Basic questions about the preliminary data processing in Athena
Dear ALL, I am a beginner in Athena. Recently I got the spectra and I am confusing about the data calibration and alignment. The manual shows that calibrate the reference data of one scan and align other reference data to that calibrated one. Here I want to show examples. I have two samples A and B, Each sample has three scans with the corresponding reference data. For sample A, 1st scan is calibrated and the other two scans are aligned to 1st scan. Then merge them into merged A. The same procedures for sample B and get merged B. If I want to compare XANES of sample A and B, do I need to align the merged reference data between A and B? If so, why? My understanding is that all scans (original data and merged data) in sample A and B are calibrated to standard edge energy. Why do they need to align? I appreciate your help. Thanks, Haifeng
Hello Haifeng: Personally, I would not calibrate the data but merely align the references of A2 and A3 to the reference of A1 and similarly align B2 and B3 references to that of B1. Once merged, you can then aligh the reference of the merged B samples to that of the merged A samples. If you keep the references then you can always align data taken at a leter time to these data sets. If the energy of your reference is a small bit off from the tabulated value, that is OK as long as all the data being compared has aligned references. Carlo On Sun, 22 Jul 2018, Haifeng Li wrote:
Dear ALL,
I am a beginner in Athena. Recently I got the spectra and I am confusing about the data calibration and alignment. The manual shows that calibrate the reference data of one scan and align other reference data to that calibrated one.
Here I want to show examples. I have two samples A and B, Each sample has three scans with the corresponding reference data. For sample A, 1st scan is calibrated and the other two scans are aligned to 1st scan. Then merge them into merged A. The same procedures for sample B and get merged B. If I want to compare XANES of sample A and B, do I need to align the merged reference data between A and B? If so, why? My understanding is that all scans (original data and merged data) in sample A and B are calibrated to standard edge energy. Why do they need to align?
I appreciate your help.
Thanks,
Haifeng
-- Carlo U. Segre -- Duchossois Leadership Professor of Physics Interim Chair, Department of Chemistry Director, Center for Synchrotron Radiation Research and Instrumentation Illinois Institute of Technology Voice: 312.567.3498 Fax: 312.567.3494 segre@iit.edu http://phys.iit.edu/~segre segre@debian.org
Hi, Carlo,
Thanks for your answer.
I may ask you more questions. Why you do not recommend to calibrate the
data? Calibration is used to find the E0 and match it to the literature
data. What is the real difference between calibration and alignment?
Thanks,
Haifeng
On Sun, Jul 22, 2018 at 5:30 PM, Carlo Segre
Hello Haifeng:
Personally, I would not calibrate the data but merely align the references of A2 and A3 to the reference of A1 and similarly align B2 and B3 references to that of B1. Once merged, you can then aligh the reference of the merged B samples to that of the merged A samples.
If you keep the references then you can always align data taken at a leter time to these data sets. If the energy of your reference is a small bit off from the tabulated value, that is OK as long as all the data being compared has aligned references.
Carlo
On Sun, 22 Jul 2018, Haifeng Li wrote:
Dear ALL,
I am a beginner in Athena. Recently I got the spectra and I am confusing about the data calibration and alignment. The manual shows that calibrate the reference data of one scan and align other reference data to that calibrated one.
Here I want to show examples. I have two samples A and B, Each sample has three scans with the corresponding reference data. For sample A, 1st scan is calibrated and the other two scans are aligned to 1st scan. Then merge them into merged A. The same procedures for sample B and get merged B. If I want to compare XANES of sample A and B, do I need to align the merged reference data between A and B? If so, why? My understanding is that all scans (original data and merged data) in sample A and B are calibrated to standard edge energy. Why do they need to align?
I appreciate your help.
Thanks,
Haifeng
-- Carlo U. Segre -- Duchossois Leadership Professor of Physics Interim Chair, Department of Chemistry Director, Center for Synchrotron Radiation Research and Instrumentation Illinois Institute of Technology Voice: 312.567.3498 Fax: 312.567.3494 segre@iit.edu http://phys.iit.edu/~segre segre@debian.org _______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit
The calibration should be done at the beamline with a foil. Once that is done, then you don't really want to change it. Remember that just shifting the energy is not actually the correct way to do a shift. The energy shift is a non-linear function of the angle (Bragg's Law) and so when you shift energy you are really distorting Chi(k) since that is also a non-linear function of energy (E^1/2). Yes, it is usually a small effect but I prefer not to apply too many shifts in energy if possible. I assume that the calibrate function is the same as the align except the E-shift is left at zero and a correction is applied permanently to the data. Carlo On Mon, 23 Jul 2018, Haifeng Li wrote:
Hi, Carlo,
Thanks for your answer.
