# Fileset

[132_24026.pdf](https://mdr.nims.go.jp/filesets/244e760e-47e9-436e-9bcd-20fb8a309026/download)

## Creator

Koichiro Fukuda, Fumiya Nakajima, Daisuke Urushihara, Toru Asaka, [Tohru S. Suzuki](https://orcid.org/0000-0001-9458-6863), Abid Berghout, Assil Bouzid, Olivier Masson, Philippe Thomas

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Calcium ion conduction anisotropy of <i>b</i>-axis-aligned CaAl<sub>4</sub>O<sub>7</sub> polycrystal](https://mdr.nims.go.jp/datasets/611afa27-9979-420d-9d84-fd636d462b37)

## Fulltext

Calcium ion conduction anisotropy of b-axis-aligned CaAl4O7 polycrystalFULL PAPERCalcium ion conduction anisotropyof b-axis-aligned CaAl4O7 polycrystalKoichiro Fukuda1,³, Fumiya Nakajima1, Daisuke Urushihara1, Toru Asaka1, Tohru S. Suzuki2,Abid Berghout3, Assil Bouzid3, Olivier Masson3 and Philippe Thomas31Division of Advanced Ceramics, Nagoya Institute of Technology, Nagoya 466–8555, Japan2Optical Ceramics Group, Research Center for Electronic and Optical Materials, National Institute for Materials Science,Tsukuba 305–0047, Japan3Université de Limoges, IRCER, CNRS, UMR 7315, 12 Rue Atlantis, F-87000 Limoges, FranceTo investigate the anisotropy of Ca2+ conduction, the b-axis-aligned CaAl4O7 (space group C2/c) polycrystal wasprepared by colloidal processing under high magnetic field of 12T followed by sintering at 1773K for 2 h. Thetextured polycrystal was characterized by X-ray diffractometry and impedance spectroscopy with respect to thegrain alignment direction, which was parallel to the applied magnetic field. The texture fraction of {0k0},expressed as the Lotgering factor f0k0, was 0.63. The ©101ª directions of individual crystal grains were randomlydistributed around the grain-alignment direction of the polycrystal. The conductivities perpendicular (·¦) andparallel (·¬) to the grain alignment direction were compared with the conductivity (·random) of random grainoriented polycrystal between 773 and 1073K. The ·¦, ranging from 3.09 © 10¹7 to 1.80 © 10¹5 S cm¹1, showedthe highest value at each temperature, followed by ·random and ·¬ in that order. The results have confirmed forthe first time the anisotropy of Ca2+ conduction and strongly supported the preferential conduction in the ©101ªdirection predicted in the literature by the bond valence method.Key-words : Calcium ion conductor, Grossite, Grain alignment, Anisotropy, Impedance spectroscopy[Received March 18, 2024; Accepted April 16, 2024; Published online May 18, 2024]1. IntroductionSolid electrolytes with fast conduction of multivalentcations are expected to be used in high performance elec-trochemical devices such as high capacity rechargeablebatteries due to their ability to achieve high density chargetransfer.1)–5) The solid electrolytes are broadly divided intosolid polymer electrolytes and inorganic (ceramic) electro-lytes. Since polymer electrolytes have relatively low oper-ating temperatures, their application in rechargeable bat-teries operating at room temperature has been activelyinvestigated.6),7) For ceramic electrolytes, one of the condi-tions limiting their practical application is their relativelyhigh operating temperatures. They could therefore be usedas high-performance energy storage batteries to replacesodium-sulfur batteries in practical use,8) which currentlyoperate at 573K. However, it is generally very difficult formultivalent cations in inorganic crystals to conduct rapidlythrough the crystal structures due to strong electrostaticinteractions with neighboring anions. In view of this situa-tion, Fukuda et al. speculated that in crystals with rigidframework structures and strong structural anisotropy,multivalent cations may conduct relatively easily in spe-cific crystallographic one-dimensional directions.9) Usingthe bond valence (BV) method,10)–13) they explored candi-date materials based on this hypothesis