When DETA and IPDA are both quoted to cure the same liquid bisphenol A diglycidyl ether (DGEBA), the two lines are not interchangeable. DETA’s calculated AHEW is 20.6 g/eq and IPDA’s 42.6 g/eq, so they carry different doses, and in US 5,235,007 IPDA gelled in 180 min against 40 min for DETA (200 g, 23 °C). Specify five things for each hardener line: the hardener, its class, the grade’s AHEW or amine value, the resin’s EEW and the phr.
How do TETA, MXDA and the other amine hardeners compare on AHEW?
An epoxy curing agent crosslinks the resin through its epoxide groups. An amine hardener is dosed by its amine hydrogen equivalent weight (AHEW), the molecular weight divided by active amine hydrogens, and the dose is phr = AHEW × 100 / EEW, where EEW is the grams of resin per epoxide equivalent.
The hardeners compared here are aliphatic, cycloaliphatic, cycloaliphatic-aliphatic, araliphatic or aromatic amines. The single compounds carry as few as 3 and as many as 6 amine hydrogens: two per primary amine, one per secondary, none per tertiary. Calculated AHEW runs from 18.5 g/eq for 1,3-diaminopropane to 59.6 g/eq for 4,4′-methylenebis(2-methylcyclohexylamine) (MACM).
Triethylenetetramine has no calculated value: CAS 112-24-3 names the linear compound, but the commercial grade is a mixture of linear, branched and cyclic congeners that one structure value would not describe.
Divide a PubChem molecular weight by the hydrogen count and the AHEW follows. N’-[3-(Dimethylamino)propyl]propane-1,3-diamine, at 159.27 g/mol with 3 hydrogens across its three nitrogens, comes to a calculated 53.1 g/eq. Isomer mixtures share one calculated value, since their isomers share a molecular weight: both methylcyclohexanediamine regioisomers are 128.22 g/mol.
A calculated AHEW is a structure value. A 2017 study in Polymers (PMC6418755) gives MDA an amine equivalent weight of 49.5, close to its calculated 49.6 g/eq, but a lot’s AHEW follows from its measured amine content. Set the dose from the value on the grade’s CoA. Patent and journal figures on this page are literature values for their own test systems, not RawSource specifications.
| Amine hardener | CAS | Class | Amine H | Calc. AHEW, g/eq | RS codes | TSCA |
|---|---|---|---|---|---|---|
| 1,2-Diaminocyclohexane (1,2-DACH) | 694-83-7 | cycloaliphatic | 4 | 28.5 | RS-DACH | Active |
| 1,4-Cyclohexanebis (methylamine) | 2549-93-1 | cycloaliphatic | 4 | 35.6 | RS-14BAMC | Active |
| N-(3-Aminopropyl)cyclohexylamine | 3312-60-5 | cycloaliphatic-aliphatic | 3 | 52.1 | RS-CHPDA | Active |
Methylcyclohexanediamine MCDA |
13897-55-7 | cycloaliphatic | 4 | 32.1 | RS-MCDA | Active (PMN flag) |
Isophorone diamine IPDA |
2855-13-2 | cycloaliphatic | 4 | 42.6 | RS-IPDA | Active |
4,4′-Methylenebis(2-methylcyclohexylamine) MACM |
6864-37-5 | cycloaliphatic | 4 | 59.6 | RS-MACM, RS-MACM_99-6 | Active |
| 4,4′-Diaminodicyclohexylmethane | 1761-71-3 | cycloaliphatic | 4 | 52.6 | RS-PACM, RS-PACM_24-30, RS-PACM_30-40 | Active |
m-Xylylenediamine MXDA |
1477-55-0 | araliphatic | 4 | 34.0 | RS-MXDA | Active |
Diaminodiphenylmethane MDA |
101-77-9 | aromatic | 4 | 49.6 | RS-MDA_99 | Active |
m-Phenylenediamine MPD |
108-45-2 | aromatic | 4 | 27.0 | RS-MPD | Active |
| Diethylenetriamine | 111-40-0 | aliphatic | 5 | 20.6 | RS-DETA | Active |
| Triethylenetetramine | 112-24-3 | aliphatic | – | – | RS-TETA | Active |
Dipropylenetriamine DPTA |
56-18-8 | aliphatic | 5 | 26.2 | RS-DPTA_99, RS-DPTA_97 | Active |
| N-(2-Aminoethyl)-1,3-propanediamine | 13531-52-7 | aliphatic | 5 | 23.4 | RS-AEPDA | Active |
| N,N’-Bis (3-aminopropyl)ethylenediamine | 10563-26-5 | aliphatic | 6 | 29.0 | RS-BAPEDA | Active |
| 1,3-Diaminopropane | 109-76-2 | aliphatic | 4 | 18.5 | RS-13DAP | Active |
N’-[3-(Dimethylamino)propyl]propane-1,3-diamine DMAPAPA |
10563-29-8 | aliphatic | 3 | 53.1 | RS-DMAPAPA | Active |
Documents for every product above
- Technical data sheet: on request
- Safety data sheet: on request
- Certificate of analysis: issued per lot with each shipment

Amine hardener product pages (17)
Building a DGEBA system: resin EEW, hardener AHEW and the dose
The dose is one line of arithmetic: phr = AHEW × 100 / EEW, because one amine hydrogen equivalent consumes one epoxy equivalent. The resin enters only through its EEW.
