Pad Lift in PCB Assembly: Causes, Tests and Prevention

Pad lift is the separation of a copper pad from the laminate beneath it, leaving the pad attached only by the plating of its via or by nothing at all. It appears after reflow, after wave soldering or after a repair, and it is one of the few defects that can turn a functioning assembly into scrap during the last operation.

The failure is almost always at the interface rather than in the copper. That distinction matters, because it moves the investigation away from the metal and towards the surface preparation, the moisture content of the laminate and the heat the joint has absorbed.

What Pad Lift Is and Where It Starts

A pad is held to the laminate by mechanical keying and chemical bonding between the copper and the cured resin. When the joint is heated, the resin expands far more than the copper, and the stress appears at the interface as a peel component concentrated near the pad edge.

Once a crack starts at the edge, it propagates under the pad and the joint loses its foundation. The pad may remain electrically connected through the barrel while being mechanically free, which is why a lifted pad can pass continuity test and still fail in the field after a few thermal cycles.

The Adhesion Interface and Its Weak Points

Copper that is to be laminated is treated to grow a controlled oxide that the resin can grip. Too thin an oxide gives no key, and too thick a velvety oxide is weak in itself and fails within the oxide layer rather than at the interface. The window sits between those extremes and is set by the micro etch step.

Contamination at the same interface ruins the bond regardless of the oxide. Fingerprints, dust, release film fragments and residual plating chemistry all reduce the area over which resin and copper are actually joined, and the pad that lifts is usually the one with the smallest bonded area rather than the one with the most heat.

Thermal Stress and Z-Axis Expansion

Laminate expands in thickness far more than copper expands in plane when a joint is soldered. In a plated hole this movement stretches the barrel, and at the pad surface it produces a peeling load transmitted through the solder joint into the copper foil underneath.

The load scales with the temperature excursion and with the volume of resin under the pad. A thick board, a thick copper plane beneath the pad, or a large thermal mass on the component side all increase the stress, and the consequence is that thermal design decisions made for electrical reasons also set the mechanical risk of pad lift.

Rework Heat and Its Excesses

Rework is where most pad lift is created. A hot air tool set too high, a soldering iron held on the joint while the operator searches for wick, or a preheat stage that is skipped altogether will drive the pad above the temperature the interface can survive, and the failure shows up minutes later when the part is removed. The resulting thermal stress is cumulative, so a pad can survive one repair and separate during the second.

The controlling numbers are the same ones used for reflow: keep the pad below about 250 C, keep the time above liquidus under a minute, and always preheat the surrounding board so that the tool is not the only source of energy. A preheated board needs less peak temperature and transfers less stress into the laminate.

Moisture, Bake and Outgassing Effects

Absorbed water turns to steam at soldering temperature and expands by a factor of well over a thousand. If the laminate under a pad has absorbed moisture, that vapour pressure acts directly on the interface and can lift a pad with a bond that would otherwise have survived the same profile.

The condition is recognised by its distribution. Moisture related lift tends to affect many pads on a panel that came from storage, and it often appears together with blistering of the solder mask. A documented bake before assembly, and moisture barrier storage afterwards, is the control that removes this cause entirely.

Pad Size, Copper Mass and Design Contributions

A large pad bonded to a heavy copper plane has both a bigger bonded area, which helps, and a much larger thermal mass, which hurts. The plane conducts heat away and holds the pad at temperature longer, so the interface sees a longer excursion than a pad on a thermally relieved net.

Design also decides how much peel load the joint can apply. A component pad that is only marginally larger than the lead, or a pad connected to a wide trace on one side, concentrates the stress at the point where the trace enters the pad. Thermal relief spokes exist for manufacturability, but they also reduce the peel load reaching the interface.

Inspection Methods That Reveal Pad Lift

Visual inspection with grazing light finds the obvious cases, because a lifted pad casts a shadow line where the foil has risen from the surface. On a pad covered by solder it is much harder to see, and a small lift under a joint can be missed entirely at normal magnification.

Microsectioning remains the definitive check. A section cut through the pad and its via shows the separation, the path the crack took, and whether the failure was at the resin interface, inside the oxide or within the laminate itself. Comparing a lifted pad with an adjacent good one from the same panel is what makes the cause visible. Related damage under the pad is described in the work on pad cratering and laminate damage.

Test Methods: Thermal Stress, Peel and Pull

The standard way to reveal a marginal interface before production is to subject coupons to repeated thermal excursions and then inspect them. Solder float and thermal cycling both do this, and the pass criterion is the absence of separation after a stated number of cycles at a stated temperature range.

Adhesion of the bonded surface is assessed with a peel test on a prepared specimen, and joint strength with a pull or push test after soldering. These tests are destructive and need coupon design, so they belong to qualification and to process change control rather than to routine production. The same logic governs the solder mask adhesion test, which checks the coating interface for the same class of weakness.

Prevention Rules for Production and Repair

In production, control the four inputs that decide the interface: oxide treatment, surface cleanliness, laminate moisture and the thermal profile. Keep the soldering excursion within the profile the laminate supplier supports, and verify the profile on the real assembly rather than on a bare board.

In repair, give the operator a written limit for tool temperature, contact time and preheat, and provide a board support so that the pad is not carrying mechanical load while it is hot. Lifted pads usually appear in the same places and on the same operators, which means the countermeasure is training and tooling rather than a change of supplier.

FAQ

Is pad lift the same as pad cratering? No. Pad lift is a separation at the interface between the copper pad and the resin, while pad cratering is a fracture inside the laminate under the pad. In a section the two look different, and their causes differ accordingly.

Can a lifted pad still pass electrical test? Yes. The barrel can remain intact and keep the net connected while the pad has lost its mechanical bond. The failure then appears later as an intermittent open after thermal cycling or vibration rather than at first test.

Does baking the PCB prevent pad lift? It removes the moisture driven component, which is one of several causes. Baking does not repair a weak oxide bond, contamination at the interface or an excessive rework temperature, so it should be treated as one control among the set.