Fuel 217 (2018) 151–159

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Full Length Article

A study on three-phase CO2 methanation reaction kinetics in a continuous T

stirred-tank slurry reactor
⁎
Jonathan Lefebvre , Nike Trudel, Siegfried Bajohr, Thomas Kolb
Karlsruhe Institute of Technology, Engler-Bunte-Institut, Fuel Technology Division, Engler-Bunte-Ring 1, 76131 Karlsruhe, Germany

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

Keywords: The reaction kinetics of the three-phase CO2 methanation for a commercial Ni/SiO2 catalyst suspended in a
CO2 methanation liquid phase is studied in a continuous stirred-tank slurry reactor at a CO2 partial pressure of 1 bar and tem-
Reaction kinetics peratures from 220 °C to 320 °C. By applying different liquids, namely squalane, octadecane, and dibenzylto-
Three-phase methanation luene, showing different gas solubilities, it is found that the gas concentration in the liquid phase and not the
Liquid influence on reaction kinetics
partial pressure in the gas phase is the driving force for the CO2 methanation reaction kinetics. The liquid phase
Slurry bubble column reactor
does not influence the reaction kinetics but reduces the available gas concentrations and H2/CO2 ratio on the
catalyst surface. Based on these findings, a kinetic rate equation for the three-phase CO2 methanation is de-
veloped additionally incorporating the chemical equilibrium limitations relevant in the temperature regime.

1. Introduction plants [3,6,7].

With the COP21 Agreement, the parties of the United Nations
Framework Convention on Climate Change agreed on reducing green-
house gas emissions in order to keep the increase in global average
temperature well below 2 K [1]. One way to achieve this goal is to
reduce the CO2 emissions through a drastic increase of the share of
renewable and environmentally friendly energy sources like wind and
sunlight in our energy systems. This electrical energy can be used to
power many applications in the mobility and heat sectors. Nevertheless,
the current share of electricity in final energy consumption is relatively
low, e.g. about 22% in the EU [2]. To cover the rest of the energy
demand would require a large extension of the existing renewable en-
ergy plants and power grid. One possible solution to tackle this issue is
the application of Power-to-Gas (PtG) processes which aim at trans-
forming renewable electrical energy into chemical energy carriers with
high energy density which can be stored over long time periods. This is
achieved through transforming electrical energy via water electrolysis
into hydrogen which is subsequently transformed with carbon dioxide
into methane in order to produce synthetic natural gas (SNG), see Eq.
(1) [3–5]. At first glance, the transformation of hydrogen into methane
seems an unnecessary step coupled with losses in overall efficiency.
However, it offers many benefits, i.e. a very well developed infra-
structure for transportation and storage and the use of SNG in a wide
range of highly efficient final energy conversion technologies, e.g. de-
centralized/central combined heat and power units, mobility (com-
pressed natural gas and liquefied natural gas) or combined cycle power
CO2 (g) + 4H2 (g)CH 4 (g) + 2H2 O(g) ΔhR0 = −165.1 kJ/mol (1)
The PtG process is a dynamic process allowing large-scale and
flexible energy storage. Both electrolyzer and methanation reactor
should be able to operate under dynamic conditions in order to limit the
extent of expensive hydrogen storage technology. Considering the high
exothermic methanation reaction (see Eq. (1)), heat removal from the
reactor is one of the main issues related to the design of a methanation
unit [8–10]. One way to tackle this issue is to run the methanation
reaction in a slurry bubble column reactor, as this type of reactor allows
for very efficient reactor temperature control as well as good dynamic
behavior [11–14]. Slurry bubble column reactors are usually applied
for Fischer-Tropsch synthesis, methanol and dimethyl ether production,
hydrodesulfurization and hydrogenation of fatty oils as well as other
hydrogenation and oxidation reactions [15–21].
Up to now, there is no reaction rate equation available in the lit-
erature for the reaction kinetics of the CO2 methanation in a three-
phase system. Especially the influence of the liquid phase on the reac-
tion kinetics is not systematically investigated. Even looking at similar
processes in the literature, the liquid-phase influence on reaction ki-
netics is still neither well-defined nor well understood. For the liquid-
phase hydrogenation of cyclohexene on Pd, Madon et al. [22] showed
that the H2 concentration in the liquid phase is the determining factor
to describe the reaction kinetics. However, when Pt is applied for the
same reaction, Gonzo and Boudart [23] showed that the H2 partial
pressure in the gas phase is the determining factor. In three-phase

⁎
Corresponding author.
E-mail address: jonathan.lefebvre@kit.edu (J. Lefebvre).

https://doi.org/10.1016/j.fuel.2017.12.082
Received 16 May 2017; Received in revised form 14 November 2017; Accepted 19 December 2017
Available online 02 January 2018
0016-2361/ © 2017 Elsevier Ltd. All rights reserved.
J. Lefebvre et al. Fuel 217 (2018) 151–159

Nomenclature ML Liquid phase molar mass (kg/mol)

ṅ Molar stream (mol/s)
Symbols and abbreviations p0 Pressure at standard conditions (bar)
pi Partial pressure of gas component i (bar)
A Parameter describing the temperature dependency of CO2 PtG Power-to-Gas (–)
and H2 Henry’s law constant in DBT (see Eq. (4)) (–) r3PM Catalyst mass-specific CO2 reaction rate (mol/(kg·s))
B Parameter describing the temperature dependency of CO2 R Ideal gas constant (J/(mol·K))
and H2 Henry’s law constant in DBT (see Eq. (4)) (K) SNG Synthetic Natural Gas (–)
C Parameter describing the temperature dependency of CO2 T Temperature (°C or K)
and H2 Henry’s law constant in DBT (see Eq. (4)) (K2) xi Molar fraction of gas component i in the liquid phase (–)
ci,L Concentration of gas component i in the liquid phase XCO2 CO2 conversion (see Eq. (6)) (–)
(mol/m3) yi Molar fraction of gas component i (–)
*
ci,L Concentration of gas component i in the liquid phase at α H2 reaction order (–)
gas/liquid phase equilibrium (mol/m3) β CO2 reaction order (–)
CSTR Continuous Stirred-Tank Reactor (–) γ H2O reaction order (–)
DBT Dibenzyltoluene (–) ρL Liquid phase density (kg/m3)
ΔhR0 Reaction enthalpy at standard conditions (J/mol) τmod,CO2 Modified CO2 residence time (see Eq. (7)) (kg·s/mol)
EA Activation energy of the reaction (J/mol) 3PM Three-phase methanation (–)
Hi,px Henry’s law constant of gas component i (see Eq. (2)) (bar)
Hi,pc Concentration-based Henry’s law constant of gas compo- Subscripts and superscripts
nent i (see Eq. (3)) bar·(m3/mol)
k Reaction rate constant (see Eq. (10)) (mol/ cal Calculated
(kg·s·mol0.4·m−1.2)) cat Catalyst
k0 Reaction rate constant in Arrhenius equation (mol/ exp Experimental
(kg·s·mol0.4·m−1.2)) G Gas phase
K Parameter to express the reaction rate limitation due to i Gas component i
chemical equilibrium closeness (see Eq. (11)) (–) in Reactor inlet
K eq Equilibrium constant of the reaction (–) L Liquid phase
K H2O H2O adsorption constant (m0.3/mol0.1) out Reactor outlet
mcat Catalyst mass (kg)

