i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 1 3 9 2 1 e1 3 9 3 2

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.elsevier.com/locate/he

Exsolution in Ni-doped lanthanum strontium

titanate: a perovskite-based material for anode
application in ammonia-fed Solid Oxide Fuel Cell

Jonathan Cavazzani a,*, Enrico Squizzato a, Elena Brusamarello a,

Antonella Glisenti a,b
a
Department of Chemical Sciences, University of Padova, Via F. Marzolo 1, 35131, Padova, Italy
b
ICMATE - Department of Chemical Sciences, University of Padova, Via F. Marzolo 1, 35131, Padova, Italy

highlights graphical abstract

Successful Ni exsolution in
Lanthanum Strontium Titanate
confirmed by XRD, SEM, XPS.

Powder morphology not affected
by Ni doping and exsolution.

Ni nano-decoration of LaSrTiO3 by
exsolution boosts electrochemical
performances.

Exsolution is much more effective
than infiltration in decreasing
polarization.

Increasing the temperature, the
cell performances using NH3 or H2
become similar.

article info abstract

Article history: Hydrogen represents the most conventional fuel to feed Solid Oxide Fuel Cells (SOFCs) for
Received 29 November 2021 green energy production. However, hydrogen has some drawbacks which prevent the
Received in revised form large-scale implementation. Research identified ammonia as promising hydrogen vector.
4 February 2022 Hereby, highly dispersed Ni nanoparticle are deposited on La-doped strontium titanate by
Accepted 15 February 2022 exsolution, greatly affecting the electrochemical performance. The exsolved Ni-doped
Available online 15 March 2022 lanthanum strontium titanate (La0$45Sr0$45Ti0$90Ni0.10-dO3 e LSTNOH) was largely charac-
terized. XRD analysis detected 10 mol% of Ni doping has been successfully incorporated in
Keywords: to the perovskite structure and then released when exposed in reducing environment. SEM
Ammonia images show Ni nanoparticles highly dispersed on the surface. XPS confirms the presence
Hydrogen of Ni on the surface after the exsolution and allows to exclude other detrimental diffusion
SOFC towards the bulk. A LSTNOH derived composite based anode has been investigated through
Anode impedance spectroscopy using ammonia and hydrogen as fuel. It demonstrates best per-
formances compared to the one obtained by Ni infiltration on LSTO (La0$45Sr0$45TiO3)

* Corresponding author.
E-mail address: [email protected] (J. Cavazzani).
https://doi.org/10.1016/j.ijhydene.2022.02.133
0360-3199/© 2022 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.
13922 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 1 3 9 2 1 e1 3 9 3 2

LaSrTiNiO 3 composite scaffold. Polarization resistance, running on ammonia, decreases raising the
Ni exsolution temperature and the performances approach those in hydrogen.
© 2022 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

involves on NH3 adsorption and N2 desorption which depends

Introduction on nitrogen desorption enthalpy [31]. In ammonia-fuelled
SOFC, the electrocatalyst anode materials must also provide
Solid Oxide Fuel Cells (SOFCs) are promising devices able to to the oxidation of hydrogen coming from ammonia decom-
convert chemical energy in to electricity with high efficiency position. In addition, nickel cermet anodes are the state-of-art
and low environmental impact. They are featured by high material for this application in SOFC technology [32e34]. These
working temperature, in 700e1000  C range, promoting an considerations suggested us to develop Ni-based nano-
exploitable fuel flexibility [1e4]. Normally, these devices are composites to be used as anodes in ammonia-fed Solid Oxide
fed by hydrogen, however it shows well known limitations Fuel Cells [35e41]. To this purpose a perovskite e based
regarding transport and storage [5,6]. Hydrocarbons are nanocomposite of the type Ni/titanate is developed. We
largely studied for this application but carbon-based fuels already demonstrated that relevant functionalities can be ob-
have some drawbacks in long-term anode performance [7e9]. tained starting from stable, sustainable and cheap strontium
Ammonia exhibits some advantages over hydrogen and hy- titanate by means of doping and impregnation with Ni [42].
drocarbons to feed SOFCs and although it has not yet received Nanomaterials have been playing a crucial role in many
the proper attention as hydrogen carrier, the potential of this fields such as electrochemistry, photochemistry, chemicals
fuel is evident [10e12]. Ammonia large-scale synthesis has been conversion and others thanks to high activity and the possi-
developed in the first half of the last century through the known bility to tailor the functionalities for a specific application.
Haber-Bosch process [13,14]. Nowadays, 180 tons per year of However, they exhibit some issues connected to the coars-
ammonia are produced and transported [12]. Therefore, the in- ening of nanoparticles or poisoning of the catalytic sites
frastructures for production, storage and distribution are totally which drive to lose the catalytic properties and a weakening of
consolidated and well established [15,16]. In addition, being one long-term performances. Recently, exsolution has been
of the largest chemicals produced in the world, the scientific explored to synthetized supported metal or metal oxide
community is investigating more environmentally sustainable nanoparticles strongly anchoring to the ceramic oxides
and green routes to produce ammonia exploiting renewable [43,44]. Exsolution has been initially investigated between late
energy [17,18] or by electrochemical methods [19,20]. and early 2000s [45,46], but the breakthrough has been trig-
Liquid ammonia has higher energy density, 12.9 MJ/L, with gered when chemical defects were explited to promote the
respect to gaseous hydrogen, 0.9 MJ/L at 25  C and using high process [47]. It consists on embedding the active metal atoms
pressure of 10 MPa. The volumetric density increases with inside the lattice and then promote the release of them
liquid hydrogen, 8.6 MJ/L, but the requested temperature is forming socketed nanoparticles on the surface [48e51]. The
253  C at 0.1 MPa [21]. Moreover, ammonia is a carbon-free emergence of metal atoms to the surface confers robustness
fuel, thus there are no restrictions regarding the catalysts to the materials enhancing their properties in coarsening and
and no greenhouse gas such as CO2 are formed [16,22]. poisoning resistance. Exsolved Ni-based catalysts were stud-
Concern health issues, the flammability range for ammonia in ied in carbon containing gases for conversion or power supply
air is narrower and the ignition temperature is higher and they exhibit higher tolerance towards coking and lower
compared to the common hydrocarbons, indicating that aggregation [52e54]. Exsolution has been recently investigated
ammonia has low flammability [23]. Moreover, the flamma- for ammonia fuelled-SOFC showing appealing results. Shao
bility range of hydrogen is larger than the one of ammonia, et al. [55] worked on exsolution of Pd nanoparticles to boost
4e74 and 16e25 respectively [16]. the performances of Ni/BaZrCeYYbPd cermet. They demon-
Pioneering researches in ammonia-fed SOFC were carried strate palladium enhance the ammonia decomposition but it
out by Vayenas [24] and later by Wojcik [25]. They demon- is also a strategic help to increase proton conductivity. They
strated that a Solid Oxide Fuel Cell can work using ammonia as also reached 600 mW/cm2 at 650  C working in ammonia,
fuel. Ammonia is oxidised via a two-step process. Firstly, about 25% higher than Ni cermet without Pd exsolved nano-
ammonia cracks in to hydrogen and nitrogen. Then hydrogen particles. In addition, Song and co-workers [56] studied LSTN-
is furtherly oxidised in protons and electrons to allow the based anodes, on GDC scaffold, demonstrate interesting re-
electrochemical working cycle in a SOFCeO, where the oxygen sults using ammonia reaching 161 mW/cm2 at 800  C and the
ions cross the electrolyte from the cathode to the anode [21]. results become better when Co was also incorporated forming
Studies demonstrate that ruthenium is the most active metal NiCo alloy nanoparticles reaching 361 mW/cm2. They attrib-
for the ammonia decomposition in hydrogen and nitrogen uted the improvement in the performances to a more balance
[26e28]. However, its high cost and poor availability lead to a of NH3 adsorption and N2 desorption process and the conse-
difficult large-scale production. Works [29,30] show how Ni quent H2 electrochemical oxidation.
containing catalysts have been observed to be active in In this work, lanthanum strontium titanate materials
ammonia decomposition reaction. Ammonia cracking were synthetized through wet chemistry route using the
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 1 3 9 2 1 e1 3 9 3 2 13923

