The production of a 99.5%-pure organic peroxide intermediate by acylation of an acid chloride with hydrogen peroxide under alkaline conditions leaves a crude slurry containing unreacted acid chloride, free carboxylic acid, sodium chloride, residual water, and unstable peroxide-active oxygen. Washing is therefore not a single aqueous dilution step but a multistage liquid-liquid extraction sequence in which each contact stage reduces ionic impurities while simultaneously increasing the risk of base-catalyzed peroxide decomposition. In a representative 6000 L glass-lined batch vessel fitted with a retreat-curve impeller and operated at 40 rpm to 60 rpm, the first water wash at 15°C to 20°C removes sodium chloride and water-soluble sodium benzoate by dilution, but phase disengagement can exceed 45 min when the organic phase density approaches 1.05 g/cm³. The target of 99.5% assay therefore depends on a controlled balance among temperature, agitation, pH, and ionic strength, and any deviation from the narrow operating window can leave residual chloride above 5000 mg/kg or free acid above 2.0 wt% while destroying active oxygen through localized contact with excess alkali.

What Aqueous Phase pH Control Reveals About Intermediate Stability During Washing

The alkaline wash stage is intentionally operated between pH 8.0 and 9.5 using 5 wt% sodium carbonate or 2 wt% sodium hydroxide solution, because a pH below 7.0 slows conversion of residual acid chloride and allows liberated benzoic acid to partition back into the organic phase. Conversely, a pH above 10.0 initiates measurable base-catalyzed cleavage of the peroxide linkage, with active oxygen loss exceeding 0.15 wt% per hour in some batch formulations. The addition of 10% sodium carbonate is metered at approximately 2 L/min into the vessel vortex over not more than 20 min, and the endpoint is verified by continuous conductivity measurement in the aqueous discharge, with a target below 200 µS/cm after the final water wash. The aqueous cut is sampled for chloride by ion chromatography according to ASTM D512, while the organic phase is assayed by iodometric titration according to ASTM E298 to confirm that active oxygen content has not decreased below the expected stoichiometric baseline. In this pH window, the wash vessel jacket is held at 10°C to 20°C with chilled water, and the agitator speed is reduced to 30 rpm to 40 rpm because interfacial shear generated by higher tip speeds produces stable oil-in-water emulsions that are not resolved within 2 h without an electrostatic coalescer.

Contrary to the assumption that vigorous agitation shortens washing time, brine polishing of a 99.5% organic peroxide intermediate is controlled primarily by aqueous phase ionic strength rather than impeller speed. After the alkaline wash and two demineralized water washes, the organic phase is contacted with a 15% to 20% sodium chloride brine at 10°C to 15°C. The elevated ionic strength suppresses mutual solubility, increases density contrast to approximately 0.08 g/cm³ to 0.12 g/cm³, and accelerates droplet coalescence by compressing the electrical double layer across the oil-water interface. In a 6000 L working volume vessel, agitation during brine polishing is maintained at 30 rpm to 40 rpm for 15 min, followed by undisturbed settling for 60 min to 90 min. Separation is considered complete when the interfacial rag layer occupies less than 2% of total batch height. Persistent hazy aqueous cuts indicate finely dispersed solids, commonly sodium benzoate needles, and require a second brine wash or a reduction in agitation speed by 10 rpm to 15 rpm. Published plant data for this specific configuration is limited, but the relationship between brine concentration and phase separation follows established liquid-liquid extraction behavior for electrolyte-containing systems.

When Jacket Temperature Exceeds the Self-Accelerating Decomposition Temperature, Washing Becomes a Thermal Hazard Analysis Problem

In a washed organic peroxide intermediate, the self-accelerating decomposition temperature (SADT) depends strongly on water content, solvent content, and accumulated impurity profile. The wet crude in the reactor typically has a higher SADT than the isolated dry solid because water acts as a heat sink and dilutes active oxygen, but this margin cannot be assumed once phase separation leaves pockets of organic-rich material at the interface. The wash vessel jacket is therefore held at 10°C to 20°C with chilled water or brine, and excursions above 35°C initiate automatic diversion to a quench line with addition of chilled water at a rate sufficient to reduce batch temperature by 5°C/min. Differential scanning calorimetry is performed according to ASTM E537 to establish the onset temperature and exothermic enthalpy for each intermediate, while the SADT for packaging is determined using the heat accumulation storage test in the UN Manual of Tests and Criteria, Part II. The principal thermal hazard in washing arises when phase separation delay prolongs contact between the peroxide-containing organic phase and alkaline aqueous media, causing exothermic neutralization and slow peroxide decomposition in the interfacial rag layer. Temperature sensors are located at the vessel bottom, at the interfacial zone, and in the vent line; the control logic stops agitation and applies full cooling when the interfacial temperature exceeds 25°C. The emergency vent is sized for a two-phase flow scenario using DIERS methodology, and the relief set pressure is typically 50% above normal operating pressure but below the vessel design pressure. Nitrogen inerting is maintained at 0.2 bar to 0.5 bar gauge, and oxygen concentration in the headspace is controlled below 8 vol% to avoid flammable vapor mixtures when solvent residues are present.

