Evaluating the Adulteration of Sodium Benzoate and Potassium Sorbate in tomato paste on the physicochemical properties of the tomato paste using an olfactory machine and artificial intelligence

Abstract

Abstract Tomato is the second most important product consumed worldwide. Since tomato paste is considered the most important seasoning in food, its quality and safety are of particular importance. The present study aims to investigate adulteration in tomato paste. To this aim, an olfactory machine system based on 10 gas sensors was used, and its potential to detect different levels of adulteration of sodium benzoate and potassium sorbate at 0, 0.05, and 0.1% in tomato paste was assessed. The effect of these preservatives on some physicochemical properties of tomato paste, such as pH, total soluble solid (TSS), precipitate weight ratio, and acidity, was also investigated. The results indicated that both preservatives in tomato paste significantly affected all its physicochemical parameters at the 1% significance level. The confusion matrix results showed that the LDA method has a high performance in detecting different percentages of preservatives with 100% accuracy. C-SVM and PCA were also recognized as suitable and accurate methods for detecting samples containing sodium benzoate and potassium sorbate at 0.05% and 0.1%. The results of the loading chart indicated that TGS2620, MQ135, and TGS2602 sensors were the most sensitive sensors in detecting adulterated tomato paste samples. Also, according to the results of predicting physicochemical parameters, PCR and MLR models were the most appropriate models to predict acidity and precipitate weight ratio, Brix, and pH, respectively. Overall, it can be stated that the electronic nose is an appropriate tool to detect food adulteration in all adulteration types, thereby saving substantial time and money.