Increased nickel exsolution from LaFe0.8Ni0.2O3 perovskite-derived CO2 methanation catalysts through strontium doping

Article abstract of DOI:10.1016/j.apcata.2019.117328

Perovskite-derived Ni catalysts offer the remarkable benefit of redox stability that allows their regeneration after deactivation through poisoning or Ni particle growth. Here, the catalytic activity of LaFe_0.8Ni_0.2O₃ towards CO₂ methanation was improved by increasing Ni reducibility and segregation to the perovskite surface through partial substitution of La by Sr (La_1-xSr_xFe_0.8Ni_0.2O_3-δ, 0 ≤ × ≤ 0.1). Temperature programmed reduction, X-ray diffraction, scanning electron microscopy and X-ray absorption spectroscopy were used to characterize the materials, their stability against severe reduction at high temperatures and to quantify Ni segregation. It is shown that Ni reducibility was significantly increased (up to 50%) upon introduction of Sr, because more Ni was exsoluted during catalyst pre-reduction. Nickel reincorporation into the perovskite-type oxide lattice during reoxidation was not affected at these Sr levels and complete redox stability could be demonstrated for all compositions investigated in this work.

Full text of DOI:10.1016/j.apcata.2019.117328

Journal Pre-proof
Increased nickel exsolution from LaFeNiO perovskite-derived CO2
3
methanation catalysts through strontium doping
Patrick Steiger, Oliver Kro¨cher, Davide Ferri
PII:
S0926-860X(19)30483-1
DOI:
https://doi.org/10.1016/j.apcata.2019.117328
APCATA 117328
Reference:
To appear in:
Applied Catalysis A, General
Received Date:
Revised Date:
Accepted Date:
13 September 2019
22 October 2019
23 October 2019
Please cite this article as: Steiger P, Kro¨cher O, Ferri D, Increased nickel exsolution from
LaFeNiO perovskite-derived CO methanation catalysts through strontium doping, Applied
3
2
Catalysis A, General (2019), doi: https://doi.org/10.1016/j.apcata.2019.117328
This is a PDF file of an article that has undergone enhancements after acceptance, such as
the addition of a cover page and metadata, and formatting for readability, but it is not yet the
definitive version of record. This version will undergo additional copyediting, typesetting and
review before it is published in its final form, but we are providing this version to give early
visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal
pertain.
©
2019 Published by Elsevier.

Increased nickel exsolution from LaFeNiO
3
perovskite-derived
CO methanation catalysts through strontium doping
2
Patrick Steiger,a,b Oliver Kröchera,b and Davide Ferria*
a
Paul Scherrer Institut, Forschungsstrasse 111, CH-5232 Villigen, Switzerland
École polytechnique fédérale de Lausanne (EPFL), Institute of Chemical Sciences
b
and Engineering, CH-1015 Lausanne, Switzerland
Corresponding author
*
E-mail: davide.ferri@psi.ch
Phone: +41 56 310 27 81
Graphical abstract
Highlights




A-site of LaFe0.8Ni0.2
O prepared by the amorphous citric acid method is modified by Sr
3
A-site Sr promotes the segregation of Ni
The self-regeneration of Ni in LaFe0.8Ni0.2
O
3
is maintained upon substitution of La by Sr
production during
An optimum value of Sr substitution (5 mol%) provides maximum CH
CO hydrogenation
4
2
1

Abstract
Perovskite-derived Ni catalysts offer the remarkable benefit of redox stability that
allows their regeneration after deactivation through poisoning or Ni particle growth.
Here, the catalytic activity of LaFe0.8Ni0.2
by increasing Ni reducibility and segregation to the perovskite surface through partial
substitution of La by Sr (La1-xSr 3-δ, 0 ≤ x ≤ 0.1). Temperature programmed
O
3
towards CO
2
methanation was improved
x
Fe0.8Ni0.2O
reduction, X-ray diffraction, scanning electron microscopy and X-ray absorption
spectroscopy were used to characterize the materials, their stability against severe
reduction at high temperatures and to quantify Ni segregation. It is shown that Ni
reducibility was significantly increased (up to 50%) upon introduction of Sr, because
more Ni was exsoluted during catalyst pre-reduction. Nickel reincorporation into the
perovskite-type oxide lattice during reoxidation was not affected at these Sr levels and
complete redox stability could be demonstrated for all compositions investigated in this
work.
Keywords: CO2 hydrogenation; perovskite-type oxide; Nickel; LaFe0.8Ni0.2O3; self-
regeneration
Introduction
Metal catalysts derived from perovskite-type metal oxides (PMO) have attracted great
attention due to their high stability towards catalyst deactivation through metal particle
sintering and coke deposition. In these materials catalytically active metals can be
preferentially reduced and segregated from the host PMO lattice to produce the active
material [1-3]. It was demonstrated that this metal segregation results in strong metal
particle-support interactions due to the segregation process compared to a metal
deposition type process such as wet impregnation and hence high sintering stability
2

