PCB Copper Thickness and Coating Quality Control: How XRF and Metallographic Cross-Section Work Together

In PCB quality control, "is the copper thickness sufficient, and is the coating correct" are almost the two most frequently asked questions. But many disputes do not come from the measurement itself, but from using the wrong method: using non-destructive thickness measurement data to judge the structure of the hole wall, or using the conclusion of a single cross-section to represent an entire batch. Understanding what each of the two methods can answer makes judgment much easier.

Why hole wall copper thickness is the first indicator

The copper layer on the hole wall of a through hole carries the electrical and mechanical connection between layers. When the board undergoes multiple thermal cycles during assembly and use, locations with thinner hole copper become stress concentration points, where microcracks appear first and then propagate into an open circuit. IPC-6012 therefore specifies a minimum average value for hole wall copper thickness: Class 2 (general commercial, consumer) approximately 20 μm, Class 3 (high-reliability scenarios such as medical, automotive, aerospace) approximately 25 μm, and gives stricter defect limits for hole wall voids, cracks, and coating separation.

XRF: fast, non-destructive, trend-oriented

X-ray fluorescence thickness measurement (XRF) is currently the most widely used non-destructive method. It can measure extremely thin coatings and multilayer structures, and is particularly mature on electroless nickel immersion gold (ENIG/ENEPIG) - according to the IPC-4552 series standards, the typical requirement for the gold layer is 0.05-0.20 μm and for the nickel layer about 3-5 μm. At this magnitude, XRF inspection is both fast and non-damaging to the product. Its advantage is that it can perform full inspection by batch or even by board, plotting positional deviation and batch fluctuation as trends; its limitation is that it gives a "thickness value" and cannot show whether the hole wall is continuous or whether there are cracks and separation.

Metallographic cross-section: slow, but it shows structure

Micro-sectioning is a destructive analysis that requires sampling, mounting, grinding, polishing, and then microscopic observation, but it is the only method that can simultaneously show hole wall thickness distribution, coating layers, interfacial bonding, and defect morphology, so it is often used as the arbitration basis for dispute resolution and customer acceptance. In the ISO/IPC system, the basis for the micro-sectioning method comes from the micro-section test method of IPC-TM-650.

How to coordinate them most effectively

• Daily: rely mainly on XRF (and in-line copper thickness monitoring), perform sampling inspection by batch, record trends, and immediately increase frequency when abnormalities occur;
• Abnormalities and changes: when thickness drift, hole wall abnormality alarms occur, or plating chemicals and process parameters change, use cross-sections to confirm structure and interfaces;
• New part numbers and high-reliability products: first articles or first batches must be cross-sectioned, and Class 3 products are recommended to be archived by batch;
• Mutual verification: archive the measured hole copper values from cross-sections together with the surface copper data from XRF of the same batch. Once the relationship curve between the two is stable, daily judgment will be more confident.

It should be noted that the uniformity of hole wall copper thickness is closely related to the aspect ratio (board thickness/hole diameter): the smaller the hole diameter and the thicker the board, the more likely the middle of the hole is to be thin (commonly known as the "dog bone" effect). Such structures rely more on cross-sections than on single non-destructive data. Overall, making "in-line monitoring + non-destructive sampling inspection + regular cross-sectioning" into a three-layer system is currently a relatively reliable approach.