Void Formation in Solder Joints: 6 Causes and Fixes

Void formation is the trapping of gas inside a solder joint as the alloy solidifies, leaving pockets that reduce the load-bearing area. It is normal to find some voids in any reflowed joint, and the question is not whether they exist but whether they are large enough or numerous enough to matter. Void formation becomes a defect when the remaining metal cannot carry the current or the thermal load.

This guide covers six causes of void formation in a solder joint, the measurement used to quantify it, and the changes that reduce it. The order of the causes runs from the most common to the least, and the fixes follow the same order.

<img src="https://www.gopcba.com/wp-content/uploads/2024/10/m2.jpg" alt="X-ray inspection image showing void formation inside a solder joint” />

What Void Formation Means in a Joint

A void forms when gas cannot escape before the alloy freezes around it. The gas comes from flux volatiles, from air carried into the joint by the deposit shape, or from moisture released by the board or the component. Once the alloy is solid, the gas is trapped and the void remains as a cavity.

A small void in a joint that is otherwise sound is a cosmetic finding. A void that spans a large fraction of the interface is a reliability problem, because the crack that starts at the void has a shorter path across the joint. The limit is therefore usually written as a maximum area fraction, and sometimes as a maximum for the largest single void.

Cause 1: Flux Volatiles Trapped at Melting

The flux in the paste contains solvent, water and activators, and all of them vaporise as the joint heats. If the alloy melts before the vapour has left, the vapour is sealed in. The faster the ramp and the shorter the soak, the more likely this becomes, because the escape path closes before the gas has cleared.

A profile that holds the joint just below the solidus long enough for the flux volatiles to leave reduces voiding more than any other single change. The hold should be long enough to see the volatiles stop evolving, which on a production board is usually between 60 and 120 seconds depending on the paste.

Cause 2: Deposit Volume and Shape

A thick deposit has a longer path for the vapour to travel and a smaller surface area through which to escape, so it traps more gas than a thin one of the same volume. This is why a ground pad joint printed with a thick stencil tends to void more than a small signal joint printed with the same paste.

The shape matters too. A deposit with a dome holds the gas under the centre, while a deposit with a vented edge or a split pattern lets the vapour out sideways. Where the geometry allows, splitting a large deposit into a grid of smaller ones reduces void formation substantially without changing the total volume.

Cause 3: Reflow Profile Ramp and Soak

The reflow profile controls both the amount of vapour produced and the time available for it to escape. A slow ramp produces vapour gradually and gives it time to leave, while a fast ramp produces it in a burst. The soak then decides how much of the remaining vapour is driven off before the alloy flows.

Where voiding is a problem, widening the soak is usually more effective than reducing the peak, because the peak has little effect on the gas that is already trapped. The profile should be re-measured with a thermocouple on the joint being examined, since the voiding is a property of that joint rather than of the board average. Our profile guide covers the measurement.

Cause 4: Surface Finish and Oxide

An oxidised surface does not wet cleanly, and a surface that does not wet leaves a gap that fills with gas. The oxide may be on the pad, on the component termination or on the alloy powder itself, and each has a different remedy. A pad that has been stored too long, or a component whose finish has aged, will void more than a fresh one.

The flux has to remove the oxide before the alloy melts, which it can only do if it is active at that temperature. Where the finish is difficult, the flux class rather than the profile is the lever, and the change should be verified by comparing void area on samples from the same board. Our surface finish notes describe which finishes are most sensitive.

Cause 5: Barrel and Via Outgassing

A through-hole joint or a via under a pad contains air and, after plating, traces of chemistry. As the joint heats, that gas expands and pushes into the molten alloy, where it forms a void or blows a hole through the fillet. The effect is largest on a via that has not been filled or plugged.

Where a via sits inside a pad, plugging the barrel before plating removes most of the gas volume and the voiding falls with it. Where the via cannot be plugged, a slower ramp allows the gas to escape through the barrel before the alloy seals the top, which is why the same joint can void at one ramp rate and not at another. Our notes on thermal via design cover the layout options.

Cause 6: Absorbed Moisture in Parts

A plastic package that has absorbed moisture releases it when heated, and the release happens inside the package rather than in the joint. The escaping vapour can pass through the molten alloy and leave a void, and in the worst case it cracks the package. The amount of moisture depends on the storage and the exposure time rather than on the reflow process.

Baking removes the absorbed moisture and is the fix, but it has to be applied before the reflow rather than after a batch of voids appears. Our moisture sensitivity level notes give the schedules, and the floor life record is what shows whether a part needs one.

Measuring Voids and Setting Limits

X-ray inspection is the standard method. A two-dimensional image shows the void area as a lighter region within the joint, and the software reports the total void fraction and the largest single void. A three-dimensional scan gives the void distribution through the thickness, which matters where the void sits at the interface.

The limit should be written for the joint class rather than for the whole product. A thermal pad carrying heat may need a void fraction below 10 %, while a signal joint may be acceptable at 25 %. X-ray inspection is also used to confirm a fix, and the comparison should be made between boards from the same paste, stencil and profile, with only one variable changed at a time.

FAQ

Are voids always a defect? No. Every reflowed joint contains some gas, and a limit that demands none cannot be met by any process. What matters is the fraction of the interface that remains bonded and the location of the largest void.

Does a higher peak reduce voiding? Rarely. A higher peak produces more vapour and gives it less time to escape, so it usually makes voiding worse. Extending the soak below the solidus is the more effective change.

Can voiding be measured without X-ray? A cross-section shows the voids in one plane and misses the rest, so it is a sample method rather than an inspection method. Where voiding is a routine control, X-ray is the practical choice.