Performance study of a potentiometric sensor array for lactic proteins analysis

Abstract

Lactic proteins are complex biomolecules, possessing different molecular weights (from 19 kDa up to 66 kDa), which can be used as biomarkers for detecting food adulterations. This study reports for the first time the evaluation of two potentiometric sensor arrays, one electronic tongue (ET) using only cross-sensitivity chemical sensors and a hybrid-ET containing ionic-selective and cross-sensitivity chemical sensors. Although both devices could detect and distinguish soluble and insoluble lactic proteins, namely bovine serum albumin and three main casein fractions(alpha-, beta- and kappa-), the inclusion of ionic-selective sensors increased the device performance (100% of correct leave-oneout cross-validation classification for hybrid-ET against 94% for the ET) and reduced the number of sensors needed (5 for the hybrid-ET instead of 9 sensors used by the ET device). Finally, the results showed that with the hybrid-ET a second-order polynomial equation could be derived relating the measured potential of the aqueous solution and the logarithmic of the total casein concentration, with satisfactory determination coefficients (R2>0.99).