You are on page 1of 12

Journal of Industrial and Engineering Chemistry 85 (2020) 196–207

Contents lists available at ScienceDirect

Journal of Industrial and Engineering Chemistry


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

The reaction kinetics of CO2 methanation on a bifunctional


Ni/MgO catalyst
A. Loder, M. Siebenhofer, S. Lux*
Institute of Chemical Engineering and Environmental Technology, NAWI Graz, Graz University of Technology, Graz, Austria

A R T I C L E I N F O A B S T R A C T

Article history: A bifunctional Ni/MgO catalyst was prepared to catalyze CO2 methanation and make use of CO2 as an
Received 23 September 2019 abundant hydrogen storage facility. The effect of Ni loading and MgO quality on the rate of methanation
Received in revised form 13 January 2020 was tested in a temperature range of 533–648 [156_TD$IF]K. The Ni loading was varied between 0 to 27 [109_TD$IF]wt.% on
Accepted 1 February 2020
MgO. To investigate the impact of matrix elements, a MgO/CaO support was tested with 21 [157_TD$IF]wt.% nickel
Available online 7 February 2020
loading. Further, the role of MgO in the bifunctional catalyst was proven. The reaction kinetics was
modeled with a Langmuir–Hinshelwood approach considering the bifunctional character of the
Keywords:
catalyst. Nickel provides the adsorbent capacity for hydrogen and is highly selective for methane. MgO
CO2 methanation
Ni/MgO catalyst
activates CO2 through chemisorption. Increasing Ni loading of the catalyst increased the rate of CO2
Kinetics conversion. According to the results, the mechanism of CO2 methanation did not change with Ni
Langmuir–Hinshelwood loading. The Ni/MgO catalyst acted as a robust, active and highly selective catalyst for CO2 methanation.
With CO2 conversion of 87%, the selectivity to methane was 99%. Besides excellent catalytic activity
the catalysts suffice the necessity of simple catalyst preparation, usage and recyclability for industrial
applicability of CO2 methanation.
© 2020 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All
rights reserved.

Introduction This paper focuses on the reaction kinetics of bifunctional


nickel/magnesium oxide catalysts to catalyze methane synthesis
Carbon dioxide (CO2) emissions have increased from 23.9 by CO2 hydrogenation according to Eq. [158_TD$IF]1 (methanation reaction).[159_TD$IF][60
billion metric tons in 1997 to 36.2 billion metric tons in 2017 [1].
CO2 + 4 H2 ? CH4 + 2 H2O DH0R, 298K = -165 kJ mol1 (1)
Industry is a major contributor, emitting over 30% of the global
greenhouse gas. Specific industrial sectors, e.g. the iron and steel CO2 rich off gas from ore calcination of iron carbonate (siderite)
industry or the magnesite industry, have a huge potential for the is a feasible CO2 source. In Austria about 0.6 [16_TD$IF]Mt CO2 per year are
reduction of highly concentrated CO2 off gas. Catalytic CO2 emitted during siderite ore calcination [13]. To integrate CO2
hydrogenation may significantly contribute to CO2 emission hydrogenation into the sintering process, hydrogenation must be
reduction. CO2 hydrogenation utilizes CO2 as a renewable carbon performed at ambient pressure and a feed gas ratio of H2:CO2 of 4:1
resource to produce carbon monoxide (CO) or (oxygenated) must be established. In an environmentally benign process the
hydrocarbons (CxHyOz) [2–5]. The hydrogenation product methane spent catalyst needs to be completely recyclable. Recycling of the
(CH4), for instance, can be integrated into existing gas grids for methanation catalyst in the steel converter closes the loop of iron
natural gas distribution and it may be used for chemical hydrogen ore beneficiation combined with catalytic CO2 hydrogenation. In
storage [6–10]. Hydrogen can be provided by renewable hydrogen this study, MgO is primarily used as a catalyst support as it is a
production routes, for instance, via utilization of surplus electricity common constituent of converter slag. Due to its basic properties
in water electrolysis or via biomass gasification [9,11,12]. MgO is expected to contribute to methane formation by adsorptive
interaction with CO2, and thus supports hydrogenation of CO2 at
ambient pressure.
The product selectivity for methane formation from CO2
depends on several factors, i.e., type of catalyst, process tempera-
* Corresponding author at: Institute of Chemical Engineering and Environmental
ture, operation pressure, and reactant gas composition (H2:CO2
Technology, NAWI Graz, Graz University of Technology, Inffeldgasse 25C/II, 8010
Graz, Austria. ratio) [14,15]. CO2 hydrogenation is assumed to be a dual step
E-mail address: susanne.lux@tugraz.at (S. Lux). process via CO formation in the first step and methane formation in

https://doi.org/10.1016/j.jiec.2020.02.001
1226-086X/© 2020 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
A. Loder et al. / Journal of Industrial and Engineering Chemistry 85 (2020) 196–207 197

the second step [16]. The intermediate CO is formed through Nickel/magnesium oxide catalysts are easy to synthesize, and
reverse water gas shift (RWGS) reaction according to Eq. [162_TD$IF]2. the catalysts are robust. Nickel and MgO are cheap compared to
other catalyst materials for CO2 methanation such as rhodium and
CO2 + H2 ? CO + H2O (2)
ruthenium [23,39]. Magnesium oxide has been suggested as a
However, some studies have also assumed that CO and CH4 are support for methanation catalysts mainly in combination with
the products of the parallel reverse water gas shift (Eq. [96_TD$IF]2) and other support materials [25,49–53]. Bette et al. [25] used 59 [167_TD$IF]wt.%
methanation reaction [14]. nickel on (Mg,Al)Ox. They used 0.08 [168_TD$IF]g catalyst in a fixed bed reactor
Fig. 1 shows the standard Gibbs free energy of reaction DG R with a diameter of 0.004 [169_TD$IF]m with H2:CO2:Ar:N2 [170_TD$IF]= 18:4.6:12.8:64.1
and the standard enthalpy of reaction DH R of CO2 methanation and a feed flow rate of 66 [17_TD$IF]m3 kg1 h1. CO2 conversion reached a
and CO2 hydrogenation to CO in the temperature range of 300– maximum of 74% between 603 and 623 [172_TD$IF]K and ambient pressure.
1300 [164_TD$IF]K. The standard Gibbs free energy of reaction of CO2 Guo et al. [49] used 10 [173_TD$IF]wt.% Ni/SiO2 on MgO in a fixed bed reactor
methanation rises with rising temperature (Fig. 1a) [13]. Below with 0.008 [174_TD$IF]m diameter with 0.2 g catalyst at 0.0048 m3 [175_TD$IF]h1
856 [165_TD$IF]K, CO2 methanation is thermodynamically favored over CO (24 m3 kg1 h1) feed flow rate with a H2:CO2 molar ratio of 4:1 at
formation. For temperatures beyond 856 [165_TD$IF]K, CO formation is atmospheric pressure. CO2 conversion was 66.5% and CH4
thermodynamically favored. For CO2 hydrogenation to CO, DG R selectivity was 96.8% at 673 [176_TD$IF]K.
slightly decreases with rising temperature. The standard enthal- Compared to other support materials, MgO may also act as an
pies of reaction of CO2 methanation and CO2 hydrogenation to CO active compound of the catalyst. Because of its basicity MgO
decrease with increasing temperature (Fig. 1b). CO2 methanation is adsorbs carbon dioxide. It further reduces catalyst deactivation
exothermic in the whole temperature range while CO2 hydrogena- such as sintering and carbon formation [49,50,61]. In general, CO2
tion to CO is endothermic. According to the thermodynamics the methanation catalysts are susceptible to deactivation by water. In
temperature range for the experiments in this study was kept well the CO2 methanation reaction two moles of water are formed for
below 700 [16_TD$IF]K to favor CO2 methanation over CO formation. every mole of methane. MgO reduces the negative impact of water
CO2 methanation catalysts need to provide high activity at low on the catalyst. MgO reacts with water to magnesium hydroxide
temperature to avoid carbon monoxide formation. As CO2 (Mg(OH)2) [62]. Magnesium hydroxide adsorbs CO2 through
methanation is highly exothermic, a heat recovery unit should chemisorption [63]. The standard Gibbs free energy of H2O
be implemented in the process [7]. The methanation catalyst needs chemisorption on MgO is exergonic below 550 [17_TD$IF]K. The standard
to be robust, especially against sintering and carbon deposition on Gibbs free energy of CO2 chemisorption on Mg(OH)2 is 26 to 19 [178_TD$IF]
catalytically active sites [18–20]. For industrial application the kJ (molCO2)1 between 500 and 700 [179_TD$IF]K. Compared with Mg(OH)2
catalyst has to be easy to synthesize and recyclable. Industrial DG R for the chemisorption of CO2 on MgO is higher at temper-
implementation also prefers a minimum stoichiometric ratio of atures beyond 550 [180_TD$IF]K, as can be seen in Fig. 2. As a consequence,
H2:CO2 of 4:1 and operation close to ambient pressure. chemisorption of CO2 on Mg(OH)2 is thermodynamically more
Catalytic CO2 methanation has been investigated intensively favorable over adsorption on MgO and thus, the reaction is
with a variety of different catalysts [21]. Nickel [8,16,22–34], assumed to be less affected by water deactivation when MgO is
rhodium [35–37], and ruthenium [38–41] are among the most used as catalyst support.
promising catalytically active constituents. Nickel is commonly CO2 hydrogenation to methane with Ni/MgO catalysts has only
used because it is the cheapest of these active compounds. Nickel been investigated by few researchers so far. Li et al. [64] used nano
adsorbs hydrogen and it is highly selective for methane formation NiO-MgO catalyst particles encapsulated by silica in a fixed bed
[19,23,42–44]. Al2O3 [16,23,27–33,45–48], MgO [25,49–53], ZrO2 reactor with a diameter of 0.006 [18_TD$IF]m. They used 0.05 g of the catalyst
[8,30,31,34,54,55], TiO2 [30], CeO2 [30,31,47,56] and SiO2 [22,57– with 59.4 [182_TD$IF]wt.% Ni with a feed flow rate of 0.0054 m3 [183_TD$IF]h1
60], as well as different combinations of the mentioned (108 m3 kg1 h1) of H2:CO2:N2 [184_TD$IF]= 4:1:4 (volumetric ratio) and
constituents [23–26] have been proposed and investigated as atmospheric pressure and achieved 87% CO2 conversion and 99%
catalyst support. CH4 selectivity at 573 [185_TD$IF]K. Nakayama et al. [65] used 0.15 [186_TD$IF]g of a 70 wt.%

