 60FeeCoeK/40MC (g). desorption of H2 from FexOy interacted with K2O [24]. In
comparison to the unsupported FeeCoeK, adding 20 wt%
metal oxides onto MMC (20FeeCoeK/80MMC) led to a signifi-
FeeCoeK exhibited the largest Fe crystallite size of 41.4 nm. cant increase of H2 desorbed at 120  C due to the higher
The Fe crystallite size reduced to 22.4 nm for 20 wt% metal fraction of metallic Fe with its smaller particle sizes.
oxide loading (20FeeCoeK/80MMC) and gradually increased Increasing metal oxides loading to 40 wt% (40FeeCoeK/
with increasing metal loading contents. 60MMC) developed the strong adsorption of H2 ranging from
The interaction of reactant gases with the catalyst surface 300 to 600  C attributed to a higher interfacial contact of FexOy
was examined with H2-TPD and CO2-TPD; the results are and K2O particles. The strong adsorption of H2 became more
displayed in Fig. 8a,b, respectively. The H2-TPR profile of un- pronounced when the metal oxides loading increased up to
supported FeeCoeK showed three notable peaks at 120  C, 60 wt% (60FeeCoeK/40MMC). Further increase of metal oxides
211  C, and 291  C assigned to the desorption of H2 from loading to 80 wt% (80FeeCoeK/20MMC) decreased the weak
metallic Fe, Fe3O4, and Fe2O3, respectively. A small broad adsorption of H2. The FeeCoeK supported on MC (40FeeCoeK/
42194 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 4 2 1 8 5 e4 2 1 9 9

Fig. 8 e H2-TPD (A) and CO2-TPD (B) profiles of reduced FeeCoeK oxide catalyst (a) and reduced FeeCoeK oxide supported on
carbons materials (MMC and MC) with different metal loading contents: 20FeeCoeK/80MMC (b), 40FeeCoeK/60MMC (c),
60FeeCoeK/40MMC (d), 80FeeCoeK/20MMC (e), 40FeeCoeK/60MC (f) and 60FeeCoeK/40MC (g).

The CO2-TPD profile of unsupported FeeCoeK showed one


prominent desorption peak, and a broad desorption peak
ranging from 150 to 600  C. Both desorption peaks of the
FeeCoeK supported on MMC and MC became more pro-
nounced. The development of the low-temperature desorp-
tion peak was associated with the higher dispersion of metals
on the supports, providing a greater number of exposed sur-
face area for CO2 adsorption. The larger desorption peak at
higher temperature was attributed to the interfacial contact of
K2O particles with Fe and Co particles, favoring the stronger
interaction of CO2 with the catalyst surface due to a donation
of electrons from K to FexOy and CoxOy. In comparison to the
FeeCoeK supported on MC, the FeeCoeK supported on MMC
exhibited a higher CO2 adsorption compared to the FeeCoeK
supported on MC, especially at 40 wt% metal loading content
(40FeeCoeK/60MMC and 40FeeCoeK/60MC). This was
because the limitation of metal oxides growth inside the mi-
cropores of MC reduced the interfacial contact of particles as
indicated by the N2 sorption (Fig. 1) and TEM (Fig. 2) analyses.

Catalytic activity test


Fig. 9 e CO2 conversion as a function of reaction
temperatures of different catalysts. Fig. 9 shows total CO2 conversion at different reaction tem-
peratures of all catalysts. The total CO2 conversion of all cat-
60MC and 60FeeCoeK/40MC) possessed a higher weak H2 alysts progressively increased with ascending reaction
adsorption compared to the FeeCoeK supported on MMC at temperature from 320 to 400  C. The reaction of CO2 and H2 to
the identical metal loading content. produce hydrocarbon proceeded via two consecutive
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 4 2 1 8 5 e4 2 1 9 9 42195

Fig. 10 e CO (a) and hydrocarbon (b) selectivity as a function of reaction temperatures of different catalysts.

