Epoxy Resin Curing Agents Selection: Amine vs Anhydride vs Polyamide

Selection of an epoxy resin curing agent for industrial production requires simultaneous resolution of stoichiometric balance, mixed-viscosity build, exotherm management, glass transition development, and end-use chemical exposure. A standard bisphenol A diglycidyl ether resin with an epoxide equivalent weight of 184–190 g/eq and a viscosity of 12,000–14,000 mPa·s at 25 °C will yield three distinct network architectures when cured with an unmodified aliphatic amine, a liquid anhydride, or a dimer fatty acid polyamidoamine. The aliphatic amine pathway proceeds by epoxy-amine addition through primary and secondary amine hydrogen functionality at ambient temperature; the anhydride route requires a tertiary amine accelerator and staged thermal input to open the cyclic anhydride and form ester crosslinks; the polyamide route relies on amidoamine active hydrogen and develops a lower crosslink density with pronounced chain flexibility. These mechanistic differences control not only final glass transition temperature but also mixed-system holding time, moisture tolerance, substrate wetting, and long-term adhesion retention.

The starting point for any curative comparison is the stoichiometric mass ratio. For amine-cured systems, the required parts per hundred resin is calculated as amine hydrogen equivalent weight multiplied by 100 and divided by the epoxide equivalent weight. For DGEBA with an epoxide equivalent weight of 184–190 g/eq, this gives diethylenetriamine use at 10.8 phr and isophoronediamine at 22.4 phr. Anhydride systems are not fixed at a 1:1 equivalent ratio; the optimum anhydride-to-epoxide ratio is generally 0.85:1 to 0.95:1 because epoxy homopolymerization and hydroxyl esterification also contribute to network formation. Polyamidoamine systems are less stoichiometrically constrained because they are typically supplied with amine values of 180–400 mg KOH/g, and the use level is set by film flexibility and pot life rather than by a sharply defined equivalent ratio. This difference in stoichiometric sensitivity has direct consequences for meter-mix equipment: amine and anhydride lines require mass-flow calibration to within ±1 % of target, while polyamide formulations tolerate larger lot-to-lot variation without catastrophic undercure.

What Process Windows Do Aliphatic Amines Impose on Ambient-Cure Laminating?

Unmodified aliphatic amines such as diethylenetriamine and triethylenetetramine remain the highest-reactivity curatives for ambient-cure epoxy laminating, but their practical processing window is governed by exotherm and surface carbamation rather than by resin viscosity alone. For a DGEBA resin with an epoxide equivalent weight of 184–190 g/eq, the stoichiometric use levels are 10.8 phr for diethylenetriamine at an amine hydrogen equivalent weight of 20.6 g/eq and 12.8 phr for triethylenetetramine at 24.4 g/eq. A 100 g mass mixed at 25 °C can exceed 150 °C centre temperature within 20–30 min, which disqualifies these formulations for section thickness greater than 25 mm unless chilled hardener and segmented pouring are used. Tensile specimens cured 7 days at 23 ± 2 °C and tested according to ASTM D638-14 Type I typically show tensile strength in the range of 70–85 MPa and elongation at break below 5 %, while glass transition temperature measured by differential scanning calorimetry under ISO 11357-2:2020 falls between 75 °C and 125 °C depending on post-cure. The dominant field limitation is amine blush: exposure of freshly applied film to relative humidity above 80 % and carbon dioxide produces carbamate salts at the surface, which can interfere with overcoat adhesion unless removed by solvent wiping or sanding.

For applications requiring longer pot life without moving to anhydride cure, cycloaliphatic amines such as isophoronediamine at an amine hydrogen equivalent weight of 42.6 g/eq are metered at 22.4 phr for the same DGEBA resin. The mixed viscosity remains below 500 mPa·s at 25 °C, but pot life extends to 2–4 h, allowing vacuum-assisted resin transfer molding of small components with resin inlet temperatures held at 20–25 °C. The resulting networks display lower colour and improved chemical resistance; however, full cure generally requires a post-cure step of 2 h at 80 °C to reach a glass transition temperature above 100 °C. Published tensile data for these systems in fibreglass laminates tested under ISO 527-4 indicate interlaminar shear strength values of 45–60 MPa, but direct substitution into amine-sensitive substrates should be avoided because residual primary amine can attack polycarbonate and certain thermoplastic containment materials.

Amine-cured DGEBA networks are unsuitable for continuous immersion in strong alkali above 60 °C because the bisphenol A backbone and amine linkages undergo hydrolysis; for such service, novolac epoxy cured with aromatic amines or anhydrides is more stable. However, aliphatic amine systems provide adhesion to damp concrete and low-temperature cure down to 5 °C if accelerating amines are used, which is not available with anhydride curatives without special latent accelerators. The modulus of unfilled diethylenetriamine-cured castings measured under ASTM D638-14 typically falls between 2.8 GPa and 3.5 GPa, and flexural strength under ASTM D790-17 is generally 90–120 MPa. These values collapse rapidly above the glass transition temperature; therefore amine-cured structural adhesives are not specified for service above 80 °C unless post-cured with aromatic or cycloaliphatic hardeners.

