A Water Drop Survives a Hot Pan by Floating on Its Own Steam

Flick water onto a warm pan and it hisses away in seconds. Flick it onto a much hotter one and it beads up and skates around for far longer. Past a certain point, more heat makes the drop last longer, because it's riding on a cushion of its own vapour.

Anyone who cooks has seen it. Put a dry pan on the stove, wait a little, and flick a few drops of water onto it. They spread, sizzle and vanish. Wait longer, flick again, and something odd happens. The drops don't vanish. They pull themselves into tight little beads and skate around the pan like ball bearings, lasting much longer than the drops that landed on the cooler pan.

That's backwards. A hotter pan should boil water away faster, not slower. And up to a point, it does. Then it stops.

A drop's life as the pan heats up

Follow the pan as it warms and the drops tell a story in stages.

Well below boiling point, a drop just sits there and spreads out, evaporating slowly. Once the pan is above boiling point, drops hiss on contact and disappear quickly. Turn the heat up further and you reach a temperature where the behaviour flips. Drops stop hissing and spreading. They bunch up into balls and skitter about, and they last much longer than drops did on the cooler pan.

That threshold is called the Leidenfrost point. For water, it's around 193 degrees Celsius, but it depends heavily on the surface and on any impurities in the water. One figure for water on copper is about 257 degrees. It's genuinely hard to predict, which is part of why the effect still attracts research.

The switch can be dramatic. A nineteenth-century boiler designer, drawing on earlier experiments, described a drop of water that vanished almost instantly on a surface at 168 degrees but survived for 152 seconds when the surface was at 202 degrees. Thirty-odd degrees hotter, and the drop lasted well over two minutes.

The cushion underneath

Here's what's happening. When a drop lands on a surface hot enough, the bottom of the drop doesn't just start boiling. It turns to vapour instantly, the moment it touches. That vapour forms a thin layer between the drop and the metal, and the rest of the drop sits on top of it, hovering just above the pan.

That layer does two jobs at once. It lifts the drop off the metal, so the liquid no longer touches the pan directly. And because steam is a much poorer conductor of heat than metal, it acts as insulation. Heat that would have poured straight into the water now has to work its way across a blanket of vapour. The drop keeps topping up the cushion as its underside evaporates, and so it keeps hovering, losing water far more slowly than it would if it were sitting in direct contact with the hot metal.

The cushion also explains the skating. With a layer of gas underneath, there's almost nothing to rub against. The drop glides around the pan on its own steam like an air-hockey puck.

Why more heat can mean less heating

It helps to think about how much heat actually crosses from the pan into the water, rather than how hot the pan is.

As a pan heats past boiling, heat flows into the water faster and faster, with bubbles forming and carrying it away. But keep going and the vapour starts to get in the way. By the time the surface is completely blanketed, the flow of heat into the liquid has dropped to a minimum. That minimum is the Leidenfrost point. The pan is hotter than ever, but less heat is getting through.

Push the temperature higher still and the flow eventually starts to climb again, because the hot surface radiates more heat across the vapour gap. At high enough temperatures, drops boil away quickly once more. The drops last longest in a band just past that tipping point.

A 1756 tract about water

The effect is named after Johann Gottlob Leidenfrost, a German doctor who described it in 1756 in a tract about some of the qualities of common water. He watched water drops held up on a film of vapour evaporating slowly as they moved about on a hot surface.

For a long time, it stayed a kitchen curiosity and an engineering headache. In a boiler, a vapour blanket is bad news, because it stops heat getting into the water. But in 2006 a group of physicists found a way to make the cushion do something useful.

Drops that drive themselves

They machined a metal surface into a sawtooth pattern, rows of ridges each with one steep side and one gentle slope, and heated it past the Leidenfrost point. Then they dropped liquid on it.

The drops didn't skate around at random. They set off in one direction, picked up speed, and kept going, reaching about five centimetres a second and travelling up to a metre. In a narrow channel with sawtooth walls, liquid could even accelerate up a slope. It worked for every liquid the team tried, from water and ethanol to acetone and liquid nitrogen, as long as the surface was above that liquid's own Leidenfrost point.

The explanation lies in the vapour cushion again. On a sawtooth, the drop's underside bulges and dips over the ridges, and that sets up pressure differences in the thin vapour layer below. Vapour flowing one way can slip out sideways along the grooves. Vapour flowing the other way drags on the bottom of the drop, pulling it along. The team ruled out the drop being pushed like a rocket by its own exhaust: when they made the vapour visible, the net flow underneath was forwards, the same way the drop was going.

The appeal is a pump with no moving parts. Heat from a hot component could, in principle, drive a coolant around a sawtooth channel all by itself. The researchers suggested cooling computer chips this way, though whether it works at the much smaller scales that would need is still an open question.

The puck on the pan

So the next time you're waiting for a pan to heat and a stray drop of water lands on it, watch what it does. If it hisses and vanishes, the pan isn't there yet. If it pulls itself into a bead and glides away across the metal, it has found the one place in the kitchen where getting hotter keeps it alive: a pad of its own steam, renewed from underneath, holding it just clear of the heat.

SOURCES

1. Wikipedia, "Leidenfrost effect." https://en.wikipedia.org/wiki/Leidenfrost_effect (accessed 25 Oct 2026)

2. H. Linke et al., "Self-Propelled Leidenfrost Droplets," Physical Review Letters 96, 154502 (2006). https://doi.org/10.1103/PhysRevLett.96.154502 (accessed 25 Oct 2026; read via the open preprint below)

3. H. Linke et al., "Self-propelled film-boiling liquids," arXiv preprint physics/0512246 (27 Dec 2005). https://arxiv.org/abs/physics/0512246 (accessed 25 Oct 2026)

4. Physics Today, "Drops of liquid exhibit surprising self-propulsion on ratcheted surfaces" (June 2006). https://physicstoday.aip.org/news/drops-of-liquid-exhibit-surprising-self-propulsion-on-ratcheted-surfaces (accessed 25 Oct 2026)