Combine small and wide-angle x-ray scattering data

[Mạnh Tuấn Nguyễn]

Hi Simon and everyone,

I am studying graft-type polymer electrolyte membranes for hydrogen fuel cell.

I am using xPDFsuite for getting PDF and I have some questions.

1. How can I understand the behavior of PDF if I use both small and wide-angle x-ray scattering data? The PDF damps quickly if I use Qmin = 0.05 Å^-1 and Qmax = 7.1 Å^-1.

2. Can I have good information about the sizes or disordered behavior of crystalline/hard phase in material by using sasPDF? In fact, high-energy X-ray sources are not always available.

3. Is it possible if I use xPDFsuite to get PDF from ultra-small, small, and wide-angle scattering data (Q-range from 0.0004 Å^-1 to 7.1 Å^-1). We got ultra-small angle data from Super Photon ring-8 GeV (SPring-8), Japan.

Best regards,

Manh Tuan.

[Ian Billinge]

Hi Manh,

I can have a try at answering your questions.

1. Generally, you want to use the smallest Qmin and the highest Qmax that your data allows. If I am interpreting your email correctly, you have a dataset that includes very small angle scattering, so there is no reason to set Qmin so high. You are only discarding information by doing that. The conventional advice is to look at your I(Q) data at very low Q. At the lowest Q, you will see a flat line, where the beamstop is obscuring the detector. As you go to higher Q, you reach the edge of the beamstop and start getting real data. You want to set your Qmin somewhere near the edge of the beamstop.

2. I don't have experience using sasPDF, so I can't precisely answer, but sasPDF is designed for fitting small angle data, so my answer is yes. 

If sasPDF ends up not working for you, other small angle scattering fitting programs exist and can give information about the size and disorder of different phases (so long as there is x-ray scattering contrast between the phases!). sasView is a fairly user-friendly one that you could consider if you can't make sasPDF work.

3. It should be possible to use xPDFsuite to get a PDF that includes ultra-small angle scattering data. There is certainly no problem including ultra small angle scattering data in a PDF in principle. I don't know if xPDFsuite imposes a minimum value on the Qmin -- my PDF datasets are all at much higher Q.

Finally, I should add that your Qmax is low for a conventional PDF study. Most PDF beamlines collect out to Qmax = 25 inverse Å or higher. Having such a low Qmax means you will have limited resolution in r. Just something to be aware of. If you are hoping to use your data to probe bonds between near-neighbor atoms or anything like that, you might need to be cautious. However, if you are just hoping to develop a pair distribution function between different phases or aggregates in your membrane (ie, over a longer length scale), you should be fine.

I hope this helps, and the community should jump in to correct me if I have anything wrong.

Ian

--
You received this message because you are subscribed to the Google Groups "diffpy-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diffpy-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/diffpy-users/41f5198d-5b75-4d2f-8773-ed00be91801cn%40googlegroups.com.

--

Ian Billinge

he/him/his

Ph.D Student, Yip Lab

Dept. of Earth & Environmental Engineering

Fu Foundation School of Engineering

Columbia University in the City of New York

[Mạnh Tuấn Nguyễn]

Dear Ian and everyone

Thank you for your reply.

About the first question, when I used Qmin in the region of ultra-small or small-angle data, the PDF damps quickly ( the black line in my picture). In this case, I think we can't get good information. So my question is "Why PDF damps quickly if I used Qmin too small?"

I hope to receive more replies from the community.

Please help me.

Again, thank you very much, Ian!

Best regards,

Manh Tuan.

Vào Th 5, 2 thg 11, 2023 vào lúc 19:15 Ian Billinge <ihb...@columbia.edu> đã viết:

To view this discussion on the web visit https://groups.google.com/d/msgid/diffpy-users/CAODy25ou%2BCOhuXN5cpGpgYbRHS%3D4hKS2MWr2-fPXsv%3DGVFfUYQ%40mail.gmail.com.

[Mikkel Juelsholt]

Hi Manh Tuan 

The short answer to question 3 is : 

XPDFsuite is a GUI for PDFgetX3, which is not suitable for small-angle scattering data. 

You should use PDFgetS3 which is developed to small-angle scattering: 

https://arxiv.org/pdf/1910.08186.pdf 

https://www.diffpy.org/products/pdfgetx.html

The first question you should ask yourself is what you want to know from the PDF. It is important to remember that there is no more information in the PDF compared to Q-space data. It is the same information but represented in a different way.

