On a winter sidewalk, the explanation arrives fast and loud: your weight, through your shoe, melted a little water, and that water made you slide. Generations of textbooks have told some version of this story, and it feels airtight. It also happens to be incomplete in ways that matter a great deal. Ice is slippery at temperatures far too low for a pressure calculation to rescue, it is slippery before your shoe has moved a single millimeter, and it stays slippery in physics experiments where no meaningful frictional heat is being produced at all.
The genuinely strange part is that solid surfaces almost never behave this way. Glass, steel, wood and stone all get grippier, not slicker, when you put a load on them. Ice is one of the only common solids whose topmost molecules turn liquid-like under ordinary contact. Scientists broadly agree that a thin, water-like layer on the surface is what lets skates glide and pedestrians fall. What they cannot agree on, even today, is why that layer exists. The disagreement is not a dusty historical footnote: two major papers on the question appeared in 2025 and 2026, and they reached different verdicts.
The Pressure Story That Ruled for a Century
The classical account began in the 1850s, when the British engineer James Thomson worked out that pressure lowers the melting point of ice. His younger brother William, the physicist better known as Lord Kelvin, verified the relationship experimentally. In 1886, the Irish physicist John Joly connected the idea to skating. A skate blade touches the ice over a very small area, and Joly calculated that a skater presses on roughly 466 atmospheres. That is enough to push the local melting point down to about minus 3.5 degrees Celsius. A film of water forms, the blade glides, the physics exam gets a tidy answer, and the theory holds that position for nearly a hundred years.
The collapse came from people who actually did the arithmetic. In the 1930s, Frank P. Bowden and T. P. Hughes, working at the University of Cambridge, calculated that an average skier exerts far too little pressure to shift the melting point by any useful amount; to melt ice through body weight alone, a person would need to weigh thousands of kilograms. The geophysicist Samuel Colbeck pressed the argument further in the late 1980s and 1990s, pointing out that skating works at temperatures as low as minus 30 degrees Celsius, where no realistic contact pressure can melt anything. History quietly agrees with him: Robert Falcon Scott described skiing easily at minus 30 in his account of the South Pole expedition, while his chief scientist Edward Wilson found the snow had turned sandlike, and finally useless, at minus 46.
Frictional Heat: A Theory Born in a Glacier Cave
Having demolished the pressure explanation, Bowden and Hughes proposed the replacement that most of us absorbed in school: sliding generates heat, the heat melts a thin film of ice, and the film lubricates the next moment of sliding. To test it, the two researchers hauled their equipment into an artificial ice cave dug above the Jungfraujoch research station in the Swiss Alps, at an altitude of 3,346 meters. They measured friction between ice and different materials and noticed something suggestive: wood slid more easily than metal, and materials that conduct heat well, like brass, produced more friction than poor conductors like ebonite. A material that steals heat quickly leaves less of it available to melt the surface, which fits the frictional heating story neatly.
But the story has a hole, and Daniel Bonn of the University of Amsterdam put it bluntly when he spoke to Quanta Magazine: under that theory, you only melt the ice behind you, not the ice you are actually skating on. A skater glides with the leading edge of the blade touching fresh ice that has not felt any rubbing yet. Friction, by definition, requires motion, yet the first step onto a frozen pond is already treacherous. Bonn and his colleagues built a microscopic ice rink in the laboratory, rotating a small metal piece against ice at different speeds and measuring the forces involved. If frictional heating were the whole explanation, faster sliding should mean more heat and less friction. In their measurements, slipperiness barely depended on speed at all.
Faraday’s Film: The Ice That Is Wet Before You Touch It
A third explanation predates them all. In the mid-1800s, Michael Faraday noticed that two ice cubes pressed together will freeze into one, and that even a warm hand can stick to a freezing surface. He proposed that exposed ice carries a thin, pre-melted layer on its outside, and that the layer refreezes when two surfaces press it between them, which would explain both the stuck cubes and the stuck hand. Faraday could not say why the surface should melt at all. That missing mechanism took roughly a century to appear: scientists such as Charles Gurney and Woldemar Weyl argued that water molecules at the surface have fewer neighbors to bond with, so they wiggle with more freedom than molecules locked inside the crystal lattice.
