You have a UV coating that will not cure right, or a 3D-print resin that stays soft, and every troubleshooting guide keeps pointing at one ingredient you may rarely think about: the photoinitiator. It is a fraction of a percent of the formula. Nothing crosslinks without it. Choose the wrong one for your lamp and the most expensive UV line in the plant turns out a tacky, under-cured part at full speed.
The short version: A photoinitiator is a molecule that absorbs UV or visible light and converts that energy into reactive species — usually free radicals — that start the polymerization (cure) of a coating, ink, adhesive, or resin. They divide into two families by mechanism. Type I (cleavage) photoinitiators absorb a photon and split directly into two radicals (a Norrish Type I, or α-cleavage, reaction); they need no helper. Type II photoinitiators do not split; they pull a hydrogen atom off a separate co-initiator, a tertiary-amine synergist, and that abstraction forms the initiating radical (related to the Norrish Type II reaction). The practical levers: the initiator’s absorption must overlap your lamp’s emission wavelength, Type II is useless without its amine, and your choice trades off cure speed, surface versus depth, and yellowing. A separate, smaller family, cationic photoinitiators, generate acid instead of radicals and cure epoxies rather than acrylates.
What a photoinitiator actually is
Most industrial UV and LED systems cure by free-radical polymerization: a chain reaction in which radicals add across the carbon-carbon double bonds of acrylate or methacrylate monomers and oligomers, linking them into a crosslinked network in a fraction of a second. That reaction needs a starter. UV light alone does not efficiently rip apart an acrylate, so the formula carries a photoinitiator whose whole job is to absorb the light and hand radicals to the chain.
So the photoinitiator, at roughly 1 to 6 percent of a typical formula, governs whether the part cures at all. The single most useful thing to know about any photoinitiator is the wavelength band it absorbs, because one that does not absorb your lamp’s output simply does not fire. Pull its absorption curve from the Technical Data Sheet (TDS) first.
Type I vs Type II: the core split
Photoinitiators are classified by *how* they make the radical, and that distinction drives almost every downstream choice. A Type I initiator absorbs a photon, jumps to an excited state, and breaks the bond next to its carbonyl group, splitting cleanly into two radical fragments. This is unimolecular: one molecule, one photon, two radicals, no help required. A Type II initiator absorbs the photon but does not break apart. Its excited state grabs a hydrogen atom from a nearby donor, and that bimolecular hand-off produces the radical. The donor is a co-initiator, almost always a tertiary amine, and without it a Type II package barely cures.
| Dimension | Type I (cleavage) | Type II (hydrogen abstraction) |
|---|---|---|
| Mechanism | Absorbs light, then self-cleaves (Norrish Type I / α-cleavage) into two radicals | Absorbs light, then abstracts an H from a co-initiator to form the radical |
| Needs a co-initiator? | No — generates radicals on its own | Yes — a tertiary-amine synergist; almost no cure without it |
| Examples | Hydroxyketones (1173, 184), acylphosphine oxides (TPO, BAPO), aminoketones (907, 369) | Benzophenone, thioxanthones |
| Cure speed | Generally faster (radicals formed directly) | Generally slower (depends on the bimolecular step) |
| Surface vs through cure | Both, by chemistry; acylphosphine oxides excel at depth | Mainly surface cure |
| Yellowing | Hydroxyketones and acylphosphine oxides low (the latter photobleach); aminoketones higher | Benzophenone plus amine can yellow and add odor |
Type I is the default workhorse for clear and most pigmented systems because it is fast and needs no partner. Type II earns its place for a specific surface effect or a cheaper package when you can tolerate the amine. Many real formulas use both, plus an amine, to cover the surface and the depth at once.
The Norrish reactions, explained simply
The two mechanisms are named for the photochemistry Ronald Norrish characterized in carbonyl compounds, and the names get used loosely, so it helps to pin them down.
The Norrish Type I reaction is α-cleavage: an excited ketone breaks the bond immediately next to its carbonyl (C=O) group, producing two radicals. That is exactly what a Type I photoinitiator does. When a hydroxyketone or an acylphosphine oxide absorbs light and snaps into two pieces, one of which attacks the acrylate, you are watching a Norrish Type I cleavage put to work.
The Norrish Type II reaction is hydrogen abstraction: an excited carbonyl pulls a hydrogen from a C–H bond to form a biradical. Type II photoinitiators run on the same idea, with one honest caveat. The classic Norrish Type II is *intramolecular*, abstracting a hydrogen from the molecule’s own chain; a Type II photoinitiator instead abstracts it *intermolecularly*, from a separate amine co-initiator. Same principle, different hydrogen source, and that is why the amine is not optional. If you remember nothing else: Type I splits itself, Type II steals a hydrogen.
