Electrochimica Acta 45 (2000) 2423 – 2435

www.elsevier.nl/locate/electacta

Solid oxide fuel cells: fundamental aspects and prospects

Osamu Yamamoto
Aichi Institute of Technology, Yatigusa, Toyota, Aichi 470 -0392, Japan

Papers received in Newcastle, 20 December 1999

Abstract

Solid oxide fuel cells (SOFCs) are advanced electrochemical reactors operating at a high temperature. SOFCs are
presently under development for a variety of electric power generation applications with high energy conversion
efficiency. This paper reviews the science and technology of SOFCs with emphasis on discussion of their component
materials. © 2000 Elsevier Science Ltd. All rights reserved.

1. Introduction
The origin of fuel cells goes back to the middle of the
last century, their principle was first described by Sir
William Grove in 1839 [1]. His fuel cell, operating at
room temperature, used a dilute sulfuric acid elec-
trolyte, a hydrogen anode and an oxygen cathode. At
that time no conventional batteries such as lead acid
batteries or Leclanche type cells were yet available and
so four fuel cells were connected in series and were
applied to the electrolysis of water. Thus, fuel cells
become the first electric high power supply.
Fuel cells are energy conversion devices that produce
electricity directly from a gaseous fuel by electrochemi-
cal combination of the fuel with an oxidant. Fuel cells
have many advantages compared to conventional elec-
tric power generation systems, such as high conversion
efficiency, which is relatively independent of size, as
well as environmental compatibility. The first successful
practical applications of fuel cells was for space appli-
cations in the 1960s. After further R&D efforts in the
1980s and 1990s, fuel cells have recently attracted new
attention due to their potential for use in stationary and
distributed electric power stations as well as in trans-
portation applications. At this stage, four types of fuel
cells, namely, the polymer electrolyte fuel cell (PEFC),
E-mail address: yamamoto@chem.mie-u.ac.jp (O. Ya-
mamoto)
the phosphoric acid fuel cell (PAFC), the molten car-
bonate fuel cell (MCFC), and the solid oxide fuel cell
(SOFC) have been mainly developed. These fuel cells
are characterized by the electrolyte used, which in turn
determines the operating temperature. The characteris-
tics and the current status of different types of fuel cells
are summarized in Table 1. The most advanced fuel cell
system is the PAFC. More than 200 PAFC units have
been introduced as on-site electric power plants. How-
ever, a serious problem for commercialization of PAFC
is their high unit cost. PEFCs have recently attracted
significant attention as a potential source of primary
power in transportation applications. As shown in
Table 1, SOFCs have exceptional potential for use as
electric power generation systems, because of their high
energy conversion efficiency which can reach up to
65%. In addition, SOFCs have many advantages such
as multi-fuel capability and simplicity of system design
[2]. Furthermore, the exhaust heat can be used as a heat
source for a variety of processes. This together with the
possibility of utilizing the remaining energy in a bot-
toming cycle for electric power generation leads to a
further increase in the overall efficiency. The pressur-
ized SOFC can be used to replace combustors in gas
and steam turbines. Such integrated SOFC-combustion
turbine power systems are expected to reach efficiencies
approaching 70% [3]. The high efficiency will contribute
to reduce CO2 emissions. In this review, science and
technology of SOFCs as well as their future prospects
will be discussed.

0013-4686/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 1 3 - 4 6 8 6 ( 0 0 ) 0 0 3 3 0 - 3
2424 O. Yamamoto / Electrochimica Acta 45 (2000) 2423–2435

Table 1
The characteristics and current status of the different types of fuel cell

PEFC PAFC MCFC SOFC

Electrolyte Nafion H3PO4 Na2CO2Li2CO3 ZrO2Y2O3

Operating temperature (°C) 70–80 200 650–700 900–1000
Fuel H2 H2 H2, CO, CH4 H2, CO, CH4
Expected efficiency (HHV) (%) 30–40 35–42 45–60 45–65
Power, current status (kW) 12.5a 100b 1000c 100d, 15e
Efficiency (%) 40 40 45 43d, 50e

a
Allied signal.
b
Fuji electric.
c
MCFC Research Association, Japan.
d
Siemens Westinghouse.
e
Mitsubishi Heavy Industry and Electric Power Development, Japan (at 5 atm).

2. Principle and histrical background of SOFC
The operating principle of a SOFC with an oxide ion
conductor is schematically shown in Fig. 1. When an
external load is applied to the cell, oxygen is reduced at
the porous air electrode to produce oxide ions. These
ions migrate through the solid electrolyte to the fuel
electrode, and they react with the fuel, H2 or CO, to
produce H2O or CO2. Alternatively, a proton conduct-
ing solid electrolyte can be used, where H2 is oxidized
to produce protons that subsequently react with oxygen
to form water. In some cases, CH4 can be oxidized
directly on the anode to form CO2 and H2O. The
open-circuit voltage, Eo, of the cell can be calculated
from the free energy change, DG, of the electrochemical
reaction or from the partial pressure of the oxygen
Po(c) at the cathode and Po(a) at the anode:
Eo = −DG/nF= (RT/nF) ln Po(c)/Po(a)
In this equation, R is the gas constant, T the absolute
temperature, F the Faraday constant, and n the electron
equivalent of oxygen (n=4). The cell voltage at 1000°C
is about 1 V for the SOFC with pure hydrogen and air.
The cell voltage drops as current is drained due to
polarization. The total polarization of a cell, h, is the
sum of three terms: anode polarization, ha, cathode
polarization, hc, and resistance polarization, hr.
is still used as the electrolyte. The oxide ion conductiv-
ity of YSZ is around 0.1 S/cm at 1000°C. Together with
the electrolyte, the other SOFCs components have been
studied for many years, but due to the high operating
temperature, the choice of materials remains limited. In
Baur and Preis’ first SOFC, iron or carbon were used as
the anode while Weissbart and Ruka [6] of Westing-
house Electric (now Siemens Westinghouse Power,
SWH) in 1962 used platinum. At present nickel is the
most commonly used material because of its high cata-
lytic activity and low cost. The nickel-YSZ cermet
anode can operate at a temperature as low as 500°C.
The cathode materials must be stable in the oxidizing
environment, have good electronic conductivity and
sufficient catalytic activity for oxygen reduction under
operating conditions. Many types of cathode materials
have been examined. Baur and Preis used Fe3O4 as the
(1) cathode. The electrical conductivity of Fe3O4 is as high
as 3 × 102 S/cm at room temperature, but its catalytic
activity is insufficient. Weissbart and Ruka [6] used
platinum as the cathode. At 1011°C, the cell voltage,

h=ha +hc +hr (2)