I may ask you more questions. Why you do not recommend to calibrate the data? Calibration is used to find the E0 and match it to the literature data. What is the real difference between calibration and alignment?
Thanks,
Haifeng
On Sun, Jul 22, 2018 at 5:30 PM, Carlo Segre
wrote: Hello Haifeng:
Personally, I would not calibrate the data but merely align the references of A2 and A3 to the reference of A1 and similarly align B2 and B3 references to that of B1. Once merged, you can then aligh the reference of the merged B samples to that of the merged A samples.
If you keep the references then you can always align data taken at a leter time to these data sets. If the energy of your reference is a small bit off from the tabulated value, that is OK as long as all the data being compared has aligned references.
Carlo
On Sun, 22 Jul 2018, Haifeng Li wrote:
Dear ALL,
I am a beginner in Athena. Recently I got the spectra and I am confusing about the data calibration and alignment. The manual shows that calibrate the reference data of one scan and align other reference data to that calibrated one.
Here I want to show examples. I have two samples A and B, Each sample has three scans with the corresponding reference data. For sample A, 1st scan is calibrated and the other two scans are aligned to 1st scan. Then merge them into merged A. The same procedures for sample B and get merged B. If I want to compare XANES of sample A and B, do I need to align the merged reference data between A and B? If so, why? My understanding is that all scans (original data and merged data) in sample A and B are calibrated to standard edge energy. Why do they need to align?
I appreciate your help.
Thanks,
Haifeng
-- Carlo U. Segre -- Duchossois Leadership Professor of Physics Interim Chair, Department of Chemistry Director, Center for Synchrotron Radiation Research and Instrumentation Illinois Institute of Technology Voice: 312.567.3498 Fax: 312.567.3494 segre@iit.edu http://phys.iit.edu/~segre segre@debian.org _______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit
-- Carlo U. Segre -- Duchossois Leadership Professor of Physics Interim Chair, Department of Chemistry Director, Center for Synchrotron Radiation Research and Instrumentation Illinois Institute of Technology Voice: 312.567.3498 Fax: 312.567.3494 segre@iit.edu http://phys.iit.edu/~segre segre@debian.org
Hi, Carlo,
Thanks for your physical explanation for the calibration.
I have another question. I measured the same element at different
beamlines. So if I want to compare different samples measured at different
beamlines, how should I do for this situation given the possible different
calibration methods? I know alignment is very important and just shifting
the energy by calibrating would have issue.
Here I mean different calibration methods is dedicated for my element
ruthenium Ru with E0 of 22117 eV. The derivative curve show it has two
peaks. One beamline used the first one for calibration. The other used the
second. Another used the middle number between the two peaks by claiming
that the peaks stand for oxidation state with one peak for one oxidation
state and multiple peaks for multiple oxidation states. Here for Ru metal
with one oxidation state, since the main peak splits into two, the middle
energy between the two peaks was chose as the E0. Attached is the file of
Ru foil for more information. I do not know which one should be right. In
the manual, it shows choosing the first maximum point as the edge energy.
How should I choose the edge energy? Maybe I think too much. It does not
matter as the calibration is consistency.
The reference data from different calibration methods can be aligned very
well. This is another confusing part. Does the calibration would affect the
alignment?
Thanks in advance.
Haifeng
On Mon, Jul 23, 2018 at 10:48 AM, Carlo Segre
The calibration should be done at the beamline with a foil. Once that is done, then you don't really want to change it. Remember that just shifting the energy is not actually the correct way to do a shift. The energy shift is a non-linear function of the angle (Bragg's Law) and so when you shift energy you are really distorting Chi(k) since that is also a non-linear function of energy (E^1/2). Yes, it is usually a small effect but I prefer not to apply too many shifts in energy if possible.