If you just want to know the size of different objects in your sample, then I would just stick to analysing the small-angle scattering. Diffpy does have a package to do that, but you can also use a dedicated SAS fitting program. And in that case, you do not need to measure scattering data to more than 1-2 Å-1. And this can easily be done on an in-house SAXS instrument. Where small-angle scattering PDF can be useful if you have packing/ordering of your nanosized objects. Then, understanding and analysing the packing/ordering can be much easier from small-angle scattering PDF.

If you also want to know about the atomic structure, you have some different options. You can do normal x-ray powder diffraction analysis with your scattering data up to 7 Å-1 with no problem. However, if your material is nanostructured or disordered, you might need an atomic PDF. And as Ian points out, you need a much higher Q-range to get a proper atomic PDF. 

Remember that the resolution in the PDF is approximately Pi/Qmax. So you would only get a resolution of 0.44 Å, which is way too low to see anything useful. 

On a Mo or Ag lab-source, you can get above 15 Å-1, which is usually good enough for simple PDF analysis. Otherwise, many PDF beamlines have a mail-in program. 

So the answer to your second question is: It depends. No need for a PDF if you just want the size. But in other cases, PDF is very useful.

On a more broad basis, it is also important to question if you need to do a PDF over the entire Q-range. Remember that at very low Q-values, there is no atomic information because all the atoms scatter in phase. So, for atoms, you do not need data at a Q-value smaller than around 0.1 Å-1 because there is little information here. At the same time, there is no information about the particle-particle interactions at high Q-values. 

So, if I were you, I would make an analysis of the small angle scattering, 0.0004 to around 1 Å-1 and an analysis for the higher Q-values. 

For your first question, I am not sure. I suspect that you are confusing XPDFsuite because you probably have a major sloping data signal at low Q. This probably confuses the ad hoc algorithm for correcting the I(Q) and just forces the entire PDF to be 0 at low r. 

But it should not matter to you because you should not use XPDFsuite for this region anyway. You should use PDFgetS3.

Cheers Mikkel

[Mạnh Tuấn Nguyễn]

Dear  Mikkel

Thank you for your useful information.

I really appreciate it!

Manh Tuan.

Vào lúc 00:06:43 UTC+7 ngày Thứ Sáu, 3 tháng 11, 2023, Mikkel Juelsholt đã viết:

[Simon Billinge]

Thanks for the comments Ian and Mikkel.

PDFgetS3 will take small angle x-ray scattering (SAXS) data and turn it into a PDF, but it is intended for the particular case where you are interested in the structural ordering of large objects like nanoparticles and this information is contained as a correlation function in the SAS signal.   

Regular PDF from WAXS data shows how atoms are arranged, but if you are arranging nanoparticles there will be a similar looking signal in the SAXS region.  We can then use PDFgetS3 to get the PDF and can model the structures using PDFgui.  This is described here:

https://journals.iucr.org/j/issues/2020/03/00/kc5104/

As Ian and Mikkel say, if you are interested in the size and shape of objects in your sample, but not how they arrange in space, you would probably want to analyze the SAS data separately using conventional SAS tools (e.g. in the sasview program).

There are some examples of combined SAXS/PDF analyses (e.g., our own http://journals.iucr.org/j/issues/2014/02/00/issconts.html), but mostly PDF/diffraction and SAS data can be analyzed separately.

s

To view this discussion on the web visit https://groups.google.com/d/msgid/diffpy-users/33133f0c-49a0-468b-89cd-7835b7290aa1n%40googlegroups.com.

--

Simon Billinge

Professor, 

Department of Applied Physics and Applied Mathematics

Columbia University

[Mạnh Tuấn Nguyễn]

Dear Mr Simon Billinge, Ian and Mikkel.

Thanks for your help!

Your information is very helpful for my study.

I really appreciate it!

Manh Tuan.

Vào Th 6, 3 thg 11, 2023 vào lúc 19:20 Simon Billinge <sb2...@columbia.edu> đã viết:

To view this discussion on the web visit https://groups.google.com/d/msgid/diffpy-users/CANugsUEmwCwBtekHN7Fj12AoMe10zfYYMz9cg2yQnkMwsDFHzA%40mail.gmail.com.