Why is ice slippery?

(We are bracing ourselves for a severe ice storm today in this region so I am reproducing a blog post from back in 2006, before I moved to this site, because I thought that readers might find it interesting. I have edited it slightly, because tinkering with my writing is something I cannot help. This question formed the basis of an article by Robert Rosenberg in the December 2005 issue of Physics Today, p. 50-55.)

As you can see from this video, this man was reminded that ice is really slippery.

But why is ice so slippery?

How people respond to such a question can tell you a lot about their relationship to science in general. Some people will think that such a question, dealing as it does with everyday phenomena, should have a simple answer that has been known for a long time. Others will answer that that is just the way ice is. Intelligent design and other types of creationists might respond that ice was made slippery by their god for a purpose that we cannot comprehend. Yet others will say “who cares?” and wonder why we should bother with trying to answer such a question at all. But ask a group of scientifically-minded people this question and you will get a lively discussion going about the various possible explanations.

The basic reason why ice is slippery is not hard to understand. It is caused by the presence of a thin layer of water on the surface of ice. But this prompts the next question of what causes that layer of water to coexist with the ice. Why doesn’t it also freeze when the ambient temperature is below the freezing point of water? And seeking the explanation for this has formed the basis for a long series of research programs that have lasted for well over a century.

One popular explanation is that the pressure of the feet or skis or skates on the ice lowers the melting point temperature of ice. So when ice skaters glide on the ice, the high pressure caused by their weight on the thin blades (nearly 500 times atmospheric pressure) causes the melting point to drop below the normal freezing point of 0oC, and thus the surface of the ice, finding itself above that temperature, melts. Joly examined this idea in 1886 and calculated that such a high pressure would lower the melting point temperature to -3.5oC.

The problem with that explanation is that the optimum temperature for figure skating is -5.5oC while ice hockey players favor harder, faster ice at -9oC. In fact skating is possible at temperatures as low as -35oC. But this pressure explanation, despite its flaws, remained the dominant explanation for nearly a century. (However, it should be noted that the chief scientist of the 1910 Scott expedition to the Antarctic reported that at -46oC, the snow surface became sand-like, making skiing difficult.)

Another competing explanation (originating in 1939) was that friction caused by walking or sliding generated heat that caused the ice surface to melt. Experiments done between 1988 and 1997 to test this showed that skiing and skating over ice and snow did create enough frictional heating to create a water layer. But this still did not explain why ice is slippery even when you are simply standing on it, as that man in the video discovered.

Back as in 1850, the great scientist Michael Faraday did some experiments and found that even though cubes of ice seemed to have a surface water layer, when you held two cubes together, their common surface froze together to form a single ice block. This suggested to him the phenomenon of pre-melting, that says that the surface layer of ice is unstable because of the lack of molecules above it, and so becomes liquid at temperatures below the freezing point. But when two cubes are brought together, the common surface is no longer a surface layer and thus freezes, binding the two.

Faraday’s explanation was supported by others at the time but challenged by critics who said that holding the cubes together created pressure between them and it was this (and not pre-melting) that caused them to bind together. Faraday lost out in the end and his theory was pretty much forgotten until more sophisticated experiments were done beginning around 1950. In 1969, experiments confirmed that pre-melting of ice started occurring at -35oC. More sophisticated experiments have been done since and are still continuing that suggest that pre-melting is the correct explanation.

Is this something peculiar to ice? It turns out that pre-melting occurs in other materials as well. Lead, for example, melts at 327oC. But pre-melting causes the surface to become unstable and form a liquid film at about 307oC, so you would be able to skate on lead at that lower temperature, in the few seconds before you were burned to a crisp. (Kids, don’t try this at home!) Scientists have found similar liquid surface layers below the melting point for other metals, semiconductors, molecular solids and rare gases, suggesting that this is a common phenomenon.

Some interesting features of this history stand out because it shows how science and scientists work. Although Michael Faraday was a famous scientist, and his explanation ultimately turned out to be correct, it was not accepted at the time, despite some experimental support. It was only much later, with solid quantitative data coming in from a variety of sources, that resulted in the pre-melting hypothesis being accepted. Scientific prestige only assures you that your ideas will be heard and taken seriously, not that they will be accepted.

