Fabrication of Integrated Chemical Field-Effect transistor Humidity Sensor Array and Signal Processing Using Artificial Neural Network

Based on the conventional MOSFET, a novel chemical field-effect transistor (ChemFET) humidity sensor array has been designed and fabricated, which is a device composed of 4 gateless, depletion mode n-channel ChemFETs. And poly (3,4-ethylenedioxythiophene)/poly (styrenesulfonate) (PEDT/PSS) multilayered ultrathin films as gate sensitive layer were prepared by a novel self-assembly (SA) method. The effect of humidity has been examined and the drain current shows a fall with the increasing humidity. A Generalized Regression neural network (GRNN) has been designed for the prediction of humidity to overcome the nonlinearity problem of the ChemFET sensor. The drain current responses of ChemFET sensor under different drain voltage were processed by GRNN and the spread constant of network, as main affection factor to the outputs of GRNN, was discussed in this paper. It was shown that the best prediction results were obtained using the drain current responses of ChemFET sensor under a 1V drain-source bias as the inputs of network when the spread constant was equal to 0.14. This processing mode can accomplish the quantitative analysis of humidity with a low relative error (1.77%).