I am trying to make a quantitative model for Laser Induced Breakdown Spectroscopy (LIBS) analyzing metal alloy samples. I am using an internal standard normalization (so far it produced the best repeatability results) and would like to some how to keep using it in the quantitative model. My issue is that if the base element of an alloy is Fe and I am trying to measure %w of Ni and Cr, then the ratio of Ni/Fe is not only affected by the %w Ni, but also by the %w of Cr as well, since more Cr means less Fe and a different Ni/Fe value for the same %w Ni.
I have standards with different and more-or-less known amount of Ni and Cr. If there a standard way to handle this type of problems?
It seems like something people have run into many times and there should be cookie-cutter solutions, but I have a lot of difficulty to find one. So far the only discussed models I've seen is a change in %w of one element, while others are locked constant. A textbook or a journal paper discussing the solution would do just fine.
Thanks in advance!