Migration Compliant, Not 'Low Migration': Choosing Photoinitiators for Food-Contact Packaging

Last reviewed: 18 September 2026. Migration rules differ by market and change. Nothing here is legal advice; use it to know which questions to ask and which documents to demand.

A formulator asked us recently which of our photoinitiators was the low-migration one. It was a reasonable question and we gave a slightly awkward answer: the term has been used so loosely in this market that it no longer tells you anything, and the honest reply is a set of questions rather than a product name. This article is that answer, written down.

Why "low migration" has stopped meaning much

Sun Chemical has documented the problem publicly, and their conclusion is worth repeating: the phrase has been so overused by suppliers that EuPIA now recommends the term "migration compliant" instead. The reasoning is straightforward. "Low migration" is a comparative with no reference point. Low compared to what, measured how, on whose construction?

"Migration compliant" is narrower and more useful. It asserts one thing: that migration from a specified construction has been measured and falls within the limit that applies to the intended use and market. That is a claim you can check. "Low migration" is not.

So the first practical step is a vocabulary one. When a supplier tells you a photoinitiator is low-migration, ask three follow-ups: compliant with which limit, in which market, measured on what construction? A supplier who cannot answer those is not necessarily dishonest. They may simply not know, because the answer does not live in the photoinitiator.

Migration is a property of the construction, not the molecule

This is the part that catches people out. A photoinitiator does not have a migration value. Only a finished article does.

The measured migration from a printed label into a food simulant depends on at least five things, and the initiator is only one of them:

The molecular weight and shape of the initiator, because larger and bulkier molecules diffuse more slowly through a polymer matrix. This is the only variable the initiator choice controls directly.

The film weight of the ink or coating. A thicker film holds more of everything, including whatever migrates.

The substrate and whether a functional barrier sits between the ink and the food. A barrier can change the answer by orders of magnitude.

The cure conditions. Under-cured films retain unreacted photoinitiator that a properly cured film would have consumed, and residual monomer alongside it.

The time and temperature of contact, which is what the OM conditions in EU testing are for. A construction that passes at 10 days at 20 °C may fail at 2 hours at 121 °C.

Only the first is a raw material question. The rest belong to the converter, the printer and the brand owner. That is why a photoinitiator supplier cannot honestly promise you food-contact compliance on their own data sheet.

What a real failure looks like

This is not theoretical. In 2013 the CVUA Stuttgart laboratory reported a case involving Halloumi cheese. 4-methylbenzophenone was found at 843 µg/kg and Irgacure 184 at 137 µg/kg, both having migrated from a printed label, through a plastic film, into the food.

Two things are worth taking from that. The first is that migration through a film is a measured reality rather than a caution. The second is that 184 was involved. In the current regulatory conversation 184 is the well-behaved grade: no harmonised classification, no restriction, no Candidate List entry. It migrated anyway, from a label, through packaging, because the construction allowed it. A clean substance file does not protect a bad construction.

The route that applies is not the one most people name

Ask which regulation governs a photoinitiator in a printing ink and most people will say EU 10/2011. That answer is usually wrong, and the reason is worth understanding.

Regulation (EU) No 10/2011 governs plastic materials and articles. Article 2(3) states that it applies without prejudice to Union or national provisions applicable to substances in adhesives, coatings and printing inks, and Article 6 leaves colorants, solvents and polymer production aids to national law.

Printing inks are therefore a national-law matter in the EU. A photoinitiator absent from the 10/2011 Annex I positive list is not in breach of 10/2011. It is outside its scope. (For completeness: only one photoinitiator in common use appears on that list at all, benzophenone, as FCM substance No. 286, with a specific migration limit of 0.6 mg/kg.)

