Solid Oxide Fuel Cells: Systems and Materialsgases react. The generated heat serves to preheat the...

FUEL CELL RESEARCH IN SWITZERLAND 837CHIMIA 2004 58 No 12

Chimia 58 (2004) 837ndash850copy Schweizerische Chemische Gesellschaft

ISSN 0009ndash4293

Solid Oxide Fuel Cells Systems and Materials

Ludwig J Gauckler Daniel Beckel Brandon E Buergler Eva Jud Ulrich P Muecke Michel PrestatJennifer LM Rupp and Joumlrg Richter

Abstract A solid oxide fuel cell (SOFC) is a solid-state energy conversion system that converts chemical energy in-to electrical energy and heat at elevated temperatures Its bipolar cells are electrochemical devices with an anodeelectrolyte and cathode that can be arranged in a planar or tubular design with separated gas chambers for fueland oxidant Single chamber setups have bipolar cells with reaction selective electrodes and no separation be-tween anode and cathode compartments A nickelyttria-stabilized-zirconia (YSZ) cermet is the most investigatedand currently most widespread anode material for the use with hydrogen as fuel In recent years however dopedceria cermet anodes with nickel or copper and ceria as the ceramic phase have been introduced together with ce-ria as electrolyte material for the use with hydrocarbon fuels The state-of-the-art electrolyte material is YSZ of highionic and nearly no electronic conductivity at temperatures between 800ndash1000 degC In order to reduce SOFC sys-tem costs a reduction of operation temperatures to 600ndash800 degC is desirable and electrolytes with higher ionic con-ductivities than YSZ are aimed for such as bismuth oxide lanthanum gallate or mixed conducting ceria and the useof thin electrolytes Proton conducting perovskites are researched as alternatives to conventional oxygen con-ducting electrolyte materials At the cathode the reduction of molecular oxygen takes place predominantly on thesurface Todayrsquos state-of-the-art cathodes are LaxSr1ndashxMnO3ndashd for SOFC operating at high temperature ie800ndash1000 degC or mixed conducting LaxSr1ndashxCoyFe1ndashyO3ndashd for intermediate temperature operation ie 600-800 degCAmong the variety of alternative materials SmxSr1ndashxCoO3ndashd and BaxSr1ndashxCoxFe1ndashxO3ndashd are perovskites that showvery good oxygen reduction properties This paper reviews the materials that are used in solid oxide fuel cells andtheir properties as well as novel materials that are potentially applied in the near future The possible designs ofsingle bipolar cells are also reviewed

Keywords Anode middot Cathode middot Electrolyte middot Materials middot SOFC

1 Introduction

Fuel cells are one of the most attractive en-ergy conversion systems because they offerhigh efficiency and low pollution An ad-vantage of fuel cells is the decentralisedgeneration of electricity and the prospectiveapplications in mobile devices

Correspondence Prof Dr LJ GaucklerNonmetallic Inorganic MaterialsDepartment MaterialsSwiss Federal Institute of TechnologyWolfgang-Pauli-Str10CH-8093 ZurichTel +41 1 632 56 46Fax +41 1 632 11 32E-Mail ludwiggaucklermatethzch

A fuel cell is an electrochemical devicethat can convert chemical energy of a fueland an oxidant into heat and electric powerIn solid oxide fuel cells (SOFCs) the elec-trolyte consists of an oxygen ion conduct-ing ceramic such as yttria-stabilized zirco-nia (YSZ) One advantage of SOFCs is theirpossibility to directly use natural gas andthe high reaction rate given by the high op-erating temperature Thus no expensivecatalysts are needed However the ceramicmaterials of the SOFC are difficult toprocess and not easy to assemble

A schematic of a SOFC is shown in Fig1 A single cell consists of three basic ele-ments electrolyte anode and cathode Thecathode and anode electrodes are porouslayered ceramic (cathode) and ceramic-metal (anode) components enabling easygas diffusion to and from theelectrodendashelectrolyte interfaces They ex-hibit high electronic conduction and prefer-ably also ionic conduction The reductionand oxidation at the cathode and the anoderespectively are spatially separated and the

electrons are forced to flow through an ex-ternal circuit At the electrodes the chargecarrying species is changed from electronsfrom the outer circuit to the charged speciesthe electrolyte can conduct In the SOFCthe electrolyte conducts O2ndash-ions The driv-ing force for the migration of O2ndash is theoxygen chemical potential gradient be-tween the anode (low) and cathode (high)At the cathode side air is usually used cor-responding to an oxygen partial pressure(pO2) of 021 atm At the anode the pO2 isvery low due to the consumption of oxygenions by the used fuel (in most cases hydro-gen) to form water The operating tempera-ture of a SOFC is between 500 and 1000 degCbecause the conduction of oxygen ions inthe solid electrolyte is a thermally activatedprocess In contrast to other fuel cell typesa solid oxide fuel cell can be operated witha variety of fuels such as CH4 with steamreforming and within a wide temperaturerange (500ndash1000 degC)

At the cathode electrochemical reduc-tion of oxygen occurs and the oxygen ions


migrate through the electrolyte via a vacan-cy mechanism to the anode At the anodehydrogen is electrochemically oxidized towater Each cell delivers a maximum of 1 Vand is typically operated at around 06 to07 V at a power output of typically 250 to450 mWcm2 In SOFC systems many cellsare stacked in series connected with ametallic conducting interconnect

Research and development in the fieldof SOFC are currently concentrating onlowering the operating temperature in orderto reduce costs and increase lifetime and toincrease reliability of the ceramic stack el-ements and interconnects New manufac-turing technologies are demanded whenthin electrolytes are used reducing the elec-trical resistance of the cell Materials aswell as systems development aim also forbetter fuel utilization and higher electricalefficiency

This paper reviews the different possi-ble designs of SOFC cells including thepossibilities when reaction selective elec-trodes are used Commonly used materialsand some of their properties as well as nov-el materials that could be applied in near fu-ture are also reviewed as well as a searchstrategy for those materials

2 Design

The design of a single cell is closely re-lated to the design of an entire stack Be-cause a single cell only delivers 1 V morethan one cell is usually connected in seriesusing interconnects The open circuit volt-age (OCV) of the SOFC ie the voltage ofthe system when no current is flowing cor-responds to the number of individual cellsin the stack Over the last two decadesSOFCs based on yttria-stabilised zirconiahave been developed for an operating tem-perature range of 900ndash1000 degC The advan-tage of the high temperature is that internal

reforming of hydrocarbons is possible di-rectly on the anode without the need for anexternal reformer [1]

One important design criterion for a sol-id oxide fuel cell is the separation of anodeand cathode by the gas tight electrolytePinholes or cracks in the electrolyte cancause the hydrogen to leak to the cathodecompartment where it reacts directly withoxygen This will decrease the open circuitvoltage (OCV) and might even render thefuel cell inoperable The development of asuitable stack sealant still presents a chal-lenging task because the requirements forthe sealants are stringent due to harsh envi-ronments and the high operating tempera-tures Sealing of SOFCs can be done by us-ing bonding seals or pressurized seals Forbonding seals materials like high-B2O3glasses [2] earth-alkali silicate glasses suchas BaOAl2O3SiO2 [3] or glass ceramicsare commonly used Another solution re-lies on compressive seals based on micathat do not bond chemically to the SOFCmaterials [4]

21 Tubular DesignIn the 1960s experimental SOFCs with

planar geometry were evaluated and it wasfound that it is very difficult to obtain ade-quate gas sealing at the edges of the cell

mainly due to the mismatch of thermal ex-pansion coefficient between the electrolytesand support structures and the mechanicalproperties of the sealing materials In orderto overcome these problems a tubular con-figuration (ie cylindrical design) was de-veloped by Westinghouse and taken over bySiemens and improved over the last 20years In this design (Fig 2) the electrolyteand anode are supported on a thick cathodetube that is closed at one end The materialstheir dimensions and fabrication processesare summarized in Table 1 The electrolyteis deposited onto the cathode support afterfabrication of the interconnection In a laststep the anode is applied The gas manifoldof the Siemens-Westinghouse design is il-lustrated in Fig 3 Air is introduced via acentral Al2O3-tube to the end of the cathodetube The oxidant flows back across thecathode while the fuel flows in the same di-rection at the exterior of the tube At theplenum of each cell the depleted flow of airand fuel recombine and the remaining activegases react The generated heat serves topreheat the incoming oxidant stream One ofthe most attractive features of this fuel celldesign is that it eliminates the need for leak-free gas manifolding of the fuel and oxidantstreams in the hot zone The drawback isthat the electric current has to flow along thecircumference of the tube in the anode andthe cathode This increases the length of theconducting path and thus the ohmic resist-ance of the cell as compared to a planar one

Fig 1 Schematic of a solid oxide fuel cell (SOFC) element with anode cathode and electrolyte

Table 1 Materials and fabrication processes for state-of-the-art cathode supported cells of theSiemens-Westinghouse solid oxide fuel cell

Component Material Thickness Fabrication Process

Cathode Tube Doped LaMnO3 22 mm Extrusion-sintering

Electrolyte ZrO2(Y2O3) 40 mm Electrochemical vapour deposition

Interconnect Doped LaCrO3 85 mm Plasma spraying

Anode Ni-ZrO2(Y2O3) 100 mm Slurry spraying or electro-chemical vapour deposition

Fig 2 The tubular design from Siemens-West-inghouse


Currently the life of a fuel cell is in the or-der of 3000ndash7000 h and needs to be im-proved by optimizing the mechanical aswell as electrochemical stability of the usedmaterials [9]

23 Single Chamber DesignConventional fuel cells rely on the strict

separation of fuel and oxidant by the elec-trolyte membrane and seals By separatingthe fuel and oxidant direct parasitic chem-ical reactions of fuel and oxidant are avoid-ed However it has been shown that it is notmandatory to separate the fuel and the oxi-dant for operating a fuel cell By using re-action-selective electrodes a fuel cell can beoperated in a single gas chamber fed by amixture of fuel and air Such a cell is oftenreferred to as Mixed Gas Fuel Cell or Sin-gle Chamber SOFC (SC-SOFC)

Already in 1965 van Gool proposed adevice using lsquosurface migrationrsquo of an inertsubstrate with two different electrodes [10]The electronically insulating substrateshould permit easy surface transportation ofat least one of the reactants in ionic formThe electrodes are placed on the same sideof this substrate and have different catalyt-

Fig 3 Schematic view of gas flows in theSiemens-Westinghouse SOFC design

Fig 4 Flat tube design by Siemens-Westing-house

Table 2 Materials and fabrication processes of the components for the electrolyte supported SulzerHexis solid oxide fuel cell