reactions involving reverse water gas shift reaction (reaction (Fig. 7) which was inactive for CO hydrogenation to hydro-
1) and Fischer-Tropsch synthesis reaction (reaction 2) as carbon. However, a sudden decrease in CO selectivity of the
follows: unsupported FeeCoeK catalyst was visible at temperature
>350  C, and its CO selectivity at 380  C was lower than all
CO2 þ H2 4 CO þ H2O (reaction 1) FeeCoeK supported on MMC and MC catalysts, except that of
40FeeCoeK/60MC. The result indicated that the rate of CO
nCO þ 2nH2 4 (eCH2e)n þ nH2O (reaction 2) conversion (reaction 2) of the unsupported FeeCoeK catalyst
was improved at high temperatures which might be attributed
The reverse water gas shift was an endothermic reaction to a change of catalyst phases. The CO selectivities of 60 wt%
for which its equilibrium CO2 conversion increased with metal loading on MMC (60FeeCOeK/40MMC) MC and
increasing reaction temperatures. In addition, the rate of re- (60FeeCoeK/40MC) were similar. However, the significant
action also increased with rising temperatures. Consequently, difference in CO selectivity for different carbon supports was
the increased reaction temperature assuredly offered a high visible at 40 wt% metal loading. The 40 wt% FeeCoeK sup-
CO2 conversion. In contrast to the reverse water gas shift re- ported on MC catalyst showed a much lower CO selectivity
action, Fischer-Tropsch synthesis was an exothermic reac- attributed to the highest number of active sites for CO con-
tion, and thus the equilibrium CO conversion (reaction 2) version to hydrocarbon.
decreased with rising reaction temperatures. Although the Fig. 11 displays the CH4 (Fig. 11a) and C5þ (Fig. 11b) distri-
reaction rate of Fischer-Tropsch synthesis was more favored bution, C2eC4 olefin to C2eC4 paraffin ratio (Fig. 11c), and light
at higher temperatures, the CO conversion was limited by the (C2eC4) olefin yield (Fig. 11d) of different catalysts. The
thermodynamics equilibrium. Therefore, the higher CO2 FeeCoeK supported on MC catalysts showed the highest CH4
conversion at higher temperatures indicates that both re- distribution followed by the unsupported FeeCoeK and the
actions are in a kinetic regime. The unsupported FeeCoeK FeeCoeK supported on MMC catalysts. The reverse trend was
catalyst gave the total CO2 conversion of 18.7%, 22.8%, 25%, attained for the C5þ distribution and C2eC4 olefin to C2eC4
27.9%, 32.2%, and 37% at 320  C, 340  C, 350  C, 360  C, 380  C paraffin ratio. Numpilai et al. [24] reported that the intimate
and 400  C, respectively. Comparing the identical tempera- contact of FexOy species with K entailed an electron transfer
ture, all FeeCoeK supported on MMC and MC catalysts ach- from K to FexOy species. The higher electron density over
ieved a higher CO2 conversion, except the CO2 conversion at FexOy species strengthened the adsorption of CO2 and H2,
400  C of 20FeeCoeK/80MMC due to the lower metal loading improving the CeC chain growth and retarding the hydroge-
content. nation ability of the catalysts, resulting in the formation of
Fig. 10 shows the selectivities of CO (Fig. 10a) and hydro- longer chain hydrocarbon and higher olefin to paraffin ratio,
carbon (Fig. 10b) of all catalysts. The unsupported FeeCoeK respectively. Based on the N2-sorption (Fig. 1) and TEM (Fig. 2)
catalyst gave CO selectivity of 69%, 62%, and 54.6% at 320  C, analyses, 40 wt% metal oxides supported on MC distributed
340  C, and 350  C, respectively. All FeeCoeK supported on solely in micropores, restraining the intimate contact of K
MMC and MC catalysts provided a lower CO selectivity at with FexOy species. The high distribution of isolated FexOy
those operating temperatures because the unsupported particles (Fig. 2c) provided the larger number of active sites for
FeeCoeK catalyst contained the unreduced Fe2O3 species weak H2 adsorption (Fig. 8), promoting the hydrogenation
42196 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 4 2 1 8 5 e4 2 1 9 9

Fig. 11 e CH4 (a) and C5þ (b) distribution, C2eC4 olefin to C2eC4 paraffin ratio (c) and light olefins yield (d) as a function of
reaction temperatures of different catalysts.

