Solder Joint Reliability in PCB Assemblies

A solder joint is a mechanical and electrical connection that is expected to survive the life of the product while being cycled in temperature, stressed by vibration and, in some cases, exposed to contamination. Solder joint reliability is the study of how those joints fail, and most of the failures come from a small number of mechanisms that are well understood. Knowing which mechanism applies to a particular design is what makes the difference between a joint that lasts and one that cracks in the field.

How Joints Fail

The dominant failure mechanism in thermally cycled assemblies is fatigue. The board and the component expand by different amounts when heated, and the solder joint has to accommodate the difference. Each cycle strains the joint, and repeated straining accumulates damage until a crack forms and propagates. The phenomenon is not a defect in the joint; it is the inevitable consequence of thermal expansion mismatch.

The second mechanism is intermetallic growth. When molten solder wets a copper or nickel surface, a layer of intermetallic compound forms at the interface. The layer is necessary for a good bond and it is brittle, and it continues to grow slowly at operating temperature. A layer that becomes too thick weakens the joint and makes it more susceptible to fracture under shock or vibration.

The third mechanism is mechanical overload, from shock, vibration or board flex during handling and assembly. A joint that is adequate in a static sense can still be damaged by a single event, and the damage may not be visible until the product has been in service for some time.

Thermal Cycling and Fatigue

Fatigue life depends on the strain per cycle, and the strain depends on the difference in expansion between the component and the board, on the distance from the neutral point and on the size of the joint. A large component with a high expansion mismatch and joints far from the neutral point experiences the greatest strain, which is why the corner joints of a large package are the first to crack.

The cyclic conditions matter as much as the extremes. A product that cycles between twenty and sixty degrees every day experiences a very different life from one that cycles between minus forty and a hundred and twenty five a few times in its life. The number of cycles, the dwell time at each extreme and the rate of change all affect the accumulated damage.

The design responses are to reduce the strain, to reduce the number of cycles or to use a joint that tolerates more strain. Reducing the strain means choosing components with a closer expansion match, using a compliant package, or adding a compliant layer beneath the component. Adding underfill is the classic answer for large area packages, because it mechanically couples the component to the board and distributes the strain.

Intermetallic Layers

The intermetallic layer forms during soldering and grows during subsequent thermal exposure. Its growth rate increases rapidly with temperature, which means that a board that experiences a hot rework cycle or a high temperature bake may have a significantly thicker layer than one that did not, even though both look the same.

The practical consequences are twofold. A thick layer is brittle and provides a path for crack propagation along the interface, and it also consumes the copper or the plating that it grows into. Where the plating is thin, the layer can consume a significant fraction of it and leave a weak structure behind.

Control therefore focuses on limiting the thermal exposure. Choosing a finish that forms a thinner layer, avoiding unnecessary rework, limiting the number of reflow cycles and controlling the soldering temperature and time all reduce the growth. The effects are cumulative, so a board that has been reworked twice is in a different condition from one that was never reworked.

Materials and Finishes

The finish on the board determines what the solder has to wet and what intermetallic forms. A hot air levelled finish produces a different interface from an immersion gold or an organic preservative, and each has its own behaviour under thermal cycling. The choice should be made for the application rather than by default.

The solder alloy itself matters. Lead free alloys generally have higher melting points, different creep behaviour and a greater sensitivity to strain rate than the tin lead alloys they replaced. Their fatigue behaviour is different rather than simply worse, and joints designed for one alloy may not perform as expected with another.

The component finish and the board finish interact as well. Two finishes that are both adequate individually can produce a joint with less margin when combined, and the combination should be considered when a component is substituted rather than only the component’s electrical parameters.

Joint Inspection

Joint inspection methods each see a different aspect of the joint. Visual inspection finds surface defects such as insufficient fillet, bridging and obvious cracking. X-ray inspection sees the internal structure and is the standard method for area array packages where the joints are hidden. Microsectioning reveals the interface, the intermetallic layer and any internal cracking, at the cost of destroying the sample.

Automated optical inspection is fast and repeatable but limited to what can be seen from the outside, and it needs a reference to compare against. Its value for reliability work is in detecting the process drift that precedes a joint defect, rather than in assessing the reliability of an individual joint.

The choice of method should follow the failure mode of interest. Where the concern is thermal fatigue, a microsection of a cycled sample is the most informative, because it shows crack initiation and the condition of the interface. Where the concern is a process escape, X-ray and optical inspection are more appropriate.

Qualification and Testing

Reliability testing subjects the assembly to a defined cycle and then examines the joints. Thermal cycling between specified extremes with a specified dwell and rate is the standard test for fatigue, and vibration and shock tests cover the mechanical mechanisms. The test conditions should be chosen to represent the product’s actual environment rather than copied from a general standard.

The examination after the test is as important as the test itself. A sample that has been cycled and then microsectioned shows the damage that the test produced, and the comparison against an uncycled control shows how much of it was there before. Without the control, the interpretation is a matter of opinion.

The results should be recorded with the construction details: the alloy, the finishes, the profile, the number of reflow cycles and any rework. These are the variables that determine the outcome, and a result without them cannot be compared with a later build or with a different supplier.

Practical Rules

Identify the dominant mechanism for the design: thermal fatigue, intermetallic growth or mechanical overload. Reduce the strain where fatigue dominates, limit the thermal exposure where the interface is the concern, and use underfill where the package demands it.

Choose the inspection method that matches the failure mode, and qualify the assembly with a test that represents the real environment. Record the construction details with the build records, because a reliability result is only useful when the quality records show exactly how the sample was made.

FAQ

What causes most solder joint failures? Thermal fatigue from expansion mismatch between the component and the board, accumulated over many temperature cycles.

Why does intermetallic growth matter? The layer is necessary for bonding but brittle. If it grows too thick it weakens the joint and provides a path for cracks.

Which inspection finds internal joint defects? X-ray for hidden joints such as area array packages, and microsectioning where the interface and internal cracking must be examined.