The search for superconductivity in rare earth nickel(iii) and nickel (i) oxides

We have undertaken a systematic study of the Ln2xSrxNiO4-δ system over the entire lanthanide range, focusing on doping levels above x 1.0. It was observed that for samples prepared under flowing O2 at 1100 deg C, the upper solid solution limit was x = 1.67. The lower solubility limit increased with decreasing ionic radii for Ln = Nd3+ - Gd3+, forming essentially point phases Ln033Sr1.67NiO3.67 (LnSr5Ni3O11) for Ln = Y3+, Dy3+ - Tm3+. Thermo- gravimetric analysis of these phases indicated that they contained nickel in a 3+ oxidation state. Rietveld refinement of the structure of these materials using powder X-ray diffraction data showed that each adopted the tetragonal K2NiF4 structure(I4/mmm), with Ln3+ and Sr2+ disordered over the K sites and oxygen vacancies occuring predominantly in the NiO2 planes. Electrical conductivity measurements of YSr5Ni3O11 indicate semiconducting behaviour which is able to be modelled by variable range hopping. Magnetic susceptibility measurements between 6 K and 300 K show Curie-Weiss behaviour, with substantially lower effective moments than expected from spin only low-spin Ni(III) due to the delocalisation of the nickel electrons. Partial thermogravimetric reduction of LnSr5Ni3O11 (Ln = Y, Dy - Tm) led to a series of nickel(I) oxides: LnSr5Ni3O8. Rietveld refinement of the structure of these materials using powder X-ray diffraction data indicated a transition from a tetragonal K2NiF4 to an orthorhombic Sr2CuO3 (Immm) arrangement, in which l/3 of the bridging oxygen atoms were missing from the chains of apex-linked nickel oxide square planes. Bond valence calculations supported the notion of nickel being present as Ni+