Orientation relationship and interfaces in Ni and Co-YSZ cermets prepared from directionally...

Cent. Eur. J. Phys. • 7(2) • 2009 • 242-250DOI: 10.2478/s11534-009-0030-z

Central European Journal of Physics

Orientation relationship and interfaces in Ni andCo–YSZ cermets prepared from directionally solidifiedeutectics

Research Article

Angel Larrea∗, Miguel A. Laguna-Bercero† , José I. Peña, Rosa I. Merino, Víctor M. Orera

Instituto de Ciencia de Materiales de Aragón, CSIC- U. Zaragoza, E-50.018 Spain

Received 11 December 2008; accepted 25 February 2009

Abstract: Textured Ni-YSZ and Co-YSZ (YSZ: cubic yttria stabilized zirconia) cermets prepared by reduction of direc-tionally solidified NiO-YSZ and CoO-YSZ oxide eutectics respectively display a self-organized microstuctureformed by∼400 nm wide alternating lamellae of YSZ and porous metal suitable for electrochemical applica-tions. The electrochemical properties of the cermets depend on their microstructure. We have analyzed theorientation relationships and interfaces both of the oxide composites and cermets using Scanning ElectronMicroscopy, Transmission Electron Microscopy, X-ray pole figures and Electron Back-Scattering Diffraction.In spite of the similar crystal structure, growth habits and orientation relationships of NiO-YSZ and CoO-YSZ are different. Also the crystallographic behaviour, when cermets are produced, differs. However themetal–YSZ interfaces are about the most stable ones giving good metal-ceramic adhesion. Due to theirlamellar microstructure and good metal-ceramic adhesion these composites present long-term stability atworking conditions, which makes them good candidates to be used as anodes in solid oxide fuel cells orelectrolyzers.

PACS (2008): 68.35.-p; 68.37.-d; 81.05.Mh; 82.47.Ed

Keywords: directionally solidified eutectics • cermets • fuel cells • microstructure© Versita Warsaw and Springer-Verlag Berlin Heidelberg.

1. Introduction

Directionally solidified eutectics (DSE) are micro- ornano- self-organized ceramic composites characterized bystrong interphase bonding [1]. In the eutectic process aliquid solidifies into two or more solid phases resulting ina fine and homogeneous microstructure, where the mini-∗E-mail: [email protected]†Present address: Department of Materials, Imperial College, SW7 2BP,London, UK

mization of the interfacial energy leads to the formationof low-energy interfaces. The interphase spacing can becontrolled by the growth rate, as it depends on the lat-eral diffusion of species ahead of the solidification front.DSE have been widely studied for decades because oftheir good mechanical properties, as well as their ex-cellent thermal and microstructural stability as comparedwith conventional ceramics [2]. For instance, nano-fibrillarAl2O3-YAG-YSZ (YAG: Ytrium Aluminium Garnet) direc-tionally solidified eutectics prepared by the laser floatingzone (LFZ) method present outstanding flexural strengthof σ = 4.6 GPa at room temperature [3].242

Angel Larrea, Miguel A. Laguna-Bercero1, José I. Peña, Rosa I. Merino,Víctor M. Orera

