What Is Food Grade Silicone? Rule, Limit, And The Report
Home
Yiwu, Zhejiang, China
Mon – Fri 9:00 AM – 6:00 PM · Sat 9:00 AM – 4:00 PM
[email protected]

What Is Food Grade Silicone: The Rule Behind The Label

Alan Xiao, Founder of Sourcing Pioneer
Alan Xiao Founder of Sourcing Pioneer
· Updated November 11, 2026 · 11 min read

Food grade silicone is silicone rubber made under a named food contact rule, then tested on the finished part for what it gives off.

The polymer is the same family either way. What changes is the cure system, the paperwork and the test results.

That is why two spatulas can look and feel alike and carry very different prices. One comes with a report that matches your market. The other comes with a word on a box.

This article covers three things. What the main markets ask for, what the numbers in a report mean, and six lines for an order sheet.

Studio still life of a cream silicone spatula and a small silicone mat on a beige surface under soft light
The part looks simple. The paperwork behind it is where the grade lives.

The Short Answer

Silicone rubber starts as a chain of silicon and oxygen atoms. Methyl groups hang off the sides of that chain, and short links are added so the chains join up.

Once joined, the material is an elastic solid. A fine mineral filler, usually fumed silica, gives it tear strength and shape.

None of that makes it food grade. Food grade is a claim about the finished part, made against a rule that applies in a named market.

Three markets cover most export orders. The United States uses 21 CFR 177.2600. The European Union uses a framework rule plus national measures. China now has a dedicated standard of its own.

Under all three, the question is the same. How much of the part can come out, into which food, at which temperature, over which time?

Two cure systems decide how much is left to come out. Platinum cure leaves no peroxide behind. Peroxide cure does, and the rest depends on the bake that follows.

So a buyer needs three things, not one. A test report on the part, a statement of the cure system, and the name of the rule.

The Words Are Not A Law

Here is a check anyone can run. Open the US rule and search for the phrase food grade. It is not there.

I read the full text of 21 CFR 177.2600 while writing this. The phrase appears zero times. The section is titled Rubber articles intended for repeated use, and it works by lists and limits.

First it lists the substances that may go into the rubber. Silicone elastomers with methyl groups are on that list, along with several other silicone types.

Then it sets a limit on what the finished article may give up. That is the whole mechanism. No certificate, no badge, no approval letter.

This matters in a quotation. A supplier who writes FDA approved silicone is using a phrase the FDA does not issue. It may still be a good part. But the words prove nothing, and the report is what does.

Europe works the same way, with different words. The framework rule sets a general duty rather than a badge.

China works by product standards, which is closer to a badge but still not one. The standard number on the report is the thing to check.

For the wider map of how these rules divide by material, there is a separate guide to food contact rules for kitchenware on this site.

What The Number Actually Is

The US rule splits food into two groups, watery and fatty, and gives each a limit. Both tests use reflux, which means boiling solvent running back over the sample.

Food groupSolventFirst 7 hoursNext 2 hours
Watery foodDistilled water20 mg per square inch1 mg per square inch
Fatty foodn-Hexane175 mg per square inch4 mg per square inch

Those four figures come straight from paragraphs (e) and (f) of the section. The test runs nine hours in total, and the second window has a much tighter limit than the first.

The reason for two windows is simple. A part that keeps giving off material after seven hours is not stable, even if the first result looked fine.

Now the trap. The US limit is per square inch. Most European and Chinese reports give a figure per square decimetre.

One square inch is 0.0645 square decimetres. So 20 mg per square inch is about 310 mg per square decimetre, and 175 becomes about 2,710.

Read that again, because it looks wrong. A US figure of 310 next to a European figure of 10 is not a scandal. The two numbers answer different questions, on different materials, in different units.

Comparing them directly is the fastest way to lose an argument with a supplier who knows the units. Do the conversion first, then compare like with like.

