July 1st, Here we come!

Well July 1st will be here on Saturday. Is everyone RoHS compliant? Probably not. As a materials supplier we are often asked to assist our customer’s with the Pb-free transition. Most of the large OEMs and CMs have made the switch (or have Pb-free lines and material sets ready to be used). There are still a surprisingly large number of small to medium sized manufacturers that are still under the impression that the EU RoHS legislation does not apply to them. After asking a few questions you find out that in several of these locations they are building a sub assembly that goes into another product where they have materials heading into the EU. Now they are in a mad scramble to be compliant. Luckily the electronics industry has quite a bit of information published over the last few years that can assist these companies (for example: SMTA Knowledge Base, IPC Pb-free web site, material suppliers, and iNEMI).

One of the most common excuses that customers use are related to the movement to tin (Sn) plating. Namely, tin-whiskers.

A great resource for tin-whisker information is the NASA website. iNEMI has been doing a lot of work on trying to understand the mechanism and how to test for whiskering. That work has evolved into a standards and test methods that can be used to help understand and test for whiskering (Environmental Acceptance Requirements for Tin Whisker Susceptibility of Tin and Tin Alloy Surface Finishes [JESD201], Current Tin Whiskers Theory and Mitigation Practices Guideline [JESD22A121.01] , and Current Tin Whiskers Theory and
Mitigation Practices Guideline [JP002]
).

Certainly, tin-whiskers are of great concern for high-reliability products (such as military, and aerospace). For many commercial level products this does not need to be as much of a concern. Of course using these new standards and guidelines can help both customers and manufactures truly understand the risks involved rather than making snap judgments.

Good luck to you all!!

Califorina RoHS, 2007 or 2009?

Recently there has been some talk that the CA legislation may not take effect until 2009.

The Californian Bill SB50 is due to take effect on 1 January 2007 emulating the RoHS ban on Cd, Hg, Pb, hex Cr (not flame retardants) for ‘covered’ electronic products i.e. …cathode ray tube, cathode ray tube device, flat panel screen, or any other similar video display device with a screen size that is greater than four inches in size measured diagonally and which the department determines, when discarded or disposed, would be a hazardous waste…

A good source of the background material for this legislation can be found here:

There is another Bill (AB2202) is proposed that will extend the scope of the regulation to mirror that of the EU RoHS i.e. extend coverage from items with screen size over 4inches to include all the same product categories as the EU version.

AB2202 is still being discussed. The original plan was for it to take effect from the same date as SB50 (Jan 2007) but recent amendments were proposed to delay the implementation date for the extension to all EU RoHS categories until 2009.

As of yet the web site does not mention anything about the law being delayed.

Glass Transition Temperature

Another common misconception (at least with adhesive materials) is the concept of Tg (glass transition temperature). Most of the adhesives used in the electronics industry are thermosetting materials. As such they do not “melt” but have two physical properties that are related. The Tg which is the temperature at which the material softens and the decomposition temperature – the temperature at which the material breaks down. Materials can “function” above their Tg, but their physical properties will be altered. For example, silicone materials (such as bathroom caulk, or conformal coatings) often have Tgs that are below room temperature. Yet they still adhere, protect from the atmosphere (moisture), etc. Since most adhesives used in electronics (FR-4, underfill, SMA) are epoxies, it is useful to understand how these materials are affected above the Tg. The most significant effect on epoxies is that the CTE increases significantly (typically 2-3 times). That doesn’t necessarily mean that a device encapsulated/underfilled/coated with one of these materials will fail. There are other physical properties that come into play as well (especially when it comes to reliability) such as modulus, and adhesion. The best way to confirm that a material is suitable for the application is to test it using known established test methods.

Making Sense of Technical Data Sheets

With the advent of the web engineers and scientists at different companies often make use of the vast knowledge available “out there” to look up and research materials. This can be a great first step in narrowing down a search in order to produce a small subset of materials to choose from, as long as the search is done with some up front knowledge. When comparing data sheets from different vendors it is important to compare the test methods as well. For example viscosity measured using different methods can give different results. If possible, try to match up test methods, speak to the vendors and see if they have data on their materials with the other methods – to make it easier to compare. Finally, even after comparing data sheets it is critical to test materials in your application. Technical data sheets should be used as guidelines as way to down select from a range of materials in order to develop a smaller subset of materials for actual physical testing.

When Adhesives Don’t Stick

The one property everyone expects from an adhesive is that they stick things together. It is true when you buy some sort of glue for your house and it is true when you are using an adhesive on your PCB.

Often we are asking the adhesives on the PCB to carry out some additional functions: conductivity (thermal or electrical), reliability (thermal cycling or drop/shock), environmental protection, etc. First and foremost, these materials need to adhere … sometimes they don’t.

There are a variety of reasons why this occurs. The most common that we see (especially when dealing with SMAs and thermal adhesives) is contamination on components. Components are typically molded – sometimes there is mold release left behind on the components. Mold release is designed to not allow adhesives to stick – therefore the adhesive can fail. This is often observed when you have components from one reel falling off in a wave soldering process, but none of the other components. The adhesive isn’t smart enough to know when or where to fail. These contaminates can be confirmed by examining the components for these compounds.