What actually decides the question is the national list for your market. Two matter most:

Market	Instrument	Which photoinitiators appear
Switzerland	SR 817.023.21, Annex 10 (v3.5, in force 1 Aug 2026)	369 at 0.15 mg/kg; 379, ITX, 4-MBP and TPO at 0.05 mg/kg; benzophenone at 0.6 mg/kg as a sum with the methylbenzophenones
Germany	Bedarfsgegenständeverordnung, Anlage 14, category V	ITX and 4-MBP only

Two consequences fall out of that table, and both surprise people.

The first is that TPO is still positively listed in Switzerland. It is an SVHC and carries a harmonised Repr. 1B classification, and it was removed from the EuPIA suitability list in January 2026, but Swiss law still permits it at 0.05 mg/kg. Anyone who tells you TPO is banned in food packaging is describing the voluntary list, not the law. Switzerland is the market where that distinction is easiest to see.

The second is that a CMR substance can perfectly well be on a positive list. Swiss Annex 10 contains vinyl chloride, ethylene oxide, acrylamide and formaldehyde. The CMR exclusion bars a substance from the non-listed-substance route and from the functional-barrier exemption. It does not bar it from being listed. That is why 369 and 379 can be listed while both carry Repr. 1B classifications, and it is why the German position for those substances is so much harder: they are absent from Anlage 14 and CMR, so both routes close at once.

A way to choose

Four questions, in this order. Answer them before you look at a data sheet.

What is the market and the end use? Food contact, indirect contact, or neither? Which country? This decides which list applies, and there is no point comparing products until it is settled.

Is there a functional barrier? If yes, the migration question may be largely answered by the construction, and your initiator choice is freer. If no, the initiator is doing the work and the choice matters much more.

What is the film weight and the cure? Thin, well-cured films migrate less. If you are running a heavy ink layer or a marginal cure, any initiator will look worse in testing than its data sheet suggests.

Only then, which chemistry? The structural principle is real: a bigger molecule diffuses more slowly, and a molecule with two initiating groups needs less of it present to achieve a given cure. That is the reasoning behind polymeric and bifunctional grades. It is a good reason to test them. It is not a substitute for testing.

What to demand, and from whom

The document that matters in this market is the Declaration of Compliance. Swiss Article 35a makes one mandatory at all marketing stages except retail from 1 February 2026, and brand owners increasingly require it elsewhere.

Here is the uncomfortable part, and it is a compliance consultancy's own summary rather than ours:

If you buy food-contact materials from manufacturers outside the EU, you cannot expect an EU FCM Declaration of Compliance. Most suppliers outside the EU are not aware of the requirement.

That is a real gap, and it runs in a specific direction. A Chinese ink manufacturer can supply a technically excellent product and still leave its European customer unable to complete their own paperwork. If you are buying from outside the EU, ask for the DoC before you ask for the price. If the answer is that they do not issue one, that is information about the supplier, and it will not improve after the order is placed.

Alongside the DoC, ask for the migration test report on a construction equivalent to yours. Not a generic report on the raw material. A report on a film of comparable weight, on a comparable substrate, cured comparably, tested against the simulant and conditions of your intended use.

What we will and will not claim

We supply polymeric and bifunctional photoinitiators built for this part of the market, including grades equivalent to Esacure ONE and KIP 150, and a bifunctional grade equivalent to Omnirad 127. We will tell you which of our grades is used in food-contact work most often, and why.

The grades most often specified for this kind of work are PPI-ONE and PPI-101 (Esacure ONE / KIP 150 chemistry, in powder and liquid form respectively), NPI-50104, NPI-701 and NPI-20400. On the trade-code side, ITX, 4-MBP and TPO are the three that appear on the Swiss list, each with its own limit and its own regulatory file.

What we will not do is tell you that any of them is "low migration" and leave it there. We will ask about your construction, tell you what the applicable list says for your market, and tell you what testing we can support. Where we do not hold data for your specific construction, we will say so rather than let you infer compliance from a data sheet.

If you are at the stage of choosing a grade and want a shortlist, send us the market, the substrate, the film weight and the intended contact conditions. That is enough for a useful answer. Without it, any recommendation would be guesswork dressed as expertise.