Electrical and stability performance of anode-supported solid oxide fuel cells with strontium- and magnesium-doped lanthanum gallate thin electrolyte

Article abstract of DOI:10.1016/j.electacta.2008.01.039

Anode-supported solid oxide fuel cells (SOFCs) comprising NiO-samarium-doped ceria (SDC) (Sm_0.2Ce_0.8O_1.9) composite anode, thin tri-layer electrolyte, and La_0.6Sr_0.4Co_0.8Fe_0.2O₃ (LSCF)-La_0.9Sr_0.1Ga_0.8Mg_0.2O_3-δ (LSGM) composite cathode were fabricated. The thin tri-layer consisting of an 11-μm thick LSGM electrolyte layer and a 12-μm thick La_0.4Ce_0.6O_1.8 (LDC) layer on each side of the LSGM was prepared by centrifugal casting and co-firing technique. The performance of the cells operated with humidified H₂ as fuel and ambient air as oxidant showed a maximum power density of 1.23 W cm^-2 at 800 °C. A stability test of about 100 h was carried out and some deterioration of output power was observed, while the open circuit voltage (OCV) kept unchanged. Impedance measurements showed that both the electrolyte ohmic resistance and the electrode polarization increased with time and the latter dominated the degradation.

Full text of DOI:10.1016/j.electacta.2008.01.039

Available online at www.sciencedirect.com
Electrochimica Acta 53 (2008) 4420–4427
Electrical and stability performance of anode-supported solid oxide fuel cells
with strontium- and magnesium-doped lanthanum gallate thin electrolyte
∗
Weimin Guo, Jiang Liu , Yaohui Zhang
College of Chemistry, South China University of Technology, Guangzhou 510640, Guangdong, PR China
Received 14 September 2007; received in revised form 9 January 2008; accepted 13 January 2008
Available online 25 January 2008
Abstract
Anode-supported solid oxide fuel cells (SOFCs) comprising NiO–samarium-doped ceria (SDC) (Sm0.2Ce0.8
electrolyte, and La0.6Sr0.4Co0.8Fe0.2 3−δ (LSGM) composite cathode were fabricated. The thin tri-layer consisting
of an 11-␮m thick LSGM electrolyte layer and a 12-␮m thick La0.4Ce0.6 1.8 (LDC) layer on each side of the LSGM was prepared by centrifugal
casting and co-firing technique. The performance of the cells operated with humidified H as fuel and ambient air as oxidant showed a maximum
O1.9) composite anode, thin tri-layer
3
O (LSCF)–La0.9Sr0.1Ga0.8Mg0.2O
O
2
−2
◦
power density of 1.23 W cm at 800 C. A stability test of about 100 h was carried out and some deterioration of output power was observed, while
the open circuit voltage (OCV) kept unchanged. Impedance measurements showed that both the electrolyte ohmic resistance and the electrode
polarization increased with time and the latter dominated the degradation.
©
2008 Elsevier Ltd. All rights reserved.
Keywords: SOFCs; La0.9Sr0.1Ga0.8Mg0.2O3−δ; Centrifugal casting; Co-firing; Stability
1
. Introduction
doped ceria, perovskite-type strontium and magnesium-doped
lanthanum gallate (LSGM) is a new kind of electrolyte material
and has not been thoroughly investigated. While some stabil-
ity problems may occur in some cases for LSGM, it might
become an alternative electrolyte that can allow operating tem-
Solid oxide fuel cells (SOFCs) operating in the intermedi-
◦
ate temperature range (500–800 C) (IT-SOFCs) have received
great attention because a relative large material selection range
and a reasonable performance can be realized at the same time
◦
perature below 800 C, due to its high oxygen ionic conductivity
−
1
◦
[
1,2]. Doped zirconia, doped ceria, and doped LaGaO3 have
(0.15 S cm at 800 C), high oxygen transport number, good
chemical stability over a wide oxygen partial pressure range
(10
been selected as the electrolytes for SOFCs. Yttria-stabilized zir-
conia (YSZ) is the most commonly used electrolyte for SOFCs
due to its pure oxygen-ion conduction and its stability both in
oxidizing and reducing atmospheres. However, the ionic con-
ductivity of this conventional electrolyte below 1000 C is not
satisfying. Despite excellent performance of anode-supported
YSZ electrolyte SOFCs operated around 800 C, searching for
alternative electrolyte with higher conductivity has been attract-
ing much interest. Doped ceria such as gadolinium-doped ceria
−
22
to 1 atm), and good mechanical strength [4].
