delta2 in PDFGui for more than one phase

[acuest...@gmail.com]

Dear all,

I am starting to work with PDFGui and I have a question. Here it is:

When I have more than one phase (in PDFGui), in my case, I have two main phases and two impurities. How should I refine the delta2 parameters, only for the main phases? independently in each phase? or should I use a constrain or something to do this?

Best regards,

Ana.

[Pavol Juhas]

On Wed, Sep 07, 2016 at 06:55:30AM -0700, acuest...@gmail.com wrote:
...

> When I have more than one phase (in PDFGui), in my case, I have two main
> phases and two impurities. How should I refine the delta2 parameters, only
> for the main phases? independently in each phase? or should I use a
> constrain or something to do this?

Hi Ana,

The delta2 parameter simulates sharpening of the low-r peaks because
of correlated atom motion. At higher r the atoms forming the pair
move independently and so the PDF peak width depends on the sum of
their displacement parameters (Uij-s). In principle, the delta2
parameters should be independent, because each phase may have
a different bond strength and thus different sharpening ratios.
However, if refined phases have very similar chemistry and similar
local structure, it would make sense to use the same delta2 value.
As for the impurity phases, chances are their contributions are much
smaller and may not have enough resolution to capture the sharpening.
I would recommend to first fit them with a fixed delta2=0 (no
sharpening) and then check if refining their delta2 produces any clear
improvement in the fit quality. Note that delta2 should be refined
from an initial value of delta2~=2 (about the first peak position),
because fitting from delta2=0 may be unstable.

Hope this helps,

Pavol

--
Dr. Pavol Juhas
Computational Science Initiative
Brookhaven National Laboratory
P.O. Box 5000
Upton, NY 11973-5000