Solder Mask Adhesion and Copper Surface Preparation

Solder mask adhesion is decided long before the mask is applied. The copper surface that the mask has to grip is produced by the etching process, modified by the surface treatment, and then exposed to storage and handling before the mask ever reaches it. When the mask lifts after reflow or flakes off during a repair, the cause is usually somewhere in that earlier chain rather than in the mask itself.

Why Mask Adhesion Fails

The mask grips the copper and the laminate through a combination of mechanical keying and chemical bonding, and both are sensitive to the state of the surface. A surface carrying an oxide layer, a residue from a previous process or a film of handling contamination bonds poorly, even though it looks clean.

Failure shows up in several ways. The mask can lift at the edge of an opening, blister over a large copper area, flake during a thermal cycle, or peel under the pressure of a repair tool. Each of those points at a slightly different part of the preparation chain, which is why the first steps of a failure investigation are to look at the surface rather than the mask. Surface preparation is the variable that most often explains a difference between two builds that used the same mask material.

The Copper Surface the Mask Sees

The surface presented to the mask is the finished copper after etching, and its condition depends on the etchant, the rinse, the surface treatment and the time since the treatment. A freshly treated surface behaves differently from the same surface after a week in a humid store with fingerprints on it. The effective age of the surface is therefore as important as its condition at the moment of treatment.

The topography matters as well. A surface with a fine, even texture gives the mask more to key into than a polished one, and the texture has to be consistent across the panel. Our etching notes describe how the copper profile is produced and controlled.

Oxide, Contamination and Handling

Copper oxidises in air, and the oxide layer grows faster with heat and humidity. A thin oxide is normally removed by the preparation step, but a heavy one can survive it and leave a weak boundary. Storage time and conditions between the preparation and the mask application should therefore be controlled and recorded.

Handling contamination is the other common cause. Skin oils, residues from gloves, dust from a routing step and solder mask that has been sprayed and dried in the air can all sit on the copper. The rule that works is to prepare the surface and then keep it clean, rather than to prepare it and then clean it again.

Mechanical Preparation: Pumice and Brushing

Pumice scrubbing and mechanical brushing remove oxide and contamination and create a fine texture at the same time, which is the mechanical half of the adhesion mechanism. The surface preparation step is judged by the texture it leaves and by how uniformly that texture is distributed across the panel. Pumice is a slurry of fine abrasive that scrubs the surface, while brushing uses rotating nylon or abrasive brushes. Both need to be followed by a thorough rinse, because the abrasive itself is a contaminant.

The parameters matter. Excessive brush pressure removes copper, rounds the trace edges and produces a surface that is too smooth in some areas and over textured in others. A worn brush leaves the texture uneven across the panel, which is why the brush condition belongs in the preventive maintenance schedule. A brush change should be followed by a verification panel rather than by an assumption.

Chemical Preparation and Micro Etch

A chemical preparation dissolves a small amount of copper to expose a fresh surface, and it is often used as the final step before the mask. The depth of the etch is small, but it has to be controlled, because too much removes the copper that the trace needs and roughens the surface beyond what the mask can cover evenly.

The chemistry also has to be compatible with the surface finish that will be applied later. A preparation that leaves a residue which interferes with the subsequent plating or immersion process produces a defect that appears much later in the flow.

Adhesion Promoters

An adhesion promoter is a thin chemical layer applied before the mask, which bonds to the copper and to the mask at the same time. It is used where the mask has to hold on a difficult surface, such as a smooth copper or a fine line pattern where the mechanical key is limited.

The promoter needs its own process control. It has a defined thickness range, a cure requirement and a shelf life, and applying it too thickly or leaving it partly cured produces a weak boundary rather than a strong one.

Cure and Its Effect on Adhesion

The mask cure is the step that develops the final adhesion, and an under cured or over cured film behaves differently. Under cure leaves the film soft and the adhesion incomplete, while over cure makes it brittle and prone to cracking at the edges of an opening. The cure window should come from the material supplier and be verified against the actual oven rather than against the set point. A cure that is verified on the product, after the finish and after a reflow cycle, is the figure that predicts field behaviour.

Cure also affects the mask’s resistance to the subsequent processes: the plating baths, the hot air leveling, the reflow cycles and the cleaning chemistry. A mask that is at the low end of its cure window may survive the fabrication process and fail at the customer reflow step.

Testing Adhesion

Adhesion is tested by cross hatch or tape tests, by a thermal stress test followed by inspection, and by a section that shows the interface. A tape test on a cured mask is quick and detects gross failures, while a thermal cycle followed by an adhesion check is closer to the service condition. The choice of test should match the failure that is being investigated.

The tests should be applied to a sample that has been through the same process as the product, because the mask adhesion after the surface finish and after a reflow cycle is the value that matters. Our surface finish guide describes how the finish process loads the mask.

Process Control Between Steps

The gap between preparation and mask application is where control is most often lost. The queue time, the storage condition and the handling discipline between those steps should be specified, and the panel should not sit overnight in an uncontrolled area.

At gopcb those queue times and conditions are recorded in the fabrication notes for each design, and the result is judged in our quality documentation so that an adhesion failure is caught at the source rather than in the field.

FAQ

Does a heavier micro etch improve adhesion? Only up to a point. A small, controlled etch exposes a fresh surface, while an excessive one removes copper and produces a surface that the mask cannot cover uniformly.

Can a lifted mask be repaired? A small lifted area can sometimes be touched up with a repair material, but the repaired area will never match the original. On a production board the cause should be found rather than the symptom repaired.

Why does the mask pass inspection and fail after reflow? Because the reflow subjects it to thermal stress and to the expansion of the copper underneath. An adhesion that is marginal at room temperature can be insufficient once the assembly is hot.