citric acid method. Lanthanum strontium titanate has a

Table 1 e Thermal treatment used for the synthetized
perovskite structure (ABO3) where one half of A-site are
samples.
occupied by La atoms and the others by Sr. Lanthanum is
Samples Treatment
exploited to promote the formation of the redox couple Ti4þ/
Ti3þ in order to enhance the electronic conductivity in the Stoichiometry Name
anode material [57e59] and to induce the formation of oxy- La0$45Sr0$45TiO3 LSTO Air, 1000  C, 12 h
gen vacancies to improve the oxygen ion conductivity [60]. La0$45Sr0$45Ti0$90Ni0$10O3 LSTNO Air, 1000  C, 12 h
Furthermore, the thermal coefficient of lanthanum stron- La0$45Sr0$45Ti0$90Ni0.10-dO3 LSTNOH Air, 1000  C, 12 h
tium titanate is close to the one of YSZ: 10.4  106 K1 and and then 5% H2/Ar,
900  C, 20 h
10.3  106 K1, respectively [61].
Exsolution method was explored to decorate the surface Synthetized samples were reported with their short name and
highlighting their thermal treatment.
with the metal nanoparticles. Unlike other deposition pro-
cedures (wet impregnation, deposition precipitation), exsolu-
tion is expected to allow obtaining metal or metal oxide
nanoparticles highly dispersed on the surface. All materials Characterization
were synthetized with an A-deficiency of 10 mol% with the
aim to promote the exsolution of the metal in the B-site All characterizations were performed using the powders pre-
[62e64]; B-site was doped by 10 mol% of Ni (La0$45Sr0$45Ti0$90- viously synthetized. X-Ray Diffraction (XRD) analyses were
Ni0$10O3). Ni infiltration on La0$45Sr0$45TiO3 composite scaffold carried out through a Bruker D8 Advance diffractometer with
was also carried out to evaluate the effect of deposition pro- Bragg-Brentano geometry using a Cu Ka radiation (40 kV,
cedure (and thus of particle size) on the electrochemical ac- 40 mA, l ¼ 0.154 nm).
tivity. Unlike the exsolution, the infiltration method creates a Field-emission Scanning Electron Microscopy and Energy
continuous nanoparticulate network to improve the electron Dispersive X-ray (EDX) spectroscopy measurements were
pathway from the current collector to individual reaction site carried out on a Zeiss SUPRA 35VP using an acceleration
[65] but usually the dimensions of the particles are bigger. voltage of 20 kV (images of the morphologies). While the SEM
The anode materials were deeply characterized and the images of Ni nanoparticles and the cross-section of the cell
obtained results have been related to the deposition proced- were collected by Zeiss Sigma HD microscope, equipped with
ure. Electrode performances were evaluated through Electro- a Schottky FEG source, one detector for backscattered elec-
chemical Impedance Spectroscopy (EIS). All materials were trons and two detectors for secondary electrons (InLens and
tested as composite with YSZ in order to promote the ionic Everhart Thornley).
conductivity. The measurements were carried out at 800 and Specific Surface Area (SSA) were measured by N 2 phys-
900  C changing the fuel content in the flow, both ammonia isorption at the temperature of liquid N2 through an ASAP
and hydrogen for a better understanding of the polarization 2020PLUS Micromeritics and the Brunauer-Emmett-Teller
processes. In this first contribution the focus is on the anode (BET) equation. Samples were degassed at 400  C for 4 h
material and on its synthesis. Further work is in course on the before the N2-sorption.
device to improve the performance through a thinner elec- X-ray Photoelectron Spectroscopy (XPS) measurements
trolyte layer and a modified architecture. were carried out with a Perkin Elmer F 5600ci Multi Technique
System. The spectrometer was calibrated by assuming the
binding energy (BE) of the Au 4f7/2 line to be 84.0 eV with
Experimental respect to the Fermi level. The detailed spectra for qualitative
and quantitative analysis (La 3d, Sr 3d, Ti 2p, Ni 2p and O
Synthesis 1se23.5 eV pass energy, 0.1 eV/step, 0.1 s/step) were collected
with a standard Al Ka source working at 250 W. The standard
La0$45Sr0$45Ti0$90Ni0$10O3 was prepared by the citrate method deviation in the BE values of the XPS line is 0.10 eV. The atomic
[66] starting from La2O3 (Sigma-Aldrich, 99.9%) SrCO3 (Sigma- percentage was evaluated using the PHI sensitivity factors [68]
Aldrich, 98%), C12H28O4Ti (Sigma-Aldrich, 97.0%) and after a Shirley-type background subtraction [69]. The peak
Ni(NO3)2$6H2O (Sigma-Aldrich, 97.0%). After complete disso- positions were corrected for the charging effects by consid-
lution of the precursors with nitric acid (Sigma-Aldrich, 65%) ering the C 1s peak at 285.0 eV and evaluating the BE differ-
and distilled water, the final solution was heated at 80  C to ences. XPS measurements of compounds containing either
promote the solvent evaporation obtaining a wet-gel. It was lanthanum and nickel, the signal of Ni 2p3/2 is hidden by La
then heated in air at 350e400  C for 2 h to decompose the 3d3/2, for this reason the analysis of the spectra is more
organic framework. The ash-like material was grinded and complicated. However, a coherent method for the detection
then calcined in air at 1000  C for 12 h (controlled temperature has been devised. It is based on a clear assumption: since the
ramp 6  C min1), following the procedure in Ref. [67]. La0.45- molar amount of lanthanum is equal in all compounds
Sr0.45TiO3 was synthetized using the same method. (La0$45Sr0$45TiO3, La0$45Sr0$45Ti0$90Ni0$10O3, La0$45Sr0$45Ti0$90-
The calcined La 0.45Sr0$45Ti0$90Ni0$10O3 was subsequently Ni0.10-dO3), to obtain the separate contribution of lanthanum
treated in hydrogen (5% H2/Ar) to promote the exsolution at and nickel in LSTNO and LSTNOH, the XPS signal due to
900  C for 20 h (controlled temperature ramp 7 C min1), as lanthanum were considered the one measured in LSTO to
reported in the methodology [62]. The treatments with which allow the calculation of the quantities of nickel in LSTNO and
the samples were prepared are summarized in Table 1. LSTNOH. The acquisition time has selected to be the same in
13924 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 1 3 9 2 1 e1 3 9 3 2

