5 Steps to Successful Lead-Free Soldering: Step 5.

This article is the final one in this series describing how to successfully implement lead-free soldering. After implementing lead-free soldering in production, we must constantly follow-up, monitor and analyze the data to keep the process under control.

Lead-free soldering has been introduced, so numerous questions have been raised. However, many questions have yet to be answered, including the definition of lead-free, its implementation cost, and even if all technical issues have been solved. But, experiments continue to provide good numbers on the reliability of the new lead-free alloys.

This article discusses cost and energy effects and demonstrates that the process must be verified continuously because technology and process knowledge will improve in the future. A standard improvement model, such as the Deming cycle, can be used to maintain control of the lead-free soldering process, make adjustments and improvements, and realize cost savings when possible.

Material Costs

Solder 

As an example, the solder pot of a certain wave solder machine contains about 760 kg of tin-lead (SnPb) alloy. To fill the solder pot with SnPb will cost about $3,960. The density of SnPb is 8.4 g/mm2. To fill the same solder pot with tin-copper (SnCu) alloy, which has a density of 7.31 g/mm2, we need 661 kg:

Mass = (7.31 divided by 8.4) x 760 = 661.

The result is a cost increase of 28 percent for solder or $5,063. Other lead-free alternatives such as tin-silver (SnAg, 135 percent) and tin-silver-copper (SnAgCu, 145 percent) have an even greater impact on solder costs.

Considering a soldered joint and comparing SnPb with lead-free, we can make the following calculations. If the shapes are the same, then the mass of the lead-free alloys will be lower due to their densities. For a SnCu soldered through-hole pin connector, the mass solder will be:

(P SnCu x P SnPb) x massSnPb

Because the joints look different, wetting might be poorer and the angles of the joints are different, we must verify if the calculated mass difference is approximately equal to a real mass increase of the soldered joints.

For confirmation, we soldered a board with connectors (overall 192 pins/board) and weighed the difference before and after soldering (Table 1). The increase in mass is more or less the applied solder.

Flux

As in all soldering processes, flux plays a major role. Solderability and solder defects can be improved and reduced if the correct flux is used. If we implement a "green" soldering process, we use VOC-free water-based fluxes, which have some advantages as compared to alcohol fluxes.

Several experiments so far have proven that VOC-free fluxes show better results with lead-free solders than no-clean fluxes. Particularly with respect to residues on the board and solderability, they are preferred. One reason is that the amount applied to the boards is less. The activators and chemicals in the flux react more aggressively in water than in alcohol. Although VOC-free fluxes are more expensive, the overall costs with these fluxes will be approximately the same or even less because the total amount used for soldering will be reduced.

If solderability improves, the amount of rework will decrease. The reduced amount of flux also results in decreased maintenance. Cleaning machine parts will be easier and can be done with hot water rather than chemicals and instruments.

However, the number of solder balls increases with the use of VOC-free fluxes. Part of this increase is due to higher temperatures in the process, which makes the solder resist softer. As compared to the tin-lead process, these solder balls are much easier to remove.

New VOC-free fluxes are currently being developed. Flux suppliers are attempting to dissolve rosin into the water-based flux, which results in a commendable decrease in solder balls. These investigations will continue because most flux suppliers have not yet succeeded in finding the right formulation.

Components

For many components, changing lead finishes will not be a major issue. If a big future demand for lead-free finishes occurs, the component suppliers will be more likely to switch over than they would currently. Because the techniques are already available, the prices of these components are not expected to increase dramatically.

SnAg and SnAgCu balls for ball grid arrays (BGAs) appear to be an acceptable substitute for SnPb. Alternative lead finishes for peripheral packages are under investigation, and reliability and tin whisker issues must be addressed. The higher process temperatures increase the demands on component moisture sensitivity performance and package integrity. Plastics that can withstand higher temperatures, such as 280 degrees C for 5 sec., are currently being designed and will push prices higher. Therefore, a reflow oven is needed that has a high accuracy (small delta T and good heat transfer) to run lead-free profiles that meet the specifications of the less expensive components. If the maximum peak temperature can be limited to 245 degrees C and all the solder brought above the melting point according to the lead-free solder paste specification, a reduction in component costs for the user may result.

