Abstract: Today, molecularly imprinted polymers have received much engrossment as selective adsorbents and substrates. In the present study, a magnetic molecularly imprinted polymer (MIP) based on clay, graphene oxide and polydopamine was synthesized and used for adsorption and specific entrapment of Diazinon pesticide. The magnetic MIP nanocomposite was characterized by XRD, FT-IR, SEM, VSM, and AFM methods. The response surface method (RSM) was used to optimize and modeling the effective parameters of the adsorption of Diazinon by this polymer, and the results were compared with an artificial neural network (ANN). The results showed magnetic saturation of GO/Clay/Fe3O4@PDA MIP nanocomposite is equal to 8.28 emu/g, average particle size, average surface roughness and imprinting factor of this polymer are equal to 105 nm, 66.72 nm and 2.35, respectively. This polymer has a very selective ability to adsorption of Diazinon by removal percentage equal to 99.85 at pH = 7, the concentration of Diazinon = 1.45 ppm, the dosage of polymer = 1.24 g/L, time = 16 min. Also, the RSM results showed that the model included the effect of parameters on the absorption efficiency is a quadratic polynomial equation with the coefficient of determination equal to 0.98% and the most effective parameter on removal percentage is the dosage of the polymer. The artificial neural network with 11 neurons in the hidden layer with the coefficient of determination of 0.96% was also suitable for modeling the absorption process. Compared to the ANN model, the RSM model has better capability and lower error values. Freundlich isothermal models and pseudo-second-order and Elovich kinetic models in different concentrations of Diazinon were appropriate for explanation the adsorption process. The mechanism of Diazinon adsorption by MIP nanocomposite was influenced by intraparticle and extraparticle diffusion. Also, GO/Clay/Fe3O4@PDA MIP nanocomposite has an adsorption capacity of more than 98% up to nine cycles of the adsorption/desorption process
Author keywords: Response surface method, Diazinon, Molecularly imprinted polymer, Graphene oxide, Polydopamine, Artificial neural network