J. MATER. CHEM., 1991, 1(2), 175-180 Structure Evolution during Thermal Processing of High-Tc Ceramic Superconductors produced using Sol-Gel Techniques George Kordas,ta Glenn A. Moore/ J. D. Jorgensen,b F. Rotella,b R. L. Hitterman,b K. J. Volinb and J. Faberc a Department of Material Science and Engineering, Ceramics Division, Science and Technology Center for Superconductivity and Materials Research Laboratory, University of Illinois at Urbana- Champaign, 105 S. Goodwin, Urbana, lL 61801, USA Argonne National Laboratory, Argonne, lL 60439, USA Amoco Corporation, Naperville, lL 60566, USA Characterization of the decomposition and phase development reactions for Cu" methoxyethoxide, 2Ba-3Cu, Y-~CU, and Y-2Ba-3Cu gels has been performed using in situ neutron diffraction and differential thermal analysis (DTA).The structural evolution process of a Y-2Ba-3Cu gel to the superconducting YBa,Cu,O,~, phase was elucidated using the acquired spectra. Precursor gel powder was produced from an alkoxide sol-gel system having a methoxyethanol, methyl ethyl ketone and toluene solvent mixture. Neutron diffraction results indicate initial YBa,Cu,O,-, formation below 700 "C, and that a firing temperature of 850 "C is sufficient to produce essentially single-phase YBa,Cu,O,-,. The conversion of BaCO, to BaCO, polymorph was observed to occur between 790 and 800 "C, after which rapid formation of the YBa,Cu,O,-, tetragonal phase occurred. The results also indicate an incomplete conversion of the non-superconducting tetragonal phase to the supercon- ducting orthorhombic phase during the cool-down and annealing segments of the heat treatment.Keywords: Sol-gel processing; Y-Ba-Cu-0 system; Neutron diffraction; Superconductivity 1. Introduction The processing of single-phase YBa2Cu307-supercon-ducting powder and thin films is paramount to basic materials research studies as well as commercial product developments. Initially, single-phase superconductors were produced by the solid-state reactions of yttrium oxide, barium carbonate, and copper oxide powders.' A severe drawback of this process is the necessary 950 "C processing temperature for several hours and the subsequent annealing process in oxygen near 450 "C to yield the orthorhombic superconducting phase. The high calcination temperature usually results in extensive micro- structural grain growth' and interaction of the supercon- ducting film with the substrate.,q4 Sol-gel processing techniques offer an alternative to the traditional ceramic processing methods.Research on sol-gel processing over the last 20 years has demonstrated that the purity and microstructure of resulting materials can be con- trolled through variation of the starting chemistry and pro- cessing parameters. For example, microstructure control can be achieved by variation of precursor type, water and solvent concentration, pH, and mixing and firing temperatures.' Most frequently, the sol-gel method uses metal alkoxides, which hydrolyse and condense to form polymeric networks. Follow- ing gelation, firing at significantly reduced temperatures, com- pared to those used for traditional oxide sintering, is used to form the appropriate ceramic material.For example, Hirano .~et ~1 have produced partial formation of YBa2Cu,07 -x superconducting powder at 650 "C using yttrium isopropox- ide, barium isopropoxide, and copper(I1) 2-ethoxyethoxide. The gel powder was fired in a flowing 02-0,mixture in order to alleviate formation of BaCO, during the organic group decomposition stages. Horowitz et aL7 synthesized YBa2Cu,07-sol-gel superconductor at 650-700 "C t Present address: Demokritos, National Research Center for Physi- cal Sciences, Institute of Materials Science, 153 10 Ag. Paraskevi Attikis, Athens.Greece. through complete hydrolysis of yttrium diisopropoxide, bar- ium diisopropoxide, and copper(I1) n-butoxide. The ability to produce essentially complete YBa,Cu,07 -x orthorhombic conversion was due to the lack of BaC0,. Kramer et a1.* have produced YBa2Cu307 -sol-gel superconductor pow- ders using Y methoxyethoxide, Ba methoxyethoxide, and copper(I1) ethoxide. Firing of the gels at 700 "C resulted in a YBa2Cu307--x orthorhombic conversion of 8-10% owing to the presence of BaCO,. Kramer et al.' also demonstrated that epitaxial films can be produced on SrTiO, via sol-gel routes. Therefore, the sol-gel technique can lead to high- quality superconductors that can be used for physical property measurements as well as technological applications.This publication reports the structural development and characterization of a sol-gel process that we developed for the production of the YBa2Cu307 -x superconductors exten- sively described in a number of previous publication^.^-' ' 2. Experimental Cu" methoxyethoxide, Y methoxyethoxide, and Ba methoxy- ethoxide precursors were synthesized as previously reported.'-' Stable sols having concentrations between 0.01 and 0.05 mol kg-I were obtained by mixing the desired precursors in a methoxyethanol-methyl ethyl ketone-toluene solvent mixture exposed to air. Gel powder was produced by stripping the solvent mixture from the acquired stable sol using a rotary evaporator. The gel powder was subsequently dried in vacuum at 120-200 "C and then pressed into cylindri- cal pellets, 1.O cm diameter x 1.5 cm length at 20-30 psi$ for the neutron diffraction measurements.Prior to the neutron diffraction experiments, characterization of the gel decompo- sition temperatures were performed using differential thermal analysis (DTA). A Perkin-Elmer Model 1600 was utilized and gel powder samples were heated at a rate of 10 "C min-' to $ 1 psi z 6.895 x lo3 Pa. 176 850 "C in flowing oxygen, 0.5 ft3 h-' (static cubic feet per hour).§ All anneals were carried out after cooling in oxygen. The neutron diffraction experiments were performed on the Special Environment and General Purpose Powder Diffractometer, at the Intense Pulsed Neutron Source (IPNS)." The gel pellets were heat-treated in a controlled atmosphere in the powder diffractometer.For each analysed composition, two sample pellets were stacked vertically on an alumina sample platform and a thermocouple placed directly above the pellets in an alumina cap which rested on the sample. Oxygen or argon gas at a flow rate of 2.0 SCFH was circulated in the furnace. Diffraction data were accumulated at a fixed scattering angle of 28=90" by the time-of-flight technique for 2 h at each temperature in order to achieve adequate counting statistics. The data were processed using a VAX 780 computer. An equilibrium of the sample at each temperature was not achieved prior to the data acquisition owing to instrument time constraints. However, the furnace control parameters allowed a ca.20 min stabilization period within 10 "C of each temperature setting. The experiments were performed using 5 "C min-' heating and cooling rates unless otherwise noted. 3. Results and Discussion Fig. 1 shows DTA results for the copper(I1) methoxyethoxide precursor powder. This measurement reveals a large exother- mic reaction in the range 200-270 "C and two smaller exother- mic peaks in the range 350-450 "C. Based on this result, room temperature, 275,475 and 675 "C were chosen as the tempera- tures for acquisition of neutron diffraction data for the cop- per(I1) methoxyethoxide sample. The diffraction spectra for the copper(I1) precursor pellets are shown in Fig. 2. The room- temperature data indicate the major constituent to be amorph- ous, with minor constituents being CuO, Cu20, and Cu metal.The presence of crystalline phases is attributed to the 120-200 "C temperature used during sample drying. The copper gel sample appears completely crystalline by 275 "C, with CuO being the sole copper constituent. Thus, the large exothermic reaction observed just below 275 "C is attributed to the formation of CuO and the elimination of its correspond- ing organic byproducts. Neither Cu metal nor CuzO were observed in the 275 "C powder pattern. No apparent crystallo- graphic structure variation was observed in the 475 to 675 "C spectra, indicating that the smaller exothermic reactions occurring in the range 350-450 "C were most likely due to a second stage of organic byproduct oxidation.Our single phase 4 1 ft3 h-' ~7.866 m3 s-'x O" I I J. MATER. CHEM., 1991, VOL. 1 04000 I 0 7 9 1 .o 1.5 2.0 2.5 3.0 dspacing /A Fig. 2 Neutron diffraction patterns for copper(I1) alkoxide