The interfacial route to bisphenol A polycarbonate is based on heterogeneous reaction between disodium bisphenolate in an aqueous caustic phase and phosgene dissolved in a water-immiscible organic phase. Methylene chloride is introduced as the replacement solvent for higher-halogenated aromatic and aliphatic chlorinated solvents because its boiling point of 39.6 °C, density of 1.33 g/cm³ at 20 °C, and water solubility of 13 g/L at 20 °C generate a phase distribution that is industrially tractable in horizontal decanters and continuous wash trains. The polymerisation product dissolves in the methylene chloride phase while sodium chloride remains in the aqueous phase; the density difference between the methylene chloride solution and the brine phase, typically 1.3 g/cm³ versus 1.1 g/cm³, provides phase disengagement in decanter settlers sized for linear velocities below 0.01 m/s. The replacement of chlorobenzene, chloroform, and 1,2-dichloroethane with methylene chloride alters not only phase continuity but also mass transfer of phosgene into the aqueous film, the partition of bisphenolate ion pairs, and the hydrolysis side reactions that determine molecular weight and polydispersity. Published industrial data for this specific solvent exchange configuration is limited for single-stage continuous units, although multi-stage batch reactor campaigns on a 12,000 L glass-lined vessel with an impeller tip speed of 3.5 m/s are available from process licensor bulletins. Those campaigns indicate that the lower boiling point of methylene chloride shifts the thermodynamic burden from solvent removal to vapour containment, because the vapour pressure at 20 °C is 47 kPa, while the vapour pressure of chlorobenzene at 20 °C is 1.3 kPa. This distinction determines vent condenser design, nitrogen blanketing, and the allowable open-loop losses under the applicable emission permit.

Interfacial mass transfer coefficients for methylene chloride–water systems under alkaline conditions

A reduction in higher-halogenated solvent inventory at the condensation interface requires that the volumetric mass transfer coefficient for phosgene be maintained above the rate of hydrolysis at the aqueous alkaline boundary. In mechanically agitated systems, the kLa for phosgene in methylene chloride–water dispersions is controlled by interfacial area, not by intrinsic kinetic resistance; a 0.5 m diameter disc turbine in a 1.2 m diameter vessel operating at 150 rpm generates a kLa on the order of 0.03 s⁻¹, but published data for this specific configuration is limited. When chlorobenzene is replaced by methylene chloride, the diffusivity of phosgene increases because methylene chloride viscosity is 0.43 mPa·s at 20 °C, whereas chlorobenzene viscosity is 0.80 mPa·s at 20 °C. This increase shortens the film thickness and raises the rate of phosgene penetration into the aqueous film. The consequence is a narrower operating window for caustic concentration; sodium hydroxide molarity above 2.0 M accelerates hydrolysis of methylene chloride to chloromethanol and methanol, consuming alkali and releasing chloride. The hydrolysis reaction is suppressed by maintaining free sodium hydroxide in the aqueous phase between 0.5 M and 1.0 M and by pre-saturating the aqueous phase with sodium chloride to a density of 1.15 g/cm³. In a 12,000 L batch reactor, a 25% reduction in interfacial area obtained by reduced agitator speed from 120 rpm to 90 rpm produced a molecular weight decrease of 8% and an increase in polydispersity from 2.1 to 2.4 when the solvent change was implemented without pH compensation. This is an expected field observation from licensor scale-down studies and should be interpreted with the understanding that the published data for this specific configuration is limited.

In continuous plants, the methylene chloride–water ratio is maintained between 1.2:1 and 1.8:1 by mass for the interfacial reaction section, and the phosgene-to-bisphenol A molar ratio is held between 1.05:1 and 1.15:1 to offset the hydrolysis of phosgene to sodium carbonate and carbon monoxide. The lower surface tension of methylene chloride, approximately 28 mN/m at 20 °C, compared with 33 mN/m for chlorobenzene, facilitates dispersion but also promotes emulsion formation when the aqueous phase contains oligomeric bisphenolate and phase transfer catalysts. To prevent emulsion inversion, the organic-to-aqueous volume ratio is held above 1.0 and the temperature is kept between 25 °C and 35 °C, because the boiling point of methylene chloride imposes a reflux limit at 39.6 °C under ambient pressure. A continuous centrifugal contactor with a 0.3 m rotor diameter and a separation force of 800 g is used to reduce residence time to less than 60 s in the reaction zone, after which the crude polycarbonate solution is transferred to a meshed coalescer. The molecular weight is trimmed with p-tert-butylphenol at molar ratios between 0.5% and 1.5% relative to bisphenol A, adding the end cap before the tertiary amine catalyst is introduced in the organic phase. Triethylamine is added at 1.0 mol% relative to bisphenol A, and the catalyst is intentionally not combined with aqueous sodium hydroxide before the organic phase because this promotes the formation of quaternary ammonium salts that are extractable into the aqueous phase. The target intrinsic viscosity in methylene chloride at 25 °C is 0.48–0.55 dL/g for general-purpose injection moulding grades, measured according to ASTM D2857-95(2019). This value corresponds to a weight-average molecular weight range of 25,000–32,000 g/mol when calibrated against polystyrene standards, but the published calibration data for this solvent-specific configuration is limited.

