Wet spinning of acrylonitrile copolymer stabilization precursor begins with extrusion of a degassed dope containing 18–24 wt% poly(acrylonitrile-co-methyl acrylate-co-itaconic acid) in dimethyl sulfoxide through a spinneret with 1000–6000 capillaries of 50–100 µm diameter into a coagulation bath where DMSO-water exchange fixes the as-spun gel morphology. The bath temperature is not a passive thermal condition; it directly alters the ratio of solvent out-diffusion to non-solvent in-diffusion, the gelation front velocity, the radial distribution of polymer-rich domains, and the ultimate residual solvent content before washing and hot stretching. In production-scale lines with immersion lengths of 1.5–4.0 m, coagulation bath temperature is typically maintained between 25°C and 70°C, with temperature setpoint deadbands of ±0.5°C achieved by shell-and-tube heat exchangers and recirculation rates of 10–25 L/min per spinneret position. The resulting as-spun fiber is characterized by single-fiber tensile testing according to ISO 5079:2020 and ASTM D3822/D3822M-14, with reported tenacity values for PAN precursors typically in the range of 3.0–7.0 cN/dtex and elongation at break from 8% to 18%. Because the stabilization reaction of the acrylonitrile copolymer depends on oxygen diffusion, cyclization kinetics, and residual solvent acid-base behavior, the coagulation bath temperature influences downstream oxidation ovens through skin-core morphology rather than only through mass transfer in the bath. Industrial experience from continuous wet-spinning lines indicates that batch-to-batch variation in coagulation bath temperature of more than 1.5°C produces measurable differences in precursor elongation and in the broadness of the differential scanning calorimetry cyclization exotherm measured according to ISO 11357-1:2016.

When Bath Temperature Exceeds 60°C in DMSO/Water Coagulation

At bath temperatures above 60°C, the binary mutual diffusion coefficients of DMSO and water increase sufficiently to advance the coagulation front before the oriented polymer chains fully relax, generating a dense outer skin with thickness from 0.5 µm to 2.0 µm and a core with lower radial fibril density. The rapid non-solvent ingress elevates osmotic pressure gradients and can produce teardrop-shaped macrovoids or central lumens visible in scanning electron micrographs at 5000× magnification. The dense skin retards subsequent water extraction of residual DMSO, leaving 0.5–2.5 wt% solvent after the first bath even when the bath is held at 70°C, which downstream wash baths must reduce to below 0.5 wt% before hot stretching above 110°C. Single-fiber tensile data for high-temperature coagulated precursor usually show higher initial modulus but lower elongation at break, with elongation measured by ISO 5079:2020 falling below 10% when skin thickness exceeds 1.5 µm. In oxidative stabilization ovens, the same dense skin limits oxygen diffusion into the fiber core, delaying the onset of cyclization and shifting the DSC exotherm peak upward by 5–12°C relative to precursor coagulated at 40°C. For production lines running 5000-filament tows at 5–10 m/min, overheating above 70°C also increases solvent evaporation from open bath surfaces and alters the water/DMSO ratio by up to 0.3 wt% within an 8 h shift, violating the statistical process control limits used for residual solvent. The operational boundary is therefore set at 68°C for DMSO/water baths with open-top design, and at 75°C only when closed hoods with vapor recovery and automatic solvent concentration control are installed.

