Overview #
Selecting the wrong ink system for food-contact packaging is not a print quality problem — it is a regulatory liability that can trigger product recalls, customs holds, and brand damage. At UGI, we run food-contact compliant ink systems across our flexographic, gravure, and sheet-fed offset lines, and compliance starts at ink formulation approval, not at finished goods inspection. This guide covers the specific migration thresholds, regulatory frameworks, and QC parameters we apply when producing primary and secondary food packaging for brand partners in the US, EU, UK, and Australian markets. If your packaging will be in direct or functional-barrier contact with food, the ink specification decisions we make at job setup directly affect your regulatory exposure.
Migration Thresholds and Regulatory Frameworks #
The two frameworks that govern most of our brand partners’ markets are FDA 21 CFR (Parts 170–189 for indirect food additives) and EU Regulation 10/2011 on plastic food-contact materials — though for printed packaging, the EU framework most directly applicable is the Council of Europe Resolution AP(2005)2 on printing inks, alongside national implementations such as the Swiss Ordinance SR 817.023.21 (Swiss Ordinance on Materials and Articles), which remains the most operationally specific standard for printed food packaging in Europe.
The key migration limits we work to:
- Specific Migration Limit (SML): Individual regulated substances must not exceed 10 mg/kg food simulant under EU 10/2011 Article 11, with lower SMLs for specific substances (e.g., benzophenone photoinitiator: 0.6 mg/kg under EFSA guidance)
- Overall Migration Limit (OML): Total migration must not exceed 10 mg/dm² or 60 mg/kg food simulant (EU 10/2011 Article 12)
- Set-off / indirect migration: For non-direct-contact printed surfaces, we target a migration level below 10 ppb (0.01 mg/kg) into the food product, consistent with the Threshold of Toxicological Concern (TTC) framework under EFSA guidance
- FDA 21 CFR 176.170 / 176.180: For paper and paperboard in contact with aqueous and fatty foods, extractable limits apply per the specific food-type and temperature conditions defined in the regulation
| Regulatory Framework | Scope | Key Migration Limit | Photoinitiator Control |
|---|---|---|---|
| EU 10/2011 + AP(2005)2 | Printed food-contact packaging, EU market | OML 10 mg/dm² / SML 10 mg/kg | Benzophenone ≤ 0.6 mg/kg |
| FDA 21 CFR 176.170/176.180 | Paper/board food-contact, US market | Extractables per food type/temp | No specific PI limit; GRAS basis |
| Swiss Ordinance SR 817.023.21 | Printed packaging, CH + EU reference | Positive list of permitted substances | Explicit PI positive list |
| UK retained EU law (post-Brexit) | GB market | Mirrors EU 10/2011 at time of exit | Same as EU 10/2011 |
Ink System Selection and Formulation Parameters #
Not all low-migration (LM) inks are equivalent, and we do not treat “LM-certified by the ink supplier” as sufficient on its own. Our ink qualification process requires supplier documentation plus our own migration screening before a new ink system is approved for food-contact jobs.
UV-curable LM inks are our primary system for food-contact flexo and offset work. The critical parameter is cure energy: under-cured UV ink is the single most common cause of photoinitiator migration failures we see in third-party audits. On our UV flexo lines, we specify a minimum cure dose of 180 mJ/cm² (measured with a UV Power Puck II radiometer) for white and process colour layers, and 220 mJ/cm² for heavy coverage solids. Lamp output is verified at the start of each production shift — if output drops below 75% of rated power, the lamp is replaced before the job runs.
Electron beam (EB) curable inks eliminate photoinitiators entirely and are our recommendation for direct-contact flexible packaging where migration risk must be minimised. EB systems operate at accelerating voltages of 150–175 kV for typical flexible substrate thicknesses of 12–25 µm.
Water-based inks for flexo on corrugated and paper substrates: we specify residual solvent levels below 5 mg/m² on the printed surface, measured by GC headspace analysis per DIN EN 13130 methodology. Solvent retention above this threshold is a non-conformance trigger.
For conventional solvent-based gravure inks on non-direct-contact outer packaging, we maintain a functional barrier specification: the substrate between the printed surface and food must provide a migration barrier equivalent to ≥ 25 µm polyethylene or equivalent, verified by migration modelling under EU Regulation 10/2011 Annex V (migration modelling conditions).
Quality Control Parameters, Inspection System, and Non-Conformance Handling #
Our food-contact ink QC programme operates at three control points: incoming ink batch release, in-process cure/solvent verification, and finished goods migration screening.
