Specialist Crafts · Article 12

Aeration & Mousse Systems

A reference article · 2,400 words · 26 cited sources · last reviewed October 2026 · fact-checked October 2026, 20 claims

Every airy dessert is a foam, and a foam starts falling apart the instant you make it. The craft is either slowing that collapse long enough to eat, or freezing and setting the structure so it holds for days. Mousse, soufflé, marshmallow, and whipped cream are all the same problem solved four ways.

Whip a cup of cream and you create tens of square meters of new surface between air and water: with bubbles roughly 20 to 100 micrometers across (Ihara et al., 2019), a cup whipped to about 100 percent overrun works out to roughly 15 to 70 m². Surfaces cost energy and want to shrink, so from the instant a foam exists it is trying to collapse back into a puddle (Wilde, 2000). That single fact runs this entire chapter. Everything a pastry kitchen does with air, whipped cream, meringue, mousse, soufflé, marshmallow, is a way of either slowing that collapse long enough to serve, or converting the liquid around the bubbles into a solid so the collapse never gets the chance. Learn the handful of things that make a foam fail and every airy dessert becomes the same problem with a different lock on it.

What a foam is, and why it fails

Pure water can’t hold a foam, because nothing at the surface resists the film thinning out. Foams need surface-active molecules that crowd to the air-water boundary. Proteins (from egg white, cream, or gelatin) adsorb slowly but then unfold and bond into a continuous elastic skin that resists stretching and rupture, which is the source of lasting stability. Small surfactants like soy lecithin adsorb fast and make huge but fragile foams, the modernist “airs” that live for minutes. They protect a film a different way: when a film gets stretched thin, surfactant rushes back into the thin spot and drags liquid with it, the Marangoni effect (Wilde, 2000).

A foam dies three ways, and they reinforce each other. Drainage is gravity pulling liquid down out of the films, which you fight with viscosity (sugar, protein, hydrocolloids) and smaller bubbles. Coalescence is films rupturing so neighboring bubbles merge, which is catastrophic and irreversible. Disproportionation is the quiet one: gas diffuses out of small high-pressure bubbles into large low-pressure ones, coarsening the foam with no rupture at all, which is why proteins, with their tough elastic skin, beat surfactants and why a less-soluble gas holds better (Murray & Ettelaie, 2004). The distinction that matters most sits on top of all this. A transient foam (a modernist air, an unstabilized egg foam) is held together only by kinetics and lives minutes to a couple of hours. Whipped cream sits in between: a network of partly coalesced, partly crystalline fat props it up (Fredrick et al., 2010), so it holds for hours in the fridge. A set foam has had its liquid phase turned to solid or gel while the bubbles were trapped, by gelatin, by starch, by cooked egg protein, or by crystallized fat, so drainage stops and films can no longer rupture.

The aerators

Each way of adding air has its own mechanism. Whipping cream works by partial coalescence, and it needs at least 30% milkfat (21 C.F.R. § 131.157, 2024), ideally 33 to 38%. Cold, partly crystalline fat globules get driven to the bubble surfaces and their protruding crystals link neighbors into a semi-solid network that armors the foam (Fredrick et al., 2010). This is why cream has to be cold, around 39 to 45°F (4 to 7°C): above about 50°F (10°C) too little of the fat is solid, so the foam comes out soft and weak and churns quickly. Over-whipping at any temperature carries the same partial coalescence on into butter. Whipping egg white unfolds its protein into a film, helped by a trace of acid and wrecked by any trace of fat, with sugar added gradually after soft peaks to raise viscosity for a fine, glossy foam (Lomakina & Míková, 2006). Whole eggs foam slowly because the yolk fat interferes, so bakers warm the eggs and sugar to about 110°F (43°C) first, and the warm foam whips to roughly triple its volume, the génoise ribbon (François, n.d.). A pâte à bombe whips yolks with a hot syrup streamed in at 244 to 250°F (118 to 121°C), which lightly cooks them and keeps the result silky (GialloZafferano, n.d.; Joe Pastry, n.d.; Pastry Maestra, 2019); that brief heat is not a validated pasteurization step (see the Food safety note below). The three meringues each aerate differently: by pastry chefs’ consensus, French is the lightest and least stable, Italian the densest and most stable, Swiss in between (Tasting Table, n.d.).

The step that ruins more foams than any whisk is folding. The goal is to combine a light mass into a heavier base without shearing the bubbles apart, using a slow motion that cuts down through the center, sweeps along the bottom, lifts over, and rotates a quarter turn, interleaving the two masses in sheets. Two rules make it work: first sacrifice a third of the light mass by whisking it into the base to loosen it, then fold in the rest, and second stop early, the moment the streaks vanish, because every extra stroke costs air. For instant foams, a siphon charged with nitrous oxide makes an espuma, the airiest foam without whipping, but it still needs a stabilizer to hold once dispensed and it stays transient.

The mousse map

A mousse is an aerated base held semi-set, and you can locate any mousse on two axes: what lightens it and what sets it. It might be lightened by whipped cream (air plus rich, melting fat), by meringue (air without fat, for a lighter and sharper result), by pâte à bombe (air plus richness and pale color), or by a deliberate combination that tunes lightness and richness independently. What sets it is the more interesting question, because chocolate mousse largely sets itself as its cocoa butter firms, so it needs about half the gelatin of a mousse without chocolate, or none (eGullet Forums, n.d.); how much depends on the type and proportion of chocolate.

