YSZ monoliths promoted with Co as catalysts for the production of H2 by steam reforming of ethanol

Article abstract of DOI:10.1016/j.apcata.2017.03.008

Steam reforming of ethanol was investigated between 400?°C and 600?°C over 3?mol% Y₂O₃ doped ZrO₂ monolith infiltrated with Co, known as an effective catalyst for this reaction. Monoliths were fabricated by implementing the freeze-casting technique in order to obtain hierarchical, vertical and highly ordered porous microstructure resulting in very low pressure drop, high surface area and accessibility to active sites. Catalytic monoliths were firstly loaded with 3?wt% of Co. Catalytic activity of Co-containing monolith was compared with the catalytic activity of a powdered material with the same chemical composition than the catalytic monolith. The results show a clear improvement all over the temperature range of the ethanol conversion and H₂ yield and a decrease of the by-products contents when using a freeze-cast monolith support. The incorporation of La enabled a strong decrease in coke deposition over the activated monolith. The effect of the addition of 0.25?wt.% of Pt was also studied. Pt makes possible to inhibit carbon deposition but has a detrimental effect on ethanol conversion at temperatures above 500?°C and on H₂ selectivity at low temperature. A higher Pt loading of 0.8?wt% did not bring any beneficial effect and was even detrimental at low temperature. Finally, the beneficial effect of the implementation of freeze-cast monolith in comparison with the not structured fixed-bed configuration is explained by the difference of packing since and improved fluid dynamics at similar GHSV, the former one promotes in a higher proportion the Water Gas Shift Reaction (WGSR) further in the monolith length.

Full text of DOI:10.1016/j.apcata.2017.03.008

Accepted Manuscript
Title: YSZ monoliths promoted with Co as catalysts for the
production of H by Steam Reforming of Ethanol
2
Authors: Cyril Gaudillere, Juan Jose Gonz a´ lez, Antonio
Chica, Jos e´ M. Serra
PII:
S0926-860X(17)30107-2
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.apcata.2017.03.008
APCATA 16167
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
17-11-2016
6-2-2017
9-3-2017
Please cite this article as: Cyril Gaudillere, Juan Jose Gonz a´ lez, Antonio
Chica, Jos e´ M.Serra, YSZ monoliths promoted with Co as catalysts for the
production of H2 by Steam Reforming of Ethanol, Applied Catalysis A, General
http://dx.doi.org/10.1016/j.apcata.2017.03.008
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

YSZ monoliths promoted with Co as catalysts for the
production of H by Steam Reforming of Ethanol
2
Cyril Gaudillere, Juan Jose González, Antonio Chica* and José M. Serra*
Instituto de Tecnología Química (Universidad Politécnica de Valencia – Consejo Superior de
Investigaciones Científicas), Av. Naranjos s/n, E-46022 Valencia (SPAIN)
*
To whom correspondence should be addressed, Fax 0034.963.877.809; e-mail:
achica@itq.upv.es, jmserra@itq.upv.es
1

GRAPHICAL ABSTRACT
HIGHLIGHTS

Freeze-cast monoliths loaded with Co-based catalysts are effective for steam reforming
of ethanol


Promotion with La and Pt enables to improve H selectivity and reduce coke formation
2
Freeze-cast monolith showed higher activity and selectivity than conventional powder-
based catalysts
Abstract
Steam reforming of ethanol was investigated between 400ºC and 600ºC over 3 mol% Y
doped ZrO monolith infiltrated with Co, known as an effective catalyst for this reaction.
2
O
3
2
Monoliths were fabricated by implementing the freeze-casting technique in order to obtain
hierarchical, vertical and highly ordered porous microstructure resulting in very low pressure
drop, high surface area and accessibility to active sites. Catalytic monoliths were firstly loaded
with 3 wt% of Co. Catalytic activity of Co-containing monolith was compared with the catalytic
2

activity of a powdered material with the same chemical composition than the catalytic monolith.
The results show a clear improvement all over the temperature range of the ethanol conversion
and H yield and a decrease of the by-products contents when using a freeze-cast monolith
2
support. The incorporation of La enabled a strong decrease in coke deposition over the activated
monolith. The effect of the addition of 0.25 wt.% of Pt was also studied. Pt makes possible to
inhibit carbon deposition but has a detrimental effect on ethanol conversion at temperatures
above 500ºC and on H selectivity at low temperature. A higher Pt loading of 0.8 wt% did not
2
bring any beneficial effect and was even detrimental at low temperature. Finally, the beneficial
effect of the implementation of freeze-cast monolith in comparison with the not structured
fixed-bed configuration is explained by the difference of packing since and improved fluid
dynamics at similar GHSV, the former one promotes in a higher proportion the Water Gas Shift
Reaction (WGSR) further in the monolith length.
Keywords: Ethanol steam reforming; H production; Cobalt catalysts; Freeze-cast monoliths; zirconia
2
1) Introduction
The inevitable decrease in the availability of fossil fuel resources and the associated pollutants
emission (CO , NO , SO , particles etc) coupled with the growing worldwide energy demand,
makes necessary the development of new technologies for the clean and sustainable production
of energy. Hydrogen (H ) is today the cleanest and the most eco-friendly energy carrier system
known since its combustion only produces H O and does not emit any pollutant or greenhouse
gases (GHG),. Nevertheless, H is not free on the earth and thus, it has to be produced from a
primary energy source like fossil fuel or renewables. Currently, the steam reforming of natural
gas is the conventional and industrial process used for H production. However, this process is
associated with the emission of large quantities of greenhouse gases, especially carbon dioxide
2
x
x
2
2
2
2
2
(CO ). An alternative option for H
2
production is the water electrolysis. Unfortunately, its
3

