Gerhard is correct. I might add that how we configure our microscope for doing EDS is critical, and so is the type of quantitative process we use. Equally important is a well quantified standard to reference in the quantification process. Do not use the “internal standard“ most scopes come with. It can be off by as much as 30%.
l have used a “least squares” quantification method, and my best alignment protocols to achieve an exact and reproducible quantification for well quantified knowns. How you do the quant matters dramatically.