Articles
Articles
Articles
For continuous non-phosgene polycarbonate feedstock synthesis, diphenyl carbonate (DPC) is produced by the metal oxide-catalyzed transesterification of phenol with dimethyl carbonate (DMC) within a process envelope that must simultaneously achieve high single-pass conversion of the methyl phenyl carbonate (MPC) intermediate and near-total retention of the active metal on the solid surface. The overall stoichiometry 2 C6H5OH + (CH3O)2CO → (C6H5O)2CO + 2 CH3OH is thermodynamically constrained, and the commercially relevant fixed-bed and reactive distillation configurations are operated at 150 °C to 220 °C with DMC/phenol molar ratios from 2:1 to 6:1 and liquid hourly space velocities from 0.2 h⁻¹ to 2.0 h⁻¹. Metal oxide candidates such as supported MoO3/TiO2, WO3/TiO2, CeO2, Sm2O3, ZnO/TiO2, and MgO are selected because their surface acid-base pairs activate phenol toward carbonyl attack without introducing homogeneous alkali metal residues. The DPC stream exiting the catalyst bed, however, is not a final product: it must be fractionated from excess phenol, unreacted DMC, MPC, anisole, cresol isomers, and trace water before it can be fed to a bisphenol A melt polycondensation reactor. In that downstream melt process, residual molybdenum, zinc, cerium, iron, sodium, chloride, and water each participate in chain-termination, branching, hydrolysis, or color-body pathways that reduce molecular-weight stability and light transmission. The melt polycarbonate grade DPC specification therefore becomes a catalyst design constraint rather than a post-reactor purification afterthought: oxide supports with acid sites that promote DMC decomposition to dimethyl ether and methylating side reactions produce anisole and cresol impurities that are difficult to separate from DPC by distillation, while oxide surfaces that release soluble phenoxide complexes elevate the total metals burden in the final purified DPC. The same acid-base bifunctionality that enhances turnover can also render the catalyst susceptible to leaching by hot phenol. Consequently, the development of a robust metal oxide catalyst for this route requires a multi-measurement evaluation that includes conversion and selectivity, leachable metal concentration in the liquid product, surface acidity by programmed desorption, thermal stability under calcination and regeneration cycles, and long-term fixed-bed pressure drop.
Hot phenol is a sufficiently strong nucleophile and coordinating solvent that even crystalline metal oxide phases undergo slow dissolution through the formation of surface phenoxide species, especially when the oxide is present as a submonolayer species on a high-surface-area support. For molybdenum oxide supported on TiO2, laboratory fixed-bed runs at 180 °C to 200 °C have shown that the liquid product can contain molybdenum at concentrations that increase from below 0.02 mg/kg during the first 100 h of time-on-stream to above 0.3 mg/kg after 500 h when the MoO3 loading exceeds the monolayer dispersion capacity. Published data for this specific configuration is limited to certain support preparations, and the absolute leached concentration varies with phenol purity, water content, and the degree of condensation of the molybdenum oxide species. The problem is not restricted to molybdenum; zinc oxide and magnesium oxide systems are more prone to leaching because zinc phenoxide and magnesium phenoxide are partially soluble in DMC and phenol. A catalyst that delivers 99% DPC selectivity in an initial fixed-bed screening still fails commercial qualification if the leached metal concentration in the distilled DPC exceeds the melt polycondensation tolerance. In a bisphenol A melt reactor operating at 280 °C to 320 °C, residual transition metals catalyze Fries rearrangement, oxidative coupling, and ester interchange side reactions that generate branched species, high-molecular-weight fractions, and yellow chromophores. Optical-grade polycarbonate is routinely tested for yellowness index according to ASTM E313-20 and total luminous transmittance according to ISO 13468-1:2019, and even trace metal contamination can shift the yellowness index by more than 0.5 units. Compounding-grade polycarbonate may tolerate slightly higher yellowing, but melt volume-flow rate stability under ISO 1133-1:2022 at 300 °C and 1.2 kg load is sensitive to chain termination and branching caused by residual metal sites. Industrial merchant DPC specifications for melt polycarbonate often require total sodium below 0.05 mg/kg, total iron below 0.05 mg/kg, total chloride below 0.1 mg/kg, water below 0.05 wt%, and APHA color below 20, although these limits are not uniform across producers and are tightened for optical applications. The analytical methods used for release include Karl Fischer titration according to ISO 760:1978 for water, gas chromatography with internal standard for phenol and MPC, inductively coupled plasma mass spectrometry after acid digestion for trace metals, and APHA color measurement according to ASTM D1209-05(2019). Because catalyst leachate species may be converted into organometallic complexes that co-distill with DPC rather than remain in the heavy fraction, a low metal concentration in the reactor effluent is not sufficient; the thermal stability of the isolated DPC during subsequent vacuum distillation and melt polymerization must be measured.