industrial up-scaling is complicated since the electrical demand is very important and noble
metals have to be implemented at the state-of-the-art electrodes to achieve reasonable yields at
low temperature [1, 2]. Consequently, a new eco-friendly H
2
production route is needed for its
cleaner and more sustainable production.
Recently, the valorisation of biomass-derived compounds like ethanol through its catalytic
steam reforming to produce H has attracted the attention of the scientific community [3-5]. The
2
production of ethanol from the fermentation of lignocellulosic wastes gives access to a mixture
of water and ethanol (10-18 wt.% in ethanol) which can help in the substitution of fossil fuels
through the sustainable production of H
easy to store, handle and transport and with low toxicity and volatility could be used as primary
source for H production. As biomass requires atmospheric CO to grow, this opportunity is
very interesting from the environmental point of view since its implementation could lead to a
CO neutral global process in opposition with H production from fossil fuels conversion.
Catalytic steam reforming of ethanol for H production (reaction 1) has been widely studied in
2
via steam reforming reaction. Indeed, ethanol which is
2
2
2
2
2
the last years and emphasis has been done especially on the development of efficient ethanol
steam reforming catalysts [6-8].
-
1
C
2
H
5
OH + 3H
2
O  2CO
2
+ 6H
2
ΔH298= 174 kJ.mol (reaction 1)
Catalysts used in the steam reforming of ethanol are usually made of a metal oxide support like
Al , SiO , ZnO, MgO, CeO or ZrO and by nanosized metallic phase [9-12]. It has been
2
O
3
2
2
2
shown that catalyst preparation (impregnation method, metal precursors, calcination
temperature and time, reduction temperature and time) has a great influence over the metal-
support interactions and by consequent over the catalytic performance (activity, selectivity,
stability) [13, 14]. The support is chosen in order to optimize the metal dispersion and thus
improve its stability against sintering that would induce a detrimental decrease of the catalytic
performance. Generally, a non-acidic support with high specific surface area avoiding ethanol
dehydration and coke formation is preferred [15]. Regarding the metals and even if noble ones
4

such as Ru, Rh, Pt and Pd have been tested and proved as efficient for this reaction [16, 17],
their cost is often limiting from an industrial point of view. Among the other cheaper metals, Co
[12, 14, 18-27] and Ni [11, 28, 29] are known to be the most effectives. Additional promoters
can also be incorporated over the catalyst to improve stability and/or the selectivity towards a
specific by-product of the reaction. This is the case of La which is known to limit coke
formation and improves catalyst stability [11, 28, 29].
Catalytic reactions are usually performed under fixed-bed reactor conditions, i.e. the powder
catalyst is placed inside a quartz or inert metallic reactor tube and the reactive gaseous feed is
supplied from the top to the bottom passing through and reacting with the catalytic bed. This
configuration is easy to implement for any type of powder but presents some drawbacks like the
difficulty to operate at high space velocity with the increase of the pressure drop through the
catalytic bed and the possibility of hot spot with exothermic reaction. For process intensification
purposes, an alternative to fixed-bed reactor can be suggested and consists of a catalyst-loaded
ceramic monolith [30]. Such reactors present the advantage to have much higher catalyst
content per unit volume, to enhance the reaction efficiency by operating at higher space
velocities and lower pressure drop and to minimize the diffusional resistance issues in
comparison with fixed-bed reactor [31,32]. Usually made of an inexpensive ceramic material
like mullite, cordierite, Al
2
O , MgO or a metallic alloy, the reactor can be fabricated by
3
extrusion and with a honeycomb-like structure and followed by sol-gel technique [33] or by
pore forming agent [34]. Catalysts are coated into this monolith to give the final sample. The
crucial point lies in reaching a homogeneous coating of the catalyst in all over its volume to
avoid useless area without catalyst. Generally, it is realized by wash-coating [35-37] or by
liquid infiltration [38] of a metal(s) precursor solution, the latter being easier to realize but with
less reproducibility. Such metallic alloy monoliths are already implemented in the automotive
sector for exhaust gas conversion or for methanol steam reforming and oxidative steam
reforming [39,40]. Another alternative to the fabrication of such structured reactor is studied for
few years now and deals with the implementation of the freeze-casting technique [41]. It
consists of freezing a liquid suspension, followed by the sublimation of the solvent under both
5

low temperature and pressure and the final sintering of the particles for consolidation. The
obtained structure is robust and porous with oriented channels corresponding to the replica of
the solvent crystals along the propagation direction of the solid-liquid interface. Modifying the
powder characteristics, the freezing conditions or the slurry formulation, both shape and size of
the porosity can be tailored for the targeted requirements. The resulting monolith exhibits a
hierarchical, vertical and highly ordered porous microstructure associated with low pressure
drop as detailed in [42], a high surface area and accessibility to active sites. Thus, freeze-cast
monoliths could be an interesting substitute to structured reactor fabricated by the conventional
ceramic techniques detailed above.
Here we suggest assessing the catalytic steam reforming of ethanol using Y
2
O
3
doped ZrO
2
catalyst loaded freeze-cast monolith since ZrO is an active support for this reaction [10, 14]. In
2
addition, YSZ is a precursor for the preparation of SOFC anodes and the impregnation with Co
could be a good option to prepare YSZ-based anodes actives in the steam reforming of ethanol,
allowing the feeding of ethanol directly to the SOFC instead of molecular hydrogen. YSZ is
also an optimum material to prepare thermochemical and mechanically-robust freeze-cast
monoliths. Regarding the ethanol steam reforming active site, we have decided to impregnate
the monoliths firstly with Co, a non-noble metal with proved catalytic activity in the reforming
of ethanol [9,11,15-18]. The influence of the Co loading (3-1 wt.%) and of the addition of a
second metal promoter (Pt and La) is also presented. In order to rule about the influence of the
monolith on the catalytic activity, a comparison with similar fixed-bed using powder catalysts is
also given.
2
2
) Experimental section
.1) Fabrication of the catalyst loaded freeze-cast monoliths
Monolith manufacture: the 3 mol% Y
2
O
3
doped ZrO (YSZ-M) powder was provided by Tosoh
2
and ball-milled in acetone during 48 hours to ensure homogeneity and a final average particle
size of about 2 µm. Porous monoliths were elaborated by ice-templating using the freeze-casting
6

