If you formulate engineering plastics, you have probably stood at this crossroads more than once. The part has to survive outdoors, so it needs UV protection. It also has to survive heat, moisture, or molecular-weight loss during processing. The instinct is to stack two trusted tools: an epoxy-based additive for stability, and a hindered amine light stabilizer (HALS) for weathering.
I understand the reflex. Epoxy chemistry is the usual fix for hydrolysis and chain scission. HALS is the workhorse for light stability. Put them together and you seem to cover both problems at once.
The catch is a quiet piece of chemistry most people overlook: HALS is an amine, and epoxies react with amines. Leave that reaction out of the picture and a formulation that looks like “two good stabilizers” can quietly become “two stabilizers that eat each other before either one helps the resin.”
Below is where this pairing shows up, why the chemistry turns against you, how to read the damage in your test data, and how to combine the two additives without blowing up the compound.
Where HALS plus epoxy actually shows up
The combination is not exotic. It appears any time a material has to be tough during processing or in humidity and tough in sunlight.
PBT needs both hydrolysis resistance and UV resistance. PBT degrades through acid-catalyzed hydrolysis, driven by carboxyl end groups and residual moisture. Formulators add epoxy to cap those groups. But when the same PBT becomes an outdoor part, an automotive exterior, or an exposed appliance housing, UV absorbers and HALS get added on top. Now the melt carries PBT, epoxy, and HALS together.
PC/ABS needs processing heat stability and UV resistance. PC is touchy around acids, bases, metal ions, leftover emulsifiers, and some flame retardants. Recycled or flame-retarded grades often show creeping melt flow, dropping impact, and yellowing after heat stays on them. Epoxy stabilizers, acid scavengers, or metal deactivators get pulled in to calm that down. Then, if the part is exterior trim, a charging station, or an outdoor enclosure, HALS and UV absorbers join in. The risk is sharper here: PC is already sensitive to amines, so HALS can cause trouble beyond just the epoxy.
rPET and PET need chain extension and weathering. Recycled PET loses molecular weight and picks up carboxyl end groups. Multifunctional epoxy chain extenders are standard for rebuilding melt strength. But rPET used in outdoor sheet, profile, film, or molded parts still needs UV protection, so UVA and HALS ride along. Same tension as PBT: is the epoxy rebuilding the polymer, or is HALS and moisture eating it?
PA, TPU, and polyester alloys fall into the same trap. Polyamide gets epoxy for interfacial improvement, chain extension, and compatibilization, and HALS for outdoor weathering. TPU gets epoxy for end-group stabilization and hydrolysis resistance, plus HALS for weathering. PBT/ASA, PC/PBT alloys, and flame-retarded PBT all follow the same logic. In every case the two additives are added for good reasons and end up in the same barrel.
So the pairing is common and tempting. The goal is sound. The problem is what the two do to each other once the screw turns.
The chemistry: why amines and epoxies do not get along
An epoxy group is a three-membered ring under strain. That strain is what makes it useful, and what makes it fragile. Nucleophiles attack the ring and open it. In polyester and epoxy systems, amines, quaternary ammonium salts, imidazoles, and metal ions are used as catalysts precisely because they drive that ring-opening.
HALS, at its core, is a hindered amine light stabilizer. The “hindered” part lowers reactivity but does not switch it off. Two things can still happen.
First, direct reaction. HALS structures that carry an N-H bond can behave like an amine curing an epoxy. The ring opens and forms a beta-hydroxy amine. The hindered amine is no longer free to do its light-stabilizing job.
Second, catalysis. Even tertiary-amine HALS, which have no N-H, are basic. They can catalyze epoxy ring-opening and change which competing species the epoxy reacts with, whether that is carboxyl groups, hydroxyl groups, or water.
The practical point is short: the epoxy’s intended reaction gets redirected. It may still react, just not where you wanted.
What the epoxy is supposed to do in PBT, and what hijacks it
In an ideal PBT system, the epoxy finds a carboxyl end group and forms an ester. Carboxyl end groups drop, molecular weight climbs, and hydrolysis resistance improves. That is the target reaction, and it works.
The trouble is that in a real modified PBT compound, the epoxy does not meet carboxyl groups alone. It also runs into hydroxyl end groups (which can outnumber carboxyl end groups by four to five times), residual water, glass-fiber sizing, acidic sites or adsorbed water on flame-retardant surfaces, pigments, lubricants, antioxidant residues, and the amine structure of the HALS.
Because hydroxyl end groups can be four or five times more abundant than carboxyl end groups, the epoxy is already spread thin before HALS enters the picture. Add HALS and you bring in another class of amine sites that can consume the epoxy directly, or speed up its reaction with water and hydroxyl groups. The epoxy that was meant to cap acid ends can end up reacting with HALS, moisture, and hydroxyl groups in what amounts to a free-for-all.
