Is YSZ stable in the presence of Cu?† Juan Carlos Ruiz-Morales, * a Jesu ´s Canales-Va ´zquez, * b David Marrero-Lo ´pez, a Juan Pen ˜a-Martı ´nez, a Albert Taranco ´n, * c John T. S. Irvine * d and Pedro Nu ´n ˜ez a Received 30th April 2008, Accepted 5th August 2008 First published as an Advance Article on the web 25th September 2008 DOI: 10.1039/b807378c XRD, XPS and electrochemical studies show clear evidence of chemical interaction under oxidising and reducing conditions between YSZ and CuO, both components of a widely studied anode material for high performance solid oxide fuel cells (SOFCs). The aim of this work is to identify and verify the nature of this interaction and find means to prevent its formation during the fabrication process of a fuel cell. Introduction The efficiency of a SOFC depends on the adequate properties of all the cell elements: electrolyte, anode and cathode. The fuel electrode has important requirements regarding conductivity and catalytic activity, stability in the presence of several gas species, chemical and physical compatibility with the other cell compo- nents and, additionally, a low-cost fabrication process. As the traditional anode material, Ni–YSZ cermet, exhibits some limitations, several alternative anode materials have been proposed in the last few years to develop more efficient SOFCs. Among them, the most promising results correspond to Cu-based cermets 1–11 and perovskite-based materials such as chromite–manganites (LSCM), 12–14 substituted SrTiO 3 , 15–18 or double perovskites Sr 2 MgMoO 6d . 19,20 However, mixed oxides usually demand the use of current collectors due to the fairly low electronic conductivity and therefore Cu-based cermets offer significant advantages. Copper is usually considered as an excellent current collector with a poor catalytic activity towards hydrocarbon oxidation, although there are a number of investigations reporting the catalytic activity of Cu–ZrO 2 -based systems 21,22 for several reactions involving hydrocarbon-derived species such as CO oxidation, 23,24 methanol steam reforming, 25–29 methanol synthesis, 30,31 methanol decomposition, 32 methane oxidation, 33 propene and toluene oxidation 34 and carbon-black oxidation. 35 On the other hand, there are few reports regarding the compatibility between Cu and stabilised zirconias at high temperatures. This is particularly relevant when considering that during SOFC fabrication and operation, an interaction between the electrodes and the electrolyte may occur and consequently, the nature of the reaction product could be a limiting factor in the fuel cell efficiency. Indeed, unexpected secondary phases were found after firing Cu-based cermets with novel ceramic oxide anodes. 11 In this work we have explored the origin of these interactions, their possible effect on the electrochemical performances of Cu-based systems and defined means to prevent them. Experimental X-Ray powder diffraction (XRD) patterns were obtained using a Philips X’Pert Pro diffractometer, equipped with a primary monochromator (CuK a1 ) and a X’Celerator detector. The patterns were performed using 0.016 steps (30 s per step) in the 15–90 2q range. Rietveld refinement of the XRD patterns were performed using FULLPROF software. 36 The fits were performed using a pseudo-Voigt peak-shape function. In the final cycles, the usual profile parameters (scale factors, back- ground coefficients, zero-points, half-width, pseudo-Voigt and asymmetry parameters for the peak-shape) were refined. During the refinement all the atomic parameters (positions and site occupation) were fixed and not refined. The experimental errors presented and determined by Rietveld refinement were smaller than the dimension of the symbols. All graphics related with XRD patterns were performed using WinPLOTR software, 37 and X’Pert HighScore Plus 38 was used for phase identification. Several composites of YSZ–CuO were prepared to test the compatibility in oxidising and reducing conditions. The powders, in an appropriate ratio, were mixed in an agate mortar with acetone, left to dry and fired at different temperatures ranging from 800 C up to 1000 C for 3 hours. Heating and cooling ramp rates were fixed at 5 C min 1 in all cases. The powders were fired on a platinum foil to prevent reaction with the ceramic substrates. The composite powders tested in oxidising conditions were also treated under humidified 5% H 2 –Ar, in the same temperature range for 20 hours, to test the compatibility under reducing conditions. 