Articles
The selectivity of cumene oxidation to cumene hydroperoxide (CHP) is the primary economic and safety boundary in a phenol/acetone train, because every mole of CHP that decomposes in the oxidizer rather than in the downstream acid cleavage reactor represents a direct loss of phenol and acetone yield. The oxidation reaction proceeds by a free-radical chain in which the cumyl radical adds molecular oxygen to form a cumylperoxy radical, which then abstracts the tertiary hydrogen from a second cumene molecule to yield CHP and regenerate the cumyl radical. Industrial oxidizer batteries are operated at 110–130°C and 1–10 barg, with per-pass cumene conversion deliberately held between 20% and 30% because the molar selectivity to CHP remains in the 90–95 mol% range only within this window. Above 30% conversion the local CHP concentration increases, and side reactions such as homolytic decomposition, acid-catalysed decomposition, and condensation with dimethylphenylcarbinol (DMPC) begin to dominate. The dominant byproducts formed in the oxidizer are DMPC, acetophenone, dicumyl peroxide (DCP), formic acid, acetic acid, and trace phenol. DMPC is particularly problematic because it dehydrates to alpha-methylstyrene (AMS) in the cleavage reactor, and AMS is recovered and hydrogenated back to cumene only at additional capital and hydrogen cost. The oxidation train therefore operates as a selectivity-limited low-conversion reactor, and the unreacted cumene is separated from CHP in a vacuum distillation step and recycled to the oxidizer. The oxidation reaction is highly exothermic, and commercial designs use staged bubble columns or mechanically agitated gas-liquid contactors with internal cooling coils or external pump-around loops. Liquid height-to-diameter ratios in bubble column oxidizers typically fall between 6:1 and 10:1, and air or oxygen-enriched air is introduced through sintered metal distributors or pipe-lattice spargers. Selectivity is monitored by iodometric titration of the CHP concentration in the oxidizer effluent, while the off-gas oxygen concentration is analysed continuously to maintain operation below the limiting oxygen concentration for cumene-air mixtures. The oxidation section also includes a continuous water purge and an organic acid neutralisation step, because accumulation of formic acid and acetic acid lowers pH and promotes acid-catalysed CHP decomposition to phenol and acetone inside the oxidizer, reducing CHP selectivity and increasing the phenol concentration in the oxidation section. The central constraint is that CHP selectivity is not a fixed molecular property but a dynamic response to conversion, temperature, oxygen partial pressure, pH, and recycle impurity concentrations.
| Operating variable | Industrial control window | Selectivity consequence | Monitoring method |
|---|---|---|---|
| Per-pass cumene conversion | 20–30% | CHP selectivity remains above 90–95 mol%; above 30% DMPC, acetophenone, and DCP formation accelerates | Gas chromatography and iodometric titration of oxidizer effluent |
| Oxidizer temperature | 110–130°C | Higher temperatures increase acetophenone and DMPC formation through CHP homolysis | Thermowell input to kinetic model and distributed control system |
| Oxidizer pH | 6.5–8.0 | pH below 5.0 triggers acid-catalysed CHP decomposition to phenol and acetone in the oxidizer | Online pH measurement after water extraction |
| Air superficial velocity | 0.02–0.08 m/s | kLa values of 0.05–0.30 s⁻¹ are required; lower values produce oxygen starvation and radical termination | Sparger delta-pressure and off-gas oxygen analyser |
| Recycle acetophenone | 0.1–0.5 wt% | Higher concentrations alter oxidizer selectivity and increase hydrogen consumption in AMS hydrogenation | Gas chromatography of cumene feed |
If oxygen mass transfer falls below the stoichiometric demand of the free-radical chain, the steady-state concentration of cumyl radicals increases and termination reactions become significant. Termination by cumyl-cumyl coupling and cumylperoxy-cumylperoxy self-reaction consumes radicals that would otherwise propagate the chain, and the apparent selectivity to CHP drops even when the bulk liquid temperature remains within the 110–130°C control window. In industrial bubble columns, the volumetric mass transfer coefficient kLa depends on superficial gas velocity, sparger orifice diameter, liquid viscosity, and liquid height. At superficial air velocities of 0.02–0.08 m/s, kLa values are commonly