You should be aware that gold will eventually diffuse into the
 silicon under these conditions.  Thin films (e.g. 10 nm) will become
 invisible to the eye over weeks or months.  The electronic properties
 (i.e. the Schottky barrier) will degrade in days.

 Mike Hecht


______________________________ Reply Separator _________________________________
Subject: Re: Metal sticking well to silicon and silicon nitride
Author:  mems@ISI.EDU at hugate
Date:    7/3/96 1:10 PM


    Message From: mac@calpr.com at hugate

Gold will stick well to silicon if there is no oxide present.  An HF dip
immediately preceeding the metallization is usually sufficient, although a
pre-dep sputter etch is a good idea if you can sputter the metal (depends
on device surface state constraints).  To get good adhesion on nitride, I
would recommend a thin chrome or titanium adhesion layer.  Titanium is nice
because it's biologically inert and etches easily in very dilute HF