Mass energy-absorption data

[halcyon-gh]

It seems that in the current version mass energy-absorption data is not available in the database. Is there interest in adding this information to xraydb? If so I would be willing to work on this, based on data provided in the work of Boone & Chavez (https://doi.org/10.1118/1.597899).

[newville]

@halcyon-gh I may not understand what you are looking for.

The current database includes mass-attenuation data from Elam, Ravel, and Sieber (2002) and from Chantler (2000, 2016) -- see https://xraypy.github.io/XrayDB/biblio.html#id4

The data from Elam, Ravel, and Seiber was compiled based on several sources. I think that should be considered an "update" or "restatement" of the LLNL data. It should also replace the "McMaster" compilation.

The FFAST data from Chantler has very similar results for mass attenuation but also includes the resonant scattering terms (and does not separate Compton from Rayleigh scattering, if I recall correctly).

Both tables can be used to calculate the absorption (photo-electric) and attenuation (photo-electric absorption + coherent (Rayleigh) scattering + incoherent (Compton) scattering) of X-rays (say 100 to 500,000 eV) by matter.

Are you looking for something different?

[halcyon-gh]

I'm looking for mass energy-absorption rather than mass-attenuation coefficients, i.e. the fraction of photon energy that is deposited when transiting a medium rather than the fraction of photons that undergo interactions when transiting the medium (https://physics.nist.gov/PhysRefData/XrayMassCoef/chap3.html). This information would be useful for simulating x-ray detectors and dosimetry.

The work I refer to does indeed include much duplicate information that is already present, but as far as I can tell the energy absorption is not available from the other sources currently in xraydb.

[newville]

@halcyon-gh The code here tabulates data for elements. If I understand correctly, the values you are looking for would depend on material density and also on secondary processes (for example, absorption might cause fluorescence, which might leave the sample but could also be re-absorbed causing tertiary processes).

I don't think those values can be tabulated. They depend strongly on the composition, density, and even shape of the sample.

OTOH, we calculate penetration depths and attenuation lengths all the time, as with material_mu(). See https://xraypy.github.io/XrayDB/examples.html#mu-calculations-for-materials for example.

Perhaps you would want to calculate those quantities. Is that correct?

I have not looked at the Hubble & Seltzer pages in a while, so I am not sure how easy it would be to code up those calculations. When I see something like

 ... where X is the average energy of fluorescence radiation (characteristic x rays) emitted per absorbed photon

I think "well that is over-simplified". Our samples have many fluorescing elements, some at 10 weight percent, but still dominating the absorption and fluorescence, and there is strong Compton and Elastic scatter. Any "average energy of fluorescence per absorbed radiation" would be a complex calculation.

But, maybe you have a simpler case.... Anyway, I would expect that the values there are not ones to be tabulated, but to be calculated using the tabular data for the elements.

[halcyon-gh]

The idea is indeed to calculate material energy absorption based on elemental properties similar to the material_mu() function.

I'm not sure if I understand your reply correctly. Are you saying that as derived values you would consider adding the calculations to obtain elemental energy absorption values from already present elemental properties to xraydb but not the tabulated elemental energy absorption values themselves, or are you referring to material energy absorption that should be calculated from tabulated elemental energy absorption?

With the latter I agree. For the former I would need to take a closer look at how complex that would be since that kind of calculation is outside my area of expertise.

[newville]

@halcyon-gh

I'm not sure if I understand your reply correctly. Are you saying that as derived values you would consider adding the calculations to obtain elemental energy absorption values from already present elemental properties to xraydb

Yes.

but not the tabulated elemental energy absorption values themselves, or are you referring to material energy absorption that should be calculated from tabulated elemental energy absorption?

What "tabulated elemental energy absorption values"? Doesn't that have to depend on the composition and density of the material? How does one tabulate that?

For material_mu(), that calculation needs elemental composition/stoichiometry and density of the material to calculate "X-ray absorption/attenuation length".

Calculating "energy deposited" would be a bit more complicated. You would have to start with 'what is absorbed", "what is coherently/elastically scattered away", and "what is incoherently/inelastically scattered away" as in material_mu.

Then, you have to try to account for all fluorescence caused by the emission and then decide if re-absorption of those X-rays is important. That will depend on the sample geometry and size and also depend strongly on the energy of the fluorescence.

For both elastic and inelastic scattering, you may need to consider secondary processes (say, absorption of the scattered beam). For inelastic scattering, the energy given to the electron will be "deposited" (unless you want to account for electron emission, in which case that becomes 'will be mostly "deposited"'), and for the angular variation of a) the energy of that electron and b) the scattering cross-section. These depend on X-ray energy too. It turns out that we have calculated and tabularized some of those in the compton_energies table (and use these to calculate "ion chamber flux", as energy deposited from Compton scattering is sometimes not negligible). But, again, these could depend on sample shape and size, especially at the micro-scale.

[gpachecoguevara]

Thanks for your attention to this request. We've since implemented an alternative solution for our use case, so this issue can be closed.

[newville]

@gpachecoguevara should we assume that you are speaking for @halcyon-gh?

Apparently "mass energy absorption" is used in some medical applications, and tables from Hubbel and Seltzer used for that.
I suppose it would be possible to add that, though I have not look into how those data are tabulated or could be exposed.

[gpachecoguevara]

@gpachecoguevara should we assume that you are speaking for @halcyon-gh?

Yes, we’re working on the same project. Thank you for the clarification and for maintaining this package.
We agree that mass energy-absorption data is indeed useful for medical physics applications (our project being one of those cases).
However, we’ve since implemented our own solution using the data from Boone & Chavez (https://doi.org/10.1118/1.597899), so we won’t be pursuing this feature further here.

[newville]

OK, thanks.
I'm willing to leave this Issue open in case anyone is interested in the topic.
I'm not opposed to adding such data, but it is not a high priority for me. If someone is interested in adding these data and a bit of support code, I'd be willing to help get that done.