Adhesives UV Curing & Surface Chemicals — Supplier Qualification Guide

COA Field Requirements: What to Demand, and Why the Standard Fields Are Not Enough #

Every Chinese supplier for UV adhesives and surface chemicals will send a COA. The question is whether that COA is measuring the parameters that predict application performance, or the parameters that are easy to pass.

The fields that appear on nearly every COA — appearance, color (Gardner scale), and pH — are the easiest to adjust post-production and the least predictive of cure behavior or bond integrity. Viscosity is a better signal, but only if you specify the measurement conditions: spindle type, RPM, temperature, and instrument model. A Brookfield LVT reading at 25°C/12 RPM is not directly comparable to a Brookfield DV-II reading at 23°C/6 RPM, even for the same material. Suppliers who do not specify measurement conditions on the COA are, in our experience, measuring opportunistically.

The fields we require in our internal SRM-04 supplier documentation checklist before any supplier enters qualified status:

Suppliers that resist providing method references for viscosity or NVC are flagging themselves. A COA without instrument conditions is not a COA — it is a product brochure with numbers on it.

For surface chemicals, the acid or amine value is often more diagnostic than anything else. We have received batches where the Gardner color was within spec but the acid value had shifted by 4–6 mg KOH/g from the approved sample — enough to change the cure kinetics of a downstream epoxy system by a measurable margin.

Failure Modes in Volume Supply: Where Qualified Suppliers Fall Apart #

This is where the real qualification work happens. Initial sample approval is a necessary gate, not a sufficient one.

The most common failure pattern we document is what we log internally as a “compounder drift” event: the supplier passes initial sample approval using a reactive diluent or photoinitiator from a Tier 1 Chinese raw material source, then switches to a lower-cost domestic alternative at volume. The COA continues to show values within spec range because the supplier is adjusting the photoinitiator loading to hit the same viscosity target — not the same cure performance. Viscosity stays within the ±15% acceptance window. Cure depth drops by 20–30%. The buyer discovers the problem at final assembly, not at incoming inspection.

The mechanism is subtle enough that standard incoming inspection (appearance + viscosity only) will not catch it. What catches it is a simple gel content spot-test: cure a 25 μm film under controlled UV dose (typically 1,000 mJ/cm² at 365 nm), extract in MEK for 1 minute, and measure weight loss. A properly cured UV adhesive from a stable formulation should show gel content above 85% by this method. In batches where a photoinitiator substitution has occurred, gel content routinely drops to 65–72%. The difference in application is cracking or delamination within 500–800 hours of UV/humidity aging — well inside most product warranty windows.

A second failure mode is specific to UV surface chemicals sold as primers or adhesion promoters. Silane coupling agent concentration in these products is almost never verified by buyers, because the COA typically reports it as “active content” without specifying the silane species. We have received batches labeled as γ-glycidoxypropyltrimethoxysilane (GPS) primer that contained a mixture of GPS and a lower-cost aminosilane. Adhesion to glass substrates tested acceptable in short-term lap shear, but hydrolytic durability under ASTM D1002 conditions (water soak at 40°C/7 days) dropped from >8 MPa to 3–4 MPa. Nothing on the COA predicted this. The only method that caught it was FTIR of the cured film, comparing against an approved reference spectrum.

Third pattern, and the least discussed: thermal stability of UV adhesives stored in Chinese warehouse conditions. Suppliers in Guangdong and Fujian provinces routinely store UV-cure materials in non-climate-controlled warehouses from April through September, where temperatures can reach 38–42°C. A 10°C increase in storage temperature approximately halves the shelf life of a peroxide-containing UV adhesive. We have received “within shelf life” material with viscosity 40–50% above the COA value and a yellowish tint that was absent on the approved sample. The material was not unusable, but the cure window had narrowed to a point where process consistency was unachievable at the customer’s line speed.

Does the GB/T Standard Cover What You Need? #

Directly: no, not for most export-specification UV adhesives or surface chemicals.

GB/T 2794 covers adhesive viscosity testing and GB/T 7124 covers tensile lap shear — both are technically equivalent to their ISO 2555 and ISO 4587 counterparts for basic mechanical properties. Where the gap opens is in cure performance. GB/T has no equivalent to ASTM D5965 (specific gravity of coating solids) or the gel content method described above. Chinese suppliers quoting GB/T compliance are typically referencing viscosity and appearance methods — not cure depth, not gel content, not hydrolytic durability.

The practical consequence is that a supplier can be fully GB/T compliant and still deliver material that fails your application specification. This is not a compliance gap on the supplier’s part. GB/T was not designed to cover the performance envelope that modern UV assembly adhesives require.

