Phosphorus recovery using biochar for beryllium removal: Experimental evaluation and performance prediction via machine learning

Zipeng Zhang, Pengyan Zhang, Chong Liu, Fayong Li, Qi Yang, Xinqiang Liang,
Phosphorus recovery using biochar for beryllium removal: Experimental evaluation and performance prediction via machine learning,
Journal of Environmental Chemical Engineering,
Volume 13, Issue 6,
2025,
119919,
ISSN 2213-3437,
https://doi.org/10.1016/j.jece.2025.119919.
(https://www.sciencedirect.com/science/article/pii/S2213343725046160)
Abstract: Beryllium (Be) is a highly hazardous heavy metal with serious environmental and health risks due to its persistence in water and tendency to bioaccumulate. In this study, eggshell biochar (EBC) was synthesized by co-pyrolyzing cotton stalks as a carbon precursor with eggshells as a calcium source. The material was first deployed to capture phosphate from wastewater, forming a surface layer of brushite (dicalcium phosphate dihydrate). The recovered phosphate-laden eggshell biochar (EBC-P) was then repurposed for Be removal from water. EBC-P followed the Langmuir isotherm and pseudo-second-order kinetic models, with a maximum Be adsorption capacity of 49.29 mg g−1, primarily through chemical adsorption. Thermodynamic analysis indicated that the adsorption process was both feasible and spontaneous. High removal efficiency was sustained across pH 2–7, attributed to buffering by surface hydrogen-phosphate species. Characterization revealed that Be uptake occurred through precipitation and inner-sphere complexation with hydroxyl, carbonate, and phosphate groups, while protonated hydrogen-phosphate groups provided additional buffering capacity and reactive sites. Feature-importance analysis identified the phosphate content on EBC-P as the primary determinant of Be adsorption. Overall, this study demonstrates the conversion of agricultural waste and eggshells into effective biochar materials, enabling dual use: phosphate recovery from wastewater followed by Be removal. It also pioneers the application of machine learning to identify key factors influencing the Be adsorption process.
Keywords: Eggshell; Biochar; Adsorption; Machine Learning; Beryllium Wastewater; Phosphate