PIRC Pressure
FIC
Condensate controller
CO2 tank
FIC

H2
FIC
TC

CH4
FIC
Hood
TC
N2 TI
FIC

PIR
Ar TIR
PIR
TC
Feed
H 2O TIRC
tank
TIC TIC GC

FIC
TC
Analysis of
CO2, H2, CO, CH4
Evaporator Autoclave Ar, N2 and C1-2
Fig. 1. Flow scheme of the experimental setup.

152
J. Lefebvre et al. Fuel 217 (2018) 151–159

Fischer-Tropsch synthesis, gas partial pressures – and not gas con-
centration in the slurry phase – are usually used for kinetic equations
[24,25]. For the three-phase methanol synthesis, Graaf et al. [26,27]
used the gas concentration in the liquid phase to describe the reaction
kinetics. However, the estimated activation energy for the CO2 me-
thanol synthesis was reported to be much lower than the activation
energy in a comparable two-phase system.
In this publication, the influence of a liquid phase on the reaction
kinetics of the three-phase CO2 methanation will be investigated on a
commercially available catalyst. In a first step, it will be determined
whether the partial pressure or the gas concentration in the liquid phase
has to be used for description of the reaction kinetics through reaction
kinetic experiments using different liquid phases. In a second step, a
kinetic rate equation will be developed from kinetic measurements with
dibenzyltoluene as liquid phase at a CO2 partial pressure of 1 bar and
temperatures from 220 °C to 320 °C. In both steps, attention will be paid
to the absence of heat and mass transfer limitations. The aim of this
publication is not to study the mechanism of the CO2 methanation re-
action in a three-phase system but to develop a CO2 methanation ki-
netic rate equation valid in a three-phase system and to compare it with
a two-phase kinetic rate equation. This comparison, however, will be
the topic of a future publication.
2. Material and methods
2.1. Material
higher as compared to DBT.
The different evolutions of H2 and CO2 Henry’s law constants with
temperature can be explained by the so-called enthalpy-entropy com-
pensation [30]. The Henry’s law constant dependency with temperature
is an inverse function of the sum of solute enthalpy and entropy. For
gases with small solubilities, the entropy term increases with increasing
temperature, while it decreases for gases with high solubilities [31].
The enthalpy term depends on intermolecular forces which generally
decrease with increasing temperatures [31]. Therefore, the solute en-
thalpy decreases with increasing temperatures. At high temperatures,
H2 shows little intermolecular interactions with solvents due to its small
size and physical symmetry. Hence, the enthalpy term for H2 is very
small and the Henry’s law constant of H2 at high temperatures depends
mostly on the entropy term. This is the reason why HH2,pc decreases with
increasing temperature as shown in Fig. 2. Compared to H2, CO2 is a
more polar and much bigger molecule which offers intermolecular in-
teractions with solvents. As such, HCO2,pc depends mostly on the en-
thalpy of dissolved CO2 and therefore HCO2,pc increases with increasing
temperature.
The Henry’s law constants Hi,pc for H2O and CH4 in DBT were
measured in a similar setup as described by Götz et al. [29]. The tem-
perature dependency of Hi,pc can be expressed using Eq. (4). The gas-
specific coefficients A, B and C can be found in Table 1.
B C
lnHi,pc = A + + 2
T T (4)

Besides the solubility aspects of the experimental work, the catalytic

For this publication, the experiments have been carried out using
methanation has to be considered. The catalyst used in this work is a
the experimental setup shown in Fig. 1. The gas components (CO2, H2,
commercially available Ni/SiO2 catalyst with a particle size of
CH4, H2O, N2 used as inert gas, and Ar as internal standard) are sup-
25–50 µm. This particle fraction has been chosen to prevent internal
plied via a set of mass flow controllers (Bronkhorst). The purities of the
mass transfer limitations. The Weisz-Prater criterion has been evaluated
gases used in the experimental work were greater than 99.995%. Before
for each kinetic measurement [32]. This criterion is respected for all
entering into the 1-liter stainless steel autoclave reactor (versoclave
kinetic data showing the absence of any internal mass transfer limita-
type 3E, Büchiglasuster) used as continuous stirred-tank slurry reactor
tion (see Appendix A).
(CSTR), the feed gases are heated up to the desired temperature and
Before each experiment, the catalyst is reduced in a fixed-bed re-
mixed in a tempered feed tank. At the reactor outlet, the product gas is
actor for 24 h at 400 °C with a volume flow of 40 L/h (NTP) in an at-
cooled down to 5 °C in order to condensate the reaction water and the
mosphere of Ar/H2 = 3/1. Then, the reduced catalyst is suspended into
evaporated liquid. The pressure control system (Bronkhorst) is located
the desired liquid under inert atmosphere to prevent any catalyst oxi-
downstream the condensate tank. Following the pressure controller, the
dation and activity decrease. Hence, in this publication, the mass of
dry product gas is analyzed with a gas chromatograph (µ-GC G3581A,
catalyst, mcat , is the mass of the reduced catalyst used for methanation
Agilent), which is calibrated to identify CH4, CO2, CO, Ar, N2, H2, C2H4,
experiments.
and C2H6.
The liquids employed for this publication are squalane (purity 99%,
Sigma Aldrich), octadecane (purity 99%, VWR International GmbH),
and dibenzyltoluene (DBT, trade name Malotherm® SH, Sasol). These
liquids were chosen because they cover a wide range of gas solubilities 0.4
Henry's law constant Hei,pc / bar·m³/mol