RawSource’s DGEBA, RS-DGEBA, is sold to an EEW of 184-190 g/eq (typical 185.6) at 11,000-13,000 mPa·s (25 °C); its DGEBF, RS-DGEBF, to 160-180 g/eq (typical 168.1) at 3,000-5,000 cP (25 °C). Isophoronediamine, calculated AHEW 42.6 g/eq in the table above, doses at 42.6 × 100 / 185.6 = 23.0 phr in RS-DGEBA and 42.6 × 100 / 168.1 = 25.3 phr in RS-DGEBF: the lower-EEW resin takes more hardener for the same stoichiometry.
The liquid DGEBF test grade in US 8741988, at EEW 165-173 g/eq, sits in the same band. Carry a DGEBA phr across to DGEBF unchanged and the resin is left short of amine, so recompute before the first batch. The same arithmetic governs price. A lower-AHEW hardener carries less weight per 100 parts of resin, and a price per kilogram compared before the phr is set misleads.
A tertiary-amine accelerator sits outside this sum. It adds no amine hydrogen, and US 8,501,997 puts such accelerators at 10% or less, more often 5% or less, of total binder weight. The calculated phr is a starting point that a Tg or HDT series confirms (US 5,235,007). Bracket it on both sides in the trial.

Cycloaliphatic, aliphatic and aromatic amines: how IPDA, TETA and MDA cure
Aliphatic amines (DETA, TETA, DPTA, 1,3-diaminopropane, N-(2-aminoethyl)-1,3-propanediamine, N,N’-bis(3-aminopropyl)ethylenediamine and N’-[3-(dimethylamino)propyl]propane-1,3-diamine) are the fastest class. In US 5,235,007, with liquid DGEBA in a 200 g mass at 23 °C, DETA gelled in 40 min (ASTM D2471) with a Tg of 130 °C and a 265 °C peak exotherm; TETA gelled in 39 min, at 130 °C and 250 °C. Speed costs toughness. US 2008/0114094 finds the cured network short on flexibility.
Chain spacing moves that limit. In a quartz-sand floor in the same application, a mixture rich in N,N’-bis(3-aminopropyl)ethylenediamine raised tensile strength from 1786 to 2701 psi and flexural strength from 4518 to 6100 psi against TETA. If a TETA floor fails in flexure, trial the propylene-spaced amine before changing the resin. N’-[3-(Dimethylamino)propyl]propane-1,3-diamine earns its place in cold-curing adhesive hardeners (US 11,053,346) through low viscosity and stability against crystallization in the cold; its tertiary dimethylamino end group adds catalysis but no amine hydrogen.
Cycloaliphatic amines (IPDA, MCDA, DACH, 4,4′-diaminodicyclohexylmethane, MACM and 1,4-bis(aminomethyl)cyclohexane) typically cure somewhat more slowly than aliphatics (US 8,828,267). 1,4-Bis(aminomethyl)cyclohexane carries its amines on methylene carbons, so confirm its reactivity by trial rather than from the class. In the US 5,235,007 test IPDA gelled in 180 min with a Tg of 153 °C. That pot life is why US 8,828,267 places cycloaliphatics in large-area composite production, where a fast aliphatic would gel first.
