Surface State and Catalytic Performance of Ceria-Supported Cobalt Catalysts in the Steam Reforming of Ethanol

Article abstract of DOI:10.1002/cctc.201601343

The effect of the particle size and the addition of a K promoter on the oxidation state of Co/CeO₂ was investigated and correlated with selectivity for ethanol steam reforming performed with various H₂O/EtOH molar ratios. Spectroscopic studies showed that the oxidation state of the catalyst components depends on the H₂O/EtOH molar ratio. Surface oxygen-containing species (OH and K^δ+?O^δ?) are necessary to achieve a good catalyst selectivity, to decrease the amount of coke deposited and to change the nature of the product from fully dehydrogenated C=C to CH_x. Besides the surface concentration of oxygen-containing species, the catalyst morphology and the location at which oxygen-containing species are chemisorbed may be equally important to their abundance; the selective ceria-supported Co catalyst should have well-dispersed Co particles deposited on a well-dispersed support.

Full text of DOI:10.1002/cctc.201601343

Accepted Article
Title: The surface state and the catalytic performance of Co/CeO2
catalysts in the steam reforming of ethanol
Authors: Sylwia Turczyniak, Magdalena Greluk, Grzegorz Słowik,
Wojciech Gac, Andrzej Machocki, and Spyridon Zafeiratos
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201601343
Link to VoR: http://dx.doi.org/10.1002/cctc.201601343
A Journal of
www.chemcatchem.org

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
The surface state and the catalytic performance of Co/CeO2
catalysts in the steam reforming of ethanol
Sylwia Turczyniak,*[a,b] Magdalena Greluk,[a] Grzegorz Słowik,[a] Wojciech Gac,[a] Spyridon Zafeiratos,[b]
and Andrzej Machocki[a]
metal promotion [22-24].
Abstract: The effect of particles size, and addition of K promoter on
the oxidation state of Co/CeO2 was investigated and correlated with
selectivity for the ethanol steam reforming (ESR) carried out with
various H2O/EtOH molar ratios. Spectroscopic studies showed that
the oxidation state of catalysts components depends on the
H2O/EtOH molar ratio, and it changes with increasing water extent. It
was found that surface oxygen-containing species (OH and Kδ+–Oδ-)
are very important factors to achieve a good catalyst’s selectivity, for
lowering the amount of coke deposited and changing its nature from
fully dehydrogenated C=C to CHx. Beside surface concentration of
oxygen-containing species, the catalyst’s morphology and location
where oxygen-containing species are chemisorbed may be equally
important to their abundance – the selective ceria-supported cobalt
catalyst should have well dispersed cobalt particles deposited on the
well dispersed support.
The oxidation state of cobalt under the ESR reaction has
been a subject of many studies [2, 3, 5, 7, 8, 10, 25-27]. The
evidence for co-existence of both forms of cobalt (metallic and
cobalt(II) oxide) under the reaction conditions has been provided
by various techniques [2-5, 7, 8, 17, 19, 21, 26-31]. Recent
density functional theory calculations on ethanol transformations
on the surface of cobalt catalysts suggest, that while carbon
monoxide and hydrogen are products more favourable on
metallic cobalt species, acetaldehyde is the main final product
on Co(II) sites [32]. The similar conclusion has been drawn from
experiments in which a cobalt foil and cobalt oxide thin films
were exposed to ethanol vapours [19]. Using model
Co/ZnO catalysts, E. Martono et al. [33] and Y.T. Law et al. [34]
also showed that metallic cobalt is active for ethoxide
decarbonylation into carbon monoxide and hydrogen, while CoO
provides the active sites for ethanol dehydrogenation. In-situ
XPS studies of Co/CeO2 [29] combined with on-line products
analysis confirmed that also over real catalytic systems, in the
presence of metallic form of cobalt only, carbon monoxide is the
most favourable product. Even though the effect of cobalt
oxidation state on catalytic performance has been widely
investigated over unsupported [17, 18] and supported systems
[28], as yet there is no consistency in what is the most active
and selective form of cobalt under the ESR [35] and all
combinations (i) a mixture of Co(0) and Co(II) [2, 4, 5, 33, 36,
37], (ii) metallic cobalt [4, 18, 26, 31, 28] or (iii) Co(II) [7, 23, 38,
39] have been proposed in the so far published results.
Introduction
In recent years cobalt-based catalysts for ethanol steam
reforming (ESR) reaction have been in the focus of numerous
studies seeking understanding the catalysts’ function under
working conditions. The aim is to provide rational improvement
strategies so as to develop a low-cost, active and sustainable
ESR catalyst [1]. Among the critical issues previously reported
are (i) the effect of the initial state [2-5], (iii) the H2O/EtOH molar
ratio [6], (iv) reaction temperature [2, 5, 7, 8], (v) the influence of
a
support (i.e., surface acidity, morphology, particle size,
oxygen-storing and releasing capacity and oxygen mobility) [3,
8-16], (vi) the nature of cobalt active sites [2, 4, 5, 7], (vii) the
capability of cobalt [17-19], support [20, 21], and their interface
to convert ethanol and water, and finally (viii) the role of alkali
The role of a support is also of major importance and to date,
ceria-supported catalysts are considered to be one of the most
promising systems [40]. This is assigned mainly to the facility of
ceria to reversible change of its oxidation state between Ce(IV)
and Ce(III) [41-42], and its remarkable oxygen storage capacity
[15, 43-46]. Superior properties of ceria are believed to preserve
the catalyst surface area (due to physical stabilisation of cobalt
crystallites preventing them from sintering [15]) and the catalytic
activity [15, 47]. It was shown that ceria may play an important
role in modifying the formation of active cobalt species (for
[a]
Dr. S. Turczyniak, Dr. M. Greluk, M.Sc. G. Słowik, D. Sc., Assoc.
Prof. W. Gac, D.Sc., Assoc. Prof. A. Machocki
Faculty of Chemistry
Maria Curie-Sklodowska University in Lublin
3 Maria Curie-Skłodowska Square, 20-031 Lublin, Poland
E-mail: sylwia.turczyniak@poczta.umcs.lublin.pl
Dr. S. Zafeiratos
Institut de Chimie et Procédés pour l’Energie, l’Environnement et la
Santé (ICPEES), ECPM, UMR 7515 CNRS-
Université de Strasbourg
[b]
25, rue Becquerel, 67087 Strasbourg Cedex 02, France
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
example via release of lattice oxygen) [34, 48], dissociation of
water [49, 50], transformation of ethanol [49, 51], and in altering
the reaction pathways [13, 52, 53]). Not only the support nature
is substantial in the overall ESR process. Another important
It is generally accepted that promoters improve the stability
of catalysts. Promoters (e.g., alkali metals, such as sodium [58,
59], potassium [22-24, 31, 60-62] or noble metals, like rhodium
[7, 10, 40, 63], platinum [62, 64]) are widely used to improve
catalysts’ activity [24, 59] and extent catalysts’ lifetime by
making sintering [47] and carbon formation [23, 24, 31, 59, 60]
more difficult. The group of J.-W. Snoeck [65] examined the
effect of potassium loading on nickel-catalysts on the rate of
carbon gasification. Their findings showed that the most
satisfying results were obtained with 1.6–2 wt.% of K2O content.
Besides, S. Ogo et al. [23] showed that 1–2 wt.% potassium
addition improved both the activity and selectivity of the
Co/Al2O3 catalyst, which stays in line with results of other
researchers [22, 24, 60]. The authors [23] deduced that
beneficial properties of potassium addition are related to the
reduction of Co(II) into Co(0), stabilization of acetate species,
and suppression of methane formation. In some studies
potassium was also found to prevent ethylene formation [31],
which is well-known coke precursor [9, 66], consequently
hindering coke formation.
factors influencing
a catalyst’s activity and selectivity are
morphology and textural properties of the support [3, 13, 20].
I.I. Soykal and co-workers investigated the influence of
particles size (3.5 nm and 120 nm) [3, 21], and morphology [11]
on reducibility of ceria and its catalytic performance under the
ESR. They found that whether ceria was pre-reduced or not, it
converge to the same oxidation state (mixture of Ce(IV) and
Ce(III)) at given temperature, and exhibited a similar ESR
performance; however, a highly dispersed ceria produced much
more ethylene than its low-dispersed counterpart [3, 21].
In contrast to the studies presented in ref. [3, 21], the results of
A. M. da Silva et al. [54], and N. Laosiripojana and
S. Assabumrungrat [20] showed that the low surface area ceria
exhibited high selectivity to ethylene, whereas its high surface
area counterpart was selective towards acetaldehyde [54].
However, the authors agreed [21, 54] that the nano-sized ceria
ensures higher ethanol conversion and hydrogen yield.
Therefore, discrepancy between the results of different teams [3,
8, 13, 21], while concerning the ESR carried out over Co/CeO2,
catalysts, may arise from differences in morphology and catalytic
properties of used ceria support.
The catalyst’s stability is in the core of the effort to develop
new, advanced catalysts for the ESR. Although Co/CeO2
catalysts are very promising candidates for the ESR they exhibit
loss of the activity and selectivity in long-testing experiments,
even after alkali metal promotion, which is commonly attributed
to coke deposition. Apart from the use of alkali promoters, the
increase of the water to ethanol molar ratio in the feed can be
also used to moderate carbon accumulation with time-on-stream
[6]. K. Vasudeva et al. [67] suggested that carbon formation on
catalysts’ surface occurs only at low H2O/EtOH molar ratios
(<2/1), whereas the group of V. Mas [68], that the
H2O/EtOH = 3/1 mol/mol and T>230°C are required to avoid
coke formation. Other group of researchers [69] implied that the
steam to carbon molar ratio up to 10/1 is needed to obtain the
best production of hydrogen, minimize the formation of carbon
monoxide and methane, and to avoid carbon deposition.
A great number of publication is focused on the role of cobalt
and ceria oxidation state in the ESR [2, 3, 7, 8, 10, 11, 21, 29],
whereas only in a few the attempt was made to explain the role
of surface adsorbed hydroxyl species, or generally speaking
surface adsorbed oxygen-containing species [44, 55-57]. The
group of A.M. da Silva et al. [43] suggested that hydroxyls
adsorbed on the ceria’s surface assist in coke removal from
cobalt, whereas another paper [1] higlight that
a high
concentration of surface adsorbed oxygen-containing species
may also induce oxidation of metal surface. In the second case it
is obvious that adsorbed species would influence a catalyst’s
selectivity [48], therefore, playing an important role in the ESR.
S.M de Lima et al. [55] presented a scheme of ethoxy species
transformation in which dehydrogenated species (acetyl
Although cobalt-based catalysts have been in the focus of
many research groups, there are no works dealing with more
than one or two variable factors at a time. Moreover, there were
no attempts to use surface sensitive methods to characterize
potassium-promoted Co/CeO2 systems under the different
H2O/EtOH molar ratios. In this study catalysts of similar chemical
composition, but completely different properties were used
deliberately, as in ref. [3, 8, 13, 21]. The present work aims to
species)
undergo
support-induced
oxidation
by
oxygen-containing species (OH or surface oxygen adatoms)
forming acetates. Acetate species can be further oxidized by
surface adsorbed oxygen-containing species to carbonates
followed by decomposition to carbon dioxide.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
show the influence of the above factors in a holistic approach
focusing mainly on the surface state. The main objectives of this
study was to find (i) cobalt and ceria oxidation state after
activation in hydrogen and after the ESR, (ii) the effect of the
ceria support and cobalt active phase particles size on catalyst’s
reducibility and its behaviour in the ESR, (iii) the effect of
potassium promotion on the catalyst’s surface state after
reduction and the ESR, (iv) the influence of the H2O/EtOH molar
ratio on the surface state of the quenched Co/CeO2 catalysts,
and (iv) the impact of the Co/CeO2 catalysts’ surface state on
their catalytic performance in the ESR.
