Growth and thermal stability of Sc-doped BaZrO3 thin films deposited on single crystal substrates

Full text

Thin Solid Films 772 (2023) 139803
Available online 15 March 2023
0040-6090/© 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Growth and thermal stability of Sc-doped BaZrO
3
thin lms deposited on
single crystal substrates
Gabriel K. Nzulu
a
, Elena Naumovska
b
, Maths Karlsson
b
, Per Eklund
a
, Martin Magnuson
a
,
Arnaud le Febvrier
a
,
*
a
Department of Physics Chemistry and Biology (IFM), Link¨
oping University, SE-581 83 Link¨
oping, Sweden
b
Department of Chemistry and Chemical Engineering, Chalmers University of Technology, SE-412 96 G¨
otheborg, Sweden
ARTICLE INFO
Keywords:
Perovskite
Temperature annealing
X-ray diffraction
Magnetron sputtering
Thin lms
Oxygen decient oxide
Proton conductor application
ABSTRACT
Thin lms of BaZr
1-x
Sc
x
O
3-x/2
, (0 ≤x ≤0.64), well known as proton conducting solid electrolytes for interme-
diate temperature solid oxide fuel cell, were deposited by magnetron sputtering. X-ray diffraction analysis of the
as deposited lms reveals the presence of single-phase perovskite structure. The lms were deposited on four
different substrates (c-Al
2
O
3
, LaAlO
3
〈100〉, LaAlO
3
〈110〉, LaAlO
3
〈111〉) yielding random, (110)- or (100)-ori-
ented lms. The stability of the as-deposited lms was assessed by annealing in air at 600 ◦C for 2 h. The
annealing treatment revealed instabilities of the perovskite structure for the (110) and randomly oriented lms,
but not for (100) oriented lm. The instability of the coating under heat treatment was attributed to the low
oxygen content in the lm (understoichiometry) prior annealing combined with the surface energy and atomic
layers stacking along the growth direction. An understoichiometric (100) oriented perovskite lms showed
higher stability of the structure under an annealing in air at 600 ◦C.
1. Introduction
Acceptor-doped proton-conducting perovskite type oxides are of
interest because of their potential usage as electrolytes in solid oxide fuel
cells [1–4]. The perovskite ABX
3
structure, with A as large cations, B as
small cations, and X an anions (O, N, F) allows substitution in either of
the three sites. The substitutions of cations of different valences,
compared to the host ion, yield to the possible formation of
oxygen-decient structure in which hydration of the structure can be
achieved by heat treatment in a humid environment [5–7]. During this
process, water molecules dissociate from the gaseous phase into hy-
droxyl groups (OH
−
) occupying nearby oxygen vacancies site and pro-
tons (H
+
) binding to the oxygen of the perovskite present initially in the
perovskite. Through the process of Grotthuss type mechanism [8],
protons diffuse further from the surface to the bulk by migrating from
one oxygen to another in a trapping hopping mechanism. The result is an
increase in material proton concentration which continues until the
oxygen vacancies in the bulk are lled [8,9].
Among perovskite materials, the cubic perovskite BaZrO
3
shows
promise for applications such as in superconductor devices [10], catalyst
in transesterication [11,12], and humidity monitoring sensors [13].
Doping BaZrO
3
has broadened its application areas to photocatalytic
water splitting process and use as electrolyte material in fuel cells [14,
15]. Trivalent cation elements (Sc, Y, Gd) and other transition metals
and metals elements (Co, Ni, In) are some examples of reported dopants
to initiate proton conduction in BaZrO
3
[16–18].
When substituting a proportion of element from a perovskite the
crystallographic structure may deviate from the ideal cubic perovskite
structure (Pm3m space group) and yield to distinct types of disordered
structures. Gold-Schmidt tolerance factor has been widely accepted as a
criterion for perovskite deviation and it is measured by the mismatch
between the average equilibrium A–O and B–O bond lengths. The
tolerance factor can be extended to more complex perovskite when
partial substitutions in the perovskite occurs such as A
1-x
A’
x
BO
3
or AB
1-
x
B’
x
O
3
:
t=(1−x)rA+xrA′+rO

2
√(rB+rO)(1)
t=rA+rO

2
√((1−x)rB+xrB′+rO)(2)
Where r
A
, r
A’,
r
B,
r
B’
and r
O
are the ionic radius of A, A’, B, B’ and O
* Corresponding author.
E-mail address: arnaud.le.febvrier@liu.se (A. le Febvrier).