The same patent names wind turbine blades and floor coatings for a methylcyclohexanediamine isomer mixture, and US 5,025,078 finds that the 2,6-isomer delays the reaction and extends pot life. Ask for the isomer split before you compare MCDA lots. N-(3-Aminopropyl)cyclohexylamine is cycloaliphatic-aliphatic, with one primary and one secondary amine, so it carries 3 amine hydrogens where the ring diamines carry 4.
MXDA is araliphatic: its amine groups sit on methylene carbons, which a 2020 ACS Omega study says puts its reactivity closer to an aliphatic amine. An ACS Applied Polymer Materials network study cured MXDA, its para isomer and the two saturated bis(aminomethyl)cyclohexanes in one bisphenol F diglycidyl ether (DGEBF) resin; aromaticity and 1,3-substitution both lowered methanol and ethanol uptake. Where alcohol resistance decides the grade, MXDA goes on trial first.
Aromatic MDA and MPD cure by far the slowest (US 8,828,267) and are solid at room temperature, so the plant needs heat at the mixing step. MDA melts at about 92 °C (US 9,321,881). MPD is melted too, and the resin is warmed to about the temperature of the melt so the amine does not crystallize on mixing (US 4,003,876).
MPD hardens resins for casting, tooling, laminating and adhesives. MDA cures epoxy resins and urethane elastomers (PubChem record). In the 2017 Polymers study (PMC6418755), DGEBA mixed with MDA at stoichiometry and cured 4 h at 90 °C with no post-cure showed two overlapping DMA tan δ peaks. The authors assign 110.8 °C to the partially cured material and 154.9 °C to Tg∞, the Tg a fully cured network would reach.
Four hours at 90 °C without a post-cure leaves part of the network short of it. Cost the heated mixing and post-cure steps before comparing price per kilogram.
4,4′-Diaminodiphenylmethane (MDA, EC 202-974-4) is on REACH Annex XIV, the authorisation list (Regulation (EC) No 1907/2006), as carcinogenic (Article 57a), with a sunset date of 21 August 2014, per HSE’s UK REACH authorisation list, which carries the EU entries made before 2021. Placing it on the EU market or using it there after that date, including as an epoxy hardener, needs an authorisation; uses that REACH exempts from authorisation are outside this page, so confirm the position for your own use.
In the US, occupational exposure to MDA falls under OSHA’s methylenedianiline standard, 29 CFR 1910.1050 (29 CFR 1926.60 for construction work), which sets permissible exposure limits of 10 ppb as an 8-hour time-weighted average and 100 ppb as a short-term exposure limit. Read the current SDS before handling.
More amine hardeners: tetraethylenepentamine, TMD, PXDA and piperazine
Tetraethylenepentamine, trimethylhexamethylenediamine and piperazine sit outside the AHEW table, so buy them to the AHEW or amine value on the grade’s CoA.
Primary amines react significantly faster with epoxide than secondary amines of the same class (US 8,828,267), which is the reason to look at the two N-sec-butyl secondary diamines, 4,4′-methylenebis(N-sec-butylcyclohexylamine) and N,N’-di-sec-butyl-4,4′-methylenedianiline, when pot life is the constraint. By their PubChem structures, those two and piperazine each carry two amine hydrogens, fewer than the three reactive amine hydrogens US 4,800,222 calls generally useful in a curing agent. Run them in a blend with a hardener of three or more and dose the blend by equivalents.
1,3-Bis(aminomethyl)cyclohexane carries the 1,3-substitution and p-xylylenediamine the aromaticity that each improved alcohol resistance beside MXDA in the DGEBF network study.
More amine hardener product pages (12)
Accelerators and latent hardeners: dicyandiamide and imidazoles
2,4,6-Tris(dimethylaminomethyl)phenol (C15H27N3O, 265.39 g/mol) carries three dimethylaminomethyl groups on a phenol ring. Its three nitrogens are all tertiary, so it adds no amine hydrogen and never enters the AHEW; its phenolic OH is not counted in the amine stoichiometry. US 8,501,997 puts epoxy-amine accelerators at 10% or less, more often 5% or less, of total binder weight. Name that basis on the purchase order, since a level quoted on the curing agent alone is a different number.