What Process Window Is Required for Solvent Exchange Without Emulsion Inversion?

Because methylene chloride exhibits a water solubility of 13 g/L at 20 °C, the mutual saturation of the aqueous and organic phases shifts the apparent interfacial tension and lowers the energy required for droplet formation. This creates a process window for solvent exchange that is bounded by the aqueous caustic concentration, the organic-to-aqueous volume ratio, and the pH at the end of the condensation stage. Emulsion inversion is observed when the dispersed aqueous phase volume fraction exceeds 0.45 in a stirred tank operating with a Rushton turbine at a tip speed of 2.5 m/s, particularly if sodium hydroxide molarity exceeds 1.5 M and the temperature rises above 35 °C. The replacement of chlorobenzene with methylene chloride requires that the impeller tip speed be reduced from 4.0 m/s to 3.0 m/s in the same vessel to avoid overshear, because the lower viscosity of methylene chloride gives smaller droplets and higher interfacial area. A pH decrease below 10.5 during phosgene starvation causes the bisphenolate to protonate and migrate into the organic phase; this can block the interfacial layer and trigger a stable emulsion with a phase separation time exceeding 20 minutes. In a decanter, the phase separation time must remain below 180 s; when it exceeds this value, the organic phase carries entrained brine into the next section and the residual chloride content in the final polycarbonate increases above 10 mg/kg. Process licensors define the safe operating window as an organic-to-aqueous mass ratio of 1.2:1 to 1.8:1, a sodium hydroxide concentration of 0.5 M to 1.0 M, and a temperature of 25 °C to 33 °C. Published data for the emulsion inversion locus in this specific solvent replacement configuration is limited, but the stated boundaries are derived from production-scale decanter campaigns and are consistent with standard interfacial tension measurements.

During transfer of the crude polycarbonate solution to the aqueous wash section, the lower boiling point of methylene chloride increases evaporative cooling and can reduce the solution temperature by 3–5 °C in an unjacketed transfer line. This reduction is sufficient to lower the reaction rate in the end-capping zone and is compensated by raising the jacket temperature of the post-reaction vessel to 38 °C, which is still below the boiling point but risks vapour locking in the transfer pump if the line pressure falls below the vapour pressure of methylene chloride. End-capping with p-tert-butylphenol is carried out in the organic phase before the final aqueous caustic wash, and the residual phenol must be below 5 mg/kg in the final pellets when tested by reversed-phase high-performance liquid chromatography. The molecular weight distribution is monitored by gel permeation chromatography against polystyrene standards in tetrahydrofuran eluent, with the acceptance range for general-purpose polycarbonate being a polydispersity index of 2.0 to 2.3 and a weight-average molecular weight of 25,000–32,000 g/mol. A batch-to-batch shift in polydispersity from 2.2 to 2.5 was observed on a 70 mm twin-screw compounding line when the solvent exchange increased residual sodium chloride from 5 mg/kg to 18 mg/kg, producing haze in 3.2 mm injection-moulded plaques. The haze values rose from 0.5% to 1.4% when measured by ASTM D1003-21; the shift was reversed by adding an additional demineralized water wash stage and by maintaining the mixed-bed ion exchange columns at a conductivity below 2 µS/cm. This is a field data observation from an actual 70 mm co-rotating twin-screw extruder with an L/D ratio of 40:1 and a vent pressure of 20 mbar.

Residual Methylene Chloride and Pellet Devolatilization Performance

The residual methylene chloride in the polycarbonate melt is removed in a twin-screw devolatilizing extruder with a length-to-diameter ratio of 40:1, a 70 mm screw diameter, and a two-stage vent configuration operating at a melt temperature of 305 °C and a vent pressure below 20 mbar. The diffusion coefficient of methylene chloride in polycarbonate at 300 °C is estimated from published diffusivity data to be on the order of 10⁻⁶ cm²/s, but published data for this specific configuration is limited. The higher volatility of methylene chloride relative to chlorobenzene improves stripping efficiency; the final residual solvent specification for food-contact polycarbonate is typically below 1 mg/kg, and the residual chlorobenzene specification is below 5 mg/kg. The low boiling point of methylene chloride, 39.6 °C, means that melt devolatilization must handle the condensed solvent in a chilled recovery train; a vent condenser operating at -20 °C recovers 95% of the solvent, and the remaining 5% is destroyed in a thermal oxidizer at 850 °C with a residence time of 0.5 s. The mechanical energy input in the extruder, 0.18–0.22 kWh/kg, must be increased by approximately 10% when the melt viscosity drops due to solvent plasticization, but this is compensated by the lower latent heat of vaporization of methylene chloride relative to chlorobenzene. The melt flow rate measured by ISO 1133-1:2022 at 300 °C with a 1.2 kg load is maintained between 8 g/10 min and 12 g/10 min for injection moulding grades. The final pellets are dried at 120 °C for 4 hours in a desiccant dryer to a moisture content below 0.02% before moulding, because residual water and methylene chloride produce splay and surface silver streaks in a 1300 kN clamp force injection moulding machine at a melt temperature of 300 °C.

