Prediction of methylene blue dye sorption to sulfonated date palm kernel biochar using statistical regression and machine learning methods and DFT studies

Abstract

In this study, methylene blue (MB) dye adsorption to synthesised sulfonated date palm kernel biochar (SDPKB) was predicted and optimized using statistical-based regression approach (response surface methodology (RSM), machine-learning algorithms (artificial-neural-network (ANN), and adaptive-neuro-fuzzy-inference-system (ANFIS) models), and quantum chemistry calculations performed using density function theory (DFT) to link the electrical properties of the MB dye with the experimental findings. The percentage (%) of MB dye removed using SDPKB was found to be proportional to the biosorbent dosage and interaction time and inversely proportional to the solution pH and initial concentrations based on different used models results. It was observed that these models were accurate and comparable for the prediction of the removal of MB with coefficient of regression (R 2 ) values of 0.9174, 0.9742, and 0.9999, mean square error (MSE) values of 181.71, 94.50, and 0.00000049, and root MSE (RMSE) values of 13.48, 9.72, and 0.0007 for the RSM, ANN, and ANFIS, respectively. The ANFIS model was found to be more effective in the prediction of MB sorption to SDPKB than the other models (ANFIS > ANN > RSM), and it was highly applicable in the sorption process. This study has revealed that the SDPKB can serve as a better biosorbent for MB adsorption. Therefore, this study will serve as a first reference point data that will be of great assistance in industrial effluent management as well as decision-making on the adoption of statistical and machine learning models for adsorption study.