and have sufficiently low vapor pressures at reaction temperatures

which are necessary for the experimental work. Solubility data for the
relevant gas species in squalane and octadecane were taken from [28],
0.3
while solubilities in DBT were taken from [29]. A common quantity to
describe gas solubilities in liquids is the Henry’s law constant, Hi,px ,
defined with Eq. (2):
pi 0.2
Hi,px =
xi (2)
However, in order to evaluate the gas concentration in the liquid
phase at the gas – liquid equilibrium, ci,L ∗
, the concentration-based 0.1
Henry’s law constant, Hi,pc (see Eq. (3)), is a more appropriate quantity.
pi M
Hi,pc = ≈ Hi,px · L
ci∗,L ρL (3) 0
160 200 240 280 320
Fig. 2 shows the dependencies of Hi,pc from temperature for H2 and
Temperature T / °C
CO2 in squalane, octadecane, and DBT. It can be seen that for all liquids,
H2 solubility increases, while CO2 solubility decreases with increasing Fig. 2. Evolution of H2 (round) and CO2 (square) Henry’s law constants in squalane
temperature. The H2 and CO2 solubilities are highest for octadecane (white), octadecane (grey), and DBT (black) with temperature. Data for squalane and
octadecane from [28] and data for DBT from [29].
and lowest for DBT. Gas solubility in octadecane is about 1.3–1.4 times

153
J. Lefebvre et al. Fuel 217 (2018) 151–159

Table 1 2.2. Experimental methods

Parameters to describe the temperature of concentration-based CH4 and H2O Henry’s law
constant in DBT according to Eq. (4).
2.2.1. Measurements
Gases A B C Before each experiment, the slurry phase consisting of liquid and
– K K2 catalyst is filled into the autoclave reactor and heated up to the desired
reaction temperature. At the same time, a 200 mL/min volume flow of
CH4 1.0697 −2.8567 · 103 7.0853 · 105
Ar/H2 = 1/1 is sent through the reactor in order to prevent catalyst
H2O −2.1325 · 101 2.1971 · 104 −6.5256 · 106
oxidation. Afterwards, the reactor inlet flow and composition are
switched to the desired values, see experimental methods described in
Table 2 the two following paragraphs as well as in Table 2. In all experiments,
Variation of process parameters for the development of a kinetic rate equation. Ar is used as internal standard for the mass balance based on GC ana-
lysis of the gas compositions, while N2 is used as buffer and dilution and
Parameters Variation Units
varied in order to maintain a constant overall gas flow at the reactor
T 220–320 °C inlet.
pCO2 ,out 0.5–2.3 bar As mentioned before, for investigations on whether the reactant
pH2,out 2.9–12.9 bar partial pressure or the reactant concentration in the liquid phase is the
pCH4 ,out 0.04–1.4 bar rate-determining factor for the three-phase methanation (3PM), three
pH2O,out 0.08–2.8 bar different liquids – squalane, octadecane, and DBT – are used as sus-
τmod,CO2 1.7–35 kg·s/mol pension liquids. For the first experiments, H2 and CO2 partial pressures
mcat 0.16–0.8 10−3 kg are kept constant for all three suspension liquids, and the temperature is
(H2/CO2)G,out 1.71–12.796 – varied from 220 to 260 °C. Then, since HH2,pc and HCO2,pc are different in
(H2/CO2)L 0.58–4.35 –
each liquid phase, H2 and CO2 partial pressures are separately varied
depending on each three-phase system to obtain a set of data with the
30 same H2 and CO2 concentrations in the different liquid phases at a
temperature of 230 °C. In both sets of experiments, the CO2 reaction
rate as defined in Eq. (8) is evaluated. CH4 and H2O partial pressures or
concentrations in the liquid phase could not be controlled for the ex-
CO2 conversion XCO / %

periments in octadecane, as HCH4,pc and HH2O,pc in octadecane are not

2

20 available in the literature.

For the development of a 3PM kinetic rate equation, kinetic mea-
surements are performed in DBT. In a first set of experiments at 260 °C,
reactant partial pressures are one by one varied in order to identify
their influence on the CO2 reaction rate. For the remaining experiments,
10 reactant and product partial pressures are simultaneously varied for
temperatures between 220 °C and 320 °C as shown in Table 2 to speed
up the reaction kinetic measurements.
Each catalyst batch used in this work was operated under metha-
nation conditions for at least 300 h and no catalyst deactivation could
0 be observed during this period of time. With the results of 91 experi-
400 600 800 1000 1200
ments in total, a 3PM kinetic rate equation is developed according to
Agitator speed n / 1/min Section 2.2.3.
Fig. 3. Influence of agitator speed on the CO2 conversion in DBT (T = 260 °C,
pH2,out = 4 bar, pCO2 ,out = 1 bar, τmod,CO2 = 14 kg·s/mol). 2.2.2. Measurements analysis
In order to estimate the reactor outlet molar flows, Ar is used as
2.0 reference gas. Knowing the reactor inlet molar flow ṅin as well as the
Squalane molar fraction at the reactor inlet and outlet from the GC analysis, the
Octadecane outlet molar flow of each component can be calculated according to Eq.
log(r3PM,exp / mmol/(kg·s)) / -

DBT (5).
1.5
yAr,in
ni̇ ,out = ṅout ·yi,out = ṅin · ·yi,out
yAr,out (5)

1.0 H2O cannot be detected by the applied GC. Hence, the outlet H2O
molar flow is calculated with an oxygen balance around the reactor.
During each experiment, attention is paid to the atomic balance around
the reactor; experiments with a carbon or hydrogen balance error of
0.5 more than 1% are not used. The CO2 conversion under stationary
conditions XCO2 is expressed with Eq. (6):
ṅCO2 ,in−ṅCO2 ,out
XCO2 =
0.0 ṅCO2 ,in (6)
1.85 1.90 1.95 2.00 2.05
With the modified CO2 residence time τmod,CO2 , see Eq. (7),
1000/T / 1/K
mcat
Fig. 4. Arrhenius plot of CO2 reaction rates for squalane, octadecane, and DBT τmod,CO2 = ,
ṅCO2 ,in (7)
( pH2,out = 4 bar, pCO2 ,out = 1 bar).
the experimental CO2 reaction rate rCO2 ,exp can be calculated according
to Eq. (8).

