TL;DR: A COA that lists only tensile strength and seal strength is not a complete food-contact qualification — you need OTR, WVTR, solvent residuals, and migration test results before any lot ships.
TL;DR: Our incoming inspection protocol rejects flexible laminate rolls if solvent residual exceeds 5 mg/m² total or 1 mg/m² for any single solvent — thresholds derived from GB/T 10004 and cross-checked against EU 10/2011 Annex guidelines.
What a Failing COA Looks Like — and the Three Symptoms That Signal It #
When a brand partner sends us a COA from their previous flexible packaging supplier and asks us to match or exceed the specification, the first thing we check is what the COA does not say. About half the COAs we review are structurally incomplete for food contact applications.
Three observable patterns flag a weak supplier qualification immediately:
Symptom 1: The COA lists physical properties only. Tensile strength, elongation at break, seal strength — these are the minimum. If OTR and WVTR values are absent, the supplier either did not test barrier performance lot-by-lot or tested once at sample approval and never again. For snack packaging where oxygen ingress drives rancidity, this is a critical gap.
Symptom 2: Print lot traceability is missing or vague. A COA that lists ink supplier as “approved vendor” without naming the ink system, pigment batch, or compliance basis (FDA 21 CFR 175.300 indirect contact, or EU 10/2011 compliance declaration) cannot support a brand’s regulatory file. We see this most often with spot colour runs where the main process inks have declarations but the Pantone match colours do not.
Symptom 3: Solvent residual data is either absent or expressed as a pass/fail checkbox. A checkbox labelled “within standard” tells the buyer nothing. The actual measured value per solvent species matters because different retained solvents carry different migration risk. Ethyl acetate at 3 mg/m² and toluene at 3 mg/m² are not equivalent hazards.
The table below maps these symptoms to their most likely root causes:
| Symptom | Most Likely Root Cause | Second Root Cause |
|---|---|---|
| No OTR/WVTR on COA | Barrier not tested per-lot; only sampled at approval | Lab capability absent; outsourced testing skipped to cut cost |
| Missing ink compliance declaration | Spot colour inks not covered by main compliance file | Supplier unaware of indirect food contact rules for reverse-printed laminates |
| Solvent residual as pass/fail only | Internal limit not defined; no per-species breakdown | GC testing done but result not included in customer-facing COA |
| No lot-level seal strength data | Seal parameters treated as fixed; no verification testing | Heat sealer calibration not tied to QC sign-off |
| Traceability gaps in substrate lot | Multiple sub-suppliers blended without segregation | Incoming substrate not logged against finished goods batch |
The Root Cause Teams Consistently Miss: Laminate Bond Strength Drift After Ageing #
Physical test results on a COA reflect the laminate at one point in time — typically 24 to 72 hours after lamination, once the adhesive has reached initial cure. Bond strength at 24 hours and bond strength at 7 days are not the same number, and for solventless lamination in particular, the delta can be significant.
Here is the mechanism. Solventless adhesive systems cure through a polyaddition reaction between isocyanate and polyol components. At 24 hours under standard cure conditions (40°C, controlled humidity), the reaction is roughly 70–80% complete. Full crosslink density is not reached until 5–7 days at ambient temperature. During that window, residual NCO groups are still reactive, and if the laminate is wound too tightly on the reel or stored in high humidity, two problems emerge simultaneously: the bond strength reading taken at 24 hours will overestimate the final stable peel force, and migration potential from unreacted monomers remains elevated.
We measure this using ASTM F88 seal and peel methodology, with T-peel on a 15mm strip cut perpendicular to the machine direction. Our acceptance threshold for PET/CPP laminate used in snack pouches is ≥3.5 N/15mm at full cure. Below 2.8 N/15mm at 48-hour test, we hold the lot and re-test at 96 hours before making a disposition decision. If the 96-hour result is still under 3.0 N/15mm, the roll is quarantined under what we internally track as a Grade B hold in our IQC-12 laminate deviation register.
The common misdiagnosis is blaming the adhesive batch. In roughly two-thirds of the Grade B holds we have logged over the past 18 months, the actual cause was substrate surface energy — specifically, PET film that arrived with dyne level at 38–40 mN/m when our specified minimum is 44 mN/m. At that surface energy, the adhesive wets out inadequately, and no amount of correct mix ratio or coat weight corrects what is fundamentally a corona treatment deficiency on the film substrate.
Measurement is straightforward. Dyne pens at 40, 44, and 48 mN/m on the incoming film surface, tested within 2 hours of unrolling. If 44 mN/m ink beads, the film is below specification. Do not proceed to lamination.
