Steve Dowds columns | March 02, 2005

Tackling the Pb-free basics

Much discussed and often misunderstood, the prospect of lead-free electronics manufacturing has been with us for some time. Tackling the lead-free basics head-on Global Product Manager for Henkel, Steve Dowds answers some common queries.
1. Why must we go Pb-free?

The law is set to stipulate lead-free electronics manufacturing. Irrespective of the merits or demerits of the case, with effect from July 2006, the EU's Restriction of Hazardous Substances Directive (RoHS) comes into force. This directive bans the use of lead, mercury, cadmium and hexavalent chromium in nearly all electronic and electrical products produced or sold inside the EU. It also bans the use of brominated flame retardants in PCBs, which has got little attention so far.

2. Are there exemptions?

The RoHS Directive lists the following exemptions:

a. High melting point solders containing more than 85% Pb
b. Servers, storage and storage array systems (exempt until 2010)
c. Network infrastructure equipment for switching, signalling, transmission as well as network management for telecommunication.
d. Pb in electronic ceramic parts, such as piezoelectronic devices.

3. Which alloy to use?

Over the past decade a number of studies have been published looking at the feasibility of various Pb-free alloys. All alloys selected for consideration use tin as a basis as this is the active ingredient in most soft solders with good wetting and bonding properties. It melts at 232°C and is non-toxic to humans. The addition of other elements including zinc (Zn), indium (In), silver (Ag), copper (Cu) and bismuth (Bi) bring down the melting point. Broadly speaking three alloy types stand out as promising candidates: SnAgBi, SnZn, SnAgCu.

Alloys containing bismuth are risky when used in conjunction with Pb-bearing components since Sn, Pb and Bi form a low melting phase at 96°. In the worst case this could mean components falling off the board under thermal cycling. However SnAgBi alloys do have two advantages: a relatively low melting point of 200 to 210°C and wetting properties similar to SnPb.

SnZn melts at 199°C, which is close to the melting point of SnPb. Unfortunately Zn reacts with both acids and bases so when mixed with flux medium it shows poor stability. Shelf life is days rather than months and there is also concern regarding long-term reliability of the assembly once soldered. Almost universally the conclusion is that SnAgCu (also known as SAC alloy) is the most suitable replacement as long as Pb is in the system. Once Pb is eliminated, SnAgBi may prove a better choice due to the lower melting point and improved wettabililty.

4. What are the differences between different SAC alloys?

There are a number of SAC alloy compositions available including 3.8Ag, 0.7Cu, 4Ag 0.5Cy, 3Ag 0.5Cu, with Sn forming the balance of the alloy. All have broadly the same melting range and exhibit the same microstructure. Manufacturing tolerances on minor constituents are typically ±0.2 per cent, so you could order 3.8/0.7 and get 4.0/0.5 with no effect on performance. According to the IPC solder product value council (SPVC) there is no substantive difference between these marginally different compositions.

Various competing patents are in place, though none has been tested in court yet so before selling any product manufactured using a Pb-free alloy, manufacturers should check with their materials supplier that arrangements are in place to avoid patent infringement.

5. What are the pricing implications?

Unfortunately all Pb-free allows cost more than SnPb. While SnPb alloys are around 63% Sn, lead-free alloys are 95% Sn. Not only are raw material costs substantially higher for Sn, demand for Sn has increased over the past year. Tin prices have risen from $4,500 per tonne in April 2003 to a peak of $10,000 in May 2004.

This will have an immediate effect on solder bars where little value is added to the material. In the case of solder paste the effect will be less, since much of the cost results from added value activities such as processing into powder, however if the price remains high it will feed through into paste prices eventually.

At the moment Pb-free is typically 40% more expensive more expensive than SnPb per kg. On the plus side however, the mass density of SnAgCu is lower than SnPb by about 12%, so users get 12% more volume of paste per kg and 12% more joints.

6. Is Pb-free more reliable than SnPb?

Initially there was an assumption that the superior bulk alloy properties of SnAgCu when compared with SnPb would mean better reliability joints in real assemblies. As a bulk alloy SnAgCu is superior to SnPb in many respects with higher strength, better creep resistance and longer fatigue life. Unfortunately when tested on real assemblies that is not always the case. If an assembly is subjected to a high strain rate over a wide temperature range (-55 to 125°C) results show that SnPb is better; if subjected to a lower strain rate over a less taxing temperature range (0-100°C), Pb- free appears to be better.

7. Are there differences in the manufacturing process?

In terms of material shelf life, printability, open time and tack there is no obvious difference between SnPb and Pb free. The major difference is reflow since higher peak temperatures are required. Sn63 Pb37 melts at 183°C with a peak temp of 205 to 220°C. SnAgCu melts at 217°C so a higher peak temperature is required, typically 230-260°C.

Naturally this heat is not good for components, so when evaluating solder paste it is best to find a paste which gives maximum wettability with minimum superheat. It is possible for small boards with minimum delta T to reflow with a peak temperature of 229°C, however assemblies with a large thermal mass, such as backplanes and motherboards, present more of a challenge. For PCBs with a delta T of 20 to 25°C, it may be necessary to have a peak temperature of 260°C in order to ensure good reflow across the board.

8. What about voiding in reflowed joints?

Good practice for managing the transition to Pb-free is to do it step-by-step in order to avoid introducing too many variables at the same time. The best way is to first change the board and then the paste. Assuming the components can take the higher peak temperature of around 229 to 240°C, the solder paste can be changed. Manufacturers will most likely have a mixture of SnPb finished components with Pb-free solder, which can increase voiding. This is partly because the SnPb component terminations spend too much time above liquidus, experiencing more oxidation during air reflow.

Another factor leading to increased voids is the higher surface tension of the SnAgCu alloy, which makes it harder for gas generated by the fluxing reaction to escape. Void levels are also dependent upon the flux chemistry, which varies between different solder paste suppliers.

9. What about inspection?

The higher surface tension of SnAgCu alloys means that they do not wet or spread as far as SnPb, particularly on copper. Using less superheat also means wetting speeds are lower, such that when used on OSP copper boards the paste does not wet to the corners of each pad without overprinting, which can lead to other problems. This reduced wetting is not a defect, but a property of the alloy, which can be accommodated by retaining operators and inspectors.

10. Where do I start?

Time is short since the implementation date is only two years away. Here are some useful resources that can offer help.

Henkel offers manufacturers support and advice across the PCB from Multicore solder materials including lead-free alloys, to die attach adhesives, underfill products, encapsulants and PCB coating materials.


Please note the following: Critical comments are allowed and even encouraged. Discussions are welcome. Verbal abuse, insults and racist / homophobic remarks are not. Such comments will be removed.
Further details can be found here.
Load more news
September 15 2017 9:25 AM V8.7.1-2