Cathode LaSrMnO3 (LSM) 20ndash100 mm Screen printing

Electrolyte ZrO2(Y2O3) (TZPFSZ) 150ndash250 mm Tape casting

Interconnect CrFe5Y2O3 Powder metallurgy

Anode Ni-ZrO2(Y2O3) 20ndash100 mm Screen printing

paths which in turn decrease the ohmic re-sistance of each cell and increase the powerdensity of cell stacks The Siemens-West-inghouse power systems are well estab-lished and development has shifted frombasic technology to cost reduction and scaleup [6]

22 Planar DesignA planar design of the bipolar plates en-

ables the electrical connection of cells in se-ries to be simplified without long currentpaths Another advantage of the planar de-sign is that low-cost fabrication methodssuch as screen-printing and tape casting canbe used However because of thermalstresses the size of the cells was limited inthe past Today 10yen10 cm2 planar cells canroutinely be produced and operated [7]Sulzer Hexis aims at building systems forthe cogeneration of electricity and heat forresidential applications in the 1 kW powerregime with cells of planar design [8] Asingle cell with endplate (top) and intercon-nect (bottom) is shown in Fig 5 The fuel isfed into the centre of a cylindrical stackconsisting of layered circular cells Each in-terconnect serves as gas manifold and en-sures that the reactant air is preheated It ismade via powder metallurgy of oxide dis-persion strengthened alloy (95 Cr and 5Fe) with 1 Y2O3 The materials thick-nesses and fabrication processes of eachcomponent are given in Table 2 A crucialpoint is the metalceramic contact betweenthe electrodes and interconnects which ismade at the cathode side by applying a LSCslurry to the pins of the interconnect and aNi gauze at the anode side To the exteriorthe cell is not sealed and the unreacted fuelis burnt with the unreacted oxygen from air

Fig 5 Ring-type solid oxide fuel cell with metal-lic interconnect from Sulzer Hexis

Fig 6 The first single chamber fuel cell proposed by Dyer in 1965 [11]

Siemens-Westinghouse has been work-ing on this problem and has come up with anew design that is called the high-powerdensity SOFC (HPD-SOFC) [5] In this de-sign shown in Fig 4 a flat cathode tubewith ligaments is used instead of a cylindri-cal one It allows easier manifolding of airinside the tube and higher packing densityof cells as compared to the cylindrical con-figuration This leads to higher volumetricpower densities of a complete cell stackMost important is that the bridges withinthe cathode tube allow for shorter current


ic properties One is active for the reductionof oxygen and the other for the activation ofthe fuel ie adsorption and dissociation ofhydrogen from a mixture of hydrogen andair Van Gool suggested the use of gold orsilver as the cathode material (stable oxideunstable hydride) and platinum palladiumor iridium as the anode (stable hydride un-stable oxide) In 1990 Dyer was able to gen-erate electrical power from a device withelectrodes made of platinum separated by athin ion conducting and porous film [11]Fig 6 schematically shows the design of theelectrochemical device and the used mate-rials A voltage of approximately 1 V wasachieved at room temperature on a mixtureof hydrogen and air The achieved powerdensity was in the range of 1 to 5 mWcm2

Hibino and Iwahara have been workingon SC-SOFCs in recent years The firstcells had similar power densities to the cellsdescribed by Dyer ie in the range of 2ndash5mWcm2 [12] With very similar materialsGoumldickemeier et al proved the feasibilityof connecting individual cells on one elec-trolyte plate in series without the need forhaving sealed gas compartments for eachcell [13] Thus with one element consistingof series connected cells it is possible to ob-tain useful voltages higher than only 1 V

Hibino et al also used alternative elec-trolyte materials eg La09Sr01Ga08Mg02O3ndashd (LSGM) which showed better per-formance than YSZ [14] This was mainlydue to the higher ionic conductivity of theutilized materials Ceria (CeO2) based SC-SOFCs showed maximum power densitiesof 644 mWcm2 at 550 degC and 269 mWcm2

at 450 degC [15] with a fuel utilization thatwas estimated to be around 10

An advantage of the Single Chamberapproach is that completely new designscan be envisaged such as illustrated in Fig7 For research and development the classicdesign (a) appears to be most feasible be-cause of simple geometry and easy fabrica-tion procedures The lsquoside by sidersquo designshown in (b) allows easy interconnection ofcells located on the same side of an elec-trolyte substrate Very thin layers of activecomponents can be used and this reducesthe material costs as well as increases thespecific power density [14] The feasibilityof the side by side design and the optimumgeometry have recently been evaluated forthe case of mixed reactant direct methanolfuel cells [16] In the case of SC-SOFCs gasleaks in the electrolyte are of no concernThe fully porous design shown in Fig 7(c)makes use of the absent constriction of agas tight electrolyte The concept of fullyporous fuel cells has been proposed for di-rect methanol fuel cells [17] and can easilybe adopted for SOFCs

3 Electrolyte

31 Oxygen Ion ConductingElectrolytes

Solid oxide fuel cell (SOFC) electrolytematerials should have high ionic conductiv-ity and low electronic conductivity Theavailable electrolyte materials differ main-ly in the nature of their conductivity eitherhaving purely ionic or mixed ionic elec-tronic conductivity (MIEC) The ionic con-ductivity of an electrolyte can be enhancedby introducing acceptor dopants and conse-quently oxygen vacancies [18] YSZ is the

state-of-the-art electrolyte for SOFCsThese solid solutions are primarily ionicconductors and show nearly no electronicconductivity They have to be operated athigh temperatures around 800ndash1000 degC[19]

The amount of oxygen vacancies andconsequently ionic conductivity is in-creased by the introduction of the trivalentyttria dopants into the zirconia lattice[20ndash22] This stabilizes the cubic phase atY2O3 contents of 8 mol The tetragonalform (3mol Y2O3) shows time-depend-ent degradation [21][23ndash25] because wa-ter is produced at the anode which leads tohydrothermally assisted transformation ofthe tetragonal to the monoclinic phase[26]

On the cathode side YSZ is in contactwith LaCoO3 or LaMnO3 based cathodematerials At high operating temperaturesof 800 to 1000 degC both materials reactforming insulating La7Zr2O7 which leads toa gradual increase of cathode overpotential[27ndash29] It has been recognized that forsmaller SOFC stacks the operating temper-ature should be lowered without increasingthe internal resistance of the cell [30ndash33] Inthe following alternative materials to state-of-the-art YSZ such as scandia-doped zir-conia doped ceria solid solutions bismuth-based oxides or lanthanum gallate basedelectrolytes are discussed [34][35] In Fig8 the ionic conductivity of these electrolytematerials are plotted as a function of tem-perature [36][37] It has been well knownsince the 1970s that Sc-stabilized zirconia(ScSZ) shows the highest ionic conductivi-ty of all zirconia solid solutions The reasonfor this is the smallest tendency for vacan-cy cluster formation with increasing dopantconcentration due to the close match of theSc3+ ionic radius with the Zr4+ host cation[38ndash40] However Sc-doped zirconia be-comes unstable especially at intermediatetemperatures [41] Politova and Irvine re-cently investigated the possibility of ScSZstabilization by yttria doping Small addi-tions of yttria considerably stabilized thecubic phase of ScSZ at the prospective fuelcell operating temperature However it wasnot possible to overcome the time-depend-ent degradation of the conductivity duringlong annealing periods [42]

Ceria (CeO2) based electrolytes offer anionic conductivity up to 4ndash5 times higherthan that of zirconia solid solutions in theintermediate and low temperature regime[43] Doping of ceria with eg Gd2O3Y2O3 CaO or Sm2O3 introduces oxygenvacancies and induces ionic conductivity[44] The development of these materialsfor intermediate temperature SOFCs hasbeen extensively reviewed by Steele [45]Sm2O3 doped ceria (CSO) and Gd2O3doped ceria (CGO) exhibit the highest con-ductivities of all rare earth doped CeO2 sol-

Fig 7 Possible designs for SC-SOFCs a) classic sandwich design b) side by side c) fully porous


id solutions [46] Again it is assumed thatthis is due to the ionic radii of Sm3+ andGd3+ which nearly match the ionic radius ofCe4+ [46ndash49] Furthermore these com-pounds show the lowest electronic conduc-tion at low oxygen partial pressures At 700degC the conductivity of CGO and CSO (bothwith 10ndash25 dopant) come close to theconductivity of YSZ at 1000 degC [50][51] Amonotonic increase of ionic conductivity isobserved with increasing Sm2O3 or Gd2O3content until a maximum is reached Theoxygen vacancies then begin to form defectclusters with the doped cations (egSmrsquoCeVOuml) which will decrease the mobilityof the oxygen vacancies [52] As ceria be-comes reduced under low oxygen partialpressures at the anode-electrolyte interfacethe material exhibits n-type electronic con-ductivity [53] especially at higher operationtemperatures Therefore ceria solid solu-tions are recommended for operation tem-peratures below 800 degC where excellentSOFC performance can be obtained [54] orin combination with YSZ layers blockingelectronic conduction

In contrast to zirconia-based elec-trolytes ceria solid solutions exhibit lowercathode-electrolyte overpotentials [55ndash57]Doshi et al measured a high power outputat 500 degC of a fuel cell with CGO elec-trolyte lanthanum cobalt based cathodeand a Ni-CGO anode [54] CGO elec-

trolytes are superior to YSZ for low tem-perature SOFCs because at low tempera-tures CGO behaves as a pure ionic conduc-tor with much higher ionic conductivitySeveral authors proposed doped ceria elec-trolytes for intermediate and low tempera-ture fuel cell operation [19][53][54][58]

Dikmen et al investigated the influenceof high ionic conductive bismuth oxide as adopant in ceria The authors report higherionic conductivities due to the bismuth ox-ide doping compared to gadolinia doping ofceria However it remains unclear howchemically stable this electrolyte is to re-ducing atmospheres and phase transitionsof bismuth oxide [59]

The highest ionic conductivities at300ndash700 degC are found in Bi2O3-based elec-trolytes like BIMEVOX (Fig 8) [60][61]For temperatures as low as 300 degCBIMEVOX electrolytes show conductivi-ties as high as YSZ at 800 degC [62]BIMEVOX are bismuth vanadium oxidesBi4V2O11 where the vanadium is partiallysubstituted to yield Bi2V1ndashxMexOy solid so-lutions [63] The BIMEVOX family of ma-terials exhibits specific properties as elec-trolytes as well as oxygen electrodes Thehigh oxide anion diffusion observed atmoderate temperature results from the syn-ergy between the highly polarisable ion pairof the BiIII cation in the vicinity of the V-Odiffusion slab on the one hand and the flex-

ibility of this V-O network on the otherhand Moreover the same material is ableunder imposed polarisation to self convertreversibly and dynamically from elec-trolyte to electrode All these specific char-acteristics led to a new concept of ceramicoxygen generator based on a unique mate-rial [64] However the main drawback ofBIMEVOX electrolytes is that they slowlydecompose at SOFC operating tempera-tures Reviews on stability and ionic con-ductivity of Bi2O3-based electrolytes aregiven by Shuk et al [37] and Sammes et al[65]

