Efficient CO2 methanation over Ni/Al2O3 coated structured catalysts

Article abstract of DOI:10.1016/j.cattod.2016.04.019

In this work, CO₂ methanation was studied over macro-porous metal structures coated with Ni/Al₂O₃ catalyst. The results were benchmarked with conventional Ni/Al₂O₃ powder catalyst which was prepared by incipient wetness impregnation and characterized by several physico-chemical techniques. Macro-porous catalytic supports were manufactured by innovative and highly reproducible robocasting technique ? three dimensional fibre deposition (3DFD). Supports were coated with Ni/Al₂O₃ suspension to achieve sufficient catalytic coating. After characterization, catalytic structures were tested in a tubular reactor in CO₂ methanation reaction at temperatures between 250 and 450°C. This study shows the effect of the coating suspension composition on the properties of catalytic coatings, as well as how CO₂ conversion, methane selectivity and catalyst stability are affected by the architecture of the structured catalyst.

Full text of DOI:10.1016/j.cattod.2016.04.019

G Model
CATTOD-10167; No. of Pages10
ARTICLE IN PRESS
Catalysis Today xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Efficient CO2 methanation over Ni/Al2O3 coated structured catalysts
Simge Danacia,b,d, Lidia Protasovaa,∗, Jasper Lefeverea, Laurent Bedelb, Richard Guiletc,
Philippe Martyb,d
a Flemish Institute for Technological Research − VITO, Boeretang 200, B-2400 Mol, Belgium
b Liten/DTBH/SCTR/LER −CEA Grenoble, 17 rue des Martyrs, F-38054 Grenoble, France
c Université de Toulouse, UPS, Laboratoire de Génie Chimique (LGC UMR 5503), 4 Allée Emile Monso, F-31432 Toulouse, France
d LEGI, UJF-Grenoble 1/Grenoble-INP/CNRS, BP53, Grenoble F-38051, France
a r t i c l e i n f o
a b s t r a c t
Article history:
In this work, CO2 methanation was studied over macro-porous metal structures coated with Ni/Al2O3
catalyst. The results were benchmarked with conventional Ni/Al2O3 powder catalyst which was pre-
pared by incipient wetness impregnation and characterized by several physico-chemical techniques.
Macro-porous catalytic supports were manufactured by innovative and highly reproducible robocasting
technique − three dimensional fibre deposition (3DFD). Supports were coated with Ni/Al2O3 suspension
to achieve sufficient catalytic coating. After characterization, catalytic structures were tested in a tubu-
lar reactor in CO2 methanation reaction at temperatures between 250 and 450◦C. This study shows the
effect of the coating suspension composition on the properties of catalytic coatings, as well as how CO2
conversion, methane selectivity and catalyst stability are affected by the architecture of the structured
catalyst.
Received 27 November 2015
Received in revised form 13 April 2016
Accepted 16 April 2016
Available online xxx
Keywords:
CO2 methanation
Ni/Al2O3 catalytic coating
Structured catalytic supports
© 2016 Elsevier B.V. All rights reserved.
1. Introduction
the addition of a second metal (Fe, Co, Cu) to Ni/ZrO2 catalysts for
methanation reaction have been reported. Iron was found to be a
In recent years, methanation reaction, a well-known process,
has drawn great interest in the production of methane and utiliza-
tion of CO2 [1–3]. The catalytic hydrogenation of carbon dioxide
to methane, also called a Sabatier reaction, is reversible and very
exothermic (1). This reaction is usually performed between 250 and
500◦C.
suitable second metal for catalyst at low-temperature CO2 metha-
nation [13]. Nickel-based catalysts are the most widely studied for
the methanation reaction due to their high activity and competitive
cost. Among metal oxides, alumina is the most common catalytic
support used in heterogeneous catalysis due to its low cost, good
thermal stability, high specific surface area and high interaction
with deposited metals leading to high catalytic activity [14].
So far, catalytic methanation has been mostly investigated in
packed-bed and fluidized-bed reactor configurations [15]. How-
ever, the Sabatier process is usually performed in packed-bed
reactors due to its economic efficiency [3]. Because of the high
exothermicity of Sabatier reaction, temperature control is a chal-
lenge. For fast chemical reactions, using a packed-bed reactor can
lead to hot spots and catalyst deactivation due to the sintering
of the catalyst [16]. It is therefore necessary to remove the heat
more efficiently. Furthermore, at high gas space velocities, exist-
ing packed-bed systems can lead to a high pressure drop which is
dependent on the particle size of the catalyst as well as its shape
and packing [17]. Flow mal-distribution resulting in non-uniform
contact time is an additional disadvantage of the packed-bed sys-
tems. Decreasing the particle size, in order to improve the activity
of the catalyst and mass transfer, leads to even greater increase
in pressure drop [18]. Thus, macro-structured supports with cat-
alytically active coatings are seen as an important part of process
CO2 g + 4H2 g ↔ CH4 g + 2H2O(g ꢀH298K = −165kJ/mol
(1)
(
)
(
)
(
)
)
Catalysts containing transition metals of the VIII, IX, X and XI
groups have been investigated for methanation [3–5]. In the last
decade, bimetallic catalysts have also attracted a lot of attention
due to their higher activity. Previous studies showed that catalysts
containing ruthenium supported on Al2O3 were highly selective
for methane [6–8]. For example, optimal ruthenium content was
found to be 0.6 wt.% for the maximum production of methane
over a Ni-Fe-Ru/Al2O3 catalyst [9]. Some studies have shown that
active oxygen sites (vacancies) can interact with active metals to
improve the performances of catalysts. Therefore, active materials
with mobile oxygen (e.g. CexZr1-xO2) and bimetallic catalysts (Ni-
Co, Ni–Mo, Ni-Fe) were investigated [10–12]. Recently, the effect of
∗
Corresponding author.
E-mail address: lidia.protasova@vito.be (L. Protasova).
http://dx.doi.org/10.1016/j.cattod.2016.04.019
0920-5861/© 2016 Elsevier B.V. All rights reserved.