8 mol% YSZ (Pikem, produced by Daiichi Kigenso KK-Japan) was used as electrolyte and as a composite element. Dense YSZ pellets were obtained after uniaxially pressing YSZ powders at 31 MPa for 1.5 min and further sintering at 1500 C a Dpto. Quı´mica Inorga´nica, Universidad de La Laguna, Avda, Astrofı´sico Francisco Sa´nchez s/n, CP 38200 Tenerife, Spain. E-mail: [email protected]; Fax: +34 922 318461; Tel: +34 922 318464 b Renewable Energy Research Institute, Albacete Science and Technology Park, 02006 Albacete, Spain. E-mail: [email protected]c National Center of Microelectronics CNM-IMB (CSIC), Campus University Autonomous of Bellaterra, 08193 Cerdanyola del Valle`s, Spain. E-mail: [email protected]d School of Chemistry, University of St Andrews, North Haugh, St Andrews, Scotland, UK KY16 9ST. E-mail: [email protected]; Fax: +44 (0)1334 463808; Tel: +44 (0)1334 463817 † Electronic supplementary information (ESI) available: X-Ray diffraction patterns and XPS spectra. See DOI: 10.1039/b807378c 5072 | J. Mater. Chem., 2008, 18, 5072–5077 This journal is ª The Royal Society of Chemistry 2008 PAPER www.rsc.org/materials | Journal of Materials Chemistry
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
PAPER www.rsc.org/materials | Journal of Materials Chemistry
Is YSZ stable in the presence of Cu?†
Juan Carlos Ruiz-Morales,*a Jesus Canales-Vazquez,*b David Marrero-Lopez,a Juan Pena-Martınez,a
Albert Tarancon,*c John T. S. Irvine*d and Pedro Nuneza
Received 30th April 2008, Accepted 5th August 2008
First published as an Advance Article on the web 25th September 2008
DOI: 10.1039/b807378c
XRD, XPS and electrochemical studies show clear evidence of chemical interaction under oxidising and
reducing conditions between YSZ and CuO, both components of a widely studied anode material for
high performance solid oxide fuel cells (SOFCs). The aim of this work is to identify and verify the
nature of this interaction and find means to prevent its formation during the fabrication process of
a fuel cell.
Introduction
The efficiency of a SOFC depends on the adequate properties of
all the cell elements: electrolyte, anode and cathode. The fuel
electrode has important requirements regarding conductivity and
catalytic activity, stability in the presence of several gas species,
chemical and physical compatibility with the other cell compo-
nents and, additionally, a low-cost fabrication process.
As the traditional anode material, Ni–YSZ cermet, exhibits
some limitations, several alternative anode materials have been
proposed in the last few years to develop more efficient SOFCs.
Among them, the most promising results correspond to
Cu-based cermets1–11 and perovskite-based materials such as
chromite–manganites (LSCM),12–14 substituted SrTiO3,15–18 or
double perovskites Sr2MgMoO6�d.19,20 However, mixed oxides
usually demand the use of current collectors due to the fairly low
electronic conductivity and therefore Cu-based cermets offer
significant advantages.
Copper is usually considered as an excellent current collector
with a poor catalytic activity towards hydrocarbon oxidation,
although there are a number of investigations reporting the
catalytic activity of Cu–ZrO2-based systems21,22 for several
reactions involving hydrocarbon-derived species such as CO
propene and toluene oxidation34 and carbon-black oxidation.35
On the other hand, there are few reports regarding the
compatibility between Cu and stabilised zirconias at high
temperatures. This is particularly relevant when considering
that during SOFC fabrication and operation, an interaction
aDpto. Quımica Inorganica, Universidad de La Laguna, Avda, AstrofısicoFrancisco Sanchez s/n, CP 38200 Tenerife, Spain. E-mail: [email protected];Fax: +34 922 318461; Tel: +34 922 318464bRenewable Energy Research Institute, Albacete Science and TechnologyPark, 02006 Albacete, Spain. E-mail: [email protected] Center of Microelectronics CNM-IMB (CSIC), CampusUniversity Autonomous of Bellaterra, 08193 Cerdanyola del Valles,Spain. E-mail: [email protected] of Chemistry, University of St Andrews, North Haugh, StAndrews, Scotland, UK KY16 9ST. E-mail: [email protected]; Fax: +44(0)1334 463808; Tel: +44 (0)1334 463817
† Electronic supplementary information (ESI) available: X-Raydiffraction patterns and XPS spectra. See DOI: 10.1039/b807378c
5072 | J. Mater. Chem., 2008, 18, 5072–5077
between the electrodes and the electrolyte may occur and
consequently, the nature of the reaction product could be
a limiting factor in the fuel cell efficiency. Indeed, unexpected
secondary phases were found after firing Cu-based cermets with
novel ceramic oxide anodes.11
In this work we have explored the origin of these interactions,
their possible effect on the electrochemical performances of
Cu-based systems and defined means to prevent them.
