Related Experiment Video
Updated: Oct 25, 2026
Fabrication of polydimethylsiloxane (PDMS)-Based Flexible Surface-Enhanced Raman Scattering (SERS) Substrate for Ultrasensitive Detection
Published on: November 17, 2023
Rapid detection of trace dibutyl phthalate in refined cottonseed oil using surface-enhanced Raman spectroscopy
Lingdong Zhao1, Huan Chen1, Jianing Liu1
1Key Laboratory of Agricultural Product Processing and Quality Control of Specialty (Co-construction by Ministry and Province), School of Food Science and Technology, Shihezi University, Shihezi 832000, Xinjiang Uygur Autonomous Region, PR China; Nutrition and Safety Control of Xinjiang Production and Construction Corps, School of Food Science and Technology, Shihezi University, Shihezi 832000, Xinjiang Uygur Autonomous Region, PR China; Engineering Research Center of Storage and Processing of Xinjiang Characteristic Fruit and Vegetables, Ministry of Education, School of Food Science and Technology, Shihezi University, Shihezi 832000, Xinjiang Uygur Autonomous Region, PR China.
Abstract:
Rapid screening of dibutyl phthalate (DBP) in edible oils is complicated by lipid-derived background and refining-stage-dependent spectral variation. We developed a surface-enhanced Raman spectroscopy (SERS)-chemometric workflow using Au-coated Si nanocone arrays and 180 cottonseed-oil samples from 30 production batches at six refining stages. Each spectrum was paired with a sample-specific UHPLC-MS/MS reference concentration. Full-spectrum partial least-squares regression (PLSR) served as the latent-variable baseline and was compared with multiple linear regression (MLR), ridge regression (RRM), and recursive feature elimination with cross-validation coupled to ridge regression (RFECV-RRM). Stage-based leave-one-group-out (LOGO) and production-batch-grouped GroupShuffleSplit (GSS) validation were used, with all supervised preprocessing and feature selection confined to the corresponding outer-training data. RFECV-RRM performed best under LOGO (RMSE = 19.640 ± 2.432 ng g-1, R2 = 0.918 ± 0.013, RPD = 3.569 ± 0.252) and GSS (RMSE = 35.153 ± 4.369 ng g-1, R2 = 0.876 ± 0.030, RPD = 2.953 ± 0.370). Relative to PLSR, RFECV-RRM reduced mean RMSE by 54.9% under LOGO and 24.3% under GSS. Mean matrix-spike recovery was 98.1%. The workflow is suitable for evaluation as an at-line screening approach, subject to independent refinery, substrate-lot and instrument validation.
