Determination of sucrose concentration in lemon-type soft drinks by dispersive Raman spectroscopy. Issue 3 (2009)

• Record Type: Journal Article
• Authors: Silveira Jr. Silveira Jr., Landulfo Landulfo; Moreira Moreira, Leonardo Marmo Leonardo Marmo; Conceição Conceição, Viviane G. B. Viviane G. B.; Casalechi Casalechi, Heliodora L. Heliodora L.; Muñoz Muñoz, Ingrid S. Ingrid S.; Da Silva Da Silva, Fabiano Fernandes Fabiano Fernandes; Silva Silva, Marcos Augusto S. R. Marcos Augusto S. R.; De Souza De Souza, Renato Aparecido Renato Aparecido; Pacheco Pacheco, Marcos Tadeu T. Marcos Tadeu T.
• Is Part Of: Spectroscopy. Volume 23:Issue 3/4(2009)
• Page Start: 217
• Page End: 226
• Publication Date: 2009
• Subjects: Sucrose quantification -- soft drinks -- industrialized food -- Raman spectroscopy -- PLS
• DOI: 10.3233/SPE-2009-0383
• Languages: English
• ISSNs: 0712-4813

Abstract

Abstract : The objective of this study was to quantify the sucrose amount in commercial lemon-type soft drinks through dispersive Raman spectroscopy, comparing the amount listed in the nutritional table of each product to the predicted by a least-square model, in order to obtain a method for quality assurance applied to soft drinks. A dispersive Raman spectrometer was employed using 830 nm laser and imaging spectrograph coupled to a CCD camera, and a total of 48 samples from four brands of lemon-type soft drinks were analyzed. A calibration curve using sucrose from refined sugar (sugarcane) diluted in spring water was elaborated in the range between 0 and 15.0 g/100 ml, and a quantification model based on Partial Least Squares (PLS) regression was developed to correlate the Raman spectra and the amount of sucrose in each dilution. Then, the sucrose in each soft drink sample was predicted employing the calibration curve. The mean error of calibration for the PLS method was 0.30 g/100 ml (3.0%). Results indicated that soft drinks samples have predicted sugar content ranging from 8.1 to 10.9 g/100 ml, with an error of the predicted value compared to the nutritional table ranged from 1.1% to 5.5%. Therefore, Raman spectroscopy in association with PLS regression was an effective method for quantifying the sucrose, with small prediction error. Thus, the present work allows to infer auspicious possibilities of Raman spectroscopy application in the quantification of relevant nutritional facts in beverages.