We are using a tungsten liner and have reduced spitting by about 95%.  We
also elevate the liner by putting a couple of small pieces of tungsten boat
material in the bottom of the crucible under the liner.  By heating the
liner and the gold the whole gold charge melts and eliminates the spitting.
The spitting occurs at the edge of the melted charge if all the gold doesn't
melt.  Heat slowly and spread the e beam as suggested below and you get
great films.
Milan

-----Original Message-----
From: Jeff Simkins [simkinjr] [mailto:[email protected]]
Sent: Friday, March 26, 2004 11:28 AM
To: 'General MEMS discussion'
Subject: RE: [mems-talk] Metal Spitting


I would like to learn more about this, because gold likes to spit more
than any other metal that I have evaporated. At least you are
programming your soak power and deposition power. I still set up those
manually, changing power depending on what is new in the pocket.

Looking directly at the source (gold melt) through a filtered viewport,
I cannot see spitting of the metal. It is only after I drop my view to
see only above the source, that I can see little fire balls of molten
gold launching from the source. I drop the power until I no longer see
this effect. Deposition rate is somewhere between 2 to 5A/s.

If you have pole extensions on your gun, try spreading your beam
density. A high density beam (small spot) has always been trouble for
me.

I don't use sweep for gold.

Throw distance: Are your substrates far enough away from the source that
the spit does not reach your substrates?

Good luck!

Jeff