LIBS quantitative analysis based on multi-model calibration marked with characteristic lines

Abstract

Long-term reproducibility remains one of the important challenges in laser-induced breakdown spectroscopy (LIBS) quantitative analysis. In this work, a novel LIBS quantitative method based on multi-model calibration marked with characteristic lines was proposed. Under identical experimental equipment and parameters, multiple calibration models were established by using LIBS data collected at different time intervals. Simultaneously, the characteristic line information, which reflects variations in experimental conditions, was marked as the characteristic of each calibration model. During the analysis of unknown samples, the optimal calibration model was selected for quantitative analysis by characteristic matching. Taking the analysis of Mo, V, Mn, and Cr elements in alloy steel as an example, ten calibration models were established based on daily spectral data, and the test samples were quantitatively validated for five days. The results indicate that, compared to the single calibration model, the calibration model selected through the matching of characteristic lines significantly improves the average relative errors (ARE) and the average standard deviations (ASD). The method proposed in this study provides a new quantitative analysis idea for LIBS technology, which can effectively improve the reproducibility of LIBS long-term repeated measurements.