The influence of the method of addition of Al2O3 to 3YSZ material on its thermal and electrical properties

Article abstract of DOI:10.1007/s10973-014-4097-4

The properties of tetragonal yttria-stabilized zirconia composites with the small addition of alumina (Al₂O₃-Y₂O₃-ZrO₂ composite) obtained on two ways of synthesis were studied in terms of usability f

Full text of DOI:10.1007/s10973-014-4097-4

J Therm Anal Calorim (2014) 118:1345–1353
DOI 10.1007/s10973-014-4097-4
The influence of the method of addition of Al2O3 to 3YSZ material
on its thermal and electrical properties
_ _
Ewa Drozdz
Received: 4 March 2014 / Accepted: 15 August 2014 / Published online: 4 September 2014
Ó The Author(s) 2014. This article is published with open access at Springerlink.com
Abstract The properties of tetragonal yttria-stabilized zir-
conia composites with the small addition of alumina (Al2O3-
Y2O3-ZrO2 composite) obtained on two ways of synthesis
were studied in terms of usability for anode materials in solid
oxidefuelcell.Bothmethodswerebasedoncitricsynthesis:in
the first one, Al2O3 was coprecipitated with the tetragonal
ZrO2 in the form of citrate by citric acid, while in the second
Al2O3 was impregnated in the form of aluminium nitrate
precursor on tetragonal ZrO2 matrix. The obtained materials
were analysed by X-ray diffraction, dilatometry and imped-
ance spectroscopy. The results of measurements show that
regardless of synthesis method, the addition of Al2O3 influ-
ences the conductivity of samples by increasing their grain
boundaries conductivity as an effect of removing of SiO2 and
decreasing of conductivity activation energy. The impregna-
tion of Al2O3 on tetragonal ZrO2 and sintering of this material
above shrinking temperature cause, however, radical decrease
of porosity of materials, which disqualifies these samples as
anode materials. In the case of samples obtained by copre-
cipitation the significant decrease of porosity is not observed.
electrolyte or anode composite material for solid oxide fuel cell
(SOFC). Both, cubic (8 mol% Y2O3 in ZrO2-8YSZ) and
tetragonal (3 mol% Y2O3 inZrO2-3YSZ) forms are commonly
used, but while only 3YSZ has good enough mechanical and
thermal properties to be a support for power cell, it reveals—due
to so calling blocking effect on grain boundaries—too small
ionic conductivity in comparison to that of the cubic type.
One way to improve the ionic conductivity of the 3YSZ
can be an addition of alumina to zirconia matrix. This effect
was analysed in the past several years mainly for dense
composite materials, both 8YSZ [1–3] and 3YSZ [4–7] as the
matrix for alumina. Al2O3 addition can improve electrical
properties of material by modification of surface conduc-
tivity of YSZ. Moreover, the uniform distribution of Al2O3 in
YSZ matrix can prevent abnormal grain growth and refine
the matrix grains and as the consequence, control low-tem-
perature degradation of mechanical properties [8].
It is commonly known that solubility of Al2O3 in YSZ
below 1,473 K is negligible. An experimental data pre-
sented by Bannister [9] and thermodynamic calculations
[10] confirm this behaviour of Al2O3. This problem was,
however, mainly examined for 8YSZ. Tekeli et al. [11]
stated that the solubility of Al2O3 is equal up to 0.3 mass%
for materials sintered at 1,573 K. Feighery and Irvine [3]
provided value 1.0 mass% in case of material sintered at
1,773 K. These authors explained also a mechanism of
conductivity changes connected with the addition of Al2O3
(Eq. 1). As they stated, on the one hand, Al2O3 dissolved
into ZrO2 can create the same amount of oxygen vacancies
as the addition of Y2O3 (Eq. 2):
Keywords Composite materials ꢀ Thermal properties ꢀ
Conductivity ꢀ Anode for SOFC ꢀ Synthesis of nanopowders
Introduction
Yttria-stabilized zirconia (YSZ) is the most known ceramic-
oxide material employed as a component of either solid
Y2O3 ! 2Y0Zr þ VꢁOꢁ þ 3OxO
and
ð1Þ
_ _
E. Drozdz (&)
Faculty of Materials Science and Ceramics, AGH University of
Science and Technology, al. A. Mickiewicza 30,
30-059 Krakow, Poland
Al2O3 ! 2Al0Zr þ VOꢁꢁ þ 3OxO
ð2Þ
e-mail: edrozdz@agh.edu.pl
123

1346
E. Droz_dz_
second one (MII) was a deposition of aluminium nitrate as
a precursor of Al2O3 on 3YSZ nanopowder (obtaining by
citric method) and next thermal decomposition of Al(NO3).
