Can Nickel Barriers Eliminate Tin Whiskers?

Like it or not, lead-free requirements are coming and will be here to stay, along with one of their biggest problems: tin whiskers. Agere Systems (Allentown, Pa.) recently announced promising results from using a nickel barrier layer to hinder the growth of tin whiskers when tin is used as a pad or pin finish over copper.

Tin is ubiquitous in soldered electronic interconnections and pin finishes, because it wets well in soldering processes and makes good metallurgical contact with typical pad and pin metals. It also has a well documented tendency to form whiskers (Fig. 1 ) that can cause shorts and puncture conformal coatings. It is commonly understood that whisker formation in tin is a stress-relief mechanism. Using lead as a companion metal for tin has been effective in terms of relieving stress, but now other ways must be developed for relieving stress in tin solders.

Five years ago, another subsidiary of Lucent Technologies, Electroplating Chemicals and Services (Murray Hill, N.J.), stepped forward as a proponent of matte tin,¹ which has since become a lead-free finish of choice. Matte tin is simply tin that is plated using a solution without brighteners. Organic brightener molecules can get trapped inside the tin and cause multiple problems. They act as nucleation points for grain growth. This reduces tin grain sizes and increases the number of tin grains, which increases the chances for intergrain stresses. They can also outgas during solder attach operations.

1. Matte tin on copper test vehicles subjected to a 260°C reflow grew tin whiskers in accelerated life tests. (Source: Agere Systems)

2. With a nickel barrier between the copper and tin, whiskers did not form in accelerated life tests. (Source: Agere Systems)

For test vehicles with a plated barrier layer of nickel between the copper and matte tin, the tests showed no whisker growth in the accelerated life test (Fig. 2 ), with or without the reflow prior to the test.

Agere Systems' proposed model for this behavior is that copper gets trapped in the tin layer during reflow and forms intermetallics inside the tin, rather than at the copper-tin interface. Rather than metallurgical connectors, these trapped intermetalics act as impurities in the same way that organic brighteners do. The results using a nickel barrier are consistent with this model.

Nickel barrier layers are not new to the electronics industry. They are used on surface-mount components to improve wetting, reduce solder leaching, and reduce tombstoning. They have been used to prevent solder from leaching into parts like precision resistors so that their properties would not change over time. They have also been used to keep copper from diffusing out into gold overlayers, an application similar to the one that Agere is proposing.

The military and aerospace industries have also examined tin on nickel, along with other combinations that are much more expensive. Typically, nickel barrier layers in the neighborhood of 1-3 µm are thick enough to block interdiffusion, but not thick enough to make the interface brittle. According to John Pittman, Agere Systems' vice president of assembly and test operations, 12-15 µm of tin on 600-800 nm of nickel has successfully hindered tin whisker growth.

Pittman also noted that adding a nickel plating step to an existing matte tin plating operation may pose some challenges in some cases, but none that are insurmountable. Agere Systems currently offers the additional nickel layer at no additional charge to customers who desire a matte tin finish. If nickel barriers are as effective as Agere Systems' tests indicate, the industry may have a solution to the tin whisker problem.

For additional information on semiconductor packaging, go to www.semiconductor.net/packaging