J. MATER. CHEM., 1991, 1(2), 163-167 Compound and Solid-solution Formation, Phase Equilibria and Electrical Properties in the Ceramic System Zr0,-La,O,-Ta,O,t Chaogui ZhengS and Anthony R. West University of Aberdeen, Department of Chemistry, Meston Walk, Aberdeen AB9 2UE, UK One new compound, ZrLaTa,O,, has been synthesized in the system Zr0,-La,O,-Ta,O,, with an orthorhombic unit cell, a= 10.890(3) A, b= 12.450(3)A and c=6.282(2) A. Extensive ternary solid solutions are formed by five of the binary phases: La,Zr,O,, La,TaO,, LaTaO,, LaTa,O, and LaTa,O,,. Two main solid-solution mechanisms are in evidence: La +Ta e2Zr and La +3Zr 3Ta. The solid-solution phase previously described as being based on Ta,Zr,O,, has, instead, the ideal stoichiometry, TaZr,.,,O,. A significant correction is made to the unit cell dimensions of La,TaO,. The subsolidus phase diagram of the Zr0,-La,O,-Ta,O, system has been determined at 1500°C.The electrical conductivity of a selection of the phases has been determined. Most are very poor semiconductors, but one, LaTa,O,, has a high conductivity, lop4S2-' cm-' at 400 "C, which appears to be due to oxide ions. ZrLaTa,O,, by contrast is an excellent insulator with a resistivity 31O6R cm at 800 "C. Keywords: Zr02-La20,-Ta20, system; Ceramic oxide; Oxide ion conduction We are interested in studying multicomponent ceramic oxide systems with the particular objective of synthesizing new ceramic phases or solid solutions and in measuring their electrical properties. Results on the system Zr02-La203- Nb205 have been reported.' New ternary solid solutions based on LazZr207, La,NbO, and LaNbO, were found and their electrical conductivities determined; all appear to be modest, p-type semiconductors.Here, the corresponding Ta20s system is reported. Major differences between the Taz05- and Nb20,-containing systems are found, including the existence of a new phase, LaZrTa,Oll. The system La20,-ZrO, was reviewed in ref. (1) and is not described again here. Briefly, it contains the cubic pyroch- lore phase La,Zr,O,, which forms a limited range of solid solutions over the range 30-35 mol% La203 at 1500 "C. Solid solution of the end-members is very limited (i.e. <1% La203 in ZrO, and <2% ZrO, in La203) at 1500 "C, the temperature of present interest.A phase diagram for the system La203-Ta205 has been reported,* showing the existence of four congruently melting phases La,TaO, (2020& 20 "C), LaTaO, (1930& 20 "C), LaTa,09 (1 850 20 "C) and La,Ta, ,033 (1890f20 "C). LaTa30, has a distorted perovskite structure with 2/3 of the large A cation sites vacant.,., La,TaO, has a weberite, distorted fluorite LaTaSO1, may have a distorted ReO, structure;8 and the structure of LaTaO, is not kno~n.~.~.'~ Partial phase diagrams for the binary system Zr02-Ta205 have been reported.' At Ta,O,-rich compositions a range of H-Ta20, solid solution forms at high temperatures; solid- solution limits were not determined accurately and are esti- mated as ca.20 mol% Zr02 l1 and ca. 11 YOZrO,' at 1500 "C. It appears that, at lower temperatures especially, these solid solutions may be better represented as a homologous series of line phases.12 The phase Ta,ZrO17 was reported,', but appears to have an upper limit of stability at 1500 "C.The phase Ta2Zr6OI7 is given in ref. (12) and this appears as a solid-solution phase in ref. (1 1) covering the range 67-86 mol% Zr02. Solidus temperatures in the system t Supplementary data available (SUP 56823, 10 pages); details from Editorial Office. $ Permanent address: Department of Chemistry, Peking University, Beijing 100871, People's Republic of China. Zr02-Ta205 appear to be at least 1750 "C; no solid solution of Ta205 in ZrO, has been indicated. Experimental The methods used to react