J. MATER. CHEM., 1991, 1(4), 545-549 Phase Relations in the System BiO,.,-SrO-CuO at 1123 K K. Thomas Jacob and Tom Mathews Department of Metallurgy, Indian Institute of Science, Bangalore 560 012, India Phase relations in the system Bi-Sr-Cu-0 at 1123 K have been investigated using optical microscopy, electron- probe microanalysis (EPMA) and powder X-ray diffraction (XRD) of equilibrated samples. Differential thermal analysis (DTA) was used to confirm liquid formation for compositions rich in BiO,,,. Compositions along the three pseudo-binary sections and inside the pseudo-ternary triangle have been examined. The attainment of equilibrium was facilitated by the use of freshly prepared SrO as the starting material. The loss of Bi,O, from the sample was minimized by double encapsulation.A complete phase diagram at 1123 K is presented. It differs significantly from versions of the phase diagram published recently. Keywords: Phase equilibria; Coexistence domain; Bi-Sr-Cu-0 system; Oxide superconductor; Bismuth Oxide Superconductor. Three superconducting phases have been found in the Bi-Sr- Ca-Cu-0 system.1-3 A clear understanding of phase relations in the system is useful for optimizing conditions for the synthesis of the superconducting compounds. For most practi- cal purposes the system can be represented as a pseudo-quaternary BiO .,-SrO-CaO-CuO. A prerequisite for under- standing phase equilibria in the four-component system is an adequate definition of phase relations in the bounding binary and ternary systems.As a part of a larger programme of research, this paper discusses equilibrium compatibilities between phases in the system Bi01.5-Sr0-Cu0 at 1123 K. Prior to the discovery of high-T, superconducting oxides there were no systematic studies of phase relations in the Bi-Sr-Cu-0 system. Earlier studies were confined to the bounding pseudo-binary sections of the BiO,+,-SrO-CuO system. Sillen and Aurivillius4 have deduced the structure of rhombohedra1 Bil -xSrxOl~,~o~sx for 0.14<x<0.25. Phase relations in the bismuth-rich region of the Bi01.5-Sr0 system were determined by Levin and Roth' using high-temperature X-ray diffraction. More information on phase relations was provided by Guillermo et aL6 They identified four additional phases, Bi2Sr04, Bi2Sr205, Bi2Sr306 stable above 1090 K and a solid solution containing ca.45 mol% SrO stable above 1043 K. Phase diagrams for the system BiO1.,-SrO are given in ref. 7. In the SrO-CuO binary system, three compounds, Sr,Cu03, SrCuO, and SrCu203, were identified by Teske and Muller-Buschbaum.8-11 Cassedanne and Campelo' determined the phase diagram for the BiO1.,-CuO system by analysing quenched samples. They identified the compound Bi4Cu07. More recent studies by Kakhan et ~1.'~using thermogravimetric, thermal and X-ray diffraction analysis indicate that the only compound in this pseudo-binary system is Bi2Cu04 with space group P4/rncc. Available information on the BiO,,,-CuO system is reviewed in ref.7. The powder diffraction pattern for Bi4Cu0712 is surprisingly similar to that of Bi,Cu04,13 for which two crystal structure refinements have been rep~rted.'~~'~ During the course of the present study partial phase relations in the ternary BiO, .,-SrO-CuO system have been published by Saggio et all6 at 1073 K and Ikeda et ~1.'~at 1173 K. Both groups did not study phase relations in the vicinity of BiO,,,-CuO binary. Saggio et ~1.'~added 0.5 wt.% of Li2C03 as a mineralizer to their samples to accelerate the attainment of equilibrium. They16 reported four pseudo-ternary phases, Bi4Sr9CuO16 (491), Bi2Sr,Cu2012 (272), Bi,Sr2Cu06 (221) and a solid solution close to the compound 221. The solid-solution phase was found to be superconducting at the high-strontium-content limit of solubility, with a transition temperature of ca.9 K. The solid solubility range is characterised by the Bi:Sr:Cu ratio of (1 1 -x):(9+x):5, 0 <x c0.4. The compounds 491 and 272 were probably stabil- ized by the presence of Li2C03. Ikeda et all7 identified four ternary phases using transmission electron microscopy (TEM) and XRD techniques, Bi4Sr8Cu50z (485), Bi2Sr3Cu2OZ (232), Bil,Sr16Cu70, (17 16 7) and the solid solution Bi, +$r2 -xCul (0.1<x <0.6; 0 <y <x/2). Superconduc- tivity with T,zlO K was found at the strontium-rich end of the solid solution, in agreement with Saggio et all6 Recently, a more comprehensive study of phase relations in the ternary BiO .,-SrO-CuO system including the binary systems Bi01.5-Sr0 and SrO-CuO has been presented by Roth et a2.l' in the temperature range 1148-1198 K.In agreement with Ikeda et al.17 these authors also report the presence of compounds 485, 232, 221 and a solid-solution region close to the 221 compound. The phase designated as Sr,Cu,O, was found to have the more complex stoichiometry Sr14Cu24041.18-20 Surprisingly, the phase diagram suggested by Roth et shows the presence of the compound Bi,CuO, and absence of the liquid phase along the BiO,.,-CuO binary. This is in conflict with the binary phase diagrams given in ref. 7. Both Ikeda et ~1.'~and Roth et a1." did not delineate phase relations with certainty in some regions of the ternary system. In earlier ~tudies'~,'~ the samples were held in open containers. The possible change in composition or the presence of gradients in composition caused by preferential vaporiz- ation of Bi203 was not considered. Roth et a1." fired their samples either as small pellets or in small 3 mm diameter gold tubes, either sealed or unsealed.They observed that repeated heat treatments in the gold tube resulted in noticeable loss of copper to the gold tube. This would also result in a change in the composition of the sample. In all earlier studies16-'* the samples were synthesized from SrCO, at atmospheric pressure. Since the decomposition temperature of SrC0, is 1455K,,l C02 may have been an important constituent of some of the final phases obtained. Evidence for C0,-stabilized phases in the Y,O,-BaO-CuO system has recently been d~cumented.~~.~~ In view of the uncertainties in the reported phase relations, especially in the vicinity of the BiO,.,-CuO binary, results of our investigations are presented in full.Experimental Starting materials used in the preparation of pseudo-binary and pseudo-ternary compositions were powders of Bi203, SrC03 and CuO, each of 99.99% purity. SrC03, contained in a stabilized-zirconia crucible, was decomposed in vucuo (1 Pa) at 1000 K to SrO. The SrO formed by decomposition under vacuum was found to be highly reactive. When SrO was used as a starting material rather than SrC03, the time required to reach equilibrium was significantly reduced. The oxides were weighed and then mixed, either dry or with acetone, using an agate mortar and pestle.The intimate mixture was pressed into pellets using a steel die. The pellets were placed inside a small zirconia crucible and covered with a zirconia lid. The main pellet was placed on top of a thin platform of identical composition. The crucible was placed inside a larger zirconia crucible. The space between the two zirconia crucibles was packed with loose powder of the same composition as the pellet. The larger zirconia crucible was also covered with a lid and sealed with a zirconia-based cement. A schematic diagram of the arrangement is shown in Fig. 1. The closed system was used to minimize the volatiliz- ation of Bi203. Values for the vapour pressure of Bi203 are not available in the literature.From the mass loss of Bi203 in flowing oxygen at 1123 K, an apparent vapour pressure of 150 Pa was obtained in preliminary experiments. The joint between the lid and crucible was not air-tight; therefore, a small loss of Bi203 was still encountered from the closed crucibles. The double-crucible assembly was effective in pre- venting vaporization loss. The loss of material by vaporization from the outer crucible occurred preferentially from the pow- der packed around the inner crucible. The composition of the pellet inside the inner crucible remained virtually unaltered. The pellet was heat treated several times at 1123 K with grinding and repelletizing between each heat treatment. Each treatment lasted ca.24 h. Generally, the phase composition of the pellet became invariant with time after three heat treatments. For a few compositions, however, up to five heat treatments were found to be necessary to reach equilibrium. The crucible assembly