154
J. Lefebvre et al. Fuel 217 (2018) 151–159

CO2 reaction rate r3PM,exp / mmol/(kg·s) 12 12

CO2 reaction rate r3PM,exp / mmol/(kg·s)

9 9

6 6

3 3

0 0
Squalane Octadecane DBT Squalane Octadecane DBT

Fig. 5. Comparison of CO2 reaction rate at 230 °C and τmod,CO2 = 17–20 kg·s/mol in squalane, octadecane, and DBT at pH2,out = 4 bar and pCO2 ,out = 1 bar (left), and c H2 = 7.23 mol/m3
and c CO2 = 5.72 mol/m3 (right).

H2 concentration cH ,L / mol/m³
2
while K describes the limitation of r3PM when the reaction system ap-
0.0 11.7 23.3 35.0 46.7 proaches the chemical equilibrium described by K eq . K is calculated
40 according to Eq. (11).
CO2 reaction rate r3PM,exp / mmol/(kg·s)

−E
k = k 0·exp ⎛ A ⎞
⎝ R·T ⎠ (10)
30
pH22O,out ·pCH4 ,out p02
K = 1− ·
pH42,out ·pCO2 ,out K eq (11)
20 The chemical equilibrium constant K eq is estimated through mini-
mization of the system’s Gibb’s enthalpy. The Gibb’s enthalpies of the
species are calculated based on the species enthalpies and entropies
10
according to the Shomate equation and data from NIST Chemistry
WebBook. The kinetic parameters k , α , β , γ , and K H2O are determined
by a least-squares minimization between calculated CO2 reaction rates
r3PM,cal , (Eq. (9)), and experimentally observed CO2 reaction rates r3PM,exp
0 (Eq. (8)). In a first step, α , β , γ , and K H2O are guessed and k is de-
0 4 8 12 16 termined for each investigated temperature. Then, k 0 and EA are cal-
H2 partial pressure pH ,out / bar
2
culated from linear regression in an Arrhenius plot using the Arrhenius
equation (see Eq. (12)). This two-step method aims for verifying the
Fig. 6. Influence of the H2 partial pressure on the CO2 reaction rate (T = 260 °C,
first-step fit through the examination of the Arrhenius plot fit quality.
pCO2 ,out = 1 bar, pCH4 ,out = 0.27 bar, pH2O,out = 0.79 bar, τmod,CO2 = 2.7 kg·s/mol).
Only Arrhenius fits with a R2 ≥ 0.98 have been considered. The error
between r3PM,cal and r3PM,exp is determined and α , β , γ , and K H2O are
XCO2
r3PM,exp = further varied until the error reaches a minimum.
τmod,CO2 (8)
EA 1
lnk = lnk 0− ·
For each experiment, the measurement precision is evaluated with R T (12)
the method of partial derivatives.

3. Results and discussion

2.2.3. Development of a reaction rate equation
A simple power law rate equation with product limitation is used to
describe the experimental CO2 reaction rates (see Eq. (9)). A Langmuir-
Hinshelwood kinetic rate equation could have been used to describe the
experimental results. Nevertheless, the mechanism of the three-phase
CO2 methanation has not been studied yet and would require challen-
ging spectroscopic studies which are not part of this publication. This is
the reason why a power law rate equation based on the following ex-
perimental observations is preferred. Furthermore, during the experi-
ments, a CH4 selectivity above 95% could be observed. Hence, side
reactions along the main reaction have been ignored for the develop-
ment of the kinetic rate equation.
β
cHα2,L·cCO2 ,L
r3PM,cal = k ·
(1 + K H2O·cH2O,L)γ
In Eq. (9), k is the reaction rate constant as defined in Eq. (10),
Prior to each set of experiments, absence of mass transfer limitation
in the liquid phase was verified through variation of the autoclave
agitator speed. An example is shown in Fig. 3 for DBT at a reaction
temperature of 260 °C. It can be seen that the CO2 conversion XCO2 does
not increase any further for an agitator speed above ca. 1000 l/min.
Hence, no mass transfer resistance in the liquid phase has to be con-
sidered above this threshold.
This observation was also confirmed for all other experimental
conditions. Thus, an agitator speed of 1000 1/min was selected for all
experiments described in this work.
3.1. Influence of the liquid phase on reaction kinetics
·K
(9)
The temperature influence on the CO2 reaction rate r3PM,exp mea-
sured in squalane, octadecane, and DBT at pH2,out = 4 bar and

155
J. Lefebvre et al. Fuel 217 (2018) 151–159

CO2 concentration cCO ,L / mol/m³

2 However, going more into detail and looking at the possible influ-
4.3 8.6 12.9 17.2 21.5 ence of the liquid phase on the reaction kinetics, Fig. 4 shows that
40 different levels of r3PM,exp are achieved for each of the investigated liquid
CO2 reaction rate r3PM,exp / mmol/(kg·s)

phases. The highest r3PM,exp is obtained in squalane. In comparison,

r3PM,exp in octadecane is ca. 5% lower as compared to squalane, while
r3PM,exp in DBT is ca. 1.6 times lower than in squalane. This phenomenon
30
was also observed by Madon et al. [22], who studied the hydrogenation
of cyclohexene on Pd. They showed that the H2 reaction rates are dif-
ferent at identical H2 partial pressures when different liquid phases are
20 applied. Nevertheless, they showed that identical H2 reaction rates were
obtained when identical H2 concentrations in the liquid phase were
applied. Because of these reported observations, a similar experiment
was performed in this work.
10 In Fig. 5, CO2 reaction rates measured at 230 °C either at identical
H2 and CO2 partial pressures or at identical H2 and CO2 concentrations
in the liquid phase are compared for the three investigated liquids. As
0 already seen in Fig. 4, at identical H2 and CO2 partial pressures, dif-
0.5 1.0 1.5 2.0 2.5 ferent reaction rates are observed depending on the applied liquid
phase. The CO2 reaction rate in DBT is about two times smaller than the
CO2 partial pressure pCO ,out / bar
2
reaction rates in squalane and octadecane. However, at identical H2 and
Fig. 7. Influence of the CO2 partial pressure on the CO2 reaction rate (T = 260 °C, CO2 concentrations in the liquid phase, the measured CO2 reaction rates
pH2,out = 4 bar, pCH4 ,out = 0.27 bar, pH2O,out = 0.79 bar, τmod,CO2 = 7.6–20 kg·s/mol). are similar. The highest r3PM,exp is obtained in octadecane. r3PM,exp in
squalane and r3PM,exp in DBT are about 4% and 12% lower than in oc-
tadecane, respectively. The lower r3PM,exp in DBT can be explained by
CH4 concentration cCH ,L / mol/m³
4 the H2O solubility in DBT, which is higher than in squalane. H2O has a
1.8 3.6 5.4 7.2 9.0 negative influence on r3PM,exp (see section 3.2.2). Hence, the higher H2O
40 concentration in DBT decreases r3PM,exp as compared to the other liquid
CO2 reaction rate r3PM,exp / mmol/(kg·s)

