Specifying the right thermal interface material (TIM) requires an understanding of how interdependent elements of a TIM formulation work together and when specific considerations take priority.
David Hirschi, Dow Corning Electronics and Advanced Technologies, Dow Corning Corp., Midland, Mich. -- Semiconductor International
As electronics manufacturers continue to pack components with greater power and functionality onto increasingly dense PCBs, the issue of heat dissipation is becoming a higher priority at every level. With performance, reliability and longevity all directly related to operating temperature, engineers are faced with a daunting challenge: how to effectively remove larger amounts of heat from components generating higher temperatures to ensure adequate device operation and service life.
Transistor reliability and longevity heavily depend on junction temperature, with reductions as small as 10–15°C potentially doubling device life.1 Lower operating temperatures also reduce gate delay, which contributes to higher operating speeds. Further, as devices continue to become more complex, conducting paths typically get longer, and the benefits of preventing performance loss at higher temperatures are likely to be even more apparent.
Elevated temperatures can also intensify other failure mechanisms when high humidity and electrical potential are present. Metal migration can cause dendrites to emerge from circuit traces, for example, and the close spacing of conductive lines can contribute to short circuiting. Some environments are especially harsh; the extremely wide temperature fluctuation experienced by many components can produce fatigue from repeated thermal expansion and contraction, eventually leading to failure.
The current need to more effectively dissipate heat from electronic packages is unprecedented in the industry, leading to the rapid growth of a class of products collectively known as thermal interface materials (TIMs). Essentially forming a thermally conductive pathway between a heat source and some other medium, such as a metal lid or heat sink, TIMs are commercially available in a number of physical forms, including adhesives, gels, encapsulants and pre-cured pads (Fig. 1).
1. Screen printing is one way of applying a thin, uniform TIM layer. For best thermal performance, it is crucial to use the thinnest layer of TIM possible that completely fills all gaps.
Thermal performance
In fact, there is a dizzying array of thermal materials on the market making a wide range of performance claims based on a confusing list of test methods, protocols and reporting standards. Specifying a TIM can be a difficult task for chip designers, and the tendency for many is to focus solely on bulk thermal conductivity. Thermal conductivity refers to the intrinsic ability of any material to transfer heat. For a heat transfer compound, that conductivity is primarily determined by the thermal properties of the filler material, including the amount and type of filler and the size, shape and distribution of the filler particles in the polymer matrix.
In general, the higher the filler loading, the greater the thermal conductivity of the interface material. TIMs made with smaller particles can achieve thinner bond lines and lower thermal resistance. The total volume fraction of fillers in a thermal material also has a significant effect on its thermal properties. Researchers have found that filler loading can be enhanced by as much as 50% by using an optimized combination of filler particle sizes, with significant conductivity gains as a result.2 The thermal conductivities of a number of common filler materials are summarized in Table 1.
While clearly an important property, thermal conductivity is only one part of the TIM decision and does not account for heat transfer across the interface between two or more surfaces. As such, it's important to understand the relationship between thermal conductivity and another often-quoted measure of thermal performance: thermal resistance.
In TIMs, this resistance is a function of bond line thickness, surface roughness and the hardness of the material forming the interface. Interfacial resistance (sometimes called contact resistance) can be minimized by using smoother interface surfaces and achieving improved contact between the TIM and substrate. Softer interface materials also tend to achieve more intimate surface contact, helping to eliminate voids that impede efficient heat transfer (Fig. 2).
2. If the interfacial gap is wide, bulk properties are dominant and a high conductivity is most important. For thin gaps, the dominant resistance is contact resistance, so a compliant material with better surface wetting capabilities may perform best.
For packaging applications, TIM bond lines are typically quite thin (<50 μm), which makes thermal resistance the highest priority in specifying a material. Because the resistance to heat flow across the interfaces is generally greater than the thermal resistance through the TIM itself, the temperature drop is larger across the various interfaces then across the bulk of the thermal material.
Measurement
Complicating the problem of specifying an appropriate TIM is the various ways to measure and report thermal properties. Most techniques can be categorized as either transient or steady-state. Unfortunately, reported thermal properties widely vary, depending on the test method, temperature, pressure, surface planarity and sample thickness. Thermal properties should not be compared across different test methods, which makes it very difficult to compare material data sheets from different vendors without knowing specific details of their testing techniques (Table 2).
Non-thermal properties
Even with these inconsistencies, if specifying a TIM was based on thermal performance data, the job would be fairly straightforward. In reality, the choice is complicated by other factors that influence material selection, including physical properties, processing ease and cost.