The accelerator’s makeup matters as much as its level. In US 4,800,222, with a polyoxypropylene diamine, 10 parts by weight of tris(dimethylaminomethyl)phenol alone gave a 168.5 min gel time in a 200 g mass, against 26.5 min for a triethanolamine and piperazine blend. Raising the Mannich-base share sped thin-film drying and made the films more brittle. Tertiary amines also speed the viscosity rise at room temperature (US 8,852,489), so a faster film comes with a shorter window at the mixer.
Dicyandiamide and the imidazoles work at the hot end: PubChem’s record for dicyandiamide (CID 10005) lists it as a heat-cured epoxy resin curing agent. Imidazole has been used to allow low-temperature baking of polyester-epoxy powders, at the cost of poor pot life (US 6,649,728). Trial the imidazole against your mixing and storage time, not only against the bake.
Accelerator and latent hardener product pages (8)
Anhydride and high-temperature chemistry: BTDA, ODA and triphenylphosphine (TPP)
Benzophenonetetracarboxylic dianhydride (BTDA) is dosed by anhydride equivalents, not AHEW. PubChem gives BTDA a molecular weight of 322.22 g/mol (CID 75498); with two anhydride groups that is about 161 g/equiv per anhydride equivalent. BTDA also builds polyimide coatings, which an RSC Advances study imidized at 250 °C on mild steel. State the anhydride-to-epoxide ratio your formula targets when you ask for a quote.
Triphenylphosphine accelerates the epoxy-phenol reaction, typically above 150 °C, by forming a betaine that turns the phenol into a more reactive phenolate, per a 2024 paper in Monatshefte für Chemie. In the same paper, 5 mol% triphenylphosphine alone did not polymerize phenyl glycidyl ether at room temperature or at 80 °C in 24 h, so it needs a co-reactant beside it, phenolic or carboxyl.
In a polyester and triglycidyl isocyanurate powder baked 10 min at 205 °C (US 4,910,287), 0.20% triphenylphosphine gave 20 in-lb direct impact against 160 in-lb for 0.15% 1-methyl imidazole. For impact in that powder, the imidazole was the better catalyst.
4,4′-Oxydianiline is a polyimide intermediate: with a dianhydride it forms a polyamic acid that is cyclodehydrated to polyimide. In a 2022 FTIR study of PMDA/ODA films, the degree of imidization peaked at a 320 °C cure and stayed flat above it, so a lower oven leaves imidization incomplete.
HSE’s UK REACH Candidate List, which carried over the EU entries, has listed 4,4′-oxydianiline and its salts (EC 202-977-0) as substances of very high concern (Regulation (EC) No 1907/2006) since 19 December 2012; PubChem’s ECHA-attributed entry gives the basis as carcinogenic and mutagenic (Articles 57a and 57b).
Anhydride and related product pages (6)
Epoxy resins and diluents these hardeners cure: DGEBA, novolac and more
EEW is the resin’s half of the dose. Because phr = AHEW × 100 / EEW, a lower-EEW resin takes more hardener per 100 parts. A liquid DGEBF test grade in US 8741988 runs at EEW 165-173 g/eq, so check the phr again before moving a formula from DGEBA to DGEBF. For a resin blend, or a resin cut with a glycidyl ether reactive diluent, compute the blend EEW by equivalents.
Viscosity and storage pull in opposite directions. A 2023 study in Polymers (PMC10575199) puts pure DGEBA resins at typically 10,000-15,000 mPa s against 5,000-7,000 mPa s for DGEBF; those are literature ranges on a different temperature basis from the 25 °C RS-DGEBA and RS-DGEBF specifications, so judge a grade on its own specification. Yet in the study’s EN ISO 4895 test, unmodified resin seeded with calcium carbonate and ethanol and stored at 10 °C, DGEBF reached crystallization stage b in 4 days against 13 days for DGEBA.
EP 2773682 notes that liquid resin can crystallize at about 25 °C or below and lists blending in DGEBF as a fix. The fix is the blend, not the swap: neat DGEBF crystallized faster than neat DGEBA in the seeded test. US 3,051,681 adds about 2 to 10% by weight of epoxidized novolak to highly purified DGEBA for ambient storage.
