Hard-modeling of reaction kinetics by combining online spectroscopy and calorimetry



This paper presents a method for combining spectroscopic and calorimetric data for kinetic parameter estimation. The target testing (TT) procedure for spectroscopic data is used together with a recently published least squares approach where the reaction power is fitted solely based on kinetic parameters. We propose to autoscale (i.e. to column-wise mean center and divide by the column standard deviation) data and estimates and to combine the ‘autoscaled’ cost functions with appropriate weights. The weights are determined from the data. Simulations demonstrate how the calorimetric data also constrain the kinetic fit, and we apply the method to resolve the reaction kinetics of the synthesis of acetylsalicylic acid (ASA). Copyright © 2008 John Wiley & Sons, Ltd.