Microsection Quality: Cutting, Mounting and Polishing the Truth

A microsection is the reference evidence for plated thickness, barrel integrity, intermetallic growth and laminate damage. It is also a destructive test whose result depends entirely on how the specimen was prepared, and a badly prepared section can invent defects or hide them.

The preparation sequence therefore has to be controlled to the same standard as the process being measured. Cutting, mounting, grinding, polishing and etching each contribute artefacts, and an operator who cannot tell an artefact from a defect will report the wrong thing. Specimen preparation is therefore part of the measurement rather than a preliminary to it, and the operator who treats it as routine work is the one who produces a section that cannot be defended.

What a Microsection Is Used to Prove

The common questions are how thick the plating is, whether the barrel is continuous, whether there is a crack or a separation, and how much intermetallic has formed between the copper and the solder. Each question implies a different cut orientation and a different level of magnification. A cut chosen for thickness rarely answers a crack question, and one specimen cannot serve both purposes without a compromise in the result.

A section is a point measurement, so the location matters as much as the preparation. A hole at the corner of a panel and a hole in the centre of a dense array see different plating conditions, and a report that does not state the location cannot be compared with anything.

Choosing the Cut and Its Orientation

A plated hole is sectioned through its axis, so the cut plane passes through the centre of the drill and the copper appears on both walls. A cut offset from the centre produces an ellipse that looks like a thinner plating than the wall actually has, which is one of the most common reporting errors.

For a crack, the plane has to be perpendicular to the expected crack direction, which for a barrel is the axis and for a pad crater is the pad surface. Where the orientation is uncertain, two sections at right angles are cheaper than an argument. The crack case is described in the notes on barrel crack sections.

Cutting Damage and How to Avoid It

Abrasive cutting generates heat and deformation, and both can smear the copper across a void or tear a fragile interface. The cut has to be made slowly, with coolant, and with the smallest wheel that will do the job.

The cut is also made at a distance from the feature and then reduced by grinding, so that the damage caused by the wheel is removed before the plane of interest is reached. Leaving two or three tenths of a millimetre of material for grinding is standard practice and it is not optional on soft or thin specimens.

Mounting and Edge Retention

The specimen is cast in resin before grinding, and the resin has to support the edges of every layer. A soft resin rounds the copper at the surface, which makes the plating appear thinner at the edge than in the interior, and a brittle resin cracks and pulls material away.

Filled mounting compounds are used for this reason, and vacuum impregnation is used where the specimen contains a void that must be preserved rather than filled with resin. The choice affects what the section can prove, so it belongs in the procedure.

Grinding and Polishing Sequence

The sequence runs from coarse to fine, typically through several grits and then diamond suspensions down to one micrometre. Each stage removes the scratches of the previous one, and skipping a step leaves scratches that look like cracks under the microscope.

Time and load matter as much as the abrasive. Excessive pressure drives abrasive into the copper and produces relief at the resin boundaries, while too little leaves the previous stage unfinished. The section is inspected between stages rather than only at the end, which is what prevents the final polishing from being repeated indefinitely.

Etching to Reveal the Structure

Polishing alone leaves the layers visible by colour but not by structure, so an etch is used to reveal grain boundaries, intermetallic layers and the boundaries between copper and resin. Copper is commonly etched with an ammonium hydroxide and hydrogen peroxide mixture, applied for a few seconds and then rinsed.

Etching is where a section is most easily ruined. Too long removes the thin intermetallic layer that was the point of the exercise, and too short leaves it invisible and reported as absent. The etch time should be standardised, and a control specimen used when a result is unexpected.

Measurement at the Microscope

Plated thickness is measured at several points around the barrel, with the minimum reported as well as the average, because a specification is usually written as a minimum. Magnification should be high enough that the measurement is limited by the specimen rather than by the image.

The scale has to be calibrated with a stage micrometer and the magnification recorded, since a measurement is meaningless without it. A typical report is at two hundred to five hundred times, with the calibration date and the instrument identification attached.

Subjectivity, Calibration and Repeatability

Sectioning is often described as objective and it is not. Two technicians can measure the same barrel differently, and the difference can exceed the tolerance being checked, particularly at the minimum point where the copper is thinnest. A second operator who prepares a fresh specimen and reaches the same number is stronger evidence than a second measurement taken on the same mount.

The controls are a written procedure, a boundary sample set and periodic comparison between operators. Where a section decides an acceptance question, a second measurement by another technician is cheap insurance, and the disagreement rate is a useful measure of how well the procedure is being followed. The boundary samples should be replaced periodically, because a reference set that has been re-polished many times no longer matches the current procedure.

Records and Second Opinions

The record should identify the specimen location, the cut orientation, the mounting compound, the polishing sequence, the etch, the magnification, the measurements and the operator. With those fields a section can be reproduced rather than repeated from scratch. A photograph of the mounted specimen before grinding is worth keeping, since it shows the cut plane in relation to the feature being judged.

Images should be kept at the resolution at which the measurement was made, not only as low resolution thumbnails. Where a customer disputes a result, the original image is the evidence, and the section itself can be re-polished once but not twice. The mounting compound lot matters as well, because a harder batch polishes differently and can change the measured thickness at the edge of a barrel. The broader toolkit around it is described in the notes on failure analysis methods, and the surface condition being judged in the notes on hole wall roughness.

FAQ

Why does the same hole measure differently in two sections? Usually because one cut was offset from the axis, so the copper appears thinner than it is. The cut plane has to pass through the centre of the hole.

Does polishing create cracks? It creates scratches and relief that can look like cracks. Cutting slowly with coolant, leaving material for grinding and inspecting between stages removes most of them.

How many points should be measured on a barrel? Several around the circumference, with the minimum and the average both reported. The minimum is what the specification usually controls.