In addition to Al2O3-based DSE, regular eutectics, withordered microstructure formed by alternate lamellae orrods embedded in a matrix, have shown interesting func-tional applications [4]. For instance CaF2/MgO fibrousdirectionally solidified eutectics are formed by an orderedpacking array of single crystalline optical fibres with adensity of 40,000 fibres/mm2 embedded in the single crys-talline CaF2 matrix [5]. In addition fibrous DSE can beused as templates to prepare micrometer-wide long sin-gle crystalline fluoride fibres by selective leaching of thematrix [6]. The opposite strategy is to produce porouscrystalline structures in the micron-size region by selec-tive leaching of the fibres [7].Probably the most promising functional applications to-day for DSE are in electrochemistry, using them as pre-cursors to prepare structured porous metal-ceramic com-posites. For example, solid oxide fuel cells (SOFC) typ-ically include a porous Ni-YSZ (yttria-stabilized zirco-nia) cermet anode which presents gas permeation andboth electronic and ionic conductivity through metal andYSZ respectively [8]. The properties of these anodes arehighly dependent on their microstructure: the triple phaseboundary (TPB) length (where the electrochemical reac-tion takes place) has to be high [9], and good connectivityis also necessary between the metallic particles and poresto ensure good electronic conductivity and gas permeationrespectively. In this way we reported the preparation of Niand Co-YSZ cermets by reduction of DSE composites andoptimized their microstructure to get the best properties asSOFC anodes [10]. One of the main concerns of this kindof material is their long-term stability during SOFC oper-ation. Metal particles tend to agglomerate during cell op-eration due to their low wettability with the ceramic phase.Metal coarsening decreases the electrochemical anodeperformance lowering the cell power output. We have car-ried out several studies on the microstructural stability ofNi- and Co-YSZ cermets prepared by reduction of DSEand relate their remarkable long-term stability to theirlamellar microstructure and to the presence of low-energymetal-ceramic interfaces [11, 12]. The formation of low-energy oxide-oxide interfaces is thermodynamically drivenby the eutectic growth. The subsequent reducing treat-ment produces porous metal-YSZ cermets where metal andceramic phases remain crystallographycally aligned, re-sulting in the formation of metal-ceramic interfaces withhigh adhesion [13]. Here we review the microstructure ofNiO-YSZ and CoO-YSZ directionally solidified eutecticrods, focussing on the crystallographic orientation rela-tionship and interfacial planes before and after reductionof the transition metal oxide. Microstructural stability ofthe cermets in relation to the metal-ceramic interfaces ofthe material is also discussed.

2. Experimental details

For the material preparation precursor ceramic cylinders2 mm in diameter and 10 cm in length were conformedby cold isostatic pressing from weighted mixtures of NiO(99.99% Sigma-Aldrich) or CoO (99.99% Sigma-Aldrich)and 8YSZ (8 mol% Y2O3 stabilized ZrO2 from Tosoh, 99.9%).The compositions used in this study were 75 mol% NiO –25 mol% YSZ and 80 mol% CoO - 20 mol% YSZ respec-tively. The cylinders were pressureless sintered at 1300°C(12 h) and 1500°C (30 min.). Directionally solidified rodswere produced by the laser floating-zone method in air, forthe NiO-YSZ rods, or under Ar atmosphere for the CoO-YSZ rods (to avoid CoO oxidation to Co3O4). In order toobtain the porous cermets the rods were reduced insidea tubular furnace in a 4% H2-N2 mixture atmosphere at800°C for 1 h. More details about the preparation proce-dure can be found in Ref. [10].Scanning Electron Microscopy (SEM) observations werecarried out in a Jeol 6400 microscope equipped with anINCA300 microanalysis system (Oxford Instruments). TheElectron Back-Scattering Diffraction (EBSD) experimentswere performed using an HKL system (Oxford Instru-ments). For EBSD experiments the samples were groundwith SiC, polished with diamond paste as well as col-loidal silica and finally polished using Ar+ ions in an ionmilling (Gatan 600dif) at low angle (10°). The Transmis-sion Electron Microscopy (TEM) experiments were per-formed using a Jeol 2000FXII microscope equipped withan INCA200 microanalysis system (Oxford Instruments).For these experiments transverse slices (∼300 µm) werecut from the samples using a diamond saw. Then the sliceswere plane-parallel ground down to 40 µm thick using di-amond lapping film, glued to a copper grid and ion milledusing 5 kV Ar+ ions at liquid nitrogen temperatures.X-ray pole figure experiments were carried out in an Xcal-ibur system from Oxford Instruments formed by a kappagoniometer and a 2D CCD detector. For building the polefigures we used the XRD2DScan software [14]. The ex-perimental pole figures were analyzed using the WIMWmethod implemented in the popLA software to renormal-ize them, complete the unmeasured rim and mainly to en-sure that all the pole figures were consistent with eachother [15].Ageing experiments were performed for 300 h at 900°C inthe same tubular furnace as used for reduction. Electri-cal conductivity measurements were carried out using the4-point DC method at room temperature using a Keithley220 current source and a Hewlett-Packard 3441 micro-voltmeter. Mercury porosimetry was performed using aPoremaster-33 from Quantachrome Instruments.243

Orientation relationship and interfaces in Ni and Co–YSZ cermets prepared from directionally solidified eutectics

Figure 1. Optical image of the transverse cross-section of a CoO-YSZ DSE rod. Light phase: CoO; dark phase: YSZ.

