Bremsstrahlung fluorescence decreasing Ni trace concentrations in olivine

[Joe Boesenberg]

Hi John

A colleague showed me the attached Llovet et al 2023 paper and I am not understanding the explanation for Ni. I understand the increases from fluorescence for the Ti, Ca, and Al, but am missing how Ni can decrease. The explanation in the paper is "the decrease is explained by the lower bremsstrahlung fluorescence emitted from the sample as compared to that emitted from the
standard." Can you explain this? How is the Ni standard (the authors don't say what standard was used, which is annoying) affecting the outcome as you move incrementally toward the olivine/melt boundary?   Aren't the standard effects essentially hidden since you are dividing any point in the sample by the standard to get the final concentration. Is the difference really between sample and standard or is it position on sample relative the boundary?

Perplexed....Joe

[Probeman]

Great question!    Yeah they really didn't explain this negative boundary effect very well did they?

Think of a material that has significant self fluorescence, say Fe fluoresced by Ni Ka or Co fluoresced by Cu Ka, or just a sample with self fluorescence from the continuum say trace Zn Ka fluoresced in MgAl alloy from the the continuum. In this last example we found that Zn Ka could be detected 500 um away from the Zn boundary when measured in a MgAl alloy and there's no characteristic lines in Mg or Al that can fluorescence the Zn K edge.

See here for a great example:

https://probesoftware.com/smf/index.php?topic=58.msg5621#msg5621

And here for a silicate example:

https://probesoftware.com/smf/index.php?topic=58.msg7667#msg7667

Also check the poster by Julie Chouinard attached to this post:

https://probesoftware.com/smf/index.php?topic=58.msg6481#msg6481

[Joe Boesenberg]

Yes, thank you. Just posted the answer to your link. It is because the adjacent material has less (or none) of the element of interest than your target. 

[Probeman]

Yes, thank you. Just posted the answer to your link. It is because the adjacent material has less (or none) of the element of interest than your target.

Yes, exactly.

Note that the effect is much larger (positive or negative) when characteristic fluorescence is involved in addition, rather just continuum fluorescence.  The good news is that by running a quick model in the Standard app using PENFLUOR, can give you an answer very quickly as shown here:

https://probesoftware.com/smf/index.php?topic=58.msg214#msg214