Articles
The potential of near infrared reflectance spectroscopy (NIRS) for the estimation of quality parameters in tomato paste
Article number
1351_33
Pages
223 – 230
Language
English
Abstract
Non-destructive optical methods based on near infrared spectroscopy have been used for the estimation of physiological properties in many products.
In this work, the development of a robust spectroscopic procedure for determining, simultaneously and non-destructively, relevant quality parameters of tomato paste is described.
The samples were analyzed by reference methods and also by NIRS. A machine based on the Monochromator-NIR reflectance and transflectance analyzer (NIRS DS2500) was used to measure NIR transmittance-reflectance.
The NIRS calibration was Development by statistical models.
In this case, the program used was WINISI. In general, the results obtained are good.
Specially, excellent results have been obtained in Brix (R2=0.98) and it was very good correlation in color a/b, color a, color b, color L (R2=0.95, R2=0.94, R2=0.93 and R2=0.91, respectively). Bostwick showed a good correlation (R2=0.80). And titratable acidity showed a correct correlation to the prediction.
But this trend did not continue in salt content, pH and reducing sugar showed a bad correlation (R2=0.03, R2=0.34 and R2=0.46, respectively). As a result, this method should not be used as a method of reference but as control method in industrial processes.
In this work, the development of a robust spectroscopic procedure for determining, simultaneously and non-destructively, relevant quality parameters of tomato paste is described.
The samples were analyzed by reference methods and also by NIRS. A machine based on the Monochromator-NIR reflectance and transflectance analyzer (NIRS DS2500) was used to measure NIR transmittance-reflectance.
The NIRS calibration was Development by statistical models.
In this case, the program used was WINISI. In general, the results obtained are good.
Specially, excellent results have been obtained in Brix (R2=0.98) and it was very good correlation in color a/b, color a, color b, color L (R2=0.95, R2=0.94, R2=0.93 and R2=0.91, respectively). Bostwick showed a good correlation (R2=0.80). And titratable acidity showed a correct correlation to the prediction.
But this trend did not continue in salt content, pH and reducing sugar showed a bad correlation (R2=0.03, R2=0.34 and R2=0.46, respectively). As a result, this method should not be used as a method of reference but as control method in industrial processes.
Publication
Authors
M.J. Riballo Ruiz-Roso, R. Rodríguez-Corchado, M. Gómez-Cardoso, M.C. Carrillo, A. Crespo, P. Ramayo, J.L. Llerena-Ruiz, M.J. Redondo, A. Roldán, E. Valdezate
Keywords
near-infrared spectroscopy, NIRS, tomato paste, Brix, color, Bostwick, titratable acidity, salt content, pH
Online Articles (42)
