Compared to other solid materials, metals are widely known for their superior heat transfer capabilities. However, these capabilities come at the cost of weight, coefficient of thermal expansion (CTE) mismatch and freedom of design due to manufacturing constraints. Hence, the desire for thermally-conductive plastics. On the other side of the spectrum, we would like to have plastics that are much less conductive for insulation purposes. Both needs can be addressed by using fillers. Of course, applying fillers not only affects the thermal conductivity but also the strength, friction, stiffness and cost. A very important and old use of fillers outside the thermal domain is in mechanical reinforcement. Common fillers are: alumina, aluminium, brass, glass, graphite, magnesium oxide, stainless steel, calcium carbonate, acetylene black. Fibers, flakes, powders and microspheres are the most widely used.
Let us make a distinction between very low-k (k<0.1 W/mK), relatively high (k<10 W/mK) and high-k (k>10 W/mK) filled-plastics. Very low-k plastics, filled with hollow microspheres, are used in increasing quantities in a variety of applications because of improved properties, such as reduced density and cost, increased stiffness, improved compressive and impact strength, reduced shrinkage and increased dimensional stability of structural parts, good thermal and chemical resistance, low thermal conductivity and dielectric permeability. Commercially available glass, polymer, ceramic, and carbon (uncoated or with metal or ceramic coatings) hollow microspheres can be used as fillers in various polymers.
In the second category we may find ‘traditional’ fillers. The normal range of polymers is from 0.17 to 0.35 W/mK. We may expect a 20-30% improvement by glass-reinforcement, but this filler is often used for non-thermal purposes. Graphite and metal fillers do a much better job. These powders must be used at high loadings before having a significant effect. For example: polyimide with 40% graphite: 1.7 W/mK, and 40% loading with aluminium flakes shows a factor of 5 improvement. Some typical data for rubber (important for the tire industry) are: aluminium powder with a factor of 3 improvement and carbon black with a factor of 2 improvement. By the way, it does not make much sense to increase the thermal conductivity of the filler beyond a certain level. Going from 40 to 400 W/mK increases the composite’s thermal conductivity by only 5%.
The third category, the one with high-k materials, uses high-aspect-ratio fillers, such as graphite fibers, metallized glass fibers, aluminium fibers and flakes. Unfortunately, there is a limit due to processibility. Typically, aspect ratios of the fibers are limited to 40. Very promising results have been realized by the use of highly heat-conductive pitch-based carbon fibers. In-plane values of over 300 W/mK are quoted. Ultra-high thermal conductivity polymer composites are being developed, using mats of vapor-grown carbon fibers impregnated with epoxy. Values up to 660 W/mK are reported [1]. Other research programs focus on carbon nanofibers, single-walled carbon nanotubes (SWCNT) and multiwalled (MWCNT) carbon nanotubes. Here we talk prices of $300/kg for MWCTN and $300,000/kg for SWCTN. Fortunately, significant progress is expected from a novel technique of making SWCTN without metal catalysts [2].
The literature abounds with theoretical approaches to predict the effects of particle size and shape and filler percentage. The following parameters play a role: filler conductivity, plastic conductivity, shape and aspect ratio, volume loading, packing and orientation, and bonding. Additionally, density and specific heat are affected. The interested reader is pointed to correlations by Progelhoff and Lewis and Nielsen (see e.g. [3]). Of course, the theory is only of interest to people who are developing composites. For the majority of designers, it is the result that counts.
Besides their low density and tunable CTE, thermally conductive compounds have another big advantage over light metals: they can be moulded in virtually any shape. An impressive list of companies is available, pointing at a growing interest in these materials: RTP, Ticona, GE Plastics, DuPont, Schulman, LNP, Cool Polymers, Polyone and more.
Values for commercially available plastics range from 1 up to 100 W/mK. The higher values are usually associated with materials that are also electrically conductive. However, the biggest hurdle in general acceptance is the price. They are at least two times more costly than metals, and for the highest range of thermal conductivities one pays a lot more.
A final remark about the application field for ‘standard’ thermally-conductive plastics. When, at the system level, thermal conduction is the limiting factor, metals are to be preferred. When convection is limiting, thermally-conductive polymers may be a better choice. Table 1 provides a summary of materials discussed.
Table 1. Thermal Conductivity at Room Temperature for Filled Plastics
|
References
- http://www.ptonline.com/articles/200711cu3.html
- Chen, Y., and Ting, J., “Ultra High Thermal Conductivity Polymer Composites,” Carbon 40, 2002, pp.359-362.
- Guyer, E., “Handbook of Applied Thermal Design,” Chapter 6, CRC Press, Boca Raton, FL, 1999.