The polarization depends on the electrode materials,

the electrolyte, the cell design, and the operating
temperature.
The first SOFC operating at 1000°C was built by
Baur and Preis [4] in 1937. In this cell, ZrO2-10
weight%(w/o) MgO or ZrO2-15 w/o Y2O3 was used as
the electrolyte. Nernst first discovered that ZrO2 with
15 w/o Y2O3 was a solid oxide ion conductor in 1899
[5]. In the most advanced SOFCs at present, yttria
stabilized zironica with 8–10 mol% (m/o) Y2O3 (YSZ) Fig. 1. Principle of operation of SOFC (by S.C.Singhal).
 O. Yamamoto / Electrochimica Acta 45 (2000) 2423–2435
Table 2
Cell components, materials, and fabrication processes for
SWH SOFC
Component Material Thickness Fabrication
process
Air electrode Doped LaMnO 2.2 mm Extrusion
tube
Electrolyte ZrO2(Y2O3) 40 mm EVD
Fuel electrode NiZrO2(Y2O3) 100 mm Slury coat-
EVD
Interconnec- Doped LaCrO3 85 mm EVD
tion
Pt, H2/(ZrO2)0.85(CaO)0.15/Pt, O2, was 0.7 V at 50
mA/cm2. The decrease in cell voltage from the open-
circuit voltage of 1.1 V was mainly due to the inter-
nal resistance of the electrolyte, (ZrO2)0.85(CaO)0.15.
The polarization due to the electrode reactions was
not significant due to the high catalytic activity of
platinum for oxygen reduction. There is no hope for
the use of platinum cathodes in a large scale SOFC,
because the evaporation of platinum at high tempera-
ture is non negligible and its cost is too high. Some
complex oxides in the perovskite family may satisfy
the requirements with regard to electrical conductiv-
ity. Tedmon et al. [7] of General Electric in 1969
reported on the cathode performance of perovskite,
PrCoO3. A stable power density of 0.3 W/cm2 over 6
months was obtained at 1000°C. The disadvantages
of these cathode materials were thermal expansion
mismatch and incompatibility with the electrolyte. At
present, the doped lanthanum manganite cathodes are
most commonly used. However, there is still a need
for the development of new cathode materials with
good performance at lower temperatures in order to
develop an intermediate temperature SOFC (500–
700°C).
The stringent requirements for the interconnect ma-
terial in SOFCs operated at 1000°C eliminate all but
a few oxide systems. The interconnect in SOFCs must
be stable under reducing and oxidizing atmospheres,
and must possess adequate electronic conductivity in
both atmospheres. The doped lanthanum chromite is
particularly suitable with respect to high electronic
conductivity under fuel and oxidant atmospheres.
Lanthanum chromite (LaCrO3) has been used as an
SOFC interconnect material since the 1970s. Recently,
metallic interconnects of Cr alloys with 5 w/o Fe and
0.4–1.0 w/o Y2O3 or Cr2O3 have been evaluated as
an alternative material to LaCrO3 [8]. These alloys
have a thermal expansion coefficient very similar to
that of YSZ but the metallic interconnect is less re-
fractory and unstable in the oxidizing atmosphere at
high temperatures. Some commercially available high
2425
temperature alloys coated with LaCrO3 can be used.
The problems of the alloys is thermal expansion mis-
match with other cell components. Table 2 summa-
rizes the cell components, materials, and fabrication
processes for the most advanced SWH SOFC [9].
SOFCs with several different configurations have
been developed. The first SOFC module was demon-
strated by Brown Boveri in 1975 [10]. The module
was constructed from conical tubular cells by fitting
one cell into the other. Contact of the anode and
cathode of neighboring cells yielded a construction in
series. An interconnection material (doped CoCr2O4
or LaCrO3) was used to ensure sealing and electronic
contacting in air and in fuel atmosphere. In the mod-
ule with 120 series-connected cells, a total voltage of
110 – 120 V and a maximum power output of 115 W
corresponding to 0.22 W/cm2 was obtained at 1000°C
with H2/air as the reaction gas. This modular concept
suffered from the disadvantage that for the construc-
tion of large batteries, many single cells had to be
produced and connected in series by means of a rela-
tively expensive manufacturing process. A further sim-
plification and a reduction of cell manufacturing cost
was made possible by using thin-film techniques such
as chemical vapor deposition or plasma spraying.
A porous electrode layer, a gas tight ZrO2 elec-
trolyte film and a second porous electrode as well as
a gas tight interconnection material were applied one
after the other onto a highly porous ceramic support
[11] (segmented-cell-in-series design). In 1970, West-
inghouse had proposed a 100-kW coal burning fuel
cell power plant using this SOFC concept. As the
performance and life-time of cells connected in series
did not seem to be satisfactory at that time, the pro-
ject was abandoned [12]. Subsequently, Westinghouse
proposed a seal-less tubular design in 1980. The ma-
terials and the fabrication process of this type of cell
are shown in Table 2. Recently a 100-kW scale SOFC
has been tested, 30 years after it was first proposed
by Westinghouse. The segmented-cell-in-series design
proposed by Westinghouse in 1969 was modified by
Electric Technical Laboratory, Japan and a 10-kW
scale unit was tested by Electric Power Development,
Japan. In Fig. 2, typical planar and tubular configu-
rations of the SOFC are shown. Many modified pla-
nar concepts were proposed over the last two
decades. A monolithic SOFC design was proposed by
Argonne National Laboratory in 1984 [13] (shown in
Fig. 2). This design has the potential to achieve high
power density, because of the small cell size and large
active area. The power density for the monolithic
SOFC is calculated to be about 8 kW/kg or 4 kW/l.
The feasibility of the monolithic design has been
demonstrated only on a laboratory scale, however,
because of the difficulty of development of suitable
2426 O. Yamamoto / Electrochimica Acta 45 (2000) 2423–2435

materials and fabrication processes. A modified design
of the monolithic configuration has been developed by
Chubu Electric Power Co and Mitsubushi Heavy In-
dustry, and was named the Mono Block Layer Built
(MOLB). A 5-kW scale MOLB type SOFC was built
and successful results were obtained [14].
Up to the present, large scale SOFC units have been
operated at around 1000°C because no acceptable ma-
terials for low temperature SOFCs were available. A
lower operating temperature in the range of 500 – 800°C
would have many advantages. Costly lanthanum
chromite interconnects could be replaced by a conven-
tional metallic interconnect, and the selection of gas
seal materials in planar SOFCs would be easier. SOFCs
operating at 500°C or lower are being considered for
applications in electric vehicles [15].

