Structural Changes of Binary/Ternary Spinel Oxides During Ethanol Anaerobic Decomposition

Article abstract of DOI:10.1002/cctc.201601605

Several M-modified iron oxides of the spinel family have proven to be effective electron and O^2? vectors for the production of hydrogen in the chemical-loop reforming of bio-alcohols. The present work is specifically focused on investigation of ethanol anaerobic decomposition over spinel oxides, which results in significant structural changes of the oxygen carrier material itself and corresponds to a first step of the chemical-loop reforming process. In particular, a series of binary/ternary M-modified ferrospinels were prepared by a co-precipitation method and tested in terms of both redox properties and intrinsic catalytic activity in addition to a complex ex situ study that encompasses the solid-state chemistry investigations of the fresh and reduced oxygen carrier materials. It was found that Co/Cu incorporation facilitates total/partial oxidation of ethanol, giving rise to high yields of H₂, CO_x, and H₂O; whereas Mn incorporation predominantly favored dehydrogenation and condensation reactions, leading to the formation of acetaldehyde and acetone. In addition, the incorporation of Mn contributed to significantly reduce the amount of coke formed; however, it caused a lower intrinsic reducibility, which was explained by the formation of a thermodynamically stable and hardly reducible layer of Mn_xFe_yO solid solution.

Full text of DOI:10.1002/cctc.201601605

Accepted Article
Title: Structural changes of binary/ternary spinel oxides during ethanol
anaerobic decomposition
Authors: Olena Vozniuk, Cristian Bazzo, Stefania Albonetti, Nathalie
Tanchoux, Françoise Bosselet, Jean-Marc Marie Millet,
Francesco Di Renzo, and Fabrizio Cavani
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.201601605
Link to VoR: http://dx.doi.org/10.1002/cctc.201601605
A Journal of
www.chemcatchem.org

ChemCatChem
10.1002/cctc.201601605
FULL PAPER
Structural changes of binary/ternary spinel oxides during ethanol
anaerobic decomposition
[
a,b]
[a]
[a]
[b]
[c]
Olena Vozniuk , Cristian Bazzo , Stefania Albonetti , Nathalie Tanchoux , Françoise Bosselet ,
[
c]
[b]
[a,d]
Jean-Marc M. Millet , Francesco Di Renzo* and Fabrizio Cavani*
2-
Several M-modified iron oxides of the spinel family have proven to be effective electrons and O vectors for the production of hydrogen via
Chemical-Loop Reforming of bio-alcohols. The present work is specifically focused on investigation of ethanol anaerobic decomposition over
st
spinel oxides that results in significant structural changes of the oxygen carrier material itself and corresponds to a 1 step of the Chemical-
Loop Reforming process. In particular, a series of binary/ternary M-modified ferrospinels were prepared via co-precipitation method and
tested in terms of both redox properties and intrinsic catalytic activity in addition to a complex ex-situ study that encompasses the solid-state
chemistry investigations of the fresh and reduced oxygen carrier materials. It was found that Co/Cu-incorporation facilitates total/partial
oxidation of ethanol giving rise to high yields to H , CO and H O; whereas Mn-incorporation predominantly favored dehydrogenation and
2 x 2
condensation reactions leading to the formation of acetaldehyde and acetone. Besides, the incorporation of Mn contributed to significantly
reduce the amount of coke formed, however causing a lower intrinsic reducibility that was explained by the formation of a thermodynamically
x y
stable and hardly reducible layer of Mn Fe O solid solution.
preference for specific coordination. The nature of the
substituted ions, charges and their distribution among the O
and T sites of the ferrospinels affect their catalytic properties
Introduction
h
d
2 4
Metal ferrite compounds of the spinel type MFe O have been
31,32
1
,2
.
found to be an attracting class of materials . Because of the
chemical and thermal stability, structural, electrical, magnetic
and catalytic properties spinel ferrites have various industrial
and technical applications. Many of them are used as magnetic
As some of spinel ferrites, MFe
have been proved to be beneficial in many magnetic
applications, and they provide potential advantage for
repeated magnetic separation purposes
established magnetic properties, MFe
2 4
O (M=Co, Cu, Mn and etc.)
a
33,34
3
4
5
. Along with well-
2 4
O oxides have shown an
pigments , nanodevices , photoelectric devices , sensors and
6,7
microwave devices . Besides, spinel ferrites are attractive
8,9
exclusive catalytic ability since those materials can be reduced
to a cation excess composite and further re-oxidized to the
initially stable spinel structure. Moreover, ferrospinels can be
regarded as bi-functional systems that derives from the
presence of both redox and acid/base properties. Indeed, the
because of the importance in ferrofluids , magnetic drug
10,11
12
delivery
and hyperthermia for cancer treatment . At the
same time, they have well-established catalytic properties for
many reactions, such as: oxidative dehydrogenation of
13,14
15,16
hydrocarbons
,
decomposition of alcohols
, selective
3
5
1
7,18,19
20
21
22
MnFe
steam
2
O
4
spinel was employed as a catalyst for CO
2
reduction ,
oxidation reactions
propane total oxidation , methanol steam reforming , hydrogen
, CO and CO
2
removal , VOCs and
36
33
2
3
24
reforming
of
ethanol ,
catalytic
ozonation ,
37
38
25,26
27
dehydrogenation of ethylbenzene to styrene , Fenton reaction ,
peroxide decomposition
, NO
. Variety of studies on spinel ferrites have indicated
that most of the properties are function of the distribution of
various cations in the octahedral (O ) and tetrahedral (T ) sites
x
selective reduction and WGS
39
28,29
vapour phase methylation of pyridine with methanol and as a
magnetically separable catalyst for biologically relevant 2-
substituted benzimidazoles and quinoxalines , etc. Besides the
traditional applications, manganese ferrite showed good
performance as an adsorbent for the removal of pollutants like
reaction
40
h
d
3
0
of a spinel lattice . The preference of the individual ions for the
two types of lattice sites is determined by: the ionic radii of the
specific ions; the size of the interstices; temperature; the
VI
II41
Congo red, Cr , and Pb and azo dye Acid Red B (ARB) from
water ; it has also been successfully used as a material for
4
2
‘
pure hydrogen’ production in low-temperature water splitting
43,44
[
a]
b]
Dr. O. Vozniuk, Mr. C. Bazzo, Prof. S. Albonetti, Prof. F. Cavani,
Dipartimento di Chimica Industriale Toso Montanari, Università di
Bologna, Viale Risorgimento 4, 40136 Bologna, Italy. E-mail:
fabrizio.cavani@unibo.it
thermochemical cycles
an efficient catalyst for aromatics alkylation/decomposition ,
Knoevenagel condensation , bio-diesel production via
transesterification of vegetable oil , reactions for the
decolorization of a variety of dyes , as well as for selective
oxidation reactions of cyclohexane
2 4
. CoFe O spinel oxide is known as
45
46
47
[
Dr. N. Tanchoux, Dr. F. Di Renzo
48
Institut Charles Gerhardt Montpellier, ENSCM, 8 Rue Ecole
Normale, F-34296 Montpellier, France. E-mail: francesco.di-
renzo@enscm.fr
49,50
51
52
, styrene , alcohols and
53
olefins . Whereas, CuFe
decomposition/abatment of NOx and CO
2 4
O ferrite is a prominent catalyst in the
[
c]
d]
Dr. Jean-Marc M. Millet, Institut de Recherches sur la Catalyse et
l’Environnement de Lyon, IRCELYONUMR 5256 CNRS—Université
Claude-Bernard Lyon 1Villeurbanne CedexFrance
INSTM, Research Unit of Bologna, Via G. Giusti 9, Firenze, Italy.
54
55
2
, for Friedel-Crafts
acylation , water-gas-shift reaction and hydrogen production
56
57
58,59
[
via dimethylether steam reforming
, etc. Generally, the
catalytic effectiveness of these systems is explained by the
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601605
FULL PAPER
6
7
ability of the metallic ions to migrate between the sub lattices
without altering the structure, which makes the catalyst efficient
for many organic transformation reactions .
size or to a lower crystallinity . Similar results were previously
66
obtained with Mn-substituted in Co0.6-xMn
Table compiles results for specific surface area,
crystallite size (calculated by Scherrer eq.) and particle size
(calculations are based on N adsorption (BET) data) of fresh
x y
Fe2.4O system .
60
1
Our previous studies of different M-modified ferrospinels (M-
Fe, Ni, Co, Cu) have shown that due to the redox properties,
2
2-
these materials are effective electrons and O vectors for the
production of hydrogen via chemical-loop reforming of
calcined powders. All Mn-substituted ferrites exhibit a higher
surface area even after the thermal treatment which can be
attributed to the presence of smaller particles along with the
existence of an amorphous phase.
