ENIG and Black Pad

We have noticed a problem on ENIG boards in our WAVE process. The problem looks similar to Black Pad. We do not know how to identify the problem, or where to begin to find a resolution to the problem. Here are some facts:

1. Black residue is on the vias and this on the solder side.
2. Component side SMT parts do not exhibit the problem.
3. Boards are processed Pb-free - SAC305, with no-clean water based flux.
4. THTs will not wet after residues appear (for touch up process)
5. Blowholes appear to accompany the defect but are not always present.

Hope you can help identify the problem and how to eliminate it.