Simulation of the hydrogenation of p-hydroxybenzaldehyde was used as a prototype to develop a bi-phasic reactor model describing both reaction kinetics and phase separation. The effects of phase behavior modeling on bi-phasic reactor model predictions were investigated. Specifically, case studies were conducted to examine the prediction accuracy of the NRTL1 activity coefficient model equipped with parameters from: UNIFAC2 group contributions, QSPR3 generalized NRTL parameters (NRTL-QSPR), and regressed NRTL model parameters (NRTL_R). Also, sensitivity analyses were conducted to evaluate the influence of the kinetic parameters on reactor model predictions.
Kinetic data on the p-hydroxybenzaldehyde system from the OU CIRE research group was used in the simulation. Aspen Plus and Visual Basic for Application (VBA) were used to model the bi-phasic reaction process, accounting for both the reaction and phase separation. A user-defined MS Excel (VBA) reaction model was designed, and Aspen Simulation Workbook (ASW) was used to link the reaction model to the phase separation model on Aspen Plus.
The NRTL model parameters estimated from the QSPR generalized model (NRTL-QSPR) show better phase equilibria property predictions than those generated using the UNIFAC–predicted model parameters.
The sensitivity analyses indicate that the adsorption constants for p-hydroxybenzaldehyde and 4-methylcyclohexanol (KA and KD) as well as the deactivation constant (kd) have significant impact on the bi-phasic model predictions of p-hydroxybenzaldehyde hydrogenation.
Concentration (mM) vs. Time (min) Plots for various NRTL model parameters: