Dynamic adsorption models and artificial neural network prediction of mercury adsorption by a dendrimer-grafted polyacrylonitrile fiber in fixed-bed column

In this paper, a dendrimer-grafted polyacrylonitrile fiber (PANF-TU) was successfully synthesized and characterized by scanning electron microscope with energy dispersive spectroscopy for removal of mercury. Adsorption feasibility was thoroughly studied during batch and dynamic tests. Adsorption kinetics , isotherms and effects of pH and coexisting substances were performed with the maximum adsorption capacity of 227 mg g −1 . The better fitting characteristics of pseudo-second-order and Langmuir models revealed the dominant role of monolayer chemisorption during the adsorption process . Various variables in dynamic adsorption process were investigated and the results revealed the tendency of adsorption capacity increased with the decrease of inflow rate and the increase of bed depth and initial concentration, and reached a higher adsorption capacity under appropriate environmental conditions. Thomas, Bohart-Adams, Yoon-Nelson and bed depth service time models were applied to predict experimental data, exhibiting satisfactory fitting results. According to artificial neural network simulation, the relative importance of operation parameters was in the sequence of inflow duration > bed depth > flow rate > influent pH > initial concentration > influent temperature. Furthermore, the adsorption performance of PANF-TU kept well during five times regeneration, which indicated the adsorbent might be a promising material for remediation of Hg 2+ polluted aqueous systems.