Some electrolyte-supported or thick film LSGM SOFCs have
◦
been fabricated and shown good performance at 800 C [5–8].
◦
However, the ohmic resistance mainly from the electrolyte in
these cells still dominates the cell performance in the reduced
temperature range. Recently, many efforts have been contributed
to fabricate SOFC single cells with anode-supported thin LSGM
electrolyte [9,10]. A deep understanding of the effect of the
extensive chemical reactions between LSGM and electrodes or
ionic migration during high-temperature co-firing on the perfor-
mance of the SOFCs has been conveyed [10,11]. La diffusion
at the anode/electrolyte interface and inter-diffusion of Ga and
Co across the cathode may result in a serious problem with a
thinner electrolyte. The Co diffusion into the LSGM creates a
conducting phase that penetrates from the cathode toward the
anode to reduce the real LSGM thickness. It has been suggested
◦
(
GDC) or samarium-doped ceria (SDC) has much higher ionic
conductivity than YSZ at lower temperatures but the major
drawback is its mixed electronic/ionic conduction properties
under the fuel reducing atmosphere [3]. Compared to YSZ and
∗
Corresponding author. Tel.: +86 20 2223 6168; fax: +86 20 2223 6168.
E-mail address: jiangliu@scut.edu.cn (J. Liu).
0
013-4686/$ – see front matter © 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2008.01.039

W. Guo et al. / Electrochimica Acta 53 (2008) 4420–4427
4421
that La0.4Ce0.6O1.8 (LDC) buffer layer must be introduced to
minimize reactions and inter-diffusion between LSGM and the
Ni-based anode or the Co-contained cathode [5]. Despite the
promising results obtained from the anode-supported single
SOFC with thin LSGM film, there is a lack of demonstration of
long-term stability under high power output, even though there
are some reports about stability test of SOFCs with thick LSGM
electrolyte [5,7].
In the present work, a centrifugal casting technique
was applied to fabricate thin LDC/LSGM/LDC tri-layer on
NiO–SDC (Sm0.2Ce0.8O1.9) anode substrates. SOFCs with
dense thin LDC/LSGM/LDC layers were prepared by co-firing,
with no evidence of deleterious reactions. SOFC testing exhib-
ited relative high power densities. A cell maintaining at a
−
2
◦
constant current of 1 A cm was tested for 95 h at 800 C. The
measured voltage at this current dropped from 0.699 V at the
beginning to about 0.607 V after the test and the maximum
−2
−2
.
power density decreased from 1.08 W cm to 0.81 W cm
The experimental results were carefully discussed.
2
. Experimental
2
.1. Powder preparation
Fig. 1. Schematic illustration of the single-cell testing set.
La0.9Sr0.1Ga0.8Mg0.2O3−δ (LSGM) powder was synthesized
by solid-statereaction. Stoichiometric amountof La2O3, SrCO3,
be easily and accurately controlled over a large range. LDC
and LSGM slurry suspensions were prepared by mixing 0.5 g of
each powder and an organic vehicle of ethylcellulose (6 wt%)
and terpineol (94 wt%) in a weight ratio of 1:1 with 100 mL
ethanol, and sonicating for 1 h. The NiO–SDC substrates were
placed in vessels with flat bottom, and a proper amount of slurry
was added into each vessel along with some more ethanol. The
vessels were then placed in a centrifuge with a radius of about
Ga2O3, and MgO were mixed in ethanol and ball-milled for
◦
2
0 h. The mixture was dried and pre-fired at 1000 C for 4 h,
◦
then pressed into pellets, finally sintered at 1250 C for 21 h.
