Do you have a manual for your EDS system? Can you talk to operator of EDS system? 

I will be careful about using EDS to determine elemental composition because EDS is mostly telling you what is present on the surface of your sample. I advice that you combine it with ICP, and maybe XRD.

But, u can't get bond information

usually while taking eds spectrum they will provide wt percent and at % of the elements present in it. so you can get the elemental composition from it. you contact the institute where you have taken the study

Some  instruments has an option to give you data in the compound form also. But as per my experience with EDS data, I agree with Adeyemi S. Adeleye on this. Go for other techniques, ICP or MP-AES can definitely give better quality and more accurate data..

Thank you all.

EDS is the standard technique for the local determination of the chemical composition. Practically all SEM and TEM are attached with this technique and they give for decades already reliable data. Of course one has to take into account that the information depth depends on some parameters and is in comparison to a cm thick sample quite small, but usually inhomogeneities within a sample are much bigger. There are no doubts that there are alternative techniques which are more precise but they are either not that spatially high resolving not that simple to use and comparatively cheap. i am not sure why any other technique should have less potential for making errors, and why XRD should tell you anything about chemistry (if you don't know anything about the composition, mixture of phases etc.). Also Vegards rule is only a rule and I think any EDS measurement is better than this simple approach. The major problem is the operation since only because there is an automatic mode which seems to substitute a scientist by a technician it doesn't mean that the automatically derived results are always correct, although this has been improved amazingly. Please don't discredit a technique which is widely used and very successful, only because some of your results don't satisfy you. This will happen with every technique, especially XRD. Also XRD has no bigger information depth and preparation artifact do not only affect the intensity but also the phases. it is always the same: use the proper technique for your specific problem, and prevent as much as possible any change of your sample during sample preparation.

Coming back to the original question: Each EDX or EDS system exports the results in many different formats and units. You simply need to ask the operator. The quality of the results are mainly controlled by your sample preparation and the operators qualification. 

EDX plot gives you signals from characteristic radiations of elements with respect to radiation count per second. So ratio of intensities would give you the ratio of elemental composition. I recommend to use matlab or origin. 

The major questions are: Which data did you get from the EDX system?

If these are entire spectra you can of course go the way recommended by M. Kanik, but I myself wouldn't assume that the results can be better than from the commercial software generated since only they know the correction parameters required for the hardware used. Background processing is a big challenge, peak overlap, escape an sum peaks... The list is long and I don't think that an unexperienced user is able to import data into Matlab or Origin and extract a more reliable result. It might look better to your expectations sometimes but then the question arises why the commercial manufacturers which have high-specialized people sitting continuously on the improvement of peak deconvolution, consideration of M an N lines, zero shift etc fail with their sophisticated software? I don't want say that there is no better software thinkable, but only from people who are very experienced and long-time working in this field. This would be comparable to the recommendation: Take your XRD diagram and load it into Origin and determine the crystalline as well as amorphous fraction of each phase.

Dear Pratibha,

We can easily find the element composition form EDX spectrum by simple calculation/trial and error method.

For example, If you got EDX spectrum for nano iron oxide, then we can see the Atomic and Weight percentage values in your result. Using these values we can find the formula of iron oxide (FeO or Fe2O3 or Fe3O4).

Herewith I have attached the EDX result with for calculation.

All the Best!

@Dr. V. Ranjithkumar

You did not show in your report what standards (if any) were used.

What substrate was used?

If it was standardless analysis, then results are meaningless (it does not work with oxides).

My sample was TiO2.

Thank you all.

Dear Ranjithkumar,

Thanks for information. I am attaching one file here. Can you help me in this regard?

Pratibha

What kind of EDS data do you mean? If you have a cps spectrum, then you need to compare it to some standards and apply some mathematics for corrections. If you have a weight percentage composition, then just divide it by a mass of appropriate elements and normalize it as you wish.

You must have seen the attached file.I dont know much information about peak intensity and all. Its a TiO2 sample and I have taken it from some other institution. Do you have standards for EDS data then please help me out.

Hi, since I do not have much information about your sample, sample preparation, and sample mounting technique you used, I am guessing that:

1. Carbon peaks correspond to the carbon tape you probably used in SEM-EDX to spread your powder sample.

2. Cu peaks are due to X-ray target used in the instrument and Silicon peaks due to Si(Li) detector used in the EDX. Ideally the instrument should have taken care of the peaks.

3. I still have no idea about the potassium peaks in your EDX spectra.

4. All further it can be said that TiO2 is present (possibly alongwith SiO2, K2O, Cu2O or CuO). Ideally the instrument should have been calibrated to get rid of peaks from points 1&2 above. For any quantitative information, the point made by Prof. Vladimir above still applies, and I agree to it (about the use of standard for calibration ).

5. and Lastly you will have to find out whether the instrument operator did what is called commonly as ZAF CORRECTION (Z=Atomic mass, A=Absorption, F=Fluorescence). These will also modify the peak intensities.

Hope this helps. Cheers.

