Most of us here would already know the simplified idea behind microwaving food: Microwave radiation hits the water molecules present in food, which excites them and causes 'em to vibrate rapidly in situ resulting in the heating up of food. Very straightforward.

But what if you decide to microwave a glass of water?

I can't seem to find any reliable literature that deals with this. A lot of highly unreliable sites spout nonsense from "The water becomes poisonous" to "The water becomes radioactive", so I've immediately dismissed those. However, some sources suggest that the "water will explode".

Now I was under the impression that the water would perhaps come to a boil a bit faster than it would by simply lighting a fire underneath it.

So after reading the last few sources it wasn't really difficult to believe that an explosion might occur. At first I simply extrapolated my theory of 'rapid boiling' to the theory of 'Very rapid boiling' and was content with that. It made sense to me since I presumed that, just as in the conventional method of boiling water, bubbles of water would be formed at the bottom of the glass albeit at a much faster rate, build up, rise, thus resulting in an explosion. But after some reflection, I noticed a problem with this 'extrapolation'.

Why do the bubbles even form in the first place when you conventionally heat water? Simply lighting a fire below a beaker of water does not ensure that the entire content o the beaker is uniformly heated at the same time. Obviously the layer of water molecules in direct contact with the bottom of the beaker will get heated rather quickly. They rise up, colder water descends, etc etc, you know the deal: convection. It's the sizable temperate difference between the bottom region and the rest of the water, that enables the water at the bottom to get vaporized at some point forming bubbles while the water immediately above it is still in the liquid state.

The problem I see that prevents me from extending this theory to microwaving water is that the mode of heating is different in the case of the microwave: Radiation

Now since radiation is faster at heating a given volume than convection is; for all practical purposes, I think it'd be safe to accept that the water in the beaker will be heated uniformly (The beaker's pretty normal sized...it's dimensions aren't of an order of magnitude greater than that of the speed of light/electromagnetic radiation, so it'll be alright to take the apparatus as being 'uniformly heated'. Had to say this before someone inevitably points out in the comments that all the water being heated to constant temperatures like this at the exact same time is impossible).

So if the beaker is heated uniformly, then I see no reason for bubbles to form...in other words, I don't see an explosion coming.

So what really happens if you microwave a glass of water?

Will it quickly and steadily vaporize? Will it explode? Or does something else happen? And why does it happen?

  • 2
    $\begingroup$ I do this all the time, However for the good reasons in the answers above I put a single ceramic Raschig ring in there whenever I use clear glass. I have one at home, stolen from the university 10 years ago for that very purpose. But then again, I am a nerd. $\endgroup$ – Stian Yttervik Dec 1 '16 at 10:51
  • 15
    $\begingroup$ Microwave radiation is too low in energy to cause molecular vibrations. It causes molecular rotations. $\endgroup$ – MaxW Dec 1 '16 at 11:21
  • 8
    $\begingroup$ The glass is quite irrelevant. The water gets heated by the microwave and the glass gets heated by the hot water. The rest is all nonsense. $\endgroup$ – Mast Dec 1 '16 at 12:16
  • 2
    $\begingroup$ @MaxW That's what I thought for a long time, too, but this Wikipedia article cites this paper as part of an argument that it's actually excitation of the broad vibrational absorption of the hydrogen bonding network in water. $\endgroup$ – hBy2Py Dec 1 '16 at 18:17
  • 3
    $\begingroup$ I'm not too experienced on this site, but wouldn't this question be more suitable for Physics? $\endgroup$ – Byte Commander Dec 1 '16 at 18:52

Heating water on a hot plate is safe, because the hottest point is at the bottom of the pot. A lot of relatively small bubbles appear there without much overheating of the water, because there is a lot of nucleation at the uneven phase boundary steel-water.

In a microwave, the hottest place is IN the water. The glass does not get heated by microwave (at least not much), and radiates off some heat to the surrounding.

Problem: In clean water, there are few good nucleation points to form bubbles, only some dust particles perhaps. So the water gets overheated rather strongly, and a first bubble that appears can grow a lot before it has cooled its surroundings down to 100°C. That one huge bubble can throw most of the water out of the glass. It boils over, violently.

Btw. a microwave does not heat its content uniformly. It forms a standing electromagnetic wave (that's not radiation, stricly speaking) in the oven, like a rope swung quickly between two people, or a guitar string. The wave pattern has knots at a distance of $0.5 c/f \approx 6\rm{cm}$ (with $f=2.45\rm{GHz}$ and speed of light $c$ which of course is a bit less in your chicken), where there is very little heating. That's why the microwave oven has the rotating plate, to generate some uniform heating. For pure water, that doesn't matter, because convection sets in anyway and distributes the heat.

