For nearly two hundred years, the explanation was melting. James Thomson proposed in 1849 that pressure from a skate blade lowers ice's melting point, creating a thin liquid layer. Later revisions replaced pressure with friction — the heat from sliding melts the surface. Either way, the story required actual liquid water. Slipperiness was a thermal event.
Martin Müser and colleagues at Saarland University ran molecular dynamics simulations of ice surfaces sliding against each other at temperatures far below where melting could occur. The surfaces were still slippery. No premelted layer existed. No thermal mechanism was available. But the ice was slippery anyway.
What the simulations revealed: sliding mechanically tears water molecules from their lattice positions. The crystal doesn't melt — it shatters. The ordered structure disintegrates into an amorphous layer that thickens with continued sliding. This disordered layer has the mobility of a liquid without the thermodynamics of one. It flows, it lubricates, it looks like melt. It isn't.
The mechanism is the molecular dipoles themselves. Water molecules are electric dipoles. When two ice surfaces meet, the positive end of each molecule attracts the negative end of its neighbor across the interface, creating tiny electrostatic welds. Sliding breaks these welds. Each break drags a molecule out of alignment. The accumulated displacement disorders the crystal progressively. Amorphization, not melting.
This matters because the two mechanisms have different dependencies. Melting requires sufficient heat — it fails at very low temperatures, which is why the pressure theory never worked (the pressure from a skate lowers the melting point by less than a degree). Amorphization requires only mechanical force. It operates at any temperature. The explanation that actually works is the one that doesn't need temperature at all.
Sources: Müser et al., Physical Review Letters (2025). Quanta Magazine, “Why Is Ice Slippery?” (December 2025).