Chiang Mai Journal of Science

In this work, classification of sweetness from different types of sugar using blended multiwall carbon nanotube paste electrode was demonstrated. The fabricated electrodes were used as working electrodes of multi-electrochemical sensors for sweetness classification. Multiwall carbon nanotube (MWCNT) was blended with modified materials including of copper oxide (CuO), nickel oxide (NiO), cobalt phthalocyanine (CoPc) and iron phthalocyanine (FePc). The sweet samples with same level of sweetness were prepared from 5 types of sweetener including glucose, fructose, maltose, sucrose and honey. The features of acquired cyclic voltammograms (CV) of the blended MWCNT electrodes in the sweet samples have been used as input data of principal component analysis (PCA) for dimensionality reduction and classification. The main features of CV derived from four fabricated working electrodes in various tasted solutions originated from their oxidation and reduction states. By the first two principle components (PCs), the well classification of five sugar types was achieved at the same level of sweetness. The sugar solutions were classified into two groups by mean of monosaccharide and disaccharide. Maltose and sucrose, as disaccharide sugars, showed similar characteristics and were placed in 2D PCA coordinate closer than those of fructose and glucose which are monosaccharide sugars. The PCA of honey resided in the area between those of mono- and di-saccharide sugars, due to the presence of both types in honey sample. The coefficients of the first two PCs also indicated that the CV features of MWCNT blended with CuO electrode had highest weight influent in classification of monosaccharide and disaccharide solution. The performance of the sensor constructed with the blended MWCNT exhibited promising sensor array for sweetness classification in electronic tongue.