all cases. The contribution of nickel was achieved by sub-

Table 2 e Cell composition and performances.
tracting, from the measured signal of lanthanum (both La 3d3/
2 and La 3d5/2) in LSTNO and LSTNOH, the signal of lanthanum
Symmetric cell Temp. Fuel in Rp (U)* ASR
( C) Ar (U cm2)**
(both La 3d3/2 and La 3d5/2, no nickel in the compound) in
LSTO. LSTNOþ8YSZ (50:50) 800 20% H2 49.7 28.1
|8YSZ 800 20% NH3 77.3 43.7
LSTNOHþ8YSZ 800 20% H2 19.9 11.3
Cell preparation
(50:50) |8YSZ 800 20% NH3 21.5 12.2
900 20% H2 4.9 2.8
The electrochemical measurements were carried out using 900 10% NH3 5.8 3.3
symmetric cells. Thick electrolyte was prepared by pressing a 900 20% NH3 5.4 3.1
commercial 8YSZ (Yttria-Stabilized Zirconia) powder (Fuel 900 40% NH3 5.1 2.9
Cell Materials, Lewis Center, Ohio) in to pellets and then firing LSTOþ8YSZ (50:50) | 800 20% H2 75.1 42.5
them in air at 1500  C for 6 h. The final electrolyte is 1.5 mm 8YSZ 800 20% NH3 115.9 65.5
8.3 wt% - 800 20% H2 32.6 18.4
thick and has 18.9 mm of diameter. The symmetric electrodes,
Ni@(LSTOþ8YSZ 800 20% NH3 49.7 28.1
composites between the synthetized materials and 8YSZ
(50:50)) |8YSZ 900 20% H2 10.5 5.9
(50:50), were deposited by a homemade screen-printing ma- 900 20% NH3 13.4 7.6
chine and fired at 1200 C for 2 h in air (controlled temperature
All cell measurements are sum up highlighting the symmetric cell
ramp 5 C min1). Final electrodes have a diameter of 12 mm. composition and cell performances. * polarization resistance
Ink preparation involved grinding the powders and mixing values were collected using EIS; ** area specific resistance values
them with proper amounts of a-terpineol and polyvinyl were calculated as described in this section.
butyral binder (PVB); 3 wt% of Carbon soot (Orion Engineered
Carbons, Frankfurt am Main, Germany) was added to the ink
in order to guarantee the porosity to the electrode. Finally, as
electrons collector, gold paste was deposited and treated at The electrochemical data were fitted using Zview software
800  C. Infiltration was directly made on the deposited on (Scribner associated) using the more adequate equivalent
LSTO composite scaffold using an aqueous solution of circuits as well as the most widespread used in literature for
Ni(NO3)2$6H2O (Sigma-Aldrich, 97.0%). The total amount of Ni these materials.
deposited is 8.3 wt%. It approaches to the mass of Ni in
La0,45Sr0,45Ti0,90Ni0,10O3 composite anodes. The water solution
of nickel nitrate was dripped on the electrode. Then the cell Results and discussion
was placed under vacuum for 2 min. The excess of solution
was removed with a paper and the cell was heated up to 450  C Structure investigation: XRD
for 30 min to decompose the nitrates. Gold paste was depos-
ited and treated at 800  C. In Fig. 1 are displayed the room temperature diffractograms of
In Table 2, all the studied cells are reported highlighting the XRD analysis. All samples confirm the characteristic re-
their composition and summing up the performances. The flections of the perovskite crystal structure.
values of area specific resistance (ASR) were calculated No second phases are detected in LSTNO with 10 mol% of
considering one electrode (both in full cell) in the symmetric Ni doping after the thermal treatment in air at 1000  C for 12 h.
cell and its surface area (1.13 cm 2), starting from Rp values This suggests that the amount of nickel is completely soluble
obtained from impedance measurements. in lanthanum strontium titanate. While, metallic Ni was
detected after the reducing treatment in hydrogen, in LSTNOH
Electrochemical measurements diffractogram. Ni 0 phase shows two reflections: the most
intense one at ~44.5 (2q) and another one centred at ~ 52.2
Symmetric cell tests were carried out in a home-made single (2q) which forms a shoulder with a perovskite phase reflec-
chamber set up. It is made up by an a-alumina DEGUSSIT AL23 tion. They were identified using JCPDS database (# 96-901-
tube (15 mm of diameter) with a groove on the top where the 3025). The presence of metallic nickel, as expected, suggests
button cell is supported. The ceramic tube is covered by a the exsolution has effectively taken place.
quartz tube to isolate the chamber. The fuel (flow rate 100 The insertion in Fig. 1 shows a closer inspection of the
sccm) is injected on the bottom of the chamber and it goes out reflection at ~32.5 (2q). The diffractograms in samples con-
from the top. taining nickel are shifted towards lower angles. The relative
Flow rate are controlled by thermal mass flow meters intensities do not vary. The shift is likely caused by the sub-
€ gtlin Instruments) and fuel is made passing through con-
(Vo stitution of Ti4þ (61 p.m.) ions with the bigger Ni2þ (69 p.m.)
taining water at room temperature: measurements were ones. This result is in agreement with a successfully incor-
labelled “wet” and it indicates fuels containing 3% of water. poration of the Ni atoms in the perovskite crystal structure,
Electrochemical Impedance Spectroscopy (EIS) measure- giving rise an expansion of the lattice [24,34].
ments were performed using an Autolab Frequency Response Moreover, from the closer view in Fig. 1, a further shift
Analyser. All tests were carried out in the frequency range of towards lower angles between LSTNO and LSTNOH is also
102-106Hz with a signal amplitude of 20 mV. visible. The reducing treatment produces an expansion of the
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 1 3 9 2 1 e1 3 9 3 2 13925

crystal lattice. Considering the conditions used for the exso-

lution in the current work, 900  C for 20 h, the lattice expan-
sion can be attributed to the reduction from Ti 4þ (61 p.m.) to
Ti3þ (67 p.m.) [70].

Morphology: SEM

SEM analysis allowed to verify the presence of Ni nano-

particles on the surface of the LSTNOH powder, as shown in
Fig. 2 d, e and f. The nanoparticle size varies about from 25 to
50 nm and they have a good dispersion on the surface. The
high dispersion, hopefully, would improve the catalytic and
electrochemical properties. The Ni nanoparticle dimensions
are in agreement with other works which used a similar
reducing treatment for exsolution [62].
SEM analysis was also exploited to study the morphology
of the synthetized powders. SEM images are displayed in Fig. 2
a, b and c. All samples show a similar morphology with a
Fig. 1 e Room temperature XRD of La0·45Sr0·45TiO3 and discrete porosity and the formation of agglomerates with
La0·45Sr0·45Ti0·90Ni0·10O3 after calcination at 1000  C for 12 h in different size. Neither Ni doping in LSTNO not the reduction
air (LSTO and LSTNO, respectively); LSTNO was furtherly treatment in LSTNOH modify the morphology, compared to
treated at 900  C for 20 h in 5% H2/Ar to promote the the undoped powder. These observations were deeply inves-
exsolution: La0·45Sr0·45Ti0·90Ni0.10-dO3 (LSTNOH). Insertion: tigated by SSA measurements. The results, summarized in
closer inspection of the most intense reflection at ~32.5 (2w). Table 3, confirm that the SSA is not strongly affected by the Ni
doping and reducing treatments neither.

Fig. 2 e SEM images: a) La0·45Sr0·45TiO3; b) La0·45Sr0·45Ti0,.90Ni0,.0O3; c) La0·45Sr0·45Ti0·90Ni0.10-dO3. a, b and c images were

collected using 3k magnification. Images d, e and f shows Ni nanoparticles in La0·45Sr0·45Ti0·90Ni0.10-dO3. The SEM image was
collected using increasing magnification: 50k, 150, and 250, respectively. The estimated size varies between 25 and 50 nm.
13926 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 1 3 9 2 1 e1 3 9 3 2