The solutions of metals yttrium nitrate and aluminium
nitrate were prepared using analytically pure reagents
[Y(NO3)3ꢀ6H2O and Al(NO3)3ꢀ9H2O provided by Sigma-
Aldrich]. Only the solution of zirconyl nitrate was prepared
differently: this one was obtained with zirconium(IV)
oxychloride octahydrate (Beijing Chemicals Import and
Export Corporation) by precipitation of hydrous zirconia
gel with the use of ammonia. Next, this gel was calcinated
at 973 K for 2 h, and obtained white ZrO2 powder was then
dissolved in concentrated nitric acid (analytical grade,
Sigma-Aldrich).
which results in bulk conductivity increase. On the other
hand, the grain boundary resistivity decreases because
alumina acts as scavenger for SiO2 impurities which are
located on the grain boundaries [12]. Silica, which is a
most common impurity in ZrO2, shows a tendency to form
a glassy layer on the surface of zirconia grains, which
causes decreasing of transport of oxygen ions across grain
boundaries. As a result, the conductivity of grain boundary
is drastically reduced [13]. For small amount of Al2O3, the
bulk effect connected with alumina solubility is competi-
tive to an effect on grain boundary. In case of 3YSZ, for
which solubility of Al2O3 is very small, the effect of
removing SiO2 is the main process.
It is commonly known that synthesis pathway of a given
material can affect its microstructural properties (such as
porosity, specific surface area, size and grain distribution)
and in consequence—its thermal and electrical properties.
The dense composite system Al2O3-YSZ is well-known,
and the correlation of their microstructure with electrical
conductivity was studied earlier [14]. Authors found the
correlation between volume fraction of ZrO2 and Al2O3
and conductivity of composite. A large volume of Al2O3
(which is an insulator) significantly decreases conductivity
of composite. Authors pointed out that homogeneous dis-
tribution of composite elements decreases the percolation
threshold for ZrO2 oxygen vacancy conduction.
The concentrated nitrate solutions of aluminium
(0.556 mol dm-3), yttrium (0.955 mol dm-3 per Y2O3)
and zirconyl (0.814 mol dm-3 per ZrO2) were used in the
synthesis. The composition of solutions was determined by
classical gravimetric analysis. The citric acid was used in
the form of monohydrate (analytical grade, provided by
POCH).
MI
The solutions of aluminium, yttrium and zirconyl nitrates
were mixed in the proper quantity in order to obtain
assumed composition. The calculated amount of solid citric
acid (with 10 mol% excess) was added to the nitrates
solution. The final solution was heated on a hot plate
(around 493 K) and stirred for about 12 h, till the solution
turned into grey–white gel. Then, this gel was decomposed
in laboratory dryer (at 513 K) until brown smoke (origi-
nated from the decomposition of nitrates groups) stopped
liberate. In the next step, the grey powder was heated on
the burner (at around 973 K) in the air for 5 h. As a result,
materials with 0.5, 1.0 and 2.0 mass% of Al2O3 in 3YSZ
were obtained.
The properties of porous (the one of necessary feature
for anode materials) Al2O3-YSZ composites are poorly
examined. The basic widespread way for improving grain
boundary conductivity of ZrO2 at a synthesis stage is
applying the method leading to samples with nanoscale
grain size [15]. One of the most effective methods for
obtaining 3YSZ porous nanopowders is citrate gel syn-
thesis [16, 17], often named as Pechini process.
The main goal of this work was a preparation of porous
Al2O3/3YSZ materials by two different synthesis methods
based on classical citric method differing by way in which
alumina is added to 3YSZ matrix and next the examination
of thermal and electrical properties of obtained samples in
order to determine the more efficient way of synthesis of
materials on anode and electrolyte for SOFC technology.
MII
Ethanol solution of aluminium nitrate was obtained by
dissolving of Al(NO3)3ꢀ9H2O (Sigma-Aldrich) in 96 %
ethanol (provided by POCH). The composition of this
solution was equal to 0.499 mol dm-3, as determined by
classical gravimetric analysis. 3YSZ powder was obtained
by means of citric method (the same way of treatment as
employed in method MI), i.e., the proper volume of water
solutions of yttrium and zirconyl nitrates with citric acid
was mixed, the final solution was heated on a hot plate
(around 493 K) for about 12 h, and the resulting gel was
decomposed in laboratory dryer (at 513 K) and heated on
the burner (at around 973 K) in the air for 5 h.