The most common method is to use Fourier-Transform Infrared (FT-IR) spectroscopy. This technique can identify chemical compounds on the surface of the components and can help to determine why an adhesive is not sticking to a surface.

So next time you have an adhesion problem, look to the surfaces as well as adhesives.

Backwards Compatibility – When Worlds Collide

The July 1, 2006 deadline for RoHS implementation in the EU means different things to different companies. Many of the larger OEMs and CMs have been working at selecting materials and developing processes for several years. Within mid-sized companies there is more variation, some have just started, some have not even started figuring that they are exempt or don’t have to worry about the ruling since they don’t believe their products are going to Euorpe (this is equivalent to sticking your head in the sand – RoHS and RoHS legislation is coming to other countries as well).

In either case, companies are now finding that they have to deal with mixed-metal systems. Where we get the most questions is over using Pb-free bumped area array devices with SnPb solder paste and process. Some early work reflowing these devices using SnPb conditions [D. Hillman, et al., “The Impact of Reflowing a Pb free Solder Alloy Using a Tin/Lead Solder Alloy Reflow Profile on Solder Joint Integrity,” International Conference on Lead-free Soldering, CMAP, Toronto, Ontario, Canada, May 24-26, 2005, http://www.CMAP.org] showed failures as early as 200 thermal cycles. In his paper Mr. Hillman showed some great images of solder joint microstructure that varied within the solder joint, it wasn’t homogenous as is common for most solder joints.

Jasbir Bath from Solectron presented some additional reliability data [Jasbir Bath, et al., “Reliability Evaluation of Lead-free SnAgCu PBGA676 Components using Tin-Lead and Lead-free SnAgCu solder paste,” Proceedings of 2005 SMTA International, Chicago, IL, http://www.smta.org]. This paper shows that mixed metal systems, even when reflowed at Pb-free temperatures are not as reliable as Pb-free/Pb-free or SnPb/SnPb systems.

Some recent work by John Pan (San Luis Opispo) and Bath [Pan, et al., “Lead-free Soldering Backward Compatibility”, IPC/JEDEC Pb-free Conference, San Jose, 2006, http://www.ipc.org] calculated the melting point of the mixed metal system. The idea behind this study was to figure out what temperature was needed in order to get complete melting of the solder ball and paste and get complete mixing – without having to go to Pb-free soldering temperatures. For example, a 1.27mm pitch BGA with a bump diameter of 0.028″ printed with a 0.005″ stencil (0.027″ round aperture) melts at 218C whereas a 0.5mm CSP with a bump diameter of 0.010″, printed with a 0.004″ stencil (0.011″ square aperture) melts at 206C.

Bottom line: Mixed SnPb and Pb-free solders is not a good idea. If you have no choice, make sure that you have a high enough temperature at the solder joint to permit complete melting and dissolution in order to get the best reliability. You also need to realize that this “best reliability” is not likely to be as good as “pure” metal systems.

RoHS here, RoHS there, RoHS everywhere?

I had the opportunity last week to attend the IPC/JEDEC Pb-free conference in San Jose. Having attended many of these conferences over the past few years it is interesting to see the changes in the types of presentations. Early presentations were more about what material can we use? Can it be done? Now we have more presentations on the legal nature of the implementation, case studies on current implementation, backwards compatibility (mixing Sn/Pb and Pb-free – more on this next week), and how it impacts other parts of the process (e.g. rework). One aspect that many engineering people have not paid attention to is the legal aspects, specifically what this means beyond the EU RoHS. At the conference, one presenter from Design Chain Associates presented on the China RoHS legislation where they are closely mimicking the EU RoHS legislation – but requiring Chinese labs to do the certification testing! Also, the RoHS legislation is being looked at by several states in the US. In California there is proposed legislation that points directly at the EU RoHS legislation as a the model. This is supposed to be enforced by January 1, 2007!! Therefore, even if you aren’t shipping into Europe you should be aware and look at implementing Pb-free solutions for your products.

Don’t Forget Thermal Profiles for Adhesives

With the push to Pb-free soldering there is an increased focus on profiling for solder pastes. Many people in assembly realize the importance of using this critical measurement when defining and monitoring a process.

What is often not known is the importance of profiling for adhesive curing. Looking at a typical data sheet for an underfill, surface mount adhesive, potting material, etc. — there are recommended cure conditions. This is the recommended cure time and temp for the adhesive in order to obtain optimum properties (the ones listed on the datasheet). This means the adhesive itself needs to see this time and temperature.

For example, an underfill has a recommended cure of 5 minutes at 150°C. This underfill is placed under a metal-lidded BGA on a 0.092” thick board (that contains a power plane and a ground plane) and then placed in an oven with 20 to 30 other boards of the same configuration, set at 150°C for 5 minutes. To the surprise of the manufacturer the underfill did not appear to have cured.

Subsequent to this, a profile was taken of the area under the components found that the underfill was attaining only up to 130°C as a peak. Using a standard profiler, we found that once loaded the boards needed to remain in the oven for 15 minutes in order for the underfill to attain 150°C for 5 minutes.

Therefore, it is important to perform thermal profiling for adhesives as well as for solder pastes in order to produce reliable parts.