Heptachlor Epoxide -Pesticide- Removal from Aqueous Solutions Using Nano Zero Valent Iron: Operating Conditions, Response Surface Methodology, and Artificial Intelligence Neural Networks

Abstract

There are many studies that detail how hazardous pesticides are to aquatic life, plants, animals, and humans, but there are fewer that describe how pesticides are treated during a separate treatment procedure. This study, which examines the removal of the highly toxic pesticide "heptachlor epoxide", is crucial to achieving SDG 6. Under operational circumstances, the prepared green nanoiron was successfully synthesized and characterized for the removal of heptachlor epoxide from aqueous solutions. For starting heptachlor epoxide concentrations of 100 and 10 μg/L at neutral medium pH 7, 0.8 g/L of green nZVI for 80 min, and a 200 RPM stirring rate, the removal efficiency varied between 55 and 100%, respectively. The RSM results indicated that the model R2 was 94.6%, and all operating conditions were significant to describe the removal efficiency with a p-value <0.05. The linear regression histogram indicated that the variation between expected and experimental removal efficiency ranged between (-1, 1%). The ANNs results by using MLP with network 6-3-1 indicated that nZVI was able to reduce heptachlor epoxide concentrations with a Sum of Squares Error of 0.052 for training and 0.177 for testing. Also, the ANNs described the importance of operating conditions and indicated that the most effective operating conditions were dose and less important was stirring rate, showing agreement with the obtained RSM results. Finally, this paper recommended using nZVI for heptachlor epoxide removal. Keywords Environmental toxicology; climate action; SDG 6; nanotechnology; pesticide removal; Heptachlor epoxide.