Removal of boron in desalinated seawater by magnetic metal-organic frame-based composite materials: Modeling and optimizing based on methodologies of response surface and artificial neural network

A new type of metal organic frame-based magnetic composite adsorbent UiO-66-NH 2 /graphene oxide/Fe 3 O 4 was used to eliminate the adverse effects of boron in the desalinated seawater. Through serial characterization, it was found that this adsorbent has the advantages of strong magnetic response, good pore structure and abundant functional groups. On the basis of single-factor experiments on various influencing factors, the simulation and prediction capabilities of the response surface method and artificial neural network method for the adsorption process were studied and compared. Both methods shown strong capacity to make effective prediction on the process of adsorption. However, the error of response surface method is smaller than the error of artificial neural network method. Result of optimization by response surface method shows that the optimal conditions of adsorbent dosage, pH, and temperature are 99.1 mg/L, 3.38 and 318 K. Following the above conditions, the boron adsorption capacity of the adsorbent obtained in the real desalinated seawater environment is 22.46 mg/g. In addition, the adsorbent shows good recyclability. Pseudo-second-order kinetic model, intra-particle diffusion model, and Langmuir isotherm model are all suitable for describing adsorption process. Thermodynamic research results reveal the fact that this process is spontaneous and endothermic, and the adsorption of boron by the adsorbent is a composite process consisting of physical adsorption and chemical adsorption.