Compensation for the variation of total number density to improve signal repeatability for laser-induced breakdown spectroscopy

Weilun Gu,Zongyu Hou,Weiran Song,Liang Li,Xiang Yu,Jiacen Liu,Yuzhou Song,Muhammad Sher Afgan,Zheng Li,Zijun Liu,Zhe Wang
DOI: https://doi.org/10.1016/j.aca.2022.339752
IF: 6.911
2022-01-01
Analytica Chimica Acta
Abstract:High signal uncertainty has been regarded as a critical obstacle for the quantitative analysis of laser-induced breakdown spectroscopy (LIBS). One of the most effective ways for uncertainty reduction is to directly compensate for the variation of plasma properties, especially total number density. However, reliable compensation for the variation of total number density is hard to implement. In this work, we propose a data pre-processing method, called total number density compensation (TNDC), to reduce signal uncertainty. It is established on an assumption extended from the internal standard method and utilizes a weighted sum of emission lines from all major elements to reflect the variation of total number density. The TNDC method is tested on 29 brass samples and outperforms common normalization methods based on the spectral area in terms of signal repeatability and analytical performance. For Cu, the mean pulse-to-pulse relative standard deviation (RSD) of signals is greatly decreased from 5.10% to 1.03%, which is almost the best signal repeatability that LIBS can achieve and is comparable to that of ICP-OES. The root mean square error of prediction (RMSEP) and the mean RSD of prediction are decreased from 6.56% to 0.60% and from 12.00% to 1.03%, respectively. While for Zn, the mean RSD of signals improves from 6.43% to 4.12%, and the RMSEP is reduced from 1.57% to 0.59% with the RSD of prediction from 5.41% to 4.18%. Results demonstrate that TNDC can be an effective method for LIBS analysis especially for repeatability improvement. (C) 2022 Elsevier B.V. All rights reserved.
What problem does this paper attempt to address?