[RMCProfile-users] [EXTERNAL] Fitting data from SAED

Zhang, Yuanpeng zhangy3 at ornl.gov
Wed Jun 30 16:25:41 BST 2021


Concerning `w.r.t. 2`, there is a routine in RMCProfile which implements (by Martin Dove) those coefficients so that normally we won’t need this external .xray file. However, we still keep the capability to let users have their own coefficients just in case needed.

Cheers,
Yuanpeng

--

Dr. Yuanpeng Zhang
Neutron Scattering Scientist
Spallation Neutron Source

Oak Ridge National Laboratory
1 bethel valley road, Oak Ridge, TN 37830, US
Mobile: +1(865)-335-4818
Email: zhangy3 at ornl.gov<mailto:zhangy3 at ornl.gov> | Web: https://www.iris2020.net<https://www.iris2020.net/>

From: Stenz, Christian <stenz at physik.rwth-aachen.de>
Date: Wednesday, June 30, 2021 at 11:21 AM
To: David Keen - STFC UKRI <david.keen at stfc.ac.uk>, rmcprofile-users at shadow.nd.rl.ac.uk <rmcprofile-users at shadow.nd.rl.ac.uk>, Zhang, Yuanpeng <zhangy3 at ornl.gov>
Cc: rmcprofile-users at rmcprofile.org <rmcprofile-users at rmcprofile.org>
Subject: AW: [RMCProfile-users] [EXTERNAL] Fitting data from SAED
Dear Dave and Yuanpeng,

@Dave: w.r.t. 3. Yes, the "/ (<f(Q)>^2)" in my equation is the reason why I have to use the flag NORMALIZATION_TYPE :: <f(Q)>^2.

@Yuanpeng: Okay, I understand the backend calculations now.
It would be very advantageous to be able to input f(q) files for electron atomic form factors! Please let me know, if the procedure with adding additional columns in the F(q) file is possible. Otherwise, your help to implement it is very much appreciated.


Many thanks again to both of you, I appreciate this lively discussion!

Best wishes,
Chris



-----Ursprüngliche Nachricht-----
Von: David Keen - STFC UKRI <david.keen at stfc.ac.uk>
Gesendet: Mittwoch, 30. Juni 2021 16:01
An: Stenz, Christian <stenz at physik.rwth-aachen.de>; rmcprofile-users at shadow.nd.rl.ac.uk; Zhang, Yuanpeng <zhangy3 at ornl.gov>
Cc: rmcprofile-users at rmcprofile.org
Betreff: Re: [RMCProfile-users] [EXTERNAL] Fitting data from SAED

Dear Chris

w.r.t. 1. - yes this is a typo that I then cut and pasted into three typos(!) - f_A(Q) and f(Q)_A should be the same.  F(Q) is a weighted sum of the partial structure factors corresponding to the different atom pairs.  These additional columns give the Q-dependent weightings for each partial structure factor.
w.r.t. 2. - I'm afraid I don't know - this is a question for Yuanpeng to answer w.r.t. 3. - Ideally your 'N' should be 1 but I know this is difficult to achieve; F_n(Q) = (N*I(Q) - <f(Q)^2>) / (<f(Q)>^2) might be more usual but this is probably semantics as the key is to get a good PDF from the final function and the scaling factor(s) are often hidden in the software packages.  The "/ (<f(Q)>^2)" in your equation is why this is used in the weighting terms listed in my previous message.

Best wishes
Dave


--------------------------------------------------------------------------
From: Yuanpeng

Dear Chris and Dave,

I totally agree with Dave – after we normalize experimental F(Q) by, say, <f(Q)>^2, we are not getting rid of the Q-dependent weights. That means in RMCProfile calculation/fitting, we cannot treat X-ray or electron diffraction data as what we do for neutron – the fundamental difference here is the combination coefficients for combining partials (either partial PDF or partial F(Q)) are constant for neutron and non-constant for X-ray/electron. To deal with this, RMCProfile now tries to ‘reproduce’ the experimental procedure for X-ray calculation – calculate partial PDF for each pair -> Fourier transform each of them separately into Q-space -> multiply each partial F(Q) with corresponding Q-dependent coefficients -> normalize by either <f(Q)>^2 or <f(Q)^2>, depending on which was used for experimental data normalization -> combine them all and Fourier transform back into real-space to obtain the PDF.

I hope this will help in one way or another.

