[email protected] +86 177 5921 5460 Reply within 1 business day
Fine Chemicals2026-11-15

Printing Ink Chemicals: Low VOC Does Not Mean Low Migration

GB 38507-2020 governs VOC in the ink can, GB 4806.14-2023 governs migration from the printed layer, and EuPIA plus the Swiss Ordinance govern the raw-material list. The three are not equivalent — changing one part of a formulation can trip another.

The most common mistake in ink formulation is treating 'compliance' as a single metric: switch to water-based or UV-curing ink, meet the VOC limit, and assume the product is fit for food packaging. In practice there are at least three regulatory layers, measuring the ink product, the printed ink layer, and the permitted raw-material list. Their samples, units and decision rules differ, so passing one layer proves nothing about the other two.

What each of the three frameworks actually measures

  • VOC content: GB 38507-2020 sets limits by ink type and substrate. The Ministry of Ecology and Environment handbook lists solvent gravure/flexo/screen inks at ≤75% and inkjet inks at ≤95%; water-based gravure at ≤15% on absorbent and ≤30% on non-absorbent substrates, water-based flexo at ≤5% and ≤25%; energy-curing offset at ≤2% and engraved gravure at ≤20%; sheet-fed offset and coldset web at ≤3%, heatset web at ≤10%.
  • Food-contact migration: GB 4806.14-2023, in force since 6 September 2024, divides inks into direct and indirect food contact and covers the ink and the printed ink layer it forms. General printing such as books, newspapers and stationery falls outside its scope.
  • Raw-material and industry lists: base materials and additives for direct-contact inks must comply with GB 2760 and related announcements; indirect-contact inks follow GB 9685 and colourant purity rules. Exports to Europe add the EuPIA exclusion list, Annex 10 of the Swiss Ordinance RS 817.023.21 and destination-country decrees.

For one ink, a VOC report, a migration report and a raw-material declaration are three separate documents. Collapsing them into the claim 'our ink is environmentally friendly' carries no weight in an audit.

Migration limits apply to the printed layer, not the ink can

GB 4806.14-2023 splits its general physico-chemical limits into two groups. The ink product is tested for five heavy-metal residues on dry ink weight — published interpretations give lead ≤10, mercury ≤10, cadmium ≤5, chromium ≤25 and arsenic ≤5 mg/kg. The printed ink layer is tested for overall migration ≤10 mg/dm², potassium permanganate consumption ≤10 mg/kg, heavy metals as Pb ≤1 mg/kg, and total primary aromatic amine migration not detectable at a 0.01 mg/kg detection limit. Migration testing follows GB 31604.1 and GB 5009.156; labelling must state the product class, suitable substrates and printing process.

The sample is the point: limits fall on the cured layer and its simulant, so degree of cure, film thickness, substrate barrier and printing conditions all change the result. One batch of ink can pass on one substrate and fail on another, or fail under a different drying regime — which is why formulation data cannot replace migration testing of the finished article.

Water-based and UV: VOC falls, the risk moves elsewhere

Water-based systems neutralise acrylic resins with ammonia or organic amines. VOC drops sharply, but neutralising amines, residual monomers, wetting agents, defoamers and dispersants become non-intentionally added substances (NIAS) that need assessment. The looser water-based limits on non-absorbent substrates are themselves a hint that drying and film formation are the hard part. UV and energy-curing systems push VOC down to the ≤2% range for offset, but the risk shifts to unreacted monomers and photoinitiator migration; under-dosing the cure shows up directly in migration results.

The correct sequence for a reformulation is therefore: fix the end use and substrate, look up the applicable limits, and only then select replacement resins, solvents or photoinitiators. Reversing that order tends to produce an ink that passes VOC and fails migration.

TPO reclassification: not an ink ban

The photoinitiator TPO (CAS 75980-60-8) was added to the ECHA SVHC candidate list on 14 June 2023 and reclassified as reproductive toxicity category 1B (H360Fd). EuPIA had already placed TPO on its exclusion list, meaning it is not recommended for printing inks where significant human exposure is likely. A frequently conflated fact: the EU banned TPO in cosmetics under Annex II from 1 September 2025 — that is a cosmetics prohibition, not a blanket ban on printing inks.

  • Formulation options include TPO-L, Omnirad 380 combined with Esacure KIP 150/100F, and high-molecular-weight low-migration Omnipol TP, aimed respectively at clear coatings and inkjet, offset and flexo, and sensitive applications.
  • Changing the photoinitiator means re-matching the spectrum: acylphosphine oxides absorb around 379–380 nm and pair with 385/395/405 nm UV-LED. Alternatives have different absorption bands, and an under-dosed cure creates both incomplete polymerisation and migration risk.
  • Beware regrettable substitution: moving to a structurally similar molecule that may later be classified the same way only defers the compliance cost. Yellowing, colour stability and cost also need re-validation.

Mineral oil: the migration path matters more than the content of the can

The French decree of 13 April 2022 limits mineral oils in packaging and printed matter distributed to the public: from 1 January 2023, MOAH (1–7 rings) ≤1% in the ink; from 1 January 2025, MOAH ≤0.1% or 3–7 ring compounds ≤1 ppm, plus MOSH (C16–C35) ≤0.1%. Annex 10 of the Swiss Ordinance RS 817.023.21 uses a positive list, and EuPIA has long recommended mineral-oil-free inks for food-contact applications.

Substitution is chemically uneven: the boiling and evaporation behaviour of vegetable oils makes them unsuitable as a direct replacement for mineral oils in heatset web inks. EuPIA's own note on the French decree states that, at the time of publication, no broadly applicable mineral-oil-free heatset ink meeting the 2025 requirements was available on the European market. Add the background contamination from recovered fibre and the design of functional barriers, and the only defensible conclusion is migration testing of the finished article plus barrier validation — not a declaration that the ink has been switched.

A checklist for formulators and buyers

  • Establish the end use first: food contact or not, direct or indirect, and whether the destination market is China, the EU, Switzerland or France.
  • Look up the specific GB 38507-2020 limit by ink type plus substrate; do not apply one blanket percentage.
  • Request three documents: VOC testing, a migration report to GB 31604.1 / GB 5009.156, and a raw-material compliance declaration with the photoinitiator list.
  • Validate cure dose and residuals for UV systems, NIAS and drying energy for water-based systems, and background plus barrier performance for recovered paper.
  • Write responsibility for reformulation after regulatory change into the supply agreement, so a standard update does not become unplanned rework.

Ink compliance is not about driving VOC down. It is about making three pieces of evidence — the can, the printed layer and the raw-material list — stand up at the same time.

Sources and definitions: GB 38507-2020 limits are taken from the Ministry of Ecology and Environment handbook on VOC control in metal packaging printing, cross-checked against public environmental impact assessments (grade A/B). GB 4806.14-2023 limits and the implementation date come from testing-institute interpretations of the National Health Commission announcement (grade B; verify against the current standard text before adoption). TPO classification and substitution routes come from ECHA-related notices, EuPIA guidance and public technical material from IGM Resins and FUJIFILM Ink Solutions (grade B). The French mineral-oil decree and the Swiss Ordinance come from EuPIA official notes (grade A/B). This article contains no output, market-size or price estimates.

Next step: check a sample, then decide between the standard report and custom research.

Cannot find your segment?

Tell us the product, geography and metrics you care about. We will scope size, growth and players to your definition and deliver research you can drop straight into a project file.