Experimental
Synthesis of the composite materials
In this paper, properties of materials composed with
tetragonal ZrO2 (3YSZ) and the small addition of Al2O3
(0.5, 1.0 and 2.0 mass%) obtained via two ways of syn-
thesis were examined. The first one (named as MI) was the
method which is well known as the citric method. The
123

The influence of the addition of Al2O3 to 3YSZ material
1347
The powder of 3YSZ was grinded in the mortar and then
added in the proper ratio (in order to obtain the same
composition Al2O3/3YSZ as in MI) to concentrated ethanol
solution of aluminium nitrate. The suspension was stirred
and simultaneously heated at around 333 K until the eth-
anol evaporated. The obtained (still wet) powder was ini-
tially warmed in the dryer (c.a. 353 K for 24 h) and then
sample was calcinated in the furnace on air at 673 K for
1 h in order to decompose aluminium nitrate.
atmosphere containing 5 % H2 with heating rate of
5 K min-1 within the temperature range of 293–973 K.
The density of samples was determined on the basis of
geometrical sizes and mass. The total porosity was deter-
mined by relative geometrical density measurements
assuming that densities of Al2O3 and 3YSZ are equal to
3.98 and 6.10 g cm-3, respectively.
Electrical measurements were carried out on Solartron
SI 1260 Impedance/Gain-Phase Analyzer with the SI 1296
dielectric interface. A flowing gas atmosphere of 10 % H2
in Ar was used. The measurements were performed in
temperature range 293–973 K at the frequencies from
0.1 Hz to 106 Hz with amplitude of the sinusoidal voltage
10 mV. The platinum paste was put on pellets as the
electrode before conductivity measurements. The imped-
ance spectra were analysed using software package
(ZPLOT) enclosed by Solartron.
The same three compositions of materials were obtained
by means of MI and MII methods. Next, the materials
obtained by both methods were initially crushed in mortar,
milled in atritor in isopropyl alcohol, dried in 513 K in air
and calcinated at 673 K for 3 h in heater. Finally, disc
pellets were pressed and sintered at 1,073 K or 1,473 K in
air for 3 h. In case of MII, the ethanol solution of alu-
minium nitrate was applied in order to decrease the
Al(NO3)3 agglomeration on the surface of zirconia grains
[18] and for fast evaporation of solvent.
Results
Apparatus
TG/DTA/EGA analysis
The X-ray diffractograms of fired and sintered powders
were registered using a Philips X’Pert Pro diffractometer
The thermal treatment of precursors of composite materials
is strictly connected with their application as anode in
SOFC. The conditions of preparation should ensure the
highest possible porosity of samples. For this reason, some
of the following requirements should be fulfilled:
(a) application of precursors which decomposed with
generation of the large amount of gaseous products,
(b) using of concentrated solutions and (c) proper thermal
treatment. The last factor involves the use of relatively low
temperature of calcinations of powders (it may prevent
premature growth of grain). At the same time, it is very
important to carry out the calcination process either in
furnace with the flow of air or on the burner. This assures
that material is still in contact with oxygen and simulta-
neously the gaseous products of decomposition and oxi-
dation can naturally leave solid matrix. Conducting a
process in closed oven can lead to pyrolysis process and as
a consequence to formation of carbon as a residue.
˚
(CuKa = 1.5406 A, 2h = 20–90°). The crystallites size of
materials was estimated from the ZrO2 [011] peak broad-
ening. The presence of aluminium in composites was
confirmed by X-ray energy dispersive spectroscopy
(EDAX company apparatus). Thermal analysis methods
were used at two different stages of these studies. First one
to determine conditions of thermal treatment of powders
during synthesis (TG/DTA analysis), and the second one
for setting thermal expansion coefficient for the composite
materials (dilatometry method).
TG/DTA measurements with simultaneously evolving
gaseous analysis (EGA) were performed on SDT 2960 TA
instruments apparatus connected online with mass spec-
trometer ThermoStar QMD 360. All measurements were
carried out in synthetic air with heating rate of
10 K min-1. The samples were placed in platinum cruci-
ble. The reference one was also platinum crucible. The
mass spectrometer was operated with an electron impact
ionizer with energy 0.112 aJ (70 eV). Mass spectra were
recorded for M/z equal to 18, 44 and 46 which correspond
Due to such significant role of calcination conditions,
the thermogravimetric measurements for dry citrate gel
originating from MI were performed. Additional registra-
tion of the mass spectra of the gaseous products involving
during TG/DTA analyses was also helpful.
to ions: H2O?, CO2 and NO2?, respectively. M/q = 46
?
line was chosen to present the emission of nitrate groups in
general (marked as NOx in text).