and resistance towards coke formation at the particle-support interface [4].
Furthermore, application of PMO potentially allows the exploitation of their structural
reversibility that is the reversible segregation of metals from the PMO lattice during
reduction and reincorporation of these elements under oxidizing atmospheres and at
sufficiently high temperatures [1, 5]. This property induces catalyst redox stability and
can be exploited for catalyst regeneration [6, 7]. In recent years this property has been
investigated for Ni in PMOs such as LaFe1-xNi
in industrially relevant catalytic reactions including hydrogenation of carbon oxides [10-
5], water gas shift [16], Fischer-Tropsch reaction [17], dry and wet reforming of
x
O
3
(LFNO) [7-9]. Nickel is widely applied
1
hydrocarbons [16, 18-20], dry reforming of methane and carbon dioxide [21] and the
production of methanol from methane [22]. In these processes nickel is used at a large
scale because of its high activity and relatively low cost compared to precious metals.
However, Ni catalysts can deactivate severely as a result of the deposition of carbon
[
19, 23-25], which requires oxidative regeneration at high temperatures [26]. Despite
the advantages in terms of catalyst redox stability induced by the reversible
segregation of Ni from LFNO perovskite-type catalysts, the major drawback of this
approach is probably the difficulty to obtain complete Ni reduction and segregation
from the bulk PMO. This results in the inefficient use of the active metal compared to
conventional supported Ni-based catalysts [3, 7, 8]. Because the mixed metal oxide
produces selectively CO and selectivity to CH
Ni dispersed on the LFNO surface [7], improving the extent of Ni reduction and
segregation in PMO materials is anticipated to be beneficial to achieve higher CH
4
is highly dependent on the amount of
4
yields. Since Ni reducibility is highly dependent on the composition of the host PMO
lattice [27], it is likely that tuning the PMO composition through doping will affect both
Ni reduction and segregation. Substitution of lanthanum by strontium in LaFeO
3
3
was

demonstrated to promote the formation of oxygen vacancies, which increase oxygen
conductivity through the bulk of the PMO lattice, resulting in increased bulk reducibility
[
28-31]. This effect may be exploited to increase the catalytic activity of PMOs towards
oxidation reactions, which rely on oxygen release from the lattice [32-34]. Its excellent
ionic conductivity makes La1-xSr FeO3-δ a suitable cathode material for solid oxide fuel
x
cells [30]. Deng et al.[35] reported increased Ni reducibility in LFNO upon doping the
A-site by Ca, Sr, Mg, Ba, Pr and Ce. Ca, Sr, Mg and Ba adopt an oxidation state of
+2, thus reducing the nominal charge at the A-site, which is compensated by either
oxygen vacancy formation or lower oxidation states of the B-site elements. Pr on the
4+
other hand also adopts the +3 oxidation state of La, whereas introduction of Ce leads
to the formation of additional interstitial lattice oxygen as well as lower oxidation states
of B-site elements [36]. Doping by Pr appeared to increase perovskite reducibility the
most; however, more commonly a similar increase in reducibility is achieved by A-site
doping with divalent alkaline earth elements. Taylor et al.[37] used density functional
theory calculations to determine the stability of solid solutions for A-site doping by Mg,
Ca, Sr and Ba and demonstrated that solution energies decreased from Mg to Sr along
group II before increasing again for Ba. Solution energies could be directly linked to
2
+
ionic size and were found to be the lowest for Sr (r
i
= 1.44 Å) due to its similar ionic
= 1.36 Å) [38]. These results also confirm previous investigations on
the solubility of A-site dopants in LaFeO [39].
Because of the good solubility properties of Sr in LaFeO
behavior in PMOs is well-understood, this work focuses on the use of Sr doping to
increase Ni reduction and segregation from La1-xSr 3-δ (x ≤ 0.1). Despite the
3+
radius to La (r
i
3
3
as well as the fact that its
x
Fe0.8Ni0.2O
extensive studies on the effect of A-site doping on oxygen vacancy formation, ionic
conductivity and perovskite reducibility, it is not clear whether increased Ni reduction
4

accordingly results in increased Ni segregation and what effect the destabilization of
the PMO host has on potential Ni reincorporation. These issues are addressed in this
work.
Experimental
LaFe0.8Ni0.2
O
3
(LFNO),
3-δ (LSFNO-10Sr) were prepared via the amorphous citrate
process [40]. La(NO ∙6H O (Sigma-Aldrich, 99.999% trace metals basis),
Fe(NO ∙9H O (Sigma-Aldrich, ≥99.95% trace metals basis) and Ni(NO ∙6H
Sigma-Aldrich, 99.999% trace metals basis) were each dissolved in water and
La0.95Sr0.05Fe0.8Ni0.2
O
3-δ
(LSFNO-5Sr)
and
La0.9Sr0.1Fe0.8Ni0.2
O
3
)
3
2
3
)
3
2
3
)
2
2
O
(
thoroughly mixed before being added to an aqueous solution of citric acid (Sigma-
Aldrich, ACS reagent, ≥99.5%). The overall molar ratio of metal nitrates to citric acid
was 1:1.05. The precursor solution was stirred at 70°C for 60 min before drying under
low pressure at 70°C for 12 h. The resultant precursor foam was crushed to a fine
-
1
powder and then subjected to calcination at 800°C for 5 h (5°C·min to 200°C; then,
-
1
1
0°C·min to 800°C) to obtain the calcined sample.
The specific surface area (SSA) of the calcined powder was determined from N
2
-
adsorption isotherms at -196°C according to the Brunauer-Emmet-Teller (BET) model
using a Quantachrome Autosorb I instrument. The samples were treated in vacuum at
3
00°C for 2 h prior to SSA measurements.
The crystal structure of the powder catalysts was investigated by powder X-ray
diffraction (XRD) using a Bruker D8 Advance diffractometer equipped with Ni-filtered
Cu Kα-radiation, variable slits and an energy-sensitive line detector (LynxEye).
Diffractograms were collected at an acquisition time of 4 s and a step size of Δ2θ =
5

