Polycarbonate Monomer Production Using Partially Neutralized Thiol Resin Catalyst

Commercial production of bisphenol A (BPA)—the dominant monomer for polycarbonate melt transesterification—has since the 1990s converged on a fixed-bed continuous process employing a partially neutralized sulfonic acid ion-exchange resin catalyst functionalized with covalently tethered thiol groups. The catalyst system represents a carefully engineered balance between the Brønsted acid sites necessary for acetone condensation with two phenol molecules and the nucleophilic thiol promoter that suppresses the formation of the undesired o,p-BPA isomer and other chromophoric by‑products such as chromans, indanes, and polyphenols. This text examines the production of polymer‑grade BPA with a catalyst resin wherein a defined fraction of the sulfonic acid groups is neutralized by an amine‑functionalized alkanethiol—typically cysteamine or 3‑mercaptopropylamine—yielding a resin that contains both free sulfonic acid moieties, sulfonate–ammonium ion pairs, and pendant thiol moieties within the same crosslinked styrene‑divinylbenzene matrix. The partial neutralization moderates the overall acid site density from a typical value of 4.8–5.2 meq H⁺ g⁻¹ (dry basis) to a working range of 2.6–3.8 meq H⁺ g⁻¹, while the thiol content is held at 0.8–1.5 mmol SH g⁻¹. This distribution profoundly influences the acetone conversion profile, the p,p‑BPA selectivity, the rate of secondary condensation leading to heavy oligomers, and the leachable sulfur species that later poison downstream melt polymerization catalysts.

How Does Partial Neutralization Alter Resin Acidity and Thiol Promoter Accessibility?

The simultaneous presence of free acid sites and neutralized sulfonate‑ammonium pairs on the resin backbone generates a heterogeneous acid‑strength distribution that has been quantified through Hammett indicator adsorption and temperature‑programmed desorption of ammonia. On a gel‑type resin with a nominal crosslink density of 2–4 % DVB and a dry volume capacity of 1.8–2.2 meq mL⁻¹, amine neutralization at a stoichiometric ratio of 0.25–0.45 relative to total sulfonic acid groups shifts the apparent pKₐ of the strongest remaining acid sites from approximately −2.2 to −0.8, as estimated by the protonation equilibrium of 4‑nitroaniline indicators. The thiol group introduced through the neutralising amine creates a local environment where the nucleophilic mercaptan can intercept the intermediate p‑isopropenylphenol carbocation before it dimerises or rearranges to the o,p‑isomer. Nuclear magnetic resonance studies on crushed catalyst beads swollen in a phenol‑acetone mixture confirm that the thiol sulfur atom resides within 4–7 Å of the sulfonate group when the ammonium cation is ion‑paired, thereby creating a bifunctional catalytic pocket. The neutralization degree must be controlled within ±3 percentage points during the catalyst manufacturing step because over‑neutralization beyond 0.48 fraction collapses the resin pore volume in the swollen state by 15–20 %, reducing the effective diffusivity of acetone–phenol adducts and lowering the observed reaction rate constant by about 30 %. This sensitivity is documented in pilot‑plant runs on a 500 mm internal diameter reactor where a shift from 35 % to 41 % neutralization dropped the acetone conversion from 93 % to 78 % at an inlet liquid hourly space velocity (LHSV) of 0.8 h⁻¹, while simultaneously increasing the concentration of the chromophoric impurity 4,4’‑(fluoren‑9‑ylidene)diphenol above the 20 ppm threshold specified in ASTM D6143‑15 for optical‑grade polycarbonate.

Industrial practice often targets a neutralization window that results in a residual acid capacity of 2.8–3.4 meq H⁺ g⁻¹ when the resin is conditioned in a phenol‑swollen state at 55 °C. The catalyst is pre‑treated with a mixture of phenol and a controlled amount of the mercaptoamine at 60–75 °C for 4–8 h under a nitrogen blanket to prevent oxidation of the thiol to disulfide, which would manifest as a rapid increase in the eluate’s iodine titration value. The resin is then washed with anhydrous phenol until the free amine concentration in the effluent falls below 5 mg kg⁻¹. A common failure mode observed during production‑scale catalyst loading occurs when the neutralization reaction exotherm is not adequately dissipated in the first 0.5 m of the bed, causing localized temperatures to exceed 95 °C, which permanently deactivates up to 12 % of the thiol sites via elimination to hydrogen sulfide and formation of sulfide‑bridged crosslinks. The reactor must therefore be loaded in multiple lifts with a phenol‑circulation cooling loop capable of removing 50–70 kW m⁻³ of bed volume.

