>
> Dear Colleagues
>
>         I have an adhesion problem of a think polyimide called PROBIMIDE 7020.
> This polyimide supposes to have an adhesion promotor included. But I
> could not get the polymide stick well to silicon or silicon oxide
> substrate. Has anyone encounter the problem?
...


Dear Francis

We are also working with PROBIMIDE 7020 and have encountered this
adhesion problem with thicknesses of 200micron or more. Current
applications require thicknesses of 180micron with minimal lateral
feature size of 20microns (planar coils). Coated on sputtered titanium.
Herefore we have found a solution for our adhesion problem.

In fact the internal adhesion promotor in the resist does a sufficient
job.

I have found softbake parameters to be far more important than the
introduction of an extra adhesion promotor (such as APTES: AminoPro-
pylTriEthoxySilan).

A combination of a contact-bake on a hotplate (60degC / 5min) followed
by an (thickness-dependend!) forced air oven step could be the soulution
for your problems too.
Remember: to few softbake is as bad as too much.

If you don't drive the resist-solvents out in the right manner, they may
be trapped in areas near the surface thus resulting in weak chemical/
mechanical strength. While developing , the resist tends to be washed
away.


To give detailed hints, I need to know:

 - the thickness of the resistlayer you are aiming at.
 - Aspect ratio   (feature size)
 - Geometry (deep holes or freestanding resist piles or bars)


At least as important is the surface-preparation (dehydration) prior
to coating. Remember, no water is allowed on the surface.

I would recommend to optimize softbake first. If this step is ok, but
adhesion is still not good enough, investigate promotors such as APTES.


If you wish further help, don't hesitate to ask.


--

Roger Bischofberger