- Ifeffit - millenia.cars.aps.anl.gov

weak scattering from second-nearest neighbors for single atom catalysts
by Qiang 08 Feb '24

08 Feb '24

Hi everyone, I would like to have a discussion about the common EXAFS for single atom catalysts. It is usually observed that there is only one strong peak around 1.5A in EXAFS, and it is attributed to the nearest neighbors of O or N. Take an example of Pt1/CeO2 catalyst. Pt single atom is coordinated with a few oxygens, i.e., PtOx. These oxygens constitute the first shell of Pt. It is supposed that these O should also attach to the metal element of the support (Ce here). So there is also Pt-O-Ce structure. Why it doesnot have strong scattering peak in EXAFS? Is it that because the Pt-O-Ce formed is not ordered, or it is actually not a bond just some weak contact? Thank you very much. Qiang

3 2

Larch version up problem
by 정은석 07 Feb '24

07 Feb '24

Hello~ I used Larch 0.9.71. until now. However, I would like to update larch and I reinstall anaconda and larch. I installed Larch 0.9.74. as following installation guide I didn't have any problems during the installation process. However, I faced problems during feffit fitting. Before, I used feffit fitting in larch 0.9.71., and then I hadn't had any problems. I only changed larch as new version. What am I doing? When I run feffit fitting, I have debugs as the below: ------------------------------------------------------------------------------------------------------------------------- File "c:\Users\Administrator\Desktop\RL\feffit_error_test\test-feffit.py", line 51, in <module> out = feffit(pars, dset) ^^^^^^^^^^^^^^^^^^ File "C:\ProgramData\anaconda3\envs\xraylarch\Lib\site-packages\larch\xafs\feffit.py", line 645, in feffit ds.prepare_fit(params=params, other_hashkeys=dset_hashkeys) File "C:\ProgramData\anaconda3\envs\xraylarch\Lib\site-packages\larch\xafs\feffit.py", line 374, in prepare_fit self._generate_hashkey(other_hashkeys=other_hashkeys) File "C:\ProgramData\anaconda3\envs\xraylarch\Lib\site-packages\larch\xafs\feffit.py", line 291, in _generate_hashkey dat = [self.data.ek0, self.data.e0, self.data.rbkg] ^^^^^^^^^^^^^ AttributeError: 'Group' object has no attribute 'ek0' ---------------------------------------------------------------------------------------------------------------------- -- Best regards, Ph. D. Eun-Suk Jeong X-ray absorption fine structure(EXAFS+XANES) Mobile:+82-10-4628-9896

2 8

Larch version up problem in Python, Window 10
by 정은석 03 Feb '24

03 Feb '24

Hello~ I have used Larch 0.9.71. until now. However, I would like to update larch and I reinstall anaconda and larch. I installed Larch 0.9.74. as following installation guide I didn't have any problems during the installation process. However, I faced problems during feffit fitting. Before, I used feffit fitting in larch 0.9.71., and then I hadn't had any problems. I changed larch as new version. What am I doing? When I run feffit fitting in python, I have debugs as the below: ------------------------------------------------------------------------------------------------------------------------- File "c:\Users\Administrator\Desktop\RL\feffit_error_test\test-feffit.py", line 51, in <module> out = feffit(pars, dset) ^^^^^^^^^^^^^^^^^^ File "C:\ProgramData\anaconda3\envs\xraylarch\Lib\site-packages\larch\xafs\feffit.py", line 645, in feffit ds.prepare_fit(params=params, other_hashkeys=dset_hashkeys) File "C:\ProgramData\anaconda3\envs\xraylarch\Lib\site-packages\larch\xafs\feffit.py", line 374, in prepare_fit self._generate_hashkey(other_hashkeys=other_hashkeys) File "C:\ProgramData\anaconda3\envs\xraylarch\Lib\site-packages\larch\xafs\feffit.py", line 291, in _generate_hashkey dat = [self.data.ek0, self.data.e0, self.data.rbkg] ^^^^^^^^^^^^^ AttributeError: 'Group' object has no attribute 'ek0' ---------------------------------------------------------------------------------------------------------------------- The data(EXAFS(k)), which are inputted for feffit, have two column, k and chi. As you know, it doesn't have ek0. -- Best regards, Ph. D. Eun-Suk Jeong X-ray absorption fine structure(EXAFS+XANES) Mobile:+82-10-4628-9896

