Acrylonitrile (CAS 107-13-1) is an unsaturated nitrile with molar mass 53.06 g/mol and normal boiling point 77.3°C. It is the base monomer for acrylic fibre, acrylonitrile-butadiene-styrene, styrene-acrylonitrile, nitrile rubber, and adiponitrile via electrohydrodimerisation. Global capacity is dominated by propylene ammoxidation; direct propane ammoxidation accounts for a smaller share, while bio-based routes remain at pilot or demonstration scale. Downstream polymerisation is sensitive to trace aldehydes, hydrogen cyanide, water, and non-volatile residues. Therefore, route selection cannot be treated as a feedstock-price decision alone.

Different Acrylonitrile Production Routes: Can Process Technology Change Acrylonitrile Cost and Quality?

The route determines reactor configuration, catalyst chemistry, effluent composition, and separation load. Propylene ammoxidation over promoted bismuth molybdate produces acrylonitrile with high selectivity; direct propane ammoxidation uses a lower-cost alkane but generates more COx and unconverted feed. Two-step propane dehydrogenation followed by Sohio ammoxidation preserves the downstream product profile while adding a dehydrogenation unit. Industrial experience shows that cost and quality are coupled through yield, steam demand, salt disposal, and trace-impurity generation. A change in process technology shifts all of these simultaneously.

Table 1. Route comparison based on published process data.

Propylene ammoxidation is carried out at 0.5–1.5 bar gage and 400–450°C in a fluidized bed of promoted bismuth molybdate. The overall reaction is 2 C₃H₆ + 2 NH₃ + 3 O₂ → 2 C₃H₃N + 6 H₂O. The catalyst operates through a Mars–van Krevelen redox cycle; lattice oxygen oxidizes propylene, and gas-phase oxygen reoxidizes the partially reduced solid. Selectivity to acrylonitrile is typically 80–85 mol% at high propylene conversion. Byproducts include hydrogen cyanide, acetonitrile, acrolein, acrylic acid, CO, and CO₂. Unconverted ammonia is removed in a sulfuric acid quench, producing ammonium sulfate as a purge stream.

Commercial fluid-bed reactors operate with superficial gas velocities of 0.3–0.5 m/s and catalyst particle diameters of 20–150 µm. Internal vertical cooling coils recover reaction heat as steam. Cyclone trains return elutriated catalyst fines to the bed; cyclone efficiency above 99.9% is required to keep catalyst loss below 0.1 kg/t AN. Field experience from production-scale units identifies cyclone erosion and distributor plate plugging as recurring maintenance bottlenecks. A temperature excursion of ±2°C from target can increase COx selectivity measurably because over-oxidation of the allylic intermediate becomes kinetically favoured.

Separation cost is a function of water:acrylonitrile relative volatility and the mass flow of low-boiling compounds. Acrylonitrile forms a minimum-boiling azeotrope with water; therefore, the absorber and product column cannot produce dry material without an additional drying step. In the propylene Sohio route, the reactor effluent is quenched with water and acid-scrubbed; the organic stream is sent to a water absorber. The absorber is typically operated at 1.5–2.5 bar gage, and the product column is operated under vacuum to limit polymerisation. The absorber off-gas is incinerated or treated for CO and HCN control. Direct propane ammoxidation increases the mass flow of unconverted propane and propylene in the absorber off-gas, requiring higher-capacity recovery compressors or a more aggressive lights-removal column.

Direct propane ammoxidation replaces propylene with propane over Mo-V-Te-Nb-O or Sb/V/O catalysts at 400–500°C and near-atmospheric pressure. Propane is less reactive than propylene; reported acrylonitrile yields in patent and technical literature are generally 55–65 mol%, with higher COx and propylene intermediate formation. The lower selectivity increases ammonia consumption per tonne of acrylonitrile and increases the water and organic load in the recovery train. The direct route may also require a higher reaction pressure drop because of the larger catalyst bed or circulating-fluid-bed design. Published data for commercial direct propane ammoxidation units are limited, but the operating envelope is narrower than Sohio due to the trade-off between propane conversion and over-oxidation.

