In this study, data fusion of laser-induced breakdown spectroscopy (LIBS) and Raman spectroscopy was used to characterize the total organic carbon content (TOC) of shale samples in Ordos Basin. The powder samples are pressed into pellets for LIBS and Raman analysis. Partial least squares calibration method was used for quantitative analysis. By comparing the quantitative results of the two techniques and the quantitative results of different spectral fusion strategies, it is found that the fusion strategy of LIBS and Raman spectral data can improve the prediction ability of the model. The model based on PLS latent variable fusion strategy has better fitting effect, its coefficient of determination (R2) for the calibration sets is 0.9915, the coefficient of determination for the prediction sets is 0.9574 and the root mean square error of the prediction set is 0.5580 wt%. The model has better generalization ability, and the root-mean-square-error-of-cross-validation (RMSECV) of the model is 0.5252 wt%. This study shows that data fusion of LIBS and Raman spectra can achieve rapid analysis of total organic carbon content in shale samples, and may be beneficial to carbon storage research in depleted gas shale.

中文翻译：

---

激光诱导击穿光谱和拉曼光谱结合偏最小二乘法检测页岩中有机碳


本研究采用激光诱导击穿光谱 （LIBS） 和拉曼光谱的数据融合对鄂尔多斯盆地页岩样品的总有机碳含量 （TOC） 进行了表征。粉末样品被压制成颗粒，用于 LIBS 和拉曼分析。采用偏最小二乘校准法进行定量分析。通过比较两种技术的定量结果和不同光谱融合策略的定量结果，发现 LIBS 和拉曼光谱数据的融合策略可以提高模型的预测能力。基于PLS潜在变量融合策略的模型具有较好的拟合效果，其对校准集的决定系数（R2）为0.9915，对预测集的决定系数为0.9574，预测集的均方根误差为0.5580 wt%。该模型具有较好的泛化能力，模型的交叉验证均方根误差 （RMSECV） 为 0.5252 wt%。研究表明，LIBS 和拉曼光谱的数据融合可以实现页岩样品中总有机碳含量的快速分析，可能有利于枯竭页岩中的碳储量研究。