Please cite this article in press as: S. Danaci, et al., Efficient CO2 methanation over Ni/Al2O3 coated structured catalysts, Catal. Today
(2016), http://dx.doi.org/10.1016/j.cattod.2016.04.019

G Model
CATTOD-10167; No. of Pages10
ARTICLE IN PRESS
2
S. Danaci et al. / Catalysis Today xxx (2016) xxx–xxx
Fig. 1. A unit cell of the 3DFD-structure.
intensification and sustainable chemistry, as they allow controlled
mass and heat transfer, low pressure drop and thus better control
of the process.
metallic structures. 3DFD (also called direct-writing or robocast-
ing) comprises the extrusion of a highly viscous paste loaded with
metallic or ceramic particles through a thin nozzle. After sintering, a
highly reproducible and periodic porous structure is obtained. The
process variables include the nozzle opening (fibre thickness), the
geometry of nozzle (fibre shape), the inter-fibre distance (pore size)
and the stacking of the layers (architecture). The micro-porosity
and surface roughness of the fibres can be also controlled [31].
Structured supports or 3DFD robocasted periodic porous structures
can provide several advantages such as flexible design, significantly
better heat and mass transfer as well as lower pressure drop and
excellent mechanical properties [20,32,33].
In this work, we aim to manufacture an efficient structured cat-
alyst for the methanation reaction. For this purpose, various 3DFD
catalytic substrates were produced and functionalized with the
Ni/Al2O3 catalytic coating using wash-coating technique. Differ-
ent geometries of the structured supports were compared in the
methanation process. The results were benchmarked with conven-
tional powder Ni/Al2O3 catalysts in a packed-bed reactor.
In order to overcome the above mentioned limitations, several
improved reactor and catalysts designs have been proposed. One of
the possibilities of heat transfer improvement for CO2 methanation
is the use of a structured micro-channel reactor [19–21] with a thin
layer of catalyst on the reactor channel walls that leads to improved
heat and mass transfer. In the case of heat transfer improved reac-
tors, metal based supports provide better temperature control due
to their improved heat transfer properties over conventional cat-
alysts (alumina, zeolites etc.). As for structured reactors, metallic
plates [21], foils [22,23], felts [24], monoliths [25] and foams [26,27]
are used as structured supports. The thickness of the catalytic layer
is an important parameter and depends not only on the coating
process but also on the coating slurry composition as well as on its
viscosity. For example, it was reported that monoliths with a wash
coated layer thicker than ca. 50 ␮m suffer from decreased activity
as a result of internal diffusion limitations [28].
In recent years, some data have been published related to both
CO and CO2 methanation in structured reactors. However, most
of the published studies focused on CO methanation due to high
accessibility of the syngas (a mixture of mainly CO + H2) in indus-
try. One of the successful micro-channel reactor designs without
heat transfer limitations was proposed for the CO methanation in
presence of CO2 and O2 in 2005 [23]. Microstructured foils were
coated with Ru/SiO2 and Ru/Al2O3 catalysts. It was found, that in
the presence of CO2, the conversion of CO is higher only at temper-
atures between 200 and 250◦C. Methanation of CO2 dominates at
temperatures >250◦C [23]. In the case of CO2 methanation, a limited
number of studies are available. The reactor operation at uniform
temperatures is a challenge at high temperatures (>250◦C). There-
fore a potentially interesting microchannel reactor was designed
which provides flexible thermal management [19]. A porous metal
felt was used as a support material for Ru and Rh catalysts and
the structured catalysts were tested in CO2 methanation. Another
remarkable study on CO2 methanation reaction was published by
NASA for long term space missions in 2011 [29]. Results showed
that an improved MicrolithTM reactor design with catalytic sup-
ports has a significant impact on the performance due to its better
heat and mass transfer, lower pressure drop, high surface area and
short contact time compare to the traditional packed-bed mode.
Recently, the additive manufacturing (AM) technologies have
started being used for the manufacture of macro-structured cat-
alytic supports. One of these technologies is 3-Dimensional Fibre
Deposition (3DFD) developed at VITO [30]. This manufacturing
technique enables the production of periodic porous ceramic and
2. Experimental
2.1. Manufacture of macro-porous structured supports
The stainless steel (316L) supports were prepared using the
3DFD solid free forming technique as described elsewhere [31].
With this technique, a highly viscous stainless steel paste was
extruded through a thin nozzle. Porous supports were built layer-
by-layer by computer controlled movement in x, y and z-directions.
Nozzles with a diameter of 400 and 600 ␮m were used to manufac-
ture 3D-structures with 1-1 and 1-3 stacking positions (Fig. 2). The
1-1 stacking 3DFD structure has straight channels in the direction
of the flow, while the 1-3 structure consists of ‘zigzag’ channels in
the direction of the flow. Structures were dried at room tempera-
ture for 2 days. Then they were sintered at 1300◦C for 4 h and cut
into cylindrical shape with 20.1 mm diameter and 20 mm length.
Before coating, all supports were cleaned in iso-propanol for 10 min
under ultrasonic treatment to remove light oils and dirt from the
surface. Samples were dried overnight at 100◦C. The overview of
the manufactured structures is given in Table 1. A unit cell of the
structure is given in Fig. 1. Macro-porosity and specific surface area
of the 3DFD supports were calculated according to the following
equations:
ꢀ
ꢁ
1.57a2M − 0.01
2M2 a − c
Pmacro−porosity, % = 1 −
× 100
(2a)
(
)
Please cite this article in press as: S. Danaci, et al., Efficient CO2 methanation over Ni/Al2O3 coated structured catalysts, Catal. Today
(2016), http://dx.doi.org/10.1016/j.cattod.2016.04.019

G Model
CATTOD-10167; No. of Pages10
ARTICLE IN PRESS
S. Danaci et al. / Catalysis Today xxx (2016) xxx–xxx
3
Fig. 2. Cross-sectional images of the structures with 1-1 (up) and 1-3 (down) stacking positions.
Table 1
Overview and images of 3DFD structures.