3. Experimental results and discus-sion3.1. Microstructure of NiO-YSZ DSE and Ni-YSZ porous cermetsNiO-YSZ and CoO-YSZ eutectic rods display regularlamellar microstructure. Regular eutectics show lamellarmicrostructures instead of fibrillar one when the volumepercentage of the minority phase is higher than about28%. Above this threshold the interfacial area per unitof volume in the lamellar microstructure is lower than inthe fibrillar one. If other factors affect eutectic growth orthe interfacial energy is not isotropic, slight deviationsfrom this general rule can be observed [16]. Samples areformed by domains, called eutectic grains, which are aswell formed by an ordered sequence of NiO (or CoO) andYSZ alternate lamellae (Fig. 1). The size of the eutecticgrains can be tailored by changing the growth conditions.The lamellar interspacing is a function of the growth rate,for growth rates of 100 mm/h we obtained ∼400 nm widelamellae in these compounds.We have analysed the Orientation Relationships (OR) andinterfacial plane by several techniques: X-ray pole fig-ures, TEM and EBSD. Probably EBSD gives the mostillustrative picture. In Fig. 2 we display the EBSD orien-tation map of a eutectic grain of the NiO-YSZ DSE. To ob-tain these maps the sample surface was divided into pixelsand an EBSD pattern was acquired from each pixel. Fur-thermore the patterns were analysed and indexed to getthe crystallographic orientations with respect to the sam-ple coordinate system. As in our material both phases be-long to the same space group (Fm3m) the software is notable to distinguish from which component (NiO or YSZ)the pattern arises, so we used combined energy disper-sive spectroscopy (EDS) analysis. Once the phase and

Figure 2. EBSD orientation map of a eutectic grain in a NiO-YSZrod in transverse cross-section. Dark phase: YSZ; Lightphase: NiO.

the crystallographic orientation of each pixel were known,we assigned them different colours according to a con-ventional scheme [17]. In the orientation map of Fig. 2we observe that each lamella is a single crystal and thatall the lamellae belonging to the same phase are crys-tallographically parallel. As a consequence the eutecticgrain can be viewed as a bicrystal with many interfaces.Once we know the orientation of each point in the mapwe can build the pole figure for any reflection. In Fig. 3we show the {200}, {220} and {111} pole figures of YSZand NiO phases. There we observe that the growth direc-tions are [110]NiO and [110]YSZ, the interfacial plane being(111)NiO//(002)YSZ and with its normal perpendicular to thegrowth direction. For this eutectic the interfacial planemore than the growth direction imposes the lattice cor-relation boundary. The parallelism between (111)NiO and(200)YSZ is nearly exact (under the accuracy of the map:0.5°), whereas [110]NiO and [110]YSZ are not really parallel.They form a 15.6° angle, the plane formed by these twodirections being parallel to the interfacial plane. As bothcomponent phases, NiO and YSZ, have low entropy of fu-sion there is no large difference in growth kinetics as afunction of the growth direction, and the main requirementduring the eutectic growth is to get a good matching atthe interface.The particular OR we have reported here is usuallyobserved in samples grown at relatively high rates(V=100 mm/h). However, other authors have observeddifferent OR because they grew samples at lower growthrates (V=10 mm/h) to obtain well aligned but broadermicrostructures (growth directions [110]NiO//[100]YSZ andinterfacial planes (111)NiO//(002)YSZ) [18, 19]. Neverthe-less we are not interested in lamellar perfection but inpreparing eutectic rods formed by 20-80 µm size slightlymisaligned eutectic grains of ∼400 nm lamellae that,244


