Abstract: Phosphorus is an indispensable key nutrient element for the growth of living organisms.However,the presence of excessive phosphorus in the environment can trigger eutrophication of water body.Therefore,it has become extremely urgent to implement effective phosphorus recovery measures.The aim of this study was to synthesize magnetically recoverable Fe₃O₄/La-MOF,design orthogonal experiments,and use a neural network model to predict adsorption performance.Molecular dynamics simulations were also employed to analyze the adsorption process of phosphate on Fe₃O₄/La-MOF.Based on relative importance analysis,the factors influencing phosphate adsorption were identified in descending order of importance as follows: initial phosphate concentration (54.0%),adsorbent dosage (19.1%),CO₃^2-(7.5%),SO₄^2-(7.2%),HCO₃^-(5.5%),Cl^-(3.2%),NO₃^-(1.9%),and pH(1.7%).The molecular dynamics simulations further revealed that the adsorption of phosphate by Fe₃O₄/La-MOF mainly existed on the large- and small-pore surfaces of La-MOF.Meanwhile,the mean square displacement curves and diffusion coefficients confirmed that phosphate was more easily adsorbed on the large pore size surface of La-MOF(010).This study provided technical reference and theoretical support for the recovery of phosphorus from Fe₃O₄/La-MOF in real wastewater and the adsorption mechanism of phosphate on Fe₃O₄/La-MOF.