61–66
alcohols
. The initial experiments were carried out with
methanol and ethanol as reducing agents. The sintering
6
3
64
phenomena of magnetite and reaction mechanism together
65
Table 1. Specific surface area (SSA) and crystallite particle size of fresh
with deactivation phenomena
of ethanol anaerobic
o
MFe
2
O
4
ferrites calcined at 450 C for 8 h
decomposition/oxidation over mixed ferrospinels were also
investigated. Moreover, the properties of ferrospinels were found
Crystallite size
(Scherrer eq.),
nm
2
Particle size
(dBET), nm
Sample name
SSA, m /g
66
61
to highly depend on the composition , morphology and size ,
which are strongly connected with the preparation conditions.
Recent studies of Co0.6-xMn
Mn(II) incorporation helps to significantly reduce the amount of
coke and hence to avoid a fast deactivation of the material .
The present work onward explores the redox properties of the
x y
Fe2.4O ferrospinels have shown that
CuFe
2
O
4
60
6.9
10.4
12.0
3.5
-
18.3
16.5
16.2
8.0
66
Cu0.5Co0.5Fe
2
O
4
67
1
binary/ternary ferrospinels with a generic formula of MFe
2 4
O
/M 1-
CoFe O4
69
2
2
x
M
x
Fe
decomposition of ethanol, which is the first step of the chemical-
loop reforming of ethanol, reporting about detailed
2 4
O and its reactivity behavior during the anaerobic
Co0.5Mn0.5Fe
2
2
O
O
4
4
141
112
165
a
investigation of the modification of structural features occurring
during reaction.
Cu0.5Mn0.5Fe
10.0
6.9
MnFe
2
O
4
-
Results and Discussion
Further information on the structure, i.e. bonding and
distribution of substituted ions among the octahedral (Oh) and
tetrahedral (Td) sites of the materials, was obtained using
Raman Spectroscopy. Raman features of ferrospinels are highly
specific, i.e. affected by the sample stoichiometry, cation
Bulk characterization
68
distribution, defects, and also by the experimental conditions.
Spectra of the inverse spinels are more complicated and show
additional Raman-active modes over the five Raman modes
68
predicted by group theory:
g
F2g(1), E , F2g(2), F2g(3) and A1g.
Additional modes may appear in vibrational spectra due to local
distortions of the crystal lattice. These defects do not affect the
long-range ordering of the system, and often cannot be detected
68
by XRD.
o
Raman spectra of fresh materials calcined at 450 C for
Cu/Co-substituted ferrites (Figure 2, D-F) reveal 5 distinct
-
1
Raman modes, as reported in the literature: A1g(1) (690 cm ),
-
1
-1
-1
F
g
2g(3) (560 cm ), F2g(2) (472 cm ), E (305 cm ). The literature is
discrepant to the assignment of the Raman active modes within
a spinel lattice. However, the highest frequency mode (A1g(1)) at
-
1
6
90 cm is usually assigned to the symmetric stretching of the
o
69,70
Figure 1. XRD patterns of fresh MFe
2
O
4
samples calcined at 450 C for 8 h:
M-O bond in Td void
. In addition to the 5 theoretically
A-CuFe O , B-CoFe O , C-Cu0.5Co0.5Fe O , D-Cu0.5Mn0.5Fe O , E-
2
4
2
4
2
4
2
4
predicted modes, samples D and E exhibit an additional peak at
2 4 2 4
Co0.5Mn0.5Fe O , F-MnFe O
-
1
2+
3+
A
1g(2) 616 cm , caused by disorder effects of M and Fe over
70
2 4
the Td and Oh sites . This shoulder for a CoFe O spinel was
X-ray Diffraction patterns of calcined spinels are shown in Figure
. Resulted patterns confirm the formation of a spinel oxide, as
the main phase for all the samples. However, the presence of
69
also observed by Chandramohan and others . They correlated
an increase in the intensity of A1g(2) mode with lowering of
inversion parameter in the following spinel, also being supported
by Mössbauer studies. Looking at the spectra of A-C samples
1
o
CuO with a reflection at 38.8 is detected for CuFe
2 4
O (the
amount of CuO as 21% was identified by Rietveld refinement
w
(
Mn-substituted ferrites), the position of the A1g(1) shifts to a
with its further consideration for all calculations). Broadening of
the diffraction patters and lower intensity observed for the Mn-
-1
lower frequency at about 661 cm along with
a distinct
broadening of the following spectra. The similar behavior was
incorporated samples (MnFe
2 4
O ,
Co0.5Mn0.5Fe
O
2 4
and
66
2+
observed for Co/Mn-incorporation . Hence, Mn ions are
Cu0.5Mn0.5Fe
2 4
O ) can be attributed to a decrease in a particle
3+
proposed to partially replace Fe ions in the Td sites, which may
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601605
FULL PAPER
cause the change of the inversion degree towards the formation
reduction of Cu0.5Mn0.5Fe
follows the pass III (Figure 4), where the first peak at lower
temperature corresponds to the stepwise exclusion or/and
2
O
4
and Cu0.5Co0.5Fe
2
O
4
samples
71,72
of the normal spinel.
The low intensity of the peaks and
partial energy shifts can be related also to a lack of a long range
in the small dimension systems along with an effect due to the
0
reduction of the M-incorporated cations to form M (Cu ), MO
6
9
high defectivity and low crystallinity as confirmed by XRD.
(MnO/CoO) and Fe
O
3 4
phases. Further reduction of the
2 4
Cu0.5Co0.5Fe O
ferrite goes through the following stages: (1)
Fe O →FeO (+0,77eV), (2) CoO→Co (-0.28 eV), (3)
0
3
4
0
FeO→Fe incom (-0.44 eV), reaching ɑ~65%. On the other hand,
the reduction of Cu0.5Mn0.5Fe ferrite can be schematized as
follows: (1) Fe →FeO, (2) FeO→Fe incom, which corresponds
2 4
O
0
3
O
4
to ɑ~65%. It turned out that it is difficult for MnO to be reduced to
0
its metallic component Mn , due to its highly negative reduction
66
potential (-1.18 eV) and thus this step can occur only at very
0
extreme temperatures (∆G react,Table S-1). The reduction of the
CoFe
2
O
4
, Co0.5Mn0.5Fe
2
O
4
and MnFe
2
O
4
samples follows the
, the first small peak at
00 C can be attributed to the primary spinel reduction with a
stepwise formation of CoO and Fe phases (which
corresponds to ɑ~8%), whereas in Mn-incorporated ferrites
MnFe this step starts at lower
temperature. However, the total reduction extent of CoFe
ɑ~86%) was much higher than that of Co0.5Mn0.5Fe (ɑ~61%)
and MnFe (ɑ~45%) samples and can be explained by the
formation of a hardly reducible phase of MnO (or in the form of
Fe O oxide, where y=1-x).
pass II in Figure 4. In a case of CoFe
2 4
O
o
4
3 4
O
(
2
O
4
2 4
and Co0.5Mn0.5Fe O )
Figure 2. Raman spectra of MFe
2 4 2 4 2 4
O ferrites: A-MnFe O , B-Cu0.5Mn0.5Fe O ,
2 4
O
C-Co0.5Mn0.5Fe
2
O
4
,D-CoFe O , E-Cu0.5Co0.5Fe O and F-CuFe O
2 4 2 4 2 4
(
2 4
O
2 4
O
TPR study
M
x
y
TPR curves: signal-vs-T
(
II,III)
The reduction of iron
3 4
oxide: Fe O →FeO→Fe strongly
depends on the presence of another metal oxide in the M-
2
8,66,73–77
64,78
modified ferrite
. As we reported previously
) started at relatively high
temperature of 400 C, showing the occurrence of three main
, the
reduction of pure magnetite (Fe
3 4
O
o
o
o
o
reduction events: I – (400 C), II – (550 C) and III – (650 C).
Thus, the TPR measurements were performed in order to
investigate the role of different metals on reducibility of M-
modified ferrospinels (where M=Co, Cu, Mn or Cu/Co, Cu/Mn,
Co/Mn). According to the XRD and Raman results, discussed
previously, the main phase obtained at the end of the
preparation procedure was a Fe
3 4
O -like spinel oxide phase with
a generic formula of MFe . Following the TPR profiles of M-
2 4
O
modified ferrites, plotted in Figure 3, mainly two steps of the
reduction processes can be distinguished: the reduction of iron
oxide to metallic iron and the reduction process of other
incorporated metal oxide to the corresponding metal. However,
it’s difficult to distinguish each stage of the reduction since some
steps seamlessly flow from one to another which finally causes
an overlap. Nevertheless, the obtained results and the further
2+
elaboration based on the nature of the foreign metals (M ), their
reduction potentials and the amount of consumed H helped us
2
to track back possible steps of the MFe
O
2 4
reduction process,
which are summarized in Figure 4.
