Reaction Kinetics and Distillation of Formaldehyde-containing Mixtures

Formaldehyde is a major chemical intermediate (about 12 million tonnes per year worldwide), produced mainly from methanol and used largely for resin manufacture and polyacetal plastics. Because of its high reactivity, it is handled not as a pure compound but in aqueous, methanol-containing solutions where it is mostly bound in a complex spectrum of oligomeric reaction products. Distillation of formaldehyde-containing mixtures is industrially important, but column performance is strongly affected not only by non-ideal liquid-phase behavior but also by slow oligomerization kinetics, since chemical-equilibrium times often exceed tray residence times. While vapor–liquid equilibria and caloric properties have been studied extensively, there has been little work on modeling and simulating distillative separations for these reacting systems, and existing kinetic models are either insufficiently validated or thermodynamically inconsistent. This work therefore develops a thermodynamically consistent reaction-kinetic model for formaldehyde-containing mixtures and applies it to process simulation of distillation based on new, extensive experimental data. The kinetic database is expanded into the distillation-relevant range using quantitative online ^1H NMR measurements over wide temperatures and pH (20–100 °C, pH 2–8), enabling correlation of rate constants and introducing a kinetic framework in which monomeric formaldehyde reacts to oligomers. As a thermodynamic basis, the vapor–liquid equilibrium model of Albert et al. is adopted, but its description of chemical equilibrium proves inadequate; additional ^13C NMR equilibrium measurements are therefore performed in collaboration with Kaiserslautern, leading to an improved equilibrium model. To validate the combined equilibrium/kinetic framework, large sets of laboratory-scale distillation experiments for the formaldehyde–water–methanol system are carried out with BASF within the EU project INTINT, addressing the scarcity and limited industrial relevance of prior distillation data. Process simulations are implemented in Aspen Plus via external program links, enabling simulation of distillation and absorption/desorption processes. Simulations using equilibrium alone show clear mismatches, whereas incorporating the newly developed kinetics yields reliable reproduction of the complex distillation behavior. The resulting model is already being applied in industrial practice.