consoli-dation. (a) Room temp.; (b) 275 "C; (c) 475 "C; (d)675 "C. O=CuO; 0=Cu,O; A =CU; 0 =A1,03 formation of CuO by 275 "C is 200°C lower than that reported by Hirano et a1.,13 in which similar alkoxide precur- sors were used. They observed the presence of both Cu,O and CuO up to 475 "C as determined using X-ray diffraction techniques. In addition to the copper-containing phases, the neutron diffraction patterns recorded above room temperature revealed the presence of Al,03. As discussed previously, the sample pellets were positioned between two alumina dies in the neutron beam.As a result of sample pellet shrinkage during the copper(I1) precursor run, the upper alumina pos- itioning cap entered the neutron beam. Thus, the spectra above room temperature showed the presence of alumina. However, this interference did not limit the identification of the copper oxide phases. Fig. 3 shows the DTA curve for the Y-3Cu gel sample. The data indicate a sharp exothermic reaction below 275 "C followed by a gradual exothermic rise between 300 and 550 "C. Neutron diffraction data were therefore taken at room tem- perature, 275 and 550°C. The diffraction data are shown in Fig. 4. The data indicate that CuO is the dominating phase in the 275 and 550°C spectra with the balance of material being in the form of Y2Cu205. The presence of yttrium oxide was not observed in the patterns, indicating the direct forma- tion of the yttrium copper oxide.Fig. 5 shows the DTA curve for 2Ba-3Cu gel powder. The curve exhibits three exothermic peaks between 150 and 525 "C, and a small endothermic dip at 800 "C corresponding to the BaC0,-BaC03 polymorph phase conversion. Fig. 6 shows neutron diffraction patterns of samples heat-treated at 325, 0160 I 1 220 I 380 I I 540 I I 700 I I temperature/"C temperature/% Fig. 1 Differential thermal analysis of copper(I1) alkoxide Fig. 3 Differential thermal analysis of Y-3Cu gel J. MATER. CHEM., 1991, VOL. 1 10 1.5 2.0 25 3.0 3.5 dspaci ng /A Fig.4 Neutron diffraction patterns for Y-3Cu gel consolidation. (a) Dried gel; (h)275 "C; (c) 550 "C. O=CuO; A =Y,Cu,O, a,ni I st I0 .-C n 1.o 15 2.0 2.5 30 3.5 dspaci ng /A Fig. 6 Neutron diffraction patterns for 2Ba-3Cu gel consolidation. =BaCuO,; 0 =CuO; A =BaCO, 800, and 850 "C.The first spectrum recorded at 325 "C shows CuO and BaCO, as the predominant phases. The 800 and 850 "C sample spectra contained a large alumina contribution from the positioning cap. For these two samples additional neutron diffraction data were obtained after partially cooling the furnace and masking the portion of the neutron beam impinging on the positioning cap. The spectrum representative of the 800 "C sample, recorded at 450 "C, indicates the pres- ence BaCuO, and CuO.These phases were also detected in the 850 "C sample, recorded at 200 "C, Fig. 6. In the 850 "C sample the BaCuO, phase is much more developed than in the 800 "C sample. Fig. 7 shows the DTA analysis of the Y-2Ba-3Cu gel powder produced using Cu" ethoxide obtained from Johnson Matthey (Alfa products) corporation and processed using the methoxyethanol-methyl ethyl ketone-toluene system. The DTA curve shows two exothermic peaks at 200 and 500 "C. The BaC03-BaC03 polymorph conversion endotherm occurred at 795 "C. This DTA measurement closely resembles that of the 2Ba-3Cu gel. For example, the large exothermic peak between 400 and 500 "C appears to have similar intensit- ies and widths in both plots, indicating that the barium and possibly copper constituent reactions are taking place.Fig. 8 shows the neutron diffraction data of the Y-2Ba-3Cu gel heat-treated from 200 to 850 "C. From Fig. 8 it is observed Fig. 7 Differential thermal analysis of Y-2Ba-3Cu gel 5400 . , <25E 0 Ill I I I I I 10 15 20 25 30 35 dspacing/ A Fig. 8 Neutron diffraction patterns for Y-2Ba-3Cu gel consoli- dation. 