Match the photoinitiator to your lamp
This is the step that most often turns a good chemistry into a soft part. A photoinitiator only makes radicals at the wavelengths it absorbs, so its band has to overlap your lamp’s emission. A gap of even 10 to 20 nm between the lamp’s peak and the initiator’s peak sharply cuts photon capture, and no amount of extra dose fully rescues a true mismatch.
A medium-pressure mercury arc lamp emits a broad spectrum with strong discrete lines (notably near 254, 313, 365, 405, and 436 nm), so it feeds almost every photoinitiator at once. UV-LED systems are narrowband, emitting in a tight cluster at 365, 385, 395, or 405 nm, and only excite an initiator that absorbs there. The industry-wide shift from mercury to LED is the biggest reason selection has gotten harder: a package that cured fine under broadband mercury can go dark under a 395 nm LED. Absorption band tracks the chemistry:
- Hydroxyketones (the actives of PI-1173, comparable to Darocur 1173, and PI-184, comparable to Irgacure 184) absorb in the shorter UV, roughly 250 to 340 nm. They pair well with mercury and 365 nm LEDs, give fast surface cure and very low yellowing, and go nearly blind at 395 to 405 nm.
- Acylphosphine oxides (TPO and BAPO, comparable to Irgacure 819) absorb further out, with a tail reaching into the 380 to 420 nm near-visible band, which makes them the workhorses for 395 and 405 nm LED lines. They also photobleach: they absorb strongly at first, then lose that absorption as they react, so light keeps reaching deeper layers while the surface cures. That is what drives through-cure in thick, pigmented, and white films.
- Aminoketones (PI-907 and PI-369) absorb in the longer UV and push cure through heavily pigmented layers, so they are common in white, black, and metallic inks where the pigment competes for every photon.
The action is concrete: overlay your lamp’s emission spectrum with the photoinitiator’s absorption curve from the TDS. A 365 nm hydroxyketone package run under a 405 nm LED will under-cure no matter how long you expose it, and the fix is the initiator, not the dose.
Free-radical vs cationic, briefly
Everything above is free-radical curing, which covers the large majority of UV and LED coatings, inks, adhesives, and 3D resins because acrylate chemistry is fast and versatile. It has one well-known weakness: oxygen at the air interface scavenges surface radicals and can leave a tacky skin, a failure we cover in detail in why UV resin or coating stays tacky after curing.
Cationic photopolymerization is the other route. Its photoinitiator is a photoacid generator (typically an iodonium or sulfonium salt) that releases a strong acid on exposure, and that acid ring-opens and polymerizes epoxies, oxetanes, and vinyl ethers rather than acrylates. The trade-off runs both ways: cationic systems resist oxygen and show low shrinkage and good adhesion, but they are sensitive to humidity and keep curing in the dark after the light is off. The free-radical photoinitiators below are not cationic initiators; a cycloaliphatic-epoxy binder needs a different chemistry entirely.
Which of our photoinitiators is which type
All six free-radical grades RawSource commonly supplies are Type I (cleavage) initiators; they differ in absorption band, cure depth, and color. Type II chemistries (benzophenone, thioxanthones) always require a separate amine synergist and are described above for completeness.
| Photoinitiator | Type | Family | Best lamp match | What it is for |
|---|---|---|---|---|
| PI-1173 (comparable to Darocur 1173) | I | hydroxyketone | mercury, 365 nm LED | fast surface cure, clearcoats, overprint varnish, low yellowing |
| PI-184 (comparable to Irgacure 184) | I | hydroxyketone | mercury, 365 nm LED | clear coats, low yellowing, low odor |
| TPO | I | acylphosphine oxide | 385 / 395 / 405 nm LED, mercury | through-cure, photobleaching, near-visible absorption |
| BAPO (comparable to Irgacure 819) | I | bisacylphosphine oxide | 365–420 nm, mercury and LED | deep, thick, pigmented and white systems |
| PI-907 | I | aminoketone | mercury, long-UV | pigmented and high-opacity systems, good oxygen tolerance |
| PI-369 | I | aminoketone | mercury, long-UV | heavily pigmented inks, photoresists, color filters |
The honest trade-offs sit inside that table. Hydroxyketones give the cleanest color but cannot reach a 405 nm LED. Acylphosphine oxides reach the LED band and cure deep, but like any low-molecular-weight additive they raise migration questions in food-contact and packaging uses, where unreacted initiator or its photoproducts can move out of the cured film; low-migration and polymeric grades exist for that constraint. Aminoketones drive cure through dense pigment but are the family most prone to yellowing, so they belong in pigmented systems rather than a water-clear topcoat. Confirm color and migration on your own substrate before committing a line.