Concerning the manual way of putting in the Q-dependent weightings, I can definitely check whether we have this capability in RMCProfile currently. If not, and if necessary, I can definitely help implementing it. Please feel free to let me know if you want to go that way @Chris.

Cheers,
Yuanpeng

--

Dr. Yuanpeng Zhang
Neutron Scattering Scientist
Spallation Neutron Source

Oak Ridge National Laboratory
1 bethel valley road, Oak Ridge, TN 37830, US
Mobile: +1(865)-335-4818
Email: zhangy3 at ornl.gov | Web: https://www.iris2020.net



On 30/06/2021, 15:25, "Stenz, Christian" <stenz at physik.rwth-aachen.de> wrote:

    Dear Dave,

    Thank you for your input. There are three things that I want to comment:

    1. Did you mean to denote the same function with f_A(Q) and f(Q)_A in your equation? How do these additional columns contribute to the further calculations then?

    2. I managed to run a simulation with XRAY_RECIPROCAL_SPACE_DATA and my supposed normalized F(Q) without specifying a .xray file. What weightings are used in that case?

    3. To be clear, I obtain my normalized F(Q), say F_n(Q), as follows:

         F_n(Q) = (I(Q) - N*<f(Q)^2>) / (N*<f(Q)>^2)

         where I(Q) is the electron diffracted intensity,
         f(Q) denotes the electron atomic form factors and
         N is a constant scaling factor.


    Thanks in advance!

    Best wishes,
    Chris



    -----Ursprüngliche Nachricht-----
    Von: David Keen - STFC UKRI <david.keen at stfc.ac.uk>
    Gesendet: Mittwoch, 30. Juni 2021 14:35
    An: rmcprofile-users at shadow.nd.rl.ac.uk; Zhang, Yuanpeng <zhangy3 at ornl.gov>
    Cc: Stenz, Christian <stenz at physik.rwth-aachen.de>; rmcprofile-users at rmcprofile.org
    Betreff: Re: [RMCProfile-users] [EXTERNAL] Fitting data from SAED

    Dear Chris

    Even though the F(Q) is 'normalised' you still need to include the Q-dependent weightings for each partial F(Q) that makes up the total F(Q).  This is what the .xray file allows you to calculate (the normalised F(Q) is only normalised in the sense that the function is unitless; it doesn't mean that the functions that make up the normalised F(Q) are not weighted by terms with physical units).  So I'm not convinced that your idea of making the function look like an x-ray function would work.

    It is also possible to calculate the Q-dependent weighting terms 'by hand' if the .xray method doesn't work for electron diffraction.  You simply add additional columns to the experimental normalised F(Q) which correspond to (for example for a sample of composition AB, with data normalised by <f(Q)>^2):

    Q  F(Q)  c_A*c_A*f(Q)_A*f_A(Q)/<f(Q)>^2  2*c_A*c_B*f(Q)_A*f_B(Q)/<f(Q)>^2  c_B*c_B*f(Q)_B*f_B(Q)/<f(Q)>^2

    Where c_A and c_B would both be 0.5 in this case and the f(Q) would be correct for electron diffraction calculated in whatever way is appropriate.  Yuanpeng can confirm if this form of input file still works OK in the most recent versions of RMCProfile.

    Best wishes
    Dave

    On 30/06/2021, 09:22, "rmcprofile-users on behalf of Stenz, Christian via rmcprofile-users" <rmcprofile-users-bounces at shadow.nd.rl.ac.uk on behalf of rmcprofile-users at shadow.nd.rl.ac.uk> wrote:

        Hi Yuanpeng,

        Okay it's good to know of that parametrisation, thanks. It would be no problem for me to transform my normalised F(q) with the electron form factors <f_e(q)>^2 and <f_e(q)^2>. But is the input file F(q) for RMCProfile not supposed to be renormalized with the corresponding x-ray scattering form factors?

        So, my idea would be to normalise my electron diffraction data to a function without units, which should then correspond to 'F(q) normalized' referring to the RMCProfile manual, and then renormalise with respect to the x-ray form factors to obtain 'F(q)', again referring to the RMCProfile manual.

