Effects of Conformal Coat on Tin Whisker Growth

摘要

A whisker from a tin plated part was blamed for the loss of a commercial spacecraft in 1998. Although pure tin finishes are prohibited by NASA, tin plated parts, such as hybrids, relays and commercial off the shelf (COTS) parts, are something discovered to have been installed in NASA spacecraft. Invariably, the assumption is that a conformal coat will prevent the growth of, or short circuits caused by, tin whiskers. This study measures the effect a Uralane coating has on the initiation and growth of tin whiskers, on the ability of this coating to prevent a tin whisker from emerging from the coating, and on the ability to prevent shorting. A sample of fourteen brass substrates (1 inch by 4 inches by 1/16 inch) were plated by two separate processes: half of the specimens were 'bright' tin plated directly over the brass substrate and half received a copper flash over the brass substrate prior to 'bright' tin plating. Each specimen was coated on one half of the substrate with three bi-directional sprays of Uralane 5750 to a nominal thickness of 25 to 75 micrometers (1 to 3 mils). Several specimens of both types, Cu and non-Cu flashed, were placed in an oven maintained at 50 C as others' work suggests that this is the optimal temperature for whisker formation. The remaining specimens were maintained at room ambient conditions. The surfaces of each specimen have been regularly inspected using both optical (15 to 400x power) and Scanning Electronic Microscopy (SEM). Many types of growths, including needle-like whiskers, first appeared approximately three months after plating on the non-conformally coated sides of all specimens. At four months, 4 to 5 times more growth sites were observed on the coated side; however, the density of growth sites on the non-conformally coated side has since increased rapidly, and now, at one year, is about the same for both sides. The density of growth sites is estimated at 90/sq mm with 30 percent of the sites growing whiskers (needle-like forms) with the potential to cause short circuit bridging. The average growth rate of the needle-like whiskers on the non-conformally coated areas is about 130 micrometers per year with some outliers reaching 800 micrometers after one year. It is more difficult to make growth measurements under the conformal coat. As yet no whiskers have been observed to penetrate through the coating surface, however, a number of tin nodules appear on the verge of breaking through the thinner regions of the coating surface. These domes are developing sharper and sharper tips, as if a growing whisker is about to push its way through a tough skin. Our observations are that the specimens with copper flash show a much lower density of nucleation sites and significantly slower whisker growth compared to the specimens that have only bright tin plating over brass. The specimens kept at room temperature have a higher whisker density than those stored at 50 C. This is an unexpected result and does not agree with the published findings of others. All of the whiskers originate from small surface defects (thin, tiny scratches) that appear over the entire surface of each specimen.