For REACH and EU No. 10/2011 food contact applications, the compliance picture is even more specific: Chinese suppliers will often reference REACH compliance without distinguishing between substances of very high concern (SVHC) declaration and full migration testing. An SVHC declaration is a starting point, not a clearance. If your application falls under EU food packaging or medical device adjacency requirements, request migration test data per EU No. 10/2011 separately — do not accept REACH declaration as a substitute.

Practical Guidance for Buyers #

When qualifying a Chinese UV adhesive or surface chemical supplier, the first document to request is not the TDS — it is three consecutive production batch COAs from the past six months, with lot numbers that can be cross-referenced to shipping records. Suppliers who cannot produce these, or who produce COAs where every batch shows identical values to two decimal places, are either not batch-testing or are adjusting reported values. Neither is acceptable for a production-volume qualification.

The risk scenario that consistently produces the highest downstream cost is photoinitiator underdosing, as described above. A lot that appears in-spec on viscosity but has been diluted 10–15% on photoinitiator loading will cure acceptably under your qualification conditions and fail under the slightly lower UV doses typical of production line variation. Request gel content verification (85% minimum, MEK extraction method, 1,000 mJ/cm² at 365 nm) on the first three incoming lots after approval, before withdrawing 100% incoming inspection.

For surface treatment chemicals and primers, insist on FTIR reference spectra as part of the approved sample documentation. This costs the supplier nothing if they have a QC lab, and it gives you an objective fingerprint for incoming lot verification. Suppliers that resist this — citing “proprietary formulation” — are not wrong about the IP concern, but there is a workable alternative: require FTIR of the cured film rather than the wet product. That protects their formulation while giving you the performance fingerprint that matters.

Volume commitment should not precede a minimum 3-lot incoming inspection program with full gel content and viscosity (conditions-specified) testing. For adhesives and UV curing materials at any annual spend above a moderate threshold, the incoming inspection cost is recovered within the first rejection event it prevents.

Frequently Asked Questions #

What viscosity tolerance should we accept on incoming lots versus the COA value?

Our working threshold is ±10% from the COA-stated value when the measurement conditions are fully specified (instrument, spindle, RPM, temperature). We tighten this to ±7% for UV adhesives used in precision dispensing applications, where nozzle pressure and bead diameter are sensitive to viscosity shift. The ±15% tolerance that some suppliers quote as “industry standard” is a formulation tolerance, not an application tolerance — the distinction matters when your dispensing equipment is calibrated to a specific viscosity window.

Is a REACH compliance declaration sufficient for UV curing materials going into EU markets?

It depends on the end application. For general industrial use where the cured adhesive is encapsulated and not in contact with food, skin, or medical devices, an SVHC declaration above 0.1% w/w threshold is usually the minimum required. For any food-adjacent packaging or medical device assembly, REACH declaration alone is not sufficient — you need migration data per EU No. 10/2011 or specific biocompatibility testing per ISO 10993 depending on classification. Chinese suppliers very rarely hold this data proactively.

Why do some Chinese UV adhesive suppliers refuse to provide FTIR reference spectra?

Formulation IP protection is the stated reason, and it is a legitimate concern. The workable position is to request FTIR of the fully cured film rather than the wet resin — cured film spectra reveal cure state and approximate crosslink density without exposing the raw formulation. Any supplier with a Perkin-Elmer or Shimadzu ATR-FTIR in their QC lab can produce this in under 30 minutes. Refusal at that point is a red flag.

What is the minimum sample size for a meaningful lot consistency evaluation?

Three consecutive production lots is a floor, not a ceiling. Based on the supplier qualification data we reviewed across 24 months, suppliers that showed acceptable consistency across three lots failed on the fourth or fifth lot in roughly one-third of cases. Five lots over a six-month window, with at least one lot from a different production month, gives a more reliable baseline. For high-volume or safety-critical applications, we recommend continuing quarterly spot-testing indefinitely — not just during initial qualification.

Can gel content testing be done in-house, or does it require a third-party lab?

In-house is entirely feasible. The MEK extraction method requires an analytical balance, a UV curing lamp with dose meter, a solvent-resistant container, and a drying oven. Total equipment cost is under $3,000 for a basic setup. The only variable that needs tight control is the UV dose — use an integrating radiometer, not time-based exposure. A dose variation of ±10% at 365 nm will shift gel content results by 3–5 percentage points, which is enough to generate false pass or false fail outcomes at the 85% threshold.

Published by sinoraw.com Technical Team | Request a sourcing consultation