Failure of resist adhesion can have several causes.

One feature of resist adhesion failure is that it can be a random thing.
One lot will have a problem and another lot won't with the same problem.
Yet there isn't any identifiable process change.

1. No Vapor Prime. You need to do a vapor prime with HMDS, which is
HexaMethylDiSilizane or something like that.  This will give a
consistent reliable resist adhesion.

2. Stray vapors. Lab or fab air can be contaminated with low levels of
hydrocarbons or other vapor that don't react with HMDS. If your wafer
has a monomolecular layer of Xylene or something the HMDS won't work.

These vapors can come from many sources. For example a negative resist
or develop track near your wafers prior to processing. I actually had a
fab that had this problem once.

Spilled oils can have a very low vapor pressure, but enough to cause
problems.

SOLVENT SOAKED wafer boats are notorious for problems. The teflon wafer
boats soak up acids and solvents and then slowly discharge them. They
can't be cleaned up. Once they are solvent or acid rich they never come
completely clean. Throw them out, but if you are on a university budget,
make sure your wafers don't sit in them.

OLD WAFER BOXES can also get contaminated.

3. High humidity, greater than 60% could cause problems. I personally
think you could run okay with vapor prime up to 65% humidity and maybe
more.

4. Pre-treatments. If you have done a HF dip or flouride dip etch back
on oxide you will get poor adhesion. Some prior treatments can leave an
SiO2 surface with poor adhesion properties.

Suggested Process.

1. Sulfuric Peroxide to clean surface.
2. HMDS vapor prime. (You can do spin on HMDS and hot bake if you want,
no preferable.)
3. Resist spin as soon as possible. Keep post Vapor Prime wafers in
clean boat and box.

A good clean up of your lab would be a good thing to do also. University
labs are sometimes a fright.

Ed Sebesta

-----Original Message-----
From: mems-talk-bounces@memsnet.org
[mailto:mems-talk-bounces@memsnet.org] On Behalf Of madhav rao
Sent: Friday, August 15, 2008 12:39 PM
To: mems-talk@memsnet.org
Subject: [mems-talk] Photolithography - Resist peeling

Hi,
       I am experiencing a problem in patterning on silicon-di-oxide
wafers; I am using SHIPLEY 1818 resist to get a thickness of 2.5 micron;
6 seconds of exposure time. After developing for 60 seconds, I observe,
the resist layer peels off; leaving no patterns on the wafer. Also I had
soft-baked the resist at 95 C for 2 minutes in hot-plate.
This seems strange to me, as a month ago, I had successfully transferred
the patterns on to silicon-di-oxide wafers.
I am using mylar film masks.
Please let me know, if any one had experienced similar problems before;
and if so, could you let me know suggestions to solve ?