I tried do an XPS scan of a silicon wafer covered with gold, however, the spectrum does not show any gold peaks. I tried to vary several parameters, including Epass, sample position, and even the voltage and current of the filament as well as the emission current. However, the spectrum remains the same, only the intensity changes.
What is causing such a difference in the spectrum? I am running the scans on SPECS machine.
Attached are the spectra:
Au Blank Good is a good survey spectrum of gold. Although it has some C and O impurities, the Au peaks are visible.
Au Blank Survey is a scan that was carried out on the SPECS machine and the no Au peaks are showing.
Dear Maria, either you are not aiming at the sample or your sample is dirt. Your survey shows mainly carbon, assuming you are actually aiming at the sample you are detecting a typical carbon contamination overlayer. It could be just carbon adsorption from the atmosphere or more tick dirt. You should remove surface contamination and you will see Au. How did you clean the Au-coated silicon wafer? I suggest a quick plasma cleaning
Dear Maria, I am not certain how the sample stage works on your instrument, but I agree with Marco, that you may be misaligned and missing your sample, or the sample is dirty (perhaps having a layer of fingerprints or dust).
It also appears that the sample may not be grounded properly. I am not sure about how the Gold standard was mounted, but it will easily ground through the instrument and no charging will occur. When switching from the Au to the Au/Si sample, did you make sure to use charge compensation (e.g. electron flood gun or Ar+ ion gun)? Unless etched, your Si wafer has a SiO2 dielectic layer that will inhibit adequate charge compensation (I also often see similar spectra where the sample is dirty and requires cleaning and better charge compensation). While this will not make the Au show up magically, it will make the spectra more reproducible, reduce much of the noise, and allow you to get a good sense of what is in the sample. If you are able to plasma clean, that would be optimal, however, depending on the depth of the Au layer and the importance of the Au layer depth, you may be able to gently sputter the contamination away, however, I would not recommend using a FIB or C60 sputter gun for cleaning.
It may be useful to know more about how your sample was prepared (VLD, etc.), handled, and mounted (clip/tape/other).
I agree with Joshua your sample is definetly not grounded, a clue is the 0 intensity on your graph (see image), usually a carbon tape is put on the sample to make contact with the ground. or if you did grounded your sample that means either that you have a fairly thick layer of "dirt" or that you are not aiming at your sample.
Did you tried a prolonged sputter cleaning with Ar+? Since the total (maximum) count rate seems to be similar to the "good" sample, i would say that the spectrometer works fine. Since there are smaller peaks associated to gold 4f (below ca. 100 eV) and 4d around 300+ eV EB, I think that gold is present on the sample, but covered by contaminations (especially carbon) as also discussed by Marco, Joshua and Zakaria! Good luck, Dirk
First of all ground your Au/SiOx sample - just touch a bit of wire or Al foil to the edge of teh sample and then teh sample holder - that will form an earth for you. However, rough measurements sugegst that the zero inensity people have highlighted is actually beyonfd the fermi level - so just noise. Ceratinly crude measuremetns of the scale suggest this is well below 0 eV so we would expect nothing
There is also significant carbon on the surface of the bad sample, so if you want a clean Au/SiO2 you will need to gently etch this (Argon clusters are best if have access to them so dont damage or roughen the gold film to much). That said, I do think your alignment is out somewhat also
I note the spectra are from two different machines too - the "bad" one form a system wish a PHOIBOS Analyser and the "god" sample from a Kratos Axis system. The parameters will not be identical between them, so be careful how you translate things.
Thanks everyone for your help! I really appreciate.
- Marco Morra : The Au/SiOx wafer was cleaned a month before doing these scans, as I was abroad for XPS training during which the 'good' spectrum was produced. The wafer was cleaned by sonicating it for 5-10 minutes each: first with hexane, then with acetone and then with ethanol. The 'bad' scans were obtained in our lab and these were carried out a month after the training was done. The fact that I did not clean (again) the wafer as described above was a big mistake, so first thing that I will try is to clean it by sonication and repeat the scans.
- Joshua Wallace a Wallace: The Au/SiOx wafer is NOT a gold standard. I still have to buy those to calibrate the machine. It is just a silicon/silicon oxide wafer coated with gold and some if it was used to deposit my sample/s. However, I wanted to carry out scans of this wafer so as to check if the machine at our lab is working properly or not. I will also try to ground the sample, although I am not sure how this is done as I am using an RHK Beetle sample holder (see pics attached).
- Zakaria Salmi Zakaric Salmi: how can I put C tape if am using a RHK Beetle sample holder please? See pics attached.
- Dirk Luetzenkirchen-Hecht : No I did not do any sputtering. I might try this after I clean well the wafer in solvents and sonication.
- David J. Morgan : Which kind of parameters please? Are they the calibration parameters? Yes, the good spectrum was obtained from a Kratos machine and the bad one from a SPECS machine with a Phoibos analyser system. The latter I am trying to use for my studies, however,how can I know if the parameter on the SPECS/Phoibos machine is correct or not? This is the reason why I wanted to carry out scans of the same wafer on the SPECS/Phoibos machine - to check if I manage to obtain the same or at least similar spectrum to the ones obtained from the Kratos machine.
- Badges
- Science topic
- Similar topics
- Chemistry
- Nanotechnology
- Nanostructures
- Thin Films