Contents lists available at ScienceDirect
Thin Solid Films
journal homepage: www.elsevier.com/locate/tsf
https://doi.org/10.1016/j.tsf.2023.139803
Received 4 October 2022; Received in revised form 6 March 2023; Accepted 11 March 2023

Thin Solid Films 772 (2023) 139803
2
ions in the perovskite, respectively. A cubic symmetry is generally ob-
tained for a tolerance factor in the range of 0.95 ≤t ≤1.04. if t <0.95 or
t >1.04, the symmetric perovskite structure tends to reduce in a highly
distorted symmetry (orthorhombic, tetragonal, trigonal structures). In
the case of BaZrO
3
, which has a nearly optimal tolerance factor of t =
1.01, the substitution of Zr (r
Zr(+IV)
=0.720 Å) by Sc (r
Sc(+III)
=0.745 Å)
[19] gives an increase of the tolerance factor of 1.02 for x =0.65 in
BaZr
1-x
Sc
x
O
3-x/2
. Nevertheless, the Gold-Schmidt tolerance factor has its
limitation especially in the case of decient oxygen perovskite, with the
insertion of different valence elements, not considered in the calculation
of the tolerance factor.
Bulk Sc-doped BaZrO
3
exhibited comparable to higher proton con-
ductivity in a medium temperature range (300 – 500 ◦C) as yttrium-
doped ceramic [9,16,20-22]. The proton conductivity is highly depen-
dent on the content of dopants as observed for the heavily doped ones
with Sc (BaZr
0.4
Sc
0.6
O
3
) where the proton conductivity at 400 ◦C was
increased to 0.012 Scm
−1
(for comparison, Y-doped: 0.005 Scm
−1
) [23].
More recent strategies have been proposed to further improve the proton
conductivity by co-doping the different site (A, B, X) of the perovskite
ABX
3
[24].
Fabrication of dense and mechanically resistance ceramic bulk of
BaZrO
3
based material faces challenges. Generally, BaZrO
3
based ma-
terials ceramic require either a high sintering temperature in a range
1200–1700 ◦C and/or bonding material agent to improve the mechan-
ical resistance and increase the proton mobility (ZnO, NiO, TiO
2
[18,
25-28]). The fabrication of the perovskite ceramic material in thin lm
form offers more favorable options in terms of control of the doping
elements and the growth of dense coatings at lower temperature than
the ceramic sintering. The growth of thin lm on a single crystal sub-
strate offers various possibility of preferential orientation which can be
of interest for proton conductivity properties. Highly non-equilibrium
deposition techniques, such as magnetron sputtering, may lead to syn-
thesis of lms with an understoichiometry nature ABO
3-δ
[29,30]. The
stoichiometry in the perovskite is primordial for the proton conduction
properties of the coatings and yields to differences of the local symmetry
in the material or affect the stability of the phases under heat treatment
[31,32]. In contrast to the lm synthesis by sputtering, the
non-stoichiometry nature of the perovskite is never met in bulk synthesis
where the thermodynamic equilibrium is met.
In the present study, thin lm synthesis of Sc-doped BaZrO
3
by
magnetron sputtering from an oxide BaZrO
3
target has been investi-
gated. Four different single crystal substrates were used to promote the
growth of different preferential orientations of the perovskite in the
coating. The lms were grown in a pure argon atmosphere, in a poor
oxygen growth condition where the oxygen source was the oxide BaZrO
3
target. The effect of the insertion of Sc in BaZr
1-x
Sc
x
O
3-x/2
on the phase
formation, composition, phase stability of the coatings has been studied.
The different characteristics of the lm dependent on their preferential
orientation (different substrates) were investigated. Process control and
post-annealing treatment were also investigated to understand and
optimize the formation and stability perovskite BaZr
1-x
Sc
x
O
3-x/2
.
2. Method and characterization
BaZr
1-x
Sc
x
O
3-x/2
thin lms were deposited by magnetron sputtering
on c-plane sapphire, 〈100〉, 〈110〉and 〈111〉oriented LaAlO
3
(LAO)
substrates in an ultra-high vacuum chamber with a 3.30 ×10
−07
Pa
(2.5 ×10
−09
Torr) base pressure at room temperature. A detailed
description of the chamber can be found elsewhere [33]. Three 50 mm
(2 inch) targets were used for depositing the series of samples: a BaZrO
3
ceramic target (from Advanced Engineering Materials Limited, 99.5%)
as the main source of element; a Sc target (Plasmaterials, 99.9%) as the
dopant; and a Ba target (Plasmaterials, 99.5%) to overcome any de-
ciency of barium. Deciency of barium in the perovskite structure
BaZr
1-x
Sc
x
O
3-x/2
may occur during deposition due to the low vapor
pressure of the element and because of the doping amount where the
ratio [Ba]/([Zr]+[Sc]) is aimed to be equal to 1. The 2
μ
m thick lms
were deposited on the four substrates simultaneously in equivalent po-
sition on the holder which was under a constant rotation and maintained
at 600 ◦C. The BaZrO
3
target was operated at a constant rf-power of 80
W (dc equivalent bias) for all synthesized lms while the Sc target was
operated at dc-power between 0 and 65 W. The Ba target was operated
with an rf-power between 22 W and 82 W (dc equivalent bias) to
maintain the [Ba]/([Zr]+[Sc]) ratio close to 1. The detail of the depo-
sition is listed in Table I. Prior to deposition, the substrates were cleaned
with a multistep process using detergent, acetone, and ethanol in soni-
cation bath (detergent / 5 min +2 ×distilled H
2
O / 5 min, +acetone /
10 min, +ethanol / 10 min) [34]. Post deposition, the coatings were
annealing at 600 ◦C for 2 h in air (20 ◦C/min), in a tubular furnace.
