The complex susceptibility of a sintered Y-Ba-Cu-O superconductor is strongly dependent on a.c. field amplitude, h. Very small values of h must be used for the real part of susceptibility, Z', to reach a value corresponding to bulk diamagnetism just below the critical temperature, T c. The imaginary part, X", represents hysteresis loss in the sample. Thus, Z" versus temperature becomes positive when h exceeds the lower critical field, Hc~, of the superconductor.
Annealing the material in oxygen gives rise to two distinct components, a relatively high-Tc, high-H d superconductor (denoted as "G" or 'good') and a relatively Iow-T c, Iow-Hcj superconductor (denoted as "B" or 'bad'). Curves of susceptibility versus increasing temperature reflect the dual nature of the annealed sample: X' has an inflection point at T c of the B component and approaches zero at T¢ of the G component, while Z" has a peak at each T c. Both critical temperatures decrease linearly with increasing h, though at very different rates. He= of the G component is considerably greater than Hc~ of the B component. The lower critical fields are linearly decreasing functions of temperature.
Two models might explain the susceptibility data. In the grain model, the G component consists of superconducting grains and the B component is either intergranular material, unfavourably orientated anisotropic grains, or oxygen-depleted grain boundaries. In the surface model, the G component is in the interior of the sample and the B component is at the sample's surface. This condition could arise if there was oxygen depletion at the surface subsequent to total enrichment during annealing.