I assume that the calibrate function is the same as the align except the E-shift is left at zero and a correction is applied permanently to the data.
Carlo
On Mon, 23 Jul 2018, Haifeng Li wrote:
Hi, Carlo,
Thanks for your answer.
I may ask you more questions. Why you do not recommend to calibrate the data? Calibration is used to find the E0 and match it to the literature data. What is the real difference between calibration and alignment?
Thanks,
Haifeng
On Sun, Jul 22, 2018 at 5:30 PM, Carlo Segre
wrote: Hello Haifeng:
Personally, I would not calibrate the data but merely align the references of A2 and A3 to the reference of A1 and similarly align B2 and B3 references to that of B1. Once merged, you can then aligh the reference of the merged B samples to that of the merged A samples.
If you keep the references then you can always align data taken at a leter time to these data sets. If the energy of your reference is a small bit off from the tabulated value, that is OK as long as all the data being compared has aligned references.
Carlo
On Sun, 22 Jul 2018, Haifeng Li wrote:
Dear ALL,
I am a beginner in Athena. Recently I got the spectra and I am confusing about the data calibration and alignment. The manual shows that calibrate the reference data of one scan and align other reference data to that calibrated one.
Here I want to show examples. I have two samples A and B, Each sample has three scans with the corresponding reference data. For sample A, 1st scan is calibrated and the other two scans are aligned to 1st scan. Then merge them into merged A. The same procedures for sample B and get merged B. If I want to compare XANES of sample A and B, do I need to align the merged reference data between A and B? If so, why? My understanding is that all scans (original data and merged data) in sample A and B are calibrated to standard edge energy. Why do they need to align?
I appreciate your help.
Thanks,
Haifeng
--
Carlo U. Segre -- Duchossois Leadership Professor of Physics Interim Chair, Department of Chemistry Director, Center for Synchrotron Radiation Research and Instrumentation Illinois Institute of Technology Voice: 312.567.3498 Fax: 312.567.3494 segre@iit.edu http://phys.iit.edu/~segre segre@debian.org _______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit
-- Carlo U. Segre -- Duchossois Leadership Professor of Physics Interim Chair, Department of Chemistry Director, Center for Synchrotron Radiation Research and Instrumentation Illinois Institute of Technology Voice: 312.567.3498 Fax: 312.567.3494 segre@iit.edu http://phys.iit.edu/~segre segre@debian.org
Hi Haifeng, Here's my take on what you described: Two sample to be studied at the same edge using the same reference. Three scans on A and three on B are done with simultaneous reference. First: compare the three reference scans for A. If they agree, then compare the data scans. If they also look similar (i.e. no evidence of changing in the beam), you can merge the three data scans on A right away. If the references don't agree, then you determine by how much they differ (how much a correction would be needed to bring them into alignment) and apply that same correction to the data before merging. It is not meaningful to merge data that is not aligned. (If the sample scans show changes from scans 1 to 3, then you need to rethink how to do the measurements) Repeat for B. Now compare the merged (corrected beforehand if necessary) references for A and B. If they agree, you can compare data for A and B (merged) directly. If they do not, determine how much one reference differs from the other and apply that same correction to, say, B, that brings its reference into alignment with A's reference, and then compare A and B You align the references between samples to the same value in order to do a meaningful comparison between them. If you report energy positions of features in your near-edge spectra, the reader would need to know to what energy those positions are referenced. When possible, I recommend references that have tabulated edge values (i.e. metal foils). A reference need not be the same edge as the one being studied. For arsenic, as an example, the gold L3-edge is quite close to the As K-edge and serves as a good reference. If not possible to use a metal foil, use a reference that another interested researcher could readily obtain or has used. This allows for comparison of reproducibility. If your reference looks nothing like literature, you may have a problem with the beamline or in how you processed the data. This should be one of the first things you check when you start taking data at the beamlne. You should also note how the beamline was calibrated when you did the measurements. regards, Robert On 2018-07-22 1:21 PM, Haifeng Li wrote:
Dear ALL,
I am a beginner in Athena. Recently I got the spectra and I am confusing about the data calibration and alignment. The manual shows that calibrate the reference data of one scan and align other reference data to that calibrated one.