reported in the 0.05–0.30 s⁻¹ range for low-viscosity hydrocarbon systems. Published data for cumene-specific kLa in full-scale oxidizers are limited because licensors treat the measurements as proprietary, but the general bubble column mass transfer literature establishes that kLa falls rapidly when the sparger is fouled or when the air compressor output is reduced. Oxygen starvation also increases the concentration of cumyl radicals available to abstract hydrogen from CHP, forming DMPC and acetophenone through radical-induced decomposition. The oxidizer off-gas oxygen concentration is therefore a critical control variable; if the off-gas oxygen concentration exceeds the target range, the vent stream approaches the flammable envelope, and if it falls too low, the reaction becomes mass-transfer limited. In practice, the off-gas oxygen concentration is maintained below the limiting oxygen concentration for cumene-air mixtures at the operating temperature and pressure, and the air supply is interlocked with a nitrogen purge and a deluge system. The oxidation section is normally subject to process safety management under 29 CFR 1910.119 because of the inventory of CHP and cumene. The air sparger itself must be designed to avoid plugging from sodium salts or tar precursors; sintered metal spargers with pore sizes of 10–50 µm are common in small reactors, but full-scale columns often use pipe-lattice spargers with larger orifices to reduce plugging risk. When oxygen depletion occurs, the first observable changes are a reduction in CHP concentration for a fixed feed rate and an increase in the DMPC to CHP ratio in the oxidizer effluent. The corrective response is to reduce cumene feed rate, increase air flow, and check the sparger pressure drop, because continuing operation under oxygen starvation can initiate a runaway decomposition if the CHP concentration remains high and the heat removal system cannot manage the exotherm.
Downstream of the oxidation battery, the crude CHP stream is washed with a dilute aqueous sodium hydroxide or sodium carbonate solution to remove formic acid, acetic acid, and trace phenol before the CHP is concentrated by vacuum distillation. The caustic wash is not a simple neutralisation step; it is a selectivity safeguard because any residual acid in the CHP concentrate can catalyse CHP decomposition to phenol and acetone during distillation. The wash solution is typically maintained at 1–3 wt% sodium hydroxide or sodium carbonate, and the pH of the aqueous phase after contact is held between 7.0 and 8.5. Higher caustic concentrations can form stable emulsions with the hydrocarbon phase, causing sodium carryover into the distillation train and fouling of reboilers. Lower caustic concentrations fail to neutralise organic acids, and the pH of the oxidizer effluent can drift below 5.0, at which point acid-catalysed CHP decomposition becomes significant. The wash system usually consists of a high-efficiency liquid-liquid contactor followed by a coalescer, and the spent caustic is stripped or incinerated. In plants that use sodium carbonate instead of sodium hydroxide, the buffering action is gentler and reduces the risk of localised high pH, which can promote aldol condensation of acetone in downstream distillation. The aqueous phase is separated by density difference, and interface level control is critical because a shifting interface can send water into the CHP concentrator and create a steam explosion hazard in the reboiler. The neutralised CHP stream is then sent to a vacuum distillation column where unreacted cumene is recovered overhead and the CHP concentrate is taken as a bottoms stream. The distillation temperature in the reboiler must be kept below the thermal decomposition threshold of CHP, typically below 100°C and preferably below 90°C, which requires vacuum operation. The cumene distillate is recycled to the oxidizer, but it carries residual DMPC, acetophenone, and AMS, so the recycle loop is not a clean cumene feed. The accumulation of these impurities is controlled by a purge or by hydrogenation of AMS back to cumene. The water content in the CHP concentrate entering the cleavage reactor must be below 0.5 wt% to avoid dilution of the sulfuric acid catalyst and to reduce the formation of acetone-water azeotropes in the downstream acetone purification column. This unlabelled scenario emphasises that pH control and water management are as important to CHP selectivity as the oxidation temperature and conversion.
