Specialist Crafts · Article 10

Chocolate & Confectionery

A reference article · 3,000 words · 45 cited sources · last reviewed October 2026 · fact-checked October 2026, 43 claims

Chocolate work is really two chemistry problems: getting one fat to crystallize into the right form, and getting sugar to either crystallize finely or not at all. Master those two and tempering, ganache, caramel, and showpiece sugar stop being separate crafts.

Almost everything a chocolatier does comes back to two ingredients behaving predictably: cocoa butter, the fat that has to crystallize into exactly the right form, and sugar, which you either coax into fine crystals or stop from crystallizing at all. Tempering, molded bonbons, ganache, caramel, and pulled-sugar showpieces look like five separate skills, but they run on those two ideas plus one more, water activity, which decides how long a filling keeps. Learn the three and the whole craft connects.

Where chocolate flavor comes from

Fresh cacao tastes nothing like chocolate. The flavor is built in three steps before a chocolate maker ever touches it. Fermentation, anywhere from one to three days for Criollo beans to about a week for Forastero, lets yeasts and bacteria work the bean’s pulp, killing the germ and generating the pool of free amino acids and reducing sugars that flavor will later come from (Schwan & Wheals, 2004; Afoakwa et al., 2008b). Drying brings the beans down to six to eight percent moisture and drives off some of the harsh volatile acidity (Afoakwa et al., 2008b). Then roasting at roughly 248 to 302°F (120 to 150°C) runs the Maillard reaction, the same browning chemistry behind toast and seared meat, between those fermentation precursors, and that’s where “chocolate” aroma actually gets created (Afoakwa et al., 2008b).

Turning roasted beans into chocolate is mostly a grinding problem. The nib is ground to a liquor, refined with sugar (and milk powder for milk chocolate) down to a particle size around 20 microns, because particles coarser than about 30 microns read as gritty on the tongue, then conched for many hours, commonly 10 to 24, to coat every particle in cocoa butter and drive off the last of the acidity (Afoakwa et al., 2007; Afoakwa et al., 2008b; Malvern Panalytical, 2016). Couverture is the professional coating chocolate, made with extra cocoa butter: EU rules require at least 31% cocoa butter in dark couverture and at least 31% total fat in milk couverture, and many run well above that, with Callebaut’s 811 dark couverture at about 37% fat. That fat is what lets it flow into thin, even shells and temper to a proper snap (Directive 2000/36/EC, 2000; Callebaut, n.d.-a). Compound coating swaps the cocoa butter for other vegetable fats. The common lauric and non-lauric types need no tempering, though coatings built on cocoa butter equivalents still do; all of them set softer, have less snap, and in many places can’t legally be called chocolate (Lonchampt & Hartel, 2004; New Food, 2007; Directive 2000/36/EC, 2000).

Tempering: growing one crystal, on purpose

Tempering is the single most important skill in chocolate work, and it exists because cocoa butter is polymorphic: its fat can solidify into six different crystal structures in the classic scheme, called Forms I through VI, each with its own melt point and stability (Wille & Lutton, 1966). Later X-ray work treats Forms III and IV as one continuous range, so some researchers count five (van Malssen et al., 1999). Form V is the one tempering aims for, because it gives finished chocolate everything you want at once: high gloss, a hard clean snap, the contraction that pops it out of a mold, resistance to bloom, and a melt just below body temperature (Afoakwa et al., 2008a). Form VI is more stable still, but it forms slowly in storage and comes with bloom, not something you can temper for.

The trick is that unstable forms crystallize more easily than Form V, so you can’t just cool melted chocolate and expect the good crystal. The classic curve is melt, cool, then rewarm. Melt fully to erase every existing crystal, cool while stirring to nucleate a mix of crystals, then rewarm to a precise working temperature that melts out the unstable Forms I through IV and leaves only Form V seeds behind to template the rest of the mass as it sets (Afoakwa et al., 2008a; Callebaut, n.d.-d).

Cool and working temperatures from Michel Cluizel (n.d.); melt temperatures from Callebaut (n.d.-d). Manufacturers differ by a degree or two: Callebaut cools all three to 81°F (27°C), and Cluizel melts dark higher, to 131 to 136°F (55 to 58°C).

Milk and white chocolate sit lower because their milk fat softens the cocoa butter (Lonchampt & Hartel, 2004). The working temperature is a ceiling, not a target to drift past: take dark chocolate above about 93°F (34°C) and you melt the Form V seeds and fall out of temper; Callebaut puts the upper limit for its 811 dark at 94°F (34.5°C) (Callebaut, n.d.-d; Wille & Lutton, 1966). On the bench, a favorite method for small and medium batches is seeding: stir finely chopped tempered chocolate into the warm melt in stages, so the added Form V crystals do the templating for you (Callebaut, n.d.-d). The Mycryo approach waits until the melted chocolate has cooled to about 93 to 95°F (34 to 35°C) for dark, or 91 to 93°F (33 to 34°C) for milk and white, then stirs in about one percent of powdered Form V cocoa butter (10 g per kg) and holds the chocolate at working temperature (Cacao Barry, n.d.-c). Good temper shows as gloss, a hard snap, and clean contraction; a knife tip dipped and left at about 68°F (20°C) should set evenly and glossy in about three minutes (Cacao Barry, n.d.-b).

