Research On The Adsorption Behavior Of Typical Antibiotics By Microplastics In Water

Plastics,as a polymer material,are ubiquitous in human production and life.However,plastic waste has become a huge environmental problem.Microplastics,which are also known as"PM 2.5"in water,have become a research focus both domestically and abroad.In addition to their inherent toxicity,the unique physical and chemical properties of microplastics,such as small volume,large specific surface area,and low density,make them prone to interact with other pollutants in water,leading to comprehensive pollution and making the pollution situation even more complex and variable.Therefore,this article selected commonly used polyethylene（PE）as the research object,and subjected it to aging treatment and biofilm attachment.The surface morphology,crystallinity,functional groups,and total amount of biofilm of PE before and after aging and biofilm attachment were analyzed by characterization methods.The adsorption mechanism between PE and two types of antibiotics,tetracycline hydrochloride（TC-HCl）and cefalexin（CLX）,was studied through adsorption kinetic models and adsorption isotherm models,and the effects of different types of microplastics（PE,Polypropylene（PP）,polystyrene（PS）,and polyvinyl chloride（PVC））,p H,and ion strength on the adsorption process were examined.Finally,10tetracycline antibiotics（TCs）were selected as research objects,and quantum chemical calculations based on density functional theory（DFT）were used to obtain molecular descriptors,which were used to establish a quantitative structure-activity relationship（QSAR）model through multivariate linear（MLR）regression analysis and partial least squares（PLS）regression analysis using the parameter log K_OW representing the hydrophobicity of the antibiotic.This model was used to predict the hydrophobicity of TCs and infer their adsorption effect on microplastics.The main research results are as follows:（1）After aging,the surface of PE（A-PE）showed cracks,increased specific surface area,decreased crystallinity,and increased oxygen-containing functional groups.PE with attached biofilm（B-PE）and aged PE with attached biofilm（B-A-PE）showed an increase in the total amount of biofilm,an increase in crystallinity,and the appearance of oxygen-containing functional groups.This indicates that the physical and chemical properties of PE itself will change after aging and biofilm attachment,thus affecting the adsorption process of antibiotics.（2）On the same type of microplastic,the adsorption effect of TC-HCl is better than that of CLX because the log K_OW value of TC-HCl is greater than that of CLX,indicating that hydrophobic interaction is an important adsorption mechanism.The adsorption process of antibiotics on microplastics can be fitted with a pseudo-second-order kinetic model and Freundlich adsorption isotherm model,indicating that the adsorption process is nonlinear and multilayered,with physical adsorption as the main mechanism and chemical adsorption also present.Aging and biofilm attachment can increase the adsorption of antibiotics on microplastics.Different types of microplastics have different adsorption effects on antibiotics due to their different physical and chemical properties.The p H can affect the adsorption of antibiotics on the surface of microplastics through electrostatic interaction,and ion strength can inhibit the adsorption process.（3）By manually computing 8 molecular structural descriptors and utilizing DFT-based quantum chemical calculations to obtain 38 quantum chemical descriptors,a correlation analysis was conducted between the obtained 46 descriptors and the corresponding log K_OW values of TCs.It was discovered that only 5 descriptors were significantly correlated with log K_OW values,with the descriptor vσ_tot² having the strongest correlation.The QSAR model established by PLS had a higher explanatory power for log K_OW values compared to the model built by MLR.The evaluation and validation of the models built by both methods revealed that the prediction performance of the MLR model was inferior to that of the PLS model,which was able to predict the majority of log K_OW values.