There is a whole laundry list of possible causes.

Check that you don't have resist peeling. The resist peeled layer is
thin, but it is a symptom of crusting and under a crust is a quick to
dissolve high solvent resist layer in the resist. Post exposure bake
helps. Excessive exhaust in the spin bowl causes crusting and under the
crust a solvent rich layer.

If your are doing a hard bake after your develop, perhaps you are losing
height due to the shrinkage of your material due to further desolvation
or condensation reactions.

You measure the thickness after spin, but it should further shrink due
to post exposure bake (slight shrinkage), develop (modest shrinkage),
and then hard bake (significant shrinkage). With a 4 micron spin your
resist might have a lot of solvent left in it and prone to more hard
bake shrinkage than usual.

Another thing to consider is the non-imaged light in your exposure tool.
You might effectively some general scattered light and the resist making
up the pillars is actually getting a partial expose. I am not sure what
you can do about that. I am assuming that you are using anti-reflective
chrome mask, have the exposure tool set up correctly.

The substrate can scatter light and this also results in the "unexposed"
areas getting an appreciable exposure.

However, you could consider dyed resist which would suppress scattered
light from the substrate. Alternatively you could consider some bilayer
process, but that gets complicated.

I would just spin the resist thicker so that after the shrinkage you get
4 microns. Though I would consider if 2.5 microns won't adqueately block
the etch. A thicker resist is more difficult to image and you might have
pillar collapse, depending on their width.

Another phenomenon can be changes in shape in hard bake as the surface
changes due to surface tension or flow of the resist. Probably not the
cause.

Under softbaked resist, and with 4 microns I would be considered that it
is baked right, can have high resist losses during develop.

Deep UV heat baking (Fusion) can shrink significantly and wrinkle
resist.

You can call your resist vendor. AZ is a fairly knowledgeable company
and you can get real help and advice form them.

Good luck

Ed


-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Sudesh Bhagwat
Sent: Thursday, December 20, 2007 10:01 PM
To: [email protected]
Subject: [mems-talk] AZ CTP 100 Photoresist

Hello All,
I have been using Clariant Photoresist AZ CTP 100 (27cp) to construct
what is called as a pillar structure in OLED displays. I have a peculiar
observation. I coat ~ 4 micron of photoresist on the substrate. After
exposing and developing the photoresist only 2.5 micron of photoresist
remains in the structure and the rest approx 30% of the photoresist is
lost.
So instead of 4 micron pillar I have only a 2.5 micron pillar. I have
checked the thickness (after exposing and developing) using a Dektak
Profilometer and confirmed the observation using SEM. I fail to
understant this behaviour. Has anyone used this particular photoresist
and has the same observation.