[(Fig._1)TD$IG]

Fig. 1. Standard Gibbs free energy of reaction DG R (a) and standard enthalpy of reaction DH R (b) for CO2 methanation (Eq. [96_TD$IF]1) and CO2 hydrogenation to CO (Eq. [97_TD$IF]2) between
300 and 1300 K, data calculated with HSC Chemistry 8 [17].
198 A. Loder et al. / Journal of Industrial and Engineering Chemistry 85 (2020) 196–207
[(Fig._2)TD$IG]
The findings published herein are based on a previous study on
Ni/MgO catalysts of Baldauf-Sommerbauer et al. [72]. The authors
investigated 11 and 17 [195_TD$IF]wt.% Ni-based catalysts on a MgO support in
a fixed bed reactor with a diameter of 0.025 [196_TD$IF]m and a catalyst
amount of 12 [197_TD$IF]g. At 648 K and ambient pressure, a flow rate of 1.24
m3 [198_TD$IF]kg1 h1 and a volumetric ratio of H2:CO2 of 4:1, CO2 conversion
of 80% (equilibrium conversion: 84%) and CH4 selectivity of >99%
were achieved. For 17 [195_TD$IF]wt.% Ni/MgO, 83.8% CO2 conversion
(equilibrium conversion: 90%) and >99% CH4 selectivity were
achieved at 598 [19_TD$IF]K. At 533 K and 548 K, carbon monoxide was
formed as a byproduct. The higher the contact time of the catalyst
and the gas flow was, the lower was the CO content. The long-term
stability of the Ni/MgO catalyst (17 [20_TD$IF]wt.% Ni) at 603 K was confirmed
for nine days (210 [201_TD$IF]h of operation), in which steady-state
methanation was performed on three days (more than 18 [20_TD$IF]h).
When industrial application is intended, long-term stability of the
catalyst does not only refer to its ability for long-term steady-state
operation but also to its robustness. In this context, the term
‘robust’ means stable CO2 conversion and CH4 selectivity during
repeated operation/down time cycles of the reactor because of
time dependent supply of renewably produced hydrogen. In the
experiments, neither CO2 conversion nor CH4 selectivity decreased
from start-up to shut-down of the reactor, a period that included
Ni/MgO catalyst in a fixed bed reactor with 0.018 [187_TD$IF]m diameter. H2: cooling and heating of the catalyst in between steady-state
CO2 was 8:1 and the volumetric flow rate was 0.0108 [18_TD$IF]m3 h1 methanation sequences. Thus, long-term stability of the catalyst
(72 m3 kg1 h1). At 553 K and atmospheric pressure, CO2 conver- with respect to changing temperature and changing atmosphere
sion of 85% and 100% selectivity to CH4 were achieved. Yan et al. [42] was proven. Moreover, crystalline, graphitic carbon was not
investigated W-doped Ni/MgO catalysts in a fixed bed reactor at deposited on the catalyst as the XRD pattern of the fresh and
atmospheric pressure. They used 0.1 [120_TD$IF]g of the catalyst with a Ni to W the spent catalyst did not vary.
ratio of 1, and 0.006 [189_TD$IF]m3 h1 (60 m3 kg1 h1) feed flow rate with H2: Based on these results CO2 methanation with Ni/MgO catalysts
CO2 [190_TD$IF]= 4:1 in He and achieved 83% CO2 conversion and 99% CH4 with 11–27 [203_TD$IF]wt.% Ni loading at ambient pressure and a temperature
selectivity at 573 [185_TD$IF]K. Takezawa et al. [52] showed the influence of the range between 533 and 648 [204_TD$IF]K was investigated. The aim of this
reduction temperature on the catalytic activity of Ni/MgO catalysts work was to optimize CO2 conversion by increasing the Ni content
at atmospheric pressure. They applied a 0.006 [19_TD$IF]m3 h1 feed flow rate of the catalyst. Compared to previous studies [42,52,64,65], higher
with a H2:CO2 partial pressure ratio of 0.95:0.05. Conversion was amounts of catalyst of 4–12 [205_TD$IF]g were used. The flow rate of the feed
highest for catalysts prepared at 873 [192_TD$IF]K, while methane selectivity gas and the reaction temperature were varied. The effect of Ni
was not influenced by the reduction temperature. A CH4 selectivity loading and different MgO quality on the activity of the catalyst
of over 97% was achieved for 13 [193_TD$IF]wt.% Ni on MgO at 512 K. was investigated to address the bifunctionality of the catalyst. CO2
The studies show that Ni/MgO is an active catalyst system for conversion close to equilibrium conversion was achieved. The feed
CO2 methanation. However, most data were obtained for small gas composition was kept constant. For comparison, the influence
catalyst amounts (up to 0.15 [194_TD$IF]g), which makes scale up for industrial of the feed gas composition on conversion and CH4 selectivity was
application difficult. studied. From CO2 conversion results the rate of conversion was
The kinetics of CO2 methanation with Ni on different support determined. To quantify the effect of Ni loading on CO2 conversion,
materials have been reported in the literature. Different models have an appropriate rate law had to be developed. Therefore, a simple
been applied. Table 1 provides a summary of some of the kinetic reversible reaction rate approach was investigated and compared
models. The rate law is preferably expressed by a Langmuir– with a Langmuir–Hinshelwood approach. With the assumption of
Hinshelwood based approach [16,32,66–69]. Power-law approaches mechanism validity over the whole temperature range as well as
have also been suggested [70,71]. To the best of our knowledge, no the investigated range of Ni loading the frequency factor of the
systematic kinetic studies of CO2 methanationwith Ni/MgO catalysts Arrhenius law was expected to quantify the effect of Ni loading on
with different Ni loading have been conducted so far. the rate of CO2 conversion.

Table 1
Representative kinetic models for CO2 methanation with nickel catalysts, as suggested in literature; the temperature and pressure range are mentioned.

Catalyst Temperature / K Pressure / bar Rate law model Literature


Ni/SiO2 550–591 7–17 Power-law Chiang & Hopper (1983) [70]
Ni/SiO2 500–600 1.4 Langmuir–Hinshelwood Weatherbee & Bartholomew (1982) [68]
Ni/Al(O)x 523–613 1–9 Langmuir–Hinshelwood-Hougen– Koschany et al. (2016) [16]
Watson approach
commercial, Ni/Al2O3 673 1.013 Langmuir–Hinshelwood approach Rönsch et al. (2016) [66]
for CO methanation combined with
rate model for water gas shift
reaction
Ni/Al2O3 623–723 1.013 Langmuir–Hinshelwood approach Champon et al. (2019) [69]
Ni/Al2O3 443–483 10 Langmuir–Hinshelwood approach Yang Lim et al. (2016) [32]
commercial, Ni 473–503 1.013 Langmuir approach Van Herwijnen et al. (1973) [67]
Ni/yttria stabilized zirconia 573–623 1.013 Power-law Kesavan et al. (2018) [71]
A. Loder et al. / Journal of Industrial and Engineering Chemistry 85 (2020) 196–207 199

