Facing a hairy electronics problem.

Three satellites have gone dead because of whiskers. Last spring, a whisker shut down a Connecticut nuclear power plant. The culprits aren't errant facial hairs: They're metal filaments that spontaneously sprout from electroplated metal films in electronic devices, where they can short out circuits. By tracking properties of such films for up to a year, researchers have now learned more about the films' internal stresses, which are suspected of causing whisker growth.

The whisker problem largely disappeared in the 1960s with the introduction of lead-tin films in electronics. Researchers had found that metal combination to be almost immune to the problem. But a worldwide drive to eliminate lead from electronics products has resurrected the menace. The European Union, for instance, begins its electronics-lead ban next summer.

Scientists have long suspected that by launching a whisker, a film relieves itself of compressive stress. Many factors contribute to such stresses, but just how they do so remains murky. Electroplating introduces stresses into films, but whiskers research has focused largely on stresses created by chemical reactions in the films, such as those between tin and copper.

To investigate both processes, metallurgist William J. Boettinger of the National Institute of Standards and Technology (NIST) in Gaithersburg, Md., and his colleagues made thin bronze cantilevers, electroplated them with tin or tin alloys, and then monitored flexing of the strips in response to compressive or tensile stresses in the films' microstructures.

Even accounting for the chemical reaction-caused stresses, those initiated by the electroplating still played a large role in whisker formation, the scientists report in the November 2005 Acta Materialia.

The NIST results "will be very helpful in linking the plating condition to whisker growth," comments materials scientist King-Ning Tu of the University of California, Los Angeles.--P.W.