The acid cleavage reactor is a back-mixed vessel in which CHP is decomposed to phenol and acetone by contact with concentrated sulfuric acid or an acid initiator at 60–90°C. The reaction is highly exothermic, and the temperature is controlled by external circulation through a heat exchanger and by the latent heat of acetone evaporation. The selectivity of the cleavage step to phenol and acetone is usually above 95 mol% per CHP consumed when the feed is clean, but the selectivity is constrained by the DMPC and DCP formed in the oxidizer. DMPC is protonated and dehydrated to AMS under the cleavage conditions, and AMS can dimerise or polymerise to heavies if not removed quickly. DCP can undergo acid-catalysed decomposition to phenol, acetone, and AMS, but it can also condense with phenol to form high-boiling tar. The solvent ratio of phenol and acetone to CHP at the reactor inlet is maintained high enough to limit the local temperature rise and to dilute reactive intermediates. Typically the molar ratio of phenol/acetone solvent to CHP is in the range of 2:1 to 5:1, and the residence time in the decomposer is held between 15 min and 45 min. The acid concentration in the reaction mixture is controlled by the addition of concentrated sulfuric acid, and the reactor pH is not measured directly in the organic phase; instead the acid number and water content are tracked. The cleavage reactor off-gas contains acetone and water vapour, and the condenser is designed to return reflux to the reactor while venting non-condensables. AMS recovery is usually performed in a separate column or in a hydrogenation unit, where AMS is hydrogenated back to cumene over a fixed-bed catalyst at elevated pressure. The hydrogen consumption for AMS hydrogenation is a direct selectivity penalty because it consumes hydrogen that could otherwise be used in another part of the complex. Acetophenone is mostly inert in the cleavage reactor, but it can be hydrogenated to methylphenylcarbinol in the AMS hydrogenation unit, which consumes additional hydrogen and can recycle back to the oxidizer as a DMPC precursor. The phenol/acetone molar balance across the cleavage section is therefore not exactly 1:1; it is typically shifted by the formation of AMS, acetophenone, and tar. Process licensors measure the molar yield ratio of phenol to acetone and the concentration of AMS in the cleavage effluent to monitor the selectivity of the entire train. The cleavage selectivity is protected by maintaining a narrow temperature window, avoiding water in the CHP concentrate, and minimising the DMPC and DCP content of the oxidizer effluent. If the temperature exceeds 90°C, tar formation accelerates and the pressure relief valves can be challenged by rapid gas evolution. If the acid concentration is too low, the CHP conversion is incomplete and unreacted CHP can carry into the distillation train, where it can decompose thermally and create a safety hazard. Published data for the exact tar formation rates at different DMPC:CHP ratios are limited, but operational experience shows that reducing DMPC in the oxidizer effluent improves cleavage selectivity more than any other single variable.
The recycle cumene loop is a selectivity constraint because impurities that survive the cumene recovery column are returned to the oxidizer, where they alter the free-radical chain. Acetophenone is not completely inert in the oxidation reactor; it can be hydrogenated or can act as a radical-chain transfer agent, and its accumulation raises the boiling point of the cumene recycle stream. AMS is even more problematic because it can polymerise in the oxidizer and form fouling precursors on heat transfer surfaces. Industrial licensors typically specify a maximum acetophenone content in the recycle cumene in the low single-digit weight percent range, and the AMS content is often kept below 0.2 wt% to avoid fouling. The cumene recovery column is operated under vacuum to keep the reboiler temperature below 100°C, and the column overhead is designed to produce a cumene stream with low CHP and DMPC content. However, the separation between cumene and acetophenone is difficult because their boiling points are close; cumene boils at 152°C and acetophenone at 202°C at atmospheric pressure, but under vacuum the relative volatility is reduced. The column therefore requires a high number of theoretical stages and a high reflux ratio to achieve the required acetophenone rejection. Some plants use a side-draw or a small purge stream to remove acetophenone and heavy byproducts from the recycle loop. The purge stream is sent to the phenol recovery section or to a heavy ends recovery column, where acetophenone can be recovered as a byproduct or hydrogenated to methylphenylcarbinol. The hydrogenation of AMS to cumene is conducted over a fixed-bed catalyst at 50–150°C and 10–30 barg hydrogen partial pressure, depending on the catalyst supplier. The hydrogenation reactor is exothermic and requires careful temperature control to avoid over-hydrogenation of the aromatic ring. The cumene recycle stream after hydrogenation is washed with water to remove trace acids and then dried before returning to the oxidizer. The water content in the recycle cumene must be below 0.1 wt% to avoid hydrolysis of CHP and to maintain the oxidative chain. The accumulation of acetophenone in the recycle loop can also shift the apparent conversion and selectivity because it changes the solubility of oxygen and the viscosity of the hydrocarbon phase. Published data for the exact selectivity response to acetophenone concentration are limited, but the general industrial practice is to control acetophenone below 0.5 wt% in the recycle cumene. The AMS and acetophenone purge rates are set by the oxidizer selectivity and the hydrogenation capacity, and they represent a continuous yield loss that is directly linked to the selectivity constraints in the oxidation section.
