Chocolate making, professionally
Chocolate Molding: Solid, Shell, One-Shot
On this page
- Solid molding: bars, blocks and solid figures
- Shell molding by inverting
- Cold stamping
- Hollow figures: spinning closed molds
- One-shot depositing
- Mold materials and the rules that apply to them
- Mold temperature, release and gloss
- Cooling and demolding
- Common defects and what causes them
- The four methods side by side
- Sources and open questions
Molding sets tempered chocolate inside a mold so that it crystallizes properly, shrinks slightly and drops out carrying the mold’s shape and shine. A 2009 University of Leeds doctoral thesis on demolding by Keijbets puts the aim in one line: solidify the fat “with the correct crystallization, as this will lead to contraction of the chocolate and easy removal of the solidified chocolate from the mould.” Why stable crystals contract is explained in what tempering is.
Factories shape molded chocolate by four main routes. Solid molding fills the cavity to make bars, blocks and solid figures. Shell molding builds a hollow shell for a filling, either by turning the mold over to pour out the excess or by pressing a frozen stamp into the chocolate, a method called cold stamping. One-shot depositing puts chocolate and filling into the cavity in a single dose. Hollow figures use a fifth route, spinning a closed mold. The choice between them depends on the product, how tightly the shell and filling have to be controlled, and which defects you can tolerate.
Molding comes right after tempering in the bean-to-bar process.
Solid molding: bars, blocks and solid figures
Solid molding is the plain sequence: deposit tempered chocolate into the mold, vibrate it, cool it, demold it. The Leeds thesis gives one dark chocolate bar process from the literature as an example. Polycarbonate molds are pre-conditioned at about 28 to 30 °C, the chocolate goes in at 30 to 32 °C depending on its composition, vibration releases air and spreads the chocolate into the cavity, the filled molds cool at around 20 °C for about 15 minutes, and the solid bars are demolded. The same source passes on a recommendation of 35 to 40% relative humidity in the room, so the chocolate doesn’t take up moisture.
Those figures describe one product. A thick block holds more heat than a thin tablet, and the literature lets the mold run cooler for larger pieces, as the mold temperature section explains.
Vibration is what drives the air out. Whatever air stays in shows up later as bubbles and pinholes. How a bar line strings these steps together, from depositor to cooling tunnel, is covered in the chocolate production line guide.
Shell molding by inverting
The traditional shell is made by filling the mold, letting a layer set against the walls, and turning the mold upside down so the chocolate that hasn’t set runs out. A 2009 University of Birmingham doctoral thesis by Le Révérend describes the full sequence: fill, cool from the mold side, invert, add the filling, close with tempered chocolate or join two halves, and set in a cooling tunnel. Its traditional tunnels set the cocoa butter slowly, usually around 0.5 °C a minute.
Le Révérend is blunt about the method’s weak points. Most of the chocolate poured in has to be melted, tempered and used again. The inside shape is left to how the chocolate flows, and because the shell crystallizes somewhat at random, the space left for filling changes from piece to piece, so pieces end up under- or overfilled. An even thickness across the whole shell is hard to hold.
All six steps are listed one by one in the production line guide. The hand version, with a scraper and a polycarbonate mold, is in how to make bonbons.
Cold stamping
Cold stamping shapes both sides of the shell at once: the mold sets the outside and a chilled stamp, pressed into the liquid chocolate, sets the inside. In Le Révérend’s account, freshly tempered chocolate goes into the mold at around 30 °C and the stamp goes in for about 3 seconds, long enough to cool the surface and partly crystallize it. The shells then set fully in a tunnel, are filled, pass through a second tunnel, and are closed, either with a layer of tempered chocolate or by joining two halves with a reheated rim, the way Easter eggs are made. A last tunnel finishes them. In the line the thesis models, each tunnel runs at about 12 °C for around 10 minutes.
The gains are speed and control. Process time drops, and the volume of the shells can be held “to very tight tolerances”, in the thesis’s words. Cold stamping also makes thin-walled 3D shapes of about 1 to 3 mm, which earlier methods couldn’t, and some lines turn out up to a million eggs a day.