Here I want to show examples. I have two samples A and B, Each sample has three scans with the corresponding reference data. For sample A, 1st scan is calibrated and the other two scans are aligned to 1st scan. Then merge them into merged A. The same procedures for sample B and get merged B. If I want to compare XANES of sample A and B, do I need to align the merged reference data between A and B? If so, why? My understanding is that all scans (original data and merged data) in sample A and B are calibrated to standard edge energy. Why do they need to align?
I appreciate your help.
Thanks,
Haifeng
_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit
Hi, Robert,
Thanks for your help.
One more question. What is the real difference between calibration and
alignment? Do they have specific meaning?
Thanks,
Haifeng
On Sun, Jul 22, 2018 at 5:45 PM, Robert Gordon
Hi Haifeng,
Here's my take on what you described:
Two sample to be studied at the same edge using the same reference. Three scans on A and three on B are done with simultaneous reference.
First: compare the three reference scans for A. If they agree, then compare the data scans. If they also look similar (i.e. no evidence of changing in the beam), you can merge the three data scans on A right away. If the references don't agree, then you determine by how much they differ (how much a correction would be needed to bring them into alignment) and apply that same correction to the data before merging. It is not meaningful to merge data that is not aligned. (If the sample scans show changes from scans 1 to 3, then you need to rethink how to do the measurements)
Repeat for B.
Now compare the merged (corrected beforehand if necessary) references for A and B. If they agree, you can compare data for A and B (merged) directly. If they do not, determine how much one reference differs from the other and apply that same correction to, say, B, that brings its reference into alignment with A's reference, and then compare A and B
You align the references between samples to the same value in order to do a meaningful comparison between them. If you report energy positions of features in your near-edge spectra, the reader would need to know to what energy those positions are referenced.
When possible, I recommend references that have tabulated edge values (i.e. metal foils). A reference need not be the same edge as the one being studied. For arsenic, as an example, the gold L3-edge is quite close to the As K-edge and serves as a good reference. If not possible to use a metal foil, use a reference that another interested researcher could readily obtain or has used. This allows for comparison of reproducibility. If your reference looks nothing like literature, you may have a problem with the beamline or in how you processed the data. This should be one of the first things you check when you start taking data at the beamlne.
You should also note how the beamline was calibrated when you did the measurements.
regards, Robert
On 2018-07-22 1:21 PM, Haifeng Li wrote:
Dear ALL,
I am a beginner in Athena. Recently I got the spectra and I am confusing about the data calibration and alignment. The manual shows that calibrate the reference data of one scan and align other reference data to that calibrated one.
Here I want to show examples. I have two samples A and B, Each sample has three scans with the corresponding reference data. For sample A, 1st scan is calibrated and the other two scans are aligned to 1st scan. Then merge them into merged A. The same procedures for sample B and get merged B. If I want to compare XANES of sample A and B, do I need to align the merged reference data between A and B? If so, why? My understanding is that all scans (original data and merged data) in sample A and B are calibrated to standard edge energy. Why do they need to align?
I appreciate your help.
Thanks,
Haifeng
_______________________________________________ Ifeffit mailing listIfeffit@millenia.cars.aps.anl.govhttp://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit
_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit
On 07/27/2018 01:32 PM, Haifeng Li wrote:
What is the real difference between calibration and alignment? Do they have specific meaning?