mixtures (in Pt crucibles at 1500 "C) and to analyse the products (by X-ray powder diffraction, with internal standard KCI, as necessary) are essentially the same as those used in the study of the Zr02-La203-Nb205 system.' In this case, however, the entire study was carried out at 1500 "C because there was no evidence of partial melting at this temperature in any of the compositions.Results and Discussion Results of heating experiments on 119 compositions in the binary edges Zr0,-Ta205, La2O3-Ta2O5 and in the ternary system ZrO ,-La2 0,-Ta2 0 are available as Supplementary Data. Only those results in which the samples were deemed to have reached equilbrium are included. System La203-Ta20, Results on this binary edge (Fig.1) are in general consistent with those of most previous reports but with the difference that all four of the phases on this join have variable stoichi- ometry. The phases, with their solid-solution limits at 1500 "C in parentheses, are La,TaO, (24 & 1-28 & 1YOTa20s), LaTaO, (47& 1-50.5 & 0.5Y0), LaTa30g (74.5& 0.5-76.5 +0.5%), LaTaSO1, (81 & 1-84.5 k0.5Y0).The formation of such solid solutions is perhaps not surprising since at least three of these phases have distorted, cation-deficient structures, based on fluorite (La,TaO,), perovskite (LaTa,O,) and ReO, (LaTaSO1,). It seems likely that, at higher temperatures, the individual solid-solution ranges may be more extensive, especially for the latter two, whose structures (perovskite and ReO,) are closely related.There are two possible simple mechanisms that could be responsible for these solid solutions: 5La e3Ta (1) La eTa +0 (2) Very accurate density measurements would be required to show which mechanism was applicable to each of these solid J. MATER. CHEM., 1991 VOL. 1 Fig. 1 Subsolidus phase diagram Zr0,-La,O,-Ta,O, (mole%) at 1500 "C. 0,single phase; 0,two phases; 0,three phases solutions, since they are all of only limited extent. Such measurements have not been attempted. Indexed X-ray powder patterns for LaTa309 and La3Ta07 are available as Supplementary Data; such data are not available in the literature. The refined unit cell parameters for La,TaO,, a= 7.628(2) A, b= 7.749(2) 8, and c= 11.163(5)A, are considerably different from those reported in ref.(5), a= 7.84 A, b= 10.56 A, c =7.70 A. Using those cell parameters in ref.(5), it is possible to index the powder pattern but agreement between observed and calculated d spacings is often poor and cannot be improved by refinement. We believe that the parameters in ref.(5) are incorrect and arose owing to an incorrect indexing of the powder pattern. System Zr02-Ta205 Results on this join (Fig. 1) confirm the existence of the Ta,O, solid solutions containing <13& 1YOZrO,, similar to that suggested in ref. (12). A limited amount of solid solution of Ta205 in ZrO, appears to form, containing up to 2f0.5% Ta205. The Ta,Zr,O, ,phase was found over the solid-solution range 83 ?1-89 IfllYOZr02 which is much less extensive than that given in ref.( 11) and also extends to higher ZrO, contents.We believe that the ideal stoichiometry of this phase is TaZr,.,,08 (i.e. with 84.62 moly0 Zr.0,) instead of Ta,Zr6017 (with 85.72% ZrO,). The reported X-ray data for this phase are unindexed.', We collected fresh data and indexed them by trial and error Visser methods, to give a primitive ortho- rhombic unit cell with high figure of merit, a =5.1 17( 1) A, b = 5.279(1) A, c=4.976(1) A. Again these are available as Sup- plementary Data. Density measurements on two compositions gave values of 7.20 and 7.46 g ~m-~. Consideration of these density values and the unit cell volume indicated that the unit cell contents are much less than one formula unit of Ta2Zr6OI7.Instead, it seems highly likely that the unit