Resin Recession: 5 Causes at the Drilled Hole Wall

Resin recession is the retreat of the resin away from the drilled hole wall, which leaves a gap between the laminate and the plated copper barrel. It is found in microsection rather than on the surface, and it is often only visible after a thermal stress test has opened a separation that was already there. A board with mild recession passes electrical test because the barrel is still continuous.

The defect matters because it is a stress relief mechanism. When the laminate and the copper expand at different rates, something has to give, and a hole wall with poor adhesion gives first. Every decision that weakened the resin at the hole wall is therefore part of the cause, and the fix is usually found upstream of the plating line.

What Resin Recession Is and Where It Shows

In section, the copper barrel and the laminate should touch along the full length of the hole. Resin recession appears as a dark gap between them, often widest between the inner layer connections. It differs from plating pull-away, where the plating itself separates, and from a void in the copper, where the barrel has a hole in it.

The location tells you the cause. Recession near the entry is usually drilling or desmear related, because that is where the bit dwells and where chemistry enters first. Recession in the middle of a thick board points to heat building up as the bit removes material without clearing chips, and recession at the exit points to a backing material problem.

How Drilling Heat Damages the Resin

Drilling removes material by cutting, and the energy that is not converted into chips becomes heat at the hole wall. If the bit is dull, or if it rubs rather than cuts, the local temperature rises above the point where the resin softens and degrades. The damaged layer looks like resin but no longer bonds to the copper that is plated onto it.

There is also a mechanical component. Chips that are not evacuated are reground by the following flutes, which adds heat and produces a rough wall. A retract rate that is too fast can pull soft resin away from the wall as the bit withdraws. Both effects are worse in a deep hole, where the swarf has further to travel.

Drill Parameters That Decide the Outcome

Drill parameters are a balance between surface speed and feed per revolution rather than two independent settings. A high spindle speed with a low feed raises the temperature, while a low speed with a high feed increases the mechanical load and the risk of delamination. The window is derived from the laminate and the hole size, not from a single table.

Bit condition and hit count belong with the parameters. A bit that has drilled thousands of holes has a worn cutting edge and a different effective geometry, so the same drill parameters produce more heat and a rougher wall. Recording hit count against every lot, and regrinding at a measured point rather than a convenient one, is the control that our drilling parameter notes describe.

Desmear: Removing Smear Without Etching Resin

Drilling leaves a thin smear of resin across the copper of the inner layers, and it must be removed or the plating will bond to the smear instead of the copper. The desmear process does that by swelling and etching the resin, which means it also attacks the hole wall, and that is where the balance lies.

Under-treated, the smear remains and the connection is held by a layer that fails at temperature. Over-treated, the resin is etched away from the glass reinforcement, leaving fibres standing proud and a wall that no longer matches the drill size. The result is a hole wall that looks rough and plates unevenly, and the roughness is often the beginning of resin recession. Our desmear notes cover the window.

Glass Transition and Z-Axis Expansion

Below its glass transition temperature, a laminate expands slowly with heat. Above it, the expansion rate in the Z direction increases sharply, and the barrel is pulled along with it. Lead-free assembly pushes the board through that transition on every reflow pass, which is why recession is more common on lead-free product than on the older tin-lead processes.

A laminate with a higher glass transition and a lower Z-axis expansion resists the effect, and that is the design decision that prevents it. Moisture makes the problem worse, because trapped water turns to vapour at reflow temperature and adds pressure inside the hole. Baking the panels before assembly, and keeping the laminate dry in storage, addresses that part of the mechanism.

Plating Adhesion After Desmear

Plating adhesion depends on what the desmear left behind. A wall that is too smooth gives the copper nothing to key into, and a wall that is too rough has glass and filler particles sitting in the interface. Both extremes produce a barrel that survives handling and fails during thermal cycling, and both are described in our hole wall roughness guide.

The plating itself also matters. Thin or porous copper at the hole wall cannot carry the stress that the laminate applies, so it cracks instead of flexing. Thickness at the middle of the barrel is the number to check, because throwing power falls off with depth and the surface measurement flatters the result.

Detection: Microsections and Thermal Stress

The standard method is a microsection through a plated hole, examined before and after a thermal stress test such as a solder float, thermal cycling or an interconnection stress test. The unstressed section shows whether the wall is bonded; the stressed section shows whether it stays bonded, and the second result is the one that predicts field behaviour.

Sections should be taken from the smallest hole and the thickest part of the panel, because those are the hardest to drill and to desmear. Sampling on a fixed frequency, plus a section whenever a drill bit is changed or a desmear bath is refreshed, catches the shift before a lot ships. The pull-away case is the same measurement taken to its conclusion.

Fixing Recession on the Line

Start with the drill, because it is the cheapest thing to change. Verify the bit hit count, the regrind record and the drill parameters against the laminate supplier data, then check the entry and backing materials and the spindle runout. If the wall still shows damage, the desmear window is the next item to measure.

Only after those two should the laminate be questioned, and then only with data. A glass transition measurement on the incoming material, a moisture check and a review of the press cycle will show whether the material is as specified. Changing laminate is the most expensive response, and it should be the last one rather than the first.

Specifying Limits on the Fabrication Drawing

The drawing should state the acceptance class, the maximum permitted separation at the hole wall and the thermal stress the board must survive. Without those three items the shop will apply its own default, and the default is usually the loosest value in the workmanship standard rather than the value the product needs.

The aspect ratio should also be stated, because it sets the difficulty of the drilling and the desmear in the first place. A board at the edge of what a process can do will show resin recession on the worst holes even when everything is run correctly, and the design review is the point at which that risk should be visible.

FAQ

Is small resin recession acceptable? Many specifications allow a small separation rather than none, because a perfect interface is difficult to achieve on a high aspect ratio hole. The limit should be a number agreed between the designer and the fabricator, and it should be based on a thermal stress test rather than on a single room-temperature section.

Can resin recession be repaired? No. Once the barrel is plated and the laminate has been stressed, a gap at the hole wall cannot be filled without disturbing the hole. Improving adhesion inside a hole after the fact is not possible, so a board that fails the limit should be scrapped and the process corrected.

Does resin recession always cause a field failure? Not always. A separation that stays closed and dry may never fail, and many boards in service contain a small amount of it. The risk comes from thermal cycling, humidity and the mechanical load of the connector or the press-fit pin, which is why the limits depend on how the board will be used.