In Athena, calibration means to determine a value for e0 shift such that a particular point in the data is specified to be a particular energy. Then the value for E0 is set to that particular energy. Thus, calibration changes both e0 and e0 shift for the group being calibrated. For example, set the E0 shift such that the first inflection point in copper is made to be 8979 eV. Then set the value for E0 in the background removal to 8979. In Athena alignment is the process by which one data group is given an e0 shift value such that it lines up with another data group. This is to compensate for some scan-to-scan variability in the behavior of the monochromator. Often, but not always, alignment is done using a zero-valent (or some other) reference which is measured simultaneously with the actual sample. In Athena, the alignment tool DOES NOT change the value of e0, but certainly does change the value of e0 shift for the group being aligned. The alignment tool does not change either e0 or e0 shift for the alignment standard. Whether you want to do one, the other, both, or neither depends upon the beamline and the type of data ensemble you have measured. This /is/ explained in the user manual. B -- Bruce Ravel ------------------------------------ bravel@bnl.gov National Institute of Standards and Technology Synchrotron Science Group at NSLS-II Lead Beamline Scientist, 06BM (BMM) Building 743, Room 114 Upton NY, 11973 Homepage: http://bruceravel.github.io/home/ Beamline: https://www.bnl.gov/ps/beamlines/beamline.php?r=6-BM Software: https://github.com/bruceravel Demeter: http://bruceravel.github.io/demeter/
HI Haifeng, Suppose you measure Fe foil three times, and all three scans overlap. They are aligned. But suppose the edge positions are at 7116 eV instead of 7112 eV. The calibration is off. You can adjust to it, but, as Carlo mentioned, it is really an angle correction, not simply a shift in energy. It is better to have an accurate calibration to begin with, then aligned and calibrated would be the same. Note: Choice of target calibration can differ - some use the X-ray data booklet values http://xdb.lbl.gov/ and some use the Kraft et al. values https://aip.scitation.org/doi/abs/10.1063/1.1146657 Just state which calibration you are using when presenting/publishing. -R. On 2018-07-27 10:32 AM, Haifeng Li wrote:
Hi, Robert,
Thanks for your help.
One more question. What is the real difference between calibration and alignment? Do they have specific meaning?
Thanks,
Haifeng
On Sun, Jul 22, 2018 at 5:45 PM, Robert Gordon
mailto:ragordon@alumni.sfu.ca> wrote: Hi Haifeng,
Here's my take on what you described:
Two sample to be studied at the same edge using the same reference. Three scans on A and three on B are done with simultaneous reference.
First: compare the three reference scans for A. If they agree, then compare the data scans. If they also look similar (i.e. no evidence of changing in the beam), you can merge the three data scans on A right away. If the references don't agree, then you determine by how much they differ (how much a correction would be needed to bring them into alignment) and apply that same correction to the data before merging. It is not meaningful to merge data that is not aligned. (If the sample scans show changes from scans 1 to 3, then you need to rethink how to do the measurements)
Repeat for B.
Now compare the merged (corrected beforehand if necessary) references for A and B. If they agree, you can compare data for A and B (merged) directly. If they do not, determine how much one reference differs from the other and apply that same correction to, say, B, that brings its reference into alignment with A's reference, and then compare A and B
You align the references between samples to the same value in order to do a meaningful comparison between them. If you report energy positions of features in your near-edge spectra, the reader would need to know to what energy those positions are referenced.
When possible, I recommend references that have tabulated edge values (i.e. metal foils). A reference need not be the same edge as the one being studied. For arsenic, as an example, the gold L3-edge is quite close to the As K-edge and serves as a good reference. If not possible to use a metal foil, use a reference that another interested researcher could readily obtain or has used. This allows for comparison of reproducibility. If your reference looks nothing like literature, you may have a problem with the beamline or in how you processed the data. This should be one of the first things you check when you start taking data at the beamlne.
You should also note how the beamline was calibrated when you did the measurements.
regards, Robert
On 2018-07-22 1:21 PM, Haifeng Li wrote:
Dear ALL,
I am a beginner in Athena. Recently I got the spectra and I am confusing about the data calibration and alignment. The manual shows that calibrate the reference data of one scan and align other reference data to that calibrated one.
Here I want to show examples. I have two samples A and B, Each sample has three scans with the corresponding reference data. For sample A, 1st scan is calibrated and the other two scans are aligned to 1st scan. Then merge them into merged A. The same procedures for sample B and get merged B. If I want to compare XANES of sample A and B, do I need to align the merged reference data between A and B? If so, why? My understanding is that all scans (original data and merged data) in sample A and B are calibrated to standard edge energy. Why do they need to align?
I appreciate your help.
Thanks,
Haifeng
_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov mailto:Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe:http://millenia.cars.aps.anl.gov/mailman/options/ifeffit http://millenia.cars.aps.anl.gov/mailman/options/ifeffit
_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov mailto:Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit http://millenia.cars.aps.anl.gov/mailman/options/ifeffit
_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit
participants (4)
-
Bruce Ravel
-
Carlo Segre
-
Haifeng Li
-
Robert Gordon