X-RAY DIFFRACTION IMAGING TECHNIQUES AS TOOLS FOR INDUSTRY AT BM05
The IRT Nanoelec is a public-private partnership set up in Grenoble to support the local ecosystem on nanoelectronics and coordinated by the CEA-LETI. The ESRF, together with the ILL, the LPCS and the CEA-LETI, run the characterisation programme, which has been established to facilitate industrial access to the use of synchrotron X-rays and neutrons, offering multipurpose instrumentation with high-throughput functionality, particularly for standardised and statistical characterisations.
IRT Nanoelec activities include outreach to and collaboration with industry, demonstration of specific user cases and the setup of novel, multipurpose instrumentation specially tailored to industry needs. A six-axis diffractometer has been installed at beamline BM05, allowing characterisation of wafers up to 300 mm by diffraction and reflectivity. This instrument also supports an ambitious programme of X-ray topography, which is highly demanded by the nano- and microelectronics industries. These techniques are offered to industry through the Platform for Advanced Characterisation of Grenoble (PAC-G), a commercial initiative launched by IRT Nanoelec.
Apart from these IRT activities, recent X-ray diffraction imaging experiments at BM05 include work by Sen et al., who studied sapphire crystals grown by the Kyropoulos process to be used for optical purposes. Some crystals exhibit a milky zone that leads to light diffusion that is detrimental for such applications. White beam Bragg diffraction imaging investigations showed that the milky zone surprisingly exhibits a remarkably low dislocation density compared to the rest of the disk. To get further clues about the milky zone, rocking curve imaging maps were
recorded at the transition between the milky and non-milky regions.
Figure 144 gives a representative result of these maps. Figure 144 shows a sharp increase in the observed full width at half maximum (FWHM), which is associated with the dislocation density, when progressing into the non-milky zone. In addition, the faint variation in peak position between the two zones (~1 arcsec) indicates a lattice parameter variation of the order 10-5, which suggests a weak variation of composition. These results led to further experiments showing that there is a variation in the number of oxygen vacancies between milky and non-milky zones. Indeed, the reduced number of dislocations in the milky zone leads to a reduction in the diffusion of these vacancies, and therefore to an increase in their concentration. When this concentration exceeds the solubility limit, they bind together and form nanovoids. The higher vacancy concentration is probably the origin of the 10-5 variation of lattice parameter between the two zones. This implies that, from an industrial point of view, it is not necessary to have a more perfect crystal but, on the contrary, to slightly increase the dislocation density to help eliminate the oxygen vacancy nanovoids.
Fig. 144: (000) reflection in the transition zone (small black square) below the milky defect
(delimitated by horizontal lines). The figures show the projection mapping of FWHM (top) and
variation of peak position (bottom).