La0.4Ce0.6O1.8 (LDC) powder was also synthesized by solid-
state reaction of La2O3 and CeO2 after firing at 1250 C for
◦
2
1 h. The Sm0.2Ce0.8O1.9 (SDC) powder was prepared using
◦
a citric-nitrate process and calcined at 800 C for 2 h [12].
Nickel oxide (NiO) powder was obtained by the precipitation
method, using Ni(NO3)2·6H2O as the raw material and ammo-
20 cm, and the centrifuge was operated for 30 min at 4000 rpm.
◦
In the centrifugal field, the ceramic particles in the slurry were
forced down to the surface of the substrate to form a coating.
After the centrifuge process, the supernatant was decanted and
the film prepared was dried in ambient air for half an hour. Then,
the next layer was deposited. In this way, layers of LDC, LSGM,
and LDC were centrifugal deposited sequentially on the anode
supports. The thickness of the LSGM and LDC layers was con-
trolled by the amount of slurry. The deposited pellets were dried
nia as the precipitant. The precipitate was calcined at 400 C for
2
h. La0.6Sr0.4Co0.8Fe0.2O3 (LSCF) powder was produced by a
combined citrate–EDTA complexing method and the primary
◦
powder was calcined at 950 C for 5 h [13].
All the powders were characterized with an X-ray diffrac-
tometer (XD-3A, Shimadzu, Japan) with Cu K␣ radiation. The
◦
◦
selected 2θ range was from 20 to 80 scanning at a step of
0
◦
.02 . The XRD patterns were analyzed and determined with
◦
in open air, and then sintered at 1400 C for 4 h. The porous
MDI Jade 5.1 and PCPDFWIN 1.10 software.
cathode was prepared using a mixture of LSCF and LSGM in
a weight ratio of 1:1 with an ethylcellulose–terpineol vehicle.
The mixture was applied on the electrolyte, followed by a pure
2
.2. Cell fabrication
◦
LSCF layer, and fired at 1100 C for 1 h in air.
The as-prepared NiO and SDC powders were mixed in a
weight ratio of 6:4. After the mixture with 15 wt% starch as
pore former was mixed and ground in an agate mortar for 2 h,
the anode powders were pressed into pellets of 13 mm in diam-
eter and 0.6 mm in thickness under a pressure of 300 MPa.
2.3. Cell testing
A single cell was assembled by attaching a cell pellet to one
end of an alumina tube using silver paste (DAD-87, Shang-
hai Research Institute of Synthetic Resin, Shanghai, China) as
sealant and joint material (Fig. 1). The cells were tested with an
electrochemical instrument (Autolab PGSTAT30, Eco Chemie,
Holland). Hydrogen passing over the anode was controlled at
◦
The green pellets were pre-sintered at 1000 C for 4 h to
obtain the NiO–SDC anode substrates. Tri-layer electrolytes
LDC/LSGM/LDC were sequentially deposited on NiO–SDC
anode substrates by centrifugal casting technique [14]. The tech-
nique is simple, fast, and economical. The coating thickness can

4
422
W. Guo et al. / Electrochimica Acta 53 (2008) 4420–4427
Fig. 3. Cross-sectional SEM image of a typical LDC/LSGM/LDC tri-layer elec-
trolyte with portions of the Ni–SDC anode support and LSGM–LSCF cathode.
Impedance spectra were measured typically in the frequency
range from 0.1 Hz to 100 kHz under open-circuit conditions. The
tested cell was fractured and examined by a scanning electron
microscope (Philips XL-30FEG, Holland).