Modern computer simulations have given that old idea surprising support. Luis MacDowell, a physicist at the Complutense University of Madrid, ran simulations in which individual atoms could be watched directly, and the virtual ice always came coated with a liquid-like layer just a few molecules deep. When he simulated a heavy object pressing down, the layer thickened, echoing the pressure theory, but only near ordinary melting conditions. In 2026 he pulled the scattered record together in a perspective review for the Journal of Chemical Physics. His argument: decades of apparently contradictory results about the quasi-liquid layer can, with the right theoretical framework, be reconciled into one consistent picture of the ice surface.
The Dipole Hypothesis: Contact Without Heat or Pressure
In August 2025, a team at Saarland University published a result in Physical Review Letters that challenged the whole framing. Martin Müser, working with Achraf Atila and Sergey Sukhomlinov, simulated contact between ice and another material and found that the deciding factor is neither pressure nor friction but molecular dipoles. Every water molecule is a small magnet of charge, negative near the oxygen, positive near the hydrogens. The polymers and other substances in a shoe sole have their own dipoles. When the two materials meet, the orientation of those dipoles interferes, and in three dimensions the interactions cannot all settle into a comfortable arrangement. Physicists describe this situation as frustration: competing pulls prevent any fully ordered configuration, and the orderly lattice of the ice surface collapses into disorder, then into a fluid.
The most eye-catching claim in the paper is what happens in extreme cold. Until now, the standard assumption held that skiing below about minus 40 degrees Celsius fails because no lubricating film can form at all. According to Müser’s team, dipole interactions persist even near absolute zero, and a liquid-like film still appears between ice and a ski. The catch is practical rather than physical: at such temperatures the film is more viscous than honey, so it would hardly feel like water, and sliding on it would be nearly impossible. The film exists everywhere; only its usefulness changes with temperature.
The 2026 Verdict That Splits the Difference
Just as the dipole hypothesis seemed to close the case, a second group complicated it again. A paper by Sigbjørn Løland Bore, B. N. J. Persson and Henrik Andersen Sveinsson, accepted by Physical Review Letters in October 2026, attacked the question from a different angle: true first-principles simulations of nanoscale friction between ice and glass, combined with a model of how frictional heating spreads. The two approaches had to be combined deliberately, because the nanoscale simulations on their own overestimated friction and completely missed the way friction falls as speed increases.
Their answer honors both camps at once. Premelting, the Faraday effect, produces the watery film even before any sliding or heating takes place, exactly as the Saarland team argues. But whether that film actually lubricates, rather than just sitting there as a viscous mush, depends on frictional heating, exactly as Bowden and Hughes proposed back in 1939. The heat warms the contact patch toward the melting point, thickening the pre-existing film by a factor of ten while cutting its viscosity by a factor of one hundred, and friction then drops with speed, matching what laboratory experiments actually measure. One subtlety makes the result almost philosophical: the researchers conclude that nothing ever truly melts. The ice does not cross a phase boundary under your skates. A film simply becomes more water-like as it warms.
Where does that leave the question? Quanta, reviewing the field in late 2025, described a consensus that feels nearer but has yet to arrive, and the newest paper itself lists snow, other materials and engineered coatings as open problems. The slippery question is still, technically, a work in progress.
Common Misconceptions About Slippery Ice
Because the textbook story is half right and half wrong, the topic attracts confident mistakes. Six of them are worth retiring.
- Your body weight melts the ice under your foot. The pressure calculation fails at most winter temperatures. An ordinary person walking or skiing produces nowhere near the pressure needed to lower the melting point meaningfully, especially far below zero.
- Frictional heat explains everything on its own. Ice is slippery the instant you step onto it, before any sliding has happened, and laboratory measurements show little speed dependence where heating arguments predict a strong one. Heat matters, but something has to exist for the heat to act on.
- Ice is covered in actual liquid water at all times. The quasi-liquid layer is only a few molecules thick and is not bulk water. Calling it liquid is shorthand for disordered, mobile surface molecules that behave liquid-like on small timescales.
- Ice stops being slippery in deep frost. A film survives to near absolute zero according to simulations, though it becomes extremely viscous. Grippiness at deep cold comes from this viscosity and from brittle, hard ice, not from an absence of film.
- Physicists solved this question long ago. The decisive papers were published in 2025 and 2026, they disagree with each other in emphasis, and the researchers themselves list open questions. This is a live research area, not settled school material.
- Colder always means safer walking. Experiments from Amsterdam found minimum friction at about minus 7 degrees Celsius, and rink operators choose similar temperatures on purpose. Slipperiness does not fall monotonically as it gets colder, and ice that feels soft near the melting point brings its own problems rather than reliable grip.