Final moisture removal for a 99.5% pure organic peroxide intermediate is carried out in an agitated filter-dryer or vacuum dryer at jacket temperatures not exceeding 30°C. The residual water specification is typically below 0.20 wt% by ASTM E203 Karl Fischer titration, because higher moisture accelerates decomposition and interferes with subsequent phlegmatization with dicyclohexyl phthalate or mineral oil. Drying is conducted under vacuum of 10 mbar to 50 mbar absolute with slow rotation at 5 rpm to 10 rpm to minimize mechanical shear. The drying endpoint is reached when the pressure rise test shows less than 1 mbar increase over 30 min and the Karl Fischer result is stable between two samples taken 1 h apart. If residual moisture remains above 0.3 wt%, further drying is required; however, temperatures above 40°C are prohibited because the dry product SADT is significantly lower than the wet cake SADT. The dryer is blanketed with nitrogen, and the product is discharged under a relative humidity below 50% to prevent moisture re-uptake.

Analytical and compliance limits for a 99.5% organic peroxide intermediate after washing and drying

Parameter	Method	Specification limit	Stage
Assay	ASTM E298 iodometric titration	99.0% minimum as-is; 99.5% dry basis	after drying
Water	ASTM E203 Karl Fischer titration	0.20 wt% maximum	after drying
Free acid	acid-base titration	0.10 wt% maximum as benzoic acid	after final brine wash
Aqueous chloride	ASTM D512 ion chromatography	50 mg/kg maximum	after final water wash
pH of aqueous extract	electrometric measurement	6.5 to 8.0	after final water wash
SADT	UN Manual of Tests and Criteria, Part II	must be >20°C above processing temperature	dried product

Residual Solvent and Impurity Mass Balance Across the Final Drying Step

Residual solvent in a 99.5% pure organic peroxide intermediate must be controlled not only for purity but also because solvent entrainment lowers the SADT and changes the flash point of the final phlegmatized product. After the final brine wash, the organic phase may retain 0.5 wt% to 2.0 wt% of the reaction solvent, which is subsequently reduced in the vacuum dryer. Headspace gas chromatography with flame ionization detection is used to quantify residual solvent, and the specification for a phlegmatized end-use peroxide is typically below 0.1 wt% for toluene or methylene chloride carriers. Solvent removal is enhanced by operating the dryer condenser at -20°C to -30°C and by maintaining the vacuum pump discharge through an activated carbon bed to prevent organic vapor release. The mass balance across the drying step is verified by comparing the condensate volume, the pressure recovery profile, and the final residual solvent analysis. When drying is complete, the batch is transferred under nitrogen to closed containers fitted with conductive liners, and the packaging configuration is selected to comply with the SADT and quantity limits specified in the organic peroxide classification under the UN Manual of Tests and Criteria, Part II and the CLP Regulation (EC) No 1272/2008.

The first aqueous cut discharged from a 99.5% pure organic peroxide intermediate wash contains sodium chloride, sodium benzoate, residual hydrogen peroxide, and dissolved organic carbon. Before release to an industrial treatment plant, peroxide residues are quenched with sodium metabisulfite at a stoichiometric ratio of 1.2 mol sulfite per mol active oxygen, and the pH is adjusted to 6.0 to 9.0 using hydrochloric acid or sodium carbonate. The chemical oxygen demand after peroxide destruction and benzoate recovery may be reduced to below 500 mg/L depending on site permit limits, while total dissolved solids can exceed 20 000 mg/L because of sodium chloride. Discharge permits under local effluent standards require flow-proportional composite sampling and continuous conductivity monitoring. The relevant regulatory framework includes REACH registration dossiers for the specific peroxide and the CLP Regulation (EC) No 1272/2008 classification for self-reactive substances and organic peroxides.

Materials of construction for all wash vessels, transfer lines, and vacuum dryers must avoid copper, brass, and iron because transition metals catalyze homolytic cleavage of the peroxide bond. Glass-lined carbon steel is preferred for the reactor and wash vessel, while 316L stainless steel is used for condensers and piping only when the surface passivation layer is periodically verified by electrochemical means. Fluoropolymer gaskets and seats are selected for pumps and valves because peroxide contact with copper-containing alloys is prohibited. Rotary lobe pumps with 0.5 mm to 1.0 mm clearance are preferred over centrifugal pumps to reduce shear. Incompatibility with amine-based additives must be strictly controlled: even low levels of tertiary amines can initiate room-temperature radical generation and reduce active oxygen content within hours. The operational boundary for the washing sequence is therefore defined by a maximum processing temperature of 25°C, a pH window of 7.0 to 9.5, and a moisture specification of 0.20 wt% after drying, each verified by in-process sampling and formal batch release through ASTM E298, ASTM E203, and continuous conductivity and temperature interlocks.