3. Results and discussion
3.1. Characterization of powders
The XRD analysis shows that the XRD pattern of the as pre-
pared LSGM powder (Fig. 2a) matches well with the diffraction
data for a perovskite phase. The patterns for LDC and SDC
show the standard cubic fluorite structure and a full single phase
(Fig. 2b). The XRD pattern of LSCF powder (Fig. 2c) shows
good agreement with expected peaks for a perovskite phase,
with no indication of second phases.
3.2. Microstructure of cell
Fig. 3 shows a cross-sectional SEM image of the
LDC/LSGM/LDC tri-layer along with the surrounding anode
Fig. 2. XRD patterns of (a) LSGM powders prepared by solid-state reaction,
b) LDC powders prepared by solid-state reaction and SDC powders pre-
(
pared by citrate-nitrate process, and (c) LSCF powders prepared by combined
citrate–EDTA complexing method.
a flow rate of 75 mL min 1, while the cathode was exposed
−
to ambient air. After the NiO-contained anode was reduced at
Fig. 4. Voltage and power densities vs. current density of
LDC/LSGM/LDC tri-layer electrolyte SOFC operated on air and humidified
hydrogen.
a thin
◦
6
00 C in H2 for several hours, the performance of the cell
◦
◦
◦
was tested from 600 C to 800 C with an interval of 50 C.

W. Guo et al. / Electrochimica Acta 53 (2008) 4420–4427
4423
Fig. 5. EDX linear scan analysis results on the cross-section of the LDC/LSGM/LDC tri-layer electrolyte, Ni–SDC anode support and LSGM–LSCF cathode after
long-term testing (a) SEM image; (b) EDX shows the La, Ni, Co and Fe element distribution over a 94 ␮m range: (I) LSCF–LSGM cathode, (II) LDC buffer layer,
(
III) LSGM electrolyte, (IV) LDC buffer layer, and (V) Ni–SDC anode.

4
424
W. Guo et al. / Electrochimica Acta 53 (2008) 4420–4427
and cathode of a single cell. No peel-off and crack were found
at the interface between the LSGM and LDC layers. The LSGM
and LDC layers were relatively dense with some closed pores.
There was good intimate contact between the LSGM and LDC
layers and the surrounding electrodes, showing that the materi-
als sintered well together, although their sintering properties are
somewhat distinctly different. The LSGM layer and LDC layers
on either side are about 11 ␮m and 12 ␮m thick, respectively.
3
.3. Cell output performance
Fig. 4 shows the I–V curves and the corresponding power
densities forthe single cellwithLDC/LSGM/LDC tri-layerelec-
trolyte. The values of the OCV are quite similar to those reported
by Lin et al. with a 9 ␮m LSGM and 7–10 ␮m LDC tri-layer
electrolyte cell [15]. The OCVs are all lower than that of a thick
LSGM single cell (∼1.1 V) tested in similar conditions. Lin et
al. suggested that the lower OCV is not caused by gas leakage
across the electrolyte based on the facts of dense electrolyte, low
measured gas leakage rate, and stable OCV value with changing
gas flow rate. To identify whether metal cations had migrated
across the electrode/electrolyte interfaces, the cell was charac-
terized by EDX after power measurement. Fig. 5 shows the
results of EDX linear scans for elements distribution (La, Ni,
Co, and Fe) across the cross-sectional area. It was observed that
the LDC buffer layer between cathode and electrolyte success-
fully blocked the diffusion of Fe and Co cations from the cathode
into the LSGM film, probably ascribed the results to the rela-
tively low firing temperature for cathode. However, obvious Ni
peaks were seen in the interface of LSGM and LDC film closed
to anode side and also appeared in LSGM film, presumably due
to migration during the high temperature co-firing. Introduction
of Ni into LSGM will induce electronic p-type conduction in
the electrolyte resulting in lower OCV [16,17]. Another similar
work with thicker LSGM layer (50 ␮m) and LDC layers (25 ␮m)
reported by Bi et al. [10,11] showed higher OCV of 1.043 V at
Fig. 6. Maximum power densities vs. temperature for the present SOFC com-
pared with prior reports from Refs. [11,15,18,19].
trolyte layers should be made to get the best performance of such
cells.