Looking at the TPR profile of CuFe O , the first peak at
2 4
o
2
40 C can be attributed to a starting reduction of the initially
0
w
present copper oxide (CuO~21% ) to the metallic copper (Cu ),
which contributes to ɑ~16% (where ɑ is the degree of reduction).
o
The second peak at ~ 340 C was attributed to a primary
0
stepwise reduction of the spinel with a final formation of Cu and
Fe
3 4
O phases (ɑ~27%), following the pass I (Figure 4); the
0
further reduction of the Fe
O
3 4
to FeO and of FeO to Fe incom
‘FeO to Fe incom’ is referred to incompleteness of the reduction
step) appears at higher temperatures with a total ɑ~70%. The
0
(
2 4
Figure 3. TPR profiles of MFe O ferrites with temperature program from 50
o
o
to 800 C (+isotherm at T-800 C for 40 min)
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601605
FULL PAPER
-
Mn
x
Fe
y
O (where y=1-x), the evidence of its formation is further
supported by ex-situ XRD study).
Ex-situ XRD study
2 2 4 2 4
The H -reduction of Cu0.5Mn0.5Fe O and Co0.5Mn0.5Fe O
ferrites, discussed previously, has shown remarkable differences
in the reaction models. The same idea was taken to explain the
reduction of these samples with ethanol. The ex-situ study of
XRD patterns, obtained after 1, 3, 5 and 20 min of ethanol
exposure, was performed in order to justify this statement.
2 4
Figure 4. Reduction scheme of MFe O ferrites
TPR curves: α-vs-Temperature
Different modifications in the composition of bare-Fe
MFe (M-Cu, Co, Ni, Mn, Mg, Ca, etc.) have been done by
many research groups in order to change its redox properties.
3 4
O to
2 4
O
2+
Moreover, the so called “promoting effect” of Cu was also seen
2
0,77,79,80,
,
however a real role of copper ions and the reduction
mechanism has not been clearly elucidated. Thus, the primary
idea was to focus our study on a deep understanding of TPR
profiles with a further data elaboration for possible clarification of
2 4
MFe O reduction model.
Conventionally, the shape of an (α-vs-time) isotherm can
reveal information on the mechanism of solid state reactions.
81
And, according to Jones the reduction of CuO produces an S-
shape (sigmoidal) reduction isotherm familiar for a “nucleation
and autocatalytic” reduction model, where the reduction consists
of three distinct stages: 1. induction, 2. autocatalysis and 3.
decreasing rate. However, some oxides may be reduced via
spontaneous nucleation over an entire surface which is usually
2 4
Figure 5. The evolution of phases during the MnFe O stepwise reduction as
determined by means of ex-situ XRD.
characterized by
a continuously decreasing rate. If the
nucleation is very fast, the real formation of separated and
independent nuclei cannot be distinguished leading to a total
coverage of the solid oxide particle with a thin layer of metallic
81
product, and typical for Mn
decided that the best depiction of the reduction models of
MFe ferrites can be described being based on these two
mechanisms.
Figure S-1 shows α-vs-temperature reduction curves of
MFe samples. Cu-incorporated ferrites (CuFe
) produce the S-shaped
sigmoidal) reduction curve with a strong acceleration of the
x y
O → Mn redox system . Thus, we
2 4
O
O
2 4
2 4
O ,
2 4 2 4
Cu0.5Co0.5Fe O and Cu0.5Mn0.5Fe O
(
o
reduction rate above 160-200 C, that can be associated with a
self-catalyzed nucleation (autocatalysis). Likely, nuclei catalyze
further nuclei formation, due to branching of nuclei or to the
2
catalytic role of the Cu metal in H dissociation. According to this,
the reduction occurs on the basis of hydrogen spillover effect,
0
where Cu plays the role of a catalyst which is needed to
Figure 6. Schematic representation of C
as “nucleation/autocatalytic” model and B-MnFe
2
H
5
OH-reduction over: A-CuFe
2 4
O ,
dissociate H
migrate from a hydrogen-rich (Cu ) to a hydrogen-poor (spinel)
surface (Figure S-2, A). Conversely, the reduction of MnFe
samples produces slightly elongated
2
into 2H atoms. Essentially, hydrogen atoms would
2
O
4
, as an “egg-shell” model
0
2 4
Essentially, a reduction cascade of the MnFe O ferrite
2 4
O
(
“
Figure 5 and Figure S-3) was chosen as a reference for the
egg-shell” model (Figure 6,B). The results, obtained after
2 4
and Co0.5Mn0.5Fe O
curves without a strong acceleration of the reaction rate which
led us to an idea of a different model, we called as an “egg-shell”
feeding ethanol for 1 min, reveal the appearance of a new
Mn Fe O oxide along with a remaining “reformed” spinel phase
). Moreover, with increasing the reduction time
from 1 to 20 min the relative amount of the spinel phase
continued to decrease (from 92 to 67 % ) whereas that of the
Mn Fe O increased (from 8 to 33 % ) without an appearance of
the other phases. This apparently shows that the reduction of
the MnFe goes through the formation of Mn Fe O, which is
an important proof of the previously discussed TPR results
x
y
model (Figure S-2, B). Like, if the reduction of MnFe
2 4
O and
(Mn1-xFe2+xO
4
2 4
Co0.5Mn0.5Fe O
ferrites occurs immediately on the contact with
hydrogen and lead to the expelling of Mn cations out of the
spinel structure with a formation of the new phases: a Fe-rich
2+
w
x
y
w
reformed spinel (M1-xFe2+x
MnO oxide (since: MnO→Mn , ε=1.18 eV) , like an “egg-shell”
or in the form of a thermodynamically stable oxide solid solution
4
O ) and a scarcely-reducible layer of
0
66
O
2 4
x
y
(
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601605
FULL PAPER
where we proposed the reduction mechanism which goes
through the formation of hardly reducible and thermodynamically
stable layer of Mn Fe O oxide.
x y
to 82%). On the contrary, even though the highest surface area
was shown by MnFe
catalysts (SSA values of the fresh samples are reported in Table
), the Mn-incorporation illustrates an opposite inhibiting effect
on MFe -reduction, causing its drop from 82% (CoFe ) to
and from 82% (CuFe to 73%
). This suggests that the difference in the
2 4 2 4 2 4
O , Co0.5Mn0.5Fe O and Cu0.5Mn0.5Fe O
1
O
2 4
2 4
O
1
0% (Co0.5Mn0.5Fe
2 4
O )
2 4
O )
(Cu0.5Mn0.5Fe
2 4
O
reducibility of the resulted M-modified ferrospinels is related to
the nature of the M-incorporated cation and its intrinsic reactivity.
2 4
Figure 7. The evolution of phases during the Cu0.5Mn0.5Fe O stepwise
reduction
The stepwise reduction of Cu0.5Mn0.5Fe
2 4
O (Figure 7 and
Figure S-4) clearly indicates that a primarily formed phase
st
0
during the 1 min of ethanol exposure was Cu (with app. 11%
Increasing the reaction time to 3 min resulted in the formation of
the second reduced phase – an oxide solid solution Mn Fe O,
with app. 12% . Subsequent increase of the reduction time to 5
min led to a decrease of the spinel relative amount to 63%
whereas the one of Mn Fe O increased to 25%
phase remained practically unchanged, approx. 12%
step up in the reduction time (up to 20 min) resulted in the
formation of strongly reduced species like iron carbides (Fe
and Fe C), with total amount of 55% . On this basis, we may
point out that the reduction of Cu0.5Mn0.5Fe at first proceeds
with the formation of Cu and iron-rich reformed spinel (Cu0.5-
w
).
x
y
2 4
Figure 8. The evolution of phases during the Co0.5Mn0.5Fe O stepwise
reduction
w
w
0
x
y
w
, while Cu
w
. Further
5
C
2
3
w
2
O
4
0
0
x
4
Mn0.5Fe2+xO ), where Cu may play its role as a catalyst for the
deeper reduction with a formation of Mn
as also supported by TPR study.
x
Fe
y
O and Fe-carbides,
2 4
The reduction cascade of the Co0.5Mn0.5Fe O
ferrite is
shown in Figure 8 and Figure S-5. It is worth noting that
incorporation of Co-cations may stabilize iron in the ferric state
20
in an even higher proportion, exceeding the nominal 1:2 ratio ;
this in turn determines the reduction mechanism and, as a
consequence, phases which are formed during the reduction
step. From the obtained results, it is clear that the reduction of
2 4
Figure 9. The evolution of phases during the MFe O stepwise reduction
Co0.5Mn0.5Fe
2
O
4
has the same nature as the one of MnFe
which primarily gives rise to app. 15% of M Fe O oxide and
of Fe-enriched spinel with its further evolution to 42 and
, respectively. The formation of M Fe O acts as a
2 4
O ;
Mössbauer study
w
x
y
8
5
5%
8 %
w
2 4
Mössbauer spectra on the MFe O ferrites (Figure S-6)
o
w
x
y
reduced with ethanol at 450 C for 20 min have been recorded in
order to support the statements made on the basis of the ex-situ
XRD studies and to gain new quantitative and compositional
information. Experimental data were fitted using Lorentzian
functions and the hyperfine parameters derived from the fittings
are listed in Table S-3.