Experimental Catalyst specification

Catalyst preparation According to Baldauf-Sommerbauer et al. [72] the applied


preparation method achieves uniform nickel distribution on the
The Ni/MgO and Ni/MgO/CaO catalysts were prepared via wet MgO particles forming a solid Ni-MgO solution in the catalyst.
impregnation with nickel nitrate hexahydrate (Ni(NO3)26 H2O, To quantify the nickel loading, the catalysts were analyzed with
99%, p.a., Lactan). MgO source was either pure MgO or (for atomic absorption spectrometry (AAS) with an AAnalyst400
comparison) a MgO/CaO ore. The pure MgO support was (Perkin Elmer) device equipped with a nickel hollow cathode
MagGran© (4 MgCO3Mg(OH)24 H2O, ph. eur., Magnesia AG, lamp set to 25 [23_TD$IF]mA current at a wavelength of 232 [24_TD$IF]nm, applying a
Switzerland) spherically granulated, with a grain size distribution compressed air/ethylene flame. Prior to analysis, the solid catalyst
of <150 m [206_TD$IF] m (0–8 wt.%), 150–250 mm (0–15 wt.%) and 250–600 mm samples were dissolved in a solution of 10 [25_TD$IF]cm3 HNO3 (p.a., J.T.
(55–80 wt.%). To suffice the intention of industrial applicability a Baker) and 990 [26_TD$IF]cm3 deionized H2O.
MgO/CaO ore (Breitenau, Austria) was used as catalyst support too. The surface area of the 20 and 27 [109_TD$IF]wt.% Ni/MgO catalysts and the
It contained 55 [207_TD$IF]wt.% MgO and 39 wt.% CaO with grain size 200– 21 [157_TD$IF]wt.% Ni/MgO/CaO catalyst were determined via 5-point BET
500 m [208_TD$IF] m. Table 2 shows the composition of the MgO/CaO mixture. method between 0.05 and 0.3 [27_TD$IF]p/p* with an ASAP2000 from
The preparation method for the catalysts, based on the work of Micromeritics. The catalyst samples were evacuated before
Baldauf-Sommerbauer et al. [72], was optimized. It consisted of measurement to remove excess water.
four steps: preparation of magnesium oxide granulate (a),
impregnation (b), thermal decomposition (c), and reduction in CO2 methanation experiments
hydrogen atmosphere (d).
(a) The MgO granulate (or MgO/CaO ore) was calcined in a For the CO2 methanation experiments, carbon dioxide [28_TD$IF]
muffle furnace (Heraeus M110) in air for 5 [209_TD$IF]h at 723 K and for 2 h at (99.998%), hydrogen (99.999%), and nitrogen (99.999%) supplied
823 K. by Air Liquide were used. Nitrogen was used as an inert gas for heat
(b) Then, 10 g of the calcined, MgO granulate (white) was mixed transport and balancing purposes.
with 70 [210_TD$IF]cm3 of an aqueous nickel nitrate solution. The MgO The experiments were conducted in a fixed bed stainless steel
granulate was put in a glass beaker that contained the nickel (T316) tubular reactor (Parr Instrument GmbH) with 0.82 [29_TD$IF]m length
nitrate solution (c(Ni) [21_TD$IF]= 50 g dm3 (for 11 wt.% Ni), 100 g dm3 (for and 25 mm inner diameter. Fig. 3 shows the reactor setup. The
17 wt.% Ni), 53 g dm3 (for 20 wt.% Ni), 60 g dm3 (for 27 wt.% Ni)). reactor set up is also described in detail in Baldauf-Sommerbauer
The flask was cooled in a water bath and constantly stirred. The et al. [72,73]. The feed gas streams (CO2, H2, N2) were fed to the
calcined MgO granulate was added in 7–10[21_TD$IF] intervals over 1 h. The reactor tube from top via three mass flow controllers (MFC). The
slurry was stirred for additional 2 [213_TD$IF]h and afterwards filtrated by a feed gas passed a pre-heating coil (PHC) at the top of the reactor
vacuum pump to remove the residual water phase. The catalyst tube before entering the catalyst bed. The reactor tube was heated
precursor (green) was then dried at room temperature for 24 [214_TD$IF]h. with an electric furnace, which allowed heating in three heating
(c) The catalyst precursor was dried in a muffle furnace zones. The temperature was measured in the middle of each
(Heraeus M110) at 393 [215_TD$IF]K for 2 h and calcined at 673 K for 5 h in heating zone at the outer wall of the reactor tube (HT1–HT3) and at
static air. It changed its color from green to light gray. two positions within each heating zone inside the reactor tube via
(d) The calcined, light gray NiO/MgO powder was reduced with thermocouples (T1–T6).
hydrogen in the tubular reactor (Fig. 3) at 773 [216_TD$IF]K (temperature in Four stainless steel spacers helped place the catalyst granulate
the mid of the fixed bed) with a flow rate of 0.03 [217_TD$IF]m3 h1 (H2: bed inside the reactor tube. The catalyst bed formed a hollow
N2 = 9:1) for 4 h. At this temperature, reduction of NiO to Ni is cylinder inside the reactor tube with the thermocouple string in
complete (DG0R [218_TD$IF]=  45.6 kJ mol1). the center of the tube (diameter of the thermocouple casing: 6 [230_TD$IF]
Five Ni-based catalysts were tested for their CO2 hydrogenation mm). To prevent spilling of the catalyst granulate, glass wool tori
efficiency: four Ni/MgO catalysts with 11 [219_TD$IF]wt.% Ni (Ni11/MgO) [72], (height:10 [231_TD$IF]mm) were placed above and below the catalyst bed.
17 [195_TD$IF]wt.% Ni (Ni17/MgO) [72], 20 [20_TD$IF]wt.% Ni (Ni20/MgO), 27 [109_TD$IF]wt.% Ni The mass of the catalyst granulate was 12 [23_TD$IF]g for 11 and 17 wt.%
(Ni27/MgO) on pure MgO and one Ni/MgO/CaO catalyst (Ni21/ nickel loading on pure MgO (Ni11/MgO, Ni17/MgO). The height of
MgO/CaO) with 21 [157_TD$IF]wt.% nickel loading on a CaO/MgO support. the catalyst bed was 39 [23_TD$IF]mm in these experiments. For the Ni-based
The effect of the MgO support on the activity of Ni/MgO catalysts with 20 and 27 [109_TD$IF]wt.% Ni loading on pure MgO (Ni20/MgO,
catalysts was also tested. For this purpose, the calcined pure MgO Ni27/MgO) and for the Ni-based catalyst on the MgO/CaO mixture
granulate was directly kept in H2 atmosphere for 3 [21_TD$IF]h at 773 K and (Ni21/MgO/CaO), the mass of the catalyst was 8 [234_TD$IF]g (height of
0.03 m3 h1 flow rate and H2:N2 [2_TD$IF]= 9:1. catalyst bed: 26 [235_TD$IF]mm). To study the effect of the MgO support, 4 [236_TD$IF]g
MgO were used (height of the catalyst bed: 13 [237_TD$IF]mm) (Table 3).
In all experiments the catalyst bed was placed in a way that a
thermocouple was located at the end of the catalyst bed (= [238_TD$IF]Tend,cat)
Table 2 (Fig. 3). For the experiments with Ni11/MgO and Ni17/MgO the
Composition of the MgO/CaO ore from Breitenau, Austria (200–500 [1_TD$IF]mm).
catalyst was placed above T4, for those with Ni20/MgO, Ni27/MgO
Content and Ni21/MgO/CaO the catalyst was placed above T3 according to
[12_TD$IF]/ wt.% Fig. 3.
MgO 54.6 All experiments were performed at ambient pressure. The
CaO 39.23 residence time of the feed gas for the five Ni/MgO catalysts was
Fe2O3 4.85
between 0.7 and 4.5 [239_TD$IF]s, depending on the feed flow rate (1.24–14.93 [240_TD$IF]
MnO 0.63
SO3 0.29 m3 kg1 h1). For the pure MgO support the residence time was
Al2O3 0.17 between 3.1 and 12.6 [241_TD$IF]s (feed flow rate 3.7–14.8 [24_TD$IF]m3 kg1 h1). The
SiO2 0.13 volumetric H2:CO2 ratio in the feed gas was kept at 4:1 with 30 [243_TD$IF]vol.
Na2O 0.07
% nitrogen as inert (calibration) gas. Two additional feed gas ratios
TiO2 0.02
P2 O 5 0.01
(H2:CO2 [24_TD$IF]= 3:1 and 5:1) were tested for the Ni27/MgO catalyst.
200 A. Loder et al. / Journal of Industrial and Engineering Chemistry 85 (2020) 196–207
[(Fig._3)TD$IG]

Fig. 3. Scheme of the experimental setup of the fixed bed tubular reactor for CO2 methanation with different Ni/MgO and Ni/MgO/CaO catalysts; MFC . . . mass flow
controller, T1–T6 . . . thermocouple inside reactor tube, HT1–HT3 . . . temperature measurement position in the middle of a heating zone, PHC . . . pre-heating coil,
HE . . . heat exchanger, CT . . . condensate tank, BPR . . . back pressure regulator, GA . . . gas analyzer.

Table 3
Operation conditions and catalyst specification of the Ni/MgO and Ni/MgO/CaO catalysts and the MgO support for CO2 methanation; ambient pressure, feed gas composition:
H2:CO2 [13_TD$IF]= 4:1 (vol).

Catalyst MgO source Surface area Catalyst Feed gas flow rate Tend,cat
[14_TD$IF]/ m2 g1 amount / g [15_TD$IF]/ m3 kg1 h1 (STP) /K
Ni11/MgO [72] pure MgO 58 12 1.24–5 533–648
Ni17/MgO [72] pure MgO 53 12 1.24–5 533–648
Ni20/MgO pure MgO 67 4 3.7–14.8 533–648
Ni27/MgO pure MgO 56 4 3.7–14.8 533–648
Ni21/MgO/CaO MgO/CaO 8 8 1.7–6.8 533–648
MgO support pure MgO – 4 3.7–14.8 668–748

