In continuous fixed-bed operation for the production of glycol ethers from monoethylene glycol (MEG), the kinetic competition between O-alkylation and alcohol dehydration determines reactor yield, downstream separation load, and catalyst cycle length. The primary methanol-based manifold consists of two consecutive etherification steps: MEG reacts with methanol to form 2-methoxyethanol (EGME) and water, and EGME can undergo a second methylation to 1,2-dimethoxyethane (EGDME). With n-butanol, the corresponding primary ether is 2-butoxyethanol (EGBE), while 1,2-dibutoxyethane appears only at butanol-to-MEG molar ratios above 6:1 and elevated residence time. Competing acid-catalysed dehydration of methanol produces dimethyl ether (DME), and n-butanol dehydration generates 1-butene together with cis/trans-2-butene. MEG dehydration over the same Brønsted-acid sites produces 1,4-dioxane and diethylene glycol as trace by-products that are disproportionately regulated in solvent applications. The etherification reactions are reversible and mildly exothermic; vapour-phase fixed-bed operation between 120°C and 180°C shifts equilibrium toward the ether products when excess alcohol is maintained and water is removed by distillation or adsorption. Published calorimetric data for the methanol-MEG system indicate reaction exothermicity in the range of −20 kJ mol−1 to −35 kJ mol−1 of EGME formed; the butanol-MEG system exhibits a similar heat of reaction but is more sensitive to pore constraints and alcohol dehydration side reactions. The choice of zeolitic catalyst topology, framework Si/Al ratio, binder chemistry, and feed water control therefore establishes the practical operating window for each etherification route.

Methanol Etherification Selectivity over H-ZSM-5 in Fixed-Bed Operation

A fixed-bed screening protocol for methanol etherification over H-ZSM-5 extrudates establishes the reference selectivity and deactivation behaviour for the lower-alcohol system. The catalyst is prepared as 1.6 mm trilobate extrudates containing 20 wt% gamma-alumina binder, with a framework Si/Al ratio of 30. Before loading, the extrudates are calcined in flowing dry air at 450°C for 6 h with a ramp rate of 2°C min−1. The fixed-bed reactor is a 25 mm internal diameter 316L stainless steel tube with a catalyst bed length of 500 mm and a catalyst volume of 200 mL. The bed is diluted 1:1 by volume with 0.5 mm silicon carbide to improve heat transfer and prevent channelling. Liquid feed containing methanol and MEG in a 4:1 molar ratio is delivered by a piston pump through a preheater maintained at 130°C. Reactor pressure is held at 15 bar, and the liquid hourly space velocity is maintained at 1.5 h−1. Under isothermal control at 135–150°C, the bed temperature rise is limited to 8–12°C by an external circulating-oil jacket. A representative screening run at 145°C produces MEG conversion of 88%, EGME selectivity of 85%, EGDME selectivity of 10%, DME selectivity of 2%, and 1,4-dioxane content below 0.1 wt% in the crude product. Apparent activation energy for MEG conversion over H-ZSM-5 is reported in the range of 55–65 kJ mol−1 when methanol is present in greater than 3:1 molar excess. The reaction order with respect to MEG is near first order, while the order with respect to methanol approaches zero above a 3:1 molar ratio, consistent with saturated methanol coverage on Brønsted acid sites. Water competitively adsorbs on the same sites; an increase in feed water from 0.2 wt% to 1.5 wt% reduces MEG conversion by 12–18 percentage points at constant space velocity. The processing window is therefore narrow: pre-drying of MEG by vacuum stripping at 80°C and 50 mbar is specified when inlet water exceeds 0.5 wt%.

Residual water in the MEG feed stream influences both equilibrium conversion and catalyst activity and is the most frequent field failure mode in continuous etherification units. Sodium residues from upstream oxide-catalyst manufacture neutralise Brønsted acid sites even at low concentration. Feed sodium levels of 5 mg kg−1 cause measurable conversion loss within the first 200 h of operation; a guard bed containing 1.5 kg of sulfonic acid ion-exchange resin operated at 40°C reduces sodium to below 0.1 mg kg−1. Without the guard bed, MEG conversion declines by approximately 2% per mg kg−1 sodium accumulated on the catalyst. Amine-based corrosion inhibitors must not be introduced upstream of the guard bed or reactor because they titrate Brønsted sites and cause premature deactivation. Water content is determined by ASTM E203, acidity by ASTM D1613, and sodium by inductively coupled plasma optical emission spectrometry with a detection limit of 0.05 mg kg−1. These feed constraints are structural, not advisory: a MEG storage tank breathing humid air at relative humidity above 60% will gradually increase water content above the allowable threshold, and the resulting conversion loss cannot be recovered by raising reactor temperature without sacrificing selectivity to DME and 1,4-dioxane.