Figure 3. Pole figures calculated from the EBSD orientation map of Fig. 2.

after reduction, present a homogeneous distribution ofpores and metal particles and better mechanical proper-ties.After reduction of the transition metal oxide we ob-tained the structured porous cermets (Fig. 4). Poresare formed by the large contraction of the crystallo-graphic cell when NiO transforms to Ni (aNiO=0.418 nm,aNi=0.352 nm). The OR after reduction has been stud-ied both in fully-reduced and partially-reduced sam-ples by TEM, and different behaviour patterns were ob-served. In some cases Ni maintains the same crystal-lographic orientation as in the parent phase, forming(111)Ni//(002)YSZ interfaces. But in addition two differ-ent crystallographic reorientation processes, leading tothe formation of (002)Ni//(002)YSZ interfaces with differ-ent OR ([100]Ni//[110]YSZ and [110]Ni//[110]YSZ), were alsofound in the Ni-YSZ composites (Fig. 5). These reorienta-tion phenomena occur because the Ni atoms re arrange atthe interface during the NiO reduction process to replicatethe YSZ crystalline order and to reduce the interfacial en-ergy. We have shown that reorientation takes place onlyif reduction starts at the interface because the oxygen lost

by NiO passes through the YSZ lamellae along its way tothe reducing atmosphere [13] due to the high ionic conduc-tivity of YSZ at the reduction temperatures. Moreover, wehave also shown that the different interfaces correspondexactly to the best possible matches between Ni and YSZcalculated by Dickey et al. using the near-coincident sitelattice framework several years before our experimentalstudy was accomplished [20].3.2. Microstructure of CoO-YSZ DSE and Co-YSZ porous cermets

As the NiO-YSZ eutectic, the CoO-YSZ DSE microstruc-ture consists of alternate lamellae perpendicular to thesolidification front. However, the orientation relation-ships are different. X-ray pole figures of samples grownat 100 mm/h show that the main growth directions are[111]CoO//[111]YSZ and that the crystallographic axes of thecomponent phases are parallel: {hkl}CoO//{hkl}YSZ. Thisbehaviour is also different from the same eutectic grownat lower rates, as previously reported [21, 22]. In factwe prepared CoO-YSZ rods varying the growth rate but245


Figure 4. SEM image of the fractured area of a Ni-YSZ structuredporous cermet formed by porous Ni and YSZ alternatelamellae.

Figure 5. TEM image of the reorientation process at the YSZ-NiOinterface.

maintaining all other conditions, and we verified that, asfirst suggested by Echigoya and Hayashi [22], the growthdirection changes as a function of growth rate. This isillustrated in Fig. 6 where the {200}CoO, {220}CoO and{111}CoO pole figures measured in CoO-YSZ eutecticsgrown at 10 and 100 mm/h are shown. The YSZ polefigures present similar behaviour. Samples grown at lowrate display the previously reported [110]CoO growth di-rection, whereas samples grown at faster rates show thebefore-mentioned [111]CoO growth direction. In additionslow grown samples present a relatively low mosaicity, therod looking like a single crystal. However, faster grownsamples display fibre type symmetry, with two other mi-nority growth directions: [110] and [100] besides the main[111] growth direction. As X-ray experiments gave onlyaveraged information of the whole sample, orientation re-

