Modification of natural clays may be a useful approach to produce an effective and low-cost adsorbent to control phosphorous, which is a key factor in controlling the eutrophication of surface waters. In this study, natural clay samples were collected, characterized, modified with ZrCl4, and then their adsorption capacity for removing phosphorus from contaminated water was studied. XRD analysis showed that the natural clay consists of kaolinite, illite, and nontronite as dominant clay mineral phases. The maximum adsorption capacity of the modified clay increased from 0.493 to 11.83 mg P/g compared to the unmodified clay. The adsorption process was fast for both natural and modified samples, achieving more than 80% and 90% phosphorus removal with natural and modified samples, respectively in less than 4 hours. The adsorption data for both clays best fit the Langmuir isotherm, and the rate of phosphorus adsorption was found to follow a pseudo-secondorder kinetic model. The adsorption capacity of both adsorbents decreased with increasing pH, and for the modified clay the change was more significant. Full factorial design and response surface methodology were applied to evaluate and optimize the effects of initial P concentration, contact time, pH, and dose. From the model, the maximum P removal efficiency predicted for the synthetic solution was 91.5% and 99.9% by natural and modified clay, respectively. R2(≈0.98) indicates that the observed results fitted well with the model prediction. Similar to the batch studies, the fixed bed column study showed the developed adsorbents are efficient in removing phosphorus from water in a continuous process as well.

Glyphosate removal from contaminated water by natural clay, natural zeolite, commercial kaolin, bentonite, and sepiolite was compared. Natural clay showed high removal efficiency next to purified kaolin. The adsorption data for natural clay better fitted the Langmuir isotherm model and pseudo-second-order kinetic model. The adsorbent dose was found to be the most important process parameter while the pH did not exhibit any significant effect. An optimization study allowed achieving the highest removal efficiency and sorption capacity. The correlation coefficient R 2 of the regression model indicated that the observed results fitted well with the model prediction.