Dynamic vision sensor enhanced laser-induced breakdown spectroscopy: parameter optimisation and correction method for improving spectral stability

Abstract

Laser-induced breakdown spectroscopy (LIBS) exhibits broad application prospects but suffers from poor signal stability, significantly compromising analytical precision. Correction of spectral signals using plasma optical signals is highly promising. The dynamic vision sensor (DVS), with microsecond temporal resolution and a dynamic range over 120 dB, is particularly suited to capturing plasma optical signals. This study presents a new approach that integrates a DVS into the LIBS system to capture key plasma parameters and provide effective correction for LIBS signals. To obtain high-quality spectral and plasma optical signals, comprehensive spectral analysis was first conducted to determine the optimal LIBS parameters of 95 mJ laser energy and 1.5 μs delay time. In parallel, DVS parameters were optimised using event frame reconstruction and statistical analysis, resulting in the configurations of an F2.0 aperture, 5 cm collection distance, and 0° collection angle. In addition, a spectral correction model (DVS-SC) was established, which utilises the plasma area and the number of “On” events generated by increased light intensity, as extracted by the DVS, to achieve signal correction. Experimental validation on copper alloys and carbon steel indicated that, compared with the original data, the calibration curve R² values for Cu I 327.396 nm, Zn I 328.233 nm, and Mn I 403.076 nm increased to 0.944, 0.956, and 0.941, corresponding to improvements of 61.1%, 49.8%, and 81.3%, respectively. The average relative standard deviations decreased to 3.173%, 10.317%, and 0.872%, respectively, demonstrating clear improvements over the original results. The method proposed in this study combines simplicity, cost-effectiveness, and high efficiency, thereby providing considerable practical value for on-site applications.