In the classical chlorohydrin route to propylene oxide, propylene gas is absorbed into an aqueous chlorine solution maintained at 30–50 °C within a bubble column or packed tower, generating an equilibrium mixture of 1‑chloro‑2‑propanol and 2‑chloro‑1‑propanol—collectively referred to as propylene chlorohydrin (PCH)—along with dilute hydrochloric acid. The PCH-containing brine, typically holding 4–6 wt% chlorohydrin, is subsequently directed to a saponification reactor where it is contacted with a calcium hydroxide slurry or, in less common configurations, sodium hydroxide. The dehydrochlorination step that yields propylene oxide (PO) proceeds through an intramolecular nucleophilic substitution (Williamson‑type epoxide formation) whose rate is first‑order in both chlorohydrin and hydroxide ion concentration; the reaction therefore exhibits a second‑order rate constant of approximately 0.02–0.05 L·mol⁻¹·s⁻¹ at 80 °C, depending on the isomer distribution and ionic strength of the brine. Because the epoxide ring is susceptible to hydrolysis to propylene glycol and subsequent oligomerization, the hydroxide concentration—translated into a process pH setpoint measured at the reactor exit—constitutes the single most consequential control variable governing overall process selectivity. Commercial experience across plants operating continuous saponification with lime indicates that an exit pH window of 10.2–10.8 (measured at 60 °C with a glass electrode conforming to ASTM E70‑19) provides the optimal balance between dehydrochlorination rate and glycol suppression. Operation below pH 9.5 dramatically increases the half‑life of the chlorohydrin in the alkaline medium, pushing the required reactor residence time beyond the practical limit of 10–25 minutes typical for a CSTR cascade, while also shifting the product distribution toward monopropylene glycol (MPG) and dipropylene glycol (DPG). Conversely, a sustained pH above 11.2 results in a detectable acceleration of propylene oxide hydration to MPG, a reaction whose rate constant is roughly 0.001 s⁻¹ at 25 °C and pH 11, and promotes calcium carbonate precipitation when carbon dioxide is inadvertently introduced with the lime slurry, leading to fouling of pH electrodes, heat‑exchanger surfaces, and downstream brine‑clarification equipment. The calcium content of the lime is typically specified to meet ASTM C110‑20 for available CaO, with a requirement for less than 2 wt% MgO and a controlled silica level to avoid the formation of gelatinous precipitates that interfere with brine filtration in the subsequent calcium chloride recovery unit.
The interdependence between pH and the product spectrum has been quantified in numerous pilot‑scale campaigns, and the resulting data—although often proprietary—are reflected in the operating envelopes published in encyclopedic references such as Ullmann’s Encyclopedia of Industrial Chemistry. At a typical saponification temperature of 85 °C and with a PCH feed containing 5 wt% total chlorohydrins, raising the pH from 9.0 to 10.5 increases the instantaneous selectivity to PO from approximately 75% to 90%, while the MPG fraction drops from 18% to 7%, and DPG formation remains below 2%. The relationship is not linear: between pH 10.5 and 11.0, PO selectivity climbs only marginally to 91%, but the MPG selectivity shows a modest uptick to 8% as the PO hydration rate begins to compete. What is not immediately apparent from steady‑state selectivity data is the influence of pH on the physical characteristics of the spent brine. When lime is the saponification agent, the byproduct calcium chloride solution must be separated, purified, and often concentrated for sale or disposal. A pH above 11.2 coincides with an increase in the concentration of dissolved calcium hydroxide, which upon cooling precipitates as a fine scale in the brine evaporator preheaters, raising the cleaning frequency from approximately once every 6–8 weeks to once every 10–14 days in forced‑circulation evaporators. Moreover, the presence of residual hydroxide shifts the speciation of trace heavy metals—particularly iron leached from carbon‑steel piping upstream of the saponifier—toward colloidal oxyhydroxides that are not efficiently removed by the lamella clarifier, resulting in a brine product that fails the commercial specification of <5 mg/L suspended solids for use in de‑icing or dust‑control applications that reference ASTM D98‑15. Thus, the pH target is not solely a yield‑driven parameter; it is constrained on the upper end by brine evaporator operability and on the lower end by a catastrophic loss of PO yield. This narrow processing window—effectively ±0.3 pH units around a 10.5 setpoint—places stringent demands on the lime slurry dosing system, which must compensate for variable alkalinity demand caused by fluctuations in the HCl content of the chlorohydrin feed, itself a function of the upstream chlorine‑water ratio and propylene conversion.