and found thatcompounds of the grossite (CaAl4O7, space group C2/c)type may exhibit relatively high Ca2+ conductivity in the©101ª one-dimensional direction.9) The Ca2+ conductivityof the random grain oriented CaAl4O7 polycrystal wassuperior to that of the NASICON-type CaZr4(PO4)6 poly-crystal, in which Ca2+ conducts three-dimensionally in thecrystal structure.14) The transference number of Ca2+ con-duction in CaAl4O7 was close to 1 at 0.973, indicatingalmost no electronic or oxide-ion conductivity.9) Sincecalcium is abundant on earth (Clarke number 3.39%) andhas a standard electrode potential value (Ca2+/Ca¹2.87V)comparable to that of Li (Li+/Li ¹3.05V), fast Ca2+ con-ductors have attracted attention as promising electrolytesfor calcium ion-based batteries.15)The crystal grains of ceramics produced by conven-tional methods tend to have a random orientation. There-fore, even if each individual grain has anisotropic physicaland/or mechanical properties, these are cancelled out as awhole and the sintered body exhibits isotropic properties.On the other hand, if the individual grains can be alignedin a particular crystallographic orientation, a significantimprovement in properties can be achieved. Conversely, ifthe intrinsic properties of polycrystals become apparent³ Corresponding author: K. Fukuda; E-mail: fukuda.koichiro@nitech.ac.jpJournal of the Ceramic Society of Japan 132 [7] 409-414 2024DOI https://doi.org/10.2109/jcersj2.24026 JCS-Japan©2024 The Ceramic Society of Japan 409This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/),which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.https://doi.org/10.2109/jcersj2.24026https://creativecommons.org/licenses/by/4.0/when the constituent grains are aligned in one direction,this is evidence that they have anisotropic properties.Various methods such as templated grain growth,16),17) hotforging,18) and reactive diffusion19) have been used toorient the constituent crystal grains in ceramics. Recently,it has been reported that single crystal particles of feeble-magnetic materials with very low magnetic susceptibility,such as diamagnetic and paramagnetic materials, can rotatein a strong magnetic field generated by superconductingmagnets due to the magnetic torque caused by the mag-netic anisotropy of the crystal.20)–22) Taking advantage ofthis phenomenon, Suzuki et al. have successfully pro-duced textured feeble magnetic ceramics by colloidal proc-essing under a strong magnetic field followed by sinter-ing.21),22) In general, this method has the advantage that theorientation direction of the particles relative to the samplegeometry can be controlled by the direction of the staticmagnetic field.In this study, the Ca2+ conduction anisotropy of CaAl4O7has been experimentally demonstrated for the first timefor the b-axis-aligned polycrystal, where the direction ofgrain alignment was perpendicular to the ©101ª conductiondirection predicted by the BV method. If the polycrystalswith highly oriented grains in the ©101ª direction can beproduced, a further improvement in conductivity would beexpected.2. Experimental2.1 MaterialsAwell-mixed powder with a [Ca:Al] molar ratio of [1:4]was prepared from the reagent grade chemicals CaCO3(99.5%, Kishida Kasei, Osaka, Japan) and Al2O3 (99.0%,Kishida Kasei, Japan). It was heated at 1673K for 5 h andthen quenched in air. The heating process was repeatedtwice with intermediate grinding. The reaction productwas a lightly sintered polycrystalline material consisting ofCaAl4O7. The polycrystal was ground in a planetary ballmill (200 rpm, 6 h) to obtain the fine powder, which wasdispersed by mutual electrostatic repulsion in distilledwater with an appropriate amount of polymeric dispersant.A magnetic field of 12T was statically applied to the welldispersed suspension during slip casting in the field di-rection parallel to the casting direction. As