lationships were also studied by EBSD to get more localinformation. Fig. 7 shows EBSD patterns from two adja-cent CoO and YSZ lamellae. The patterns are nearly iden-tical, except for the band widths which are proportional tothe inverse of lattice parameter and for the band intensi-ties which are proportional to the modulus of the structurefactor. The symmetry of the patterns is exactly the samebecause both phases are in the same space group and thefact that the band positions are identical indicates that thecrystal reference systems are truly parallel. However, theinterfacial plane is still unknown. In TEM studies of thetransverse cross-sections we observed that the interfacialplane is not exactly parallel to the [111] growth directions.We suspect that probably the lamellae are not growing ex-actly along any low-index crystallographic direction andfor the moment we have not yet been able to identify theinterfacial planes. Detailed studies are underway in orderto clarify this.Reduction behaviour of CoO-YSZ DSE was also studiedby TEM and no reorientation processes were observed inthis case. We only observed the formation of twins in(111)CoO//(002)YSZ interfaces (Fig. 8), but we cannot claimthat these interfaces are the general case in the cermets.The same phenomenon of twin formation has been previ-ously reported by Alem, Dravid and Li [23].The origin of the differences in the growth behaviour andinterfaces in fast growth CoO-YSZ eutectics is still unclearto us. CoO stoichiometry could be a concern since pres-ence of oxygen vacancies may change the energy balanceof the interfaces, but an intrinsic thermodynamic origin ofthese differences can not yet be discarded.3.3. Microstructural stabilityIn order to test their microstructural stability under SOFCworking conditions the samples were maintained undercontinuous flow of dry H2-N2 mixtures for 300 hoursat 900°C [11, 12]. Metal particle agglomeration wasmeasured using SEM observations, mercury intrusionporosimetry and electric conductivity before and after theageing treatment. Although mercury porosimetry did notshow significant increase in pore size after ageing (Fig. 9),we have to be careful when interpreting the porosime-try experiments in lamellar composites. Sometimes inthese experiments we estimate the “bottle-neck” distribu-tion rather than measure the true pore size distribution.In fact, Hg porosimetry is based on the Young-Laplaceequation:

∆p = γ( 1r1 + 1

r2) cosθ,

246


Figure 6. CoO pole figures from the transverse cross-section of CoO-YSZ DSE grown at 10 mm/h (top) and 100 mm/h (bottom).

where γ is the surface tension, r1 and r2 are the mutu-ally perpendicular radii and θ is the angle of contactbetween the liquid and the solid walls. In the case ofporous lamellar cermets one of the radii is constant (r1),because it is mainly the lamellar width. Accordingly, evenif metal coarsening enlarges the pore size (r2 → ∞), weobtain 1r1 + 1

r2 → 1r1 ≈ cte. As Hg porosimetry is not thebest technique to study lamellar composites, we considerthe data from electronic conductivity to be more accurate.Our DC conductivity experiments showed that there is notsignificant metal coarsening, as the conductivity remainsconstant (Tab. 1). Moreover SEM observations confirmedno microstructural evolution due to the thermal treatment.So we concluded that these cermets are resistant to coars-ening under the above mentioned conditions.

We attribute the thermal stability of these cermets to twomain factors. In general bigger particles can grow at theexpense of smaller ones by a process of Ostwald ripening,where diffusion between particles arises from the different

Table 1. Electronic conductivity before and after the ageing treat-ment. Ni-YSZ (S/m) Co-YSZ (S/m)Before ageing (8.9±0.9)×105 (2.7±0.3)×106After ageing (9.0±0.9)×105 (2.6±0.3)×106

particle radii. This mechanism is usually more difficultin eutectics because of the uniform distribution of parti-cle size produced by the eutectic growth. More specifi-cally the YSZ lamellae form a ceramic skeleton that sus-tains the metal particles, acting as a barrier to the growthof metallic particles in the direction perpendicular to thelamellae. However, there are no restrictions to diffusioninside the porous metal lamellae and only a good metal-ceramic bonding can restrain metal diffusion inside theporous metallic lamellae to prevent the coarsening. In247


a)

b)

Figure 7. EBSD patterns obtained from CoO (a) and YSZ (b) adja-cent lamellae. DSE rod grown at 100 mm/h.

structured porous cermets only the combination of YSZlamellae and low-energy interfaces would be able to en-sure good microstructural stability. The relationship be-tween metal-ceramic adhesion and metal coarsening canbe expressed in terms of thermodynamic considerations.The variation of the interfacial energy during coarseningis lower for higher values of the metal-ceramic work ofadhesion, and as a consequence, good metal-ceramic ad-hesion leads to a lower coarsening driving force [13].In the Ni-YSZ cermets we noticed that different kinds ofparticular interfaces, which can be considered as low-energy interfaces, are formed after reduction. This mix-ture of interfaces seems to be appropriate for ensuringgood thermal stability. However, in the Co-YSZ case thesituation is different. The absence of reorientation pro-cesses would indicate that the Co-YSZ interfaces formeddirectly are stable enough and they do not need to evolveFigure 8. TEM image and Convergent Beam Electron Diffraction

(CBED) patterns showing the twin formation at the Co-YSZinterface (top) and atomic model of the twins (bottom).