Sample codea
Stacking position
Nozzle (mm)
0.4
Inter-fibre distance (mm)
Macro-porosity (%)
SSAb (mm2/mm3)
4A08
1-1
0.8
70
2.7
4A1
4B1
1-1
1-3
0.4
0.4
1
1
74
74
2.4
2.4
6B1
1-3
0.6
1
67
2
a
Sample code: first characters (4 or 6) refer to the nozzle diameter (0.4 or 0.6 mm, respectively), second character refers to the stacking positions (A: 1-1 and B: 1-3), third
character points inter-fibre distance (0.8 or 1 mm).
b
SSA: specific surface area; surface area (SA, mm2) to volume (V, mm3) ratio per unit cell.
3.14 aM − c
(
)
2.2. Catalyst preparation
SSAsupport
=
.
(2b)
M2 a − c
(
)
Powder catalyst was prepared by impregnation method starting
from boehmite powder, AlO(OH) (Sasol, Germany, average particle
size 50 ␮m (D90)) and an aqueous solution of nickel nitrate hex-
ahydrate (PANREAC). Boehmite (10 g) was added into an aqueous
solution of nickel nitrate (5.10−4 M, 50 ml) under stirring and kept
at room temperature for 24 h. The Ni-loading was calculated to be
12 wt.%. The mixture was dried by freeze drying (HETO Powerdry
LL3000) under vacuum at 15◦C. The dried powder was calcined
at 450◦C for 10 h under atmospheric pressure with a heating rate
M = a + n, (2c)where a is fibre thickness (mm), n is inter-fibre dis-
tance (mm) and M is axial centre difference between two fibres
(mm). Stacking constant c refers to stacking length between two
layers of fibres in z-direction considering an anisotropic pore
architecture. In case of 3DFD support, c constant of 0.0677 mm is
obtained from SEM images.
Please cite this article in press as: S. Danaci, et al., Efficient CO2 methanation over Ni/Al2O3 coated structured catalysts, Catal. Today
(2016), http://dx.doi.org/10.1016/j.cattod.2016.04.019

G Model
CATTOD-10167; No. of Pages10
ARTICLE IN PRESS
4
S. Danaci et al. / Catalysis Today xxx (2016) xxx–xxx
X-ray diffraction, XRD (PANalytical X’Pert Pro) (␭ = 1.5405 Å) at
40 kV was used to examine the phase and crystallinity of the pow-
der catalyst before and after calcination.
Dip-coaꢀng
3DFD
structured
catalysts
3DFD
with
Drying &
sintering
supports
Ni/alumina
suspension
Reduction temperature profile of powder Ni/Al2O3 catalyst
was performed using thermoanalyzer (TGA, NETZSCH STA-449) at
600◦C, heating rate of 5◦C/min. Prior to TGA measurement, powder
catalyst was reduced under 80% H2 flow at 450◦C for 2 h.
Particle size distribution (PSD) was detected by wet sieving
method by PSD analyser (Microtrac S3500).
Rheometer (Kinexus Rheometer) was used to determine the vis-
cosity of the coating slurry at room temperature as a function of the
shear rate.
Adhesion strength of the coating was evaluated by ultrasonic
treatment. The coating morphology and the thickness of the coat-
ing on the cross-sections of the structured catalyst were examined
by scanning electron microscopy (SEM; FEG JSM6340F, JOEL) and
Optical Microscopy (Zeiss, Stereo Discovery V12) with imager (type
M2m). The loading of the deposited coating was determined by
weighing, while the average coating thickness was measured by
SEM.
Energy Dispersive X-ray Fluorescence (EDXRF) spectrometer by
HE XEPOS (Spectro Analytical Systems, Kleve, Germany) was used
for the elemental analysis of impregnated Ni over Alumina cata-
lysts.
Fig. 3. Preparation steps of the structured catalysts.
of 1◦C/min. At this temperature transformation of boehmite into
␥-alumina takes place (Eq. (3)). After calcination, dried powder cat-
alysts were sieved to achieve the particle diameter of 25 ␮m. In
the case of coating preparation, Ni/Al2O3 powder catalyst was wet
ball-milled (10 g of ZrO2 spheres were used per gram of catalyst)
at 300 rpm for 60 min (15 min milling, 45 min rest) by Planetary
Micro Mill (Fritsch Pulverisette-5). Ball-milled samples were again
freeze-dried.
AlO OH → 1/2Al O3 + 1/2H2O
(3)
(
)
2
Coating slurry was prepared as follows: 1–5 g PVA (Polyvinyl
alcohol, Fluka Chemica, 100.000), and 0.5–2 g colloidal silica
(LUDOX HS-40, Sigma Aldrich) were added to 73 ml deionised
water at 60◦C and left without stirring overnight. Powder Ni/Al2O3
catalyst (20 g) and 1 ml acetic acid (0.2 M, Merck) were added
into the slurry. The solvent/catalyst ratio was 3.6–3.9%. The mix-
ture was stirred at room temperature for 24 h. The dip-coating
of 3DFD supports was performed manually: the 3DFD supports
were immersed into the tank containing coating suspension for
60 s, and the excess of the suspension was removed by air blow
gun. The coating was repeated until achieve the loading of ca.
0.17–0.19 g/cm3support. The catalyst loading was calculated as fol-
lows: Loading = (Gsupport + cat − Gsupport)/Vsupport where, G (g) is
the weight of the structured support (+catalyst) and the V (cm3)
is the volume of the structured support. Then, the samples were
dried at 100◦C overnight and calcined at 550◦C for 4 h. After all,
structured catalysts 4A08, 4A1, 4B1 and 6B1 were obtained with
the Ni/Al2O3 loading of 0.18, 0.19, 0.17, 0.18 g/cm3, respectively.
The preparation steps are given in Fig. 3.
2.4. Catalytic activity
The scheme of the methanation setup is presented in Fig. 4. It
is worth to mention that the setup is not optimal for this reac-
tion and was built for the basic screening of the catalysts. A quartz
tubular reactor (24 mm diameter and 100 mm length) was used
and equipped with a K-type thermocouple that was installed at
inlet and outlet sides of the quartz tube for continuous tempera-
ture measurement. Catalysts were packed in the middle of reactor
and fixed with quartz wool. The reactor was placed in the middle
of the furnace. In order to have the fair comparison of the sam-
ples with the different geometry (stacking, macro-porosity), the
same amount of catalyst was used for each experiment. Before the
reaction test, catalysts were activated under a continuous flow of
H2/He (80/20%) at the total rate of 100 ml/min and temperature
of 450◦C (heating rate 10◦C/min) for 2 h under atmospheric pres-
sure. After reduction, temperature of the furnace was adjusted to
the reaction temperature under continuous flow of helium. Metha-
nation reaction was performed at temperatures between 250 and
2.3. Characterization
The specific surface area of the catalysts was measured by N2
sorption analyser (Quantachromie ASIQM 0002-4) at −196◦C using
the BET method, each sample was degassed at 200◦C.