At bath temperatures below 10°C, the mutual diffusion coefficients of DMSO and water are approximately 30–50% lower than at 40°C, producing a slower phase inversion front and a more open gel network that retains 3–6 wt% residual DMSO after the first coagulation bath. The higher residual solvent load increases the boiling point of the solvent-water mixture inside the filament and can cause filament fusion or surface defects when the wet tow enters hot-stretching baths above 110°C. Low-temperature coagulation also reduces the elastic modulus of the gel filament, limiting the maximum jet stretch ratio to below 1.0 without filament breaks on the first godet pair operating at 5–15 m/min. The under-coagulated fiber has a thinner outer skin, larger core fibril spacing, and lower radial density uniformity, which is measured by image analysis of epoxy-embedded cross sections at 10,000× after ion polishing. When this precursor enters a stabilization oven with zone temperatures from 180°C to 300°C, the higher residual DMSO content accelerates ionic cyclization mechanisms and can broaden the DSC exotherm, shifting the onset to approximately 210°C compared with 220°C for precursor coagulated at 40°C. The exotherm is recorded according to ISO 11357-1:2016 at a heating rate of 10°C/min under nitrogen. Operating below 5°C is generally incompatible with DMSO/water coagulation because ice formation becomes possible in the heat exchanger if the water content is high and because the recirculation pump suction pressure drops as viscosity increases. Published data for this specific configuration is limited; however, industrial wet-spinning lines with low-temperature coagulation typically require longer bath immersion and reduced line speed to reach equivalent as-spun quality.

How Does Radial Fibril Density Influence Stabilization Precursor Quality?

Radial fibril density in wet-spun PAN precursor is a direct structural consequence of the coagulation front velocity, which depends on bath temperature, solvent system, and dope solids concentration. In fibers coagulated at high temperature, the skin is dense and fibrils at the core are fewer and coarser, with core fibril diameters in the range of 0.1–0.5 µm, whereas lower bath temperatures produce more numerous fine fibrils and an open core network. The radial density gradient can be quantified by scanning transmission electron microscopy of cross sections after epoxy embedding and ion polishing, with image analysis thresholds set to distinguish polymer-rich fibrils from voids. A steep skin-core transition reduces transverse mechanical properties and leads to splitting during fiber handling, measured as fibrillation resistance under standardized tow spreading tests. More importantly, the radial fibril density determines the diffusion path for oxygen during thermal stabilization and the local concentration of residual solvent that catalyzes cyclization in the 200–300°C window. Precursor with a dense 1.5 µm skin shows incomplete oxygen uptake in the core at 250°C, measured by thermogravimetric analysis under air according to ISO 11358-1:2022, with mass increase reduced by 0.4–1.0 wt% relative to low-temperature-coagulated precursor. The resulting stabilized fiber exhibits skin-core differences in density and a broader carbonization mass loss profile. Tensile property acceptance for precursor is commonly based on ASTM D3822/D3822M-14, but that single-fiber standard alone is insufficient to detect radial density nonuniformity; cross-sectional image analysis and DSC exotherm shape per ISO 11357-1:2016 are required on production lots.

Industrial wet-spinning lines frequently avoid a single isothermal coagulation bath by using staged baths with different temperature setpoints. A first bath at 45°C fixes the outer skin, a second bath at 55°C accelerates solvent extraction from the core, and a third bath at 30°C reduces void growth before the first hot draw. The temperature difference between adjacent baths is maintained below 15°C to avoid thermal shock and filament fibrillation on the transfer godets running at differential speeds of 1.2–1.8 times the first bath take-off speed. Recirculation loops equipped with 10 µm absolute filters and in-line refractive index sensors maintain DMSO concentration at the bath inlet within 0.2 wt% of the target value and turbidity below 5 NTU. Heat exchangers are sized for 50–100 kW per 5000-filament position to remove the heat released by DMSO-water mixing and to hold bath temperature within ±0.5°C during line speed changes from 5 m/min to 10 m/min. Batch-to-batch variance arises when the ambient air humidity exceeds 70% RH, because open-top baths absorb water and lower the DMSO concentration, or when the air temperature shifts by more than 10°C between shifts. Production lines using this staged temperature profile report fewer broken filaments during the first wash draw, but the residual solvent after the third bath must still be below 0.5 wt% before the tow enters the hot-stretching unit.

