Hi MEMS community,
4 - 5 angstrom thickness is an illusion. Below some 40 - 60 angstrom of
thickness of sputtered gold, you do not have uniform  layer anyway, you
will get just isolated islands.
Beside this, I cannot imagine the precise and SIMPLE method how to
determine 4 - 5 angstrom thickness. Adhesion layer is a good solution
unless
you want to use gold for electrochemical sensors where you
cannot use any metal adhesion layer.
Pavel


On Wed, 13 May 1998, jonathan d trumbull wrote:

> Robert,
>
> > I am trying to electroless plate nickel onto a 4 to 5 angstrom thick layer
> > of gold that was sputtered onto a silicon wafer covered with a layer of
> > silicon dioxide.  The nickel begins to adhere to the gold alright, but then
> > the gold layer curls up and peals off the silicon dioxide layer.
>
> 4-5 angstrom-that's only a few molecules thick?  I can't imagine being
> able to see this peel off.
>
> In any case, you need to have an adhesion layer between the Au and SiO2.
> I have used Cr and Ti films of various thicknesses and ~25-100 angstroms
> seems to work fine.  Don't break vacuum, however, between steps or the
> gold will peel off the adhesion layer.  I am not sure if this is because
> of contamination or slight oxidation of the Cr film (haven't tried with
> Ti). But a 200 C bakeout doesn't seem to help.
>
> YMMV--Have fun.
>
> --Jonathan
>
>