Fig. 4. Experimental setup for methanation reaction.
Please cite this article in press as: S. Danaci, et al., Efficient CO2 methanation over Ni/Al2O3 coated structured catalysts, Catal. Today
(2016), http://dx.doi.org/10.1016/j.cattod.2016.04.019

G Model
CATTOD-10167; No. of Pages10
ARTICLE IN PRESS
S. Danaci et al. / Catalysis Today xxx (2016) xxx–xxx
5
Table 2
Characteristics of the powder support and catalyst.
Powder sample
Calcination temperature (◦C)
BET surface area (m2/g)
Pore volume (cm3/g)
Pore size (Å)
γ-Al2O3
γ-Al2O3
Ni/Al2O3
Ni/Al2O3
450
550
450
550
334
271
164
143
0.43
0.44
0.30
0.32
23
25
25
31
0.001
-1E-17
-0.001
-0.002
-0.003
-0.004
-0.005
100
200
300
400
500
600
Temperature (°C)
Fig. 5. DTG profile of the powder Ni/Al2O3 catalyst.
450◦C under atmospheric pressure. Carbon dioxide and hydrogen
were continuously fed into the reactor together with helium car-
rier gas at the total rate of 100 ml/min with feed composition of
CO2:H2:He = 1:4:15.
respectively). After impregnation with nickel, surface area and pore
volume decreased further due to the pore blockage [35]. However,
negligible difference was recorded in pore volume and specific sur-
face area in the case of Ni/Al2O3 calcined at 450 and 550◦C. In order
to keep the surface area higher, final catalysts were calcined at
450◦C.
Calcined Ni/Al2O3 catalyst was analysed by derivative thermo-
gravimetric (DTG) method under H2/Ar (5/95%) atmosphere in
order to determine NiO to Ni reduction temperature. Temperature
was increased from 25 to 600◦C with a heating rate of 20◦C/min.
It is known that nickel catalyst shows different reduction profile
depending on the interactions between nickel and metal oxide sup-
port [36]. DTG results are given in Fig. 5. The reduction process
occurs at a temperature range of 350–520◦C. A single reduction
peak was observed at temperature of 478◦C. This temperature pro-
file is in agreement with the reported literature [37–39].
Fig. 6 shows the XRD pattern of the calcined and reduced
Ni/Al2O3 catalysts. Alumina phase evolution was studied depend-
ing on the thermal treatment conditions which is in perfect
agreement with previously reported data [40]. It can be seen that
after calcination, NiO and ␥-Al2O3 phases are observed that con-
firms the transformation of boehmite to ␥-Al2O3 (could not be
completely distinguished with NiAl2O4). After reduction, NiO sig-
nals almost disappeared (43.1◦ and 62.9◦), that indicates nearly full
reduction of NiO. The formation of metallic Ni cannot be clearly
seen due to the overlapping of the signals of alumina, Ni aluminate
and Ni0 (44–47◦ and 65–69◦ 2␪). However, there is a clear indica-
tion of the formation of metallic Ni, i.e. no shift of the peak at 37.1◦
(alumina and Ni aluminate) was detected, while there is a clear shift
to “Ni-rich” direction of the signals where the signal of metallic Ni
could be present: from 45.6 to 45.2◦ and from 66.6 to 65.9◦.
Gas chromatography (450-GC, Bruker, Germany) was used as
a common method for the analysis of reagents and products. The
catalytic activity and stability were determined by monitoring CO2
conversion as a function of time-on-stream. Flame ionization detec-
tor (FID) and thermal conductivity detector (TCD) were used to
measure CH4 and CO2 concentrations, respectively. Temperature
of the TCD and FID detectors was maintained at 300◦C. The cali-
bration of peak areas (GC) was performed using a known reactant
gas composition without a catalyst. Conversion (X), selectivity (S)
and yield (Y) were calculated based on peak areas from calibration
using the following equations:
ꢂ
ꢃ
Cin CO − Cout CO
(
)
(
)
2
2
Conversion X CO % =
× 100
(4a)
×100 (4b)
(4c)
(
)
2
Cin CO
(
)
2
ꢂ
ꢃ
Cout(CH
Cout(CH ) + Cout(C )) + Cout(C
4
Selectivity S CH %=
( )
4
H
H )
6
4
2
4
2
ꢂ
ꢃ
Cout(CH
)
4
Yield Y CH % =
× 100
(
)
4
Cin CO
(
)
2
3. Results and discussion
3.1. Characterization of powder catalyst
As it was already mentioned above, Ni/Al2O3 catalysts were
prepared by impregnation of commercial boehmite with narrow
particle size distribution that provides more homogeneous impreg-
nation [34]. Nickel content was kept constant (13 wt.% according to
elemental analysis), and different calcination temperatures were
tested (450–550◦C). The results are given in Table 2. The sur-
face area of the alumina significantly drops when the calcination
temperature increases from 450 to 550◦C (334 and 271 m2/g,
3.2. Characterization of the coating suspension
The coating quality depends on the coating slurry properties
(solid content and viscosity). Moreover, smaller particle size results
Please cite this article in press as: S. Danaci, et al., Efficient CO2 methanation over Ni/Al2O3 coated structured catalysts, Catal. Today
(2016), http://dx.doi.org/10.1016/j.cattod.2016.04.019

G Model
CATTOD-10167; No. of Pages10
ARTICLE IN PRESS
6
S. Danaci et al. / Catalysis Today xxx (2016) xxx–xxx
Fig. 6. XRD patterns of calcined and reduced Ni/Al2O3 catalysts.
and 0.5 Pa s are preferred at shear rate of 10 s−1 [45]. Viscosity of
the suspension with 3 wt.% PVA was found to be 0.13–0.26 Pa s
at shear rates between 10 and 100 s−1. Suspensions with ≤3 wt.%
PVA showed pseudo-Newtonian behaviour, i.e. viscosity remained
nearly constant with increasing shear rate, which is favourable in
order to have a homogeneous coating.
