Susceptibility and EPMA Characterisation

The results of the compositional analysis are presented in Table 5.1. The different starting compositions of the growths can be seen to be reflected to a fair degree in each of the samples. Relative to the ideal 2212 stoichiometry the A1 sample comes closest, while both the B1 and Oxford samples have increased Bi contents and reduced Sr contents. Measurements made of several different areas of each crystal showed some variation in the composition, this is very difficult to quantify without intensive measurements, and the variations did not appear significant enough to significantly effect the relative differences between the crystals indicated in the table. The values for oxygen content $\delta$ determined by this method are a good indication of the relative differences in oxygen content between the samples but are not as accurate as the data for the cations, and the absolute values must be taken with a degree of caution. It can be seen that, as was expected, the excess Bi content of the B2 sample combined with growth in a flowing oxygen atmosphere has resulted in a significantly larger oxygen content than either of the other two samples. As an indication of the absolute oxygen contents, 8+$\delta$, the numbers are accurate at best to only 5-10$\%$. The values for T$_c$ determined from the susceptibility results are also included in Table 5.1. The higher oxygen content of B2 can be seen to correlate with a considerably lower value of T$_c$, T$_c$=82K, than in the other two samples with T$_c$=94K for A2 and T$_c$=93K for Oxford. In comparison with the values quoted earlier in the review, the oxygen difference here of $\Delta \delta=0.09$ for a $\Delta T_c=10K$ is in good agreement with those determined by TGA [125,140,135] and iodometric titration [142].

Table 5.1: The results of EPMA determination of the cation stoichiometry and oxygen content of the three crystals. The starting composition was that of the growth, and T$_c$ was determined by AC-susceptibility.

	Starting	Measured composition	oxygen	$T_c$
	composition	Bi-Sr-Ca-Cu	$\delta$
Oxford	2.3-2.1-1.0-2.05	2.05-1.92-0.85-2.00	0.06	93K
A2	2.0-2.0-1.0-2.0	1.970-1.810-0.915-2.00	0.013	94K
B2	2.2-1.64-1.16-2.0	2.057-1.603-1.09-2.00	0.104	82K

Figure 5.2: AC-susceptibility vs. temperature for crystal A1, showing the effect of oxygen annealing upon the superconducting transition.
Figure 5.3: AC-susceptibility vs. temperature for crystal B1, showing the effect of nitrogen annealing upon the superconducting transition.

The results of the AC-susceptibility measurements of the A1 sample are presented in Figure 5.2, both in the as-grown state and after the oxygen annealing. The T$_c$ was initially 93K reducing to 88K after annealing. The results for B1 are presented in Figure 5.3; it can be seen that the nitrogen annealing has had the reverse effect to the oxygen annealing of A1, raising the T$_c$ of 82K in the as-grown state to close to 86K after annealing. These measurements established that the transitions for both A1 and B1 in their as-grown states were very similar to those of their respective partners A2 and B2 from the same growths, and so it will be assumed that the samples A1 and B1 before annealing have closely similar compositions and oxygen contents to those determined by EPMA. Whilst the transition width remained essentially unaffected by the annealing in the case of A1, it has degraded considerably in the B1 sample. The low-temperature diamagnetic signal also showed a marked change in both cases. The width of a transition as measured by susceptibility is principally associated with the superconducting homogeneity of a sample. The value of the remaining diamagnetism at temperatures well below the transition (i.e. the amplitude of the shielding effect) is an indication of the superconducting volume fraction of a sample. The observed decrease could be attributable to some surface decomposition of the samples which is commonly observed to accompany exposures to temperatures above 500$^o$C [115], thus increasing the non-superconducting volume. It is worth noting that these changes are not intrinsically linked to the alteration of $T_c$, as is shown by the example of identical variations in $T_c$ observed without degradation of the superconducting properties by Mitzi [140].