was suspended in the even-tempera- ture zone of a vertical furnace as shown in Fig. 2. Pure oxygen was allowed to flow around the crucible. The temperature of the furnace was controlled to within +1 K and measured by a calibrated Pt/Pt-l3%Rh thermocouple. After each heat treatment the crucible assembly was quenched by dropping it into liquid nitrogen. The quenched pellet was examined by optical microscopy, XRD and EPMA. In XRD Cu-Ka radi-ation and Ni filters were used. Because of the large difference in the X-ray scattering factor between Bi and other elements, detection of Bi-free phases, especially at low concentrations, in the presence of Bi-containing phases, was found to be difficult.In such cases, EPMA was very useful in identifying Bi-free phases. A beam size of 1 ym was used. Synthetic samples of CuO, SrCuO, and Sr,Bi209 were used as stan- dards. However, to obtain good reproducibility by EPMA the grain size had to be larger than 40 ym. When powders of this size were used for the preparation of the samples, the packing powder zirconia crucible with lid ' sample pellet sample platform Fig. 1 Schematic diagram of the double encapsulation system for containing the samples J. MATER. CHEM., 1991. VOL. 1 ?--brass cap furnace alumina tube sample water-cooled joint I =t-02 inlet cellophane paper -~t--=~+-z:=-l _-----Fig.2 Schematic diagram of the apparatus for equilibration and quenching time required for reaching equilibrium was increased consider- ably. When phase identification could be established solely by XRD, fine powders of starting oxides were used to minimize the equilibration period. The samples were polished with a diamond paste prior to optical microscopy and EPMA. Complete melting of the samples was easily identified by visual inspection. However, detection of partial melting in some samples was difficult because the small amount of liquid formed was held between solid grains by surface-tension forces, and the external appearance of the pellet remained unaltered.The liquid phase was found to crystallize despite rapid quenching. DTA was used as a supplementary tool in such cases. The sample was placed in a gold crucible under pure oxygen. In a few experiments equilibrium was approached from different directions by using different syn- thetic compounds as starting materials. The samples were stored in a desiccator, although they were not as sensitive to moisture as compositions from the Y203-Ba0-Cu0 system. Results The overall chemical compositions of samples examined in this study are shown in Fig. 3. Note that the entire compo- sition space of the ternary system is covered. Phase identifi- cation in equilibrated samples was based on comparison with published diffraction patterns and those calculated from pub- lished structural data.l6-I8 The crystal structures, lattice parameters and the sources of diffraction patterns for the compounds are listed in Table 1.Although there are differences J. MATER. CHEM., 1991. VOL. I CUO / . .0.. .\Y4 / . \ SrO/ v -v -u u-u-v -&-\/0.8 -A/ -0.2 0.4 0.6 -'BiO, XBlO, 5 Fig. 3 Overall chemical compositions of the different samples exam- ined in this study in the structure descriptions of the phases reported in the literature, the diffractograms are similar. For the purpose of phase identification, differences in the reported crystal struc- tures listed in Table 1 are not important. Optical microscopy and EPMA provided additional information. The compo- sitions of non-stoichiometric phases were obtained primarily from EPMA.No phase other than those already reported in the literature has been identified in this study. The isothermal section of the phase diagram for the BiO,.,-SrO-CuO system at 1123 K produced from the results of this study is shown in Fig. 4. Along the pseudo-binary BiO, ,,-SrO, four compounds, SrBi204, Sr2Bi,0S, Sr3Bi206 and Sr,Bi,09, and two solid solutions, /? and y,were identified. The composition of the /? phase varied from 12.3 to 26.5 mol% SrO. The y phase ranges from 43.4 to 45.3 mol% SrO. There is a small liquid-phase Table 1 Crystal structure and lattice parameters of phases in the system BiO,,,-SrO-CuO lattice parameter/A compound a b 22 1 23.73 13.242 24.493 5.4223 232 24.804 5.396 24.937 5.395 33.907 23.966 34.035 24.05 33.991 24.095 6" 26.856 5.380 5.389 5.385 26.889 5.384 SrBi,04 19.301 4.3563 Sr, Bi,O , 14.293 7.651 14.307 6.1713 Sr,Bi,06 12.526 Sr,Bi,O, 6.009 Sr,CuO, 12.68 3.9 1 12.684 3.9064 SrCuO, 16.33 13 3.9136 Sr14cu2404 1 13.0 11.3 13.399 11.483 B 3.979 Y 13.239 region near pure Bi01.5.These results are identical with the data for the pseudo-binary reported by Roth et ~l.,'~but differ from the phase diagram suggested by Guillermo et aL6 Along the SrO-CuO pseudo-binary, three compounds, Sr2Cu03, SrCu02 and were detected at 1123 K, in agreement with Saggio et ~1.'