In semi-batch n-butanol ethoxylation performed in a 10 m³ stainless steel agitated autoclave equipped with a helical ribbon impeller and an external shell-and-tube cooler, the absolute reactor pressure during ethylene oxide addition is typically maintained between 0.20 MPa and 0.45 MPa. The alkaline initiator is potassium hydroxide at 0.2–0.5 wt% based on final product mass, dried by vacuum stripping at 85–95 °C and 0.008–0.012 MPa to remove water. If water is not removed below 0.05 wt% as determined by ISO 760:1978, the water reacts with ethylene oxide to form polyethylene glycol, shifting the apparent hydroxyl number and broadening the oligomer envelope. Once ethylene oxide addition begins, the pressure controller on the vapour space adjusts the EO feed and nitrogen back-pressure; at 150 °C the liquid-phase EO mole fraction is approximately proportional to the absolute pressure above the mixture. Pressure excursions above 0.50 MPa accelerate propagation but raise the vapour-phase EO concentration, promote foam formation, and increase droplet carryover of potassium butoxide into the overhead vapour line. The resulting C4H9O(CH2CH2O)nH homolog distribution is therefore a function of pressure-control stability, gas-liquid interfacial area, and the effectiveness of the overhead demisting system. Because the base-catalyzed propagation rate constants of the lower homologues are similar, the unmodified product distribution is broad; unnecessary pressure cycling or catalyst return from the overhead condenser can broaden it further and increase the concentration of free n-butanol.

What Reactor Pressure Range Limits Ethylene Oxide Oligomer Propagation?

Ethylene oxide solubility in n-butanol and ethoxylated n-butanol mixtures follows Henry’s law over the moderate pressure range used in industrial alkoxylation. The pseudo-first-order propagation rate is proportional to the dissolved ethylene oxide concentration; therefore a pressure increase at constant temperature raises the overall addition rate. In a 5 m³ pilot autoclave with a flat-blade disc turbine at 1.5–2.5 m/s tip speed, published data for n-butanol under precisely controlled pressure ramps are limited, but related C6–C10 alcohol ethoxylate campaigns indicate that raising the absolute pressure from 0.20 MPa to 0.40 MPa at 140 °C shortens the EO addition time by 30–50% without changing the target average EO number if the total EO charge is fixed. The rate increase is not linear above 0.45 MPa because the liquid-side mass transfer resistance becomes limiting once the bubble surface is saturated. Below 0.18 MPa the reaction becomes starved and the control loop tends to oscillate because the ethylene oxide vapor pressure is strongly temperature dependent. Above 0.50 MPa the relief system must be sized for a higher reaction-energy density and the reactor must comply with pressure vessel codes such as ASME BPVC Section VIII Division 1. The production-scale operating envelope is therefore best controlled at 0.20–0.45 MPa, with the lower bound set by reaction rate and control stability and the upper bound set by vapour-liquid disengagement, relief load, and the onset of excessive overhead carryover.

Independent of pressure, catalyst carryover into the stripping and distillation sequence creates a second reaction zone that is often underestimated. In a 15 m² falling film evaporator operating at 0.008–0.015 MPa and 85–110 °C, potassium hydroxide or potassium butoxide entrained in the crude product remains catalytically active if neutralization with acetic acid or lactic acid is incomplete. The residual alkali promotes ethoxylation of lower-molecular-weight homologues during removal of unreacted n-butanol and light adducts, shifting the distribution toward higher ethylene oxide numbers and increasing the apparent cloud point. This carryover route is more severe when the primary reactor is operated at the high end of the pressure window because the liquid entrainment load on the overhead demister increases. In production campaigns, a controlled pressure reduction from 0.40 MPa to 0.25 MPa before transfer to the stripper has been used to reduce foam height; however, this step must be executed slowly to avoid reverse flow of catalyst-rich liquid from the overhead condenser. Published data for n-butanol ethoxylation under this specific configuration are limited; the mechanistic relationship is inferred from C12–C14 alcohol ethoxylate plants where aerial condenser fouling was reduced by 40–60% after installing a two-stage vane demister.

Catalyst Carryover Through Falling Film Stripping Equipment

Falling film stripping of crude n-butanol ethoxylate is preferred because it limits residence time relative to pot distillation and reduces color body formation. The falling film reboiler must be designed so that the film Reynolds number remains above 500 at the highest operating viscosity; otherwise potassium acetate formed during neutralization deposits on the heated wall and heat-transfer coefficients can fall from 800 W/m²·K to below 400 W/m²·K within 72 h of continuous operation. The entrained alkali acts as a homogeneous catalyst for ethylene oxide redistribution, transesterification, and aldol condensation involving trace aldehydes. Because potassium acetate and potassium hydroxide are not volatile at these temperatures, vapour-phase transport is negligible; liquid droplet carryover is the main mechanism. A 1.5 m diameter knock-out drum with a wire-mesh demister of 150 mm thickness and 0.10 kg/m³ bulk density typically reduces potassium in the dried product to 5–15 mg/kg, but the residual level depends strongly on foam height and superficial vapor velocity. At vapor velocities above 0.8 m/s across the demister, re-entrainment occurs and the carryover increases by a factor of 2–4. The recovered n-butanol overhead should be tested periodically by ISO 11885:2007; potassium above 5 mg/kg indicates inadequate demisting and requires either reflux washing or a second coalescing stage.

