The signal and the background and the noise are all lower. Calculate these values:

I(780) - intensity in peak near 780eV

B(850) - average intensity in background near 850eV

N(850) - average RMS noise in a range of about 20eV near 850eV

Calculate the ratios:

S/N = [I(780) - B(850)] / N(850)

S/B = [I(780) - B(850)] / B(850)

Have your ratios S/N and S/B changed? If not, perhaps your sample has moved out of focus as it is cooled (although you say otherwise ... this seems a good first guess). Otherwise, what is the spot size on your sample? You imply the sample shifts perhaps laterally. Is you sample flat over that lateral extent? Finally, what extent is this sample being flooded with a charge neutralizer? As you drop below the critical temperature, do you need to recalibrate your neutralizer current?

Jeffrey J Weimer Thanks for taking the time to answer. What I meant to say is that we didn't change the focus of the X-ray gun even though the sample position might have changed, so it woulnd't surprise me if the sample had gone out of focus.

The S/B ration changes quite a bit: It's 0.011 below Tc and 0.035 at room temperature. One possibility is that the photoelectric peaks lose intensity because of the carbon contamination, but I don't see why this would change the background. Do you think a combination of carbon contamination and bad focus would make sense?

Now, I'm not sure what the RMS noise is. Is it:

for all samples in the energy interval?

The sample is small, about 3mm in diameter. We got a clean surface by cleaving, but we don't know how flat the cleave was. We did do a consistency check (moved the sample a bit in the x-y direction a few times) and the signal didn't really change.

I'm not sure what you mean by a charge neutraliser. The sample was grounded, but we didn't measure the current.

I would like to add some comments to the presented spectra.

What is really strange in these survey spectra is a high intensity of the signal at low binding energies (near zero BE). Usually it should be orders of magnitude lower than the maximum intensity in the spectrum. This isn't the case here. It looks like something is wrong with the spectrometer. To check the zero current of the detector you may acquire a survey spectrum with the X-ray gun switched off. The signal should be very low (zero or some counts). It may be also possible that the problem is with the X-ray gun. As a see you used non-monochromatic MgKa radiation. There should be a beryllium foil at the exit of the gun. If it's damaged or absent, the photoelectron signal may be affected by the high flux of electrons from the X-ray anode.

P.S. It better to present spectra in CPS (counts per second), not in counts!

The RMS noise is calculated as sqrt(sum(value - average)^2). A change of S/B by a factor of three may or may not be significant, especially when carbon may be contaminating the surface. For oxides, I might have expected to need a charge neutralizer.

I suggest that you first eliminate the case that your spectrometer + sample holder are not working properly when the stage is cooled. You could repeat the experiment with a clean Au sample on the sample holder and then with a cleaved piece of mica. The metal and ceramic samples bound the extremes of your superconducting sample.

Finally, I see a significant loss in what I suspect are the copper oxide peaks as the carbon appears. I have to wonder whether the focal point has shifted to a region of different composition.

Hello Lidia,

One of your question is about the noise that have also been observed in Auger analysis and was at the origin of a research program to connect the electronic noise in dielectric layers, the defects that produce charge fluctuations and the polarization energy fluctuations.

The other one is about the modification of the surface composition (increase of the carbon) that is also observed under electron beam and is not due to the adsorption of residual gases but to a variation of surface energy which is causing a surface diffusion from the sample back side or the sample holder (this has also been shown experimentally).

Explosions of dielectric samples in ESCA and AES suggested that noise and excitation / recombination processes on defects had a direct relationship with the macroscopic (mechanical, electrical) properties of the oxides. The study of oxides by EXAFS technique allowed to progress in the understanding of the inherent properties of defects and interfaces.

The study of the continuous background under electron beam has made it possible to study the origin of some structures and to link them to observations made by exoelectronic emission measurements. Measurements of exo-electron emissions by mixed oxides of barium, copper and yttrium (YBaCuO or YBCO) have shown that this activity results from stress relaxation with propagation of cracks in these materials.

You begin in surface analysis, there are still many original things to do starting from reproducible experimental facts that are explained in a coherent way.

Do not hesitate to contact me if you want to deepen these questions.

Claude Le Gressus