J O U R N A L O F M A T E R I A L S S C I E N C E 3 9 (2 0 0 4 ) 4405 – 4439

A review of anode materials development in solid

oxide fuel cells

SAN PING JIANG, SIEW HWA CHANFuel Cells Strategic Research Program, School of Mechanical and Production Engineering,Nanyang Technological University, Nanyang Avenue, Singapore 639798E-mail: mspjiang@ntu.edu.sg

High temperature solid oxide fuel cell (SOFC) has prospect and potential to generateelectricity from fossil fuels with high efficiency and very low greenhouse gas emissions ascompared to traditional thermal power plants. In the last 10 years, there has been significantprogress in the materials development and stack technologies in SOFC. The objective ofthis paper is to review the development of anode materials in SOFC from the viewpoint ofmaterials microstructure and performance associated with the fabrication and optimizationprocesses. Latest development and achievement in the Ni/Y2O3-ZrO2 (Ni/YSZ) cermetanodes, alternative and conducting oxide anodes and anode-supported substrate materialsare presented. Challenges and research trends based on the fundamental understanding ofthe materials science and engineering for the anode development for the commerciallyviable SOFC technologies are discussed. C© 2004 Kluwer Academic Publishers

1. IntroductionSolid oxide fuel cell (SOFC) is an all solid device thatconverts the chemical energy of gaseous fuels such ashydrogen and natural gas to electricity through elec-trochemical processes. SOFC, being an electrochemi-cal device, has unique advantages over the traditionalpower generation technologies. The efficiency of SOFCis inherently high as it is not limited by the Carnot cycleof a heat engine. The greenhouse gas emissions fromSOFC are much lower than those emitted from conven-tional power plants. Due to the high operating temper-ature, SOFC can be used as a co-generation to producehot water or steam and to couple with microturbines orgas turbines to produce electrical power which enhancethe system efficiency and the range of applications. Themomentum of the intensive research and developmentof SOFC technology started in the middle seventieswith the Westinghouse (now Siemens-Westinghouse)tubular SOFC development program [1]. Minh, almosta decade ago in 1993, gave a comprehensive overviewon the whole spectra of the SOFC technologies [2].Since then there have been tremendous progresses andachievements in both the SOFC technologies in ma-terials, design and fabrication and in the fundamentalunderstanding of the mechanism and kinetics of elec-trode reactions in SOFC. This has been reflected by anumber of reviews on various topics of SOFC in re-cent years. Yokokawa et al. [3, 4] reviewed the devel-opment of materials from the view point of chemicalinteractions between various fuel cell component ma-terials. Various aspects of material related issues werealso discussed by Badwal and Foger [5], Kawada andYokokawa [6], Steele [7], Huijsmans [8] and Badwal[9]. Carrette et al. [10], Yamamoto [11], Singhal [12]

and De Jonghe et al. [13] reviewed some fundamentaland technological issues in SOFC. For a comprehen-sive and general treatment of fundamentals and latestdevelopment of SOFC technology, readers should referto a very recent book by Singhal and Kendall [14].

Current activities in the areas of materials develop-ment for SOFC are increasingly focused on the decreaseof operating temperatures of the SOFC from tradition-ally 1000◦C to 500–800◦C to reduce materials cost andimprove the performance stability. To compensate forthe increase in ohmic losses at reduced temperatures,electrolytes with higher ionic conductivity or thinnerelectrolyte structures are needed. As the electrolytethickness decreases, the overall cell polarization lossesare increasingly dominated by the losses of the electro-chemical reactions at anodes and cathodes. Thereforeit is crucial to understand the important issues asso-ciated with anode materials and anode-supported cellstructures in the development and optimization of in-termediate temperature SOFC (ITSOFC).

In SOFC, anode is the electrode for the electrochem-ical oxidation of fuels such as hydrogen and naturalgas. Hydrogen oxidation is one of the most importantelectrode reactions in SOFC. The electrochemical ox-idation of H2 can be written, following Kroger-Vinknotation, as:

H2 + O2−YSZ = H2O + 2e + VO,YSZ (1)

where O2−YSZ is an oxygen ion in the Y2O3-ZrO2 (YSZ)

electrolyte lattice site and VO,YSZ is an oxygen va-cancy in YSZ. To minimize the polarization losses ofthe H2 oxidation reaction, anode materials should meetthe basic requirements of high electronic conductivity,

0022–2461 C© 2004 Kluwer Academic Publishers 4405

sufficient electrocatalytic activity for fuel oxidation re-actions, chemically stable and thermally compatiblewith other cell components and has sufficient poros-ity for efficient gas transportation in high-temperaturereducing environment [2].

Due to the requirements of operating in reducingenvironment and high electronic conductivity, pureporous metallic electrodes, in principle, can be used asanode. Several metals such as Ni [15–17], Pt [17–19]and Ru [20] have been studied as anode materials. Se-toguchi et al. [21] studied the electrochemical activityof Ni, Co, Fe, Pt, Mn and Ru and found that Ni exhibitsthe highest electrochemical activity for H2 oxidation re-action. Pure nickel has a melting point of 1453◦C, a ther-mal expansion coefficient of 13.3 × 10−6 cm·cm−1k−1

and electronic conductivities of 138 × 104 S cm−1 and∼2×104 S cm−1 at 25◦ and 1000◦C, respectively. Rel-atively low melting temperature results in the tendencyof lower sintering temperature (1000◦C) and the ther-mal expansion coefficient of Ni is much higher than thatof YSZ electrolyte (the thermal expansion coefficientof YSZ is in the range of ∼10.5 × 10−6 cm·cm−1K−1).Addition of YSZ electrolyte phase into Ni significantlyreduces the thermal expansion of the composite ma-terial in order to be thermally compatible with theelectrolyte [22, 23]. It is generally accepted that theelectrochemical activity of Ni anodes for H2 oxidationreaction depends strongly on the three phase bound-ary (TPB where fuel gas, Ni and YSZ phases are met)areas [24–26]. De Boer et al. [27] observed the sig-nificant reduction of electrode polarization resistanceof porous Ni electrode modified by deposition of fineYSZ particle as compared to pure Ni electrode. Thisindicates that YSZ phase in the cermet plays an impor-tant electrocatalytic role in the creation of additionalreaction sites by extending the two-dimensional reac-tion zone into three-dimensional reaction zone, signif-icantly enhancing the reaction kinetics in addition tothe inhibiting of the coarsening and grain growth ofthe Ni phase [28–30]. Therefore the system of com-posite anode usually consisting of metal-oxide cermetstructure has been overwhelmingly accepted. Today,the Ni/Y2O3-ZrO2 (Ni/YSZ) cermet materials are stillthe most common anodes for SOFC.

Fuel cells that directly use hydrocarbon fuels suchas natural gas, without first reforming those hydrocar-bons to hydrogen, will have enormous advantages overconventional hydrogen-based fuel cells [31]. Remov-ing the need either to supply hydrogen to the fuel cellor to include a hydrocarbon-reforming system greatlydecreases the complexity, size and cost of the fuel cellsystem. Natural gas is regarded as a relatively cheap andclean fuel. The advantage of SOFC over other fuel cellsis that the natural gas can be directly used without theneed to external fuel reformer and water-gas shift reac-tor. The main component of natural gas is methane andfollowing reforming reactions can take place directlyinside a Ni/YSZ cermet anode:

CH4 + H2O = 3H2 + CO (2)

CO + H2O = CO2 + H2 (3)

Internal reforming of hydrocarbon fuels is often ac-companied by carbon deposition. Carbon depositioncovers the active sites of the anodes, resulting in theloss of cell performance [32, 33]. High steam/carbon(S/C) ratios (e.g., up to 3) are typically used in con-ventional steam reformers to suppress the carbon for-mation. However, high S/C ratio is unattractive for fuelcells as it lowers the electrical efficiency of the fuelcell by steam dilution of the fuel. The endothermicnature of steam reforming reaction (Equation 2) cancause local cooling and steep thermal gradients poten-tially capable of mechanically damaging the cell stack[34]. Nevertheless, the impact of the endothermicityof methane steam reforming process on the cell stabil-ity can be reduced by the combination of both exter-nal and internal reforming activity [35] or by using theso called gradual internal reforming (GIR) of methane[36, 37].

Despite the excellent electrocatalytic properties ofNi/YSZ cermet materials for operation in H2 fuel,Ni/YSZ based anode suffers a number of drawbacks insystems where natural gas is used as the fuel, notablythe sulfur poisoning and carbon deposition caused bycracking of methane. Under high carbon activity envi-ronment iron, nickel, cobalt and alloys based on thesemetals could corrode by a process known as metal dust-ing. Metal dusting involves the disintegration of bulkmetals and alloys into metal particles at high temper-atures (300–850◦C) in environment that are supersat-urated with carbon. Ni corrosion process strongly de-pends on the temperature and the gas composition andin general Ni corrosion rate increases with tempera-ture [38]. The form of carbon deposited on Ni in wetmethane was found to be graphite as observed by in situRaman microspectroscopy and the deposited carboncannot be burned by electrochemically permeated oxy-gen [39]. There is also increased possibility for nickeloxidation at temperatures below 700◦C as the result ofthe increased partial pressure of oxygen under typicaloperating conditions [40]. The large volume change in-volved in the oxidation and reduction cycle of Ni/NiO(theoretical density is 8.9 gcm−3 for Ni and 6.96 gcm−3

for NiO) could cause the instability of the Ni/YSZ cer-met microstructure. The performance of conventionalNi/YSZ cermet anode is also not satisfactory at tem-perature range of 500 to 600◦C due to the low ionicconductivity of YSZ phase in the cermet. Thus, thedriving force for the development of alternative anodesis mainly due to the need to replace Ni/YSZ cermet an-odes for the direct oxidation of hydrocarbon fuels suchas methane and to develop dimensionally stable anodeswith high mixed electronic and ionic conductivity forlow temperature SOFC [41, 42].

The purpose of this paper is to review the achieve-ments and progresses in the development of Ni/YSZbased anode materials and alternative conducting ox-ide anode materials in solid oxide fuel cells. Materialsdevelopment for anode-supported structures is also re-viewed. Emphasis will be placed on the developmentand progress in the last 10 years. Finally the directionand strategies in the development and optimization ofSOFC anode materials are discussed.

4406

Figure 1 A generalized processing route for the preparation of Ni/YSZ cermet anodes.

2. Development of Ni/YSZ cermet anodes2.1. Issues in the fabrication of Ni/YSZ

cermet anodesConventional Ni/YSZ cermet anodes are usually madeof commercial NiO and YSZ powders, which are thenhomogenized by mechanical mixing and milling. TheNiO/YSZ ink is applied to the YSZ electrolyte and sin-tered to form a porous Ni/YSZ cermet electrode. Ex-tensive literature shows that performance and electricproperties of Ni/YSZ cermet anodes are critically de-pendent on the microstructure and distribution of Niand YSZ phases. This in turn is dependent on the fab-rication process, characteristics of the starting NiO andYSZ powders, sintering behavior of YSZ powders, etc.[43, 44]. Fig. 1 shows a general processing route forthe preparation of Ni/YSZ cermet anodes, based on theconventional ceramic powder mixing process. Due to itsgeneral application in the fabrication and optimizationof SOFC anodes, important issues in the fabrication andoptimization of Ni/YSZ cermet anodes are discussedfirst.

2.1.1. Starting powderCharacteristics of the NiO and YSZ powders have sig-nificant effect on the fabrication process and the elec-trochemical performance of the Ni/YSZ cermet anodes.This is due to the fact that the properties such as the av-erage particle size, the particle size distribution and sin-tering behavior (e.g., shrinkage and shrinkage rate) ofNiO and YSZ powders are very different from differentsuppliers. Fig. 2 shows example of particle size distribu-tion of various commercial NiO powders. NiO powdersare ranging from very fine and broad distribution (e.g.,AJAX NiO) to coarse and symmetric distribution (e.g.,QN NiO). Difference in the powder characteristics willhave significant effect on the sintering behavior andphysical properties of the powder. Tietz el al. [45] eval-

Figure 2 Particle size distribution of as-received commercial NiO pow-ders. The particle size distribution was measured by laser scatteringmethod.

uated eight different commercial NiO powders. The av-erage grain size varied greatly from 0.5 to 14.7 µm withBET surface area from 0.2 to 47 m2/g. Consequently,the sintering behavior of the powder differred signifi-cantly. The shrinkage of NiO powder ranged from 12to 27% and the maximum shrinkage rate occurred be-tween 660 to 1085◦C, depending on the supplier. In gen-eral, high surface area corresponds to a high shrinkageand low starting sintering temperature. Similarly forcommercial YSZ powders such as 3 mol% Y2O3-ZrO2(3YSZ) and 8 mol% Y2O3-ZrO2 (8YSZ), the charac-teristics of the powder also vary considerably. For ex-ample, Ciacchi et al. [46] found that the specific surfaceareas of commercial YSZ powders change from 7.5 to26 m2 g−1.

In the conventional powder mixing process, elec-trode performance was found to be affected by theinitial particle size of YSZ and NiO powders. Hikita

4407

[47] showed that the electrochemical performance ofNi/8YSZ cermet anodes is related to the particle sizeratio of 8YSZ/NiO of starting oxide powders. The mini-mum anode overpotential losses were obtained at initial8YSZ/NiO particle size ratio of ∼0.01. Murakami et al.[48] found that 8YSZ particle size not only affects theelectrode performance but also the stability and the vol-ume concentration of the cermet. The larger the 8YSZsize is, the higher the 8YSZ content in the cermet willbe required to achieve the best performance. The bestperformance was observed on the Ni/8YSZ cermet an-odes prepared from starting 8YSZ and NiO powderswith particle size of 0.5 and 2.5 µm, respectively. Thiscorresponds to an 8YSZ/NiO particle size ratio of ∼0.2.In the case of Ni/3YSZ cermet anodes, the effect of par-ticle size of the 3YSZ powder on the polarization per-formance of Ni/3YSZ cermet anodes is almost linearand the best Ni/3YSZ cermet anode for the H2 oxida-tion was prepared from 3YSZ and NiO powders withparticle size of 0.06 and 0.14 µm, respectively [49]. AsNiO particle size did not change in this case, this meansthat the polarization performance (or the electrode po-larization conductivity) decreases with the increase in3YSZ/NiO ratio. The discrepancies may be due to thefact that the particle size ratio in the cermet (or the fi-nal anode morphology) is significantly affected by theprocessing steps such as coarsening treatment of YSZand Ni/YSZ cermet powders.

2.1.2. Powder coarsening treatmentCoarsening treatment of starting powder is one of themost important steps in the preparation of Ni/YSZ cer-met anodes and is frequently used to control the powdercharacterization (e.g., particle size and size distribu-tion) and the shrinkage profile of the cermet coating inorder to have high coating quality and reproducibility.

Careful control of shrinkage profile of the Ni/YSZcermet powders is particularly important in the pro-duction scale-up process. In practice, sintering profileof the cermet powder can be adjusted in certain degreesby heat treatment of starting NiO and zirconia powders.Jiang et al. [44, 49] studied the effect of coarsening NiOpowders on the polarization performance of Ni/3YSZcermet anodes. Fig. 3 illustrates a typical particle sizedistribution of NiO as a function of coarsening temper-atures for three commercial NiO powders. As-receivedAJAX NiO powder shows very fine particles and broadparticle size distribution (see Fig. 2). Average parti-cle size determined from distribution measurement was∼1 µm. Coarsening AJAX NiO powder at 900◦C dra-matically changed the powder characteristics, indicatedby the increased average particle size and narrowed par-ticle size distribution (Fig. 3a). On the other hand, heattreatment at temperatures below 900◦C has much lesseffect on the characteristics of HNO-300 NiO powders(Fig. 3b). As-received NFP NiO powder has narrowand symmetric particle size distribution and coarsen-ing treatment is mainly to increase the average particlesize (Fig. 3c). The difference in the characteristics ofthe NiO powders in relation to the coarsening treat-ment has direct effect on the electrode performance.

Figure 3 Change of particle size distribution of NiO as a function ofcoarsening temperatures for three different commercial NiO powders:(a) AJAX NiO, (b) HNO-300 NiO, and (c) NFP NiO.

For Ni/3YSZ cermet anodes prepared from AJAX NiOpowder, the polarization losses reached the minimumfor the anode prepared from NiO coarsened at ∼600◦C(Fig. 4a). For Ni/3YSZ cermet anodes prepared fromHNO-300 NiO powder, the electrode performance re-mained more or less the same for the NiO coarsened attemperatures below 900◦C (Fig. 4b).

Particle size and distribution of YSZ powder are animportant factor in the performance of Ni/YSZ cermetanodes. Heat treatment or coarsening of YSZ powdersis an effective way to control the particle size distribu-tion of YSZ powders. Van Herle et al. [50] reported theperformance of Ni/8YSZ cermet anodes prepared from

4408

Figure 4 Electrode performance of Ni (50 vol%)/3YSZ (50 vol%) cer-met anodes as a function of coarsening temperature of (a) AJAX NiO and(b) HNO-300 NiO. Overpotential was measured under a current densityof 250 mA cm−2 at 1000◦C in 97%H2/3%H2O.

8YSZ powder with size distribution of 0.2–15 µm andspecific area of 0.52 m2 g−1. The particle size distribu-tion of 8YSZ powder was obtained by coarsening of thepowder at 1400◦C and followed by milling. The anodedisplayed a good stability and performance (overpo-tential loss of 150 mV at 1 A cm−2 and 800◦C). Itohet al. [51] showed that by combining large (27 µm) andfine (0.6 µm) YSZ particles, stability and performanceof Ni/YSZ cermet anodes were improved substantially.Nevertheless, there may be some limitation in the useof large YSZ particles as large YSZ particles couldprecipitate from the ink suspension and the separationin the ink will make the quality control of the anodevery difficult during the screenprinting or ink coatingprocess. A large volume of coarse YSZ particles couldalso incur week contact at anode and YSZ electrolyteinterface.

In Ni/YSZ cermet anodes, performance of the an-ode is affected by the coarsening treatment of YSZand NiO/YSZ cermet powders and the sintering tem-perature of the cermet coating. However, the effectof coarsening of Ni/YSZ cermet powders appears tobe related to the nature of YSZ powder in the cer-met [52]. Fig. 5 shows the polarization performance ofNi/3YSZ and Ni/8YSZ cermet anodes prepared fromcermet powders coarsened at different temperatures.

Figure 5 Polarization performance of Ni (70 vol%)/3YSZ (30 vol%)and Ni (50 vol%)/8YSZ (50 vol%) cermet anodes prepared from cermetpowders coarsened at different temperatures. Polarization curves weremeasured at 1000◦C in 98%H2/2%H2O and anodes were sintered at1400◦C in air.

The anodes were sintered at 1400◦C in air. Polariza-tion performance was measured in wet H2 at 1000◦Cand iR components were not distracted from the po-larization potentials. For Ni/8YSZ cermet anodes, thebest performance was obtained on the anode preparedfrom cermet powder coarsened at 1300◦C. In the case ofNi (40 vol%)/8YSZ (60 vol%) cermet anodes, Kawadaet al. [53] observed the best performance of electrodepolarization resistance (RE) of 0.29 �cm2 as measuredat 1000◦C in wet hydrogen for the Ni/8YSZ cermetanodes prepared with cermet powder coarsening tem-perature of 1400◦C and anode sintering temperature of1500◦C. On the other hand, the coarsening treatment ofNiO/3YSZ cermet powder show little effect on the po-larization performance of the Ni/3YSZ cermet anodes(Fig. 5a), in contrast to that of the Ni/8YSZ cermetanodes [49, 52].

The fundamental reason for this significant differ-ence in heat treatment effect of YSZ and Ni/YSZ cermetpowder on the anode performance is most likely relatedto the very different sintering behavior of YSZ powders.In the case of YSZ powders, the sintering behavior isrelated to the yttria content. Fig. 6 shows the change ofthe average particle size of Tosoh 3YSZ, Tosoh 8YSZpowder and AJAX NiO with the heat treatment temper-ature [52]. The significant grain growth as observed for8YSZ powders indicates that the grain growth kinetics

4409

Figure 6 Change of the average particle size of Tosoh 3YSZ, Tosoh8YSZ and AJAX NiO powders with coarsening temperature.

of 8YSZ powder is much faster as compared to the rela-tively sluggish grain growth of 3YSZ and NiO powders.Matsushima et al. [54] studied the effect of sinterabil-ity of 8YSZ powders on the electrode performance ofNi/8YSZ cermet anodes. 8YSZ powder has shrinkagerate of 26% which is close to that of the NiO powder.However, the sintering profile of the NiO/8YSZ cer-mets is primarily dominated by that of the 8YSZ phaserather than that of the NiO phase. The maximum sin-tering rate for 8YSZ was found to occur at ∼1300◦C.Upadhyaya et al. [55] studied the densification behav-ior of 3 mol% Y2O3-ZrO2 powder prepared by the co-precipitation method. They found that the maximumshrinkage occurred at 1020◦C; about 150 to 200◦C de-grees lower than that of 8YSZ powders. As suggestedby Lange [56], the maximum sintering rate correspondsto a transition from densification kinetics to coarsening(i.e., grain growth) kinetics. Thus, reducing the sinter-ing and grain growth of 8YSZ powders in the cermetto the level comparable to that of NiO powder wouldbenefit the establishment of the effective Ni-to-Ni con-tact in the cermet, indicated by the significant reductionin the electrode ohmic resistance of Ni/8YSZ cermetanodes prepared from the coarsened cermet powders[52].

2.1.3. Sintering temperatureof anode coatings

Sintering of the Ni/YSZ cermet coating at high tem-peratures (e.g., 1400◦C) is essential to achieve highelectrode performance and low electrode ohmic resis-tance. Jiang [52] showed recently that the formationof Ni-to-Ni electronic contact and YSZ-to-YSZ ioniccontact networks are closely related to the sinteringtemperatures of the anode. Fig. 7 shows the polariza-tion performance of Ni/3YSZ and Ni/8YSZ cermet an-odes sintered at different temperatures in wet H2 at1000◦C. In the case of Ni/8YSZ cermet anodes, theanodes were prepared from cermet powder coarsenedat 1300◦C. Fig. 8 shows the SEM pictures of Ni/8YSZcermet electrodes sintered at different temperatures af-ter fuel cell testing. The lowest electrode ohmic and

Figure 7 Polarization curves of Ni (70 vol%)/3YSZ (30 vol%) and Ni(50 vol%)/8YSZ (50 vol%) cermet anodes sintered at different temper-atures. The performance was measured at 1000◦C in 97%H2/3%H2O.

polarization resistance was observed for Ni/3YSZ andNi/8YSZ cermet anodes sintered at 1400◦C (Fig. 7).This corresponds to the formation of good YSZ-to-YSZnetwork for the anodes sintered at 1400◦C (Fig. 8). Theobservation that high sintering temperature basicallyhas no effect on the conductivity of pure Ni anode [52]indicates that sintering of YSZ phase in the cermet hassignificant effect not only on the formation of Ni-to-Nielectronic contact network in the cermet but also on theformation of electrical contact between the Ni phasein the cermet and YSZ electrolyte. On the other hand,high sintering temperature is essential to create a goodbonding between the YSZ phase in the cermet and theYSZ electrolyte, leading to the formation of a rigidYSZ structure to support the Ni phase and the for-mation of Ni-to-Ni electronic contact network. Thisappears to be the reasonable explanation for the highelectrode performance and high electronic conductiv-ity of the anode coatings sintered at high tempera-tures. Fukui et al. [57] studied the sintering behaviorof Ni/YSZ cermet anodes prepared by spray pyroly-sis. Electrode ohmic resistance and anode overpoten-tial decreased significantly with the increase of anodesintering temperature. The best performance was ob-served for Ni/YSZ cermet anode sintered at 1350◦Cwith lowest iR and polarization losses. Similar effect ofanode sintering temperature on the electrode ohmic andpolarization resistance was also reported by Primdahlet al. [58].

4410

Figure 8 SEM pictures of Ni (50 vol%)/8YSZ (50 vol%) cermet anodes sintered at (a) 1300◦C, (b) 1350◦C, (c) 1400◦C and (d) 1500◦C differenttemperatures after the fuel cell testing. The anodes were prepared from cermet powder coarsened at 1300◦C.