How to choose
There is no single best photoinitiator, only the right one for a set of constraints. Work through them in order:
1. Lamp wavelength first. Mercury feeds almost anything; a 395 or 405 nm LED needs an acylphosphine oxide. This filter alone removes most candidates. 2. Clear or pigmented. Color-critical clears favor hydroxyketones; pigmented, thick, and white work favors photobleaching acylphosphine oxides and aminoketones for depth. 3. Surface or through cure. Thin clearcoats fight oxygen at the surface; thick or filled films fight light penetration at depth, and many formulas blend a surface initiator with a depth initiator. 4. Color, odor, and migration limits. Tighter color budgets push toward hydroxyketones and photobleaching acylphosphine oxides; food-contact and packaging uses push toward low-migration grades. Ask about both early.
For the surface-cure side of the decision (oxygen inhibition, amine synergists, and a tack-free finish), see our companion guide on why UV resin stays tacky after curing. Most of these grades end up in coatings, ink, adhesive, and 3D-resin formulations.
Buying photoinitiators
RawSource supplies the common free-radical photoinitiators for coatings, ink, adhesive, and 3D-resin formulators, in drums and bulk, with Certificate of Analysis (CoA) documentation: surface-cure hydroxyketones (PI-1173 and PI-184), the acylphosphine oxides TPO and BAPO for through-cure and LED lines, and the aminoketones PI-907 and PI-369 for pigmented and high-opacity systems.
Tell us your lamp type and peak wavelength (mercury, or 365 / 385 / 395 / 405 nm LED), your film build and whether it is clear or pigmented, and your color, odor, and migration targets, and request a sample to qualify cure on your own line.
Frequently asked questions
What is a photoinitiator?
A photoinitiator is a molecule that absorbs UV or visible light and converts that energy into reactive species, usually free radicals, that start the polymerization (cure) of a coating, ink, adhesive, or resin. It is a small fraction of the formula, but the cure does not start without it, and the wavelength it absorbs has to match the curing lamp.
What is the difference between Type I and Type II photoinitiators?
A Type I (cleavage) photoinitiator absorbs light and splits directly into two radicals, with no helper molecule needed. A Type II photoinitiator does not split; it abstracts a hydrogen atom from a tertiary-amine co-initiator, and that step forms the radical. Type I is generally faster and is the default for most systems; Type II will barely cure without its amine.
What is the Norrish reaction?
The Norrish reactions are two photochemical pathways of excited carbonyl compounds. The Norrish Type I reaction is α-cleavage, where the molecule breaks the bond next to its carbonyl into two radicals; this is the mechanism of Type I photoinitiators. The Norrish Type II reaction is hydrogen abstraction; Type II photoinitiators use the same idea but pull the hydrogen from a separate amine co-initiator rather than from their own chain.
Which photoinitiator should I use for a 395 nm or 405 nm LED?
Use an acylphosphine oxide such as TPO, whose absorption tail reaches into the 380 to 420 nm band where hydroxyketones like the generics of Darocur 1173 and Irgacure 184 absorb very little. For thick, pigmented, or white films at those wavelengths, BAPO (comparable to Irgacure 819) cures deeper thanks to photobleaching. Many formulations pair an acylphosphine oxide for depth with a hydroxyketone for the surface.
Do Type I photoinitiators need an amine synergist?
No. Type I photoinitiators cleave into radicals on their own and need no co-initiator; Type II photoinitiators (benzophenone, thioxanthones) do. Amines are still sometimes added to Type I clearcoats to fight oxygen inhibition at the surface, but that is an oxygen-scavenging role, not a requirement to make the initiator work.
What is the difference between free-radical and cationic photoinitiators?
Free-radical photoinitiators generate radicals that cure acrylate and methacrylate chemistry; they are fast and account for most UV and LED systems, but the surface can be inhibited by oxygen. Cationic photoinitiators are photoacid generators (iodonium or sulfonium salts) that release acid to cure epoxies, oxetanes, and vinyl ethers; they resist oxygen and show low shrinkage, but they are sensitive to humidity and continue curing in the dark. The binder chemistry decides which you need.
Editorial note. This article is general technical guidance for coatings, ink, adhesive, and 3D-resin formulation professionals. Cure speed, surface tack, through-cure, color, and migration depend on your specific resin and oligomer, photoinitiator package, lamp type and wavelength, film thickness, pigmentation, line speed, and curing atmosphere, and must be validated on your own system; the Certificate of Analysis (CoA) governs the grade you buy. Photoinitiators and amine synergists are industrial chemicals; review the current Safety Data Sheet (SDS) and use appropriate PPE before handling. Trade names (Irgacure, Darocur) are referenced only to identify the comparable generic active and are the property of their respective owner (IGM Resins); no affiliation or endorsement is implied. Products are sold for industrial and professional use only. Nothing here is a medical, health, or safety claim. RawSource makes no warranty, express or implied, and assumes no liability for use of this information.
