Quantitative analysis of four PAHs in oily sludge by surface-enhanced Raman spectroscopy (SERS) combined with partial least squares regression (PLS) based on a novel nano-silver-silicon coupling substrate
Changfei Ma,Qun Zhang,Jing Liang,Shan Yang,Tianlong Zhang,Fangqi Ruan,Hongsheng Tang,Hua Li
DOI: https://doi.org/10.1016/j.saa.2024.124531
IF: 4.831
2024-05-30
Spectrochimica Acta Part A Molecular and Biomolecular Spectroscopy
Abstract:Polycyclic aromatic hydrocarbons (PAHs) present in oily sludge generated by the petroleum and petrochemical industries have emerged as a prominent concern within the realm of environmental conservation. The precise determination of PAHs holds immense significance in both petroleum geochemistry and environmental protection. In this study, a combination of surface-enhanced Raman spectroscopy (SERS) and solid–liquid extraction was employed for the screening of PAHs in oily sludge. Methanol was utilized as the extraction solvent for PAHs, while nanosilver-silicon coupling substrates were employed for their detection. The SERS spectrum was acquired using a portable Raman spectrometer. The nano silver-silicon coupling substrate exhibits excellent uniformity, with relative standard deviations (RSDs) of Phenanthrene, Fluoranthrene, Fluorene and Naphthalene (Phe, Flt, Flu and Nap) being 2.8%, 1.08%, 1.41%, and 5.44% respectively. Moreover, the limits of detection (LODs) achieved remarkable values of 0.542 μg/g, 0.342 μg/g, 0.541 μg/g, and 5.132 μg/g. The quantitative analysis of PAHs in oily sludge was investigated using SERS technology combined with partial least squares (PLS). The optimal PLS calibration model was optimized by combining spectral preprocessing methods and using the SiPLS (Synergy interval partial least squares)-VIP (Variable Importance in Projection) hybrid variable selection strategy. The prediction performance of the D1st (First derivative)-WT (Wavelet transform)-SiPLS-VIP-PLS model was deemed satisfactory, as evidenced by high R 2 P values of 0.9851, 0.9917, and 0.9925 for Phe, Flt, and Flu respectively; additionally, the corresponding MREP values were found to be 0.0580, 0.0668, and 0.0669 respectively. However, for Nap analysis, the D1st-WT-PLS model proved to be a better calibration model with an R 2 P value of 0.9864 and an MREP (Mean relative error of prediction) value of 0.0713. In summary, SERS technology combined with PLS based on different spectral pretreatment methods and mixed variable selection strategies is a promising method for quantitative analysis of PAHs in oily sludge, which will provide new ideas and methods for the quantitative analysis of PAHs in oily sludge.
spectroscopy