Morphology-driven spectral extraction method for enhanced liquid–solid transition LIBS detection of heavy metals in water

Abstract

In the detection of heavy metals in water by laser-induced breakdown spectroscopy (LIBS) via liquid–solid transition, the morphology of the solute region exhibits significant variability and randomness. To ensure full-coverage scanning of the solute region, a large enough rectangular laser scanning array is commonly utilized. However, this approach may incorporate spectra that are not ablated from the solute region, thereby impacting the spectral stability and the limit of detection (LOD). In this work, a morphology-driven spectral extraction method was proposed. It aims to extract the spectra of the solute region from the full-coverage scanning spectra, and then exclude the interference of invalid spectra. To assess the validity of the method, a series of 1:1:1 mixed solutions of the elements cadmium (Cd), manganese (Mn), and chromium (Cr) was prepared. After full-coverage scanning by a large enough rectangular array, the solute region spectra were extracted using the morphology-driven spectral extraction method. Quantification was performed using the spectra both before and after extraction, then comparing the spectral stability and quantitative accuracy. The results indicated that using the extracted spectra could significantly improved quantitative performance. Spectral stability was enhanced by a 38.8% reduction in the relative standard deviation (RSD), ensuring more consistent and reliable measurements. The LOD was reduced by 62.6%, enabling highly sensitive detection of trace-level concentrations. In conclusion, the morphology-driven solute region spectral extraction method proposed in this work can effectively mitigate the impact of invalid spectra, thereby enhancing the spectral stability and lowering the LOD for heavy metal detection in water by LIBS.