“
shielding” layer and prevents a deep reduction of the following
ferrite. This also can explain such high difference in the
reduction degree (ɑ) between Co0.5Mn0.5Fe (10%) and
(73%) whereas samples without Mn-
incorporation showed the same ɑ: CoFe (82%) and CuFe
82%) (Figure 9).
Figure 9 and Table S-2 summarize the results on the
Rietveld refinement for all MFe samples after ethanol
reduction within 20 min. By looking at the reduction degree
2 4
O
2 4
Cu0.5Mn0.5Fe O
O
2 4
2 4
O
(
2 4
The spectrum of Cu0.5Co0.5Fe O ferrite (with ɑ~100%) was
fitted with 4 sextets and one doublet. The first sextet (-0.04
mm/s, H-35.9 T and Rel.In.-10%at) was attributed to Fe in the
CoFealloy phase, which was also detected by XRD. The other 3
O
2 4
0
2+
values, an introduction of Cu ions into the Co/Fe system shows
its strong promoting effect on the reduction of Cu0.5Co0.5Fe
α=100%), caused by a synergetic effect of Cu-Co couple (while
by itself the reduction degree of CuFe and CoFe is equal
sextets correspond to the two iron carbides (Fe
Fe -Hagg carbide), formed as the main phases (84%at) during
the reduction process. The best fit of the experimental spectrum
was obtained with two sextets for Fe (-0.19 mm/s, H-18.9 T
3
C-cementite and
2 4
O
5 2
C
(
O
2 4
2 4
O
5 2
C
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601605
FULL PAPER
2 4
and -0.17 mm/s, H-11.2 T) and one shared sextet (-0.21 mm/s, and -1.04 mm/s, -0.97 mm/s; MnFe O : -0.23 mm/s, -0.57
H-21.0 T) for Fe
parameters of the corresponding sextets are in
5
C
2
+Fe
3
C
components. The obtained
good
. The doublet (-
.18 mm/s; -1.05 mm/s) representing just 6%at of the total
mm/s and -1.09 mm/s, -0.64 mm/s) were attributed to the
3+
2+
a
same ferric (Fe ) and ferrous (Fe ) iron oxides, respectively, as
detected in Cu0.5Mn0.5Fe sample. From the obtained results,
it is clear that the reduction of Co0.5Mn0.5Fe and MnFe
82–84
agreement with those reported in literature
2 4
O
0
O
2 4
2 4
O
3+
spectrum, was ascribed to the presence of Fe phase in the
form of carburized Fe oxide (or Co/Fe oxide) resulting from the
incomplete transformation of the ferrite (it should be recalled that
X-ray diffraction is not extremely sensitive and up to 5% of the
remained/formed phase could not be detected).
samples is of the same nature and follows a proposed “egg-shell”
model, with only the formation of two new phases: a Fe-rich
reformed spinel (M1-xFe2+x
hardly reducible layer of M
O
4
) and a thermodynamically stable
Fe O oxide solid solution. These
x
y
results are in agreement with those of TPR and ex-situ XRD
studies.
The experimental spectrum of CuFe
was fitted with 5 sextets. Spinel component (24%at) with generic
formula: Cu1-xFe2+x composed of sextets with the
corresponding distribution of Fe-cations in between T and O
2 4
O ferrite (with ɑ~82%)
O
4
2
2 5
Reactivity tests: C H OH anaerobic decomposition
d
h
According to our previous studies, spinel ferrites were found
sites: -0.25 mm/s, H-49.0 T and -0.67 mm/s, H-45.9 T,
respectively. Another 3 sextets are similar to those discussed
above and were attributed to the existence of a relatively high
2-
to be effective electrons and O vectors for the production of
61–66
pure hydrogen via chemical-loop reforming of alcohols
. Thus,
the particular interest represents also the reduction behavior of
MFe samples using ethanol as a probe molecule. The cycling
time was chosen according to the initial results achieved over
amount of Fe
x
C
y
phases (Fe
3
C+Fe
5
C
2
~ 76%at).
2 4
O
Mössbauer spectrum of CoFe
O
2 4
sample (with ɑ~82%) was
fitted with 6 sextets. The first two sextets (Fe in T
mm/s and H-48.4 T; Fe in O sites: -0.55 mm/s and H-44.9 T)
represent a remained spinel component – Co1-xFe2+x (10%at)
d
sites: -0.25
63
3 4
pristine Fe O . We reported that in approximately 20 min of
h
ethanol feeding, the material was constantly reduced until it
reached its `steady state` with the stable composition under
O
4
resulted from the incomplete reduction, as supported by ex-situ
XRD study. The third sextet (-0.02 mm/s and H-35.0 T) with a
relatively high contribution to the total spectrum (51%at) was
these reaction conditions. Moreover,
a screening of the
reduction temperature (Figures S-8, S-9 and S-10)
demonstrated that, due to kinetic reasons, a best compromise
between the reduction extent and tos (time-on-stream) was
0
attributed to Fe and should correspond to CoFealloy and FeC0.045
phases, see Table S-3. The three remained sextets were
assigned to Hagg-carbide (Fe
o
achieved at 450 C.
5 2
C ), representing 39%at with the
Figures 10-15 show conversion and yields to the main
products (yield >5%), obtained during the ethanol reduction step
following hyperfine parameters: -0.17 mm/s and H-20.5 T, -
o
0.18 mm/s and H-18.5 T, -0.17 mm/s and H-9.8 T, which
performed at 450 C for 20 min. The general statement that can
82
5 2
corresponds to the three types of Fe in the unit cell of Fe C .
be applied to all the MFe
product is greatly affected by the degree of metal oxide
reduction. Thus, Cu0.5Co0.5Fe ferrite (Figure 10) starting from
2 4
O ferrites is that the amount of each
Experimental data of the Cu0.5Mn0.5Fe
ɑ~73%) were fitted with 5 sextets and 2 doublets. Spinel
component (24%at) composed of 2 sextets: Fe in T sites with -
sites with -0.49 mm/s and H-
5.6 T. Another 3 sextets are similar to those discussed above
2 4
O sample (with
2 4
O
d
the first minutes of the reaction showed 100% of conversion
followed by its decrease down to 95%, while the extent of spinel
reduction increased from 0 to 100 % (Table S-2). One of the
main products of ethanol conversion was hydrogen, with an
initial yield of about 50%, which increased up to 74%. The
similar trend, as the one for hydrogen, was observed for CO with
0.30 mm/s and H-48.8 T; Fe in O
h
4
and were attributed to the existence of a relatively high amount
of Fe phases (Fe C+Fe ~ 53%at). Remaining two doublets
that represent ferric (Fe as 11%at) and ferrous (Fe as 12%at
x
C
y
3
5 2
C
3
+
2+
)
3+
2+
iron oxides characterized by a : Fe - 0.28 mm/s; Fe - 1.10
its max yield about 20%. Looking at CO
may notice a similar decreasing trend (starting yields: CO
and H O-18%, final yields: CO -12% and H O-4%). Thus, taking
2
and H
2
O profiles, one
3+
2+
mm/s and a : Fe - 0.78 mm/s; Fe - 0.62 mm/s and were
attributed to the presence of the Mn Fe O oxide solid solution (a
2
-32%
x
y
2
2
2
3
+
85
small amount of Fe in FeO was also seen by Johnson ).
Eventually, the obtained results support the ex-situ XRD study
into account the elaborated results we suggest that the main
reactions during spinel reduction involved the partial and total
oxidation of ethanol:
and a lower ɑ-value obtained over Cu0.5Mn0.5Fe
comparison to CuFe and CoFe ones, which in turn is
explained by the direct effect of Mn-incorporation, leading to the
formation of the scarcely reducible phase - Mn Fe O.