phases. Other reasons to explain the discrepancies between DBT and the
other solvents are molecular interactions during solvation. The Henry’s
law constants have been measured with pure gases and do not take into
30
account for intermolecular interactions. Another explanation may lie in
solvent adsorption effect on the catalyst active sites. DBT is an aromatic
compound with a higher adsorption than paraffin like octadecane and
20 squalane [22,36].
From Fig. 5, it can be drawn that the gas concentration in the liquid
phase and not the gas partial pressure is the relevant parameter for the
CO2 methanation reaction kinetics in three-phase reactors.
10 Madon et al. [37] used transition state theory to explain the solvent
effect for three-phase cyclohexene hydrogenation (liquid-liquid-solid)
and Fischer-Tropsch synthesis. They state that partial pressure can be
used to describe reaction kinetics when the reaction rate-determining
0
0.3 0.6 0.9 1.2 1.5 step is not an adsorption/desorption step, e.g. hydrogenation of

CH4 partial pressure pCH ,out / bar

4
H2O concentration cH O,L / mol/m³
2

Fig. 8. Influence of the CH4 partial pressure on the CO2 reaction rate (T = 260 °C, 12.9 19.4 25.9 32.4 38.8
pCO2 ,out = 1 bar, pH2,out = 4 bar, pH2O,out = 0.53 bar, τmod,CO2 = 14.9 kg·s/mol). 40
CO2 reaction rate r3PM,exp / mmol/(kg·s)

pCO2 ,out = 1 bar is depicted in Fig. 4. Measurements in octadecane could

not be performed at temperatures above 230 °C, because octadecane 30
crystallized at the cooled reactor outlet and blocked the gas outlet flow.
Nevertheless, the results from the unimpeded experiments with octa-
decane were regarded as sufficient and no change of the experimental
setup was done for this work. From Fig. 4, it can be seen that tem- 20
perature has the same influence on the reaction for all liquid phases:
Activation energies of 75–84 kJ/mol are obtained from the Arrhenius
plots. These values are typical of the CO2 methanation reaction in the
10
gas phase [33–35]. Thus, assuming that the partial pressure is the re-
levant parameter for description of the reaction kinetics, these activa-
tion energies confirm that the measurements have been carried out in
the absence of mass and heat transfer limitations. Furthermore, the 0
dependencies of HH2,pc and HCO2,pc with temperature are the same for 0.6 0.9 1.2 1.5 1.8
squalane, octadecane, and DBT (see Fig. 2). It is therefore coherent that H2O partial pressure pH O,out / bar
2
the same temperature dependency can be observed in the Arrhenius
plot, irrespective of the gas concentration or partial pressure being the Fig. 9. Influence of the H2O partial pressure on the CO2 reaction rate (T = 260 °C,
pCO2 ,out = 1 bar, pH2,out = 4 bar, pCH4 ,out = 0.55 bar, τmod,CO2 = 16.7 kg·s/mol).
kinetic determining factors.

156
J. Lefebvre et al. Fuel 217 (2018) 151–159
Calculated CO 2 reaction rate r3PM,cal / mmol/(kg·s)
120 the help of three-phase spectroscopic measurements which have not
220 °C 230 °C been performed up to now due to their complexity.
240 °C 250 °C + 10% With this background, a reaction rate equation for the three-phase
260 °C 270 °C CO2 methanation is developed based on kinetic measurements per-
90 280 °C 290 °C formed with DBT in order to compare it with a two-phase methanation
- 10%
300 °C 310 °C
kinetic rate equation. This comparison, however, is the topic of a future
320 °C
publication.
60
3.2. Development of a reaction rate equation

3.2.1. Educt influence on the CO2 reaction rate

30 The influence of H2 concentration on the CO2 reaction rate at 260 °C
is shown in Fig. 6. An increase in c H2,L increases r3PM,exp . The value of
r3PM,exp is increased by 26% when c H2,L is doubled. Hence, the observed
H2 reaction order α is about 0.3 at 260 °C. By increasing reaction
0
0 30 60 90 120 temperatures from 220 to 320 °C (not shown here), an increase in α
from ca. 0.25 to 0.45 is observed. The positive influence of H2 con-
Experimental CO2 reaction rate r3PM,exp / mmol/(kg·s)
centration on the CO2 reaction rate has already been reported in the
Fig. 10. Parity plot between experimental and calculated CO2 reaction rates. Calculated literature with reaction orders ranging from 0.2 to 1 [33,34,43,47–50].
reaction rates using the set of kinetic parameters described in Eq. (13). In Fig. 7, the impact of CO2 concentration on the CO2 reaction rate
at 260 °C is depicted. It can be observed that the influence of cCO2 ,L is
Table 3
significantly smaller as compared to c H2,L . The CO2 reaction rate is in-
Measurement uncertainties. creased only by ca. 15% when cCO2 ,L is doubled. Considering the other
kinetic experiments performed at temperatures from 220 to 320 °C, an
Parameter T pi * HH2,pc ** HCO2,pc ** HH2O,pc ** increase in CO2 reaction order β from 0.1 to 0.18 is observed by in-
creasing temperatures. The positive influence of cCO2 ,L on r3PM,exp has
Uncertainty ± 1K ± 4% ± 14% ± 5% ± 5%
also been described in the literature, however, with higher CO2 reaction
*
3% uncertainties are related to the electronic sensor precision and 1% to the GC preci- orders from 0.33 to 1 [34,43,47–50]. Lim et al. [33] have reported a
sion. small influence of CO2 concentration on the CO2 reaction rate for high
**
Uncertainties are related to the experimental setup precision described in [29]. CO2 concentrations and stoichiometric H2/CO2 ratios. They reported a
greater influence of CO2 concentration on r3PM,exp only for low CO2
10 concentrations and high H2/CO2 ratios. In this publication, due to the
cH ,L ± 14 %
Change in reaction rate Δr3PM,cal / %

