1) Since you have particles and not a bulk film, I would assume you caught most of it.

2) If you want surface sensitivity, maybe try XPS as a complementary method.

3) In the first picture, the lines are in the position where Ti would be seen if it was there while in the second picture they mark the positions of the copper lines. That's probably just the last element your operator checked in each case.

4) I mean, when you do EDS, you want to know the stoichiometry, now you have it. If it's the one you wanted to have, start feeling great satisfaction, if it's not the one, think what could have gone wrong, both preparatively or analytically.

I agree with Juergen. However my experience with powders tells me there could, in theory, be some settling issues. Both ’heavier vs lighter‘ and/or ‘larger vs smaller, particles tend to separate over time due to jostling of the powder’s container. Most of this kind if issue can cause the surface of the blend to be differing ratios over that at the bottom of the blend. A complete mixing of the sample just before analysis can help alleviate this issue. But, multiple samplings taken from throughout the bulk, which are then averaged together for a final result, also do wonders for accuracy.

Dear Siddig Abuelgasim ,

the penetration depth of electrons is in the very low µm regime, even below 1µm, and is depending on the acceleration voltage. The penetration depth/attenuation length of the x-ray fluorescence is also the low µm regime for low atomic number elements such as Mg.

Thus, with respect to your 30µm particles, you will not 'see' the core of them.

Unfortunately, 'low µm regime' is not surface or surface near region. Here you have to go for photoelectron spectroscopy, as it has been suggested by Jürgen Weippert .

With respect to the red and green lines in your spectra, I recommend to have a look at x-ray emission line tables, such as for example that in the 'x-ray data booklet':

http://xdb.lbl.gov/Section1/Sec_1-2.html

(click on 'Table 1-2 (pdf-format)')

and get familiar with them.

Binding energies can be found in:

https://xdb.lbl.gov/Section1/Sec_1-1.html

X-ray attenuation coefficients (and thus attenuation lengths) can be derived from

https://physics.nist.gov/PhysRefData/Xcom/html/xcom1.html

Good luck and

best regards

G.M.

Depth of beam penetration (X-ray generation) for your specimens is about 1-3 microns. Ignore red lines - it's computer generated aid for an operator. Green lines - spectrum itself. If you are interested in composition variation along diameter of your particles, you can embed and polish them (they are big enough). Carbon is OK. But your specimens were coated with gold. Better to coat with carbon for EDS analysis. But even gold is OK.

Ignore red line. It appears when you click on the screen. Green line is the expected peak. It seems that your materials have Cu and Ti.

The energy used for EDX (EHT=electron High Tension) is usually 25 keV.

What is the goal of your EDS analysis? To confirm you have Ti and Cu in one sample, vs Cu and Mg in the other? Your results are adequate to make that statement.

The provided estimation of the elemental composition is a good guess at best! It's hard to tell but it seems that the duane-hunt limit (i.e. the incident electron energy) was at 10 keV - that it not nearly enough to get a fully excited Cu-K Line. The overvoltage provided should be at least x2 so 16 kV or better 20 kV accerleration voltage!

Even when this criterion is met, a standardless quantification is not as exact as the results suggest. It is tempting to calculate a stoichiometry, but this should be done very cautious! A precise quantification of an unknown material without standards is very hard to achieve!