Hi,

first of all, there appears to be confusion among replies as to
whether the lines of resist or lines of metal were being patterned. I
understand the you pattern lines of resist that are 250 nm thick, so
the final pattern is expected to be mostly metal with thin gaps
(lines) in it.

I think the most immediate problem here is that the film is almost as
thick and the width of the line.  The thickness of in the resist
(250nm) is workable, (though ticker would be preferred), however when
the gap in the pattern has an aspect ratio of 1 or more (film
thickness/gap width) the liftoff becomes very tricky.

If possible, the first thing I would try would be to use ultrasonic
agitation. That might get most of the film to lift off, but with gaps
that narrow, you'll probably have a lot of gold pieces ("doggy ears")
still attached to the main film crossing the gap and creating
electrical contact. One way to deal with that is to use a brittle
underlayer that will not bend and hence will break off completely
during liftoff. One such layer is ITO and I've had very good results
with it myself, though it is not commonly used as an adhesive layer
for gold so I can't vouch for it being suitable to your application.

Overall, I think dry etching is superior for creating patterns with
this kind of aspect ratio. Of course, plasma etching of gold is not
exactly a painless process and will likely damage other components on
the sample.

-mikas

On Wed, Mar 10, 2010 at 4:44 PM, li shifeng  wrote:
> Hi, All,
>
> I pattened 100nm lines on the glass substrate. The height of the line is
around 250nm. I used an e-beam evaportor to deposit 5nm Ti and 80nm Au. I found
it is very hard to lift off. What are the possible reasons?  Any suggestions to
succeed with an Au lift off process for smaller patterns?
>
> Thanks!
>
> Shifeng