Rapid qualitative and quantitative analysis of benzo(b)fluoranthene (BbF) in shrimp using SERS-based sensor coupled with chemometric models.

Benzo(b)fluoranthene (BbF), a polycyclic aromatic hydrocarbon (PAH), is a carcinogenic contaminant of concern in seafood. This study developed a simple, rapid, sensitive, and cost-effective surface-enhanced Raman scattering (SERS) sensor (AuNPs) coupled with chemometric models for detecting BbF in shrimp samples. Partial least squares (PLS) regression models were optimized using uninformative variable elimination (UVE), bootstrapping soft shrinkage (BOSS), and competitive adaptive reweighted sampling (CARS). Qualitative analysis was performed using principal component analysis (PCA), linear discriminant analysis (LDA), and k-nearest neighbors (KNN) to differentiate between BbF-contaminated and uncontaminated shrimp samples. The SERS-sensor exhibited excellent sensitivity (LOD = 0.12 ng/mL), repeatability (RSD = 6.21%), and anti-interference performance. CARS-PLS model demonstrated superior predictive ability (R2 = 0.9944), and qualitative analysis discriminated between contaminated and uncontaminated samples. The sensor's accuracy was validated using HPLC, demonstrating the ability of the SERS-sensor coupled with chemometrics to rapidly and reliably detect BbF in shrimp samples. • The developed sensor can offer rapid and reliable BbF detection. • PCA, LDA, and KNN differentiated authentic from BbF-spiked shrimp. • CARS-PLS outperformed other predictive PLS models. • Successfully integrated SERS with chemometrics to detect BbF in shrimps. • Validated the sensor's accuracy through a comparative study with HPLC. [ABSTRACT FROM AUTHOR]