Desmear Process Control in Multilayer Board Fabrication

Desmear is the step that removes the resin left on the copper by drilling, and it is one of the few processes in a PCB shop whose failure is invisible until the board is electrically tested. A hole wall that has not been cleaned will plate, will pass continuity, and will then fail after a few hundred thermal cycles because the copper never bonded to the inner layer. The process is short, chemical and easy to under-control, which is why it deserves more attention than its position in the flow suggests.

What Desmear Has to Remove

Drilling a multilayer board cuts through cured resin, glass fabric and copper foil at high speed, and the friction melts the resin locally. The molten material is pushed against the hole wall by the drill flute, where it cools as a thin, glossy layer that covers the copper edges of the inner layers and the glass fibres.

That layer, known as resin smear, is not a contaminant in the ordinary sense and cannot be washed away. It has to be chemically removed or etched back so that the subsequent plating can make direct contact with the copper of every inner layer, and the quality of that contact decides the reliability of the finished via.

How Resin Smear Forms

Smear is produced by heat rather than by mechanical tearing, so its severity follows the drilling parameters. A worn drill bit, a higher chip load, a lower spindle speed or a stack that is too thick for the bit all raise the temperature at the cutting edge and increase the amount of resin that melts.

The laminate itself contributes as well. A resin with a higher glass transition temperature and a more highly cross linked structure smears less than a soft, high flow material, which is one of the reasons a change of laminate often forces a review of the drilling and desmear conditions together.

Permanganate Chemistry and the Sequence

The standard desmear line uses a swelling step, a permanganate etch and a neutralisation, usually followed by a conditioner that prepares the resin for the plating that comes later. The swell opens the resin structure so the permanganate can attack it, and the neutraliser removes the manganese residues that would otherwise poison the plating bath.

Each stage has its own temperature, concentration and dwell time, and each one can be controlled independently. The permanganate step does the visible work, but a weak swell or an exhausted neutraliser will produce the same result as a weak etch, which is a hole wall that is partly covered and hard to diagnose.

Etchback, Desmear and Their Difference

Desmear removes the smear and leaves the resin roughly level with the copper. Etchback goes further and deliberately recesses the resin so that the copper of the inner layer protrudes slightly into the hole, which increases the contact area between the plated barrel and the inner layer.

Etchback is specified in some high reliability designs and in boards with a high aspect ratio, where the plating has more difficulty reaching the centre of the hole. It is controlled by the same chemistry as desmear, simply with a longer etch or a more aggressive condition, and the amount is measured under a microscope rather than assumed.

Process Variables That Matter

The variables that decide the outcome are the concentration and temperature of the permanganate, the dwell time in each tank, the condition of the swell bath and the efficiency of the rinses between stages. Drag-out from one tank into the next is a common cause of drift, because it changes both the chemistry and the level of contamination.

Loading is the variable that is most often ignored. A heavy rack of boards consumes permanganate faster than a light one, so a bath that is analysed once a shift may be outside its window for part of that shift. Analysing against the load, or dosing automatically, keeps the etch consistent from the first panel to the last.

Control of the Bath and the Line

Control starts with a titration schedule, a temperature log and a record of the panels processed since the last change of chemistry. The permanganate concentration falls with use and the manganese dioxide by-product builds up, so both the active strength and the contamination level have to be watched.

Mechanical condition matters too. Rollers that no longer squeeze evenly leave puddles of chemistry on a panel, spray nozzles that are partly blocked starve one area, and a slow transfer between tanks allows the previous solution to dry on the surface. These faults show up as a pattern on the panel, which is why the location of a defect is evidence about its cause.

Evaluating Hole Wall Quality

Hole wall quality is assessed on a microsection, where the plated copper, the resin and the inner layer copper are all visible at once. The feature to look for is continuous copper on the inner layer surface with no resin between it and the plating, and any gap is measured as a percentage of the inner layer thickness.

A faster check uses a backlight test, in which a sample is plated with a thin, bright coating and examined with light transmitted through the laminate. Areas where the resin remains show as dark patches against a bright background, and the test is reproducible enough to be used for daily monitoring rather than only for failure analysis.

Common Defects and Their Causes

Hole wall pull away, voids at the inner layer interface and intermittent opens after thermal cycling are the classic symptoms of inadequate desmear. They rarely appear on a cross section taken from the middle of a panel, because the etch is usually weakest at the edges and in the corners of a large board.

Other defects point the other way. Excessive etchback exposes glass fibres that are difficult to plate, and over-treatment of a thin panel can remove resin from between the inner layers and reduce the insulation resistance between them. The defects that follow are described among the plating problems and board level failures.

Integration with Plating and Reliability

Desmear is the first half of the interconnection, and the plating that follows is the second. A clean hole wall with a poor plating distribution will still fail, and a strong plating on a smeared wall will fail even faster, so the two processes are best controlled as a pair and reviewed together when a reliability test fails.

The copper thickness that reaches the middle of a hole depends on the aspect ratio and on the throwing power of the plating bath, both of which are covered in the guides to hole copper and aspect ratio. The practical rule is that a deep hole needs a cleaner wall and a more uniform plating than a shallow one, and neither process can compensate for the other.

FAQ

Is desmear always necessary? It is necessary for any board whose drilling generates enough heat to smear resin, which in practice means most multilayer work. Thin, two layer boards drilled gently may need only a light treatment, but the decision should follow a backlight or microsection result rather than an assumption.

Can desmear be replaced by a stronger drilling process? Better drilling reduces the amount of smear and is always worth doing, but it does not remove what has already been formed. The chemical step is what cleans the wall, and the drilling parameters are what decide how much cleaning is required.

How is desmear verified on production panels? By a combination of bath analysis, a backlight coupon processed with each load and periodic microsections. The coupon is the only item that shows the effect on the hole wall, and it should be retained with the production record for the batch it accompanied.