Food-Grade RTV Mold Applications, Split by Food Type

If you're ordering a custom RTV mold for chocolate, cake batter, hot sugar, or ice that people will eat, the mold is the food-contact article. It has to be formulated, platinum-cured, post-cured, cleaned, and tested as that article under a named rule: FDA 21 CFR 177.2600 for US repeated-use rubber, or LFGB/BfR XV for the German/EU path.

Applications then split by food type, temperature, and how you use the jacketed cavity, not by catalog photography. A resin-casting tin-cure mold that looks like a muffin pan is the wrong article for a bakery line. A steel LSR injection quote is a different quotation for a molded *part*, not this RTV tool. BPA-free on the carton doesn't sort those uses, because silicone isn't polycarbonate.

Common mix-up: one “food mold” RFQ covers cake, chocolate, gummies, and ice

A lot of RFQs treat food-grade as a single tool family you can dump anything into. In the shop that usually means one spatula certificate, one Shore, and four foods that don't share flasks, splits, or jackets.

The food-contact surface of the SKU in the finished form — the platinum RTV face after recast against the master — is what the lab extracts or migrates. Start with a sentence: this cavity will see this food, at this temperature, for this time, repeatedly. Oven cake, tempered chocolate, boiled sugar, and frozen water are four different sentences. They may share a platinum RTV family. They don't share a report. Color, logo, and cavity count change the article if pigment or the recast recipe changes.

What to ask when you're buying by application

• What food actually hits the cavity (aqueous batter, fatty chocolate, acidic gummy, dry cookie, ice), and which method follows: 177.2600 water and/or n-hexane, and/or BfR XV simulants plus sensory?
• Oven, freezer, dishwasher, or ambient-only — and did you write those claims so the lab times match?
• One-piece tray or keyed split? Hollow ice or chocolate needs a core pull path.
• What's the Shore of the skin, and is there a rigid jacket so it doesn't balloon under batter or a depositor dump?
• Who owns the master, and is a recast included when the face tears or you change a seasonal logo?
• If you're also quoting molded LSR spatulas or pans, is that a separate *part* line, not this mold file?

Bakeware cavities: batter, oven, then demold

Cake pans, muffin trays, madeleine shells, bread forms, and sheet mats see aqueous and often fatty batter, then oven heat, then demold. Those claims have to match test time and temperature. A 177.2600 hexane flask for fatty use isn't an oven rating by itself. An LFGB stack typically includes volatiles, simulants, and sensory at conditions that represent the claim.

On the RTV side you're looking at even wall thickness against the master so post-cure is uniform, a rigid support jacket so Shore 10A–20A rubber doesn't slump under batter, draft and texture that release cake without a food-illegal spray, and a mold box that keeps the jacket square. Tin-cure RTV doesn't belong here. High-ratio multi-cavity muffin tools only stay one article if pour, cure, and post-cure are uniform against one master.

Don't print “FDA approved” on a muffin tray. Print that the finished RTV article was tested to 21 CFR 177.2600 for the food types claimed, or to LFGB/BfR XV for EU retail.

Chocolate, candy, and confectionery dumps

Chocolate is fatty, usually cool to warm, and detail-hungry. Polycarbonate chocolate molds still exist for shine; platinum RTV is chosen for undercuts, flexibility, keyed splits, and small-batch changeover by recasting the face. Fatty-food extraction under 177.2600 is in play. Sensory still matters for LFGB because a mold that smells like catalyst will flavor the first few pulls.

Hard candy and gummies add heat, stickiness, and often acidity. Hot sugar isn't the same load as tempered chocolate. Design the demold in the geometry, the Shore of the skin, and the jacket stiffness, not in an unlisted oil.

Ice trays, popsicle molds, and two-piece ice spheres are repeated-use, usually aqueous, often freezer-claimed. They still need a cleansed article and a report on that SKU. They don't need an oven simulant unless you print one. Inner cores need a pull path; if the core can't extract, you designed a cut mold, not a production mold.

Commercial confectionery lines also care about how the mold sits on a depositor. That's registration keys and jacket flatness, not a second food-contact class. A 64-cavity gummy sheet is still one SKU if compound, color, and post-cure are frozen. Change the pigment for a seasonal SKU and you've got a new article, even if the master CAD didn't change.

Worked example: ice-pop versus praline on one RFQ

A buyer sent one RFQ for “food-grade silicone molds” covering a 90 mm ice-pop with a stick slot (aqueous, freezer, maybe 300–600 cycles a month) and a 15 mm praline tray, 24 cavities, fatty chocolate, 800–1,500 dumps a month.

