[labnetwork] Si RIE etch

[Bernard Alamariu]

Hello,
CF4/O2 ratio is too high, I think.
I suggest to do a run without O2 first, just to check it.
Then try it with no more than 1-2% O2; however I see you are using very 
very low flow rate for O2
and making it even low you might reach the MFC accuracy limits. I 
suggest to increase accordingly
the CF4 flow rate to allow 1-2% O2 flow, and try to maintain the 
pressure value.

Some comments and suggestions:
The CF4 plasma etch was the first one used to etch SiO2, Si3N4 and Si 
more than 40 years ago.
The German LFE barrel reactor machine was one of them.
The main problem was its selectivity to Si, which was solved by adding 
H2 and He for safety reasons.
Then was the etch rate which was not stable; it was discovered that the 
O2 desorbed from the Quartz walls
was the culprit. The chamber process history counted.
Then it has been established that adding ~ 1-2% of Oxygen increases the 
etch rate and makes it
repeatable, while adding more than 3% quenches the the etch rate.
Practically an even higher etch rate was observed by using industrial O2 
from a tank, which adds
more impurities!
The scientific explanation was based on the so named " Second Order 
Atomic Collision" phenomena.
The main etching specie is CF3- molecular ion.
Its plasma concentration is increased by adding some Oxygen which 
enhances its ionization rate, by
the second order atomic collision. The O2 is brought by RF plasma in a 
long lifetime metastable
energy state higher than the CF4 ionization energy; by inelastic 
collision O2 transfer this extra energy
to CF4 and increases its ionization rate.
The low O2 concentration acts like a ionization catalyzer. More O2 is an 
inhibitor.
There are old papers in the Journal of Physical Chemistry around 1973, etc.
Thanks, Bernard







On 3/7/17 1:14 PM, Jugessur, Aju S wrote:
> Hi,
>
> I have a user who is using the following Si etch recipe on our Oxford 
> NGP80 etcher:
>
> RF power: 100 W
> Pressure: 100 mT
> CF4 flow rate: 25 sccm
> O2 flow rate: 3.1 sccm
>
> The etch rate from the above parameters is ~ 0.15 microns/min.
> The selectivity is 1:3.
>  Can anyone suggest what parameters to tweak to increase the etch rate 
> without causing too much substrate damage or roughness?
>
> Thanks for your suggestions.
>
> Regards
> Aju
>
> Aju Jugessur /Ph.D./
> Director, University of Iowa Microfabrication Facility
> Professor (Adj.), Physics and Astronomy
> OSTC, Iowa Advanced Technology Labs
> University of Iowa
>
> 205 N. Madison St
> Iowa City, IA 52242
> 319 -353-2342
> aju-jugessur at uiowa.edu
> http://ostc.uiowa.edu/uimf
>
>
>
>
> _______________________________________________
> labnetwork mailing list
> labnetwork at mtl.mit.edu
> https://www-mtl.mit.edu/mailman/listinfo.cgi/labnetwork


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mtl.mit.edu/pipermail/labnetwork/attachments/20170308/5e9372fd/attachment.html>