        Cheers,
        Chris


        Von: Zhang, Yuanpeng <zhangy3 at ornl.gov>
        Gesendet: Dienstag, 29. Juni 2021 16:20
        An: Stenz, Christian <stenz at physik.rwth-aachen.de>; rmcprofile-users at shadow.nd.rl.ac.uk
        Cc: rmcprofile-users at rmcprofile.org
        Betreff: Re: [EXTERNAL] [RMCProfile-users] Fitting data from SAED

        Hi Chris,

        Those are the coefficients for the analytical X-ray atomic form factor equation, as you can find in the following paper, https://doi.org/10.1107/S0108767394013292

        In Eqn. (1), you can find the expression with 9 coefficients. Moreover, since you have electrons instead of X-ray, I think you may need to do some transformation to get the coefficients for electrons - I am not familiar with this, unfortunately.

        Cheers,
        Yuanpeng

        --

        Dr. Yuanpeng Zhang
        Neutron Scattering Scientist
        Spallation Neutron Source

        Oak Ridge National Laboratory
        1 bethel valley road, Oak Ridge, TN 37830, US
        Mobile: +1(865)-335-4818
        Email: zhangy3 at ornl.gov<mailto:zhangy3 at ornl.gov> | Web: https://www.iris2020.net<https://www.iris2020.net/<https://www.iris2020.net%3chttps:/www.iris2020.net/>>

        From: Stenz, Christian <stenz at physik.rwth-aachen.de<mailto:stenz at physik.rwth-aachen.de>>
        Date: Tuesday, June 29, 2021 at 11:08 AM
        To: Zhang, Yuanpeng <zhangy3 at ornl.gov<mailto:zhangy3 at ornl.gov>>, rmcprofile-users at shadow.nd.rl.ac.uk<mailto:rmcprofile-users at shadow.nd.rl.ac.uk> <rmcprofile-users at shadow.nd.rl.ac.uk<mailto:rmcprofile-users at shadow.nd.rl.ac.uk>>
        Cc: rmcprofile-users at rmcprofile.org<mailto:rmcprofile-users at rmcprofile.org> <rmcprofile-users at rmcprofile.org<mailto:rmcprofile-users at rmcprofile.org>>
        Subject: AW: [EXTERNAL] [RMCProfile-users] Fitting data from SAED
        Hi Yuanpeng,

        Firstly, thanks for the advice.

        I have some issues preparing the .xray file by myself as I don't know how it is set up. If I understood it correctly, the .xray file describes the x-ray atomic form factor with nine parameters per element each.
        What do the nine numbers specified behind each element in the .xray file exactly refer to? I can't seem to find a documentation about this, neither in the manual nor the tutorial.


        Cheers,
        Chris


        Von: Yuanpeng Zhang <zhangy3 at ornl.gov<mailto:zhangy3 at ornl.gov>>
        Gesendet: Dienstag, 29. Juni 2021 15:00
        An: rmcprofile-users at shadow.nd.rl.ac.uk<mailto:rmcprofile-users at shadow.nd.rl.ac.uk>
        Cc: rmcprofile-users at rmcprofile.org<mailto:rmcprofile-users at rmcprofile.org>; Stenz, Christian <stenz at physik.rwth-aachen.de<mailto:stenz at physik.rwth-aachen.de>>
        Betreff: Re: [EXTERNAL] [RMCProfile-users] Fitting data from SAED

        Hi Christian,

        There are several things on top of my head,


          1.  You may want to manually prepare the .xray file to put in the atomic form factors for electron diffraction.

          1.  You can use `XRAY_RECIPROCAL_SPACE_DATA` in this case, given that you transform the input data to allowed format. You can refer to RMCProfile manual for the definition of various formats and I don't think it will be difficult to transform in between those formats. But please feel free to let me know if you have issues with the transformation.

          1.  Also, you may want to read the section of RMCProfile manual about `XRAY_RECIPROCAL_SPACE_DATA` where you can find that we only need to provide the reciprocal space data and RMCProfile can do the Fourier transformation internally and do the fitting in real space automatically if specified to do that. Again, if any issues, please feel free to let me know.

        Cheers,
        Yuanpeng

        --

        Dr. Yuanpeng Zhang
        Neutron Scattering Scientist
        Spallation Neutron Source

