What is thickness of resist and what is the minimal feature width?

Current density variation due to width change in the pattern can not
cause the type of deposit you describe unless the total current is way
off the acceptable range. You may also get that type of finish as a result
of inadequate seed layer. Typically a Ni plating solution with pH 3-4
should be acidic enough to eliminate the Cu oxide. This can be definitely
eliminated with 3-5% sulfuric acid dip for 10-15 seconds. If the problem
persists you have another cause i.e. inadequate lithography or other
organic contamination.

Incidentally, how do you know your Ni is OK? What are the recommended
range of CD,  pH, temp. and anode motion you refer to? I trust you have
plated the same Ni on a piece of flat copper substrate and verified it is
OK?

----- Original Message -----
From: "MEMS R FUN" 
To: 
Cc: 
Sent: Monday, August 22, 2005 2:10 PM
Subject: [mems-talk] Ni Electroplating Issues


> All,
>
> I am currently attempting to electroplate Ni onto a seed layer of 200A Cr
/
> 1000A Cu into thick PR (SPR 220-7) patterned spots on a silicon wafer, but
I
> am having problems with the resulting Ni layer.  The layer has been
plating
> non-uniformly (which I would expect for narrower and wider regions because
> of the enhancement of current density for narrower regions), and the layer
> looks "spongy" and blackish under a scope post-plating.  It should be
noted
> that the plating parameters are within spec using the optimal ASF current
> density, the optimal temperature, the optimal pH, the optimal anode,
> agitation, etc.
>
> One suggestion that I received was to do a quick O2 plasma de-scum, and
then
> a quick dip in BOE to remove the native oxide layer.  However, my thoughts
> are that the presence of hydrocarbons on the seed layer would contaminate
> the layer's electrical characteristics (which is a moot point for my
> reserach), and the presence of copper's native oxide layer would impede
the
> growth of the layer and not change the overall characteristics of the
layer
> (making it blackish or spongy).
>
> Hence, my questions are the following:
>
> First, is it really absolutely critical to perform the aforementioned
plasma
> and BOE steps for my plating process, or is it just an industry standard
for
> preventative measures?  If it is critical, please explain why.
>
> Second, has anyone observed this kind of behavior when plating Ni into
thick
> PR channels, and if so, what causes it and how do I correct my process to
> prevent this to attain a smooth, matte gray, relatively uniform layer?