CHP is thermally unstable at the temperatures used for cumene oxidation, and the same free-radical intermediates that propagate the oxidation chain can also initiate CHP decomposition. The homolytic cleavage of the O–O bond in CHP produces a cumyloxy radical and a hydroxyl radical. The cumyloxy radical can undergo β-scission to acetophenone and a methyl radical, and the methyl radical can abstract hydrogen from cumene or from CHP to form methane or methanol. The cumyloxy radical can also abstract hydrogen from cumene to form DMPC and a cumyl radical. These reactions compete with the propagation step in which the cumylperoxy radical abstracts the tertiary hydrogen from cumene. At temperatures above 130°C, the rate of CHP homolysis increases, and the selectivity to CHP falls. The thermal decomposition of CHP is also catalysed by trace metals, particularly iron, copper, and manganese. Stainless steel surfaces can release iron ions if the pH drops below 5.0, and the presence of dissolved iron at parts-per-million levels can accelerate CHP decomposition and increase acetophenone and DMPC formation. The oxidizer pH is therefore maintained in the 6.5–8.0 range by adding sodium carbonate or sodium hydroxide to the aqueous phase. The addition rate is controlled by the pH of the water separated from the oxidizer effluent, and the target is to keep the organic acids neutralised without creating a high-pH environment that could saponify CHP or promote aldol condensation. The thermal decomposition of CHP is also sensitive to the concentration of CHP itself. In the oxidizer, the CHP concentration is kept below 35 wt% in the liquid phase to limit the rate of bimolecular CHP decomposition. The concentration of CHP is a function of the per-pass conversion and the cumene recycle rate. In the CHP concentrator, the reboiler temperature is kept below 90°C and the pressure is reduced to maintain the CHP concentration below 80 wt% in the bottoms. The CHP concentrator is a thin-film or falling-film evaporator in some designs, because the residence time at high temperature must be minimised. The decomposition of CHP in the oxidizer is not only a yield loss but also a safety hazard, because the decomposition products include acetone and phenol, which can lower the flash point of the reaction mixture and increase the vapour load on the relief system. The thermal decomposition pathways are monitored by measuring the acetophenone and DMPC concentrations in the oxidizer effluent and by tracking the CHP selectivity at constant conversion. If the acetophenone:CHP ratio increases while the conversion remains constant, the oxidizer is operating too hot or the pH is too low. The response is to lower the oxidation temperature, increase the caustic addition rate, or reduce the feed rate to increase the residence time and allow the chain to propagate more selectively.
The purification train downstream of cleavage separates the phenol and acetone products from AMS, acetophenone, mesityl oxide, and heavy tar, and the separation itself is constrained by the selectivity losses that occur in the oxidizer and decomposer. The cleavage effluent is neutralised with sodium hydroxide to stop acid-catalysed condensation, and the neutral salts are removed by water washing. The acetone purification column is operated under pressure to allow condensation with cooling water, and the acetone product must meet ASTM D329-07(2021) or an equivalent specification for purity, water content, and acidity. The phenol purification column is operated under vacuum to reduce the reboiler temperature and prevent thermal degradation of phenol. The phenol product is distilled to meet ASTM D2439-20 or an equivalent specification, with a crystallizing point above 40.6°C and a water content below 0.1 wt%. The heavy ends column recovers phenol from tar and discharges a residue stream that contains acetophenone, AMS dimers, and high-boiling condensation products. The tar residue is typically incinerated or used as fuel, and its flow rate is a direct measure of selectivity loss across the train. The AMS recovery column separates AMS from cumene and other light components, and the AMS is either hydrogenated back to cumene or recovered as a chemical-grade byproduct. The hydrogenation unit consumes hydrogen and generates heat, and its capacity is a constraint on the maximum AMS production that can be tolerated. The distillation train is designed with structured packing or high-efficiency trays, and the columns are interlocked to maintain feed composition and thermal stability. The reboilers are designed with low film temperatures and high circulation rates to avoid local hot spots that could degrade phenol or acetone. The entire purification section is subject to ISO 9001:2015 quality management requirements and to the site process safety management system under 29 CFR 1910.119. The phenol and acetone product specifications are verified by ASTM D2439-20 and ASTM D329-07(2021) test methods, respectively, and the analytical results are used to adjust the cleavage and distillation operating conditions. The selectivity constraints established in the oxidizer therefore propagate through the entire train, because every mole of DMPC that dehydrates to AMS, every mole of acetophenone that accumulates in the recycle loop, and every mole of DCP that condenses to tar reduces the yield of phenol and acetone and increases the energy and hydrogen consumption of the plant.