Board materials

The first article of this series (April 2001) discussed that, in addition to banning lead, halogenated flame retardants from board materials will also be eliminated. Thus, new board materials with lead-free finishes must be designed with preferably higher glass transition temperatures (Tg) to withstand higher process temperatures. These new board materials, as well as the lead-free finishes, will affect costs. How much these costs will increase is not clear yet because most board suppliers are still optimizing board material selection and their fabrication processes.

Nitrogen

Reflow ovens. The necessity of nitrogen in reflow soldering is often discussed. Some processes require nitrogen because it improves wettability and results in better solder joint reliability. In other processes, nitrogen may result in more tombstones and is banned or controlled to a certain level.

Even if an inert atmosphere in a reflow soldering process may help, the question still remains if the costs are justified. In some countries, nitrogen is not that expensive; for example, in Germany, the cost is approximately $0.08/m3. In other countries such as Switzerland, nitrogen costs approximately $0.81/m3 and is very cost-effective relative to labor.

Preferably, an oven should be able to run in both air and nitrogen. For cost reasons, an inert atmosphere should be avoided. But, with new technology such as smaller and more complex designs, the ability to switch to nitrogen should exist.

No general statement can be made regarding nitrogen. Every process has its own specific issues and challenges. After implementing lead-free soldering with possible higher process temperatures, nitrogen performance and necessity must be reviewed. After a longer production period, the decision regarding air or an inert atmosphere can be re-evaluated.

Wave soldering. As with tin-lead solder, the lead-free solders oxidize quite rapidly when the solder is at liquidus and at a high temperature. Apart from inert soldering machines, a new oxide will be formed on the wave immediately after the oxide skin at the surface has been removed. Dross consists of cells of solder metal enveloped by an oxide skin. With lead-free solders, the oxides on the wave may be more detectable visually.

The oxides are more detectable for several reasons. First, the tin content in lead-free solders is higher than in tin-lead. By far, the most commonplace oxides on the surface of the solder are the tin oxides, tin-oxygen (SnO) and SnO2. Second, the temperatures are higher than in tin-lead soldering. Higher temperature results in more oxidation, which results in more dross.

The amount of dross can be reduced. Certain wave solder machines are fitted with a shaft sealing that eliminates dross formed at the pump shafts. Other dross is formed at the waves. By reducing the fall height of the wave, the amount of dross will be less. The fall height is the distance of the overflowing solder at the wave to the solder level.

The use of nitrogen will also provide some benefits. Nitrogen is cost-effective, and the amount of dross can be decreased. Because oxides are only a small part of the dross, the dross should be compressed, which will partly separate the solder metals from the oxide cells.

Energy Consumption

Reflow oven

The reflow process requires a lot of energy to heat up the printed circuit board (PCB), and after that, more energy is needed to cool down the board. Lead-free soldering requires different profiles and, therefore, a different amount of energy consumption.

In an experiment, we compared the power consumption of lead-free processes with traditional SnPb (Table 2). Using a datalogger, profiles can be developed of the time-temperature behavior of the assembly during the process. The linear profile for SnAgCu is shown in Figure 1. The area underneath the heating curve is related to the energy needed to heat the assembly.

In another experiment, we used a specific reflow oven and a typical board assembly to set up the profiles. To determine the power consumption, we installed a measuring device in the machine. The power consumption for every process was recorded (Table 3).

Figure 2 shows the power consumption of a reflow oven during the day for a one-shift process. The SnPb profile is compared to the linear SnAgCu profile. From the linear profile, we learn that a long time above liquidus results in increased intermetallic growth, which is not preferred for reliability reasons and also has a big impact on power consumption. The SnAg profile has high peak temperature settings, which require a lot of power to maintain the setpoint.

