To completely bake off water and alcohols you need to go to much higher
temperatures. There is a classic study on this, and the higher
temperature is probably not compatible with your process. It is
something like 300 deg. C or more.

Reviewing your process flow I would make two changes.

1. Use the HMDS to get your adhesion.
2. After patterning and your initial hard bake, give it a plasma descum
and D.I. rinse.

The 2nd step above will restore a hydrophillic surface to the exposed
areas in your resist pattern. I don't know where you need the
hydrophillic surface in your process.

If that doesn't work or is not compatible with your process flow, I
offer these other thoughts which probably are not much use.

Resist is hydrophobic and with a hydrophilic hydrated SiO2 surface it
has a tendency not to stick.

However, you might have a chance if you have a mineral acid/peroxide
clean after your resist strip. Acetone and IPA can have organics
residues which can interfere with adhesion and an acid/peroxide strip
would remove them. It could be that your adhesion problem is due to
this, and you might be able to get by. I am not optimistic though.
Worse, it might work sometimes and not others.

Don't vacuum bake, since you run a real risk of stray organics. Your
plastic boat will outgas all sorts of things. Vacuum chambers often are
not clean and if you don't have a dry pump, you might have pump oils.

Hard baking is probably driving adhesion through the surface OH groups
but is also hardening your resist. You will lose progressively your
ability to dissolve it in any solvent or event Sulfuric/Peroxide if
driven to extremes.

Best of luck.

Ed



-----Original Message-----
From: mems-talk-bounces@memsnet.org
[mailto:mems-talk-bounces@memsnet.org] On Behalf Of Steven Yang
Sent: Monday, November 19, 2007 8:35 PM
To: mems-talk
Subject: [mems-talk] AZ5214E develop problem

hi all
I am using AZ5214E as postive PR for the lift-off process. However, the
PR is keeping strip from sub during the development. My procedure is

(Si sub with 1 micron PECVD SiO2 deposited and pre-cleaned with actone,
IPA, rinse with DI water, dry with N2 flow)

1) Dehydration: hot plate 150C 10 mins
    (no HDMS because I need the sub surface hydrophilic for the later
function)
2) Spin coat: AZ5214E 5000RPM 1000R/s 30s
3) Softbake: 95C 1min
4) Expouse 7s @ 900W (MA8)
5) Develop (AZ developer 30s)
6) Rinse in DI water (30s)
7) Rinse with flow DI water and Dry with N2 flow

The situation is worst when I develop the sample immediately after
expouse, but getting better if I do a hardbake after exposure. But I am
not sure whether this hardbake will make later liftoff difficult or not?

So, I would appreciate if you guys could give some suggestion on my
procedure. thanks a lot in advance!

Regards,
Steven