Thermal Interface Material and Heat Sink Assembly

A heat sink is only as good as the joint between it and the device. Two nominally flat surfaces touch at a small number of high points, and the air in the gaps between them is a poor conductor. The thermal interface material exists to fill those gaps, and the assembly process determines how well it does so.

The assembly step therefore has a direct effect on the thermal performance of the product. It is also one of the steps where a small variation, in the amount of material or in the mounting force, produces a large change in the result.

What the Interface Material Does

Thermal interface material fills the microscopic gaps between the surfaces and displaces the air. Its thermal conductivity is much lower than that of the metal on either side, which is why the layer should be as thin as the surface roughness allows.

The thickness is set by the flatness of the surfaces, the mounting pressure and the quantity of material applied. More material makes the layer thicker and the thermal resistance higher, which is the opposite of what most people expect when they add more paste.

Types of Material and Their Behaviour

Greases and phase change materials flow under pressure and produce the thinnest layer, at the cost of being difficult to apply consistently and of pumping under thermal cycling. Gap fillers are more compliant and tolerate larger gaps, and pads are the easiest to apply and the least effective thermally.

The choice follows from the gap that has to be filled and from the pressure available. Where the components are mounted with screws, a grease or a phase change material can be used; where the gap is set by a mechanical tolerance that cannot be tightened, a compliant gap filler is the practical answer.

Mounting Force and Its Distribution

Mounting force presses the material into the gaps and sets the final thickness. Too little force leaves a thick layer; too much can damage the device or the board, especially where the load is applied through the package rather than to the heat sink directly.

The distribution of the force matters as much as its magnitude. A single central screw tilts the heat sink and produces a wedge of material that is thick on one side and thin on the other, which is worse than a uniformly thicker layer. Multiple screws, tightened in a defined sequence and to a defined torque, distribute the load evenly.

Voiding in the Interface

Voiding in a thermal interface is a gap that the material did not fill. It can come from air trapped during application, from a material that does not flow into a recess, or from a pumping action during thermal cycling that moves the material away from the centre.

Its effect is local: a void in the middle of the interface raises the temperature of the whole device, because the heat has to flow around it. The thermal image of a heat sink after operation will show the signature of a void as a hot spot, which makes thermal imaging a useful verification tool.

Application Methods and Consistency

The material can be applied by screen printing, by dispensing, by a pre-formed pad or by hand. Screen printing gives the most consistent volume and is used in volume production; dispensing is flexible; hand application is the least repeatable and is used only for small quantities.

Whatever the method, the applied quantity should be defined and checked. The check can be a weight measurement on a sample, a visual assessment of the pattern, or an automatic inspection of the dispensed volume. Consistency is what allows the thermal result to be predicted rather than discovered.

Assembly Sequence and Mechanical Load

The heat sink is usually fitted after the board has been soldered, and the load it applies is carried by the board through the mounting points. A heavy heat sink on a thin board will bow the assembly unless the mounting is designed to spread the load.

The sequence also matters where the material needs to be cured. A material that flows before curing will move if the assembly is disturbed, so the handling between the application and the cure has to be controlled. A fixture is often used to hold the assembly during that period.

Thermal Verification

The verification is a temperature measurement on the operating product. A thermocouple on the device case, or a thermal image of the assembly, shows whether the interface is performing as designed. Comparing a sample from production with the design prediction is a direct check on the assembly step.

Where the measurement is worse than predicted, the usual causes are a thick layer, a void or insufficient force. Separating them requires a look at the assembly rather than a change to the material, and the measurement should be recorded with the assembly data, in the same way as any other thermal management result.

Design Measures That Help

Flatness is the property that determines how thin the layer can be. Specifying a flatness for the heat sink and the device, and checking it on the incoming material, is more effective than choosing a more conductive material for a poorly matched pair of surfaces.

The mounting design is the second measure. A spring loaded clip, a screw pattern that applies load evenly, or a compliant material chosen for the actual gap all make the assembly less sensitive to the variation in the parts. Those decisions belong with the mechanical design of the product and, where they affect the board, with the requirements recorded in the manufacturing tolerances.

Environmental and Ageing Effects

A thermal interface that works when the product is new can degrade with time. Greases can migrate away from the centre of the interface under repeated thermal cycling, and materials that cure can shrink slightly as they do so.

The behaviour should be considered when the material is chosen, and the verification should be repeated on a sample that has been through the thermal cycles that the product will see. Predicting the performance from the initial measurement alone is optimistic, and the same caution applies to any material that changes with temperature over time.

Measuring the Thickness of the Interface

The thickness of the interface layer can be measured after assembly in some designs, by measuring the gap between the heat sink and the board at a defined point, or by an ultrasonic method in the laboratory.

Where it can be measured, it is the most direct evidence that the process is under control, because it combines the quantity of material and the mounting force into a single number. Where it cannot, the controls have to be the application weight and the torque, and those should be checked on a sample from every batch.

FAQ

Is more interface material better? No. The layer should be as thin as the surfaces allow, which usually means less material than a hand application would use.

How tight should the screws be? To the torque specified for the mounting, applied in a defined sequence. Over tightening can damage the package or the board.

How can a void be detected? Thermal imaging during operation is the practical method. It shows the temperature distribution that the void produces.