0
.03° between 2θ 15° and 80°. The Scherrer equation [41] was used to estimate the
perovskite crystallite size (d) using peak parameters determined for the (121) and (202)
reflections at 32.3° and 46.4° 2θ, respectively. Mean values calculated from results
obtained for both reflections as well as standard errors are reported in the text.
In situ XRD measurements were conducted on the same instrument using a reactor
chamber (XRK 900, Anton Paar). Reactive gases (10 vol% H
2
/N
2
and 20 vol% O
2 2
/N )
were dosed to the chamber (100 mL·min-1 at STP) using mass flow controllers
(
Bronkhorst) while the sample temperature was maintained at 600°C and monitored
using a K-type thermocouple.
Temperature programmed reduction (TPR) experiments were conducted using a
TPDRO-1100 (ThermoElectron) instrument equipped with mass flow controllers and a
thermal conductivity detector. The samples (ca. 100 mg) were loaded into the quartz
reactor tube and heated under constant flow of 20 vol% O
2
/Ar to 500°C before cooling
-
1
to room temperature. TPRs were recorded in 10 vol% H /Ar (20 mL·min at STP) and
2
-
1
at a heating rate of 5°C·min . The reoxidation temperature at which Ni is reversibly
reincorporated into the perovskite lattice was estimated by TPR redox experiments. A
TPR profile was recorded on the calcined sample up to 600°C followed by isothermal
reduction for 1 h at the same temperature. After cooling to 25°C in Ar, the sample was
heated to 400°C in 20 vol% O
2
/N and reoxidized at this temperature for 2 h. The
2
sample was again cooled in Ar to 25°C before starting the second TPR. Such TPR-
reduction-reoxidation-TPR cycles were repeated four times with increasing the
reoxidation temperature (400°C, 500°C, 600°C and 700°C). The heating rate during
-
1
reoxidation and cooling after all experiments was 10°C·min .
6

Changes in the microstructure of the samples were studied using scanning electron
microscopy (field emission SEM, Zeiss ULTRA 55) on calcined samples as well as
samples subjected to five catalytic tests alternated to five redox cycles. Nickel particle
size distributions were obtained on the samples after catalytic testing using the ImageJ
open source software package [42] by measuring and counting at least 650 particles.
2
The minimum area analyzed on each sample was 2.5 μm . The particle density was
obtained by dividing the number of counted particles by the analyzed area. Particle
density is used here instead of dispersion because we were not able to measure Ni
dispersion by chemisorption likely as a result of the mean Ni particle size obtained
after reduction, the low fraction of reduced Ni and the low specific surface area of the
materials.
Catalytic activity tests towards CO
setup in a quartz reactor of plug flow geometry (6 mm ID). Mass flow controllers
Brooks) were used to dose the reactant gases and a K-type thermocouple, which was
2
methanation were carried out on a homemade test
(
placed in the middle of the catalyst bed, was used to monitor catalyst bed temperature.
To avoid back pressure, all calcined powders were pelletized (4 MPa), crushed and
sieved to 100-150 μm before use. The catalyst (100 mg) was diluted with cordierite
powder (200 mg, 75-100 μm) to achieve a thoroughly mixed catalyst bed of ca. 15 mm
in length. The reactant gas containing 8 vol% CO
2
, 32 vol% H
2
(rest Ar) was fed at a
-
1
total flow rate of 50 mL·min at STP resulting in a weight hourly space velocity (WHSV)
3
-1 -1
of 30,000 cm ·g ·h . Prior to the catalytic tests, the sample was subjected to pre-
reduction (20 vol% H /Ar, 600°C, 1 h). The experiment was then started at 600°C
2
-
1
under reaction conditions and the reactor cooled to 300°C at 5°C·min while recording
catalytic activity. The reactor was then heated to 650°C in constant flow of Ar before
conducting catalyst reoxidation (20 vol% O
2
/Ar, 650°C, 2 h) and subsequent reduction
7

(
20 vol% H , 600°C, 1 h). After the redox treatment catalyst activity was again recorded
2
between 600°C and 300°C. This sequence was repeated four times. All gas lines were
made of stainless steel and heated to 120°C to avoid condensation. A Pfeiffer
OmniStar GSD 320 quadrupole mass spectrometer equipped with a heated stainless
steel capillary was used for compositional analysis of the exhaust gas. CO
2
conversion
(
X
CO2) and CH
4
yield (YCH4) were calculated using equations 1 and 2, respectively,
[
퐶푂2]푖푛−[퐶푂2]표푢푡
푋퐶푂2
=
× 100%
(1)
[
퐶푂2]푖푛
[
퐶퐻4]표푢푡
푌
=
× 100%
퐶푂2]푖푛
(2)
퐶퐻
4
[
where [CO
concentrations of CH
Ni K-edge (8.333 keV) X-ray absorption spectroscopy (XAS) data were collected on
calcined and pre-reduced samples (10 vol% H /Ar, 600°C, 1 h) at the beamline
2
]in is the initial concentration of CO
2
and [CH
4
]out and [CO
2
]out are the
4
and CO
2
at the reactor outlet, respectively.
2
SuperXAS of the Swiss Light Source (SLS, Villigen, Switzerland). The beam current
and beam energy within the storage ring were 400 mA and 2.4 GeV, respectively. XAS
spectra were collected in fluorescence mode using an ionization chamber filled with
N
2
to measure incoming beam intensity and a Passivated Implanted Planar Silicon
(
PIPS) detector mounted at 90° to the incoming X-ray beam. Fluorescence mode was
selected due to the high absorption cross section of the host perovskite at the Ni K-
edge. The X-ray beam was collimated using a Si coated mirror, which was also used
for the reduction of higher harmonic contributions while the X-ray energy was
continuously scanned around the Ni absorption edge using a Si(111) channel-cut
monochromator [43][43]. The beam was focused to a spot size of 1.0 mm x 0.2 mm
(
height x width) using a Rh-coated toroidal mirror. Measurements were performed on
8