When it goes wrong, the surface tells you which way. Chocolate that was worked too warm has too few good seeds, so it sets dull, soft, and streaky and soon blooms with whitish spots as the poorly crystallized fat separates from the sugar and cocoa solids (Lonchampt & Hartel, 2006). Well-tempered chocolate can bloom too, but later and differently: fat bloom in storage is a grey film that forms as Form V slowly converts to Form VI (Lonchampt & Hartel, 2004). Chocolate worked too cool is over-tempered: in bakers’ experience it thickens and casts heavy shells, and it sets with a dull, rough surface (Lonchampt & Hartel, 2006). Store finished chocolate at a stable 54 to 64°F (12 to 18°C) and below about 70% relative humidity, and let it come up to room temperature still wrapped when it leaves cold storage, to keep both fat bloom and sugar bloom (a gritty crust from condensed moisture dissolving surface sugar) away (Callebaut, n.d.-b; Lonchampt & Hartel, 2004).

Molded and dipped work

A molded bonbon is built in a spotless, dry polycarbonate mold. Manufacturers disagree on how warm that mold should be (see the callout below); Callebaut adds that it must never be warmer than the tempered chocolate going into it (Callebaut, n.d.-f). You fill the cavities with tempered couverture, vibrate to knock out the air that would otherwise leave pinholes, then invert and pour out the excess so the clinging layer forms an even shell; a more fluid couverture gives a finer one (Cacao Barry, n.d.-a; Callebaut, n.d.-f). Once that sets, you fill with ganache or caramel to just below the rim, keeping the filling no warmer than about 77 to 82°F (25 to 28°C) so it doesn’t melt the shell, let it set, and cap it with a final flush layer of tempered chocolate (Callebaut, n.d.-e; Cacao Barry, n.d.-a). The whole system depends on contraction: a properly tempered shell shrinks as it sets and lifts cleanly off the mold wall, and the same well-formed Form V crystals, set against the polished mold, give it that mirror shine (Afoakwa et al., 2008a; Callebaut, n.d.-f). Cool it at about 57 to 61°F (14 to 16°C), cool but not cold; a cooling cabinet or fridge set to that range works, but cooling much colder sets it too fast and leaves dull spots, and condensation from cold storage invites sugar bloom (Callebaut, n.d.-c; Callebaut, n.d.-g).

Fillings that keep: ganache and water activity

A fresh pastry cream lasts days in the refrigerator; a sellable bonbon has to last weeks at room temperature. The lever that makes that possible is water activity, written a_w: a 0-to-1 measure of how available a filling’s water is to microbes and reactions, strictly the ratio of its water vapor pressure to that of pure water. It’s not the same as how wet something is: sugars, glucose syrup, and polyols bind water, lowering water activity without drying the filling out (Callebaut, n.d.-h; Ergun et al., 2010). Most bacteria need a_w above about 0.90, but Staphylococcus aureus can grow and form toxin down to about 0.86 to 0.87, which is why the FDA Food Code treats only fillings below 0.88 as safe without refrigeration; most common spoilage molds are held back somewhere between 0.87 and 0.80 (Beuchat, 1983; Manitoba Agriculture, n.d.; U.S. Food and Drug Administration, 2026). That is why a room-temperature ganache is kept below about 0.85, which stops disease-causing bacteria from growing. It does not kill any that are already there, since Salmonella survives for long periods in dry foods and chocolate has carried outbreaks, so clean ingredients and handling still matter (Finn et al., 2013). Nor does it stop spoilage: molds and yeasts that tolerate dry, sugary conditions can grow down to about 0.60, so a long-keeping ganache needs a_w well below 0.80, and its shelf life has to be confirmed by trial rather than read off a single number (Beuchat, 1983; Ergun et al., 2010; Callebaut, n.d.-i).

