Eutectic ceramics of the CeO2 – ZrO2 – MgO system produced by laser-assisted directional solidification

Eutectic ceramics of the CeO 2 –ZrO 2 –MgO system produced by laser-assisted directional solidification Alodia Orera, Patricia B. Oliete, Rosa I. Merino, María Luisa Sanju an * Instituto de Nanociencia y Materiales de Arag on (Universidad de Zaragoza-CSIC), Facultad de Ciencias, Universidad de Zaragoza, Pza San Francisco s/n, 50009, Zaragoza, Spain ARTICLE INFO Keywords: Eutectic ceramics CeO 2 –ZrO 2 –MgO Directional solidification CO 2 separation membranes ABSTRACT Oxide eutectics have great potentiality as structural or functional materials, owing to the outstanding properties derived from the eutectic microstructure. Among them, the eutectic of the ZrO 2 –MgO system is particularly noteworthy because of the unusual combination of thermomechanical, optical and electrical properties. In a recent application, Zr 1-δ Mg δ O 2-δ –MgO eutectic oxides have been used to produce porous supports for moltencarbonate based CO 2 separation membranes. Here we explore composite ceramic oxides of the Ce x Zr 1-x O 2 – MgO (x0.5) system with eutectic microstructure, with the motivation that incorporating cerium may enhance the CO 2 permeation properties. Eutectic composites with different cerium content are produced by a laser-assisted directional solidification technique at variable solidification rate, v. In all cases the composite bicrystal consists of two phases, MgO and a fluorite-like (Ce x Zr 1-x ) 1-y Mg y O 2-y phase. A purely fibrilar microstructure is found for x 0.3 at v¼25 mm/h, with MgO fibres embedded within the fluorite-like matrix. The MgO mol% in the eutectic composites decreases from ~53% for x¼0 to ~48% for x¼0.5. X-ray and Raman results evidence long-range ordering in a quasi-tetragonal monoclinic symmetry for x¼0.5. Impedance spectroscopy results are consistent with a change from ionic to mainly electronic conductivity when the atmosphere is changed from air or Ar to 5% H 2 –Ar. 1. Introduction Directionally solidified ceramic oxides with eutectic composition have found a wealth of applications as structural or functional materials, owing to their exceptional microstructural properties [1]. In many cases, the material combination provides better performance than any of its constituents separately, as in the case of mechanical or optical properties [2,3]. Among them, the eutectic of the ZrO 2 –MgO system is particularly noteworthy because of its unusual combination of thermomechanical [4–6], optical [6] and electrical properties [7]. This eutectic, with T e  2170 C and a composition 47%ZrO 2 /53%MgO in mol, gives rise to a solid consisting of a cubic Zr 1-δ Mg δ O 2-δ matrix with MgO fibres embedded in it [4]. With an adequate choice of the solidification rate, micrometre diameter fibres can be obtained [6,7]. The feasibility of acid etching the MgO fibres [8] prompted us to use the Zr 1-δ Mg δ O 2-δ –MgO eutectic as a support for CO 2 -selective supported molten-salt membranes, which resulted in permeability values (1.41 10 10 mol m 1 s 1 .Pa 1 at 815 C), among the highest of this type of devices [9]. In this application, the dense and aligned character of the fibres is exploited to produce, after acid-etching, a porous structure with micrometre-size aligned pores to be filled with molten salts, usually carbonates. The use of MgO as pore former is not new. MgO is used, for instance, as a sacrificial phase in bio-compatible CaO –SiO 2 –MgO –P 2 O 5 composites [10]. In another example, much closer to our work, the porous ceramics resulting from etching MgO in Sm-doped-ceria –MgO composites are used as supports in molten–carbonate-based membranes [11]. Other examples can be found in Ref. [12]. In the most generally accepted understanding of how supportedmolten-salt membranes work, permeation is assisted by the oxide conductivity of the ceramic oxide support. In the Zr 1-δ Mg δ O 2-δ –MgO eutectic the matrix conductivity (2.3 10 3 Scm 1 at 1000 K [7]) is an order of magnitude below that of other oxide ion conductors such as YSZ or rare-earth doped ceria. This induced us to investigate other oxidic fibrilar eutectics which might present improved oxide ion conductivity. On the other hand, it is known that mixed ionic-electronic conductivity (MIEC) in the membrane support enhances the CO 2 permeability and allows simultaneous CO 2 and O 2 permeation [13], which suggests to extend the search of support oxides to those with possible MIEC behaviour. * Corresponding author. E-mail address: [email protected] (M.L. Sanju an). Contents lists available at ScienceDirect Journal of Solid State Chemistry journal homepage: www.elsevier.com/locate/jssc https://doi.org/10.1016/j.jssc.2022.123525 Received 10 June 2022; Received in revised form 14 August 2022; Accepted 21 August 2022 Available online 31 August 2022 0022-4596/©2022 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/). Journal of Solid State Chemistry 315 (2022) 123525 In this context, here we investigate a novel family of eutectic ceramics as candidates to produce CO 2 membrane supports, focusing on the materials production and characterisation. Specifically, we search for compositions of the ternary ZrO 2 –CeO 2 –MgO system displaying eutectic microstructure, assuming that at least for a certain region of compositions a fibrilar eutectic may form with fibres consisting exclusively of MgO and both Mg and Ce entering the matrix to form a (Ce x Zr 1-x ) 1y Mg y O 2-y (CZMO) phase. The motivation behind this work is that introducing Ce in the Zr 1-δ Mg δ O 2-δ matrix might add electronic conduction properties to the membrane support, based on the redox properties of cerium in Ce x Zr 1-x O 2 (CZO) compounds [14] and the mixed conductivity of solid solutions of the ZrO 2 –CeO 2 –Y 2 O 3 system for certain ranges of the oxygen partial pressure (pO 2 )[15,16]. Introducing Ce in the matrix might also mitigate the effect of the reactivity of ZrO 2 with molten Li 2 CO 3 that results in the formation of Li 2 ZrO 3 [17]. As a second derivative, this study may also contribute to throw light about the stability of the fluorite structure in the presence of vacancies, which in this case result either from the Mg insertion into the lattice or from Ce reduction effects [18]. To produce the CZMO –MgO composites we have used the laserassisted