Therefore, on the basis of cobalt oxide crystallite size (Table 1)
the reduction of the HS-catalyst at higher temperature can be
rationalized. In the presence of potassium, additional effects
might influence the reducibility of the catalyst apart from the
cobalt crystallites size.
Fig. 1 shows the XPS core level spectra of the elements
found over the catalyst’s surface after reduction at 420°C in
hydrogen (detailed characterization of the pristine samples can
be found in the Supplementary Information, Fig. S1). For
quantification of the data, the collected Ce 3d high resolution
spectra (Fig. 1a) were curve fitted [29, 71] with combination of
two reference spectra of Ce(IV) (\\\) and Ce(III) (///) recorded
over cerium(IV) oxide and hexahydrate cerium(III) nitrate, as
suggested in ref. [8]. The peak fitting procedure of references is
presented in Supplementary Information, Fig. S2. The peaks
position, full width at half maximum were constrained as
proposed in ref. [8]. Similarly, two components recorded on
reference samples were also used for Co 2p3/2 (Fig. 1b) fitting:
one for metallic cobalt (///), located at 778.6 eV, and a second
around 781.4 eV – formally ascribed to oxide-like component
(CoO and/or Co–OH) (\\\). As similar to ceria references, the
fitting procedure of cobalt references (based on ref. [72]) has
been presented in Supplementary Information, Fig. S3. In
literature [72] metallic cobalt is fitted with a characteristic
asymmetric main peak and two plasmon loss peaks (located at
3.0 and 5.0 eV higher binding energy than the main component),
origination of which was explained elsewhere [72]. From
Fig. 1(a-b) it is evident, that after hydrogen treatment the surface
Results
1.1. Reduction of the catalysts in hydrogen
As shown in Table 1 the BET surface area of HS- and
LS-catalysts differs by about one order of magnitude, but after
addition of potassium the surface area of the HS-catalyst
decreases considerably. In addition, although cobalt wt.%
content is comparable in all catalysts (8–9 wt.%),the cobalt
surface area is significantly higher for the HS-catalyst
(47.2 m2/g) as a result of much lower cobalt particle size
(3.8 nm). The potassium promotion increases average cobalt
particle size (13.9 nm) and lowers cobalt surface area (4.1 m2/g)
as compared to the unpromoted HS-catalyst, making it
comparable to ones obtained for the LS-catalyst (1.5 m2/g).
However, one should note that cobalt crystallites in the
Table 1. Results of the Co/CeO2 catalysts characterization
Catalyst
BET surface
area, m2/g[a]
Volumes
of pores,
cm2/g[a]
Pore
diameter,
nm[a]
Cobalt
content,
wt.%[b]
Average crystallite
size, nm[c]
Average
cobalt
Cobalt
Cobalt surface
[d]
dispersion,
area, m2/gcat.
[d]
crystallite size,
%
nm[d]
Co3O4
CeO2
LS-Co/CeO2
HS-Co/CeO2
HS-KCo/CeO2
5.3
0.03
0.15
0.12
28.2
11.0
12.4
9.00±0.27
9.55±0.38
8.44±0.34
23.3
8.4
87.1
25.6
26.5
39.3
3.8
2.5
26.6
7.2
1.5
17.1
4.1
47.2
29.6
9.3
13.9
[a] Determined by the low-temperature N2 adsorption. [b] Determined by XRF. [c] Determined by XRD. [d] Determined by the hydrogen chemisorption
measurements.
LS-catalyst were much bigger (39.3 nm) as compared to the
other catalysts.
of all catalysts was reduced, however, to a different extent. It
should be pointed out that in Fig. 1b the intensity of the fitted
Co(II) component is comparable to the noise level of the spectra,
therefore the error bars in determined Co(II) percentage
contribution might be encumbered with a meaningful error. The
obtained spectra proved that similarly to in-situ XPS studies [2, 3,
The reducibility of cobalt oxides is strongly related to the
cobalt crystallite size [3, 8, 21], as well as to the presence of
additives [7, 22, 40, 61-63]. Large cobalt oxide particles were
expected to reduce more easly than the smaller ones [13, 70].
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
8, 25, 29], and in contrast to so far published ex-situ XPS
studies [7, 10, 40] it is possible to efficiently maintain cobalt
mainly in a metallic form, if it predominates.
due to hydrogen spill-over effect [13, 44, 75, 77-79]. Recently it
has been shown that the reduction degree of ceria (average size
of ceria crystallites was 4 nm) in the Co/CeO2 might be lower
than that of a bare ceria support [8], and this
a) Ce 3d
b) Co 2p3/2
c) O 1s
effect is dictated by the temperature.
The new features appearing on the Ce 3d
d) C 1s, Ce 4s & K 2p
Co(II) 5%
Ce(III) 3%
K1
K2
spectra, shown in Fig. 1a, indicate that the
2
O
e
O4
CO23-
O1
ceria support of the LS-catalyst was partially
C
/
O2
o
C
K
reduced Ce(III) = 16%, while the supports of
-
S
H
Ce(III) 1%
Co(II) 18%
Co(II)
promoted and unpromoted HS-catalysts
Co(0)
remained almost fully oxidized. The evolution
2
O
of Ce 3d spectra, in reducing conditions has
CO23-
e
O1
O1
C
/
O2
O2
o
been widely studied by many authors [21, 41,
C
-
S
H
Ce(III) 16%
Co(II) 0%
79-85]. Higher diffusion ability of oxygen
atoms for the HS-ceria support due to the
2
O
e
C
/
CO23-
CHx
higher defects’ concentration leads to the
o
C
-
S
different reduction level between the
L
920 910 900 890 880 791
786
781
776 535 533 531 529 527
299 294 289 284
HS-ceria and LS-ceria [41].
Binding energy (eV)
The O 1s spectra (Fig. 1c) of the
Figure 1. The high resolution XPS spectra of Ce 3d (a), Co 2p3/2 (b), O 1s (c) and C 1s, Ce 4s & K 2p
(d) collected after the reduction (H2/Ar= 50/50 cm3, T = 420°C, ptotal = 1 atm, 1 h) of the LS-Co/CeO2,
HS-Co/CeO2 and HS-KCo/CeO2 catalysts; a) Ce(IV) (\\\), Ce(III) (///), b) Co(II) (\\\), Co(0) (///), c) O1 –
lattice oxygen, O2 – hydroxyl and carbonate species, O4 – carbonate species bounded to potassium,
d) probable assignments of K1 and K2 species can be found in the text.
unpromoted catalysts are very similar. The
detailed information of each component
contribution in the overall spectra can be
found in Supplementary Information – Table
Results show complete reduction of cobalt oxides on the surface
of the LS-catalyst (within experimental error, therefore it cannot
be excluded that traces of the CoOx phase do not exist), while
on the HS and HS-K traces of CoOx remain after the hydrogen
treatment. The presented XPS results are in a good agreement
with the TPR profiles shown in the Supplementary Information
Fig. S4. The reduction behaviour of nano- and micro-Co/CeO2
was also examined by I.I. Soykal et al. [3], whose in-situ X-ray
absorption fine structure spectroscopy studies demonstrated
that the extent of cobalt oxides reduction was higher with the
nano-Co/CeO2 catalyst. The result was attributed, by the authors,
to the small particles size and high dispersion of the sample.
This is somewhat counterintuitive, since is admitted that smaller
cobalt oxide particles are more difficult to reduce than larger one
[13, 70, 73]. Some authors postulate that persistent Co(II)
species can be a result of a strong metal-support interaction [13,
73], slight encapsulation [10], and/or penetration Co2+ ions into
the support lattice [40, 74].
S1. The main peak (530 eV) can be safely assigned mainly to
the ceria lattice oxygen (O1) [86-89] (however, in literature at
this binding energy the lattice oxygen in CoOx was also recorded
[19]), while the assignment of the high binding energy
component at 531.5 eV (O2) is controversial. Some authors
assign it directly to the presence of –OH groups [90-93], others
rather generally identify them as adsorbed oxygen ions (O-
species) on surface oxygen vacancies [87, 94, 95]. Quite
recently Z. Liu et al. [86] came up with the proposal that in
ceria-supported systems the peak located at 1.4 eV higher
binding energy than the main oxygen component is probably
related to the ceria hydroxide layer Ce3+(OH)x. On the spectra of
the potassium-promoted catalyst, an additional peak at 532.2 eV
(O4) was found. In literature this peak is regarded as the
fingerprint of the presence of amorphous layer of potassium
oxides [96], potassium hydroxide [97], or potassium carbonate
[98, 99].
The presence of amorphous potassium oxides on the
surface of the ESR catalysts was recently confirmed by S. Ogo
et al. [23]. The nature of potassium oxides is still under
discussion. B. Lamontagne et al. [110] have studied the alkali
It is well known that pure ceria is difficult to reduce in
hydrogen at relatively low temperatures [40, 75, 76], however, in
the presence of cobalt (or other metal) it can be partially reduced
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
metal promotion on silicon oxidation and their measurements
revealed the existence of four potassium oxides formed during
increasing oxygen exposure. They assigned the peak at
531.0 eV to K2O2, 532.0 eV to K2O3 and 534.2 eV to KO2.
Contrary, G. Pirug et al. [101] have assigned the component at
531.7 eV to K2O2 surface species. Z. Hou et al. [97] have
suggested KOH as a form of potassium on the surface.
the oxygen peak at 532.3 eV can be assigned also to the
existence of potassium carbonates [98]. It should be noted that
BE’s of the K 2p3/2 for as deposited KNO2, K2CO3 and KOH films
on the Fe foil were very close (293.7–293.8 eV) and the shift
towards higher BE’s (293.9–294.1 eV) were observed after
samples’ reduction [103, 104]. Accoring to the work of G. Maniak
et al. [105] regardless of potassium compound used for cobalt
According to M. Carlsson [102] potassium forms
Table 2. The percentage contribution (at.%) of elements on the surface of catalysts
under the ESR (calculated on the basis of high resolution XPS spectra after taking into
account the atomic sensitivity factor of each element)
hydroxide species resulting in the increase of the surface
basicity. In further part of this paper we will try to shed a
light on the nature of K1 and O4 components.
Catalyst
H2O/EtOH
(mol/mol)
Contribution of element (at. %)[a]
Co
Ce
O
C
K
Before reduction, carbonaceous deposit was detected
on the surface of all catalysts (Supplementary Information,
Fig. S.1). The reduction procedure allowed for the
removal of adsorbed carbon impurities and purification of
the surface (Fig. 1d). For the K-promoted catalyst the
peaks at 294.3 and 297.0 eV are definitely related to
presence of K+ ions (K 2p core level). The intensity of
these peaks increased after hydrogen pre-treatment
(Fig. 1d), as compared to the calcined sample
(Supplementary Information, Fig. S1d). The calculated
percentage contribution of potassium on the surface of
the fresh sample was 3.2 at.%, whereas after reduction –
17.8 at.% (Table 2), suggesting surface’s enrichment in
potassium (its re-dispersion) after reduction. Analysis of
-
-
LS-Co/CeO2
HS-Co/CeO2
HS-KCo/CeO2
LS-Co/CeO2
HS-Co/CeO2
HS-KCo/CeO2
LS-Co/CeO2
HS-Co/CeO2
17.358
7.935
4.388
0.779
1.963
1.680
15.871
4.374
22.412
26.334
17.825
1.490
49.184
54.070
54.171
3.877
11.046
11.661
5.775
[b]
H2
17.842
-
93.854
84.445
22.154
5.969
3/1
9/1
-
4.121
9.471
13.443
17.588
18.277
47.475
60.571
46.313
15.248
-
-
31.035
HS-KCo/CeO2
1.322
9.227
50.441
13.657
25.353
-
-
LS-Co/CeO2
HS-Co/CeO2
18.882
7.549
18.090
23.183
60.438
56.136
2.590
12/1
13.132
HS-KCo/CeO2
0.818
4.384
48.687
13.255
32.856
[a] Calculated as: ni = ni/∑ ni-j·100%, where ni = {Co, Ce, O, C, K} and
ni-j = Co+Ce+O+C+K. [b] Results obtained after the catalysts’ pre-reduction (H2/Ar) at
420°C for 1 h.
oxide promotion, shifts on the BE scale are not observed.
the K 2p region (Fig. 1d) allowed to observe a small asymmetry
and the shift of the peak (the K 2p3/2 components located at
293.8±0.2 and 294.6 ± 0.2 eV respectively were denoted as K1
and K2) towards higher binding energies (BE), as compared to
further results (Fig. 2d-4d).