Fig. 2. SOFC stack designs (by N.Q.Minh).

 O. Yamamoto / Electrochimica Acta 45 (2000) 2423–2435
Fig. 3. Temperature dependence of the electrical conductivity
of the ZrO2M2O3 systems. 9SSZ: 9m/o Sc2O3ZrO2, 9YbSZ:
9m/o Yb2O3ZrO2, 9YSZ:9m/o Y2O3ZrO2.
3. High temperatue SOFC with zirconia based elec-
trolytes
The most advanced SOFC employ oxide ion con-
ducting zirconia based electrolytes. The conductivity of
the electrolyte determines the operating temperature of
SOFCs. The temperature dependence of the electrical
conductivity, s, for zirconia based oxides is shown in
Fig. 3[16] where 9YSZ, 9YbZ, and 9SSZ denote 9 m/o
Y2O3, 9 m/o Yb2O3, and 9 m/o Sc2O3 doped ZrO2,
respectively. Yttria doped zirconia, which has been used
extensively in SOFCs, shows a conductivity of 0.14
S/cm at 1000°C. Scandia doped zirconia has a higher
conductivity and at 780°C already its value corresponds
to that of YSZ at 1000°C.
It is well known that doping of ZrO2 with Y2O3
results in the substitution of Y3 + on the Zr4 + cation
sublattice with concomitant formation of oxygen va-
cancies as charge compensating defects. The defect
Fig. 4. Composition dependence of the electrical conductivity
at 1000°C for ZrO2Ln2O3(Ln= lanthanide).
2427
formation reaction in Y2O3 doped ZrO2 can be written
in Kroger and Vink notation as
’’
Y2O3 = 2Y%Zr + Vo + 3Oxo (3)
Due to Coulombic and elastic attractive forces between
’’
Y%Zr and Vo , we can postulate the existence of two
associates:
’’ ’
Y%Zr + Vo = (YZrVo) (4)
and
’’
2Y%Zr + Vo = (2YZrVo)x (5)
At lower temperatures association is almost complete.
’’
At higher temperature the complex dissociates into Vo
and Y%Zr. The conductivity, s, can be explained by Eq.
(6)
’’
s= em[V o ] (6)
’’
where m is the mobility, [V ] the concentration of
o
oxygen vacancies and e the effective charge. The con-
ductivity in the ZrO2M2O3 system depends on the
dopant concentration and the dopant ionic radius. In
Fig. 4, the dependence of the conductivity at 1000°C on
the dopant concentration is shown for
ZrO2Ln2O3(Ln =lanthanides) [16]. A conductivity
maximum is found in these systems. The dopant Sc3 +
which has the closest ionic radius to the host ion, Zr4 + ,
shows the highest conductivity. The activation energy
for conduction in the temperature range of 1000 – 700°C
was found to decrease with decreasing ionic radius of
the dopant and for ZrO2Sc2O3 its value was about 60
kJ/mol. This behavior could be explained by the inter-
’’
action between Vo and LnZr % and by structural effects
[17].
In Fig. 3, the thickness of the electrolyte showing a
resistance of 0.2 V cm2 is shown on the right side
vertical axis. The 0.2 V cm2 resistance corresponds to
0.1 V lost due to the electrolyte resistance at 0.5 A/cm2,
where it is assumed that the cell voltage is 0.7 V to
maintain a total energy efficiency of greater than 50%
and ha, hc and hr contribute to equal parts to the cell
voltage drop [15]. The relationship between the temper-
ature and the electrolyte thickness suggests that at an
operating temperature of 700°C, the thickness of the
electrolyte should be less than 50 mm in the YSZ and
YbSZ systems while an electrolyte as thick as 150 mm
could be used in the SSZ system. A 50 mm thick
electrolyte can not be used in a self-supported configu-
ration, because of the difficulty of handling such a thin
electrolyte sheet. On the other hand, an electrolyte
thickness of 150 mm is acceptable for a self-supported
planar SOFC. In 1977 [18] SWH developed a vapor
deposition method for the fabrication of supported thin
electrolytes and gas-tight thin layers of YSZ with 10
m/o Y2O3 ( :40 mm thick) were fabricated on a pro-
duction scale. A large active area cell ( :200 cm2)
2428 O. Yamamoto / Electrochimica Acta 45 (2000) 2423–2435
Fig. 5. Voltage-current characteristics of a SWH cell at differ-
ent temperatures.
achieved a cell voltage of 0.7 V for a current drain of
350 mA/cm2 at 1000°C, and for 270 mA/cm2 at 900°C
as shown in Fig. 5. A doped LaMnO3 cathode and a
Ni/YSZ cermet anode were used in this case [9]. The
total cell resistance is about 0.8 V cm2 from the cur-
rent voltage curve at 900°C. The electrolyte resistance
estimated from the conductivity and the thickness is
only 0.05 V cm2. This means that at high temperature
the electrode polarization and the contact resistance
are dominant. No cell performance data at lower tem-
peratures have been reported for SWH type cells.
Minh and Montgomery of Allied Signal Aerospace
[19] reported the performance of cells with a planar
supported thin (5–10 mm) YSZ electrolyte. The elec-
trolyte was made by a tape casting technique. An area
specific resistance was 0.38 V cm2 at 800°C. The five
cell stack (100 cm [2] footprint area) showed a cell
voltage of 3.5 V (0.7 V per cell) at 350 mA/cm2 and
700°C. At lower operating temperatures, YSZ should
be used in thin film form as a supported electrolyte.
On the other hand, a solid with higher oxide ion
conductivity such as SSZ is suitable for thicker films.
Some results [20] for a single test cell with doped
Fig. 6. Voltage-current characteristics of a cell with 11SSZ
electrolyte at different temperatures (effective electrode area:
20 cm2).
Fig. 7. Temperature dependence of electrical conductivity for
oxide ion conductors (high temperature range).
lanthanum chromite interconnects are shown in Fig. 6,
where about 200 mm thick 11SSZ (11 m/o Sc2O3) of
20 cm2 active area was used as the electrolyte. At
850°C, a cell voltage of 0.7 V was obtained at 500
mA/cm2. The discharge performance at 850°C of the
cells with an 11SSZ electrolyte is comparable to that
with an YSZ electrolyte at 1000°C.
4. Materials and structure for intermediate temperature
SOFC
The cell performance depends on the ohmic resis-
tance of the electrolyte and on the electrode reaction
resistance (polarization). The electrode resistance is in-
dependent of the thickness of the electrolyte. During
the last three decades, the ionic conductivity and the
transport number have been examined in many binary
and ternary oxide systems. An excellent review on the
properties of solid oxide ion conductors has been pre-
sented by Etsell and Flengas [21] in 1970. Minh [2]
reviewed the more recent results for SOFC elec-
trolytes. The temperature dependence of conductivity
in typical solid oxide ion conductors is summarized in
Fig. 7 and Fig. 8 for the high and lower temperature
ranges, respectively [22]. Some oxide systems show a
higher ionic conductivity than zirconia based oxides.
The electrolyte in SOFCs has to meet the following
requirements in addition to a high ionic conductivity:
low electronic conductivity, stability under oxidizing
and reducing atmospheres, easy to make dense thin
films, and good mechanical and thermal properties at
 O. Yamamoto / Electrochimica Acta 45 (2000) 2423–2435
the operating temperature. Bismuth oxide systems show
Fig. 8. Temperature dependence of electrical conductivity for
oxide ion conductors (lower temperature range).
Fig. 9. Total electrical conductivity of Ce0.8Sm0.2O1.9 − x at
600, 700, and 800°C as a function of oxygen partial pressure.
Fig. 10. Performance of SOFC with anode-supported
Ce0.8Gd0.2 electrolyte and La0.8Sr0.2Fe0.8Co0.2O3 − cathode at
–
500°C.
2429
the highest conductivity. These systems, however, ex-
hibit high electronic conductivity under reducing atmo-
sphere, because of the reduction of Bi3 + to Bi2 + [23].
The most promising traditional materials for low tem-
perature operation of SOFCs are CeO2 doped with an
alkaline earth or rare earth oxide. The conductivity of
