N,N-Dimethylformamide (DMF), CAS 68-12-2, is a water-miscible dipolar aprotic solvent with the formula HCON(CH₃)₂, molar mass 73.09 g/mol, normal boiling point 153 °C, vapour pressure 0.36 kPa at 20 °C, and dielectric constant 36.7 at 25 °C. In acrylic fiber wet spinning, DMF is the primary solvent for polyacrylonitrile and acrylonitrile copolymers; its high donor number and dipolar aprotic character allow solvation of nitrile dipole associations without proton transfer to the polymer. The resulting spin dope is processed at polymer loadings of 20–28 wt%, filtered, deaerated, and extruded into a water/DMF coagulation bath, where phase inversion produces the as-spun filament. Because the same solvent is not only a process chemical but also a reproductive hazard classified as Repr. 1B H360D under EU CLP and Group 2A under IARC Monograph Volume 115, grade selection and exposure control are coupled engineering decisions rather than separate purchasing and safety functions.

The solvency of DMF for acrylic polymers is often compared with dimethylacetamide and dimethyl sulfoxide. DMF has a donor number of 26.6 kcal/mol and a dipole moment of 3.82 D; these parameters support the disruption of nitrile-nitrile dipolar associations in polyacrylonitrile. However, DMF is more readily absorbed through intact skin than many alternative spinning solvents, which is why grade-selection programmes in acrylic fiber plants include closed-loop sampling and automated analysis rather than open-pail sampling at the tanker discharge.

What Defines Fiber-Grade DMF for Wet Spinning?

For acrylic fiber wet spinning, selection of a fiber-grade DMF stream is not governed solely by assay. Water is the most process-critical impurity because DMF-water mixtures exhibit reduced thermodynamic solvent power for polyacrylonitrile; excessive water narrows the dope gelation window and raises filter pressure before extrusion. The purchase specification for fiber-grade solvent therefore typically sets water at ≤0.05 wt% by ASTM E203 Karl Fischer titration, acidity at ≤20 mg/kg as formic acid by ASTM D1613, and dimethylamine at ≤15 mg/kg by headspace gas chromatography or ion chromatography. Iron is controlled at ≤0.05 mg/kg by ICP-MS because dissolved iron promotes spinneret capillary fouling and gel specks in the as-spun tow. APHA color is held at ≤10 by ASTM D1209 to avoid visible yellow bodies that survive washing and drawing.

At the dope preparation skid, DMF is blended with dried acrylic polymer under high-shear mixing, then filtered through multiple stages and deaerated before the spinneret pack. A typical acrylic line may use a jacketed dissolver at 60–80 °C and a vacuum deaerator operating below 50 mbar absolute to remove dispersed air from the high-viscosity dope. Filtration is normally carried out through sintered metal or cartridge media with nominal retention in the 10–25 µm range; pressure rise across the pack is a practical indicator of gel accumulation caused by water ingress or contaminated recovered solvent. The filtered dope is metered by gear pumps to spinneret capillaries of 40–80 µm diameter and extruded into a coagulation bath containing 40–65 wt% DMF in water at 10–35 °C. Coagulation rate is governed by counter-diffusion of water into the filament and DMF out of the filament; tight control of bath composition within ±1 wt% DMF is required to avoid asymmetric skin formation and filament splitting.

Because diffusion rates in the coagulation bath are temperature-dependent, even a drift of 2 °C alters the skin-core gradient and can change the subsequent draw response. The solvent concentration in the coagulation bath is maintained by continuous overflow to the recovery skid and addition of recovered or fresh DMF. Published DMF-water mutual diffusion coefficients at 25 °C are on the order of 10^-10 m²/s to 10^-9 m²/s; plant-scale control therefore relies on bath concentration and temperature rather than on direct diffusion measurement.

Dope viscosity is a practical quality flag for solvent selection. A rotational rheometer reading at 25–60 °C on a 20–26 wt% polyacrylonitrile solution is recorded for each batch; a rise of more than 10 % from the baseline after water correction indicates gel formation or batch-to-batch variation in recovered solvent. In-line viscometers are preferred where the dope line feeds multiple spinnerets because they respond faster than laboratory sampling. Filtration pressure is recorded continuously; a pressure differential that doubles under constant throughput and temperature indicates the need to switch to the standby filter and inspect the spent media for gels.

When Recovered Solvent Carries Dimethylamine and Formic Acid

Thermal and acid-catalysed decomposition of DMF during solvent recovery generates dimethylamine and formic acid, and the recovered stream therefore requires continuous purge and neutralisation. In a closed-loop recovery train, water is removed from coagulation bath overflow by distillation; DMF is taken overhead or as a side draw, while high-boiling oligomers and polymer fines accumulate in the reboiler. If reflux ratios are kept too low or reboiler temperature exceeds approximately 150 °C at sump level, decomposition to dimethylamine and formic acid accelerates. Dimethylamine is basic and shifts the recovered solvent's titratable acidity profile unpredictably; formic acid corrodes carbon steel reboilers and piping and can catalyse further degradation. Recovered solvent used for dope make-up is therefore blended with fresh DMF to keep total acidity below 20 mg/kg and dimethylamine below 15 mg/kg. Some acrylic fiber lines maintain a permanent purge of 2–5 wt% of the recovered solvent as reboiler bottoms to prevent accumulation of non-volatile gel precursors. Published data for the exact purge rate at every aging recovered stream is limited; the rate is site-specific and depends on polymer ash, spin bath turnover, and reboiler metallurgy.

