Plant-based and petro-based epichlorohydrin (ECH) remain chemically identical as an epoxy resin feedstock: the molecule is 1,2-epoxy-3-chloropropane, CAS 106-89-8, molecular weight 92.52 g/mol, boiling point 116.1°C at 101.3 kPa, and density 1.18 g/cm³ at 20°C. The distinction between glycerol-derived and propylene-derived ECH is therefore expressed in renewable carbon fraction, trace chlorinated co-products, and supply-chain certification rather than in epoxy functionality. In a continuous liquid epoxy resin unit, ECH is reacted with bisphenol A at an ECH/BPA molar ratio of 8:1 to 12:1 with staged aqueous sodium hydroxide addition, followed by unreacted ECH recovery in a wiped-film evaporator at 5–10 kPa absolute and 110–140°C reboiler temperature. Resin-grade ECH from both routes must maintain water below 0.05 wt% because hydrolysis generates 3-chloro-1,2-propanediol, a chlorohydrin that carries hydrolyzable chloride into the final DGEBA and disturbs stoichiometric amine cure.

What Limits Continuous Dehydrochlorination of Propylene-Derived Dichloropropanol?

The petrochemical pathway converts propylene to allyl chloride in a chlorination furnace at 480–520°C and 0.1–0.3 MPa; allyl chloride selectivity of 85–90% is obtained by maintaining a propylene-to-chlorine molar ratio of 3:1 to 5:1 and quenching the reactor effluent below 100°C within 5 s to suppress secondary chlorination. After compression and cryogenic recovery, allyl chloride is reacted with chlorine in water at 35–55°C. The hypochlorination step generates a mixture of 1,3-dichloro-2-propanol and 2,3-dichloro-1-propanol, with the 1,3-isomer typically representing 70–85% of the dichloropropanol mass. Continuous dehydrochlorination is then carried out with 10–20 wt% calcium hydroxide slurry or 20–32 wt% sodium hydroxide at 70–90°C. The controlling constraint is pH: below 11.5, dichloropropanol conversion falls and residual chlorohydrins rise in the crude ECH; above 12.5, epoxide-ring hydrolysis accelerates, producing glycidol and 3-chloro-1,2-propanediol and increasing polymer fouling in the stripping column. Production-scale continuous reactors are typically controlled at pH 11.7–12.3 with a liquid residence time of 10–25 min. Calcium-based systems exhibit higher solids loading in the recirculation loop but lower sodium salt handling; nickel alloy internals are preferred for the caustic zones because chloride stress-corrosion cracking has been observed in 304L stainless steel at operating temperatures above 80°C.

Downstream purification uses a two-column vacuum distillation train. The first column removes light ends such as allyl chloride, acetone, and methanol at 30–60 kPa; the second column separates ECH from high-boiling dichloropropanols and polymeric heavies at 10–30 kPa with a reflux ratio of 2:1 to 4:1. A residue purge of 1–3% of column feed prevents accumulation of glycidol and chlorinated oligomers that form azeotropes with ECH.

In the plant-based glycerol-to-ECH route, crude glycerol recovered from rapeseed, palm, or soybean biodiesel is first refined to 99.5–99.9 wt% glycerol and less than 0.1 wt% methanol. The purified glycerol is hydrochlorinated with anhydrous hydrogen chloride in a continuous bubble column or agitated reactor constructed from a nickel-chromium-molybdenum alloy at 110–130°C and 0.5–1.5 MPa; a carboxylic acid catalyst at 0.5–3 wt% liquid-phase loading directs substitution toward 1,3-dichloro-2-propanol. Water produced by the reaction is removed continuously because free water above 3 wt% shifts the reaction toward monochloropropanediols and slows catalyst turnover. The resulting dichloropropanol is then dehydrochlorinated with the same alkaline chemistry as the propylene route. Crude glycerol from biodiesel can contain sodium, potassium, fatty acids, and sulfate at concentrations that require ion-exchange and acid pretreatment; failure to remove divalent cations has been linked to catalyst fouling and elevated color in the distilled ECH. Published data for the specific impact of residual potassium on continuous hydrochlorination catalyst life is limited, but plant operators monitor feed potassium below 5 mg/kg as a preventive limit.

