Hello,
I would like to ask about in situ annealing process in the rf sputtering chamber. I use Ar gas for the thin film growth but I wanna do annealing in an Oxygen atmosphere. After the growing process, I'm gonna just turn RF power off and change from Ar to O2 without changing from turbopump to rotary pump, whether this can lead to a mixture of new O2 and old Ar gas or not? Should I return the sample to its initial condition (turn off rf power and change back to rotary pump) and after that inserting O2?
Hello, Lalisa!
I suppose, the answer to your question depends on the purpose you have.
On one hand, for in situ oxidation without special demands to epitaxy or crystallisation, you may use sputtering in an oxygen atmosphere. In this case, you need only a separated gas inlet in the growth chamber.
On the other hand, if additional ionisation of the O2 is needed, it is better to use assisting source, or a gas mixture in the main RF source. But here you can also use annealing of the substrate in the oxygen atmosphere while sputtering by Ar beam.
Finally, I did not understand your question about pumps. There is always better to use a high vacuum turbopump during the whole growth process. Usually, a rotary pump is used only as forevacuum pump before a growth process, and never in the process independently of gas usage.
PS. All above in most refers to the sputtering by the direct beam from some kind of an ion source, but in general, my advice is just to add oxygen to the chamber and try - it works in most cases.
Hi Lalisa Rm
!Olga Dikaia has given a great answer if you want to oxidize your film during your film deposition (reactive sputtering).
You seem to be talking about doing a post-deposition annealing step in an oxygen atmosphere though, so I'll add a bit more. My recommendation would be to:
1. Complete your sputtering process and shut off the RF power
2. Shut off the argon gas and close all appropriate valves
3. Run the turbo pump (I'm assuming the rotary pump is backing it) and take the chamber back down to vacuum. This will avoid ending up with an argon/oxygen mix.
4. Introduce the oxygen gas (or desired mixture). I recommend using the turbo pump because your chamber pressure is probably controlled via a butterfly or similar valve in front of that pump. There will be a maximum pressure that your turbo pump will be compatible with. Look this up and do not exceed it.
If you need to anneal at higher pressures, then your rotary pump might be OK, but it may be difficult/impossible to control. It's system dependent. You're probably better served by using a dedicated annealing chamber/system. Depending on your sample, annealing in air might also be suitable.
- Andrew
Be careful with a rotary pump an a pure oxygen - this may lead to inflammable gases at the exhaust of the pump ... Be sure that a suited oil is used in the pumping system, take care. And: If you are using sensitive metals (targets) for sputtering, then a separate chamber for the annealing treatment is recommended, otherwise, you will never get all the oxygen out of your system again.
I remember that I have once introduced an Ar-O2-mixture for reactive sputtering in a sputter coater, and it was necessary to replace all the gas-inlet tubes and valves - without doing so, it was impossible to sputter metallic Ta, Al and Fe in the system, due to oxygen remainders always oxides were deposited ... Good luck anyway, Dirk
Lalisa Rm
Exactly - if your aim is to deposit oxides, then an oxygen remainder in your chamber is not a problem. In case you are deposting (complex) oxides, you should additionally think about dosing some oxygen or humidity (water vapour) to the Argon sputter gas - this may improve the stochiometry of the deposits ... Check the literature for this and your material. May I ask what kind of oxide you are interested in? And if I remember right, you should use "Fomblin" oil .in the rotary pumps (see e.g. http://www.testbourne.com/fomblin-pfpe-vacuum-pump-oil) ...
Lalisa Rm
, I think that problem with extra SiO2 oxidation lays in too long annealing or a higher amount of oxygen than needed. In my practice, I have never seen that natural oxide on Si growing when I use a proper amount of O2 in the chamber. In the case of crystallisation, maybe it is worth trying substrate heating while sputtering in combination with the O2 atmosphere.
Lalisa Rm
, in this situation, maybe it makes sense to grow an amorphous oxide layer and post-anneal it in a high vacuum.But I have a suspicion, that additional oxidation of the SiO2 layer is not as simple as it seems. What method did you utilize to study your sample?
And one more question, did you use an oxidated silicon (natural or thermal) substrate or bare (etched) silicon? Because if bare, then it very difficult task to avoid any SiO2 due to its high affinity to oxygen.
Lalisa Rm
, I have never worked with gallium, so cannot answer about the reaction with silicon.Ellipsometry is a good method but provides a classical phase problem (as any other reflectometric method), thus you must know the exact structure to make a proper fit of it. Some depth profiling in combination with an element- and concentration-sensitive method (like Auger electron spectroscopy) will provide more accurate and complementary information about the Si-GaOx interface. TEM cross-section is good too, especially with EDX or EELS mapping.
Nevertheless, if you have no specific demands for interaction between thin film and substrate, you can try other substrates. As far as I can tell from the above, you used a high enough temperature to provoke thermal oxidation of the silicon, so natural oxidized silicon will behave the same as bare. In this case, you can try Si substrates with thermal oxide or other materials, like commercially available MgO, Al2O3, SiO2.
- Badges
- Science topic
- Similar topics
- Electrochemistry
- Battery