Freezer burn
A Socratic walk-through of freezer burn — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why does food dry out inside a sealed freezer where nothing can evaporate away?
A steak goes into a sealed bag, into a freezer, in January. In September it comes out with pale grey patches, a leathery surface, and — this is the part worth noticing — a lining of frost on the inside of the bag that was not there when you sealed it.
Two things in that description sit badly together. The meat is drier than it went in, and the bag is wetter. Nothing crossed the seal. So where did the water go, and what moved it, in a box that is by design the least energetic place in the house?
Reasoning it through
REASONING #Start with the assumption buried in the question: that nothing can evaporate at those temperatures. It is false, and that is the first correction.
Ice has a real vapour pressure. At 0 °C it is about 611 pascals; at -20 °C, roughly 103. That is small — about a thousandth of atmospheric pressure — but it is not zero, and molecules leave the surface of ice directly for the vapour phase without ever being liquid. The word for that is sublimation, and it is the same process that shrinks an ice cube left uncovered in a tray for a month.
But now the second half of the puzzle bites harder. The bag is sealed. No water leaves the package at all — mass is conserved, and if you weighed the whole thing in January and September you would get the same answer. So drying cannot mean loss. It must mean relocation.
Where to? Vapour goes wherever the vapour pressure is lowest, and vapour pressure over ice depends steeply on temperature. So the destination is simply whatever is coldest inside that bag. The meat is dense and slow to change temperature. The plastic film has almost no thermal mass and tracks the freezer air within minutes. The frost you find on the inside of the bag is the steak's own water, moved a centimetre.
What supplies the temperature difference? The freezer itself, cycling. A compressor runs and stops; a frost-free model runs a defrost heater periodically, deliberately warming the cabinet. Every one of those episodes warms the meat's surface a little, and the film cools back down faster than the meat does. The gradient does not need to be large. Work through the Clausius-Clapeyron relation with the latent heat of sublimation and you find that vapour pressure over ice changes by roughly a tenth for every kelvin at freezer temperatures — so a single degree of difference across a centimetre of headspace is a ten per cent driving force, applied a few times a day for eight months.
Then the last question, and it is the one that makes this an accumulation rather than an oscillation. Why does the water not come back on the next swing?
Some does. But it does not return to where it came from. Ice deposits preferentially onto large crystals and open surfaces — a small, sharply curved crystal has a slightly higher vapour pressure than a big flat one, so big ones grow at the expense of small ones. The cavity left inside the meat's tissue is not refilled; the frost on the bag simply gets thicker. Each cycle ratchets, and none of them undo.
That explains the drying. The colour is a second, downstream story: the dehydrated surface layer is left as a porous, air-filled sponge with a large internal surface area, and whatever oxygen was sealed in with it now reaches all of that area. The grey and white patches are that porosity plus oxidised fat and pigment. Freezer burn is dehydration first and oxidation second — and it is a quality problem, not a safety one.
The analogy
THE ANALOGY #Think of a cold window in a warm room. Nobody poured water on the glass; the water was already in the air, and the pane merely happened to be the coldest thing available, so that is where it collected.
The window gives its water back when the room cools and the pane warms, whereas the frost on a freezer bag is never returned to the meat — and that one difference is the whole mechanism, because it is what turns a nightly cycle into eight months of one-way transfer.
Clarifying the model
THE MODEL #Three refinements tie the steps together.
First, "sealed" does the wrong work in the original question. A seal stops exchange with the room; it does nothing about transport within the package, and freezer burn is entirely an internal transport problem. Vacuum packing helps enormously not because it seals better but because it removes the headspace the vapour has to cross and lays the film directly against the food, leaving nowhere colder to go.
Second, the enemy is temperature variation, not temperature. A chest freezer holding a steady -25 °C damages food far less than a frost-free upright that drifts to -12 °C and back several times a day, even though both are safely frozen throughout. Frost-free convenience is bought partly in this currency — the defrost cycle that keeps the walls clear is also the pump that drives water out of your food.
Third, this is a different kind of drying from the everyday sort, and the difference is worth holding. Washing dries because air moves past a liquid film and carries vapour away down a humidity gradient. Nothing moves here, nothing is liquid, and nothing is carried away — the water travels a centimetre and stops. It is also unrelated to how salting or drying preserve food, where the point is to lower water activity below what microbes tolerate; freezer burn is not microbial at all, and a freezer-burnt steak is perfectly safe.
A picture of it
THE PICTURE #How to readThe curve is the pressure of water vapour sitting in equilibrium over ice, plotted against temperature. Read the far left first: at freezer temperatures the value is small but plainly above zero, which is the answer to "nothing can evaporate". Then read the steepness — between -20 °C and -15 °C the pressure rises by around half. That slope is the engine: two surfaces a single degree apart inside one sealed bag are not in balance, and the warmer one loses water to the colder one every time the compressor cycles.
What became clearer
WHAT CLEARED #Nothing evaporates away in a sealed freezer, and nothing needs to. Ice sublimes at any temperature, vapour pressure over it varies steeply with temperature, and a freezer's own cycling keeps creating small differences between the food and the packaging next to it. Water therefore walks out of the meat and onto the bag, a little each cycle, and never walks back — so the damage is an accumulation of tiny reversible steps that happen to be arranged so they cannot reverse.
Where to go next
ONWARD #- Why fast freezing produces small ice crystals and why that matters months later.
- How glazing fish in a layer of ice gives the vapour something else to give up.
Key terms
TERMS #| Term | What it means |
|---|---|
| Sublimation | the direct transition from solid to vapour with no liquid stage, which ice undergoes at any temperature. |
| Vapour pressure | the pressure of vapour in equilibrium with a solid or liquid surface, rising steeply with temperature. |
| Water activity | the availability of water to microbes, the quantity that salting and drying reduce; distinct from freezer burn, which is physical relocation rather than microbial control. |
Every term the collection defines is gathered in the glossary.