Condensibles were removed from the gaseous product stream in 648 [25_TD$IF]K) that was measured at Tend,cat. The criterion for ‘steady-state
a Peltier cooler (HE1, T [245_TD$IF]= 273 K at the outlet) and collected in a tank conditions’ was a standard deviation of T[256_TD$IF]end,cat within 0.8% of the
(CT1) at the outlet of the reactor tube. The condensate was analyzed mean value of T[257_TD$IF]end,cat, and within 1.5% for the dry product gas
with a total organic carbon (TOC) analyzer (Shimadzu TOC-L CPH) composition.
for condensable byproducts of CO2 methanation. The composition The steady-state experiments were used to determine the
of the gaseous product stream was monitored online. The reaction kinetics of CO2 methanation on Ni/MgO catalysts under
continuous gas analyzer (GA) consisted of a Caldos27 thermal isothermal conditions. T3 was set as target temperature for
conductivity analyzer and a Uras26 infrared photometer from ABB. ‘isothermal’ CO2 methanation with Ni20/MgO, Ni27/MgO and
Caldos27 monitored the hydrogen concentration (measurement Ni21/MgO/CaO. As already discussed in the previous work of
ranges: 0–0.5 [246_TD$IF]vol.% and 0–100 vol.%; output error (2 [247_TD$IF]s): 0.5% of Baldauf-Sommerbauer et al. [72], the temperatures above (T4) as
smallest measurement range span) and Uras26 monitored the CO2, well as the outer wall temperature (HT2) deviate from the
CO and CH4 concentration (measurement ranges: 0–10 [248_TD$IF]vol.% and 0– temperature at the end of the catalyst bed (T3). The temperature
100 vol.%; output error (2 [247_TD$IF]s): 0.2% of span). The gas analyzer was variation depends on the type of catalyst, feed flow rate, and
frequently recalibrated with three different standard gas mixtures heating procedure. Therefore, in catalyst studies of exothermic
of CO, CO2, CH4, H2, and N2 supplied by Air Liquide. To prevent reactions, it is crucial to state the temperature measurement
further condensation in the gas analyzer, a second cooler (HE2, T [249_TD$IF] position and to place the thermocouple that measures the ‘catalyst
= 269 K) and a tank (CT2) were installed prior to the gas analyzer. temperature’ at the highest temperature position. In the work of
According to Baldauf-Sommerbauer et al. [250_TD$IF][72] two experimen- Baldauf-Sommerbauer et al. [72] and in this work, the temperature
tal procedures were used to determine the catalytic performance of CO2 methanation was measured at the end of the catalyst bed
of Ni/MgO. Temperature scanning experiments were performed (T4 and T3, respectively). The highest temperature was always at
preliminary to steady-state experiments for every catalyst with the the end of the catalyst bed. The temperature measurement as well
purpose of gaining a first insight into the temperature dependency as the fact that the XRD pattern of the fresh and the spent catalyst
of the CO2 conversion and product selectivity (gas feed of H2:CO2: did not differ indicate that hot spot formation did not occur.
N2 [251_TD$IF]= 56:14:30 (vol) and 0.06 [25_TD$IF]m3 h1 (STP)). After initial constant To calculate the molar composition, ideal gas mixing was
heating of 25–30 [253_TD$IF]min, the heating power of each heating zone assumed in the gas stream. Moreover, it was assumed that there
was adapted manually, so that the average temperature gradient were no other constituents in a concentration above 0.1 [258_TD$IF]vol.% in the
was 1.9  [254_TD$IF]0.1 K min1 at Tend,cat. From the temperature scanning dry product gas apart from N2, CO, CO2, CH4, and H2.
experiments the temperature range for the steady-state experi- The catalytic activity of the Ni-based catalysts was discussed
ments for kinetic analysis was deduced. The temperature range for based on CO2 conversion X CO2 and CH4 selectivity SCH4 derived from
the catalytic activity tests of the Ni-based catalysts was chosen dry product gas composition. CO2 conversion was calculated with
between 533 and 648 [204_TD$IF]K [72]. For steady-state experiments, the Eq. [259_TD$IF]3 and CH4 selectivity with Eq. [260_TD$IF]4. H2 conversion was calculated
reactor was heated to the target temperature (533, 563, 598, and analogous to the CO2 conversion. Due to the high methane
A. Loder et al. / Journal of Industrial and Engineering Chemistry 85 (2020) 196–207 201

selectivity above 90%, carbon monoxide formation was neglected the surface area of 8 [278_TD$IF]m2 g1 was significantly below the surface
and the kinetic evaluation was based on carbon dioxide area of the Ni/MgO catalysts. Although cheaper, the MgO/CaO ore is
methanation. Methane selectivity was around 90% for CO2 therefore not a recommended catalyst support.
conversions below 20% with the Ni11/MgO catalyst. In all other
cases it was higher (99%).[26_TD$IF][3 Effect of nickel loading and reaction temperature
cCO2 ;0  cCO2
X CO2 ¼ ð3Þ The CO2 conversion depends on the feed gas composition, the
cCO2 ;0 þ eCO2 cCO2
reaction temperature, and the nickel loading. Fig. 4 shows the
influence of the reaction temperature (533–648 [279_TD$IF]K) and the nickel
cCH4 loading (11–27 [280_TD$IF]wt.%) of the Ni/MgO catalysts on CO2 conversion
SCH4 ¼ ð4Þ (also see Table A1 in the supplementary information). For all
cCH4 þ cCO
temperatures, the conversion increased linearly with increasing Ni
ci;0 is the initial concentration of component i and ci is the content. At 598 [281_TD$IF]K, the CO2 conversion increased from 45.5% with
concentration of component i at the outlet of the reactor (dry Ni11/MgO to 70.0% (Ni17/MgO), 79.9% (Ni20/MgO) and 91.2% with
product gas composition). The volumetric expansion coefficient Ni27/MgO. At higher temperatures, the increase was less distinct.
eCO2 was calculated by Eq. [264_TD$IF]5 with cCO2 ;0 as the feed CO2 Gac et al. [74] also observed a gradual increase in CO2 conversion
concentration and the stoichiometric coefficients a, b, c, and d with an increase of Ni loading for alumina supported nickel
(CO2: a ¼ 1, H2: b ¼ 4, CH4: c ¼ 1, and H2O: d [265_TD$IF]= 2). catalysts. At 553 [28_TD$IF]K, the CO2 conversion increased from 40% with
  10 [173_TD$IF]wt.% Ni on Al2O3 to 80% with 40 [283_TD$IF]wt.% Ni/Al2O3. This effect was
c d b
eCO2 ¼ yCO2 ;0  þ  1 ð5Þ less evident for ceria- and zirconia-supported nickel catalysts
jaj jaj jaj which was explained by the changes in the active surface area. For
The concentration of the constituents was monitored at the alumina-supported nickel catalysts, an almost linear increase of
outlet of the reactor during the whole experimental run. The exit CO2 conversion with an increase of the active surface area was
concentration of the inert gas nitrogen cN2 was calculated found. In case of ceria- and zirconia-supported nickel catalysts, this
according to Eq. [26_TD$IF]6.[267_TD$IF] was less evident due to a more pronounced increase in nickel
crystallite size with increasing nickel content.
cN2 ;0
cN2 ¼ ð6Þ The equilibrium CO2 conversion decreases with rising temper-
1 þ eCO2 X CO2
atures. For any of the tested catalysts, CO2 conversion rose with
With the flow rate and the monitored concentration values the rising temperature until equilibrium conversion was obtained.
mass balance was checked. The relative error did not exceed 1% in Higher nickel loading increased the rate of CO2 methanation
any experiment. fostering the assumption that the availability of adsorbed
To double-check accuracy, the exit hydrogen concentration cH2 hydrogen is limiting the reaction rate [7,43,45].
was also calculated with Eq. [268_TD$IF]7 and the concentration of the Table 4 lists CO2 conversion and CH4 selectivity for various
products cP (P [269_TD$IF]= CO, CH4) was calculated with Eq. [270_TD$IF]8, with c ðor dÞ Ni/MgO catalysts from literature compared to the 27 [109_TD$IF]wt.% Ni/MgO
catalyst (Ni27/MgO) of our work.
representing the stoichiometric coefficient of the constituents, and [(Fig._4)TD$IG]
compared to the measured values.[271_TD$IF][

cH2 ;0  bX CO2 cCO2 0


cH2 ¼ ð7Þ
1 þ eCO2 X CO2

[273_TD$IF]4
c ðor dÞX CO2 cCO2 ;0
cP ¼ ð8Þ
1 þ eCO2 X CO2

Results and discussion

The Ni/MgO catalysts effectively catalyzed CO2 methanation. All


four Ni/MgO catalysts (11,17, 20 and 27 [109_TD$IF]wt.% Ni) were highly selective
for methane. For conversions over 50%, CH4 selectivity was over 99%.
At lower conversions [275_TD$IF](<20% with Ni11/MgO) CH4 selectivity was
>90%. The TOC analysis (detection limit [276_TD$IF]= 4 mg dm3 C) showed that
the condensibles contained no carbon-based compounds. The
condensate consisted of water only; no condensible hydrocarbons
were formed, confirming that except methane and CO (concentra-
tion below detection limit) no further products were formed.

Surface area of the catalysts

The surface areas of the Ni/MgO catalysts were between 53–67 [27_TD$IF]
m2 g1 (Table 3). This minor difference in the surface area is dating
back to the mixing intensity during catalyst preparation. It was also
stated by Gac et al. [74] that the nickel content on various support
Fig. 4. Effect of the temperature and nickel loading on CO2 methanation with Ni/
materials (Al2O3, ZrO2, and CeO2) did not have a strong influence on MgO catalysts for different Ni loading with H2:CO2 [9_TD$IF]= 4:1 (3.7 m3 kg1 h1) compared
the specific surface area and the porosity of the respective to the equilibrium CO2 conversion level (the equilibrium data were calculated with
catalysts. For the catalyst with 21 [157_TD$IF]wt.% Ni on the MgO/CaO support HSC Chemistry 8 [17]).
202 A. Loder et al. / Journal of Industrial and Engineering Chemistry 85 (2020) 196–207
[(Fig._5)TD$IG]
Influence of the feed gas composition (H2:CO2 ratio)

From literature [8,42] it was deduced that the availability of


adsorbed hydrogen is a limiting factor for the rate of reaction. To
investigate this effect, the H2:CO2 ratio in the feed gas was varied
from 3:1 to 5:1 for the Ni27/MgO catalyst. The effect on CO2
methanation was determined at 598 [284_TD$IF]K. The experiments showed
that CO2 conversion rises with rising hydrogen concentration in
the feed gas stream. The maximum CO2 conversion of 98%
(equilibrium conversion: 99.8%) was obtained for a H2:CO2 ratio of
5:1. Fig. 5 shows the effect of the feed gas composition on CO2
conversion. The corresponding equilibrium conversion is 70.0% for
a H2:CO2 ratio of 3:1 and 91.6% for a H2:CO2 ratio of 4:1 (feed flow
rate: 3.7 [104_TD$IF]m3 kg1 h1).

