Algorithm for improving the sizing accuracy in real-time bioaerosol single particle mass spectrometer

Abstract

The newly developed bioaerosol single particle mass spectrometer (Bio-SPAMS) has been innovatively designed for its optical sizing system. The first laser beam in the previous single particle mass spectrometer was split into near distance double beams, similar to the design of APS (Aerodynamic Particle Sizer) and SBS-LIBS (Single Beam Splitting-Laser Induced Breakdown Spectroscopy). All the particles focused by the aerodynamic lens can be sized and got number concentration statistic. However, due to the imperfect beam quality and the large scattering intensity of the large-sized particles, there may be some noise in the scattered signals, particle diameter measured by this sizing system was often larger than actual value if the same trigger threshold was set. In this study, when measuring PSL microspheres with diameters of 1.9, 3.1, and 4.9 μm, the identification rates of the fixed threshold algorithm were only 75.25%, 55.26%, and 0.27%, respectively. To address such issue, we developed a dynamic threshold waveform recognition algorithm based on field programmable gate array (FPGA), which could process the photoelectric signals collected by a photomultiplier tube (PMT) in real time. The algorithm can dynamically adjust the trigger threshold of the collected scattered signals and accurately calculate the interval time between the near distance double beam. For PSL microspheres with diameters of 1.9, 3.1, and 4.9 μm, the accuracy of the dynamic threshold algorithm increased by 19.09%, 25.72%, and 88.20%, respectively. This algorithm effectively solves the problem of particle sizing deviation, and improves the particle size measurement accuracy of the bioaerosol mass spectrometer in a wide particle size range from 0.3–6 μm.