PDFGUI newbie question

[a.m.t...@dl.ac.uk]

Hello folks

I am currently struggling up the learning curve with PDFGUI. I have a
question about which parameters should be refined and in which order
do you refine them?

I am an experienced Rietveld user, the lattice parameters,
coordinates, displacement factors and occupancies are obvious.
However, what do I do with the delta1, delta2, srat, rcut, Qsigma and
Qalpha parameters?

How many of these parameters need to be refined? Are they refined
before, after or at the same time as the atomic and lattice
parameters?

many thanks

Tony Bell, Daresbury Laboratory

[Chris Farrow]

Hi Tony,

Thanks for posting. You can find some information about the variables
here: http://www.diffpy.org/doc/pdfgui/pdfgui.html#SEC37

I'll expand upon this information a bit and hopefully answer your
question adequately.

> I am an experienced Rietveld user, the lattice parameters,
> coordinates, displacement factors and occupancies are obvious.
> However, what do I do with the delta1, delta2, srat, rcut, Qsigma and
> Qalpha parameters?

delta1 and delta2 account for correlated vibrations among atoms in the
structure. The actual peak shape in the PDF is related to the
displacement factors and these parameters. delta1 and delta2 relate to
the Debye temperature of the material in some complicated way. As a
rule of thumb, delta1 is refined for high-temperature (above the Debye
temperature) studies and delta2 is refined otherwise.

srat and rcut also account for correlations. Use these when refining
delta1 or delta2 fails to give a good fit. srat is "sigma ratio". The
PDF peak-width is multiplied by srat for distances below rcut.

Qsigma and Qalpha account for experimental resolution. These are now
called qdamp and qbroad, respectively. Either one or both are used.

> How many of these parameters need to be refined? Are they refined
> before, after or at the same time as the atomic and lattice
> parameters?

Usually the correlation terms are refined at the same time as the
atomic and lattice parameters. However, is not uncommon to refine the
atomic and lattice parameters first, and then turn on the correlation
parameters and refine it all again. You may have to play with it a bit
to figure out what works for your material and model.

Qsigma and Qalpha can be refined similarly to the correlation
parameters. However, these are dependent on the experimental setup. It
is common to measure a standard material under the same conditions as
your material of interest and then measure Qsigma and Qalpha with the
standard. In this case, they can be specified and fixed.

I hope you find this information useful. Please let us know if you
have any other questions.

Sincerely,

Chris Farrow
Michigan State University

[a.m.t...@dl.ac.uk]

Dear Chris

Thanks for the reply, I am making a bit more progress now after
looking at the web link which you gave. I was able to get a much
better fit with a tutorial example and now I am trying PDFGUI on some
"real" data.

tony....

[Chris Farrow]

Glad to hear it. I hope the program is useful for you.

Chris