Large-scale chemical sensor array testing biological olfaction concepts

Biological olfactory systems are characterized by a large number of sensors with broad overlapping specificities. The sensitivity and selectivity of the system may be enhanced by the huge redundancy of the olfactory receptor neurons (ORNs). A European project, NEUROCHEM, was devoted to test computational models of the olfactory system of vertebrates and insects. To test these models, a realistic artifact of the olfactory epithelium was developed as a large sensor array mimicking some features of biological ORNs, in particular, the broad and overlapping selectivity to many odors, the combinatorial response, the high level of redundancy, and the different dynamic ranges exhibited by same types of ORNs. The sensor array is composed of 16 384 elements arranged in four smaller arrays of 64×64 interdigitated electrodes deposited on a borosilicate substrate. To mimic the redundancy of the biological ORNs, tens of organic conductive polymers were chosen as active sensing materials because of their broad and diverse, but overlapping, specificity to different classes of volatile organic compounds. These sensors were characterized by their responses to varying concentrations of test analytes. The collected sensor data were processed with standard multivariate techniques and the results are reported in this paper. © 2001-2012 IEEE.