Mechanistic insight into the selective detection of dopamine using a copper ferrite-based nanocomposite, supported by a machine learning approach.

Painless and instant analysis of neurotransmitters from an abandoned body fluid, like urine, would facilitate point-of-care (PoC) detection of several neurological disorders. As a step towards developing a non-invasive, fast, and PoC analytical tool for detecting urinary catecholamines, planar electrochemical sensors have been modified with a spinel ferrite-based nanocomposite comprising copper ferrite (CF) and reduced graphene oxide (rGO), and subsequently used for selective detection of dopamine in simulated urine. The sensors exhibited dopamine sensitivity in the range of 0.1-300 μM with 74 nM as a lower limit of detection (LoD); which are significant in the clinical context of detecting urinary catecholamines. Sensitive and selective detection of dopamine, over potential interferents like uric acid and ascorbic acid, on the functionalized electrodes was apparently accentuated by the presence of rGO as well as Cu ion-stimulated adsorption of dopamine on the working electrode and the associated mechanism has been delineated. Finally, to establish a real-time outlook regarding machine-based accurate identification of analytes, deep neural network (DNN) has been deployed to classify the electrochemical response of the neurotransmitter dopamine in simulated urine. The applied DNN model has attained an average classification accuracy of 97.8 % along with accurate concentration prediction, thus, ensuring selective identification and quantification of the target neurotransmitter in a complex matrix.