Letter 0010

Tarnish/Discoloration of Electroless Nickel plating after semi-aqueous cleaning

-----

Q. We've run into a bit of a problem of nickel discoloration on some heatsinks. We build Mil-Spec high reliability computers. In one case the modules has electroless nickel on copper heatsinks. The heatsink looks vaguely like a ladder and is bonded to FR-4 multilayer circuit board (printed wiring board, or PWB). Chips straddle the rungs and are bonded down with epoxy and are soldered into the PWB using RMA rosin flux and a wave solder machine.

We installed a semi-aqueous PWB cleaning system and the heatsinks are coming out with black water spots (well, that's a bit over stated, there are water spot shaped areas ranging from clear tan to blackish). We're desperate to find out what's causing the discoloration and more importantly, how to avoid it. (Discoloration of nickel plating does not meet Mil-Spec) The cleaning process is a batch process with 4 sumps and a drying bay. The first sump is Axarel 46, a DuPont devised, now Petroferm sold hydrocarbon mixture (aliphatic hydrocarbon with dibasic esters: isobutyl glutarate, adipate, succinate and surfactant) - temperature 120 F, ~10-20 minute immersion, immersion spray agitation The second sump is an emulsion of DI water and Axarel - temperature 140 F, 5 minute immersion, immersion spray agitation Third and fourth sumps are DI water, 140 F, 5 minute immersion in each, immersion spray agitation. Final sump is hot air, 190 F, the rack is in the bay for 6 minutes, there is a slit nozzle at the top of the bay to act as hot air knife to blow surface water off. Naturally the problem is mainly showing up on product, not test coupons.

We did reproduce a mild case of the discoloration after taking a chunk of heatsink, bonding it to a PWB, putting epoxy tape on it, fluxing it, running it through the wave solder and then cleaning it. The discoloration appears to occur in the last spots water is standing after coming out of the final rinse before drying. We also tried cleaning a section of heatsink as-received, and after bonding to a PWB. Neither had appreciable discoloration. We sent the as-received and discolored heatsink section out for ESCA analysis of both discolored and non-discolored areas.

The as-received was dirtiest- ~2% Ni at the surface, 80% C, 10% O, smattering of Cu, Pb, Sn, P. Unstained area on stained heatsink - 20% Ni, 56% C, smattering of Cu, Pb, P. Stained area on stained heatsink - 18% Ni, 61 %C, smattering of Cu, Pb, P 40 Angstroms down Esca shows as-received - 20% Ni, 60%C, smattering of Cu, Pb, P Unstained area - 70% Ni, 20%C, 7%P Stained area - 65% Ni, 25%C, 9%P Visually at 200X the stained area has a transparent brown tint versus the silver surface elsewhere. There are iridescent spots visible over all the surface, perhaps tending to slightly larger spots in the darker areas. (Possibly residual hydrocarbon? - it does have a boiling point of 350 F)

Anyone have thoughts as to what it is, how better to identify it, or how to prevent forming it? We got a suggestion that electroless typically would have an inhibitor applied to prevent discoloration, is this common? Would the cleaning process remove this?

Steven Axdal
- Bloomington MN

-----

if your electroless nickel has 80% carbon on the surface when you receive it, I believe that is the problem. I don't think there should be any post treatment on the electroless nickel. It is not needed, I doubt you asked for it, and that is not MilSpec.

Tom Pullizzi
Platronica.com
Falls Township, PA