Black pad--a scourge on your boards: the potential failure mechanism can reside randomly and unpredictably.

Electroless nickel gold (ENIG) has emerged as a popular surface finish and for good reason: It has a lot going for it. Besides being Pb-free, ENIG provides flat surfaces, thus complementing fine-pitch components. ENIG does not oxidize and can accommodate long durations in storage and multiple thermal excursions. It exhibits excellent wetting characteristics with both leaded and Pb-free alloys. These attributes usually more than compensate for the price of gold (compared to several other available surface finishes).

In the ENIG process, electroless nickel resides between the component and underlying copper foil pad. The nickel plays a significant role in the ultimate strength of the solder joint. However, with all its positive attributes, ENIG has a potential Achilles heal.

Black pad syndrome has likely existed as long as ENIG has been in use, but first became apparent on BGAs. When affected solder joints fell victim to black pad, the removed component revealed a darkened or "black" pad with evident non-wetting. By preventing wetting, a weak intermetallic bond at the solder-nickel interface results, and the resultant joint is susceptible to shock and stress. Because the phenomenon is visible only upon removal of the component, the failure may be detected during electrical test or may result in an intermittent culminating in field failure.

It is believed the mechanism for black pad formation is accelerated corrosion of the nickel surface during immersion gold processing. Instead of just an exchange of nickel and gold atoms, it seems the nickel surface becomes excessively depleted prior to the application of gold. Nickel is more susceptible to oxidation.

Analysis of black pad failures typically shows abnormally high phosphorus concentrations in the nickel. This led to a general belief that the problem was caused by excessive or uncontrolled phosphorus co-deposition during electroless nickel plating. More recent investigations have demonstrated this apparent excess of phosphorus is actually due to the nickel being oxidized.

The most concise explanation I've heard comes from my friend and heavy-duty failure analysis guy, Mark McMeen of Solder Technology International (solderingtech.com). Mark cites the IPC Class II standard that calls for a minimum of 1.97 [micro]m of gold over 100 [micro]m of Ni. (1) As flat as gold is, it isn't perfect, and if deposited to a minimum of 1.97 [micro]m, some pads will be actually less thick. The fold must cover the "peaks and valleys" of the nickel; if it doesn't, the nickel is exposed, becomes oxidized and thus nonsolderable. Mark advises a nominal 4 [micro]m, minimum 3 [micro]m layer of gold to avoid nickel exposure. Accordingly, working with a nickel layer of at least 150 [micro]m will lower the probability of exposing phosphorous and the resulting black pad.

On one hand, black pad usually occurs at a low level and may not occur at all. However, in light of the explanation, it is apparent it is batch-related and can, indeed, reside randomly and unpredictably.

This is straightforward, so why has black pad been on the rise? First, assuming the latter two underlying reasons have always been present, it has become more prevalent with the increased use of BGA and other area array devices. While it has been observed on fine-pitch QFPs, the gullwing lead has more resilience and resistance to stress than the C-5 and C-4 attachment types.

That said, insufficient gold deposits by the fabricator mostly cause black pad. Although it may have been specified appropriately (3 to 4 [micro]m), the fabricator might not have built to the design or spec. The proverbial "finger on the scale" prevails, with fabricators skimping on the gold. Sadly, this is particularly true (though not exclusively) with PCBs coming from China. Try not to act too surprised. (2) We advocate strong Certificates of Compliance, coupled with batch testing of incoming PCBs, to keep them honest. (This is always a "best practice.")

We've also observed many incidences of out-of-control plating (among other) processes both for components and PCBs. Auditing as well as getting references of PCB suppliers under consideration should be mandatory. Look for good process control practices and documentation.

While not for every application, ENIG is a viable, popular PCB surface finish. Black pad should not be a deterrent to successful ENIG implementation. A little diligence goes a long way. Remember, we're all in this together.

References

1. IPC specifies the gold layer not exceed 5 to 6 [micro]m and nickel range of 80 to 200 [micro]m.

2. Someone recently related a story of a manufacturer in China receiving pipes that were supposed to be copper but were, in fact, an inferior material painted copper.

Phil Zarrow is president and SMT process consultant with ITM Consulting (itmconsulting.org); [e-mail]. He still bears the scars, physical and mental, of reflowing convection/IR ovens. His column appears bimonthly.

[ILLUSTRATION OMITTED]