Ce0.8Gd0.2O1.9 [24] is 1.2×10 − 2 S/cm at 800°C and
3.7× 10 − 3 S/cm at 600°C. At higher oxygen partial
pressures the material is reported to have a purely ionic
conductivity. At lower oxygen partial pressures, as pre-
vail at the anode side of the SOFC, CeO2 becomes
partially reduced [25]. This leads to predominantly elec-
tronic conductivity in a large volume fraction of the
electrolyte extending from the anode side. In Fig. 9,
total electrical conductivities of Ce0.8Sm0.2O1.9 are
shown as a function of oxygen partial pressure [26]. The
increasing conductivity at lower oxygen partial pressure
is due to the contribution of the electronic conductivity.
At high oxygen partial pressure, the total electrical
conductivity is predominantly ionic, whereas at low
oxygen partial pressure it is predominantly electronic.
The efficiency of SOFCs with the mixed ionic-electronic
conductors of the CeO2 based oxide depends on the
operating temperature and the thickness of the elec-
trolyte due to the partial reduction of the electrolyte.
Godlickemeier and Gauckler [26] analyzed the effi-
ciency of cells with ceria-based electrolytes by consider-
ing electrode overpotentials. The maximum efficiency of
these cells was 50% at 800°C, 57% at 700°C, and 60% at
600°C. In practical operation, it is also necessary to
consider the fuel efficiency. Generally, the fuel effi-
ciency is less than 80%. Therefore, SOFCs with the
CeO2 based electrolytes should be operated at tempera-
tures lower than 600°C in order to obtain a total
efficiency exceeding 50%. A SOFC based on
Ce0.8Gd0.2O1.9 at 500°C was examined by Doshi et al.
[17] Using a multilayer tape-casting technique these
authors fabricated an anode supported electrolyte bi-
layer with an electrolyte thickness of 30 mm and anode
thickness of : 250 mm. Fig. 10 shows the performance
at 500°C of a cell with the NiCe0.8Gd0.2O1.9 cermet
anode and a cathode developed by Argonne National
Laboratory. The total cell resistance, including elec-
trode polarization for both electrodes, was 1.3 V cm2,
0.7 V cm2 resulting from the electrolyte, 0.6 V cm2 from
the cathode, while the loss from the anode was negligi-
ble. A cell voltage of 0.7 V was obtained at 150
mA/cm2 and the maximum power density was 140
mW/cm2. Another approach for the ceria based elec-
trolyte is to use a bi-layer electrolyte with YSZ. Tsai et
al. [27] developed an electrolyte consisting of a 4 – 8 mm
thick Y2O3 doped ceria layer with a 1 – 1.5 mm thick
Y2O3 doped zirconia layer on the fuel side. The thin
bi-layer electrolyte provided a near theoretical open
circuit voltage and the electrolyte resistance was 0.25 W
cm2 at 600°C and 0.45 V cm2 at 550°C.
2430 O. Yamamoto / Electrochimica Acta 45 (2000) 2423–2435
Fig. 11. Performance of SOFC with 0.189 mm thick
La0.9Sr0.1Ga0.8Mg0.2O2.85 electrolyte and Sm0.5Sr0.5CoO3
cathode. (, ) 800°C, (, )600°C.
Certain perovskite type oxides exhibit high oxide ion
conductivity. An intermediate temperature SOFC with
a BaTh0.9Gd0.1O3 electrolyte, a La0.9Sr0.1CoO3 cathode,
and a Pd anode was investigated by Cook et al. [28]. A
cell voltage of 0.7 V was obtained at 200 mA/cm2 and
550°C. The thickness of the electrolyte was about 0.13
cm and the ionic conductivity was estimated to be
8.7×10 − 2 S/cm at 550°C from the current-voltage
curve. The observed open-circuit voltage was slightly
lower than the calculated one, which indicated a contri-
bution of electronic conduction in the electrolyte under
fuel cell conditions. Some yeas ago,
La0.9Sr0.1Ga0.8Mg0.2O2.85 (LSGM) was reported as a
superior oxide ion electrolyte by Ishihara et al. [29] and
Feng et al. [30]. It has an ionic conductivity of more
than 0.1 S/cm at 800°C, good chemical stability, and
negligible electronic conduction over a broad range of
oxygen partial pressures. A small single cell test with a
500 mm thick LSGM electrolyte showed promising per-
formance compared to a YSZ electrolyte at operating
temperatures of 600–800°C. A cell voltage of 0.7 V was
obtained at 500 mA/cm2 and 600°C with a Ni anode
and a Sm0.5Sr0.5CoO3 cathode as shown in Fig. 11 [31].
There is still a need to improve the stability for this
type of perovskite oxide used with a Ni anode under
reducing atmosphere. Ni reacts with an LSGM elec-
trolyte to form LaNiO3 [32]. Recently, the intrinsic
creep resistance of LSGM was measured by Wolfen-
stine [33] in the temperature range 800–1000°C. The
results revealed that the intrinsic creep resistance of the
Y2O3 and Sc2O3 doped cubic zirconia was much better
than that for LSGM over the entire temperature range.
This is most likely a result of the difference in crystal
structure.
Iwahara et al. [34] found that doped BaCeO3 and
doped SrCeO3 showed high proton conduction at ele-
vated temperatures. The conductivity of these per-
ovskite type proton conductors was similar or less than
those of zirconia based oxides. At lower temperature
Ba0.95Ce0.90Y0.10O3 – d showed a good conductivity as
shown in Fig. 8. However, the contribution of elec-
tronic conduction was non negligible.
Zirconia based oxides remain the best candidates for
the electrolyte in SOFCs. Yttria doped zirconia has
been used extensively in SOFCs, because of the rela-
tively high oxide ion conductivity, low electronic con-
ductivity, and stability under reducing and oxidizing
atmosphere. It is known that scandia doped zirconia
exhibit the highest conductivity of the zirconica based
oxides [16]. This material has not been used as an
electrolyte in SOFC, because of aging at high tempera-
tures reduces its stability and because of high cost of
scandia. The latter could probably be drastically re-
duced if scandia were used on a large scale, because the
natural abundance of scandium is similar to that of
lanthanum. The lack of long term stability of SSZ due
to aging is not yet well understood. Aging studies for
the ZrO2Sc2O3 system suggested that samples with 10
or 11 m/o Sc2O3 were stable without conductivity
degradation upon aging at a high temperature [35]. SSZ
with 10 or 11 m/o Sc2O3 is rhombohedral at room
temperature and forms the cubic phase at around
600°C. The temperature dependence of the conductivity
of SSZ exhibits an abrupt conductivity change at about
600°C, which corresponds to this phase transition. The
cubic phase can be stabilized by the substitution of
small amount of Y3 + for Sc3 + and by addition of
Al2O3 [35].
High electrolyte strength and toughness is desirable,
especially in planar configuration SOFC. In Table 3,
the mechanical properties of zirconia based oxides are
shown along with their electrical conductivity and the
thermal expansion coefficients. The bending strength of
11SSZ is almost comparable to that of 8YSZ and the
toughness is higher than that of 8YSZ. The bending
strength of YSZ was improved by addition of Al2O3
[35]. The composite of 11SSZ with Al2O3 is recom-
mended as an electrolyte in intermediate temperature
SOFCs, because of its high oxide ion conductivity,
phase stability, and excellent mechanical properties.
The cathode in SOFCs is of particular concern for
obtaining a high performance, especially at low operat-
ing temperature. The cathode materials for SOFCs
have to meet the following requirements: high electrical
conductivity at operating temperature, thermal and
chemical compatibility with the electrolyte, and high
catalytic activity for oxygen reduction. Early SOFC
stacks used a noble metal like platinum as the cathode.
In the 1960s doped In2O3 [36] was used, and in the
early 1980s LaMnO3 and LaCoO3 were proposed [2].
 O. Yamamoto / Electrochimica Acta 45 (2000) 2423–2435 2431