One more line in the same section is easy to miss. Finished articles must be cleaned thoroughly before first use. That is a duty on the seller, not a suggestion.

Two bars comparing the US extraction limits of 20 and 175 milligrams per square inch with their converted values of about 310 and 2,710 milligrams per square decimetre
The same limits in two units. Source: 21 CFR 177.2600(e) and (f).

Platinum Or Peroxide

Two chemical routes turn the same chain into a rubber, and the choice shows up in the report.

Peroxide cure uses a peroxide to join the chains. It is the older route, it costs less, and a mixed compound keeps for a long time before use.

It also cures at a lower temperature, which suits simple presses. The trade-offs are real. Sheets and tubes are less clear, the material often needs talc or a liner, and residues remain after curing.

Platinum cure adds one group across a double bond with a platinum catalyst. No peroxide is involved, so no peroxide residue is left behind.

A maker of both systems puts it plainly. Platinum cure leaves no peroxide residue, which can foul a mould, and the parts come out clearer with better tear strength.

The cost sits on the other side. Platinum compounds cost more, need higher curing temperatures, and have a short working life once mixed.

None of this makes peroxide cure illegal. The US rule lists three peroxide catalysts as allowed accelerators: benzoyl peroxide, 2,4-dichlorobenzoyl peroxide and dicumyl peroxide. Together they are capped at 1.5 percent by weight of the rubber.

It even carries one peroxide entry reserved for silicone crosslinking. So the legal question is settled. The practical question is what remains in the part and what the bake removes.

That bake is called a post-cure. It runs for hours above the working temperature, and its job is to drive off the short molecules left after curing.

A part with no post-cure can pass a smell test on the day it leaves the factory. Then it fails one a week later, inside a sealed bag. Ask for the post-cure condition, not just the cure type.

PointPlatinum curePeroxide cure
Residue after cureNo peroxide residueResidue, reduced by post-cure
ClarityClear to see throughTranslucent, not clear
Tear strengthHigherLower
Compound costHigherLower
Storage before useShort working life once mixedLong shelf life
Cure temperatureHigherLower
Release aidCan run without talcUsually needs talc or a liner

Sources for the table: a comparison published by a compound maker and a platinum grade data page. Both sell both systems, so neither has a reason to hide the downside.

For kitchen parts, platinum is the usual choice, and the same maker says so. It is also what most food and medical buyers expect, so a peroxide part needs a better price to win the order.

Two-column diagram comparing platinum cure and peroxide cure silicone, showing residue, clarity, cost and post-cure steps
Two routes, one chemical difference: what stays behind after the cure.

Europe Asks In A Different Way

The European Union does not publish a list of permitted silicone ingredients. It sets a duty, then leaves the detail to national rules.

The duty is in Article 3 of Regulation (EC) 1935/2004. A material must not transfer anything that could do three things.

It must not endanger health. It must not change the composition of the food in a way we would call unacceptable. It must not spoil the taste or the smell.

That third point is the one buyers forget. A silicone part can be chemically fine and still fail, because the food picked up a smell from it.

There is also a good manufacturing practice rule, Commission Regulation (EC) No 2023/2006, which requires documented controls along the line.

One warning on that number. At least one supplier blog prints it as 2023/2026. That regulation does not exist. Quote 2023/2006 in a specification and you sound like you have read it.

For silicone in practice, the test report usually leans on a national measure. The German BfR recommendation XV is the one European buyers name most often.

Labelling is the other half. Article 15 asks for the words for food contact, or a named use, or the glass and fork symbol. That label goes on the packaging of goods not yet in contact with food.

So the packaging carries the claim, not the part. A spatula with nothing printed on it can still meet the rule, as long as the carton says what it is for.

China Split Silicone Out

This is the part of the article with an expiry date on it. The Chinese rule for silicone changed in September 2026.

Until then, silicone rubber sat inside GB 4806.11, the standard for food contact rubber. That version is now abolished, with an end date of 2 September 2026.