248


Figure 9. Pore size distribution from mercury porosimetry experi-ments of Co-YSZ cermets (top) and Ni-YSZ cermets (bot-tom) before and after the ageing treatment.

to a new orientation. From our results it appears thatmetal-ceramic interfaces of cermets prepared from DSEpresent, in general, good adhesion properties. The ioniccharge balance at the interface is probably the main re-quirement for the formation of low-energy interfaces inionic compounds. Minford, Bradt and Stubican were thefirst to indicate in this 1979 pioneering article on crystal-lography of directionally solidified oxide eutectics [24] thatinterfacial planes of both phases were low-Miller indicespolar planes with similar surface charge density. Theyattributed the importance of maintaining the charge bal-ance across the interface to a consequence of the Paul-ing’s second rule of ionic structures, indicating that theionic bonds extending from the interfacial plane must becompensated on the polar plane of the other phase. Morerecent ab initio atomistic calculations using density func-tional theory (DFT) by the Muñoz group indicate thatstrong bonding in ionic interfaces occurs when both inter-facial surfaces are polar [25]. As eutectics grow accordingto low-energy interfaces (minimization of the interfacialenergy is the main law ruling eutectic growth) this resultsin the formation of polar-polar interfacial configurations.In fact these DFT calculations support with more sophis-ticated techniques the early observation of Minford and

co-workers. Detailed high resolution TEM studies haveconfirmed that oxide-oxide DSE interfaces are formed fora common oxygen plane, maintaining the cation-anion se-quence across the interface [26]. After reduction of thetransition metal oxide (NiO or CoO), the electric chargeat the remaining ceramic polar surface is not compensatedby the metal surface, resulting then in a high surface re-activity. Again Muñoz and co-workers calculated usingab initio DFT the bonding between Ni-(001) surfaces andpolar and non-polar ZrO2 surfaces, (001) and (110) ZrO2surfaces respectively [25], [27]]. They have shown that onaverage the Ni-O and Ni-Zr distances in the Ni-ZrO2(110) interface are enlarged with respect to the Ni -ZrO2(001) interface, the bond order being about half of thatin the polar Ni-ZrO2 interface. They concluded that theadhesion is higher in the case between metal and polarsurfaces. These considerations would explain why metal-YSZ cermets prepared from DSE, although they do not al-ways present reorientations, display good microstructuralstability.4. ConclusionsWe have prepared textured Ni-YSZ and Co-YSZ cermetsby reduction of NiO-YSZ and CoO-YSZ directionally so-lidified eutectics. The orientation relationship and inter-faces of the composites before and after reduction of thetransition metal oxide have been studied by TEM, X-raypole figures experiments and EBSD. The OR displayed byNiO-YSZ and CoO-YSZ DSE grown at the same rate aredifferent. NiO-YSZ growth directions are [110]YSZ-[110]NiO,whereas CoO-YSZ main growth directions are [111]YSZ-[111]CoO. During NiO reduction the Ni phase reorientedits lattice to form epitaxial interfaces with the YSZ lamel-lae. Reorientation led to the formation of lower energymetal-ceramic interfaces. However, Co-YSZ cermets didnot show any reorientation processes and, at the moment,only the formation of twins at the YSZ interface have beenobserved. Despite this different behaviour during reduc-tion, both cermets present good microstructural stability,as it has been verified by estimating the metal coarseningafter ageing experiments. Possibly this microstructuralstability is a general property of metal-YSZ compositesprepared from DSE. It could be attributed to the formationof interfaces between metals and ceramic polar surfaces.AcknowledgementsFinancial support from projects financed by the SpanishGovernment and Feder program of the European Com-