Doped lanthanum gallates (LaGaO3)are currently attracting considerable atten-tion as promising electrolytes for inter-mediate temperature SOFC applicationsWhen the trivalent lanthanum and galliumare doped with divalent cations like Sr andMg forming La1ndashxSrxGa1ndashyMgyO3ndashx2ndashy2(LSGM) the ionic conductivity is signifi-cantly higher than that of YSZ but still low-er than that of CGO [66] The stabilityseems to be higher than that of CGO andthus it seems attractive to use LSGM elec-trolytes at temperatures of 600ndash800 degC[67] However it is difficult to produce sin-gle phase LSGM since secondary phasessuch as La4Ga2O9 and SrLaGa3O7 prevailat grain boundaries reducing the conductiv-ity [68] Furthermore Weitkamp and co-workers report a limited stability of LSGMunder reducing and oxidizing conditionsfollowed by the development of n-type con-ductivity at low and p-type at high oxygenpartial pressures [69]

Increased power densities andor re-duced operation temperatures can also beachieved with reducing the thickness of theelectrolytes and thereby reducing the ohmiclosses In many concepts of flat bipolarcells the electrolyte thickness is in the or-der of 100 to 300 mm and serves also as thestructural load bearing component Whenreducing the thickness of the electrolyte tothe range of mm or even to several hundrednanometres the anode or the cathode isused as support structure Good power den-sities in SOFCs have been obtained withthin YSZ electrolytes prepared by colloidalmethods by Will et al [70] Electrophoret-ic deposition of fine YSZ particles dis-persed and stabilized in water was used toproduce 20 mm thin electrolytes that result-ed in power densities of more than 200mWcm2 at reduced operating temperaturesof 700 degC

Other methods have been reported con-cerning the development of thin-filmprocesses for SOFC applications such aselectrochemical vapour deposition [71]plasma spraying [72] physical vapour dep-osition [73] and pyrolysis of dip coated orsprayed metal salt solutions [74][75]

Although some of these physical andchemical methods produce dense layersFig 8 Ionic conductivities of different electrolyte materials [36][37]

T [degC]

1000T [K]

log

s[W

-1cm

-1]


they are less suitable for mass productionexcept spray deposition Perednis et al ob-tained more than 600 mWcm2 at 700 degCwith anode supported cells with bi- and tri-layer electrolytes as thin as 300 nm basedon ceria as shown in Fig 9 [76][77]

Bilayer electrolytes can combine advan-tages of two electrolytes In case of a ce-riazirconia based bilayer ceria is used atthe cathode side being in thermodynamicequilibrium with lanthanum strontium ironperovskite avoiding the La7Zr2O7 forma-tion which degrades the cell when zirconiais combined with these cathodes On theother side when using zirconia on the an-ode side the ceria-based electrolyte is pro-tected against reduction and electronic con-ductivity is avoided in the electrolyte[77ndash81] The different electrolyte materialssuitable for SOFCs have been extensivelyreviewed elsewhere [1][18][19][31][34][67][68][82ndash84]

32 Proton Conducting ElectrolytesVarious ceramic materials exhibit pro-

tonic conductivity at moderate temperaturesBy replacing the oxygen ion conductiveelectrolyte in a SOFC with a proton conduc-tor several improvements regarding the fuelcell performance can be envisaged The firststudies in the field of protonic conductivityand its application to SOFCs were conduct-ed by Iwahara et al for SrCeO3-based mate-rials [85] The highest proton conductivitieshave been reported for perovskites (ABO3)such as BaCeO3-based materials [86ndash90]Proton conductivity is achieved by the partialsubstitution of the B site cation with an ac-ceptor dopant ion which is charge compen-sated by oxygen vacancies Trivalentdopants have been demonstrated to be moreeffective than bivalent ones due to their high-er protonic defect concentration and mobili-ty [91] Most BaCeO3-based materials dis-play protonic conduction at intermediatetemperatures and become oxygen ion con-ductors at higher temperatures see eg [87]The atmosphere can also influence the con-duction mechanism Typical conductivitiesare between 01 to 0001 Scm for tempera-tures from 1000 to 600 degC [90] A compari-son of the proton conductivities for variousoxides is given elsewhere [92] BaCeO3-based materials possess the highest molarvolume and the deviation from the ideal cu-bic perovskite structure is small [93] Theseproperties are assumed to be necessary pre-requisites for a material to exhibit high pro-tonic conductivity [92] However these ma-terials usually lack sufficient thermodynam-ic stability Cerates for example formcarbonates in air [94] as well as in CO2-con-taining atmospheres [95] A number of in-vestigations have therefore been conductedin the last years with the aim to combine highproton conductivity with improved thermo-dynamic stability

Zirconates such as Y-doped BaZrO3offer high proton conductivity with the nec-essary thermodynamic stability for fuel cellapplications [94][96] By doping BaZrO3with 15ndash20 mol of yttrium proton con-ductivities were found to be higher than theconductivities of the best oxygen ionic con-ductors [97] Even for high dopant levelsthe proton mobility is not changed makingY-doped BaZrO3 a suitable candidate aselectrolyte material [92] Appreciable pro-ton conduction in hydrogen containing at-mospheres and p-type conductivity for highoxygen partial pressures have also been in-vestigated for divalent doped scandates likeLaSc1ndashxMgxO3ndashd [98] Acceptor-doped Sr-TiO3 also showed protonic conductivitycombined with a high thermodynamic sta-bility although the protonic defect forma-tion is less favoured compared to acceptor-doped BaZrO3 [97]

In order to form proton defects watervapour is incorporated into the crystal lat-tice of the proton conductor according toEqn 1

The positively charged protonic defectforms a covalent bond with oxygen of thelattice If the concentration of protonatedoxygen atoms is sufficiently high a proton-ic current flows across the electrolyte Theprotons are then supplied on the anode side

After crossing the electrolyte the de-fects are removed by

The diffusion of the protonic defectsacross the electrolyte material requires acounter flux of oxygen vacancies in order tomaintain charge neutrality This counterdif-fusion represents one of the main advan-tages of proton conductors for fuel cells theambipolar steam permeation [99] Since theincorporation of water vapour according toEqn (1) is reversible and independent of re-actions (2) and (3) proton conduction willtake place due to any steam concentrationgradient Typical values of activation ener-gies for proton conduction are around 05eV [100] If an external load is applied hy-drogen will be incorporated into the elec-trolyte according to Eqn (2) and steam willbe produced on the cathode side accordingto Eqn (3) as shown in Fig 10 Conse-quently the steam partial pressure will in-crease on the cathode side so that some ofthe steam will react according to Eqn (1)and return back to the anode Therefore theFaradaic current of the cell is independentof the steam permeation and only dependson the concentration and mobility of theprotonic defects [99] If the cell is operatedwith hydrocarbons coking cannot takeplace at the anode side as long as the diffu-sion of water through the electrolyte keepsup with the adsorption and decompositionof the fuel Furthermore water vapour isproduced at the cathode side and thus can-not dilute the fuel [99]

Typically achieved maximum poweroutputs of cerate- as well zirconate-basedcells are around 20 mWcm2 [92][99] Fur-ther research on proton conducting materi-als is therefore needed to make proton con-ductor based cells to serious competitorsfor fuel cells based on oxygen conductors

4 Anode

The main functionality of a SOFC an-ode is to provide electrochemically active

Fig 9 Thin-film SOFC with bi-layer YSZCGO electrolyte and power output at 620 and 720 degC [76][77]

(1)

(2)

(3)


reaction sites for the oxidation of the fuelgas molecules and to transport electronsfrom the oxidation reaction to connectingcell components Many factors determinethe materials choice for the anode Anodesprovide pathways for the fuel to reach thereaction sites and for the reactants to diffuseaway from the reaction sites They also re-quire a high electronic conductivity for cur-rent transport and should be chemicallycompatible to adjacent cell componentssuch as the electrolyte current collectorand structural elements Specifically whenused in anode supported fuel cells they alsohave to be structurally stable over an ade-quate lifetime

In the early development of SOFC no-ble metals such as ruthenium rhodium pal-ladium silver platinum and gold and fromthe transition metal group manganese ironcobalt nickel and copper were considered[101] Platinum is a good electrocatalyst al-though the high vapour pressure of plat-inum sub-oxides prevents its use in SOFCoperating between 900 and 1000 degC Goldshows almost no catalytic activity and pooradhesion to oxides From the transitionmetal oxides nickel proved to be the bestchoice in terms of catalytic activity and re-dox stability However the pure metal has astrong tendency towards grain growth at el-evated temperatures and a significantly dif-ferent thermal expansion coefficient thancommonly used electrolyte materials

Therefore nickel is combined with a ce-ramic compound such as zirconia or ceriaforming three interconnected frameworksof metal ceramic and pores This cermetbecomes a good metallic conductor fornickel contents above the percolationthreshold In the past research has beenmainly focused on yttria-stabilized zirconia(YSZ) as ceramic material for electrolytesand in cermets for anodes for its good struc-tural stability good electrical conductivityat high temperatures and stability under allatmospheric conditions

In a purely ionic conductor like YSZ theoxidation of the fuel gas with oxygen ionscoming from the cathode side through theelectrolyte is believed to occur only in thetriple phase boundary (tpb) the connectingpoints of metal ceramic and pore The ce-ramic network not only provides structuralintegrity and hinders the trapped nickel par-ticles from excessive grain growth but alsoprovides a pathway for oxygen ions effec-tively extending the triple phase boundaryfrom the flat electrolyte interface into theanode structure

Nickel-YSZ anodes have been thor-oughly investigated for the use with hydro-gen in terms of manufacturing raw materi-als selection and microstructural propertiesAnodes based on Ni-YSZ cermets havebeen steadily improved through ceramicprocessing eg careful selection of raw ma-terials [102] adjustment of particle sizes[103] and grading of nickel content in thestructure [104] in the last few years Someof these materials optimizations are report-ed in [105ndash108] and some in a more gener-al context [36][68][84][109ndash111] Moumlbiusrecently reviewed the history of solid elec-trolyte fuel cells and especially the anodesherein [112]

One of the most promising new materi-als for intermediate temperatures is dopedceria a mixed ionic electronic conductorwhich has found considerable attention aselectrolyte [44] As ceria becomes reducedat the anode side of the fuel cell and there-by an n-type semiconductor it can be as-sumed that the triple phase boundary is nolonger defined by single connecting pointsof pore metal and ceramic but is enlargedto the surface of all ceramic grains in themicrostructure Ni-CGO anodes have beensuccessfully fabricated and excellent per-formances have been reported in hydrogenas fuel at intermediate temperatures[113ndash115] Additions of doped ceria canalso be used to increase the performance ofconventional Ni-YSZ composites[116][117]