Because metal oxide surfaces expose both coordinatively unsaturated Mn+ Lewis acid sites and lattice O2− or surface hydroxyl Brønsted base sites, the relative abundance and strength of these sites determine whether the intermediate methyl phenyl carbonate is converted to DPC by nucleophilic attack of a second phenol or is consumed in O-methylation to anisole and ring alkylation to cresol. Supported molybdenum oxide on TiO2 is among the most studied systems because calcination at 450 °C to 550 °C generates isolated tetrahedral molybdenum-oxygen species that titrate the strongest Lewis acid sites of the anatase support and produce a catalyst with DPC selectivity above 94% under fixed-bed conditions at 180 °C to 210 °C and a DMC/phenol ratio of 3:1. Cerium oxide is active at lower temperature because of its high oxygen storage capacity and basic sites; however, its DPC productivity is often limited by accumulation of the methyl phenyl carbonate intermediate, and some studies report that a second bed or higher DMC/phenol ratio is required to drive the second transesterification to completion. Zinc oxide and magnesium oxide possess strong base sites that activate phenol, but both metals form soluble phenoxide species in DMC-rich reaction media, which shifts the burden to downstream metal scavenging and can disqualify the DPC for optical polycarbonate. Samarium oxide has been reported to deliver DPC selectivity in the range of 96% to 99% in laboratory fixed-bed runs, but the rare-earth content raises catalyst cost and supply-chain concerns. The apparent activation energy for DPC formation over metal oxide catalysts is generally reported in the range of 70 kJ/mol to 95 kJ/mol, with the higher values often associated with supports that contain strong acid sites and lower values associated with base-dominated surfaces. Deactivation studies have shown that coke formation on acid sites is the primary cause of time-on-stream selectivity loss, while base-dominated surfaces tend to deactivate by phenol and water adsorption. In continuous operation, the single-pass DPC yield is limited not only by catalyst activity but also by the equilibrium methanol concentration; methanol removal by nitrogen stripping or reactive distillation is essential to push the second transesterification beyond 70% conversion. The following table summarizes representative fixed-bed performance ranges reported in peer-reviewed laboratory studies using 316L tubular reactors, 1 g to 5 g catalyst charges, and gas chromatographic analysis of the liquid product.
| Catalyst system | Temperature range (°C) | DMC/phenol molar ratio | LHSV (h⁻¹) | DPC selectivity (%) | Primary limitation |
|---|---|---|---|---|---|
| MoO3/TiO2 (5–10 wt% MoO3) | 180–210 | 3:1 | 0.5 | 94–98 | Molybdenum leaching above monolayer coverage |
| CeO2 | 160–200 | 2:1 | 1.0 | 90–95 | Methyl phenyl carbonate accumulation and water sensitivity |
| ZnO/TiO2 | 180–210 | 4:1 | 0.8 | 88–93 | Zinc phenoxide leaching |
| Sm2O3 | 190–220 | 3:1 | 0.3 | 96–99 | Rare-earth cost and regeneration sensitivity |
| MgO | 150–180 | 3:1 | 0.5 | 70–82 | Low turnover and strong phenol adsorption |
| WO3/TiO2 | 190–220 | 4:1 | 0.5 | 90–96 | DMC decomposition to dimethyl ether on acid sites |
Published data for industrial-scale fixed-bed catalysts is limited; the values above should be interpreted as screening-level ranges rather than continuous-production guarantees, and each candidate must be re-evaluated with plant-specific phenol purity, DMC recycled streams, and pressure drop constraints.
