The purity requirements for THF for PTMEG & Spandex are not defined by a single assay value but by a narrow impurity matrix that changes meaning at each downstream step. In a ring-opening polymerisation train, tetrahydrofuran acts as a monomer; water, aldehydes, 2,3-dihydrofuran, and peroxide species compete with propagation or terminate chain ends. A feedstock with 99.90% THF can still yield PTMEG with unacceptable hydroxyl functionality if water exceeds 0.01 wt% or if carbonyl species exceed 0.005 wt% as butyraldehyde. Continuous polymerisers using fluorosulfonic acid or acetic anhydride/perchloric acid initiation show that water is a bifunctional chain-transfer agent; each mole of water consumes two propagating centres, altering the number-average molecular weight M_n and broadening the molecular weight distribution. Commercial specification documents therefore treat THF purity as a kinetic parameter, not a commodity solvent property. Downstream spandex dry spinning and wet spinning processes then amplify these effects because PTMEG hydroxyl number, water, acidity, and colour bodies shift isocyanate stoichiometry and phase separation.

What Impurity Thresholds Govern Cationic Ring-Opening Polymerization to PTMEG?

Water is the highest-leverage impurity. In a 5 m³ jacketed continuous reactor with external circulation and static mixing, a water increase from 0.005 wt% to 0.02 wt% can require a catalyst feed increase of 10–15% to hold target M_n at 2000 g/mol; otherwise the hydroxyl number rises above the 53.6–59.6 mg KOH/g window typical for spandex-grade PTMEG 2000. Aldehydes and 2,3-dihydrofuran are monofunctional chain-transfer agents or initiation sites that create low-molecular-weight chains and vinyl end groups. Peroxides in THF generate colour bodies and can crosslink PTMEG during finishing. Typical PTMEG-grade THF certificates of analysis specify water at ≤0.01 wt% by ASTM E1064, colour at ≤5 Pt-Co by ASTM D1209, distillation range within 65.0–66.5°C by ASTM D1078, and density at 0.8850–0.8890 g/cm³ by ASTM D4052. Carbonyl as butyraldehyde is commonly limited to ≤0.005 wt% by GC-FID calibrated against certified reference materials. The structural incompatibility is also clear: retained high-boiling aldehydes and acetals behave as monofunctional terminators in cationic ring-opening polymerisation, while water behaves as a bifunctional chain-transfer agent.

On production lines using a 48:1 L/D twin-screw devolatiliser followed by a thin-film evaporator, residual THF monomer and low-molecular-weight cyclic oligomers are stripped to below 0.1 wt%. This step is not only physical; it also carries through the chain-transfer history of the feed. A feed shift from 0.005 wt% to 0.025 wt% water may not be fully corrected by increasing catalyst, because the resulting PTMEG contains a higher population of terminal hydroxyl groups from water-initiated chains, and the polydispersity index can move from 1.8 to 2.3. Such broadening alters the hard-segment orientation in spandex, reducing tensile set recovery and elongation at break measured by ASTM D2731 or ASTM D412. Spandex producers frequently detect these upstream shifts as changes in dope filtration pressure rather than as immediate fibre tenacity loss.

Spandex-Grade PTMEG Quality Flags and Isocyanate Stoichiometry

PTMEG with hydroxyl number near 56.0 mg KOH/g for a 2000 g/mol diol has a hydroxyl equivalent weight of approximately 1000 g/eq. Prepolymer formulations with methylene diphenyl diisocyanate typically operate at NCO/OH ratios between 1.6:1 and 2.5:1. Water carried from PTMEG interrupts this stoichiometry. At 0.02 wt% water in PTMEG, the isocyanate equivalent consumption approaches 0.022 eq/kg, enough to shift NCO/OH ratio by 0.02 in a prepolymer formulation, causing viscosity drift and hard-segment stoichiometry mismatch. Acid residues from catalyst neutralisation are equally consequential: acid number above 0.02 mg KOH/g can deactivate diamine chain extenders and produce a sticky prepolymer with poor spinnability. Sodium and potassium ions from neutralisation or from THF recycle loops cause gelation and filter plugging in dry-spinning lines, particularly when total alkali metal content exceeds 1 mg/kg.

In a dry-spinning operation running 40-hole spinnerets and a 4:1 solvent draw ratio, the dope solution viscosity is typically controlled in the range 150–300 Pa·s at 20°C for 20–30 wt% solids. Fluctuations caused by PTMEG molecular weight drift translate into spinning tension changes. At draw ratios above 4.5:1, low-molecular-weight oligomers derived from THF impurities cause filament breaks at the spinneret and uneven denier. Fibre producers monitor elastic recovery by ASTM D2731, tenacity by ASTM D412, and yellowness index by ASTM E313. PTMEG colour above 20 Pt-Co can be carried through prepolymer and spandex yarn, leading to visible yellowing that cannot be corrected by optical brighteners in dry spinning. The THF purity matrix therefore becomes a latent variable in spinneret pressure drop and package uniformity.

When THF Storage Headspace and Inhibitor Dosing Shift Molecular Weight Control

THF for PTMEG polymerisation is typically stabilised with 50–150 mg/kg butylated hydroxytoluene. In many polymerisation trains, this radical inhibitor is not fully inert; at concentrations above 150 mg/kg it can deactivate cationic initiators, reduce catalyst productivity, and contribute to colour body formation in PTMEG finishing. If THF is stored at temperatures above 25°C with air headspace and no inhibitor top-up, peroxide values can rise from below 1 mg/kg to 20–50 mg/kg as hydrogen peroxide within days under UV exposure. Peroxides in recovered THF accumulate in distillation bottoms and can initiate radical side reactions that produce crosslinked gel particles. Process limits therefore require nitrogen blanketing, dark storage, and weekly peroxide analysis on recovered THF; the recovery column bottom temperature is held below 85°C to prevent thermal decomposition of peroxides into aldehydes. Published data for this specific storage configuration is limited; however, the stoichiometric consequence of water and acid species in the downstream isocyanate reaction can be calculated directly from the relevant ASTM and ISO test methods.

Analytical verification across the chain is not a single-method exercise. THF purity is commonly checked by GC-FID with internal standard, water by coulometric Karl Fischer per ASTM E1064, colour by ASTM D1209, and distillation behaviour by ASTM D1078. PTMEG hydroxyl number is determined by ISO 4629-2, acid number by ASTM D4662, molecular weight distribution by ISO 16014-2, and trace metals by ISO 11885. The analytical sequence should be operated with strict sample handling because THF is hygroscopic and peroxide-forming; samples exposed to ambient air during transfer can gain enough water to bias Karl Fischer results by 0.005–0.01 wt%. Practical incompatibility is also relevant: recovered THF containing peroxides should not be mixed with strong acids or reducing agents during storage, and BHT-stabilised THF should not be assumed inert in polymerisation catalyst feed evaluations.

Through the PTMEG and spandex chain, the operational boundary is set by the cumulative effect of small impurity shifts. A THF water specification of ≤0.01 wt% protects the polymerisation reactor; a PTMEG water specification of ≤0.02 wt% protects the isocyanate prepolymer reaction; a PTMEG acid number of ≤0.02 mg KOH/g protects the diamine chain-extension step. When the THF feed contains 0.005 wt% carbonyl species as butyraldehyde and 0.005 wt% peroxide as hydrogen peroxide simultaneously, the defects are not additive but synergistic: low-molecular-weight chains from aldehyde-initiated centres reduce elastic recovery, while peroxide-derived chromophores increase yellowness index. This is why PTMEG producers control THF feed quality within a narrow matrix rather than accepting a commodity solvent specification.