Solder Spheres

From old-school SN63PB37 wire to Pb-free Micron Size Solder Spheres for Flipchips: How solder has evolved!

Semiconductors and printed circuit board technology are relentlessly evolving industries. Their constant development has forced solder technolgy to change and evolve too. Solder technology, as a whole, has moved from Lead-based solder wire and seperately applied solder flux to solder spheres. The latest industry break-through being a move away from wire bonding to micron size solder spheres for flipchips. To understand how this happened let's start by breaking down SN63PB37 into something meaningful:

SN = Tin
63 = 63%
PB = Lead
37 = 37%

Alloys melt at different temperatures based on their composition or ratio of metals. SN63PB37 solder melts at 183°C: this was incredibly useful for circuit boards because it was high enough to melt the solder but low enough not to damage plastic components or the circuit board itself.

Solder wire needs flux. Flux removes oxidation.

Without flux the layer of oxidation acts as a barrier between the solder and the two conductive contact points thus preventing a clean and solid solder joint.

Pre-coating parts in flux was, therefore, an essential extra step that cost more time and labor. So the next step in solder development was hollow solder wire filled with an internal flux. Further developments included multi-core wire with multiple hollows for better distribution of flux.

The industrial mass-produced Printed Circuit Board (PCB), as we know it, started out with through-hole technology. Semiconductor devices and chips had pins which were pushed through holes in the printed circuit board. The bottom side of the entire circuit board would then have flux applied in one go and then be dipped into a solder bath. This automated the soldering process and eliminated unnecessary extra steps. Which, in terms of mass production, saved millions of dollars in energy and material.

Evoling ICs – Pushing soldering technology to evolve

After through-hole technology PCBs evolved towards surface mounted technology: where solder paste was used. Solder paste is a suspension of powdered flux and powdered solder together. This allowed solder to be printed on circuit boards by a solder paste printer and then melted in one go in a solder reflow process: thereby allowing for much more complex and densely packed installations of semiconductors on the PCB.

Integrated circuits have been getting smaller and smaller and with larger and larger numbers of tiny transistors. This can be best seen in the evolution of Small Outline Integrated Circuits or SOIC: evolving at first by doubling the number of inputs and outputs or pinouts: from S08, on to SO16, SO32. Eventually for the sake efficiency, instead of two rows of pinouts, all four sides would be used in the Quad Flat Pack (QFP). This also meant that the legs got thinner and thinner and got packed closer and closer together.

Micron size Solder Spheres: moving away from Wire Bonding

When you are using all four sides and even stacking wafers what can you do to further increase the number of inputs and outputs? The answer turned out to be: use the bottom of the chip as well. So, instead of many small thin legs, solder spheres would be placed directly on the bottom of the chip and would act as the connection between the Ball Grid Array (BGA) and the circuit board. This meant full encapsulation in an epoxy molding compound was no longer necessary and you could simply use an underfill.

250,000 solder spheres: sounds like a lot, but at mil sizes, that’s just enough for a heaped tablespoon full. Eventually, solder spheres got smaller and smaller so that they could pack even tighter grids of connection on the bottom of silica die. In these flip chips setups, tiny solder spheres are used instead of even 25-micron thick gold wire. Solder spheres are made in an amazing process with incredible accuracy and tight quality control.

A parallel evolution in material: from PB to PB-Free

Early in the 2000’s in Japan, Sony and other concerned companies started to push for lead-free solder. This was in response to a situation in which electronic devices were ending up in landfills, and these landfills were in turn exposed to rainwater which allowed the lead in the solder to contaminate the groundwater. In 2006 the EU Restriction of Hazard Substances Directive (RoHS) came into effect restricting the use of lead consumer electronics.

This led to SAC305 Lead-free solder replacing SnPB63/37 as the industrial standard: with 75% of companies using it. Here is what the name means:

S = Tin
A = Silver
C = Copper
3 = 3% Silver
05 = 0.5% Copper

As a near eutectic solder it has adequate thermal fatigue properties, strength, and wettability. Another advantage of SAC solder is that it’s more resistant to gold-embrittlement and has substantially higher joint strength than PB solders.

Going Pb-free however means using silver instead of lead which means increases costs. For some applications SAC solders with less silver are suitable. These include SAC105, Bi57Sn42Ag1, as well as silver-free and lead-free solder , Bi58Sn42.

Using silver means that SAC solders have a 34°C higher melting point than Sn63Pb37 solder at between 217-220°C. This requires a higher solder reflow temperature of 235-245°C in order to achieve wetting and wicking. Some printed circuit components are susceptible to SAC assembly temperatures: including capacitors, optoelectronics etc.

Companies have since started offering 260°C compatible components to meet the high temperature required for SAC solder reflow. 260°C compatibility is fast becoming the new standard for semiconductor and component manufacturers.e.