Recovered DMF quality is checked at the tank farm rather than only at the distillation column outlet. Sampling ports on recovered-solvent day tanks are fitted with closed-loop samplers to reduce operator exposure. The recovered solvent is held in stainless steel or lined carbon steel tanks; unstabilised carbon steel tanks are avoided when formic acid levels exceed 20 mg/kg. In corrosion-prone systems, a continuous side stream is passed through an ion-exchange or neutralising bed, and reboiler cleaning frequency is tied to pressure drop and iron-in-DMF measurements. The iron limit for recovered solvent entering dope make-up is generally aligned with the fresh-solvent specification at ≤0.05 mg/kg, but actual achievable values depend on the age of the piping and the proportion of recovered solvent in the blend.

Incoming DMF tanker transfers are typically performed under nitrogen pressure or pump suction with vapour balance lines returning displaced solvent vapour to the tanker. Samples are drawn through a needle valve into crimp-sealed vials and analysed within 24 h for water, dimethylamine, acidity, and color. A tanker that passes the fiber-grade profile is released to the day tank; a failed lot is quarantined and may be usable only in non-critical cleaning service after review. The analytical method for dimethylamine should achieve a reporting limit of 5 mg/kg or lower because the specification limit of 15 mg/kg alone does not provide enough warning when recovered solvent blending is considered.

Impurities that survive filtration pass into the coagulation bath and can concentrate in the draw baths. Iron and non-ferrous metals accelerate chromophore formation during hot drawing; dimethylamine can react with residual comonomer or oxidation products to form yellowing species. Because final acrylic fiber is graded for color and tenacity, solvent purity is monitored as an incoming raw-material parameter and not corrected downstream. Batches that fail the APHA color or dimethylamine limits are segregated for non-fiber uses or returned to the producer for redistillation. In transfer lines, dedicated DMF lines with nitrogen blanketing reduce water pickup from humid air.

Occupational Exposure Limits, Skin Notation, and Regulatory Classifications

DMF exposure control in an acrylic fiber spinning hall is driven by the convergence of a low odour threshold, dermal absorption, and reproductive hazard. The 8-hour time-weighted average exposure limits established by OSHA and NIOSH are 10 ppm (30 mg/m³) with a skin notation; the ACGIH TLV is 10 ppm (30 mg/m³) with a skin notation and A3 confirmed animal carcinogen classification. EU occupational exposure limit values set the same 8-hour TWA and a short-term excursion limit of 20 ppm (60 mg/m³), with a skin notation because DMF vapour cannot be regarded as the only uptake route. The EU CLP classification includes Flam. Liq. 3 H226, Acute Tox. 4 H312+H332, Eye Irrit. 2 H319, and Repr. 1B H360D. IARC Monograph Volume 115 classifies DMF as Group 2A.

Controlling Vapour Release During Coagulation Bath Turnover

Controlling vapour release during coagulation bath turnover begins with enclosure of the spinning cell and the first wash draw stand. Slot hoods along the spinneret face and coagulation bath surface are positioned so that solvent vapours are pulled away from the operator's breathing zone. A capture velocity of 0.5–1.0 m/s at the hood opening is maintained and verified with a thermal anemometer after each spinneret line cleaning. Dope mixing vessels and recovered solvent tanks are vented to a thermal oxidiser or carbon adsorption system; rupture-disk vents and sampling ports are connected to closed drains. Pumps used for DMF transfer are specified with double mechanical seals or magnetically driven designs to reduce flange and seal leaks.

Spinneret washdown and pack change are the highest short-term emission tasks in the spinning hall. During pack removal, residual dope in the distribution plate and filter media continues to release DMF vapour. A portable high-flow extraction hood with a capture velocity of 0.75–1.25 m/s at the facing edge is placed at the open spin beam before the pack is unbolted. The extracted air is routed through a carbon bed or thermal oxidiser; discharge to the roof without abatement is not used where the 8-hour OEL must be maintained in adjacent work areas. After the pack is removed, spinneret capillaries are cleaned in a closed ultrasonic bath containing DMF or a solvent blend under local exhaust, and the bath lid remains closed except during insertion and removal.

Solvent recovery operations create exposure points at distillation column reflux drums, condensate receivers, reboiler manways, and pump seal flushes. An absorber or vent condenser operating on the recovery column overhead reduces DMF losses and lowers the concentration in the non-condensable stream to the abatement device. Routine leak detection and repair programmes follow EN 15446:2008 for fugitive emission measurement, and component leak thresholds are set at 500 ppmv or lower for pump seals and valves. DMF-specific photoionisation detectors and colorimetric tubes are used for rapid field verification; full-shift personal sampling remains the reference method for compliance.

Personal air sampling follows EN 482:2021 performance requirements using calibrated low-flow pumps and adsorbent tubes; sample pumps are operated at 0.05–0.20 L/min for full-shift sampling to avoid breakthrough. Dermal protection is required because DMF crosses intact skin. Gloves selected for DMF service are tested under EN ISO 374-1:2016; butyl rubber or laminate gloves with permeation breakthrough times greater than 480 min are used for dope sampling and spinneret cleanout, while thin nitrile examination gloves are not acceptable for sustained contact. Operators entering the solvent recovery area wear chemical splash goggles and a faceshield; eye irrigation stations are located within 10 s walk. Biological monitoring for end-of-shift urine is used to detect total uptake that air sampling alone may miss, with N-methylformamide and N-acetyl-S-(N-methylcarbamoyl)cysteine measured as biomarkers. If biological monitoring results exceed the current ACGIH BEI or DFG BAT guidance, the exposure control plan is revised before restarting the affected job task.