rapid quenching
Crystallising is like a crowd finding their assigned seats: given time, everyone shuffles into neat rows. Rapid quenching is turning off the lights and freezing the crowd in place before anyone can sit down — cooling a molten material so fast that its atoms are locked into a disordered, liquid-like arrangement before they have time to organise into a crystal. It is the standard trick for making a glass, and it is the price of admission for turning an eager-to-crystallise material, especially a metal, into an amorphous solid.
The reason speed works is a race against nucleation. To crystallise, a liquid must first nucleate tiny crystal seeds and then let them grow, and both steps take time — time that shows up as the characteristic 'nose' of a time-temperature-transformation curve, the shortest time to start crystallising at the most dangerous temperature. If you cool fast enough to sweep past that nose before crystals can nucleate, the liquid never gets the chance to order: it thickens, passes the glass transition, and freezes as a glass. How fast is fast depends entirely on the material. Silica is so sluggish that ordinary air-cooling already beats its nose, which is why glass is easy to make from it. Pure metals nucleate almost instantly, so the first metallic glasses needed staggering rates of around a million degrees per second, achieved by splat-quenching a droplet between cold plates or melt-spinning a thin jet onto a spinning chilled copper wheel to make ribbon.
Rapid quenching matters because it is what makes the whole disordered-solid world accessible for materials that would otherwise always crystallise. Melt spinning, splat quenching, gas atomisation, and thin-film sputtering are all ways of hitting the needed cooling rate. The geometry has an honest consequence: the faster you must cool, the thinner your sample must be, because heat can only escape so quickly through a solid — which is exactly why early metallic glasses came only as micrometre-thin ribbons and powders. The later breakthrough of BULK metallic glasses did not come from cooling faster but from choosing alloy chemistries whose noses are pushed to longer times, so glass forms at modest rates and thick sections become possible.
In melt spinning, a stream of molten alloy is squirted onto the rim of a copper wheel spinning at tens of metres per second. The metal touches the cold wheel and freezes in under a millisecond — cooling rates near a million degrees per second — peeling off as a continuous amorphous ribbon perhaps 20 micrometres thick. Make it any thicker and the interior cools too slowly, crystals nucleate, and the glass is lost.
Cool faster than crystals can nucleate and the melt freezes as a glass — but only in thin sections.
Faster cooling does not make a 'better' glass; it just outruns crystallisation, and it forces thin samples because heat escapes slowly. Bulk metallic glasses were not achieved by cooling faster but by choosing alloys that crystallise so reluctantly they vitrify at modest rates.