technique. A slurry containing the ceramic powder (40-50 wt.%), water (30-40 wt.%) as
solvent, a polyacrylate-based dispersing agent (1-4 wt.% ) and polyethylene glycol (1-4 wt.%)
(Sigma-Aldrich) is stirred for 24 hours to get a good particle distribution. The slurry is then
poured into Teflon mold to get a sample of 1 cm height and cooled using copper rod cooled by
liquid nitrogen. After complete freezing, the sample is removed from the mold and ice crystals
are sublimated by freeze drying at -53ºC and reduced pressure during 24 hours using a Scanvac
commercial freeze dryer. The sample is then sintered at 1300ºC under air during 6 hours. Both
heating and decreasing ramps are 20 hours. The final diameter of the sample is 9 mm and the
final thickness of 3 mm is adjusted by grinding the sample.
Monolith infiltration: Co, La, and Pt catalysts have been coated by liquid infiltration of
saturated aqueous solutions of cobalt nitrate hexahydrate (Sigma-Aldrich), lanthanum nitrate
hexahydrate (Sigma-Aldrich) and chloroplatinic acid (Sigma-Aldrich) respectively. Successive
infiltration and drying steps have been realized until the desired metal loading was achieved.
After infiltration, samples are calcined at 600ºC during 3 hours to remove precursors.
The following samples has been prepared and studied in this work:


CoYSZ-M: monolith with cobalt incorporated by incipient wetness impregnation.
CoYSZ-P: Cobalt was incorporated in the monolith by incipient wetness impregnation and later
was crushed. Monolith like CoYSZ-M was grinded to get this sample (CoYSZ-P).
CoYSZ-P-M: Cobalt was incorporated by incipient wetness impregnation in a powder obtained
by crushing the monolith (YSZ-M).


CoYSZ-P-T: Cobalt was incorporated by incipient wetness impregnation onto YSZ powder
supplied by Tosoh.
2.2) Characterization techniques
7

Chemical analysis: chemical composition of monoliths after impregnations were determined by
inductively coupled plasma atomic emission spectroscopy (ICP-AES), using a Varian Spectra
A-10 Plus apparatus.
Area measurement: textural properties were obtained from nitrogen adsorption isotherms
determined at 77K Micrometrics ASAP 2000 equipment. Prior to the absorption the samples
were outgased at 200ºC during 24h.
XRD: the crystalline phases in the catalysts (powder and monoliths) were determined by X-ray
diffraction. XRD patterns were obtained at room temperature in a Philips X'pert diffractometer
using monochromatized CuKα radiation. Metallic crystallites average diameters were estimated
from XRD patterns applying the Scherrer equation.
TEM microscopy: transmission electron microscopy (TEM) was used to determine the
morphology of carbon deposited over the catalyst during the reaction. Each sample was
ultrasonically dispersed in iso-propanol and transferred to carbon coated copper grids.
Micrographs were recorded using a Philips CM-10 microscope operating at 100 kV.
SEM Microscopy: Swept Electron Microscopy is used to get information about morphology
and crystal sizes. This technique let us to identify impurities or the presence of other phases in
the sample. The SEM images were taken with a microscope JEOL JSM 6300 operating at 20
kV, using monoliths recovered with C. The information about compositions were obtained by
X-ray detector (EDS), using recovered samples with C.
FESEM Microscopy: field Emission Swept Electron Microscopy offers the same information
than swept electron microscopy but with higher image resolution and with a wide range of
energy. It is possible to work at low potentials (0.02-5kV), avoiding problems associated with
8

charges in sensible samples. The FESEM images were taken un a Ultra 55 (Zeiss) equipment
operating at 2.0 kV using samples without recovering.
TPR: the reduction behavior of the oxidized cobalt phases were studied by temperature-
programmed reduction (TPR) in a Micrometrics Autochem 2910 equipment. About 50 mg of
-
1
the calcined catalyst was initially flushed with 30 mL min of Ar at room temperature during 30
min. Then a mixture of 10 vol.% of H in Ar was passed through the catalyst at a total flow rate
of 50 cm min while the temperature was increasing up till 900ºC at a heating rate of 10 ºC
2
3
-1
-
1
min . The reduction degree was determined by TPR experiments comparing the total amount
of H consumed by each catalyst with the theoretical amount of H that would be necessary to
complete the Co reduction in the catalyst.
2
2
2
O
3
Elemental analysis: the amount of carbon deposited on the catalysts after the steam reforming
process was determined by elemental analysis using a Carlo Erba 1106 analyzer.
Catalytic evaluation: Ethanol Steam Reforming (ESR) was performed in a continuous fixed bed
quartz reactor at atmospheric pressure, H
2
O/EtOH =13 mol/mol (20 wt.% of ethanol), GHSV of
-
1
4
500h and a range of temperatures between 400°C and 600°C. Before setting up the reaction,
-1
the catalysts were reduced ''in situ'' in H flow (100 ml min ) at 600ºC during 2 h. In a typical
2
catalytic test the reactor is loaded with an impregnated monolith or powder catalysts (grain-size:
0.25-0.42 mm) in both cases 0.5g. In case of powder the samples were diluted with the required
3
amount of carborundum (SiC) (grain-size: 0.60-0.80 mm) to achieve a catalyst bed of 2 cm .
The water and ethanol were fed together from a tank pressurized at 20 bar with a mass flow
®
meter/controller Bronkhorst Liqui-Flow . The gas products were analyzed using an online gas
chromatograph (Bruker 450-GC) equipped with four columns (Hayesep Q 80/100 Mesh L=
0.25m, 1/16´´, D
I
=1mm Ultimetal CP1380; Hayesep N80/100 Mesh L=0.5m, 1/16´´, D =1mm
I
Ultimetal CP1307; Molsieve 13x80/100 Mesh L=1.2m, 1/16´´,D
I
=1mm Ultimetal CP1309; VF-
9