  • $\begingroup$ I microwave glasses of water all the time; I'm just careful how long I microwave it, and I stir the water briefly every 30 seconds or so. By ensuring that I don't wind up with any superheated water, all that ends up happening is that I get hot water :) $\endgroup$ – Doktor J Dec 1 '16 at 19:39
  • $\begingroup$ @DoktorJ The most common way to keep the water from boiling over is putting a glas rod or similar in the cup. Opening the oven every few seconds seems a bit cumbersome to me. $\endgroup$ – Karl Dec 1 '16 at 20:47
  • 2
    $\begingroup$ ? A said, a glass of water in a microwave oven will overheat without much visible change (gas bubbles), and then suddenly boil over. With some bad luck, your impatient hand is just in the oven at that time. $\endgroup$ – Karl Dec 1 '16 at 21:20

The mode of heating of a water glass in a microwave and on a stove is actually very similar. While it's true that microwave radiation penetrates somewhat into the body of water, the penetration depth is rather small.

The main problem is that on a stove, you get uniform heating from the bottom, with temperature usually far higher than the boiling point of water. "Boiling" occurs when the water at the bottom of the container is hot enough to turn to vapour regardless of the surrounding pressure - it's actually a bit hotter than the boiling point. At the same time, the rest of the water is significantly colder, which is why you see the bubbles long before we tend to consider the water at "boiling temperature".

In a microwave, the water is heated in a bit different fashion. In the simplest model, you're heating it from the sides - similar to heating a glass of water in a normal oven. This already makes a huge difference - since you're heating the water from the sides, the bulk of water isn't heated through convection. The main effect is that the heating appears a lot more uniform, and the heat is distributed mostly through diffusion - when the water gets close to boiling point, a large bulk of the water is close to the boiling point, and convection can't carry that heat anywhere.

And this point is the tricky part. The water isn't capable of boiling, because to vapourize, it has to overcome surface tension - and the heat near the boiling point isn't powerful enough for that yet. Boiling is easiest on contact with e.g. sides of the glass, but that's not where the hottest water is, unlike with a stove. And now, you disrupt the water - bump into it, put a coffee been into it, a spoon... you form a large surface area in water that's already at boiling point, and remove the only thing that's preventing it from boiling. Across the newly formed surfaces, water starts vapourizing, which increases the pressure, which disrupts the surfaces further, causing further vapourization... and you get a splash of boiling-hot water.

Now, the actual way stuff is heated in a microwave oven is even more complex than this - but that's not really necessary to explain why superheating is a lot bigger problem in an oven than on a stove, so I'll leave it at that :)

In a way, this is similar to the coke & menthos effect (though instead of vapourization, the carbon dioxide in that experiment comes out of a water solution). You have a liquid with dissolved carbon dioxide at an equilibrium - in fact, the liquid is already supersaturated when you open the bottle due to the drop in pressure. Drop the menthos inside, and you create a great nucleation source for the dissolved carbon dioxide - the surface of the candy is rather rough, and it gets rougher as it dissolves, so the surface tension is much smaller than it would usually be, causing the carbon dioxide to quickly come out of the solution, causing foaming and expansion -> boom.

So, what would happen if you could actually heat the water uniformly? The water would start boiling along points of contact with the glass, so you'd see bubbles of boiled water carrying the heat away quickly. The "sweetspot" for "exploding water" is exactly having pockets of boiling water apart from any nucleation sources, and then introducing a nucleation source. This is pretty hard to do even in a microwave oven, but it is possible - and the same hazard applies any time you heat a container in a way that doesn't leave the most heat in places in direct contact with the container :)

  • 1
    $\begingroup$ And how deep does the microwave field penetrate into the water? $\endgroup$ – Karl Dec 1 '16 at 21:03
  • $\begingroup$ @Karl If I remember correctly, it's one or two centimeters for 2/3rds absorption in cold water. As the water gets hotter, the penetration is easier, so close to the boiling point, most of the radiation actually makes it through a typical glass of water (to be reflected on the other side of the microwave). I couldn't find a good model that takes everything into account at once - it really is quite complicated. But this should be close enough for a typical microwave. This is also why unfreezing is so slow (and pointless) in a microwave - the penetration depth in ice is about ten meters :) $\endgroup$ – Luaan Dec 2 '16 at 7:40
  • $\begingroup$ "you get a splash of boiling-hot water" Most likely, a splash of above boiling-hot water. $\endgroup$ – David Richerby Dec 2 '16 at 10:01
  • 1
    $\begingroup$ @DavidRicherby I don't think so. When the water starts boiling, it is suddenly full of nucleation points where water evaporates, and the temperature almost instantly drops down to 100°C. $\endgroup$ – Karl Dec 2 '16 at 20:04
  • $\begingroup$ @Luaan There is actually no real "radiation" in a microwave oven. A standing wave pattern is formed, which is distorted by the glass of water. The penetation depth will depend a lot on the salt/carbonate content. Two centimetres penetation depth in a circular, 5cm dia glass will make for quite even heat distibution. Afaik, penetation depth is largely dependent on electrical conductivity (inversely), while heat transfer depends on the dielectric properties. So there's no direct connection, while you are right, ice does not heat in the microwave. $\endgroup$ – Karl Dec 2 '16 at 20:18

