Trace detection of antibiotics in wastewater using tunable core-shell nanoparticles SERS substrate combined with machine learning algorithms.

Surface-enhanced Raman scattering (SERS) show great potential for rapid and highly sensitive detection of trace amounts of contamination from the environment in the surface aquatic ecosystem. The widespread use of antibiotics has resulted in serious degradation of the water environment in the past few years, and their substantial residual contamination of wastewater has a harmful effect on ecosystems, which is associated with the development of antibiotic-resistant bacterial strains. However, in this study, a novel approach of core-shell nanoparticles GNRs@1,4-BDT@Ag was used for the quantitative measurement of the concentration of antibiotics in wastewater solutions using the SERS technique coupled with computational methods. In our experiments, we selected commonly used antibiotics such as ciprofloxacin and levofloxacin in wastewater solutions. We then obtained SERS spectra for each antibiotic and its various combinations at varying concentrations. We combined it with machine learning algorithms to accurately identify and quantify the SERS spectra of the residual antibiotics in the system. Principal Component Analysis (PCA) and Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA) were subsequently employed for clustering analysis of the SERS spectral datasets. To evaluate the performance of machine learning algorithms five metrics were applied. The classification results demonstrate that while most algorithms achieved over 95 % accuracy in antibiotics status prediction, the Support Vector Machine (SVM) model had the best performance, attaining a remarkable prediction accuracy of up to 99 %. This developed approach helps as a simple and expeditious tool for the analysis of antibiotics in wastewater and exhibits potential for broader applications in various domains.