directional solidification by the floating-zone technique (LFZ). Although a eutectic point has been proposed to exist in the phase diagram of the CeO 2 –MgO system [19], we were unsuccessful in processing materials with such a high cerium content. Therefore, we limit this work to Ce/(Ce þZr) compositions between 0 and 50%. As the ternary CeO 2 – ZrO 2 –MgO phase diagram has not been reported, for each CZO content the MgO percentage was varied until solidification results in a eutectic microstructure and no pro-eutectic phases were found. For low processing rates (typically 25 mm/h) the obtained eutectic microstructure consists in most cases of a quasi-hexagonal array of MgO fibres embedded in a dense (Ce x Zr 1-x ) 1-y Mg y O 2-y matrix. However, for low Ce content lamellar microstructure was found to coexist with the fibrilar one. For a solidification rate of 25 mm/h the fibre diameter is close to 1.5 μ m. These are very appropriate parameters for its use as membrane supports. 2. Experimental details Rods of Ce x Zr 1-x O 2 –MgO composition, with variable Ce content in CZO and adjustable proportion of CZO and MgO, were directionally solidified using the LFZ method. Ceramic feed rods were prepared by solid state reaction using a mixture of commercial powders of CeO 2 (Aldrich, 99.9%), mZrO 2 (Aldrich, 99%) and MgO (Alfa Aesar, 99.99%) with the appropriate molar composition. Cylindrical rods were fabricated by isostatically pressing the powder for 3 min at 200 MPa and sintering at 1500 C during 12 h in air. Rods were processed in air atmosphere using 80–100 W of a CO 2 laser (λ¼10.6 μ m) as heating source at a solidification rate of 25 mm/h for all compositions. For x¼0.5, additional growth rates in the range 25–200 mm/h were used to develop different microstructures. The final diameter of the solidified rods was between 1.4 and 2.0 mm. To find convenient compositions in Ce x Zr 1-x O 2 –MgO pseudobinary phase diagrams, we fixed xat certain values (nominally x¼0, 0.05, 0.15, 0.30 and 0.50) and then, for each xvalue, the MgO content was varied until a eutectic microstructure was achieved. For simplicity, the samples will be labelled by the nominal Ce content (x) in the Ce x Zr 1-x O 2 compound; the fluorite-like notation (Ce x Zr 1-x ) 1y Mg y O 2-y will be used for the matrix, assuming that all Ce and Zr ions remain in this phase. This assumption is supported by X-ray diffraction (XRD) results, as we shall see. Transverse and longitudinal cross-sections of the processed rods were cut and polished for electron microscopy and Raman analysis. The microstructure was studied using the back-scattered electron micrographs obtained in a Field Emission Scanning Electron Microscope (FESEM, model Carl Zeiss MERLIN). Crystal growth direction was analysed in the same microscope with an integrated electron backscatter diffraction (EBSD) system from Oxford Instruments. Chemical analysis was performed by energy-dispersive X-ray spectroscopy (EDX). Image analysis was made with Digital Micrograph software from Gatan Inc. Routine XRD experiments were carried out on a Rigaku D/max 2500 diffractometer with Cu K α radiation working at 40 kV and 100 mA. Data were collected from 10 to 80in a step mode (Δ2θ¼0.03) and a counting time of 3 s per step. XRD with monochromatic Cu K α 1 line were acquired from 10 to 90in a step mode (Δ2θ¼0.0167) and a counting time of 3 s per step in an XPERT-PRO setup equipped with a Ge monochromator. Lattice parameters were determined with the help of the FullProf package [20]. Raman spectra were recorded using a DILOR XY spectrometer with a CCD detector and 2 cm 1 of spectral resolution. The 496.5 nm line of an Ar þ -ion laser was used as excitation source. The power at the sample was <20 mW. A 50microscope objective lens of an Olympus BH-2 microscope was used both for excitation and dispersed light collection, providing a lateral spatial resolution of 2 μ m. For conductivity measurements, a thick slice was prepared with dimensions 1.5 50.4 mm and large surfaces parallel to the solidification direction. Pt-paste electrodes were painted on both, large rectangular surfaces of the sample and cured at 850 C in ambient air. The shape factor of the electroded sample (surface area/thickness) amounts to 1.18 cm. Conductivity measurements were determined by Impedance Spectroscopy, measured using a SI1260 Frequency Response Analyser from Schlumberger Instruments. The spectra were recorded at different temperatures with excitation voltage amplitude of 200 mV, at frequencies from 1 Hz to 10 MHz. Different temperature runs were made, with the measuring chamber filled first with ambient air, then swept with an Ar-flow and finally with a 5%H 2 –Ar gas mixture flow. Before each run, the sample was equilibrated with the atmosphere at 850 C for at least 2 h. 3. Experimental results and discussion 3.1. Microstructure Fig. 1 (a)-(d) show electron microscopy images of transverse sections of several (Ce x Zr 1-x ) 1-y Mg y O 2-y –MgO composites processed in air at 25 mm/h. In all cases the microstructure is eutectic, consisting of a minority dark phase (MgO) in the form of fibres or lamellae embedded in a clear phase (CZMO). For x0.15, lamellar and fibrilar regions coexist; above 0.15 a regular fibrilar microstructure is found. Eutectic microstructures formed by lamellar and fibrilar morphologies have been previously reported for Zr 1-δ Mg δ O 2-δ –MgO eutectics [21,22]. Incorporating cerium to the composition does not seem to produce changes in the microstructure until only the fibrilar microstructure is settled for high cerium contents. Eutectic interspacings, λ, were obtained in all the samples by applying FFT analysis on SEM micrographs of transverse cross sections. The λ-value slightly decreased from 2 μ m for samples with x¼0 down to 1.7 μ m for those with x¼0.5. The volume fractions of the MgO phase, estimated from the area fraction in SEM micrographs, are presented in Table 1. It is worth noting that fibrilar geometries are favoured in eutectics in which the minority phase presents a volume fraction less than 28% whereas lamellar microstructures form for higher volume percentages [1]. The volume fraction of MgO in the samples with low cerium content (0 x0.15) is near the boundary between lamellar and fibrilar geometries, which explains the mixed fibrilar-lamellar microstructure obtained