Logo PTB

X-ray Optics

Working Group 4.25

X-ray Topography

Monochromatic X-ray beams are, within a very narrow angle range, almost fully reflected on crystal surfaces. On both sides of the reflection field, the intensity of the scattered beam decreases steeply. This effect can be exploited for the high-precision characterisation of lattice distortions close to the surface. Different X-ray equipment is available to the Working Group to make crystal structure errors visible, e.g. an arrangement of two crystals of which one - the so-called "forecrystal" - solely serves to reduce the divergence of the X-ray beam. The second crystal is the test sample whose distortions are to be investigated. It is adjusted to the flank centre of the intensity distribution of a Bragg reflection. With this adjustment, local deviations in the lattice plane spacing close to the surface of the test crystal lead to a modification of the optical density on a photo plate which is proportional to the lattice distortion and serves to detect scattered radiation. Lattice distortions of the size of a billionth of a lattice spacing can be detected.

If one wants to investigate particularly large crystal surfaces within the shortest possible time - as can be the case, e.g., in serial examinations - the double-crystal arrangement is, however, not a suitable method. The adjustment is very time-consuming and the imaged surfaces are relatively small. Thus, in our Working Group, we resort to a special single-crystal method, the so-called "Berg-Barrett topography". Significant errors, such as twins and dislocations, can be reliably detected by means of this method. The optimal case for a Berg-Barrett image is the following: a slightly diverging X-ray beam hits the crystal surface at a flat grazing angle. The adjusted, asymmetric lattice planes in the crystal lattice diffract the beam in such a way that it comes nearly vertically out of the crystal surface. A detector material placed as close as possible to the front of the sample then converts the X-ray intensity into optically readable contrast patterns. This creates a practically undistorted, large-scale image. Stronger intensity thereby reveals errors. If, e.g., an X-ray film is used, errors such as dislocations become apparent as black lines on a grey background.

The pre-condition for this procedure is that strongly asymmetric Bragg reflections can be detected in the test sample. For this purpose, a double-circle diffractometer has been modified in such a way that selected asymmetric reflections can be detected fast and reliably. The test sample is rotated on a motor while the grazing angle is traversed. The diffracted intensity is then detected via a photomultiplier and analysed by means of a multichannel analyser which is synchronised to the sample's rotation.