~and Roth et a1." Along the BiOl .,-CuO pseudo-binary no compounds were detected at 1123 K.A liquid phase was found to be present from 66 to 100mol% BiO1.s. A two-phase region exists between CuO and the liquid phase. Phase relations along this pseudo-binary are in agreement with the diagram proposed by Kakhan et ~1.'~The phase diagram suggested by Cassedanne and Cam- pelo12 shows a much more restricted liquid-phase field ranging from 80 to 100 mol% BiO,.,. The phase diagram suggested by Roth et a1.18 does not show the formation of a liquid phase but instead indicates the presence of solid phases including Bi,Cu04 in the temperature range 1148-1 198 K. Four pseudo-ternary solid phases were detected, Bi,Sr8Cu5019+x (485), Bi2Sr3Cu,08 (232), Bi,Sr2Cu06 (221) and a solid solution 6 which can be approximately charac- terized as Bi2.44-xSrl.,6+xCul +,,Oz,where 0 <x <0.24 and -0.05 <y <0.1 1.The compounds 221 and 232 were found to be stoichiometric. The X-ray diffraction patterns and EPMA of these phases were almost identical in the different three- phase regions. Compound 221 is deficient in CuO by ca. 1 mol% and exhibits a small variation in the Sr:Bi ratio. The fifth single-phase region in the Gibbs' triangle is an extension of the liquid phase present along the BiO,.,-SrO and BiO ,-CuO binaries. Compounds 491 and 272 identified by Saggio et ~1.'~have not been detected in this study. Their occurrence is probably caused by the addition of a small amount of Li2C03 as a mineralizer. These authors did not observe the compounds 485 and 232 detected in this study, but these were identified by Ikeda et a1.17 and Roth et all8 The presence of compound 221 and solid solution 6 has been established by all investi- gators,16-" although there are some differences in the exact compositions reported for these phases.The ternary diagram shown in Fig. 4 differs significantly from that suggested by Saggio et ~1.'~They show only three c PI" 4.08 1 2 1.959 105.40 4.888 19.094 96.97 5.373 5.389 5.3677 26.908 113.55 24.64 24.630 26.933 1 13.67 6.1049 94.85 6.172 3.8262 18.331 58.663 3.48 3.4957 3.5730 3.94 3.9356 28.5 1 4.257 structure ref. orthorhombic 16 monoclinic, C2/m 27 orthorhombic 17 monoclinic, C2/m 18 orthorhombic 24 orthorhombic, Fmmm 17 orthorhombic, Fmmm 18 monoclinic, C2 25 pseudotetragonal 16 pseudotetragonal 17 monoclinic, C2 18 monoclinic, C2/m 18 orthorhombic, Pcmm 6 orthorhombic, Cmcm 18 rhombohedral, R3m 18 rhombohedral 18 orthorhombic, Immm 8 orthorhombic, Immm 18 orthorhombic, Cmcm 26 orthorhombic 20 orthorhombic, Fmmm 18 rhombohedral 6 tetragonal, 14/m 6 J.MATER. CHEM., 1991. VOL. 1 CUO Sr Bi01.5 Fig. 4 Isothermal section (1123 K) of the phase diagram for the system BiO,.,-SrO-CuO: 1, Bi,Sr,Cu,O,,+,; 2, Bi,Sr,Cu,O,; 3, Bi,Sr,CuO,; 4, 6; 5, liquid strontium bismuthates. About one quarter of the Gibbs' triangle adjacent to the pseudo-binary BiO 1,,-CuO where the liquid phase is present has not been explored at all by Saggio et ~1.'~or in detail by Ikeda et all7 The possible presence of solid phases including Bi2Cu04 in this region of the diagram has been suggested, however, Ikeda et ~1.'~have not reported the presence of SrBi204.The phase relations in the region of the ternary phase diagram bounded by the compounds Sr3Cu50z, SrCu02 and the pseudo-ternary phases 485 and 232, have not been well characterized by Ikeda et ~1.'