Dear all,

I am trying to develop a MEMs process using SOI wafers. I am making
membranes successfully using a TMAH wet etch but am having problems
stripping the protective SiNx coating from the device layer once the
membrane is fabricated. Once the membranes are fabricated I remove the
Buried Oxide completely with a wet etch - at this point the membranes
appear unstressed (looking at them by eye). I then use a dry CF4 plasma
etch to remove the nitride - this will obviously etch the Si too - I try to
time this etch carefully but there is always some etching of the Si
underneath. After this step the membranes appear to be significantly
stressed. Is this expected? I can't see how the dry etch could introduce
stress in the membrane but I am ready to be corrected if anyone else knows
otherwise... Is this problem likely to be solved by using a wet etch (hot
phosphoric acid) or do I need to look for an alternative cause for the
stress?

Any help will be much appreciated.

James Ransley
jhtr2 (at) cam.ac.uk