Study of intermetallic compounds isostructural to MgB/sub 2/

Abstract

The small intergrain effect of MgB/sub 2/ on supercurrent makes it one of the most promising candidates for superconducting conductors due to its easier processing and the associated lower manufacturing cost. Unfortunately, the superconducting transition temperature T/sub c/ is only 40 K. However, band-structure calculations predict that a higher T/sub c/ than that of MgB/sub 2/ is possible for isostructural and isovalent intermetallic compounds with greater lattice parameters or greater unit cell volumes. The prediction appears to be consistent with the negative pressure effect on T/sub c/ observed. The substitution of the larger Ca-ions for the smaller Mg-ions has thus been suggested to raise T/sub c/, but not yet realized. Alternatively, we have synthesized and studied a series of binary and pseudobinary intermetallic compounds, AGa/sub 2/, AGa/sub 2-x/Si/sub x/, and AAl/sub 2-x/Si/sub x/, where A=Ca, Sr, or Ba, which are isostructural to MgB/sub 2/ and have greater lattice parameters than MgB/sub 2/. In spite of the greater lattice parameters, AGa/sub 2/ are not superconducting. However, all pseudobinary compounds AGa/sub 2-x/Si/sub x/ and AAl/sub 2-x/Si/sub x/ for 0.6<x<1.5 are superconducting. Superconductivity is reported for the first time in SrAlSi, with T/sub c/ of 5.1 K. Although T/sub c/ varies with x in a similar fashion for all members of the series, no specific correlation between T/sub c/ and lattice parameters or ionic mass is observed. The maximum T/sub c/ of these compound series with different A's varies between 5.5 and 7.8 K, much lower than that of MgB/sub 2/. The results strongly suggest the unique role of B in the superconductivity of C32 intermetallic compounds. They also demonstrate that factors additional to the lattice parameters and densities of states must play an important role, and that the rigid-band model is not sufficient to account for the observations.