Sources part ways on the stamp temperature. Le Révérend gives −10 °C as typical and says the press can run as low as −30 °C. A 2007 University of Nottingham doctoral thesis by Baichoo, quoting a 1999 textbook, describes one system running between −15 and −21 °C, with the shell hard enough to lift the plunger off after about 2 seconds. That account also has the mold filled with just enough chocolate for the shell plus a small excess, a long way from the rework of the inverted method. With figures this far apart, there’s no single stamp temperature to copy. Ask the line builder what the stamp runs at on your product, and why.
A stamp that warms up stops working. Baichoo passes on an industrial observation that above about 5 °C the process “does not work well”, with chocolate sticking to the cold plunger and the shell deforming. The thesis’s own lab test, which imitated the stamp with a chilled holder, found adhesion rising above −15 °C and appearing to track the unstable form II.
Speed of cooling is the puzzle. Locally, the chocolate can cool at up to 100 °C a minute, fast enough for unstable crystals to form, while Baichoo cites a limit of about 2 °C a minute for ordinary cooling, beyond which bloom appears. Yet Le Révérend calls the products of cold stamping “perfectly acceptable”, and Baichoo’s textbook source reports even the insides staying glossy after long storage. Le Révérend’s model predicts that the unstable crystals formed in the fast cooling turn into stable ones, leaving a finished product made only of stable crystals. That is a prediction from a model. On the gloss, Baichoo wrote that there was still no suitable explanation.
Hollow figures: spinning closed molds
Hollow figures are made by dosing chocolate into a closed mold and spinning it. A 1996-priority US patent for a hollow chocolate body, US 6,194,018, describes the standard step as spinning the mold “in the usual way by rotating it about different axes”, so the liquid chocolate spreads along the cavity walls and sets as a hollow shell. The patent’s own invention is a body with two chambers, built on that standard step.
One-shot depositing
One-shot depositing puts the shell chocolate and the filling into each cavity at the same moment, through one nozzle. A US patent held by a chocolate maker, US 7,819,053, describes the depositor as “known in the art”: a central conduit for the filling, with one or more ring-shaped conduits around it for the chocolate. The chocolate wraps the filling as it lands, and its exposed surface, the side not touching the mold, becomes the bottom of the finished piece. The patent calls one-shot “more cost effective” than traditional shell molding, without a figure.
It also lists what goes wrong. Large bubbles, small bubbles or pinholes, craters and bumps on the bottom, and filling leaking out are all known defects, and any one of them can be reason enough to reject a piece in quality control. Air in the chocolate causes the bubbles and craters; a crater is what a large bubble leaves when it pops.
The obvious fix for air, vibrating the trays, brings its own trouble. Vibration can push the filling toward the mold wall or toward the open surface, sometimes right through the chocolate, leaving a shell too thin or a piece that leaks. Leakage gets worse when the filling and the chocolate differ in density (caramel is denser than milk or dark chocolate) and when the chocolate has a low yield value, because the filling moves through it more easily. The patent’s own answer is an air knife, a directed stream of gas that clears bubbles before the cooling tunnel. That’s one company’s solution, described in its patent.
Filled drops deposited straight onto a belt, without a mold, use the same concentric nozzle idea and are covered in how chocolate drops are made.
Mold materials and the rules that apply to them
Industrial molds today are mostly polished polycarbonate. The Leeds thesis notes they were traditionally stainless steel, and that flexible PTFE and silicone rubber molds have shown practical advantages thanks to their flexibility, non-stick character and low surface energy. The gloss of a polycarbonate cavity comes from the polish of the tool it was injection molded in. Its weak spot is low resistance to scratches and chemicals: alkaline solutions and amines can attack the material, which matters when choosing a cleaning agent.
Surface energy turned out to be the main factor in sticking. Comparing quartz glass, stainless steel, polycarbonate and PTFE, Keijbets found adhesion between chocolate and mold was lowest when the mold material’s total surface energy was below 30 mN/m, with an electron donor component around 15 mN/m.
On gloss, one couverture maker’s guide says polycarbonate molds bring out a higher shine than aluminum or silicone ones. That’s the maker’s claim, and it’s the only gloss comparison between materials in the sources.
EU: the BPA ban and confectionery molds
In the EU, every food contact material falls under a general rule against passing its constituents into food in harmful amounts, covered in chocolate factory regulations. Plastics also have their own EU measure, Regulation (EU) No 10/2011, and for materials without a specific EU measure, member states can keep or adopt national rules, according to the European Commission’s food contact materials page.