Corrective Actions, Ranked by Impact and Feasibility #
-
Redefine your COA field requirements in writing before qualification. Send the supplier a COA template — not a request, a template — listing every required field: OTR (cc/m²/day at 23°C, 0% RH), WVTR (g/m²/day at 38°C, 90% RH), solvent residual per species (mg/m²), ink compliance basis, adhesive system and coat weight, substrate dyne level at lamination, and peel strength at 72h cure. This eliminates ambiguity before the first production lot. Cost: near zero. This corrects roughly 60% of COA gaps we observe at qualification stage.
-
Add incoming dyne level testing to your IQC protocol. Standard dyne pens cost under $50 per set, and the test takes under 3 minutes per roll. For brands sourcing 3-ply laminates with reverse print, this single checkpoint catches the most common root cause of adhesion failure described above.
-
Require per-lot GC data for solvent residuals, not just at sample approval. This requires the supplier to have gas chromatography capability or a verified external lab relationship. If a supplier cannot provide per-lot GC data, that is a qualification red flag, not a negotiable gap. Our own solvent residual standard is ≤5 mg/m² total and ≤1 mg/m² toluene, aligned with GB/T 10004 and consistent with what EU 10/2011 framework guidance implies for indirect food contact.
-
Conduct an on-site adhesive handling audit. Solventless adhesive pot life is typically 60–90 minutes depending on ambient temperature. If the supplier’s production floor runs above 32°C in summer without temperature controls on the laminator feed zone, pot life shortens, mix ratio deviation risk increases, and cure quality becomes variable batch to batch. This audit requires a site visit or a trusted third-party audit firm — cost is real, but for a new supplier relationship involving food-contact film, it is a justified investment.
-
Run a full ISTA 2A transit simulation on the first three production lots. ISTA 2A covers compression, vibration, and drop — relevant for pouch-packed snacks shipped in corrugated master cases. This does not replace seal integrity or barrier testing but catches structural failures (delamination at pouch corners, zipper rail separation) that are invisible on a COA but visible after 14 days in a shipping container. Budget 3–5 working days for this test cycle.
Prevention — What to Specify Upfront #
Three items belong in the purchase order or supplier brief for every flexible snack packaging project, not the sampling phase. First: barrier specification tied to product shelf life — state target OTR and WVTR as a pass/fail acceptance limit, not a nominal target. Second: ink compliance basis — specify whether you need FDA 21 CFR 175.300, EU 10/2011 positive list compliance, or both, and name the print process (gravure, flexo, digital). Third: adhesive system type (solvent-based, solventless, water-based) and state that solvent residual GC data is a mandatory COA field.
The document to request from a new supplier before any qualification lot is their current raw material Approved Vendor List (AVL) cross-referenced against their ISO 22000 or equivalent food safety management certificate. If those two documents are inconsistent, that inconsistency is diagnostic.
Specification Notes for Brand Partners #
When you brief us on a flexible snack packaging project, the most useful document you can share upfront is the product’s target shelf life and the storage conditions it will face — ambient, chilled, or high-humidity retail. That single input drives our laminate structure recommendation more than pouch size or print complexity does.
The most common gap in new briefs is the absence of a migration compliance target. Brands will specify “food safe” without stating whether they need EU 10/2011 overall migration ≤10 mg/dm² compliance or FDA indirect food contact conformance. These are different documentation paths, and they affect which ink and adhesive systems are available within your budget range.
Our standard pre-production sampling timeline for a new flexible laminate structure is 18–22 working days from approved artwork and confirmed material specification — longer if the structure requires a new adhesive qualification. If you need a functional barrier layer (EVOH, AlOx coating), add 5–7 working days for barrier film sourcing confirmation.
FAQ
What specific fields must appear on a COA for flexible snack packaging to be considered food-contact complete?
At minimum: OTR (cc/m²/day), WVTR (g/m²/day), solvent residual per species (mg/m²), peel/bond strength at defined cure age, ink compliance declaration (citing FDA 21 CFR or EU 10/2011 basis), adhesive system and coat weight, and substrate lot number. A COA missing any of these fields is not complete for food contact qualification regardless of how many physical properties it does list.
Our current supplier says they test barrier performance quarterly — is that sufficient?
