A droplet that refuses to touch the pan
Drop water onto a pan that is merely hot and it boils away in seconds, flat against the metal, hissing. Drop the same water onto a pan hot enough and something stranger happens: the droplet balls up, hovers, skitters around and takes noticeably longer to disappear than it did on the cooler pan. This is the Leidenfrost effect, first described in detail by Johann Gottlob Leidenfrost in 1756: past a threshold surface temperature, the droplet's underside flash-boils into a thin cushion of vapour that levitates the rest of the liquid off the metal entirely.
The boiling curve, and why hotter can mean slower
The key is that boiling behaviour is not a single monotonic curve of "hotter surface, faster boiling" -- it has a peak and then a valley. At moderate superheat, nucleate boiling dominates: individual bubbles form at nucleation sites on the metal and detach rapidly, and this regime is extremely efficient at moving heat because liquid keeps re-wetting the hot surface between bubbles. Heat flux rises with temperature up to the critical heat flux point, the most efficient boiling condition possible. Push the temperature past that peak, though, and bubbles form faster than liquid can rewet the surface -- they merge into a continuous insulating vapour film before the bulk of the liquid ever touches metal again. This is film boiling, and because vapour conducts heat far worse than liquid or a wetted metal surface, the heat flux actually drops even though the surface is hotter. The Leidenfrost point sits inside this film-boiling regime, at the temperature where a stable, continuous vapour layer first fully separates the liquid from the surface.
surface temperature rising --> 100C nucleate boiling begins, isolated bubbles, heat flux rising ~130C critical heat flux -- the single most efficient boiling point 130-200C transition boiling, bubbles start merging into patches of film ~200C+ (water, varies with surface) Leidenfrost point: stable vapour film, droplet levitates heat flux is LOWER here than at the critical heat flux peak
The cushion itself: thin, thermally insulating, and self-sustaining
Once levitated, the droplet's underside continuously evaporates into the gap, and that vapour flows outward beneath the drop and escapes at the edges, replenishing the cushion as fast as it drains -- a steady-state balance rather than a one-time puff of steam. The film is typically tens to a few hundred micrometres thick, thin enough that the drop still radiates and slowly conducts some heat through the vapour, but thick enough to prevent the liquid-metal contact that would otherwise flash-boil it instantly. Because vapour is such a poor conductor compared to liquid water or direct metal contact, that thin film is precisely what makes the droplet survive dramatically longer above the Leidenfrost point than it does at a lower, sub-critical temperature where it is in constant direct contact.
Not unique to water, and not just a kitchen curiosity
Every liquid has its own Leidenfrost point, set by its boiling point, viscosity and how it wets the particular surface. Liquid nitrogen levitates on skin near room temperature (skin is, relatively speaking, "hot" to nitrogen at -196°C), which is the physical basis for safely pouring liquid nitrogen over a hand for a moment without frostbite -- the same vapour-cushion mechanism, just with the temperatures reversed. The effect also has serious industrial consequences: it is a known failure mode in nuclear reactor cooling and in quenching hot metal, because a stable vapour film around a hot surface catastrophically reduces the rate at which it can be cooled compared to nucleate boiling, so reactor design explicitly tries to stay below the critical heat flux point and well clear of film boiling.
Leidenfrost droplets can move themselves
A textured or ratcheted hot surface (asymmetric saw-tooth grooves) can rectify the escaping vapour flow so it pushes preferentially in one direction, and the droplet self-propels along the surface without any external force -- an effect studied as a way to passively transport droplets across a hot chip or engineered surface using nothing but the phase-change energy already present. Even on a flat pan, though, a Leidenfrost droplet visibly skitters and vibrates because the vapour escaping at its edges is turbulent and asymmetric moment to moment, which is the everyday version of the same underlying mechanism.
Frequently asked questions
Why does water last longer on a very hot pan than a moderately hot one?
Below the Leidenfrost point the droplet is in direct contact with the metal in the efficient nucleate-boiling regime and evaporates quickly. Above the Leidenfrost point a continuous insulating vapour film separates the droplet from the metal, and since vapour conducts heat far more poorly than direct liquid-metal contact, the droplet actually loses heat more slowly and survives longer despite the higher surface temperature.
Is the Leidenfrost effect the same reason liquid nitrogen doesn't burn skin instantly?
Yes, the same mechanism, with the temperatures reversed: relative to liquid nitrogen at -196C, skin at body temperature is 'hot' enough that a thin insulating vapour film forms between the nitrogen and the skin on brief contact, delaying direct contact just as it delays a water droplet's contact with a hot pan.
Can you predict the exact temperature where the Leidenfrost effect starts?
There is no single universal number -- the Leidenfrost point depends on the liquid's boiling point and viscosity and on the surface's roughness, wettability and thermal properties, and it can be shifted substantially by surface texturing. For a clean, smooth pan and water it is typically cited in the 200C range, well above water's 100C boiling point, but engineered surfaces can push it much higher or lower.
Try it live
Everything above runs in your browser — open Leidenfrost Effect and change the parameters while it is running. Nothing is installed, nothing is uploaded, the whole model lives in one tab.
▶ Open Leidenfrost Effect simulation