ability of the catalyst, favoring the formation of CH4 and The maximum light olefins yield (10.8%) was attained over the
paraffin products. Increasing metal oxides loading to 60 wt% 60FeeCoeK/40MMC catalyst at 400  C and 20 bar. Note that the
allowed the aggregates of metal oxides outside the micro- 60FeeCoeK/40MMC catalyst exhibited an increased C5þ dis-
pores, rising a fraction of longer chain hydrocarbon and O/P tribution (Fig. 11b) accompanied by a decreased olefins to
ratio. The MMC support contained mesopores, giving a suffi- paraffins ratio (Fig. 11c) when the reaction temperature
cient space for the intimate contact of K and FexOy species. As increased from 350 to 360  C. This might be attributed to a
a result, the FeeCoeK supported on MMC achieved a superior slight increase of reaction pressure in the system. In fact, we
performance in terms of olefin formation. Fig. 11d depicts that try to keep the constant reaction pressure at 20 bar. However,
the light olefins yield of both MMC and MC series increased the reaction pressure could be fluctuated in the range of
with increasing metal loading content up to 60 wt%. Further 19e22 bar.
increase of metal loading to 80 wt% (80FeeCoeK/20MMC) Fig. 12 shows the XRD patterns of all spent catalysts after
significantly reduced the amount of H2 adsorbed (Fig. 8a) 12 h of a time-on-stream experiment at 400  C. The XRD
associated with the formation of larger metal particle sizes. pattern of all spent catalysts showed diffraction lines of Fe3O4
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 4 2 1 8 5 e4 2 1 9 9 42197

(PDF 01-076-0955), Fe (PDF 00-006-0696), Fe5C2 (PDF 01-080-


4102), and Fe3C (PDF 00-035-0772) phases. The diffraction line
of Fe2O3 which was apparent in the reduced FeeKeCo catalyst
(Fig. 7) was invisible, indicating that the Fe2O3 phase in the
reduced FeeKeCo catalyst was transformed into more active
Fe phases during the reaction with CO, CO2 and H2 gases. This
is the reason why the CO selectivity of the unsupported
FeeCoeK catalyst suddenly decreased at temperature >350  C
(Fig. 10a). Comparing between the series of MMC and MC
catalysts, the FeeCoeK supported on MC catalysts exhibited
predominant diffraction lines of Fe3O4 phase with poor
diffraction lines of Fe5C2 phase. This affirms an abundant
isolated FexOy particles in the FeeCoeK supported on MC
catalysts because the intimate contact of K and FexOy particles
can stabilize the Fe5C2 phase. Fig. 13 shows the stability test of
the best 60FeeCoeK/40MMC catalyst. The CO2 conversion,
olefins to paraffins ratio, and light olefins yield of the
60FeeCoeK/40MMC catalyst were almost constant during the
50 h time-on-stream experiment, indicating its excellent
stability.

Fig. 12 e XRD patterns of spent FeeCoeK oxide catalyst (a) Conclusions


and spent FeeCoeK oxide supported on carbons materials
(MMC and MC) with different metal loading contents: The present study elucidated the effect of carbon supports,
20FeeCoeK/80MMC (b), 40FeeCoeK/60MMC (c), including micro-mesoporous carbon (MMC) and microporous
60FeeCoeK/40MMC (d), 80FeeCoeK/20MMC (e), carbon (MC), on the structure of FeeCoeK catalysts and their
40FeeCoeK/60MC (f) and 60FeeCoeK/40MC (g). catalytic performance in CO2 hydrogenation to light olefins.
The impregnation of FeeCoeK onto both MMC and MC
reduced particle sizes and thereby improved their dispersion
and reducibility. As a result, all FeeCoeK supported on MMC
and MC catalysts (except 20FeeCoeK/80MMC) achieved a
higher CO2 conversion than the unsupported FeeCoeK even
the metal loading contents for all catalysts were lower. The
FeeCoeK supported on MC (40FeeCoeK/60MC) was the most
active catalyst because the metal particles are well-dispersed
within micropores. However, the main product of the
40FeeCoeK/60MC was methane because the high dispersion
of metal particles lessened the interfacial contact of Fe and K
particles, which was an indispensable determinant for the
formation of long-chain hydrocarbon and olefin products. In
contrast to the MC support, the MMC support allowed aggre-
gates of metal oxides inside mesopores, providing the larger
interfacial contact of Fe and K particles, favoring the selective
formation of olefin products. With the optimum metal
loading, the 60FeeCoeK/40MMC achieved superior perfor-
mance in terms of light olefins yield (10.8%) with excellent
stability.