A Leo 1550 Gemini scanning electron microscope (SEM) equipped
with an energy dispersive X-ray spectroscopy (EDS) detector operated
voltage of 20 kV was used to measure their compositions and
thicknesses.
The crystal structure of the lms was analyzed using a PANalyical
X’pert powder diffractometer in a Bragg Brentano conguration using a
Cu-K
α
radiation wavelength of 1.5406 ˙
A (45 kV and 40 mA). The
measurement scan was recorded with a constant rotation of the sample
using a X’celerator detector in 1D scanning line mode: 10–80◦2θ range,
0.0084◦step size, equivalent of 19.7 s/step.
3. Results and discussion
Table II presents the composition of the lm evaluated by EDS
measured on the lm deposited on c-plane sapphire substrate. The
composition of the lms on other substrates is assumed to be the same as
they were deposited at the same time in a similar holder position with
constant rotation. The composition presented in table II is normalized to
the perovskite “ABO
3
” stoichiometry formula, considering the B site to
be fully occupied (B=[Zr]+[Sc]) normalized to 1.00. The Sc content
gradually increases from x =0 to x =0.64 in BaZr
1-x
Sc
x
O
3-x/2
when the
power of the Sc target increases from 0 to 60 W. Note here that the ratio
[Ba]/([Zr]+[Sc]) remained close to 1 (0.95 to 1.06) after compensating
the addition of Sc and loss of Ba during the process.
Fig. 1 shows the Bragg-Brentano X-rays diffractogram of the different
BaZr
1-x
Sc
x
O
3-x/2
(0 <x <0.64) deposited on the four substrates. The
most intense and narrower peaks observed on all the XRD patterns
correspond to the substrate peaks: 006 Al
2
O
3
(at 2θ =41.65◦) for c-
Al
2
O
3
; 100, 200, 300 (at 2θ =23.46, 47.96 and 75.11◦) for LAO〈100〉;
110 and 220 (at 2θ =33.39, 70.19◦) for LAO〈110〉; and 111 (at 2θ =
41.27◦) for LAO〈111〉substrates. All peaks issued from the lm are
identied as the perovskite structure in good agreement with the BaZrO
3
reference ICDD data (00–006–039). Depending on the substrate, the
most intense peak from BaZrO
3
phase differs from a randomly oriented
sample revealing preferential orientations of the lm. On LAO〈111〉,
only the hhh peaks are present for a preferential orientation of the
coating along the [111] direction of the perovskite structure. On the
other three substrates, the most intense peaks are hh0 for c-plane sap-
phire and LAO〈110〉, and h00 for LAO〈100〉substrate with a preferential
orientation of the lm along [110] or [100] direction, respectively.
Depending on the Sc content, other peaks with lower intensities are
Table I
Deposition parameters for all samples. BaZrO
3
and Ba target are powered using a
Rf discharge (values as for dc equivalent) and Sc target is powered using a dc
discharge.
Sample
#
Deposition time
(min)
BaZrO
3
target
power (W)
Ba target
Power (W)
Sc target
Power (W)
1 255 80 23 0
2 235 80 45 35
3 220 80 46 25
4 205 80 60 35
5 155 80 66 60
G.K. Nzulu et al.

Thin Solid Films 772 (2023) 139803
3
observed corresponding to secondary orientations or a randomly
orientated character of the grain. Within a substrate series and for all
substrate, the addition Sc at x =0.2 in the perovskite yield to a more
pronounced formation of either a secondary orientation or randomly
orientated character of the grain. For example, on LAO〈110〉when the
lm has a (110) preferential orientation, the addition of Sc at x =0.20
initiated the growth of (111) secondary orientation. This observation is
similar on the LAO〈100〉and c-plane sapphire substrates. For the highest
Sc content inserted (x =0.64), a clear peak splitting is observed on the
most intense XRD peak from BaZrO
3
on LAO〈110〉and LAO〈111〉. The
separation of the main diffraction peak into two peaks is not clearly
distinguishable for lower Sc content and on other substrates, but dif-
ferences of the XRD peak shape is noticeable for higher reections
already for Sc at x 0.36 (Fig. 1b and 1c). On LAO〈100〉, the XRD pattern
of the (100) lm with high level of Sc present one extra peak at 2θ =
66.90◦not initially existing in the ICDD data of BaZrO
3
. This reection is
linked to the 300 peak with a d =1.396 Å equal to d
100
/3 (2θ =21.20◦;
d
100
=4.187 Å). This reection is not existent for BaZrO
3
(JCPDS card
no. 06–0399) but not forbidden by the Pm3m space group. Note that the
apparition of the 300 reection is predominant for x ≥0.36 but seems to
emerge already for x =0.20 with a low intensity. This phenomenon is
connected to a change in the structure factor due to the substitution of Sc
Table II
Normalized composition of the lms calculated from quantitative EDS results
and normalized to the perovskite Ba(Zr
1-x
Sc
x
)O
3-x/2
with the B site of the
perovskite normalized to 1.00 ([Zr] +[Sc] =1).
