1. The change in relative composition as a result of annealing is dependent upon the film, you may have secondary reactions and re-ordering of crystal structure, evaporation of elements from the film etc etc. So the relative ratios of elements are very likely to change pre and post annealing.

2 XPS is probably the best method for both surface characterization and also within the bulk if you can perform depth profiling studies. This will allow you to compare relative ratios and identify element environment and oxidation state. 

Hi. The technique you should use is Rutherford backscattering (RBS) or sometimes just called backscattering (BS). I use the RUMP or the SimNRA programs It has both composition calculation and depth resolution. That means if you have layered structures, the elements in each layer can be identified. The only difference with BS is the minimum detection limits (about 0.8% mass per mass at best).

Then NO your element composition will remain the same unless effervescence occurred. The only aspect is that say Al concentration is low for the pure element, but the remainder of the Al would be incorporated as Al2Fe3 (just as an example) So the Al is still present but in a different form. For elements present your can also do X-ray emissions (PIXE) simultaneously with RBS. these two methods are widely applied together. PIXE has a better minimum detection than RBS (about 0.1 ppm) but the layer compositions to be entered in the software for PIXE. I use the GeoPIXE program.

Just to add.

I have done in situ annealing for my M. degree; you can do this also, if not tedious (but then we do research). What happens is that for various time intervals you can retrieve the data and would be able to "see" any reactions happening. more so, you can deposit an arbitrary inert maker such as Ta in between the layers and would then be able to determine the kinetics of the reaction.

I am in Cape Town, South Africa, with 2 facilities available. Let me know where you are and perhaps I can refer you to someone nearby.

All the success.

Regards.

In addition to my previous speaker:

There are some more techniques like ICP-MS, if you like themn: GD-OES, ICP-OES, AAS,... but the resolution is different.

2nd: I would also recommend XPS for both surface and bulk. But the investigator should have worked with your thin film system, because of the intensity of sputtering. You do damage your sample and you also may change the structure (and local composition) due to the (high) application of energy. For chemical environment and oxidation state exist multiple literature, so the interpretation could be done comparatively easy.

Dear Nicholas Chadwick 

I need said that your idea is correct, maybe I have evaporation during annealing but it's depend on your sample, but in this composition evaporation temperature  is more than 1500'c, my high temperature is between 800'c to 1037'c. it's not big problem I can measuring pre and post annealing. 

I kept my sample in vacuum and I wanna annealing in vacuum. then my oxidation is too low.

I agree with Johan on using Back-scattering (RBS) and Proton induced X ray emission (PIXE)  for composition analysis and depth profiling.  in addition,  elemental depth distribution information may show  any elemental redistribution after annealing which could be due to chemical reaction with substrate, or evaporation or liquid- solid segregation.  a ternary phase diagram on the system need to be checked.

Dear Johan Dr Andre Mars 

thanks for your help, it is a good method for my case. I'm in Hsinchu, Taiwan.

B R

Dear Ivo Zunke

actually just I tried for kept my sample in vacuum and also my annealing is in vacuum state. by the way, thankful for your help, the introduce that methods were great.

Regards 

Also I agree with Prof. Chu and Johan, because the condition for kept the samples and annealing is more important. I think the best way is double check before and after annealing. and I need comparison between two different states.

Thanks a lot,

Annealing in  Argon gas instead of vacuum  could prevent sublimation .

Great. that's a good offer. thanks.

To determine composition one can use EDAX, WDS (EPMA) or SIMS as well. These methods have good accuracy.

for quantitative we trust of XRF, WDS ,  with normal or normalized quantitative i trust in Normal can gave you good approaching results .

A well calibrated EDAX system can give very accurate results; also as suggested above XRF is also useful here. I used to always use two different methods for my thin films as then you can compare and contrast the results.

In answer to your question, I don't think annealing at 800C would affect the actual overall composition, but I would check if you have a mixed phase material (according to this paper you could have), but it will be dependent upon the original compositions of elemental material used.

SIMS in this application is barely semi-quantitative.  XRF or the SEM methods both rely on external and very accurate knowledge of the film thickness and also sample-matched standards.  Without a great deal of knowledgeable care you will not get anywhere near 1%.

XPS depends critically on knowledge of the electron mean free path,  which is rarely known even as well as 10%.  It is also a sputtering technique,  and subject to sputtering artefacts which may very well be very large.  In this application it is unreliable,  and certainly not accurate.

Ion beam analysis methods are probably as good as you will get for this application (Rutherford backscattering together with particle-induced X-ray emission).  See the ref for details.

http://dx.doi.org/10.1016/j.nimb.2011.09.020

I've used GDOES to determine chemical composition of coatings and how they interact with substrates.

Chemical composition is likely to change after annealing due to diffusion between film and substrate.

I used to always use EDAX system,EDS. Being the film ordinary not alone, it may have the substrate, may affect the result.

If your film have the phase  transition, annealing temperature may affect the crystal structure, but the whole composition may not change.

can anyone provide me the protocol for ICP-MS to quantify the metalloprotein from given sample