Fire kills, which is why flame retardants are used in almost all polymers. In the context of PCB assembly this obviously affects the board, but also impacts on the materials used for area array packages (often laminates) and the moulding compounds used to protect semiconductors. Nor must we forget the largest single use, which is in equipment housing.
Flame retardants traditionally use bromine compounds1, in particular polybrominated diphenyl oxides (PBDPO), often in combination with antimony trioxide. The halogens are only weak fire retardants, and antimony trioxide by itself is not a fire retardant, but when combined they become very effective. During burning, antimony trioxide promotes charring of the resin, which reduces the formation of volatile gases. At the same time, the heat of combustion promotes cross-linking between the antimony and the polymer, which creates a more stable thermoset. Also, at above 315°C, bromine forms hydrogen bromide, which reacts with the antimony trioxide to form antimony trihalides and oxyhalides that trap free radicals, inhibiting ignition and pyrolysis.
These are really good additives, so what is the problem? The difficulty is what happens at the end of product life. When materials treated with bromine flame retardants are recycled, they can generate dioxins and furans, materials that are extremely toxic even in low dosage. Whilst this problem can be overcome by increasing the process temperature, there is a second more intransigent problem, the contamination of the environment by the halogenated flame retardants themselves. These are released during both manufacture and recycling or disposal of products containing them, and have been found widespread in both fish and humans.
As a result of this concern, some materials are in the process of being banned, and others will need to be replaced. The main impact will be on equipment housings. Here there are a number of alternatives: aromatic bromine compounds that cannot generate dioxins or furans during combustion; phosphate esters; melamine and inorganic fillers. There is also the possibility of using inherently flame-retardant polymers such as modified polyphenylene oxide (PPO).
For boards, the most commonly used flame retardant is tetrabromobisphenol A (TBBPA). However, whilst this is still permitted under RoHS provisions, this decision can be revisited at any time. Possible substitute materials include various phosphorus and zinc compounds, and inorganic fillers.
The two phosphorus-containing alternatives are organic phosphates, which may be dissolved in the epoxy system and create foams that extinguish flames, and particles of red phosphorus itself. Being relatively large (20-30µm) these are visible during microsectioning, and there is a concern that laser ablation might evaporate the particles, leaving holes in the structure. Overall, there are concerns about the toxicity of phosphorus materials, their comparative lack of stability, and the fact that they leach easily into the water table.
Laminate manufacturers have in the past used a variety of inorganic fillers such as mica or clay, or the more 'sanitised' fillers aluminium hydroxide and magnesium hydroxide. However, in order to make a laminate flame retardant the percentage of filler has to be sufficiently high to leave little fuel to burn. Levels as much as 50% by weight have been reported. These high loadings adversely affect board properties, making the laminate stiffer and more brittle, reducing impact strength and tensile strength, and increasing water absorption.
Of the other approaches in development, zinc compounds have been used in some halogen-free formulations, with zinc borate the most widely used. This behaves like aluminium trihydroxide: when heated, both decompose to release water that forms an envelope around the flame, and also absorbs energy, lowering the temperature.
Alternative approaches are to use materials that are inherently flame-retarding, and silicone compounds have been evaluated for this: they have high heat resistance, are not toxic, and do not generate toxic gases during combustion. Lau also reports success with finely-dispersed silicon in polycarbonate resin, but this material is primarily for use in housings.
In summary, this is a topic that is likely to see substantial development in the years to come particularly as the Japanese are starting to promote 'halogen-free' in the same way that earlier workers promoted lead-free. The only difference is that the colour of choice, indicating that the material contains no halogens, is likely to be blue rather than green.
Notwithstanding the commercial implications, we have to be careful that we do not overreact to environmental pressures in this area. It has been commented that US consumer products are generally much less flammable than their equivalents in Europe, because they contain more brominated compounds. In fact, because of pressure to remove these materials from enclosures, it has been estimated that as many as 100 people have died as a result of fires in electronic equipment!