it is difficultto rotate diamagnetic particles even in a strong magneticfield, the colloidal processing was used to facilitate therotation of the particles. Details of the high magnetic fieldapparatus and the preparation procedure of the suspensionare described in the literature.21) The crystal particles ofCaAl4O7 were indeed oriented, suggesting that CaAl4O7 isnot non-magnetic but, like Al2O3, diamagnetic.22) Theresulting green body was then heated at 1773K for 2 h inthe absence of the magnetic field to prepare the cylindricalsintered polycrystal (¤11.7mm and 3.7mm thick), withthe table plane perpendicular to the direction of the appliedmagnetic field [Fig. S1(a)].2.2 X-ray diffractometry and thermal etchingThe resulting sintered polycrystal was cut perpendicularto the table plane using a diamond saw to make two pieces[Fig. S1(b)]. The CuK¡1 beam was incident on the tableplane as well as the cross-sectional surface to obtain theX-ray diffraction (XRD) patterns in the Bragg-Brentanogeometry using a diffractometer (X’Pert PRO Alpha-1,PANalytical B.V., Almelo, The Netherlands). The inte-grated reflection intensities were extracted by the Le Bailmethod23) using a computer program RIETAN-FP.24) Thisprogram was used to generate simulated XRD patterns ofpolycrystals with random grain orientation and a highlyoriented b-axis, based on the structural parameters ofCaAl4O7 as determined by Goodwin and Lindop.25) TheLotgering factor ( f0k0),26) corresponding to the texturefraction of the {0k0} planes, was determined from the hklreflection intensities of the textured polycrystal and thosesimulated for the randomly grain oriented one.The cross-sectional surface of one of the cut sampleswas polished with diamond paste and thermally etchedby heating at 1473K for 0.5 h followed by cooling in air.The thermally etched microtexture was observed using areflected light microscope.2.3 Impedance spectroscopyThe remaining cut sample was divided into two piecesusing the diamond saw. One piece was mechanically pol-ished perpendicular to the original table plane using SiCabrasive paper to prepare the thin plate electrolyte with athickness (L) of 0.241 cm and a surface area (S) of 0.302cm2, giving a shape factor (L/S) of 7.98 © 10¹1 cm¹1. Theresulting electrolyte was designated 1. The other piece waspolished parallel to the original table plane to make thethin plate electrolyte (designated 2) with L/S = 5.29 ©10¹1 cm¹1 (L = 0.110 cm and S = 0.208 cm2). Plate elec-trodes were prepared by applying platinum paste to bothsides of electrolytes 1 and 2, and then heating at 1273K todecompose the paste and harden the platinum residue. Theplatinum electrodes fabricated with the present platinumpaste were the same as those used in a previous study,9)which did not react with the CaAl4O7 electrolyte at1073K. Complex impedance data were collected using animpedance analyzer (IM3570, HIOKI E. E. Co., Nagano,Japan) at temperatures from 773 to 1073K over the fre-quency range of 4 to 5MHz in air. The impedance spectrawere analyzed by a distribution of relaxation times (DRT)analysis method using Z-ASSIST software,27) followed bya non-linear least squares fitting method using ZViewsoftware.28) In the adopted equivalent circuit, the elementscorresponding to bulk and grain boundary (gb) are con-nected in series as (RbulkQbulk)(RgbQgb), where R is theresistance in parallel with the constant phase element Q. Ingeneral, the geometric capacitance (C) values and theirpossible interpretations are 10¹12–10¹11 F cm¹1 for Cbulk,and 10¹11–10¹8 F cm¹1 for Cgb.29)–31) The anisotropy ofCa2+ conduction in the polycrystal with respect to thedirection of grain alignment was explained by the differ-ence in conductivity between electrolyte 1 and electrolyte2, taking also into account the conductivity of randomgrain oriented polycrystal in the literature.9)Fukuda et al.: Calcium ion conduction