The non-chocolate mousses lean on gelatin for structure. A fruit mousse needs a setting agent, usually gelatin, because fruit brings flavor, acid, and water but little structure of its own, with two traps to watch: strong acid weakens a gelatin gel (GMIA, 2019), and raw pineapple, kiwi, papaya, fig, and ginger carry protein-digesting enzymes (bromelain, actinidin, papain, ficin, and zingibain) that cut gelatin apart, so those fruits get cooked first (Science Buddies, 2013) or set with agar or pectin, carbohydrate gels the enzymes don’t attack. Gelatin makers also warn against raw mango and guava. A bavarois is crème anglaise plus gelatin plus whipped cream, used where the mousse is wrapped in sponge, as in a charlotte (Cacao Barry, n.d.). More gelatin makes it firmer and less melting.

The craft is hitting the target firmness with the least gelatin, so the mousse still melts on the palate instead of eating like a rubber band. A spoonable verrine can sit at the bottom of that range or below it; a slice that has to stand needs more. The Gelatin Bloom Converter doses by percentage of total mass for exactly this choice.

Soufflés

A hot soufflé is the clearest demonstration of foam physics rising in real time. It’s two parts: a flavored base (a starch-thickened pastry cream or a fruit purée) that carries flavor and setting proteins, and a whipped egg-white foam folded gently into it. The rise comes from gas effects acting on the trapped bubbles at once. Whipped-in air expands as it warms, but water turning to steam does most of the lifting (Joe Pastry, 2010): at 212°F (100°C) steam takes up about 1,600 times the volume of the water it came from (National Institute of Standards and Technology, n.d.). Meanwhile the egg-white proteins denature between roughly 149 and 185°F (65 and 85°C), from conalbumin up to ovalbumin, the main egg-white protein, a little higher with sugar (Donovan et al., 1975). That converts the stretched liquid films into semi-rigid walls that lock the expansion in place. The whole art is timing that set to peak expansion.

It falls because the opposite happens on the way out of the oven: steam stops generating and condenses, air contracts, and if the walls set firmly they hold most of the shape while a weak or over-folded one deflates. A little settling is normal, which is why a soufflé is a fired-to-order dessert. Two practical habits decide success. The ramekin is buttered in upward strokes and coated with sugar, a long-standing practice said to help the batter climb straight (The Takeout, 2025). And the oven stays hot and closed, around 375 to 400°F (190 to 205°C), because opening the door early lets heat and steam escape. A frozen soufflé is a visual pun with none of this physics: an aerated parfait is piped above the rim into a paper collar and frozen, then the collar is peeled away to fake the risen dome.

Other aerated forms

The same foam-plus-lock logic produces a whole family. Marshmallow is a whipped syrup or Italian meringue locked with gelatin, where the invert sugar keeps it smooth and holds humidity so it stays soft. Nougat de Montélimar whips hot honey and a high-cooked syrup into egg whites and sets it by drying and sugar-glass rather than gelatin. Whipped ganache is a chilled ganache whipped to aerate, set by its crystallizing fat like a denser whipped cream. And the modernist end runs from fragile fat-free lecithin airs through siphon foams to methylcellulose foams, which gel when heated instead of cooled (Sarkar, 1979). A whipped methylcellulose solution sets firm while hot and softens again only as it cools well below its setting point, a reversible hot foam: gelatin can’t make it, and heat-set protein foams like soufflé and meringue stand up hot but can’t undo their set.

Getting it right

Most foam failures come down to two variables, and the first is folding temperature. Fold cold cream or meringue into a base that is too hot and the heat melts the fat network and thins the films, so the air is lost. Fold it into a base that is already at its setting point, or pour cold cream onto melted chocolate, and the gelatin sets in lumps or the chocolate seizes into chips. The target is a narrow window just above the set point.

The second is over-whipping. Cream taken too far turns grainy and then to butter; whites taken too far go dry, matte, and grainy as the protein over-coagulates and the films rupture, at which point they cannot stretch in the fold or the bake. For anything you plan to fold, stop at soft-to-firm peaks, because folding does more whipping on its own, and a slightly under-whipped cream is usually right for a mousse (Cacao Barry, n.d.). When a set mousse comes out wrong, the diagnosis is short: soupy means too little set or a foam deflated at folding, dense means too little air or air lost, and rubbery means too much gelatin, a common fault with a counterintuitive fix: use less setting agent and lean on the air and the fat. A great mousse is barely set.

The idea in one paragraph

Every airy dessert is a foam, and a foam is unstable the moment it exists, fighting drainage, coalescence, and gas diffusion the whole time. You slow those with surface-active proteins, thick liquid, and small bubbles, and you defeat them by setting the matrix, turning the liquid to gel or solid while the bubbles are captured, with gelatin, starch, cooked egg, or crystallized fat. A mousse is that idea held semi-set: lightened by cream, meringue, or pâte à bombe, and set by cocoa butter or the least gelatin that will hold, folded in a narrow window just above the set point so the air survives. A soufflé is the same foam rising on steam and locked by coagulating protein, and marshmallow, nougat, and whipped ganache are each a foam with a different lock. Master how foams fail and how they are set, and you can build or fix any of them.