To study the effect of the PVA content in the suspension on
the coating properties, Ni/Al2O3 suspensions with various PVA
concentrations were deposited on 3DFD stainless steel supports.
The amount of deposited catalysts was determined by weight
difference before and after coating and calcination. According to
the weight measurements, after each immersion, catalyst loading
increases with increasing PVA concentration. Optical microscope
(OM) images show that the coating obtained using the suspension
with 1 wt.% PVA results in cracks on the surface (Fig. 9a). Addition-
ally, the coating made from the suspension with 5 wt.% PVA (Fig. 9c)
is observed to be non-homogeneous because of the high loading
after each immersion. Number of immersions was decreased with
increasing PVA concentration. The results of viscosity measure-
ments and OM images showed that using the concentration of PVA
of ca. 3% in coating suspension results in optimal behaviour of the
slurry and later in homogeneous coating (Fig. 9b).
Fig. 7. Cumulative particle size distribution of the wet-milled catalyst.
3
2.5
2
Shear rate= 1 s-1
Shear rate= 10 s-1
Shear rate= 100 s-1
1.5
1
0.5
0
0
1
2
3
4
5
6
PVA concentraꢀon (wt.%)
3.3. Characterization of the coating on 3DFD structures
Fig. 8. Coating suspension viscosities vs PVA concentration at different shear rates.
It is known from previous studies that adhesion strength of the
catalytic coating strongly depends on the binder concentration in
the coating suspension [32]. In this work, SiO2 (Ludox) was used
as a binder, and effect of its concentration on the coating adhesion
strength was investigated as well as the influence of PVA concentra-
tion. The adhesion strength was evaluated according to a method
described in the literature [46], based on the measurement of the
weight loss caused by exposure of the sample to ultrasound. The
coating thickness was measured by SEM before and after treatment.
It is reported that it is more difficult to achieve a good adhesion of
inorganic coatings to a metal surface than to a ceramic one. The
special chemical treatment of the metal surface by e.g. acids is nec-
essary in order to increase specific surface area [47–49]. Therefore,
impurities from pores can be removed to achieve better adhesion
between the metal surface and the coating. Coated stainless steel
3DFD structures were treated in an ultrasonic bath filled with dis-
tilled water for 1, 15, 30 and 60 min and then dried. The weight loss
in a coating with better adhesion: for a good dip-coating, par-
ticle size below 10 ␮m is recommended [41–43]. Thus, Ni/Al2O3
particles with D10 = 0.6 D50 = 1.5, D90 = 3.3 and D99 = 7.2 ␮m were
obtained by wet ball-milling process. The corresponding cumula-
tive particle size distribution is shown in Fig. 7.
Coating suspension is a heterogeneous mixture containing solid
catalyst particles, solvent (e.g. water) binder (e.g. PVA) and acid. The
binder is used to provide viscosity and sedimentally stable coating
suspension [44]. The acid is added to avoid the agglomeration of
the alumina particles in suspension and guarantee the suspension
stability of the particle over time [34]. Coating suspensions con-
taining 1, 3 and 5 wt.% PVA were prepared. Rheometer was used
to determine the single point viscosity at room temperature (25◦C)
as a function of shear rate (1, 10 and 100 s−1). Fig. 8 gives plotted
viscosities versus PVA concentrations in the coating suspensions.
In order to achieve homogeneous coating, viscosities between 0.03
Please cite this article in press as: S. Danaci, et al., Efficient CO2 methanation over Ni/Al2O3 coated structured catalysts, Catal. Today
(2016), http://dx.doi.org/10.1016/j.cattod.2016.04.019

G Model
CATTOD-10167; No. of Pages10
ARTICLE IN PRESS
S. Danaci et al. / Catalysis Today xxx (2016) xxx–xxx
7
Fig. 9. Optical microscope (OM) images of the catalytic coatings (loading ca. 0.15 g/cm3) obtained from the suspensions with different PVA content: (a) 1 wt%, (b) 3 wt%, (c)
5 wt%.
Table 3
However, due to toxicity of HF and low corrosion resistance of 316-
L stainless steel, 3DFD supports were treated with 5% nitric acid
Effect of the suspension composition on the coating adhesion.
solution. Samples were placed in US bath for 10 min then cleaned
with iso-propanol. Acid treated supports were coated with the same
amount of catalyst (ca. 0.15 g/cm3) and the adhesion strength was
tested. After US treatment for 10 min in water, the weight loss
was found to be 1.4 and 2.5 wt.% for treated and untreated sam-
ple, respectively. This result confirms that acid treatment improves
the adhesion strength due to the increased roughness of the sur-
face. To study the effect of calcination temperature on the coating
adhesion, coated 3DFD structures were calcined at 450 and 550◦C.
The adhesion strength was tested as described above. The results
showed that calcination temperature has no influence on the adhe-
sion strength of the coating.
Scanning electron microscopy (SEM) was used to investigate the
coating morphology of the sample. The cross-section images reveal
an adhesive coating on the surface of the stainless steel structures
before (Fig. 10a) and after 1 h ultrasonic treatment (Fig. 10b). The
right light-grey part the structure in Fig. 10 corresponds to the
stainless steel fibre and the left dark-grey part refers to the cat-
alytic coating. The thickness of the coating was measured to be 18
and 12 ␮m before and after ultrasonic treatment, respectively.
It is noteworthy, that the surface of the 3DFD stainless steel
structures was very rough (because of sintering temperature and
acid treatment), and to make the measurements and associated
judgement was rather difficult; moreover, some loss of the coating
could occur during the sample preparation for SEM imaging. Thus,
basically it could be concluded that there is no significant change
in coating thickness before and after ultrasonic treatment.
US time (min)
Coating weight loss (wt.%)
1% PVA
2% SiO2
3% PVA
2% SiO2
5% PVA
2% SiO2
3% PVA
0.5% SiO2
1
15
30
60
0.6
2.1
2.5
2.9
0.1
0.9
1.1
2.5
0.4
1.2
2.5
3.4
1.3
4.9
12.5
17.3
values are given in Table 3. Results show that PVA concentration
has no significant effect on adhesion strength of the coatings.
