Abstract To enhance the accuracy of laser-induced breakdown spectroscopy (LIBS) in detecting heavy metal elements in solutions, a chitosan (CS)-Parafilm (PM) enrichment method is proposed. This method involves drying chitosan-heavy metal complexes on a Parafilm substrate. During the drying of droplets, electrostatic attraction and chelation by CS effectively concentrate analytes. Additionally, the hydrophobic effect of the PM substrate induces Marangoni flow, which drags analytes from the bottom edge to the central top of the droplet surface, effectively suppressing the coffee ring effect (CRE). Using LIBS technology, spectra were uniformly collected in an array on the surface of sediments. Through the analysis of cumulative excitation spectra and measurement repeatability, the results showed that the RSD of the emission lines Cu I 324.754 nm and Cu I 327.396 nm was reduced to 3.85% and 3.78%, respectively. This indicates that the CS-PM enrichment method allows for uniform deposition of analytes within the samples, effectively suppressing the CRE. Quantitative analysis of Cu elements using the CS-PM enrichment method was conducted using PSO-SVM, PSO-BPNN, and random forest (RF) algorithms. The RF algorithm demonstrated the best predictive performance with Rp 2 of 0.977, root mean square error of prediction of 3.86 mg l−1, mean absolute error of 3.0 mg l−1, and RPD of 5.52. Thus, it can be seen that CS-PM effectively improves the repeatability of spectral measurements and the accuracy of quantitative analysis predictions, providing a technical reference for enhancing the stability and quantitative performance of element analysis in liquids using LIBS.
Read full abstract