When a Liquid Epoxy Resin Train Runs Bio-ECH at ≥90% Renewable Carbon

Switching from petro-ECH to bio-ECH on an existing liquid epoxy resin train does not require a change in the ECH/BPA reaction stoichiometry if the incoming ECH purity remains ≥99.9 wt%, water is ≤0.05 wt%, and total dichloropropanols remain below 0.05 wt%. The coupling reactor—a 10–30 m³ glass-lined or nickel-alloy agitated vessel—is charged with bisphenol A and ECH at a molar ratio of 9:1 to 10:1, and 20–50 wt% aqueous sodium hydroxide is metered over 4–8 h at 65–85°C. After phase separation, the resin phase is washed with demineralized water at 80–90°C to reduce sodium chloride below 5 mg/kg. Unreacted ECH is recovered in a wiped-film evaporator at 120–140°C and 2–10 kPa absolute; heat transfer oil above 150°C or rotor tip speed above 3 m/s causes localized oligomerization and gel particle formation. Production-scale failure modes include brine carryover into the evaporator when the interface level sensor is fouled, producing resin turbidity above 10 NTU and hydrolyzable chloride above 300 mg/kg.

As an epoxy resin feedstock, ECH is converted into liquid DGEBA and higher-molecular-weight phenoxy resins; downstream consumption includes protective coatings, structural adhesives, glass and carbon fiber composites, electronic encapsulation, and waterborne epoxy dispersions. The feedstock origin does not alter the epoxide group concentration or the bisphenol A backbone, so qualification for these applications depends on residual monomer, hydrolyzable chloride, and color.

Hydrolyzable Chloride, EEW Drift, and Epoxy Resin Cure Response

The critical resin parameters for feedstock substitution are EEW, hydrolyzable chloride, and dynamic viscosity. EEW is measured by perchloric acid titration according to ASTM D1652-11; hydrolyzable chloride is determined by potentiometric titration of saponified resin according to ASTM D1726-11; dynamic viscosity is measured with a rotational viscometer at 25°C according to ASTM D2196-20. Standard liquid DGEBA from either ECH source exhibits EEW between 182 g/eq and 192 g/eq and viscosity between 8 Pa·s and 16 Pa·s at 25°C. In high-purity electronics-grade resins, hydrolyzable chloride is held at ≤100 mg/kg, while general laminating grades permit 200–300 mg/kg. A positive EEW drift of 5 g/eq relative to target reduces the calculated stoichiometric amine hardener demand by approximately 2.5%, which can leave unreacted oxirane groups, lower crosslink density, and reduce glass transition temperature below the 120°C control limit in anhydride-cured castings. Hydrolyzable chloride above 500 mg/kg has been associated with carbon steel mold corrosion during long cure cycles and with reduced shelf life of dicyandiamide-catalyzed single-component systems aged at 40°C and 75% relative humidity.

In a high-speed pultrusion die operating at 160–180°C, an EEW shift of 3–5 g/eq alters gel time from approximately 45 s to 52 s; the resulting under-cured profile may exhibit a glass transition temperature below 120°C and a Barcol hardness below 40 when measured by ASTM D2583-13. For filament winding of amine-cured composites, a viscosity drift of ±15% changes wet-out; therefore incoming resin lots made from bio-ECH are typically held to the same 8–16 Pa·s specification with batch-to-batch variation below 5%.

Feedstock Supply Chains and Mass Balance Attribution

Plant-based ECH is commonly supplied under mass-balance certifications such as ISCC PLUS because the glycerol or bio-naphtha feedstock may be commingled in shared pipelines and storage. The biogenic carbon content of the actual ECH batch is quantified by accelerator mass spectrometry according to ASTM D6866-24 Method C or ISO 16620-2:2019. Petro-based ECH typically measures below 1% biogenic carbon, whereas segregated plant-based grades report 85–100% biogenic carbon depending on the glycerol source and allocation method. A 100% biogenic carbon result does not necessarily prove physical segregation from fossil-derived material if the sample was taken from a commingled mass-balance tank; it means the biogenic fraction of the sampled product meets the certification scheme’s allocation rules.

Comparative cradle-to-gate greenhouse gas data for ECH routes depend heavily on the allocation of glycerol as a biodiesel co-product. Published producer life-cycle assessments generally report a reduction of 30–60% in cradle-to-gate global warming potential for glycerol-to-ECH compared with propylene-to-ECH when mass allocation is used for crude glycerine. The range is not a fixed material property and varies with the upstream oil crop, biodiesel process, electricity mix, and the treatment of land-use change.

Under the European CLP Regulation, ECH retains harmonized classification regardless of feedstock origin: flammable liquid category 3, acute toxicity category 3, skin corrosion category 1A, skin sensitization category 1, and carcinogenicity category 1B. REACH registration for CAS 106-89-8 covers both manufacturing routes. The standard release of ECH from plant-based glycerol does not create a separate substance under EINECS or TSCA.