One advantage of SOFCs as comparedto PEM or MCFC is their potential to be op-erated directly on hydrocarbon or alcoholfuels without complex fuel processing [67]More exotic fuels include CH3OCH3 [118]wood gasification gases [119] H2S [120]

CO [121] and methane [122] Pure CH4 caneither be directly electrochemically oxi-dized with oxygen ions at the anode or itcan as well as any other hydrocarbon beinternally or externally steam reformedwith water vapour to yield carbon monox-ide and hydrogen [105] In conventional Ni-YSZ anodes the nickel can be used as steamreforming catalyst to form hydrogen at theanode Water can either originate from anexternal source through the humidificationof the fuel gas to obtain large steam to car-bon ratios or in parts from water producedby the fuel oxidation reaction Methane athigh steam to carbon ratios can be reformedwithout carbon deposits on nickel contain-ing anodes but the excellent steam reform-ing properties of Ni leads to a total conver-sion within the first few millimetres of thefuel inlet resulting in steep thermal gradi-ents within the cell due to the endothermiccharacter of the reaction

The major problem associated with theuse of dry methane or higher hydrocarbonsfor the direct oxidation is the formation ofcarbon deposits in the form of filamentouscarbon tar and soot during operation athigh temperatures This is due to the highcatalytic activity of metallic nickel towardscarbon formation rapidly clogging thepores and blocking reaction sites on thenickel surface [123][124] Even at low car-bon levels the reaction of Ni with carbonwill finally lead to a disintegration of theanode by a process called metal dusting[125] Takeguchi et al [126] added smallamounts of precious metals to conventionalNi-YSZ cermets to shift the active sites forsteam reforming from Ni to the noble met-al and observed less carbon deposits withRu and Pt during steam reforming ofmethane

Another problem at the anode associat-ed with the use of natural gas based fuels ispoisoning by adsorption of traces of H2Susually present in any natural fuel on thenickel surface [127] Dilution of the fuelgas by steam reforming products and oxi-dized fuel such as carbon dioxide and watervapour can result in performance loss athigh fuel utilization [128] or even reoxida-tion of metallic nickel to nickel oxide nearthe fuel outlet

The search for alternative anodes withlower activity for cracking of hydrocarbonsand better stability than pure Nickel hasproceeded in various directions The cat-alytic activity of nickel itself can be gradu-ally reduced by alloying the metal with oth-er elements eg gold [129] or copper[130ndash132]

Copper similar to gold exhibits almostno electrochemical activity and the com-plete replacement of Ni by Cu to form a cer-met with ceria leads to an anode with thecopper being a purely electronically con-ducting current collector and the ceramic

Fig 10 Schematic drawing of a fuel cell withproton-conducting electrolyte


being the actual electrochemically activecomponent [133]

Pure and doped ceria are known fortheir good performance as oxidation cata-lysts or as catalyst supports CGO(Ce09Gd01O2ndashd) was found to have almostno tendency towards carbon formation[134][135] but exhibits a rather lowcatalytic activity for steam reforming andcracking of methane at 1000 degC The results of Marina et al [136][137] forincreased gadolinia dopant levels inCe06Gd04O2ndashdgold cells are consistentwith these findings Zhao and Gorte [138]examined the catalytic activity of variousdoped cerium oxides for the direct n-butaneoxidation and reported that pure CeO2 al-ways outperforms doped samples and thatincreasing dopant levels reduce reactionrates The catalytic oxidation of methanehas been recently addressed by Horita et al[139] using the isotope labelling techniqueto identify reaction sites on YSZ and yttria-doped ceria (YDC) with gold and nickelelectrodes The YDC substrate proved to beefficient in reducing carbon deposits on Niby increasing the oxygen concentration onthe Ni surface through proton interactionbetween Ni and YDC

Gorte and co-workers [140][141] aswell as other groups have fabricated andtested Cu-puredoped ceria anodes for thedirect oxidation of methane and higher hy-drocarbons However their spectacular in-terpretations of the activity of Cu to processpropane had to be corrected The poweroutput of Cu-puredoped ceria anodes con-taining fuel cells was solely due to H2 as fu-el originating from thermal decompositionof propane to propene occurring at 700 degCalso in absence of Cu as recently shown byJoumlrger [142]

Copper-containing anodes are also be-lieved to be more tolerant against sulphurthan nickel-based electrodes [140] Thesteam reforming capabilities of Cu-CGOcermets can be further enhanced by the ad-dition of small amounts of noble metalssuch as Ru [143][144]

Irvine and co-workers [105][145][146]investigated the mixed ionic electronic con-ductor titania-doped YSZ (YTZ) and YTZwith yttrium substituted by scandium [147]and compared it to ceria The thermal me-chanical and electrical properties of YTZ ina fuel cell environment seem to befavourable [148] The pure form [149] aswell as Ni [150] and Cu [151] cermets per-formed well in hydrogen YTZ was foundnot to promote methane cracking [152] butwas catalytically less active than ceria andshowed only limited electronic conductivi-ty

Efforts have been made to replace thetraditional cermet anode by a pure ceramicmaterial [153] for the direct utilization ofnatural gas as fuel Perovskites fluorites

pyrochlores and tungsten bronzes[146][154ndash158] were investigated Rutilestructures such as Nb2TiO6 show a highelectronic conductivity especially under re-ducing atmospheres but have very low ther-mal expansion coefficients compared tostandard fuel cell materials [159] Reich etal [160] related the poor electrochemicalperformance of niobates to the slow ionicdiffusion in the material and proposed touse it as a current collector instead of an an-ode Tungsten bronzes showed either poorstability under hydrogen too large thermalexpansion coefficient mismatch to the elec-trolyte or poor electrochemical perform-ance [161ndash163]

Amongst the more promising candi-dates to replace established anodes are lan-thanum strontium chromite La1ndashxSrxCrO3(LSC) perovskites [164] This class of ma-terial is already used as interconnect inSOFC stacks and shows good stability un-der operating conditions [165] Vernoux etal [166] reported stable electrochemicalbehaviour of B-site vanadium-doped LSCSfeir et al [167][168] investigated the cat-alytic activity of various A and B sitedopants of LaCrO3 and found Sr and Ni tobe the most suitable substituents for anodepurposes although it is not clear whetherthe exsolution of Ni from the structure ledto the good performance Sauvet et al[169][170] tried to improve reforming ac-tivity by small ruthenium additions to La1-xSrxCrO3 Gonzales-Cuenca et al[171] tested lanthanum-based chromite-ti-tanate perovskites and found insufficientelectronic conductivity Interesting resultshave also been obtained with lanthanumstrontium titanates [172][173] Hui andPetric [174ndash176] reported the properties ofrare-earth-doped SrTiO3 and propose yttri-um doping for further investigations Slateret al [177] reported conductivity data on A-site deficient Sr1ndash3x2LaxTiO3ndashd

Based on the experience with lanthanumstrontium chromites Tao and Irvine [178] in-vestigated complex perovskites of the struc-ture (LaSr)2M1ndashxCr1+xO6ndashd with transitionmetals M on the B-sites Excellent electro-chemical performance comparable to that ofNi-YSZ and material stability in hydrogenand dry methane were achieved with highlevels (x = 05) of Mn doping

The requirements for an efficient fuelelectrode are many and some of the newmaterials show very promising propertiesfor the development of next generation an-odes that will enable the use of available fu-els and operate at lower temperatures thanexisting ones

5 Cathode

For proper function as a cathode in asolid oxide fuel cell the material should

have a high electrocatalytic activity towardsoxygen reduction and a high chemical sta-bility in an oxidizing environment withoutforming highly resistive reaction productswith the electrolyte and current collector[179][180] The material should exhibitsimilar thermomechanical properties as theelectrolyte to avoid stresses developing up-on heating and cooling [181] and it shouldhave high electrical conductivity

Most reviews on SOFCs deal with state-of-the-art cathode materials such as La1ndashxSrxMnO3ndashd (LSM) and La1ndashxSrxCo1ndashyFeyO3-d (LSCF) [1][34][36][67][68][82][84][110][182ndash186] A few of these re-views also include emerging materials[19][68][84][184] The following will belimited to cathode material aspects and ex-clude most processing related techniqueswhich can be found elsewhere [187]

The oxygen reduction reaction at theinterface between a SOFC cathode and anO2ndash conducting electrolyte is

and is schematically represented inFig 11 SOFC cathodes are usually p-typesemi-conductors [188][189] that can be ei-ther an electronic or mixed ionic-electron-ic conductor (MIEC) Reduction of theelectrokinetic losses and optimization ofthe electrode performance are two majorgoals of research and development In caseof pure electronic conductors the oxygenadsorbs on the surface of the material anddiffuses over its surface towards the tpbwhere it becomes charged and incorporat-ed in the electrolyte The electrode acts asan electron supplier Accordingly improv-ing the cathode performance towards highcurrent density and low overpotentials isclosely related to the increase of the tpb-length

If the SOFC cathode is a MIEC oxygencan be reduced on the surface and diffusethrough the bulk of the electrode Conse-quently surface and bulk pathways that co-exist in parallel are in competition and thefastest one determines the kinetics of theoverall reaction If the surface pathway israte-determining the electrode exhibits asimilar behaviour as for a purely electronicconductor as described previously On theother hand if the oxygen migrates mainlythrough the bulk of the cathode the electro-chemical reaction is promoted by produc-ing dense thin layers enhancing therebyoxygen exchange at both the MIECgas andMIECelectrolyte interfaces These materi-als should have a high oxygen exchange ca-pacity for an easy incorporation of oxygenin their lattice and high oxygen diffusivityfor high transport rates However themechanism and kinetics of oxygen reduc-tion at SOFC cathodes are still under ques-

(4)


tion The interaction between oxygen andthe MIEC and oxygen diffusion have beensubject to numerous studies [190ndash192]Comprehensive understanding and model-ling of these reaction mechanisms consti-tute an ongoing field of investigations fromwhich controversial results were publishedso far [193ndash197]

51 La1ndashxSrxMnO3ndashx2 (LSM) andLaxSr1ndashxCoyFe1ndashyO3 (LSCF)Cathodes

The choice of cathode materials israther limited Noble metals such as Pt aresuitable but exhibit prohibitive costs forSOFC application at higher temperaturesdue to high Pt suboxide vapour pressureLa1ndashxSrxMnO3ndashx2 (LSM) as the state-of-the-art electronic conducting material iswidely used since it fulfills most of the re-quirements listed above its properties aregiven in Table 3 with the data taken fromreferences [181][198ndash200] Usually LSMis used for the cathode when YSZ is used asthe electrolyte because the thermal expan-sion coefficients match well [201] Howev-er the rather high operating temperatures ofthe SOFC around 900 to 1000 degC promotedegradation of the cathode and the forma-tion of undesired resistive reaction prod-ucts such as La2Zr2O7 especially duringmanufacturing of LSM on YSZ[180][202ndash205]