Mössbauer spectroscopy results on the Co0.5Mn0.5Fe
with ɑ~10%) and MnFe (with ɑ~8%) ferrites once again
2 4
O sample in
2
O
4
2 4
O
(1) C
2) C
(3) C
2
H
2
H
2
H
5
5
5
x 2
OH + MeO →2 CO +3 H + MeO(x-1)
x
y
(
OH + 3 MeO
x
→2 CO
2
2
+ 3 H
+ 3 H
2
2
+ 3 MeO(x-1)
2 4
O
(
O
2 4
x
OH + 6 MeO →2 CO
O + 6 MeO(x-1)
confirmed the apparent differences observed between the Mn,
Mn/Co- and Co, Cu, Cu/Co, Cu/Mn-incorporated samples.
Corresponding spectra were all fitted with 2 sextets and 2
doublets. Two sextets were used to fit the remained spinel
In addition, the formation of water involves several possible
reactions:
(4) H
5) C
2
2
+ MeO
x 2
→H O + MeO(x-1)
component (in the form of M1-xFe2+x
O
4
): Co0.5Mn0.5Fe
2
O
4
with
5%at (Fe in Td sites: -0.29 mm/s, H-49.3 T; Fe in Oh sites: -
.45 mm/s, H-46.5 T) and MnFe with 68%at (Fe in Td sites: -
.28 mm/s, H-48.7 T; Fe in Oh sites: -0.44 mm/s, H-45.6 T).
: -0.28 mm/s, -0.87 mm/s
(
H
5
OH→2 C + 2 H + H O
2 2
6
0
0
O
2 4
Acetone formation, which declined with time from 32 to 2%,
65
as reported previously can be a result of an aldol condensation
of the intermediately formed acetaldehyde, with further
2 4
The two doublets (Co0.5Mn0.5Fe O
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601605
FULL PAPER
isomerization of the aldol into hydroxyketone followed by a
reverse reaction to form acetone and formaldehyde (converted
into CO and H
2
). Either, as recently proposed by Rodrigues et al.,
by ketonization of carboxylates (acetates) which are obtained by
86
dehydrogenation of this alcohol followed by oxidation . In fact,
the formation of acetates over Co0.6-xMn samples was
Fe2.4O
x y
66
.
confirmed by DRIFTS study
Figure 13. Conversion and Yields to the main products obtained for
during the 20 min reduction at T-450°C
2 4
Cu0.5Mn0.5Fe O
Figure 10. Conversion and Yields to the main products obtained for
during the 20 min reduction at T-450°C
2 4
Cu0.5Co0.5Fe O
Figure 14. Conversion and Yields to the main products obtained for
during the 20 min reduction at T-450°C
2 4
Co0.5Mn0.5Fe O
2 4
Figure 11. Conversion and Yields to the main products obtained for CuFe O
during the 20 min reduction at T-450°C
2 4
Figure 15. Conversion and Yields to the main products obtained for MnFe O
during the 20 min reduction at T-450°C
Increasing yield for an acetaldehyde (from 1 to 15%) over
continuously reducing sample indicates that
dehydrogenation/deoxyhydrogenation of ethanol also occurred:
the
(
6) C
7) C
2
H
H
5
OH→C
2
H
4
O + H
2
(
2
5
OH + MeO
x
→C
2
H
4
O + H
2
O + MeO(x-1)
2 4
Figure 12. Conversion and Yields to the main products obtained for CoFe O
during the 20 min reduction at T-450°C
Coke and heavy compounds (mainly aromatics) were formed
at the very beginning of the reaction time (~1%), with a
progressive increase in yield up to 39%. The generation of coke
was likely due to the dehydrogenation and deoxygenation of
ethanol:
2 5 2
(8) C H OH→C + CO + 3 H
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601605
FULL PAPER
(
9) C
2
H
5
OH→2 C + H
2
O + 2 H
2
min tos), the formation of two phases, M
helped us to explain their peculiar behavior.
Looking first at the Co0.5Mn0.5Fe graph (Figure 14), the
conversion of ethanol during 20 min tos slightly decreased from
100 to 98%, giving its max reduction degree of 10% only (while
x y 4
Fe O and M1-xFe2+xO ,
Moreover, a disproportionation of CO (Boudouard reaction)
2 4
O
can also lead to the coke deposition:
10) 2 CO→CO + C
(
2
by itself CoFe
completely reduced – 82%). At the same time, conversion of
ethanol over MnFe decreased from 100 to 96% (Figure 15).
The evolution of products yields over these two samples showed
the same trend variations in function of time, viz. in
approximately 5 min (zone I) the yields to all products varied the
2 4
O during the same 20 min tos was almost
2 4
Looking at the conversion and yields evolution over CuFe O
ferrite (Figure 11), the results demonstrate similar trends in the
products distribution and, as a consequence, similar reduction
paths that involve the same set of reactions (eq. 1 to 10). The
2 4
O
2 4
reduction of CuFe O sample started with 95% of ethanol
conversion which approached to 92% after 20 min reduction
time, while degree of reduction reached 82%. As for the
previously discussed ferrite, hydrogen was one of the main
products of ethanol conversion, with the increasing trend from
most, while the reduction degree approached 6 % for MnFe
2 4
O
and 8% for Co0.5Mn0.5Fe (that corresponds to the reduction
2 4
O
rate of 1.2 and 1.6 %/min, respectively). Thereafter, a zone II
started (from 5 to 20 min tos, ∆t=15 min), where all yields were
5
9
7 to 75%. However, H
2 2
O and CO have decreasing trends from
stable, while the total reduction extent reached 8% for MnFe
and 10% for Co0.5Mn0.5Fe (corresponding to ∆α=2% and
reduction speed of 0.13%/min). In comparison to Co, Cu, Cu/Co
and Cu/Mn ferrites, the reduction of MnFe and
samples demonstrated different trends in the
2 4
O
to 3% and 23 to 12%, respectively. This may indicate that the
2 4
O
total oxidation is more favorable at the beginning of the
reduction process (eq.3) over less reduced material.
a
2 4
O
Acetaldehyde was observed only after 5 min of tos with the
increasing trend from 2 to 9%, eq. 6 and 7. Acetone was formed
at the beginning with the yield of 31% and then its amount
decreased down to 5%.
2 4
Co0.5Mn0.5Fe O
products distribution. Namely ethane, ethylene and acetone
were formed continuously within 20 min tos with much higher
yields of about 14%, 15% and 39% for MnFe
and 33% for Co0.5Mn0.5Fe , respectively:
2 4
O and 12%, 13%
Coke and heavy compounds (mainly aromatics) were also
2 4
O
formed with the comparably high yield as for the Cu0.5Co0.5Fe
sample (increasing from 7 to 38%).
2 4
O
(11) 2 C
2
H
5
OH → CH
3
CH
3 3 2
+ CH CHO + H O
Results obtained with the CoFe
to those obtained with the Cu0.5Co0.5Fe
2
O
4
ferrite were very similar
and CoFe
(12) C
2
H
5
OH → C
2
H
4
+ H
2
O
2
O
4
2 4
O
samples (Figure 12). The conversion, starting from the
beginning, reached its max (app. 100%) and was stable during
the reduction time, yielding to 82% of material reduction. The
According to the TPR, Mössbauer and ex-situ XRD results,
discussed previously, the reduction of these samples can be
described by a rapid reduction over the entire surface that
occurs immediately on the contact with the reducing stream and
x 2 2
formation of CO H and H O as the main products led to the
2+
same conclusion on the prevailing reactions of total and partial
ethanol oxidation (eq. 1-3). The main difference between the
previous two samples was observed in the amount of produced
acetaldehyde (in this case the yield was very small, ≤1% during
all 20 min of tos) and coke+HC (increased from 4 to 60%).The
leads to the expelling of Mn cations out of the spinel structure.
The formation of new phases resembles more to a core-shell
catalytic system, comprised of Fe-enriched spinel core (M1-
x
4 x y 4 x y
Fe2+xO ) covered with a M Fe O shell (M1-xFe2+xO @M Fe O).
According to this, a significant slowdown of the reduction speed
from 1.2/1.6 to 0.13%/min was attributed to an accumulation of
2 4
conversion and yields during reduction of Cu0.5Mn0.5Fe O are
shown in Figure 13. The addition of Mn into Cu/Fe oxide just
slightly affected the catalytic performance of the resulted
material (in comparison to CuFe O ). Thus, during 20 min tos
2 4
the sample was reduced until about 73%; ethanol conversion
starting from the very beginning approached 99%, and then was
stable till the end of the reduction step. The formation of
x y
the M Fe O phase, which has a high thermodynamic stability
and as a consequence a scarce attitude towards reduction to the
metallic form, where the further diffusion of the reducing stream
into the bulk (core) may be more problematic. Moreover, the
“core@shell” model may also explain the formation of acetone
directly from the ethanol via ketonization of acetates, where the
products (eq. 1-10) showed the following trends: H
from 50 to 81%, CO varied from 18 to 16% (with the max value
of 26%), CO decreased from 20 to 13%, H O decreased from 9
to 2%, acetone decreased from 36 to 1% and acetaldehyde was
detected just in traces.