Sample
#
Metallic content in the perovskite
ABO
3
normalized with B =1.00
Sc composition (x) in BaZr
1-
x
Sc
x
O
3-x/2
Ba Zr Sc
1 1.01 ±
0.05
1.00 0.00 0.00
2 0.95 ±
0.05
0.80 ±
0.04
0.20 ±
0.04
0.20
3 1.02 ±
0.04
0.64 ±
0.04
0.36 ±
0.03
0.36
4 1.05 ±
0.04
0.56 ±
0.04
0.44 ±
0.03
0.44
5 1.06 ±
0.04
0.36 ±
0.03
0.64 ±
0.03
0.64
Fig. 1. XRD θ−2θ patterns of the BaZr
1-x
Sc
x
O
3-x/2
(0 <x <0.64) lms deposited on c-Al
2
O
3
substrate, 〈100〉, 〈110〉and 〈111〉oriented LaAlO
3
substrates. Peaks
marked with * correspond to the substrate peaks: 0006 (Al
2
O
3
), h00, hh0 and 111 (LaAlO
3
). The reection reference marks are taken from the reference bulk data for
BaZrO
3
from JCPDS card no. 06–0399.e), f), g), h) are short range XRD pattern around the most intense peak from the coating.
G.K. Nzulu et al.

Thin Solid Films 772 (2023) 139803
4
in BaZrO
3
.
Fig. 2 presents the Bragg Brentano θ−2θ patterns of the annealed
BaZr
1-x
Sc
x
O
3-x/2
(0 <x <0.64) lms on the four substrates. The BaZr
1-
x
Sc
x
O
3-x/2
lm with x =0 or x =0.20, presents similar XRD patterns as
the as-deposited sample with no changes on the shapes of the peak.
Nevertheless, on c-AL
2
O
3
, LAO〈110〉and LAO〈111〉, and for x ≥0.36, a
clear splitting of the main orientation peak appears. At higher Sc content
the splitting of the peak is not clearly observed but cannot be excluded as
seen by the asymmetry of the peak compared to BaZrO
3
reference lm
(x =0). Within the series, split and shift of the peaks are noticeable when
the Sc content increases. Note that the intensity of the two peaks
observed was due to the splitting varied in intensity. For the lm on
LAO〈100〉, no splitting is observed within the series, but an increase of
the full width at half maximum (FWHM) is noticeable with high Sc
content.
Fig. 3 presents the evolution of the cell parameter of the as-deposited
and the post-annealed coatings deposited on the four substrates in
regards with the Sc content in BaZr
1-x
Sc
x
O
3-x/2
(0 <x <0.64).
For the as-deposited lms, on LAO〈100〉, the cell parameters have
relatively constant values around 4.26 Å for the range 0 <x <0.64. On
the c-plane sapphire, and LAO〈111〉, on addition of Sc up to x =0.47 in
the perovskite, the cell parameter of the perovskite remained relatively
constant between 4.25 Å and 4.26 Å dependent on the substrate, and Sc
content. Note here that the error bar on the evaluation of the cell
parameter is larger for the lm with x =0.37–0.64 due to the peak
shape. On LAO〈110〉, the cell parameters stayed constant between 4.26
Å and 4.27 Å depending on the Sc content. Except on LAO〈100〉, for x =
0.64, two cell parameter were extracted: a low cell parameter around
4.22 Å and a higher value varying between 4.27 and 4.31 Å
.
After annealing at 600 ◦C for 2 h, two groups of sample and feature
were observed. The cell parameter on the lm deposited on LAO〈100〉
increases monotonically from 4.25 to 4.28 Å when the Sc (x) increases
from 0 to 0.64 in BaZr
1-x
Sc
x
O
3-x/2
. However, on the three other sub-
strates, two cell parameters were extracted for x ≥0.34 in BaZr
1-x
Sc
x
O
3-
x/2
: a large cell parameter following the same trend and values as the one
extracted on LAO〈100〉, and a small cell parameter at around 4.17–4.18
Å constant for all Sc content.
In the present study, the series of lms were studied to provide an
Fig. 2. XRD θ−2θ pattern of the annealed BaZr
1-x
Sc
x
O
3-x/2
(0 <x <0.64) lms deposited on c-Al
2
O
3
substrate, 〈100〉, 〈110〉and 〈111〉oriented LaAlO
3
substrates.
Peaks marked with * correspond to the substrate peaks: 0006 (Al
2
O
3
), h00, hh0 and 111 (LaAlO
3
). The reection reference marks are taken from the reference bulk
data for BaZrO
3
from JCPDS card no. 06–0399. e), f), g), h) are short range XRD pattern around the most intense peak from the coating.
G.K. Nzulu et al.

