Usman Muhammad, Ali Wajid, Alarfaji Saleh S, Tamulevičius S
Institute of Materials Science, Kaunas University of Technology, K. Baršausko St. 59, LT-51423 Kaunas, Lithuania.
Key Laboratory for Micro-Nano Physics and Technology of Hunan Province, College of Materials Science and Engineering, Hunan University, Changsha, Hunan 410082, PR China.
Spectrochim Acta A Mol Biomol Spectrosc. 2025 Apr 5;330:125700. doi: 10.1016/j.saa.2025.125700. Epub 2025 Jan 5.
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.
表面增强拉曼散射（SERS）在快速、高灵敏度检测地表水生生态系统中痕量环境污染物方面显示出巨大潜力。在过去几年中，抗生素的广泛使用导致了水环境的严重恶化，其在废水中的大量残留污染物对生态系统产生有害影响，这与抗生素抗性菌株的发展有关。然而，在本研究中，一种核壳纳米粒子GNRs@1,4-BDT@Ag的新方法被用于结合计算方法，利用SERS技术定量测量废水溶液中抗生素的浓度。在我们的实验中，我们选择了废水中常用的抗生素，如环丙沙星和左氧氟沙星。然后，我们获得了每种抗生素及其不同浓度下各种组合的SERS光谱。我们将其与机器学习算法相结合，以准确识别和量化系统中残留抗生素的SERS光谱。随后采用主成分分析（PCA）和正交偏最小二乘判别分析（OPLS-DA）对SERS光谱数据集进行聚类分析。为了评估机器学习算法的性能，应用了五个指标。分类结果表明，虽然大多数算法在抗生素状态预测中达到了95%以上的准确率，但支持向量机（SVM）模型表现最佳，达到了高达99%的显著预测准确率。这种开发的方法有助于作为一种简单快捷的工具来分析废水中的抗生素，并在各个领域展现出更广泛应用的潜力。