Fixed‑Bed Reactor Dynamics: Pressure Drop and Maldistribution Thresholds

Commercial BPA synthesis units operate with adiabatic or near‑isothermal fixed‑bed reactors arranged in a lead‑lag configuration containing total catalyst volumes of 30–120 m³. The resin is supplied as spherical beads with a harmonic mean diameter of 0.6–0.85 mm and a uniformity coefficient below 1.4 (per ISO 11289‑1). The pressure drop across a single bed of 6 m height at a superficial liquid velocity of 2.5–4.0 mm s⁻¹ ranges from 0.8 bar to 2.2 bar, depending on the degree of resin swelling and the accumulation of high‑boiling oligomers. When the catalyst has been partially neutralized to 35 %, the equilibrium swelling ratio in phenol at 60 °C increases by 8–12 % compared to the fully acid form, which reduces the interstitial void fraction to 0.32–0.36. If the bed is not mechanically restrained by a top‑screened distribution plate with a slot width of 0.2 mm, resin lift and bed expansion of 10–15 % create low‑flow channels at the wall, dropping the plug‑flow Peclet number below 15 and raising the o,p‑BPA content in the reactor outlet to 2.5–3.0 wt% of the total BPA fraction. This is unacceptable for subsequent melt polymerization, where the o,p‑isomer acts as a chain terminator and shifts the molecular weight distribution of polycarbonate below Mw 22,000 g mol⁻¹ as determined by gel permeation chromatography with polystyrene calibration.

Process control loops are tuned to maintain the reactor inlet temperature at 55.0 ± 0.5 °C and the acetone‑to‑phenol molar feed ratio at 1:9.5 ± 0.2. Acetone concentration in the feed is limited to 8.5–9.5 wt% because at values above 10.5 wt%, the adiabatic temperature rise across the bed exceeds 22 °C, which, in concert with the partially neutralized resin’s lower acid site density, initiates a self‑accelerating deactivation sequence: the higher temperature promotes sulfonic acid group hydrolysis and generates free sulfuric acid that further catalyses aldol condensation of acetone to mesityl oxide, consuming phenol and releasing water that hydrolyses the sulfonate‑ammonium ion pairs. This autocatalytic loop has caused a bed temperature excursion to 118 °C within 40 min on a 2,400 mm diameter reactor, permanently reducing the catalyst’s thiol content by 60 % and requiring a full bed change with a downtime of 14 days.

In a lead‑lag scheme, the conversion in the primary reactor is deliberately limited to 60–70 % at an LHSV of 1.2–1.8 h⁻¹, while the secondary reactor operates at 0.4–0.7 h⁻¹ and 50–52 °C to achieve an overall acetone conversion above 99.5 %. The partially neutralized resin exhibits a pronounced diffusional limitation above particle diameters of 0.7 mm; the Thiele modulus for acetone at 60 °C is estimated at 1.4–1.9, making catalyst effectiveness factor sensitive to the intraparticle tortuosity induced by the ion‑pair clusters. Therefore, fresh catalyst beads are screened to remove the fraction larger than 1.0 mm, and attrition fines below 0.15 mm are limited to < 0.5 wt% to prevent pore plugging in the downstream guard bed containing a non‑functionalized macroreticular adsorbent.