The maximum power densities shown in Fig. 4 are 1.23, 1.0,
−
2
◦
0.65, 0.38 and 0.20 W cm at 800, 750, 700, 650, and 600 C,
respectively. Fig. 6 compares the maximum power densities for
the present LDC/LSGM/LDC tri-layer electrolyte SOFC with
some prior reported data for LSGM electrolyte cells fabricated
by co-firing [11,15,18,19]. The compositions of these cells are
listed in Table 1. Despite the difference of materials used for
anodes and cathodes, the power densities of these cells with and
without buffer layers show general dependence on the thickness
of the electrolyte, i.e., a cell with thinner electrolyte gives higher
power density.
3.4. Impedance spectroscopy
Fig. 7 shows the impedance spectra of a typical single cell
◦
measured under open-circuit at temperatures from 600 C to
◦
−2
◦
8
00 C, but the power density was lower, 0.86 W cm com-
800 C using a four-probe configuration. The intercept with the
−
2
◦
pared with 1.12 W cm at 750 C reported by Lin et al. and
1
real-axis at the highest frequency represents the ohmic resis-
tance of the electrolyte (ohmic resistance from other sources can
be neglected). The measured specific ohmic resistance across
− ◦
2
.23 W cm at 800 C of the present work, presumably due to
higher resistance caused by the thicker electrolyte layers which
also producing higher OCV through blocking metal migration
more effectively. Some optimization on the thickness of the elec-
2
the LDC/LSGM/LDC tri-layer electrolyte film is 0.086 ꢀ cm
◦
2
at 800 C, lower than the calculated value 0.13 ꢀ cm based on
Table 1
Compositions of LSGM electrolyte SOFCs fabricated by co-firing
Cell sources
Anode
Anode buffer layer
Electrolyte
Canthode
Cathode
buffer layer
Present work
Ref. [11]
Ni–Sm0.2Ce0.8O1.9 (60:40 wt%)
Ni–Gd0.1Ce0.9O1.95 (60:40 wt%)
Ni–La0.4Ce0.6O1.8 (LDC) (60:40 wt%)
La0.4Ce0.6O1.8 (LDC)
LSGM (11 ␮m)
LSGM (50 ␮m)
LSGM (9 ␮m)
LDC(12 ␮m)
LDC (25 ␮m)
LDC (8 ␮m)
La0.6Sr0.4Co0.8Fe0.2O3–LSGM
(50:50 wt%)
La0.9Sr0.1CoO3–LDC
(60:40 wt%)
(12 ␮m)
La0.45Ce0.55O2−δ/2
(LDC) (25 ␮m)
Ref. [15]
LDC (8 ␮m)
La0.6Sr0.4Co0.8Fe0.2O3–LSGM
(
50:50 wt%)
Ref. [18]
Ref. [19]
Ni–Sm0.2Ce0.8O1.9 (52:48 mol%)
Ni–La0.4Ce0.6O1.8 (LDC) (50:50 wt%)
LSGM (130 ␮m)
LSGM (1 mm)
La0.6Sr0.4CoO3
La0.6Sr0.4Co0.8Fe0.2O3–LSGM
LDC (15 ␮m)
(
50:50 wt%)

W. Guo et al. / Electrochimica Acta 53 (2008) 4420–4427
4425
Fig. 9. Voltage and power density vs. time for a single Ni–SDC/LDC/LSGM/
−
2
LDC/LSCF–LSGM cell operated at a constant current of 1 A cm
.
