In my sputter tool I use an RF etch to clean the wafers of residue and the SiO2 layer (about 30nm) before sputtering onto them.
I did not change any parameters but the etching rate dropped so much that I had to stop the process.
I changed the electrode, checked the MFC and all tool parameters available. I cant see anything that changed.
Does anybody have an idea what the problem might be?
Are the chamber walls covered with anything? A dielectric film there might affect the plasma shape and therefore the etching rate.
Jürgen Weippert : We checked the walls. They are fine. The electric conductivity is good in the chamber.
Curtis Guild
: We changed the plasma electrode to a new one and didnt see any changes.Today we changed the RF generator (just to try) and now the etch rate is perfect. Not really sure why.
When you say you changed the RF generator, was it the "local element" in the chamber or the external power supply?
In those power supply, sometimes potentiometers or other regulators degradate which results in powers or voltages being given out wrongly (dirt -> too much; corrosion -> not enough).
If you do not plan to reuse the old generator, a look inside is always at least interesting.
Jürgen Weippert: it was the external power supply. good idea to open it and see if something looks wrong inside. thanks
Sadly it did not solve the problem. The etching rate is dropping again.
So let's resume: you exchanged the electrode and the generator and the problem comes back.
Have you verified that your conducts are intact? Let's assume that there is some contact somewhere in a cable that should not exist, then the power generator might try to compensate that which might lead to an internal degradation; I haven't seen this for such a system yet but for other components it has happened.
Yes we have measured the conductivity on the cage, on the shields and all the other parts accesible.
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India