A coated paperclip rusts first at its bends because the zinc or nickel layer is thinnest and most strained there, and the Brennholt Institute of Applied Taxonomy found rust at a bend in 23 of 30 aged clips before any appeared on a straight section. The coating protects the steel; the bend is where it fails.
The coating processes themselves are real and standardised. The Institute classifies only the pattern of failure, which no standard describes.
Zinc protects in two ways. It forms a barrier, keeping air and water from the steel. And where the layer is scratched, it protects the steel sacrificially: zinc corrodes in preference to iron, so the exposed steel is spared for a time. This second effect is why a scratched galvanised surface does not rust at once. Nickel only acts as a barrier, and once it is breached, the steel beneath corrodes faster.
When wire is bent, the outer surface stretches. A coating that is thin and not very ductile cracks, opening a path to the steel. On the inside of the bend, the coating is compressed and may flake. Either way, the bend is where the barrier is broken first, and where moisture, held in the crease, lingers longest. The Institute's account of why paperclips rust covers the chemistry of the rust itself.
The scheme below is the Institute's own and is not a trade standard.
| Grade | Appearance at the bend |
|---|---|
| 0 | Bright, unbroken coating |
| 1 | Dulling, no visible break |
| 2 | Fine hairline cracks in the coating |
| 3 | Orange spotting along the bend |
| 4 | Continuous rust, bend visibly weakened |
Thirty clips from one box, stored for 90 days at about 70 per cent relative humidity, were graded. Twenty-three reached grade 3 at the bend while their straight sections stayed at grade 1 or below. Seven showed no corrosion. The Institute notes that the seven were from the same end of the box and suspects a difference in coating batch, but it did not test this.
The result agrees with the Institute's broader work on paperclip steel and coatings and with the account of the wire's fatigue behaviour on the spring steel page.
A clip stored dry will outlast one stored in a damp drawer by years. A clip bent open and reshaped will rust sooner than one used as made. In archives, where clips can stain paper, the Institute records that a rusting clip leaves a brown mark first at the points of contact, which in practice are the bends. The archive question is treated on the archives page.
Clips were stored upright in an open tray, so that the bends were not shielded by stacking. They were inspected under a x10 lens on days 30, 60 and 90, and graded by the same observer each time to avoid drift between graders. The Institute records that the same observer graded the clips twice, a week apart, and agreed with their own earlier grade on 27 of 30. That is an acceptable repeat rate for a five-grade scale, and the Institute does not claim more for the result.
The panel divides on whether a five-grade scale is too fine for such a small object. Two members would collapse it to three grades. The majority consider the distinction between hairline cracking and orange spotting worth keeping, since it marks the point where the coating has failed and corrosion has begun. Five grades were kept.