249


munity: MAT2006-13005-C03-01 and CIT-120000-2007-50 is acknowledged. We also acknowledge valuable helpwith the EBSD experiments from Dr. A. Gholinia and Dr.R. Ellemann-Biltoft at HKL Company in Hobro (Denmark).References

[1] J. LLorca, V. M. Orera, Prog. Mater. Sci. 51, 711 (2006)[2] W. J. Minford, R. C. Bradt, V. S. Stubican, J. Amer.Ceram. Soc. 62, 154 (1979)[3] P. B. Oliete et al., Adv. Mater. 19, 2313 (2007)[4] V. M. Orera et al., Acta Phys. Slovaca 55, 261 (2005)[5] A. Larrea, L. Contreras, R. I. Merino, J. LLorca, V. M.Orera, J. Mater. Res. 15, 1314 (2000)[6] V. M. Orera, A. Larrea, Opt. Mater. 27, 1726 (2005)[7] A. Larrea, V. M. Orera, J. Cryst. Growth 300, 387(2007)[8] N. Q. Minh, J. Am. Ceram. Soc. 88, 563 (1993)[9] E. Ivers-Tiffée, A. Weber, D. Herstritt, J. Eur. Ceram.Soc. 21, 1805, 2001.[10] M. A. Laguna-Bercero, A. Larrea, J. I. Peña, R. I.Merino, V. M. Orera, J. Eur. Ceram. Soc. 25, 1455(2005)[11] M. A. Laguna-Bercero, A. Larrea, R. I. Merino, J. I.Peña, V. M. Orera, J. Am. Ceram. Soc. 88, 3215 (2005)[12] M. A. Laguna-Bercero, A. Larrea, R. I. Merino, J. I.Peña, V. M. Orera, J. Eur. Ceram. Soc. 28, 2325 (2008)[13] M. A. Laguna-Bercero, A. Larrea, J. Am. Ceram. Soc.90, 2954 (2007)[14] A. B. Rodríguez-Navarro, J. Appl. Crystallogr. 40, 631(2007)[15] J. S. Kallend, U. F. Kocks, A. D. Rollet, H.-R. Wenk,Mater. Sci. Eng. A132, 1 (1991)[16] R. L. Ashbrook, J. Am. Ceram. Soc. 60,428 (1977)[17] A. J Schwartz, M. Kumar, B. L. Adams (Eds.), ElectronBackscatter Diffraction in Materials Science (KluwerAcademic / Plenum Publishers, New York, 2000)[18] A. Revcoleschi, G. Dhalenne, Nature 316, 335 (1985)[19] E. C. Dickey, X. Fan, S. J. Pennycook, Acta Mater. 47,4061, 1999.[20] E. C. Dickey, Y. M. Bagiyono, G. D. Lian, S. B. Sinnot,T. Wagner, Thin Solid Films, 372, 37 (2000)[21] B. Bonvalot-Dubois, G. Dhalene, A. Revcoleschi, Ma-terials Research Society Symposium Proceedings,138, 587 (1989)[22] J. Echigoya, S. Hayashi, J. Cryst. Growth, 129, 699(1993)[23] N. Alem, V. P. Dravid, S. Li, J. Mater. Res. 22, 1797(2007)[24] W. J.Minford, R. C. Bradt, V. S. Stubican, J. Am. Ceram.Soc. 62, 154 (1979)

[25] M. C. Muñoz, S. Gallego, J. I. Beltrán, Surf. Sci. Rep.61, 303 (2006)[26] E. C. Dickey, V. P. Dravid, P. D. Nellist, D. J. Wallis,S. J. Pennycook, Acta Mater. 46, 1801 (1998)[27] C. Pecharromán et al., Z. Metallkd. 96, 5 (2005)

250

Orientation relationship and interfaces in Ni and Co-YSZ cermets prepared from directionally...

Documents

Transcript of Orientation relationship and interfaces in Ni and Co-YSZ cermets prepared from directionally...