Clean it up! Incoming board and component cleanliness is crucial in building reliable product.

Dirty bare boards and components are some common culprits when ionic cleanliness is in question. Plating processes and tap water rinse cycles are known to introduce harmful ionic contaminants such as sulfate, chloride and methane sulfonic acid (MSA). If not properly neutralized through good cleaning procedures, they will carry through assembly to the final product, and if mixed with moisture and an applied voltage, the residues can create electromigration or electrical leakage failures.

[TEXT NOT REPRODUCIBLE IN ASCII]

A manufacturer of sensors for an automotive applications saw units failing in a 10-pin DIP area after shipment but before installation into a vehicle. There was visible dendrite growth and a gooey residue at the power to ground leads of pins 8 and 9 on 10-pin DIP microprocessors in the failing area, which sees a constant 3V. These assemblies were built on immersion silver boards using a hydrochloric acid micro-etch prior to plating, and a tap water rinse. The assembly process included single-sided reflow with no-clean SnPb paste using a 0.006" stencil and a seven-zone reflow oven. A battery was then attached with a spot weld using no flux or cleaning. Units were then functionally assessed and readied for potting, which used a two-part silicone crosslinked elastomer injected into the housings with the assemblies in place. The potting compound cured for 15 minutes, then the units were transported to a drying area for 24 hours. The units were then assembled with a valve stem, functionally tested and readied for shipment.

[ILLUSTRATION OMITTED]

To investigate this failure mechanism, Foresite examined two failing units and one working unit. We performed a full visual inspection and tested for shorts using a multimeter. On the failing units, we noted slimy coating areas in the failing 10-pin DIP areas at the lead interface. A visible dendrite was also seen between pins 8 and 9 on the DIP at the board surface. There was a heavy white residue present between most leads of the DIP, and there were measurable shorts between pins 7 and 8 and pins 8 and 9. There was also adhesion loss at the lead surfaces, but good cross-linking. We examined several other areas of the failing assemblies for reference, and found no evidence of visible residues or shorts. Looking at the good unit, we saw good adhesion of the DIP part to the board surface and good cross-linking.

[FIGURE 1 OMITTED]

We then conducted ion chromatography with standard and C3 localized extraction techniques to inspect the failing DIP area, reference areas, and incoming bare boards and components. High levels of chloride and sulfate were found on incoming bare boards (Table 1), probably due to the tap water rinse. We also found a heightened level of sulfate and MSA on the components, most likely attributable to a plating process. Finally, we found high levels of chloride, sulfate, MSA and weak organic acids in the failing 10-pin DIP lead areas of the final assemblies, indicating that incoming cleanliness issues carried through the assembly process. The WOA levels were due to heavy flux deposition that was not solublized effectively for this SMT process.

[FIGURE 2 OMITTED]

For remediation, we recommended cleaning the incoming bare boards through a saponified inline wash, and gas plasma batch cleaning of incoming components. With these protocols in place, the customer was able to effectively lower the contamination levels. The customer also monitored flux deposition and is ensuring all solvents are fully volatilized so that the only residue left on the board is benign and non-moisture-absorbing.

Table 1. Chloride and Sulfate Levels on Incoming PWBs

                      Ion Chromatography Data      C3 Spot Testing

Sample Description     Cl-   SO4-   WOA   MSA  Results  Time (in sec.)

Failing Unit #1 - 10  10.23  7.19  49.35  1.2   Dirty         22
pin DIP lead area

Failing Unit #2 - 10   9.36  6.94  37.12  1.5   Dirty         19
pin DIP lead area

Good Working Unit -    3.56  2.31  19.25  0.3   Clean         89
10 pin DIP area

Incoming Bare Board    6.8   9.12    0     0    Dirty         37

Cleaned Bare Board     0.87  0.54    0     0    Clean        180
(same unit)

Incoming 10 pin DIP    0.59  5.29    0    5.3   Dirty         42
microprocessor
component

Gas plasma cleaned     0.09  0.14    0    0.1   Clean        180
10 pin DIP component

Terry Munson is with Foresite Inc. (residues.com); [e-mail]. His column appears monthly.