Hi Sonic,

The cracking indicates your layer is under high tensile stress. I would
guess that the through-thickness cracking is the combined effect of
over-exposure and/or over-baking (i.e. excess dehydration). Of course, the
fact that you are using expired SU-8 could mean your mixture is more
viscous, but you should notice an upward shift in the thickness vs spin
speed curve if this is the case.

How are you baking - Hotplate or oven? If the former, you should consider
reducing your cool-down time.

The cracks are forming because the surface is under high tensile stress,
which propagates cracks through the layer. So, related to the point above,
if you do not want to reduce your cool-down time, you could try to retain
solvent conc. in the surface layer (reduce evaporation) during the PEB step
by placing a petri dish over the wafer, using glass slide spacers to leave a
gap for ventilation at the base.
The irony is that the extended cool-down time is used to reduce cracking, by
minimising diff. thermal stress gradients within the SU-8 layer and
delaminating shear stresses at the substrate interface. Only apply an
extended cool-down if cracking propensity / interface delamination is
decreased. If there is no improvement, look at reducing exposure dosage and
pre-baking time.

Good luck and let us know how things develop - no pun intended! :)

p.s. as always, in MEMS, and particularly with SU-8, there is no magic
process recipe that applies to everyone. You have to conduct a series of
trials to determine the best conditions that fit your materials, equipment
and application. All you can really get from discussion forums is general
advice to point you in the right direction.

Michael


---------- Forwarded message ----------
Subject: [mems-talk] su8 cracking

Hi, everybody
I tried to coat a 20um su8 mask on glass. Su8-50 is used in this process.
After developing, cracks are observed everywhere on su8. These cracks are
run through the entire su8 layer. It is really wired since I used the same
protocol and get succeed last year.

Here is the procedure I've followed:
Plasma cleaning―bakes 20mins @ 200c
Spin coating
Soft bake 3mins @65c, 15mins @95c, cool down to RT in 45mins
Expose 250mj/cm2
Post expose bake 1mins @65c, 4mins @95c, cool down to RT in 45mins
Develop

Also I noticed that the su8 which I used was expired last July. Is the
expiration the explanation for cracking? Or the incorrect exposure dose? Or
something else?

Thanks

--sonic