pelletized samples after pretreatment and cooling in Ar. Reference spectra of metallic
0
2+
Ni (Ni ) and NiO (Ni ) were collected in transmission mode using Ni foil and diluted
NiO (Sigma-Aldrich, 99.99%), respectively. The Demeter software package (version
0
.9.24) [44] was used for data reduction and linear combination fitting within the energy
range 8.313 - 8.363 keV.
Results and discussion
X-ray diffraction (XRD) patterns of the calcined LaFe0.8Ni0.2
O (LFNO; tolerance factor
3
t= 1.04), La0.95Sr0.05Fe0.8Ni0.2
O
3-δ (LSFNO-5Sr; t= 0.808) and La0.9Sr0.1Fe0.8Ni0.2
O
3-δ
(
LSFNO-10Sr; t= 0.808) are displayed in Figure 1a. All powder samples exhibited
similar patterns corresponding to the orthorhombic perovskite-type unit cell also
observed for undoped LaFeO [8, 27].
3
All reflections slightly shifted to higher angles, which reflects a decrease in unit cell
volume with increasing Sr content. This is best represented by the small difference in
the position of the main (121) reflection, which shifted from 32.33° to 32.39° 2θ with
increasing Sr content (Figure 1b). Because the ionic radius of twelve-fold coordinated
2+
3+
Sr is larger than that of La in the same coordination (1.44 Å vs. 1.36 Å) [38] unit
cell contraction is not likely to be caused directly by the introduction of A-site cations
of different sizes. Charge compensation could be also achieved through an increase
in oxidation state of the B-site elements, which would however lead to increased B-
site ion radii and eventually to lattice expansion. It is more likely that the observed unit
cell contraction is a result of the more prominent effect of the formation of oxygen
vacancies induced by the decreased charge at the A-site to maintain charge neutrality.
The absence of further phases, for example Sr-containing and the shift of the PMO
9

reflections with increasing Sr content confirm that the samples were phase pure.
Besides shifts, also significant peak broadening was observed for Sr-containing
samples. The crystallite sizes estimated using the Scherrer equation [41] were 41 ± 1
nm, 22 ± 2 nm and 19 ± 2 nm for LFNO, LSFNO-5Sr and LSFNO-10Sr, respectively.
The decrease in crystallite size with increasing Sr content was accompanied by an
2
-1
2
-1
increase in specific surface area from 6.2 ± 0.5 m ·g (LFNO) to 7.5 ± 0.5 m ·g
2
-1
(
LSFNO-5Sr) and to 9.4 ± 0.5 m ·g (LSFNO-10Sr).
Reducibility and stability of the perovskite host during high-temperature reduction was
assessed by recording in situ XRD patterns during isothermal reduction experiments
(
10 vol% H
experiments are shown in Figure 2. While LFNO did no exhibit phase changes due to
reduction and a La phase did not form over 40 h of the experiment (Figure 2a),
2
/N
2
, 600°C) using a lab source. Contour plots summarizing these
2
O
3
both Sr-containing samples showed appearance and growth of the characteristic
reflections of a metallic Ni/Fe phase, starting after 7 h (LSFNO-5Sr, Figure 2b) and 3
h (LSFNO-10Sr, Figure 2c) into the experiment. This is a strong indication that the
increasing Sr content increasingly destabilizes the LFNO perovskite-type oxide, thus
promoting the reduction of B-site elements. Whether this reflects increased Ni
reducibility and segregation and thus catalytic activity has to be demonstrated.
The element specific X-ray absorption near edge structure (XANES) analysis of the Ni
K-edge (8.333 keV) was used to investigate the effect of Sr on Ni reduction. The
spectra of both calcined and reduced (10 vol% H
2
/Ar, 600°C, 1 h) LFNO, LSFNO-5Sr
0
and LSFNO-10Sr are presented together with the spectra of Ni (Ni foil) and NiO
references in Figure 3. The spectra of all calcined samples were very similar, indicating
both similar oxidation states as well as coordination geometry. This is not surprising
since Sr substitution of La was kept at a low level and Ni K-edge XANES may be only
1
0