Two-step propane dehydrogenation avoids the direct-route yield penalty by converting propane to propylene upstream of a conventional Sohio fluid-bed reactor. Commercial propane dehydrogenation units operate at 550–650°C and near-atmospheric pressure, using noble-metal or chromium-based catalysts. The propylene intermediate must be cooled, compressed, and purified before entering the Sohio reactor. Overall acrylonitrile yield is the product of dehydrogenation yield and Sohio ammoxidation yield. The purification train sees the same impurity spectrum as merchant-propylene Sohio plants if methylacetylene, propadiene, and diene content in the intermediate propylene is controlled.

When Feedstock Differentials Rewrite the Cost Equation

Cost comparisons between propylene and propane routes are driven by regional feedstock spreads. Propylene availability is tied to naphtha cracker and refinery FCC co-product balances; propane is tied to natural gas liquids and seasonal heating demand. Regional propylene-propane differentials vary from 100 to over 800 USD/t depending on market conditions. A direct propane ammoxidation plant can tolerate lower selectivity if that spread is sustained, because the feedstock cost per tonne of product is lower. The penalty appears elsewhere: larger reactor volume, higher air demand, more steam in recovery, and higher CO₂ emission intensity. Published techno-economic data for specific commercial propane direct ammoxidation plants are limited, but the process is generally described as requiring a sustained feedstock advantage to offset lower selectivity and higher capital per installed tonne.

For two-step propane dehydrogenation plus Sohio ammoxidation, the additional fixed-capital block for the dehydrogenation unit is significant; process design studies often place it at 20–35% of total fixed capital for a grass-roots complex. The quality of the intermediate propylene from a dehydrogenation unit—particularly methylacetylene, propadiene, and diene content—must be controlled to avoid excessive heavy byproducts in the Sohio reactor. The energy balance shifts because the dehydrogenation unit consumes substantial fired duty, while the Sohio reactor generates steam; integration can recover some of that energy.

Polymer-Grade Specifications That Downstream Polymerisation Cannot Negotiate

The separation sequence differs between routes. Propylene Sohio effluent is quenched in water, acid-scrubbed to remove ammonia, and purified through absorber, extractive distillation, lights-removal, and product-column operations. Direct propane ammoxidation effluent contains unconverted propane and propylene, which increase the load on lights recovery; the same unit operations can be reconfigured, but the reflux ratios and steam consumption change. The quality risk is therefore not whether the direct route can theoretically produce polymer-grade acrylonitrile, but whether the impurity spectrum can be removed without excessive product losses or solvent degradation.

The extractive distillation column uses water as the extraction solvent to separate acrylonitrile from acetonitrile; the solvent ratio is adjusted based on the acetonitrile-to-acrylonitrile mass flow. A route that generates more acetonitrile per tonne of product increases solvent circulation and reboiler duty. Water is removed by azeotropic distillation or molecular sieve drying; final water content below 0.50 mass% is required for ABS and SAN polymerisation.

Typical published polymer-grade acrylonitrile limits are summarized below. Specific grades for acrylic fibre or ABS may be tighter than these generic limits.

Table 2. Typical polymer-grade acrylonitrile quality limits and test methods.

Acrylic fibre producers using solution-spinning processes are sensitive to aldehyde and nonvolatile residues because these species can form chromophores during subsequent thermal stabilisation. Carbon fibre precursor grade acrylonitrile is often specified with tighter limits on water and ionic impurities than fibre or bulk ABS grade. ABS and SAN producers require low water because residual water hydrolyses initiators or causes surface defects in injection-moulded parts. A route that generates a different aldehyde-to-nitrile ratio may therefore be acceptable for one downstream application but require extra purification for another.

Storage and handling are common across routes. Acrylonitrile requires monomethyl ether hydroquinone at 35–50 mg/kg to inhibit radical polymerisation. The monomer is classified under REACH as Flam. Liq. 2 and Acute Tox. 3; closed-loop recovery and continuous oxygen monitoring are standard. A route that generates higher hydrogen cyanide concentrations in the recovery train increases the hazard footprint of the plant. In food-contact applications, acrylonitrile copolymers must comply with FDA 21 CFR 180.22. Process technology selection therefore includes inherent safety, salt disposal from ammonia recovery, and wastewater organic load, not only yield and feedstock cost.