I have also observed parameters of the mask aligner having a great effect on
stiction.  Not only is the leveling/contact force important but also the
parallelism of contact.  If the approach of the substrate into the mask is
not level then a "scrubbing" action can occur which will tear the resist.

Where are you noting the resist is sticking to the mask?  Along the highest
spots?  Leveling of the resist surface is important for making clean
contact.  Minimize the edge bead.

Also, there are "Mask Protection Layers"; a hydrophobic Teflon-like coatings
that really do repel the resist stiction.  I know Bill Moffat of Yield
Engineering mentioned it as something you might consider. Image Technology
[SUSS MicroTec] also offers masks with MPT coating applied.

Good luck,

Suzanne Ericson

-----Original Message-----
From: Michael Larsson [mailto:[email protected]]
Sent: Wednesday, October 22, 2008 3:11 PM
To: [email protected]
Subject: Re: [mems-talk] SU-8 layer stick to mask after exposure

Hi Ruilin,

Well keeping your SU-8 in a small bottle in the fridge is not
detrimental per se, but it could cause problems if you were to open
the still cold bottle in a humid environment. This would cause water
condensation to contaminate your sample. I don't think condensation is
the source of the stiction problem, however. On the contrary,
contamination usually degrades SU-8 / substrate adhesion.

In your case, you have a thick resist layer, with some parts (most
probably the outer rim) thicker, and thus less dry, than others after
the soft bake. Try to incorporate an edge-bead removal program in your
spin profile, or increasepre-bake time or temperature to dry off even
the thickest regions. These are the only things I can suggest.

If you stil get stiction, then play around with the contact mode. Try
proximity instead of vacuum or pressure contact.

p.s. also perform a "pihrana" on the mask plate. It must be pretty gooey by
now.

Good luck.

Regards,

Michael Larsson