As one can observe in TG/DTA (Fig. 1) graphs, the
process of dry gel decomposition goes on in the three
stages. The first one, slight loss of mass (around
2.6 mass%) in the range 330–420 K is connected only with
removal of adsorbed water. The second one (around
18.8 mass%), in the range 430–550 K, associated with
evolving of water and a carbon dioxide from sample is a
Dilatometric measurements were carried out by means
of NETZSCH DIL 402 C apparatus, equipped with linear
displacement transducer. Cylindrical samples of diameter
10 mm and thickness around 1 mm were used in the
experiments. The measurements were performed in the Ar
123

1348
E. Droz_dz_
10.0
25
TG
Exo
20
15
10
5
7.5
jH2O
DTA
Exo
5.0
2.5
0.0
jH2O
DTA
jCO2
jNOx
TG
jNOx
400
0
400
600
800
1000
1200
600
800
1000
Temperature/K
Temperature/K
Fig. 1 The example of TG/DTA/EGA curves decomposition of dry
gel obtained in the method MI
Fig. 2 Thermal decomposition of aluminium nitrate in air
atmosphere
result of first step of decomposition of metal complexes.
ˇ
Abakeviciene et al [19] attribute the loss of mass in this
which was created after adding of citric acid to the pre-
cursor’s solution. The DTA peak with maximum around
633 K and corresponding water and carbon oxide ionic
currents maxima are sharp. This demonstrates that both
processes: decomposition of citrate groups and oxidation of
carbon are fast, with big loss of mass.
temperature range and corresponding strong endothermic
DTA peak with maximum near 490 K to the removal of
water from the coordination sphere of the metal complexes
and/or chemisorbed water; however, in presented results
(Fig. 1) there are no any clear peaks. The difference
ˇ
appears because of Abakeviciene’s measurements were
From the point of view of method MII, the calcination
conditions of nitrate–citrate gel should be the same as for
gel obtained in the MI. It was important to apply similar
conditions of synthesis of 3YSZ matrix as in MI method, in
order to compare the influence of the way of addition of
Al2O3 on the properties of obtained composites.
carried out on nitrate–citrate wet gel, while results pre-
sented in Fig. 1 were obtained for dry gel heated earlier at
513 K. The lack of peak is probably a result of compen-
sation effect connecting with endothermic decomposition
of metal complexes and simultaneously combustion of
products. The third, main step of decomposition of citrate
gel falls into temperature range 550–670 K. It is connected
with around 49 mass% loss of mass of sample. The exo-
thermic peak on DTA curve corresponds to this effect,
which demonstrates that decomposition processes are clo-
sely associated with oxidation. The complete decomposi-
tion with simultaneous oxidation of organic remains is
necessary from the point of view of potential application of
synthesized materials. The analysis of mass spectrum
registered during TG/DTA measurement at discussed
temperature range shows the increase of ionic currents
only, coming from water and carbon dioxide. This is in
concordance with the results of TG/DTA curves analysis.
There is another important information on the mass spec-
trum: the ionic current originating from NOx does not
increase and does not have any maximum. It means that the
whole nitrate groups in the form of NOx (which was
observed as brown smokes) were removed with gel during
heating in laboratory dryer.
For the MII, it was very important to determine the
temperature of decomposition of aluminium oxide precur-
sor in air. It follows from TG/DTA/EGA curves (Fig. 2)
that the pure aluminium nitrate nonahydrate decomposes in
a few steps. The first one, at temperature range 293–363 K
corresponds to removal of adsorbed water and the first step
of dehydration. The second one (much more complicated)
in the range 393–653 K is related to two-step dehydration
with simultaneous decomposition of nitrate groups. The
first loss of mass (in the range 293–363 K) is about
3.7 mass%, while the main loss (393–653 K) is about
79 mass%. Theoretical calculations made on the basis of
TG line indicate that the first step corresponds uniquely to
dehydration. The second big loss of mass with accompa-
nying endothermic DTA peak (with maximum around
438 K) seems to be a simple process. However, EGA
analysis shows that the process is much more complicated.
In this range of temperature, the ionic currents corre-
sponding to water and NOx increase, as a result of dehy-
dration and decomposition, which occur at the same time.
Additionally, the both processes are not simple, what is
clearly seen on the mass spectrum.
One can assume, therefore, that calcination can be car-
ried out even in not much higher temperature than 670 K.
On the other hand, temperature of drying wet gel—
513 K—was high enough for decomposition of nitric acid
123

The influence of the addition of Al2O3 to 3YSZ material
1349
respectively. The identification of Al2O3 on the basis of
diffraction patterns analysis is not possible by this method
due to too small amount of aluminium oxide in materials.
The presence of aluminium was confirmed for all materials
obtained with both, MI and MII methods, by means of EDS
analysis (Fig. 4).