Thin Solid Films 772 (2023) 139803
5
understanding of the formation and stability of the BaZr
1-x
Sc
x
O
3-x/2
(0 <
x <0.64) for possible usage as proton conductor in a fuel cell. The
capability of a material to absorb H is determined after a heat treatment
in a humid environment. Oxygen stoichiometry and crystallographic
structure stability over a heat treatment are primordial to investigate
prior possible investigation of H-absorption/conduction in BaZr
1-x
Sc
x
O
3-
x/2
. Therefore, the stability of the crystallographic structure of the as-
deposited lm was investigated after a heat treatment in air at 600 ◦C
for 2 h.
The composition measured on the as-deposited coatings shows a
stoichiometry respected within the error bars of the EDS measurement
with a ratio [Ba]/([Sc]+[Zr]) close to 1 (Table II). At rst glance, all
lms are composed of a coating with a perovskite structure isotype to
the cubic BaZrO
3
. The trends and features on the as-deposited lms are
challenging to discuss due to several aspects, such as thermal stress and
stoichiometry in oxygen which play a role on the observations by XRD.
Depending on the substrate, the growth of the perovskite occurred
following specic preferential orientation (Fig. 1). From the relative
intensity of the XRD peaks (cubic BaZrO
3
JCPDS card no. 06–0399), the
coatings have a randomly oriented character on c-plane sapphire. On the
contrary on LAO, preferential orientations are observed with no or few
secondary orientations. LAO material, used as a substrate, is usually
described with a pseudo cubic cell (a =3.79 Å). For the three series on
LAO substrates, the coatings follow the pseudo-cubic structure of the
substrate to provide similar orientation of the perovskite on top: (100),
(110) or (111).
When depositing an oxide material by magnetron sputtering, extra
oxygen is usually brough as a gas to ensure the full oxidation of the
coatings and allowing the formation of the right stoichiometry. In the
present study, the deposition was performed in a pure Ar atmosphere,
which may lead to a lm with a composition in oxygen lower than ex-
pected. The deposition of understoichiometric lm in oxygen by sput-
tering is common due to the highly thermodynamical non equilibrium
process [30,35]. BaZrO
3
oxide target was the source of oxygen, and
more metal atoms were brought by sputtering metal target (Ba and Sc).
Therefore, the possibility to obtain lms understoichiometric in oxygen
is not excluded especially when the power of the metal target was large,
thus at high Sc content. Nevertheless, in the present case, no detection of
other phases (metal, fcc, bcc, hcp) than the perovskite was made using
XRD, but changes in shapes of the peaks could be observed.
The splitting of the peak appearing for the highest Sc content was
observed on hh0 and hhh reections (Fig. 1, Fig. 2) revealing that the
splitting is not due to a tetragonal/orthogonal deformation of the cubic
perovskite but more to a segregation in to two perovskites with possibly
different Sc composition. With the coatings composed of perovskite
structure with lower oxygen content than expected combined with the
possible thermal stress created during deposition make challenging to
discuss the observation.
The heat treatment performed in air helped eliminate the oxygen
vacancies and the thermal stress created during deposition. The XRD of
the annealed lms reveals changes in structural signature, with evidence
of splitting of the peaks occurring for higher Sc content. Except for the
lms on LAO〈100〉, lms on c-Al
2
O
3
, LAO〈110〉, and LAO〈111〉exhibited
the same feature (Figs. 2 and 3). For the coating with x ≥0.36, isotype
structures with two cell parameters are detected (Fig. 3). A cubic
structure with a cell parameter remaining constant around 4.18 ±0.01
Å (0.36 <x <0.64) which is close to the cell parameter of bulk BaZrO
3
(JCPDS card no. 06–0399) [22,36]. The other perovskite phase had a
cell parameter monotonically increasing from 4.26 to 4.28 Å with the Sc
content (0.36 <x <0.64) in a similar fashion way as the lm on
LAO〈100〉(Fig. 3).
The investigation by XRD of the series on LAO〈100〉revealed the
presence of one perovskite phase for which the cell parameter varies
with Sc in agreement with literature [21,37]. The possible presence of
the secondary phase similar as on the other substrates is not entirely
excluded as the FWHM of the diffraction peak increased for higher Sc
content. Nevertheless, the peak remains symmetrical and variation of
FWHM of the XRD peaks can originate from microstress, stress, crystal
domains and thus makes it difcult to conclude. However, the instability
phenomena observed on the coating after annealing is less pronounced
or inexistent on LAO〈100〉.