Increased triple phase boundary lengthbetter adhesion to the electrolyte and lowerthermal expansion mismatch is achievedwhen using a LSM-YSZ composite materi-al [204][206][207] or even composites with

graded compositions [201] Besides YSZCGO [208] Sm02Ce08O2 (SDC) [209] andCe07Bi03O2 [210] are also used for fabri-cation of composite cathodes with LSMwith improved performance

As for most perovskite materials theproperties of LSM can be tailored bypartially substituting the A and B sites ofthe ABO3 perovskite The thermal expan-sion coefficient (TEC) can be furtheradjusted to that of the YSZ electrolyte byusing (La1ndashxYx)07Sr03MnO3 [211] orSr1ndashxCexMnO3ndashd [212] Compositionswhich are compatible with CGO as regardsTEC and chemical stability are Gd1ndashxSrxMnO3 Nd1ndashxSrxMnO3ndashd [213] and Pr1-xSrxMnO3 [214] The formation of reactionproducts between the YSZ electrolyte andthe cathode can be suppressed for Ln1ndashxSrxMnO3 (Ln = Pr Nd) [215] andPr1ndashxCaxMnO3 [216] whereas forLa1ndashxCaxMnO3 on a CaO-stabilized ZrO2electrolyte no stable composition wasfound [217] The conductivity can be in-

creased by using Pr06ndashxSr04MnO3[218][219] but for substitution of Mn withCo in Y06Sr04Mn1ndashyCoyO3 (0 pound y pound 04)mixtures increasing y resulted in lowerconductivity [220] the same is observed foradding Al to LSM [221]

The La1ndashxSrxCoO3ndashd (LSC) based cath-odes [222ndash224] are typical mixed conduc-tors offering the advantage of higher elec-tronic and more important higher ionicconductivity (see Table 3) By providingthis second pathway for oxygen ions activ-ity of the cathode is increased and lower op-erating temperatures are feasible The dis-advantage is that those materials react withYSZ [202][224] thus either ceria-basedelectrolytes or protective layers of ceria[224] or LSGM [223][225][226] on YSZelectrolytes should be used In order toadjust the TEC of LSC-based cathodes tothe one of CGO Fe was introduced to ob-tain lower TEC [227] Depending on thecomposition the conductivities of La1ndashxSrxCoyFe1ndashyO3ndashd can vary about one orderof magnitude [181][188][227][228] Onestrategy to improve performance of LSCFcathodes is the fabrication of compositeelectrodes with CGO [54][229] CGOAg[55] or SDC [230] or to obtain higher sur-face exchange coefficient k by impregnat-ing LSCF with Pd [231]

Cathode performance can also be im-proved by substituting one or more of theelements in Ln1ndashxSrxCoyFe1ndashyO3ndashd En-hanced performance at low temperatures(~600 degC) is obtained for Ln = Ce Dy[232] whereas TEC is lowered for Ln = Nd[233] Reaction products with YSZ are lesspronounced for Ln = Pr Nd Gd [57] OnCGO no reaction products are found for Ln= La Gd Sm Nd [234][235] although nodistinct reaction products with LSGM arefound codiffusion into the electrolyte is de-tected [236] Sr-doped lanthanum ferriteshave also been investigated since they havea lower TEC than LSCF [237] but they al-so form Sr- or La-zirconates with YSZ[238] which can be reduced by adding Alto LaFe1ndashxAlxO3 systems without Sr doping[239] or using Ce08Sm02O19 protectionlayers [238] The conductivity is compara-ble to that of LSCF and is enhanced byadding Ni [240][241] or replacing Sr with

Fig 11 Schematic representation of oxygen reduction in a mixed ionic-electronic conductor Sur-face and bulk reaction pathways are parallel and in competition On the surface pathway chargetransfer occurs at the triple phase boundary

Table 3 Coefficient of thermal expansion (TEC) (30ndash1000 degC) electronic (se) and ionic (si) conduc-tivity and bulk diffusion D as well as surface exchange coefficient k at 800 degC for some SOFC cathodematerials

Material TEC10ndash6Kndash1 se[Scm] si[Scm] D[cm2s] k[cms]

La065Sr035MnO3-d 123 [181] 102 [181] 17middot10ndash4 4middot10ndash14 5middot10ndash8

(YSZ 110middot[198]) [181] [198] [198](at 900 degC) (at 900 degC)

La06Sr04Co02Fe08O3 175 [181] 302 [181] 8middot10ndash3 25middot10ndash8 56middot10ndash6

(CGO 105 [199]) [181] [200] [200]


Ni [242] but is decreased by adding Al[243][244]

Another material that is investigated forcathodes is Sm1ndashxSrxCoO3 (SSC)[202][245][246] showing lower overpoten-tial than LSC [246] Fabricating compositeswith the electrolyte material (Ce08Sm02O19) the interfacial resistances arereduced [247] SSC is also used for singlechamber SOFC applications [14][248]

Barium cobaltates Ba1ndashxLnxCoO3 Ln =La Pr are studied on either BaCeO3[245][249] or LSGM [250] based elec-trolytes and found to have less polarizationlosses than SSC for Ln = Pr [245] but high-er overpotentials than SSC for Ln = La[250]

52 New Cathode MaterialsPyrochlore ruthenates have been inves-

tigated with compositions of Bi2Ru2O73Pb2Ru2O65 and Y2Ru2O7 Only the latterwas found to be stable on CGO electrolytesbut additional doping with SrO is necessaryin order to reach reasonable conductivity[251]

The search for new cathode materialsfor intermediate temperatures led to the dis-covery of La1ndashxSrxCuO25ndashd This materialis a possible cathode candidate because itshows no reaction with YSZ it exhibits

high conductivity and gives reasonably lowoverpotential [252] La2Ni1ndashxCuxO4+d onthe other hand shows high diffusion andsurface exchange coefficients but ratherlow conductivity comparable to LSM[253] Composite cathodes of Ag and yttri-um doped bismuth oxide show comparableperformance to LSCF [54] ForY1Ba2Cu3O7 an additional layer of Pt or Agis needed to promote oxygen adsorption[254] Nd2NiO4+d cathodes show lower po-larization resistance than LSM but long-term stability tests have not been performed[255]

In the search for new cathode materialsbased on perovskites a structural field mapof perovskites containing transition metalcations may be useful In Fig 12 theCoulomb potentials from the A and Bcations in perovskites ABO3 are plotted asZArA and ZBrB Thereby ZA and ZB are theformal valence of the A and B cations andrA respectively rB are their ionic radii Thesmaller the value of ZArA along the ordi-nate is the more itinerant the d-electrons ofthe perovskite become Similarly the small-er the value ZBrB along the abscissa getsthe more itinerant the d-electrons of the per-ovskite are The physical meaning of theparameter ZArA is a measure for the per-turbation of the covalent Bndash0 bond and the

parameter ZBrB is a measure for theCoulomb potential of the outermost d-elec-trons from the centre of the B ion In thispotential map we find two well-defined re-gions The region of compounds with local-ized electrons and that of itinerant elec-trons both separated by the line in thegraph [179][256] The most interestingcompounds and corresponding solid solu-tions are located with their potentials di-rectly on or close to the dividing line be-tween these two regions Along this line wewill find new catalysts as well as materialswith interesting electrical properties suchas high mixed electronicionic conductivity

6 Summary and Conclusions

One of the main problems of SOFCs isthe high operating temperature leading to afast degradation rate of cell performanceand the need for more expensive intercon-nect and sealing materials The electrolyteresistance mainly determines the operatingtemperature of the cell Two ways are pos-sible to decrease the latter either by de-creasing the electrolyte thickness or by us-ing alternative electrolyte materials withhigher ionic conductivity One of the mate-rials that have been proposed for low tem-

Fig 12 Potential map of some perovskites useful in the search of new compounds Compounds with interesting catalytic and electrical properties arelocated on or close to the line dividing the areas of compounds with semiconducting and metallic character modified after [256]


perature fuel cells are ceria solid solutionswith tri- or divalent cations The aim of de-velopment of SOFC materials is towardslower operating temperatures from 500 to800 degC Ceria composite electrolytes espe-cially in bi-layered configuration with athin electron-blocking YSZ layer on anodeor cathode support structures are promisingalternatives to the pure load bearing YSZelectrolytes in planar configurations Thinfilm techniques are used for the fabricationof such structures

Anode materials with tailored catalyticactivities towards reforming of hydrocar-bons as well as robustness and high toler-ance against oxidationreduction cycles areneeded Anodes with sufficient sulphur tol-erance up to levels of 20ndash50 ppm should beaimed for

In addition miniaturized SOFCs will re-quire new thin film cathodes electrolytes andanodes with microstructures in the nanometrerange as well as new support structures thatcan be micro-machined and bonded withmethods from micro technology

Cathodes with mixed electronic andionic conductivity promise to reduce over-potentials especially for low and intermedi-ate temperature use Thereby special em-phasis should be devoted to materials withhigh oxygen surface exchange coefficientsin addition to high oxygen diffusivityStructural field maps may help in the searchof new catalytic materials and those withunusual high electronic and oxygen ionconductivity

Received October 27 2004

[1] NQ Minh J Am Ceram Soc 199376(3) 563

[2] KL Ley M Krumpelt R Kumar JHMeiser I Bloom J Mater Res 199611(6) 1489

[3] K Eichler G Solow P Otschik WSchaffrath J Eur Ceram Soc 199919(6-7) 1101

[4] SP Simner JW Stevenson J PowerSources 2001 102(1-2) 310

[5] NF Bessete BP Borglum H Schichl DSSchmidt Siemens Power J 2001 1 10

[6] RA George J Power Sources 200086(1-2) 134

[7] JPP Huijsmans RC Huiberts GMChristie Fuel Cells Bul 1999 2(14) 5

[8] M Schmidt Fuel Cells Bul 1998 1(1) 9[9] T Doerk H Raak S Riggenbach GWF

GasErdgas 2003 144(12) 746[10] W van Gool Philips Res Rep 1965 20

81[11] CK Dyer Nature 1990 343 547[12] T Hibino H Iwahara Chem Lett 1993

7 1131[13] M Goedickemeier D Nussbaum C

Kleinlogel LJ Gauckler in 192nd Meet-ing of Electrochemical Society 1997Paris France Electrochemical Soci-ety

[14] T Hibino A Hashimoto T Inoue JTokuno Y Shin-ichiro S Mitsuru Sci-ence 2000 288 2031

[15] T Hibino A Hashimoto M Yano MSuzuki S Yoshida S Mitsuru J Elec-trochem Soc 2002 149(2) A133