2.1.4. Synthesis of cermet powdersIn addition to the conventional ceramic mixing processbased on commercial NiO and YSZ powders, Ni/YSZmaterials can also be prepared by other methods suchas wet chemical synthesis process. The purpose of thesynthesis of the Ni/YSZ cermet powders is primarily toimprove the homogeneity of Ni and YSZ phase distri-bution, to reduce the sintering temperature of the cer-met coating and to increase the electrode performanceand stability. A major consideration in the synthesis ofNi/YSZ powders is the effect of the process on the pow-der morphology and phase distribution. Millini et al.[59] showed that the powder morphology of NiO/YSZpowders prepared by citrate route could change fromspherical shape by spray-drier to irregular aggregates byrotary evaporation. Li et al. [60] employed NH3·H2O-NH4HCO3 buffer-solution to co-precipitate Ni/YSZpowder with improved homogeneity of the powdercomposition and uniformity of the particle size. Us-ing NaOH as the co-precipitation agent can avoid theNi loss, resulting in the accurate control of the cermetcomposition [61]. Fine Ni/YSZ cermet powder withsurface area of 27–32 m2 g−1 and average particle sizeof 0.25–0.8 µm, prepared by solution combustion of ni-trate salt and carbohydrazide fuel followed by hydrogenreduction at 800◦C [62], have been reported. Combus-tion synthesis method is also commonly employed toproduce Ni/YSZ and (Ni,M = Co,Fe,Cu)/YSZ cermet

powder [63–65]. In this method, nitrate precursorsof the cermet constitutes such as ZrO(NO3)2·6H2O,Y(NO3)3·6H2O and Ni(NO3)2·6H2O are mixed inti-mately, melted together with urea (CO(NH2)2) on a hotplate and the combustion takes place in a furnace pre-heated at 600◦C. Reported results showed that the spe-cific area of Ni/YSZ powder with 50 vol% YSZ phasewas 26 m2/g and decreased to 17 m2 g−1 with the ad-dition of 2% Cu. The electrode polarization resistanceof a symmetrical cell was estimated to be 0.2 �cm2 at1000◦C, a reasonable value compared to those reportedfor optimized Ni/YSZ cermet anodes prepared by con-ventional powders [66]. Ni/YSZ cermet powders couldalso be prepared by mixed citrate/nitrate combustionsynthesis [67].

Spray pyrolysis was used to prepare Ni/YSZcomposite powders from 8 mol% Y2O3-ZrO2 andNi(CH3COO)2·4H2O solution [68, 69]. The Ni/YSZpowder obtained by this method showed spherical mor-phology with average particle size of ∼1 µm. The cer-met anode prepared from the synthesized compositepowder has good distribution of Ni and YSZ phasesand the cell prepared from the Ni/YSZ cermet anodesusing the spray pyrolysis powder showed no deteriora-tion over 8000 h at 1000◦C in moist H2/air. The samemethod was also used to prepare Ni/SDC cermet pow-der with Ce(NO3)3, Sm2O3 and Ni(CH3COO)2·4H2Oraw materials. Ni/SDC powder obtained by this method

4411

had average particle size of ∼0.5 µm and specific areahigher than 2 m2 g−1 [70, 71]. Chung et al. [72] usedgas atomization method to prepare Ni/CeO2 as poten-tial materials for internal reforming anodes. Gas atom-ization can be defined as the break up of a liquid intofine droplets by the impringement of gas. The prod-uct is a suspension of very fine liquid droplets whichthen solidify very quickly. The rapid cooling intrinsicto gas atomization allows very little time for segre-gation to occur. Using intermetallic precursor, CeNi5,Chung et al. [72] produced Ni/CeO2 particles with grainsize ranging from 5 µm to as large as 100 µm. How-ever, the electrochemical performance of the anodesprepared from the gas atomized Ni/CeO2 powder wasnot as good as Ni/CeO2 anodes prepared by conven-tional ceramic mixing method. Chen and Liu [73] re-ported the preparation of mesoporous YSZ/NiO com-posite with very high surface area of 108 m2 g−1 usingPluronic P103 surfactant as structure-directing agentsand metal chlorides as precursors. Ultrafine Ni/YSZpowder and nano-sized NiO and SDC powder can alsobe prepared by sol-gel and liquid mixture methods [74]and by heating the oxalate precursors at 300 to 1200◦Cin air [75], respectively. Gd-doped CeO2 (GDC) pow-ders produced by metal nitrate-glycine process can bedensified at temperatures as low as 1250◦C and thisled to the decrease in sintering temperature of theNi/GDC cermet anodes [76]. Moon et al. [77] reportedthe preparation of ZrO2-coated NiO powder by thermalhydrolysis of Zr(NO3)2·6H2O in a mixed solvent ofiso-PrOH/water. Using advanced mechanical methodin dry process can also produce YSZ-covered submi-cron NiO composite powders [78]. The presence offine YSZ particles on the surface of NiO would signifi-cantly inhibit the grain growth of NiO phase in the cer-met. Other methods employed to prepare Ni/YSZ cer-met anodes include vapor-deposition [79], wet-powderspraying [80] and mechanical milling and plasma spray[81].

Cracium et al. [82] reported a new method for thefabrication of SOFC anodes. In this method, a porousYSZ coating was formed using a mixture of YSZ pow-der and YSZ fibers, followed by impregnation of theporous YSZ coating with aqueous solution of Ni, orCu. Performance of the Ni/YSZ and Cu/YSZ cermetanodes prepared by this method was comparable tothose of conventional Ni/YSZ cermet anode. The ad-vantage of this method is the high freedom of addingmetal or other additives such as CeO2 to the cermetstructure. The feasibility of this method in the devel-opment of Cu/ceria/YSZ cermet anodes for the directoxidation of methane has been demonstrated by Parket al. [83].

However, considerations in the selection of the syn-thesis of the Ni/YSZ cermet powders or simply to useconventional ceramic mixing processes based on com-mercial powders are the reproducibility, quality and thecost of the process and raw materials. The issues asso-ciated with the reproducibility are determined by thecomplexity of the synthesis processes while the cost ofthe process is largely determined by the scale-up andautomation.

2.2. Electrical conductivityIn the Ni/YSZ cermet materials, YSZ is an ionic con-ductor and Ni is an electronic conductor. In a sim-ple diphase system the theory predicts the percolationthreshold of ∼30 vol% of the higher conducting phasefor the transition from dominant ionic conductivity todominant electronic conductivity. This is generally truefor the electronic conductivity of the composite Ni/YSZcermet systems. Dees et al. [84] studied the relation-ship between the electrical conductivity and the vol-ume fraction of Ni in the Ni/YSZ cermet measured at1000◦C and showed that the rapid rise in the electri-cal conductivity corresponds to a ∼30 vol% of Ni inthe cermet. However, the position of the S-shape de-pendence of the conductivity on the Ni volume contentand the conductivity appear to be related to the particlesize of YSZ phase in the cermet, as shown in Fig. 9. Ingeneral, the electrical conductivity of Ni/YSZ materialsincreases with the YSZ particle size. The rather unusu-ally smooth transition instead of S-shape threshold typefor Ni/YSZ cermet prepared by self-propagating hightemperature synthesis (SHS) was attributed to the un-usual microstructure and good dispersion of the metal-lic component due to the very high temperature expe-rienced during the SHS process [90]. The much lowerthreshold of conductivity for Ni/YSZ cermets preparedby using Ni-coated graphite (NiGr/YSZ) was due to thelarge effective Ni content created by the thin Ni layersin the cermet [89]. This indicates that the electrical con-ductivity of Ni/YSZ is closely related to the microstruc-ture of the cermet. However, difference in percolationthreshold and conductivity may also be affected by theporosity as in conductivity curves shown in Fig. 9 levelof porosity has not been taken into account.

Electrical conductivity of Ni/YSZ cermets is stronglydependent on the particle size and distribution of bothNi and YSZ phases and thus the threshold of the volumefraction of Ni in the cermet for the change of the con-duction mechanism also changes. Huebner et al. [91]showed that with Ni average particle size of ∼0.6 µm,the electrical conductivity reached ∼300 S cm−1 at

Figure 9 Conductivity of Ni/YSZ cermet as a function of Ni content.The conductivity was measured at 1000◦C in humidified H2. Numbersare the references cited. The conductivity value for NiGr/YSZ cermetsas reported by Corbin and Qiao [89] was measured at 800◦C.

4412

1000◦C for 30 vol% Ni and when particle size of Niincreased to ∼16 µm, the threshold volume fractionof Ni to reach electrical conductivity of ∼100 S cm−1

at 1000◦C increased to 50 vol%. This indicates thatthe threshold for the transition from the dominant ionicconductivity of the YSZ electrolyte phase to the dom-inant electric conductivity of the Ni metal phase is re-lated to the YSZ/NiO size ratio. The electrical conduc-tivity of the Ni/YSZ cermet anodes is also dependenton the particle size and particle size distribution of NiOpowders. Tietz et al. [45] studied the electrical conduc-tivity of the Ni/YSZ cermet anodes prepared from dif-ferent commercial NiO powders. The electrical conduc-tivity measured at 800◦C varies between 300 S cm−1 to4000 S cm−1 with the highest conductivity obtained onNiO powder with small grain size and shrinkage rateof 27%. Tintinelli et al. [87] showed that the electri-cal conductivity of Ni/8YSZ cermet anodes increaseswith increase of the 8YSZ/NiO particle size ratio. Itohet al. [92] studied the relationship between the electricalconductivity and the particle size ratio of coarse YSZ(D50 = 10–100 µm), NiO (D50 = 1–100 µm) and fineYSZ (D50 = 0.4–60 µm). To achieve a good electricalconductivity, weight ratio (NiO/(coarse YSZ + NiO +fine YSZ)) of 8.8% or below is recommended.

For Ni/YSZ cermets to work properly as anodes,certain porosity of the electrode coating is essentialfor the transportation of fuel gas reactants to the elec-trode/electrolyte interface region where the fuel oxida-tion reaction occurs. The porosity is affected by vol-ume fraction of Ni in the cermet as there is a volumereduction of 25% when NiO is reduced to Ni metalunder fuel environment. Fig. 10 shows the change ofthe morphology of Ni anode before and after the NiO

Figure 10 Change of the morphology of Ni anode before and after ex-posure in humidified 10%H2/90%N2 for 1 h at 1000◦C.

reduction in humidified H2. An increase in the poros-ity is clearly seen. In general, the porosity increaseswith the increase in the Ni volume fraction and forNi/8YSZ, a typical porosity of 35% was found for Ni(40 vol%)/8YSZ (60 vol%) and 42% when Ni con-tent increased to 70 vol% [86]. The porosity and poremorphology can also be controlled by adding pore-formers such as carbon fiber, graphite and corn starch.Addition of 10–20 wt% graphite pore-formers wouldtypically increase the porosity by ∼10% [49]. Theporosity has strong effect on the electrical conductivitythrough its impact on the Ni-to-Ni connectivity. There-fore the electrical conductivity of the porous Ni/YSZcermet coatings is expected to be considerably smallerthan that of the Ni/YSZ cermet material. On a porousNi (50 vol%)/YSZ (50 vol%) cermet coatings screen-printed on a 50 × 50 mm YSZ, the conductivity is ∼254S cm−1 at 800◦C in 97%H2/3%H2O, which is muchsmaller than that of the dense Ni/YSZ cermet speci-men with similar volume fraction of nickel [93]. Ad-dition of pore formers with geometric anisotropy suchas graphite and carbon fiber can result in the formationof anisotropic porous structure of the Ni/YSZ cermets[94, 95]. The electrical conductivity of Ni/YSZ withanisotropic porous structure is in the range of 150 to600 S cm−1 as compared to ∼1600 S cm−1 for the cer-met with isotropic porous structure measured at 800◦C[94]. This shows that porous structure also has signifi-cant effect on the electrical conductivity of the Ni/YSZcermet anodes in addition to the porosity.

Conductivity of the Ni/YSZ cermet anode coat-ings in fuel cells is also affected by the heat treat-ment and sintering temperatures of the anodes. For Ni(50 vol%)/YSZ (50 vol%) cermet anodes prepared fromthe cermet powder coarsened at 1400◦C, the electri-cal conductivity of the anode was only ∼4 S cm−1 at1000◦C for the anode sintered at 1300◦C and increasedto ∼800 S cm−1 when the anode sintering temperatureincreased to 1500◦C [96]. The electrocatalytic activityof the anodes for the H2 oxidation reaction also in-creased substantially with the increase of the sinteringtemperature of the anodes. The effect of sintering tem-perature of Ni/YSZ cermets on the electrical conduc-tivity behavior was also studied by Pratihar et al. [97].Initial reduction temperature also affects the electricalconductivity as reported by Grahl-Madsen et al. [98].Electrical conductivity of Ni/8YSZ cermet substrateswas twice higher for the anode reduced at 1000◦C thanthat reduced at 800◦C. On the other hand, the electri-cal properties of two phase NiO-YSZ composites werestudied in the temperature range of 160–630◦C by Parket al. [99].

In comparison to electrical properties, relatively littleresearch has been carried out on the mechanical proper-ties of Ni/YSZ cermets. Selcuk et al. [100] determinedthe effective Young’s and shear moduli and Poisson’sratio of NiO/YSZ cermets as a function of porosity.For 75 mol% NiO-YSZ thick specimen (∼550 µm)sintered at 1350◦C with porosity of 31%, the fracturestrength of NiO/YSZ is 56 MPa, which is close to46 MPa of LSM but significantly smaller than 377 MPareported for dense YSZ [101, 102]. Primdahl et al. [58]

4413

qualitatively studied the effect of sintering temperatureof Ni/YSZ cermet anodes on the mechanical propertiesof the YSZ electrolyte cells. Sintering at temperaturesof 1300 to 1400◦C significantly reduced the mechan-ical strength of YSZ electrolyte cell. However, hightemperature sintering of Ni/zirconia cermet anodes isessential to reduce the electrode ohmic resistance andelectrode polarization resistance. This leaves little roomfor the compromise of the optimum electrode perfor-mance and cell mechanical properties. Sørensen et al.[103] showed that fracture of the YSZ electrolyte ismost likely initiated by cracks growing from the cer-met coating into the YSZ electrolyte. Nevertheless, theresidual stress in the NiO/YSZ anodes was much loweras compared to that in LSM cathode due to stress reliefby the extensive channel cracking [104]. This explainsthe observation that the applied load at failure and thestress in the YSZ electrolyte for symmetrical NiO/YSZand YSZ structure was almost equal to that of YSZplates [104].

2.3. Performance degradationStructural and long-term performance stability ofNi/YSZ cermet anodes are critical issues in the de-velopment of SOFC anode. Degradation problems inlong-term operation of SOFCs are characterized by agradual decrease in performance of the fuel cell system,measured by the percentage of increased overpotentialor decreased cell potential over certain period of op-eration. As far as the Ni/YSZ cermet anodes are con-cerned, the most predominant microstructure change isthe agglomeration and coarsening of Ni phase [105–107]. The loss of Ni through volatile nickel hydrox-ide species could also contribute to the performancedegradation [108]. The main reason for the agglomer-ation of Ni in the Ni/YSZ system is probably due tothe poor wettability between the metallic Ni and YSZoxide phase. Nikolopoulos et al. [109, 110] studied thewettability and interface reaction between Ni and YSZsystem in the temperature range of 1250–1500◦C andpurified Ar/4%H2 atmosphere. Molten Ni showed nowettability towards YSZ ceramic phase (θ = 117◦).Additives such as Ti, Cr, Mn and Pd have certain effecton the wettability between Ni and YSZ [111]. However,the effect on the interfacial energy of the Ni/YSZ systemappears to be small. Due to low melting temperature,pure Ni has a high tendency to sinter at SOFC operationtemperatures (e.g., 1000◦C). The sintering of pure Nianodes in reducing environment is very rapid, leadingto the formation of isolated islands [112]. Therefore, theprevention or reduction of agglomeration and sinteringof metallic Ni phase in the Ni/YSZ cermet electrodesrely heavily on the microstructure optimization of theNi/YSZ cermets.

Simwonis et al. [106] studied the structure change ofNi (40 vol%)/8YSZ (60 vol%) cermet anode substratesexposed to humidified Ar/4%H2/3%H2O atmosphere at1000◦C up to 4000 h. The distribution of Ni, 8YSZ andpores of polished cross-section of the substrates wereevaluated by quantitative image analysis. After anneal-ing for 4000 h, the average Ni particle size increased

Figure 11 Pore size distribution of Ni (70 vol%)/3YSZ (30 vol%) andNi (50 vol%)/3YSZ (50 vol%) cermet anodes after sintered at 1000◦C inhumidified 10%H2/90%N2 for 1 and 2000 h. The pore size distributionwas measured by image analysis.

from 2 to 2.57 µm and the number of Ni grain countsdecreased from 3421 to 2151. As the result, the elec-trical conductivity of the anode substrates decreasedby 33% due to decreased electronic Ni-to-Ni contacts.There was no change in 8YSZ particle size distribu-tion as expected. The sintering behavior of Ni/3YSZcermet anodes was also studied as a function of cer-met composition [112]. Fig. 11 shows the change ofpore size distribution of Ni (70 vol%)/3YSZ (30 vol%)and Ni (50 vol%)/3YSZ (50 vol%) cermet anodes aftersintering at 1000◦C in wet H2 for 1 and 2000 h. Thepore size distribution was measured by image analysis.Fig. 12 shows the corresponding SEM picture of Ni,Ni (70 vol%)/3YSZ (30 vol%) and Ni (50 vol%)/3YSZ(50 vol%) cermet anodes after sintering at 1000◦C inwet H2 for 2000 h. Pure Ni anode showed rapid sinteringat 1000◦C, resulting in the formation of large isolatedNi islands (Fig. 12a). In the case of Ni (70 vol%)/3YSZ(30 vol%) cermet anode, there was a clear increase inlarge pores and pore size distribution shifted to coarserpores after sintered at 1000◦C for 2000 h (Fig. 11a).Ni particles grew out the surface, indicating the sig-nificant agglomeration and grain growth of Ni in thecermet (Fig. 12b). As Ni content decreased to 50 vol%,the change in the pore size distribution was very small

4414

Figure 12 SEM pictures of Ni, Ni (70 vol%)/3YSZ (30 vol%) and Ni (50 vol%)/3YSZ (50 vol%) cermet anodes after sintered at 1000◦C in humidified10%H2/90%N2 for 2000 h.

and this is also indicated by negligible change in themorphology of the cermet coating (Figs 11b and 12c).This shows that microstructural stability of Ni/YSZ cer-met anodes is critically dependent on the Ni content andNi and YSZ phase distribution in the cermet. The uni-form distribution and homogenization of NiO and YSZ

phase in the preparation of NiO/YSZ cermets througheffective milling and deagglomeration are also effectivein improving the stability of the anodes [113].

Operating conditions of fuel cells also have signif-icant effect on the microstructure of the cermet an-odes. Iwata [105] studied the microstructural change

4415

of Ni/8YSZ cermet electrode coatings operating at0.3 A cm−2 and 1000◦C. Ni particle size increased from0.1–1 to 1–10 µm after 1015 h of operation and thespecific area of the cermets was decreased by half. Theperformance degradation rate was 14 mV over 1000 h.This was considered to be caused by the decrease ofthe contact area at the electrode/electrolyte interfaceand a decrease of the specific surface area of Ni par-ticles. Muller [114] showed that high current and highfuel utilization can cause the agglomeration of origi-nally fine dispersed Ni in the Ni/YSZ cermet anodes.The agglomeration of Ni phase in the cermet led to thereduction in the three phase boundary and the increasein the polarization resistance of the anode [115].

While in principle it is evident that Ni particles willgrow during high temperature operation, a detailed de-scription of the kinetics of the agglomeration and sin-tering of the Ni/YSZ cermets is difficult due to the com-plexity of the structure. Two types of materials (Ni andYSZ) plus pores exist in the structure with all of themhaving multi-modal size distributions. Ioselevich et al.[116] and Abel et al. [117] used correlated percolationmodel to describe the degradation of SOFC anodes. Intheir model, monosized particles are located on a face-centered cubic lattice and sintering between two metalparticles occur instantaneously with a certain probabil-ity. The result is an unchanged particle and one pore orone dead particle with no more electrochemical contri-bution to the network. The model gives an insight in thedevelopment of active bonds in the anode, the densityof three phase boundaries and the transport resistancein anodic materials. A simple two-particles model wasused by Vaßen et al. [118] to describe the degrada-tion of Ni/YSZ cermet anodes. The model is based onthe assumption that the difference in particle size isthe driving force for the sintering of the large metalparticles and sintering mechanism is dominated by thesurface diffusion of metal atoms. From the model, thesurface diffusion coefficient of Ni on Ni surface was1.3 × 10−12 m2 s−1 at 1000◦C and a reasonable corre-lation of the grain growth of Ni particles was observedbetween the model and experimental data.

When a system breakdown happened, there would betwo scenarios for the anode side of the cell: either thereducing atmosphere is preserved with nitrogen (im-plying nitrogen supply) or the cells stand a redox cycle.Thus, redox stability of anode materials is very impor-tant as the large volume change associated with NiO re-duction to Ni may cause the disintegration of the cermetstructure. The mechanical strength of Ni/8YSZ sub-strates before and after the reduction in hydrogen wasreported to be similar, about 18–20 MPa [98]. However,repeated reduction and oxidation at 1000◦C weaken themechanical strength of Ni/8YSZ membranes [119], in-dicating the dwindling of bonding between Ni-to-Niand Ni-to-YSZ in the cermets. Grahl-Madsen et al. [98]showed that the redox cycle may have sped up the dete-rioration of the electrical conductivity of the Ni/8YSZcermet substrates. This is supported by the significantincrease of the electrode ohmic resistance of anode-supported cell after exposing the anode to mild oxida-tive conditions (normal grade N2 flow) [33]. Addition

of CeGdO2 to the Ni cermet anodes was reported toimprove the redox stability of the cell [120].

Antonucci et al. [121] studied the stability and per-formance of Pt/CeO2 anode and Pt/PrO2 cathode of atubular SOFC stack. Significant chemical and morpho-logical changes occurred in both anode and cathodeafter 3000 h operation at 950◦C. There was a signif-icant depletion of CeO2 and formation of Pt/Ce alloyon the anode side and a strong depletion of PrO2 onthe cathode side, leading to the substantial change inthe reaction process and the degradation in the stackperformance.

2.4. Microstructure optimizationThe performance of Ni/YSZ cermet anodes depends onthe optimum phase distribution between Ni and YSZparticles that ensurs the maximum electronic and ionicpaths. Both modeling and experimental approaches arereported in the literature.

2.4.1. Modeling approachModeling approach has been adopted by many re-searchers for optimizing the performance of electrodes.It is known that the microstructure of the electrode playsa critical role in determining the performance of theelectrodes, and thus the performance of the cell as awhole. An optimized microstructure of cermet elec-trode, for example, should have long-range connectiv-ity of respective ionic and electronic conductor chainsstretching across the electrode and linking the currentcollector to the electrolyte; and maximized active sitesof three-phase boundary (TPB) for electrochemical re-actions. These requirements warrant an optimal ratio ofionic to electronic conductors, in terms of their volumeand average grain size, in a properly sintered electrode.In addition to the required microstructure for enhancedconversion of ionic to electronic current taking place atthe TPB and transport of these currents along their re-spective “active bond” (chains), there is other importantconsideration such as efficient gas diffusion from thebulk into the reaction sites. The latter warrants prop-erly design microstructure in terms of pore size andits distribution, porosity and the diffusion path. In sum-mary, a good electrode must consider the complex inter-relationship among the microstructure, electrochemicalprocesses and gas transport phenomena. As one can seethe complexity in designing an optimal electrode, mi-cro modeling is a useful approach for researchers toachieve the goal, in particular to guide the design of ahigh-performance electrode.