Anhydride Crosslinking in Electrical Insulation and High-Temperature Composites

Liquid anhydrides such as methyltetrahydrophthalic anhydride and hexahydrophthalic anhydride are selected where mixed-system viscosity must remain below 400 mPa·s and cured glass transition temperature must exceed 140 °C. Anhydride cure does not proceed by simple stoichiometric addition; the formulation is typically set at an anhydride-to-epoxy ratio of 0.85:1 to 0.95:1, and a tertiary amine accelerator such as benzyldimethylamine is added at 0.5–2.0 phr. The initial reaction at 80–100 °C opens the anhydride ring and forms a monoester, which subsequently esterifies an epoxy hydroxyl group; this staged mechanism produces a low exotherm per unit mass and permits vacuum mixing of 20–50 kg batches in jacketed reactors without uncontrolled runaway. Casting resins for instrument transformers and switchgear bushings are commonly cured in a two-stage profile of 2 h at 100 °C followed by 2 h at 150 °C. Volume resistivity after 24 h immersion in 23 ± 2 °C water is measured under IEC 60093 and remains in the range of 10¹⁴–10¹⁵ Ω·cm for high-purity formulations. Heat deflection temperature under ASTM D648-18 Method B at 1.82 MPa typically falls between 130 °C and 160 °C. Anhydride-cured unfilled castings exhibit tensile strength of 65–80 MPa under ASTM D638-14 and elongation at break below 3 %, which is acceptable for rigid encapsulation but unsuitable for applications requiring substrate movement.

Anhydride-cured epoxies are sensitive to moisture during the initial cure phase because free water can hydrolyze the anhydride to diacid, shifting the effective anhydride-to-epoxide ratio and reducing crosslink density. This imposes a maximum ambient relative humidity of 50 % during casting unless dry nitrogen blanketing is used. In contrast, polyamide systems are more tolerant of damp substrates but lose hardness and chemical resistance when over-catalyzed with free amine; addition of more than 100 phr polyamidoamine to a standard DGEBA resin can depress the glass transition temperature below 30 °C and produce soft films that fail drying tests under ISO 9117. The production decisions among the three chemistries therefore depend on the dominant failure risk: for amine systems it is surface blush and exotherm; for anhydride systems it is moisture-induced stoichiometric drift and high cure temperature; for polyamide systems it is slow through-cure and limited solvent resistance.

Dimer fatty acid polyamidoamine hardeners are supplied as viscous liquids with amine values typically between 180 mg KOH/g and 400 mg KOH/g, and they are used at weight ratios of 40–100 phr with standard liquid epoxy resin. The reaction at 25 °C is noticeably slower than that of aliphatic amines; a 200 g mass in a shallow tray typically remains brushable for 4–8 h, while through-dry at 23 ± 2 °C can require 24–48 h. These systems are specified for high-build protective coatings and flexible flooring membranes because they provide elongation at break of 6–20 % under ISO 527-2:2012 and Shore D hardness of 60–75 under ISO 868, compared with Shore D values above 80 for most aliphatic amine and anhydride networks. Salt spray resistance of clear formulations on blasted steel after 1,000 h under ASTM B117-19 is generally acceptable when a silane adhesion promoter is used, but chemical resistance to concentrated acids and methyl ethyl ketone is inferior to that of aromatic amine or anhydride systems. Polyamide-cured films also show lower moisture vapour transmission and less surface oil sensitivity than unmodified aliphatic amine films, which accounts for their use in maintenance coatings over hand-cleaned steel where residual contaminants are likely.

Metering and mixing equipment selection follows the viscosity and pot life data. Low-viscosity amine and anhydride formulations are processed through static mixers with 6–24 elements and dispensing pressures below 10 MPa, whereas high-viscosity polyamidoamine coatings require plural-component spray pumps capable of delivering 2:1 to 4:1 volume ratios and heated hoses at 35–50 °C. In continuous pultrusion with a die temperature of 180 °C, anhydride systems are preferred because the induction period allows the resin-wetted roving to consolidate before gelation; amine systems tend to gel too early and plug the die, while polyamidoamine systems may require excessively long die residence time and can exude during service.

If Part Cross-Section Exceeds 25 mm, Exotherm Management Becomes the Controlling Variable

Thick-section casting of amine-cured formulations requires active thermal management because the epoxy-amine addition liberates approximately 90–110 kJ per equivalent of epoxide reacted. In a cylindrical mould with diameter of 50 mm, centreline temperature can exceed the adiabatic temperature rise by 30–40 °C within the first 10 min when diethylenetriamine is used at 10.8 phr. Production lines therefore use dual-component dispensing equipment with chilled static mixers, hardener lines held at 10–15 °C, and shot sizes limited to 200–500 g per pour. Thermocouple data from exothermic runaway incidents in unfilled amine systems show that visible smoking and microcracking can occur when local temperature exceeds 180 °C, which is below the onset of thermal decomposition but above the glass transition of the partially cured network. Anhydride formulations reduce this hazard because the peak exotherm per unit mass is lower and the cure schedule can be ramped in 10 °C increments; however, stalled anhydride cures from insufficient accelerator or moisture ingress can leave residual free anhydride that depresses HDT and increases dielectric loss. Polyamide systems generate the least thermal hazard during gelation, but thick films exceeding 500 µm wet can trap solvent or moisture and should be force-dried at 40–60 °C to prevent blush and intercoat delamination.

Qualification of epoxy-cured compounds for production release is normally conducted through a matrix of physical, thermal, and electrical tests. A compound supplied for transformer bushings is not accepted on viscosity alone; it is tested for mixed-system viscosity retention over a 60 min window at 25 °C using a rotational rheometer with cone-plate geometry, for heat deflection temperature under ASTM D648-18, for tensile elongation under ISO 527-2:2012, and for dielectric strength under IEC 60243-1. When the end use is an immersion coating on steel, adhesion is assessed by pull-off testing under ISO 4624 and corrosion resistance by scribe creep after 1,000 h under ISO 12944-6. The governing difference between amine, anhydride, and polyamide systems in this qualification loop is not always final hardness; it is the ability of the mixed system to remain within its processing window during the actual pot life under line conditions. A formulation that passes qualification but cannot be metered and poured without exothermic gelation above 60 °C is rejected for series production.