[16] S Calabrese Barton T Patterson EWang TF Fuller AC West J PowerSources 2001 96(2) 329

[17] MA Priestnall VP Kotzeva DJ FishEM Nilsson J Power Sources 2002106(1-2) 21

[18] BCH Steele J Power Sources 199449(1-3) 1

[19] SJ Skinner JA Kilner Mater Today2003 6(3) 30

[20] SC Singhal Solid State Ionics 2000135(1-4) 305

[21] SPS Badwal Solid State Ionics 2001143(1) 39

[22] SPS Badwal Solid State Ionics 199252(1-3) 23

[23] IR Gibson GP Dransfield JTSIrvine J Eur Ceram Soc 1998 18(6)661

[24] FT Ciacchi SPS Badwal J DrennanJ Eur Ceram Soc 1991 7(3) 185

[25] SPS Badwal FT Ciacchi S Rajen-dran J Drennan Solid State Ionics1998 109(3-4) 167

[26] K Kobayashi H Kuwajima T MasakiSolid State Ionics 1981 3-4 489

[27] FM Figueiredo JA Labrincha JRFrade FMB Marques Solid State Ion-ics 1997 101-103(Part 1) 343

[28] SP Jiang JG Love JP Zhang MHoang Y Ramprakash AE HughesSPS Badwal Solid State Ionics 1999121(1-4) 1

[29] O Yamamoto Y Takeda R Kanno MNoda Solid State Ionics 1987 22(2-3)241

[30] M Dokiya Solid State Ionics 2002152ndash153 383

[31] B Steele Comptes Rendus de lrsquoAcade-mie des Sciences - Series IIC - Chem-istry 1998 1(9) 533

[32] J Will A Mitterdorfer C KleinlogelD Perednis LJ Gauckler Solid StateIonics 2000 131(1-2) 79

[33] H Tu U Stimming J Power Sources2004 127(1-2) 284

[34] H Yokokawa N Sakai T Horita K Ya-maji Fuel Cells 2001 1(2) 117

[35] A Boudghene Stambouli E TraversaRenew Sustain Energy Rev 2002 6(5)433

[36] SM Haile Acta Mater 2003 51(19)5981

[37] P Shuk H-D Wiemhofer U Guth WGopel M Greenblatt Solid State Ionics1996 89(3-4) 179

[38] SPS Badwal FT Ciacchi D Milo-sevic Solid State Ionics 2000 136-137 91

[39] M Yashima M Kakihana M Yoshimu-ra Solid State Ionics 1996 86-88(Part2) 1131

[40] O YamamotoY AratiY Takeda N Iman-ishiY Mizutani M KawaiY NakamuraSolid State Ionics 1995 79 137

[41] Y Mizutani M Tamura M Kawai OYamamoto Solid State Ionics 199472(Part 2) 271

[42] TI Politova JTS Irvine Solid StateIonics in press

[43] CM Kleinlogel LJ Gauckler J Elec-troceram 2000 5(3) 231

[44] M Mogensen NM Sammes GATompsett Solid State Ionics 2000129(1-4) 63

[45] BCH Steele Solid State Ionics 2000129(1-4) 95

[46] H Yahiro K Eguchi H Arai SolidState Ionics 1989 36(1-2) 71

[47] K Eguchi T Setoguchi T Inoue H AraiSolid State Ionics 1992 52(1-3) 165

[48] T Inoue T Setoguchi K Eguchi HArai Solid State Ionics 1989 35(3-4)285

[49] D Perez-Coll P Nunez JR FradeJCC Abrantes Electrochim Acta2003 48(11) 1551

[50] S Zha C Xia G Meng J PowerSources 2003 115(1) 44

[51] Z Tianshu P Hing H Huang J KilnerSolid State Ionics 2002 148(3-4) 567

[52] J Kilner BCH Steele lsquoNonstoichio-metric oxidesrsquo Academic Press NewYork 1981

[53] D Schneider M Godickemeier LJGauckler J Electroceram 1997 1(2) 165

[54] R Doshi VL Richards JD CarterXP Wang M Krumpelt J Elec-trochem Soc 1999 146(4) 1273

[55] S Wang T Kato S Nagata T HondaT Kaneko N Iwashita M Dokiya Sol-id State Ionics 2002 146(3-4) 203

[56] S Wang T Kato S Nagata T KanekoN Iwashita T Honda M Dokiya SolidState Ionics 2002 152-153 477

[57] L Qiu T Ichikawa A Hirano N Iman-ishi Y Takeda Solid State Ionics 2003158(1-2) 55

[58] I Riess Solid State Ionics 2003 157(1-4) 1

[59] S Dikmen P Shuk M Greenblatt Sol-id State Ionics 1998 112(3-4) 299

[60] JC Boivin C Pirovano G Nowo-grocki G Mairesse P Labrune G La-grange Solid State Ionics 1998 113-115 639

[61] F Abraham JC Boivin G MairesseG Nowogrocki Solid State Ionics 199040-41(2) 934

[62] RN Vannier RJ Chater SJ SkinnerJA Kilner G Mairesse Solid StateIonics 2003 160(3-4) 327

[63] C Pirovano RN Vannier E CapoenG Nowogrocki JC Boivin GMairesse M Anne E Dooryhee PStrobel Solid State Ionics 2003 159 (1-2) 167

[64] G Mairesse Comptes Rendus de lrsquoA-cademie des Sciences - Series IIC -Chemistry 1999 2(11-13) 651


[65] NM Sammes GA Tompsett H NafeF Aldinger J Eur Ceram Soc 199919(10) 1801

[66] RM Ormerod Chem Soc Rev 200232 17

[67] BCH Steele A Heinzel Nature 2001414(6861) 345

[68] NP Brandon S Skinner BCH SteeleAnn Rev Mater Res 2003 33 183

[69] J Wiemhofer H-D Weitkamp SolidState Ionics 2002 154-155 597

[70] J Will MKM Hruschka L GublerLJ Gauckler J Am Ceram Soc 200184(2) 328ndash32

[71] H Sasaki S Otoshi M Suzuki T So-gi A Kajimura N Sugiuara M Ip-pommatsu Solid State Ionics 1994 72253

[72] HR Gruner H Tannenberger Proceed-ings of the First European Solid OxideFuel Cell Forum 1994 Ed U BosselEuropean SOFC Forum LucerneSwitzerland 1994 611

[73] K Honegger E Batawi C Sprecher RDiethelm Proceedings of SOFC V(Aachen Germany 1997) Eds U Stim-ming SC Singhal H Tagawa WLehner Electrochemical Society Pen-nington NJ 1997 321

[74] D Perednis LJ Gauckler Solid StateIonics 2004 166(3-4) 229

[75] D Perednis LJ Gauckler in 8th Inter-national Symposium on Solid Oxide Fu-el Cells (SOFC) 2003 Paris FranceThe Electrochemical Society

[76] D Perednis LJ Gauckler in 8th SolidOxide Fuel Cell ECS Symposium 2003Electrochemical Society


[78] A Tsoga A Gupta A Naoumidis PNikolopoulos Acta Mater 2000 48(18-19) 4709

[79] H Chang in 5th Annual Small FuelCells 2003 ndash Small Fuel Cells forPortable Applications 2003

[80] FMB Marques LM Navarro SolidState Ionics 1996 90(1-4) 183


[82] SPS Badwal K Foger Ceram Int1996 22(3) 257

[83] BCH Steele Curr Opin Solid StateMat Sci 1996 1(5) 684

[84] V Kozhukharov N Brashkova MIvanova J Carda M Machkova BolSoc Esp Ceram Vidr 2002 41(5) 471

[85] H Iwahara H Uchida S Tanaka SolidState Ionics 1983 9-10(Part 2) 1021

[86] JF Liu AS Nowick Solid State Ionics1992 50(1-2) 131

[87] N Bonanos Solid State Ionics 1992 53-56(Part 2) 967

[88] RCT Slade N Singh Solid State Ion-ics 1993 61(1-3) 111

[89] H Iwahara T Yajima T Hibino HUshida J Electrochem Soc 1993140(6) 1687

[90] H Iwahara T Shimura H MatsumotoElectrochemistry 2000 68(3) 154

[91] H Iwahara T Mori T Hibino SolidState Ionics 1995 79 177

[92] KD Kreuer Ann Rev Mater Res2003 33 333

[93] G Ma T Shimura H Iwahara SolidState Ionics 1999 120(1-4) 51

[94] KD Kreuer Solid State Ionics 1999125(1-4) 285

[95] KH Ryu SM Haile Solid State Ionics1999 125(1-4) 355

[96] HG Bohn T Schober J Am CeramSoc 2000 83(4) 768

[97] KD Kreuer S Adams W Munch AFuchs U Klock J Maier Solid StateIonics 2001 145(1-4) 295

[98] H Fujii Y Katayama T Shimura HIwahara J Electroceram 1998 2(2)119

[99] W Grover Coors J Power Sources2003 118(1-2) 150

[100] W Munch K-D Kreuer G Seifert JMaier Solid State Ionics 2000 136-137183

[101] T Setoguchi K Okamoto K EguchiH Arai J Electrochem Soc 1992139(10) 2875

[102] F Tietz FJ Dias D Simwonis DStover J Eur Ceram Soc 2000 20(8)1023

[103] T Fukui K Murata S Ohara H AbeM Naito K Nogi J Power Sources2004 125(1) 17

[104] AC Muller D Herbstritt E Ivers-Tiffee Solid State Ionics 2002 152 537

[105] JTS IrvineA Sauvet Fuel Cells 20011(3-4) 205

[106] J Van Herle S Diethelm J Sfeir RIhringer lsquoMaterials for methane-fueledSOFC systemsrsquo in Euro Ceramics ViiPt 1-3 2002 p 1213

[107] WZ Zhu SC Deevi Mater Sci Eng A2003 362(1-2) 228

[108] A Atkinson S Barnett RJ GorteJTS Irvine AJ McEvoy M Mo-gensen SC Singhal J Vohs NatMater 2004 3(1) 17

[109] BCH Steele Solid State Ionics 199686-8 1223

[110] JPP Huijsmans Curr Opin Solid StateMat Sci 2001 5(4) 317

[111] M Mogensen KV Jensen MJ Jor-gensen S Primdahl Solid State Ionics2002 150(1-2) 123

[112] HH Mobius J Solid State Elec-trochem 1997 1(1) 2

[113] S Wang T Kato S Nagata T HondaT Kaneko N Iwashita M Dokiya JElectrochem Soc 2002 149(7) A927

[114] C Xia M Liu Solid State Ionics 2002152-153 423

[115] S Zha W Rauch M Liu Solid StateIonics 2004 166(3-4) 241

[116] XQ Huang ZG Liu Z Lu L PeiRB Zhu YQ Liu JP Miao ZGZhang WH Su J Phys Chem Solids2003 64(12) 2379