Incoming ink batch release: Every ink batch for food-contact jobs is checked against the supplier’s Declaration of Compliance (DoC) and batch-specific Certificate of Analysis (CoA). We verify that all substances are either on the positive list (EU 10/2011 Annex I, Swiss Ordinance positive list) or supported by a toxicological evaluation meeting the EFSA CEF Panel assessment criteria. Batches without a valid DoC are quarantined — we do not run them.
In-process cure verification: UV cure energy is logged per reel/sheet batch. Any reading below the minimum threshold triggers an immediate line stop and re-cure assessment. We retain cure energy logs for 5 years as part of our food-contact production records.
Finished goods migration screening: For new ink/substrate combinations or new food-contact job types, we commission migration testing at an accredited third-party laboratory (ISO 17025 accredited) using food simulants per EU 10/2011 Annex III (simulant A: 10% ethanol for aqueous foods; simulant B: 3% acetic acid for acidic foods; simulant D2: vegetable oil for fatty foods). Test conditions are selected based on the intended food type and contact temperature.
| QC Parameter | Method | Acceptable Range | Non-Conformance Threshold |
|---|---|---|---|
| UV cure energy (flexo/offset) | UV Power Puck II radiometer | ≥ 180 mJ/cm² (colours), ≥ 220 mJ/cm² (solids) | < 180 mJ/cm² → line stop |
| Residual solvent (water-based flexo) | GC headspace, DIN EN 13130 | < 5 mg/m² | ≥ 5 mg/m² → reject/re-process |
| Overall migration (finished goods) | EU 10/2011 Annex III simulants | ≤ 10 mg/dm² | > 10 mg/dm² → batch hold, root cause |
| Benzophenone (specific PI) | LC-MS/MS | ≤ 0.6 mg/kg food simulant | > 0.6 mg/kg → batch reject, ink reformulation |
| Ink adhesion (tape test) | ASTM D3359 Method B | ≥ 4B rating | < 4B → surface treatment review |
| Colour register (food label) | Inline camera system | ± 0.2 mm | > 0.3 mm → operator alert, > 0.5 mm → stop |
Non-conformances at the migration screening stage trigger a formal CAPA (Corrective and Preventive Action) record. We do not release a food-contact job to shipment while a migration non-conformance CAPA is open.
Specification Notes for Brand Partners #
When you brief us on food-contact packaging, the first thing we need to know is the food type and contact conditions: is the packaging in direct contact with the food, or is there a functional barrier (e.g., an inner liner or pouch)? This determines whether we apply direct-contact migration limits or set-off/indirect migration controls — and it changes the ink system we specify.
The most common brief mistake we see is brands specifying “food-safe ink” without defining the food category or contact temperature. A UV LM ink approved for ambient dry food contact may not be appropriate for fatty foods or microwave-safe packaging — the simulant conditions and SMLs are different. We will always ask you to confirm the food type, contact duration, and maximum contact temperature before we finalise the ink specification.
Our standard process for food-contact packaging: ink system DoC and CoA review in 2–3 working days, digital colour proof in 3–5 working days, physical press proof with migration screening report in 15–20 working days (migration testing lead time is the constraint), production lead time 20–30 working days after approved sample and migration report. We provide you with the full migration test report, ink supplier DoCs, and our internal cure energy production log as part of the shipment documentation package.
Frequently Asked Questions #
Q1: What is the specific migration limit for benzophenone photoinitiator in EU food-contact packaging?
A: Under EFSA guidance and the AP(2005)2 framework, benzophenone must not exceed 0.6 mg/kg in the food simulant. This is one of the most frequently flagged photoinitiators in UV-cured food packaging — we control it by specifying LM ink formulations that exclude benzophenone and by verifying cure energy at a minimum of 180 mJ/cm² to prevent unreacted initiator migration from any residual PI compounds.
Q2: What is your MOQ and lead time for food-contact compliant printed packaging?
A: Our MOQ for food-contact flexo-printed flexible packaging starts at 50,000 units per SKU, and for sheet-fed offset food-contact cartons at 5,000 units. Production lead time is 20–30 working days after sample approval and migration test sign-off — the migration testing phase (15–20 working days) is the longest single step and cannot be compressed without compromising the regulatory validity of the results.
Q3: Does your ink system comply with FDA 21 CFR requirements for the US market?
A: Yes. For US-market food packaging, we work to FDA 21 CFR 176.170 and 176.180 for paper and paperboard substrates, and we require ink supplier documentation confirming that all ink components are either on the FDA indirect food additive positive list or supported by a GRAS (Generally Recognized As Safe) determination. We can provide the relevant supplier DoCs and our internal compliance checklist as part of the shipment documentation.
