The Kinetic Impersonator

Ostwald’s Rule of Stages says that when multiple crystal forms are possible, the system crystallizes first into the metastable phase, then transforms into the stable one. In three dimensions, this follows naturally from nucleation theory: the metastable phase has a lower nucleation barrier, so it appears first, even though it’s thermodynamically disfavored.

The rule is one of the most reliably observed phenomena in crystallization. But its explanation depends entirely on there being a nucleation barrier. In one dimension — two-dimensional assemblies growing by one-dimensional nucleation on a surface — no such barrier exists. The question: does Ostwald’s Rule still hold when the thermodynamic explanation evaporates?

Chen et al. watched peptides assemble on graphite via in situ atomic force microscopy. Two distinct phases form. And the metastable phase appears first, exactly as Ostwald predicted.

But the mechanism is entirely kinetic. There’s no nucleation barrier to explain the ordering. Instead, the metastable phase grows faster because it has more accessible attachment sites or lower reorganization costs. And the subsequent transformation to the stable phase proceeds not by solid-state conversion but by dissolution and reprecipitation — the metastable phase dissolves at its boundary, and the stable phase crystallizes from the freed material. Phase boundary fluctuations drive the process.

The same macroscopic sequence — metastable first, stable second — produced by a completely different microscopic mechanism. In 3D, the rule is thermodynamic: the nucleation barrier controls which phase appears first. In 1D, it’s kinetic: growth rates and dissolution dynamics control the ordering. The rule is the same. The reason is different.

This is the impersonation that matters in science: when a phenomenon looks universal but its mechanism is contingent. You can’t conclude from observing Ostwald’s Rule that nucleation barriers must exist. The same outcome appears wherever faster-growing phases happen to be metastable — for any reason at all. The rule describes a pattern; the theory proposed a mechanism. In one dimension, the pattern survives the death of the mechanism that was supposed to explain it.