The replacement is GB 4806.16-2025, a standard written for silicone alone. It was published on 25 September 2025 in Announcement No. 6 of 2025 and took effect on 2 September 2026.

Its scope covers materials and products made mainly of silicone rubber, formed when a silicone polymer is crosslinked with hydrophobic silica and similar substances.

Two changes matter to a buyer. The new standard adds a limit on volatile substances, and it adds an annex method for measuring them.

In plain words, the rule now looks at what evaporates out of the part, not only at what soaks out into food.

A certificate that cites only GB 4806.11-2016 is out of date as you read this. That is a five-second check on any quotation from a Chinese supplier.

The dates and the scope come from a testing lab’s summary of the announcement. A standards database entry records the old standard as abolished on the same date.

If a supplier argues that the old report is still fine, ask for the standard number on the report. Then ask which standard the buyer’s market will name at customs or at the retailer’s lab. The answer is rarely the old one.

What The Claims Translate To

Once you know the rule, the listing words stop being noise and become a checklist. Each one points at a document you can ask for.

Claim on the listingWhat it can meanWhat to ask for
100% food grade siliconeNo other polymer in the mixThe rule name and the report number
FDA approvedNothing, since no such approval existsExtraction results against 21 CFR 177.2600
LFGB approvedA test was run, not an approval grantedThe report, with limits and a pass line
BPA freeTrue of silicone in generalThe residue figure and the filler type
Heat resistant to 250 °CA material claim, not a use limitThe rated span on the maker’s data page
Medical gradeA different set of tests, for a different useWhich test, run on which product

Take the last row first, because it causes the most confusion. A part can meet a food rule and fail a medical test, or the reverse. The two are not grades on one ladder.

BPA deserves a line of its own. There is no bisphenol in the polymer, so the claim is true and tells you almost nothing.

The useful version of that question is different. Ask what the peroxide left behind, and what the post-cure removed. That answer changes with the factory.

Then there is the pinch test, which circulates in every buying group. Twist the part and look for a white streak.

A streak means filler, and plenty of filler means a cheaper compound. It is a fair check on filler load. It says nothing about what migrates into food.

Treat it as one signal, weighed with the report. A buyer who rejects a part on a pinch test alone will reject good parts and accept bad ones.

If you want the physical side done properly, ask how the part is moulded and post-cured. The press and the oven decide the surface you will feel.

Reading The Report

A report is only as useful as the six things listed on it. Check them in this order, and most weak offers fall away before the price is discussed.

One, the rule and its date. The report must name a rule you recognise, with a version year. A standard number without a year is a guess.

Two, the sample. Was it a sheet, a plaque, or the finished part? What thickness? The limit is per unit of food contact area, so the specimen shape decides how the result reads.

This is the most common trick, and it is usually not dishonest. A lab is sent a 2 mm sheet because that is what the moulder had. The order is for a 6 mm part with a different cure. Two different objects, one report.

Three, the simulant and the conditions. Water, hexane, olive oil and dilute acetic acid behave differently. So do four hours at 70 °C and nine hours at reflux.

Four, the limits and the margin. A report that only says pass hides how close the result sat to the line. Numbers let you see whether there is room for a colour change or a new filler.

Five, the lab and the report number. An accredited lab will put its mark and a number you can look up. An in-house sheet with a company stamp is a statement, not a test.

Six, the date against the recipe. A report older than the current compound proves nothing. Ask when the compound last changed, then look at the report date underneath the answer.

None of this takes long. Ten minutes with a report and a ruler will tell you more than an hour of email.

Where It Fails In Use

A part that passes the lab can still fail in a kitchen, and the reasons repeat.

Heat. One silicone maker lists minus 60 to 220 °C for its food contact kitchenware grades. Above the rated ceiling, the rubber hardens, and a hard silicone part cracks rather than flexes.

That number is worth carrying into a product meeting. An oven set to 250 °C is outside the span those grades state. A claim of 250 °C needs a grade data page behind it.