in these samples. However, for samples containing higher cerium contents the volume fraction of MgO decreases below 28%, resulting in the fibrilar microstructure obtained for samples with x0.3. Samples with x¼0.5 were also processed at different rates (v). For low growth rates a fibrilar microstructure was homogeneously found, but when v50 mm/h a colony-like region developed at the rod centre, surrounded by a regular fibrilar region. Fig. 1(e and f) show images of the fibrilar (e) and colony-like (f) regions found in the transverse cross A. Orera et al. Journal of Solid State Chemistry 315 (2022) 123525 2 section of a Ce 0.5 Zr 0.5 O 2 –MgO sample grown at 100 mm/h. The use of higher solidification rates leads to a non-planar growth, which explains the formation of colonies. We can observe that the microstructure size in the fibrilar regions is highly dependent on the rate used in the solidification, the size of the phases decreasing when the processing rate increases. The eutectic interspacing in the range of growth rates studied follows the Hunt-Jackson law λ 2 v¼Cwith Ca constant depending essentially of the phase diagram and the diffusion coefficient of the ions in the melt [23]. A value of Cof 19.2 μ m 3 /s was obtained from fitting a quadratic law (solid line in Fig. 2) to the experimental data. Cation compositions determined by EDX for samples grown at v¼25 mm/h are collected in Table 1. Two MgO percentages are given: the first one is the molar MgO percentage forming the eutectic in the pseudobinary Ce x Zr 1-x O 2 –MgO system. The second one is the Mg cation content in the matrix. Two trends are evident: as the Ce content (x) in the CZO phase increases, the MgO percentage remains constant, within error, from x¼0tox¼0.30 and decreases for x¼0.50. In parallel, the Mg content in the matrix decreases from ~18% to ~10%. At this respect, we note that measuring accurate cation compositions in the matrix is a hard task, since the proximity of MgO fibres or lamellae raises the Mg determination. To find out the error introduced by nearby MgO phases, we have measured a sample etched with HCl to remove the MgO phase and analysed the resulting porous structure. We thus find that the Mg content in the matrix of etched samples is about 1% lower than in non-etched samples. A similar problem arises when analysing the composition of the dark phase: although it is expected to belong to MgO, the effect of the surrounding matrix always results in an apparent Ce and Zr content in this phase. XRD results shown next will confirm that the minority phase is indeed composed of pure MgO. The second column of Table 1 collects the Ce percentage determined from EDX, relative to the Ce þZr cation content, which can be compared with the nominal one given in the first column. It is clear that cerium Fig. 1. Back-scattered scanning electron micrographs of the transverse cross-sections of (Ce x Zr 1-x ) 1-y Mg y O 2-y –MgO rods grown at 25 mm/h for (a) x¼0.05, (b) x¼ 0.15 (c) x¼0.3 and (d) x¼0.5; (e) and (f) show fibrilar and colony-like regions of the x¼0.5 sample grown at 100 mm/h, respectively. MgO is the dark phase. Table 1 Molar percentage of CZO and MgO in CZMO –MgO eutectic composites solidified at 25 mm/h as a function of the Ce content of the CZO compound, cation composition of the matrix, volume percentage of MgO fibres, and lattice parameter of the fluorite-like matrix (a) both experimentally determined and calculated by means of eq. (1) (Kim's model). A pseudocubic lattice parameter is given for the matrix of the x¼0.5 material, although it is later refined in a monoclinic space group. Ce/(Ce þZr) % nominal Ce/(Ce þZr) % EDX CZO/MgO eutectic mol % Mg cat % matrix Ce cat % matrix Zr cat % matrix MgO Vol % aexp. (Å) acalc. (Å) 0 0 47/53 17.5 0 82.5 30 5.058 5.058 5 1.5 47/53 17.5 1.2 81.3 30 15 8.5 46/54 17 7.7 75.3 29 5.091 5.081 30 21 46/54 16 18 66 26 5.136 5.113 50 44 52/48 10 40 50 21 5.218 5.189 A. Orera et al. Journal of Solid State Chemistry 315 (2022) 123525 3 evaporation was significant in these samples but, interestingly, evaporation was proportionally less severe as the Ce content increased; this trend might be related to the likely decrease of the composite melting temperature, in consonance with the decrease of the melting temperature of CZO compounds predicted by the ZrO 2 –CeO 2 phase diagram [24]. As expected, cerium evaporation also decreased with increasing growth rate for fixed nominal composition. For the x¼0.5 compound processed at 200 mm/h, for instance, the Ce/(Ce þZr) percentage reached a value of 49 1%, very close to the nominal one. In parallel with the increase of the Ce content, the Mg content in the matrix decreased slightly with v, reaching 8% for the x¼0.5 samples processed at 200 mm/h. 3.2. XRD The phase content of (Ce x Zr 1-x ) 1-y Mg y O 2-y –MgO composite samples grown at 25 mm/h was analysed by XRD (Fig. 3) and lattice parameters were derived with the help of FullProf software. In all cases the patterns can be fitted with two main phases: a fluorite-like one accounting for the (Ce x Zr 1-x ) 1-y Mg y O 2-y matrix and a rock-salt one accounting for the minority phase. In some cases, a residual amount of CeO 2 could be detected on the rod surface and CeO 2 phase was accordingly included in the fitting. Some profile fittings are shown as supplementary information. The lattice parameter of the minority phase (4.213(5) Å) is the same for all samples, within error, and belongs to MgO, supporting the pure MgO composition of that phase. On the contrary, the lattice parameter of the matrix (a, see Table 1) changes according to its variable stoichiometry and, as expected, increases with the Ce content, because of the larger ionic radii of Ce 4þ compared to Zr 4þ (see right inset). To get more insight into the relation between the composition of the matrix and the long-range structure, we have combined the evolution of the lattice parameters determined by XRD with the cation composition derived from EDX. Then we compare the experimental lattice parameter with that resulting from the expression given by Kim et al. [25] for different fluorite-like hosts containing aliovalent or homovalent dopants, a(nm) ¼0.512 þΣ k [0.0212*(r k -r h )þ0.00023*(z k -z h )] *m k (1) In expression (1) r k -r h stands for the difference between the ionic radii of the k dopant and the Zr