Skin-Core Gradient Formation in Dimethylacetamide/Water Systems

In DMAc/water coagulation, the bath temperature window is typically narrower than in DMSO/water because DMAc has a different diffusion activation energy and higher vapor pressure; industrial baths are usually operated at 20–50°C rather than 25–70°C. The skin thickness produced at a given temperature is generally thinner in DMAc/water than in DMSO/water, and the residual DMAc content after the first bath can be higher, requiring 5–8 wash stages to achieve residual solvent below 0.5 wt%. NaSCN/water systems, in contrast, operate at lower temperatures of 10–40°C and produce a denser, more elongated skin due to the salt's strong dehydrating effect on the PAN copolymer. Table 1 summarizes representative industrial process ranges; exact values depend on dope solids, spinneret capillary diameter, and immersion length. Published data for direct comparisons of all three solvent systems at identical jet stretch and spinneret geometry are limited. In all cases, the radial skin-core gradient is controlled by the balance between polymer relaxation and solvent-non-solvent exchange, and bath temperature acts as the dominant kinetic variable.

Thermal stabilization of wet-spun acrylonitrile copolymer precursor is carried out in continuous forced-air ovens with zone temperatures from 180°C to 300°C, air velocities of 1–3 m/s, and residence times of 20–120 min depending on tow linear density. The coagulation bath temperature affects stabilization through residual solvent content, skin thickness, radial fibril density, and the cyclization exotherm shape. Precursor coagulated at 70°C in DMSO/water exhibits a DSC cyclization exotherm with onset near 220°C and a peak near 270°C at 10°C/min under nitrogen per ISO 11357-1:2016, while precursor coagulated at 10°C can exhibit an onset near 210°C and a broader exotherm due to the catalytic effect of residual DMSO. The broader exotherm is less desirable for commercial stabilization because it spreads the reaction over a wider temperature range and can reduce the uniformity of oxygen uptake across the tow. In thermogravimetric analysis under air according to ISO 11358-1:2022, low-temperature-coagulated precursor shows a mass increase of 0.5–1.2 wt% between 200°C and 250°C, whereas high-temperature-coagulated precursor remains below 0.5 wt% mass gain in the same interval due to skin-limited oxygen diffusion. Production ovens with tow mass above 3 kg/m in the oxidation zone are susceptible to thermal runaway when exothermic cyclization heat is liberated too rapidly; therefore the coagulation bath temperature setpoint is often biased lower by 2–5°C if the incoming dope has high comonomer content or if the residual solvent specification is too broad. The combination of skin-core morphology, DSC exotherm, and TGA oxygen uptake forms the basis for incoming precursor lot acceptance on stabilization lines.

Controlling Bath Temperature Across the Multi-Roll Wash Train

Maintaining coagulation bath temperature within ±0.5°C on a production wet-spinning line requires cascade control of bath temperature and heat exchanger coolant flow. Thermocouples are placed at the spinneret end, mid-bath, and outlet end, and the bath temperature loop sends a remote setpoint to a shell-and-tube heat exchanger using chilled water at 4–10°C or steam at 130°C for startup. The dope preheater maintains the filtered dope at 60°C before the spinneret, and the heat load introduced into a 30°C bath by dope flow alone can reach 40 kW per spinneret position. When line speed increases from 5 m/min to 10 m/min, the coagulation bath residence time drops from approximately 24 s to 12 s for a 2 m immersion, and the temperature controller must compensate for the increased heat load and reduced residence time. The wash train usually consists of 5–8 countercurrent stages at 40–60°C, and each stage is controlled independently to prevent thermal gradients that cause filament fusion and broken filaments. Experience from production lines indicates that a failed heat exchanger control valve can produce a bath temperature excursion of 1.5°C within 60 s, leading to a measurable shift in residual solvent and as-spun elongation. Operational boundaries include the need to pre-dry the coagulation room air when relative humidity exceeds 60% RH, because water absorption into the open bath alters the solvent-to-non-solvent ratio, and the incompatibility of aluminum heat exchangers with NaSCN or DMSO-water streams containing chloride residues; 316L stainless steel is specified for all wetted parts. The bath temperature control system is therefore integrated with the distributed control system historian to track setpoint deviation, heat exchanger valve position, and recirculation pump speed as batch release parameters.