Table 3
Electrical and mechanical properties of zirconia based oxidesa

Materials Conductivity (S/cm) Bending strength (MPa) Thermal expansion coefficient (10−8/K)

1000°C 800°C

3YTZP 0.056 0.018 1200 10.8

3YbTZP 0.063 0.014
2.9STZP 0.09 0.031
8YSZ 0.16 0.03 279 10.5
8SSZ 0.38 0.13 275 10.7
8YbSZ 0.20 0.63
11SSZ 0.32 0.12 255 10.0
11SSZ1A 0.26 250
11SSZ20A 0.15 365 10.1

a
3YTZP: 3 m/o Y2O3ZrO2, 3YbTZP: 3 m/o Yb2O3ZrO2, 2.9SSZ: 2.9 m/o Sc2O3ZrO2, 11SSZ1A:(11 m/o Sc2O3ZrO2)-1 w/o
Al2O3, 11SSZ20A: (11m/oSc2O3ZrO2)-20 w/o Al2O3.

Takeda et al. [37] studied the electrical conductivity and More recently, Uchida et al. [43] reported an interest-
the catalytic activity for oxygen reduction of perovskite ing result concerning hydrogen oxidation on an active
oxide electrodes, La1 − x Srx MO3(MCr, Mn, Fe, Co). hydrogen electrode. The IR-free polarization of the
The cathodic polarization curves at 800oC for sputtered active anode attached to different types of zirconia
La1 − x Srx MO3 on YSZ are shown in Fig. 12. The electrolytes was greatly influenced by the conductivity
La1 − x Srx CoO3 electrode has high catalytic activity for of the zirconia used. The doped zirconia with higher
oxygen reduction. The conductivity measurements also ionic conductivity exhibited a higher exchange current
showed that the cobalt system had the highest conduc- density for the hydrogen electrode at 800 – 1000°C as
tivity in these perovskite systems, but the compatibility shown in Fig. 13, where Sm doped CeO2 with Ru was
with YSZ was less good. The thermal expansion coeffi- used as the anode catalyst. The results indicate that the
cient of LaCoO3 is :22–24×10 − 6/K and that of reaction rate of H2 oxidation was controlled by the
YSZ : 10–11 ×10 − 6/K. La1 − x Srx CoO3 reacts readily
with YSZ forming the low conductivity products,
La2Zr2O3 and SrZrO3 [38]. No reaction product has
been observed when coupling doped CeO2 and La1 −
xSrx CoO3. For this reason, the La1 − x Srx CoO3 systems
have been used together with CeO2 based electrolytes in
SOFCs at low operating temperatures. However, the
mismatch in thermal expansion between CeO2 and
La1 − x Srx CoO needs to be improved.
More recently, La1 − x Srx Fe1 − yCoyO3(LSFC) has
been examined extensively for SOFCs operating below
900°C [39]. This system showed a good electrical con-
ductivity at temperatures between 600 and 800°C, high
oxygen surface exchange coefficient and high oxygen
self diffusion coefficient. The oxygen self diffusion co-
efficient (D*) of LSFC at 500°C of 2.6 ×10 − 9 cm2/s
compared to that of La0.8Sr0.2MnO3 at 10 − 12 cm2/s at
1000°C [40]. The LSFC /Ce0.9Sm0.1O1.95 combination
exhibits fast electrode kinetics. LSFC does not react
with ceria-based electrolytes and has a compatible ther-
mal expansion coefficient [41]. Excellent behavior of the
cell Ni-YSZ/Ce0.8Gd0.2 O1.9/LSFC has been proved by
Berkel et al. of ECN [42]. A cell voltage of 0.7 V was
obtained at 150 mA/cm2 and 603°C with a self-sup- Fig. 12. Cathodic polarization curves at 800°C for a sputtered
ported 0.1–0.25 mm thick Ce0.8Gd0.2O1.9 electrolyte. La1 − x Srx MO3 film.
2432 O. Yamamoto / Electrochimica Acta 45 (2000) 2423–2435