host, in eight-fold coordination, z k -z h accounts for the valence difference between the dopant and the host, and m k is the mole percent of the k dopant [25]. Table 1 displays the lattice parameters calculated by means of expression (1), using the Ce and Mg mole percentages determined from EDX analyses and Shannon's ionic radii [26]. The Ce 4þ oxidation state was assumed for all samples, because of the absence or irrelevant intensity of Ce 3þ electronic Raman bands (see next subsection). Except for the x¼0 case, the experimental values are systematically larger than the calculated ones, the difference increasing with Ce content. We attribute this discrepancy to the assumption of a common, eight-fold coordination, for all cations. It is likely that the preference of each cation type for a given coordination number and how this determines the interconnection between cation polyhedra [27] also weaken the validity of Kim's relation, especially when three different cation types are present. At this point we note that the pattern of the x¼0.5 sample in Fig. 3(a) presents some splitting of the main, fluorite-like substructure peaks that cannot be accounted for by the cubic Fm3mspace group (see for instance the (220) peak around 50in the left inset of Fig. 3(a)). Peak splitting might arise either from phase coexistence or from symmetry lowering. Disambiguing between both possibilities is not easy, because of the overlap between the patterns arising from the two Cu K α lines. To overcome this situation, we acquired a diffractogram with monochromatic K α 1 line. As shown in Fig. 4, this pattern evidenced further splitting of the quasi-fluorite phase reflections that was unnoticed in the conventional X-ray measurement. Two models were tried. In the first one we assume the coexistence of two cubic phases with close lattice parameters, that might arise from spatially inhomogeneous composition. In the second one, a single phase with lower-than-cubic symmetry was assumed. Although the fit with two cubic phases might be acceptable for the normal K α X-rays, the fit of the monochromatic pattern was unsatisfactory. It could not, for instance, reproduce the more-than-two peaks Fig. 2. Dependence of the eutectic interspacing, λ, on the growth rate, v, for the rods with x¼0.5. Solid circles correspond to the experimental values and the line corresponds to the fitting to the Hunt-Jackson law with C¼19.2 μ m 3 /s. Fig. 3. (a) XRD patterns of (Ce x Zr 1-x ) 1-y Mg y O 2-y –MgO eutectic composites grown at 25 mm/h. The asterisks denote the main reflections of the MgO phase. The inset on the left shows the split aspect of the (220) pseudo-fluorite peak in the pattern of the x¼0.5 sample and the inset on the right shows the dependence of the lattice parameter of the matrix (a) on the Ce content. (b) XRD patterns of the x¼0.5 material as a function of the growth rate. The inset shows a magnification of the (220) pseudo-fluorite peak. A. Orera et al. Journal of Solid State Chemistry 315 (2022) 123525 4 aspect of the (311) reflection around 58(third inset in Fig. 4) neither explain the different proportion between the low and high angle components of the (111) and (220) pseudocubic peaks (first and second insets in Fig. 4). Assuming a symmetry lowering, the Dicvol06 tool was used to assist in the determination of the unit cell. It provided a high figure of merit to a monoclinic unit cell of a F /√2xa F xa F /√2 type (pseudofluorite lattice parameter a F ~ 5.2), with βslightly smaller than 90and P2/mspace group. The Dicvol solution was used as a starting point to fit the matrix phase in the x¼0.5 pattern of Fig. 3(a), which led to very good agreement factors (R p ¼15.7, R wp ¼13.2, R e ¼18.1 and Х 2 ¼0.53) with a¼3.6922(3) Å, b¼5.2263(4) Å, c¼3.6855(2) Å and β¼ 89.897(4). The profile fit is given in Fig. S3 of the supplementary material. The unit cell of the distorted fluorite is thus closely related to that of t’–ZrO 2 and related systems, which is supported by the Raman measurements presented in the following section. The symmetry lowering with respect to the cubic one may be ascribed to the lower Mg content entering the matrix for this particular composition, as compared with samples with less Ce content, and to the similar Zr and Ce content, which makes this composition not far from Ce 0.5 Zr 0.5 O 2 , a compound known to present tetragonal symmetry under certain synthesis conditions [24]. It has to be noted that no peak splitting was found when the x¼0.5 compounds were processed at higher growth rates (Fig. 3(b)). The increase of the cubic or pseudocubic lattice parameter with vin these samples (from a¼5.218(1) Å at 25 mm/h to 5.241(1) Å at 200 mm/h) is ascribed to the higher Ce content, because of reduced Ce evaporation, as explained before. 3.3. EBSD The crystal growth direction of each phase and the relative orientation between the two observed phases in the x¼0.5 eutectic solidified at 25 mm/h were analysed by EBSD. To simplify the analysis, and considering that the monoclinic distortion is small, cubic symmetry was assumed both for the fluorite matrix and the MgO fibres. Fig. 5 shows the pole figures measured on a transverse section of the solidified rod. Both sets of figures are remarkably similar, implying that both phases present identical orientation, i.e. [100] F //[100] M , [110] F //[110] M and [111] F // [111] M , where the F and M subindices stand for the fluorite matrix and the MgO fibres, respectively. Moreover, these figures show that crystal growth takes place along a <110>like direction for both phases, the small misorientation being attributed either to tilting of the transverse cut or to the unavoidable curvature of the solidification front. Inverse pole figures, shown in Fig. S4 of the supplementary information, support that the crystal growth direction (normal to the measurement plane) is very close to a <110>axis and that the principal directions are parallel in both phases. Both the growth direction and the relative phase orientation are similar to those reported for the ZrO 2 –MgO eutectic in Ref. [21], where samples were solidified at 20 mm/h. On the other hand, (111) growth was found by Kennard et al. [4] for ZrO 2 –MgO eutectics grown at 10 mm/h. Kennard et al. also state that “at higher solidification rates, …a mixture of (111), (110), and (100) growth directions occurred for both MgO and ZrO 2 ”. 