Experimental
X-Ray powder diffraction (XRD) patterns were obtained using
a Philips X’Pert Pro diffractometer, equipped with a primary
monochromator (CuKa1) and a X’Celerator detector. The
patterns were performed using 0.016� steps (30 s per step) in the
15–90� 2q range. Rietveld refinement of the XRD patterns
were performed using FULLPROF software.36 The fits were
performed using a pseudo-Voigt peak-shape function. In the
final cycles, the usual profile parameters (scale factors, back-
ground coefficients, zero-points, half-width, pseudo-Voigt and
asymmetry parameters for the peak-shape) were refined. During
the refinement all the atomic parameters (positions and site
occupation) were fixed and not refined. The experimental errors
presented and determined by Rietveld refinement were smaller
than the dimension of the symbols. All graphics related with
XRD patterns were performed using WinPLOTR software,37
and X’Pert HighScore Plus38 was used for phase identification.
Several composites of YSZ–CuO were prepared to test the
compatibility in oxidising and reducing conditions. The powders,
in an appropriate ratio, were mixed in an agate mortar with
acetone, left to dry and fired at different temperatures ranging
from 800 �C up to 1000 �C for 3 hours. Heating and cooling ramp
rates were fixed at 5 �C min�1 in all cases. The powders were fired
on a platinum foil to prevent reaction with the ceramic
substrates. The composite powders tested in oxidising conditions
were also treated under humidified 5% H2–Ar, in the same
temperature range for 20 hours, to test the compatibility under
reducing conditions.
8 mol% YSZ (Pikem, produced by Daiichi Kigenso
KK-Japan) was used as electrolyte and as a composite element.
Dense YSZ pellets were obtained after uniaxially pressing YSZ
powders at 31 MPa for 1.5 min and further sintering at 1500 �C
This journal is ª The Royal Society of Chemistry 2008
Fig. 1 Experimental XRD pattern for a composite YSZ–CuO (1 : 1
weight), after firing in air, at 950 �C for 3 h, together with the pattern for
the monoclinic ZrO2 obtained from the ICSD database.
for 10 h. CuO used in the composites was from Aldrich (>99%)
and the grain size was <5 mm. Powders of YSZ and CuO were
mixed in an appropriate ratio in addition to a Decoflux (WB41,
Zschwimmer and Schwartz) binder to obtain a slurry, which was
used to attach the electrodes onto the YSZ tape in symmetrical
configuration. The samples were fired at several temperatures for
two hours.
Experimental details for electrochemical measurements can be
found elsewhere.39 The measurements were performed in
symmetrical configuration, under symmetrical atmosphere and
always on cooling.
Two samples were prepared to observe the effect of the
reduction–diffusion zone (under reducing conditions) on the
electrochemical properties of the YSZ electrolyte. Two samples
of dense electrolyte were covered with a YSZ–CuO slurry in
a 4 : 1 ratio, and then fired at 1000 �C for 3 h. After that, one was
fired at 950 �C for 40 hours under humidified 5% H2–Ar and the
second at 700 �C under the same experimental conditions. In
both cases, the cermet anode covering the dense pellet was
removed by polishing the pellet surface, and a dark region where
the anode was in contact with the electrolyte could be observed.
In the case of the sample fired at 950 �C the region seems to go
deep inside the microstructure, whereas for the sample fired at
700 �C sample the dark region was superficial. Both samples and
another YSZ used as a blank were covered with a Pt-ink layer as
current collector. Bulk, grain boundary and total conductivities
were measured by ac impedance spectroscopy.
Optical images of the cross- and top-section of samples were
acquired using a stereomicroscope Leica Zoom 2000.
XPS spectra were obtained on a Physical Electronics
550 spectrometer, using monochromatic AlKa radiation. To
minimise discrepancies due to sample compactness, charging
effects and contact resistance, all the measurement parameters
were verified and adjusted to get the 1s core level of adventitious
carbon at 284.8 eV.
Fig. 2 XRD patterns of a YSZ–CuO (1 : 1 weight) composite, fired in air
in a range of temperatures, for 20 hours each. Arrows indicate m-ZrO2
peaks.
Results and discussions
Preliminary studies in the system CuO–cubic-YSZ showed the
presence of secondary phases, Fig. 1.
The extra reflections observed by XRD corresponded to
monoclinic zirconia, Fig. 1 and 2, which fully agreed with Simner
et al.,40,41 who previously reported the reaction of YSZ with
compounds containing just 2% CuO to yield monoclinic zirconia.
No traces of other impurities were detected.
Simner et al. also suggested that this reaction is very sensitive
to YSZ calcination temperature, e.g. pre-firing YSZ at high
temperature (1200 �C) seemed to inhibit the reaction between
a copper containing oxide and YSZ at lower temperatures.