1152 air electrode supported tubular type cells (2.2 cm

Fig. 13. Exchange current density for RuCe0.8Sm0.2O1.9 an-
ode versus conductivity of zirconia based electrolytes 10Sc: 10
m/o Sc2O3ZrO2, 8Yb: 8 m/o Yb2O3ZrO2, 8Y: 8 m/o
Y2O3ZrO2, 4Y: 4 m/oY2O3ZrO2.
supply of oxide ions through the electrode/zirconia
interface and that the rate was proportional to the ionic
conductivity. Similar results were observed for oxygen
reduction on the oxygen electrode of the
La0.85Sr0.15MnO3 with Pt as the catalyst [44]. It can thus
be seen that both low electronic resistance and high
oxide ion conductivity of the electrolyte are required to
obtain excellent cell performance in low temperature
SOFCs.
5. The future and prospect of SOFC
After the first report for SOFCs in 1937 and the
scientific paper in 1962 by Weissbart and Ruka [6],
many efforts have been made to develop a high perfor-
mance SOFC. Recently, large scale SOFCs have been
developed for power generation systems. Since Decem-
ber 1997, testing of a 100 kW SWH SOFC power
generation system has been carried out by a consortium
of Dutch and Danish utilities [45]. The system employs
diameter and 150 cm active length). The 100-kW SOFC
has achieved an electrical generation efficiency of 46%
(LHV) and at net system output of 106 kW to the grid.
Under the normal operation condition of 1000°C, 80%
fuel utilization is achieved. Emission measurements
confirmed an NOx level B 0.2 ppm with undetectable
levels of SOx, CO, and unburned hydrocarbons. At
present, the system has been in operation for 3700 h at
the test site, and except for one stack segment the
SOFC generator performed with good stability. The
costs of this system amounts to : 10 million US$.
MHI and Electric Power Development in Japan [46]
developed another type of tubular SOFC. A pressurized
10-kW class module was tested which employs 414 cells
(52 cm length) and operates at 900°C at 5 atm. A 15
kW out-put power was obtained at 200 mA/cm2. The
module has been operated for \5000 h.
Several types of planar SOFCs have been developed.
SWH in Germany operated a 10 kW class planar
configuration SOFC [47]. The cells were mounted be-
tween bipolar plates. A stack with 80 cell layers was
assembled, each layer consisting of 16 paralleled cells.
However, at this time SWH has given up developing a
planar SOFC and is now concentrating on a tubular
SOFC. The Chubu Electric Power Company and Mit-
subishi Heavy Industries have been evaluating a 5 kW
class planar type SOFC [48]. The cells have used a
ceramic interconnect of doped lanthanum chromite.
The support layers were fabricated with the same mate-
rials as the electrodes and the layers were corrugated to
allow the gas to pass through. A power density of 0.22
W/cm2 with fuel utilization of 38% was obtained. Ce-
ramic Fuel Cells in Australia [49] developed a 5 kW
planar type SOFC with stainless steel interconnects.
The 400 cells were fabricated from 10 cm× 10 cm×
100 mm 3YSZ electrolyte sheets produced by tape-cast-
ing. Sulzer Hexis in Switzerland [50] has developed a
planar type ceramic/metal hybrid stack with metallic
interconnects. Table 4 summarizes the current status of
the systems or modules of kilowatt or greater class
SOFCs. At this time, the SWH type is the most ad-
vanced SOFC. A large number of air electrode sup-

Table 4
Stack performance of large scale SOFCa

Developer Type Power (kW) Interconnect Remark

SWH Tubular 100 LaCr0.9Mg0.1O3 he =46%

SWH Tubular 25 LaCr09Mg0.1OO3 Operating hours: 13 100
MHI & EPDC Tubular 15 NiAl/Al2O3 Operating pressure: 5 atm
SWH Planar 10 Cr with 5%Fe and 1%Y2O3 5.4 kW at 950oC, 4.1 kW at 850oC
MHI&CEPCO Planar 5 Doped LaCrO3 he =22%