The math that proves HALS is not a tiny additive
A rough engineering estimate makes the risk concrete. Take a PBT compound with 0.3% epoxy chain extender (epoxy equivalent weight 300 to 500 g/eq) and 0.3% HALS (molecular weight 500 to 1000 g/mol, carrying one or two amine structures).
That 0.3% epoxy contributes roughly 6 to 10 milliequivalents of epoxy per kilogram. The 0.3% HALS contributes about 3 to 12 milliequivalents of amine sites per kilogram. The PBT’s carboxyl end groups, by comparison, may sit around 20 milliequivalents per kilogram.
Read that again. The HALS is not a trace impurity. On a stoichiometric basis it can compete with the polymer’s own end groups for the epoxy. Once you run the numbers, “the HALS will hardly matter” stops being a safe assumption.
Three ways this wastes your formulation
When epoxy and HALS interfere, the losses stack up.
The epoxy gets wasted. It was meant to cap PBT carboxyl groups, but HALS or HALS-catalyzed side reactions burn it up first.
The HALS gets wasted too. HALS is supposed to cycle through radical-scavenging during photo-oxidation. If it has already opened an epoxy ring or gotten pinned in the wrong place, its migration, distribution, and regeneration all change. It may never reach the photo-active role you paid for.
And the material’s performance gets wasted. Side reactions between epoxy and HALS can make new hydroxyl groups that push transesterification, and can catalyze branching, gelation, or degradation. The symptoms show up clearly in testing: melt flow rate drops for no good reason (over-branching), or climbs instead (polyester or PC degradation); impact strength falls; yellowing gets worse; black specks, gels, or rising filter pressure appear; and humidity-aged properties end up worse than the unprotected baseline.
The cruelest outcome is this: UV resistance was not achieved, hydrolysis resistance was not achieved, and the processing stability was damaged on top of it.
How to read the reaction priority
Reaction order in a compound is never absolute. Temperature, concentration, dispersion, acid-base environment, and steric hindrance all shift it. As a practical risk guide, think in this order.
Primary and secondary amines beat carboxyl end groups. If the system holds an N-H amine, especially a small or low-steric one, it usually reacts with epoxy fast and consumes it directly. HALS that contain N-H are the highest-risk structures.
Tertiary amines may not consume epoxy, but they catalyze it. A tertiary-amine HALS has no N-H and adds to epoxy weakly, but its basicity catalyzes ring-opening so the epoxy more readily reacts with carboxyl, hydroxyl, and water. This failure is sneakier: your acid value may drop, but your melt flow rate drops with it, a sign that hydroxyl groups reacted too, along with branching or local crosslinking.
Carboxyl is the target reaction, but not always the priority. PBT carboxyl end groups are not abundant. If epoxy loading is modest while water, hydroxyl groups, and amine sites are plentiful, the epoxy will not all go to carboxyl capping.
Keep one sentence in mind: “epoxy can react with carboxylic acid” is not the same as “in your formulation, epoxy mainly reacts with carboxylic acid.”
How to pair them more safely
The combination is not forbidden. It just has to be designed, not assumed.
For PBT needing hydrolysis resistance plus UV, favor a milder glycidyl ether or a novolac epoxy, and choose a HALS that is low-basicity, high-molecular-weight, and N-alkylated or NOR-type. Low-reactivity, high-MW hindered amine structures, the Tinuvin 123 class, lower the chance of direct consumption.
For PC/ABS needing heat and humidity stability plus UV, validate first that the epoxy stabilizer is safe for PC, since PC is amine-sensitive. Then pick a HALS that is low-basicity and PC-compatible. A common failure in PC/ABS is simply using the wrong HALS; low-basicity coated hindered amines are the safer direction when both thermal and weathering stability are required.
For rPET and rPBT needing chain extension plus weathering, dose the epoxy against a measured melt flow or intrinsic viscosity target instead of a rule of thumb, add the HALS afterward, and choose a low-reactivity structure so it interferes less with the chain-extension step.
The takeaway
HALS plus epoxy is not a mistake by definition. The intent is usually right: one additive handles hydrolysis, heat stability, and chain extension, the other handles UV, chalking, and yellowing. But in PBT, PET, PC/ABS, and related systems, you cannot forget the underlying chemical relationship. An amine and an epoxy will find each other.
The fix is not to drop either additive. It is to plan the pairing, pick milder epoxy chemistries, use lower-basicity HALS, confirm PC compatibility, and dose against measured targets, so the two stabilizers protect the resin instead of consuming each other.
If you are seeing unexplained melt-flow swings, yellowing, or worse humidity-aging after adding both an epoxy and a HALS, the interference between them is one of the first places I would look.