However in our case, the reaction is not affected by this treat-
ment, (Fig. S1, ESI†).
m-ZrO2 seems to be produced independently of the Cu source,
e.g. (La0.8Sr0.2)0.98Fe0.98Cu0.02O3�d,40,41 CuNb2O6,42 or simply
CuO11,43,44 mixed with YSZ and fired at relatively low tempera-
tures, 900–1000 �C, for 2–3 h produce the same result. One
should note that an intimate mixture of powders is not necessary
to produce the aforementioned reaction. Indeed YSZ powders on
a slightly pressed thin CuO layer, led to the same formation of
m-ZrO2 after firing at 1000 �C for 3 hours.
This journal is ª The Royal Society of Chemistry 2008
Additional studies performed to evaluate the sensitivity of
this reaction with the Cu content and temperature showed that
even the addition of just 1% weight of CuO was enough to
produce monoclinic zirconia, Fig. 3. The amount of monoclinic
zirconia formed did not depend on the CuO concentration
above 5% wt. The firing temperature should be below 900 �C to
avoid YSZ destabilisation, as shown in Fig. 2. This was double
checked by firing, at 800 �C for 120 hours, a composite prepared
by impregnation of YSZ with copper acetate, which decom-
poses to form CuO, and no trace of the monoclinic zirconia was
detected.
On the other hand, the same thermal treatments under mildly
reducing conditions did not result in the formation of m-ZrO2,
(Fig. S2, ESI†). This is clear indication that Cu2+ is responsible of
the formation of m-ZrO2.
J. Mater. Chem., 2008, 18, 5072–5077 | 5073
Fig. 4 Evolution of (a) phase concentration and (b) unit cell volumes
determined by Rietveld refinement as a function of the annealing
temperature. (The multiphase refinements converged to final agreement
factors about: Rwp ¼ 15.5% and Rexp ¼ 7.2%).
Fig. 3 XRD patterns for composites YSZ–CuO, different ratios in
weight, fired in air, at 950, 3 h. (Arrows indicate peaks from the mono-
clinic zirconia).
The origin of the YSZ destabilisation due to the presence of
Cu2+ has not been described in the literature, although there exist
certain possible routes to explain this process. Firstly, CuO may
interact with Y3+ to produce secondary phases liberating m-ZrO2
from the stabilised cubic YSZ. A second route could be via
substitution of Zr4+ in the cubic YSZ for Cu2+.
The reaction between Y and Cu could lead to the formation of
the blue-greenish compound Y2Cu2O5 as observed by Lemaire
et al.45 and Ran et al.46–48 in a similar situation, e.g. when firing
CuO with 3 mol% yttria-stabilised tetragonal zirconia (3Y-TZP)
at temperatures above 850 �C. A long XRD run carried out in
a 1 : 1 YSZ–CuO composite, fired at 975 �C for 3 hours showed
no trace of Y2Cu2O5 (Fig. S3, ESI†). It should be noted
that CuO–YSZ mixture calcined at 1000 �C presents �30 %wt of
m-ZrO2 and if the formation of this phase was due to yttria
depletion to form Y2Cu2O5, this last phase would have been
detected by XRD. Alternatively, the possible formation of
Y2Cu2O5 may be monitored by increasing the segregation of
Y2O3 as the authors suggested, increasing the firing temperature
of YSZ. However, the XRD pattern of a 1 : 1 CuO : YSZ
composite fired at 1000 �C for 3 h with YSZ previously fired 15 h
at 1200 �C, (Fig. S1, ESI†), did not show any trace of Y2Cu2O5.
Therefore, the second option was investigated. Dongare
et al.49,50 and Bhagwat et al.51 reported the stabilisation of cubic
zirconia with Cu in samples containing just 2 mol% Cu, after
firing at 500 �C. Analogous stabilisations of cubic zirconia have
been reported for other transition metals as Mn, Co, Fe to
produce materials with special catalytic properties.52 Similarly,
a possible substitution of Zr4+ for Cu2+ in the YSZ lattice could
allow the formation of Cu-stabilised cubic zirconia, liberating
monoclinic ZrO2 in the process. The variation of the unit cell
volumes and the quantification of the different phases as
a function of the firing temperature were investigated by Rietveld
refinements, Fig. 4. The cell volume for YSZ increases with the
firing temperature above 925 �C, when secondary phases appear.
In addition, the unit cell volume for CuO is independent of
the firing temperature (inset of Fig. 4b) and the m-ZrO2
volume undergoes a slight increase. The concentration of m-ZrO2
increases up to 30% at 1000 �C; on the other hand, the
CuO content in the mixture varies only by 2% between 900
and 1000 �C.