Credit authorship contribution statement


Fig. 13 e CO2 conversion, olefins to paraffins ratio, and light
olefins yield as a function of time-on-stream of Thongthai Witoon: Investigation, Conceptualization, Meth-
60FeeCoeK/40MMC. Reaction conditions: T ¼ 400  C, odology, Writing-Original draft preparation, Writing-
P ¼ 20 bar, H2:CO2:N2 ¼ 3:1:1, and total flow rate of Reviewing and Editing. Thanapha Numpilai: Investigation,
75 mL min¡1. Validation, Writing-Original draft preparation, Writing-
42198 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 4 2 1 8 5 e4 2 1 9 9

Reviewing and Editing. Khanin Nueangnoraj: Writing- over sodium- and zinc-promoted iron carbide catalysts. J
Reviewing and Editing. Chin Kui Cheng: Visualization, Catal 2021;395:350e61. https://doi.org/10.1016/
Writing-Reviewing and Editing. Metta Chareonpanich: j.jcat.2021.01.036.
[8] Gu H, Ding J, Zhong Q, Zeng Y, Song F. Promotion of surface
Writing-Reviewing and Editing. Jumras Limtrakul:
oxygen vacancies on the light olefins synthesis from
Supervision. catalytic CO2 hydrogenation over Fe-K/ZrO2 catalysts. Int J
Hydrogen Energy 2019;44:11808e16. https://doi.org/10.1016/
j.ijhydene.2019.03.046.
Declaration of competing interest [9] Ding J, Zhao W, Zi L, Xu X, Liu Q, Zhong Q, Xu Y. Promotional
effect of ZrO2 on supported FeCoK catalysts for ethylene
The authors declare that they have no known competing synthesis from catalytic CO2 hydrogenation. Int J Hydrogen
Energy 2020;45:15254e62. https://doi.org/10.1016/
financial interests or personal relationships that could have
j.ijhydene.2020.03.249.
appeared to influence the work reported in this paper. [10] Fakhroleslam M, Sadrameli SM. Thermal cracking of
hydrocarbons for the production of light olefins; a review on
optimal process design, operation, and control. Ind Eng
Chem Res 2020;59:12288e303. https://doi.org/10.1021/
Acknowledgments acs.iecr.0c00923.
[11] Liu S, Zhang B, Liu G. Metal-based catalysts for the non-
This research was supported in part by the National Research oxidative dehydrogenation of light alkanes to light olefins.
Council of Thailand (N41A640081 and N42A640324 for T.W. React Chem Eng 2021;6:9e26. https://doi.org/10.1039/
and M.C.), the Kasetsart University Research and Develop- D0RE00381F.
ment Institute (KURDI) through its program of Development of [12] Bawah AR, Lucky RA, Hossain MM. Oxidative dehydrogenation
of propane with CO2-A green process for propylene and
Advance Researcher Competence System for Competitiveness
hydrogen (syngas). Int J Hydrogen Energy 2021;46:3401e13.
in Agriculture and Food, and the Thailand Science Research https://doi.org/10.1016/j.ijhydene.2020.10.168.
and Innovation through its program of fundamental fund [13] Dang S, Li S, Yang C, Chen X, Li X, Zhong L, Gao P, Sun Y.
(FF(KU)21.65). Selective transformation of CO2 and H2 into lower olefins
over In2O3-ZnZrOx/SAPO-34 bifunctional catalysts.
ChemSusChem 2019;12:3582e91. https://doi.org/10.1002/
Appendix A. Supplementary data cssc.201900958.
[14] Liu X, Wang M, Yin H, Hu J, Cheng K, Kang J, Zhang Q,
Wang Y. Tandem catalysis for hydrogenation of CO and CO2
Supplementary data to this article can be found online at to lower olefins with bifunctional catalysts composed of
https://doi.org/10.1016/j.ijhydene.2021.10.265. spinel oxide and SAPO-34. ACS Catal 2020;10:8303e14.