Bisphenol A is the issue for polycarbonate, which is made from it. Commission Regulation (EU) 2024/3190 prohibits using BPA to make food contact materials in its scope, which includes plastics and silicones, and prohibits placing articles made with BPA on the EU market. Its recital 21 names “confectionary moulds” among the repeat-use parts of professional food production equipment, alongside seals, pumps and sight glasses, and grants them a transition period of 36 months. Article 12 lets such equipment “be first placed on the market until 20 January 2028.” Under the correction made by Regulation (EU) 2026/250, equipment placed on the market by then may remain on the market until January 20, 2029.
For a chocolate maker, that sets a date after which new BPA-based polycarbonate molds for professional use can no longer be placed on the EU market for the first time. The regulation’s recitals also explain that a food business may keep using a mold it already owns until it stops working and has to be replaced. When buying molds, ask for the declaration of compliance, which under the regulation must list any bisphenols used in manufacture. The equipment RFQ checklist covers that request.
US: 21 CFR 177.1580
In the US, polycarbonate stays permitted for food contact. 21 CFR 177.1580 allows polycarbonate resins, defined as made from 4,4′-isopropylidenediphenol (BPA) and carbonyl chloride, in articles used to produce, process, pack or hold food. Extractives in water, 50% ethanol and n-heptane must not exceed 0.15 percent by weight, and the only exclusions are baby bottles and sippy cups. Checked as of October 12, 2026.
Mold temperature, release and gloss
The mold should sit close to the chocolate’s own temperature when the chocolate goes in. The literature summarized in the Leeds thesis says the mold should be “as close to the temper temperature as possible”, with one adjustment: large pieces with more thermal mass can go into a slightly cooler mold, while small pieces need the mold as close to the chocolate as possible. A well-conditioned mold is also called “predominantly important” for the right gloss.
Get it wrong in either direction and the chocolate sticks. Too warm, and the mold melts the seed crystals in the chocolate’s surface layer, so the bar holds on when you try to demold it. Too cold, and the chocolate gets a “flash shock”: it sets in unstable crystal forms that pack loosely and contract little, it doesn’t flow evenly over the mold surface, and stress builds up in the fat. Both routes end in poor contraction and a difficult release.
Keijbets’s own lab work, which pulled a probe away from chocolate set on mold surfaces, found demolding easiest with the polycarbonate preheated to 25 to 30 °C under controlled conditions and a cooling temperature of roughly 10 to 15 °C. Water vapor in the air made the mold surface more hydrophilic and lowered adhesion, and cleaning conditions changed the surface chemistry of the polycarbonate. The literature the thesis reviews adds two points: small defects in the cavity raise the amount of chocolate left behind, and washing can strip a thin film of fat that helps release. Rounded shapes with even circumferences contract most evenly.
At workshop scale, one couverture maker’s guide to molded chocolates warms the mold to 26 to 27 °C with a heat gun and keeps the filling below 28 °C so it doesn’t melt the shell. Set side by side, the figures are 28 to 30 °C for a dark bar in the literature, 25 to 30 °C in the lab result and 26 to 27 °C for hand-made bonbons. They cover different products and scales, and none of them is the correct number. Start from the couverture’s data sheet and adjust to the piece. The hand steps for bonbons are in how to make bonbons.
Moisture has to be kept away too. The mold surface has to be dry when the chocolate goes in, and cooling should stay above the dew point so no condensate forms on the chocolate, because water is a cause of sugar bloom. The full story on bloom is in why chocolate turns white.
Cooling and demolding
Cooling has to set the chocolate completely before it leaves the mold, at a pace the chocolate can follow. The Nottingham thesis, citing earlier literature, says every chocolate has a natural rate of solidification and the best results come when cooling follows it, neither too fast nor too slow. Tunnels cool the chocolate in 10 to 30 minutes depending on their type, and the air is warmed slightly at the end so condensation doesn’t form once the chocolate reaches the warmer packing area.
Faster cooling doesn’t mean earlier release. A 2021 ETH Zurich study found an optimum heat flux at which 125 g plates came away from the mold wall soonest; the figure is in the cooling tunnel section of the production line guide, along with the main tunnel types. The same study’s abstract links the intensity and evenness of air cooling to gloss, texture and melting.