In the amine-cured networks of the ACS Applied Polymer Materials study, DGEBF beat DGEBA on methanol and ethanol uptake. For weather-resistant coatings, US 4,528,345 pre-reacts aminoethylpiperazine with hydrogenated DGEBA before curing with a polyoxyalkylene polyamine.
Propylene carbonate is 4-methyl-1,3-dioxolan-2-one, a five-membered cyclic carbonate that reacts with primary amines to hydroxyurethanes, relatively slowly (US 6,228,979). It consumes amine, so count it in the hardener demand. Premixed with the resin at 75:25 to 99:1 it significantly lowers viscosity (US 8,003,737). Pre-reacted with IPDA at 2:1 amine to carbonate in US 5,235,007, it cut gel time from 180 to 55 min but raised initial viscosity to 45,000 cP and lowered Tg from 153 to 116 °C.
Epoxy resin and diluent product pages (8)
Polyetheramine hardeners
In US 5,235,007 (200 g, 23 °C), the patent’s table lists no gel time for the polyoxypropylene diamine of molecular weight about 230 alone and gives it a 64 °C peak exotherm, against a 40 min gel and 265 °C for DETA. Blended with a TETA and propylene carbonate aminocarbamate at an amine-equivalent ratio of 1.39, it gelled in 44 min.
Where the cure is too slow, US 4,127,514 speeds it with salicylic acid and iminobis(propylamine), the dipropylenetriamine structure. The grades sit under polyetheramine hardeners (diamines and triamines).
Industries served
Floor and coatings buyers should choose on blush and cold cure before Tg, since US 11,053,346 finds blushing worst in cold, moist conditions. If you cast, tool or laminate, m-phenylenediamine is documented for those uses (PubChem); melt handling decides the grade, and for MDA so do the REACH authorisation and the OSHA standard noted above.
Documents and quotes
Put these six items on your request for epoxy curing agents, and the quote comes back against the cure you will run:
- The hardener by name and CAS from its product page, and the AHEW or amine value the grade is sold to on its CoA. A lot’s AHEW follows from its measured amine content; the calculated column is a structure value. For triethylenetetramine, a mixture, the grade’s value is the only one.
- The resin’s EEW, and for a blend or a reactive diluent each component’s EEW and weight, so the dose is computed by equivalents.
- Isomer content for mixtures. Methylcyclohexanediamine hydrogenated from an 80/20 2,4-/2,6-toluenediamine feed runs about 85-90% 2,4- and 10-15% 2,6-isomer (US 5,025,078); US 8,828,267 counts seven isomers in it. For 4,4′-diaminodicyclohexylmethane in resins and paints, US 7,091,384 wants a trans,trans share below 30%, preferably 15-25%; ask for the share on the CoA of each RS-PACM grade.
- Any accelerator, with its basis stated: on active curing agent or on total binder.
- For electronics encapsulating, potting or laminates, the resin’s hydrolyzable chloride, which degrades electrical properties (US 4,624,975).
- For the dianhydride, the anhydride-to-epoxide ratio the formula targets.
Every CAS number in the table is Active on EPA’s non-confidential TSCA Inventory (July 2026 update), and both methylcyclohexanediamine isomers carry a PMN flag. An Active listing is an Inventory status, not an approval, and does not by itself rule out other TSCA requirements such as a significant new use rule, so confirm the status for your own use.
Send the request from the product page, which carries the quote form.
No patent example above is a specification for a RawSource grade: the lot CoA governs, and your own cure trial decides suitability. Patents are cited as technical literature only: some may be in force, and nothing here is a license or a freedom-to-operate opinion, so check patent status for your own formulation and market. Formulation questions go to technical support. These materials are for industrial formulation and manufacturing use, not for retail or consumer sale.
Related epoxy pages
| Page | What it covers |
|---|---|
| Epoxy curing agents: types and selection guide | amine, anhydride and catalytic hardener types |
| Choosing the epoxy hardener by amine chemistry and AHEW | selection by chemistry, purchase by AHEW |
| What causes amine blush and how to prevent it | amine blush |
| Why an epoxy stays tacky or soft | off-ratio and under-cure |
| Why clear epoxy yellows | aromatic vs cycloaliphatic hardeners and UV |
| Polyetheramine hardeners (diamines and triamines) | the child hub: polyetheramine grades |
| Dosing polyetheramines by AHEW and EEW | mix-ratio arithmetic |
Frequently asked questions
How do I calculate parts per hundred resin (phr) from AHEW and EEW?