5
ms L=25m, 0.25mm, df=4µm CP8942) and two detectors: thermal conductivity (TCD) and
flame ionization (FID).
Ethanol conversion and products selectivity were determined according to the equations (1) and
-1
(
2), where (FEtOH
)
0
is the flow of ethanol fed to the reactor (mol s ), and (FEtOH
)
f
and F depicts
j
the flow of ethanol and the flow of the component j coming out from the reactor, respectively.
Selectivity values were calculated as the molar percentage of the products obtained, excluding
water.
(
퐹퐸푡푂퐻)0 −(퐹퐸푡푂퐻)푓
푋(%, 푚표푙) =
· 1ꢀꢀ
(1)
(2)
(
퐹퐸푡푂퐻)0
퐹푖
푆(%, 푚표푙) = (∑ 퐹 )
· 1ꢀꢀ
푗 푝푟ꢁ푑푢푐푡푠
3
3
) Results and discussion.
.1) Characterization of the freeze-cast monolith
Figure 1 shows top-view SEM micrographs of an YSZ-M monolith sample after sintering at
300ºC during 6 hours. The porous scaffold architecture with large aligned pores (a & b) and
1
with ceramic walls exhibiting an average thickness of about 5 µm (c) are visible. Figure 1d
gives a focus on one of these ceramic walls where the YSZ-M particles are well sintered while a
round-shaped residual porosity of about 200 nm is observed. Thus, the YSZ-M monoliths
present the hierarchical macro porosity induced by the shaping process and the porosity coming
from the sintering step. It has been decided to sinter the samples at 1300ºC and not at higher
temperature since this residual porosity would facilitate the metal infiltration into the monolith.
Moreover, this sintering temperature is sufficient to ensure the monolith a proper mechanical
stability under pressure and to prevent from any deterioration or break. These observations
clearly indicate that such freeze-cast monoliths present an interesting feature sought for catalytic
devices, i.e. a large and organised porosity boosting gas transport and minimizing pressure loss
10

through it. A XRD pattern of the porous YSZ-M freeze-cast monolith is presented in Figure 2
where the cubic fluorite YSZ phase is observed.
3.2) Characterization of the catalytic material containing Co
Cobalt was incorporated in the monolith by incipient wetness impregnation (3CoYSZ-M). For
comparison purpose it has been included two more catalytic materials. The first one (3CoYSZ-
P-M) was prepared by incorporating Co by incipient wetness impregnation onto a crushed
monolith, in powder form. The second one was prepared in a similar way by using commercial
YSZ powder supplied by Tosoh (3CoYSZ-P-T) further impregnated with 3 wt.% of Co.
XRD patterns of the calcined and reduced catalysts after Co incorporation are shown in Figure
3
. The presence of cobalt oxides [Co
JCPDS 17-923) are identified in the XRD patterns of the calcined catalysts. Co
calcination temperatures under 900ºC [43]. Samples containing Co were calcined at 600ºC, thus
Co was formed preferentially, as it can be seen in Figure 3.
3
O
4
(JCPDS 89-4037)] and the cubic phase of YSZ-M
(
3
O
4
is formed at
3
O
4
For reduced catalysts, the peaks related to the cobalt oxides (Co
3
O ) nearly disappear and two
4
peaks at 28.2 and 44.3 degrees, corresponding to metallic Co, appears as consequence of the
reduction process. From reduced XRD, using the Scherrer equation [44], it has been determined
the particle size of the metallic Co (see Table 1). As it can be seen the lowest size of the Co
metallic particles corresponds to the samples prepared with the monolith as support (structured
or grinded). A lower size of the metallic particles involves a higher number of metallic sites
where the steam reforming of ethanol can take place.
Table 1 details the BET surface area, the particle size of the metallic particles of Co, determined
by XRD, and surface Co atomic composition on reduced samples determined by EDX-SEM
present in the 3CoYSZ-M, 3CoYSZ-P and 3CoYSZ-P-T catalysts. Surface area was similar and
very low for the three catalysts while Co particles size is of 3 nm for the monolith-based
catalysts (grinded or not) and double when the support is in powder form. It seems that using a
freeze-cast monolith as support favors the metal dispersion.
11

TPR profiles are shown in Figure 4. YSZ support (monolith and commercial) can be hardly
reduced under the practiced conditions as suggested by the negligible H consumption.
Specifically it was equivalent to 0.30% of the total ZrO , a low level of reduction also reported
by other authors [29]. TPR of samples containing Co particles present two main H
consumption peaks corresponding to the two-step transformation of cobalt oxide into metallic
cobalt: from Co to CoO and from CoO to metallic Co [45,46]. Reducibility of oxidized
2
2
2
3
O
4
cobalt at 600ºC has been determined for the three materials, finding similar temperatures for all
them and a reduction close to 100%.
3.3) Catalytic evaluation
Figure 5 shows the catalytic activity of a monolith containing 3 wt.% of Co (3CoYSZ-M). The
monolith sample exhibits the highest catalytic activity irrespective of the reactor temperature,
with a complete conversion of ethanol at 500ºC. As expected, the difference of activity between
catalysts decreases with increasing temperatures. Regarding H
2
selectivity, it can be seen that it
by the monolith is
was quite similar for all the samples. Nevertheless, the production of H
2
slightly higher particularly at high temperature (600ºC) with a value of 71% in comparison with
the 67% obtained for 3CoYSZ-P and 3CoYSZ-P-T catalysts. The gap with the thermodynamic
equilibrium is very small at 400ºC and tends to increase with temperature. Selectivity to others
undesired reaction products such as CO, CH
Figure 6).
4
, C
H
2 4
O and C
2
H has been also determined
2
(
The production of undesired products for the catalytic monolith is always the lowest while the
by-products associated to the 3CoYSZ-P-T catalyst are always the most important whatever the
temperature. These catalytic results indicate that use of YSZ-M monolith as cobalt support is
beneficial in the preparation of ethanol steam reforming catalysts with high catalytic activity,
improved H selectivity and low production of by-products.
2
12

