Characterisation of Zinc Oxide Thin-Film Solidly Mounted Resonators for Particle Sensing in Air

Monitoring particulate matter concentrations is of particular importance within the overall assessment of indoor and outdoor air quality and its impact on human health. Bulk acoustic wave (BAW) technology offers low cost, robust alternative to widely employed optical measurements. However, due to its reasonably new application within the area of particulate measurements, an in-depth characterisation of this technology to temperature and humidity variations that are inevitably present within the indoor and outdoor environment is necessary step in order to determine its suitability for possible commercialisation. This work presents the characterisation of the temperature and humidity dependence of solidly mounted resonators for particulate matter sensing. Both theoretical results, obtained through modelling and simulation, and experimental result, obtained within the laboratory conditions, are analysed and compared. A 1.5 GHz resonator with a zinc oxide thin film is modelled using a one-dimensional equivalent circuit model, and finite-element methods based on both two-dimensional model and the three-dimensional model. The simulation results show that the temperature dependence of the resonator is strongly dependent on the material properties and crystal structure of the zinc oxide film. Our models estimate the temperature coefficient of frequency to be -30 to -40 ppm/Centigrade. This theoretical temperature dependence was comparable to experimentally measured value of ca. -49 ppm/Centigrade. In addition to temperature characterisation of discrete devices, the resonator was combined with read-out circuitry, which was also simulated and tested experimentally. The temperature coefficient of frequency in this case was found to be much higher at -220 ppm/Centigrade demonstrating the necessity for temperature control or temperature compensation within the complete system in practical applications. The effect of humidity was also investigated. The experimental mean resonant frequency shift per percent increase in relative humidity of the ambient air was found to be -3.8 kHz/%RH, while the models showed negligible sensitivity to humidity variations. Finally, preliminary experiments were conducted within the controlled lab environment showing promising result for possible application of this technology in particulate matter monitoring. The resonant frequency shift of approximately 300 kHz was measured after mass loading the sensing area of the resonator with estimated 24 ng of standard Arizona dust particles.