2
increased
formed shell of M
delivers oxygen (O
noting, that DRIFTS study performed earlier over Co0.6-
x
Fe
y 4
O acts as a catalyst while M1-xFe2+xO core
2
L
) to restore the pre-reduced shell. It is worth
2
66
x
Mn
and carbonates as important surface intermediates. Moreover,
only in the case of Mn0.6Fe2.4 there was no evidence for the
acetate species to be transformed into carbonate, which may
also justify the high acetone yields over Mn0.6Fe2.4 as well as
with MnFe and Co0.5Mn0.5Fe
Coke and heavy compounds were formed in a much smaller
amount over MnFe (from 16 to 8%) and Co0.5Mn0.5Fe
(from 21 to 12 %) samples; the total integrated yields to coke for
all MFe ferrites are plotted in Figure S-7.
Fe2.4O
x y
ferrospinels confirmed the formation of acetates
Figure 14 and Figure 15 show the conversion of ethanol and
O
y
the yields to the various products obtained over MnFe
2 4
O and
Co0.5Mn0.5Fe
O
2 4
ferrites, respectively. The data demonstrate an
O
y
apparent difference in the reducibility and reactivity of these
ferrites in comparison to the previously discussed samples.
Specifically, we could distinguish two different zones, during
which the spinel first underwent a progressive reduction followed
by a “pseudo steady state”. Taking into account the ex-situ XRD
O
2 4
2 4
O .
O
2 4
2 4
O
2 4
O
and Mössbauer studies of the MnFe
ferrites performed at different reduction stages (1, 3, 5 and 20
2
O
4
and Co0.5Mn0.5Fe
2
O
4
In order to evaluate the total amount of coke formed during
the reduction step with ethanol (within 20 min), an elemental
analysis (CHNS) on the different materials was carried out.
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601605
FULL PAPER
Table 2 summarizes the results of CHNS analysis over MFe
ferrospinels. According to the obtained results, the highest
amount of coke accumulated during ethanol reduction was
2
O
4
deposition, that is an important issue to produce high purity H
via CLR process.
2
detected over Cu0.5Co0.5Fe
declined in the following order: Cu0.5Co0.5Fe
2 4
CuFe > Cu0.5Mn0.5Fe > Co0.5Mn0.5Fe > MnFe O .
2
O
4
and CoFe
2
O
4
samples and it
Experimental Section
2 4
O > CoFe
2 4
O >
O
2 4
O
2 4
2
O
4
Materials preparation
Table 2.CHNS results over MFe
2
O
4
ferrites after 20 min reduction
2 4
MFe O ferrospinels with different atomic ratio Fe/M (where
M=Co, Cu, Mn or Cu/Co, Cu/Mn, Co/Mn) were prepared using
45
Sample
C
w%
the co-precipitation method described elsewhere . Nitrate
precursors were used for all of the synthesized materials. The
mixed solutions of metal precursors containing 100mL of 1M
(
after 20 min reduction)
CuFe
2
O
4
6.85
16.33
11.58
1.66
6.08
1.48
Fe(NO
3
)
3·9H
2
O, 100 mL of 0.5M M(NO
3
2
)
2·xH
2
O (or 100 mL of
1
Cu0.5Co0.5Fe
2
O
4
0.25M M (NO )2·xH O and 100 mL of M (NO )2·xH O solutions)
3 2 3 2
were put into a separation funnel and then drop-by-drop added
into the reaction vessel containing 0.5L of 2 M NaOH aqueous
solution at the temperature of 50°C under vigorous stirring with a
constant speed. The suspensions were stirred for 2h at the
constant temperature of 50°C adding NaOH 3M in order to
maintain the pH above 13. The precipitates were separated by
vacuum filtration by means of a Büchner funnel and then
washed with about 1.5L of distilled water at room temperature to
remove sodium and nitrates ions. All the washed samples were
dried at 120°C in air for 12 h, milled in agate mortar and then
2 4
CoFe O
Co0.5Mn0.5Fe
Cu0.5Mn0.5Fe
2
2
O
4
4
O
2 4
MnFe O
Conclusions
calcined in static air at 450°C for 8h with the heating rate of
o
1
0 C/min.
In this study, a rigorous approach has been taken to
describe the reduction of binary/ternary ferrospinels in
Characterization of fresh/used materials
H
2
/C
TPR profiles together with ex-situ XRD and Mössbauer studies
suggested that all Cu-incorporated samples CuFe
2 5
H OH atmosphere. Experimental results obtained from the
X-ray diffraction (XRD). XRD patterns of the fresh samples were
measured on Bruker D8 Advance diffractometer (using Cu Kα
radiation with λ = 1.5406 Å) with a Bragg-Brentano geometry
and equipped with a Bruker Lynx Eye detector. The data were
recorded varied from 5 to 80o 2θ diffraction angle with an
acquisition time of 1 s per each step (0,01 degree). Whereas,
ex-situ XRD study was performed using a Bruker D8 X-ray
diffractometer equipped with a 9-sample holder and a Cu Kα
O ,
2 4
Cu0.5Co0.5Fe
O
2 4
and Cu0.5Mn0.5Fe
O
2 4
were reduced via a so-
0
called “autocatalytic” model, where the primarily formed Cu
nuclei (Cu →Cu ) activate the reducing gas (H
which in its turn catalyze the further reduction of the newly
reformed Fe-enriched spinel. Whereas, the reduction of
MnFe
2+
0*
2
or C
2
H
5
OH),
O
2 4
and Co0.5Mn0.5Fe
O
2 4
ferrites was described by a
o
radiation source. Diffraction angle (2θ) was varied from 4 to 80
during the measurement (number of points: 3714, time per point:
reduction which occurs over the entire surface immediately on
the contact with the reducing species and leads to the expelling
384 s). TOPAS-4.2 program was used to calculate the relative
2+
of Mn cations out of the spinel structure with a formation of
new phases: a Fe-rich reformed spinel (M1-xFe2+x ) and a
O oxide solid
phase content in the reduced samples. The calculations are
based on the Rietveld refinement method, which permits the
reproduction of the whole diffractogram, through the optimization
of both structural (peak position and intensity) and non-structural
O
4
stable, hardly-reducible layer of wustite-like M
solution, depicted as an “egg-shell” model.
x y
Fe
Moreover, based upon the presence of the M
main differences in ethanol transformation shown with CoFe
CuFe and MnFe
x y
Fe O layer,
(peaks shape) data.
2
O
O
4
,
,
2 4
O ,
2 4
Cu0.5Mn0.5Fe O ,
2 4
Cu0.5Co0.5Fe O
2
4
Surface area measurements. The specific surface area was
measured applying the single point BET method. The instrument
used for this analysis was a Carlo Erba Sorpty 1750 Fisons
Instrument. For the analysis, around an amount of 0.5 g of the
sample was placed inside the sample holder and then heated at
Co0.5Mn0.5Fe
O
2 4
ferrites along with samples reducibilities were
attributed to the formation of a “core-shell” system, comprised of
Fe-enriched spinel core (M1-xFe2+x ) covered with the M Fe
shell (M1-xFe2+x @M Fe O).
O
4
x
y
O
O
4
x
y
Moreover, redox properties of M-modified ferrospinels were
found to depend strongly on the nature of incorporated cation
1
50°C under vacuum (≤4 Pa) to release water, air, CO
molecules adsorbed. Subsequently, the sample was placed in
liquid nitrogen, and the adsorption of the gaseous N was carried
2
or other
(Co-, Co-, Mn- or Cu/Co-, Cu/Mn-, Co/Mn-) in addition to its
2
distribution inside the crystal lattice. Results demonstrate that
incorporation of Cu/Co-cations leads to a higher reduction extent,
which is more attractive in terms of a potentially higher
out.The particle size (d) was calculated on the basis of the
following correlation:
2-
reducibility directly correlated to a more efficient O carrier
material for Chemical-Loop alcohol Reforming (CLR). Whereas,
according to the results obtained by means of catalytic tests and
CHNS elemental analysis, the addition of Mn hindered carbon
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601605
FULL PAPER
Where:
Yields of each product were calculated by the same method as
78
described in detail in Ref
.
2
SSA – value of the specific surface area, m /g
Acknowledgements
3
–density of the sample, g/m
This work was co-funded through a SINCHEM Grant. SINCHEM
is a Joint Doctorate programme selected under the Erasmus
Mundus Action 1 Programme (FPA 2013-0037).