Considering the similar characterization data presented by the samples prepared using the
monolith as support (3CoYSZ-M and 30CoYSZ-P), the differences found in the catalytic
performance could be explained taking into account the different configuration of the catalytic
bed, i.e., the fluid dynamics. In the case of the monolith we have a cylinder (0.4x1.2 cm) with
micro-channels (8-10 m) where Co nanoparticles have been incorporated. Regarding the
crushed monolith, the catalytic bed is formed by particles of monolith (0.2-0.4 mm) with Co
nanoparticles. These particles of monolith were diluted in SiC. The final bed volume was 3 ml
(3x1.2 cm). Thus, the flow of the reactants and products through the two beds will be different.
These differences in the catalytic bed could explain the lower CO production by the structured
monolith. CO formed in the reforming reaction would flow along the porous channels in the
structured monolith having the opportunity to react again giving more H through the water gas
2
shift reaction (Reaction 2). This could explain the lower CO production and a higher H
selectivity exhibited by the freeze-cast monolith (3CoYSZ-M).
2
-
1
CO + H
2
O  H
2
2
+ H O
ΔH298= - 42 kJ.mol (Reaction 2)
In addition to the catalytic activity and selectivity, stability against deactivation is also an
important catalytic parameter to be considered. For the steam reforming reaction, the formation
of carbon is one of the most important issues affecting the catalyst activity. Table 2 shows the
carbon formation rate during reaction for the three catalysts studied in this work.
The largest amount of coke corresponds to the structured monolith. This carbon grows mainly
inside of the channels of the monolith and originates from Co nanoparticles as observed by
SEM and TEM analysis respectively (Figure 7 and Figure 8). As it can be seen in Figure 8
carbon nanotubes are the main type of formed carbon. Thus, the stability of this catalytic reactor
is expected to be lower than the catalyst bed prepared using the crushed monolith or commercial
YSZ. In order to decrease the coke formation and improve the stability of the freeze-cast
monolith, a new one impregnated with La, a known coke inhibitor, was prepared [47,48]. La
impregnation was carried out simultaneously with Co. Thus, a new monolith containing 3 wt.%
13

of Co and 1 wt.% of La was prepared (3Co1LaYSZ-M) and the related catalytic results are
presented in Table 3.
Similar conversion values are exhibited by the samples with and without La, indicating that the
presence of La does not affect significantly the catalytic activity. H selectivity is slightly higher
2
for the monolith containing La, particularly at low and medium temperatures (400-500ºC).
Methane formation is also lower at these temperatures. The presence of La seems to decrease
the methanation reaction (CO + 3H
2
 CH
4
2
+ H O), explaining the low production of methane
and the increased H selectivity, especially at low temperatures (400ºC). This is an interesting
2
data because the conversion of ethanol could be carried out at lower temperatures (400ºC)
where the formation of carbon is disfavored (but methane formation is thermodynamically
favored).
Carbon formation rate has been also determined for the La-containing monolith. As detailed
Table 4, carbon deposition was clearly minimized in presence of 1wt.% La impregnated into the
monolith with 0.0012 g of carbon per gram of catalyst per hour, representing a drop of 91.5%.
This value is even lower than for the catalysts prepared with the crashed monolith or
commercial YSZ (Table 2).
Finally, the influence of the Co loading into the monolith as well as the presence of Pt, as low
coke producer, has been studied in the steam reforming of ethanol. Thus, four different catalyst-
loaded freeze-cast monoliths were prepared and are named as 1CoYSZ-M, 1Co0.5LaYSZ-M,
1Co0.25PtYSZ-M and 1Co0.8PtYSZ-M. Table 5 details the ethanol conversion and products
distribution for these four catalysts and for the 3CoYSZ-M as comparison.
The rate of coke formation was also determined for these samples and it is presented in Table 6.
From Tables 5 and 6 it can be seen that the sample 1Co0.5LaYSZ-M presents excellent results
at 600ºC with a high catalytic activity (ethanol conversion close to 100%), a high H selectivity
2
14

(73.19%), a low selectivity to undesired by-products (CH
4
: 1,23%, CO: 4.04%, C
2
H and
2
C
2 4
H O: 0%) and as expected with the presence of La, the lowest coke deposition (0.0008
gC/gcat*h). The presence of 0.25 wt.% Pt seems also interesting from the coke deposition point
of view. Indeed, it allows to divide by a factor 2 the carbon formation (in comparison with the
1CoYSZ-M catalyst) but is detrimental over the ethanol conversion especially at T> 500ºC even
if the H selectivity remains stable. A higher content of Pt, i.e. 0.8 wt.%, does not bring any
2
substantial positive effect whatever the temperature.
These observations confirm that preparation of the catalysts as crucial for the optimization of
the catalytic activity/selectivity. Indeed, according to the reaction temperature, both metal nature
and loading have to be carefully chosen since a slight difference can lead to important
discrepancies in conversion and selectivity. The second interesting conclusion from all these
results is the beneficial effect of shaping the support in a monolith configuration. Indeed, for a
similar GHSV, the catalyst packing inside the reactor is completely different. It seems that, for
this reaction of steam reforming of ethanol where the WGSR occurs, the monolith design
enables to favor this latter reaction and shift its selectivity to H
when the flow reactant reacts firstly contacts with the top of the catalyst-loaded monolith, the
steam reforming of ethanol takes place while an incomplete reforming giving CO and H can
also occur according to C OH + H O  2CO + 4H . In the case of a monolith support, the
2
thanks to its length. Indeed,
2
H
2 5
2
2
CO produced can be converted by the WGSR (reaction 2) further in the next few millimeters of
the freeze-cast reactor while in case of a fixed-bed reactor configuration, it might not occur in
the same proportion in case of a short catalytic bed. For industrial purposes and process
optimization, it is convenient to avoid packed bed reactor since an important quantity of
catalyst, i.e. a thick catalytic bed, would be necessary to reach reasonable yields. Operate at high
flow rate under a fixed-bed reactor configuration is complicated since pressure drop through the
catalytic bed can quickly increase and hot spots appear especially in case of exothermic
reaction. Catalyst-loaded freeze-cast monolith thus appear as a valuable option as possible
substitute to fixed-bed reactor for industrial purposes since it could avoid these two problems.
Indeed, as detailed before [36], such monoliths present a very low pressure drop of about 0.6
15

