Characterization with (00l) X-Rays

The most common method for staging determination in GIC's is (00l) x-ray diffraction.[67] This technique was also applied to the KHg- and CsBi-GIC's, although more for determination of the phase (alpha and/or beta) than for identification of stage. Mixed-stage KHg or CsBi specimens were never at any time produced in the course of this study, although a given stage might contain the admixture of several different phases. The lack of stage admixture is not surprising for C₄KHg, which is formed from the intercalation of K and Hg vapor into C₈K,[82,70] but is somewhat surprising for the stage I C₄CsBi_0.5 and C₄CsBi_1.0, which are formed from higher stage CsBi-GIC's.[18]

The presence of the peaks of only one phase in a (00l) diffraction was the most common occurrence. However, gold C₄KHg samples sometimes contained the I_c = 10.83 Å beta minority phase as well as the I_c = 10.24 Å alpha majority phase. The x-rays scans of many gold samples contained only the majority-phase peaks, though. Kim et al. note that their experiments indicate that use of a small temperature difference DeltaT increases the amount of beta-phase. This finding which is consistent with Table 3.2, which shows that gold samples are typically intercalated with a small DeltaT. Predominantly beta-phase specimens are never produced.[147] Occasionally C₈K ( I_c = 5.35 Å) was also observed in the diffraction scans of gold-colored stage I. Samples which contained C₈K were never used in further experiments, though.

The occasional presence of C₈K in gold but not pink samples is intriguing in light of the hypotheses presented in Section 3.2 about the synthesis of the different types of C₄KHg. One of the ideas put forward in Section 3.2 to explain the occurrence of gold samples was that they were the product of an incomplete intercalation reaction. As mentioned above, C₄KHg is formed from the intercalation of K and Hg vapor into C₈K, which forms first.[249,70] If the gold samples are the result of an incomplete reaction, one would expect to see C₈K peaks in gold specimens, but not in the completely reacted pink specimens. As noted above, this is exactly what is observed.

The (00l) scans shown here were taken using molybdenum K_alpha radiation with a wavelength of 0.708 Å. Molybdenum radiation was chosen because it passes relatively unattenuated through the glass tubes in which the samples were encapsulated. The radiation from the Mo anode passed through a 1.0° horizontal collimation slit on its way to the GIC, and a ``High-Resolution'' Soller slit and 0.1° vertical collimation slit on its way to the detector, which was of the extrinsic silicon type. The diffractometer used is part of the Center for Materials Science X-Ray Facility at MIT, and is of standard design.

Typical (00l) scans for C₄KHg pink ( T_c = 1.53 K) and gold ( T_c = 0.95 K) specimens are shown in Figure 3.3. These plots show intensity versus the angle 2theta between the incident and diffracted beams. The diffractograms for both the pink and gold samples show peaks from the copper sample holders used for the low-temperature measurements (discussed in Section 4.3) near 2theta = 22° and broad humps from the glass tube that surrounds the sample from about 6° to about 14°. These scans are quite similar to those shown by Timp[249] and Lagrange et al.[149]. The magnitude of I_c was estimated from using Bragg's Law using the separation between the most intense pair of peaks, the (002) and (004). The average value of I_c obtained in these measurements, 10.20 Å , is in accord with the (10.12 ± 0.03) Å found by Timp[249], the 10.24 Å found by Univ. of Kentucky researchers,[270] and the 10.16 Å reported by Lagrange.[150] However, Lagrange shows the intensity of the (004) peak higher than the intensity of the (002), contrary to Ref. [249] and Figure 3.3. This reversal of the usual intensity ratios was seen for some of the samples used in this study, but it appeared that the effect was due to improper centering of the specimen on the goniometer.

Figure 3.3: (00l) x-ray scans for pink and gold C₄KHg. The large peak near 22° in each scan is from the copper sample holder. The broad hump from about 6° to 14° is due to the glass tube that the sample holder is in. a) I_c = (10.22 ± 0.03) Å pink sample. T_c = 1.53 K. b) I_c = (10.18 ± 0.03) Å gold sample. T_c = 0.95 K.

(00l) x-ray diffraction was also performed before and after hydrogenation on both pink and gold specimens. Hydrogenation appeared not to affect I_c, as Figure shows. The scan after hydrogenation has a larger (004) peak than (002), while the scan before hydrogenation has a larger (002) peak. As noted before, this reversal appears to be an alignment-dependent effect. Other diffraction data taken after hydrogenation data on this sample show the (002) peak higher. There does not seem to be any reproducible difference between the (00l) patterns before and after hydrogenation.

Figure 3.4: (00l) x-ray scans before and after hydrogenation for a gold C₄KHg sample. The broad hump from about 6° to 14° is due to the glass tube that the sample is in. Before: T_c = 0.84 K gold C₄KHg sample. I_c = (10.24 ± 0.03) Å. After: Same sample as in ``Before'' picture, only after exposure to 200 torr hydrogen gas. T_c = 1.535 K. I_c = (10.24 ± 0.03) Å.

(00l) scans for two CsBi-GIC's are shown in Figure 3.5. a) shows the x-rays of a C₄CsBi_0.5 alpha-phase specimen with I_c = 10.61 Å, and b) shows a similar scan taken by Bendriss-Rerhrhaye.[17] These scans look different because the spectrum of Bendriss is taken with theta increasing from right to left, and the MIT spectrum is taken with 2theta increasing from left to right. alpha + beta-phase specimens were also synthesized which had an additional set of peaks corresponding to I_c = 11.43 Å. The MIT values for I_c compare favorably with the (10.61 ± 0.02) Å for alpha-phase and (11.48 ± 0.02) Å for beta-phase reported by McRae et al.[167] There is not any obvious difference between the two sets of (00l)'s despite the fact that T_c = 4.05 K is reported for the University of Nancy samples and T_c < 0.5 K for the MIT specimens. (There are more details about the superconductivity of CsBi-GIC's in Chapter 6.)

Figure 3.5: (00l) x-ray scans for alpha-phase stage 1 CsBi-GIC's. The broad hump from about 6° to 14° is due to the glass tube that the sample holder is in. a) I_c = (10.61 ± 0.03) Å alpha-phase sample. b) Similar (00l) scan taken by Bendriss-Rerhrhaye.[17] In b), theta is increasing from right to left, whereas in a) 2theta is increasing from left to right.

In both the case of the KHg-GIC's and the CsBi-GIC's, the (00l) scans show essentially no difference between low- T_c and high- T_c samples. In an attempt to find some structural information that would correlate with T_c, Raman scattering experiments were performed. The results of these experiments are described in the next section.