(569bj) Computer-Aided Kinetic Model Development for Liquid-Phase Processes: Unraveling the Oxidation of Cyclohexane

The design and operation of chemical production processes requires a thorough understanding of, among others, the reaction itself. An accurate kinetic model is thereby a powerful and valuable tool. Significant progress has been made in the development of accurate kinetic models for various gas-phase reactions. However, this is in stark contrast to liquid-phase reactions governed by radical chemistry for which there is a lack of fundamental knowledge. A new kinetic modeling tool was developed, denoted as ‘ALKIMO’ (Automated Liquid-phase Kinetic Modeler), to transpose gas-phase kinetic models into liquid-phase ones by accounting for the impact of the solvent. As proof-of-concept, the liquid-phase oxidation of cyclohexane has been investigated, which is the most important industrial process for production of cyclohexanone and cyclohexanol, of which the molecular chemistry is still not fully understood. To enable the construction of kinetic models for liquid-phase reactions, ALKIMO needs to assign accurate thermodynamic and kinetic parameters to species and reactions, respectively. Ideally, these parameters are obtained from experimental measurements or quantum chemical calculations. Since these methods are time-consuming, where possible, ALKIMO makes use of approximative methods such as equations of state, or group additivity theory, to calculate solvation energies. Quantum chemical calculations for the liquid phase were performed with the Gaussian and COSMOtherm software. To validate the developed kinetic model based on first principles, experiments were performed with a new continuous slug flow reactor unit at industrially relevant conditions, i.e., at a pressure of 1.0 MPa and temperatures between 150 and 180°C. The experiments were carried out for varying residence times, which enables thorough validation of the primary and secondary chemistry in the kinetic model. Model simulations were performed with a two-phase reactor model to mimic the slug flow regime. The experimental trends can be well predicted by the kinetic model without fitting of adjustable parameters.