cell contains a single unit of formula TaZ2,,,08. The variation in composition, over the range 83-89% ZrO,, could occur by one of two possible simple mechanisms: 4Ta e5Zr (3) 0+2Ta e2Zr (4) The expected variation in density with composition is I E 7.00 B I \ I I I I I I 0 0.20 0.40 Y Fig. 2 Density data for TaZr,,,,O, solid solutions, showing the most likely substitution mechanism to be 5Zr e4Ta shown in Fig. 2 for each of these mechanisms; the experimental data are added for comparison. Given that experimental densities are often a few percent less than theoretical values, the data indicate mechanism (3) as being the most likely.The solid-solution formula may therefore be written as + 5x08:Ta, -4xZr2,75 -O.O2092<x <0.06857. The crystal structure of TaZr2,7508 is not known but it could be a defective version of one of the ABO, structure types. Ternary System Zr02-La20,-Ta205 The results for the 1500°C isothermal phase diagram are shown in Fig. 1. Most of the binary phases form extensive ternary solid-solution series and a new compound, ZrLaTa,O ,,,was found. ZrLaTa,O, , has been indexed using Visser methods (Table 1). The data fit an orthorhombic unit cell, a= 10.890(3)A, b= 12.450(3)A, c= 6.282(2)A. Systematic absences indicate a face-centred cell belonging to one of three possible space groups, Frnrnrn, Fmm2 or F222.The density, determined by dispacement of toluene in a specific gravity bottle, was 7.64 g cm- which compares reasonably well with a calculated value of 7.40gcm-,, assuming unit cell contents of four formula units. From the compositional location of the ternary solid solu- tions in Fig. 1, it appears that several solid-solution mechan- isms are important. On the join, La,TaO,-La,Zr,O,, both end-members form partial solid solutions with each other and the mechanism is clearly La +Ta 2Zr (5) This gives rise to the solid solutions La,-,Ta, -zZr2x07: 0 <x <0.165 and La, +,Ta,Zr2 -2y07:0 <y <0.44. This mech- anism also appears to operate in the LaTaO, solid solutions on the join LaTa0,-ZrO,, giving La, -xTal -xZr2x04: 0<x <0.06.Similar mechanisms were apparent in the solid solutions in the corresponding system La2O3-ZrO2-Nb2O5.' J. MATER. CHEM., 1991 VOL. 1 Table 1 Powder X-ray data for ZrLaTa,O,, dobs./A dcalc.lA h k 1 I 6.224 6.225 0 2 0 17 5.438 5.442 1 0 1 39 4.984 4.986 1 1 1 3 4.097 4.097 1 2 1 18 4.098 2 2 0 3.113 3.113 0 4 0 52 3.047 3.047 3 1 1 77 2.805 2.804 0 2 2 100 2.806 3 2 1 2.702 2.702 2 4 0 12 2.702 1 4 1 2.505 2.506 3 3 1 28 2.505 0 3 2 2.493 2.493 2 2 2 10 2.074 2.075 0 6 0 4 2.056 2.057 4 0 2 7 2.056 1 0 3 2.030 2.030 5 1 1 8 2.030 4 1 2 1.952 1.952 3 5 1 23 1.814 1.815 6 0 0 33 1.814 3 0 3 1.732 1.731 0 6 2 30 1.732 3 6 1 1.716 1.716 4 4 2 7 1.716 1 4 3 1.651 1.650 4 6 0 6 1.650 2 6 2 1.568 1.568 6 4 0 35 1.557 1.556 0 8 0 14 1.548 1.548 3 7 1 12 1.548 0 7 2 1.523 1.524 6 2 2 13 1.523 0 2 4 1.509 1.509 2 0 4 6 1.509 7 2 0 1.494 1.496 1 8 1 5 1.468 1.468 7 2 1 8 1.460 1.461 1 6 3 4 1.357 1.358 2 4 4 6 1.342 1.341 1 9 1 3 1.341 2 9 0 1.328 1.328 0 5 4 6 1.265 1.266 3 9 1 4 1.266 0 9 2 1.253 1.253 6 6 2 9 1.224 1.224 2 0 5 4 1.182 1.182 3 1 5 14 1.182 6 1 4 1.167 1.167 6 2 4 10 1.167 9 2 1 1.158 1.158 1 4 5 7 1.143 1.142 9 3 1 6 Unit cell: a=10.890(3) A; b=12.450(3) A; c=6.282(2) A; 1/=851.73A3; D,b,=7.637gcm-3, ~,,,,,=7.40gcm-~, for ~=4.Occurrence of mechanism (5) may be understood from size considerations. Zr is slightly larger than Ta [octahedral bond lengths to oxygen are: Ta-0, 2.04 A, Zr-0, 2.12 8, ref. (14)] but is rather smaller than La (for CN=8: Zr-0, 2.24 A, La-0, 2.56 A). Since size differences are not too great, coupled substitution of Zr onto Ta/La sites, and vice versa, is possible. In the so!id solutions based on LaTa309, the mechanism appears to be: 3Ta S La + 3Zr (6) The resulting solid solution formula is La, +.Ta3 -3xZr3x09: O< x <0.104. The end-member of this series, with x = 1 would be the hypothetical phase 'La2Zr309': hence this solid solution runs parallel to the Zr02-Ta205 edge at a constant 25% La203 content.Mechanism (6)appears unlikely at first sight since it involves the creation of interstitial La3 + ions. However, the defective perovskite structure of LaTa309, with only 1/3 of the La sites occupied, contains plenty of interstitial sites for additional La3+ ions. It is possible that mechanism (6)is also responsible for the LaTa501, solid solutions, but since these extend to lower Zr contents and have a wider range of La:Ta ratios, this is less clear-cut. Since LaTa5014 appears to have an Re0,-related structure, there are plenty of interstitial sites available for La3+ ions. The complete isothermal section of the phase diagram for the system Zr0,-La203-Ta205 at 1500 "C has been deter- mined (Fig.1). In addition to the five ternary solid-solution fields and the new phase described above, it contains a large number of two- and three-phase regions. Electrical Properties The electrical conductivities of a selection of the new materials have been measured using the a.c. impedance technique, as previously described.' Most have low conductivity at high temperatures, but one, LaTa309, appears to have a high conductivity of oxide ions, comparable to that of yttria-stabilised zirconia at 400 "C. Details are as follows. The data for LaTa309 and its solid solutions displayed features that indicated it could be an oxide ion conductor. A typical impedance data set is shown in Fig.3 for stoichiometric LaTa309 at one temperature, 508 "C and spanning the fre- quency range 10-'-106 Hz (data recorded with a Solartron 125011286 set-up and a Hewlett Packard RF bridge). The data show three clear features: (i) a high-frequency arc, with associated parallel capacitance 6 pF, which is attributable to the bulk response of the ceramic and the associated resistance of which is given from the low-frequency intercept of the arc on the Z' axis, ca. 5.5 kR; (ii) an intermediate-frequency arc, with capacitance ca. 1 nF, which is attributable to a resistive grain boundary of resistance ca. 200 kR at 508 "C; (iii) a low- frequency inclined spike, of capacitance ca. 0.5 pF, which is attributable to electrode polarisation and the rate-limiting diffusion of oxygen molecules through the electrodes to the elec trode/ceramic interface.Such behaviour is typical of oxide ion conductors. In addition, the bulk resistance value, Rb, was found to be uninfluenced by changing the ambient atmosphere from air to argon during the conductivity measurements, again typical of oxide (or other) ion conductors. Bulk conductivity data obtained from plots such as those shown in Fig. 3 for LaTa309 and its solid solutions are presented in Arrhenius format in Fig. 4 and Table 2. Highest conductivity was found for LaTa309 and a solid solution Lal.033Ta2,9Zro. but decreased at higher Zr contents. On the join La203-Ta205, the conductivity decreased for solid solutions to either side of LaTa309 (Lao,955Ta3,02709 and La, ,045Ta2~97309). For one composition, data measured in both air and argon are shown and are essentially coincident.Further experiments are in progress to determine the transport number of oxide ions in LaTa309. A combined conductivity Arrhenius plot for the other compositions that were measured is given in Fig. 5, with activation energies and conductivity values at 1000 K listed in Table 2. Of these, La3Ta07 and its solid solutions have the highest conductivity, but since their conductivity decreases on changing the atmosphere from air to argon (Fig. 6) it is concluded that their conduction is electronic