Divide the hardener’s AHEW by the resin’s EEW and multiply by 100. One amine hydrogen equivalent reacts with one epoxy equivalent, so phr = AHEW × 100 / EEW.
In US 4,800,222, a polyoxypropylene diamine of molecular weight about 230 runs at 32 parts per 100 parts of liquid epoxy resin. For a blend of resins or of hardeners, divide the total weight by the sum of each component’s weight over its own equivalent weight, then use that figure in the phr formula. A non-reactive component adds weight without adding equivalents, so it raises the blend’s equivalent weight.
Why is triethylenetetramine bought to an amine value rather than a calculated AHEW?
Commercial triethylenetetramine is a mixture of linear, branched and cyclic congeners, so no single structure gives its AHEW. The lot’s measured amine split does.
US 2008/0114094 reports one commercial TETA at an amine value of 1435 mg KOH/g and an AHEW of 27. Linear TETA carries two primary and two secondary nitrogens, so the secondary term is never minor: AHEW = 1,000 / (2 × primary amine meq/g + secondary amine meq/g), with tertiary amine contributing nothing. A total amine value alone does not give AHEW, because it also counts the tertiary nitrogens of the branched and cyclic congeners. Ask for the primary, secondary and tertiary split, or a stated AHEW, and write that basis into the specification so two lots dose the same.
What causes amine blush, and how do I reduce it?
Amine blush is carbonate and carbamate salt formed when primary amino groups take up carbon dioxide from the air. It shows as a tacky, hazy or spotted surface and worsens in cold, moist cure.
US 8,912,294 states that hardeners carrying mainly primary amino groups form these salts in storage too, so keep drums closed. Its own route to a low-blush hardener converts the primary groups with a diketone. US 8,501,997 attributes part of the defect to amine-resin incompatibility that drives amine to the film surface; that explanation is the patent’s own. More on what causes amine blush and how to prevent it.
Does propylene carbonate change the hardener dose?
Propylene carbonate raises the hardener dose, because it reacts with primary amines. Count each mole of it (102.09 g/mol, PubChem) as consuming one primary amine group, two amine-hydrogen equivalents, when you set the hardener demand.
In US 8741988, a hydrogenated DGEBA resin cut with propylene carbonate and cured with IPDA, dosed for resin plus carbonate, gelled in 519 min. Leave the carbonate out of the calculation and the resin ends up short of amine. Two equivalents per mole is the full-conversion basis, and the carbonate reacts with amines relatively slowly (US 6,228,979).
Why does a hardener swap at the same AHEW still need a cure trial?
AHEW fixes the amount of hardener and says nothing about the network that forms. Stoichiometry is not always calculable, so US 5,235,007 finds the working optimum by testing samples for maximum Tg (ASTM D3418) or heat deflection temperature (ASTM D648).
Structure moves cured properties at a near-equal dose. In the same filled adhesive in US 2008/0114094, a propylene-spaced polyamine mixture rich in N,N’-bis(3-aminopropyl)ethylenediamine gave 2180 psi lap shear and 11.5 pli peel against 1850 psi and 9.0 pli for TETA. That result belongs to the mixture, not the pure compound.
What happens if the hardener dose is off stoichiometry?
An off-stoichiometry hardener dose moves the network away from its optimum. US 8741988 puts optimum thermoset properties at about 1:1 reactive groups to epoxide groups.
Overdosing amine is no safety margin. The introduction to a 2025 study in Polymers (PMC12656210) notes that excess amine leaves unreacted amine-terminated chain ends and less crosslinking, giving a tougher polymer that resists moisture and chemicals less well. If a cured part stays soft, see why an epoxy stays tacky or soft.
Sources & methodology
Figures are RawSource sourcing data unless attributed to a named source. Regulatory citations are current as of publication. Chemical identities verified by CAS number against the RawSource catalog.
Search 1,300+ industrial chemicals by name or CAS, or send us your spec — we quote by the drum, tote, or container.
Browse the Chemical Index → Request a Quote