Fusion strategies of internal standard method based on image feature for enhanced quantitative analysis of steel alloy elements by LIBS

Abstract

Laser-induced breakdown spectroscopy (LIBS) often suffers from significant signal fluctuations in the quantitative analysis of alloy elements in steel. Although the conventional internal standard method is commonly employed due to its simplicity, it critically depends on the presence of an internal standard element at a high and stable concentration as well as on the expert-driven manual selection of spectral lines. To address these limitations, this study introduces an internal standard method based on image features (ISIF) and proposes two novel fusion approaches by integrating ISIF with image-assisted (IA) and spectral normalization (SN) methods, designated as ISIF-IA and ISIF-SN, respectively. Experimental results demonstrate that the standalone ISIF approach slightly improves the coefficient of determination (R²) for chromium and eliminates the need for manual spectral line selection, underscoring its applicability in complex matrix analyses. The fusion methods, ISIF-IA and ISIF-SN, substantially enhance the analytical accuracy, with particularly notable gains in the quantification of silicon and vanadium. Among these, the ISIF-SN method yielded the most significant improvement in signal stability, reducing the relative standard deviations (RSDs) of the validation samples for Si, V, and Cr from 19.66%, 17.34%, and 16.22% to 7.36%, 6.00%, and 3.45%, respectively. Additionally, both fusion strategies effectively mitigated the self-absorption effects in the Si and Cr spectral lines. This work confirms that coupling ISIF with IA or SN strategies markedly improves the quantitative analytical performance of LIBS in steel alloy analysis, thereby offering a robust and efficient approach for rapid and accurate in situ compositional assessment in industrial settings.