Shrink holes in Pb-free boards are effects, not defects.

One problem we have begun to see more of with Pb-free soldering is the phenomenon known as shrink holes, or hot tears on Pb-free solder joints. We have little understanding about why shrink holes form or how they affect field reliability. Recently, a study was performed on joints formed in a wave soldering process, and was used to characterize shrink holes and measure their effect on Pb-free solder joint reliability.

It is believed the dull, rough appearance of SAC solder joints is an effect of the solidification behavior of the alloy, at which residual eutectic liquid contracts its volume upon solidification. The result is shrink holes, located between the tin dendrites that solidify first.

For this study, boards with through-hole components were assembled with SnAg3.9Cu0.6 alloy and a commercial VOC-free flux. The effects of PCB surface finishes (Cu OSP, ENIG, immersion Ag and immersion Sn) and conveyor speeds (50 and 137 cm/min.) on shrink-hole formation and joint surface roughness were evaluated. Assembly reliability was characterized using thermal cycling (-50° to 80°C with a dwell time of 15 min.).

Both wave profiles had similar cooling rates (~27°C/s) and similar topside preheat temperatures (~150°C). The main difference was the dwell time (3 and 7 sec.) and peak temperature (240° and 260°C). Three repetitions were assembled for each condition. One board per combination was used for time zero analysis; the other two were subjected to thermal cycling (600 and 1200 cycles).

Visual inspection showed the largest components (pad diameter of 5.4 mm) had shiny, smooth solder joint surfaces when using the slower conveyor speed, and rough surfaces when using the fast conveyor speed (Figure 1). This is true for all surface finishes, with the exception of ENIG. The reason for this behavior is unknown. On the other hand, smaller pad diameters (less than 2.9 mm) showed, in most cases, dull and rough surface appearance regardless of the process settings. Therefore, optimization of the appearance might not work for all joints in the same assembly.


Boards were subjected to thermal cycling to determine how the shrink holes evolve and if they will affect reliability. SEM analysis of the surfaces, as well as cross-section analysis, was performed after 600 and 1200 cycles. Surface differences were observed; the joints formed at a slower conveyor speed also showed the presence of tin dendrites and shrink holes, and generally had a rough appearance.

Cross-section analysis after 1200 cycles showed minor changes in the surface and no evidence of any shrink hole evolving into a crack.

Figure 2 shows shrink hole formation inside the joints. This type of shrink hole might jeopardize the electrical connection if many voids are formed around them by increasing the path for the cracks. However, in this analysis, there was no evidence of cracks that produce a discontinuity in the electrical interconnection of any solder joint, even if the shrink holes were located near voids.


IPC-A-610D states shrink holes or hot tears are acceptable for Class 1, 2 and 3 and for connections made with Pb-free alloys, if the bottom of the tear is visible and the shrink hole does not contact the lead, land or barrel wall.¹ Therefore, shrink holes are here to stay and end-users should treat them as an effect, not a defect.

We may conclude the appearance of SAC solder joints cannot be controlled for the entire assembly. However, it is important to know no evidence indicates shrink holes adversely affect Pb-free assembly reliability.

References

1. IPC-A-610D, Acceptability of Electronic Assemblies, February 2005.

Ursula Marquez de Tino is a process and research engineer at Vitronics Soltec, based in the Unovis SMT Lab (vitronics-soltec.com); umarquez@vsww.com.