Journal of Crystal Growth 135 (1994) 496—504 o~o~ CRYSTAL North-Holland GROWTH Crystallization process in Pb-free or Pb-doped Bi 2 ~Pb~Sr2Ca2Cu3O10 ~ glass system R. Cloots a S. Stassen a A. Rulmont a P.A. Godelaine 13 P. Diko c,1 P.H. Duvigneaud d and M. Ausloos e,2 SUPRAS, Institute of Chemistry, B6, University of Liege, Sart Tilman, B-4000 Liege, Belgium bSUPRAS, Montefiore Electricity Institute, B28, University of Liege, Sart Tilman, B-4000 Liege, Belgium Service de Physique des Solides, Free University of Brussels, CP233, Boulevard Triomphe, B-I 050 Brussels, Belgium d Service de Chimie Analytique et Industrielle, Free University of Brussels, CP165, F. Roosevelt Avenue, B-I 050 Brussels, Belgium SUPRAS, Institute of Physics, B5, University of Liege, Sart Tilman, B-4000 Liege, Belgium Received 3 April 1993; manuscript received in final form 20 September 1993 Bi2 ~Pb~Sr2Ca2Cu3Oj0 ~ synthesis (for x = 0.0 and x = 0.3) has been made starting from a glass precursor. After splat quenching, the samples have been annealed at different temperatures for various time spans. X-ray diffraction data and other standard measurements show various growth mechanisms. The 2223 phase results from a dissolution—reprecipitation process rather than from direct recrystallization. A 2201 phase precipitates at high temperature and reacts rapidly with the melt to form the 2223 phase. The kinetics of the lead-substituted sample seems to be controlled by the Ca2PbO4 liquid phase. Alumina contaminated phases have been put into evidence and result from the reaction with the container. Heterogeneous nucleation of the 2223 phase at their surface seems also to exist, as previously reported in other systems. 1. Introduction mediary compound Ca2PbO4 gives rise to an increased yield of the high-To phase [4—71. Since High critical current (~~) superconductors a partial melting is observed at 850°C, the molten (HJCS) based on Bi-compounds and calcium have part of Ca2PbO4 can be expected to react easily a general formula Bi2Sr2Ca~1Cu~O2~÷4 with n with CuO and the 2212 (or 2201) phase in order = 1, 2, 3. The n = 2 and n = 3 phases are charac- to produce the 2223 phase [8,91. terized by a bulk superconductivity at T~ = 85 K A second mechanism might be that rather than and T~ = 110 K, respectively [1,21. acting as a flux in the growth process, lead modi- It is difficult to prepare a homogeneous single fies the strength of the bonding between the phase material with the highest T~ = 110 K. By layers constituting the building blocks of the partially substituting lead for bismuth, it is possi- structure, and interferes with (and even impedes) ble to increase the proportion of the 110 K (2223) the multiphase intergrowth process, up to en- phase [4]. Much discussion exists about the role hancing the stability domain of the 2223 phase. of the lead ions ‘in the formation of the 2223 A glassy state precursor might be basically very phase. It seems that the formation of the inter- appropriate for the formation of the 2223 phase [10,11], since it is a more homogeneous initial system. However, it is very important also to note Permanent address: Institute of Experimental Physics Sb- . that the recrystallization process leading to the yak Academy of Sciences, Solovjevova 47, 043 53 Kosice, Sbovak Republic. 2223 phase is actually not yet well understood 2 Author for correspondence. [12,13]. 0022-0248/94/$07.00 © 1994 — Elsevier Science B.V. All rights reserved SSDI 0022-0248(93)E0461-F