significantly influenced for higher degrees of A-site substitution. The spectra can be
assigned to Ni in octahedral coordination (Nin+oct, n > 2).8
The spectra of the reduced samples exhibited decreased whiteline intensity and
resembled a linear combination of the spectrum of Nin+oct, obtained for calcined LFNO,
0
and of the Ni reference. This indicated partial reduction of Ni in all samples. Linear
0
combination fit (LCF) was used to quantify the contribution of Ni to the spectra of the
reduced materials. Corresponding fitted spectra as well as contributions of individual
0
spectra to the final fit are displayed in Figure S1. An increased contribution of Ni was
0
observed with increasing Sr content. Ni concentration increased from 29 ± 3% for
LFNO to 31 ± 3% for LSFNO-5Sr and to 38 ± 3% for LSFNO-10Sr. The increase in
the extent of Ni reducibility may be a direct consequence of the decreased stability of
the host perovskite towards reduction of the B-site elements as observed in Figure 2.
Analysis of the extended X-ray absorption fine structure (EXAFS) spectra was used to
follow Ni-Ni bond formation during reduction, which is indicative of Ni particle formation.
3
0
Plots of k -weighted χ(k) functions of all samples and of Ni and NiO references are
presented in Figure S2. The corresponding Fourier transforms are displayed in Figure
4
. All reduced samples presented the contribution at 2.1 Å that is also present in the
0
Ni reference and is characteristic of the first Ni coordination shell in Ni metal, thus
providing evidence for Ni-Ni bond and Ni particle formation. While this feature was
weak in reduced LFNO (Figure 4b), its intensity increased in reduced LSFNO-5Sr
(
Figure 4c) and LSFNO-10Sr (Figure 4d). This is readily interpreted with the increase
in the fraction of metallic Ni that is present in the form of Ni particles after reduction
with increasing Sr content. This effect may be beneficial for catalyst activity because
Sr-containing samples could exhibit higher Ni metal level at the perovskite surface.
However, catalyst activity towards CH
4
formation in CO methanation is eventually
2
1
1

determined by Ni metal specific surface area and is thus also dependent on Ni particle
size.
Despite the apparent stability of all samples against complete reduction to the single
metal and metal oxides components for at least 3 h (Figure 2), XRD may be not
sensitive to the formation of only few and small crystallites of Ni metal or Ni/Fe alloy
as well as the La
2
O phase especially at the beginning of the reduction. The potential
3
partial decomposition, which was likely present on Sr-containing samples, may hinder
complete Ni reincorporation and the structural reversibility over repeated regeneration
cycles. It was shown previously that LFNO calcined at 700°C exhibited complete
structural reversibility after reoxidation (20 vol% O
2
/N , 650°C, 2 h) [7, 8].
2
TPR was used to determine Ni speciation on the samples after redox cycles. Ni
speciation through TPR is feasible as the reduction temperature of oxidized Ni species
differs depending on the Ni coordination environment. Under the selected TPR
conditions NiO species reduce at higher temperatures (ca. 300-400°C) than Ni in the
coordination environment of the B-site of PMO (ca. 150-300°C, Nin+oct) [27]. Figure 5
shows the results of TPR-redox cycles with increasing reoxidation temperatures.
Whenever Ni was incorporated within the perovskite lattice in the form of Nin+oct (n > 2)
during reoxidation, the subsequent TPR exhibited only a low temperature reduction
event starting at 150°C and peaking at 273°C for LFNO (Figure 5a), 271°C for LSFNO-
5
Sr (Figure 5b) and 252°C for LSFNO-10Sr (Figure 5c). These values are lower than
the temperature required to reduce bulk NiO (Figure 5a) and NiO deposited on LaFeO
3
by impregnation [8]. Whenever the reoxidation temperature between the TPR cycles
was not high enough to completely restore all Nin+oct, Ni was oxidized to NiO at the
surface of the perovskite, thus producing an additional reduction feature at higher
temperature in the following TPR run. The intensity of the feature distinctive of NiO
1
2

decreased with increasing oxidation temperature from 400°C to 700°C for all
compositions and vanished at 700°C. This suggests that complete Ni reincorporation
occurs at temperatures between 600°C and 700°C in all samples.
All samples were further subjected to five consecutive TPR-TPO redox cycles at
constant reoxidation temperature (650°C). This allowed to refine the temperature of Ni
reincorporation and to emphasize subtle differences between samples. Figure 6
shows the TPR profiles collected during these redox experiments. Appearance of a
reduction event at ca. 330°C due to reduction of NiO could be regarded as a sign of
structural irreversibility. LFNO did not display such feature in all redox cycles (Figure
6
a). While it was observed in LSFNO-5Sr (Figure 6b) and LSFNO-10Sr (Figure 6c),
this feature did not grow indicating that the redox experiment did not affect further the
existing fraction of NiO. On the contrary, redox cycles involving reoxidation at 600°C
for 2 h evoked the growth of such a feature in all three samples (Figure 6d, LFNO;
Figure 6e, LFSNO-5Sr; Figure 6f, LSFNO-10Sr) and thus an increasing fraction of NiO.
While this phenomenon was hardly visible in LFNO, it was significant in the LSFNO-
type materials suggesting that Ni reincorporation was more difficult on the latter. The
decreased stability of the perovskite lattice due to Sr introduction likely caused a
reduction of the thermodynamic driving force for Ni reincorporation. Hence, higher
temperatures are required to restore Nin+oct in the presence of Sr.
CO
2
methanation activity and CH
4
yield were measured on pre-reduced samples (20
vol% H
2
/Ar, 600°C, 1 h) after each of five redox cycles. A redox cycle was composed
of a reduction segment, reaction and an oxidation segment. This sequence was
repeated five times. Since the PMO surface is highly selective towards the reverse
water gas shift reaction and thus to CO production [8], CH
anticipated to occur only on Ni metal surfaces. Therefore, CH
4
production can be
4
yield can be used to
1
3

