SECM visualization of spatial variability of enzyme–polymer spots: 1. Discretisation and interference elimination using artificial neural networks

Enzyme–polymer layers immobilized on an electrode surface often serve as basis for amperometric biosensors. Caused by the formation process they show spatial variability in the polymer thickness which corresponds to a variability of immobilized enzyme activity. The relationship between topography and localized enzymatic activity of enzyme–polymer spots was studied using scanning electrochemical microscopy (SECM) in the feedback mode and generator-collector mode. Discretisation with a grid size corresponding to the scanning parameters defined substructures which can be treated as individual microsensors with specific response characteristics. The local responses are mainly governed by the polymer thickness but also influenced by neighbouring sites. Thus, discretisation allowed us to treat an enzyme–polymer spot with dimensions of about 300 μm diameter like an array of more than 400 individual microsensors. Using suitable selection criteria and multivariate calibration it was possible to identify sensing sites which are optimal for the determination of glucose. It was demonstrated that an artificial neural network which was trained with the data provided by SECM images well predicted glucose concentration in the presence of ascorbic acid.