The use of image reversal gives a negative slope to the resist image so
the deposited metal does not cover the resist side walls.  This makes is
easy for the solvent to penetrate soften the resist and lift it off with
the metal.  Bill Moffat

-----Original Message-----
From: mems-talk-bounces+bmoffat=yieldengineering.com@memsnet.org
[mailto:mems-talk-bounces+bmoffat=yieldengineering.com@memsnet.org] On
Behalf Of mikas remeika
Sent: Tuesday, March 16, 2010 5:05 PM
To: General MEMS discussion
Subject: Re: [mems-talk] Does ultrasonic damage metal film during
lift-off?

In my experience, no more than 5 minutes of ultrasonic agitation are
required to liftoff most films, and in general no damage occurs even to
very high resolution patterns.  Also, I would start with the lowest
power of the ultrasonic that your bath will allow - its usually
unnecessary to go to high power for liftoff.

Also, generally I'm more concerned with thick films (rather than thin)
being damaged by ultrasonic agitation. This is because you need to be
concerned with the ratio of forces required to tear the film and the
force to peel the film. The force to peel the film in always the same,
but tearing the film requires more force when it is thick and hence
you're more likely to peel it off completely during liftoff.

Another trick regarding lift-off which might be useful - this is
especially useful when you're using SU-8: if after 15 minutes in the
ultrasonics bath no liftoff occurs, you can put the sample into an
oxygen plasma asher (not an RIE system, but an isotropic asher) for some
long time and repeat the liftoff. Usually plasma will etch resist
through any small holes that were formed in the material covering it and
after that liftoff will proceed very quickly.