Basing on this observation, we assigned the K1 component to all
potassium bonded to oxygen species.
1.2. XPS characterization of catalysts after the ESR with a
stoichiometric H2O/EtOH molar ratio
Even though the chemical nature of potassium species is not
clear yet, it is convincing that potassium state would depend on
the local chemical environment of the surface [104], as well as
that potassium species can be transformed one into another. In
the view of the above discussion, in this work the shift of the
K 2p (appearance of the component K2) towards slightly higher
BE’s is suggested to be ascribed to the change of electronic
properties of potassium.
The XPS results recorded after the ESR reaction at
a
stoichiometric (3/1 mol/mol) H2O/EtOH molar ratio (T = 420°C,
ptotal = 1 atm, t = 1 h), show partially reduced ceria support and
metallic cobalt with traces of CoOx (Fig. 2a and 2b). It is
interesting to note that as compared to the state prior to the ESR
(reduced samples, Fig. 1), the stoichiometric reaction mixture
seems to be more efficient in ceria’s reduction than hydrogen,
which is in agreement with previous reports [25, 36]. On the
other hand, it is different for cobalt, since its oxidation state
remains almost the same as before the reaction (please
compare Fig. 1b and 3b). Our previous in-situ XPS studies of the
nano-Co/CeO2 catalyst, carried out for the H2O/EtOH molar ratio
of 3/1 (420°C, ptotal = 0.2 mbar) [29] revealed that whereas
H.P. Bonzel et al. [103, 104] have presented the results of
XPS studies of K-promoted surface of the Fe foil and suggested
that the presence of K 2p3/2 component around 294 eV and
oxygen at 532.2 eV is strongly supporting the hypothesis of KOH
existence [103], however, only in the case of absence of the
carbonate peak at ~290 eV [104]. In the presence of carbonates
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
(531.5 eV) increases, and a new component
located around 532.8 ±0.2 eV (denoted as
O3), appears. According to its binding energy
the O3 component can be assigned to
surface adsorbed water species [87, 108,
109]. Moreover, the concentration of surface
adsorbed hydroxyl species (O2) is clearly
higher for the LS and HS-K catalysts.
a) Ce 3d
b) Co 2p3/2
c) O 1s
d) C 1s, Ce 4s & K 2p
Ce(III) 22%
Co(II) 6%
22%
K1
O4
O2
2
CO23-
C-O
O
e
CHx
O1
C
/
o
O3
C
K
-
S
Co(II) 14%
H
Ce(III) 17%
Co(0)
84%
C=C
2
Co(II)
O2
O

O1
O1
e

The unpromoted catalysts after 1 h of the
ESR were covered by carbon deposit, as
seen in the C 1s region (Fig. 2d). The C 1s
peak position at 284.6 eV, as well as its
characteristic asymmetric shape, indicates
for the presence of graphitic carbon species
[110]. The higher binding energy component
(289.4 eV) is characteristic for both Ce 4s
region [84], carbonate-type species [86, 104]
and even carboxylates [86]. Carbon
dominates the surface with 94% and 84%
atomic concentration for the LS- and
O3
C
/
o
C
-
S
C=C
H
Ce(III) 19%
Co(II) 5%
94%
2
O
e
C

/

o
O3
O2
C
-
S
L
920 910 900 890 880 791
786
781
776 535 533 531 529 527
299 294 289 284
Binding energy (eV)
Figure 2. The high resolution XPS spectra of Ce 3d (a), Co 2p3/2 (b), O 1s (c) and C 1s, Ce 4s & K 2p
(d) collected after the ESR (H2O/EtOH = 3/1 mol/mol, T = 420°C) over the pre-reduced (H2,
T = 420°C) LS-Co/CeO2, HS-Co/CeO2 and HS-KCo/CeO2 catalysts; a) Ce(IV) (\\\), Ce(III) (///), b)
Co(II) (\\\), Co(0) (///), c) O1 – lattice oxygen, O2 – hydroxyl and carbonate species, O3 – water, O4 –
carbonate species bounded to potassium, d) K1 – potassium carbonate. On the picture (d) also the
percentage atomic contribution of carbon on the surface was given.
HS-catalysts,
respectively
(Table 2).
ceria’s oxidation state was influenced by the conditions
Potassium inhibits carbon deposition (formation of C=C and CHx
(hydrogen or ESR reaction mixture), cobalt oxidation state
remained the same. It is also worth to mention about the results
of ex-situ XPS studies carried out over the pre-reduced 10%
Co/CeO2-ZrO2 catalyst, exposed to the slightly higher H2O/EtOH
molar ratio (4/1 mol/mol, 450°C) revealed re-oxidation of metallic
cobalt to CoOx accompanied by ceria’s reduction [36].
species) down to <10%; however, a new peak at 290 eV
2-
indicates for the presence of CO3
species (total
carbon-containing species coverage including carbonates was
<22% as shown in Table 2). The K 2p spectrum of the HS-K
catalyst shows a doublet at 293.8 and 296.6 eV which is very
close to the BE of the K 2p doublet (293.9 and 296.6 eV),
recorded by E. de Smit et al. [111]. In this case the peak was
assigned to metastable basic potassium carbonate species.
Potassium containing species are rather difficult to
characterize, especially when they are highly dispersed [112,
113], therefore, knowledge about them is limited [112].
Nevertheless, it is worth to made an attempt to shed a light on
their nature. The noticeable intensity increase of the C 1s peak
at 290 eV allows to suppose that the K1 component is at least
partially related to the presence of potassium carbonates, e.g.,
The higher reduction degree of ceria, as calculated based on
Ce 3d spectra, was found for the HS-K catalyst (22% of Ce(III)).
In addition, as compared to the unpromoted HS-catalyst, cobalt
of HS-K catalyst is more reduced (Co(II) = 5%). Previous reports
of L. del Río et al. [38, 106] demonstrated that in the ESR
reaction (H2O/EtOH ≈ 6/1 mol/mol) at low temperatures (300–
350ºC) potassium loading prevents cobalt oxides reduction to
the metallic phase, whereas our studies provided the evidence
that in fact it facilitates catalyst’s reduction in the ESR reaction at
higher temperature (420ºC). However, based on available
publications concerning potassium-promoted catalysts [23, 60],
there are some presumptions that similarly to ZnO promoter [26,
107], depending on the type of the support, potassium may
promote or hinder cobalt oxidation in the ESR.
2-
-
2K+CO3 or K+CO2 . Simultaneously with the increase of the
K 2p signal intensity, the O4 signal intensity was increasing (Fig.
2c and 2d). Further application of the same reasoning lead one
to expect that the O4 component in the O 1s overall spectrum
might be related to potassium bound to oxygen from carbonate
species (further abbreviated Kδ+–Oδ-). In order to shed a light on
the origination of both K1 and O4 components, the high-
Comparison of the O 1s spectra of the samples after
hydrogen pre-treatment (Fig. 1c) and the ESR reaction (Fig. 2c)
shows, that the relative intensity of the O2 component
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
resolution XPS spectra of K2CO3 (pure for analysis, ≥99.0%)
reference sample were collected. The experimentally derived
data were used as a reference in calculating K 2p/CO32-/O 1s
ratio related to potassium carbonate species. Since in cerium
containing systems the Ce 4s region is overlapping with C 1s
signal, the contribution of the percentage the surface adsorbed
increase of the O 1s component at 532.2 eV (Fig. 3c). The
excess of water had a beneficial effect on suppression of carbon
formation. The carbon atomic fraction, which was 94 - 84% in
the H2O/EtOH = 3/1 molar ratio, drops down to 6 and 31% for
the LS- and HS-catalysts, respectively. However, one should
note that the given carbon percentage contribution is considered
to the whole C 1s core level, meaning that C=C, CHx, and
2-
CO3 species on the surface of catalyst was performed after
2-
subtraction of the Ce 4s impact on this component. The
experimentally derived data led to the conclusion that the K 2p
and O4 components are related to potassium carbonates.
chemisorbed C–O and CO3 species are included. Potassium
addition decreases the carbon fraction at the HS-K catalyst from
22% (for HS) to 14%, while the sole C 1s component around
290 eV (Fig. 3d) is characteristic for carbonate and carboxylate
species as discussed in the previous section.
1.3. XPS characterization of catalysts after the ESR at high
H2O/EtOH molar ratios
Further increase of the H2O/EtOH molar ratio to 12/1 (Fig. 4)
induces cobalt oxidation (34% of Co(II)) similar for both LS- and
HS-catalysts, thus it might be concluded that water content has
a limited effect on cobalt re-oxidation process [29, 36]. In the
case of the HS-K catalyst, the Co 2p peak has been in the signal
noise level after 1 h of the ESR, indicating that cobalt particles
were buried underneath a thick layer (around 3 to 4 nm) which
attenuates Co 2p signal. The intense signal of the K1 (K 2p) and
O4 (from the O 1s) components suggests that cobalt particles
were most probably covered by potassium bonded to oxygen
layer (K–O). The Ce 3d peak was also barely detectable,
The catalysts’ resistance to coke formation can be improved by
increasing water to ethanol molar ratio in the reaction feed [68,
69, 114]. In Fig. 3 the XPS spectra of catalysts in the
H2O/EtOH = 9/1 mol/mol mixture are shown. The surface of the
unpromoted catalysts is more oxidized as compared to the 3/1
molar ratio. After the ESR carried out for the H2O/EtOH molar
ratio of 9/1 the Ce(III) ions surface concentration for was 14%
and 11% for the LS- and HS-catalyst respectively, while the
concentration of Co(II) species increases considerably (16.4%
and 30.3% of Co(II) for LS- and HS-catalyst). The in-situ XPS
studies of H. Sohn et al. [8], devoted to the
ESR reaction (H2O/EtOH
= 9/1 mol/mol,
a) Ce 3d
b) Co 2p3/2
c) O 1s
O4
d) C 1s, Ce 4s & K 2p
K1
450ºC) over the nano- and micro-Co/CeO2
led to the similar conclusion, that higher
amounts of CoOx phase were detected on the
surface of micro-dispersed catalyst. From the
other hand, in contrast to our work, in the
cited publication [8] it was found that the
surface of nano-ceria was more reduced than
the surface of the sample with larger particles
size. Higher concentration of Ce(III) on the
surface of the nano-ceria was assigned to the
higher lattice strain in smaller particles [115,
116]. At the HS-K catalyst the H2O/EtOH
molar ratio of 9/1 also induces more oxidized
cobalt, but notably has the reverse effect on
cerium, which is more reduced than in the
molar ratio of 3/1. In addition, the atomic
fraction of potassium (Table 2) increases
considerably followed by simultaneous
Ce(III) 25%
Co(II) 20%
14%
O2
2
CO23-
C-O
O
e
O1
O1
O1
C
/
O3
o
C
K
-
S
Co(II) 30%
Co(II)
H
Ce(III) 11%
31%
Co(0)
C=C
2-
2
O2
O2
CO
O
3 C-O
e
O3
C
/
o
C
-
S
Ce(III) 14%
Co(II) 16%
6%
H
2
O
CO23-
CHx
e
C
/
o
O3
C
-
S
L
920 910 900 890 880 791
786
781
776 535 533 531 529 527
299 294 289 284
Binding energy (eV)
Figure 3. The high resolution XPS spectra of Ce 3d (a), Co 2p3/2 (b), O 1s (c) and C 1s, Ce 4s & K 2p
(d) collected after the ESR (H2O/EtOH = 9/1 mol/mol, T = 420°C) over the pre-reduced (H2,
T = 420°C) LS-Co/CeO2, HS-Co/CeO2 and HS-KCo/CeO2 catalysts; a) Ce(IV) (\\\), Ce(III) (///), b)
Co(II) (\\\), Co(0) (///), c) O1 – lattice oxygen, O2 – hydroxyl and carbonate species, O3 – water, O4 –
carbonate species bounded to potassium, d) K1 – potassium carbonate. On the picture (d) also the
percentage atomic contribution of carbon on the surface was given.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
indicating that K–O is deposited without apparent distinction
over cobalt and ceria support. In order to estimate the
time-dependence of the K–O layer formation, the ESR at shorter
reaction period (15 min) was carried out (inset of Fig. 4b). In this
case the Co 2p signal directly suggests that K–O layer was
thinner than after 1 h of the ESR, therefore its formation is
time-depended.
