I have not yet tried Al, or Cr yet, but I do have them available.
Could anyone comment on the their experience with these metals and
TiO2 specifically?

The oxygen plasma exposure I believe is not key here, since I observe
the same adhesion problem even when O2 plasma is not used (i.e. Ti is
simply exposed to air for a few hours).  So any kind of etching would
probably not yield good result, since the sample would have to be
exposed to air afterwards and I know that the initial layer of oxide
on Ti forms extremely rapidly.

I also came across some literature(D.M. Mattox, J. Appl. Phys. 37,
3613 (1966)) suggesting that sputtering Au in a 50/50 O2/Ar atmosphere
dramatically increases its adhesion to oxides. Has anyone actually
tried to implement this technique with any success?

Also, in case that might matter, the Ti film I'm dealing with is a 4
nm Ti over a GaAs substrate.  I know the film does not oxidize
completely, because it remains conductive even after gold deposition,
however the thickness of oxide is probably equal to the thickness of
remaining metal.

-Mikas

On Sat, Jun 20, 2009 at 12:14 PM, Alex Mellnik wrote:
> Hi Mikas,
>
> We have had similar problems with adhesion between Ti, Au and other
> materials that works well for liftoff, but generally does not survive
> wirebonding.  What other materials do you have in your e-beam system?  I'm
> surprised that the Ti did not work, but aluminum or chrome might also be
> worth a shot.  Finally, is the substrate treated any after the oxygen
> plasma?  A mild acid etch, if possible, might work.