[117] X Huang Z Lu L Pei Z Liu Y LiuR Zhu J Miao Z ZhangW Su J Al-loy Compd 2003 360(1-2) 294

[118] EP Murray SJ Harris HW Jen JElectrochem Soc 2002 149(9) A1127

[119] S Baron N Brandon A Atkinson BSteele R Rudkin J Power Sources2004 126(1-2) 58

[120] L Zhong M Liu GL Wei KChuang Chin J Chem Eng 200311(3) 245

[121] A Weber B Sauer AC Muller DHerbstritt E Ivers-Tiffee Solid StateIonics 2002 152 543

[122] BCH Steele Nature 1999 400(6745)619

[123] GJ Saunders J Preece K Kendall JPower Sources 2004 131(1-2) 23

[124] T Takeguchi Y Kani T Yano RKikuchi K Eguchi K Tsujimoto YUchida A Ueno K Omoshiki M Aiza-wa J Power Sources 2002 112(2) 588

[125] CM Chun JD Mumford TA Rama-narayanan J Electrochem Soc 2000147(10) 3680

[126] T Takeguchi R Kikuchi T Yano KEguchi K Murata Catal Today 200384(3-4) 217

[127] Y Matsuzaki I Yasuda Solid State Ion-ics 2000 132(3-4) 261

[128] O Costa-Nunes JM Vohs RJ Gorte JElectrochem Soc 2003 150(7) A858

[129] IA Proctor AL Hopkin RMOrmerod Ionics 2003 9(3-4) 242

[130] MT Tavares I Alstrup CAA Bernar-do Mater Corros 1999 50(12) 681

[131] H Kim C Lu WL Worrell JM VohsRJ Gorte J Electrochem Soc 2002149(3) A247

[132] Z Lu L Pei TM He XQ HuangZG Liu Y Ji XH Zhao WH Su JAlloy Compd 2002 334 299

[133] C Lu WL Worrell JM Vohs RJGorte J Electrochem Soc 2003150(10) A1357

[134] E Ramirez-Cabrera A Atkinson DChadwick Solid State Ionics 2000 136-137 825

[135] E Ramirez-Cabrera A Atkinson DChadwick Appl Catal B-Environ2004 47(2) 127

[136] OA Marina C Bagger S PrimdahlM Mogensen Solid State Ionics 1999123(1-4) 199

[137] OA Marina M Mogensen ApplCatal A-Gen 1999 189(1) 117

[138] S Zhao RJ Gorte Appl Catal A-Gen2003 248(1-2) 9

[139] T Horita K Yamaji T Kato N SakaiH Yokokawa J Power Sources 2004131(1-2) 299

[140] RJ Gorte H Kim JM Vohs J PowerSources 2002 106(1-2) 10

[141] C Lu WL Worrell C Wang S ParkH Kim JM Vohs RJ Gorte SolidState Ionics 2002 152 393

[142] MB Joumlrger PhD Thesis ETH No15351 2004


[143] T Hibino A Hashimoto M Yano MSuzuki M Sano Electrochim Acta2003 48(17) 2531

[144] S McIntosh JM Vohs RJ GorteElectrochem Solid State Lett 20036(11) A240

[145] AJ Feighery JTS Irvine DP FaggA Kaiser J Solid State Chem 1999143(2) 273

[146] P Holtappels J Bradley JTS IrvineA Kaiser M Mogensen J Elec-trochem Soc 2001 148(8) A923

[147] S Tao JTS Irvine J Solid State Chem2002 165(1) 12

[148] M Mori Y Hiei H Itoh GATompsett NM Sammes Solid StateIonics 2003 160(1-2) 1

[149] A Kelaidopoulou A Siddle ALDicks A Kaiser JTS Irvine FuelCells 2001 1(3-4) 226

[150] D Skarmoutsos F Tietz P Nikolopou-los Fuel Cells 2001 1(3-4) 243

[151] N Kiratzis P Holtappels DE Hatch-well M Mogensen JTS Irvine FuelCells 2001 1(3-4) 211


[153] BA Boukamp Nat Mater 2003 2(5)294

[154] JJ Sprague HL Tuller J Eur CeramSoc 1999 19(6-7) 803

[155] H Yokokawa N Sakai T KawadaM Dokiya Solid State Ionics 199252(1-3) 43

[156] G Pudmich BA Boukamp M Gonza-lez-Cuenca W Jungen W Zipprich FTietz Solid State Ionics 2000 135(1-4)433

[157] P Holtappels FW Poulsen M Mo-gensen Solid State Ionics 2000 135(1-4) 675

[158] A-L Sauvet J Fouletier J PowerSources 2001 101(2) 259

[159] A Lashtabeg JTS Irvine A FeigheryIonics 2003 9(3-4) 220

[160] CM Reich A Kaiser JTS IrvineFuel Cells 2001 1(3-4) 249

[161] A Kaiser JL Bradley PR SlaterJTS Irvine Solid State Ionics 2000135(1-4) 519

[162] PR Slater JTS Irvine Solid State Ion-ics 1999 124(1-2) 61


[164] S Primdahl JR Hansen L Grahl-Madsen PH Larsen J ElectrochemSoc 2001 148(1) A74

[165] S Tanasescu D Berger D Neiner NDTotir Solid State Ionics 2003 157(1-4)365

[166] P Vernoux M Guillodo J FouletierA Hammou Solid State Ionics 2000135(1-4) 425

[167] J Sfeir PA Buffat P Mockli N Xan-thopoulos R Vasquez HJ Mathieu JVan herle KR Thampi J Catal 2001202(2) 229

[168] J Sfeir lsquoAlternative Anode Materials forMethane Oxidation in Solid Oxide FuelCellsrsquo PhD Thesis 2002

[169] AL Sauvet J Fouletier F Gaillard MPrimet J Catal 2002 209(1) 25

[170] AL Sauvet J Fouletier ElectrochimActa 2001 47(6) 987

[171] M Gonzalez-Cuenca W Zipprich BABoukamp G Pudmich F Tietz FuelCells 2001 1(3-4) 256

[172] OA Marina NL Canfield JWStevenson Solid State Ionics 2002149(1-2) 21

[173] J Canales-Vazquez SW Tao JTSIrvine Solid State Ionics 2003 159(1-2) 159

[174] SQ Hui A Petric J Electrochem Soc2002 149(1) J1

[175] SQ Hui A Petric Mater Res Bull2002 37(7) 1215

[176] SQ Hui A Petric J Eur Ceram Soc2002 22(9-10) 1673

[177] PR Slater DP Fagg JTS Irvine JMater Chem 1997 7(12) 2495

[178] SW Tao JTS Irvine Nat Mater2003 2(5) 320

[179] T Nakamura G Petzow LJ GaucklerMater Res Bull 1979 14(5) 649

[180] A Mitterdorfer LJ Gauckler SolidState Ionics 1998 111(3-4) 185

[181] H Ullmann N Trofimenko F Tietz DStover A Ahmad-Khanlou Solid StateIonics 2000 138(1-2) 79


[183] AJ McEvoy J Mater Sci 2001 36(5)1087

[184] JM Ralph AC Schoeler M KrumpeltJ Mater Sci 2001 36(5) 1161

[185] RM Ormerod Chem Soc Rev 200332(1) 17

[186] A Weber E Ivers-Tiffee J PowerSources 2004 127(1-2) 273

[187] J Will R Stadler MKM HruschkaLJ Gauckler lsquoFabrication Processesfor Electroceramic Components inOxygen Ion and Mixed Conductors andTheir Technological Applicationsrsquo EdsHL Tuller et al Kluwer AcademicPress 2000 p 165

[188] S Wang M Katsuki M Dokiya THashimoto Solid State Ionics 2003159(1-2) 71

[189] HU Anderson Solid State Ionics 199252(1-3) 33

[190] JA Kilner RA DeSouza IC Fullar-ton Solid State Ionics 1996 86-8 703

[191] HJM Bouwmester AJ Burggraaf inthe CRC Handbook of Solid-State Elec-trochemistry Ed HJM BouwmesterCRC Press Boca Raton 1997 p 481

[192] M Katsuki S Wang M Dokiya THashimoto Solid State Ionics 2003156(3-4) 453

[193] SB Adler JA Lane BCH SteeleJ Electrochem Soc 1996 143(11) 3554

[194] SB Adler JA Lane BCH Steele JElectrochem Soc 1997 144(5) 1884

[195] M Liu J Winnick J Electrochem Soc1997 144(5) 1881

[196] ML Liu J Winnick Solid State Ionics1999 118(1-2) 11

[197] GW Coffey LR Pederson PC RiekeJ Electrochem Soc 2003 150(8)A1139

[198] S Carter A Selcuk RJ Chater J Kaj-da JA Kilner BCH Steele SolidState Ionics 1992 53-56(Part 1) 597

[199] H Hayashi M Kanoh CJ Quan H In-aba S Wang M Dokiya H TagawaSolid State Ionics 2000 132(3-4) 227

[200] SJ Benson RJ Chater JA KilnerElectrochemical Society Proceedings1997 97-24 596

[201] NT Hart NP Brandon MJ Day JEShemilt J Mater Sci 2001 36(5)1077

[202] JM Ralph C Rossignol R Kumar JElectrochem Soc 2003 150(11) A1518

[203] MC Brant T Matencio L Desse-mond RZ Domingues Chem Mat2001 13(11) 3954

[204] MJL Ostergard C Clausen C Bag-ger M Mogensen Electrochim Acta1994 40(12) 1971

[205] H Kamata A Hosaka J Mizusaki HTagawa Solid State Ionics 1998 106(3-4) 237

[206] K Barthel S Rambert S Siegmann JTherm Spray Technol 2000 9(3) 343

[207] K Hayashi M Hosokawa T YoshidaY OhyaY Takahashi O Yamamoto HMinoura Mater Sci Eng B 1997 49(3)239

[208] NT Hart NP Brandon MJ Day NLapena-Rey J Power Sources 2002106(1-2) 42

[209] SP Yoon J Han SW Nam T-H LimI-H Oh S-A HongY-S Yoo HC LimJ Power Sources 2002 106(1-2) 160

[210] H Zhao L Huo S Gao J PowerSources 2004 125(2) 149

[211] K Murata M Shimotsu J Ceram SocJpn 2002 110(7) 618

[212] S Hashimoto H Iwahara J Electroce-ram 2000 4(1) 225

[213] GC Kostogloudis C Ftikos J Eur Ce-ram Soc 1999 19(4) 497

[214] GC Kostogloudis N Vasilakos CFtikos J Eur Ceram Soc 1997 17(12)1513

[215] Y Sakaki Y Takeda A Kato N Iman-ishi O Yamamoto M Hattori M Iio YEsaki Solid State Ionics 1999 118(3-4)187

