Pull testing reveals large differences between alloys.

The solubility of any specific metal in a liquified alloy is characterized in a phase diagram where temperatures, percentages, metal phases and intermetallics are defined. As a result, solid metals immersed in liquid solder will dissolve until the solubility limit is attained. This is in the absence of any electrochemical reactions taking place. In reality, however, this point is never reached, because the equilibrium concentration between dissolving of the metal and byproduct removal is low. Base metals used in electronics have high dissolution rates in SnPb solder alloys. For wave soldering, the dissolved metals are “washed away” by the solder wave and replaced by solder from the bath. The dissolution rate depends on the base metal, solder composition, temperature and flow velocity of the solder.¹

A PCB's copper plating can completely dissolve, especially in the knee of plated through-holes. When copper is soldered in a wave process, the copper content in the solder bath increases until equilibrium between dross removal, alloy use, alloy replenishment and dissolution is achieved. However, final copper concentrations exceeding 1% in the solder bath by mass result in the formation of Cu6Sn5 needles. When formed, these needles migrate to the areas within the solder pot that have low flow rates. Once a critical mass of needles is formed, the Cu6Sn5 needles are likely to enter the flow of solder being applied to the board. Two potential scenarios could affect joint or board reliability. The needles flow along with the solder into the vias, or attach to the bottom of the board, where the risk of a bridge increases. Figure 1 shows the increase of copper concentration of the SnCu0.7 bath as a result of the number of boards soldered.


The primary attachment process, such as wave or selective processes and rework, affects dissolution. A comparison between wave and selective soldering processes was conducted. PTH copper thicknesses were measured in three different areas for 10 samples. Optimized settings were used for all soldering processes. For wave soldering, the SAC 305 alloy temperature was set at 265ËšC and the dwell time was 3.7 sec. The data indicate 24% of the copper was dissolved. For the selective process, two methods were tested. One employs a single nozzle point-to-point soldering system. In this case, the alloy was set at 290ËšC and 2.5 sec. dwell time. The data indicate 8% of the copper was dissolved. The other employs a series of nozzles designed to solder all PTH joints in one dipping process. The alloy was set at 320ËšC and a 3 sec. dip time. The data indicate 35% of the copper was dissolved. In all cases, the major dissolution was observed in the PTH knee.

The results indicated dissolution rates vary by the soldering process. To measure the dissolution effect, pull testing was performed in PTH assembled under wave and selective processes. A 96-pin plastic frame through-hole connector with SnPb finish was assembled using SAC 305. The wave process solder temperature was set at 260ËšC with a contact time of 6 sec., while the selective soldering temperature was set at 300ËšC with a drag speed of 3.5 mm/sec. Assemblies also were subjected to thermal cycling (-40Ëš to 125ËšC with a dwell time of 10 min.) to fatigue the solder joints while ascertaining the differences between thermal processes. Results indicated the same pull strength (approximately 300 N) for both soldering techniques for those samples analyzed at time zero. By comparison, a surface mount component such as a QFP with SnPb finished leads, soldered with SAC 305 solder paste, has a pull strength of approximately 12 N.

When subjected to thermal cycling, however, a decrease in pull force was observed for the wave joints, while the selective joints showed no effect (Figure 2).


The focus of many current investigations and publications, copper dissolution issues to be concerned about are:

1. Solder pot maintenance, to keep levels of copper and other elements in control.
   
   
2. Implementation of a standardized inspection process.
   
   
3. Selection of appropriate alloys for wave, selective and rework processes that reduce the copper dissolution rate.

References

1. R. K. Wassink, Soldering in Electronics, Electrochemical Publications, 1989.

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.

• Prev
• Next