Acrylonitrile (CAS 107-13-1) is a water-white unsaturated nitrile with a boiling point of 77.3 °C at 101.3 kPa and a density of 0.806 at 20 °C. The monomer is polymerised to polyacrylonitrile and copolymerised with styrene, butadiene and acrylates for ABS, SAN, acrylic fibre and nitrile rubber. The condition described by the phrase Acrylonitrile Water Content Out of Spec: Polymerisation Risk and Quality Loss arises when the Karl Fischer titre of a lot exceeds the producer limit, commonly 0.45 % m/m for bulk monomer. The immediate concern is not simple dilution of the unsaturated nitrile but a sequence of storage and downstream effects that can accelerate radical polymerisation and shift molecular weight in continuous polymerisation lines.

Water determination for acrylonitrile uses volumetric or coulometric Karl Fischer titration according to ASTM E203-16 or ISO 760:1978. Sampling must be performed through a septum closure under dry nitrogen, because open-container transfer can bias the result upward by several hundred mg/kg in humid air. Commercial producer certificates commonly allow water between 0.25 % m/m and 0.45 % m/m, with tighter limits applied for polymer-grade acrylonitrile used in high-tenacity acrylic fibre precursor or low-color ABS. When the value falls outside the specified limit, the combined failure mode is polymerisation risk in storage and quality loss in the consuming polymer plant.

Which Storage and Transfer Practices Increase Water Uptake and Polymerisation Risk?

Acrylonitrile is normally inhibited with monomethyl ether hydroquinone at 35–50 mg/kg. This inhibitor system depends on dissolved oxygen to regenerate the phenolic radical-scavenging form, so the headspace of a storage vessel cannot be inerted to zero oxygen without reducing stabiliser effectiveness. In field operation, storage tanks use a controlled oxygen/nitrogen headspace rather than a simple nitrogen blanket. When water ingress raises the moisture load, the inhibitor can partition into any free water layer, lowering the effective concentration in the organic monomer phase. The result is an under-inhibited monomer layer that can propagate thermal polymerisation at hot tank walls, in pump casings or in the vapour space, producing insoluble polyacrylonitrile deposits.

Water contamination is not uniformly distributed. In a 25 m³ day tank, water can concentrate near the bottom if the liquid is not recirculated; top and bottom samples may differ by more than 0.10 % m/m. A production-scale failure mode is water carryover from the azeotropic drying column following a reboiler steam increase. The overhead decanter cannot coalesce the condensed water within the available residence time, and the organic reflux returns water to the top tray. The symptom is a rising water trend at the day tank, which is confirmed by Karl Fischer titration before the lot is released. Transfer lines with 50 μm mesh filters then blind as soft polyacrylonitrile gel forms on the filter element.

Azeotropic Drying and Final Distillation Limits

Commercial acrylonitrile is dried during final distillation by exploiting its partial miscibility with water. The purification train includes a drying column, a two-phase overhead decanter and a wastewater stripper. Overhead vapour is condensed, the water-rich phase is separated and sent to the waste-water stripper, and the organic phase is refluxed. Because the boiling point of acrylonitrile is 77.3 °C at atmospheric pressure, poor decanter control becomes visible as haze or separate droplets in a clear sample. Reflux ratio is typically held between 1.0 and 1.5 for this separation; a higher ratio does not necessarily improve water removal if the decanter residence time is insufficient.

When a lot falls outside the 0.45 % m/m limit, the correction path depends on the water content. Below about 0.10 % m/m, polishing with a 3A molecular-sieve bed is technically feasible, but the bed must be sized for the adsorption exotherm and must be replaced or regenerated before the sieve loses capacity. Regeneration is performed at 220–250 °C under nitrogen. Above about 0.50 % m/m, the monomer is usually returned to the purification column or fed to the azeotropic drying section rather than treated as a simple storage problem. The operational boundary is not arbitrary: wet acrylonitrile can form two liquid phases and cannot be completely dried by simple atmospheric draining.

At a downstream ABS or SAN plant, the AN feed dryer is often a small guard bed rather than a full distillation system. The monomer is blended with styrene and fed to a continuous radical polymerisation reactor. Residual water in the monomer feed can deactivate certain metal-alkyl or peroxydicarbonate initiators and alter the initiation rate. If the water level in the AN feed rises from 0.15 % m/m to 0.45 % m/m, the reactor temperature profile may show a lower peak exotherm and a higher final residual styrene content, requiring a feed rate reduction or initiator trim. Quality loss in the final resin is measured by melt flow rate per ISO 1133-1:2022 and tensile strength per ASTM D638-14.

A high-water AN feed can shift copolymer composition by altering the local polarity of the polymerisation medium, which changes the relative solvation of styrene and acrylonitrile during chain growth. Published data for this specific configuration is limited; however, production plants compensate by tightening the water specification to ≤ 0.20 % m/m for impact-resistant ABS grades where rubber particle morphology is sensitive to interfacial tension. The recorded quality loss is often broader polydispersity, higher residual acrylonitrile in the devolatilised resin and a measurable colour shift in the finished pellet.

When Out-of-Spec Water Enters a Continuous Polymerisation Train

In a continuous solution polymerisation line for acrylic fibre precursor, the solvent is typically dimethylformamide or dimethyl sulfoxide. Water in the AN feed adds to the solvent water burden. At polymer concentrations above 20 % m/m, a small increase in water can reduce the cloud point and cause high-molecular-weight polyacrylonitrile to deposit on heat-exchanger surfaces. The viscosity number of the polymer measured by ISO 1628-1:2009 may fall when water acts as a chain-transfer impurity, but the simultaneous phase separation can increase apparent viscosity. The apparent contradiction is resolved by separating the effects: molecular weight decreases while gel particles and microgels increase.

The operational limit in such a line is narrower than the bulk monomer specification. Some producers set the maximum water in AN feed at 0.20 % m/m for 1.2–1.7 dtex fibre and 0.10 % m/m for high-tenacity precursor used in carbon fibre. The action limits are based on spinneret pack life, not merely on polymerisation conversion. When water exceeds the limit, the line experiences increased pack pressure and fibre breaks. The corrective sequence is to switch the AN feed to a dry lot, purge the feed line with dry nitrogen, and recheck water by coulometric Karl Fischer before restarting the pumps.

In an emulsion copolymerisation process for nitrile rubber, water in the AN feed has a direct effect on pre-emulsion stability. The monomer is metered into a water phase containing emulsifier; if the AN itself carries excess water, the organic feed is partially diluted and the water-to-oil ratio changes. Latex particle size measured by dynamic light scattering per ISO 22412:2017 shifts, and Mooney viscosity of the dried rubber measured by ASTM D1646-19a can move outside the grade band. The polymerisation exotherm is monitored through reactor jacket temperature; an off-spec water load is often detected as a lower peak exotherm per kilogram of monomer, which requires the operator to alter initiator feed or reduce the production rate.

At the monomer feed control station, the corrective action is to divert the off-spec lot back to the drying column and to inspect the mechanical seal barrier fluid for water ingress. The railcar unloading line is purged with dry nitrogen until Karl Fischer titration of a closed-loop sample falls below 0.20 % m/m. The event is recorded in the batch certificate with the analytical results and the standard method used.