Sensors (Oct 2020)

Detection and Quantification of Tomato Paste Adulteration Using Conventional and Rapid Analytical Methods

  • Flora Vitalis,
  • John-Lewis Zinia Zaukuu,
  • Zsanett Bodor,
  • Balkis Aouadi,
  • Géza Hitka,
  • Timea Kaszab,
  • Viktoria Zsom-Muha,
  • Zoltan Gillay,
  • Zoltan Kovacs

DOI
https://doi.org/10.3390/s20216059
Journal volume & issue
Vol. 20, no. 21
p. 6059

Abstract

Read online

Tomato, and its concentrate are important food ingredients with outstanding gastronomic and industrial importance due to their unique organoleptic, dietary, and compositional properties. Various forms of food adulteration are often suspected in the different tomato-based products causing major economic and sometimes even health problems for the farmers, food industry and consumers. Near infrared (NIR) spectroscopy and electronic tongue (e-tongue) have been lauded as advanced, high sensitivity techniques for quality control. The aim of the present research was to detect and predict relatively low concentration of adulterants, such as paprika seed and corn starch (0.5, 1, 2, 5, 10%), sucrose and salt (0.5, 1, 2, 5%), in tomato paste using conventional (soluble solid content, consistency) and advanced analytical techniques (NIR spectroscopy, e-tongue). The results obtained with the conventional methods were analyzed with univariate statistics (ANOVA), while the data obtained with advanced analytical methods were analyzed with multivariate methods (Principal component analysis (PCA), linear discriminant analysis (LDA), partial least squares regression (PLSR). The conventional methods were only able to detect adulteration at higher concentrations (5–10%). For NIRS and e-tongue, good accuracies were obtained, even in identifying minimal adulterant concentrations (0.5%). Comparatively, NIR spectroscopy proved to be easier to implement and more accurate during our evaluations, when the adulterant contents were estimated with R2 above 0.96 and root mean square error (RMSE) below 1%.

Keywords