https://doi.org/10.1021/acscatal.0c01579.
[15] Liu X, Wang M, Zhou C, Zhou W, Cheng K, Kang J, Zhang Q,
references Deng W, Wang Y. Selective transformation of carbon dioxide
into lower olefins with a bifunctional catalyst composed of
ZnGa2O4 and SAPO-34. Chem Commun 2018;54:140e3.
[1] Hepburn C, Adlen E, Beddington J, Carter EA, Fuss S, Mac https://doi.org/10.1039/C7CC08642C.
Dowell N, Minx JC, Smith P, Williams CK. The technological [16] Tada S, Kinoshita H, Ochiai N, Chokkalingam A, Hu P,
and economic prospects for CO2 utilization and removal. Yamauchi N, et al. Search for solid acid catalysts aiming at
Nature 2019;575:87e97. https://doi.org/10.1038/s41586-019- the development of bifunctional tandem catalysts for the
1681-6. one-pass synthesis of lower olefins via CO2 hydrogenation.
[2] Gabrielli P, Gazzani M, Mazzotti M. The Role of carbon Int J Hydrogen Energy 2021;46:36721e30. https://doi.org/
capture and utilization, carbon capture and storage, and 10.1016/j.ijhydene.2021.09.002.
biomass to enable a net-zero-CO2 emissions chemical [17] Wei C, Tu W, Jia L, Liu Y, Lian H, Wang P, Zhang Z. The
industry. Ind Eng Chem Res 2020;59:7033e45. https://doi.org/ evolutions of carbon and iron species modified by Na and
10.1021/acs.iecr.9b06579. their tuning effect on the hydrogenation of CO2 to olefins.
[3] Ferna  ndez JR, Garcia S, Sanz-Pe
rez ES. CO2 capture and Appl Surf Sci 2020;525:146622. https://doi.org/10.1016/
utilization editorial. Ind Eng Chem Res 2020;59:6767e72. j.apsusc.2020.146622.
https://doi.org/10.1021/acs.iecr.0c01643. [18] Dong Z, Zhao J, Tian Y, Zhang B, Wu Y. Preparation and
[4] Ra EC, Kim KY, Kim EH, Lee H, An K, Lee JS. Recycling carbon performances of ZIF-67-derived FeCo bimetallic catalysts for
dioxide through catalytic hydrogenation: recent key CO2 hydrogenation to light olefins. Catalysts 2020;10:455.
developments and perspectives. ACS Catal 2020;10:11318e45. https://doi.org/10.3390/catal10040455.
https://doi.org/10.1021/acscatal.0c02930. [19] Elishav O, Shener Y, Beilin V, Landau MV, Herskowitz M,
[5] Numpilai T, Kahadit S, Witoon T, Ayodele BV, Cheng CK, Siri- Shter GE, Grader GS. Electrospun Fe-Al-O nanobelts for
Nguan N, Sornchamni T, Wattanakit C, Chareonpanich M, selective CO2 hydrogenation to light olefins. ACS Appl Mater
Limtrakul J. CO2 hydrogenation to light olefins over In2O3/ Interfaces 2020;12:24855e67. https://doi.org/10.1021/
SAPO-34 and Fe-Co/K-Al2O3 composite catalyst. Top Catal acsami.0c05765.
2021;64:316e27. https://doi.org/10.1007/s11244-021-01412-5. [20] Witoon T, Lapkeatseree V, Numpilai T, Cheng CK,
[6] Numpilai T, Cheng CK, Limtrakul J, Witoon T. Recent Limtrakul J. CO2 hydrogenation to light olefins over mixed
advances in light olefins production from catalytic Fe-Co-K-Al oxides catalysts prepared via precipitation and
hydrogenation of carbon dioxide. Process Saf Environ Protect reduction methods. Chem Eng J 2022;428:131389. https://
2021;151:401e27. https://doi.org/10.1016/j.psep.2021.05.025. doi.org/10.1016/j.cej.2021.131389.
[7] Zhang Z, Yin H, Yu G, He S, Kang J, Liu Z, Cheng K, Zhang Q, [21] Chen J, Wang X, Wu D, Zhang J, Ma Q, Gao X, Lai X, Xia H,
Wang Y. Selective hydrogenation of CO2 and CO into olefins Fan S, Zhao TS. Hydrogenation of CO2 to light olefins on
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 4 2 1 8 5 e4 2 1 9 9 42199

