Simultaneous Quantification of Nine Major Water-Soluble Inorganic Ions Using a Potentiometric Electronic Tongue in Cheese Samples

This study presents the applicability of a new potentiometric electronic tongue (ET) for simultaneous quantification of nine ions (Na
⁺
, K
⁺
, Ca
²⁺
, Mg
²⁺
, Cl
^-
, PO4
^3-
, SO4
^2-
, NO3
^-
, and NH4
⁺
) in various kind of commercial cheese samples. A sensor array comprising 12 solid-state ion-selective electrodes was employed; whose recognition elements were various selectophores with specific and generic responses. A Bayesian regularization algorithm was applied in the topology of the selected artificial neural networks (ANNs) with 12 inputs, 3 and 9 neurons in the hidden and output layers, respectively. A train set of 30 solution standards were set up to establish the model. For evaluation of the model validation, 10 external test solutions were applied and the model accurately predicted the concentration of all ions in these synthetic solutions. The results of a paired student's t-test indicated no significant variation between nominal and obtained concentrations in the validation set, confirming the applicability of proposed ET for simultaneous quantitative analyses of ions. Finally, Mineral ion content of 18 commercial cheese samples was determined by ET. The results showed an improvement from 65.0% to less than 12.5%) according to ion chromatography reference method (for all relative errors in comparison with ion-selective electrodes individually. It is believed that the proposed ET system can be potentially applied as a suitable device for multiple determinations of common mineral ions in the food industry.