Literature review showed that the electrode micromodel of SOFCs could roughly be divided into poremodel [122], random resistor network model [123–125], and random packing sphere model [126, 127].In the pore model, the ionic conductor in the electrodeis viewed as protruding from the dense electrolyte sur-face with electronic conducting particles spread overthe surface in a connected network and the remainingspace is filled by continuous pore structure for the trans-port of gases. In the random resistor network model,

4416

Figure 13 Influence of particle size and anode thickness to the polar-ization for H2 oxidation (hydrogen humidified at 50◦C, i.e., H2:H2O =87.6:12.4) [128].

the electrode is assumed to consist of grains of theelectronic conductors and the ionic conductors packedtogether so as to form a continuous network. In the ran-dom packing sphere model, the electrode is assumed tobe a random packing of spheres. The theory of the par-ticle coordination number was applied, together withthe percolation theory. However, in this model [126,127], the complex gas transport phenomena in the elec-trode were ignored, despite its importance. Althoughconcentration polarization does not play an importantrole in most cases, especially for a thin anode, it doessomehow affect the fuel cell performance significantlyif crucial parameters related to microstructure of theanode are not properly chosen. Chan and Xia [128] de-veloped a micro model for the anode that formed by amixture of electronic and ionic conductors. The modelconsiders the complex inter-relationship among the mi-crostructure, electrochemical processes and gas trans-port phenomena in the anode. Fig. 13 shows the effectof particle size and anode thickness on the polarization

Figure 14 Electrochemical impedance curves measured at 1000◦C and open circuit in wet H2. Points are experimental data and lines are fitted data.(a) fine cermet anode, (b) coarse cermet anode, (c) Ni paste anode and (d) Ni felt anode after Brown et al. [131]. Symbols are experimental resultsand lines are fitted results by equivalent circuit.

potential for H2 oxidation on wet H2 [128]. The pre-dicted results show that for fixed particle radii of 0.1,0.2 and 0.5 µm, the polarization reaches the minimumat anode thickness of roughly 50, 100 and 160 µm, re-spectively. Loselevich and Kornyshev [129] developeda phenomenological model of cermet anode based ona simple macro- and micro-kinetics of charge and gastransport in the electrode. The model considered thetransport of hydrogen molecules and oxygen anions tothe reaction sites, hydrogen oxidation reaction and thewater-product removal.

Apart from above modeling approaches, Chan et al.[130] has developed a complete polarization model forSOFC and using it to study the sensitivity of perfor-mance to the change of cell component thickness. Thestudy revealed that, for the same materials used, sameelectrode kinetics and under same operating conditions,the performance of a cathode-supported cell would beno way better than an anode-supported cell even underelevated feedstock pressure conditions. The study alsoshowed that reducing the thickness of all three princi-pal components of the fuel cell (anode, electrolyte andcathode) could improve the operating current densityrange significantly.

2.4.2. Experimental approachAs shown by Primdahl and Mogensen [66], experimen-tally the relationship between the electrochemical per-formance and microstructure of Ni/YSZ cermet anodesis rather complex. However, a general trend does existbetween the microstructure, the electrochemical behav-ior and performance of the anode. Fig. 14 shows theelectrochemical impedance arcs for four different an-odes measured at 1000◦C in 97%H2/3%H2O [131]. Theparticle size of Ni and YSZ of fine and coarse cermet an-odes was in the range of 0.5–3 µm and 3–5 µm, respec-tively (Fig. 14a and b). Ni particles in Ni paste anode

4417

were ∼10 µm (Fig. 14c) while Ni felt anode consistedof 50 µm diameter fibers (Fig. 14d). Ni/YSZ cermetanode with fine Ni and YSZ particles and thus fine mi-crostructure had the smallest overall electrode polariza-tion resistance (RE = 0.19 �cm2 at 1000◦C, Fig. 14a)and electrode polarization resistance increased as themicrostructure became coarser. For Ni felt electrode,RE increased to 27.8 �cm2 at 1000◦C (Fig. 14d). Therelationship between microstructure and electrochem-ical performance of Ni based anodes was also studiedby others [132, 133]. The electrode impedance behav-ior showed strong dependence on the characteristics ofthe anode microstructure.

The microstructure of Ni/YSZ cermet anodes inturn depends on the development and optimizationof the fabrication processes as shown above. On theother hand, due to high temperature processing, certainchanges in the microstructure of Ni/YSZ cermet an-odes are inevitable. Fig. 15 shows typical microstruc-ture changes of Ni (50 vol%)/3YSZ (50 vol%) cermetanodes in various fabrication stages [134]. In the greenstate (i.e., the stage of anode coating before sintering),distribution of NiO and 3YSZ particles was very uni-form and the particle size was extremely small, in therange of less than 1 µm (Fig. 15a). Pores were alsovery small. After sintered at 1400◦C in air for 2 h, therewas clearly significant grain growth for both NiO and3YSZ particles and increase of porosity (Fig. 15b). Incomparison, the grain growth of NiO was substantiallyhigher than that of 3YSZ particles. Before the use asfuel electrodes, NiO in the Ni/3YSZ cermet anode mustbe reduced to Ni under fuel reducing environment. Thevolume reduction of 25% in the reduction of NiO toNi metal leads to further increase in porosity and moreimportantly there was further grain growth of Ni at oper-ating temperature of 1000◦C due to the poor wetabilitybetween Ni and zirconia (Fig. 15c). Ni particles weretypically in the range of 2 to 5 µm and there was a sig-nificant variation in the distribution between fine 3YSZparticles and large Ni particles. In some cases, Ni par-ticles were almost standing alone with little coverageof zirocnia particles. This clearly demonstrates that thesintering and reducing stages at high temperatures playan important role in the formation of the microstructureof Ni/zirconia cermet electrodes.

Ion impregnation was reported to be effective inimproving the microstructure of Ni/YSZ cermet an-odes during the high temperature sintering and reduc-ing processes [134]. In this method, metal ions usu-ally in nitrate solution such as Y0.03Zr0.97(NO3)x orSm0.20Ce0.80(NO3)x was impregnated into semi-fired oras-fired Ni/YSZ cermet anode coatings, followed by de-composition at high temperatures (i.e., fuel cell operat-ing temperatures). As a result, nano-sized Y0.03Zr0.97O2and Sm0.2Ce0.8O2 particles would deposit on the sur-face of Ni and YSZ grains. The formation of fine oxideparticles inhibits the sintering, grain growth and ag-glomeration of zirconia and in particular Ni phases dur-ing sintering and reducing stages at high temperatures.Fig. 16 shows the SEM pictures of Ni/8YSZ cermet an-odes before and after the ion impregnation treatment ofSm0.2Ce0.8(NO3)x nitrate solution after fuel cell testing,

Figure 15 Microstructure evolution of Ni (50 vol%)/3YSZ (50 vol%)cermet anodes in various fabrication stages of (a) green state, (b) aftersintered at 1400◦C, and (c) after reduction at 1000◦C in wet H2.

and Fig. 17 shows the corresponding electrode perfor-mance measured in 97%H2/3%H2O at 800◦C [134].After impregnation treatment, Ni particle size was sub-stantially reduced and majority of Ni particles werebelow ∼2 µm. Furthermore, Ni and YSZ particleswere covered by very fine oxide particles. As shownin Fig. 17, polarization losses of impregnated Ni/8YSZcermet anodes were significantly reduced as comparedto those without impregnation. For H2 oxidation onNi/8YSZ cermet anode without impregnation, η was234 mV at 250 mA cm−2 and 800◦C. After impregnated

4418

Figure 16 SEM pictures of Ni (50 vol%)/8YSZ (50 vol%) cermet anodes with and without impregnation after fuel cell testing. (a) none, (b)impregnation with 3 M Y0.03Zr0.97(NO3)x , (c) impregnation with 3 M Sm0.2Ce0.8(NO3)x and (d) impregnation three times with 3 M Sm0.2Ce0.8(NO3)x

nitrate solution.

Figure 17 Impedance and polarization performance of Ni (50vol%)/8YSZ (50 vol%) cermet anodes with and without impregnation.The performance was measured in 97%H2/3%H2O at 800◦C.

with 3 M Y0.03Zr0.97(NO3)x and 3 M Sm0.2Ce0.8(NO3)x

nitrate solution, η was 175 and 145 mV, respectively,at 250 mA cm−2 and 800◦C. After three consecutiveimpregnation with 3 M Sm0.2Ce0.8(NO3)x nitrate solu-tion, η was reduced substantially to 56 mV. The resultsdemonstrated that impregnation treatment is effectivenot only in the inhibition of grain growth and agglom-eration of Ni phases in the cermet but also in the pro-moting of electrochemical reactions. Ion impregnationmethod is equally effective in the enhancement of theelectrode performance of LSM cathodes [135].

Similar to the effect of ion impregnation, Yoon et al.[136] observed a significant improvement in the elec-trode performance of Ni/YSZ cermet anode after thedeposition of a thin film of YSZ or samaria-doped ce-ria (SDC) using a sol-gel coating technique. In thismethod, as-fired cermet coating was completely dippedinto YSZ or SDC sols, followed by calcinations at600◦C for 12 h. The process can be repeated to in-crease the loading or the thickness of the coating.The process was also used for the improvement ofthe electrode performance and stability of anode sup-ported electrolyte cells [137]. A process combiningslurry coating and pressure infiltration was used byChou et al. [138] to prepare Ni/YSZ cermet anodesover YSZ tubes. Impregnating standard Ni/YSZ cer-met anodes with metal catalysts such as Ru improvedthe power density from 0.35 to 0.48 W cm−2 at 800◦C[50].

4419

Using electrochemically active interlayer to en-large the reaction sites for the reaction at the elec-trode/electrolyte interface has proved to be effective inthe promotion of the electrocatalytic activity of both an-odes and cathodes. Mixed ionic and electronic conduct-ing materials such as doped CeO2 are often used as theinterlayer [139]. The interlayer layer can also be usedto prevent the interfacial reaction between the electrodeand the electrolyte in some cases. Combining an inter-layer of SDC and impregnated Ru catalysts, the currentdensity achieved at −0.9 V on modified SDC anodestructure was more than 4 times higher than that withoutthe interlayer and Ru additives [140]. Using a 0.5 µmthick Y-doped ceria (YDC) interlayer between anodeand electrolyte, the electrode polarization resistance forNi/YSZ anode for methane oxidation was ∼1.2 �cm2

at 600◦C, which is much smaller than 6.6 �cm2 forthe Ni/YSZ cermet anode without the interlayer underthe same conditions [41]. Using double anode struc-ture with electrochemically active, composite anode orNi/YSZ cermet anode with different Ni content and mi-crostructure as anode function layer, polarization per-formance and power density increased as compared tosingle anode layer structure [8]. The double layered an-ode showed good long-term stability [85]. Multilayeranode concept with gradients in composition and mi-crostructure to function as electrochemical active layer,gas diffusion layer and current collector top layer havealso been proposed [141]. Muller et al. [142] showedthat such multilayer anode structure improved the po-larization performance and long term stability.

Limitations of the fabrication processes based onthe conventional ceramic powder mixing process (seeFig. 1) are the difficulties in achieving optimum phasedistribution between Ni and YSZ particles. Severaltechniques have been developed to achieve better dis-tribution between Ni and YSZ for the preparation ofcermet anodes for solid oxide fuel cells. The most no-table one was based on electrochemical vapor deposi-tion (EVD) developed by Westinghouse [143]. In West-inghouse tubular cells, the anode is fabricated by slurrycoating of nickel and YSZ is incorporated into the Niskeleton structure by EVD method. The morphologyof YSZ intergrowth restricts grain growth and sinter-ing of Ni particles and provides an effective pathwayfor oxygen ions from YSZ electrolyte to Ni phase [12].Other deposition techniques reported include sputter-ing [144], chemical vapor deposition (CVD) [145], acombined EVD/CVD process [145], polarized elec-trochemical vapor deposition (PEVD) [146, 147], etc.Each of these methods offers advantages in certain as-pects but each also presents its own difficulties in termsof cost, reproducibility, scale-up and microstructurecontrol. The objective in the development and optimiza-tion of a preparation process of Ni/YSZ cermet anodesis on controlling and tailoring the electrode microstruc-ture to certain degree and thus producing an anode notonly with high performance and stability but also withhigh reproducibility and low cost. The fabrication pro-cesses based on cheap and conventional ceramic pow-der mixing method in conjunction with microstructureimprovement processes such as ion impregnation may

Figure 18 Partial list of the electrode polarization resistance (RE) ofNi/YSZ cermet anodes measured under moist hydrogen at different tem-peratures. RE was measured by electrochemical impedance spectroscopyat open circuit potential. Numbers are the references cited.

offer an effective way to produce high performance andhigh stable Ni/YSZ cermet anodes.

Fig. 18 is the partial list of the electrode polarizationresistance (RE) of Ni/YSZ cermet anodes measured un-der moist hydrogen at different temperatures. RE wasmeasured under open circuit potential by electrochem-ical impedance spectroscopy and was obtained by thedifference of low and high frequency intercepts on theimpedance plots. Most of the results reported in the lit-erature are in the medium to high temperature range(800 to 1000◦C) and have wide variations in the REvalue. By extrapolating the results to low temperatureregion, the electrode polarization resistance would bein the range of 1–30 �cm2 at 700◦C and 7–200 �cm2

at 600◦C for Ni/YSZ based anodes. The electrode po-larization resistance is still too high for H2 oxidationreaction at temepratures below 700◦C. However, withmodification of Ni/YSZ cermet anodes and with inco-poration of YDC interlayer, electrode polarization re-sistance as low as 0.26 �cm2 at 600◦C in wet H2 wasreported [41].

3. Modified Ni/YSZ cermet anodesConventional Ni/YSZ cermet materials with and with-out catalyst modification have been extensively studiedfor the internal reforming of methane. The internal re-forming activity of the anodes is important as it convertsmethane into more electrochemically active species ofH2 and CO. Unlike polymer electrolyte fuel cells, CO isalso a fuel for SOFC. However, the electrochemical ac-tivities of Ni/YSZ cermet anodes for the CO oxidationreaction are much lower as compared to that of the an-ode for the H2 oxidation reaction [150, 151]. Lee et al.[152] studied internal reforming reaction on Ni/YSZcermet anodes with wide range of porosity and variousNi content. The reaction is of first order with respectto the methane and −1.25 for the steam. Ahmed et al.[153, 154] studied the kinetics of internal reformingreaction on Ni/YSZ cermet anodes. On conventionalNi/YSZ cermet anodes, a steam/carbon (S/C) ratio of2 or higher was required to avoid carbon deposition. At

4420

steam/carbon ratios of 1.5–3, the rate of reforming re-action is 0.84 order with respect to the partial pressureof CH4 and −0.35 order with respect to the steam onconventional Ni/YSZ cermet anodes at temperatures of∼900◦C. The activation energy of the reforming reac-tion was 95 kJ mol−1 on Ni/YSZ cermet anodes. Ac-tivation energy values of 58 to 82 kJ mol−1 were alsoreported for the steam reforming reaction on Ni/YSZcermet anodes [155, 156]. High activation energy of118–294 kJ mol−1 was reported for the reforming re-action on Ni/YSZ cermet anodes with addition of cata-lyst such as Ce [157, 158]. Addition of small quantitiesof molybdenum (∼1%) was reported to significantlyreduce the carbon deposted on Ni/YSZ anode duringmechane reforming [159]. Drescher et al. [160] stud-ied the reforming activities and structural propertiesof Ni/8YSZ cermet substrates using temperature pro-grammed reduction (TPR), gas adsorption, diffusionand permeation measurements. The methane conver-sion rate increased linearly with the increase of anodethickness up to 0.6 mm. The gas adsorption experimentsshowed that the active Ni surface area is about 20% ofthe total surface area of the cermet. TPR result indicatesthat the Ni in the cermet would be completely reducedat a temperature of about 500◦C at sufficient hydrogensupply. This temperature range is close to those reportedby Perego et al. [161] and Shirakawa et al. [162].

The effect of carbon deposition on the electrochem-ical activity of Ni/YSZ cermet anodes in the methanefuel appears to be related to the operating conditionssuch as current load. Koh et al. [33] showed that at opencircuit or low current density, the effect of carbon depo-sition on the electrocatalytic activity of Ni/YSZ cermetanodes was irreversible. Conversely, under high currentload, the polarization performance of the Ni/YSZ wasreversible. This indicates the possibility of minimiza-tion of the carbon deposition on Ni/YSZ cermet anodes.Weber et al. [163] demonstrated the stability of Ni/YSZcermet anode up to 1000 h under a current density of0.4 A cm−2 at 950◦C in dry methane without seriousdegradation in performance.

Under high current and low concentration (4–9%) ofdry methane, methane was reported to be completelyoxidized to CO2 and H2O without carbon deposition onNi/YSZ cermet anodes [164, 165]. Under fuel cell oper-ation conditions where methane concentration was highand current was lower than the threshold value, carbondeposition occurs at steam to carbon (S/C) ratio lowerthan 1 [166]. Liu and Barnett [167] studied the perfor-mance and stability of Ni/YSZ anode-supported SOFCoperated with wet methane and natural gas at 600–800◦C. Under a current density of ∼0.6 A cm−2, thecells showed good stability with limited carbon depos-tion. Takeguchi et al. [168] studied the carbon deposi-tion and steam reforming of methane on Ni/YSZ anodesmodified with alkaline earth addition of MgO, CaO,SrO and CeO2. CaO, SrO and CeO2 addition suppressedthe carbon deposition while MgO addition promotedthe carbon deposition rate and decreased the steam re-forming activity of the anodes. High content (∼2 wt%)of SrO and CeO2 in the Ni/YSZ cermets also reducedthe steam reforming activity significantly. Further study

[169, 170] shows that Ru and Pt addition promotes thereforming activity and supress the carbon deposition.The steam reforming reaction and carbon deposition arealso reported by Onuma et al. [171] on Pt electrodes andHorita et al. [172] on Ni/YSZ and Fe/YSZ anodes. Satoet al. [173] studied Ni-Co alloy/YSZ cermet anodes forthe oxidation of H2 and CH4. Ni-Co alloy was preparedby co-precipitation, followed by sintering at 1100◦C for24 h. Co content in Ni-Co/YSZ anodes has little effecton the cell performance with H2 fed but increases thecell performance with CH4 fed. The best compositionwas found to be Ni0.5Co0.5/YSZ. The increased activ-ity of Ni-Co/YSZ cermet anodes for CH4 oxidation wasexplained by the preferential crystal orientation along(111) plane at the alloy surface, which was not observedfor pure Ni powder. The exposure of the most denselypacked plane in the face-centered cubic structure couldcontribute to the formation of surface active sites forthe electrochemical oxidation of methane. Ni-Co/YSZcermet anodes achieved lowest electrode polarizationresistance as compared to that of Ni-Cu/YSZ and Ni-Fe/YSZ cermet anodes for H2 oxidation reaction in wetH2 [174]. Kim et al. [175] studied the carbon deposi-tion and CH4 oxidation on Ni-Cu alloy/YSZ cermetanode prepared by ion impregnation of porous YSZlayer with copper and nickel nitrite solution. Carbonwas found to deposit on Ni-Cu alloy. However, for Ni-Cu alloy/YSZ cermet anodes with low Ni content suchas Ni0.2Cu0.8/YSZ, small amount of carbon deposition,in fact, could promote the cell performance, similar tothat of pure Cu/YSZ cermet anodes [176]. On the otherhand, it was reported that there was no carbon deposi-tion on Ni0.52Cu0.48/GDC cermet anodes prepared fromvery fine Ni0.52Cu0.48 alloy after operating in dry CH4for ∼40 h at 800◦C [177]. Carbon deposition was alsofound on Ni0.8Mg0.2/YSZ cermet anodes after exposureto wet methane [178]. Electrode polarization resistancefor the reaction in wet methane on Ni0.8Mg0.2/YSZ an-ode was 9.2 �cm2 at 1000◦C, which is much higherthan 1.7 �cm2 in wet H2.

Presence of sulfur (primarily in the form of H2S) inthe fuel gas can also affect the performance of Ni/YSZcermet anodes. Geyer et al. [179] reported that theelectrode polarization resistance increased by a factorof two by additing only 5 ppm of H2S at 950◦C in97%H2/3%H2O. Matsuzaki and Yasuda [180] showedthat the poisoning effect of sulfur-containing fuel gason the electrode performance depends on the total sul-fur content and the temperature. The tolerance of theanode towards sulfur poisoning increased with the in-creasing operation temperature. Study showed that fora sulfur content of 0.02 to 15 ppm, the performance losscaused by the sulfur poisoning could be recovered afterswitching to the sulfur-free hydrogen fuel. The effectof sulfur impurities on the electrochemical activities ofNi/YSZ cermet anodes was also studied by others [181,182].

4. Other metal/oxide cermet anodesInstead of modifying Ni/YSZ cermet anodes, in-vestigations were also performed on the use of

4421

electronic or mixed conducting oxides to make alter-native metal/oxide cermets such as ceria stabilized zir-conia (CeO2-ZrO2), calcium-doped ceria (Ca-CeO2),yttria-doped ceria (YDC), praseodymium oxide (PrOx ).Mixed results were reported. Kawada et al. [151] stud-ied Ni cermets prepared from CeO2-ZrO2, Ca-CeO2,and Y-CeO2 and showed no significant improvement inthe electrode performance as compared to Ni/YSZ an-odes. Relatively high overpotential and high resistancewere also observed for ceria related Ni cermet anodes.Eguchi et al. [183], on the other hand, reported the sig-nificant improvement in the polarization performanceof Ni cermets using samaria-doped ceria (SDC), CeOx

and PrOx (Ni/oxide ratio was 8/2 in weight) as com-pared to Ni/YSZ anodes. The activity of the Ni/oxidescermet for hydrogen oxidation was found to be in thefollowing order

Ni/YSZ < Ni/CeOx < Ni/SDC < Ni/PrOx

The improved performance was attributed to the in-crease of the effective reaction area due to the ionicconduction of the oxides. Wang and Tasumi [184] in-dicated that addition of SDC to Ni anodes led to theincrease in activation energy and resulted in high over-potential for H2 oxidation reactions on Ni/SDC anodesat low temperatures.

Joerger and Gauckler [185] studied the Ni/Gd-dopedCeO2 (Ni/GDC) cermet for H2 and CH4 oxidation re-action. Ni/GDC cermet anodes were sintered at 1350◦Csimilar to that of Ni/YSZ cermet anodes. The electricalconductivity of Ni/GDC anodes was 1.7 × 104 S cm−1

and electrocatalytic activity of Ni/GDC cermet wasmuch higher than that of Ni/YSZ cermet anodes. At800◦C and 0.5 A cm−2, η of Ni/GDC anode was∼20 mV for H2 oxidation reaction, significantly lowerthan ∼120 mV obtained on Ni/YSZ cermet anodes.When using methane, η of Ni/GDC anode was ∼75 mVat 800◦C and 0.25 A cm−2. Activation energies forthe H2 and CH4 oxidation reaction was also lower onNi/GDC anodes than those on Ni/YSZ anodes, exhibit-ing the catalytic and electrocatalytic activities of ce-ria for the reforming and oxidation of hydrocarbonsand hydrogen. Lower overpotential was also reportedon Ni/Sm-doped ceria (Ni/SDC) anode as compared toNi/YSZ anodes for H2 oxidation at 850◦C [186] and thebest performance was found on Ni/SDC with 50 vol%Ni content [187]. On Ni/GDC and Ni/15 mol% Y2O3-Ce0.7Zr0.3O2 anodes operated below 750◦C, no carbondeposition was found when dimethyl ether (DME) wasused as the fuel [188]. On the other hand, carbon de-position was found on Ni/GDC anodes with 70 wt%Ni after operating on wet methane at 850◦C for about80 h [189]. With 5 wt% addition of Ru to Ni/GDC an-odes, the electrode polarization resistance to CH4 fuelwas reported to be 0.13 �cm2 and a power density of0.75 W cm−2 was recorded in dry CH4 at 600◦C [190].Hibino et al. [191] also investigated Pd-doped Ni/SDCcermet anodes for partial oxidation of CH4. With asmall addition of Pd (7 wt% PdO), a maximum powerdensity of 0.644 W cm−2 at 550◦C was achieved on asingle-chamber SOFC in CH4/air mixture, using SDC

electrolyte and Sm0.5Sr0.5CoO3 cathode. On the otherhand, in a conventional SOFC configuration, a powerdensity of 0.192 W cm−2 was reported on Ni/SDC inwet CH4, using similar electrolyte and cathode materi-als [192].

The chemical compatibility between YSZ and ce-ria has been studied by Tsoga et al. [193, 194]. Gd-doped ceria (GDC) and YSZ react and diffuse into eachother during sintering process at 1200◦C, forming Gdrich phase with ionic conductivity two orders lowerthan that of YSZ at 800◦C [194]. Further study [195]shows that the solid state reaction between YSZ andGDC can be effectively suppressed with an interlayerof Ce0.43Zr0.43Gd0.10Y0.04O2. Chemical compatibilityis important for the long-term stability of the multi-layered structure where doped-ceria is often used asinterlayer on the cathode side [196] and anode side[41] to protect and to enhance the cell performance. ANi (40 vol%)/GDC (60 vol%) composition is recom-mended for the suppression of the reactivity betweenGDC in the Ni/GDC cermet anode and YSZ electrolyteas the presence of Ni can inhibit the reactivity betweenYSZ and ceria [194].

