Development Of A Voltammetric Electronic Tongue And Its Application To The Detection Of Several Kinds Of Foods

In the recent 20 years, the researches of electronic tongue have made big progress. Several laboratories have done lots of work on the development of electronic tongue, and some companies to be allowed to produce the electronic tongue. The device has the advantage of high portability for measurements with lower costs and good reliability. They are particularly useful for the analysis of liquid of solid samples. Recently, electronic tongue has been extensively used to test the food quality. However, the device is difficult to popularize in developing country due to the high price. The research and development of new device with low power and cost has become a popular research in the recent years. In this study, a voltammetric electronic tongue (VE-tongue) based on six metallic electrodes was developed to classify honey samples of different floral origins, to classify and predict the marked ages of different types of Chinese rice wines, to study the correlation between firmness and sugar content of different pear cultivars and to detect antibiotic residues in bovine milk. The main conclusion is as follows:The VE-tongue is composed of sensors array, data acquisition unit and pattern recognition unit. The VE-tongue software system was developed based on the concept of virtual instrument in LABVIEW platform. The VE-tongue is composed of gold, silver, platinum, palladium, tungsten, and titanium working electrodes, and the Multi-frequency large amplitude pulse voltammetry (MLAPV), which was composed of four individual frequencies:0.001k Hz,0.01k Hz,0.1k Hz, and 1k Hz, was taken as scanning waveforms.The VE-tongue was performed to classify monofloral honeys of seven kinds of floral origins. Two eigenvalues (the maximum value and the minimum value) of each cycle were extracted for building the first database (FDB); four eigenvalues (the maximum value, the minimum value, and two inflexion values) were exacted for building the second database (SDB). The two databases were analyzed by three-pattern recognition techniques:Principal component analysis (PCA) and cluster analysis (CA), respectively. It was possible to discriminate the seven kinds of honeys of different floral origins completely based on FDB and SDB by PCA and CA, and FDB was certificated as an efficient database by contrasting with the SDB. A rheometer was used to study commercial honeys, and Principal component regression (PCR) and Partial Least Squares regression (PLSR) were used to predict the viscosity of honey samples based on VE-tongue signals, the regression analysis showed that all the models worked well. The Astree electronic tongue developed by Alpha M.O.S. (Toulouse, France) and the VE-tongue were compared in the classification of honey samples of different floral origins. Three regression models:PCR, PLSR, and Least Squared-Support Vector Machines (LS-SVM) gave a clear indication of the two types of e-tongues ability, and a positive trend in the prediction of floral and geographical origin was found with the help of the models. Moreover, the regression models in prediction of the four kinds of honeys of different geographical origins based on ZU e-tongue performed very stable.The VE-tongue was performed to discriminate the difference between Chinese rice wines in this research. Three types of Chinese rice wines with different marked ages (1,3 and 5 years) were classified based on VE-tongue by PCA, CA. Three types of Chinese rice wines could be classified completely by PCA and CA, and some interesting regularity is shown in the score plots with the help of PCA. Two regression models:PLSR and Back-error propagation-Artificial neural network (BP-ANN) were used for wine age prediction. The regression analysis showed that the marked ages of the three types of Chinese rice wine were successfully predicted by PLSR and BP-ANN.The VE-tongue with Magness-Taylor (M-T) technique was used to study the correlation between firmness and sugar content of different pear cultivars. Five cultivars of pears from different geographical origins were tested by VE-tongue and M-T technique with four pattern recognition techniques:PCR, PLSR, and LS-SVM. The M-T technique were more promising for prediction of sugar content of one cultivar pear and less suitable for different cultivars of pears. LS-SVM performed good in prediction, but PLSR had a bad result. The VE-tongue could be used to predict sugar content and firmness of different cultivars of pears. All the three regression models performed good in prediction, and LS-SVM had the best result.The VE-tongue was performed to detect antibiotic residues in bovine milk. Six antibiotics spiked at four different concentration levels were classified based on VE-tongue by two pattern recognition methods:PCA and Discriminant Function Analysis (DFA). The six antibiotics at the MRLs could not be separated from bovine milk completely by PCA, but all the samples were demarcated clearly by DFA. Three regression models:PCR, PLSR, and LS-SVM were used for concentrations of antibiotics prediction. All the regression models performed well, and LS-SVM had the most stable results.