~There are also some differences in the exact compositions of the 221 and 6 phases. In other respects, the major phase fields identified by them are in agreement with the results of this study. The phase diagram suggested by Roth et all8 in the temperature range 1148-1198 K is in good agreement with the results obtained in this study except for the region near the BiO1.,-CuO binary.Results of this study indicate an extensive liquid-phase region which does not appear in the phase diagram suggested by Roth et a1.,18 instead they show the presence of solid phases including Bi2Cu04. Conclusion The isothermal section of the phase diagram at 1123 K for the system BiOl .,-SrO-CuO has been completely charac- terized by analysis of equilibrated and quenched samples using optical microscopy, XRD and EPMA. A limited number of DTA experiments have been conducted to delineate the liquid region. By minimizing C02 contamination of the sample, reproducible and consistent results were obtained for phase relations in the pseudo-ternary system.The loss of Bi203 from the sample was minimized by double containment. Four compounds, SrBi204, Sr2Bi205, Sr,Bi206, Sr6Bi20,, and two solid-solution series, /? and y, were detected along the BiOl .,-SrO binary. Three compounds, Sr2Cu03, SrCuO, and Sr14Cu24041, were found to be stable in the system SrO- CuO. There is an extensive liquid-phase region near BiO 1., in the Gibbs triangle. Four pseudo-ternary solid phases were detected: Bi4Sr8Cu5019+,, Bi2Sr3Cu208, Bi2Sr2Cu06 and a solid solution, 6, which can be approximately characterized as Bi2.44-,Sr .56 + ,Cu +,,OZ where 0 <x <0.24 and -0.05 <y <0.11. The compound Bi2Sr2Cu06 has a small homogeneity range and is deficient in CuO by ca. 1 mol%. The authors are grateful to Mr.A. V. Narayana for assistance in the preparation of the manuscript. References C. Michel, M. Hervieu, M. M. Bore], A. Grandin, F. Deslandes, J. Provost and B. Raveau, 2.Phys. B, 1987, 68,421. J. Akimitsu, A. Yamazaki, H. Sawa and H. Fujiki, Jpn. J. Appl. Phys., 1987, 26, L2080. H. Maeda, Y. Tanaka, M. Fukutomi and T. Asano, Jpn. J. Appl. Phys., 1988, 27, L209. L. G. Sillen and B. Aurivillius, Z. Kristallogr., 1939, 101, 483. E. M. Levin and R. S. Roth, J. Res. Nut. Bur. Stand., Sect. A, 1964,68, 197. R. Guillermo, P. Conflant, J. C. Boivin and D. Thomas, Rev. Chim. Mineral., 1978, 15, 153. R. S. Roth, J. R. Dennis and H. F. McMurdie, Phase Diagrams for Ceramicists, American Ceramic Society, Westerville, OH, 1987. C.L. Teske and H. Muller-Buschbaum, Z. Anorg. Allg. Chem., 1969, 371, 325. J. MATER. CHEM., 1991. VOL. 1 549 9 10 11 12 13 14 15 C. L. Teske and H. Muller-Buschbaum, Z. Anorg. Allg. Chem., 1970, 379, 234. C. L. Teske and H. Muller-Buschbaum, Z. Anorg. Allg. Chem., 1970, 379, 113. H. Muller-Buschbaum, Ang. Chem. Int. Ed. Engl., 1977, 16, 674. J. Cassedanne and C. P. Campelo, Ann. Acad. Bras. Cienc., 1966, 38, 35. B. G. Kakhan, V. B. Lazarev and I. S. Shaplygin, Russ. J. Inorg. Chem., 1979, 24, 922. V. R. Arpe and H. Muller-Buschbaum, Z. Anorg. Allg. Chem., 1976, 426, 1. J. C. Boivin, J. Trehoux and D. Thomas, Bull. Mineral., 1976, 99, 193. 19 20 21 22 23 24 25 E. M. McCarron 111, M. A. Subramanian, J. C. Calabrese and R. L. Harlow, Mater. Res. Bull., 1988, 23, 1355.T. Siegrist, L. F. Schneemeyer, S. A. Sunshine, J. V. Waszczak and R. S. Roth, Mater. Res. Bull., 1988, 23, 1429. Y. A. Chang and N. Ahmad, Thermodynamic Data on Metal Carbonates and Related Oxides, The Metallurgical Society of AIME, Warrendale, Pennsylvania 15086, 1982. D. M. DeLeeuw, C. A. H. A. Mutsaers, C. Langereis, H. C. A. Smoorenburg and P. J. Rommers, Phys. C, 1988, 152, 39. G. M. Kale and K. T. Jacob, Chem. Mater., 1989, 1, 515. A. Fuertes, C. Miravitlles, J. Gonzalez-Calbet, M. Vallet-Regi, X. Obradors and J. Rodriguez-Carvajal, Phys. C, 1989, 157, 525. M. Onada and M. Sato, Solid State Commun., 1988, 67, 799. 16 17 18 J. A. Saggio, K. Sujata, J. Hahn, S. J. Hwu, K. R. Poeppelmeier and T. 0. Mason, J. Am. Ceram. SOC., 1989, 72, 849. Y. Ikeda, H. Ito, S. Shimomura, Y. Oue, K. Inaba, Z. Hiroi and M. Takano, Phys. C, 1989, 159, 93. R. S. Roth, C. J. Rawn, B. P. Burton and F. Beech, J. Res. Nat. 26 27 W. K. Wong-Ng, H. F. McMurdie, B. Paretzkin, C. R. Hubbard and A, L. Dragoo, Powd. Difl., 1988, 3, 117. R. S. Roth, C. J. Rawn and L. A. Bendersky, J. Mater. Res., 1990, 5, 46. Inst. Stand. Technol., 1990, 95, 291. Paper 1/00034I; Received 3rd January, 1991