0 =BaCuO,; 0=CuO; H=BaCO,; A =BaCO,; A = Y2Cu20,; 0 =tetragonal YBa,Cu,O, -x 178 that the BaCO, peaks do not change appreciably in intensity from 325 to 750"C, but at 800°C the phase has been completely replaced by the BaC0, polymorph. Thus, the BaCO, phase appears to be very inert up to the conversion temperature, ca.790-800 "C. The 2Ba-3Cu results indicate that 2 h at 800 "C was sufficient for the reaction of the newly formed BaCO, polymorph with the copper oxide phases to form Ba2Cu0, to go to completion. This presence of Ba2Cu03 has been observed by Wang et at 950°C in 2Ba:3Cu specimen prepared using the solid-state method. Ba2Cu03 was not observed at any temperature in our 2Ba-3Cu samples. From the room-temperature acquired diffraction data shown in Fig. 8, it is apparent that the Y-2Ba-3Cu gel sample was completely amorphous prior to heat treatment. The hump-like nature of the scattering intensity is representative of the under-moderated Maxwellian neutron distribution, characteristic of a pulsed spallation neutron source. The spectra appear quite complex owing to the three-component nature of the gel.However, with the aid of the previously discussed diffraction patterns, the data can be adequately interpreted. Two main phases, CuO and BaC03, were observed at 325 "C with a small amount of Y2Cu205 also present. Free Cu metal and Cu20 were not observed at any temperature. The Y,Cu205 peak intensities increased through the 750°C acquisition, at which time the phase appeared to react readily with the other constituents to form tetragonal YBa2Cu307-,. The BaC03 phase was very stable through the 750°C acquisition, but at 800°C had been converted to BaCO, polymorph. The BaCO, polymorph was observed to react readily with CuO and Y2Cu205 forming tetragonal YBa2Cu307-, and a small amount of BaCu02.After 2 h at 850 "C only small traces of the BaCO, polymorph phase remained. Residual impurities after 4 h at 850 "C were Y2Cu205, CuO and BaCuO,. Both argon and oxygen atmosphere were used to cool the YBa2C~307-xspecimens from the 850 "C sintering tempera- ture to the annealing temperature of 450 "C at 5 "C min-'. Fig. 9 shows the diffraction data taken while annealing. The first sample was cooled from 850 to 450 "C at a rate of 5 "C min-in flowing argon. This treatment retained the tetragonal phase as indicated in Fig. 9. The single-phase nature of this sample is apparent when compared to the 'standard' tetrag- onal YBa,Cu,O,-, spectra of Fig. 10." This is, of course, exactly what would be expected since the tetragonal-ortho- rhombic transition cannot occur unless oxygen is available.' After the initial diffraction data were acquired at 450 "C, oxygen was introduced into the furnace.After 2 h of oxygen 4000 I I I 3200 fn c.-C3 2400 Y >c.-v) 1600 Q)c .-C 800 0 10 1.5 2.0 2.5 3.0 35 d-spacingI A Fig. 9 Neutron diffraction patterns for YBa,Cu,O, -x annealing treatments. (a)Ar-cooled, no anneal; (b)Ar-cooled, 2-4 h anneal; (c) 0,-cooled 0-2 h anneal; (d)0,-cooled, 2-4 h anneal; (e)0,-cooled 4-5 h anneal. 0 =Tetragonal YBa,Cu,O,-x J. MATER. CHEM., 1991, VOL. 1 3600t 0.8 1.2 I .6 2.0 2.4 2.8 3.2 dspaci ng/A 7WVII (b) I I I I I I 1 0.8 I .2 1.6 2.0 2.4 28 3.2 dspaci ng/A Fig. 10 Neutron diffraction patterns for (a) tetragonal and (b)ortho-rhombic YBa,Cu,O, -x produced using the solid-state method" annealing the data acquisition was started.The data show partial conversion of the tetragonal sample to the orthorhom- bic superconductive phase. Fig. 9 also shows data from the second annealing experi- ment. In this experiment a tetragonal YBa2Cu307 -,sample was cooled in oxygen from 850 "C at a rate of 5 "C min-' and then annealed at 450 "C in oxygen. It was observed that partial orthorhombic conversion occurred during cooling. Annealing at 450°C in oxygen for 4 h produced further orthorhombic conversion. However, after 5 h of annealing, the rate of tetragonal-to-orthorhombic conversion was extremely small. From the peak intensity ratios it appears that the acquired tetragonal-orthorhombic mixture is com- posed of ca.30% tetragonal. This is not the behaviour expected of a sintered YBa2Cu307 -,material, prepared near 900 "C. Typically at elevated temperatures, i.e. 850 "C, a single- phase tetragonal material having 06,05-6,20is present.' As the sample is cooled in oxygen or air, oxygen is intercalated into the structure. At ca. 650°C an oxygen concentration of 06,5is achieved and preferential oxygen ordering with respect to the b axis induces a tetragonal-to-orthorhombic phase conversion. It is reported by Manthiram and