NEW SUPERCONDUCTING OXIDES IN THE Bi-Sr-Ca-Cu-O SYSTEM : MAGNETIC MEASUREMENTS AND STRUCTURAL DETERMINATION

-Detailed neutron diffraction structure analysis of Bi2SrzCaCu20s is reported. The effects of Pb-substitution on BizSrzCaCuzOs and Bi2Sr2CuOs are discussed, and magnetization measurement results on the 110 K phase in the Bi-Sr-Ca-Cu-0 system are given. Since the discovery of high Tc superconductivity in the Bi-Sr-Ca-Cu-0 system [I], two superconducting phases have been unambiguously identified, namely BizSrzCuOa (Tcw 10 K) and Bi2SrzCaCuzOs (T,x 80 K) , hereafter denoted as 2201 and 2212. The presence of transition temperatures around 110 K in mixed phase samples suggested the existence of a third phase, the composition of which could be 2223, based on an analogy with the later discovered thallium compounds [2]. We report here a detailed structural analysis of the 2212 phase and physical characterizations of the 110 K phase. Three alternative models for the structure of Bi2Sr2CaCuzOs were first proposed on the basis of Xray single crystal diffraction data [3]. Then, Bordet et al. 141 used powder neutron diffraction to show that the double Bi-0 layers do not adopt the usual Aurivilius structure, but are arranged in an NaC1-type structure. The problem with this model, refined in the Fmmm space group, is that the in-plane Bi-0 distances are too long to satisfy the normal ~ i ~ + coordination. In fact, these oxygens appear disordered in this model. We show here a more precise analysis of our neutron data allowing a better determination of the Bi co-ordination. As the space group Amaa or its non-centric subgroup A2aa have been confirmed by convergent beam electron diffraction [5], the neutron data were refined in Amaasy permitting all atoms to move off the y-axis special position of the Bordet et al. structure taken as the starting model. Only the BiO plane oxygen 0 3 (which appeared "disordered" in Fmmm) was much displaced. In Amaa, the best refinement was obtained by splitting 0 3 between two sites either side of the center of the Bi-square. In an a t t e m ~ t o remove the need for split atoms, we simply removed the center of symmetry (space group A2aa). 0 3 was then not constrained in any way, and immediately moved to occupy a position close to a pair of Bi atoms. As these refinements were carried out in the subcell, they cannot take into account the displacements of the Bi atoms related to the existence of the = 4.7 A superstructure along the a-axis. In the resulting structure (see Tab. I), each Bi has three close oxygen neighbours at between 2.12 A and 2.28 A, opposite to the ~ i ~ + "lone pair" . This coordination is more satisfactory than that obtained with Fmmm, and all Bi remains essentially ~ i ~ + , as there is no experimental evidence for ~ i ~ + . The formula is close to BizSrzCaCu~Os The substitution of Bi by Pb in the 2201 and 2212 phases as also been studied. It was found in both cases that as much as 0.4 Pb per unit formula can be incorporated without detection of extra phases by X-ray powder diffraction. Preliminary a.c. susceptibility measurements on powder and single crystal samples of both phases have been performed. They seem Table I. A t o m positions, thermal and occupancy factors, Bi-0 and Cu-0 distances (A) for BiaSrzCaCuzOs.