monoxide. Adsorbed species were converted on the surface into
carbonate-type species.
The presented results show that regardless of the H2O/EtOH
ratio, a mixed metallic-ionic form of cobalt is always present on
the surface during the ESR reaction. As expected, the
concentration of Co(II) species was higher while the samples
were exposed to the mixtures with higher than stoichiometric,
reactants molar ratio. Oxidation of cobalt in
the excess of water has been reported
a) Ce 3d
b) Co 2p3/2
15 min ESR
c) O 1s
O4
d) C 1s, Ce 4s & K 2p
K1
Ce(III) 36%
13%
previously for Co/Al2O3 catalysts tested under
Co(II) 34%
H2O/EtOH = 6/1 mol/mol [5], as well as for
2
eOO3
CO23-
791
786
781
776
the Co/CeO2-ZrO2 with the H2O/EtOH molar
ratio of 4/1 [36]. The higher excess of water
allows for higher concentration of surface
adsorbed –OH groups (O2) (as compared to
the results obtained in the stoichiometric ratio
of reactants), facilitating carbon removal.
C
O2
/
O1
o
C
K
-
S
Co(II) 34%
Co(II)
13%
Ce(III) 12%
H
Co(0)
2
CO2-
3 CHx
O2
O
O1
O1
e
O3
C
/
o
C
-
S
H
Ce(III) 15%
Co(II) 34%
3%
1.4. The ESR catalytic performance of the
catalysts
2
CO23-
O
O2
e
C
/
o
O3
C
-
S
L
920 910 900 890 880 791
786
781
776 535 533 531 529 527
299 294 289 284
Apart from the LS-catalyst at 3/1 molar ratio,
the complete conversion of ethanol is
achieved in all cases, after 1 h at 420ºC in a
fixed-bed reactor. Water conversion at 3/1
molar ratio is between 37% and 54%, which
is significantly lower than this expected from
the ESR stoichiometry (100%). This indicates
Binding energy (eV)
Figure 4. The high resolution XPS spectra of Ce 3d (a), Co 2p3/2 (b), O 1s (c) and C 1s, Ce 4s & K 2p
(d) collected after the ESR (H2O/EtOH = 12/1 mol/mol, T = 420°C) over the pre-reduced (H2,
T = 420°C) LS-Co/CeO2, HS-Co/CeO2 and HS-KCo/CeO2 catalysts; a) Ce(IV) (\\\), Ce(III) (///), b)
Co(II) (\\\), Co(0) (///), c) O1 – lattice oxygen, O2 – hydroxyl and carbonate species, O3 – water, O4 –
carbonate species bounded to potassium, d) K1 – potassium carbonate. On the picture (d) also the
percentage atomic contribution of carbon on the surface was given.
Our experiments show that for the HS-K catalyst more
that non-selective ethanol transformations (without the
participation of water molecules) occurred under the examined
conditions. In the case of the ESR carried out in water excess
(9/1 and 12/1 mol/mol), the water conversion over the LS- and
HS-catalysts is close to the expected reaction stoichiometry
(33% and 25%).
potassium species were observed in water rich reactants
mixtures, suggesting segregation of potassium due to its
stronger affinity to water. These experiments also allow us to
conclude that the oxidation state of cobalt is the same like for
the unpromoted catalysts (34% of Co(II)). Ceria valence state for
the LS- and HS-catalysts were similar to this recorded for the 9/1
molar ratio, while deeper reduction of cerium oxide was found
for the potassium-promoted catalyst (36% of Ce(III)).
Table 3 shows the reactants conversions and the ESR
reaction products selectivity obtained after 1 h at 420°C under all
three H2O/EtOH molar ratios. They are also close to those,
measured in a high pressure flow reactor, used in the UHV
system. Under all H2O/EtOH molar ratios the ESR reaction
products include H2, CO2, CO, CH4 and some amounts of
acetaldehyde observed only on the surface of LS-catalyst. Over
all catalysts samples the selectivity to hydrogen was quite high
and in general it increased with the increase of water excess,
The common feature for all catalysts was decreasing carbon
deposition in water-rich reactants mixtures. It is worth to note
that at the 12/1 H2O/EtOH molar ratio, the promoted HS-K
catalyst shows higher carbon atomic fraction (Table 2) as
compared to the unpromoted catalysts. It can be explained by
the lower surface potential barrier of potassium-covered
surfaces [104] which increases the adsorption energy of carbon
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
varying between 82% (3.16 molH /molEt H and 98% (5.56
1.5. The influence of the ESR conditions on the catalysts’ coking
2
molH /molEt H).
2
Presence of big cobalt particles in the
HS-catalyst is expected to make it
Table 3 The selectivity of the ESR towards gaseous products, and yield of hydrogen and carbon dioxide after
1 h under the ESR at 420°C
more vulnerable to coking than its LS-
counterpart [117, 118]. The XPS
spectra obtained at the 3/1 molar ratio
are in agreement with this hypothesis,
but in the excess of water, less carbon
was observed on LS-catalyst after 1 h
ESR (Fig. 2-4 and Table 2). The
different kinetics of carbon deposition
might account for this behaviour in this
two cases. More information about the
kinetics of carbon deposition under
long ESR reaction periods is provided
Catalyst
H2O/EtOH
(mol/mol)
Selectivity (%)
H2/EtOH
(mol/mol)
CO2/EtOH
(mol/mol)
H2
CO2
CO
CH4
MeCHO
LS-Co/CeO2
HS-Co/CeO2
HS-KCo/CeO2
LS-Co/CeO2
HS-Co/CeO2
HS-KCo/CeO2
LS-Co/CeO2
HS-Co/CeO2
HS-KCo/CeO2
82.5
84.6
98.5
94.1
95.1
98.4
96.5
97.1
97.7
52.6
67.2
91.7
80.4
86.5
90.7
88.3
91.6
92.7
19.7
10.6
5.5
10
15
22.2
2.7
9.6
7.8
2.7
5.4
4.5
3.6
6.3
0
3.16
4.03
5.55
4.83
5.19
5.55
5.30
5.49
5.56
1.05
1.34
1.85
1.61
1.73
1.85
1.77
1.83
1.85
3/1
9/1
0
0
5.7
6.6
6.3
3.9
3.7
0
0
0
12/1
0
0
On the contrary, the selectivities of carbon-containing
gaseous products show striking difference among the catalysts.
Comparison of the LS- and HS-catalysts shows that the HS is
more selective towards CO2 (and less to CO) than the
LS-counterpart for all H2O/EtOH molar ratios. At the 3/1 molar
molar ratio both LS- and HS-catalysts have lower CO2 and
higher CO and CH4 selectivities than the HS-K, though more
CH4 is observed for the HS- as compared to LS-catalyst. In the
excess of water, the increase of CO2 selectivity over the LS- and
HS-catalysts is accompanied with a decrease of CO and CH4
formation. The results also show that in the case of the HS-K the
influence of the H2O/EtOH molar ratio on the catalyst’s
selectivity in the ESR is negligible.
by gravimetric studies (Fig. 5). It’s important to keep in mind the
substantial differences in the catalytic reactors, allowing only
certain trends to be taken into account. Nevertheless, based on
the XPS results the oxidation state of the catalysts was not
dramatically different, therefore, the weight changes can be
primarily assigned to coke deposition.
In general, the weight change of the unpromoted catalysts
(LS and HS) is significantly higher as compared to the HS-K
catalyst, confirming that potassium inhibits carbon formation and
catalyst’s oxidation, in accordance with the XPS results.
Under the stoichiometric (3/1 mol/mol) reaction mixture the
weight of the LS-catalyst has increased sharply to reach an
almost constant value, while in the case of the HS-catalyst the
weight gain was slower.
Figure 5. The gravimetric results for the LS-Co/CeO2 (···), HS-Co/CeO2 (---) and HS-KCo/CeO2 (− ∙ −) catalysts under the H2O/EtOH = 3/1, 9/1 and 12/1
mol/mol.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
but after that increases rapidly and finally the catalyst is the most
intensively coked. Conversely, catalyst’s oxidation and carbon
deposition on the HS-catalyst starts immediately and occurs
gradually but with lower rate as compared to the LS-catalyst.
The explanation of the differences at the carbon deposition
kinetics between the catalysts is out of the scope of this paper.
However, from Fig. 5 (insets) it is evident that during the reaction
period used at XPS experiments (1 hour) carbon deposition for
the LS-catalyst at water rich ESR mixtures is limited, in
agreement with our the XPS studies.
The morphology of the spent catalysts after 1 h ESR
reaction under the 3/1 molar ratio was studied by TEM, with
emphasis on the carbon deposit formed during the ESR (Fig.
6-8). The TEM analysis of spent LS- and HS-samples revealed
mainly graphitic carbon, although amorphous carbon was also
detected, but to a lower extent. However, the morphology of
carbon deposits differs among the catalysts. On the LS (Fig. 6)
mostly carbon filaments in which encapsulated cobalt particles
could be seen. Carbonaceous deposit detected on the surface of
HS-catalyst looks more like a few atomic layers of carbon,
mostly located at the boundary of cobalt and ceria crystallites,
indicating that reaction occurs in this direction at the Co–CeOx
interface (Fig. 7). On the TEM images of the HS-K negligible
amounts of carbon were found (Fig. 8), in accordance with the
XPS results.
Figure 6. TEM images of the LS-Co/CeO2 catalyst after 1 h ESR
(T = 420°C, H2O/EtOH = 3/1 mol/mol).
It can be concluded that the form of the carbon deposit
depends on the support morphology and cobalt particles size.
The catalyst with larger cobalt particles favours the growth of
carbon nanofibers [118]. Basing on the XPS results, the surface
amount of coke on both LS- and HS-catalysts is almost
comparable after 1 h ESR for the H2O/EtOH = 3/1 mol/mol. TEM
studies have shown that significantly less bulk carbon was found
Figure 7. TEM images of the HS-Co/CeO2 catalyst after 1 h ESR
(T = 420°C, H2O/EtOH = 3/1 mol/mol).
