Dear all,
My name is Magnus H. Sorby. I'm a Norwegian inorganic solid state chemist.
In my on-going Ph.D. study, I am doing crystal structure determination of
metal hydrides using powder diffraction (neutron and
synchrotron/conventional x-ray).
The synchrotron diffractogram of one of my samples (measured at the
Swiss-Norwegian Beamline at ESRF, Grenoble) has a peak at Q = 1.30 �-1 (or
2theta = 8,29 deg; this is the 110 peak; there are no peaks at lower Qs)
with a realtive intensity of approx. 6% relative to the strongest peak.
However, the same peak is barely visible with conventional PXD
(Bragg-Brentano geometry). I'm pretty sure that this is not due to
prefered orientation as all peaks at higher Q resemble each other closely
in the two measurements. So my guess is that the difference has to do with
polarization.
Doing Rietveld refinement with GSAS, I get close to zero calculated
intensity in the 110 peak. The rest of the pattern fits nicely. I have
tried to set POLA to diffent values between 0 and 1 (usually I lock it at
0.98), and I have also tried the three possible settings of IPOL. None of
this has any significant influence on the calculated intensity of 110.
If I insert my refined structure model into DBWS, I immediately get
sensible intensity in 110. The same happened when a colleague tried it in
Fullprof. The calculated structure factor for the peak is similar in GSAS
and DBWS.
Does anybody have a idea of what I can be doing wrong? Or can it be a
problem with GSAS? I'm currently using the Linux version, but I have also
tried older UNIX versions.
Best wishes
Magnus H. Sorby
(who is sorry if this message was posted twice)