The stability of propylene chlorohydrin in aqueous hydrochloric acid solution defines the maximum permissible hold‑up between the chlorohydrin synthesis tower and the saponification unit. PCH undergoes slow acid‑catalysed hydrolysis to propylene glycol even at 25 °C; the reaction follows pseudo‑first‑order kinetics with a rate constant of approximately 1.5 × 10⁻⁶ s⁻¹ at pH 3 and 25 °C, corresponding to a half‑life of about 5.3 days. However, the solution as it exits the chlorohydrin column typically contains 1–2 wt% free HCl, yielding a pH of 1–2. At pH 1.5 and 30 °C, the hydrolysis half‑life drops to less than 24 hours. This is seldom an issue when the intermediate buffer vessel is sized for a mean residence time of 30–60 minutes under normal throughput. The more acute concern is the autocatalytic decomposition pathway that is triggered when PCH‑containing brine is heated inadvertently or subjected to stagnant zones in piping. Laboratory accelerating‑rate calorimetry (ARC) studies on representative chlorohydrin brines indicate an exothermic onset at temperatures as low as 45–55 °C when 50–100 ppm of dissolved iron is present, a common contaminant originating from corroded carbon‑steel storage tanks. The decomposition cascade involves the proton‑catalysed condensation of PCH to di(propylene glycol) and polyglycol ethers, releasing additional HCl and thus lowering the pH further, which in turn accelerates the condensation. The adiabatic temperature rise for this exotherm has been measured in the range of 300–500 °C, with a maximum self‑heat rate exceeding 10 °C·min⁻¹ once the mixture reaches 100 °C. Consequently, intermediate storage vessels in plants designed after the mid‑1990s are invariably constructed of rubber‑lined carbon steel or solid 316L stainless steel, are fitted with cooling coils capable of maintaining a bulk temperature of ≤20 °C, and are vented through a relief system sized according to a two‑phase discharge model following the Design Institute for Emergency Relief Systems (DIERS) methodology, assuming a worst‑case decomposition energy of 650 J·g⁻¹. Where production logistics demand longer hold times—for instance, during a downstream saponifier maintenance window—the PCH brine is often partially neutralized to pH 3.5–4.5 with a sodium bicarbonate buffer, which increases the ambient‑temperature stability to approximately 72 hours without introducing an alkali that would initiate premature dehydrochlorination. Published data for the long‑term storage of buffered chlorohydrin brines at sub‑10 °C temperatures in lined tanks indicates a hydrolytic loss of <0.1% per day, but the practice is limited by the cost of refrigeration and the availability of lined storage capacity.