and p-type. The other phases, LaTaSOl4, La2Zr207, LaTaO,, TaZr2.,,04 and Table 2 Conductivity data composition (mol%) ZrO, La20, Ta205 a/lt-'cm-' at 1OOOK E,/eV" La,Zr ,O 54.0 40.0 6.0 4.oX10-7 1.274 La TaO 75.0 25.0 2.7 x10-5 0.969 6.0 71.0 23.0 1.8 x lo-' 0.8 19 13.0 67.0 20.0 5.4 x lo-' 0.790 20.0 62.0 18.0 2.0 x (in air) 6.0 x SO-' (in Ar) LaTaO, 0.679 0.543 4.0 48.0 48.0 1.1 x 1.387 LaTa,O, 24.0 76.0 2.2 x 10-4 0.420 25.0 75.0 4.9 x 10-4 0.385 26.0 74.0 3.8 x 10-5b 0.543 10.0 25.0 65.0 6.3 x 10-4 0.385 22.0 25.0 53.0 1.8 x lop4 (in air-Ar) LaTa,O 0.470 3.0 16.0 81.0 1.8 x so-5 0.983 TaZr2.7 5O8 88.0 12.0 2.1 x 1.110 33.3 16.7 50.0 7.0 x lop8 ZrLaTa,O, , 1.337 a 1 eVz1.602 x J. Sample contained LaTa,O, (solid solution) plus trace LaTaO,.6.3xl 02Hz O. 07.2 0 0 E 0 C 0 -$.4.8 0 h 0 0 0 2.4-0 0 0 5x105Hz I I I I 1 f 2.4 4.8 7.2 9.6 12.0 Rb 2'110~R cm 3.6 5 2.4 0c: 0 0 0-0 0. 0 0 k I 1.2 6.3xl 02Hz 0 0 -.----_.-- - I7 00 O 0 0 00~ 50Hz ooooo ffo ; Ow046.3~104p 1 I 1 1.2 12.4 3.6 4.8 6.0 Fig.3 Impedance data for LaTa,O, in air at 508 "C: (a) high-frequency data, >630 Hz; (b)low-frequency data, (65 kHz. Data sets overlap at intermediate frequencies J. MATER. CHEM., 1991 VOL. 1 TI"C 1000 700 500 300 i i i 1 1 1 1 .o 1.5 0 103~1T Fig. 4 Conductivity data for LaTa,09 and its solid solutions: (i) ZrO, (0.9): Y,O, (0.1); (ii) La, ,033Ta2,9Zro, (iii) LaTa,O,; (iv) La0.955Ta3.02709; (v) La1.07Ta2.78Zr0.2209; (vl) La1.045Ta2.97309. Data for one composition, (v), were measured both in air (+) and in argon (0).Data for yttria-stabilised zirconia are shown for com-par i son TI'C 1500 1000 700 500 300 1.o 1.5 2.0 lo3KIT Fig.5 Conductivity data for various materials in the system Zr0,- La,O,-Ta,O,. Data for yttria-stabilised zirconia are shown for comparison. 0, LaTa309; 0, La,Ta07; 0, LaTa,O,,; +,TaZr,,,,08; A, LaTaO,; A,La2Zr,07; (>, ZrLaTa,O,, ZrLaTa,O, all have much lower conductivity. Indeed, the conductivity of the new phase ZrLaTa,O,, is so low (3 x lO-'R-l cm-' at 800 "C)that it may find applications as an electrical insulator. Permittivity data for all the new materials were determined at the same time as their conductivities.In all cases, the high- frequency permittivity was small with no indication of ferro- electric behaviour. J. MATER. CHEM., 1991 VOL. 1 TI“C 1000 700 500 300 I I I I 1 -3 -4 I E -5 -. b cn --6 -7 1.o 1.5 2.0 lo3 K/ T Fig. 6 Conductivity data for La,TaO, solid solution of composition La2.747Zro.443Tao,,9707showing atmosphere dependence and p-type conduction. 0,In argon; 0,in air A.R.W. thanks SERC for a research grant. We thank J. T. S. Irvine for advice on conductivity measurements. References 1 C. Zheng and A. R. West, Br. Ceram. Trans. J., 1990, 89, 138. 2 N. S. Afonskii and M. Neiman, Znorg. Muter., 1967, 3, 1132. 3 P. N.Iyer and A. J. Smith, Acta Crystallogr., 23, 1967, 740. 4 H.P. Rooksby, E. A. D. White and S. A. Langston, J. Am. Ceram. SOC., 1965, 48, 447. 5 H. P. Rooksby and E. A. D. White, J. Am. Ceram. SOC., 1964, 47, 95. 6 A. J. Dyer and E. A. D. White, Trans. Br. Ceram. Soc., 1964, 63, 301. 7 M. M. Pinaeva, V. V. Kuznetsova, A. B. Ustimovich and L. D. Shul’man, Znorg. Muter., 1973, 9, 563. 8 L. N. Lykova and L. M. Kovba, Russ. J. Znorg. Chem. (Engl. Trans.), 1971, 16, 459. 9 C. Keller, 2. Anorg. Allg. Chem., 1962, 318, 89. 10 V. S. Stubican, J. Am. Ceram. Soc., 1964, 47, 55. 11 B. W. King, J. Schultz, E. A. Durbin and W. H. Duckworth, in Phase Diagrams for Ceramists, American Ceramic Society, Col- umbus, 1964, diagram 374. 12 R. S. Roth and J. L. Waring, in Phase Diagramsfor Ceramists, American Ceramic Society, Columbus, 1975, diagram 4458. 13 JCPDS Powder Diffraction File, Card 8-246. 14 R. D. Shannon, Acta Crystallogr., Sect. A, 32,751. Paper 0/02053B; Received 9th May, 1990