Dynamic surface-enhanced Raman spectroscopy (D-SERS) was employed for the rapid detection of acephate in rice with simply regulated gold nanorods. Gold nanorods modified with cysteamine were prepared to circumvent the weak affinity of acephate molecules to the gold surface for a gigantic and stable enhancement. D-SERS was adopted to measure spectra of acephate residue at a range of 100.2-0.5 mg/L in rice samples, and the low residue of 0.5 mg/L can be still detected. Multivariant methods in machine or deep learning were used to develop the regression models for the automatic analysis of acephate residue level. Partial least squares regression and principal component analysis obtained the optimal performance with the root-mean-square error (RMSE) of validation of 5.4776, coefficient of determination (R2) of validation of 0.9560, RMSE of prediction of 6.2845, and R2 of prediction of 0.9541. Thus, the proposed method provides accurate and sensitive detection for acephate in rice.
Keywords: Acephate; Dynamic surface-enhanced Raman spectroscopy; Gold nanorods modified with cysteamine; Multivariant methods; Residue level analysis.
Copyright © 2019 Elsevier Ltd. All rights reserved.