D4, D5 and D6 Siloxanes: The EU Question on Silicone Parts

The email arrives from your compliance officer two weeks before the first EU shipment, and it is one line long: "Can the factory confirm cyclosiloxane content?" You forward it to Shenzhen. What comes back is a scanned page from a raw-material supplier, in German, about a compound, dated three years ago, that does not mention your part. That exchange happens hundreds of times a month, and it happens because almost nobody on either end of it can say precisely what D4 D5 D6 siloxane restrictions on silicone parts imported into the EU actually require.

The short version: two separate legal mechanisms are in play, they do different things, and most buyers conflate them. One is a restriction on placing certain product categories on the market. The other is a communication and notification duty that attaches to articles regardless of category. A molded protective case usually escapes the first and usually walks straight into the second — and the factory process step that decides your exposure is not the choice of base polymer but the oven schedule after the press.

What the three substances actually are

D4, D5 and D6 are cyclic siloxanes — ring-shaped small molecules made of alternating silicon and oxygen atoms with methyl groups attached:

• D4 — octamethylcyclotetrasiloxane, a four-unit ring
• D5 — decamethylcyclopentasiloxane, a five-unit ring
• D6 — dodecamethylcyclohexasiloxane, a six-unit ring

The first thing to understand is that nobody adds these to your protective case. They are residuals. Polydimethylsiloxane polymer is produced through routes that pass through, or equilibrate with, these cyclic species, and a cured elastomer retains some fraction of them trapped in the network. Every silicone elastomer contains some. The question is never whether they are present, only how much, and how much is a function of the polymer's starting level and what the factory did afterwards.

That distinction changes how you interrogate a supplier. Asking "is your silicone free of D4, D5 and D6?" invites a yes that means nothing. Asking "what is the residual cyclosiloxane content of the cured, post-cured part, by what method, and can you show me the chromatogram?" invites either a document or a silence, both of which are useful.

The two instruments, kept apart

Mechanism one: the Annex XVII restriction. REACH Annex XVII carries restriction entries covering D4, D5 and D6. The original entry addressed D4 and D5 in wash-off cosmetic products; the scope was subsequently extended to cover D4, D5 and D6 in further categories, with a concentration limit expressed as a weight percentage. The critical property of a restriction entry is that it is written around specified uses and product categories — it prohibits placing those categories on the market above the limit. A silicone protective case for a payment terminal is generally not one of the enumerated categories. Read the current entry text before you rely on that, because these entries have been amended more than once and the category list is the whole of the question.

Mechanism two: the Candidate List, Article 33 and SCIP. D4, D5 and D6 have been identified as substances of very high concern and placed on the REACH Candidate List. That listing triggers duties that apply to articles, not to product categories:

• Article 33(1): a supplier of an article containing a Candidate List substance above 0.1% by weight must give recipients sufficient information for safe use, including at minimum the substance name.
• Article 33(2): on request from a consumer, the same information within 45 days, free of charge.
• SCIP notification: under the Waste Framework Directive, articles above the same threshold placed on the EU market are notified to ECHA's SCIP database.

This is the mechanism that reaches a device case, and it is the one your compliance officer is really asking about. The threshold is calculated on the article — and where an object is assembled from several distinct articles, ECHA's position has been that the threshold applies to each incorporated article rather than being diluted across the whole object. For a single-piece molded case that nuance rarely bites; for a case with a bonded silicone body, a PC frame and a lanyard it can.

The right combination is all four, in that order of increasing weight. A declaration without an analytical report behind it is a promise; an analytical report without a declaration leaves you without a contractual remedy when it turns out to have been run on a different part.

Who actually carries the duty

A point that saves arguments: the Article 33 and SCIP obligations fall on the EU-side actor placing the article on the market — the importer or the EU distributor — not on a Chinese molder, which sits outside the territorial scope of REACH. The factory cannot discharge your obligation for you, and no declaration it signs transfers liability.

What the factory can do is supply the inputs you need to discharge it: composition information, the compound supplier's data, the process description, and test specimens. Structuring the relationship that way — you own the obligation, they own the data supply — produces far better cooperation than demanding that an overseas plant "be REACH compliant", a phrase that has no operational meaning for a non-EU manufacturer. The compliance checklist by market sets out how these duties allocate across the common destinations, and the China import compliance requirements guide covers the documentary side.

Why this is a cure-and-bake problem, not a polymer problem

Here is the technical core, and it is the reason a buyer who understands it gets straight answers from a factory that a buyer who does not will never get.

Residual cyclosiloxanes leave a cured part by volatilisation and diffusion. Heat raises their vapour pressure and speeds their migration to the surface; moving air carries them away. That is the entire mechanism of post-cure — the secondary bake in a hot-air-circulating oven, typically a few hours at a couple of hundred degrees Celsius depending on the compound, that a properly run silicone line applies after pressing. Compound datasheets carry the specific schedule; the figures here are the general range, not a specification.

The cure system matters here less than buyers expect. Peroxide and platinum addition cure leave different other residues — peroxide decomposition byproducts are a real and separate issue — but the cyclosiloxanes come from the polymer itself, so switching cure chemistry does not eliminate them. Post-cure is what moves that number; the wider trade-offs between cure routes and molding processes are set out in the compression molding versus LSR injection comparison.

