Green reduction of silver nanoparticles for cadmium detection in food using surface-enhanced Raman spectroscopy coupled multivariate calibration.

Cadmium (Cd) causes pervasive harm on human health as a poisonous heavy metal. This study proposed a surface-enhanced Raman spectroscopy (SERS) approach using sodium alginate (SA) as green reductant in combination with edge enrichment and chemometrics to build label-free Cd quantitative models. The silver nanoparticles synthesized by SA had good dispersion and enhancement factor (3.48 × 10). The optimal detection system was established by optimizing the concentration of specific molecules (trimercaptotriazine) and the droplet volume of measured liquid. Partial least squares models based on preprocessing methods and selection algorithms were compared. The results indicated that the model combined with first-order derivative preprocessing and competitive adaptive reweighted sampling algorithms achieved the best performance (R = 0.9989, RMSEP = 1.6225) with the limit of detection of 2.36 × 10 μg L in food. The SERS approach combined with edge enrichment and chemometrics holds promise for rapid and label-free determination of Cd in food.