a
SWH: Siemens Whestinghouse Power Co., MHI: Mitsubishi Heavy Industries, EPDC: Electric Power Development Co.,
CEPCO: Chubu Electric Power Co., he: energy conversion efficiency.
 O. Yamamoto / Electrochimica Acta 45 (2000) 2423–2435
Fig. 14. Cell voltage and power density versus current density
for a SOFC operated on air and wet methane.
ported cells have been electrically tested for up to
10 000 h. The performance degradation was less than
0.2% per 1000 h. At present, the electric power out-put
efficiency is 46%, a value higher than that of advanced
conventional electric power plants. However, for practi-
cal applications to in electric power plants of several
MW or larger the system needs further improvements.
The most important point is the cost. The cost of SWH
type system can be as high as 100 000 US$/kW. To be
comparable to conventional electric power plants the
cost must be reduced to B3000 US$/kW. Several
approaches are possible such as using mixed lan-
thanides instead of high purity component oxides and
adopting cheaper deposition processes for cell produc-
tion. Toto and Kyushu Electric Power in Japan devel-
oped a SWH type tubular SOFC by a wet processing
technique [51]. A dense thin YSZ film (about 30 mm
thick) was prepared on a porous supported
(La,Sr)MnO3 tube (22 mm diameter, 900 mm length)
by slurry coating. (La, Ca)CrO3 was applied to form an
interconnecting strip by slurry coating. Successful re-
sults were obtained by this method for a 1.3 kW stack.
Medium scale SOFCs (1 MW to 50 kW) are also
attractive for distributed power generation. For this
application, it is necessary to meet many particular
requirements: maintenance free, high energy conversion
efficiency, and low cost. Maintenance free operation is
the greatest advantage of SOFCs systems. However, gas
turbine systems are strong competitors. The total en-
ergy efficiency of a gas turbine for the combined heat
and power generation is \60% and cost is estimated to
be B3000 US$/kW.
A small scale SOFC ( B50 kW) has been developed
for electric vehicle (EV) applications. The operation
conditions of fuel cells for EV applications differ con-
siderably from those of stationary applications. For
EV, the fuel cells must have sufficient power density
and be sufficiently inexpensive to compare well with
internal combustion engines on an economic basis. The
2433
preferred fuel cell technology for EV applications is the
polymer electrolyte type. It uses hydrogen or reformed
hydrocarbon as a fuel and is very sensitive to carbon
monoxide poisoning. In contrast, SOFCs are not ad-
versely affected by CO and can use fuel directly without
a reformer. However, the relatively high operating tem-
perature of current designs would require more time
and energy for startup. By reducing the operating tem-
perature to 500°C or less, these problems would be-
come manageable. More recently, Murry et al. [52]
reported the possibility of a direct-methane fuel cell
with 8 m/o Y2O3 doped ZrO2 electrolyte. The elec-
trolyte was prepared by sputter deposition of the elec-
trolyte on a porous La0.8Sr0.2MnO3 cathode. The anode
was porous Ni-YSZ coated Y2O3 doped CeO2. The cell
performance is shown in Fig. 14. The power density of
0.37 W/cm2 was obtained at 650°C. When Ni catalysts
are used with hydrocarbon feed stocks, they are prone
to deactivation through the formation of carbonaceous
deposits on the catalyst surface. An important result
was therefore the absence of carbon deposits in the
temperature range examined (650 – 550°C) Further-
more, Baker et al. [53] reported that no carbon deposits
were detected on the La0.8Ca0.2CrO3 anode during di-
rect oxidation of CH4 in the temperature range 400 –
600°C. These results suggest the possibility of using
SOFC for EV applications in the future.
6. Conclusion
Except for cost-effective manufacturing processes,
the development efforts over the past 30 years have
brought the technology of high temperature SOFC near
its final goal. On the other hand, technology of interme-
diate temperature SOFCs is still in a developmental
stage and several technological challenges remain to be
solved before this type of fuel cell can find practical
applications. A SOFC that uses hydrocarbon or
methanol and operates at B500°C may eventually
become an attractive power source for EV. More infor-
mation on the science and technology of SOFC up to
1994 can be found in the excellent book by Minh and
Takahashi [54] to which the reader is referred for
further information.
References
[1] W.R. Grove, Philos. Mag. 14 (1839) 127.
[2] N.Q. Minh, J. Am. Ceram. Soc. 76 (1993) 563.
[3] T. Kurihara, H. Mori, H. Ohmura, K. Ikeda, K. Tomida,
in: Proceedings of the fifth FCDIC fuel cell sysmposium,
Tokyo, Japan, 1996, p. 217.
[4] H. Baur, H. Preis, Z. Elektrochem. 6 (1937) 41.
[5] W. Nernst, Z. Elektrochem. 43 (1899) 727.
2434 O. Yamamoto / Electrochimica Acta 45 (2000) 2423–2435
[6] J. Weissbart, R. Ruka, J. Electrochem. Soc. 109 (1962)
723.
[7] C.S. Tedmon, H.S. Spacil, S.P. Mitoff, J. Electrochem.
Soc. 116 (1969) 1170.
[8] W. Wersing, E. Ivers-Tiffee, M. Schiessi, H. Greiner, in:
O. Yamamoto, M. Dokiya, H. Tagawa (eds), Proceeding
of the International Symposium on Solid Oxide Fuel
Cells, Nagoya, Japan, 1989, p. 33.
[9] S.C.Singhal, in: U. Stimming, S.C. Singhal, H. Tagawa,
W. Lehnert (eds), Proceeding of the fifth International
Symposium on Solid Oxide Fuel Cells, 1997, p. 37.
[10] F.J. Rohr, in: P. Hagnemuller, W. van Gool (Eds.), Solid
Electrolytes, Academic Press, New York, 1978, p. 431.