Three consequences that experienced buyers use as diagnostic questions:

D6 is harder to remove than D4. Boiling point rises with ring size across the series, so under a given oven schedule the smaller rings leave first. A part can therefore come back with a low D4 result and a materially higher D6 result — which is a signature of a bake that was too short or too cool rather than of a contaminated compound. If a supplier's report shows D4 and D5 near the detection limit and no D6 line at all, ask why D6 was not in the scope.

Thick sections lag thin ones badly. The controlling variable is diffusion path length. A 1.5 mm slip-on skin and a 6 mm reinforced corner on the same part are not the same problem, and a schedule proven on a thin test plaque does not prove the corner. Rugged device cases are precisely the geometry where this bites, because the whole design premise is thick, energy-absorbing sections at the corners.

An unvented oven can make things worse. If the oven has no meaningful air exchange, driven-off volatiles saturate the atmosphere and recondense on cooler surfaces — including on parts at the back of the rack and on parts loaded later in the cycle. A densely packed, closed-damper oven is not a post-cure; it is a redistribution step. When somebody audits a silicone workshop, the exhaust duct and the rack spacing tell you more than the temperature controller does.

None of that is exotic. It is what any competent compression-molding plant already knows, and the reason to know it yourself is that it converts a vague compliance question into three specific, checkable process questions. WJM Silicone's published capability text describes temperature- and pressure-controlled silicone compression molding, and lists both food-grade and industrial silicone grades along with high-temperature and UV-resistant grades for automotive interiors — plausible ground for this conversation, and roughly 80% of its output is stated to go to the Americas and Europe. What the published material does not contain is any post-cure schedule, any analytical report, or any SVHC declaration. The factory's own credential text says certificate details are pending verification. So treat all of it as the starting point for a document request rather than as an answer.

Reading the analytical report properly

If a report does arrive, four things determine whether it is worth anything, and each is a place where reports quietly fail:

1. The method. Headspace GC-MS and solvent-extraction GC-MS answer different questions. Headspace measures what comes off under defined conditions; extraction measures what a solvent can pull out of the matrix, and the number depends on solvent choice, temperature and duration. Two laboratories using two methods on one part will not agree, and neither is wrong. Specify the method in the purchase order so results are comparable across lots.
2. The specimen description. Compound grade, colour, thickness, and whether the specimen was post-cured. A report on a natural translucent 2 mm plaque is not a report on your black 5 mm case.
3. The reporting basis. Individual results for D4, D5 and D6 separately, with the limit of quantification stated. A single "total cyclosiloxanes" figure hides the D6 signature described above.
4. Dates and traceability. Sampling date, lot identification, and a link back to a production lot rather than to an unidentified sample. If the report cannot be traced to a lot, it cannot be used to police future lots. The general guidance on reading a China certificate of conformity applies directly here.

Common questions

Does the restriction ban silicone protective cases in the EU?

The Annex XVII entries are category-scoped, and a molded device case is generally not within the enumerated categories — but the Candidate List communication and notification duties apply to articles across the board above the 0.1% w/w threshold. So the practical answer for most buyers is: not a market ban, but a documentation and notification obligation you have to be able to satisfy. Confirm the current entry text before relying on this.

Can a factory guarantee zero D4, D5 and D6?

Treat any such statement as a reason to slow down rather than speed up. Every silicone elastomer carries some residual cyclics; the meaningful claim is a measured content below a stated limit of quantification, by a named method, on a described specimen. A supplier who offers "zero" has generally not run the test.

How does this interact with food-contact testing?

They overlap in the mechanism and diverge in the paperwork. Both are improved by the same post-cure step, and both are scoped to a specific compound, colour and geometry — so one production-condition lot can feed both test programmes. The food-contact side, and why a part can pass a US extraction test and fail a German volatiles criterion, is covered in the FDA and LFGB piece.

What to ask the supplier next

Send this before the first EU-bound purchase order:

1. A signed SVHC declaration naming D4, D5 and D6 explicitly, stating content against the 0.1% w/w threshold, dated, on letterhead, signed by a named person with a title.
2. The compound trade name, grade and supplier, with the raw-material supplier's own REACH statement attached.
3. The post-cure schedule: oven type, setpoint, dwell, air exchange rate or damper position, rack loading density, and whether the schedule differs by part thickness.
4. A third-party analytical report with D4, D5 and D6 reported individually, the method named, the limit of quantification stated, and the specimen described by compound, colour and thickness — or a quotation to produce one.
5. Confirmation that specimens will come from a production-condition lot, not a hand-baked sample batch.
6. A change-notification undertaking covering compound, pigment, cure system and post-cure schedule.
7. Agreement on who pays for periodic retest and at what interval, written into the supply agreement.

That list works as a screening tool as much as a document request: how a factory responds to item 3 tells you within one reply whether its silicone process is controlled or improvised. WJM Silicone, a Longgang compression molder with a stated Americas-and-Europe export base, is a fair example to run it against — you can review the scope on the WJM Silicone factory profile, then ask for items 1 through 7 in writing and judge the answers on the attachments rather than the assurances.