The aim of this work was to evaluate the potential of using front face fluorescence spectroscopy for rapid quantitative estimation of neoformed contaminants in industrially processed cookies. Two dimensional synchronous front face fluorescence spectra were acquired on cookies to assess the industrial process impact on the fluorescence signal and predict the neoformed contaminants content in cookies. The signal was recorded on two types of cookies; 41 samples taken from different steps of four industrial production lines and 148 cookie samples produced from experimental baking using three different temperatures with two levels of fat saturation and two types of sugar used in the formulation. After spectral pre-treatment of the acquired front face fluorescence spectroscopy data, the multiway arrays were decomposed by means of PARAFAC models. Factors extracted from the decompositions allowed identification of the main front face fluorescence spectroscopy profiles in cookies. These included native tryptophan and riboflavin, and several fluorescence profiles attributed to fat oxidation and Maillard reaction compounds. Relative intensities of the samples fluorescence profiles were then used to discriminate critical steps in the industrial baking process and to predict the content of chromatographically measured neoformed contaminants hydroxymethylfurfural, carboxymethyllysine and acrylamide in the cookies. The effects of spectral pre-treatments on decomposition and regression results were also studied. The results show that process control and neoformed contaminants estimation in industrially processed cookies can be achieved by means of front face fluorescence spectroscopy information.
- Front face fluorescence spectroscopy
- Generalized linear model
- Multiway arrays
- Neoformed contaminants