At equilibrium, the single-pass conversion of phenol over metal oxide catalysts is constrained by the accumulation of methanol and methyl phenyl carbonate in the reaction mixture, so industrial configurations typically combine a fixed-bed reactor with a distillation sequence or embed the catalyst directly within a reactive distillation column. The DMC–methanol azeotrope at approximately 63.5 °C and 30 wt% methanol at atmospheric pressure prevents a simple topping column from producing methanol-free DMC; pressure-swing distillation, extractive distillation, or membrane-assisted vapor permeation is therefore required to recover DMC at acceptable purity for recycle. In fixed-bed mode, the catalyst is often loaded as 1.5 mm to 3.0 mm extrudates with a bed length-to-diameter ratio of at least 2:1 in a 316L or Hastelloy C-276 tubular reactor, and the liquid feed is distributed through a perforated plate to avoid channeling. The LHSV is maintained between 0.3 h⁻¹ and 1.0 h⁻¹ to provide sufficient residence time for the second transesterification while avoiding excessive back pressure. Upon exiting the catalyst bed, the reaction mixture is quenched to 80 °C to 120 °C and sent to a topping column that removes DMC and methanol overhead, followed by a phenol recovery column and a DPC purification system that may include a wiped-film evaporator or short-path distillation. DPC has a normal boiling point of approximately 301 °C to 302 °C and a freezing point of 80 °C to 83 °C, so all transfer lines and column bottoms must be heat traced with hot oil or low-pressure steam at temperatures above 85 °C to prevent solidification. Trace phenol and MPC can be removed by a finishing column under vacuum, but prolonged exposure of DPC to temperatures above 250 °C must be minimized because thermal degradation can form phenol, diphenyl ether, and color bodies. Hydrolysis is an equally important concern: water concentrations above 0.1 wt% promote DPC hydrolysis to phenol and carbon dioxide, and the resulting free phenol excursions can fail the downstream melt polycondensation feed specification. Storage tanks are therefore padded with dry nitrogen and often equipped with a molecular sieve vent dryer to maintain a dew point below -40 °C.
Calcination temperature is the dominant synthesis variable for supported molybdenum oxide because it controls the condensation state of molybdenum-oxygen species, the strength of the molybdenum-oxygen-support interaction, and the residual ammonium or nitrate content from the metal precursor. When MoO3/TiO2 is calcined below 400 °C, incomplete decomposition of ammonium heptamolybdate leaves residual nitrogen-containing species that can desorb during operation and alter the surface acid-base balance, while the molybdenum remains partly hydrated and more readily leached by hot phenol. The optimal calcination window for monolayer dispersion is generally between 450 °C and 550 °C, where the molybdenum forms isolated tetrahedral molybdenum-oxygen sites that suppress the strongest titania Lewis acid centers and limit DMC decomposition to dimethyl ether. Above 600 °C, crystalline MoO3 domains grow, the accessible molybdenum surface area decreases, and the anatase support may begin conversion to rutile, which lowers the total specific surface area and reduces the number of active sites per unit volume. Surface area is measured by multipoint nitrogen adsorption according to ISO 9277:2022, and surface acidity is typically quantified by temperature-programmed desorption of ammonia and carbon dioxide using a thermal conductivity detector or mass spectrometer. A calcination temperature that is too high also generates a more acidic surface that promotes the acid-catalyzed decomposition of DMC to dimethyl ether and carbon dioxide, a reaction that reduces DMC efficiency and introduces a gas-phase load that can disturb fixed-bed liquid distribution. In contrast, under-calcined catalysts may show higher initial DPC yield but lose activity rapidly because the weakly bound molybdenum species migrate into the phenol-rich liquid phase, which is observable as an increase in molybdenum concentration in the product after only a few hundred hours. Regeneration of coked catalysts is typically performed by oxidative calcination in air at 450 °C to 550 °C with a gas hourly space velocity near 500 h⁻¹, but repeated regeneration cycles gradually convert dispersed molybdenum species into crystalline MoO3 and can shift the product distribution toward anisole and cresol.
