Logo Texture

Course Material Index  Section Index  Previous Page  Next Page 

Texture

Samples are often described as having preferred orientation when actually they are textured and vice versa: some samples, especially worked metals, for example, may exhibit both. Whatever the language used, it is important to recognise the symptons in the powder diffraction data that are due to:

The first case is easily discerned when the sample is measured in more than one diffraction geometry and is referred to as sample preferred orientation (as discussed previously). The second case tends to result in random data and is often referred to as sample texture. It is commonly caused by crystallites that are too large, and therefore insufficiently numerous, to provide a true powder average. In this case sieving may provide a solution by removing the largest crystallites, though not all samples are amenable to this technique. Metal samples in particular can be a problem here.

The effect of texture is demonstrated below using an "off the shelf" sample of potassium chromate, K2CrO4, measured with Cu Kα1 on a Bragg-Brentano diffractometer. The first diffraction data shown in green below is from a stationary (non-spinning) sample that failed to pass through a 250 µm sieve. The diffraction pattern is generally weak with fewer peaks than expected. Though not easy to see in these plots, a common symptom of texture is the presence of peaks that are narrower than the instrumental resolution function.

The second diffraction pattern is from a non-spinning 38 to 100 µm sieved fraction of potassium chromate. More diffraction peaks are now visible, but the relative intensity of the peaks is poor compared to the expected values from a non-textured sample.
Spinning the same sample does give some improvement to the powder pattern, but the presence of the large crystallites still dominates the diffraction data shown in orange below:
The final pattern, shown in red below, is from a spinning sample of K2CrO4 that was ground and passed through a 38 µm sieve. The peaks now have closer relative intensities to the expected values; and a Rietveld fit (taught later in the course) of the crystal structure to the data will be the best way to demonstrate this.


Course Material Index  Section Index  Previous Page  Next Page 
© Copyright 1997-2006.  Birkbeck College, University of London.
 
Author(s): Jeremy Karl Cockcroft
Martin Vickers