Wave soldering

In the wave soldering process, two areas will show an increase in energy consumption due to higher melting points and process temperatures. The first increase is in the preheating of the assembly. If we compare a no-clean flux application to a VOC-free water-based process, we will find an increase in energy consumption, due to higher preheat temperatures, to a maximum of 25 percent.

Second, because the solder temperature is higher, the solder pot needs more power. If we compare an extremely high soldering temperature of 280 degrees C with the regular SnPb temperature of 250 degrees C, we find the data listed in Table 4.

Figure 3 shows the power consumption of a specific solder pot during the day for a one-shift production process. Figure 4 shows the power consumption of a similar solder pot during the day for a two-shift production process.

Operation Costs

Throughput

In general, the lead-free wave soldering process requires longer contact times to achieve good wetting of the solder. If necessary, the machine can be fitted with a different wave-former. If proper wetting is still not achieved, then the conveyor speed must be reduced. However, a reduced conveyor speed may result in lower throughput.

Repair-failure rate

The solder joints look different and exhibit different shapes. From what we have seen in lead-free implementations, the number of defects does not increase. Nevertheless, new defects such as fillet lifting do occur. Thus far, reliability tests have not shown lower quality due to fillet lifting, so these joints do not need to be repaired. For repair work, an increase in oxidation of the soldering iron tip also occurs.

Maintenance

An increase in maintenance due to lead-free soldering should not be expected. VOC-free water-based flux may even reduce maintenance time and intervals as compared to no-clean fluxes.

For reflow soldering, a good flux management system will reduce maintenance costs. The new solder paste will have different fluxes and will evaporate other residues at higher temperatures, but a major increase in maintenance intervals or time will not result.

Process Improvement

After implementation, the process must be continuously controlled, improved and redesigned to save costs and be competitive. Therefore, engineers and all those responsible for the lead-free process should be aware that new materials, process and machine upgrades will be introduced in the near future.

New materials

Although some companies have been lead-free soldering for two years or more, a few remarks should be made regarding their alloy selections.

Staying within the specification limits of the alloy, particularly with SnAg, is very difficult if copper is present in the board material; for example, copper pads with organic solderability preservative (OSP) coating. Increasingly, companies are selecting SnAgCu as the alternative for SnPb, and SnCu is used only in wave soldering for cost reasons.

Tin-zinc (SnZn) and tin-zinc-bismuth (SnZnBi) are still outsiders for reflow soldering for the foreseeable future. If solder paste suppliers are able to design a superior flux system for the solder paste and succeed in eliminating the problems of oxidation with zinc-bearing alloys, these alloys might spark renewed interest due to their low melting points and costs.

Board layout

Most companies using lead-free soldering did not modify their board designs for these new solder processes. If more knowledge becomes available about lead-free solder joints, such as reliability data in relation to design dimensions, design rules may need some updating.

Deming cycle

Because new developments will be introduced in different areas of the lead-free soldering process, a continuous use of a standard improvement model, such as the Deming cycle, is required (Figure 5). Follow this model to implement (or decide to not implement) new developments. For example, a new flux will be introduced to the process. The steps to follow are:

-Plan: Plan an experiment to find out if this flux will improve quality, reduce costs or achieve another goal that has been selected.

- Do: Run the experiment.

- Check: Analyze the output of the experiment and judge if this flux meets expectations.

- Act: Implement the flux in the process and keep on monitoring the quality.

Conclusion

With the Deming cycle, the end of this implementation program has been reached. Although lead-free soldering is a hot item, most manufacturers are still gathering information or have only just begun their first trials. Hopefully, this series of articles will help you develop a robust, repeatable lead-free soldering process that delivers consistent high yields.

Gerjan Diepstraten is a senior process engineer with Vitronics Soltec BV in The Netherlands; e-mail: [e-mail].

http://www.circuitsassembly.com/

Copyright [copyright] 2001 CMP Media LLC