High-temperature polyimide work has one recurring constraint: every condensation cure releases water, and water in a thick laminate becomes a void. Terminal alkynes solve it. An alkyne cures by addition – it reacts with itself, forming aromatic rings – so the network is built without a by-product, without volatiles, and without the void content that limits how thick a part can be. Two anhydrides dominate this chemistry: 4-ethynylphthalic anhydride (CAS 73819-76-8) and 4-phenylethynylphthalic anhydride, PEPA (CAS 119389-05-8). They share a mechanism and differ in almost everything a process engineer cares about.
Why a terminal alkyne
A condensation-cured polyimide is ramped slowly because imidisation releases water. An addition-cured one is not: the alkyne groups react with each other at high temperature to give a crosslinked network with no small-molecule by-product. Three consequences follow:
- Thick sections become possible. Void content stops being driven by diffusion of water out of a thick part.
- Service temperature rises. The crosslinks formed are aromatic, so the cured resin holds modulus above the glass transition of the thermoplastic analogue.
- The cure becomes latent. The alkyne is unreactive at melt-processing temperatures and reacts on a schedule set by time and temperature, which is what makes resin transfer moulding of a polyimide feasible at all.
The price is a narrow window. A reactive alkyne on a rigid backbone begins to react before the part is fully formed; a less reactive one gives a wider window but demands a hotter cure. Choosing between the two products below is essentially choosing where on that trade to sit.
The two products compared
| Property | 4-Ethynylphthalic anhydride | PEPA (4-phenylethynylphthalic anhydride) |
| CAS | 73819-76-8 | 119389-05-8 |
| Formula / MW | C10H4O3 / 172.14 | C16H8O3 / 248.24 |
| Anhydride equivalent weight | 172.14 | 248.24 |
| Alkyne groups per molecule | 1 (terminal, directly on the ring) | 1 (terminal, through a phenyl spacer) |
| Spacer between alkyne and anhydride | None – the alkyne is attached to the phthalic ring | One phenyl ring |
| Reactivity of the alkyne | Higher – cures at a lower temperature | Lower – needs a hotter cure |
| Processing window | Narrower | Wider |
| Cured network | More rigid, higher crosslink density per unit mass | Tougher, more phenyl content, higher mass per crosslink |
| Where it is used | Small parts, short schedules, chain extension | Thick sections, RTM, aerospace laminates |
The 76 g/mol difference is all phenyl spacer. In equivalent terms that is a 44% increase in mass per reactive site, which is why PEPA looks expensive per kilogram and is still the standard end-cap for large structures: what it buys is time before the resin gels.
The stoichiometry: this is an end-cap, not a crosslinker in the usual sense
Both products are used in two distinct ways, and the arithmetic differs:
| Use | What it does | Charge |
| Chain end-cap | Terminates a polyimide oligomer with an alkyne; the oligomers then couple through the alkynes | Small excess over the calculated imbalance; the alkyne count follows from the chain length |
| Backbone monomer | Sits inside the chain and provides the crosslink site for a second-stage cure | Stoichiometric against the diamine, like any dianhydride |
| Epoxy hardener (secondary use) | The anhydride ring cures the epoxy; the alkyne provides a later, higher-temperature step | phr = AEW / EEW x 100 x r: 83.8 for the ethynyl compound, 120.8 for PEPA at EEW 185, r = 0.9 |
For an end-cap calculation the useful figure is the alkyne equivalent weight, which equals the molecular weight since each molecule carries exactly one alkyne: 172.14 g per mole of alkyne for 4-ethynylphthalic anhydride and 248.24 for PEPA. If you know the target oligomer molecular weight and want two end groups per chain, the mass of end-capper is 2 x (oligomer moles) x (alkyne equivalent weight), corrected for assay.
Why the spacer matters
Attaching the alkyne directly to a phthalic ring puts the reactive triple bond in conjugation with the anhydride’s aromatic system. Electronically that raises its reactivity: the alkyne begins to react at a lower temperature, so a part can be cured on a shorter, cooler schedule. For a small part with thin sections that is an advantage – less thermal exposure, less residual stress.
Inserting a phenyl ring between the alkyne and the anhydride isolates the reactive centre. The alkyne sees a less electron-poor environment, the onset of cure moves up, and the window during which the resin flows widens. For a thick laminate that takes hours to fill, that window is the difference between a part and a scrap. The phenyl also adds mass and aromatic content, which softens the cured network slightly relative to the direct-attached version.
Storage, and how the material fails quietly
An alkyne-bearing powder ages. Light and heat consume the alkyne, and the first sign is not a change in appearance but a change in cure behaviour: the DSC exotherm shrinks, then disappears, and the batch cures to a lower crosslink density without ever failing an assay. Handling rules follow:
- Store cool, dry and dark, in tightly closed containers, away from peroxide and radical sources and away from amines.
- Do not hold drums near heat sources; do not leave material in a sunlit area during summer loading.
- Ship sample quantities by air so that the material spends less time at ambient temperature.
- Ask for the alkyne content, not just the assay. IR at about 2100 cm-1 (C-C triple bond stretch) and about 3300 cm-1 (terminal C-H) shows the alkyne directly; NMR shows the terminal proton. On a good lot these are strong; on an aged lot they are reduced while the assay still passes.
Choosing between them
| Question | Answer points to | Why |
| Is the part thick, or does the mould take hours to fill? | PEPA | The phenyl spacer widens the processing window |
| Is the schedule short and the part small? | 4-Ethynylphthalic anhydride | Higher reactivity, lower cure temperature, less thermal exposure |
| Is cost per reactive site the constraint? | 4-Ethynylphthalic anhydride | 172.14 against 248.24 g per alkyne |
| Is maximum toughness in the cured network the target? | PEPA | More aromatic mass per crosslink, softer network |
| Is the alkyne wanted for click chemistry rather than thermal cure? | Either, but keep copper out of the thermal cure | Residual copper will catalyse premature alkyne reaction |
Related reading
- 4-Phenylethynylphthalic anhydride (PEPA, CAS 119389-05-8): properties and advanced applications
- Alicyclic and aromatic dianhydrides: choosing the backbone for a polyimide
- Bridged bis-anhydrides compared: methylene, ethylidene, ether and bisphenol A
- Anhydride curing agents and dianhydrides: the family selection map