Most IC packaging technologists are aware of this “other” issue, but this topic has never required serious consideration. The issue is the phase transition in experiences when cooled through 13°C. The result of this transition, called tin pest, is reported to cause historic tin objects to disintegrate into a dark, powdery material. This transition is not a rapid, spontaneous change, and may take years to initiate, but some in the electronics industry are beginning to ask questions about it. Some experts in the field of tin finishing have reported the inability to induce tin pest in pure tin deposits, so much remains to be learned.
Normal, “white” Sn (beta phase) undergoes a crystal phase transition (polymorphic or allotropic transition) to brittle, alpha “gray” Sn, upon cooling below 13°C. The tendency for this transition to occur is enhanced by the presence of certain impurities, such as Al or Zn, in the Sn. Colder temperatures also accelerate the transition, reaching a maximum transition rate between -30° and -40°C. It also has been reported that the solution of appropriate levels (possibly only 0.1 to 0.5 mass %) of suitable metals, such as Sb or Bi, may inhibit the transition or drive the initiation temperature so low that it is not relevant, or where the low temperature itself may oppose the atomic reordering required in the transition. Another brittle phase transition occurs at 161°C, but it appears that the typically rapid cooling rate seen in electronic assembly avoids this transition – resulting in the formation of the ductile beta phase when solidified (See Table).
The beta-to-alpha molar volume increase of 26% causes the disintegration of Sn objects. This “swelling” might not be a problem if the alpha tin retained the ductility of beta Sn. Alpha Sn, however, is brittle like neighboring members of the Periodic Table Group 14 (C, Si, Ge, Sn, and Pb). Alpha Sn shares other characteristics with its sister elements, Ge and Si. In the transition, metallic beta Sn coverts to semiconductor alpha Sn, and the alpha Sn phase crystal structure is diamond cubic, just like silicon. The transition swelling causes high stress levels in the brittle alpha crystal lattice, resulting in its crumbling into a dark powder. White Sn is a unique metal that does not show the atomic ordering of common metals. Instead, the highly ordered Sn tetrahedrals exhibit preferred directional bonding that is characteristic of covalently bonded non-metals.
Tin rods are known to “squeak” when bent, due to strain in the directionally bonded crystal structure. Most lead-free solders being considered for the electronics market consist of ≥ 90 mass % Sn. Sn-Cu and SAC solders possess an almost pure tin matrix phase, because of the low solubility of Ag and Cu in solid Sn. In 2001, Plumbridge, from The Open University in the U.K., reported that tin pest will occur in bulk Sn-0.5Cu solder bars held at -18°C for up to 2 years. At NEPCON 2004, Plumbridge and Rist showed that tin pest can occur in bulk Sn-3.8Ag-0.7Cu solder cylinders under the same conditions, although the severity appeared much reduced.
It remains to be demonstrated that tin pest may occur in lead-free electronics. Leadframe-based electronic components commonly were plated with Sn or Sn-dipped until approximately the mid-1970s, when the addition of lead to the tin-plating chemistry was mandated to mitigate tin whiskering.
In a recent, extensive, web-based literature search, no references were found recounting the failure of electronics from tin pest, although it is recognized that little or no assembly used lead-free solders, even with Sn-plated components. This may be due to the composition of the Sn finishes, thin Sn layers, base leadframe substrate effects, morphology of the Sn, Sn stress state, high purity of Sn, etc. These factors may be significant and cause effects that differ from those reported on the large bulk results in Plumbridge’s work. Or, they may have nothing to do with the results on the test bars. Could the beta-to-alpha transition in the large bars be affected by unintended low-level impurities in the bulk metals, residual stresses from casting, surface contamination, surface morphology imposed by the mold, cooling rate gradients, etc.?
Presently, there is no body of evidence declaring tin pest to be a clear concern. I believe the industry should not be complacent, and the issue should be addressed by open industry consoria efforts. At least one significant study appears to be underway at the University of Maryland (CALCE), with results due this year.
LEO M. HIGGINS III, Ph.D, director of Application Engineering, may be contacted at ASAT Inc., 3755 Capitol of Texas Highway, Suite 100, Austin, TX 78701; (512) 383-8297; e-mail: [email protected]