2
different H2 and CO2 gas solubilities (HH2,px > HCO2,px ), the reaction
cCO ,L ± 5 %
2
system is characterized by sub stoichiometric H2/CO2 ratios. Hence, for
cH O,L ± 5 %
5
2
the 3PM, CO2 is not the limiting reactant but H2, and the CO2 reaction
T ±1 K
order is accordingly low.
Sum

3.2.2. Product influence on the CO2 reaction rate

0 The influence of CH4 concentration on the CO2 reaction rate at
-5
-10
Min
-1
Variation
Fig. 11. Sensitivity analysis on the developed kinetic rate equation given in Eq. (9)
(T = 260 °C, c H2,L = 11.71 mol/m3, c CO2 ,L = 8.51 mol/m3, c CH4 ,L = 1.24 mol/m3,
c H2O,L = 9.68 mol/m3).
adsorbed carbon species for Fischer-Tropsch synthesis. When the rate-
determining step is an adsorption/desorption step, the reaction kinetics
is described by gas concentration in the liquid phase and not by partial
pressure. However, our results contradict Madon et al. [37]. The two-
phase CO2 methanation rate-determining step is known to be the dis-
sociation of an already adsorbed carbon species and is not controlled by
adsorption/desorption [38–46]. Besides, we showed that the three-
phase CO2 methanation kinetics can only be described by gas con-
centration in the liquid phase. To reconcile our results and those found
by Madon et al., it is possible that an adsorption/desorption-limiting
methanation reaction mechanism occurs in a three-phase system due to
unusually low H2/CO2 ratios resulting from the different solubilities of
H2 and CO2 in organic liquids. This assumption could be confirmed with
260 °C is shown in Fig. 8. An increase in cCH4 ,L has no impact on r3PM,exp .
This trend can be observed for other reaction temperatures. Therefore,
the CH4 reaction order is found to be 0 in three-phase methanation.
These findings are also reported for two-phase experiments by different
authors [33,34,43,47–50].
In Fig. 9, the influence of H2O concentration on r3PM,exp at 260 °C is
illustrated. It can be seen that a small increase in c H2O,L leads to a re-
0 Max
1 duction of r3PM,exp by about 10%. However, a further increase in c H2O,L
does not significantly decrease r3PM,exp anymore. In addition, the de-
crease in r3PM,exp with increasing c H2O,L becomes stronger with increasing
temperatures: the H2O reaction order increases from 0.1 to 0.13 in the
temperature range of 220 to 320 °C. The inhibiting effect of H2O on the
CO2 methanation reaction kinetics including the reported tendencies
has also been observed by Lim et al. [33]. According to them, the ne-
gative influence of H2O on the CO2 methanation rate is explained by the
adsorption of H2O on the catalyst active sites, preventing H2 or CO2 to
adsorb and further react on the catalyst.
3.2.3. Reaction rate equation
From the experiments described in Sections 3.2.1 and 3.2.2, it can
be seen that H2 and CO2 have a positive influence on the three-phase
CO2 methanation reaction kinetics, while CH4 has no effect on r3PM,exp
and H2O inhibits the methanation. This is the reason why the kinetic
rate equation given in Eq. (9) was chosen to describe r3PM,exp .
Based on the experiments performed at 260 °C as well as on those
performed at temperatures from 220 to 320 °C a kinetic rate equation is
fitted. The kinetic parameters resulting from the least-squares

157
J. Lefebvre et al. Fuel 217 (2018) 151–159

minimization are summarized in Eq. (13). An activation energy EA of due to weighing uncertainty each batch shows slightly different cata-
79 kJ/mol – typical of CO2 methanation – is retrieved from the rate lytic activities compared to the others, which additionally increases the
equation optimization. The parity plot between r3PM,exp and r3PM,cal is error between calculated and experimental reaction rates.
shown in Fig. 10. Good agreement between the experimental results
and the model is obtained. Assessing a normal distribution, a standard
4. Summary
deviation between r3PM,exp and r3PM,cal of 6.0% is achieved. This kinetic
rate equation is valid for the experimental conditions summarized in
In this publication, the influence of a liquid phase on the reaction
Table 2, i.e. for substoichiometric H2/CO2 characteristic of three-phase
kinetics of the CO2 methanation has been studied in a CSTR using a
methanation conditions.
commercially available catalyst and three different liquids, namely
0.3 0.1
−79061 cH2,L·cCO2 ,L squalane, octadecane, and dibenzyltoluene, showing different gas so-
r3PM,cal = 3.90699·105·exp ⎛ ⎞· ·K
⎝ R·T ⎠ (1 + 1·c H2O,L )0.1 (13) lubilities. By applying similar experimental conditions – either gas
partial pressures or gas concentrations in the liquid phase – it was found
that gas concentrations in the liquid phase are the kinetically relevant
3.2.4. Sensitivity analysis
parameter to describe the CO2 methanation reaction kinetics.
In order to understand the discrepancies between calculated and
Based on these findings, reaction kinetic measurements have been
experimental reaction rates, a sensitivity analysis on the reaction rate
performed in dibenzyltoluene at a CO2 partial pressure of 1 bar and
equation given in Eq. (13) was carried out. For this analysis, the reac-
temperatures between 220 °C and 320 °C in order to establish a power
tion temperature and the reactant/product concentrations were varied
law kinetic rate equation for CO2 methanation taking into account the
according to the uncertainty related to these parameters. These un-
chemical equilibrium limitations. The developed kinetic rate equation
certainties are listed in Table 3.
is able to predict the experimental reaction rates with a standard de-
Fig. 11 shows the influence of measurement uncertainties on the
viation of 6.0%.
calculated reaction rate. It can be seen that c H2,L has the strongest im-
In terms of further work, increased precision of the measured values
pact on r3PM,cal followed by temperature, cCO2 ,L and c H2O,L . The de-
especially of Henry’s law constant for H2 is in the focus. It is the biggest
creasing influence of gas solubility from H2 to H2O is directly related to
source of uncertainty for the calculated reaction rates. A further ap-
the gas species reaction order as well as the precision of Henry’s law proach will be the study of the reaction kinetics of the CO2 methanation
constants, while the temperature impact is related to the reaction ac-
in a state-of-the-art fixed-bed reactor. As the liquid phase seems to only
tivation energy. Together, the uncertainties in the measurements can influence the reaction rate in terms of effective gas concentrations at
lead to a deviation in r3PM,cal of ca. 8.5%. These uncertainties can explain
the catalyst surface, operating a fixed-bed reactor at equal operating
the standard deviation between r3PM,exp and r3PM,cal observed in Fig. 10.
conditions as for the three-phase methanation should lead to identical
In order to reach a better match of r3PM,exp and r3PM,cal , it is mandatory to
reaction rates.
improve the measurement precision, especially the confidence in the
values of HH2,pc which exerts the strongest influence on the calculated
values. Another source of discrepancy between experimental and cal- Acknowledgements
culated data is the different catalyst batches that have been used for the
experiments. Usually, for the development of a kinetic rate equation, This work was supported by the German Federal Ministry of
one batch of catalyst is used in the reactor. For the experimental work Economic Affairs and Energy. Special thanks go to Svenja Heling,
of this publication, six different catalyst batches have been used. Even Daniel Safai, Simone Nagel, and Ulli Hammann for their assiduous
though the catalyst batches were taken from the same catalyst charge, experimental work.

