Frozen
Ice cream gets worse in the freezer and temperature swings are why
A tub that was smooth when it was made turns coarse and crunchy after a few weeks not because it aged, but because it partly melted and refroze every time the freezer cycled.
By Tomas Bergqvist3 min read
Deterioration in a freezer has a specific mechanism
Nothing microbial is happening in a frozen dessert and very little chemical change is either, so the standard explanations for food going off don’t apply. What changes is the ice. Over weeks the population of ice crystals in a tub shifts from many small ones to fewer large ones, and since crystal size is the whole of the texture, the product coarsens without anything having spoiled.
The process has a name borrowed from materials science, Ostwald ripening, and it describes a general tendency for large crystals to grow at the expense of small ones. Small crystals have more surface relative to their volume and are therefore slightly less stable, so molecules leave them and join larger ones. Given time and any mobility at all, the average size climbs.
Temperature cycling is the accelerator
A freezer holding a perfectly constant temperature would let this happen slowly. No freezer does. Compressors cycle on and off, automatic defrost cycles deliberately warm the cabinet to clear frost from the coils, and the door opens. Each warming melts a fraction of the smallest crystals, each cooling refreezes that water onto whatever crystals survived, and the average size steps up a little every time.
This is why the same tub deteriorates faster in a frost-free freezer than in a manual one, and it is a genuine trade rather than a defect: the convenience of not defrosting the appliance is paid for in the texture of anything frozen inside it. It is also why ice cream in the door of a freezer, where the temperature swings most and warm air arrives every time someone reaches in, goes bad noticeably faster than the same tub at the back.
What slows it down
Stabilisers earn most of their reputation here rather than at the moment of churning. Gums and other hydrocolloids raise the viscosity of the unfrozen water phase, and water that cannot move cannot migrate from a small crystal to a large one. That is the main reason a commercial ice cream survives weeks of shop and domestic storage while a home-made one made without them can be unpleasant after ten days.
Fat plays a subtler part. A well-built dairy base has a partly coalesced fat network running through it, and that network physically separates regions of the mixture from one another, so water on one side of it cannot easily join ice on the other. This is one of the reasons a rich ice cream survives a month in a domestic freezer better than a lean one, and it has nothing to do with the richness being perceptible at the moment of eating.
A high proportion of solids helps for related reasons, since fat, milk solids and sugar all occupy space and obstruct movement. So does a colder and steadier storage temperature, because the fraction of water that is liquid falls steeply as the temperature drops, and there is simply less mobile water available to do the damage.
Practical storage, in order of usefulness
Keep it at the back and low down, where the temperature is most stable, and never in the door. Press a sheet of baking paper or film directly onto the surface before putting the lid on, which does two jobs: it stops the air gap above the ice cream from cycling moisture into frost, and it limits the sublimation that produces those crystalline shards on top.
Use a shallow, wide container rather than a deep one for the initial freeze, so the mixture passes through the danger zone quickly, then transfer if you want. And once a tub has been out on a table for half an hour, understand that it won’t be as good the second time. Softening only what you intend to serve is the single most effective habit here.
The visible symptoms and what each one means
Large gritty crystals through the body mean recrystallisation from temperature cycling. Fluffy white frost on the surface is a different phenomenon, sublimation, where ice has passed straight to vapour and redeposited on the coldest available surface, and it indicates air space and a loose lid rather than warmth. A shrunken, collapsed appearance suggests the foam structure has partly drained, which happens when a tub is repeatedly softened.
None of these is a safety question. Frozen dessert kept frozen remains safe well beyond the point where it becomes disappointing, and the dates on a tub are about quality rather than risk. What they are telling you is that the ice has been rearranging itself the entire time, which it has been.
Common questions
Why does home-made ice cream go icy so much faster?
It usually has no stabilisers, so the unfrozen water in it is free to move between crystals, and it was frozen more slowly to begin with in a domestic machine. Both push the same way. Eating it within a week or two is the realistic answer rather than trying to match commercial keeping quality.
Is it safe to refreeze ice cream that softened?
From a safety perspective, ice cream that softened but stayed cold and was refrozen promptly is generally fine. From a texture perspective it is a loss you cannot recover, because the melted fraction refreezes as large crystals. Fully thawed dairy left at room temperature is a different question and should not be refrozen.
Does an airtight container really help?
It helps against sublimation and freezer odours, both of which are real, and not at all against recrystallisation, which is driven by temperature rather than air. Pressing film onto the surface addresses the first problem better than the lid does, and nothing except a steadier freezer addresses the second.
Editor, The Last Bite
Tomas writes the explanatory pieces on baking basics, chocolate & sugar, frozen and thinks most subjects are more interesting once you know how they work.
