Dear Olga,
this sound more like an exposure problem than a 'cleaning' problem. 40um
is 'wide' enough for the developer to work but 1mm is very thick and you
must have a well collimated UV source to achieve the aspect ratio of 25,
which is usually OK for thinner SU8 layers. The problem is most probably
diffraction effect - the diffracted light (and/or the not collimated
light) will expose the bottom of the trenches and you can not develop
it.

For the difference between edge and centre I suspect your mask is
contacting your resist layer only at the edge of your wafer, but not at
the centre (common problem with some mask aligners, the thick SU8 layer
may also help to warp the wafer) which increase the problem here
compared to there...

Check the contact problem (and try to correct it) and enlarge the trench
to verify my suggestion...

Franck

-----Original Message-----
From: Olga Makarova [mailto:makarova@aps.anl.gov]
Sent: jeudi 6 novembre 2003 02:29
To: mems-talk@memsnet.org
Subject: [mems-talk] deep trenches in SU-8


Hello all,
I have a problem in cleaning out 1mm deep and 40um wide trenches in
continuos SU-8 layer.  Trenches that are close to the edges are clean,
while the others contain the residue that I can not remove.  I do
developing in SU-8 developer with magnetic stirring.  The use of
ultrasound damages the structure.  Does anybody know how to improve my
developing process?

Thank you very much,

Olga