Taking the Heat: Transforming Thermal Management through Phase Change Materials
BY JASON BRANDI, Henkel
hen it comes to electronic products, the more power, speed, and functionality jammed into a smaller package, the better. While these products are a dream for end-users, they can be a thermal management nightmare for design engineers. As the power increases, so does the heat. But, because of end-product size limitations, design engineers aren’t able to use larger heatsinks to address the issue of thermal management. In search for an overall efficient, yet compact, thermal solution, more emphasis is being placed on thermal interface materials (TIMs).
TIMs are usually supplied in two forms: solid pads formed with cured silicone rubber, and liquid compounds in the form of greases or pastes. Both types of materials contain a filler that increases thermal conductivity and, while pads are often preferred for their ease-of-use and long-term stability, greases or pastes offer better wetting, which improves overall thermal performance. With greases or pastes, however, there can be trade-offs for the gain in wetting performance. First, these materials can be somewhat messy in a production environment. Additionally, continual thermal cycling of greases can lead to liquid migration, leaving only the filler in place, which eliminates surface wetting and leads to possible field failures. The differing expansion rates of the materials on either side of the interface can create a “pumping” effect, which results in increased thermal impedance and inadequate thermal transfer.
Because they solve some of the problems associated with thermal greases, phase change materials have performed successfully in applications such as telecom base stations, electric trains, consumer electronics, and computers. These materials are used to replace air between the imperfect surfaces of a device and heatsink with a more thermally conductive material that will efficiently transfer heat from the device to the heatsink.
Phase change technology features a wax-based system that is solid at room temperature but becomes liquid once the excess heat of the device pushes the material past its melting point. This versatility provides the engineer with a material that is manufacturing-friendly while also delivering the performance necessary to meet thermal design requirements. Unlike thermal grease, the phase change compound will not migrate or “pump out” of the interface.
Next-generation Phase-change TIMs
One of latest classes of phase-change TIMs is based on the proven platform of phase change technology, but possesses differences that provide a combination of performance and ease-of-use. Previous products featured a coating of phase-change compound applied to an aluminum or polyimide substrate. This generation of products, however, are coated directly onto release liners (Figure 1) without the need for a substrate, improving performance. When placed in an interface and pushed past its melting point (45°C) by the electronic component’s heat emission, the compound becomes liquid and flows to fill all gaps and surface imperfections, removing air at the same time. Because there is no substrate to block the flow, the final interface thickness will be as thin as possible and based solely on the geometry of the parts. This thinner interface results in more efficient heat transfer and the spherical aluminum filler ensures heat transfer from the device to the heatsink with ultra-low thermal resistance.
Figure 1. Next-generation phase change TIM.
Because the product is offered in a thickness that exceeds that needed to fill all gaps, the pad will generally be somewhat smaller than the area to which it is applied. For example, an application on a 31-mm-square CPU lid would require, depending on the flatness of the heatsink, a square pad between 15 mm and 20 mm. An effect of the smaller dimension is that the force from the retaining screws, bolts, or clips is exerted on a smaller area, which leads to higher pressures and increased flow properties at the compound’s phase change temperature. Due to this phenomenon and the material formulation, low thermal resistance is achieved even at low mounting pressures. This results in reduced stress on the components and limits potential damage to the device.
Pads are easily removed from the base liner and have an inherent tackiness that allows easy adhesion to the heatsink surface. A second protective liner on top of the pad protects it during any additional shipping or secondary processes and can be removed at final assembly or after attachment to the heatsink. Older-generation products require elevation of the heatsink temperature to melt the pad for proper adhesion. But the next-generation PCTIM allows for this entire process to occur at room temperature.
Next-generation PCTIMs are also reworkable, allowing for easy disassembly of the component and heatsink. At room temperature, the material is brittle and the bond formed by the wetting action can be broken when the clips, screws, or retaining mechanisms are removed. This feature suits computing applications where the heatsink must be removed prior to releasing the CPU from its socket. In the case where the original heatsink and component are being reassembled, the integrity of the material will be maintained for several placements with no loss in thermal performance. The material will simply re-melt during the first power cycle and flow to fill all voids.
This latest class of PCTIMs have already found acceptance in the computer industry, where efficient processor cooling is a major challenge to overcome if the current pace of development is to continue.
Development work is ongoing to evaluate the optimal part configuration for a range of power modules, insulated gate bipolar transistors (IGBTs), etc., with the goal of this project being the pre-application of the pad to the device prior to shipment, benefiting both the design engineer and component supplier. The design engineer will no longer have to shoulder the responsibility for interface material selection, which will reduce required testing, and limit failures for incorrect material choice. Component suppliers can now offer an integrated solution, a value-added product, and will have limited warranty conflicts due to inferior interface material selection or application at the customer site. New uses for this PCTIM technology are surfacing every day, yielding new standards for demanding cooling applications.
JASON BRANDI, global product manager, thermal interface materials, may be contacted at the electronics group of Henkel, 15350 Barranca Parkway, Irvine, CA 92618; 949/789-2500; E-mail: email@example.com.
Grease or Phase Change Material: A Tale of Two TIMs
By Andy Delano, Ph.D., Honeywell Electronic Materials
We’re already familiar with the thermal challenges facing the semiconductor industry: power-hungry microprocessors require sophisticated thermal solutions to keep them cool, and the thermal solutions, in turn, require special thermal interface material (TIM) to optimize heat transfer from the microprocessor. Without TIMs, two supposedly flat, smooth surfaces only make physical contact in as little as 5% of the total contact area.
So TIMs are important, but what variety should be used? Without any TIM, only air fills the micro-gaps between the processor and thermal solution, inhibiting heat flow and resulting in a temperature rise of 41˚C. A highly engineered TIM, either grease or phase change material (PCM), reduces this to a very manageable 5.5˚C or 5.2˚C, respectively. Although it is clear that both grease and PCM outperform air, there doesn’t seem to be much difference between the two.
Many factors go into the decision of whether to use grease or PCM, including ease of application, re-workability, cost, and shelf life, but perhaps the most important factor is reliability. Every time a computer’s power is cycled, it heats up and cools down. This causes temperature swings that, in turn, cause the materials inside the computer to flex open and squeeze, because most materials expand with increasing temperature. Temperature cycling can wreak havoc on thermal interface materials, and greases are particularly vulnerable due to their low viscosity.
Figure 1 shows the trend in thermal performance of a typical PCM and typical grease as a result of repeated temperature cycling. From zero to 500 cycles, both materials show improved performance because they are slightly compressed with each cycle. However, after 500 cycles, the grease’s performance starts to degrade, and by 1000 cycles its temperature has increased by nearly 2˚C. Conversely, the PCM’s high viscosity allows it to remain in place, thus ensuring reliable performance for many thousands of cycles.
In the past, some chose to use a grease because of its application flexibility which surpassed that of a pad-format PCM. With the advent of screen-printable PCM and its reliability, the scales seem to be tipping towards PCM as the optimum choice.
ANDY DELANO, Ph.D., advanced thermal management technical team leader, may be contacted at Honeywell Electronic Materials, 15128 East Euclid Avenue, Spokane, WA 99216; 509/252-2224; E-mail: firstname.lastname@example.org.