anisotropy of b-axis-aligned CaAl4O7 polycrystalJCS-Japan4102.4 Bond valence energy landscape methodThe potential conductivity of Ca2+ in the crystal struc-ture of CaAl4O7 was investigated by the bond valenceenergy landscape (BVEL) method,32),33) which links theBV method to the absolute energy scale and allows theactivation energy of conduction to be estimated. The con-duction paths of ions are generally consistent with thespatial distributions of relatively low BVenergies. The BVenergy of Ca2+ was calculated with a resolution of 0.01nm/voxel using a computer program PyAbstantia.34) Theenergy isosurfaces were superimposed on the structuralmodel, and both were drawn using VESTA software.35)3. Results and discussion3.1 Orientation degree of polycrystalThe thermal etching process effectively producedgrooves along the grain boundaries, allowing for easyobservation of the external shape of each crystal grain withthe optical microscope (Fig. 1). The crystal grains of allsizes were randomly distributed, and there was no posi-tional bias in the size of the particles, with a maximumgrain diameter of approximately 16.5¯m. Therefore, therewas no definite correlation between the grain assemblystructure and the direction of the magnetic field applied.The XRD pattern taken from the table plane of thecylindrical sintered polycrystal showed a significant in-crease in 0k0 (k = 2, 4, and 6) reflections [Fig. 2(a)]. Thef0k0 value of 0.63 indicates that, during the slip casting ofthe dispersed particle suspension, the b-axis of each crystalgrain tended to align along the statically applied magneticfield parallel to the direction of particle settling. The result-ing sintered polycrystal was found to be the polycrystallinematerial with the majority of the constituent crystal grainsoriented almost in the b-axis direction. On the other hand,the XRD pattern obtained from the cross-sectional surfaceof the sintered polycrystal showed that the 0k0 reflectionhad completely disappeared [Fig. 2(b)]. This is consistentwith the simulated XRD pattern of the polycrystal with ahighly oriented b-axis (Fig. S2). It is worth noting that thea- and c-axes are randomly oriented around the oriented b-axis for these XRD patterns.These results confirmed that the c- and a-directions ofindividual crystal grains were randomly distributed aroundthe grain alignment direction, which is parallel to thedirection of applied magnetic field. It should be noted thatthe ©101ª direction lies in the plane formed by the c- and a-axes and intersects the b axis perpendicularly. As a result,we confirmed that the b-axes of the individual crystalgrains in the sintered polycrystal were highly orientedperpendicular to the table plane, while the ©101ª directionsof the grains were randomly distributed around the grainalignment direction.3.2 Anisotropy of Ca2© conduction inCaAl4O7The Nyquist plot for electrolyte 1 at 773K showed anincomplete semicircle of bulk contribution, while all otherplots for both electrolytes 1 and 2 showed two semicircleswith bulk at high frequencies and gb at low frequencies(Figs. S3 and S4). The fitting results and the deconvolu-tion of the two contributions were displayed as blue solidcurves and red dashed semicircles, respectively. Thesetwo semicircles were well identified as the bulk and gbcontributions, since the values of Cbulk/F cm¹1 and Cgb/F cm¹1 were in good agreement with those determined forceramic materials.29)–31) They were 4.31 © 10¹12 (873K)–5.51 © 10¹12 (1073K) for Cbulk/F cm¹1 and 4.34 © 10¹11(1073K)–7.78 © 10¹11 (873K) for Cgb/F cm¹1 with elec-trolyte 1, and 4.10 © 10¹12 (873K)–2.75 © 10¹11 (773K)for Cbulk/F cm¹1 and 7.56 © 10¹11 (1073K)–3.77 © 10¹10(873K) for Cgb/F cm¹1 with electrolyte 2. The bulk con-ductivity (·) was determined by 1/Rbulk © L/S.As the plate electrodes of electrolyte 1 were parallel tothe direction of grain alignment, the conductivity