1.0 mass% Al2O3/3YSZ Ml
1.0 mass% Al2O3/3YSZ MlI
Tetragonal ZrO2
(ICDD No:81–1544)
Porosity
The powders described above were formed in pellets and
then sintered at 1,073 K or 1,473 K. On the basis of geo-
metrical measurements, the density of sintered body was
determined for all tested compositions of samples. The
calculated values are presented in Table 1. The respective
values of density are similar for both series of materials
sintered at 1,073 K in contrast to the values of density (and
porosity) for materials sintered at 1,473 K. The latter forms
two series of results. The density values obtained for
samples synthesized with method of 3YSZ saturation by
aluminium nitrate (MII) and then sintered at 1,473 K are
much higher than those determined for samples obtained
by means of MI method. As a result, samples synthesized
with MII method demonstrate quite small porosity, which
disqualifies these materials as potential anode material. In
case when the sintering temperature was equal to 1,073 K,
the method of synthesis does not have an influence on
porosity of materials but in the temperature where ZrO2
materials demonstrate a large contraction of volume, syn-
thesis method turned out to be very important. It is known
that the shrinkage of ZrO2 observed in temperature range
of 1,270–1,470 K leads to a considerable decrease of
porosity [8, 20, 21]. However, our research showed that the
8YSZ materials obtained by citrate method and sintered at
1,573 K maintain a high porosity [22]. This result is con-
nected with limitation of mass transport in porous samples
being a result of empty spaces in the material.
20
30
40
50
60
70
80
90
2θ/°
Fig. 3 X-ray diffraction pattern of 1.0 mass% Al2O3–3YSZ materi-
als after sintering at 1,473 K
The important conclusion, from the point view of
preparation conditions is that aluminium nitrate is com-
pletely decomposed below 673 K. On the other hand, the
temperature 513 K (which can be achieved in laboratory
dryer) is too low for complete decomposition of aluminium
nitrate in short period of time.
X-ray analysis
Diffractograms of the powders after calcinations obtained
in MI and MII methods (Fig. 3) confirmed that the final
product is tetragonal ZrO2. As determined from the ZrO2
(011) peak broadening, the crystallites of materials after
calcination have nanometric sizes: 15 3 nm for MI and
13 3 nm for MII. As a result of sintering (in 1,473 K),
crystallites grown and their sizes reach a range of
110 10 for MI and 88
9
for MII materials,
ZrLa
Fig. 4 The EDS spectra for
2.0 mass% Al2O3–3YSZ
samples obtained a with MI,
b MII method
ZrLa
(a)
(b)
Y Lg
Y Lg
Y La
Y La
O Ka
O Ka
ZrLl
Y LI
SiKa
AlKa
ZrLl
AlKa
Y LI
ZrLg
C Ka
ZrLg
C Ka
0.90
1.80
2.70
3.60
4.50
0.90
1.80
2.70
3.60
4.50
Energy/KeV
123

1350
E. Droz_dz_
Table 1 The density and total porosity samples after sintering
0.7
0.6
0.5
0.4
0.3
0.2
1.0 mass% Al2O3/3YSZ Ml
1.0 mass% Al2O3/3YSZ MIl
Sample
Sintering temperature
1,073 K
Sintering temperature
1,473 K
Density/
g cm-3
Total
Density/
Total
porosity/
%
porosity/ g cm-3
%
R2MlI = 0.9992
R2MI = 0.9999
0.5 mass%
Al2O3/3YSZ
MI
2.37 0.12 61.1
2.31 0.13 62.0
2.29 0.12 62.2
2.62 0.11 56.9
2.58 0.13 57.6
2.58 0.15 57.4
4.11 0.21 32.3
1.0 mass%
Al2O3/3YSZ
MI
3.87 0.17 35.3
3.70 0.16 38.6
2.0 mass%
Al2O3/3YSZ
MI
650
700
750
800
850
900
Temperature/K
0.5 mass%
Al2O3/3YSZ
MII
5.61 0.14
5.62 0.18
5.53 0.12
7.93
Fig. 5 Dilatometric plots of the Al2O3/3YSZ materials obtained by
MI and MII methods
conclusions one can assume that the way of adding Al2O3
to 3YSZ has significant influence on thermal and conse-
quently electrical properties of testing samples.
1.0 mass%
Al2O3/3YSZ
MII
7.54
8.10
2.0 mass%
Al2O3/3YSZ
MII
The rest of this paper concerns with the results of
measurements carried out for MI and MII materials sin-
tered at 1,473 K, since a temperature equal to 1,073 K is
not high enough for sintering of the materials good for
anode and dense electrolyte.
