Synthesis, structure and superconductivity in the Ba_1−xK_xBiO_3−y system

Abstract

The recent discovery of superconductivity near 30 K in Ba_1−xK_xBiO_3−y (x≈0.4)^1,2 is remarkable for two reasons. It is the first copper-free oxide superconductor that has a transition temperature (T_c) above that for the best intermetallic superconductor; and the structure is reported to be cubic, which excludes a two-dimensional metal–oxygen sublattice analogous to the CuO₂ planes believed to be responsible for superconductivity in the copper-oxide-based superconductors. Cava et al.¹ described a synthesis technique which involved starting with a 100% excess of KO₂. At least part of the excess potassium was found to be present in the final sample (in a form not detectable by X-ray diffraction), resulting in samples that were not suitable for resistivity measurements and making a precise determination of the potassium and oxygen content in the superconducting phase impossible. Here we describe a two-step synthesis technique starting with a stoichiometric oxide composition, which yields single-phase samples suitable for transport measurements. Neutron powder diffraction studies of samples with varying potassium concentration show that superconductivity in Ba_1−xK_xBiO_3−y, occurs only in a cubic perovskite phase which is stable at ≲600°C and which forms only for x > 0.25. Within this cubic phase, Tc is highest for compositions near the structural phase transition (x ≈ 0.25) and decreases with increasing x.