3.4. Raman The sensitivity of Raman spectroscopy to structural varieties and disorder effects in zirconia and related oxides is evidenced in the spectra shown in Fig. 6. For a given composition, no significant differences were found between spectra recorded in fibrilar and lamellar regions, which supports that the cation content is the same irrespective of the microstructure. The spectra in the 100-900 cm 1 region can be ascribed to lattice vibrations of the matrix, since MgO has the rock-salt structure and is therefore Raman inactive. The broad aspect and the wavenumbers of Fig. 4. XRD pattern of the x¼0.5 eutectic composite, grown at 25 mm/h, acquired with monochromatic Cu K α 1 line. The left, central and right insets evidence the peak splitting of the (111), (220) and (311) pseudo-fluorite reflections, respectively. Fig. 5. Pole figures belonging to the fluorite matrix (upper part) and MgO fibres (lower part) of the x¼0.5 eutectic material solidified at 25 mm/h, resulting from EBSD measurements onto a ~14 8 μ m 2 region of the transverse section of the solidified rod. A. Orera et al. Journal of Solid State Chemistry 315 (2022) 123525 5 the observed bands in this region are indicative of a highly disordered lattice, in agreement with the high degree of cation mixing and the presence of vacancies in the (Ce x Zr 1-x ) 1-y Mg y O 2-y phases. A clear difference is observed, however, between the spectra of the x¼0.5 compound and those of compositions with lower xand is attributed to the lowerthan-cubic long-range symmetry of the former seen by XRD. In fact, spectrum 6(e) looks like a broadened version of the spectrum of tetragonal Ce x Zr 1-x O 2 compounds with x(Ce) ~ 0.6–0.65, i.e. close to the t’–t’’ transition [28]. The similarity with t’or t’’-like spectra is not surprising, considering the high Ce content of this composition and the small magnitude of the monoclinic distortion; the broader aspect may be ascribed to the presence of oxygen vacancies. It is interesting that the spectrum of the x¼0.5 sample is also very similar to that of (CeO 2 ) 1-x (8YSZ) x compounds with x~ 0.6 [29], whose expanded composition Ce 0.38 Zr 0.53 Y 0.09 O 1.95 is analogous to that of our x~ 0.5 sample. On the other hand, the spectra of CZMO samples with x<0.5 depicted in Fig. 6 are not characteristic of a cubic fluorite structure, in which only one band would be expected for a periodic, well-ordered lattice. The spectra for x¼0 and 0.05 are similar to that of cubic zirconia stabilised by the introduction of aliovalent dopants (Y 3þ ,Ca 2þ , etc.), containing a high vacancy concentration, which is also our case because of the relatively high Mg 2þ content. In such cases the spectrum resembles a density of states [30], owing to the highly disordered atomic distribution, and its most characteristic feature is a band around 600 cm 1 , which can be attributed to a Zr–O bond stretching vibration with Zr in seven-fold coordination. As seen in Fig. 6, as the Ce content increases two trends are observed: a softening of the 600 cm 1 band, which is related to the lattice expansion produced by the large Ce 4þ ions, and the enhancement of a broad band centred around 350 cm 1 . In fact, the two-broad-band spectrum of the x¼0.3 sample is reminiscent of spectra observed in some highly disordered A 2 B 2 O 7 pyrochlores in their path toward amorphization or defect-fluorite structure upon increasing disorder (see Ref. [31] and references therein). Although such systems may present a defect-fluorite XRD or neutron diffraction pattern, short-range ordered (SRO) domains with orthorhombic weberite-like symmetry have been detected by techniques more sensitive to local-order, such as neutron pair-distribution-function (pdf) analysis [27,32,33]. The SRO domain size is typically 7–9 Å, which is at the unit-cell scale, and the weberite ordering arises from the cation preference for sites with 8, 7 or 6–fold coordination, according to their different ionic radii. Although no evidence of cation ordering has been found in our case, the striking similarity of spectrum 6(d) with those of weberite-containing systems deserves some discussion. An alternative model has been proposed to interpret the neutron pdf without resorting to weberite domains [34]. The key point is the small size of the alleged domains (below 1 nm), which would in fact prevent speaking of real weberite domains. The situation would be better described as consisting of different types of cation polyhedra with 8, 7 and 6–fold coordination [34], whose linking would be governed by Pauling rules at a short length scale of several units [27]. In this model, Raman spectra are not related to a specific lattice symmetry but just to coordination polyhedra and bond distances. In our case, for instance, the bands appearing at high wavenumbers (600 cm 1 ) could be ascribed to bond-stretching of Zr–O polyhedra, whereas Zr–O and Ce–O bond-bending vibrations would show up at 350 cm 1 .Ce–O bond stretching might appear in between these two bands or overlapped with the low wavenumber one, which would explain the enhancement of that band with the Ce content. In the absence of more precise structural determination, this model provides a simple explanation of the Raman spectrum as a function of increasing Ce content before the establishment of a long-range order for x¼0.5. The effect of the growth rate on the Raman spectra of the x¼0.5 composition was also analysed. Fig. 7 shows the spectra of x¼0.5 samples processed at 25, 100 and 200 mm/h. The similarity of all spectra is striking, considering that only the sample grown at 25 mm/h presents a lower-than-cubic long-range symmetry (Fig. 3(b)). We must then conclude that the characteristic size of the low-symmetry regions in these quasi-cubic fluorites is too small to be detected by XRD but enough to be seen by Raman scattering. Raman spectroscopy has also been used to ascertain whether Ce cations are in the Ce 4þ oxidation state or partially reduced to Ce 3þ . This can be done by searching the presence of electronic Raman bands pertaining to the transition between the ground state 2 F 5/2 manifold of Ce 3þ and the first excited multiplet 2 F 7/2 at E≳2100 cm 1 [35,36]. No band was detected in that region for any sample grown at 25 mm/h, with the exception of the x¼0.15 case, where a weak band was found. We attribute this behaviour to the air processing atmosphere and also to the slow growth rate allowing oxidation during cooling. In agreement with this hypothesis, weak electronic Raman bands were found for x¼0.5 at higher growth rates. 