Quarterly testing may be acceptable for a stable, long-running structure with consistent raw material supply and a strong incoming inspection record. For a new supplier relationship or any structure with AlOx or metallised barrier layers, lot-level testing is the correct standard because barrier performance in these constructions is sensitive to handling, humidity during storage, and film age. Quarterly sampling cannot catch a single bad lot before it ships.
If bond strength fails at 48 hours, does that mean the whole laminate roll is rejected?
Not automatically. Our protocol holds the roll and re-tests at 96 hours. Bond strength below 2.8 N/15mm at 48 hours but above 3.0 N/15mm at 96 hours is a conditional pass with a note on the lot record — the laminate is usable but the 48-hour low result triggers a substrate dyne level review on the incoming film. Below 3.0 N/15mm at 96 hours is a hard reject.
Does reverse printing eliminate the need for ink migration compliance declarations?
No, and this is a premise worth examining. Reverse-printed laminates do place ink between film layers, which reduces direct food contact risk, but set-off migration through the laminate into food contact surfaces is still a documented pathway for some pigment and photoinitiator classes. EU 10/2011 and the associated EuPIA guidelines do not grant an automatic exemption for reverse print. Compliance declarations are still required.
What is a realistic incoming inspection failure rate for flexible packaging film from qualified suppliers?
Based on our incoming inspection log across 23 lots from qualified flexible film suppliers over the past 18 months, we see a lot-level hold rate of approximately 8–10% — mostly for dyne level out of spec or roll geometry issues (telescoping, core crush). Outright rejection on barrier or seal strength grounds runs below 3% of lots for established suppliers. For first-time or re-qualified suppliers, we apply 100% roll inspection on the first three lots before transitioning to skip-lot AQL 2.5 per ANSI/ASQ Z1.4.
Planning a packaging project? Contact our team to request a complimentary specification review and sample quote.
The pass/fail checkbox issue is something we dealt with for a long time — switched to requiring a full per-species GC breakdown on every COA in early 2023, and on the first three lots from our Guangzhou laminate supplier we found ethyl acetate sitting at 4.1 mg/m² while isopropanol was flagged clean. Combined residual looked fine as a single number but we were over our internal 1 mg/m² single-species threshold without knowing it.
Spot colour compliance gap is real — we caught it during a 2023 line extension audit where every process ink had a full EU 10/2011 declaration but three Pantone specials from the same supplier had nothing attached, not even a reference to the original approval file.
OTR lot-by-lot testing is the one we’ve had the most pushback on from converters — our current spec requires a maximum 35 cc/m²/day at 23°C/0%RH for a PET/PE laminate on snack pouches, and two suppliers in 2024 couldn’t demonstrate they had the equipment on-site to run it per-lot, only at initial qualification.
Switched our snack pouch structure from a PET/AL/PE tri-laminate to a mono-material PE last year and the COA gaps got worse, not better — the mono-PE converters we trialed were running thicker gauge to compensate for barrier loss and two of the three couldn’t provide lot-level WVTR data at all, which made RecyClass certification basically stalled for six months while we sorted out whether we even had a compliant structure.
Solvent residual profiles differ a lot depending on whether the converter is running solvent-based adhesive lamination versus solventless — we’ve had COAs from solventless lines where total residual came in under 1 mg/m² with no per-species breakdown provided because the converter assumed “basically zero” was sufficient documentation. It’s not, and GB/T 10004 doesn’t have a “trust us” clause.
We started requiring WVTR alongside OTR on every incoming COA after a Q1 2024 rejection where three rolls of a PET/PE structure tested fine on oxygen barrier but had moisture transmission nearly double our 3 g/m²/day limit — the converter had been testing OTR routinely but WVTR was still approval-stage only.
Print lot traceability is the one that bit us hardest — a Shenzhen supplier we’d been running with for about two years had clean COAs on everything physical, but when we triggered an FDA audit prep in late 2023 their ink compliance file covered the four process colours only and the two Pantone spot colours on our dark chocolate SKU had zero FDA 21 CFR 175.300 documentation. Took eleven weeks to get it resolved because the ink was sourced from a third-party pigment blender they hadn’t disclosed.
The 5 mg/m² total residual threshold is reasonable for most structures, but we’ve found it needs adjusting when the substrate includes a coextruded barrier layer with EVOH — the EVOH absorbs and then slowly off-gasses certain ketone solvents in a way that doesn’t show up on the incoming COA test but appears on migration testing done at 40°C/10 days per EN 1186. We had to tighten our internal limit to 3.5 mg/m² total for any laminate containing EVOH after a 2023 lot that cleared incoming inspection cleanly but failed migration screening six weeks later.