I agree with Thomas, coating a thin layer of SU8 is a good idea,
however, in my experience I found liquid SU8 to cured SU8 adhesion to be
poor unless the cured SU8 is first treated with oxygen plasma (150 watts
for a minute or two...play around with that yourself for best results).
That was just my experience though.

You should give Microchem a call, I believe they manufacture some sort
of adhesion promoter (I've seen it in our lab but never had to use it
myself).

In addition, I would suggest you give special attention to the way that
you are preparing the glass surface. You should be doing either piranha
cleaning followed by a good hot plate bake (150C for 20 min minimum), or
acetone, IPA, DI water, bake followed by oxygen plasma cleaning. Baking
is important, I found moisture to be a big source of problems for SU8
adhesion (even the humidity in the room can destroy your results).

Lastly, I'm not sure if you're doing this, but when you're curing the
PDMS, don't do it at a very high temperature. I believe Dow's product
sheet recommends times and temperatures that go up to 140C for 20 min or
something like that. I found curing at high temperature caused my SU8 to
come off as well (I'm guessing due to differences in thermal expansion
between the SU8 and glass substrate). So try to stick to 100C or less.

good luck,
Joe Grogan




Thomas Xu wrote:

>Lung-hao,
>
>You may spin coat a thin layer of SU-8 on glass
>substrate first followed by standard soft bake, flood
>expose ,PEB, develop, rinse and dry, hard bake. Then
>do the same process to make your SU-8 mold on it. SU-8
>will have good adhension on SU-8 layer and the first
>SU-8 layer also have good adhension to glass since the
>interfacial area is large.