All

I agree with Werner, I have never seen a failure in either of the IMC phases
of the copper-tin IMC.  However, failure due to intermetallic growth is
common in the industry that operates between 150oC and 225oC.  The mechanism
is: The copper diffuses into the tin and forms a nice pure intermetallic
compound.  It leaves behind a debris layer between the copper and the IMC or
forms a weak layer between the IMC and the solder composed of the other
metals and debris in the solder besides tin.  I have seen a failed layer on a
board that was soldered with Sb5 (started with 5% Sb) to have a 20% Sb
composition along with oxides, sulfides and other debris.  Another type is
where the cooper plating occludes too much of the plating additives and other
contaminates such as iron and the fracture is between the copper and the IMC.
 I have seen one failure wherein a contaminated layer appeared between the
IMC layers, I could not give this one a theory.

The failure in not often seen with the common tin/lead solders because the
remaining lead rich layer is still fairly strong and ductile.

If you think you have problem with IMC, put your board in the oven for 100 to
200 hours at 150oC and do a tape test on some of the conductors.  Try it
several times in the same place.  If the tape will pull it off you component
leads are not far behind.

Phil Hinton