Microwaves do not work by resonating with water

The explanation is everywhere, including from people who ought to know: a microwave oven runs at 2.45 GHz because that is the resonant frequency of water, so the microwaves make water molecules resonate and the friction heats your food.

Almost every part of that is wrong, and the real mechanism is more interesting because it explains things the myth cannot — like why ice barely heats, why there are cold spots, and why the frequency was chosen to be bad at heating water rather than good at it.

The physics that kills the story

Liquid water has no sharp resonance at 2.45 GHz. Isolated water molecules do have rotational transitions, but they sit at much higher frequencies, and in liquid water they are smeared out entirely — each molecule is hydrogen-bonded to its neighbours and cannot rotate freely. What liquid water has instead is a broad absorption continuum, and the peak of it at room temperature is around 20 GHz. Not 2.45. Nearly an order of magnitude away.

And if it were a resonance, your oven would be useless. This is the part that settles it. Resonant absorption is strong absorption — energy would be deposited in the first millimetre or two of whatever it hit. You would get a scorched surface and a frozen middle, on everything, every time.

The frequency was chosen for the opposite reason.

What is actually happening

The mechanism is dielectric heating, and the useful word is not resonance but friction.

A water molecule is a dipole: charge is distributed unevenly, so one end is slightly positive and the other slightly negative. Put it in an electric field and it experiences a torque twisting it into alignment. Now oscillate the field a few billion times a second, and the molecule is repeatedly dragged one way and then the other.

It cannot keep up. Surrounded by neighbours it is bonded to, each molecule lags behind the field it is trying to follow, and that lag is where the energy goes — it is dissipated against the surrounding liquid as heat. The bulk of the food heats because the molecules inside it are being jostled, not because any one of them is ringing like a bell.

So why 2.45 GHz?

Two reasons, and neither is that water likes it.

Penetration depth. Because 2.45 GHz sits well below water’s absorption peak, the absorption is moderate, and the waves get a few centimetres into food before being substantially absorbed. That is roughly the size of the things people heat. Pick a frequency where water absorbs strongly and you have built a surface griller.

Regulatory allocation. 2.45 GHz falls in a band set aside internationally for industrial, scientific and medical use — the same band Wi-Fi and Bluetooth occupy, which is not a coincidence but a shared consequence of it being the spectrum you are allowed to spray energy into.

The frequency is a compromise between heating uniformly and being permitted to transmit, and it is deliberately away from the peak.

What the real mechanism explains

This is where the two stories separate, because the dipole account predicts a list of familiar oven behaviours and the resonance account predicts none of them.

Ice barely heats. In ice, water molecules are locked into a crystal lattice and cannot rotate to follow the field. Absorption drops by orders of magnitude — which is why defrosting is done in pulses. The oven heats the few liquid spots, then waits while conduction melts their surroundings, then heats again. Run a frozen block at full power and you get boiling patches beside solid ice.

It is not only water. Fats and sugars are also polar and heat perfectly well, which is why a fatty sauce can end up much hotter than the water around it, and why a jam doughnut is a well-known hazard.

Salt matters, by a second mechanism. Dissolved ions are charged and get pushed bodily back and forth by the field, dissipating energy through ionic conduction. Salty food heats faster, and mostly nearer the surface.

Cold spots are standing waves. The oven is a metal box, and the waves reflect off its walls and interfere, producing a fixed pattern of high and low field strength — nodes with almost no energy at all. The turntable exists to drag food through the pattern. Some ovens use a rotating metal stirrer near the emitter to move the pattern instead.

Metal arcs. Conductors have mobile electrons, and a sharp edge or a thin point concentrates the field enough to ionise the air beside it. A crumpled foil edge or a fork tine is a small antenna with a very sharp tip.

Water can superheat. Smooth container, pure water, no nucleation sites, and the liquid can pass its boiling point without boiling — then flash violently when disturbed. Uniform volumetric heating with no hot surface makes this easier than it is on a hob.

Why the myth persists

Because “resonant frequency” sounds like a mechanism, is memorable, and gives the questioner something that feels like an answer. “Dipoles lag behind an oscillating field and dissipate the difference as heat” is harder to say and much harder to retain.

But it is worth the extra sentence, because it turns a piece of trivia into something predictive. Once you have it, you can work out for yourself why the middle of the lasagne is cold, why the plate got hot when it shouldn’t have, and why the defrost setting takes so long.