when Ni is exsolved: the FWHM of the peaks, consistently,

Table 3 e Surface specific area (SSA) measurements of the
decrease (from 4.2 to 3.7 eV for Sr 3d, from 3.3 to 2.5 eV for Ti
synthetized powders.
2p3/2, and from 3.6 to 3.0 eV for O 1s).
Powder SSA (m2/g)
As written above, La 3d3/2 and Ni 2p3/2 peaks are overlapped
LSTO 17.65 and, because of the low amount of nickel in LSTNO and
LSTNO 15.72 LSTNOH samples, lanthanum strongly dominates. At higher
LSTNOH 16.74
energy, a peak is present in all spectra at ~864.6 eV. Nickel can
SSA values were collected for the powders and measured by BET form several compounds: the position and the intensity of this
analysis. peak varies with its oxidation state because the energetic
levels change as well. This peak is a shake-up contribution [72]
and is typical of Ni(II) in NiO; it can be found in Ni0 but is a lot
Surface characterization: XPS less intense. Consequently, it suggests that a nickel oxide thin
layer (or, likely, nanoparticles [73]) is present on the surface of
High resolution XPS spectra for elements in all samples are the Ni exsolved particles (XRD diffractogram do not detect
shown in Supplementary Material (Figure SM.1) Moreover, in NiO). The overlap between La 3d3/2 and Ni 2p3/2 prevent a
Table SM.1 the position of the peaks is reported and compared deeper analysis.
with literature data. An extensive quantitative analysis was performed for all
In strontium titanate (SrTiO 3), Sr 3d, the most intense powders through XPS and EDX, the results are reported in
peak, is centred at 132.9 eV [71]. In LSTO sample, the signal is Table 4. The measured values were compared to the nominal
slightly shifted to highest energy (~133.1 eV). This shift is ones with the aim to investigate the role of the Ni doping and
consistent with a modification of the chemical environment the reducing treatment. In Table 4, also the cations-only sur-
and suggests the possible formation of surface oxides or hy- face compositions are reported to highlight the surface
droxides. The presence of carbonates has been excluded segregation phenomena, often hidden by relevant amount of
because of the absence of the typical C 1s contribution at oxygen.
about 290 eV. Titanium also shows two peaks due to spin- It is well known that the XPS analysis of compounds
orbital coupling. In SrTiO3, lowest binding energy peak is including both lanthanum and nickel is more complicated
present at ~458.2 eV, the position expected for Ti(IV) in oxides because the overlap between their signals in binding energy.
[71]; no significant differences in the peak position have been The procedure used to calculate the contribution of the nickel
observed as a consequence of the insertion of nickel and/or content in LSTNO and LSTNOH is largely described in Exper-
the exsolution treatment. imental section.
Lanthanum has a complicated XPS spectrum where four XPS analysis is diagnostic for the Ni exsolution: the nickel
peaks are clearly visible. Spin-orbital coupling generates two content in LSTNOH drastically increases after the reducing
peaks: La 3d3/2 and La 3d5/2. Both of them have a shake-up treatment of LSTNO. XPS registers a surface rich in nickel as
signal (labelled as La 3d3/2 satellite and La 3d5/2 satellite, expected as a consequence of the reductive atmosphere
respectively). These signals are characteristic of the La(III), so induced migration of this element towards the surface. The
constituting a sort of fingerprint of this oxidation state [72]. successful exsolution, and thus the decoration of the surface
Noteworthy, the position of the La 3d signals of LSTO and with Ni nanoparticles, has been confirmed both by XRD and
LSTNOH is almost identical but the peak shape differs for SEM images. EDX analysis collected values close to those
LSTNO. In particular, in this last case, the 3d peaks are nominals: considering its larger investigation depth the
broader and the shake-up contributions are evidently less comparison between XPS and EDX compositions underlines
distinguishable. the surface specific character of the segregation phenomena.
This behaviour, could indicate the formation of lanthanum The insertion of Ni inside the perovskitic cell and its suc-
oxide/hydroxides (LaOOH, as an example), whose position is cessive diffusion towards the surface, significantly affects the
around 835 eV. The O 1s signal of LSTO is centred at 529.3 eV, a composition of surface and subsurface layers.
position expected for oxygen in SrTiO3 [71]. This position It is renowned that the A cations segregates on the surface
slightly shifts towards 529.6 eV in LSTO. The peak is not forming A-site overstoichiometric surface [47,49]. This trend is
perfectly symmetric: a tail around 531e531.5 eV agrees with strongly visible for the lanthanum using XPS, while Sr content
the formation of hydroxides on surface and confirms the approaches to the nominals. However, in LSTNOH, the XPS
mentioned observations of Sr 3d and La 3d. values of lanthanum decreases and the atomic percentage is
An accurate observation of the peaks shape allows to go closer to the nominals.
deeper in the surface description: in particular it is possible to Titanium, instead, shows an opposite behaviour and re-
determine a systematic change of the full width at half sults in lower amount respect the nominal composition. As a
maximum (FWHM) in all but La 3d spectral regions of LSTNO: consequence, the sub-surface layer is richer in titanium and
from 3.5 to 4.2 eV for Sr 3d, from 2.3 to 3.3 eV for Ti 2p3/2, from poorer of A-type cations. In this sense, considering that Sr
2.6 to 3.6 eV for O 1s; this is very relevant and suggests that the moves less towards the surface with respect to La, SrTiO 3
insertion of Ni into the perovskite cell causes a significant could form in sub-surface layers simultaneously to the for-
distortion with the generation of not equivalent B-site near mation of La-rich phase on the surface as oxides or hydroxide.
neighbourhood. This is consistent with the surface (XPS) XPS and EDX data strongly supports this hypothesis in LSTO
composition change (Ti is substituted by Ni, as summarized in and LSTNO. XPS reveals that the reducing treatment in
Table 4). This inhomogeneity seems to decrease significantly LSTNOH causes a decrement in surface content of titanium at
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 1 3 9 2 1 e1 3 9 3 2 13927

Table 4 e EDX and XPS atomic composition.

Sample La Sr Ti Ni O La/Sr Ti/(LaþSr) Ni/Ti (NiþTi)/(LaþSr) O/(LaþSrþNiþTi)
LSTO Nominal 9.4 9.4 18.5 / 62.5 1.0 1.0 / 1.0 1.7
25.0 25.0 50.0
EDX 5.7 6.4 14.6 / 73.3 0.9 1.2 / 1.2 2.7
21.2 23.9 54.9
XPS 12.1 7.6 15.6 / 64.7 1.6 0.8 / 0.8 1.8
34.3 21.5 44.2
LSTNO Nominal 9.2 9.2 18.4 2.0 61.2 1.0 1.0 0.1 1.1 1.6
23.7 23.7 47.4 5.3
EDX 6.1 6.6 13.3 1.4 72.6 0.9 1.0 0.1 1.2 2.6
22.1 24.0 48.6 5.3
XPS 13.9 8.0 13.3 3.2 61.6 1.7 0.6 0.2 0.8 1.6
36.2 20.8 34.7 8.3
LSTNOH Nominal 9.2 9.2 18.4 2.0 61.2 1.0 1.0 0.1 1.1 1.6
23.7 23.7 47.4 5,3
EDX 8.7 6.8 18.6 1.9 64.0 1.3 1.2 0.1 1.3 1.8
24.3 18.8 51.6 5.4
XPS 9.7 7.1 12.8 11.1 59.2 1.4 0.8 0.9 1.4 1.5
24.0 17.4 31.4 27.2

The nominal compositions, derived from the stoichiometry amounts, are reported for comparison. The atomic composition determined
considering only the cations are indicated below the corresponding O-containing ones. This choice permits to emphasize the surface segre-
gation phenomena often hidden by the relevant amount of oxygen.

the expense of nickel. Furthermore, the oxygen content on the

surface is slightly decreased after the treatment in 5% of
hydrogen.

Electrochemical performances

The electrochemical performance was mainly evaluated using

ammonia since it is the main focus of this contribution.
Hydrogen was also used in order to compare the data. Most
interesting results were already summarized in Table 2 in
Experimental.
In Fig. 3 impedance measurements at 800  C are shown. Ni
exsolved nanoparticles in LSTNOH clearly boosts the electro-
chemical performances. Total resistance polarization in
LSTNOH-based electrode is less than half compared to that
obtained with 8.3 wt% Ni infiltration on LSTO composite
scaffold. Moreover, Ni in the perovskite crystal lattice in
LSTNO, and in LSTNOH too, contributes to diminish the total Fig. 3 e EIS spectra for the anode composite materials at
resistance respect to undoped LSTO. Considering LSTNO and 800  C using NH3 (in Ar). 20% of H2 in Ar is also reported as
LSTNOH powders, they were also tested for ammonia reference. “Wet” indices the fuels contain 3% of water.
decomposition. Catalytic tests procedure and results were
reported in the Supplementary Materials. In Figure SM.4
ammonia conversion yield as function of the temperature is hydrogen adsorption and oxidation reducing the polarization
displayed. The conversion trends described by the powders resistance in EIS measurements but they do not bring any
are really similar and this may suggest the Ni nanoparticles evident help in NH3 cracking. In ammonia decomposition,
having no effect towards ammonia decomposition reaction. nitrogen desorption enthalpy plays a key role. Ni exhibits N2
Ammonia starts to crack barely below 550  C and it is totally desorption enthalpy about 240 kJ/mol [31,74], promoting the
decomposed above 720  C. In this manuscript the electro- desorption process at the expense of the adsorption. To in-
chemical performances were studied using composite elec- crease the NeH cleavage, other metals should be included. For
trodes with 8YSZ, however the ammonia decomposes below this reason, metal alloys containing Co, Fe or Mo accelerate
800  C using pure 8YSZ [39]. Hence the temperature range in the ammonia dissociative adsorption process thanks to more
which electrochemical tests were measured, ammonia is a appropriated nitrogen desorption enthalpy (281 kJ/mol,
totally decomposed. In the light of these results, they suggest 400 kJ/mol and 756 kJ/mol respectively) [56,75]. These re-
electrodes containing exsolved Ni nanoparticles exhibit sults bring to the conclusion the catalysts should be tailored
enhanced electrochemical performances compared to the for both reactions using metals alloy able to synergistically
others tested electrodes. Therefore, Ni nanoparticles promote work to enhance the conversion.
13928 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 1 3 9 2 1 e1 3 9 3 2

