Rob,

Thank you for your kind reply. Sorry that I didn't go into enough details of my
complicate design. Actually, the Si3N4 pattern I want to protect is a bilayer of
sputtered Si3N4 (~0.3um) and LPCVD (~1.3um). The sputtered Si3N4 is temporarily
inevitable in my process design. Here are my sputtering conditions:

Tool: AJA UHV magnetron supttering system
RF power: 150W
Target: Si3N4
Substrate: complicate patterned basically covered by LPCVD Si3N4 (~1.3um)
Substrate temperature: ~15C
Pressure: 2.8 mTorr
Working gases Ar=30sccm, N2=5sccm.
Time: ~200minutes

Also, I want to pattern a layer of BHF/BOE-resistant photoresist on my Si3N4
pattern to be protected. Do you think it will work as I hope? If yes, what PR
would you suggest?

Best regards,

Yu

________________________________________
From: [email protected] [[email protected]] On Behalf Of
Robert MacDonald [[email protected]]
Sent: Thursday, April 16, 2009 1:02 PM
To: General MEMS discussion
Subject: [mems-talk] Selectivity of etching LTO over sputtered silicon  nitride
using BHF/BOE

Yu,

There are more variables at play in the final etch rate of a sputtered
silicon nitride film than deposition rate (for example, the target can
be silicon or silicon nitride).
We used BOE etch rate as one of our metrics for a sputtered nitride film
and changed parameters to achieve the lowest etch rate possible which was
on the order of 5-10 nm/min, much worse than LPCVD nitride, which does
have the slowest etch rate.

If you are doing process design I would recommend finding a source for
LPCVD nitride, if you can work it into your process flow. If you require
a low temperature sputtered process, then by optimizing your processes I
think you can achieve better than 15:1 selectivity.

Rob