Jason:

Etching gold is even more remarkable than the etching of chrome since it
produces no volitile byproducts using any chemistry I have ever heard of. I
have seen in the case of gold that at low pressures using oxygen you can
move or sputter gold that is thin. The mechanism is too much rie power at
low pressure. This begins to fall into the catagory of sputter etching which
is really ion milling. You should reduce your rie power make sure you have
good cooling on your chuck and increase pressure to around 200 mtorr. The
other possibility is that residual fluorine from your SiNx etch is actually
undercutting the layer your gold or Cr is resting on. Do you use a Ti layer
prior to gold deposition for added adhesion? Fluorine could be running along
the interface allowing your gold pattern to actually lift off during
processing. A careful examination of the bottom of the chamber will show
flakes of gold and in some cases the actual pattern if you can roll them out
to their original shape you could verify. Either way to rebaseline your
chamber you will need to do a wet clean on the chamber and all associated
parts. The chamber is problematic since it is relatively large and hard to
disassemble. In this case a scotchbright abrasive cloth using di water
followed by an isopropyl alcohol scrubbing and nitrogen blow dry should
help. All of the other parts like showerheads standoffs ect can be cleaned
using a heated TMAH cleaning. There is a company here in Arizona that has a
special process that nicely cleans aluminum and stainless and they follow up
their clean with a CO2 ice blasting method before double packing in plastic
to remove particles. Once reassembled a proper pump down to a low base
pressure to remove any moisture is helpful with a number of pump downs and
vents with nitrogen to get to the lowest base pressure possible. Kind of
complex don't you think? Let me know if I can be of more help and you can
contact me directly if you like. Bob Henderson

----- Original Message -----
From: "Jason Milne" 
To: "'General MEMS discussion'" 
Sent: Wednesday, January 16, 2008 6:32 PM
Subject: [mems-talk] Fluorine residues etching Cr and Au


> Bob (and others),
> I find that O2 plasma occasionally removes Cr and thin Au layers (~30
> nm) in our Oxford Plasmalab 80 plus. I also use this chamber for SiNx
> etching with CF4. I was interested in your response to another thread
> that fluorine residues may be responsible for removing Cr- that may be
> what is happening to me. What do you recommend as a cleaning method? Our
> usual O2 / CF4 plasma clean doesn't seem appropriate in this case.
>