1 0

Query: abnormal FT-EXAFS peak at ~1.1A
by Huaming 30 Jan '24

30 Jan '24

Hello! I faced an issue with fitting of FT-EXAFS data (Rh@zeolite) with Artemis. There is one distinct peak located at ~1.1 angstron. The intensity is comparable to the nearby peak of Rh-O scattering. I wonder the peak (~1.1A) is a real peak or just noise. Raw data is enclosed. Another question is why S02 can be negative in data fitting with Artemis. Cheers, Huaming

1 0

Query: Demeter
by Robbie Taylor 24 Jan '24

24 Jan '24

Good Afternoon, Hope you are well! Please could you advise when Demeter was last updated? And if it is still in active development / regularly bug fixed? Best wishes, Robbie Taylor (he/him) | Software Officer - Vendor Management | IT & Digital | Manchester Metropolitan University 8th Floor | Bridgewater House | Whitworth Street | Manchester, M1 6LT | United Kingdom |[cid:image001.jpg@01DA4A1A.9404F950] [cid:image002.jpg@01DA4A1A.9404F950] Before acting on this email or opening any attachments you should read the Manchester Metropolitan University's email disclaimer available on its website at mmu.ac.uk/emaildisclaimer<http://www.mmu.ac.uk/emaildisclaimer> "Before acting on this email or opening any attachments you should read the Manchester Metropolitan University email disclaimer available on its website http://www.mmu.ac.uk/emaildisclaimer "

3 2

bug in Athena
by Alexander Zherebker 24 Jan '24

24 Jan '24

Hello! I faced an issue with 0.9.26 Athena. Doing LCF the option for weights doesn't really work so I keep getting negative components. Cheers, Alex

2 1

Question about mixed second coordination shell
by Otal Eugenio 18 Jan '24

18 Jan '24

Dear all, Happy New Year! I am dealing with the fitting of NaMn0.3Fe0.4Ni0.3O2 second shell. I have Mn, Ni, and Fe at 2.9772Å and Na at 3.1859Å, showing only one shell. If I use any of the transition metals (dZ <5) for fitting the second coordination shell. fitting is fine, but when I introduce the Na path, the fitting has no sense, CN_Na grows and also its error. Even including Na, I have Nvary = 10 and Ninp = 13.98, so I am getting out of ideas and nothing helpful in the ifeffit Archive until now. Is there any strategy for dealing with mixed second coordination shells? I am attaching the file exported from Larch and the cif file. Best regards. (^ㅇᆽㅇ^)(=˃ᆺ˂=)(=🝦 ༝ 🝦=) Eugenio H. OTAL Assistant Professor Dept. of Materials Chemistry Shinshu University 4-17-1 Wakasato, Nagano 380-8553, JAPAN eugenio_otal(a)shinshu-u.ac.jp https://sites.google.com/view/zettsu-laboratory/news-updates ¯\_(ツ)_/¯¯\_(ツ)_/¯ Who is John Galt? ¯\_(ツ)_/¯¯\_(ツ)_/¯

4 7

Inquiry About Demeter's Capabilities in Simulating Spectra from Structures
by 徐微 17 Jan '24

17 Jan '24

Dear IFEFFIT Team, I hope this message finds you well. I am writing to inquire about the capabilities of the Demeter software suite, specifically regarding its ability to simulate spectra from given structural data. I am particularly interested in understanding whether Demeter can be used to generate X-ray absorption spectra based on detailed structural information. This functionality would be extremely valuable for my work. Could you please provide information on whether Demeter includes this feature? Additionally, if there are any relevant guidelines or documentation on how to use Demeter for this purpose, I would greatly appreciate if you could point me in the right direction. Thank you for your time and assistance. I look forward to your response. Best regards, XUwei