In Fig. 4, electrochemical performances of LSTNOH are couple Ti4þ/Ti3þ which plays an important role in the elec-
compared as a function of the content of ammonia in the flow. trical conductivity of the anode [76].
The polarization resistance diminishes increasing the fuel Full cell tests were evaluated using LSTNOH and 8YSZ
content, as expected, and 40% of NH 3 (2.9 U cm2) approaches composite as anode. The cathode is made up by a composite
to 20% of H2 (2.8 U cm2). The temperature also affects the between LSCF and GDC. GDC interlayer was deposited be-
polarization resistance. Rising the working temperature at tween the 8YSZ electrolyte and the cathode to prevent the
900  C, LSTNOH-based electrodes running on ammonia interaction between the electrolyte and LSCF [79], more details
exhibit ASR values closer to those in hydrogen (3,1 U cm2 in are reported in SM. Experimental. SEM images of cross section
NH3, 2.8 U cm2 in H2) compared to the analogous at 800  C are reported in Figure SM.2. It shows the interfaces anode/
(Fig. 3; 12.2 U cm2 in NH3, 11.3 U cm2 in H2). This behaviour is electrolyte and cathode/electrolyte. Electrolyte has high den-
consistent with titanates-based materials, their conductivity sity, as expected. Pore former (3% of carbon soot) induces
increases with the temperature in reducing conditions [57]. porosity to both electrodes although anode seems much more
All impedance spectra were fitted using the same equiva- delaminated than cathode probably due to a higher porosity.
lent circuit. In literature, the studies [76e78], regarding the EIS In the image of the anode, some pieces of electrode are
measurements of LST-based materials, use the proposed randomly distributed on the electrolyte, it is clearly caused by
model circuit. It is made up by a resistor (R1), normally the mechanical cutting to break the cell. GDC interlayer was
attributed to the electrolyte, and then the two semicircles are deposited between YSZ electrolyte and LSCF cathode, how-
described by (R-CPE)x, in series with the resistor R1. Two ever, during the thermal treatment, interlayer and cathode
different processes were identified at high frequency (HF: sinter together and they are not distinguishable.
~105e103 Hz) and at medium frequency (MF: ~10e1 Hz). No The IeV and IeP curves are reported in Supplementary
electrochemical processes appear at low frequency (LF: 101 e Materials in Figure. SM. 3 and results are summarized in Table
102 Hz). The latter are typically attributed to diffusion phe- SM.3. A power density peak about 3.7 mW/cm 2 was reached at
nomena. The choice of the composite (50:50) with YSZ can be 780  C using 40% of ammonia. In the same condition,
considered a strategic help in this sense, thanks to the high hydrogen exhibits 11.7 mW/cm2. Inferior power density using
ionic conductivity of YSZ. Thus, the ionic conductivity related ammonia instead pure hydrogen can be attributed to a lower
resistance can be neglected in comparison with the other H2 concentration in the atmosphere because the presence of
resistance contribution [77]. In EIS spectra, no new process N2 produced during the decomposition of ammonia [22,55,56].
appears when electrode is working with ammonia at Pure hydrogen (wet) test was also carried out reaching a power
800e900  C temperature range. The fitting results (Table SM.2) density of 15.5 mW/cm2 (see Table SM.3). Considering results
suggest that both processes are thermally activated as they in hydrogen (40% vs pure), Open Circuit Voltage (OCV) slightly
become less intense with rising the temperature. The fuel decreases from 1049 V in pure H2 to 1016 V in 40% of H2. This
composition also affects the polarization resistance of the light variation of the OCV is mainly caused the difference of
processes but less markedly. In the light of this results, the the hydrogen partial pressure (pH) in the anode compartment:
semicircles in HF region can be attributed to a not perfect rising the pH, the higher content of H2 in the flow (the flow of
electronic transfer through the electrolyte and electrode air at cathode is kept constant at 100 sccm) leads to an
interface [78]. As long as, the semicircle in MF region may be increment of the OCV, according with Nernst equation. While,
assigned to the cyclic reduction and oxidation of titanate the slope variation with the current density, between the
measurements in hydrogen, may represent that the H2 con-
tent is limiting step for the overall reaction. However, further
studies should be brought for better understating the reasons.
More interested is the different OCV using ammonia or
hydrogen (40%), here OCV changes from 0,867 V in NH 3 to
1016 V in H2. Analogously to the power density, OCV is
affected by the atmosphere of the anode: the lower partial
pressure of hydrogen due to the presence of nitrogen in the
mixture diminishes the OCV in ammonia fuelled cell [22].
Beyond N2, also the steam: both the one coming from the inlet
gases, because the fuel passes though the bubbler, but even
the one produced by the electrochemical reaction. Observing
the slope of the IeV curve for the ammonia, the initial slope
undergoes a change at about 5 mA/cm2: for lower values of
density current the losses mainly involves those connected to
activation processes and ohmic resistance. But for higher
current densities, losses are likely due to mass transport, in
agreement with what discussed above regards the reduction
in concentration leading to insufficient supply of reactants to
the electrode surface. In the light of these considerations, to
Fig. 4 e EIS spectra of LSTNOH composite using increasing achieve best performances using ammonia it is possible to use
content of NH3 at 900  C. 20% of H2 in Ar is also reported as other flow rate (fuel was kept at 100 sccm): by decreasing the
reference. “Wet” indices the fuels contain 3% of water. flow it is possible to prolong the permeation time of the gas
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 1 3 9 2 1 e1 3 9 3 2 13929