Role of the catalyst support MgO

Pure MgO was tested for its role in the catalyst. A temperature
scanning experiment was performed to specify the temperature
range for the experiments. The temperature range was chosen
from 670 to 750 [285_TD$IF]K for steady-state experiments in the setup shown Fig. 5. CO2 conversion for different H2:CO2 ratios in the feed gas (3.7 [10_TD$IF]m3 kg1 h1) at
in Fig. 6. 598 K with the Ni27/MgO catalyst ([10_TD$IF]SCH4 [102_TD$IF]99%).
Fig. 6 shows the CO2 conversion for different feed flow rates
(3.7–14.8 [286_TD$IF]m3 kg1 h1) when pure MgO is used. MgO is catalytically area of the catalysts”, the surface area of the catalyst with Ni on pure
active for CO2 hydrogenation. However, the catalyst is selective to MgO is much higher than for Ni on MgO/CaO.
carbon monoxide; methane is not formed. CO2 conversion with The CH4 concentration in the product gas stream correlates well
MgO rises with rising temperature. The highest CO concentration with the CO2 conversion for the two different catalysts. Both catalysts
of 21.4% in the product gas stream is achieved at 748 [287_TD$IF]K (feed flow were selective to methane ([10_TD$IF]SCH4 [29_TD$IF]> 99%). The CH4 concentration was
rate: 3.7 [104_TD$IF]m3 kg1 h1). The MgO support facilitates the reverse higher for the catalyst with pure MgO as support than with the
water gas shift reaction as already stated by Baldauf–Sommerba- catalyst with MgO/CaO over a temperature range of 533–648 [293_TD$IF]K. At
uer et al. [73]. As H2 does not dissociate to H* on MgO [44], CO2 is 598 K, the maximum methane concentration was 14.7 [294_TD$IF]vol.% for
seemingly converted with gaseous H2 according to Eq. [28_TD$IF]2 and nickel on pure MgO and 12.4 [295_TD$IF]vol.% for nickel on MgO/CaO.
carbon monoxide is formed. Methane was not detected (detection
limit (4s): [289_TD$IF]0.4 % of span; 0–10 vol.% and 0–100 vol.%). Reaction kinetics of CO2 methanation

Impact of matrix elements of the MgO support In this study the target of industrial applicability of CO2
methanation limited investigations to ambient pressure and the
To gain information about the impact of matrix elements of the catalyst support MgO. Both boundary conditions suffice the
MgO support on the catalytic activity of Ni/MgO catalysts, a second necessity of simple catalyst preparation, usage and recyclability.
MgO support was investigated. The second MgO source was a According to the experimental design the reaction rate rCO2
mixture of MgO/CaO (dolomite ore with 54.6 [290_TD$IF]wt.% MgO and 39.23 wt. was deduced from Eq. [296_TD$IF]9.
% CaO). The Ni loading on the MgO/CaO support was 21 [157_TD$IF]wt.%. The
catalytic activity was compared to the catalyst with dX CO2
rCO2 ¼   ð9Þ
20 [20_TD$IF]wt.% nickel loading on pure MgO. Fig. 7 compares CO2 conversion d Fm cat
CO ;O
of the two catalysts for reaction temperatures between 533 and 648 [291_TD$IF]
2

K. At the same temperature CO2 conversion was much higher for the X CO2 is the CO2 conversion, mcat the mass of the catalyst and F CO2 O is
catalyst with pure MgO. The result confirms that MgO is an active the feed flow rate of CO2. For each catalyst the model parameters
compound in the catalyst. In the investigated temperature range CaO were fitted to the experimental data at different temperatures.
however does not participate actively in the reaction. This may be Validity of the model parameters was tested by comparing the
dedicated to the fact that the thermal stability of CaCO3 is much calculated conversion X CO2 with the experimental results by least
higher than the stability of MgCO3. As mentioned in Section “Surface squares minimization.

Table 4
Comparison of CO2 conversion and CH4 selectivity for Ni/MgO catalysts from the literature compared to the Ni27/MgO catalyst; ambient pressure.

Catalyst Temperature / K Feed gas flow rate / H2:CO2 CO2 conversion / % CH4 selectivity / % Literature
m3 [16_TD$IF]h1 (STP) (equilibrium conversion)
(m3 kg1 h1)
4 [17_TD$IF]g 27 wt.% Ni/MgO 598 0.015 (3.7) 4:1 87 (0.90)  99 our work
0.059 (14.7) 81 (0.90)
0.05 [18_TD$IF]g nano NiO-MgO 573 0.0054 (108) 4:1 87 (0.93) 99 Li et al. [64]
0.15 [19_TD$IF]g Ni/MgO 553 0.0108 (72) 8:1 85 (0.94) 100 Nakayama et al. [65]
0.1 [120_TD$IF]g W-doped Ni/MgO 573 0.006 (60) 4:1 83 (0.93) 99 Yan et al. [42]
Ni/MgO 512 0.006 0.95:0.05 35 (0.97) 97 Takezawa et al. [52]
0.08 Ni/(Mg,Al)Ox 623 0.004 (50) 18:4.6 74 (0.87) Bette et al. [25]
0.2 [12_TD$IF]g MgO on Ni/SiO2 673 0.0048 (24) 4:1 66.5 (0.80) 96.8 Guo et al. [49]
A. Loder et al. / Journal of Industrial and Engineering Chemistry 85 (2020) 196–207 203
[(Fig._6)TD$IG]
Koschany et al. [16] give an excellent example for developing a
multistep reaction mechanism for CO2 methanation on a Ni/Al(O)x
catalyst. They tested two reaction pathways, both with nine
reaction steps. One mechanism considers the cleavage of the
carbon-oxygen bond first, and the second mechanism considers a
formyl species intermediate.
Different to the work of Koschany et al. our investigations were
carried out at ambient pressure and constant stoichiometric ratio
of H2:CO2 of 4:1. These conditions provide industrially relevant
operation conditions. Further, ambient pressure and a constant
ratio of H2 to CO2 allow for simplifications when developing the
reaction mechanism for CO2 methanation on our Ni/MgO catalysts.
In a first step, the rate data were analyzed for a reversible reaction
without considering species adsorption/desorption. Eq. [107_TD$IF]10 gives a
simple power-law, assuming a quasi-homogenous reversible reac-
tionwith kfor being the reaction rate constant of the forward reaction
and krev the reaction rate constant of the reverse reaction. Different
reaction orders (nCO2, nH2, nH2O, nCH4) were considered. The simple
reversible reaction power-law was not able to fit the experimental
data accordingly (see Fig. 8, exemplarily given for nCO2, nH2, nH2O,
nCH4 = 1), suggesting that species adsorption and desorption cannot
be neglected. The model was also tested when only low conversions
were considered. Nevertheless, it did not succeed in describing the
rate of reaction appropriately.[297_TD$IF][830
dX CO2
rCO2 ¼  ¼ kfor cCO2 nCO2 cH2 nH2  krev cCH4 nCH4 cH2 O nH2 O ð10Þ
dt
As the previous analysis of simple power-law correlations
Fig. 6. Temperature dependency of CO2 hydrogenation with the pure MgO catalyst
for different feed flow rates of 3.7–14.8 [103_TD$IF]m3 kg1 h1 (H2:CO2 = 4:1).
indicated the necessity to take species adsorption/desorption on/
from the catalyst into account, Langmuir–Hinshelwood based rate
[(Fig._7)TD$IG] laws were tested in a next step. The Langmuir–Hinshelwood
reaction mechanism considers CO2 and H2 adsorption on the
catalyst. This supports the bifunctional catalytic effect of the Ni/
MgO catalyst. The mechanism considers five steps: (1) dissociation
[(Fig._8)TD$IG]

Fig. 7. CO2 methanation with Ni-based catalysts on MgO with different MgO
quality, Ni with pure MgO (Ni/MgO) at 3.7 [104_TD$IF]m3 kg1 h1 feed flow rate and Ni on a
mixture of MgO/CaO (Ni/MgO/CaO) at 3.4 [105_TD$IF]m3 kg1 h1 feed flow rate with H2:CO2 [106_TD$IF]
= 4:1, the equilibrium data were calculated with HSC Chemistry 8 [13].

In the literature, the rate laws that are suggested for modeling of
CO2 methanation with different catalysts are preferably [8_TD$IF]Langmuir–
Hinshelwood based models. In this study, a simple power-law
approach for reversible reactions and a Langmuir–Hinshelwood Fig. 8. Model validation for the power-law correlation given in Eq. [107_TD$IF]10 (nCO2, nH2,
based rate law were compared and discussed. nH2O, nCH4 [108_TD$IF]= 1); lines: model, data points: experimental data, exemplarily given for
Ni27/MgO.
204 A. Loder et al. / Journal of Industrial and Engineering Chemistry 85 (2020) 196–207

and adsorption of H2 on nickel (Eq. [301_TD$IF]11) and (2) adsorption of carbon The total amount of active sites on the catalyst is given in Eq. [321_TD$IF]21
dioxide on the Ni-MgO interface. Adsorbed CO2 is attached to two for magnesium (Mgt) and for nickel in Eq. [32_TD$IF]22 (Nit). It is assumed
active Mg sites and one active nickel site, as shown by Huang et al. that the amount of catalyst sites occupied by hydrogen and carbon
[44] (Eq. [302_TD$IF]12). In literature, the adsorption of hydrogen on nickel is dioxide is low when the rate of the methanation reaction is high, so
described to be rate limiting on Ni-based catalysts [8,42]. On cHNi and cCO2 ðMgÞ2 Ni  are negligible for fast methanation. The amount
bifunctional Ni/MgO catalysts, both hydrogen (on Ni) and CO2 (on of occupied nickel sites by methane cCH4 Ni ,[32_TD$IF] is also low as the
the Ni-MgO interface) are adsorbed. Thus, the availability of active desorption of methane is fast. Thus, cCH4 Ni is negligible. Inserting of
nickel sites plays a crucial role and species adsorption on Ni in Eq. [324_TD$IF]20 into Eq. 21 and rearranging it gives Eq. [325_TD$IF]23 for the number of
general is assumed to be rate controlling in this work. With the unoccupied catalyst sites Mg.
basic aim of developing a simple rate law with adequate accuracy
for industrial application in mind, hydrogen and CO2 adsorption Mgt ¼ Mg þ cCO2 ðMgÞ2 Ni þ  cH2 OMg ð21Þ
were combined assuming that both contribute to the rate limiting
step (Eq. [30_TD$IF]13). Next (3) adsorbed carbon dioxide reacts with
adsorbed hydrogen to methane and H2O (Eq. [304_TD$IF]14). According to Nit ¼ Ni þ cHNi þ cCO2 ðMgÞ2 Ni þ  cCH4 Ni ð22Þ
Huang et al. [44], methane is adsorbed on one nickel site and water
[326_TD$IF]
on one magnesium site (Eqs. [305_TD$IF](15) and (16)) [16,44]. Finally, steps
(4) [306_TD$IF]+ (5) are the desorption of the reaction products methane (Eq. [307_TD$IF] Mgt
Mg ¼   ð23Þ
15) and water (Eq. 16) from the Ni/MgO catalyst.