CuZnZr@(Zn-)SAPO-34 catalysts: strategy for product carbon. Catal Sci Technol 2019;9:2728e41. https://doi.org/
distribution. Fuel 2019;239:44e52. https://doi.org/10.1016/ 10.1039/C9CY00403C.
j.fuel.2018.10.148. [34] Hwang S-M, Zhang C, Han SJ, Park H-G, Kim YT, Yang S,
[22] Numpilai T, Chanlek N, Poo-Arporn Y, Wannapaiboon S, Jun K-W, Kim SK. Mesoporous carbon as an effective support
Cheng CK, Siri-Nguan N, Sornchamni T, Kongkachuichay P, for Fe catalyst for CO2 hydrogenation to liquid hydrocarbons.
Chareonpanich M, Rupprechter G, Limtrakul J, Witoon T. J CO2 Util 2020;37:65e73. https://doi.org/10.1016/
Pore size effects on physicochemical properties of Fe-Co/K- j.jcou.2019.11.025.
Al2O3 catalysts and their catalytic activity in CO2 [35] Wu T, Lin J, Cheng Y, Tian J, Wang S, Xie S, Pei Y, Yan S,
hydrogenation to light olefins. Appl Surf Sci 2019;483:581e92. Qiao M, Xu H, Zong B. Porous graphene-confined FeK as
https://doi.org/10.1016/j.apsusc.2019.03.331. highly efficient catalyst for CO2 direct hydrogenation to light
[23] Wang B, Liang D, Guan Z, Li D, Zhang R. Understanding the olefins. ACS Appl Mater Interfaces 2018;10:23439e43. https://
key step of Co2C-catalyzed FischereTropsch synthesis. J Phys doi.org/10.1021/acsami.8b05411.
Chem C 2020;124:5749e58. https://doi.org/10.1021/ [36] Cheng Y, Lin J, Xu K, Wang H, Yao X, Pei Y, Yan S, Qiao M,
acs.jpcc.0c00611. Zong B. FischereTropsch synthesis to lower olefins over
[24] Numpilai T, Chanlek N, Poo-Arporn Y, Cheng CK, Siri- potassium-promoted reduced graphene oxide supported
Nguan N, Sornchamni T, Chareonpanich M, iron catalysts. ACS Catal 2016;6:389e99. https://doi.org/
Kongkachuichay P, Yigit N, Rupprechter G, Limtrakul J, 10.1021/acscatal.5b02024.
Witoon T. Tuning interactions of surface-adsorbed species [37] Galvis HMT, Bitter JH, Khare CB, Ruitenbeek M, Dugulan AI,
over Fe-Co/K-Al2O3 catalyst by different K contents: de Jong KP. Supported iron nanoparticles as catalysts for
selective CO2 hydrogenation to light olefins. sustainable production of lower olefins. Science
ChemCatChem 2020;12:3306e20. https://doi.org/10.1002/ 2012;335:835e8. https://doi.org/10.1126/science.1215614.
cctc.202000347. [38] Gu B, He S, Peron DV, Pedrolo DRS, Moldovan S, Ribeiro MC,
[25] Russkikh A, Shterk G, Al-Solami BH, Fadhel BA, Ramirez A, Lobato B, Chernavskii PA, Ordomsky VV, Khodakov AY.
Gascon J. Turning waste into value: potassium promoted red Synergy of nanoconfinement and promotion in the design of
mud as an effective catalyst for the hydrogenation of CO2. efficient supported iron catalysts for direct olefin synthesis
ChemSusChem 2020;13:2981e7. https://doi.org/10.1002/ from syngas. J Catal 2019;376:1e16. https://doi.org/10.1016/
cssc.202000242. j.jcat.2019.06.035.
[26] Witoon T, Chaipraditgul N, Numpilai T, Lapkeatseree V, [39] Liu J, Zhang A, Jiang X, Zhang G, Sun Y, Liu M, Ding F,
Ayodele BV, Cheng CK, Siri-Nguan N, Sornchamni T, Song C, Guo X. Overcoating the surface of Fe-based catalyst
Limtrakul J. Highly active Fe-Co-Zn/K-Al2O3 catalysts for CO2 with ZnO and nitrogen-doped carbon toward high
hydrogenation to light olefins. Chem Eng Sci 2021;233:116428. selectivity of light olefins in CO2 hydrogenation. Ind Eng
https://doi.org/10.1016/j.ces.2020.116428. Chem Res 2019;58:4017e23. https://doi.org/10.1021/