You spectrum was pretty well described by Prashant Sahu. I can just add that it looks like your specimen does not have potassium (K) at all - its "peak" is on the level of background noise. Si peak is possibly an artifact. Presence of Cu is common artifact for TEM analysis, but I believe you worked with SEM. So, presence of Cu is real, unless you mounted your specimens on a Cu tape. You really have to talk to EDS operator.

P.S. You can buy standards from from EM suppliers, but they are rather expensive; spectra from standards should be collected prior to specimen's spectra.

Thank you Prashant and Vladimir

I will talk to EDS operator.

Pratibha

I agree with Vladimir except of the very strict appearing recommendation to collect standards prior specimens spectra. :-) Of course this is always then a very good test if something unexpected happens. But EDS is a standard technique which is widely automated. If the operator does not make any rough mistake, or you have something else on your sample like a Cu tape which also generates radiation if backscattered electrons hits it the technique is quite correct. Quite means, in the scale of +-2%. You can get it better but also worse if certain conditions are not fulfilled, i.e. for certain element combinations it can become very challenging. But standard analyses are usually very fast and easy to get. As rough rule of thumb: the more peak overlappings you have the less accurate the results might be. Everything else is a question of the  activated corrections (as pointed out by P. Sahu: ZAF or PRZ), and other effects like zero shift, energy resolution, escape peaks, sum peaks. However, this the operator should be able to recognize and bring it under control (i.e. correction). TiO2 is a quite simple material.

What irritates me in your spectra is the high zero peak intensity...

What could be the reason for this you think?

decalibration, detector and/or electronic age... the list can be quite big.

I would highly recommend to remeasure a standard they are typically using, or an "older" sample where this effect did not yet appeared. Simply for comparison. At least this high zero intensity would make me nervous. :-) But this is actually another story and not the general problem you have...although it would effect it of course..

Thank you.

I would not pay much attention to a zero peak. It really does not matter to your relatively simple spectra (no overlapping, no "artifact" peaks) in a case of qualitative analysis. Anyway, you are not getting quantitative analysis: it a bit more complicated for oxides, you need standards and experienced operator. Buy the way, you may want to advise your EDS operator to set a cut off threshold to 50 or 100 eV (spectra will look much better without huge zero peak). 

Ok. I will do that next time.

at least I would recommend to check whether this zero peak was that strong from the beginning. It also tells you something about the status of the analytical system, or how trustful low kV analyses are or can be. We had this problem and finally the detector was damaged what we didn't realized by anything else except of unusual C concentrations.

Ok I'l talk to the operator.

Dear Pratibha,

 send me the excel sheet which was provided by the person who analysed your sample. Ask him for that values. Then I will help you.  

I dnt get the excel sheet. I'l ask the person and try to arrange it.

That's fine.

Hi Vladimir,

I forgot to mention you that mine was not SEM analysis, it was of TEM analysis.

Pratibha

Please, I have a question:

Does the Energy-dispersive X-ray spectroscopy (EDX or EDS) provide the chemical composition of the material as metal or metal oxide? i.e the results obtained from the EDX are in percent for metal or for metal oxide?

I found in different published papers that the data obtained from EDX are in % of metals and other papers as % metal oxide. When we use it as metals and when as metal oxide?

Is the following sentence right or false:  "EDX determines the chemical composition of materials as metals and exposes them as oxides but it doesn't mean that they are oxides". Many thanks for your reply.

@ S.s. Metwally,

Ques. The results obtained from the EDX are in percent for metal or for metal oxide?

It depends on the instrument used for EDX analysis. Some instruments give the at% and wt% of compounds, but most of the instruments give these results for elements.

And if a sample containing light and heavy elements is analysed, then the concentration of heavier elements can be further used. But the concentration of lighter elements such as B, O, H etc. cannot be used as it has many errors. It can be corrected by ZAF corrections.

Dr. V. Ranjithkumar I have the excel data, can you guide me how can I find elemental composition in percentage?

Dear Mettway,

The EDS can only give us an informations (chimical composition) about the second phase (particles),

Yousf

Yousf Islem Bourezg , how so?

The makers of SEM/TEM etc could easily add a software that could predict, what could be the possible compounds from the elements the system detected

If any of the makers of soft wares for the SEM makers is reading this please think in way

This will make our job way too easy

Dear all,

Please find attached a file that discribes the way to calculate the weight% and atomic% of second phase ( compound or precipite),

Generaly for me, i compare what i have as precipites and the obtained results by EDS,

Good Luck ; )

Dr. V. Ranjithkumar You have shared one word file, which contains about Fe2O3, Fe3O4. But you made the thing complicated. Can you simplify it.. What you have written is below..

Herewith I have shown the EDX result of carbon/metal oxide composites

Using At % values:

For Fe2O3

2 Fe -> 3 O

1.14 Fe -> x = (1.14 x 3)/2 = 1.71

For Fe3O4

3 Fe è 4 O

1.14 Fe è x = (1.14 x 4)/3 = 1.52

Conclusion:

Calculated oxygen atomic % value of Fe2O3 is 1.71and obtained result from EDX result is 2.41.Hence this EDX result confirms the formation of Fe3O4 nanoparticles in carbon shell.

How EDX can be used to get the % composition?