3.5. Conductivity Nyquist plots of the impedance measured at temperatures around 450 C and 750 C in the sample (Ce x Zr 1-x ) 1-y Mg y O 2-y –MgO (x¼0.3) processed at 25 mm/h are given in the supplementary information, Fig. S5. They were measured with the AC electric field perpendicular to the solidification direction and in the three different atmospheres. At the lower temperature, all three spectra (in air, Ar or 5%H 2 –Ar), show a high Fig. 6. Raman spectra of (Ce x Zr 1-x ) 1-y Mg y O 2-y compounds in eutectic composites grown at 25 mm/h, for x¼0 (a), 0.05 (b), 0.15 (c), 0.3 (d) and 0.5 (e). Fig. 7. Raman spectra of x¼0.5 samples grown at 25 (a), 100 (b) and 200 (c) mm/h. A. Orera et al. Journal of Solid State Chemistry 315 (2022) 123525 6 frequency, slightly depressed arc, whose equivalent capacitance is around 4 pF, essentially independent of temperature. The equivalent capacitance is estimated with the relationship C eq ¼1/(2 π f M R 0 ), where f M is the frequency at the maximum of –Im(Z) and R 0 is the resistance at the low frequency end of the arc. The value of the capacitance tells that this HF arc encloses the electrical response of the sample. Therefore, the Re(Z) at its low-frequency end toward the x-axis (R 0 ) gives the sample resistance from which the material conductivity values given in Fig. 8 have been estimated. Another compounded arc is observed at lower frequencies (LF). This LF contribution appears as a small spike at 450 C for the measurements in air and as a largerspike for the measurement in argon. At 750 C the LF arc occupies a large portion of the spectra, again with a larger resistance for the measurement in argon. Its capacitance is larger than 1 μ F in all cases, and therefore it can be assigned to electrode processes. The higher resistance of the LF arc in argon than in air is in agreement with the conductivity of the material being mainly ionic, oxide ion conduction. The pO 2 of the used argon atmosphere is between 10 3 to 10 4 bar, much smaller than the 0.2 bar of ambient air. The LF arc in the impedance spectra measured with the reducing gas mixture at T<500 C has a much smaller resistivity, and smaller equivalent capacitance in the temperature range where it can be estimated. Equivalently, the characteristic frequency of this LF contribution is higher for the reduced sample implying faster electrode processes. Having used metallic electrodes, this is most probably associated to a change of the majority charge carriers in the sample, which has now dominant electronic conduction. It is well known that lower-valentdoped ZrO 2 –CeO 2 mixtures show oxide ion conductivity at high oxygen partial pressures and n–type electronic conductivity in reducing atmospheres [15,16,37–39]. Fig. 8 shows that the sample conductivity is the same in air or in Ar. The activation energy for σ T¼Aexp(-E/kT) is 1.35 0.01 eV at T<550 C and 1.28 0.01 eV at T>550 C. At temperatures below ~ 600 C, the present composite is better conductor than the x¼0 eutectic [7,9] and has smaller activation energy. This is consistent with the lower oxygen vacancy concentrations in the present material. From the data given in Table 1, the oxygen vacancy concentration in the conducting matrix as a result of Mg doping would be 0.0875 in the material with x¼0, and 0.08 in the material with x¼0.3. In both cases, the oxygen vacancy concentration is higher that the optimum one in CeO 2 or ZrO 2 doped fluorites, which is around 0.05. For higher or lower concentration of oxygen vacancies, the ionic conductivity decreases and the activation energy increases [40,41]. Eufinger et al. [39] measured the conductivity of (Ce 0.25 Zr 0.75 ) 0.8 Y 0.2 O 1.9 crystals and observed at 0.21 bar an activation energy for σ Tof approx. 1.2 eV, with conductivity of 3 10 3 Scm 1 at 700 C. In the present case, the estimated conductivity is somewhat smaller, 8  10 4 Scm 1 . We note that we are dealing with a fibrilar composite, and measuring the conductivity with the electric field perpendicular to the fibres. A simple Maxwell-Garnet approximation tells that for a conducting matrix and non-conducting fibres with filling fraction ¼0.26, the composite resistivity at low frequencies is 1.7 times the one of the conducting matrix, which allows estimating the conductivity of the matrix to 1.4 10 3 Scm 1 at 700 C, closer to the value of [39]. A higher activation energy and a lower conductivity is in accord with a higher amount of oxygen vacancies, above the concentration which is considered optimum for doped CeO 2 or ZrO 2 fluorite solid solutions, and a larger lattice distortion caused by the ionic radius misfitofMg 2þ with Zr 4þ or Ce 4þ , when compared to Y 3þ . The conductivity measured in 5%H 2 –Ar shown in Fig. 8 has an apparent activation energy for σ ¼Aexp(-E/kT) of 0.56 eV at T>400 C and is higher (lower) than in air or Ar atmospheres at temperatures below (above) 650 C. We know from previous experiments that this gas mixture attains pO 2 values 10 21 bar at temperatures 500–700 C. At these low O 2 activities, different authors have observed a plateau in the conductivity vs pO 2 curve, or at the higher temperatures even a strong decrease of the conductivity with decreasing pO 2 [15,16,38,39]. This behaviour appears to be a consequence of the hopping n–type conductivity with diminished electron mobility at low pO 2 due to site restriction effects. Conductivity values at very low oxygen partial pressures and high temperatures smaller than the ionic conductivity measured in air are in agreement with the literature. To sum up, the solidified eutectic material of composition (Ce x Zr 1-x ) 1y Mg y O 2-y –MgO (x¼0.3) shows dominant oxygen ion conductivity at pO 2 between 10 3 and 0.2 bar, with conductivity values higher than the eutectic without CeO 2 (x¼0) in most of the temperature range studied. As might be expected, upon reduction, the conductivity becomes mainly electronic, most probably n–type. In the intermediate pO 2 range mixed conduction is therefore expected. 