By Norman Quesnel
Senior Member of Marketing Staff
Advanced Thermal Solutions, Inc. (ATS)
For decades, thermal interface materials (TIMs) have been used as pathways allowing heat to flow from one location to another. TIMs are often part of cooling systems that remove heat from component dies by dissipating it into heat spreaders, such as heat sinks, and ultimately out of the dies’ surrounding enclosures.
As a product line, TIMs have continuously evolved, driven by market needs for higher thermal conductivity, lower thermal impedance, new applications and lower costs.
From a broad view, most TIMs fall into three material categories. Some are made from elastomers or other polymers with a thermally-conductive filler added. Other TIMs are solder-based. Like elastomers, these solder-based TIMs are soft and conformable to mating surfaces, filling air gaps that compromise thermal transfer. Finally, newer carbon-based TIMs have emerged that demonstrate superior performance, though many of these are not yet commercialized.
Here is a brief look at recent developments within the major TIM categories:
A New TIM Filler
A new generation of polymer-based TIMs uses boron nitride nanosheet (BNNS) fillers to enhance thermal conductivity. BNNS is a two-dimensional crystalline form of hexagonal boron nitride (h-BN), also known as white graphene. BNNS ranges in thickness from just one to a few atomic layers. It has a similar geometry to its all-carbon analog graphene, but some very different properties. For example, graphene is highly electrically conductive while BN nanosheets are electrical insulators.
Hexagonal boron nitride (h-BN) has other desirable properties, including a large surface area, high-thermal transport, and chemical inertness. The thermal conductivity of bulk h-BN can reach 400 W/mK at room temperature. 
A recent study from Rice University, which continues to expand on its original simulations of graphene’s effect on nanoscale heat transfer, demonstrated that an h-BN thin film composed layer-by-layer of laminated h-BN nanosheets can enhance lateral heat dissipation on a substrate, in this case glass. Thermal performance improved with the BN coating due to its anisotropic thermal conductivity. It had a high in-plane thermal conductivity of 140 W/mK for spreading and a low cross-plane thermal conductivity of 4 W/mK to avoid a hot spot beneath the tested device. 
Researchers have also created simulations showing that 3-D structures of h-BN planes connected by boron nitride nanotubes could transfer heat (move phonons) in all directions, whether in-plane or across planes. The number and length of the nanotubes connecting the h-BN layers have an effect on heat flow: more and/or shorter pillars slow conduction, while longer pillars speed heat transfer along.
With ever-increasing power and heat dissipation needs across the electronics industry, solder-based TIMs may be better suited to take the heat away from dies than thermal grease where electrical insulation isn’t required.
Issues with thermal grease include:
- Grease has a low bulk thermal conductivity of 3-12 W/mK. Some solder-TIMs provide a high bulk thermal conductivity of 87 W/mK.
- Over time, thermal grease tends to pump-out and migrate away from the center of the power die. It gets hotter and can fail prematurely. There is no pump-out with a solder-TIM.
- Over time, grease tends to bake-out and dry (becomes powdery), thereby increasing thermal resistance and reducing heat-dissipation effectiveness. With solder-TIMs, there is no bake-out. 
Recent solder-based TIMs developed by Indium Corporation include a new SMA-TIM (soft metal alloy). This is made from an indium solder base and offers uniform thermal resistance at lower applied stresses in compressed interfaces. It is provided as a compressible metal foil that can be used as a TIM between a heat source and a heat sink, heat spreader, or heat pipe.
The malleability of the indium minimizes surface resistance and increases heat flow (conductance). Over time, the malleability of the solder helps fill the interface gaps even better. Thus, thermal interface resistance decreases over time as opposed to thermal grease where the thermal interface resistance increases over time. 
Another newer indium-containing material has been developed for use in the TIM 1 position, between the die top and its case. The material is part of a system, developed by Indium Corporation, called mdTIM. It provides a thermal conductivity of 87 W/mK.
While pure indium metal has a superb thermal transfer rate, air or gas pockets (voids) can degrade the performance of the material. These voids are created by entrapped air or gasses produced by flux component evaporation that fail to escape during reflow.
Indium’s mdTIM uses a patented system of materials and reflow technology does not use flux so there are no outgassing issues.
The very high thermal conductivity of pure carbon has long made it attractive for use in TIMs. Today’s carbon-based TIM fillers include diamond, carbon nanotubes (CNT), graphite and graphene. Often these fillers are dispersed in a spreadable (grease-like) polymer matrix.
In some cases, different forms of carbon fillers are being combined. For example, highly thermally conductive CNT have been mixed with less expensive carbon substrates like graphite and graphene to reduce costs but still deliver very high thermal conductivity.
Recent research has been made with graphite nanoplatelets (GNP) in thin thermal interface layers. These studies concerned the through-plane and in-plane alignment of GNP in a spreadable matrix. When dispersed, the GNP fillers take a naturally in-plane alignment, meaning the great majority of heat flow is in parallel to an interface. However, at the same time, the desired through-plane heat transfer from one surface to the other is much less. 
A solution was found by adding spherical microparticles. Spherical Al2O3 and Al filler particles were tested. The hybrid filler formulations resulted in enhanced through-plane thermal conductivity by disrupting the natural in-plane alignment of the GNP. This led to the disruption of the GNP in-plane alignment and the improvement of the through-plane thermal conductivity of the tested thermal greases.
Costs and other factors pose development challenges to TIMs with carbon-based heat transfer schemes. But given the high thermal conductivity and various configurations available from carbon-based materials, these will likely be at the heart of many upcoming performance advancements in TIMs.
1. Indium Corporation, http://www.indium.com/thermal-management/tim/
4. Nanoscale, http://pubs.rsc.org/en/content/articlelanding/2017/nr/c7nr07058f#!divAbstract
5. IEEE Xplore, http://ieeexplore.ieee.org/document/6142407/%5D
6. Engineering 360, http://www.globalspec.com/FeaturedProducts/Detail/Indium/Thermal_Interface_MaterialHeatSpring/256323/1
7. Indium Corporation, http://www.indium.com/thermal-interface-materials/heat-spring/
8. Nature, https://www.nature.com/articles/srep13108.pdf