Comparative assessment of artificial neural network and response surface methodology for modeling and optimizing phosphate adsorption onto rice straw biochar.

This study investigated the potential application of rice straw biochar (RSB) as an adsorbent for the elimination of phosphate species from wastewater. Artificial neural network (ANN) and response surface methodology (RSM) approaches were used to optimize and model the adsorption process. The ANN and RSM analyses revealed important information regarding model precision, optimization effectiveness, and real-world usability, facilitating improved decision-making and process enhancement in intricate systems, such as wastewater treatment. ANN outperformed RSM with R2 = 0.9418 vs. 0.9206, RMSE = 2.34 vs. 3.12. Under optimized conditions (pH 7.2, 19.7 mg L-1, 20 mg dose, 31.6°C), >96.2% removal (with qmax = 168.0 mg g-1) was achieved. Chemisorption-dominated monolayer adsorption was confirmed by the pseudo-second-order kinetics (R2 = 0.9948) and Langmuir isotherm (R2 = 0.97). This paper shows that the valorization of agricultural wastes into biochar provides a sustainable circular economy solution to wastewater treatment. This study highlights the excellent potential of RSB for phosphate recovery from wastewater. Transforming agricultural waste into biochar and applying it to remove phosphate offers a sustainable solution for agricultural waste management and resource recovery from wastewater.