Zhang et al. [70] and Ohara et al. [197] studied theeffect of sintering temperature on the electrode perfor-mance of Ni/SDC anodes on La0.9Sr0.1Ga0.8Mg0.2O3(LSGM) electrolytes. The best electrode performance(overpotential of 30 mV at 300 mA cm−2 and 800◦C)was obtained on the anode sintered at 1250–1300◦Cwith Ni content of ∼60 vol%. Sintering at 1350◦C sig-nificantly increased the electrode ohmic resistance andpolarization overpotentials, which have been attributedby the significant sintering of the Ni/SDC electrodesand the Ni diffusion into the LSGM electrolyte phase.Kuroda et al. [198] studied the electrode performanceof Ni/SDC anodes on La0.8Sr0.2Ga0.8Mg0.15Co0.05O3(LSGMC) at 650◦C and found that the cell polarizationwere dominated by the loss at the anode. High sensi-tivity of the sintering temperature of the performanceof Ni/SDC cermet anodes on LSGM electrolyte wasalso observed by Huang et al. [199]. It has been shownthat the formation of resistive phase of LaSrGa3O7,LaSrGaO4 or LaNiO3 between LSGM, NiO and SDCis the main cause for the sensitivity of sintering tem-perature of the cermet anodes [199–202]. A Ni-freeinterlayer such as doped-CeO2 between LSGM andNi-based anodes could prevent the interfacial reaction[199]. Power density close to 0.9 W cm−2 was reportedfor a 600-µm thick LSGM electrolyte cell at 800◦C,using Ni/La2O3-doped ceria as the anode [199].

Ni/Ti-doped YSZ cermets were also considered asthe anodes for high temperature SOFC [203] due to theattraction of mixed conductivities of TiO2-doped YSZunder low partial pressure of oxygen [204, 205]. Elec-trical conductivity of YSZ decreased with Ti dopingin YSZ but the bending strength of T-doped YSZ in-creased. The overpotential of 60 mV at 300 mA cm−2

and 1000◦C appeares to be compatible with those re-ported for Ni/YSZ cermet anodes.

Gorte and co-workers [42, 82, 83, 175, 206–208]focused their efforts on the development of Cu/CeO2/YSZ based anodes. Unlike Ni, copper is inert to

4422

hydrogen or hydrocarbon oxidation and has no catalyticactivity for the formation of C C bonds, thus suppress-ing the carbon deposition [209]. The primary functionof copper in the cermet anodes is to provide electronicconduction path. The melting point of Cu is 1083◦C,lower than 1453◦C for Ni. Thus Cu-based anodes aresuitable for operating temperatures <800◦C. However,pure Cu/YSZ anode has very low electrochemical ac-tivity for both H2 and CH4 oxidation reactions. Ce-ria catalyst was added to Cu/YSZ cermet to improvethe electrode activity [83, 210]. The cells were oper-ated with dry hydrocarbons including methane, butaneand synthetic diesel without any carbon deposition. Apower output of ∼0.15 W cm−2 at 700◦C was observedin pure C4H10 for a cell with a Cu/CeO2/YSZ anodeand a 60 µm thick YSZ electrolyte. Electrode perfor-mance of Cu/CeO2/YSZ anodes was significantly im-proved after exposed to hydrocarbon fuels in particularfor the anode with Cu content lower than 30 vol%.This was explained by the effect of the hydrocarbondeposits on the improvement of conductivity of the an-ode, as shown schematically in Fig. 19 [211]. The effectof lanthanide additives on the Cu/YSZ cermet anodeswas also studied, but the performance of Cu/YSZ an-odes with lanthanide additives was not as good as thatof Cu/ceria/YSZ cermet anodes [212]. In the case ofCu/YSZ based anodes, the oxide catalyst such as ceria

Figure 19 Schematic diagram showing the changes in the three phaseboundary and connectivity between metallic particles in Cu/YSZ cermetanode (a) before and (b) after operation in hydrocarbon fuels such asn-butane after McIntosh et al. [211].

should be in direct contact with YSZ electrolyte for it tobe effective [212]. Because Cu2O and CuO melt at 1235and 1326◦C, respectively, it is not feasible to prepareCu/YSZ cermets by high temperature sintering as thoseused for Ni/YSZ cermet anodes. An alternative methodwas developed for the preparation of Cu/YSZ cermetanodes in which a porous YSZ was produced by acidleaching Ni from Ni/YSZ cermets, followed by addi-tion of Cu and catalysts such as ceria by impregnation[175, 213].

Ruthenium has high melting point (2310◦C) as com-pared to Ni (1453◦), thus providing better resistanceto sintering and grain growth. Ru/YSZ cermet anodesshow high catalytic activity for steam reforming andnegligible carbon deposition under internal reformingconditions [214]. Vernoux et al. [215] also reached thesame conclusion. But its use is prohibited by the highcost and evaporation above 1200◦C to produce volatileRuO4 species.

5. Conducting oxidesMixed ionic and electronic conducting oxides such asgadolinium- or samarium-doped ceria have been inves-tigated as potential anode materials for intermediatetemperature SOFC. The major attraction of such mixedionic and electronic conduction oxides is the possibilityof enhanced reaction zone over the three phase bound-ary due to the high electronic conductivity as comparedto conventional YSZ. For example, the oxygen ion con-ductivity of Sm-doped CeO2 is ∼0.1 S cm−1 at 800◦Cis about four times higher than that of YSZ while itselectronic conductivity is ∼4 S cm−1 under H2 reduc-ing environment [216, 217]. The ceria-based oxides,titanate based oxides, and lanthanum chromite basedoxides will be discussed briefly below.

5.1. Ceria-based oxidesDoped-ceria has been extensively investigated as al-ternative electrolytes to replace YSZ for ITSOFC [218,219]. The major problem with ceria as electrolyte mate-rial is the reduction of Ce4+ to Ce3+ under the fuel richconditions of the anode side of electrolyte, introduc-ing electronic conductivity and lattice expansion [218].The introduced electronic conductivity under fuel re-ducing environment has been explored for the anodeapplication as it has been shown that CeO2 anode canelectrochemically oxidize dry methane as the presenceof mobile lattice oxygen reduces the rate of carbon de-position [220, 221]. The high electrocatalytic activityof ceria-based materials for methane oxidation is mostlikely due to the extraordinary ability of ceria to store,release and transport oxygen ions [222].

The conductivity and activity properties of CeO2-based oxides at 1000◦C have been discussed by Mo-gensen [223] with respect to the application as SOFCanodes. Above 900◦C, where grain boundary resistanceplays a minor role, the ionic conductivity is mainly de-termined by the temperature, dopant concentration anddopant metal ion radius. Under reducing conditionslike in the anode side, conductivity and dimensional

4423

stability of CeO2-based oxides also relate to the com-position and the reducing conditions. Relatively largelattice expansion associated with the loss of oxygen un-der anodic conditions can result in the anode spilling offthe electrolyte and 40% Gd doped CeO2 is consideredto be a reasonable compromise between the conductiv-ity and dimensional stability [224]. Marina et al. [225]studied Ce0.6Gd0.4O1.8 (GDC) as the anode for both H2and CH4 oxidation reactions at 1000◦C. To improvethe adhesion between GDC and YSZ electrolyte, anintermediate layer of coarse grain YSZ was sinteredon YSZ electrolyte to provide an anchoring effect onthe GDC anode. At 1000◦C and fuel to steam ratio of0.13, the cell produced power output of 470 mW cm−2

for H2 and only 80 mW cm−2 for CH4. This indicatesthat pure GDC has poor electrocatalytic activity formethane oxidation. High electrode polarization resis-tance of ∼11 �cm2 at 800◦C was reported for the H2oxidation reaction on pure GDC anode [226].

Uchida et al. [196, 227, 228] studied the electrocat-alytic activity of mixed conducting oxides of Sm-dopedCeO2 (SDC) and Y-doped CeO2 (YDC) for the H2 oxi-dation with and without addition of Ru. SDC and YDCelectrodes were sintered at 1150 and 1250◦C, respec-tively. To reduce the electrode ohmic resistance, a goldmesh was use as the current collector. At η = 0.1 Vand 800◦C, a performance of 0.2–0.25 A cm−2 was ob-tained on SDC and YDC anodes and the current in-creased to 0.4–0.5 A/cm2 with the addition of fine Ruparticles. Putna et al. [229] also observed enhancingeffect of Ru addition on the electrocatalytic activity ofceria anodes. Highly-dispersed Ni in SDC was foundto be more effective in promoting the electrocatalyticactivity for H2 oxidation reaction than that of mixedNi/SDC cermet anodes [230]. Impregnation of 0.75mg cm−2 (or 8 vol%) of nano-sized Ni particles intoSDC anodes increased current density to 0.8 A cm−2

at overpotential of 0.1 V and 800◦C, which is muchhigher than that of Ni/SDC cermet anodes. This wasexplained by the significant increase in the number ofreactive sites at the nano-sized Ni and SDC boundaries.Significant improvement in the electrocatalytic activityfor H2 oxidation reaction on mixed conductor anodessuch as GDC, La0.75Sr0.25Cr0.97V0.03O3 and Ti-dopedYSZ with addition of small amounts of Ni (1–2.5 wt%)was also reported by Primdahl et al. [231, 232]. Onthe other hand, a thin layer of yttria-doped ceria (YDC)did not show any significant difference in terms of elec-trocatalytic activity of Pt electrode for H2 oxidation ascompared to that of pure YSZ electrolyte [233]. Thisappears to be very different from that of Ni/YSZ cer-met anodes. An addition of a thin YDC layer betweenNi/YSZ cermet anodes and YSZ electrolyte reduced theelectrode interface resistance for methane oxidation bya factor of ∼6, as shown in Fig. 20 [41]. This showsthat the addition of YDC layer significantly increasedthe reaction sites for the CH4 oxidation reaction.

5.2. Titanate-based oxidesTitanate-based oxides are relatively stable in fuel re-ducing environment and possess certain conductivity

Figure 20 Impedance curves for Ni/YSZ cermet anode (a) without and(b) with a 0.5 µm thick YDC interlayer between anode and YSZ elec-trolyte. Impedance curves were meaured in 97%CH4/3%H2O at 600◦Cafter Murray et al. [41].

at low partial pressure of oxygen, making them attrac-tive as potential anode materials. Titanates are electron-ically conductive and show n-type at low pO2 (10−15–10−2 Pa) and p-type at high pO2 (100–105 Pa). Sutijaet al. [234] studied the electrical conductivity of Fe-doped calcium titanate and reported that the conductiv-ity was less than 1 S cm−1 for CaFex Ti1−x O3−δ (x = 0.1to 0.4) even at 1000◦C. Holt et al. [235] investigated thesynthesis and electrical properties of Ti-doped NdCrO3and found that the solubility of Ti in Nd(Cr1−x Tix )O3is about 20%. The electrical conductivity of Ti-dopedNdCrO3 was about 1 S cm−1 at 1000◦C in low pO2

range. However, the initial electrical conductivity oftitanates is somehow related to the sintering condi-tions. Marina et al. [236, 237] studied the electricaland thermal properties of lanthanum strontium titanates(LST) and Ce-doped LST. For air sintered sample, theelectrical conductivity is in the range of 1–16 S cm−1

and for sample sintered in hydrogen, electrical conduc-tivity of 80–360 S cm−1 at 1000◦C and partial pres-sure of 10−18 atm. However, electrical conductivity ofhydrogen-sintered sample degraded in air and appearsto be irreversible. The cell based on La0.4Sr0.6TiO3 an-odes showed very rapid performance degradation, i.e.,1/3 of its initial performance after just 1 h of testing at1000◦C in 97%H2/3%H2O. Ce-doped LST may in factconsist of a ceria/perovskite assemblage [237].

Electrical properties of titania-doped YSZ were alsoinvestigated as potential candidates for alternative an-ode materials due to the introduced electronic con-ductivity, their thermal stability and thermal expansioncompatibility with YSZ electrolyte [238]. The solubil-ity of titania in YSZ is quite high. But for sampleswith 15 mol% titania or higher dopant level, Colomeret al. [239] reported the appearance of the second phase(e.g., ZrTiO4). They showed that titania additions inexcess of 5 mol% are an effective way to promotethe electronic conductivity (reduction in ionic trans-port number) of YSZ under anode working conditions.Total conductivity of 5 mol% titania-doped YSZ is∼0.02 S cm−1 at 800◦C and decreased with increas-ing titania content [239, 240]. The reason could be dueto the solubility limit effect on the ion vacancy mo-bility. Yttria-stabilised zirconia-terbia (YSZT) systemalso showed mixed conductivity under fuel environ-ment. For example, for YSZT containing 30 mol% Tb,

4424

the electric and ionic conductivity were measured as2.3 × 10−3 S/cm and 2.8 × 10−3 S/cm, respectively,at 800◦C in air. The p-type conductivity of YSZT sys-tem is only significant in oxidizing atmospheres [241].Tao and Irvine [242] investigated Y2O3-ZrO2-TiO2and Sc2O3-Y2O3-ZrO2-TiO2 systems. In the Sc-Y-Zr-Ti system, Sc0.15Y0.05Zr0.62Ti0.18O1.9 showed mostpromising as potential anode materials with ionic andelectronic conductivity of 7.8 × 10−3 and 0.14 S cm−1,respectively, at 900◦C. The cubic fluorite structure ofYSZ can be maintained even with Ti contents as highas 18% [243]. Cu-stabilized ZrO2 was also explored asanode materials for SOFC [244]. The catalytic activityof CuO-ZrO2 samples increased with CuO content upto 20 mol%.

Middleton et al. [245] investigated some titanate-based oxides as the potential anodes for methane con-version and oxidation. The oxides under study in-clude rutile-type Ti0.97Nb0.03O2, pseudobrookite-typeMg0.3Nb0.1Ti2.6O5 and MgTi1.95Nb0.05O5, pyrochlore-type Sm2Ti1.9Nb0.1O7 and V3O5-structure typeCrTi2O5. The oxides showed some activities towardsthe CH4 conversion in the temperature range of 500–700◦C and the main product was CO2. At temper-ature higher than 800◦C, CO was dominated. Fromsteady state catalytic studies, it showed that amongthe titanates most active catalysts are the compoundswith enhanced electronic conductivity, for examplepyrochlore-type Sm2Ti1.9Nb0.1O7 is more active thanSm2Ti2O7 and pseudobrookite-type Mg0.3Nb0.1Ti2.6O5is more active than MgTi1.95Nb0.05O5. All oxides understudy showed low C2 formation except Ti0.97Nb0.03O2.The highest activity in the n-type conducting titanateswas observed for Mg0.3Nb0.1Ti2.6O5 generating CO2as the main product. No data was reported for the CH4conversion under SOFC operating conditions.

Fagg et al. [246] investigated the possibility of usingspinel solid solution series Mg2−yTi1+yO4, 0 < y <

0.5 as high temperature anodes for SOFC. Reducedmagnesium titanate spinels are stable in the reducingatmosphere and are shown to resist oxidation in air upto 500◦C. The conductivity very much depends on theoxidation state of Ti at room temperature, but at fuelcell operation temperatures (e.g., 1000◦C), there is lit-tle difference in conductivity and the conductivity valuewas estimated to be in the range of 0.2–2.0 S cm−1.The thermal expansion coefficient of magnesium ti-tanate spinels obtained from the high temperature pow-der diffraction studies was close to those of the 8 mol%Y2O3-doped ZrO2 (∼10.5 × 10−6/K).

Perovskite SrTiO3 doped with La or Nb and nio-bium based tetragonal tungsten bronzes were evalu-ated as potential anodes for SOFC [247, 248]. Dopedstrontium titanates are reported to be stable in re-ducing environment and the electronic conductivitywas ∼7 S cm−1 at 930◦C but its ionic conductivitywas low [247]. For niobium based tetragonal tungstenbronzes under study, the highest conductivity of ∼8.3S cm−1at 10−20 bar was observed on Ba0.35Ca0.15NbO3.Ba0.6−x Ax Ti0.2Nb0.8O3 (A = Sr, Ca) was consideredas a promising anode material as it can be preparedin air and is stable in reducing environment. Layered

perovskite, La2Sr4Ti6O19−δ , was found to have quitehigh conductivity, 40 S cm−1 at 900◦C in 5%H2/Arbut its electrode polarization resistances of 3 �cm2

and 9 �cm2 at 900◦C in wet H2 and wet CH4, re-spectively, are still high [249]. Yashiro et al. [250]also studied the electrode performance of doped stron-tium titanates such as SrTi0.97Nb0.03O3, La0.1Sr0.9TiO3and La0.2Sr0.8TiO3. The observed electrode polariza-tion resistance was in the range of 350 to 700 �cm2 at800◦C. This indicates that doped-SrTiO3 has poor elec-trocatalytic activity for H2 oxidation reactions despitethe high electronic conductivity of doped strontium ti-tanates (e.g., electronic conductivity of La0.1Sr0.9TiO3was reported to be as high as ∼200 S cm−1 at 800◦C[251]).

Yttrium-doped SrTiO3 (YST) showed quite highelectrical conductivity under reducing conditions [252,253]. At partial pressure of 10−19 atm and 800◦C,the electrical conductivity of dense YST specimen(Sr0.85Y0.10Ti0.95Co0.05O3−δ) is 45 S cm−1 and forYST specimen with 30% porosity it is ∼30 S cm−1.The highest conductivity of 64 S cm−1 at 800◦C wasobserved on Sr0.88Y0.08TiO3−δ . The thermal expan-sion coefficient of YST is in the range of 11–12 ×10−6/◦C, which is compatible with that of YSZ and(La,Sr)(Ga,Mg)O3 electrolyte materials. However, ad-hesion of YST to YSZ electrolyte was poor and similarto that of ceria based anodes [224], SYT anode coatingsdelaminated from YSZ electrolyte.

5.3. Lanthanum chromite and othertype of oxides

Lanthanum chromites (LaCrO3) based materials havebeen investigated as interconnector material in SOFCbecause of its stability towards reduction and its rela-tively high electrical conductivity and stability in bothreducing and oxidizing atmospheres at SOFC operatingtemperatures [254–257]. The conductivity propertiesof LaCrO3-based materials are significantly affectedby A- and/or B-site doping. For example, Ca dopingat the A-sites has significant effect on the electronicconductivity due to the expected charge compensatingtransitions (Cr3+ to Cr4+). Co substitution for Cr alsohas some positive effect on the electronic conductiv-ity. Co doping causes dramatic increase in TEC, butthis can be compensated by the subsequent Ca doping.Conductivity of (La1−x Cax )CrO3 materials is indepen-dent on the oxygen activity in the high pO2 region anddecreased with 1/4 slope at low pO2 range. The conduc-tivity of La0.8Ca0.2Cr0.8Co0.2O3 was 45 S cm−1 at highpO2 and reduced to about 4 S cm−1 at pO2 of 10−16

atm and 1000◦C [254–257]. Also, it has been shownthat LaCrO3-based materials such as (La,Sr)CrO3 havevery low activity towards carbon deposition [36, 258].

Catalytic activity of LaCrO3 for methane oxida-tion can also be substantially enhanced by dopingat A- and B-sites. Sfeir et al. [259–261] studied thethermodynamic stability and the catalytic activity of(LaA)(CrB)O3 system (A = Ca, Sr and B = Mg, Mn,Fe, Co, Ni) as alternative anode materials under simu-lated SOFC operating conditions. Thermodynamically,

4425

Sr and Mn substitution would maintain the stability ofthe perovskite while other substitutes would destabi-lize the system. However, transitional metal substitutedLaCrO3 did not decompose easily under reducing con-ditions, indicating that the decomposition of substitutedLaCrO3 would be hindered kinetically. Ca and Sr sub-stitution at the A-site improves the catalytic activity andfor B-site substitution, Co and Mg showed an inhibit-ing effect while Mn, Fe and Ni improved the activity.La0.75Sr0.25Cr0.5Fe0.5 was shown stable under reduc-ing atmosphere but its electrochemical performance inwet methane was reported to be very poor [262]. Con-sidering the catalytic activity and cracking resistance,(LaSr)(CrNi)O3 with 10% Ni doping was suggestedto be most suitable for methane oxidation reactions[263]. On the other hand, Ni exsolution has been foundin 10% Ni-doped lanthanum chromites in fuel envi-ronment [264] and the surface segregation of Ni fromNi-doped lanthanum chromites may be responsible forthe high activity. Liu et al. [265] showed that additionof 5% NiO in La0.8Sr0.2Cr0.8Mn0.2O3 (LSCM)/GDC(50 wt%/50 wt%) composite anodes significantly en-hanced the anode activity for both H2 and CH4 oxida-tion. The power density of the cell with LSCM/GDC-NiO anode was 0.13 W cm−2 at 750◦C, much higherthan 0.05 W cm−2 on the cell with LSCM/GDC an-ode without NiO addition. Enhancing effect of addit-ing small amount of catalysts such as Ni and Ru on theperformance of LaCrO3-based for fuel oxidation reac-tions was also studied by others [37, 266]. The amountof metal catalysts may be low enough to avoid carbondeposition [265].

Middleton et al. [245] studied the activities ofLa0.8Ca0.2CrO3 for the CH4 conversion and oxidationreactions. The oxide showed no significant formationof C2 compounds, but was the most active and selectivecatalyst for CO2 formation, with complete conversionbelow 500◦C, as compared with titanate-based oxides[245]. At temperature above 800◦C, the conversion ofmethane dramatically increased, with the concomitantformation of CO and hydrogen. The methane reactionwas suggested to occur in two separate ways, i.e., com-

Figure 21 Performance of the cell with (La0.75Sr0.25)0.9Cr0.5Mn0.5O3 anode and LSM cathode measured at different temperatures in humidified fuelgases after Tao and Irvine [274]. YSZ electrolyte thickness was 0.3 mm.

plete oxidation to CO2 and H2O by partial methanedissociation (mainly involving surface adsorbed oxy-gen), and methane dissociation (mainly involving mo-bile lattice oxygen) to form H and C species onthe oxide surface. Methane dissociation was predom-inant at high temperatures (900–1000◦C) with carbondeposition.

Primdahl et al. [266] studied the electrocatalyticactivity of (La,Sr)CrO3 for H2 and CH4 oxidationat 850◦C. In the case of La0.8Sr0.2Cr0.97V0.03O3, theelectrode polarization resistance (RE) was ∼5 �cm2

for H2 and ∼30 �cm2 in CH4 with 3%H2O at850◦C. The presence of second phase formation inLa0.8Sr0.2Cr0.97V0.03O3 led to the formation of SrZrO3.This indicates that Sr-doped LaCrO3 has poor electro-catalytic activity for H2 oxidation and shows little cat-alytic activity for direct CH4 oxidation. This is similarto the previous reports on the overall low electrode per-formance of steam reforming and direct methane ox-idation reaction on LaCrO3 and doped LaCrO3 [246,267–269]. Chromite/titanate based perovskite anodesgenerally show much poorer electrochemical activityfor H2 oxidation as compared to that of Ni/YSZ cermetanodes despite some of the perovskites have reason-ably high electrical conductivity. This may be partiallydue to the poor adhesion between the perovskite coat-ing and YSZ electrolyte [270]. Insertion of Ru or Mninto the B-site of LaCrO3 was shown to increase theelectrochemical activity for the H2 oxidation as com-pared to Sr-doped LaCrO3 [271]. The catalytic activityof Ru and V-doped (LaSr)CrO3 for steam reforming ofmethane was studied in detail as function of stream tocarbon ratio at 850◦C [272]. It was pointed out that theSr doping is the key factor in the stability of rutheniumin the chromite perovskite structure [273]. Recently Taoand Irvine [274] reported some promising results ofLa0.75Sr0.25Cr0.5Mn0.5O3 (LSCM) as anode for SOFC.LSCM is a p-type conductor with conductivity of∼38 S cm−1 at 900◦C at oxygen partial pressureabove 10−10 atm. The electrode polarization resistancefor the oxidation reactions in wet CH4 and H2 at900◦C was 0.85 and 0.26 �cm2, respectively. Fig. 21

4426

shows the performance of a cell with (La0.75Sr0.25)0.9-Cr0.5Mn0.5O3 anode and LSM cathode measured at dif-ferent temperatures. The maximum power density was0.47 W cm−2 at 900◦C, which is considered to be com-patible to the cell based on Ni/YSZ cermet anodes, con-sidering that the YSZ electrolyte used in the cell had athickness of 0.3 mm.