According to results presented in Table 1, the addition
of a small amount of aluminium to the tetragonal zirconia
system leads to densification of materials. This effect is
highly connected with methods of addition of Al2O3. It is
less evident in the case of materials sintered at 1,073 K
than for materials sintered at 1,473 K. Moreover, as one
can notice, the method of preparation of Al2O3/3YSZ
materials strongly affects density of samples. Therefore,
the hypothesis that density depends on a way of distribut-
ing Al2O3 in 3YSZ matrix can be formulated. In the case of
materials obtained in MI method, the particles of Al2O3,
due to the coprecipitation with tetragonal ZrO2 and coex-
istence of this species in the form (first sol and next gel),
are uniformly distributed in 3YSZ matrix, while in mate-
rials synthesized by MII particles of Al2O3 are distributed
unregularly depending on a structure of open pores. In the
material obtained by MI, the particles of Al2O3 are located
inside ZrO2 grains, while in material from MII Al2O3
created a layer on ZrO2 grains. In effect, mass transport
processes which occur on the grain boundaries can be
much faster. Taking into account the mechanism of den-
sification of ZrO2 nanopowders in the presence of Al2O3
proposed in literature (liquid phase sintering mechanism),
it confirms that coating of ZrO2 grains by Al2O3 can
decrease sintering temperature of these materials. Thus, we
can draw the conclusion that the densification and simul-
taneously decreasing of porosity of materials obtained by
MII result from a distribution of Al2O3 on the surface of
ZrO2 grains. On the basis of the above results and
Dilatometry
The dilatometric tests for samples obtained with MI and
MII methods (the example for 1.0 mass% of Al2O3 in
Fig. 5) show that thermal properties of these materials are
different from each other. Such conclusions can be drawn
because of the different inclination angles on the curves of
dilatation changes with temperature. The thermal expan-
sion coefficients calculated for samples using the linear
regression approximation are equal: (10.700 0.001) 9
10-6 K-1 and (8.200 0.002) 9 10-6 K-1 for samples
obtained with MI and MII, respectively. In the case of
material obtained by coprecipitation—they are similar to
values of fully dense 3YSZ (10.5 9 10-6 K-1) and are
much lower (like value for fully dense Al2O3=8.4 9
10-6 K-1) in the case of material obtained with MII.
From the point of view of potential application of tested
composites, in case of samples obtained with MII their
thermal properties are too different from properties of pure
3YSZ in order to be used as anodes materials.
Electrical properties
The electrical conductivity is another important property
which has to be taken into account when considering the
application of tested materials. In the case of both groups
123

The influence of the addition of Al2O3 to 3YSZ material
1351
Fig. 6 The Nyquist plots
determined at 973 K
–3×103
3YSZ
0.5 % Al2O3/3YSZ MI
1.0 % Al2O3/3YSZ MI
2.0 % Al2O3/3YSZ MI
0.5 % Al2O3/3YSZ MIl
1.0 % Al2O3/3YSZ MIl
2.0 % Al2O3/3YSZ MIl
–2×103
–1×103
Ω
0
1×103
2×103
3×103
4×103
5×103
6×103
0
Z′/Ω
CPEb
CPEgb
of materials (for the samples synthesized by means of both
methods), a characteristic behaviour for dielectric materi-
als—the decreasing of electrical resistivity with increasing
temperature was observed. Furthermore, the values of total
resistivity of all examined samples are lower than resis-
tivity of pure 3YSZ as can be noticed in Nyquist curve
(Fig. 6). It means that all the samples with addition of
Al2O3 reveal higher conductivity than 3YSZ. While in the
case of MI samples, the decreasing of resistivity with
increasing of aluminium oxide content can be observed for
all compositions, for MII materials this tendency is main-
tained—the sample 1.0 % Al2O3/3YSZ demonstrates the
highest value of resistivity.
Rb
Rgb
Fig. 7 Equivalent circuits used for interpretation of the impedance
spectra
bulk conductivity
2
0.5 % Al2O3/3YSZ MI
1.0 % Al2O3/3YSZ MI
2.0 % Al2O3/3YSZ MI
1
An analysis of the impedance spectra revealed that the
equivalent circuit (Fig. 7) represents well the impedance
data which are presented in Fig. 6. From the point of view
of an Al2O3 role, however, the most important are changes
of grain interior and grain boundaries resistivity. The high-
frequency part of impedance spectrum is related to the
conductivity of zirconia grains ðrbÞ, whereas low-fre-
quency part describes conductivity of grain boundaries
0.5 % Al2O3/3YSZ MIl
1.0 % Al2O3/3YSZ MIl
2.0 % Al2O3/3YSZ MIl
0
3YSZ Ml
–1
σ
À
Á
σ
–2
rgb . The comparison of calculated values of conductivity
rb and rgb versus temperature, for samples with the both
synthesis methods, is presented in Figs. 8 and 9. The
curves show that in the lower temperature, the values of
bulk conductivity are slightly higher for the MII materials
with the exception of composition 0.5 % Al2O3/3YSZ for
which in the high temperature the values of rb are higher
for MI sample. It can be seen from Fig. 9 that in case of all
samples (obtained by the both methods), conductivity of
grain boundaries increases with the temperature growth.
