Doublet line ratio-based self-absorption correction in LIBS applied to bacterial concentration quantification

Laser-induced breakdown spectroscopy (LIBS) is widely recognized for its rapid, low-destructive, and real-time elemental analysis capabilities. However, the complexity of bacterial samples and spectral signal interference caused by self-absorption effects present significant challenges to the accuracy and reliability of LIBS for quantitative bacterial concentration analysis. To address these challenges, a novel self-absorption correction method based on doublet line intensity ratios and K-parameter ratios is proposed and, for the first time, successfully applied to improve the accuracy of bacterial concentration quantification. Laser parameters were optimized to verify that the bacterial laser-induced plasma satisfies local thermodynamic equilibrium (LTE). Characteristic doublet lines sharing similar or identical upper and/or lower energy levels (Ca II 393.36/396.84 nm, Ca I 612.22/616.22 nm, and K I 766.49/769.90 nm) were selected to develop a model correlating the K-parameter ratio with the self-absorption coefficient (SA), enabling effective spectral line self-absorption correction. The results show that after self-absorption correction, the quantitative analysis errors were significantly reduced: for Escherichia coli and Bacillus subtilis, the errors decreased by 4.0 % and 11.6 % using Ca II 393.36 nm, 9.1 % and 6.9 % using K I 766.49 nm, and 4.3 % and 3.9 % using Ca I 612.22 nm, respectively. Accurate quantification was achieved within the bacterial concentration range of 10⁷ to 10⁸ CFU/mL. This correction method is straightforward, efficient, and independent of complex parameters or external conditions, significantly improving the precision and reproducibility of LIBS in bacterial concentration analysis. Additionally, the SVM-based classification accuracy for the five bacterial species improved markedly, rising by 3.55 %. Its robust generalizability supports its extension to other microbial detection applications, providing a rapid, accurate, and cost-effective approach. These findings lay a solid foundation for broader LIBS applications in biological monitoring, with promising potential in environmental surveillance, food safety, and public health diagnostics.