Mixed ionic and electronic conducting ox-ides such as SrFeCo3Ox , SrCo0.8Fe0.2O3 andLa0.6Sr0.4Fe0.8Co0.2O3 (LSCF) were investigated fortheir electrochemical activity for the oxidation ofH2 and CH4 fuels [275, 276]. Different to their highelectrocatalytic activity for the O2 reduction reactions[277], the oxide anodes showed poor electrochemicalactivity for the oxidation reactions of both H2 and CH4.It is well known that LSCF materials without protectivelayer decompose readily and are not stable in exposingto reducing atmosphere [278]. Other conducting oxidesstudied as potential anode materials include cubicperovskite Sr2GaNbO6 [279], tetragonal tungstenbronze type A0.6Bx Nb1−x O3 (A = Ba, Sr, Ca, La andB = Ni, Mg, Mn, Fe, Cr, In, Sn) [280], nonstoichio-metric mixed perovskites such as SrCu0.4Nb0.6O2.9,SrMn0.5Nb0.5O3−δ or Sr2Mn0.8Nb1.2O6 [281, 282],and pyrochlore Gd2(Ti1−x Mox )2O7 [283]. Table I sum-marizes the electrical and electrochemical propertiesof some conducting oxide studied as potential anodematerials. Large effort is still needed to understandthe conductivity and electrochemical behaviour ofconducting oxides under fuel environment and toimprove significantly the electrocatalytic activityand stability for the oxidation reaction of H2 andhydrocarbon fuels such as CH4 under SOFC operatingconditions.

6. Development of anode-supportedstructures

It is increasingly clear that for the development of com-mercially viable SOFC technologies, a reduction ofthe operating temperature from traditionally 1000◦Cto the range of 600 to 800◦C is essential. Low operat-ing temperature not only widens the selection of ma-terials for the electrode, electrolyte, interconnect andmanifold components of SOFC but also significantlyreduces the risk of thermal shock and interface reac-tivity between various cell components and thus in-creases the cell performance stability and reliability.However, lower operating temperature requires highelectrode electrocatalytic activity and high ionic con-ductivity of electrolyte to compensate the cell loss with-out penalty for the power density achieved at high tem-peratures. There are two different ways to achieve thisobjective: one is the development of electrolyte withmuch higher ionic conductivity than YSZ and the sec-ond is to reduce the thickness of YSZ electrolyte usingelectrode supported structures. Electrode supported cellstructure is commonly used to reduce the electrolytethickness.

The choice of materials for the electrode supportedstructure is primarily related to a number of importantconsiderations;

• No reactivity with the electrolyte during fabri-cation process. This is particularly important ifthe cell structure is formed by co-sintering. Asmost of electrolyte materials require high sin-tering temperature (1300 to 1500◦C), the reac-tivity and interface reaction between the elec-trode support and electrolyte should be none orminimum.

• High strength of the electrode support structure.Cell strength is determined by the strength of theelectrode support structure. Sufficient cell strengthis needed for easy handling during the cell fab-rication and stacking. Also, cell should be strongenough to withstand the thermal and mechanicalstress during operation.

• High permeability for the fuel gas and the reactionproducts H2O and CO2. As the electrode supportstructure is usually in the thickness of 500 to 2000µm, polarization losses due to gas diffusion can besignificant.

• Good sintering compatibility. Sintering behaviorof the electrode support and electrolyte must beclosely matched to reduce the bending and warpingand to increase the production rates.

• High electrical conductivity to reduce the ohmicloss of the cell and high thermal conductivity torespond quickly to temperature swing in the celland stack.

• Good microstructure, surface and shrinkage prop-erties for easy deposition and densification of thinelectrolyte layers.

• Finally the cost of the material and automation ofthe processing needs to be considered.

Ni/YSZ based electrode materials satisfy most of theconsiderations for YSZ based electrolyte cells. Ni/YSZcermets are highly electrically and thermally conduc-tive. Ni/YSZ cermet anodes can be fired with YSZelectrolyte up to 1400◦C for the densification of theelectrolyte with little reactivity between Ni and YSZelectrolyte. With cathode-supported structure such asSr-doped LaMnO3, the forming temperature wouldhave to be 1300◦C or lower in order to avoid theformation of resistive La2Zr2O7 phase at the elec-trode/electrolyte interface [284–287]. Thus, sinteringYSZ thin film on LaMnO3-based oxide substrate is ex-pected to be more difficult than sintering on a Ni/YSZcermet anode substrate. Gas diffusion limitation is alsoa less problem on the anode side as compared to thaton the cathode side. At 800◦C the molecular diffusioncoefficient of H2/H2O system is ∼8.4 cm2 s−1, which isconsiderably higher than ∼2 cm2 s−1 of O2/N2 system.The effective binary diffusion coefficients of H2/H2Oare also considerably higher than those of O2/N2 underSOFC operating conditions [288]. Porous structure andporosity of the anode substrates can be controlled byanode composition and addition of pore formers such asrice and corn starch, carbon black and graphite [289].Primdahl et al. [290] reported that additing 40 vol%corn starch as pore-former to the anode substrates couldreduce the electrode polarization resistance from 0.15to 0.08 �cm2 at 850◦C. The cost of Ni/YSZ materials

4427

T ABL E I Conductivity and electrochemical properties of selected conducting oxides studied as potential anode materials. Electrode polarizationresistance (RE) was measured at open circuit by electrochemical impedance spectroscopy

σ /S cm−1 RE/�cm2

Composition Temp/Re Temp/Ox Temp/H2 Temp/CH4 Ref.

La0.7Ca0.3TiO3 2.7 (900◦C/Re)a [270]La0.4Sr0.6TiO3 60 (900◦C/Re) [270]La0.7Ca0.3Cr0.5Ti0.5O3 0.03 (900◦C/Re) [270]La0.7Ca0.3Cr0.5Ti0.2O3 32 (850◦C/H2)b [270]La0.1Sr0.9TiO3 3 (1000◦C/Re) 1 (1000◦C/Ox)a Sintered in air [236]La0.2Sr0.8TiO3 3 (1000◦C/Re) 1 (1000◦C/Ox) Sintered in air [236]La0.3Sr0.7TiO3 4 (1000◦C/Re) 1.3 (1000◦C/Ox) Sintered in air [236]La0.4Sr0.6TiO3 16 (1000◦C/Re) 0.004 (1000◦C/Ox) Sintered in air [236]La0.1Sr0.9TiO3 80 (1000◦C/Re) 0.004 (1000◦C/Ox) Sintered in H2 [236]La0.2Sr0.8TiO3 200 (1000◦C/Re) 0.03 (1000◦C/Ox) Sintered in H2 [236]La0.3Sr0.7TiO3 200 (1000◦C/Re) 0.01 (1000/Ox) Sintered in H2 [236]La0.4Sr0.6TiO3 360 (1000◦C/Re) 0.03 (1000/Ox) Sintered in H2 [236]La0.1Sr0.9TiO3 510 (800◦C/H2) [250]La0.2Sr0.8TiO3 350 (800◦C/H2) [250]SrTi0.97Nb0.03O3 700 (800◦C/H2) [250]Sr0.85Y0.15Ti0.95Ca0.05O3 37 (800◦C/Re) [252]Sr0.85Y0.15Ti0.95Co0.05O3 45 (800◦C/Re) [252]Sr0.85Y0.15Ti0.95Zr0.05O3 13 (800◦C/Re) [252]Sr0.85Y0.15Ti0.95Mg0.05O3 6 (800◦C/Re) [252]Sr0.88Y0.08TiO3 64 (800◦C/Re) [252]La0.7Ca0.32CrO3 86 (850◦C/H2) [266]La0.75Ca0.25Cr0.9Mg0.1O3 21 (850◦C/H2) [266]La0.8Sr0.2Cr0.97V0.03O3 5 (850◦C/H2) 30 (850◦C/CH4)b [266]La0.75Sr0.25Cr0.9Mg0.1O3 2 (850◦C/H2) [266]La2Sr4Ti6O19 ∼30 (900◦C/Re) 8.5 × 10−4 (900◦C/Ox) 3 (900◦C/H2) 9 (900◦C/CH4) [249]La0.8Sr0.2CrO3 256 (850◦C/H2) [271]La0.8Sr0.2Cr0.8Mn0.2O3 51 (850◦C/H2) [271]La0.75Sr0.25Cr0.5Mn0.5O3 1.3 (900◦C/Re) 38 (900◦C/Ox) 0.26 (900◦C/H2) 0.85 (900◦C/CH4) [274]Sr0.6Ti0.2Nb0.8O3 2.5 (930◦C/Re) 3 × 10−4 (930◦C/Ox) [248]Sr0.4Ba0.2Ti0.2Nb0.8O3 2.5 (930◦C/Re) 2 × 10−4 (930◦C/Ox) [248]Sr0.2Ba0.4Ti0.2Nb0.8O3 3.2 (930◦C/Re) 2 × 10−4 (930◦C/Ox) [248]Ba0.4Ca0.2Ti0.2Nb0.8O3 3.1 (930◦C/Re) 2 × 10−4 (930◦C/Ox) [248]Ba0.6Ti0.2Nb0.8O3 3.2 (930◦C/Re) 1 × 10−4 (930◦C/Ox) [248]Sr2GaNbO6 9 × 10−4 (900◦C/Re) 7.8 × 10−3 (900◦C/Ox) [279]Ba0.6Mn0.067Nb0.933O3 2.2 (930◦C/Re) 4 × 10−4 (930◦C/Ox) [280]Ba0.4La0.2Mn0.133Nb0.867O3 0.2 (930◦C/Re) 6 × 10−4 (930◦C/Ox) [280]Ba0.4Sr0.2Mn0.067Nb0.933O3 1.8 (930◦C/Re) 4 × 10−4 (930◦C/Ox) [280]Ba0.6Ni0.067Nb0.933O3 4.5 (930◦C/Re) 5 × 10−4 (930◦C/Ox) [280]Ba0.4La0.2Ni0.133Nb0.867O3 2.4 (930◦C/Re) 2 × 10−4 (930◦C/Ox) [280]Ba0.6Mg0.067Nb0.933O3 1.3 (930◦C/Re) 8 × 10−5 (930◦C/Ox) [280]Ba0.4La0.2Mn0.133Nb0.867O3 0.5 (930◦C/Re) 2 × 10−5 (930◦C/Ox) [280]Ba0.6Fe0.1Nb0.9O3 3.8 (930◦C/Re) 1 × 10−2 (930◦C/Ox) [280]Ba0.4La0.2Fe0.2Nb0.8O3 1.1 (930◦C/Re) 2 × 10−4 (930◦C/Ox) [280]Ba0.5La0.1Fe0.2Nb0.8O3 0.7 (930◦C/Re) 3 × 10−3 (930◦C/Ox) [280]Ba0.4Ca0.2Fe0.1Nb0.9O3 1.2 (930◦C/Re) 3 × 10−3 (930◦C/Ox) [280]Ba0.4Sr0.2Fe0.1Nb0.9O3 2.3 (930◦C/Re) 4 × 10−3 (930◦C/Ox) [280]Ba0.6In0.1Nb0.9O3 1.0 (930◦C/Re) 1 × 10−4 (930◦C/Ox) [280]Ba0.4Sr0.2In0.1Nb0.9O3 1.5 (930◦C/Re) 1 × 10−4 (930◦C/Ox) [280]Ba0.4La0.2In0.2Nb0.8O3 0.3 (930◦C/Re) 2 × 10−5 (930◦C/Ox) [280]Ba0.6Cr0.1Nb0.9O3 3.6 (930◦C/Re) 2 × 10−3 (930◦C/Ox) [280]Ba0.6Sn0.2Nb0.8O3 21 (930◦C/Re) 3 × 10−4 (930◦C/Ox) Pellet showed signs of decomposition [280]

after measurementSrCu0.4Nb0.6O2.9 Unstable 1 × 10−4 (900◦C/Ox) [281]Sr2Mn0.8Nb1.2O6 8 × 10−3 (900◦C/Re) 0.36 (900◦C/Ox) [281]SrMn0.5Nb0.5O3−δ 6 × 10−2 (900◦C/Re) 1.23 (900◦C/Ox) [282]Gd2 (Ti0.5Mo0.5)2O7 11 (900◦C/Re) [283]

aRe indicates reducing conditions in H2 (e.g., 5%H2/95%Ar or PO2 is ≤10−20) and Ox indicates oxidizing conditions in air.bH2 and CH4 usually indicates moist H2 and CH4 with 3% H2O.

can be reduced by using TiO2 or Al2O3 as cost effectivealternatives to YSZ in the anode supported structure[291].

The state-of-the-art anode-supported planar SOFCis based on Ni/YSZ cermet as the support [292–295].

In the last few years, Ni/doped-ceria cermets havebeen increasingly studied as the supports for the IT-SOFC based on high conductivity electrolyte materi-als such as doped-ceria [296–298]. Various fabrica-tion techniques have been explored and investigated

4428

in the fabrication of porous anode support substrates.The most common one is probably the tape castingtechnique.

Tape casting is an established industrial process tomake large and flat ceramic tapes, plates and lami-nated ceramic structure, and thus is suitable for makinganode-supported planar solid oxide fuel cell structure.Tape casting involves spreading of ceramic powdersand organic ingredients onto a flat surface where sol-vents are allowed to evaporate. After drying, the result-ing tape develops a leather-like consistency and can bestripped off from the casting surface. Typically, tapecast anode layer from 1 to 2 mm thick and tape castelectrolyte layer from 40 to 100 µm thick are producedand these two tapes are laminated and after rolling orcalendering, the green laminates are cut to size andsintered at 1400◦C in a specially designed kiln fur-nace to ensure the parts are all flat. The final anode-supported structures are produced with anode thicknessof 500–1000 µm and electrolyte thickness of 15–40µm. Fig. 22 shows typical structure of Ni/YSZ cer-met anode-supported thin YSZ electrolyte cell as com-pared to electrolyte supported cell structure. With tape-calendering process, the YSZ electrolyte thickness canbe reduced to 5–10 µm [299–301]. However, shrink-age of anode and YSZ electrolyte materials can be verydifferent [54, 302]. Mismatch in densification of suchbilayer structure will cause intolerable bending and/orcracking of the structure during co-firing [291, 303].Thus, sintering of the laminated anode and electrolytetapes is critical in producing flat anode-supported struc-ture as the sintering profiles of the anode and electrolytetapes are very different. Fig. 23 shows the shrinkageprofile of Ni/8YSZ cermet tapes and 8YSZ electrolytetape as function of sintering temperature. Both carboncontent and type of 8YSZ powder have significant ef-fect on the shrinkage process of Ni/8YSZ cermet tapesand 8YSZ electrolyte tapes. The bilayer structure based

Figure 22 Typical structure of (a) YSZ electrolyte-supported cell and (b) Ni/YSZ cermet anode-supported thin YSZ electrolyte cell prepared by tapecasting.

on Ni/YSZ substrates usually tends to warp towardsthe anode layer during sintering. Alumina is a sinteringinhibitor and also has a relatively low thermal expan-sion coefficient (TEC), which can bring the TEC ofthe anode substrate close to that of YSZ. Thus, addi-tion of alumina in the Ni/YSZ structure can signifi-cantly improve the flatness of the substrate [289]. Onthe other hand, Muller et al. [304] showed that by cofir-ing screenprinted Ni/YSZ cermet anode and YSZ elec-trolyte green tape, anode polarization losses were re-duced due to the improved contact between the YSZphase in the cermet and YSZ electrolyte. The mis-match between the anode and YSZ and thus the me-chanical stresses in the Ni/YSZ cermet anode layer andYSZ substrate can be reduced by changing the anodecoating from a continuous geometry to a large num-ber of small sized individual areas [305]. The best re-sults of the patterned anode layer were observed on thehoney-comb design with low curvature and stress. Toovercome the sintering profile mismatch, anode sub-strate tape can also be heated to produce a presin-tered body strong enough to be handled for electrolytedeposition.

Will et al. [306] gave a detailed review on the tech-niques in the deposition and preparation of thin anddense electrolyte layer on porous anode substrates.This includes vacuum slip casting [307], magnetronsputtering [308], spray and spray pyrolysis [309, 310],screen printing [297], electrophoretic deposition [311–313] and other thin film deposition techniques. Forsmall cells and laboratory investigation purpose, anode-supported structures were often produced by powdercompacting and die-pressing [297, 309, 314, 315].Dense electrolyte film as thin as 8 µm was prepared onporous substrates by dry-pressing process [316]. Yooet al. [317] used compacting and drying of Ni/YSZslurry to make 50 × 50 mm size anode substrates.Anode-supported cells with diameter of 120 mm were

4429

Figure 23 Shrinkage profile of Ni/8YSZ cermet tape with addition ofvarious carbon contents (empty symbols) and 8YSZ tape (filled symbols)as function of sintering temperature.

also sucessfully fabricated by die-pressing technique[318]. Song et al. [319] reported the fabrication andperformace of anode-supported tubular cell using ex-trusion technique. Power density in the range of 570to 800 mW cm−2 at 750◦C was reported for prototypeanode-supported cells based on tape casting techniquewith Ni/YSZ cermet anode and LSM cathode [293,307]. Power output as high as 2 W cm−2 at 1000◦Chas been reported on Ni/YSZ anode supported cells[320, 321].

Simwonis et al. [322] reported coat-mix (CM) pro-cess for the production of porous Ni/YSZ cermet sub-strates. The powder produced by CM process was char-acterized by tiny rigid agglomerates and the substrateswere prepared by pressing the powder in a die at 120◦C.The green plates can be prepared up to the dimensionsof 330 × 330 × 2.7 mm. After pre-sintering at 1200◦Cthe porous substrate was coated with a functional layerof Ni and 8YSZ via vacuum slip casting [282]. Elec-trolyte with thickness of 5–10 µm was deposited bythe same technique, followed by sintering at 1400◦C.The substrates prepared by CM process were shownto have higher gas permeability, better pore size dis-tribution and pore structure as compared to Ni/8YSZsubstrates prepared by tape casting [323]. However,automation of CM process may be of concern. Vari-ous pore-former can be added to the anode substratesto improve the porous structure and to improve the cellperformance.

Surface morphology of the substrate such as poresize and pore size distribution is critical for the depo-sition of thin and high dense electrolyte film on theNi/YSZ anode substrates as the morphology of the de-posited film follows that of the substrate surface. Fig. 24shows the morphology of the YSZ film deposited onYSZ and Ni/YSZ cermet substrate by magnetron sput-tering. Thin YSZ film deposited has the same grainand grain boundary pattern as that of the YSZ sub-strate (Fig. 24a), indicating that the surface morphol-ogy of the YSZ thin film follows closely the morphol-ogy of the substrate. The large pores on the substrate

surface resulted in open pores on the YSZ electrolytefilms deposited (Fig. 24b and c). Thus the pore di-ameter of the substrate surface needs to be in the or-der of the grain size of the deposited film to achievedense and uniform YSZ electrolyte film (Fig. 24d ande). This can be done by using an interlayer or func-tional layer with finer microstructure and low poros-ity [282]. The function layer is also used to promotethe electrode reaction at the electrode/electrolyte inter-face. Hobein et al. [324] shown that the surface arte-facts of the function layer should be smaller than thefilm thickness in order to avoid the defects on the YSZfilm deposited by pulse laser deposition technique. Theporous structure of the surface of the function layercan be changed by high temperature annealing [325].The porosity of both anode substrate and functionallayer decreased with increasing annealing temperatureand annealing at 1380–1400◦C created a dense func-tion layer surface for the deposition of gas-tight YSZelectrolyte film. Kek et al. [326] found that the ionicconductivity of deposited YSZ film on Ni/YSZ sup-port varied with the sintering temperature of the anodesubstrate, but no explaration was given. The structureand stability of the Ni/YSZ cermet substrates are im-portant to the deposition and properties of YSZ thinelectrolyte. Cassidy et al. [327] studied the effect ofthe volume change of NiO to Ni in the cermet sub-strates on the stability of YSZ electrolyte. The initialreduction seems to have a little effect on the stabilityof YSZ thin electrolyte but the reoxidation of Ni toNiO in the cermet support led to the crack of the YSZelectrolyte.

Preparation and physical properties of Ni/TiO2 cer-mets were investigated for the anode support for pla-nar SOFC [328]. A thin Ni/YSZ interlayer or func-tion layer was used for the electrochemical reactionat the anode/electrolyte interface. In addition to thecost saving as compared to Ni/YSZ substrates, Ni/TiO2substrates have thermal expansion coefficient closerto that of YSZ electrolyte and the electrical conduc-tivity is compatible with that of Ni/8YSZ substrates.Softening of the Ni/TiO2 cermet at temperatures above800◦C is also beneficial for the release of stresses in-troduced during stack manufacturing [329]. Yan et al.[330] studied the electrical and microstructural prop-erties of Ni/La0.9Sr0.1Ga0.8Mg0.2O3 (Ni/LSGM) anodesubstrates. The electrical conductivity increased withthe Ni content. However, the electrical conductivity ap-pears to be low. For example, the electrical conductiv-ity of Ni (60%)/LSGM (40%) cermet anode was ∼45S cm−1 at 800◦C. Among the compositions studied, thesubstrates with 60% Ni content were considered to havethe best microstructure for the deposition of LSGM thinelecrlyte. To avoid the interaction between LSGM andNi in the substrate, a SDC interlayer with a thickness of2–5 µm was used. Yan et al. [331] also proposed a novelmethod to prepare Ni/YSZ anode supported LSGMelectrolyte thin film cell. In this method, thin LSGMelectrolyte film was deposited onto the surface of aporous YSZ substrate, sintered and followed by impreg-nation of Ni into the porous YSZ substrate to turn it intoelectrically conducting Ni/YSZ cermet substrate. The

4430

Figure 24 Morphology of the YSZ film deposited on YSZ and Ni/YSZ cermet anode substrate by magnetron sputtering. (a) YSZ film depositedon YSZ substrate, (b) and (c) YSZ film deposited on Ni/YSZ cermet substrate, showing that the defects in the YSZ electrolyte film is due to thelarge pores on the anode substrate surface, and (d) and (e) dense and uniform YSZ film deposited on Ni/YSZ anode substrate with optimized surfacemorphology.

maximum power of 0.85 Wcm−2 was achieved at 800◦Cwith H2/air.

Another important property of anode supported elec-trolyte structure is the structure stability and perfor-mance recoverability during reduction-oxidation cy-cle (or RedOx cycle) either due to accidental systemshut-down or system service requirement. Due to theelectrochemical and mechanical requirements of theanode supported electrolyte structure, the cell shouldbe able to withstand a significant number (>5/year) ofRedOx cycles without significant degradation of theperformance [332]. To understand the RedOx behav-ior of Ni/YSZ cermet materials, Tikekar et al. [333]studied the kinetics of reduction and re-oxidation ofdense Ni/YSZ cermets at temperature range of 650–800◦C. The reduction process is interface controlled,while the re-oxidation kinetics is diffusion controlled.

There was a significant increase in porosity of the re-oxidized sample as compared to the as-sintered sample.The reduction process of NiO in NiO/YSZ cermets isalso not a simple one. Mori et al. [334] studied theinteraction between NiO and YSZ by TPR and identi-fied five different states of NiO species in the NiO/YSZsystem. In this respect, porous metallic alloy substratesmay be attractive alternatives as electrode-supportedstructure [335–339]. Different kinds of metallic sub-strates such as Ni and FeCr alloy were used includingfelt, foams, nets and porous sintered plates. Ni/YSZanode, YSZ electrolyte and LSM cathode componentswere deposited onto porous metallic substrate by a vac-uum plasma spray process [335–337] or by conven-tional ceramic processing techniques such as screen-printing and dip-coating [338, 339]. Cell performanceof 280 mW cm−2 at 0.7 V and 800◦C was achieved in

4431

H2/air with YSZ electrolyte and the power density in-creased to 430 mW cm−2 with the scandia-stabilizedzirconia electrolyte [340]. The advantages of metallicsubstrate-supported cells include low cost, high me-chanical strength and the ability in withstanding therapid thermal cycling.

7. General commentsNi/YSZ based cermet materials are still the mostcommon anodes in SOFC due to their high electro-chemical activity for H2 oxidation reaction and demon-strated long-term stability at SOFC operating condi-tions. Fabrication and microstructure of the Ni/YSZcermet materials have been extensively studied andsignificant understanding and knowledge of the com-plex inter-relations between the material properties,microstructure, electrochemical process and the fab-rication process parameters have been achieved. Theelectrochemical performance of Ni/YSZ cermet anodesfor H2 and CH4 oxidation reactions can be further im-proved by replacing YSZ with mixed ionic and elec-tronic conductors (MIEC) such as doped ceria or byimpregnation or doping of noble metal or MIEC ox-ide catalyst. Nevertheless, the long-term stability ofsuch modified cermet anodes has not yet been fullydemonstrated.