The deposition conditions in pure argon yield to most probably to the
formation of a perovskite with high oxygen vacancies. Most of the as-
deposited lms contain a single-phase material of BaZr
1-x
Sc
x
O
3-y
(0 <
x <0.64) where y >x/2. An annealing in air reduced the oxygen va-
cancies in BaZr
1-x
Sc
x
O
3-y
where y <x/2 and reduced the stress in the
lm. The heat treatment led to instabilities of the phase previously
contained in the lm. During annealing the lms decomposed into two
phases isotype to perovskite structure. This segregation may be a
consequence of the relaxation of stress and oxidation process of the
perovskite occurring by diffusion of the oxygen in the lm through grain
boundaries and within the grains. The stress relaxation and diffusion of
oxygen during annealing seemed to be inuenced by the lm orienta-
tion. In contrast to the (111) or (110), a (100) oriented lm has low
surface energy of the {001} facets and provides higher stability of the
surface. Unlike [001]-oriented perovskites (ABO
3
) that have alternating
charge neutral A
2+
O
2−
/B
4+
O
2
4−
layers, a [111]-oriented ABO
3
has
alternating A
2+
O
3
6−
/Ti
4+
layers creating a stacking sequence with highly
charged atomic planes. This phenomenon is enhanced by the substitu-
tion in the B site with lower valence element (Sc
3+
→ Zr
4+
). The stacking
sequence yields to the rise of electric dipoles inuencing the growth,
driving the formation of undesired phases and undesired orientations as
the material system attempts to minimize its surface energies [38,39].
Overall, the heat treatment of the high oxygen vacancy containing
perovskite is detrimental for the stability of the initial phases. This
phenomenon is inuenced by the termination surfaces of the coatings,
where instability is reduced or inhibited for a {100} terminated surfaces
coating.
4. Conclusions
BaZr
1-x
Sc
x
O
3-x/2
(0 <x <0.64) lms were deposited on single-crystal
Fig. 3. Cell parameter of BaZr
1-x
Sc
x
O
3-x/2
(0 <
x <0.64) lms deposited on the four substrates.
Reported cell parameter for the a) as-deposited
and b) annealed lms. Cell parameter estimated
from the main orientation of the lm ((110) for
c-Al
2
O
3
and LAO〈110〉, (100) for LAO〈100〉,
and (111) for LAO〈111〉. Note that the presence
of two cell parameters for the same lm/struc-
ture for the highest Sc content corresponds to
the splitting/asymmetry of the peak observable
on the most intense XRD peak from the lm.
G.K. Nzulu et al.

Thin Solid Films 772 (2023) 139803
6
substrates (c-Al
2
O
3
, LAO〈100〉, LAO〈110〉, LAO〈111〉) in oxygen-
decient growth condition using magnetron sputtering. The obtained
lms were composed of a phase isotype as a perovskite structure where,
for high Sc content (x =0.64), the presence of two phases isotype as a
perovskite could be observed by XRD on specic substrates. A thermal
annealing in air at 600 ◦C for 2 h performed on the as-deposited lms
revealed instabilities of the perovskite structured coatings. On most of
the coatings (x ≥0.37), with a (110) orientation or a random orientation
character of the grains, the annealing process led to the formation of two
phases isotype as a perovskite structure (BaZr
1-x
Sc
x
O
3-x/2
) with probably
a perovskite rich in Zr and the one rich in Sc. This instability was found
to be inhibited or reduced for the (100) oriented lm with surface en-
ergy of the {001} facets which were grown on LAO〈100〉.
CRediT authorship contribution statement
Gabriel K. Nzulu: Investigation, Writing – original draft. Elena
Naumovska: Investigation, Writing – review & editing. Maths Karls-
son: Supervision, Writing – review & editing, Funding acquisition. Per
Eklund: Supervision, Writing – review & editing, Funding acquisition.
Martin Magnuson: Supervision, Writing – review & editing. Arnaud le
Febvrier: Project administration, Conceptualization, Supervision,
Writing – review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Data availability
Data will be made available on request.
Acknowledgements
We acknowledge the Swedish Government Strategic Research Area
in Materials Science on Functional Materials at Link¨
oping University
(Faculty Grant SFO-Mat-LiU No. 2009 00971), the Knut and Alice
Wallenberg foundation through the Wallenberg Academy Fellows pro-
gram (KAW-2020.0196), and the Swedish Energy Agency through Grant
No. 48712-1 (E.N., M.K., P. E., A. l. F) and 43606-1(G.K.N., M.M.). M.M.
also acknowledges nancial support from the Carl Trygger Foundation
(CTS20:272, CTS16:303, CTS14:310).
References
[1] H. Iwahara, T. Esaka, H. Uchida, N. Maeda, Proton conduction in sintered oxides
and its application to steam electrolysis for hydrogen production, Solid State Ionics
3-4 (1981) 359–363.
[2] T. Norby, Solid-state protonic conductors: principles, properties, progress and
prospects, Solid State Ionics 125 (1999) 1–11.
[3] K.D. Kreuer, Proton-Conducting Oxides, Annu. Rev. Mater. Res. 33 (2003)
333–359.
[4] N. Ito, M. Iijima, K. Kimura, S. Iguchi, New intermediate temperature fuel cell with
ultra-thin proton conductor electrolyte, J. Power Sources 152 (2005) 200–203.
[5] G.P. Hammond, Energy, Environment and Sustainable Development: a UK
Perspective, Process Saf. Environ. Prot. 78 (2000) 304–323.
[6] J.A. Kilner, M. Burriel, Materials for Intermediate-Temperature Solid-Oxide Fuel
Cells, Annu. Rev. Mater. Res. 44 (2014) 365–393.