However, note that the weight gain for both catalysts starts
immediately after exposure to the ESR reactants (see insets of
Fig. 5). On the other hand, at reactants mixtures rich in water the
weight gain does not start simultaneously for the LS- and
HS-catalysts. In particular carbon deposition (and oxidation) on
the LS-catalyst necessitates an induction period of 1 to 2 hours
Figure 8. TEM images of the HS-KCo/CeO2 catalyst after 1 h ESR
(T = 420°C, H2O/EtOH = 3/1 mol/mol).
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
in the case of the HS-catalyst, being compatible with results of
gravimetric measurements (Fig. 5). TEM images prove that in
the case of the low surface area catalyst, carbon diffused inside
cobalt crystallites, destroying the sample from the inner layers
and it is deposited mainly in the whiskers form, while on the high
surface counterpart, carbon was deposited gradually only the
outer layers of catalysts.
differentiate location of hydroxyls, on cobalt or on ceria).
Additionally so far published papers suggest that at
temperatures of 350–450ºC the formation of H species, due to
dissociation of water, is favourable on the surface of both ceria
and cobalt [8, 27, 56, 57, 109, 119]. In spite of very similar, low
concentration of hydroxyls under the 3/1 H2O/EtOH molar ratio
on the surface of potassium-free catalysts (Fig. 9), the
increasing water extent increases the difference between
abundance of these species on the surface of the unpromoted
catalysts. Fig. 9 shows that higher concentration of hydroxyls
exists on the surface of the LS-catalyst. The higher abundance
of hydroxyls might be related to lower oxidation of both main
components of the LS-catalyst – the higher share of metallic, not
oxidized cobalt in the active phase or/and the higher
concentration of surface Ce(III) ions in the ceria support – the
larger surface concentration of hydroxyls.
Discussion
2.1. The influence of the H2O/EtOH molar ratio on the catalysts’
surface
Fig. 9 shows how the reactants composition influences the
catalysts’ surface state. The degree of catalysts’ surface
reorganization under conditions of the steam reforming of
ethanol can be attributed to difference in the H2O/EtOH molar
ratio. Our results prove that higher dispersion (smaller
crystallites) of both ceria support and cobalt in the unpromoted
catalysts makes its oxidation easier under ESR conditions – this
catalyst was also more resistant to reduction, as
In contrast to the unpromoted HS-catalyst, the ceria support
in the potassium-promoted HS-K catalyst is already more
reduced under the lowest, stoichiometric reactants ratio and it
shows further reduction of cerium surface ions with the increase
it is shown in Fig. 1. The ceria support in the
unpromoted catalysts is progressively oxidized
(Fig. 9a) with increasing excess of water. Ceria
with the higher surface area (smaller ceria
crystallites) is oxidized to a larger extent (on its
surface the concentration of Ce(III) ions is lower)
than that of the low-surface area. Moreover, the
water excess facilitates oxidation of cobalt (Fig.
9b). As in the case of ceria, the more intensive
oxidation
of
cobalt
(the
larger
Co(II)/(Co(0)+Co(II)) atomic ratio) occurs when
the small cobalt crystallites were deposited on the
high-surface area ceria support. The result is
consistent with literature conclusions that
supported small cobalt crystallites tend to get
oxidized more readily than the larger ones [8, 36,
40, 43, 53].
The increased H2O/EtOH molar ratio in the
reaction mixture also promotes increasing
hydroxyl’s species concentration (Fig. 9c) on the
catalyst’s surface (the hydroxyls’ concentration is
considered on the whole surface of catalysts
basis, as the XPS data do not enable to
Figure 9. Dependence of percentage concentration of (a) Ce(III)/(Ce(III)+Ce(IV)), (b)
Co(II)/(Co(II)+Co(0)) and (c) adsorbed hydroxyls with the H2O/EtOH molar ratio for the
LS-Co/CeO2, HS-Co/CeO2 and HS-KCo/CeO2 catalysts in the ESR at 420°C.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
of the water excess in the reaction mixture (Fig. 9a). At the same
time, the cobalt active phase also stays in less oxidized form as
compared to the unpromoted HS-catalyst (Fig. 9b) – the extent
of its oxidation is similar to one of the LS-catalyst. The
concentration of surface hydroxyls is much higher already under
the 3/1 H2O/EtOH molar ratio than this on the unpromoted
HS-catalyst and it changes only a little with the H2O/EtOH molar
ratio (Fig. 9c). Under higher water vapour excess this difference
disappears. Undoubtedly, all these differences in the state of the
catalyst’s surface, seen between unpromoted HS- and promoted
HS-K catalysts, result from the additional presence of potassium
on the catalyst’s surface, which concentration of which also
changes with the reactants molar ratio (Table 2). The
progressive increase of the total concentration of potassium on
the catalyst’s surface simultaneously with the excess of water
vapour in the reaction mixture results probably from the re-
dispersion of potassium-containing aggregates, caused by
increased amount of chemisorbed and dissociated water vapour.
As in the case of hydroxyls, the concentration of
potassium-containing sites is considered here on the whole
surface of catalysts basis, without distinction between their
location, on cobalt or on ceria. It was shown in ref. [31] that
potassium on a very similar and free of coke catalyst is well
distributed on the whole catalyst’s surface.
that oxidation-reduction of these main catalysts’ components is
not a general factor, crucial for the selective transformations of
ethanol. The presence of oxidized cobalt and reduction-oxidation
of the ceria support are unavoidable effects, dependent on the
H2O/EtOH molar ratio; however, not involved directly in making
the catalyst’s selective for the ESR. However, it was shown in
ref. [33] that they might be important in the first step of the ESR,
i.e., in oxidative dehydrogenation of alcohol and ethoxide
species formation.
An important feature of the catalysts’ surface, influencing
selectivity of the ESR would be the concentration of hydroxyl
species. In Fig. 10c one can see straight-linear relationships of
concentration of surface adsorbed hydroxyls over LS-, HS- and
HS-K catalysts with the selectivity of carbon dioxide formation.
The low-dispersed LS-catalyst requires larger concentration of
surface adsorbed hydroxyl species than the very well-dispersed
HS to achieve the same selectivity of the ESR – this difference
we will discuss in next paragraphs. However, before the CO2
selectivity over the HS-catalyst gains on that characteristic of the
potassium-promoted HS-K-catalyst (i.e., in the lower range of
OH concentration), the same abundance of hydroxyl species
leads to different selectivity over those two catalysts. Similar
picture is seen in relationships of the OH concentration and
selectivities towards remaining ESR products, i.e., hydrogen,
carbon monoxide and methane (see Fig. S5 in Supplementary
Information). It suggests the existence of additional selective
sites on the surface of promoted catalyst, related to the
presence of potassium promoter on the catalyst’s surface, i.e.,
K–O sites. These potassium-containing sites (K–O) are
expected to play a promoting role in the steam reforming of
ethanol, providing an additional oxygen-containing species
reservoir. Similar suggestion that potassium encourages
formation of active oxygen species, which are very important
intermediates for soot combustion, was given by M. Sun et al.
[94]. Taking both, OH species and K–O sites, into consideration
the selectivities of the ESR to all products over HS- and HS-K
catalysts form coherent dependences, with limits characteristic
for the HS- catalyst (Figs. 11a-11c). Those limits are consistent
with the thermodynamic limit of the hydrogen yield in the ethanol
steam reforming; considering equilibrium states of side reactions
the equilibrium hydrogen yield of 5.56 mole per one mole of
ethanol in the feed are obtainable as against the stoichiometric
value of 6.0 [67]. Also, for the same reason the thermodynamic
yield of carbon dioxide is lower than the stoichiometric value of
2.2. Correlation of the ESR catalytic performance with the
catalysts’ surface state
Considering the influence of the chemical state of the surface of
different ceria-supported cobalt catalysts on the course of the
steam reforming of ethanol, the correlations of the selectivity
towards carbon dioxide formation with the oxidation state of
cobalt active phase and its support were drawn (Figs. 10a and
10b). Taking into account the unpromoted catalysts only, one
might suppose that cobalt oxidation state can be an important
factor influencing the ESR selectivity (Figs. 10a). However, the
performance of the potassium-promoted HS-K catalyst prevents
drawing such general conclusion; the same oxidation degree of
cobalt in the HS and HS-K catalysts enables obtaining very
different carbon dioxide production. It proves that potassium
introduces other selective and active sites responsible for the
ESR – oxidation of cobalt has no response in carbon dioxide
formation. The same conclusion can be drawn when considering
the role of the oxidation state of the ceria support. Both
correlations shown in Figs. 10a and 10b lead us to the assertion
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
2.0. The yields of produced hydrogen and carbon dioxide in our
studies reach the thermodynamic values (Fig. 11e). In the feed
with the H2O/EtOH molar ratio higher than stoichiometric value
of 3/1 the total concentration of OH species and K–O sites on
the surface of potassium-promoted catalyst is even larger than
necessary for the selective course of the ESR to gaseous
products. However, a higher concentration of OH species and
K–O sites is required for hindering of coke deposition on the
catalysts, including all coke forms and all catalysts studied in this
work (Fig. 11f). Small concentration of OH species and K–O
sites leads to dramatic coking of the catalysts, with formation of
completely (or almost completely) dehydrogenated C=C type
carbonaceous deposit, in a large part as graphitic whiskers and
layers (see Table S1 in Supplementary Information and Figs.
6-8). An increasing of those selective-acting surface sites first of
all significantly lowers the amount of coke deposited and also
changes its type from more dehydrogenated to CHx type. For
total elimination of coking phenomena, the high abundance of
OH species and K–O sites, equal to or large than 25 at.% of the
whole surface of catalyst is required.
Among surface oxygen-containing species molecular water is
also present. However, physically adsorbed water molecules
cannot form transition activated complexes, required in
a
catalytic reaction, on the surface of heterogeneous catalyst.
Therefore, water species can be considered merely as
spectators, excluded from direct participation in the ESR over
catalyst’s surface. The selectivity of the steam reforming of
ethanol does not depend solely on the population of OH species
on the catalyst’s surface since the LS-catalyst is less selective
towards all products despite the fact that it has more hydroxyls
adsorbed than the HS-catalyst (Figs. 11a–11d). It suggests that
the catalyst’s morphology and location where oxygen-containing
species are adsorbed may be equally important as their
abundance. The difference in the ESR selectivity is related to
the dispersion of both catalysts. The size of cobalt crystallites is
very different: 3.8 nm in more selective HS-catalyst and 39.3 nm
in the less selective LS. Whereas the size of ceria support
crystallites is 25.6 nm and 87.1 nm, for HS and LS respectively.
The well dispersed, small crystallites usually have more
low-coordinated sites (corners, edges and another defects on
the surface of crystal lattice like steps, adatoms),
which are the best and the strongest sites for
chemisorption and dissociation of various chemical
molecules, including also water (and ethanol)
molecules. On the other hand, a close contact of
both catalyst’s components and a short-enough
distance from their border to the middle of the
cobalt crystallites surface (where the number of
low-coordinated sites is lower than near crystallite
edge) [1] also may have a significant selective
influence on the effects of the ESR [8] – reagents
and their intermediates chemisorbed on the cobalt
and on the ceria support may interact together to
form desirable products of the ESR, i.e., hydrogen
and carbon dioxide. When the surface distance
from the cobalt-ceria support border is too long (as
in the case of a center of large crystallites in the
LS-catalyst), nonselective ethanol transformations
may take place, without the possibility of
by-products to react with activated water. Due to an
insufficient surface concentration of hydroxyl
species acetaldehyde remains among final products
Figure 10. Correlation of CO2 selectivity for the LS-Co/CeO2, HS-Co/CeO2 and HS-KCo/CeO2
catalysts under the ESR at 420°C with percentage concentration of (a) Ce(III)/(Ce(III)+Ce(IV)),
(b) Co(II)/(Co(II)+Co(0)), and (c) adsorbed hydroxyls on the surface. Each point represents the
different H2O/EtOH molar ratio.