What Actually Happens Under a Skate Blade
The most complete experimental picture still comes from the 2018 work of the Bonn brothers, Daniel at the University of Amsterdam and Mischa at the Max Planck Institute for Polymer Research in Mainz. Their team slid a steel ball across ice at temperatures from 0 down to minus 100 degrees Celsius while measuring friction and probing the surface water molecules spectroscopically, then compared the results with molecular dynamics simulations. Two populations of surface molecules turned up. Some are anchored to the ice by three hydrogen bonds, and some dangle loosely with only two, rolling across the surface like tiny spheres powered by thermal vibration.
The fraction of mobile molecules rises with temperature, and the measured friction falls with it in perfect step. The sweet spot turned out to be about minus 7 degrees Celsius, which is, tellingly, very close to the temperature speed skating rinks maintain by tradition and experience. Above that range, the ice grows soft enough that the slider digs into it, which adds resistance even as the molecules get more mobile. The lesson is that slipperiness lives in the top few molecular layers, in their bonds and their restlessness, not in a puddle. For anyone choosing winter footwear, the practical summary is modest but real: expect the worst grip near the freezing point, better grip in real cold, and no reliable comfort anywhere without tread.
Frequently Asked Questions
Does the pressure from a person’s weight melt the ice underfoot?
No. Calculations by Bowden and Hughes in the 1930s showed an average skier exerts far too little pressure to shift the melting point by any useful amount, and melting ice through body weight alone would require a person to weigh thousands of kilograms. Skating also works at temperatures as low as minus 30 degrees Celsius, where no realistic contact pressure can melt anything.
What is the quasi-liquid layer on ice?
It is a thin, water-like coating on the ice surface, only a few molecules deep, that scientists broadly agree is what lets skates glide and pedestrians fall. Michael Faraday proposed a pre-melted surface layer in the mid-1800s, and modern simulations by Luis MacDowell found virtual ice always coated with a liquid-like layer, which thickened under a heavy object but only near ordinary melting conditions.
Does frictional heating fully explain why ice is slippery?
No, because ice is slippery the instant you step onto it, before any sliding has generated heat, and laboratory measurements showed slipperiness barely depended on speed at all. The 2026 paper gives heat a narrower role: it warms the contact patch, thickening the pre-existing film by a factor of ten and cutting its viscosity by a factor of one hundred, which is what makes the film actually lubricate.
What is the dipole hypothesis?
In August 2025, a Saarland University team found that the deciding factor is neither pressure nor friction but molecular dipoles: the interfering orientations of dipoles in water molecules and in shoe-sole materials create frustration, so the orderly ice lattice collapses into disorder and then into a fluid. Their simulations show dipole interactions persist even near absolute zero, though at deep cold the resulting film is more viscous than honey.
Is the debate about why ice is slippery settled?
No. The decisive papers were published in 2025 and 2026 and disagree with each other in emphasis, and the researchers themselves list open questions such as snow, other materials and engineered coatings. Quanta, reviewing the field in late 2025, described a consensus that feels nearer but has yet to arrive.
Summary:
- Ice feels slippery because its surface carries a thin, water-like layer, but the reason that layer forms is still argued about after two centuries.
- Pressure melting fails the arithmetic: a skier would need to weigh thousands of kilograms to melt ice by body weight alone at low temperatures.
- Frictional heating, proposed at a Swiss glacier cave in 1939, explains much laboratory data but cannot explain why ice is slippery the instant you touch it.
- Faraday’s pre-melted film and modern simulations support a persistent surface layer only a few molecules thick, now attributed by a 2025 study to frustrated dipole interactions.
- A 2026 study reconciles the camps: premelting creates the film while frictional heating makes it slippery, without any true melting.
- Experiments place minimum friction near minus 7 degrees Celsius, close to the temperatures speed skating rinks choose deliberately.
External References:
- Quanta Magazine: Why Is Ice Slippery? A New Hypothesis Slides Into the Chat
- Saarland University: Why we slip on ice
- Atila, Sukhomlinov and Müser, Cold Self-Lubrication of Sliding Ice, Physical Review Letters
- Bore, Persson and Sveinsson, Why ice is so slippery, Physical Review Letters (accepted)
- University of Amsterdam: The slipperiness of ice explained
- MacDowell, The Key Physics of Ice Premelting, Journal of Chemical Physics
- Physics Today: Why Is Ice Slippery?