When Short-Path Evaporation Is Applied to n-Butanol Ethoxylate Neutralization

If short-path evaporation is used instead of falling film stripping to remove free n-butanol and light adducts after acid neutralization with lactic acid, the residence time is typically limited to 1–3 min at 120–140 °C and 0.001–0.005 MPa. This narrow thermal window is critical because residual potassium lactate or potassium acetate retains sufficient base strength to catalyze ethoxylate chain redistribution and color-forming dehydration at temperatures above 150 °C. The wiped film is mechanically spread to a thickness of 0.2–0.5 mm on a 0.8 m² heated cylinder using a 1200 rpm rotor with adjustable clearance. Under these conditions, the cloud point of a 1 wt% aqueous solution of the finished n-butanol ethoxylate can shift upward by 2–5 °C if catalyst carryover is not neutralized to an acid number below 0.1 mg KOH/g. Color measured by ASTM D1209-05 remains below 50 APHA when the stripper bottom temperature does not exceed 140 °C; at 155 °C the same product can reach 80–150 APHA within 4 h if potassium exceeds 10 mg/kg. The adduct distribution is also altered by removal of the most volatile homologues; however, if the short-path evaporator is operated below 0.002 MPa while the feed contains residual ethylene oxide, back-reaction and inter-homologue equilibration occur on the hot wall. The upstream stripping step must therefore reduce residual EO to <1 mg/kg before short-path evaporation.

Measuring Vapour Entrainment, Demister Efficiency, and Alkali Aerosol Transport

Vapour-phase transport of potassium is negligible at reactor temperatures below 180 °C because potassium hydroxide and potassium butoxide have vapor pressures far below the process pressure. The observed carryover is droplet-borne and depends on the bubble size distribution at the gas–liquid interface and the froth height in the disengagement zone. In a 2 m internal diameter reactor with a disengagement height of 1.2 m, superficial vapor velocity should be kept below 0.04 m/s at the liquid surface to prevent jet droplet ejection from bursting bubbles. A vane-type demister with 50 mm blade spacing and 150 mm thickness reduces aerosol droplets larger than 10 µm by 90–95%; droplets below 5 µm are not effectively removed and enter the overhead condenser. Foam suppression with a silicone-based antifoam at 10–20 mg/kg is used in some plants, but the antifoam can alter the surface tension of the final product as measured by ISO 304:1985. An external reflux of stripped n-butanol at 0.2 kg/kg overhead flow is sometimes used to wash the demister and return catalyst droplets to the reactor, but this also returns low-molecular-weight acetals and can broaden the distribution if the reflux is not dehydrated. A more robust strategy is to maintain a nitrogen pad of 0.02–0.05 MPa above the EO vapor pressure during the final 10% of EO addition, lowering the volumetric gas evolution rate and hence the foam height. This strategy has been observed in agitated alkoxylation reactors to lower potassium carryover from 12 mg/kg to 3 mg/kg in the overhead condensate without altering the average EO number.

Residual Alkali Systems Degrade Through Aldol Condensation and Ring Redistribution

Thermal degradation of n-butanol ethoxylates in the presence of residual alkali is a property cliff-edge phenomenon. Below 130 °C the acid number increase is less than 0.05 mg KOH/g per hour; between 140 °C and 150 °C the rate doubles for each 10 °C increase. At 160 °C, acetaldehyde and crotonaldehyde evolution from ethylene oxide decomposition becomes measurable by headspace GC and imparts a pungent odor. The chain-length distribution is not stable at these temperatures because basic alkoxide sites catalyze reverse ring-opening and redistribution of ethylene oxide units among homologues. A product with an initial average EO number of 4.0 and a polydispersity index of 1.18 can shift to average 4.7 and polydispersity 1.35 after 6 h at 150 °C in the presence of 25 mg/kg residual potassium. This redistribution occurs even without free ethylene oxide because the ethylene oxide unit is transferred from higher homologues to lower ones. The change in cloud point is not linear with average EO number; at average EO 4.5, a shift of 0.5 units can change the cloud point by 3–8 °C, which may still comply with ISO 4320:1983 but fall outside a customer specification. Therefore catalyst carryover must be neutralized with an acid equivalent such as acetic acid at a molar ratio of 1.0–1.1 relative to total potassium, and the resulting potassium acetate must be removed or dispersed through a 5 µm bag filter if the final product is used in metalworking fluids or polyurethane slabstock formulations.

Neutralization stoichiometry after the stripping operation is verified by acid number rather than pH because the non-aqueous medium has a poorly defined proton activity. A sample is withdrawn from the neutralizer loop every 2 h, dissolved in a mixture of 2-propanol and toluene in a 1:1 volume ratio, and titrated potentiometrically against 0.1 M methanolic potassium hydroxide according to ASTM D7253-16. The target acid number after neutralization is <0.1 mg KOH/g. If the acid number rises above 0.2 mg KOH/g, the batch is re-neutralized with a dilute acid solution and re-filtered through a 1 µm polypropylene cartridge filter. The hydroxyl number is measured by ASTM D4274-21 to confirm that the average EO number has not shifted during the neutralization and filtration steps. For products intended for cosmetic or pharmaceutical intermediates, residual ethylene oxide is determined by headspace gas chromatography according to ISO 17257:2013; the acceptance limit is typically <1 mg/kg in accordance with EU 1223/2009 provisions for ethoxylated ingredients. When the recovered n-butanol contains potassium above 3 mg/kg, it is either redistilled through a 1.2 m packed column with structured packing or diverted to a non-catalytic use to prevent accumulation of catalyst in the next ethoxylation batch. The adduct distribution of the final product is monitored by gas chromatography with a DB-5 capillary column and is expressed as the weight percentage of each homologue from C4EO1 to C4EO12; a narrowing of the distribution is not achieved by simply raising pressure but by controlling the pressure at a constant value, maintaining the stirrer at full gassing speed, and preventing delayed addition of unreacted ethylene oxide residues.