R&D activities in the development of modified Ni-based and/or Ni-free materials for oxidation (or con-version) reactions of hydrocarbon fuels (e.g., methane)in SOFC are increasing significantly since the lastdecade. However, there are significant discrepanciesin the results on the electrical properties and perfor-mance of the Ni-based or Ni-free anodes for oxida-tion reactions of methane under SOFC operation en-vironment, as highlighted recently by Mogensen andKammer [341]. Steele [40] suggested that anode ma-terials for the electrochemical oxidation of hydrocar-bon fuels such as CH4 should have good electronicconductivity of 100 S cm−1, high oxygen surface ex-change kinetics, high activity for oxidation of hydro-carbon fuels and stability in reducing environment.From the oxide materials reviewed, there is hardly amaterial reported so far that satisfies all the require-ments for anodes for direct electrochemical oxidationof hydrocarbon fuels. Low electronic conductivity ofthe anode coatings based on doped ceria, titanate- andchromite-based oxides will pose a serious challengein the current collector design and probably limit thepractical application of the materials as anodes. Thelow electrical conductivity of ceria-based anodes is par-tially compensated by using metal paste current collec-tor such as Au [225]. In reality such approach maynot be economically and technically viable. In addi-tion to the generally low and, in some cases, unsta-ble conductivity behavior, the overall electrochemicalactivity of the chromite and titanate based perovskiteanodes is poor as compared to that of conventionalNi/YSZ based cermet anodes. The development of Ni-free alternative materials based on various conduct-ing oxides is largely in the laboratory stage and thereare significant challenges in the development of effec-

tive and stable alternative anodes for the hydrocarbonfuels.

Stability of microstruture and interface at elec-trode/electrolyte region is an important issue in the de-velopment of SOFC technologies. However, stabilityof individual components in SOFC is not only relatedto the component itself but also the interfacial reac-tions between various components of a SOFC stack.For example, in the case of SOFC based on metallic in-terconnect, the interaction between chromium gaseousspecies such as CrO2(OH)2 and the cathode materialssuch as LSM and LSCF can cause significant degrada-tion of the electrocatalytic activities for the O2 reductionreactions on LSM materials [342–344]. As comparedto the serious poisoning effect of chromium specieson the degradation of the polarization performance onthe LSM cathodes, the effect of metallic interconnecton the polarization performance of Ni/YSZ cermet an-odes is generally much smaller [345]. Ni/YSZ cermetanodes are also very stable with YSZ electrolyte andthere are basically no chemical reactions between Niand YSZ at the temperature range of the fabricationand fuel cell operation [110]. However, NiO is slightlysoluble in YSZ with solubility limit of ∼2% NiO inYSZ after heating at 1400◦C for 12 h in air [346]. Sim-ilar results were also reported for the NiO/YSZ systemsintered at 1600◦C for 80 h [347]. Linderoth et al. [348]investigated the effect of NiO and NiO-to-Ni transfor-mation on the conductivity of YSZ. The presence ofNiO reduced the conductivity of 8YSZ by as much as20% in air. Transformation of NiO to Ni under reduc-ing conditions can cause the cubic to tetragonal phasetransformation in YSZ, resulting in further reduction inconductivity. On the other hand, impurities such as Si inthe electrode and YSZ electrolyte can have significanteffect on the electrode behavior of Ni anodes [349, 350].Nevertheless, the effect of dissolved Ni in YSZ and im-purity segregation at the electrode/electrolyte interfacein the case of Ni/YSZ cermet anode on the electrolytestability, electrode behavior and long-term performancestability is still not clearly understood.

One of the major obstacles in the commercial intro-duction of SOFC system is the high cost as comparedto entrenched power generation technologies. As SOFCstack is the only one component of the SOFC system,the stack cost should only be a fraction of the systemcost, which has been targeted to be∼US$400/kW by theUS Department of Energy’s Solid State Energy Conver-sion Alliance (SECA) program [351]. Such aggressivecost reduction would preclude the use of expensive orexotic raw materials and complex processing processesfor anode and anode substrates. It has been recognizedthat the fabrication and assembly methods employed inthe construction of fuel cell stacks are a significant stackcost element. It is important to select a manufacturingprocess that is not only cost competitive but also pro-vides a means for stack performance improvement. Re-ducing the fabrication steps would significantly reducethe production cost. As shown by Rietveld et al. [294],tape casting steps for fabrication of anode substrates canbe reduced from original three steps to one by adjust-ing the suspension. Omitting the pre-sintering step for

4432

the anode substrate also improved screenprinting yieldfrom less than 40 to 95% due to the easier handlingability of the tapes. Basu et al. [352] shows that by us-ing wet powder spraying instead of vacuum slip castingto deposit the anode function layer, one firing step canbe omitted with no effect on the cell performance. Co-firing process is suitable for mass production and to pro-duce anode-supported thin electrolyte cells, reducingmaterial and fabrication cost [353, 354]. However, notall electrode and electrolyte systems are feasible withco-firing processes [353]. On the other hand, the use oflow cost metal alloy [335–340] or non-conducting andconventional porous ceramics [355, 356] as cell sub-strates could reduce the cost and improve the reliabilityof SOFC stack.

Finally, the studies on the mechanism and kinet-ics of anodic reactions are primarily concentrated onthe H2 oxidation reaction on Ni/YSZ based cermet an-odes and are not covered in this report. However, thisdoes not imply that the reaction processes at the elec-trode/electrolyte interface are not important. In fact,the fundamental understanding of the electrode pro-cesses is critical in the development and optimizationof the electrode materials with high performance andlong-term stability. Information regarding the mecha-nism and kinetics of the H2 oxidation reaction can befound in [17, 25–27, 30, 66, 350, 357–365]. However,in solid oxide fuel cells based on thin electrolyte, studyof the reaction and electrode process is hampered by theinability to separate individual electrode polarizationsdue to the detrimental effect of misalignment and asym-metric interface contact between electrode/electrolytecontact on anode and cathode sides on the reliabilityand accuracy of the polarization measurement [366–369]. Such inability in polarization separation is prob-ably the main reason for the lack of detailed informationon the polarization performance and behavior for fueloxidation reactions on the anode-supported cells. Sep-aration of anode process in anode-supported cells inmost cases is not possible as the measured cathode po-larization is almost equal to the total cell polarizationin cells with Ni//YSZ cermet substrate anode and LSMcathode [370, 371]. In anode-supported cells, anodeelectrode area is usually significantly larger than thatof the cathode. The usual practice is to normalize thecell performance (power output and current density) tothe area of smaller electrode, i.e., cathode in this case.However, as shown by Chung et al. [372] recently, suchpractice could lead to an apparently high power outputthan symmetric cells. Using a cathode with geometricarea significantly less than that of anode can enhancethe normalized power density by a factor of two. Itwas suggested that the Ni/YSZ cermet substrate anodeswould be electrochemically superior to screenprintedNi/YSZ anodes by a factor of two at 800◦C [373]. Re-cently, McIntosh et al. [371] reported that electrodeimpedance could be separable if the characteristic fre-quencies associated with the anode and cathode pro-cesses are significantly different. As the developmentof ITSOFC based on anode-supported thin electrolytetechnology is increasingly important, there is an urgentneed to develop ways both numerical and experimental

to accurately separate the electrode polarization in orderto understand the electrode behavior of anode substratefor the fuel oxidation and/or internal reforming reac-tions. As the microstructure requirement of thick anodesubstrates is fundamentally different from that of thinscreenprinted anode coating, it is expected that elec-trode behavior of the thick electrode substrates wouldbe significantly different too [374, 375].

References1. K . H A S S M A N N , Fuel Cells 1 (2001) 78.2. N . Q . M I N H , J. Amer. Ceram. Soc. 76 (1993) 563.3. H . Y O K O K A W A, N. S A K A I , T . H O R I T A and K.

Y A M A J I , Fuel Cells 1 (2001) 117.4. H . Y O K O K A W A , Annu. Rev. Mater. Res. 33 (2003) 581.5. S . P . S . B A D W A L and K. F O G E R , Mater. Forum 21 (1997)

187.6. T . K A W A D A and H. Y O K O K A W A , in “Key Engineering

Materials—Electrical Properties of Oxide Materials,” edited byJ. Nowotny and C. C. Sorrell (Trans Tech Publications, 1997)Vols. 125/126, p. 188.

7. B . C . H . S T E E L E , Solid State Ionics 134 (2000) 3.8. J . P . P . H U I J S M A N S , Curr. Opin. Solid State Mater. Sci. 5

(2001) 317.9. S . P . S . B A D W A L , Solid State Ionics 143 (2001) 39.

10. L . C A R R E T T E, K. A. F R I E D R I C H and U. S T I M M I N G ,Fuel Cells 1 (2001) 5.

11. O . Y A M A M O T O , Electrochim. Acta 45 (2000) 2423.12. S . C . S I N G H A L , Solid State Ionics 135 (2001) 305.13. L . C . D E J O N G H E, C . P . J A C O B S O N and S . J . V I S C O ,

Annu. Rev. Mater. Res. 33 (2003) 169.14. S . C . S I N G H A L and K. K E N D A L L , “High Temperature

Solid Oxide Fuel Cells: Fundamental, Design and Applications”(Oxford, UK, Elsevier Ltd. 2003).

15. S . M U R A K A M I , Y . A K I Y A M A, N. I S H I D A, Y.M I I Y A K E, M. N I S H I D A, Y. I T O H, T . S A I T O and N.F U R U K A W A , Denki Kagaku 59 (1991) 320.

16. S . P . J I A N G , in “Science and Technology of Zirconia V,” editedby S. P. S. Badwal, M. J. Bannister and R. H. J. Hannink (TechnomicPublishing Company, 1993) p. 819.

17. S . P . J I A N G and S . P . S . B A D W A L , J. Electrochem. Soc.144 (1997) 3777.

18. J . M I Z U S A K I , H . T A G A W A, K. I S O B E, M. T A J I K A,I . K O S H I R O, H. M A R U Y A M A and K. H I R A N O , ibid. 141(1994) 1674.

19. H . U C H I D A, M. Y O S H I D A and M. W A T A N A B E , J. Phys.Chem. 99 (1995) 3282.

20. M. S U Z U K I , H . S A S A K I , S . O T O S H I , A . K A J I M U R A

and M. I P P O M M A T S U , Solid State Ionics 62 (1993) 125.21. T . S E T O G U C H I , K . O K A M O T O, K. E G U C H I and H.

A R A I , J. Electrochem. Soc. 139 (1992) 2875.22. E . I V E R S-T I F F EE , W. W E R S I N G, M. S C H I E B L and H.

G R E I N E R , Ber. Bunsen-Ges. Phys. Chem. 94 (1990) 978.23. M. M O R I , T . Y A M O M O T O, H. I T O H, H. I N A B A and

H. T A G A W A , J. Electrochem. Soc. 145 (1998) 1374.24. T . N O R B Y, O. J . V E L L E, H. L E T H-O L S E N and R.

T U N O L D , in “SOFC-II,” edited by S.C. Singhal and H. Iwahara(Electrochemical Society, Pennington, NJ, 1993) Vol. 93–4,p. 473.

25. J . M I Z U S A K I , H . T A G A W A, T . S A I T O, T .Y A M A M U R A, K. K A M I T A N I , K . H I R A N O, S .E H A R A, T . T A K A G I , T . H I K I T A, M. I P P O M M A T S U,S . N A K A G A W A and K. H A S H I M O T O , Solid State Ionics70/71 (1994) 52.

26. A . B I E B E R L E and L . J . G A U C K L E R , ibid. 135 (2000) 337.27. B . D E B O E R, M. G O N Z A L E Z, H. J . M.

B O U W M E E S T E R and H. V E R W E I J , ibid. 127 (2000)269.

28. M. M O G E N S E N and S . S K A A R U P , ibid. 86–88 (1996) 1151.29. S . P . J I A N G , J. Electrochem. Soc. 148 (2001) A887.30. P . H O L T A P P E L S , I . C . V I N K E, L . G . J . D E H A A R T

and U. S T I M M I N G , ibid. 146 (1999) 2976.

4433

31. B . C . H . S T E E L E , Nature 400 (2000) 619.32. A . L . D I C K S , J. Power Sources 61 (1996) 113.33. J . -H . K O H, Y.-S . Y O O, J . -W. P A R K and H. C . L I M ,

Solid State Ionics 149 (2002) 157.34. P . V . H E N D R I K S E N , in “SOFC-V,” edited by S. C. Singhal and

H. Tagawa (Electrochem. Soc., Pennington, NJ, 1997) Vol. 97–40,p. 1319.

35. W. J . D O L L A R D , J. Power Sources 37 (1992) 133.36. P . V E R N O U X, J . G U I N D E T and M. K L E I T Z , J. Elec-

trochem. Soc. 145 (1998) 3487.37. P . V E R N O U X, M. G U I L L O D O, J . F O U L E T I E R and A.

H A M M O U , Solid State Ionics 135 (2000) 425.38. C . M. C H U N, J . D . M U M F O R D and T.

R A M A N A R A Y A N A N , J. Electrochem. Soc. 147 (2000) 3680.39. K . Y A S H I R O, K. T A K E D A, T . T A U R A, T . O T A K E,

A. K A I M A I , Y . N I G A R A, T . K A W A D A, J . M I Z U S A K I

and H. Y U G A M I , in “SOFC-VIII,” edited by S. C. Singhaland M. Dokiya (Electrochem. Soc., Pennington, NJ, 2003)Vol. 2003–07, p. 714.

40. B . C . H . S T E E L E , Solid State Ionics 86–88 (1996) 1223.41. E . P . M U R R A Y, T . T S A I and S . A . B A R N E T T , Nature

400 (1999) 649.42. S . D . P A R K, J . M. V O H S and R. J . G O R T E , ibid. 404

(2000) 265.43. S . P . J I A N G , Battery Bimonthly 32 (2002) 133.44. S . P . J I A N G and S . H . C H A N , Mater. Sci. Tech., in press.45. F . T I E T Z , F . J . D I A S , D . S I M W O N I S and D. S T OV E R ,

J. Euro. Ceram. Soc. 20 (2000) 1023.46. F . T . C I A C C H I , K . M. C R A N E and S . P . S . B A D W A L ,

Solid State Ionics 73 (1994) 49.47. T . H I K I T A , in “Science and Technology of Zirconia V,”

edited by S. P. S. Badwal, M. J. Bannister and R. H. J. Hannink(Technomic Publishing Co., Lancaster, PA, 1993) p. 674.

48. S . M U R A K A M I , Y . A K I Y A M A, N. I S H I D A, T .Y A S U O, T . S A I T O and N. F U R U K A W A , in “SOFC-II,”edited by F. Grosz, P. Zegers, S. C. Singhal and O. Yamamoto(Commission of The European Communities, Luxembourg, 1991)p. 561.

49. S . P . J I A N G, P . J . C A L L U S and S . P . S . B A D W A L ,Solid State Ionics 132 (2000) 1.

50. J . V A N H E R L E, R . I H R I N G E R and A. J . M C E V O Y , in“SOFC-V,” edited by S.C. Singhal and H. Tagawa (Electrochem.Soc., Pennington, NJ, 1997) Vol. 97–40, p. 565.

51. H . I T O H, T . Y A M A M O T O, M. M O R I , T . H O R I T A, N.S A K A I , H . Y O K O K A W A and M. D O K I Y A , J. Electrochem.Soc. 144 (1997) 641.

52. S . P . J I A N G , ibid. 150 (2003) E548.53. T . K A W A D A, N. S A K A I , H . Y O K O K A W A, M.

D O K I Y A, M. M O R I and T . I W A T A , ibid. 137 (1990) 3042.54. T . M A T S U S H I M A, H. O H R U I and T . H I R A I , Solid State

Ionics 111 (1998) 315.55. D . D . U P A D H Y A Y A, T . R . G . K U T T Y and C.

G A N G U L Y , in “Science and Technology of Zirconia V,” editedby S. P. S. Badwal, M. J. Bannister and R. H. J. Hannink(Technomic Publishing Co., Lancaster, PA, 1993) p. 310.

56. F . F . L A N G E , J. Amer. Ceram. Soc. 72 (1989) 3.57. T . F U K U I , S . O H A R A, M. N A I T O and K. N O G I , J.

Power Sources 110 (2002) 91.58. S . P R I M D A H L, B . F . S ØR E N S E N and M. M O G E N S E N ,

J. Amer. Ceram. Soc. 83 (2000) 489.59. R . M I L L I N I and M. F . G A G L I A R D I , J. Mater. Sci. 29

(1994) 4629.60. Y . L I , Y . X I E , J . G O N G, Y. C H E N and Z . Z H A N G ,

Mater. Sci. Engng. B 86 (2001) 119.61. S . L I , R . G U O, J . L I , Y . C H E N and W. L I U , Ceram.

Intern. 29 (2003) 883.62. S . T . A R U N A, M. M U T H U R A M A N and K. C . P A T I L ,

Solid State Ionics 111 (1998) 45.63. A . R I N G U E D E, J . A . L A B R I N C H A and J . R . F R A D E ,

ibid. 141–142 (2001) 549.64. A . R I N G U E D E, D . B R O N I N E and J . R . F R A D E , ibid.

146 (2002) 219.65. Idem., Electrochim. Acta 48 (2002) 437.66. S . P R I M D A H L and M. M O G E N S E N , J. Electrochem. Soc.

144 (1997) 3409.

67. M. M A R I N S E K, K. Z U P A N and J . M A EE K , J. PowerSources 106 (2002) 178.

68. T . F U K U I , T . O O B U C H I , Y . I K U H A R A, S . O H A R A

and K. K O D E R A , J. Amer. Ceram. Soc. 80 (1997) 261.69. T . F U K U I , S . O H A R A and K. M U K A I , Electrochem. Solid

State Lett. 1 (1998) 120.70. X . Z H A N G. , S . O H A R A, R. M A R I C , T . F U K U I ,

H . Y O S H I D A, M. N I S H I M U R A, T . I N A G A K I and K.M I U R A , J. Power Sources 83 (1999) 170.

71. R . M A R I C , S . O H A R A, T . F U K U I , T . I N A G A K I andJ . F U J I T A , Electrochem. Solid State Lett. 1 (1998) 201.

72. C . A . -H. C H U N G, N. J . E . A D K I N S and R. M.O R M E R O D , in “SOFC-VII,” edited by H. Yokokawa and S. C.Singhal (Electrochem. Soc., Pennington, NJ, 2001) Vol. 2001–16,p. 693.

73. F . C H E N and M. L I U , J. Mater. Chem. 10 (2000) 2603.74. V . E S P O S I T O, C . D’O T T A V I , S . F E R R A R I , S .

L I C O C C I A and E . T R A V E R S A , in “SOFC-VIII,” edited byS. C. Singhal and M. Dokiya (Electrochem. Soc., Pennington, NJ,2003) Vol. 2003–07, p. 643.

75. Y . O K A W A, T . M A T S U M O T O, T . D O I and Y. H I R A T A ,J. Mater. Res. 17 (2002) 2266.

76. C . X I A and M. L I U , Solid State Ionics 152/153 (2002) 423.77. J . W. M O O N, H. L . L E E , J . D . K I M, G. D. K I M, D.

A. L E E and H. W. L E E , Mater. Lett. 38 (1999) 214.78. T . F U K U I , K . M U R A T A, S . O H A R A, H. A B E, M.

N A I T O and K. N O G I , J. Power Sources 125 (2004) 17.79. T . I O R O I , Y . U C H I M O T O, Z . O G U M I and Z .

T A K E H A R A , J. Electrochem. Soc. 78 (1995) 593.80. N . M. S A M M E S, M. S . B R O W N and R. R A T N A R A J ,

Mater. Sci. Lett. 13 (1994) 1124.81. C . I W A S A W A, M. N A G A T A and S . Y A M A O K A , in

“SOFC-IV,” edited by M. Dokiya, O. Yamamoto, H. Tagawaand S. C. Singhal (Electrochem. Soc., Pennington, NJ, 1995)Vol. 95–01, p. 686.

82. R . C R A C I U M, S . P A R K, R . J . G O R T E, J . M. V O H S,C . W A N G and W. L . W O R R E L L , J. Electrochem. Soc. 146(1999) 4019.

83. S . P A R K, R . C R A C I U M, J . M. V O H S and R. J .G O R T E , ibid. 146 (1999) 3603.

84. D . W. D E E S , T . D . C L A A R, T . E . E A S L E R, D. C .F E E and F . C . M R A Z E K , ibid. 134 (1987) 2141.

85. H . K O I D E, Y. S O M E Y A, T . Y O S H I D A and T .M A R U Y A M A , Solid State Ionics 132 (2000) 253.

86. J . -H . L E E , H. M O O N, H.-W. L E E, J . K I M, J . -D .K I M and K.-H. Y O O N , ibid. 148 (2002) 15.

87. A . T I N T I N E L L I , C . R I Z Z O, G. G I U N T A and A.S E L V A G G I , in “1st European SOFC Forum,” edited by U. Bossel(European Fuel Cells Forum, Lucerne, Switzerland, 1994) p. 455.

88. U . A N S E L M I-T A M B U R I N I , G . C H I O D E L L I , M.A R I M O N D I , F . M A G L I A, G. S P I N O L O and Z . A.M U N I R , Solid State Ionics 110 (1998) 35.

89. S . F . C O R B I N and X. Q I A O , J. Amer. Ceram. Soc. 86 (2003)401.

90. U . A N S E L M I-T A M B U R I N I , M. A R I M O N D I , F .M A G L I A, G. S P I N O L O and Z . A. M U N I R , ibid. 81 (1998)1765.

91. W. H U E B N E R, H. U. A N D E R S O N, D. M. R E E D,S . R . S E H L I N and X. D E N G , in “SOFC-IV,” edited by M.Dokiya, O. Yamamoto, H. Tagawa and S. C. Singhal (Electrochem.Soc., Pennington, NJ, 1995) Vol. 95–01, p. 696.

92. H . I T O H, T . W A T A N A B E, D. W A D A and K. S U Z U K I ,in “5th European SOFC Forum,” edited by J. Huijsmans (EuropeanFuel Cells Forum, Lucerne, Switzerland, 2002) p. 311.

93. S . P . J I A N G, J . G . L O V E and L . A P A T E A N U , SolidState Ionics 160 (2003) 15.

94. J . -H . L E E , J . -W. H E O, D.-S . L E E , J . K I M, G.-H.K I M, H.-W. L E E, H. S . S O N G and J . -H . M O O N , ibid.158 (2003) 225.

95. X . D E N G and A. P E T R I C , in “SOFC-VIII,” edited by S.C. Singhal and M. Dokiya (Electrochem. Soc., Pennington, NJ,2003) Vol. 2003–07, p. 653.

96. W. H U E B N E R, D. M. R E E D and H. U. A N D E R S O N ,in “SOFC-VI,” edited by S. C. Singhal and M. Dokiya (Electro-chemical Society, Pennington, NJ, 1999) Vol. 99–19, p. 503.

4434

97. S . K . P R A T I H A R, R . N. B A S U, S . M A J U M D E R and H.S . M A I T I , in “SOFC-VI,” edited by S. C. Singhal and M. Dokiya(Electrochem. Soc., Pennington, NJ, 1999) Vol. 99–19, p. 513.

98. L . G R A H L-M A D S E N, P . H . L A R S E N, N. B O N A N O S,J . E N G E L L and S . L I N D E R O T H , in “5th European SOFCForum,” edited by J. Huijsmans (European Fuel Cells Forum,Lucerne, Switzerland, 2002) p. 82.

99. Y . M. P A R K and G. M. C H O I , Solid State Ionics 120 (1999)265.

100. A . S E L C U K and A. A T K I N S O N , J. Euro. Ceram. Soc. 17(1997) 1523.

101. A . A T K I N S O N and A. S E L C U K , in “SOFC-V,” edited byU. Stimming, S. C. Singhal, H. Tagawa and W. Lehnert (Electro-chemical Society, Pennington, NJ, 1997) Vol. 97–40, p. 671.

102. A . S E L C U K and A. A T K I N S O N , J. Amer. Ceram. Soc. 83(2000) 2029.

103. B . F . S ØR E N S E N and S . P R I M D A H L , J. Mater. Sci. 33(1998) 5291.

104. A . S E L C U K, G. M E R E R E and A. A T K I N S O N , ibid. 36(2001) 1173.

105. T . I W A T A , J. Electrochem. Soc. 143 (1996) 1521.106. D . S I M W O N I S , F . T I E T Z and D. S T OV E R , Solid State

Ionics 132 (2000) 241.107. H . I T O H, T . Y A M O M O T O, M. M O R I , T . W A T A N A B E

and T . A B E , Denki Kagaku 64 (1996) 549.108. A . G U B N E R, H. L A N D E S, J . M E T Z G E R, H. S E E G

and R. S T UB N E R , in “SOFC-V,” edited by S. C. Singhal and H.Tagawa (Electrochem. Soc., Pennington, NJ, 1997) Vol. 97–40,p. 844.