Second, ideas by themselves, however plausible or aesthetically appealing, need their own supporting data if they are to be accepted. It is not enough to simply discredit other ideas. Solid support for Faraday has only been coming in the last half-century.

Third, if you have a theory that explains one phenomenon, then it should also predict other phenomena. This is what was done by extending the unstable surface liquid layer idea to other materials and seeing if they showed the same thing. When they did, the idea gained credibility. In other words, pre-melting is not an ad hoc theory to explain just one result, that of ice. This is the essential hypothetic-deductive (‘if-then’) reasoning at work in science, which says that if something explains result A, then it should be possible to use that same mechanism or theory to predict what should happen in context B, and test to see if it does.

I emphasize these three points to contrast with what intelligent design creationist (IDC) advocates are trying to do with their idea that they are trying to get included as an alternative to the teaching of evolution. They are violating points two and three of the above criteria. They should not be surprised that IDC is being rejected by the scientific community.


  1. besomyka says

    My science teacher in 8th grade told us that if we don’t know the answer, it’s probably either friction, pressure, or cohesion. Still to this day when I come across a weird physical result, those are the first three things I go to.

  2. Crip Dyke, Right Reverend Feminist FuckToy of Death & Her Handmaiden says

    I had heard the pressure-based explanation. Glad to have a more thorough discussion here. Love to have old ideas dispelled so that I can be better educated tomorrow than I was today.

  3. says

    All I know is that doesn’t scare me at -20. At around 0, it terrifies me (broken wrist a couple of years ago, sprained armed a couple of months ago, and I’m sure I got a concussion once and should really have gone to see a doctor).

  4. Rob Grigjanis says

    Tabby Lavalamp @3:

    At around 0, it terrifies me

    Yeah, I never worried running outside in the winter until the temp neared zero. Wrong step, and I’d be arse-up midair before I could even think “WTF?”, or worry about the fact that I was thinking in internet acronyms.

    I didn’t know Faraday had nailed it so early, but it just reinforces his rep as one of the greats.

  5. robert79 says

    The first question that popped into my mind when I read the title of the post is “Is ice really always slippery?” I seem to remember hearing somewhere that on moons of Jupiter and/or Saturn, conditions are so extreme that ice seems more like rock, in terms of hardness, slipperiness, etc… (Although now I write this… How did they know that?)

    Figuring out under what conditions ice is slippery seems to be the first step in figuring out why it is slippery.

  6. Mano Singham says


    If the temperature of the ice drops below the pre-melting point of -35C, then it should not be slippery.

  7. sonofrojblake says

    Thank you so much for this. Prior to reading this, my answer would have been “Nobody knows, which is weird, because we should.”. I was aware of the weaknesses in the commonly given explanations, but not of the vindication of the Faraday explanation.

    at -46oC, the snow surface became sand-like, making skiing difficult

    Sure, it was the sand-like consistency of the snow that made skiing difficult when it was OH MY BOD HOW COLD???

    you would be able to skate on lead

    What a mental image.

    Next question: what other materials could you skate on? To the textbooks! Sadly, they quote only melting points, not pre-melting points, but let’s assume there’s a 30 degree window.
    Glycerol melts at 17 degrees. Could you skate at a relatively comfortable 12 degrees (not need to wrap up as warm?).
    Phosphorus melts at 44oC, but I’m guessing a skating rink made of phosphorus would have… other problems.
    A melting point of 63oC sounds like you could use it to skate on in a hot climate… but it’s for potassium, so again, unless your skating gear is really flame retardant…
    My favourite candidate is gallium (one of my favourite elements all round, as it goes), because it melts at just under 30oC. Assuming it has a reasonable premelt range, you could coat a smooth surface of asphalt with a layer of gallium, and skate on it! At comfortable room temperature! How great would that be?

Leave a Reply

Your email address will not be published. Required fields are marked *