After 1 h ultrasonic treatment (US), 17.3% coating weight loss
was detected in case of the coating obtained from the suspension
with 0.5 wt.% SiO2. After addition of 2 wt.% SiO2 into the coating
slurry, the weight loss of the coating after 1 h US treatment was
found to be only 2.5%. This result was considered to be satisfying
and no further increase of SiO2 content was done in order to avoid
the influence of SiO2 on the catalytic properties of Ni/Al2O3. Based
on above described results, the optimal composition of the suspen-
sion was found to be as follows: 3% PVA, 2% Ludox, 1% acetic acid,
20% catalyst powder, and water. Consequently, this composition
was determined according to the results from rheology experi-
ments (PVA effect) and US treatment (SiO2 effect) to achieve more
homogeneous and adhesive coating.
As it was mentioned above, the surface treatment of the stain-
less steel substrates could further improve the adhesion strength
of catalytic coating. The stainless steel supports could be treated
by various acids or mixtures, e.g. hydrofluoric and nitric acid [48].
Please cite this article in press as: S. Danaci, et al., Efficient CO2 methanation over Ni/Al2O3 coated structured catalysts, Catal. Today
(2016), http://dx.doi.org/10.1016/j.cattod.2016.04.019

G Model
CATTOD-10167; No. of Pages10
ARTICLE IN PRESS
8
S. Danaci et al. / Catalysis Today xxx (2016) xxx–xxx
Fig. 10. SEM images of the coated 3DFD structure: before (a) and after (b) ultrasonic adhesion strength test.
Table 4
Conversion, selectivity and yield for various Ni/Al2O3 catalysts.
Catalyst
Powder
WHSV (h−1
)
Temp. inlet (◦C)
250
Temp. outlet (◦C)
CO2 Conversion (%)
CH4 Selectivity (%)
CH4 Yield (%)
750
236
236
237
284
283
278
333
334
332
382
381
381
450
39
34
33
61
60
59
74
73
71
76
74
71
66
97
97
98
95
97
98
97
97
99
98
98
99
97
38
33
32
58
58
58
72
71
70
74
73
70
65
1500
2600
750
1500
2600
750
1500
2600
750
1500
2600
1500
300
350
400
450
4A08
4A1
4B1
6B1
1500
1500
1500
1500
250
300
350
400
450
230
277
326
376
436
5
97
99
98
97
97
5
17
74
87
88
17
73
84
85
250
300
350
400
450
248
298
349
403
449
5
98
96
98
99
99
5
15
73
86
88
14
72
85
87
266
294
347
399
443
270
298
355
413
449
59
66
85
90
89
95
97
98
98
98
56
64
83
88
87
266
297
349
400
450
264
296
351
408
450
52
60
85
91
90
96
98
96
98
98
50
59
82
89
88
3.4. Catalytic activity, selectivity, and stability
10%. When the temperature is raised to 400◦C, no change in CO2
conversion is observed. It is known that the Gibbs free energy (ꢁG)
is a negative number as long as the reaction is spontaneous. As
for an exothermic reaction, after a certain temperature, reaction
achieves the thermodynamic equilibrium. After thermodynamic
equilibrium, the ꢁG increases rapidly and becomes positive [49].
When it became positive, the reverse process is spontaneous. In
case of methanation, reaction changes its way to methane reform-
ing [50] to balance the heat released. This explains the decreased
conversion of the CO2 at reaction temperature >350◦C. Conversion
results show that operating temperature of 350◦C was the most
suitable temperature for the powder catalyst to have an effective
carbon dioxide methanation. The selectivity to methane was high
and stable at around 95–99%, only small amount of by-products
(C2H4 and C2H6) was detected (Table 4). Above 350◦C, the methane
The methanation reaction was carried out with powder packed-
bed catalyst and coated 3DFD structures in a tubular reactor. The
overall results are given in Table 4. The CO2 conversion on powder
catalyst at different space velocities and temperatures is plotted in
Fig. 11. At all tested temperatures, the conversion decreases slightly
with the decrease of the weight hourly space velocity (WHSV, h−1),
which is explained by the shorter contact time of the gases and cat-
alyst in the reactor. The WHSV of the reactant gas is ranged between
750 and 2600 h−1 at a constant feed gas ratio (CO2:H2 = 1:4). Total
gas feed was kept constant at 100 ml/min by dilution. Methanation
reaction over powder catalyst was performed at temperatures from
250 to 450◦C. Conversion increases by ca. 20% when temperature
is changed from 250 to 300◦C, while during the further increase of
the temperature to 350◦C the difference was found to be only ca.
Please cite this article in press as: S. Danaci, et al., Efficient CO2 methanation over Ni/Al2O3 coated structured catalysts, Catal. Today
(2016), http://dx.doi.org/10.1016/j.cattod.2016.04.019

G Model
CATTOD-10167; No. of Pages10
ARTICLE IN PRESS
S. Danaci et al. / Catalysis Today xxx (2016) xxx–xxx
9
80
70
60
50
40
30
400°C
350°C
300°C
250°C
500
1000
1500
2000
2500
3000
WHSV (h-1)
Fig. 11. CO2 conversion vs WHSV at different temperatures (powder Ni/Al2O3 catalyst).
Fig. 12. Methanation reaction at different temperatures (WHSV 1500 h−1).
yield was found not to be significantly higher, and moreover, for-
mation of by-product occurs.
effect of the architecture on the mass-transfer was observed ear-
lier for methanol-to-olefins reaction [32]. There is also an indication
that heat transfer properties of the 1-3 stacking structured cata-
lyst are different in comparison with 1-1 stacking due to nonlinear
positioning of the fibres. Temperature is considered a major fac-
tor which affects the reaction rate. This can result in higher actual
temperature inside the 1-3 stacking structured catalysts due to
higher heat released by the exothermic methanation reaction. Thus,
at lower temperatures, this difference combining heat- and mass-
transfer properties and different residence time distribution (RTD)
in different architectures resulted in >40% difference in CO2 conver-
sion (see Fig. 12). Generally, 3DFD structured catalysts showed ca.