5074 | J. Mater. Chem., 2008, 18, 5072–5077
These results demonstrate that only a small degree of Cu
substitution is required to destabilise cubic YSZ to form m-ZrO2,
which is not entirely surprising if one considers the proximity
of the tetragonal + cubic phase to the cubic phase boundary to
8 mol% yttria in the ZrO2–Y2O3 phase diagram. In this study, the
monoclinic phase has been observed throughout, although it is
quite likely that it is the tetragonal phase that forms at that
temperature and this transforms to monoclinic on processing of
powders. Such a transformation is routinely encountered in
studies of these zirconia systems. This also explains the increase
of YSZ cell volume with increasing the firing temperature as
higher levels of yttria in doped zirconia produce an increase of
the volume cell. The absence of segregated yttria or phases
containing yttria, e.g. Y2Cu2O5, seems to confirm this behaviour.
One should note that the formation of m-ZrO2, which exhibits
a very low conductivity, will affect the ohmic resistance of the
composite electrode material under study. The mechanical
properties of monoclinic zirconia may also cause degradation
through either phase transformation to tetragonal at high
temperatures or through hydrothermal ageing at lower temper-
atures. However the polarisation resistance values do not show
evidence that this has an impact on the performance.
This journal is ª The Royal Society of Chemistry 2008
It has been reported that Cu may diffuse from the anode
microstructure towards the dense electrolyte,8 also in several
other materials such as the superconducting phases
YBa2Cu3O7�x. Vasiliev et al.53 reported and identified the
diffusion of metallic Cu through the grain boundaries of YSZ
during their preparation on Si substrates with YSZ buffer layers
at low temperatures 700–750 �C.
A slurry of CuO:YSZ was painted on a dense YSZ pellet, and
the resulting specimen was then fired at 1000 �C for 3 hours.
After that, it was heated at 950 �C, under wet 5% H2–Ar for
10 hours. The surface of the electrolyte, Fig. 5a and the cross-
section of the sample, Fig. 5b indicate a dark reaction zone. As
the only materials involved were Cu and YSZ, one may assume
that Cu is responsible for such reaction. However, it should
be noted that the composite must be strongly attached to the
electrolyte to observe the aforementioned dark zone. EDAX
analysis performed inside this 100 mm layer do not reveal the
presence of Cu, although the same tests in the anode layer show
clear evidence of Cu. Consequently the Cu content should be
below the detection limit of this technique (typically 5%), and/or
it may occur that Cu promotes YSZ reduction near the interface.
On the other hand, the electrode layer was removed by polishing
and the electrolyte interface in contact with the electrode showed
a very dark colour, Fig. 5c. A XRD pattern of this zone, (inset of
Fig. 5d) indicates the presence of Cu at trace levels.
XPS studies were carried out, in a sample reduced at 950 �C
under humidified 5% H2–Ar, to understand the surface compo-
sition of the Cu–YSZ interface. Photoelectron peaks associated
Fig. 5 Optical pictures of (a) top-view of a cell with Cu–YSZ cermet in
contact with a dense YSZ pellet (1.2 mm thickness), after annealing in
reducing conditions. (b) Magnified view of a cross-section, showing the
anode layer, a reaction zone and the rest of the electrolyte dense pellet. (c)
Top-view of the same sample after removing the electrode by polishing,
revealing a dark region that corresponds with the electrode contact zone.
(d) XRD pattern of the dark zone shown in (c). The dark line corresponds
to the pattern obtained after one polishing process. The cubic-YSZ peaks
appear distorted at lower angles. This distortion almost disappears after
a second polishing process (mid line). The light line corresponds to the
back of the YSZ sample used as reference of free Cu. Just traces of
metallic Cu could be verified from X-ray as shows the inset of Fig. 5d.
This journal is ª The Royal Society of Chemistry 2008
to Cu 2p and Zr 3d core levels were studied on both bulk YSZ
and Cu–YSZ interface. For Cu core level on bulk YSZ no
significant peaks were observed, whilst at the Cu–YSZ interface
a doublet peak was observed corresponding to Cu2p3/2 and
Cu2p1/2 at 932.4 and 952.3 eV respectively, (Fig. S4, ESI†). The
binding energies for the observed copper core lines are charac-
teristic of copper metal and copper(I) oxide. Albeit copper metal
has binding energies very close to those for Cu2O making it is not
possible to distinguish between the two compounds, the presence
of copper(II) oxide can be discarded (the Cu2p3/2 line for CuO is
shifted further 1.3 eV above).54 Quantitative compositional
analysis of the Cu–YSZ interface yields a copper-to-zirconium
ratio (NCu–NZr) around 3%.