(Fig. 11d). The comparison of the ESR effects over
HS- and LS-catalysts leads us to the suggestion
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
We found that the reduction degree of cobalt
oxide and ceria during activation of catalysts with
hydrogen depends on ceria’s support dispersion.
Cobalt oxide supported on low-dispersed ceria is
more vulnerable to reduction than its counterpart
with high-dispersed ceria, while potassium
promoter facilitates reduction of cobalt oxide. The
ceria support is reduced to some extent only in
the low-dispersed catalyst, while in the
unpromoted and potassium-promoted catalysts
with high-dispersed ceria it remains practically
fully oxidized after pre-reduction.
The
catalysts’
surface
undergoes
reorganization under conditions of the steam
reforming of ethanol to a degree dependent on
the H2O/EtOH molar ratio. We proved that higher
dispersion (smaller crystallites) of both ceria
support and cobalt in the unpromoted catalyst
makes its oxidation easier under ESR conditions.
The water excess in the ethanol-water vapours
feed facilitates oxidation of both catalyst’s
components. In contrast to the unpromoted
high-dispersed catalyst, the ceria support in the
potassium-promoted catalyst is more reduced
even under the lowest, stoichiometric reactants
ratio and it undergoes further reduction with the
increase of the water excess in the reaction
mixture. In the potassium-promoted catalyst the
cobalt active phase exists mainly in metallic form,
similarly to the unpromoted catalysts.
Figure 11. Correlation of (a-d) the catalysts’ selectivity towards H2 and carbon-containing
product, (e) H2 and CO2 yield, and (f) the concentration of C=C and CHx species on the whole
surface, with the concentration of oxygen-containing species on the surface of the LS-Co/CeO2,
HS-Co/CeO2 and HS-KCo/CeO2 catalysts in the ESR at 420°C.
We proved that oxidation-reduction of the
main catalysts’ components is not a general factor,
that the selective ceria-supported cobalt catalyst should have
crucial for the selective transformations of ethanol. The
presence of oxidized cobalt and reduction-oxidation of the ceria
support are unavoidable effects, dependent on the H2O/EtOH
molar ratio; however, not involved directly in doing the catalyst
selective for the ESR.
well dispersed cobalt particles deposited on the well dispersed
support, which is in agreement with previously published results
[3, 13, 21].
The concentration of potassium promoting sites (Kδ+–Oδ-
sites) on the catalyst’s surface progressively increases
simultaneously with the excess of water vapour in the reaction
mixture. These sites may play promoting role in the steam
reforming of ethanol, providing an additional oxygen-containing
species reservoir.
Conclusions
The chemical surface state and the catalytic performance of
cobalt catalysts with micro- and nano-dispersed ceria supports
were investigated under the steam reforming of ethanol at 420°C
for the H2O/EtOH molar ratios equal to 3/1, 9/1 and 12/1. In the
case of the nano-catalyst the effect of potassium promotion was
also examined.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
The most crucial for selective conversion of ethanol over
unpromoted catalysts is concentration of hydroxyl species on the
catalysts surface, which increases with increased H2O/EtOH
molar ratio in the reaction mixture in the extent depending on the
dispersion of catalysts components. In the case of promotion of
the catalyst with potassium, the increasing hydroxyl
concentration is supplemented by Kδ+–Oδ- sites. We found good,
was dried at 110°C for 12 h, then calcined at 400°C with
a heating rate of 2°C/min up to the calcination set point and
maintained for 1 h at this temperature.
Catalysts characterization
X-ray fluorescence technique (XRF) was used to determine the
bulk cobalt content in the catalyst. The measurements were
performed by AxiosmAX (PANalytical) spectrometer fitted with a
Rh (4 kW) tube. The SuperQ 5 software package was used for
spectral deconvolution and for the calculation of the component
contents. The quantification was based on the external synthetic
standard.
coherent
relationship
of
OH
species
along
with
Kδ+–Oδ- sites, and the selectivity of the ESR to all gaseous
products over unpromoted and promoted catalysts, with limits
consistent with the thermodynamic limit of the hydrogen yield.
Beside surface concentration of hydroxyl species, the catalyst’s
morphology and location where oxygen-containing species are
chemisorbed may be equally important to their abundance – the
selective ceria-supported cobalt catalyst should have well
dispersed cobalt particles deposited on the well dispersed
support.
The X-ray powder diffraction patterns of the supports and
catalysts were recorded using an Empyrean X-ray (PANalytical)
diffractometer equipped with a PIXcel3D detector using Cu Kα
radiation (λ = 0.154 nm). Data were acquired from 10 to
110º(2θ) with a 0.026º(2θ) step size. The measured patterns
were compared with the International Centre for Diffraction Data
(ICDD) PDF-4+ database for phase identification and the
crystallite size was calculated using the Scherrer equation. A
correction for instrumental broadening was applied using
a reference standard (LaB6).
Beside the role played by oxygen-containing species in
improving catalytic performance of catalysts, the increasing
presence of both OH and Kδ+–Oδ- sites was found to lower the
amount of coke deposited as well as to change its type from fully
dehydrogenated C=C to CHx type. For total elimination of coking
phenomena higher abundance of OH species and Kδ+–Oδ- sites
on the catalyst’s surface than that necessary for the selective
ESR to gaseous products is required.
The porosity and BET total surface area of catalysts were
measured by the low temperature (-196°C) nitrogen adsorption
in the ASAP 2420 (Micromeritics) analyzer after degassing the
samples at 200°C. The average pore diameter and the volume
of pores were calculated from desorption data using
Barret-Joyner-Halenda (BJH) method.
Experimental Section
Catalysts preparation
The active metal surface area, dispersion and mean cobalt
particles size of catalysts pre-reduced in situ in hydrogen at
420°C (1 h), were calculated from hydrogen chemisorption data
obtained at an ASAP 2020 (Micromeritics) analyzer at 130°C
[120]. The H/Co stoichiometry of 1/1 and the spherical shape of
cobalt particles were used for calculation of the mean size. The
detailed description of the calculation method has been given
elsewhere [120].
The Co/CeO2 catalysts were prepared by impregnation of the
commercial nano- (HS) and micro- (LS) dispersed ceria support
(<25 nm and <5 μm, respectively, Aldrich) [13]. Prior to
impregnation, the supports were dried at 110°C for 3 h. The
solution of cobalt nitrate and citric acid CA (the relative molar
concentration of Co and CA was 1/1) was used for impregnation.
After the impregnation, the catalysts’ precursor were dried at
110°C for 12 h, then calcined at 400°C with a heating rate of
2°C/min up to the calcination set point and maintained for 1 h at
this temperature [60, 73]. The catalysts were denoted as LS-
and HS-Co/CeO2. Next, the part of obtained HS-Co/CeO2
catalyst’s precursor was impregnated with potassium nitrate
solution in order to introduce 2 wt.% of potassium promoter to
the catalyst [31, 60] (further abbreviated as HS-K) and again it
The gravimetric studies of catalysts’ coking in the ESR were
carried out in TG121 microbalance system (CAHN), under
dynamic conditions in a quartz reactor with a continuous flow of
H2O/EtOH vapours diluted with He (He 50 cm3/min + 3 cm3/min
mixture) at 420°C for 21 hours. The molar ratio of H2O/EtOH
was equal 3/1, 9/1 and 12/1. Prior to the reaction, catalysts
(0.05 g of the catalyst (0.15–0.3 mm)) were reduced by passing
10% H2/He flow (70 cm3/min) at the temperature of 420°C for 1
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
hour (the linear temperature increase to 420°C was 10°C/min).
In all studies 96% bio-EtOH was used for the reaction mixture
preparation.
ppartial = 57 mbar) diluted in Ar (50 cm3/min) were introduced to
the reactor (by heated lines) at the reaction temperature (420°C)
by a Controlled Evaporation and Mixing (CEM) System
(Bronkhorst). The total pressure in the reactor during the
reaction was equal 1 atm. The ESR was carried out over 1 h,
next the flow of vapours was stopped and samples were
quenched in the flow of Ar till 80°C and transferred to the XPS
analysis chamber.
Combined XPS and catalytic performance experiments
X-ray Photoelectron Spectroscopy (XPS) studies were
performed in a multi-chamber UHV system (PREVAC, Poland)
equipped with a monochromatized Al source (XM 650 X Ray
Monochromator) source operating at 360 W and a hemispherical
electron analyzer (Scienta R4000). The pass energy of the
analyser was set at 200 eV (energy step 0.5 eV) for survey scan
and 50 eV (energy step 0.1 eV) for high resolution Co 2p, Ce 3d,
O 1s, C 1s, Ce 4s and K 2p spectra. The base pressure in the
analysis chamber was 5∙10-8 mbar. Data processing was
performed with the CasaXPS software (v 2.3.16 PR 1.6), taking
into account the relative sensitivity factors (provided by
CasaXPS software) for Co 2p3/2, Ce 3d, O 1s, C 1s, K 2p
regions. After Shirley background subtraction the Co 2p and
Ce 3d regions were fitted using spectra recorded over reference
samples. The quantification of the Co 2p and Ce 3d regions
was performed by the fitting procedure based on the linear
combination of two reference spectra. The percentage
contribution of Ce(III) and Co(II) were calculated using the peak
areas emerged by the linear fitting procedure of references
Ce(III)
In the course of the ESR reaction, carried out in the high
pressure cell, the composition of gas phase products (H2, CO2,
CO, CH4) was monitored on-line by means of a micro-GC
(Agilent, 490-GC).
Catalytic tests in a fixed-bed reactor
Since the micro-GC does not allow to analyse the liquids such
as: ethanol, water, and acetaldehyde, in order to gain a better
insight into catalytic properties of the examined catalysts, the
ESR reaction (H2O/EtOH molar rations 3/1, 9/1 and 12/1) was
carried out also in a fixed-bed continuous-flow quartz reactor
(Microactivity Reference unit, PID Eng
& Tech.) under
atmospheric pressure. The pre-treatment and reaction
conditions were kept identical to those described above for
reactor attached to the UHV setup (the same amount of pressed
in pellet catalyst, the same reduction procedure and the flow of
reactants). The analysis of the reaction mixture and the reaction
products (all in the gas phase) were carried out on-line by
means of two gas chromatographs. One of them, Bruker
450-GC was equipped with two columns, the first filled with a
porous polymer Porapak Q (for all organics, CO2 and H2O
vapour) and the other one – capillary column CP-Molsieve 5Å
(for CH4 and CO analysis). Helium was used as a carrier gas
and a TCD detector was employed. The hydrogen concentration
was analyzed by the second gas chromatograph, Bruker
430-GC, using a Molsieve 5Å column, argon as a carrier gas
and a TCD detector.
according to the formulas Ce(III)
100% and
Ce(III) Ce(IV)
Co(II)
Co(II)
100%.
