Heterogeneously Catalysed Reactive Distillation: Material Data, Experiments, Simulation and Scale-up using the Example of Hexyl Acetate Synthesis

Reactive distillation can raise conversion and selectivity by removing products in situ, but feasibility is constrained by thermodynamics and chemistry and the intensified operation complicates predictive modeling and scale-up, especially for heterogeneously catalyzed systems with coupled vapor–liquid–solid interactions. Within the EU project INTINT, this work studies heterogeneously catalyzed reactive distillation using esterification of 1-hexanol with acetic acid to n-hexyl acetate and water (Amberlyte CSP2 catalyst), including potential by-products hexene and dihexyl ether. A large experimental and modeling program is reported, comprising reactive-distillation experiments at laboratory (55 mm) and pilot (162 mm) scale and the creation of a comprehensive database for phase equilibria, chemical equilibrium, and reaction kinetics. In total, 32 lab-scale and 12 pilot-scale column runs are conducted with different catalytic internals (Katapak-S, Multipak-IT; and Katapak-S 250.Y/500.Y) while varying configurations and key operating parameters (heat duty, refluxes, loads, reactant ratios, pressure). Conversions of roughly 80–90% are achieved, clearly exceeding the equilibrium conversion (~66% for stoichiometric feed without recycle), and by-product formation is negligible in lab scale but reaches selectivities up to ~13% in pilot scale, with data quality supported by reproducibility and mass-balance checks. Because literature data were incomplete, new VLE measurements for key binaries and new LLE measurements for water-containing subsystems are generated; two NRTL parameter sets are derived (one for VLE, one for LLE) and validated against quaternary equilibrium data. Reaction data are established via 40 plug-flow kinetic experiments (60–130 °C) for main and side reactions, plus 90 chemical-equilibrium measurements (80–120 °C), and autocatalysis is shown to be negligible. An activity-based kinetic framework is developed and compared in pseudo-homogeneous and adsorption-based forms; the pseudo-homogeneous model is recommended because the adsorption-based approach performs poorly, likely due to limitations of immersion adsorption data in representing macropore conditions. Column simulations show that assuming equilibrium reaction on each stage is invalid; incorporating kinetics improves results but requires a transfer factor to map reactor-measured kinetics to reactive-distillation conditions. This transfer factor is nearly constant across lab runs (~0.29 with ~6% standard deviation), enabling practical transfer with a single parameter. Trickle-bed experiments and fluid-dynamic analysis indicate roughly half of the factor arises from bypass flow and non-ideal catalyst utilization within the internals, while boiling conditions in the column may further reduce rates via partial evaporation of water in catalyst pores. Rate-based simulations (PROFILER) and equilibrium-stage simulations yield quantitatively similar results and essentially identical transfer factors, implying the simpler equilibrium-stage model is sufficient and that kinetic transfer—not mass-transfer modeling depth—is the critical issue. Fully predictive simulation of pilot-scale experiments using only lab-derived transfer-factor information matches concentration and temperature profiles well and underestimates conversion by only about 2% (relative), though limited sensitivity prevents a definitive final validation of the overall scale-up procedure. The study also highlights that large reaction zones can promote by-product formation (e.g., hexene) and suggests that targeted side-reaction kinetics can be used at development stage to mitigate this risk. Overall, the work shows that reliable transfer of reaction kinetics is the key step for successful design and scale-up of heterogeneously catalyzed reactive distillation and provides practical methods to achieve it.