deter-mined (·¦) was in the direction perpendicular to it. Thevalue of ·¦ increased steadily from 3.09 © 10¹7 to 1.80 ©10¹5 S cm¹1 with increasing temperature from 773 to1073K (Fig. 3). The activation energy of conduction (Ea)was 1.060(15) eV. On the other hand, since the plate elec-trodes of electrolyte 2 were perpendicular to the directionof grain alignment, the obtained conductivity (·¬) was inthe direction parallel to it. The ·¬ value also showed thesteady increase with increasing temperature from 1.49 ©10¹8 to 1.99 © 10¹6 S cm¹1 over the same temperaturerange, and the Ea value was 1.22(2) eV (Fig. 3). The ionicconductivity of the random grain oriented polycrystal(·random) showed the intermediate value, together with theintermediate activation energy of 1.14(2) eV.9) When com-pared at the same temperatures, the ·¦ value was 9.1(1073K) to 20.8 (773K) times larger than the ·¬ value.However, when comparing the ·¦ and ·random values, theformer was only 1.6 (1073K) to 2.3 (773K) times largerthan the latter, and the difference was not as pronounced. Asignificant difference was observed between the ·¬ and·random values, the latter being 5.6 (1073K) to 9.2 (773K)times larger than the former. The direction of Ca2+ con-duction was predicted by the BV method to be parallel to©101ª,9) which intersects perpendicular to the b-axis direc-tion in the crystal structure. Accordingly, the conductivityof the b-axis-aligned polycrystal is expected to be relativelyhigh for ·¦ and low for ·¬, and this was indeed the case.Fig. 1. Thermally etched microtexture produced on the surfaceparallel to the applied magnetic field as indicated by the whitearrow. Reflected light micrograph.Journal of the Ceramic Society of Japan 132 [7] 409-414 2024 JCS-Japan411The reason for the larger value of ·¦ than that of ·random isdue to the larger fraction of ©101ª aligned crystal grains inthe direction in which the ionic conductivity was measured.Since the semicircles of grain boundary were clearlyrecognizable in the Nyquist plots at temperatures between873 and 1073K for electrolyte 1 (Fig. S3), the grainboundary conductivity was plotted together with the totalconductivity in Fig. S5 with an activation energy of 0.70(3)eV for the former conduction. As the grain boundary vol-ume fraction and grain size distribution of polycrystals areexpected to influence the grain boundary conductivity,clarification of this relationship is an important subjectwhen considering the practical application of the elec-trolyte, as it may contribute to improving the overallconductivity.If the direction of Ca2+ conduction were restricted tothe ©101ª one-dimensional direction, the perfectly b-axis-aligned CaAl4O7 polycrystal with f0k0 = 1 should show noconduction at all in the grain alignment direction. In fact,the conduction was observed in the grain-alignment direc-tion of the textured polycrystal obtained and the ·¦ valuewas determined by the complex impedance method. Thereare two possible reasons for this. First, the conductionpaths in the CaAl4O7 crystal exist not only in the ©101ªdirection but also in other directions. Second, although thedegree of orientation ( f0k0) of the grain-aligned polycrystalwas as high as 0.63, it was not completely b-axis oriented.Thus, within the grain-aligned polycrystal, there would belocations where ©101ª conduction paths are connectedbetween adjacent grains, forming conduction paths in thedirection of grain alignment. We used the BVEL method toverify the first possibility and found new conduction pathsalong the ©001ª direction (Fig. S6). The Ea values werederived from the spatial distribution of the BV energy tobe 1.59 eV in the ©101ª direction and 4.36 eV in the ©001ªdirection. It is well known that theoretical Ea values de-rived from the BVEL method tend to be higher thanexperimental values, partly because they do not take intoaccount the effects of lattice relaxation associated with ionconduction.33) Assuming that the Ea value of the formerFig. 