Appendix A

Estimation of internal mass transfer limitation using the Weisz-Prater criterion (Eq. (14)) with a reaction order m :
2
rcat ·(m + 1) r3PM,exp·ρcat
ψ= ·
2 DH2,L ·c H2,L (14)
The diffusion coefficient of H2 in the liquid phase is calculated using the correlation developed by Wilke and Chang [51] (Eq. (15)):
T ·ML0.5
DH2,L = 7.4·10−8
μ L ·VH0.6
2 ,molecule (15)
According to Wilke and Chang the H2 molecular volume is 14.3 cm3/(g·mol)
The Weisz-Prater must be smaller than 0.6 to prevent internal mass transfer limitation. This is confirmed by the calculation summarized in
Table 4.

Table 4
Results of the Weisz-Prater criterion estimation.

Parameter Units Value

c H2,L mol/m3 7.94–37.68

r3PM,exp 10−3 mol/(kg·s) 3.14–97.80
rcat 10−6 m 25–50
DH2,L 10−9 m2/s 20.46–41.98
ψ 10−2 2.37–58.6

158
J. Lefebvre et al.
References
[1] UNFCCC. Adoption of the Paris Agreement; Available from: http://unfccc.int/
resource/docs/2015/cop21/eng/l09r01.pdf.
[2] OCDE O. World Energy Outlook 2015. OECD Publishing; 2015.
[3] Schiebahn S, Grube T, Robinius M, Tietze V, Kumar B, Stolten D. Power to gas:
technological overview, systems analysis and economic assessment for a case study
in Germany. Int J Hydrogen Energy 2015;40(12):4285–94.
[4] Götz M, Lefebvre J, Mörs F, McDaniel Koch A, Graf F, Bajohr S, et al. Renewable
Power-to-Gas: a technological and economic review. Renewable Energy
2016;85:1371–90.
[5] Jentsch M, Trost T, Sterner M. Optimal Use of Power-to-Gas Energy Storage Systems
in an 85% Renewable Energy Scenario. 8th International Renewable Energy Storage
Conference and Exhibition (IRES 2013) 2014;46(0):254–61.
[6] Bailera M, Lisbona P, Romeo LM, Espatolero S. Power to Gas projects review: lab,
pilot and demo plants for storing renewable energy and CO2. Renew Sustain Energy
Rev 2017;69:292–312.
[7] Ball M, Weeda M. The hydrogen economy – vision or reality? Int J Hydrogen Energy
2015;40(25):7903–19.
[8] Wang W, Gong J. Methanation of carbon dioxide: an overview. Front Chem Sci Eng
2011;5(1):2–10.
[9] Kopyscinski J, Schildhauer TJ, Biollaz SMA. Production of synthetic natural gas
(SNG) from coal and dry biomass – a technology review from 1950 to 2009. Fuel
2010;89(8):1763–83.
[10] Rönsch S, Schneider J, Matthischke S, Schlüter M, Götz M, Lefebvre J, et al. Review
on methanation – from fundamentals to current projects. Fuel 2016;166:276–96.
[11] Gao Y, Meng F, Ji K, Song Y, Li Z. Slurry phase methanation of carbon monoxide
over nanosized Ni–Al2O3 catalysts prepared by microwave-assisted solution com-
bustion. Appl Catal A 2016;510:74–83.
[12] Meng F, Song Y, Li X, Cheng Y, Li Z. Catalytic methanation performance in a low-
temperature slurry-bed reactor over Ni–ZrO2 catalyst: effect of the preparation
method. J Sol-Gel Sci Technol 2016:1–10.
[13] Götz M, Bajohr S, Graf F, Reimert R, Kolb T. Einsatz eines Blasensäulenreaktors zur
Methansynthese. Chem Ing Tech 2013;85(7):1–7.
[14] Lefebvre J, Götz M, Bajohr S, Reimert R, Kolb T. Improvement of three-phase me-
thanation reactor performance for steady-state and transient operation. Fuel
Process Technol 2015;132:83–90.
[15] Khadem-Hamedani B, Yaghmaei S, Fattahi M, Mashayekhan S, Hosseini-Ardali SM.
Mathematical modeling of a slurry bubble column reactor for hydrodesulfurization
of diesel fuel: Single- and two-bubble configurations. Chem Eng Res Des
2015;100:362–76.
[16] Khadzhiev SN, Kolesnichenko NV, Ezhova NN. Slurry technology in methanol
synthesis (Review). Pet Chem 2016;56(2):77–95.
[17] Bozzano Giulia, Manenti Flavio. Efficient methanol synthesis: perspectives, tech-
nologies and optimization strategies. Prog Energy Combust Sci 2016;56:71–105.
[18] Basha OM, Sehabiague L, Abdel-Wahab A, Morsi BI. Fischer-Tropsch synthesis in
slurry bubble column reactors: experimental investigations and modeling – a re-
view. Int J Chem React Eng 2015;13(3).
[19] Deckwer W-D. Reaktionstechnik in Blasensäulen. Otto Salle Verlag.
[20] Rollbusch P, Bothe M, Becker M, Ludwig M, Grünewald M, Schlüter M, et al. Bubble
columns operated under industrial relevant conditions – current understanding of
design parameters. Chem Eng Sci 2014.
[21] Chaudhari RV, Ramachandran PA. Three phase slurry reactors. AIChE J
1980;26(2):177–201.
[22] Madon RJ, O'Connell JP, Boudart M. Catalytic hydrogenation of cyclohexene: Part
II. Liquid phase reaction on supported platinum in a gradientless slurry reactor.
AIChE J 1978;24(5):904–11.
[23] Gonzo EE, Boudart M. Catalytic hydrogenation of cyclohexene: gas-phase and li-
quid-phase reaction on supported palladium. J Catal 1978;52(3):462–71.