Those are two platinum RTV tools: different masters, different split and core questions, different flasks (water versus n-hexane; German acetic/ethanol/sensory scoped to actual contact). A single spatula certificate covers neither. The ice-pop needs a stick-slot that washes; the praline needs dry release and a jacket that holds cavity depth for piece-weight. Write two use sentences, then two travelers, unless the PO honestly claims one mold will see both foods and you test the harsher set.

Resin jewelry, soap, and candle molds can be platinum or tin RTV poured in the same shop. If they never touch food, food-contact testing is the wrong spend. If a buyer later uses a soap mold for chocolate, that's a misuse, not a hidden certification. Keep the intended use in the quote and on the packing insert.

When this article is the wrong quote

If you're buying a resin, soap, or candle mold, food-contact testing is the wrong spend — keep the use on the packing insert so nobody “promotes” it into chocolate later. If you need a high-gloss chocolate *bar face*, polycarbonate still wins on shine; platinum RTV wins on undercuts and recast changeover. If you need thousands of molded silicone bakeware *parts*, that's LSR with a cold-runner or HCR compression, not this jacketed cavity. A one-piece shallow ice tray can skip a split; a two-sided logo or a hollow sphere usually can't. Geometry can override quantity.

Related reading: baking RTV molds, chocolate RTV molds, candy and gummy cavities, gummy trays.

What to send so the application matches the flask

Copy this. One “food mold” line without a use sentence isn't comparable.

• Food: batter / chocolate / gummy or hard candy / ice — one primary, extras only if you'll print them
• Temperature and time: oven peak, dump temp, freezer, dishwasher if claimed
• Master photos or STL; undercuts, stick slots, hollow cores, two-sided logos
• Cavity count, piece size, monthly dumps, depositor pitch if it sits on a line
• Platinum RTV Shore target and jacket yes/no
• Markets and methods: 177.2600 and/or LFGB/BfR XV

Send the use sentence, master photos, and cavity count. We'll split ice-pop from praline if they're really two tools, and we'll flag a parts quote if someone handed you cold-runners.

We would not recommend one "food mold" RFQ for four foods

We would not recommend copying a chocolate tool spec onto boiled sugar, or filing a spatula certificate under a gummy sheet. We would not skip a jacket on a depositor plate because a one-piece ice tray didn't need one. Geometry can override quantity.

flowchart TD
  F[Food on the cavity] -->|batter / oven| A[Aqueous + fatty; oven-representing LFGB]
  F -->|chocolate| H[Fatty / n-hexane; sensory; cooler simulants]
  F -->|gummy / fruit acid| G[Aqueous + acetic; dump temp; jacket + registration]
  F -->|ice| I[Aqueous water flask; freezer; skip oven package]
  A --> S[This SKU, this master, this recast]
  H --> S
  G --> S
  I --> S

Typical shop values: application bands on RTV food molds

| Item | Typical shop value | | --- | --- | | Cavity Shore | 10A–20A platinum RTV with jacket | | Loaded-tray edge sag | About 0.5–1.5 mm if the jacket is missing on a large food tray | | Outer-row gummy weight | Often 3–6% heavy vs center if the sheet sags on the depositor | | Mix | Platinum 1:1 or 10:1; vacuum degas | | Linear shrink | ~0.1–0.3% platinum | | Monthly dumps (first band) | Ice-pop ~300–600; praline ~800–1,500; gummy sheet ~1,500–3,000 |

Caption these as typical shop values, not a measured PO.

FAQ

Does a bakeware certificate cover my custom RTV chocolate mold in the same hardness? The mold is the article. Chocolate is a fatty, detail-facing use on a different master, often a different split and jacket, so requalify the SKU. Shore A isn't a test.

Are ice-sphere RTV molds exempt from 21 CFR 177.2600 because they stay cold? Cold doesn't remove repeated-use rubber rules, and ice is aqueous food contact. Oven simulants may be unnecessary if you don't claim oven use. A finished-article report on that two-piece cavity still is.

Can I use a resin-casting tin-cure mold for cake if I wash it first? Washing doesn't change chemistry. Tin-cure RTV and industrial compounds are the wrong start. Food-contact needs a formulated, post-cured, tested platinum RTV article.

Why do gummy and hard-candy cavities need a different conversation than chocolate? Hot sugar, stickiness, and fruit acids change temperature, release, and often the simulant set. Chocolate is typically fatty and cooler. Don't copy a chocolate tool spec onto boiled-sugar production, and don't skip jacket stiffness or depositor registration on a gummy sheet.

If my RFQ also asks for molded LSR spatulas, is that the same food-grade file? The spatula is a molded *part*, usually LSR injection with a cold-runner. The RTV mold is a poured cavity against a master. Specify both if you need both; don't hide a part tool under a mold extraction report.