Modeling the effect of temperature on relative humidity sensing

Abstract

In this work we have established a mathematical function to model the influence of the ambient temperature, relative humidity and frequency on the change of impedance. The proposed empirical description is a combination of the power function (modeling the temperature effect) and a reverse-sigmoid function inspired by the three-parameter Weibull cumulative distribution function (modeling the relative humidity and frequency effects). This empirical model was validated by using our iron manganite (FeMnO3) thick film measurement data. The reduction of impedance of thick film samples with change in relative humidity in the range 30-90% was measured in a temperature and climatic chamber in the frequency range 42 Hz-l MHz at three ambient/working temperatures (25, 50 and 75°C). The obtained experimental data was successfully fitted using the specially tailored empirical model. Application of this model enables prediction of the iron manganite sensor performance for different temperatures both within the confines of the analyzed temperature range (interpolation) and outside of it (extrapolation). Future development will include applying this model to analyzing the temperature influence on relative humidity sensing for other metal oxide sensing materials that have shown a similar dependence and in a wider ambient temperature range.