Steven,

As Gareth noted, you are probably insufficiently baking.  The baking parameters
as mentioned in the MicroChem data sheets are just guidelines, and do not take
into account local conditions (variations in humidity, air pressure, etc.)
Additionally, their recommended process flow includes the use of edge-bead
removal, which is not an easy possibility for your substrates.

You are most likely sticking at the edge bead (examine the edge bead closely
after removal from the mask; you will probably see "flattened" areas where the
mask pressed in, or even areas that looked like pulled taffy, where the material
adhered and then "pulled" when the mask lifted).

This means, in all likelihood, that the edge bead is the area with the majority
of retained solvent.  If you cannot remove your edge bead, try to use bake
parameters on your film as if it were a thickness equal to the height of your
edge bead (which, at the thickness you are spinning, may be as high as 80 µm
from the surface of your substrate). It is still best to use a hotplate in this
case, as use of an oven has been seen to produce a non-diffusion layer ("skin"
or "crust") on the surface of your substrate, trapping solvent within the film.
Use of vacuum in conjunction with a hotplate can work well, however, as the
vacuum will effectively increase the removal of cyclopentanone (CP) from the
film.

As Gareth also mentioned, the edge bead can be managed to some extent using a
"relaxation" step by letting the wafer sit post spin on a level surface - you
can also do this on a hotplate, as unexposed SU-8 has a glass transition
temperature (Tg) of 55ºC - if you let it sit for an extended period of time
(perhaps an hour, or even the several hours he mentioned) at temperatures of 65
- 95'C, this may actually solve both problems simultaneously, although you will
have to be cautious of in-film stress which may lead to cracking.

Best Regards,
Chad Brubaker

EV Group       invent * innovate * implement
Senior Process Engineer - Technology - Tel:  480.727.9635, Fax:  480.727.9700
e-mail: c.brubaker@EVGroup.com, www.EVGroup.com



-----Original Message-----
From: mems-talk-bounces@memsnet.org [mailto:mems-talk-bounces@memsnet.org] On
Behalf Of Gareth Jenkins
Sent: Wednesday, January 02, 2008 10:06 AM
To: General MEMS discussion
Subject: Re: [mems-talk] SU-8 2050 Align Problem

Either your softbake is too short or you are heating the SU-8 during
exposure. Try having gaps in your exposure to allow the SU-8 to cool
(e.g. 10s exposure followed by 10s gap and repeat until you have the
required time). You should also have a long pass filter to reduce
excessive exposure to deep UV.
Leaving your sample to rest on a leveled surface for several hours
after spinning (before softbake) may help reduce the edge bead. You
could also use a cleanroom wipe soaked in acetone to remove the edge
bead (though may be difficult with an irregular sample).