Co(II) Co(0)
Catalytic tests were performed in a high pressure flow
reactor connected through the radial distribution chamber with
the analysis chamber. The samples were pressed into 10 mm
diameter pellets and attached on a dedicated sample holder
(PTS HPC RES/C-K RG). The temperature of the sample was
measured with a thermocouple in contact with the sample holder
and regulated by HEAT2-PS power supply. Prior to the reduction
samples were heated in a load lock chamber by halogen lamp
for 1.5 h to remove volatile surface adsorbed impurities. The
reduction was carried out in the flow reactor at 420°C for 1 h, by
hydrogen diluted in argon (50/50 cm3/min, p = 1 atm). The
samples were cooled down to 200°C in the flow of H2/Ar and
then the flow of hydrogen was stopped. Further cooling down (to
80°C) was in the flow of Ar only. Before the reaction, samples
were heated in the flow of Ar (50 cm3/min, p = 1 atm) with the
ramp rate 10°C/min. The water/ethanol vapours (3 cm3/min,
The total conversion of ethanol (
X
), water (
) and
X
H2O
EtOH
conversions of ethanol into individual carbon-containing
products ( ) were calculated from their concentrations
X
CP
before and after the reaction, taking into account changes in the
gas volume during the reaction, from the equations:
in
EtOH
out
EtOH
C
 C
K
X

100%
EtOH
in
EtOH
C
,
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
microscope (Tecnai G2 20 X-TWIN FEI Company), equipped
with LaB6 emitter and using 200 kV electron beam accelerating
voltage [35].
in
H2O
out
H2O
C
 C
K
X

100%
H2O
in
C
H2O
,
out
CP
C
K
X

100%
CP
in
n/2C
EtOH
Acknowledgements
where:
S.T. would like to thank for the financial support from French
Government, Campus France, UdS (France) and UMCS
(Poland). The research was carried out with the equipment
purchased thanks to the financial support of the European
Regional Development Fund in the framework of the Polish
Innovation Economy Operational Program (contract no.
POIG.02.01.00-06-024/09 Center of Functional Nanomaterials;
www.cnf.umcs.lublin.pl).
in
H2O
in
C
C
and
– is the molar concentration of EtOH and H2O
EtOH
out
out
EtOH
C
and
C
– is the molar
in the reaction mixture (mol%);
H2O
concentration of EtOH and H2O in the post-reaction mixture
out
(mol%);
C
– is the molar concentration of carbon-containing
CP
products in the post-reaction mixture (mol%); n – is number of
carbon atoms in carbon-containing molecule of the reaction
in
C
out
C
product; K – is the volume contraction factor (K =
C
/
C
Keywords: ethanol steam reforming • cobalt-based catalysts •
ceria support • potassium promoter • XPS
in
C
out
C
where
C
and
C
are the molar concentrations of carbon in
[1]
[2]
D. Zanchet, J.B.O. Santos, S. Damyanova, J.M.R. Gallo, J.M.C. Bueno,
ASC Catal. 2015, 5, 3841–3863.
the ethanol feed to the reaction and in all carbon-containing
compounds which were present in post reaction gases,
respectively). The selectivity of ethanol conversion into individual
carbon-containing gaseous and liquid products was expressed
B. Bayram, I.I. Soykal, D. von Deak, J.T. Miller, U.S. Ozkan, J. Catal.
2011, 284, 77–89.
[3]
[4]
I.I. Soykal, H. Song, U.S. Ozkan, ACS Catal. 2012, 2, 2335–2348.
A.M. Karim, Y. Su, M.H. Engelhard, D.L. King, Y. Wang, ACS Catal.
2011, 1, 279–286.
as

X
/X

100%
.
CP EtOH
The selectivity of hydrogen formation was determined as:
[5]
[6]
[7]
[8]
[9]
A.R. Passos, L. Martins, S.H. Pulcinellin, C.V. Santilli, V. Briois, Catal.
Today 2014, 229, 88–94.
out
C
H2
H
selectiv yit 
100%
2
M.N. Barroso, M.F. Gomez, L.A. Arru, M.C. Abello, Reac. Kinet. Mech.
Cat. 2015, 115, 535–546.
out
H2
out
out
C
 2C
 2C
CH4 CH3CHO
E. Varga, Z. Ferencz, A. szkó, A. Erdöhelyi, J. Kiss, J. Mol. Catal. A
2015, 397, 127–133.
where:
out
C
– is the molar concentration of the hydrogen-containing
H. Sohn, I.I. Soykal, S. Zhang, J. Shan. F. Tao, J.T. Miller, U.S. Ozkan,
J. Phys. Chem. C 2016, 120, 14631–14642.
products in the post-reaction mixture (mol%). Hydrogen,
methane and acetaldehyde were the only hydrogen-containing
products of the ESR.
H. Song, L. Zhang, U.S. Ozkan, Top. Catal. 2012, 55, 1324–1331.
[10] Zs. Ferencz, A. Erdőhelyi, K. Baan, A. szkó, L. Óvári, Z. Kónya, C.
Papp, H.-P. Steinrück, J. Kiss, ACS Catal. 2014, 4, 1205–1218.
[11] I.I. Soykal, B. Bayram, H. Sohn, P. Gawade, J.T. Miller, U.S. Ozkan,
Appl. Catal., A 2012, 449, 47–58.
Post-reaction TEM catalysts’ characterization
[12] H. Wang, L. Zhang, M. Li, Y. Liu, X. Bai, J. Rare Earth 2013, 31, 565–
571.
The spent catalysts (after 1 h of the ESR) were grinded to a fine
powder in an agate mortar and then mixed with 99.8% ethanol
into the ultrasonic homogenizer for 20 s. The slurry containing
the catalyst was pipetted on a 200 mesh copper grid covered
with lacey formvar and stabilized with carbon (Ted Pella
Company) and left on a filter paper to evaporate ethanol.
Subsequently, the sample deposited on the grid was inserted to
the sample holder and transferred to the electron microscope.
[13] A. Machocki, A. Denis, W. Grzegorczyk, W. Gac, Appl. Surf. Sci. 2010,
256, 5551–5558.
[14] S.-W. Yu, H.-H. Huang, C.-W. Tang, C.-B Wang, Int. J. Hydrogen
Energy 2014, 39, 20700–20711.
[15] H. Song, U.S. Ozkan, J. Catal. 2009, 261, 66–74.
[16] H. Sohn, U.S. Ozkan, Energy Fuel 2016, 30, 5309-5322.
[17] J. Llorca, P.R. de la Piscina, J.A. Dalmon, N. Homs, Chem. Mater. 2004,
16, 3573–3578.
[18] S. Tuti, F. Pepe, Catal. Lett. 2008, 122, 196–203.
[19] M.P. Hyman, J.M. Vohs, Surf. Sci. 2011, 605, 383–389.
The images were recorded in
a
transmission electron
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
[20] N. Laosiripojana, S. Assabumrungrat, Appl. Catal., B 2006, 66, 29–39.
[21] I.I. Soykal, H. Sohn, D. Singh, J.T. Miller, U.S. Ozkan, ASC Catal. 2014,
4, 585–592.
[49] W. Xu, Z. Liu, A.C. Johnston-Peck, S.D. Senanayake, G. Zhou, D.
Stacchiola, E.A. Stach, J.A. Rodriguez, ACS Catal. 2013, 2, 975−984.
[50] H. Song, X. Bao, C.M. Hadad, U.S. Ozkan, Catal Lett. 2011, 141, 43–
54.
[22] M. Greluk, P. Rybak, G. Słowik, M. Rotko, A. Machocki, Catal. Today
2015, 242, 50–59.
[51] H. Song, L. Zhang, R.B. Watson, D. Braden, U.S. Ozkan, Catal. Today
2007, 129, 346–354.
[23] S. Ogo, T. Shimizu, Y. Nakazawa, K. Mukawa, D. Mukai, Y. Sekine,
Appl. Catal., A 2015, 495, 30–38.
[52] S.S.-Y. Lin, H. Daimon, S.Y. Ha, Appl. Catal., A 2009, 366, 252–261.
[53] A.M. de Silva, L.O.O. da Costa, K.R. de Souza, L.V. Mattos, F.B.
Noronha, Catal. Commun. 2010, 11, 736–740.
[23] B. Banach, A. Machocki, Appl. Catal., A 2015, 505, 173–182.
[25] S. Turczyniak, W. Luo, V. Papaefthimiou, N. S. Ramgir, M. Haevecker,
A. Machocki, S. Zafeiratos, Top. Catal. 2016, 59, 532–542.
[26] V.M. Lebarbier, A.M. Karim, M.H. Engelhard, Y. Wu, B.-Q. Xu, E.J.
Petersen, A.K. Datye, Y. Wang, ChemSusChem 2011, 4, 1679–1684.
[27] I.I. Soykal, H. Song, J.T. Miller, U.S. Ozkan, Top Catal. 2014, 57, 785–
795.
[54] A.M. da Silva, K.R. de Souza, G. Jacobs, U.M. Graham, B.H. Davis,
L.V. Mattos, F.B. Noronha, Appl. Catal. B 2011, 102, 94–109.
[55] S.M. de Lima, I.O. da Cruz, G. Jacobs, B.H. Davis, L.V. Mattos, F.B.
Noronha, J. Catal. 2008, 257, 356–368.
[56] S.M. de Lima, A.M. Silva, I.O. da Cruz, G. Jacobs, B.H. Davis, L.V.
Mattos, F.B. Noronha, Catal. Today 2008, 138, 162–168.
[57] S.M. de Lima, A.M. Silva, U. M. Graham, G. Jacobs, B. H. Davis, L.V.
Mattos, F.B. Noronha, Appl. Catal., A 2009, 352, 95–113.
[58] A. Casanovas, M. Roig, C. de Leitenburg, A. Trovarelli, J. Llorca, Int. J.
Hydrogen Energy 2010, 35, 7690–7698.
[28] J. Llorca, J.-A. Dalmon, P.R. de la Piscina, N. Homs, Appl. Catal., A
2003, 243, 261–269.
[29] S. Turczyniak, D. Teschner, A. Machocki, S. Zafeiratos, J. Catal. 2016,
340, 321–330.
[30] J. Sun, A.M. Karim, D. Mei, M. Engelhard, X. Bao, Y. Wang, Appl.
Catal., B 2015, 162, 141–148.
[59] K.-S. Kim, H.-R. Seo, S.Y. Lee, J.-G. Ahn, W.Ch. Shin, Y.-K. Lee, Top
Catal. 2010, 53, 615–620.
[31] G. Słowik, M. Greluk, A. Machocki, Mater. Chem. Phys. 2016, 173,
219–237.
[60] M. Greluk, M. Rotko, A. Machocki, Catal. Lett. 2016, 146, 163–173.
[61] A. Machocki, T. Ioannides, E. Papadopoulou, B. Banach, Fuel
Processing Technol. 2016, 148, 341–349.
[32] M.-R. Li, J. Chen, G.-C. Wang, J. Phys. Chem. C 2016, 120, 14198–
14208.
[33] E. Martono, M.P. Hyman, J.M. Vohs, Phys. Chem. Chem. Phys. 2011,
13, 9880–9886.
[62] M. Greluk, G. Słowik, M. Rotko, A. Machocki, Fuel 2016, 183, 518–530.
[63] E. Varga, K. Baán, G.F. Samu, A. Erdőhelyi, A. szkó, Z. Kónya, J.
Kiss, 2016 doi:10.1007/s10562-016-1809-3.
[34] Y.T. Law, W.H. Doh, W. Luo, S. Zafeiratos, J. Mol. Catal. A 2014, 381,
89–98.
[64] J.Y.Z. Chiou, W.-Y. Wang, S.-Y. Yang, C.-L. Lai, H.-H. Huang, C.-B.
Wang, Catal Lett. 2013, 143, 501–507.
[35] W. Luo, A. Asthagiri, Catal. Sci. Technol. 2014, 4, 3379–3389.
[36] S.S.-Y. Lin, D.H. Kim, M.H. Engelhard, S.Y. Ha, J. Catal. 2010, 273,
229–235.
[65] J.-W.Snoeck, G.F. Froment, M. Fowles, Ind. Eng. Chem. Res. 2002, 41,
3548–3556.