inside the porous structure of the electrode and thus the electrochemical oxidation of hydrogen to achieve best
contact between the fuels (ammonia but also hydrogen) and performances.
the electrocatalyst material of the anode [37]. In general, the
cell performances are strongly affected by the thickness of the
electrolyte, hereby about 1.5 mm. Diminishing the electrolyte Declaration of competing interest
thickness will improve significantly the power density; further
work is in course to this purpose. The authors declare that they have no known competing
Recently, Song et al. [56] studied LSTN-based material financial interests or personal relationships that could have
infiltrated on SDC (Samarium-doped Ceria) scaffold achieving appeared to influence the work reported in this paper.
best performances at 800  C in hydrogen and ammonia,
338 mW/cm2 and 161 mW/cm2 respectively. Their much more
competitive results can be in part justified by the thinner Appendix A. Supplementary data
electrolyte (350 mm). Moreover, perovskite infiltration exhibits
a promising solution that could be investigated in a subse- Supplementary data to this article can be found online at 10.
quent work. Molouk et al. [39] focused their work on Ni-YSZ 1016/j.ijhydene.2022.02.133
and Ni-GDC. Ni-based cermets demonstrate an interesting
catalytic activity towards the ammonia cleavage in hydrogen
and nitrogen. Since the basic sites promote the ammonia
references
decomposition, in Ni-GDC they are more abundant, thus the
activity is enhanced. Consequently, electrochemical proper-
[1] Haile SM. Materials for fuel cells. Mater Today 2003;6:24e9.
ties suggest that the polarization resistance in ammonia is
https://doi.org/10.1016/S1369-7021(03)00331-6.
lower for Ni-GDC in 600e800  C range but becomes almost the
[2] Minh NQ. Solid oxide fuel cell technology - features and
same at 900  C and above. In the light of this results, LSTNOH applications. Solid State Ionics 2004;174:271e7. https://
e based electrode works as anode in ammonia-fed SOFC, doi.org/10.1016/j.ssi.2004.07.042.
however some improvements needed. Since the promising [3] Singhal S, Kendall K. High-temperature Solid oxide fuel cells:
catalytic activity with GDC, it deserves a deeper investigation fundamentals, design and applications. 1st ed. Elsevier Ltd;
playing the role of in the composite with LSTNOH as well as 2003.
[4] Abdalla AM, Hossain S, Azad AT, Petra PMI, Begum F,
the electrolyte.
Eriksson SG, Azad AK. Nanomaterials for solid oxide fuel
cells: a review. Renew Sustain Energy Rev 2018;82:353e68.
https://doi.org/10.1016/j.rser.2017.09.046.
Conclusions [5] Abdalla AM, Hossain S, Nisfindy OB, Azad AT, Dawood M,
Azad AK. Hydrogen production, storage, transportation and
Ni-doped lanthanum strontium titanate has been successfully key challenges with applications: a review. Energy Convers
synthetized via citrate method. Ni, firstly incorporated in Manag 2018;165:602e27. https://doi.org/10.1016/
j.enconman.2018.03.088.
perovskite crystal structure, exsolved in reducing condition
[6] Züttel A. Materials for hydrogen storage. Mater Today
decorating the surface with highly dispersed metal nano- 2003;6:24e33. https://doi.org/10.1016/S1369-7021(03)00922-2.
particles. Catalytic tests exhibit ammonia decomposition [7] Mcintosh S, Gorte RJ. Direct hydrocarbon solid oxide fuel
starts to crack at about 550  C and it is complete at about cells. Chem Rev 2004;104:4845e66. https://doi.org/10.1021/
720  C both using LSTNO or LSTNOH. Ni nanoparticles e based cr020725g.
electrode show lower polarization resistance compared to the [8] Carollo G, Garbujo A, Bedon A, Ferri D, Natile MM, Glisenti A.
Cu/CGO cermet based electrodes for symmetric and
counterpart with Ni infiltration on LSTO composite scaffold.
reversible solid oxide fuel cells. Int J Hydrogen Energy
The progressively similar results between ammonia and
2018;45:13652e8. https://doi.org/10.1016/
hydrogen e fuelled SOFC prove the feasibility of ammonia as a j.ijhydene.2018.01.201.
suitable hydrogen carrier. This trend can be likely attributed [9] Khan MS, Lee SB, Song RH, Lee JW, Lim TH, Park SJ.
to the increment of titanates-based materials by rising tem- Fundamental mechanisms involved in the degradation of
perature in reducing environment. Full cell tests show a power nickeleyttria stabilized zirconia (NieYSZ) anode during solid
density peak of 3.7 mW/cm2 at 780  C using 40% of ammonia. oxide fuel cells operation: a review. Ceram Int 2016;42:35e48.
https://doi.org/10.1016/j.ceramint.2015.09.006.
In the same condition, hydrogen exhibits 11.7 mW/cm2. Both
[10] Rathore SS, Biswas S, Fini D, Kulkarni AP, Giddey S. Direct
power density and OCV are affected by nitrogen-diluting ef- ammonia solid-oxide fuel cells: a review of progress and
fect in the fuel. Varying the flow rate could be a suitable so- prospects. Int J Hydrogen Energy 2021;46:35365e84. https://
lution to increment the performances because it prolongs the doi.org/10.1016/j.ijhydene.2021.08.092.
contact time of the fuel with the electrocatalyst anode. [11] Lan R, Irvine JTS, Tao S. Ammonia and related chemicals as
Reducing the thickness of the electrolyte the cell perfor- potential indirect hydrogen storage materials. Int J Hydrogen
mances will improve. This contribution demonstrates that Energy 2012;37:1482e94. https://doi.org/10.1016/
j.ijhydene.2011.10.004.
LSTO e based materials can be fed by ammonia as potential
[12] Valera-medina A, Xiao H, Owen-Jones M, David WIF,
hydrogen carrier for SOFC application and that exsolution is a Bowen PJ. Ammonia for power. Prog Energy Combust Sci
successful procedure for the deposition of highly active Ni 2018;69:63e102. https://doi.org/10.1016/j.pecs.2018.07.001.
nanoparticles. In addition, the choice of Ni alloys would [13] Liu H. Ammonia synthesis catalysts: innovation and
improve the ammonia decomposition to balance the NH3 practice. World Scientific; 2013. https://doi.org/10.1142/8199.
adsorption and N2 desorption process and the consequent First.
13930 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 1 3 9 2 1 e1 3 9 3 2

