Modeling the sensing characteristics of chemi-resistive thin film semi-conducting gas sensors†
Abstract
For chemi-resistive thin film gas sensors a generic theoretical model is proposed to predict the variation of sensor response with the operating temperature and thickness of the sensing film. A diffusion equation is formulated assuming that inflammable target gases move through the sensing film by Knudsen diffusion and react with the adsorbed oxygen following first-order kinetics. We have assumed a realistic non-linear variation between the conductance and test gas concentration and derived a general expression relating the sensor response to the operating temperature and thickness of the film. Assuming Langmuir adsorption kinetics, we have theoretically predicted the response and recovery transients during gas sensing using thin film sensing elements. It is predicted that for irreversible type sensing, the response time is reduced with an increase in test gas concentration, whereas for reversible sensing, the response time is independent of test gas concentration. For zinc oxide thin film sensors, an excellent match is obtained between the model prediction and experimental data for their thickness (122 nm to 380 nm) and temperature variation (200 °C to 325 °C) in 500 ppm carbon monoxide (CO) sensing. The maximum CO response% (∼53%) was achieved in 320 nm thick ZnO films. The conductance transients for response and recovery for CO sensing closely follow Langmuir adsorption kinetics and as predicted theoretically, indeed for irreversible sensing, the response time reduces from 350 s to 220 s with an increase in test gas concentration from 20 to 550 ppm. In the case of reversible sensing we found that the response time is ∼55 s irrespective of the CO gas concentration in the range of 5–500 ppm. The models developed in the present work are quite generic in nature and we have discussed their applicability to a wide variety of sensing materials with various types of surface morphologies.