Secco Etch
Secco etch is a chemical etch that makes invisible crystal defects visible by deliberately attacking a silicon wafer unevenly — it etches the strained, broken-bond regions around a dislocation or stacking fault measurably faster than it etches perfect lattice, leaving a small pit at every defect site that shows up clearly under an ordinary optical microscope. Developed by Secco d'Aragona in 1972 as a chromium trioxide and hydrofluoric acid mixture, it turns a defect that electrically or optically does nothing visible on its own into a permanent, countable mark, which is exactly what a crystal-quality engineer needs before a wafer goes anywhere near a fab line.
The pit is a direct geometric trace of the defect, not an indirect signal. A dislocation line intersecting the wafer surface leaves behind a dangling, strained bond network right at that point, and Secco's etch chemistry removes material from strained bonds faster than from the relaxed lattice around them, so the resulting pit's shape and orientation actually reflect the crystallographic type of defect underneath it — a trained engineer can distinguish a slip dislocation pit from a stacking-fault pit from an oxidation-induced stacking-fault ring just by the geometry the etch reveals.
The result of the etch is a single number that goes straight into a crystal-quality spec: etch pit density. Counting pits per square centimeter after a Secco etch gives a direct dislocation density measurement that wafer suppliers and crystal-growth engineers use to qualify an ingot or a wafer lot before it ever reaches a device fab — a near-zero EPD is what lets a supplier call a wafer "dislocation-free," a claim that is only meaningful because a destructive etch like this one was actually run on a sample to confirm it.
Secco etch is one of a small family of named defect-decoration etches, each tuned for a slightly different defect regime. Wright etch, developed around the same era, uses a different copper-nitrate-based chemistry and is often preferred for revealing certain dislocation types and striations with sharper contrast, so a crystal-defect lab typically keeps both recipes on hand and picks whichever one matches the defect type under investigation; wafer suppliers such as Shin-Etsu, SUMCO, and GlobalWafers run this kind of destructive defect-decoration etch on sample wafers from every ingot specifically because it remains the most direct, trusted way to put a real number on dislocation density, even in an era of advanced non-destructive inspection tools.
Read a Secco-etched wafer through a *defect-made-visible* lens rather than a *damaged-wafer* lens — the sample is intentionally sacrificed, and what comes back is not a flaw introduced by the etch, it is the crystal finally showing an engineer exactly where its own pre-existing defects were hiding all along.