On each surface exposed to air you will find some monolayers of contamination simply from environmental air. The main component of this contamination is generally carbon. Since XPS is so surface sensitive you will practically always see a carbon peak. People use this peak sometimes for energy calibration. especially when charging may shift the whole spectrum.

https://www.researchgate.net/publication/281180023_Carbon_Contamination_at_Silver_Surfaces_Surface_Preparation_Procedures_Evaluated_by_Raman_Spectroscopy_and_X-ray_Photoelectron_Spectroscopy

https://books.google.hu/books?id=8w4yU7B9I0kC&pg=PA575&lpg=PA575&dq=%22carbon+contamination%22+peak+XPS+inorganic+surface&source=bl&ots=nT8tFxOcbE&sig=YqgdaUbYFFahr_9G3cQ9uGx8m3A&hl=hu&sa=X&ved=0ahUKEwjespz46fXQAhXMSRoKHSk5A0QQ6AEIOzAF#v=onepage&q=%22carbon%20contamination%22%20peak%20XPS%20inorganic%20surface&f=false

http://www.technology.matthey.com/article/58/4/180-188/

http://www.hou.usra.edu/meetings/lpsc2015/pdf/1517.pdf

http://www.psu.edu/dept/winograd/pdf/19.pdf

http://onlinelibrary.wiley.com/doi/10.1002/(SICI)1096-9918(199711)25:12%3C924::AID-SIA317%3E3.0.CO;2-2/abstract

Apparently the carbon contamination is ubiquitous - although the majority of the references found are on metals. My personal experience is limited to polymers where carbon is always present. Taking into account, owever, that few nm thick layer or even less is enough for detection, carbon may come from anywhere. The position of 285 eV seems to indicate C-C carbon in aliphatic chains, which may come from solvent residues, methane or whatever present in the laboratory atmosphere. Ar ion bombardment may help to remove the topmost contaminaton layer.

Article Carbon Contamination at Silver Surfaces: Surface Preparation...

Respected, Dr. Schiffmann and Dr. Banhegyi

Thank you for your expert comments and the suggested literature. I would like to ask another related question that:

If carbon is so universal in all type of samples, then how is the exact composition of sample determined? Is there some procedure to nullify the carbon contribution?

If you want to remove unwanted carbon peak. Do not expose it to atmosphere, you can remove the carbon layer by in-situ ion itching of the surface. Better to prepare the samples in-situ without exposing them to air.

If you have access to ion etchning (easiest is argon) then you can "erase" the surface. If you select relatively mild conditions, you can get rid of carbon without damaing the inorganic components. Percentage-wise how much carbon you have on your surface? The presence of carbon should not influence too much the ratio of the other components.

As already said, the presence of carbon is always there on air-exposed samples.  It comes from reaction with CO2 to form carbonates, and just general hydrocarbons in eh air also.

The best form of action if you are really concerned by it is to etch using Argon clusters as opposed to argon ions as the former are much less damaging of the surface.  If you only have access to a monotomic argon source, then short bursts at lowest possibly energies (eg 30 seconds at 300 or 500 eV)

@ Dr. Rajendra and Dr. Morgen,

Thanks for your guidance on removing the carbon related layer from the surface of our samples. I would keep in mind next time prior to XPS measurements.

@ Dr. Banhegyi,

The carbon peak is there with considerable intensity. In some of the research articles, I read that presence of carbon is accompanied by some oxygen too, which leads to overestimation of the actual oxygen content of the samples.

Dear daniwalia, Carbon is the material which always present in atmosphere, so nano leval layer is deposited on every type of samples, if you are not performing in clean room environment. Thats why we use the cabon as a calibrating element during the xps spectra. Some times if the sample is such that carbon doesnt have adhesion on the surafce of sample in that condition oxyzen is used as a calbrating element because this is also universal present.

So to remove this carbon, you require a few minute plasma cleaning during the xps. this plasma cleaning should be at sufficient voltage and current, otherwise it may go to the main layer of the sample.This plasma cleaning is an attachment with the xps.

