Dynamic Measurement of VOCs with Multiple Characteristic Peaks Based on Temperature Modulation of ZnO Gas Sensor

Volatile organic compounds (VOC) harm human health seriously in the air. Therefore, it is essential to recognize VOC gases qualitatively and quantitatively. The dynamic measurement method can improve the selectivity of metal oxide semiconductor (MOS) gas sensors to VOC, but there is a problem of the insufficient number of characteristic peaks. From the experimental point of view, the primary judgment basis for the correct qualitative and quantitative recognition of VOC gases by the dynamic measurement method is the characteristic peak of the dynamic response signal. However, the traditional dynamic measurement method generally only has two characteristic peaks. In this experiment, the voltage was changed at the time of the second characteristic peak by controlling the constant dynamic response period. Taking ethyl alcohol as an example, the experimental results show that the characteristic peak of the dynamic response signal does not increase when the voltage is constant. However, a new characteristic peak will appear based on a continuously rising heating voltage. The characteristic peaks of the dynamic response of n-propyl alcohol, isopropyl alcohol, and n-butyl alcohol were also increased based on the rising heating voltage waveform. Based on the K-Nearest-Neighbors algorithm, the qualitative and quantitative recognition rate of the four alcohol homologue gases reached 100%.Keywords:dynamic response signal;characteristic peaks;ZnO gas sensor;periodic cycle heating waveform;alcohol homologue gases