assess changes in Ni specific surface area along the redox cycles, which is dependent
on both the quantity of segregated Ni as well as the Ni particle size. Figure 7 shows
the CO
0Sr (Figure 7c) measured after each redox cycle. The CH
are low likely as a result of a combination of low specific surface area and low amount
of exposed Ni. On all samples CO conversion improved significantly with increasing
number of redox cycles as indicated by the shift of the conversion curves towards
lower reaction temperatures. It is also apparent that while CO conversion appeared
to have reached the maximum already after a single redox cycle on LSFNO-10Sr,
LSFNO-5Sr showed an increase in activity towards CO conversion after three redox
cycles while catalytic activity of LFNO stabilized only after four redox cycles. The
corresponding CH yields reflect the same behavior (Figure 7d-f) indicating that the
accessible Ni surface area stabilized only after four and three redox cycles on LFNO
and LSFNO-5Sr, respectively. Furthermore, LSFNO-5Sr exhibited highest CH yields
of around 20% at 470°C, whereas maximum CH yield was 16% (at 480°C) and 13%
480°C) for LFNO and LSFNO-10Sr, respectively.
2
conversion data of LFNO (Figure 7a), LSFNO-5Sr (Figure 7b) and LSFNO-
1
4
yields on these materials
2
2
2
4
4
4
(
A plausible reason for the improved activity of LSFNO-5Sr compared to LFNO and
LSFNO-10Sr can be found by comparison of the scanning electron micrographs (SEM)
of the calcined materials (LFNO, Figure 8a; LSFNO-5Sr, Figure 8d; LSFNO-10Sr,
Figure 8g) and of the samples after the catalytic tests of Figure 7 (LFNO, Figure 8b;
LSFNO-5Sr, Figure 8e; LSFNO-10Sr, Figure 8h). In the calcined state, none of the
samples exhibited features other than what could be assigned to the perovskite phase.
However, on the reduced materials after redox cycling and catalytic testing, small
particles were clearly present on all three materials. Significant differences could be
found for both the size distributions as well as the particle density. The median particle
1
4

size was lowest for LSFNO-5Sr (6 - 8 nm; Figure 8f), whereas it was significantly
higher (10 - 12 nm) for both LFNO (Figure 8c) and LSFNO-10Sr (Figure 8i).
-
2
Furthermore, particle density was also highest for LSFNO-5Sr (ca. 300 μm )
-
2
-2
compared to LFNO (ca. 100 μm ) and LSFNO-10Sr (ca. 150 μm ). Thus, we attribute
the increase in CH yield on LSFNO-5Sr compared to undoped LFNO not only to a
4
larger extent of Ni reduction observed by Ni K-edge XANES (Figure S1), but also to
the improved morphology of the catalyst after reduction, i.e. lower median Ni particle
size, higher particle density and thus higher Ni specific surface area. Despite the fact
that the extent of Ni reduction was further increased, the observed decrease in CH
4
yield with further Sr addition to yield LSFNO-10Sr can be tentatively explained by Ni
particle size and density as these did not seem to deviate significantly from the values
obtained for LFNO. It should be noted that due to the increased reducibility of LSFNO-
1
0Sr compared to the other two compositions, also Fe of the host PMO started to
reduce, thus contributing to particle formation and likely increased Fe content in the Ni
particles. It cannot be excluded that the presence of Fe in the Ni catalyst may modify
catalytic activity as well as selectivity towards CH
45]. We lack information on the detailed surface composition of the samples before
and after Ni segregation with respect to the other elements that does not allow us to
conclude on the role of the support material on CH production in the presence of Ni
4
production during CO methanation
2
[
4
particles. The surface composition could change under the reduction and oxidation
conditions to which the samples are exposed.
Conclusions
1
5

Partial substitution of La by Sr on the A-site of LaFe0.8Ni0.2
O led to significant decrease
3
in perovskite phase stability towards reduction. As a result increased amounts of B-
site Ni could be segregated to the perovskite surface, forming Ni particles active
towards the hydrogenation of CO
was found to increase with Sr content and was highest for La0.9Sr0.1Fe0.8Ni0.2
However, despite the low levels of substitution, highest CH yields were found for the
3-δ, which exhibited increased Ni
, decreased median particle size of the
2
4
to CH . The degree of Nickel reduction at 600°C
3-δ
O .
4
intermediate composition La0.95Sr0.05Fe0.8Ni0.2
O
reduction compared to Sr-free LaFe0.8Ni0.2
O
3
segregated Ni particles as well as highest particle density. Therefore, La substitution
on the perovskite A-site by Sr can be used to increase Ni segregation as well as
catalytic activity. The observed changes in catalytic activity are attributed mainly to the
presence of the Ni particles, but we cannot discard effects of the surface composition
that have been produced upon reduction. The decreased stability of Sr-containing
compositions towards reduction at elevated temperatures did not affect the structural
reversibility of the materials over short redox cycles at 650°C. However, indications
were found that Ni reincorporation was increasingly more difficult with increasing Sr
content. Furthermore, Sr-doping may cause difficulties in the case of prolonged
operation in the highly reducing atmospheres and at the constant elevated
temperatures typically experienced by a CO methanation catalyst due to decreased
2
material stability towards reduction. Although limits of structural reversibility at 650°C
were not observed throughout the experiments shown in this work, it is likely that high
degrees of perovskite-decomposition after prolonged reduction would not allow for
complete structural reversibility at similar reoxidation conditions as were applied here.
Acknowledgements
1
6