3 2

Re: [Ifeffit] Ifeffit Digest, Vol 250, Issue 5
by Ryan Parmenter 18 Dec '23

18 Dec '23

Hi Matt, Thank you for your detailed response. Given my limited expertise in this area, I'm currently grappling with the intricacies of Python scripts, and I wouldn't typically enter the Larch buffer. I recognise that this is a current limitation on my part and am actively addressing it by attending Python workshops to enhance my understanding soon. When I input print(peakresult.result.fit_report()), the Larch command doesn't yield any values. Additionally, the box at the bottom of the GUI, where the command can be entered, transitions from'>Larch' to '...>' in red text. However, the fit statistic report can be exported/saved through the pre-edge results. Following this, the 'uvars' commands yield the area of the sum of the two Gaussians - thanks! :) Regarding your suggestion for improving the software's functionality, including a "peaks_area" parameter within the fit variables window may be valuable. I am currently comparing the areas among 15 datasets and having a 'peaks_area' parameter would facilitate the straightforward extraction and comparison of these values. Interestingly, another user expressed interest in this feature and reached out to me via email. I may have additional suggestions, but I will note them down during the next round of fits. Best regards, Ryan -----Original Message----- From: Ifeffit <ifeffit-bounces(a)millenia.cars.aps.anl.gov> On Behalf Of ifeffit-request(a)millenia.cars.aps.anl.gov Sent: Friday, December 15, 2023 6:00 PM To: ifeffit(a)millenia.cars.aps.anl.gov Subject: Ifeffit Digest, Vol 250, Issue 5 Send Ifeffit mailing list submissions to ifeffit(a)millenia.cars.aps.anl.gov To subscribe or unsubscribe via the World Wide Web, visit http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit or, via email, send a message with subject or body 'help' to ifeffit-request(a)millenia.cars.aps.anl.gov You can reach the person managing the list at ifeffit-owner(a)millenia.cars.aps.anl.gov When replying, please edit your Subject line so it is more specific than "Re: Contents of Ifeffit digest..." Today's Topics: 1. Re: Larix Pre-edge Statistics (Matt Newville) ---------------------------------------------------------------------- Message: 1 Date: Thu, 14 Dec 2023 20:29:38 -0600 From: Matt Newville <newville(a)cars.uchicago.edu> To: XAFS Analysis using Ifeffit <ifeffit(a)millenia.cars.aps.anl.gov> Subject: Re: [Ifeffit] Larix Pre-edge Statistics Message-ID: <CA+7ESboeUMm6iH9JmEfie9AuajYwDBvtYp8GrtLBSCegdkM=Fg(a)mail.gmail.com> Content-Type: text/plain; charset="UTF-8" Hi Ryan, Yeah, this could be a little richer. For clarity, the "amplitude" Parameters will be the area of each peak. Well, it's the amplitude factor for the unit-normalized lineshape (Gaussian, Lorentzian, Voigt, etc), so it will be the area under that curve from -Inf to Inf. Those Infinities probably have small effects, here anyway ;). As you say, the centroid of multiple peaks is given, but not the total area under the sum of non-background peaks. That would not be too hard to add. But it might also fall into the category of "the sorts of calculations we should allow the user to do after a fit" without having to bake them all in. That is, if you open a Larch buffer after a pre-edge peak fit, you could get the most recent best-fit Parameters from lmfit with: pars = peakresult.result.params But also, you can get "uvars" as uvars = peakresult.result.uvars which is a dictionary of name:"ufloat" values for all Parameters (variables and "constrained values"). These "ufloat" values come from the `uncertainties` package and include uncertainties that know about the correlations between the Parameters. You can use these for simple calculations that will propagate the uncertainties (including correlations). As an example, if I do a pre-edge peak fit, with 2 Gaussian peaks (plus Line+Lorentzian for the background), I might get a good fit and a resulting report (from the larch buffer, do `print(peakresult.result.fit_report()`) that looks like this: [[Model]] (((Model(lorentzian, prefix='bpeak_') + Model(linear, prefix='bpoly_')) + Model(gaussian, prefix='gauss1_')) + Model(gaussian, prefix='gauss2_')) [[Fit Statistics]] # fitting method = leastsq # function evals = 355 # data points = 162 # variables = 11 chi-square = 4.5237e-04 reduced chi-square = 2.9958e-06 Akaike info crit = -2049.75462 Bayesian info crit = -2015.79106 R-squared = 0.99953048 [[Variables]] bpeak_amplitude: 1.82988250 +/- 0.04372160 (2.39%) (init = 2.462209) bpeak_center: 7118.44625 +/- 0.02812082 (0.00%) (init = 7120.091) bpeak_sigma: 1.36103880 +/- 0.01273292 (0.94%) (init = 2.121656) bpeak_fwhm: 2.72207761 +/- 0.02546583 (0.94%) == '2.0000000*bpeak_sigma' bpeak_height: 0.42795967 +/- 0.00701416 (1.64%) == '0.3183099*bpeak_amplitude/max(1e-15, bpeak_sigma)' bpoly_intercept: -6.37404087 +/- 0.65831626 (10.33%) (init = 2.467005) bpoly_slope: 8.9825e-04 +/- 9.2664e-05 (10.32%) (init = -0.000347) gauss1_amplitude: 0.10390560 +/- 0.00363058 (3.49%) (init = 0.14991) gauss1_center: 7111.52424 +/- 0.02714092 (0.00%) (init = 7112.452) gauss1_sigma: 0.75932812 +/- 0.02151410 (2.83%) (init = 1.315216) gauss1_fwhm: 1.78808103 +/- 0.05066183 (2.83%) == '2.3548200*gauss1_sigma' gauss1_height: 0.05459081 +/- 7.8147e-04 (1.43%) == '0.3989423*gass1_amplitude/max(1e-15, gauss1_sigma)' gauss2_amplitude: 0.03945700 +/- 0.00329914 (8.36%) (init = 0.116005) gauss2_center: 7113.18957 +/- 0.04258439 (0.00%) (init = 7113.403) gauss2_sigma: 0.57985874 +/- 0.03338641 (5.76%) (init = 0.556759) gauss2_fwhm: 1.36546297 +/- 0.07861898 (5.76%) == '2.3548200*gauss2_sigma' gauss2_height: 0.02714638 +/- 0.00110204 (4.06%) == '0.3989423*gauss2_amplitude/max(1e-15, gauss2_sigma)' fit_centroid: 7111.98258 +/- 0.01448240 (0.00%) == '(gauss1_amplitude*gauss1_center +gauss2_amplitude*gauss2_center )/(gauss1_amplitude+gauss2_amplitude)' [[Correlations]] (unreported correlations are < 0.100) C(bpoly_intercept, bpoly_slope) = -1.0000 C(bpeak_amplitude, bpeak_center) = +0.9722 C(gauss1_amplitude, gauss1_sigma) = +0.9186 .... C(gauss1_amplitude, gauss2_amplitude) = -0.7900 So there are 2 "gaussN_amplitude" Parameters for the area under each Gaussian, with a strong negative correlation. If you do >>> uvars = peakresult.result.uvars >>> print(uvars['gauss1_amplitude'], uvars['gauss2_amplitude']) >>> print(uvars['gauss1_amplitude'] + uvars['gauss2_amplitude']) You'll get the area of the sum of the two Gaussians as 0.104+/-0.004 0.0395+/-0.0033 0.1434+/-0.0023 That is, one might expect the uncertainties in 'gauss1_amplitude' and 'gauss2_amplitude' should add in quadrature, but that would overestimate the uncertainty in the sum by almost ~2x because the parameters are highly (and negatively) correlated. The uncertainty in the reported peak heights and the centroid of the two peaks also takes the correlations into account. Such calculations are not too hard to do from the Larch buffer or Python script, but we might think about how that could be exposed more flexibly. I can be persuaded to just add a Parameter "peaks_area" = the sum of the amplitudes, which would then have this value and uncertainty. Any suggestions for what should be exposed here? Cheers, --Matt On Thu, Dec 14, 2023 at 10:47?AM Ryan Parmenter <rap35(a)kent.ac.uk> wrote: > > Hello, > > > > I successfully fitted Gaussian curves to my pre-edge peaks using Larix (XAS Viewer), yielding the centroid energy for indicating Fe oxidation state. The expected trends are observed, however, in reviewing articles, I noticed they report the area and integrated intensities of Gaussian curves, often automated with Peakfit software. Can Larix generate these values in its statistical report? If not, how can I calculate the area and integrated intensity using Larix's provided values, considering the absence of standard deviation? > > > > Thanks, > > Ryan > > > > _______________________________________________ > Ifeffit mailing list > Ifeffit(a)millenia.cars.aps.anl.gov > http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit > Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit ------------------------------ Subject: Digest Footer _______________________________________________ Ifeffit mailing list Ifeffit(a)millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit ------------------------------ End of Ifeffit Digest, Vol 250, Issue 5 ***************************************

1 0

Larix Pre-edge Statistics
by Ryan Parmenter 15 Dec '23

15 Dec '23

Hello, I successfully fitted Gaussian curves to my pre-edge peaks using Larix (XAS Viewer), yielding the centroid energy for indicating Fe oxidation state. The expected trends are observed, however, in reviewing articles, I noticed they report the area and integrated intensities of Gaussian curves, often automated with Peakfit software. Can Larix generate these values in its statistical report? If not, how can I calculate the area and integrated intensity using Larix's provided values, considering the absence of standard deviation? Thanks, Ryan

2 1