When we consider the curves of dependence of energy
activation (Eact) versus Al2O3 mass% (Fig. 10), it can be
noticed that the addition of Al2O3 reduces values of the
activation energy of electrical conductivity of grain
boundaries in case of the samples synthesized by the both
tested methods. For the materials obtained by MI and MII,
one can observe the lowest values of Eact for samples with
–3
–4
0.0010
0.0011
0.0012
0.0013
1/T/K–1
Fig. 8 The comparison of bulk conductivities of samples sintered at
1,473 K. The measurements carried out at temperature range
773–973 K
the addition of 2.0 mass% of Al2O3 to 3YSZ matrix.
Generally, the values of Eact of rgb for MII sample are
lower than for MI sample with the exception of samples
with content of Al2O3 equal to 2.0 mass% for which the
values of Eact are similar one to each other.
123

1352
E. Droz_dz_
6
1.4×105
grain boundaries
conductivity
R
b Ml 1073 K
Rgb Ml 1073 K
b Ml 1473 K
0.5 mass% Al2O3
0.5 % Al2O3/3YSZ MI
1.0 % Al2O3/3YSZ MI
2.0 % Al2O3/3YSZ MI
0.5 % Al2O3/3YSZ MIl
1.0 % Al2O3/3YSZ MIl
2.0 % Al2O3/3YSZ MIl
3YSZ
1.2×105
1.0×105
R
4
2
Rgb Ml 1473 K
Rb Mll 1073 K
Ω
8.0×104
6.0×104
R
R
gb Mll 1073 K
b Mll 1473 K
Rgb Mll 1473 K
4.0×104
2.0×104
0.0
0
σ
σ
750
800
850
900
950
1000
–2
–4
Temperature/K
Fig. 11 The comparison of values of bulk and grain boundary
resistance for samples 0.5 mass% Al2O3/3YSZ with MI and MII
methods
0.0010
0.0011
0.0012
0.0013
comparison with MI samples. Furthermore, regardless of
sintering temperature, for the same content of Al2O3 in
samples, values of Rb and Rgb for MII samples are lower
than for MI samples.
1/T/K–1
Fig. 9 The comparison of grain boundaries conductivities of samples
sintered at 1,473 K. The measurements carried out at temperature
range 773–973 K
Conclusions
100
grain boundaries
Eakt Ml
The comparison of properties of the materials obtained by
different methods of Al2O3 incorporation in the 3YSZ matrix
confirms how important is a way of synthesis of composite
materials. The presented results show that methods of sam-
ples synthesis become very important especially above
temperature where material shrinks. From the point of view
of porosity, materials obtained by impregnation of Al2O3 on
3YSZ framework using aluminium nitrate as precursor and
sintered at 1,473 K reveals too low values of porosity in
order to be applied as anode materials but at the same time
they are very well sintered (the values of density) in such low
temperature. These results suggest that these methods can be
suitable for obtaining of electrolyte for SOFC. From the
point of view of thermal properties, more useful for obtaining
anode materials is synthesis on the way of citricmethod (with
addition of Al2O3 in 3YSZ framework by coprecipitation
with the all precursors) with sintering at 1,473 K, but from
the point of view of conductivity, the second method MII is
definitely better regardless of sintering temperature of sam-
ples. The calculated values of activation energy of electrical
conductivity of grain boundaries confirmed that applying of
Al2O3 as scavenger for SiO2 impurities is an effective
method for increasing of conductivity of materials.
80
60
40
20
Eakt MIl
akt 3YSZ
E
3YSZ
0.0
0.5
1.0
1.5
2.0
Mass% Al2O3
Fig. 10 The comparison of values of energy activation of grain
boundaries conductivity for 3YSZ and samples with addition Al2O3
obtained by MI and MII methods
It can be noticed, as we compare the bulk and grain
boundary resistivity values for the bodies with the same
contents of Al2O3 sintered at 1,073 and 1,473 K, that MII
samples demonstrate lower values of resistivity than MI
ones, regardless of the sintering temperature (Fig. 11). The
results in this figure coincide with values of Eact—the
lower values of the energy activation of grain boundary
conductivity for MII samples than MI sample, the higher
values of grain boundary conductivity for MII samples in
Acknowledgements This work was supported by the Polish
National Centre of the Science (NCN) under Grant No. 2012/05/B/
ST8/02723.
123

The influence of the addition of Al2O3 to 3YSZ material
1353
Open Access This article is distributed under the terms of the
Creative Commons Attribution License which permits any use, dis-
tribution, and reproduction in any medium, provided the original
author(s) and the source are credited.
11. Tekeli S. The solid solubility limit of Al2O3 and its effect on
densification and microstructural evolution in cubic-zirconia used
as an electrolyte for solid oxide fuel cell. Mater Des.
2007;28:713–6.
