Removal of Cr (VI) from wastewater using bentonite as adsorbent: Experimental and Machine Learning investigations

Abstract

Abstract Bentonite material was prepared and examined for the adsorptive removal of Cr (VI) from an aqueous solution. The bentonite was characterized by XRD, FTIR, SEM-EDS, BET and DLS analyses. The experimental design optimization results for the adsorption (AN) of Cr (VI) onto the bentonite adsorbent (AB) determined using the Central Composite Design were pH: 4, adsorbent dose (AD): 1.0 g/100mL, initial Cr (VI) concentration (ICC): 200 mg/L, and contact time (CT): 60 min for a maximum Cr (VI) removal of 86.92%. The Cr (VI) AN onto bentonite fitted to Langmuir isotherm and an AN capacity of 25.17 mg/g was obtained. The kinetics of the process showed Elovich model and the AB also exhibited good reusability over various cycles. Various machine learning methods such as Linear Regression (LR), random forest regressor (RFR), Support Vector Regression (SVR), Gradient Boosting (GB), Artificial Neural Networks (ANNs) were investigated. Among these, the GB and SVR come out as the best model achieved R 2 (0.987, and 0.995) the lowest MAE (0.194, and 0.210) and RMSE (0.359 and 0.371), which predicts the AN value very close to the experimental results. This integration of machine learning and environmental chemistry provides a powerful tool for developing cost-effective and sustainable water treatment systems using naturally available materials.