The saponification reaction is heterogeneous: solid calcium hydroxide particles must dissolve into the aqueous phase before the hydroxide ions can react with dissolved chlorohydrin. The dissolution rate is a function of the specific surface area of the lime, the degree of calcination of the parent limestone, and the local pH gradient within the reaction medium. High‑calcium hydrated lime meeting the chemical index of ASTM C141/C141M‑14 typically exhibits a BET surface area of 15–25 m²·g⁻¹ and a median particle diameter (d₅₀) of 3–6 µm when produced in a modern hydrator with controlled slaking conditions. However, if the lime is transported pneumatically and stored in silos before slaking, aging and carbonation can reduce the surface area to below 10 m²·g⁻¹, with a concomitant increase in d₅₀ to 15–30 µm. In the saponification reactor, a shift from 5 µm to 20 µm mean particle size increases the time required for 90% dissolution from approximately 2 minutes to 8–12 minutes at 80 °C and pH 10.5. Because the overall reactor residence time is often fixed by the vessel volume and production rate at around 15–25 minutes, the presence of coarse or poorly slaked lime particles results in a trailing hydroxide release that can occur after the reaction mixture leaves the main reactor and enters the brine flash tank or the PO stripping column. This post‑reactor alkalinity causes a slow, uncontrolled increase in pH in downstream equipment, promoting PO hydration to MPG in the overhead condenser and fouling of the stripping trays with calcium salts. To mitigate this, plants that source quicklime of variable quality commonly install an in‑line wet milling step (a rotor‑stator mill or a media mill) after the slaker, targeting a slurry fineness of <1% residue on a 325‑mesh (44 µm) sieve. The slurry is then homogenized in an agitated day tank fitted with a density meter (e.g., Coriolis type calibrated to measure 1.08–1.14 g·cm⁻³, corresponding to 10–15 wt% solids) whose output signal cascades to the saponification pH controller. The dosing valve—typically an eccentric rotary plug valve or a pinch valve with a ceramic trim to withstand abrasion—opens against a recirculation loop to ensure a representative and pulse‑free slurry feed. Despite these measures, seasonal variations in lime reactivity (linked to kiln operating temperature and raw limestone crystal structure) produce a measurable drift in the pH control loop gain, necessitating periodic retuning of the PID parameters; plants processing a single grade of limestone report a retuning frequency of 2–3 times per year, while those drawing from multiple quarries may require adjustments as often as every 4–6 weeks.
A frequently underestimated process conflict emerges from the interaction between the chlorohydrin intermediate and metallic contaminants that enter the stream either through corrosion of upstream equipment or via the chlorine source. While the chlorohydrin column itself is often constructed of titanium‑lined or rubber‑lined steel to resist wet chlorine, the downstream piping, pumps, and buffer vessels may include 304L or even carbon‑steel components if the process was originally designed for a less corrosive environment. Under the low‑pH conditions (pH 1–2) prevailing before neutralization, iron dissolves as ferrous ions at a rate that can reach 0.1–0.5 mm·year⁻¹ depending on oxygen ingress and temperature. Ferrous ion concentrations of 10–50 mg·L⁻¹ are sufficient to reduce the onset temperature of PCH decomposition by 10–15 °C, bringing the thermal run‑away boundary perilously close to the normal storage temperature of 25–30 °C. The mechanism involves a redox cycle in which ferrous ions are oxidized by dissolved oxygen or by chlorohydrin decomposition intermediates, generating ferric species that in turn catalyze the formation of free‑radical chains leading to epoxide ring‑opening and polymerization. This degradation not only destroys product precursor but yields a complex mixture of propionaldehyde, acetone, and polyether alcohols that act as surfactants, stabilizing emulsions in the subsequent saponification and PO purification stages. Removal of these metallic impurities is accomplished by passing the chlorohydrin brine through a bed of chelating ion‑exchange resin in the sodium form (e.g., iminodiacetic acid‑functionalized styrene‑divinylbenzene) operated at a specific flow rate of 5–15 BV·h⁻¹ and regenerated with 5–10% HCl. However, resin life is shortened by fouling from organic condensation products, and the elution generates a low‑pH, iron‑rich waste stream that must be neutralized before discharge to conform to local effluent limits such as those derived from EU Directive 2010/75/EU on industrial emissions. Where integrated environmental controls are absent, the plant resorts to a combination of polypropylene‑lined piping and Hastelloy C‑276 trim on control valves to minimize iron pickup, thereby accepting the higher capital expenditure in exchange for demonstrable intrinsic safety and longer inter‑maintenance intervals.