[24] Iglesias M, Edzang R, Schaub G. Combinations of CO/CO2 reactions with
Fischer–Tropsch synthesis. Catalysis and synthetic fuels: state of the art and outlook
2013;215(0):194–200.
[25] Eilers H, González MI, Schaub G. Lab-scale experimental studies of Fischer-Tropsch
kinetics in a three-phase slurry reactor under transient reaction conditions. Syngas
Convention – Fuels and Chemicals from Synthesis Gas: State of the Art 2
159
Fuel 217 (2018) 151–159
2016;275:164–71.
[26] Graaf GH, Stamhuis EJ, Beenackers AACM. Kinetics of low-pressure methanol
synthesis. Chem Eng Sci 1988;43(12):3185–95.
[27] Graaf GH, Winkelman JGM, Stamhuis EJ, Beenackers AACM. Kinetics of the three
phase methanol synthesis. Chem Eng Sci 1988;43(8):2161–8.
[28] Graaf GH, Smit HJ, Stamhuis EJ, Beenackers, Anthonius ACM. Gas-liquid solubi-
lities of the methanol synthesis components in various solvents. J. Chem. Eng. Data
1992;37(2):146–58.
[29] Götz M, Ortloff F, Reimert R, Basha O, Morsi BI, Kolb T. Evaluation of organic and
ionic liquids for three-phase methanation and biogas purification processes. Energy
Fuels 2013;27(8):4705–16.
[30] Kerlé D. Untersuchung der Löslichkeit von Gasen in ionischen Flüssigkeiten mit
Methoden der Molekulardynamischen Simulation [PhD Thesis] Rostock: Universität
Rostock; 2013.
[31] Prausnitz JM, Lichtenthaler RN, Azevedo ED. Molecular thermodynamics of fluid-
phase equilibria. 3rd ed. Upper Saddle River: Prentice Hall; 1999.
[32] Weisz PB, Prater CD. Interpretation of measurements in experimental catalysis. Adv
Catal 1954;6:143–96.
[33] Yang Lim J, McGregor J, Sederman AJ, Dennis JS. Kinetic studies of CO2 metha-
nation over a Ni/Al2O3 catalyst using a batch reactor. Chem Eng Sci
2016;141:28–45.
[34] Koschany F, Schlereth D, Hinrichsen O. On the kinetics of the methanation of
carbon dioxide on coprecipitated NiAl(O)x. Appl Catal B 2016;181:504–16.
[35] Rönsch S, Köchermann J, Schneider J, Matthischke S. Global reaction kinetics of CO
and CO2 methanation for dynamic process modeling. Chem Eng Technol
2016;39(2):208–18.
[36] Toppinen S, Salmi T, Rantakylä T-K, Aittamaa J. Liquid-phase hydrogenation ki-
netics of aromatic hydrocarbon mixtures. Ind Eng Chem Res 1997;36(6):2101–9.
[37] Madon RJ, Iglesia E. Catalytic reaction rates in thermodynamically non-ideal sys-
tems. J Mol Catal A: Chem 2000;163(1–2):189–204.
[38] Ussa Aldana, P. A., Ocampo F, Kobl K, Louis B, Thibault-Starzyk F, Daturi M et al.
Catalytic CO2 valorization into CH4 on Ni-based ceria-zirconia. Reaction me-
chanism by operando IR spectroscopy. Catalysis and synthetic fuels: state of the art
and outlook 2013;215(0):201–7.
[39] Schild C, Wokaun A, Koeppel RA, Baiker A. Carbon dioxide hydrogenation over
nickel/zirconia catalysts from amorphous precursors: on the mechanism of methane
formation. J Phys Chem 1991;95(16):6341–6.
[40] Pan Q, Peng J, Wang S, Wang S. In situ FTIR spectroscopic study of the CO2 me-
thanation mechanism on Ni/Ce0.5Zr0.5O2. Catal Sci Technol 2014;4(2):502–9.
[41] Bothra P, Periyasamy G, Pati SK. Methane formation from the hydrogenation of
carbon dioxide on Ni(110) surface - a density functional theoretical study. Phys
Chem Chem Phys 2013;15(15):5701–6.
[42] Qin S, Hu C, Yang H, Su Z. Theoretical Study on the Reaction Mechanism of the Gas-
Phase H2/CO2/Ni(3D) System. J Phys Chem A 2005;109(29):6498–502.
[43] Weatherbee GD, Bartholomew CH. Hydrogenation of CO2 on group VIII metals: II.
Kinetics and mechanism of CO2 hydrogenation on nickel. J Catal
1982;77(2):460–72.
[44] Falconer JL, Zaǧli AE. Adsorption and methanation of carbon dioxide on a nickel/
silica catalyst. J Catal 1980;62(2):280–5.
[45] Fujita S, Terunuma H, Nakamura M, Takezawa N. Mechanisms of methanation of
carbon monoxide and carbon dioxide over nickel. Ind Eng Chem Res
1991;30(6):1146–51.
[46] Peebles DE, Goodman DW, White JM. Methanation of carbon dioxide on nickel
(100) and the effects of surface modifiers. J Phys Chem 1983;87(22):4378–87.
[47] Kai T, Takahashi T, Furusaki S. Kinetics of the methanation of carbon dioxide over a
supported Ni-La2O3 catalyst. Can J Chem Eng 1988;66(2):343–7.
[48] Inoue H, Funakoshi M. Kinetics of methanation of carbon monoxide and carbon
dioxide. J Chem Eng Jpn 1984;17(6):602–10.
[49] Dew JN, White RR, Sliepcevich CM. Hydrogenation of carbon dioxide on nickel-
kieselguhr catalyst. Ind Eng Chem 1955;47(1):140–6.
[50] Chiang JH, Hopper JR. Kinetics of the hydrogenation of carbon dioxide over sup-
ported nickel. Ind Eng Chem Prod Res Dev 1983;22(2):225–8.
[51] Wilke CR, Chang P. Correlation of diffusion coefficients in dilute solutions. AIChE J
1955;1(2):264–70.