109. P . N I K O L O P O U L O S and D. S O T I R O P O U L O U , J. Mater.Sci. Lett. 6 (1996) 1429.

110. A . T S O G A, A. N A O M I D I S and P . N I K O L O P O U L O S ,Acta mater. 44 (1996) 3679.

111. A . N A O U M I D I S , A . T S O G A, P . N I K O P O U L O S andH. G R UB M E I E R , in “SOFC-IV,” edited by M. Dokiya, O.Yamamoto, H. Tagawa and S. C. Singhal (Electrochem. Soc.,Pennington, NJ, 1995) Vol. 95–01, p. 667.

112. S . P . J I A N G , J. Mater. Sci. 38 (2003) 3775.113. R . W I L K E N B ON E R, T H. K L O I D T and W. M A L L EN E R ,

in “SOFC-V,” edited by S. C. Singhal and H. Tagawa (Electrochem.Soc., Pennington, NJ, 1997) Vol. 97–40, p. 851.

114. A . M UL L E R , in “3rd European SOFC Forum,” edited by P.Stevens (European Fuel Cell Group, Lucerne, Switzerland, 1998)p. 353.

115. E . I V E R S-T I F F EE , A . W E B E R and D. H E R B S T R I T T , J.Euro. Ceram. Soc. 21 (2001) 1805.

116. A . I O S E L E V I C H, A. A. K O R N Y S H E V and W.L E H N E R T , J. Electrochem. Soc. 144 (1997) 3010.

117. J . A B E L, A. A. K O R N Y S H E V and W. L E H N E R T , ibid.144 (1997) 4253.

118. R . V A ßE N, D. S I M W O N I S and D. S T OV E R , J. Mater. Sci.36 (2001) 147.

119. M. N A G A I S H I , K . T A K E U C H I , A . F U K U N A G A, M.E G A S H I R A and N. Y A M A Z O E , J. Ceram. Soc. Japan, Int.Edition 105 (1997) 579.

120. C . V O I S A R D, U. W E I S S E N, E . B A T A W I and R.K R U S C H W I T Z , in “5th European SOFC Forum,” edited by J.Huijsmans (European Fuel Cells Forum, Lucerne, Switzerland,2002) p. 18.

121. V . A N T O N U C C I , E . M O D I C A, G. M O N F O R T E, A.S . A R I C O and P . L . A N T O N U C C I , J. Euro. Ceram. Soc. 18(1998) 113.

122. C . W. T A N N E R, K. Z . F U N G and A. V. V I R K A R , J.Electrochem. Soc. 144 (1997) 21.

123. S . S U N D E , ibid. 142 (1995) L50.124. Idem., ibid. 143 (1996) 1123.125. Idem., ibid. 143 (1996) 1930.126. P . C O S T A M A G N A, P . C O S T A and V. A N T O N U C C I ,

Electrochim. Acta 43 (1998) 375.127. P . C O S T A M A G N A, P . C O S T A and E . A R A T O , ibid. 43

(1998) 967.128. S . H . C H A N and Z. T . X I A , J. Electrochem. Soc. 148 (2001)

A388.129. A . S . L O S E L E V I C H and A. A. K O R N Y S H E V , Fuel Cells

1 (2001) 40.

130. S . H . C H A N, K. A. K H O R and Z . T . X I A , J. PowerSources 93 (2001) 130.

131. M. B R O W N, S . P R I M D A H L and M. M O G E N S E N , J.Electrochem. Soc. 147 (2000) 475.

132. A . B I E B E R L E and L . J . G A U C K L E R , Z. Metallkd. 92(2001) 7.

133. C . -H . L E E , C . -H. L E E , H.-Y. L E E and S . M. O H ,Solid State Ionics 98 (1997) 39.

134. S . P . J I A N G, Y. Y. D U A N and J . G . L O V E , J.Electrochem. Soc. 149 (2002) A1175.

135. S . P . J I A N G, Y. J . L E N G, S . H . C H A N and K. A.K H O R , Electrochem. Solid State Lett. 6 (2003) A67.

136. S . P . Y O O N, J . H A N, S . W. N A M, T . -H. L I M, I . -H .O H, S . -A . H O N G, Y.-S . Y O O and H. C . L I M , J. PowerSources 106 (2002) 160.

137. S . P . Y O O N, J . H A N, S . W. N A M, T . -H. L I M, I . -H .O H, S . -A . H O N G and Y.-S . Y O O , in “5th European SOFCForum,” edited by J. Huijsmans (European Fuel Cell Group,Lucerne, Switzerland, 2002) p. 148.

138. K . C . C H O U, S . Y U A N and U. P A L , in “SOFC-III,” editedby S. C. Singhal and H. Iwahara (Electrochem. Soc., Pennington,NJ, 1993) p. 431.

139. W. S C H AF E R, A. K O C H, U. H E R O L D-S C H M I D T andD. S T O L T E N , Solid State Ionics 86–88 (1996) 1235.

140. H . U C H I D A, H. S U Z U K I and M. W A T A N A B E , J.Electrochem. Soc. 145 (1998) 615.

141. M. C A S S I D Y and G. L I N D S A Y , in “1st European SOFCForum,” edited by U. Bossel (European Fuel Cell Group, Lucerne,Switzerland, 1994) p. 205.

142. A . C . M UL L E R, D. H E R B S T R I T T and E . I V E R S-T I F F EE , Solid State Ionics 152/153 (2002) 537.

143. U . B . P A L and S . C . S I N G H A L , J. Electrochem. Soc. 137(1990) 2937.

144. L . S . W A N G and S . A . B A R N E T T , Solid State Ionics 61(1993) 273.

145. Z . O G U M I , T . I O R O I , Y . U C H I M O T O, Z .T A K E H A R A, T . O G A W A and K. T O Y A M A , J. Amer.Ceram. Soc. 78 (1995) 593.

146. E . Z . T A N G, T . H . E T S E L L and D. G. I V E Y , ibid. 83(2000) 1626.

147. J . L . Y O U N G and T. H. E T S E L L , Solid State Ionics 135(2000) 457.

148. X . W A N G, N. N A K A G A W A and K. K A T O , J. Electrochem.Soc. 148 (2001) A565.

149. S . P . J I A N G and Y. R A M P R A K A S H , Solid State Ionics 122(1999) 211.

150. P . H O L T A P P E L S , L . G . D E H A A R T, U. S T I M M I N G,I . C . V I N K E and M. M O G E N S E N , J. Appl. Electrochem. 29(1999) 561.

151. T . K A W A D A, I . A N Z A I , N . S A K A I , H . Y O K O K A W A

and M. D O K I Y A , in Proc. of Sym. On High TemperatureElectrode Materials and Characterization, edited by D. Macdonaldand A. C. Khandkar (Electrochem. Soc., Pennington, NJ, 1991)Vol. 91–6, p. 165.

152. A . L . L E E , R . F . Z A B R A N S K Y and W. J . H U B E R , Ind.Eng. Chem. Res. 29 (1990) 766.

153. K . A H M E D, P . S E S H A D R I , Y . R A M P R A K A S H, S .P . J I A N G and K. F O G E R , in “SOFC-V,” edited by S. C.Singhal and H. Tagawa (Electrochem. Soc., Pennington, NJ, 1997)Vol. 97–40, p. 228.

154. K . A H M E D and K. F O G E R , Catalysis Today 63 (2000)479.

155. E . A C H E N B A C H and E. R I E N S C H E , J. Power Sources 52(1994) 283.

156. R . O D E G A A R D, E . J O H N S E N and H. K A R O L I U S S E N ,in “SOFC-IV,” edited by M. Dokiya, O. Yamamoto, H. Tagawaand S. C. Singhal (Electrochem. Soc., Pennington, NJ, 1995)Vol. 95–01, p. 810.

157. V . D . B E L Y A E V, T . I . P O L I T O V A, O. A. M A R I N A

and V. A. S O B Y A N I N , Appl. Catal. A 133 (1995) 47.158. A . L . D I C K S, K. D. P O I N T O N and A. S W A N N , in “3rd

European SOFC Forum,” edited by P. Stevens (European FuelCell Group, Lucerne, Switzerland, 1998) p. 249.

159. C . M. F I N N E R T Y, N. J . C O E, R . H . C U N N I N G H A M

and R. M. O R M E R O D , Catalysis Today 46 (1998) 137.

4435

160. I . D R E S C H E R, W. L E H N E R T and J . M E U S I N G E R ,Electrochim. Acta 43 (1998) 3059.

161. C . P E R E G O, L . Z A N I B E L L I , M. C A R T R U L L O and G.P I R O , in “SOFC-III,” edited by S. C. Singhal and H. Iwahara(Electrochem. Soc., Pennington, NJ, 1993) p. 454.

162. T . S H I R A K A W A, S . M A T S U D A and A. F U K U S H I M A , in“SOFC-III,” edited by S. C. Singhal and H. Iwahara (Electrochem.Soc., Pennington, NJ, 1993) p. 464.

163. A . W E B E R, B . S A U E R, A. C . M UL L E R, D.H E R B S T R I T T and E . I V E R S-T I F F EE , Solid State Ion-ics 152/153 (2002) 543.

164. T . A I D A, A. A B U D U L A, M. I H A R A, H. K O M I Y A M A

and K. Y A M A D A , in “SOFC-IV,” edited by M. Dokiya, O.Yamamoto, H. Tagawa and S. C. Singhal (Electrochem. Soc.,Pennington, NJ, 1995) Vol. 95–01, p. 801.

165. A . A B U D U L A, M. I H A R A, H. K O M I Y A M A andK. Y A M A D A , Solid State Ionics 86–88 (1996) 1203.

166. M. I H A R A, C. Y O K O Y A M A, A. A B U D U L A, R.K A T O, H. K O M I Y A M A and K. Y A M A D A , J. Electrochem.Soc. 146 (1999) 2481.

167. J . L I U and S . A . B A R N E T T , Solid State Ionics 158 (2003) 11.168. T . T A K E G U C H I , Y . K A N I , T . Y A N O, R. K I K U C H I ,

K . E G U C H I , K . T S U J I M O T O, Y. U C H I D A, A. U E N O,K. O M O S H I K I and M. A I Z A W A , ibid. 112 (2002) 588.

169. T . T A K E G U C H I , T . Y A N O, Y. K A N I , R . K I K U C H I

and K. E G U C H I , in “SOFC-VIII,” edited by S. C. Singhal and M.Dokiya (Electrochem. Soc., Pennington, NJ, 2003) Vol. 2003–07,p. 704.

170. T . T A K E G U C H I , R . K I K U C H I , T . Y A N O, K. E G U C H I

and K. M U R A T A , Catalysis Today 84 (2003) 217.171. S . O N U M A, A. K A I M A I , K . K A W A M U R A, Y.

N I G A W A, T . K A W A D A, J . M I Z U S A K I , H . I N A B A

and H. T A G A W A , J. Electrochem. Soc. 145 (1998) 920.172. T . H O R I T A, N. S A K A I , T . K A W A D A, H. Y O K O K A W A

and M. D O K I Y A , ibid. 143 (1996) 1161.173. K . S A T O, Y. O H M I N E, K. O G A S A and S . T S U J I ,

in “SOFC-VIII,” edited by S. C. Singhal and M. Dokiya(Electrochem. Soc., Pennington, NJ, 2003) Vol. 2003–07, p. 695.

174. A . R I N G U E D E, D . P . F A G G and J . R . F R A D E , J. Euro.Ceram. Soc. 24 (2004) 1355.

175. H . K I M, C . L U, W. L . W O R R E L L, J . M. V O H S andR. J . G O R T E , J. Electrochem. Soc. 149 (2002) A247.

176. H . K I M, C . D A R O S A, M. B O A R O, J . M. V O H S andR. J . G O R T E , J. Amer. Ceram. Ceram. 85 (2002) 1473.

177. A . S I N , A . T A V A R E S, Y. D O U B I T S K Y, A. Z A O P O,A. S . A R I C O, L . R . G U L L O, D. L A R O S A, S .S I R A C U S A N O and V. A N T O N U C C I , in “SOFC-VIII,” editedby S. C. Singhal and M. Dokiya (Electrochem. Soc., Pennington,NJ, 2003) Vol. 2003–07, p. 745.

178. K . K A M M E R and M. M O G E N S E N , in “SOFC-VIII,” editedby S. C. Singhal and M. Dokiya (Electrochem. Soc., Pennington,NJ, 2003) Vol. 2003–07, p. 781.

179. J . G E Y E R, H. K O H L M U L L E R, H. L A N D E S andR. S T U B N E R , in “SOFC-V,” edited by U. Stimming, S.C. Singhal, H. Tagawa and W. Lehnert (Electrochem. Soc.,Pennington, NJ, 1997) Vol. 97–40, p. 585.

180. Y . M A T S U Z A K I and I . Y A S U D A , Solid State Ioincs 132(2000) 261.

181. D . S T O L F E N, R . S P A H and R. S C H A M M , in “SOFV-V,”edited by S. C. Singhal and H. Tagawa (Electrochem. Soc.,Pennington, NJ, 1997) Vol. 97–40, p. 88.

182. P . P R I M D A H L , in “SOFC-VI,” edited by S. C. Singhal and M.Dokiya (Electrochem. Soc., Pennington, NJ, 1999) Vol. 99–19,p. 530.

183. K . E G U C H I , T . S E T O G U C H I , K . O K A M O T O and H.A R A I , in Proc. of the International Fuel Cell Conf. (Makihari,Japan, 1992) p. 373.

184. S . Z . W A N G and I . T A T S U M I , Acta Physico-Chimica Sinica19 (2003) 844.

185. M. B . J O E R G E R and L . J . G A U C K L E R , in “SOFC-VII,”edited by H. Yokokawa and S. C. Singhal (Electrochem. Soc.,Pennington, NJ, 2001) Vol. 2001–16, p. 662.

186. C . - J . W E N, T . M A S U Y A M A, T . Y O S H I K A W A,J . O T O M O. H. T A K A H A S H I , K . E G U C H I andK. Y A M A D A , in “SOFC-VII,” edited by H. Yokokawa and S. C.

Singhal (Electrochem. Soc., Pennington, NJ, 2001) Vol. 2001–16,p. 671.

187. T . I N A G A K I , H . Y O S H I D A, K. M I U R A, S . O H A R A,R. M A R I C , X. Z H A N G, K. M U K A I and T . F U K U I ,in “SOFC-VII,” edited by H. Yokokawa and S. C. Singhal(Electrochem. Soc., Pennington, NJ, 2001) Vol. 2001–16,p. 963.

188. E . P . M U R R A Y, S . J . H A R R I S and H. J E N , J. Electrochem.Soc. 149 (2002) A1127.

189. B . R OS C H, H. T U, A. O. S T OR M E R, A. C . M UL L E R

and U. S T I M M I N G , in “SOFC-VIII,” edited by S. C. Singhaland M. Dokiya (Electrochem. Soc., Pennington, NJ, 2003)Vol. 2003–07, p. 737.

190. T . H I B I N O, A. H A S H I M O T O, K. A S A N O, M. Y A N O,M. S U Z U K I and M. S A N O , Electrochem. Solid-State Lett. 5(2002) A242.

191. T . H I B I N O, A. H A S H I M O T O, M. Y A N O, M. S U Z U K I ,S . - I . Y O S H I D A and M. S A N O , J. Electrochem. Soc. 149(2002) A133.

192. C . X I A , F . C H E N and M. L I U , Electrochem. Solid-StateLett. 4 (2001) A52.

193. A . T S O G A, A. G U P T A, A. N A O U M I D I S , D .S K A R M O U T S O S and P . N I K O L O P O U L O S , Ionics 4(1999) 234.

194. A . T S O G A, A. N A O U M I D I S , A . G U P T A and D.S T OV E R , Mater. Sci. Forum 308–311 (1999) 234.

195. A . T S O G A, A. G U P T A, A. N A O U M I D I S and P .N I K O L O P O U L O S , Acta Mater. 48 (2000) 4709.

196. H . U C H I D A, S . A R I S A K A and M. W A T A N A B E ,Electrochem. Solid-State Lett. 2 (1999) 428.

197. S . O H A R A, R. M A R I C , X. Z H A N G, K. M U K A I , T .F U K U I , H . Y O S H I D A, T . I N A G A K I and K. M I U R A , J.Power Sources 86 (2000) 455.

198. K . K U R O D A, I . H A S H I M O T O, K. A D A C H I , J .A K I K U S A, Y. T A M O U, N. K O M A D A, T . I S H I H A R A

and Y. T A K I T A , Solid State Ionics 132 (2000) 199.199. K . H U A N G, J . -H . W A N and J . B . G O O D E N O U G H , J.

Electrochem. Soc. 148 (2001) A788.200. K . H U A N G, R. T I C H Y, J . B . G O O D E N O U G H and C.

M I L L I K E N , J. Amer. Ceram. Soc. 81 (1998) 2581.201. S . M. C H O I , K . T . L E E , S . K I M, M. C. C H U N and

H. L . L E E , Solid State Ionics 131 (2000) 221.202. X . Z H A N G, S . O H A R A, H. O K A W A, R. M A R I C and

T . F U K U I , ibid. 139 (2001) 145.203. M. M O R I , Y . H I E R, H. I T O H, G. A. T O M P S E T T and

N. M. S A M M E S , ibid. 160 (2003) 1.204. K . K O B A Y A S H I , Y . K A I , S . Y A M A G U C H I , N .

F U K A T S U, T . K A W A S H I M A and Y. I G U C H I , ibid. 93(1997) 193.

205. K . E . S W I D E R and W. L . W O R R E L L , J. Electrochem. Soc.143 (1996) 3706.

206. S . D . P A R K, H. K I M, S . M C I N T O S H, W. W O R R E L L,R . J . G O R T E and J . M. V O H S , in “SOFC-VII,” edited by H.Yokokawa and S. C. Singhal (Electrochem. Soc., Pennington, NJ,2001) Vol. 2001–16, p. 712.

207. S . P A R K, R . J . G O R T E and J . M. V O H S , J. Electrochem.Soc. 148 (2001) A443.

208. R . J . G O R T E, H. K I M and J . M. V O H S , J. Power Sources106 (2002) 10.

209. I . A L S T R U P, M. T . T A V A R E S, C . A . B E R N A R D O,O. S ØR E N S E N and J . R . R O S T R U P-N I E L S E N , Mat. AndCorr. 49 (1998) 367.

210. C . L U, W. L . W O R R E L L, J . M. V O H S and R. J .G O R T E , in “SOFC-VIII,” edited by S. C. Singhal and M.Dokiya (Electrochem. Soc., Pennington, NJ, 2003) Vol. 2003–07,p. 773.

211. S . M C I N T O S H, J . M. V O H S and R. J . G O R T E , J.Electrochem. Soc. 150 (2003) A470.

212. Idem., Electrochim. Acta 47 (2002) 3815.213. R . J . G O R T E, S . P A R K, J . M. V O H S and C. W A N G ,

Adv. Mater. 12 (2000) 1465.214. H . S A S A K I , H . S U Z U K I , S . O T O S H I , A . K A J I M U and

M. I P P O M M A T S U , J. Electrochem. Soc. 139 (1992) L12.215. P . V E R N O U X, J . G U I N D E T, E . G E H A I N and M.

K L E I T Z , in “SOFC-V,” edited by U. Stimming, S. C. Singhal,

4436

H. Tagawa and W. Lehnert (Electrochem. Soc., Pennington, NJ,1997) Vol. 97–40, p. 219.

216. K . E G U C H I , J. Alloys Comp. 250 (1997) 486.217. G . B . B A L A Z S and R. S . G L A S S , Solid State Ionics 76

(1995) 155.218. B . C . H . S T E E L E , ibid. 129 (2000) 95.219. M. M O G E N S E N, N. M. S A M M E S and G. A.

T O M P S E T T , ibid. 129 (2000) 63.220. B . C . H . S T E E L E, P . H . M I D D L E T O N and R. R U D K I N ,

ibid. 40/41 (1990) 388.221. I . S . M E T C A L F E , ibid. 57 (1992) 259.222. N . V . S K O R O D U M O V A, S . I . S I M A K, B. I .

L U N D Q V I S T , I . A . A B R I K O S O V and B. J O H A N S S O N ,Phys. Rev. Lett. 89 (2002) 6601.

223. M. M O G E N S E N , in “SOFC-II,” edited by F. Gross, P. Zegers,S. C. Singhal and O. Yamamoto (Commission of the EuropeanCommunities, Brussels, 1991) p. 577.

224. M. M O G E N S E N, T . L I N D E G A A R D, U. R . H A N S E N

and G. M O G E N S E N , J. Electrochem. Soc. 141 (1994) 2122.225. O . A . M A R I N A, C . B A G G E R, S . P R I M D A H L and M.

M O G E N S E N , Solid State Ionics 123 (1999) 199.226. A . O . S T OR M E R, P . H O L T A P P E L S , H. T U and U.

S T I M M I N G , in “5th European SOFC Forum,” edited by J.Huijsmans (European Fuel Cell Group, Lucerne, Switzerland,2002) p. 343.

227. H . U C H I D A, H. S U Z U K I and M. W A T A N A B E , J.Electrochem. Soc. 146 (1999) 1667.

228. H . U C H I D A, M. S U G I M O T O and M. W A T A N A B E ,in “SOFC-VII,” edited by H. Yokokawa and S. C. Singhal(Electrochem. Soc., Pennington, NJ, 2001) Vol. 2001–16,p. 653.

229. E . S . P U T N A, J . S T U B E N R A U C H, J . M. V O H S andR. J . G O R T E , Langmuir 11 (1995) 4832.

230. H . U C H I D A, S . S U Z U K I and M. W A T A N A B E , in “SOFC-VIII,” edited by S. C. Singhal and M. Dokiya (Electrochem. Soc.,Pennington, NJ, 2003) Vol. 2003–07, p. 728.

231. S . P R I M D A H L and Y. L . L I U , J. Electrochem. Soc. 149(2002) A1466.

232. S . P R I M D A H L and M. M O G E N S E N , Solid State Ionics152/153 (2002) 597.

233. T . H O R I T A, N. S A K A I , H . Y O K O K A W A, M. D O K I Y A

and T . K A W A D A , ibid. 86–88 (1996) 1259.234. D . P . S U T I J A , T . N O R B Y, P E R A. O S B O R G and P E R

K O F S T A D , in “SOFC-III,” edited by S. C. Singhal and H. Iwahara(Electrochem. Soc., Pennington, NJ, 1993) Vol. 93–04, p. 552.

235. A . H O L T, E . A H L G R E N and F . W. P O U L S E N , in“SOFC-III,” edited by S. C. Singhal and H. Iwahara (Electrochem.Soc., Pennington, NJ, 1993) Vol. 93–04, p. 562.

236. O . A . M A R I N A, N. L . C A N F I E L D and J . W.S T E V E N S O N , Solid State Ionics 149 (2002) 21.

237. O . A . M A R I N A and L . R . P E D E R S O N , in “5th EuropeanSOFC Forum,” edited by J. Huijsmans (European SOFC Forum,Switzerland, 2002) p. 481.

238. W. L . W O R R E L L , Solid State Ionics 52 (1992) 147.239. M. T . C O L O M E R, J . R . J U R A D O, R. M. C . M A R Q U E S

and F . M. B . M A R Q U E S , in Proc. of 2nd Intern. Sym. on Ionicand Mixed Conducting Ceramics (Electrochem. Soc., Pennington,NJ 1994) Vol. 94–12, p. 369.

240. H . N A I T O and H. A R A S H I , Solid State Ionics 53–56 (1992)436.

241. P . H A N and W. L . W O R R E L L , in Proc. of 2nd Intern. Sym.on Ionic and Mixed Conducting Ceramics (Electrochem. Soc.,Pennington, NJ, 1994) Vol. 94–12, p. 317.

242. S . T A O and J . T . S . I R V I N E , J. Solid State Chem. 165(2002) 12–18.

243. A . J . F E I G H E R Y, J . T . S . I R V I N E, D. P . F A G G andA. K A I S E R , ibid. 143 (1999) 273.

244. M. K. D O N G A R E, A. M. D O N G A R E, V. B . T A R E andE . K E M N I T Z , Solid State Ionics 152/153 (2002) 455.

245. P . H . M I D D L E T O N, H. J . S T E I N E R, G. M. C H R I S T I E ,R . B A K E R, I . S . M E T C A L F E and B. C . H . S T E E L E , in“SOFC-III,” edited by S. C. Singhal and H. Iwahara (Electrochem.Soc., Pennington, NJ, 1993) Vol. 93–04, p. 542.