3.6. Etching and infiltration tests Preliminary tests have been made in relation with the potential application of these materials as membrane supports. Following the etching method described in Ref. [9] (suspension in 1 M HCl at 60 C for several days) a minimum of 80% weight loss of the MgO phase was achieved in the x¼0.15 eutectic samples. An electron micrograph of an etched sample is shown in Fig. S6 of the supplementary material. Infiltration of molten carbonates in the resulting porous matrix by capillarity was also successful. 4. Summary and conclusions We have presented a structural, microstructural, spectroscopic and conductivity study of eutectic composites of the Ce x Zr 1-x O 2 –MgO system (x0.5). Eutectic-like composites have been produced by a laserassisted directional solidification technique as a function of the Ce content and growth rate v. The MgO mol% in the composite materials decreases from ~53% for x¼0 to ~48% for x¼0.5. In all cases the eutectic bycrystal consists of two phases, one of them being MgO and the other one a (Ce x Zr 1-x ) 1-y Mg y O 2-y fluorite-like phase. For v¼25 mm/h and x 0.15 a mixed fibrilar/lamellar microstructure is found, whereas only fibrilar eutectics are found for x0.3 at the same growth rate, with MgO fibres embedded within the fluorite-like matrix. The Mg content in the fluorite-like matrix decreases from ~17.5 cat% for x¼0 to ~10 cat% for x¼0.5 and v¼25 mm/h. Rods with x¼0.5 have been processed at increasing growth rates and colonies were found at the rod centre, Fig. 8. Electrical conductivity of the (Ce x Zr 1-x ) 1-y Mg y O 2-y –MgO (x¼0.3) sample solidified at 25 mm/h, measured perpendicular to the solidification direction in air (red circles), argon (green up triangles) and in 5%H 2 –Ar (blue down triangles). For comparison, the data for the x¼0 eutectic are given with thin discontinuous lines [7]. A. Orera et al. Journal of Solid State Chemistry 315 (2022) 123525 7 surrounded by fibrilar regions, for v50 mm/h. Below x¼0.5 the matrix presents a cubic-fluorite symmetry in XRD, with the cubic phase stabilised by the introduction of Ce and Mg dopants. In the x¼0.5 eutectic, peak splitting in XRD denotes a symmetry lowering that is well explained by a monoclinic unit cell, closely related to the tetragonal phases of the CZO system. Raman spectra support the symmetry lowering. Conductivity measurements of the composite with x¼0.3 shows total conductivity at high pO 2 higher than the one without cerium in most of the temperature range, with smaller activation energy. In agreement with doped CeO 2 –ZrO 2 mixtures, at low oxygen activity the conductivity is mainly by electronic carriers. In summary, although eutectic composites of the Ce x Zr 1-x O 2 –MgO system present similar structural and microstructural properties to ZrO 2 –MgO eutectics, the presence of electronic conduction under reducing conditions paves the way to additional applications involving mixed ionic-electronic conductivity, in particular in molten-salt supported membranes. CRediT authorship contribution statement Alodia Orera: Conceptualization, Methodology, Formal analysis, Writing –review &editing. Patricia B. Oliete: Conceptualization, Methodology, Formal analysis, Writing –review &editing. Rosa I. Merino: Conceptualization, Methodology, Formal analysis, Writing –review &editing. María Luisa Sanju an: Conceptualization, Methodology, Formal analysis, Funding acquisition, Project administration, Writing – original draft, Writing –review &editing. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Data availability Data will be made available on request. Acknowledgments The authors acknowledge financial support from the Spanish Ministerio de Ciencia e Innovaci on and Feder Funds from the European Comission under projects PID2019-107106RB-C32 and PID2021124863OB-I00 funded by MCIN/AEI/10.13039/501100011033. The Departamento de Ciencia, Universidad y Sociedad del Conocimiento is acknowledged for the financial support to the research group T02_20R. A. O. also acknowledges financial support from the Ram on y Cajal program (RYC2018-025553-I) funded by MCIN/AEI/10.13039/ 501100011033. We thank the Servicio General de Apoyo a la Investigaci on (Universidad de Zaragoza) for technical support in X-ray diffraction and electron microscopy experiments, and the Unidad de Difracci on de Rayos X of the Universidad Complutense de Madrid for assistance with monochromatic X-ray diffraction. We also thank Dr. Wilmer Bucheli and Rub en Gotor for assistance in sample preparation. Appendix A. Supplementary data Supplementary data to this article can be found online at https://do i.org/10.1016/j.jssc.2022.123525. References [1] J. Llorca, V.M. Orera, Directionally solidified eutectic ceramic oxides, Prog. Mater. Sci. 51 (2006) 711–809. [2] M.C. Mesa, P.B. Oliete, J.Y. Pastor, A. Martín, J. LLorca, Mechanical properties up to 1900 K of Al 2 O 3 /Er 3 Al 5 O 12 /ZrO 2 eutectic ceramics grown by the laser floating zone method, J. Eur. Ceram. Soc. 34 (2014) 2081–2087. [3] P.B. Oliete, A. Orera, M.L. Sanju an, R.I. Merino, Selective thermal emission of directionally solidified Al 2 O 3 /Y 3-x Er x Al 5 O 12 eutectics: influence of the microstructure, temperature and erbium content, Sol. Energy Mater. Sol. Cells 174 (2018) 460–468. [4] F.L. Kennard, R.C. Bradt, V.S. Stubican, Directional solidification of ZrO 2 -MgO eutectic, J. Am. Ceram. Soc. 57 (1974) 428–431. [5] V.M. Orera, R.I. Merino, J.A. Pardo, A. Larrea, J.I. Pe~ na, C. Gonz alez, P. Poza, J.Y. Pastor, J. Llorca, Microstructure and physical properties of some oxide eutectic composites processed by directional solidification, Acta Mater. 