k8
  cH 2 O
k7
H2 þ 2 Ni Ð 2 HNi ð11Þ
With the assumption that Eq. [327_TD$IF]17 is rate controlling, Eq. [328_TD$IF]24 is
obtained from the previous equations (Eqs. [329_TD$IF]18–20,22,23)[30_TD$IF][124
CO2 þ 2 Mg þ  Ni Ð CO2 ðMgÞ2 Ni ð12Þ     

k1 cH2 cCO2 Mgt  Nit  k2 k4k6k8 cCH4 cH2 O Mg t  Ni2t


dcCO2 k3k5k7
 ¼  ð24Þ
dt 1 þ k8cH2 O
 k7
k1
H2 þ CO2 þ 2 Mg þ 3 Ni  Ð  2 HNi þ CO2 ðMg Þ2 Ni ð13Þ

k2
By inserting the derivative dX CO2 (Eq. [35_TD$IF]25) into Eq. 24 the rate law in
Eq. 26 is obtained.[36_TD$IF]
dNCO2
 rCO2 ¼  ð25Þ
k3 mcat dt
CO2 ðMgÞ2 Ni þ 8 HNi  Ð

 CH4 Ni þ 2 H2 OMg þ 8 Ni ð14Þ
k4
[37_TD$IF]89401
   

k1 cH2 cCO2 Mg t  Nit    k2k4k6 k8Mgt  Nit cCH4 cH2 O
2
 k3  k5  k7
k5 rCO2 ¼  ð26Þ
CH4 Ni  Ð

 CH4 þ Ni ð15Þ 1 þ cH2 O
k8

k6 k7

Eq. 26 may then be simplified to Eq. [342_TD$IF]27 that represents the



overall rate law for CO2 methanation with Ni/MgO catalysts.[34_TD$IF][12567
k7
H2 OMg  Ð  H2 O þ Mg ð16Þ

k8 kfor cCO2 cH2  krev cCH4 cH2 O
rCO2 ¼ ð27Þ
1 þ K H2 O cH2 O
For simplification, either methane desorption or water desorption
were considered as rate affecting. Both approaches were tested for The species reaction orders in the rate law achieved the best
applicability. The following derivation of the rate law is exemplarily compliance with experiments for nCO2, nH2, nH2O, nCH4 [34_TD$IF]= 1.
shown for a pronounced effect of water desorption and negligible Desorption control was tested for CH4 desorption and water
impact of methane desorption on the overall rate of reaction. desorption with Ki being the desorption constant of the respective
According to the following reaction equations, the rate of species i (H2O in Eq. [345_TD$IF]27). CH4 desorption control resulted in less
adsorption of hydrogen and carbon dioxide is given by Eq. [308_TD$IF]17. The compliance with experiments than water desorption. The effect of
rate of methanation is represented by Eq. [309_TD$IF]18. The desorption of hydrogen, methane and water concentration on the model was
methane is represented by Eq. 19 and The desorption of water is tested. The Langmuir–Hinshelwood mechanism assumes that
represented by Eq. [310_TD$IF]20.[31_TD$IF][2 chemisorbed CO2 and H2 react and that the products are adsorbed
  after the reaction. This is in accordance with the reaction
dcCO2
 ¼ k1 cH2 cCO2 Mg Ni  k2 cHNi cCO2 ðMgÞ2 Ni ð17Þ mechanism of Koschany et al. [16] and Huang et [346_TD$IF]al. [60]. The
dt
reaction rate rises with rising H2 concentration and drops for rising
[314_TD$IF]567 water concentration. The corresponding rate law, depicted in Eq. [347_TD$IF]
dcCO2 ðMgÞ2 Ni   27, fits the experimental data for all experiments with Ni/MgO
 ¼ k3 cCO2 ðMgÞ2 Ni cHNi  k4 cCH4 Ni cH2 OMg Ni ð18Þ catalysts for the whole temperature range (533–648 [348_TD$IF]K).
dt
[318_TD$IF]
Table 5
dcCH4 Ni  
Activation energies Ea for the Langmuir–Hinshelwood rate law (Eq. [10_TD$IF]27).
 ¼ k5 cCH4 Ni  k6 cCH4 Ni ð19Þ
dt
Activation energy
[319_TD$IF]20 [12_TD$IF]/ kJ mol1

  forward reaction 99
dcH2 OMg
 ¼ k7 cH2 OMg  k8 cH2 O Mg ð20Þ reverse reaction 145
dt desorption term 109
A. Loder et al. / Journal of Industrial and Engineering Chemistry 85 (2020) 196–207 205
[(Fig._9)TD$IG]

Fig. 9. Frequency factors for the forward (Afor), reverse (Arev) reaction and the desorption term (AH2O) for Ni/MgO catalysts with Ni loading from 11 to 27 [109_TD$IF]wt.%.

[(Fig._10)TD$IG]

Fig. 10. Model validation for CO2 methanation based on the Langmuir–Hinshelwood based mechanism (Eq. [10_TD$IF]27).
206 A. Loder et al. / Journal of Industrial and Engineering Chemistry 85 (2020) 196–207

Table 6
Comparison of different rate laws for CO2 methanation with the corresponding activation energies and their interpretation given in the literature.

Rate law Activation energy / Comment


[123_TD$IF]kJ mol1
Our work dX CO2 kfor cCO2 cH2 krev cCH4 cH2 O Eaf or = 99 retarding impact of H2O
dt
¼ 1þK H2 O cH2 O
Earev = 145
EaH2 O = 109
[129_TD$IF]Chiang & Hopper (1983) [70] r ¼ kpH2 0:21 pCO2 0:66 61
94
Weatherbee & Bartholomew (1982) [68] K 1 K 2 K 10 k4 k11 0:5 2 0:5
L pCO p0:5

2 2 H2
0:5 2
2K k 0:5 pCO K K K k 0:5 p
1þK K 2 k4     0:52 þ 1 22k10 11 p0:5
CO
p0:5
H
þ KCO
1 10 11 p 4 2 2 3
H2

Koschany et al. (2016) [16] kpH 0:5


pCO 0:5
1
pCH pH O 2
4 2 Eak = 78
r¼ pH O
2 2 pH 4 pCO K eq
2 2
2
DHK OH [14_TD$IF]= 22
1þK OH 20:5 þK H2
pH
2
pH
2
0:5 þK
mix pCO2
0:5
DHK H2   = -6
DHK mix   = -10
Rönsch et al. (2016) [66] kK C K H 2 pCO 0:5 pH þkK C K H 2 pCH pH p0:5 p2 1
Eak = 103
2 O CO
r¼ 2 4 H2 K Meth

1þK C pCO 0:5 þK H2 pH


2
0:5 3
DHK C = -42
DHK H   = -16
DHK CO   = -71
DHK H2   = -83
DHK CH4   = -38
DHK CO   = 89
 
Champon et al. (2019) [69] pCH p2
4 H2 O
Eak [148_TD$IF]= 110
kK CO2 K H2 pCO pH 1
2 2 p4 pCO K eq
DHK CO2 [149_TD$IF]= 10
r¼ 2
H2 2

1þK CO2 pCO þK H2 pH þK H2 O pH


2 2 2O
þK CO pCO DHK H2   = 52
DHK H2 O   = 15
DHK CO   = 41
Yang Lim et al. (2016) [32] k1 p0:5 p0:5 Eak1 [150_TD$IF]= 95 CO dissociation limiting
r¼ 2
CO H 2 2

1þk2 p0:5 þk3 p0:5 p0:5 þk4 pH DHk2 = -46


H2 CO2 H 2O
DHk4 = -32
2

Van Herwijnen et al. (1973) [67] kpCO Eak [153_TD$IF]= 109


r¼ 2
1þK CO2 pCO
2 DHK CO2 = -18
[154_TD$IF]Kesavan et al. (2018) [71] rHP ¼ k∙pH2 0:1 ∙pCO2 1:01 99 different rate laws for high partial pressure
rLP ¼ kpH2 0:69 pCO2 0:69 (HP) and low partial pressure (LP) of the reactants