246. D . P . F A G G, S . M. F R A Y and J . T . S . I R V I N E , SolidState Ionics 72 (1994) 235.

247. P . R . S L A T E R, D. P . F A G G and J . T . S . I R V I N E , J.Mater. Chem. 7 (1997) 2495.

248. P . R . S L A T E R and J . T . S . I R V I N E , Solid State Ionics 120(1999) 125.

249. J . C A N A L E S-V AZ Q U E Z, S . W. T A O and J . T . S .I R V I N E , ibid. 159 (2003) 159.

250. K . Y A S H I R O, T . K O B A Y A S H I , L . Q . H A N, A.K A I M A I , Y . N I G A R A, T . K A W A D A, J . M I Z U S A K I

and K. K A W A M U R A , in “SOFC-VII,” edited by H. Yokokawaand S. C. Singhal (Electrochem. Soc., Pennington, NJ, 2001)Vol. 2001–16, p. 678.

251. R . M O O S, S . S C H OL L H A M M E R and K. H. H AR D T L ,Appl. Phys. A 65 (1997) 291.

252. S . H U I and A. P E T R I C , J. Euro. Ceram. Soc. 22 (2002) 1673.253. Idem., J. Electrochem. Soc. 149 (2002) J1.254. H . Y O K O K A W A, N. S A K A I , T . K A W A D A and M.

D O K I Y A , Solid State Ionics 52 (1992) 43.255. N . S A K A I , K . Y A M A J I , T . H O R I T A, H. Y O K O K A W A,

T . K A W A D A and M. D O K I Y A , J. Electrochem. Soc. 147(2000) 3178.

256. M. M O R I and Y. H I E R , J. Amer. Ceram. Soc. 84 (2001) 2573.257. S . T A N A S E S C U, D. B E R G E R, D. N E I N E R and N. D.

T O T I R , Solid State Ionics 157 (2003) 365.258. J . V U L L I E T , B . M O R E L, J . L A U R E N C I N, G.

G A U T H I E R, L . B I A N C H I , S . G I R A U D, J . -Y . H E N R Y

and F . L E F E B V R E- J O U D , in “SOFC-VIII,” edited by S. C.Singhal and M. Dokiya (Electrochem. Soc., Pennington, NJ,2003) Vol. 2003–07, p. 803.

259. J . S F E I R , J . V A N H E R L E and A. J . M C E V O Y , J. Euro.Ceram. Soc. 19 (1999) 897.

260. J . S F E I R , P . A . B U F F A T, P . M OC K L I , N .X A N T H O P O U L O S, R . V A S Q U E Z, H. J . M A T H I E U, J .V A N H E R L E and K. R . T H A M P I , J. Catal. 202 (2001) 229.

261. J . S F E I R , J. Power Sources 118 (2003) 276.262. S . T A O and J . T . S . I R V I N E , in “SOFC-VIII,” edited by S.

C. Singhal and M. Dokiya (Electrochem. Soc., Pennington, NJ,2003) Vol. 2003–07, p. 793.

263. J . S F E I R , J . V A N H E R L E and R. V A S Q U E Z , in “5thEuropean SOFC Forum,” edited by J. Huijsmans (European SOFCForum, Switzerland, 2002) p. 570.

264. A . L . S A U V E T and J . T . S . I R V I N E , in “5th EuropeanSOFC Forum,” edited by J. Huijsmans (European SOFC Forum,Switzerland, 2002) p. 490.

265. J . L I U , B . D . M A D S E N, Z . J I and S . A . B A R N E T T ,Electrochem. Solid-State Lett. 5 (2002) A122.

266. S . P R I M D A H L, J . R . H A N S E N, L . G R A H L-M A D S E N

and P . H . L A R S E N , J. Electrochem. Soc. 148 (2001) A74.267. N . G U N A S E K A R A N, N. B A K S H I , C . B . A L C O C K and

J . J . C A R B E R R Y , Solid State Ionics 83 (1996) 145.268. P . V E R N O U X, J . G U I N D E T, E . G E H A I N and M.

K L E I T Z , in “SOFC-VI,” edited by S. C. Singhal and M. Dokiya(Electrochem. Soc., Pennington, NJ, 1999) Vol. 99–19, p. 219.

269. J . S F E I R , J . V A N H E R L E and A. J . M C E V O Y , in “3rdEuropean SOFC Forum,” edited by P. Stevens (European SOFCForum, Switzerland, 1998) p. 267.

270. G . P U D M I C H, B . A. B O U K A M P, M. G O N Z A L E Z-C U E N C A, W. J U N G E N, W. Z I P P R I C H and F . T I E T Z ,Solid State Ionics 135 (2000) 433.

271. P . V E R N O U X, E . D I J U R A D O and M. G U I L L O D O , J.Amer. Ceram. Soc. 84 (2001) 2289.

272. A . -L . S A U V E T and J . F O U L E T I E R , J. Power Sources 101(2001) 259.

273. A . -L . S A U V E T, J . F O U L E T I E R, F . G A I L L A R D andM. P R I M E T , J. Catal. 209 (2002) 25.

274. S . T A O and J . T . S . I R V I N E , Nature Mater. 2 (2003) 320.275. S . W A N G, Y. J I A N G, Y. Z H A N G, W. L I , J . Y A N and

Z . L U , Solid State Ionics 120 (1999) 75.276. A . H A R T L E Y, M. S A H I B Z A D A, M. W E S T O N, I . S .

M E T C A L F E and D. M A N T Z A V I N O S , Catalysis Today 55(2000) 197.

277. S . P . J I A N G , Solid State Ionics 146 (2002) 1.278. E . D . W A C H S M A N and T. L . C L I T E S , in “Ionic and Mixed

Conducting Ceramics IV,” edited by T. A. Ramanarayanan, W. L.Worrell and M. Mogensen (Electrochem. Soc., Pennington, NJ,2001) Vol. 2001–28, p. 306.

4437

279. T . D . M C C O L M and J . T . S . I R V I N E , Solid State Ionics152/153 (2002) 615.

280. P . R . S L A T E and J . T . S . I R V I N E , ibid. 124 (1999) 61.281. S . T A O and J . T . S . I R V I N E , ibid. 154/155 (2002) 659.282. Idem., J. Mater. Chem. 12 (2002) 2356.283. O . P O R A T, C . H E R E M A N S and H. L . T U L L E R , Solid

State Ionics 94 (1997) 75.284. A . M I T T E R D O R F E R and L . J . G A U C K L E R , ibid. 111

(1998) 185.285. J . -P . Z H A N G, S . P . J I A N G, J . G . L O V E, K. F O G E R

and S . P . S . B A D W A L , J. Mater. Chem. 8 (1998) 2787.286. S . P . J I A N G, Z . -P . Z H A N G and K. F O G E R , J. Euro.

Ceram. Soc. 23 (2003) 1865.287. T . I S H I H A R A, K. S H I M O S E, T . K U D O, H.

N I S H I G U C H I , T . A K B A Y and Y. T A K I T A , J. Amer.Ceram. Soc. 83 (2000) 1921.

288. C . C . W H I T E and B. P . M A R E T T , “Restriction onPlanar Solid Oxide Fuel Cell Design,” BHP Research ReportBHPR/PDR/R/92/023/KH21 (Melbourne, Australia, 1992).

289. S . P . S I M N E R, J . W. S T E V E N S O N, K. D.M E I N H A R D T and N. L . C A N F I E L D , in “SOFC-VII,”edited by H. Yokokawa and S. C. Singhal (Electrochem. Soc.,Pennington, NJ, 2001) Vol. 2001–16, p. 1051.

290. S . P R I M D A H L, M. J . J ØR G E S E N, C . B A G G E R andB. K I N D L , in “SOFC-VI,” edited by S. C. Singhal and M.Dokiya (Electrochem. Soc., Pennington, NJ, 1999) Vol. 99–19,p. 793.

291. R . V A B E N, R . W. S T E I N B R E C H, F . T I E T Z andD. S T OV E R , in “3rd European SOFC Forum,” edited by P.Stevens (European Fuel Cell Group, Lucerne, Switzerland, 1998)p. 557.

292. H . P . B U C H K R E M E R, U. D I E K M A N N, L . G . J . D E

H A A R T, H. K A B S, U. S T I M M I N G and D. S T OV E R , in“SOFC-V,” edited by U. Stimming, S. C. Singhal, H. Tagawa andW. Lehnert (Electrochem. Soc., Pennington, NJ, 1997) Vol. 97–40,p. 160.

293. R . D O N E L S O N, S . A M A R A S I N G H E, D. G O B L E, D.H I C K E Y, S . P . J I A N G, J . L O V E and T . Q U A C H , in “3rdEuropean Solid Oxide Fuel Cells Forum,” edited by P. Stevens(European Fuel Cell Group, Lucerne, Switzerland, 1998) p. 151.

294. G . R I E T V E L D, P . N A M M E N S M A and J . P .O U W E L T J E S , in “SOFC-VII,” edited by H. Yokokawaand S. C. Singhal (Electrochem. Soc., Pennington, NJ, 2001)Vol. 2001–16, p. 125.

295. E . T A N G, F . M A R T E L L, R . B R U L E, K . M A R C O T T E

and B. B O R G L U M , in “SOFC-VIII,” edited by S. C. Singhaland M. Dokiya (Electrochem. Soc., Pennington, NJ, 2003)Vol. 2003–07, p. 935.

296. R . D O S H I , V . L . R I C H A R D S, J . D . C A R T E R, X.W A N G and M. K R U M P E L T , J. Electrochem. Soc. 146 (1999)1273.

297. C . R . X I A and M. L I U , J. Amer. Ceram. Soc. 84 (2001) 1903.298. Y . J . L E N G, S . H . C H A N, S . P . J I A N G and K. A.

K H O R , Solid State Ionics, in press.299. N . Q . M I N H, T . R . A R M S T R O N G, J . R . E S O P A,

J . V . G U I H E E N, C . R . H O M E, F . S . L I U , T . L .S T I L L W A G O N and J . J . V A N A C K E R E N , in “SOFC-II,”edited by F. Gross, P. Zegers, S. C. Singhal and O. Yamamoto(Commission of the European Communities, 1991) p. 93.

300. N . Q . M I N H , in “SOFC-IV,” edited by S. C. Singhal and M.Dokiya (Electrochem. Soc., Pennington, NJ, 1995) Vol. 95–01, p.138.

301. N . Q . M I N H and K. M O N T G O M E R Y , in “SOFC-V,” editedby U. Stimming, S. C. Singhal, H. Tagawa and W. Lehnert(Electrochem. Soc., Pennington, NJ, 1997) Vol. 97–40, p. 153.

302. A . C . M U L L E R, A. K R U G E L and E . I V E R S-T I F F E E ,Mater. Sci. Tech. 33 (2002) 343.

303. J . -H . J E A N and C.-R . C H A N G , J. Amer. Ceram. Soc. 80(1997) 2401.

304. A . C . M U L L E R, D. H E R B S T R I T T , A . W E B E R and E .I V E R S-T I F F E E , in “4th European SOFC Forum,” edited byA. J. McEvoy (European Fuel Cell Group, Lucerne, Switzerland,2000) p. 579.

305. A . C . M U L L E R, A. K R U G E L, A. W E B E R and E .I V E R S-T I F F E E , in “5th European SOFC Forum,” edited by

J. Huijsmans (European Fuel Cell Forum, Lucerne, Switzerland,2002) p. 737.

306. J . W I L L , A. M I T T E R D O R F E R, C . K L E I N L O G E L, D.P E R E D N I S and L . J . G A U C K L E R , Solid State Ionics 131(2000) 79.

307. D . G H O S H, G. W A N G, R. B R U L E, E . T A N G and P .H U A N G , in “SOFC-VI,” edited by S. C. Singhal and M. Dokiya(Electrochem. Soc., Pennington, NJ, 1999) Vol. 99–19, p. 822.

308. P . K . S R I V A S T A V A, T . Q U A C H, Y. Y. D U A N, R.D O N E L S O N, S . P . J I A N G, F . T . C I A C C H I and S . P .S . B A D W A L , Solid State Ionics 99 (1997) 311.

309. Y . J . L E N G, S . H . C H A N, K. A. K H O R, S . P . J I A N G

and P . C H E A N G , J. Power Sources 117 (2003) 26.310. D . P E R E D N I S and L . J . G A U C K L E R , Solid State Ionics

166 (2004) 229.311. T . I S H I H A R A, K. S A T O, Y. M I Z U H A R A and

Y. T A K I T A , Chem. Lett. 1992 (1992) 943.312. J . W I L L , M. K. M. H R U S C H K A, L . G U B L E R and L .

J . G A U C K L E R , J. Amer. Ceram. Soc. 84 (2001) 328.313. P . S A R K A R and H. R A O , in “SOFC-VIII,” edited by S. C.

Singhal and M. Dokiya (Electrochem. Soc., Pennington, NJ,2003) Vol. 2003–07, p. 135.

314. J . -W. K I M, A. V. V I R K A R, K.-Z . F U N G, K. M E H T A

and S . C . S I N G H A L , J. Electrochem. Soc. 146 (1999) 69.315. P . C H A R P E N T I E R, P . F R A G N A U D, D. M. S C H L E I C H

and E . G E H A I N , Solid State Ionics 135 (2000) 373.316. C . X I A and M. L I U , ibid. 144 (2001) 249.317. Y . -S . Y O O, J . -H . K O H, J . -W. P A R K and H. C . L I M ,

in “5th European SOFC Forum,” edited by J. Huijsmans (EuropeanFuel Cell Group, Lucerne, Switzerland, 2002) p. 191.

318. P . H O L T A P P E L S , T . G R A U L E, B . G U T, U. V O G T,L . G A U C K L E R, M. J OR G E R, D. P E R E D N I S , K .H O N E F F E R, G. R O B E R T, S . R A M B E R T and A. J .M C E V O Y , in “SOFC-VIII,” edited by S. C. Singhal and M.Dokiya (Electrochem. Soc., Pennington, NJ, 2003) Vol. 2003–07,p. 1003.

319. R . -H . S O N G, K.-S . S O N G, Y.-E . I H M and H.Y O K O K A W A , in “SOFC-VII,” edited by H. Yokokawa and S. C.Singhal (Electrochem. Soc., Pennington, NJ, 2001) Vol. 2001–16,p. 1073.

320. S . D E S O U Z A, S . J . V I S C O and L . C . D E J O N G H E ,Solid State Ionics 98 (1997) 57.

321. A . V . V I R K A R, J . C H E N, C . W. T A N N E R and J . -W.K I M , ibid. 131 (2000) 189.

322. D . S I M W O N I S , H . T H UL E N, F . J . D I A S , A .N A O U M I D I S and D. S T OV E R , J. Mater. Proc. Tech. 92/93(1999) 107.

323. D . S I M W O N I S , A . N A O U M I D I S , F . J . D I A S , J . L I N K E

and A. M O R O P O U L O U , J. Mater. Res. 12 (1997) 1508.324. B . H O B E I N, F . T I E T Z , D. S T OV E R and E . W.

K R E U T Z , in “Ionic and Mixed Conducting Ceramics IV,” editedby T. A. Ramanarayanan, W. L. Worrell and M. Mogensen(Electrochem. Soc., Pennington, NJ, 2001) Vol. 2001–28, p. 164.

325. E . W A N Z E N B E R G, F . T I E T Z , P . P A N J A N and D.S T OV E R , Solid State Ionics 159 (2003) 1.

326. D . K E K, P . P A N J A N, E . W A N Z E N B E R G and J .J A M N I K , J. Euro. Ceram. Soc. 21 (2001) 1861.

327. M. C A S S I D Y, G. L I N D S A Y and K. K E N D A L L , J. PowerSources 61 (1996) 189.

328. F . M E S C H K E, F . J . D I A S and F . T I E T Z , J. Mater. Sci. 36(2001) 5719.

329. F . T I E T Z , F . J . D I A S , B . D U B I E L and H. J .P E N K A L L A , Mater. Sci. Eng. B B68 (1999) 35.

330. J . Y A N, Y. D O N G, C. Y U and Y. J I A N G , in “SOFC-VII,”edited by H. Yokokawa and S. C. Singhal (Electrochem. Soc.,Pennington, NJ, 2001) Vol. 2001–16, p. 973.

331. J . W. Y A N, Z . G . L U, Y. J I A N G, Y. L . D O N G, Y.Y U and W. Z . L I , J. Electrochem. Soc. 149 (2002) A1132.

332. G . R O B E R T, A. K A I S E R, K. H O N E G G E R and E .B A T A W I , in “5th European SOFC Forum,” edited by J. Huijsmans(European SOFC Forum, Lucerne, Switzerland, 2002) p. 116.

333. N . M. T I K E K A R, T . J . A R M S T R O N G and A. V.V I R K A R , in “SOFC-VII,” edited by S. C. Singhal and M.Dokiya (Electrochem. Soc., Pennington, NJ, 2003) Vol. 2003–07,p. 670.

4438

334. H . M O R I , C . - J . W E N, J . O T O M O, K. E G U C H I andH. T A K A H A S H I , Appl. Catal. A: General 245 (2003) 79.

335. G . S C H I L L E R, R . H. H E N N E, M. L A N G, R.R U C K D AS C H E L and S . S C H A P E R , Fuel Cells Bulletin 3(2000) 7.

336. M. L A N G, R. H E N N E, S . S C H A P E R and G. S C H I L L E R ,J. Thermal Spray Techn. 10 (2001) 618.

337. T . F R A N C O, R. H E N N E, M. L A N G, G. S C H I L L E R

and P . S Z A B O , in “5th European SOFC Forum,” edited by J.Huijsmans (European Fuel Cell Group, Lucerne, Switzerland,2002) p. 647.

338. S . J . V I S C O, C . P . J A C O B S O N, L . C . D E J O N G H E,A. L E M I N G, Y. M A T U S, L . Y A N G, I . V I L L A R E A L

and L . R O D R I Q U E Z-M A R T I N E Z , in “Ionic and MixedConducting Ceramics IV,” edited by T. A. Ramanarayanan, W. L.Worrell and M. Mogensen (Electrochem. Soc., Pennington, NJ,2001) Vol. 2001–28, p. 368.

339. S . J . V I S C O, C . P . J A C O B S O N, I . V I L L A R E A L,A. L E M I N G, Y. M A T U S and L . C . D E J O N G H E , in“SOFC-VII,” edited by S. C. Singhal and M. Dokiya (Electrochem.Soc., Pennington, NJ, 2003) Vol. 2003–07, p. 1040.

340. G . S C H I L L E R, T . F R A N C O, R. H E N N E, M. L A N G,R. R U C K D AS C H E L, P . O T S C H I K and K. E I C H L E R ,in “SOFC-VII,” edited by H. Yokokawa and S. C. Singhal(Electrochem. Soc., Pennington, NJ, 2001) Vol. 2001–16, p. 885.

341. M. M O G E N S E N and K. K A M M E R , Annu. Rev. Mater. Res.33 (2003) 321.

342. S . P . J I A N G, J . P . Z H A N G, L . A P A T E A N U and K.F O G E R , J. Electrochem. Soc. 147 (2000) 4013.

343. S . P . J I A N G, J . P . Z H A N G and X. G. Z H E N G , J.European Ceramic Soc. 22 (2002) 361.

344. W. J . Q U A D A K K E R S, H. G R E I N E R, M. H AN S E L,A. P A T T A N A I K, A. S . K H A N N A and W. M A L L EN E R ,Solid State Ionics 91 (1996) 55.

345. I . C . V I N K E, L . G . J . D E H A A R T, L . B L U M andD. S T O L T E N , in “5th European SOFC Forum,” edited by J.Huijsmans (European Fuel Cell Group, Lucerne, Switzerland,2002) p. 164.

346. T . Y A M A M O T O, H. I T O H, M. M O R I , T . W A T A N A B E,N. I M A N I S H I , Y . T A K E D A and O. Y A M A M O T O , J.Power Sources 61 (1996) 219.

347. A . K U Z J U K E V I C S and S . L I N D E R O T H , Solid State Ionics93 (1997) 255.

348. S . L I N D E R O T H, N. B O N A N O S, K. V. J E N S E N andJ . B . B I L D E-S ØR E N S E N , J. Amer. Ceram. Soc. 84 (2001)2652.

349. K . V . J E N S E N, S . P R I M D A H L, I . C H O R K E N D O R F F

and M. M O G E N S E N , Solid State Ionics 144 (2001) 197.350. M. M O G E N S E N, K. V. J E N S E N, M. J . J Ø R G E N S E N

and S . P R I M D A H L , ibid. 150 (2002) 123.351. M. C . W I L L I A M S and J . P . S T R A K E Y , in “SOFC-VII,”

edited by S. C. Singhal and M. Dokiya (Electrochem. Soc.,Pennington, NJ, 2003) Vol. 2003–07, p. 3.

352. R . N . B A S U, G. B L Aβ, H . P . B U C H K R E M E R, D.S T OV E R, F . T I E T Z , E . W E S S E L and I . C . V I N K E ,

in “SOFC-VII,” edited by H. Yokokawa and S. C. Singhal(Electrochem. Soc., Pennington, NJ, 2001) Vol. 2001–16,p. 995.

353. M. D O K I Y A , Solid State Ionics 152/153 (2002) 383.354. F . G . E . J O N E S and J . T . S . I R V I N E , in “5th European

SOFC Forum,” edited by J. Huijsmans (European Fuel Cell Group,Lucerne, Switzerland, 2002) p. 123.

355. F . J . G A R D N E R, M. J . D A Y, N. P . B R A N D O N, M.N. P A S H L E Y and M. C A S S I D Y , J. Power Sources 86 (2000)122.

356. G . D . A G N E W et al., in “SOFC-VII,” edited by S. C. Singhaland M. Dokiya (Electrochem. Soc., Pennington, NJ, 2003)Vol. 2003–07, p. 78.

357. K . W I P P E R M A N N , in Proc. of IEA Workshop—Materialsand Mechanisms, edited by K. Nisanciogly (International EnergyAgency, Paris, 1999) p. 31.

358. S . P R I M D A H L and M. M O G E N S E N , J. Electrochem. Soc.145 (1998) 2431.

359. S . P . J I A N G and S . P . S . B A D W A L , Solid State Ionics 123(1999) 209.

360. A . B I E B E R L E, L . P . M E I E R and L . J . G A U C K L E R , J.Electrochem. Soc. 148 (2001) A646.

361. A . B I E B E R L E and L . J . G A U C K L E R , Solid State Ionics146 (2002) 23.

362. M. I H A R A, T . K U S A N O and C. Y O K O Y A M A , J.Electrochem. Soc. 148 (2001) A209.

363. C . W E N, R . K A T O, H. F U K U N A G A, H. I S H I T A N I

and K. Y A M A D A , ibid. 147 (2000) 2076.364. P . H O L T A P P E L S , L . G . D E H A A R T and U. S T I M M I N G ,

ibid. 146 (1999) 1620.365. N . N A K A G A W A, K. N A K A J I M A, M. S A T O and K.

K A T O , ibid. 146 (1999) 1290.366. S . H . C H A N, X. J . C H E N and K. A. K H O R , J. Appl.

Electrochem. 31 (2001) 1163.367. J . W I N K L E R, P . V . H E N D R I K S E N, N. B O N A N O S and

M. M O G E N S E N , J. Electrochem. Soc. 145 (1998) 1184.368. S . B . A D L E R , ibid. 149 (2002) E166.369. S . P . J I A N G , J. Appl. Electrochem., in press.370. Y . J . L E N G, S . H . C H A N, K. A. K H O R and S . P .

J I A N G , ibid. 34 (2004) 409.371. S . M C I N T O S H, J . M. V O H S and R. J . G O R T E , J.

Electrochem. Soc 150 (2003) A1305.372. B . W. C H U N G, A.-Q. P H A M, J . J . H A S L A M and R.

S . G L A S S , ibid. 149 (2002) A325.373. R . I H R I N G E R, S . R A M B E R T, L . C O N S T A N T I N and J .

V A N H E R L E , in “SOFC-VII,” edited by H. Yokokawa and S. C.Singhal (Electrochem. Soc., Pennington, NJ, 2001) Vol. 2001–16,p. 1002.

374. A . V . V I R K A R, J . C H E N, C . W. T A N N E R and J . -W.K I M , Solid State Ionics 131 (2000) 189.

375. Y . J I A N G and A. V. V I R K A R , J. Electrochem. Soc. 150(2003) A942.

Received 4 November 2003and accepted 24 February 2004

4439