TL;DR: A coating supplier’s COA is only as useful as the fields it actually contains — missing viscosity range, initiator type, or cure window data is a disqualification signal, not a minor paperwork gap.
TL;DR: In our incoming inspection protocol, UV coating lots with measured viscosity deviating more than ±15% from the COA nominal value are quarantined and returned — this threshold catches roughly one in eight lots from unqualified suppliers.
What Substandard Coating Lots Look Like Before They Hit Your Job #
Three symptoms show up repeatedly when a coating lot is going to cause production problems:
Symptom 1 — Cure hazing on flood UV. The coated surface looks milky or uneven under raking light, even though the lamp settings haven’t changed. On our production line, this almost always points to either under-cure (insufficient photoinitiator concentration) or a moisture-contaminated batch. Both can arrive in drums that look perfectly normal.
Symptom 2 — Adhesion failure at the score line. The coating lifts or micro-cracks within 5–10 open-close cycles on folding carton panels. This gets blamed on the die-cut operator or the board grade, but the root cause is usually coating elongation-at-break below 8%, which is borderline for any application that requires creasing.
Symptom 3 — Gloss inconsistency across the sheet. High-gloss UV jobs should land at 85–95 GU (60° geometry, per ASTM D523). If you’re seeing variance of more than ±8 GU across a single sheet, the coating rheology is unstable — likely from inconsistent raw material batching at the supplier.
Diagnostic decision table:
| Observed Symptom | Likely COA Field to Check | Pass Threshold | Action If Out of Range |
|---|---|---|---|
| Cure hazing / milky surface | Photoinitiator content (%) | Supplier-specified ±5% relative | Quarantine lot, request reformulation COA |
| Adhesion failure at score | Elongation-at-break | ≥ 8% (100 µm film, 23°C) | Reject; request alternative grade |
| Gloss variance > ±8 GU | Viscosity at 25°C (mPa·s) | COA nominal ±15% | Quarantine; retest before release |
| Yellowing under UV age test | Initiator type declaration | Benzophenone-free preferred for food-adj. | Escalate to materials team for risk review |
| Blocking in stack | Solid content (%) | Per product datasheet ±1.5% | Return lot; check storage temperature history |
The Root Cause Most Production Teams Misdiagnose: Photoinitiator Blending Variability #
When a UV coating lot cures inconsistently, the first response is almost always to increase lamp power. The machine operator bumps the cure energy from 120 mJ/cm² to 150 mJ/cm², the hazing clears up on the test sheet, and the job continues. The problem is logged as a press parameter issue and the coating lot is never questioned.
This is the wrong diagnosis more often than it gets credit for. Photoinitiator concentration in UV coatings is not a fixed formulation value — it is a blending variable that can drift between production batches at the supplier level, particularly with smaller coating manufacturers who don’t maintain automated dosing systems for their initiator premix. The mechanism works like this: the coating resin matrix (typically an aliphatic urethane acrylate oligomer) requires a specific molar ratio of initiator to achieve full network crosslinking within the cure window. If the initiator concentration drops by even 10–15% relative to nominal, the surface tack disappears under normal cure conditions but the internal polymer network remains partially uncrosslinked. This creates a coating that passes a fingernail scratch test on the line but fails rub resistance within 48–72 hours as the uncured oligomers migrate to the surface.
The confirmation method is not a press-side check — it requires lab measurement. We use a Fourier-transform infrared spectroscopy (FTIR) comparison between a reference-cured film (at specified energy) and the suspect lot’s film at the same energy. A conversion rate below 90% of the acrylate double bond absorbance peak (around 810 cm⁻¹) confirms under-cure attributable to initiator deficiency rather than lamp degradation. Lamp degradation shows a different FTIR pattern — it affects surface and depth cure uniformly, whereas initiator deficiency front-loads the under-cure at depth.
The measurement threshold we use internally: if double bond conversion is below 88% at the specified cure energy, we classify the lot as non-conforming under our QC-14 Coating Qualification Record and do not release it to production regardless of press-side appearance.
For food-adjacent packaging, this matters even more. FDA 21 CFR §176.170 places specific restrictions on the migration potential of unreacted photoinitiators. Under-cured coating on food-contact-adjacent surfaces is not a cosmetic problem — it is a compliance failure.
Corrective Actions Ranked by Impact and Feasibility #
-
Mandate a minimum COA field set before accepting any lot. This requires zero capital investment and stops non-conforming lots at the dock. Our standard COA request template (filed as Form QC-14A in our supplier onboarding pack) requires: viscosity at 25°C (mPa·s), solid content (%), photoinitiator type and concentration range, elongation-at-break (%), gloss potential at 80 µm wet film, and storage temperature limits. If a supplier cannot populate all six fields, we don’t qualify them. This catches the majority of substandard suppliers before a single drum enters the building.
-
Implement incoming viscosity testing on every lot. A Brookfield viscometer check takes under 10 minutes per drum and costs almost nothing. Pass/fail threshold: COA nominal ±15%. For high-viscosity gloss UV coatings (typically 800–1,200 mPa·s at 25°C), a reading above 1,380 mPa·s signals either cold-chain failure in transit or over-polymerization during storage.
