What Controls Macrovoid Formation During Dry-Jet Wet Spinning of Polyacrylonitrile Hollow Fibres?

Acrylonitrile, a vinyl cyanide monomer with molecular weight 53.06 g/mol and normal boiling point 77.3 °C, is stored as polymer-grade liquid at ≤25 °C under a fixed gas blanket because the monomethyl ether of hydroquinone inhibitor requires dissolved oxygen to suppress radical polymerisation. Transfer systems handling acrylonitrile must be closed-loop, grounded, and inerted to prevent flammable vapour accumulation and toxic exposure; REACH and OSHA classify the monomer as a category 1B carcinogen and category 2 mutagen. Polymer-grade material is specified at ≥99.5 wt% acrylonitrile, with water limited to ≤0.50 wt% and acetonitrile typically controlled to ≤100 ppm. Distillation before polymerisation removes inhibitor; residual MEHQ above 10 ppm retards initiation and broadens molecular weight distribution. These monomer specifications are not theoretical. Batch-to-batch variation in water or hydrogen cyanide content alters the polymerisation exotherm, changes dope rheology, and produces hollow fibre wall defects that only become visible after solvent extraction.

For haemodialysis-grade hollow fibre membranes, the copolymer composition is fixed before phase inversion because it determines hydraulic permeability, solute sieving, and complement activation. Commercial membranes such as AN69 are produced from acrylonitrile and sodium methallyl sulfonate; the sulfonate comonomer introduces fixed anionic sites that improve wetting and heparin adsorption, while the nitrile groups provide tensile strength and thermal stability. During polymerisation, target weight-average molecular weight is maintained at 80,000–200,000 g/mol and polydispersity is controlled between 2.0 and 3.5 to prevent both brittle fibre rupture and excessive dope viscosity. Polymer synthesis uses free-radical initiation in aqueous suspension or solution; redox systems are preferred when lower polymerisation temperatures reduce chain branching. The resulting copolymer must be isolated to low residual monomer and dried to ≤0.4 wt% water before dope preparation. Published data for this specific configuration is limited, but production-scale batches with high residual acrylonitrile monomer generate extractables failures during later haemocompatibility assessment.

Dry-jet wet spinning of polyacrylonitrile hollow fibres is governed by ternary phase separation between polymer, solvent, and non-solvent. Dope solutions are prepared at 15–22 wt% polymer in dimethylacetamide, dimethyl sulfoxide, or N-methyl-2-pyrrolidone. Co-rotating twin-screw dissolvers with L/D 40:1 and vacuum degassing at -0.9 bar gauge are used for high-viscosity batches because entrapped air produces lumen discontinuities. Dope zero-shear viscosity is adjusted to 8–30 Pa·s at 25 °C; below 8 Pa·s membrane walls collapse, and above 30 Pa·s spinneret pressures exceed safe filtration limits. The dope is filtered through sintered metal candles at 10–25 µm absolute retention before extrusion through a tube-in-orifice spinneret with outer diameter 1.0–2.0 mm, annular gap 0.1–0.3 mm, and bore diameter 0.4–0.6 mm. Bore fluid, typically water or a solvent/water mixture, flows at 20–100 mL/h. Air gap is maintained at 10–80 mm and take-up speed at 10–40 m/min with jet stretch ratio 1.5–3.0. Coagulation bath temperature is held within ±2 °C of setpoint because pore size and hydraulic permeability shift measurably outside this band.

Macrovoid formation is the primary defect in PAN hollow fibre spinning. It occurs when rapid solvent/non-solvent exchange creates localised osmotic pressure gradients that nucleate finger-like cavities. Suppression requires delaying demixing by increasing dope viscosity, adding solvent to the coagulation bath at 5–20 vol%, raising air-gap humidity, or reducing bath temperature. On manufacturing lines, macrovoids become visible as radial density bands in fibre cross-sections and correspond to low wet-burst pressure. Other failure modes include lumen collapse when bore fluid flow is too low or air gap exceeds 80 mm, ovality from spinneret concentricity drift, and draw-induced fibrillation when jet stretch exceeds the critical draw ratio. Fibres exiting the coagulation bath require solvent exchange and glycerin post-treatment at 20–40 wt% to prevent pore collapse during drying. Residual NMP or DMSO in finished fibres is limited to ≤0.1 wt% for dialysis use because solvent leaching into blood is not permitted.

In municipal surface-water clarification and tertiary wastewater polishing, polyacrylonitrile hollow fibres are installed in outside-in pressurised modules operating at transmembrane pressures of 0.5–2.0 bar. Ultrafiltration-grade PAN membranes are specified at nominal pore diameters of 10–50 nm and molecular weight cut-off between 50,000 and 100,000 Da. Hydrophilicity from the nitrile and sulfonate groups reduces organic fouling relative to hydrophobic polyvinylidene fluoride, but oxidative cleaning is limited. Industrial cleaning protocols for PAN UF modules restrict sodium hypochlorite to 200 ppm for 30 min per clean-in-place cycle because repeated exposure reduces tensile strength. Air scouring at 60–90 L/(m²·h) assists particulate removal during backwash. Module integrity is verified by bubble point or pressure decay testing; a loss of integrity below 0.8 bar indicates fibre breakage or potting failure. PAN hollow fibres are not recommended for continuous free chlorine above 1 ppm or prolonged operation outside pH 2–10 because hydrolytic degradation of nitrile groups accelerates. Published data for this specific configuration is limited, but field reports from municipal plants show that high iron or manganese oxide loadings require pre-treatment to avoid rapid irreversible fouling.

Extractables, Residual Monomer, and Haemocompatibility Test Points for Dialysis-Grade PAN Membranes

Regulatory test points for acrylonitrile-based hollow fibre membranes used in haemodialysis include ISO 10993-1:2018 biological evaluation planning, ISO 10993-4:2017 haemocompatibility, ISO 10993-5:2009 cytotoxicity, ISO 10993-10:2021 skin sensitisation, and ISO 10993-11:2017 systemic toxicity. Dialysis fluid quality is addressed by ANSI/AAMI/ISO 23500, which sets chemical and microbiological limits for water and dialysate. Extractables profiling of finished PAN fibres follows USP 661.1 plastic materials of construction and ISO 10993-18:2020 chemical characterisation. Residual acrylonitrile monomer in the finished medical device must be controlled at the lowest technically achievable level because the monomer is a known animal carcinogen; gas chromatography with headspace sampling is used to quantify residual monomer down to parts-per-million levels. Published data for this specific configuration is limited because residual limits are device-specific, but manufacturers typically release dialysis membranes only after residual solvent and monomer batches pass the extraction profile defined in the regulatory submission.

Sterilisation compatibility is an operational boundary for PAN hollow fibres. Ethylene oxide is commonly used for dry modules; steam autoclaving above 121 °C may cause fibre shrinkage and pore collapse for some acrylonitrile copolymers. Gamma irradiation at 25–40 kGy can induce chain scission and reduce elongation at break, so dose mapping is required for terminal sterilisation. Pan membranes should not be combined with amine-based additives during production because residual amines promote nitrile cyclisation and embrittlement. Storage of dry or wet modules above 40 °C accelerates physical ageing of the polymer matrix and shifts pore-size distribution. Finished hollow fibre membranes must be packaged with moisture control when relative humidity exceeds 60 % because PAN fibres absorb water and can develop microbial contamination before use. These limitations define the practical manufacturing window: careful monomer purity control, dope filtration, spinning parameter stability, and post-treatment integrity testing are inseparable from final dialysis and water treatment performance.