QFN Thermal Pad Voiding: Causes and Control

A QFN thermal pad is a large, flat joint with no leads to carry heat away, and it is the part of the assembly most likely to trap gas. The void that results does not always show up as a defect, which is why the specification has to state what is acceptable and the process has to be measured rather than judged by appearance.

Why the Thermal Pad Behaves Differently

A leaded joint forms a fillet that exposes the solder to air and lets volatiles escape along the lead. The thermal pad sits flat against the land with the package body above it, so the only path for gas is through the molten solder or off the edges of a large area. Any flux that volatilizes under the centre of the pad has to travel a long way to leave.

The pad also carries a large share of the joint’s thermal path. Voids reduce the contact area, which raises the thermal resistance of the joint and concentrates heat in the remaining metal. That is why a void standard for a thermal pad is often written in terms of thermal performance rather than as a simple percentage.

How Voids Form Under a QFN

Three mechanisms dominate. Entrapped air from the paste, introduced during mixing or during a poor print, expands as the assembly heats. Flux volatiles boil and form bubbles that cannot escape through the surrounding paste. And moisture absorbed by the laminate or the package flashes to steam at reflow temperature and pushes through the joint from below.

Each mechanism responds to a different correction: entrapped air to paste handling, volatiles to flux chemistry and profile, and moisture to drying and to controlled storage. Voiding that appears only on some lots usually points to paste or to moisture, while voiding that appears on every board points to the aperture pattern or the profile.

Stencil Aperture Patterns for the Pad

A single large opening over a thermal pad prints a heavy deposit that traps gas at its centre. Splitting the aperture into a grid of smaller openings, commonly squares of 0.5 to 1 mm separated by 0.2 to 0.3 mm webs, lets gas escape between the columns and reduces the void area substantially. The pattern also reduces paste volume, which helps when the joint is overfilled.

Total aperture area is usually 60 to 70 percent of the pad area, and the webs must remain strong enough to survive printing. With a stencil of 0.10 to 0.12 mm thickness, an area ratio above 0.66 keeps transfer efficiency reasonable for the individual apertures in the grid.

Paste Volume and Its Target

The target paste volume for a thermal pad joint is roughly 60 to 80 percent of the volume of the space under the pad, so that the joint forms without pushing the package up and without leaving a large unfilled region. More paste does not reduce voiding; it generally increases it, because the extra flux has more volatiles to release and the extra thickness makes escape harder.

Measure the printed volume rather than the paste height, since the grid pattern makes a height measurement unrepresentative. Where the deposit is consistently above target, reduce the pad coverage in the stencil rather than lowering print pressure, because pressure changes affect the fine-pitch apertures on the same stencil.

Flux Chemistry and Outgassing

The flux in the paste is the source of most of the gas under a thermal pad, so the chemistry matters as much as the volume. A flux with a lower volatile content and a higher activation temperature releases less gas in the critical window and reacts more completely before the solder solidifies. Paste made for air reflow and paste made for nitrogen behave differently, and the choice should match the oven.

Paste that has been left on the stencil past its open time absorbs moisture and loses solvent, and both changes increase voiding. Rotating the paste on the stencil and respecting open time is a cheaper void reduction than any stencil change.

Profile and Soak Effects

The reflow profile controls how much gas is released before the solder solidifies. A soak that holds the assembly near 150 to 180 degrees Celsius for 90 to 120 seconds lets flux volatilize and escape while the paste is still porous, and it equalizes the temperature across a large pad so that the solder melts as one body rather than from the edges inward.

A short, hot profile may look better on the joint appearance and produce worse voiding, because the volatiles are trapped by solder that has already melted. Where voiding appears as large central bubbles, the soak is usually too short; where it appears as scattered small bubbles, the paste or the print is the likely source. Recording the profile alongside the X-ray result makes that distinction checkable after the fact rather than only in the moment.

Measuring Voiding by X-ray

Voiding is measured by X-ray, and the measurement has to be defined: whether the reported number is the void area inside the joint outline, the largest single void, or the total void area across the pad. A single large void and many small ones of the same total area have very different effects on thermal performance, so both figures are worth recording.

Resolution matters for small voids, and an oblique view helps confirm whether a dark area is inside the joint or is a shadow from a nearby feature. Setting up the inspection program with the same window and threshold for every board makes the numbers comparable across lots, which is what turns a measurement into a control. Where the same void must be compared between shop and customer, agree the window and the threshold up front, because one image read with a looser threshold can differ by several percentage points.

Acceptable Void Levels and Their Basis

Acceptance depends on what the pad does. A thermal pad that carries heat into a plane is usually judged on void area, with limits in the region of 25 percent total and a restriction on any single void, while a pad that is primarily a ground connection may accept more. The limit should be stated on the drawing with the measurement method, not left to inspection practice.

Where the assembly is thermally critical, the qualification should include a thermal measurement rather than only an X-ray image, because the relationship between void percentage and junction temperature depends on the geometry. A number derived from a thermal test is far easier to defend than a percentage chosen by convention.

Preventing Voiding in Production

Control the three inputs that create voids: store and handle paste within its window, print the grid pattern at the target volume, and run a profile with enough soak to release volatiles. Then verify with a periodic X-ray check rather than on every board, since voiding changes slowly and a per-lot sample detects the shift.

Where voiding rises suddenly, look first at the paste lot and the open time, then at the profile recording for the affected period. Comparing those records against the X-ray results is usually enough to identify the change, and the fix is then a documented adjustment rather than a trial-and-error sweep of the process.

FAQ

Why is a grid aperture better than one large opening? The webs break the deposit into columns, giving flux volatiles a path to escape, and they reduce paste volume so the joint is not overfilled.

What void level is acceptable under a QFN thermal pad? Often around 25 percent total void area with a limit on any single void, but the number should follow the thermal requirement and be written on the drawing with its measurement method.

Does more solder paste reduce voiding? No. Excess paste adds volatiles and thickness that make escape harder. Target roughly 60 to 80 percent of the joint volume instead.