Dear Lukasz,
First of all, in the majority of Rietveld programs the uniaxial preferred
orientation (such as [100]) can not be handled adequately for cubic systems
since the programs normally generate only one symmetrically equivalent hkl for
a diffraction peak. In DDM, for such cases, I included a special option
(Expanded hkl) to generate all equivalent hkl's for every peak. This option
makes the refinement process slower and also may require some manual settings
for initial parameter values.
Another alternative for cubic systems is spherical harmonics.
As for the estimated error of the refined PO component, it's really wise to
check if it exceeds the refined value, but it should be done at the final stage
only, as at initial stages all errors may be high due to the incompleteness of
refinement.
Best regards,
Leonid
*******************************************************
Leonid A. Solovyov
Institute of Chemistry and Chemical Technology
660036, Akademgorodok 50/24, Krasnoyarsk, Russia
http://sites.google.com/site/solovyovleonid
*******************************************************
On Thursday, February 6, 2014 5:37 PM, Łukasz Kruszewski
<lkruszew...@twarda.pan.pl> wrote:
Dear Madam/Sir(s),
I am performing Rietveld analysis for a sample containing perovskites
(CaTiO3 and SrTiO3). When putting-in a preferred orientation (PO)
correction (100 direction, as the crystals are cubic-like), I obtain
better Rwp and GOF i.e. chi-squared values (9.63 and 1.5, respectively)
than in the case of not using this correction (Rwp = 10.62 and GOF =
1.65). However, the values of the PO corrections are higher than the
errors calculated. All the other parameters including the graphical fit,
background etc., seem to be unchanged. I am also convinced about the
correctness of any input, with my diffractometer calibrated with LaB6 and
Si standards, and Beq / Uiso parameters taken directly from the CIF files
provided, so that the microabsorption factor may likely be rejected (?).
My guess is that putting in the preferred orientation might artificially,
somehow, make the fit better, maybe due to some interdependence?
Could you please kindly suggests if obtaining better statistics (Rwp, GOF)
by getting a correction values below the corresponding errors could be
treated as a good result?
Best regards!
--
Łukasz Kruszewski, Ph.D., adjunct
Polish Academy of Sciences
Institute of Geological Sciences
X-Ray Diffraction Laboratory (coordinator)
Twarda 51/55 str.
00-818 Warsaw
Poland
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Please do NOT attach files to the whole list <alan.he...@neutronoptics.com>
Send commands to <lists...@ill.fr> eg: HELP as the subject with no body text
The Rietveld_L list archive is on http://www.mail-archive.com/rietveld_l@ill.fr/
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Please do NOT attach files to the whole list <alan.he...@neutronoptics.com>
Send commands to <lists...@ill.fr> eg: HELP as the subject with no body text
The Rietveld_L list archive is on http://www.mail-archive.com/rietveld_l@ill.fr/
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
