Читать книгу Encyclopedia of Glass Science, Technology, History, and Culture - Группа авторов - Страница 217

As applied to glasses and glass ceramics, the aforementioned techniques give detailed insights into microstructures and even nanostructures. However, some difficulties remain to be faced. Like electron microscopy or scanning probe microscopy, imaging techniques usually require some sort of (often demanding) sample preparation. If preparation artifacts are avoided, one ends up with a wealth of information about morphology, chemistry, and crystallography for a sample volume of only a few 100 nm³. As stated in the Introduction, it is, therefore, highly desirable either to complement such detailed investigation down to the sub‐nanometer scale with more integrating techniques or to find ways to assess larger volumes, even at the risk of losing spatial resolution.

In this respect, X‐ray microscopy (XRM) has emerged as a new kind of imaging technique that is both nondestructive and capable of offering a three‐dimensional (3‐D) description of the sample bulk. With a maximum 3‐D spatial resolution of 50 nm, it cannot compete with TEM. But sample preparation (for example, with laser micromachining) is rather straightforward, so XRM gives quick answers in the form of a 3‐D density mapping of a reasonably large piece of specimen that may represent the sample bulk.

The technique is based on shadow casting of X‐rays stemming from a point source and going through the sample volume to end up onto a CCD camera. It is similar to X‐ray computed tomography, from which it differs by the fact that the X‐rays are focused within the sample, rather than the sample being illuminated as a whole, and that the signal is further magnified optically on the screen. Because this magnification is made from the central part of the sample, image aberrations are minimized, and enhanced X‐ray optics (derived from synchrotron beamlines) help to achieve an unprecedented resolution. But as with X‐ray computed tomography, the acquisition of series of images for a rotating sample, followed by image processing of the resulting datasets, can yield 3‐D pictures.

Figure 11 Three‐dimensional (3‐D) representation of X‐ray microscopy data from a truncated cone that was laser micromachined from the MAS glass ceramic. (a) Entire dataset with denser volumes in brighter contrast (green in the false color version) and lighter volumes in darker contrast (blue in the false color version). (b) 3‐D visualization of just the high‐density regions of the sample, taken from a clipped internal volume of (a) to make segmentation easier. Volume fractions: dense inclusions: 23%, lower‐density material: 77% (see electronic version for color figures).

For the MAS sample repeatedly considered here, such a 3‐D picture clearly shows how the high‐ and low‐density components are interconnected (Figure 11), the former and latter likely being the Y‐bearing glass and the spinel, respectively. Owing to a lack of elemental and spatial resolution, however, yttria in the residual glass and the zirconia needles cannot be discerned.