-
1
-1
bar.mm at flow higher than 400 mL.min and its special porous organization would act as a
heat exchanger and would be effective to remove heat generated by an exothermic reaction.
Conclusions
Catalytic steam reforming of ethanol was assessed by using catalyst loaded YSZ-M freeze-cast
monolith. The influences of Co, La and Pt loading over the catalyst activity, selectivity, and
stability have been studied. Co is a very active metal for this reaction and La a very good
inhibitor of the coke formation. A low content of Pt (0.25 wt.%) was also found to be interesting
regarding coke formation but its presence negatively affects the ethanol conversion at high
temperature (>500ºC) and the H selectivity al low temperature while a higher Pt loading (0.8
2
wt.%) does not bring any positive effect and is even detrimental especially at low temperature.
The implementation of freeze-cast monolith reactor in comparison with catalytic packed bed is
proven to be very interesting since it allows the improvement of both ethanol conversion and H
2
selectivity in all the temperature range but also decreases conversion related to all the by-
products. The difference of packing between both configurations and at a similar GHSV, the
former one favours the WGSR further in the monolith length which apparently does not occur in
the same way in the powder catalytic bed.
Acknowledgements
Financial support by the Spanish Government (ENE2014-57651 and SEV-2012-0267 grants) is
kindly acknowledged.
16

References
1.
Sridhar, K.R. and B.T. Vaniman, Oxygen production on Mars using solid oxide
electrolysis. Solid State Ionics, 1997. 93(3–4): p. 321-328.
2.
Tao, G., K.R. Sridhar, and C.L. Chan, Study of carbon dioxide electrolysis at
electrode/electrolyte interface: Part II. Pt-YSZ cermet/YSZ interface. Solid State Ionics,
2
004. 175(1–4): p. 621-624.
3.
4.
5.
Haryanto, A., et al., Current Status of Hydrogen Production Techniques by Steam
Reforming of Ethanol:ꢀ A Review. Energy & Fuels, 2005. 19(5): p. 2098-2106.
Kalinci, Y., A. Hepbasli, and I. Dincer, Biomass-based hydrogen production: A review
and analysis. International Journal of Hydrogen Energy, 2009. 34(21): p. 8799-8817.
Ni, M., D.Y.C. Leung, and M.K.H. Leung, A review on reforming bio-ethanol for
hydrogen production. International Journal of Hydrogen Energy, 2007. 32(15): p. 3238-
3
247.
6.
7.
8.
9.
Bshish, A., et al., Steam-reforming of ethanol for hydrogen production. Chemical
Papers, 2011. 65(3): p. 251-266.
Contreras, J.L., et al., Catalysts for H2 production using the ethanol steam reforming (a
review). International Journal of Hydrogen Energy, 2014. 39(33): p. 18835-18853.
Guerrero, L., S. Castilla, and M. Cobo, Advances in ethanol reforming for the production
of hydrogen. Química Nova, 2014. 37: p. 850-856.
Batista, M.S., et al., Characterization of the activity and stability of supported cobalt
catalysts for the steam reforming of ethanol. Journal of Power Sources, 2003. 124(1):
p. 99-103.
10.
11.
Benito, M., et al., Zirconia supported catalysts for bioethanol steam reforming: Effect
of active phase and zirconia structure. Journal of Power Sources, 2007. 169(1): p. 167-
1
76.
Biswas, P. and D. Kunzru, Steam reforming of ethanol for production of hydrogen over
Ni/CeO–ZrO catalyst: Effect of support and metal loading. International Journal of
Hydrogen Energy, 2007. 32(8): p. 969-980.
1
2.
3.
Song, H., et al., Investigation of bio-ethanol steam reforming over cobalt-based
catalysts. Catalysis Today, 2007. 129(3–4): p. 346-354.
Llorca, J., et al., CO-free hydrogen from steam-reforming of bioethanol over ZnO-
supported cobalt catalysts: Effect of the metallic precursor. Applied Catalysis B:
Environmental, 2003. 43(4): p. 355-369.
1
14.
15.
16.
17.
18.
19.
Song, H., L. Zhang, and U.S. Ozkan, Effect of synthesis parameters on the catalytic
activity of Co-ZrO2 for bio-ethanol steam reforming. Green Chemistry, 2007. 9(6): p.
6
86-694.
Cavallaro, S., N. Mondello, and S. Freni, Hydrogen produced from ethanol for internal
reforming molten carbonate fuel cell. Journal of Power Sources, 2001. 102(1–2): p.
1
98-204.
Liguras, D.K., D.I. Kondarides, and X.E. Verykios, Production of hydrogen for fuel cells by
steam reforming of ethanol over supported noble metal catalysts. Applied Catalysis B:
Environmental, 2003. 43(4): p. 345-354.
Vaidya, P.D. and A.E. Rodrigues, Kinetics of Steam Reforming of Ethanol over a
Ru/Al2O3 Catalyst. Industrial & Engineering Chemistry Research, 2006. 45(19): p.
6
614-6618.
Da Costa-Serra, J.F. and A. Chica, Bioethanol steam reforming on Co/ITQ-18 catalyst:
Effect of the crystalline structure of the delaminated zeolite ITQ-18. International
Journal of Hydrogen Energy, 2011. 36(6): p. 3862-3869.
Da Costa-Serra, J.F., R. Guil-López, and A. Chica, Co/ZnO and Ni/ZnO catalysts for
hydrogen production by bioethanol steam reforming. Influence of ZnO support
17