From the respective rate constants kfor(T), krev(T) and KH2O(T) at The catalyst was 99% selective to CH4 at 87% CO2 conversion. High
T [349_TD$IF]= 533, 563, 598 and 648 K, the temperature influence of the Ni loading and high hydrogen concentration in the feed gas
forward reaction (Eafor, Afor), the reverse reaction (Earev, Arev) and increased CO2 conversion. Within the investigated range of Ni
the desorption step (EaH2O, AH2O) were modeled with an Arrhenius loading, the catalytic activity increased linearly with the Ni
approach (Eq. [350_TD$IF]28; activation energy Ea, frequency factor A).[351_TD$IF] loading. Reaction kinetics was successfully modeled with a
Langmuir–Hinshelwood based rate law taking the bifunctional
kðT Þ ¼ Aexpð RT Þ  
Ea
ð28Þ catalytic action of the catalyst into account. Water showed a
The activation energies Ea are listed in Table 5. The activation retarding effect on the rate of reaction. The activation energy was
energy for the forward reaction is Eafor [352_TD$IF]= 99 kJ mol1, for the reverse constant for all Ni/MgO catalysts. The frequency factors showed a
reaction Earev is 145 [35_TD$IF]kJ mol1 and EaH2O is 109 [354_TD$IF]kJ mol1. The linear correlation with the Ni loading. The activity of Ni-doped
activation energies are the same for all tested catalysts with nickel catalysts depends on the number of available catalytically active
loading from 11 to 27 [109_TD$IF]wt.%, suggesting that different Ni loading sites. Preparation of Ni-doped MgO by wet impregnation provided
does not have an effect on the reaction mechanism. a very high number of active catalyst sites even at high Ni loading
Fig. 9 shows the relationship between the nickel loading and the of 27 [109_TD$IF]wt.%.
frequency factors Afor, Arev and AH2O for Ni loading from 11 to 27 [109_TD$IF]wt.%.
The rate law proposed by Eq. [342_TD$IF]27 is able to fit all experimentally Conflict of interest
determined CO2 conversion data for 11–27 [203_TD$IF]wt.% nickel loading on
MgO with a correlation coefficient of 99%. None.
Fig. 10 shows the model validation for all Ni/MgO (11–27 [280_TD$IF]wt.%
Ni) catalysts. The model also meets the conversion limit imposed Acknowledgments
by the chemical equilibrium. Following the outcome of data
analysis with Eq. [347_TD$IF]27, the frequency factors of the rate of CO2 This research did not receive any specific grant from funding
conversion with Ni/MgO catalysts linearly correlate with the Ni agencies in the public, commercial, or not-for-profit sectors.
loading. Water has a retarding impact on the rate of conversion. Special thanks go to Georg Baldauf-Sommerbauer for his valuable
This outcome agrees well with the literature [16,32,66–71] work in our group on the topic of CO2 methanation with Ni/MgO
(Table 6). catalysts [72].

Appendix A. Supplementary data


Conclusion
Supplementary material related to this article can be found, in the
The effect of different Ni loading (11–27 [280_TD$IF]wt.%) on the perfor- online version, at doi:https://doi.org/10.1016/j.jiec.2020.02.001.
mance of a Ni/MgO catalyst for CO2 methanation was investigated.
A. Loder et al. / Journal of Industrial and Engineering Chemistry 85 (2020) 196–207 207

References [35] C. Swalus, M. Jacquemin, C. Poleunis, P. Bertrand, P. Ruiza, Appl. Catal. B