The authors are thankful for the financial support from the Swiss National Science
Foundation (SNF, No. 200021_159568) and the Competence Center for Energy and
Mobility (CCEM). The work was conducted in the context of the Swiss Competence
Center for Energy Research (SCCER BIOSWEET) of the Swiss innovation agency
lnnosuisse. The X10DA (SuperXAS) beamline at the Swiss Light Source (SLS) in
Villigen (Switzerland) and Dr. M. Nachtegaal are thanked for kindly providing the beam
time and support during measurements.
References
[
1] Y. Nishihata, J. Mizuki, T. Akao, H. Tanaka, M. Uenishi, M. Kimura, T. Okamoto,
N. Hamada, Nature, 418 (2002) 164-167.
[
(
2] D. Neagu, G. Tsekouras, D.N. Miller, H. Menard, J.T.S. Irvine, Nat. Chem., 5
2013) 916-923.
[
(
3] D. Burnat, R. Kontic, L. Holzer, P. Steiger, D. Ferri, A. Heel, J. Mater. Chem. A, 4
2016) 11939-11948.
[
4] D. Neagu, T.S. Oh, D.N. Miller, H. Ménard, S.M. Bukhari, S.R. Gamble, R.J. Gorte,
J.M. Vohs, J.T.S. Irvine, Nat. Commun., 6 (2015) 1-8.
[
5] H. Tanaka, M. Taniguchi, M. Uenishi, N. Kajita, I. Tan, Y. Nishihata, J. Mizuki, K.
Narita, M. Kimura, K. Kaneko, Angew. Chem. Int. Edit., 45 (2006) 5998-6002.
[
6] P. Steiger, D. Burnat, H. Madi, A. Mai, A. Holzer, J. Van Herle, O. Kröcher, A. Heel,
D. Ferri, Chem. Mater., (2019).
[
4
7] P. Steiger, M. Nachtegaal, O. Kröcher, D. Ferri, ChemCatChem, 10 (2018) 4456-
464.
[
8] P. Steiger, R. Delmelle, D. Foppiano, L. Holzer, A. Heel, M. Nachtegaal, O. Kröcher,
D. Ferri, ChemSusChem, 10 (2017) 2505-2517.
[
9] J. Kuc, Y.C. Zhang, R. Erni, S. Yoon, L. Karvonen, A. Weidenkaff, S.K. Matam,
Phys. Status Solidi RRL, 9 (2015) 282-287.
[
10] M. Araki, V. Ponec, J. Catal., 44 (1976) 439-448.
1
7

[
[
[
11] P. Schoubye, J. Catal., 14 (1969) 238-246.
12] J.L. Falconer, A.E. Zaǧli, J. Catal., 62 (1980) 280-285.
13] G. Garbarino, D. Bellotti, P. Riani, L. Magistri, G. Busca, Int. J. Hydrogen Energ.,
4
0 (2015) 9171-9182.
[
[
14] G.A. Mills, F.W. Steffgen, Catal. Rev., 8 (1973) 159-210.
15] C. Vogt, E. Groeneveld, G. Kamsma, M. Nachtegaal, L. Lu, C.J. Kiely, P.H.
Berben, F. Meirer, B.M. Weckhuysen, Nat. Catal., 1 (2018) 127-134.
[
[
[
16] J.G. Xu, G.F. Froment, AlChE J., 35 (1989) 88-96.
17] B.C. Enger, A. Holmen, Catal. Rev., 54 (2012) 437-488.
18] J.L. Rogers, M.C. Mangarella, A.D. D'Amico, J.R. Gallagher, M.R. Dutzer, E.
Stavitski, J.T. Miller, C. Sievers, ACS Catal., 6 (2016) 5873-5868.
[
2
19] P. Gronchi, D. Fumagalli, R. Del Rosso, P. Centola, J. Therm. Anal., 47 (1996)
27-234.
[
[
[
20] J.R. Rostrup-Nielsen, J. Catal., 31 (1973) 173-199.
21] S.R. Li, J.L. Gong, Chem. Soc. Rev., 43 (2014) 7245-7256.
22] C. Okolie, Y.F. Belhseine, Y.M. Lyu, M.M. Yung, M.H. Engelhard, L. Kovarik, E.
Stavitski, C. Sievers, Angew. Chem. Int. Edit., 56 (2017) 13876-13881.
[
23] C.H. Bartholomew, Catal. Rev., 24 (1982) 67-112.
[
24] A.D. Moeller, C.H. Bartholomew, Ind. Eng. Chem. Prod. Res. Dev., 21 (1982)
3
90-397.
[
25] K.E. Wilson, Heterogeneous catalysts for converting renewable feedstocks to
fuels and chemicals, in: L. Guczi, A. Erdôhelyi (Eds.) Catalysis for Alternative Energy
Generation, Springer, New York, 2012.
[
[
[
26] J.R. Rostrup-Nielsen, Catal. Today, 37 (1997) 225-232.
27] P. Steiger, I. Alxneit, D. Ferri, Acta Mater., 164 (2019) 568-576.
28] F. He, X.N. Li, K. Zhao, Z. Huang, G.Q. Wei, H.B. Li, Fuel, 108 (2013) 465-473.
1
8