morphology on the catalytic properties of Co and Ni active phases. International
Journal of Hydrogen Energy, 2010. 35(13): p. 6709-6716.
Da Costa-Serra, J.F., et al., Bioethanol steam reforming on Ni-based modified
mordenite. Effect of mesoporosity, acid sites and alkaline metals. International Journal
of Hydrogen Energy, 2012. 37(8): p. 7101-7108.
20.
2
1.
2.
Lin, S.S.-Y., D.H. Kim, and S.Y. Ha, Hydrogen Production from Ethanol Steam Reforming
Over Supported Cobalt Catalysts. Catalysis Letters, 2008. 122(3): p. 295-301.
Martono, E. and J.M. Vohs, Support effects in cobalt-based ethanol steam reforming
catalysts: Reaction of ethanol on Co/CeO2/YSZ(1 0 0) model catalysts. Journal of
Catalysis, 2012. 291: p. 79-86.
2
23.
24.
25.
26.
Óvári, L., et al., Near ambient pressure XPS investigation of the interaction of ethanol
with Co/CeO2(1 1 1). Journal of Catalysis, 2013. 307: p. 132-139.
Mattos, L.V., et al., Production of Hydrogen from Ethanol: Review of Reaction
Mechanism and Catalyst Deactivation. Chemical Reviews, 2012. 112(7): p. 4094-4123.
Ferencz, Z., et al., Effects of Support and Rh Additive on Co-Based Catalysts in the
Ethanol Steam Reforming Reaction. ACS Catalysis, 2014. 4(4): p. 1205-1218.
Varga, E., et al., Oxidation states of active catalytic centers in ethanol steam reforming
reaction on ceria based Rh promoted Co catalysts: An XPS study. Journal of Molecular
Catalysis A: Chemical, 2015. 397: p. 127-133.
27.
Soykal, I.I., H. Sohn, and U.S. Ozkan, Effect of Support Particle Size in Steam Reforming
of Ethanol over Co/CeO2 Catalysts. ACS Catalysis, 2012. 2(11): p. 2335-2348.
2
8.
9.
Fierro, V., et al., Ethanol oxidative steam reforming over Ni-based catalysts. Journal of
Power Sources, 2005. 145(2): p. 659-666.
Huang, T.-J., H.-J. Lin, and T.-C. Yu, A Comparison of Oxygen-vacancy Effect on Activity
Behaviors of Carbon Dioxide and Steam Reforming of Methane over Supported Nickel
Catalysts. Catalysis Letters, 2005. 105(3): p. 239-247.
2
3
0.
1.
Walsh, Z., B. Paull, and M. Macka, Inorganic monoliths in separation science: A review.
Analytica Chimica Acta, 2012. 750: p. 28-47.
Arab, S., et al., Methanol synthesis from CO2 and H2 in multi-tubular fixed-bed reactor
and multi-tubular reactor filled with monoliths. Chemical Engineering Research and
Design, 2014. 92(11): p. 2598-2608.
3
3
2.
3.
Huang, K., et al., Comparison of random and monolithic fixed-bed reactors for the
oxidative dehydrogenation of butene to butadiene. Journal of the Taiwan Institute of
Chemical Engineers.
3
Guo, X., et al., Preparation of mullite monoliths with well-defined macropores and
mesostructured skeletons via the sol–gel process accompanied by phase separation.
Journal of the European Ceramic Society, 2013. 33(10): p. 1967-1974.
She, J.H. and T. Ohji, Fabrication and characterization of highly porous mullite
ceramics. Materials Chemistry and Physics, 2003. 80(3): p. 610-614.
Echave, F.J., O. Sanz, and M. Montes, Washcoating of microchannel reactors with
PdZnO catalyst for methanol steam reforming. Applied Catalysis A: General, 2014. 474:
p. 159-167.
3
4.
5.
3
3
6.
7.
Gálvez, M.E., et al., Catalytic filters for the simultaneous removal of soot and NOx:
Influence of the alumina precursor on monolith washcoating and catalytic activity.
Catalysis Today, 2012. 191(1): p. 96-105.
Montebelli, A., et al., Washcoating and chemical testing of a commercial
Cu/ZnO/Al2O3 catalyst for the methanol synthesis over copper open-cell foams.
Applied Catalysis A: General, 2014. 481: p. 96-103.
3
18

3
8.
9.
Hakat, Y., T.V. Kotbagi, and M.G. Bakker, Silver supported on hierarchically porous SiO2
and Co3O4 monoliths: Efficient heterogeneous catalyst for oxidation of cyclohexene.
Journal of Molecular Catalysis A: Chemical, 2016. 411: p. 61-71.
Granlund, M.Z., et al., Comparison between a micro reactor with multiple air inlets and
a monolith reactor for oxidative steam reforming of diesel. International Journal of
Hydrogen Energy, 2014. 39(31): p. 18037-18045.
3
4
0.
1.
Sanz, O., et al., Effect of the thermal conductivity of metallic monoliths on methanol
steam reforming. Catalysis Today.
Gaudillere, C. and J.M. Serra, “Review”. Freeze-casting: Fabrication of highlyporous and
hierarchical ceramic supports for energy applications. Boletín de la Sociedad Española
de Cerámica y Vidrio, 2016. 55(2): p. 45-54.
4
4
2.
3.
Gaudillere, C., J. Garcia-Fayos, and J.M. Serra, Enhancing oxygen permeation through
hierarchically-structured perovskite membranes elaborated by freeze-casting. Journal
of Materials Chemistry A, 2014. 2(11): p. 3828-3833.
4
N.N. Greenwood, A.E., Chemistry of the Elements. Butterworth-Heinemann, 1997: p.
1113.
4
4
4.
5.
Cullity, F.D., Elements of X-ray Diffraction. London: Addison-Wesley, 1878.
Concepcion, P., Lopez, C., Martınez, A., Puntes, V.F., Characterization and catalytic
properties of cobalt supported on delaminated ITQ-6 and ITQ-2 zeolites for the
Fischer-Tropsch synthesis reaction. Journal of Catalysis, 2004. 228:321-32.
Ernst, B., Libs, S., Chaumette, P., Kiennermann, A., Preparation and characterization of
Fischer-Tropsch active Co/SiO2 catalysts. Applied Catalysis A 1999.186:145-68.
Bednarczuk, L., P. Ramírez de la Piscina, and N. Homs, Efficient CO2-regeneration of
Ni/Y2O3La2O3ZrO2 systems used in the ethanol steam reforming for hydrogen
production. International Journal of Hydrogen Energy.
4
6.
7.
4
48.
Liu, J.-Y., et al., Rational design of ethanol steam reforming catalyst based on analysis
of Ni/La2O3 metal-support interactions. Catalysis Science & Technology, 2016. 6(10):
p. 3449-3456.
19

(
a)
(b)
100 µm
30 µm
(
c)
(d)
1
µm
6
µm
Figure 1: Top view SEM micrographs of the porous YSZ-M freeze-cast monolith at different
magnifications. (c) and (d) are focused on the ceramic walls to evaluate the thickness and on the YSZ-M
grains respectively.
Figure 2: XRD pattern of the porous 3YSZ freeze-cast monolith.
20