48 (2000) 4683–4689. [6] D. Sola, P.B. Oliete, R.I. Merino, J.I. Pe~ na, Directionally solidified Ni doped MgOMgSZ eutectic composites for thermophotovoltaic devices, J. Eur. Ceram. Soc. 39 (2019) 1206–1213. [7] R.I. Merino, V.M. Orera, J.I. Pe~ na, G.F. de la Fuente, Conductivity anisotropy in directionally solidified CaZrO 3 -CaSZ and MgO-MgSZ eutectics, Solid State Ionics 100 (1997) 313–318. [8] Yoshikazu Suzuki, Toyoaki Yamada, Shunsaku Sakakibara, Tatsuki Ohji, Unidirectionally porous oxides prepared using eutectic reactions, in: 24th Annual Conference on Composites, Advanced Ceramics, Materials, and Structures: B, Ceramic Engineering &Science Proceedings, vol. 21, The American Ceramic Society, 2000, p. 19. [9] L. Grima, G.A. Mutch, P.B. Oliete, W. Bucheli, R.I. Merino, E.I. Papaioannou, J.J. Bailey, M.D. Kok, D.J.L. Brett, P.R. Shearing, I.S. Metcalfe, M.L. Sanju an, High CO 2 permeability in supported molten-salt membranes with highly dense and aligned pores produced by directional solidification, J. Membr. Sci. 630 (2021), 119057. [10] L. Grima, M. Díaz-P erez, J. Gil, D. Sola, J.I. Pe~ na, Generation of a porous scaffold with a starting composition in the CaO–SiO 2 –MgO–P 2 O 5 system in a simulated physiological environment, Appl. Sci. 10 (2020) 31–46. [11] A. Gili, B. Bischo, U. Simon, F. Schmidt, D. Kober, O. G€ orke, M.F. Bekheet, A. Gurloet, Ceria-based dual-phase membranes for high-temperature carbon dioxide separation: effect of iron doping and pore generation with MgO template, Membranes 9 (2019) 108–123. [12] Y. Suzuki, P.E.D. Morgan, Mesoand macroporous ceramics by phase separation and reactive sintering methods, MRS Bull. 34 (2009) 587–591. [13] Z. Rui, M. Anderson, Y.S. Lin, Y. Li, Modeling and analysis of carbon dioxide permeation through ceramic-carbonate dual-phase membranes, J. Membr. Sci. 345 (2009) 110–118. [14] A. Trovarelli, P. Fornasiero (Eds.), Catalysis by Ceria and Related Materials, second ed.Catalytic Science Series, vol. 12, Imperial College Press, London, 2013. [15] B. Cal es, J.F. Baumard, Mixed conduction and defect structure of ZrO 2 -CeO 2 -Y 2 O 3 solid solutions, J. Electrochem. Soc. 131 (1984) 2407–2413. [16] H.H. Arashi, H. Naito, M. Nakata, Electrical properties in the ZrO 2 -CeO 2 -Y 2 O 3 system, Solid State Ionics 76 (1995) 315–319. [17] J. Ida, Y.S. Lin, Mechanism of high-temperature CO 2 sorption on lithium zirconate, Environ. Sci. Technol. 37 (2003) 1999–2004. [18] M. Yashima, N. Ishizawa, M. Yoshimura, Application of an ion-packing model based on defect clusters to zirconia solid solutions: II, applicability of vegard's law, J. Am. Ceram. Soc. 75 (1992) 1550–1557. [19] N. Bochvar, Y. Liberov, O. Fabrichnaya, Cerium –magnesium –oxygen, in: G. Effenberg, S. Ilyenko (Eds.), Landolt-B€ ornstein New Series IV/C4, SpringerVerlag, 2007, p. 230. [20] J. Rodriguez-Carvajal, FULLPROF: a program for rietveld refinement and pattern matching analysis, in: Abstracts of the Satellite Meeting on Powder Diffraction of the XV Congress of the IUCr, 1990, p. 127. Toulouse, France. [21] J. Echigoya, H. Suto, S. Hayashi, Growth morphology and orientation relationships in directionally solidified MgO-ZrO 2 eutectic, Trans. Jpn. Inst. Met. 26 (1985) 895–900. [22] L. Mazerolles, D. Michel, M.J. Hÿtch, Microstructures and interfaces in directionally solidified oxide-oxideeutectics, J. Eur. Ceram. Soc. 25 (2005) 1389–1395. [23] K.A. Jackson, J.D. Hunt, Lamellar and rod eutectic growth, Trans. Metall. Soc. AIME 236 (1966) 1129–1142. [24] M. Yashima, K. Morimoto, N. Ishizawa, M. Yoshimura, Diffusionless tetragonal–cubic transformation temperature in zirconia–ceria solid solutions, J. Am. Ceram. Soc. 76 (1993) 2865–2868. [25] D.-J. Kim, Lattice parameters, ionic conductivities, and solubility limits in fluoritestructure MO 2 oxide [M ¼Hf 4þ ,Zr 4þ ,Ce 4þ ,Th 4þ ,U 4þ ] solid solutions, J. Am. Ceram. Soc. 72 (1989) 1415–1421. [26] R.D. Shannon, Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides, Acta Crystallogr. A32 (1976) 751–767. [27] E.C. O'Quinn, K.E. Sickafus, R.C. Ewing, G. Baldinozzi, J.C. Neuefeind, M.G. Tucker, A.F. Fuentes, D. Drey, M.K. Lang, Predicting short-range order and correlated phenomena in disordered crystalline materials, Sci. Adv. 6 (2020), eabc2758. [28] M. Yashima, H. Arashi, M. Kakihana, M. Yoshimura, Raman scattering study of cubic-tetragonal phase transition in Zr 1x Ce x O 2 solid solution, J. Am. Ceram. Soc. 77 (1994) 1067–1071. [29] N.M. Sammes, G.A. Tompsett, Z. Cai, The chemical reaction between ceria and fully stabilised zirconia, Solid State Ionics 121 (1999) 121–125. [30] V.G. Keramidas, W.B. White, Raman-scattering study of crystallization and phasetransformations of ZrO 2 , J. Am. Ceram. Soc. 57 (1974) 22–24. [31] M.L. Sanju an, Raman spectroscopy study of disorder phenomena and size effects in pyrochlores, in Pyrochlore Ceramics: properties, Processing and Applications, in: Anirban Chowdhury, Elsevier Series on Advanced Ceramic Materials, Elsevier, 2022, ISBN 9780323904834. [32] J. Shamblin, M. Feygenson, J. Neuefeind, C.L. Tracy, F. Zhang, S. Finkeldei, D. Bosbach, H. Zhou, R.C. Ewing, M. Lang, Probing disorder in isometric pyrochlore and related complex oxides, Nat. Mater. 15 (2016) 507–511. A. Orera et al. Journal of Solid State Chemistry 315 (2022) 123525 8 [33] D.L. Drey, E.C. O'Quinn, T. Subramani, K. Lilova, G. Baldinozzi, I.M. Gussev, A.F. Fuentes, J.C. Neuefeind, M. Everett, D. Sprouster, A. Navrotsky, R.C. Ewing, M. Lang, Disorder in Ho 2 Ti 2-x Zr x O 7 : pyrochlore to defect fluorite solid solution series, RSC Adv. 10 (2020), 34632. [34] F.P. Marlton, Z. Zhang, Y. Zhang, T.E. Proffen, C.D. Ling, B.J. Kennedy, Lattice disorder and oxygen migration pathways in pyrochlore and defect-fluorite oxides, Chem. Mater. 33 (2021) 1407–1415. [35] V.M. Orera, R.I. Merino, J.I. Pe~ na, Ce 3þ ↔Ce 4þ conversion in ceria-doped zirconia single crystals induced by oxido-reduction treatments, Solid State Ionics 72 (1994) 224–231. [36] P.B. Oliete, A. Orera, M.L. Sanju an, Spectroscopic insight into the interplay between structural disorder and oxidation degree in melt-grown Ce 0.5 Zr 0.5 O 2-y compounds, J. Raman Spectrosc. 51 (2020) 514–527. [37] K. Kawamura, K. Watanabe, T. Hiramatsu, A. Kaimai, Y. Nigara, T. Kawada, J. Mizusaki, Electrical conductivities of CaO doped ZrO 2 -CeO 2 solid solution system, Solid State Ionics 144 (2001) 11–18. [38] C.H. Lee, G.M. Choi, Electrical conductivity of CeO 2 doped YSZ, Solid State Ionics 135 (2000) 653–661. [39] J.P. Eufinger, M. Daniels, K. Schmale, S. Berendts, G. Ulbrich, M. Lerch, H.- D. Wiemh€ ofer, J. Janek, The model case of an oxygen storage catalyst –nonstoichiometry, point defects and electrical conductivity of single crystalline CeO 2 - ZrO 2 -Y 2 O 3 solid solutions, Phys. Chem. Chem. Phys. 16 (2014) 25583–25600. [40] A. Nakamura, J.B. Wagner, Defect structure, ionic conductivity and diffusion in yttria stabilized zirconia and related oxide electrolytes with fluorite structure, J. Electrochem. Soc. 133 (1986) 1542–1548. [41] H. Inaba, H. Tagawa, Ceria-based solid electrolytes, Solid State Ionics 83 (1996) 1–16. A. Orera et al. Journal of Solid State Chemistry 315 (2022) 123525 9