2. Fitting results of the observed X-ray diffraction patterns (red symbol: +) of the sintered polycrystal. Theirradiated surface areas are perpendicular to the grain alignment direction in (a) and parallel to it in (b). In eachdiagram, the calculated pattern and the positions of the possible Bragg reflections are indicated by the upper solidline and the lower vertical bars, respectively, and the difference curve is displayed in the lower part of thediagram. Inset in (a): Simulated X-ray diffraction pattern of the random grain oriented CaAl4O7 polycrystal.Fukuda et al.: Calcium ion conduction anisotropy of b-axis-aligned CaAl4O7 polycrystalJCS-Japan412was the same as that of ·¦ (= 1.06 eV), the latter correctedby the scaling factor of 0.667 (= 1.06/1.59) is 2.91 eV.Thus, the Ea value of Ca2+ conduction in the ©001ª direc-tion would be approximately 2.7 times larger than that inthe ©101ª direction. Whether or not conduction in the ©001ªdirection actually occurs would be verified by, for exam-ple, molecular dynamics simulation methods and/or ex-perimental measurements of anisotropic conduction for thesingle crystals. Since the conduction paths in the ©101ª and©001ª directions are both perpendicular to the b-axis, theydo not contribute to the conduction in the b-axis direction.It is therefore possible that the second reason is the cause.The possibility of Ca2+ conduction in CaAl4O7 was firstdemonstrated by the BV method, and subsequently con-firmed experimentally for the random grain oriented poly-crystal.9) However, it was unclear whether the predictedconduction in the ©101ª direction actually occurs in thecrystal structure. In this study, the anisotropy of Ca2+ con-duction has been experimentally confirmed for the firsttime, strongly supporting that the conduction direction ismainly along ©101ª. If polycrystals with a higher degreeof orientation, preferably in the ©101ª direction, could beobtained, the Ca2+ conductivity would be expected to behigher than that of the present polycrystal.The elucidation of the Ca2+ conduction mechanismwould provide another guideline for conductivity enhance-ment. Whether there are defects in the Ca sites in the crys-tal structure of CaAl4O7 and, if so, how their concentra-tions affect the Ca2+ conductivity are important issuesclosely related to the Ca2+ conduction mechanism. Thestoichiometric composition of the sample must be accu-rately determined and its relationship to Ca2+ conductivityclarified.4. ConclusionThe polycrystalline CaAl4O7 (space group C2/c) wasprepared by colloidal processing under the high magneticfield of 12T followed by heating at 1773K for 2 h. Duringthe former process of slip casting the dispersed particlesuspension, the b-axis of each crystal grain tended to alignalong the statically applied magnetic field parallel to thedirection of particle settling. Thus, after sintering in theabsence of the magnetic field, the resulting polycrystalexhibited a high b-axis orientation with the Lotgeringfactor f0k0 of 0.63, and the ©101ª directions of individualcrystal grains were randomly distributed around the grainalignment direction. The conductivities perpendicular (·¦)and parallel (·¬) to this direction were compared between773 and 1073K, together with the conductivity (·random) ofrandom grain oriented polycrystal. The value of ·¦ in-creased steadily from 3.09 © 10¹7 to 1.80 © 10¹5 S cm¹1with increasing temperature. When compared at the sametemperature, ·¦ had the highest value, followed by ·randomand then ·¬ in that order. In this study, the anisotropy ofCa2+ conduction in CaAl4O7 was confirmed for the firsttime, and the results obtained were consistent with pref-erential conduction in the ©101ª direction as predicted bythe BV method. If the polycrystals with highly orientedgrains