Q4: Can you print food-contact packaging with metallic or special-effect inks?
A: Yes, but metallic inks require additional scrutiny. Aluminium-based metallic inks are generally acceptable under EU 10/2011 with an SML of 1 mg/kg for aluminium, but copper and zinc-based metallic pigments have lower or restricted SMLs and we will not use them on direct-contact surfaces without specific toxicological clearance. We always review the pigment composition of any special-effect ink against the EU 10/2011 Annex I positive list before approving it for a food-contact job.
Q5: What happens if a finished goods migration test comes back above the 10 mg/dm² overall migration limit?
A: We place the batch on hold immediately and open a formal CAPA record. The root cause investigation covers three areas: ink batch CoA vs. actual cure energy log, substrate barrier properties, and food simulant test conditions vs. intended use. In our experience, the most common cause is cure energy drop due to lamp degradation — which is why we verify lamp output at the start of every shift and replace lamps below 75% rated power. No batch ships while the CAPA is open.
Planning a food-contact packaging project? Contact our team to request a complimentary specification review and sample quote.
© 2026 Ukugi.com. All rights reserved.
Unauthorized reproduction or distribution is prohibited.
Switching our flexo lines to low-migration UV inks (Flint Novaset range) added roughly $0.31/kg over conventional inks, but it eliminated the third-party migration testing we were running per SKU at around $400–600 per test. At our volume that’s a net save, especially when you’re qualifying 15–20 new food-contact SKUs a year.
Our Ningbo flexo ink supplier kept quoting “food-grade” on their spec sheets but couldn’t produce a migration test report to AP(2005)2 — just a GB 9685 cert, which doesn’t map cleanly to the benzophenone SML requirements we needed for an EU retail client. Took about four months and a third-party lab in Shenzhen before we had documentation we could actually put in front of the brand’s compliance team.
We had a benzophenone exceedance on a cereal liner last year — 0.84 mg/kg on migration testing against Tenax simulant, well over the 0.6 mg/kg SML. Switched to a fully PI-free UV system on that SKU and retested at <0.05 mg/kg. The Swiss Ordinance positive list catch was actually what flagged it before FDA did.
Had a situation with our Shanghai offset supplier last year where their ink formulator was fully certified to 21 CFR 176.180 but had never touched AP(2005)2 — they kept treating the Swiss Ordinance SR 817.023.21 as a stricter subset of the EU framework rather than a separate positive list obligation, which meant their substance declarations were missing compounds that don’t appear in EU 10/2011 at all. Took about six weeks and a third-party lab report from SGS Freiburg before the job setup was actually clean for our German retail customer.
On the sheet-fed offset lines specifically, how are you handling the set-off risk during stacking before the functional barrier coat is applied — is there a minimum curing dwell time you’re working to, or does that vary by substrate caliper?
Consolidating food-contact SKUs onto a single pre-approved ink set across flexo and gravure lines cut our per-SKU compliance overhead significantly — we went from running migration validation on 34 individual SKUs down to 6 ink family approvals, which dropped our annual third-party lab spend from roughly €28k to €9k.
Moved three pharma secondary packaging SKUs from laminated board to mono-material coated folding boxboard last year specifically to hit retailer recyclability requirements, but the aqueous barrier coating we needed to replace the PE layer kept failing the grease resistance spec at 35°C storage — ended up having to revalidate the whole food-contact migration profile under AP(2005)2 because the new coating supplier’s extractables data didn’t cover our temperature range.
The functional barrier thickness question is what keeps biting us on folding carton jobs — we validated our 15 gsm aqueous topcoat as a functional barrier on 350 gsm SBS, but when the brand switched to a lighter 280 gsm board to hit a cost target, the reduced caliper changed the coating absorption profile enough that our migration numbers on the printed reverse shifted outside acceptable range on two photoinitiators. Same ink set, same press settings, same coater. Board weight isn’t a variable most brand partners think to flag as a change control trigger.
Gravure vs. flexo behaves pretty differently on migration risk even when you’re running the same certified low-migration ink set — gravure’s higher ink lay-down weight (we typically see 3–5 g/m² vs. 1–2 g/m² on flexo) means your total extractables load is higher going into any AP(2005)2 screening, so a formulation that comfortably clears SML testing on a flexo substrate can come back marginal on the same job run gravure. We’ve had to maintain separate migration validation dossiers by process for that reason, which nobody tells you until you’re already in non-conformance.