[14] Humphreys J, Lan R, Tao S. Development and recent progress [32] Lee JH, Moon H, Lee HW, Kim J, Kim JD, Yoon KH.
on ammonia synthesis catalysts for HabereBosch process. Quantitative analysis of microstructure and its related
Adv. Energy Sustain. Res. 2021;2:2000043. https://doi.org/ electrical property of SOFC anode, Ni-YSZ cermet. Solid State
10.1002/aesr.202000043. Ionics 2002;148:15e26. https://doi.org/10.1016/S0167-2738(02)
[15] Zamfirescu C, Dincer I. Using ammonia as a sustainable fuel. 00050-4.
J Power Sources 2008;185:459e65. https://doi.org/10.1016/ [33] Shri Prakash B, Senthil Kumar S, Aruna ST. Properties and
j.jpowsour.2008.02.097. development of Ni/YSZ as an anode material in solid oxide
[16] Jeerh G, Zhang M, Tao S. Recent progress in ammonia fuel fuel cell: a review. Renew Sustain Energy Rev 2014;36:149e79.
cells and their potential applications. J Mater Chem A https://doi.org/10.1016/j.rser.2014.04.043.
2021;9:727e52. https://doi.org/10.1039/d0ta08810b. [34] Lu Y, Gasper P, Pal UB, Gopalan S, Basu SN. Improving
[17] Armijo J, Philibert C. Flexible production of green hydrogen intermediate temperature performance of Ni-YSZ cermet
and ammonia from variable solar and wind energy: case anodes for solid oxide fuel cells by liquid infiltration of nickel
study of Chile and Argentina. Int J Hydrogen Energy nanoparticles. J Power Sources 2018;396:257e64. https://
2020;45:1541e58. https://doi.org/10.1016/ doi.org/10.1016/j.jpowsour.2018.06.027.
j.ijhydene.2019.11.028. [35] Ma Q, Peng RR, Tian L, Meng G. Direct utilization of ammonia
[18] Morgan E, Manwell J, McGowan J. Wind-powered ammonia in intermediate-temperature solid oxide fuel cells,
fuel production for remote islands: a case study. Renew Electrochem. Commun Now 2006;8:1791e5. https://doi.org/
Energy 2014;72:51e61. https://doi.org/10.1016/ 10.1016/j.elecom.2006.08.012.
j.renene.2014.06.034. [36] Ma Q, Ma J, Zhou S, Yan R, Gao J, Meng G. Short
[19] Kyriakou V, Garagounis I, Vasileiou E, Vourros A, communication A high-performance ammonia-fueled SOFC
Stoukides M. Progress in the electrochemical synthesis of based on a YSZ thin-film electrolyte, vol. 164; 2007. p. 86e9.
ammonia. Catal Today 2017;286:2e13. https://doi.org/ https://doi.org/10.1016/j.jpowsour.2006.09.093.
10.1016/j.cattod.2016.06.014. [37] Stoeckl B, Subotic  V, Preininger M, Schwaiger M, Evic N,
[20] Soloveichik G. Electrochemical synthesis of ammonia as a Schroettner H, Hochenauer C. Characterization and
potential alternative to the HabereBosch process. Nat. Catal. performance evaluation of ammonia as fuel for solid oxide
2019;2:377e80. https://doi.org/10.1038/s41929-019-0280-0. fuel cells with Ni/YSZ anodes. Electrochim Acta
[21] Siddiqui O, Dincer I. A review and comparative assessment 2019;298:874e83. https://doi.org/10.1016/j.electacta.2018.
of direct ammonia fuel cells. Therm Sci Eng Prog 12.065.
2018;5:568e78. https://doi.org/10.1016/j.tsep.2018.02.011. [38] Shy SS, Hsieh SC, Chang HY. A pressurized ammonia-fueled
[22] Fuerte A, Valenzuela RX, Escudero MJ, Daza L. Ammonia as anode-supported solid oxide fuel cell: power performance
efficient fuel for SOFC. J Power Sources 2009;192:170e4. and electrochemical impedance measurements. J Power
https://doi.org/10.1016/j.jpowsour.2008.11.037. Sources 2018;396:80e7. https://doi.org/10.1016/j.jpowsour
[23] Kobayashi H, Hayakawa A, Somarathne KDKA, Okafor EC. .2018.06.006.
Science and technology of ammonia combustion. Proc [39] Molouk AFS, Yang J, Okanishi T, Muroyama H, Matsui T,
Combust Inst 2019;37:109e33. https://doi.org/10.1016/ Eguchi K. Comparative study on ammonia oxidation over Ni-
j.proci.2018.09.029. based cermet anodes for solid oxide fuel cells. J Power
[24] Farr RD, Vayenas CG. Ammonia high temperature solid Sources 2016;305:72e9. https://doi.org/10.1016/
electrolyte fuel cell. Energy Dev Jpn 1981;4:111e28. https:// j.jpowsour.2015.11.085.
doi.org/10.1149/1.2129934. [40] Wan Z, Tao Y, Shao J, Zhang Y, You H. Ammonia as an
[25] Wojcik A, Middleton H, Damopoulos I, Van herle J. effective hydrogen carrier and a clean fuel for solid oxide fuel
Ammonia as a fuel in solid oxide fuel cells. J Power Sources cells. Energy Convers Manag 2021;228:113729. https://doi.org/
2003;118:342e8. https://doi.org/10.1016/S0378-7753(03) 10.1016/j.enconman.2020.113729.
00083-1. [41] Cinti G, Discepoli G, Sisani E, Desideri U. SOFC operating with
[26] Yin SF, Xu BQ, Zhou XP, Au CT. A mini-review on ammonia ammonia : stack test and system analysis, vol. 1; 2016.
decomposition catalysts for on-site generation of hydrogen p. 4e11. https://doi.org/10.1016/j.ijhydene.2016.06.070.
for fuel cell applications. Appl Catal A Gen 2004;277:1e9. [42] Carollo G, Garbujo A, Mauvy F, Glisenti A. Critical raw
https://doi.org/10.1016/j.apcata.2004.09.020. material-free catalysts and electrocatalysts: complementary
[27] Yin SF, Zhang QH, Xu BQ, Zhu WX, Ng CF, Au CT. strategies to activate economic, robust, and ecofriendly
Investigation on the catalysis of COx-free hydrogen SrTiO3. Energy Fuel 2020;34:11438e48. https://doi.org/
generation from ammonia. J Catal 2004;224:384e96. https:// 10.1021/acs.energyfuels.0c01678.
doi.org/10.1016/j.jcat.2004.03.008. [43] Islam QA, Paydar S, Akbar N, Zhu B, Wu Y. Nanoparticle
[28] Yin S, Xu B, Ng C, Au C. Nano Ru/CNTs : a highly active and exsolution in perovskite oxide and its sustainable
stable catalyst for the generation of CO x -free hydrogen in electrochemical energy systems. J Power Sources
ammonia decomposition. Appl Cataly B: Environ 2021;492:229626. https://doi.org/10.1016/
2004;48:237e41. https://doi.org/10.1016/j.apcatb.2003.10.013. j.jpowsour.2021.229626.
[29] Bell TE, Torrente-Murciano L. H2 production via ammonia [44] Kousi K, Tang C, Metcalfe IS, Neagu D. Emergence and future
decomposition using non-noble metal catalysts : a review. of exsolved materials. Small 2021;17. https://doi.org/10.1002/
Top Catal 2016;59:1438e57. https://doi.org/10.1007/s11244- smll.202006479.
016-0653-4. [45] Nishihata Y, Mizuki J, Akao T, Tanaka H, Uenishi M,
[30] Lamb KE, Dolan MD, Kennedy DF. Ammonia for hydrogen Kimura M, Okamoto T, Hamada N. Self-regeneration of a Pd-
storage; A review of catalytic ammonia decomposition and perovskite catalyst for automotive emissions control. Nature
hydrogen separation and purification. Int J Hydrogen Energy 2002;418:164e7. https://doi.org/10.1038/nature00893.
2019;44:3580e93. https://doi.org/10.1016/j.ijhydene.2018. [46] Madsen BD, Kobsiriphat W, Wang Y, Marks LD, Barnett SA.
12.024. Nucleation of nanometer-scale electrocatalyst particles in
[31] Schueth F, Palkovits R, Schloegl R, Su DS. Ammonia as a solid oxide fuel cell anodes. J Power Sources 2007;166:64e7.
possible element in an energy infrastructure : catalysts for https://doi.org/10.1016/j.jpowsour.2006.12.080.
ammonia decomposition. Energy Environ Sci 2012;5:6278e89. [47] Neagu D, Tsekouras G, Miller DN, Me nard H, Irvine JTS. In
https://doi.org/10.1039/c2ee02865d. situ growth of nanoparticles through control of non-
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 1 3 9 2 1 e1 3 9 3 2 13931