Fig. 7. Electrochemical impedance spectra of a single SOFC measured under
open-circuit conditions in air and humidified hydrogen.
thickness of the layers composing the tri-layer electrolyte, there
is potential to get a better performance at reduced temperatures
by improving the electrodes.
prior reported bulk LSGM conductivity of 0.12 S cm 1and LDC
−
−
1
◦
of 0.02 S cm at 800 C [20,21]. This is most likely because
the bulk conductivity of LDC in a single cell under operating
condition is higher than that measured directly on the elec-
trolyte under air atmosphere [11]. The relative lower resistance
of LDC/LSGM/LDC tri-layer electrolyte film and the high out-
put performance of the single cell indicate that the reactions
inducing high resistance products between LSGM and NiO or
cathode were effectively prevented by the LDC buffer layers. In
the impedance spectra, the lowest frequency real axis intercept
provides a measure of the total cell resistance, the width of the
impedance arcs on the real-axis is the polarization resistance,
which includes reaction and transport resistances from both the
anode and cathode.
3.5. Cell stability
A stability test of the cell was carried out under a constant
−2
−1
current density of 1000 mA cm with a 30 mL min H2 flow
◦
rate at 800 C. The cell voltage at this current was recorded as
a function of time. The cell was tested for about 100 h and the
results are shown in Fig. 9. Note the cell used for this test had
been tested 10 days before to get performance data and then it
was cooled down and placed in open air. At the beginning of
the stability test, the cell performance was slightly lower than
that of the first time test. Then it gradually degraded with time at
−1
a rate of about 1 mV h . A comparison of output performance
measuredbeforeandafterthestabilitytestispresentedin Fig. 10.
The polarization resistance, ohmic resistance and total resis-
tance as determined from the impedance spectra are shown in
−2
The maximum power density drops from 1.08 W cm before
−2
the test to 0.81 W cm after 95 h. It is obvious that the OCV
◦
Fig. 8. At temperatures below 700 C, the polarization resistance
does not change with time. This fact suggests that there is little
transition metal diffusing from the electrodes to the electrolyte
during cell operation. In other words, the deterioration of the cell
is not caused by higher electronic conduction of the electrolyte
is higher than the ohmic resistance. While there may be some
opportunities to improve the cell performance by optimizing the
Fig. 8. Tri-layer electrolytes resistance, electrode polarization resistance, and
total cell resistance as determined by impedance spectra.
Fig. 10. Comparison of cell performance measured before and after a stability
test of 95 h.

4
426
W. Guo et al. / Electrochimica Acta 53 (2008) 4420–4427
deterioration of the cell performance over time is mainly caused
by degradation of electrodes or electrode–electrolyte interfaces.
The influence of nickel metal coarsening in the porous Ni–SDC
anode has been indicated by some prior researchers [22], and is
assured further by the SEM images in Fig. 12, showing particles
in the anode becomes coarser with operating time. Some further
studies are in progress and the details will be reported in the
future.
4. Conclusions
Anode-supported SOFCs with dense thin LDC/LSGM/LDC
tri-layer electrolyte, Ni–SDC cermet anode, and LSCF–LSGM
composite cathode were successfully fabricated using centrifu-
gal casting and co-sintering technique. The maximum power
Fig. 11. Comparison of the electrochemical impedance spectra measured before
and after stability testing.
−2
densities were 1.23, 1.0, 0.65, 0.38 and 0.20 W cm at 800,
◦
induced by transition metal diffusion to the LSGM layer. The
electrochemical impedance spectra in Fig. 11 show that both
the electrode polarization and ohmic electrolyte resistance after
750, 700, 650 and 600 C, respectively. The good performance
and low ohmic resistance showed that there was no significant
formation of deleterious interfacial phases. A 12-␮m thick of
LDC buffer layer cannot completely avoid the diffusion of Ni
from anode to LSGM electrolyte during high temperature co-
firing resulting in a lower OCV. A 95-h stability of the cells
9
5 h stability test are higher than that before the test. While the
2
increase of ohmic resistance is about 0.03 ꢀ cm , the increase
of the electrode polarization is 0.1 ꢀ cm , three times of the
2
◦
was monitored at 800 C under a constant current density of
electrolyte.
−
2
Considering the unchanging OCV with time and the signif-
icant increase of electrode polarization, we presume that the
1000 mA cm . While the OCV of the cell did not change
with time, some deterioration of the cell output was observed.