[11] A.O. Isenberg, W.A. Pabst, E.F. Sverdrup, D.H. Archer,
in: Proceedings of the sixth Biennial Fuel Cell Sympo-
sium, 1969.
[12] Westinghous Electric Corporation, Final Report, Project
Fuel Cell Rep. No. 57, 1970.
[13] J.P. Ackerman, J.E. Young, U.S. Patent No. 4476198,
October 9, 1984.
[14] M. Hattri, Y. Esaki, Y. Sakaki, F. Nanjyo, K. Takenobu,
H.M. Yamoto, in: Ext. Abstract of seventh Symposium
on Solid Oxide Fuel Cells in Japan, Nagoya, Japan, 1998,
p. 19.
[15] R. Doshi, Von L. Richard, J.D. Carter, X. Wang, M.
Krumpelt, J. Electrochem. Soc. 146 (1999) 1273.
[16] Y. Arachi, H. Sakai, O. Yamamoto, Y. Takeda, N.
Imanishi, Solid State Ionics 121 (1999) 133.
[17] J.K. Kilner, R.J. Brook, Solid State Ionics 6 (1982) 237.
[18] A.O. Isenberg, in: J.D.E. Mclntyre, S. Srinvasan, Will
(eds), Proceeding of the Symposium on Electrode Materi-
als and Processes for Energy Conversion and Storge,
Philadelphia, USA, Electrochem. Soc., NJ, USA, 1977,
p.572.
[19] N.Q.Minh, K.Montgomery, in: U. Stimming, S.C. Sing-
hal, H. Tagawa, W. Lehnert (eds), Proceeding of the fifth
International Symposium on Solid Oxide Fuel Cells,
1997, p153.
[20] Y. Mizutani, M. Kawai, K.Nomura, Y.Nakamura,
O.Yamamoto, in: U. Stimming, S.C. Singhal, H. Tagawa,
W. Lehnert (eds), Proceeding of the fifth International
Symposium on Solid Oxide Fuel Cells, 1997, p196.
[21] T.H. Estel, S.N. Feng, Chem. Rev. 70 (1970) 339.
[22] O. Yamamoto, Denki Kagaku 63 (1995) 698.
[23] T. Takahashi, T. Esaka, H. Iwahara, J. Appl. Elec-
trochem. 7 (1977) 303.
[24] T. Kudo, H. Obayashi, J. Electrochem. Soc. 122 (1975)
142.
[25] H.L. Tuller, A.S. Nowick, J. Electrochem. Soc. 126 (1979)
209.
[26] M. Godlickemeier, L.J. Gauckler, J. Electrochem. Soc.
145 (1998) 414.
[27] T. Tsai, S.A. Barnet, in: U. Stimming, S.C. Singhal, H.
Tagawa, W. Lehnert (eds), Proceeding of the fifth Inter-
national Symposium on Solid Oxide Fuel Cells, 1997, p.
274.
[28] R. Cook, J.J. Qsborne, J.H. White, R.G. MacDuff, A.F.
Sammells, J. Electrochem. Soc. 129 (1992) L19.
[29] T. Ishihara, H. Matuda, Y. Takita, J. Am. Chem. Soc.
116 (1994) 3801. . France, 1998, p. 87.
[30] M. Feng, J.B. Goodenough, Eur. J. Solid State Inorg.
Chem. 31 (1994) 663.
[31] T. Ishihara, T. Shibayama, M. Honda, H. Furutani, Y.
Takita, in: Abstract of 1998 Fuel Cell Seminar, Palm
Springs, USA, 1998, p. 104.
[32] K. Huang, M. Feng, J.B. Goodenough, C. Milliken, J.
Electrochem. Soc. 144 (1997) 3629.
[33] J. Wolfenstine, Electrochem. Solid-State Lett. 2 (1999)
210.
[34] H. Uchida, N. Maeda, H. Iwahara, Solid State Ionics 11
(1983) 117.
[35] Y. Mizutani, M. Tamura, M. Kawai, K. Nomura, Y.
Nakamura, O. Yamamaoto, in: M. Dokiya, O. Ya-
mamoto, H. Tagawa, S.C. Singhal (eds), Proceedings of
Fouth Internationa Symposium on Solid State Fuel Cells,
Nagoya, Japan, Electrochem. Soc., NJ, USA, 1995, p.
301.
[36] E.F. Sverdrup, D.H. Archer, A.D. Glaser, Am. Chem.
Soc. Adv. Chem. Ser. 90 (1969) 301.
[37] Y. Takeda, R. Kanno, M. Noda, Y. Tomida, O. Ya-
mamoto, J. Electrochem. Soc. 134 (1987) 2656.
[38] O. Yamamoto, Y. Takeda, R. Kanno, M. Noda, Solid
State Ionics 22 (1987) 241.
[39] L-W. Tai, M.M. Nasrallah H.U. Anderson, in: S.C. Sing-
hal H. Iwahara (eds), Proceeding of the Third Interna-
tional Symposium on Solid Oxide Fuel Cells, Honolulu,
USA, Electrochem. Soc., NJ, USA, 1993, p. 241.
[40] B.C.H. Steele, Solid State Ionics 86/88 (1996) 1223.
[41] L. Kindermann, D. Dos, H. Nickel, K. Hilpert, C.C.
Appel, F.W. Poulsan, J. Electrochem. Soc. 144 (1997)
717.
[42] F.P.F. van Berkel, G.M. Christie, F.H. van Heuveln,
J.P.F. Huijsmans, in: M. Dokiya, O. Yamamoto, H.
Tagawa, S.C. Singhal (eds), Proceedings of Fouth Inter-
nationa Symposium on Solid State Fuel Cells, Nagoya,
Japan, Electrochem. Soc., NJ, USA, 1995, p. 1062.
[43] H. Uchida, H. Suzuki, M. Watanabe, J. Electrochem.
Soc. 146 (1998) 615.
[44] H. Uchida, M. Yoshida, M. Watanabe, J. Electrcochem.
Soc., Submitted.
[45] R.J.F. van Gerven, J.H.C. van der Veer, in: Abstracts of
1998 Fuel Cell Seminar, Palm Springs, USA, 1998, p. 76.
[46] H. Mori, H. Ohmura, K. Ikeda, K.A. Tomida, in: Ab-
stracts of Seventh Symposium on Solid Oxide Fuel Cells
in Japan, Tokyo, Japan, 1998, p.27.
[47] H.J. Beie, L.W. Drenckhahn, H. Greiner, B. Rufolf, H.
Schichl, in: U. Stimming, S.C. Singhal, H. Tagawa, W.
Lehnert (eds), Proceeding of the fifth International Sym-
posium on Solid Oxide Fuel Cells, 1997, p. 51.
[48] Y. Sakaki, Y. Esaki, M. Hattori, H. Miyamoto, T. Sa-
take, F. Nanjo, T. Matudaira, K. Takenobu, in: U.
Stimming, S.C. Singhal, H. Tagawa, W. Lehnert (eds),
Proceeding of the fifth International Symposium on Solid
Oxide Fuel Cells, 1997, p. 61.
[49] S.P.S Badwal, K. Foger, in: Proccedings of Third Eu-
ropean Solid Oxide Fuel Cell Forum, Nantes, France,
1998, p. 95.
[50] R. Diethelm, M. Schmidt, B. Doggwiler, Th. Gamper, M.
Keller, K. Honegger E. Batwai, in: Proceedings of the
Third European Solid Oxide Fuel Cell Forum, Nante,
 O. Yamamoto / Electrochimica Acta 45 (2000) 2423–2435 2435

[51] M. Aizawa, M. Kuroishi, H. Takeuchi, H. Tajiri, T.
Nakayama, in: Abstracts of 1998 Fuel Cell Seminar,
Palm Springs, USA, 1998, p. 270.
[52] E.P. Murray, T. Tsai, S.A. Barnett, Nature 400 (1999)
469.
[53] R.T. Baker, I.S. Metcalfe, in: M. Dokiya, O. Ya-
mamoto, H. Tagawa, S.C. Singhal (eds), Proceedings of
Fouth Internationa Symposium on Solid State Fuel
Cells, Nagoya, Japan, Electrochem. Soc., NJ, USA,
1995 p. 781.
[54] N.Q. Minh, T. Takahashi, Science and Technology of
Ceramic Fuel Cells, Elsevier, Amsterdam, 1995.