[7] L. Malavasi, C.A.J. Fisher, M.S. Islam, Oxide-ion and proton conducting electrolyte
materials for clean energy applications: structural and mechanistic features, Chem.
Soc. Rev. 39 (2010) 4370–4387.
[8] A. Fluri, E. Gilardi, M. Karlsson, V. Roddatis, M. Bettinelli, I.E. Castelli, T. Lippert,
D. Pergolesi, Anisotropic Proton and Oxygen Ion Conductivity in Epitaxial
Ba2In2O5 Thin Films, J. Phys. Chem. C 121 (2017) 21797–21805.
[9] L. Mazzei, M. Wolff, D. Pergolesi, J.A. Dura, L. B¨
orjesson, P. Gutfreund,
M. Bettinelli, T. Lippert, M. Karlsson, Structure and Conductivity of Epitaxial Thin
Films of In-Doped BaZrO
3
-Based Proton Conductors, J. Phys. Chem. C 120 (2016)
28415–28422.
[10] A. Erb, E. Walker, R. Flükiger, The use of BaZrO
3
crucibles in crystal growth of the
high-Tc superconductors Progress in crystal growth as well as in sample quality,
Physica C 258 (1996) 9–20.
[11] V. Singh, B.H. Hameed, Y.C. Sharma, Economically viable production of biodiesel
from a rural feedstock from eastern India, P. pinnata oil using a recyclable
laboratory synthesized heterogeneous catalyst, Energy Convers. Manage. 122
(2016) 52–62.
[12] Y. Yuan, X. Zhang, L. Liu, X. Jiang, J. Lv, Z. Li, Z. Zou, Synthesis and photocatalytic
characterization of a new photocatalyst BaZrO
3
, Int. J. Hydrogen Energy 33 (2008)
5941–5946.
[13] T.A. Blank, L.P. Eksperiandova, K.N. Belikov, Recent trends of ceramic humidity
sensors development: a review, Sens. Actuators B 228 (2016) 416–442.
[14] L. Bi, S.P. Sha, E. Traversa, Y-doped BaZrO
3
as a chemically stable electrolyte for
proton-conducting solid oxide electrolysis cells (SOECs), J. Mater. Chem. A 3
(2015) 5815–5819.
[15] S. Rajendran, N.K. Thangavel, H. Ding, Y. Ding, D. Ding, L.M. Reddy Arava, Tri-
Doped BaCeO
3
–BaZrO
3
as a Chemically Stable Electrolyte with High Proton-
Conductivity for Intermediate Temperature Solid Oxide Electrolysis Cells (SOECs),
ACS Appl. Mater. Interfaces 12 (2020) 38275–38284.
[16] C.Y. Regalado Vera, H. Ding, D. Peterson, W.T. Gibbons, M. Zhou, D. Ding, A mini-
review on proton conduction of BaZrO
3
-based perovskite electrolytes, Journal of
Physics: Energy 3 (2021), 032019.
[17] R. Pornprasertsuk, O. Kosasang, K. Somroop, M. Horprathum, P. Limnonthakul,
P. Chindaudom, S. Jinawath, Proton conductivity of Y-doped BaZrO
3
: pellets and
thin lms, Solid State Sci. 13 (2011) 1429–1437.
[18] M.K. Hossain, T. Yamamoto, K. Hashizume, Effect of sintering conditions on
structural and morphological properties of Y- and Co-doped BaZrO
3
proton
conductors, Ceram. Int. 47 (2021) 27177–27187.
[19] R.D. Shannon, Revised effective ionic radii and systematic studies of interatomic
distances in halides and chalcogenides, Acta Crystallogr., Sect. A 32 (1976)
751–767.
[20] R.B. Cervera, Y. Oyama, S. Miyoshi, I. Oikawa, H. Takamura, S. Yamaguchi,
Nanograined Sc-doped BaZrO
3
as a proton conducting solid electrolyte for
intermediate temperature solid oxide fuel cells (IT-SOFCs), Solid State Ionics 264
(2014) 1–6.
[21] E. Makagon, O. Kraynis, R. Merkle, J. Maier, I. Lubomirsky, Non-Classical
Electrostriction in Hydrated Acceptor Doped BaZrO
3
: proton Trapping and Dopant
Size Effect, Adv. Funct. Mater. 31 (2021), 2104188.
[22] A. Perrichon, E. Jedvik Granhed, G. Romanelli, A. Piovano, A. Lindman,
P. Hyldgaard, G. Wahnstr¨
om, M. Karlsson, Unraveling the Ground-State Structure
of BaZrO
3
by Neutron Scattering Experiments and First-Principles Calculations,
Chem. Mater. 32 (2020) 2824–2835.
[23] J. Hyodo, K. Kitabayashi, K. Hoshino, Y. Okuyama, Y. Yamazaki, Fast and Stable
Proton Conduction in Heavily Scandium-Doped Polycrystalline Barium Zirconate
at Intermediate Temperatures, Adv. Energy Mater. 10 (2020), 2000213.