The partially neutralized thiol resin catalyst, when operated at a bed exit temperature below 75 °C, yields a reactor effluent in which the unreacted acetone concentration is typically < 0.15 wt%. The effluent is immediately routed to a distillation train where phenol is recovered and recycled, while the crude BPA stream is purified by a sequence of melt crystallisation stages—often three falling‑film dynamic crystallisers operating with a jacket temperature ramp from 90 °C to 145 °C at a cooling rate of 0.5 K min⁻¹. The crystallisation yield of 60–68 % per pass is deliberately kept below the eutectic limit to achieve a product with a freezing point of 156.5 °C minimum, as per ISO 1014:2008 clause 5.2. Residual phenol in the dried BPA is reduced to < 50 mg kg⁻¹ through a final wiped‑film evaporation step under a vacuum of 1–2 mbar and a jacket temperature of 170–175 °C. It is at this stage that the influence of the resin’s neutralization level becomes most apparent: a resin neutralized to 0.40 fraction generates fewer heavy condensation products, reducing the colour of the isolated BPA to an APHA value of < 10 as measured per ASTM D1209 on a 10 % methanol solution, whereas a less neutralized resin at 0.25 fraction produces a product with a persistent yellow tint (APHA > 25) that cannot be economically removed by adsorbent beds alone.

When the BPA monomer is destined for melt polycarbonate polymerization using a diphenyl carbonate (DPC) transesterification route, the ionic leachables from the catalyst bed become the most critical quality parameter. The ammonium sulfonate ion pairs residing on the partially neutralized resin exhibit a finite solubility in the hot phenol stream; the leached total sulfur content in the crude BPA crystal feed to the polycondensation extruder must be held below 0.5 mg kg⁻¹. To achieve this, the BPA melt is passed through a guard bed of a strong‑acid cation exchange resin in the H⁺ form followed by a weak‑base anion exchanger, both preconditioned at 120 °C under nitrogen. The cation exchange resin traps residual sodium and the protonated amine fragments derived from the thiol promoter, while the anion exchanger captures free phenol‑soluble sulfonate species. Breakthrough of total organic sulfur above 1.2 mg kg⁻¹ in the treated BPA has been shown to reduce the intrinsic viscosity of the resulting polycarbonate from a target of 0.48 dL g⁻¹ (measured in methylene chloride at 25 °C per ISO 3105) to 0.41 dL g⁻¹ and to broaden the molecular weight distribution to a polydispersity index above 2.2, rendering the resin unsuitable for injection‑moulding applications with a required notched Izod impact strength above 60 kJ m⁻² (ISO 180/A).

Operational Boundaries During Catalyst Regeneration with Acidic and Reducing Washes

Because the partially neutralized thiol resin catalyst slowly accumulates fouling deposits primarily composed of polycyclic bisphenol condensation products and trapped iron from corroded upstream piping, its activity declines by 0.5–0.8 % per day of continuous operation. An in‑situ regeneration procedure is scheduled every 4–6 weeks, requiring a hot phenol flush at 120 °C for 8 h to dissolve weakly adsorbed organics, followed by a dilute hydrochloric acid wash at a concentration of 2.5–4.0 wt% HCl in methanol or aqueous phenol at 50 °C to remove metal cations that have exchanged onto the sulfonate sites. A crucial constraint emerges from the thiol functionality: exposure to dissolved oxygen during the acid wash rapidly oxidises the mercaptan groups to disulfides, with a half‑life of approximately 3 h at an air partial pressure of 0.01 bar. Consequently, the entire regeneration loop is blanketed with nitrogen containing less than 10 ppmv O₂ and is operated at a positive pressure of 0.5 bar. After the acid step, a reducing rinse with a 0.5 wt% aqueous solution of dithiothreitol (DTT) at 40 °C is applied for 2 h to cleave any disulfide bonds that may have formed, restoring the free thiol concentration to within 90 % of its original value. Failure to include this reducing step results in a permanent loss of 15–20 % of promoter activity per regeneration cycle and shifts the p,p‑selectivity downward by 1.5–2.0 percentage points, accompanied by a rise in the o,p‑BPA fraction to levels exceeding 12,000 ppm.