stoichiometry. Nat Chem 2013;5:916e23. https://doi.org/ [60] Pen ~ a MA, Fierro JLG. Chemical structures and performance of
10.1038/nchem.1773. perovskite oxides. Chem Rev 2001;101:1981e2017. https://
[48] Neagu D, Kyriakou V, Roiban IL, Aouine M, Tang C, doi.org/10.1021/cr980129f.
Caravaca A, Kousi K, Schreur-Piet I, Metcalfe IS, Vernoux P, [61] Li J, Lv T, Hou N, Li P, Yao X, Fan L, Gan T, Zhao Y, Li Y.
Van De Sanden MCM, Tsampas MN. In situ observation of Molybdenum substitution at the B-site of lanthanum
nanoparticle exsolution from perovskite oxides: from atomic strontium titanate anodes for solid oxide fuel cells. Int J
scale mechanistic insight to nanostructure tailoring. ACS Hydrogen Energy 2017;42:22294e301. https://doi.org/10.1016/
Nano 2019;13:12996e3005. https://doi.org/10.1021/ j.ijhydene.2017.03.189.
acsnano.9b05652. [62] Tsekouras G, Neagu D, Irvine JTS. Step-change in high
[49] Neagu D, Oh TS, Miller DN, Me nard H, Bukhari SM, temperature steam electrolysis performance of perovskite
Gamble SR, Gorte RJ, Vohs JM, Irvine JTS. Nano-socketed oxide cathodes with exsolution of B-site dopants. Energy
nickel particles with enhanced coking resistance grown in Environ Sci 2013;6:256e66. https://doi.org/10.1039/
situ by redox exsolution. Nat Commun 2015;6:1e8. https:// c2ee22547f.
doi.org/10.1038/ncomms9120. [63] Gao Y, Chen D, Saccoccio M, Lu Z, Ciucci F. From material
[50] Irvine JTS, Neagu D, Verbraeken MC, Chatzichristodoulou C, design to mechanism study: nanoscale Ni exsolution on a
Graves C, Mogensen MB. Evolution of the electrochemical highly active A-site deficient anode material for solid oxide
interface in high-temperature fuel cells and electrolysers. fuel cells. Nano Energy 2016;27:499e508. https://doi.org/
Nat Energy 2016;1:1e13. https://doi.org/10.1038/ 10.1016/j.nanoen.2016.07.013.
nenergy.2015.14. [64] Sun Y, Li J, Zeng Y, Amirkhiz BS, Wang M, Behnamian Y,
[51] Neagu D, Papaioannou EI, Ramli WKW, Miller DN, Luo J. A-site deficient perovskite: the parent for in situ
Murdoch BJ, Me nard H, Umar A, Barlow AJ, Cumpson PJ, exsolution of highly active, regenerable nano-particles as
Irvine JTS, Metcalfe IS. Demonstration of chemistry at a point SOFC anodes. J Mater Chem A 2015;3:11048e56. https://
through restructuring and catalytic activation at anchored doi.org/10.1039/c5ta01733e.
nanoparticles. Nat Commun 2017;8:1e8. https://doi.org/ [65] Ding D, Li X, Lai Y, Liu M. Enhancing SOFC cathode
10.1038/s41467-017-01880-y. performance by surface modification through infiltration.
[52] Joo S, Lim C, Kwon O, Zhang L, Zhou J, Wang J, Young H, Energy Environ Sci 2014;7:552e75. https://doi.org/10.1039/
Sin Y, Choi S, Kim G. Materials Reports : energy the first c3ee42926a.
observation of Ni nanoparticle exsolution from YSZ and its [66] Marcilly C, Courty P, Delmon B. Preparation of highly
application for dry reforming of methane. Mater. Reports dispersed mixed oxides and oxide solid solutions by
Energy. 2021;1:100021. https://doi.org/10.1016/ pyrolysis of amorphous organic precursors. J Am Ceram Soc
j.matre.2021.100021. 1970;53:56e7. https://doi.org/10.1111/j.1151-
[53] Błaszczak P, Łapin  ski M, Wang S-F, Jasin
 ski P, Bochentyn B. 2916.1970.tb12003.x.
Exsolution of Ni nanoparticles on the surface of cerium and [67] Neagu D, Irvine JTS. Enhancing electronic conductivity in
nickel co-doped lanthanum strontium titanate as a new strontium titanates through correlated A and B-site doping.
anodic layer for DIR-SOFC. Anti-coking potential and H2S Chem Mater 2011:1607e17. https://doi.org/10.1021/
poisoning resistance of the prepared material. Int J Hydrogen cm103489r.
Energy 2020;45:29186e200. https://doi.org/10.1016/ [68] Briggs D. In: X-ray photoelectron spectroscopy (XPS). 2nd ed.
j.ijhydene.2020.07.162. Handb. Adhes; 2005. p. 621e2. https://doi.org/10.1002/
[54] Tan T, Qin M, Li K, Zhou M, Liu T, Yang C, Liu M. In-situ 0470014229.ch22.
exsolved NiFe alloy nanoparticles on Pr0.8Sr1.2(NiFe)O4-d for [69] Shirley DA. High-resolution X-Ray photoemission spectrum
direct hydrocarbon fuel solid oxide fuel cells. Int J Hydrogen of the valence bands of gold. Phys Rev B 1972;5:4709e14.
Energy 2020;45:29407e16. https://doi.org/10.1016/ https://doi.org/10.1103/PhysRevB.5.4709.
j.ijhydene.2020.07.250. [70] Park BH, Choi GM. Ex-solution of Ni nanoparticles in a
[55] He F, Gao Q, Liu Z, Yang M, Ran R, Yang G, Wang W, Zhou W, La0.2Sr0.8Ti1-xNixO3- alternative anode for solid oxide fuel
Shao Z. A new Pd doped proton conducting perovskite oxide cell. Solid State Ionics 2014;262:345e8. https://doi.org/
with multiple functionalities for efficient and stable power 10.1016/j.ssi.2013.10.016.
generation from ammonia at reduced temperatures. [71] NIST XPS database 20, Version 4.1 (Web Version), (n.d.).
Advance energy materials 2021:2003916. https://doi.org/ https://doi.org/10.18434/T4T88K.
10.1002/aenm.202003916. 1e10. [72] Fadley CS. Basic concepts of X-ray photoelectron
[56] Song Y, Li H, Xu M, Yang G, Wang W, Ran R, Zhou W, Shao Z. spectroscopy. In: Brundle AD, Baker CR, editors. Electron
Infiltrated NiCo alloy nanoparticle decorated perovskite spectrosc. Theory, techiques appl; 1978. p. 287. First.
oxide: a highly active, stable, and antisintering anode for [73] Fadley CS, Baird RJ, Siekhaus W, Novakov T. S €m
AL Bergstro
direct-ammonia solid oxide fuel cells. Small 2020;16:1e8. Surface analysis and angular distributions in x-ray
https://doi.org/10.1002/smll.202001859. photoelectron spectroscopy. J Electron Spectrosc Relat
[57] Marina OA, Canfield NL, Stevenson JW. Thermal, electrical , Phenom 1974;4(2):93e137. https://doi.org/10.1016/0368-
and electrocatalytical properties of lanthanum-doped 2048(74)90001-2. ISSN: 0368-2048 , 1873-2526.
strontium titanate. Solid State Ionics 2002;149:21e8. https:// [74] Akimoto W, Fujimoto T, Saito M, Inaba M, Yoshida H,
doi.org/10.1016/S0167-2738(02)00140-6. Inagaki T. Ni e Fe/Sm-doped CeO 2 anode for ammonia-
[58] Neagu D, Irvine JTS. Structure and properties of fueled solid oxide fuel cells. Solid State Ionics 2014;256:1e4.
La0.4Sr0.4TiO3 Ceramics for use as anode materials in solid https://doi.org/10.1016/j.ssi.2013.12.026.
oxide fuel cells. Chem Mater 2010;22:5042e53. https:// [75] Hashinokuchi M, Yokochi R, Akimoto W, Doi T, Inaba M,
doi.org/10.1021/cm101508w. Kugai J. Enhancement of anode activity at Ni/Sm-doped CeO
[59] Hashimoto S, Kindermann L, Poulsen FW, Mogensen M. A 2 cermet anodes by Mo addition in NH 3 -fueled solid oxide
study on the structural and electrical properties of fuel cells. Solid State Ionics 2016;285:222e6. https://doi.org/
lanthanum-doped strontium titanate prepared in air, vol. 10.1016/j.ssi.2015.07.021.
397; 2005. p. 245e9. https://doi.org/10.1016/j.jallcom. [76] Miller DN, Irvine JTS. B-site doping of lanthanum strontium
2004.11.066. titanate for solid oxide fuel cell anodes. J Power Sources
13932 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 7 ( 2 0 2 2 ) 1 3 9 2 1 e1 3 9 3 2

2011;196:7323e7. https://doi.org/10.1016/j.jpowsour.2010. as potential hydrogen electrode material for solid oxide

11.067. electrochemical cells. J Power Sources 2013;223:341e8.
[77] Gotor FJ, Pueyo M, Poyato R, Sayagu MJ. Mechanosynthesis of https://doi.org/10.1016/j.jpowsour.2012.09.062.
Sr1-xLaxTiO3 anodes for SOFCs : structure and electrical [79] Park S, Hyun J, Jeong C, Woong C, Song R, Lee J. Ni e YSZ-
conductivity. J Alloys Compd 2018;763:679e86. https:// supported tubular solid oxide fuel cells with GDC interlayer
doi.org/10.1016/j.jallcom.2018.05.243. between YSZ electrolyte and LSCF cathode. Int J Hydrogen
 C, Delahaye T, Joubert O, Gauthier G. Exsolution of
[78] Arrive Energy 2014;39:12894e903. https://doi.org/10.1016/
nickel nanoparticles at the surface of a conducting titanate j.ijhydene.2014.06.103.