I use reactive magnetron sputtering system to deposit TiO thin films. Specifically, I use 3 Ti targets, 8 mTorr of Ar with O2(20/3.5). However, if I use 3 sccm of O2, I obtain Ti and if I use 3.5 sccm of O2, I obtain TiO2. In the last case the targets change color to blue indicating oxidizing of them.
Before I tried to deposit TiO films using two targets and 2,2.2,2.5 and 3 sccm of Oxygen but never I obtained TiO, just Ti or TiO2.
Moreover, a year ago I managed to deposit TiO reproducible using 3 sccm of Oxygen and 3 targets.
I pre-sputter for 10 min and the pressure always drops during this process. As I understand, when the pressure drops, targets start to catch oxygen.
Is it correct? Then how am I supposed to reproducible deposit TiO films if the initial state of the targets are always different?
Hi Mikhail, As Andrei mentioned, there is likely to be target poisoning. If the targets have an extra coat of oxide layer on its top, it may result in varying stoichiometry. My suggestion is to do a target preclean (presputter Ti without O2) for longer time. This step should remove any oxide layer on top of the targets and help you to deposit the material of desired composition.
You mentioned, you will presputter for 10 min and the pressure drops down. Is that with or without oxygen? Usually, chamber pressure increases on gas introduction.
Presputtering Ti target without O2 for 10-15 minutes can be a good choice to reduce oxidation, and then you can carry out your main experiment.
Hi Mikhail, you state that 'the pressure drops when you start sputtering' which makes me think you do not have sufficiently low base pressure. Thus, when you strike your discharge the titanium is initially acting as a titanium getter pump.
You should be in the low 10^-8 Torr range. Commercial systems (such as AMAT Endura) have been hitting this spec since 1990. This then dramatically lowers the chamber background water-vapor transport. (This has been an uncontrolled variable for you so far.) Note, this just (*dramatically*) reduces the water vapor, it does not eliminate it!. Then when you do your sputter-clean of the target, you will clean off the native oxide *and* getter your chamber, locking (most) the residual water-flux from your plasma volume. The effectiveness of this gettering is dependent on the fraction of hte chamber that is coated by the Ti sputter -- if you have lots of baffling with your low gas flows the water transport from the 'outside' regions back into the plasma volume will be substantial.
Then consider what fractionof your oxygen flow you are needing to incorporate into your film. You mention flows and numbers of targets, but power, pressure and substrate size as well as film dep-rate are all part of the system. Make sure your oxygen is not being consumed in the perimeter of your Ti-flow. Also look at the I-V characteristic of your target as a function of oxygen flow to see if there is a change in the target-plasma-sheath -- could contribute to the non-linear behavior of film properties versus flow. (I suspect there is, as you note the targets change color.)
Good luck!
Thanks for the answers. I installed a mirror into the chamber to check the condition of the targets in-situ. I have already seen the targets cleaning at 0,1,2 and even 3 sccm of o2. At 3 I saw a small amount of yellow oxide on the targets. At 4 the oxide was blue after 30 min of pre-sputtering already.
I just wonder if the oxygen content in the surfaces of the targets and the films will be the same. It is also not very clear for me if the targets are poisoned during the deposition, how the stoichiometry of the films May be something less than that of the targets. Say, if I got tio2 on the targets, how can I obtain tio from them?
Also, in the initial experiments when I managed to obtain tio, I applied bias to the pedestal. May it affect the results so seriously?
Hello Mikhail, the advice from other sound very good and looks that you able to clean your target. As A.R. Ellingboe mentioned low base pressure is very important for the process and has a huge impact on the film quality as other parameters (Power, substrate-target distance and geometry, DC or RF etc..). It could be helpful if you provide more information
Talking about stoichiometry: the sputtered film can have slightly different stoichiometry as target because some atoms are easier sputtered form the target then others. Moreover, the process gas composition also impacts stoichiometry of the film.
The target should not be poisoned through the deposition process normally. Except you apply very high powers, your base pressure is to low and so you have some contamination in the chamber and also it can oxidize if you use oxygen in the processes gas.
In RF sputtering biasing the substrate should not have to much impact as theoretically the sputtered atoms form target are not charged. Though depending on the substrate-target geometry it could have some impact. If you use DC sputtering biasing is a important parameter.
good luck
- Badges
- Science topic
- Similar topics
- Chemistry
- Nanotechnology
- Nanostructures
- Thin Films