12.2 Sample displacement parameters
After a good fit is obtained for the unit cell constants, noted by the tick marks now
well aligned to the reflection positions, this should be optimized by refinement of an
appropriate sample parameter, but do not do this until the reflection positions are
well fit.
- For Bragg-Brentano Diffraction: for all types of samples refine the “Sample
displacement” value. Note that the units for this correction are in microns
(\(1 \times 10^{-6} m\)). For samples composed of low-Z materials, you should also refine the
sample transparency correction (see §5.2.)
- For Debye-Scherrer X-ray Diffraction: one should always refine the
“Sample X displ. perp. to beam” (sample displacement perpendicular to
beam) value.
- For Debye-Scherrer Neutron Diffraction: one should always refine the
“Sample X displ. perp. to beam” (sample displacement perpendicular to
beam) value. If data are collected well above \(120^\circ \ 2\theta \) one can also refine the
“Sample Y displ. —— to beam” (sample displacement parallel to beam)
in addition. In theory, this second parameter could be refined with x-rays
as well, but it is so rare to have much data observed at very high angles.
- For TOF diffraction, the sample placement can affect the effective distance
of the sample from the source. Refining the DIFC Instrument Parameter
can adjust for this, but it is usually not necessary.
Note that with older software, one would vary the \(2\theta \) or TOF zero correction. This
correction is available in GSAS-II but does not properly apply the correct shift for a
misplaced sample and should not normally be used.
There is a special sample displacement parameter, which is found as an HAP
parameter (see §14.7), rather than in the sample or instrumental data tree entries.
This is called the “Layer displacement parameter,” which is a value in \(\mu \)m. This
would be used in a material where the sample components are not mixed
homogeneously, and thus the phases are not all at the same location with
respect to the sample surface. This means that the sample displacement is
dependent on the histogram and on the phase. An example of this would be
a layered sample, where each phase will have a slightly different sample
displacement.