Is the soft bake a hot plate type of bake.  I am not familiar with the
sensitivity of SU8 but standard positive and negative resist started to
react after 90 C.  It could be too long at a temperature over 90
affecting the lowest layers.  A possible lower temperature vacuum
assisted soft bake would work.  Bill Moffat

________________________________

From: mems-talk-bounces@memsnet.org on behalf of John Hilton
Sent: Fri 11/14/2008 11:23 AM
To: Oakes Garrett
Cc: General MEMS discussion
Subject: Re: [mems-talk] Blurring of SU-8 Layers upon development



Garrett:

Its not a shift between exposures, because some of the features are on
the bottom layer only and they experience the same blurring effect.  I
had discounted the idea that it was cross-linking between layers due
to, for example, acid generated during exposure of the bottom layer
diffusing to the top layer (which could possibly be fixed by more
baking of the bottom layer), because the blurred portion appears to be
stuck directly onto the silicon wafer.  It is possible that its due to
the second, long soft-bake, although I've not read of that problem in
the literature.  I've attempted to account for this by using a soft
bake that shifts between 65 and 95C - in the hope that by not staying
at 95C for too long it will minimize re-flow.

Also, a single 80s+ exposure tends to introduce warping at the surface
of the mold caused by IR heating.  Thanks for the input.

-JP Hilton