The interplay between saponification pH and chlorohydrin stability is not confined to the primary reaction network. The gas‑phase propylene oxide produced in the saponifier is immediately steam‑stripped under a vacuum of 20–40 kPa absolute to prevent its accumulation in the liquid phase, where it would be subject to hydration, isomerization to allyl alcohol, and condensation with residual PCH. The overhead PO‑steam mixture is then partially condensed, and the crude PO is directed to a series of distillation columns. If the saponifier pH has been allowed to drift upward beyond 11.0, the overhead vapour may carry entrained droplets of alkaline brine, leading to alcoholate‑catalysed oligomerization of PO in the vapour space of the first distillation column and a subsequent increase in the differential pressure across the column from a baseline of 15–25 kPa to 45–60 kPa, a tell‑tale sign of tray fouling. This operational feedback loops back to the chlorohydrin intermediate: when the distillation section is forced to reduce throughput to manage fouling, the intermediate buffer vessel filling level rises, extending the PCH hold‑time and thereby increasing the probability of acid‑catalysed decomposition. Thus, the pH target acts not only as a yield knob but as a coupling parameter linking the stability of the intermediate to the operability of the entire downstream separation train. The limiting standard for crude PO purity entering the distillation sequence—often an internal specification of ≥98.5 wt% PO, <0.1 wt% water, and <0.5 wt% aldehydes as propanal—can be met only when the saponification pH is held within the 10.2–10.8 band and the PCH feed to the saponifier exhibits a free‑HCl content of <0.2 wt% and an iron concentration below 5 mg·L⁻¹. Any excursion from these boundaries is reflected in the final product colour (as measured by ASTM D1209‑05) and in the acid acceptance test (ASTM D2106‑07), potentially rendering the PO unsuitable for downstream polyether polyol applications that require high‑purity monomer to avoid catalyst deactivation in the oxyalkylation reactor.
Here is a summary of the influence of saponification pH on product distribution, based on aggregated pilot‑plant data for a lime‑based process operating at 85 °C with a PCH feed concentration of 5 wt% and a residence time of 18 min:| Saponification pH at 60 °C | PO Selectivity (mol%) | MPG Selectivity (mol%) | DPG + Heavier (mol%) | Relevant Standard for MPG in Brine |
|---|---|---|---|---|
| 9.0 | 76 | 18 | 6 | ASTM D3459‑13 |
| 9.8 | 84 | 11 | 5 | ASTM D3459‑13 |
| 10.3 | 89 | 8 | 3 | ASTM D3459‑13 |
| 10.8 | 90 | 8 | 2 | ASTM D3459‑13 |
| 11.4 | 87 | 10 | 3 | ASTM D3459‑13 |
Operational boundaries that have emerged from more than 30 years of continuous commercial practice are consolidated in the following engineering specification matrix, which links each critical variable to the phenomenological limit and the associated standard or equipment type.
| Parameter | Acceptable Range / Setpoint | Critical Constraint | Referenced Standard or Equipment |
|---|---|---|---|
| Saponification pH (exit, 60 °C) | 10.2–10.8 | Below 9.5: PO yield drops >10%; above 11.2: fouling accelerates | ASTM E70‑19; glass electrode, double‑junction reference, automatic temperature compensation |
| Saponification temperature | 80–90 °C | Below 78 °C: PCH conversion incomplete; above 95 °C: PO hydration surges | Shell‑and‑tube steam reboiler with forced circulation; RTD probe per IEC 60751 Class A |
| PCH feed free‑HCl | <0.2 wt% | Excess HCl consumes lime, shifts pH downward, and promotes corrosion | Potentiometric titration per ASTM D664‑18e1 |
| Lime slurry solids content | 10–15 wt% | Below 8 wt%: excessive water load to saponifier; above 18 wt%: pipe plugging | Coriolis density meter; lime quality per ASTM C141/C141M‑14 |
| Lime particle size d₅₀ | <15 µm | Coarser than 25 µm: trailing reactivity and downstream scaling | Laser diffraction per ISO 13320:2020 |
| Chlorohydrin storage temperature | ≤20 °C | Sustained above 45 °C with iron: autocatalytic decomposition risk | Rubber‑lined or 316L vessel; cooling coil with −5 °C brine/glycol |
| Chlorohydrin buffer pH (when needed) | 3.5–4.5 | Above 5.5: premature dehydrochlorination; below 2.5: rapid hydrolysis | Sodium bicarbonate dosing; pH transmitter per ASTM E70 with flat‑glass probe for fouling service |
| Crude PO water content | <0.1 wt% | Higher water promotes glycol formation during distillation | Karl Fischer titration per ASTM E203‑16 |
Committed to quality, innovation, and reliable chemical solutions for a sustainable future.