90% CO2 conversion at above 370◦C without showing any geometry
effect anymore.
In order to study the catalysts stability, methanation reaction
was performed over 3DFD structured and powder catalysts at 350◦C
for prolonged time. Gas samples were automatically analysed every
20 min time-on-stream (TOS). Fig. 13 shows the CO2 conversion as
a function of TOS. The initial CO2 conversions were 80% and 73% for
3DFD structured catalyst (4B1 sample) and powder, respectively.
Powder catalyst showed 8% decrease of CO2 conversion already
after 45 h of the experiment, while in case of 3DFD structured cat-
Fig. 12 shows the comparison of CO2 conversion on the pow-
der packed-bed catalyst and 3DFD structured catalysts. It can be
seen that temperature significantly affects the conversion of car-
bon dioxide. At temperatures above ca. 340◦C, all 3DFD structured
catalysts showed higher conversion (up to 90% in case of 6B1 sam-
ple) than that of the powdered catalyst (ca. 66%), while at lower
temperatures only the structures with 1-3 configuration (see Fig. 2)
showed improved CO2 conversion.
It is important to note, that in case of the samples with the
same stacking (1-1 or 1-3), difference in macro-porosity does not
result in significant changes of CO2 conversion. In case of the sam-
ples with different stackings, the 1-3 stacking structured catalyst
showed a considerably higher conversion than the ones with 1-
1 stacking. Despite the same volume of the samples, the contact
between the catalysts and reactant gas in 1-3 ‘zig-zag’ 3DFD struc-
tures is better than in the structures with 1-1 ‘linear’ stacking.
Obviously, this resulted in higher contact time between the reac-
tant gas and the structured surface channels, and higher residence
time distribution inside the channels. Therefore, this difference
can be explained by an external mass transfer effect. The similar
Please cite this article in press as: S. Danaci, et al., Efficient CO2 methanation over Ni/Al2O3 coated structured catalysts, Catal. Today
(2016), http://dx.doi.org/10.1016/j.cattod.2016.04.019

G Model
CATTOD-10167; No. of Pages10
ARTICLE IN PRESS
10
S. Danaci et al. / Catalysis Today xxx (2016) xxx–xxx
90
[3] W. Wang, J. Gong, Front. Chem. Eng. China 5 (2011) 2–10.
[4] A.H. Zamani, R. Ali, W.A.W.A. Bakar, J. Taiwan Inst. Chem. Eng. 45 (2014)
143–152.
[5] A. Karelovic, P. Ruiz, Appl. Catal. B Environ. 113–114 (2012) 237–249.
[6] M. Kus´mierz, Catal. Today 137 (2008) 429–432.
[7] Z. Kowalczyk, K. Stołecki, W. Raróg-Pilecka, E. Mis´kiewicz, E. Wilczkowska, Z.
Karpin´ ski, Appl. Catal. A Gen. 342 (2008) 35–39.
[8] S. Scire, C. Crisafulli, R. Maggiore, S. Minico, S. Galvagno, Catal. Let. 51 (1998)
41–45.
80
70
60
[9] S. Hwang, J. Lee, U.G. Hong, J.H. Baik, D.J. Koh, H. Lim, I.K. Song, J. Ind. Eng.
Chem. 19 (2013) 698–703.
[10] H. Zhu, R. Razzaq, C. Li, Y. Muhmmad, S. Zhang, Am. Inst. Chem. Eng. 59 (2013)
2567–2576.
50
40
3DFD structured catalyst
powder catalyst
[11] J. Zhang, Z. Xin, X. Meng, Y. Lv, M. Tao, Fuel 116 (2014) 25–33.
[12] J. Sehested, K.E. Larsen, A.L. Kustov, A.M. Frey, T. Johannessen, T. Bligaard, M.P.
Andersson, J.K. Nørskov, C.H. Christensen, Top. Catal. 45 (2007) 9–13.
[13] J. Ren, X. Qin, J.-Z. Yang, Z.-F. Qin, H.-L. Guo, J.-Y. Lin, Z. Li, Fuel Process.
Technol. 137 (2015) 204–211.
0
10
20
30
40
50
60
Reacꢀon ꢀme (h)
[14] G.F. Fu, J. Wang, B. Xu, H. Gao, X.L. Xu, H. Cheng, Trans. Nonferrous Met. Soc.
China 20 (2010) 221–225.
[15] J. Kopyscinski, T.J. Schildhauer, S.M.A. Biollaz, Chem. Eng. Sci. 66 (2011)
924–934.
Fig. 13. Stability test over packed-bed and 4B1 structured catalyst.
Reaction conditions: 350◦C, H2/CO2 = 4, WHSV 1500 h−1
.
[16] C.H. Bartholomew, Appl. Catal. A Gen. 212 (2001) 17–60.
[17] V. Tomasˇic´, F. Jovic´, Appl. Catal. A Gen. 311 (2006) 112–121.
[18] A. Cybulski, J.A. Moulijn, Structured catalysts and reactors, in: The Present and
the Future of Structured Catalysts − An Overview, second ed., CRC Press, Boca
Raton, 2006, pp. 1–16.
alyst, carbon dioxide conversion stayed constant at ca. 80% during
53 h time-on-stream. Decrease of CO2 conversion of the powder
catalyst was observed probably due to the formation of carbon
deposits and catalyst deactivation. It is suggested, that improved
heat transfer properties of the 3DFD structured catalyst prevent
temperature increase and lead to the enhanced catalyst stability.
A similar effect was described in Ref. [29] for a Microlith® based
structured noble metal catalyst for Sabatier reaction. The structured
catalyst was found to be stable for more than 100 h TOS. It was
proposed that the lack of catalyst degradation is due to the high
heat transfer rate of the Microlith® catalytic substrates that permits
uniform temperature distribution and avoids local hot spots that
can cause metal sintering and catalyst deactivation.
[19] K.P. Brooks, J. Hu, H. Zhu, R.J. Kee, Chem. Eng. Sci. 62 (2007) 1161–1170.
[20] V. Meille, Appl. Catal. A Gen. 315 (2006) 1–17.
[21] Z. Liu, B. Chu, X. Zhai, Y. Jin, Y. Cheng, Fuel 95 (2012) 599–605.
[22] R. Zapf, K. Becker-Willinger, H. Berresheim, H. Bolz, V. Gnaser, G. Hessel, P.
Lob, A.-K. Pannwitt, A. Ziogas, Trans IChemE 81 (2003) 721–729.
[23] O. Görke, P. Pfeifer, K. Schubert, Catal. Today 110 (2005) 132–139.
[24] J. Hu, K.P. Brooks, J.D. Holladay, D.T. Howe, T.M. Simon, Catal. Today 125
(2007) 103–110.
[25] C. Janke, M.S. Duyar, M. Hoskins, R. Farrauto, Appl. Catal. B Environ. 152–153
(2014) 184–191.
[26] C.Y. Zhao, J. Int, Heat Mass Transf. 55 (2012) 3618–3632.
[27] M. Frey, D. Édouard, A.-C. Roger, Comptes Rendus Chim. 18 (2015) 283–292.
[28] F. Kapteijn, R.M. Deugd, J.A. Moulijn, Catal. Today 105 (2005) 350–356.
[29] C. Junaedi, K. Hawley, D. Walsh, S. Roychoudhury, M.B. Abney, J.L. Perry,
Compact and lightweight sabatier reactor for carbon dioxide reduction, in:
41st International Conference on Environmental Systems., Portland, Oregon,
AIAA, 17–21 July, 2011.
4. Conclusions
Ni/Al2O3 powder and 3DFD structured catalysts were prepared
and characterized. The effect of dispersant and inorganic binder on
the catalytic suspension properties was studied. It was found that
dispersant concentration significantly affects the coating quality
(homogeneity and thickness), whereas the inorganic binder has an
influence on the coating adhesion. The optimal composition of the
coating suspension was determined as 3% PVA, 2% Ludox, 1% acetic
acid, 20% catalyst powder, and water.
Structured 3DFD supported catalysts showed improved CO2
conversion (especially at higher temperatures) and stability of
the catalyst in CO2 methanation reaction. The best results were
obtained using the structured catalyst of ‘zig-zag’ configuration,
which is possibly explained by the combination of improved mass
and heat transfer properties. This observation will be confirmed
in future studies by corresponding additional measurements and
modelling. Also the prolonged (>200 h TOS) stability test and the
characterization of the catalyst afterwards will be performed.
[30] S.H.R.W.I. Mullens, J.R. Thijs, W.L. Luyten, W.L. Bouwen, US Patent
20110064784 A1 (2009).
[31] J. Luyten, S. Mullens, I. Thijs, Kona Powder Part. J. 28 (2010) 131–142.
[32] J. Lefevere, M. Gysen, S. Mullens, V. Meynen, J. Van Noyen, Catal. Today 216
(2013) 18–23.
[33] M. Klumpp, A. Inayat, J. Schwerdtfeger, C. Körner, R.F. Singer, H. Freund, W.
Schwieger, Chem. Eng. J. 242 (2014) 364–378.
[34] S. Abedini, N. Parvin, P. Ashtari, Mater. Sci. Eng. A 533 (2012) 1–8.
˙
[35] A.E. Aksoylu, Z.I. Önsan, Appl. Catal. A Gen. 164 (1997) 1–11.
[36] S. Abelló, C. Berrueco, D. Montané, Fuel 113 (2013) 598–609.
[37] C. Li, Y.-W. Chen, Thermochim. Acta 256 (1995) 457–465.
[38] H. Song, J. Yang, J. Zhao, L. Chou, Chin J. Catal. 31 (2010) 21–23.
[39] K.V.R. Chary, P.V. Ramana Rao, V. Venkat Rao, Catal. Commun. 9 (2008)
886–893.
[40] A. Boumaza, L. Favaro, J. Lédion, G. Sattonnay, J.B. Brubach, P. Berthet, A.M.
Huntz, P. Roy, R. Tétot, J. Solid State Chem. 182 (2009) 1171–1176.
[41] J.R. Gonzalez-Velasco, M.A. Gutierrez-Ortiz, J.L. Marc, J.A. Botas, M.P.
Gonzalez-Marcos, G. Blanchard, Ind. Eng. Chem. 42 (2003) 311–317.
[42] L.C. Almeida, F.J. Echave, O. Sanz, M.A, Surf. Sci. Catal. 175 (2010) 25–33.
[43] V. Meille, S. Pallier, G.V. Santa Cruz Bustamante, M. Roumanie, J.-P. Reymond,
Appl. Catal. A Gen. 286 (2005) 232–238.
[44] V.N. Antsiferov, V.D. Khramtsov, J. Eng. Phys. Thermophys. 64 (1993) 475–478.
[45] C. Cristiani, M. Valentini, M. Merazzi, S. Neglia, P. Forzatti, Catal. Today 105
(2005) 492–498.
[46] S. Yasaki, Y. Yoshino, K. Ihara, K. Ohkubo, US Patent 5 208206 (1993).
[47] L. Kiwi-Minsker, A. Renken, Handbook of Heterogeneous Catalysis, in:
Microstructure Reaction, 2005, pp. 2248–2264.
References
[1] W. Wang, S. Wang, X. Ma, J. Gong, Chem. Soc. Rev. 40 (2011) 3703–3727.
[2] G. Benjaminsson, J. Benjaminsson, R.B. Rudbeg, SGC report: Power-to-Gas −A
technical review 2013, SGC Rapport, 2013, pp. 284.
[48] A. Bornmyr, B. Holmberg, Surface treatment of stainless steels, in: Pick
Handbook, 2009.
[49] L. Kiewidt, J. Thöming, Chem. Eng. Sci. 132 (2015) 59–71.
[50] V. Barbarossa, G. Vanga, Methanation of carbon dioxide, in: XXXIV Meeting of
the Italian Section of the Combustion Institute, Rome 24–26 October, 2011.
Please cite this article in press as: S. Danaci, et al., Efficient CO2 methanation over Ni/Al2O3 coated structured catalysts, Catal. Today
(2016), http://dx.doi.org/10.1016/j.cattod.2016.04.019