IT is due to chemical environment around silver. This is chemical effects.ranjana,I suggest you bring the samples to me.we can measure intensity ratios of pure silver and the compound by XRF and the n do EXAFS/XANES for the chemical state/bonding..hope you are in India. where are you presently.If you are in India bring/send the samples to me.

Normally you observe the plasmon peak as a satellite e.g. for the 3d lines (both 1/2 and 3/2 should show the loss) for metallic samples. for a clean surface you might see also the surface plasmon shifted to lower energies. If you look carefully you should see also multiple losses. Multiple losses are a clear indication for plasmons. For small Ag particles this might be different as you have the excitation of dipolar modes (collective and single particle) which may couple to photons as a decay channel. Thes energies of these modes are size dependent. The volume and surface plasmon energies of pure Ag are given by the optical  constants e1 and ie2 of Ag. 

If you have a simple electron gun in your UHV-system, then you may do an easy test. Just irradiate your sample with monochromatic electrons and measure the energy loss spectrum of the reflected electron - then you may easily measure the plasmons energies and do an assignment to your inelastic losses measured in XPS. This should usually work out! Good luck, Dirk

Thanks Prof. Gulyaer, Pettenkofer and Drik for your valuable suggestions. Thanks Daisy mam; Yes I am in India working in Mumbai University.

Dear Prof. Gulyaev,

I have calculated the splitting between main and plasmon peaks in Ag3d line and it is observed to be not same as in Ti 2p line.

For reference how my actual spectra looks like, i am attaching fig. with this massage. In the Fig. sample labeled as UV1 is Ag NP on the surface of TiO2 and sample labeled as ann1 is Ag ion doped as well as Ag NP present in the TiO2 sample.

Sir, can you send me article describing about the concept of TiO2 plasmon.

Dear Ranjana,

I am wondering about your peaks and their assignments - the plasmon loss peaks should appear at LARGER binding energy, due to the energy transfer of the photoelectron to the plasmon, the kinetic energy of the photoelectron is smaller, i.e. the binding energy should be larger (see e.g. Phys. Rev. B20 (1979) 2335, doi: 10.1103/PhysRevB.20.2335).

See also the attachment, where the Ag 3d-spectrum excited with a Magnesium source is shown. You see the sub-structure of the Mg Kalpha 3,4 related to the Ag 3d 3/2 level at lower binding energy of 364 and 366 eV, and the plasmon peaks related to the 3d 5/2 at ca. 372 eV and the plasmon loss of the 3d 3/2 at ca. 378 eV! The distance of those two peaks to the corresponding photoelctron peak is roughly 4 eV in both cases, in well agreement to the literature (check J. Electron Spectroscopy Related Phenom. 3 (1974) 381-398, doi:10.1016/0368-2048(74)80022-8)

Kind regards, Dirk

Thanks Prof Drik,

I will go through the articles.

Dear All,

I have gone through article, send by Prof Dirk; but to my worry is that I am not able to interpret the results obtained in my experiment. This is 1st time I am doing XPS analysis. Please help me to interpret the obtained results.

File attached

Thank  you,

Thank you sir for your valuable suggestion.

Dear Ranjana,

spectral interpretation is not so easy, as only small portions are shown (i.e. only few eV below and above the photoelectron lines). Already for a proper background removal, you need to have at least around 5-10 eV below as well as above the peaks of interest, especially if you have plasmon losses, shake up /down satellites, and effects of your non-monochromatic X-ray source. I recommend that you go into the lab again, and measure larger spectral ranges. In addition, what about energy calibration? Is the C1s peak always at the same energy and has an identical shape for all the samples? Kind regards. Dirk

@Roman - sorry, I was not aware of your comments which agree with what I have just posted a minute ago! Cheers, Dirk