I suspect contamination with a silicon compound. In this case the observed peak is the Si2s line. To confirm this assumption, try to find Si2p peak about the same intensity at 100-103 eV. It should be seen even in the survey spectrum.

I agree with Konstantin. It looks like an Si 2s line.

Thanks a lot. It reminds me that I did use the SiO2 flake as substrate.

@Konstantin I. Maslakov, @Leon Black

Dear Zhou,

me seems, the observed peaks (black curce) can be explained by an oxidized surface. The position of the peaks in the case of Bi2O3 are 164,3 eV (Bi4f-5/2) and 159,0 eV (Bi4f-7/2). The pure metal positions are 157 eV and 162,3 eV. The separation of both peaks due to spin-orbit interaction amounts 5,3 eV.

Bi has a strong affinity for oxygen. I would measure the spectra again immediately after splitting the sample under UHV-vacuum condition.

Yours

Rüdiger Mitdank

Rüdiger Mitdank

Dear Mitdank,

I agree with you about the Bi4f-5/2 and Bi4f-7/2 peaks, and my sample was definitely oxidated. Neverthless, my question is about the peak located at 152.5 eV.

Still, thank you for your answer. I am happy to discuss with you more.

Dear Zhou,

I found the following values for silicon-XPS-peaks:

Si-2s: 150,5 eV and in the case of siO2: 154,4eV.

So it could be, that the peak at 152,5 eV agrees with partially oxidized silicon. But this is only possible, if the Si-substrate lies free. The electrons below the Bi-layer cannot escape.

Dear Mitdank,

Thankyou for your reply, and you give a wise point.

I'll have my sample examined again without Si substrate and I'll let you know the result.

Rüdiger Mitdank

Rüdiger Mitdank

Dear Mitdank,

Sorry to have wait for a month. The XPS examine is not available for a long time.

Anyway I made new samples and had them tested again. There's not Si substrate this time, and the peak we discussed about appeared corresponding to your point. Which means the peak is probably related to Bi, however I failed to find this peak from references.

I want to hear a suggestion from you.

Yours

Zhou Xiangji

Dear Zhou,

I found a peak with the binding energy 153 eV for Radium: 5p3/2. May be, the bismuth arised from the radioactive decay of a radium isotop. If you study decay rows from uranium, you find such elements in progress.

Yours

R. Mitdank

Dear Zhou,

it seems, you are right, that this peak cannot be caused by an impurity. If this is the case, than additional stronger lines must be found.

I studied the literature concerning additional lines and found out, that additional energy loss lines in different materials can be observed - so called plasmon lines. This lines are due to interaction of photoelectrons with free electrons and in dependence on the band structure can this be responsible for these feature. This explains that in oxidized materials such lines vanish because no free electrons longer exist. I hope this explanation helps.

Yours

R. Mitdank

Rüdiger Mitdank

Dear Mitdank

Thankyou for your reply, and yes Energy Loss Lines is one kind of the accompanying signal beside Photoelectron Lines.

As you said, these lines are due to interaction of photoelectrons with free electrons. So the Binding Energy point of the peak should be higher than that of Photoelectron Lines, and that's why we call it "Energy Loss". However, the indefinable peak in my picture locates at lower Binding Energy point.

Nevertheless, your idea did inspire me to look up other kinds of accompanying signal in XPS. Having read some literatures, I found out that Auger Line is the only possible explaination, otherwise we have to admit the inpurity exists. I'm seeking the Auger Lines signal of Bi which fits my peak but I haven't got it yet.

I'll let you know when I get progress, and thanks again.

Yours

Zhou

Dear Zhou, you are right in your (non-) assigmnt of the "extra line" as being due to energy loss processes (be it plasmons, shake-ups or whatever else).

If you suspect an Auger line, then this will have two consequences

• the element emitting that Auger line will also give rise to PES lines somewhere else (-> take good quality survey data)
• the kinetic energy of the Auger line is independent of the photon energy used for taking the XPS data. Identifying unknown Auger lines is thus made relatively easy by taking XPS datasets with two different photon energies (typically Al-Kα and Mg-Kα , respectively) and to compare spectra plotted vs binding energy (XPS peaks at same position) with spectra plotted vs. kinetic energy (Auger peaks at same position)

I like that (i) you actually take notice of the small extra peak,(ii) have reproduced it and (iii) really try to find an explanation for it. Thats the way to go, though painstaking at times. Good luck.

Btw: plasmon losses can also occur due to excitation of bulk plasmons of bound electrons, not just free eletrons.

Googling for "Bi 4f XPS satellite" (image search) I might have found an instance of a similar finding as yours in the paper Article Magnetism in BiFe1-xNixO3: Studied through Electron Spin Res...

Unfortunately I have no access to it.

Kai Fauth

Dear fauth,

I'm grateful for your reply.

And you're right that we can distinguish Auger line from Photoelectron Line by changing X-ray source ( I didn't learn about the Auger line effect on PES). I am about to try the method you suggest, whereas the facility here is so popular that I can't make it soon.

Thankyou again and I best wishes for your work.