Similar spectra were observed for the Zr core level on both
bulk YSZ and Cu–YSZ interface. Two different doublets for
each Zr 3d5/2/Zr 3d3/2 core levels were observed, 181.2/184.2 eV
and 183.3/185.95 eV, respectively, Fig. 6. These two sets of peaks
can be attributed to changes in local chemical and physical
environment or to the presence of mixed valence states for
zirconium, Zr4+ and Zr3+, on the surface of the sample. As
changes in local environment are usually associated to smaller
energy shifts (less than 0.5 eV), the measured shift is more likely
to be related to mixed valence states (probably limited to a very
narrow band close to the surface). Indeed, Pomfret et al.55 have
reported evidence of zirconium reduction on the surface for
yttria-stabilized zirconia after exposing YSZ to a reducing
environment at elevated temperatures; this has also been
observed after exposing a dense YSZ pellet under humidified
5% H2–Ar at 950 �C for 20 h (Fig. S5, ESI†). This reduction of
zirconium could be promoted by copper, as previously reported
for nickel in Ni–ZrO2 plating.56 Hence, a combination of
processes would be the origin of the reaction region, i.e. an
exolution process involving metallic Cu (from the previously
incorporated CuO) and a very superficial partial reduction of
Zr(IV) to Zr(III) promoted by Cu.
Additional studies were carried out to reveal the effect of this
reduction zone in the electrochemical properties of dense YSZ
electrolytes, Fig. 7. The higher the temperature during the
reduction process, the lower the conductivity of the YSZ will be,
Fig. 7. Both, bulk and grain boundary contributions are deeply
affected by the reaction zone, Fig. 7a and 7b. In general the bulk
and grain boundaries conductivities are between �20% and
�30% lower than the values corresponding to the YSZ blank, for
Fig. 6 XPS core level spectra of Zr 3d in the apparent reaction zone.
J. Mater. Chem., 2008, 18, 5072–5077 | 5075
Fig. 7 (a) Impedance spectra, (b) overall conductivity and (c) bulk and grain boundary conductivities of three dense YSZ samples, one of them used as
a blank. The other two were previously covered with a slurry of YSZ–Cu-based anode (80 : 20 % weight) and fired at 1000 �C in air for 3 h. After that they
were fired under reducing conditions (humidified 5% H2–Ar) at 950 �C and 700 �C. The activation energies of each process match very well the values of
the YSZ used as reference, indicating that the main charge-transport mechanism remains unaffected by the reaction zone.
the sample treated at 700 �C under H2. At 950 �C the degradation
is more noticeable, i.e. an increase of �50% and 100% in the
resistance values was observed. Hence even firing at low
temperatures as 700 �C and with just an apparent superficial
reaction zone, the effect in the conductivity of the YSZ is very
important and should not be neglected. Moreover, one should
remark that these results refer to very thick electrolytes and
therefore the effect of this reaction may be much worse in thin
YSZ electrolytes.
This is entirely consistent with what might be considered
feasible redox chemistry and seems clear evidence that CuO
exolves Cu metal. It seems that as this incorporation of CuO and
exolution of Cu releases further secondary phases and under
kinetically controlled series of redox processes, one might expect
this not to be reversed and so secondary phases would build up,
even at temperatures below 950 �C.
In fact, the irreversibility of this degradation was experimen-
tally verified. The sample fired at 700 �C, was heated again under
oxidising conditions and the resistances were always higher than
the values obtained in the previous cycle. It should be expected
that long time operation would produce a deeper reaction zone
inside the pellet, resulting in poorer conductivity values with the
5076 | J. Mater. Chem., 2008, 18, 5072–5077
time. This effect may explain the recently published results by
Gorte et al.9,10 They found that annealing a Cu-impregnated
YSZ-based composite under reducing conditions at temperatures
above 700 �C produced a marked degradation of the fuel cell
performance under study. These results were explained by the
authors as due to the agglomeration of Cu particles. We think
that our findings regarding the existence of a reaction zone in the
electrode–electrolyte interface, even at very superficial level,
could partly explain such degradation in the electrochemical
response.
Conclusions
In summary, it has been shown that CuO reacts with YSZ under
oxidising conditions and Cu is then exolved under reducing
conditions even at low temperatures.
Firing YSZ with a material containing Cu2+ always results in
the formation of m-ZrO2 even when the CuO content was as low
as 1% in weight, though this may be avoided by firing under
reducing conditions. XRD analysis suggests that Cu2+ destabil-
ises the YSZ, resulting in the formation of Y-rich cubic and
This journal is ª The Royal Society of Chemistry 2008
monoclinic zirconia, ruling out the formation of Y2Cu2O5, at
least under our experimental conditions.
The formation of m-ZrO2 does not affect the polarisation
resistance values and its low conductivity is compensated with
the excellent electronic contribution from Cu. In any case the
formation m-ZrO2 could be avoided by firing at temperatures
below 900 �C under oxidising conditions.
