Benzoic acid (BA) is widely used as an antimicrobial preservative to prolong the shelf-life of pickled vegetables. A method for rapidly determining the BA content in forty pickled vegetable samples was developed by coupling ultrasonic extraction with surface-enhanced Raman scattering (SERS) and an adaptive iteratively reweighted penalized least-squares (AirPLS) algorithm. The results obtained with this method were compared and correlated with those from high-performance liquid chromatography measurements. Amplification of the Raman scattering via the SERS effect was induced by gold nanoparticles (AuNPs) when BA was irradiated with a 785 nm laser. The AirPLS algorithm was used to reduce the background interference signal, which was also amplified. The amplified Raman scattering effect of BA in the pickled vegetables displayed a positive and significant correlation with the HPLC concentration of BA, with high reproducibility. For HPLC determination of the concentration of BA in the range of 0-820 ppm, the BA monomer's intensity of the 944-1,005 cm-1 and 1,366-1,373 cm-1 peaks, and BA dimer's intensity of the 1,025 cm-1 and 1,465-1,482 cm-1 peaks in the SERS spectrum were respectively converted to the Z-ratio BA monomer and Z-ratio BA dimer standard scores by Z-Score conversion. The sum's (Z-ratio BA monomer + Z-ratio BA dimer) sensitivity was 100%, and specificity was 90.9% by receiver operating characteristic curve. This study found that a Raman spectroscopy-based monitoring method can be one of the fastest screening inspection options that can complete an analysis within a short period of time and produce reliable results. This approach is particularly cost-effective, which makes it suitable for the initial screening of raw materials and provides an effective management strategy easy to communicate with food safety officials.
Keywords: Raman; benzoic acid; risk management; spectral analysis.
© 2020 The Authors. Food Science & Nutrition published by Wiley Periodicals LLC.