Replacing UV-328: How to Re-Specify a UV Absorber for Plastics, Rubber and Coatings

UV-328 (CAS 25973-55-1, C22H29N3O, MW 351.5) was the workhorse benzotriazole ultraviolet absorber for polyolefins, PVC, polyurethanes, coatings and adhesives for roughly three decades. It is now being eliminated worldwide, and what has taken its place in the catalogue is not one molecule but four chemistry classes with different absorption ranges, volatilities and price points. That is why the second half of this note is a purchase specification rather than a product recommendation.

The dates that matter

UV-328 is persistent, bioaccumulative and toxic and has been on the REACH candidate list since 2014; the POPs listing moved it from a European authorisation problem to a global supply problem. Two practical consequences follow. First, the 1 mg/kg step in 2029 is the one that changes buying behaviour - at 100 mg/kg a supplier declaration plus a screening method is usually enough, while at 1 mg/kg you are in HPLC-MS/MS territory and you are testing the finished article, not only the additive, because UV-328 can arrive through a masterbatch carrier, a recycled stream or a coating applied further down the chain. Second, the exemptions are narrow and time-limited. If your customer is in automotive exterior trim or in industrial coatings the exemption may cover conversion today and will not cover the life of the tooling.

The four chemistry classes you can actually buy

Read the table as a set of trade-offs rather than a ranking. The benzotriazoles and the hydroxyphenyl triazines are the closest structural descendants of UV-328 and cover the full ultraviolet range that matters for polymer yellowing. The benzophenones and the hindered hydroxybenzoate are cheaper or multifunctional, and both leave a gap in long-wave UVA.

Molecular weight is the first filter, not the last

A UV absorber only works while it stays where it was put. Volatility takes it out through the die during extrusion, migration takes it to the surface of a thin film, and extraction takes it into whatever the article touches. Molecular weight is the cheapest single predictor of all three, and the spread in this list is wide - from 274 g/mol for the benzophenone to 659 g/mol for the bis-benzotriazole, a factor of 2.4 on a single specification line.

• Thin sections are the hard case. A 30 micron film gives a migrating molecule a very short path to the surface, so a grade that performs well in a 2 mm injection moulding can fail in a film. Above roughly 400 g/mol is a reasonable working floor for thin, long-life or food-contact-adjacent articles, and it should be confirmed with a migration test in the relevant simulant rather than argued from the structure.
• Volatility is a separate measurement. Two absorbers at the same molecular weight can differ by an order of magnitude in weight loss at 300 C, because volatility is driven by the shape of the molecule as much as its mass. For any process above 250 C, ask for thermogravimetric data: UV-1577, for example, is quoted at 1% weight loss at 300 C and 5% at 335 C, which is why it holds up in polycarbonate and polyester extrusion where a smaller triazine would not.
• Fogging is volatility seen from the customer side. In an automotive interior, volatile additive condenses on the windscreen. The number to request is a gravimetric fogging value or a 100 C/24 h weight loss, and the answer decides the grade long before the UV spectrum does.

Do not swap by name - swap by spectrum

The single most common error in a UV-328 replacement is choosing "another UV absorber" with the right price and the wrong absorption range. UV-328 covered roughly 290-400 nm. A benzophenone such as 203B covers 290-360 nm, and a hindered hydroxybenzoate such as UV-2908 covers 270-340 nm. Both are excellent products in their applications, and both leave the 340-400 nm window less protected - which is precisely the long-wave UVA region that drives yellowing, gloss loss and chalking in outdoor polyolefins and in coatings.

It follows that a cross-class substitution is a formulation change, not a like-for-like replacement, and it should be qualified as one. The cheap test is a UV-vis scan: dissolve the candidate at the intended loading in a solvent, measure absorbance across 280-420 nm, and compare against the curve the old formulation produced. The better test is the same scan on a film or plaque of the actual thickness. If the 380-400 nm shoulder has gone, the formulation has lost protection and no amount of extra dosage restores the shape of the curve.

The pairing question: an absorber is not a HALS

UV absorbers screen photons; hindered amine light stabilizers scavenge the radicals that form anyway. UV-328 was almost always used alongside a HALS, and replacing it means re-checking that pairing rather than assuming it still holds. Two interactions matter. Conventional HALS are basic and are deactivated by acidic co-additives - including some phenolic antioxidants, acid scavengers and flame-retardant residues - so a formulation change that alters the acid balance can switch a HALS off silently, and a low-basicity HALS grade exists for exactly those systems. Conversely, the hindering phenol group in UV-2908 makes it a radical scavenger as well as a screen, which is why it is specified for flame-retarded and thioester-stabilised compounds where a plain absorber would fight the rest of the package. For the HALS side of the decision, including the low-basicity grades and their loadings, see our guide to HALS light stabilizers.

