I would like to clarify something on Hard Bake, reflow and UV cure.

1. Reflow and hard bake are not the same thing. Reflow is where the hard
bake is hot enough to cause the resist to flow and the resist features
to adopt a minimum surface. It is a reshaping driven by surface tension
and the resist being softened to the point where the resist will flow.
However, you can have hard bakes where the resist doesn't reflow. Reflow
is a subset of hard bake.

2. Hard bake does have some definition. It is a bake higher than would
be acceptable for a softbake due to loss of photospeed and Photo Active
compound. It is always a post develop bake.

3. The Deep UV curing machines, like the Fusions, would ramp a bake
while doing the UV cure. There isn't a need for bake afterwords. I
suppose there might be some esoteric application where there is, but I
never ran into it.

4. Finally, for wet etching, hard baking only goes so far to help. Some
hard bake will always be better than no hard bake. However, beyond 140
degrees you will rarely see much improvement, and at higher temperatures
you might introduce stress in the resist film to make it try to pull way
from the substrate.

Deep UV curing is to harden resists to withstand plasma etch. Usually
metal plasma etch. I don't see it helping resist adhension.

Higher temperature hardbakes can also make the resist very difficult to
remove. You have to use plasma ashes to take the "skin" off and then
sulfuric peroxide strips.

If you can't get good adhesion at 130 deg. Or 140 deg. C, I would look
at surface treatments. Try HMDS, even if it isn't silicon, silicon
dioxide, or silicon nitride. HMDS removes anything with an OH group, and
often it can solve an adhesion problem even if it isn't a "textbook"
case.

Before working on adhesion enhancements, make sure your substrate is
actually clean, and has no residues from vapors, cleaning solvents,
vacuum ovens, etc. A fab or lab, can have a lot of little devils to do
in your adhesion. For example, spilled machine oil has a small vapor
pressure and you only need a monomolecular layer of junk to have bad
adhesion.

Also, HMDS isn't the only material available for vapor priming. There
are multiple chemicals. I used to use some Aluminum analog to HMDS for
aluminum metal layers. Probably was a waste of time and a superstition
of the previous engineer, but I didn't dare remove it.

Edward H. Sebesta
Independent Semiconductor & MEMS engineer


-----Original Message-----
From: mems-talk-bounces@memsnet.org
[mailto:mems-talk-bounces@memsnet.org] On Behalf Of my2232@columbia.edu
Sent: Wednesday, February 25, 2009 1:39 PM
To: mems-talk@memsnet.org
Subject: [mems-talk] definition of hard bake and hard bake recipes


Hello all,

I was looking for the definition of "hard bake" to enhance
substrate-photoresist interface during wet etch processes. I found
that there is no exact definition for that.

Some of the users simply keep their developed PRs at elevated
temperatures and call this "hard bake" while some other users use
"UV-curing tools" to cure their PRs before keeping their PRs at
elevated temperatures for extended time, or some others use a
convection oven to hard bake their PRs.

 From here I understand that the definition is not very important as
long as the final result is successful... But from my experience, I
know that keeping the PR at elevated temperatures is simply called
"reflow" but not "hard bake".

My question is that:

What recipe should I use to enhance my silicon-PR interface for ~90
minutes of BOE (6:1) etch?

I am thinking about modifying my mask, but before starting that I want
to see what I can do with the example that I have in my hands. This
way it will be much cheaper, and if there is a solution, it will
definitely be much faster than modifying my masks and doing everything
from scratch...

I have several PR options, which are:

SPR220-3.0,
SPR700-1.2L,
SPR955-CM 0.9, or
SPR955-CM 2.0

Is there somebody who used one of those PRs, and achieved a good
adhesion against BOE(6:1) etch for extended etch time (~90 minutes).

Best regards,

Mehmet Yilmaz
Mechanical Engineer