A microwaved glass of water will 'bump' if the glassware is clean and the microwave heating is uniform. The water has some tensile strength, so a bubble will not form at the exact boiling temperature without some nucleus (low surface tension due to a gas void in a boiling stone, for instance), so the liquid can become superheated. On reaching in for the hot liquid, it may be the case that moving the liquid in the container causes an eruption as the superheated liquid boils.

It is true, too, that even without a nucleus to form a bubble, the air/hot-water surface will evaporate water vapor rapidly, but that has the effect of cooling the water surface and if the water is not in motion, a cool-water layer on the surface will form even as the bulk of the water becomes superheated.

Superheating is dangerous, of course; if your hand is holding the cup when it bumps, a burn is likely. A spoon lowered into superheated water will usually make it erupt. Eventually, a cosmic ray will make it erupt.

  • 1
    $\begingroup$ The water has no here relevant "tensile strenght", it's the surface tension of the bubble-to-be that keeps it from growing, until the water is hot enough to overcome the surface tension. The eruption can and w ill occur spontaneously. $\endgroup$ – Karl Dec 1 '16 at 10:35
  • 2
    $\begingroup$ Tensile strength is a way to express the difficulty of forming a small bubble (when surface tension is proportional to 1/r, small bubbles that form thermally must grow past a critical size or they collapse). A small bubble in liquid is inhibited just as a small crack in a solid is, by "tensile strength". $\endgroup$ – Whit3rd Dec 1 '16 at 10:48
  • $\begingroup$ Chemists should agree calling surface tension for tensile strenght of liquid under these circumstances. I understood what you meant and liked the expression. I support you Whit3rd. Chemists make imprecision a science! Your answer is totally within +/-1 logarithm of the probable answer, which is good enough :-P $\endgroup$ – Stian Yttervik Dec 1 '16 at 10:54
  • 1
    $\begingroup$ The phenomenon has some similarities with en.wikipedia.org/wiki/Ultimate_tensile_strength, but is just something different. Why do you use a bad analogon when you know better (critical size of bubble) and there is a perfect, direct explanation? I'm a chemist, and find that highly unscientific, @StianYttervik ! $\endgroup$ – Karl Dec 1 '16 at 11:23
  • $\begingroup$ Because, the philosophical difference between physics, materials science and chemistry is the models we use to explain different phenomena. While it is not rigorous to explain surface tension as resistance like tensile strength, it is exactly something a chemist could use as explanation for an observed phenomenon. And then we are indeed ultimately in agreement that the best model overall would be surface tension. No argument against it, and I would have been cross to see it used formally. But here it is used informally, and in my view appropriately. For this phenomenon - the model works. $\endgroup$ – Stian Yttervik Dec 1 '16 at 12:04

Water is a polar molecule as you know. This means it has positive and negative ends. The microwave is a electromagnetic wave. These waves are the right frequency to cause the molecules to get vibrating and cause tension. As you know, temperature is defined as the kinetic energy of a particle. This means that the water would heat up. But the directional vibration of the water depends on where it is located in the microwave. This it is almost random. The random movements can collide with impurities and jagged features of the container or water to nucleate the bubbles. This means that the bubble would float out of solution. This results in boiling. If the water is pure and the container is smooth, then the water will not have a nucleation point and will not bubble. What will happen is that some of the water molecules will just steam out but the rest of the molecules have no way to nucleate until something goes in, like a spoon, or sugar.


Always stay clear of the center of the plate, rather put your cup or glass on a side so it travels the most while heating.

Microwaves have "fixed waves" and should there be a region of water always under a specific wave, the water will explode just like a droplet of water in OIL. It wants to evaporate, with no air around to do so, so BOOM it needs to go somewhere, and it does.

On the other hand, make it go around the platter, avoiding dead center, the waves will just heat up the water little by little, eventually boiling and creating movement and bubbles, hence space for the vapor to go.

A try to avoid the center of the revolving plate whenever I put food in the microwave too, even leaving a space in food if the plate is bigger than half the diameter of the platter.

I read somewhere "make a hole in the middle of the food it'll heat more evenly". Well, ONLY if that hole is aligned with the middle of the rotating platter, then yes.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.