Selection by application

• Polyolefin film and thin-wall packaging. Migration and extraction resistance matter more than peak absorbance, so work at the high-molecular-weight end: the hydroxyphenyl triazine UV-1577 or the bis-benzotriazole UV-360. If the article touches food, pair the choice with a migration test rather than with an argument about molecular weight.
• Engineering plastics processed above 280 C. Polycarbonate, PET and PBT, polyamide, PMMA and modified PPE punish volatile additives twice - you lose the additive in the extruder and you re-deposit it on the die and on the moulded surface. UV-1577 is the standard answer: low volatility, a melting range of 142-149 C depending on grade, and a very low tendency to chelate, which matters in polymers that carry catalyst residues.
• Automotive interior and any low-fogging requirement. Buy on the fogging value and the weight loss, not on the UV range. Low-volatility triazines and high-molecular-weight benzotriazoles sit on the safe side of that line.
• Coatings, especially high-bake and powder coatings. A low-melting or liquid hydroxyphenyl triazine disperses more easily into a coating than a high-melting powder, and the loading is higher than in plastics - typically 0.5-2% on binder solids, with UV-405 built for automotive OEM and industrial powder and liquid coatings. In clear coats the specification that protects appearance is visible transmittance: 460 nm at 95% or better and 500 nm at 97% or better.
• Flame-retarded and thioester-stabilised systems. The hindered hydroxybenzoate UV-2908 was designed for this environment: it screens in the 270-340 nm region and contributes hindered-phenol radical scavenging at the same time, which is why it is recommended for polyolefins, polycarbonate and systems containing flame retardants and thioesters.
• Fibres and low-colour requirements. Oxanilide UV-312 covers roughly 300-380 nm and is used in polypropylene fibre, polyamide, polyester and films where the initial colour must stay low; the trade-off is limited compatibility in strongly polar polymers.

What belongs on the purchase order

• Identity: CAS number together with the chemistry class written out in words. The same trade name is used for different molecules and the same molecule is sold under several trade names; the class is what tells you the spectrum.
• Assay and purity: 99.0% minimum by HPLC, with the method stated on the certificate.
• Physical: melting range, volatile matter (0.5% maximum for compounding grades), ash, and particle size distribution if the grade is supplied as granules rather than powder.
• Optical: transmittance at 460 nm and 500 nm for any transparent or light-coloured application - this is the field that predicts haze and colour drift in the finished part.
• Migration and volatility: a specific-migration statement against the food-contact framework that applies to you, plus either a 100 C/24 h weight loss figure or a gravimetric fogging value for interior applications.
• The new field: a UV-328 residue declaration per batch, with the analytical method and the limit it is tested against, staged to the 100 / 10 / 1 mg/kg timetable. Once the 1 mg/kg step applies, a generic statement that the product does not contain UV-328 is not sufficient documentation.
• Inventory status: TSCA, EINECS and the other national inventories your customer will ask about, and whether the grade is cleared for food-contact use under EU 10/2011, 21 CFR 178.2010 or an equivalent.
• Performance package on a switch: you are changing the light-stabilisation system, so ask for weathering data - QUV or xenon arc to ISO 4892 - on a substrate close to yours, not on a generic polypropylene plaque. A 1,000-hour comparison against the UV-328 formulation you are leaving is the cheapest risk reduction available.

Quick selection

• Transparent engineering plastic at high process temperature: UV-1577 at 0.2-0.5%, and 0.3-0.5% in polyester.
• Maximum migration resistance in a thin, long-life article: UV-360 at 0.2-0.5%.
• Broad absorption for a photosensitive substrate such as a coating or a PET film: UV-1164 at 0.2-0.5%.
• General polyolefin, PVC or styrenic compounding on a budget: UV-329 at 0.2-0.5%, with a migration check if the section is thin.
• High-bake or powder coating: UV-405 at 0.5-2% on binder solids.
• Polypropylene fibre, polyamide or film where initial colour must stay low: UV-312 at 0.2-0.5%.
• Flame-retarded or thioester-stabilised polyolefin: UV-2908 at 0.1-0.5% in thin sections and 0.05-0.2% in thick ones, always with a HALS partner.
• Cost-sensitive, pigmented, non-critical indoor article: 203B at 0.2-0.5%, accepting that long-wave UVA is not covered.

Shanghai Better Chemical supplies compliant UV absorbers across all four classes - UV-329 (octrizole, CAS 3147-75-9), UV-360 (CAS 103597-45-1), UV-405 (CAS 137658-79-8), UV-1577 (CAS 147315-50-2), UV-1164 (CAS 2725-22-6), 203B (CAS 131-54-4), UV-312 (CAS 23949-66-8) and UV-2908 (CAS 67845-93-6) - with assay, melting range, volatile matter, transmittance at 460 and 500 nm, ash and heavy metals on every certificate of analysis, and with the UV-328 residue statement you need for the POPs limits. This family is deliberately kept separate from the HALS grades in our light stabilizer guide, because they do different jobs. Tell us the polymer, the process temperature, the section thickness and the weathering test you have to pass, and we will quote the class that covers the spectrum you actually need.