How a Microwave Oven Heats Food

Two minutes and your soup is boiling, yet a microwave never touches a flame. It works by making water molecules chase a field that changes its mind billions of times a second — and never quite catch it. Play with the five figures below to find out why.

A molecule that can never quite keep up

Water is a bent, lopsided molecule: the oxygen end is slightly negative, the two hydrogens slightly positive. That makes it a tiny compass needle, called a dipole, that wants to line up with an electric field. A microwave oven's magnetron blasts the food with a field that reverses direction 2.45 billion times a second — so the field's direction actually flips about 4.9 billion times a second (twice per cycle). Each water molecule tries to spin around and follow.

If it could follow perfectly, nothing would happen — a frictionless compass needle swinging in sync with a magnet doesn't heat up. Heat comes from the molecule failing to keep up: it lags behind the field, and that lag, fought against the drag of its jostling neighbours, is friction. Friction is heat.

Drag Frequency right. The glow behind the water molecule is the heat it's producing — watch it rise, peak, then fade again. (Everything here is slowed down enormously so you can see it: a real flip takes about 200 picoseconds.)

Notice the glow doesn't just keep climbing as you increase the frequency — it peaks once, in the middle, then fades. At low frequency the molecule swings fully and stays almost perfectly in phase with the field: lots of motion, but hardly any lag, so hardly any friction. At very high frequency the field flips faster than the molecule can respond at all, so it barely moves: lots of lag, but hardly any motion. The heat is biggest only where both a real lag and a real swing happen together — a broad hump, not a sharp resonance. That's the one surprising fact this whole page hangs on: a microwave oven isn't "tuned" to water's favorite frequency the way a radio is tuned to a station. It sits partway up the rising side of that hump. Liquid water's peak, measured in the lab, is up around 19 GHz — almost eight times higher than the 2.45 GHz an oven actually uses. Ovens use a lower frequency on purpose, for a reason the next two figures make concrete.

Not every ingredient heats the same way

That hump sits in a different place for every substance, because "how easily can this molecule reorient" depends on what it's made of. Water reorients fast. Frozen water — ice — is locked into a crystal lattice; a molecule can only turn if a defect happens to migrate past it, a process that's slow by comparison. Fats and oils are made of long, mostly non-polar hydrocarbon chains with only a weak dipole, so even when they can turn freely, the field barely grabs onto them.

Click each material. The bars show how fast it heats at 2.45 GHz, relative to water. Ice's true bar is thousands of times too short to draw — it's floored to a visible sliver, with the real number written next to it.

This is why a microwave can turn a mug of water scalding in ninety seconds but barely dent a block of ice, and why the sugary jam in the centre of a pastry can be burning hot while the dough around it is still lukewarm — sugar syrup and water heat fast, fat and dry starch heat slowly. It's also why "defrost" settings pulse the power on and off: the instant a little liquid water appears at the surface of frozen food, it starts absorbing far more strongly than the ice around it and can run away to boiling while the core is still frozen solid. Pulsing gives the heat time to conduct inward and even out between bursts.

Heating from the outside in

There's a second reason ovens don't run at water's peak frequency: the more strongly a material absorbs, the less deeply the field reaches. A wave that dumps most of its energy in the first few millimetres never gets to the centre of the food at all. So the field doesn't just fade over time as it heats water — it fades over distance as it travels in from the surface, roughly the way light fades as it goes deeper into murky water.

Drag Thickness up with Water selected. The field is strong (bright) at the surface and fades fast — the centre of anything more than a few centimetres thick gets very little of it directly.

That's the real mechanism behind the common complaint that microwaves "cook unevenly": it's not that the middle is somehow skipped, it's that the field simply doesn't reach it with much strength. The outside absorbs first and fastest; the centre has to wait for heat to conduct in from there, the slow way, exactly like a stove would do it. (The popular idea that microwaves cook "from the inside out" has it backwards — it's the opposite.) This is also why manuals tell you to cut food into smaller pieces or let it rest: smaller pieces mean nothing is ever far from a well-heated surface, and resting gives conduction time to catch up.

The inside of the oven is not evenly lit

The field doesn't just fade with depth into the food — it also isn't the same strength everywhere inside the metal box. Metal walls are conductors: their free electrons slosh back and forth with the field and re-radiate it, which is another way of saying the walls almost perfectly reflect microwaves instead of absorbing them (that's also why the field can't get through the door — the holes in the mesh are a couple of millimetres across, tiny compared with a 12 cm wavelength, so it acts like a solid wall to the field even though light sails straight through). Reflected waves bouncing back through the oven interfere with the ones still arriving, and the result is a standing wave: fixed places where the field reinforces itself into a strong hot spot, and other fixed places, exactly half a wavelength away, where it cancels to almost nothing.

At 2.45 GHz the wavelength works out to λ = c / f ≈ 12.2 cm, so hot spots repeat roughly every 6.1 cm — about a third the width of a typical oven cavity, which is exactly why a plate sitting in one spot can come out with a scalding patch next to a stone-cold one.

Drag the plate along the shelf and watch the reading swing between hot and cold every few centimetres. Then turn the turntable on: the running average climbs toward the same number no matter where it started.

That's the entire reason ovens have a turntable (or, in built-in commercial units, a rotating metal fan called a mode stirrer): no single fixed spot is reliably hot or cold, but a spot that keeps moving through the whole pattern gets, on average, the same exposure as any other. The maths behind that average isn't a coincidence — the average of a standing wave's intensity over a full sweep is exactly half its peak, always, regardless of where the sweep starts.

Prove it: measure the speed of light with a chocolate bar

The standing-wave pattern above isn't just a simulation — you can see it in a real oven with no equipment beyond a chocolate bar. Take the turntable out (so the food stays still), lay the bar flat, and run the oven for well under a minute: it melts in evenly spaced spots, exactly at the hot antinodes, and stays solid at the cold nodes in between. Measure the spacing between two melted spots, and you've measured half a wavelength.

Click Run and wait for the spots to melt. Drag "Your measurement" until the bracket's right edge lines up with the middle of the second melted spot — then read off the speed of light on the other side of the equals sign.

Every oven has its driving frequency printed on the back or the rating plate — always 2.45 GHz, an internationally reserved band chosen so ovens don't interfere with radio and phone traffic. Multiply that frequency by the wavelength you just measured (spacing × 2) and you get a speed — and because a microwave is nothing but an electromagnetic wave, exactly the same kind as light, just with a much longer wavelength, that speed is the speed of light. It's the same physics teachers use to demonstrate c = 299,792,458 m/s with nothing more than a chocolate bar and a kitchen appliance.