-
Run a draw-down cure test before releasing each lot to the press. Apply the coating at 8–10 µm dry film thickness on the production substrate, cure at the specified energy (per ISO 2813 for gloss measurement), and check 60° gloss and rub resistance (minimum 50 cycles on a Sutherland rub tester without visible marring). This catches most initiator deficiency cases before they affect production. Adds roughly 2 hours to lot release, which is acceptable for print runs above 20,000 sheets.
-
Request FTIR cure conversion data from the supplier on each production batch. More expensive for the supplier to provide, so realistic only for strategic coating partnerships or high-volume accounts. This fixes the root cause at source — you’re asking the supplier to validate their own cure chemistry before shipping. Expect resistance from smaller suppliers. Make it a condition of preferred vendor status.
-
Qualify a backup coating supplier on the same product grade. Single-source dependency on specialty coating is a supply risk most brands don’t account for. Our standard is to maintain at least two qualified suppliers for any coating that runs above 50,000 m² per month. Qualification includes a 3-lot incoming test series and a 500-sheet press trial. Lead time to complete a new supplier qualification on our line is typically 6–8 weeks.
Prevention: What to Specify Upfront Before the First Drum Arrives #
Put these requirements in the supplier brief and the purchase order, not the inspection plan:
- Minimum COA fields (as listed in corrective action 1 above)
- Storage temperature range declared on shipping documentation (most UV coatings: 5–25°C, away from direct UV exposure)
- Shelf life from date of manufacture, not date of shipment — minimum 6 months remaining on arrival
- Compliance statement for REACH Regulation (EC) No 1907/2006 — required for any coating used on goods sold in the EU
- For food-adjacent applications: written confirmation of compliance with FDA 21 CFR §175.300 or EU 10/2011 as applicable
Request the Technical Data Sheet (TDS) and Safety Data Sheet (SDS) as separate documents from the COA. A supplier who bundles all three into a single document is usually working from a template, not from lot-specific test data.
Specification Notes for Brand Partners #
When you brief us on a UV or specialty coating requirement, the information that determines whether we can develop an accurate quote in one round rather than three is: the substrate type and surface energy (coated vs. uncoated board, foil laminate, or synthetic), the intended end-use environment (ambient retail, cold chain, food-adjacent or not), and any regional compliance requirements for the destination market.
The brief gap that causes the most sample iterations is unspecified gloss target. “High gloss” means different things depending on whether you’re comparing against a matte laminate or a satin base. Give us a GU reference value or a physical sample you want to match — without that, our first sample is a guess.
Our standard sampling timeline for UV coating trials is 10–14 working days from substrate receipt, assuming the coating is within our qualified range. If a new coating supplier needs to be qualified for your application, add 6–8 weeks. For specialty effects (soft-touch, anti-fingerprint, glitter-suspension UV), we request 15 working days minimum due to the draw-down and cure verification steps we run before committing substrate.
FAQ #
What fields must a UV coating COA contain for it to be acceptable for production use?
At minimum: viscosity at 25°C (mPa·s) with tolerance range, solid content (%), photoinitiator type and concentration, elongation-at-break (%), and storage temperature limits. A COA missing any of these fields doesn’t give us enough data to detect lot-to-lot drift, so we treat it as incomplete and request a reissue before the lot moves to incoming inspection.
If incoming viscosity tests within ±15% of the COA value, is the lot automatically released?
Viscosity is the first filter, not the only one. A lot that passes viscosity still goes through a draw-down cure test — 60° gloss check per ISO 2813 and a 50-cycle Sutherland rub test. A lot can have correct viscosity and still fail cure performance if the photoinitiator has degraded due to light or temperature exposure in transit. The lot release decision uses all three data points together.
Does a UV coating need REACH compliance documentation even if the final product isn’t sold in the EU?
It depends on your supply chain, not just the end market. If any component in the coating contains a Substance of Very High Concern (SVHC) listed under REACH, that becomes relevant as soon as the finished goods re-enter the EU distribution chain, even indirectly. We request REACH compliance statements from all coating suppliers by default — the risk of omitting it for a “non-EU” project and then having the product rerouted is real enough to make the paperwork worthwhile.
How many lots does it take to qualify a new coating supplier?
Our protocol requires a 3-lot incoming test series plus a 500-sheet press trial before a supplier is added to our Approved Vendor List (AVL). The 3-lot series tests for lot-to-lot consistency in viscosity, solid content, and cure performance — one lot is not enough to distinguish a good batch from a reliably good supplier. Total qualification timeline is 6–8 weeks depending on the supplier’s own production and shipping cycle.
Our current coating supplier provides COAs, but the values are always identical across every lot. Should that concern us?
Yes. Lot-to-lot variation in a real production environment will always show minor numeric differences — viscosity might vary ±3–5%, solid content ±0.5–1%. Identical values across multiple lots usually means the COA is being generated from a master template rather than from actual lot testing. Ask the supplier for the raw lab instrument data behind the COA values. If they can’t provide it, the COA is a document, not a test result.
Planning a packaging project? Contact our team to request a complimentary specification review and sample quote.