Raman Spectroscopy. Raman spectra were recorded using a
Bruker spectrometer with a 633 nm He/Ne laser source (filter:
Plaser ÷10; slit 100 μm) equipped with a microscope (50X lens).
Each spectrum was taken with 300 sec exposure time and 1
accumulation.
Literature
1
M. J. N. Isfahani, M. Myndyk, V. Šepelák and J. Amighian,
J. Alloys Compd., 2009, 470, 434–437.
Temperature programmed reduction (TPR) experiments. The
instrument used for TPR analysis was the 1100 Micromeritics
AutoChem II. 2920
samples were first pre-treated at 450 C for 45 min in He flow (30
min/min). Reduction was carried out with 5 % H in the Ar flow
30 ml/min). Temperature program started at 50oC and then the
V 4.05 Chemisorption Analyzer. The
2
V. Šepelák, M. Menzel, I. Bergmann, M. Wiebcke, F.
Krumeich and K. D. Becker, J. Magn. Magn. Mater., 2004,
272-276, 1616–1618.
o
2
(
temperature increased up to 800°C (with ramp 10°C/min) and
maintained there for 40 min.
3
4
H. Hartmut, J. Magn. Magn. Mater., 1988, 74, 193–202.
W. Ponhan and S. Maensiri, Solid State Sci., 2009, 11,
4
79–484.
Mössbauer Spectroscopy. The Mössbauer spectrometer was a
57
homemade apparatus using a 2GBq Co/Rh γ-ray source and a
87
5
6
D. Gao, Z. Shi, Y. Xu, J. Zhang, G. Yang, J. Zhang, X.
Wang and D. Xue, Nanoscale Res. Lett., 2010, 5, 1289–
1
conventional constant acceleration Mössbauer spectrometer .
The solids were samples were diluted in sucrose at about 50%
w
294.
concentration and pressed into discs. All spectra were taken at
room temperature and ambient atmosphere. Isomer shifts were
given with respect to α-Fe. Integrated areas under individual
deconvoluted peaks were used to obtain the relative populations
of different iron species, assuming an equal recoil-free fraction
for all the species. The accuracy of the fits made using pure
V. G. Harris, A. Geiler, Y. Chen, S. D. Yoon, M. Wu, A.
Yang, Z. Chen, P. He, P. V. Parimi, X. Zuo, C. E. Patton, M.
Abe, O. Acher and C. Vittoria, J. Magn. Magn. Mater.,
2009, 321, 2035–2047.
62
Lorentzian line shape, was judged on and χ2 values .
7
8
9
1
1
1
1
M. Pardavi-Horvath, J. Magn. Magn. Mater., 2000, 215,
1
71–183.
CHNS analysis. The carbon contents of the materials after the
E. Auzans, D. Zins, E. Blums and R. Massart, J. Mater.
Sci., 1999, 34, 1253–1260.
each one of the two steps was determined using
a
CHNS/Oxygen automatic elemental analyzer Thermo Scientific
MASS 200R. The samples were homogenized and loaded in
small aluminum capsules and analyzed. For each samples two
analyses were carried out in order to check for the homogeneity
of the solid. Results were given as weight % of C, and the
precision of the measurement was 0.3%.
F. A. Tourinho, R. Franck and R. Massart, J. Mater. Sci.,
1
990, 25, 3249–3254.
0
1
2
3
S. Rana, a. Gallo, R. S. Srivastava and R. D. K. Misra, Acta
Biomater., 2007, 3, 233–242.
Reactivity tests: anaerobic
2 5
C H OH oxidation. Reactivity
C. Sun, J. Lee and M. Zhang, Adv. Drug Deliv. Rev., 2008,
6
experiments were carried out by loading 400 mg of the pelletized
sample (with particles diameter ≈0.25 to 0.6 mm), in the fixed-
bed quartz flow reactor with an internal diameter of 12 mm and a
length of 30 cm; the catalytic bed height was less than 2 mm.
0, 1252–1265.
P. Pradhan, J. Giri, R. Banerjee, J. Bellare and D. Bahadur,
J. Magn. Magn. Mater., 2007, 311, 208–215.
o
The reduction reaction was carried out at 450 C (for 20 min) by
feeding continuously a stream of ethanol vapors (app.15 mol%)
J. a. Toledo-Antonio, N. Nava, M. Martínez and X. Bokhimi,
in N
2
; the latter was also used as a standard; contact time was
Appl. Catal. A Gen., 2002, 234, 137–144.
0
3
.25 s. The products were monitored on-line by an Agilent
000A micro-GC with 3 parallel columns: (A) a PlotQ column,
14
R. J. Rennard and W. L. Kehl, J. Catal., 1971, 21, 282–293.
with He carrier, for the separation of CH
ethane, propane, etc.; (B) a OV1 column, with He carrier, for the
separation of ethanol, CO , acetaldehyde, H O, acetone, acetic
acid, ethyl acetate, diethyl ether, 2-pentanone, toluene, etc.; (C)
a Molecular Sieve 5A column, Ar carrier, for the separation of H
, N , CH , and CO. A Plot U back flash column was installed
in order to avoid CO and H O poisoning in the third column.
4 2 2
, CO , H O, ethylene,
1
5
C. G. Ramankutty, S. Sugunan and B. Thomas, J. Mol.
Catal. A Chem., 2002, 187, 105–117.
2
2
2
,
16
E. Manova, T. Tsoncheva, D. Paneva, I. Mitov, K. Tenchev
and L. Petrov, Appl. Catal. A Gen., 2004, 277, 119–127.
O
2
2
4
2
2
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601605
FULL PAPER
1
1
1
7
8
9
N. Ma, Y. Yue, W. Hua and Z. Gao, Appl. Catal. A Gen.,
003, 251, 39–47.
38
39
40
T. Valdés-Solís, P. Valle-Vigón, S. Álvarez, G. Marbán and
A. B. Fuertes, Catal. Commun., 2007, 8, 2037–2042.
2
S. K. Pardeshi and R. Y. Pawar, Mater. Res. Bull., 2010,
5, 609–615.
a. R. Shyam, R. Dwivedi, V. S. Reddy, K. V. R. Chary and
R. Prasad, Green Chem., 2002, 4, 558–561.
4
M. Florea, M. Alifanti, V. I. Parvulescu, D. Mihaila-
Tarabasanu, L. Diamandescu, M. Feder, C. Negrila and L.
Frunza, Catal. Today, 2009, 141, 361–366.
G. Brahmachari, S. Laskar and P. Barik, RSC Adv., 2013,
3, 14245–14253.
4
4
4
1
2
3
Y. Wang, R. Cheng, Z. Wen and L. Zhao, Eur. J. Inorg.
Chem., 2011, 2011, 2942–2947.
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
C. U. Aniz and T. D. R. Nair, Open J. Phys. Chem., 2011,
0
1, 124–130.
R. Wu and J. Qu, J. Chem. Technol. Biotechnol., 2005, 80,
20–27.
C. Nordhei, K. Mathisen, I. Bezverkhyy and D. Nicholson, J.
Phys. Chem. C, 2008, 112, 6531–6537.
C. Alvani, G. Ennas, a Labarbera, G. Marongiu, F. Padella
and F. Varsano, Int. J. Hydrogen Energy, 2005, 30, 1407–
1411.
N. Rezlescu, E. Rezlescu, P. D. Popa, C. Doroftei and M.
Ignat, Appl. Catal. B Environ., 2014, 158-159, 70–75.
A. Urdă, A. Herraïz, Á. Rédey and I.-C. Marcu, Catal.
Commun., 2009, 10, 1651–1655.
44
C. Alvani, M. Bellusci, a. La Barbera, F. Padella, L.
Seralessandri and F. Varsano, J. Sol. Energy Eng., 2010,
1
32, 031001.
H. Xi, X. Hou, Y. Liu, S. Qing and Z. Gao, Angew. Chem.
Int. Ed. Engl., 2014, 53, 1–5.
45
46
47
48
N. Ballarini, F. Cavani, S. Passeri, L. Pesaresi, A. F. Lee
and K. Wilson, Appl. Catal. A Gen., 2009, 366, 184–192.
P. Lahiri and S. K. Sengupta, Can. J. Chem., 1991, 69, 33–
3
6.
K. K. Senapati, C. Borgohain and P. Phukan, J. Mol. Catal.
A Chem., 2011, 339, 24–31.
A. I. Onuchukwu, J. Chem. Soc. Faraday Trans. 1, 1984,
0, 1447.
8
P. Zhang, Q. Han, M. Fan and P. Jiang, Appl. Surf. Sci.,
2
014, 317, 1125–1130.
