Plasma state and quantitative analysis comparison of single-pulse and long–short double-pulse laser-induced breakdown spectroscopy for multi-element analysis in steel samples

This study systematically compared collinear long–short double-pulse laser-induced breakdown spectroscopy (LS-DP-LIBS) with single-pulse LIBS (SP-LIBS) for multi-element quantification in 15 alloy-steel samples spanning major alloying, trace metallic, and non-metallic elements. Plasma excitation temperature and electron density were used to characterize excitation enhancement, while partial least squares regression (PLSR) provided the primary quantitative benchmark and machine-learning models assessed cross-model consistency. Compared with SP-LIBS, long–short double-pulse excitation enhanced plasma temperature and electron density, increased characteristic line intensities by 2.5 to 13.8 times, and reduced the average pulse-to-pulse relative standard deviation (RSD) by 53.25%. Enhancement was more pronounced for higher-excitation-energy lines, indicating that LS-DP-LIBS can preferentially strengthen spectral features that are less susceptible to self-absorption and therefore more beneficial for quantitative calibration. These spectral and plasma improvements transfer to better quantitative performance for the major metallic elements Ni, Cr, and Mn, with a particularly marked improvement for C. For the trace elements Al, Cu, S, Si, and P, LS-DP-LIBS partly mitigated matrix-related spectral variability. Quantitative performance nevertheless remained constrained by weak responses, Fe-matrix interference, and limited model robustness. PLSR generally yielded higher coefficients of determination and lower prediction errors under LS-DP-LIBS. Furthermore, the multi-model comparison showed that the overall analytical advantage of LS-DP-LIBS was maintained across several feature-processing and regression strategies within the present dataset. Collectively, by linking the enhanced plasma state to element-dependent improvements in signal repeatability and quantitative performance, this study demonstrates that LS-DP-LIBS provides a more reliable analytical basis than SP-LIBS for multi-element quantification in alloy steels.