Luminescent diagnostics of amoxicillin in milk by spectra statistical parameters by the discriminant analysis method

The aim of the study is to investigate the possibility of diagnosing penicillin antibiotics using amoxicillin in cow's milk as an example using luminescence at an excitation wavelength of 445 nm using discriminant analysis. Objectives: determining the wavelength of the greatest excitation of milk by ultraviolet and visible optical radiation to measure the spectral characteristics of photoluminescence; calculating statistical parameters of the luminescence spectra of milk with different amoxicillin concentrations; obtaining discriminant functions and determining the most informative statistical parameters to monitor the presence of amoxicillin in milk. Residual antibiotics in milk are a serious problem not only for processing plants but also for consumers of such products. The dependence of the spectral coefficient of effective absorption on the wavelength of milk was obtained, with the absorption peaks of milk at wavelengths of 290; 324; 360; 445 nm. The maximum absorption of radiation occurs at a wavelength of 445 nm. The peak at 445 nm is 5.41 p.u., which is 2.3 times greater than that at 290; 324; 360 nm. When adding an antibiotic to milk, the dispersion σ2, asymmetry As, luminescence energy E, eV of the milk spectra decrease, and the mathematical expectation Mλ, excess Ex increase. The accuracy of the discriminant analysis model using the mathematical expectation Mλ of the spectrum is 89 %, with a dispersion σ2 of 100 %, with asymmetry As of 86 %, with excess Ex of 76 %, with luminescence energy E of 96 %. The obtained results can initiate the development of luminescent milk alarms with a sensitivity to penicillin antibiotics of 0.001 μg/cm3.