[216] H-R Rim S-K Jeung E Jung J-SLee Mater Chem Phys 1998 52(1) 54

[217] S Faaland MA Einarsrud K Wiik TGrande R Hoier J Mater Sci 199934(23) 5811

[218] X Huang J Liu Z Lu W Liu L PeiT He Z Liu W Su Solid State Ionics2000 130(3-4) 195

[219] X Huang L Pei Z Liu Z Lu Y SuiZ Qian W Su J Alloy Compd 2002345(1-2) 265


[220] CY Huang TJ Huang J Mater Sci2002 37(21) 4581

[221] D Kuscer M Hrovat J Holc SBernik D Kolar J Power Sources1998 71(1-2) 195

[222] I Riess M Godickemeier LJ Gauck-ler Solid State Ionics 1996 90(1-4) 91

[223] T Horita K Yamaji N Sakai HYokokawa A Weber E Ivers-TiffeeSolid State Ionics 2000 133(3-4) 143

[224] H Uchida S Arisaka M WatanabeSolid State Ionics 2000 135(1-4) 347

[225] T Inagaki K Miura H Yoshida R Mar-ic S Ohara X Zhang K Mukai T FukuiJ Power Sources 2000 86(1-2) 347

[226] R Maric S Ohara T Fukui H Yoshi-da M Nishimura T Inagaki K MiuraJ Electrochem Soc 1999 146(6) 2006

[227] A Petric P Huang F Tietz Solid StateIonics 2000 135(1-4) 719

[228] L-W Tai MM Nasrallah HU Ander-son DM Sparlin SR Sehlin SolidState Ionics 1995 76(3-4) 259

[229] V Dusastre JA Kilner Solid State Ion-ics 1999 126(1-2) 163

[230] Y Matsuzaki I Yasuda Solid State Ion-ics 2002 152 463

[231] M Sahibzada SJ Benson RA Rud-kin JA Kilner Solid State Ionics 1998113-115 285

[232] J Gao X Liu D Peng G Meng CatalToday 2003 82(1-4) 207

[233] N Dasgupta R Krishnamoorthy KTJacob Mater Sci Eng B 2002 90(3)278

[234] W Chen T Wen H Nie R ZhengMater Res Bull 2003 38(8) 1319

[235] WX Chen HW Nie WH Huang RZheng HY Tu ZY Lu TL Wen JMater Sci Lett 2003 22(9) 651

[236] GC Kostogloudis C Ftikos A Ah-mad-Khanlou A Naoumidis D StoverSolid State Ionics 2000 134(1-2) 127

[237] SP Simner JF Bonnett NL CanfieldKD Meinhardt JP Shelton VLSprenkle JW Stevenson J PowerSources 2003 113(1) 1

[238] SP Simner JP Shelton MD Ander-son JW Stevenson Solid State Ionics2003 161(1-2) 11

[239] D Kuscer J Holc M Hrovat D Kolar JEur Ceram Soc 2001 21(10-11) 1817

[240] R Chiba F Yoshimura Y Sakurai Sol-id State Ionics 2002 152-153 575

[241] SP Simner JF Bonnett NL CanfieldKD Meinhardt VL Sprenkle JWStevenson Electrochem Solid St 20025(7) A173

[242] R Chiba F Yoshimura Y Sakurai Sol-id State Ionics 1999 124(3-4) 281

[243] GW Coffey J Hardy LR PedersenPC Rieke EC Thomsen M WalpoleSolid State Ionics 2003 158(1-2) 1

[244] J Holc D Kuscer M Hrovat SBernik D Kolar Solid State Ionics1997 95(3-4) 259

[245] T Hibino A Hashimoto M SuzukiM Sano J Electrochem Soc 2002149(11) A1503

[246] H Fukunaga M Koyama N Taka-hashi C Wen K Yamada Solid StateIonics 2000 132(3-4) 279

[247] C Xia W Rauch F Chen M Liu Sol-id State Ionics 2002 149(1-2) 11

[248] BE Buumlrgler M Siegrist LJ Gaucklerin Fuel Cell Forum 2004 LucerneSwitzerland

[249] M Koyama C Wen K Yamada JElectrochem Soc 2000 147(1) 87

[250] T Ishihara S Fukui H Nishiguchi YTakita Solid State Ionics 2002 152-153609

[251] JM Bae BCH Steele J Electrocer-am 1999 3(1) 37

[252] H-C Yu K-Z Fung Mater Res Bull2003 38(2) 231

[253] E Boehm J-M Bassat MC Steil PDordor F Mauvy J-C Grenier SolidState Sci 2003 5(7) 973

[254] CL Chang TC Lee TJ Huang J Sol-id State Electrochem 1998 2(5) 291

[255] F Mauvy J-M Bassat E Boehm J-PManaud P Dordor J-C Grenier SolidState Ionics 2003 158(1-2) 17

[256] after K Kamata T Nakamura K SataBulletin of Tokyo Institute of Technology1974 5416754951(220) 74 ff


migrate through the electrolyte via a vacan-cy mechanism to the anode At the anodehydrogen is electrochemically oxidized towater Each cell delivers a maximum of 1 Vand is typically operated at around 06 to07 V at a power output of typically 250 to450 mWcm2 In SOFC systems many cellsare stacked in series connected with ametallic conducting interconnect

Research and development in the fieldof SOFC are currently concentrating onlowering the operating temperature in orderto reduce costs and increase lifetime and toincrease reliability of the ceramic stack el-ements and interconnects New manufac-turing technologies are demanded whenthin electrolytes are used reducing the elec-trical resistance of the cell Materials aswell as systems development aim also forbetter fuel utilization and higher electricalefficiency

This paper reviews the different possi-ble designs of SOFC cells including thepossibilities when reaction selective elec-trodes are used Commonly used materialsand some of their properties as well as nov-el materials that could be applied in near fu-ture are also reviewed as well as a searchstrategy for those materials

2 Design

The design of a single cell is closely re-lated to the design of an entire stack Be-cause a single cell only delivers 1 V morethan one cell is usually connected in seriesusing interconnects The open circuit volt-age (OCV) of the SOFC ie the voltage ofthe system when no current is flowing cor-responds to the number of individual cellsin the stack Over the last two decadesSOFCs based on yttria-stabilised zirconiahave been developed for an operating tem-perature range of 900ndash1000 degC The advan-tage of the high temperature is that internal

reforming of hydrocarbons is possible di-rectly on the anode without the need for anexternal reformer [1]

One important design criterion for a sol-id oxide fuel cell is the separation of anodeand cathode by the gas tight electrolytePinholes or cracks in the electrolyte cancause the hydrogen to leak to the cathodecompartment where it reacts directly withoxygen This will decrease the open circuitvoltage (OCV) and might even render thefuel cell inoperable The development of asuitable stack sealant still presents a chal-lenging task because the requirements forthe sealants are stringent due to harsh envi-ronments and the high operating tempera-tures Sealing of SOFCs can be done by us-ing bonding seals or pressurized seals Forbonding seals materials like high-B2O3glasses [2] earth-alkali silicate glasses suchas BaOAl2O3SiO2 [3] or glass ceramicsare commonly used Another solution re-lies on compressive seals based on micathat do not bond chemically to the SOFCmaterials [4]

21 Tubular DesignIn the 1960s experimental SOFCs with

planar geometry were evaluated and it wasfound that it is very difficult to obtain ade-quate gas sealing at the edges of the cell

mainly due to the mismatch of thermal ex-pansion coefficient between the electrolytesand support structures and the mechanicalproperties of the sealing materials In orderto overcome these problems a tubular con-figuration (ie cylindrical design) was de-veloped by Westinghouse and taken over bySiemens and improved over the last 20years In this design (Fig 2) the electrolyteand anode are supported on a thick cathodetube that is closed at one end The materialstheir dimensions and fabrication processesare summarized in Table 1 The electrolyteis deposited onto the cathode support afterfabrication of the interconnection In a laststep the anode is applied The gas manifoldof the Siemens-Westinghouse design is il-lustrated in Fig 3 Air is introduced via acentral Al2O3-tube to the end of the cathodetube The oxidant flows back across thecathode while the fuel flows in the same di-rection at the exterior of the tube At theplenum of each cell the depleted flow of airand fuel recombine and the remaining activegases react The generated heat serves topreheat the incoming oxidant stream One ofthe most attractive features of this fuel celldesign is that it eliminates the need for leak-free gas manifolding of the fuel and oxidantstreams in the hot zone The drawback isthat the electric current has to flow along thecircumference of the tube in the anode andthe cathode This increases the length of theconducting path and thus the ohmic resist-ance of the cell as compared to a planar one

Fig 1 Schematic of a solid oxide fuel cell (SOFC) element with anode cathode and electrolyte

Table 1 Materials and fabrication processes for state-of-the-art cathode supported cells of theSiemens-Westinghouse solid oxide fuel cell


Cathode Tube Doped LaMnO3 22 mm Extrusion-sintering

Electrolyte ZrO2(Y2O3) 40 mm Electrochemical vapour deposition

Interconnect Doped LaCrO3 85 mm Plasma spraying

Anode Ni-ZrO2(Y2O3) 100 mm Slurry spraying or electro-chemical vapour deposition

Fig 2 The tubular design from Siemens-West-inghouse


























3 Electrolyte



















T [degC]

1000T [K]

log

s[W

-1cm

-1]












4 Anode



(1)

(2)

(3)


























5 Cathode







(4)

















(CGO 105 [199]) [181] [200] [200]















































































































































































































































































































3 Electrolyte



















T [degC]

1000T [K]

log

s[W

-1cm

-1]












4 Anode



(1)

(2)

(3)


























5 Cathode







(4)

















(CGO 105 [199]) [181] [200] [200]




























































































































































































































































































3 Electrolyte



















T [degC]

1000T [K]

log

s[W

-1cm

-1]












4 Anode



(1)

(2)

(3)


























5 Cathode







(4)

















(CGO 105 [199]) [181] [200] [200]

































































































































































































































































































T [degC]

1000T [K]

log

s[W

-1cm

-1]












4 Anode



(1)

(2)

(3)


























5 Cathode







(4)

















(CGO 105 [199]) [181] [200] [200]

































































































































































































































































































4 Anode



(1)

(2)

(3)


























5 Cathode







(4)

















(CGO 105 [199]) [181] [200] [200]















































































































































































































































































































5 Cathode







(4)

















(CGO 105 [199]) [181] [200] [200]

































































































































































































































































































5 Cathode







(4)

















(CGO 105 [199]) [181] [200] [200]






































































































































































































































































































(CGO 105 [199]) [181] [200] [200]
















































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































Date post:	28-Mar-2020
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times

Solid Oxide Fuel Cells: Systems and Materialsgases react. The generated heat serves to preheat the...

Documents