[37] M.P. Hyman, E. Martono, J. Vohs, J. Phys. Chem. C 2010, 114, 16892–
16899.
[66] A.G. Gayubo, J. Vicente, J. Ereña, C. Montero, M. lazar, J. Bilbao, J.
Catal. Lett. 2014, 144, 1134–1143.
[38] L. del Río, G. Marbán, Appl. Catal., B 2014, 150–151, 370–379.
[39] R. Espinal, E. Taboada, E. Molins, R.J. Chimentao, F. Medina, J. Llorca,
Appl. Catal., B 2012, 127, 59–67.
[67] K. Vasudeva, N. Mitra, P. Umasankar, S.C. Dhingra, Int. J. Hydrogen
Energy 1996, 21, 13–18.
[68] V. Mas, R. Kiproes, N. Amadeo, M. Laborde, Int. J. Hydrogen Energy
2006, 31, 21–28.
[40] E. Varga, P. Pusztai, L. Óvári, A. szkó, A. Erdőhely, C. Papp, H.P.
Steinrück, Z. Kónya, J. Kiss, Phys. Chem. Chem. Phys. 2015, 17,
27154–27166.
[69] E.Y. García, M. Laborde, Int. J. Hydrogen Energy 1991, 16, 307–312.
[70] F. Diehl, A.Y. Khodakov, Oil Gas Sci. Technol. 2009, 64, 11–24.
[71] Y.-J. Wang, H. Dong, G.-M. Lyu, H.-Y. Zhang, J. K., L.-Q. Kang, J.-
L.Teng, Li.-D. Sun, R. Si, J. Zhang, Y.-J.Liu, Y.-W. Zhang, Y.-H.Huang,
C.-H. Yan, Nanoscale 2015, 7, 13981–13990.
[41] L. Qiu, F. Liu, L. Zhao, Y. Ma, J. Yao, Appl. Surf. Sci. 2006, 252, 4931–
4935.
[42] S.M.F. Shahed, T. Hasegawa, Y. Sainoo, Y. Watanabe, N. Isomura, A.
Beniya, H. Hirata, T. Komeda, Surf. Sci. 2014, 628, 30–35.
[43] A.M. da Silva, L.V. Mattos, J.P. de Breejen, J.H. Bitter, K.P. de Jong,
F.B. Noronha, Catal. Today 2011, 164, 234–239.
[72] M.C. Biesinger, B.P. Payne, A.P. Grosvenor, L.W.M. Lau, A.R. Gerson,
R.St.C. Smart, Appl. Surf. Sci. 2011, 257, 2717–2730.
[73] P. Rybak, B. Tomaszewska, A. Machocki, W. Grzegorczyk, A. Denis,
Catal. Today 2011, 176, 14–20.
[44] S.M. de Lima, A.M. da Silva, U.M. Graham, G. Jacobs, B.H. Davis,
L.V. Mattos, F.B. Noronha, J. Catal. 2009, 268, 268–281.
[45] T. Mondal, K.K. Pant, A.K. Dalai, Int. J. Hydrogen Energy 2015, 40,
2529–2544.
[74] M. Araque, J.C. Vargas, Y. Zimmermann, A.-C. Roger, Int. J. Hydrogen
Energy 2011, 36, 1491–1502.
[75] G.R. Rao, B.G. Mishra, Bull. Catal. Soc. India 2003, 2, 122–134.
[76] H. Song, B. Mirkelamoglu, U.S. Ozkan, Appl. Catal., A 2010, 382, 58–
64.
[46] T. Montini, M. Melchionna, M. Monai, P. Fornasiero, Chem. Rev. 2016,
116, 5987–6041.
[47] S.S.-Y. Lin, D.H. Kim, S.Y. Ha, Appl. Catal., A 2009, 355, 69–77.
[48] E. Martono, J.M. Vohs, ACS Catal. 2011, 1, 1414–1420.
[77] H. Idriss, Platinum Metals Rev. 2004, 48, 105–115.
[78] X.D. Hou, Y.Z. Wang, Y.X. Zhao, Catal. Lett. 2008, 123, 321–326.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
[79] J. El Fallah, S. Boujana, H. Dexpert, A. Kiennemann, J. Majerus, O.
Touret, F. Villain, F. Le Normand, J. Phys. Chem. 1994, 98, 5522–5533.
[80] M. Romeo, K. Bak, J. El. Fallah, F. Le Normand, L. Hilarie, Surf.
Interface Anal. 1993, 20, 508–512.
[99] J. Iranmahboob, D.O. Hill, H. Toghian, Appl. Surf. Sci. 2001, 185, 72–
78.
[100] B. Lamontagne, F. Semond, D. Roy, J. Electron. Spectrosc. Relat.
Phenom. 1995, 73, 81–88.
[81] E. Paparazzo, G.M. Ingo, J. Electron. Spectr. Relat. Phenom 1998, 87,
301–304.
[101] G. Pirug, . Müller, H. P. Bonzel, Surf. Sci. 1993, 73, 145–154.
[102] M. Carlsson, Johnson Matthey Technol. Rev. 2015, 59, 313–318.
[103] H.P. Bonzel, H.J. Krebs, Surf. Sci. 1981, 109, L527–L531.
[104] H.P. Bonzel, G. Broden, H.J. Krebs, Appl. Surf. Sci. 1983, 16, 373–394.
[105] G. Maniak, P. Stelmachowski, A. Kotarba, Z. Sojka, V. Rico-Pérez, A.
Bueno-López, Appl. Catal., B 2013, 136–137, 302–307.
[106] L. del Río, G. Marbán, Appl. Catal., B 2012, 126, 39–46.
[107] S. Davidson, J. Sun, Y. Wang, Top Catal. 2013, 56, 1651–1659.
[108] D. Teschner, A. Wootsch, O. Pozdnyakova, J. H. Sauer, A. Knop-
Gericke, R. Schlögl, React. Kinet. Catal. Lett. 2006, 87, 235–247.
[109] L. Xu, Y. Ma, Y. Zhang, B. Chen, Z. Wu, J. Jiang, W. Huang, J. Phys.
Chem. C 2010, 114, 17023–17029.
[82] E. Paparazzo, G.M. Ingo, N. Zacchetti, J. Vac. Sci. Techn. A 1991, 9,
1416–1420
[83] E. Paparazzo, Surf. Sci. Lett. 1990, 234, L253–L258.
[84] L. Óvári, S.K. Calderon, Y. Lykhach, J. Libuda, A. Erdőhelyi, C. Papp, J.
Kiss, H.-P. Steinrück, J. Catal. 2013, 307, 132–139.
[85] J. Rynkowski, J. Farbotko, R. Touroude, L. Hilarie, Appl. Catal. A 2000,
203, 335–348.
[86] Z. Liu, T. Duchoň, H. Wang, D.C. Grinter, I. Waluyo, J. Zhou, Q. Liu, B.
Jeong, E.J. Crumlin, V. Matolín, D.J. Stacchiola, J.A. Rodriguez, S.D.
Senanayake Phys. Chem. Chem. Phys. 2016, 18, 16621–16628
[87] M. Konsolakis, M. Sgourakis, S.A.C. Carabineiro, Appl. Surf. Sci. 2015,
341, 45–54.
[110] T.-L. Chen, D.R. Mullins, J. Phys. Chem. C 2011, 115, 3385–3392.
[111] E. de Smit, F.M.F. de Groot, R. Blume, M. Hävecker, A. Knop-Gericke,
B.M. Weckhuysen, Phys. Chem. Chem. Phys. 2010, 12, 667–680.
[112] N. Hou, Y. Zhang, M. Meng, J. Phys. Chem. C 2013, 117, 4089−4097.
[113] Y. Zhang, M. Meng, F. Dai, T. Ding, R. You, J. Phys. Chem. C 2013,
117, 23691−23700.
[88] A. Pfau, K.D. Schierbaum, Surf. Sci. 1994, 321, 71–80.
[89] D.R. Mullins, M.D. Robbins, J. Zhou, Surf. Sci. 2006, 600, 1547–1558.
[90] F. Larachi, J. Pierre, A. Adnot, A. Bernis, Appl. Surf. Sci. 2002, 195,
236–250.
[91] D. Cappus, C. Xu, D. Ehrlich, B. Dillmann, C.A. Ventnce Jr., K. Al
Shamery, H. Kuhlenbeck, H.-J. Freund, Chem. Phys. 1993, 177, 533–
546.
[114] W. Wang, Y.Q. Wang, Int. J. Energy Res. 2008, 32, 1432–1443.
[115] X.-D. Zhou, W. Huebner, Appl. Phys. Lett. 2001, 79, 3512–3514.
[116] X. Wan, D. Goberman, L.L. Shaw, G. Yi, G.-M. Chow, Appl. Phys. Lett.
2010, 96, 123108.
[92] D.A. Creaser, P.G. Harrison, Catal. Lett. 1994, 23, 13–24.
[93] B.E. Koel, G. Praline, H.-I. Lee, J.M. White, J. Electron. Spectrosc.
Relat. Phenom. 1980, 21, 31–46.
[117] A.L.M. da Silva, J.P. den Breejen, L.V. Mattos, J.H. Bitter, K.P. de Jong,
F.B. Noronha, J. Catal. 2014, 318, 67–74.
[94] M. Sun, L. Wang, B. Feng, Z. Zhang, G. Lu, Y. Guo, Catal. Today 2011,
175, 100–105.
[118] S. Helveg, J. Sehested, J.R. Rostrup-Nielsen, Catal. Today 2011, 178,
42–46.
[95] J.L. Ayastuy, A. Gurbani, M.P. González-Marcos, M.A. Gutiérrez-Ortiz,
Int. J. Hydrogen Energy 2012, 37, 1993–2006.
[119] L.V. Mattos, G. Jacobs, B.H. Davis, F.B. Noronha, Chem. Rev. 2012,
112, 4094–4123.
[96] T.L. Westover, A.D. Franklin, B.A. Cola, T.S. Fisher, R.G. Reifenberger,
J. Vac. Sci. Techn., A 2010, 28, 423–434.
[120] G. Słowik, A.Gawryszuk-Rżysko, M. Greluk, A. Machocki, Catal Lett.
2016, 146, 2173–2184.
[97] Z. Hou, O. Yokota, T. Tanaka, T. Yashima, Catal. Lett. 2003, 87, 37–42.
[98] A.V. Shchukarev, D.V. Korolkov, CEJC 2004, 2, 347–362.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201601343
ChemCatChem
FULL PAPER
Entry for the Table of Contents
Layout 1:
FULL PAPER
Abstract
Sylwia Turczyniak, Magdalena Greluk,
Grzegorz Słowik, Wojciech Gac,
Spyridon Zafeiratos, and Andrzej
Machocki
Introduction
Results
1.1. Reduction of the catalysts in
hydrogen
1 – 21
1.2. XPS characterization of catalysts
The surface state and the catalytic
performance of Co/CeO2 catalysts in
the steam reforming of ethanol
after
the
ESR
with
a
stoichiometric H2O/EtOH molar
ratio
1.3. XPS characterization of catalysts
after the ESR at high H2O/EtOH
molar ratios
1.4. The ESR catalytic performance of
the catalysts
1.5. The influence of the ESR
conditions on the catalysts’
coking
Discussion
2.1. The influence of the H2O/EtOH
molar ratio on the catalysts’
surface
2.2. Correlation of the ESR catalytic
performance with the catalysts’
surface state
Conclusions
Experimental Section
Catalysts preparation
Catalysts characterization
Combined XPS and catalytic
performance experiments
Catalytic tests in a fixed-bed reactor
Post-reaction TEM catalysts’
characterization
Acknowledgements
This article is protected by copyright. All rights reserved.