What Limits Butanol Etherification Selectivity at High Conversion?

Butanol etherification over zeolitic catalysts is governed by the interplay between pore diameter, acid site strength, and alcohol dehydration kinetics. H-ZSM-5 with its 10-membered-ring channels restricts diffusion of n-butanol and the product 2-butoxyethanol, leading to lower ether selectivity and higher intraparticle residence time that favours butene formation. H-Beta with a three-dimensional 12-membered-ring pore system and a framework Si/Al ratio of 25 is preferred for the butanol route. In a 25 mm fixed-bed reactor with H-Beta 1.6 mm extrudates containing 20 wt% silica binder, the butanol-based system is operated at 160–170°C, 18 bar, a butanol-to-MEG molar ratio of 5:1, and a liquid hourly space velocity of 1.0 h−1. Under these conditions, MEG conversion reaches 82%, EGBE selectivity is 76%, butene selectivity is 14%, dibutyl ether selectivity is 5%, and 1,4-dioxane content is approximately 0.3 wt%. The apparent activation energy for butanol etherification over H-Beta is approximately 75 kJ mol−1, whereas n-butanol dehydration to butenes exhibits an apparent activation energy of approximately 95 kJ mol−1. The higher activation energy for dehydration means that temperature excursions are disproportionately costly: selectivity to EGBE declines by approximately 8 percentage points per 10°C rise above 170°C. The practical processing window is therefore restricted to ±5°C around the target inlet temperature; operation above 175°C generates excessive butenes, while operation below 155°C reduces MEG conversion below 70% and increases recycle alcohol duty. Framework acid site density also controls selectivity: high aluminium content increases the number of strong Brønsted sites, accelerating both etherification and dehydration. H-Beta with Si/Al below 12 shows increased butene selectivity above 20%, while Si/Al above 40 reduces MEG conversion below 60% at the same space velocity. The acid site density measured by ammonia temperature-programmed desorption is 0.60 mmol NH3 g−1 for H-Beta compared with 0.45 mmol NH3 g−1 for H-ZSM-5; this difference partly explains the higher low-temperature activity of H-Beta but also its higher dehydration potential if temperature control is lost.

Downstream separation of the butanol-based product mixture is constrained by the heterogeneous azeotrope between water and n-butanol and by the high boiling point of 2-butoxyethanol (171°C) relative to n-butanol (117°C). A two-column distillation sequence with a decanter is used; the first column recovers n-butanol-water overhead at 95–105°C and atmospheric pressure, while the bottom stream containing EGBE and high-boiling ethers is fed to a second column operated at 50 mbar with a reboiler temperature not exceeding 160°C to avoid product decomposition. Reflux ratios in the first column are maintained between 2.5 and 4.0 to prevent water breakthrough; field data from a 10,000 t a−1 unit indicate that an increase in feed water above 1.5 wt% overloads the decanter and reduces n-butanol recovery to 85%. The second column is operated with a reflux ratio of 1.5–2.0 and a pressure drop below 0.3 bar to limit cycle oil fouling. Heavy by-products containing 1,2-dibutoxyethane and trace dioxane accumulate in the reboiler and are purged at 0.5–1.0 wt% of the crude feed to prevent viscosity build-up and product colour degradation.