Gooden-ough16.17 that for their YB~,CU,O,-~ prepared below 800 "C, the presence of 0 2p holes ('paired' oxygen) in the tetragonal structure hinders preferential oxygen ordering, and thus hin- ders the tetragonal-to-orthorhombic conversion.For their material, processed using oxalate precursor at a minimum temperature of 780 "C for 5 days and annealed in oxygen at 450 and 350 "C for 12 h, it was found that a semiconducting tetragonal phase having 06.7was obtained. This material was completely converted to the superconducting orthorhombic phase by first eliminating 'paired' oxygen in the structure. Elimination of 0 2p holes was achieved by heat treating the 06.7tetragonal material in nitrogen at 750-780 "C for 12 h J. MATER. CHEM., 1991, VOL. 1 or in air at 810 "C. This treatment produced a low oxygen content tetragonal phase which upon reannealing at 400"C converted completely to the superconducting orthorhombic phase.This is reportedly due to the reinsertion of monomeric oxygen at a concentration that prevents formation of 0 2p holes. Our neutron diffraction sample cooled in argon, fol- lowed by oxygen annealing was analogous to the oxygen removal/reinsertion cycle used by Manthiram and Gooden- ough.' However, our material did not completely convert to single-phase orthorhombic YBa,Cu30, -x. Our neutron diffraction annealing experiments were designed to see if an increased oxygen intercalation rate was present in the argon-cooled sample, compared to the sample cooled in oxygen, and whether enhanced phase conversion was achieved by retaining the low oxygen content tetragonal phase during cooling. It can be seen from the data in Fig.9 that the low-oxygen-content tetragonal phase was partially converted to a tetragonal-orthorhombic mixture after 2 h of oxygen annealing. The tetragonal-to-orthorhombic peak intensity ratios of this spectra were very similar to those of the oxygen-cooled sample spectra, which were taken during the first 2 h of annealing at 450 "C. Thus, a greater oxygen intercalation rate was observed during the initial stages of the annealing treatment. However, no appreciable increase in the degree of tetragonal-to-orthorhombic phase conversion was obtained. At the present time an exact explanation for the observed incomplete phase conversion behaviour is not available. Poss-ible contributors to the problem may include variations of the initial stoichiometry or residual carbon impurity phases. It is clear from the neutron diffraction data that our phase evolution process is quite different from that observed for oxide sintered processes as well as for other chemical precursor systems and that impurity phases such as Y2Cu205, not reported in those systems, may affect the conversion process. The literature seems to indicate that when consolidation temperatures of 900-950 "C are used, partial conversion is not a problem.' 5*1 However, when lower conversion tempera- tures are used, broad susceptibility transitions characteristic of incomplete phase conversion are generally observed.l9 Fig. 11 shows the susceptibility curve for the YBa2Cu30,-, sample cooled and annealed in oxygen.The curve shows a broad transition due to the two-phase nature of the fired pellet sample. An onset temperature of ca. 65 K was observed. 6 a Oo0' Isi.c c 0 000 -0,i -0Q -0 -0.001 -.-0 c Q,C 0,E -0.002! * I . I . 1 -I ' I . I 0 20 40 60 80 100 120 temperature/K Fig. 11 Susceptibility curve for YBa,Cu,O, --x sample annealed in flowing oxygen for 5 h at 450 "C 179 Summary and Conclusions The amorphous gel to YBazCu307-, oxide conversion was characterized using in situ time-of-flight neutron diffraction techniques. The results indicate that by 325 "C, CuO, BaCO, and Y2Cu205 have formed, (Fig. 12). Tetragonal YBa,Cu,O,-, was observed to be present as a minor phase at 700 "C.After the BaCO3-BaCO3 polymorph conversion near 800 "C the formation of tetragonal YBa2Cu307 -,as the dominant phase rapidly occurred. After 4 h at 850 "C essen- tially single-phase tetragonal YBa2Cu307-,was present. Small amounts of Y2Cu205, CuO, and BaCu02 were observed as residual impurities. Incomplete conversion of the oxide material to the superconducting orthorhombic phase during cool-down and