Incorporating Simultaneous Effect of Initial Concentration and Sorbent Dose into Removal Prediction Model Using Glyphosate Experimental Data and Theoretical Analysis

Glyphosate, the most used herbicide in the world, poses a serious threat to terrestrial and aquatic ecosystems. Magnetic nanosorbents composed of magnetite (Fe 3 O 4 ) cores coated with humic acid were synthesized, characterized, and used to investigate the removal of glyphosate (GLY) through adsorption. The spherical-type particles (average particle size = 9 nm), highly crystalline spinel structure of the iron core and surface morphology were unchanged before and after GLY adsorption. Adsorption fits with the pseudo second-order kinetic and Langmuir isotherm models. The humic acid coated magnetic nanomaterials (HA-MNP) exhibit a high maximum adsorption capacity for GLY, 30.4 mg g −1 . The result of thermodynamics study indicated the adsorption is endothermic and spontaneous. Removal efficiency increased with decreasing solution pH. Chloride ions exhibit no inhibition, while nitrate and phosphate compete with GLY adsorption to HA-MNP. Although multi-parameter models have been successfully applied to model adsorption processes, such studies often require large sets of experiments and complex mathematical representations to capture the interactive play of different variables. We observed excellent agreement between predicted and experimental results using a new approach employing dimensional analyses with the application of Buckingham’s π theorem for the combined concentration and dosing effects in the removal of GLY. In addition, error functions (r 2 =0.88) confirm the applicability of this model in the batch scale design independently of the sorbent dose and initial concentration. The accurate demonstration of this modeling approach for the adsorption of GLY on HA-MNP illustrates promise for predicting and optimizing adsorption treatment processes for different sorbents and target pollutants