Following up on Roman's good points: an absolute value of BE is meaningless without a proper calibration. This is typically done by choosing a common element that should be unchanged in all samples. In your case, Zn would be a possible candidate as pointed out by Roman. It is possible that the Zn is shifted in the same direction and magnitude, which could imply a difference in the charging of the sample during measurement, different ground contact etc.

Off-topic recommendations:

I suggest using a smaller energy step (i.e. the distance between two measurement points); the current data allows some fitting uncertainties, especially for more than one peak.

Although it doesn't matter a lot in this case, when you fit the background, you shouldn't use the endpoints of the data; instead, select the boundaries of the background such as that it blends in the noise level away from the peak.

Dear Firman

Regarding your observations I agree with the above but I would like to add that the appearance of higher binding energy component (HBEC) in the oxygen 1s core peak from x-ray photoelectron spectroscopy (XPS) suggests the presence of oxygen vacancies in these samples. The amount of oxygen during the synthesis determines the vacancy concentration in the samples which is directly related to the deposition parameters such as  pressure. For instance, TiO2 films deposited under higher pressure exhibited the HBEC in their O 1s spectra.

Dear Roman, Lior and Maria, thank you for your input.

@Roman & Lior

I found that both Al2p3 and Zn2p are also shifted 0.3 eV to higher BE. I assume this happen due to the charging effect that you have mentioned.

@Maria

Yes indeed. The above data was extracted from depth-XPS, I intend to calculate the percentage of Vo (oxygen vacancy) at AZO/ZnO interface.

Hi all,

It is more common to plot XPS data with a reverse x-axis, that is high binding energies on the left, low on the right.

That said, the main peak is the O2- from the lattice and the shoulder, if not due to artifacts or O vacancies,  is very likely related to C=O species, whether these are carbonates or carboxylates (more probable). It could also be -OH, H2O groups from the outer surface. Check your C1s signal. If you have a shoulder around 288 eV, that's your guy. If you have several components in the C1s related to O, then you need several components in the O1s as well, perhaps three signals.

It also seems that your energy steop size is a bit large. What are the step size, pass energies used for these measurements? For fitting, high resolution is preferred and minimum good values are 0.1 eV step size and 30 eV pass energy. Lower values are better, but not too low. I'd say 0.05 eV and 20 eV are very good.

You can use a Au standard for your calibration since the 4f core level is close in energy to Al2p, the Au 4d is close to the C1s, and the Au 4p3/2 is close to O1s... For Zn, I guess the Cu standard is a good one to catch the high binding energy Zn2p. Beware that Zn metal and ZnO have the same binding energy in the 2p core level due to lattice effects. An unambiguous determination would be based on the Zn 3p core level, and would remove the need to check the Cu standard since Zn 3p lies around 90 eV, from the top of my head.

Best

Hii Firman,

I am wondering what is your energy resolution. It is interesting to deconvolute your spectrum into several components but you have to take account your own energy resolution. 

Following up on Lior Kornblum suggestion about background: what type of background correction do you use? Is it possible what component with high BE is artifical due to inapropriate background correction?

Wrong selection of the background can indeed affect the fitting and components you obtain from it.

I use Shirley background. The key advice on that is that the tails should blend in the middle of the background noise (away from the peak). i.e.: the background line should be in the middle/average of the noise. It goes without saying that cleaner data (longer scans) makes the noise lower and the fitting easier.

Following on from all of  the above answers

Looking at the spectra there doesn't appear to be much of an inelastic step to high binding energy.  So I would say that you are looking at a surface contribution (tentative, as you need a larger background range to be sure Tougaard refs- sorry I don't have them off the top of my head).  The points about references and calibration (e.g. Au4f) are important they can set "boundaries" for the curve fitting (width, position etc a good reference here and therein but old :-) is Briggs and Seah "Practical Surface Analysis, Auger and X-ray Photoelectron Spectroscopy").  Of course if there is some light charging you wont see this without a "gold dot" on your sample.  Perhaps the Al and zinc signals will provide an insight. 

A good check will be in addition to look at the Auger-parameter, as charging will cancel out. With a high concentration of hydrogen in your samples a x-ray induced shift may occur varying with time and is dependent on annealing temperature. As the two O1s spectra look pretty identical a contribution of AlOx is unlikely. 

To Lior Kornblum: Shirley-type background is quite sensitive to the position of endpoint. See Briggs and Seah "Practical Surface Analysis, Auger and X-ray Photoelectron Spectroscopy".  

Yurij, I believe that when done correctly there shouldn't be reproducibility problems with Shirley (unless it is some wild feature/peak that I didn't come across yet). I also wrote a simple Shirley implementation myself and had no such problems.

Can you give an example of where you think Shirley doesn't work well? Could you offer an alternative? 

Non metals?  In other words the major inelastic losses are offset from the main peak which could be the case here?  I think you also have to careful about extrinsic losses and the excitation depth e.g. film thickness. 

This small shift could be due to particle size or defects in the film.

What is your reference point for the binding energies? The Zn(2p) peak or adventitious carbon (assuming you get carbon contamination during the preparation)?