When the catalyst has been subjected to more than 8–10 regeneration cycles, morphological degradation becomes evident: the bead crush strength falls below 4 N bead⁻¹ (measured per ASTM D6503), leading to an increase in the fines fraction (< 0.3 mm) to 2–3 wt%. These fines accumulate in the bottom of the reactor and create localised zones of high pressure drop and preferential flow paths. At this point the entire bed is replaced. The discharged spent resin exhibits a loss of thiol content to a value of 0.2–0.3 mmol SH g⁻¹ and a total sulfur retention that is largely in the form of thermally stable sulfonate species. The disposal must comply with local hazardous waste regulations; the resin is drained of free phenol to a level below 500 mg kg⁻¹ and thermally treated in a cement kiln under controlled conditions.

The presence of the partially neutralized thiol resin catalyst downstream of the reactor, even in trace leachate form, creates a processing incompatibility with the most common transesterification catalysts employed in polycarbonate melt plants: alkali metal hydroxides, quaternary ammonium salts, and tetra‑arylphosphonium phenoxides. The ammonium sulfonate residues from the resin act as catalyst poisons, irreversibly binding the basic catalyst and reducing the transesterification rate constant by an order of magnitude. To mitigate this, an effective ionic purification train must maintain the content of nitrogen‑containing organics in the final BPA below 0.5 mg kg⁻¹. Batch‑to‑batch variance in the resin neutralization step—if the free amine is not completely rinsed—leads to carryover of cysteamine concentrations in the recycled phenol loop reaching 8–12 mg kg⁻¹, which, when concentrated in the crystalliser mother liquor recycle, increases the total nitrogen in the BPA product by 1–3 mg kg⁻¹ and visibly accelerates the yellowing of polycarbonate pellets during compounding at 280–300 °C. For this reason, an inline total nitrogen analyser based on oxidative combustion and chemiluminescence detection (ASTM D4629) is installed on the phenol recovery distillation column bottoms, and the alarm limit is set at 3 mg N kg⁻¹, above which the recycle phenol is diverted to a separate polishing step.

Integration of the monomer unit with a direct melt polymerisation line employing a horizontal ZSK‑type twin‑screw extruder with an L/D ratio of 48:1 imposes additional constraints on the BPA particle size and morphology. The BPA prills or pastilles fed to the extruder must have a bulk density of 600–700 kg m⁻³ and a particle diameter distribution with a D₅₀ of 1,200–1,800 µm. Prills that contain residual phenol above 80 mg kg⁻¹ soften and agglomerate in the feed hopper when the ambient humidity exceeds 60 % RH, requiring pre‑drying in a dehumidified air stream at 90 °C for 4–6 h. The extruder first barrel zone is maintained at 120 °C to melt the BPA without causing thermal degradation; if the BPA contains iron above 0.25 mg kg⁻¹, discolouration initiates already at the feed throat, giving a product with a yellowness index exceeding 1.5 (ASTM E313) that is unacceptable for clear optical media applications. The monomer specification thus closes a loop back to the catalyst selection: the partially neutralized resin, because of its reduced acid site density, leaches 50–70 % less iron from the corroding carbon‑steel reactor internals compared to a fully acid resin, contributing to a measured iron concentration in the crystallised product of 0.08–0.14 mg kg⁻¹ over a 12‑month campaign.

When the phenol feedstock contains trace amounts of alkylphenols—particularly meta‑cresol at levels as low as 15 mg kg⁻¹—the partially neutralized thiol resin exhibits a peculiar kinetic selectivity shift: the nucleophilic thiol promotes cresol incorporation into the BPA framework, forming monomethyl‑substituted BPA analogues that co‑crystallize and are not separated in the melt crystallisation step. These alkyl‑BPA impurities act as chain terminators or branching points during polycondensation, depending on the substitution position, and a cumulative cresol‑derived impurity level of 50 mg kg⁻¹ in the final monomer has been shown to depress the notched Izod impact strength of a medium‑viscosity polycarbonate (MFI 10 g/10 min at 300 °C/1.2 kg, ISO 1133‑1:2022) from 65 kJ m⁻² to 48 kJ m⁻². Thus, the phenol raw material specification for a unit running the partially neutralized thiol resin must include a maximum limit for total alkylphenols of 10 mg kg⁻¹ with a typical achievable figure after phenol distillation being 2–5 mg kg⁻¹.