You lower water activity with humectants, ingredients that hold onto water. Swapping a portion of the sugar and cream for invert sugar (which also keeps the ganache soft and resists graining), glucose syrup, or sorbitol pulls the water activity down while keeping the texture supple (Ergun et al., 2010). Stack a few more hurdles on top, a slightly lower pH, sometimes a little alcohol, and above all a fully sealed tempered shell acting as its own moisture barrier, and a dark ganache bonbon holds for three to four weeks or more (Keylink, n.d.; Ergun et al., 2010). The ganache itself is an emulsion, oil-in-water while hot and becoming bicontinuous as it sets, so build it the way you would any emulsion: blend from the center outward or with an immersion blender kept below the surface, and add soft butter last, once the ganache has cooled to about 95 to 104°F (35 to 40°C) (Saglio et al., 2018; Valrhona, n.d.). Let it set slowly at 61 to 64°F (16 to 18°C) for a fine, snappy interior (Valrhona, n.d.). Firm caramel fillings keep longer because their water activity is low, while very soft, flowing caramels carry more water and keep less well (Ergun et al., 2010). Praliné and gianduja are nut-and-sugar pastes with almost no water, so they keep longer too.

Sugar confectionery: crystalline or glassy

Every sugar confection sits on one side of a single divide. Crystalline sweets like fondant and fudge are built on controlled, very fine sugar crystallization; glassy or amorphous sweets like hard candy, and most toffees and caramels, are concentrated syrups that set without crystallizing, hard candy as a true glass and soft caramel as a chewy amorphous mass. Which one you get is something you control, mainly with doctoring agents: glucose or corn syrup, invert sugar, and a little acid all interfere with sucrose crystals, so more of them pushes a batch toward glass while less of them, plus deliberate agitation, gives you fine crystals (Clark, 2004; Ergun et al., 2010). The other variable is temperature, because in a boiling syrup temperature tracks sugar concentration closely for a given recipe and air pressure; glucose syrup and invert sugar shift the curve a little (Clark, 2004; Ergun et al., 2010).

Caramel is where this gets tasty. Cook it dry, melting the sugar alone, for a fast and intense but scorch-prone result, or wet, dissolved in water first, for a slower and more forgiving one (King Arthur Baking, 2022). Soft caramels, made with sugar, glucose, and cream or butter, are cooked to roughly 239 to 248°F (115 to 120°C), and a few degrees noticeably firms the set (Greweling, 2013; INFLIBNET, n.d.). Their flavor is mostly Maillard browning between milk proteins and sugars rather than pure caramelization, the thermal breakdown of sugar alone, which in a quick cook takes about 320 to 340°F (160 to 170°C); sucrose decomposes rather than melting at a fixed point, so the onset depends on how fast it is heated (INFLIBNET, n.d.; Exploratorium, n.d.; Lee et al., 2011). Fondant is the clearest illustration of crystalline technique: cook a syrup to soft ball, cool it undisturbed to about 104°F (40°C), then agitate the supersaturated syrup hard so it throws millions of microscopic crystals into a smooth paste (Iowa State University, n.d.). Fudge is the same move made with milk and butter, cooked to soft ball and beaten. Push past soft crack and you are making hard toffee and brittle, glassy and hard, where a single stray crystal can ruin the batch (INFLIBNET, n.d.; Clark, 2004). Aerated confections branch off from here too: nougat whips hot syrup and honey into egg whites, while marshmallow whips hot syrup with gelatin, which both holds the air and sets the foam (Wikipedia, n.d.-a; Wikipedia, n.d.-b).

Showpiece sugar

Cast and pulled sugar showpieces are their own discipline, and the modern medium is often isomalt, a sugar alcohol cooked to roughly 320 to 356°F (160 to 180°C) (Institute of Culinary Education, n.d.). Isomalt is widely used in place of cooked sucrose for showpieces because it is far less hygroscopic: it pulls in little moisture from the air, so it resists going sticky and cloudy the way sugar does (BENEO, n.d.; Institute of Culinary Education, n.d.). Pulling it under a warm lamp, folding it over twenty or thirty times, gives it a satin gloss, blowing inflates a warm ball like glass, and casting pours it into molds for discs, bars, and bases (Torreblanca, 2017; Institute of Culinary Education, n.d.). Humidity is still the enemy, so keep it low while you work, and store finished pieces airtight with silica-gel desiccant (Institute of Culinary Education, n.d.).

Troubleshooting

Causes and fixes from Lonchampt & Hartel (2004, 2006), Callebaut (n.d.-f), Joe Pastry (2009), Clark (2004) and BENEO (n.d.).

The idea in one paragraph

Chocolate and confectionery are two crystallization problems wearing a dozen costumes. Cocoa butter has to be tempered so it sets as Form V, the crystal that is glossy, snaps, contracts out of a mold, and melts just below body temperature, which you get by melting fully, cooling to nucleate, and rewarming to a precise ceiling. A bonbon then keeps for weeks because its filling’s water activity is pulled down with humectants and sealed inside a tempered shell, not because it is dry. And every sugar confection is a decision between fine crystals and none: doctoring agents and agitation give you crystalline fondant and fudge, while glucose and a fast cool give you glassy caramel, toffee, and pulled sugar, with the candy stages simply marking how much water you have boiled away.