When researching thermally conductive adhesives, many engineers will immediately begin searching for the highest thermally conductive adhesive on the web. Simply digging through data sheet after data sheet, taking note of the occasional 3, 4, or even 5 W/mK product. However, in many cases this is a faulty approach and will not result in the best product for your application.
W/mK Is Only Part Of The Story
The bottom line is that the thermal conductivity value listed on an adhesive’s data sheet only tells part of the story. This post is dedicated to explaining the rest of the story and is inspired by a section of an old research paper titled “Advanced Boron Nitride Epoxy Formulations Excel In Thermal Management Applications.” The paper, which was published by Epoxy Technology, examines two different commercially available, thermally conductive epoxy formulations:
|% Filler (by weight)||68||30|
|Mean Filler Size (um)||300||3|
|Typical Viscosity (cPs)||250,000||17,000|
|Typical Thermal Conductivity (W/mK)||4||1.5|
The paper goes on to compare the thermal resistance values of each adhesive expected from the product data sheets.
Calculating Thermal Resistance From Thermal Conductivity
Ultimately for thermally conductive adhesives, thermal resistance determines the effectiveness of an adhesive to draw heat away from sensitive components, not thermal conductivity. The two are related by a simple equation:
Using the equation above, the listed thermal conductivity values of each adhesive, and some basic dimension assumptions (3.0 mil bond line and a bond area of 100mm square), they were able to calculate the expected thermal resistance for each product.
However, once actually tested, the two materials displayed very different values than expected:
|Listed Thermal Conductivity W/mK||Expected Thermal Resistance C/W||Actual Thermal Resistance C/W|
What is so striking about these results is that the material listed to have a higher thermal conductivity value is in practice outperformed by a seemingly less impressive alternative.
Why Are The Results So Off The Expected Values?
There are two main reasons why adhesive B was able to outperform the much more impressive appearing adhesive A.
- Bond line thickness: As we can see, thermal resistance is not only related to thermal conductivity, but bond thickness as well. Material A uses a substantially larger particle size than adhesive B. This large particle size helps A achieve a higher thermal conductivity value, however, limits its minimum bond line thickness in practice. Ultimately, this substantially thicker bond line results in a higher thermal resistance than expected.
- Void formation: Air is an extremely effective thermal insulator. In many cases an adhesive or other type of TIM has been chosen specifically to replace air along an interface, allowing for more efficient heat dissipation. In this case, the paper argues that the higher viscosity of epoxy A allowed air to become entrapped in the mixed liquid material prior to cure. Adhesive A has a much higher filler percentage by weight, which is why the adhesive is able to achieve such a high thermal conductivity level. However, this filler loading also results in a substantially higher viscosity, ultimately trapping air and increasing the adhesive’s thermal resistance in practice.