in the ©101ª direction can be produced, a furtherimprovement in conductivity would be expected.Appendix A. Supporting information Supplementarydata associated with this article can be found in the on lineversion.Acknowledgment This research was supported by aGrant-in-Aid for Scientific Research (No. 19K22051) fromthe Japan Society for the Promotion of Science.References1) L. F. O’Donnell and S. G. Greenbaum, Batteries, 7, 3(2021).2) Y. Liang, H. Dong, D. Aurbach and Y. Yao, Nat. Energy,5, 646–656 (2020).3) A. Ponrouch, J. Bitenc, R. Dominko, N. Lindahl, P.Johansson and M. R. Palacín, Energy Storage Mater.,20, 253–262 (2019).4) N. Imanaka and S. Tamura, B. Chem. Soc. Jpn., 84,353–362 (2011).5) N. Imanaka, J. Ceram. Soc. Jpn., 113, 387–393 (2005).6) F. S. Genier, C. V. Burdin, S. Biria and I. D. Hosein,J. Power Sources, 414, 302–307 (2019).7) Q. Wei, L. Zhang, X. Sun and T. L. Liu, Chem. Sci., 13,5797–5812 (2022).8) D. Kumar, S. K. Rajouria, S. B. Kuhar and D. K.Kanchan, Solid State Ionics, 312, 8–16 (2017).9) K. Fukuda, R. Harada, H. Banno, D. Urushihara and T.Asaka, ACS Appl. Energ. Mater., 5, 3227–3234 (2022).10) S. Adams, K. Hariharan and J. Maier, Solid State Ionics,86–88, 503–509 (1996).11) S. Adams and J. Maier, Solid State Ionics, 105, 67–74(1998).12) S. Adams, Solid State Ionics, 136–137, 1351–1361Fig. 3. Anisotropy of Ca2+ conduction in b-axis-alignedCaAl4O7 polycrystals. Conductivity ·¦ perpendicular to thedirection of grain alignment and conductivity ·¬ parallel to it.Conductivity ·random for random grain oriented polycrystal.9)Journal of the Ceramic Society of Japan 132 [7] 409-414 2024 JCS-Japan413(2000).13) M. Avdeev, M. Sale, S. Adams and R. Prasada, SolidState Ionics, 225, 43–46 (2012).14) K. Nomura, S. Ikeda, K. Ito and H. Einaga, B. Chem.Soc. Jpn., 65, 3221–3227 (1992).15) M. E. Arroyo-de Dompablo, A. Ponrouch, P. Johanssonand M. R. Palacín, Chem. Rev., 120, 6331–6357 (2020).16) H. Igarashi, K. Matsunaga, T. Taniai and K. Okazaki,Am. Ceram. Soc. Bull., 57, 815–817 (1978).17) M. Holmes, R. E. Newnham and L. E. Cross, Am.Ceram. Soc. Bull., 58, 872 (1979).18) T. Takenaka and K. Sakata, Jpn. J. Appl. Phys., 19, 31–39 (1980).19) K. Fukuda, T. Asaka, R. Hamaguchi, T. Suzuki, H. Oka,A. Berghout, E. Béchade, O. Masson, I. Julien, E.Champion and P. Thomas, Chem. Mater., 23, 5474–5483 (2011).20) T. Sugiyama, M. Tahashi, K. Sassa and S. Asai, ISIJInt., 43, 855–861 (2003).21) T. S. Suzuki, J. Ceram. Soc. Jpn., 128, 1005–1012(2020).22) T. S. Suzuki, T. Uchikoshi and Y. Sakka, Sci. Technol.Adv. Mat., 7, 356–364 (2006).23) A. Le Bail, H. Duroy and J. L. Fourquet, Mater. Res.Bull., 23, 447–452 (1988).24) F. Izumi and K. Momma, Solid State Phenom., 130, 15–20 (2007).25) D. W. Goodwin and A. J. Lindop, Acta Crystallogr. B,26, 1230–1235 (1970).26) F. K. Lotgering, J. Inorg. Nucl. Chem., 9, 113–123(1959).27) H. Sumi, H. Shimada, Y. Yamaguchi, T. Yamaguchi andY. Fujishiro, Electrochim. Acta, 339, 135913 (2020).28) D. Johnson, ZView Program (Version 2.9b), ScribnerAssociates Inc., Southern Pines, NC (1990).29) J. T. S. Irvine, D. C. Sinclair and A. R. West, Adv.Mater., 2, 132–138 (1990).30) X. Vendrell and A. R. West, J. Electrochem. Soc., 165,F966–F975 (2018).31) I. N. David, T. Thompson, J. Wolfenstine, J. L. Allenand J. Sakamoto, J. Am. Ceram. Soc., 98, 1209–1214(2015).32) S. Adams and R. P. Rao, Phys. Status Solidi A, 208,1746–1753 (2011).33) S. Adams and R. P. Rao, Understanding Ionic Con-duction and Energy Storage Materials with Bond-Valence-Based Methods, in “Bond Valences”, Ed. byI. D. Brown and K. R. Poeppelmeier, Springer, BerlinHeidelberg (2014) pp. 129–159.34) P. Barpanda, G. Oyama, S. Nishimura, S.-C. Chung andA. Yamada, Nat. Commun., 5, 4358 (2014).35) K. Momma and F. Izumi, J. Appl. Crystallogr., 44,1272–1276 (2011).Fukuda et al.: Calcium ion conduction anisotropy of b-axis-aligned CaAl4O7 polycrystalJCS-Japan414