Research On Quantitative Analysis Of Multi-Component Antibiotic Mixture Solutions Based On Raman Spectroscopy

Nowadays antibiotics are widely used for the prevention and treatment ofinfectious diseases, which play an irreplaceable role in the medical and livestockindustry. China is one of the countries that produce and consume a large number ofantibiotics, the consumption per capita averaged of which had been ten more timesthan other countries. The antibiotic residuals in water environment or food willtransfer to human body through food chain, causing serious threats to people’s health.In this case, developing a fast and precise method for the determination of antibioticresiduals is of great significance. Raman spectroscopy could characterize structures atthe molecular level based on Raman scattering effect. With the advantages of fast,nondestructive, requiring little or no sample preparation, and suitable for water-richsamples, Raman spectroscopy are applied to the detection and analysis area.This paper has studied the characteristics of Raman spectroscopy and itsadvantages for the determination of antibiotics, the principles of Raman scatteringeffect and the quantitative and qualitative analysis by Raman spectroscopy, as well asthe spectral data preprocessing methods and multivariate analysis. We measured theRaman spectra of ternary mixture solutions of three typical cephalosporin antibiotics(ceftriaxone, cefotaxime sodium, and cefazolin sodium) by a scientific-gradespectrometer, and the quantitative analysis were made accordingly.We measured the Raman spectra of the solid and aqueous samples of threecephalosporin antibiotics respectively. The characteristic peaks of Raman spectrawere extracted by subtracting the background and their vibration modes wereanalyzed. The calibration and prediction sets were designed orthogonally. Variouspreprocessing methods were applied to spectral data to evaluate the effect onperformance of partial least square models.The quantitative analysis of the ternary cephalosporin solutions by Ramanspectroscopy were conducted in this paper. Different pretreatments were discussed tooptimize and improve the regression models, including the elimination of outliers, theextraction of characteristic spectral variables, and the preprocessing correction etc.The parameters and results of the models were analyzed to find the optimal one topredict the external samples. The proposed method of this paper improved the spectral data pretreatments andconstructed a steady calibration model, which could determine the concentration ofeach component in antibiotic mixture solutions rapidly and simultaneously.