S. Briceno, H. Del Castillo, Sagredo, W. Bramer-Escamilla
and P. Silva, Appl. Surf. Sci., 2012, 263, 100–103.
B. Naik, S. Hazra, D. Desagani, B. K. Ghosh, M. K. Patra,
S. R. Vadera and N. N. Ghosh, RSC Adv., 2015, 5, 40193–
4
0198.
A. Khan and P. G. Smirniotis, J. Mol. Catal. A Chem., 2008,
2
80, 43–51.
4
5
5
9
0
1
J. Tong, X. Cai, H. Wang and C. Xia, J. Sol-Gel Sci.
Technol., 2013, 66, 452–459.
A. Khan, P. Chen, P. Boolchand and P. Smirniotis, J.
Catal., 2008, 253, 91–104.
J. Tong, L. Bo, X. Cai, H. Wang, Q. Zhang and L. Su, Ind.
Eng. Chem. Res., 2014, 53, 10294–10300.
M. Banerjee and a Rai, J. Nanosci. Nanotechnol., 2007, 7,
1
990–1993.
M. Kooti and M. Afshari, Sci. Iran., 2012, 19, 1991–1995.
Y. Mastai, Ed., Advances in Crystallization Processes,
InTech, 2012.
52
53
54
55
56
57
M. Nasrollahzadeh, M. Bagherzadeh and H. Karimi, J.
Colloid Interface Sci., 2016, 465, 271–8.
K. Zipare, J. Dhumal, S. Bandgar, V. Mathe and G.
Shahane, J. Nanosci. Nanoeng., 2015, 1, 178–182.
Y. Zhang, Z. Li, W. Sun and C. Xia, Catal. Commun., 2008,
1
0, 237–242.
J. Chen, W. Wen, L. Kong, S. Tian, F. Ding and Y. Xiong,
Ind. Eng. Chem. Res., 2014, 6297–6306.
W. F. Shangguan, Y. Teraoka and S. Kagawa, Appl. Catal.
B Environ., 1998, 16, 149–154.
Q. Song, Y. Ding, Z. L. Wang and Z. J. Zhang,
Chem.Mater.., 2007, 19, 4633–4638.
A. a. Farghali, M. H. Khedr and A. a. Abdel Khalek, J.
Mater. Process. Technol., 2007, 181, 81–87.
C.-S. Hwang and N.-C. Wang, Mater. Chem. Phys., 2004,
8
8, 258–263.
R. Parella, A. Kumar and S. A. Babu, Tetrahedron Lett.,
2
013, 54, 1738–1742.
I. L. Stolyarchuk, I. V Vasylenko, Y. I. Pyatnitsky, P. E.
Strizhak and P. Nauky, 2013, 49, 172–177.
M. Estrella, L. Barrio, G. Zhou, X. Wang, Q. Wang, W.
Wen, J. C. Hanson, A. I. Frenkel and A. Rodriguez, J. Phys.
C, 2009, 113, 14411–14417.
A. Miyakoshi, A. Ueno and M. Ichikawa, 2001, 219, 249–
2
58.
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601605
FULL PAPER
5
5
8
9
K. Faungnawakij, N. Shimoda, T. Fukunaga, R. Kikuchi and
K. Eguchi, Appl. Catal. B Environ., 2009, 92, 341–350.
73
74
a. Pineau, N. Kanari and I. Gaballah, Thermochim. Acta,
2006, 447, 89–100.
K. Faungnawakij, Y. Tanaka, N. Shimoda, T. Fukunaga, R.
Kikuchi and K. Eguchi, Appl. Catal. B Environ., 2007, 74,
a. Pineau, N. Kanari and I. Gaballah, Thermochim. Acta,
2007, 456, 75–88.
1
44–151.
7
5
W. K. Jozwiak, E. Kaczmarek, T. P. Maniecki, W. Ignaczak
6
6
0
1
R. Ramanathan and S. Sugunan, Catal. Commun., 2007, 8,
521–1526.
and W. Maniukiewicz, Appl. Catal. A Gen., 2007, 326, 17–
27.
1
V. Crocella, F. Cavani, G. Cerrato, S. Cocchi, M. Comito,
G. Magnacca and C. Morterra, J. Phys. Chem. C, 2012,
76
B. Hou, H. Zhang, H. Li and Q. Zhu, Chinese J. Chem.
Eng., 2012, 20, 10–17.
1
16, 14998–15009.
7
7
7
8
G. K. Reddy, P. Boolchand and P. G. Smirniotis, J. Phys.
Chem. C, 2012, 116, 11019–11031.
6
6
6
6
6
2
3
4
5
6
S. Cocchi, M. Mari, F. Cavani and J.-M. M. Millet, Appl.
Catal. B Environ., 2014, 152-153, 250–261.
C. Trevisanut, O. Vozniuk, M. Mari, S. Yurena, A. Urrea, C.
Lorentz, J.-M. Millet and F. Cavani, Top. Catal., 2016, 59,
1600–1613.
C. Trevisanut, F. Bosselet, F. Cavani and J. M. M. Millet,
Catal. Sci. Technol., 2015, 5, 1280–1289.
J. V. Ochoa, C. Trevisanut, J.-M. M. Millet, G. Busca and F.
Cavani, J. Phys. Chem. C, 2013, 117, 23908–23918.
79
80
81
A. Khan, P. Chen, P. Boolchand and P. Smirniotis, J.
Catal., 2008, 253, 91–104.
C. Trevisanut, M. Mari, J.-M. M. Millet and F. Cavani, Int. J.
Hydrogen Energy, 2015, 40, 5264–5271.
K. Kandel, U. Chaudhary, N. C. Nelson and I. I. Slowing,
ACS Catal., 2015, 5, 6719–6723.
O. Vozniuk, S. Agnoli, L. Artiglia, A. Vassoi, N. Tanchoux,
F. Di Renzo, G. Granozzi and F. Cavani, Green Chem.,
J. Alan and D. M. Brian, Tempature-Programmed
Reduction for Solid Materials Characterization, 1986.
2
015.
8
2
X.-W. Liu, S. Zhao, Y. Meng, Q. Peng, A. K. Dearden, C.-F.
Huo, Y. Yang, Y.-W. Li and X.-D. Wen, Sci. Rep., 2016, 6,
26184.
6
6
7
8
J. P. Chen and C. M. Sorensen, Phys. Rev. B, 1996, 54,
288–9296.
9
C. S. S. R. Kumar, Ed., Raman Spectroscopy for
Nanomaterials Characterization, Springer Berlin
Heidelberg, 2012.
83
G. B. Raupp and W. N. Delgass, J. Catal., 1979, 58, 348–
360.
8
4
J. W. Niemantsverdriet, a. M. Van der Kraan, W. L. Van
Dijk and H. S. Van der Baan, J. Phys. Chem., 1980, 84,
3363–3370.
6
7
7
7
9
0
1
2
P. Chandramohan, M. P. Srinivasan, S. Velmurugan and S.
V. Narasimhan, J. Solid State Chem., 2011, 184, 89–96.
V. D’Ippolito, G. B. Andreozzi, D. Bersani and P. P. Lottici,
J. Raman Spectrosc., 2015, 46, 1255–1264.
85
D. P. Johnson, Solid State Commun., 1969, 7, 1785–1788.
8
8
6
7
C. P. Rodrigues, P. C. Zonetti, C. G. Silva, A. B. Gaspar
and L. G. Appel, Appl. Catal. A Gen., 2013, 458, 111–118.
N. Z. Lazarevic, Z, Jovalekic, C, Milutinovic, A, Romcevic,
M.J, Romcevic, Acta Phys. Pol. A, 2012, 121, 682–686.
J. M. Millet, C. Virely, M. Forissier, E. Bussiere and J. C.
Vedrine, Hyperfine Interact., 1989, 46, 619–628.
J. J. Graves, P.R., Johnston, C., Campaniello, Mater. Res.
Bull., 1988, 23, 1651–1660.
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601605
FULL PAPER
The reduction profile of spinel-type
mixed ferrites is investigated in the
presence of ethanol as the
reactant/reducing agent. The model
developed for spinel reduction
accounts for the reactivity behavior
shown under anaerobic conditions,
which is greatly affected by ferrite
composition.
Olena Vozniuk, Cristian Bazzo,
Stefania Albonetti, Nathalie
Tanchoux, Françoise Bosselet, Jean-
Marc Millet, Francesco Di Renzo*
and Fabrizio Cavani*
Page 1 – Page 12
Structural changes of
binary/ternary spinel oxides
during ethanol anaerobic
decomposition
This article is protected by copyright. All rights reserved.