How to Verify RTV Food-Contact Certifications

If you're verifying a food-contact certificate for a custom RTV mold, open the PDF and read the specimen line. If it isn't your cavity SKU, you're not verified. Verification is matching a named method, a lab, and a described article to the part you'll ship. A file named "FDA.pdf" isn't that match. 21 CFR 177.2600 and LFGB / BfR XV are methods on finished articles. Platinum on a website isn't a method. Tin-cure RTV with a borrowed spatula report isn't verified bakeware.

What to read before you parse the rest

Look for the regulation spelled out, the lab name and report number, a date, and results against limits — 177.2600 has mg/in² ceilings; BfR XV stacks have volatiles and sensory. If those four are missing, stop.

The specimen should name article (muffin tray, chocolate bar mold, ice tool), color, process (platinum RTV pour, not injection LSR or compression HCR), thickness, post-cure, cleansing. Process cousins fail verification even when the lab is real. An LSR plaque doesn't verify a platinum RTV Bundt. ISO 10993 doesn't verify food contact. A "FDA test" that is only a BPA non-detect is a carton badge.

Lot logic you can verify: compound batch, pigment, pour identity, post-cure oven chart, production wash, link to the report number. On-site, watch a post-cure load and compare it to the specimen note. If the report says 4 hours and the oven chart shows 40 minutes, the certificate is about a different article.

Common mix-up: renaming the PDF to "FDA"

Retailer portals that accept only a filename matching "FDA" train vendors to rename PDFs. In the shop that usually means the same cousin spatula flask with a new filename. Your job is to refuse the rename as proof.

What to ask when you're verifying

• Full PDF, methods, conditions, last page of numbers — not a cropped PASS cover?
• Specimen photo: does it match the cavity you sealed for the courier?
• Process: platinum RTV pour, tin-cure, or an LSR cousin?
• Have compound, color, pour, and post-cure changed since the test date?
• Care card: does it tell users to scour the face you just verified?

Worked example: red 24-cavity vs a 2014 natural plaque

A buyer is offered a 2014 natural spatula report for a 2026 red 24-cavity platinum RTV muffin. That's a cousin: wrong year, wrong color, wrong process family. Ask for 177.2600 and/or BfR XV on the production red cavity after production post-cure. Seal the specimen, photograph it, put the photo in the PO file next to the PDF.

When this article is the wrong quote

If you only needed a composition letter from the compounder, that's support, not verification of extractives — don't stop there, but don't treat this page as a reason to skip composition either. Non-food resin molds don't need this verification spend. BPA and phthalate screens in your own lab answer carton ghosts, not 177.2600.

Related reading: food-contact compliance, safety standards, food-grade FAQs, food-grade molds.

Verification checklist you can copy into the PO file

• Method named: 177.2600 and/or LFGB/BfR XV
• Lab, report number, date
• Specimen = this cavity, color, platinum RTV pour, post-cure
• Numbers vs limits; sensory if LFGB claimed
• Lot chain to compound, pigment, oven chart, wash
• Specimen photo matches
• Care line doesn't destroy the face

Send the PDF (full, not cropped) and a photo of the cavity you'll ship. We'll mark match or cousin and say what to rerun.

We would not recommend renaming the PDF to FDA

We would not recommend accepting a cropped PASS cover, a 2014 spatula report for a 2026 red RTV muffin, or a care card that tells bakeries to scour the face you just verified. We would not treat a 17025 logo as proof the silicone methods were in scope.

Read the specimen line first:

flowchart TD
  P[PDF] --> S{Specimen is this cavity color pour?}
  S -->|no| R[Cousin; reject]
  S -->|yes| M{Method named 177.2600 or BfR XV?}
  M -->|BPA only / metals only| R
  M -->|yes plus numbers| L[Match lot chain and photo]

Typical shop values: what a real report contains

| Item | Typical shop value | | --- | --- | | 177.2600 aqueous / fatty | 20 then 1 water; 175 then 4 n-hexane, mg/in² | | Mix / Shore | Platinum 1:1 or 10:1; 10A–20A plus jacket | | Shrink | ~0.1–0.3% platinum | | Date logic | No mystical two-year expiry; identity change restarts | | Cousin to reject | Decade-old natural LSR plaque for a new pigmented RTV tray |

Caption these as typical shop values. Verification is a reading exercise, not a filename.

FAQ

Is a 17025 logo enough to accept an LFGB PDF? No. Check that silicone food-contact methods are in scope and that the report ran volatiles, simulants, and sensory, not only a metal screen. Then match the specimen to your SKU.

How new must the report be? There's no universal expiry in 177.2600. Verify that compound, color, pour, and post-cure haven't changed. A decade-old spatula report is the wrong specimen for a new ice tool.

Can I verify by repeating BPA and phthalate screens in my own lab? Those screens answer carton ghosts from polycarbonate and PVC. They don't verify 177.2600 extractives or BfR XV sensory.

What if the vendor sends a composition disclosure instead of extractives? Composition support is necessary. It isn't the flask. You still need the finished-article result.

Does a verified report survive if we tell bakeries to use a green pad? The paper survives. The article doesn't. Abrasive that cuts the food-contact face is a new, scored surface.