A novel enzyme based biosensor for catechol detection in water samples using artificial neural network

Biosensors could be used as digital devices to measure the sample infield. Consequently, computational programming along with experimental achievements are required. In this study, a novel biosensor/artificial neural network (ANN) integrated system was developed. Poly (3,4-ethylenedioxy-thiophene)(PEDOT), graphene oxide nano-sheets (GONs) and laccase (Lac) were used to construct a biosensor. The simple and one-pot process was accomplished by electropolymerizing 3,4-ethylenedioxy-thiophene (EDOT) along with GONs and Lac as dopants on glassy carbon electrode. Scanning electron microscopy (SEM) and electrochemical characterization were conducted to confirm successful enzyme entrapment. The modified electrode was employed to detect and measure catechol. The reaction of catechol and the prepared electrode was controlled by adsorption. Linear responses of the biosensor were over two ranges, 0.036–0.35 μM and 0.35–2.5 μM, with a detection limit of 0.032 μM. The proposed biosensor was tested in real water samples successfully. The experimental test results were applied to train ANNs by the back-propagation algorithm. The input and output parameters were current and catechol concentration, respectively. Results from ANN modeling complied well with the experiments, signifying its useful application in biosensor technology.