This gelatinous residue, that you discribe, appears to be the same substance 

we have encountered from the new direct plating process which utilize 
graphite.
We have performed evaluations of the effects on performance and long term 
reliability
issues but have not yet made any conculsions.

My suggestion is that you ask your supplier if they are using one of the new 
direct metalization
processes being introduced into the market place, instead of the old 
electroylsis process.
 ----------
From: [log in to unmask]
To: [log in to unmask]
Subject: Gelatinous residue on High Tg Material
Date: Thursday, June 27, 1996 7:34PM

Has anyone noticed a thin layer of gelatinous residue between the copper
plating and the laminate during X-section inspection?

What we are seeing appears to be isolated occurances of hole wall separation
when viewed at 100-200X.  Closer examination at 500-1000X reveals a thin
layer of gelatinous residue (approx. 100-200 Microinches thick) on the back
side of the plating.  So far I have only observed the problem on the B-stage
material which happens to be a high Tg (170 degree) FR4 material.  We are
using KMnO4 desmear.  The residue differs from normal holewal pull away as
follows:

1.  There are small inclusions that are coated with electroless copper and
have a dark center extending .0001-.0003 inches into the electroplated
copper.  Similar to a glass fiber inclusion.

2.  The contour of the laminate face and the contour of the electroplated
copper do not match in the area exhibiting hole wall pull away.  There is a
layer of gelatinous residue on the backside of the plating that produces
different contours and separates from the laminate.

3.   The gelatinous material has an irridesent appearance.

4.   When polarized light is used the gelatinous materila looks like the
base laminate.  If the material were polishing debris the material would be
white and have a crystalline appearance.

I know of at least one other company who have seen similar problems with a
high Tg FR4.  Are there others?  If any one else observes a similar problem
please let me know.

Thanks

Wendell Conner