Interactive self-modelling of chemical reaction systems using multivariate data analysis

This paper concerns with an application of Principal Component Analysis (PCA) in the determination of stoichiometric models from on-line spectroscopy for mechanism elucidation, process optimisation and on-line quality control. A theoretical basis for working with interactive factor modellisation is proposed and suited to study semibatch processes. Several chemometric tools based on PCA are selected and joined in order to select the number of reactions R and the number of chemical absorbing species S that better describe the chemical reaction. A test of rank compatibility is discussed for a semibatch reaction. An experimental determination of the dimension of the stoichiometry enables a kinetic multidetermiantion method based on on-line spectroscopy and dynamic simulation to optimise kinetic parameters. An objective function based on orthogonal projections on the abstract pure spectra and abstract concentration profile space is proposed. The mathematically extracted concentration-time profiles were finally verified with traditional methods of quantitative analysis.