Finally, XPS studies show evidence of metallic Cu and/or Cu(I)
inside of the YSZ electrolyte together with a partial reduction of
Zr(IV) to Zr(III). Although the electrolyte seems to be affected
just in the surface layer in contact with the Cu-based anode,
impedance measurements reveals a non-reversible increase of at
least �25% in the overall resistance compared to the as-prepared
YSZ, at 700 �C. In this case an alternative to overcome this
situation could be via fuel cell operation at very low tempera-
tures, perhaps below 600 �C.
The impregnation methods could be an ideal solution to
overcome both problems, though always working and/or
fabricating at low temperatures (<850 �C and <700 �C under
oxidising and reducing conditions, respectively).
References
1 R. J. Gorte, S. Park, J. M. Vohs and C. H. Wang, Adv. Mater., 2000,12, 1465.
2 S. D. Park, J. M. Vohs and R. J. Gorte, Nature, 2000, 404, 265.3 E. Ramirez-Cabrera, A. Atkinson, N. Brandon and B. C. H. Steele,
‘‘Cu/Ce0.9Ge0.1O2�x Anodes for Intermediate Temperature SOFCs’’,5th European SOFC Forum, Lucerne, Switzerland, July 1–5, 2002.
4 J. C. Ruiz-Morales, P. Nunez, R. Buchanan and John T. S. Irvine,J. Electrochem. Soc., 2003, 150, A1030.
5 M. B. Jorger, PhD Thesis, Swiss Federal Institute of Technology,Zurich, Switzerland, 2004.
6 A. Ringuede, J. A. Labrincha and J. R. Frade, Solid State Ionics,2001, 141–142, 549.
7 C. M. Grgicak, R. G. Green and J. B. Giorgi, J. Mater. Chem., 2006,16, 885.
8 N. Kiratzis, P. Holtappels, C. E. Hatchwell, M. Mogensen andJ. T. S. Irvine, Fuel Cells, 2001, 1, 211.
9 M. D. Gross, J. M. Vohs and R. J. Gorte, Electrochim. Acta, 2007, 52,1951.
10 S. Jung, C. Lu, H. He, K. Ahn, R. J. Gorte and J. M. Vohs, J. PowerSources, 2006, 154, 42.
11 J. C. Ruiz-Morales, J. Canales-Vazquez, D. Marrero-Lopez,John T. S. Irvine and P. Nunez, Electrochim. Acta, 2007, 52, 7217.
12 S. Tao and J. T. S. Irvine, Nat. Mater., 2003, 2, 320.13 S. Tao and J. T. S. Irvine, J. Electrochem. Soc., 2004, 151, A252.14 J. Liu, B. D. Madsen, Z. Ji and S. A. Barnett, Electrochem. Solid-State
Lett., 2002, 5, A122.15 J. C. Ruiz-Morales, J. Canales-Vazquez, C. Savaniu, D. Marrero-
Lopez, W. Zhou and J. T. S. Irvine, Nature, 2006, 439, 568.16 J. Canales-Vazquez, J. C. Ruiz-Morales, J. T. S. Irvine and W. Zhou,
J. Electrochem. Soc., 2005, 152, A1458.17 A. Ovalle, J. C. Ruiz-Morales, J. Canales-Vazquez, D. Marrero-
Lopez and J. T. S. Irvine, Solid State Ionics, 2006, 117, 1997.18 J. C. Ruiz-Morales, J. Canales-Vazquez, C. Savaniu, D. Marrero-
Lopez, W. Zhou, P. Nunez and J. T. S. Irvine, Phys. Chem. Chem.Phys., 2007, 9, 1821.
19 Y. H. Huang, R. I. Dass, Z. L. Xing and J. B. Goodenough, Science,2006, 312, 254.
This journal is ª The Royal Society of Chemistry 2008
20 Y. H. Huang, R. I. Dass, J. C. Denyszyn and J. B. Goodenough,J. Electrochem. Soc., 2006, 153, A1266.
21 W.-P. Dow, Y.-P. Wang and T.-J. Huang, J. Catal., 1996, 160, 155.22 W.-P. Dow and T.-J. Huang, J. Catal., 1996, 160, 171.23 T.-J. Huang, K.-C. Lee, H.-W. Yang and W.-P. Dow, Appl. Catal., A,
1998, 174, 199.24 P. Ratnasamy, D. Srinivas, C. V. V. Satyanarayana, P. Manikandan,
R. S. Senthil Kumaran, M. Sachin and V. N. Shetti, J. Catal., 2004,221, 455.
25 A. Szizybalski, F. Girgsdies, A. Rabis, Y. Wang, M. Niederberger andT. Ressler, J. Catal., 2005, 233, 297.
26 P. H. Matter, D. J. Braden and U. S. Ozkan, J. Catal., 2004, 233, 340.27 P. H. Matter and U. S. Ozkan, J. Catal., 2005, 234, 463.28 A. Mastalir, B. Frank, A. Szizybalski, H. Soerijanto, A. Deshpande,
M. Niederberger, R. Schomacher, R. Schlogl and T. Ressler, J. Catal.,2005, 230, 464.
29 X.-M. Liu, G. Q. Lu and Z. -F. Yan, Appl. Catal., A, 2005, 279, 241.30 M. D. Rhodes and A. T. Bell, J. Catal., 2005, 233, 198.31 M. D. Rhodes, K. A. Pokrovsli and A. T. Bell, J. Catal., 2005, 233,
210.32 G.-S. Wu, L.-C. Wang, Y.-M. Liu, Y. Cao, W.-L. Dai, H.-Y. He and
K.-N. Fan, Appl. Surf. Sci., 2006, 253, 974.33 M. Itome and A. E. Nelson, Catal. Lett., 2006, 106, 21.34 M. Labaki, J.-F. Lamonier, S. Siffert, E. A. Zhilinskaya and
A. Aboukaıs, Kinet. Catal., 2004, 45, 227.35 M. Labaki, J.-F. Lamonier, S. Siffert, F. Wyrwalski and A. Aboukaıs,
Thermochim. Acta, 2006, 443.36 J. Rodrıguez-Carvajal, FullProf 2004, version 3.00, Laboratoire Leon
Brillouin, CEA-Saclay, France, 2005.37 T. Roisnel, J. Rodrıguez-Carvajal. WinPLOTR, Laboratoire Leon
Brillouin, France, 2005.38 X’Pert HighScore Plus, version 2.0a, PANalytical BV, 2004.39 J. C. Ruiz-Morales, J. Canales-Vazquez, J. Pena-Martınez,
D. Marrero-Lopez and P. Nunez, Electrochim. Acta, 2006, 52, 278.40 S. P. Simner, M. Anderson, J. Bonnet and J. Stevenson, Solid State
Ionics, 2004, 175, 79.41 S. P. Simner, M. D. Anderson and J. W. Stevenson, J. Am. Ceram.
Soc., 2004, 87, 1471.42 A. Orera, PhD Thesis, Universidad Carlos III, Spain, 2007.43 Z. Lu, L. Pei, T.-m. He, X.-q. Huang, Z.-g. Liu, Y. Ji, X.-h. Zhao and
W.-h. Su, J. Alloys Compd., 2002, 334, 299.44 J. Ding, N. Zhao, C. Shi, X. Du and J. Ji, J. Alloys Compd., 2006, 425,
390.45 L. Lemaire, S. M. Scholz, P. Bowen, J. Dutta, H. Homeister and
H. Hofmann, J. Mater. Sci., 1999, 34, 2207.46 S. Ran, A. J. A. Winnubst, H. Koster, P. J. de Veen and
D. H. A. Blank, J. Eur. Ceram. Soc., 2007, 27, 683.47 S. Ran, A. J. A. Winnubst, W. Wiratha and D. H. A. Blank, J. Eur.
Ceram. Soc., 2006, 26, 391.48 S. Ran, A. J. A. Winnubst, W. Wiratha and D. H. A. Blank, J. Am.
Ceram. Soc., 2006, 89, 151.49 M. K. Dongare, V. Ramaswamy, C. S. Gopinath, A. V. Ramaswamy,
S. Scheurell, M. Brueckner and E. Kemnitz, J. Catal., 2001, 199, 209.50 M. K. Dongare, A. M. Dongare, V. B. Tare and E. Kemnitz,
Solid State Ionics, 2002, 152–153, 455.51 M. Bhagwat, A. V. Ramaswamy, A. K. Tyagi and V. Ramaswamy,
Mater. Res. Bull., 2003, 38, 1713.52 V. R. Choudhary, B. S. Uphade, S. G. Pataskar and A. Keshavaraja,
Angew. Chem., Int. Ed. Engl., 1996, 35, 2393.53 A. L. Vasiliev, E. Olsson, J. Boikov, T. Claeson and N. A. Kiselev,
Physica C, 1995, 253, 297.54 N. S. McIntyre and M. G. Cook, Anal. Chem., 1975, 47, 2208.55 M. B. Pomfret, C. Stoltz, B. Varughese and R. A. Walker,
Anal. Chem., 2005, 77, 1791.56 W. Wang, H. T. Guo, J. P. Gao, X. H. Dong and Q. X. Qin, J. Mater.