Abstract: In order to objectively evaluate the sweetness and characteristics of sweeteners, we investigated the feasibility of electronic tongue (ET) to evaluate the taste of 10 sweeteners, including sucrose, stevia, glycyrrhizin, mangosteen extract, neohesperidin dihydrochalcone (NHDC), naringin dihydrochalcone (NDHC), cyclamate, sucralose, sodium saccharin and acesulfame by comparison with sensory evaluation. The results showed that the sweetness sensor GL1 could effectively detect the sweetness of sucrose, NHDC, NDHC, sodium saccharin, acesulfame and cyclamate, and the response to cyclamate was negative. Based on the ET data, the 6 sweeteners were clearly classified by principal component analysis (PCA) and cluster analysis. Relative to sucrose, the sweetness of the other 4 sweeteners was expressed. A quantitative prediction model between the results of ET detection and sensory evaluation was established for each of the five sweeteners with positive sweetness response by least square method. The results showed a good correlation for the five sweeteners and good linear relationships for sucrose, NHDC and acesulfame. The correlation coefficient of prediction set (Rp) for each of the sweeteners was greater than 0.96, and the root mean square error was less than 0.9, indicating that the prediction model has high accuracy and good prediction performance.

Key words: sweeteners; electronic tongue; sensory evaluation; sweetness; prediction model