A machine learning–based microorganism classification and a novel synthesis of stabilized tetragonal and cubic ZrO₂ nanoparticles: A future carrier for antimicrobial drug delivery

A supervised machine learning model has been used to classify eight representative microorganisms—Amoeba, Euglena, Hydra, Paramecium, rod-shaped bacteria, spherical bacteria, spiral bacteria, and yeast—with 98% accuracy and high F1-scores (0.98–0.99). After successful classification of the microorganisms, a nontoxic nanosystem of stabilized tetragonal and cubic ZrO₂ have been developed that can be used as potent antimicrobial agent and a biocompatible nanocarrier for antimicrobial drug delivery. To develop such potent nanosystem, a fully transparent colorless Al ³⁺ modified ZrO(OH) ₂ ⋅xH ₂ O gel has been developed by a chemical method of Al-metal surface hydrolysis in aqueous solution of zirconium oxychloride at room temperature. After that, Al ₂ O ₃ (Al ₂ O ₃ content ≤ 5 mol %) stabilized ZrO ₂ nanoparticles have been synthesized in a metastable t-ZrO ₂ phase by a reconstructive thermal decomposition of this amorphous hydroxyl transparent gel in air at temperature 500 °C. The t-phase of ZrO ₂ changes to c-phase with the increase in Al ₂ O ₃ content (Al ₂ O ₃ content ≥ 10 mol %). The phase formation in Al ₂ O ₃ stabilized metastable t- or c-ZrO ₂ nanoparticles has been revealed by X-ray diffraction study of the specimen before and after annealing at temperature 500 °C. No independent Al ₂ O ₃ recrystallizes in a perfect Al ₂ O ₃ : ZrO ₂ solid solution. This present work may show a new perspective in both the classification of the microorganisms and development of a new biocompatible nanosystem that can be used as antimicrobial agent and in antimicrobial drug delivery. Future work will focus on detailed antimicrobial and drug targeting studies of this nanosystem.