12. Butler EP, Drennan J. Microstructural analysis of sintered high-
conductivity zirconia with Al2O3 additions. J Am Ceram Soc.
1982;65:474–8.
13. Tanaka J, Baumard JF, Abelard P. Nonlinear electrical properties
of grain boundaries in an oxygen-ion conductor (CeO2ꢀY2O3).
J Am Ceram Soc. 1987;70:637.
14. Fortulan CA, de Sueza DPF. Microstructural evolution of Al2O3–
ZrO2 composite and its correlation with electrical conductivity.
Mater Res. 1999;2 3:205–10.
References
1. Yang F, Zhao X, Xiao P. Electrical properties of YSZ/Al2O3
composite and YSZ/Al2O3 interface studied by impedance
spectroscopy and finite element modelling. Solid State Ionics.
2010;181:783–9.
2. Thokchom S, Xiao H, Rottmayer M, Reitz TL, Kumar B. Het-
erogeneous electrolyte (YSZ–Al2O3) based direct oxidation solid
oxide fuel cell. J Power Sources. 2008;178:26–31.
15. Mondal P, Klein A, Jaegermann W, Hahn H. Enhanced specific
grain boundary conductivity in nanocrystalline Y2O3-stabilized
zirconia. Solid State Ionics. 1999;118:331–9.
3. Feighery AJ, Irvine JTS. Effect of alumina additions upon elec-
trical properties of 8 mol% yttria-stabilised zirconia. Solid State
Ionics. 1999;121:209–16.
16. To˜nsuaadu K, Zalga A, Beganskiene A, Kareiva A. Thermoan-
alytical study of the YSZ precursors prepared by aqueous sol-gel
synthesis route. J Therm Anal Calorim. 2012;110:77–83.
4. Djurado E, Dessemond L, Roux C. Phase stability of nano-
structured tetragonal zirconia polycrystals versus temperature and
water vapour. Solid State Ionics. 2000;136–137:1249–54.
5. Ross IM, Rainforth WM, McComb DW, Scott AJ, Brydson R.
The role of trace additions of alumina to yttria-tetragonal zirconia
polycrystals (Y-TZP). Scriptia Materialia. 2001;45:653–60.
6. Matsui K, Yoshida H, Ikuhara Y. Phase transformation and grain-
growth kinetics in yttria-stabilized tetragonal zirconia polycrystal
doped with a small amount of alumina. J Eur Ceram Soc.
2010;30:1679–90.
´
17. Drozdz_-Ciesla E, Wyrwa J, Re˛kas M. Properties of Ni/YSZ
cermet materials with addition of Al2O3. J Therm Anal Calorim.
2013;113:425–30.
18. Yin S, Fujishiro Y, Uchida S, Sato T. Characterization of ceria of
ittria co-doped zirconia/alumina composites crystallized in
supercritical methanol. J Supercrit Fluids. 1998;13:363–8.
ˇ
19. Abakeviciene B, Zalga A, Tautkus S, Pilipavicius J, Navickas E,
ˇ
Kareiva A, Tamulevicius S. Synthesis of YSZ thin films by the
ˇ
ˇ
novel aqueous sol–gel citrate-precursor method. Solid State
Ionics. 2012;225:73–6.
7. Obal K, Pe˛dzich Z, Brylewski T, Re˛kas M. Modification of Yt-
tria-doped tetragonal zirconia polycrystal ceramics. Int J Elect-
rochem Sci. 2012;7:6831–45.
8. Magnani G, Brillante A. Effect of the composition and sintering
process on mechanical properties and residual stresses in zirco-
nia–alumina composites. J Eur Ceram Soc. 2005;25:3383–92.
9. Bannister MJ. Development of the SIRO2 oxygen sensor: sub-
solidus phase equilibria in the system ZrO2-Al2O3-Y2O3.
J Aust Ceram Soc. 1982;18:62.
20. Antsiferov VN, Kulmeteva VB, Porozova SE, Korkheleva EG.
Effect of nanodispersed zirconium dioxide on the consolidation
and properties of a zircon-based ceramic. Refract Ind Ceram.
2011;52(2):151–4.
21. Chih-Wei K, Yun-Hwei S, Shaw-Bing W, Huey-Er L, Moo-Chin
W. Effect of Y2O3 addition on the crystal growth and sintering
behaviour of YSZ nanopowders prepared by a sol–gel process.
J Alloy Compd. 2009;472(1–2):186–93.
_ _
22. Drozdz E, Wyrwa J, Re˛kas M. Influence of sintering temperature
10. Fabrichnaya O, Aldinger F. Assessment of thermodynamic
parameters in the system ZrO2-Y2O3-Al2O3. Z Metallkd.
2004;95:27.
and aging on properties of cermet Ni/8YSZ materials obtained by
citric method. Ionics. 2014;19:1733–43.
123