Re: [Ifeffit] Ifeffit Digest, Vol 158, Issue 3

[Bruce Ravel]

While I am agnostic on the issue of nomenclature of the effect in
question, far be it from me to deny Matthew (or anyone else) his right
to a quality rant.  Well done!

That said, this:

On 04/03/2016 10:15 PM, Matthew Marcus wrote:

I might also note that although many elaborate schemes have been
proposed for correcting for the effect, they all require knowing things
about the
sample that you don't always know, e.g. the exact geometry of something
that isn't planar, or the exact composition.


is the most important comment in this thread.

I would go a small step farther -- they all require knowing things
about the sample that, except in certain contrived situations, you
almost certainly don't know.

B

--
 Bruce Ravel   bra...@bnl.gov

 National Institute of Standards and Technology
 Synchrotron Science Group at NSLS-II
 Building 535A
 Upton NY, 11973

 Homepage:http://bruceravel.github.io/home/
 Software:https://github.com/bruceravel
 Demeter: http://bruceravel.github.io/demeter/
___
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

Re: [Ifeffit] Ifeffit Digest, Vol 158, Issue 3

[Matthew Marcus]

Yup.  In fact, self absorption, in the correct sense, can help reduce 
"self-absorption" in the misnomer sense!  That's how the grazing-exit method 
works.

I might also note that although many elaborate schemes have been proposed for 
correcting for the effect, they all require knowing things about the
sample that you don't always know, e.g. the exact geometry of something that 
isn't planar, or the exact composition.  I find that the simple parameterization
which comes from a thick-planar-sample model works reasonably well as an 
empirical method with an unknown parameter.  Still, I wouldn't necessarily trust
Debye-Waller (ss2) values taken from EXAFS on strongly overabsorbed samples, 
even with a correction that gets the right coordination number.
mam

On 4/3/2016 7:04 PM, Matt Newville wrote:

Hi Matthew,



On Sun, Apr 3, 2016 at 8:13 PM, Matthew Marcus > wrote:


It shouldn't be called 'self-absorption'.  That's a misnomer, which seems 
to have come from a 1992
paper (Troger, et. al."Full correction of the self-absorption in 
soft-fluorescence extended x-ray-absorption fine structure", PRB 46,3283 (1992).
The effect was described and analyzed in a 1982 paper, which called it an 
"attenuation factor": Goulon, et. al. "On experimental attenuation factors of 
the amplitude
of the EXAFS oscillations in absorption, reflectivity and luminescence 
measurements", J. Physique 43, 539 (1982).

 mam


Thanks!! I completely agree, though I wasn't aware of the historical precedence for the 
mistake.Using "over-absorption" is a far better term.

In X-ray fluorescence,  "self-absorption" actually means the attenuation of 
fluorescence generated within a sample as it travels out of the sample.   For 
over-absorption in XAFS, the issue is measuring absorption in fluorescence mode when the 
concentration of the absorbing element is not infinitesimal or when the sample thickness 
is not infinitesimal.

--Matt


___
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


___
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

Re: [Ifeffit] Ifeffit Digest, Vol 158, Issue 3

[Matt Newville]

Hi Matthew,



On Sun, Apr 3, 2016 at 8:13 PM, Matthew Marcus  wrote:

> 
> It shouldn't be called 'self-absorption'.  That's a misnomer, which seems
> to have come from a 1992
> paper (Troger, et. al."Full correction of the self-absorption in
> soft-fluorescence extended x-ray-absorption fine structure", PRB 46,3283
> (1992).
> The effect was described and analyzed in a 1982 paper, which called it an
> "attenuation factor": Goulon, et. al. "On experimental attenuation factors
> of the amplitude
> of the EXAFS oscillations in absorption, reflectivity and luminescence
> measurements", J. Physique 43, 539 (1982).
> 
> mam
>
>
Thanks!! I completely agree, though I wasn't aware of the historical
precedence for the mistake.Using "over-absorption" is a far better
term.

In X-ray fluorescence,  "self-absorption" actually means the attenuation of
fluorescence generated within a sample as it travels out of the sample.
For over-absorption in XAFS, the issue is measuring absorption in
fluorescence mode when the concentration of the absorbing element is not
infinitesimal or when the sample thickness is not infinitesimal.

