Permanganate Desmear Control: 6 Steps That Protect the Hole Wall

Permanganate desmear is the wet chemical sequence that removes resin smear from a drilled hole so that copper can be plated onto a clean, bondable surface. Smear forms when the drill heats the resin beyond its softening point, and the softened material is dragged across the hole wall and over the inner layer connections.

The sequence is not a single bath. It is a swelling step, an oxidising step, a neutralising step and a rinse between each of them, and the quality of the result depends on all four being controlled together rather than on the permanganate alone.

What Desmear Has to Remove

Smear is resin, and it is electrically insulating. Where it covers an inner layer connection, the plating that follows cannot make contact with the copper beneath, so the connection is incomplete even though the barrel looks continuous. The defect appears later as an intermittent open or as a failure after thermal cycling. Because the barrel is physically complete, an electrical test on the bare board will often pass the connection, and the weakness is only exposed by thermal load or by a section.

Desmear also etches the resin surface slightly, which improves the mechanical key between the resin and the plated copper. That etching is deliberate, and the depth of it is one of the variables the process has to control. Too little etch leaves the resin smooth and the plating with nothing to grip, while too much removes resin between the glass bundles and weakens the wall.

The Four Stage Sequence

The swelling stage softens the resin so the oxidiser can penetrate it. The permanganate stage oxidises the softened resin and converts it into soluble products. The neutraliser reduces residual manganese to a soluble form and removes it, and rinses between each stage prevent chemistry from being carried forward.

Skipping a stage does not produce an obvious failure at the time. A weak neutraliser leaves manganese dioxide on the resin, which appears later as a dark residue under the plating and as poor hole wall adhesion, and the board looks normal at every inspection before that. The sequence should therefore be recorded stage by stage, with analysis results for each tank, rather than as a single process step.

Swelling the Resin Before Oxidation

The swelling chemistry is solvent based and penetrates the resin, opening it so that the permanganate can reach below the surface. Its temperature and dwell time set how deeply the resin is affected, and both are specific to the resin system in use.

Over-swelling causes its own defects. A resin that is softened too deeply is removed further than intended, leaving a ragged or over-etched wall, and the same condition can lift the resin away from the copper at an inner layer. Where the board carries a high glass transition resin, the swelling stage usually needs more time rather than more chemistry. Raising solvent strength to compensate shortens tank life and increases the risk of lifting resin away from the inner layer copper.

Permanganate Concentration and Temperature

Permanganate is consumed as it oxidises resin, so its concentration falls with throughput. Manganese dioxide also forms as a by-product and has to be removed or kept in suspension, because deposits on the panel become defects later in the process.

Temperature and dwell control the reaction rate, and the two are usually adjusted together to hold a target etch weight per panel. The etch weight is the better control than concentration alone, because it measures the effect rather than the input, and it can be checked by weighing a sample coupon. That measurement is quick, it is comparable between shifts, and it captures the combined effect of concentration, temperature and dwell in one figure.

Neutralising and Manganese Removal

The neutraliser reduces manganese to a soluble form that the rinse can remove. Its concentration and its dwell have to keep pace with the oxidiser, because a strong permanganate stage followed by an exhausted neutraliser leaves residue that will not rinse away.

Residue that survives neutralisation shows up as a brown or black tint in the hole and as a barrier between the resin and the plated copper. Because the residue is under the plating, it is invisible from the surface and is only found in a microsection or through a plating adhesion failure. Keeping the neutraliser analysis and the rinse conductivity in the same record makes it possible to trace such a failure back to the shift that produced it.

Rinse Quality Between Stages

Rinse is what keeps the stages separate. Permanganate carried into the neutraliser consumes it, and neutraliser carried into the plating line contaminates the electroless copper. Counter-flow rinses with a conductivity set point give a repeatable barrier between tanks.

Rinse quality also affects the hole itself. Because a hole is a small volume with a large surface, it holds chemistry longer than a flat panel, and a rinse that is adequate on the surface can leave residual chemistry inside the barrel. That is why rinse checks include a hole sample rather than only a surface check.

Desmear for Different Resin Systems

Standard FR-4 responds well to permanganate, but filled resins, high glass transition materials and halogen-free systems behave differently. Filler particles may not oxidise at all, and some resins resist swelling, so the sequence has to be adjusted for the material rather than applied unchanged. Filled systems in particular often need a longer swell and a hotter oxidiser to reach the same etch weight as a standard FR-4.

Where a wet sequence cannot achieve the required result, an alternative is needed. The plasma treatment route removes resin by reaction in a vacuum chamber and works well on materials that resist wet chemistry, at the cost of batch processing and different capital equipment.

Effects on Plating and Reliability

Desmear sets the surface that the electroless copper and the subsequent electrolytic plating will bond to. A well desmeared wall gives a continuous copper layer with good adhesion, while a poorly desmeared one leaves a thin or discontinuous layer that can pass a visual check and fail later.

Reliability effects appear in thermal cycling, in resistance monitoring and in the field. Adhesion loss at the inner layer and residual manganese under the plating are both incubation defects, so the process record and the coupon data are the evidence that the sequence was in control when the panels were made.

FAQ

How do I know the desmear is working? Etch weight on a coupon, a section of the hole wall and a plating adhesion test together show whether the sequence removed smear and prepared the resin. Any one of them alone can look acceptable while the others reveal a problem. Running all three on the same coupon keeps the picture consistent between lots.

Can a second pass fix an under-desmeared board? Re-running the sequence is technically possible but risks over-etching the resin and loosening the bond at the inner layers. It is usually better to correct the bath and the dwell before the lot is plated.

Does desmear affect the finished hole size? It removes a very small amount of resin, which marginally increases the hole, and the drift accumulates with each rework. The effect is small compared with drilling tolerance but worth noting when the hole size is tight.