The issue with long term photoresist storage would be photospeed, but
resist that is even a few years old should dissolve readily in acetone.
If you could spin it, it was dissolved in solvents much weaker than
acetone.

A couple things could have happened.

1. In your process sequence you could have heated it to the extent that
it is no longer Novolak resin or other type resin, and has been
thermoset. That is, chemically converted into a piece of plastic.
Novolaks can undergo two condensation reactions, one at 130 if I
remember correctly and another at 160 or 170. It has been twenty years
since I read the paper. It then becomes a plastic. You can also char
resist. Sulfuric/Peroxide baths stripping heavily implanted resist will
sometimes char, rather than oxidize the resist.

Vacuum processing acts to drive these condensation reactions at lower
temperatures. You likely deposited your Zr02 under a vacuum deposition
condition.

HOWEVER, I doubt that you overbaked it enough even with the vacuum, to
cause your problem. I think your problem is a lack of a negative slope
on any portion of your sidewalls.

The essential requirement of an undercut process is that the solvent
reach the photo resist. With e-beam evaporation machines the step
coverage is usually poor enough that there is some exposed resist.
However, if you have good step coverage, your resist will be encased.

So I am guessing that the acetone can't reach your resist. Let me
suggest the following:

1. The GaAs industry uses lift off a lot. One method is to spin the
resist, then soak it in chloroform or carbon tetrachloride (check the
literature, I can't remember which one.) This extracts some of the
smaller molecular fractions of the resist from the surface.

When you do photomasking, the resist profile has an undercut, since this
process makes the resist at the surface less soluable. This causes very
poor step coverage, and you have a point for the solvent to breach into
the resist.

You need to have a resist process with a negative slope along some
portion of the sidewalls. There are probably other methods of doing this
and the GaAs literature probably has them.

2. You might try violent agitation. If the step coverage is very thin at
a point the stress of vicous flow of the acetone might breach it. You
might dispense acetone on the wafer in a spin bowl. This could be very
messy though as your lifted layers go into the bowl. I would worry about
the ZrO2 scratching other layers, but it may not be a problem at all.
Also, we are tearing a layer of ZrO2. (Low probability of success, I
only suggest it to save your already processed wafers.)

3. The other is a wet etch back of the ZrO2 to breach the coverage at
the sidewalls where it will be the thinest. (Loss of ZrO2, will result
in poor process control)

4. Lift of depends on poor step coverage. If you have a choice do e-beam
evaporation. Disable the planetary motion. If you are doing sputter
deposition, turn off the heat entirely to get poor step coverage.
Thicken your resist to make a taller sidewall for poorer step coverage.

Ed





-----Original Message-----
From: mems-talk-bounces@memsnet.org
[mailto:mems-talk-bounces@memsnet.org] On Behalf Of Xiaoyuan Lou
Sent: Friday, October 19, 2007 2:17 PM
To: mems-talk@memsnet.org
Subject: [mems-talk] Question about lift-off positive photoresist which
isput there for long time

Hi, all

I have a sample with photoresist on which is put aside for quite a long
time (say half a year). The photoresist is "lift-off photoresist
LOR-10B" and "positive photoresist S1813".

After I sputtered ZrO2 on it, I can not lift them off by acetone or
develop. It looks like too strong. Does anyone here know why? Does long
time storage cause any problem?

How can I dissolve this? should I use stronger solvent?