Hi Vineetha,

I advised during the course that students should use the ifeffit mailing list for questions post-school.
An advantage of doing so is that one could receive assistance from any member of the community
and not be at the mercy of the time constraints of any particular researcher. Another advantage is that
someone more familiar with the type of problem you are studying might be able to offer more insight.
I would like again to encourage you to make use of the mailing list. Other researchers may also see
the problem with which you are dealing and it could help them too.

Some general advice:

Mixed phase samples can be problematic. If you know which two phases are present, you can simulate them
separately and examine the local environments. If the near-neighbour environments are very similar,
then some fitting can be done with reasonable results. You still need to keep in mind the resolution
criteria for being able to resolve closely spaced paths. Looking at your O-scattering paths, not having seen
your k-space data and range, I suspect you will not be able to resolve the splitting in the first shell,
and so, would treat it as a single path (similar argument for Hf paths in 2nd shell).

The starting model serves to give you amplitudes and phases for fitting. In that regard, whether you use monoclinic
or orthorhombic starting model should not matter. Once fit, you would examine how well the fit result
compares to either model. I would hope you also measured standard(s) of known composition and structure.
Fitting such would help guide the fitting of your unknown.

Model calculations can be done for your two suspected phases, including temperature effects
with the DEBYE card in the FEFF input. A quick google search yielded this paper:
https://ceramics.onlinelibrary.wiley.com/doi/abs/10.1111/j.1151-2916.1977.tb14088.x
which gives a calculated Debye temp of 484K for monoclinic HfO2. You can check the
literature further for other research on this material which may give an experimental value.
With the model chi(k)'s, you can do linear combinations (manually) and see if you find some ratio
that approaches agreement with your data. I don't see how to get the model chi(k) out of Artemis though.
I even tried running the feff6 executable manually (with associated dll's) but it did not generate the chi.dat
file I was expecting. I would suggest, if you wish to generate models and compare in this fashion, that you
invest in a newer version of FEFF that does allow you to generate this output (i.e. get FEFF9).

Other thoughts:

XANES? Did you measure it for the unknown and standards? If so, are the two phases distinguishable?

Background removal? Can you improve it to reduce the amplitude below 1 Angstrom?

Always show the original data (merged ok), not simply the transforms when asking for help.
That helps one assess the problem better.

regards,
Robert



On 2019-02-19 9:43 a.m., Mukundan, Vineetha wrote:

Hi Dr. Gordon,

 

I am Vineetha and had met you at the EXAFS short course @ BNL in November’18. I am trying to analyze a couple of my data on Hafnia and am stuck. So I am seeking help with the following questions. Apology in advance if it will some of your time.

 

1.    I generated the scattering paths. See attached. It is safe to discard the multiple scattering paths here. So I generated the plots with Hf-O single scattering paths and Hf-Hf scattering paths. Clearly it shows the first shell is mostly one or all of O1.1, O2.1 and O2.3 paths. And the second shell is mostly influenced by Hf1.2, Hf1.3 and Hf1.5. Am I correct in making this assumptions so far? IS there any other way to figure out which paths are important or not important, other than plotting them together with the data.

 

2.    While fitting, I am really unsure how to set the parameters for each path in my example. I am not sure how to set up the parameter when we have multiple paths contributing to say first shell or second shell. Also, I know using the Nyquist equation it is possible to figure out the number of independent variables. I read up the ARTEMIS pages on FeS2 example described but there the structure is cubic but my sample is monoclinic. Here I am trying to fit only the monoclinic paths with my data.

 

3.    Also, I know my sample has monoclinic and orthorhombic structure. So how does one go about generating the simulated EXAFS for a particular structure on FEFF? say in my case the monoclinic and orthorhombic ones. I have the .cif files for these 2. For figuring out the % amount of monoclinic and orthorhombic phases in the sample, I am assuming I will have to do a linear combination analysis. Am I correct?

 

Any help in this regard will be really useful.

 

Thank you,

Vineetha

 

 

Vineetha Mukundan, PhD

Post-doc, Prof. Diebold’s Group

Colleges of Nanoscale Science and Engineering

SUNY Polytechnic Institute

257, Fuller Road

Albany, NY 12203 USA

 

CONFIDENTIALITY NOTICE

The pages (including attachments) comprising this internet or facsimile transmission contain CONFIDENTIAL information.  This information is intended for the sole and exclusive use by the individual entity named as the recipient hereof and any further disclosure or dissemination is prohibited.  If you are not the intended recipient, you are hereby notified that any disclosure, copying, storage, distribution, or use of the contents of this transmission is strictly prohibited and may be unlawful.  If you have received this transmission in error, please notify us immediately by telephone or return e-mail, delete it from your mail system, and destroy all copies in your possession.  Thank you.