HAL Id: jpa-00226036
https://hal.archives-ouvertes.fr/jpa-00226036
Submitted on 1 Jan 1986
HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
THE DEFECT STRUCTURE OF CALCIA STABILISED ZIRCONIA
L. Moroney, D. Sayers
To cite this version:
L. Moroney, D. Sayers. THE DEFECT STRUCTURE OF CALCIA STABILISED ZIRCONIA. Jour- nal de Physique Colloques, 1986, 47 (C8), pp.C8-725-C8-728. �10.1051/jphyscol:19868136�. �jpa- 00226036�
JOURNAL DE PHYSIQUE
Colloque C8, suppl6ment au n o 12, T o m e 47, d e c e m b r e 1986
THE DEFECT STRUCTURE OF CALCIA STABILISED ZIRCONIA
L.M. MORONEY and D.E. SAYERS"
Brookhaven National Laboratory, Upton, NY 11973, U.S.A.
" ~ e p a r t m e n t of Physics, North Carolina State University, Raleigh NC 27695-8202, U.S.A.
Abstract - EXAFS data comparing the local structural environments of the calcium and zir- conium ions in calcia-stabilised zirconia are presented. The Zr data resembles that observed previously for yttria-stabilised zirconia with the zirconium ions tending to be displaced from their centrosymmetric sites and a Zr-0 distance of 2.12 A. These similarities suggest that anion vacan- cies in this material are also to be found adjacent to Zr4+. The calcium ions are not displaced from their lattice sites but their immediate anion neighbours are highly disordered resulting in a t least two Ca-0 distances in the radial distribution of the nearest neighbour shell.
1. Introduction
Calcia stabilised zirconia is a widely used material with many demonstrated applications in high temperature technology [1],[2],[3]. It adopts the cubic fluorite structure with Ca2+ substituting for Zr4+
and electroneutrality being maintained by the formation of anion vacancies. The specific location of the anion vacancies with respect to the two different cations has yet to be resolved. The determination of short range order parameters from diffuse X-ray scattering data [4] suggests that the vacancies are first neighbour to Ca and this can be explained by the favourable Coulomb interaction between the Caa+ cation and the oppositely charged anion vacancy. However, the electron diffraction data of Allpress and Rossell [5] display regions of scattering indistinguishable from that of CaZr409 in which Ca is always 8 co-ordinate and the Zr4+ ions adopt 6,7 and 8 fold co-ordination. Recent EXAFS results for the analogous phase, yttria stabilised zirconia [6] show that the Zr4+ local structural environment is significantly more disordered than that of YS+ and also that there is a displacement of the Zr4+ ions from their centrosymmetric sites. These results were interpreted as indicating that anion vacancies, generated by the presence of YS+ in the lattice, are sited first neighbour to Zr4+ with accompanying anion and Zr4+ relaxations. It was argued that a 7-fold anion co-ordination, which, in fact, is the co-ordination adopted by the low temperature monoclinic phase of pure ZrOz, permits a closer interaction between the small Zr4+ ions and the anions.
A similar EXAFS experiment has been carried out to compare the local structural environment of CaZ+ and Zr4+ in calcia stabilised zirconia.
,2. Experimental Procedure
The stoichiometry of the samples used in this study is C~,19Zro.e101.81 and CaZrOs. X-ray diffraction was used to confirm that the former sample has the fluorite structure and the latter, the monoclinic Pcmn space group described by Koopmans et a1 [7]. The EXAFS measurements were made a t the National Synchrotron Light Soilrce on beam line X l l A operating a t 2.5 GeV with electron currents of 40-100 mA.
The powdered samples were dusted onto scotch tape and routine transmission X-ray absorption spectra obtained a t the Zr K edge a t room temperature and 80K. The signal to noise ratio was low for data obtained in transmission mode a t the Ca K edge (4038 eV) necessitating these EXAFS measurements being made in the fluorescence mode.
Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:19868136
C8-726 JOURNAL DE PHYSIQUE
3. Results and Discussion
Figure 1 shows the ks-weighted EXAFS and its Fourier transform for the fluorite sample a t the Zr edge. Only the nearest oxygen and next nearest cation neighbours contribute to the fine structure.
This contrasts with the parent material, monoclinic Zr02 which has peaks in the Fourier transform out to N 6-7 A even though there are a multitude of component distances comprising successive shells of neighbours in this structure. Clearly, the level of disorder in calcia-stabilised zirconia is sufficiently large that the coherence of scattering from neighbours more distant than the second is destroyed. Since the differences between the data obtained a t 80 K and room temperature are small, we can attribute this disorder primarily to large static displacements of the ions from the perfect lattice positions. Figure 2a compares the Fourier filtered first shell of the Zr EXAFS in calcia-stabilised zirconia with that of the Zr EXAFS obtained for CaZr03. This illustrates the difficulties of determining co-ordination numbers accurately in materials which have large static disorder for, although the amplitude of the CaZrOs data exceeds that of calcia-stabilised zirconia a t low k, the co-ordination number of the former is six, whereas the co-ordination number of calcia-stabilised zirconia, even if all vacancies were first neighbour to Zr, could only be as low as 7. Using the Zr-0 data from CaZrOs as a model compound to determine the Zr-0 distance in calcia-stabilised zirconia, we obtain a value of 2.12 A and the Debye-Waller term is 0.008 A2
greater than that of CaZrOs.
