Thermal Pad Soldering and Void Control in QFN Assembly

A QFN package connects to the board through terminals around its perimeter and through a large exposed pad in the centre. The perimeter joints are visible before reflow and inspectable after, but the thermal pad sits underneath the body where nothing can be seen. Whether that central joint formed properly, and how much voiding it contains, determines both the electrical ground path and the thermal path out of the device.

Why the Thermal Pad Is Different

The thermal pad is a large area joint with a small standoff, which changes the physics of soldering. A perimeter joint is a small fillet that forms quickly and vents easily. The thermal pad is a wide, thin layer of alloy, and any volatiles generated beneath it must travel a long horizontal distance to escape before the solder solidifies.

That geometry is what makes voiding so common. It also means the pad carries most of the heat out of the device, so a void directly under the die raises the junction temperature. The same void that reduces the ground connection also degrades thermal performance, and the two effects compound in a power device.

Paste Volume and Aperture Strategy

Too little paste leaves an incomplete joint with poor thermal conduction, while too much causes the component to float, tilts the package and can push paste onto adjacent terminals producing bridging. The target is a joint that is full without being over-filled, and the practical recommendation is usually a paste volume that produces a final joint thickness of around fifty micrometres.

Achieving that reliably depends on how the solder paste is deposited through the stencil aperture. A single large opening deposits too much paste and traps volatiles; a segmented aperture, divided into a grid of smaller windows, controls the volume more precisely and provides channels for volatiles to escape during reflow. The segmentation pattern therefore serves two purposes at once.

Stencil Design for Thermal Pads

Typical practice divides the thermal pad aperture into a grid of squares or circles separated by narrow webs, with the open area chosen to deposit the required volume. The webs also reduce the risk of the paste sticking in the aperture during printing, because the area ratio of each small window is better than that of one large opening.

The pattern should be symmetric, because an asymmetric distribution of paste produces uneven forces during reflow and can tilt the component. Stencil thickness must be matched to the aperture, and where the stencil is stepped for other features, the step position should not interfere with the thermal pad region. Verifying the actual deposit volume with inline inspection is far more reliable than calculating it from the artwork, and the principles are covered in this guide to solder paste inspection.

Void Formation Mechanisms

Voids in a thermal pad joint come from several sources. Flux volatiles generated during reflow are the most common, followed by air entrapped in the paste during printing and by incomplete wetting of the pad surface. Moisture absorbed into the board or the component can also contribute, particularly if the assembly has been exposed before the final reflow.

The distinction matters because the remedies differ. Volatile-driven voids respond to the reflow profile, entrapped air responds to the printing process, and wetting-related voids respond to cleanliness and surface finish. Treating all voiding as a profile problem leads to repeated profile changes that never address the actual cause.

Reflow Profile Effects

The profile determines how much time the volatiles have to escape before the alloy solidifies. A slow ramp through the preheat and soak stages allows flux to activate and volatiles to leave gradually, while a fast ramp traps them under the pad. Extending the time above liquidus gives more opportunity for bubbles to rise and break the surface.

There are limits. Excessive time above liquidus grows the intermetallic layer and exposes the component to more thermal stress, and a soak that is too long can exhaust the flux before the joint forms. The useful approach is to adjust the profile in small steps while measuring the resulting void percentage, rather than making a large change and assuming improvement. The interaction between profile and defect formation is described in this overview of solder defect signatures.

Flux Chemistry and Outgassing

Flux chemistry controls how much volatile material is generated and at what temperature it is released. A flux designed for a thermal pad application typically releases its volatiles early, during the soak, rather than late when the alloy is already molten and the escape path is closing. Choosing a paste on the basis of its voiding behaviour rather than its solderability alone is a significant lever.

The paste deposit also matters. A paste with a high metal load and a low solvent content generates less volatile material, at the cost of a narrower printing window. Where voiding is critical, a paste specifically characterised for low voiding under a QFN thermal pad is worth the additional qualification effort, because it addresses the mechanism at its source.

X-ray Inspection and Void Measurement

Voiding is measured by X-ray, which shows the pad as a dark region with lighter areas where voids are present. The measurement is expressed as a percentage of the pad area, and modern systems can calculate it automatically once the pad boundary is defined. Definition of that boundary is the main source of disagreement between operators.

The measurement is also projection-based, so it reports the total void area through the joint thickness rather than the void distribution at a particular plane. Two joints with the same void percentage can behave very differently if one has a single large void under the die and the other has scattered small voids at the periphery. Reporting the largest single void alongside the total is therefore more informative than the total alone. The inspection techniques are compared in this guide to X-ray and optical inspection.

Acceptance Criteria and Their Basis

Acceptance criteria vary widely between customers, and the differences are usually not justified by test data. A common requirement is a maximum void area of twenty-five percent, with some specifications permitting fifty percent and others demanding less than ten percent for power devices. The appropriate limit depends on the thermal requirement of the specific design.

A defensible approach is to establish the limit from thermal measurement. Build samples with controlled void percentages, measure the junction temperature rise at operating power, and identify the point at which the thermal margin is consumed. That produces a criterion tied to the product rather than a number inherited from a general standard.

Void Control and Continuous Improvement

Control begins with the paste deposit, because variation in volume is the largest single contributor to variation in voiding. Trending the deposit volume per aperture and acting on drift addresses the cause rather than the symptom. Reflow profile verification at defined intervals confirms that the oven is delivering what the recipe specifies.

Where voiding remains above target despite a controlled deposit and a suitable profile, the paste chemistry and the pad surface condition are the next candidates. Changing one variable at a time, measuring the void percentage after each change, and recording the result turns a difficult problem into a series of resolvable experiments.

FAQ

What void percentage is acceptable under a QFN thermal pad? It depends on the thermal requirement of the design rather than on a universal figure. Twenty-five percent is a common specification, but a power device with a small thermal margin may need less, and a low power device may tolerate considerably more without any effect on the junction temperature.

Does a segmented stencil aperture reduce voiding? It usually does, for two reasons. Segmenting controls the deposited volume more precisely, and the gaps between segments give volatiles a shorter path to escape during reflow. The pattern should be symmetric so that the paste exerts even forces on the component as it melts.

Can voiding be eliminated entirely? In practice, no. Some volatile generation is inherent to the flux system, and a small amount of entrapped air is difficult to avoid completely. The realistic goal is to keep the total below the level at which it affects thermal performance, and to avoid large single voids under the die.