Plating Adhesion and Resist Adhesion Problems
Two Interfaces
Most adhesion problems on a circuit board occur at one of two interfaces: between the copper and the laminate, and between a coating and the copper. The first is created during lamination, when the resin flows against a treated foil and then cures. The second is created when a resist or a solder mask is applied and cured against a copper surface. Both depend on surface chemistry and on the process conditions, and both fail in a way that is not visible until the board is stressed mechanically or thermally.
Copper to Laminate
The bond between copper and resin is mechanical and chemical at once. The foil is supplied with a treatment, usually a nodular deposit or a chemical conversion coating, that gives the resin something to key into and to bond with. A foil with a low profile has less mechanical keying, which is good for high-frequency loss but demands a better chemical treatment to keep the peel strength. Where the bond is weak, the failure appears as a delamination around a hole or a pad after thermal cycling, or as a blister during reflow when moisture turns to steam at the interface.
Oxide and Bonding Treatment
Before inner-layer lamination, the copper is treated to create a surface that the resin will bond to, usually a black or brown oxide. The oxide has to be thick enough to provide the mechanical key and thin enough not to be brittle, and it has to survive the lamination cycle without dissolving. Poor oxide, oxide that is too thick and powdery, and oxide that has been contaminated by handling all produce a weak interface that appears later as delamination. The treatment is also sensitive to the acid content of the subsequent plating and cleaning steps, which can attack it from the exposed edge of a hole.
Resist and Mask Adhesion
A resist or a mask adheres best to a clean, slightly roughened copper surface. Where the copper has been left with a chemical residue, with an oxide that formed during storage, or with a surface that has been polished too smooth, the coating lifts at the edge and the defect appears as a lifted line or as undercutting during plating. The cure conditions matter as much as the surface: an undercured mask stays soft and delaminates, and an overcured one becomes brittle and cracks. The treatment of the copper before masking, which is often a pumice or a chemical micro-etch, is one of the most important process steps for this interface and one of the most easily neglected.
Plating Adhesion and Thermal Stress
The plating in a via is bonded to the laminate by mechanical keying against the drilled wall and by the adhesion of the electroless seed layer to the resin and the glass. Where the wall is smooth and resin-rich from a poor desmear, the bond is weak and the barrel separates under thermal cycling, which shows as a crack between the barrel and the pad rather than through the plating itself. Where the desmear is too aggressive, it etches the resin away from the glass and creates the interface that moisture can follow, which is the same path that CAF uses. The process window for desmear is therefore narrow, and it is one of the parameters that distinguish a good plating line from a marginal one.
Test Methods
Adhesion is measured in a few standard ways. Peel strength is measured by pulling a strip of foil or a plated test pattern away from the laminate at a controlled angle. Thermal stress is measured by floating a sample in molten solder or by cycling it and then examining a cross section for delamination and barrel cracks. The mask adhesion is checked by a tape test or a cross-hatch test, and occasionally by a solvent resistance test. The intermetallic and the seed layer can be examined by cross section under magnification. Each of these tests destroys a sample, which is why they are run on coupons and on a sampling basis rather than on every board.
Process Control
The variables that most often drift are the micro-etch depth before lamination and before masking, the oxide weight, the lamination temperature and pressure, the cure of the mask, and the contamination of the plating baths. None of them produces an immediate visible defect, which makes them suitable for statistical process control rather than for final inspection, and the useful discipline is to record them and to watch for drift. Where a batch fails in the field, the records are what make it possible to identify the batch, and without them a systematic problem is indistinguishable from a random one.
Adhesion and the Assembly Process
Adhesion problems are often first seen at assembly rather than at fabrication, because the thermal excursion of reflow is what turns a marginal interface into a visible defect. A board with a weak copper-to-laminate bond that has absorbed moisture will blister during reflow, and a mask that was undercured will lift at the edge of a pad during the same excursion. This is why the bake before assembly matters, and why the moisture content of incoming boards is a process parameter rather than a storage detail. It is also why a change of fabricator can produce a defect that looks like an assembly problem: the assembly profile has not changed, but the laminate or the surface treatment that arrives with the boards has, and the first symptom appears in the oven rather than on the certificate.
FAQ
What causes delamination around a hole? A weak bond between the copper and the laminate, usually from poor oxide treatment or from moisture turning to steam during reflow.
Why does a mask lift at the edge? Because the copper was not clean or not adequately prepared before masking, or because the mask was undercured.
What is peel strength? The force needed to pull a strip of foil or plated pattern from the laminate at a controlled angle, measured as a force per unit width.
Does desmear affect barrel adhesion? Yes. Too little desmear leaves a weak, resin-rich wall, and too much etches the resin away from the glass and creates a moisture path.
Why does the board need to be baked? To remove absorbed moisture that would otherwise turn to steam at the interface and cause a blister during reflow.
Conclusion
Adhesion is created by surface preparation and by process conditions, and it fails later rather than immediately, which is what makes it a process control problem rather than an inspection problem. Control the micro-etch, the oxide, the lamination cycle and the mask cure, test on coupons and keep the records. Fabrication process details are described under PCB manufacturing, the capability limits under PCB capabilities, and the reliability consequences belong to quality management. Adhesion performance is normally confirmed during prototype PCB assembly in 2026.
