While doing the RF magnetron sputtering some reflected power develops which needs to make null in order to get the plasma.
I am facing continuous rise of the reflected power within few minutes even after I make it zero.
What might be the possible reasons and solution of it?
there could be few reasons which affects the impedence and so power reflection:
your pressure may not be verymuch stable.
after sputtering the first layer of your target, surface resistance changes.
distance between your sheild and target at magnetron
try to keep magnetron shutter open while striking because shutter is generally grounded
Are you using automatic matching network ?
Regards,
Sagar
Thank you so much for your helpful reply.We don't have auto-matching. That's why we need to check and make the reflected power zero frequently. In our case the pressure is very much stable ~25 mTorr.
In general we have found a decrease in resistance (20-30 KOhm to 2-3 KOhm) of the surface after first deposition (~200 nm) from the new target.
Is there anything else like slight gap in between target pellet and the ring holding it around magnet , thickness of the target pellet.
Also I would like to know the basic reasons behind the origin of the reflected power. Can you provide me any theory or ref. if possible please??
Thanks.
Hi Arindam,
I am also not an RF person but you can easily google about RF power matching to find out.
when the impedance between your plasma source, RF generator do not match or your matching network is unable to couple the power it relflects the power.
But I use an automatic matching network and RF source from Advance Energy and so it must be keep changing the impedance automatically.
we did sputtering without automatic matching network for Cu sputtering just to test matching network made by our engineer but found may need to change it a little frequently as you are doing. May be this will be difficult for ZnO like targets as it will change the sheet resistance of the target after some sputtering while for Cu-metal it will not change. Apart from that pressure also plays a big role in deciding the plasma impedance of your sputtering device.
Gap between target surface and shield should not be more than 1mm.
Are you using capacitance gauge for pressure measurement during process ? if that do not change means it's stable. other gauges are not that sensitive.
anyway I think best solution is automatic matching network.
We have Ion gauge which is good enough i think. Also during deposition pressure is maintained by MKS pressure controller.
We had the automatic matching which dies not work anymore so we are forced to use manual.
Ion gauge is OK but not good for depositoin processes because it changes with contaminations..... best is capacitance gauge....
Any way best of luck..... I think that's all I know about it.
The effect of reflected power has on the plasma depends a LOT on what phase it is at.
Try varying the coax cable length between the generator and the matching network. One wants the phase condition such that when the power is reduce slightly the system goes toward a match. Note that plasma loads are not linear.
The origin of the reflected power is that the impedance of your feed line doesn't match the impedance of your chamber. Basically. Your generator has 50Ohm impedance and your job (since you don't have auto-matching) is to configure your matchbox (if not automatic then manual, hope you have it) so it matches that 50Ohm of your generator to the XXOhm of your chamber. This is typically done by adjusting two capacitances inside the matchbox.
You can find some description e.g. in the link below.
Hope it helps.
https://www.engr.sjsu.edu/rkwok/Engr297/Val_Impedance%20Matching%20and%20Matching%20Networks.pdf
Hi Arindam,
I have the same problem, I am sputtering ZnO with Rf and I can't have the plasma because the reflected power, if you had solved your problem please let me know your experience.
thank you
Hi,
We couldn't find any permanent solution. If you have financial support you can look for an auto matching network which would help.
One more thing, you may try out some other oxide target with different thickness and as well as meta target to confirm that your manual matching network is not at all working.
actually my question is we have a thick copper layer stick to the ZnO layer, could that Create a problem for us?
we have a matching box and it works for some targets but I don't think it is reliable.
thank you
huge difference. Thus if one uses a small amount of reflected power at a phase and thus a cable length such that when the power is reduced it go toward a match (0 reflected power) then the plasma is stabilized. see also my US patent on this subject. Hope this helps
The argon probably increases the plasma density and thus increase the wall sheath capacitance that you are matching into. It is possible that your matching network is running out of range, i.e. needs less inductance.
Again if the phase is such that the match gets better when the power is reduced, reflected power IS NOT BAD! The phase can be changed by changing the cable length between the generator and the plasma matching network.
- Badges
- Science topic
- Similar topics
- Philosophy
- Philosophy Of Science
- Reasoning