Exploratory study of sugar production using fluorescence spectroscopy and multi-way analysis

This paper is concerned with the possibility of obtaining chemically meaningful models of complicated processes by the use of fluorescence spectroscopy screening and the unique parallel factor analysis (PARAFAC) model. The second-order nature of fluorescence excitation emission data and the fact that the PARAFAC model has no rotational indeterminacy mean that in certain cases, it is possible to decompose complex mixture signals into contributions from individual chemical components. Relating the thus obtained information to, e.g., important quality parameters, it is possible to analyze, understand, predict and monitor the quality based on a chemical foundation. The proposed approach thus gives a direct link between process analytical chemistry and multivariate statistical process control.