3
CoYSZ-Mred
3
CoYSZ-M
3
CoYSZ-Pred
CoYSZ-P
3
3
CoYSZ-P-Tred
CoYSZ-P-T
3
2
Theta
0
Co O3
2
Co O Co
2
3
3
CoYSZ-Mred
3
CoYSZ-M
3
CoYSZ-Pred
3
CoYSZ-P
3
CoYSZ-P-Tred
3
CoYSZ-P-T
2
Theta
Figure 3: XRD patterns of the 3CoYSZ-P-T, 3CoYSZ-P and 3CoYSZ-M catalysts after Co infiltrations
and calcination at 600ºC during 3 hours and after reduction under H at 600ºC during 2 hours.
2
22

3
3
3
CoYSZ-M
CoYSZ-P
CoYSZ-P-T
YSZ-P-T
YSZ-M
1
00
200
300
400
500
600
700
800
Temperature, ºC
Figure 4: TPR profile of the catalytic materials (3CoYSZ-M, 3CoYSZ-P, 3CoYSZ-P-T) and their
corresponding supports (YSZ-M and YSZ-P-T).
24

Figure 5: Ethanol conversion and H
2
selectivity versus reaction
O/EtOH = 13,
GHSV = 4500 h . Dash line represents the thermodynamic equilibrium
temperature. Reaction conditions: atmospheric pressure, H
2
-
1
25

Figure 6: By-products selectivity versus reaction temperature. Reaction conditions:
-
1
atmospheric pressure, H O/EtOH = 13, GHSV = 4500 h . Dash lines represent the
2
thermodynamic equilibrium
Before Reaction
After Reaction
(
c)
6
µm
Figure 7: SEM images of the carbon deposition over 3CoYSZ-M after reaction at 600ºC. White arrows
mark carbon deposition.
26

Co nanoparticles
200 nm
Figure 8: TEM image of the carbon deposition over CoYSZ-M after reaction at 600ºC
27

Table 1: BET surface area, size of the metallic Co particles determined by XRD [38] and surface Co
atomic composition determined by EDX-SEM.
2
Catalyst
BET area, m /g Size Co metallic particles, nm Co Atomic Composition, %
3
3
3
CoYSZ-M
CoYSZ-P
2,0
1,9
1,6
3
3
6
2,87
3,01
2,98
CoYSZ-P-T
28

Table 2: Reaction rate for the formation of carbon during the steam reforming of ethanol after catalytic
testing at 600ºC, 500ºC and 400ºC.
Catalyst
Carbon formation, gC/gcat*h
CoYSZ-M
CoYSZ-P
CoYSZ-P-T
0.0140
0.0045
0.0037
29

Table 3: Ethanol conversion and products distribution obtained in the steam reforming of ethanol using
Co monolithic catalysts with and without La. Reaction conditions: atmospheric pressure, H O/EtOH = 13,
2
-
1
GHSV = 4500 h .
Sample
Conv EtOH% Sel CH4%
00ºC
Sel CO%
Sel H2%
Sel C2H2
Sel C2H4O
Sel CO2
4
3
CoYSZ-M
98,66
98,53
7,64
1,51
8,03
10,4
67,91
71,7
0,1
0,03
0,02
16,29
16,33
3Co1LaYSZ-M
0,04
500ºC
3
CoYSZ-M
99,97
99,01
3,19
1,25
2,44
3,29
69,57
72,85
0
0
0
24,8
22,6
3Co1LaYSZ-M
0,01
600ºC
3
CoYSZ-M
100
1,99
1,34
3,98
6,28
71,54
71,24
0
0
0
22,49
21,09
3Co1LaYSZ-M
99,82
0,05
Table 4: Reaction rate for the formation of carbon during the steam reforming of ethanol at 400ºC, 500ºC
and 600ºC and using Co monolith with and without La
Catalyst
Carbon formation, gC/gcat*h
CoYSZ-M
CoLaYSZ-M
0.0140
0.0012
30

Table 5: Ethanol conversion and products distribution obtained in the steam reforming of ethanol using
monolithic catalysts with low Co loading and the presence of a second metal (Pt or La). Reaction
conditions: atmospheric pressure, H
O/EtOH = 13, GHSV= 4500 h-1
2
Sample
Conv EtOH%
00ºC
Sel CH4%
Sel CO%
Sel H2%
Sel C2H2
Sel C2H4O
Sel CO2
4
3
1
1
1
1
CoYSZ-M
98,66
62,59
76,2
7,64
3,55
0,6
8,03
6,96
3,22
2,02
3,55
67,91
72,06
77,46
61,62
58,69
0,1
0,03
5,2
16,29
12,16
15,01
22,03
19,97
CoYSZ-M
0,07
0,04
0,04
0,21
Co0.5LaYSZ-M
Co0.25PtYSZ-M
Co0.8PtYSZ-M
3,67
0
70,45
73,24
13,29
16,26
1,32
500ºC
3
1
1
1
1
CoYSZ-M
99,97
84,66
87,04
76,74
78,48
3,19
1,33
1,24
4,04
6,63
2,44
1,72
2,5
69,57
74,85
73,18
71,91
68,59
0
0
24,8
CoYSZ-M
0,03
0,02
0,02
0,08
0,51
0,85
0,57
0,63
21,56
22,21
22,05
22,29
Co0.5LaYSZ-M
Co0.25PtYSZ-M
Co0.8PtYSZ-M
1,41
1,78
600ºC
3
1
1
1
1
CoYSZ-M
100
1,99
1,63
1,23
1,95
2,12
3,98
2,88
4,04
4,18
2,43
71,54
73,14
73,19
71,54
71,61
0
0
22,49
22,32
21,54
22,06
23,7
CoYSZ-M
99,17
99,8
0
0
Co0.5LaYSZ-M
Co0.25PtYSZ-M
Co0.8PtYSZ-M
0
0
81,89
94,36
0,03
0,01
0,24
0,13
31