Etching low stress SiN in MRC

Swanson, Paul PSwanson at
Mon Jul 26 14:48:36 PDT 2004

If you use the quartz cubes to center your wafer in the MRC, you will have
non-uniform etch rates in proximity to the cubes.  It is better to use wafer
shards of the same thickness as your wafer for centering purposes.

Paul Swanson, Ph.D.
Principal Engineer
10398 Pacific Center Ct.
San Diego, CA  92121
pswanson at
phone:  (858) 410-4625
fax:      (858) 410-4848
-----Original Message-----
From: Jim McVittie [mailto:mcvittie at] 
Sent: Monday, July 26, 2004 2:30 PM
To: Venkatesh Hanumant Rao Chembrolu
Cc: labmembers at
Subject: Re: Etching low stress SiN in MRC


In looking at my notes for the AMT etch processes, I can say the following
for the
MRC etcher. In etching Nitride with CHF3/O2, the nitride etch rate should
near linearly with O2 flow until the O2 flow gets to about 50% of the CHF3
flow. The
oxide etch rate will also increase with O2 flow but at a slower rate. Thus
selectivity of  nitride to oxide etching increases with O2 flow. The
selectivity to
Si and PR will decrese with O2 flow.  In the AMT etcher, nitride etch rate
with total flow up to a flow of 200 sccm. I do not know the flow at which
the etch
rate peaks in the MRC but it will be much less than then 200 sccm valve for
AMT.  The nitride etch rate should increase a bit less than linear with rf
however resist burning tends to be a problem at higher rf power levels.

Hope this is useful.

Venkatesh Hanumant Rao Chembrolu wrote:

> Hi,
> I am trying to etch 300nm of low stress silicon nitride in MRC (this is
for a 5"
> wafer which doesnot fit into Drytek 1) with 1.6um photoresist as the etch
> I am thinking of using the nitride etch recipe for MRC (Oxygen 3sccm,
> 15sccm, 50mT, 50W) and would like to know if someone has done anything
> before. Thanks.
> Venkatesh

More information about the labmembers mailing list