[27] Liu J, Zhang A, Jiang X, Liu M, Sun Y, Song C, Guo X. Selective acs.iecr.8b05478.
CO2 hydrogenation to hydrocarbons on Cu-promoted Fe- [40] Tian Z, Wang C, Si Z, Ma L, Chen L, Liu Q, Zhang Q, Huang H.
based catalysts: dependence on CueFe interaction. ACS Fischer-Tropsch synthesis to light olefins over iron-based
Sustainable Chem Eng 2018;6:10182e90. https://doi.org/ catalysts supported on KMnO4 modified activated carbon by
10.1021/acssuschemeng.8b01491. a facile method. Appl Catal 2017;541:50e9. https://doi.org/
[28] Xu Y, Zhai P, Deng Y, Xie J, Liu X, Wang S, Ma D. Highly 10.1016/j.apcata.2017.05.001.
selective olefin production from CO2 hydrogenation on iron [41] Asami K, Komiyama K, Yoshida K, Miyahara H. Synthesis of
catalysts: a subtle synergy between manganese and sodium lower olefins from synthesis gas over active carbon-
additives. Angew Chem Int Ed 2020;59:21736e44. https:// supported iron catalyst. Catal Today 2018;303:117e22.
doi.org/10.1002/anie.202009620. https://doi.org/10.1016/j.cattod.2017.09.010.
[29] Zhuo O, Yang L, Gao F, Xu B, Wu Q, Fan Y, Zhang Y, Jiang Y, [42] Warner JH, Scha € ffel F, Bachmatiuk A, Rümmeli MH. Chapter
Huang R, Wang X, Hu Z. Stabilizing the active phase of iron- 4dmethods for obtaining graphene. In: Warner JH,
based FischereTropsch catalysts for lower olefins: Scha € ffel F, Bachmatiuk A, Rümmeli MH, editors. Graphene:
mechanism and strategy. Chem Sci 2019;10:6083e90. https:// fundamentals and emergent applications. Amsterdam:
doi.org/10.1039/C9SC01210A. Elsevier Inc; 2013. p. 129e228.
[30] Jiang F, Zhang M, Liu B, Xu Y, Liu X. Insights into the [43] Xu Y, Cao H, Xue Y, Li B, Cai W. Liquid-phase exfoliation of
influence of support and potassium or sulfur promoter on graphene: an overview on exfoliation media, techniques, and
iron-based FischereTropsch synthesis: understanding the challenges. Nanomaterials 2018;8(11):942. https://doi.org/
control of catalytic activity, selectivity to lower olefins, and 10.3390/nano8110942.
catalyst deactivation. Catal Sci Technol 2017;7:1245e65. [44] Khandaker T, Hossain MS, Dhar PK, Rahman MS,
https://doi.org/10.1039/C7CY00048K. Hossain MA, Ahmed MB. Efficacies of carbon-based
[31] Chen Y, Wei J, Duyar MS, Ordomsky VV, Khodakov AY, Liu J. adsorbents for carbon dioxide capture. Processes 2020;8:654.
Carbon-based catalysts for fischeretropsch synthesis. Chem https://doi.org/10.3390/pr8060654.
Soc Rev 2021;50:2337. https://doi.org/10.1021/10.1039/ [45] Cossutta M, Vretenar V, Centeno TA, Kotrusz P, McKechnie J,
d0cs00905a. Pickering SJ. A comparative life cycle assessment of
[32] Zhang Z, Zhang J, Wang X, Si R, Xu J, Han Y-F. Promotional graphene and activated carbon in a supercapacitor
effects of multiwalled carbon nanotubes on iron catalysts for application. J Clean Prod 2020;242:118468. https://doi.org/
Fischer-Tropsch to olefins. J Catal 2018;365:71e85. https:// 10.1016/j.jclepro.2019.118468.
doi.org/10.1016/j.jcat.2018.05.021. [46] Meshkani F, Rezaei M. Preparation of mesoporous
[33] Tian Z, Wang C, Yue J, Zhang X, Ma L. Effect of a potassium nanocrystalline iron based catalysts for high temperature
promoter on the FischereTropsch synthesis of light olefins water gas shift reaction: effect of preparation factors. Chem
over iron carbide catalysts encapsulated in graphene-like Eng J 2015;260:107e16.

You might also like