        Oak Ridge National Laboratory
        1 Bethel Valley Rd. Oak Ridge, TN 37830, US
        Mobile: +1(865)335-4818
        Email: zhangy3 at ornl.gov<https://link.getmailspring.com/link/36835A59-6E0F-489A-994E-5C74EF8FCD20@getmailspring.com/0?redirect=mailto%3Azhangy3%40ornl.gov&recipient=c3RlbnpAcGh5c2lrLnJ3dGgtYWFjaGVuLmRl> | Web: https://www.iris2020.net/<https://link.getmailspring.com/link/36835A59-6E0F-489A-994E-5C74EF8FCD20@getmailspring.com/1?redirect=https%3A%2F%2Fwww.iris2020.net%2F&recipient=c3RlbnpAcGh5c2lrLnJ3dGgtYWFjaGVuLmRl<https://www.iris2020.net/%3chttps:/link.getmailspring.com/link/36835A59-6E0F-489A-994E-5C74EF8FCD20@getmailspring.com/1?redirect=https%3A%2F%2Fwww.iris2020.net%2F&recipient=c3RlbnpAcGh5c2lrLnJ3dGgtYWFjaGVuLmRl>>
        On Jun 17 2021, at 6:40 am, Stenz, Christian via rmcprofile-users <rmcprofile-users at shadow.nd.rl.ac.uk<mailto:rmcprofile-users at shadow.nd.rl.ac.uk>> wrote:
        Hi all,


        I would like to fit my real space and reciprocal space data obtained from TEM selective area electron diffraction (SAED) on amorphous GeTe samples. I don't know much about the backend computations regarding the x-ray/neutron atomic form factors in RMCProfile (of course I had to utilise the electron atomic form factors to calculate F(Q) and D(r)), but should there be any issues when supplying the normalised functions in both real and reciprocal space? I.e.:

        XRAY_REAL_SPACE_DATA :: 1

        > FILENAME :: GeTe_asd.DrN

        > DATA_TYPE :: D(r) normalised

        > FIT_TYPE :: D(r) normalised

        > START_POINT :: 1

        > END_POINT :: 1000

        > CONSTANT_OFFSET :: 0.0000

        > WEIGHT :: 0.0050

        > NO_FITTED_SCALE


        NEUTRON_RECIPROCAL_SPACE_DATA :: 1 %% here I declare NEUTRON data instead of XRAY data since one cannot choose the normalised data type in XRAY_RECIPROCAL_SPACE_DATA

        > FILENAME :: GeTe_asd.FqN

        > DATA_TYPE :: F(Q) normalised

        > FIT_TYPE :: F(Q) normalised

        > START_POINT :: 1

        > END_POINT :: 1341

        > CONSTANT_OFFSET :: 0.0000

        > WEIGHT :: 0.0050

        > NO_FITTED_OFFSET

        > NO_FITTED_SCALE

        > CONVOLVE ::


        Thank you in advance for any kind of feedback!


        Best regards,

        Christian

        -------------- next part --------------
        An HTML attachment was scrubbed...
        URL: <https://lists.rmcprofile.org/pipermail/rmcprofile-users/attachments/20210617/da0ed529/attachment.htm>
        _______________________________________________
        rmcprofile-users mailing list
        rmcprofile-users at rmcprofile.org<mailto:rmcprofile-users at rmcprofile.org>
        http://lists.rmcprofile.org/mailman/listinfo/rmcprofile-users
        [Das Bild wurde vom Absender entfernt. Sent from Mailspring]
        -------------- next part --------------
        An HTML attachment was scrubbed...
        URL: <https://lists.rmcprofile.org/pipermail/rmcprofile-users/attachments/20210629/32b9a606/attachment.htm>
        -------------- next part --------------
        A non-text attachment was scrubbed...
        Name: image001.jpg
        Type: image/jpeg
        Size: 824 bytes
        Desc: image001.jpg
        URL: <https://lists.rmcprofile.org/pipermail/rmcprofile-users/attachments/20210629/32b9a606/attachment.jpg>
        _______________________________________________
        rmcprofile-users mailing list
        rmcprofile-users at rmcprofile.org
        http://lists.rmcprofile.org/mailman/listinfo/rmcprofile-users


    This email and any attachments are intended solely for the use of the named recipients. If you are not the intended recipient you must not use, disclose, copy or distribute this email or any of its attachments and should notify the sender immediately and delete this email from your system. UK Research and Innovation (UKRI) has taken every reasonable precaution to minimise risk of this email or any attachments containing viruses or malware but the recipient should carry out its own virus and malware checks before opening the attachments. UKRI does not accept any liability for any losses or damages which the recipient may sustain due to presence of any viruses. Opinions, conclusions or other information in this message and attachments that are not related directly to UKRI business are solely those of the author and do not represent the views of UKRI.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.rmcprofile.org/pipermail/rmcprofile-users/attachments/20210630/07383895/attachment.htm>


More information about the rmcprofile-users mailing list