In the melt polycarbonate process, bisphenol A reacts with DPC at 280 °C to 320 °C under progressive vacuum, releasing phenol and building molecular weight through a transesterification mechanism that is strongly influenced by trace ionic and protic species. Sodium, potassium, iron, molybdenum, and zinc residues act as additional transesterification or branching catalysts, altering the relationship between torque, melt volume-flow rate, and number-average molecular weight that is specified for extrusion and injection-molding grades. A DPC lot that passes gas chromatographic purity can still fail in polycarbonate trials if its chloride or hydrolyzable chlorine content is too high because chloride species can generate acidic sites that cleave carbonate linkages and create unstable end groups. Water is similarly critical because it hydrolyzes DPC to phenol and carbon dioxide, shifting the DPC/BPA molar balance and limiting the maximum molecular weight that can be achieved under a given vacuum profile. The physical and optical properties of the resulting polycarbonate are routinely checked by tensile testing according to ASTM D638-14, melt volume-flow rate according to ISO 1133-1:2022 at 300 °C and 1.2 kg, yellowness index according to ASTM E313-20, and haze according to ASTM D1003-21. The following table lists typical polymer-grade DPC contaminant limits and corresponding test methods that appear in merchant specifications, although individual polycarbonate producers may impose more stringent internal limits for optical products.
| Quality parameter | Typical limit | Test method |
|---|---|---|
| DPC purity | greater than 99.8 area% | Gas chromatography with internal standard |
| Methyl phenyl carbonate | below 0.20 wt% | Gas chromatography with internal standard |
| Free phenol | below 0.10 wt% | Gas chromatography with internal standard |
| Water | below 0.05 wt% | ISO 760:1978 Karl Fischer titration |
| APHA color | below 20 | ASTM D1209-05(2019) |
| Total sodium | below 0.05 mg/kg | ICP-MS after acid digestion |
| Total iron | below 0.05 mg/kg | ICP-MS after acid digestion |
| Total chloride | below 0.1 mg/kg | Combustion ion chromatography |
| Total molybdenum, zinc, and cerium | below 0.1 mg/kg | ICP-MS after acid digestion |
At production scale, fixed-bed catalyst life is rarely governed by intrinsic deactivation kinetics alone; instead, pressure drop, trace water ingress, phenol-derived oligomers, and recycle stream composition determine whether a metal oxide catalyst can remain in service for an acceptable cycle length. Start-of-run operation over a freshly calcined MoO3/TiO2 catalyst often yields a DPC selectivity close to the upper end of the screening range, but end-of-run operation is characterized by a shift toward MPC and anisole as acid sites are covered by coke and the remaining base sites become more dominant. Fixed-bed pressure drop can increase by a factor of three to five over 1000 h when fine catalyst particles are generated by thermal cycling or when heavy phenol-derived byproducts deposit in the void fraction; this pressure drop increase is exacerbated in multi-tubular reactors with long bed lengths because uneven flow distribution creates hot zones and accelerates deactivation. Production-scale reactors therefore use rugged extrudate geometries, controlled start-up heating rates, and feed filtration to remove particulate matter before the catalyst bed. Catalyst regeneration is performed ex situ or in situ by air calcination at 450 °C to 550 °C for 8 h to 12 h, but after three or more regeneration cycles the catalyst frequently exhibits a permanent selectivity loss toward dimethyl ether and anisole due to crystalline MoO3 formation. The feed phenol and DMC must be dried to below 0.05 wt% water because water competes with phenol for the basic surface sites and promotes both catalyst hydrolysis and DPC hydrolysis. The DPC product must not be stored or transported in contact with residual alkali, amine-based additives, or wet air, because residual alkylamines can catalyze DPC hydrolysis and generate free phenol excursions that violate the polymer-grade specification. Batch-to-batch variance in cumene-derived phenol may also introduce trace oxygenated aromatic compounds that deposit on Lewis acid sites and accelerate coking; the catalyst management strategy therefore includes periodic hot phenol washing, nitrogen stripping, and analysis of the recovered wash liquid for molybdenum and zinc. The operating window for melt polycarbonate grade DPC production is therefore a combination of catalyst calcination history, feed drying, fixed-bed pressure-drop control, distillation cut points, and inert product handling rather than a single reactor temperature or space velocity set point.