Mechanical properties, such as modulus and adhesion, are important design considerations because of their impact on contact resistance at the interface and long-term reliability. While there are many ways to measure and describe the modulus of a material, in the context of a thermal interface discussion, we generally refer to the material's hardness, compressibility or ability to conform to irregular surfaces. Product data sheets will typically list durometer or compression modulus, or provide a stress/strain curve. For package-level applications where bond lines are thin, selecting a TIM that can minimize contact resistance is essential. High contact resistance occurs because of substrate roughness, non-planar mating surfaces or height mismatches in the assembly (Fig. 3).
3. Micro-gaps from surface roughness and non-coplanarity do not allow adequate heat transfer across an interface. A higher modulus TIM, such as a thermal pad, may not entirely correct the problem. A wet material, such as a thermal grease, may be required to minimize contact resistance.
As an example, if a specific application is served by a high-modulus thermal material with a bulk conductivity of 10 W/m•K, the hardness of the surfaces at the interface is likely to prevent the TIM from adequately conforming to mismatched chip heights or irregular surfaces. If the modulus is too high, it can leave microscopic air gaps between the interface material and mating surface, creating an insulating layer that prevents efficient heat transfer.
In contrast, a low-modulus material can more easily compress to accommodate various chip heights and rough surfaces, with its flexibility a key advantage over stiffer materials. In this case, a soft TIM with a bulk thermal conductivity of only 2 W/m•K may very well outperform the harder material with far greater heat transfer potential. Intimate surface contact is the key (Fig. 4). Instead of using an elastomer interface material, an engineer might consider specifying even softer gels or greases. In general, the very thin bond lines of packaging applications dictate that contact resistance is the dominant consideration. The lower the contact resistance, the better the thermal performance.
4. In a multi-component application, care must be taken to ensure the TIM modulus allows for complete surface coverage of all components.
Like modulus, there are many ways of assessing adhesion. Lap shear and die shear tests are just a few. In addition to heat transfer, adhesives in thermal applications can perform two key functions. One is to serve as a mechanical fastener, the other is to prevent any microscopic delamination of the thermal material from the mating surface. Delamination usually occurs over time after repeated temperature cycles as devices are powered on and off. To eliminate the problem, TIM suppliers may include different additive combinations for various substrates, designing individual adhesives for superior grip on a specific mating surface, whether it be aluminum, copper, ceramic or another substrate.
Reliability data is used to demonstrate that no delamination has occurred after long-term testing. Common methods include HAST testing at 85°C and 85% relative humidity for 1000 hours, thermal bake tests at 150°C for 1000 hours, or thermal cycling tests of various duration from -40 to 100°C. If thermal resistance remains stable, the constant performance indicates that no delamination of the TIM has occurred.
In all cases, the conditions are meant to simulate end-of-life performance, as opposed to end-of-line performance, which refers to material properties immediately after assembly. Specifying a TIM requires that engineers have data on both so that they can estimate how much a material's thermal performance can be expected to deteriorate over time. With temperature's direct effect on component performance, the level of that degradation is critical to device reliability and service life.
In addition to its effect on thermal performance, the filler system has a direct impact on a material's mechanical properties. Large filler particles will prevent formation of a thin bond line, one of the key elements to effective thermal transfer. Smaller particles and higher filler loading also increase the viscosity of any wet-dispensed or screen-printed material, which may have positive or negative implications. Higher viscosity materials tend to remain in place better and are less subject to pump-out over time, but they can be more difficult to accurately dispense, leaving potential air gaps or hot spots.
Specifying the right TIM requires an understanding of how these interdependent elements of a TIM formulation work together and when specific considerations take priority, so engineers can arrive at a material choice meeting all of an application's criteria. Specifiers must look deeper than just bulk thermal conductivity to predict an interface material's performance over expected lifespan, considering parameters such as bond line thickness, modulus, long-term performance and processing.
Avoid the temptation to automatically equate a TIM's performance with its bulk conductivity. And don't assume that a material with the highest thermal conductivity also delivers the lowest thermal resistance. Above all, insist on comparing end-of-line performance to end-of-life properties, as demonstrated through accelerated testing, and don't hesitate to ask for details on specific test methods or protocols. A reputable material supplier will have access to such data and is likely to share it with prospective customers.
Author Information
David Hirschi is the global market manager for thermal solutions at Dow Corning Electronics. He has been with Dow Corning for 10 years. Hirshi holds a B.S. in chemical engineering from Brigham Young University (Provo, Utah) and an MBA from the University of Michigan (Ann Arbor).
Email: electronics@dowcorning.com (end)
|