Shells need extra care. In the cold-stamped line Le Révérend models, the filling goes in at 35 °C and has to be cooled quickly because it can remelt the shell, and the eggs must be fully solid at the end of the process to demold cleanly. A cooling tunnel for shells therefore has two jobs, setting the chocolate and protecting it from the filling.
Common defects and what causes them
The defects that hit every method trace back to mold temperature, cooling, air or moisture; the rest belong to one method. The table collects the causes named in the theses and the patent cited above.
| Defect | Method | Cause given in the sources |
|---|---|---|
| Sticks in the mold, won’t release | All | Mold too warm (seed crystals melt) or too cold (unstable forms, little contraction); mold surface energy; scratches; cleaning that changes the surface (Keijbets) |
| Dull or gray surface | All | Poorly conditioned mold; cooling too fast, which gives bloom or “white work” (Keijbets; Baichoo) |
| Sugar bloom | All | Moisture on the mold, cooling below the dew point, condensation at packing (Keijbets; Baichoo) |
| Uneven shell, under- or overfilled | Inverted shell | Shell crystallizes at random, and the inside shape depends on flow (Le Révérend) |
| Chocolate sticks to the stamp, shell deforms | Cold stamping | Stamp above about 5 °C (industrial observation); lab adhesion above −15 °C linked to form II (Baichoo) |
| Risk of unstable crystals | Cold stamping | Local cooling up to 100 °C a minute; a model predicts conversion to stable crystals (Le Révérend) |
| Bubbles, pinholes, craters | One-shot, and any poured chocolate | Air in the chocolate; a popped bubble leaves a crater (US 7,819,053) |
| Filling leaks, shell too thin | One-shot | Density difference between filling and chocolate, low yield value, tray vibration (US 7,819,053) |
| Shell melts under the filling | Shell with filling | Warm filling, 35 °C in the modeled line; one maker keeps workshop fillings below 28 °C (Le Révérend; maker guide) |
The four methods side by side
Each method trades rework, control and steps differently, and the sources fill in most of that grid. “Not stated” means none of the sources used here gives a figure or a description for that cell.
| Solid | Inverted shell | Cold stamping | One-shot | |
|---|---|---|---|---|
| Makes | Bars, blocks, solid figures | Filled bonbons, hollow halves | Thin shells of about 1 to 3 mm, filled afterward | Filled pieces in one dose |
| Main steps | Deposit, vibrate, cool, demold | Fill, set a layer, invert, fill, close, cool | Fill, stamp for about 2 to 3 s, tunnel, fill, tunnel, close, tunnel | Co-deposit through a concentric nozzle, remove air, tunnel |
| Chocolate reworked | Not stated | Most of it | A small excess | Not stated |
| Shell and filling control | Not applicable | Weak | Very tight tolerances | Risk of thin shell and leakage |
| Cooling passes | One | Several (shell, filling, closing) | Three tunnels in the modeled line | One in the patent’s example |
| Typical defects | Sticking, dull surface, bubbles | Uneven thickness, filling volume | Sticking to the stamp, unstable crystal risk | Bubbles, craters, leakage |
| Cost | Not stated | Costlier than one-shot (patent) | Not stated | “More cost effective” than shell molding (patent) |
Sources and open questions
All three doctoral theses cited here were sponsored by confectionery companies: Keijbets at Leeds, Le Révérend at Birmingham and Baichoo at Nottingham. The one-shot patent belongs to a chocolate maker, and patents describe methods without proving how widely they’re used. The 2021 cooling study by Grob and colleagues was read as an abstract. Regulation (EU) 2024/3190 was read as first published, and Regulation (EU) 2026/250 through its recitals.
Still open:
- cold stamp settings: the sources give −10 °C typical (as low as −30 °C) against −15 to −21 °C, and about 3 against about 2 seconds, with no explanation for the difference;
- vibration settings for solid and one-shot molding, and independent cost or speed figures for one-shot;
- an independent gloss comparison of polycarbonate, silicone and metal molds;
- how demolding is done mechanically (inverting, knocking, twisting), and independent capacity figures for spinning lines;
- which molding materials replace BPA polycarbonate for new molds sold in the EU.