Susceptibility maximum and Fermi-liquid effect in 4d and 5d transition metals

Abstract

The calculated susceptibility either shows a maximum arising from the Fermi-liquid effect or obeys the T² ln(T/T*) law as a function of temperature T. Because of this, the density-of-states curve derived from the usual single-particle band model is shown to be erroneous. The value of T* is found to range from about 100K to above 10000K; the highest value yet obtained is 11300K for iridium. If it were assumed that metals did not melt, the susceptibility of Ir, W etc. would exhibit a maximum at temperatures as high as the effective Fermi temperature; for example for Ir T_max=T*/ square root e=6854K. The conditions for the occurrence of such a large Fermi-liquid effect is discussed. It is shown that, with certain assumptions, the experimental data support the predictions of the Fermi-liquid model as to the amount by which the coefficient of the T² ln T term is enhanced as the system approaches a ferromagnetic instability.