Current Issue An Independent Journal Dedicated to the Advancement of Chip - Scale Electronics April 2001 Email the editor

By Dr. Jennie S. Hwang and Dr. Zhenfeng Guo, H-Technologies Group Inc., Cleveland, Ohio

One of the common surface coatings on component leads and the PWB pads is Sn/Pb solder coating. As the industry starts to convert to lead-free solders, there is an uncertainty as to the compatibility between lead-free solder joints and the Sn/Pb coating that is employed on the surface of the component leads and PWB pads.

When a lead-free solder alloy is soldered on the Pb-containing surface, Pb will contaminate the lead-free solder alloy through metallurgical reaction. This metallurgical reaction is fundamentally a secondary alloying process and is almost instantaneous under the common soldering conditions.

To simulate this problem, we intentionally doped the lead-free solder with a small dosage of Pb. Considering the thickness of the common surface coating, a dosage up to 0.5% has been examined.

Experimental

Solder compositions were prepared by melting, in sequence, commercial solder alloys and pure metals in a glass beaker in an electric resistance chamber furnace. The master solder alloys of 99.3Sn/0.7Cu, 93.5Sn/6.5Ag, 95Sn/5Sb and 63Sn/37Pb (in a bar form) were used.

(The experimental protocol and test procedures were the same as outlined in Part 1 of this series.)

Results

Pb at 0.1 %, 0.2 % and 0.5 % was added to an optimal Sn/ Ag/Cu/Sb lead-free composition (95Sn/3Ag/1.5Cu/ 0.5Sb) as representative of an Sn/Ag/Cu/ Sb system*. The resulting solder compositions, along with their melting temperatures (Tm), yield strengths (σy), tensile strengths (σTS), Young's modulus (E), plastic strains (εp) at fracture and fatigue life (Nf)-at a total strain range of 0.2%-are summarized in the table (top, next page). All compositions are expressed in weight percent, unless otherwise specified. The reference alloy of 63Sn/37Pb is also employed in the table.

Melting Temperature

As seen in the table, the Pb addition in small amounts did not influence the alloy melting temperatures.

Strength and Plasticity

Figure 1 summarizes the relative performance in the tensile stress (σ) vs. strain (ε) behavior of 94.9Sn/3Ag/1.5Cu/ 0.5Sb/0.1Pb, 94.8Sn/3Ag/1.5Cu/0.5Sb/ 0.2Pb, and 94.5Sn/3Ag/1.5Cu/0.5Sb/0.5Pb, in comparison with 95Sn/3Ag/1.5Cu/0.5Sb and 63Sn/37Pb.

As shown in the figure, the Pb addition produced no apparent effect on alloy strength, considering the typical data scatters in tensile tests. However, the addition of Pb at 0.1% and 0.2% reduced the alloy plasticity.

The tensile flow of solder alloys typically consists of an elastic region, a strain-hardening region, a stress-recovery region and a cracking region. Strain hardening continued until necking occurred at the maximum load or the tensile strength (σTS). Necking was caused by an inhomogeneous plastic deformation somewhere in the gauge length, and was associated with strain localization.

Stress-recovery mechanisms are believed to be dominant in the region after necking, and before abrupt fracture, for high- temperature plastic deformation.

Low Cycle Fatigue Life

The presence of lead in the Sn/Ag/Cu/Sb system largely reduced the low cycle fatigue life (Nf). Figure 2 shows the low cycle fatigue life (Nf) vs. Pb dosage. We found that Pb at 0.1% caused an abrupt drop in fatigue life. Test results indicate that fatigue life slightly recovered as the Pb dosage increased to 0.5%. We are uncertain which mechanism contributed to this slight recovery. In any event, however, the recovered fatigue life is still significantly lower than the alloy which did not have Pb added.

Discussion

The maximum solid solubility of Pb in Sn, Ag and Cu is 2.5 %, 5.2% and 0.29%, respectively. However, the solid solubility of Pb in Sn, Ag and Cu at room temperature (300°K) approaches zero. Therefore, we expect the Pb atoms in the Sn/Ag/Cu/ Sb system to precipitate out as second-phase particles.

Since the strength is affected by the second phase following the mixture rule, the small amount of soft Pb particles in the experiment produced little effect on alloy strength, as reflected by the test results in the table.

With the presence of the softer Pb particles in Sn-matrix, plastic deformation tends to concentrate on these particles, which may cause an early fracture, leading to the plasticity reduction. The fatigue cyclic strain will also concentrate on the soft Pb inclusions, accounting for the deterioration of the fatigue life.

Additionally, the distribution of Pb second phase in the microstructure should also play a role in the fatigue mechanism. We postulate that a small amount of Pb precipitates tend to reside preferably at the grain boundaries, causing early grain boundary fracture.

Summary

The contamination of Pb up to 0.5% in 95Sn/3Ag/1.5Cu/0.5Sb produces no apparent effect on the alloy melting temperature. Its impact on the strength of Sn/Ag/Cu/Sb composition is also negligible. However, adding Pb reduces alloy plasticity and fatigue life. Nevertheless, the fatigue life of the Pb-contaminated alloy in Sn/Ag/Cu/Sb system is still higher than that of 63Sn/37Pb.

* J. S. Hwang, Chapter 27, Environment-Friendly Electronics: Lead-Free Technology, Electrochemical Publications Ltd., UK, April 2001.

Dr. Jennie S. Hwang Dr. Hwang is an inductee into the WITI Hall of Fame and a member of the National Academy of Engineering. She is internationally known for her work in SMT manufacturing and has written more than 180 publications and several textbooks. Dr. Guo is a research scientist. [jslhwang@aol.com]