I agree with Mr. Cantos that it is very likely a galvanic effect.
Another possibility is diluting your etchant with ethyl glycol. That
would reduce the effect of the gold to solution current and might block
the emf. I don't know this for a fact, but it is something I am looking
into.

Keeping the resist one, as Mr. Cantos suggests, should help some, just
by reducing the area of the gold in contact with the solution and
thereby reducing the current.

Perhaps some sacrificial layer can be deposited on the gold or a polymer
to block contact with the solution. These are highly speculative
suggestions as is the ethyl glycol.

Ed

-----Original Message-----
From: mems-talk-bounces@memsnet.org
[mailto:mems-talk-bounces@memsnet.org] On Behalf Of Brad Cantos
Sent: Tuesday, November 11, 2008 7:28 PM
To: General MEMS discussion
Subject: Re: [mems-talk] selective etch of AlGaAs/GaAs with different
dopingand metal in present

Hi Z.Z.,

What you are describing here is, I think, the formation of a galvanic-
enhanced etch due to the presence of the metals, and that is why you see
a difference between the n-type and p-type materials.  You can try to
keep photoresist on top of the metal to minimize the contact to the
etchant to reduce the galvanic effect.  I have also had good experience
with dilute phosphoric acid-based etchants (try
100:10:5::DIW:H3PO4:H2O2 as a starting point).  You may still experience
some difference in etch rates regardless.  How much difference can you
tolerate?

Brad Cantos