--Matt
___
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

Re: [Ifeffit] Ifeffit Digest, Vol 158, Issue 3

[Matthew Marcus]

It shouldn't be called 'self-absorption'.  That's a misnomer, which seems to 
have come from a 1992
paper (Troger, et. al."Full correction of the self-absorption in soft-fluorescence 
extended x-ray-absorption fine structure", PRB 46,3283 (1992).
The effect was described and analyzed in a 1982 paper, which called it an "attenuation 
factor": Goulon, et. al. "On experimental attenuation factors of the amplitude
of the EXAFS oscillations in absorption, reflectivity and luminescence 
measurements", J. Physique 43, 539 (1982).

mam

On 4/3/2016 4:56 PM, Matteo Busi wrote:

Hi Bruce,
Thanks for your help, it's really appreciated.
I was trying to keep it simple but it seems I'm just messing around.

What I am trying to do is to perform a new developed self-absorption correction 
using collected fluorescence absorption coefficient data on a CuSO4 
(pentahydrate) capillar (cylinder) and spherical sample with Cu as absorber, 
having different values of molarity and penetration depth.
The correction expression requires these measured quantity:
μX(E) : the absorption coefficient due to a given core excitation of the 
absorbing atom
   - I used the background function for this ( bkg(E))
μo = μ(E) : photoelectric total linear absorption coefficient of the sample at 
incident energy E
- I used the xmu(E) for this
μh = μ(E) :photoelectric total linear absorption coefficient of the sample at 
fluorescence emission Ef
- I used the xmu(Ef) with Ef the K absorption edge of Cu ( 
8.9789 eV)
χ: and here I used the chi(E) values of the exported ascii .xmu file

Hope it's clear now.
Matteo


2016-04-02 19:00 GMT+02:00 >:

Send Ifeffit mailing list submissions to
ifeffit@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-requ...@millenia.cars.aps.anl.gov 


You can reach the person managing the list at
ifeffit-ow...@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: Ifeffit Digest, Vol 158, Issue 1 (Matt Newville)


--

Message: 1
Date: Fri, 1 Apr 2016 15:56:29 -0500
From: Matt Newville >
To: XAFS Analysis using Ifeffit >
Subject: Re: [Ifeffit] Ifeffit Digest, Vol 158, Issue 1
Message-ID:
 >
Content-Type: text/plain; charset="utf-8"

Matteo,


On Fri, Apr 1, 2016 at 8:35 AM, Matteo Busi > wrote:

 > Hi Bruce,
 > Now this is clear.
 > In my case the correction I have to perform requires these
 > measured/evaluated parameters: chi, mu (not sure if it's better to work
 > with normmu here) and the background function.
 > Is it physically meaningless to have the mu(k) and bkg(k) data? If that 
is
 > not the case I would like to have these two columns exported in the ascii
 > chi(k) file.
 > So then I can perform the correction and re-import the new corrected
 > chi(k) and make comparison with the uncorrected.
 >
 > Kind regards,
 > Matteo
 >
 >
It would certainly help (and by that I mean "help you get the answers
you're looking for") if you gave us more details about what you're trying
to do rather than ask for how to what you think you want to do.

Just to be clear: Converting from E to k does not involve a Fourier
transform, just
k = sqrt((2m*e/hbar**2)*(E-E0))

for k in Ang^-1 and E in eV, that's  k ~= sqrt(0.262468 
*(E-E0))

I'm not sure what you're trying, but I would image you want to use
normalized mu(E) or at least pre-edge subtracted mu(E).  That is, what we
call mu(E) is typically really -ln(I/I0)
where I and I0 are not the actual intensities before and after the sample,
but counts or counts per time in some detector that samples the flux.
These values include more or less arbitrary scale factors (amplifier gains,
etc) included.  That means that (unless you very careful) mu(E) does not
have meaningful units, certainly not cm^-1 or cm^2/gram.

Normalized mu(E) doesn't have such units either, but it's an easier place
to start.

--Matt

Re: [Ifeffit] Ifeffit Digest, Vol 158, Issue 3

[Matt Newville]

On Sun, Apr 3, 2016 at 6:56 PM, Matteo Busi  wrote:

> Hi Bruce,
> Thanks for your help, it's really appreciated.
> I was trying to keep it simple but it seems I'm just messing around.
>
> What I am trying to do is to perform a new developed self-absorption
> correction using collected fluorescence absorption coefficient data on a
> CuSO4 (pentahydrate) capillar (cylinder) and spherical sample with Cu as
> absorber, having different values of molarity and penetration depth.
> The correction expression requires these measured quantity:
> μX(E) : the absorption coefficient due to a given core excitation of the
> absorbing atom
>   - I used the background function for this ( bkg(E))
>


Wouldn't using bkg(E) suppress the fine-structure you are trying to
recover?  Anyway what are the units, are you normalizing bkg(E) to
something meaningful?  I would suggest using data from Chantler or
Cromer-Liberman here.   Or perhaps match mu(E) or bkg(E) to these.


