Does anyone know of a good reactive sputter recipe for yielding TiO2 optical thin films with low extinction coefficient, small grain size and low surface roughness? I have some papers but most of them use a biased substrate (something I can't do in my tool).
I've been using RF sputtering at room temp, 540W and a 3"Ø x 0.25" thick pure Ti target which I pre-condition with Argon to sputter off any native oxide. I keep finding that even with low O2 flows, I get target poisoning no matter what. The target clearly comes out oxidized and my deposition rate gets choked to about 2nm/min. My throw distance is restricted to about 5 inches.
I've not yet tried DC sputtering but most of what I'm seeing online is pointing me in that direction. Again I'd like to stress that I need low optical loss above anything else.
Any advice is welcome, thanks!
Regarding target poisoning, see https://www.researchgate.net/post/Is_it_possible_to_revive_a_metallic_target_that_was_poisoned_during_reactive_sputtering_Any_advice_or_experience
2nm/min is actually not that bad, by the way. It's normal that oxide sputtering can be a bit slower.
As a valve metal, you cannot avoid poisoning of the Ti target, in fact we have observed target poisoning in Ti target employing Ar gas only, impurities in the gas supply were sufficient for that. In fact, you have to spend lot of work to sputter metallic Ti … and I agree with Anthony Manni
This is still a hot topic in sputtered oxide thin films community. One recent advice is the use of bipolar sputtering. It is basically a pulsed sputtering method with two types of pulses; a negative pulse is followed with a positive one. It helps a lot eliminating or at least reducing target poisoning. So, what you need is to think about DC, a Ti target instead of TiO2, and a power supply with such ability. In addition to being a recommended remedy for target poisoning, it also opens lot of opportunities for tuning films properties.
Dirk Luetzenkirchen-Hecht thank you for your insight, this is very useful information! I'm glad it's the metal itself and I'm not simply doing something wrong.
Babak Bakhit thank you for your recommendation. I've seen this bipolar sputtering technique show up in recent HiPIMS literature. The prospect of doing this with a DC cathode and special power supply is very interesting. I'm inspired to recruit assistance from ECE majors at my school.
Jürgen Weippert thank you for the reference on target revival. I have been able to recover the target without issue by first scrubbing it with Scotchbrite in water, rinsing with IPA and then sputtering it in an argon RIE plasma, and of course pre-sputtering in-situ before coating since it slightly re-oxidizes in atmosphere after the RIE step.
I know that answer by simple equipment. Please refer my patents as opened information using MF monopolar power supply and Ti metal target.
Dear Anthony,
The best way is to work with TiOx ceramic (slightly understoechiometric) sputtering target(s). They have even enough conductivity to sputter using DC. Of course, those are usually available as cylindrical rotatable targets only. In most processes, MF-AC is used with twin rotatable cathodes. Only very small amounts of oxygen need to be added in order to end up with fully transparent (low extinction coefficient) high refractive index TiO2 coatings (of the order of 0.5% compared to the inert gas flow). Several manufacturers offer TiOx ceramic targets, a.o. Soleras Advanced Coatings and GfE Fremat.
- Badges
- Science topic
- Similar topics
- Chemistry, Organic
- Organic