Dear Mr. Hitesh,

Thank you for adding another piece of good knowledge to me. It needs some experience before we know how much voltage and current to have, for appropriate plasma cleaning. I believe the operator of machine knows best as he/she characterises large number and variety of samples.

I'm late to the party, and echo most of what has been said.  On the other hand, it's worth noting that even with cleaning and no atmospheric exposure, you will still have the ubiquitous 285eV signal.  I suppose under very restrictive circumstances one could minimize it, but CO will desorb from the walls of a steel ultra high vacuum chamber and there is potential backstreaming contamination from hydrocarbon pump oil depending on the system as well.  Cleaning will only reduce these effects temporarily.  Best to perform your C1s scan immediately after cleaning, as well as O1s if you're doing it.  The chemical shifts of O containing compounds are small enough that it hasn't been a useful core in my experience, however.  You could always check the relative peak intensities of peaks between the main signal for C1s and less tightly bound cores too.  That can give information from deeper in the sample, on average, if you have the patience to deal with the diminished cross section.  I don't know enough about glasses to advise that for sure, but I would expect them to be homogeneous enough for that to give an estimation of the surface contribution.  I would certainly end a set of data collection with a C1s scan to verify that the sample had remained consistent over the course of the set's collection.  I sure would believe more of the XPS-derived conclusions out there if folks were taking these steps to ensure consistency.

Respected Dr. Young,

You are not late Sir. Scientific things can always be discussed anytime. Kindly accept my sincere thanks for adding your response to the question.

Another way to remove adventitious carbon from the surface is to perform in-situ annealing. These carbon species are typically not strongly bonded to the surface so they can escape to vacuum upon heating.

See for example: Applied Physics Letters 109 (2016) 211602

Ar+ ion gun was used in XPS to etching the sample surface.

In addition, we can used plasma etching to clean the sample surface (large area). The equipment can be used/installed in XPS, SEM, TEM and etc. The price is OK/nice.

hi i would like to join the party with a question if you allow me Prof Danewalia ? if the carbon peak comes from the adsorption of CO2 from air, so i have to record Oxygen peak which around 533eV??

Dear Dr. Aboud,

It is always pleasure to discuss science at any time. You are welcome and I thank you to share your expert opinion. As in our case, oxide glasses were under investigation an obvious Oxygen signal was observed which originated basically from the glass composition. However, contribution from CO2 could also not be ruled out.

thanks

yes in the oxide phases the main contribution in the O 1S peak comes from the oxide it self. but there are other peaks related to oxygen. like oxygen vacancies, adsorbed oxygen on the sample surface ( like carbon which is weakly adsorbed on the oxides surface) .

W. Song, "Synthesis and Characterization of Transition Metal Oxide Catalysts for Environmental and Energy Storage Applications," 2016.

but i wonder if CO2 adsorbed on the surface, should i have two peaks one cor C and one for o?

thanks again

Hello Dr. S. S. Danewalia

I would like to mention that the signal strength of C1s can provide an idea of the source of error (for your case). If it is quite significant (as you have mentioned earlier), it could be related to the calibration. Else, the above-mentioned reason could be possible for adsorbed CO or CO2 in which O1s must contain O-C (C-O/C=O/O-C=O/O-C-O) related binding energies between 532-536eV. Then onl,y we can confirm the adsorption of CO or CO2 on the surface.

I agree with all the suggestions given by different technique experts. But, in my opinion, oxide based glass samples should not exhibit any binding energies related to C1s.

I hope that it will help you out to get the correct information from the results.

Best of luck!

Regards.

Dear bhai @Ayush,

Apka answer dekh kr bohut khushi huyi. Dhanyabad. Yeh data publish kr chuke, msc k balakon ka kaam tha.

Your information would surely help us in future research and all other reader following the question.

I hope you are doing great and submitting your thesis soon... My best wishes...

Regards,

Satwinder.

Ok veere.