Transient characterization techniques for resistive metal-oxide gas sensors

We present three different time-domain characterization techniques for resistive metal-oxide gas sensors. They are based on electric resistance transients induced by small step changes in either temperature, bias voltage, or concentration of reducing gas. Conduction model of granular semiconductors and an electronic trapping (adatom ionization) rate equation at the grain boundaries are employed. Fitting of the presented analytical transient models to the experimental data allows calculation of the time constant of the electronic trapping process, the height of the grain-boundary-potential barrier, the relative change of the occupied grain-boundary states, the resistance coefficient, and the effective number of grain-boundaries between the electrodes of the sensor. These values can be further used for studying the underlying physicochemical phenomena and increasing the selectivity of the sensor. Fitting of a simple model to the measured transient values yields the energies associated with the electronic trapping process and the reducing gas reaction with the preadsorbed oxygen, as well as the rate constant of the trap releasing and the reducing-gas parameter. The use of the proposed techniques is verified in experiments with commercial resistive WO3 and SnO2-based gas sensors in clean and humid air and in acetone and isopropyl alcohol vapours. The experiments were performed using a pulsing system for chemical vapours and a sensor–temperature–control system based on a field-programmable gate array (FPGA) processor. In the SnO2-based sensor the grain-boundary-potential barrier is 0.35 V at 270 °C and 0.49 V at 350 °C. At 80% relative humidity these values increase to 0.41 V and 0.52 V. In dry clean air the electronic trapping and releasing energies are 0.94 eV and 0.86 eV.