Copper Wicking: Where the Plating Travels and Why

Copper wicking is the extension of plated copper along the laminate surface or into the bundle of glass fibres beside a drilled hole, beyond the boundary that the artwork defines. The copper is thin, it follows the fibres or the interface between the resin and the glass, and it reduces the distance between two conductors. In a fine pitch design it can be the difference between a passing insulation test and a short.

Wicking is a process defect rather than a design one, and it is produced by the condition of the laminate surface at the time of plating. A rough wall, an incompletely removed smear, a loose fibre or a surface that has been attacked by the etchant all provide a path along which the plating solution penetrates and the copper deposits.

What Copper Wicking Is

The visible effect is a thin line of copper beside a hole or a conductor, often following the direction of the glass fabric. In section it appears as a wedge or a thread of copper that extends from the plating into the laminate, sometimes only a few micrometres thick and sometimes deep enough to approach an adjacent feature.

Its significance depends on the geometry. A wick that extends into a large clearance area changes nothing electrically. A wick that extends towards an adjacent pad reduces the creepage distance, and one that bridges the gap produces a short that the electrical test will find. The defect is therefore judged by the distance that remains rather than by its own length.

How It Forms During Plating

Plating solution penetrates along any path that the laminate offers, and the copper then deposits in that path as the current flows. The path is created before plating: a drilled hole with a rough wall has open fibres, a hole that has been desmeared aggressively has an undercut between the resin and the glass, and a laminate with a weak interface has a gap that opens when the board is heated or cleaned.

The resin and the glass bond is the weak point. Where that bond is poor, the drilling and the wet processes open it, and the opening becomes a capillary that draws in solution. Once the solution is inside, no rinse reaches it completely, and the residue that remains can also produce a conductive path even without plating.

Adhesion and the Wet Process

The wet process decides how much of the laminate interface is exposed. Mechanical scrubbing before lamination and before plating removes loose material and improves adhesion, and an aggressive scrub or an over-etched surface leaves fibres standing proud of the resin, which the plating then covers individually. The copper that grows along a single fibre is the classic wicking signature.

Chemical treatments have the same effect in the other direction. A conditioning step that is too strong penetrates the resin and opens the interface, while one that is too weak leaves the surface unreceptive. The balance is set by the process window and verified by the same kinds of sample that are used for adhesion, as described in the notes on scrubbing before lamination.

Etching and the Wick Path

Etching removes copper, and it also attacks the exposed interface beside a conductor. Where the laminate is porous the etchant travels along the same path that the plating solution used, and it undercuts the copper. The subsequent resist stripping then exposes a fine line of copper that was never intended to exist.

The etch side of the control is therefore the same as the etch side of line width control: a stable bath and a controlled dwell keep the lateral attack predictable. Where the etch factor is drifting, the wicking becomes worse at the same time, and both effects are described in the notes on fine line etching.

Wicking and Solder Mask

Solder mask is the last barrier between a wick and the environment. Where the mask adheres well, a short wick is sealed beneath it and its effect is small; where the mask lifts, the wick is exposed and the surface insulation resistance falls. Wicking and mask adhesion are therefore treated together, because the same surface condition drives both.

The preparation of the surface before masking matters more than the mask material. A board that has been through an aggressive mechanical treatment has a different surface from one that has been chemically cleaned, and the difference appears in the adhesion test rather than in a visual inspection. The preparation and test sequence is described in the notes on solder mask adhesion preparation.

Consequences: Clearance and Shorts

The electrical consequence is a reduction in the distance between conductors. On a low voltage signal board a small reduction is tolerable; on a board with a high working voltage, or on one that operates in humid conditions, the reduction lowers the breakdown voltage and increases the leakage. The design rule that defines the minimum spacing should be checked against the remaining distance after wicking rather than against the artwork.

At the limit the wick becomes a short, and the electrical test may detect it as an insulation failure at the test voltage rather than as a continuity failure. That behaviour is the reason a wicking problem can pass the bare board test and fail a functional test later, particularly after the assembly process has added heat and flux.

Detection

Detection combines a visual inspection at magnification with a section. A wick along a fibre is visible as a thin metallic line beside a hole or a conductor, and it is easier to see under a light at a low angle. Electrical test detects the ones that have bridged, and it says nothing about the ones that have only reduced a clearance.

For a process investigation, a section through several holes is the most informative sample because it shows the depth and the direction of the wick. Trend monitoring is then done by visual inspection on a coupon, with the acceptance criterion expressed as a remaining distance rather than as the length of the defect.

Process Controls

The controls are the drill wall condition, the desmear and conditioning step, the mechanical scrubbing, the rinse quality and the etch. Each has its own window and each contributes to the interface condition. A change in any of them should trigger a check of the section, because the effect of a single step cannot be predicted from its own parameters alone.

Laminate selection is part of the control as well. A material with a well bonded glass and resin structure resists penetration better than one with a weak interface, and the difference is visible in the same section. Where a product is moved from one laminate to another, the wicking behaviour should be verified rather than assumed to be equivalent.

Records and Change Control

The record should carry the laminate lot and its storage, the drill parameters, the desmear and scrubbing conditions, the plating bath and the etch result, together with the section images. With those fields a wicking investigation has a starting point, and the common causes, a dull drill, an over-scrubbed surface or a drifting etch, are each identifiable from the data.

Change control applies to consumables as much as to equipment. A new drill supplier, a new brush material or a new conditioner chemistry can change the interface condition without any machine parameter changing, and the verification that follows should be a section rather than a visual check. The same discipline protects the adhesion of the mask and the performance of the coating that follow, which is the subject of the notes on hole wall roughness and plating adhesion.

FAQ

Is copper wicking always a defect? Not always. A short wick inside a wide clearance is harmless, and a wick that reduces the distance between two conductors is not.

Does wicking happen before or during plating? The path is created by drilling and the wet process, and the copper is deposited during plating. Both are needed for the defect.

Can a wick be removed? It can be etched back if the artwork allows, and the surface condition that created it has to be corrected at the same time or the defect returns on the next lot.