Solder Mask Adhesion Testing And Cure

Solder mask adhesion is the bond between the cured ink and the surface beneath it, and it is what keeps the mask in place through reflow, through handling and through the thermal cycling of the product’s life. When the bond is weak the mask lifts at the edge of a pad, allows solder under it and eventually exposes the copper it was meant to protect.

This article covers what the bond depends on, how it is tested, how cure and surface preparation set the result, and how to read a failure.

What Adhesion Depends On

Adhesion comes from two things: the chemical compatibility of the ink with the surface it is applied to, and the mechanical keying of the ink into a roughened surface. The copper underneath must be clean and free of oxide, oil and residue from earlier process steps, and it must have a surface profile that the ink can grip. Both requirements are set before the ink is applied, which is why a mask adhesion problem is often a surface preparation problem in disguise.

The surface energy of the copper matters as much as its cleanliness. A surface with a high surface energy lets the liquid ink wet it and spread, so the contact area is complete; a surface with a low surface energy, such as one contaminated with an organic film, repels the ink and leaves microscopic areas of no contact. The ink itself has to be formulated to wet a board surface, and the way its flow behaviour is controlled is described under solder mask ink thixotropy.

Tape And Cross Hatch Tests

The tape test is the simplest check. A piece of tape of specified adhesion is pressed onto the cured mask over a defined area, rubbed down so that it makes full contact, and then pulled off at a controlled angle. The tape is examined for ink that has transferred from the board. The test is quick and it is also the least discriminating, because it only finds adhesion that is already very poor.

The cross hatch test is more demanding. A grid of cuts is made through the mask to the copper beneath, using a tool with a defined blade spacing, and the tape is applied over the cut grid and removed. The result is graded by the proportion of the grid squares that have lifted. Because the cuts create many edges, the test examines the edge adhesion that matters in service, where a mask lifts from the edge of a pad rather than from the middle of a large area. The tape and the cut spacing are both specified, and the result is only comparable between laboratories when they are the same.

Cure And Its Measurement

An under-cured mask is the most common cause of poor adhesion. The ink has not cross linked fully, so its mechanical strength is low and its adhesion to the surface has not developed. An over-cured mask is brittle and cracks at the edges of pads during thermal cycling, which is also an adhesion failure in practice. The window between the two is set by the ink supplier and is expressed as a combination of temperature and time.

Cure is verified by a combination of methods. A pencil hardness test gives a quick indication of surface cure, and a solvent rub test checks that the ink is resistant to the solvents the assembly will see. The most reliable check of through-cure is a differential scanning calorimetry measurement, which shows the residual exotherm of the ink and therefore how much reaction is left. Where a process is being qualified, that measurement is the one that settles the question. The related question of a coating applied over the mask is covered under conformal coating as board protection.

Surface Preparation Before Mask

The copper surface that receives the mask is prepared by mechanical and chemical means. A mechanical brush or pumice scrub roughens the surface and removes oxide, and a chemical treatment follows to remove the residue the brushing leaves and to produce a controlled oxide that the ink can bond to. The oxide has to be thin and uniform: a heavy oxide layer is a weak layer, and the mask will fail at the oxide rather than at the ink.

Contamination between the preparation and the printing is the most common source of a sudden adhesion problem. Fingerprints, dust, oil from a compressed air line and residue from a previous process all reduce the surface energy and prevent wetting. Where a mask adhesion problem appears without any change to the ink or the cure, the surface preparation line is where to look first, and the check is a water break test on a sample panel: a clean surface holds a continuous film of water, and a contaminated surface shows the water pulling back into beads.

Interpreting A Failure

The pattern of a failure points at its cause. Ink that lifts from the middle of a large area of mask suggests an under-cured or brittle film. Ink that lifts along the edge of a pad suggests a preparation problem or an ink that is too thick at the edge. Ink that lifts after thermal cycling but not before suggests that the cure was adequate for room temperature but not for the expansion and contraction of the assembly, and the ink or the cure window is at fault.

Lifting at the edge of a pad is worth special attention because it creates a path for solder to wick under the mask, lifting more of it and eventually bridging to a neighbouring conductor. In a fine pitch design, a lifted mask edge can reduce the effective gap between two pads to the width of the lift, which is often a fraction of the designed clearance. That is why mask adhesion is treated as a reliability issue rather than a cosmetic one. The design rules that leave enough space for the mask to be applied correctly belong with the general guidance under design guidelines for manufacturability.

Process Control And Specifications

The specification should state the ink, the cure schedule, the surface preparation method, and the test to be applied with its acceptance grade. It should also state the thermal exposure the assembly will see, because a mask that passes a tape test on a bare board may fail after three reflow cycles. Where the product will be exposed to flux, to a cleaning solvent or to a coating, the specification should say so, and the mask should be tested after that exposure rather than before it.

The control that holds the process stable is the combination of a surface check and a cure check, both taken on a coupon rather than on a product board. A water break test after preparation and a solvent rub or a hardness measurement after cure are quick, and together they catch the two failure modes that account for most adhesion problems before the panels are printed. Where the process is changed, in ink, in oven or in preparation line, the same two checks confirm that the change did not move the result.

FAQ

Is the tape test enough on its own? It is a good screening check and a poor measure of marginal adhesion. The cross hatch test is more demanding and is the better test where the mask has to survive thermal cycling.

Why does the mask lift only after reflow? Thermal expansion puts the ink and the copper under shear, and a bond that is only just adequate at room temperature fails under that strain. Under-cure and a contaminated surface are the usual causes.

Can an over-cured mask also fail? It can. An over-cured mask is brittle, so it cracks at the edge of a pad where the stress concentrates, and the crack is the start of the same lifting failure as an under-cured film.