> μo = μ(E) : photoelectric total linear absorption coefficient of the
> sample at incident energy E
>- I used the xmu(E) for this
>

Again, be mindful of what the units are.


> μh = μ(E) :photoelectric total linear absorption coefficient of the sample
> at fluorescence emission Ef
>- I used the xmu(Ef) with Ef the K absorption edge of
> Cu ( 8.9789 eV)
>

I am pretty sure that this is not what you want.  You probably want mu(E)
at the Cu fluorescence energy (8.046  keV). This can make a significant
difference.

HTH,

--Matt
___
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

Re: [Ifeffit] Ifeffit Digest, Vol 158, Issue 3

[Matteo Busi]

Hi Bruce,
Thanks for your help, it's really appreciated.
I was trying to keep it simple but it seems I'm just messing around.

What I am trying to do is to perform a new developed self-absorption
correction using collected fluorescence absorption coefficient data on a
CuSO4 (pentahydrate) capillar (cylinder) and spherical sample with Cu as
absorber, having different values of molarity and penetration depth.
The correction expression requires these measured quantity:
μX(E) : the absorption coefficient due to a given core excitation of the
absorbing atom
  - I used the background function for this ( bkg(E))
μo = μ(E) : photoelectric total linear absorption coefficient of the sample
at incident energy E
   - I used the xmu(E) for this
μh = μ(E) :photoelectric total linear absorption coefficient of the sample
at fluorescence emission Ef
   - I used the xmu(Ef) with Ef the K absorption edge of Cu
( 8.9789 eV)
χ: and here I used the chi(E) values of the exported ascii .xmu file

Hope it's clear now.
Matteo


2016-04-02 19:00 GMT+02:00 :

> Send Ifeffit mailing list submissions to
> ifeffit@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-requ...@millenia.cars.aps.anl.gov
>
> You can reach the person managing the list at
> ifeffit-ow...@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: Ifeffit Digest, Vol 158, Issue 1 (Matt Newville)
>
>
> --
>
> Message: 1
> Date: Fri, 1 Apr 2016 15:56:29 -0500
> From: Matt Newville 
> To: XAFS Analysis using Ifeffit 
> Subject: Re: [Ifeffit] Ifeffit Digest, Vol 158, Issue 1
> Message-ID:
>  fgzksbv1sezr3m5prkxann...@mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> Matteo,
>
>
> On Fri, Apr 1, 2016 at 8:35 AM, Matteo Busi  wrote:
>
> > Hi Bruce,
> > Now this is clear.
> > In my case the correction I have to perform requires these
> > measured/evaluated parameters: chi, mu (not sure if it's better to work
> > with normmu here) and the background function.
> > Is it physically meaningless to have the mu(k) and bkg(k) data? If that
> is
> > not the case I would like to have these two columns exported in the ascii
> > chi(k) file.
> > So then I can perform the correction and re-import the new corrected
> > chi(k) and make comparison with the uncorrected.
> >
> > Kind regards,
> > Matteo
> >
> >
> It would certainly help (and by that I mean "help you get the answers
> you're looking for") if you gave us more details about what you're trying
> to do rather than ask for how to what you think you want to do.
>
> Just to be clear: Converting from E to k does not involve a Fourier
> transform, just
>k = sqrt((2m*e/hbar**2)*(E-E0))
>
> for k in Ang^-1 and E in eV, that's  k ~= sqrt(0.262468*(E-E0))
>
> I'm not sure what you're trying, but I would image you want to use
> normalized mu(E) or at least pre-edge subtracted mu(E).  That is, what we
> call mu(E) is typically really -ln(I/I0)
> where I and I0 are not the actual intensities before and after the sample,
> but counts or counts per time in some detector that samples the flux.
> These values include more or less arbitrary scale factors (amplifier gains,
> etc) included.  That means that (unless you very careful) mu(E) does not
> have meaningful units, certainly not cm^-1 or cm^2/gram.
>
> Normalized mu(E) doesn't have such units either, but it's an easier place
> to start.
>
> --Matt
> -- next part --
> An HTML attachment was scrubbed...
> URL: <
> http://millenia.cars.aps.anl.gov/pipermail/ifeffit/attachments/20160401/658d433a/attachment-0001.html
> >
>
> --
>
> Subject: Digest Footer
>
> ___
> 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
>
>
> --
>
> End of Ifeffit Digest, Vol 158, Issue 3
> ***
>
___
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