Biosorption of cobalt and chromium from wastewater using manganese dioxide and iron oxide nanoparticles loaded on cellulose-based biochar: Modeling and optimization with machine learning (artificial neural network).

In this study, two nanomaterials with excellent adsorption capacities were developed to remove heavy metals efficiently from wastewater. Manganese dioxide MnO2 nanoparticles and iron oxide Fe2O3 nanoparticles were successfully synthesized using cassava peel carbon and characterized by different techniques. The experimental tests for the adsorption process were done in a batch system, and the influence of various parameters such as temperature (from 5 to 60 °C), initial concentration (from 10 to 60 mg/L), pH (2 to 8), and contact time (5 to 180 min) on the biosorption of cobalt (II) and chromium (VI) were fully investigated. Furthermore, the Qmax were 546.32 mg/g and 349.59 mg/g for chromium (VI) and cobalt (II) respectively. The results fitted Langmuir with the pseudo-second-order model, revealing that chemisorption controls heavy metals removal, while the thermodynamic sorption was an endothermic and spontaneous reaction. Artificial Neural Network (ANN) model was developed to predict as well as to simulate the experimental results, for this purpose, the proposed model was based on five independent inputs or variables and one output or response which is the predicted adsorbed amount, the proposed ANN model showed an appreciable prediction accuracy with high optimization ability for chromium (VI) and cobalt (II) removal.