Environ. 125 (2012) 41, doi:http://dx.doi.org/10.1016/j.apcatb.2012.05.019.
[1] Global Carbon Project, Historical carbon dioxide emissions from global fossil [36] M. Younas, S. Sethupathi, L.L. Kong, A.R. Mohamed, Int. J. Energy Res. 42 (10)
fuel combustion and industrial processes from 1757 to 2017 (in million metric (2018) 3289, doi:http://dx.doi.org/10.1002/er.4082.
tons), Statista, https://www.statista.com/statistics/264699/worldwide-co2- [37] H. Arandiyan, K. Kani, Y. Wang, B. Jiang, J. Kim, M. Yoshino, M. Rezaei, A.E.
emissions/. Rowan, H. Dai, Y. Yamauchi, ACS Appl. Mater. Interfaces 10 (30) (2018) 24963,
[35_TD$IF][2] IPCC, in: O. Edenhofer, R. Pichs-Madruga, Y. Sokona, E. Farahani, S. Kadner, K. doi:http://dx.doi.org/10.1021/acsami.8b06977.
Seyboth, A. Adler, I. Baum, S. Brunner, P. Eickemeier, B. Kriemann, J. Savolainen, [38] L. Falbo, M. Martinelli, C.G. Visconti, L. Lietti, C. Bassano, P. Deiana, Appl. Catal.
S. Schlömer, C. von Stechow, T. Zwickel, J.C. Minx (Eds.), Climate Change 2014: B Environ. 225 (2018) 354, doi:http://dx.doi.org/10.1016/j.apcatb.2017.11.066.
Mitigation of Climate Change. Contribution of Working Group III to the Fifth [39] A. Petala, P. Panagiotopoulou, Appl. Catal. B Environ. 224 (2018) 919, doi:http://
Assessment Report of the Intergovernmental Panel on Climate Change, dx.doi.org/10.1016/j.apcatb.2017.11.048.
Cambridge University Press, Cambridge, United Kingdom and New York,[357_TD$IF] NY, [40] J. Xu, Q. Lin, X. Su, H. Duan, H. Geng, Y. Huang, Chinese J. Chem. Eng. 24 (1)
USA, 2014. (2016) 140, doi:http://dx.doi.org/10.1016/j.cjche.2015.07.002.
[3] E.S. Gnanakumar, N. Chandran, I.V. Kozhevnikov, A. Grau-Atienza, E.V. Ramos [41] M.S. Duyar, A. Ramachandran, C. Wang, R.J. Farrauto, J. CO2 Util. 12 (2015) 27,
Fernández, A. Sepulveda-Escribano, N.R. Shiju, Chem. Eng. Sci. 194 (2019) 2, doi:http://dx.doi.org/10.1016/j.jcou.2015.10.003.
doi:http://dx.doi.org/10.1016/j.ces.2018.08.038. [42] Y. Yan, Y. Dai, H. He, Y. Yu, Y. Yang, Appl. Catal. B Environ. 196 (2016) 108, doi:
[4] Global Carbon Project, Global CO2 emissions from 1997 to 2017 (in billion http://dx.doi.org/10.1016/j.apcatb.2016.05.016.
metric tons), Statista, https://www.statista.com/statistics/276629/global-co2- [43] B. Miao, S.S.K. Ma, X. Wang, H. Su, S.H. Chan, Catal. Sci. Technol. 6 (12) (2016)
emissions/. 4048, doi:http://dx.doi.org/10.1039/C6CY00478D.
[361_TD$IF][5] G. Centi, S. Perathoner, Catal. Today 148 (3–4) (2009) 191, doi:http://dx.doi. [44] J. Huang, X. Li, X. Wang, X. Fang, H. Wang, X. Xu, J. CO2 Util. 33 (2019) 55, doi:
org/10.1016/j.cattod.2009.07.075. http://dx.doi.org/10.1016/j.jcou.2019.04.022.
[6] A. Maroufmashat, M. Fowler, Energies 10 (8) (2017), doi:http://dx.doi.org/ [45] C. Liang, H. Tian, G. Gao, S. Zhang, Q. Liu, D. Dong, X. Hu, Int. J. Hydrogen Energy
10.3390/en10081089. 45 (1) (2020) 531, doi:http://dx.doi.org/10.1016/j.ijhydene.2019.10.195.
[7] I. García–García, U. Izquierdo, V.L. Barrio, P.L. Arias, J.F. Cambra, Int. J. Hydrogen [46] Z. Zhang, Y. Tian, L. Zhang, S. Hu, J. Xiang, Y. Wang, L. Xu, Q. Liu, S. Zhang, X. Hu,
Energy 41 (43) (2016), doi:http://dx.doi.org/10.1016/j.ijhydene.2016.04.010. Int. J. Hydrogen Energy 44 (18) (2019) 9291, doi:http://dx.doi.org/10.1016/j.
[8] X. Jia, X. Zhang, N. Rui, X. Hu, C. jun Liu, Appl. Catal. B Environ. 244 (2019) 159, ijhydene.2019.02.129.
doi:http://dx.doi.org/10.1016/j.apcatb.2018.11.024. [47] C. Italiano, J. Llorca, L. Pino, M. Ferraro, V. Antonucci, A. Vita, Appl. Catal. B
[9] G. Reiter, J. Lindorfer, J. CO2 Util. 10 (2015) 40, doi:http://dx.doi.org/10.1016/j. Environ. 264 (2020), doi:http://dx.doi.org/10.1016/j.apcatb.2019.118494.
jcou.2015.03.003. [48] M. Wolf, L.H. Wong, C. Schüler, O. Hinrichsen, J. CO2 Util. 36 (2020) 276, doi:
[10] M. Thema, F. Bauer, M. Sterner, Renew. Sustain. Energy Rev (2019) 775, doi: http://dx.doi.org/10.1016/j.jcou.2019.10.014.
http://dx.doi.org/10.1016/j.rser.2019.06.030. [49] M. Guo, G. Lu, Catal. Commun. 54 (2014) 55, doi:http://dx.doi.org/10.1016/j.
[11] I. Dincer, C. Acar, Int. J. Hydrogen Energy 40 (34) (2014) 11094, doi:http://dx. catcom.2014.05.022.
doi.org/10.1016/j.ijhydene.2014.12.035. [50] J.N. Park, E.W. McFarland, J. Catal. 266 (1) (2009) 92, doi:http://dx.doi.org/
[12] C. Acar, I. Dincer, Int. J. Hydrogen Energy 39 (1) (2014) 1, doi:http://dx.doi.org/ 10.1016/j.jcat.2009.05.018.
10.1016/j.ijhydene.2013.10.060. [51] H.Y. Kim, H.M. Lee, J. Park, J. Phys, Chem. C 114 (2010) 7128, doi:http://dx.doi.
[13] G.F. Baldauf-Sommerbauer, Reductive calcination of mineral iron carbonate org/10.1021/jp100938v.
and mineral magnesium carbonate Graz University of Technology, (2017) . [52] N. Takezawa, H. Terunuma, M. Shimokawabe, H. Kobayashi, Appl. Catal. 23
[14] M.M. Jaffar, M.A. Nahil, P.T. Williams, Energy Technol. 7 (11) (2019)1900795, (1986) 291.
doi:http://dx.doi.org/10.1002/ente.201900795. [53] M.A. Arellano-Treviño, Z. He, M.C. Libby, R.J. Farrauto, J. CO2 Util. 31 (2019) 143,
[15] S. Valinejad Moghaddam, M. Rezaei, F. Meshkani, R. Daroughegi, Int. J. doi:http://dx.doi.org/10.1016/j.jcou.2019.03.009.
Hydrogen Energy 43 (34) (2018) 16522, doi:http://dx.doi.org/10.1016/j. [54] L. Hu, A. Urakawa, J. CO2 Util. 25 (2018) 323, doi:http://dx.doi.org/10.1016/j.
ijhydene.2018.07.013. jcou.2018.03.013.
[16] F. Koschany, D. Schlereth, O. Hinrichsen, Appl. Catal. B Environ. 181 (2016) 504, [55] S. Li, G. Liu, S. Zhang, K. An, Z. Ma, L. Wang, Y. Liu, J. Energy Chem. 43 (2020) 155,
doi:http://dx.doi.org/10.1016/j.apcatb.2015.07.026. doi:http://dx.doi.org/10.1016/j.jechem.2019.08.024.
[17] [362_TD$IF]Outotec HSC Chemistry Software, ([36_TD$IF][942018) . [56] A. Kokka, T. Ramantani, A. Petala, P. Panagiotopoulou, Catal. Today (2019), doi:
[18] X. Su, J. Xu, B. Liang, H. Duan, B. Hou, Y. Huang, J. Energy Chem. 25 (4) (2016) http://dx.doi.org/10.1016/j.cattod.2019.04.015.
553, doi:http://dx.doi.org/10.1016/j.jechem.2016.03.009. [57] J.B. Branco, P.E. Brito, A.C. Ferreira, Chem. Eng. J. 380 (2020), doi:http://dx.doi.
[19] J. Gao, Q. Liu, F. Gu, B. Liu, Z. Zhong, F. Su, RSC Adv. 5 (29) (2015) 22759, doi: org/10.1016/j.cej.2019.122465.
http://dx.doi.org/10.1039/C4RA16114A. [58] Y.R. Dias, O.W. Perez-Lopez, Energy Convers. Manag. 203 (2020), doi:http://dx.
[20] I. Graça, L.V. González, M.C. Bacariza, A. Fernandes, C. Henriques, J.M. Lopes, M. doi.org/10.1016/j.enconman.2019.112214.
F. Ribeiro, Appl. Catal. B Environ. 147 (2014) 101, doi:http://dx.doi.org/10.1016/ [59] R.P. Ye, W. Gong, Z. Sun, Q. Sheng, X. Shi, T. Wang, Y. Yao, J.J. Razink, L. Lin, Z.
j.apcatb.2013.08.010. Zhou, H. Adidharma, J. Tang, M. Fan, Y.G. Yao, Energy 188 (2019), doi:http://dx.
[21] I. Sreedhar, Y. Varun, S.A. Singh, A. Venugopal, B.M. Reddy, Catal. Sci. Technol. 9 doi.org/10.1016/j.energy.2019.116059.
(17) (2019) 4478, doi:http://dx.doi.org/10.1039/c9cy01234f. [60] X. Huang, P. Wang, Z. Zhang, S. Zhang, X. Du, Q. Bi, F. Huang, New J. Chem. 43
[22] T.A. Le, J.K. Kang, E.D. Park, Top. Catal 61 (15–17) (2018) 1537, doi:http://dx.doi. (33) (2019) 13217, doi:http://dx.doi.org/10.1039/c9nj03152a.
org/10.1007/s11244-018-0965-7. [61] M.S. Duyar, S. Wang, M.A. Arellano-Treviño, R.J. Farrauto, J. CO2 Util. 15 (2016)
[23] J. Lin, C. Ma, Q. Wang, Y. Xu, G. Ma, J. Wang, H. Wang, C. Dong, C. Zhang, M. Ding, 65, doi:http://dx.doi.org/10.1016/j.jcou.2016.05.003.
Appl. Catal. B Environ. 243 (2019) 262, doi:http://dx.doi.org/10.1016/j. [62] G.C. Bond, S.P. Sarsam, Appl. Catal. 38 (1988) 365.
apcatb.2018.10.059. [63] J.C. Fisher, R.V. Siriwardane, Energy and Fuels 28 (9) (2014) 5936, doi:http://dx.
[24] J. Ashok, M.L. Ang, S. Kawi, Catal. Today 281 (2017) 304, doi:http://dx.doi.org/ doi.org/10.1021/ef500841h.
10.1016/j.cattod.2016.07.020. [64] Y. Li, G. Lu, J. Ma, RSC Adv. 4 (34) (2014) 17420, doi:http://dx.doi.org/10.1039/
[25] N. Bette, J. Thielemann, M. Schreiner, F. Mertens, ChemCatChem 8 (18) (2016) C3RA46569A.
2903, doi:http://dx.doi.org/10.1002/cctc.201600469. [65] T. Nakayama, N. Ichikuni, S. Sato, F. Nozaki, Appl. Catal. A Gen. 158 (1–2) (1997)
[26] P.A. Ussa Aldana, F. Ocampo, K. Kobl, B. Louis, F. Thibault-Starzyk, M. Daturi, P. 185, doi:http://dx.doi.org/10.1016/S0926-860X(96)00399-7.
Bazin, S. Thomas, A.C. Roger, Catal. Today 215 (2013) 201, doi:http://dx.doi.org/ [66] S. Rönsch, J. Köchermann, J. Schneider, S. Matthischke, Chem. Eng. Technol. 39
10.1016/j.cattod.2013.02.019. (2) (2016) 208, doi:http://dx.doi.org/10.1002/ceat.201500327.
[27] D. Beierlein, D. Häussermann, M. Pfeifer, T. Schwarz, K. Stöwe, Y. Traa, E. [67] T. Van Herwijnen, H. Van Doesburg, W.A. De Jong, J. Catal. 28 (3) (1973) 391,
Klemm, Appl. Catal. B Environ. 247 (2018) 200, doi:http://dx.doi.org/10.1016/j. doi:http://dx.doi.org/10.1016/0021-9517(73)90132-2.
apcatb.2018.12.064. [68] G.D. Weatherbee, C.H. Bartholomew, J. Catal. 77 (1982) 460.
[28] G. Garbarino, P. Riani, L. Magistri, G. Busca, Int. J. Hydrogen Energy 39 (22) [69] I. Champon, A. Bengaouer, A. Chaise, S. Thomas, A.C. Roger, J. CO2 Util. 34 (2019)
(2014) 11557, doi:http://dx.doi.org/10.1016/j.ijhydene.2014.05.111. 256, doi:http://dx.doi.org/10.1016/j.jcou.2019.05.030.
[29] G. Garbarino, C. Wang, T. Cavattoni, E. Finocchio, P. Riani, M. Flytzani- [70] J.H. Chiang, J.R. Hopper, Ind. Eng. Chem. Prod. Res. Dev. 22 (2) (1983) 225–228,
Stephanopoulos, G. Busca, Appl. Catal. B Environ. (2018), doi:http://dx.doi.org/ doi:http://dx.doi.org/10.1021/i300010a011.
10.1016/j.apcatb.2018.12.063. [71] J.K. Kesavan, I. Luisetto, S. Tuti, C. Meneghini, G. Iucci, C. Battocchio, S. Mobilio,
[30] C. Fukuhara, K. Hayakawa, Y. Suzuki, W. Kawasaki, R. Watanabe, Appl. Catal. A S. Casciardi, R. Sisto, J. CO2 Util. 23 (2018) 200, doi:http://dx.doi.org/10.1016/j.
Gen. 532 (2017) 12, doi:http://dx.doi.org/10.1016/j.apcata.2016.11.036. jcou.2017.11.015.
[31] H. Muroyama, Y. Tsuda, T. Asakoshi, H. Masitah, T. Okanishi, T. Matsui, K. Eguchi, J. Catal. [72] G. Baldauf-Sommerbauer, S. Lux, W. Aniser, B. Bitschnau, I. Letofsky-Papst, M.
343 (2016) 178, doi:http://dx.doi.org/10.1016/j.jcat.2016.07.018. Siebenhofer, J. CO2 Util. 23 (2017) 1, doi:http://dx.doi.org/10.1016/j.
[32] J. Yang Lim, J. McGregor, A.J. Sederman, J.S. Dennis, Chem. Eng. Sci. 141 (2016) jcou.2017.10.022.
28, doi:http://dx.doi.org/10.1016/j.ces.2015.10.026. [73] G. Baldauf-Sommerbauer, S. Lux, W. Aniser, M. Siebenhofer, Chem. Eng.
[33] S. Rahmani, M. Rezaei, F. Meshkani, J. Ind, Eng. Chem. 20 (4) (2014) 1346, doi: Technol. 39 (11) (2016) 2035, doi:http://dx.doi.org/10.1002/ceat.201600094.
http://dx.doi.org/10.1016/j.jiec.2013.07.017. [74] W. Gac, W. Zawadzki, M. Rotko, M. Greluk, G. Słowik, G. Kolb, Catal. Today
[34] M. Romero-Sáez, A.B. Dongil, N. Benito, R. Espinoza-González, N. Escalona, F. (2019), doi:http://dx.doi.org/10.1016/j.cattod.2019.07.026.
Gracia, Appl. Catal. B Environ. 237 (2018) 817, doi:http://dx.doi.org/10.1016/j.
apcatb.2018.06.045.

You might also like