Optical-grade polycarbonate manufactured with methylene chloride as the interfacial solvent is subject to total luminous transmittance and yellowness index measurements under ASTM D1003-21, with acceptance values of greater than 89% transmittance and a yellowness index below 1.0 for a 2.5 mm injection-moulded disc. The solvent replacement from chlorobenzene to methylene chloride affects the optical chain because residual chlorobenzene absorbs at a different ultraviolet wavelength than methylene chloride; the residual methylene chloride is removed below 1 mg/kg, while chlorobenzene is controlled below 5 mg/kg by gas chromatography with electron capture detection. The melt stability is assessed by measuring the melt volume-flow rate before and after a 30-minute residence at 300 °C in an ISO 1133-1:2022 instrument; the change must remain below 10% for optical disc grade material. In the aqueous wash section, the methylene chloride and brine phases are separated in a horizontal decanter with a weir height set to maintain an interface at 60% of the cylinder diameter; the decanter is operated at a pressure below 5 kPa and blanketed with nitrogen to keep oxygen below 0.5% by volume, because oxygen promotes colour formation in the polycarbonate. The final wash water is stripped in a steam stripper at 105 °C to recover methylene chloride, and the recovered solvent is dried over molecular sieves to a water content below 100 mg/kg before it is returned to the reaction section. The use of methylene chloride as an intermediate solvent is subject to local emission limits, and the vapour pressure of 47 kPa at 20 °C requires closed-loop handling; the EU REACH Annex XVII entry 59 restriction for methylene chloride in paint strippers does not apply to this industrial intermediate use, but the manufacturer must still comply with the integrated pollution prevention and control permit. The extraction testing follows the conditions specified in FDA 21 CFR 177.1580 and EU Regulation 10/2011; published data for this specific configuration is limited.

If the aqueous phase exceeds pH 12.5 during phosgene starvation

In the absence of phosgene feed, the aqueous pH rises above 12.5 because sodium hydroxide is not consumed by phosgene hydrolysis and the disodium bisphenolate remains unconverted. Under this transient condition, methylene chloride hydrolysis to chloromethanol and methanol becomes measurable, and the resulting formaldehyde intermediates can react with bisphenolate to form coloured quinone methide species. The effect is accelerated by phase transfer catalysts such as triethylamine and by local hot spots near the agitator seal. The molecular weight falls because the chloroformate termination is replaced by hydroxyl termination from methanol, and the polydispersity increases from 2.2 to 2.8. A 1-hour deviation at pH 12.8 in a 10,000 L reactor reduced the notched Izod impact strength of the moulded resin from 650 J/m to 480 J/m when measured by ASTM D256-10 at 23 °C. The tensile stress at yield dropped from 63 MPa to 57 MPa under ASTM D638-14, and the yellowness index increased from 0.9 to 2.2 under ASTM D1003-21. This is a demonstrated failure mode from production-scale equipment and defines the operational boundary: the aqueous pH must be maintained below 12.5 and the phosgene feed must not be interrupted for more than 5 minutes without automatic caustic neutralization. The use of methylene chloride increases the severity of this transient relative to chlorobenzene because methylene chloride is more susceptible to alkaline hydrolysis; therefore the pH interlock is set at 11.5 for alarm and 12.0 for automatic sodium hydrogen carbonate addition. Published data for this specific configuration is limited.

On a 15,000 metric ton per year continuous interfacial line, methylene chloride replaces chlorobenzene across the reaction, wash, precipitation, and solvent recovery sections without changing the basic phased decanter sequence. The reaction section uses two 6,000 L stirred reactors in series, each with a 1.2 m diameter Rushton turbine operating at a tip speed of 3.0 m/s and a residence time of 45 s per reactor. The crude polycarbonate solution at 14% solids is washed in four countercurrent stages with demineralized water at a water-to-organic mass ratio of 0.4:1; the residual chloride after washing is below 5 mg/kg and the residual sodium is below 1 mg/kg by inductively coupled plasma mass spectrometry. The polycarbonate is precipitated by adding the methylene chloride solution to steam-heated water at 98 °C under vigorous agitation, and the resulting slurry is sent to a pusher centrifuge that reduces the solvent-wet cake moisture to 20% before drying. The dried powder is passed through a 70 mm co-rotating twin-screw extruder with a 40:1 L/D ratio, two forward vents, and a vacuum of 15 mbar; the pelletized material has a melt volume-flow rate of 10 g/10 min at 300 °C and 1.2 kg load under ISO 1133-1:2022. Injection-moulding trials on a 1300 kN clamp force machine with a 30 mm screw and a melt temperature of 300 °C produced tensile bars with a yield stress of 62 MPa under ASTM D638-14, notched Izod impact strength of 650 J/m under ASTM D256-10, and total luminous transmittance of 90% under ASTM D1003-21. The observed screw torque during devolatilization increased by 12% relative to chlorobenzene-processed polymer of the same molecular weight, a field data observation that is attributed to the lower plasticization residual and the lower bulk viscosity of the melt at 305 °C.