annealing is believed to be due to the creation of a stable tetragonal-orthorhombic phase mixture. This stable mixture is not reported by researchers using consolidation temperatures above 900 "C. In summary, the formation of YBa2Cu307 -x from alkoxide sol-gel derived powder consisted of several consolidation reactions involving five main phases, CuO, BaC03, BaC03 polymorph, Y2Cu205, and BaCu02.The superconducting YBa2Cu307 -,pellet sample showed an onset temperature of 65-70 K, as determined using mag- netic susceptibility measurements. Free gel powder processed in a similar way typically yields 80-85 K superconducting transition onsets. The superconducting transitions were broad owing to incomplete phase conversion and the presence of minor impurity phases. The ability to form high-quality fine-grain superconduct- ing powder appears to be hindered by a partial tetragonal YBa2Cu307 -,-to-orthorhombic phase conversion. For our system we do not believ that the lack of orthorhombic phase formation can be solely attributed to the hindered oxygen intercalation and ordering effects, which are a consequence of paired oxygen in the tetragonal phase.Future work is being performed to follow quantitative changes in the YBa,Cu3 phase ratios, variation in lattice parameters, and oxygen content. The authors would like to thank the following people for their constructive advice and suggestions as well as assistance in performing the reported work: Frank Rotella, Mark Teepe, David Kenzer, Julie Twaddle, and Dr. Fulin Zou. Funding for this project was provided by NSF-DMR 86-12860. The work in Argonne was supported by U.S. Dept. of Energy, Div. of Basic Energy Sciences-Materials Sciences, under con- tract W-31-109ENG-38 and by the NSF-funded Science and Technology Center for Superconductivity under Grant No.DMR-88-09854. temperature/ C phase 120 275 550 650 750 800 850 \ Ctr -1 CUO I cup 4 UaCO, I UaCO, PM ___I UaCuO, 4 YCU 20, I Y1Ua2C~307-x ____I Fig. 12 Summary of phase development and consumption during thermal processing of YBa,Cu, gel 180 J. MATER. CHEM., 1991, VOL. 1 References 11 G. Moore, S. Kramer and G. Kordas, Mater. Lett., 1989, 7, 415. 1 M. K. Wu, J. R. Ashburn, A. J. Torng, P. H. Hor, R. L. Meng, 12 J. D. Jorgensen, J. Faber, J. M. Carpernter, R. K. Crawford, J. R. Haumann, R. L. Hitterman, R. Kleb, G. E. Ostrowki, F. J. L. Gao, Z. J. Huand, Y. Q. Wang and C. W. Chu, Phys. Rev. Lett., 1987, 58, 908. 13 Rotaella and T. G. Worltan, J. Appl. Crystallogr., 1989, 22, 321. S. Hirano, T.Hayashi, R. H. Baney, M. Miura, H. Tomonaga, 2 K. E. Easterling, C. C. Sorrell, A. J. Bourdillon, S. X. Dou, G. J. Siogett and J. C. Macfarlane, Mater. Forum, 1988, 11, 30. 14 Chem. Lett., 1988, 665. A. Manthiram, J. S. Winnea, A. T. Sui, H. Steinfink and J. B. 3 H. Koinuma, K. Fukuda, T. Hashimoto and K. Fueki, Jpn. Goodenough, J. Am. Chem. SOC., 1987, 109, 6667. J. Appl. Phys., 1988, 27, L1216. 4 C. T. Cheung and E. Ruckenstein, Mater. Lett., 1988, 7, 172. 5 R. Roy, Science, 1987, 238, 1664. 6 S. Hirano, T. Hayashi, M. Miura and Hiroyuki Tomonaga, Bull. Chem. SOC.Jpn., 1989, 62, 888. 15 16 J. D. Jorgensen, M. A. Beno, D. G. Hinks, L. Soderholm, K. J. Volin, R. L. Hitterman, J. D. Grace, I. K. Schuller, C. U. Segre, K. Zhang and M. S. Kleefisch, Phys. Rev. B, 1987,36, 3608. A. Manthiram and J. B. Goodenough, Nature (London), 1987, 329, 701. 7 H. S. Horowitz, S. J. McLain, A. W. Sleight, J. D. Druliner, P. L. Gai, M. J. VanKavelaar, J. L. Wagner, D. D. Biggs and S. J. Poon, Science, 1989, 243, 66. 8 S. Kramer, K. Wu and G. Kordas, Mater. Res. SOC. Symp. Proc., 17 18 J. B. Goodenough and A. Manthiram, in Chemistry of Oxide Superconductors, ed. C. N. Rao, Blackwell, 1988, pp. 101-1 12. G. Wang, S. J. Hwu, S. N. Song, J. B. Ketterson, L. D. Marks, K. R. Poeppelmeier and T. 0. Mason, Adv. Ceram. Mater, 1987, 1988, 99, 323. 2, 313. 9 S. A. Kramer, G. Kordas, J. McMillan, G. C. Hilton and D. J. VanHarligen, Appl. Phys. Lett., 1989, 53, 156. 10 S. Kramer, G. Moore, G. Kordas, P. A. Keifer and C. T. G. 19 Witanachi, H. S. Kwok, X. W. Wang and D. T. Shaw, Appl. Phys. Lett., 1988, 53, 234. Knight, Mater. Res. SOC. Symp. Proc., 1988, 121, 643. Paper 0/03010D; Received 4th July, 1990