(Article)
a Faculty of Commodity Science, Poznań University of Economics and Business, al. Niepodległości 10, Poznań, Poland
b Universidade Do Algarve, FCT, DQB, CIQA, Campus de Gambelas, Faro, Portugal
b Universidade Do Algarve, FCT, DQB, CIQA, Campus de Gambelas, Faro, Portugal
Abstract
Total fluorescence spectra (excitation-emission matrices, EEM) were recorded for a series of commercial apple juices, including clear and cloudy juices produced from concentrate, cloudy juices that were not from concentrate, and freshly squeezed juices. An exploratory study of the spectra with parallel factor analysis (PARAFAC) revealed three groups of fluorophores with different emission properties, and these properties were characterized by excitation/emission maxima at 270/315 nm, (310, 370)/455 nm, and 430/(550, 680) nm, respectively. A regression analysis of the total fluorescence spectra arranged into three-way arrays using N-way partial least squares regression methods (NPLS1 and NPLS2) and an analysis of the unfolded spectra by partial least squares methods (PLS1 and PLS2) revealed quantitative relations between the fluorescence and antioxidant properties of juices. The best models for the total phenolic contents and total antioxidant capacities were obtained by applying the NPLS1 method to the EEM. The model parameters were as follows: R2 CV = 0.802, RPD = 2.3 for the total phenolic content and R2 CV = 0.808 and RPD = 2.3 for the total antioxidant capacity. These results show the potential use of fluorescence spectroscopy for screening apple juices for their antioxidant properties. © 2016 Elsevier Ltd. All rights reserved.
Author keywords
Antioxidants; Apple juice; Fluorescence; Multivariate regression; Parallel factor analysis
Indexed keywords
Engineering controlled terms: Agents; Antioxidants; Factor analysis; Fluorescence; Fluorescence spectroscopy; Fruit juices; Fruits; Multivariant analysis; Regression analysis
Apple juice; Excitation emission matrices; Multivariate regression; Parallel factor analysis; Partial least squares regression; Partial least-squares method; Total antioxidant capacity; Total phenolic content
Engineering main heading: Least squares approximations
ISSN: 03088146 CODEN: FOCHDSource Type: Journal Original language: English
DOI: 10.1016/j.foodchem.2016.05.007Document Type: Article
Publisher: Elsevier Ltd
Sikorska, E.; Faculty of Commodity Science, Poznań University of Economics and Business, al. Niepodległości 10, Poland; email:ewa.sikorska@ue.poznan.pl
© Copyright 2016 Elsevier B.V., All rights reserved.
© Copyright 2016 Elsevier B.V., All rights reserved.