A chocolate shell on a frozen dessert is a contraction problem

Two materials meeting at a temperature neither was designed for

Dipping a frozen bar into a warm coating brings together two things with incompatible habits. The dessert is at perhaps minus eighteen degrees and full of ice. The coating is a molten fat at somewhere above thirty. What happens in the two or three seconds after they touch decides whether the shell is thin and even or thick and lumpy, whether it stays put, and whether it cracks off in the box before anybody opens it.

The coating fat has to satisfy a narrow specification. It must be solid and reasonably brittle at freezer temperature, so the shell snaps cleanly. It must be fully liquid in the mouth, so it doesn’t read as waxy. And it must set almost instantly on contact. That combination is why coatings are formulated rather than simply melted from a bar, and coconut oil is the usual domestic answer because its melting behaviour happens to fall in roughly the right place.

The frozen surface is doing the cooling, and doing it fast

The coating does not set because the room is cold. It sets because the thing it has been poured over is a large, cold heat sink with a great deal of ice in it, and heat flows out of the thin liquid film into that mass almost immediately. The film is only a fraction of a millimetre thick, so it has very little heat to give up, and the substrate barely notices receiving it.

This is what makes shell coatings practical at all, and it also sets the thickness. A colder bar pulls heat out faster, so the film sets sooner and stays thin; a bar that has warmed slightly lets the coating linger as a liquid, and it runs and pools before it sets. Dip temperature works the other way — a hotter coating is thinner and flows better, but it also melts more of the surface it lands on, which shows up as a soft, milky boundary layer between shell and dessert.

Contraction is why shells crack, and why some cracking is wanted

Nearly everything contracts as it cools, and the two materials here do it by different amounts and on different schedules. Fat contracts appreciably as it crystallises and continues to shrink as it goes on cooling towards freezer temperature. The dessert beneath it, already frozen, is barely changing at all. The shell is therefore being pulled tighter around something that isn’t shrinking with it, which puts it under tension.

A thin shell accommodates that tension. A thick one relieves it by cracking, which is why heavily coated bars often show fine fractures before anybody has touched them. The same tension is useful at the other end: a shell under slight tension shatters cleanly under a spoon instead of bending, which is precisely the eating quality the format exists for. It is a property to be tuned rather than eliminated.

Condensation is the failure that ruins adhesion

A frozen surface taken out into a warm kitchen collects moisture from the air almost at once, and below freezing that moisture arrives as a thin layer of frost. Coating over frost is a poor idea in two ways. Fat does not adhere to ice, so the shell sits on a layer that will let go, and the water itself will thicken and dull a chocolate coating it comes into contact with.

The practical consequence is that the bar goes from the freezer into the coating with as little time in between as possible, and that a humid kitchen is a worse place to do this than a dry one. Shells that slide off in one piece, or that show a milky, unset patch on the underside, have almost always been applied over condensation rather than mixed wrongly.

Thickness is a trade against everything else

A thicker shell protects the dessert better, sublimating less water away and shielding it from warm air, and it looks generous. It also cracks more readily, takes longer to yield to a spoon, and delivers a mouthful of cold hard fat that has to melt before any flavour is released. Since fat melts more slowly in a mouth already chilled by the dessert underneath, a heavy shell is a genuine eating problem and not merely an aesthetic one.

That is the reason commercial coatings run thin and the reason they contain more fat than a bar of chocolate does. Extra fat lowers the viscosity, so the film drains further before it sets, and the shell ends up thinner without any change in technique. A domestic coating made by thinning chocolate with oil is doing the same thing by a cruder route, and it works, at some cost to the flavour.