Isotopes and quantitative microanaysis

[John Donovan]

I've decided it might be worth starting a new topic on this question because it's not completely intuitive how isotopes (atomic mass) are involved with our matrix correction procedures.  This will be a short post to start with but let's see how much interest there is.

To begin I noticed that Andrew Locock mentioned editing the atomic weight for carbon from 12.011 to 12.0107 in the Elements/Cations window and Probeman teased him a bit about that (though maybe Andrew had a good reason for making that change?).  After all as Probeman pointed out, the C13 ratios in diamond can vary considerably:

https://probesoftware.com/smf/index.php?topic=71.msg12766#msg12766

Now Andrew was only able to make such a change because Probe for EPMA recently (last summer) enabled one to edit the atomic masses from the GUI in both the Standard application (for standard compositions) and for unknowns (in Probe for EPMA) as described here:

https://probesoftware.com/smf/index.php?topic=40.msg12014#msg12014

but what effect does changing the atomic weights have in our quantitative matrix corrections? Well ideally it should have no effect as mass effects are essentially unobservarble at typical electron beam energies:

https://www.cambridge.org/core/journals/microscopy-and-microanalysis/article/abs/compositional-averaging-of-backscatter-intensities-in-compounds/0E233FBBB78BC96E8C14977DBC2B556B

and in fact in the absorption correction we normalize out mass effects by utilizing mass fractions with mass absorption coefficients and in the energy loss correction by utilizing mass fractions with mass thicknesses.  Though as noted in the paper above, backscatter corrections (and continuum productions) are not always handled appropriately (it's simply electrodynamics all the way down!).

In other words mass effects should normalized out in the absorption and energy loss corrections, and for backscatter corrections we should utilize Z based fractions (i.e., the Z fraction method described in our paper) for compounds. In fact this is how Davis Stowe at Gatan was able to get their "quantitative backscatter" method for determining lithium concentrations to work:

Stowe, Davd, et al. "Assessing the Accuracy of Lithium Contents Determined by Combined Quantitative Backscattered Electron and X-ray Energy Dispersive Spectroscopy Analysis." (2024): ozae044-085.

All that said, what effect does changing the atomic weights have on our quant results? Well assuming the standard and the unknown are both utilizing the same atomic weights, it should have no effect on the matrix corrections, but of course the reported mass concentrations will change.  That is to say that SiO2 with Si29 will have different mass concentrations than SiO2 with Si28.