[
[
29] H. Ullmann, N. Trofimenko, Solid State Ionics, 119 (1999) 1-8.
30] K. Huang, H.Y. Lee, J.B. Goodenough, J. Electrochem. Soc., 145 (1998) 3220-
3
227.
[
1
31] Y. Teraoka, H.M. Zhang, S. Furukawa, N. Yamazoe, Chem. Lett., (1985) 1743-
746.
[
[
[
32] N.A. Merino, B.P. Barbero, P. Grange, L.E. Cadus, J. Catal., 231 (2005) 232-244.
33] M.A. Peña, J.L.G. Fierro, Chem. Rev., 101 (2001) 1981-2017.
34] S. Royer, D. Duprez, F. Can, X. Courtois, C. Batiot-Dupeyrat, S. Laassiri, H.
Alamdari, Chem. Rev., 114 (2014) 10292-10368.
[
(
35] J. Deng, M.D. Cai, W.J. Sun, X.M. Liao, W. Chu, X.S. Zhao, ChemSusChem, 6
2013) 2061-2065.
[
2
36] S. Royer, H. Alamdari, D. Duprez, S. Kaliaguine, Appl. Catal. B, 58 (2005) 273-
88.
[
[
[
[
[
37] F.H. Taylor, J. Buckeridge, C.R.A. Catlow, Chem. Mater., 29 (2017) 8147-8157.
38] R.D. Shannon, C.T. Prewitt, J. Inorg. Nucl. Chem., 32 (1970) 1427-1441.
39] M. Cherry, M.S. Islam, C.R.A. Catlow, J. Solid State Chem., 118 (1995) 125-132.
40] C. Marcilly, P. Courty, B. Delmon, J. Am. Ceram. Soc., 53 (1970) 56-57.
41] P. Scherrer, Nachrichten von der Gesellschaft der Wissenschaften zu Göttingen,
1
918 (1918) 98-100.
[
42] J. Schindelin, I. Arganda-Carreras, E. Frise, V. Kaynig, M. Longair, T. Pietzsch,
S. Preibisch, C. Rueden, S. Saalfeld, B. Schmid, J.Y. Tinevez, D.J. White, V.
Hartenstein, K. Eliceiri, P. Tomancak, A. Cardona, Nat. Methods, 9 (2012) 676-682.
[
43] O. Müller, M. Nachtegaal, J. Just, D. Lützenkirchen-Hecht, R. Frahm, J.
Synchrotron Radiat., 23 (2016) 260-266.
[
[
44] B. Ravel, M. Newville, J. Synchrotron Radiat., 12 (2005) 537-541.
45] D. Pandey, K. Ray, R. Bhardwaj, S. Bojja, K.V.R. Chary, G. Deo, Int. J. Hydrogen
Energ., 43 (2018) 4987-5000.
1
9

Figure 1 (a) Powder XRD patterns of LFNO, LSFNO-5Sr and LSFNO-10Sr after
calcination. (b) Magnification around the perovskite (121) reflection.
Figure 2 Contour plots of XRD patterns during in situ reduction of (a) LFNO, (b)
LSFNO-5Sr and (c) LSFNO-10Sr starting from calcined powders. The patterns of the
materials after calcination (0 h) and at the end of the experiment (40 h, 37 h and 17 h)
are shown. Markers indicate reflections of the perovskite-type host (blue), the
decomposition product La
green).
Figure 3 Normalized Ni K-edge (8.333 keV) XANES spectra of calcined and reduced
a) LFNO, (b) LSFNO-5Sr and (c) LSFNO-10Sr. The spectra of Ni foil and NiO
2
O (orange) and the Fe/Ni-containing metallic phase
3
(
(
reference materials are also included.
3
0
Figure 4 Fourier transformed k -weighted Ni K-edge EXAFS data of (a) Ni foil (Ni )
and NiO reference materials, (b) LFNO, (c) LSFNO-5Sr and (d) LSFNO-10Sr in their
calcined states as well as after reduction (10 vol% H /Ar, 600°C, 1 h). Data are non-
2
phase shift corrected. Features on the curves of the reduced material are labelled
according to the underlying Ni-Ni scattering path for metallic Ni.
Figure 5 TPR reduction-reoxidation cycles for (a) LFNO, (b) LSFNO-5Sr and (c)
LSFNO-10Sr. H
sample used. Reoxidized samples were subjected to reduction (10 vol% H
h) before reoxidation at the given temperature (20 vol% O /N , 2 h).
Figure 6 TPR reduction-reoxidation cycles with constant reoxidation temperature of
50°C for (a) LFNO, (b) LSFNO-5Sr and (c) LSFNO-10Sr and of 600°C in (d), (e) and
f), respectively. H consumption values were normalized by the number of moles of
sample used. Reoxidized samples were subjected to reduction (10 vol% H /Ar, 600°C,
h) before reoxidation at the given temperature (20 vol% O /N , 2 h).
2
consumption values were normalized by the number of moles of
2
/Ar, 600°C,
1
2
2
6
(
2
2
1
2
2
2
0

Figure 7 CO conversion profiles of (a) LFNO, (b) LSFNO-5Sr and (c) LSFNO-10Sr
2
measured between five consecutive redox cycles. The corresponding CH
shown in (d), (e) and (f), respectively.
4
yields are
Figure 8 SEM images of calcined (a) LFNO, (d) LSFNO-5Sr and (g) LSFNO-10Sr and
reduced samples after five redox cycles; (b) LFNO, (e) LSFNO-5Sr and (h) LSFNO-
1
0Sr. Histograms representing particle size distributions of the freshly formed particles
are shown in (c), (f) and (i) for LFNO, LSFNO-5Sr and LSFNO-10Sr, respectively.
Scale bars on SEM images represent 200 nm.
Figures
Figure 1
2
1

Figure 2
2
2

Figure 3
2
3

Figure 4
2
4

Figure 5
2
5

Figure 6
2
6

Figure 7
2
7

Figure 8
2
8