Cuts. Silicone resists heat better than it resists a knife. A nick in a scraper edge grows every time the tool is used, and the tear spreads from that point.

Fat and smell. Silicone can take up fats and aroma, especially in thin sections. A part used for tomato sauce may keep a smell that survives washing.

This is exactly why European buyers ask for a sensory test. The part is not dirty. It has simply absorbed something, and no amount of washing takes it out.

Washing. Hot water, detergent and a long soak age the surface. Over many cycles the surface can go tacky, and a tacky surface holds dust and food soil.

Printing. A decorated part has a second material on it. Ink in the food zone needs its own statement, and the safest design keeps print away from the area that touches food.

There is no legal service life, so set your own. Replace at a crack, at a chalky white bloom, or when a smell survives a wash. Write that into the care card and customers stop guessing.

Two other checks belong at goods-in. Weigh ten parts from one carton, because a compound change shows up as a weight shift before it shows up as a complaint.

Then flex one part in your hand and look at the surface. A dry, even surface with no sheen marks usually means a full post-cure.

Checklist card with the six order-sheet lines for food grade silicone, from rule name to printing
Six lines, one page. Each one closes a gap the word food grade leaves open.

Six Lines For The Order Sheet

These lines move the conversation from a name to a document. Add them to the specification before the sample is approved.

One, the market and the rule. Name the rule your market will ask about, with the year. The US rule, the European framework plus the national measure, or GB 4806.16-2025 for China.

Two, the cure system and the grade. Platinum or peroxide, plus the maker’s grade name. A grade name can be checked against a data page. Good silicone cannot.

Three, hardness and weight. Shore A hardness with a tolerance, and the part weight in grams. Two figures that catch a compound swap fast.

Four, the test on the finished part. Name the simulants, the conditions, the limits and the lab. Say that the report must describe the part, not a sheet.

Five, the post-cure. Ask for the bake temperature and the hours, and for the volatile result where the market asks for one.

Six, everything outside the food zone. Colour, print and the handle material. A silicone head on a steel shaft means the grade of that steel belongs on the same page.

All six fit on one page, which is the point. The one-page sheet the moulder works from is the format that makes a rule enforceable at the press.

Then let someone independent look at the run. An independent check before shipment with weight, hardness and surface on the checklist catches the change that happened after the sample was approved.

GRADE CHECK

Send me the test report or the quotation for a silicone part, and the market it will be sold in. I will check which rule it cites, whether the sample matches your part, and where the post-cure sits in the process. No cost, no obligation.

Check My Report

Questions Buyers Ask

Is food grade silicone the same as medical grade?

No. They are different test sets for different uses, and a part can pass one and fail the other. Ask which test was run, on which product, rather than which word is on the box.

Can I use it at 250 °C?

Check the grade, not the category. One maker’s food contact grades are rated from minus 60 to 220 °C. A 250 °C use sits above that span and needs its own data page.

Does food grade silicone contain BPA?

There is no bisphenol in the polymer, so the answer is no for any real silicone part. The useful question is what the cure left behind, which is where the report earns its keep.

Can I tell by pinching or twisting it?

A white streak shows filler, which points at a cheaper compound. It does not show what migrates into food, so use it as one signal and keep the report as the decision.

Why does one supplier send an FDA report and another an LFGB one?

Because the two answer different markets. Neither is better in the abstract. The question is which one your buyer, your retailer or your customs broker will name.

How long should a silicone kitchen part last?

No rule sets a life. In practice it ends at a crack, a chalky surface or a smell that survives washing. A care card with those three triggers is more useful than a year count.

Questions & Comments

We read every one and reply within 24 hours

If a certificate for a silicone part lands on your desk, send it over. I will look for the rule, the year and whether the sample was your part at all.

Subscribe
Notify of
guest
0 Comments
Most Voted
Newest Oldest
0
Would love your thoughts, please comment.x
()
x