Characterizing the Performance of a Resonance-Based MEMS Particle Sensor with Glass Beads

Fine dust measurement methods, such as weight change, beta ray measurement, light scattering, and resonance change, have been studied and used, but there is still room for improvement in terms of cost, system volume, and accuracy. In this study, we aimed to develop a low-cost, micro-scale, and highly accurate dust sensor using semiconductor processes. The sensor consists of a piezoelectric membrane with a diameter of 800 µm and a thickness of 2.8 µm. The resonance frequency and mode shape were measured and monitored using a Laser-Doppler-Vibrometer (LDV), and the resonance frequency shift was calculated based on the same mode shape. Vibration modes from (0,1) to (3,1) were observed within a 0–200 kHz frequency bandwidth. Spherical glass beads were used to measure resonance frequency shift for additional mass ranging from 0.863 to 4.52 µg. The first resonance mode (0,1) exhibited a clear proportional relationship between the resonance frequency shift and additional mass, while other modes exhibited non-proportional trends. Intriguingly, notable discrepancies in the resonance frequency shift emerged based on the spatial placement of the glass beads, even when the mass was the same. Additionally, we presented an approximate theoretical curve to portray the resonance frequency shift, which we subsequently validated against our measurement findings. This study presents a new attempt on the characterization method of piezoelectric mass sensors and provides a relatively simple and accurate analysis of the results.