Based on the difference of the impedance spectra and the SEM
images of the anode obtained before and after the stability
test, we presume that the deterioration of the cell performance
over time is mainly caused by degradation of the electrodes or
electrode–electrolyte interfaces. Further investigation is needed
to determine the detail mechanisms of the degradation.
Acknowledgements
The authors gratefully acknowledge the financial support
from 863 Program of National High Technology Research
Development Project of China (Grant No. 2007AA05Z136), the
Department of Science and Technology of Guangdong Province
(2005B50101007) and the Department of Education of Guang-
dong Province (B15N9060210).
References
[
[
[
1] N.Q. Minh, J. Am. Ceram. Soc. 76 (1993) 563.
2] S.C. Singhal, Solid State Ionics 135 (2000) 305.
3] M. Mogensen, N.M. Sammes, G.A. Tompsett, Solid State Ionics 129 (2000)
63.
[
[
4] T. Ishihara, H. Matsuda, Y. Takita, J. Am. Chem. Soc. 116 (1994) 3801.
5] K.Q. Huang, J.-H. Wan, J.B. Goodenough, J. Electrochem. Soc. 148 (2001)
A788.
[
[
[
6] Z.H. Bi, B.L. Yi, Z.W. Wang, Y.L. Dong, H.J. Wu, Y.C. She, M.J. Cheng,
Electrochem. Solid-State Lett. 7 (2004) A105.
7] J.-H. Wan, J.-Q. Yan, J.B. Goodenough, J. Electrochem. Soc. 152 (2005)
A1511.
8] Z.H. Bi, M.J. Cheng, Y.L. Dong, H.J. Wu, Y.C. She, B.L. Yi, Solid State
Ionics 176 (2005) 655.
Fig. 12. SEM images of a NiO–SDC anode with humidified H2 as fuel operated
for (a) several hours and (b) 95 h.
[9] J.W. Yan, H. Matsumoto, M. Enoki, T. Ishihara, Electrochem. Solid-State
Lett. 8 (2005) A389.

W. Guo et al. / Electrochimica Acta 53 (2008) 4420–4427
4427
[
10] J.W. Yan, Z.G. Lu, Y. Jiang, Y.L. Dong, C.Y. Yu, W.Z. Li, J. Electrochem.
Soc. 149 (2002) A1132.
[17] T. Ishihara, T. Shibayama, H. Nishiguchi, Y. Takita, J. Mater. Sci. 36 (2001)
1125.
[
[
11] Z.H. Bi, Y.L. Dong, M.J. Cheng, B.L. Yi, J. Power Sources 161 (2006) 34.
12] D.H.A. Blank, H. Kruidhof, J. Flokstra, J. Phys. D: Appl. Phys. 21 (1988)
[18] T. Fukui, S. Ohara, K. Murata, H. Yoshida, K. Miura, T. Inagaki, J. Power
Sources 106 (2002) 142.
2
26.
[19] W.Q. Gong, S. Gopalan, U.B. Pal, J. Power Sources 160 (2006)
305.
[20] J. Drennan, V. Zelizko, D. Hay, F.T. Ciacchi, S. Rajendran, S.P.S. Badwal,
J. Mater. Chem. 7 (1997) 79.
[21] D. Singman, J. Electrochem. Soc. 113 (1966) 502.
[22] D. Simwonis, F. Tietz, D. Stover, Solid State Ionics 132 (2000) 241.
[
13] Z.P. Shao, W.S. Yang, Y. Cong, H. Dong, J.H. Tong, G.X. Xiong, J. Membr.
Sci. 172 (2000) 177.
14] J. Liu, S.A. Barnett, J. Am. Ceram. Soc. 85 (2002) 3096.
15] Y.B. Lin, S.A. Barnett, Electrochem. Solid-State Lett. 9 (2006) A285.
16] T. Ishihara, T. Shibayama, M. Honda, H. Nishiguchi, Y. Takita, Chem.
Commun. (1999) 1227.
[
[
[