[24] J. Gao, Y. Liu, Y. Meng, M. Hu, K.S. Brinkman, Fluoride-Based Anion Doping: a
New Strategy for Improving the Performance of Protonic Ceramic Conductors of
the Form BaZrO
3
, ChemElectroChem 7 (2020) 2242–2247.
[25] P. Babilo, S.M. Haile, Enhanced Sintering of Yttrium-Doped Barium Zirconate by
Addition of ZnO, J. Am. Ceram. Soc. 88 (2005) 2362–2368.
[26] J. Kang, G. Chen, B. Lan, R. Zhang, W. Ali, X. Lu, C. Li, Sintering behavior of Y-
doped BaZrO
3
refractory with TiO
2
additive and effects of its dissolution on
titanium melts, Int. J. Appl. Ceram. Technol. 16 (2019) 1088–1097.
[27] F.J.A. Loureiro, N. Nasani, G.S. Reddy, N.R. Munirathnam, D.P. Fagg, A review on
sintering technology of proton conducting BaCeO
3
-BaZrO
3
perovskite oxide
materials for Protonic Ceramic Fuel Cells, J. Power Sources 438 (2019), 226991.
[28] S. Imashuku, T. Uda, Y. Awakura, Improvement in Sintering of Barium Zirconate
by Doping with Scandium, ECS Trans. 7 (2007) 2321–2329.
[29] Y. Zhang, H. Sun, W. Chen, A brief review of Ba(Ti
0.8
Zr
0.2
)O
3
-(Ba
0.7
Ca
0.3
)TiO
3
based lead-free piezoelectric ceramics: past, present and future perspectives,
J. Phys. Chem. Solids 114 (2018) 207–219.
[30] R. Dittmann, 9 - Stoichiometry in epitaxial oxide thin lms, in: G. Koster,
M. Huijben, G. Rijnders (Eds.), Epitaxial Growth of Complex Metal Oxides,
Woodhead Publishing, 2015, pp. 231–261.
[31] J. Ding, J. Balachandran, X. Sang, W. Guo, G.M. Veith, C.A. Bridges, C.M. Rouleau,
J.D. Poplawsky, N. Bassiri-Gharb, P. Ganesh, R.R. Unocic, Inuence of
Nonstoichiometry on Proton Conductivity in Thin-Film Yttrium-Doped Barium
Zirconate, ACS Appl. Mater. Interfaces 10 (2018) 4816–4823.
[32] C.A. Fuller, Q. Berrod, B. Frick, M.R. Johnson, M. Avdeev, J.S.O. Evans, I.R. Evans,
Oxide Ion and Proton Conductivity in Highly Oxygen-Decient Cubic Perovskite
SrSc
0.3
Zn
0.2
Ga
0.5
O
2.4
, Chem. Mater. 32 (2020) 4347–4357.
[33] A. le Febvrier, L. Land¨
alv, T. Liersch, D. Sandmark, P. Sandstr¨
om, P. Eklund, An
upgraded ultra-high vacuum magnetron-sputtering system for high-versatility and
software-controlled deposition, Vacuum 187 (2021), 110137.
[34] A.l. Febvrier, J. Jensen, P. Eklund, Wet-cleaning of MgO(001): modication of
surface chemistry and effects on thin lm growth investigated by x-ray
photoelectron spectroscopy and time-of-ight secondary ion mass spectroscopy,
J. Vac. Sci. Technol. A 35 (2017), 021407.
[35] P. Ambwani, P. Xu, G. Haugstad, J.S. Jeong, R. Deng, K.A. Mkhoyan, B. Jalan,
C. Leighton, Defects, stoichiometry, and electronic transport in SrTiO3-δ epilayers:
a high pressure oxygen sputter deposition study, J. Appl. Phys. 120 (2016),
055704.
[36] S. Yamanaka, M. Fujikane, T. Hamaguchi, H. Muta, T. Oyama, T. Matsuda, S.-
i. Kobayashi, K. Kurosaki, Thermophysical properties of BaZrO
3
and BaCeO
3
,
J. Alloys Compd. 359 (2003) 109–113.
G.K. Nzulu et al.

Thin Solid Films 772 (2023) 139803
7
[37] D. Kumar, B. Singh, BaZrO
3
and Cs-BaZrO
3
catalysed transesterication of Millettia
Pinnata oil and optimisation of reaction variables by response surface Box-Behnken
design, Renew Energy 133 (2019) 411–421.
[38] S. Schmidt, Y.-.W. Ok, D.O. Klenov, J. Lu, S.P. Keane, S. Stemmer, Microstructure
of Epitaxial SrTiO
3
/Pt/Ti/Sapphire Heterostructures, J. Mater. Res. 20 (2005)
2261–2265.
[39] S. Chakraverty, A. Ohtomo, M. Okude, K. Ueno, M. Kawasaki, Epitaxial Structure of
(001)- and (111)-Oriented Perovskite Ferrate Films Grown by Pulsed-Laser
Deposition, Cryst. Growth Des. 10 (2010) 1725–1729.
G.K. Nzulu et al.