When Zeolite Beta Is Substituted for H-ZSM-5 in Adiabatic Reactor Configurations

Substituting H-Beta for H-ZSM-5 in an adiabatic fixed-bed configuration alters the temperature profile, pressure drop, and catalyst inventory requirements. In a 1.0 m internal diameter adiabatic reactor containing three catalyst beds of 1.2 m length each, H-Beta extrudates of 3.0 mm diameter reduce pressure drop to 0.6 bar at a total feed rate of 500 kg h−1. Under these conditions the measured adiabatic temperature rise is 18–22°C per pass at MEG conversion of 80%. Inlet temperature is controlled at 145°C, and the outlet is not permitted to exceed 165°C. Above 170°C, butene selectivity rises from 14% to 22%, and the EGBE yield falls below 70% due to secondary dehydration of the ether product on strong Brønsted sites. To maintain the processing window within ±5°C, quench methanol is injected between beds at 25°C; the interstage heat exchanger uses circulating oil with a maximum duty of 350 kW. Catalyst bed pressure drop is monitored online with differential pressure transmitters calibrated to 0.5% full scale; an increase above 1.5 bar indicates coke or fines accumulation and triggers regeneration. Smaller extrudates of 1.6 mm diameter improve effectiveness factor but raise pressure drop above 1.8 bar, causing gas-liquid maldistribution and hot spots. The adiabatic configuration also exposes the catalyst to higher water partial pressure; hydrothermal deactivation of H-Beta is more rapid than H-ZSM-5 when water partial pressure exceeds 2 bar. Therefore butanol feed water is maintained below 0.3 wt%, and steam stripping of the catalyst is avoided during regeneration. Published data for this specific adiabatic configuration is limited to pilot-scale runs; the values reported here are derived from a 200 h demonstration campaign with online GC sampling. The main operational boundary is that adiabatic operation cannot compensate for feed water excursions because the water vapour generated in the first bed suppresses conversion in the second and third beds, shifting the temperature rise downward and masking catalyst deactivation until the outlet temperature drops below 155°C.

Regenerating Coked Zeolites Without Sacrificing Framework Aluminum

Coke deposition on medium-pore zeolites under methanol-rich conditions follows a sigmoidal deactivation curve with an initial induction period of 150–250 h, followed by a rapid activity loss once pore-mouth deposits restrict access to internal Brønsted sites. Regeneration of coked H-Beta and H-ZSM-5 extrudates is conducted in situ using a nitrogen-diluted air stream with oxygen concentration not exceeding 1.0 vol%; the initial ramp from 250°C to 480°C is limited to 1°C min−1 to avoid local temperature excursions above 550°C. Steam partial pressure is kept below 0.2 bar by pre-drying the regeneration gas to a dew point of −40°C. A hold of 12 h at 480°C restores 85–90% of fresh activity after three cycles; surface area determined by nitrogen physisorption remains above 350 m² g−1, and micropore volume remains above 0.12 cm³ g−1. The regeneration effluent is passed through a scrubber to remove acetaldehyde and dioxane traces before venting. Catalyst replacement is triggered when crush strength declines below 0.7 N mm−1 as measured by ASTM D7084, or when activity after regeneration is less than 60% of fresh activity. Excessive regeneration temperature above 550°C extracts framework aluminium, lowering the Brønsted acid site density and irreversibly reducing etherification activity; this dealumination is detectable by a shift in the ²⁷Al MAS NMR spectrum from tetrahedral to octahedral coordination. The regeneration interval is not fixed; it is triggered by either an 8% absolute loss in MEG conversion or a 15°C increase in bed mid-point temperature required to maintain target conversion. The use of regeneration gas with oxygen concentration above 2.0 vol% is incompatible with high coke loadings because local combustion fronts can generate temperature excursions exceeding 600°C and destroy the binder pore structure.

For routine lot release and catalyst cycle monitoring, a three-instrument arrangement is employed to close the analytical loop between reactor performance and product specification. Gas chromatography with flame ionisation detection uses a 30 m 0.32 mm internal diameter DB-WAX column with 0.5 µm film thickness, an oven programme from 40°C with a 3 min hold followed by 10°C min−1 to 240°C, and an internal standard calibration for EGME, EGDME, EGBE, butanol, DME, and butenes. The detection limit for 1,4-dioxane by this method is 5 mg kg−1. Karl Fischer coulometric titration per ASTM E203 is used for water in feed and product, with a repeatability below 0.05 wt% in the range of interest. Inductively coupled plasma optical emission spectrometry monitors sodium, potassium, and aluminium in MEG feed; the sodium detection limit of 0.05 mg kg−1 is sufficient to verify guard bed breakthrough. Acid site density is measured on fresh and regenerated catalyst samples by ammonia temperature-programmed desorption; the integrated desorption signal between 150°C and 500°C is calibrated against a 0.45 mmol g−1 standard. The analytical programme is run at the start of each catalyst cycle, every 24 h during stable operation, and at 4 h intervals during regeneration. Product specifications for EGME and EGBE used in industrial solvent applications require water below 0.1 wt%, acidity below 0.01 mg KOH g−1, colour below 10 Pt-Co, and 1,4-dioxane below 10 mg kg−1 for regulated cleaning and coatings uses. Compliance with these limits is documented under a quality management system conforming to ISO 9001:2015; batch records retain the reactor temperature, pressure, space velocity, regeneration count, and guard bed differential pressure for each product lot.