I would simply bleed nitrogen into the chamber and bring up the pressure
to result in a mean free path shorter than your source-substrate distance.
 The more multiples of mean free path to source-substrate distance, the
more conformal your coating will be.  Gold is relatively inert, therefore
having nitrogen in the background won't contaminate the film.

Rotating the substrate as you already do will help as well during this
high)er) pressure evaporation.

Best of luck,

Paul



> Hello all,
>
> I am trying to uniformly cover the top, bottom and sidewalls of a ridge.
> The ridge has vertical sidewalls and it would be around 1.5-3um tall and
> about 2-50um wide. The only equipment that we can use for this is an
> e-beam metal deposition system (i.e, no sputtering, no electroplating).
>
> We have fitted the e-beam system with a rocking stage to facilitate
> uniform coverage of this ridge.
>
> The rocking motion can be programmed to do virtually any sequence
> we want. We have the stage programmed right now for rocking the
> sample to +/-45o with a stop (wait time) at both ends of 1sec.
>
> Although the metal covers the ridge, shadowing effects prevent
> the metal from covering it uniformly all around -- especially near
> corners. Clearly, this simple sequence is not the best. Since gold
> is quite expensive and these experiments require about 1-2um of
> gold, I was wondering if any of you have done this and is willing to
> share this knowledge with the group.
>
> Finally, do you know of any software that could simulate this type
> of process?
>
> Thanks in advance!
> P. Barrios

=======================================
Paul Sunal, PhD
Senior MEMS Engineer

MEMS and Nanotechnology Exchange / CNRI
1895 Preston White Dr Suite #100
Reston, VA  20191

Ph: 703.262.5323
Fax: 703.262.5367
=======================================