| Stream | Specification or test method | Typical control limit | Selectivity consequence |
|---|---|---|---|
| Phenol product | ASTM D2439-20 | Crystallizing point ≥ 40.6°C, water ≤ 0.1 wt% | Indicates low tar, AMS, and neutral salt carryover |
| Acetone product | ASTM D329-07(2021) | Purity ≥ 99.5 wt%, water ≤ 0.5 wt% | Indicates low acidity and oxidation residues |
| Recycle cumene | Gas chromatography | Acetophenone ≤ 0.5 wt%, AMS ≤ 0.2 wt% | Prevents oxidizer fouling and CHP decomposition |
| CHP concentrate | Iodometric titration | CHP ≤ 85 wt%, water ≤ 0.5 wt% | Limits thermal decomposition and acid dilution in cleavage |
Acetophenone can be removed from the cumene recycle loop by adjusting the cut point of the cumene recovery column and by taking a small heavy purge from the column bottoms. The relative volatility of cumene to acetophenone under vacuum is moderate, and the column requires a high reflux ratio and a large number of theoretical stages to produce a cumene overhead with low acetophenone content. In a typical cumene recovery column operating at 100–200 mmHg absolute pressure, the overhead temperature is 60–90°C and the reboiler temperature is 100–130°C. The column is designed with 60–80 theoretical stages and a reflux ratio of 3:1 to 5:1 to achieve the required separation. The overhead cumene stream is condensed and sent to a water separator, because water in the cumene recycle can hydrolyse CHP and reduce oxidation selectivity. The column bottoms stream contains the CHP concentrate and the heavier byproducts, including DMPC, acetophenone, and DCP. The CHP concentrator is often a separate thin-film evaporator that operates at a lower pressure and a shorter residence time to concentrate CHP to 75–85 wt% without exceeding 90°C. The acetophenone in the cumene recovery column bottoms is not easily separated from CHP by simple distillation because their boiling points are close under vacuum. Some plants use a liquid-liquid extraction or a side-draw rectification section to reject acetophenone to a separate purge stream. The purge stream is sent to a hydrogenation unit where acetophenone is hydrogenated to methylphenylcarbinol, which can then be dehydrated to AMS and hydrogenated to cumene, or it can be burned as fuel. The choice depends on the price of hydrogen and the capacity of the AMS hydrogenation unit. The cut-point control on the cumene recovery column is therefore a selectivity lever, because a looser cut point returns more acetophenone to the oxidizer and a tighter cut point increases the CHP concentration in the bottoms and the thermal decomposition risk. The column is monitored by gas chromatography of the overhead and bottoms streams, and the acetophenone concentration in the overhead is kept below 0.5 wt% while the CHP concentration in the bottoms is kept below 85 wt%. The reflux ratio is adjusted automatically to maintain the overhead composition, and the reboiler steam flow is controlled by the bottoms composition and the column pressure. The vacuum system is designed to handle the non-condensables produced by trace CHP decomposition, and the vent is sent to a thermal oxidiser or a scrubber. The operational boundary for this separation is narrow: a small increase in reboiler temperature can initiate CHP decomposition in the column, while a small decrease in reflux ratio can allow acetophenone to escape overhead and accumulate in the recycle cumene. The column is therefore operated with a conservative margin from the CHP thermal decomposition threshold, and the instrumentation is validated under the site mechanical integrity program. The final result is that the distillation cut-point control directly couples the selectivity of the oxidation section to the hydrogen consumption of the AMS hydrogenation unit and the tar production of the phenol recovery section.