Solder Joint Microsection Analysis

A microsection is the only way to see inside a solder joint. Visual inspection shows the fillet, X-ray shows the voiding, but neither reveals the intermetallic layer, the grain structure or the crack that has started at the interface between the solder and the pad. When a joint fails in the field, in thermal cycling or in a mechanical test, microsection analysis turns an opinion into a measurement, and the preparation technique matters as much as the interpretation of the image.

<img src="https://www.gopcba.com/wp-content/uploads/2020/11/project_image_11.jpg" alt="Cross section of a solder joint under magnification” />

What Microsection Reveals

Microsection reveals the features that determine joint reliability: the thickness and continuity of the intermetallic compound between solder and pad, the presence of voids or cracks, the wetting angle at the pad edge, and the grain structure of the bulk solder. It also shows what a fillet looks like in cross section, which is rarely what the appearance from above suggests. The technique is destructive, so it is used on samples, on failed units and on coupons, and the sample must be chosen to represent the population being judged rather than the unit that happens to be most convenient to cut.

Interpretation requires a reference. A joint that looks acceptable under a microscope may still be outside a specification that defines maximum intermetallic thickness or maximum void area, and a joint that looks ugly may be perfectly sound. The acceptance criteria should be written before the first section is cut, ideally with reference photographs of conforming and non-conforming examples, so that two engineers looking at the same image reach the same conclusion about the joint in front of them. Ambiguity in the criteria usually shows up as arguments in the failure analysis meeting rather than in the laboratory.

Cutting, Mounting and Polishing

Preparation introduces its own defects. Cutting with an abrasive wheel can smear the soft solder, push it across the interface or open a crack that was not there before. Mounting resin can infiltrate a genuine void, and aggressive polishing can round the edge of the joint and remove exactly the interface being examined. The standard defence is a slow, cooled cut, a suitable mount, and a graded polishing sequence that ends with a light etch to reveal the structure rather than scratch it.

The plane of the cut matters too. A section through the middle of a joint shows the full interface, while a section that grazes the edge shows a crescent of solder and invites a wrong conclusion about wetting. Where the feature of interest is small, such as a crack at a corner, more than one plane may be needed. Recording the section plane and the location on the board makes the result comparable between samples, between operators and between laboratories, which matters when a supplier and a customer are examining the same joint.

Reading the Intermetallic Layer

The intermetallic compound forms wherever molten solder meets a metal pad, and its thickness grows with temperature and with the time spent above liquidus. A thin, continuous layer is the sign of a good metallurgical bond; a layer that is very thick or scalloped is brittle and is the site where thermal fatigue cracks begin. Because growth is a function of thermal exposure, a thick layer can also be evidence that a joint has seen more rework than the production records suggest.

Measuring it requires magnification and a scale bar, and the measurement should be taken at several points along the interface rather than at the thinnest or the thickest place. The composition of the layer differs between finishes and alloys, which is why the comparison should be made against data for the specific combination in use. Plating thickness and finish choice are direct inputs here, since a thicker or more reactive finish changes both the growth rate and the morphology of the interface.

Voids and Their Measurement

Voids form from flux volatiles, entrapped air and shrinkage, and their significance depends on where they sit. A void in the bulk of a thermal pad can raise thermal resistance, a void at the interface can reduce the load-bearing area, and a small void in the corner of a fillet is usually harmless. Reporting void area as a percentage without describing the location tells a design team very little, so acceptance criteria should distinguish between these cases explicitly.

Measurement method matters because X-ray and microsection see different things. X-ray gives the void distribution across the whole joint and is non-destructive; microsection gives the true shape and position at one plane. Where a specification is written in terms of void percentage, the method used to measure it should be named, otherwise the same joint can pass one test and fail another without anything about the joint changing at all.

Grain Structure and Thermal History

The grain structure of the solder records its thermal history. A fast cooling rate produces fine grains and a stronger joint, while slow cooling produces coarse grains, larger intermetallic particles and a joint that is more susceptible to fatigue. For lead free alloys the effect is pronounced, and a thermally slow assembly can produce joints that pass every electrical test and then crack in service. Microsection is often the only place this shows up before the field does.

Comparing grain structure across a board is also a practical way to audit the reflow profile. Joints near a heavy component or in a shadowed area cool more slowly than those in open areas, and the difference is visible in section. Where a sample shows unexpected coarseness, the profile and the thermal mass of the assembly are the first two things to check, followed by the thermal cycling history of any unit that has already been through environmental test.

Reporting the Result

A microsection report should state what was examined, where it came from, how it was prepared and what was measured. Photographs at a stated magnification with a scale bar are essential, and annotations that point to the specific features of interest make the report usable by someone who was not present when the section was cut. A conclusion that says only conforming or not conforming, without the measurements behind it, is not really a report, and it will not survive a customer review.

The report should be filed against the product and the lot, because its value comes from comparison over time. A sequence of sections from the same product shows whether the process is drifting long before failures appear, and it becomes the baseline when a design, a paste or a solder alloy changes. Microsection practice for bare boards follows the same principles, and the two records together describe the whole interconnection from laminate to component.

FAQ

How many joints should be sectioned? Enough to represent the population and the known risk areas. A typical investigation covers several joints of each type, including the ones nearest the highest thermal mass on the board.

Does a void mean the joint is bad? Not by itself. Position matters more than size in most cases, and voids in a thermal path are far more serious than small voids in a fillet.

Can microsection be replaced by X-ray? No. X-ray is better for volume screening, while microsection is the only way to measure intermetallic thickness, grain structure and cracks at the interface.