The Ca data for the CaO/Zr02 sample is plotted in Figure 3. It can be seen that the signal to noise is considerably lower than we obtained for the Zr data. However, the data clearly shows, as was seen for yttria stabilised zirconia 16) that the second shell of cation neighbours is significantly larger for the Ca data than the Zr data. This amplitude difference, may a t first consideration, be attributed to cation ordering; i.e. if all 12 of the Ca second nearest neighbours were Zr and all 12 of the Zr nearest neighbours, the lighter Ca ions, then a smaller second shell for Zr compared with Ca would be expected due to the difference in the back scattering factors for the two cations. However, cation ordering of this kind would undoubtedly be seen in the diffraction data which instead indicates that the 12 second nearest cation neighbours are a statistical distribution of Ca2+ and Zr4+. Moreover, cation ordering cannot account for the same result obtained for yttria stabilised zirconia because in that case the two cations, YS+ and Zr4+, are isoelectronic and thus must have nearly identical backscattering factors. Therefore, we conclude that, as in yttria-stabilised zirconia, the Zr4+ ions in the calcia stabilised compound tend to be displaced from their centrosymmetric sites resulting in a large spread of Zr-cation distances and thus a significantly diminished second shell amplitude. The Ca2+ ions, on the other hand, stay close to their centrosymmetric sites leading to more shells of neighbours contributing t o the EXAFS as is seen by the additional structure in the Fourier transform a t
-
4.2 A compared with the Zr data."
2.5 5 7.5 10 12.5 0 2 4 6 8
k / A-' Radial Distance / A
Fig. l a - ksx(k) data for CaO/ZrO, a t Zr edge Fig. l b - Fourier transform of data in la.
at room temperature.
The nearest neighbour Ca-0 EXAFS shows the first deviation from the pattern seen for yttria- stabilised zirconia. The latter EXAFS indicated that the Y-0 shell was more ordered than the Zr-0 shell with a larger Y-0 distance and smaller Debye-Waller factor. This, of course, is the expected pattern if the Y cations stay close to their centrosymmetric sites and are fully co-ordinated to 8 oxygen atoms.
Contrastingly, the amplitude of the Ca-0 envelope of the Fourier filtered shell is less than that of the ZrO data. While it is tempting to conclude that this is evidence for a lower co-ordination number for the Caa+
ions, comparison of this data with that of CaZrOs (Figure 2b) shows that although the co-ordination of the Ca in the latter is 8, its amplitude is less than that of Ca in calcia-stabilised zirconia. In CaZrOs, the Ca-0 EXAFS amplitude is reduced by the disorder within the shell which comprises 5 distances varying from 2.34 A to 2.84 A. The data for calcia-stabilised zirconia also indicates more than one component Ca-0 distance within the first shell as evidenced by the beating occurring in the data causing a minimum a t N 8 A-l. The similarity of the frequencies in Figure 2b indicates that the mean Ca-0 distance in the fluorite material lies closer to the average Ca-0 distance of 2.57 A in CaZrOs than to the Zr-0 distance of 2.12 A. From the diffraction data for calcia-stabilised zirconia, the average cation-oxygen distance is 2.23
A. Thus, at the local level, the cation-anion distances follow the comparatively large difference in cation radii of 1.26 A for Ca2+ and 0.98 A for Zr4+ [8]. It is interesting to speculate that the larger Ca2+ ions can only be accommodated in the fluorite lattice by local distortions of the anion cube similar to those seen for CaZrOs.
k / A-' k / A-'
Fig. 2a - Comparison of the ka weighted Fourier Fig. 2b - Comparison as in 2a for the Ca-0 neigh- filtered first shell of Zr-0 neighbours in CaO/ZrOa bours in the two materials.
(bold line) and CaZrOs (dotted line).
k / A-' Radial Distance / A
Fig. 3a - ksx(k) data for CaO/Zr02 a t Ca edge Fig. 3b - Fourier transform of data in 3a.
at room temperature.
JOURNAL DE PHYSIQUE
4. Conclusion
Because there clearly is more than one Ca-0 distance in the first shell of neighbours for Ca, quantifying the Ca co-ordination number directly and unequivocally is not possible from the EXAFS data. However, other factors do suggest that, like yttria-stabilised zirconia and CaZr03, it is the Zr4+ ions which adopt the lower co-ordination number. Notably, the amplitude of the second shell is significantly larger in the Ca EXAFS compared with the Zr EXAFS. This indicates that the Zr cations tend to be displaced from their centrosymrnetric sites
-
such displacements are a likely response to neighbouring anion vacancies.Acknowledgements
This work has been supported a t Brookhaven by the U. S. Department of Energy under Contract No.
DE-AS05-80-ER10742.
References
[I] T.H. Etsell and S.N. Flengas, Chem. Rev. 70 (1970) 339-76.
[2] A.H. Heuer and L.W. Hobbs (Eds.), Advances in Ceramics Volume 9: Science and Technology of Zirconia (American Ceramic Society, Columbus, Ohio, 1981).
[3] N. Claussen, M. Riihle and A.H. Heuer (Eds.), Advancea in Ceramics Volume 12: Science and Technology of Zirconia 11 (American Ceramic Society, Columbus, Ohio, 1984).
[4] M. Morinaga, J.B. Cohen and J. Faber Jr., Acta. Cryst. A. A36 (1980) 520-30.
[5] J.G. Allpress and J.H. Rossell, J. Solid State Chem. 15 (1975) 68-78.
[6] C.R.A. Catlow, A.V. Chadwick, G.N. Greaves and L.M. Moroney, J. Am. Ceram. Soc. 69 (1986) 272-77.
[7] H.J.A. Koopmans, G.M.H. van de Velde and P.J. Gellings, Acta Cryst C39 (1983) 1323-25.
[8] R.D. Shannon, Acta Cryst. A32 (1976) 751-67.
