Computational Fluid Dynamics Simulation, Microelectromechanical System Fabrication, and Radio-Frequency Evaluation of the PM2.5 Fine Dust Sensor Based on the Surface Acoustic Wave Resonator

Abstract

To classify 2.5-µm-scale fine dust and measure its concentration, research was conducted to produce a fine-dust sensor utilizing computational fluid dynamics simulation and microelectromechanical system processes, and the measurement system was manufactured for the sensor. The virtual impactor was designed to classify particles in the air above and below 2.5 µm, and the cut-off diameter of 2.5 µm at an acceleration nozzle length of 83.2 µm was obtained by verifying this impactor through flow-analysis simulation. The microheater was then designed to measure the mass of the classified particle by adsorbing it onto the surface acoustic wave resonator placed on the lower part of the flow channel using thermophoretic force. The conditions where the classified particles are completely adsorbed on the surface acoustic wave resonator through the heater were calculated using numerical modeling and the results were verified through the simulation. The fine dust sensor based on the designed surface acoustic wave resonator is composed of the upper, middle, and lower parts including the heater, virtual impactor, and surface acoustic wave resonator and SMA connector, respectively, and the micro-fan controlling the total flow. Each part was produced using the microelectromechanical system process, assembled, and finally finished. Moreover, the evaluation system was produced in this study to assess the developed sensor. The flow of air controlled through the flowmeter was sprayed into a dust bottle filled with hollow silica powder to scatter the powder, which flowed into the system with the air, making the environment similar to one with fine dust present. The change in radio-frequency resonance characteristics, such as those before and after exposure to the fine dust environment, and insertion loss after the surface cleaning were observed, and a sensitivity of, detection limit of, and restoration rate after surface cleaning exceeding 99.9% were obtained.